Table VI.
Calculated Effects of Changing Experimental Conditions on Range of Absorption Measurable by Monochromatic Technique
Xone CuKa Geiger
Radiation Detector
t
P1 Pn
Pa
- PI
% Cl increase in absorption range
0.15 38.5 56.2 0.46
0
results comparable in accuracy to those obtained by either the polychromatic or the internal standard technique. When compared with the polychromatic technique i t is not susceptible to the errors introduced by the presence in the sample of elements on the high atomic number side of the first absorption edge of the characteristic radiation used. These errors may be fairly large before there is a n y reason to suspect them. The monochromatic technique is not affected by the random errors associated with approximate proportionality of absorption coefficients a t different wave lengths. The main limitation of the monochromatic method is the restricted range of absorption it can measure, but samples possibly subject to this limitation will be detected by the absorption measurement. About 5% of samples received in this laboratory for quartz analysis fall in this category. When the diffraction-absorption methods and the internal standard method are compared, preparation and analyt-
Changes from Original Conditions Detector Sample and sample thickness thickness Radiation CuKa CuKa RIoKa Proportional Geiger Geiger 0.10 38.5 65.0 0.69 50
0.10 38.5 84.3
0.15 4.08 21.8
1.19
4.34
160
840
ical times are about the same. If the amount of sample available is much less than 500 mg., i t should be mixed with starch to produce the necessary bulk to fill the slide cavity completely for diff raction-absorption analysis. Internal standard analysis may be made with as little as 100 mg. of sample by filling the remainder of the cavity with starch. I n general, the choice of method for other quantitative problems will be determined by the expected range of sample absorption, the presence or absence of elements affected b y absorption edges, and m-hether i t is desirable to dilute or adulterate the sample. For determining only one component of mixtures, any one of the three techniques may be suitable. I n some cases i t may be difficult to find a suitable internal standard and this problem will become more serious if more than one component is to be determined. Black ( 1 ) has described a method for determining a number of components in multicomponent mixtures. This
method is fast and accurate but is probably limited in its application, as it was used for a specialized analytical problem, in which a number of factors combined to produce the speed and accuracy. I n the general case of multicomponent analysis, diffraction-absorption may be the only suitable method of obtaining an accurate quantitative determination. ACKNOWLEDGMENT
The author wishes to acknowledge the constructive criticism and suggestions of Leroy Alexander of the Mellon Institute and of Kingsley K a y and Jean Leroux of this laboratory. The assistance of Claire Powers with sample preparation and experimental determinations is also gratefully acknowledged. LITERATURE CITED
Black, R. H., ANAL. CHEU. 2 5 , 743 (1953).
Kay, K Am. Ind. Hyg.Assoc. Quart. 1 1 , 125 (1950). Klug, H. P., A N ~ LC. H E K 2 5 , 704 (1953).
Klug, H. P., Alexander, L. E., “X-ray Diffraction Procedures,” p. 311, Wiley, New York, 1954. Leroux, J., Lennox, D. H., Kay, K., ANAL.CHEW2 5 , 740 (1953). Lonsdale, K., Acta. Cryst. 1, 12 (1948):
( 7 ) Parrish, W.,Hamacher, E. A, Lowitzsch, K., Philips Tech. Rev. 16, 123 (1954). (8) Taylor, J., Parrish, W , Rev. Sci. Instr. 2 6 , 367 (1955). (9) Wilchinsky, Z. W.,Acta Cryst. 4, 1 (1951).
RECEIVED for review July 6, 1956. ACcepted January 25, 1957. Pittsburgh Conference on Analytical Chemistry and Applied Spectroscopy, Pittsburgh, Pa., February 1956.
Mass Spectra of Trimethylsilyl Derivatives A. G. SHARKEY, Jr., R. A. FRIEDEL, and S. H. LANGER Bureau o f Mines, U. S. Departmenf o f the Interior, Region V , Bruceton, Pa.
,The mass spectra of 26 aliphatic trimethylsilyl ethers and eight related silicon compounds have been obtained. Fragmentation peaks have been correlated with molecular structures. Structures have been proposed for rearrangement ions appearing in the spectra. By the use of trimethylsilyl derivatives, compounds of several classes including alcohols, phenols amines, and thiols can b e determined in the presence of compounds that normally interfere.
