652
Vol. 17, No. 10
INDUSTRIAL AND ENGINEERING CHEMISTRY
The diffusion method offers better control of the crystal size and shape than preparing crystals by direct mixing. Although the crystal size depends upon the concentration, the habit is only slightly affected. A considerable number of crystals were prepared and electron micrographs taken. When the habit of the crystals was compared with that observed for micro- and macroscopic crystals, good agreement was observed. Calculations for a typical reaction indicate that quantities of the order of 10-7 gram of silver ion can be detected in reaction with a 0.1 M potassium chromate solution. The theoretical limitation is approximately 10-8 or 10-9 gram of silver ion in the same reaction. Electron diffraction analysis is utilized to help in identification of the crystal. The application of the electron microscope to chemical microscopy is suggested. A useful role may be expected, especially in
the study of precipitates which are not perceptible under the optical microscope. The practical details of individual reactions have yet to be developed. They would include a study of concentration limits, interferences, trials in the identification of real unknowns, etc. LITERATURE CITED
(1) Boricky, E.,“Elemente einer neuen chemischmikroskopischen Mineral- und Gestein-Analyse”, b a g , 1877. (2) Burton, C. J., Barnes, R. B., and Rochow, T. G., IND. ENQ. CHEM.,34, 1429 (1942). (3) Chamot and Mason, “Handbook of Chemical Microscopy”, Ne’w York, John Wiley & Sons, 1940. (4) Dana, E. S., “Textbook of Mineralogy”, New Pork, John , Wiley & Sons, 1915. (5) Iqid., revised by W. E. Ford, 1940. (6) Picard, R. G., J . Applied Phya., 15, 678-84 (1944). (7) Winchell, “Microscopic Characters of Artificial Minerals”, New York, John Wiley & Sons, 1931.
A n Automatic Differential Manometer D. J. LEROY Department of Chemistry, University of Toronto, Toronto, Canada
HE problem of measuring small pres-
T
The author describer a gage for measuring pressure changes with an accuracy of 0.01 mm. when the total pressure is 50 mm. or more. The sensitivity of the instrument is independent of the total pressure and mercury is used as the confining liquid. Pressure changes are read directly off a scale graduated in millimeters.
sure changes in gaseous reactions has always presented considerable difficulty to the physical chemist. The McLeod gage has probably been used more than any other instrument for this purpose but it can be used only a t relatively small total pressures. Bourdon and spoon gages have been used in many cases but they are of necessity very fragile if pressure changes of the order of 0.01 mm. are to be detected. In the opinion of the author the most satisfactory differential manometer for general laboratory use is that due to Pearson ( 4 ) . It has the advantage of using mercury as the confining liquid and its sensitivity is independent of the total pressure.
The princi le on which the gage operates is simple. Consider three verticarglass tubes, TI,2’2, and Ta, joined together at the bottom and containing mercury. If T1is evacuated and TZis connected to the system whose pressure chan es are to be measured, and if ht, h, and ha are the coresponjing heights of the mercury above a fixed horizontal plane, then the pressure in the system is obviously hl- hz. Furthermore if, by some method, we are to maintain level hl constant, any change in pressure in the system must be compensated by a pressure change in tube Ts. Since there is a fixed amount of mercury in the gage, Aha/ Ah2 must be equal to (&/d#, where dz and da are the internal diameters of tubes Tt and 2’3, respectively; or A p = (d8/dz)zX Aha, where p is the gas pressure in TI.In practice & may be 20 to 25 mm. and ds 1 to 2 mm. The pressure change in TZcan thus be magnified several hundred times. The two limiting factors are the method of indicating a constant level in TI and the method of controlling the pressure in Ta.
