January 19731 ORIGISAL PAPERS IN PHXRMACECTICAL ANALTSIS Original Papers in Pharmaceutical Analysis 9 The following are summaries of three of the papers presented a t a Xeeting of the Joint Pharmaceutical Analysis Group held on June 8th 1972. Polarography as a Method of Quality Control in the Assay of Selected Pharmaceutical Products BY W. FRANKLIN S~II’TH (Cheniistvy Depavtment Chelsea College University of Londo+i 41anvesa Road L o d o n S. LV.3) -1 S U R V E ~ of the literature relating to the polarographic assay of compounds of pharma- ceutical importance will reveal the frequency of use of the technique in the 1980s with a rapid decline in the following decade. Perusal of more recent literature reveals however a renewal of interest in the applicability of polarography not only to formulation analysis but particularly to low-level monitoring of certain drugs and their metabolites in body fluids.The usefulness of polarography in quality control hitherto relatively unappreciated is illustrated by the following facts. 1 . Nany active constituents of formulations in contrast to the other ingredients are electroactive and can be oxidised or reduced under polarographic conditions e.g. compounds containing functional groups such as thiol azoniethine (which as well as existing isolated in a ring system can be formed by enolisation of carbonyl or thiocarbonyl functions in the x-position to the nitrogen atom) carbonyl thiocarbonyl and various valency states of nitrogen such as nitro and X-oxide. The method possesses an appropriate selectivity that is particularly important in tlie differentiation of mixtures when the active constituents are electroactive.An elegant example of this appeared recent1y.h the literature with the simultaneous determination of vitamins B and B and Niacinamide in a multi-vitamin preparation by cathode-ray polaro- graphy.l This selectivity can also be used to determine the degree of decomposition in those products where the breakdown product of the active material is reduced a t a different potential. 3. The method designed for the selected pharmaceuticals mentioned below is rapid and does not involve any separation step. The active constituent is dissolved out of the formulation into a suitable solvent medium (a hydrochloride can be quickly dissolved in water - methanol and many others in an aprotic solvent such as dimethyl formamide acetonitrile or dimethyl sulphoxide) while being ground up in a mortar.h n aliquot of the mixture is then sampled dissolved in an appropriate medium which may be a traditional aqueous supporting electrolyte or the aprotic solvent - tetraalkylammonium salt and the current - potential curve recorded after deaeration. By reference to a pre-determined calibration graph or by the standard addition method a single or multi-tablet assay can be carried out in a minimum of 10 minutes with an error of 2 to 3 per cent. well within the limits allowed by the British Pharmacopoeia or National Formulary. The ability to carry out polarographic measurements in an entirely non-aqueous medium increases the number of preparations that can be assayed by this means. 5J’- Disubstituted barbiturates such as for example phenobarbital are electroinactive under aqueous conditions but are reduced in a different way in aprotic media resulting in well defined reversible waves.A wide variety of electroactive compounds of pharmaceutical importance have been subjected to polarographic study and all have given waves suitable for analytical purposes. This includes hypnotics and tranquillisers such as the barbiturates mentioned above 1,4-benzodiazepines (nitrazepam medazepam chlordiazepoxide diazepam flurazepam oxazepam and lorazepam) and quinazolines (methaqualone). Further additions to the list are intravenous anaesthetics such as thiobarbiturates antibiotics such as the tetracyclines (tetracycline hydrochloride oxytetracycline chlortetracycline) and some steroid hormones such as testosterone.Very common reducible functional groups encountered in this series are azomethine and carbonyl two which seem to appear frequently in molecules of biological significance. Many of these compounds possess non-aromatic heterocyclic ring systems and their behaviour parallels that of other similar organic substances being modified only by 2 . 4. 10 [Proc. SOC. -4 ualyt. Chevz. the presence of the heterocyclic ring. In addition to this concept there is one additional factor that occurs with most heterocycles-their adsorbability on the mercury drop either of the depolariser itself or of reaction intermediates or products This can complicate the wave pattern in aqueous media in the concentration range to 10-~ RI (frequently adsorption effects are absent in an entirely aprotic medium) but is an added advantage a t the lower concentrations encountered in body fluid analysis.This has recently been exploited2 by polarography carried out directly in blood serum containing “spiked” benzo- diazepines and obtaining measurable adsorption currents even in the presence of the other biological constituents. The method of approach used to ascertain ideal conditions for the assay of a formulation involves the determination of the polarographic behaviour of the active constituent(s) in the pH range 2 to 12 in Britton-Robinson buffers. Non-aqueous niedia are used only when no or badly defined waves are observed in aqueous solutions. The limiting current is plotted against pH and the pH region where the best defined and/or largest wave exists is chosen as the medium for the assay.Graphs are constructed of the limiting current iwsus concentra- tion in the range M