940
ANA
peaks. One or more rearrangement peaks (masses 29, 43, and 5 7 ) were found in the spectra of the four cyclic ketones. IDENTIFICATION OF UNKIVOWN KETONE
Analytical Significance of Rearrangement Peaks. Both normal fragmentation and rearrangement peaks can be useful in identifying an unknown ketone. It is apparent that ions corresponding in ma88 t o parent peaks of low molecular neight ketones, but resulting from high molecular weight ketones, can lead to erroneous analyses. For methyl, propyl, butyl) and higher types without branching on a carbon atom adjacent t o the carbonyl group, mass 58 rearrangement peak intensities are more than 20% of the 2-propanone parent sensitivity. The values shown in Table IV for the above ketones range from about twice the 2-propanone sensitivity down to one fifth, a i t h the majority near 80%. A similar chance exists for confusing higher mass ethyl-type ketones v i t h 2-butanone. For 3-heptanone and higher molecular weight ethyl types, the sensitivity is 70% or more of the 2-butanone sensitivity. It should be possible to determine the molecular weight of any ketone through 2-undecanone from the relatively intense parent peaks of these compounds. The empirical rules, as stated in the discussion of normal fragmentation, should aid in determining the mass of each of the two hydrocarbon groups attached t o the carbonyl group. Rearrangement peaks will in many instances verify these side-chain mass assignments. Even more important, rearrangement peaks should aid in determining the position of branching. Table V summarizes useful information obtained from rearrangement peaks. Four evamples of the identification of ketones are given in Table VI, utilizing normal fragmentation and the information contained in Table V.
L Y T I C A L C H E M 16 T R Y
Identification of straight-chain ketones should be relatively easy. I n many cases branching can be determined, especially if the branching is on the smaller hydrocarbon group. Although positive identification cannot always be made, the choice is limited to two or three possibilities. ACKNOWLEDGMENT
The authors Rish to acknonledge the asmtance of A. F. Logar, B. M. Thames, and Joseph Malli in obtain~ngthe mass spectra. LITERATURE CITED
Bloom, E. G., Nohler, F. L., Tengel. J. H., Kise, C. E., J . Research A'atl. B u r . Standards 41, 129 (1948).
Brown, R. A,, ANAL.CHEY.23, 430 (1951). Charlet, E. hl., Consolidated Engineering Corp., Pasadena, Calif., hlass Spectrometer Group Kept. 74 (1950). Eden, M., Burr, B. E.. Pratt. -1.IT.. -1.v.i~. CHEM.23, 1735 (1951)
F., Shulta. J. L . . I p p l . Spectroscopy 6 , No. 5, 24 (1952). Friedel, R. A,, Sharkey, A. G., Jr.. .LYAL. CHEM. 28, 940 Friedel, R. A., Logar. (1956).
Friedel, R. h.,Sharkey, -1.G., Jr.,J . Chem. P h y s . 17,
584
(1949).
Friedel, R. A , , Shulta. J. L., Sharkey. -1.G . , J r . , ABAL.CHEM. 28, 926 (1956).
Happ, G. P., Stewart, D. X-., J . A n i . C h e w . Soc. 74, 4404 (1952). Kinney, I. W., Jr., Cook, G . L.. ah-.^^. CHEM.24, 1391 (1952). Langer, dlois, J . P h y s . & Colloid Chern. 54, 618 (1950). O'Keal, SI. J., Jr., \Vier, T. P., J r . , ASAL. CHEM.23, 830 (1951). Washburn, H. W., in "Physical lIethod.5 in Chemical Analysis," p. 592. W. G. Berl. ed.. -1cademic Press, New York, 1950. RECEIVED for review Sovember 5 , 195.5. -4ccepred hIarch 20, 1956. Presented a t A S T M E-14 Committee l l e e t i n g on l l a j s Spectrometry, New Orleans, L a . , May 1954. A contribution from t h e Central Experiment Station, Bureau of Mines, Rruceton, P a .