770
ANALYTICAL CHEMISTRY
A
mass spectrometric methods have been used to analyze mixtures of alcohols ( 1 , 6, 7 , 9, 16, 17), determination of individual CS to Clo primary alcohols in the presence of hydrocarbons requires preliminary separation before chemical or spectroscopic methods can be applied. A method, utilizing the trimethylsilyl ether derivatives of alcohols, has been developed for rapid, direct analysis of individual alcohols in hydrocarbon solutions (12). The trimethylsilyl ether derivatives of LTHOUGH
alcohols, (CH3)3SiOR,are readily prepared (11) and produce distinct mass spectra free of interference b y hydrocarbons. The mass spectra of 26 aliphatic trimethylsilyl ethers and eight related silicon compounds mere obtained. Trimethylsilyl ethers have higher volatility than hydrocarbons and oxygenated compounds containing fewer carbon atoms. For example, the mass spectrum of a Clo alcohol from a conventional room-temperature mass spectrom-
Table I.
Mass Spectra of Trimethylsilyl Derivatives
Compounda Carbons Molecular weight
-
Methylb 4 104
Ethyl 5 I18
Propyl 6 132
8 7 . 6c 40.7
138 30.2 210 334 82.1 333 83.4
198 29.0 207 296 73.0 130 83.7 ... 1.8 1.6 567 1040f 1.0 5.1 5.3 12.5 I99 19.5 5.3 9570 103
[RQ-OSi( CH,),] Primary Straight-Chain I-Butyl 1-Pentyl ]-Hexyl I-Heptyl 7 8 9 10 146 160 174 188
I-Octyl 11
1-Nonyl 12 216
1-Decyl 13 230
278 17.0 194 201 75.6 117 107
174 8.9 137 112 42.8 65.7 68.4
113 6.1 92.0 58.5 26.4 38.3 40.8
54.9 2.8 65.0 34.9 15.1 22.5 24.4
...
. . .
...
...
1.6 36.4 490 984 5.4 40.2 170 67.8 277 25.3 47.9 4.0 0.7 ... 15.1 I .4 1.3 ... 1.3 4.4 ... 1.6
1.4 133 290 569 14.8 20.4 97.5 46.0 164 15.9 21.3 2.4 0.3 1.9 2.7 0.2 1.2
0.1 70.0 I73 352 46.7 11.4 54.7 29.2 103 10.1 15.4 1.7 0.2 0.6 1.7 *.. 1.3
0.7 26.8 105 212 35.4 6.6 33.5 20.8 60.4 5.6 9.1
202
Mass/charge
29 31rd 43 45re 47r 59r 61r 65.5 66 69 73 75r 83 87
89 101 103 104
115
117 118
130 131 132 145 146 147 I59 160 161
162 173 174 187 188 201 202 215 216 230 Sensitivity of base, div./micromole
180
149 11.4 867 39.5
...
1.1
0.7 41.2 19.6
...
1.1 1.1
313 750
...
...
1236j.0 ~
. .
...
P 3.9h
...
12.1 2.7 11.0 114Of.Q 109
...
. . . .
5.5 P 9.8 ...
. .
I
... ... ... ... ...
, . .
...
...
... ...
...
...
... ...
. .
...
, . .
...
...
...
...
... ...
. . .
...
. .
315
...
20.4 114 9.7 217 21.1
12.2 2.5 ... ...
.
947. g 112 5.0 P 8.1
...
...
...
...
...
... ...
...
...
...
2.0 484 1000f
, . .
... ...
, . .
...
...
4.4 P 6.1 ...
...
. . ..
.
179 23.6 151 262 79.0 109 86.0 ..
.
.
...
t
... ...
...
... ... ...
...
...
...
... ... ...
... ... ...
107
110 ...
0.9 84.4 12.7 551 528 1080f 1050 ~I .6 78.8 12.3 45.6 192 179 60.3 54.9 282 281 27.0 27.5 8.9 45.6 2.8 9.6 1.2 0.4 1 .o
1.9
...
1060g 140 35.7 4.4 I' 13.0 1.7 0.4
...
261
248
Compound" Carbons Molecular weight
275 20.5 189 241 87.8 125
...
...
270
267 33.5 I77 269 75.4 7 28
2-Me1-propyl 7 146
2-MeI-butyl 8 160
3-MeI-butyl
130 21 225 245 65 92 80
231 17.8 141 208 66.5 84.3 80.9
150 17.3 191 219 70 8 115. 78.6
8
160
...
2.2 ...
4.4
...
7.4 11301sg 160 50.1 4.4
]09& ___
... ... ...
P 10.4
. .
...
165 2.8 P 6.6
...
...
...
... ... ... ...
...
...
286
... ...
, . .
...
...
...
291
Primarv Branched 2,2,Dime2-MeI-propyl I-pentyl 8 9 160 174
8.0 0.5
...
... ...
3.2 ... 686f3Q 112 ... P 2.3
80
...
.. ,
... ...
...
...
