J. Phys. Chem. 1994,98, 3930-3932
3930
Infrared Detection of Matrix-Isolated, Mass-Selected Ions Thomas M. Halasinski, Jerry T. Godbout, John Allison,* and George E. Leroi' Department of Chemistry, Michigan State University, East Lansing, Michigan 48824- 1322 Received: January 14, 1994; In Final Form: February 18, 1994"
Sufficient quantities of mass-selected cations have been isolated in inert matrices for vibrational spectroscopic observation for the first time. When 15-25-nA beams of CF3+ from CF3C1, CFsBr, or CF3H are co-deposited for 10-25 h with neon or argon a t 5 K, the antisymmetric stretching vibration (u3) of the cation can be observed by FTIR spectroscopy. The mechanism by which the matrix maintains the necessary approximate neutrality is not certain; however, both positive and negative charges are detected when the matrices containing massselected cations are allowed to warm.
Introduction In mass spectrometry, molecular structures are determined by considering the ionic species formed by electron impact ionization. The applicability of mass spectrometry is widespread among chemists, yet few ionic structures have actually been determined directly. Vibrational spectroscopy,the most direct approach for determining molecular structures, is not sufficiently sensitive to be used for the characterization of ions at the concentrations encountered in mass spectrometry. A number of approaches for generating ionic species to be entrapped in inert matrices for spectroscopic studies have been proposed.' Because preselection of the desired ion is lacking, or full structural characterization is not provided by the methodology, the mechanisms developed to date for spectroscopic characterization of ions by matrix isolation do not provide general approaches for studying the types andvariety of ions that areof interest to thechemistry community. We have proposed instrumentation for the accumulation and infrared spectroscopiccharacterization of mass-selected ions using matrix-isolation techniquesq2In the method, both mass-selected cations and anions can be deposited with an inert host on a window held at low temperature for subsequent spectroscopicexamination. Spectroscopicdetection of mass-selected,matrix-isolated cations was first demonstrated by Maier and co-workers using UV-visible absorption' and later extended to laser-induced fluorescence in our laboratorya2 Our ultimate goal is to couple mass-selective ion sources that produce sufficient numbers of both positively and negatively charged species with the appropriate matrix isolation hardware to allow the determination of ionic structures by vibrational spectroscopy. Upon completion of an ion source that produces suitable currents of mass-selected CF3+, we found that these cations can be trapped in a matrix, allowing for infrared measurements to be carried out without the concurrent deposition of negatively charged species. The mechanism by which the necessary approximate overall neutrality of the matrix is realized is incompletely understood. We report here the FTIR identification of CF3+, mass selected from several precursors and isolated in neon or argon matrices. Experimental Section Although the principles on which it is based remain unchanged, the experimental apparatus used in this study (Figure 1) has been modified from that described previously.2 The Air Products Displex 202 cryostat which cooled the matrix substrate to 15 K has been replaced with an APD HS-4B Heliplex system with a - 4 K capacity to permit the use of neon matrices. The CsI sample window temperature is measured with a Lakeshore Cryotronics 330 controller and a DT-470 Si diode sensor; the *Abstract published in Advance ACS Abstracts, April 1, 1994.
