ANALYTICAL CHEMISTRY
1914 where mole percentage of GeHl in H? null-point reading, divisions, after subtracting the zero reading 2’ = temperature, degrees Kelvin P = barometric pressure, mm. of mercury
T7 d
= =
The accuracy of this method, with the germane-hydrogen mixture made up as described and the 500-mm, is
*0.05’ % germanel and the range is to 15% germane’ The main advantage of this method is the rapidity with which analyses can be performed, about 10 minutes per determination,
an important matter when a large number of samples have to be analyzed. LITERATURE CITED
(1) Adams, L. H., J . Wash. A c a d . Sci., 5, 265 (1915). ( 2 ) Candler, C., “Modern Interferometers,” London, Hilger and watts, 1951. (3) Paneth, F. .1.,and Rabinonitsch, E., Ber., 58B, 1138 (1924). RECEIVED for review June 4. 1953. Accepted .kuguSt 10. 1953. Research supported jointly b y Army. S a r y , and Air Force under contract with hlassacliusetts Inbtitute of Technology. Work done as part of a thesis submitted by Paul H. Robinson in partial fulfillment of the requirements for the degree of ma-ter of >ripnce.
Mass Spectrometric Analyses of Some Six- and Seven-Carbon Alcohols V. -4. YARBOROUGII Carbide and Carbon Chemicals Co., Division of Union Carbide and Carbon Corp., South Charleston, ECEST
advances in the design of commercial mass spectrom-
R eters---e.g., close temperature control of the source, suppression of metastable ions, fabrication of metal inlet systems that eliminate the use of stopcock lubricant, and accurate measurement of pressure in the 0- to 100-micron range (12)-permit the examination and analysis of compounds that, heretofore, could not be studied. The last-mentioned improvement, particularly, has widened the scope of masc spectrometers considerably. Previously, more cumbersome micromanometers ( 1 4 ) or calibrated micropipets ( 3 ) were required, which often were not entirely dependable for mixtures with vapor pressures less than 50 microns.
K’. Va.
test that method, its accuracy, and its reproducibility, three mixtures containing knom-n amounts of 3-heptanol, 2-heptanolj 4-heptanol, P-eth>-l-l-butanol, and 1-hexanol were analyzed. These results are summarized in Table I. The reproducibility is within +5% of the value of the contained component. The average deviation of a given component from its known value (for all three mixtures) is 8.2%. The maximum single deviation is 19.3% of its known value. The mass patterns for these five alcohols are presented in Table 11.
APPARATUS AND PROCEDURE
A Consolidated mass spectrometer Model 21-101, convei ted to Nodel 21-103, was used for all analyses. The samples were introduced into the instrument through a mercury orifice (21, which was modified in these laboratories by &I. L. RlcTeer and G. E. Mellen to the design presented in Figure 1. A capillary micropipet (S),pictured in Figure 2, was used to introduce the samples through this mercury orifice into the preleak bulb. When a sample was to be introduced, the Teflon plug was removed, and the container holding the liquid to be analyzed was inserted in the orifice. The Teflon plug was broken away from the vacuum seal by the disk threaded to the shaft of thp plug. The pressure of the mercury above forced the liquid out of the pipet into the inlet system of the spectrometer. The size of the micropipet (about 0.0005 ml.) was such that, upon complete expulsion and expansion of the liquid into the 3-liter, preleak bulb, the desired pressure (about 30 microns) was attained. From two viewpoints this system had a decided advantage over the older technique which employed a sinteredglass disk under a mercury seal (10): The time of introduction was shortened considerably; and difficulties due to adsorption in the sintered glass and loss from fractionation, both of which are characteristic of the sintered glass disk, c ere greatly minimized. The sample passed directly from the orifice to the 3liter, preleak bulb, in which the pressure m-as measured r ~ i t ha micromanometer ( I d ) . Compounds containing a hydroxyl group tend to be adsorbed on the walls of the inlet system through hydrogen bonding. For samples containing alcohols, therefore, the system had to be saturated several times with the mixture in question. This was accomplished by introducing the sample into the mass spectrometer, leaving it there for 2 minutes, and then pumping the sample out again. Usually, after several such “saturation” treatments, the mixture could be reproducibly analyzed. A “pump-out” period of 10 minutes between runs is adequate.
