tion of apparent molecular weight with molality. This appears t o be the case with the petroleum distillate fractions analyzed using octadecane in benzene reference solution (Table 111). When deviations from ideality occur, more reliable determinations can be made by determining the apparent molecular weights a t several finite concentrations and extrapolating to zero concentration, as was done for azobenzene and trilaurin (Figures 8 and 9). As more than one concentration of standard reference solution can be used with the same test material, a new sample does not have to be weighed and transferred. The test sample must dissolve completely in the solvent employed. I n general, solvents in which the sample shows the greatest solubility should exhibit the least deviation from ideality. The solvent employed must have a vapor pressure low enough to make evaporation losses insignificant during handling. High vapor pressures are desirable for a more rapid equilibration. Molecular weight standards above molecular weight 400 were unavailable; therefore, the method n-as evaluated by comparing molecular weight values obtained by isothermal distillation and twin thermistor ebullioscopic methods. The mean molecular weights obtained by the two procedures agreed within 2% relative (Table 111). Under the conditions employed, the method is applicable to materials that boil above 275’ C.; otherwise, the vapor pressure is sufficient to cause sample loss. Loss of benzoic acid, boiling point 249’ C., was observed by a progressive decrease in weight on the helix pan after the system was equilibrated. KO loss of azobenzene, boiling point 293’ C., occurred even after it was left in the system for pro-
longed periods. Unknowns can be checked for volatility in the same manner by observing whether or not the equilibrated unknown and solvent gets progressively lighter over a few days a t equilibrium. Because straight-line plots were obtained in Figures 5, 6, and 7, molecular weights were calculated from data taken before equilibrium was reached, using the expression
where a = weight gain a t equilibrium, mg.
m = slope, hours-’ t = time, hours w = weight gain a t time t, mg. Both a and m are unknown and are obtained graphically from the intersection of a plot of a us. m values and a second set of values upon substitution of t3 and w 3 for tz and w2. This method of calculation of molecular weight from data taken prior to equilibration of the system has been applied to the data shown in Figures 5, 6, and 7. The average relative error of seven determinations was 8.5%. With average values of w and t taken from a mean curve drawn through a plot of w us. t values, this error is expected to be less than 5%. This calculation method has the advantage that values can be obtained in one working day instead of 3 days; however it is less accurate and requires more operator time. CONCLUSIONS
The results obtained in this investigation indicate that molecular weights can be determined by the proposed procedure with a fraction of a milligram of test material. An accuracy to 3% is attainable. The temperature of the
equilibrium system must be controlled for prolonged periods to =k0.005° C. The method is inapplicable to materials boiling below 275’ C. Three days of elapsed time are required for a determination. The operator time per single determination, however, is only 2 hours. ACKNOWLEDGMENT
The author is grateful to Max Blunier for proposing use of the helical spring balance in the isothermal distillation method and for conducting initial experiments which indicated the feasibility of the method. LITERATURE CITED
Barger, G., J . Chem. SOC.85, 286 (1904). Childs, C. E., ANAL. CHEM. 26, 1963-4 (1954). Clark, E. P., IND.ENG. CHEM., ANAL.ED.13,820-1 (1941). Hover. Herbert. Mikrochemie 36/37, i169-73 (1951). McBain, J. W., Bakr, A. &I.,J . Am. Chem. SOC.48,690-5 (1926). Madorsky, L., Rev. Sei. Instr. 21, 393-4 (1950). Nash. L. K., ANAL. CHEM. 19, 799-802 (1947). Niederl, J. B., Kasanof, D. R., Risch, G. K., Rao, D. S., Mikrochemie 34, 132-41 (1948). I Parrette, R. L., J . Polymer Sei. 15,447-58 (1955). (10) Puddington, I. E., Can. J . Research 27B, 151-7 (1949). (11) Rast, Karl, Ber. 3727-8 (1922). (12) Signer, R., Ann. Chem., 478, 246-66 (1930). (13) Taylor, G. B., Hall, M. B., ANAL. CHEM.23, 947-9 (1951). RECEIVEDfor review May 27, 1957. Accepted September 5, 1957. Division of Analytical Chemistry, Symposium on Analytical Contributions to Research in Petroleum Geochemistry, 131st Meeting, ACS, Miami, Fla., Bpril 1957. Publication 111, Shell Development Co., Exploration and Production Research Division. I
Multipurpose Standard for Microchemical Analysis W. H. SMITH National Bureau of Standards, Washingfon 25;
b As a standard for use in ultimate analysis of organic compounds, 5chloro 4 hydroxy 3 methoxybenzylisothiourea phosphate is suggested.
