2054
Anal. Chem. 1980, 52, 2054-2057
(14) Stanley, W. L.; Vannier, S. H.; Gentilli. B. J. Assoc. Off. Anal. Chem. 1957, 40, 282. (15) Bower, D. E.; Leat. W. M. F.; Howard, A. N.;Gresham, G. A. Biochem. J . 1983, 189. 24P. (16) Brown, T. L.; Benjamin, J. Anal. Chem. 1984, 36, 446. (17) Amos, R. J. Chromatogr. 1970, 48, 343. (18) Bevenue, A.; Kelley, T. W.; Hylin, J. W. J . Chromatogr. 1971, 54, 71. (19) Hutzinger, 0.; Safe, S.; Ziko, V. ”The Chemistry of PCBs”; Chemical Rubber Co. Press: Cleveland, OH, 1974. (20) Poole, C. F.; Wibberly, D. G. J. Chromatogr. 1977, 132. 51 1. (21) . . Lamoarski. L. L.: Mahle, N. H.: Shadoff. L. A. J. Auric. - Food. Chem. 1978, 2 6 , 1113. (22) Peel, D. A. Nature (London) 1975, 254, 324. (23) Eiceman, G. A.; Clement, R. E.; Karasek, F. W. Anal. Chem. 1979, 57, 2343. (24) Clement, R. E.; Eiceman, G. A.; Karasek, F. W.; Bowers, W. D.;Parsons, M. L. J. Chromatogr. 1980, 169, 53. (25) Matsumura, F.; Ward, C. T. Project No. OWRT A-058-W1S; Wisconsin University: Madison, WI, 1976. (26) Buser, H. R.: Bosshardt, H. P. J. Chromatogr. 1974, 9 0 . 71.
(27) Buser. H. R.; Bosshardt, H. P. J. ASSOC. Off. Anal. Chem. 1976, 59, 562. (28) Rappe, C.; Marklund, S.;Buser, H. R.; Bosshardt, H. P. Chemosphere 1978, 3, 269. (29) Buser, H. R. Anal. Chem. 1977, 49, 918. (30) Buser, H. R . Chemosphere 1979, 8 , 251. (31) The Dow Chemical Co.,Trace Chemistries of Fire, 1978. (32) Blair, E. H., Ed. Adv. Chem. Ser. 1973, No. 120. (33) Boer, F. P.; Van Remoofiere, F.; Muelder, W. W. J. Am. Chern. Soc. 1972, 9 4 , 1006. (34) Nestrick, T. J.; Lamparski, L. L.; Stehl, R. H. Anal. Chern. 1979, 51, 2273. (35) Aniline, 0. Adv. Chern. Ser. 1973, No. 120, 126. (36) Bredeweg, R. A.; Rothman, L. D.; Pfelffer, C. D. Anal. Chem. 1979, 51, 2061. (37) Firestone, D. J. Assoc. Off. Anal. Chern. 1977, 60. 354.
RECE~VED for review January 28,1980. Resubmitted May 14, 1980. Accepted July 31, 1980.
Secondary Ion Mass Spectra of Diquaternary Ammonium Salts Timothy M. Ryan, Robert J. Day, and R. Graham Cooks” Department of Chernisrty, Purdue University, West Lafayette, Indiana 47907
Molecular dications emitted by momentum transfer processes are observed in secondary ion mass spectra (SIMS) of dlquaternary ammonium salts. The relationship between molecular structure and the observation of dlcatlons Is explored. Large Intercharge separations, correspondlng to lessened lntramolecular coulombic repulslons, are observed to correlate with dlcatlon detectlon. Fragmentatlon with charge separation is facllitated by small intercharge distances and can preclude observation of the dicatlon. Electron attachment to yield the monocatlon Is an alternative to dlcation emission when the structure of the dication facilitates reduction. This occurs, for example, for the herblclde diquat ( N,N’-ethylene-2,2’-bipyrldyl dlbromlde) which is detected as Its monocation. Complete spectra of diquaternarles can be taken wlth nanogram size samples.