Mass Spectra of Acetal-Type Compounds R. A. FRIEDEL and A. G. SHARKEY, JR. Bureau o f M i n e s ,
U. S. D e p a r t m e n t o f t h e Interior, Bruceton, Pa.
.\lass spectra of various acetals have been determined and correlated with molecular structure, and the analytical applications are discussed. Evidence for the temporary existence of hemiacetals in the vapor state has been found in the form of anomalous mass peaks which appear in spectra of mixtures of oxygenated conipounds. The decomposition of hemiacetals in the vapor phase is a first-order reaction; specific reaction rate constants for the decompositions have been measured by following decreases in spectral peak heights with time. Analytical interference by unstable hemiacetals can be circumvented b: proper procedure.
MASS
ectra of mixtures of Oxygenated compounds frequently -p produce mass peaks that are not attributable to common alcohols, acids, aldehydes, or the like. Some of these peaks are at odd masses and are, therefore, fragmentation peaks. They are consistently two mass units higher than fragmentation peaks from other oxygenated compounds. T o produce such peaks, the oxygenated hydrocarbons must contain two oxygen atoms and no unsaturated bonds. Acetals and hemiacetals are logical possibilities. Mixtures of aldehydes and alcohols produce hemiacetals, most of which are rather undable at room temperature ( 1 , 6).
A series of acetals of formaldehyde, xetaldehyde, and propionaldehyde was investigated to correlate mass spectra and to use the data for analyzing mixtures of oxygenated compounds. The general formula of the acetals is
H OR: R1-C
'/ \
OR? where R1 is hydrogen for diosymethane.; (formais) from the reaction of formaldehyde yith alcohols, or CH, for dioxyethanes (acetals) from the reaction of acetaldehydp with alcohols, or C2H6for dioxypropanes (propionale) froni the reaction of propionaldehyde with alcohols. EXPERIMEYTAL
Spectra were run on a Consolidated 21-103 mass spectrometer equipped s i t h a Consolidated micronirinometer. Effusion rate measurements were made with valve open to the mass spectrometer leak. The effusion rate TTith n-butane was the reference for calculation of apparent molecular weights For the calculation of specific reaction rate constants, the curves obtained for mass spectral peak heights decreasing n*ith time were corrected for normal decay of peak heights.
94 1
V O L U M E 28, NO. 6, J U N E 1 9 5 6 -
Table I.
hlass Spectral Patterns of Acetals RiCH(0Rdz
m/e
18 19
DiDiDi-nDiDiDi-nrnethoxy- ethoxy- propoxy- methoxy- ethoxy- propoxymethane methane methane ethane ethane ethane 1 .OS 0.52 2.38 0.90 1.41 7.62 0.08 2.00 0.65 2.41 0.10 2.78
26 27 28 29 30 31 32 33
0.20 0.47 2.82 45.6 3.35 12.6 0.83 0.23
39 40 41 42 43 44 45 46 47
0.27 0.06 0.19 0.41 0.96 2.32 100CPd 2.26 3.52
55 56 57 58 59 60 61 69 70 71 72 73 74 75 76 83 84 85 86 87 88 89 90
... ...
3.83 9.58 8.49 55.9 3.10 39.7 3.09 1.64
5.98 23.8 6.60 38.9 2.83 19.2 0.41 0.20
4.17 27.5 7.07 25.0 1.36 7.70 0 10 0 99
6.51 45.8 18.5 68.5 2.63 29.5 0.72 1.82
8.58 46.5 15.3 75.3 2.30 41.8 1 33 0.26
6.42 27.2 10.85 51.5 3.85 28.4 2.21 1.53
13 3 60.2 25.7 97.0 8.34 67.7 1.23 0.24
0.18 10.61 0.0J 1.86 2.87 22.8 1.40 6.79 6.13 g o d 3.47 11.5 13.7 15.2 0.60 0.34 14.4 0.93
0.26 0.10 2.17 3.14 25.3 1.12 1.96 0.04 10.60
0.29 0.17 1.02 2.94 2.20 7.89
7.68 1.31 18.2 5.92 52.4 10.7
17.3 3.06 57.9 12 6 45.7 8 99 83.5 1.99 1.08
23.1 4.76 53.7 43.8 20.9 8.83 0.38 0.19
8.61 1.26 33.0 5.15 11.7 1.11 44.3 0.99 20.8
6.82 2.19 17.4 5.87 12.3 2.17 19.7 5.08 63.6
46.3 75.0 11.2 1.21 1.21 0.11 1.13
2.43 1.07 7.22 1.61 2.49 0.27 0.26
4.43 1,12 18.4 13.5
19.2 6.57 5.32 0.79 2.57 0.23
0.15 0.06 3.25 3.06 73.9C 3.57 100d,a 3.33
...
