Catalytic Methods in Analytical Chemistry G. SVEHLA Department of Analytical Chemistry Queen's University Belfast Contents Introduction Quantitative catalytic analysis Classification of the methods-Analytical classification -Chemical classification -Kinetical classification -Experimental classification Determination of metals-Aluminium -Barium -Beryllium -Bismuth -Calcium -Chromium -Cobalt -Copper -Germanium -Iron -Magnesium -Manganese -Mercury -Molybdenum -Nickel -Osmium -Ruthenium -Silver -Strontium -Thallium -Tungsten -Vanadium -Zinc 235 Determination of non-metallic substances and anions-Bromide -Cyanide -Hydrogen peroxide -Iodide -Oxygen -Phosphate -Selenium -Sulphide Catalytic and kinetic determination of organic substances-Amines -Ascorbic acid -0-Chloronitrobenzene -Complexing agents -Dihydric phenols -Esters -Ethanolamides -Ketones -Nap11 thols -Oxidative enzymes -Pesticides -Phenols -Serum enzymes and their substrates Qualitative catalytic analysis Application of catalytic waves in polarography Kinetochromic spectrophotometry Catalymetric titrations New techniques and instruments Theory of catalytic analysis reviews 23 CATALYTIC METHODS IN ANALYTICAL CHEMISTRY 237 Introduction Catalytic or kinetic methods of chemical analysis are based on the well known fact that small amounts of catalysts can initiate or accelerate certain chemical reactions.Some reactions take place or are accelerated only in the presence of the catalyst and by the kinetic examination of the reaction the presence or absence of the catalyst substance can be ascertained by qualitative tests.As the amount of catalyst required for the initiation or acceleration of the reaction in usually very small these tests are very sensitive. The selectivity and specificity of such tests can vary according to the nature of the reactions involved; in some cases the catalytic action is specific in others a number of chemically related species e.g., transition metals show almost the same catalytic action and the selectivity of the tests is poor. The importance of qualitative tests made by catalytic methods is far surpassed by the importance of quantitative determinations which may offer cheap and sensitive alternatives to methods that require expensive instrumentation.These quantitative determinations are generally based on slow reactions proceeding in solution containing a suitable catalyst. In the absence of the catalyst the slow uncatalysed reaction proceeds alone; if the catalyst is added a new route for the over-all reaction in which the catalyst also takes part becomes available. The catalyst is then regenerated by another reaction that is the catalyst is involved in a reaction cycle. There are therefore two competing reactions proceeding at the same time the uncatalysed reaction and the catalysed one. The reaction rate depends on the contribution of these competing reactions to the over-all process. If the concentration of the catalyst is too low the share of the catalysed reaction in the over-all process becomes negligible and therefore there is no measureable change in the over-all reaction rate compared to that obtained in the absence of the catalyst.If on the other hand the concentration of the catalyst becomes so high that the contribution of the uncatalysed reaction to the over-all process becomes negligible and the catalytic reaction becomes predominant the rate of the over-all process becomes very high and somewhat insensitive to changes in the concentra-tion of the catalyst. Between these two extremes however there exists a con-centration range in which altering the concentration of the catalyst produces changes in the contribution of the two competing reactions to the over-all process. Accordingly the reaction rate will vary with the concentration of the catalyst.By measuring the reaction rate as a function of the catalyst concentration a calibra-tion curve can be obtained and used for practical analytical purposes. This review deals with the latest development of catalytic methods of analysis. The literature of the period January 1967 to June 1970 is covered. The earlier literature is well reviewed in the excellent monograph of Mark and Rechnit2.l Another monograph on the subject by Yatsimirskii2 contains selected procedures for the determination of certain ions and can also be used with advantage. Because of the importance of quantitative determinations these methods will be reviewed first and in most detail. This is followed by a short discussion of new Q 238 SVEHLA qualitative catalytic tests.The application of catalytic waves in polarography is followed by reviews of kinetochromic spectrophotometry and of catalymetric titrations. The review is concluded with accounts on new techniques and instru-ments and on the development of the theory of catalytic methods. Quantitative Catalytic Analysis Classification of the methods The literature of quantitative catalytic analysis is so vast and is growing so fast that it is very important to classify the various methods into separate groups. This classification can be made from various points of view. In this review the material is arranged so that it is most suitable for a practising analyst who wishes to choose a method for his particular purpose. This is based on the chemical nature of the analyte.The chemical classification is based on the nature of reactions involved in the method. A kinetical classification on the other hand is based on the principle involved in the measurement of the reaction rate while an experi-mental classification can be made according to the techniques used in measuring the concentration of the reactants. As the principles of quantitative catalytic methods can be elucidated most conveniently according to these principles a more detailed description of these classes is given. Analytical classification. From the practical point of view it is found convenient to classify the substances that can be determined by quantitative catalytic methods into three groups metals non-metallic inorganic substances and organic substances.Within each group substances are dealt with in alphabetical order. Chemical classification. This method of classification is based on the type of chemical reaction that is involved in the process. Thus we can distinguish between (a) redox (or electron-exchange) reactions (b) ligand-exchange reactions, (c) enzyme-catalysed reactions and (d) catalysed electrode reactions. Though the last of the classes might fall into category (a) it is logical to treat them separately as they are applied for quite special purposes. Redox reactions are by far the most common processes among catalytic reactions. In these reactions two redox couples take part. Considering homo-geneous redox systems only these can be characterised with the equilibria and and with the oxidation reduction potential El and E for systems 1 and 2 respec-tively.If El > E a reaction between Ox and Red will take place-n,Ox + n,Red + 2(m,n,-m2n,)Hf -+ %,Red + n,Ox + (m,n,-m,n,)H,O If lzl and/or n are higher than 1 more electrons have to be exchanged during the process. Such a process however cannot take place in one single step and so the mechanism of such processes includes a series of reactions with the transfer of Ox + 2m,H+ + n,e- =+ Red + m,H,O Ox + 2m,H+ + n,e- + Red + m,H, CATALYTIC METHODS IN ANALYTICAL CHEMISTRY 239 one electron in each step. Some of these steps axe often slow. If a catalyst is present which reacts with one of the original reactants in such a way that an alternative reaction mechanism becomes available the over-all reaction rate increases considerably.This mechanism involves the product of the above reaction reacting with the other reactant one of the products of this process being the original form of the catalyst. Thus the catalyst is involved in a reaction cycle. As the catalytic reaction is also an oxidation - reduction reaction the catalyst itself must be a part of a redox system with an oxidation-reduction potential lying between those of the two reactant systems. This is the reason why the transition metal ions are so often found to be good catalysts of such reactions. Ligand-exchafige reactions represent a new so far rather neglected field of catalytic analysis. It is known that many ligand-exchange reactions between complexes MfL1 and M2L2 (where M1 and M2 represent metals L-l and L2 ligands) are slow-MIL1 + MaLB+MIL* + MaL1 The reason for this is that these over-all reactions take place as a series of dis-sociation and re-combination steps such as-MIL1 + MI+ L1 MaL8 +t M8 + L8 M1+ L* +MIL' M* + L1 + MaL1 Because of the high stability of the original complexes the over-all reaction is slow.Any of the free metal ions Mf or M2 or the ligands L1 or L2 will catalyse the reaction. If for example a small amount of the free ligand L1 is available the reactions L1 + MaL* 3 M4L1 + L* and L2 + MIL1 3 M1L2 + L1 will proceed and the over-all reaction rate increases. The ligand L1 is regenerated and a new reaction cycle can be initiated. Only a few of these systems have been investigated in the past but there is no doubt however that ligand-exchange reactions represent a potentially important field of catalytic analysis.Enzyme-catalysed reactiorts are frequently used in clinical food and pharma-ceutical laboratories mainly as a preliminary treatment for analysis. They can be used directly however for kinetic determination of the enzymes themselves or of the substrates activators or inhibitors. These reactions proceed in two steps. First the enzyme E and the substrate S form a transitional species ES in an equilibrium reaction-E + S t E S The transitional species then decomposes in a slow reaction to form the product P, and the enzyme is re-combined-ES+P+E The concentration of both the enzyme and substrate influence the reaction rate as do certain activators and inhibitors. Thus there are many possibilities for th 240 SVEHLA analytical applications of enzyme-catalysed reactions.One of the great advantages of enzyme-catalysed reactions is their specificity both towards the enzymes themselves and towards the activators and inhibitors. As enzymes of guaranteed purity can be purchased easily the importance of such reactions is growing steadily. CataZysed electrode reuctions known as the catalytic currents in polarography, are themselves redox processes but as they occur on the electrode surface and therefore need special equipment it is better that they are treated separately. To understand the nature of such processes let us consider the polarographic reduction of a catalyst (most often a metal ion) if present alone (with a suitable supporting electrolyte) If the potential applied is adequately chosen the reduction will proceed on the electrode giving rise to a diffusion current id which is governed by the Ilkovic equation.If however another substance A is present which re-oxidises the product of the above reaction C(m-n)+ + A +- Cm+ + B (where B is another reaction product) the current will increase considerably. (Note that neither A nor B is reduced or oxidised on the electrode at the given potential.) This catalytic current i, depends on the concentration of the catalyst in a rather complex way”; by its measurement the concentration of the catalyst can be determined by using calibration curves. It is not therefore the reaction rate that is measured and none of the reactions involved need be slow. The process can be considered however as the ‘catalytic’ reduction of A to B on the electrode and in the absence of C the reaction would not be thermodynamically possible.The product B is often hydrogen gas and therefore the role of the catalyst is the decrease of the overvoltage of hydrogen on the electrode (mainly the drop-ping mercury electrode). The role of the catalyst is in the initiation rather than the acceleration of the reaction. References to the above classes are frequently made in this review. Kinetical classification. Following the rate of any chemical reactions entails the measurement of two parameters concentration and time at certain intervals while keeping the temperature constant. The results with the aid of correlations of chemical kinetics can then be interpreted and used for analytical purposes.Although almost each author suggests an individual method of measur-ing the reaction rate these methods fall broadly into four groups; the initial rate method ; fixed Concentration methods ; fixed time methods and the simultaneous comparison method. The initial rate method is recommended by a number of authors for reactions where only the initial period of the reaction can be interpreted simply. This is the case if the product of the slow process is involved in a second slow reaction Le., consecutive reactions occur or if as the reaction product builds up the reaction is reversed Le. the reaction leads to an equilibrium. By plotting the concentration of one of the reactants or products as a function of time only the initial part of this Cm+ + ne- j C(m-n) CATALYTIC METHODS IN ANALYTICAL CHEMISTRY 241 curve will be meaningful as regards evaluation (though it is necessary to take measurements over a longer period so as to reproduce the initial part of the curve in a proper way) By definition the slope of such curves taken at zero time is equal to the initial rate.This initial rate depends linearly on the concentration of the catalyst and the latter can be evaluated easily from a calibration