V O L U M E 2 2 , NO. 6, J U N E 1 9 5 0
835
with which several drops of. dye were diffused throughout the cell after being placed inside the glass shield, as well as by the rate with which diffusion current readings became constant in performance of titrations. The stability of current readings Was tested by titration of sulfate ion with lead nitrate solution, according to the procedure of Kolthoff and Pan (S),and found to be satisfactory. Results were within the limits of accuracy specified for the method. Current readings with gas flowing through the cell were not noticeably higher than those obtained with no gas flowing.
LITERATURE CITED
(1)
Hurne, D. N., and Harris, U'. E . , rND. ENO;. CHEM.,ANAL.ED.,
15, 465 (1943). (2) Kolthoff, I. M., and (1940).
Langer, A , , J . Am. Chem. Soc., 62, 211
(3) Kolthoff, I. M., and Pan, Y. D., Ibid., 62, 3332 (1940). (4) Laitinen, H. A.. Higuchi, T., and Czuha, M., Ibid., 70, 561 (1948). (5) Philbrook, G. E., and Grubb, H. M., ANAL.CHEM.,19, 7 (1947). (6) Wise, 1%'.S.,Chetnislry and Industry, 1948, 37.
RECEIVED Deeemher 5 ,
1849.
Pressure Pycnometer J. D. PARK, W. E. JOHNSON, AND J. R. LACHER, University of Colorado, Boulder, Cob. URING an investigation in this laboratory it was necessary
D to devise some means of determining the composition of binary liquid mixtures. Measurement of the density of the liquid mixture was chosen as the analytical tool. Because one or both of the components in the determinations were extremely volatilee.g., CF,CI*, CHC12F-ordinary pycnometric methods would not suffice. The pycnometer used by Walden (IO), somewhat similar to that used by Goodhue and Hazen (4), consisted of a glass bulb on the lower end of a graduated capillary tube. The upper end was sealed by mechanical pressure t o a concentric valve of the aerosol type. The maintenance, operation, and cleaning of this model were rather difficult, and in general it was not rugged enough for the purpose. The possibility of an all-metal construction was then investigated. The steel pycnometer used by Eilarts, Smith, and Cook .. e( 9 ) for specific volume determinations on oil-well samples wa
S E C T I O H A-A
signed for use at extremely high pressures (up to 4500 pounds per square inch) and in design and operation was deemed to be too involved for adaptation to the authors' use. The pycnometer as finally designed and used is constructed chiefly of brass, and has been used successfully a t pressures up to 100 pounds per square inch. APPARATUS
The body of the pycnometer and the expansion chamber were turned from brass bar stock. The plug threaded into the body a t the lower end was of the same material, as was the late a t the top of the expansion chamber, which was secured wit{ three screws. The Hoke-type needle valves were threaded into the body of the pycnometer and the upper plate of the expansion chamber. Adapters (not shown in Figure 1) were threaded onto the valve exits for use with flare-type fittings. All the metal-to-metal joints were soldered. The reading tube was a section of borosilicate glas? capillary inch in inside diamtubing, 6/16 inch in outside diameter and eter, which had been raduated a t 2-mm. intervals over a span of 5 cm. The ends o f the tube were ground down flat, using a drill- re= and corundum powder. The tube was sealed in place with %eflon gaskets by mechanical pressure from three bolts extending from the body of the pycnometer through the expansion chamber assembly. Two pycnometers of this type were constructed, with capacities of ca. 36 and 40 ml. The capacity of the expansion chambers was ca. 3 ml. The tare weights of the two pycnometers were approximately 500 and 700 grams. CALIBRATION
The pycnometers were calibrated with water a t 0' C., the temperature at which all samples were taken. The pycnometer was cleaned thoroughly with acetone, followed by an alcohol rinse, and evacuated with a Hyvac pump. The outside was then cleaned with acetone, and the pycnometer was allowed to come to constant weight. It was then connected at the lower valve to the container of water by means of an L-shaped len t h of copper tubing. %e liquid containers used (bombs), were of the small (500-ml.) aerosol type, fitted with Hoke-type valves and adapters for flare fittings. The blow-out plugs could be removed for cleaning purDoses. or to furnish an air inlet if the contained liauid had too low ;a 0; pressure to force i t from the bomb. $he pycnometer was then immersed in an ice-water bath, so t h a t the water level came to a mark on the stem 1 cm. above the body of the pycnometer. The lower valve was opened by means of a specially constructed long-handled wrench, and the system was evacuated through the top valve. Both valves were closed, and the vacuum line was disconnected. The bomb valve was then opened, and the pycnometer was allowed to fill to the reference mark, the flow being controlled with the pycnometer valve. About 15 to 20 minutes were usually required for temperature equilibrium. The pycnometer was removed from the ice bath and disconnected, and the outside was thoroughly washed with acetone and then allowed to come to constant weight. After each calibration run the pycnometer was again immersed in the ice bath. In each case the liquid level came back to the reference mark. Figure 1. Diagram of Pressure Pycnometer
The results are shown in Table I.
