APPENDIX I
It is given that y = AX
(11)
where y is an m X 1 vector, 1: is an n x 1 vector, A is an m x n matrix, and m > 72. It is desired to find 2: such that //ellz=
(Y - A$, Y - A T )
lle(dIl2
=
(y
Y
+
Hence
33, 30 (1962). (3) Lee, F., Sears, F. W., Turcotte,
Hildreth, C., Nav. Res. Log. Q . 4 pp. 79-85 (1957). (2) Kendall, B. R. F., Rev. Sci. Instr.
Ji;
1963. ( 4 ) Luenberger, D. G., "Re,solution of
(14)
Mass Spectrometer Data, Stanford Electronics Laboratories Report SUSEL-64-129,(Nov. 1964). (5) i\lc,powell, C. A,, "Mass Spectrometry, pp. 29-34, McGraw-Hill, New
( A T A 2 - ATy, d ) = 0 (15) Since Equation 15 must hold for all 4 i t follows that
York, 1963. (6) Wiley, W. C., McLaren, I. H., Rev, Sci. Instr. 26, 1150 (1965).
2(Ai, A l ) - 2 (y, A!) and
(Ai
- y, AC)
=
- ATy
0
(16)
i = (ATA)-'ASy
(17)
=
and (13)
LITERATURE CITED
D. L., Statistical Thermodynamics," p. 53, Addison-Wesley, Reading, Mass.,
ATAi
- A(Z + Ed),
- A(% + 4 2 lIe(0)112
(1)
(12)
is a minimum. Let the vector x which minimizes lle/j2 be 2 . Suppose x is replaced by Z e l where E is a parameter and is an arbitrary n x 1 vector. Then, since Z is the optimal solution
!.
The squared error lie(€) is a continuous function of E and achieves its minimum a t E = 0. Therefore, its derivative with respect to E must be zero at t = 0.
RECEIVED for review December 6, 1965. Accepted March 14, 1966. The work reported in this paper was supported in part by XASA Grant NsG 80-60 and by Office of Naval Rpsearch Contract Nonr225(24), NR 373 360.
Mass Spectra of Isocyanates JOHN M. RUTH and ROGER J. PHILIPPE' Research Department, liggeft and Myers Tobacco Co., Durham,
b The mass spectra of isocyanic acid and 14 isocyanates were examined. Jmportant ions in the spectra of normal alkyl isocyanates are (CH2NCO)+ and, where the alkyl chain is long enough, (C5HaNO)+. Molecular ion and (M-1 ) ion intensities vary widely, with M+ reaching a minimum where the alkyl chain length is most favorable for the formation of the rearrangement ion (CsHgNO)+. Aromatic isocyanates give strong peaks for the molecular ion, for (M-28), and for (M-SS), as well as others which depend more on the structures of individual cornpounds.
N
o
STUDY of the mass spectra of isocyanates has been published. Investigations of the spectra of the cyanides, isocyanides, and isothiocyanates were reviewed recently by Budzikiewicz, Djerassi, and Williams ( 3 ) . EXPERIMENTAL
The mass spectra in Table I were recorded with a Model 14-101 Bendix time-of-flight instrument using 70-volt electrons. The ion source wall was at room temperature in every case except t h a t of octadecyl isocyanate, when i t was at 74" C. The chromatographed samples were admitted from the glass chromatograph traps through a stain1 Present address, Tabacofina, s. A., Service de Recherches, Merxem, Antwerp, Belgium.