graph. The weakness of the method lies in the difficulty in accurately drawing or measuring the initial slope. The fixed reaction time method involves the measurement of the concentra-tion of a reactant or product after a pre-determined lapse of time. If the order of the reaction and the rate constant are known these results can easily be inter-preted kinetically.However even if the kinetics of the reaction have not been investigated the results can be applied for analysis provided that all the experi-mental circumstances are kept constant and the only parameter varied during the experiments is the concentration of the catalyst. The calibration graph where the measured quantity (which is proportional to the concentration or the decrease or increase of concentration of a reactant or product) is plotted against the concentra-tion of the catalyst is a straight line in most cases. These calibration curves can be used in the analysis of unknown samples. The fixed concentration method is in many ways similar to the previous one, but instead of measuring the concentration after a fixed time the time is measured during which the concentration of a reactant or product changes to a pre-deter-mined value.These results are again easy to interpret kinetically if the order of the reaction and the rate constant are known; the integrated rate equations contain all of the parameters measured but again even if the kinetics of the reaction have not been investigated the method may be used for analytical purposes provided that all experimental circumstances are kept constant with the exception of the con-centration of the catalyst. Plotting the reciprocal values of these time values (the so-called reaction times) against the concentration of the catalyst a straight-line relationship is obtained in most cases and this calibration curve can be used in the analysis of unknown samples.One special form of the fixed concentration method is the application of Landolt reactions for analysis. As the method is based on the measurement of time it is often referred to as a ‘chronometric method.’ The method of simultaneous comparison is a simple technique that is suitable if the reaction is accompanied by a visible change of colour or optical density. A set of reagent solutions is prepared and placed into identical beakers or test-tubes. One of these contains the unknown sample while the others contain known amounts of the catalyst. With a special device a so-called ‘starter pipette,’ equal amounts of the other reagents are added to each of the solutions mixed and the colours of the solutions watched.As the time goes on the differences between colours of solutions with different amounts of catalyst in them becomes more and more apparent. The colour of the solution with the unknown concentration is compared to those of the others. The unknown concentration will be equal to the concentra-tion of the catalyst in the vessel that has a similar colour to that containing the sample. By repeated experiments in which the concentrations of the compariso 242 SVEHLA solutions are chosen according to the findings of the first one the concentration of the unknown solution can be reliably determined. The accuracy of the measure-ment is highly dependent on the reliability of the starter pipette and the skill of the operator. As quantitative measurements are not made the results are somewhat subjective and they cannot be interpreted kinetically.Experimental classification. As already mentioned catalytic analyses are based on the measurement of concentration and time. There is little variation in the methods for the measurement of time a stop-watch generally being used. There are several methods for the measurement of concentration and the instru-mentation differs substantially from one to another. The main methods of measuring the concentration are visual photometric potentiometric and gas volumetric although some special methods cannot be so classified. A visual method is the simplest and cheapest as it does not need any instru-ment. Naturally one cannot measure concentrations by the eye and therefore the visual method is restricted to certain but important special determinations.The method of simultaneous comparisons discussed above applies visual examination and provides a high degree of reliability. The Landolt reaction method uses a rapid colour change as the indication of the end of the period to be measured and this can be seen easily by the human eye. Some of the older catalytic techniques which are not included in the present review applied a visual technique when the time of complete decolourisation of a solution had to be measured. These methods, though simple offered only a very limited precision and have merely historical importance. The photometric method of monitoring the concentration of a reactant or product offers a high degree of precision. The solutions are mixed in an ordinary beaker or test-tube and the measurement of time is started at the same instant.The mixture is then poured into the cell of a photometer and its optical density measured. The main disadvantage of the method is that it is not easy to keep the temperature of the mixture constant which is an essential requirement of any catalytic method. The danger of heating the solution is especially high if it has to be irradiated by the light beam of the instrument without interruption (if for example the optical density is recorded continuously during this period). Spectro-photometers with thermostated cells can be used with advantage. The potentiometric method is based on the measurement of a reversible electrode potential in the solution. As defined by the Nernst equation the electrode potential depends on the concentration of a particular species.Its application is somewhat restricted as there are only a few really reversible electrodes available that respond instantaneously to changes in concentration. The hydrogen ion Concentration can be monitored easily with the glass electrode. As there are many slow oxidation - reduction reactions in which hydrogen ions are involved the potentiometric method might be the obvious choice especially if there is no visible change of colour occurring at the same time. The weakness of the potentiometric method lies in the high error built into such measurements because of the logarith CATALYTIC METHODS IN ANALYTICAL CHEMISTRY 243 mic correlation between electrode potential and concentration.It is easy to show with a simple error calculation that if the electrode potential is measured with a precision of &l mV (or the pH monitored with a precision of &0*02) the relative error committed in the measurement of the concentration is about 4 per cent., which is quite high even for the purpose of catalytic analysis. The gas volumetric method is based on the measurement of volumes of gases evolved in the course of a reaction. This technique is frequently used in connection with processes involving the catalytic decomposition of hydrogen peroxide where the product is oxygen gas. This method is somewhat clumsy and is sensitive to changes in environmental temperature and pressure and of course is restricted to reactions where there is a gas among the products of the reaction.Besides these main types there are quite a few other methods suggested for the monitoring of the concentration of reagents or products. Fluorimetric monitoring is possible if one of the reagents or products emits fluorescence radiation when irradiated with an ultraviolet light. If saccharides are involved in the reaction, polarimetric monitoring can be used. The occurrence or disappearance of chemi-luminescence can be used if chemiluminescent materials are involved in the reac-tion. As these are good indicators of hydrogen peroxide in alkaline media they can also be used if hydrogen peroxide is one of the reactants. These methods however, are limited to certain substances and are applied only occasionally. Determination of metals The metals are listed in alphabetical order and the information on each method is presented in a condensed way including whenever available the classification of the technique according to the principles described above the reaction and the catalytic effect on which the method is based the concentration range within which the determination can be carried out information on the calibration curve and data on accuracy and precision and finally interferences and their elimination.When the original paper was not available the information is from the abstract. Aluminium. Continuing their work on the examination of the catalytic decarboxylation of oxaloacetic acid with various substituted pyridines Michaylova and Bontchev4 worked out a method for the determination of aluminium.A fixed reaction time method was used and the pressure of the carbon dioxide gas collected into a vessel is measured 35 minutes after the initiation of the reaction. The minimum amount of aluminium that can be determined is 06pg and 25-pg amounts of copper iron or zinc 250 pg of lead magnesium and cadmium and 4000pg of calcium can be tolerated. The calibration curve is constructed by plotting the pressure of carbon dioxide versus concentration of aluminium. Though not based strictly on catalytic action the kinetic method of Bognar and Pataky-Szabo5 for the determination of 0.1 to 1 pg of aluminium in 5 ml of solution has to be mentioned here. The slow reaction between aluminium and Pontachrome-Violet SW or Pontachrome-blue black R at pH 5 is monitored by examining the fluorescence of the product by a simultaneous comparison method.The error of the determination is reported to be below 10 per cent 244 SVEHLA Barium. A kinetic although not strictly catalytic method for the deter-mination of barium calcium magnesium and strontium was described by Pausch and Margerum.6 The ligand-exchange reaction between lead(I1) and the alkaline earth complexes of t~ans-l,2-diaminocyclohexane-N,N,N',N'-tetraacetate (CyDTA) was used. A photometric measurement of the optical density of the lead complex was coupled with an oscilloscopic observation the oscillograms were photographed and used for evaluation. The method is suitable for the simultaneous determination of the four alkaline earth ions. Beryllium. Two methods have been described recently for the determination of beryllium both based on the inhibition of enzyme-catalysed reactions.Towns-hend and Vaughan7 recommend the use of calf intestinal alkaline phosphatase, which catalyses the hydrolysis of phosphate esters. This hydrolysis is inhibited by 20 to 90-ng amounts of beryllium in 7 ml of solution. A fixed time method was used; the reaction is stopped by the addition of sodium hydroxide solution after 3 minutes followed by a spectrophotometric measurement. Interference from other metal ions is prevented by adding sodium diethyldithiocarbaminate to the mixture. Another method similar in principle was developed by Guilbault Sadar and Zimmer.8 Both acid phosphatase (working at pH 7) and alkaline phosphatase (pH 8) can be applied.The substrate used is umbelliferone phosphate which is cleaved by the enzyme to produce the highly fluorescent umbelligerone. The reaction rate is followed by measuring and recording the fluorescence activity as a function of time. In a blank experiment the blank background fluorescence is recorded. The percentage inhibition is calculated from the results and is plotted against the concentration of beryllium to obtain a calibration graph. Between 0.01 and 0.1 pg ml-1 of beryllium can be determined with an error of between +3.5 and -2-5 per cent. Larger amounts of fluoride chloride bromide iodide, phosphate sulphate dichromate aluminium cadmium copper lead lithium, magnesium manganese mercury nickel potassium silver and sodium can be tolerated.Bismuth interferes and can be determined in a similar way. Bismuth. The method of Guilbault Sadar and Zimmer,8 described for the determination of beryllium can also be used for the determination of 1 to 70 pg ml-1 of bismuth with an error of between -5 and +1 per cent. An initial rate method was described by Hargi~.