-
Blacet and his associates (I, 9) have used this type of gage successfully in their work on photochemical gas reactions. The correct height of the mercury in T Iwas indicated by the making and breaking of an electrical contact between the mercury surface and a fine tungsten wire, the circuit being completed through a second tungsten wire sealed through the tube below the mercury surface. The pressure in Ta was controlled manually by turning a three-way stopcock, either evacuating or admitting air to a large ballast volume a t the top of 7‘s. I n the author’s experience this method of control was unsatisfactory for three
reasons: (1) the turning of the stopcock communicated unde sirable vibrations to the mercury surfaces; (2) it was difficult to adjust the pressure in 7‘s to the correct value manually; and (3) it would be a great advantage to be able to read the pressure a t any time without having to manipulate the controls. The gage in its present form is illustrated in Figure 1. It is similar in construction t o that of Blacet et al. (1, 2) with the exception of the automatic feature and a few minor changes.
The t u n p t e n needle, W , was made from a piece of 1-mm diameter wire shaped to a fine point by treatment with sodium nitrite. Accurate centering of the needle was facilitated b seging it as shown into a 10/30 ground joint lubricated wit$ Apiezon N grease. The two electrical contacts were connected to a vacuum tube rela of the type described by Serfass (6). No time delay was u s e J The relay operated a fixed solenoid, S. When the solenoid is actuated the lass-enclosed iron plun er is raised and air enters the 500-cc. b u k , B , through the thin dum disk D,sealed to the bottom of the 6-mm. diameter Pyrex tube. d e n the rela current is cut off the plunger falls, shutting off the air supply. $he success of the instrument is largely due to the needle valve, A-, which controls the rate of evacuation. After some investigation Hoke high-vacuum valve S o . 318 was found to be most satisfactor For a porous disk, D, of f i e d dimensions and with continuous pumping through N (a Hyvac pump was used) there is an equilibrium pressure for B where the rate at which air enters through D is equal to the rate a t which it is pumped out through N . If this equilibrium pressure happens to be in the range where the gage is supposed to operate, control will be impossible. Consequently, it is necessary to adjust N so that the rate a t which air is pumped away will be somewhat less than the rate at which it can enter (with D open) over the whole range of pressures to be encountered. If the pressure change during an experiment is large, an occasional slight adjustment of the needle valve wiy assure control at all times. In most of the author’s experiments no readjustment was necessary. In one model TZwas approximately 22 mm. and Ta 2.0 mm. in inside diameter. TI was calibrated over a range of 100 mm. before sealing it into the apparatus. Since the pressure change in the system during a run was only of the order of 1 mm., the value of dz would remain practically constant. Ts was calibrated over a range of about 300 mm., corresponding t o a pressure in B of from 250 to 550 mm. For a nonuniform bore the expression for the pressure change becomes Ap = Sd32 where the
.bun-
dz
October, 1945
ANALYTICAL EDITION
integral is taken over the range of variation of ha, from the value a t the beginning of the run to the value a t the time of observation. Although the internal diameter of TIdoes not appear in the expression for A p , its value does &ect the behavior of the instrument. Since the mercury level in T1 continually fluctuates as contact with the needle is made and broken, it is desirable to have dl small to avoid large fluctuations in ha; however, if it is too small the mercury will have a tendency t o stick, and if it is too large, vibrations on the mercury surface will be serious. I n the instrument described here dl was approximately 18 mm. in inside diameter. The only other dimension that need be mentioned is the distance from the tip of the needle, W , to the top of the mercury surface in R. This was approximately 800 mm., so that only a slight pressure had to be applied to bring the mercury in contact with the needle before closing P, as described below. Valve N and the U-tube and solenoid can be placed in any convenient position. The construction of the top of bulb B is such as to prevent mercury from capillary 2‘1, from getting on top of the -4lundum disk, D,or into valve hTthrough an accidental large increase in pressure in the system. When not in use the manometer was isolated from the rest of the system by a mercury cutoff. PROCEDURE