720
ANALYTICAL CHEMISTRY
N. C.
less steel taper joint and Hoke 2PY280 needle valve. Carbon dioxide was usually observed at first, either contained in the sample trap or resulting from hydrolysis by adsorbed water on the walls, after which the isocyanate replaced it. I n all cases except those of the acid and the methyl isocyanate, a liquid phase was present in the sample trap, permitting such a separation of a more volatile impurity. The absence from the ion source gas phase of the amines, which would have been the other product of hydrolysis, is shown by the absence of the molecular ions (4). T'ne allyl and isopropyl isocyanates were introduced through the Bendix molecular leak inlet system, and the 1-naphthyl and octadecyl by means of the direct inlet probe. Isocyanic acid was prepared by the thermal decomposition of cyanuric acid under vacuum. The methyl, cyclohexyl, phenyl, o-tolyl, m-tolyl, and p-tolyl isocyanates were Eastman Organic Chemicals highest purity grade, with boiling ranees of 37'-39' C., 166'-168" 2.,55'-56" C./13 mm.; 77'-i8' C./12 mm.. 75'-76' C.112 mm., - a n d '68'-70' 'C./lO mm., 'respectively. The ethyl, butyl, and octyl isocyanates were Eastman practical grade, with boiling ranges of 57'61" C.. 110'-114' C.. and 77'-78' C./6 mm., respectively. The propyl isocyanate, b.p. 83'44' C., was Carwin Organic Chemicals, 97.5y0 minimum purity. All mere treated by gasliquid chromatography (6). Later, when the chromatograph with the trapping system was not available, the allyl, isopropyl, 1-naphthyl, and
octadecyl isocyanates were obtained from Chemetron Chemicals and used without further purification. The high resolution mass measurements were done a t the Mass Spectroscopy Laboratory of the Mellon Institute with an A.E.I. AIS-9 doublefocusing instrument. The cold inlet system was used at room temperature, while the ion source was a t approximately 230' C. RESULTS
Table I contains the time-of-flight mass spectra. Successive scans of the methyl isocyanate showed a small rate of reservoir pressure drop. Two successive scans were used to correct the intensities for this decrease. Some of the other spectra in Table I1 are averages of four to ten scans. The averaging was done in two or three cases because of intensity fluctuations detectable on the oscilloscope and in successive recorder charts, but such behavior was not observed in most of the measurements. The recording of a number of scans in succession for each sample also permits observation of the removal of such volatile contaminants as C 0 2from the traps, as described above. High resolution mass measurements of selected peaks in the spectrum of octyl isocyanate are recorded in Table 11. Only the molecular ion was measured in the spectrum of octadecyl isocyanate, giving 295.2873, in agreement with the composition ClgHnNO+.