290
280
3-MeI-pentyl 9 I74
2-EtI-butyl 9 174
229 22.3 145 193 69.3 111 78.6
254 18.2
1.o
...
... 1.7
...
2.9 ... 1.0 3.1
, . .
... 0.6
... ... ...
... ...
...
... ...
... ...
1.0
...
... ...
438; 92.9 ... P 2.5 IO
, . .
2651 __ 9 0
49.3 P 1.0
...
322
309 2-EtI-hexyl 11
202
Mass/charge
29 31rd 43 45re 47r 5% 61r 65 5 66 60 73 r .
i or
83 87 80
101 103 104 115 I17 118 130 131 132
5 2 1 2
1 1 1 6 810
looof
1.4 4 1 10 4 6 2 444 42 2 7 1
1.1 7636 92 4
...
0.8
10.9 687 943, 1.6 4.4 47.2 15.8 495 47.6 4.2 1.6 0.3 6.4
...
...
197 17 150 194 59 83 G5
...
0.9 128 517 674 1.4 21 .o 523 31.3 630 30.6 9.3 2.5 0.4 1.1 . .
1.8
9
6 7 8
0 7
3 3
031f -__
669 1.6 4 1 52 0 28 1 572 59 6
1.1
..
... ...
1.0
176 17.1 208 186 64.3 80.7 82.4
...
1. o 20.0 675 897f 22.6 5.2 54.2 13.0 525 51.8 13.3
-
8.1
0.9 1.4 2.4 0.3
... 1. o
4.1 455 536 619f 19.1 326 24. I 274 26.5 23.8 5.5 I .o 1.o 1 .o
...
304 14.1 211 184 207 I77 74.7 68.3 86.7 92.4 84.3 98.1 ... ... 1.2 1.7 18.0 74.0 747 761 1010f IOOOf -13.8 34.8 6.6 8.7 53.5 73.8 13.4 18.1 557 686 56.3 69.7 9.7 13.6 15.6 14.9 1.5 1.2 1.7 3.3 4.3 1.5 *.. 0.3 (Continued on paye 7 7 2 )
~____
VOL. 29, NO. 5 , M A Y 1957
0
771
Table I.
Mass Spectra of Trimethylsilyl Derivatives (Continued)
Compounda 145 146 147 159 160 161 162 173
2-Me1-propyl 6 4 P6 3
2-Me1-butyl 73l e 95 9
32 5 3 6 P9 5 1 4 0 .4
3-Me1-butyl 797f.g __104 34 4 2 5 P3 1 0 4 ...
174
202 215 216 230 Sensitivity of base, div./micromole
6lr
65.5 66 69 73 75, 83 87 89 101 103 104 115 117
3-Me1-pentyl 3 5
1 8 3 0 423' 59.7 18 9 1 .7
1 2 P5 9
, .
.. ...
1 6 584s 82 8 26 2 2 .2 2 0 P3 0
243
194
...
...
...
, . .
... ...
. .
. . .
...
...
...
...
.,. .
I
.
208
255
Secondary 3-Me2-Pentyl 2-butyl 8 8 160 160
2-Propyl 6 132
2-Butyl 7 146
89 6 22 6 215 313
150 21.9 178 284 74 6 92 3
116 18 I 190 257 80 5 88 1
84 8
881
103 17 7 225 239 63 7 86 6 68 7
...
...
1.0 3.6 1010 818 0.8 19.4 3 1 29 3 17 5 2 2
0.8 ?J , 0 1160 565 0.8 24.0 2.9 29.2 3.9 0.7 1.5 1 13i 1.7 0 i
800
82 3 88 G ... 0.9
1 5 699 12001 0.7 5.3 1 6 14 5 0.7 5.1 94308i 102
...
15.1 I' 1 9
...
1.0 1.4 10301 82 1 0.6 23.6 2.4 19.9 2.7 0.6 6.2 831i 104 1.6 3150 42.4
4.7 1070f. ___
127
6 1 1 4 2790 40 2 15 3
l i O J 3 '
292
2-Heq 1 9 174 260 19 5 175 328 94 5 87 1 110 ...
0.9
25.6 1000 834 3.4 27.4
3-Heptyl
2-Et1-hexyl 13 4 2 5 4 2 1 8 0.4 3 6 6230 102
10
ANALYTICAL CHEMISTRY
r
101 18 2 158 254 76 3 58 5 91 8
107
...
1.7 45.9 1310' 747 3.0
9.2
26.2 28.0 3.2 28.1 1300f 142 l i 1 3
37.5 83.4 8.1 56.2 6.9 0.7 12.6 891i 105 50.2 6.5 6 8 570 80.4 25.5 2.3 1129 17.2
... 305
Tertiarj 2-hie2-Me2-propyl 2-butyl
349
R. .. 6
1 2
308
146
20 6 141 253 86 7 117
4
..