~ The reaction is a combined ligand-exchange and redox process. Phosphate bismuth and molybdate ions react in a slow equilibrium process to form a product that is then reduced by ascorbic acid to form the blue 18-molybdo-bismuthophosphoric acid. The formation of the latter can be monitored by spectrometry and the initial rate expressed in change of absorbancy per minute is found to be proportional to the concentration of bismuth. For the determination of 1 to 5 pg ml-l of bismuth the error is between -1.2 and +0.9 per cent.while at the lowest concentration level at which deter-minations can still be made 0.04 pg ml-l the error is below 310 per cent. A great number of metal ions and anions do not interfere and these are tabulated in the original paper CATALYTIC METHODS IN ANALYTICAL CHEMISTRY 246 Calcium. The kinetic method of Pausch and Margerum6 for the deter-mination of alkaline earth metals has already been mentioned in connection with barium. The method is suitable for the determination of calcium alone or in binary or ternary mixtures. Townshend and VaughanlO applied an interesting enzyme-catalysed method for the determination of 1 to 4 pg of calcium in 7 ml of solution. The apo-enzyme of calf-intestinal alkaline phosphatase can be activated in the presence of zinc by small amounts of calcium.p-Nitrophenyl phosphate is used as a substrate which is hydrolysed by the enzyme. A fixed reaction time method is applied the hydrolysis being quenched after 4 minutes by the addition of sodium hydroxide solution. The absorbance of the solution is measured at 410 nm and is plotted on the calibration curve against the concentration of calcium. The only disadvantage of the method is the fact that although the preparation of the apo-enzyme is simple it is some-what time-consuming and the product is not too stable. Chromium. Jasinskiene and Bilidiene recently described two methods for the determination of chromium. The first of thesell is based on the oxidation of indigocarmine with hydrogen peroxide in acid medium.The oxidation is catalysed by chromium(V1) ions. If 2,2’-bipyridil is added to the mixture as little as 6 ng ml-1 of chromium can be determined; in its absence the sensitivity is some-what lower. The extinction of the mixture is measured at 597 nrn and plotted as a function of time; the reaction rate is then determined from these plots. The calibration curve is constructed by plotting the reaction rate against the concen-tration of chromium. Copper interferes with the determination. Their second method12 utilises the catalytic action of chromium(V1) on the rate of the oxidation of methyl orange by hydrogen peroxide. Citric acid acts both as an activator and as a complexing agent. At pH 2.8 (obtained simply by the addition of hydro-chloric acid) between 14 alid 200 ng ml-l of chromium can be determined by a spectrophotometric monitoring of the reaction rate.The extinction at 490 nm is measured and plotted as a function of time; from these graphs the reaction rates can be determined. Calibration graphs are obtained by plotting the rate as a function of the concentration of chromium. The coefficient of variation of the method is below jC9 per cent. and about twenty common metals do not interfere. Other activators may be used instead of citric acid. Cobalt. Cobalt(I1) ions have a catalytic effect on the oxidation of various organic substances by hydrogen peroxide and all of the recent methods suggested for the determination of cobalt are based on this principle. Kucharkowski and DOgel3 applied the reaction between hydrogen peroxide and tiron at pH 10.3 to the determination of cobalt(I1).The reaction product has a slightly yellow colour and its concentration can be monitored by measuring the extinction of the mixture at 336 nm. The initial rate method can then be applied although more recently Kucharkowski14 has modified the technique by applying the fixed reaction time method. Between 0.6 and 10 ng ml-1 of cobalt can be determined although with a fixed reaction time of 24 hours the sensitivity can be extended to 0.03 ng ml-l 246 SVEHLA The method has successfully been applied to the determination of trace amounts of cobalt in molybdenum metal. Popa and Costache16 used 2,6,7-trihydroxy-9-phenylxanthen-3-one in ethanolic solution as a reagent to be oxidised by hydrogen peroxide and by measuring the concentration of the reagent by spectrophotometry at 460nm the initial rate method can be applied.The reaction takes place in alkaline medium whose pH is maintained by a sodium hydroxide - sodium borate buffer. The calibration graph obtained by plotting the initial rate against the concentration of cobalt is linear within the range of 0.08 to 0.64 ng ml-l and the error is less than &5 per cent. More recently Popa and Costache16 replaced their reagent by the more readily available Bordeaux S (C.I. Acid Red 27). Spectrophotometric monitoring is done at 496 nm. Between 3 and 16ng ml-l of cobalt can be determined with errors below 20 per cent. Kreingold and Bo~hevolnov~~ determined cobalt in germanium tetrachloride and trichlorosilane applying the reaction between hydrogen peroxide and salicylfluorone in an alkaline medium.The sensitivity of the method is 0-05 ng per 5 ml. A detailed study of interferences has also been made. The special merit of their paper is the fact that they recommend four different catalytic methods for the determination of traces of four metals in germanium tetrachloride and tri-chlorosilane samples. Copper. The catalytic action of copper on various slow redox reactions is well known and is the reason why so many methods have recently been proposed for its catalytic determination. Pall Svehla and ErdeylS determined copper by using a process based on the Landolt reaction between peroxidisulphate and iodide ions in acidic medium with sodium thiosulphate as delaying agent.In this visual, fixed concentration technique reaction times are measured and the reciprocal values are plotted against the concentration of the catalyst. For 1 to 100 pg ml-l of copper the calibration graph is linear and the error is below 10 per cent. The interference of iron(II1) can be eliminated by adding sodium fluoride to the mixture. The theoretical backgrounds have been described.19 In a series of papers Hessel-barth20-22 described the determination of copper based on its catalytic action on the reaction between iron(II1) ions and thiosulphate. Thiocyanate ions are used to produce the iron(II1) thiocyanate complex the colour of which can be examined visually when the method of simultaneous comparison is adapted. Over a certain level of acidity the results are independent of the concentration of hydrochloric acid in the solution.Between 0.1 and 100 pg ml-l of copper can be determined by the method. Although many ions interfere with the determination copper can easily be separated by cation exchanger22 or co-precipitated with mercury sulphide21 before analysis. Dittelzs based his method on the catalytic effect of copper on the reaction between hydrogen peroxide and quinol in the presence of pyridine at pH 7 which is maintained by a phosphate - tartrate buffer. The rate of the reaction is monitored by measuring the absorbance of the mixture at 470nm. Unknown samples are analysed with the aid of a calibration graph in which the rate is plotted against the concentration of copper.Up to 10 ng of copper can be determined with an error o CATALYTIC METHODS IN ANALYTICAL CHEMISTRY 247 less than &5 per cent. the limit of detection being 4-2 ng under normal circum-stances although in the absence of tartrate ions this can be reduced to 049ng. Colovos and Papadopoulos24 have described an automatic method very similar to the above in which 2,4-diaminophenol was used as a reagent with hydrogen peroxide. The extinction at 500 nm is measured and reaction times are determined from the recorded charts on the basis of the fixed concentration method. Reci-procal reaction times increase linearly with copper concentration within the range of 0.2 to 12.0 pg ml-1. Tin molybdenum and iron interfere seriously with the determination. N e d ~ e d ~ ~ applied the reaction between hydrogen peroxide and the dipotassium salt of Congo Red (C.I.Direct Red 28) which is catalysed by copper. The concentration of the dye is monitored by spectrophotometric measurements at 501 nm. At pH 12 between 1 and 20 ng ml-l of copper can be determined in this way. Cerium cobalt silver vanadium thiocyanate hexacyanoferrate(II1) and EDTA interfere seriously. The interference by manganese can be eliminated by the addition of oxalate ion to the mixture. The method was adapted for the determination of copper in samples of technical potassium hydroxide lithium chloride caesium carbonate and rubidium sulphate. Heller and Guyon26 based their very sensitive method on the catalytic action of copper on the reduction of iso-propylmolybdate by ascorbic acid to molybdenum blue.The fixed time method is applied the extinction of the solution being measured at 750 nm after 60 minutes. At pH 1.85 between 50 and 300 p.p.b. of copper can be determined. The inter-ference of a number of ions can be overcome by a preliminary solvent extraction of copper. The inhibiting action of copper of an enzyme-catalysed reaction was applied by Guilbault Kramer and Hackley27 for the determination of 0.2 to 6 pg ml-l of copper. The hydrolysis of indol-3-yl-acetate by hyaluronidase can be followed by the fluorimetric measurement of the product of the hydrolysis. By recording the variation of the fluorescent intensity with time the initial rate method can be applied. A monochromatic ultraviolet radiation (395 nm) is used for excitation while the fluorescence is measured at 470 nm.The hydrolysis proceeds slower in a phosphate - citrate buffer of pH 6.4. The coefficient of variation of the determination is k2.3 per cent. For the determination of traces of copper in zinc metal Gyganok Cujko and Reznik28 applied a preliminary separation by co-precipitation of the copper wth a 0.02 per cent. amount of the zinc present in the sample by using hydrogen sulphide. After filtration washing and dissolution the catalytic determination of copper is carried out by applying the reaction between phosphomolybdate ions and thiourea in sulphuric acid medium. The extinction of the mixture is measured at 680nm at regular intervals the initial rate method being applied for evaluation. Between 7 and 33 x per cent.of copper in zinc metal or zinc sulphate can be determined with an error of less than 5 per cent. Kreingold and Bozhevolnov17 determined amounts as low as 5ng of copper in 5 ml of solution on the basis of its catalytic action on the dirnerization of lumocup-ferron [a-(4-dimethylaminobenzylidene) hippuric acid]. The method was applied to the analysis of copper in samples of high purity germanium tetrachloride and trichlorosilane 248 SVEHLA Germanium. The slow reaction between molybdate and iodide ions in acid medium which is catalysed by traces of germanium was applied for its deter-mination by Michalski and Ge10wa.~~ The concentration of iodine formed in the reaction can be measured by amperometry or biamperometry. The initial rate method can be applied for evaluation.The sensitivity under normal circumstances is 50 ng ml-l but with special precautions even 8 ng ml-l of germanium can be determined. The average error of the determination is 5 per cent. and ammonium, barium calcium potassium sodium and zinc ions (up to 0 . 0 5 ~ concentration) do not interfere. Iron. A number of new methods have been suggested for the determination of iron. Thompson and Svehla3* applied the perborate - iodide Landolt reaction with ascorbic acid as a delaying agent. At pH 4 maintained by an acetate buffer 1 to 10 pg ml-l of iron can be determined by measuring the reaction time that elapses from mixing the reagents to the occurrence of the Landolt effect. Reciprocal reaction times gave a linear calibration curve when plotted as a function of the concentration of iron.The deviations at the 95 per cent. level of significance were also calculated and plotted together with the calibration curve. Molybdenum and osmium interfere with the determination. The interference of molybdenum can be eliminated by masking it with tartrate ions. The theoretical backgrounds of the method were de~cribed.~~ Another Landolt reaction18 that is catalysed by iron ions is the slow reaction between peroxidisulphate and iodide ions in acid medium. Sodium thiosulphate is used as a delaying agent. The reaction time is measured between mixing the reagents and the liberation of iodine which occurs instan-taneously. Between 1 and 100 pg m1-1 of iron can be determined with an error of less than 10 per cent. The calibration curve made up by plotting reciprocal reac-tion times as a function of concentration of iron is linear.Copper interferes with the determination. The theoretical backgrounds of the method are discussed in a thesis.19 A fixed reaction time method was recommended by Orav and K ~ k k ~ ~ for the determination of as little as 0.1 pg ml-l or iron. The slow reaction between triethylene tetramine and hydrogen peroxide is used. The optimum pH is 10 which is maintained by a borate buffer. At the fixed reaction time of 10 minutes the catalytic reaction is stopped by acidifying with sulphuric acid and the amount of unreacted hydrogen peroxide is determined by iodimetric titration. In a calibra-tion graph the amount of hydrogen peroxide decomposed during the process is plotted as a function of catalyst concentration.The error of the method is high, but was kept below 28 per cent. in the cases examined by the authors. Kreingold and BozhevolnoP based their method on the catalytic effect of iron on the oxida-tion of Acid Chrome Dark Blue dye by hydrogen peroxide in acid medium. The reaction rate is followed photometrically; the initial rate when plotted against the concentration of iron gives a linear calibration curve. Amounts as low as 2 ng ml-I of iron can be determined. Antimony gallium molybdenum niobium, tin titanium and tungsten interfere with the determination. The method was applied for the determination of iron in lanthanum oxide when the coefficient o CATALYTIC METHODS IN ANALYTICAL CHEMISTRY 249 variation was 18 per cent.Later Kreingold Bozhevolnov and Antonov3* recom-mended the oxidation of H-acid [4-amino-5-hydroxynaphthalene-2,7-disulphonic acid] by hydrogen peroxide in the presence of hydrochloric and acetic acids cata-lysed by traces of iron for the determination of the latter. The fixed times method was used the extinction of the solution being measured at 508 nm after 15 minutes. The calibration graph is constructed by plotting the difference of extinction between the test solution and a blank as a function of the concentration of iron. The method is sensitive for up to 5 ng of iron in the final solution. Alkali metals alkaline earth metals aluminium cobalt copper lanthanum manganese nickel and zirconium do not interfere. A third method of Kreingold and Bozhev~lnov~~ applies the reaction between Stilbexon [4,4'-bis(carboxymethyl)amino-stilbene-2,2'-disulphonic acid] and hydrogen peroxide in acid medium.The method was successfully applied for the determination of iron traces in high purity germanium tetrachloride and tri-chlorosilane. The limit of detection is 2.5 ng of iron in 5 ml of solution. Kharlamov, Dodin and Mantsevich35 based their method on the photo-oxidation of methyl orange by atmospheric oxygen in the presence of traces of iron. A fixed time method is applied with photometric measurement. The reagents are mixed at pH 2 (which is obtained by the addition of suitable amounts of oxalic sulphuric or perchloric acids) exposed for 10 minutes to ultraviolet radiation and before and after irradiation the extinction of the mixture is measured at 350 nm.The differ-ence of extinction is plotted as a function of catalyst concentration. 