0022-3654/94/2098-3930$04.50/0
controlleris also utilized to warm the sample window for annealing experiments. The cryostat chamber is evacuated with an APD-8 cryopump, and the radiation from a Nicolet 520P FTIR spectrometer is transferred through the chamber to a remote, narrow-band mercury-cadmium-telluride (MCT) detector. A modified Finnigan 3200 mass spectrometer provides massselected ion currents that are at least an order of magnitude larger than those obtained from themodified residual gas analyzers used previously$ and the flux of neutrals has been reduced considerably, owing to the differentially pumped ion source housing. Cations are created by a Finnigan electron impact (EI) ion source with kinetic energies of 100-110 eV, using 1OO-eV electrons. Because of its favorable infrared spectroscopic properties, CF3+ was chosen as the mass-selected ion for this initial work. Also, its relatively low electron affinity (9.2 eV5) allows observation in both argon and neon matrices without significant spectral perturbation. The mass spectra of CF3X (X = C1, Br, H) show that the resolution of the quadrupole can be degraded to less than unity to increaseion transmission,without introducing ionic species other than CF3+ into the experiment. Ion optics (Figure 1) focus the mass-selected ions and direct them to the CsI sample window. A dc octopole focuses6the ions exiting the mass filter through an aperture, which limits the diffusion of neutral precursors and diffusion pump oil, into the cryopumped vacuum chamber. At this point, a set of plates deflects the ion beamonto the sample window. Typical CF3+currents(measured at the sample window with a Faraday plate) in the experiments reportedrangedfrom 1.5 t o 2 5 (X 1W)A. Thisratecorresponds to a flux of 0.6-0.9 nmol/h. A wide range of pressures are present throughout this instrument. The cryopumped cryostat vacuum chamber has a base pressure of 1.O X 1O-Q Torr. While precursor gas pressures in the ion source are typically 5-1 5 mTorr, differential pumping provides quadrupole mass filter chamber pressures of -2 X 10-6 Torr. When ions are being formed, some precursor gas molecules enter the vacuum chamber, raising the pressure to -2 X Torr. When matrix gas is introduced, the overall chamber pressure rises to -2.5 X le7Torr. The beam of mass-selected ions was co-deposited with excess neon or argon onto the sample window held at - 5 K. Highpurity neon (AGA) and argon (Matheson), both 99.9995%,were used as the matrix gases, with flow rates through the sprayer located in front of the window of 0.5 mmol/h in all experiments. All halocarbon gases (CFsCI, CF'Br, CF3H) were used as received. FTIR spectra were recorded between 900 cm-' (limited by the CaFz windows on the cryostat vacuum chamber) and 4000 cm-' with 1-cm-1 resolution by averaging 256 scan files. The spectra taken before the deposition were used as background for correction of the absorption after matrix formation. The movable Faraday plate, used to measure the mass-selected
-
0 1994 American Chemical Society
The Journal of Physical Chemistry, Vol. 98, No. 15, 1994 3931
Letters
Figure 1. Schematicillustrationof the experimental chamber constructed to isolate mass-selected ions in a low-temperature inert gas matrix for
subsequent structural analysis via infrared spectroscopy: (A) ion source; (B) quadrupole mass filter;(C) dc octopole;(D) einzel lens; (E) deflection plates; (F) diffusion pump; (G) aperture; (H)cryopumped UHV sample chamber; (I) FTIR source; (J) MCT detector; (K) matrix gas inlet; (L) Heliplex cryostat radiation shield; (M) CsI window; (N) window holder; (0)Faraday plate on linear motion feedthrough; (P) CaFz window on conflat flange; (Q) future negative-ion source.
J
10.002 A.U.
C
I
I
I
I
1700
1680 1660 1640 Wavenumbers (cm") Figure 2. Infrared absorption spectra in the 1700-1640-~m-~region of neon matrices after 11 h of deposition of (a) m / z 69 (CF3+, 20 nA) generated from electron impact of CFsCI, (b) m / z 120 (no ion current) while subjecting CF3C1 to electron impact, and (c) m / z 50/51 (CF2+/ CF*H+, 15 nA) generated from electron impact of CF3H. ion current at the sample window, was also used to collect and measure the current of both positively and negatively charged species as they were released from the matrix upon warming. The current from the Faraday plate was converted to a voltage signal with a Keithley 480 picoammeter, digitized with a Keithley DAS-8 data acquisition board, transferred to an IBM 486 compatible microcomputer for storage, and analyzed with Keithley Easyest LX 1.0 software. Results and Discussion
Co-deposition of 15-25-nA beams of CF3+ from CF3C1, CF3Br, or CF3H and neon for up to 25 h onto the cryogenic substrate resulted in the appearance of two new infrared absorptions, at 1670 and 1651 cm-I. Only these bands, shown for generation of CF3+from CF3C1in Figure 2 (trace a), among absorptions due to matrix-isolated HzO, COz and precursor molecules, have intensities that correlate with the integrated ion current. (The