-
1
TT‘8”
WAX JOIN-
1
8n
RESULTS
The analyses of mixtures of oxygenated compounds-e.g., alcohols, ketones, aldehydes, etc.-have been reported ( 4 4 , 9, 11, 13). The present paper describes a method developed for the analysis of mixtures of six- and seven-carbon alcohols. To
Figure 1. Diagram of Mercury-Covered Orifice
V O L U M E 25, NO. 12, D E C E M B E R 1 9 5 3
1915
Another, somewhat less accurate but much faster, (Weight per cent) method is to use the 87 and I II - ~ _ _ I11 _ _ _ _ ~ 7 3 mass peaks to remove the MeasDeviaIleasDepiaMeasDeliamajor portion of the contriComponent Known ured tion Kno\!n ured tion Knoan ured tion bution of 3-heptanol and 432.5 +1.8 12.4 30.7 3-Heptanol 8.0 7.6 -0.4 10.9 -1.j 11.9 iO.4 4.1 1.7 11.5 4-Heptanol 8.0 8.9 +0.9 -0.6 heptanol to the other peaks. -0.8 30.7 39.9 16.6 1-Hexanol 67.9 66.8 -1.1 13.4 -3.2 These two peaks are virtually 12.8 il.1 63.3 11.7 2-Ethyl-1-butanol 8.1 8.4 +0.3 67.6 i4.3 15.4 12.9 -2.5 3.G 3.4 +0.4 2-Heutanol 8.0 8.4 -0.2 free of interference from the Average deviation 0.62 1.32 1.96 other constituents. Simultane________~_ ~ _ _ ous equations involving the 56, Table 11. $lass Patterns for Some Six- and Seven-Carbon 70, and 45 peaks next are em4lcohols ployed to separate 1-hexanol, 2-ethyl-1-butanol, and 2-teptanol, .?za5s to and a correction for their small contribution t o the 73 and 87 Charge 12-Ethyl243Elexanol I-butanol Heptanol Heptanol Heptanol n-Butane Ratio peaks then is made. 14 1.75 3.12 0.85 1.19 1.19 15.63 Under electron-beam bombardment, secondary alcohols rupture 85.88 13 8.58 14.91 5,59 8.44 7.40 preferentially a t the carbon-to-carbon bond immediately ad16 0.40 0 60 0.44 0.36 1.72 1.72 3.84 3.26 1.21 0.41 19 jacent to the carbon to which the hydroxyl group is attached. 72 28 26 5.46 10.22 1.55 3 23 4.34 The resulting galvanometer deflection is not always the most 358 13 27 95.65 17.32 46.72 39.62 4;. 82 intense in the mass spertrum, but is characteristic for a specific 278 03 28 19.18 3.26 12.42 8.95 o.29 3.51 25 26.76 29 98,55 15.09 44.61 37.74 secondary alcohol. 1-Hexanol is characterized by large 56 and 7 50 1.03 30 1.97 3.59 2.55 0.47 31 41.51 0 31 51.38 28.26 77.54 4.67 31 peaks; 2-ethyl-l-butanol, on the other hand, exhibits a large 32 0.45 1.38 1.47 1.65 0.12 0.42 70 peak as well as substantlal galvanometer deflection at a mass0.05 33 0.14 0.44 1.68 0.41 to-charge ratio of 84. The following equations represent plaus2.15 23.75 4.58 0 65 1.42 1.14 38 141.66 lL43 24.45 51.68 4.98 16.61 39 ible explanations for the formation of some of these fragments: 4.82 18.81 2.37 1 44 6.59 2.22 40 41 50.00 16.22 56.77 257.19 94.38 23.95 CH,(CHJ,C'H,OH e --f [C4H8]+ CHICHZOH 2e (1) Table I.
