摘要:

122 Analyst, February, 1973, Vol. 98, PP. 122-125 A Gas-chromatographic Method for the Determination of Low Concentrations of Acrylic Acid in Mixtures of C2 to C5 Fatty Acids in Biological Materials BY R. C. NOBLE AND J. W. CZERKAWSKI (Hannah Research Institute, Ayr, Scotland, KA6 6HL) The presence of acrylic acid in mixtures of short-chain (C, to C,) fatty acids can be determined by gas chromatography by using a composite column technique. By varying the proportions of the total column length occupied by non-polar and polar liquid phases, the acrylic acid peak could be “moved” to a predetermined position on the chromatogram and complete separation from the other acids could be achieved. Suitable placing of the peak enabled the concentration of acrylic acid to be accurately and quickly determined.MANY micro-organisms convert pyruvic acid into propionic acid. This conversion can be achieved in two ways, either through the metabolic pathway, which involves succinate its an intermediate, or by the direct reduction pathway, in which lactic acid produced from pyruvic acid is dehydrated to acrylic acid, which in turn is reduced to propionic acid.1 In the rumen, where microbial fermentation is such that when readily fermentable carbohydrate is fed there is a large accumulation of propionic acid, the reduction pathway may predominate and acrylic acid may well accumulate., Although the concentrations of the main short-chain fatty acids (C, to C,) that occur in the rumen can readily be measured by gas - liquid chromato- graphic methods, the determination of acrylic acid presents some problems that do not appear to have been resolved.The work of Whanger and Matronel and later that of Czerkawski and Breckenridge3 has shown that when these short-chain fatty acids are subjected to gas - liquid chromatography on polar columns, the retention volume of acrylic acid is indistinguishable from that of butyric acid; hence, in such circumstances, acrylic acid remains undetected. The apparent increase in the concentration of butyric acid when lactic acid was fermented by rumen micro-organisms, as observed by other worker^,^ may be explained by this inability of the column to separate butyric and acrylic acids. Details are presented in this paper of a simple and sensitive gas - liquid chromatographic method for the determination of acrylic acid in complex mixtures of short-chain fatty acids.EXPERIMENTAL Samples of rumen contents were obtained from sheep that were receiving a diet of molassed sugar-beet pulp, and the short-chain (C, to C,) fatty acids were prepared for analysis according to the method of Cottyn and Boucque.6 Standard solutions were prepared by dissolving known amounts of the pure free acids in water. Aliquots (2 to 3 p1) of the solutions were analysed by on-column injection into a standard gas - liquid chromatograph (Pye 104 or Perkin-Elmer Fll) fitted with glass columns and maintained under isothermal temperature conditions. Two types of liquid phases were used: 6 + 95 m/m Carbowax 20M TPA on Chromosorb G (80 to 100 mesh, acid washed and silanised) and 15 + 85 m/m Apiezon L on Chromosorb W (100 to 120 mesh, acid washed, silanised and containing 0.5 per cent.of phosphoric acid). The detector response factors for each acid were obtained by using the standard solutions. The amount of each fatty acid present was calculated by automatic integration and the results obtained were expressed as moles per 100 ml by multiplying this value by the appropriate factor. RESULTS AND DISCUSSION Acrylic acid is co-chromatographed with butyric acid on columns that contain Carbo- wax 20M or a similar polar liquid phase (see Fig. 1) over a wide range of temperatures. An initial attempt was made to utilise this fact to estimate the concentration of acrylic acid by difference after bromination or hydrogenation of the samples in the presence of small amounts of palladised charcoal.Both of these methods proved to be satisfactory, provided 0 SAC and the authors.NOBLE AND CZERKAWSKI 123 that the concentration of acrylic acid was equal to or greater than that of propionic acid and butyric acid. In the rumen, however, the concentration of acrylic acid is usually very much less than that of butyric acid, which in turn is normally about half of the concentration of propionic acid. Furthermore, the detector response factor for acrylic acid is only about half of that for butyric acid. Hence, attempts to determine the concentration of acrylic acid by measuring the decrease in the butyric acid peak or the increase in the propionic acid peak following hydrogenation proved to be insufficiently accurate to be acceptable.t Injection Fig. 1. Gas - liquid chromato- graphy of short-chain (C, to C,) fatty acids containing acrylic acid: (a) 16 per cent. Apiezon L and 0-5 per cent. phosphoric acid on Chromosorb W a t 90 "C; (b) 6 per cent. Carbowax 20M TPA on Chromosorb G a t 146 "C; and (c) composite column consisting of 86 parts of Apiezon L and 16 parts of Carbowax 20M TPA a t 96 "C. The carrier gas was nitrogen a t a flow-rate of 30 ml min-l Further exploratory work showed that the separation of acrylic acid from butyric acid could be achieved on non-polar columns such as Apiezon L (see Fig. 1). However, with non-polar columns, the separation of the other short-chain fatty acids was not as satisfactory as that achieved on the polar phases; consistent tailing of peaks was obtained, which reduced the accuracy of quantitative measurements.Furthermore, the retention volume of acrylic acid on the non-polar phases was only slightly less than that of propionic acid and it proved to be impossible to detect relatively small concentrations of acrylic acid in the presence of propionic acid. Isobutyric acid, which on a Carbowax 20M column emerged soon after propionic acid, behaved in a similar manner on the Apiezon L column. It therefore seemed124 NOBLE AND CZERKAWSKI: GAS CHROMATOGRAPHY OF [Analyst, Vol. 98 TABLE I CONCENTRATIONS OF ACRYLIC ACID AND BUTYRIC ACID IN RUMEN FLUID AFTER THE ADDITION OF KNOWN CONCENTRATIONS O F ACRYLIC ACID Concentration/mequiv per 100 ml A r I Acrylicacid . . .. 2.0 1.6 1.2 0.8 0.4 Relative peak areas of Butyric acid .. .. 1.0 1.0 1.0 1.0 1.0 acrylic to butyric acid (mean of 4 determina- tions f standard error) 2.10 f 0.18 1.72 f 0.16 1-28 f 0.11 0.86 f 0-14 0.49 & 0.09 logical that a composite column that consists partly of a polar liquid phase and partly of a non-polar liquid phase might enable the acrylic peak to be “placed” in any desired position between isobutyric and butyric acids. Initially, a 2-m x 0.6-cm glass column was packed with the Apiezon L phase for the first 1-2 m of its length. A small plug of glass-wool was introduced and the remainder of the column space was packed with the Carbowax 20M phase. The column was conditioned at 160 “C. This 3: 2 ratio of non-polar to polar phase, however, resulted in only a partial separation of acrylic acid from butyric acid and hence the relative proportion of the Apiezon L phase had to be increased.It could be calculated that the acrylic acid peak could be placed midway between the isobutyric acid and butyric acid peaks when the proportion of Apiezon L phase to Carbowax 20M phase was 4: 1 (see Fig. 1). However, in samples of rumen fluid the butyric acid peak is normally larger than that of isobutyric acid and greater resolution could be achieved when the acrylic acid peak was placed nearer to the isobutyric acid peak. This positioning could be readily achieved by increasing the proportion of the Apiezon L phase to about 85 to 90 per cent. of the total volume of the column. Further studies showed that the quality and resolution of the peaks was considerably improved when the cross- sectional area of the composite column was reduced by using a column of i.d.0.3 cm. When a 2-m x 0.3-cm glass column containing the Apiezon L phase for 85 per cent. of its length and the Carbowax 20M phase for the remaining 15 per cent. of its length was maintained under isothermal temperature conditions at 95 “C, the time taken for the complete analysis of the short-chain fatty acids up to isovaleric acid was about 20 minutes. The reproducibility and sensitivity of the method are illustrated by the results shown in Table I. These results were obtained by carrying out the analysis on samples of rumen fluid to which had been added various amounts of acrylic acid. The range of results obtained TABLE I1 CONCENTRATIONS OF THE SHORT-CHAIN (C, TO C,) FATTY ACIDS IN THE RUMEN O F A SHEEP FOLLOWING THE INFUSION OF 5 g OF ACRYLIC ACID Time/hours Units of f A \ Acid* concentration 0 0.25 0.50 1.0 2.0 4.0 6.0 per cent.65-5 61-8 61.8 63.0 62.1 65.3 67.9 per cent. 19-7 19.6 21.0 23.1 24.5 22.4 20-3 2 : 0 mequiv per 100 ml 6-08 4.61 4-35 6.08 4.53 5.49 6.24 3 : 0 mequiv per 100 ml 1-53 1.46 1.48 1-86 1.79 1.88 1.87 4 : 0 (iso) mequiv per 100 ml - 0.02 - - - - - - - - - per cent. - 0.20 4 : 0 mequiv per 100 ml 1.13 5 : 0 (iso) mequiv per 100 ml 0.01 per cent. 0.1 6 : 0 mequiv per 100 ml 0.02 per cent. 0.2 3 : 1 mequiv per 100 ml - per cent. - per cent. 14.6 0.65 8.7 0.02 0.2 0.03 0.4 0.66 8.8 0.63 8.9 0.01 0.1 0.02 0.2 0.55 7.8 0.73 0.76 0.84 0.95 9.1 10-3 10.0 10.3 0.03 0.03 - - 0.4 0.4 0.03 0.01 0.02 0.03 0.4 0.1 0.2 0-3 0-33 0.18 0.16 0.10 4-1 2.6 1.9 1.1 - - * The first figure represents the number of carbon atoms in a molecule of the acid, and the second figure the number of double bonds.February, 19731 ACRYLIC ACID IN SHORT-CHAIN FATTY ACIDS 125 for each sample, together with the relative ratios of the peak areas of butyric and acrylic acids at the various concentrations, clearly show that the method is capable of permitting the accurate determination of even relatively low concentrations of acrylic acid in the mixture of short-chain fatty acids.At the present time, investigations in this laboratory are concerned with the possible r6le of acrylic acid in the formation and accumulation of propionic acid in the rumen under certain dietary conditions. The present method provides a simple, quick and direct means of tracing the metabolic inter-relationships between the various volatile f atty-acid constituents to be found in the rumen. For instance, after the infusion of 5 g of acrylic acid dissolved in 10 ml of water into the rumen of a sheep, the rapid removal of acrylic acid and the subse- quent changes that occurred in the concentrations of the other fatty acids, in particular propionic acid, could easily be traced (see Table 11). We acknowledge the assistance given by Miss G. Breckenridge, Miss A. S. Wallace and Mr. L. West. REFERENCES 1 . 2. 3. 4. 5. Whanger, P. D., and Matrone. G., Biochim. Biophys. A d a , 1967, 136, 27. Ladd, J . N., and Walker, D. J . , Biochem. J., 1959, 71, 365. Czerkawski, J . W., and Breckenridge, G., Br. J. Nutr., 1972, 27, 147. Satter, L. D., and Esdale, W. J., Afifil. Microbiol., 1968, 16, 680. Cottyn, B. G., and Boucque, C. V., J . Agric. Fd Chem., 1968, 16, 105. Received June 19th, 1972 Accepted October 6th, 1972

ISSN:0003-2654    

DOI:10.1039/AN9739800122

出版商:RSC    

年代:1973

数据来源: RSC

摘要:

126 Analyst, February, 1973, Vol. 98, fifi. 126-132 Automatic Logging and Processing of AutoAnalyzer Peaks with an Off -line, Time-sharing Computer* BY J. D. CAISEY AND B. D. RIORDAN (Toxicology Department, Glaxo Research Ltd., Fulmer, Buckinghamshire) A complete data acquisition and processing system for use with Technicon AutoAnalyzers is described. The system involves the use of a Solartron Data- logger and a teleprinter terminal connected to the Honeywell G266 Time- Sharing Computer Service by a Post Office telephone line. AutoAnalyzer recorders are fitted with re-transmitting slide-wires to produce voltages that are proportional to peak height. Laboratory-built circuits detect the occurrence of peaks and generate digitise-command signals to the logging system. Peak-height voltages are output to a tape punch and a printer.A manual-entry unit enables coding data to be punched at the beginning of each tape. Ten AutoAnalyzer channels can be monitored simultaneously. The channel of origin of each value is identified by a number that precedes each peak voltage. The computer is used to process the raw data generated by the Auto- Analyzers and datalogger. The limited area of core store available to each computer user necessitates the use of several programs, which are called sequentially into the core store, to process the data. The programs were written in the BASIC language by members of the authors’ Department. BEFORE a new drug can be marketed, its manufacturer must satisfy the Committee on Safety of Medicines that the drug has no unacceptable side-effects.The safety of a new drug is investigated by administering it to animals for periods of up to 2 years. At intervals during an experiment, blood and urine samples are obtained from the animals and assayed for various constituents. As each experiment involves a large number of animals, such a workload is most economically completed by automating the various analytical techniques. Most of our biochemical analyses are carried out with Technicon Mark I AutoAnalyzers, of which we have ten channels. These instruments, operating at rates of up to sixty samples per hour, produce large numbers of raw results that normally have to be transformed into useful units by the technical staff. As our interest lies in the interpretation of a statistical analysis of the final results, there is a considerable amount of clerical work associated with our assays.Every manual procedure is liable to be affected by human error, and this is particularly true of the manipulation of numbers. In October, 1967, a computer terminal was installed in the Department, linked to the De La Rue - Bull (now Honeywell G265) time-sharing service. Programs were written to perform the statistical procedures that had hitherto been carried out on desk calculators. At the same time, a simple program was written for converting AutoAnalyzer peak heights into concentrations. This procedure reduced the amount of clerical work required, but the peak heights had still to be read and the data tapes punched manually. To eliminate this slow and error-prone step, we investigated the possibility of modifying our AutoAnalyzers to produce punched tape automatically.An automatic data acquisition system, manufactured by Elliot Medical Automation Ltd., was described by Flynn, Piper and Roberts1 in 1966. However, when we became interested in the problem in 1968, this system was no longer available, having been replaced by a complete system on-line to its own dedicated computer. This and subsequent systems, described by Gray and Owen2 and Griffiths and Carter,3 were designed to meet the needs of hospital laboratories, where “instant” results are required for individual samples. As we were primarily interested in batch-processing groups of results, and already had access to an off-line time-sharing computer, the considerable expense of an on-line system could not For this purpose an on-line computer is virtually essential. * Presented at the Third SAC Conference, Durham, July 12th to lSth, 1971.@ SAC and the authors.CAtSEY AND RIORDAN 127 be justified. All that we required was a system that would measure the peak heights and punch the measured values on to paper tape. After comparing several systems, we selected the Compact-2 Datalogger manufactured by the Solartron Electronic Group Ltd. As normally supplied, this instrument could not be used with several AutoAnalyzers simultaneously, but the Solartron engineers were confident that the necessary modifications were possible. The need for modification of the datalogger was caused, in part, by the limitations of the time-sharing computer system.With an on-line computer, recorders are usually scanned at regular time intervals and the computer calculates the shape and height of each peak from these readings. However, because the time-sharing system reads paper tape at the rate of only ten characters per second and the amount of available computer store is limited, such a multiplicity of readings was unattractive. The system would be feasible only if we limited the amount of data to one reading per peak. Because of this limitation, we had to construct detectors to identify the top of each peak, and the datalogger had to be modified to sample a peak from any channel at any time. PEAK DETECTOR Each recorder is fitted with a re-transmitting slide-wire, the output of which is a voltage that is proportional to peak height. This voltage is applied to two parts of the circuit shown in Fig.1 : firstly, to one input of the differential amplifier A,, and secondly, via the operational amplifier A, with unity gain, to the capacitor C, that is used to store the maximum peak voltage. Because operational amplifiers invert their input signals, the voltage * stored on capacitor C , is applied to a second operational amplifier A, and then to the second input of the differential amplifier A,. The differential amplifier drives the relay RL, that supplies the digitise-command contact-closure to the datalogger. closure Fig. 1. Circuit diagram of peak detector. The components are listed in the Appendix While the recorder pen is moving upscale, the voltages at the two inputs of A, increase in unison.As the pen changes direction, the voltage applied to the direct input starts to fall, but that from A, remains constant. After a short period of time, the amplified difference between the two inputs becomes sufficient to close the digitise-command relay RL,. The storage capacitor C , must be discharged between peaks so that a small peak following a large peak will not be masked. This discharge is effected by a timing circuit, which is activated by a contact-closure generated within the datalogger at the completion of the digitisation sequence. By means of this contact-closure, a charge is placed on capacitor C,..128 CAISEY AND RIORDAN: AUTOMATIC PROCESSING OF [Analyst, Vol. 98 Provided that the charge remains on C,, the output of the differential amplifier A, will hold the relay RL, closed, thus shorting capacitor C3 to earth.Capacitor C , discharges through resistors R,,, R13 and R,, and variable resistor R16. While the storage capacitor is short- circuited to earth, the peak detector circuit is inoperative. By adjusting resistor R,,, the duration of this “dead period” can be set so as to prevent any spurious results being recorded, except those which occur near the top of a peak when the detector circuit is operational. A brief visual inspection of the recorder tracing is sufficient to identify any badly formed peaks that are likely to have caused erroneous results. DATALOGGER A block diagram of the datalogger is shown in Fig. 2. The central part of the datalogger is a digital voltmeter.Voltages measured by the voltmeter are converted into binary-coded decimal form and output to a strip printer and tape punch. The punch has a manual keyboard for typing coding data on to the tape. In order to log from several AutoAnalyzers simul- taneously, two extra modules are added. The first is a peak occurrence store that remembers on which channels peaks have been detected but not logged. The memory consists of ten bistables, one for each channel, which are set by the closure of the digitise-command relay RL, (Fig. 1). The second module is a scanner that examines continuously the bistables in the peak occurrence store at the rate of twenty-five per second. Manual entry occurrence store I Strip printer Fig. 2. Block diagram of datalogger To summarise, the sequence of events is as follows.The recorder pen moves upscale in response to a signal from the colorimeter. The output from the re-transmitting slide-wire is applied to the peak detector circuit. When the pen changes direction, a peak is detected and the digitise-command relay sets the appropriate bistable in the peak occurrence store. When the scanner reaches that channel, it stops scanning and the voltage stored on capacitor C3 is measured by the digital voltmeter. When the punch and printer have output the data, they send signals back to the datalogger to re-start the scanner, re-set the bistable in the peak occurrence store, and energise the storage-capacitor discharge circuit. The output of the printer gives a visual record of the data on the punch tape and is used to monitor the functioning of the datalogger.February, 19731 AUTOANALYZER PEAKS WITH AN OFF-LINE COMPUTER 129 The tape punch generates a seven-character word with the format shown in Fig.3. The word starts with a comma to act as a delimiter between data items in the computer. This is followed by a single digit of channel identity and four digits that correspond to the digital voltmeter reading. The final character is a space. 9 xxxxx Channel Voltage Space I Fig. 3. Tape-punch word format A word counter in the peak occurrence store initiates the automatic punching of the characters for carriage-return and line-feed on the tape after every tenth word. This pro- cedure is necessary because a teleprinter is used to input the data-tape to the computer.The teleprinter is connected to a telephone through a Post Office Modem unit. The com- puter is accessed by switching on the teleprinter, dialling the computer’s telephone number and pressing a data-button on the telephone when the computer responds. Instructions and data can then be transmitted to the computer by typing on the teleprinter, or by running a tape through the tape reading attachment. To assess the performance of the peak detector and datalogger, the linear equation for the regression of datalogger voltage (volts) on peak height (millimetres) was calculated for each channel. The smallest regression coefficient found was 0.065 and the largest 0,068. A Student’s t-value was calculated from each regression coefficient and its sample standard deviation; in every instance, the probability of such a value occurring by chance was less than 0.001.(The constant term of each regression equation was within the range -1.377 to +0.698.) The smallest correlation coefficient between peak height and datalogger voltage was 0.998. These figures show an almost perfect correlation between punched voltage and peak height, and discrimination by the peak detector between very small changes in peak height. A change of 0.01 V, the smallest change in voltage that the datalogger can represent, corre- sponds to a change in peak height of 0.15 mm or 0.054 percentage transmission units. DATA PROCESSING The area of core store available to each user of the time-sharing system is not large enough to permit a datalogger tape to be completely processed in one operation.We have therefore divided the operation into three stages, storing the initial data and intermediate results on disc files between the stages. The programs that control the stages are also stored on disc and are called sequentially into core as required. All programs were written in the BASIC language by members of the authors’ Department. Two information files are also stored on disc. One file contains details of all toxicity experiments that are currently being performed in the Department, listed against identifying experiment numbers. The other file contains the names and units of measurement of the AutoAnalyzer assays, listed against identifying code numbers. The datalogger tape starts with a series of manually punched code numbers that provide the computer with the information necessary to process the data. The first number indicates the number of AutoAnalyzer channels recorded on the tape.This number is followed by separate lists of code numbers for each channel, identifying the experiment and assay, the number of standards and their concentrations, whether drift correction is required, the positions of the drift standards, and whether the samples have been assayed in the correct order. The first program reads the channel codes and peak-height voltages from the main data file and writes them into individual channel files according to their channel identities.130 CAISEY AND RIORDAN: AUTOMATIC PROCESSING OF [Analyst, vol. 98 The names of the files and the number of peaks written into each are printed by the teleprinter, to indicate the presence of any spurious or unlogged peaks.The second program operates on each channel file in turn. It calls a subroutine to copy the data in a channel file into a temporary storage file. The subroutine then reads the numerical coding data from the beginning of the temporary file and, by referencing the two information files, writes the details of the experiment and assay, in alphanumeric form, back into the channel file. Control then returns to the main program. The main program reads the standard and sample peak heights from the temporary file. It applies a linear drift correction to the peak heights and calculates the concentrations of the samples by linear interpolation between the standard peak heights.The concentrations are written into the channel file and printed on the teleprinter, together with the values for the quality control samples. If the peak height of any sample falls outside the range of the standards, the computer prints the serial number of the sample and requests either a replace- ment value for the sample or confirmation that it should be treated as a missing observation. If a replacement value is input, the computer asks for any dilution factor applicable to it. EXPT NO: 3 9 TABLE NO: EFFECT OF REPEATED SUBCUTANEOUS ADMINISTRATION OF N Y / l 0 0 4 FOR 3 MONTHS ON FEMALE CHARLES RIVER CD RATS HAEMOGLOB I N ( G / I 00ML) MEASURED A T 8 WEEKS DOSAGES GPI: I ML VEHICLE /KG/DAY GP2: 8 MG NY/1004 /KG/DAY GP3: 16 MG NY/1004 /KG/DAY GP4t 3 2 MG NY/1004 /KG/DAY (SEE TEXT) GP5: 6 4 MG N Y / I 0 0 4 /KG/DAY (FOR 2 WEEKS ONLY) RESULTS CAGE I 2 3 4 5 6 7 8 9 10 GP I 14.5 16.0 15.9 14.3 16.2 14.3 15.8 15.3 15.0 14.6 GP 2 15.3 15.2 15.1 14.8 13.9 16.3 14.1 15.6 14.9 15.1 GP 3 GP 4 1 GP 5* S 15.2 14.7 14.9 14.7 14.4 13.9 14.1 14.9 15.0 15.3 14.9 15.6 14.8 15.4 15.6 MEAN 15.2 ls.O 14.9 S MISSING OBSERVATION * TREATMENT OMITTED FROM ANALYSIS (CALCULATED RESULT INCLUDED I N MEAN) ANALYSIS OF VARIANCE SOURCE SUM OF SQ D.F.MEAN SQ F P TREATMENT - 5 2 6 4 3 6 2 - 2 6 3 2 1 8 .479 .373 CAGE 1.81451 9 e201612 *367 4 6 5 ERROR 9.33337 1 7 -5 49022 TOTAL 11 - 6 7 4 3 28 S .D. -74 1 D .9 Fig. 4. Typical computer-produced table of results The program is run by typing the names of the channel files that contain the data to be processed.Additional instructions can be entered when required. All, or any, of the concentrations can be multiplied by dilution factors; spurious peaks can be discarded; the serial numbers of missing samples can be input, and the program will write an identifyingFebruary, 19731 AUTOANALYZER PEAKS WITH AN OFF-LINE COMPUTER 131 code into the file in place of each missing observation. By typing a list of the sample serial numbers, in the order in which they were assayed, the program will sort the samples into their correct sequence before writing the concentrations into the file or printing them. Finally, extra standard curves can be introduced against which all subsequent samples will be compared. Various statistical programs have been written to analyse the results contained in the channel files.These programs perform analyses of variance on different experiment designs. They print the results and statistical analyses in the form of fully entitled tables that are suitable for inclusion in our Departmental reports and in submissions to the Committee on Safety of Medicines. A typical table is shown in Fig. 4. A comparison of the data acquisition and processing procedures used in this Department, before and after the introduction of the present system, is given in Fig. 6 . Before the intro- duction of this system, all tables of results had to be prepared by typists and subsequently checked by a member of the scientific staff. Thus, all members of the Department have benefited from the reduction in the amount of clerical work.AutoAnal yzer AutoAnalyzer Chart reader t Desk calculator Tables prepared Stencils cut I Stencils checked I I J Tables duplicated Datalogger I Time-sharing computer system I Tables photocopied Fig. 5. Comparison of manual and automatic data acquisition and processing procedures We now have a system that, apart from pressing the appropriate keys on a teleprinter, requires no human intervention between loading the AutoAnalyzer sample plates and photo- copying the final tables of results. We can now obtain statistically analysed results on the day that the animals are bled, which was impossible before the introduction of the system described in this paper. REFERENCES 1. 2. 3. Flynn, F. V., Piper, K. A., and Roberts, P. K., J. Clin. Path., 1966, 19, 033. Gray, P., and Owen, J . A., Clin. Chim. Ada, 1969, 24, 389. Griffiths, P. D., and Carter, N. W., J. Clzn. Path., 1969, 22, 609. Received January 19th, 1972 Revised Sefitamber 1 lth, 1972 Accepted October llth, 1972132 CAISEY AND RIORDAN Appendix LIST OF COMPONENTS Components were obtained from R. S. Components Ltd. except when stated otherwise. Amplifiers- From Computing Techniques Ltd. A1 A, A, A4 = Operational d.c. amplifier, Type A6-1 = Operational d.c. amplifier, Type A6-1 = Differential amplifier, Type A1-5 = Differential amplifier, Type A1-5 Resistors- R, R, R, R4 R, RB R, R, Capacitors- = 100 kR R, = lMR = 100 kR R,, = 120 kSZ = 50a R,, = 120 kR = 100 kR = 100 kR R,, = 100 kR = 100 kR = 100 kR R,, = 500 kR, variable = 10R R,, = 51 SZ R,, = lMR, pre-set Cl = 8200 pF, silvered mica C , = 8200 pF, silvered mica c, = 2000 pF, 25 V, electrolytic working c4 = 0.001 pF, ceramic c, = 50 pF, 25 V, electrolytic working Diodes- Dl, D,, D,, D,,D,,D, = IB40 D, = REC50A Relays- RL,, RL, = Type 45, 27 V d.c., 700 i2

