V O L U M E 21, NO. 7, J U L Y 1 9 4 9
827
genin effect similar to that described by Sprince and Woolley (14) and Wright and Skeggs (19) for a variety of other bacteria. Differences in the sensitivity of response to pantoic acid have been observed, however, with different cultures. These variations appear to be associated with the time of incubation required for attainment of suitable levels of growth. As previously noted, a decrease in the growth rate has been observed after repeated serial transfers of the culture from slant to slant. The concomitant increase observed in the relative activity of pantothenic acid as compared to pantoic acid appears to indicate that this rffect is due to a decrease in the efficiency of the organism in bynthesixing pantothenic acid from the latter. Kith the present culture carried as directed, the use of a heavy iiioculum (G = 15 to 20), plus the modified medium, has permitted shortening the time of incubation in stationary flasks from 60 to 40 hours. The use of continuous shaking in Evelyn tubes during incubation permits assays in 20 to 24 hours. Invlusion of Tween 80 in bacterial media was reported by Dubos and Davis (4)to enhance the rate and abundance of growth of tubercle bacilli and has been applied extensively with other organisms (6). The stimulatory effect of the lactate is similar to that observed by Tosic ( I @ , who attributed it t o the buffering effect of the bicarbonate formed from the lactate during growth of '4 cetohacter turbidans. SUMMARY
.1 microbiological procedure is described which permits assay of panthenol, the biologically active hydroxy analog of pantothenic acid, in the presence of pantoyl lactone as well as pantoic acid. The lactone is removed quantitatively from water solution by continuous extraction with ethyl ether for 1 to 2 hours. Panthenol in the aqueous residue is then hydrolyzed to pantoic acid for microbiological assay with Acetobacter suboxydans. ilssay before hydrolysis provides a correction for preformed pantoic acid and a partial correction for pantothenic acid. Hence, small percentages of pantothenic acid (up to 10%) will not seriously affect the specificity of the panthenol assay. B y modification of
the assay medium of Sarett and Cheldelin and by setting up the assay in Evelyn tubes with continuous shaking, the incubation time has been shortened to 20 to 24 hours. Good agreement with bioassays in rats has been demonstrated. ACKNOWLEDGMENT
Many of the early exploratory trials were carried out by J. M. Cooperman and J. M. Scheiner. M.Weiss and Miss I. Kellermann performed many of the later microbiological experiments. The rat bioassays were done bv Miss R. Pankopf and I. Korman. LITERATURE CITED
(1) Burlet, E.,2.Vitaminforsch.,14,318(1944). (2) De Ritter, E.,Jahns, F. W., and Rubin, S. H., J . Am. P h u m . Assoc., in press. (3) Drekter, L.,Drucker, R., Pankopf, R., Scheiner, J., and Rubin. S . H., Ibid., 37,498 (1948). (4) Dubos, R. J.,and Davis, B. D., J . Ezptl. Med., 83,409 (1946). (5) F r o s t , D.V.,IND.ENG.CHEM., ANAL.ED.,15,306 (1943). (6) Glassman, H. N.,Bact. Rev., 12,105 (1948). (7) Pfaltz, H., Z . Vitaminforsch.,13,236 (1943). (8) Roberts, E.C., and Snell, E. E., J. Biol. Chem., 163,499 (19461 (9) Rubin, S. H., J . Am. Pharm. Assoc., 37,502 (1948). (10) Rubin, S.H., Cooperman,J. M., Moore, M.E., and Scheiner, J., J . Nutrition, 35,499 (1948). (11) Rubin, S. H., Scheiner, J., and Hirschberg, E., J . BbZ. Chem.. 167,599 (1947). (12) Sarett, H. P.,and Cheldelin, V. H., Ibid., 159,311 (1945). (13) Snell, E.E., and Strong, F. M., IKD.ENG.CHEM.,ANAL.ED.. 11,346(1939). (14) Sprince, H., and Woolley, D. W., J . A m . Chem. Soc., 67, 1734 (1945). (15) Tosic, J.,Biochem. J . , 40,209 (1946). (16) Underkofler, L.9.,Bants, A . C., and Peterson, R'. H., J . Bat.. 45,183 (1943). (17) Walter, M., Jubilee Vol., Emil Barell (Hoffmann-LaRoche. Inc., Basle) 1946,98. (18) Walter, M., 2. Vituminforsch.,18,228 (1947). (19) Wright, L.D., and Skeggs,H. R., J . Bact., 48,117 (1944). RECEIVEDKovember 17, 1948. Presented before the Division of Biologioa Chemistry at the 114th hfeeting of the . ~ ? I E R I c A N C R E I r I C A L SOCIETY Washington. D. C. Piihlication 1.38
Microdetermination of Carbon and Hydrogen FRIEDRICH 0. FISCHER' Chemical Institute, University of Heidelberg, Heidelberg, Termany 4 modification of the proved Pregl micromethod for carbon and hJdrogen eniploys a fully automatic combustion procedure applicable to micro-sized samples of materials of widely varying combustion characteristics. The values obtained are more accurate than the usually accepted * 0 . 3 7 ~absolute. With 10- to 20mg. samples and semiautomatic operation, an accuracy of *0.025%absolute can be attained; this furnishes a method of high accuracy requiring relatively small amounts of sample for such difficult problems as the determination of the cnnstitutinn of high molecular weight compounds.
