As the light beam passing through the sample converges, a loss of energy results when the microcell is used in the thermostated adapter. This loss was measured and found t o he between 77 and 78% of the energy transmitted by the microcell in the standard adapter. Despite the loss of energy, instrument performance remained satisfactory because the signal t o noise ratio is still sufficiently large. Data on the performance of the microcell adapter are given by Bennett (1). ACKNOWLEDGMENT
The authors wish t o thank Monfred J. Weiss for construction of the cell and Richard Bennett for selection of the wire screen.
eter, equipped with the Model 85 microscope attachment (5). The entire optical path of the instrument was continuously flushed with a stream of air dried over activated alumina. A. Measurements on Self-Supporting Films, Suitable for Nonvolatile Liquids. A relatively nonvolatile liquid is measured as a thin film contained in a small hole drilled or cut in a thin sheet of suitable material. For example, a hole approximately 0.5 mm. in diameter was carefully drilled with a diamond-tipped pencil in a glass coverslip of 0.09-mm. thickness. After thorough cleaning of the glass surface, 4 p l . of a 0.4% solution of n,-decyl alcohol in carbon tetrachloride was delivered over this hole from a micropipet. The amount of solute actually delivered was 13 y (0.016 pl.). As the solvent evaporated, a thin film of the pure alcohol co1lect.e.d in hole nsed .._ . ..the ..... ....nnd w2.s _ . . ... to ..oh.. tain a satisfactory spectrum. I n like mann er, a satisfactory spectrum was ) benzyl obtafined from 20 y (0.018 ~ 1 of beneslate. Support for this film was provnded by a Teflon sheet of 0.03-mm. thick]less with a hole 0.5 mm. in diameter. The end of a thin d a s s cauillarv was a metri,
in the top disk, B. The larger a-ire loop was used around the inner sample loop as a guide for refitting the cell together and to keep the stopcock grease, which was used to seal the cell, away from the sample loop. It was necessary to use insoluble stopcock greases whenever volatile solvents or solutes were placed in the cell. An esterification resin (7) and a glycerol-bentonite clay mixture were both satisfactory greases for this purpose. The cells were filled by touching the sample loop with the tip of a glass capillary containing the sample, and then covering it immediately with the top pellet.
~
LITERATURE CITED
(1) Bennett, R. T., Ph.D. thesis, Rutgers
University, New Brunmick, PIT. J.,
1956. (2) Trenner, N. R., Arison, B., Walker, R. W.. ANAL. CHEM. 28, 530
(1956). ( 3 ) Trenner, N. R., Arison, B., Walker R. W., Appl. Spectroscopy 7, 166 (1953).
Sample Handling for Qualitative Infrared Microspectroscopy
E.
D. Black, J. D. Margerum, and
G. M. Wyman Quartermaster Research 8. Development Center, Natick, Mass. o IDENTIFY minute amounts of Tmaterials, a need arose for simple techniques for handling liquids and solutions in infrared microspectroscopy. While several ingenious methods have been described ( I ) , they usually require special equipment. For example, in one very elegant technique the samples are contained in capillary tubes made from silver chloride (2). These specially prepared capillaries were about 0.075 mm. in diameter and held only about 0.005 pl., although a somewhat larger volume was needed for manipulation of the sample. Another type of microcell in which the sample fills an etched-out space between polishedsalt flats has been described (4, 6 ) . These latter cells used minimum sample volumes ranging between about 6 p l . for a cell of 0.1-mm. thickness and about 0.2 ul. for a 0.01-mm. cell. APPARATUS AND PROCEDURE
All measurements were made using
a Perkin-Elmer Model 112 Spectrom-
The effective film thickness of both samplcs was estimated to be 0.04 mm. by comparing their spectra with those obtained by conventional means in a 0.03-mm. cell. The films remained intact for a t least 1hour. Some liquids form satisfactory films more readily on one supporting material than on another. It is evident that such factors as the surface tension of the liquid and the wettability of the solid influence the formation of these films. Delivery of a liquid t o the hole via a volatile solvent was preferred over direct delivery of the liquid from a micropipet or rod, for in the latter case the film thickness was a p t to be too great. B. Measurements on Liquids Contained in Potassium Bromide “Sandwich” Cells, Suitable for Volatile Liquids a n d for Solutions. The sample cell consisted of a loop of thin platinum wire sandwiched between potassium bromide disks, and was very easily constructed with the aid of a press and a die used for making potassium bromide pellets (6). The cells were made in the following mannei : Sufficient potassium bromide powder for a thin disk was packed firmly with hand pressure into the pellet die. Ta.0 loops of wire, approximately 1 and 4 mm. in diameter, were formed from platinum wire 0,010 inch in diameter, placed on top of the packed powder and covered with a thin potassium bromide pellet. All this was pressed together (25,000pound total load) to form a unit, which was then separated at the interface with a razor blade and used as a demountable cell, as indicated in Figure 1. The wire was thus partially embedded in the bottom disk, A , and caused indentations
Figure 1. Potassium bromide infrared microcell
Because of its relatively high boiling point and good transparency in the 2to 13-micron region, bromoform (Eastman Kodak, spectrograde) was found to be the most convenient solvent for use in these cells. (On exposure to room light this diphenylamine-stabilized bromoform turned yellow. Homever, this color change did not cause a detectable change in its infrared spectrum.) A typical cell used only 0.075 p l . of bromoform t o fill it, and gave a solvent thickness of approximately 0.08 0.01 mm. T o check this met.hod, satisfactory spectra were obtained for solutions of ethyl propionate, isobutyraldehyde, and aaohenaene in bromoform. Each filling of the cell required about 5 t o 20 y of solute, depending upon the concentration of the solution. Carbon tetrachloride can also he used as a solvent (demonstrated by obtaining a satisfactory spectrum of benzoic acid in this medium), but, because of its greater volatility, carbon
*
VOL. 29, NO. I, JANUARY 1957
169
tetrachloride is more difficult to handle in such minute volumes. Best results were obtained by placing a n excess of a precooled carbon tetrachloride solution in the area of the sample loop, thus using 0.1 to 0.2 pl. of solution. The main advantage of the above techniques is their relative simplicity. An additional advantage lies in the ease of locating and focusing on the samples, which had a larger cross section than that of the infrared beam at its focus point. Hoivever, because the thickness of the samples is not accurately reproducible, these methods are suitable only for qualitative measurements. The sample volumes required are small; approximately 0.02 111. of liquid by Method A and about 0.08 pl. of solution by Method B. The relative ease of handling tends to compensate for the larger 1-olumes that are required here as compared to the silver chloride capillary method.