m/e 12 13 14 15 16 17 18 20.5 21.5 22 24 25 26 26.5 27 27.5 28 29 30 31 31.5 32.5 33.5 36 37 38 39 40 41 42 43 43.5 44 44.5 45 45.5 46 48 49 50 51 51.5 52 52.5 53 54 55 55.5 56 56.5 57 57.5 58 59 59.5 60 61 62 63 64 65 66 66.5 67 68 69 69.5 70 70.5 71 72 73 74 75 76 77 78 79 80 81
5 0.4 3 14 2 1 3
5 3 6 11 2 0.9 3
1
7
6
5
43 6 100 13 1
12 27 3
26 100 10 0.5
3 2 1 0.3 4 0.5 7
4 5 14 42 2 0.2 0.3 0.1
2 2 8 25 2 0.3 0.8
2 11 67 1 0.1 0.1
0.6 3 18
0.4 3 28
0.3 2 14
42
87
37
67 11 0.7
79 70 30 0.4
24 5 0.9
0.1 0.1 6 8 14 8 18 9 3
0.3 3 6 17 11 19 39 4 8
1
1 1 4
4 0.4 0.1
0.8 6 35 0.3 53 0.6 100 7 0.4
3 0.9 6 26 5 0.3 2
2
0.8 6 11 33 6 9 4 0.6 0.3
0.9 17 100 70 28 25
0.2 3 7 33 12 82 24 82, 22
0.3
6
0.2 2 2
2 23 19
1 3 9
6 7 43
34
100
6
46
63
69
0.2
5.4 54 2.7 0.2
1 0.9
100 5 0.4
69 26 2 0.2
3 9 72 21 81 35 11 2
0.6 2
0.3
0.4 4 7
0.3
6
4
31
39
9
18 3 0.6
31 6 0.3 0.2
15 3 0.3 0.1
20 2 0.4 0.2
8 0.8 0.1
1 24 42 85 14 7 3 3
0.7 17 29 61 9 7 2 2
0.7 13 22 46
2
1
3
0.3 0.1
0.2
0.3 0.2
0.2 7 15 26 3 0.9 0.4 1 0.3 0.7 0.2 0.1
0.8 8 58 87 10 61 10 12 5 0.8
0.4 4 29 41 5 32 6 7 3 1
0.5 5 35 54 7 38 8 8 4 0.6
0.9
2
0.6
54
2 2 0.4
30
22
1
20
36
4
6 2
5 2
0.2
0.5
0.3 0.5 0.4
1
0.2 0.4 1 0.8 1 2
6 5 21
19 6 23
8
9
9
4
8
3 0.4 0.3
12 2 0.6
0.2
0.3
0.3 0.2 2
0.6 0.5 3
1
0.1
0.1 0.7
7 5 2 2
0.1 2 21 18 6 0.9 0.7 0.9 0.6 1
3 5 5
0.4
0.5
3 4 14
0.3 0.6
1
0.5 0.4
0.4 0.2
11 26 17
7
4
0.6 0.7 0.1 0.3
64
19
0.4 4 4
0.1 0.4 2
20
0.3
2 5
0.2
100
0.8
3 7
0.4
0.2 5
0.8
1
0.1
0.3 2 26
1
2
0.2 0.1
0.1 0.9 15
0.4
5
2
0.2 2 36
6
2
0.3
0.3
5
3
3 0.4
5 0.4 0.1 0.2
0.9 0.4 4 7 2 0.4 2
3 35 56 52 10 11 3 2 0.7 2 4 1 5 1 0.8 6 26 29
0.3 2 0.2 40 7 10 2 100 75 40 19 58 100
0.4
2
0.5 3 19
0.2
0.8 4
0.2
0.3
0.4 21
0.8
8 45 7 0.1
0.3 2
28
0.9 0.6 3 10 1 0.3 0.9
45 82 9 0.2
0.4 2
3
2 0.7 0.8 2 0.6 0.1 0.5
15
0.2 0.7
81
0.8 3
0.3 0.2 4 20 1 0.3 0.7
77
0.2 1 2
100
0.2
0.2
0.4
0.5 4 10 29 22 33 82 17
0.2 0.1 3 12 2 2 7
7 3 10 14 38 65 18 2 0.8
0.9 8 15 33 18 16 4 7 2 0.3 0.2
0.5 5 10 22 13 13 4 5 2 0.4 0.3
0.6 5 10 22 11 12 4 5 2 0.4 0.3
0.4 7 18 28 10 3 0.6
2 4
0.2
13*
0.7 0.2 0.1 0.2 0.9 0.4 1 2 3
2 5 3 3 2 0.4
2 9 13 18 32 81 17 2
0.1 1 5 7 11 20 39 9 1
1 4 8 11 23 35 7 1
2 12 12 7 3 0.8
(Continued)
VOL. 38, NO. 6, MAY 1966
e
721
Table 1.
(1)"
(2)"
(3)"
82 83 84 84.5 85 ..
