Chapter one: A NA LY TI CA L CHEMISTRY, I T S FU N C TI O N S A N D A P P LI C A T I O N S

Key Notes
Definition Analytical chemistry is a scientific discipline used to study the chemical
composition, structure and behavior of matter.

Purpose The purpose of chemical analysis is to gather and interpret chemical

information that will be of value to society in a wide range of contexts.

Scope and Ǫuality control in manufacturing industries, the monitoring of clinical

applications and environmental samples, the assaying of geological specimens, and
the support of fundamental and applied research are the principal
applications.

Related topics Analytical problems and Computer control and data

procedures (A2) collection (H3)
Chemical sensors and biosensors Data enhancement and databases
(H1) (H4)
Automated procedures (H2)

Definition Analytical chemistry involves the application of a range of techniques and

methodologies to obtain and assess qualitative, quantitative and structural
information on the nature of matter.
● Qualitative analysis is the identification of elements, species and/or
compounds present in a sample.
● Quantitative analysis is the determination of the absolute or relative amounts
of elements, species or compounds present in a sample.
● Structural analysis is the determination of the spatial arrangement of atoms in
an element or molecule or the identification of characteristic groups of atoms
(functional groups).
● An element, species or compound that is the subject of analysis is known as an
analyte.
● The remainder of the material or sample of which the analyte(s) form(s) a part
is known as the matrix.

Purpose The gathering and interpretation of qualitative, quantitative and structural infor-
mation is essential to many aspects of human endeavor, both terrestrial and extra-
terrestrial. The maintenance of, and improvement in, the quality of life throughout
the world, and the management of resources rely heavily on the information
provided by chemical analysis. Manufacturing industries use analytical data to
monitor the quality of raw materials, intermediates and

pg. 1
finished products. Progress and research in many areas is dependent on estab-
lishing the chemical composition of man-made or natural materials, and the
monitoring of toxic substances in the environment is of ever increasing impor-
tance. Studies of biological and other complex systems are supported by the
collection of large amounts of analytical data.

Scope and Analytical data are required in a wide range of disciplines and situations that
applications include not just chemistry and most other sciences, from biology to zoology, but
the arts, such as painting and sculpture, and archaeology. Space exploration and
clinical diagnosis are two quite disparate areas in which analytical data is vital.
Important areas of application include the following.
● Quality control (QC). In many manufacturing industries, the chemical
composition of raw materials, intermediates and finished products needs to
be monitored to ensure satisfactory quality and consistency. Virtually all
consumer products from automobiles to clothing, pharmaceuticals and food-
stuffs, electrical goods, sports equipment and horticultural products rely, in
part, on chemical analysis. The food, pharmaceutical and water industries in
particular have stringent requirements backed by legislation for major compo-
nents and permitted levels of impurities or contaminants. The electronics
industry needs analyses at ultra-trace levels (parts per billion) in relation to the
manufacture of semi-conductor materials. Automated, computer-controlled
procedures for process-stream analysis are employed in some industries.
● Monitoring and control of pollutants. The presence of toxic heavy metals
(e.g., lead, cadmium and mercury), organic chemicals (e.g., polychlorinated
biphenyls and detergents) and vehicle exhaust gases (oxides of carbon,
nitrogen and sulfur, and hydrocarbons) in the environment are health hazards
that need to be monitored by sensitive and accurate methods of analysis, and
remedial action taken. Major sources of pollution are gaseous, solid and liquid
wastes that are discharged or dumped from industrial sites, and vehicle
exhaust gases.
● Clinical and biological studies. The levels of important nutrients, including
trace metals (e.g., sodium, potassium, calcium and zinc), naturally produced
chemicals, such as cholesterol, sugars and urea, and administered drugs in the
body fluids of patients undergoing hospital treatment require monitoring.
Speed of analysis is often a crucial factor and automated procedures have been
designed for such analyses.
● Geological assays. The commercial value of ores and minerals is determined
by the levels of particular metals, which must be accurately established.
Highly accurate and reliable analytical procedures must be used for this
purpose, and referee laboratories are sometimes employed where disputes
arise.
● Fundamental and applied research. The chemical composition and structure of
materials used in or developed during research programs in numerous
disciplines can be of significance. Where new drugs or materials with potential
commercial value are synthesized, a complete chemical characterization may
be required involving considerable analytical work. Combinatorial chemistry
is an approach used in pharmaceutical research that generates very large
numbers of new compounds requiring confirmation of identity and structur
NALYTICAL PROBLEMS AND
PROCEDURES

Analytical problems

Analytical
procedures

collection (H3)
Data enhancement and databases

(H1)

Analytical The most important aspect of an analysis is to ensure that it will provide useful
problems and reliable data on the qualitative and/or quantitative composition of a material or
structural information about the individual compounds present. The analyt- ical
chemist must often communicate with other scientists and nonscientists to
establish the amount and quality of the information required, the time-scale for
the work to be completed and any budgetary constraints. The most appropriate
analytical technique and method can then be selected from those available or new
ones devised and validated by the analysis of substances of known composition
and/or structure. It is essential for the analytical chemist to have an appreciation
of the objectives of the analysis and an understanding of the capabilities of the
various analytical techniques at his/her disposal without which the most appro-
priate and cost-effective method cannot be selected or developed.

Analytical The stages or steps in an overall analytical procedure can be summarized as

procedures follows.
● Definition of the problem. Analytical information and level of accuracy
required. Costs, timing, availability of laboratory instruments and facilities.
● Choice of technique and method. Selection of the best technique for the
required analysis, such as chromatography, infrared spectrometry, titrimetry,
thermogravimetry. Selection of the method (i.e. the detailed stepwise instruc-
tions using the selected technique).
● Sampling. Selection of a small sample of the material to be analyzed. Where
this is heterogeneous, special procedures need to be used to ensure that a
genuinely representative sample is obtained (Topic A4).

pg. 3
● Sample pre-treatment or conditioning. Conversion of the sample into a form
suitable for detecting or measuring the level of the analyte(s) by the selected
technique and method. This may involve dissolving it, converting the analyte(s)
into a specific chemical form or separating the analyte(s) from other components
of the sample (the sample matrix) that could interfere with detec- tion or
quantitative measurements.
● Qualitative analysis. Tests on the sample under specified and controlled
conditions. Tests on reference materials for comparison. Interpretation of the
tests.
● Quantitative analysis. Preparation of standards containing known amounts
of the analyte(s) or of pure reagents to be reacted with the analyte(s). Calibration
of instruments to determine the responses to the standards under controlled
conditions. Measurement of the instrumental response for each sample under the
same conditions as for the standards. All measurements may be replicated to
improve the reliability of the data, but this has cost and time implications.
Calculation of results and statistical evaluation.
● Preparation of report or certificate of analysis. This should include a
summary of the analytical procedure, the results and their statistical assess-
ment, and details of any problems encountered at any stage during the
analysis.
● Review of the original problem. The results need to be discussed with regard
to their significance and their relevance in solving the original problem.
Sometimes repeat analyses or new analyses may be undertaken.
ANALYTICAL TECHNIǪUES AND METHODS

Analytical There are numerous chemical or physico-chemical processes that can be used to

Analytical
techniques

Analytical methods A method is a detailed set of instructions for a particular analysis using a
specified technique.

Method validation

Ǫuality in analytical laboratories (A6)

techniques provide analytical information. The processes are related to a wide range of
atomic and molecular properties and phenomena that enable elements and
compounds to be detected and/or quantitatively measured under controlled
conditions. The underlying processes define the various analytical techniques.
The more important of these are listed in Table 1, together with their suitability for
qualitative, quantitative or structural analysis and the levels of analyte(s) in a
sample that can be measured.
Atomic and molecular spectrometry and chromatography, which together
comprise the largest and most widely used groups of techniques, can be further
subdivided according to their physico-chemical basis. Spectrometric techniques
may involve either the emission or absorption of electromagnetic radiation over
a very wide range of energies, and can provide qualitative, quantitative and
structural information for analytes from major components of a sample down
to ultra-trace levels. The most important atomic and molecular spectrometric
techniques and their principal applications are listed in Table 2.
Chromatographic techniques provide the means of separating the compo-
nents of mixtures and simultaneous qualitative and quantitative analysis, as
required. The linking of chromatographic and spectrometric techniques, called
hyphenation, provides a powerful means of separating and identifying
unknown compounds (Section F). Electrophoresis is another separation tech-
nique with similarities to chromatography that is particularly useful for the
separation of charged species. The principal separation techniques and their
applications are listed in Table 3.

Analytical An analytical method consists of a detailed, stepwise list of instructions to be

methods followed in the qualitative, quantitative or structural analysis of a sample for one
or more analytes and using a specified technique. It will include a summary and

pg. 5
Table 1. Analytical techniques and principal applications

Technique Property measured Principal areas of application

Gravimetry Weight of pure analyte or compound Quantitative for major or minor
of known stoichiometry components
Titrimetry Volume of standard reagent solution Quantitative for major or minor
reacting with the analyte components
Atomic and molecular Wavelength and intensity of Qualitative, quantitative or structural
spectrometry electromagnetic radiation emitted or for major down to trace level
absorbed by the analyte components
Mass spectrometry Mass of analyte or fragments of it Qualitative or structural for major
down to trace level components
isotope ratios
Chromatography and Various physico-chemical properties Qualitative and quantitative
electrophoresis of separated analytes separations of mixtures at major to
trace levels
Thermal analysis Chemical/physical changes in the Characterization of single or mixed
analyte when heated or cooled major/minor components
Electrochemical analysis Electrical properties of the analyte Qualitative and quantitative for major
in solution to trace level components
Radiochemical analysis Characteristic ionizing nuclear Qualitative and quantitative at major
radiation emitted by the analyte to trace levels

Table 2. Spectrometric techniques and principal applications

Technique Basis Principal applications

Plasma emission spectrometry Atomic emission after excitation in high Determination of metals and some
temperature gas plasma non-metals mainly at trace levels
Flame emission spectrometry Atomic emission after flame excitation Determination of alkali and alkaline
earth metals
Atomic absorption spectrometry Atomic absorption after atomization Determination of trace metals and
by flame or electrothermal means some non-metals
Atomic fluorescence Atomic fluorescence emission after Determination of mercury and
spectrometry flame excitation hydrides of non-metals at trace
levels
X-ray emission spectrometry Atomic or atomic fluorescence Determination of major and minor
emission after excitation by electrons elemental components of
or radiation metallurgical and geological samples
-spectrometry -ray emission after nuclear excitation Monitoring of radioactive elements in
environmental samples
Ultraviolet/visible spectrometry Electronic molecular absorption in Quantitative determination of
solution unsaturated organic compounds
Infrared spectrometry Vibrational molecular absorption Identification of organic compounds
Nuclear magnetic resonance Nuclear absorption (change of spin Identification and structural analysis
spectrometry states) of organic compounds
Mass spectrometry Ionization and fragmentation of Identification and structural analysis
molecules of organic compounds
Table 3. Separation techniques and principal applications
Technique Basis Principal applications

}
Thin-layer chromatography Qualitative analysis of mixtures
Differential rates of migration of
Gas chromatography Quantitative and qualitative
analytes through a stationary phase
determination of volatile compounds
by movement of a liquid or gaseous

High-performance liquid mobile phase Quantitative and qualitative

chromatography determination of nonvolatile
compounds

Electrophoresis Differential rates of migration of Quantitative and qualitative

analytes through a buffered medium determination of ionic compounds

lists of chemicals and reagents to be used, laboratory apparatus and glassware, and
appropriate instrumentation. The quality and sources of chemicals, including solvents, and
the required performance characteristics of instruments will also be specified as will the
procedure for obtaining a representative sample of the material to be analyzed. This is of
crucial importance in obtaining mean- ingful results (Topic A4). The preparation or pre-
treatment of the sample will be followed by any necessary standardization of reagents
and/or calibration of instruments under specified conditions (Topic A5). Ǫualitative tests
for the analyte(s) or quantitative measurements under the same conditions as those used for
standards complete the practical part of the method. The remaining steps will be concerned
with data processing, computational methods for quantitative analysis and the formatting
of the analytical report. The statistical assessment of quantitative data is vital in
establishing the reliability and value of the data, and the use of various statistical
parameters and tests is widespread (Section B).
Many standard analytical methods have been published as papers in analyt-
ical journals and other scientific literature, and in textbook form. Collections by trades
associations representing, for example, the cosmetics, food, iron and steel, pharmaceutical,
polymer plastics and paint, and water industries are available. Standards organizations
and statutory authorities, instrument manufacturers’ applications notes, the Royal Society
of Chemistry and the US Environmental Protection Agency are also valuable sources of
standard methods. Often, labora-
tories will develop their own in-house methods or adapt existing ones for specific purposes.
Method development forms a significant part of the work of most analytical laboratories, and
method validation and periodic revalidation is
a necessity.
Selection of the most appropriate analytical method should take into account the
following factors:

● the purpose of the analysis, the required time scale and any cost constraints;
● the level of analyte(s) expected and the detection limit required;
● the nature of the sample, the amount available and the necessary sample preparation
procedure;
● the accuracy required for a quantitative analysis;
● the availability of reference materials, standards, chemicals and solvents,
instrumentation and any special facilities;
● possible interference with the detection or quantitative measurement of the analyte(s)
and the possible need for sample clean-up to avoid matrix interference;

pg. 7
● the degree of selectivity available − methods may be selective for a small
number of analytes or specific for only one;
● quality control and safety factors.

Method validation Analytical methods must be shown to give reliable data, free from bias and suit-
able for the intended use. Most methods are multi-step procedures, and the
process of validation generally involves a stepwise approach in which optimized
experimental parameters are tested for robustness (ruggedness), that is sensi-
tivity to variations in the conditions, and sources of errors investigated.
A common approach is to start with the final measurement stage, using cali-
bration standards of known high purity for each analyte to establish the perfor-
mance characteristics of the detection system (i.e. specificity, range, quantitative
response (linearity), sensitivity, stability and reproducibility). Robustness in
terms of temperature, humidity and pressure variations would be included at
this stage, and a statistical assessment made of the reproducibility of repeated
identical measurements (replicates). The process is then extended backwards in
sequence through the preceding stages of the method, checking that the optimum
conditions and performance established for the final measurement on analyte
calibration standards remain valid throughout. Where this is not the case, new
conditions must be investigated by modification of the procedure and the process
repeated. A summary of this approach is shown in Figure 1 in the form of a flow
diagram. At each stage, the results are assessed using appropriate statistical tests
(Section B) and compared for consistency with those of the previous stage. Where
unacceptable variations arise, changes to the procedure are implemented and the
assessment process repeated. The performance and robustness of the overall
method are finally tested with field trials in one or more routine analytical
laboratories before the method is considered to be fully validated.
Step 1 Performance characteristics of detector
for single analyte calibration standards

Step 2 Process repeated for mixed analyte

calibration standards

Step 3 Process repeated for analyte calibration

standards with possible interfering
substances and for reagent blanks

Step 4 Process repeated for analyte calibration

standards with anticipated matrix
components to evaluate matrix
interference

Step 5 Analysis of 'spiked' simulated sample

matrix. i.e. matrix with added known
amounts of analyte(s), to test recoveries

Step 6 Field trials in routine laboratory with

more junior personnel to test ruggedness

Fig. 1. Flow chart for method validation.

pg. 9
S AMPLING AND S AMPLE HAN DL IN G

Key Notes

Representative A representative sample is one that truly reflects the composition of the
sample material to be analyzed within the context of a defined analytical
problem.

Sample storage Due to varying periods of time that may elapse between sample
collection and analysis, storage conditions must be such as to avoid
undesirable losses, contamination or other changes that could affect the
results of the analysis.

Sample Preliminary treatment of a sample is sometimes necessary before it is in a

pre-treatment suitable form for analysis by the chosen technique and method. This may
involve a separation or concentration of the analytes or the removal of
matrix components that would otherwise interfere with the analysis.

Sample preparation Samples generally need to be brought into a form suitable for
measurements to be made under controlled conditions. This may involve
dissolution, grinding, fabricating into a specific size and shape,
pelletizing or mounting in a sample holder.

Related topic Analytical problems and procedures (A2)

Representative The importance of obtaining a representative sample for analysis cannot be

sample overemphasized. Without it, results may be meaningless or even grossly
misleading. Sampling is particularly crucial where a heterogeneous material is to be
analyzed. It is vital that the aims of the analysis are understood and an appro-
priate sampling procedure adopted. In some situations, a sampling plan or
strategy may need to be devised so as to optimize the value of the analytical
information collected. This is necessary particularly where environmental
samples of soil, water or the atmosphere are to be collected or a complex indus-
trial process is to be monitored. Legal requirements may also determine a
sampling strategy, particularly in the food and drug industries. A small sample
taken for analysis is described as a laboratory sample. Where duplicate analyses
or several different analyses are required, the laboratory sample will be divided
into sub-samples which should have identical compositions.
Homogeneous materials (e.g., single or mixed solvents or solutions and most
gases) generally present no particular sampling problem as the composition of
any small laboratory sample taken from a larger volume will be representative of the
bulk solution. Heterogeneous materials have to be homogenized prior to obtaining
a laboratory sample if an average or bulk composition is required. Conversely,
where analyte levels in different parts of the material are to be
measured, they may need to be physically separated before laboratory samples
are taken. This is known as selective sampling. Typical examples of hetero-
geneous materials where selective sampling may be necessary include:
● surface waters such as streams, rivers, reservoirs and seawater, where the
concentrations of trace metals or organic compounds in solution and in sedi-
ments or suspended particulate matter may each be of importance;
● materials stored in bulk, such as grain, edible oils, or industrial organic chem-
icals, where physical segregation (stratification) or other effects may lead to
variations in chemical composition throughout the bulk;
● ores, minerals and alloys, where information about the distribution of a partic-
ular metal or compound is sought;
● laboratory, industrial or urban atmospheres where the concentrations of toxic
vapors and fumes may be localized or vary with time.
Obtaining a laboratory sample to establish an average analyte level in a highly
heterogeneous material can be a lengthy procedure. For example, sampling a
large shipment of an ore or mineral, where the economic cost needs to be
determined by a very accurate assay, is typically approached in the following
manner.
(i) Relatively large pieces are randomly selected from different parts of the
shipment.
(ii) The pieces are crushed, ground to coarse granules and thoroughly mixed.
(iii) A repeated coning and quartering process, with additional grinding to
reduce particle size, is used until a laboratory-sized sample is obtained. This
involves creating a conical heap of the material, dividing it into four equal
portions, discarding two diagonally opposite portions and forming a new
conical heap from the remaining two quarters. The process is then repeated
as necessary (Fig. 1).

Fig. 1. A diagrammatic representation of coning and quartering (quarters 1 and 3, or 2 and 4 are discarded each time).

pg. 11
The distribution of toxic heavy metals or organic compounds in a land rede-
velopment site presents a different problem. Here, to economize on the number
of analyses, a grid is superimposed on the site dividing it up into approximately
one- to five-metre squares. From each of these, samples of soil will be taken at
several specified depths. A three-dimensional representation of the distribution
of each analyte over the whole site can then be produced, and any localized high
concentrations, or hot spots, can be investigated by taking further, more closely-
spaced, samples. Individual samples may need to be ground, coned and
quartered as part of the sampling strategy.
Repeated sampling over a period of time is a common requirement. Examples
include the continuous monitoring of a process stream in a manufacturing plant
and the frequent sampling of patients’ body fluids for changes in the levels of
drugs, metabolites, sugars or enzymes, etc., during hospital treatment. Studies of
seasonal variations in the levels of pesticide, herbicide and fertilizer residues in
soils and surface waters, or the continuous monitoring of drinking water supplies
are two further examples.
Having obtained a representative sample, it must be labeled and stored under
appropriate conditions. Sample identification through proper labeling, increas-
ingly done by using bar codes and optical readers under computer control, is an
essential feature of sample handling.

Sample storage Samples often have to be collected from places remote from the analytical labora- tory
and several days or weeks may elapse before they are received by the labo-ratory
and analyzed. Furthermore, the workload of many laboratories is such that
incoming samples are stored for a period of time prior to analysis. In both
instances, sample containers and storage conditions (e.g., temperature, humidity,
light levels and exposure to the atmosphere) must be controlled such that no
significant changes occur that could affect the validity of the analytical data. The
following effects during storage should be considered:

● increases in temperature leading to the loss of volatile analytes, thermal or

biological degradation, or increased chemical reactivity;
● decreases in temperature that lead to the formation of deposits or the precipi-
tation of analytes with low solubilities;
● changes in humidity that affect the moisture content of hygroscopic solids and
liquids or induce hydrolysis reactions;
● UV radiation, particularly from direct sunlight, that induces photochemical
reactions, photodecomposition or polymerization;
● air-induced oxidation;
● physical separation of the sample into layers of different density or changes in
crystallinity.

In addition, containers may leak or allow contaminants to enter.

A particular problem associated with samples having very low (trace and
ultra-trace) levels of analytes in solution is the possibility of losses by adsorp-
tion onto the walls of the container or contamination by substances being
leached from the container by the sample solvent. Trace metals may be depleted
by adsorption or ion-exchange processes if stored in glass containers, whilst
sodium, potassium, boron and silicates can be leached from the glass into the
sample solution. Plastic containers should always be used for such samples.
Conversely, sample solutions containing organic solvents and other organic
liquids should be stored in glass containers because the base plastic or additives
such as plasticizers and antioxidants may be leached from the walls of plastic
containers.

Sample pre- Samples arriving in an analytical laboratory come in a very wide assortment of sizes,
treatment conditions and physical forms and can contain analytes from major constituents
down to ultra-trace levels. They can have a variable moisture content and the matrix
components of samples submitted for determinations of the same analyte(s) may
also vary widely. A preliminary, or pre-treatment, is often used to condition them
in readiness for the application of a specific method of analysis or to pre-concentrate
(enrich) analytes present at very low levels. Examples of pre- treatments are:

● drying at 100°C to 120°C to eliminate the effect of a variable moisture content;

● weighing before and after drying enables the water content to be calculated or
it can be established by thermogravimetric analysis (Topic G1);
● separating the analytes into groups with common characteristics by dis-
tillation, filtration, centrifugation, solvent or solid phase extraction (Topic
D1);
● removing or reducing the level of matrix components that are known to cause
interference with measurements of the analytes;
● concentrating the analytes if they are below the concentration range of the
analytical method to be used by evaporation, distillation, co-precipitation, ion
exchange, solvent or solid phase extraction or electrolysis.

Sample clean-up in relation to matrix interference and to protect special-

ized analytical equipment such as chromatographic columns and detection
systems from high levels of matrix components is widely practised using solid
phase extraction (SPE) cartridges (Topic D1). Substances such as lipids, fats,
proteins, pigments, polymeric and tarry substances are particularly detri-
mental.

A laboratory sample generally needs to be prepared for analytical measurement

by treatment with reagents that convert the analyte(s) into an appropriate chem-
Sample
ical form for the selected technique and method, although in some instances it is
preparation
examined directly as received or mounted in a sample holder for surface
analysis. If the material is readily soluble in aqueous or organic solvents, a simple
dissolution step may suffice. However, many samples need first to be decom-
posed to release the analyte(s) and facilitate specific reactions in solution. Sample
solutions may need to be diluted or concentrated by enrichment so that analytes
are in an optimum concentration range for the method. The stabilization of solu-
tions with respect to pH, ionic strength and solvent composition, and the removal
or masking of interfering matrix components not accounted for in any pre-treat-
ment may also be necessary. An internal standard for reference purposes in
quantitative analysis (Topic A5 and Section B) is sometimes added before adjust-
ment to the final prescribed volume. Some common methods of decomposition
and dissolution are given in Table 1.

pg. 13
Table 1. Some methods for sample decomposition and dissolution
Method of attack Type of sample
Heated with concentrated mineral Geological, metallurgical
acids (HCl, HNO3, aqua regia) or
strong alkali, including microwave
digestion
Fusion with flux (Na2O2, Na2CO3, Geological, refractory materials
LiBO2, KHSO4, KOH)

Heated with HF and H2SO4 or HClO4 Silicates where SiO2 is not the analyte
Acid leaching with HNO3 Soils and sediments
Dry oxidation by heating in a furnace Organic materials with inorganic analytes
or wet oxidation by boiling with
concentrated H2SO 4 and HNO3 or HClO4
CALIBRATION AND STANDARDS

Chemical standard

Reference material A reference material is a material or substance, one or more properties of which
are sufficiently homogeneous and well established for it to be used for the
calibration of apparatus, the assessment of a measurement method or for
assigning values to materials.

Calibration and linear regression (B4)

Calibration With the exception of absolute methods of analysis that involve chemical reac -
tions of known stoichiometry (e.g., gravimetric and titrimetric determinations), a
calibration or standardization procedure is required to establish the relation
between a measured physico-chemical response to an analyte and the amount or
concentration of the analyte producing the response. Techniques and methods
where calibration is necessary are frequently instrumental, and the detector
response is in the form of an electrical signal. An important consideration is the
effect of matrix components on the analyte detector signal, which may be
supressed or enhanced, this being known as the matrix effect. When this is
known to occur, matrix matching of the calibration standards to simulate the
gross composition expected in the samples is essential (i.e. matrix components
are added to all the analyte standards in the same amounts as are expected in the
samples).
There are several methods of calibration, the choice of the most suitable
depending on the characteristics of the analytical technique to be employed, the
nature of the sample and the level of analyte(s) expected. These include:
● External standardization. A series of at least four calibration standards
containing known amounts or concentrations of the analyte and matrix
components, if required, is either prepared from laboratory chemicals of guar-
anteed purity (AnalaR or an equivalent grade) or purchased as a concentrated
standard ready to use. The response of the detection system is recorded for
each standard under specified and stable conditions and additionally for a blank,
sometimes called a reagent blank (a standard prepared in an identical

pg. 15
fashion to the other standards but omitting the analyte). The data is either plotted
as a calibration graph or used to calculate a factor to convert detector responses
measured for the analyte in samples into corresponding masses or
concentrations (Topic B4).
● Standard addition.
● Internal standardization.
The last two methods of calibration are described in Topic B4.
Instruments and apparatus used for analytical work must be correctly main-
tained and calibrated against reference values to ensure that measurements are
accurate and reliable. Performance should be checked regularly and records kept
so that any deterioration can be quickly detected and remedied. Microcomputer
and microprocessor controlled instrumentation often has built-in performance
checks that are automatically initiated each time an instrument is turned on.
Some examples of instrument or apparatus calibration are
● manual calibration of an electronic balance with certified weights;
● calibration of volumetric glassware by weighing volumes of pure water;
● calibration of the wavelength and absorbance scales of spectrophotometers with
certified emission or absorption characteristics;
● calibration of temperature scales and electrical voltage or current readouts
with certified measurement equipment.

Chemical Materials or substances suitable for use as chemical standards are generally
standard single compounds or elements. They must be of known composition, and high
purity and stability. Many are available commercially under the name AnalaR.
Primary standards, which are used principally in titrimetry (Section C) to
standardize a reagent (titrant) (i.e. to establish its exact concentration) must be
internationally recognized and should fulfil the following requirements:
● be easy to obtain and preserve in a high state of purity and of known chemical
composition;
● be non-hygroscopic and stable in air allowing accurate weighing;
● have impurities not normally exceeding 0.02% by weight;
● be readily soluble in water or another suitable solvent;
● react rapidly with an analyte in solution;
● other than pure elements, to have a high relative molar mass to minimize
weighing errors.
Primary standards are used directly in titrimetric methods or to standardize
solutions of secondary or working standards (i.e. materials or substances that do
not fulfill all of the above criteria, that are to be used subsequently as the titrant in
a particular method). Chemical standards are also used as reagents to effect
reactions with analytes before completing the analysis by techniques other than
titrimetry.
Some approved primary standards for titrimetric analysis are given in Table 1.

Reference Reference materials are used to demonstrate the accuracy, reliability and com-
material parability of analytical results. A certified or standard reference material (CRM or
SRM) is a reference material, the values of one or more properties of which have
been certified by a technically valid procedure and accompanied by a trace- able
certificate or other documentation issued by a certifying body such as the
Table 1. Some primary standards used in titrimetric analysis
Type of titration Primary standard
Acid-base Sodium carbonate, Na2CO3
Sodium tetraborate, Na2B4O7.10H 2O
Potassium hydrogen phthalate, KH(C8H4O4)
Benzoic acid, C6H5COOH
Redox Potassium dichromate, K2Cr2O7
Potassium iodate, KIO3
Sodium oxalate, Na2C2O4
Precipitation (silver halide) Silver nitrate, AgNO3
Sodium chloride, NaCl
Complexometric (EDTA) Zinc, Zn
Magnesium, Mg
EDTA (disodium salt), C10H14N2O8Na2

Bureau of Analytical Standards. CRMs or SRMs are produced in various forms

and for different purposes and they may contain one or more certified compo-
nents, such as
● pure substances or solutions for calibration or identification;
● materials of known matrix composition to facilitate comparisons of analytical
data;
● materials with approximately known matrix composition and specified
components.
They have a number of principal uses, including
● validation of new methods of analysis;
● standardization/calibration of other reference materials;
● confirmation of the validity of standardized methods;
● support of quality control and quality assurance schemes.

pg. 17
ǪUALITY IN ANALYTICAL LABORATORIES

Quality control

Quality assurance Ǫuality assurance (ǪA) is the combination of planned and systematic
actions necessary to provide adequate confidence that the process of
quality control satisfies specified requirements.

Accreditation
system

Ǫuality control and chemometrics

(B5)

Quality control Analytical data must be of demonstrably high quality to ensure confidence in the
results. Quality control (QC) comprises a system of planned activities in an
analytical laboratory whereby analytical methods are monitored at every stage to
verify compliance with validated procedures and to take steps to eliminate the
causes of unsatisfactory performance. Results are considered to be of sufficiently
high quality if
● they meet the specific requirements of the requested analytical work within
the context of a defined problem;
● there is confidence in their validity;
● the work is cost effective.
To implement a ǪC system, a complete understanding of the chemistry and
operations of the analytical method and the likely sources and magnitudes of
errors at each stage is essential. The use of reference materials (Topic A5) during
method validation (Topic A3) ensures that results are traceable to certified
sources. ǪC processes should include:
● checks on the accuracy and precision of the data using statistical tests (Section
B);
● detailed records of calibration, raw data, results and instrument performance;
● observations on the nature and behavior of the sample and unsatisfactory
aspects of the methodology;
● control charts to determine system control for instrumentation and repeat
analyses (Topic B5);
● provision of full documentation and traceability of results to recognized
reference materials through recorded identification;
● maintenance and calibration of instrumentation to manufacturers’ specifica-
tions;
● management and control of laboratory chemicals and other materials including
checks on quality;
● adequate training of laboratory personnel to ensure understanding and
competence;
● external verification of results wherever possible;
● accreditation of the laboratory by an independent organization.

Quality assurance The overall management of an analytical laboratory should include the provision
of evidence and assurances that appropriate ǪC procedures for laboratory activ-
ities are being correctly implemented. Quality assurance (QA) is a managerial
responsibility that is designed to ensure that this is the case and to generate
confidence in the analytical results. Part of ǪA is to build confidence through the
laboratory participating in interlaboratory studies where several laboratories
analyze one or more identical homogeneous materials under specified condi- tions.
Proficiency testing is a particular type of study to assess the performance of a
laboratory or analyst relative to others, whilst method performance studies and
certification studies are undertaken to check a particular analytical method or
reference material respectively. The results of such studies and their statistical
assessment enable the performances of individual participating laboratories to be
demonstrated and any deficiencies in methodology and the training of personnel
to be addressed.

Accreditation Because of differences in the interpretation of the term quality, which can be
system defined as fitness for purpose, ǪC and ǪA systems adopted by analyical labora-
tories in different industries and fields of activity can vary widely. For this
reason, defined quality standards have been introduced by a number of organi-
zations throughout the world. Laboratories can design and implement their own
quality systems and apply to be inspected and accredited by the organization for
the standard most appropriate to their activity. A number of organizations that
offer accreditation suitable for analytical laboratories and their corresponding
quality standards are given in Table 1.

Table 1. Accreditation organizations and their quality standards

Name of accreditation organization Quality standard

Organization for Economic Co-operation Good Laboratory Practice (GLP)
and Development (OECD)
The International Organization for ISO 9000 series of quality standards
Standardization (ISO) ISO Guide 25 general requirements for
competence of calibration and testing
laboratories
European Committee for Standardization EN 29000 series
(CEN) EN 45000 series
British Standards Institution (BSI) BS 5750 quality standard
BS 7500 series
National Measurement Accreditation NAMAS
Service (NAMAS)

pg. 19
Chapter two: ERRORS IN ANALYTICAL MEASUREMENTS

Key Notes

Measurement errors All measurement processes are subject to measurement errors that affect
numerical data and which arise from a variety of sources.

Absolute and An absolute error is the numerical difference between a measured value
relative errors and a true or accepted value. A relative error is the absolute error divided
by the true or accepted value.

Determinate errors Also known as systematic errors, or bias, these generally arise from
determinate or identifiable sources causing measured values to differ
from a true or accepted value.

Indeterminate errors Also known as random errors, these arise from a variety of uncontrolled
sources and cause small random variations in a measured quantity when
the measurement is repeated a number of times.

Accumulated errors Where several different measurements are combined to compute an

overall analytical result, the errors associated with each individual
measurement contribute to a total or accumulated error.

Related topic Assessment of accuracy and precision (B2)

Measurement The causes of measurement errors are numerous and their magnitudes are vari-
errors able. This leads to uncertainties in reported results. However, measurement
errors can be minimized and some types eliminated altogether by careful exper-
imental design and control. Their effects can be assessed by the application of
statistical methods of data analysis and chemometrics (Topic B5). Gross errors
may arise from faulty equipment or bad laboratory practice; proper equipment
maintenance and appropriate training and supervision of personnel should
eliminate these.
Nevertheless, whether it is reading a burette or thermometer, weighing a
sample or timing events, or monitoring an electrical signal or liquid flow, there
will always be inherent variations in the measured parameter if readings are
repeated a number of times under the same conditions. In addition, errors may
go undetected if the true or accepted value is not known for comparison
purposes.
Errors must be controlled and assessed so that valid analytical measurements can
be made and reported. The reliability of such data must be demonstrated so that an
end-user can have an acceptable degree of confidence in the results of an
analysis.
Absolute and The absolute error, EA, in a measurement or result, xM, is given by the equation
relative errors
EA = xM -xT

where xT is the true or accepted value. Examples are shown in Figure 1 where a
200 mg aspirin standard has been analyzed a number of times. The absolute errors
range from 4 mg to 10 mg.
The relative error, ER, in a measurement or result, xM, is given by the equation
ER = (xM - xT)/xT
Often, ER is expressed as a percentage relative error, 100ER. Thus, for the aspirin
results shown in Figure 1, the relative error ranges from 2% to 5%. Relative
errors are particularly useful for comparing results of differing magnitude.

Aspirin (mg)
195 200 205 210

–5 0 5 10
Absolute error (EA; mg)
–2.5 0 2.5 5
Relative error (ER; %)
Fig. 1. Absolute and relative errors in the analysis of an aspirin standard.

Determinate There are three basic sources of determinate or systematic errors that lead to a
errors bias in measured values or results:
● the analyst or operator;
● the equipment (apparatus and instrumentation) and the laboratory environ-
ment;
● the method or procedure.
It should be possible to eliminate errors of this type by careful observation and
record keeping, equipment maintenance and training of laboratory personnel.
Operator errors can arise through carelessness, insufficient training, illness or
disability. Equipment errors include substandard volumetric glassware, faulty
or worn mechanical components, incorrect electrical signals and a poor or
insufficiently controlled laboratory environment. Method or procedural errors are
caused by inadequate method validation, the application of a method to samples or
concentration levels for which it is not suitable or unexpected varia- tions in sample
characteristics that affect measurements. Determinate errors that lead to a higher
value or result than a true or accepted one are said to show a positive bias; those
leading to a lower value or result are said to show a nega- tive bias. Particularly
large errors are described as gross errors; these should be easily apparent and
readily eliminated.
Determinate errors can be proportional to the size of sample taken for analysis. If so,
they will have the same effect on the magnitude of a result regardless of the size of the
sample, and their presence can thus be difficult to detect. For example, copper(II) can be
determined by titration after reaction with potassium iodide to release iodine according to
the equation

2Cu2 + 4I 2CuI +I2

However, the reaction is not specific to copper(II), and any iron(III) present in the sample
will react in the same way. Results for the determination of copper in an alloy containing
20%, but which also contained 0.2% of iron are shown inFigure 2 for a range of sample sizes.
The same absolute error of 0.2% or relative error of 1% (i.e. a positive bias) occurs
regardless of sample size, due to the presence of the iron. This type of error may go undetected
unless the constituents of the sample and the chemistry of the method are known.

21
Copper found (%)

20 bias
True value

19
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
Sample size (g)
Fig. 2. Effect of a proportional error on the determination of copper by titration in the

presence of iron.

Constant determinate errors are independent of sample size, and therefore become
less significant as the sample size is increased. For example, where a visual indicator is
employed in a volumetric procedure, a small amount oftitrant is required to change the
color at the end-point, even in a blank solution(i.e. when the solution contains none of the
species to be determined). This indicator blank (Topic C5) is the same regardless of the
size of the titer when the species being determined is present. The relative error, therefore,
decreases with the magnitude of the titer, as shown graphically in Figure 3. Thus, for an
indicator blank of 0.02 cm3, the relative error for a 1 cm3 titer is 2%, but this falls to only
0.08% for a 25 cm3 titer.

Indeterminate Known also as random errors, these arise from random fluctuations in measured
errors quantities, which always occur even under closely controlled condi- tions. It is impossible
to eliminate them entirely, but they can be minimized by careful experimental design and
control. Environmental factors such as temper- ature, pressure and humidity, and electrical
properties such as current, voltage and resistance are all susceptible to small continuous
and random variations described as noise. These contribute to the overall
indeterminate error in any

pg. 23
2.5

Relative error (%)

1.5

0.5

0
0 10 20 30
Size of titer (cm3)

Fig. 3. Effect of a constant error on titers of differing magnitudes.

physical or physico-chemical measurement, but no one specific source can be

identified.
A series of measurements made under the same prescribed conditions and
represented graphically is known as a frequency distribution. The frequency of
occurrence of each experimental value is plotted as a function of the magnitude
of the error or deviation from the average or mean value. For analytical data,
the values are often distributed symmetrically about the mean value, the most
common being the normal error or Gaussian distribution curve. The curve
(Fig. 4) shows that
● small errors are more probable than large ones,
● positive and negative errors are equally probable, and
● the maximum of the curve corresponds to the mean value.
The normal error curve is the basis of a number of statistical tests that can be applied
to analytical data to assess the effects of indeterminate errors, to compare values
and to establish levels of confidence in results (Topics B2 and B3).
Frequency of occurrence
of each deviation

– 0 +
Deviation from mean, 

Fig. 4. The normal error or Gaussian distribution curve.

Accumulated Errors are associated with every measurement made in an analytical procedure, and
errors these will be aggregated in the final calculated result. The accumulation or
propagation of errors is treated similarly for both determinate (systematic) and
indeterminate (random) errors.
Determinate (systematic) errors can be either positive or negative, hence some
cancellation of errors is likely in computing an overall determinate error, and in
some instances this may be zero. The overall error is calculated using one of two
alternative expressions, that is
● where only a linear combination of individual measurements is required to
compute the result, the overall absolute determinate error, E T, is given by
ET = E1 + E2 + E3 + …….
E1 and E2 etc., being the absolute determinate errors in the individual
measurements taking sign into account
● where a multiplicative expression is required to compute the result, the
overall relative determinate error, ETR, is given by
ETR = E1R + E2R + E3R + …….
E1R and E2R etc., being the relative determinate errors in the individual measure-
ments taking sign into account.
The accumulated effect of indeterminate (random) errors is computed by
combining statistical parameters for each measurement (Topic B2).

pg. 25
A SSESSMENT OF ACCURACY AND
PRECISION

Key Notes

Accuracy and Accuracy is the closeness of an experimental measurement or result to

precision the true or accepted value. Precision is the closeness of agreement
between replicated measurements or results obtained under the same
prescribed conditions.

Standard deviation The standard deviation of a set of values is a statistic based on the normal
error (Gaussian) curve and used as a measure of precision.

Relative standard Relative standard deviation (coefficient of variation) is the standard

deviation deviation expressed as a percentage of the measured value.

Pooled standard A standard deviation can be calculated for two or more sets of data by
deviation pooling the values to give a more reliable measure of precision.

Variance This is the square of the standard deviation, which is used in some
statistical tests.

Overall precision An estimate of the overall precision of an analytical procedure can be

made by combining the precisions of individual measurements.

Confidence interval This is the range of values around an experimental result within which
the true or accepted value is expected to lie with a defined level of
probability.

Related topic Errors in analytical measurements (B1)

Accuracy and These two characteristics of numerical data are the most important and the most
precision frequently confused. It is vital to understand the difference between them, and
this is best illustrated diagrammatically as in Figure 1. Four analysts have each
performed a set of five titrations for which the correct titer is known to be
20.00 cm3. The titers have been plotted on a linear scale, and inspection reveals
the following:
● the average titers for analysts B and D are very close to 20.00 cm3 - these two
sets are therefore said to have good accuracy;
● the average titers for analysts A and C are well above and below 20.00 cm3
respectively - these are therefore said to have poor accuracy;
● the five titers for analyst A and the five for analyst D are very close to one
another within each set – these two sets therefore both show good precision;
● the five titers for analyst B and the five for analyst C are spread widely
within each set -these two sets therefore both show poor precision.
Correct
result

D
19.70 20.00 20.30

Titer (cm3)

Fig. 1. Plots of titration data to distinguish accuracy and precision.

It should be noted that good precision does not necessarily produce good accuracy
(analyst A) and poor precision does not necessarily produce poor accuracy (analyst B).
However, confidence in the analytical procedure and the results is greater when good
precision can be demonstrated (analyst D).
Accuracy is generally the more important characteristic of quantitative data to
be assessed, although consistency, as measured by precision, is of particular concern in some
circumstances. Trueness is a term associated with accuracy, which describes the closeness of
agreement between the average of a large number of results and a true or accepted reference
value. The degree of accuracy required depends on the context of the analytical problem;
results must be shown to be fit for the purpose for which they are intended. For example, one
result may be satisfactory if it is within 10% of a true or accepted value whilst it may be
necessary for another to be within 0.5%. By repeating an analysis a number of times and
computing an average value for the result, the level of accuracy will be improved, provided that
no systematic error (bias) has occurred. Accuracy cannot be established with certainty
where a true or accepted value is not known, as is often the case. However, statistical tests
indicating the accuracy of a result with a given probability are widely used (vide infra).
Precision, which is a measure of the variability or dispersion within a set of
replicated values or results obtained under the same prescribed conditions, can be assessed
in several ways. The spread or range (i.e. the difference between the highest and lowest value)
is sometimes used, but the most popular method is to estimate the standard deviation of the
data (vide infra). The precision of results obtained within one working session is known as
repeatability or within-run precision. The precision of results obtained over a series of
working sessions is known as reproducibility or between-runs precision. It is sometimes
necessary to separate the contributions made to the overall precision by within-
run and

pg. 27
between-runs variability. It may also be important to establish the precision of
individual steps in an analysis.

Standard This is the most widely used measure of precision and is a parameter of the normal
deviation error or Gaussian curve (Topic B1, Fig. 4). Figure 2 shows two curves for the
frequency distribution of two theoretical sets of data, each having an infinite number
of values and known as a statistical population.
Frequency of occurrence

sd = 2
of each deviation

1 > 2

sd = 1

– m +
Deviation from mean

Fig. 2. Normal error or Gaussian curves for the frequency distributions of two statistical
populations with differing spreads.