0.025 pg ml-1 of iron can be determined with an error not exceeding 3.4 per cent. Copper interferes. The method has successfully been applied for the determination of iron in nickel metal. Guilbault Kramer and Hackley2' applied the same enzyme-catalysed reaction for the determination of iron as the one described for copper. Between 0.2 and 12 pg ml-l of iron can be determined on the basis of its inhibiting action on the enzymatic hydrolysis of indol-3-yl-acetate. The coeficient of varia-tion of these determinations is better than 52.3 per cent. Though not based strictly on catalysis the photochemical kinetic method of Lukasiewicz and Fitz-gerald36 has also to be mentioned.They have determined 0.1 to 1 pg ml-1 amounts of iron by a spectrophotometric monitoring of the rate of the reduction of iron(II1) ions by oxalate in the presence of 1 ,lo-phenanthroline which not only complexes iron(I1) ions and so speeds up the reduction but is also an excellent spectro-photometric reagent for the monitoring of the process. The procedure requires irradiation by ultraviolet light or the reduction does not take place. The initial rate method is applied; the rate is directly proportional to the concentration of iron. Magnesium. The kinetic method of Pausch and Margerum,6 described in connection with barium and calcium can be applied for the determination of magnesium alone and in binary and ternary mixtures with other alkaline earth metals .Manganese. The catalytic method for the determination of manganese, based on the oxidation of &ethyl aniline with potassium periodate has been know 250 SVEHLA for some time.37 Hadjiioannou and Kephalas3* automated the method by applying the instruments described by Malmstadt and Hadjiioannou3Q earlier. The method is based on the principle of fixed concentrations but both the measurements of extinction and time are made automatically and the results are recorded and printed out. Between 3 and 40-ng amounts of manganese can be determined with an error of less than 2 per cent. and a coefficient of variation of &l per cent. The method was successfully applied for the determination of manganese in natural waters. Janjic Milanovic and Celap40 used the oxidation of Alizarin S with hydro-gen peroxide in sulphuric acid medium.The reaction is catalysed by 2 to 10 ng ml-l of manganese. With spectrophotometric measurement of the extinction at 335 nm, both the fixed time method and the initial rate method can be applied though a procedure applying the method of simultaneous comparison is also described. The latter however is applicable only for higher amounts (0-1 to 1 pg ml-l of manganese. The influence of foreign ions has been investigated thoroughly and the kinetics of the reactions involved were also examined. Ditte123 recommended the use of the oxidation of diethylaniline by periodate ions. This reaction is catalysed by as little as 10 ng of manganese. The solution contains a phosphate -tartrate buffer of pH 6 that also acts as a complexing agent to eliminate the effect of some metals.The progress of the reaction is monitored by measuring its extinction at 471 nm. The initial rate method can be applied for evaluation. At the 10-ng level the coefficient of variation is 7.1 per cent. Kreingold and Bozhevol-determined 0-3ng of manganese in 5ml of solution on the basis of its catalytic effect on the reaction between hydrogen peroxide and lumomagneson [5-(5-chloro-2-hydroxy-3-sulphophenylazo)-barbituric acid]. Antimony bismuth, cobalt lanthanum lead magnesium nickel platinum silver thorium zinc and zirconium as well as citrate EDTA and phosphate interfere with the determination. Mercury. Mealor and Townshend41 used an enzyme-catalysed reaction, inhibited by traces of mercury for the determination of the latter within the to molar concentration range.The hydrolysis of sucrose by invertase proceeds with decreasing rate with increasing concentration of mercury. The fixed time method can be applied and the concentration of sucrose after a period of 60 minutes can be measured by polarimetry. A suitable buffer (of pH 4.0 or 5.5) has to be applied the pH being chosen according to other metals present. If silver is present a pH of 4 is the optimum level of acidity for there to be no inhibition by silver. In the absence of silver and some other interfering ions such as lead and zinc a pH of 5.5 should be used. The effects of various metal ions were examined. There is virtually no interference from cadmium copper lead silver uranium and zinc if present in 100 to 1000 fold amounts.More recently Townshend and V a ~ g h a n ~ ~ described a method based on a particularly interesting principle. The dehydrogenation of ethanol by nicotinamide adenine dinucleotide is catalysed by yeast alcohol dehydrogenase to yield acetaldehyde and the hydrogenated form of the nicotinamide adenine dinucleotide. The reaction leads to an equilibrium. The inverse process is also catalysed by the same enzyme. Metal ions such as mercury and silver inhibit both processes. The degrees of inhibition in both directions ar CATALYTIC METHODS IN ANALYTICAL CHEMISTRY 251 much the same if mercury is present but there is an appreciable change however, in the behaviour of silver. On this basis a certain degree of selectivity can be obtained.The forward process is simpler and more practical and this is recom-mended as a basis for the determination of 2 to 20 ng of mercury. A fixed time method can be applied and the fluorescence of the hydrogenated nicotinamide adenine dinucleotide is measured both in a blank and in the test solution. The relative activities calculated from these results can be plotted as function of the catalyst concentration. Although the results obtained were reproducible in un-buffered solutions it is reported to be more advisable to work at constant pH values (8 for the forward and 5.6 for the reverse process). The main source of error, working at these concentration levels is the adsorption of mercury on to the walls of glassware. Soaking with a solution of dithizone in carbon tetrachloride removes the adsorbed mercury ions from the glass surface.Rodriguez and Pardue4' determined mercury by its inhibiting effect on the reaction between cerium(1V) and arsenic(III) catalysed by iodide and osmium. Details of the method will be described in connection with the determination of silver. Molybdenum. The catalytic action of molybdenum on redox reactions involving peroxy compounds has been known for some time. Thompson and Sveh1a3O based their method on the perborate - iodide Landolt reaction with ascorbic acid as a delaying agent. At pH 4 maintained by an acetate buffer 1 to 10 pg ml-l of molybdenum can be determined. In this fixed concentration method, reaction times between mixing the reagents and the appearance of iodine are measured.Reciprocal reaction times plotted as a function of the concentration of molybdenum give linear calibration graphs. The deviations around the calibration curve with a 95 per cent. level of significance were also calculated. Iron and osmium interfere with the determination. The interference of iron can be eliminated by adding EDTA to the solution but the effect of osmium remains unchanged. The theoretical backgrounds of the method have been discussed.31 The catalytic effect of molybdenum on the hydrogen peroxide - iodide reaction was utilised by Weisz Klockow and Ludwig.43 The novelty of their method lies in the interesting principle applied for the measurements. The authors used a potentiostat i e . an automatic titrimeter which can be used to maintain a pre-determined electrode potential in the cell by adding a reagent from an automatic burette.The volume of the reagent added is recorded as a function of time. If iodine and iodide are both present from the start the oxidation - reduction potential of the iodine - iodide couple can be measured in the cell. If some of the iodide is oxidised to iodine the potential will become more positive. To maintain the potential during the course of the reaction iodide has to be added. With a potentiostat this can be solved easily. Provided that both the iodide and iodine concentrations were large enough at the beginning the constancy of the electrode potential means that the concen-tration of iodide is kept constant during the measurements. By this the order of the reaction is reduced to zero (for a more detailed explanation the original paper should be consulted) and the reaction rate becomes constant.The latter will b 252 SVEHLA proportional to the rate of addition of iodine which is equal to the slope of the recorded curve. Thus a calibration curve can be drawn and samples containing 1 to 10 pg ml-l of molybdenum can be analysed with an error of less than -4 per cent. and those with 0 to 1 pg ml-l of molybdenum with an error of below -6 per cent. Lazarev44 applied a turbidimetric method for the kinetic determination of molybdenum on steel. The reaction applied is that which occurs between selenate and tin(I1) ions leading to the formation of selenium and which is catalysed by molybdenum. After the dissolution of the sample the solution is acidified with hydrochloric acid and gum acacia solution is added to protect the suspension of selenium.The reaction is started and after a fixed time of 30 minutes the extinction of the solution (caused by the turbidity in the mixture) is measured against a blank at 390 nm. In the final solution 1 mg of molybdenum can be determined with an error of -7 per cent. Oxidising agents and rhenium interfere. Nickel. Mealor and Town~hend*~ described the determination of 0.1 to 0.7 p.p.m. of nickel based on the decomposition of permanganate in an alkaline solu-tion that also contains 1-hydroxyethylidene-1,l-diphosphoric acid. A fixed concentration method with spectrophotometric measurement is applied and the time required for the extinction (at 600 nm) of the mixture to reach 0.300 against a water blank is measured.The calibration graphs are linear if the reciprocal reaction times are plotted as a function of the squares of concentrations of nickel. The measurement can be automated. Interference by cobalt copper iron and silver can be eliminated by removing these ions by solvent extraction. Phosphate and oxalate decrease the rate of reaction. Osmium. Osmium is a well known catalyst of redox reactions. Gregorowicz and S ~ w i n s k a ~ ~ applied its effect on the reaction between persulphate ions and 3-amino-4-hydroxy-benzenesulphonic acid for its catalytic determination. The concentration of the organic reagent can be monitored during the process by measuring the extinction of the mixture at 530 nm and results are evaluated on the principles of the initial rate method.Between 7 and 35 pg of osmium per 50 ml can be determined with an average error of 0.7 pg. Working at a pH of 2.2 and 20°C, cadmium mercury molybdenum silver tungsten vanadium zinc and zirconium do not interfere copper has a slight effect and iron oxidants and reductants have a marked interference on the rate of the reaction. The possibility of the deter-mination of osmium based on the reaction between cerium(1V) and arsenic(II1) ions was pointed out recently by Rodriguez and Pardue!' while Thompson and Svehla30 and Thompson3l reported its effect on the Landolt reaction between perborate and iodide ions with ascorbic acid as a delaying agent. None of these authors gives however detailed procedures for the determination of osmium.Ruthenium. Ottaway Fuller and Allan48 recommended the use of the slow oxidation of the tris(lJ0-phenanthroline) - iron(I1) complex by periodate cata-lysed by ruthenium for the determination of the latter. The extinction of th CATALYTIC METHODS IN ANALYTICAL CHEMISTRY 253 iron(I1) complex can be measured at 505 nm. After a fixed time of 10 minutes at 25°C the extinction reading is taken. Its reciprocal value when plotted against the concentration of the catalyst gives a linear calibration curve within the region of to 2 x 1 0 - 8 ~ ruthenium. At higher concentrations the curve bends towards the concentration axis. The detection limit of ruthenium was found to be 10-lo M. A number of ions were investigated for interferences but only iridium, osmium and rhenium showed serious effects on the accuracy of ruthenium deter-minations.Silver. The