...
0.42 7.7: 91.7 3.20 0 97
0.09
...
...
... ...
0.03 0.06
...
43.5c P 1.471
.,. 0.27 0.62 5,30 0.18
4.10 2.03 15.4 1.93 6.94 1.086 0.70 0.18 0.12 0.74 2.26 60.5C 2.79 0.52
...
... , . .
0.51 2.61 lOOCtd 3.37 0.28
-
... ... ... ... ... 0.10 34.65 1.15
4.04 17.0
lOOd 2.75
0.076 0.05 0.19 0.50 2.98 0.31 3.18 0.08 0.16 0.13 1.09 48.4C 2.17 5.00 0.22
...
.. 3.32 0.15
6.54 0.37 2.31 0,093
...
1.40
,
..
101 102 103 104 117 118
... .. ... . . . . ...
3Y.Qe P1.63
..,
3.99 4.46 0.28 4.44 0.24
129 131 132
. . . .
...
...
.,.
E O d
1 71
1.99 1.21 2 75 3 21 11 0 9
1 56 0.43 0.12 0.78 0 27 1.60 0.35
...
... .
.
...
5.0lC PO.17
...
, ,
..
1.61 PO.07
0.28 0.09 1.02
...
...
0.14 0.39 30 7 ' 1.74 3.49 0.20 1.84
... ... ...
... ...
, . .
14.28 0.68 0.59 P...
...
9.28 33 5 LOOd
7 63 1.40 0 54 3 74
...
..,
1.60 1.69 2.50 0.47 0.18
lOOd -
.
, ,
.., ...
. .
. .
Di-nDiDi2-1Ie. hexoxy- rnethoxy- ethoxy1,31,3IA5ethane propane propane Dioxolanea dioxolane Trioxane 2.15 0.51 2.00 ... 0.73 4.91 4.56 1.47 0.11 ... 1 64
4.08 29.8 9.30 27.0 4.49 22.7 0.70 0.19
7.58 31.0 11.2 58.7 4.87 100d 1.68 0.29
0.05 0.09 0.08 0.02 0.014
...
Di-nbutoxyethane 0.68 1.71
..
, .
0.45
...
...
5.67e 0.20
2.32 0.14 1.98 0.19 0.31
...
0.34
...
75.5c 4.62 4.30 0.22 0 27
1o:is 16.7 1.10 0.30
...
...
..
... ,, ,
0.14 0.03 1.01 0.51 0.08
...
., ..
1.33 PO.06
...
...
, . .
...
...
9.77c
0.38
-
... ...
...
146 ... P . 159 ... ii.9. 174 . . P... ... , . . 215 ... . . .. 0.94' ... 280 .. . . ... P... ... Sensitivity of base peakb 69.5 68.8 67.4 52,2 69.0 03.3 49.0 33.6 47.3 .i.I'.I. Project 44 a t Carnegie Institute of Technology. Serial N o . 87, Contr.: Union Oil Co. I Sensitivity of n-butane mass 43 = 50.4 divisions per micron. C Parent minus ORt. d Rase peak. Parent minus R I . i P = parent mass. Y Spectrum corrected for 1-propanol impurity. detected by infrared.
Compounds used and their sources ai e listed below. Dimethoxpmethane (dimethylformal), Celanese Corp Dirthoxymethane (diethylformal), Eastman Kodak Co. Di-n-propouymethane (di-n-prop! lformal), fractionated, Celanrse Corn. 1,l-Dikthoxyethane (dinieth? lacctal), Celanew Coil). 1,l-Diethoxvethane (diethvlacetal or acetal), Eastman Kodak co. I ,1-Di-n-propoxyethane (tli-n-propylacetal), synthesized at Bureau of Mines. 1,l-Di-n-butoxyethane (di n-butylacetal), Celanese Corp. 1,l-Di-n-hexoxyethane (di-n-hexylacetal), fractionated, Carbide and Carbon Corp. I ,1-Dimethoxypropane (dimethylpropional), fractionated, Celanese Corp. 1,l-Diethoxypropane (diethylpropional), synthesized at Bureau of Mines.
lOOd -
3.87 1.30
1.32 6.78 1.00 0.11 0.68 0.49 32.6 1.05
...