The maximum in each curve corresponds to the population mean, which for
these examples has the same value, m. However, the spread of values for the two
sets is quite different, and this is reflected in the half-widths of the two curves at
the points of inflection, which, by definition, is the population stan- dard
deviation, s. As s2 is much less than s1, the precision of the second set is much
better than that of the first. The abscissa scale can be calibrated in absolute units
or, more commonly, as positive and negative deviations from the mean, m.
In general, the smaller the spread of values or deviations, the smaller the value
of s and hence the better the precision. In practice, the true values of m and s
can never be known because they relate to a population of infinite size. However, an
assumption is made that a small number of experimental values or
a statistical sample drawn from a statistical populatio n is also distributed
normally or approximately so. The experimental mean, x, of a set of values x1,
x2, x3,…….xn is therefore considered to be an estimate of the true or population
mean, m, and the experimental standard deviation, s, is an estimate of the true
or population standard deviation, s.
A useful property of the normal error curve is that, regardless of the magni-
tude of m and s, the area under the curve within defined limits on either side of
m (usually expressed in multiples of ±s) is a constant proportion of the total area.
Expressed as a percentage of the total area, this indicates that a particular
percentage of the population will be found between those limits.
Thus, approximately 68% of the area, and therefore of the population, will be
found within ±1s of the mean, approximately 95% will be found within ±2s and
approximately 99.7% within ±3s. More practically convenient levels, as shown in Figure 3,
are those corresponding to 90%, 95% and 99% of the population, which are defined by
±1.64s, ±1.96s and ±2.58s respectively. Many statistical tests are based on these
probability levels.

80% 90%
frequency (y/N)

frequency (y/N)
Relative

Relative
–4 –3 –2 –1 0 1 2 3 4 –4 –3 –2 –1 0 1 2 3 4

95% 99%
frequency (y/N)

frequency (y/N)
Relative

Relative

–4 –3 –2 –1 0 1 2 3 4 –4 –3 –2 –1 0 1 2 3 4

Fig. 3. Proportions of a population within defined limits of the mean.

The value of the population standard deviation, s, is given by the formula

i=N
¯
√
Σ (x − m) 2
i

s= — — i =1
(1)
N

where xi represents any individual value in the population and N is the total number of
values, strictly infinite. The summation symbol, , is used to show that the numerator of
the equation is the sum for i 1 to i N of the squares ofthe deviations of the individual x
values from the population mean, m. For verylarge sets of data (e.g., when N >50), it may
be justifiable to use this formula as the difference between s and s will then be negligible.
However, most analytical
data consists of sets of values of less than ten and often as small as three. Therefore, a
modified formula is used to calculate an estim  ated standard
deviation, s, to replace s, and using an experimental mean, x, to replace the
population mean, m:

¯
√ Σ (x − x) 2
i=N _
i
s= —
i= 1 — (2)

N−1

pg. 29
Note that N in the denominator is replaced by N - 1, which is known as the
number of degrees of freedom and is defined as the number of independent
deviations (xi − x) used to calculate s. For single sets of data, this is always one
less than the number in the set because when N- 1 deviations are known the last
i=N 
one can be deduced as, taking sign into account, Σ(xi − x) must be zero (see
i=1
Example 1 below).
In summary, the calculation of an estimated standard deviation, s, for a small
number of values involves the following steps:

● calculation of an experimental mean;

● calculation of the deviations of individual xi values from the mean;
● squaring the deviations and summing them;
● dividing by the number of degrees of freedom, N - 1, and
● taking the square root of the result.

Note that if N were used in the denominator, the calculated value of s would be
an underestimate of s.
Estimated standard deviations are easily obtained using a calculator that
incorporates statistical function keys or with one of the many computer soft- ware
packages. It is, however, useful to be able to perform a stepwise arithmetic
calculation, and an example using the set of five replicate titers by analyst A (Fig.
1) is shown below.

Example 1
xi/cm3 (xi − x) (xi − x)2
20.16 0.04 1.6  10 3

20.22 0.02 4  10 4
20.18 0.02 4  10 4
20.20 0.00 0
20.24 0.04 1.6  10 3

Â
_ 101.00 4 ¥ 10 3
x 20.20

s
= √¯
4  10
— — ¯ = 0.032 cm
−3
3
4

Relative The relative standard deviation, RSD or sr, is also known as the coefficient of
standard variation, CV. It is a measure of relative precision and is normally expressed as
deviation a percentage of the mean value or result

_
sr = (s/x)  100 (3)
It is an example of a relative error (Topic B1) and is particularly useful for
comparisons between sets of data of differing magnitude or units, and in calcu- lating
accumulated (propagated) errors. The RSD for the data in Example 1 isgiven below.

Example 2
0.032
sr = ——  100 = 0.16%
20.20
Pooled standard Where replicate samples are analyzed on a number of occasions under the sameprescribed
deviation conditions, an improved estimate of the standard deviation can be obtained by pooling the
data from the individual sets. A general formula for the pooled standard deviation, spooled,
is given by the expression

¯ ¯ ¯ (x − x ) + ...¯
i=N 1 i=N 2 i =N 3 i=N k
_ _ _ _
Σ (x − x ) + Σ
i 1 (x − x ) + Σ
2
i 2
2
+ Σ (x − x )
i 3
2
i k
2

√ i=1 i=1 i=1 i=1

spooled = ———i=k
——— (4)
Where n1,n2,n3,nk ΣNi = k
where N , N , N …N i=1 _ _
are the numbers of results in each of the k
sets, and x1, x2,
_

x3, . . . x k, are the means for each of the k sets.

Variance The square of the standard deviation, s 2, or estimated standard deviation, s 2, is used in a
number of statistical computations and tests, such as for calculating accumulated
(propagated) errors (Topic B1 and below) or when comparing the precisions of two sets of
data (Topic B3).

Overall precision Random errors accumulated within an analytical procedure contribute to the overall precision.
Where the calculated result is derived by the addition or subtraction of the individual
values, the overall precision can be found by summing the variances of all the
measurements so as to provide an estimate of the overall standard deviation, i.e.
soverall = √( 1¯ 22
s 2 + s¯ + s 32¯
+ . . .)

Example
In a titrimetric procedure, the buret must be read twice, and the error associated with each
reading must be taken into account in estimating the overall preci- sion. If the reading error
has an estimated standard deviation of 0.02 cm 3, then the overall estimated standard
deviation of the titration is given by

soverall = (¯ ¯
0.022 + 0.022̄) = 0.028 cm3
√
Note that this is less than twice the estimated standard deviation of a single reading. The
overall standard deviation of weighing by difference is estimated in the same way.
If the calculated result is derived from a multiplicative expression, the overall relative
precision is found by summing the squares of the relative standard devia- tions of all the
measurements, i.e.

sr(overall) = ¯ sr12 + r2s2¯

+ sr32 ¯
+ . . .)
√(

Confidence The true or accepted mean of a set of experimental results is generally unknown except
interval where a certified reference material is being checked or
analyzed for calibration purpos_es. In all other cases, an estimate of the accu- racy of the
experimental mean, x,_must be made. This can be done by defining a range of values on
either side of x within which the true mean, m, is expected to
lie with a defined level of probability. This range, which ideally should be as narrow as
possible, is based on the standard deviation and is known as the

pg. 31
confidence interval, CI, and the upper and lower limits of the range as confi-
dence limits, CL. Confidence limits can be calculated using the standard devia-
tion, s, if it is known, or the estimated standard deviation, s, for the data. In either
case, a probability level must be defined, otherwise the test is of no value.
When the standard deviation is already known from past history, the confi-
dence limits are given by the equation
_ zs
CL(m) = x  — (5)
√N̄

where z is a statistical factor related to the probability level required, usually 90%,
95% or 99%. The values of z for these levels are 1.64, 1.96 and 2.58, respec- tively,
and correspond to the multiples of the standard deviation shown in Figure 3.
Where an estimated standard deviation is to be used, s is replaced by s, which
must first be calculated from the current data. The confidence limits are then
given by the equation
_ ts
CL(m) = x  — (6)
√N̄

where z is replaced by an alternative statistical factor, t, also related to the prob-

ability level but in addition determined by the number of degrees of freedom
for the set of data, i.e. one less than the number of results. It should be noted that
(i) the confidence interval is inversely proportional to √N̄, and (ii) the higher
the selected probability level, the greater the confidence interval becomes as both
z and t increase. A probability level of 100 percent is meaningless, as the
confidence limits would then have to be  .
The following examples demonstrate the calculation of confidence limits
using each of the two formulae.

Example 3
The chloride content of water samples has been determined a very large number
of times using a particular method, and the standard deviation found to be
7 ppm. Further analysis of a particular sample gave experimental values of
350 ppm for a single determination, for the mean of two replicates and for the
mean of four replicates. Using equation (5), and at the 95% probability level,
z 1.96 and the confidence limits are:

1 determination — 7 350 ± 14 ppm

CL(m) = 350 ±1.96
√1̄

2 determinations — 7 350 ± 10 ppm

CL(m) = 350 ±1.96
√2̄

4 determinations — 7 350 ± 7 ppm

CL(m) = 350 ±1.96
√4̄

Example 4
The same chloride analysis as in Example 3, but using a new method for which
the standard deviation was not known, gave the following replicate results,
mean and estimated standard deviation:
Chloride/ppm Mean Estimated standard deviation
346 351.67 ppm 6.66 ppm
359
350
Using equation (6), and at the 95% probability level, t = 4.3 for two degrees of
freedom, and the confidence limits are:

3 determinations CL(m) = 352 ± 4.3

——  6.66 352 ± 17 ppm
√3̄
The wider limits given by equation (6) when the standard deviation is estimated
with only three results reflects the much greater uncertainty associated with this
value, which in turn affects the confidence in the degree of accuracy. To demon-
strate good accuracy, the confidence interval, CI, should be as small as possible
and increasing the number of replicates will clearly achieve this. However, due
to the √N̄ term in the denominator, to reduce the interval by, say, a factor of
two requires an increase in the number of replicates by a factor of four as shown
by Example 3. Unfortunately, the law of diminishing returns applies here, so if
the CI is to be halved again, the number of replicates must be increased from four
to sixteen. Similarly, in Example 4, the number of replicates would have to be
increased from three to twelve to halve the CI, which would represent an
unacceptable amount of time and money for most analytical laboratories.

pg. 33
Chapter three: SIGNIFICANCE TESTING

Significance tests These are statistical tests used to compare individual values or sets of
values for significant differences.

Outliers

Q-test The Ǫ-test is used to determine whether to reject or retain a suspected

outlier.

F-test The F-test enables the precisions of two sets of data to be compared using
their variances.

t-test

Analysis of variance

Assessment of accuracy and precision (B2)

Significance Significance tests involve a comparison between a calculated experimental factor

tests and a tabulated factor determined by the number of values in the set(s) of
experimental data and a selected probability level that the conclusion is correct.
They are used for several purposes, such as:
● to check individual values in a set of data for the presence of determinate errors
(bias);
● to compare the precision of two or more sets of data using their variances;
● to compare the means of two or more sets of data with one another or with
known values to establish levels of accuracy.
Tests are based on a null hypothesis an assumption that there is no signifi-
cant difference between the values being compared. The hypothesis is accepted
if the calculated experimental factor is less than the corresponding tabulated
factor, otherwise it is rejected and there is said to be a significant difference
between the values at the selected probability level. The conclusion should
always be stated clearly and unambiguously.
Probability levels of 90%, 95% and 99% are generally considered appropriate
for most purposes, but it should be remembered that there are also corre-
sponding 10%, 5% or 1% probabilities, respectively, of the opposite conclusion
being valid. For example, if a test indicates that the null hypothesis is correct and
that there is no significant difference between two values at the 95% proba- bility
level, it also allows the possibility that there is a significant difference at the 5%
level.
Separate tabular values for some significance test factors have been compiled
for what are described as one-tailed and two-tailed tests. The exact purpose of
the comparison that is to be made determines which table to use.
● The one-tailed test is used EITHER to establish whether one experimental
value is significantly greater than the other OR the other way around.
● The two-tailed test is used to establish whether there is a significant differ-
ence between the two values being compared, whether one is higher or lower
than the other not being specified.
The two-tailed test is by far the most widely used. Examples are given below.

Outliers Inspection of a set of replicate measurements or results may reveal that one or more
is considerably higher or lower than the remainder and appears to be outside the
range expected from the inherent effects of indeterminate (random) errors alone.
Such values are termed outliers, or suspect values, because it is possible that they
may have a bias due to a determinate error. On occasions, the source of error may
already be known or it is discovered on investigation, and the outlier(s) can be
rejected without recourse to a statistical test. Frequently, however, this is not the
case, and a test of significance such as the Ǫ-test should be applied to a suspect value
to determine whether it should be rejected and therefore not included in any further
computations and statistical assessments of the data.

Q-test Also known as Dixon’s Q-test, this is one of several that have been devised to
test suspected outliers in a set of replicates. It involves the calculation of a ratio,
Ǫexptl, defined as the absolute difference between a suspect value and the value
closest to it divided by the spread of all the values in the set:
Ǫexptl - suspect value − nearest value /(largest value − smallest value)
Ǫexptl is then compared with a tabulated value, Ǫtab, at a selected level of proba-
bility, usually 90% or 95%, for a set of n values (Table 1). If Ǫexptl is less than Ǫtab,
then the null hypothesis that there is no significant difference between the
suspect value and the other values in the set is accepted, and the suspect value
is retained for further data processing. However, if Ǫexptl is greater than Ǫtab, then the
suspect value is regarded as an outlier and is rejected. A rejected value should
NOT be used in the remaining calculations.

Table 1. Critical values of Q at the 95% ( P = 0.05) level for a two-tailed test

Sample size Critical value

4 0.831
5 0.717
6 0.621
7 0.570
8 0.524

Example 1
Four replicate values were obtained for the determination of a pesticide in river
water
0.403, 0.410, 0.401, 0.380 ug dm 3

pg. 35
Inspection of the data suggests that 0.380 g dm 3 is a possible outlier.
Ǫexptl = 0.380 - 0.401 /(0.410 - 0.380) = 0.021/0.03 = 0.70
Ǫtab = 0.83 for four values at the 95% probability level
As Ǫexptl is less than Ǫtab, 0.380 g dm 3 is not an outlier at the 95% level and
should be retained.

Example 2
If, in Example 1, three additional values of 0.400, 0.413 and 0.411 g dm 3 were
included, 0.380 g dm 3 is still a possible outlier.
Ǫexptl = 0.380 - 0.400 /(0.413 - 0.380) =0.020/0.033 = 0.61
Ǫtab = 0.57 for seven values at the 95% probability level
Now, as Ǫexptl is greater than Ǫtab, 0.380 g dm 3 is an outlier at the 95% level and
should be rejected. Note that because the three additional values are all around
0.4 g dm 3, the suspect value of 0.380 g dm 3 appears even more anomalous.

F-test This test is used to compare the precisions of two sets of data which may origi-
nate from two analysts in the same laboratory, two different methods of analysis
for the same analyte or results from two different laboratories. A statistic, F, is
defined as the ratio of the population variances, s 12/s 22, or the sample variances,
s 2/s 2, of the two sets of data where the larger variance is always placed in the
1 2
numerator so that F 1.
If the null hypothesis is true, the variances are equal and the value of F will be one
or very close to it. As for the Ǫ-test, an experimental value, Fexptl, is calcu- lated
and compared with a tabulated value, Ftab, at a defined probability level, usually 90%
or 95%, and for the number of degrees of freedom, N- 1, for each set of data. If
Fexptl is less than Ftab, then the null hypothesis that there is no significant difference
between the two variances and hence between the preci- sion of the two sets of data,
is accepted. However, if Fexptl is greater than Ftab,there is a significant difference
between the two variances and hence between the precisions of the two sets of
data.
Some values of Ftab at the 95% probability level are given in Table 2. The
columns in the table correspond to the numbers of degrees of freedom for the
numerator set of data, while the rows correspond to the number of degrees of
freedom for the denominator set. Two versions of the table are available,
depending on the exact purpose of the comparison to be made: a one-tailed F-
test will show whether the precision of one set of data is significantly better than
the other, while a two-tailed F-test will show whether the two precisions are
significantly different.

Table 2. Critical values of F at the 95% ( P = 0.05) level for a two-tailed test

n1 5 7 9

n2
5 7.146 6.853 6.681
7 5.285 4.995 4.823
9 4.484 4.197 4.026

n1 number of degrees of freedom of the numerator. n2 number of degrees of freedom of the

denominator
The application of a two-tailed F-test is demonstrated by the following
example.

Example 3
A proposed new method for the determination of sulfate in an industrial waste
effluent is compared with an existing method, giving the following results:

Method Mean/g dm 3
No. of No. of degrees s/mg dm 3

replicates of freedom
Existing 72 8 7 3.38
New 72 8 7 1.50

Is there a significant difference between the precisions of the two methods?

⁄
s2 existing (3.38)2
Fexptl = 2 2
s new
= — = 5.08
(1.50)

The two-tailed tabular value for F with 7 degrees of freedom for both the
numerator and the denominator is
F7,7 = 5.00 at the 95% probability level
As Fexptl is greater than Ftab, the null hypothesis is rejected; the two methods are
giving significantly different precisions.

t-test This test is used to compare the experimental means of two sets of data or to
compare the experimental mean of one set of data with a known or reference
value. A statistic, t, is defined, depending on the circumstances, by one of three
alternative equations.
_ _
Comparison of two experimental means, xA and xB
_ _
NM 1⁄ 2
( )
(xA − xB)
t — — (1)
spooled N+M
where spooled is the pooled estimated standard deviation (Topic B2) for sets A and
B, and N and M are the numbers of values in sets A and B respectively. If N M,
1

then the second term reduces to (N/2) ⁄ . A simplified version of equation (4),
2

Topic B2, can be used to calculate spooled as there are only two sets of data.

{[(N − 1)s + (M − 1)s ]/[N + M − 2]}

1
⁄2
spooled 2
A
2
B (2)

In some circumstances, the use of equation (1) may not be appropriate for the
comparison of two experimental means. Examples of when this may be the case
are if
● the amount of sample is so restricted as to allow only one determination by
each of the two methods;
● the methods are to be compared for a series of samples containing different
levels of analyte rather than replicating the analysis at one level only;
● samples are to be analyzed over a long period of time when the same experi-
mental conditions cannot be guaranteed.
It may therefore be essential or convenient to pair the results (one from each
method) and use a paired t-test where t is defined by
pg. 37
x_d 1
⁄
t= — N 2
(3)
sd
_
xd being the mean difference between paired values and sd the estimated
standard deviation of the differences.

Comparison of one experimental mean with a known value, m

_
(x − m) ⁄ 1

t=—N 2
(4)
s
Using the appropriate equation, an experimental value, texptl, is calculated and
compared with a tabulated value, ttab, at a defined probability level, usually
between 90 and 99%, and for N- 1 degrees of freedom (equations (3) and (4)) or (N
M - 2) degrees of freedom (equation (1)). If texptl is less than ttab, then the null
hypothesis that there is no significant difference between the two experimental
means or between the experimental mean and a known value is accepted, i.e.
there is no evidence of a bias. However, if texptl is greater than ttab, there is a
significant difference indicating a bias.
Both one-tailed and two-tailed t-tests can be used, depending on circum- stances,
but two-tailed are often preferred (Table 3). The application of all three t-test
equations is demonstrated by the following examples.

Table 3. Critical values of t at the 95% and 99% ( P 0.05 and 0.01) levels for a
two-tailed test
Number of degrees of freedom 95 percent level 99 percent level
2 4.30 9.92
5 2.57 4.03
10 2.23 3.10
18 2.10 2.88

Example 1
Two methods for the determination of polyaromatic hydrocarbons in soils were
compared by analyzing a standard with the following results:
No. of determinations by each method: 10
No. of degrees of freedom: 1_8
UV spectrophotometry: x_ = 28.00 mg kg 1 s = 0.30 mg kg 1
Fluorimetry: x = 26.25 mg kg 1 s = 0.23 mg kg 1
Do the mean results for the two methods differ significantly?
Equation (2) is first used to calculate a pooled standard deviation:

{[ ][ ]} = {(9 ¥ 0.3 +9 x0.23 )/18}

1
⁄2
spooled = (N − 1)s2A+ (M − 1)s2B / N + M − 2
1
2 2 ⁄2

spooled = 0.267 mg kg−1

Then equation (1) is used to evaluate texptl
_ _ ⁄ 1

(xA − xB) NM
( )
2

texptl = ——  — = {(28.0 -26.25)/0.267} ¥ 5 ⁄

1
2
14.7
spooled N+M

For 18 degrees of freedom, the two-tailed value of ttab at the 95% probability
level is 2.10, and at the 99% level it is 2.88.
As texptl is greater than ttab at both the 95 and 99% probability levels, there is a
significant difference between the means of the two methods.

Example 2
A new high performance liquid chromatographic method for the determination
of pseudoephedrine in a pharmaceutical product at two different levels was
compared with an established method with the following results:
Pseudoephedrine per dose (mg)
Method 1 Method 2
59.9 58.6
59.3 58.3
60.4 60.5
30.7 29.4
30.2 30.4
30.1 28.9

Do the means of the two methods differ significantly?

Because the two levels of pseudoephedrine differ considerably, equation (3)
for a paired t-test is used to calculate texptl. The differences between the pairs of
values are 1.3, 1.0, 0.1, 1.3, 0.2 and 1.2 mg per dose, and the estimated
standard deviation of the differences from their mean of 0.750 mg per dose is
0.706 mg per dose. Substitution of these values into the equation gives
_
xd  N ⁄
texptl ——
1

(0.750/0.706) ¥ 6 ⁄
1
2
2.60 2

sd

For 5 degrees of freedom, the two-tailed value of ttab at the 95% probability levelis 2.57. As
texptl is greater than ttab, there is a significant difference between the means of the two
methods. (Note: using equation (1) would give a texptl value of
0.08 and an incorrect conclusion.)

Example 3
A method for the determination of mercury by atomic absorption spectrometry gave values of
400, 385 and 382 ppm for a standard known to contain 400 ppm. Does the mean value differ
significantly from the true value, or is there any evidence of systematic error (bias)?
_
x = 389 ppm s = 9.64 ppm m = 400 ppm

Using equation (4) to evaluate texptl

_
(x − m) 1⁄ (389 − 400) 1⁄
texptl —  N 2 ——  3 2 1.98
s 9.64

For 2 degrees of freedom, the two-tailed ttab value is 4.30 at the 95% probability level. As
texptl is less than the two-tailed value of ttab, the mean is not significantly different from the
true value. There is, therefore, no evidence of a systematic error, or bias.

Analysis of Analysis of variance, also known as ANOVA, is a statistical technique for inves- tigating
variance different sources of variability associated with a series of results. It enables the effect of each
source to be assessed separately and compared with the other(s) using F-tests.
Indeterminate or random errors affect all measure- ments, but additional sources of variability
may also arise. The additional sources can be divided into two types:

pg. 39
● additional random effects, described as random-effect factors;
● specific effects from determinate sources, described as controlled or fixed-
effect factors.
Where one additional effect may be present, a one-way ANOVA is used, whilst
for two additional effects, two-way ANOVA is appropriate. Both involve much
lengthier calculations than the simpler tests of significance, but facilities for
these are available with computer packages such as Microsoft Excel and Minitab.
Typical examples of the use of ANOVA are:
● analysis of a heterogeneous material where variation in composition is an
additional random factor;
● analysis of samples by several laboratories, methods or analysts where the
laboratories, methods or analysts are additional fixed-effect factors;
● analysis of a material stored under different conditions to investigate stability
where the storage conditions provide an additional fixed-effect factor.
C ALIBRATION AND LINEAR
REGRESSION

Key Notes
Calibration Calibration is the process of establishing a relation between a detection or
measurement system and known amounts or concentrations of an analyte
under specified conditions.

Correlation The coefficient is used to assess the degree of linearity between two
coefficient variables, e.g. an instrument response and an analyte mass or
concentration.

Linear regression Calculations to define the best straight line through a series of calibration
points represented graphically are described as linear regression.

Limit of detection The smallest mass or concentration of an analyte that can be measured
quantitatively at a defined level of probability defines a limit of detection.

Standard addition This is a calibration procedure that avoids matrix interference by

measuring instrument response for an analyte in both the sample and a
sample to which known amounts of an analyte standard have been
added.

Internal This is a calibration procedure where the ratio of the instrument response
standardization for an analyte to that of an added standard is measured for a series of
analyte standards and samples.

Internal Internally normalized results give the relative composition of a mixture

normalization by expressing the instrument response for each analyte as a fraction or
percentage of the sum of the responses for all of the analytes.

Related topic Calibration and standards (A5)

Calibration Many quantitative analytical procedures rely on instrumental measurements where

a property of the analyte(s) is monitored by a suitable detection system. The
detector generates an electrical signal, the magnitude of which is deter- mined by
the mass or concentration of the analyte. Before using a particular analytical
procedure to analyze samples, it is first necessary to establish the detector
responses to known amounts of the analyte (calibration standards) over a
selected mass or concentration range, a process known as calibration. The
relation between the two variables is often linear (directly proportional), but
there is generally an upper limit to the range of values beyond which a curved or
curvilinear relation is observed. In some instances, there may be no direct linear
relation at all, or a logarithmic or more complex mathematical correlation may be
found.

pg. 41
Calibration data are generally used to construct a calibration graph, where
detector response is plotted on the ordinate axis (y-values) and mass or concen-
tration of the analyte on the abscissa axis (x-values) as shown in Figure 1.
The graphs are often linear, being defined by the equation
y bx a (1)
where b is the slope and a the intercept on the y-axis. In some cases, it is prefer- able
to plot a logarithmic function of the detector response or analyte concentra- tion to
obtain a linear calibration curve.
Unknown levels of the analyte are determined from the graph by interpola-
tion. Where a linear relation has been established, a calibration factor can be used
to convert detector response to mass or concentration of analyte when analyzing
samples.
Theoretically, the graph should pass through the origin, but frequently in
practice there is a small positive intercept due to traces of analyte in the reagent
blank or contributions to the detector signal by other components in the
standards. Calibration points also show a degree of scatter due to the effects of
experimental errors in preparing the standards, or noise in the measuring
circuitry. A line of best fit through the points, known as a regression line, is
therefore drawn or computed.
Calibration graphs may show curvature, particularly at higher mass or
concentration levels, but this does not invalidate their use if the data are repro-
ducible. However, it is advisable to prepare additional standards to define thecurve
more closely, and the use of a factor to compute analyte levels in samples is
precluded.
Statistical methods are used to assess calibration data
● for linearity or otherwise;
● to calculate the parameters defining a calibration curve;
● to assess the effects of determinate and indeterminate errors on standards and
samples.

25

20
Detector response

Sample response
15

10

5 Interpolated sample
mass/concentration
0
0 2 4 6 8 10 12
Analyte mass/concentration
Fig. 1. A typical calibration graph.

Correlation The correlation coefficient, r, indicates the degree of linearity between x and y
coefficient and is given by the expression
i=N _ _
Σ{(xi − x)(yi − y)}
r= —i=1 —— (2)
i=N i=N
_ (yi − y) ]}
{[Σ (xi − x) ][Σ
2 2 _ 1
⁄2
i=1 i=1 _ _
where x y ; x y ; x y ;….x ,y are the co-ordinates of the plotted points, and
1 1 2 2 3 3 n n x y
are the means of the x and y values respectively, and  indicates sums of terms (see
standard deviation equations (1), (2) and (4), Topic B2).
The range of possible values for r is -1 ≤ r ≤ +1. A value of unity indicates a perfect linear
correlation between x and y, all the points lying exactly on a straight line, whilst a value of
zero indicates no linear correlation. Values may
be positive or negative depending on the slope of the calibration graph. These alternatives are
illustrated in Figure 2 (a) to (c).
Most calibration graphs have a positive slope, and correlation coefficients frequently exceed
0.99. They are normally quoted to four decimal places. (Note that graphs with a slight
curvature may still have correlation coefficients exceeding about 0.98 (Fig. 2(d)), hence great
care must be taken before concluding that the data shows a linear relation. Visual inspection
of the plotted points is the only way of avoiding mistakes.)

Linear When inspection of the calibration data and the value of the correlation coeffi- cient show
regression that there is a linear relation between the detector response and the

(a) (b)
10
10
8
Detector response

Detector response

8
6 r = –1
6
4
4
2
2
0
0 0 2 4 6 8 10
0 2 4 6 8 10
Analyte mass/concentration
Analyte mass/concentration

(c) (d)
10
10
8
Detector response

Detector response

8 r = 0.9850
6
6
4
4
2
2
0
0 0 2 4 6 8 10
0 2 4 6 8 10
Analyte mass/concentration
Analyte mass/concentration

Fig. 2. Examples of correlation coefficients. (a) Perfect positive correlation; (b) perfect negative correlation; (c) no
correlation, and (d) curved correlation.

pg. 43
mass or concentration of the analyte, it is necessary to draw a line of best fit through
the plotted points before they can be used as a working curve. Although this can be
done by eye, a more accurate method is to employ linear regression. It is invariably
the case that, due to the effects of indeterminate errors on the data, most of the
points do not lie exactly on the line, as shown in Figure 1. Linear regression enables
a line of best fit through the points to be defined by calculating values for the slope
and y-axis intercept (b and a respectively in equation (1)), and the method of least
squares is commonly used for this purpose. An assumption is made that only errors
in the detector responses (y-values) are significant, any errors in the values for the
mass or concentration of the analyte being neglected. The deviations in the y-
direction of the individual plotted points from the calculated regression line are
known as y-residuals (Fig. 3) and the line repre- sents the regression of y upon x.
The method of least squares minimizes the sum of the squares of the y-residuals
by equating them to zero in defining equations for the slope and intercept of the
regression line.
For the slope, b
i=N _ _
Σ{ (xi − x)(yi − y )}
b= —
i=1 i=N — (3)
2 _
Σ
i=1
(xi − x)

For the y-axis intercept, a

_ _
a = y − b.x (4)
N
_ .B
_ . As equation (4) is a re-arrangement of equation (1), it follows that the point
x, y, known as the centroid, must lie on the regression line.

1.2

1
Absorbance at 325 nm

0.8 Slope = 0.00878

y-residuals
0.6

0.4

0.2

Intercept = 0.0686
0
0 20 40 60 80 100 120
Concentration (mg cm–3)
Fig. 3. Calibration graph, regression line, slope and intercept values for the UV spectrophoto-
metric de_ter_mination of an active ingredient in a sun cream. ——— regression line; o
centroid, x, y; ------------ confidence limits lines at the 99 percent level; ]——} confidence
limits for sample concentration of 30 mg cm 3. Inset: Illustration of y-residuals.
Example
A calibration graph was prepared as part of a validation procedure for a new method to
determine an active constituent of a sun cream by UV spectrophoto- metry. The following data
were obtained:

Analyte conc. (mg cm 3) 0 20 40 60 80 100 120

UV absorbance at 325 nm 0.095 0.227 0.409 0.573 0.786 0.955 1.123

The data is first checked for linearity by calculation of the correlation coefficient, r, and
visual inspection of a plotted curve. Some calculators and computer soft- ware can perform
the computation from the raw data, but it is instructive to show the full working, for which
tabulation is preferable.
_ _2 _
x y (x − _2 _ _
x) (xi − x) (yi − y) (yi − y) (xi − x )(yi − y)
i i
0 i
0.095 −60 3600 −0.5004 0.2504 30.024
20 0.227 −40 1600 −0.3684 0.1357 14.736
40 0.409 −20 400 −0.1864 0.0347 3.728
60 0.573 0 0 −0.0224 0.0005 0
80 0.786 20 400 0.1906 0.0363 3.812
100 0.955 40 1600 0.3596 0.1293 14.384
120 1.123 60 3600 0.5276 0.2784 31.656

_ 420 _ 4.168 0 11200 0 0.8653 98.340

x = 60 y 0.59543

Substitution of the totals in columns 4, 6 and 7 in equation (2) gives

r = 98.340/(11200 x 0.8653)1/2 = 98.340/98.445 =0.9989

Figure 3 and the correlation coefficient of 0.9989 show that there is a good linear relation
between the measured UV absorbance and the analyte concentration.
The slope and y-axis intercept of the regression line, given by equations (3) and (4)
respectively are

b = 98.340/11200 = 0.00878 a = 0.59543 = (0.00878 x 60) = 0.0686

The y-axis intercept, slope and analyte masses or concentrations calculated by

interpolation from the regression line are all affected by errors. Additional equations can be
used to obtain the following statistics:

● estimated standard deviations for the slope and intercept;

● estimated standard deviations for analyte masses or concentrations deter- mined from the
calibration graph;
● confidence limits for analyte masses and concentrations at selected proba- bility levels;
● limit of detection of the analyte (vide infra).
Confidence limits (Topic B2) over the entire range of the calibration graph at selected
probability levels, e.g. 95 or 99 percent, can be displayed (dashed curves, Fig. 3). A
horizontal line drawn through a given experimental point on the regression line and
intersecting the confidence limits lines on either side
gives the upper and lower limits for that particular mass or concentrat_ion_.
Figure 3 shows the 99% limits, the narrowest interval being at the centroid, x, y,
of the graph, and widening steadily towards each end.
Some calculators and computer packages have the ability to perform the regression
calculations described. Where there is a nonlinear relation between

pg. 45
the detector response and the mass or concentration of the analyte more
complex curvilinear or logarithmic regression calculations are required.

Limit of For any analytical procedure, it is important to establish the smallest amount of
detection an analyte that can be detected and/or measured quantitatively. In statistical
terms, and for instrumental data, this is defined as the smallest amount of an
analyte giving a detector response significantly different from a blank or back-
ground response (i.e. the response from standards containing the same reagents
and having the same overall composition (matrix) as the samples, where this is
known, but containing no analyte). Detection limits are usually based on esti-
mates of the standard deviation of replicate measurements of prepared blanks.
A detection limit of two o_r three times the estimated standard deviation of the
blanks above their mean, xB, is often quoted, where as many blanks as possible
(at least 5 to 10) have been prepared and measured.
This is somewhat arbitrary, and it is perfectly acceptable to define alternatives
provided that the basis is clear and comparisons are made at the same proba-
bility level.

Standard Where components of a sample other than the analyte(s) (the matrix) interfere
addition with the instrument response for the analyte, the use of a calibration curve based
on standards of pure analyte may lead to erroneous results. Such matrix
interference effects can be largely if not entirely avoided by preparing calibra-
tion standards where known amounts of pure analyte are added to a series of
equal sized portions of the sample, a procedure known as spiking. In addition,
one portion of sample is not spiked with analyte. (Note: if spiking sample solu-
tions with analyte changes the volume significantly, volume corrections must be
applied.)
The effects of the matrix on measurements of the analyte in both the spiked
and unspiked samples should be identical. The instrument responses are then used
to construct a calibration graph where the x-axis values are the added amounts of
analyte and the response for the unspiked sample is at x 0 (i.e., the curve does NOT
pass through the origin). The regression line is calculated and extrapolated back to
give a negative intercept on the x-axis at y 0, which corre- sponds to the amount
of analyte in the sample (Fig. 4).
The less reliable procedure of extrapolation rather than interpolation is
outweighed by the advantage of eliminating or minimizing matrix interference.
The method of standard addition is widely used, particularly when the
composition of the sample matrix is variable or unknown so that the response of
a reagent/matrix blank would be unreliable. At least three and preferably more
spiked samples should be prepared, but if the amount of sample is limited, as few
as one sample must suffice. It is especially useful with such analytical tech -
niques as flame and plasma emission spectrometry and potentiometry (Topics
E4, E5 and C8).

Example
The calcium level in a clinical sample was determined by flame emission
spectrometry using a standard addition method, which gave the following data:
Spiked calcium (ppm) 0 10 20 30 40 50
Emission intensity 0.257 0.314 0.364 0.413 0.468 0.528
at 423 nm
0.7

0.6 Slope = 0.005349

Emission intensity at 423 nm

0.5

0.4
Spiked samples

0.3
Unspiked
sample
0.2
Sample concentration
48 ppm
0.1
Intercept = 0.2569

–60 –50 –40 –30 –20 –10 0 10 20 30 40 50 60

Calcium (ppm)

Fig. 4. Standard addition calibration graph, regression line, slope and intercept values for the
flame emission determination of calcium in a clinical sample.

Detailed calculations of the correlation coefficient, r, and the slope and intercept
values have not been given, but should be set out as in the previous example if a
suitable calculator or computer program is not available.
The amount of calcium in the sample can be read from the extrapolated graph
or calculated from the slope, b, and the intercept, a
Calcium concentration in sample = a/b =0.2569/0.005349 = 48 ppm

Internal For some analytical techniques, particularly chromatography, variations in

standardization experimental conditions can adversely affect the precision of the data. A
calibration procedure whereby a constant amount of a selected substance, the
internal standard is added to all samples and analyte standards alike
compensates for variations in sample size and other parameters. The ratio of the
detector response for the analyte in each standard to the corresponding
response for the added internal standard is plotted on the y-axis of a calibration
graph against the mass or concentration of the analyte on the x-axis. The corre-
lation coefficient, slope and intercept values can be computed as shown previ-
ously, and response ratios for the analyte and added internal standard in the
samples can then be used to determine the amount of analyte in the samples by
interpolation on the graph. If only one or two analyte standards are prepared,
the amount of analyte in a sample can be calculated by simple proportion, i.e.
analyte in sample response ratio for sample
— — —— = — — — ——
analyte in standard response ratio for standard

Internal For some purposes, only the relative amounts of the analytes in a multicomponent
normalization mixture are required. These are normalized to 100 or 1 by expressing each as a

pg. 47
percentage or fraction of the total. Internal normalization is of particular value in
quantitative chromatography where several components of a sample can be
determined simultaneously, and absolute levels are not of interest. The relative
composition is calculated from the instrument response, peak area in the case of
a chromatographic analysis, for each component in the mixture using the formula
Ax
%xi = —
i=n  100
Σ Ai
i=1

where xi is one of n components and A is the measured area or response.

Example
Figure 5 is a chromatographic record (chromatogram) of the separation of a 5-
component mixture. The measured peak areas (using electronic integration with
a computing-integrator, computer and chromatography data processing software
or geometric construction such as triangulation, 1⁄2 ¥ base ¥ height) and percent-
ages by internal normalization, which must total 100 percent, are given in Table 1
(e.g., for component 1, relative percent =(167.8/466.94) x100 = 35.9 percent).

Table 1. Peak areas and percentage composition by internal normalization for a

5-component mixture
Component Measured peak area Relative percent
(arbitrary units)
1 167.8 35.9
2 31.63 6.8
3 108.5 23.2
4 80.63 17.3
5 78.38 16.8
Totals 466.94 100.0
Inject

0 3 6 9 12 15 18
Time (min)

Fig. 5. Chromatogram of a 5-component mixture.

Ǫ UALITY CONTROL AND
CHEMOMETRICS

Control charts Graphical representations of quantitative data from an ongoing series of

measurements can be used to monitor the stability of a system for quality
control (ǪC) purposes.

Collaborative testing Schemes have been devised to compare results for the analysis of
homogeneous samples or standard materials from groups of analytical
laboratories to test specified methods of analysis.

Multivariate
statistics

Ǫuality in analytical laboratories (A6)

Control charts The purpose of a control chart is to monitor data from an ongoing series of quantitative
measurements so that the occurrence of determinate (systematic) errors (bias),
or any changes in the indeterminate (random) errors affecting the precision of
replicates can be detected and remedial action taken. The predomi- nant use of
control charts is for quality control (ǪC) in manufacturing industries where a
product or intermediate is sampled and analyzed continually in a process stream
or periodically from batches. They may also be used in analytical laboratories,
such as those involved in clinical or environmental work, to monitor the
condition of reagents, standards and instrument components, which may
deteriorate over time.
Shewart charts consist of a y-axis calibrated either in the mass, concentration
or range of replicated results of an analyte or in multiples of the estimated stan-
dard deviation, s, of the analytical method employed, and the sample number
along the x-axis. An averages or X-chart, the most common type, is pre-
prepared with a series of five parallel horizontal lines, the centre one being posi-
tioned along the y-axis to correspond to the true, accepted or target value for the
analyte (Fig. 1). The other four lines are positioned in pairs on either side of the
target value line and act as critical levels that, when exceeded, indicate probable
instability in the system. The inner pair are defined as warning levels and the
outer pair as action levels.
An averages chart is used to monitor the level of an analyte either as single values
or as means of N replicates to check for determinate errors (bias) in the results. Chart
criteria and decisions are based on the following:
● plotted values have a Gaussian or normal distribution (Topic B2);
● warning lines are positioned to correspond to a selected probability level

pg. 49
3s/N1/2

2s/N1/2

Mass or concentration
of analyte, x x

2s/N1/2

3s/N1/2

0 5 10 15 20
Sample number

Fig. 1. A Shewart averages control chart.

multiple of the estimated standard deviation of the method, usually 95% or

±1.96s/N1/2;
● action lines are positioned to correspond to a selected probability level
or multiple of the estimated standard deviation of the method usually 99.7% or
±3.01s/N1/2;
● the pairs of warning lines will move inwards towards the target line as the
number of results per plotted point, N, increases, so the value of N must be
fixed before the chart is used;
● in the absence of determinate errors (bias), 95% of values should fall within
the upper and lower warning lines. The system is then considered to be stable
or under control;
● two or more consecutive values falling outside the warning lines but within
the action lines indicate a possible loss of control;
● two or more consecutive values falling outside the action lines indicate the
occurrence of one or more determinate errors; the system is then considered
to be out of control and remedial action should be taken;
● trends in plotted values are an indication of incipient problems, as are ten or
more consecutive values on one side of the target line.
The chart shows that the system is in control during the first 6 samples analyzed,
but the upper warning level has been breached by result 7. However, the next 8
results are within the warning levels, but results 17 to 20 indicate a downward trend
culminating with both lower limits being breached indicatinga loss of control due
to one or more determinate errors. At this point, the causes should be sought and
remedial action taken.

Collaborative The principal purpose of collaborative testing is to assess the accuracy of

testing results from a group of laboratories performing one or more specific quantita-
tive determinations. Interlaboratory studies are also used to assess other perfor-
mance characteristics of a particular method or the competence of individual
analysts in relation to a specified technique, analyte or group of analytes. They
are frequently used by trade associations, standards organizations or agencies
with stringent analytical requirements to develop, test and validate analytical methods.
Proficiency testing schemes are designed to test the competence of individual
laboratories to perform specific analyses or a range of determinations of analytes in
various matrixes and possibly using alternative techniques.
Some typical examples of collaborative studies and proficiency testing are

● alcohol in beverages;
● metals or volatile organics in soils;
● hazardous airborne substances;
● trace metals or pesticides in domestic water supplies;
● drugs of abuse by chromatographic methods;
● adulterants in foodstuffs;
● additives in polymers and composites.
In statistical terms, collaborative trials are designed to reveal the presence of a bias in the
results arising from determinate errors that may occur in addition to inherent indeterminate
(random) errors. Statistics that are used in these schemes include estimated standard
deviation and coefficient of variation, confidence limits (Topic B2) and tests of significance
including ANOVA (Topic B3). Of particular value is a z-score to indicate the accuracy of
results from the partici- pating laboratories. The z-value is a statistical factor based on a
Gaussian or normal distribution, which is included in the equation for confidence limits of
an experimental value in relation to a true or accepted value (Topic B2, equation (5)).
Rearrangement of this equation, for a single determination, xi, (N 1) gives
xi − m
z = |—|
s
where m is the true or accepted mass or concentration of the analyte and s is a value for the
standard deviation for the method selected by the organizers of the study (strictly an
estimated value, s). A typical chart of z-scores for a collabora- tive study involving 22
laboratories is shown in Figure 2.
Results that have z-scores of between 1.96 and 1.96 are considered to have acceptable
accuracy, as these values correspond to confidence limits at the 95% probability level (Topic
B2, Fig. 3). In the example, only laboratories 12 and 18

1
z-score

–1

–2

–3
11 8 6 14 2 17 7 4 9 3 12 18
19 13 21 22 20 10 5 15 16 1

Laboratory number

Fig. 2. z-score chart for results from 22 laboratories.

pg. 51
have results with unacceptable accuracy, whilst laboratories 5 and 15 have results
closest to the true or accepted value. In practice, the situation is oftenworse than
this, with some laboratories incurring worryingly large determinate errors that need
to be identified and rectified.
Although accuracy is of prime importance in collaborative studies and pro-
ficiency testing, precision should also be monitored. The level to be expected varies
considerably with the concentration of an analyte and type of sample. A useful guide
for different levels of an analyte as measured by the coefficient of variation (Topic
B2) is exemplified in Figure 3.