ISSN:0003-2654    

DOI:10.1039/AN9739800126

出版商:RSC    

年代:1973

数据来源: RSC

摘要:

Analyst, February, 1973, Vol. 98, $9. 133-136 133 Non-aqueous Atomic-absorption Spectrophotometric Analysis of Organonickel Complexes by a Ligand Exchange Method BY M. A. LEONARD AND W. J. SWINDALL (Department of Analytical Chemistry, The Queen's University of Belfast, Belfast, BT9 6AG, Northern Ireland) A procedure is described in which diethylammonium diethyldithiocar- bamate is used to complex nickel in a mixed organic solvent, thus eliminating interferences due to different bonding. The interferences can also be removed by using a nitrous oxide - acetylene flame and a method has been devised that requires the use of only a single solvent for dissolution. The ligand exchange method should have wide application in the elimination of bonding interferences. THE inorganic research section of this Department produces organonickel compounds typified by 'those shown in Fig.1. A rapid method was therefore sought for the determination of nickel in 5-mg samples. Direct dissolution of the compounds in a suitable non-aqueous solvent followed by atomic-absorption spectrophotometry seemed appr~priate,l-~ but it was found that the standard (nickel cyclohexylbutyrate) and different samples yielded different absorbances for equal nickel concentrations. This phenomenon had been noted previously,6#6 and was due to the variety of environments in which the nickel atom was situated. Fig. 1. Examples of organonickel cornpouncls An apparently obvious solution to this difficulty was chelation of the nickel with a strongly complexing reagent that is compatible with organic solvents.The use of diethyl- ammonium diethyldithiocarbamate in conjunction with an air - acetylene flame proved to be successful. Good results were also obtained by using a nitrous oxide - acetylene flame and no ligand exchange, but in this method the advantage of having stable sample solutions after the addition of the reagent was lost. EXPERIMENTAL APPARATUS- A Unicam SP90 atomic-absorption spectrophotometer with an SP94 nitrous oxide accessory, a Unicam high-intensity nickel hollow-cathode lamp and a recorder with a full-scale deflection of 10mV were used. 0 SAC and the authors.134 LEONARD AND SWINDALL : ATOMIC-ABSORPTION SPEGTROPHQTOMETRIC [AWZt!ySt, VOl. 98 The operating conditions used were as follows: burner height, 1.1 cm; slit width, 0.075 mm; wavelength, 341.5 nm; current, 15 mA; flames, (a) acetylene 1-0 1 min-l - air 5.0 1 min-l and (b) acetylene 3.0 1 min-l - nitrous oxide 5.0 1 min-1. REAGENTS- Nickel cyclohexylbutyrate standard-BDH Chemicals Ltd.The sample was of ques- tionable purity and the nickel content was checked by wet oxidation followed by aqueous atomic-absorption spectrophotometry. ChZoroform-Analytical-reagent grade , Hopkin and Williams. Isobutyl methyl ketone-Reagent for atomic-absorption spectrophotometry, Hopkin and Williams. Diethylammonium diethyldithiocarbamate-Hopkin and Williams. PROCEDURE- Calibration graph-Dissolve 33.83 mg of pure nickel cyclohexylbutyrate standard in 30 ml of chloroform containing 113 mg (6 molar proportions) of diethylammonium diethyl- dithiocarbamate. Add approximately 65 ml of isobutyl methyl ketone, cool the mixture to room temperature, then dilute it accurately to 100.0 ml with isobutyl methyl ketone. The solution contains 50 p.p.m.of nickel. Dilute this solution with 3 + 7 chloroform - isobutyl methyl ketone so as to give solutions that contain 5 to 26 p.p.m. of nickel, then spray these solutions into the flame under the given conditions in order to produce a cali- bration graph. Sam$e-Dissolve a weighed amount of sample in isobutyl methyl ketone so as to give a solution that contains approximately 0.5 mg of nickel. Add 15 ml of chloroform containing a three-fold molar excess of diethylammonium diethyldithiocarbamate, then dilute the solution to 50.0 ml with isobutyl methyl ketone but adjust to volume after cooling the solution to room temperature.Spray the solution into the flame under the conditions used when obtaining the calibration graph. RESULTS AND DISCUSSION Good agreement was obtained between the experimental and calculated results for the nickel content of the samples, as shown in Table I. The samples analysed also gave satis- factory results for the determination of other constituents. Precision was assessed by the analyiis of five samples of the compound Ni[P(C,H,),],Br, by different methods, as shown in Table 11. TABLE I COMPARISON OF RESULTS OBTAINED VY DIFFERENT METHODS Nickel found, per cent. Organic solution + diethyl- ammonium diethyldithio- carbamate, air - acetylene flame 10.0 13.0 13.0 10.9 16-1 8-0 Organic solution, nitrous oxide - acetylene flame _.8.0 13.6 The compounds to be analysed dissolved in isobutyl methyl ketone, but nickel cyclo- hexylbutyrate, which dissolves in chloroform, would not dissolve in the ketone. A mixed solvent was therefore used and it was necessary to dissolve the standard and samples in their respective solvents followed by addition of the other solvent. When the solvents were mixed, heat was produced, so the flasks were cooled before making the volume up to the mark.February, 19731 ANALYSIS OF ORGANONICKEL COMPLEXES TABLE I1 STATISTICAL RESULTS ON THE DETERMINATION OF NICKEL IN ---- --=-------- Ni[P(C,H,) 31 2Br2 Theoretical nickel content = 7.90 per cent. Nickel found by photometric titration, per cent. 7-47 7-89 7-87 7.69 7.78 Standard deviation (S) .. 0.17 Mean (3) . . .. . . 7-74 .. S GT loo - - . . 2.2 Range at 96 per cent. confidence .. . . 0.21 Nickel found by using diethylammonium diethyldithiocarbamate and air - acetylene flame, per cent. 7-82 7-72 7.94 7.76 7.82 7.81 0.083 1.1 0.10 1HE LUMYUUNU 136 Nickel found by using nitrous oxide - acetylene flame, per cent. 7-94 7.73 7-68 7.80 7.78 7.79 0-098 1.3 0.12 When absorbance was plotted against burner height for different nickel compounds in chloroform - isobutyl methyl ketone solution, maxima were obtained that occurred at different burner heights for different compounds (Fig. 2). Sastri, Chakrabarti and Willis' proposed that the different absorbances were due to the different types of bonding of the nickel to the ligand, and that when metal - oxygen bonds were involved, metal oxides would be more readily formed in the flame.These oxides were more difficult to dissociate to the free metal atoms that are necessary for absorption to occur. This effect would give rise to the observed difference between the burner height for maximum absorption for nickel cyclohexylbutyrate, which contains metal-oxygen bonds, and that for the samples, which do not contain such bonds. a, C m 2 2 n 6 " 0,031 I I I I I I I I 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.5 1.6 Burner heightkm Fig. 2. Effect of burner height on absor- bance for different organonickel compounds in chloroform - isobutyl methyl ketone (3 + 7) solution : I , nickel cyclohexylbutyrate containing 10 p! ml-l of nickel: 11, Ni[P(C,H,),C,H6]2C12 containing 9.6 pg ml-l of nickel; 111, NiP(C,H,,),C,H,Cl containing 7.0 pg ml-l of nickel ; and IV, NiP(C,H,),C,H,GeCl, containing 7.4 pg ml-l of nickel 0.051 I I I 1 0.8 '0.9 1.0 1.1 1.2 Burner height/cm Fig.3. Effect of burner height on absorbance for different organonickel compounds in chloroform - isobutyl methyl ketone (3 + 7) solution with diethylammoniurn diethyldithiocar- bamate added : I , nickel cyclohexyl- butyrate containing 16pgml-l of nickel; 11, Ni[py(CH,),S(CH,),py]Br, contain- ing 10.6 p g ml-' of nickel; and 111, NiP(C,H,,),C,H,Cl containing 10.4 pg ml-1 of nickel136 LEONARD AND SWINDALL A solution to the problem seemed to be to obtain the standard and the samples in the same form, when the ease of dissociation of the metal-ligand bond would be irrelevant.Ligand exchange with diethylammonium diethyldithiocarbamate was tried and it was observed visually that this compound complexed the nickel from the standard and all the samples. The nickel - diethyldithiocarbamate complex was stable and soluble, which was an advantage as some sample solutions underwent precipitation on standing. Graphs of absorbance against burner height for complexed standard and samples showed maximum absorbances at the same burner height (Fig. 3). This ligand exchange process should have numerous applications; in fact, just after this work was carried out a paper was published by Kashiki, Yamazoe and Oshima,* in which the use of iodine in the determination of lead, present as a variety of alkyl derivatives, in petrol was described. The addition of iodine yielded identical absorbances for all compounds.This work was later extended9 to the determination of calcium, barium, zinc and copper in petrol. Another approach to the problems posed by the organonickel compounds is through the use of the nitrous oxide - acetylene flame where the higher temperature minimises the effect of different bonding. Sample and standard solutions were made up in a 30 per cent. V/V solution of chloroform in isobutyl methyl ketone only and gave the results shown in Table I. The need for a mixed solvent can be avoided by dissolving the nickel cyclohexylbutyrate in isobutyl methyl ketone containing diethylammonium diethyldithiocarbamate, which com- plexes the nickel and readily gives a solution. This procedure also avoids the need to cool the solution so as to remove the heat produced by mixing the solvents. The standard deviation for five samples analysed in this way was found to be 0.091 per cent.of nickel. This simpli- fication was appreciated, however, only when the work had been virtually completed. CONCLUSION We found that both the ligand exchange and the nitrous oxide - acetylene flame methods gave correct results. The ligand exchange technique is preferred in this application, however, owing to the stability of the nickel - diethyldithiocarbamate complex, which overcomes the difficulty of handling unstable solutions of the samples. REFERENCES 1. 2. 3. 4. 6. 6. 7. 8. 9. Allan, J. E., Nature, Lond., 1960, 187, 1110. Gatehouse, B. M., and Walsh, A.. Spectrochim. Acta, 1960, 16, 602. Robinson, J. W., Analytica Chim. Acta, 1960, 23, 479. Lockyer, R., Scott, J . E., and Slade, S., Nature, Lond., 1961, 189, 830. Chakrabarti, C. L., and Singhal, S. P., Spectrochim. Acta, 1969, 24B, 663. Takeuchi, T., Suzuki, M., and Yanagisawa, M., Analytica Chim. Ada, 1966, 36, 268. Sastri, V. S., Chakrabarti, C. L., and Willis, D. E., Can. J . Chem., 1969, 47, 687. Kashiki, M., Yamazoe, S., and Oshima, S., Analytica Chim. Acta, 1971, 53, 96. -,-,- , Ibid., 1971, 54, 633. Received May 8th, 1972 Accepted October 4th, 1972

