Glassware

The manufacturers’ tolerances for the volumes of a number of items of glassware are given in Chapter 1. The larger the volume measure the smaller the tolerance percentage is of the nominal volume. Thus, for a Grade A 1 ml pipette the volume is within ± 0.7% of the nominal volume, whereas for the 5 ml pipette the volume is within ± 0.3% of the nominal volume. If greater accuracy than those guaranteed by the tolerances is required, then the glassware has to be calibrated by repeated weighing of the volume of water contained or delivered by the item of glassware. This exercise is also useful for judging how good one's ability to use a pipette is, since weighing of the volumes of water dispensed correctly several times from the same pipette should give weights that agree closely.

Primary standards and standard solutions

Primary standards are stable chemical compounds that are available in high purity and which can be used to standardise the standard solutions used in titrations. Titrants such as sodium hydroxide or hydrochloric acid cannot be considered as primary standards since their purity is quite variable. So, for instance, sodium hydroxide standard solution may be standardised against potassium hydrogen phthalate, which is available in high purity. The standardised sodium hydroxide solution (secondary standard) may then be used to standardise a standard solution of hydrochloric acid. Table 3.1 lists some commonly used primary standards and their uses.

Table 3.1 Primary standards and their uses

EDTA, Ethylenediamine tetracetic acid.

Strong acid/strong base titrations

Figure 3.1 shows the titration curve obtained from the titration of a strong acid with a strong base. The pH remains low until just before the equivalence point, when it rises rapidly to a high value. In many titrations a coloured indicator is used, although electrochemical methods of end-point detection are also used. An indicator is a weak acid or base that changes colour between its ionised and un-ionised forms; the useful range for an indicator is 1 pH either side of its pKa value. For example, phenolphthalein (PP) pKa 9.4 (colour changes between pH 8.4 and pH 10.4) undergoes a structural rearrangement as a proton is removed from one of its phenol groups when the pH rises, and this causes the colour change (Fig. 3.2). Methyl orange (MO) pKa 3.7 (colour changes between pH 2.7 and pH 4.7) undergoes a similar pH-dependent structural change. Both these indicators fall within the range of the inflection of the strong acid/strong base titration curve.

Fig. 3.1 The change in pH as 25 ml of 1 M HCl are titrated with 1 M NaOH.

Fig. 3.2 The structural rearrangement responsible for the colour change in phenolphthalein.

There are only a few direct strong acid/strong base titrations carried out in pharmacopoeial assays.

Strong acid/strong base titrations are used in pharmacopoeial assays of: perchloric acid, hydrochloric acid, sulphuric acid and thiamine hydrochloride.

Weak acid/strong base and weak base/strong acid titrations

On addition of a small volume of the strong acid or strong base to a solution of the weak base or weak acid, the pH rises or falls rapidly to about 1 pH unit below or above the pKa value of the acid or base. Often a water-miscible organic solvent such as ethanol is used to dissolve the analyte prior to the addition of the aqueous titrant.

Figure 3.3 shows a plot of pH when 1 M NaOH is added to 25 ml of a 1 M solution of the weak acid aspirin.

Fig. 3.3 Titration curve for 25 ml of a 1.0 M solution of aspirin (pKa 3.5) titrated with 1.0 M NaOH.

In the case of aspirin, the choice of indicator is restricted by where the inflection in its titration curve lies; PP is suitable as an indicator whereas MO is not.

In the example of the titration of quinine with hydrochloric acid (Fig. 3.4), MO is a suitable indicator because it falls within the inflection of the titration curve whereas PP is not suitable.

Fig. 3.4 25 ml of a 1.0 M solution of quinine (pKa 8.05) titrated with 1.0 M hydrochloric acid.

Some acids or bases can donate or accept more than one proton, i.e. 1 mole of analyte is equivalent to more than 1 mole of titrant. If the pKa values of any acidic or basic groups differ by more than ca 4, then the compound will have more than one inflection in its titration curve. Sodium carbonate is a salt of carbonic acid and it can accept two protons. The pKa values of carbonate and bicarbonate are sufficiently different (pKa 10.32 and 6.38) for there to be two inflections in the titration curve. The two stages in the titration are:

In a titration of sodium carbonate, the first inflection is indicated by PP and the whole titration by MO (Fig. 3.5).