50
Coefficient of variation (%)

40

30

20

10

0
10% 1% 0.1% 0.01% 10 ppm 1 ppm 0.1 ppm 0.01 ppm 1 ppb

Analyte concentration (log scale)

Fig. 3. Guide to typical coefficients of variation as a function of analyte concentration.

Note that values of less than 1% should be attainable for major components of
a sample, whilst at ppb (parts per billion) levels, over 50% is acceptable.

Multivariate Computerized and automated analytical instrumentation facilitates the collec-

statistics tion of large amounts of data and the simultaneous monitoring of numerous
experimental parameters. To maximize the useful information that can be
extracted, sophisticated multivariate chemometric techniques are employed.
The mathematical computations involve matrix algebra and vectors, and rely
on the availability of specialized computer software. Typical applications include
the characterization of substances from profiles based on spectral,
chromatographic and other data, and quantitative analysis based on multiple
simultaneous measurements. Two important applications of multivariate
statistics are pattern recognition and multivariate modeling.

Pattern recognition
Sets of measurements characterizing a sample, e.g. the position of prominent
infrared absorption bands (Topic H4), significant mass spectral fragments,
levels of particular analytes, and selected physical properties, are described as
patterns. These can be used to classify substances or to identify unknowns by
pattern matching. Figure 4 shows the distribution of trace levels of copper and
manganese in twelve geological samples where three clusters are evident. Just
50

40

Manganese (ppm)
30

20

10

0
0 5 10 15
Copper (ppm)

Fig. 4. Copper and manganese distribution in geological samples showing three clusters with
differing proportions of each metal.

as two parameters can be plotted as a single point with specified x and yco-ordinates on a
two-dimensional graph, the values of n parameters can be represented by a point in n-
dimensional space. Although n-co-ordinate graphs cannot be visualized, they can be
studied through appropriate computer processing and manipulation. Where a number of
substances have similar sets of n co-ordinates, and therefore similar characteristics, they
produce closely-spaced groups of points described as clusters, the interpreta- tion of this
data being described as cluster analysis. Mathematical procedures to detect clusters
include principal component analysis (PCA) and factor analysis (FA), which seek to
simplify the data by projection from n dimen- sions onto a line, plane or 3-D graph to
reduce the number of dimensions without losing information.
Cluster analysis can be used in many ways, e.g. to monitor clinical specimens from hospital
patients where, for example, the levels of pH, glucose, potassium, calcium, phosphate and
specific enzymes vary according to the absence, pres- ence or severity of a particular disease.
It can also be applied to the characteriza- tion of glass fragments for forensic purposes
through profiling and comparisons of their trace metal contents, for identifying the source
of a crude oil spillage on the basis of the proportions of minor organic compounds and metals,
and in the classification of organic compounds with similar structural features to facilitate
the elucidation of unknown structures.

Multivariate modeling
Ǫuantitative analysis for one or more analytes through the simultaneous measurement of
experimental parameters such as molecular UV or infrared absorbance at multiple
wavelengths can be achieved even where clearly definedspectral bands are not discernible.
Standards of known composition are used to compute and refine quantitative calibration
data assuming linear or nonlinear models. Principal component regression (PCR) and
partial least squares (PLS)
regression are two multivariate regression techniques developed from linear
regression (Topic B4) to optimize the data.

pg. 53
Multivariate modeling is the basis of quantitative analysis by near infrared
spectrometry (Topic E11) to determine moisture or fats in cereals, meat and
other foodstuffs. It is particularly applicable to environmental analysis where
complex and inter-related variables affect the distributions of a wide range of
organic compounds and elements occurring naturally or as pollutants.
Chapter four :SOLUTION EǪUILIBRIA

Key Notes
Solvents The major component of a solution is referred to as the solvent, and there
is a wide range of inorganic and organic solvents used in analytical
chemistry. Their properties determine their use.

Solubility When a substance called the solute is dissolved in a solvent to form a

solution, its behavior is often altered. Reactions in solution are faster than
in the solid state. The amount of substance that can dissolve in a given
amount of solvent at a stated temperature and pressure is called the
solubility and is determined by the nature of the materials and the laws
governing the solubility equilibrium.

Ions in solution Some substances form ions, which are species possessing a charge. These
behave in a distinct way in solution. They may attract molecules of
solvent, may associate together, and may react with other species to form
complexes or a precipitate.

The pX notation Since concentrations vary over a very wide range, they are often
represented by the logarithmic pX notation where pX = - log(X), where X is
the concentration or activity of an ion, or an equilibrium constant.

Equilibria in The laws of thermodynamics govern the behavior of all species in

solution solution. Every reaction depends upon the thermodynamic properties of
the species involved. Where those properties are changed by the solvent
by association, by reaction or temperature, the behavior will alter.
Physical and chemical equilibria in solution are most important.

Related topics Other topics in Section C Separation techniques (D1 D9)

(C2-C10)

Solvents The use of solvents for analytical work is determined by their properties, as shown
in Table 1.
Solvents with high dielectric constants (er > 10), for example, water and ammonia,
are referred to as polar and are ionizing solvents, promoting the formation and
separation of ions in their solutions, whereas those where er is about 2, such as
diethyl ether, tetrachloromethane and hexane are nonpolar and are nonionizing
solvents. There are also many solvents whose behavior is inter- mediate between
these extremes.
The solution process in a liquid may be represented by a general equation:
A(l) + B = B(sol) solution
solvent solute

The action of solution changes the properties of both solute and solvent. The
solute is made more mobile in solution, and its species may solvate by attraction

pg. 55
Table 1. Properties of some solvents

Solvent Boiling point ( C) Density, Dielectric

(g cm 3) constant, er
Water 100 1.00 78.6
Ammonia 34 0.68 22.0
Ethanol 78 0.79 24.3
n-hexane 69 0.66 1.88
Diethyl ether 34 0.71 4.33
Note: density at 25 C or at BP; dielectric constant relative permittivity

to the solvent. The solvent structure is also disrupted by the presence of species
different in size, shape and polarity from the solvent molecules.
Ideally, the behavior should depend on the concentration m (in molarity, mole
fraction or other units), but often this must be modified and the activity, a, used:

a = m y = p/pn
where g is called the activity coefficient. The vapor pressure of the solution is p,
and that in the standard state is pn. Activities are dimensionless.
Solvents, such as water, with high dielectric constants (or relative permittivi-
ties) reduce the force F between ions of charges z1e and z2e a distance r apart:
F = z1z2 e2/eoer r2
where eo is the permittivity of free space. Also, they will solvate ions more
strongly and thus assist ion formation and separation.
Hexane, diethyl ether and tetrachloromethane (CCl 4) all have low dielectric
constants and are nonpolar. They are very poor at ionizing solutes. However,
they are very good solvents for nonpolar substances.

Solubility The equilibrium amount of solute which will dissolve in a given amount of solvent
at a given temperature and pressure is called the solubility. The solu- bility
may be quoted in any units of concentration, for example, mol m 3, molarity,
mole fraction, mass per unit volume or parts per million (ppm).
There is a general ‘rule of thumb’ that ‘like dissolves like’. For example, a
nonpolar hydrocarbon solvent such as hexane would be a very good solvent for
solid hydrocarbons such as dodecane or naphthalene. An ester would be a good
solvent for esters, and water or other polar solvents are appropriate for polar
and ionic compounds.
● Gases dissolve in solvents according to Henry’s Law, provided they do not
react with the solvent:
pB = x B K
where xB is the mole fraction of solute gas B which dissolves at a partial pres- sure
pB of B, and K is a constant at a given temperature. This is analytically important
for several reasons. For example, nitrogen is bubbled through solutions to
decrease the partial pressure of oxygen in electrochemical experi- ments.
Similarly, air is removed from liquid chromatography solvents by
passing helium through them, or by boiling them, since gas solubility
decreases as the temperature is increased.
● Liquids. When different liquids are mixed, many types of behavior may
occur. If the molecules in the liquids are of similar size, shape, polarity and
chemical nature they may mix in all proportions. For example, benzene and
methylbenzene (toluene) mix completely. In such ideal solutions, obeying
Raoult’s law, the activity coefficient is close to 1:

a = p/pn = x

If the component molecules differ gratly in polarity, size or chemical nature (e.g.,
water and tetrachloromethane) they may not mix at all. This is an important
condition for solvent extraction (Topic D1). The distribution of a solute between
a pair of immiscible liquids depends primarily on the solubility of the solute in
each liquid.
● Solids generally follow the ‘like dissolves like’ rule. Nonpolar, covalent
materials dissolve best in nonpolar solvents. Solid triglycerides such as
tristearin are extracted by diethyl ether, but are nearly insoluble in water.
Salts, such as sodium chloride are highly soluble in water, but virtually
insoluble in ether.

Ions in solution The behavior of ions in solution may be summarized as follows.

(i) Solids whose structure consists of ions held together by electrostatic forces (e.g.
NaCl) must be separated into discrete ions when they dissolve. These ions
often gain stability by solvation with molecules of the solvent. Such solutions
are described as strong electrolytes.
(ii) Some covalent molecules, such as ethanoic acid, may form ions by the
unequal breaking of a covalent bond, followed by stabilization by solvation.
This occurs only partially and these are called weak electrolytes.

H : OCOCH3 [ H + OCOCH3

(iii) In some cases ions do not separate completely. In concentrated solutions,

oppositely charged ions may exist as ion-pairs, large ions of surfactants may
aggregate into micelles, which are used in capillary electrophoresis (Topics
D8 and D9), and the dissociation of covalent molecules may be only partial.
(iv) At ‘infinite dilution’, that is, as the concentration approaches zero, ions are truly
separate. However, even at quite low concentrations, the attractions between
ions of opposite charge will cause each ion to become surrounded by an
irregular cloud, or ionic atmosphere. As the solution becomes even more
concentrated, the ionic atmosphere becomes more compact around each ion
and alters its behavior greatly.
In very dilute solutions, the effects of the ionic atmosphere may be approxi- mated
by using the Debye-Hückel theory, which predicts that the mean ionic activity
coefficient, g± is, for an electrolyte with positive ions with charge z , negative
ions with charge z , is given by:

log(g±) = -A (z . z ) |(I)

where I is the ionic strength = 1⁄2 £ (ci zi2) for all the ions in the solution.
For more concentrated ionic solutions, above 0.1 M, no general theory exists,

pg. 57
but additional terms are often added to the Debye-Hückel equation to compen-
sate for the change in the activity.

The pX notation The concentration of species in solution may range from very small to large. For example
in a saturated aqueous solution of silver chloride, the concentration of silver ions is
about 10-5 M, while for concentrated hydrochloric acid the concentration of
hydrogen and chloride ions is about 10 M. For convenience, a logarithmic scale
is often used:
pX = -log (X)
where X is the concentration of the species, or a related quantity. Thus, for the
examples above, pAg = 5 in saturated aqueous silver chloride and pH = -1 in
concentrated HCl.
Since equilibrium constants are derived from activities or concentrations as noted
below, this notation is also used for them:
pK = -log (K)

Equilibria in Most reactions will eventually reach equilibrium. That is, the concentrations of
solution reactants and products change no further, since the rates of the forward and reverse
reactions are the same.
From the above arguments concerning solutions, and from the laws of
thermodynamics, any equilibrium in solution involving species D, F, U and V:
D+F[U+V
will have an equilibrium constant , KT, at a particular temperature T given by:
KT = (aU . aV)/(aD . aF)
where the activities are the values at equilibrium. It should be noted that KT changes
with temperature. The larger the equilibrium constant, the greater will be the
ratio of products to reactants at equilibrium.

There are many types of equilibria that occur in solution, but for the impor-
tant analytical conditions of ionic equilibria in aqueous solution, four examples
are very important.
(i) Acid and base dissociation. In aqueous solution, strong electrolytes (e.g., NaCl,
HNO3, NaOH) exist in their ionic forms all the time. However, weak electrolytes
exhibit dissociation equilibria. For ethanoic acid, for example:
HOOCCH3 + H2O [ H 3O +CH3COO
Ka = (aH . aA)/(aHA . aW) +1.75 ¥ 10 5

where HA, W, H and A represent each of the species in the above

equilibrium. In dilute solutions the activity of the water aW is close to 1.
For ammonia:
NH3 +H2O [ NH4 +OH
Kb = (aNH4 . aOH )/(aNH3 . aW) = 1.76 x 10 5
Water behaves in a similar way:
2 H 2O [ H3O + OH
KW =(aH3O . aOH ) = 10 14
(ii) Complexation equilibria. The reaction between an acceptor metal ion M and
a ligand L to form a complex ML is characterized by an equilibrium constant.
This is discussed further in Topic C7, but a simple example will suffice here:
M(aq) L(aq) [ ML(aq)
Kf (aML)/(aM . aL)
For example, for the copper EDTA complex at 25oC: Kf 6.3 ¥ 1018
(iii) Solubility equilibria. If a compound is practically insoluble in water, thisis
useful analytically because it provides a means of separating this compound
from others that are soluble. The technique of gravimetric analysis has been
developed to give very accurate analyses of materials by weighing pure
precipitates of insoluble compounds to give quantitative measurements of their
concentration. For the quantitative determination of sulfate ions, SO 2 , the
solution may be4 treated with a solution of a soluble barium salt such as
barium chloride BaCl2, when the following reaction occurs:
Ba2 SO 2
4 [ BaSO 4(s)
Conversely, if solid barium sulfate is put into water:
BaSO4 (s) Ba2 SO 42
The solubility product, Ksp, is an equilibrium constant for this reaction
Ksp a(Ba2 ) . a(SO42 ) 1.2 ¥ 10 10
bearing in mind that the pure, solid BaSO4 has a 1. This means that a solution
of barium sulfate in pure water has a concentration of sulfate ions of only 1.1
¥ 10 5 M. The concentration of the barium ions is the same.
(iv) Redox equilibria. When a species gains electrons during a reaction, it
undergoes reduction and, conversely, when a species loses electrons it
undergoes oxidation. In the total reaction, these processes occur simultane-
ously, for example:
Ce4 Fe2 Ce3 Fe3
The cerium is reduced from oxidation state 4 to 3, while the iron is oxidized
from 2 to 3. Any general ‘redox process’ may be written:
Ox1 Red2 Red1 Ox2
The equilibrium constant of redox reactions is generally expressed in terms
of the appropriate electrode potentials (Topics C5, C8), but for the above
reaction:
K (a(Ce3 ) . a(Fe3 ))/(a(Ce4 ) . a(Fe2 )) 2.2 ¥ 1012

Summary
For ionic equilibria in solution, which are widely used in analytical chemistry,
a large equilibrium constant for the reaction indicates that it will proceed
practically to completion. If the equilibrium constant is of the order of 1010,
then the ratio of products to reactants will be much greater than 1000 to 1. For
example:

pg. 59
H OH H2 O K =1014
C(aq) + A(aq) +CA(solid) K = 1010
M(aq) + L(aq) = ML(complex) K =1010

Ox1 + Red2 =Red1 +Ox2 K = 1012

Therefore, these reactions may be used for quantitative measurements, for
example by volumetric or gravimetric techniques (Topics C5, C7 and C8).
It should be noted that, in calculations involving solution equilibria, certain
rules should always be considered.
● Electroneutrality. The concentrations of positive and negative charges must
be equal. Sometimes, ions that do not react are omitted from the equations,
although they must be present in the solution.
● Stoichiometry. The total amounts of all species containing an element must
be constant, since no element can be created or destroyed.
● Equilibria. All possible equilibria, including those involving the solvent,must
be taken into account.
E LECTROCHEMICAL
REACTIONS

Electrochemical
reactions

Electrochemical cells

Electrode potentials The potential difference between electrodes depends on the properties of
the electrodes and the concentrations in the solution.

Solution equilibria (C1)

Electrochemical Important electrochemical reactions have already been noted in Topic C1, for
reactions example:
● acid base reactions, where the acid donates a proton to the base;
● precipitation reactions, where the reactants form an insoluble product;
● complexation reactions, where a ligand coordinates to an acceptor;
● oxidation reduction reactions, where the oxidizing agent gains electrons
from the reducing species.
All of these reactions involve charged species and all may be studied by electro-
chemical methods and used for analysis.

Electrochemical In order to study an electrochemical reaction, the appropriate cell must be

cells constructed. It is impossible to measure the electrical properties using a single
contact. For example, connecting to one end of a resistor will not allow measure-
ment of its resistance. Connections must be made to two electrodes, and a cell must
be constructed. Electrical connection to the solution, whether to measure a cell emf
or to conduct electrolysis reactions, must be made through two electrodes.
Cells with two similar, inert electrodes placed in the same solution may be used
for measuring conductance, discussed in Topic C10. Cells where electro- lytic
reactions occur are used in voltammetry, which is discussed in Topic C9. For
potentiometric methods, discussed here and in other parts of this section, two
dissimilar electrodes are used to construct a cell whose emf depends on the solutions
and electrodes used.
Many different types of electrodes are available and the most important are
described in Topic C3. The simplest are made of a metal, such as zinc, copper, or

pg. 61
platinum. Consider a cell set up with a zinc electrode dipping into a zinc sulfate
solution, and connected through a porous disc to a copper sulfate solution in
contact with a copper electrode, as shown in Figure 1.
The cell is conventionally written as:
Zn Zn2 SO4 2 Cu2 SO 42 Cu
The cell reaction will result in simultaneous reduction or addition of electrons to
a species in solution at the one electrode and oxidation, or removal of electrons
at the other.
Cu2 +2e = Cu reduction (note: ‘reduction on the right’)
Zn - 2e = Zn2 oxidation (note: ‘oxidation on the left’)
Adding these gives the cell reaction (excluding the sulfate anions).
Cu2 + Zn =Cu + Zn2
If the cell reaction is allowed to take place, by connecting the copper and zinc
electrodes electrically, then it acts as a galvanic cell or battery, and does work. If
the potential difference is measured with no current flow, it is defined as the
electromotive force, or emf, E, and this is equivalent to reversible conditions. If
an external voltage is applied to the cell it can cause electrolysis, driving the cell
reaction.
In order to prevent the zinc sulfate and the copper sulfate mixing, a porous
disc, or a conducting salt bridge of potassium sulfate in a tube, or a wick can be used
to connect the solutions. This helps to reduce the liquid junction potential, which
occurs when two solutions of unequal concentration, or containing dissimilar ions
are placed in contact. Because of the different rates of diffusion of the ions, a liquid
junction (or diffusion) potential, EL, is set up and this affects the total cell emf.
For example, between a solution of 0.1 M HCl and a solution of 0.01 M HCl, EL is
about 38 mV.
By using a third solution, which must not react, and in which the cation and
anion have similar mobilities and diffusion characteristics, to connect the two
electrode solutions the liquid junction potential may be minimized. Concentrated
potassium chloride, about 4 M, is often used, since K and Cl ions have almost the
same mobility (see Topic C10). Potassium nitrate or ammonium nitrate is also
suitable if chloride ions react. Use of a salt bridge reduces the effective EL to about
1 mV.
The laws of thermodynamics, which are described in detail in textbooks of
physical chemistry, provide the connection between the chemical reaction and
the electrochemical quantities that are measured in analysis.

Meter

Zinc electrode Copper electrode

Zinc sulfate solution Copper sulfate solution

Porous plate

Fig. 1. Daniell cell.

The free energy G depends on the activity (or approximately the concentra-
tion) of the reactants and products. For a general reaction:
Ox1 + Red2 = Red1 + Ox2
G = Gn + RT ln [(a(Ox2).a(Red1))/(a(Ox1).a(Red2))]
where the superscript, n, means that it applies to the standard state. For gases,
this is 1 atmosphere pressure; for liquids and solids it is less than 1 atmosphere
pressure for the pure material; and for solutions, it is an activity of 1.
The electromotive force measures the free energy change under reversible
conditions when n Faradays of electrical charge occur in the cell reaction and is
related to it by the equation:
G = -nFE
For example, in the Daniell cell reaction, 2 moles of electrons are transferred,and
the cell emf is about 1.1 V. Therefore:
G = -2 x 96485 x 1.1 = --212.3 kJ mol 1
Combining the above equations, for the Daniell cell:
E = En + (RT/2F) ln [(a(Zn2 ).a(Cu))/(a(Zn).a(Cu2 ))]
This is often referred to as the Nernst equation for the cell. (Note: the value of
En measures the cell emf and relates to the free energy change under standard
state conditions, while the value of E relates to the free energy change under
other, specified conditions. It is important to recognize that these equations
apply at any temperature, not just a standard temperature.)

Electrode In order to establish any scale it is necessary to have a reference point. Since
potentials measurement of emf is made using a cell, one electrode must be taken as a reference
electrode. The standard hydrogen electrode (SHE) has been chosen as the primary
reference.
A hydrogen electrode may be constructed using a platinum metal plate as
contact, and bubbling hydrogen gas through so that it makes contact with both
solution and platinum. When the gas is at 1 atmosphere pressure and the solu-
tion has an activity of hydrogen ions of 1, this is a standard hydrogen electrode
(SHE).
By convention, the standard hydrogen electrode is assigned an electrode
potential of 0.000 V and all other electrode potentials are compared to it.
A cell can be constructed with a copper electrode in copper sulfate and a standard
hydrogen electrode, as shown in Figure 2.

H2 supply Copper electrode

Pt electrode

Sintered disk
Solution with Copper sulfate solution
H+ (a=1)

Fig. 2. Copper-hydrogen electrochemical cell.

pg. 63
The cell reaction is the reduction of copper ions by hydrogen:
Cu2 + H 2= Cu + 2 H
Writing the Nernst equation for this cell:
E -En - (RT/nF) ln [(a (Cu).a (H )2)/(a (Cu2 ).a (H2))]
The superscript, n, means the standard value, when the activities are all 1. Note that
the quantity (RT/F) has the value 0.02569 V at 25 C. Here, n =2, since two electrons
are transferred at each electrode. For a standard hydrogen electrode, a (H ) 1
and a (H2) = p(H2) = 1, so, we may write:
E = En - (RT/2F) ln [a (Cu)/(a (Cu2 )]
which is the equation for the copper electrode potential. It may also be written:
E = En + (RT/2F) ln [a (Cu2 )/(a (Cu)]
and for a pure copper electrode, a(Cu) 1, so at 25 C:
E = En +0.01284 ln [a (Cu 2 )]
This means that the electrode potential of a copper electrode depends logarith-
mically on the activity of the copper ions in solution. This is known as the
Nernst equation. Standard electrode potentials are listed in Table 1 of Topic C3.
Although the SHE is essential to define electrode potentials, it is not suitable
for many routine analytical uses, and the saturated calomel electrode (SCE) is
often used in practical cells. This is described in Topic C3.
The Nernst equation is very useful in analytical measurements, since it allows
the analyst to measure variations in concentration during reactions or experi-
ments. It is important to note that, strictly, electrochemical cells measure the
activity, although in many cases this may be related to the concentration.

Electrolysis If an electric current is passed through a cell consisting of two electrodes in

contact with an electrolyte solution, the current causes the ions to migrate
towards the electrode of opposite charge, where a reaction takes place. For
example, when acidified water is electrolyzed:
(i) the hydrogen ions are attracted to the negative electrode (cathode) and gain
electrons to form hydrogen gas:
2H +2e = H 2
(ii) the hydroxyl ions are attracted to the positive electrode (anode) and lose
electrons to become water and oxygen gas:
2OH + 2e = H2O + 1⁄2 O2
The total reaction is therefore:
2H2O = 2H + 2OH = H 2O + H2 + 1⁄2 O2
In order to cause 1 mole of reaction involving the transfer of n electrons, then n moles
of electronic charge must be transferred or nF Coulombs cause 1 mole of such a
reaction, where F is the Faraday equal to 96458 C.
Electrolysis is the basis of the analytical techniques of polarography, voltam-
metry, amperometry and coulometry (see Topic C9). Electrolysis may also be
used for the deposition, production and purification of materials. For example,
the copper electrolysis cell has two electrodes placed in an acidified aqueous
solution of a copper salt.

Example
If a current of 0.123 amps is passed through a copper electrolysis cell for one
hour, how much copper will deposit?
0.123 A = 0.123 C s 1

Since 1 hour = 3600 s, therefore, It 442.8 C are passed.

1 mole of copper is 63.5 g, which requires 2F = 2 x 96485 C, so 442.8 C deposit
63.5 x 442.8/(2 x 96485) g = 0.1457 g.
It is possible to deposit metals onto an electrode quantitatively, using the
technique of electrogravimetry, although this is not often used. Electrolysis may
also be used to generate reagents to react with analytes, and this is referred to as
coulometry.

pg. 65
P OTENTIOMETRY

Key Notes

Cells In order to make measurements of electrode potentials, or to study the

changes that take place in a solution reaction, an appropriate
electrochemical cell must be set up.

Indicator electrodes The indicator electrode makes electrical contact with the solution and acts
as a sensor, which responds to the activity of particular ions in solution
and acquires a potential dependent on the concentration of those ions.

Selectivity The ideal electrode should respond to a single ion, but this is not often
the case. The effectiveness of any indicator electrode is determined by its
selectivity.

Direct potentiometry The direct measurement of concentrations is possible using electrodes of

high selectivity and reproducibility. The measurement may also be used
to follow titrations.

Related topics pH and its control (C4) Titrimetry II: complexation,

Titrimetry I: acid base titrations precipitation and redox titrations
(C5) (C7)

Cells When a cell is set up, but not connected to any outside circuit, no reaction should
take place. If the cell is now ‘shorted out’ by connecting the electrodes on right and
left, electrons will flow and the cell reaction will occur. This reaction changes the
concentrations of the original solutions. Therefore, to measure the original sample
concentrations within the cell, a device must be used that does not allow current,
and hence electrons, to flow. While older systems used a potentiometer, where the
potential difference was balanced by adjusting an elec- trical circuit so that an
external source gave exactly the same potential difference detected by the null-point
of a galvonometer, modern potentiometry uses digital voltmeters (DVM), where
the current used to take the measurement is extremely small. A suitable
experimental arrangement is shown in Figure 1.
If a check is needed on the correctness of the measured value for an experi- mental
cell, a standard cell, such as the Weston cadmium cell, may be used as a calibration,
since the value of its emf is accurately known over a range of temperatures. The
electrode potential is defined using the standard hydrogen electrode as reference,
as described in Topic C2.

Indicator There are many types of indicator electrode used in analyses to construct electro-
electrodes chemical cells. They may be classified as shown in Table 1.
When two electrodes are combined in a cell, the measured emf may be sepa-
rated into ‘half-cell emfs’ that relate to the individual electrodes. For a Daniell
cell discussed in Topic C2:
Digital voltmeter

Electrode A Electrode B

Salt bridge Test solution

Fig. 1. Experimental arrangement with salt bridge junction and DVM.

Table 1. Potentiometric indicator electrodes

Class Description Example
Class 1 Metal/metal ion Ag/Ag (cation reversible)
Class 2 Metal/saturated metal salt/anion Ag/AgCl/Cl (anion reversible
Redox Inert metal/redox couple Pt/Ce4 , Ce3
Pt/H , H2
Membrane Inner electrode/solution/ Glass electrode
ion selective membrane Fluoride electrode
ISFET Coated field-effect transistor pH-sensitive
Gas-sensing electrodes pH-electrode + membrane For CO2, NH3

E(cell) ,=E(right) - E(left) = E(Cu2 , Cu) - E(Zn2 , Zn) Dividing the

emf equation into separate electrode equations:
E(Cu2 , Cu) = En(Cu2 , Cu) + (RT/2F) ln (a (Cu2 )/(a (Cu))

E( Zn 2 , Zn) =En(Zn 2 , Zn) + (RT/2F) ln (a (Zn2 )/(a (Zn))

If these equations are obeyed, then the electrodes show Nernstian response.

Class 1 electrodes
These are the simplest electrodes, but not necessarily the easiest to use. A metal rod is
immersed in a solution of ions of the same metal, for example silver with a solution
containing silver ions. With some ions it is important to prevent hydrolysis or
complexation taking place.

Ag (solution) + e Ag(solid)

E(Ag /Ag)= En(Ag , Ag) +RT/F ln (a (Ag )/(a (Ag)) if a pure

silver rod is used, a(Ag) 1, so we may write:
E(Ag /Ag) = En(Ag , Ag) + RT/F ln (a (Ag ))

This is therefore an electrode reversible to silver ions. An example of the use of this
electrode is given later. En(Ag , Ag) = 0.800 V at 25 C is the standard electrode potential
of the silver electrode.

Class 2 electrodes
When an insoluble salt of a metal (see also Topic C8) is present, the concentra- tion of the
metal ion depends on the concentration of the anion and on the

pg. 67
solubility product Ksp. Therefore, if we put insoluble silver chloride, AgCl, in
contact with a silver rod,
a (Ag ) = Ksp/(a (Cl ))
so that the concentration of silver ions is controlled by the silver chloride
solubility.
E(AgCl/Ag ) = En(AgCl/Ag) - RT/F ln (a(Cl ))
Since Ksp (AgCl) ~ 1.6 x10 10 at 25 C,
En (AgCl/Ag) = 0.22 V
This electrode is reversible to chloride ions.
Another important electrode of this class is the calomel electrode, which is
discussed below.
In order to make accurate potentiometric measurements, a cell must be
constructed that is reproducible and reliable. The emf should depend chiefly on
the particular species in the sample that it is intended to measure, and if neces-
sary the system must be calibrated.
As noted in Topic C2, electrode potentials are conventionally referred to the
standard hydrogen electrode (SHE), for which the standard electrode potential,
En 0.000V. Another electrode, for example the silver electrode, may be com-
bined with the SHE to form a cell. Table 2 gives a representative list of standard
electrode potentials at 25 C.

Table 2. Standard reduction electrode potentials at 25 C

Electrode reaction En/V

Li+ + e = Li -3.04
K+ + e =K -2.92
Ca2+ + 2e = Ca -2.87
Na+ + e = Na -2.71
Mg2+ + 2e = Mg -2.37
Al3+ + 3e = Al -1.66
Mn2+ + 2e = Mn -1.18
Zn2+ + 2e = Zn -0.76
Fe2+ + 2e = Fe -0.44
Sn2+ + 2e = Sn -0.14
Pb2+ + 2e = Pb -0.13
H+ + e = 1 ⁄2 H 0.0000 exactly
AgBr + e = Ag + Br +0.10
Sn4+ + 2e = Sn2 +0.15
AgCl + e = Ag Cl +0.22
Cu2+ + 2e = Cu +0.34
Hg2Cl2 + 2e = 2Hg + 2Cl +0.27
1
⁄2 I2 +e =I +0.54
Fe3+ + e = Fe2 +0.76
Ag+- + e = Ag +0.80
IO 6H 5e = 1⁄2 I 3H O +1.19
3 2 2
⁄2 O2
1
2H + 2e = H 23O +1.23
⁄2 Cr O
1 2-
7H + 3e = Cr 7/2 H O +1.33
2 7 2

⁄2 Cl 2
1
+ e = Cl +1.36
Ce4 + e = Ce3 +1.44
MnO4- +8H + 5e = Mn 2 + 4H2O +1.52
The SHE is rather inconvenient to use, since it requires a supply of inflam- mable
hydrogen and has a tendency to change emf slightly as the bubbles of hydrogen cover the
metal.
Reference electrodes should have a constant potential, should not react with the sample,
and should have a very small liquid junction potential with the sample solution. Two
reference electrodes are commonly used.
The calomel reference electrode is shown in Figure 2(a). This is an electrode of Class 2, with
liquid mercury in contact with mercury(I) chloride, or calomel, in a solution of potassium
chloride of high, fixed concentration.
The electrode reaction is
Hg2Cl2 (solid) + 2e =2Hg (liquid) +2Cl- (aq) The
electrode potential is given by:
E(cal) = En(cal) - RT/2F ln (a(Cl-)2)

If a concentrated solution of KCl is used, either saturated or 3.5 M, then this elec- trode has a
constant potential at 25 C. Changes in the solution outside the electrode have a very small effect
on the potential of this electrode, since the chloride con- centration is high and is not altered
significantly by the external solution. Additionally, the concentrated KCl acts as a salt bridge.

E(cal, sat) = 0.244 V (saturated calomel electrode, SCE)

E(cal, 3.5M) = 0.250 V (calomel electrode, 3.5M KCl)

The silver/silver chloride reference is similar, having a silver wire coated with silver
chloride and in contact with concentrated KCl solution.

E (AgCl, 3.5M) 0.204 V

(a) (c)

Contact Contact Contact

Fill hole

Hg/Hg2Cl2 AgCl/Ag
paste electrode

KCI solution

Buffer solution
Porous plug Selective surface
Glass membrane

Fig. 2. (a) Calomel reference electrode; (b) glass electrode; (c) solid-state electrode.

pg. 69
If the sample solution might react with chloride ions, for example silver or lead
salts, then a double junction reference electrode may be used, with an
additional liquid junction of KNO3.

Redox electrodes
If an inert wire, usually platinum, is placed into a solution containing both
oxidized and reduced species, it will take up an equilibrium potential depen-
dent on the ratio of their concentrations:
Ox + ne- = Red
E(Ox, Red) = En(Ox, Red) + (RT/nF) ln (a(Ox)/a (Red))

Ion-selective electrodes (ISE)

All of the above electrodes are sensitive to particular ions, but have certain
disadvantages, such as loss of response due to poisoning, or a tendency to
mechanical or other failure.
A large variety of special electrodes have been developed to test for a verywide
range of ions. Many involve an insoluble membrane, capable of electrical conduction,
whose surface potential depends chiefly on the concentration of a particular ion.
The glass electrode for measuring pH, shown in Figure 2(b) is an early
example. A thin membrane of a conducting glass separates the inner solution of
a silver–silver chloride electrode in a chloride-containing solution from the
sample solution outside. Both sides of the glass membrane become hydrated,
and the outer surface exchanges its cations with the hydrogen ions of the sample
solution. The potential of the glass electrode depends on the ratio of the activi-
ties of H ions inside and out, on the potential of the inner silver halide electrode
and on the asymmetry potential, characteristic of each particular electrode.
This gives a total electrode potential:
E(glass) = E*(glass) +RT/F ln (a(H ))
= E*(glass) -(2.303 RT/F) pH
The unknown constant E* must be determined by calibration with known buffers,
as described in Topic C4. Modern glass electrodes for pH measurement often
incorporate a reference electrode.
Although these electrodes work very well in the pH range 1 9, at high pHthey
suffer from an alkaline error due to the effect of other cations. This is discussed in
the next section. Glass electrodes are also fragile and may suffer from slow
response times.
Crystalline membrane electrodes, constructed either as the glass electrode, or
with a direct contact as shown in Figure 2(c), have an outer crystal surface which
responds to particular ions. The fluoride electrode has a crystal of LaF 3, treated
with Eu(II) to increase conductivity, which responds selectively to the
adsorption of free F ion on its surface. The selectivity is very good, due to the
ability of the small F on to fit to the LaF 3 crystal lattice to the exclusion of larger
ions. However, OH ions are also small and can interfere, and F may also form
undissociated hydrofluoric acid. Therefore, it is necessary to use this electrode
in a buffer solution at about pH 6.
Interferences occur with metals such as Al3 , Fe3 and Ca2 , which combine
with the fluoride. With a SCE reference electrode, the cell emf is:
E(cell) = E*(F ) -(RT/F) ln (a (F )) - E(SCE)
or, at 25oC:
E(cell) = k - 0.0257 ln (a(F ))

If a cell is constructed with a fluoride ISE combined with a SCE, calibration with solutions of
fluoride ions of known concentration enable the value of k to be found. When the same cell is
filled with an unknown solution, the fluoride ion concentration may be determined. This is an
example of direct potentiometry.
Other crystalline ion-selective electrodes use AgCl for Cl , Ag2S for Ag and S2 , Ag2S +
CuS for Cu2 and many more.
Liquid membrane electrodes are shown in Figure 3(a). Typically an
organophilic porous plastic membrane separates the internal reference from the sample
solution. Another reservoir contains a liquid ion-exchanger, which satu-rates the pores in
the membrane. The calcium electrode uses calcium dodecyl phosphate [(RO)2 PO2]2 Ca
where R is the dodecyl radical, dissolved in dioctyl phenylphosphonate.
The electrode response is given by:

E(Ca2 ) = k + (RT/2F) ln (a(Ca 2 ))

Ion selective field effect transistors (ISFETs)

One disadvantage of the glass electrode for measuring pH is its fragility. A modern
development uses a field effect transistor where the gate surface is coated with a material,
often aluminum oxide, whose surface is sensitive to pH. As the potential at the surface
changes with pH, the current induced through the transistor varies. A temperature
diode, incorporated in the electrode

Leads to meter
(a) (b)

Ag electrode

Reference
Liquid organic electrode
ion exchanger
Internal
solution
Aqueous solution Glass electrode
saturated with
AgCl and MCl2 O-ring

Porous plastic Gas-permeable

membrane with membrane
liquid ion exchanger

Fig. 3. (a) Liquid ion exchange membrane electrode for M2 ions. (b) Gas sensing electrode
using glass ISE.

pg. 71
simultaneously measures the temperature so that a rapid reading of pH,
compensated for temperature is obtained. The electrodes are robust and the
response time is a few seconds. An example is given in Topic H1.

Gas-sensing electrodes
If gases such as CO2 or NH3 are allowed to diffuse into the solution surrounding
a pH electrode, they will alter the pH. The construction is shown in Figure 3(b).
The pH electrode, often incorporating a reference electrode as well, is separated
from the sample solution by a microporous hydrophobic membrane, which
will allow gases but not water to diffuse through rapidly. For CO 2 the overall
equilibrium occurs in 3 stages:
(i) carbon dioxide gas diffuses from the outer solution through the membrane
until inner and outer solutions are at the same concentration;
(ii) the solution of CO2 forms the acid H2CO3, which dissociates to form
hydrogen ions:
CO2 +2H2O =H2CO3+ H2O = H3O +HCO 3

K a(H3O ). a(HCO 3 )/(p(CO 2(sample)))

(iii) if the internal solution within the membrane has a constant activity of
HCO 3 , for example sodium hydrogen carbonate, then the pH may be
calculated:
pH = -log (H3O ) = log (K. p (CO2 (sample)))
so that by measuring the pH we can find the concentration of CO2.
Similar arguments apply with an ammonia-sensing electrode.

Selectivity The ideal electrode should respond only to changes in concentration of a single
ion i. This is rarely true and the response of a real electrode system can be described
by the Nikolsky Eisenmann equation:
Ei = En +i S log[a(1) +∑K (a(2))
pot1,2
(z1/z2)]

where S is the slope of the emf -log(a) plot, which for Nernstian behavior should
be 0.0592/z(1) at 25 C, a(1) is the activity of the ion selected of charge z(1), a(2)
is the activity of the interfering ion of charge z(2) and Kpot1,2 is the selec- tivity
coefficient for ion 1 with respect to ion 2. The smaller the value of this selectivity
coefficient, the smaller the interference by the second ion.

Direct By calibration, and the use of appropriate electrode response equations, it is

potentiometry possible to measure concentrations directly, as indicated in the example above for
fluoride ion in tap water.
A calibration plot should indicate three things:
(i) whether the electrode responds to the correct ion;
(ii) whether that response is Nernstian;
(iii) what range of concentrations may be studied.

Figure 4 shows the calibration plot for a copper ion selective electrode, where a total
ionic strength adjustment buffer (TISAB), such as 1M NaNO 3, has been added to
each solution so that the response is effectively at constant ionic strength and
constant activity coefficient.
150

100

Electrode potential (mV)

50

–50

–100
–7 –6 –5 –4 –3 –2 –1 0
log (a)

Fig. 4. Electrode potential response for a copper ISE. (a) Against activity (solid line); (b)
against concentration (dashed line).

This graph shows that over the concentration range 10-1 to 10-5 M the calibra-
tion is linear with a slope of 0.029 V/log (a(Cu 2 )). Below about 10-6 M the line
curves, since the solubility of the crystal becomes significant. Interferences from
Cu(I), Ag(I) and Hg(II) are troublesome.
If a calibration curve is constructed, direct measurement of solution concen-
trations within that range may be made.

Example
Calibration performed with a ISE selective to Mg2 ions gave a value of S =
0.0296, and E = 0.411 V for a solution of a(Mg 2 ) = 1.77 x 10 3 M.
What is the activity of Mg2 ions when E =0.439 V?
Substitution gives a(Mg2 ) = 1.998 x 10-4 M.

The method of standard additions (Topic B4) has the advantage that comparison
is made in the same matrix solution, so that interferences and other effects are
the same for the unknown and for the standard.
Potentiometric titrations are discussed in Topic C5.

pg. 73
Chapter five : pH AND ITS CONTROL

Key Notes
Definition of pH Since the concentrations of ions in aqueous solution vary over an
enormous range from greater than 1 molar down to 10--14 molar or less, it
is convenient to use a logarithmic function to describe them. For
hydrogen ions in aqueous solution, denoted by H3O , often called the
hydronium ion, this is defined as:
pH = -log(a(H3O )) ≈ log(c( H3O )/mol dm-3)

The pH scale Pure water contains equal amounts hydrogen ions and hydroxyl ions
OH , at a concentration of 10 7 molar. Therefore, the pH of neutral water
is -log (10-7) = 7. Acids have pH values less than 7, down to less than 1,
while for alkalis, the pH is greater than 7, up to 13 or more.

Buffers For many analytical measurements it is necessary to control the pH so

that it varies as little as possible. Solutions that resist changes in pH are
known as buffer solutions. They usually consist of a mixture of a weak
acid and its salt for lower pH values, or a weak base and its salt for
higher pH values.

pH measurement pH may be estimated approximately using visual indicators, but more

accurate determination requires the use of pH meters. These must be
calibrated using standard buffers.

pH control This is most usually achieved with buffers. Enzyme reactions,

electrophoretic separations and also spectrometric and electrochemical
determinations may all require pH control.