13.7
12.5 4.8 57.0 3.6 7.6 0.2
28 8 35.G 50.6
5,01
12.9 0.37 0 11
...
...
0 40 0.27 1.91 8.35 71.6 3.29 64.6 1.46 0.27
0.1 0.6 4.5 70.0 57.9 26.4 0.7 0.1
...
...
0.29
0.12 0.06 0.31 0.03
lOOd -
P6.0 0.5
8.4i 0.44
, . .
...
...
0.12 0.13 9.72 P0.49 0.05
...
...
...
...
... ..,
. . ...
...
, . .
...
... ...
, . .
34.8
29.3
...
... ... ...
6.20 0.20
...
...
2.32 P0.06
...
... ...
...
... ... ...
,..
...
37.9
0.38 11.7 0.49
...
...
...
...
8.89
17 0 PO 62
...
...
...
0.03 0.07
, . .
. .
...
...
...
,..
,..
.
...
l0Od -
... ... ... ...
... . .
0 23 0 36 24 2
...
0.47 0.41 43.0e 2.03 0.36
. .
0 la
...
...
.
0.05 0.05 0.02 0.01 0.14 0.10 0.033
0.25
...
,
2.27 0.27
0.3 0.7
...
...
,
lOOd -
, . .
0.24 5.8$ 60.4 3.38 0.43
, . .
...
0.12 0.07 0.60 3.62 54.5 5.11
...
... ...
0.43
0.38 P...
2.52 7.40 3.04 17.8 0 32 0.91 0.12
0.28 1.41 30.3 2.22 0.23 0.17
...
...
0 34 0 32 0 38 1 29 0 26
0.19
0.1 0.1 0.2
...
...
0 13 0 32 2 70 33 4 9 58 l s d 6 33 0 60
2,4,6-Trime-1,3,5trioxane 0.39 1.85
73 4
110.2
1,3,5-Trioxane, sublimed, Eastman Kodak Co. 2,4,6-Trimethyl-1,3,5-trioxane (paraldehyde), Eastman Kodak Co. 2-Methyl-1,3-dioxolane, Carbide and Carbon Corp. Formaldehyde solution, 37%, Fisher Scientific Co. ACETALS
Mass Spectra. Spectral patterns of various avetals are given in Table I. The preferred fragmrntations involve rupture of (1) a C-0 bond with loss of -OR, and 12) the C-Rl bond with loss of Ri; the residual mass frsgmellts are easy to identify because of the mass difference between carhori and oxygen. Loss of -OR* produces peaks a t masses 101, 87, 73. etc., whereas IOSS of Ri produces peaks a t 103, 89, 7; et?. These are the two strongest mass peaks in the upper half of each spectrum; they
942
ANALYTICAL CHEMISTRY
are indicated in the spectra in Table I. A single acetal can be identified by application of a simple equation involving the mass fragments from loss of -R1 or -OR!:
2
x
(mass of fragment from loss of -OR,) (mass of fragment from loss of -R1)
=
constant
The numerical value of the constant depends on the aldehyde from which the compound is derived: 15, 43, 71, etc., reepectively, for formddehyde, acetaldehyde, propionaldehyde, etc. After the aldehyde is determined, the alcohol portion of the molecule can be identified.
