V O L U M E 23, NO. 2, F E B R U A R Y 1951
361
balance are desirable, because they reach equilibrium much more quickly than do larger samples. The use of the balance in such a determination is best shown by an example. Figure 8 is the well-known graph of the va or pressure plotted against water of hydration of the various hylrates of magnesium sulfate. To illustrate a use of the balance this hydrate system n-as studied in the following way.
I u)
z 4
550
-
500
-
corresponding to the monohydrate. In like manner a t times C and D baths of partially frozen dioxane and aceto henone, having temperatures of 8.6" and 16.6" C., respectivJy, were placed around the trap. Breaks corresponding to the formation of di- and tetrahydrates were obtained. With a salt of undetermined hydrate composition, a larger number of trap temperatures would be necessary for a complete hydrate study. To extend the temperature range, the system, excluding the trap, can be fitted with jackets through which a heated bath liquid may be circulatrd. Determination of Gas Density. The balance may be used in the determination of gas densities by mounting a buoyancy bulb on one end of the beam. Reactions in Closed Systems. Reactions may be carried out on the pans of the microbalance in closed systems under wide conditions of temperature and pressure and monitored by following the accompanying weight changes. Any gaseous reagent tolerated by the all-quartz construction may be used.
TETRA HYDRATE)^
a
450 LITERATURE CITED
C (MONO-HYDRATE)
400
350
I
\ L j B (ANHYDROUS STATE) .
1
~
2
I
I
~
3 4 5 6 7 TIME IN HOURS
I
8
I
I
I
9
Figure 9. Experimental Curve of Waters of Hydration of Magnesium Sulfate
&-I sample of partially hydrated magnesium sulfate 1% as placed on one pan of the microbalance. Thr balance was inserted into a closed system, the temperature of which was fixed near room temperature. A trap containing water was in the system. By using a propriate baths, the temperature of this trap was fixed a t vakes lower than the temperature of the rest of the system. The vapor pressure of watrr in the closed system was thus estahlished by the temperature of the coldest part of the system, which was the trap. Changing the temperature of the trap changed the vapor pressure of water in the system. Changcls in the weight of the sample could he interprrted in terms of hydrate formation, as shown in Figure 9. I t time A on the curve a dry ice-acetone bath was placed around the trap, and thr systrm was evacuated and closed. --I rapid 108s of weight rcw~lted until the anhydrous state was reached. A t time H a n ice bath a t a tmiperature of 0 " C was placed around the trap, producing a water vapor pressure of 4 6 mm. inside the system Thi, weight of thts sample increased by an amount
Angstrom, "Oebversigt Kongl. Vetonskaps," Stockholm, Akadamie forhandlinger,Femtion deandra hrghnden S 643,1895 Corwin, A.,IND.ENG.CHEM., . ~ N A L .ED.,16,258(1944). Emich, F., in Abderhalden,"Handbuch der biologischen Arbeitamethoden," Abteil I, Teil 3, p. 183, Berlin, Urban am1 Schwareenberg, 1921. Gorbach, G., Mikrochemie, 20,254(1936). .Johns, I. B., hfonsanto Chemical Co., Dayton, Ohio, unpuhlished work. Kirk, P., Craig, R., Gullberg, J., and Boyer, R., ANAL.C H E V . , 19,427(1947). Lindner, J., Mikrochemie, 34,74 (1948). Petterson, H., "New Microbalance and Its Use," dissertation, Stockholm, University of Goteborg, 1914. Steele, B., and Grant, K., Proc. Roy. Soc., London, A82, 580 (1909). Stock, A., "Hydrides of Boron and Silicon," p. 185 et seq., George Fisher Baker Non-Resident Lectureship in Chemistry, Cornell University Press, 1933. Stock, A,, 2.phys. Chem., 139,47(1928). Strong, J., "Procedures in Experimental Physics," Chap. \-, New York, Prentice-Hall, Inc., 1943. Wiesenberger, E., Mikrochemie, 4,10 (1931). RECEIVED August 22, 1949. Presented before the Division of Analytical Chemistry a t the 118th Meeting of the AMERICAN CHEMICALQOCIETI. Chicago, Ill. Contribution 67 from the Institute for Atomic Research, Iowa State College. Work performed under Contract W-7405-eng 82, Manhattan District, U. S.Corps of Engineers.
low Temperature Separation of Ethane from Methane and Air RORERT L. SLOBOD, The Atlantic Refining Co., Dallas, Tex.