Related topics Titrimetry I: acid-base titrations (C5)

Definition of pH The acidity or alkalinity of a reaction mixture is most important. It can control the
rate of reaction, the nature of the species present and even subjective proper- ties
such as taste. The original definition of pH (Sorensen, 1909) related it to the
concentration of hydrogen ions. Two facts should be recognized. First, like many
ions in solution, the hydrogen ion does not exist in aqueous solutions as a ‘bare’ H
species, but as a variety of hydrated ions, such as H3O . Second, the
determination of pH is often carried out by methods that measure the activity of the
hydrogen ions, a(H3O )
a(H3O ) = c(H3O )g± or pH = -log[a(H3O )]
where c(H3O ) is the molar concentration of hydrogen ions, and g± is the mean
ionic activity coefficient of the ions in solution (see Topic C1).
At low concentrations (<10 2 molar), g± is close to 1, and the difference
between concentration and activity is small for uni-univalent electrolytes.
The practical (or operational) definition of pH recognizes that it is deter- mined
using electrochemical cells having an electrode selective to hydrogen ions. This
has been discussed in Topics C2 and C3, but a typical cell is:
Reference electrode Solution X glass electrode.
This gives an output e.m.f , EX,
EX =E* + (RT/F) ln [a(H3O )X]
The constant E* depends on the exact nature of the reference and glass elec- trodes,
and is best eliminated by calibration with a standard solution S whichhas a pH
that is accurately known.
ES =E* + (RT/F) ln [a(H3O )s]
Subtracting these and converting the logarithms gives a practical definition of
pH:
pH(X) = pH(S) + (ES-E X)/(RT ln(10)/F)
Typical calibration buffers are discussed below.

The pH scale In all aqueous solutions, pH values may range between about 0 and 14 or more
as shown in Figure 1. Molar solutions of strong mineral acids, such as HCl,
HNO3 or H2SO4 have pH values less than 1. Weak acids, such as ethanoic or
citric acid in decimolar solution have a pH of around 3.
A useful standard is 0.05 M potassium hydrogen phthalate which, at 15C has a
pH of 4.00. Although pure water is neutral and has a pH of 7.00, freshly distilled
water rapidly absorbs carbon dioxide from the air to form a very dilutesolution
of carbonic acid, and therefore has a pH of around 6.

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14

Very acid Acidic Neutral Alkaline Very alkaline

Fig. 1. The pH scale.

Another standard occasionally used is 0.05 M borax (sodium tetraborate,

Na2B4O7), which has a pH of 9.18 at 25 C.
Dilute alkalis such as ammonia or calcium hydroxide (lime water) have pH
values near to 12, and for molar caustic alkalis, such as NaOH, the pH is over 13.

Buffers As many reactions depend greatly upon the concentration of hydrogen ions inthe
solutions being used, it is important to control the pH. This is usually achieved by
using a solution which has a pH that is accurately known and that resists any change
in pH as solvent for the experiment. Such solutions are called buffers.
The equilibria that govern the reactions of weak acids or bases in aqueous
solution will resist attempts to change them. This is known as Le Chatelier’s
principle. For example, the dissociation of ethanoic acid obeys the equation:

pg. 75
CH3COOH [ CH3COO- + H3O
and an equilibrium constant is written (in terms of concentrations)
Ka = c(Ac-) xc(H3O )/c(HAc) = 1.75 x10-5
using the abbreviation Ac for the CH3COO group. Converting to logarithmic
form, and recalling that pK = -log(K):
pH = pKa + log [c(salt)/c(acid)]
This is the Henderson-Hasselbalch equation.
If we make a mixture containing both the free acid, HAc, and its salt sodium
ethanoate, NaAc, then the equilbrium and the concentrations of acid and salt will
determine the concentration of hydrogen ions and the pH.

Example 1
For a mixture of 50 cm 3 of 0.1 M HAc with 40 cm 3 of 0.1M NaAc, giving a total
volume of 90 cm3,
c(H3O ) = 1.75 x10-5 x [(50 x 0.1/90)/(40 x 0.1/90)]
= 2.19 x 10-5 M, so that pH
=4.66
Addition of acid to this buffer shifts the above equilibrium to the left and most of
the added hydrogen ion combines with the anion. Adding 10 cm3 of 0.1 M HCl
lowers the pH only to about 4.45. If this amount of acid were added to 90
cm3 of water, the pH would be 2.0. Similarly, when alkali is added, the hydroxyl
ions react with the acid to produce more salt. 10 cm 3 of 0.1 M NaOH will raise
the pH only to around 4.85. If this amount of alkali were added to 90 cm3 of
water, the pH would rise to 12.

Weak bases and their salts behave in much the same way. For example,
ammonia and ammonium chloride:
NH3 + H2O NH4 + OH +
Kb = c(OH+) x c(NH )/c(NH
4 )3= 1.75 x10-5 or,
rewriting the Henderson-Hasselbalch equation:
pOH = pKb + log[c (salt)/c(base)]or,
since pH +pOH =14.0
pH = 14.0 - pKb - log[c(salt)/c(base)]
For a mixture of equal amounts of 0.1 M ammonia and 0.1 M ammonium
chloride
pH = 14.0 - 4.75 =9.25
A most useful range of buffers is obtained by using salts of a dibasic (or
tribasic) acid such as phosphoric acid, H3PO4 - for example, potassium
dihydrogen phosphate, KH2PO4, and disodium hydrogen phosphate, Na 2HPO4.
The equilibrium involved here is:
H2PO4- H3O HPO42-
For this equilibrium, the second dissociation constant of phosphoric acid, Ka2, is
close to 1 x 10-7, or pKa2 = 7. Figure 2 shows the effect of adding acid or alkali on
the pH of a mixture containing 50 cm3 each of 0.1 M KH2PO4 and 0.1 M Na2HPO4,
which originally has a pH of 7.0. Adding 0.001 moles of acid to 100 cm3 of
water would lower the pH to 2.
The more concentrated the buffer, the greater will be its buffer capacity. This
is the amount of acid (or alkali) that, when added to 1 liter of buffer, will change
its pH by 1 unit. In the above example, the buffer capacity is about 0.04 moles. If
we had used a more concentrated buffer, the capacity would be greater. Very dilute
buffers have little buffer capacity, and hence have limited use.
Table 1 gives a selection of buffers and standard solutions that are useful for
pH control. These solutions and others are often used to calibrate pH meters.

pH measurement Two important methods exist for pH measurement: visual, using indicators,
and potentiometric, by means of electrochemical cells.
Indicators for pH measurement are weak acids (or bases) where the color of
the acid form is different from that of the salt.

10

8
pH

4
–4 –3 –2 –4 0 1 2 3 4
Millimoles of acid (+) or alkali (–)

Fig. 2. The effect of adding acid ( ) or alkali ( ) to a phosphate buffer mixture, originally at
pH 7.

Table 1. Buffer solutions

Solution pH at 25 C
0.05 M potassium hydrogen phthalate 4.008
0.1 M ethanoic acid, 0.1 M sodium ethanoate 4.640
0.025 M potassium dihydrogen phosphate, 6.865
0.025 M disodium hydrogen phosphate
0.01 M borax 9.180
0.1 M ammonia, 0.1 M ammonium chloride 9.250
0.025 M sodium bicarbonate, 10.012
0.025 M sodium carbonate
Saturated calcium hydroxide 12.454

pg. 77
HIn H3O - In-
Colour 1
Colour 2
The dissociation constant of the indicator, KIn, is given by
KIn =c(H3O ) x c(In-)/c(HIn)
For example, for methyl orange (Fig. 3):
Acid form, red Alkaline form, yellow + H+
As with buffers, the equilibrium and the concentration of hydrogen ions will
govern the ratio of color 2 to color 1.

–SO NH N N(CH3 )2 + = –
SO3 N N N(CH3)2 + H+
3

Red (pH<3) Yellow (pH>5)

Fig. 3. Structures of methyl orange at low and high pH (only one tautomeric form is shown).

Example 2
With the indicator bromocresol green, where KIn = 1.6 x 10-5, or pKIn = 4.8 and
color 1 (acid) is yellow, while color 2 (salt) is blue, a solution of pH 4.0 will give:
log [c(In-)/c(HIn)] = log [c(color 2)/c(color 1)] = pH -pKIn = -0.8
Therefore, c(color 2)/c(color 1) 0.16, which means about 14% blue, 86% yellow,
or visually a very yellowish green. When the pH is equal to the pKIn, there are
equal amounts of each form making a green color.
A wide range of indicators is available for titrations and other purposes and
these are discussed further in Topics C5 and C7.
This provides a useful and rapid method of estimating pH by eye; for
example, using litmus paper which is red below about pH 6 and blue above
pH 8. Both wide range and narrow range indicator papers are available to enable
a rapid estimation of pH. However, to determine the pH accurately using
indicators, careful spectrometric comparison would be needed and this is a time-
consuming method that is rarely used.

The pH meter uses a reference electrode and a glass electrode with a high-
resistance voltmeter and affords a rapid and accurate method of measuring pH
(Fig. 4). The calomel reference electrode is decribed fully in Topic C3.
The glass electrode is an example of a membrane ion-selective electrode and
is described in Topic C3. It responds to hydrogen ions:
E(glass) = E* + (RT/F) ln [a(H3O )X]
Therefore the complete cell
Pt Hg Hg 2Cl2 (s) KCl (sat, aq) Solution X glass membrane AgCl Ag has
an e.m.f. at 25 C equal to:
E (cell) = (E* - 0.241) + ( RT/F) ln [a(H3O )X]
pH meter

Hg–Hg2Cl2 paste Ag–AgCl electrode

Saturated KCl Buffer solution

Porous ceramic plug Glass membrane

Unknown solution

Fig. 4. pH measurement system.

By using one of the standards described above, for example, 0.05 M potassium
hydrogen phthalate, which has a pH at 25 o C of 4.008, we may eliminate E* and
measure the activity of hydrogen ions, and hence the pH, in the unknown
solution X.
pH(X) = pH(S) +(E S- E X)/(RT ln(10)/F)

pH control It is often necessary to control the pH of a solution, especially if hydrogen ions

are being generated or consumed, or if the nature of the species being analyzed
changes with the pH. A few examples will illustrate this problem.
(i) In recording the UV spectrum of a solution of a weak acid, such as a phenol,
the peak maxima occur at different wavelengths in an acid medium
compared with those in a basic medium. Comparisons may best be made
using a constant pH buffer.
(ii) In complexometric titrations (see Topic C7), such as the determination of
magnesium by EDTA, the complex is formed readily and completely at
high pH, so the titration is carried out using an ammonia-ammonium
chloride buffer to keep the solution at pH 10.
(iii) As noted in Topic C3, the fluoride electrode detects free F ions very well,
but OH ions interfere, and H ions form undissociated H2F2. It is therefore
essential to make measurements in a buffer of about pH 5 6.
pH is most often controlled by performing the analysis in a suitable buffer
solution. Occasionally, where a reaction produces an acid (or alkali) the tech-
nique of pH-stat titration may be used. Here, the acid produced is detected, and
sufficient alkali added to return the pH to the optimum value.

pg. 79
T ITRIMETRY I: ACID – BASE
TITRATIONS

Key Notes

Titrimetry Titrimetry is an analytical technique for the determination of the

stoichiometry of a reaction by the addition of controlled amounts of a
standard reagent.

Standard solutions Titrations usually involve the addition of controlled volumes of a

standard solution, whose concentration is known accurately, to a solution
of reactant of unknown concentration.

Equivalence points The theoretical amount of solution that must be added until the reaction
and end points is just complete is the equivalence point and the end point in a titration is
the point at which change is detected accurately. In an ideal case, these
points should be the same.

Indicators In order to detect the end point, a visual indicator may be added.
Instrumental methods may also be used.

Potentiometric The use of an electrochemical cell with an indicator electrode and a

titrations reference electrode to measure the concentration at each stage in the
titration facilitates the detection of the end point and the automation of
titrations.

Applications The major applications for acid-base reactions are the determination of the
concentrations or amounts of acids or bases.

Related topics Other topics in this Section (C1-C4, C6-C10)

Titrimetry In order to obtain accurate quantitative data for a reaction in solution, it is

necessary that the reaction be fast, complete and occur in fixed, reproducible
amounts. The requirement for fast reaction is achieved readily when ionic
species are involved, although in some other cases, it is necessary to warm the
solutions or add a catalyst. The reaction will be complete provided the equilib-
rium constant is large (see Topic C1).
The technique of volumetric analysis is the simplest type of titrimetry, and
involves the addition of controlled volumes of a reagent solution, the titrant, to
a known volume of another solution, the titrand in a volumetric titration. This
procedure may be automated, and the changes detected instrumentally (Topics
C2, C3, C9 and C10). In some cases, excess of a reagent is added and the excess
measured by back titration.
The volumes and concentrations can be measured with high accuracy.
Calibration of volumetric glassware by discharge into weighed containers
allows the determination of volumes to 0.01 cm3. This corresponds to about 0.05%
for a typical titration volume of 20 cm3.
Weighing solutes prior to dissolution gives comparable accuracy. Pre-deter-
mined volumes, for example 25 cm 3 are measured with manual or mechanical
pipets, while cumulative volumes of titrant during addition are measured by a
manual or automated buret.
Any reaction used in titrimetry will cause the concentrations of the species in
solution to change. For acid base reactions, the concentration of hydrogen ions, and
hence the pH, will alter and similar changes in the ionic concentrations occur with
other reactions.
Figure 1 shows the pH values in the titration of a strong base by a strong acid
as a function of the volume of acid added. pH is discussed in Topic C4. The figure
shows that there is only a small change in pH before about 24 cm3 where there is
still an excess of base, and also after 26 cm3 where there is an excess of acid.
However, in the region around 25 cm3, where the acid exactly neutralizes the
base, the change in pH is very large. This signifies the end point of the titration.
Provided that the end point coincides with the equivalence point or the stoi-
chiometry of the reaction studied, the amounts of titrant and titrand measured
should correspond to the actual amounts present.
For a general reaction:
aA+bB = cC + dD
a moles of A react with b moles of B to produce c moles of C and d moles of D.
A molar solution contains 1 mole of solute per 1000 cm3 (or 1 l) of solution.
Therefore, 1 cm3 of a 1 M solution contains 1 mmol of solute.

14

12

10

8
pH

0
0 10 20 30 40 50
Volume of acid/cm3

Fig. 1. Titration of 25 cm3 of 0.1 M NaOH with 0.1 M HCl (pH against volume added).

pg. 81
Example 1
A 25.00 cm3 aliquot of a solution of a base of known concentration 0.1057 M is
titrated with an acid of unknown concentration. The reaction involved 1 mole of base
and 1 mole of acid. The end point was determined as 24.88 cm3 of acidadded. What
is the concentration of the acid?
25.00 cm3 of the base solution contained 25.00 x 0.1057 = 2.6425 mmol base.
From the known reaction, this should be equivalent to 2.6425 mmol of acid. Since
the volume of acid at the end point was 24.88 cm3, the concentration must be
(2.6425/24.88) = 0.1062 M.
In some cases, the end point detected does not correspond exactly with the
equivalence point. This may be due to problems with the reaction, or to the small
amount of reagent needed to react with additional materials (for example, the
added indicator) present in the titrand. In these cases, a blank titration must be
performed, or allowance made for the titration error.

Standard In order to achieve the highest accuracy, it is necessary to use well-established

solutions standard materials as reagents in the primary calibration or standardization of
the reacting solutions (see Topic A5).
The most important of these are called primary standards, and should be
easy to obtain, purify and dry, should be stable and not hygroscopic, but should
be readily soluble and react rapidly and stoichiometrically. They should ideally
have a high relative molecular mass to minimize weighing errors.
The above criteria mean that reagents such as sodium hydroxide, which is
hygroscopic and may react with carbon dioxide from the air, and potassium
permanganate, which slowly decomposes in air, are unsuitable as primary
standards.
Solutions used for quantitative analysis need to be checked and calibrated
frequently. For example, hydrochloric acid solutions should be checked against
sodium carbonate solution and sodium hydroxide against potassium hydrogen
phthalate.

Equivalence The end point of a titration is based upon experimental observation, whereas the
points and end equivalence point is the theoretical value dependent upon the reaction equa- tion. In an
points ideal case, these should be the same, but a check may be needed to ensure that factors such
as blank errors do not affect the results.
In any titration, the end point corresponds to rapid changes in the concentra-tion of
species. This may be detected in many ways. Instrumental methods are discussed later,
and visual indicators are discussed below.
For reactions such as a strong acid neutralizing a strong base (Fig. 1), the change at the
end point is large, and the rate of change with volume is very great, as shown by the
derivative plot in Figure 2(b). In other cases, such as the reaction of weak acids with weak
bases, the titration curve shows a much less pronounced change, and the derivative plot
may be needed to confirm the end point, as in Figure 2(b). The second derivative plot is
also useful, but relies greatly on very accurate measurements. Plots of the concentration
of a species (for example H , or OH ), against the volume added, give straight lines inter-
secting at the end point.
If a mixture of acids, or a polybasic acid such as maleic acid, is titrated, then two end
points are obtained. Similarly, for mixtures of two titrands, two end points will be detected,
provided the species have sufficiently different equilib- rium constants for the reaction
(e.g. Ka1 Ka2).
(a) (b)
12 0.35

0.3
10

0.25
8
0.2

dpH/dV
pH

6
0.15

4
0.1

2 0.05

0 0
0 10 20 30 40 50 60 0 10 20 30 40 50 60
Volume of acid (cm3) Volume of acid (cm3)

Fig. 2. (a) Titration of 25.00 cm3 of 0.105 M ammonia with an ethanoic acid solution. (b) Derivative plot of d (pH)/dV
showing the end point as 26.80 cm3.

Indicators In general, indicators have two forms, which possess different colors. Indicators for acid-
base titrations are themselves weak acids or bases where the two forms differ in color as
shown in Figure 3 of Topic C4 for methyl orange.

HIn (color 1) = H + In- (color 2)

The choice of indicator depends upon the reaction to be studied. As noted in Topic C2,
the equilibrium constant of the indicator must match the pH range, or electrode potential
range of the species being titrated. Table 1 shows a selection of indicators for acid base
reactions.
As a general rule it is noted that the color change takes place over the range:
pH = pKIn ± 1
Similarly for other reactions, the concentration at which the indicator changes must match the
concentrations at the end point.
For example, in the titration of the weak dibasic maleic acid with sodium hydroxide, the
first end point, corresponding to sodium hydrogen maleate occurs at pH = 3.5, while the
second end point for disodium maleate is at pH = 9. Since pKIn = 3.7 for methyl orange,
this will change at the first end point, while phenolphthalein would change color around pH
= 9.

Table 1. Typical visual indicators for acid-base titrimetry

Acid–base Low pH High pH pKIn

color color
Thymol blue Red Yellow 1.7
Methyl orange Red Yellow 3.7
Bromothymol blue Yellow Blue 7.0
Phenolphthalein Colorless Red 9.6
Alizarin yellow R Yellow Orange 11

pg. 83
Potentiometric The use of electrodes, particularly the glass electrode for pH measurements and
titrations the wide range of other ion selective electrodes (ISE) described in Topic C3, enables
titrations to be studied throughout the addition of titrant, so that small changes may
be detected. It also allows automation of the titration.
To compare potentiometric titrations with those where the end point is detected
visually, it is useful to think of the ion selective electrode as an indi- cator
electrode responding to the ion to be detected, and to remember that it must be
combined with a reference electrode, the potential of which must not be affected
by the titration reactions.
In general, the titration will produce a graph such as Figure 2(a) where pH,
pIon or E(cell) is plotted against the volume of titrant added. The sharp change
at the end point is readily observed. Derivative plots such as Figure 2(b) are an
aid to finding the end point.
Three types of potentiometric titration may be recognized. Detailed discus-
sion of the reactions is given in Topics C6 and C7, but the principles will be
discussed here.
● R, or reagent sensed. A reagent is added, for example, some copper(II) EDTA
complex, which is sensed by a copper ISE. Addition of EDTA to a solution
containing Ca2 (or Mg2 , Ni2 etc.) ions gives an abrupt change at the end point
where the excess EDTA reduces the concentration of Cu2 ions.
● S, or sample sensed. The ISE senses the ion in the solution being titrated. pH
titrations using a glass electrode belong to this type, as do titrations of fluoride
with La3 .
● T, or titrant sensed. The titrant added to the sample is sensed by an ISE
responsive to that ion; for example a silver ion ISE for the titration of halide
ions by silver nitrate.
Potentiometric titrations for complex, precipitation and redox systems are discussed
in Topic C7.

Applications There are many applications for acid-base titrations, several of which are
routinely used analytical methods described in the appropriate topics.
● The determination of the concentration of acid in foods and pharmaceuticals.
● The measurement of acid number (or base number) during the course of a
reaction. For example, in the production of polyester resins by the reaction of
a glycol with maleic and phthalic acids, the total acid remaining is deter-
mined by titration of a weighed sample with potassium hydroxide using
phenolphthalein as indicator.
● The Kjeldahl method for nitrogen determination is a good example of a back
titration. The sample (for example, a food product) is oxidized by concen-
trated sulfuric acid to remove carbonaceous matter. Excess sodium hydroxide
solution is then added, and the ammonia released is carefully distilled off into a
known volume of standard acid, such as 0.1 M boric acid. The excess acid is
then titrated with standard alkali.
Automated titrations are important in producing rapid, reproducible results
in commercial and research laboratories. Samples may be prepared and loaded
using mechanical pipets or direct weighing and dissolution methods. Titrant is
added to the sample titrand solution using peristaltic pumps, or burets driven
by pressure or piston systems. The addition is very reproducible after accurate
calibration. The progress of the titration is most often followed by potentio-
metric measurements, as outlined above (see also Topic H2).
C OMPLEXATION , SOLUBILITY AND
REDOX EǪUILIBRIA

Redox equilibria

Titrimetry II (C7) Other topics in this Section (C1-C5,

C8-C10)

Complexation The formation of stable compounds and complexes is important in analytical

chemistry, since many species may be formed in a real sample. The amounts and
nature of the species present are analyzed to study speciation. For example, in a
natural water sample, the metal ions may form complexes with water molecules,
carbonate species, plant acids or pollutants. Complexes may be used for
titrations, both directly and for masking unwanted reactions.
The formation of a complex compound between an acceptor species, most
usually a metal ion, M, and a coordinating species, or donor ligand, L, involves
the formation of coordinate bonds, for example, hexamino cobalt (III)
Co3 + 6NH3 =Co (NH ) 33 6
The formation of such complexes involves interactions between the orbitals ofthe
central atom and suitable orbitals or lone pair electrons of the ligands. The structure
and stability of the complexes are discussed more fully in textbooks of inorganic
chemistry.
The formation constant for the equilibrium may be represented in two ways.
The stepwise formation constants, Kf, relates to each addition of a ligand
molecule:
M + L = ML Kf1 = c (ML)/(c (M) . c (L))
ML + L = ML2 Kf2 = c (ML2)/(c (ML) . c (L))
or, generally:
MLn 1 + L = MLn Kfn = c (MLn)/(c (MLn–1) . c (L))
The overall formation constant, b, relates to the formation of the entire complex
in one equation, so that for the complex with n ligands:

pg. 85
M + nL = MLn
bn = c (MLn)/(c (M) . c (L) n)
The overall formation constant is the product of the stepwise formation constants:
bn = Kf1 . Kf2. … Kfn or log bn = log Kf1 +log Kf2 +…+ log Kfn
When a ligand is used that can bond to several sites, it is referred to as a multi-
dentate ligand. One of the most important examples is ethylenediamine
tetracetic acid (EDTA):
(HOOCCH2)2-N-CH2-CH2-N- (CH2COOH)2
This tetrabasic acid, abbreviated to H4Y, has four acetate group and two nitro-
gens, which may complex to the central metal ion, as shown in Figure 1. It is
important to remember that for a satisfactory titration, the equilibrium constant
K of the reaction must be greater than 1000.
Since the concentrations of the various species containing Y (H4Y, H3Y-, H2Y2-
, HY3-, Y- ) will vary with the pH, a formation constant K MY may be written:
Mn- + Y4- = MY(n-4)+
KMY = c(MY(n-4 )/(c(Mn ).c(Y4-))
and c(Y4-) cL x 4

where 4 depends on the pH and the acid dissociation constants (K1, K2, K3, K4) of
EDTA, as shown in Table 1, and cL is the total concentration of all the ligand species.

2–
O
C
OO H2CH2
C C
O N
M
O CH2
N CH2
O C C

H2 O
CH2
C
O

Fig. 1. The structure of a metal-EDTA chelate showing its octahedral geometry.

Table 1. Values of 4 as a function of pH

pH α4 pH α4
2 3.7 x 10 14 8 5.4 x 10 3

3 2.5 x 10 11 9 5.2 x 10 2

4 3.6 x 10 9 10 3.5 x 10 1

5 3.5 x 10 7 11 8.5 x 10 1

6 2.2 x 10 5 12 9.8 x 10 1

7 4.8 x 10 4
To compare the formation constants at similar pH conditions the equilibrium
may be written:
K’MY = KMY 4 = c(MY(n-4) )/(c(Mn+) . cL)

Example
The formation constant KMY for magnesium-EDTA is 5 x 108. Should pH 5, or
pH 10 be used to titrate Mg?
At pH 5, for Mg:
K’MY -5 x 108 x 3.5 x 10-7 = 1.75 x 102
At pH 10, for Mg:
K’MY = 5 x 108 x 3.5 x 10-1 = 1.75 x108
Therefore, magnesium could be titrated only at pH 10, as the value of the equi-
librium constant is too small below this.
The fact that EDTA forms a number of ring systems adds considerably to the
stability of the complex. This is called the chelate effect.

Solubility The formation of insoluble compounds by reaction between two soluble species
is discussed in detail in Topic C8, where the amount of an insoluble product is
measured gravimetrically (Table 2).
The solubility equilibrium is described by the solubility product, Ksp:
Mn (solvated) +An-(solvated) =MA(solid) + solvent

Ksp = a (Mn+) . a (An-)

(Note: This equilibrium constant is written in the inverse way to most others,
and thus a very small solubility product is desirable for complete reaction. The activities
of the pure solid MA and of the solvent are taken as 1.)
There are some general rules governing which cations will form precipitates with which
anions. Nitrates and perchlorates are generally soluble.
Various other reagents are also useful for gravimetric analysis, for example
dimethylglyoxime (a) for nickel and oxine (b) for aluminum and magnesium (Fig. 2(a) and
(b)).
If an excess of a precipitating ion is present, in order to keep the equilibrium constant
unchanged, the concentration of the other ion must decrease. This is called the common
ion effect. Occasionally, complex formation may occur, so that the precipitate formed
redissolves when excess reagent is added.
Since equilibrium constants are thermodynamic quantities, they should really be written
in terms of activities and the activity coefficient, depending on the ionic strength taken into
account.

Table 2. Ions forming insoluble products (Ksp ~ 10 5)

Anions Cations

Cl , Br , I
- - -
Ag+, Pb2+, Cu+, Hg22+
CO32- Ca2+, Sr2+, Ba2+, Mg2+
SO42- Ca2+, Ba2+, Pb2+
OH- Al3+, Co 3+, Cr3+, Fe3+, Mg2+, Mn,2+ Ni2+, Zn2+
S2- Ag+, Cd2+, Cu2+, Fe2+, Hg2+, Mn2+, Pb2+, Zn2+

pg. 87
(a)
CH3C CCH3

NOH NOH

(b)

OH
Fig. 2. Reagents for the precipitation of metal ions. (a) Dimethylglyoxime. (b) Oxine.

Redox equilibria In order to establish a scale of oxidative power, it is necessary to have a stan- dard, and since
these reactions involve electrons, measurement of the reduction electrode potential is a
convenient way to do this. The details are given morefully in Topic C3.
Some standard reduction electrode potentials, where the reagents are at unit activity, at
25 C are given in Table 3. These potentials allow the prediction of which ions will oxidize
other ions, under standard conditions, that is when the concentrations are molar. A more
poisitve electrode potential will oxidize a more negative potential.
It was shown in Topic C2 that the electrochemical cell e.m.f. is related to the free energy
change, and hence to the equilibrium constant:

En = (RT/nF) ln K

Therefore, the larger the cell e.m.f, the larger the equilibrium constant, and the more
complete the reaction.

Example
For the reaction of cerium(IV) ions with iron(II) ions, what is the likely reaction, and what
is the equilibrium constant? Which reagent is the oxidizing agent, and which the reducing
agent?
Cell: Pt Fe2 , Fe3 Ce4 , Ce3 Pt
En (cell) = En (rhs) – En (lhs) =1.44 = 0.77 = 0.67 V

Table 3. Standard reduction electrode potentials of some common redox systems at

25∞C
Reaction E n/V

H2O2 + 2H+ + 2e- = 2H2O 1.77

MnO + 8H + 5e = Mn2 + 5H O 1.51
4 2
Ce4 + e-= Ce3 1.44
Cr2O 2 +14H + 6e = 2Cr3 +7H O 1.33
7 2
I2 + 2e- = 2I 0.54
Fe3 + e- = Fe2- 0.77
S4O 2-
6 + 2e = 2S O 2 3
- 2-
0.08
2CO2 + 2e- = C2O 2- 4 -0.49
ln K = 0.67/(8.314 x298/(1 x 96485)) = 26.09
or K = 2.1 x1011
Cerium(IV) oxidizes iron(II) almost completely.
Examination of Table 3 shows that permanganate will oxidize iron(II) and
oxalate, iodine will oxidize thiosulfate to dithionite, but iodide will be oxidized
by iron(III) to iodine.

pg. 89
T ITRIMETRY II:
COMPLEXATION ,
PRECIPITATION AND REDOX
TITRATIONS

Complexation
titrations

Precipitation
titrations

Redox titrations

Complexation During complexation reactions the concentration of the analyte ion (for example,
titrations a metal ion) changes most rapidly at the end point. As noted in Topic C6, the
most widely used complexing agent is ethylenediaminetetracetic acid or EDTA, and
Table 1 gives a selection of metal EDTA formation constants.
Using the values of 4 given in Topic C6, Table 1, we may calculate the practical,
or conditional, formation constant at a particular pH
K’MY =KMY α4
From the data in the tables, it can be calculated that magnesium could be
titrated at pH 10, but not at low pH. This has already been discussed in Topic C6.

Table 1. Metal-EDTA formation constants at 25 C

Cation KMY log (KMY)

Ag+
2.0 x 107
7.3
Mg2+ 4.9 x 108 8.7
Ca2+ 5.0 x1010 10.7
Fe2+ 2.1 x 1014 14.3
Fe3+ 1.0 x1025 25.1
Zn2+ 3.2 x 1016 16.5
Cd2+ 2.9 x 1016 16.5
V3+ 8.0 x 1025 25.9
It is possible to titrate two cationic species in a solution by performing the
titration at different pH values. However, if a solution of high pH must be used,
this might cause precipitation of metal hydroxides or other insoluble species. In
order to prevent this, secondary complexing agents can be added to retain the
metal ion in solution. Ammonium chloride and triethylamine are typical
reagents for this purpose.
Zinc ions, which might otherwise form insoluble Zn(OH) 2 at pH of 10, may be
converted to soluble zinc amine complexes. These are less stable than the EDTA
complex and the zinc may then be reacted quantitatively.
Secondary complexing agents may also act as masking agents. Examples
of this are the use of cyanide ions to form stronger complexes with heavy
metal ions so that magnesium can be titrated, or masking Fe2 and Mn2 using
hydroxylamine in water hardness determinations.
Standard solutions of EDTA may be prepared from the dry disodium salt
(Na2H2Y, RMM 336 or the dihydrate, RMM 372), by dissolving a known amount
in water free of heavy metals. Alternatively, the solution may be standardized
by a standard magnesium salt solution.
In complex and precipitation titrations, as in others, the end point corre- sponds
to a rapid change in the concentration of species. This may be detected by
instrumental methods, particularly potentiometry (see Topic C3) and by visual
indicators discussed below. Using suitable indicators, or potentiometric
measurements, it is possible to detect two or more end points.
Complexometric indicators behave in a similar way to titrating complexing
agents such as EDTA. They generally change color with pH, but one species, for
example HIn2 , will react with excess metal ions Mn :
HIn2 + Mn = MIn(n-3)+ + H+
blue red
Some selected indicators for complexometric titrations are given in Table 2.
In order to use potentiometric methods to study complexometric titrations,
an electrode specific to the metal ion may be used (see Topics C3 and C5), for
example, a copper ISE to follow the reaction of copper with EDTA.
Alternatively, a ‘J’-shaped electrode with a small mercury pool may be used
together with a small amount of added Hg-EDTA complex. This acts in a similar
way to the Class 2 electrodes, where the complexes determine the concentration
of ions in contact with the mercury pool:
Mn + HY3 = MY(n-4)+ + H +
Hg2 + HY3- = HgY2- + H
Hg2 + 2e- = Hg

Table 2. Indicators for complexometric titrations

Indicator Free color Complex color* Metal ions

Eriochrome black T Blue Red Ba, Cd, Ca, Pb, Zn

Pyrocatechol violet Yellow Blue Al, Bi, Cd, Co, Cu
Fe, Mg, Mn, Ni, Zn
Xylenol orange Yellow Red Bio, Cd, Pb, Th, Zn
Calcon carboxylic acid Blue Red Ca, Cd, Mg, Mn, Zn
*varies with metal and pH

pg. 91
The major application of complexation titrations is for the determination of the
concentrations or amounts of metallic elements in water, food and other industrial
samples.

Example
The calcium and magnesium ions in hard water may be determined. The solution
is adjusted to pH 12 with NaOH, when Mg(OH) 2 is precipitated. The calcium is
then titrated with EDTA using calcon carboxylic acid as indicator. Both calcium
and magnesium are then determined in by titrating a sample with EDTA at pH 10
using eriochrome black T, and finding the magnesium by difference.

Precipitation For precipitation reactions, the change in the concentration of either ion forming
titrations the precipitate may be considered. Since the changes often involve many orders
of magnitude of concentration, it is again convenient to use the pX notation. For
example, for the reaction of silver ions with chloride to form an insoluble silver
chloride precipitate
Ag + Cl- = AgCl (s)
the concentration may be expressed as:
pAg = -log (a (Ag )) ~ -log (c (Ag ))
Figure 1 shows the pAg values in the titration of sodium chloride by silver nitrate
as a function of the volume of silver nitrate added. This figure shows that, before
the end point pAg is very high (that is, the concentration of silver ions is small)
and changes little, because there is still an excess of chloride and the silver is
almost completely removed as precipitate. After the end point, there is an excess
of silver ions, the concentration increases and pAg decreases. In the region around
the end point, where the amounts are nearly equal, the change in pAg with volume
added is very large. If a mixture of iodide and chloride ions is titrated (dashed line),
the iodide, which is less soluble, precipitates first and pAg is even higher than for
chloride. Then the chloride precipitates. Both end points can be found.
As noted in Table 1 of Topic A5, silver nitrate, sodium chloride and potassium
chloride are primary standards for silver halide precipitation reactions. Other

14

1
pAg

(a)

0
0 5 10 15 20 25 30 35
Volume AgNO3 added
Fig. 1. Silver-halide titrations. (a) Chloride alone (solid line); (b) iodide plus chloride (dashed
line).
precipitation titrations (e.g., barium with sulfate, zinc with ferrocyanide) are less
commonly performed.
Indicators for silver-halide precipitation titrations are of two types. The first react
specifically when an excess of titrant becomes present immediately afterthe end
point for example, if a small amount of potassium chromate is added, it will
react with excess silver ions to produce deep red silver chromate in neutral
solutions (Mohr’s method). In acid solutions, the silver is titrated with potassium
thiocyanate (KCNS) solution (Volhard’s method). Iron (III) ammo- nium sulfate
solution is added and reacts with an excesss of thiocyanate to
produce a deep red iron thiocyanate species.
Adsorption indicators such as fluorescein adsorb onto the precipitate when
excess silver ions are present and the precipitate takes on a pinkish color.
As with other indicators, the change of color is detectable by eye over a range: log
(c(ion)) ±1
A selection of indicators for precipitation titrations is given in Table 3.

Table 3. Indicators for precipitation titrations

Precipitation Color 1 Color 2 Ions detected

CrO 42 Yellow Deep red Ag
Fe3 Light brown Deep red CNS
Fluorescein Green-yellow Pink Ag+

It is considerably easier to titrate mixed chloride and bromide in solution by

potentiometry. The use of a silver ion selective electrode, or even a silver wire,
together with a double junction reference electrode, since the chloride ions from
a calomel electrode would react, allows the determination of the silver ion
concentration. Other precipitation titrations may be followed using suitable ion
selective electrodes.
The major applications of precipitation titrations involve the determination of
halides with silver, or the reverse, or the determination of silver in acid solutions
with thiocyanate.

Example
An insecticide containing chlorine was digested in nitric acid to convert the
chlorine to soluble chloride. Silver nitrate was added in excess, and the excess
titrated with potassium thiocyanate by Volhard’s method. It is important to
know whether all the chlorine is converted to chloride.

Redox titrations Oxidation reduction or redox titrations are used for determining metals with two
well-defined oxidation states, and indirect methods for the determination of
organic compounds.
For redox reactions the concentrations both of the oxidized species, Ox, and
of the reduced species, Red, will change simultaneously. Considering a cell with
a redox electrode and a reference electrode:
SCE a (Ox), a (Red) Pt
the cell emf is given by:
E = En + (RT/F) ln (a (Ox)/a (Red)) - ESCE

pg. 93
Therefore, as the reaction proceeds during titration, the ratio of the concentra-
tions will change and the emf will alter. The potentiometric titration curve will
resemble those described in Topic C5. A summary list of redox reagents is given
in Table 3 of Topic C6.
For standard solutions, sodium oxalate and iron(II) ammonium sulfate and
potassium iodate are suitable, but potassium permanganate and iodine solu-
tions decompose on standing and must be standardized before use.
The indicators for redox reactions are reagents whose oxidized and reduced
forms differ in color:
In(Ox) (color 3) + ne- = In(Red) (color 4) An
example of this is 1,10-phenanthroline iron (II)
[Fe (C12H8 N2)3]3 + e- [Fe (C12H 8 N2)3]2
oxidized form, pale blue reduced form, deep red
In several cases, the indicator reaction additionally involves hydrogen ions, so
the change is pH dependent. Table 4 lists commonly used redox indicators.

Table 4. Indicators for redox titrations

Redox Oxidized Reduced EIn/V Solution

color color
1,10-phenanthroline iron(II) complex Pale blue Red 1.11 1 M H2SO4
Diphenylamine Violet Colorless 0.76 Dilute acid
Methylene blue Blue Colorless 0.53 1M acid
Phenosafranine Red Colorless 0.28 1M acid

For a redox indicator where one electron is involved, at 25 C, the color

change takes place at electrode potentials in the range
E EIn ± 0.059
One further useful indicator employed in redox titrations involving iodine is
starch, or more synthetic equivalent materials. The starch forms a blue-black
complex with iodine, which is rendered colorless when all the iodine has been
removed.
The applications of redox titrations include the determination of metals, with
two well-defined oxidation states, which are present in metallurgical samples
and ores. In order to dissolve the material, it may be necessary to use oxidative
conditions, for example, concentrated nitric acid. This will convert the majority
of the ions into their higher oxidation state, and in order to titrate them they
must first be reduced quantitatively. This may be done by passing the acidified
solution through a Jones reductor, which contains a zinc-mercury amalgam. The
effluent may then be titrated using a suitable oxidant. Some organic compounds,
such as phenols, may be determined by bromination with a bromate/bromide
mixture, followed by back titration of the excess using thiosulfate.
Chapter six : GRAVIMETRY

Key notes

Gravimetry Gravimetry is the analytical technique of obtaining a stable solid

compound, of known stoichiometric composition so that the amount of
an analyte in the sample may be found by weighing.

Precipitation This involves treatment of an analytical sample, usually in solution, to

obtain a quantitative amount of an insoluble compound of known
composition.

Purification The precipitate must be as pure as possible. Substances that are similar
and might precipitate under the same conditions must be removed, and
the analysis must be carried out so that no impurities are co-precipitated.

Drying and heating If precipitation is carried out from solution, the solid precipitate will have
solvent associated with it. This must be removed. Heating near the
boiling point of the solvent will do this, and further heating may be
needed to obtain a more stable compound whose formula is known.

Weighing The procedures of weighing the container initially and with the final
sample are most important.

Related topics Complexation, solubility and Thermogravimetry (G1)

redox equilibria (C6)

Gravimetry Gravimetry is one of the ‘classical’ techniques of analysis, and although less
frequently used now, it is of value when an accurate reference method is
required for comparison with an instrumental technique.
If an element is present in a mixture, for example, silver in a sample of nickel,
one way of separating it is to dissolve the metal completely in a suitable solvent.
In this example, the metal mixture could be dissolved in concentrated nitric acid
and a reagent added that would react with the silver to produce a precipitate,
which for silver might be a sodium chloride solution:
Ag(s) + Ni(s) +HNO3(sol) = AgNO3(sol) + Ni(NO3)2(sol)
AgNO3(sol) + NaCl(sol) =AgCl(s) + NaNO3(sol)
The silver chloride is precipitated completely, and may be filtered off since both
nickel nitrate and nickel chloride are very soluble in water. The precipitate will
be wet and may contain traces of nickel in solution, so must be thoroughly
washed and dried, as discussed below.
Since weighing may be carried out readily and accurately in almost all labora-
tories, gravimetry is often used as a reference method. Analysis of major compo-
nents of metal samples such as steel, and of minerals and soils may be carried
out by gravimetric methods, but they often involve lengthy separations and are

pg. 95
time consuming. Newer instrumental methods may determine several compo-
nents simultaneously, rapidly and are generally applicable down to trace levels.

Precipitation As noted in Topic C1, many elements will form compounds insoluble in water or other
solvents. Provided the compound is stable, or may be converted into a stable compound
easily, these insoluble precipitates may be used for analysis. The technique for obtaining a
precipitate may be summarized as follows:

(i) The sample should be dissolved as completely as possible in a suitable solvent. Any
residue that does not dissolve (for example, silica present in the metal sample of the
above example), may be filtered off at this stage.
(ii) Unless there are undesirable changes when the sample is heated, the solu- tions
should be warmed. This speeds up reactions and helps to form a more granular
precipitate.
(iii) The precipitating reagent must be chosen to give as insoluble a precipitate as
possible. Preferably, a reagent that will produce the largest mass of precipitate should
be used. For example, aluminum may be precipitated and heated to give the oxide,
Al2O3 when 10.0 mg of aluminum will produce 18.9 mg precipitate. If ‘aluminon’ (8-
hydroxyquinoline, C 9H6ON) is used, 170.0 mg of precipitate results.
(iv) The precipitating reagent should be added slowly, with stirring to the warm solution.
To check whether precipitation is complete, the precipitate is allowed to settle, and
more reagent added. If further precipitate does not form, the reaction is complete. If
the solution appears cloudy, it is possible that a colloidal form of the solid is present.
This may be coagulated by further warming or adding more reagent.
(v) The reaction mixture is filtered. Various means may be used for this. The simplest is a
quantitative filter paper (ashless), which has been dried and weighed previously. These
may be dried, or burnt (‘ashed’).

Another useful filter is a sintered glass or porcelain crucible, dried and weighed as before.
Glass will withstand heating to about 300 C and porcelain to over 800 C.

Purification The precipitate should be washed to remove traces of solution. This may cause difficulties,
as the washing may redissolve the solid. Using a wash liquid with a common ion reduces
the solubility.
For example, if 0.18 g of a precipitate of lead sulfate is washed with 1 dm3 of distilled water,
it will dissolve 0.046 g of precipitate or 25%, as the solubilityproduct is Ksp = 2.3 x10- 8 (mol
dm 3)2. If the precipitate is washed with 1 dm3 of
0.01 M sulfuric acid, then the amount dissolved is much less, 0.7 mg or 0.4%.