By connecting reservoir R to a vacuum line by means of the three-way stopcock the mercury level in TI and Tzis brought down to a point somewhat below the junction of these two tubes. Stopcock P is open. TIand T2are then evacuated and flamed, if necessary, together with the remainder of the system connected to the top of T2. When a stable vacuum of 10-6 to 10-6 mm. is attained, the mercury is admitted to T1 and TZby connecting the three-way stopcock to the atmosphere. The gas IS then intro-
653
duced into the reaction system a t the desired pressure, forcing the mercury down in T2 and up in 2’1. The pressure in B i s adjusted until the mercury level in 23‘ is in a suitable part of the tube. By means of an aspirator bulb connected to the three-way stopcock on R the mercury level in TI is raised until it coincides approximately with the tip of the needle. Stopcock P is now closed and the relay circuit switched on. When the needle valve is properly adjusted the gage is completely automatic. T h e mercury in Ta will rise or fall until contact is made in TI. Small vibrations on the mercury surface cause the relay to go on and off a hundred or more times a minute, but the level in Ts remains sensibly constant to within a fraction of a millimeter. The position of the mercury in T3 is noted and the run can then be started. Subsequent changes in ha are related to pressure changes in the d2 reaction system by the relation A p = X Ah3 or, if Ta is not
2
uniform, Ap =
J ds2 X dha d22
‘
The use of this type of gage by the author has already been referred to ( 3 ) . Its use in other researches as well has shown it t o be almost indispensable in the investigation of gaseous reactions involving small pressure changes. The maximum value of ( & / d ~ used ) ~ so far is about 100, although larger ratios could be used (1, 3 ) . Temperature changes must, of course, be takcn into consideration if the measurements are to have any meaning. The author has done this in two ways: (1) by immersing the system, not including the manometer, in a thermostatically controlled bath, and (2) by observing the temperature of the system at the time of each pressure reading and making the necessary corrections. I n addition to actual pressure changes in the system due to its temperature, the manometer itself IS subject to an error resulting mom the change in volume of the mercury. The magnitude of this latter error can be calculated as follows:
If V Ois the volume of mercury in the gage (not including that in reservoir R ) when the pressure in the system is zero and if Vis the corresponding volume when the pressure is p cm., then V = Vo
- pA2
where A2 is the cross-sectional area of TD. The increase in ha for a temperature increase At is the sum of two quantities resulting from the increase in volume of the mercury and the dtcrease in the density of the mercury in a column of length p . The fist, 1 dV A quantity is - - X At and the second is As -2 X 2 V dt X At Aa dt dV For mercury = 1.8 X 10-4 Y dt 1.8 x 10-4 vo Therefore *d- h = 1.8 X (V ~442)= As dt A3
-
+
In the manometer described here V Owas approximately 115 cc. and A , 0.03 sq. cm., the error in ha due to a change in temperature of the mercury IS about 6.6 mm. per degree, corresponding to a pressure error of approximately 0.08 mm. per degree. It follows that the temperature of the manometer must not be permitted to change more than a few tenths of a degree during the course of a run and that the volume of mercury must be kept a t a minimum. The latter is accomplished by using small-diameter tubing in all places except in the part of TI adjacent to the tip of the needle and in the part of T , covering the range of pressures to be encountered.
I1
LITERATURE CITED
(1) Blacet, F. E.,and T’olman, D. H., J . A m . Chem. Soe., 61, 582 (1939). (2) Leighton, P.A . , and Blacet, F. E., Ibid.,54,3165(1932). (3) LeRoy, D. J., and Steacie, E. W. R., J. Chem. Phys., 12, 117 (1944). (4) Pearson, T.G.,2. p h y s i k . Chem., A156, 86 (1931). (5) Serfass, E.J., IND.ENO.CHEM.,ANAL.ED.,13, 262 (1941).
Figure 1.
Gage
CORRECTION. The article “Microdetermination of Carbon and Hydrogen” by Ralph 0. Clark and Gordon H. Stillson [IND. ENG.CHEM.,ANAL.ED., 17, 520 (1945)] was presented a t the 107th Meeting of the AMERICANCHEMICAL SOCIETY, Cleveland, Ohio, and not a t the New York Meeting, as stated in the footnote.