catalytic action of silver on various redox reactions was utilised by a number of authors for its determination. Bontchev and Alexiev4O applied the oxidation of hydrochloric acid by cerium(1V) ions. This reaction is catalysed by as little as 0.06 pg ml-l of silver. The pH of the solution must be just over 2 nitric acid being used to keep the acidity at this level. A fixed reaction time of 20 minutes is allowed at 30"C and the extinction of the solution which is proportional to the concentration of cerium(1V) is measured at 420 nm. From these results a calibra-tion graph can be constructed. The error of the method is about 13 per cent. at 95 per cent. confidence limits.The method can be used for the determination of silver in high purity zinc and cadmium. Copper and manganese interfere with the determination. Alexiev and Bontchevso later examined the possibility of deter-mining silver on the basis of its catalytic effect on the reaction between perosi-disulphate ions and sulphanilic acid at pH 4.5. The product of the reaction has a yellow colour and its extinction can be measured at 420 nm even with a filter instrument. When using the initial rate method for evaluation 0 to 10 x M of silver can be determined with a linear calibration curve. Eontchev Alexiev and Dimitrovasl later modified the previous procedure by adding 2,2'-bipyridyl , which acts as an activator to the mixture. The sensitivity of the determination was increased by this modification to 4 x pg ml-l of silver.The coefficient of variation of the method is jz7.6 per cent. The inhibiting action of silver on some enzyme catalysed reactions was used for its determination by Mealor and Towns-hend4I and Townshend and Vaughan.42 The principles are identical to those described in connection with the determination of mercury. The sensitivities for silver are 0.2 pmole and 1 ng ml-1 respectively which is twice as good as those quoted for mercury. Rodriguez and Pardue4' utilised the inhibiting action of silver on the reaction between cerium(1V) and arsenic(II1) in the presence of iodide as catalyst. The inhibition is caused by the formation of silver iodide. The extinction of the solution caused by the presence of cerium(IV) can be measured at 407 nm and the results can be evaluated on the basis of the initial rate method.There are modifications of the method applying iodide alone or mixtures of iodide and os-mium as well as iodide and iodate as catalyst for determinations of silver in the con-centration range of 0 to 25 x 1 0 - 8 ~ . Menghani and Bakore52 examined the catalytic effect of silver on the reaction between peroxidisulphate ions and pinacol. Although their main task was to clarify the kinetics and mechanism of the reactions involved their findings can also be utilised for the catalytic determination of silver 254 SVEHLA Strontium. The method of Pausch and MargerumJ6 discussed in more detail in connection with barium can also be used for the determination of strontium alone or in binary or ternary mixtures with other alkaline earth metals.Thallium. The reaction between cerium(1V) ions and thallium(1) in acid medium is slow in the dark but the rate of reduction of cerium(IV) which can be followed by spectrophotometric measurements at 425 nm increases considerably in daylight. The phenomenon is interpreted by Schenk and B a ~ z e l l e ~ ~ as the catalysis of thallium on the photo-reduction of cerium(1V). Although their main concern was to accelerate reactions for the cerimetric titration of thallium(1) ions, they drew attention to the possibility of analysing mixtures of thallium(1) and other metals kinetically. The strong catalytic action of manganese was also noted. Tungsten. Hadjiioannou and Valkana54 applied the catalysis of the reaction between hydrogen peroxide and iodide for the determination of tungsten.The formation of iodine in acid medium can be monitored by photometric measure-ments. A fixed concentration method was applied and reaction times lying between 20 and 100 s were measured. Between 0.5 and 3 pg of tungsten can be determined with a coefficient of variation of between 1 and 2 per cent. Interferences of other ions were studied and the maximum amounts of these causing errors of less than 5 per cent. were determined and tabulated. Vanadium. The catalytic action of vanadium on redox reactions between halates and halides is well known. Thompson and S ~ e h l a ~ ~ utilised the Landolt reaction which is based upon the bromate - iodide reaction in acid solution by using ascorbic acid as a delaying agent.By applying a citric acid - sodium citrate buffer (pH 2.2) copper iron molybdenum and titanium which would otherwise interfere are also complexed and the determination of 0 to 10 pg ml-f of vanadium becomes specific only for this ion. The technique involves a fixed concentration method with measurement of the reaction time from mixing the reagents up to the appearance of the colour of iodine. Reciprocal reaction times result in a linear calibration curve when they are plotted against the concentration of vanadium. Another method based on a Landolt reaction was developed by Bognar and J e l l i ~ ~ e k . ~ ~ They apply the reaction between chlorate and chloride with tin(I1) chloride as a delaying agent o-Tolidine is used as an indicator for chlorine.As the reaction is slow the temperature should be kept at 40°C. For 1 to 1Opg of vanadium the coefficient of variation is below 5 per cent. Later Bognar and Sarosij' replaced tin(I1) chloride by hydrazine sulphate as a delaying agent. The temperature in this case has to be kept at 65°C. More recently Bognar and Jellinek,58 modifying their earlier method based on the bromate - bromide -ascorbic acid Landolt reaction recommended the use of dimethyl or tetraethyl-9-phenylene diamine as an indicator for the appearance of bromide. A highly sensitive method has recently been recommended by Fuller and O t t a ~ a y . ~ ~ The method utilises the oxidation of Bordeaux with bromate in acid medium. The rate of the reaction is rather sensitive to variations in acidity.A fixed concentratio CATALYTIC METHODS IN ANALYTICAL CHEMISTRY 255 method was applied combined with the spectrophotometric monitoring of the concentration of Bordeaux at 515nm the time elapsing from the mixing of the reagents until the transmission had increased to 70 per cent. was measured. Between 0-005 and 0.2 pg ml-l amounts of vanadium can be determined with a standard deviation of 3 per cent. (at the 0.05 pg m1-1 level). The effects of thirty-seven cations anions and complexing agents were examined. Thiocyanate nitrite, iodide ruthenium and copper have a marked interference. The calibration curve, which is constructed by plotting reciprocal reaction times as a function of the concentration of the catalyst bends towards the concentration axis. Tanaka and Awata6O applied the reaction between 4-hydrazinobenzenesulphonic acid and chlorate ions at pH 2.5 to the determination of vanadium.After a fixed time the product of the reaction 4-diazobenzenesulphonic acid is coupled with l-naphthyl-amine and the concentration of the latter is determined spectrophotometrically at 530 nm. From 0.02 to 1 p.p.m. of vanadium can be determined. A preliminary extraction with 8-hydroxyquinoline at a pH of 4-45 into chloroform and re-extrac-tion at pH 10 into an ammonia - ammonium chloride buffer increases the selec-tivity. Iron can be masked with 1,2-diamino-cyclohexane-N,N,N',N'-tetraacetic acid but antimony bismuth silver and tin interfere seriously. Zinc. Townshend and Vaughan described two methods for the determination of zinc both based on enzyme-catalysed reactions.The first of these' is based on the same principle as described in connection with beryllium and with it 0.6 to 6 pg of zinc can be determined. The second methodlo is identical to that described for calcium; procedures for 6 to 65 and 65 to 650-ng amounts of zinc are given. Both methods are accurate and highly selective. Determination of non-metallic substances and anions Bromide. For the determination of as little as 3ng ml-l of bromide, Toropova and Tamarchenko61 recommended the use of potassium bromate in acid medium. On adding methyl orange to the solution its colour fades slowly because of the formation of bromine. The initial rate of decolourisation which can be determined spectrophotometrically at 490 nm is proportional to the amount of bromide present.Arsenite thiosulphate sulphide thiocyanate nitrite and iron interfere. Cyanide. The method of Guilbault Kraxner and Hackley,27 described in connection with copper can also be used for the determination of 0-1 to 4 pg ml-l amounts of cyanide with a coefficient of variation of 2-3 per cent. Hydrogen peroxide. A n interesting catalytic method for the determination of traces of hydrogen peroxide has been described by Krueger Warriner and Jaselkis.62 Xenon trioxide and t-butyl alcohol axe used as reagents in neutral solution. The reaction is preceded by an incubation period the length of which depends on the concentration of the catalyst. The ultraviolet absorption of xenon trioxide can be monitored at 200nm and from the inflexion of the extinction 256 SVEHLA vemw time curves which resemble a potentiometric titration curve the incubation period can be determined.Alternatively the half-life of xenon trioxide can be used for evaluation. From 36 to 360 parts per lo9 of hydrogen peroxide can be deter-mined with an error of below 10 per cent. Krause Zielinski and WeimannM as well as Krause Domka and Marciniecs4 described kinetic methods for the determination of traces of hydrogen peroxide based on the rate of decolourisation of certain dyes. These are not strictly speaking catalytic methods although the techniques used are related to those usually applied in catalytic determinations. Iodide. A number of new methods have been suggested for the catalytic microdetermination of iodide and a number of authors proposed modifications of older established techniques.Ballzos6 has utilised the iodide-catalysed oxidation of @,$'-tetramethyldiamino-diphenyl-methane with chloramine T. The formation of a reaction product can be monitored by spectrophotometry and the whole process can be automated. As little as 10-14-g amounts of iodide can be determined in 1 1 of drinking water. Toropova and TamarchenkoSg recommended the use of the reaction between arsenite and iodate which is catalysed by 5 ng ml-l or more of iodide. The rate of the reaction can be monitored by the amperometric measurement of the dif-fusion current of iodate. Proskuryakovas7 applied the oxidation of hexacyanofer-rate(I1) ions with nitrite to the determination of 0.02 to 0.1 pg ml-l of iodide while Bogurth and Schaeg68 made use of the oxidation of arsenite by manganese(II1) ions.A photometric monitoring is possible in both cases. Bognar and nag^^^ recom-mend a simultaneous comparison method for the determination of 0 to 1 pg ml-l amounts of iodide. The determination is based on the catalytic effect of iodide on the oxidation of 3,3 -dimethyl-naphthidine with hydrogen peroxide. A spectro-photometric monitoring is also possible. Ottaway Fuller and R o w ~ t o n ~ ~ re-examined the method originating from Iwasaki Utsumi and Ozawa.'l They pointed out the weaknesses of the procedure and suggested some important improvements. The well known method by Sandell and K ~ l t h o f f ~ ~ ~ ~ ~ based on the catalytic action of iodide on the reaction between cerium(1V) and arsenite was applied modified or automtaed by several auth~rs!~J*-~~ Oxygen.A method based on a combined redox and ligand-exchange reaction has been reported79 for the determination of dissolved oxygen in water. The method is strictly speaking not catalytic as oxygen is used up in the reaction but the rate of reaction can be monitored and the evaluation is made on a kinetic basis. Pentacyanocobaltate(I1) is oxidised first and the excess of the reagent is subjected to various ligand-exchange reactions. As little as 3 x lo-* moles per litre of dissolved oxygen can be determined. Phosphate. The method described by Crouch and Malmstadts0 is based on the fact that under certain conditions the initial rate of the formation of molyb-denum blue from phosphate molybdate and ascorbic acid is proportional to the concentration of phosphate.The formation of molybdenum blue can be monitore CATALYTIC METHODS IN ANALYTICAL CHEMISTRY 257 spectrophotometrically at 650 nm. The reaction is quite fast readings having to be taken as soon as 10s after the mixing of the reagents. The method can be automated and applied to the routine determination of phosphate in serum. More recently, Javier Crouch and Malmstadtsl made some improvements and innovations in the instrumentation and their new procedure is now based on an automatic reading carried out 10 ms after mixing the reagents. Selenium. West and Ramakrishnas2 recommended the use of the reduction of methylene blue with sodium sulphide at pH 10-8 for the catalytic determination of selenium.If sodium sulphite is present the interference by polysulphide can be eliminated. Formaldehyde and disodium ethylenediamine tetraacetate must also be added making the method specific for selenium. The time for complete decolourisation is measured for various concentrations of selenium and is used to prepare the calibration curve. Between 0-1 and 1 pg of selenium can be determined by the method. Bognar and Sarosim recommended a Landolt reaction method based on the chlorate - chloride - hydrazine sulphate system by which 0.1 to 1 pg per 5 ml of selenium can be determined with a coefficient of variation of 10 per cent. The simultaneous comparison technique can also be applied. Larger amounts of some metals can be tolerated. Kawashima and TanakaM have determined 0 to 0*15 pg amounts of selenium by the catalytic reduction of 1,4,6,11-tetra-azo-naphthacene with phosphorous acid in hydrochloric acid medium.The reaction has to be started at 50"C and after a fixed time of 30 minutes the mixture is cooled to 0°C to stop the reaction. The extinction of the solution must be measured at 600nm within 30 s. Interferences were carefully examined and listed. More recently another method based on the oxidative coupling reaction of phenyl hydrazine fi-sulphonic acid with 1-naphthylamine has been reported.