188
5.... 0
, .
182g 30.4 11.2
210
P ii:9 146 ... 147 ... 2370 159 I' 11 . :3 33.7 P 10 8 160 10.6 ... 161 0.7 ... 162 ... 173 P 9.4 174 .. 187 P'8:8 .. 188 ... 20 1 ... ... 202 ... ... ... ... 215 ... ... ... ... ... 216 ... . . ... ... ... ... 230 ... Sensitivity of base, div./micromole 278 256 282 230 345 325 a Source 0, synthesized a t Bureau of Mines. * Alcohol from which trimethylsilvl ether is derived. Peak height in divisions per liqiid volume (0.00068 cc.) when mass 27, n-heptane = 492 div. Rearrangement peaks except for methyl. a Rearrangement peaks. f Base peak, most intense peak in spectrum. Parent mass minus 15. Parent peaks indicated by P. Ion resulting from fragmentation at functional carbon-for secondary alcohol derivatives.
772
2-Et1-butyl 3 5
P 10 ..
Compoundb Carbons Molecular R eight Mass/charge 29 31rd 43 45re 47r 59r
Primary Branched 2,2,Dime2-Me1-propyl I-pentyl 4969 2 6 65 1 22 7 2 0 5500 2 2 P7 7 78 6 1 2 24 7 0 .33 2 .1 2 0 P9 5
...
1.5 790
1190' 1.0
2.0 2.7 15.1 0.8
...
16.6 0.9 3.5 773s 86.8
...
P ...
8 160
200 19 2 200 277 85 2 15 0 104 2.0 1.5 3.3 I 140f 947 0 9
3 2 32 1
3 5 2 6
19. i 1 6 0 2 5 8 754 90 0 2920 39 3 1 7 PO 8
...
... ...
...
...
...
... 234
...
209
~
Table II.
Mass Spectra of Trimethylsilyl Derivatives and Related Compounds
[Ra-OSi( CHB)B] Compound Sourceb Carbons Molecular weight Masslcharge 29 31rd
43 45r 47r 59r 61r 65.5 66
66.5 73 75r 75.5 77 87 89 91 101
I03 104 105 115 117
n-Butyl5 Mercaptan 0 7 162 27gC 22 4 177 188 43 4 90 6 47 8 4 0 696 212 62 2 6 938' 3 5 4 GO 1
27.5 68.3 18.9 .
.
8.6
5 1
P-"
m-a Cresol
Cresol
0
0 IO
10 180
180
52.5 13.1 I14
159 29.3 54.4 20.1
...
9.9
...
76.8 16.4 166 192 37.1 42.8 38.3
...
...
8.1
...
122 120
154 108
...
...
...
65.8 2.9 28.2 192 2.1 14.2 5.7 72 5
69.3 2.7 25.0 I55 1.9 12.9 7.0 66 8 1.2 1.3
76.8 2.4 34.0 545 2.0 14.7
5.7 40.3
1.6
1.4
1.2
3.5 13.6
1.4 .
,
1.9 7.6
..
4.4
6.8
, . .
1.2 ...
8 9 1 6 717'
1 4
...
135 131 111 183 1.2 11.5
128 3 ,5
3.2 1.0 10.0 1.2
2.9
2.4
6.2
3.1 1820'pJ
... . .
...
10.8
8.5 lji&,f
. . 7.3 ...
... ... ... ...
111o"J
P 523 ...
P 5i5 ...
1.6 1.3 132 22.6 4.0 4.7 80 7 .52.3 7.0 :3 . 6
11
21 0 605
77 1 34 0 0 9
2 32 7 TI 5
5
386 58 5
'7 1 6 8
37.8 6.3 ...
11.6
... 320
1 1
114
162 75.9 24.9 109 190 16.4 189 12.2 6.6 286 42.5 33 1 40.2
18209%' 299 ... . .
1, 0
41,Z 7.4 10.8 1'7.1
21 . 0 3 2
18.3
17.0 28.8 6,7 8.8 0,3 103
16.1 39.7 2.5 .. 14.9 2390c1J ~ _ _ 4 7 , . .
1. O
P45.1 '7.9 . .
. . .
P'O.8 ..
. .
... ...
327
834 2 1
P 199 ...
n ti
. . .
0: 2 9 .5
...
, . .
...
185
109
4.4 245 38.5 18.1 200
...
. . .
...
...
131 4.7
~~
...
P 532
101
8 7 11 4 12 4 0 8 23 1 108 11 8
...