- -
I
- -
1946 Ogg and Willets (3) proposed benzylisothiourea hydrochloride as a “new standard for use in ultimate analysis of organic compounds especially suited for microprocedures.” It has N
D. C.
been extensively used by microchemists. Benzylisothiourea hydrochloride contains carbon, hydrogen, nitrogen, chlorine, and sulfur. A compound, 5 - chloro - 4 hydroxy - 3 - methoxybenzylisothiourea phosphate, which is related in structure and contains also phosphorus, oxygen, and the methoxyl group, is proposed as an alternative standard. This compound contains all the elements in the seven microchemi-
-
cal standards now issued by the National Bureau of Standards, except iodine. “/2
CH2. S . C
0
C1
“€I. H3P0d
OCHB OH
C9HI4C1S2OeSP,molecular weight -344.73 VOL. 30, NO. 1, JANUARY 1958
149
"70 Carbon Hydrogen Chlorine hTitrogen Oxygen Sulfur Phosphorus
31.35 4.09 10.29 8.13 27.85 0.30 8.99 100.00
llethoxyl
9.00
PREPARATION
The following procedures are suggested because they are readily conducted and do not require elaborate equipment. Vanillin is chlorinated by sodium hypochlorite to form 5-chlorovanillin. The aldehyde group is reduced by sodium borohydride and the resulting alcohol is condensed in one step with thiourea and phosphoric acid.
temperature, and filter to remove resinous material. Acidify the filtrate with about 55 nil. of 85% phosphoric acid, and dilute with 2 volumes of water. Separate the crystalline needles and wish them with water until free from acid. The yield is about C J O ~ ~Crystal. lize from hot water with the addition of adsorbent carbon. Afterward crystallize twice from benzene to remove residual resin. The capillary melting point of a product prepared by the procedure Fvas 118.5" c. 5-CHLORO-4-HYDROXY-3METHOXYBENZYLISOTHIOUREA PH0S PHATE
Heat the following mixture to the boiling point under a reflux condenser. 5-Chlorovanillyl alcohol, g. Thiourea,. g. Phosphoric acid, 85%, ml. Ethyl alcohol, 99.5Tc,, ml.
47 21 17
400
5-CHLOROVANILLIN
-4 method for the chlorination of vanillin by aqueous sodium hypochlorite has been developed by Hopkins and Chisholm ( 9 ) , Mix 60.8 grams of vanillin and a solution of 16 grams of sodium hydroxide in 500 ml. of water. Stir until the vanillin has been dissolved. Establish a temperature of about 20" C. and maintain it during addition of 28.4 grams of chlorine available in the form of sodium hypochlorite dissolved in 1200 ml. of water. Stir at intervals. When a negative test is obtained with starch-iodide paper, acidify with a mixture of equal volumes of concentrated hydrochloric acid and water. Separate the precipitate of crystalline plates and wash it with water until free from acid. The minimum yield is 75%. Some resin is formed during the reaction. Purify by crystallization from a mixture of boiling acetone and water and afterward, with the addition of adsorbent carbon, from a mixture of boiling ethyl alcohol and water. Dry the product and recrystallize it from benzene. The capillary melting point of a preparation made by this procedure was 165" C. This value agrees with that obtained by Hann ( I ) , mho prepared this compound by passing chlorine into a solution of vanillin in glacial acetic acid in the presence of fused sodium acetate. 5-CHLOROVANILLYL ALCOHOL
Mix 45 grams of 5-chlorovanillin with a solution of 45 grams of sodium hydroxide in 1 liter of distilled water. Stir until the vanillin is in solution and heat to 80" C. Add 3 grams of sodium borohydride in several successive portions. Stir occasionally, allow to cool to room 150
ANALYTICAL CHEMISTRY
After about 6 hours a portion of the phosphate separates. Allow the contents of the container t o cool to room temperature and separate the crystalline material by filtration. Reflux the filtrate as before and again remove the product which separates. When the filtrate is again heated, a third portion will appear after the mixture is allowed to cool. The minimum yield is approximately 60% of the theoretical value. If the reaction products are not removed as described. the mixture may bump violently.