Secondary ion mass spectrometry (SIMS) has recently been shown to be a sensitive method for the characterization of organic salts (1-4). Reported here is the observation of intact organic dications emitted from diquaternary ammonium salts upon sputtering. This constitutes the first observation of multiply charged organic molecular ions in SIMS. The result is of interest with regard to both analytical applications of SIMS and the fundamentals of ionization during sputtering. Specifically, some biologically important compounds, such as the herbicides paraquat and diquat and the curare alkaloids, have the diquaternary structure, so that SIMS may facilitate their characterization. In addition, studies on organic dications reflect the degree to which electron attachment occurs during sputtering. This process yields observable charged products for dications, but neutrals are sputtered when monocations are reduced during ion bombardment.
EXPERIMENTAL SECTION All compounds were synthesized by using standard methods for the preparation of quaternary ammonium salts. The organic salts were burnished onto a 1 cm2roughened foil of either silver 0003-2700/80/0352-2054$01.00/0
or platinum prior to SIMS analysis using argon primary ions at 5 keV and 0.3-0.5 nA primary ion current. Beam diameter was approximately 1 mm and pressures in the ultra-high-vacuum chamber remained below 1 X lo-* torr during the course of the experiments. All spectra were taken with Riber SIMS system using a quadrupole mass analyzer, Channeltron electron multiplier, and pulse-counting electronics. Intercharge distances were measured by using Dreiding modeh; charge localization on nitrogen was assumed and the maximum distance in the unstrained molecule is reported. Intercharge distances ( r )were used to calculate coulombic repulsive energies ( r ) from T (eV) = 14.6/r (A).
RESULTS AND DISCUSSION The SIMS spectrum of N,N’-bis(dimethyl)-4,4’-trimethylenedipiperidine diiodide (1) is shown in Figure 1. This spectrum provides both the molecular weight (inferred from the highest mass doubly charged ion, 2682f) and structural information on the compound. Emission of the doubly charged species is confirmed by the observation of the 13C isotope peak one-half mass unit above the dication peak (m/z 134.5 in Figure 2). Changing the counterion does not affect the SIMS spectrum; for example, the dibromide and diiodide of compound 1 gave identical SIMS spectra. Analogous results were obtained for N,N’-bis(ethy1methyl)-4,4’-trimethylenedipiperidinediiodide (2) and for the aromatic compounds N,N’-bis(trimethyl)-4,4’-methylenedianiline diiodide (3) and N,N’-bis(dimethylethy1)-4,4’methylenedianiline diiodide (4).The spectrum of compound 3 is shown in Figure 3; the dication, 2842+at m / z 142 is of relatively low abundance, but its 13C isotope is well resolved in high-resolution scans. A considerable number of diquaternary salts (5-19, Table I) did not exhibit observable dications. Compounds 18 and 19, while they did not yield molecular dications, did show the corresponding singly charged ions in their SIMS spectra. Compounds 5-17 may fail to exhibit dications because they fragment by a favorable charge separation route, M2+ MI+ + M2+. This is indicated by the absence of both singly and doubly charged molecular ions for these samples.
-
0 1980 American Chemical Soclety
ANALYTICAL CHEMISTRY, VOL. 52, NO. 13, NOVEMBER
1980
2055
I79
1
164 ~
58 X IO 112
,
126
I36
!g
‘,
‘50
149
, ,
/,
.
I
.
i1,
‘93
,
-
”c
m
I70
150
190
I
0
n
I
c 3
n
I
182
Flgure 1. SIMS spectrum of the diquaternary diiodide 1.
i, -
P
I
L
m/z
130
110
150;
170
-
190
--
-
Flgwe 4. Comparison of the SIMS specira of the diquaternary 14 and its monomethylated analogue (intact monocation at m l z 179).
DOUBLY
CHARGED
183
MOLECULAR CATION
156
130
I32
I34
I36
138
I40
I20
I40
I60
I00
200
m/z Flgure 2. Portion of the spectrum of 1 at higher resolution showing
the partially resolved
I3C isotopic
peak of the intact dication at m l z
134.5.
134
m
Flgure 3. SIMS spectrum of the diquaternary diiodide 3.
It is noteworthy that the dications detected have larger calculated intercharge distances (Table I) than those which do not survive the sputtering process. Coulombic repulsion may therefore contribute significantly to driving the frag-
Flgwe 5. SIMS spectrum, taken on 1 ng of the dquatemary dibromide 19.