Fig. 3.5 Titration of 1 M sodium carbonate with 1 M HCl.

Self-test 3.3

How many inflections do the following substances have in their titration curves when titrated with a strong base? Draw the predominant forms of these substances which would exist at pH 14.

Answers: oxalic acid 1, fumaric acid 1, citric acid 1, phenylalanine 2

Weak acid/strong base titration is used in the pharmacopoeial assays of: benzoic acid, citric acid, chlorambucil injection, mustine injection, nicotinic acid tablets and undecanoic acid.

Estimation of esters by back titration

Excess of sodium hydroxide is added to the ester. The following reaction occurs:

The XSNaOH is back titrated with HCl using PP as an indicator.

This procedure is used in pharmacopoeial assays of: benzyl benzoate, dimethyl phthalate, ethyl oleate, methyl salicylate, cetostearyl alcohol, emulsifying wax, castor oil, arachis oil, cod liver oil and coconut oil.

Saponification value

The assay of fixed oils provides a special case of ester hydrolysis since they are triesters of glycerol. The saponification value for a fixed oil is the number of mg of potassium hydroxide (KOH) equivalent to 1 g of oil. A high value means rancidity, a low value possible adulteration with mineral oil. Almost all edible oils have a saponification value between 188 and 196. Hydrolysis of the fixed oil is carried out with ethanolic KOH.

This procedure is used in the pharmacopoeial assays of: castor oil, cod liver oil, cotton seed oil, almond oil and sesame seed oil.

Acid values are also determined for fixed oils. The acid value for a substance is the number of mg of KOH required to neutralise 1 g of the test substance when it is titrated with 0.1 M ethanolic KOH to a PP end-point. This value is quoted for many fixed oils in order to eliminate rancid oils, which contain large amounts of free fatty acid. Typically acid values for fixed oils are in the range of 1–2.

Estimation of alcohols and hydroxyl values by reaction with acetic anhydride (AA)

Alcohols can be determined by reaction with excess acetic anhydride (AA) (Fig. 3.7). This is a useful titrimetric method because the alcohol group is difficult to estimate by any other means.

Fig. 3.7 Estimation of benzyl alcohol by reaction with acetic anhydride.

The excess AA and acetic acid may be back titrated with NaOH using PP as an indicator.

In a related assay, a hydroxyl value is determined for a fixed oil. A 1:3 mixture of AA in pyridine is used in the determination; the pyridine is present as a catalyst. The hydroxyl value may be defined as:

Reaction with acetic anhydride is used in pharmacopoeial assays of benzyl alcohol and dienestrol, and determination of hydroxyl values of castor oil, cetosteryl alcohol and cetomacrogol.

Theory

Non-aqueous titration is the most common titrimetric procedure used in pharmacopoeial assays and serves a double purpose, as it is suitable for the titration of very weak acids and bases and provides a solvent in which organic compounds are soluble. The most commonly used procedure is the titration of organic bases with perchloric acid in acetic acid. These assays sometimes take some perfecting in terms of being able to judge the end-point precisely.

The theory is, very briefly, as follows: water behaves as both a weak acid and a weak base; thus, in an aqueous environment, it can compete effectively with very weak acids and bases with regard to proton donation and acceptance, as shown in Figure 3.8.

Fig. 3.8 Competition of water with weak acids and bases for proton acceptance and donation.

The effect of this is that the inflection in the titration curves for very weak acids and very weak bases is small, because they approach the pH limits in water of 14 and 0 respectively, thus making end-point detection more difficult. A general rule is that bases with pKa < 7 or acids with pKa > 7 cannot be determined accurately in aqueous solution. Various organic solvents may be used to replace water since they compete less effectively with the analyte for proton donation or acceptance.