6.9
He~iiniethyldisiloxane
53.3 26.4
196 28 5 944e 136
3 2
15.9
Hexamethyldisilazane 0 6 161
71 2 22 5 132 239 54 0 138 35 3
2.6 675
6.9 60.0
8&',f
220
..
5 5 48 6 3 8
11.5
9
. . .
4 1 48 0 4 4
...
Propylenea Glycol 0
, . .
, . .
230
16.9
16.7 185 3.8
... I' 218e
Phenol0 0 9 166
54.3 12.8 119 166 25. I 48.7 17. I
101
...
180 191
'
4 8 7
7 1 3 0 0 3 6
159
R
170
198
131
205 220 Sensitivity of base, div./micromole
56.6 20.9 128
...
28 8 2 5
I60 161 162 165 166
Cresol 0 10 180
t
118 119 121 130 132 133 135 145 I46 147 151
0-a
..
.
101
32 6f PO 8 300
.
. .
...
404
tiG8
Cornpound from which trimrthvlsilyl ether is derived. 0 , synthesized by Bureau of Mines; D, Dow Corning Corp., DC-200. Peak height in divisions per liquid volume (0.00068 cc.) when mass 27, n-heptane = 492 div. Rearrangement peaks. Base peak, most intense in spectrum. Parent mass minus 15. Parent peak indicated by P.
eter is not usable, but t'he spechum of a tririietliylsilyl ether prepared from a Cl0 alcohol can be obtained n-ithout difficulty. Trinietliylsilyl ethers show several intense rearr:ingeriient, peaks in addition to the expected fragmentation peaks. Sunierous other instances of intense rearrangement peaks in mass spectra, mainly of oxygen-containing compountk, have been reported (8, 10, 14, 16). Earlier work by Dibeler on a similar silicon compound, hexamethyl-
disilosane, (CH3)6Si20, also shon-ed many rearrangement peaks (3). EXPERIMENTAL
The spectra xvxe obtained on a Con'Ohdated E1ect'roti\-nalnics'Orp.* 'lode' 21-103~ mass ples were introduced from a constantyolume pipet (o~ooo6s cc.) through a mercury orifice ( 2 ) . preparation of the alcohol derivatives has been described b y Langer (11). Physical constants for compounds previously re-
ported agree with literat'ure values ( I S ) ' Heart cuts were used in all instances for mass spectrometer calibrations. MASS SPECTRAL CORRELATIONS
Fragment Ions. PRIMARY STRAIGHTCHa1N TRIMETHTLSILYL ETHERS.Important peaks in t h e spectra of t,he trimethylsilyl ethers are given in Tables I and 11. A characteristic mode of fragmentation was found for all t h e normal (C, to Clo) aliphatic trimethylsilyl ethers : Parent VOL. 29, NO. 5, MAY 1957
e
773
Fragmentation Peaks in Mass Spectra of Normal Aliphatic Trimethylsilyl Ethers
Table 111.
Example. 1-hexyl derivative Relative Intensities
Structures
d e
43
CHI
\
-Si-(
or -C3H7)
/
17
CHI
\
CH3-Si-
73
45
/
CH3
CH3
\
CH3-Si-0-
89
16
/
CH3 CHI 103
'/ A
CH3-Si-0CHa CHa
Parent minus 15
H 25
I
H
\
- -1 CH3
CH3-Si-O-CH2-
100
x2
for the ethyl and propyl trimethylsilyl ethers, mass 89 is much less intense than for any of the C4 to Clo normal alcohol derivatives. Thus, the ion (CH3)&3iO+ is unimportant unless the hydrocarbon chain consists of four or more carbon atoms. The most intense (base) peak is the parent-minus-15 ion for the derivatives of the normal alcohols C1, C2, and C6 to Clo (Table I). Mass 75, a rearrangement ion, is the base peak for the Ca, C4, and C5normal alcohol derivatives, although the parent-minus-15 peaks are almost equally intense. P R I M A R Y BR.4NCHED TRIJIETHYLSILYL ETHERS.Below mass 103 the same major peaks appear in the spectra of the primary branched- and straight-chain alcohol derivatives. However, the pattern variation is larger for the branched than for the normal derivatives. The t n o derivatives from ?-branched alcohols, 3-methyl-l-butanol and 3methyl-1-pentanol, produce more mass 89 positive ions than other branched- or straight-chain derivatives (Table I).
CH3
or CH, Parent
\
- -CHI
CH3-Si-O-CH2-
0.9
/
A strong peak corresponding in mass to the alkyl radical minus tn o hydrogen atoms, is also in the spectra of the 7-branched derivatives.
CH3
SECONDSRY TRIMETHY LSILYL ETHERS.