Drying and Drying can be done in stages. To remove water, the filtered solid in its container is placed
heating in a desiccator and left for a few hours. A vacuum desiccator is even more efficient for
removing solvents at low temperature. Heating in ovens, furnaces or directly with burners
will raise the temperature to remove materials or to decompose the precipitate to a more
stable form. For example, ‘basic aluminum succinate’ is a good precipitate for aluminum, but
must be ignited to constant weight at about 1200 C to convert to aluminum oxide.
The stages involved in drying and decomposition can be studied using
thermogravimetry (see Topic G1).
Weighing Modern balances can readily weigh samples directly, and masses from several
grams down to a few micrograms can be weighed accurately and quickly. It is
important that the conditions are the same for the initial weighing (crucible,
filter paper) as for the final weighings. Temperature is especially important and
hot samples should never be placed directly onto a balance pan.

pg. 97
VOLTAMMETRY AND
AMPEROMETRY

Instrumentation The cell uses a working microelectrode, a reference electrode, and a

counter electrode, and a controlled voltage supply.

Principles Voltammetric techniques involve the electrolysis of the solution to be analyzed

using a controlled external power source and measuring the resultant current -
potential or current time curves to obtain information about the solution.
The species to be determined undergoes oxidation or reduction at a working
electrode. The voltage between the working electrode and an auxil- lary or
counter electrode is controlled by the external circuitry in order to maintain a
preselected potential difference at the working electrode, with respect to the
reference electrode, as a function of time. A typical voltam- metric cell, shown
in Figure 1, has a working electrode, reference electrode
and an auxillary electrode and contains the solution to be analyzed. Often the
solution is deaerated with nitrogen to prevent interference due to the reduc- tion
reactions of oxygen.
If there is no reaction at the working electrode, the potential changes greatly for a
very small increase in current. A mercury drop electrode, for example, has a
polarization range between +0.3 V and -2.7 V against the SCE and in the absence of
oxygen so that many reactions that occur in that range may be studied.
By controlling the potential of the working electrode, a particular reaction
may be selected. Suppose a cell has two inert, solid electrodes and a reference
electrode, which dip into an aqueous solution containing copper ions.
In order to cause any reaction, the applied potential must exceed the decom-
position potential. This may be calculated by considering the reactions at each
electrode, and adding the extra potentials or overpotentials due to polarization
effects at the electrodes and to the voltage needed to drive the current against
the resistance of the solution. In this example, in order to drive the cell reaction,
a voltage greater than about –2.54 V must be applied.
Potentiostat

V A

Nitrogen in Nitrogen out

Working
electrode Reference electrode

Auxillary electrode
Sample
solution Stirrer

Fig. 1. Basic voltammetric cell.

The reaction that occurs is:

Cu2 + H2O = Cu(s) + 1⁄2 O2 + 2H
The copper is then deposited on the cathode and oxygen is evolved at the anode.
This is the basis for electrogravimetry, where the copper is completely deposited
from solution and the increase in weight of the cathode determined. The analysis
may be conducted using either a controlled potential or a controlled current.
Coulometric methods of analysis involve measuring the quantity of electricity
in coulombs needed to convert the analyte to a different oxidation state. If the
electrolysis occurs at 100% efficiency, Faraday’s laws may be applied and each
96485 C will bring about the reaction corresponding to 1 mole of electron transfer.
For example, using a silver anode, the passage of a current produces silver
ions, which react with any chloride in the solution. Bromine and acids may also
be generated coulometrically.
Polarographic methods employ a microelectrode, often a dropping mercury
electrode (DME), as the working electrode, plus a reference electrode (SCE) and a
mercury pool as auxillary electrode. The simplest potential time regime, where the
potential increases regularly (linear potential sweep dc voltammetry) is applied
to the cell containing the analyte and a supporting electrolyte to carry the
majority of the current.
In these methods, the transport of ions to the electrodes depends on three
factors: diffusion, convection or stirring, and conduction. The effects of conduc-
tion of the ion that reacts at the electrode is minimized by using a concentration
of supporting electrolyte such as KCl about 50-times higher than that of the
analyte. Stirring and convection are minimized. The resulting polarographic
curve shows three regions.
(i) If a potential difference is applied across a cell and no reaction occurs, only
the residual current Ir will flow.
(ii) If a reducible ion, say Cd 2 , is present, it will migrate to the dropping
mercury cathode. If the applied potential exceeds its decomposition poten-
tial, ED, it will be reduced to the metal which dissolves in the mercury:
Hg
Cd2 + 2e- = Cd(Hg)

pg. 99
As the cadmium plates out, the layer around the electrode is depleted and
more cadmium ions must diffuse in from outside through the diffusion layer
of thickness d. This will cause a current, I, to flow, which depends on the
concentration gradient between the bulk solution and the surface. Eventually,
the surface concentration becomes zero, and the limiting diffu- sion current
is reached:
Id = constant (c(bulk))/d -= kS (c(bulk))
The constant, kS, depends on the number of electrons transferred, the diffu- sion
coefficient of the ion in the solution, and the characteristics of the cathode.
(iii) If the potential is increased further, the current does not increase unless other
reducible ions are present. These three regions are shown in Figure 2.
The potential difference, E, across the cell at any stage is:
E= ESCE -EDME or
E =ESCE - (En + (RT/2F) ln [(a(Cd2 , surface)/a(Cd(Hg))]
Cd

From the equations above, the concentrations may be substituted by the currents,
since the concentration of reduced species in the mercury depends on the current
I and the diffusion constant in the amalgam, kA
I = kA(c(Cd(Hg))
E = E SCE - (En +Cd(RT/2F) ln (k /k ))A +S(RT/2F) ln [(I - I)/I]
d

When I = 1⁄2 Id, that is at the half-wave position, the DME has the half-wave
potential, E ⁄ 1
2

EDME, ⁄ = ESCE - (E +⁄ (RT/2F) ln [1⁄2 Id/1⁄2 Id] =E

1
2
1
2
1
⁄2

The half-wave potential is usually quoted relative to the SCE, and, like the
standard electrode potential, is characteristic of the electrode reaction. Typical
values are shown in Table 1.

Limiting current
Current, I

Residual current
E1/2

–0.5 –0.7 –0.9 –1.1

E/(V)

4
Fig. 2. Current-voltage curve for 10 M cadmium sulfate solution.
Table 1. Value of the half-wave potential
Ion E ⁄ /V with respect to the SCE
1
2

in 0.1 M KCl in 1 M NH3, NH4Cl

Cd2+ -0.60 -0.81
Cu2+ +0.04 -0.24
-0.22 -0.51
Zn2+ -1.00 -1.35
K+ -2.14
C6H5NO2 -0.22
Note that copper(II) is reduced in two stages.

Instrumentation For most voltammetric and amperometric methods, the instrumentation includes
a working microelectrode, a reference electrode and an auxillary or counter
electrode, together with electronic equipment to control the voltage and voltage
sweep, plus a computer or recorder to collect data.
The earliest microelectrode used was the dropping mercury electrode
(DME), where pure mercury flows through a fine capillary, either due to gravi-
tational force, or by applied pressure. Drop times of a few seconds are usual. This
electrode has the advantages that:
● the surface area is small and is constantly refreshed so that products of
electrolysis do not accumulate;
● mercury has a high overpotential for hydrogen formation, which allows the
reduction of other species.
Other electrodes used are the static, or hanging mercury drop electrode, where
the drop is dislodged at a particular time and size, and solid micro- electrodes, such
as platinum and glassy carbon, which may be incorporated into a rotating disc
electrode.
Dissolved oxygen in the sample solution must be removed, since oxygen may
be reduced in two steps, giving waves that overlap with those of the sample.
O2 + 2H +2e -= H2O2 E ⁄ = -0.05 V
1
2

H2O2 + 2H + 2e- = H 2O E ⁄ = -0.9 V

1
2

This is usually done by passing oxygen-free nitrogen through the sample

solution during the experiment.
Maxima on the waves are due to surface effects, and may be suppressed by
adding a small amount of surface-active agents, such as gelatin or Triton-X100.
Anodic stripping voltammetry is designed to measure trace amounts by
preconcentrating them onto a suitable electrode. The experiment has two stages:
(i) The sample is electrolyzed onto a hanging mercury drop, or a mercury film
deposited on a carbon electrode. By Faraday’s laws, (Topic C2), passing a
current of I amps for t seconds will produce a concentration cR in a mercury film
of thickness l, area A:
cR = It/nF lA
Because the current is limited by diffusion:
I = mnFD cB A
where m is a mass transfer coefficient, D the diffusion coefficient and cB the
bulk concentration.

pg. 101
(ii) The reduced species (that is, the metal) is then oxidized out of the film by
making the electrode increasingly anodic. A peak appears on the current-
potential plot, and the peak current can be shown to be:
Ip = k(cB nt)
where the constant k includes the diffusion and other constants, and n is the
rate of increase of the anodic potential. The peak potential at which an active
species is oxidized is characteristic of that species, and is close to its half-wave
potential.

Applications Polarographic techniques may be used in both qualitative and quantitative modes.
Since the half-wave potential is characteristic of the particular reaction that is
occurring at that potential, it is possible to identify the species involved. A
simple case is shown in Figure 3 where a mixture of metal ions was analyzed.
The two reduction waves for copper occur at –0.1 and –0.35 V, cadmium at
–0.69, nickel at -1.10 and zinc at –1.35 V. This illustrates an analysis that may
identify the species qualitatively and, by using a standard addition method, can
also determine the ions quantitatively.
Organic substances may be determined either in an aqueous or a nonaqueous
medium. For example, the concentration of nitrobenzene in commercial aniline
may be found by studying the reaction:
C6H 5NO2 + 4H + 4e = b C6H5NHOH + H2O
The oxygen electrode is based on voltammetric principles and depends on
the diffusion and reduction of oxygen. It is also called the Clark sensor.
The cell has a lead anode and a silver cathode set close together in an alkaline
solution, often 1M KOH. At the anode, the reaction is
Pb(s) + 4OH- (aq) = PbO 2- 2(aq) + 2H O 2 2e-
The silver cathode is inert, unless oxygen or another reducible species can diffuse
to it. A semipermeable membrane through which only oxygen can diffuse
surrounds the electrodes, and then the reduction reaction takes place.
O2(aq) + 2H2O + 4e- = 4OH
Current, I

0 0.3 0.6 0.9 1.2 1.5

E (applied) (V)

Fig. 3. Polarogram of Cu2+, Cd2+, Ni2+ and Zn2+ ions at ~10-4 M.

Since the current depends on the diffusion of the oxygen to the electrode from
the external solution, and this diffusion is proportional to the concentration of
oxygen in the external solution, this electrode may be used to measure dissolved
oxygen.
Amperometric titrations are used to determine substances by measuring the
limiting diffusion current of a species as a function of the volume of a reagent
added to react with that species. Since Id is proportional to the concentration, it
will decrease as a species is used up, or increase as the excess of a species becomes
greater. For example, for the determination of Pb2 with Cr2O 2 : 7

2Pb2 + Cr2O 27- + H O2 = 2PbCrO + 2H

4

At an applied potential of 0.0 V, and at pH4, dichromate is reduced, but Pb 2+ is

not, giving the graph shown in Figure 4.
Applications of anodic stripping voltammetry, are chiefly for the determina-
tion of trace amounts of amalgam-forming metals (Fig. 5), while cathodic strip-
ping voltammetry is used for determining species that form insoluble salts with
mercury. The preconcentration stage allows determination in the concentration
range 10 6 to10 8 M.

12

10

8
Current I (A)

0
0 5 10 15 20 25 30 35 40
Volume dichromate added (ml)

Fig. 4. Amperometric titration of Pb(II) with dichromate at pH 4 and 0.0 V.

Current I (A)

Cu

Pb

0 –0.2 –0.4 –0.6

E /V
Fig. 5. Anodic stripping voltammogram for 7.5 ppb Cu(II), 2.6 ppb Pb(II) and trace amounts
of Cd(II).

pg. 103
C ONDUCTIMETRY

Increasing the number of ions increases the conductance.

Conductivity Since analysts are concerned with concentrations, it is preferable to

compare molar conductivities, which depend on the chacteristics of the
ions.

Conductimetric
titrations

Movement of ions When an electric field is applied across a solution, a force is exerted on the ions
that will cause them to move. Positive ions will move toward the more negative
electrode, negative ions toward the more positive, but each will carry current.
The speed with which the ions move usually depends on:
(i) the electric field strength (V m 1);
(ii) the charge z on the ion;
(iii) the size of the ion in solution; and
(iv) the viscosity of the solvent.
In water, the hydrogen ion, H3O , and the hydroxyl ion, OH , are small, and
move most rapidly by a very fast exchange with the molecules of water. In dilute
solutions, the ions move independently of each other.
For an electric field of 1 V m 1, the ionic speed is called the mobility, ui. Table
1 gives the ionic size and mobility of selected ions in aqueous solutions at 25 C.
It is worth noting that despite their larger size in crystals, hydrated potassium
ions are smaller than hydrated sodium ions when in solution and move faster.

Table 1. Selected vales of mobility and ionic radius

Ion 108 ui/m2 s 1

V 1
ri /nm
Na+ 5.2 0.18
K+ 7.6 0.13
F+ 5.5 0.17
Cl+ 7.9 0.12
Br+ 8.1 0.12
Similar rules apply to anions. Also, potassium ions and chloride ions have very
similar mobilities. This is useful when ‘salt bridges’ are needed (see Topic C2).
If ions are present in other solvents, such as liquid ammonia, they will conduct
in a similar way, but their mobilities will be different.

Conductance Ohm’s law still applies to relate the current, I, to the applied voltage, E, and the
resistance, R, or to the conductance, G = 1/R.
I = E/R = EG
G is related to the concentration of ions and has units of ohms-1 or siemens (W-1
or S).
In order to measure the conductance of a solution, a conductance cell is used.
It has two electrodes, often platinum, coated with platinum black, separated by
a fixed distance. Connecting these to a conductance meter, or more simply to a
Wheatstone bridge, the conductance may be measured directly. In order to prevent
electrolysis (see Topics C2, C9) taking place, which would change the concentrations
in the solution, the bridge uses alternating current. Plastic and flow-through
conductance cells are also available
In any conductor, the conductance depends on the cell dimensions and on the
characteristics of the conducting medium. For a cell with electrodes of area A,set
a distance l apart:
G = k/(l/A)
where k is called the conductivity of the solution with units of W 1 m 1 or S m 1.
So that the results from different cells and solutions may be compared, the cell
constant, K = (l/A) is measured using a solution of known conductivity, often
0.1 M aqueous potassium chloride, which has a conductivity k =1.288 α-1 m- 1 at
25 C.

Conductivity Although it is often possible to work directly with conductance, it is useful to

know the conductivity and molar conductivity when results are to be compared.
From the equations above:
k = G(l/A)
Since the analyst is concerned with molar concentrations, and the conductivity
depends mostly on the amounts of ions present, the molar conductivity Lm is
defined by:
Lm = k/c
where Lm has units of W 1 m2 mol 1 provided that the concentration c is
expressed in mol m-3 (= 1000 M).

Example
When immersed in a 0.1 M aqueous solution of KCl at 25 C, a conductance cell
had G = 8.59 x 10-3 α-1. For 0.1 M HCl, using the same cell, the conductance was
measured as 2.57 x 10-2 W 1. Calculate the conductivity and molar conductivity of
the HCl.
For the KCl calibration:
k =G(l/A) =8.59 x 10-3 x (l/A) = 1.288 α-1m-1
Thus (l/A) =149.9 m 1 and for HCl, k =3.85 α-1 m- 1

pg. 105
Lm =k/c =3.85/100 = 0.0385 α-1 m2 mol 1
For strong electrolytes, it must be noted that, because of ionic interaction and the
effects of the ionic atmosphere, the molar conductivity decreases as the
concentration increases, according to the Onsager equation:
Lm = Lom - K c
where Lo mis the molar conductivity at infinite dilution, and K is a constant,
dependent on temperature and the solute.
For weak electrolytes, the ionic concentration depends on the dissociation
constant, and since the relationship of the dissociation constant, K to the degree
of ionization is, for a weak acid, approximately:
K = c 2/(1 – α)
The molar conductivity of a weak electrolyte may be written:
Lm = αLo m

These equations show that, while Lm for a strong electrolyte decreases almost
linearly with the (c)1/2, for weak electrolyes such as ethanoic acid, it decreases
approximately as (1/c)1/2. More details will be found in textbooks of physical
chemistry.
Direct measurement of concentration is possible. The presence of sodium
chloride, or other ionic species will increase conductance, and by knowing the
molar conductivity, the concentration may be found by measuring the
conductivity of the solution.
When very pure water is required for the preparation of pure organics, or the
manufacture of semiconductor components, conductance measurements are
most useful. Provided dissolved CO 2 and other intrinsic factors are eliminated,
extremely pure water should have a conductivity of around 150 mS m 1 (or
1.5 mS cm 1) at 25 C. In order to comply with regulations for some work, it
must be less than 200 mS m 1.

Conductimetric Whenever titrations of ionic solutions are carried out, the number and nature of
titrations the ions change throughout the entire titration. If a strong base, say NaOH, is
titrated with a strong acid, HCl, the reaction
Na + OH + H + Cl- = Na + Cl-+ H2O
will first of all remove the OH as nonconducting water, so the conductance will
decrease until the end point. Then excess HCl will increase the conductance. This
gives a ‘V’ shaped graph of conductance versus titer. If a mixture of acids is used,
they will be titrated in the order of their strength. Figure 1(a) shows the
conductimetric titration of a mixture of HCl, boric acid and ammonium chloride
using NaOH.
For precipitation reactions:
Ag + NO
3
- + Na +Cl = AgCl(s) + Na + NO -
3

The sodium and silver ions have similar conductivities, so adding NaCl solution
does not change the conductance much until the end point is reached. Then the
conductance rises with excess NaCl. This is shown in Figure 1(b). The titrant is
often about ten-times more concentrated than the sample solution, to avoid
dilution errors.
(a) (b)
1.4 4

12 3.5
1000 x conductance

3
1

Conductance
2.5
0.8
2
0.6
1.5
0.4 1
0.2 0.5
0 0
0 0.5 1 1.5 2 0 1 2 3 4 5
Volume added (cm3) Volume added (cm3)

Fig. 1. (a) Conductimetric titration of mixed acids with 1.0 M NaOH. (b) Conductimetric titration of 0.1 M AgNO3 with
1.0 M KCl.

Conductimetric titrations may be carried out in nonaqueous media. For

example, phenols, which are very weak acids in water, have enhanced acidity
when liquid ammonia is used as the solvent, and may be titrated conductimetri-
cally with an ammoniacal solution of KOH.

pg. 107
Chapter seven: SOLVENT AND S OL I D -P H AS E
E XT R A C T I O N

Key Notes

Extraction techniques Solvent and solid-phase extraction are two techniques for separating
mixtures of substances, either by selective transfer between two
immiscible liquid phases or between a liquid and a solid phase.

Extraction efficiency Extraction efficiency is defined as the fraction or percentage of a

and selectivity substance that can be extracted in one or more steps. Selectivity is the
degree to which a substance can be separated from others in a mixture.

Solvent extraction Procedures are based on the extraction of nonpolar, uncharged species
from an aqueous solution into an immiscible organic solvent, or the
extraction of polar or ionized species into an aqueous solution from an
organic solvent.

Solid-phase sorbents Sorbents are particulate materials such as silica, chemically-modified

silica, alumina and organic resins that can interact with and retain
substances from solutions. Retained substances can be subsequently
released back into a suitable solvent under controlled conditions.

Solid-phase Sample solutions are passed through the sorbent under conditions where
extraction either the analyte(s) are retained and matrix components washed through
or the reverse. Retained analytes are removed with an alternative solvent
before completing the analysis.

Related topic Solution equilibria (C1)

Extraction The purpose of an extraction technique is physically to separate components of

techniques a mixture (solutes) by exploiting differences in their relative solubilities in two
immiscible liquids or between their affinities for a solid sorbent. Substances
reach an equilibrium distribution through intimate contact between the two
phases, which are then physically separated to enable the species in either phase
to be recovered for completion of the analysis. An equilibrium distribution of
the solutes between the two phases is established by dissolving the sample in a
suitable solvent, then shaking the solution with a second, immiscible, solvent or
by passing it through a sorbent bed or disk. Where the equilibrium distributions
of two solutes differ, a separation is possible. The principal factors that deter-
mine how a solute will distribute between two phases are its polarity and the
polarities of each phase. Degree of ionization, hydrogen bonding and other
electrostatic interactions also play a part.
Most solvent-extraction procedures involve the extraction of solutes from an
aqueous solution into a nonpolar or slightly polar organic solvent, such as
hexane, methylbenzene or trichloromethane, although the reverse is also
possible. Readily extractable solutes, therefore, include covalent neutral mole- cules
with few polar and no ionized substituents; polar, ionized or ionic species will
remain in the aqueous phase. A wide range of solid sorbents with different polarities
are used in solid phase extraction making this a very versatile tech-nique, but the
factors affecting the distribution of solutes are essentially the same as in solvent
extraction. Two examples illustrate these points.

Example 1
If an aqueous solution containing iodine and sodium chloride is shaken with a
hydrocarbon or chlorinated hydrocarbon solvent, the iodine, being a covalent
molecule, will be extracted largely into the organic phase, whilst the completely
ionized and hydrated sodium chloride will remain in the aqueous phase.

Example 2
If an aqueous solution containing a mixture of weakly polar vitamins or drugs
and highly polar sugars is passed through a hydrocarbon-modified silica sorbent
which has a nonpolar surface, the vitamins or drugs will be retained on the
surface of the sorbent whilst the sugars pass through.

Solvent extraction is governed by the Nernst Distribution or Partition Law. This

states that at constant temperature and pressure, a solute, S, will always be
distributed in the same proportions between two particular immiscible solvents.
The ratio of the equilibrium concentrations (strictly their activities (Topic C1)) in the
two phases defines a distribution or partition coefficient, KD, given by the
expression
[S]org
KD — (1)
[S]aq

where [] denotes concentrations (activities) of the distributing solute species, S,

and the value of KD is independent of the total solute concentration. In practice,
the solute often exists in different chemical forms due to dissociation (ioniza-
tion), protonation, complexation or polymerization, so a more practically useful
expression which defines a distribution or partition ratio, D, is
(CS)org
D — (2)
(CS)aq

where (CS) represents the total concentration of all forms of the distributing solute S in each
phase. If no interactions involving S occur in either phase, then D and KD would be identical.
However, the value of D is determined by the experimental conditions and it can be adjusted
over a wide range to suit the requirements of the analytical procedure.
A distribution ratio can also be defined for solid phase extraction, i.e.
(CS)sorb
D= — (CS)liq
(3)

where the numerator represents the solute concentration in the solid sorbent and the
denominator the solute concentration in the liquid phase.
Solutes with large values of D (e.g. 10 4 or more) will be essentially quantita- tively
extracted into the organic phase or retained by the sorbent, although the nature of an
equilibrium process means that 100% extraction or retention can never be
achieved.
Extraction The efficiency of an extraction depends on the value of the distribution ratio, D.
efficiency and For solvent extraction, it also depends on the relative volumes of the two liquid
selectivity phases and for solid-phase extraction on the surface area of the sorbent. With
solvent extraction, the percentage of a solute extracted, E, is given by the expres-
sion
100D
E=
—— (4)
[D + (Vaq/Vorg)]

where Vaq and Vorg are the volumes of the aqueous and organic phases, respec-
tively.
For solutes with small values of D, multiple extractions will improve the
overall efficiency, and an alternative expression enables this to calculated
n
(Caq) n = C aq[V aq/(DV org+ V aq)] (5)

where Caq and (Caq)n are the amounts of solute in the aqueous phase initially and
remaining after n extractions, respectively.
The following example demonstrates the use of these formulae.

Example 1
A water sample contains 10 mg each of a halogenated pesticide and an ionic herbicide
which are to be separated by extraction of the pesticide into methyl- benzene. Given that
the pesticide distribution ratio, D, for methylbenzene/ water is 50, calculate the extraction
efficiency for

(i) one extraction of 20 cm3 of water with 10 cm3 of methylbenzene

(ii) one extraction of 20 cm3 of water with 30 cm3 of methylbenzene
(iii) three extractions of the same 20 cm3 of water with 10 cm3 portions of
methylbenzene (30 cm3 in total)

(i) Substitution of the values for D, Vaq and Vorg in equation (4) gives
100 x 50
E=— —= 96.15%
[50 + ( 20/10
)]

(ii) Substitution of the values for D, Vaq and Vorg in equation (4) gives
100 x50
E=— —= 98.68%
[50 + ( 20/30
)]

(iii) Substitution of the values for D, Vaq, Vorg, Caq and (Caq)n in equation (5) gives

(Caq)3 = 10 x [20/((50 x10) 20)]3

= 10 x [0.03846]3 5.6896 x 10 4 mg
E = 99.99%

An extraction of 99.99%, as achieved in (iii), would be considered quantita-tive, although

the lower efficiencies obtained in (i) and (ii) might be acceptable in the context of a defined
analytical problem. It is clear that increasing the volume of organic solvent, or extracting with
the same volume divided into several smaller portions, increases the overall efficiency of an
extraction.
NB the ionic herbicide has a negligibly small distribution ratio, being very polar and
highly water soluble.
Selectivity in extraction procedures is the degree to which solutes in a
mixture can be separated by virtue of having different distribution ratios. For
pg. 111
two solutes with distribution ratios D1 and D2, a separation or selectivity factor,
b, is defined as
b = D1/D2 where D1 ≥ D2 (6)
Selectivity factors exceeding 104 or 105 (log b values exceeding four or five) are
necessary to achieve a quantitative separation of two solutes, as for most prac-
tical purposes, a separation would be considered complete if one solute could be
extracted with greater than 99% efficiency whilst extracting less than 1% of
another. The extraction of many solutes can be enhanced or supressed by
adjusting solution conditions, e.g. pH or the addition of complexing agents (vide
infra).

Solvent extraction Solvent extraction (SE) is used as a means of sample pre-treatment or clean-up to
separate analytes from matrix components that would interfere with their detection
or quantitation (Topic A4). It is also used to pre-concentrate analytes present in
samples at very low levels and which might otherwise be difficult or impossible to
detect or quantify. Most extractions are carried out batchwise in a few minutes
using separating funnels. However, the efficient extraction of solutes with very
small distribution ratios ( 1) can be achieved only by continu- ously exposing the
sample solution to fresh solvent that is recycled by refluxing in a specially designed
apparatus.
Broad classes of organic compounds, such as acids and bases, can be sepa- rated
by pH control, and trace metal ions complexed with organic reagents can be
separated or concentrated prior to spectrometric analysis (Topic E10).

Extraction of organic acids and bases

Organic compounds with acidic or basic functionalities dissociate or protonate
in aqueous solutions according to the pH of the solution. Their extraction can,
therefore, be optimized by pH adjustments. The relation between pH and the
distribution ratio, D, of a weak acid can be derived in the following way:
A weak acid, HA, dissociates in water according to the equation
HA =H+ + A (7)
The acid dissociation constant, Ka, is defined as
[H+]aq[A-]aq
Ka = —— (8)
[HA] aq

Only the undissociated form, [HA], can be extracted into a nonpolar or slightly
polar solvent such as diethyl ether, the distribution or partition coefficient, KD,
being given by
[HA]ether
KD = — (9)
[HA]aq

However, the distribution ratio takes account of both the dissociated and
undissociated forms of the acid in the aqueous phase and is given by
[HA]ether
D = —— (10)
([HA]aq + [A ]aq)

Re-arrangement of equation (8) and substitution for [A-]aq in equation (10) gives
[HA]ether
———
D [HA]aq (1 + Ka/[H ]aq) (11)
+

and substituting KD for [HA]=ether/[HA]aq from equation (9) gives

KD
D = —— (12)
(1 + Ka[H ])

Equation (12) shows that at low pH, when the acid is undissociated, it is extracted with the
greatest efficiency as D KD, whereas as the pH is increased the value of D decreases until
at high pH the acid is completely dissociated into the anion A , and none will be extracted.
This is shown graphically in Figure 1 as pH versus E plots for two weak acids (curves 1 and
2) with different acid dis- sociation constants (Ka or pKa values), curve 1 being for the
stronger of the two acids. For a weak organic base, such as an amine, protonation occurs
at low pH according to the equation

RNH 2 H RNH 3 (13)

The relation between pH and the distribution ratio for a weak base (curve 3) is therefore
the opposite of a weak acid so that it is possible to separate acids from bases in a mixture
either by extracting the acids at low pH or the bases at high pH. Separating mixtures of
acids or mixtures of bases is possible only if their dissociation or protonation constants
differ by several pK units.

Extraction of metals
Metal ions in aqueous solutions are not themselves extractable into organic solvents, but many
can be complexed with a variety of organic reagents to form extractable species. Some
inorganic complex ions can be extracted as neutral ionic aggregates by association with
suitable ions of opposite charge (counter ions). There are two principal metal extraction
systems:

Uncharged metal chelate complexes (ring structures that satisfy the coordination
requirements of the metal) with organic reagents
e.g. 8-hydroxyquinoline (oxine)
acetylacetone (AcAc)

100

80
Percent extracted (E)

60

40

20

0
0 1 2 3 4 5 6 7 8 9
pH

Fig. 1. Solvent extraction curves for two organic acids having different pKa values (curves 1
and 2) and a base (curve 3). Curve 1 represents the acid with the larger Ka (smaller pKa).

pg. 113
1-(2-pyridylazo)-2-naphthol (PAN)
sodium diethyldithiocarbamate (NaDDTC)
(see also Table 1);
Electrically neutral ion-association complexes
e.g. Cu(neocuproine) , NO
2 3
(C6H5CH2)3NH , GaCl 4
[(C2H5)2O]3H , FeCl 4
(see also Table 1).
The Nernst distribution law applies to metal complexes, but their distribution
ratios are determined by several interrelated equilibria. As in the case of organic
acids and bases, the efficiency of extraction of metal chelates is pH dependent,
and for some ion-association complexes, notably oxonium systems (hydrogen ions
solvated with ethers, esters or ketones), inorganic complex ions can be extracted
from concentrated solutions of mineral acids.
Reagents that form neutral metal chelate complexes (5- or 6-membered ring
structures) are weakly acidic and contain one or more additional co-ordinating
sites (O, N or S atoms). Protons are displaced according to the general equation

Mn+(H2O)x + nHR = MRn + nH+ + xH2O (14)

not extractable extractable

where HR is a weakly acidic and co-ordinating reagent (ligand), and the metal
ion Mn has a formal valency n. The removal of hydrogen ions is necessary to
drive the reaction to completion, and pH control, which is essential, is achieved
by buffering the aqueous solution. It is sometimes possible to improve the selec-
tivity of the procedure by adding an additional reagent, known as a masking agent,
that reacts preferentially with one of the metals to form a nonextractable water-
soluble complex. Typical masking agents include EDTA (ethylenedi-
aminetetraacetic acid), citrate, tartrate, fluoride, cyanide and thiourea. Solvents
commonly used to extract metal chelate complexes include trichloromethane, 4-
methyl-pentan-2-one and methylbenzene.
Electrically neutral ion-association complexes consist of cationic (positively
charged) and anionic (negatively charged) species that form an overall neutral
aggregate extractable by an organic solvent. Either the cations or the anions

Table 1. A selection of reagents and extraction systems for the solvent extraction of
metals
Reagent Type of metal complex

8-Hydroxyquinoline (oxine) Neutral metal chelate complexes,

Di-alkyldithiocarbamates extractable into organic solvents. Intense
e.g. sodium diethyldithiocarbamate color of many facilitates colorimetric
(NaDDTC) determinations.

1,10-Phenanthroline (o-phen) Ion-association complexes. Metals as

2,9-Dimethyl-1,10-phenanthroline cationic or anionic chelated complexes
(neocuproine) extracted with suitable counter ion.
Ethylenediaminetetraacetic acid (EDTA)
Oxonium systems: Ion-pairs with anionic metal halide or
i.e. protons solvated with alkyl ethers, thiocyanate complexes. Chloride complexes
ketones, esters or alcohols extractable from strong HCl solutions.
should be bulky organic hydrophobic groups to provide high solubility in
nonpolar solvents. The metal ion can be a cationic or an anionic complex, and
can be an inorganic species such as FeCl 4, MnO 4 or a chelated organic complex
such as Fe(1,10-phenanthroline) 2 or UO (oxine) . Suitable counter-ions of
3 2 3
opposite charge to the metal-complex ions include (C4H9)4N , ClO 4
(C6H5CH2)3NH , [(C4H9O)3P O]H and an oxonium ion such as [(C2H5)2O]3H .
A list of metal chelate complexing agents and ion-association complexes is given
in Table 1. Most complexing agents react with a large number of metals (up to
50 or more), but pH control, the use of masking agents and a variety of ion-
association systems can enable the selective extraction and separation of just one or
two metals to be accomplished.

Solid-phase These are generally either silica or chemically-modified silica similar to the
sorbents bonded phases used in high-performance liquid chromatography (Topics D6 and
D7) but of larger particle size, typically 40–60 mm diameter. Solutes interact with
the surface of the sorbent through van der Waals forces, dipolar interac- tions,
H-bonding, ion-exchange and exclusion. The four chromatographic sorp- tion
mechanisms described in Topic D2 can be exploited depending on the sorbent
selected and the nature of the sample. Sorbents can be classified according to the
polarity of the surface. Hydrocarbon-modified silicas are nonpolar, and
therefore hydrophobic, but are capable of extracting a very wide range of organic
compounds from aqueous solutions. However, they do not extract very polar
compounds well, if at all, and these are best extracted by unmodified silica,
alumina or Florisil, all of which have a polar surface. Ionic and ionizable solutes
are readily retained by an ion-exchange mechanism using cationic or anionic
sorbents. Weak acids can be extracted from aqueous solu- tions of high pH when
they are ionized, and weak bases from aqueous solutions of low pH when they
are protonated. It should be noted that this is the opposite way around compared
to solvent extraction into non-polar solvents. However, by suppressing
ionization through pH control, extraction by hydrocarbon- modified silica
sorbents is possible. Sorbents of intermediate polarity, such as cyanopropyl and
aminopropyl modified silicas may have different selectivities to nonpolar and
polar sorbents. Some SPE sorbents are listed in Table 2 along with the
predominant interaction mechanism for each one.

Solid-phase Compared to solvent extraction, solid-phase extraction (SPE), is a relatively

extraction new technique, but it has rapidly become established as the prime means of
sample pre-treatment or the clean-up of dirty samples, i.e. those containing high
levels of matrix components such as salts, proteins, polymers, resins, tars etc. In
addition to being potential sources of interference with the detection and quan-
titation of analytes, their presence can be detrimental to the stability and perfor-
mance of columns and detectors when a chromatographic analysis is required.
The removal of interfering matrix components in general and the pre-concentra-
tion of trace and ultra-trace level analytes are other important uses of SPE which
is versatile, rapid and, unlike solvent extraction, requires only small volumes of
solvents, or none at all in the case of solid phase microextraction (vide infra).
Furthermore, SPE sorbents are cheap enough to be discarded after use thus
obviating the need for regeneration. The analysis of environmental, clinical,
biological and pharmaceutical samples have all benefited from the rapid growth
in the use of SPE where it has largely replaced solvent extraction. Specific exam-
ples include the determination of pesticides and herbicides in polluted surface

pg. 115
Table 2. Typical SPE sorbents and interaction mechanisms

Sorbent Polarity Interaction mechanisms

Silica SiO2 Polar Adsorption;
H-bonding
Florisil, MgSiO3 Polar H-bonding
alumina Al2O3 Polar H-bonding
Bonded phases (modified silica)
-C18H37 (C18 or ODS) Nonpolar Van der Waals interactions
-C8H17 (C8 or octyl) Nonpolar Van der Waals interactions
-C6H5 (phenyl) Nonpolar Van der Waals interactions
and p–p
interactions
-(CH2)3CN (cyanopropyl) Polar Polar interactions;
H-bonding
-(CH2)3NH2 (aminopropyl) Polar H-bonding
-(CH2)3C6H4SO3H Ionic Cation exchange
-(CH2)3N(CH3)3Cl Ionic Anion exchange
Chiral (cyclodextrin) Polar Adsorption;
H-bonding
dipolar interactions
steric effects
Styrene/divinyl benzene co-polymer Nonpolar Size exclusion

waters and soils, polycyclic aromatic hydrocarbons (PAHs) in drinking water,

polluted industrial and urban atmospheres, and drugs in biological fluids.
Sorbents are either packed into disposable cartridges the size of a syringe barrel,
fabricated into disks or incorporated into plastic pipette tips or well
plates. Most SPE is carried out using a small packed bed of sorbent (25–500 mg)
contained in a cartridge made from a polypropylene syringe barrel, the sorbent
being retained in position by polyethylene fritted disks. The sorbent generally
occupies only the lower half of the cartridge, leaving space above to accommo-
date several millilitres of the sample solution or washing and eluting solvents. A
typical cartridge procedure is illustrated in Figure 2 and consists of four distinct
steps:
● Sorbent conditioning. The cartridge is flushed through with the sample solvent
to wet the surface of the sorbent and to create the same pH and solvent
composition as those of the sample, thus avoiding undesirable chemical changes
when the sample is applied.
● Sample loading or retention. The sample solution is passed through the
cartridge with the object of either retaining the analytes of interest whilst the
matrix components pass through or retaining the matrix components whilst
the analytes pass through. In some procedures, the analyte(s) and one or
more of the matrix components are retained whilst the remainder of the
matrix components pass through.
● Rinsing. This is necessary to remove all those components not retained by
the sorbent during the retention step and which may remain trapped in the
interstitial solvent.
● Elution. This final step is to recover retained analytes, otherwise the matrix- free
solution and rinsings from the second and third steps are combined for
quantitative recovery of the analytes before completion of the analysis.
Conditioning Retention
Conditioning the sorbent prior to sample Adsorbed isolate
application ensures reproducible retention Undesired matrix constituents
of the compound of interest (the isolate) Other undesired matrix components

Rinse Elution
Rinse the columns to remove undesired Undesired components remain
matrix components Purified and concentrated isolate ready
for analysis
Fig. 2. Diagrammatic representation of a cartridge-based solid phase extraction procedure.

SPE can be semi- or fully-automated to increase sample throughput and to improve both
precision and accuracy. The degree of automation ranges from the parallel off-line
processing of batches of up to about 10 samples using a vacuum manifold to provide
suction, to on-line autosamplers, xyz liquid handlers and robotic workstations.
Several alternative formats for SPE are available. These include

● Disks, which have relatively large cross-sectional areas compared to packed- bed
cartridges and thin sorbent layers (0.5–1 mm thick) containing about 15 mg of material.
This reduced bed-mass results in low void volumes (as little as 15 ml) thus minimizing
solvent consumption in the rinsing and elution steps, improving selectivity and
facilitating high solvent flow rates when large volumes of sample are to be processed.
● Plastic pipette-tips incorporating small sorbent beds designed for processing

pg. 117
very small volumes of sample and solvents rapidly, and having the advan-
tage of allowing flow in both directions if required.
● Well plates containing upwards of 96 individual miniature sample-
containers in a rectangular array and fitted with miniature SPE packed beds
or disks. Well plates are used in xyz liquid handlers for processing large
numbers of samples prior to the transfer of aliquots to analytical instruments,
particularly gas and liquid chromatographs and mass spectrometers.
● Solid-phase microextraction (SPME) is an important variation of SPE that
allows trace and ultra-trace levels of analytes in liquid or gaseous samples to
be concentrated. The sorbent is a thin layer of a polymeric substance such as
polydimethylsiloxane (PDMS) coated onto a fused-silica optical fibre about
1 cm long and attached to a modified microsyringe (Fig. 3). The fibre is
exposed to the sample and then inserted directly into the injection port of a
gas or liquid chromatograph to complete the analysis. An advantage of SPME
over SPE is the avoidance of solvents, but good precision for quantita- tive
determinations is more difficult to achieve, and automated systems are only
just being developed. SPME is finding particular use in water analysis, the
analysis of fragrances and volatiles in foodstuffs by headspace sampling
(Topic D5), the detection of drugs and their metabolites in urine, blood and
breath samples, and the monitoring of air quality in working environments.

Polymer coated Septum piercing

fused silica fiber needle Barrel Plunger

Stainless steel Viewing Z-slot

tube Adjustable needle window
guide/depth gauge

Fig. 3. Diagram of a solid phase microextraction device.

PRINCIPLES OF CHROMATOGRAPHY

Key Notes

Chromatographic Chromatography is the process of separating the components of mixtures

separations (solutes) that are distributed between a stationary phase and a flowing
mobile phase according to the rate at which they are transported through
the stationary phase.

Solute migration Solutes migrate through a stationary phase at rates determined by their
and retention relative affinities for each phase, and are characterized by defined
retention parameters.

Sorption processes During a chromatographic separation, solute species are continually

transferred back and forth between the mobile and stationary phases by
the process of sorption followed by desorption. Several mechanisms by
which this occurs give rise to different modes of chromatography.

Peak profiles and Individual solutes migrating through a stationary phase develop an
peak broadening approximately symmetrical band or peak profile which becomes broader
as a function of time and distance travelled.

Peak asymmetry Peak profiles are fundamentally symmetrical but can become
asymmetrical, or skewed, as solutes migrate, due to changes in sorption
behavior.

Efficiency and The quality of a chromatographic separation is measured in terms of the

resolution degree of band broadening, or efficiency, measured for individual peaks,
and the degree of separation, or resolution, of adjacent peaks.

Qualitative and Chromatographic analysis can provide qualitative information in terms

quantitative analysis of characteristic retention parameters and quantitative information in
terms of peak areas or peak heights.

Related topics Thin-layer chromatography (D3) High-performance liquid

Gas chromatography: principles and chromatography: principles
instrumentation (D4) and instrumentation (D6)

Chromatographic Chromatography was originally developed by the Russian botanist Michael

separations Tswett in 1903 for the separation of colored plant pigments by percolating a
petroleum ether extract through a glass column packed with powdered calcium
carbonate. It is now, in general, the most widely used separation technique in
analytical chemistry having developed into a number of related but quite
different forms that enable the components of complex mixtures of organic or
inorganic components to be separated and quantified. A chromatographic

pg. 119
separation involves the placing of a sample onto a liquid or solid stationary
phase and passing a liquid or gaseous mobile phase through or over it, a
process known as elution. Sample components, or solutes, whose distribution
ratios (vide infra) between the two phases differ will migrate (be eluted) at
different rates, and this differential rate of migration will lead to their separa-
tion over a period of time and distance.
Chromatographic techniques can be classified according to whether the sepa-
ration takes place on a planar surface or in a column. They can be further sub-
divided into gas and liquid chromatography, and by the physical form, solid or
liquid, of the stationary phase and the nature of the interactions of solutes with
it, known as sorption mechanisms (vide infra). Table 1 lists the most important
forms of chromatography, each based on different combinations of stationary and
mobile phases and instrumental or other requirements.
Paper chromatography (PC) is simple and cheap but lacks the separating
power and versatility of thin-layer chromatography (TLC) which has largely
replaced it. Both require only inexpensive equipment and reagents, and, unlike
the various forms of column chromatography, comparisons can be made
between a number of samples and standards chromatographed simultaneously.
Gas (GC) and high performance liquid chromatography (HPLC) are comple-
mentary techniques best suited to the separation of volatile and nonvolatile
mixtures, respectively. Both these techniques are instrumentally-based and
computer-controlled, with sophisticated software packages and the ability to
separate very complex mixtures of up to 100 or more components. HPLC is
particularly versatile, having several alternative modes suited to different types
of solute. For example, ion-exchange (IEC) and ion chromatography (IC) are

Table 1. A classification of the principal chromatographic techniques

Technique Stationary Mobile Format Principal
phase phase sorption
mechanism
Paper chromatography (PC) Paper (cellulose) Liquid Planar Partition (adsorption,
ion-exchange,
exclusion)
Thin-layer chromatography (TLC) Silica, cellulose, Liquid Planar Adsorption (partition,
ion-exchange resin, ion-exchange,
controlled porosity exclusion)
solid

Gas chromatography (GC)

Gas-liquid chromatography (GLC) Liquid Gas Column Partition
Gas-solid chromatography (GSC) Solid Gas Column Adsorption
Liquid chromatography (LC)
High-performance liquid Solid or bonded-phase Liquid Column Modified partition
chromatography (HPLC) (adsorption)
Size-exclusion chromatography (SEC) Controlled porosity Liquid Column Exclusion
solid
Ion-exchange chromatography (IEC) Ion-exchange resin or Liquid Column Ion-exchange
Ion chromatography (IC) bonded-phase
Chiral chromatography (CC) Solid chiral selector Liquid Column Selective adsorption
modes that enable mixtures of either anionic or cationic solutes to be separated. Size-
exclusion (SEC) and chiral chromatography (CC) are two additional modes used for
separating mixtures of high relative molecular mass solutes and enantiomers, respectively.
Earlier forms of liquid chromatography, used for relatively large scale separa- tions and
known as classical column chromatography, are based on large glass columns through
which the mobile phase flows by gravity compared to the pressurized systems used in
HPLC.

