V O L U M E 21, NO. 5, M A Y 1 9 4 9
633
Table 1. Distillation of Limonene-Aniline-Benzyl Alcohol Mixture
Volatile fraction Middle fraction Residual fraction
Redistillation of Middle Fraction Weight, Refractive gram index 0,008 1.5770 0.012 1.6814 0.013 1.5761
First Distillation Weight, Refractive gram index 0.215 1 .'ii82 0.053 0.148
...
The usual function of distillation is to separate a reaction product from other substances, some of which may have higher. others lower, boiling points. This use is illustrated by distillation of a ternary mixture containing by volume 25y0 &limonene (boiling point 177.8' C.; n*O 1.4743), 50% aniline (boiling point 184.35" C.; nno1.5863), and 25% benzyl alcohol (bqilin point 205.5"C.; nZo1.5396). A 0.435-gram sample was distillecfat 10-mm. pressure and a rate of 0.01 gram per minute, and 0.045 gram of the middle fraction was then redistilled under the same conditions. The data are given in Table I. The presence of either limonene or benzyl alcohol as impurity in the aniline would lower the refractive index. Preliminary tests showed that aniline solutions
containing less than 10% by volume of limonene, benzyl alcohol, or both gave refractive indexes within 1% of values calculated on the assumption of linear relationships between refractive index and composition by volume. The distillation data indicate that the middle fraction from the first distillation contained 83 to 9370 aniline, and that from the redistillation contained 90 to 96y0 aniline by volume, depending on whether the impurity is assumed to be benzyl alcohol or limonene. A second distillation not only serves to purify the material further, but gives an indication of the homogeneity and the trend in the physical constants as the substance is purified. Thus, in the above euample, the refractive index data show that the middle fraction from the first distillation was not homogeneous and that the impuritie- were substances of lower refractive index. LITERATURE CITED
(1) Gould, C. W., Jr., Holaman, G., and Niemann, C., ANAL.CHEM., 20, 361-3 (1948). (2) Rose, A,, I n d . Eng. Chem., 28, 1210-12 (1936). (3) Sohneider, F., "Qualitative Organic Microanalysis," New York, John Wiley & Sons, 1946. RECEIVED May 28, 1948.
Buret for Precise Measurement of Small Volumes of Gases S. S. BURKE, JR., University of Wisconsin, Madison, V i s .
1
1
1
1
I
l
l
I l 1 l l l
I l l
634
ANALYTICAL CHEMISTRY Table I. Variahility of Micrometer Readings
Trial
Micrometer Readingso R, R,
0,7747 0.7747 0.7748 0.7744 0.7747 0.7283 0.7284 0.7285
0.2608 0.2608 0.2608 0.2608 0.2008 0.2148 0.2148 0.2148 Av.
a
b
volume' 0.5139 0.5139 0.5140 0.5136 0.5139 0.5135 0.5136 0.5137 0.51375
Errorb +0.00015 +0.00016
+a ,00025 -0.OOC15 +0.04015
- 0.00025 -0.CUO15 -0.
oooo5
*0.00016
Miorometer scale units. 0.1000 = 45.5 cu, mm. Calculated &Q deviation from average volume. ~
Figure 2. Sequence Diagram Showing Course of Gas Sample T h r o u g h Process of Measurement A.
