187
Anal. Chem. 1981, 53, 187-189 (4) "Petroleum Products and Lubricants (I)"; American Society for Testing Materials: Philadelphia, PA, 1976; p 24. (5) Ozubko, R. Anal. Cbem., in press. (6) Brown, J. K.; Ladner, W. R. Fuell960,39, 87. (7) Williams, R. 8.; Chiamberlaln, N. F. "Proc.-WorM Pet. Congr., 6th 1963, Section V, 17. (8) Sternberg, H. W.; Raymond, R.; Akhtar, S. ACS Symp. Ser. 1975,
No. 20, 111. (9) Snape, C. E.; Ladner, W. R.; Battle, K. D. Anal. Chem. 1979, 57, 2189.
RECEIVED for review April 2,1980. Accepted October 20,1980.
Medium-Resolution Mass Spectrometry as a Nitrogen Compound Specific Detector Emilio J. Gallegos Chevron Research Conrpany, Richmond, Callfornia 94802
A gas chromatograph or a direct-insertion probe interfaced to a medlum-resolution mass spectrometer set at about 3000 resolution is used in the multiple Ion detection (MID)mode
to monltor the Intensity of the CH2N+ ion at m / q 28. This conflguratlon converts the system Into a nltrogen compound specific detector. Practicability of this system in terms o? levels of detection, quantitation, etc., is demonstrated by use of authentic nitrogen compound mixtures, gasoiines, and coals. The significance of CO+ and C2H,+ Ion monitoring, also found at m / q 28, Is dlscussed.
Nitrogen compounds contribute to gum formation in fuels and on the pistons, valves, and rings of internal combustion engines. They are ah0 process catalyst poisons. The use of increasingly heavy fractions of petroleum and, in the near future, the refining of coal and shale oils which introduce considerably higher levels of nitrogen will increase the importance of isolating and identifying nitrogen compounds both in the finished product and in the feedstocks to refinery processes. A number of element-specificdetectors have been described ( I , 2). They use a bead of silica doped with an alkali metal. Nitrogen, phosphorus, and sulfur compounds have been detected, although the mechanism of the detector is not well understood, and quantitative reproducibility is difficult to achieve. The method descrilbed here depends upon the detection of the CHzN+ion using a medium-resolution mass spectrometer. This will allow detection of all volatile nitrogen-containing compounds with N-C-H bonding.
EXPERIMENTAL SECTION Two different gas chromatograph-mass spectrometer-computer (GC-MS-C)systems were used in this work. One used a 305-m squalane-coated 0.5 mm i.d. stainless steel capillary column coupled to an AEI M13-9 Nier Johnson double focusing mass spectrometer through a 0.0025-mm vinyl methyl silicon membrane interface. The other system used a 75-m, Dexil400 coated 0.25 mm i.d. glass capillary column coupled directly to the source of a Nuclide 12-90-Gsingle focusing mass spectrometer. The mass resolution of both systems was set to about 3000. An all-quartz direct-insertion probe described elsewhere (3) was used in conjunction with the MS-9 mass spectrometer for analysis of the solid coal samples. Data were gathered at 5-s intervals in the multiple ion detection, MID, mode for the probe work. The data taken from GC-MS-C were gathered at 2-s intervals in the MID mode. 0003-2700/81/0353-0187$01.00/0
Magnetic scanning for fuel mass spectra were gathered at 4.553 intervals on the Nuclide-Dexil system.