Solute migration The rate of migration of a solute through a stationary phase is determined by its
and retention distribution ratio, D, which in turn is determined by its relative affinity for the two phases.
In the context of chromatography, D is defined as the ratio of the total solute concentration,
CS, in the stationary phase to that in the mobile phase, CM, i.e.
CS
D= — (1)
CM

Thus, large values of D lead to slow solute migration, and small values of D lead to rapid
solute migration. Solutes are eluted in order of increasing distribu- tion ratio. The larger
the differences between the distribution ratios of the solutes in a mixture, the more easily
and quickly they can be separated. Because the interaction of solutes with the stationary
phase slows down their rate of migration relative to the velocity of the mobile phase, the
process is described as retardation or retention.

● Column separations (GC and LC). For the separation of mixtures on columns, solutes
are characterized by the length of time they take to pass through, i.e. their retention time,
tR, or by a retention factor, k, that is directly proportional to D. The retention time and
the retention factor are related by the expression
tR = tM(1 + k) (2)

where t M (sometimes written as t 0 and known as the dead time) is the time taken by a
nonretained solute to pass through the column. A nonretained solute migrates at the
same velocity as the mobile phase, having a distribu- tion ratio and retention factor of
zero; hence t R tM. Solutes whose D and k values are greater than zero are
proportionately retarded, having retention times longer than t M, e.g.

if k = 1, tR = 2 tM
if k = 2, tR = 3 tM etc.

Chromatographic conditions are generally adjusted so that k values are less than
about 20, otherwise retention times become unacceptably long. Practical values of k are
easily calculated using a re-arranged equation (2)
(tR tM)
k — (3)
tM

For size-exclusion chromatography (Topic D7), the retention volume, VR, is used to
characterize solutes. This is the volume of mobile phase required toelute the solute from
the column. Retention times are directly proportional to retention volumes at a
constant flow rate, so equation (2) can be re-written as

pg. 121
VS
(
VR = V M 1 + D—
VM ) (4)

or VR = VM + DVs (5)
where D has been substituted for k using the relation
VS
k=D —
VM ( ) (6)

and the volumes of the stationary and mobile phases in the column are VS
and VM, respectively.
Equation (5) is regarded as a fundamental equation of column chromato-
graphy as it relates the retention volume of a solute to its distribution ratio.
● Planar separations (PC and TLC). Separations are normally halted before the
mobile phase has travelled completely across the surface, and solutes are char-
acterized by the distance they have migrated relative to the leading edge of the
mobile phase (solvent front). A solute retardation factor, Rf, is defined as
1
Rf = — (7)
1+k

The maximum value of Rf is 1, which is observed for a solute having a distri-

bution ratio and retention factor of zero, and therefore migrating at the same
velocity as the mobile phase. Solutes whose D and k values are greater than
zero are proportionately retarded, the minimum value for Rf being zero,
which is observed when the solute spends all of the time in the stationary
phase and remains in its original position on the surface. Practical values of
Rf are evaluated from the ratio
Rf = (distance moved by solute) / (distance moved by the solvent front)

Sorption Sorption is the process whereby solute species are transferred from the mobile
processes to the stationary phase, desorption being the reverse process. These processes
occur continually throughout a chromatographic separation, and the system is
therefore described as being in a state of dynamic equilibrium. A solute is
repeatedly re-distributed between the phases as the mobile phase advances, in
an attempt to maintain an equilibrium corresponding to its distribution ratio, D.
There are four basic sorption mechanisms, and it is common for two or more
to be involved simultaneously in a particular mode of chromatography, viz:
adsorption; partition; ion-exchange; exclusion.
● Adsorption is a surface effect, not to be confused with absorption, which is
a bulk effect. Surface adsorption involves electrostatic interactions such as
hydrogen-bonding, dipole–dipole and dipole-induced dipole attractions.
Solute species compete with the mobile phase for a limited number of polar
sites on the surface of the adsorbent of which silica gel is the most widely
used. Its surface comprises Si-O-Si and Si-OH (silanol) groups, the latter
being slightly acidic as well as being polar, which readily form hydrogen
bonds with slightly-polar to very-polar solutes. Water in the atmosphere can
de-activate an adsorbent surface by itself being adsorbed, thereby blocking
adsorption sites. This can be overcome by drying the adsorbent if a more
active material is required, although reproducibility may be difficult to
achieve unless ambient temperature and humidity are carefully controlled.
Some common adsorbents are listed in Table 2. Suitable
Table 2. Adsorbents for chromatographic separations
(listed in order of decreasing polarity)
Alumina (most polar)
Charcoal
Silica gel
Molecular sieve
Magnesium silicate
Cellulose
Polymeric resins (least polar)
(styrene/divinyl benzene)

mobile phases for TLC and HPLC are to be found listed in Table 1, Topic D6.
The more polar the solute, the more tenaciously it will be adsorbed onto
the surface of an adsorbent. Nonpolar solutes (e.g. saturated hydrocarbons)
have little or no affinity for polar adsorbents, whilst polarizable solutes (e.g.
unsaturated hydrocarbons) have weak affinities arising from dipole/induced
dipole interactions. Polar solutes, especially those capable of hydrogen-
bonding, are adsorbed strongly and require fairly polar mobile phases to
elute them. An approximate order of increasing polarity and therefore order
of elution (increasing distribution ratio) among classes of organic solutes is
alkanes alkenes aromatics ethers esters ketones and
aldehydes thiols amines and amides alcohols phenols acids
Adsorption-based chromatography is particularly useful for separating mixtures
of positional isomers such as 1,2-, 1,3- and 1,4-di-substituted aromatic
compounds with polar substituents, whereas members of a homolo- gous series
have similar polarities and cannot generally be separated at all. Chiral
chromatography, a mode of HPLC, depends on differences in the adsorptive
interactions of two or more enantiomers of a compound with a chiral stationary
phase. Gas solid chromatography (GSC) is an adsorption- based mode of GC.
● Partition is a sorption process analogous to solvent extraction (Topic D1),
the liquid stationary phase being thinly coated or chemically bonded onto
an inert solid. Where the liquid is bonded to the supporting solid, it is
debateable as to whether it behaves as a liquid and whether the sorption
process should be described as modified partition, because adsorption may
also be involved. In true partition, solutes are distributed according to their
relative solubilities in the mobile and stationary phases, but the exact
mechanism for bonded phases is not clear. The use of bonded phases has
become widespread in all forms of chromatography, and a pure partition
mechanism probably occurs only in gas liquid chromatography (GLC)
where the stationary phase is not chemically bonded to the column wall
(Topic D4).
Bonded phases are described in more detail in Topics D4 and D6.
● Ion-exchange is a process whereby solute ions in the mobile phase can
exchange with counter-ions carrying the same charge and associated with
oppositely charged groups chemically bound to the stationary phase. The
stationary phase is a permeable polymeric solid, such as an insoluble organic
resin or a chemically modified silica, containing fixed charge groups and

pg. 123
mobile counter-ions. Both cationic and anionic ion-exchangers are available, the
exchange processes being represented by the following equations
cation-exchange: nR-H + Xn = (R-)nXn + nH
anion-exchange: nR Cl- + Yn- = (R )nYn- + nCl
where R represents the polymeric resin or silica, and Xn and Yn are solute
cations and anions respectively of valency n.
The factors affecting ion-exchange equilibria and selectivity are described
in Topic D7.
● Exclusion differs from the other sorption mechanisms in that no specific inter-
actions between solute species and the stationary phase are necessary or desir-
able. The stationary phase is a controlled-porosity silica or polymer gel with a
range of pore sizes, and solutes remain in the mobile phase throughout the
separation, merely diffusing through the porous structure to different extents
depending on their size and shape. Solutes whose size exceeds the diameter of
the largest pores are entirely excluded from the structure and migrate at the
same rate as the mobile phase. Solutes smaller than the diameter of the smallest
pores can diffuse throughout the structure and have the slowest rate of migra-
tion. Solutes of an intermediate size can diffuse through some pores but not
others, migrating at rates between those of the largest and smallest species.
Size-exclusion chromatography (SEC) is a mode of HPLC (Topic D7) and
is also a classical technique employing large columns of silica or polymericgel
particles and gravity flow of the mobile phase. It is sometimes described as
gel permeation or gel filtration chromatography.

Peak profiles and During a chromatographic separation, individual solutes develop a symmetrical
band broadening or Gaussian concentration profile (Topic B2) in the direction of flow of the
mobile phase. The profiles, known as bands or peaks, gradually broaden and
often become asymmetrical as the solutes continue to migrate through the
stationary phase. The principal underlying reasons accounting for the peak
shapes and the observed broadening can be summarized as follows:
● continual sorption and desorption of a solute between a mobile and a
stationary phase inherently produces a Gaussian concentration profile which
broadens as the solute migrates further. (This can be demonstrated by a math-
ematical treatment of a solvent extraction procedure to separate mixtures,
developed in 1952, and known as Craig Countercurrent Distribution);
● solute species travel slightly different total distances through a particulate
stationary phase, causing concentration profiles to broaden symmetrically, this
being known as the multiple-path effect;
● solute species spread by diffusion in all directions when they are in the
mobile phase. Diffusion in both the direction of flow of the mobile phase and
directly counter to it (longitudinal or axial diffusion) contributes to the
symmetrical broadening of the peak profile;
● sorption and desorption, or mass transfer, between the stationary and mobile
phases, are not instantaneous processes, and are sometimes kineti- cally slow.
Because the mobile phase moves continuously, a true equilibrium distribution of
a solute is never established, and the concentration profile in the stationary
phase lags slightly behind that in the mobile phase causing further peak
broadening. Slow desorption can also result in the peak becoming
asymmetrical or skewed (vide infra);
● variations in the distribution ratio of a solute with its total concentration also leads to
asymmetrical or skewed peaks (vide infra).

Figure 1 illustrates the symmetrical nature of a chromatographic peak and symmetrical

broadening. Figure 2 illustrates the mutiple-path, longitudinal diffusion and mass transfer
effects.

Flow

Mobile phase

Interface

Stationary phase

Equilibrium concentration
Actual concentration
Fig. 1. The symmetrical nature and broadening of a chromatographic peak.

(a) Stationary phase particles

Mobile phase

(b) Concentration
profile of band Forward and backward
diffusion in mobile phase
as band moves along

Mobile phase

Analyte band

(c) Mobile phase

Movement off SP Movement onto SP

Stationary
phase (SP) Analyte attracted onto SP

Fig. 2. Illustration of the three principal causes of band broadening: (a) multiple-path effect; (b) longitudinal diffusion
effect; (c) mass-transfer (non-equilibrium) effect. Reproduced from A. Braithwaite & F.J. Smith, Chromatographic
Methods, 5th edn, 1996, first published by Blackie Academic & Professional.

pg. 125
Peak asymmetry The concentration profile of a migrating solute is fundamentally symmetrical
(Gaussian), only if the solute distribution ratio, D (defined by eqn. (1)), remains
constant over the concentration range of the entire peak, as shown by a linear
sorption isotherm, which is a plot of the solute concentration in the stationary
phase, CS, against that in the mobile phase, CM (Fig. 3(a)). However, curved
isotherms, resulting from changes in the solute distribution ratio at higher
concentration levels, lead to two types of peak asymmetry, or skew, described
as tailing and fronting.
Both tailing and fronting are undesirable, as closely eluting peaks will be less
well separated and retention data less reproducible. Where either occur,
reducing the amount of solute chromatographed will generally improve the peak
shape, but slow de-sorption may still cause some tailing.

(a) Linear (b) Convex (c) Concave

CS CS CS

CM CM CM

Gaussian Tailing Fronting

t t t

Fig. 3. Sorption isotherms and the resulting peak profiles. (a) Linear isotherm, symmetrical
peak. (b) Curvature towards the CM axis, tailing peak. (c) Curvature towards the C S axis,
fronting peak.

Efficiency and Two means of assessing the quality of a chromatographic separation are to measure the
resolution extent of band broadening of individual solute peaks (efficiency) and the degree of
separation of adjacent peaks (resolution).
For column chromatography, a plate number, N (based on the theoretical plate concept
of distillation columns), is used as a measure of efficiency. Assuming a Gaussian peak
profile, N is defined in terms of the solute retention
time, t R, and the peak width as given by the standard deviation, st (Fig. 4), i.e.

tR 2
N = (— ) (8)
st
In practice, it is much easier to measure either the baseline width of a peak, Wb, or the
width at half height, Wh/2, and two alternative expressions derived from equation (8) are
tR 2

(
N = 16 —
Wb ) (9)

and
tR

Concentration axis
Peak height
Injection
Wh/2

One-half
peak height
Wb
(4s)
Time axis

Fig. 4. Measurement of chromatographic efficiency from a Gaussian peak.

N = 5.54 tR 2
(— (10)
Wh/ )
2

Some laboratories favor the use of eqn. (9), but some favour eqn. (10) on the grounds
that peak width at half height can be measured with greater accuracy than the base width.
To make valid comparisons, the same formula should always be used.
An alternative measure of efficiency, which is independent of the length of a
chromatographic column, is the plate height, H (or Height Equivalent of a Theoretical
Plate, HETP), and given by
L
H= — (11)
N

where L is the column length, usually expressed in millimetres or centimetres.

NB N is a dimensionless number and to ensure correct computation, the units of tR and
Wb or Wh/2 must be the same. Most chromatography data systems have software to perform
the calculations. Plate numbers for solutes separated by GC and HPLC are of the order of
several thousands to several hundreds of thousands.
Columns giving very high plate numbers are capable of separating multicom-ponent
mixtures, but it is their resolving power, as measured by the resolution, RS, that is of prime
importance. This is defined as the difference between the retention times of two adjacent
solute peaks, t R, divided by their average base- widths, (W1 + W2)/2 (Fig. 5)
2▼tR
RS = ——(W1 + W2)
(12)

As in the case of the efficiency formulae, Gaussian peaks are assumed and all
measurements must be in the same units as RS is a dimensionless number. Values of RS
approaching or exceeding 1.5, which is defined as baseline resolu- tion, are deemed
satisfactory for most purposes.
The following is an example of the calculation of efficiency and resolution using the
above formulae:

pg. 127
tR,1 tR, 2

Concentration axis

W1 W2

Time axis

Fig. 5. Measurement of the resolution of two adjacent Gaussian peaks.

Example. A gas chromatographic method for the separation of a mixture of

cyclohexane, t-butanol and benzene on a 10 m capillary column gave the follow-
ing data:

Parameter Cyclohexane t-Butanol Benzene

tR 3 m 20 s 3 m 30 s 3 m 45 s
Wb 8s 9s 11 s
Wh/2 4.6 s 5.1 s 6.2 s

Calculate (a) the plate numbers, using both plate number formulae, for each
solute, (b) the plate heights, and (c) the resolution between adjacent pairs of solutes.
tR 2 tR 2
(a) plate numbers N = 16 —
Wb ( ) N = 5.54 —
Wh/2 ( )
( ) = 10 000
200
( ) = 10 473
200
2 2
cyclohexane N = 16 — N = 5.54 —
8 4.6

N = 16( — ) = 8711 N = 5.54( —) = 9393

210 2 210 2
t-butanol
9 5.1
N = 16(
—)
N = 5.54(
—)
225 2 225 2
benzene = 6694 = 7296
11 6.2
10 000
(b) plate heights H=—
N

cyclohexane H = 1.0 mm H = 0.95 mm

t-butanol H = 1.15 mm H = 1.06 mm
benzene H = 1.49 mm H = 1.37 mm
2▼tR
(c) resolution RS = ——
(W1 + W 2)
2▼tR
cyclohexane/t-butanol —— = 20/17 = 1.2
RS = (W
1 + W 2)
2▼tR
t-butanol/benzene RS = —— = 30/20 = 1.5
(W1 + W 2)

Note that (i) The plate numbers are slightly higher when half-height peak widths
are used to calculate them due to peak tailing increasing W b,
hence comparisons of efficiencies are valid only if the same
formula is used throughout.
(ii) The cyclohexane and t-butanol peaks are not fully resolved (RS
1.2), but the t-butanol and benzene peaks have baseline
resolution (RS 1.5).

Qualitative and There are three approaches to qualitative chromatographic analysis, viz
quantitative
● Comparison of retention data for unknown solutes with corresponding data for
analysis
standards (known substances) obtained under identical conditions.
For planar chromatography (PC and TLC), retardation factors, Rf values,
for standards and unknowns are compared by chromatographing them
simultaneously so as to eliminate variations in laboratory materials and
conditions. For column separations, retention times, tR, or volumes, VR, are
compared by chromatographing standards and unknowns sequentially under
stable conditions with as little time between runs as possible.
● Spiking samples with known solutes.
For column separations where samples are known to contain certain
solutes, a comparison is made between two or more chromatograms run
under identical conditions. The first is of the original sample, and subsequent
ones are obtained after adding a spike of one of the known solutes. Any peak
in the original chromatogram that is of increased size in a subsequent one can
then be identified as the corresponding spiked solute. However,
unambiguous identifications may not be possible.
● Interfacing the chromatograph with a spectrometer (Section F).
For column separations, this provides spectral information for each sepa-
rated solute in addition to retention data. Spectra of unknown solutes can be
compared with those in computerized library databases or interpreted
manually, even when pure standards are not available.
Comparisons of retention data alone are not always reliable as many
substances can have identical chromatographic behavior. Two or more
comparisons made under different chromatographic conditions, e.g. different
stationary or mobile phases, reduce the chances of making an incorrect identifi-
cation.
For quantitative chromatographic analysis, in addition to ensuring stable
and reproducible conditions for sample preparation and chromatography,
further specific requirements must also be met viz:
● the analyte (solute) must be identified and completely separated from other
components in the chromatogram;
● standards of known purity must be available;
● a recognized calibration procedure must be used.
For planar chromatography, solute spot areas or densities can be measured in
situ, or the solute spots can be removed, dissolved and measurements made by
another analytical technique such as UV spectrometry (Topic E9).
For column separations, quantitation can be by peak area, peak height or

pg. 129
peak height ¥ retention time. Integrated peak areas are directly proportional to
the amount of analyte chromatographed when working within the linear range
of the detector, and are the most reliable. They can be measured by computing-
integrators, by triangulation (1/2-base ¥ height of the triangle that approximates
to the Gaussian peak profile) or by cutting out and weighing peaks drawn by a
chart recorder. Detector responses for an analyte are established by the prepara-
tion of a calibration graph using chromatographed standards, with or without
the addition of an internal standard, by standard addition or by internal
normalization. These calibration procedures are described in Topics A5 and B4.
THIN-LAYER CHROMATOGRAPHY

Key Notes

Principles and Thin-layer chromatography is a technique where the components of

procedures mixtures separate by differential migration through a planar bed of a
stationary phase, the mobile phase flowing by virtue of capillary forces.
The solutes are detected in situ on the surface of the thin-layer plate by
visualizing reagents after the chromatography has been completed.

Stationary phase A variety of finely-divided particulate sorbents are used as thin-layer

stationary phases. These include silica-gel, cellulose powder, ion-
exchange resins, restricted pore-size materials, and chiral selectors.

Mobile phase Single solvents or blends of two or more solvents having the appropriate
overall polarity necessary to achieve the required separation are used as
mobile phases. They range from nonpolar hydrocarbons to polar
alcohols, water, and acidic or basic solvents.

Solute detection Methods of visualizing solutes include spraying the surface of the thin-
layer plate with a chromogenic reagent, or viewing it under a UV lamp if
the sorbent has been treated with a fluorescent indicator.

Alternative TLC Alternative development procedures aimed at improving

procedures chromatographic performance have been introduced, and new stationary
phases are becoming available.

Applications of TLC Thin-layer chromatography is used primarily as a qualitative analytical

technique for the identification of organic and inorganic solutes by
comparisons of samples with standards chromatographed
simultaneously. Ǫuantitative analysis is possible but precision is
relatively poor.

Related topic Principles of chromatography (D2)

Principles and Thin-layer chromatography is a form of planar chromatography similar to

procedures paper chromatography, but the stationary phase is a finely-divided sorbent
spread as a thin layer on a supporting flat plastic, aluminum or glass plate.
Solutes migrate through the stationary phase at rates determined by their distri-
bution ratios (Topic D2), those with the largest values moving the least, if at all,
whilst those with the smallest values moving with the advancing mobile phase,
or solvent front. A typical TLC procedure consists of the following steps:
● sufficient mobile phase to provide about a 0.5 cm depth of liquid is poured
into a development tank, or chamber, which is then covered and allowed to

pg. 131
stand for several minutes to allow the atmosphere in the tank to become
saturated with the solvent vapor;
● small volumes of liquid samples and standards, or solutions, are spotted onto
the sorbent surface of a TLC plate along a line close to and parallel with one
edge (the origin). The plate is then positioned in the tank with this edge
in contact with the mobile phase and the cover replaced (Fig. 1(a));
● the mobile phase is drawn through the bed of sorbent from the edge of the
plate, principally by capillary action, and this development process is halted
shortly before the solvent front reaches the opposite side of the plate. Sample
components and standards migrate in parallel paths in the direction of flow
of the mobile phase, separating into discrete zones or spots;
● the plate is removed from the development tank, dried in a current of warm
air, and solute spots located by appropriate methods (vide infra);
● each solute is characterized by the distance migrated relative to the solvent
front, i.e. its Rf value, which will lie between 0 and 1 (Topic D2), and
unknowns are identified by comparisons with standards run simultaneously.
Figure 1(b) illustrates a developed and visualized TLC plate with Rf values
shown alongside. Note that the shapes of some spots have become slightly
elongated in the direction of flow of the mobile phase, which is an example of
tailing (Topic D2). This is caused by slow desorption as the solute migrates, or
saturation of adsorption sites by high concentrations of the solute leading to a
convex sorption isotherm, and is most likely to occur where adsorption is the
principal chromatographic sorption mechanism (Topic D2).

Stationary phase Stationary phases used in TLC are microparticulate sorbents with particle diameters
of between 10 and 30 mm. The smaller the mean particle size and the narrower the
size range, the better the chromatographic performance in terms of band spreading
(efficiency) and resolution (Topic D2). Thin-layer chromatog- raphy plates are
prepared by coating sorbents onto rectangular plastic, aluminum or glass sheets
in adherent uniform layers approximately 250 mm

(b)

Rf
1.00
(a)
Lid

TLC plate
Developing 0.59
coated with
absorbent tank

Origin line Sample spots 0.27

Mobile phase

Fig. 1. TLC plates (a) during and (b) after development and visualization; Rf values are 0
shown alongside. a, Reproduced from R.J. Hamilton & S. Hamilton, Thin-Layer
Chromatography, 1987. b, Reproduced from R.M. Smith, Gas and Liquid Chromatography in
Analytical Chemistry, 1988. © John Wiley & Sons Ltd. Reproduced with permission.
thick. Commercially produced plates are available in several sizes between 5 cm
and 20 cm square and may incorporate an insoluble fluorescent reagent to facili-
tate the detection of solute spots (vide infra). The most commonly used sorbents
are silica and powdered cellulose, and the corresponding sorption mechanisms are
adsorption and partition, respectively. Thin layers can also be made of
chemically-modified silicas, ion-exchange resins, exclusion gels and cyclo-
dextrins that display chiral selectivity. Some of these sorbents are similar to the
bonded phases used in HPLC which are discussed in Topics D6 and D7. Some
TLC sorbents are listed in Table 1.

Table 1. Stationary phases (sorbents) for thin-layer chromatography

Sorbents Chromatographic mechanism Typical applications

Silica gels Adsorption Amino acids, hydrocarbons, alkaloids,
vitamins
Hydrocarbon modified silicas Modified partition Nonpolar compounds
Cellulose powder Partition Amino acids, nucleotides, carbohydrates
Alumina Adsorption Hydrocarbons, alkaloids, food dyes,
lipids, metal ions
Kieselguhrs (diatomaceous earths) Partition Sugars, fatty acids
Ion-exchange celluloses Ion-exchange Nucleic acids, nucleotides, halide and
metal ions
Sephadex gels Exclusion Polymers, proteins, metal complexes
b-Cyclodextrins Stereo-adsorptive interactions Mixtures of enantiomers

Mobile phase The range of mobile phases used in TLC is extremely wide and they are often
selected empirically. Blends of two solvents are common because the solvent
strength, or eluting power, can be easily adjusted to optimize a separation by
altering solute distribution ratios. Some general guidelines in selecting and
optimizing the composition of a mobile phase are
● Solvents should be of the highest purity as TLC is a very sensitive analytical
technique;
● Mobile phase eluting power should be adjusted so that solute Rf values fall
between 0.2 and 0.8 so as to maximize resolution;
● For separations on silica gel and other polar adsorbents, the overall polarity
of the mobile phase determines solute migration rates and hence their Rf
values; small additions of a slightly polar solvent, such as diethyl ether, to a
nonpolar solvent, such as methylbenzene, will significantly increase Rf values;
● Polar and ionic solutes are best separated using a blend of a polar organic
solvent, such as n-butanol, with water; the addition of small amounts of
ethanoic acid or ammonia to the water increases the solubilities of basic and
acidic solutes, respectively.
A helpful guide to solvent strength in adsorption and partition-based separa-
tions is an eluotropic series, an example of which is given for HPLC in Table 1
of Topic D6. The solvents are listed in order of increasing solvent strength for
adsorption-based separations. The order for partition-based separations is
broadly similar.

pg. 133
Solute detection Solutes separated by TLC remain on the surface of the plate after development.
As the majority of solutes are colorless, their spots must be located using a chemical
or physical means of visualization. Chemical methods include:
● Spraying the plate with a locating, or chromogenic reagent that reacts chemi-
cally with all solutes or those containing specific functional groups or struc-
tural features to give colored spots. The plates sometimes need to be warmed
to accelerate the chromogenic reaction and intensify the spots. Examples are
given in Table 2;
● Viewing the plate under a UV lamp set at an emission wavelength of 254 nm
or 370 nm to reveal the solutes as dark spots or bright fluorescent spots on a
uniform background fluorescence. Commercial plates can be purchased with
an insoluble fluorescent substance incorporated into the stationary phase to
provide the background fluorescence, or the plate can be sprayed with a
fluorescent reagent after development;
● Spraying the plate with concentrated sulphuric or nitric acid and heating to
oxidize and char organic solutes which are revealed as brown to black spots;
● Exposing the plate to iodine vapor in a sealed chamber when many organic
solutes develop dark brown colorations;
● Scanning across the surface of the plate with a densitometer, an instrument
that measures the intensity of radiation reflected from the surface when irra-
diated with a UV or visible lamp. Solutes that absorb radiation are recorded
as peaks on a recorder trace or VDU screen;
● Radiolabelled solutes can be detected by autoradiography (blackening of a
photographic film sensitive to X-rays), liquid scintillation counting on
scraped-off areas of the stationary phase, or monitoring the surface of the
plate with a Geiger-Müller tube.

Table 2. Some examples of TLC locating reagents

Method of detection Color of solute spots Application

General reagents
Phosphomolybdic acid heat Dark blue Many organics
Conc. sulphuric acid heat Charred brown-black All organics
Iodine vapor Brown Many organics
Selective reagents
Ninhydrin Pink to purple Amino acids and amines
2,4-Dinitrophenylhydrazone Orange/red Carbonyl compounds
Bromocresol green/ blue Yellow Organic acids
2,7-Fluorescein Yellow-green Most organics
Vanillin/ sulphuric acid Blue, green, pink Alcohols, ketones
Rhodamine-B Red fluorescence Lipids
Anisaldehyde/antimony trichloride Various Steroids
Diphenylamine/zinc Various Pesticides

Alternative TLC Variations of the basic development procedure and new stationary phases aimed
procedures at improving resolution, sensitivity, speed, reproducibility and selectivity have
been developed from time to time. Two of the more significant and useful ones
are:
● Two-dimensional TLC. This is a means of improving resolution in samples
where the component solutes have similar chemical characteristics and hence
Rf values, e.g. amino acids. A single sample is spotted close to one corner of the plate
and, after development, there is a partial separation of the solutes along one edge. The
plate is dried, turned through 90 and developed a second time, but using a different
mobile phase composition for which the solutes have different distribution ratios compared
to those for the first mobile phase. This results in the partially-separated solutes
separating further because of changes to their migration rates. Visualization gives a 2-D
map or fingerprint of the sample components. The procedure for the separa- tion of 14
amino acids is shown diagrammatically in Figure 2.
● High-performance plates (HPTLC). These are coated with a thinner layer
(100 mm thick) of a 5 mm particle diameter stationary phase with a very narrow range
of sizes. Sensitivity and resolution are improved because solute spots are compact,
development is much faster, partly because smaller plates can be used, and solvent
consumption is reduced. HPTLC plates with silica and bonded-phase silicas are
commercially available.

Applications of TLC is applicable to a very wide range of mainly organic solutes. It is used primarily in the
TLC biochemical, pharmaceutical, clinical and forensic areas for qualitative analysis by the
comparison of Rf values of sample solutes with those of standards run on the same plate. It
is especially useful for checks on purity, to monitor the course of reactions and production
processes, and to characterize complex materials such as natural products. The screening
of samples for drugs

1st Solvent front 2nd b Solvent front

solvent solvent

Sample Origin Turn 90 Origin a

After 1st development After 2nd development

LEU

VAL Ph-ALA
-ABA Solvent 1 n-butanol/acetic
GLU acid/water (12:3:5)
GLY THR ALA PRO-OH
PRO

Sample TAU
application

Solvent 2 Phenol/H2O

Fig. 2. Two-dimensional TLC of a mixture of 14 amino acids. Top panel, reproduced from R.J. Hamilton & S. Hamilton,
Thin-Layer Chromatography, 1987. © John Wiley & Sons Ltd. Reproduced with permission. Bottom panel, reproduced from
A. Braithwaite & F.J. Smith, Chromatographic Methods, 5th edn, 1996, first published by Blackie Academic & Professional.

pg. 135
in clinical and forensic studies, and testing for the presence or absence of specific
substances (limit tests) are additional applications. Tables 1 and 2 include further
examples.
Quantitative TLC by the measurement of spot areas or by computerized
scanning reflectance densitometry (Topic E8) is possible, but, unless HPTLC
plates are used, the relative precision attainable is generally only 5–10%. The
principal source of error is in applying sample spots to the plate, although auto-
mated systems can reduce this.
TLC has a number of advantages over GC and HPLC:
● the ability to run 10–20 or more samples simultaneously for immediate and
direct comparison with standards, which represents a considerable saving in
time;
● the basic technique is very cheap, versatile and quick;
● all solutes, including those that do not migrate from the origin, are detectable.
Disadvantages are the limited reproducibility of Rf values due primarly to
changes in the composition of the mobile phase during development of the plate,
and increasing spot diffusion (band broadening) because the flow rate of the
mobile phase slows as it travels across the plate.
Chapter eight:GAS CHROMATOGRAPHY:
PRINCIPLES AND
INSTRUMENTATION

Key Notes

Principles Gas chromatography is a technique for the separation of volatile

components of mixtures by differential migration through a column
containing a liquid or solid stationary phase. Solutes are transported
through the column by a gaseous mobile phase and are detected as they
are eluted.

Mobile phase The mobile phase is an inert gas, generally nitrogen or helium, supplied
from a cylinder via pressure and flow controls, and passing through
purification cartridges before entering the column.

Sample injection Gaseous, liquid and solid samples are introduced into the flowing mobile
phase at the top of the column through an injection port using a
microsyringe, valve or other device.

Column and Columns are either long, narrow, capillary tubes with the stationary
stationary phase phase coated onto the inside wall, or shorter, larger diameter tubes
packed with a particulate stationary phase. Stationary phases are high-
boiling liquids, waxes or solid sorbents.

Temperature control The column is enclosed in a thermostatically-controlled oven that is

maintained at a steady temperature or programmed to increase
progressively during a separation.

Solute detection Solutes are detected in the mobile phase as they are eluted from the end
of the column. The detector generates an electrical signal that can be
amplified and presented in the form of a chromatogram of solute
concentration as a function of time.

Instrument control A dedicated microcomputer is an integral part of a modern gas

and data processing chromatograph. Software packages facilitate the control and monitoring
of instrumental parameters, and the display and processing of data.

Related topics Principles of chromatography (D2) Gas chromatography: procedures

and applications (D5)

Principles Gas chromatography (GC) is a separation technique where volatile, thermally

stable solutes migrate through a column containing a stationary phase at rates
dependent on their distribution ratios (Topic D2). These are inversely propor- tional
to their volatilities, which in turn are determined by their partial vapor

pg. 137
pressures and hence their boiling points. Solutes are therefore generally eluted
in order of increasing boiling point, except where there are specific interactions with
the stationary phase. The gaseous mobile phase elutes the solutes from the end of
the column where they pass through a detector that responds to each one. An
elevated temperature, usually in the range 50–350 C, is normally employed
to ensure that the solutes have adequate volatility and are therefore eluted
reasonably quickly.
There are two modes of gas chromatography:
Gas-liquid chromatography (GLC), which employs a liquid stationary phase in
which solutes can dissolve, the sorption process being partition. Specific
interactions of solutes with the stationary phase may alter the order of elution
from that of increasing boiling points. GLC is by far and away the more widely
used mode of GC, the large number of alternative stationary phases enabling
many types of sample to be analyzed.
Gas-solid chromatography (GSC) employs a solid, sometimes polymeric,
sorbent as the stationary phase, the sorption process being surface adsorp- tion.
GSC has limited specialist applications, being used mainly for analyzing
mixtures of gases or solvents with relatively low relative molecular masses.
A schematic diagram of a gas chromatograph is shown in Figure 1. It consists of five
major components:

Microliter syringe
(or autosampler,
sample valve)
Sample
Gas flow control system

Carrier gas Detector

Septum Amplifier

Injector
Makeup flow

Capillary column

Split flow

Column oven

Chromatogram
Fig. 1. Schematic diagram of a capillary column gas chromatograph. Reproduced from D.W. Grant, Capillary Gas
Chromatography, 1996. © John Wiley & Sons Ltd. Reproduced with permission.
● gas supply and controls;
● sample injection port;
● column housed in a thermostatically-controlled oven;
● detection and recording system;
● microcomputer with control and data processing software.
These are described in the following sections.

Mobile phase The mobile phase is known as the carrier-gas because its sole purpose is to
transport solutes through the column, thus not contributing to chromatographic
selectivity. It should be inert, non toxic, non flammable and preferably cheap.
Helium and nitrogen are invariably used routinely, the former with capillary
(open tubular) columns and the latter with packed columns (vide infra). Helium
gives better chromatographic efficiency (reduced band broadening) due to faster
mass transfer (Topic D2). The carrier gas must be purified by passing it through
suitable adsorbents so as to avoid undesirable chemical changes to sample
components and stationary phases, or adverse effects on detector performance.
The most common contaminants and the means of removing them are:
● Air or oxygen at levels above about 10 ppm, which can oxidize sample
components and liquid stationary phases, especially at high column tempera-
tures. These can be removed by a cartridge containing molecular sieve.
● Hydrocarbons, which affect detector performance by contamination or
producing a large and constant background signal. These can be removed by
a cartridge containing activated carbon.
● Water vapor, which can affect some solid and bonded-liquid stationary
phases, and the performance of some detectors. This can also be removed
with molecular sieve.
The carrier gas is supplied from a cylinder via a pressure-reducing valve at 10–
45 psi (0.7–3 bar) which provides flow rates of between 1 and 50 cm3 min 1
depending on the type of column in use. A mass-flow controller ensures constant
flow rates regardless of back-pressure and temperature (the viscosity of gases
increases with temperature).

Sample injection Samples should preferably be injected into the flowing mobile phase rapidly so as
to occupy the smallest possible volume when vaporized. This ensures a narrow
initial sample band that maximizes column efficiency and resolution.
There are a number of methods of injection and designs of injection port available,
and the choice is determined by the type of column in use and the nature of the
sample. Small volumes of liquids or solutions (0.1–10 ml) are gener- ally injected
into a heated injection port, through which the carrier gas continu- ously flows, from
a calibrated microsyringe used to pierce a self-sealing silicone-rubber septum. Gases
are introduced via a gas sampling valve or gas syringe, and solids with volatile
components as solutions. Septa must be replaced regularly to avoid leakage and
can be a source of contamination by previously injected samples or the bleeding of
plasticizers into the gas stream, especially when operating at very high
temperatures. A separate septum-purge gas stream vented to the atmosphere, or
septumless valves can overcome the latter problem. Capillary (open tubular)
columns require specially designed injection ports to prevent overloading them with
samples, which can severely impair efficiency and resolution.

pg. 139
Alternative methods of sample injection are summarized below and in Table
1, with schematic diagrams in Figures 2 and 3.

Table 1. Gas chromatography sample injection systems

Type of Injection Method of injection Advantages or disadvantages
column system
Capillary Split Carrier-gas split 1 : 10 to 1 : 500 to Overloading of column avoided, but
reduce the amount of sample sensitivity reduced and may result in
entering the column by 90% or more discrimination between solutes with
different boiling points
Splitless Whole of injected sample condensed Increased sensitivity, but limited to low
on cooled top of column, then levels of solutes in sample, and may
released by heating broaden bands
On-column Sample condensed in cooled zone Increased sensitivity, minimal thermal
at top of column, then volatilized by degradation of solutes, no discrimination
programmed heating effects
Packed Flash-vaporization Sample injected into zone heated to Rapid volatilization of sample, but
20–50 C above column temperature thermal degradation of some solutes
may occur
On-column Sample injected onto top of Increased sensitivity, minimal thermal
packed bed degradation of solutes, no discrimination
effects

● Split injection is used with capillary columns to prevent overloading the

stationary phase with sample. The split-point is a centrally positioned hollow
needle that allows a small part of the injected sample (2% or less) to reach the
column and vents the remainder to the atmosphere via a controlvalve (Fig. 2).
Sensitivity is less than with splitless injection because of the

Dump valve
Injection Septum
purge
valve

Control
valve

Carrier
inlet

Adsorption tube

Fig. 2. Schematic diagram of a split/splitless injection

Capillary column port. Reproduced from A. Braithwaite & F.J. Smith,
Chromatographic Methods, 5th edn, 1996, first published
by Blackie Academic & Professional.
Syringe
Septum

Septum
retainer

Heated

Carrier gas Oven liner

Fig. 3. Schematic diagram of a flash vaporization
injection port. Reproduced from Instrumental
Methods of Analysis, 2nd edn, by H.H. Willard, L.L.
Merritt, J.A. Dean, F.A. Settle © 1988. Reprinted
Seal
with permission of Brooks/Cole, an imprint of the
Column Wadsworth Group, a division of Thomson Learning.

very small fraction of injected sample reaching the column. A slow purge
stream of gas prevents substances bleeding from the septum accumulating in
the injection port.
● Splitless injection is used with capillary columns when samples contain low
levels of some components, and maximum sensitivity is required. The
control valve is kept shut as a 0.5–5 ml volume of sample in a volatile solvent
is injected. By cooling the top of the column to just above the boiling point of
the solvent, the sample components are trapped while the solvent travels on
down the column. The control valve is then opened to purge any remaining
sample vapors from the injection port, and the column temperature is raised,
releasing the solutes into the gas stream.
● On-column injection is used with both capillary and packed columns to mini-
mize the possibility of the decomposition of thermally labile solutes as well as
providing increased sensitivity. It is also ideal for quantitative analysis.
● Flash-vaporization is primarily used with packed columns and is a means of
rapidly volatilizing samples in a zone heated 20–50 C above the column temper-
ature to provide as narrow a band of vapor as possible (Fig. 3). However, there
is a risk of thermally degrading labile solutes, although a glass liner inserted into
the metal block in which the samples are injected minimizes the risk.
● Automated sample injection is advantageous for improving precision and
for processing large numbers of samples loaded into autosampler trays. Most
injection ports can be adapted for this purpose, which is often under
computer control.
● Special injection techniques are employed for headspace analysis, pyrolysis
gas chromatography and thermal desorption to concentrate samples (Topic
D5).

The column is where the separation process occurs and it is, therefore, the
Column and
central component of a gas chromatograph. There are two types of GC column,
stationary phase
and a comparative summary is given in Table 2.
pg. 141
Table 2. A comparison of capillary and packed GC columns
Capillary columns (open tubular) Packed columns
Tubing Tubing
Fused quartz (SiO2) Stainless steel or glass
Very high purity (<1 ppm metals)
Length 10–100 m, coiled Length 1–3 m, coiled
Internal diameter 0.1–0.7 mm Internal diameter 2–3 mm
External protective coating of a polyimide or aluminum
Packing
Granular (0.125–0.25 mm), inert, silaceous solid
support for liquid stationary phase or porous solid
adsorbent

Stationary phase Stationary phase

Very thin liquid layer (0.1–5 mm) Thin, 1–10 percent w/w coating of liquid for GLC;
Coated or chemically bonded to inside wall (WCOT) must cover solid support completely
For GLC, bonded phases reduce column-bleed Porous solid adsorbent or polymer for GSC
considerably
Very thin finely-divided porous solid (PLOT) for GSC

Injection systems Injection systems

Split, splitless or on-column Flash-vaporization or on-column

Sample capacity Sample capacity

Narrow bore <<0.1 ml 0.1–20 ml
Megabore 0.1 to 10 ml

Preferred carrier gas Preferred carrier gas

Helium or hydrogen Nitrogen

Performance (Fig. 4) Performance (Fig. 4)

Very high efficiencies and resolving power for complex Limited efficiency and resolving power for up to
mixtures of up to 100 or more components, especially about 20 components
for narrow bore columns, but sample capacities limited;
solutes elute at lower temperatures than with
corresponding packed column

Price and source Price and source

Expensive, from specialist supplier Inexpensive, can be packed in the laboratory

● Capillary (open tubular) columns have become the most widely-used in recent
years. They consist of long, narrow bore, high-purity quartz tubing with a very
thin layer of a liquid or solid stationary phase (vide infra) coated or chemically
bonded to the surface of the inner wall. This minimizes band spreading because
of rapid mass transfer (Topic D2). Also, as there is an unrestricted flow of carrier
gas through the center of the column generating little back-pressure, very long
lengths (up to 100 m) can be used, resulting in extremely high efficiencies and
resolving power for the separation of complex mixtures.
Capillary columns are available with a range of internal diameters and
thicknesses of stationary phase. The narrower the bore and the thinner the
coating, the greater the efficiency, but the lower the sample capacity before
overloading causes peaks to tail and resolution to deteriorate. Special injec-
tion systems are therefore required, as described in the previous section, to
reduce the amount of sample injected with a conventional microsyringe. The
widest bore columns (megabore), which exceed 0.5 mm internal diameter,
are the least efficient, but have the highest sample capacities, so special injec-
tion techniques are not needed.
● Packed columns are much shorter than capillary columns, rarely exceeding
2 m, the length being limited by the back-pressure generated by the gas
flowing through a packed bed. The stainless steel or glass tubing has an
internal diameter of 2 or 3 mm and is filled with a granular material that acts
as a solid support for a thin coating of a liquid stationary phase for GLC, or as
an adsorbent for GSC. Solid supports are inert, porous silaceous materials such as
diatomaceous earths (kieselguhrs) with a large surface area. Their particle
sizes vary between 0.125 mm (US sieve mesh 120) and 0.25 mm (US sieve mesh
40), individual columns being packed with particles having a narrow range
between these limits to improve packing characteristics and chromatographic
efficiency. The smaller the particle size, and the thinner the coating of
stationary phase, the less solute bands spread by the multiple path and mass
transfer effects (Topic D2). Packed columns are much cheaper than capillary
columns but their overall efficiencies and resolving power are limited. They are
best suited to the separation of mixtures of up to ten or twenty components.
Examples of separations on capillary and packed columns are shown in Figure 4.