ISSN:0003-2654    

DOI:10.1039/AN9739800133

出版商:RSC    

年代:1973

数据来源: RSC

摘要:

Analyst, February, 1973, Vol. 98, pp. 137-145 137 A Comparison of a Spectrophotometric (Quercetin) Method and an Atomic-absorption Method for the Determination of Tin in Food BY &E ENGBERG (Department of Pesticides, Food Additives and Contaminants, National Food Institute, Copenhagen, Denmark) Procedures for the determination of tin in food, which involve a spectro- photometric method (with the quercetin - tin complex) and an atomic- absorption method, are described. The precision of the complete methods and of the individual analytical steps required is evaluated, and the para- meters that influence the precision are discussed. It is concluded that while the spectrophotometric method is to be preferred for very low tin concentrations, for instance, residues of organotin compounds, the two methods are equally useful for the determination of tin in concentrations normally found in canned foods.With both methods recoveries of added tin do not deviate significantly from 100 per cent. within their respective working ranges. THE concentration ranges of tin in the two main fields of interest in the determination of tin in food are widely different: 10 to 1000 p.p.m. in canned food and about 0.01 to 1 p.p.m. for organotin residues. Consequently, different methods are often used for the determinations in the two ranges. Recently, the Analytical Methods Committee of the Society for Analytical Chemistry1 has recommended a spectrophotometric method based on the reaction of tin( IV) with catechol violet for the determination of tin in the range “up to 30 pg”; for amounts between 30 and 150 pg, the dithiol method is under consideration and for those above 150 pg, a method that involves titration of tin(I1) with iodine is expected to be suitable.The aim of the present work was to demonstrate the wide working range of the reagent quercetin (3,5,7,3’,4’-pentahydroxyflavone), and to compare the spectrophotometric quer- cetin method with the atomic-absorption determination of tin in the hydrogen - air flame. PRELIMINARY STUDY- Agreement with Beer’s law indicates that a single tin - quercetin complex with a high stability constant is formed at least in the range 1 to 400pg (in 50ml of final solution), the absorbance being the upper limiting factor. When tin - quercetin solutions containing larger amounts of tin (at least up to 2000pg of tin per 50ml) are diluted with quercetin solution, so that their concentrations correspond to the directly measurable range (1 to 400 pg per 50 ml), there is also agreement with Beer’s law.This result indicates that the next limiting factor in food analysis will not be the formation of the quercetin - tin complex, but rather the formation of insoluble compounds during wet ashing, as found by Kirk and Pocklington.2 The absorptivity, as calculated from the slope of the standard curve obtained with a Beckman DB-G spectrophotometer, is 20 600 1 mol-l cm-l, which is about one third of the value for the two well known tin reagents, catechol violet and phenylfluorone, and about three times the value for dithiol. Accordingly, when using 1-cm cells, the optimum range of 0.1 to 1.1 absorbance units with minimum relative error in the spectrophotometric measure- ment corresponds to a tin concentration of 0.6 to 6 pg ml-l, or 30 to 300 pg in 50 ml of final solution.This range is convenient for canned food samples. For the low concentration ranges, which mostly apply to organotin residues, the use of long light paths or scale expan- sion, or both, is justified because of the linearity of the calibration graph down to at least 0.02 pg ml-l concentration or 1 pg in 50 ml of final solution. To attain the lowest possible detection limit in the atomic-absorption method, the hydrogen - air flame should be used. With a Perkin-Elmer 403 instrument equipped with a three-slot burner head, this flame gave a linear response at least in the range 0.2 to 100 pg ml-l of tin, corresponding to 4 to 2000 pg in 20 ml of final solution.With an acetylene - air flame, the linear response range was about 2 to 360 pg ml-1 or 40 to 8000 pg in 20 ml of @ SAC and the author.138 ENGBERG : COMPARISON OF SPECTROPHOTOMETRIC METHODS [AVZcZt?$St, VOl. 98 final solution. Consequently, while the range for canned food is well covered by the atomic- absorption measurement, the spectrophotometric quercetin method seems to be more suitable for the very low concentrations. PRE-TREATMENT OF SAMPLES- For the pre-treatment of food samples, wet ashing with sulphuric acid plus nitric acid or hydrogen peroxide is often recommended. With much canned food, fruit juices and other easily digested samples wet oxidation with sulphuric acid - hydrogen peroxide is usually satisfactory, according to the recommendations of the Analytical Methods Committee.1 With organotin determinations, however, Corbin3 points out that the presence of nitric acid in an amount comparable with the water content of the sample is important, especially in the initial stages of the ashing, so as to prevent losses of organotin.According to our preliminary experiments nitric acid should be used as the oxidising agent whenever certain natural or artificial colours, especially carotenoids or tartrazine, are present so as to ensure more complete oxidation. The maximum absorption of carotenoids, for instance, in carrots, as well as that of tartrazine, is very close to the maximum of the quercetin - tin complex and their complete destruction should be made a routine procedure, as described in the method below.ISOLATION OF TIN- In addition to the carotenoids and tartrazine, many different ions may interfere with the formation or measurement, or both, of the quercetin - tin complex. According to Koch and Koch-Dedic,* the interfering substances include chromium(V1) [and, according to our experience, chromium(III)], iron(III), gallium, germanium, hafnium, mercury(I), molyb- denum(VI), niobium, platinum, scandium, antimony(III), tantalum, thorium, titanium, vanadium(V), tungsten(VI), zirconium, fluoride, orthophosphate and oxalate. Of these, appreciable amounts of iron may be present in canned food, but can be effectively masked with thiourea. The presence of chromium may also be expected, bearing in mind the practice of treating cans with compounds of this metal.However, the spectral interference of chro- mium is not eliminated by the use of thiourea, which is also true for molybdenum and titanium, both of which may be present if the food samples are contaminated with soil particles. During the preliminary experiments this latter interference became a serious problem in the examination of fresh carrots for residues of organotin fungicides. With measurements with the hydrogen - air flame in atomic-absorption spectrophotometry, poten- tially interfering anions and cations are also numerous (Juliano and Harrison5 and Capacho- Delgado and Mannings). Harrison and Juliano’ described the interference from organic solvents.Although the interferents are fewer with the acetylene - air flame, Pearlman, Hefferen and Lyons mentioned calcium and phosphate, and Capacho-Delgado and Manning6 mentioned nitrate. The separation of tin(1V) by its extraction as tin(1V) iodide into toluene, which has been proposed by Newman and Jones,lo provides an effective solution to most of these interference problems. The only known exceptions are the interferences mentioned from carotenoid and tartrazine in the quercetin method and germanium and antimony( 111), which are partially co-extracted with the tin(1V) iodide. The co-extraction of antimony can be hindered by using an iodide concentration higher than 1.5 M in the acidic phase. Germanium is rarely expected to be present in food samples.FORMATION OF THE TIN - QUERCETIN COMPLEX- For the determination of tin with quercetin, the influence of acidity after digestion has been especially studied. Malkusll proposed dilution of the sulphuric acid digest with 10 per cent. hydrochloric acid. A 1-ml aliquot is partially neutralised with “6 drops of 10 per cent. sodium hydroxide,” after which thiourea is added as masking agent, followed by the ethanolic quercetin reagent. The absorption of the quercetin complex formed is measured at 437 nm. Karv&nek, Curda and Miler12 demonstrated that the complex formation is less dependent on the acidity when sulphuric acid is used than with hydrochloric acid. They add concentrated sulphuric acid to the digest to give a specific volume, dilute the acidic solution with water and add thiourea and quercetin solution.Kirk and Pocklington2 recommend partial neutral- isation of the sulphuric acid with sodium carbonate, pointing out that the use of sodium Clauss, Laugel and Hasselmanne noted the interference from iron.February, 19731 FOR THE DETERMINATION OF TIN IN FOOD 139 hydroxide for this purpose may lead to erratic results, possibly resulting from localised hydrolysis. These authors further mention that neutralised solutions that are allowed to stand for 20 minutes before being re-acidified give significantly lower results. In the method described below, the principle adopted by Karvhek, Curda and Miler12 is followed, measurement in sulphuric acid solution being combined with the extraction of tin into toluene as tin(1V) iodide in order to improve the sensitivity and specificity of the method. METHODS PRINCIPLE OF THE METHODS- After wet ashing the sample with sulphuric acid, tin is extracted as tin(1V) iodide into toluene and back-extracted as stannate with aqueous sodium hydroxide.If quercetin is used for the quantitative measurement, the solution is re-acidified with sulphuric acid and any iron that remains is masked with thiourea before complex formation. If the final measurement is by atomic absorption with a hydrogen - air flame, potassium hydroxide is used to form the stannate and the solution is re-acidified with hydrochloric acid in order to gain an absorption enhancement caused by potassium and to avoid suppression caused by sulphate. REAGENTS- Re-distilled water and analytical-reagent grade reagents were used throughout.Sulphuric acid, sp. gr. 1-84. Sulphuric acid, approximately 9 N-Cautiously mix 250 ml of sulphuric acid, sp. gr. 1.84, Nitric acid, sp. gr. 1.42. Ammonium oxalate, saturated aqueous solution. Hydrogen peroxide, 30 or 50 per cent. m/V. Potassium iodide solution, approximately 5 M-Dissolve 83 g of potassium iodide in water Toluene. Potassium or sodium hydroxide solution, 5 and 0.1 N. Hydrochloric acid, 6 N. Ascorbic acid solution, 5 per cent. m/V-Prepare freshly each week. Thiourea solution, 10 per cent. m/V. Quercetin solution, 0.2 per cent. m/V in 96 per cent. ethanol. Ethanol, 96 per cent., distilled. Tin(1V) stock solution-Dissolve 0.0951 g of tin(I1) chloride (SnC1,.2H20) in 20 ml of sulphuric acid, sp. gr.1.84, filus 30 ml of 30 per cent. hydrogen peroxide. Evaporate the solution to fumes and allow it to stand for 10 minutes. Add 43 ml of concentrated sulphuric acid and then, cautiously, 50 ml of water (use a Kjeldahl digestion flask for this procedure). Pour the solution into a 250-ml calibrated flask containing about 50 ml of water. Rinse the Kjeldahl flask and dilute the solution to volume with water. This solution remains unchanged for several months when stored in a polythene bottle. with 500 ml of water, cool to room temperature and dilute to 1 litre with water. and dilute to 100 ml. Prepare freshly each day. 1 ml of solution = 200 pg of tin, Tin(1V) dilute standard solutions-Prepare fresh standard working solutions each day by diluting the stock solution with the approximately 9 N sulphuric acid to give solutions containing 20 and 2 pg ml-1 of tin. Calibrated flasks, auto-sampler glasses and other glassware, which are used after the extraction step, should stand overnight in 6 M hydrochloric acid before being rinsed with re-distilled water.Rinse all other glassware with 6 M hydrochloric acid and then re-distilled water. SAMPLE ASHING- For the determination of tin in canned foods such as baby-food, fruit juices and other easily digestible samples, sulphuric acid and hydrogen peroxide are used according to the recommendations of the Analytical Methods C0mmittee.l For samples containing natural or artificial colours such as carotenoids and tartrazine, nitric acid is preferred as an oxidising140 ENGBERG : COMPARISON OF SPECTROPHOTOMETRIC METHODS [AndySt, VOl.98 agent together with sulphuric acid, and for complete digestion of, for example, fresh carrots, ashing is continued until the digests cease to darken during evolution of fumes of sulphur trioxide for 15 minutes. As previously mentioned, the use of nitric acid is also preferred when digestion of organotin compounds is required. PROCEDURE FOR THE SPECTROPHOTOMETRIC QUERCETIN METHOD- After wet ashing, the sulphuric acid digest is poured into a measuring cylinder and concentrated sulphuric acid is added to give a volume corresponding to one quarter of that of a suitable calibrated flask. The acid is then transferred to the flask and the ashing flask and the measuring cylinder are rinsed with water, which is added to the calibrated flask.