Non-aqueous titration of weak bases

Acetic acid is a very weak proton acceptor and thus does not compete effectively with weak bases for protons. Only very strong acids will protonate acetic acid appreciably according to the equation shown below:

Perchloric acid is the strongest of the common acids in acetic acid solution, and the titration medium usually used for non-aqueous titration of bases is perchloric acid in acetic acid. Addition of acetic anhydride, which hydrolyses to acetic acid, is used to remove water from aqueous perchloric acid. Weak bases compete very effectively with acetic acid for protons. Oracet blue, quinalidine red and crystal violet (very weak bases) are used as indicators in this type of titration. A typical analysis is shown in Figure 3.9 for L-3,4-dihydroxyphenylalanine (LDOPA).

Fig. 3.9 Analysis of L-3,4-dihydroxyphenylalanine (LDOPA) by non-aqueous titration.

When the base is in the form of a salt of a weak acid, removal of an anionic counter ion prior to titration is not necessary, e.g. for salts of bases with weak acids such as tartrate, acetate or succinate. However, when a base is in the form of a chloride or bromide salt, the counter ion has to be removed prior to titration. This is achieved by the addition of mercuric acetate; the liberated acetate is then titrated with acetous perchloric acid. This is illustrated in Figure 3.10 for the example of phenylephrine HCl.

Fig. 3.10 The analysis of phenylephrine hydrochloride by non-aqueous titration.

Non-aqueous titration with acetous perchloric acid is used in the pharmacopoeial assays of: adrenaline, metronidazole, codeine, chlorhexidine acetate, chlorpromazine, amitriptyline HCl, propranolol HCl, lidocaine (lignocaine) HCl, and HCl and quaternary amine salts, such as neostigmine bromide and pancuronium bromide.

Non-aqueous titration of weak acids

For the non-aqueous titration of weak acids, a solvent such as an alcohol or an aprotic solvent is used, which does not compete strongly with the weak acid for proton donation. Typical titrants are lithium methoxide in methanol or tetrabutyl ammonium hydroxide in dimethylformamide. End-point detection may be carried out with thymol blue as an indicator or potentiometrically (see p. 81).

Non-aqueous titration of acidic groups is carried out in pharmacopoeial assays of: barbiturates, uracils and sulphonamides.

Theory

Reduction potential is a measure of how thermodynamically favourable it is for a compound to gain electrons. A high positive value for a reduction potential indicates that a compound is readily reduced and, consequently, is a strong oxidising agent, i.e. it removes electrons from substances with lower reduction potentials. The oxidised and reduced forms of a substance are known as a redox pair. Table 3.2 lists the standard reduction potentials for some typical redox pairs.

Table 3.2 Standard reduction potential (E_o) for some redox pairs relative to the standard hydrogen electrode potential 0

A substance with a higher reduction potential will oxidise one with a lower reduction potential. The difference in potential between two substances is the reaction potential and is approximately the potential difference which would be measured if the substances comprised two halves of an electrical cell. For example Cl₂ will oxidise Br⁻ according to the following equation:

Taking values from Table 3.2, the reaction potential is given by: 1.36 – 1.065 = 0.29 V

For the reaction:

The reaction potential is given by: 1.36 – (–2.888) = 4.248 V (i.e. a large difference and calcium burns in chlorine).

In the above examples we have ignored the effect of concentration of oxidant and reductant on E_o values; in fact, E (the observed electrode potential) is stable over a wide range of concentrations. The E-value for a solution containing a redox pair is governed by the Nernst equation:

where [Ox] is the concentration of the oxidised form of a particular substance and [Red] is the concentration of the reduced form of a particular substance:

By substituting a value for the constant terms, this equation can also be written as:

where n is the number of electrons transferred during the reaction. It is clear that E is approximately equal to E_o except when there is a large difference between [Ox] and [Red].

The titration curve for Fe^2 + against Ce^4 + is shown in Figure 3.12. This curve is for a titration carried out with a standard hydrogen electrode as the reference electrode.

Fig. 3.12 Titration of 25 ml of 1 M Fe^2 + with 1 M Ce^4 +.