A N A L Y T I C A L CHEMISTRY
836 Table I.
Volume of Pycnometers at 0' C.
Pycnometer 1, MI.
Pycnometer 2, MI.
39.97 30.95 39.94 30.95
36.15 36.13 36.15 36.14
Av.
results summarized in Table I1 give a comparison of the densities obtained by this method with those Obtained by other means. No corrections were made for buoyancy, because most of the results reported are accurate to within 0.001. However, a simple buoyancy correction has been suggested by Lipkin el al. (6) where the correction is C = 0.0012 X (1 - determined density). This works only where the pycnometer has been calibrated with water as recommended here.
Table 11. Densities of Pure Liquids at 0' C. Pycnometer Pycnometer
Compound CFrCIz
1
2
1.402 1.399 1.401" 1.287 1.622
1.401 1.401 1.39ga 1.286 1.621
Other Workers 1.395 ( 2 )
LITERATURE CITED
(1) Benning, A. F., and MoHarness, R. R.,
I d . Eng. Chem., 32,
814 (1940).
(2) Bichowsky, F.R.,and Gilkey, W. K., Ibid., 23,366 (1931). (3) Eilarts, K.,Smith, R. B., and Cook, A. B., U. 5. Bur. Mines, RepLlnvest. 3474 (1939). (4) Goodhue, L. D.,and Haaen, A. C., ANAL. CAEM.,19, 248 (1947). (5) Hovorka, F.,and Geiger, F. E., J . Am. Chem. Soc., 55, 4759 Run on refractionated material. (1933). (6) Lipkin, M. R.,Mills, I. W., Martin, C. C., and Harvey, W. T., ANAL.CHEM.,21,504(1949). (7) Looke, E. G., Brode. W. R., and Henne, A. L., J . Am. Chem. DENSITIES OF PURE LIQUIDS Soc., 56, 1726 (1934). (8) Reidel, L., 2.ges. K d l t e - I d . , 45, 221 (1938). Density determinations were made on dichlorodifluoromethane, (9) Swarts, F.,J . chim. phys., 28, 622 (1931). difluoromonochloromethane, and 1,1,2-t~rifluoro-1,2,2-trichloro- (10) Walden, C. H., University of Colorado, unpublished work. CHClFt CCIFI-CCI1F
1 . 2 8 6 (calcd.) ( I ) 1 . 6 2 1 2 (calcd.) ( f ) 1,6210 ( 8 ) 1 , 6 2 0 0 (7,9) 1 . 6 1 9 5 (6)
ethane, using the same procedure as used in the calibration. The
RECEIVED June 27, 1949.
Method for Facilitating Recording, Filing, and lntercomparison of Infrared Spectra 0. D. SHREVE AND M. R. HEETHER Philadelphia Laboratory, E . I . d u Pont de Nemours & Company, Inc., Philadelphia, Pa. T H E large recorder traces, obtained in the qualitative scanning of infrared spectra, are inconvenient t o file, difficultly accessible after filing, and cumbersome to handle when simultaneous intercomparison of several records is desired. In the caae of a Beckman IR-2 infrared spectrophotometer, these disadvantages may be and operator minimized by the following procedure, which should be readily adaptable to other instruments. The marker switch, which marks the record a t intervals correspondin to each 30" of arc on the calibrated wave-length dial, is replace! with an ordinary tapping key. Using a mechanical Figure 1. 1. 2.
automatic slit drive to close the slits at an approRriate rate, the
~~~?~~?
a recorder chart s eed of 30 inches per hour, a 26-inch record is obtained. The S i t shaft is driven by the idler shaft Of the wave-length drive by means of a sprocket wheel and ladder chain ~i~~~ 1 shows the tvDe trace ob- - of bsckmound tained dvith ti& device. Records of convenient filing size and with a ermanent wavelength scale attached are then obtained aa folgws. With the instrument dark (shutter closed) t i e 1 5 to 2micron range is scanned exactly &s described above and the
Sample Record Mounted on Keysort Card for Filing
Atmospheric background Background with dry nitrogen passing through opectmmster
/-
I
II
2%;t i
& ~ ! n ~ t ~ ~ ~ ~% :r:af$::: ~ ' to 2-micron redonis then means of the tapping key. The scanned a t a slower rate (28.5 minutes) without attent>ion. With