RESULTS AND DISCUSSION Figure 1is reproduced from the mlq 28 profile for pyridine. Nz+and CO+ are due to background. This nevertheless shows all possible fragment ions at nominal mass mlq 28. These are CO+, N2+,CH2N+,C13CH3+,and CzH4+taken at about 5000 resolution. At a resolution of 3000, CHZN' and C2H4' may be monitored independently without interference from the others. Table I gives names, boiling point, structural formula, and percent of total ionization of the CHzN+ ion in the mass spectra of 21 of nitrogen-containing compound types expected in gasoline range samples. They are numbered in order of elution time through a Dexil 400 glass capillary column. Figure 2 shows the CHzN+,CzH4' mass chromatograms, and the reconstructed gas chromatograms (RGC) of a standard 21-nitrogen compound mix. These data were generated from the Nuclide-Dexil system. The oven was heated from 50 to 300 "C at 4 OC/min. The numbers on the RGC trace refer to the standard nitrogen compounds listed in Table I. The RGC represents a separate run in which total mass spectra were acquired in the magnet scan mode. The CH2N+and C2H4+ mass chromatograms were acquired by scanning only over the mlq 28 multiplet. Note that the C2H4+ ion intensity for all of the nitrogen compounds is 25% or less than that of CHZN' ion a t mlq 28. Figure 3 shows the CH2N+and CzH4+mass chromatograms and the inverted gas chromatograph flame ionization detector trace of gasoline distillation cuts 20-23. These data were generated from the 305-m MS9-squalane system. Note that the position of nitrogen compounds (see arrow) changes with respect to the hydrocarbon going from cut 20 through cut 23. These data suggest that there is mainly one nitrogen component and that it is the same one in the four cuts. The compound was identified to be Cmethylbenzenamineat a level of about 400 ppm in cut 21. Subsequent dilutions of this cut indicate that the lower limit of detection for this nitrogen compound by the method described here is approximately 5 PPm. A similar analysis was made on the pyrolytic effluent of six United States coals. The coals were finely powdered and introduced into the mass spectrometer via an all-quartz, direct-insertion probe, which places the sample 1cm from the electron beam. They were heated from approximately 100 "C to approximately 600 "C at 10 "C/min. 0 1981 American Chemical Soclety
188
ANALYTICAL CHEMISTRY, VOL. 53, NO. 2. FEBRUARY 1981
loow
CHIN'
mlq
Figm 1. Rofile of
WYODAK
I
loor
-
peaks at m I q 2 8 fa pyridine and background air. 0
C H i K MASS CHROMATOGRAM
300 400 500 600 TEMPERATURE. 'C
CO', CH#, and C&l,+ rekase from t d t y Wyodak mal, direct Insertion probe 100-600 OC at 10 'Clmln. Figun 4.
A m l q 28.031
1
I
0
I'
[
NAVAJO
r/?
100
t0.006
CO'
GASCHROMATOGRAM
1:30
-
1x00 2230 3000 3R30 4 5 0 0 RETENTION TIME. MINSEC
Figure 2. CH,N+ and C2H,* mass chromatogramsand gas chromatogram of a standard 21 nibogen component mix with Uw NffilidsC8xil GC-MS system.
0 300 400 500 600 TEMPERATURE.OC
Figure 5. CO+.CH2N+, and CIH,+ release from cretaceous Naval0 coal direct insertion probe 100-600 'C at 10 OClmin. LOVERIDQE
'oolco+ 100
0
3320
6540
10020
13320
16840
o--*L, 300 400
-
RETENTON TIME MINSEC
Figure 3. CH,N+ and Cg,+ mass chromatogramsand inerted FID trace of gasoline cuts 21-23 with ihe MS-9 Squalane system.
Figures 4-6 show the CO+, CH2Nt, and C&+ release curves for three of the six coals. The Wyodak subbituminous coal and Noonan lignite are tertiary coals, the Hiawatha and Navajo are bituminous cretaceous coals, and the Loveridge
- , L , , 500
600
TEMPERATURE. 'C
Figure 6. CO'. CH2N+. and CzH,*. and Cg,+ release from carboniferous Loveridge ccal,direct insertion probe 100-600 OC at 10 OClmin.