P
KLGL'b micromethod (10) has undergone development 111 three directions: The sample size has been decreased ( 2 , 9 ) , the combustion has been made automatic ( 4 , 6, 11, 13, 16), and the precision and accuracy have been increased (5,6-8). 111 the absence of a practicable ultramicrobalance the first and third are mutually exclusive; higher accuracy requires larger amount< of sample. B completely automatic combustion procedure rail be recommended fully for routine microdeterminations It considerablj relieves the analyst and gives more uniform results But where higher accuracy is iequired and a larger quantitv of 1
Present address, Schroderstrasse 6, Heidelberg, Germsny.
saiiiple is to be burned, a voiiipletelj. aut oinatic procedure is applicable only when the combustion characteristics of the substance in question have heeii determined. A semiautomatic procedure is useful in other cases. It is less tedious and leads to more accurate results than can be obtained by manual operation. By using a modification of the Pregl microprocedure, an automatic or semiautomatic combustion, and 10- to 20-mg. samples, it has been found possible to obtain values within *0.02%. This increase in sample size is small compared to requirements for macroprocedures (1, 12, 14, 16) or to amounts needed to carry out a number of microdeterminations for the purpose of ohtrtiuing a more accurate average value.
ANALYTICAL CHEMISTRY
828
The connection of the combustion tube with the water tube, the most delicate point of tbe whole apparatus, must be formed in a way to permit the larger amounts of water vapor and carbon dioxide t o pass without loss. Replacing the usual rubber connection with a good ground joint accomplishes this and eliminates those errors introduced by improper lubrication of the rubher and the wiping of the &sorption tubes before each weighing. Tbe absorption tubes are wiped only before the beginning of a combustion series and after that are bandled with chamois gloves and cork plate pinerr*. ! v i m srupprrwi, iiic tunes nre ronstmt iir weight.
._-
Figure 1. Apparatus
. .
APPARATUS
active filling of the combustion and absorption tubes can be attained by increasing the surface and activity of the reagents. Appreciable increase in the quantity of reagents, the dimensions of Pregl's apparatus, or the quantity or velocity of gas flow results in an increase in the magnitude of the inherent errors. Even in the p-ence of copper oxide an excess of oxygen is essential. The new automatic apparatus causes the slow vaporization of the sample far below its boiling point, and thus ensurea excess oxygen. ~~
C I
0 1 * > , 5 * " .
Figure 2.
Platinum Tube
I A
Figure 3. Pressure Regulator
C A C E C
A
C
D
B
C
A
Figure 4. A.
Silrerxool
E.