(The following pfocedures should be carried out rapidly and with as few interruptions as possible.) After cooling and filling the flask to the mark with water, an aliquot, containing 0 to 400pg of tin, is transferred by pipette into a separating funnel (at least up to 2000pg is permitted, provided that it is diluted later, as described below) and a 5 M solution of potassium iodide is added. If antimony is present, the concentration of iodide in the mixture should be 1 6 ~ , so as to prevent co-extraction of the antimony. If antimony is absent, 2 6 m l of 5~ potassium iodide are added to 25 ml of the diluted digest solution, the solution is mixed and 10ml of toluene are added. The mixture is shaken vigorously for 2 minutes and the phases are allowed to separate, the aqueous phase being discarded.The toluene layer is washed, without shaking, with 5ml of a mixture containing the same proportions of potassium iodide solution and 9 N sulphuric acid as the diluted digest (1 volume of the potassium iodide solution to 10 volumes of the acid). The aqueous phase is again discarded and the toluene layer coloured pink with extracted iodine. The washing is repeated without shaking, the inner surface of the separating funnel rinsed and the wash solution discarded; 5.0 ml of water and 060 ml of 5 M sodium hydroxide solution are added and the funnel is shaken for 30 s. The toluene layer should now be colourless. If not, 5 M sodium hydroxide solution is added from a graduated pipette until the pink colour disappears, 2 drops being added in excess.The total volume of sodium hydroxide solution added is noted and the funnel shaken for 30 s after the last addition. The sodium hydroxide phase is run off quantitatively into a 20-ml calibrated flask and the remaining toluene discarded; 5.0 ml of 9 N sulphuric acid are added to the flask and the solution is mixed. If more than 0.5 ml of sodium hydroxide solution is used for the extrac- tion, an equal volume of 9 N sulphuric acid is added in addition to the 6.0 ml. The re-acidified solution is coloured yellow by the iodine, which is reduced to iodide by the addition of 0.5 or 1 ml of 5 per cent. ascorbic acid solution. A 10-ml graduated pipette is filled with thiourea solution and partially emptied into the calibrated flask, filling the latter to the mark.The remainder is divided equally between two 25-ml calibrated flasks. After mixing the contents of the 20-ml flask, 10 ml of this solution are transferred by pipette into one of the 25-ml flasks, while the remaining 10 ml are rinsed quantitatively with ethanol into the second 25-ml flask, which is filled to the mark with ethanol. This latter solution serves during the spectrophotometric measurement as a control on residual natural or added colours that may be present in the sample. To the first 25-ml flask, 2.50ml of quercetin reagent are added and the flask is filled to the mark with ethanol. The solution is allowed to stand for half an hour, its temperature being thermostatically controlled to -&1 "C if a coefficient of variation lower than 1 per cent.is required. MEASUREMENT AND CALCULATION- The absorbance is measured at 437 nm, by using 4-cm cells if the amount of tin in the weighed sample is below 20 pg and 1-cm cells for amounts between 20 and 400pg. All measurements should be made against references of reagent blanks that are carried through the whole procedure, including the extraction steps. If a fixed reference solution is used for all measurements, an unextracted blank should be used as a reference for extracted blanks and samples. The unextracted blank solution is prepared by mixing 320 ml of water with 20ml of sulphuric acid (sp. gr. 1-84>, cooling and adding 200 ml of thiourea solution and 100 ml of quercetin solution. The mixture is then diluted to 1 litre with 96 per cent.ethanol.February, 19731 FOR THE DETERMINATION OF TIN IN FOOD 141 Thermostatic control of the cuvettes is recommended if a precision better than &l per cent. is desired. The measured net absorbance of the quercetin-free solution is subtracted from the net absorbance of the quercetin complex solution so as to correct for colour not originating from quercetin or from the complex. The amount of tin in the mass of sample taken is determined from a calibration graph, which is linear at least within the range considered here, i.e., up to about 400 pg. As the day-to-day variation is considerable compared with other contributions to the total variation, it is advisable in each series to check a few points of the calibration graph simultaneously with the sample measurement, by using the same volumes of reagents for samples and standards. PROCEDURE FOR THE ATOMIC-ABSORPTION METHOD- The procedure is as described above, except that potassium hydroxide is used instead of sodium hydroxide and that the stannate solution is re-acidified with 6 M hydrochloric acid instead of with 9 N sulphuric acid.After the re-acidification with 5 ml of the hydrochloric acid, the iodine colour is removed with 0.5 or 1 ml of ascorbic acid solution, and the calibrated flask is filled to the mark (20 ml) with water. It is unnecessary to prepare a colour control or to remove interference from iron with thiourea in this procedure, and the 20 ml of solution are used directly for the atomic-absorption measurement. It is advisable to allow the hydrogen - air flame to burn for 20 to 30 minutes so as to become stabilised before the measurement is made.In the present investigations use was made of a Perkin-Elmer 403 instrument, with which the deuterium background corrector proved useful, especially near the detection limit. The bracketing technique is used and blank measurements are made after each series of standard-sample-standard. A concentration of about 1.6 pg ml-1 of tin in the final solution corresponds to 1 per cent. absorption. RESULTS AND DISCUSSION FACTORS THAT INFLUENCE THE PRECISION OF THE SPECTROPHOTOMETRIC METHOD- The contributions of the individual analytical steps to the final precision of the method are shown in Table I, and are based on measurements, in 1-cm cells, in the range 0 to 60 pg per 50 ml of final solution, in which the absolute standard deviation is virtually independent of the concentration of tin, so that the standard deviations for single measurements, SA, are directly comparable.The degrees of freedom given are those used for the determination of the sA values. For the actual spectrophotometric measurement (step A), the error in TABLE I RELATIONSHIP BETWEEN ESTIMATED STANDARD DEVIATION , sA (FOR SINGLE MEASUREMENTS) , Concentration range 0 to 60 pg of tin per 50 ml of final solution AND ANALYTICAL STEPS INVOLVED Analytical steps Estimated standard deviation, SA P Degrees of Absorbance units pg per 60 ml freedom Not including day-to-day vaviation- A .. .. .. .. .. . . .. . . <0.0004 <0.1 6 A + B .. .. .. .. .. . * 0.0005 0.15 6 A + B + C + D + E (standards and small samples; A + B + C + D + E (100-g samples; extraction A + B + C * .. .. .. . . .. .. 0.00 15 0.5 6 extraction volume ratio 25 : 10) . . .. .. 0.0012 0.4 11 volume ratio 1OO:lO) . . .. .. .. . . 0.004 1.6 12 Including day-to-day vayiatzon- A + B +. C + D + E (standards and small samples; extraction volume ratio 25 :lo) . . .. .. 0.014 6 8 A = Spectrophotometric measurement. B = Storage of final solutions in calibrated flasks for 1 hour before spectrophotometric measurement. C = Preparation of final solution (not including extraction step). D = Extraction and back-extraction. E = Wet ashing.142 ENGBERG COMPARISON OF SPECTROPHOTOMETRIC METHODS [Andyst, Vol. 98 reading the instrument is the limiting factor. The standard deviation is low, and barely contributes to the error in the complete method, pro,vided that the glassware is thoroughly cleaned as described earlier.A certain increase in the standard deviation results from allowing the solutions to stand in the calibration flasks (step B). More marked, however, is the increase that derives from the addition of the complexing reagent and the mixing of the final solutions (step C), which probably reflects the variation in absorbance from the quercetin reagent itself. Even if the maximum absorption of unreacted quercetin occurs a t 409nm, the contribution to the absorption at 437 nm is still significant. Measured in l-cm cells against water this blank absorption amounts to 0.09 absorbance unit at 437 nm, which corresponds to a net absorbance of 30 pg of tin as the quercetin - tin complex.Accordingly, an error of 1 per cent. in the measurement of the quercetin reagent by pipette yields a contribution to the total error of 0.3 pg per 50 ml of final solution, thus partially explaining the increase noted from step C, Table I. TABLE I1 STANDARD CURVES WITH AND WITHOUT PRIOR EXTRACTION (1 TO 12 pg PER 50 ml) x = pg of Sn per 50 ml . . 1.37 2.74 4.11 6-48 6.86 8.22 9.69 10.96 y1 = absorbance 0.0223 0.0550 0-0655 0.0916 0.1107 0.1397 0.1580 0.1806 = slope . . 1.62 2.00 1.59 1-67 1.61 1.69 1-86 1.64 Without ya = absorbance 0.0176 0.0353 0-0512 - 0.0907 0.1088 0.1277 0.1375 = slope . . 