% The mixture should be kept at 50°C for 60 minutes when the reaction can be stopped by cooling the system to O"C and the extinction of the solution is measured at 525 nm. The extinction is directly proportional to the amount of selenium present.Potassium chlorate has to be present although its role in the reaction mechanism has not been studied. Sulphide. Babko and Maksimenkos* published a method for the deter-mination of 0.1 to 10 ng of sulphide. Iron(I1) ions reduce silver nitrate to silver metal in the presence of sulphide ions only. Lighting must be controlled during the reaction to avoid the photo-reduction of silver ions. The extinction of the colloidal silver solution can be measured at 530 nm and the calibration graph is prepared by plotting extinction values as a function of sulphide concentration. For 5ng of sulphide the coefficient of variation is about 4 per cent. Up to 50 pg of chloride do not interfere. Oxidising or reducing substances that interfere with the reaction must not be present.Cataltyic and kinetic determination of organic substances The principle of determining concentrations by kinetic measurements has long been extended to the analysis of organic substances. In these procedures the role o 258 SVEHLA the organic substance can only seldom be described as catalytic. The compound often takes part in the main reaction and is used up during the process. In some cases the organic species acts as an inhibitor or complexing agent and in this way influences the catalytic behaviour of a metal or enzyme. As the experimental techniques applied for the determination of organic substances by kinetic analysis are the same as those described earlier for inorganic systems this review also deals shortly with the kinetic (but not catalytic) determination of organic substances.The substances are dealt with in alphabetical order regardless of their structure or composition. A reviews7 on kinetic methods in organic analysis was published in 1967 and references to earlier works are available therein. Arnines. Shresta and DasS8 described a kinetic method for the determination of primary amines based on their reaction with salicylaldehyde. A Schiff base is formed during the reaction. After a fixed time the reaction can be stopped by the addition of acetic anhydride when the unreacted primary amines are acetylated. The Schiff base can then be titrated by perchloric acid in a non-aqueous medium. A special graphical method is used for the evaluation. Methods for the deter-mination of secondary and tertiary amines are also described.Ascorbic acid. Bognar and Jellineksg have described a kinetic method for the determinaiton of 10 to 100 pg ml-l amounts of ascorbic acid based on its delaying action on the bromate - bromide redox reaction. A linear calibration curve can be prepared by plotting reaction times as measured in other Landolt reactions against ascorbic acid. With the simultaneous comparison method the range of determin-able ascorbic acid is 5 to 50 pg ml-l although the possibility of determining as large amounts as lo00 pg ml-I is also outlined. o-Chloronitrobenzene. Legradig0 used a kinetic method for the detection and determination of o-chloronitrobenzene in the presence of $-chloronitrobenzene. The o-compound forms a red product when heated with hydroxylamine in an alkaline solution in ethanol; the extinction of the coloured product can be measured by spectrophotometry.Chlorodinitrobenzene reacts even in cold solution and can be determined in the presence of the two other compounds. Complexing agents. Mottola Haro and Freisergl determined various organic complexing agents such as cysteine EDTA and salicylic acid These substances inhibit the catalytic action of copper(I1) ions on the atmospheric oxidation of L-ascorbic acid. The rate of decomposition of ascorbic acid can be monitored by extinction measurements at 265nm. As the rate of oxidation depends on the pH of the solution which itself varies in a solution of ascorbic acid the use of a buffer (pH 7-4) is recommended. From 1 to 1Opg ml-1 amounts of the com-plexing agents can be determined.Dihydric phenols. Schenk and BrowB2 recommend the use of a kinetic method for the determination of either catechol or quinol in the presence of resorcinol. The method is based on the reaction between the free radical 2,2 CATALYTIC METHODS IN ANALYTICAL CHEMISTRY 259 diphenyl-1-picrylhydrazil and the dihydric phenols. The reaction is started and the extinction of the free radical can be monitored at 540 nm. From the extinction versus time curves the initial rate can be determined and the concentration of phenols calculated. Ethanol is used as a solvent and daylight should be excluded. Esters. A dual temperature differential kinetic method was described by Munnellyg3 for the simultaneous determination of some acetate esters (methyl, 2-hydroxyethyl and phenyl acetate) by following the rate of second-order saponifi-cations at two temperatures 15" and 30°C.The graphical method of Hanna and Siggia,w a modified version of the initial rate method can be used for evaluation. Ethanolamides. Fatty alkanolamines can be saponified in alcoholic potas-sium hydroxide solutions when 2 moles of the weak base react with 1 mole of potassium hydroxide. This difference in the stoicheiometry was applieds5 to the determination of ethanolamides in mixtures with amines amine soaps and amide esters which behave in a different way. The unreacted amount of potassium hydroxide is determined after a certain time by titrating it with hydrochloric acid. Results are evaluated from a calibration curve where the amount of saponified sample is plotted against its concentration.The calibration curve is linear and the error of the determination of 10 to 75 mol per cent. of diethanolamide generally does not exceed a few per cent. Ketones. Toren and Gnuseg6 described a method for the determination of ketones in binary mixtures based on their different rates of reduction by hydroxyl-amine. The reactions can be followed by an automatic titrimeter (operated as a pH-stat) and the volume of titrant needed for the restoration of the original pH is recorded as a function of time. The precision of the method depends on the difference in the rate constants for the reduction of each ketone. If the two constants differ by a factor of 5 components of mole of total ketones can be determined with a coefficient of variation of 1 per cent.while if the factor is 3 the coefficient of variation increases to 3 per cent. Naphthols. For the kinetic determination of 1- or 2-naphthol Babking7 recommended that bromine be liberated from potassium bromate and potassium bromide in acid solution. This reacts with 1- or 2-naphthol present in the mixture, by forming a colourless brominated product. If methyl orange is present the excess of bromine will attack the dye and decolourise it. The time needed to decolourise the methyl orange is under controlled experimental conditions proportional to the amount of naphthol present. Up to 250 mg 1-1 of 1- or 2-naphthol can be deter-mined with an error of less than &5 per cent. Up to 1 g 1-l of chloride does not interfere but sulphate phenols and other organic substances that react with bromine do.Oxidative enzymes and their substrates. Monoamine and diamine oxidase can be determined with a kinetic methodg8 that is based on the reaction of fi-hydroxyphenyl-acetic acid with the enzyme to yield a fluorescent product 260 SVEHLA Hydrogen peroxide has to be present. The fluorescence of the product is monitored by excitation at 317 nm and measurement of the emission at 414 nm. The initial rate method can be applied to calculate results. The method can also be applied to the determination of the substrates of the enzymes such as cadaverine histamine, putrescine benzylamine tyramine and furfurylamine. The error of the method for the concentration ranges quoted in the paper is generally below -3 per cent.Pesticides. The hydrolysis of umbelliferone phosphate achieved by phos-phatase enzyme is inhibited by certain pesticides such as aldrin and heptachlor. This method was applieds to the determination of 5 to 1oOpg ml-l amounts of aldrine and 50 to 700pg ml-l of heptachlore. The method is identical to that discussed in connection with beryllium. Phenols. A method similar to that of Schenk and Browng2 for the deter-mination of dihydric phenols has been described.99 Phenol o-cresol $-cresol and 2,6-xylenol can be determined up to a concentration of 5 mmoles 1-1 with an error below 3 per cent. The accuracy of the determination of 3,4-xylenol and m-cresol was less satisfactory. Serum enzymes. Kinetic methods are frequently applied in clinical labora-tories for the determination of serum enzymes.SzaszlOO described a kinetic-photometric method for the determination of serum leucine aminopeptidase. Knedel and BottgerlOl use a fixed concentration method with photometric measure-ments for the determination of cholinesterase activity in serum. The optimum conditions for the determination of serum alkaline phosphatase were investigated by Hausamen Helger Rick and Gross.lo2 The well known method of Oliver103 was modifiedlo4 for the determination of serum creatine kinase applying commercially available reagents and a spectrophotometric measurement. Qualitative Catalytic Analysis Most of the quantitative methods reviewed above are suitable for qualitative testing also. There are however a few new techniques described specifically as qualitative tests that are reviewed here.Robinsonlo5 recommends a method for the identification of alcohols. First their 3,5-dinitrobenzoate esters are prepared and are then hydrolysed in alkaline medium. The rate of hydrolysis can be measured by spectrophotometry at 285 nm. From the rate of hydrolysis and the melting point of the ester twenty-seven different alcohols can be identified. A specific and sensitive test for arsenic was described by Krause and Slawek.106 The oxidation of indigo carmine by hydrogen peroxide is catalysed by arsenic and g of arsenic can be detected by the decolourisation of the dye in a concentration limit of 1 los. For the detection of barium Hamya and Townshendlo7 applied an induced precipitation method.This is based on the fact that under the circumstances described in the paper the formation of lead sulphate precipitate is strongly accelerated by the presence of as little as 0.4 pg of barium. Winterlo* reports on the identification (and semi-quantitative determination) o CATALYTIC METHODS IN ANALYTICAL CHEMISTRY 261 catalase in plant extracts. The method is based on the gas volumetric measure-ments on the initial rate of the decomposition of hydrogen peroxide. LegradiBO recommends the method reviewed among the quantitative determinations of organic substances for the detection of o-chloronitrobenzene and related sub-stances. The detection of iodide can be achieved by using the same reaction as is used for its determinati~n.