Normal condition, filled with mercury except for sir enclosed within barometer bulb B . Internal ~ r e s ~ u zreduoed e C , D . Micrometer Diston withdrawn E . First reading, RL,barometer set arbitrarily at midaoale, Po P , C. Partial ejection 01 sample H . Second reading, Rz,vertical oonneeting tube filled with meraury I . Final ejeetian of sample J . Norms1 condition
herent in the principle of measuring the length of a segment of gas enclosed by mercury in a capillary tube between two meniscus boundaries. Aside from the difficulty in procuring a capillary tube ai sufficient length with a uniform bore or the alternate difficulties associated with calibration of one which is not of uniform bore, a troublesome problem which can he encountered is regulation of the pressure of tho mmplo itself a t the time a reading of its volume is taken. It is not justifiable to assume, under the actual conditions of analysis, that a dependable equilibrium always exists between the pressure of the sample and the pressure of the atmosphere vheu a segment of mercury is interposed in the intervening capillary, although errors from this source can be reduced by sorupulous attention to cleanliness of both mercury and glass surfaces. Furthermore, the deposition of foreign materials within a capillary during repeated readings can alter the bore appreciably in irregular fashion and render calibration cometions a t any given time more or less unreliable. With the buret described, the volume of sample is translated into the accurately measured length of a known uniform cylinder. This cylinder never comes in contact with the gas sample and consequently is relatively unaffected by the presence of the impurities. Because the regulated pressure is transmitted to the sample through not more than 2 om. of small-bore capillary, the unpredictable effects of surface forces ai-e minimized. Since only one of the two meniscus boundaries between gas sample and confining
mercury is subject to effectsof these surface forces, the errors from this source am still further reduced. The barometer iudieatiug the pressure exerted by the sample will function efficiently and dependably only if the interior of the buret is isolated from the atmosphere as a fixed and separate system. For this reason, dudng actual analysis, continuity of the mercury thread in the tip of the buret is iutgrrupted by B minute gas bubble which effectivelyblocks any flow of mercury a t the existing small gradient of pressure. It can be shown from the equation for tho general gas law that the ratio of pressure of t.he sample to pressure of the air in the barometer bulb changes insignificantly with ordinary variations in temperaturc. Far a 10" rise from 25" to 35" C., the error in reading s given volume would amount to O . l ~ o . The error involved in the thermal expansion of the mercury in the burot, tube between the micrometer and the sample was directly dotermincd by act.ual measurement. It was found for the normal 0.1" C. maximum variation in the temperature of the circulating water in the jacket of tho buret that an error of only 0.02 eu. mm. would he introduced even if the change took place within the brief period of 1 minute usually required to obtain R, and R,. The observed oonstrtnoy of R, during actual use of the buret confirms this finding. Representative tabular data have been included to show the accuracy that can bo expeoted from this apparatus under satisfactory conditions. ReadingJ are expressed in English scale units bemuse this happened to he the type of micrometer available; each decimal fraction, 0.1000 inch, is equivalent to 45.5-011. mm. volume of the micrometer piston. In actual analysis, only the relative values and not the absolute values of the micrometer readings are ordinarily required in the calculation of results. The uniform diameter of the micrometer piston was cheeked prior to assembly by dircet mcasurcment with a micrometer caliper. In addition, functional tests with a given small quantity of gas wore made by measuring its volume in different ranges of the miorometer scale without allowing tho gas to leave the interior of the buret. The close agreement of the results within 0.0001 scale unit, or 0.05 eu. mm., provided additional evidence of the linear uniformity of volume of the micrometer piston. Similar measurements of % larger volume of approximately 225 DU. mm. show the variability indicated by tho oight consecutive readings listed in Table I. The observed deviations may be attributed largely to the fact that compressible gas lies between the displacing piston and tho meniscus which is adjusted to the reference mark of the buret capillary. Althoueh the foregoing - - discussion deals solely with the intrinsic acouracy of the buret, this device has been used in the analysis of various gas mixtures according to the techniques of Rlacot and Leighton ( I ) , as modified by Seevers and Stormont (51, with a roliablo accuracy of 0.1 to 0.2% of the volume of the sample. In the experience of the author, determinations have been made with greater convcniencc and facility than with other similar instruments. Finally, the principal variable component of measurement with this apparatus is the value obtained for the first reading, R,, where the gas sample is interposed between the reference mark and the micrometer. Errors due to this variability are smaller than the deviations to be expected with measurcments in a capillary tuba; they are reduced to a minimum by interruption of the pressure continuity of the interior system with the atmasphere, by proper setting of the micrometer, and by careful maintenance of s clean mercury bath. ~
ACKNOWILEDGMENT
. ..
The author performed this work as a teilow in anesthesiology of the National Research Council. Grateful acknowledgment for construction of the machined parts of the apparatus is made to J. S. Hipple, former mechanician of the Wisconsin General Hospital.
635
V O L U M E 21, NO. 5, M A Y 1 9 4 9 LITERATURE CITED
(1) Blacet. F. E.,and Leighton. P. A.. IND.ENQ.CEEM., ANAT..E% 3,266(1931). (2) Christiansen, J. A,, J . Am. Chern. Soc., 47,109 (1925). (3) Krogh, A,. Skand. Arch. Phu.?ioZ., 20,279 (1908). (4) Soholernder, P. F.. J , Biol. Chhem., 167,235 (1947).