Where a reference electrode has a reduction potential > 0, then the predicted reading of the potential for a redox pair is obtained by subtracting the reduction potential for the reference electrode, e.g. for an Ag/AgCl reference electrode, 0.223 V is subtracted.

In carrying out redox titrations, standard Ag/AgCl or Hg/Hg₂Cl₂ electrodes are used as a reference in conjunction with an inert redox electrode, e.g platinum, which takes its potential from the particular redox pair in the solution in which it is immersed.

Redox titration is used in pharmacopoeial assays of: ferrous salts, hydrogen peroxide, sodium perborate and benzoyl peroxide by titration with KMnO₄. In the case of KMnO₄ titrations, the end-point may be detected when the purple colour of the permanganate persists.

Direct titrations

Iodine is a moderately strong oxidising agent (see Table 3.1). During oxidation iodine is reduced as follows:

It will oxidise substances with lower reduction potentials, e.g. the titration of ascorbic acid is carried out as shown in Figure 3.13.

Fig. 3.13 Oxidation of ascorbic acid by iodine.

The iodine solution used is standardised against sodium thiosulphate (see later). In addition, the end-point is detected using starch indicator, which produces a blue coloration with excess iodine.

Direct iodometric titration is used in pharmacopoeial assays of: ascorbic acid, sodium stilbigluconate, dimercaprol injection and acetarsol.

Iodine displacement titrations

These titrations involve displacement of iodine from iodide by a stronger oxidising agent followed by titration of the displaced iodine with sodium thiosulphate. For example, the available chlorine in bleach is estimated as follows:

The displaced iodine is then titrated with thiosulphate according to the following equation:

A different approach is used in the estimation of phenols. Bromine is generated by the reaction of potassium bromide with a defined volume of a standard solution of potassium bromate according to the following equation:

The bromine generated is then reacted with the phenol, and 1 mole of phenol reacts with 3 moles of bromine (Fig. 3.14).

Fig. 3.14 Reaction of phenol with bromine.

Excess bromine is used and the bromine remaining after the above reaction is reacted with iodide as follows:

The liberated iodine is then titrated with thiosulphate, thus quantifying the excess bromine.

Iodine displacement titrations are used in pharmacopoeial assays of: liquefied phenol, methyl hydroxybenzoate, propyl hydroxybenzoate and phenidione.

Iodine-absorbing substances in penicillins

A major stability problem in penicillins is the hydrolysis of the lactam ring, as shown in Figure 3.15. Penicillins with an open lactam ring are inactive as antibiotics, since it is the reactive lactam ring which kills the bacteria.

Fig. 3.15 Hydrolysis of the lactam ring in penicillins.

When the lactam ring is open it will react with iodine. 1 mole of the ring-open form of penicillin will react with 8 equivalents of iodine; the intact lactam ring will not react. In this type of titration, excess iodine solution is added to a sample of the penicillin, and the iodine that is not consumed in the reaction is estimated by titration with sodium thiosulphate. The value obtained for the amount of hydrolysed penicillin in the sample, should be no more than 5% of that obtained when all the penicillin in the same amount of sample is completely hydrolysed to the ring-open form and then reacted with iodine. Most of the pharmacopoeial monographs for penicillins indicate that this test should be carried out.

Titrations using indicator dyes

A small amount of an anionic or cationic dye is added to an aqueous solution of the analyte, which is a lipophilic cationic or anionic compound. A small amount of coloured lipophilic ion pair is formed and this is extracted into a small amount of chloroform, which becomes coloured by the ion pair. Titration of the lipophilic anion or cation is carried out with a lipophilic counter ion, e.g. benzethonium chloride or sodium dodecyl sulphate. At the end-point, excess of the titrant breaks up the coloured complex in the chloroform layer.

Ion pair titration using a coloured indicator complex is used in pharmacopoeial assays of: dicyclamine elixir, procyclidine tablets, sodium dodecyl sulphate and cetrimide emulsifying ointment.