~~ For the identification of sugars Mikkelson and Robinsonfog recommend the spectrophotometric measurement of the rate of formation of oximes with hydroxylammonium chloride.Fifteen different sugars can be identified by the method. Application of Catalytic Waves in Polarography Some new quantitative methods recommended in the past 3 years involve the use of polarographic catalytic waves. Kolthoff and MaderllO observed two catalytic hydrogen waves in solutions containing cobalt(I1) or cobalt(II1) ions and mercapto-anilines. These waves could be used for the determination of cobalt and showed different characteristics to the Brdickalll waves. Later Mader and Kolthoff llC showed that in some cases adsorption phenomena occur in the polarography of similar systems. A catalytic reduction wave of nitrate occurs in acid solutions containing traces of uranium.This phenomenon can be used for the determination of uranium.l13 Although the current measured is not a linear function of the uranium concentration uranium can be determined with a coefficient of variation of &5 per cent. Toropova and Anisimova,ll* on the other hand dealt with solutions of 8-mercaptoquinolineJ and found that in acetate buffers this substance causes the evolution of catalytic hydrogen waves that can be used for its deter-mination. They extended the method for the determination of metals that react with 8-mercaptoquinone producing a smaller catalytic wave. The theory of kinetic currents has also been developing in the past 3 years. Pence Delmastro and Boomanlfs dealt with the determination of rate constants of very fast regeneration reactions at spherical electrodes.Delmastro116 later demonstrated that in practical cases the ratio of the kinetic current to the purely diffusion controlled current at a stationary spherical electrode is always less than two. Delmastro and Booman117 also dealt with polarographic kinetic currents of the first order. The kinetics of the adsorption of the reaction product in stationary electrode polarography has been examinedll* both theoretically and experi-mentally. Kinetochromic Spectrophotometry West and his co-workers have recently developed a technique that they call kinet ochromic spec trophot ometry . Certain ligand-exchange reactions in which zirconium ions take part seem to be catalysed by substoicheiometric amounts of certain anions.The product of these reactions is coloured and can be measured by spectrophotometry. However it is not the reaction rate that is monitored (and therefore these methods form a different class among catalytic techniques) but th 262 SVEHLA extinction of the final product is measured after the elapse of a certain minimum time when it remains constant. A closer examination of these processes showed that the role of the catalyst is to achieve a certain degree of depolymerisation of the hydrolysed zirconium polymers. The amount of depolymerised zirconyl ions is proportional to the amount of the catalyst ions present. As the complex is formed from the depolymerised zirconyl ions the extinction of the solution after sufficient time is allowed for the formation of the complex becomes proportional to the catalyst concentration.On this principle a method for the determination of 0.005 to 0.05 pg ml-l of fluoride (in the final volume) by using the xylenol orange -zirconium reaction has been reported.l19 Both sulphatef20 and fluoride121 have been determined by making use of the reaction between zirconium and methylthymol blue. Interferences were studied carefully and methods of their elimination were suggested. Catalymetric Titrations The principle of catalysis can also be applied for the end-point detection of titrations although it is not generally the catalyst that is determined. The catalyst is used as a titrant and until the equivalence point is reached it is used up by the titration reaction.The excess of the catalyst initiates a catalysed reaction that is accommpanied by a change of colour (or temperature or electrode potential), which indicates the end-point. As the catalyst is present in large concentrations, these catalysed reactions at the end-point are fast almost instantaneous. Follow-ing up earlier works on the subject Weisz and Janjic122 introduced catalymetric end-point detection into complexometric titrations. The oxidation of certain organic substances by hydrogen peroxide takes place only if free metal ions are present. On adding an excess of a complexing agent to a solution of metals the excess of the reagent can be back titrated with a solution of a metal by using the catalytic reactions mentioned for end-point detection. Weisz and Kiss123 used the same principle but instead of a visual method they measured the temperature of the mixture which increases rapidly when the end-point is reached.This thermo-metric-catalytic method was used earlierx2* for the end-point detection of precipita-tion titrations. Silver mercury and palladium can be titrated with potassium iodide by using the reaction between cerium(1V) and arsenic(II1) ions for end-point detection. The method can also be applied to the indirect determination of anions that react with these metals. Earlier a potentiometric measurement for end-point detection based on the same catalytic reaction had been reported.lZ5 Fedorova and Yatsimirskii126 determined palladium on the basis of its inhibition effect on the cerium(1V) - arsenic(II1) reaction catalysed by iodide.Bognar and Sarosil27 applied Landolt reactions for the end-point detection of titrations that involve silver or mercury as well as iodide as reactants. Mottola128 determined microgram amounts of anionopolycarboxylic acids and metals with manganese(I1) ions as titrant. The reaction between malachite green and periodate ions is used as the reaction indicating the end-point and the extinction of the mixture is measured CATALYTIC METHODS IN ANALYTICAL CHEMISTRY 263 The method is entirely automatic and with the simple instrumentation described in the paper recorded titrigrams are obtained. applied the principle of differential thermometric end-point detection for catalymetric titration of cobalt, based on its catalytic action on the reaction between tartaric acid and hydrogen peroxide.Vajgand and GaaPo used an acetic acid medium for their catalytic thermometric titrations. Tertiary amines and salts of organic acids can be titrated with perchloric acid. Acetic anhydride must be present and does not react with the water that is produced by the titration reaction until free perchloric acid is present. The increase in temperature accompanying the reaction can be measured by a thermometer. Later the same principle was used for the determination of triethyl-amine brucine potassium acetate and dimethyl aniline.131 The procedure was automated at the same time. Tertiary amines and salts of organic acids have been determined by catalymetric titration in acetic acid by using a coulometrically generated tit rant .lSz New Techniques and Instruments There was considerable progress made in experimental techniques and instrument ation of catalytic analysis.Many authors recommend the use of computers for the evaluation of results. Thus CrouchlS described a small all-electronic device that computes the reciprocal reaction time for use in rate calculations. Parker Pardue and Willisla describe a miniature computer that can be used for catalytic ox kinetic analytical purposes. A larger model was recommended earlier.131 All of these were analogue computers. The application of digital computers in analytical chemistry including the solution of rate equations has been reviewed.136 Innovations in the experimental technique involving potentiometric techniques were made by several authors.An electrode capable of measuring urease enzyme activity has been recommended.la7 Guilbault Smith and Montal~ol~~ used an electrode sensitive to ammonium ions for studying the kinetics of deaminase enzyme systems. Sand and H ~ b e r l ~ ~ recommended the application of differential constant-current potentiometry for kinetic analysis. In this method the potential between two polarised cathodes or anodes is measured and recorded as a function of time. These graphs can be evaluated by using the initial rate technique. The potentiostatic methodg3 was discussed when dealing with the determination of molybdenum. Alexander and BarclaylQ0 described a new technique ‘current-cessation chronopotentiometry,’ and its analytical applications. An electrical pulse is passed through a coulometric cell containing certain metal ions and anions for 30 to 200 s and the potential of a working electrode is measured as a function of time.The time required for the potential of the working electrode to reach its maximum rate of decay is inversely proportional to the concentration of one of the substances in the cell. Olmsted and Nicholsonl*l recommended the use of a double potential step method for measuring rate constants of dimerization reactions. The method can also be applied for catalytic analyses 264 SVEHLA Interesting developments on the application of spectrophotometric methods to catalytic and kinetic analysis have also been made in the past 3 years. The automated fast reaction rate method in the millisecond range has already been reviewed in connection with the determination of phosphate.8l A semi-automatic instrument for the continuous measurement of concentrations has been reported.lg2 The core of the instrument is a spectrophotometer with a high stability and low noise.The auto-andyser was modified and adaptedlg3 for reaction rate measure-ments of multiple enzyme samples by continuous-flow analysis. The use of thermoanalytical methods in kinetic and catalytic analysis has been recommended by several authors. Besides the methods reviewed in connection with catalyrnetric titrations new techniques have been suggested for the measurement of reactions rates by thermoanalytical methods. These methods are interesting in that they can be used to measure the rate of reaction of substances in the solid (or molten) phase.Taylor and Watsonlg4 applied differential thermoanalysis to the measurement of relative reaction rates. Thermogravimetric analysis has been used for the determination of the rate constant and activation energy of various degradation rea~ti0ns.l~~ Sharp Sally and Went~orthl*~ examined and compared three methods of obtaining kinetic information from thermogravimetric curves. For the evaluation of data obtained by non-isothermal kinetic measurements, Kresze Kosbahn and WinMerlP7 described a method involving the use of a digital computer. The efficiency of mixing of reagents and the resulting instrumental delay effect on recorded kinetic curves have been examined,148 and the findings are relevant to methods involving both slow and fast reactions.Theory of Catalytic Analysis Reviews Many of the papers mentioned so far contain some discussion of the kinetics and reaction mechanism of the processes involved. There are a number of papers, however containing predominantly theoretical material in which the aim of the author is to examine a certain reaction or a principle regardless of its practical applications. Most of the theoretical papers deal with the kinetics and mechanism of a single reaction or a special group of reactions. Goldman and H a r g i ~ l ~ ~ studied the kinetics of the formation and reduction of phosphorous-bismuth dimeric heteropolymolybdate a reaction that can be applied for the determination of bismuth. Their results indicate that the heteropolymolyb-date complex contains eighteen atoms of molybdenum one atom of phosphorus and one atom of bismuth.The role of cobalt in catalytic reactions was studied by several authors. The oxidation of Tiron with hydrogen peroxide in the presence and absence of cobalt was examined,13J4 and a mechanism for the catalysed reaction was postulated. The rate constant was measured and used to explain the calibration curves obtained. The kinetics of oxidation of the cobalt(I1) bipyridyl complex by copper(I1) and iron(II1) perchlorates in anhydrous aceto-nitrile were studied by using polarographic measurements with a rotating platinum e1ectr0de.l~~ The rate equations of these processes were postulated and verified CATALYTIC METHODS IN ANALYTICAL CHEMISTRY 265 Both reactions are first order with respect to the cobalt complex but second order with respect to iron(II1) while the order with respect to copper is 0.5.Heller and Guyon2s when reporting on their catalytic-spectrophotometric determination of copper made detailed kinetic measurements on the rate of the reduction of iso-polymolybdate with ascorbic acid both in the presence and absence of the catalyst. Reactions between more than thirty metals and traws-l,2-diaminocyclohexane N,N,N',N'-tetraacetate were in~estigatedl~l and the respective rate constants of the reactions were determined. Some ligand-exchange reactions between metals and between hydrogen ion and a metal were also examined. An interesting correlation between the atomic number of the element and rate constants within the series of lanthanides has been discovered enabling a differential kinetic method to be used for their determination in binary mixtures.The 'classical' reaction between cerium(1V) and arsenic(II1) ion catalysed by iodide has also been investigated.152 It is interesting to note that although this reaction is widely used in industrial and clinical laboratories for the determination of iodine the kinetics and mechanism of the reaction have not before been thoroughly examined and interpreted. A highly complex rate equation with no less than seven rate constants has been derived and verified by experimental results. The kinetics of the reaction between perborate and iodide ions catalysed by iron and molybdenum in acid medium has been in~estigated~l; the effects of osmium were also noted.The rate constants of the reactions between hydrogen peroxide and acid chrome dark blue, both in the presence and absence of iron as catalyst have been determined.