a t this optimum pH and linearity is generally observed. The formulation is ground down dissolved in an appropriate solvent medium an aliquot diluted with buffer and the current - potential curve recorded. Small amounts of non- aqueous solvent (10 to 20 per cent.) in the sample solution rarely affect the height or linearity of the calibration graph determined in completely aqueous media. Several of the benzo- diazepines and tetracyclines have been tested by the outlined method and a good correlation between the expected content and the recovered content of the formulation has been found. ,4 suitable cell system has been designed that can be used by the quality control analyst in conjunction with standard polarographic equipment so that the criteria of selectivity simplicity and speed of assay are satisfied as far as possible.Work is in progress to operate this on a semi-automatic basis. ORIGINAL PAPERS IN PHARMACEUTICAL ANALYSIS to REFERENCES 1. 2. Schertel JI. E. and Sheppard A. J . J . Pharnz. Sci. 1971 60 1070. Halvorsen S. and Jacobsen E. A mlyticn Chinz. <4cta 1972 59 137. An Automated Method for the Determination of Thiabendazole BY T. D. \VHITE (Quality Copztrol Depai,tment. Mevck Shavp 6- Dohwic LZ‘iiiitrd Hoddesdon Hevtfovdsiiire) THE official A.O.A.C. method for determination of thiabendazole [2-(4‘-thiazolyl)benz- imidazole a pharmaceutical and veterinary anthelmintic and agricultural fungicide in animal feedstuffs was described by Szalk~wski.l-~ This method has been adapted as an automated method using Technicon AutoAnalyzer I equipment.This colorimetric method is less prone to interference than ultraviolet spectrophotometric measurement and requires less complex preparation than fluorimetric measurement. Other compounds that contain the thiazole ring interfere. A flow diagram is shown in Fig. 1. Finely ground feedstuff is shaken with ethyl acetate and the filtered extract re-extracted with 0.1 s hydrochloric acid for presentation to the system a t a nominal concentration of 5 pg ml-l. In the system the acidic solution of thiabendazole is reduced with a 0.3 per cent. slurry of zinc dust in glycerine and the products are coupled with a 0.17 per cent. solution of P-phenylenediamine dihydrochloride in 2 s sulphuric acid and a 15 per cent.solution of ammonium iron(II1) sulphate in 0.1 x sulphuric acid forming a blue complex whose absorbance is measured at 610 nm in a 15-mm flow cell During reduction the use of glass transmission tubing and the minimisation of mixing reduces aggregation of the zinc particles which is deleterious to consistent results. Complete reduction of thiabendazole concentrations to a t least 10pg ml-1 is achieved. Excess of zinc is removed by filtration through a glass-wool plug after the reduction stage. Reduction is carried out in an 80-foot glass delay coil. January 19731 ORIGINAL PAPERS IN PHARMACEUTICAL ASALYSIS 11 Owing to the high viscosity of the stream the normal polythene capillary connection between the colorirneter de-bubbler and the flow cell is replaced with standard transmission tubing.The system is calibrated by using thiabendazole standard solutions a t 2 4 6 and 8 pg ml-l and samples are analysed a t the rate of twelve per hour ( 2 minutes sampling and 3 minutes wash interval). Resolution is 0.1 pg ml-l and the coefficient of variation of ten assays of a single solution is 1.2 per cent. I 2.5 ~ 0 . 1 N HCI Sample 2.0 Air Resarnple Amine reagent 0.6 1.6 - - Iron( I I I) reagent Waste 4 - 7 Recorder Colorimeter 15-rnrn flow cell 610 nrn Fig. 1. Flow diagram 4 range of different commercial feed samples both unmedicated and containing other additives mixed with thiabendazole gave recoveries ranging from 95 to 100 per cent. at a 1 per cent. medication level. The original work on the manual method indicating low levels of interference was supported.For most formulations of thiabendazole ultraviolet spectrophotometric methods are available so that while of general applicability this method is likely to be mast useful for medicated feeds and possibly for the assay of residues in fruit currently involving a manual multiple-extraction procedure followed by fluorescence measurements. The manual colorimetric method is sensitive to changes in conditions and rigid adherence to the specified procedures is essential for accurate and reproducible results.2 The use of the AutoAnalyzer ensures uniformity reduces operztor bias and leads to an increased rate of sample processing although it is relatively slow for an automated chemical assay. REFERENCES 1. 2. 3. Szalkowski C. R. J - 4 s ~ . O f l . Agric. Chem. 1964 47 236 Szalkowski C .R. and Kanora J . I b i d . 19G5 48 288. Szalkowski C . R. I b i d . 1966 49 312. Permittivity Measurements Applied to Aqueous Solutions BY P. I. HARRISOP; (St Mary’s Hosfiital Pvncd Street London W.2) l l . 