I
X T H E literature ( 2 , 3, 5 , 7 ) the separation of ethane from air or other more volatile suhstances is generally assumed to be accomplished in a satisfactory manner by the use of liquid air or liquid nitrogen. 1,aboratory tests using high vacuum apparatus similar to that described by Prescott and Morrison (6) show, however, that the use of liquid air as a refrigerant results in a considerable loss of ethane and that even a t the colder temperature produced by liquid nitrogen ethane is not quantitatively retained in the cold trap. The vapor pressure of ethane, however, is sufficiently lor$ered when the temperature is reduced from the liquid nitrogen boiling point ( - 196 ' C.) to -210" C. to enable quantitative separation to be made. Temperatures below - 196' C. can be produced by maintaining liquid nitrogen a t reduced pressure, the lower limit (-210" C.) corresponding to the triple point where the system has a vapor pressure
of 96.4 mm. (4). Yo attempt was made to obtain temperatures below -210" C. because thr formation of solid nitrogen complicates the procedure below this point. It is obvious, therefore, that maintaining liquid nitrogen a t a reduced pressure to produce a temperature approaching -210" C. offers a possible means for the quantitative condensation of ethanr. EXPERIMEhTAL
The inability of liquid nitrogen a t atmospheric pressure (-195.8" C.) to lower the vapor pressure of ethane to an insignificant value is drmonstrated by the following experiment, A knowi amount of pure dry ethane, approximately 25 cu mm., was confined in a glass trap, which was part of an elaborate high vacuum system in which mercury cutoffs were used in place
362
ANALYTICAL CHEMISTR ~
In geochemical methods of prospecting for oil, gas samples containing methane, ethane and heavier hydrocarbons, and air are frequently obtained in which the ethane and heavier hydrocarbon fraction must be determined. A method has been developed using liquid nitrogen a t reduced pressure to separate the ethane quantitatively from the methane and air. In order to carry out this procedure conveniently, a special trap has been designed incorporating a low temperature reservoir containing liquid nitrogen
of stopcocks to avoid contamination or loss in the handling of small hydrocarbon samples. While the ethane sample was maintained a t approximately -196" C. by a liquid nitrogen bath, the sample was exposed directly to the pumps. Under these conditions in which the pressure over the ethane was reduced to a fraction of a micron by the mercury diffusion pump, ethane was lost a t an appreciable rate, amounting to approximately 14y0 of the sample in the first 10 minutes of pumping. Continued pumping over a 25-minute interval showed this loss to be approximately linear with time, as shown in Figure 1.
maintained a t reduced pressure. This low temperature source cools a container of liquid nitrogen open to the atmosphere for easy accessibility. A broader application in the field of low pressure gas analysis is suggested by the data obtained showing an inwmplete recovery of pure ethane from the gas phase when liquid nitrogen at atmospheric pressure is used as the refrigerant. Consequently, liquid nitrogen at reduced pressure is suggested for any separation of ethane where complete recovery is required.
The central tube, A , is filled with liquid nitrogen, B , which acts as a bath material and has the advantage of being a t atmospheric pressure. A4dditionalliquid nitrogen, C, is placed in the annular space between ii and the Dewar flask, D. The annular space i s sealed by a cork, E , providing a closed chamber which is evacuated through line F. Upon evacuation through F , the temperature of liquid nitrogen, C, is reduced to approximately -208" C. when a vacuum of about 25 to 27 inches of mercury is maintained. This large bath of cold liquid nitrogen cools the liquid nitrogen] B , in tube A to approximately -208" C. .4n important feature in this design, leading to great convenience in the experimental work, is the fact that the cooling bath, R, is maintained at atmospheric pressure, but a t a temperature lower than can ordinarily be obtained with liquid nitrogen a t this pressure.
Table I.
Quantitative Retention of Ethane a t -208" C.
Vol. of Pure Ethane a t Start Cu. mm. 3.7 3.7 3.7
Time of Pumping with Diffusion Pump Min.