room and by placing aluminum plates (300 X a10 X 25 mm.) The rider and a11 the weights were in the interior f, the C-S. made of aluminum and the tare bottles were filled with glass beads, These materials were chosen because their densities were nearer the densities of the samples to be weighed than the heavy metals. Only when the density of the sample differed significantly from that of aluminum a.ere vacuum correctjons applied. A significant variation in air density between weighings iS not as likely in micro as in macro work because of the shorter lapse of time. The rider and weights were carefully intercalibrated. combustion Furnace and Constant-Temperature Chamber. The combustion furnace and constant-temperature chamber, separated.by 20 mm. of asbestos, are placed in a porcelain tube 250 mm. ioug and 70 mm. in diameter (Figure 1.) This is mounted on a Pertinrtx box which contains ail the instruments for control of temperature and automatic 'regulation-voltage regulators, contact thermometer relay, and rheostats. A voltmeter, ammeter, time switcli, and microcbronometer are mounted on the front of the box. Depending on the eonstsncy of the current supply the canstrtnt-temperature chamber is contralled by a simpfe voltage regulator or contact thermometer. The constant-temperature chamber is made of aluminum and is 60 mm. long and 35 mm. in diameter. There is a central bore for the combustion tube and a pwrallei bore 8 mm. in diameter for the contact thermometer. The relay does not break the whole circuit but rather shortcircuits a small rheostat when the temperature falls below that required. The constant-temperature chamber and the combustion furnace are connected in series, 60 that the temperature of the combNustion furnace is controlled simultaneously with that of the constant-temperature chamber. Ab'solute temperatures are not so important as constancy of temperature. The constmt-temperature chamber is operated a t about 200" C. and the furnaceat about 700" C. Theconstanttemp ersture chamber should be held constant within 0.5" C. and t,hefurnece within 2 ".C. Tb e combustion tube (Figure 4) is made of Supremsn glass and fitted with polished standard-taper ground joints. The ..:--L plpswshaped enlargement, 100 nun. long and 22 mm. in diameter, provides space for the sample vapors to mix well with oxygen before entering the oombustion zone proper and for substances that detonate. The combustion tube is held firmly by clamps (Figure l), so it
Combustion Tube
Copper orids C. Aabeatos D .
Lesxdchmmale E.
leaddioxide
829
V O L U M E 21, NO. 7, J U L Y 1 9 4 9
This is filled up with quartz uool and then introduced, inverted, into the platinum tube. Hygroscopic samples are weighed in bottles with ground-in stoppers, preferably made of quartz, which are introduced directly into the platinum cylinder. 0
I
2
3
1
5 c n
Filling. The combustion tube filling (Figure 4) consists of the reagents used by Pregl (IO)but they are much more finely divided and have more effective surfaces. Figure 5.
Absorption Tubes
cannot be displaced by removal of the stopper or by the steel spring that holds the absorption tubes in place. A platinum tube (Figure 2) 7 mm. in diameter and 30 mm. long is fastened to the stopper by a platinum wire 170 mm. long and 1.5 mm. in diameter. The platinum boat or glass capillary is placed in this tube. A window, the size of the platinum boat, is cut in the tube, so the sample can be watched during heating. Seven circular platinum plates, fastened to the platinum wire normal to its axis, divide the back of the combustion tube into seven chambers and prevent the back-diffusion of vapors. Starting at the furnace and extending 220 mm. toward the sto per (Figure 1) the combustion tube is wound with a chromeniciel heating tape in such a way as to a l l o ~good observation of the sample between the windings. The windings are held in place on the underside of the combustion tube by an asbestospotassium silicate putty. This wound section is surrounded by a Supremav tube 220 mm. long and 24 mm. in inside diameter. Automatic Vaporization Apparatus. The automatic vaporization apparatus consists of the pressure regulator (Figure 3) and a circular rheostat driven by an electric motor through a reduction gear. The manometer of the pressure regulator 1s filled with slightly acidified water and the U-tube with phosphorus pentoxide on pumice, asbestos, and Ascarite. A and B are platinum contacts. C is a steel wire for narrowing the capillary and thus adjusting the sensibility. The rheostat is connected into the circuit of the heating tape. The motor turns the rheostat slowly, decreasing the resistance, thus causing a constantly rising temperature within the vaporization zone. The rate of increase can be regulated further by controlling the speed of the motor with a potentiometer connected into the circuit of the armature coil. A commutator, connected in like manner, permits reversal of the direction of the motor should the sample vaporize too rapidly. The regulation of the speed of the motor and the proportion of the reduction gear should be chosen so that the resistance is cleared in 3 minutes a t high speed and in 2 hours a t slow speed. The rise in temperatyre in the vaporization zone should be about 6 " C. per minute a t low speed. The rheostat is wound with resistance wire of different diameters in such a way as to obtain a linear rise of temperature. d circular rheostat without stops was chosen because it returns to its initial position. A pointer mounted on the axis of the rheostat indicates on a scale the temperature within the vaporization zone a t any position of the rheostat. The vaporization unit can be used in an entirely automatic manner for routine microdeterminations or for precision determinations where the combustion characteristics of the given material are known. When the vaporization is too rapid the pressure regulator acting through a relay turns off the heating system, and, through the commutator, reverses the motor so that the resistance is increased. It remains thus until the excess pressure has been dissipated. If the pressure regulator is adjusted to a high sensibility, this play continues for some time and until the sample is slowly vaporized and burned. For the precision determination where the combustion characteristics of the material are not known the apparatus is used in a semiautomatic manner by varying the speed of the motor as needed. The art of combustion consists in burning the sample very slowly, allowing it to vaporize far below its boiling or sublimation point. When the pressure in the pressure regulator has considerably decreased, the motor speed is increased, and the vaporization zone brought to red heat. Liquids are weighed in capillaries of larger than usual size. When the air bubble in the half-filled capillary is heated, the liquid is expelled below its boiling point. In order to obtain a gradual mixing of the vapors with oxygen, the capillary is fist put into a quartz tube 35 mm. long and 4 mm. in inside diameter.