The monocation (184’) is indicated as G’.
mentation reactions. The coulombic energy associated with the observed dications was calculated to be 1.3-1.5 eV, while that for unstable species was greater than 1.5 eV. As an example of a compound which readily undergoes dissociation with charge separation, Figure 4 shows the SIMS spectrum of the diquaternary 14, as well as that of its monoquaternary analogue. The similarity between these spectra indicates that facile dissociation of the dication by charge separation occurs to generate the monocation ( m / z 179) from which lower mass fragments arise. It is noteworthy that the substantial calculated coulombic repulsion in the nascent dication, 2.5 eV, can account for its instability. A second reason for failure to observe the dication is electron capture, a process expected to occur when the electron affinity of M2+ is positive. This is probably precluded in saturated systems; however, one-electronreduction does occur (although not necessarily as a gas-phase process) for the herbicides paraquat and diquat. These compounds, N,”dimethyl-4,4’-bipyridyl dichloride (18) and N,N’-ethylene2,2’-bipyridyl dibromide (19), respectively, are reduced in SIMS as evidenced by the emission of abundant intact monocations, M+., under ion bombardment. The spectrum of diquat (19), Figure 5, was obtained on a 1-ng sample which
2056
ANALYTICAL CHEMISTRY, VOL. 52, NO. 13, NOVEMBER 1980
Table I. Diquaternary Studies, Their Measured Interchange Distances and Calculated Coulombic Repulsive Energies CCMPCJhlD
STRJCTURE
WTERCdARGE
DIS-4rlCE
CCULOMBIC
ENERLY
QEPVLSIVE
ievl
I
10.9
I ,3
2
10.9
1.3
9.5
I .5
9.5
I.5
7.5
1'9
6
6.3
2,3
c
7
3.0
4.8
0
3.0
4,8
7.3
2.0
3.0
4.8
3.0
4.8
5.8
2.5
13
5.1
2.8
-14
5.8
2.5
15
5-1
2.8
16
9.5
1.5
17
5.2
2,8
7.I
2.0
2.8
5.2
rc
cc
3 c
-4 -5
c
II
c CHx
12 N
h
/v
18 N
19
lasted for several hours, demonstrating the low detection limits attainable. An interesting contrast exists between the behavior of diquat (19) and its ring opened analogue, compound 17. Both nascent cations have large coulombic energies (5.2 eV), and the absence of the intact dication is therefore expected. However, while 17 shows only singly charged fragment ions,
19 shows a singly charged intact molecular cation. The existence of low-energy fragmentation pathways, such as methyl loss in 17 but not in 19, rationalizes this observation. All the compounds, both those which emit dications and those which do not, give spectra which are rich in singly charged fragment ions. The formation of these even-electron species can be rationalized in terms of a few simple frag-
Anal. Chem. 1980, 52, 2057-2061
mentation types. In particular, dealkylation processes, as already illustrated in Figure 4, and elimination reactions accompanied by hydrogen transfer are common. These processes lead to two series of fragment ions from compound 1; viz.,m / z 112, 126, 140, and 154 and m / z 114, 128, 142, and 156. Structures of the type :Nc)-C /
n H2n+ I
and
'/ ~ 3 - c "HZn-1 accommodate this result. Direct examination of the monoquaternized analogue of 1 suggests that it does indeed serve as a precursor for these even-electron fragments. In other words, the monoquaternary and the diquaternary gave very similar spectra except that the dication m / z 134 and the ion m / z 267 were absent for the former compound. Monocations, formed by reduction, are also a source of even-electron fragment ions. These observations on intact singly and doubly charged ions and their fragment ions, when viewed with regard to the types of structures examined here, permit the following postulate: ejection of dications from the solid into free space is a favorable event only if the dication does not possess orbitals into which electrons can be donated. If it does, the monocation and its fragments are detected instead. If this partial neutralization is energetically unfavorable, then the dication will be observed unless dissociation into two singly charged ions
2057
is facile. This is often the case unless the charge separation is relatively large. In conclusion, it has been shown that SIMS is a sensitive method for the analysis of salts containing organic dications. The observation of the dication has been found to be related to its internal energy as reflected in the charge separation in the nascent dication. Even when intact dications are not detected, SIMS spectra containing structurally diagnostic fragments are obtained. These observations are significant in light of the limited number of techniques which yield high-quality mass spectrometric data for organic salts. They also provide further evidence of the phenomenological similarity between ionization by SIMS and by field desorption, doubly charged ions being recorded in the field desorption (FD) spectra of diphosphonium salts (5).