5.8 5.8 5.8
Vol. of Ethane after Pumping Cu. mm. 3.7 3.7 3.7 5.8 5.8 5.8
Result S O
No NO No No No
losa loss loss loss losa losa
DISCUSSION
I
I
I
1
I
The problem of quantitatively separating ethane by condensation from more volatile constituents is complicated by the fact that ethane has a small but significant vapor pressure at the temperature of liquid nitrogen ( -195.8" C.). The equation ( 1 ) for the vapor pressure of ethane as a function of temperature is
log P = -1050.8/T
I
/' 10 15 20 EXPOSURE TO PUMPIN MIN.
25
30
Figure 1. Loss of Ethane a t Liquid Nitrogen Temperature ( -196" C.) during Pumping
The ability to retain ethane quantitatively in a cold trap as a result of a lowered vapor pressure through the use of a lower temperature is demonstrated by the results of the following experiment.
A small volume (3.7 cu. mm.) of pure dry ethane was transferred to the cold trap and cooled to approximately -208" C. by using liquid nitrogen a t reduced pressure as the refrigerant. This sample was then exposed to the pumps fgr three 5-minute intervals under pumping conditions similar to the previous test.
+ 1.75 log T - 0.0134T + 7.102
where P is in millimeters of mercury, T is absolute temperature in degrees Kelvin, and the log is to the base 10. Assuming this equation to apply over the temperature range of interest, the vapor pressure at -196" C. ( 7 7 " E(.)is calculated to be 5.4 X
vAc'-a ,1I T?, FREEZING TUBE FILLED WITH SUPERCOOLED N I F
ATM
-
LIQUID NITROGEN INTRODUCED HERE
CORK
E D
The quantitative retention of the ethane is shown by Table I, which also includes additional data confirming this result. APPARATUS
The apparatus used to obtain the desired temperature of approximately -208' C. is shown in Figure 2.
Figure 2.
Apparatus for Producing Low Temperature Using liquid nitrogen at reduced pressure
V O L U M E 2 3 , NO. 2, F E B R U A R Y 1 9 5 1
363
10-2p. This value, although low, is significant in a vacuum system where the total pressure may easily be only one fifth of this number. Lowering the temperature of the freezing trap, therefore, is the obvious solution to prevent loss of ethane. From the above equation, the calculated vapor pressure of ethane a t -210" C. (63"K.) is 5.3 X 10-5p, which is only 0.001 of the vapor pressure a t -196" C. Experimental verification of this significant lowering of the vapor pressure is given in Table I, where the data shoiv that no loss of ethane mas observed when the condensed hydrocarbon was exposed to the pumps a t this low temperature obtained by maintaining liquid nitrogen a t a reduced pressure of 25 to 27 inches of mercury. This method (cooling a t -210" C.) has been used to recover ethane quantitatively from air mixtures containing small amounts of methane.
reported, but the amount was comparable with the experimental error of the determination. It is conceivable t'hat in the presence of large amounts of methane more of this gas niight be present in the condensed fraction with the ethane. Such contamination could be reduced by vaporizing the sample and then recondensing the ethane a t -210" C., leaving the uncondensed methane to be pumped off. Because no other refrigeralit is kno\vii which can he used to produce conveniently a temperature of approsimatelj- -210" C., liquid nitrogen a t reduced pressure s e e m to offer a unique solution to the problem of obt,aining simply the conditioiis needed for the quantitative recovery of ethanr t'roni :Lir a i d methane.
The gas mixture I{ as passed through a coiled glass trap plared in the refrigerated zone, B (Figure 2). T o effect complete removal of the condensable fraction, the gases were circulated with a Toepler pump for a t least five passes over the trap. The noncondensed gases (methane and air) were removed by exposing the system to the pumps. The amount of condensed gas was determined by warming the trap and pumping the gas into a measuring pipet identical with that described by Prescott and hlorrison (6). The measured sample was then transferred to a sample tube by circulating the gas over a 3-mm glass tube inimersed in liquid nitrogen a t approximately -210" C. The gas was scaled off in the sample tube and then introduced into a mass spectrometer where the amounts of ethane, methane, and air were determined.
The author wishes to aclinowletige the assistmcc i n the analytical work provided by Nrs. 11. E. C. Tieiles.