Copper oxide is formed in the combustion tube. Pure electrolytic copper wire 0.05 mm. in diameter is cut into pieces 5 mm. long. It is washed with benzene, alcohol, and ether and, after drying, is stuffed into the combustion tube compactly. I t s complete oxidation takes 12 to 15 hours. The resistance to gas flow is not seriously increased by this transformation. It is not advisable t o carry out the oxidation in a separate tube and transfer the copper oxide because the thin pins break too easily. Silver is used in the form of a very fine wool, and the lead chromate is used in h e grains. The increase in surface is attended by an increase in the hygroscopicity of the filling. For this reason, not only the lend dioxide hut also the rest of the filling must be held a t c o n i t m t temperature; the combustion tube must be closed immediately after each analysis. Exact hydrogen values can be obtained in the presence of lead dioxide if the temperature and the amount and velocity of the gases passed through it are held constant, and if materials of widely different hydrogen content are not burned consecutively. Absorption Tubes. The absorption tubes used by Pregl (IO) vere modified slightly in dimensions and provided with ground joints (Figure 5 ) . The joints must not be lubricated, but are ground in with fine emery powder and then polished n-ith rouge in turpentine. (All ground joints on the apparatus are treat,ed in this manner.) The Tater tube is made preferably of Supremax glass, so the parts of the joint connecting the water and combustion tubes will have the same coefficient of expansion. The .filling is phosphorus pentoxide on pumice. The carbon dioxide tube is filled Rith Ascarite followed by a short layer of phosphorus pentoxide on pumice, separated by asbestos. The absorption tubes are held together and to the combustion tube bv a single steel spring (Figure 1). To pre"vent cgndensation of the larger amounts of moisture, a semicylindrical heater 40 mm. long, cut from a copper tube 7 mm. in diameter, extends from the constant-temperature chamber under the joint and capillaries of the water tube (Figure 1). If the absorption tubes have capillaries of standard length and not more than 0.2 mm. in inside diameter, and if they are capped (stoppered), they are absolutelv constant in weight. Marioke Bottle. A modified hlariotte bottle is shoxm in Figure 6. The 200-cc measuring cylinders are arranged one above the other (Figure 1) on a rod which cp.n be rotated. By simply changing over the rubber tubing and inverting, it is ready for the next analysi-.
i! :eo
-
= 60
10 PO
0
2
4
6
a
10
c-.
Figure 6. Modified Mariotte Bottle
To obtain accurate hydrogen values, especially when the humidity is high, it is necessary to operate the air supply separate from the oxygen and t o pass it through a large drying tower containing Ascarite and phosphorus pentoxide on pumice. The air passes too slowly through the capillary of the pressure regulator (Figure 3). A rapid flow of dry air must stream from the comhustion tube when it is opened to
ANALYTICAL CHEMISTRY
830
______Table I.
Benzil
Balioylic acid
Acetanilide
p-Nitrobiphenyl
c
-910.
Mg.
Mg.
10.593 11.235 15.318 14.215
5,757 6.086 8.312 7.717
36,460 38.665 52.721 48.928 Av.