The five ethers derived from straightchain secondary alcohols show intense peaks resulting from a break a t the functional carbon. r
$F
(Rp is the larger hydrocarbon group)
Y W 4
y-
-
100-
II
Si
/I \
I-
4
CH3 CH3 CH3
E
-
'I
-
29 31
lii
434547
PI
5 9 61
mass peaks are weak, but a very intense peak resulting from the loss of one of the four methyl groups appears 15 mass units below the parent mass. Peaks a t masses 73, 89, and 103 also are characteristic of this series, and these peaks have been correlated with the trimethylsilyl ether structure. The spectrum of the 1-hexyl ether is given as an example in Table 111. Mass 43, which is an important peak in the mass spectra of many compounds, is a major peak for the trimethylsilyl ethers, ANALYTICAL CHEMISTRY
I
I' '2
73
75
Figure 1 . Pattern deviation, normal trimethylsilyl ethers
774
l+
150 O O L
89
Cb
to
~
103
CIOaliphatic
Patterns of the C, to Clc normal aliphatic trimethylsilyl ethers are similar for normal fragmentation masses 103, 89, 73, 43, and 29, and for rearrangement masses 75, 61, and 45 (Figure 1). The methyl, ethyl, branched, and secondary trimethylsilyl ethers do not show this pattern similarity. The intensity of mass 89, (CH3)3PiO+, is different for the C1, Cz, and Csnormal aliphatic compounds. For the methyl derivative, mass 89 is the garent-minus15 ion and is therefore very strong;
This fragmentation produces the base peak a t mass 117 for 2-pentyl and 2hexyl trimethylsilyl ethers and a strong peak a t mass 131 for 3-type trimethylsilyl ethers. The branched-chain secondary alcohol derivative (3-methyl-2-butanol) behaves similarly by losing the larger alkyl group and producing the second most intense peak in the spectrum at mass 117. 1Iass 73 is more intense for the straight-chain secondary than for the corresponding primary alcohol derivative and is the base peak for 2-butyl and 3-heptyl trimethylsilyl ethers. Parent-minus-15 mass peaks are less intense for the derivatives of the secondary alcohols. TERTIARY TRIMETHYLSILYL ETHERS. The mass spectra of the trro tertiary alcohol trimethylsilyl ethers are similar. The parent-minus-15 ion for tertiary butyl and parent-minus-29 ion for tertiary amy1 produce intense peaks a t
niass 131 in both spectra. All other peaks down to mass 7 5 are weak. Rearrangement Ions. Several intense rearrangement peaks appear in t h e mass spectra of all t h e normal alcohol derivatives with the exception of methyl. N a s s 7 5 is the most prominent reariangement peak and always one of the three largest peaks in the spectrum. R a t e of effcsion determinations (4, 5 ) made on mass 75 rule out the possibility of an impurity. K i t h n-butyl trimethylsilyl ether, identical effusion rates were obtained for masses 131, 75, and 73, indicating that the mass 75 fragment is derired from the parent maqs 146. T n o series of rearrangement peaks are found in the mass spectra of derivatives of normal aliphatic alcohols. The peaks in the two series are designated by ri and r2 in Figure 1. The first series, rl, is that found b y Dibeler in hexamethyl disiloxane, (CH3),Si20, and includes masses 59, 45, and 31 ( 3 ) . These same mass ions were also found by Zemany and Price in the mass spectrum of tetramethylsilane, (CH3)4Si (18). These authors concluded that the rearrangement fragments contain Si-H bonds. The second series, r2,also has peaks differing by 14 mass units and includes masses 75, 61, and 47. This series is also explained by rearrangement structures similar to those proposed b y Dibeler, with the addition of an oxygen atom ( 3 ) . Derivatives of the branched primary, secondary, and tertiary alcohols investigated show the same two series of rearrangement peaks as the normal compounds. M a s s 75 is the base peak in the spectra of all the derivatives of pbranched primary alcohols given in Table I.