Titrations using iodide as a lipophilic anion

This procedure is more widely used in pharmacopoeial assays than the dye extraction procedure. Excess potassium iodide is added to an aqueous solution of the analyte, which is a lipophilic cation. A lipophilic ion pair is formed between the cation and the iodide ion and is then removed by extraction into an organic phase such as chloroform. The excess iodide remaining in the aqueous phase is then titrated in concentrated HCl (> 4 M) with potassium iodate. The iodate oxidises iodide to I⁺, which immediately reacts with Cl⁻ to give ICl, resulting in the following equation:

A small amount of chloroform is used as an indicator and, in the presence of the reaction mixture, it becomes coloured purple by traces of iodine, which are present during titration. The purple colour disappears at the end-point, when the conversion of all I⁻ and I₂ into ICl is complete.

Ion pair formation with iodide followed by titration of excess iodide with iodate is utilised in pharmacopoeial assays of: cetrimide, cetylpyridium bromide, domiphen bromide and benzalkonium chloride.

Potentiometric end-point detection

All of the titrations discussed in the preceding sections can be carried out using a suitable electrode to measure the potential of the solution as the titration progresses. The advantage of making potentiometric measurements in order to detect end-points is that the measurements can be made in solutions which are coloured, unlike indicator-based end-point detection, and give unambiguous end-points where indicator colour changes are not clear or sudden. The disadvantage of potentiometric titrations is that they are relatively slow, since time has to be allowed for readings to stabilise, particularly near the end-point of the titration. However, potentiometric titrations can be automated and potentiometric end-point detection is used in automatic titrators, where the titrant is pumped into the sample under microprocessor control. The electrode that is usually used to make the measurements in potentiometric titrations is the pH-sensitive glass indicator electrode. This electrode consists of a pH-sensitive glass membrane bulb which encloses a phosphate buffer solution containing potassium chloride solution and saturated with silver chloride. The solution is in contact with an internal reference electrode which consists of a silver wire. The circuit is completed by a second reference electrode, which in modern combination electrodes is a second silver/silver chloride electrode, which contacts the external solution via a porous junction (Fig. 3.17). The electrode monitors the variation in the potential difference, which is largely caused by the interaction of hydrogen ions with the outer surface of the pH-sensitive glass membrane.

Fig. 3.17 pH-sensitive glass electrode.

The potential which develops on the inner and outer glass surfaces of the electrode is due to the following equilibria:

The number of Gl⁻ sites on the outer membrane increases with decreasing [H⁺] and thus its potential becomes increasingly negative with respect to the inner surface with increasing pH. The Nernst equation can be simplified and written in the following form for the glass electrode when the temperature is 20°C:

where E is the measured potential in volts and Ek is a constant composed of the sum of the various potential differences within the electrode, which do not vary appreciably. The combination electrode is constructed so that its potential is ca 0 V at pH 7.0. It can be seen from the equation above that E changes by 59.1 mV for each pH unit.

When potentiometric titration is carried out, the volume of titrant added is plotted against the measured potential. Since the electrode takes time to equilibrate, the volume of titrant required to reach the end-point is first calculated and a volume of titrant is added to within ca 1 ml of the end-point. Then the titrant is added in 0.1 ml amounts until the steep inflection in the titration curve is passed. The end-point of the titration is the point where the slope of the titration curve is at its maximum. Thus, if dE/dV is plotted for the titration, the maximum of the plot gives the end-point. The end-point can also be determined from the mid-point of the inflection in the titration curve or from the tabulated data (Table 3.3). Figure 3.18 shows a curve for the titration of 2 mmoles of aspirin with 0.1 M NaOH. The end-point corresponds to the mid-point of the inflection or, if the tabulated data are examined, it can be taken to be the mid-point between the two volumes, where dE/dV is greatest, i.e. at 20.05 ml between 20 and 20.1 ml. The actual end-point for exactly 2 mmoles of aspirin titrated with 0.1 M NaOH should be 20 ml; addition of 0.1 ml aliquots towards the end-point means that the end-point is only accurate to within ± 0.05 ml.