25
I
I
1
I
I
1
(Figu~e1) has no mass peaks above its molecular neight, 30 But a commercial formaldehyde solution vaporized into the mass spectrometer showed peaks at 31, 32, 33, 45$ 61, and 75 Methanol is added to formaldehyde solutions as an inhibitor. and is responsible for a t least some of the 31, 32, and 33 peaks. The 7 5 peak may be explained by the formation of dimethoxymethane (dimethylformal). But the 61 peak and parts of the 45 and 33 peaks could not be assigned. The only apparent explanation for the 61 peak is loss of a hydrogen atom from the hemiformal of formaldehyde and methanol, H&(OH) (OCH3’. The 45 peak also is attributed to hemiformal and results from splitting off of a hydroxyl group. Methylene glycol, CH2(0H)*, a know1 constituent of aqueous formaldehyde solutions ( 6 ) , is probably too unstable to exist in the vapor phase a t a pressure of about 0.1 mm. of mercury. Hall and Piret found indications in vapor density studies on alcoholic formaldehyde solutions that methylene glycol dissociated more completely than hemiformals on vaporization (4).
Table 11. Y
I5t
1 m/e 29 30
a
IO
I
I
I
1
1
I
Evidence of Hemiformal in Vapor Phase of Formaldehyde-Methanol-Wa ter
Peak Height 2160 1107
31
1179
32
026
33
75.7
45
7i
Methanol 552 73.3 942 823 9
Formaldehyde 1388 1033.7 15.2 2.1
9.9
Residuals 220
Proposed Structure of Fragment Ion
0
221.8
+H-
0 65 8
7,OCHi 77
-H-C
NO,
Table I also gives the spectra of two dioxolanes, cyclic acetals derived from ethylene glycol. I n the spectrum of 1,3-dioxolane, the fragmentation may be considered as occurring a t two bonds simultaneously to produce a fairly intense peak a t an even mass,
61
185
185
H-I-C !
\OH
44. 0-CH? I
/’
Table 111.
m/e 28 29
In Z-methyl-l,3-dioxolane, the same type of fragmentation produces the even-numbered mass 58. Trioxane, the trimer of formaldehyde, and paraldehyde, the trimer of acetaldehyde, may also be considered cyclic acetals. Mixtures Containing Acetals. An acetal is easily detected in the presence of alcohol and aldehyde. Rlass peaks a t 75, 89, and 103, frequently found in mixtures of aldehydes and alcohols, are usually attributable to acetals. But mass 61, common in such mixtures, has often been difficult to identify. S o fragment of mass 61 can logically occur in spectra of acetals; some acetal spectra have small 61 peaks due to rearrangement, but these are identifiable. A fragment of mass 61 must contain two oxygens, hut the smallest such fragment from acetals, -CH(OCH3)2 from dimethoxymethane (dimethylformal), has a mass of 7 5 . A possible explanation for the presence of unaccountable mass 61 peaks is the fragmentation of hemiacetals. HEMIACETALS
From Formaldehyde and Alcohols (Hemiformals) in the Vapor Phase. DECOMPOSITIOS RATECOSSTANT. The spectrum obtained from a commercial formaldehyde solution indicated the possible presence of hemiacetals. The spectrum of formaldehyde
30 3 1 .~
32 33 43
44 4.5
Peak Height 96.0 1152.0 87.5 720.0 470.0 25.3
207.0 312.0 33 .i
59
5.2
61
16.6
75
Evidence of Hemiacetal in Vapor Phase of Acetaldehyde-Methanol
1.9
Methanol 55.2
492 58.4 712 470 7.1
Acetaldehyde 27.7 67 1 7 i 1.9 180 312 9.7
Residuals 13.1
Proposed Structure of Fragment I o n
-10.0
1.4
6.1 0
18.2 27.0 0
24.0 5.2
OCHJ +CH3CH7 \OH Dimethoxyethane
16.6 C H 3 ’ y C H/ O C H 3 1.9
+
\OH Dimethoxyethane
A formaldehyde-methanol blend was prepared to check the spectrum obtained with commercial formaldehyde solution. The spectral peaks were the same, except that no indication of dimethoxymethane (mass 75) was obtained. Extensive hemiformal formation is indicated in Table I1 by the residual peaks remaining after subtraction of the spectra of methanol and formaldehyde. These residuals were larger than those from commercial formaldehyde solution, because the synthetic blend contained a much higher concentration of methanol.