Mass Spectrometric Anal3 ses of Six- and Seven-Carbon 4lcohols
~
42 43 44
4:
4.48 60.75 13.05 8.12 8.21
5.96 28.55 6.38 14.91 0.35
100.63 798.75 26 72 0.50
1.10 2.11 1.38 0 65 139.78 16.71 9.10 8.74 8.09 4 72 1.06 1.18 0.08 2 50 1.11 4 15 I 27 2.16 0.28 10.72 0 48 0-
2.16 0.82 13 74 6.17 17.04 9.21 LOC3l
i.50 1.81 8.53 6.56 20.06
7 10 19 48 8.83 100.00 ___ 2 17
51.21 77.29 5.64 4.23
39.18 289.13 11.61 10.61
3.20 3.54 56.26 100.00 9.47 0.53 0.63 24.26 3.40 2.64 0.36 1.56
8.71 3.68 91.85 43 42 14.47 0.83 4.22 24.67 100.00 S? 79 .I 13 1.82
1.28 9.04 0.00 102.17
0.76 30.76 0.00 102.17
7.33 0.78 0 00 116.20
0.23 0.11 0.07 116.20
0.63 0.88 29.28 116.20
62.84
31.12
177.76
82.04
75.93
+ CH3(CHJ,CH,OH + e -+ [C'H,OH]+ + [CbHiJ + 2e (CZH,)?C'H-CH,OH + e -+ [ C ~ H I ? ] HzO + 2e (C'?HS)?CH-CH~OH + e --+ [(',Hi01 + CHiOH + 2( +
+
+
-&
+
53 . - 4i _ D.)
i ti
57 58 .5 9 69 70 71 72 73 83 84 87 Molecular weight Sensitivity diy./ micrun
;::)
4.03
6 67 .' 72 43 2.31 0.77 2.22
58.12 10.53 (43 sensitivity = 84.11)
IS0 mm.
i
'I/"
W N D TIP PS S H M P POSSIBLE
BQ)E ff CENTER IF USED) IS UNDSTCRED LD. OF
DISCUSSION
Simultaneous equations were utilized for the analysis of mistures of 1-hesanol, 2-ethyl-1-hutmol. 2-heptanol, 3-heptanol, and 3-heptanol. The contribution of a given component to a selected mass peak 'ii'ap obtained I)?. the use of determinants. Five gencml inverse solution3 n-ere set u p \\-hioh, once solved, permitted the resolution of these six- and seven-carbon alcohol niistures in less than an hour. The pattern coefficients used for the analyses presented h c w i n :ire given i n Table 111. -_
-
-
__
.~
~.
Table 111. Pattern Coefficients m/e Tr 6 70 4R
87 73
1-
I-Iexanol 1.0000 0.0340 0.0423 0.0000 0.0156
(4)
100.00
Ionizing voltage. 70 volts Ionizing current. 10 iiiicroainperes Ion source temperature. 250' C. Magnet current. 0.605 ampere Basis of pressure nieasiirernent. Jlicroinanometer
-
(2) (3)
2-Ethyl1-butanol
2Heptanol
3Heptanol
4Heptanol
1.0000 2.3031 0.2674 0.0000 0.0419
1.0000 0.5180 12,3610
1.0000 1.009: 2 4165 4.7453 0.3647
1 0000 0.1220 0.8923 0 0077 10.9890
0,0000
0.0593
SPECIFIED NBlNO
PULLED DOWNm
600
RED-
\
BQ)E
119
/Ls POSSIBLE
Figure 2.
lricropipet
The analyses of other oxygenated compounds, and of aromatic and aliphatic hydrocarbons, have been discussed (1, 7-9, 11, 13). The use of the mercury orifice ( 2 ) and niicromanometer ( 1 2 ) described herein, in conjunction with a metal inlet system, however, makes such analjses much easier to carry out with accuracy and precision. LITER4TURE CITED
(1) Brown, R. A., Taylor, R . C., Nelpolder, F. W,, and Young, JT. S.,. ~ X A L . CHEV, 20, 5 (1948).
ANALYTICAL CHEMISTRY
1916 (2) Charlet, E. -M., and Harris, R. G., Consolidated Engineering Corp., Pasadena, Calif., Consolidated Group Rept., 74 (Feb.
17, 1950). (3) Friedel, A. R., Sharkey, A. G., and Humhert, C. R., ANAL. CHEM.,21, 1572 (1949). (4) Gifford, A. P., Rock, S. hI., and Comaford, D. J., Ibid., 21, 1026 (1949). (5) Kelley, H. M., Ibid., 23, 1081 (1951). (6) Langer, A., and Fox, R. E., Ibid., 21, 1032 (1949). (7) Lumpkin, H. E., and Thomas, B. W., Ibid., 23, 1738 (1951). ( 8 ) Johnsen, S. E. J., Ibid., 19, 305 (1947). (9) Taylor, R. C., Brown, R. A,, Young, TV. S., and Headington, C. E., Ibid., 20, 396 (1948).
