Water Vapor Saturation in Gas-Measuring Burets THEODORE W. STEIN AND ROBERT C. REID .lassachusetts Institute of Technology, Cambridge 39, M a s s . N THE
common volumetric or Orsat analysis of gas mixtures, in
1. which . the various components are determined by successive
absorption and/or combustion, the gas sample must have the same percentage of \rater vapor present each time i t is measured if the results are to br correct. The most convrnicnt n a y to accwniplish this is to keep the gas sample saturated n i t h water vapor. I n the conimonly used procedure, saturation of the gas sample is obtained 1)y keeping a few drops of water on the surface of the mercury used as the confining fluid. This water is spread iri Y thin film over the Iluret walls as the mercury level is lowered. If a salt solution n-ere used as a confining fluid, water vapor woultl I)(, av:tilttble from the solution a t a constant, but reduced, pxrti:il pressure. ;\lthough it has bren recognized that saturation of the gas sample m a y take considerable time, no figures or estimates are availxhle. The question of how long it t:il.res to saturate the gas is discussecl in this paper.
l
I"\
n-here D = diirusivity of lyater vapor into air, set ,i = ( P P , ) / ( P , - P o ) ,and the boundary conditions then become
-
4=1,e=o A = O , 8 =
m
A = 0,x = 0 A = 0,
= 2L
d
IVhere
P = partial pressure of water vapor at any value of
T
a t any
time 0 Po = vapor pressure of water in water layer P, = initial partial pressure of water vapor a t 1, = 0 the solution is of the general form
i n = 1, 3, 5 The time required t o reach saturation in the case of a buret whose walls are covered with a film of water is found to be considerably less
I
Figure 1.
:I
I
20 30 40 50 D I S T A N C E F R O M WATER L A Y E R - C M IO
1
!
I
I
I
I
I
07
06
05
04
03
I
l
l
02
01
GO
Partial Pressure of R a t e r Vapor i n Buret When Sicles i r e Not F e t t e d
An es:nnple will slio~rthe importance of saturating the gas before volume iiieasureinerits are made. Consicler a typiral . *\ i o - c c . saniplc of d r ~ gas containing lOyo carbo dr;tn-n into a measuring buret r drops of water. The water ovcr mercury covc~rotl\vi tl jacket is kept a t 2,i" ('. If the gas sample were to be completcaly saturated with \rater vapor i t would contain (23.7; 760) (100) = 3.1yo n-:ttrr vapor and the measured volume would equal (TOj0.969) = 72.5 cc. IIoxrever, if the gas were only 50% saturated, i t would contain l.55y0 water vapor and a volume of ( i o ,0.9845) = T I , 1 I-($. wnuld be measured, introducing an error of (72.5 - 71.1) = 1.2 ec. IVhen the carbon dioxide in the gas samplr. is absorbed in dilute base, the residual gas which is returned t,o the buret, saturated with water vapor (neglecting, for purpose of the r,raniplel the decrease in vapor pressure of water in caustic solutions), has a volume equal to (63/.969) = 65.1 cc. : thus the carbon dioxide percentage reported would be (T1.l - 65.1)/(71.1) X (100) = 8.4% when the correct percent:cge equals ( i 2 , : 3 - 63.1)/(72.3) X (100) = 10%. Calculations are given in this paper to show that a clean buret whose ~ r a l l sare covered with a thin film of water will saturate the gas very rapidly, whereas in a buret whose sides are not wetted b\- wat,er it is virtuall>- impossible to saturate the gas in an)' rcasonable length of time. Considering the latter case first, a buret of length L has a f water in the bottom. It is desired t o calculate the pressure of n-ater vapor along the length of the buret rent time intervals, assuming the gas is stagnant and all mass transfer occurs by molecular diffusion. Let 5 represent the distance up the buret from the liquid water. P the partial pressure of water vapor, and 0 the t h e . Then by Fick's second law of diirusion
01 08
DISTANCE FROM CENTER-CM
Figure 2. Partial Pressure of Water Vapor in Buret When Sides Are Wetted
If a buret whose radius is R has a film of water on its sides, then the relation between the radius, R, time 0, and the partial pressure of the water vapor. P , is:
- + - -1 = b*P bP- - 1 bP ai,? r dr D ae
(3)
Inserting the dimensionless parameter A as before, with the boundary conditions: ,i=O,e=
m , r s R
A=1,8=0,
r < R
A=O,B=B,
T = K
the solution is of the form.
(4) where k , is defined by the roots of the Bessel function
J,(knR)
1919
=
0
(5\
1920
ANALYTICAL CHEMISTRY
-4 typical example has been solved using these equations. A buret 61 em. in length and 16 mm. in diameter is considered to be filled initially with dry air. The water available to saturate this air is a t 30" C. where the vapor pressure, Po,equals 31.8 mm. of mercury and the diffusion coefficient of water vapor into air, D, equals 0.2 sq. cm. per second. Figure 1 illustrates the case where the sides are not wetted and water is present only a t the base of the buret, and Figure 2 shows the other extreme where the sides of the buret are covered by a film of water. A time of 2 to 3 seconds suffices to saturate the air in the clean buret, whereas in the buret whose sides are not wetted, the air is not saturated even after 4 hours. Because most burets become dirty or nonwettable because of a layer of oil or grease on the glass, it is suggested that the water
used to saturate the gas contain a small amount of some neutral detergent. This procedure should keep burets cleaner and consequently decrease any analytical error from insufficient water vapor saturation. The calculations presented in this paper are valid only for the limiting cases where all water vapor is transported through the gas in the buret by molecular diffusion. I n any actual case, the gas entering the buret creates a large amount of turbulence which will aid in equalizing any water vapor pressure gradients. Thus the time required for saturation will be less than calculated; the amount less will depend on the actual case under consideration. RECEIVED for review June
24, 1953.