LITERATURE CITED (1) Ullmann, R. Mikrocbim. Acta 1979, 221. ( 2 ) Unger. S. E.; Ryan, T. M.; Cooks, R. G. Anal. Cbim. Acta, in press. ( 3 ) Day, R . J.; Unger, S. E.; Cooks, R. G. J . Am. Cbern. SOC. 1979, 101, 501. ..
(4) Sichtermann, W.; Junack, M.; Eicke, A,; Benninghoven, A. Fresenius' 2.Anal. Chem. 1980, 301. 115-116. (5) wood, G. w.; McIntosh;J. L ~ W P. , Y. J . org. chem. 1975, 40,638.
M.I
RECEIVED for review April 9, 1980. Accepted July 28, 1980. This work was supported by the National Science Foundation CHE-78-08728 and MRL Program DMR 77-23798. R.J.D. thanks the Analytical Division of the American Chemical Society for a summer fellowship.
Determination of Tetrachlorodibenzo-p-dioxins by Mass Spectrometric Metastable Decomposition Monitoring Edward K. Chess and Mlchael L. Gross" Department of Chemistry, University of Nebraska, Lincoln, Nebraska 68588
A method of analysis of tetrachlorodlbenzo-p-dioxin (TCDD) has been developed In which the loss of COCl from the molecular Ion is monitored. Because the reactlon is antlclpated to be highly speclflc for TCDDs, the purified sample extracts were introduced in the mass spectrometer by using the direct Insertion probe Instead of gas chromatography. The precision, quality of callbration, and detectlon limit have been evaluated by analyzing standard solutions. FHty plcograms of TCDD can be detected wlth a signal-to-noise ratio of 8:l. The accuracy and specHiclty have been tested by comparlng the results of thls analysls with those using gas chromatography/hlgh-resolutlon mass spectrometry (GCIHRMS). Actual sol1 samples taken In the vicinity of a chemlcal industry landfill were used. The speciflcity Is comparable to the GC/HRMS method; however, the accuracy ls not as good. The procedure is appropriate for rapld screening of purlfied sample extracts and for valldatlng results from other analyses.
This paper is a report of a rapid screening technique specific for tetrachlorodibenzo-p-dioxin(TCDD) a t the low partsper-trillion levels using a normal Nier/ Johnson geometry double focusing mass spectrometer operated in a defocused metastable mode. This method, herein called specific reaction monitoring, has been evaluated by comparing it with a proven 0003-2700/80/0352-2057$01 .OO/O
low-resolution gas chromatography/high-resolution mass spectrometric procedure. Furthermore, the technique has been applied to the analysis of environmental soil samples taken from the vicinity of chemical landfills. The urgent need for parts-per-trillion analysis of toxic environmental contaminants, such as TCDD, has led to the development of several state-of-the-art analytical methods (1-4). For assurance of high specificity, either the sample cleanup, separation, or detection must be done by using high-performance procedures. In fact, low-resolution gas chromatography (LRGC) coupled with low-resolution mass spectrometry (LRMS) has been found to be nonspecific for TCDD and subject to false postives unless high-performance cleanup procedures are used (5). LRGC used in conjunction with high-resolution mass spectrometry (HRMS) is specific for TCDDs but lacks the chromatographic resolution necessary to distinguish each of the 22 TCDD isomers. This method has been employed extensively by scientists at Dow Chemical (I) and by Gross and co-workers (2). A totally isomer specific analysis can be done by employing high-performance liquid chromatography in the cleanup (6) or possibly by use of high-resolution capillary column GC coupled with LRMS (3) or HRMS ( 4 ) . As an alternative to HRMS, the needed specificity may be obtained by using negative ion mass spectrometry with either a chemical ionization source (7,8) or an atmospheric pressure 0 1980 American Chemical Society