Table 3.3 Potential difference values obtained for titration of 2 mmoles of aspirin (pKa 3.5) against 0.1 M NaOH

Fig. 3.18 Titration of 20 mmoles of aspirin (pKa 3.5) with 0.1 M NaOH. The end-point corresponds to the mid-point of the inflection.

Use of potentiometric titration to determine pKa values

Potentiometric titration provides the principal method for determining pKa values and it is best applied to substances with pKa values < 11. For example, the pKa of benzoic acid can be determined as follows: a 0.01 M solution of benzoic acid (50 ml) is titrated with 0.1 M KOH. The KOH is added in 0.5 ml increments; it would be expected that 5 ml of 0.1 M KOH would be required to neutralise the benzoic acid. The pH of the titration is monitored with a glass electrode and the pH of the mixture after 2.5 ml of 0.1 M KOH has been added will equal the pKa value of benzoic acid, since:

The pKa value may be checked after addition of each 0.5 ml, since the concentrations of acid and salt are known at each point on the titration curve. The slight increase in volume due to the addition of the 0.1 M KOH may be ignored. For a base, the Henderson–Hasselbalch equation is written as given in Chapter 2, page 29. Automatic titrator software will evaluate the titration curve and report half-neutralisation points. Figure 3.19 shows a potentiometric titration curve for the amino acid glycine which has two pKa values.

Fig. 3.19 Determination of the pKa values for glycine using potentiometric titration.

Automatic titration (Fig. 3.21)

Titrations can be automated and controlled by a microprocessor. The titrant is delivered via an automatic burette and the end-point is detected potentiometrically with a glass combination electrode. Alternatively, if ions other than hydrogen are being measured, another ion-selective electrode may be used. The apparatus is microprocessor controlled and can be programmed to run in various modes:

Fig. 3.21 Automatic titration apparatus.

The microprocessor control also enables the instrument to be set to calculate pKa values directly from the pH profile it obtains by titration of a sample. A sample changer can be incorporated, so that several samples can be automatically titrated.

Flow injection analysis

Flow injection analysis (FIA) represents a refinement of wet chemical methods. The basis of the technique is that the sample is injected into a continuously flowing stream of reagent. The sample reacts with the reagent and this reaction is measured with a detector. The range of detectors available is the same as that which is used in conjunction with HPLC (Ch. 12, p. 322) except that there is no chromatographic separation involved. Thus the technique is not as selective as chromatographic methods and its selectivity is dependent on the specificity of the reaction between the analyte and the reagent or the property used for detecting it. A simple schematic diagram of a flow injection analysis system is shown in Figure 3.22. The basic setup may be modified to include several manifolds that allow the introduction of the sample followed by additional reagents. The advantages of the technique are its cheapness and rapidity.

Fig. 3.22 A simple flow injection analysis system.

A precise volume of sample (1–100 μl) is injected and passed through the incubation coil, which is of sufficient length to allow the sample to disperse in the reagent but not long enough for the sample to become diluted so much that the integrity of the plug of sample is lost. The detector response is dependent on the degree of dispersion of the sample. A typical flow of the reagent + analyte is as shown in Figure 3.23.

Fig. 3.23 Segmented flow in flow injection analysis (FIA).

The parameters which have to be optimised include:

Since a number of factors are involved in the optimisation, some time is required to develop the method. However, when set up the method can replace titrations, and replicate analyses can be conducted very quickly with minimal consumption of reagents.

As in chromatography, the ideal peak shape obtained in FIA should be Gaussian, although in practice the ideal shape may not have time to develop. The mathematics governing the dispersion processes have been developed thoroughly and the process is largely analogous to the dispersion occurring in capillary gas chromatography, where longitudinal diffusion is the major factor governing band broadening.

Figure 3.24 illustrates an application of FIA to the Karl Fischer titration. The consumption of the reagent by water is detected spectrophotometrically by monitoring the stream of reagent at 615 nm. The absorbance due to the iodine in the reagent is removed by its reaction with water, which causes formation of iodide and thus negative absorbance is measured.

Fig. 3.24 Calibration of Karl Fischer flow injection analysis (FIA) method for determination of residual in water in pharmaceuticals. The analyte was methanol containing 0.05, 0.1, 0.2, 0.4 and 0.5% w/v water.