1-35 1-29 1.24 - 1.32 1.32 1.34 1.25 With It is surprising to note that the extraction and ashing procedures (steps D and E) do not contribute measurably to the standard deviation.As for the extraction step alone, this was verified by further experiments in which 4-cm cells were used for the measurements. The results obtained from these experiments are given in Table 11, in which are compared the results for absorbance and slopes of the curves for standard tin solutions with and without prior extraction (in the latter instance, the procedure according to Karvknek et aZ.12 was used). Based on these results the following curves have been calculated: Without extraction: y1 = 0.003 + 0.0162~ With extraction: y2 = 0.000 + 0.0130~ sy.x = 0.0042 absorbance unit sy.x = 0.0033 absorbance unit In both instances, the standard deviation of a single measurement corresponds to about 0-25pg of tin (in 50ml of final solution), i.e., the standard deviation is unaffected by the extraction procedure as such, within the range tested here (0 to 10 pg of tin per 50 ml).A difference, however, is noted between the two series of results in Table 11. When extraction is included in the procedure the slope of the standard curve is about 20 per cent. lower than with complex formation and measurement alone. As the extraction involves three sub-steps, namely extraction of tin(1V) iodide, back-extraction of stannate and re- acidification, it is of interest to note that a decrease in slope of the same magnitude can be obtained without extraction, but by adding directly to the tin(1V) solution the same excess amount of sodium hydroxide as was used in the extraction of tin(1V) iodide from toluene, followed by acidification with sulphuric acid, which indicates that the loss in absorbance and decrease in slope is caused by the acid-alkaline-acid changes rather than an incomplete extraction process.Also, from Table I, it appears that there is a considerable day-to-day variation, as measured by a marked increase in the standard deviation when comparable results obtained by the complete analytical procedure during consecutive days are evaluated. This variation occurs when the extraction is included, whereas it is found that if the final solution is made up according to KarvAnek et aZ.12 without prior extraction, the day-to-day variation is reduced, which indicates that the main difficulty in the quantitative determination of tin is the tendency towards hydrolysis or polymerisation, or both.These effects become apparentFebruary, 19731 FOR THE DETERMINATION OF TIN IN FOOD 143 when tin(1V) solutions that are approximately neutral, or in which no complexing agent such as quercetin is present, are left for more than a few minutes. During the extraction procedure, the solutions are adjusted to a pH beyond neutral twice, which treatment, as shown above, decreases the apparent concentration but not necessarily the standard deviation. While the considerable day-to-day variation is thus explained, the need is emphasised for the extraction procedure to be carried out rapidly and with as few interruptions as possible. For measurements of more than 60 pg per 50 ml of final solution, the absolute error ceases to be practically independent of the concentration and the error tends to become proportional to the concentration.One of the reasons for this effect is the influence of the temperature on the spectrophotometric measurement. Investigations into this temperature dependence showed that while the absorbance of the non-extracted blank solution was prac- tically independent of the temperature in the range 25 to 40 "C, the absorbance of the tin - quercetin complex decreased relatively about 1.1 per cent. per 1 "C temperature rise. The density change, amounting to 0.07 per cent. per 1 "C, was responsible for only a small fraction of this decrease. While the temperature dependence of the tin - quercetin complex solutions was found to be reversible, a small irreversible increase in absorbance was noted when the extracted blank solutions were kept for several minutes in the light beam of the spectrophotometer, This increase (about 1 per cent.relative absorbance change during 20 minutes, measured against water) could be seen, to whatever level the temperature was increased or decreased, and is probably due to spurious liberation of iodine caused by the light. Consequently, the use of unextracted blanks is recommended if fixed references are used, and thermostatic control of the cuvettes during the spectrophotometric measurement is necessary when precisions lower than 1 per cent. relative are desired. WORKING RANGE OF THE SPECTROPHOTOMETRIC METHOD- An estimate of the detection limit of the method can be based on the above observations.From the results given in Table 11, it can be seen that Beer's law is obeyed at least down to 1 pg of tin (per 50 ml of final solution), and from Table I that the standard deviation for small sample sizes (up to 1 g, or perhaps more, and containing less than 60 pg of tin) is about 0.4 pg of tin per 50 ml of final solution. If 4-cm cells are used the corresponding figure is 0.25 pg of tin. Consequently, the smallest absolute amount of tin that can be detected in small samples with a reasonably low number of replicate determinations is about 1 pg. For larger samples such as, for instance , those for organotin determinations in vegetable crops, an increased absolute standard deviation is obtained, which, however, is not proportional to the sample size.When 80-g samples of carrots are analysed, the standard deviation is 1.3 pg of tin, corresponding to 0.015 pg g l . In an investigation of the content of organotin residues in carrots, an average content of 4 pg or 0.05 pg ,q1 (corresponding to three times the standard deviation of the whole procedure) was therefore determined with a single-sided significance at the 0.5 per cent. level (with six-fold determination of differences between treated and untreated samples). The working range, i.e., the range in which no significant deviation from linearity of the standard curve is found, is about 1 to 400 pg (per 50 ml of final solution). PRECISION AND ESTIMATED WORKING RANGE OF THE ATOMIC-ABSORPTION METHOD- As the pre-treatment and extraction procedures are similar for the atomic-absorption and the spectrophotometric quercetin method, it is necessary only to consider the precision of the actual atomic-absorption measurement. Here, the sensitivity was greater after prior extraction (&%A = 1.4 pg ml-l) than without prior extraction (S,%+ = 2-0 pg ml-l) and, further, the use of the deuterium background corrector proved to be important, especially in the low concentration range (below 1 pg ml-l), when it was found experimentally to give a four to five-fold lowering of the standard deviation of a single measurement.For measurements of standard solutions at 224.6 nm this standard deviation was found to be 0.08 pg ml-l, corresponding to 1-6 pg per 20 ml of final solution. The corresponding standard deviation for the quercetin method was 0.4 pg per 50 ml of final solution, i.e., by analysis of standard solutions, the detection limit of this method is about four times lower than that of the atomic-absorption spectrophotometric method.With higher concentrations144 ENGBERG : COMPARISON OF SPECTROPHOTOMETRIC METHODS [Analyst, Vol. 98 of tin, however, the precisions of the two methods become more closely comparable or even reversed, as shown below. The upper limit of the linear part of the standard curve is about 100 pg ml-l or 2000 pg in 20 ml of final solution. COMPARATIVE DETERMINATIONS OF TIN IN CANNED BABY-FOOD BY THE QUERCETIN AND THE In Table I11 are given the comparative results obtained under the following experimental conditions by the two methods on actual food samples.Two 10-g sub-samples from each of several different canned baby-food products were ashed with 10 ml of concentrated sulphuric acid Plas the amount of 30 per cent. hydrogen peroxide necessary for complete ashing; 80 pg of tin as tin(1V) standard solution was added to one sub-sample from each can. After ashing, sulphuric acid and water were added to give a volume of 50 ml of 25 per cent. V/V sulphuric acid. Of these solutions, 25 ml were used for the quercetin determination and the other 25 ml for the atomic-absorption deter- mination, both after extraction as described above. [Because of the expected higher tin content of cans I, I1 and I11 (unlacquered) of apple sauce, only two 5-g samples were ashed, and fractions corresponding to 0.5 g of sample were extracted.] TABLE I11 All values are given in micrograms of tin per gram of sample ATOMIC-ABSORPTION METHODS- COMPARATIVE DETERMINATIONS OF TIN I N CANNED BABY-FOOD BY THE TWO METHODS Atomic-absorption Quercetin method method A I \ A Diffe:- r- Total Differ- Total Sample (a) Added Content ence (b) Added Content ence (a)-@) 8.4 Chopped fish, can I 8.4 0 8.4 8.6 0 16.5 8.0 8.5 } +"' 16.4 8.0 Chopped fish, can I1 Chopped meat and tomato, can I Chopped meat and tomato, can I1 Vegetables and meat, can I Vegetables and meat, can I1 61.6 0 + (9.6) Apple sauce, can I 71-2 0 84.4 16.0 k!l:i } -2*8 79.2 16-0 Apple sauce, can I1 70.6 0 71.6 0 - 1.0 88.6 16.0 ;::: } +2'o 86.6 16.0 ;;:! } Apple sauce, can I11 97.0 0 94.2 0 + 2-8 112.0 16-0 :;:! } -Ieo 109.4 16.0 :;:: } -'*' Apple sauce, can IV 1.3 0 k:: } -0.6 7.6 8.0 } -Os5 +Oa4 8.7 8.0 Mean difference -0.22 f 0-6 On average, three cans (or six sub-samples) were examined per day, together with reagent standards and blanks. Comparison of the absorbance increment corresponding to the tin added to the sub-samples with the slope of the reagent standard curve obtained on the same day showed no significant difference between internal and external standard curves, i.e., no significant difference from 100 per cent.recovery of added tin. Consequently, the average slope of internal and external standard curves was used for calculation of the tin contents of the sample solutions. Thus, results in the range 5 to 175 pg of tin per 50 ml of final solution were obtained including the added amounts of tin.After subtraction of these amounts, the net results expressed in micrograms per gram of sample were in the range 1 to lOOpgg-l, as shown in Table 111.February, 19731 FOR THE DETERMINATION OF TIN I N FOOD 145 As the range of directly measured concentrations, 5 to 175 pg per 50 ml, is beyond the range given in Table I, the standard deviations shown in this table cannot be directly applied to the present results as estimated standard deviations for the baby-food samples. However, conclusions concerning the accuracy and conformity of the two methods can be drawn from these results as follows. As mentioned earlier, comparison of the slopes of internal and external standard curves did not reveal significant deviation from 100 per cent. recovery of added tin, which is further confirmed by calculation of the mean of differences between the net results of paired sub- samples after subtraction of the amount of tin nominally added.Provided that approximately normal distribution of the differences obtains (which is not contradicted by the results), it can be seen that the mean differences, which are shown together with their standard deviations at the bottom of Table 111, are not significantly different from zero for either of the two methods. Further, the mean of differences between the spectrophotometric quercetin and atomic- absorption results (both without standard additions) was calculated, excluding a single, un- explained, greater deviation. Comparison with the standard deviation shows that the mean of differences does not significantly deviate from zero : the results given by the quercetin and atomic-absorption methods are not significantly different. It is interesting to note that the mean of differences and the corresponding standard deviation are smaller for the atomic-absorption than for the spectrophotometric quercetin results. Although the difference is not significant, there is a tendency for it to be in the reverse direction to that found near the detection limits of the methods. Consequently, in the con- centration range covered by the baby-food samples, the atomic-absorption method is at least as precise as the spectrophotometric quercetin method. CONCLUSION It is concluded that while the spectrophotometric quercetin method is preferred for very low tin concentrations (for example, for organotin residues), the spectrophotometric and atomic-absorption methods are equally useful for the determination of tin in concen- trations normally found in canned food. F. Bro-Rasmussen, Head of the Department, is thanked for helpful advice concerning the presentation of this work. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. REFERENCES Analytical Methods Committee, Analyst, 1967, 92, 320. Kirk, R. S., and Pocklington, W. D., Ibid., 1969, 94, 71. Corbin, H. B., J . Ass. Off. Analyt. Chem., 1970, 53, 14. Koch, O., and Koch-Dedic, G., “Handbuch der Spurenanalyse, ” Springer-Verlag, Berlin, 1964. Juliano, P. O., and Harrison, W. W., Analyt. Chem., 1970, 42, 84. Capacho-Delgado, L., and Manning, D. C . , Spectrochirn. Acta, 1966, 22, 1505. Harrison, W. W., and Juliano, P. O., Analyt. Chem., 1969, 41, 1016. Pearlman, R. S., Hefferen, J. J., and Lyon, H. W., J . Dent. Res., 1970, 49, 1437. Clauss, C., Laugel, P., and Hasselmann, M., Chim. Analyt., 1971, 53, 102. Newman, E. J., and Jones, P. D., Analyst, 1966, 91, 406. Malkus, Z., 2. Lebensmittelunters. u. -Forsch., 1957, 106, 257, KarvAnek, M., Curda, D., and Miler, V., Pram. Potravin., 1965, 16, 369; Analyt. Abstr., 1966, 13, Received April 24th, 1972 Accepted Sefitember 28th, 1972 6510.