GLC stationary phases are thin coatings of very high boiling liquids, oils or
waxes, some with a polymeric structure, e.g. polysiloxanes and polyethylene glycol.
Those for GSC are solid adsorbents and polymers. They can be classified according
to their polarity, varying from nonpolar hydrocarbons to polar poly- esters,
cyanopropyl silicones and alumina. There are special phases that show particular
selectivities for specific types of solute such as fatty acids, bases and enantiomers,
and high-temperature phases based on silicone-carborane co-poly- mers. Hundreds
of stationary phases have been investigated, many having very similar
characteristics, but most laboratories use only a few for routine work.
The most important features of stationary phases are:
● They should be nonvolatile, chemically and thermally stable over a wide
temperature range, and nonreactive towards the separating solutes;
● Most liquids have a recommended operating temperature range; beneath the
lower limit, peak shapes become badly distorted because the phases solidify.
At temperatures close to or exceeding the upper limit, the liquid gradually
bleeds from the column and/or degrades, which changes the chromato-
graphic characteristics and leads to an unstable detector signal;
● Liquids chemically bonded to the walls of capillary columns bleed much
less, and the columns can be washed through with solvents to remove
strongly retained sample residues contaminating the stationary phase;
● The thinnest coatings of stationary phase give the highest efficiencies and
resolving power, but the lowest sample capacities;
● The choice of stationary phase is determined by the sample; generally, they
should have similar polarities otherwise peaks may be distorted, but
compromise choices must be made where solutes in a mixture have a wide
range of polarities;
● Elution order can be altered by changing the stationary phase where there

pg. 143
(a) Peak indentification
1. isopentane
2. n-pentane
3. cyclopentane
4. 3-methylpentane
5. n-hexane
6. 2,4-dimethylpentane

Inject
7. benzene
8. cyclohexane
9. 3-methylhexane
10. n-heptane
10

0 3 6 9 12 15 18 21 24
Time (min)

(b) Peak indentification

1. pyridine
2. 2,4,5-trimethyloxazole
3. 3-methylpyridine 89
4. 1-pentanol
5. ethylheptanoate
6. 1-hexanol
7. 1-heptanol 10
15
8. linaloate 12
9. 1-octanol
10. 1-nonanol
11. citronellol 14
12. 1-decanol 11 17
13. 3-acetylpyridine
14. anisaldehyde 13
15. cinnamaldehyde
16. -decalactone
17. anisylalcohol

16

0 40 min

Fig. 4. Examples of gas chromatographic separations on capillary and packed columns. (a) Packed column separation
of mixed hydrocarbons; (b) capillary column separation of some flavor compounds. Reproduced from D.W. Grant,
Capillary Gas Chromatography, 1996. © John Wiley & Sons Ltd. Reproduced with permission.

are specific interactions with a solute, e.g. on a nonpolar stationary phase, t-butyl
alcohol (bp 82.6 C) elutes before cyclohexane (bp 80.8 C) because the latter,
being nonpolar itself, dissolves better in the stationary phase. However, on a
polar stationary phase with hydroxy groups, the elution order is reversed
because the alcohol can H-bond to it.
Selections of stationary phases are given in Tables 3 and 4.
Table 3. A selection of stationary phases for packed columns

Stationary phase Chemical type Polarity

Apiezon L Branched-chain alkane grease Nonpolar
OV101 Dimethyl silicone Nonpolar
OV17 50% Phenyldimethyl silicone Medium polarity
Carbowax 20M Polyethylene glycol Very polar

Table 4. A selection of stationary phases for capillary columns

Stationary phase Chemical type Polarity Applications

BP1 100% methyl Nonpolar Solvents, VOCs, petroleum

}
products
BP5 5% phenyl Nonpolar Aromatics, PAHs, drugs,
95% methyl perfumes
OV1701 14% cyanopropyl polysiloxanes Medium polarity Alcohols, phenols, esters,
BP10 86% methyl ketones, pesticides
DB17 50% phenyl Medium polarity Esters, ketones, plasticizers
RT50 50% methyl
CP-Wax Polyethylene glycol Very polar Alcohols, esters, acids,
DB-Wax amines, solvents

VOCs volatile organic compounds; PAHs polyaromatic hydrocarbons.

Temperature Temperature control is essential in ensuring reproducible separations by GC. The

control column is enclosed in an insulated and thermostatically-controlled oven with a
heater and circulating fan to maintain a uniform temperature from ambient to
about 400 C. For isothermal (constant temperature) chromatog- raphy, the
selected temperature must be maintained to ±0.1 C, as solute distrib- ution ratios
are highly temperature sensitive, e.g. a 20 increase in column temperature
results in about a two-fold decrease in distribution ratio and a corresponding
decrease in retention time, tR. Temperature programming is a procedure used to
optimize the separation of complex mixtures (Topic D5).

Solute detection The carrier-gas flows through a detector that responds to changes in a bulk physical
property, such as its thermal conductivity, in the presence of a solute vapor, or to
a specific property of the eluting solutes themselves, such as their ability to be
ionized. Detectors may be universal, responding to practically all solutes, or
selective, where they respond to solutes with particular characteris- tics, such as
specific elements or structural features. Ideally, detectors should have the
following characteristics:
● a rapid and reproducible reponse to the presence of solute vapors in the
carrier gas;
● high sensitivity, i.e. able to detect very low levels of solutes;
● stablity in operation;
● a signal directly proportional to solute concentration or mass over a wide
range (wide linear dynamic range).
Although many types of GC detector have been investigated, only four are in
widespread use. Details of these are summarized below and in Table 5.

pg. 145
Table 5. Characteristics of GC detectors
Detector Sensitivity Linear Characteristics
(g s 1) range
Thermal conductivity (TCD) 10 9
104 Robust, non-destructive, flow and temperature
sensitive, poor linear dynamic range, insensitive to
inorganic solutes
Flame ionization (FID) 10 12
107 Excellent sensitivity and linear dynamic range, best
universal GC detector
Nitrogen-phosphorus (NPD) 10 14
(N) 105 Similar to FID, but selective for N and P containing
10 15
(P) 105 solutes, limited linear dynamic range

Electron capture (ECD) 10 13

103 Excellent sensitivity for solutes with electronegative
elements, temperature sensitive, easily contaminated,
limited linear dynamic range

(i) Thermal conductivity detector (TCD). This is one of the oldest types, and
is known also as a katharometer or hot-wire detector. It is a universal
detector consisting of a heated metal block containing a reference cell
through which pure carrier gas constantly flows, and a sample cell through
which carrier gas flows after emerging from the end of the column ( Fig. 5).
The cells contain identical, heated platinum filaments whose resistances
depend on their temperatures, which in turn depend on the rates of heat
loss from their surfaces, these being a function of the thermal conductivity
of the surrounding gas. When pure carrier gas flows through both cells, the
resistances of the two filaments are the same, and a Wheatstone bridge
circuit into which they are both incorporated can be balanced to give a
stable baseline or background signal. When a solute is eluted from the
column and passes through the sample cell, its presence alters the thermal
conductivity of the carrier gas. The temperature of the filament and hence
its resistance changes, and an out-of-balance signal proportional to the
solute concentration is created in the bridge circuit. When the solute has
passed through the cell, the signal returns to the baseline value.
The TCD is robust and reliable, but has only moderate sensitivity and a
limited dynamic range making it more suitable for qualitative than quanti-
tative work.

To waste To waste

Analytical Reference
filament filament
Fig. 5. Schematic diagram of a thermal
conductivity detector, TCD. Reproduced
from I.A. Fowlis, Gas Chromatography:
Analytical Chemistry by Open Learning,
Analytical Reference 2nd edn, 1995, with permission from Her
column gas stream, in Majesty’s Stationery Office. Crown
effluent, in Copyright.
(ii) Flame ionization detector (FID). This is the most important of a group of detectors
where the signal is related to the ionization of eluting solutes. Carrier gas emerging from
the column is mixed with air and hydrogen and burnt at a small metal jet (Fig. 6). A 150–
200 V DC potential is applied between the burner jet (positive) and a collector electrode
(negative) positioned just above the micro-flame. The electrodes are connected to an
external circuit where the signal can be amplified and recorded. Eluted solutes are
combusted to yield ions which increase the electrical conduc- tivity of the flame and are
collected by the negative electrode, thus allowing a current proportional to the
concentration of ions derived from the solute to flow around the external circuit.
The FID is the most widely used universal detector, being extremely
sensitive and responding to all organic solutes except formaldehyde, formic acid and
fully halogenated compounds. It has the widest linear dynamic range of all GC
detectors, making it ideal for quantitative analysis, and its only disadvantage is the
lack of response to inorganic solutes.
(iii) Nitrogen-phosphorus detector (NPD). This is a selective detector that is basically a
flame ionization detector, modified by positioning a ceramic bead containing a
rubidium or caesium salt, electrically heated to 800 C,
between the burner jet and the collector electrode. The response to nitrogen-
containing compounds is enhanced by a factor of about 50 over that of an unmodified
FID, and the response to phosphorus-containing compounds is enhanced by a factor
of about 500. The linear dynamic range is intermediate between those of the FID and
TCD.
(iv) Electron capture detector (ECD). This is another form of ionization
detector, and shows a selective response to solutes containing halogens, sulfur and
unsaturated structures, all of which have high electron affinities.

Collector electrode

Detector body
–
Flame ignition coil Polarizing voltage
(150–300 V)

Insulated jet +

Heated detector base

Fig. 6. Schematic diagram of a flame

ionization detector, FID. Reproduced from
Air H2 I.A. Fowlis, Gas Chromatography: Analytical
Chemistry by Open Learning, 1995, 2nd
edn, with permission from Her Majesty’s
Column
Stationery Office. Crown Copyright.

pg. 147
The ECD is one of the most sensitive of GC detectors, especially if argon
replaces nitrogen as the carrier gas, but traces of air, oxygen or water in the
gas, liquid stationary phase bleeding from the column or residues from
halogenated solvents used in sample preparation are detrimental to its
performance, and its linear range is very limited. Ǫuantitative analysis is
difficult because response is dependent on solute structure.

Instrument Components of a modern gas chromatograph include a dedicated microcom- puter

control and data with analogue-to-digital convertor (ADC) to digitize the detector signal, and
processing software packages that perform the following functions (see also Section H):

● Facilitate the setting and monitoring of instrument parameters, i.e.:

(i) carrier gas flow;
(ii) oven temperature and temperature program;
(iii) automated sample injection;
(iv) detector gases, mode and sensitivity.
● Display of chromatograms and other information in real time with high-
resolution color graphics, and electronic integration of peak areas (instru-
mentation may include a separate computing integrator).
● Recording and processing of retention and calibration data, calculations and
statistical assessment of results.
● Storage and retrieval of method parameters for specific analyses.
● Diagnostic testing of the condition and performance of instrumental compo-
nents.
● Communication with the database of a laboratory information and manage-
ment system (LIMS) for further data-processing and archiving of results.
GAS CHROMATOGRAPHY :
PROCEDURES AND
APPLICATIONS

Key Notes
Temperature This is a form of gradient elution whereby the temperature of the column
programming is progressively increased during a separation to optimize
chromatographic performance.

Special procedures The analysis of nonvolatile materials, volatile components of complex

used in GC mixtures, trace levels of solutes and multicomponent samples with very
large numbers of solutes require special sampling procedures.

Qualitative analysis Unknown solutes can be identified by comparisons of retention times,

spiking samples with known substances, or using retention indices. These
may be ambiguous, and more reliable information can be provided by
interfacing GC with a spectrometric technique.

Quantitative analysis Ǫuantitative information is obtained from peak area measurements and
calibration graphs using internal or external standards, or by standard
addition or internal normalization.

Related topics Principles of chromatography (D2) Gas chromatography: principles

and instrumentation (D4)

Temperature The effect of column temperature on chromatographic retention is pronounced

programming in that there is an inverse exponential relation with the distribution ratio, D,
which results in a shortening of retention times as the temperature is increased.
Temperature programming, which is a form of gradient elution, involves
raising the oven temperature progressively during a separation to improve the
resolution of mixtures where the components have a wide range of boiling
points, and to shorten the overall analysis time by speeding up the elution of the
higher boiling compounds. Isothermal (constant temperature) conditions may
be unsatisfactory for the following reasons:
● if the isothermal temperature is too high, early eluting peaks may not be fully
resolved;
● if the isothermal temperature is too low, later eluting peaks may have unac-
ceptably long retention times, poor detection limits and small, broad and/or
skewed peaks;
● intermediate isothermal temperatures may result in part of the chro- matogram
having acceptable resolution and detection limits whilst other parts do not.

pg. 149
The separation of a series of n-alkanes isothermally at 150°C on a packed
column is shown in Figure 1(a), and illustrates the first two problems. The first
four alkanes, n-hexane (C6) to n-decane (C10), are incompletely resolved, C6 and
C7 co-eluting, whilst later eluting peaks are smaller, broader and are fronting
(Topic D2). In the temperature programmed chromatogram (Fig. 1(b)), a
complete separation of all the n-alkanes up to C21 has been achieved and in
about a third of the time taken to separate only up to C15 isothermally. C6 –C8
are fully resolved and the later peaks are sharper and more symmetrical. For
mixtures where the individual components are not, as in the case of the n-
alkanes, members of a homologous series, temperature programs are often more
complex. They may involve initial, intermediate and final isothermal periods
separated by temperature ramps of rates varying between 2°C and 30°C per
minute. The optimum conditions for a particular sample are generally estab-
lished by trial and error.

(a)
C9
C6 + C7 C10 C11 C12 C13

C8
C14

C15
Inject

0 5 10 15 20 25 30 45 50 55 90 95
Minutes

(b)
C10 C11 C12 C13 C14

C9
17

C18 C
19
C6 C7 C8
C20 21

0 4 8 12 16 20 24 28 32 36
Minutes

Fig. 1. Isothermal and temperature programmed separation of a homologous series of n-

alkanes; column: 20 ft ¥ 1/16 in, 3% Apiezon-L on 100–120 mesh Varaport 30; flow rate: 10
cm3 min 1 He. (a) Isothermal; (b) Temperature programmed.
Two potential disadvantages of temperature programming are the inevitable
delay between consecutive chromatographic runs while the oven is cooled down
and a stable starting temperature re-established, and the possible
decomposition of thermally-labile compounds at the higher temperatures.
Computer-controlled systems improve the reproducibility of temperature
programming, and the oven can be automatically force-cooled between runs to
save time.

Special Although GC is primarily a technique for analyzing mixtures of volatile solutes,

procedures used there are many compounds and materials that are either nonvolatile, have volatile
in GC components in a nonvolatile matrix, or are thermally labile. In some instances,
samples may contain solutes at such low levels that they must be pre- concentrated
prior to analysis.
Special procedures have been developed for handling such samples, e.g.
● Nonvolatile and thermally labile materials can either be pyrolyzed or
chemically derivatized to yield volatile products that can be successfully
chromatographed.
Small samples of paints, plastics, polymers and many ionic compounds
can be pyrolyzed (thermally decomposed) in a modified injection port under
controlled conditions to yield characteristic lower molecular mass and
volatile products that are swept onto the column. The resulting pyrograms
can be used as fingerprints of the original materials for identification
purposes (Fig. 2). Compounds of very limited volatility and/or thermal
sensitivity containing hydroxyl, carboxyl and amino functional groups can
be readily reacted with appropriate reagents to convert these into much less
polar methyl, trimethylsilyl or trifluoroacetyl funtionalities. Fatty acid, carbo-
hydrate, phenol and aminoacid derivatives can be chromatographed, but
often HPLC (Topics D6 and D7) is the preferred technique.
● Headspace analysis involves chromatographing the vapors derived from a
sample by warming it in a partially filled vial sealed with a septum cap. After
equilibration under controlled conditions, the proportions of volatile sample
components in the headspace above the sample are representative of those in
the bulk sample. The headspace vapors, which are under slight positive
pressure, are sampled by a modified and automated injection system or gas
syringe, and injected onto the column (Fig. 3(a)). The procedure is useful for
mixtures of volatile and nonvolatile components, such as residual monomers
in polymers, alcohol or solvents in blood samples (Fig. 3(b)), and flavors and
perfumes in manufactured products, as it simplifies the chromatograms and
protects the column from contamination by nonvolatile substances.
● Thermal desorption is a procedure where solutes can be collected on a solid
sorbent in a pre-concentration step, then thermally desorbed by rapid heating
in a unit linked to a modified injection port and through which the carrier gas is
flowing. Sorbents, such as activated charcoal or one of the gran- ular packings
used in packed column GC, are normally contained in a small tube with which
polluted industrial or urban atmospheres can be sampled by allowing passive
diffusion through the tube over a prolonged period or drawing the air through
over shorter periods. Thermal desorption can also be used in conjunction with
headspace analysis to pre-concentrate volatile solutes, and to purge and trap
volatile solutes in liquid samples using a stream of gas (Fig. 3(a)).

pg. 151
(a) (b)

0 2 4 6 8 10 12 0 2 4 6 8 10 12
Minutes Minutes

(c) (d)

0 2 4 6 8 10 12 0 2 4 6 8 10 12
Minutes Minutes

Fig. 2. Pyrograms of four common polymers. (a) Polyethylene; (b) A polyester; (c) A fluoro-
carbon copolymer; (d) Nylon.

Qualitative Methods of identifying unknown solutes separated by chromatographic tech- niques are
analysis described in Topic D2. In the case of gas chromatography, there are four alternatives:

● Comparisons of retention times (tR) with those of known solutes under iden- tical
conditions, preferably on two columns of differing polarity to reduce the chances of
ambiguous identifications;
● Comparisons of chromatograms of samples spiked with known solutes with the
chromatogram of the unspiked sample;
● Calculation of a retention index based on a set of standards that can be compared with
published or in-house library data. For a homologous series of compounds, the
logarithm of the retention time is directly proportional to the number of carbon
atoms. The Kováts system, based on the homologous
(a)
Carrier gas

Cryostat Desorption oven Automatic

(for thermal switching
desorption mode) system
Backflush
vent
Fused silica cold trap
inside tubular heater
Purge
vessel

Purge/desorb vent

Capillary column

(b)
12

Peak identification
1. diethyl ether
2. hexane
3. acetone
4. dichloromethane
5. methyl ethyl ketone
6. ethanol
7. benzene
8. methyl isobutyl ketone
9. toluene

Time (min)

Fig. 3. Combined headspace sampling, thermal desorption and purge and trap injection
system with example headspace chromatogram. (a) Injection system; (b) Headspace
chromatogram. Detection of solvents in blood sample by headspace analysis as part of an
industrial hygiene study. Sample held at 60°C. Column UCON LB 550, 25 m, at 40°C.
Produced from D.W. Grant, Capillary Gas Chromatography, 1996. © John Wiley & Sons Ltd.
Reproduced with permission.

series of n-alkanes, defines an index for each alkane as 100 ¥ the number of
carbon atoms for any column at any temperature, i.e. n-pentane is 500, n- octane
is 800 etc. The retention index of any other solute, relative to the n- alkane scale,
is calculated or read from a graph of log(retention time) against

pg. 153
Log (retention time) Unknown

860

600 700 800 900 1000

(Retention index scale (C number x100))

Fig. 4. Kováts retention index plot for n-alkanes. Retention index for unknown interpolated
from its log(retention time) as 860.

carbon number by interpolation (Fig. 4). For an unknown solute, its Kováts
index can then be checked against data bases by computer searching;
● Interfacing a gas chromatograph with a mass or infrared spectrometer. This
enables spectral information for an unknown solute to be recorded and inter-
preted. Identifications are facilitated by searching libraries of computerized
spectra (Topics F3 and F4).

Quantitative Methods used in quantitative chromatography are decribed in Topic D2 and

analysis alternative calibration procedures are described in Topics A5 and B4. Detector
response factors must be established for each analyte as these can vary consider-
ably, especially where selective detectors such as the ECD or NPD are used.
Calibration can be with external standards chromatographed separately from the
samples, by internal standardization, standard addition or internal
normalization. An internal standard should have similar chromatographic char-
acteristics to the analyte(s); homologues or isomers are often the most suitable.
The use of internal standards, where peak/area ratios of analyte to internal
standard are calculated, is preferable because a major source of variability arises
from the very small volumes injected by microsyringe, and peak area ratios are
independent of the volume injected. Autoinjectors minimize this source of error, and
quantitative analysis by GC can be expected to have an overall relative precision of
between 1 and 5%.
H IGH - PERFORMANCE LIǪUID
CHROMATOGRAPHY :
PRINCIPLES AND
INSTRUMENTATION

Key Notes
Principles High-performance liquid chromatography (HPLC) is a technique for the
separation of components of mixtures by differential migration through a
column containing a microparticulate solid stationary phase. Solutes are
transported through the column by a pressurized flow of liquid mobile
phase, and are detected as they are eluted.

Mobile phase The mobile phase is either a single solvent or a blend of two or more
having the appropriate eluting power for the sample components. It
ranges from a nonpolar liquid to aqueous buffers mixed with an organic
solvent.

Solvent delivery The solvent delivery system comprises a means of degassing, filtering
system and blending up to four solvents which are then delivered to the top of
the column under pressure by a constant flow pump.

Sample injection Liquid samples or solutions are introduced into the flowing mobile phase
at the top of the column through a constant or variable volume loop and
valve injector that is loaded with a syringe.

Column and Columns are straight lengths of stainless steel tubing tightly packed with
stationary phase a microparticulate stationary phase. The column packings are chemically-
modified silicas, unmodified silica or polymeric resins or gels.

Solute detection Solutes are detected in the mobile phase as they are eluted from the end
of the column. The detector generates an electrical signal that can be
amplified and presented in the form of a chromatogram of solute
concentration as a function of time.

Instrument control A dedicated microcomputer is an integral part of a modern high-

and data processing performance liquid chromatograph. Software packages facilitate the
control and monitoring of instrumental parameters, and the display and
processing of data.

Related topics Principles of chromatography (D2) High-performance liquid

chromatography: modes,
procedures and applications
(D7)

pg. 155
Principles High-performance liquid chromatography (HPLC) is a separation technique
where solutes migrate through a column containing a microparticulate
stationary phase at rates dependent on their distribution ratios (Topic D2).
These are functions of the relative affinities of the solutes for the mobile and
stationary phases, the elution order depending on the chemical nature of the
solutes and the overall polarity of the two phases. Very small particles of
stationary phase are essential for satisfactory chromatographic efficiency and
resolution, and the mobile phase must consequently be pumped through the
column, resulting in the generation of a considerable back-pressure. The compo-
sition of the mobile phase is adjusted to elute all the sample components reason-
ably quickly. Solutes eluted from the end of the column pass through a detector
that responds to each one. There are a number of modes of HPLC enabling an
extremely wide range of solute mixtures to be separated. The modes (Topic D7)
are defined by the type of stationary phase and associated sorption mechanism.
A schematic diagram of a high-performance liquid chromatograph is shown
in Figure 1. It consists of five major components:
● solvent delivery system;
● sample injection valve;
● column;
● detection and recording system;
● microcomputer with control and data-processing software.
These are described in the following sections.

Mobile phase The mobile phase, or eluent, is most frequently a blend of two miscible solvents
that together provide adequate eluting power and resolution. These are deter-

He He
Solvent
reservoirs column

Integrator/

Analytical
Dual-port Scavenger
valve

chamber

Pump

Chromatogram

Chromatography
Fig. 1. Schematic diagram of a high-performance liquid chromatograph. Reproduced from
control station
A. Braithwaite & F.J. Smith, Chromatographic Methods, 5th edn, 1996, first published by
Blackie Academic & Professional.
mined by its overall polarity, the polarity of the stationary phase and the nature
of the sample components. Unlike a GC carrier gas, which plays no part in
chromatographic retention and selectivity, the composition of an HPLC mobile
phase is crucial in both respects. For normal-phase separations (stationary
phase more polar than mobile phase), eluting power increases with increasing
solvent polarity, whilst for reversed-phase separations (stationary phase less
polar than mobile phase), eluting power decreases with increasing solvent
polarity. An eluotropic series of solvents, which lists them in order of increasing
polarity, is a useful guide to solvent selection for HPLC separations. Table 1 is an
example that also includes UV cut-off wavelengths as UV absorbance
detectors are the most widely used (vide infra). Elution can be under isocratic
conditions (constant mobile phase composition) or a composition gradient can
be generated by a gradient former to improve the resolution of complex
mixtures, especially if the sample components have a wide range of polarities.
The most widely used mobile phases for reversed-phase separations are
mixtures of aqueous buffers with methanol, or water with acetonitrile. For
normal-phase separations, which are less common, hydrocarbons blended with
chlorinated solvents or alcohols are typical.

Table 1. An eluotropic series of solvents for HPLC

Solvent Solvent strength Solvent strength UV cut-off

parameter, e° parameter, p¢ (nm)
(adsorption) (partition)
n-Hexane 0.01 0.1 195
Cyclohexane 0.04 0.2 200
Tetrachloromethane 0.18 1.6 265
Methylbenzene 0.29 2.4 285
Trichloromethane 0.40 4.1 245
Dichloromethane 0.42 3.1 230
Tetrahydrofuran 0.56 4.0 212
Propanone 0.56 3.9 330
Acetonitrile 0.65 5.8 190
iso-Propanol 0.82 3.9 205
Ethanol 0.88 4.3 205
Methanol 0.95 5.1 205
Ethanoic acid 1 4.4 255
Water 1 10.2 170

Solvent delivery The mobile phase is either a single solvent or a blend of two to four solvents
system delivered at pressures of up to about 5000 psi (350 bar) with a constant and
reproducible flow rate of <0.01–5 cm3 min 1. The solvent delivery system
comprises the following components:
● A mechanical pump designed to deliver a pulse-free flow of mobile phase. Most
are single or dual piston reciprocating pumps (Fig. 2) with specially designed
cams and pulse dampers to minimize or eliminate inherent flow variations, or
one-shot pulseless syringe pumps used primarily with micro- bore columns
(vide infra) requiring low flow rates. The wetted parts of the pump should be
inert to all solvents (stainless steel, titanium, industrial sapphire, ruby, and Teflon
being the principal choices) with minimal volume pumping chambers to facilitate
rapid changes of mobile phase composition.

pg. 157
Common inlet

Common outlet

Fig. 2. A typical twin-headed reciprocating pump. Reproduced from W.J. Lough & I.W.
Wainer, High Performance Liquid Chromatography, 1996, first published by Blackie Academic
& Professional.

● Solvent reservoirs with in-line filters (2 mm porosity or less) for each solvent
to remove dust and other particulate material. This reduces pump wear and
protects the column from becoming clogged which results in increased back-
pressures.
● A means of de-gassing the solvents to remove dissolved air. Air interferes with
the detector response by forming bubbles in the flow-through cell as the pressure
reduces to atmospheric at the end of the column. De-gassing is normally
accomplished by bubbling helium through each solvent to displace the air, or
by passing them through a commercial permeable-membrane de- gassing unit.
● A gradient former to generate binary, ternary or quaternary mixtures of solvents
with a pre-programmed composition profile during a separation (gradient
elution).

Sample injection Liquid samples and solutions are injected directly into the pressurized flowing
mobile phase just ahead of the column using a stainless steel and Teflon valve
fitted with an internal or external sample loop (Fig. 3). The loop, generally of
between 0.5 and 20 ml capacity, is first filled or partially filled with sample from
a syringe while the mobile phase flows directly to the column. By turning a
handle to rotate the body of the valve, the mobile phase is diverted through the
loop thus injecting the sample onto the top of the column without stopping the
flow. A disposable guard column is sometimes positioned between the injector
and the analytical column to protect the latter from a buildup of particulate
matter and strongly retained matrix components from injected samples. It
consists of a short length of column tubing, or a cartridge, packed with the same
stationary phase as is in the analytical column.
Valve injection can easily be automated, controlled by computer software and
used with autosamplers. For quantitative analysis, filled-loop injection has a
relative precision of about 0.5%.

Column and The column is where the separation process occurs and it is, therefore, the
stationary phase central component of a high-performance liquid chromatograph. There are two
(a) (b) Injection
Injection
port Waste
port Waste

Sample Sample
loop loop

To From To From
column pump column pump

Fig. 3. Sample-injection valve and loop. (a) Sample-loading position; (b) sample-injection position.

types of HPLC column, conventional and microbore, and a comparative summary is

given in Table 2.
Microbore columns have three principal advantages over conventional columns, i.e.:
● solvent consumption is about 80% less because of the much lower mobilephase
flow rate (10–100 ml min 1)
● the low volume flow rate makes them ideal for interfacing with a mass spectrometer
(Topics F3 and F4)
● sensitivity is increased because solutes are more concentrated, which is especially useful
if sample size is limited, e.g. for clinical specimens.

However, in practice, they are not as robust as conventional columns and are not
necessary for many routine applications.
Columns are connected to the sample injection valve and the detector using short
lengths of very narrow bore (~0.15 mm internal diameter) stainless steel or PEEK
(polyether ether ketone) tubing to minimize dead-volume which contributes to band
spreading in the mobile phase by diffusion.
HPLC stationary phases are predominantly chemically-modified silicas,
unmodified silica or cross-linked co-polymers of styrene and divinyl benzene. The surface
of silica is polar and slightly acidic due to the presence of silanol (Si- OH) groups. It can be
chemically modified with reagents, such as chlorosilanes, which react with the silanol
groups replacing them with a range of other func- tionalities (Fig. 4(a)). The resulting
bonded phases, which are hydrolytically stable through the formation of siloxane
(SiæOæSiæC) bonds, have different chromatographic characteristics and selectivities to
unmodified silica.
Octadecyl silica (ODS or C18) is the most widely used of all the stationary phases, being
able to separate solutes of low, intermediate and high polarities. Octyl and shorter alkyl chains
are considered to be more suitable for polar solutes. Aminopropyl and cyanopropyl (nitrile)
silicas are good replacementsfor unmodified silica, which can give variable retention times
due to traces ofwater in the solvents. Polar, and especially basic solutes, tend to give tailing

pg. 159
Table 2. A comparison of conventional and microbore HPLC columns
Conventional columns Microbore columns

Tubing Tubing
Stainless steel Stainless steel
Lengths 3, 10, 15, 20 and 25 cm Lengths 25 and 50 cm
Coupled lengths 1 m or more
1⁄4≤ outside diameter 1
⁄4≤ outside diameter
Internal diameter 4.6 mm Internal diameter 1 or 2 mm

Stationary phase (packing) Stationary phase (packing)

Porous, microparticulate silica, chemically-modified Porous, microparticulate silica, chemically-modified silicasilica
(bonded phases) or styrene/divinyl benzene (bonded phases) or styrene/divinyl benzene co-polymers
co-polymers

Mean particle diameters 3, 5 or 10 mm with a narrow Mean particle diameters 3, 5 or 10 mm with a narrow
range of particle sizes range of particle sizes

Operating pressures Operating pressures

500–3000 psi (35–215 bar) 1000–5000 psi (70–350 bar)

Typical mobile phases Typical mobile phases

Hydrocarbons chlorinated solvents or alcohols for Hydrocarbons chlorinated solvents or alcohols for
normal-phase; methanol or acetonitrile water or normal-phase; methanol or acetonitrile water or
aqueous buffers for reversed-phase aqueous buffers for reversed-phase

Flow rate 1–3 cm3 min 1

Flow rate 10–100 ml min 1

Modified instrumentation
Solvent delivery system capable of accurate flow control
down to 10 ml min 1 or less
Small volume sample injection valves
Small volume detector cells

Performance Performance
Efficiency increases with diminishing particle size, Very efficient and sensitive, but slow
but column life for 3 mm particles is shorter
Solvent consumption only a quarter that of conventional
Separations on 3 cm fast columns in less than
columns
1 minute

peaks on bonded phase silicas because of adsorptive interactions with residual silanols and
metallic impurities in the silica. The problem is reduced by end- capping, a process of blocking
the sites with trimethylsilyl ((CH3)3-Si-) groups(Fig. 4(b)), and by using highly purified silica
( 1 ppm metals).
Size exclusion and ion-exchange stationary phases are either silica or polymer based.
Sulphonic acid or quaternary ammonium groups provide cation and anion-exchange
capabilities respectively, but slow rates of exchange leading to poor efficiencies and low
sample capacities have limited their use, except for ion chromatography (Topic D7).
Chiral stationary phases have been developed for the separation of mixtures
of enantiomers but are expensive and have a very limited working life.
The availability of a wide range of bonded phases together with polymeric materials, has
resulted in the development of a number of modes of HPLC (Topic D7). The more important
stationary phases and their characteristics are summarized in Table 3.
(a)
CH3 CH3

Si OH CISi R Si O Si R

CH CH3
3
Si OH Si OH

(b) CH CH3
3
R Si O Si R
Si O Si
CH
CH + CISi(CH3)3 3
3 CH
Si OH End capping 3
Si O Si
CH
3
CH3

Fig. 4. Formation of bonded-phase silicas. (a) Monomeric bonded phases; (b) End-capping
of residual silanols. R = alkyl, aminoalkyl, ion-exchange groups.

Table 3. Stationary phases for HPLC

Stationary phase Sorption mechanism Characteristics
Unmodified silica, SiO2 Adsorption, normal-phase Polar, retention times variable due to
adsorbed water
Bonded phases
Octadecyl silica, -C18H37 (ODS or C18) Modified partition, reversed- Nonpolar, but unreacted silanol
Octyl silica, -C8H17 phase groups cause polar solutes,
Propyl silica, -C3H7 especially bases, to tail, pH range
limited to 2.5–7.5
All separate a very wide range of
solutes
Aminopropyl, -C3H 6NH2 Modified partition, normal or Polar, separates carbohydrates
reversed phase pH range limited to 2.5–7.5
Sulphonic acid, -(CH2)nSO3H Cation-exchange Slow mass transfer broadens peaks,
limited sample capacity, pH range
limited to 2.5–7.5 for silica-based
materials
Quaternary amine, -(CH2)nNR3OH Anion-exchange
Controlled-porosity silicas (some with Size exclusion Compatible with both organic and
–Si(CH3)3 groups) aqueous solvents, pH range limited
to 2.5–10

-, b-, g-cyclodextrin silicas Chiral selectivity based on Expensive, limited life, resolution
adsorptive interactions sensitive to mobile phase
composition

pg. 161
Polymer phases
Cross-linked styrene/divinyl benzene Modified partition, exclusion Nonpolar if unmodified,
co-polymers, unmodified or with or ion-exchange stable over pH range 1–13
ion-exchange groups
Solute detection Detectors are based on a selective response for the solute, such as UV- absorbance
or fluorescence, or on a bulk property of the mobile phase which is modified by
the solute, such as refractive index. Ideally, detectors should have the following
characteristics:
● a rapid and reproducible response to solutes;
● high sensitivity, i.e. able to detect very low levels of solutes;
● stability in operation;
● a small volume cell to minimize band broadening, i.e. 8 ml or less for a
conventional column, 1 ml or less for a microbore column;
● a signal directly proportional to solute concentration or mass over a wide
range (linear dynamic range);
● insensitivity to changes in temperature and flow rate;
● a cell design that does not entrap air bubbles that outgas from the mobile
phase at the end of the column.
Many types of detector have been investigated, and the most widely used are
summarized below and in Table 4.
● UV-visible absorbance detector. This type, which is the most widely used, is
based on the absorbance of UV or visible radiation in the range 190–800 nm
by solute species containing chromophoric groups or structures (Topics E8 and
E9). Detector cells are generally 1 mm diameter tubes with a 10 mm optical path
length and designed so as to eliminate refractive index effects which can alter the
measured absorbance. There are three types of UV- visible absorbance
detector:
(i) Fixed-wavelength filter-photometers, which are the simplest, employing
mercury-vapor lamp sources and optical filters to select a limited number
of wavelengths, e.g. 254, 280, 334 and 436 nm, and a phototube detector.
They have a limited use, lacking versatility, but they are cheap.
(ii) Variable-wavelength spectrophotometers (Fig. 5) are much more

Table 4. Characteristics of HPLC detectors

Detector Sensitivity g cm 3
Linear range Characteristics
UV-visible absorbance Good sensitivity, most widely used, selective
Filter-photometer 5 x10 10
104 for unsaturated groups and structures. Not
Spectrophotometer 5 x10 10
105 significantly flow or temperature sensitive.
Diode-array Can be used with gradient elution.
Spectrometer >2 x 10 10
105

Fluorescence 10 12
104 Excellent sensitivity, selective, including
fluorescent derivatives. Not flow or temperature
sensitive.
Refractive index 5 x 10 7
104 Almost universal, but only moderate sensitivity.
Very temperature sensitive (control to ±0.001°C).
Cannot be used with gradient elution.
Electrochemical
Conductimetric 10 8
104 Flow and moderately temperature sensitive.
Amperometric 10 12
105 Cannot be used with gradient elution. Detects
only ionic solutes. Excellent sensitivity, selective
but problems with electrode contamination.

pg. 163
versatile as they allow monitoring at any wavelength within the working
range of the detector to give the optimum response for each solute. They
employ deuterium and tungsten lamp sources for the UV and visible
regions, respectively, a diffraction grating monochromator for wave-
length selection and a photomultiplier detector. Many are computer-
controlled for programmable wavelength switching during a separation
to optimize sensitivity and selectivity.
(iii) Photodiode-array detectors are spectrometers with fixed optics and a
detection system consisting of one or two arrays of photodiodes on a silicon
chip positioned to receive radiation dispersed by a diffraction grating (Fig.
6). Electronic scanning, digitizing and processing of the signals by a
microcomputer enables ‘snapshots’ of the complete spec-
trum of the flowing eluent to be collected and stored every 0.1 s. The spectra
and the developing chromatogram at any wavelength can be displayed
on a VDU screen in real time and subsequently shown as a 3-D color
plot of absorbance, wavelength and time (Fig. 7). The data can be
manipulated and re-plotted on the screen, and comparisons made with
library spectra for identification purposes.

Mirror Mirror

Reference Grating Entrance slit

photo diode Prealigned lamp

Mirror

Flow cell
Beam splitter
Sample
photo diode

Fig. 5. UV-visible variable-wavelength spectrophotometric detector.

Concave Slit Lens Shutter Lens 1

holographic grating 2&3

Flow cell Aperture D2 lamp

Diode array
Spectrum
190–800 nm

Fig. 6. Diode-array detector (DAD).

5

0.20
600
Absorbance 4 500
2 3 400
1 300

200 Time
0.00 100 (seconds)
0
200

400
Wavelength (nm)

5 3 1
0.20 2 600
500
Absorbance

400

6 300
200 Time
0.00 100 (seconds)
0
200

400

Wavelength (nm)

Fig. 7. 3-D display mode for a diode-array detector (DAD).

● Fluorescence detectors are based on filter-fluorimeters or spectrofluori- meters. They are

more selective and can be up to three orders of magnitude more sensitive than UV
absorbance detectors. The detector responds selec- tively to naturally fluorescing solutes
such as polynuclear aromatics, quino- lines, steroids and alkaloids, and to fluorescing
derivatives of amines, amino acids and phenols with fluorogenic reagents such as dansyl
chloride (5- (dimethylamino)-1-naphthalene sulfonic acid).
● Refractive index (RI) monitors are the closest to being universal HPLC detectors, as
nearly all dissolved solutes alter the refractive index of the mobile phase. They are
differential detectors, generating a signal that depends on the difference between the
RI of the pure mobile phase and the modified value caused by the dissolved solute,
which can, therefore, be positive or negative.
They are several orders of magnitude less sensitive than UV absorbance detectors,
but are invaluable in the separation of saturated solutes such as carbohydrates, sugars
and alkanes. They are highly temperature sensitive and are very difficult to use with
gradient elution because the sample and reference cells cannot be continuously
matched.
● Electrochemical detectors are based on measuring either the conductance of an aqueous
mobile phase containing ionic solutes, or the current generated by the electrochemical
reduction or oxidation of solutes at a fixed applied
potential (amperometry) (Topic C9 ).

pg. 165
Conductance monitors are used in ion chromatography, a mode of HPLC
useful for separating low levels of inorganic and organic anions and cations
by ion-exchange (Topic D7).

Instrument These aspects of HPLC closely parallel those described for GC (Topic D4).
control and data Additional instrument parameters to be set and monitored include:
processing
● solvent composition, flow rate and pressure limit;
● gradient elution programs;
● wavelength(s) and wavelength-switching for UV-visible absorbance detec-
tors;
● wavelength range, sampling frequency and mode of display for a DAD.
HIGH-PERFORMANCEL IǪUID
CHROMATOGRAPHY: MODES, PROCEDURES AND
APPLICATIONS

Key Notes

Modes of HPLC Modes of HPLC are defined by the nature of the stationary phase, the
mechanism of interaction with solutes, and the relative polarities of the
stationary and mobile phases.

Optimization of After selection of an appropriate mode, column and detector for the
separations solutes to be separated, the composition of the mobile phase must be
optimized to achieve the required separation. A trial and error approach
or a computer aided investigation can be adopted.

Qualitative analysis Unknown solutes can be identified by comparisons of retention factors or

times, spiking samples with known substances or through spectrometric
data.

Quantitative analysis Ǫuantitative information is obtained from peak area or peak height
measurements and calibration graphs using internal or external
standards, or by standard addition or internal normalization.

Related topics Principles of chromatography (D2) High-performance liquid

chromatography: principles
and instrumentation (D6)

Modes of HPLC Almost any type of solute mixture can be separated by HPLC because of the wide
range of stationary phases available, and the additional selectivity provided by
varying the mobile phase composition. Both normal- and reversed- phase
separations are possible, depending on the relative polarities of the two phases.
Although these are sometimes referred to as modes of HPLC, the nature of the
stationary phase and/or the solute sorption mechanism provide a more specific
means of classification, and modes based on these and the types of solutes to
which they are best suited are summarized below.
● Adsorption chromatography. Separations are usually normal-phase with a
silica gel stationary phase and a mobile phase of a nonpolar solvent blended
with additions of a more polar solvent to adjust the overall polarity or eluting
power, e.g. n-hexane + dichloromethane or di-ethyl ether. The choice of solvent
is limited if a UV absorbance detector is to be used. Traces of water in the
solvents must be controlled, otherwise solute retention will not be
reproducible. Solutes are retained by surface adsorption; they compete with
solvent molecules for active silanol sites (Si-OH), and are eluted in

pg. 167
order of increasing polarity. This mode is not used extensively, but is suitable for
mixtures of structural isomers and solutes with differing functional groups. Members
of a homologous series can not be separated by adsorption chromatography because
the nonpolar parts of a solute do not interact with the polar adsorbent surface.
● Modified partition or bonded-phase chromatography (BPC). Most HPLC
stationary phases are chemically-modified silicas, or bonded phases, by far
the most widely used being those modified with nonpolar hydrocarbons. The solute
sorption mechanism is described as modified partition, because, although the bonded
hydrocarbons are not true liquids, organic solvent molecules from the mobile phase
form a liquid layer on the surface.
The most popular phase is octadecyl (C18 or ODS), and most separations are
reversed-phase, the mobile phase being a blend of methanol or aceto- nitrile with water
or an aqueous buffer. For weakly acidic or basic solutes,the role of pH is crucial because
the ionized or protonated forms have a much lower affinity for the ODS than the
corresponding neutral species, and are therefore eluted more quickly. The dissociation of
weak acids and the protonation of weak bases are shown by the following equations
RCOOH RCOO- + H

RNH 2 H RNH 3

Thus, at low pH, bases are eluted more quickly than at high pH, whilst the opposite
holds for weak acids (Fig. 1).

