Table I. Determination of Inorganic Phosphate in Control Blood Serum with and without Added ATP Inorganic phosphorous, mg/100 ml of serum Found by Found by Reporteda F-Sb R.R.O 4.6 4.6d 3.8 3.8e
4.5 7.1 3.8 6.9
4.48 4.54 3.95 3.98
Manufacturer’s value. Fiske-Subbarow analysis (2). Reaction rate analysis (1). d Blood serum doped with solid ATP to make 1@1M ATP. e Blood serum doped with solid ATP to make 10-2 ATP. a
b
30 seconds, phosphate can be determined with very little interference from ATP hydrolysis, even when ATP is present in about 100-fold molar excess. The experiment illustrated in Figure 1 indicates that the
h a 1 step in the analysis of a biological sample can be accomplished by the reaction rate procedure with very little hydrolysis of ATP. To test the entire analysis procedure, blood serum samples with accurately known amounts of inorganic phosphate were doped with ATP and analyzed by the rate method and the Fiske-Subbarow method. Table I shows the results obtained on control serum with and without added ATP. The highest ATP content represented about a 100-fold molar excess over phosphate. These results indicate that the reaction rate procedure for phosphate is capable of high specificity. Hydrolysis of ATP is seen to contribute only a very small error even when ATP is present in large excess. Because of convenience and the availability of control samples, only blood serum was utilized as the biological sample. However, these results can be directly related to other biological materials for the same analysis steps (deproteinization, centrifugation, and analysis of the supernatant) are required. RECEIVED for review May 13, 1968. Accepted June 17,1968.
On-Column Formation and Analysis of Trimethylsilyl Derivatives George G . Esposito U.S . Army Coating and Chemical Laboratory, Aberdeen Proving Ground, M d . 21005
TRIMETHYLSILYL (TMS) reagents are used extensively in gasliquid chromatography (GLC) to improve the chromatographing properties of various chemical types. The reaction is relatively simple and rapid; reactive species yield compounds which are more volatile, less polar and thermally stable. A number of methods have been employed for TMS derivatization. These include procedures for alcohols and amines (I), phenols (Z), sugars (3),amino acids (4) and polyols (5); Mason and Smith (6) investigated the quantitative aspects of the technique. Even though the above methods have found great utility in GLC analysis, they all suffer from one deficiency; they are not applicable to compounds mixed with reactive solvents. In his review on reaction gas chromatography (7), Beroza referred to the on-column mode of derivative formation, but the methods were primarily concerned with the “peak-shift” technique and did not mention the on-column formation of TMS derivatives. A method is presented which describes an on-column technique for the formation of TMS derivatives of certain classes of compounds and their subsequent chromatographic analysis, The principal feature of the technique is the ability to accommodate aqueous and alcoholic solutions, thus avoiding excessive sample manipulations. The concept is illustrated by application to fatty acids, carboxylic acids, and polyhydric alcohols. ~~
(1) S. H. Langer, S. Connell, and I. Wender, J. Org. Chem., 23, 50 (1958). (2) S . Friedman, C. Jahn, M. Kaufman, and I. Wender, Bur. Mines BUM.609 (1963). (3) C. C. Sweeley, R. Bentley, M. Makita, and W. W. Wells, J. Amer. Chem. SOC.,85,2497 (1963). (4) K. Ruhlman and W. Gresecke, Angew. Chem., 73,113 (1961). (5) B. Smith and 0. Carlsson, Acra Chem. Scad., 17,455 (1963). (6) P. S. Mason and E. D. Smith, J. Gas Chromatog., 4, 398 (1966). (7) M. Beroza and R. A. Coad, ibid., p 199.
1902
ANALYTICAL CHEMISTRY
A solution containing the sample is injected onto the chromatographic column and is immediately followed by an injection of TMS reagent. The interval between injections is all the time required for the water and alcohol to separate from the rest of the sample. TMS derivatives are formed as the reagents pass through the zones occupied by reactive compounds and are chromatographed as they traverse the remaining segment of the column. EXPERIMENTAL
Reagent. The trimethylsilation reagent used throughout the investigation was a commercial product recommended for the conditioning and treating of chromatographic columns. It is distributed by the Pierce Chemical Co. (Rockford, Ill.) under the trade name, SILYL.8 and consists of three different TMS donors, N,O-bis-(trimethylsily1)-acetamide, trimethylsilyldiethylamine, and hexamethyldisilazane. The details of the method used to obtain the experimental data are presented in Table I. The test is conducted by preloading the sample and reagent syringes; the sample is introduced first, immediately followed by the reagent.