and River King are bituminous A and B carboniferous coals. Table I1 summarizes the average percent release of CO+, CH2N+,and C2H4+for the tertiary, cretaceous, and carboniferous coals. The younger, tertiary coals release the most CO+ and the least C2H4+,followed by the cretaceous and then the
ANALYTICAL CHEMISTRY, VOL. 53, NO. 2, FEBRUARY 1981
189
Table I. Synthetic Nitrogen Compound Mixture Compound (Molar Massl
70 of Total Ionization for C H ~ N +at mie 28
BP, "C
4.9
12. N, N-Dimethyianiline 11211
85-91
2.7
13. N-12-AminomethyliPiperazine (1151
116
0.51
173)
11011 3. Pyridine
Compound (Molar Massl
55
1. Diethylamine
2. Triethylamine
5 Of
I791
14. 3-Amino-4-Methylpyridine ( p Amino-a-picoline) (1081
BP, "C
Total Ionization for CH?N+ at m/e 28
121
0.19
4.53
82-851
4. M)
12 m m
4. 3-Picoline CP1 1931
144
0.61
0.98
145
1.03
15. 2,4-Xylidine 11211
218
5. 4-Picoline (a1 1931
152
0.31
16. 3,4-Xylidine (1211
226
0.86
6. 2,2,6,6-Tetramethyi-
1651
0. b l
piperidine (1411 7. 2,5-Oimethylpyrrole 1951
17. 2,3-Xyiidine 740
rnm
(1211
18. N-Methylindole (1311
104-1051 12 mm
0.81
1021
0.15
5 mm
159
0.60
0.6
0.56
19. Indole Il-Benz(b1pyrroieI (1171
253
143
160-162
0.26
20. 2-Methylindoie 11311
273
0.12
21. Carbazole 1167)
255
0.44
8. 2,4-Lutidine
(1071 9. 2,5-Lutidine
11071 10. N,N-Dimethylcyciohexyiamine 11271 11. 3.4-Lutidine 1107)
163
0.40
Table 11. CO+and CH,N+ Percent Contribution to m/e 28 tertiary carboniferous coals cretaceous W:yodak, Navajo, River King, Noonan, Hiawatha, Loveridge, %
%
%
CO' CH,N+
83.5
65.4
1.6
1.3
C2H4 +
14.9
33.3
53.8 1.3 44.9
carboniferous coals. These results are consistent with the theory of coal formation (4). A well-defined trend for relase of CH2N+across geological time is not apparent from the values obtained. It is interesting to note (Figures 4-6) that some CH2N+ producing components are released about 100 "C earlier than the bulk of the CzH4+ or CO+ producing components. The early release components (dashed line area) may well be due to aliphatic amine compounds which show a high CH2N+ sensitivity; whereas, those CH2N+ producing components which evolve at the same temperature as the bulk of the CO+ and C2H4+ producing components are due to aromatic nitrogen-type compounds, which show a much lower CH2N+ sensitivity. (See Table I.) This may have important consequences in coal processing. To the author's knowledge, this is the first report of the nature of nitrogen compounds evolved
from the thermal decomposition of coal.
CONCLUSIONS
-
This work has demonstrated the practicability of using a medium-resolution, M / A M 3000, mass spectrometer as a nitrogen-specific detector. This work was done on an -18 year old vintage AEI MS-9 and 9-year old single focusing Nuclide mass spectrometers. Present high-resolution instruments would probably give a 10- to 100-fold increase in sensitivity at the same resolution. This technique has found application in monitoring organic sulfur (5) compounds by use of the CHS+ fragment and can obviously be extended to other heterocompounds which give moieties (such as COz+)that are resolvable at low mass.
ACKNOWLEDGMENT The author wishes to thank A. L. McClellan and P. R. Ballinger for useful suggestions in the preparation of this manuscript.
LITERATURE CITED (1) Kolb, B.; Auer, M.; Pospisil, P. J . Chromafogr. Scl. 1077, 75 (2),
53-63. (2) Albert, D. K. Anal. Chem. 1978, 50, 1822-1829. (3) Gallegos, E. J. Proc. World Pet. Congr., 7th 1967, 4 , 249-260. (4) Tlssot. 8. P.; W e b , D. A. "Petroleum Formation and Occurrence"; Springer-Verlag: New York, 1978; pp 202-224. (5) Gallegos, E. J. Anal. Chem. 1975, 47, 1150-1154.
RECEIVED for review April 14,1980. Accepted October 6,1980.