;rl
Rearrangement Series T? (with -0-)
previously. These can be summarized as follows.
silyl Ethers. Because of their high volatility, normal aliphatic trimethylsilyl ether derivatives of C, t o Clo alcohols can be analyzed by the mass spectrometer without difficulty. Relative intensities of t h e parent-minus15 peak, useful in such analyses, are given in Figure 2. These peaks are not fragmentation peaks in t h e spectra of hydrocarbons and osygenated compounds normally analyzed by the mass spectrometer. K i t h derivatives of normal C4 to Cg alcohols, the total contribution to the characteristic parent-miaus-15 peak of a compound by its homologs is less than 59;-, of the peak intensity. This slight interference can be corrected on the basis of the spectra of the pure compounds. Considerably less than 1% of any of the normal aliphatic tr‘methylsilyl ethers should be detectable in the presence of hydrocarbons. Mass 75 is of uniform intensity for the C3 to C, normal alcohol derivatives and therefore serves as a convenient check on the total alcohol content. Individual alcohol derivatives can be identified as to “type” by means of the mass spectral correlations described
‘
”
0
Only secondary and tertiary alcohol derivatives have strong peaks (other than the usual parent-minus-15 peak) above mass 103. The intensity ratio of mass 75.to 73 is different for primary, primary branched, and secondary alcohol derivatives according to class. The ratios are given in Table IV. Mass 89 is useful in determining the position of branching. Of the primary alcohols, only those having ybranching show intense peaks a t mass 89. Secondary and tertiary alcohols can be identified as to type from the original alcohol spectrum (6). The trimethylsilyl derivative of propylene glycol has the expected very weak parent peak and an intense parentminus-15 peak. I n contrast, both the parent and parent-minus-15 peaks are intense for the trimethylsilyl derivatives of phenol, butyl mercaptan, and the cresols. m- and p-cresol shon- similar fragmentation patterns, while o-cresol has a much less intense parent-minus-15 peak and a more intense mass 91 than m- and p-cresol. Trimethylsilyl derivatives should aid in the analyses of mix-
0
~
Apparent Structure
10
L 45
H J
Figure 2. intensities
parent-minus- 1 5 peak
61
-0-Si-H Table
r 31
Trimethylsilyl ethers,
47
IV.
Type Identification of Alcohols by Trimethylsilyl Ether Derivatives
H l +
-0-Si-H
Isotope determinations indicate that this is the correct assignment. Rearrangement peaks in the mass spectrum of hexamethyldisilazane, (CHa)&NH, parallel those in the spectrum of hexamethyldisiloxane. Analyses of Alcohols by Trimethyl-
Alcohol Type (Trimethylsilyl Derivatives) Primary-straight chain 1-Butyl to 1-decyl Primary-singly branched Butyl t o octyl including methyl and ethyl branching Secondary-(2-type) 2-Butyl to 2-hexyl a
Peak Intensities Mass 75 Mass 73a Ratio Mass 73 (Vol. Sensitivity)# Average Spread Average Spread. 2.0
0.1
513
38
1.3
0.2
665
210
0.82
0.04
1013
40
n-Heptane mass 27 = 492 divisions/liquid volume.
VOL. 2 9 , NO. 5 , MAY 1957
775
tures containing these classes of compounds as well as in the analysis of alcohols. ACKNOWLEDGMENT
The authors wish to express appreciation to Janet L. Shultz for the tabulation of the mass spectra. LITERATURE CITED
Brown, R. A., Young, W.S., Nicholaides, Nicholas, AXAL.CHEM.26, 1653 (1954). Charlet, E. M., “Mercury-OrificeType-Inlet-System,” Consolidated Engineering Carp. Mass Spectrometer Group meeting, Kew Orleans, La., 1950. Dibeler, 1’. H., Mohler, F. L.,
Reese, R. M., J . C h e w Phys. 21, 180 (1953). (4) Eden, Murray, Burr, B. E., Pratt, A. AXAL. CHEM. 23, 1735 (1951). (5) Friede1,’R. A , , Sharkey, A. G., Jr., J . Chem. Phys. 17,584 (1949). (6) Friedel, R. A., Shultz, J. L.. Sharkev. A . G,, Jr.. ~ A L .CLEM. 281 926 (1956). ( i )Gifford, A . P., Rock, S. AI., Comaford, I). J., Ibid , 21, 9 (1949). (8) Happ, G. P., Sten-ard, D. W., J . Am. Chem. SOC.74, 4402 (1952). (9) Kellev, H. lI., A k s A L . CHElf. 23, lo81 (1951). ‘ (10) Langer, A., J . Phys. R. Colloid Chem. 54. 618 (19.50’1. (11) Lanier,-S.‘ H.: konnell, S., Wendw, I., J. Org. Chem., to be published. (12) Langer, S. H., Friedel, R. A . , Render, I., Sharkey, 9.G., “Use of Trimethvlsilyl Derivatives in the
w.,
M a s s S ectrometric Ana!?&. of Fischer-Gopsch Alcohols, Division of Petroleum Chemistry, 128th Meeting, ACS, Minneapolis, Ifinn., September 1955. (13) Sauer, R. O., J . Am. Chem. SOC.66, 1707 (1944). (14) Sharkey, A . G., Jr., Shultz, J. L., Friedel, R. A., ANAL.CHEM.2 8 , 934 (1956). (15) Stevenson, D. P., Hipple, J. A . , J . Am. Chem. SOC.64, 1588 (1942). (16) Thomas, B. W.,Segfried, IT. D., ANAL.CHEAf. 21, 1022 (1949). (17) Yarborough, V. A , : Ibid., 2 5 , 1914 (1953).