% The kinetics of the oxidation which is catalysed by manganese of Alizarin S with hydrogen peroxide have been studied.40 This reaction proceeds in slightly alkaline solution. The kinetics of the ligand-exchange reaction between ethylenediamine tetraacetic acid and its nickel complex have been elucidated by using deuterated reagents and n.m.r. techniques.153 An exhaustive study of the catalytic properties of compounds of the platinum group lead to the discovery of new oxidation -reduction reactions that proceed at room temperature with a measurable rate and are catalysed by various of the compounds.lM The kinetics and mechanism of the reactions were investigated and possibilities of analytical applications outlined.Reactions involving vanadium were thoroughly studied by various groups of investigators; Fuller and O t t a ~ a y l ~ ~ investigated the oxidation of vanadium(1V) in acid medium by bromate. The rate equation contains not only a rate constant but also an equilibrium constant and while the reaction is first order with respect to vanadium(IV) a mixed order was found with respect to bromate. The oxidation of Bordeaux with bromate in acid medium was studied both in the presence156 and absence16' of vanadium. The uncatalysed reaction is a first-order process with respect to Bordeaux but the contribution of bromate to the rate cannot be explained so simply as the rate is influenced by the concentration of bromide that is inevitably present in bromate as a trace impurity.The process at the same time is autocatalytic for which the rate equation derived by the authors allows. The process catalysed by vanadium is more complex as it exhibits an induction period and the process is autocatalytic as well. The mechanism158 and the kinetics150 o 266 SVEHLA the oxidation of p-phenetidine with chlorate and the role of vanadium in these processes have been investigated. The role of vanadium is not merely that of an electron donor and acceptor (though it exists in both the tetravalent and penta-valent state in the reaction cycle) but it also forms a complex with charge transfer transition from which an arylamino radical is then formed.Bontchev and MladenowalG0 examined the kinetics of the oxidation of vanadium(1V) by chlorate and the effects of hydroxycarboxylic acids on the reaction rate. Thompson and S ~ e h l a ~ ~ and Thompson31 examined the kinetics of the oxidation of iodide with bromate in the presence and absence of vanadium. The same system was examined later by Bognar and JellineklG1 under somewhat different experimental conditions. The competing role of ascorbic acid with vanadium in the Landolt reaction was emphasised by them. While the contributions reviewed above were dealing with specified reactions, quite a number of theoretical papers tried to cover a broader field and to give explanations and theories of a more general nature. BontcheP2 dealt with reaction mechanisms in catalyitc analysis showing how the knowledge of the mechanism of a catalytic reaction can help in establishing the optimum experimental conditions and in developing new catalytic methods.The possibility of increasing the sensi-tivity of these procedures based on a knowledge of the reaction mechanism enhances the value of this paper. Svehlal@ reviewed the applications of Landolt reactions in quantitative analysis. Simple methods for the kinetic examination of the uncatalysed and catalysed reactions are outlined and a generally adaptable kinetic explanation has been given. On this basis an explanation of the particular shape of the calibration curve can be made. A general treatment on the selectivity, sensitivity and precision of Landolt reactions is also included.Catalytic methods using ligand-exchange reactions with masking agents and automatic rate measure-ments have been reviewed164 and those co-ordination chain reactions mentioned in the earlier part of this review were then first postulated. This field of catalytic analysis has hardly been explored yet the sensitivity and precision of these methods appears to be promising. The lack of selectivity of these techniques can partly be overcome by the application of a second complexing (masking) agent. Mottola165 reviewed the fast developing field of catalymetric titrations. The general theory presented in his paper is applicable to almost all existing methods and will no doubt help to find new applications of this technique. Lukasiewicz and Fitzgerald36 outlined the theory of their new technique called photochemical kinetic analysis which is distinguished by its high degree of selectivity.The reviews of Rechnitz166 and Guilbault16' on catalytic methods published among the annual reviews of AnalyticaE Chemistry can be read with interest. The author thanks Professor C. L. Wilson Mr. P. J. McKenna and Mrs. M. Scullion for assistance in the preparation of the manuscript CATALYTIC METHODS IN ANALYTICAL CHEMISTRY 267 1 2 3 4 6 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 References Mark H. B. and Rechnitz G. A.‘Kinetics in Analytical Chemistry. Chemical Analysis,’ Yatsimirskii K. B. “Kinetic Methods of Analysis. International Series of Monographs Delahay P. and Stiehl G. J. J . Amer. Chem. Soc. 1952 74 3500. Michaylova V. and Bontchev P. R. Michrochim. Acta 1970 344. Bognar J. and Pataky-Szabo M. Ibid. 1969 221. Pausch J. B. and Margerum D. W. Analyt. Chem. 1969,41 226. Townshend A. and Vaughan A. Talanta 1969 16 929. Guilbault G. G. Sadar M. H. and Zimmer M. Analytica Chim. Acta 1969 44 361. Hargis L. G. Analyt. Chem. 1969 41 597. Townshend A. and Vaughan A. Talanta 1970,17 289. Jasinskiene E. and Bilidiene E. Zh. analit. Khim. 1967 22 741. - and - Ibid. 1968 23 143. Kucharkowski R. and Doge H. G. 2 . analyt. Chem. 1968,238 241. Kucharkowski R. Ibid. 1970,249 22.Popa G. and Costache D. Revue roum. Chim. 1067 12 963. - and- Ibzd. 1968 13 573. Kreingold S. U. and Bozhevolnov E. A. Trudy Kom. analit. Khim. 1968 16 194. Pall A. Svehla G. and Erdey L. Talanta 1970 17 211. Hrabeczy-Pall A. Ph.D. Thesis Technical University Budapest 1969. Hesselbarth H. Z. analyt. Chem. 1969,248 163. Dittel F. Ibid. 1967,229 193. Colovos G. and Papadopoulos C. Mikrochem. J. 1968 13 236. Nedved J. Chem. Listy 1968 62 698. Heller R. L. and Guyon J. C. Analyt. Chem. 1968,40 773. Guilbault G. G. Kramer D. N. and Hackley E. Analyt. Biochem. 1967 18 241. Cyganok L. P. Cujko V. T. and Reznik B. E. Zav. Lab. 1967,33 5. Michalski E. and Gelowa H, Chemia Analit (Warsaw) 1967 12 147. Thompson H. and Svehla G. Z. analyt. Chem. 1969,247 244. Thompson H.Ph.14. Thesis Queen’s University of Belfast 1970. Orav M. and Kokk K. Uchen. Zap. tcartu. gos. Univ. 1968 207. Kreingold S. U. and Bozhevolnov E. A. 21%. analit. Khim. 1967 22 218. Kreingold S. U. Bozhevolnov E. A. and Antonov V. N. Zav. Lab. 1968,34 260. Kharlamov I. P. Dodin E. I. and Mantsevich A. D. Zh. analzt. Khim. 1967 22 371. Lukasiewicz R. J. and Fitzgerald J. M. Anal. Lett. 1969 2 159. Kun E. J. Biol. Chem. 1947 170 509. Hadjiioannou T. P. and Kephalas T. A. Mikrochim. Acta 1969 1215. Malmstadt H. V. and Hadjiioannou S. I. Analyt. Chem. 1962 34 452. Janjic T. J. Milovanovic G. A. and Celap M. B. Ibid. 1970 42 27. Mealor D. and Townshend A. Talanta 1968 15 747. Townshend A. and Vaughan A. Ibid. 1970 17 299. Weisz H. Klockow D. and Ludwig H. Ibid.1969 16 921. Lazarev A. I. Zh. analit. Khim. 1967,22 1836. Mealor D. and Townshend A. Analytica Chim. Acta 1967 39 235. Gregorowicz Z. and Suwinska S. Mikrochim. Acta 1967 546. Rodriguez P. A. and Pardue H. L. Analyt. Chem. 1969,41 1376. Ottaway J. M. Fuller C. W. and Allan J. J. Analyst 1969 94 522. Bontchev P. R. and Alexiev A. A. Mikrochim. Acta 1968 875. Alexiev A. and Bontchev P. R. Ibid. 1970 13. Bontchev P. R. Alexiev A. and Dimitrova B. Talanta 1969 16 597. Menghani G. D. and Bakore G. V. Bull. Chhem. SOL Jafian 1968,41 2574. Schenk G. H. and Bazzelle W. E. Analyt. Chem. 1968 40 162. Hadjiioannou T. P. and Valkana C. G. Hem. Hron. 1967 32 89. Thompson H. and Svehla G, Microchem. J. 1968 13 576. Bognar J. and Jellinek O. Mikrochim. Acta 1967 193.Interscience Vol. 24 1968. of Analytical Chemistry,’ Pergamon Press Vol. 25 1966. - Ibid. 1969,248 167. - Ibid. 1969,248 289 268 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 SVEHLA Bognar J. and Sarosi Sz. Ibid. 1967 813. Bognar J. and Jellinek O. Ibid. 1969 366. Fuller C. W. and Ottaway J. M. Analyst 1970 95 41. Tanaka M. and Awata N. Analytica Chim. Acta 1967 39 485. Toropova V. F. and Tamarchenko L. M. Zh. analit. Khim. 1967,22 576. Krueger R. H. Warriner J. P. and Jaselkis B. Talanta 1968 15 741, Krause A. Zielinski S. and Weimann T. 2. Analyt. Chem. 1968,237 193. Krause A. Domka F. and Marciniec B. Chemia Analit. (Warsaw) 1968 13 883. Ballzo H. 2. Analyt. Chem. 1969 245 20. Toropova V. F. and Tamarchenko L.M. Zh. analit. Khim. 1967,22 234. Proskuryakova G. E. Ibid. 1967,22 802. Bogurth W. and Schaeg W. Mikrochim. Acta 1967 658. Bognar J. and Nagy L. Ibid. 1969 108. Ottaway J. M. Fuller C . W. and Rowston W. B. Analytica Chim. Acta 1969 45 541. Iwasaki I. Utsumi S. and Ozawa T. Bull. Chem. Soc. Japan 1953 26 108. Sandell E. B. and Kolthoff I. M. J . Amer. Chem. Soc. 1934 56 1426. - and - Mikrochim. Acta 1937 1 9. Knapp G. and Spitzy H. Talanta 1969 16 1353. - and - Ibid. 1969 16 1361. Appleby A. and Spillett R. E. U . K . Atomic Energy Authority Report RCC-M 1967, Comoy E. Revue fr. Etud. Clin. biol. 1967 12 189. Godin J. M. and Archimbaud M. Rapp. CEA No. 3145 1967. Margerum D. W. and Stehl R. H. Analyt. Chem. 1967,39 1351. Crouch S. R. and Malmstadt H.V. Ibid. 1967 39 1090. Javier A. C. Crouch S. R. and Malmstadt H. V. Ibid. 1969 41 239. West P. W. and Ramakrishna T. V. Ibid. 1968 40 966. Bognar J. and Sarosi Sz. Mikrochim. Acta 1969 361. Kawashima T. and Tanaka M. Analytica Chim. Acta 1968 40 137. Kawashima T. Nakano S. and Tanaka M. Ibid. 1970,49 443. Babko A. K. and Maksimenko T. S. Zh. analit. Khim. 1967 22 550. Legradi L. Magy. Kern. Lap. 1967,22 488. Shresta I. L. and Das M. N. Analytica Chim. Acta 1970 50 135. Bognar J. and Jellinek O. Mikrochim. Acta 1969 312. Legradi L. 2. analyt. Chem. 1968 237 426. Mottola H. A, Haro M. S. and Freiser H. Analyt. Chem. 1968 40 1263. Schenk G. H. and Brown D. J. Talanta 1967 14 257. Munnelly T. I. Analyt. Chem. 1968 40 1494. Hanna J. G. and Siggia S.Ibid. 1962 34 547. Lohman F. H. and Mulligan T. F. Ibid. 1969 41 243. Toren E. C. jun. and Gnuse M K. Analyt. Lett. 1968 1 295. Babkin M. P. Zh. analit. Khim. 1968,23 637. Guilbault G. G. Kuan S. S. and Brignac P. J. jun, Analytica Chim. Ada 1969, Kwiatkowska I. and Kwiatkowski E. Chemia Analit. (Warsaw) 1968 13 783. Szasz G. Amer. J . Clin. Path. 1967 47 607. Knedel M. and Bottger R. Klin. Wschr. 1967 45. 325. Hausamen T. V. Helger R. Rick W. and Gross W. Clin. Chim. Acta 1967 15 241. Oliver I. T. Biochem. J. 1955 61 116. Hess J. W. MacDonald R. P. Natho G. J. W. and Murdock K. J. Clin. Chem. 1967, Robinson J. R. Analyt. Chem. 1967 39 1178. Krause A. and Slawek J. 2. analyt. Chem. 1969 245 44. Hamya J. W. and Townshend A. Analytica Chim. Acta 1969 46 312.Winter E. 2. analyt. Chem. 1969 244 248. Mikkelson T. J. and Robinson J. R. J . Pharm. Sci. 1968 57 1180. Kolthoff I. M. and Mader P. Analyt. Chem. 1969 41 924. Brdicka R. Colln Czech. Chem. Commun. 1936 8 366. Mader P. and Kolthoff I. M. Analyt. Chem. 1969 41 932. Habashi F. and Thurston G. A, Ibid. 1967 39 242. p. 210, 47 503. 13 994. 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 11 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 CATALYTIC METHODS IN ANALYTICAL CHEMISTRY 269 Toropova V.F. and Anisimova L. A. Zh. analit. Khim. 1967 22 1264. Pence D. T. Delmastro J. R. and Booman G. L. Analyt Chem. 1969,41 737. Delmastro J. R. Ibid. 1969 41 747. Delmastro J. R. and Booman G. L. Ibid. 1969.41 1409. Hulbert M. H. and Shain I. Ibid. 1970,42 162. Cabello-Thomas M. L. and West T. S. Talanta 1969 16 781. Hems R. V. Kirkbright G. F. and West T. S. Ibid. 1969 16 789. - - and - Ibid. 1970 17 433. Weiiz H. and Janjic T. 2. analyt. Chem. 1967 227 1. Weisz H. and Kiss T. Ibid. 1970 249 302. Weisz H. Kiss T. and Klockow D. Ibzd. 1969 247 248. Weisz H. and Klockow D. Ibid. 1967 232 321. Fedorova T. I. and Yatsimirskii K.B. Zh. analit. Khim. 1967 22 283. Bognar J. and Sarosi Sz. Mzkrochczm. Acta 1969 463. Mottola H. A. Analyt. Chem. 1970,42 630. Sajo I. Talanta 1968 15 578. Vajgand V. J. and Gaal F. F. Ibid. 1967 14 345. Vajgand V. J. Kiss T. A. Gaal F. F. and Zsigrai I. J. Ibzd. 1968 15 699. Vajgand V. J. Gaal F. F. and Brusin S. S. Ibid. 1970 17 415. Crouch S. R. Analyt. Chem. 1969,41 880. Parker R. A. Pardue H. L. and Willis B. G. Ibid. 1970 42 56. James G. E. and Pardue H. L. Ibid. 1968 40 796. Childs C. W. Hallman P. S. and Perrin D. D. Talanta 1969 16 1119. Montalvo J. G. jun. Awalyt. Chem. 1969 41 2093. Guilbault G. G. Smith R. K. and Montalvo J. G. jun, Ibid. 1969 41 600. Sand J. R. and Huber S. O. Ibid. 1970 42 238. Alexander W. A. and Barclay 14. J. J. Electroanal. Chem. 1967 13 137. Olmstead M. L. and Nicholson R. S. Analyt. Chem. 1969 41 851. Weichselbaum T. E. Plumpe W. H. jun. Adams R. E. Hagerty J. C. and Mark, Brown H. H. and Ebner M. R. Clin. Chem. 1967 13 847. Taylor L. J. and Watson S. W. Analyt. Chem. 1970 42 297. Farre-Rius F. Huret J. Puyo M. and Guiochon G. Analytica Chim. Ada 1969 45, Sharp J. H. and Wentworth S. A. Analyt. Chem. 1969,41 2060. Kresze G. Kosbahn W. and Winkler J. 2. analyt. Chem. 1967 231 1. Caselli M. Cavaggioni A. and Papoff P. Talanta 1968 15 1335. Goldman H. D. and Hargis L. G. Analyt. Chem. 1969 41 490. Nemec I. Kies H. L. and Nemcova I. Analytica Chim. Acta 1970 49 541. Margerum D. W. Pausch J. B. Nyssen G. A. and Smith G. F. Analyt. Chem. 1969, Rodriguez P. A. and Pardue H. L. Ibid. 1969,41 1369. Can J. D. and Reilley C . N. Ibid. 1970 42 51. Yatsimirskii K. B. Kalinina V. E. Morozova R. P. and Federova T. L. Trudy zvanovsk. Khim-tekhnol. Inst. 1968 Jubilee No. 68. Fuller C. W. and Ottaway J. M. Analyst 1969 94 32. - and - Ibid. 1970 95 28. - and - Ibid. 1970 95 34. Bontchev P. R. and Jeliazkowa B. G. Mihrochim. Acta 1967 116. -and - Ibid. 1967 125. Bontchev P. R. and Mladenowa Z. Ibid. 1968 427. Bognar J. and Jellinek Ibid. 1969 318. Bontchev P. R. Talanta 1970 17 409. Svehla G. Analyst 1969 94 513. Margerum D. W. Stehl R. H. and Latterell J. J. Analyt. Chem. 1967 39 1346. Mottola H. A. Talanta 1969 16 1267. Rechnitz G. A. Analyt. Chem. 1968,40 455R. Guilbault G. G. Ibid. 1970 42 334R. H. B. jun. Ibid. 1969 41 725. 463. 41 233