(5) Sewers, M. H.. and Storrnont, It. T., IND.END.CHEX..ANAL. Eo..9.39(1937). (6) sutton, T,c,,J , ,q& ~ ~ t15, 133 ~ (1938). ~ ~ ~ ~ (7) Swea~ingen.J. 5.. Gerbes. O., snd Ellis. E. W.. IND.EN^. CEEM.. ANAL.Eo.. 5.369 (1933). R~cslveo January 30,1948.
Simplified Blacet-Leighton Apparatus for Gas Microanalysis VICTOR M. LEWIS, University of Massachusetts, Arnherst Mass.
HE capillary buret type of gas microanalysis apparatus has Tfound extensive use in many branches of chemistry. I n its initial form, i t was described by Blacet and Leighton ( 1 ) and the technique of its use in the analysis of volumes as low as 20 cu. mm. of various gaseous mixtures has been fully described (Z-5). This piece of equipment, however, is relatively complicated, expensive to purchase, and difficult to make in the laboratory. It is also relatively tedious to operate, and the mioroburet is hard t o clean. Difficulty is also frequently experienced in getting the last small bubble of gas into the mioroburet from the analysis chamber. Swearingen et al. (6)described a horizontal modification of this apparatus which eliminated the trouble resulting from the necessity for equalizing the mercury level in the reservoir with that at the aero mark in the microburet. It also exerted less pressure on the mercury-actuating mechanism, with consequent more positive control of the mercury thread and less trouble with mercury leaks. The rubber mercury actuator described in the paper, however, has been found unsatisfactory in operation. The apparatus described here, and used with excellent results combines features described in various publications and e m he readily constructed from standard equipment.
stahdard tauer metal ioint N.;
Figure 2. Details of Gas Analysis Chamber
14/35. The pas microburet con-
ntt&hmknt to the mercury valve, and' the-opposite end of the mioroburet is fused to the gas analysis chamber (Flgure 2), whlch is a piece of glass tubing approximately 6 mm. in internal bare. The capillary tube has a right-angle bend in it, so that the andysis chamber dips into the mercury reservoir. It is iacketed by part of a 50-ml. buret, the graduations on whch serve for making readings in the microburet. The guide far the absorbent holder consists of a piece of brass rod through which is drilled a hole large enough to take the glass rod. The guide is attached to an upright support by means of a clamp holder. Additional gas reservoirs of the typo desoribed by Blaoet and Leighton (1)may be attached t o the side of the mercury reservoir by means of spring clips. The complete apparatus 1s shown m Figure 3. In operation, the gas sample is intreduced.into the analysis chamber and its volume is measured by drawlug i t up into the horizontal part of the microburet. Absorpbion takes place when
Figure 3. cue BV?S",U~"II
Assembly of Gas Microanalysis Apparatus
1s l,lll"UUurU
LYY"
Yrl
IIYI.,.YCL,
y.yll
Ly"".p-
tion the volume is again measured. Thus all operations are performed without transferring the sample; this results in both a saving of time and an increase in accuracy. If the microburet is attached to the valve by means of agroundglass joint, it may be readily removed for cleaning, or interehangeable burets of different capillmy bores may he used. LITERATURE CITED
(1) Blaoet. F. E..and Leighton, P. A., IND. ENG.CHEM.,ANAL.Eo: 3,266-9 (1931). (2) Blscet, F. E., and MacDonald. G. D., Ibid., 6,334-6 (1934). ( 3 ) Blsoet, P. E., MsoDonald, G. D., and Leighton, P. A,,Ihid., 5. 2724 (1933). (4) Blacet, P. E.,Sellers, A. L., and Blaedel, W. J.. Ibid., 12. 356-7 (1940). (5) Blaoet, F. E., and Volrnan. D. H., Zbid., 9,44-5 (1937). (6) Swearingen. J. S., Gerbes, D.. and Ellis, E. W., Ib& 5, 369-70 (1933).
Figure 1. Valve for Actuating Mercury Thread
R ~ o m v ~November n 10, 1947. t u r d Experiment Station.
Contribution 642, MQSsachmsettaAniaioul-
t