control the f l o ~rate with fluctuations not exceeding 3% for periods up to 16 hours, and it should be simple in design and easy to fabricate in the laboratory. Satisfactory control of flow rate was not obtained by either a glass stopcock or a Teflon micro needle valve.
-CHROMATOGRAPHIC COLUMN
@f -10/30
ST JOINT
FLOW R E G U L A T O R
POWDERED PYREX G I.ASS (40-60MESH)
ACKNOWLEDGMENT
The authors are indebted to D. R. A . JT‘harton and hl. L. Kharton for their helpful suggestions and to L. 11. Dogliotti for the photograph. LITERATURE CITED
Blout, E. R., “?\licrospectroscopv,” vol. I, part 111, 2nd ed., p. 2179, “Technique of Organic Chemistry,” ed. by .4. Weissberger, Interscience, New York, 1951. Blout, E. R., Parrish, JI., Jr., Bird, G. R.. -4bbate. 11. J.. J . Ont. SOC. Amer.’52, 966 (1952). Coates, V. J., Offner, h.,Siegler, E. H., Zbid., 43, 984 (1953). Davison, W. H. T., Ibid., 45, 227 (1955). Friedel, R. h.,Pelipetz, AI. G., Z b i d , 43. 1051 11953). Perk& Elmer Znstrument X e u s 5 , S o . 3, 8 (1954). Sager, T. P., ISD.ENG.CHERT., AXAL. ED.4, 388 (1932).
Device for Regulating Small Flow Rates in Chromatographic Columns L. H. Gevantman, R. K. Main, and L. M. Bryant, U. S. Naval Radiological Defense Laboratory, San Francisco 24, Calif.
of investigations involving column chromatography of true solutions, a device was needed to moderate the flow rate of eluate to approximately 3 or 4 ml. per hour. The following additional characteristics were desired: The device should be easily adjustable to columns of different inherent initial flow rates: it should
I
N THE COURSE
170
ANALYTICAL CHEMISTRY
GLASS WOOL
A satisfactory all-glass device consisted of a tube 6 em. long and 3 mm. in inside diameter, constricted a t the bottom and fused to a female 10i30 standard-taper joint a t the top. This tube was filled with fluid and packed n-itli powdered Pyrex glass (Corning Glass Co. KO.7740, 40- or 60-mesh glass ponder) to a desired height. A plug of glass wool a t the constriction fixed the powdered glass in place. This column was fitted to a male standard-taper joint which was connected through a stopcock to the bottom of the chromatographic column proper. The flow rate through the resulting flow regulator n-as adjusted by suitable changes in column diameter and length and in grain size of the powdered glass. Within limits, the facility with which the height of the powdered glass may be adjusted in the tube provides a flexibility hitherto not obtainable from a suitable capillary tube. Furthermore, excessive lengths of capillary tubes are not needed in this compact device.
Fraction Cutter for large Scale Column Chromatography Julian E. Philip and Jay R. Schenck, Abbott Laboratories, North Chicago, 111.
efficient fraction cutter for scale chromatographic columns has been used in this laboratory for over a year. This all-glass unit is especially useful for collecting fractions of 0.5 to 20 liters. SIMPLE,
A large
A complete unit is shown in Figure 1. Tubing 10 ml. in outside diameter is used, except for the 15-mm. portion which contains the float. The top surface of the float is ground against its seat to provide an efficient shutoff. Three or four indentations are made below the float, so that solutions may flow rapidly down the tube. These units are connected in series by tubing to the chromatographic column During operation, the effluent from the chromatographic column flows through the T-tube and down around the float into the container. ’IThen the liquid level in the container reaches the bottom of the straight glass tube, the solution rises in this tube and in the T-tube until the float seats and further f l o ~is diverted to the next unit. This effectively prevents mixing during the column operation. Before disassembly. the tubing is disconnected from the column and gently blon-n out with air. This removes the solution from the horizontal part of the T-tubes and the connecting tubing. The principal source of contamination of early fractions by later fractions is the volume above the float This solution enters the container nhen the stopper is removed at the end of the experiment. I t is, therefore, desirable to make this part of the T-tube as short as possible.
ti K L I N D E N T A T I O N
Figure 1.
Fraction cutting unit
The amount collected in any container is determined by the position of the bottom of the straight glass tube or T-tube (whichever is higher) and may be thereby adjusted. The top of the straight glass tube must extend above the T-tube enough to allow a considerable head of liquid if many units are used. ACKNOWLEDGMENT
Thanks are due Chester Swopes for constructing the first units.