16

14 pH range for polymeric supports

12 pH range for silica supports

Weak base
10
k

8 acid
Ampholyte

0
0 2 4 6 8 10 12 14
pH

Fig. 1. Relation between retention factor, k, and the pH of the mobile phase for weak acids,
bases and ampholytes in reversed-phase separations.

pg. 167
ODS and other hydrocarbon stationary phases will separate many
mixtures, and are invariably a first choice in developing new HPLC methods.
They are particularly suited to the separation of moderately polar to polar
solutes (Fig. 2(a)).
Aminoalkyl and cyanoalkyl (nitrile) bonded phases (the alkyl group is
usually propyl) are moderately polar. The former is particularly useful in
separating mixtures of sugars and other carbohydrates (Fig. 2(b)), whilst the
latter is used as a substitute for unmodified silica, giving more reproducible
retention factors and less tailing, especially with basic solutes. Both normal-
phase and reversed-phase chromatography is possible by appropriate choice
of eluents.
● Ion-exchange chromatography (IEC). Stationary phases for the separation of
mixtures of ionic solutes, such as inorganic cations and anions, amino acids
and proteins, are based either on microparticulate ion-exchange resins, which are
crosslinked co-polymers of styrene and divinyl benzene, or on bonded phase
silicas. Both types have either sulfonic acid cation-exchange sites(-SO H )
or quaternary
3 ammonium anion-exchange sites (-N R OH ) incor-3porated into
their structures.

(a) (b) Instrument : Du Pont HPLC

(8800 series)
Column : Zorbax™ NH2 25 cm 
4.6 mm i.d.
Flow rate : 4 cm3/min
4 5
Mobile phase : Acetonitrile/water
(75:25)
Temperature : 50C
12
Detector : R.I.
10 11
2
Peak identity
1. Unknown
2. Fructose
3. Glucose
4. Sucrose
20 min.

Antiepileptic drugs

Column : 250  4.6 mm

Packing : Spherisorb 5 ODS
Flow rate : 1 ml/min.
Eluent : Solvent A:
methanol/water 20 : 80
Solvent B: acetonitril
Gradient : 17.5% B 45% B
in 15 minutes
Detector : UV 210 nm 0 1 2 3 Commercial Cola
type soft drink
Time (min)
on Zorbax™ NH2

Fig. 2. Separations of solutes of different polarities on bonded-phases. (a) Pharmaceuticals separated on ODS; UV
absorbance detection; (b) sugars separated on aminopropyl silica; RI detection.
Ion-exchange chromatography is not that widely used. Inorganic ions and some
cations are better separated by a related mode known as ion chro- matography (vide
infra), whilst for organic ions, ion-pair chromatographyis generally preferred because
of its superior efficiency, resolution and selectivity.
● Ion chromatography (IC). This is a form of ion-exchange chromatography for the
separation of inorganic and some organic cations and anions with conductometric
detection after suppressing (removing) the mobile phase
electrolyte (Fig. 3(a)).
The stationary phase is a pellicular material (porous-layer beads), the particles
consisting of an impervious central core surrounded by a thin porous outer layer (~2
mm thick) incorporating cation- or anion-exchange sites. The thin layer results in much
faster rates of exchange (mass transfer) than is normally the case with ion-exchange and
therefore higher efficiencies. Mobile phases are electrolytes such as NaOH, NaCO3 or
NaHCO3 for the separation of anions, and HCl or CH 3SO3H for the separation of cations.
The detection of low levels of ionic solutes in the presence of high levels of an eluting
electrolyte is not feasible unless the latter can be removed. This is accomplished by a
suppressor cartridge that essentially converts the electrolyte into water, leaving the
solute ions as the only ionic species in the mobile phase.
The following equations summarize the reactions for the separation of inorganic
anions on an anion-exchange column in the HCO form using a sodium hydrogen
3
carbonate mobile phase:

(a) (b) 1
Eluent
reservoir
Delivery mode
Pump

Sample
injector

Separator
Separation mode
column

Suppressor
Detection mode
Inject

Conductivity
cell

Data mode 0 2 4 6 8 10
Recorder Electronic Computer
integrator
Time (min)

Fig. 3. Ion chromatography. (a) Schematic diagram of an ion chromatograph. (b) Anions in water separated on an anion-
exchange column. Reproduced from Dionex UK Ltd with permission.

pg. 169
Column reaction:
n(Resin-N R3HCO3 ) Xn (Resin-N R )3 X
n
n + nHCO
3

where Xn = F , Cl , NO 3 , SO 42 , PO 43 etc.

Suppressor reactions:
Na HCO3 H H 2O CO2 Na
introduced via removed via
a membrane a membrane

Nan Xn + nH H +X n + nNa
separated by introduced via detected
n by removed via
the column a membrane conductance a membrane

Similar reactions form the basis of the separation of cations. An example of the
separation of inorganic anions at the ppm level is shown in Figure 3(b).
● Size exclusion chromatography (SEC). This is suitable for mixtures of
solutes with relative molecular masses (RMM) in the range 10 2–10 8 Da. Stationary
phases are either microparticulate cross-linked co-polymers of styrene and divinyl
benzene with a narrow distribution of pore sizes, or controlled-porosity silica gels,
usually end-capped with a short alkyl chain reagent to prevent adsorptive interactions
with solutes. Exclusion is not a true sorption mechanism because solutes do not
interact with the stationaryphase (Topic D2). They can be divided into three groups:
(i) Those larger than the largest pores are excluded completely, and are eluted in the
same volume as the interstitial space in the column, Vo.
(ii) Those smaller than the smallest pores, can diffuse throughout the entire network
and are eluted in a total volume, Vtot.
(iii) Those of an intermediate size separate according to the extent to which they
diffuse through the network of pores, of volume Vp and are eluted in volumes
between Vo and Vtot.
Only those solutes in the third group will be separated from one another, and their
retention volumes are directly proportional to the logarithm of their relative
molecular mass (RMM; molecular weight). Columns can be calibrated with standards
of known RMM before analyzing unknowns.
Figure 4 shows a typical plot of elution volume against log (RMM) and a
chromatogram of a mixture with a range of solutes of differing molecular mass. SEC is
of particular value in characterizing polymer mixtures and in separating biological
macromolecules such as peptides and proteins. It is also used for preliminary
separations prior to further analysis by other more efficient modes of HPLC.
● Chiral chromatography. Chiral stationary phases (CSP) enable enantiomers
(mirror image forms) of a solute to be separated. Several types of these stereoselective
materials have been investigated and marketed commercially, some of the most useful
being cyclodextrins bonded to silica. The cyclodex- trins are cyclic chiral
carbohydrates with barrel-shaped cavities into which solutes can fit and be bound by
H-bonding, p p and dipolar interactions. Where the total adsorptive binding energies
of two enantiomers differ, they will have different retention factors and can be resolved.
Steric repulsion and the pH, ionic strength and temperature of the mobile phase all
affect the resolution. Although of great interest to the pharmaceutical industry
for the
Exclusion
107 1. Thyroglobulin 670K
2. IgA 300K
Log relative molecular mass (RMM)

3. IgG 150K
106 4. LDH 143K
5. Oralbumin 44K
6. Trypsin inhibitor 20.1K
105 Separation
according to
molecular size
104

103

102
Permeation

Vo
5 10 15 20 25 30 35 40
Vto
t Retention volume (cm3)
Retention volume (VR)

Fig. 4. A typical size exclusion calibration curve and chromatogram of the separation of a protein mixture. Column:
BIOSEP-SEC-S3000. Mobile phase: pH 6.8 phosphate buffer. Detector: UV abs. at 280 nm. Reproduced from W.J.
Lough & I.W. Wainer (eds), High Performance Liquid Chromatography, 1996, first published by Blackie Academic &
Professional.

separation of enantiomeric forms of drugs having different pharmacological activities,

chiral columns are expensive and most have very limited working lives. Capillary
electrophoresis (Topic D8) provides a cheaper alternative.

Optimization of The optimum conditions for an HPLC separation are those which give the required
separations resolution in the minimum time. A stepwise approach based on the characteristics of the
solutes to be separated, and trial chromatograms with different mobile phase
compositions is usually adopted. The following is an outline of the procedure for a
reversed-phase separation where a hydrocarbon stationary phase, usually C18 (ODS), is
the first choice.

● The mode of HPLC most suited to the structures and properties of the solutes to be
separated is selected, having regard to their relative molecular mass, polarity, ionic or
ionizable character, and solubility in organic and aqueous solvents.
● The stationary phase and column are selected (Topic D6, Tables 2 and 3). The shortest
column and the smallest particle size of stationary phase consistent with adequate
resolution should be used.
● The detector, subject to availability, should match the solute characteristics. UV-visible
absorbance detectors are suitable for many solutes except those that are fully saturated.
Fluorescence and electrochemical detectors should be considered where high sensitivity is
required.
● Mobile phase composition is optimized by obtaining and evaluating a number of trial
chromatograms, often with the aid of computer optimization software packages. A
typical series of chromatograms for a reversed-phase separation on a hydrocarbon
bonded phase column is shown in Figure 5.

pg. 171
3
2
6
1 4
5

Methanol (Meth) 7.5%

Tetrahydrofuran (THF) 42.5%
Water 50%

6
4

Meth 31.8%

Water 47%

1 32

5 4

Meth 0%

Water 65%

1 2
3

6
5
4
Meth 10%
THF 25%
Water 65%

Meth 4.5%
6 5
Water 70%

Fig. 5. Optimizing an HPLC separation using ternary mobile phases. Solutes: 1. benzyl alcohol; 2. phenol; 3. 3-phenyl-
propanol; 4. 2,4-dimethylphenol; 5. benzene; 6. diethyl o-phthalate.
Note how the elution order of the six components in the mixture alters with
the mobile phase composition.

Qualitative Methods of identifying unknown solutes separated by chromatographic

analysis techniques are described in Topic D2. In the case of HPLC, there are four
alternatives:
● Comparisons of retention factors (k) or retention times (tR) with those of
known solutes under identical conditions, preferably on two columns of
differing selectivity to reduce the chances of ambiguous identifications.
● Comparisons of chromatograms of samples spiked with known solutes with
the chromatogram of the unspiked sample.
● Comparisons of UV-visible spectra recorded by a diode-array detector with
those of known solutes. This is of limited value because most spectra have
only two or three broad peaks so many solutes have very similar spectral
features.
● Interfacing a high-performance liquid chromatograph with a mass spectro-
meter. This enables spectral information for an unknown solute to be
recorded and interpreted. Identifications are facilitated by searching libraries
of computerized spectra (Topics F3 and F4).

Quantitative Methods used in quantitative chromatography are decribed in Topic D2, and
analysis alternative calibration procedures are described in Topic A5. Peak areas are
more reliable than peak heights as they are directly proportional to the quantity
of a solute injected when working within the linear range of the detector. Most
HPLC detectors have a wide linear dynamic range (Topic D6, Table 4), but
response factors must be established for each analyte as these can vary consider-
ably. Calibration is normally with external standards chromatographed sepa-
rately from the samples, or by standard addition. Constant volume loops for
sample injection give very good reproducibility (about 0.5% relative precision),
making internal standards unnecessary, cf gas chromatography (Topic D5), and
auto-injectors are frequently employed for routine work. An overall relative
precision of between 1 and 3% can be expected.

pg. 173
ELECTROPHORESIS AND ELECTROCHROMATOGRAPHY:

PRINCIPLES ANDINSTRUMENTATION

Key Notes

Principles Electrophoresis is a technique for the separation of components of

mixtures by differential migration through a buffered medium across
which an electric field is applied. Electrochromatography is a hybrid of
electrophoresis and HPLC.

Running buffer Samples are introduced into a buffer solution that provides an electrically
conducting medium and pH stability throughout the separation.

Supporting medium In some modes of electrophoresis, solutes migrate through a solid

medium consisting of a polymeric gel which supports the running buffer.

Electro-osmosis The application of a potential gradient across the running buffer causes
hydrated buffer cations to move towards the cathode. The resulting bulk
flow of liquid is described as an electro-osmotic flow, and is of particular
significance in high-performance capillary electrophoresis.

Sample injection Samples are placed in wells formed in the gel slab or column, or
introduced into a capillary column by hydrodynamic or electrokinetic
means.

Solute detection Solutes are either all detected in situ after the separation is complete by
treating the supporting medium with a chromogenic reagent, or
sequentially in the running buffer as they migrate towards one end of a
capillary tube.

Instrument control A dedicated microcomputer is an integral part of a modern

and data processing electrophoresis system. Software packages facilitate the control and
monitoring of instrumental parameters, and the display and processing
of data.

Related topics Principles of chromatography (D2) High-performance liquid

chromatography: principles
and instrumentation (D6)

Principles Electrophoresis, in its classical form, is used to separate mixtures of charged

solute species by differential migration through a buffered electrolyte solution
supported by a thin slab or short column of a polymeric gel, such as polyacry-
lamide or agarose, under the influence of an applied electric field that creates a potential
gradient. Two platinum electrodes (cathode and anode) make contact with the electrolyte
which is contained in reservoirs at opposite ends of the supporting medium, and these are
connected to an external DC power supply. Considerable amounts of heat may be
generated during separations at higher operating voltages (Joule heating), and many
systems stabilize the operating temperature by water-cooling. Buffer solutions undergo
electrolysis, producing hydrogen and oxygen at the cathode and anode, respectively, and
the reservoirs have to be replenished or the buffer renewed to maintain pH stability.
Cationic solute species (positively-charged) migrate towards the cathode, anionic
species (negatively-charged) migrate towards the anode, but neutral species do not
migrate, remaining at or close to the point at which the sample is applied. The rate of
migration of each solute is determined by its electrophoretic mobility, m, which is a
function of its net charge, overall size and shape, and the viscosity of the electrolyte. The
latter slows the migration rate by viscous drag (frictional forces) as the solute moves through
the buffer solution and supporting medium.
The distance travelled, d, after the application of a potential, E, between two electrodes
for time, t, is given by

E
d=m¥t¥ ( )—
S
(1)

where S is the distance between the two electrodes and E/S is the potential gradient. For two
separating solutes with mobilities m1 and m2, their separation, Dd , after time, t, is given by

E
Dd = (m 1- m ) ·2t · ( ) — (2)
S
Dd is maximized by the application of a large potential gradient over a long period.
However, as in chromatographic separations, diffusion of the solute species in the buffer
solution causes band spreading which adversely affects resolution, so excessive separation
times should be avoided.
The role of buffers in electrophoresis is crucial because many solutes are weakly acidic,
basic or ampholytic, and even small changes in pH can affect their mobility. Amino acids,
peptides and proteins are particularly susceptible as their direction of migration is a
function of pH, e.g for glycine

OH H2N CH 2 COOH OH H2N CH 2 COO

H3N CH 2 COOH

low pH H or H high pHanion

cation H3N+ CH2 COO
neutral/zwitterion

Typical formats for classical electrophoresis are shown in Figure 1. In the slab gel method,
the supporting gel is pre-formed into thin rectangular slabs on which a number of samples
and standards can be separated simultaneously. Alternatively, it can be polymerized in a
set of short tubes. The whole assembly is enclosed in a protective perspex chamber for
safety reasons because of the high voltages employed (500 V–2 kV DC, or up to 50 V cm 1). A
separation may take from about 30 minutes to several hours, after which the gels are
treated with a suitable visualizing agent to reveal the separated solutes (vide infra). Slab

pg. 175
(a)
Cathode

Glass plates
Buffer solution
Sample

Gel
Anode

Buffer solution

(b)
Cathode

Buffer solution

Sample
Rubber grommet (seal)

Glass or plastic tube

Polyacrylamide gel

Buffer solution

Anode

Fig. 1. Typical formats for classical gel electrophoresis. (a) Slab gel. (b) Tube gel. Reproduced
from M. Melvin, Electrophoresis, 1987 with permission from Wiley-VCH and from M. Melvin.

gels dissipate Joule heat more efficiently than column gels and have a superior
resolving power.
High-performance capillary electrophoresis (HPCE or CE) is a relatively
recently developed form where solutes are separated in a narrow-bore fused- quartz
capillary tube, 10–75 cm in length and 25–100 mm internal diameter (Fig. 2). High
voltages (10–50 kV DC, or up to 500 V cm 1) result in rapid separations, and the
heat generated is rapidly dissipated through the capillary wall. Solutes are
detected sequentially by a variety of means as they travel towards one end of the
capillary by migration and electro-osmosis (vide infra), but, unlike clas- sical
electrophoresis, only one sample at a time can be analyzed. A comparative summary
of classical and capillary electrophoresis is given in Table 1.
Data
acquisition

Capillary Capillary
inlet Detector outlet

Anode Cathode

Electrolyte Electrolyte
buffer buffer

Reservoir Reservoir

Hight voltage
power supply

Fig. 2. Schematic diagram of a high performance capillary electrophoresis system.

Capillary electrochromatography (CEC) is an even newer technique than CE

and uses capillaries packed with 5 mm or smaller particles of a stationary phase
similar to those used in HPLC (Topics D6 and D7). Solutes are separated by
a combination of electrophoretic migration and chromatographic sorption
processes (Topic D2) giving the technique additional versatility in varying the
selectivity. Efficiencies are particularly high because the packing minimizes band
spreading by solute diffusion in the buffer solution.

Running buffer The function of the running buffer is to provide an electrically conducting medium
and pH stability. The latter is essential in ensuring that solutes have a constant
mobility throughout the separation. Typical buffers covering a wide range of pH
values are listed in Table 2(a). Concentrations of 0.05–0.5 M provide an optimum
ionic strength that allows rapid migration of solutes without the generation of
excessive heat or losses by evaporation. Buffer additives in the form of
surfactants, complexing agents and organic solvents are sometimes added to
control migration rates and selectivity. Some examples are given in Table 2(b).

Supporting Various solid media are employed to support the running buffer in traditional
medium electrophoresis. Polymeric gels, such as polyacrylamide, agarose, starch and cross-
linked dextrans (Sephadex) are the most common, although paper and cellulose
acetate have also been used. The gels are saturated with the running buffer, and have
a restricted range of pore sizes that can be controlled during polymerization. This
facilitates the separation of solutes by a size exclusion mechanism (Topic D2) in
addition to their differential electrophoretic migration. Polyacrylamide gels are the
most versatile, offering superior resolution and dissipating heat efficiently. Agarose
gels are particularly effective for the separation of mixtures of large biomolecules
such as DNA proteins, as the pore sizes are larger than those in polyacrylamide.

Electro-osmosis When a potential gradient is applied across a running buffer, hydrated buffer cations
tend to be drawn towards the cathode producing a bulk flow of liquid known
as the electro-osmotic flow (EOF). Although the effect is not pronounced in
classical electrophoresis, in capillary electrophoresis it is

pg. 177
Table 1. A comparison of classical gel and capillary electrophoresis
Classical gel Capillary
Gels Tubing
Polyacrylamide or agarose Fused-quartz capillary
slabs: length and width 5–25 cm length 25–75 cm
thickness 1–2 mm internal diameter 20–100 mm
columns: length 7–10 cm
internal diameter ~5 mm

Applied field Applied field

100 V–2 kV (5–500 V cm 1) 10–50 kV (up to 500 V cm 1)

Heat dissipation Heat dissipation

Slow from columns, quicker from slabs, Rapid due to high surface area to volume
aided by water cooling ratio, aided by air cooling

Running buffer Running buffer

0.05–0.5 M electrolyte providing As for classical gel
conductivity and pH stability

Buffer additives Buffer additives

Urea, surfactants, complexing agents, Urea, surfactants, inorganic salts, organic
organic solvents solvents, sulfonic acids, chiral
cyclodextrins, amines

Sample injection Sample injection

0.1 –1 cm3 loaded into well in slab or 1–50 nl by hydrodynamic or electrokinetic
layered onto top of column method

Solute detection Solute detection

Chromogenic reagent or staining dye e.g. Similar to HPLC detectors,
ninhydrin for amino acids, coomassie UV absorbance and fluorescence the most
brilliant blue or ponceau S for proteins, common, see Table 3
acridine orange for nucleotides including
DNA

Modes Modes
Horizontal or vertical slab, disk or column Capillary zone electrophoresis (CZE)
SDS –PAGE Micellar electrokinetic chromatography
Isoelectric focusing (MEKC)
Immunoelectrophoresis Capillary gel electrophoresis (CGE)
Capillary isoelectric focusing (CIEF)

Performance Performance
Slow with limited resolution, but many Very high efficiency, resolution and
samples run simultaneously, may require sensitivity, moderately fast, 5 min to about
several hours 1h

responsible for the movement of all species (cationic, anionic and neutral) along
the capillary towards the detector and is the basis of the technique. The strength
of the EOF in fused-quartz capillaries filled with a running buffer arises from the
ionization of the surface silanol groups (SiOH) on the inner wall above about
pH 4. Hydrated buffer cations accumulate close to the negatively charged
Table 2(a). Typical running buffers for electrophoresis
Buffer Useful pH range
Phosphate 1.1–3.1
Ethanoate 3.8–5.8
Phosphate 6.2–8.2
Borate 8.1–10.1

Zwitterionic buffers
MES 5.2–7.2
(2-(4-morpholino)ethanesulfonic acid)
TRIS 7.3–9.3
((2,3-dibromopropyl) phosphate)

Table 2(b). Typical buffer additives for electrophoresis

Additive Function
Inorganic salts Change protein conformations
Organic solvents Modify EOF, increase solute solubilities
Urea Solubilize proteins, denature oligonucleotides
Surfactants Form micelles, cationic ones reverse charge on capillary wall
Cyclodextrins Provide chiral selectivity

surface to form an electrical double layer with an associated potential (zeta

potential) (Fig. 3). Application of a voltage across the capillary initiates an EOF,
the velocity or mobility, mEOF, of which is typically about an order of magnitude
greater than those of individual solutes. Hence, the total migration rate of a solute,
mTot, is
mTot = msolute + mEOF (3)
and all solutes, are carried towards the cathodic end of the capillary by the EOF
where they pass through the detector cell.
Important characteristics of the EOF are:
● The velocity increases with increasing pH of the running buffer as more
silanol groups become ionized.
● The velocity increases with the magnitude of the applied potential.
● The velocity decreases along with the zeta potential with increasing ionic
strength.
● The velocity decreases with increasing viscosity of the running buffer.
● It has a flat flow profile across the capillary in contrast to the parabolic profile
of a pumped HPLC mobile phase (Fig. 3). This minimizes band spreading and
results in very high efficiencies (plate numbers) for the
separating solutes.
● There is no pressure drop along the capillary, as occurs when the mobile phase
is pumped through an HPLC column, because the electrically- generated
driving force acts equally along the whole length.
● The velocity can be controlled, reduced to zero or reversed by additives such as
surfactants, organic solvents or quaternary amine salts incorporated intothe
running buffer, or by modifying the inside wall of the capillary.

pg. 179
Capillary wall

Electro-osmotic
bulk flow profile

Negatively charged Hydrated cations

capillary surface due accumulating at surface
to SiO– sites

Cross-sectional flow profile

due to electro-osmotic flow

Cross-sectional flow profile

due to hydrodynamic flow

Fig. 3. EOF and comparisons of flow-profiles in a fused-quartz capillary and an HPLC

column.

Sample injection For classical electrophoresis, samples of 0.1–1 cm3 are loaded into wells formed in
gel slabs or layered onto the tops of gel columns, often with the addition of a sucrose
solution to increase the density. For CE and CEC, much smaller samples (1–50 nl)
are drawn into one end of the capillary (usually the anodic end) from a sample
vial, either hydrodynamically using gravity, positive pressure or a vacuum, or
electrokinetically by applying a voltage for a short time when the EOF causes the
sample components to migrate into the capillary. The repro- ducibility of sample
injection into capillaries, typically 0.5–3%, is variable, and electrokinetic methods
may discriminate between components of a mixture because of differences in
electrophoretic mobilities. Time, temperature, pressure drops and sample and
running buffer viscosities are all sources of variability, and automated sample
injection is preferable to minimize these effects.

Solute detection For classical gel electrophoresis, solutes are detected on the gel after the separa- tion
is complete by treating it with a chromogenic reagent similar to those used in
TLC (Section D3), or a staining dye. The gels are immersed in a reagent or dye
solution, then the excess is removed by washing with a suitable solvent to reveal the
solute bands.
Detectors for CE and CEC are similar to those used in HPLC (Topic D6), UV-
absorbance spectrometers, especially diode array detectors, and fluorimeters
being the most common. DADs are increasingly preferred because of their ability
to monitor at multiple selected wavelengths and the complete spectral information
provided. The detector cell is normally a portion of the capillary itself, sometimes
enlarged (bubble cell) or bent so as to increase the optical path and hence the
sensitivity. Table 3 summarizes the characteristics of the more important detectors.

Table 3. Characteristics of CE and CEC detectors.

Detector Sensitivity Sensitivity Characteristics
mass (moles) concentration
(molar)
UV-visible absorbance 10 13
–10 16
10 5–10 8
Good sensitivity, most widely
used. DADs are versatile and
give spectral information.
Fluorescence 10 15
–10 17
10 7–10 9
Sensitive, but many solutes
need to be derivatized.
Laser-induced fluorescence 10 18
–10 20
10 14
–10 16
Extremely sensitive, but many
solutes need to be
derivatized. Expensive.

Electrochemical Sensitive, require special

Amperometric 10 18
–10 19
10 10–10 11 electronics and capillary
Conductometric 10 15
–10 16
10 7–10 8 modification. Conductometric
almost universal.

Instrument These aspects of CE and CEC closely parallel those described for GC and HPLC
control and data (Topics D4 and D6). Additional instrument parameters to be set and monitored
processing or controlled include:
● applied potential and run-time;
● temperature (heat dissipation);
● hydrodynamic and electrokinetic sample-injection.

pg. 181
EL EC T RO PH O RE SI S A N D
ELECTROCHROMATOGRAPHY: M ODES,
PR O CE D UR E S A N D
APPLIC ATI ON S

Modes of
electrophoresis and

Qualitative analysis For classical gel electrophoresis, unknown solutes are identified by
comparisons of the distances migrated with those of standards run
simultaneously. For capillary electrophoresis (CE) and CEC, migration
times and spectrometric data are used.

Quantitative analysis

Electrophoresis and electrochromatography: principles and

Modes of There are several modes of classical gel electrophoresis. They are defined by the
electrophoresis format and the nature of the gel, the running buffer and any incorporated
and electro- additives.
chromatography
● Slab gel, column and disk electrophoresis (Topic D8) are the principal formats.
They allow multiple samples and standards to be run simultane- ously for
comparison purposes, which is comparable to separations by TLC (Topic D3).
Samples are placed in wells near the centre of horizontal slabs when both
cationic and anionic species are to be separated, or at one end if all solutes
are expected to carry the same charge. With vertical slabs, samples are placed in
wells at the top of the slabs so that only downwards migration is possible.
● Sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) is
used to determine the relative molecular masses (RMM values) of individual
proteins in a mixture. The proteins are first de-natured with mercapto-
ethanol before bonding to the SDS, which is an anionic surfactant. The
negatively-charged proteins subsequently migrate at rates inversely propor-
tional to the logarithm of their RMM, the system being calibrated with
proteins of known RMM.
● Isoelectric focusing is a mode used to separate ampholytes such as amino
acids and peptides. A stable pH gradient is first created along the gel by
polymerizing it with a mixture of polyamino-polycarboxylic acids having a
range of pKa values. Applying a potential causes them to migrate to positions
in the gel where they become stationary on account of electrical neutrality by
forming neutral molecules or zwitterions. The pH values at these positions
define their isoelectric points, pI. Electrophoresis of samples results in
ampholytic components migrating to their respective isoelectric points in the
pH gradient, after which they can be located with a suitable dye.
● Immunoelectrophoresis depends on specific antigen–antibody reactions for
the detection of separated proteins. Low-voltage electrophoresis of samples
in an agarose gel is followed by the introduction of antisera that diffuse
throughout the gel forming visible precipitates with the separated antigens.
Capillary electrophoresis separations can be performed in one of four principal
modes, each depending on a different separation mechanism. They are primarily
used in the pharmaceutical, clinical and biomedical fields for the analysis of
mixtures of amino acids, peptides, proteins and other macro- molecules, and
drugs and their metabolites in body fluids. Analyses down to nanogram (10 9 g)
or picogram (10 12 g) levels are often quicker, giving better resolution than
corresponding HPLC procedures (Fig. 2c).
● Capillary zone electrophoresis, CZE, is the simplest and currently the most
widely used mode of CE. The capillary is filled with a running buffer of the
appropriate pH and ionic strength (Topic D8, Table 1a), and all solutes are
carried towards the cathodic end of the capillary by a strong EOF. Cationic
and anionic solutes are separated, but neutral species, which migrate
together at the same velocity as the EOF, are not separated from one another.
Cationic solutes migrate faster than the EOF because their overall mobilities
are enhanced by their attraction to the cathode, whereas anionic solutes
migrate slower than the EOF because they are attracted towards the anode.
Solutes reach the detector in order of decreasing total mobility (mTot msolute
mEOF), as shown diagrammatically in Figure 1a, the individual solute mobili-
ties being determined by their size and charge, that is:
(i) cationic species first in increasing order of size;
(ii) neutral species next, but not separated;
(iii) anionic species last in decreasing order of size.
Buffer additives, such as urea, surfactants and organic solvents are some-
times used to alter the selectivity of a separation through controlling the EOF,
solute mobilities, solubilities and other effects (Topic D8, Table 1b). An
example of a CZE electropherogram, the separation of some artificial
sweetners and preservatives, is shown in Figure 1b.
● Micellar electrokinetic chromatography, MEKC or MECC, is a more versa- tile
mode than CZE because both neutral and ionic solutes can be separated. A
surfactant is added to the running buffer, forming aggregates of mole- cules,
or micelles, having a hydrophobic center and a positively or nega- tively-charged
outer surface (Fig. 2a). The micelles act as a chromatographic
pseudo-stationary phase, into which neutral solutes can partition, their
distribution ratios depending on their degree of hydrophobicity. Cationic
micelles migrate towards the cathode faster than the EOF and anionic micelles
more slowly. Neutral solutes migrate at rates intermediate between the velocity
of the EOF and that of the micelles. By analogy with chromatog- raphy, they are
eluted with characteristic retention times, tR, that depend on

pg. 183
(a)

1 Phenylalanine 7 Cyclamate
(b)
1 2 Aspartame 8 Sorbic acid
3 p-propyl hydroxybenzoate 9 Benzoic acid
80 4 p-ethyl hydroxybenzoate 10 Aspartic acid
5 p-methyl hydroxybenzoate 11 Saccharin
60 6 Dihydroxyacetic acid 12 Acesulfame

4 9
40
2
3
20 5 11
6 8 12
7 10
0

3.8 4.8 5.8 6.8 7.8

Time (min)

Fig. 1. Capillary zone electrophoresis (CZE). (a) EOF and order of solute migration. (b) Separation of some artificial
sweeteners and preservatives by CZE capillary, 65 cm, 50 mm i.d.; buffer, 0.02 M borate, pH 9.4; temperature, 25°C;
voltage, 30 kV, injection, hydrodynamic 50 mbar sec; detection, UV absorbance at 192 nm. Reproduced from D.N. Heijer,
High Performance Capillary Electrophoresis, 1992, with permission from Agilent Technologies UK Ltd and D.N. Heijer.

their distributions between the running buffer and the micelles. The general order of
elution is:
(i) cationic micelles and cationic solutes;
(ii) neutral solutes partitioning into the cationic micelles;
(iii) EOF;
(iv) neutral solutes partitioning into anionic micelles;
(v) anionic micelles and anionic solutes.
Sodium dodecyl sulfate, SDS, is a frequently used anionic surfactant for MEKC. The
most hydrophobic neutral solutes migrate at the same velocity as the SDS, which is
slower than the EOF, and the least hydrophobic neutral solutes migrate at the same
velocity as the EOF (Fig. 2a).
An example of an MEKC electropherogram, the separation of some compo-
nents of cold relief-products, is shown in Figure 2b. The addition of chiral selec- tors such
as the cyclodextrins (Topic D6), in place of a surfactant, enables mixtures of
enantiomers to be separated more cheaply than by HPLC (Fig. 2c).
(a)

= Surfactant
(negative charge) = Electroosmotic flow

= Solute = Electrophoresis

(b)
1 Acetaminophen 4
2 Caffeine
3 Sulpyrin 11
10
4 Naproxen
5 Guaifenesin 5
6 Impurity 1
2 3
7 Phenacetin 78
8 Ethenzamide
9 4-isopropylantipyrine 9
6
10 Noscapine
11 Chlorpheniramine

5 10 15 20
Time (min)

(c) Hexobarbital (0.1 mg ml–1)

mAU mAU
20 CE HPLC
80

60

40

20

0 0
6.0 7.0 8.0 9.0
Time (min) 10 30 50
Time (min)

Fig. 2. Micellar electrokinetic capillary chromatography (MEKC). (a) Formation and migration of SDS micelles;
(b) separation of some constituents of cold-relief products; (c) separation of the enantiomers of hexobarbital and
comparison with an HPLC separation on a chiral stationary phase. b and c reproduced from D.N. Heijer, High
Performance Capillary Electroporesis, 1992, with permission from Agilent Technologies UK Ltd and D.N. Heijer.

pg. 185
● capillary gel electrophoresis, CGE, is similar to classical gel electrophoresis, the
capillary being filled with a polyacrylamide or agarose gel that super- imposes
size exclusion selectivity onto the electrophoretic migration of ionic solutes. The
larger the solute species the slower the rate of migration through the gel. Solute
peaks are narrow because band spreading by diffusion in the running buffer is
hindered by the gel structure. The main applications of CGE are in separating
polymer mixtures, protein fractions and DNA sequencing.
● capillary isoelectric focusing, CIEF, is similar to classical isoelectric focusing,
a pH gradient being first formed in the capillary using carrier ampholytes having
pI values spanning the required pH range, typically 3 to 10. Sample solutes
migrate and are focused in positions along the capillary where their isoelectric
point, pI, is equal to the pH. Solute zones are self-sharpening because diffusion
away from the focal points causes the solutes to aquire a charge which results in
them migrating back towards their isoelectric point. After the separation is
complete, pressure is applied to the anodic end of the capillary to move all the
solutes sequentially through the detector cell.

Capillary electrochromatography, CEC, is a relatively new technique, and is a

hybrid of capillary electrophoresis and high-performance liquid chromatog- raphy,
combining elements of both. Particular features of CEC are:

● the capillary is packed with an HPLC stationary phase, usually a bonded- phase
silica, and filled with a running buffer (>pH 4);
● as in CE, the applied potential generates a strong EOF with a flat flow profile,
but the electrical double-layer formed is predominantly at the surface of the
individual particles of packing rather than the capillary wall;
● unlike in HPLC, there is no pressure drop because the driving force is gener-
ated throughout the length of the column;
● even higher efficiencies are observed than in CZE because the column packing
limits solute diffusion in the mobile phase. Very small particles of stationary
phase, currently 1.5 to 3 mm nominal diameter, can be used and columns of 25
to 50 cm in length are typical. Internal diameters are generally 50 to 00 mm,
but narrower columns are advantageous because the EOF is faster, thus speeding
up the separations;
● column packings can be porous, non-porous, spherical or irregular in shape
and of controlled pore size if required. In some cases, mixed-mode separa-
tions can be achieved by using both non-polar and polar or ionic bonded
phases in the same column.

CEC separations are based on both electrophoretic migration for charged

analytes and chromatographic sorption for neutral species, hence providing an
additional source of selectivity over and above differences in electrophoretic
mobility. The composition of the mobile phase can have dramatic effects on both
the EOF and the selectivity of the separation. CEC has a number of other advan-
tages over both CE and HPLC. Compared to HPLC, solvent consumption is
greatly reduced, which facilitates coupling CEC to mass spectrometry (Section
F). Furthermore, there is no need to use a micelle-forming surfactant additive as
in MEKC when separating neutral solutes. This is also an advantage when
coupling the technique to mass spectrometry because solvents containing high
concentrations of surfactants such as SDS can cause difficulties with some
ionization sources.
As yet, the number of applications is limited but is likely to grow as instru-
mentation, mostly based on existing CE systems, and columns are improved and
the theory of CEC develops. Current examples include mixtures of poly-
aromatic hydrocarbons, peptides, proteins, DNA fragments, pharmaceuticals
and dyes. Chiral separations are possible using chiral stationary phases or by the
addition of cyclodextrins to the buffer. In theory, the very high efficiencies
attainable in CEC provides high peak capacities, and therefore the possibility
of separating complex mixtures of hundreds of components. A typical CEC
separation is shown in Figure 3.
UV absorbance (254 nm)

1 2
8 9
1011 13
12 14

0 2 4 6 8
Time (min)
Fig. 3. 30
Column: cm  75 m,
Separation of 14
17explosive
cm packed with 1.5 m
compounds by capillary electrochromatography.
dp non-porous Column:
ODS-II (Micra Scientific);
15% 30methanol–85%
cm ¥ 75 µm, 17 10 cmmMpacked
MES;with 1.5 µm
applied dp non-porous
voltage: 1 s at Scientific);
ODS-II (Micra
12 kV; injection: 2 kV. Peaks:mobile
1 = oc
phase: 15% methanol–85%210
tetranitro-1,3,5,7-tetrazocine; mM MES; applied voltage: 12 kV; injection:
= hexahydro-1,3,5-trinitro-1,3,5-triazine; 1 s at 2 kV.
3 = 1,3-dinitrobenze
Peaks: 1 = octahydro-1,3,5,7-tetranitro-1,3,5,7-tetrazocine;
trinitrobenzene; 5 = nitrobenzene; 6 = 2,4,6-trinitrotoluene; 7 =2 2,4-dinitrotoluene;
= hexahydro-1,3,5-trinitro-
8 = methyl
1,3,5-triazine; 3 = 1,3-dinitrobenzene;
trinitrophenylnitramine; 4 = 1,3,5-trinitrobenzene;
9 = 2,6-dinitrotoluene; 5 = nitrobenzene;
10 = 2-amino-4,6-dinitrotoluene; 11 = 2-nitroto
6 = 2,4,6-trinitrotoluene;
nitrotoluene; 7 = 2,4-dinitrotoluene;
13 = 4-amino-2,6-dinitrotoluene; 148==3-nitrotoluene.
methyl-2,4,6-trinitrophenylnitramine;
9 = 2,6-dinitrotoluene; 10 = 2-amino-4,6-dinitrotoluene; 11 = 2-nitrotoluene;
12 = 4-nitrotoluene; 13 = 4-amino-2,6-dinitrotoluene; 14 = 3-nitrotoluene.

Qualitative Classical gel electrophoresis is primarily a qualitative technique, the identifica-

analysis tion of unknown solutes being accomplished by comparisons of migration
distances with those of standards run simultaneously or sequentially under
identical conditions. Migration distances can be expressed as Rf values
analogous to those used in TLC (Topic D3).
Solute migration times in CE and CEC separations can be equated to
chromatographic retention times, tR, and similar methods are used to identify
unknowns (Topic D2), that is:
● comparisons of retention times with those of known solutes under identical
conditions;
● comparisons of electropherograms of samples spiked with known solutes
with the electropherogram of the unspiked sample;
● comparisons of UV-visible spectra recorded by a diode-array detector with
those of known solutes;
● interfacing of a CE or CEC system with a mass spectrometer. Identifications are
facilitated by searching libraries of computerized spectra (Topics F3 and F4).

pg. 187
Quantitative Generally, only semi-quantitative information can be obtained from classical gel
analysis electrophoresis; visual comparisons can be made with standards run on the
same gel. Scanning densitometry, as used in TLC (Topic D3) provides more
reliable data, but variations in gel thickness, band spreading and reactions with
reagents and staining dyes all contribute to considerable variability in the
measurements.
For CE and CEC, the methods used for quantitative chromatography (Topic
D2) are suitable, and alternative calibration procedures are described in TopicA5.
Peak areas are more reliable than peak heights, as they are directly propor- tional
to the quantity of a solute injected when working within the linear range of the
detector.
FURTHER READING

General Reading Atkins, P.W. (1998) Physical Chemistry, 6th edn. Oxford
University Press, Oxford, UK.
Fifield, F.W. and Kealey, D. (2000) Principles and Practice of Analytical Chemistry, 5th edn. Blackwell
Science, Oxford, UK.
Harris, D.C. (1998) Ǫuantitative Chemical Analysis, 5th edn. Freeman, USA. Kellner, R., Mermet, J-M.,
Otto, M. and Widmer, H.M. (eds) (1998) Analytical
Chemistry. John Wiley & Sons, Chichester, UK.
Skoog, D.A., Holler, J.F., Nieman, T.A. (1997) Principles of Instrumental Analysis. Thomson, New York.
Whittaker, A.G., Mount, A.R., Heal, M.R. (2000), Instant Notes Physical Chemistry. Bios, Oxford, UK.
Willard, H.H., Merritt, L.L. Jr., Dean, J.A. and Settle, F.A. Jr. (1998) Instrumental Methods of Analysis,
7th edn. Wadsworth, USA.

More advanced reading

Section A Kenkel, J. (1999) A Primer on Ǫuality in the Analytical
Laboratory. CRC Press, UK.

Section B Miller, J.C. and Miller, J.N. (2000) Statistics and

Chemometrics for Analytical Chemistry, 4th edn. Ellis Horwood PTR Prentice Hall, UK.

Section C Bard, A.J. and Faulkner, L.R. (2000) Electrochemical

Methods: Fundamentals and Applications. J. Wiley & Sons, Chichester, UK.
Kissinger, P.T. and Heineman, W.R. (eds), (1995) Laboratory Techniques in Electroanalytical
Chemistry. M. Dekker, New York.
Monk, P.M.S. (2001) Fundamentals of Electroanalytical Chemistry. J. Wiley & Sons, UK.
Wang, J. (2000) Analytical Electrochemistry. J. Wiley & Sons, Chichester, UK.

Section D ACOL book on Gas Chromatography, 2nd edn (1995). John

Wiley & Sons, Chichester, UK.
Anderson, R. (1987) Sample Pre-treatment and Separation. John Wiley & Sons, Chichester, UK.
Baker, D.R. (1995) Capillary Electrophoresis. John Wiley & Sons, Chichester, UK. Braithwaite, A. and
Smith, F.J. (1996) Chromatographic Methods, 5th edn.
Chapman and Hall, UK.
Lindsay, S. (1992) High Performance Liquid Chromatography, 2nd edn. John Wiley & Sons,
Chichester, UK.

pg. 189