RESULTS AND DISCUSSION A 1 solution of polyols in a mixture of alcohol and water was analyzed and produced the chromatogram shown in Figure 1. When a water solution of the same mixture was tested, the results were essentially the same. There is an obvious decrease in yield as the boiling point of the poly01 increases. A probable explanation is that the time duration of contact between reagent and poly01 is greatest with the lower boiling polyols. . When a synthetic mixture of saturated fatty acids was injected onto the column from an alcohol solution, the oncolumn reaction with reagent produced the chromatogram
B C
D
0
15
IO
5
20 25 MINUTES
35
30
40
0
Figure 1. On-column formation and separation of TMS-polyol derivatives E. Glycerine F. Trimethylol ethane G. Trimethylol propane H. Pentaerythritol
Ethylene glycol B. Neopentyl glycol C. 1,4-Butanediol D. Diethylene glycol A.
shown in Figure 2. Figure 3 illustrates the results from the analysis of dicarboxylic acids. The same three samples were analyzed without the addition of reagent. All compounds emerged as badly "tailing" peaks and in some cases were barely distinguishable from the base line. Column Packing. Because the reagent reacts with substrates containing active hydrogen, only liquid phases without active hydrogen sites can be used. All of the work shown was performed on silicone grease columns, but other column packings were investigated. When a polyester column was used, the fatty acids were greatly disproportionated with the unsaturated fatty acids giving the lowest yields. Some batches
52 lX-
EX+
A
0
5
IO
15
MINUTES
20
I 25
Figure 3. On-column formation and separation of TMS-dicarboxylic acid derivatives A. Fumaric 13. Adipic
C. Azelaic D. Sebacic
5
10
15
20
25
30
MINUTES
Figure 2. On-column formation and separation of TMS-fatty acid derivatives A. Lauric B. Myristic
C. Palmitic D. Stearic
of reagent performed better than others. In a recent publication (8), Zinkel indicated that incongruous results were obtained when TMS esters of fatty acids and resin acids were separated on polar phases. The derivatives were prepared externally before being chromatographed. Porous polymer beads (styrene-divinyl benzene type) were investigated as a column packing because of their polarphobic properties, but they proved unsuitable because of the long retention time for the reagent. Column Length. The length of the column was most critical with the analysis of polyols. Six- and ten-foot columns (8) D. F. Zinkel, M. B. Lathrop and C. L. Zank, ibid., J . Gas Chromatog., 6,158 (1968).
Table I. Detail of Test Instrument: Model 810 F& M gas chromatograph equipped with a thermal conductivity detector Column: Copper tubing ('/(-inch) packed with 2 0 z silicone grease on 60-80 mesh Chromosorb W. General Conditions Reagent volume, pl 40 Injection port temperature, "C 300 Detector temperature, "C 300 75 Helium flow rate, cc/minute Detector cell current, mA 160 Specific Conditions DicarFatty boxylic Poly01 acids acids 3 2 2 Sample size, p1 Sample concentration, 1 10 5 Solvent, ethanol :water 75:25a 953 60:40 Length of column, ft 10 6 10 Initial column temperature, "C 40 100 50 Terminal column temperature, OC 250 260 300 Column heating rate, "C/minute 4 6 10 If the same conditions are observed, water solutions produce comparableresults (Figure 1).
VOL 40, NO. 12, OCTOBER 1968
1903
performed well; but shorter columns, 4-foot and less, gave rise to poorer yields. As the column is shortened, water solutions are adversely affected to a greater degree than alcohol solutions. Column Temperature. Low initial temperatures and programmed temperature analysis produced the best results for polyols. The fatty acids and dicarboxylic acids were not as sensitive to column temperature changes. Silylation Reagent. SILYL. 8 is composed of three powerful TMS agents which are claimed to have a synergistic effect on silylatable materials. It was developed as a column conditioner for improving column efficiency, reducing tailing, and removing silylatable residues from columns. Silylanizing agents that release HC1 into the system are absent. Some variation in the composition and performance of different batches was observed; but as the material was not developed as a reagent for derivatization, this was to be expected. Reproducibility with the same batch of reagent was good. Some impurities from the reagent show up in the poly01 range. Sample-Reagent Ratio. Decreasing the amount of reagent had the greatest effect on the poly01 analysis. Smaller injections, 10-20 pl, caused a decrease in the size of the derivative peaks; other peaks appeared which are believed to be intermediate derivatives which had not been converted to the final product due to insufficient reagent. Even though the samplereagent ratio may not be optimum, it is adequate for the qualitative study made during this investigation.