In all cascs the sample was found to be essentially free of methane. I n some instances a few per cent of methane was
.4CKNOWLEDGAl ENT
LITERATURE CITED
(1) Burrell and Iiohertso11, C . S. Bur. X n c a , Tech. P O ~ 142 W (1915). (2) D u m , T. H., C . Y.P a t e n t 2 , 2 1 2 , 6 8 1 (Aug. 27, 194Oj. 13) Hassler, G. L., Ihid., 2,230,593 (Feb. 4,1941). (4) H o d g m a n a n d Holmes, "Handbook of Cheinistry and Physics," 24th e d . , p. 1778, Cleveland, Ohio, Chemical Rubber Publishing Co., 1910. ( 5 ) Horvitz, L., U. S.P a t e n t 2,287,101 (.June 23, 1942). (6) Prescott and Morrison, IXD. ENG.C m x . , . X s . k r , , Eo., 11, 230
(1939). ( 7 ) Sanderson, R. T., U. S. I'atcnt 2,375,949 (May 15, 1945). R E C E I V E.Jul)D 11, t949
Kjeldahl Microdigestions in Sealed Tubes at 470' C. LIWREYCE RI. WHlTE ANI 3 I i R I O N C. LONG V e s t e r n Regional Research Laboratory. .Ilbany, Calif. Kjeldahl digestions for the microdetermination of heteroq clic nitrogen were carried out in heavy-walled, sealed glass tithes at 470' C. w i t h concentrated sulfuric acid and mercuric oxide cataly s t . At this temperature the digestion is complete in a fraction of the time usually reconirueiided for heteroc?clic compounds. The accuracy and precision of the method are good because there is no possibility of nitrogen loss due to thermal deconiposition of arnnioniimi bisulfate or by bumping. The pressure deleloped w ithiit the digestion tuhes is nominal.
I
S 1889 Gunning ( 5 )introduccxd the use of pota,esium sulfate to hastcri the Kjeldahl digestion by raising the boiling temperature of the digest. Its use has become general. The increased boiling temperature is especially effective in shortening the time required to obtain complete nitrogen recovery from refractory compounds. Thus Ogg and Willits ( 7 ) sholl-ed that the time required for complete digestion of nicotinic acid was approsimately halved for each IO" C. the temperature of the digest was raised, and even with as much as 625 mg. of potassium sulfate per milliliter of sulfuric acid, they found that a minimum of 3 hours was required for the complete digestion of this refractory material on the microscale. Too high a concentration of potassium sulfate has been shon 11 to cause nitrogen loss due to thermal decomposition of the animonium bisulfate formed during the digestion. Therefore, there is a practical linlit to which the digestion time may be shortened by increasing the boiling temperature of the digest through thr addition of potassium sulfate. To avoid the uncertainties incident to use of high concentrations of potassium sulfate and the long, tedious digestion required for refractory materials, the authors digest such samples with sul-
furic :Lcicl uiitl tiicsrcwric oxide in se:iled tuhw at 470' C. Gnd(>r thrw conditiolis iiitrogtin cminot be lost, through theriiial dwomposition of ammonium hisulfate or by humping. The dipstion requirvs little :ittention and it is complctd very quickly. A previously descrihrd Kjeldahl sealed-tube digestion iiiacromethod ( 6 ) ,using fumiiig sulfuric acid and requiring several hours' hcating :It 330" C., does not offer t,he advantages of a p w d and convenience inherent in thc nxthod described hpre. PROCEDURE
Weigh a 5- to 10-mg. sample into a heavy-idled borosilicate glass Carius tube ( 9 ) approsimately 7 iiiches (17.5 cm.) long. Add 40 mg. of mercuric oxide and 1.5 ml. of concentrtlted sulfuric acid and seal the tube with a gas-osygen torch. Place the sealed tube i n an inclined position on a corrugated aluniinum shelf in a welded steel box constructed to fit closely within a temperaturecontrolled muffle furnace. Close the box and insert it, into the muffle heated to 560" C. and reset the temperature control to 470" C., the desired digestion temperature. (The heat capacity of the box made it necessary to preheat the muffle to 560" C., so that t,he sample would quickly come to temperature. Under these condit,ions, the shelf that supported the tuhes reached 470"C. in approximately 15 minutes.) Heat 15 minutes at 470" C.,