93.928 93.914 93.925 93.930 93.924
Theory U G H % 70 7c 6.082 9’3.935 fi.06.5 6.062 6.073 6.075 6.073
79.970 79.976 79.990 80.005 79.985
4.820 79.983 4.815 4.812 4.811 4.814
HzO
COS
C
5%
17.863 17.431 14.320 15.195
7.694 7.500 6.158 fi.532
52.347 51.084 41.975 44.548 Av.
14.993 14.150 19.754 15.851
5.842 5.573 7.692 6.207
33.437 60.860 4.361 60.888 4.379 31.575 60.893 4.408 44.046 60.847 4.358 35.340 60.841 4.382 Av. 60.860 4.377
12.133 13.137 12.704 13.120
7.265 7.878 7.630 7.890
31.612 34.215 33.088 34.172 Av.
71.100 71.073 71.075 71.077 71.081
6.701 71.085 6.713 6.711 6.721 6.730 6.716
12.666 15.235 16.194 14.718
5.176 6.219 6.601 fi n12
33,584 40,386 42.936 WOZR
72.357 72.340 72.352 72.357 72.351
4.562
.Av.
4.573 72.349 4.568
4 795
Deviatiun C H
7c
R
-0.007 -0.021
f0.017 -0.003
1 : : ;
$:E:$’?,
-o,oll
+o,oo8
-0.013 -0,007
+0.020
+0,025
-0.008
-0.018 +0,oz9
-0,021 -o,027
+0,003 -0.021
-o,oo8
-o,oo2
f0.01.5 -0,012
-0
1 ::;
$: :’$ +0,003
+0.025
$+0,002 :E;: $::gig +O , o l ~
-o,oo.l
-0.012 OO2
$:!-E::
1 . 5 5 5 -kE ::
4.571 L56R
12,010. H = 1,008.
~____
.
Mg.
Benzene
Salicylic acid
Acetanilide
~~
Iritomatic Inalyses of Micro Samples
Samyle
H2O Mg.
2.94 4.550 3.14 5.124 3.57 5.655 3.93 5.215
5.944 4.593 4.458 3.976 4.644 5.271
‘.05 2.37 1.81 1.77
2.39 2.78 3.19 .5.629 3.39
COS
C
Mg. 14.31
‘2
H C7c
Theory ’ H
C %
70
92.21 7.77 42.2*i 15.37 92.18 7.72 17.33 92.30 7.79 19.12 92.27 7.78 Av. 92.24 7.77
7.75
60.87
-(.‘w
11.64
60.91 60.88
4.40
13.26 10.24 9.95 Av.
4.46 60.84 4.41 80.91 4.44 60.89 4.43
10.35 12.10 13.74 14.67 Av.
71.04 71.10 71.14 71.12 71.10
Deviatiuri H
C
R
%
1;::; f0.05 f0.02 -0.01
6.73 71.09 ti.71 6.70 6.77 6.74 6.71
oue hour and serial determinatioiis t a kr 45 to 50 minutes. Precision analytical results for fivr urganic compounds are given in Table 1. Routine results on micro samples whrrr fully automatic operation was employed are given in Table 11. Even substance which are very volatile, or sublime, or detonate can be burned automatically. A great number of carbon and hydrogen microdeterminations in various types of organic compounds were performrd manually and aut,omatically. The results obtained by the automatic procrdure were always as accurate as thosr obtained manually and usually wrrr more accurate. In the case of vrr!. volatile materials and those which sukilime, the automatic procedure described above proved much superior to manuill operation or the type of automatic combust,jon employing a small motor-driverl
f 0 . 0 0 3 +O.OOi C0.008 +0.016 + n nn2 +n 01.1
=
Table 11.