Evidence indicates that the reaction occurs in the liquid phase. Decreasing the liquid phase causes a substantial reduction in the yield of poly01 derivatives. Fifteen-, thirty-, forty-five-, and sixty-second intervals between injections were tried. In addition, a slow injection of reagent of 10-second duration was also used. These changes in the use of the reagent had no substantial effect on the results. In general, variations in the operating conditions had a greater effect on the poly01 analysis than on the analysis for fatty acids and dicarboxylic acids. This is probably caused by the difference in reaction rates. Moderate variations in carrier gas flow, helium pressure, and injection port temperature had no effect on the analysis. Extreme changes were not investigated. Adherence to definite operating conditions is recommended in order to avoid variable affects on the results. It may be possible to improve the method by altering the ratio of components in the reagent system or by using other silylation reagents . ACKNOWLEDGMENT
The advisory assistance of C. F. Pickett, Director of the Laboratory, and M. H. Swann, Chief of the Analytical Section, is acknowledged and appreciated. RECEIVED for review April 25, 1968. Accepted June 24,1968
Titration of Some Metal Diethyldithiocarbamates by Iodine in Chloroform Arthur F. Grand’ and Milton Tamres Department of Chemistry, University of Michigan, Ann Arbor,
SEVERAL METHODS have been developed for the analysis of water-insoluble metal dialkyldithiocarbamates ( I , 2). One common procedure is the acid decomposition of the salt with analysis of either the amine or carbon disulfide product ( I , 3-5). Other methods such as photometry ( I , &S), chromatography ( I , 9), and polarography ( I O , 11) involve use of a nonaqueous solvent. Photometry requires prior knowledge of Present address, FMC Corporation, Princeton, N. J. (1) G. D. Thorn and R. A. Ludwig, “The Dithiocarbamates and Related Compounds,” Elsevier, New York, 1962. (2) E. E. Reid, “Organic Chemistry of Bivalent Sulfur,” Chemical Publishing Co., Inc., New York, 1962, Vol. IV, pp 196-385. (3) T. Callan and N. Strafford,J. SOC.Chem. Ind., 43, IT (1924). (4) D. G. Clarke, H. Baum, E. L. Stanley, and W. F. Hester, ANAL.CHEM,, 23, 1842 (1951). (5) M. L. Shankaranarayana and C. C. Patel, ibid., 33, 1398 (1961). (6) J. M. Chilton, ibid., 25, 1274 (1953). (7) R. J. Jacoste, M. H. Earing, and S . E. Wiberley, ibid., 23, 871 (1951). (8) T. C. J. Ovenston and C. A. Parker, Anal. Chim. Acta, 4, 135 (1950). (9) K. Lu and H. P’o, Hua Hsueh Tung Pao, 11, 55 (1963); CA, 60, 12644 (1964). (10) T. Fuginaga, T. Nagai, and K. Yamashita, Nippon Kagaku Zasshi, 84,506 (1963); CA, 61, 1247 (1964). (11) T. Fuginaga and K. Yamashita, Bull. Chem. SOC.Japan, 37, 989 (1964). 1904
ANALYTICAL CHEMISTRY
Mich. the species present and their absorption properties. Chromatography has been used more for separation than for quantitative determination, and polarography has not become generally employed. Simple titration as an analytical method has been employed only in the case of the water-soluble metal dialkyldithiocarbamates, with a mild oxidizing agent such as iodine (12). The present study explores the feasibility of applying the same titration method to analyze water-insoluble metal diethyldithiocarbamates with chloroform as the solvent. EXPERIMENTAL
Reagent grade chloroform was used without further purification. The iodine was resublimed twice from a pulverized mixture of iodine and potassium iodide. All the metal complexes were prepared by mixing aqueous solutions of sodium diethyldithiocarbamate and a soluble salt containing the desired metal ion. The complexes were recrystallized from chloroform, or from methylene chlorideethanol. Carbon, hydrogen, and sulfur analyses checked with those expected for these metal diethyldithiocarbamates. Visual titrations were performed using a buret with a stopcock made of Teflon (Du Pont). Photometric titrations were performed with either a Beckman D U or a Cary 14 spectrophotometer. (12) A. L. Linoh, ANAL.CHEM., 23,293 (1951).