(18) Zemany, P. D., Price, F. P., J . Ana. Chem. SOC.70,4222 (1948). RECEIVED for review September 26, 1956. Accepted December 26, 1956. ASTM E-14 Cornmittfee Meeting on hIass Spectrometry, San Francisco, Calif., June 1955.
Infrared Spectra of Aliphatic Peroxya c id s ‘EDGAR R. STEPHENS, PHILIP L. HANST, and ROBERT C. DOERR The Franklin lnstifufe, Laborafories for Research and Development,
,The infrared spectra of peroxypropionic acid and peroxybutyric acid were recorded in the vapor phase from 2 to 15 microns. In addition to the C-H and carbonyl bands, the most prominent absorptions were found at 3.05 microns, 6.9 microns, and 8.5 microns.
I
x THE COURSE of an investigation of the photo-oxidation of hydrocarbons a t low concentration in air, samples of peroxypropionic and peroxybutyric acids nere prepared and their infrared spectra recorded in the vapor phase. GiguPre and Olmos have published the spectra of peroxyformic and peroxyacetic acids ( 2 ) . Their qpectrum of peroxyacetic acid vapor is similar to the spectra reported here. Minkoff has published spectra which he attributes to peroxyacetic, peroxypropionic, and peroxybutyric acids (3). However, as he himself indicates, his samples were impure. His spectrum of peroxyacetic acid is fimilar to the spectrum obtained b y GiguBre and Olmos, but his spectra of peroxypropionic and peroxybutyric acids are not similar either to the peroxyacetic acid spectrum or to the spectra reported in this work. The chief impurity in Minkoff’s samples appears to have been the ordinary aliphatic acids. These have been found to be persistent contaminents of the peroxyacids. THE SPECTRA
The spectra of perosypropionic and
776
0
ANALYTICAL CHEMISTRY
0th and Parkway, Philadelphia
peroxybutyric acid are sho1vn in the upper portion of Figures 1 and 2. These acids are unstable; for that reason the spectrum of a fresh sample shon-s bands of decomposition products. After the spectra had been run, the samples were allowed to decompose in the cell and the spectra were rerun. Bands of the unstable peroxyacids can be identified if a comparison is made of the spectra obtained before and after decomposition. For comparison the spectra of pure propionic and butyric acid vapors were also recorded and are shown in the lower portion of Figures 1 and 2 . The spcctra w r e recorded on a Perkin-Elmer singlp-ham, single-pass spectrometer with the use of a PerkinElmer I-meter gas absorption cell. For each of the perovyacids the vapor pressure of the absorbing gas was about 3 or 4 mm. of mercury. The acid vapors were mived in the ahsorption cell with 1 atmosphere of dry, carbon diouide-free air which served to retard the decomposition of the pcrovT-acids. The peroxyacids shorn absorption bands usually ascribed to the stretching vibrations of C = 0 (5.7 microns), 0-0 (11.5 microns), and 02-H (3.05 microns), as well as other frequencies which apparently originate in the oxygenated part of the molecule. These are given in Table I. The chief distinguishing bands of peroxyacids appear, from these spectra, to be a t wave lengths of 3.05, 6.9, and 8.5 microns. The 3.05-band in particular is a useful indication of the presence of per-
Pa.
oxyacid in complex reaction mixtures. It is probably due t o the OH vibration and may, therefore, indicate that these spectra are those of the monomer and not the dimer. Giguere and Olmos ( 2 ) discuss in some detail the assignment of bands in the spectra of peroxyformic and peroxyacetic acids. PREPARATION AND CHEMICAL PROPERTIES
The perosyacids were prepared by the method of Fischer, Dull, and Volz (1) which is essentially as fol1on.s. Aldehyde, in ice-cold carbon tetra-
Table I.
Infrared Absorption Bands of Peroxyacids
Frequency, Kave Length, Cm.-1 Microns 1ntensit.v” 3280 2940 1760 1450 1180 1000 875 805
Pcrosypropionic .kid 3.05 3.40 5.68 6.90 8.48 10.00 11.43 12.42
hf
S S s S
IT 11 11-
Perosybutyric Acid ?\I 3280 3.05 S 2940 3.40 s 1760 5.68 S 1450 6.90 S 1175 8.52 862 11.6 IV a S = strong; l f = medium; W = weak.