(10) Taylor, R. C., and Young, IT.S., IND. ENG.CHEM.,A N ~ LED.. . 17, 811 (1945). (11) Thomas, B. W., and Seyfried, W.D., Ax.4~.CHEY.,21, 1022 (1949). (12) Washburn, H. W.,Berry, C. E., Robinson, C. F., Gifford, A. P., and Rock, S. XI., Consolidated Engineering Corp., Pasadena, Calif., C.E.C. Poceedings, 5, S o . 4, 6 (1951). (13) Washburn, H. W., Wiley, H. F., Rock, S. If.,and Berry, C. E., ANAL.CHEY.,17, 74 (1945). (14) Young, W. S., and Taylor, R. C., Zhid., 19, 133 (1947). RECEIVED for r e r i e v M a y 8, 1953. Accepted August 18, 1953. Presented a t the Conference on Analytical Chemistry and Applied Spectroscopy. Pittsburgh, Pa., 1953.
Reference Electrode for Potentiometric Titrations in Glacial Acetic Acid R. A. GLENN, Coal Research Laboratory, Carnegie Institute of Technology, Pittsburgh, Pa. HE
q
titration of nitrogen bases in glacial acetic acid with a
Tstandard solution of perchloric acid in the same solvent has proved to be an indispensable technique in the analysis of petro-
leum ( r ) ,shale oil ( r ) , coal hydrogenation oils (IO),pharmaceuticals ( 8 ) , and many other organic nitrogenous compounds (1-4,6). In the titration of colorless compounds the end point may be readily determined by use of color indicators, but with colored compounds i t must be determined potentiometrically. The choice of electrodes for making potentiometric titrations in nonaqueous media, however, presents a problem. The glass electrode, in conjunction with a sleeve-type calomel reference electrode, both with and without a salt bridge, has been used by several workers (6-9), but not without difficulties resulting from the high sensitivity of the system to stray currents and from contamination of the cell by the solution being titrated. To eliminate the use of a salt bridge and the difficulties encountered in the use of the calomel electrode, Fritz ( 9 ) substituted a silver wire coated with silver chloride for the sleeve-type calomel reference electrode. This paper reports further on the use of the silver-silver chloride electrode in conjunction with the glass electrode and how the instability of the electrode pair in the region of the end point may be obviated. When the silver-silver chloride electrode is usedin conjunction with the glass electrode and the two are immersed directly into the solution of the sample being titrated potent io me t r ic a1 1 y , very erratic e.m.f. readings are observed in the region of the end point (see Figure I), thus making its determination doubtful. When the same electrodes are used in a continuous automatic titration using the apparatus of Katz and Glenn ( 5 ) ,a reversal in the change of the observed e.m.f. occurs in the region of the end point (see curve A , Figure 2), which again renders its location difficult. This reversal in the change of the observed e.m.f. results from the fact that the potential of both the glass and the silver-silver chloride electrodes in the region of the end point changes markedly but in the opposite direction.
r, I
VOLUME OF TITRANT ADDED, ml.
Figure 1. Manual Titration Curve Silver-silver chloride electrode used in conjunction with electrode Sample. Coal hydrogenation neutral distillate oil
glass
300
/Paraffin
400-
Seal
-
Commercial Silver Silver Chloride Electrode
3 sool Glacial Acetic Acid Saturated with KCI 600-
5 24/40 Silver Button Coated with Silver Chloride KCI Crystals
700-
Glass Wool Plug 0
20
40
€0
80
100
120
CHART DIVISIONS
TITRANT ADDED 1.91 ,ue/div.
Figure 2. A. B.
Continuous Automatic Potentiometric Titration Curves
Silversilver chloride electrode dipping directly into solution being titrated. Sample, quinoline Silversilver chloride electrode placed in isolation cell. Sample, coal hydrogenation distillate oil
Fignre 3. Reference Electrode Isolation Cell