Accepted August 20. 1953.
Spectrophotometric Determination of Copper in Ores with 2,Z'-Bipyridine J . P. MEHLIG
AND P. L. KOEHRISTEDT1 Oregon S t a t e College, Corvallis, Ore.
the properties of 2,2'-bipyridine (2,2'B bipyridyl)described along with its method of preparation. Tartarini L.~U
(1)
(10) in discussing new color reactions involving cuprous salts reported that it forms a cupric complex which can be reduced with hydroxylamine in ammoniacal solution to give a highly colored cuprous complex. M o s s and llfellon (8)developed a colorimetric method for the determination of iron with 2,2'-bipyridine and Mehlig and Shepherd ( 7 ) applied 2,2'-bipyridine to the spectrophotometric determination of iron in ores. In a study of the 1,lO-phenanthroline-cuproussystem Moss and Mellon (9) stated that 2,2'-bipyridine, which contains the same cyclic -N-CC-Sgrouping, is not as sensitive as 1,lO-phenanthroline in the formation of a colored cuprous comple.;. but made no copper determinations with it. The purpose of the work described in this paper was to obtain further proof that macro constituents may be satisfactorily determined spectrophotometrically by application of 2,2'-bipyridine to the determination of copper in ores. APPARATUS AND SOLUTIONS
All spectrophotometric measurements were made n-ith a Beckman Model B spectrophotometer. 2,2'-Bipyridine. A solution made by dissolving 1 gram in 0.2M hydrochloric acid and diluting to 1000 ml. with distilled water. Hydroxylamine Hydrochloride. .4n aqueous solution containing 10 grams per 100 ml. Methyl Carbitol (Diethylene Glycol Monomethyl Ether). Commercial grade. Standard Copper Solution. One gram of electrolytically pure copper pellets was dissolved in 10 ml. of concentrated hydrochloric acid and 5 ml. of concentrated nitric acid and the solution x a s transferred to a 1000-ml. volumetric flask. The solution was neutralized with 6 X ammonium hydroxide until the first indication of the formation of the blue cupric ammonia complex, diluted to the mark a t 20' C. with distilled water, and thoroughly shaken. By means of a microburet 5 ml. of this solution were transferred to a 100-ml. volumetric flask, diluted to the mark at 20" C. with distilled water, and thoroughly shaken. Each milliliter of this solution contained 0.05 mg. of copper. THE COLOR REACTION
To produce the color system the volume of the standard copper solution required to give the desired concentration of copper was measured with a microburet into a 50-ml. volumetric flask. Then, in order, were added 2 ml. of 6.44 ammonium hydroxide to form the cupric-ammonia complex, 10 ml. of 2,2'-bipyridine solution, 1 ml. of hydroxylamine hydrochloride solution to reduce the copper to the cuprous state, and 20 ml. of methyl carbitol as a stabilizer. The mixture was diluted to the mark a t 20" C . with 1
Present address, Hanford Works. General Electric Co., Riohland, Wash.
distilled water and thoroughly shaken. The orange-brown color developed immediately. ill1 transmittance measurements were made with a 1-cm. Corex glass cell a t a wave length of 430 mw, the wave length of maximum absorption, after adjustment of the instrument so that the transmittance of the blank solvent containing the reagents was lOOyo. That Beer's law is obeyed bv the color system was proved by the straight line which resulted when the extinctions for six solutions containing 1, 2, 3, 4, 5, and 6 mg. of copper per liter were plotted against the respective concentrations. Above 6 mg. per liter the line began to curve.
Table I. Sample So.
1 2 3 4
5
6 7 8 9 10 11 12
Results Obtained with 2,2'-Bipyridine
Nature
Ore Ore
Ore OX Or?
OI? Oxide Oxide Oxide
Oxide Matte Matte
Copper by Copper by 2,2'-BiIodide pyridine Method, % Method, 70 10.43 11.16 12.04 20.33 19.40 18.63 15.02 14.00 13.23 22.31 21.61 14.09
10.48 11.16 12.00 20.37 19.31 18.61 15.05 14.03 13.22 22.30 21.63 14.02
Difference,
Error,
%
70
+0.05 0.00
+0.48 0.00 -0.33
-0.04 fO.04 -0.09 -0.02 f0.03 +0.03 -0.01 -0.01
+0.02 -0.07
+0.20 -0.46
-0.11 +0.20 70.21 -0.08 -0.05 +0.09 -0.50
DETERMINATION OF COPPER IN ORES
An accurately weighed sample of ore, varying from 0.1 to 0.2 gram depending upon the copper content, was heated with a mixture of 10 nil. of concentrated hydrochloric acid and 5 ml. of concentrated nitric acid until solution was complete or only a white siliceous residue remained. Iron was removed by double precipitation with 15M ammonium hydroxide. The filtrate in a 1000-ml. volumetric flask v a s acidified with 6M hydrochloric acid, then neutralized with 6 M ammonium hydroxide to the first appearance of the cupric ammonia complex, diluted to the mark a t 20" C. with distilled water, and thoroughly shaken. An aliquot of 4 ml. of this solution was measured with a microburet into a 50-ml. volumetric flask and the procedure from this point for the determination of the transniittancy at 430 mp was the same as that described ahove. The percentage of copper was calculated by use of the extinction coefficient, which had been determined by obtaining the transmittancy a t 430 mp of a series of solutions containing 1, 2, 3, 4,5, and 6 mg. of copper per liter. RESULTS
The method was applied to the determination of copper in six ores, four oxides, and two mattes. Duplicate determinations were made for each sample and at least two aliquots were analyzed for each duplicate. The results are shown in Table I along with the values obtained by the iodide titrimetric method (3).