__
~
Analytical Results on Five Compounds
Sample Fluorene
~
+0.04
f0.03
+0.02
$: T-:;! $:: f0.02
+0.05
-0.05 +0.01 f0.05 +0.03 fO.O1
+0.02 -0.01 +0.06
+0.03 +0.03
- -
~
The automatic procedure described i r i this paper differs from many others so designated in t,hat the heating rate is -controlled directly by the combustion characteristics of the sample rather than by a prearranged schedule set by an operator. The speed of the motor is so adjusted that full temperature will be reached in 15 minutes. If this is too fast the pressure regulator interrupts the heating for a short time. In the precision procedurr where larger samples are burned the method brcomes semiautomatic in that the operator adjusts the speed of the motor, and thus the rate of increase of temperature, according to t,he volatility of the sample. The automatic features make the work of the operator easier and, more important, the obtained are nloreaccurate, LITERATURE CITED
( I ) B a x t e r , G . P.. a n d Hale. A . H . , J . .4m. (‘hern. Sor.. Picric acid 4.345 0.53 5.01 31.47 1.37 :31 4 5 1.32 4-0.02 f 0 . 0 5 58,510 (1936). +0.05 +0.03 6.810 0.82 7.86 31.50 1.38 ( 2 ) Berraz, G., Anales S O C .Cidnt. -4rgentina. Secciriri -0.08 +0.07 5.27 31.37 1.39 4.585 0.57 Santa F6, 9, 9 (1937). f0.04 fO.08 8.223 0 . 7 8 7.18 31.49 1.40 $ 3 ) Boetius, M., “Uber die Fehlerquelleii der mikrofO.O1 +o.ofi .4v. 31.46. 1.38 analytischen Bestinimung von C u n d H nach der Methode von Pregl,” Berlin, S‘erlag Chemie. Pyridine perchlorate ti ,344 1.95 7.78 33.47 3.44 83.41 3.37 f 0 . 0 3 + 0 . 0 7 -0.04 10.05 5.557 1.70 6 . 8 0 33.40 3.42 1931. +0.02 + 0 . 0 2 7.292 2.21 8.94 33.46 3 39 (4) C l a r k , R. O., a n d Stillson, G. H . , I s n . F;sr?. +0.03 +0.08 6.152 1.59 6.32 33.47 3 . 4 5 CHEM.,ANAL.ED.,1 9 , 4 2 3 (1947). 33.45 3.42 fO.O1 fO.05 ( 5 ) Hallett, L . T . , Ibid., 10, 101 (1938). (6) Kirner, W. R., Zbid., 10, 342 (1938). (7) Lieb, H . , and Soltys, A., Mikrochemie, 20. 59 (1936). (8) Lindrler, J., Ber., 59, 2561, 2806 (1926): 60, 124 (1927); 63, prevent introduction of moist room ar. (One cubic centimeter 949, 1123, 1396, 1672 (1930); 64, 1560 (1931): 65, 1696 of air at 22“ C. and a relative humidity of 70y0 contains 0.014 (1932); 67, 1652 (1934); 76, 701 (1943); Mikrochemie. 34, mg. of water.) 67 (1948). If oxygen and air are not available in requisite purity, a pre(9) Niederl, J. B., a n d M e a d o w , J. R., Mikrochemie, 9, 350 (1931) (10) Pregl-Roth, “ Q u a n t i t a t i v e organische bfikroanalyse,” Wien. heater must be used. The pressure conditions described by Pregl ( I O ) are essential for Springer Verlag, 1947; Pregl, F., “ Q u a n t i t a t i v e Organic Microanalysis,” Philadelphia, P. Blakiston’s Sons, 1937. the proper Operation Of the pressure regu1ator Of the automatic (11) Reihlen, H., a n d m e i n b r e n n e r , E,, Chem, Fabrik, 7, 63 (1934): combustion unit. To prevent loss a t the connectiom slight excess Mikrochemie, 23,285 (1937 /38). (12) Renoll, M. W., Midgley, Thomas, Jr., and Henne, A. L.. IND suction is applied during the combustion. The air is delivered at higher pressure (greater speed) but i t is not necessary to reESG.CHEM.,AXAL.ED.,9, 566 (1937). (13) Royer, G. L., N o r t o n , A R., a n d Sundberg, 0. E., Ibid.. 12. adjust the Mariotte bottle because by this time the concentrations 688 (1940). of water and carbon dioxide in the gas stream are very small. (14) Tunnicliff. D.D., P e t e r s , E. D., Lykken, Louis, and Tuemmler, F. D., Ibid., 18, 710 (1946). The combustion proper takes 15 to 20 minutes. The rate of (15) K a g m a n , D. D., a n d Rossini, F. D., J . Research l’atl. BUT.. Air is delivered at a rate of 8 oxygen flow is 5 cc. per minute. Standards, 32,9 5 (1944). to 10 cc. per minute and 140 to 150 CC. are used. Pregl’s weighing (16) Zimmermann. W.. Mikrochemie. 31, 149 (1943) procedure ( I O ) is followed. \ qingb determination takes ahout R ~ ~ Frbniarl ~ ~ 3 ~1948) ~