2 3 4
5
9.14
9.06 9.00 9.09 9.09 8.95 8.98 8.92
LITERATURE CITED
Hann, R. bl., J. Am. Chem. SOC.47, 2000 (1925).
RECEIVEDfor review March 13, 1957. rlccepted August 8, 1957.
(iinalyses by Rolf Paulson and Lorna Tregoning) Chlw Phosrine, phorus, Sulfur, Samplesa 92 70 % 10.28 10.40 10.30 10.31 10.37 10.18 10.14 10.10 10.27
The phosphate melts with decornposition a t about 175". Khen the dried salt is exposed to a humid atmosphere it absorbs moisture. In 24 hours a sample of about 5 grams in a layer about 1 cm. deep gained 0.06% in neight in an atmosphere of 50% relative humidity. Under similar conditions, in an atmosphere of SOY0 relative humidity, the gain in weight was 0.457,. It is therefore necessary to dry the phosphate to constant weight, preferably a t 105" in vacuo, before it is used as a standard. Table I indicates the consistency in coniposition that may be expected by using these methods of preparation and purification.
Hopkins, C. Y., Chisholm, M. J., Can. J . Research 24B,208 (1946). Ogg, C. L., Willets, C. O., IND.ENG. CHEM.,ANAL.ED. 18, 334 (1946).
Table 1. Microanalytical Determination of Chlorine, Phosphorus, and Sulfur in Standard Compound
1
water. This amount of solvent does not completely dissolve the crystals but is sufficient to effect purification by digestion. Heat at 50" for 1 hour, allow t o cool to room temperature, and filter. Remove the phosphate, place it in a beaker, cover it with water, and heat t o 50" for 1 hour as before. Then add at least 6 volumes of a mixture of equal volumes of acctone and ethyl alcohol to precipitate the phosphate. Separate by filtration through a sintered-glass funnel. Repeat the treatment with water, followed by precipitation. Dry the final product in vacuo a t 105" and grind to a powder in an agate mortar.
9.38 9.39 9.33 9.39 9.34 9.43 9.36 9.40
Theoretical values 10.29 8.99 9.30 I, 2, 4, and 5 . Separate preparations. 3. From same preparation as sample 2, but after storage in laboratory for about 9 months.
To purify the product; transfer it to a beaker and cover it with a mixture of 9 volumes of acetone and 1 volume of
Quality Control of Pharmaceuticals. Applications of Quantitative Paper Chromatography in Conjunction with Instrumental Methods-Correction I n the paper on ('Quality Control of Pharmaceuticals. Applications of Quantitative Paper Chromatography in Conjunction with Instrumental Methods" NIL. CHEM.49, 1649 (1957)] three errors occurred. Page 1650, column 1, 3rd paragraph, 8th line, should be deleted. Page 1651, Figure 2 , "pH 5/2" should read "pH 5.2." Page 1661, column 2, third line below Figure 2 should read "retaining the fourth as a blank." H. J. PAZDERA W.H. MCXULLEN