4 ~ 1 ~ medical and pharmaceutical techniques can benefit from analytical methods that can be incorporated into “in-line’’ systems. Ideally there should be no direct contact between the solution and the measuring apparatus. This was one of the criteria that stimulated the work outlined below. Permittivity (dielectric constant) is a measure of the electric dissymmetry of a dipole. A polar molecule placed between the plates of a charged condenser will rotate to align itself 12 ORIGINAL PAPERS IP; PHARMACEUTICAL XSALPSIS [Proc. SOC. Analyt. Clzenz. with the electric field.If the frequency is increased a region is reached when rotational drag and inertia limit complete alignment. D‘hen the frequency reaches the “microwave region,” it will often exceed the rotational frequency of the molecule. In the visible region only electrons can move sufficiently rapidly to be influenced by the field In addition to being frequency dependent (exhibiting dispersion) the permittivity of a system of molecules may be expected to change under other conditions that alter the freedom of movement of the molecules or their chemical nature for example pressure temperature solx7ent interaction viscosity and pH. This latter example in an extreme form may produce changes in the structure of the molecule (in the case of zwitterions). The capacitance of a condenser is a function of the permittivity of the molecular system between the plates of a condenser (i.e.the dielectric). In an alternating field energy may be dissipated. The permittivity of water (liquid state) decreases in the region of 10 000 NHz. At visible frequencies permittivity approximates to the square of the refractive index for most polar liquids. Bernal and Fowler1 suggested three possible structures for water a tridyinite-like structure for ice a quartz-like structure up to 200 ‘C and a close-packed ammonia structure at higher temperatures. Pauling2 proposed a pentagonal dodecahedra1 hydrate around a “guest” water molecule and based a general theory of anaesthesia on the formation of such hydrates in the cerebro-spinal fluid. The effects of the addition of electrolytes to water are explained in terms of whether or not the ions in question fit into the interstitial spaces and cavities and how much of their presence alters the lattice.Frank and \Ven3 postulated that hydrogen bonds would be continually forming and breaking giving groups of molecules with very brief half-lives. Drost-Hansen4 has produced evidence to show that many parameters including density specific heat viscosity permittivity and refractive index show anomalies at certain tempera- tures. To verify this measurements were made on water at 500 NHz and the temperature raised from 0 to 55 “C which is equivalent to lowering the frequency. The results showed a peak of dielectric loss a t about 40 “C which could be due to an alteration in water structure. The water was distilled sterilised and finally filtered into the cell through a 0.22-pm membrane filter to minimise spurious results due to micro-organisms.When a cation is added to water electrostriction occurs the water molecule crowding about the ion resulting in a smaller molar volume of water in the neighbourhood of the ion. As the concentration of the electrolyte is increased the various zones with differing degrees of order will successively overlap giving rise to the reversal of certain solution properties as observed by Good.5 An increase in pressure breaks up the local water structure near ions stripping away their hydration atmospheres and reducing their radii. An extreme case is that of a simple dehydrated ion moving through a non-associated medium. The plates of a condenser are effectively electrodes and a t low frequencies in particular are subject to electrode polarisation and unless four electrode cells or special switching bridges are used spurious results will be obtained.\Vhile Ermakov‘j has made conductance measurements in the 50 MHz region to differentiate between different ions it seems unlikely that this will form a basis for quantitative analysis. Measurements of conductivity and permittivity in this region can however be made without contact between the solution under test and the measuring electrodes. For 1 1 electrolytes conductivity plays a part up to about 2000 NHz but a t 1000 JIHz it is significantly concentration dependent. The 2 2 electrolytes show an unsymmetrical double dispersion of permittivity similar to the amino-acids. The lower frequency dispersion has been attributed to the relaxation of ion pairs.Studies of the health risks due to microwaves from radar and other equipment involve a knowledge of the permittivity and dielectric loss of tissues. lleasurements on niicro- organisms have permitted the calculation of their membrane thickness. Permittivity measurements are now being used in medicine to study the relationship between heart disease and abnormal lipoprotein^.^ The degree of hydration is sensitive to the variation in the shape of the lipoprotein molecules and this is the basis of a study of the relationships of abnormal lipoproteins associated with hyperbetalipoproteinaemia for these lipoproteins tend to be more strongly hydrated. This is shown as dielectric loss which is related to specific conductance. The permittivity of an amino-acid falls between 10 and 1000 RIHz.Many different structures probably exist simultaneously. January 19731 PARTICLE SIZE ANALYSIS 13 In addition to estimating from permittivity measurements the moisture content of powders the droplet size of water-in-oil emulsions can also be found. X further use is in the study of molecular behaviour. If functional groups are free t o rotate they will usually show a separate dispersion a t higher frequencies. This work was carried out in the Department of Pharmacy Chelsea College University of London Nanresa Road London S.W.3. The author acknowledges the assistance given by Professor Jonscher Department of Physics Chelsea College. REFEREXCES 1. 2 . 3 . 4. 5 . 6. 7 . Bernal J . D. and Fowler R. H. ,I. Chem. Plzys. 1933 1 515 Pauling L. Science 3-.1’. 1961 134 15 Frank H. S. and \$:en 1s’. Y . Discuss. Favadu-y Soc. 1955 21 133. Drost-Hansen W. Adv. Clzem. Sev. 1967 67 i 0 . Good W. E’lectvochiun. Acta 1964 9 203. Ermakov V. I. Sniirnov K. I. and Zagorets P. O. Russ. J Pizys. Chef??. 1963 37 280 Grant E. H. Sheppard K. J . Mills G. L. and Slack J . Lancet 1972 i 1159.