ISSN:0003-2654    

DOI:10.1039/AN9739800137

出版商:RSC    

年代:1973

数据来源: RSC

摘要:

146 Analyst, February, 1973, Vol. 98,p. 146 The Spectrophotometric Determination of Nitrofurantoin in Blood and Urine BY JANE HARRISON, D. A. LEWIS* AND R. J. ANCILL (Pharmacology Group, School of Pharmacy, University of Bath, Bath, Somerset, BA 2 7A Y ) THE method of Conklin and Hollifield1p2 for the determination of nitrofurantoin in blood and urine lacks sensitivity because of a progressive, non-uniform increase in absorbance at 400 nm. We have sought the cause of this increase. The addition of Hyamine 10-X hydroxide to nitromethane also produced increased absorb- ance at 400 nm, which suggested that the increased absorbance that complicated analyses was dependent on the reagents or the product of their reaction, namely methazonic acid. We therefore prepared methazonic acid in the following manner, which is essentially the method of Dunstan and G~ulding.~ A 2 M solution of Hyamine hydroxide in nitromethane was stored overnight at 4 "C.The light, viscous oil that formed was separated from excess of nitromethane and poured into a vigorously stirred excess of diethyl ether. The resulting yellow oil was dissolved in 2 N sulphuric acid, the free methazonic acid then being isolated by the method of Dunstan and Goulding3 The product was identified by infrared spectroscopy. At room temperature it rapidly became brown and absorbed strongly at 400nm, in contrast to the white, freshly prepared compound, which was transparent at this wavelength. Storage at -20 "C failed to arrest this degeneration, which led, in a few days, to a red coloration and in a few weeks to the formation of a brown resin.We suggest that the decomposition of methazonic acid into highly absorbing substances accounts for the difficulty encountered in the analytical procedure for the determination of nit rof urantoin. When the absorbances of test solutions were measured exactly 10 minutes after the addition of 0.5 ml of 0-04 M Hyamine hydroxide in methanol, reproducible results were obtained for the determination of nitrofurantoin in urine. The problem was more acute in the method for the determination of nitrofurantoin in blood, as the proportion of Hyamine hydroxide to nitromethane was greater and the concentration of drug was smaller. The modified method of Mattok, McGilveray and Charette,4 in which the concentration of Hyamine hydroxide is reduced, was adopted. Pure methazonic acid was unstable. We thank Berk Pharmaceuticals Ltd. for financial assistance. REFERENCES 1 . 2. 3. 4. Conklin, J. D., and Hollifield, R. D., Clin. Chem., 1966, 11, 926. -*- , Ibid., 1966, 12, 690. Dunstan, W. R., and Goulding, E., J . Chem. S O ~ . , 1900, 1262. Mattok, G. L., McGilveray, I. J., and Charette, C., Clin. Chem., 1970, 16, 820. Received March 16th, 1972 Accepted September 13th, 1972 * Present address : Department of Pharmacy, University of Aston in Birmingham, Gosta Green, @ SAC and the authors. Birmingham, B4 7ET.

ISSN:0003-2654    

DOI:10.1039/AN9739800146

出版商:RSC    

年代:1973

数据来源: RSC

摘要:

A nabst, February, 1973, Vol. A Method for the 98, pp. 147-148 147 Detection of Microgram Amounts of Hydroxamic Acids BY Y. K. AGRAWAL (Department of Chemistry, Indian Institute of Technology, Powai, Bombay-76, India) A rapid and simple method for the detection of small amounts of hydroxamic acids is described. The chloroform-soluble, violet-coloured vanadium complex formed between vanadium and the hydroxamic acids allows the detection of as little as 10 to 20 pg of hydroxamic acid in one drop of sample solution. A LARGE number of hydroxamic acids, which have the functional group (I), have been s ynt hesised. l-' -N-OH -c=o I (1) These acids are versatile reagents and have applications in both organic and inorganic analysis.8-10 They react with many metal ions to give complexes (11).R1-N-0 (n - k)+ [ ( R2-(!=O))?n+] The reaction of hydroxamic acids with vanadium(V) to yield a violet-coloured complex has been utilised for the detection of microgram amounts of these acids on a spot plate.llJ2 The present paper describes a procedure for the detection of hydroxamic acids that involves the use of the violet-coloured vanadium complex formed in an acidic medium. EXPERIMENTAL REAGENTS- Standard vanadium solution, approximately lo-* M-A solution containing 11.7 mg 1-1 of analytical-reagent grade ammonium metavanadate was prepared and standardised by the method of Tandon and Bhattacharyya.12 Chloro form-AnalaR. Hydrochloric acid-AnalaR. PROCEDURE- Mix one drop of vanadium solution with two drops of the concentrated hydrochloric acid, with warming if necessary, in a micro-scale test-tube.Cool the mixture, add one drop of a solution of the sample in chloroform and shake it vigorously. If a hydroxamic acid is present a violet colour that is extracted into the organic phase develops in a few seconds. The coloured complex is stable for several days. DISCUSSION Hydroxamic acids are sensitive and selective reagents for vanadium12-15 with which they form violet-coloured complexes that can be extracted into chloroform from acidic media. The limits of detection for the substituted N-arylhydroxamic acids are within the range 10 to 20pg. 0 SAC and the author.148 AGRAWAL The present method is applicable to all hydroxamic acids, which gives it an advantage over the benzoyl peroxide fusion method of Feigl and Silva, which was found in the present work to be inapplicable to unsaturated and disubstituted hydroxamic acids.ls It also provides greater limits of detection, is less tedious and time consuming and is more selective.The author thanks Dr. S. G. Tandon for his guidance and br. A. B. Biswas, Senior Professor of Chemistry at the Indian Institute of Technology, Bombay, for providing facilities. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. REFERENCES Priyadarshini, U., and Tandon, S. G., J. Chem. Engng Data, 1967, 12, 143. Bhura, D. C., and Tandon, S. G., Ibid., 1969, 14, 278. Agrawal, Y. K., and Tandon, S. G., J. Indian Chem. Soc., 1971, 48, 397. -- , Ibzd., 1971, 16, 495. Gupja, V. K., and Tandon, S. G., Ibid., 1972, 17, 248. Agrawal, D. R., and Tandon, S. G., Ibid., 1972, 17, 287. Davidson, D., J. Chem. Educ., 1940, 17, 81. Brandt, W. W., Rec. Chem. Prog., 1960, 21, 195. Agrawal, Y. K., Analyst, 1972, 97, 578. Tandon, S. G., and Bhattacharyya, S. C., Analyt. Chem., 1964, 36, 1378. -- , Ibid., 1961, 33, 1267. Ryai, D. E., Analyst, 1960, 85, 569. Bhura, D. C., and Tandon, S. G., Analytica Chim. Acta, 1971, 53, 379. Agrawal, Y. K., Analyt. Lett., in the press. Feigl, F., and Silva, E., Analyst, 1957, 82, 582. -- , , J..Chem. Engng Data, 1971, 16, 371. Received June 20th, 1972 Accepted Augzcst 7th, 1972

ISSN:0003-2654    

DOI:10.1039/AN9739800147

出版商:RSC    

年代:1973

数据来源: RSC

摘要:

Analyst, February, 1973 Book Reviews 149 METHODS OF BIOCHEMICAL ANALYSIS. Edited by DAVID GLICK. Volume 19. Pp. viii + 632. 1971. Price L9.85. H. Haglund of Stockholm deals with isoelectric focusing in pH gradients, a technique for fractionating and characterising ampholytes. Proteins that differ by as little as 0.02 pH unit at the isoelectric point can be separated. The term “isoelectric focusing” has been in use since 1967 and replaced “isoelectric separation” and several other cxprcssions. Development of the method was delayed by the technical difficulty of achieving a stable pH gradient and preventing convection in the electrolyte. Kolin in 1955 used columns stabilised with a sucrose density gradient while Svensson in 1961 to 1967 provided the theoretical and practical groundwork for further developments.In 1966, Vesterburg succeeded in producing synthetic carrier ampholytes that met Svensson’s requiremcnts. Thcn, in 1969, Valmct devised new apparatus for “zone convection clectrofocusing,” a self-stabilising dcnsity-gradient system that has proved to be of service for both analytical and preparative tasks. The carrier ampholytes now in use are aliphatic polyaminopolycarboxylic acids that display many values of pK for the same type of New York, London, Sydney and Toronto: Interscience Publishers. -CH,-N-(CH,) ,-N-CH,- I I I (CH2) 5 R NR2 molecule, where R = H or -(CH,),COOH and x = 2 or 3, and that have the desired high solubility in water. These carriers are marketed as “ampholines” and cover the pH range 3 to 10 (LKB- Produkter AB, Sweden).They are supplied at a concentration of 40 per cent. m/V in 25-ml bottles and over the pH range 3 to 10, the standard fractions refer to 3 to 6 , 6 to 8, 7 to 10, 3 to 5, 4 to 6, 5 to 7, 6 to 8, 7 to 9 and 8 to 10. When an ampholine solution in water is exposed to a suitable voltage under convection-free conditions, a “natural” pH gradient is set up by the current itself. When a protein mixture is introduced into a system such that the pH increases from the anode to the cathode, the constituents will concentrate in narrow zones, focusing occurring at the isoelectric point of each protein. Isoelectric focusing has been applied to polyacrylamide gels in order to stabilise them against convection. In density-gradient electrofocusing, ultraviolet absorption at 260 or 280 nm is used to monitor separations.Haglund, writing from the LKU laboratories, describes experimental procedures in detail. The topic already requires 161 references and there can be no doubt that, in one form or another, isoelectric focusing has a considerable future for the qualitative and quantitative analysis of proteins and peptides, as well as for preparative work. Mass spectrometry in the determination of structures of natural products containing sugars is the subject of a monograph by S. Hanessian of Montreal. After giving a good account of basic principles and definitions, he proceeds to survey the results of investigations on the mass spectra of a wide range of simple carbohydrates and their derivatives. This survey is the main contribu- tion; it occupies 70 pages and covers the ground well.I t is followed by a section on mass spectra of antibiotics and natural products containing sugars. Results on nucleosides and nucleotides are summarised before nucleoside antibiotics such as amicetin and puromycin are considered. High- resolution mass spectrometry on picromycin (found molecular ion m/e 625.3298 ; calculated for C2,H,,N0,, 525.3302), o-trimethylsilylpimaricin, lucensomycin and nystatin in all instances led to corrections of the postulated structures. Mass spectra of cardenolides and cyclitol antibiotics such as streptomycin and neomycin are also recorded. This chapter is clearly useful and is very readable. A contribution from the Department of Medicine at Bristol University, by J.R. Clamp, T. Bhatti and R. E. Chambers, deals with the determination of carbohydrates in biological material by gas - liquid chromatography. The topic was also reviewed in 1962 in this series (Volume 10) by C. T. Bishop and the present contribution should be read in the light of earlier reviews. A long table lists the carbohydrate derivatives studied and gives references to the literature. The compounds are arranged systematically and grouped under four headings according to the type of derivative prepared for gas - liquid chromatography (0-methyl, U-acetyl, 0-trimethylsilyl and “others, ” such as U-trifluoroacetyl). A second table gives detailed information from selected references so as to illustrate the separations achieved by various systems.150 BOOK REVIEWS [ATZaZyst, VOl.98 The article goes on to discuss the results for different types of derivatives. The most interesting part concerns “biological” materials and special emphasis is laid on (a) qualitative and quantitative analysis of constituent monosaccharides and ( b ) structural studies on polymers. Trimethylsilyl derivatives are favoured because they are easy to prepare and yields are quantitative. Internal standards are noted and methods of analysis of heterosaccharides are described. Finally, a recom- mended analytical procedure is given in detail and discussed fully. G. W. Leddicote of the Georgia Institute of Technology, Atlanta, contributes a comprehensive review of activation analysis of the “biological” trace elements. The method depends on the fact that bombardment with nuclear particles can induce radioactivity in an isotope (or isotopes) of many of the elements present in a sample.The aim is then to measure the radioactivity of a specific radionuclide and make use of it in analysis. About seventy elements have already been determined in this way. At least 1160 radioactive species have been recorded and sixty-six of them occur naturally while the others are artifacts made by impacts on atoms of neutrons, protons, deuterons, helium nuclei, electrons or gamma-radiations. Neutron activation is the method that is used most frequently. Leddicote lists the detection limits in parts per million, and these limits range from 6 (bismuth) to 0.1 (iron, lead, silicon, sulphur and zirconium), 0.01 (magnesium), 0.002 (mercury) and 0-000 006 (silver).Theoretical aspects, e.g. , nuclear reactions and cross-sections, the formation of radio- nuclides and the sensitivity of the activation method, are reviewed. I t is feasible for “counting” to be carried out without chemical separation (“direct” or non-destructive activation analysis). Gamma-ray spectrometry is much used and may be complemented by serial measurements in order to take into account the decay of short-lived radioisotopes. Gamma-ray spectra produced by multichannel pulse-height analysers are complex structures and numerous publications are cited that deal with computer techniques for identifying and measuring mixtures of gamma-rays. The alternative procedure is to add a known mass of a non-radioactive chemical carricr and to carry out a chemical separation prior to counting. Applications of activation analysis have been tabulated and documented and separate tables cover (i) human biochemistry, (ii) trace-element constituents of animal tissues and body fluids and (iii) activation analysis of trace elements in plants.J. Homolka of Prague deals with the polarography of proteins, a subject in which considerable advances are being made. Volume 11 (1963) contained an article by Muller on the polarographic analysis of proteins and amino-acids by means of the BrdiEka reaction. In this method, serum can be added to solutions of cobalt (111) salts [hexamminocobalt (111) chloride] without previous treatment and a double wave is obtained. This characteristic is a polarographic property that is peculiar to proteins.The reaction is based on the presence of SH and S-S groups. An addition of cysteine causes an increase in the second wave in solutions of cobalt(I1) salts. The polarographic method is particularly suitable for following the denaturation of proteins in biological fluids and important possibilities in clinical biochemistry are emerging. ‘I’he denaturation reactions have been followed in healthy persons and in abnormal subjects. The behaviour of albumin and globulin fractions has been studied and interesting and promising results have been obtained. The techniques used are complicated and specialised and the article has a twenty-page section on methodology. This section contains a confluence of advanced physical chemistry and explora- tory clinical biochemistry that presents physicians with challenges to their powers of interpretation.Homolka has made a definite contribution in assembling the newer knowledge on techniques and results. There are 260 references. In this series of volumes, the scientists asked to write on particular topics are chosen either because they originated a method or have had special and personal experience of such a method. This latest volume contains five articles, each of which comprises approximately 50 000 words, and covers an important area admirably. ATLAS OF THERMOANALYTICAL CURVES (TG-, DTG-, DTA-CURVES MEASURED SIMULTANEOUSLY). Volume 1. Edited by G. LIPTAY. Pp. xiv + 15-116. London, New York and Rheine: Heyden & Son Ltd. 1971. Price Lll.50; $28; DM103. This book contains a wealth of interesting and valuable material conveniently arranged for reference purposes.As the editor mentions in the Preface, this arrangement is not an easy task, as experimental conditions have a much greater influence in thermal analysis than in other methods. This problem is only partially resolved in the present edition, as all the results quoted were obtained There are 1027 references. R. A. MORTONFebruary, 19731 BOOK REVIEWS 151 on the derivatograph. Thus the thermogravimetric, differential thermal gravimetric and differen- tial thermal analytical curves on any one material can be compared, but comparisons with results from other equipment cannot be made. References are given, however, to some data from other sources, but these data are not plotted.In recording the data in this Atlas two extremes are chosen: Case 1, in which a small sample is used with a slow heating rate, and Case 2, in which larger samples are taken with a faster heating rate. To comprehend fully how to use this Atlas effectively, it is essential that the Preface and Appendix are read in detail. The notes that go with each material should also be read with care in order to appreciate experimental details and the explanations of the course of events described by the plotted data. This edition contains fifty entries but it is possible to envisage that some of these single entries will in future merit an entire edition to themselves-the entries under the heading of coal may be cited as an example. D. DOLLIMORE PHOTOMETRIC ORGANIC ANALYSIS.BASIC PRINCIPLES WITH APPLICATIONS. Part I. By EUGENE SAWICKI. Volume 31 in Chemical Analysis : A Series of Monographs on Analytical Chemistry and its Applications. Pp. xvi + 679. New York, London, Sydney and Toronto: Wiley- Interscience. 1970. Price L15.25. This volume is concerned mainly with the basic principles and relatively little with the .ications of photometric organic analysis, which will be discussed in a later volume. Nevertheless, it is a most useful book for practising analytical chemists, and to others for whom spectrophotometry is a working tool; indeed, in one relatively short but highly interesting section dealing with problems that are amenable to the photometric approach, a list of the various problems mentioned is sufficient to indicate the philosophy behind the book: the author lists health, including cancer, cardiovascular diseases, genetic diseases, allergy and asthma, pregnancy, medicines and drugs, toxicology, pesticides, food, water and air pollution, investigation of the brain and its functions, the biochemistry of various life forms, and the origin and synthesis of life.There are large numbers of examples of each group of compounds, including those having zwitterionic, cationic or anionic structures, and the effects of temperature and solvent environment on the spectra of compounds are dealt with in detail. The text is well illustrated with tabulated data. These data amplify, in a practical manner, many of the points that are enunciated in the text. The arrangements in the various chapters are easily followed and help to make the book useful as a source book for information. For many chemists concerned with the interpretation of results obtained from ultraviolet and visible spectrophotometry, this book will be a valuable addition to their libraries.The basic theories of photometric analysis are considered in various chapters. L. S. BARK APPLICATIONS OF INFRARED SPECTROSCOPY IN BIOCHEMISTRY, BIOLOGY AND MEDICINE. By FRANK S. PARKER. Pp. xiv + 601. London: Adam Hilger. 1971. Price k12. Fifteen years ago, infrared spectroscopy could be acclaimed as the most powerful tool available for the rapid identification and measurement of an organic compound. This claim should now, so far as compounds of biological and medical interest are concerned, be reserved for nuclear magnetic resonance spectroscopy.This change of emphasis does not mean that infrared spectro- scopy has been outmoded. It is still the preferred spectroscopic technique for the routine study of compounds in the solid state and for the qualitative examination of relatively small samples; in structural and analytical studies, it can normally provide information to complement and supplement that supplied by nuclear magnetic resonance measurements. Further, infrared spectrometers are considerably cheaper than nuclear magnetic resonance instruments and are used by the modern chemist almost as freely as an earlier generation used a melting-point apparatus ! These points should be remembered when reading the present book, which surveys applications of infrared spectroscopy in biology and medicine.Dr. Parker has produced a comprehensive monograph, in which, after describing experimental procedures and sampling techniques, he discusses the spectra of certain groups of compounds, such as carbohydrates, lipids, amino-acids, proteins, nucleic acids and steroids, that either have a biological origin or are used in biological studies, and then reviews applications in specific fields, such as enzymology and microbiology. The biochemical topics covered range from bones, teeth and minerals to paleobiochemistry and insect sex attractants. The volume includes a chapter on information sources for infrared spectro- scopy and spectra, and concludes with an appendix on methods for fractionating and purifying162 BOOK REVIEWS [Analyst, Vol.98 samples before submitting them to infrared examination. There is a general index and a chemical compound index but not an author index. The author is fairly broad-minded as regards his interpretation of biology and medicine, but there are a few gaps, such as the absence of references to the spectra of steroidal sapogenins. The detail with which individual subjects are treated varies from section to section; often an infrared method is described uncritically without any indication as to whether alternative, and possibly better, methods involving the use of other techniques are available. This volume will be of value to an analyst who wishes to know whether or not a particular problem has been studied by infrared spectroscopy; it will give him a good lead into the literature, but, per se, it will not always help him to select the best method for solving his problem.Never- theless, the book is a useful reference work that should find a place in all spectroscopic libraries. J. E. PAGE ELECTRON PROBE MICROANALYSIS. By L. S. BIRKS. Second Edition. Volume 17 in Chemical Analysis: A Series of Monographs on Analytical CJzemistry and its Applications. Pp. xiv + 190. New York, London, Sydney and Toronto: VC'iley-Interscience. 1971. Price i 7 . This book describes concisely what electron probe microanalysis is, the features of instru- mentation that potential users should be aware of, and how to present and interpret the results. It is a primer rather than an encyclopaedia and will be appreciated as such by many, especially as so much of the information is conveyed visually by excellent diagrams and photographs, which account for about half of the book.Inevitably, some matters are treated so briefly that one wonders why they were included at all, but this is more than offset by the way in which the difficult points of practical significance, such as correction factors, are handled. Most newcomers to the technique will want this volume by their side, and users of an electron-probe service will find that the short time needed to read the book is well spent. The author notes that the publishers objected to his including a colour photograph because of the cost of printing (p. 100). One is therefore astounded by the price of a second edition of a small book, the success of which is reasonably (and rightly) assured.D. BETTERIDGE THE ELECTRICAL SENSING ZONE METHOD OF PARTICLE SIZE MEASUREMENT (THE COULTER PRINCIPLE). By T. ALLEN and K. MARSHALL. Pp. viii + 105. Published by the authors at the University of Bradford. 1972. Price i 2 . The authors will probably find that the demand for this edition will excecd the supply and the reviewer confidently expects that further editions with revisions will appear in the future. An exposition of this kind is certainly required, not only on this technique but also on other techniques that are now available for particle sizing. This book is very difficult to read, however, as not enough attention has been paid to the layout of material, to a text that in part is little more than a series of lecture notes and to the use of symbols that are inadequately defined.The criticism on layout is illustrated by reference to Fig. 1, which has a caption, while Figs. 2 and 3 on the same page have no captions and no explanations are given in the figures. Tables on pages 7 and 25 have no titles or numbers. The theory on pages 5 to 8 is too inadequately presented for the reader to make a reasoned assessment of the method, and the authors would be well advised in future editions to present a more critical approach to the theory. One page on the subject of electrolytes is disappointing, and the provision of only two sentences on the major problem that in aqueous environments the presence of the electrolytes often causes flocculation has condensed a major problem in the use of this equipment almost to vanishing point. I t would need a very persuasive Irishman to contend that Chapter 15, on Industrial Applications, is anything but a list of references brought together by two sentences! The authors may well contend that their symbols are defined in a long list a t the commencement of the book, but care should still be maintained that they are all defined as they are introduced in the text. Having said all this, the reviewer must say that he expects this book to be a best seller and to be rapidly upgraded in future editions. The authors are well known experts in this field, they have a confident grasp of their subject and have produced a book on a topic that has up to now been found only in commercial manuals and leaflets. A detailed examination of the equipment described will reveal that the commercial models are constantly being improved, which provides a further reason for the expectation that a constant revision of the subject matter of this book will be required. This book is a detailed publication on a single technique for particle-size analysis. D. DOLLIMORE

ISSN:0003-2654    

DOI:10.1039/AN9739800149

出版商:RSC    

年代:1973

数据来源: RSC