(4)a
(5)"
(6)"
17
1
17
15
1
86 87
Time-of-Flight Mass Spectra (confinued)
Relative intensity, yo (8)0 (9)& (10)" (11)" 23 2 1 2 3 3 9 6 11 b 0.4 9 4 1 23 2 0.3
1
10 2 0.5
0.4 1 1
0.7
89 90 91 92 95 96 97 98 99 100. 101 102 103 104 105 106 109 110
6
74 10 0.3
0.2 10 5 1
~
1
17 67 7 0.8
2 1
111
15
112 113 114 115 116 119 120 122 123 124 125 126 127 128 132 133 134 138 139 140 141 142 154 155 168 169 182 183 196 197 _. . 210 211 224 225 238 239 252 253 266 267 280 281 295
11 1
~~
1 1
4 6 66 6
(14)"
0.8
0.9
0.7
1
1 1 1
1
2 2 9
21 2
3 4 20 2
0.8 0.4 0.1 3 5 91 62 12
0.3 2 0.3 0.1 2 3 42 19 5
8
1 1
3 4 22 3
3
11,
1
0.5 0.7
0.3 0.3 c.1 2 4 35 23 6
(15)"
5b
7 13 11
5 3 0.7 0.3 2 2 2 1 1
0.4 0.3 0.7 1
1
2 14 27
13 21 2 0.8
lob
10:
0.4
0.1
0.9
0.3 0.2 0.2 0.3
b
1
1
0.3 30 10;
0.7 0.1
2 2
0.6 0.3
1 1
0.8 0.7 0.8 1
0.7
1 0.7 1 0.6 1
0.7 1
0.7
0.7
0.5 0.8 0.3 0.3 5.4
\-,
ANALYTICAL CHEMISTRY
35 10:
51
lot
t.7
0.2
The numbers a t the tops of the columns represent the following compounds: (11) phenyl isocyanate (6) allyl isocyanate (1) HNCO (12) o-tolyl isocyanate (7) n-GHgNCO (2) CHsNCO (13) m-tolyl isocyanate (8) n-CeHi7NCO (3) CzHbNCO (14) p-tolyl isocyanate (4) n-CBHTNCO (9) n-CiaHuNCO (.5) iso-GH,NCO (10) cyclohexyl isocyanate (15) 1-naphthyl isocyanate --- - " . Small peak present as a shoulder,'bit not resolved.
722
(13)" 0.3
0.4
88
6
(12P
(7)0
43 4?
100
DISCUSSION
All of the spectra contained some air background, which has been subtracted. The ni/e 32 peak was assumed to come entirely from air and was removed completely. Proportional corrections were then calculated for the m/e 14, 16, 28, and 40 peaks, with the correction to m/e 40 being negligible. Water and possible carbon dioxide impurities were in general present in rather small concentrations, as Table I indicates. Enough normal alkyl isocyanates were examined to permit some correlation of their spectra. The identity of the base peak is rather variable, being m/e 28,56, 56, 27, 41, and 43 for the methyl, ethyl, propyl, butyl, octyl, and octadecyl compounds. For the same six, the relative intensity of the molwular ion peak decreases steadily from 81% for the methyl isocyanate to 0.3% for the octyl, and then increases again to 5.4% for the octadecyl. Since the octadecyl compound had been introduced in a silica capillary on the direct inlet probe, while the other five, including the octyl, had gone through a stainless steel needle valve and inlet tube, the octyl was reexamined by means of the probe. I t s spectrum showed no significant change, which removed any doubt associated with the stainless steel inlet system. The ratio of the (AT-1) peak to the molecular ion peak height is similarly variable. Among the first five listed above, it reaches a minimum with the ethyl compound, increases, and then for octadecyl isocyanate becomes too small to be measured with the time-of-flight instrument. All of the normal alkyl isocyanates yield an important ion at m/e 56, which is the (CH2NCO)+ fragment. I n the methyl compound it cannot possibly have any other composition. I n the octyl, this composition was verified by high resolution work, recorded in Table 11. Peaks of Iower intensity occur which represent ions heavier than 56 by multiples of 14 mass units. These peaks suggest cleavage a t other carbon-carbon bonds farther removed from the functional group. Only one branched compound, the isopropyl, was available, in which m/e 70 replaces the m/e 56 of the normal ones, in a reasonable correlation.