943
V O L U M E 28, NO. 6, J U N E 1 9 5 6 To find the origin of the mass peak a t 61, determination of the molecular weight of the parent structure was attempted by effusion rate measurements based on Graham’s law of diffusion ( 2 . 3 ) . If the parent structure of that fragment were the hemiformal of methanol, CH2(0H) (OCH,), effusion rate measurements should indicate that the 61 peak is caused by loss of one hydrogen from mass 62. The measurements might have indic*:tted that mass 61 arose from a structure of higher molecular \\eight, but obvioiislj it could not have oiiginated from a struc-
(bl
600
heights are explained by decomposition of unstable hemiformals to formaldehyde and methanol. From 30 t o 60 minutes, the decreasing peaks indicate that decomposition of hemiformal had been completed and that the normal effusion rate of the mass spectrometer had become dominant. A blend of formaldehyde and ethyl alcohol in water was also investigated briefly. -4peak corresponding to the 61 peak from the hemiformal of methanol was found, as expected, a t mass 75 for the hemiformal of ethyl alcohol. From Acetaldehyde and Alcohols in Vapor Phase. DECOJIPOSITIOX RATE CossTasr. A blend of acetaldehyde and niethanol \vas investigated for hemiacetal formation. Hemiacetal apparently formed, but only in small concentration in the vapor phase. The spectrum, given in Table 111, includes a mass 61 peak. .4n attempt to ineasuie the niolecular weight of the parent structure responsible for the mass 61 peak produced an impossible molecular weight of 0.068. This value indicates that the hemiacetal is less stable than the hemiformal. Decon~position of henliacetal, as determined by decay of t h e mass 61 peak (Figuie 3), \\as also a first-order reaction. The specific reaction rate constant, 0.052 minute-’, nnq about 7 . 5 times that of the hemiformal.
MUSS
33 9
1 2 5 t
. .L .-
I
I
I
0
30
60
I
I
\
1
0
30
Mass61
60
TIME, M’NUTES
Figure 2. \-ariation of mass pealis with time in blend of formaldehyde-methanol-water I
ture with a mass less than 61. The apparent molecular \$eight (Figure 1) was 0.64 because of the instability of the structure producing the mass 61 peak. The peak decreased rapidly during decomposition of the parent structure, hemiformal, in the vapor phase. The logarithmic decay of the peak height for mass 61 (Figure I ) illustrates that vapor-phase decomposition of hemiformal is a hrst-order reaction, The specific reaction rate constant, cald a t e d from the rateof decrease of this peak, was0.0069 minute-1. This procedure is a simple and accurate method for studying ieactions in the vapor phase. Rate data may be obtained for either decreasing or increasing components in a reaction by following changes in peak heights of individual components. Decomposition of the hemiformal of methanol was further obqerved by following temporal changes of other spectral peaks of a vaporized sample of aqueous formaldehyde-methanol. Figure 2 shows the erratic behavior of various mass peaks. Curves in Figure 2a, represent major peaks of the principal components, formaldehyde and methanol; curves in Figure 2b, represent peaks other than those attributable to the two major components. The curves of Figure 2b, may be explained by decomposition of the unstable hemiformal and the consequent decrease of ions of masses 61, 45, and 33. The structures proposed for some of these mass fragments are given in Table 11. The curves in Figure 2a, also resulted from decomposition of hemiformal. During the first 30 minutes, the increasing peak
1
I
,
TIME, MINUTES
Figure 3. Rate of decomposition of mass 61 fragment from hemiacetal of methanol
The behavior of other spectral peaks was also observed (Figui o 4). As with heniifornial (Figure 2 ) , the peaks attributable to stable structures increased during the first 30 minutes, while the peaks for the iinstable structures, masses 61, 45, and 33, decreased rapidly. The uniform rates of decrease of unstable peaks are good evidence that the peaks arose from the same structure. Structures proposed for some of these unstable peaks are given in Table 111. The temporal change of pressure in microns, shown a t the bottom of Figure 4, indicates the increased number of molecules resulting from decomposition of hemiacetal. To obtain results comparable with those from formaldehydemethanol-water, a mixture of acetaldehyde-methanol-water was investigated. Formation of hemiacetal was not as extensive as of hemiformal. The same unstable peaks were found, but the relative intensities of hemiacetal peaks were less in the presence of water. An acetaldehyde-ethyl alcohol mixture was investigated. During runs on the same sample, made 30 minutes apart, rapid
944
ANALYTICAL CHEMISTRY
decay of masses 7 5 , 61, and 47 indicated the presence of the unstable hemiacetal of ethyl alcohol. The 75 peak probably arose from loss of a methyl group. MIXTURES OF ALCOHOL AND ALDEHYDE
Apparent Formation of Acetals. The mixing a t room temperature of an aldehyde and an alcohol mass was found to produce mass 1)eaks attributable to a trace of acetal. The persistence of these peaks indicated the stability in vapor phase of the parent substance. Table IV gives several examples of apparent acetal formation The acetal m a s peaks were weak but had the correct relative intensities Mixtures of formaldehyde and methanol did not indicate formation of dimethoxymethane. All other blends, including formaldehydeethyl alcohol, did indicate formation of the expected acetal. The mass spectrometer is ideally suited for the detection of traces of acetals because they usually have two or more distinctive mass peaks above the mass range of the alcohol and aldehyde.