Determination of chloroxine

The antibiotic chloroxine was determined utilising the formation of a complex between the drug and Al^3 + in an FIA system. The complex was determined by measurement of fluorescence with 399 nm as the excitation wavelength and 496 nm as the emission wavelength. In order to ensure solubility of the complex in the aqueous reagents, a surfactant was included in the reagent mixture. The precision of the method was greater than that obtained using a laborious batch method for measuring samples manually using a fluorescence spectrophotometer.³

Determination of captopril

An FIA method for the determination of captopril was based on the oxidation of the thiol group in the molecule by Ce^4 +. This reaction results in the emission of light (chemiluminescence), which can be measured. In this example the dye rhodamine G was used to enhance the emission of light by the reaction. The method developed was rapid and precise.⁴

Determination of non-steroidal anti-inflammatory drugs

Diclofenac sodium, famotidine and ketorolac were analysed utilising their formation of a coloured charge transfer complex with 2,4 dichloro-6-nitrophenol. The complexes were detected by UV/visible spectrophotometry at 450 nm. The method was not affected by the presence of common excipients in the formulations analysed. The precision and accuracy of the method was comparable to that of HPLC methods used to analyse the same samples.⁵

Determination of promethazine

The generation of a coloured product upon the oxidation of promethazine with Ce^4 + was used in the development of an FIA method. Promethazine in tablet form could be analysed by this method with a precision of ± 1% and at a rate of 122 samples per h.⁶ In a similar method, promazine was oxidised by passing through a short column containing MnO₂ and then the oxidation product was measured.⁷

Determination of chlorocresol

Chlorocresol is a preservative commonly used in injections and its determination often involves the use of laborious extraction procedures in order to separate it from formulation components, followed by spectrophotometric measurement. An FIA method for chlorocresol was developed by utilising its reaction with nitrous acid to form a coloured nitro compound. The method was accurate to 99.5% of the true value of chlorocresol in a formulation and a precision of ± 1% was achieved.⁸

Limit test for heavy metals

Many pharmacopoeial monographs contain a limit test for heavy metals. Sometimes the metal is specified, e.g. lead, but often the test is more general. Pharmacopoeial tests often involve precipitation of the metals as their sulphides. An FIA method was developed based upon complex formation between heavy metals and diethyldithiocarbamate (DDC). Figure 3.25 shows the FIA system used for this analysis and illustrates how relatively simple components can be assembled to carry out a complex analytical task. The analysis was achieved by using a segmented flow system where alternate segments of buffer solution + reagent and carbon tetrachloride were produced. In the first extraction coil, the heavy metals in the sample are extracted as their complexes, along with some excess complexing agent, into carbon tetrachloride. In the second extraction coil, the excess reagent in the organic layer is back extracted by the borax solution, which is mixed into the carrier stream. The flow was then passed into a phase separator, which only allowed the organic solvent to flow through to the detector.⁹

Fig. 3.25 Flow injection analysis (FIA) system used to analyse heavy metal residues by extraction of their complexes with DDC into carbon tetrachloride.

Use of segmented flow in determination of partition coefficients

A system similar to the one described above was used for determination of partition coefficients. An FIA system with segmented flow was devised so that the partitioning of a drug between aqueous buffer and chloroform could be measured. The aqueous and organic phases were separated using a phase separator. The system could be set up to measure the concentration of the drug in either the organic or the aqueous phase. Such a system enables rapid repeat determinations of partition coefficient at various pH values with minimal sample consumption.¹⁰

Automated dissolution testing

FIA was used to optimise sampling from a tablet dissolution apparatus in order to determine the rate of release of iron (II) from a sustained release formulation. The dissolution medium was automatically sampled at 30-minute intervals and the 100 μl aliquots of medium were mixed with the iron-complexing agent ferrozine, diluted and then passed into a spectrophotometric detector. The system was microprocessor controlled, thus enabling unattended sampling of the dissolution medium for a prolonged period.¹¹