In the octyl and octadecyl compounds, where the length of the alkyl chain is sufficient, a strong peak is observed for an ion of m/e 99. High resolution measurement of this peak in the octyl isocyanate spectrum showed the composition (CsHgK'O)+. The strucOH
Table 11. High Resolution Mass Measurements on Octyl Isocyanate
m/e 55.0554
56,0140 56.0503 56.0629 57.0581 57.0705 99,0689 112 0754 126 0917
ture is suggested, by analogy with the corresponding sulfur-containing ion of the isothiocyanates, whose formation involves a bicyclic rearrangement requiring an alkyl chain of a t least six carbon atoms (3). High resolution measurement of the peak a t m/e 57 showed largely the composition (CaH7N)+, involving hydrogen rearrangement and loss of both ends of the molecule. A few special features may be pointed out, without further discussion. I n the spectrum of methyl isocyanate, the m/e 27.5 ion, (HCKCO) +2, corresponds to the (HCK'CS)+2 of methyl isothiocyanate, discussed by Hobrock, Shenkel, and Kiser ( 5 ) . However, the true intensity of this doubly charged ion is less than that indicated by Table I, because of poor resolution from the adjacent large peaks. The m/e 30 peak present in a number of the spectra probably represents the (KO) ion corresponding to the (NS) of the isothiocyanates. I n the case of butyl isocyanate, the relative intensity of m/e 30 in an unpurified sample was examined and found to be the same as that in the gas chromatographed sample, indicating that it belonged to the isocyanate spectrum. Although this peak might at first suggest some hydrolysis of isocyanates to primary amines, the absence of the amines is shown by the absence of their molecular ions (4). High resolution work on ethyl isocyanate indicates in the m/e 42 region an intensity ratio for (C2H4N)+/ (XCO) + of 14/1, so that in this case, at least, KCO+ contributes very little. I n the spectra of the aromatic compounds, the molecular ion always gives the most intense peak, as might be expected. Those having a methyl sub+
140.1117
Composition CaH,+ CzH*NO+ CsH&+' CdHs+ CsH?N+
'
CdHo+
C;H;NO+
1
Multiplet intensity Ratio 10/1/4 16/1
'
C~HI"&O+ C~HI~NO+ CsHldNO+
stituent on the ring show strong (11-1) peaks, interpretable in terms of a substituted tropylium ion. What is probably the tropylium ion itseIf, a t m/e 91, is also seen in the spectra of the tolyl compounds, where it may be produced by loss of the functional group NCO. Loss of CO and loss of H plus CO occurs in all, though only the (&I-28)peak is strong in the phenyl compound. Another prominent peak in all of the spectra of aromatic compounds results from loss of 55 mass units, probably CHSCO, recalling the loss of CO and HCO from phenol ( I ) and the napththols (a). ACKNOWLEDGMENT
The writers thank Robert G. Honeycutt for assistance with the gas chromatography.
+
LITERATURE CITED
(1) Aczel, T., Lumpkin, H. E., ANAL. CHEM.32, 1819 (1960). (2) Beynon, J. H., i n "Advances in Mass Spectrometry, Waldron, J. D., ed., p. 347, Pergamon Press, New York, 1959. (3). Rudzikiewics, II., Djerassi, C., Williams, D. H., "Interpretation of Mass Spectra of Organic Compounds," pp.
111-124, 208-212, Holden-Day, San Francisco, 1964. (4) Gohlke, R. S., JlcLafferty, F. W., ANAL.CHEM.34, 1281 (1962). ( 5 ) Hobrock, B. G., Shenkel, R. C., Kiser, R. W., J. Phys. Chem. 67, 1684 (1963). (6) Philippe, R. J., Unpublished data, Liggett and Myers Tobacco Co., Durham, N. C., 1965. RECEIVEDfor review October 11, 1965. Accepted March 10, 1966.
VOL. 30, NO. 6, MAY 1966
e
723