Table I\’.
Formation of Acetals in Alcohol-Aldehyde Mixtures Mole %
a
Ethyl alcohol Propionaldehyde Diethoxypropaoe 1-Propanol Acetaldehyde Di-n-propoxyethane Acetaldehyde Methanol Dimethoxyethane Acetaldehyde Methanol Water Dimethoxyethane Ethyl alcohol Acetaldehyde Diethoxyethane Apparent concentration.
Synthetic 67.6 32.4 0 64.8 35.2 0 41.7 58.3 0 5.6 7.8 86.6 0 49.9 50.1 0
Analysis 68.2 31.5 (0.3). 66.4 33.6 (0.3)a 42.4 57 6 (0.1). 5.1 7.3 87.6 (0.03)a 52.5 47.5 (0.I)O
TIME, MINUTES
Figure 4.
Variation of mass peaks with time in blend of acetaldehyde-methanol
hemicompounds after expansion of the sample into the spectrometer. After a predetermined waiting period, or after there is no further rise in micromanometer pressure, analysis may be carried out ACKNOWLEDGMENT
Evidence of the presence of acetals in alcohol-aldehyde mixtures is significant because the formation of acetals in the absence of a dehydrating agent is not expected. However, no attempt was made to isolate the acetals from the alcohol-aldehyde mixtures; therefore, the existence of acetals is not certain. Formation of Hemiacetals. Hemiacetals definitely form upon mixing an aldehyde and an alcohol (1, 5 ) . -4lthough they are unstable in the vapor phase, hemiacetals may interfere with analysis of the alcohol and aldehyde. One can check for interference by searching the spectrum for peaks suspected of instability and observing their behavior with time. Slowly rising pressure for several minutes after expansion of the sample also is a good indication of the presence of unstable molecules that are decomposing to give two or more molecules each. Interference can be circumvented by awaiting complete decomposition of the
The authors wish to acknowledge the helpful cooperation of Irving Wender and Heinz W. Sternberg and the assistance of Janet L. Shultz, A. F. Logar, B. hl. Thames, and Joseph Malli. LITERATURE CITED
(1) Ashdown, A . , Kleta, T. A., J . Cham. SOC. 1948, 1454. ( 2 ) Eden, M., Burr, B. E., Pratt, A . W., ANAL. CHEM.23, 1736 (1951).
(3) Friedel, R. A , , Sharkey, A. G., J . Chem. Phys. 17, 584 (1949). (4) Hall, M. W., Piret, E. L., I n d . Eng. Chem. 41, 1277 (1949). ( 5 ) McKenna, F. E., Tartar, H. V., Lingafelter, E. C.. J . Am. Chem. SOC.71, 729 (1949). ( 0 ) Walker, J. P., “Formaldehyde,” Reinhold, Sew York, 1953. RECEIVED for review September 22, 1955. Accepted March 27. 1956. Presented a t Pittsburgh Conference on Analytical Chemistry and Applied Spectroscopy, Pittsburgh, Pa., March 1951. Contribution of Branch of Coal-to-Oil Research, Bureau of Mines, Region V, Bruceton, Pa.