most prominent of these background absorptions are those of the precursor, CF3Cl. After 11 h of deposition, the absorbance of the most intense of these transitions, V I at 1104 cm-I,7 was 0.27.) The 1670 and 1651-cm-I peaks appeared both with, and in the absence of, CC14 mixed with the neon as a potential electron scavenger. We assign the 1670-cm-I peak to the antisymmetric stretch (v3)of mass-selected, matrix-isolated CF3+. The identity of the 165 1-cm-l feature is not certain. The frequency does not correlate with previously identified fragments of CF3Cl, and the band is not observed when argon matrices are employed. It might arise from CF3+ in a different matrix environment, although 19 cm-1 is unusually large for a site splitting. Control experiments demonstrate that the 1670- and 1651cm-I absorptions are not due to neutral species generated during the ionization process. When the mass spectrometer was set to select m / z 120, but all other experimental parameters were retained, no ions were transmitted for deposition in the matrix, yet the flux of neutrals from the ion source remained. During such experiments, the peaks at 1670 and 1651 cm-I did not appear (Figure 2, trace b), but the rest of the spectrum was unchanged. To determine if these features were due not to mass-selected CF3+ but to a secondary species formed upon collisions of the ions with solid surfaces, gas-phase species, and/or the growing matrix, mass-selected CFzH+( m / z 5 1) (along with a small amount of CF2+ ( m / z 50)) was deposited into the neon matrix. As illustrated in trace cof Figure 2, the new peaks wereagain absent. (It should be noted that bands attributable to CFzH+ or CF2+ were not observed, but based on previous assignmentsEthe CFzH+ absorptions would have been obscured by background water absorptions, and the amount of CF2+ deposited was too small to observe in this experiment.) The position of the 1670-cm-I band is consistent with a previous assignment to u3 of CF3+ matrix-isolated in neon.la In that study, split absorptions at 1670 and 1664 cm-* were reported; we did not observe the feature at 1664 cm-I. The antisymmetric stretch of CF3+ also has been reported in the range 1663-1667 cm-1 in argon matrices.E-10 When we deposited mass-selected CF3+in anargonmatrxat 5 K,anabsorptionat 1667 cm-'withanapparent shoulder at 1665cm-I was observed. If the CF3+ ions are partially neutralized during deposition, spectral features due to CFJ should be observed. An extremely weak absorption was observed in neon matrices at 1254 cm-I, which agrees with previous assignmentsla for CF3 in neon. The low intensity of this feature (absorbance = 0.0007) indicates that only barely detectable quantities of CF3 are present in the matrix. Features due to other neutralization/fragmentation products, such as CFz1l and CF," are not observed. In the absence of counter charges, one cannot accumulate a sufficient number of positively charged species in a matrix to obtain a measurable infrared absorption spectrum. By an as yet unknown mechanism, electrical neutrality must surely be maintained. Some negatively charged species are in fact present and detectable in the matrix, although their identity is not certain. When the Faraday plate, installed to monitor the impinging ion flux, was placed directly in front of the cryostat window as neon matrices containing mass-selected cations were warmed to about 20 K, current was measured. As shown in trace b of Figure 3, very small (picoamp) transient signals due to both positive and negative species entrained in the vaporizing matrix were measured when the neon matrix was warmed. For the control experiments, where the peaks attributable to CF3+ were absent, nosuch transient ion signals were observed. These results indicate that negatively charged species are being produced and matrix isolated as a result of the deposition of positive ions. This experiment also provides some insight into the composition of the matrix. If the matrix were to have a homogeneous distribution of positive and negative charge carriers, either they would recombine and be neutralized or their signals would cancel. The presence of transient currents
3932 The Journal of Physical Chemistry, Vol. 98, No. 15, 1994
Corresponding peaks are not readily observed in our experiments with neon matrices; this may indicate that the mechanism of counterion formation is matrix dependent.
20
-
0
d -20 =:
Conclusions CF3+ was chosen since it is produced in high yield from a variety of precursors, and its vibrational transitions in rare gas matrices are known and are relatively intense. The work reported here demonstrates that the FTIR spectrum of this cation can be observed when mass-selected beams of CF3+ are deposited into neon or argon matrices. We will continue to use CF3+ as a probe to more fully understand the processes occurring during ion deposition. We are also developing a similar source of massselected negative ions to further investigate questions related to the requirement of neutrality. Then we can implement this combination of mass selection, matrix isolation, and vibrational spectroscopy to determine the structure of ions of general chemical interest.
C
2 5 20
0
d0
=6
Letters
o -20
-40
5 LL -60 I
5
I
15 i5 Time (seconds)
I
35
Figure 3. Transient current observed at the Faraday plate located
-
1
cm from the substrate during warming of (a) an argon matrix after a 25-h deposition of 20-nA CF3+generated from electron impact of CF3C1, followed by 2 h of 40-nA, IO-eV electron bombardment, and (b) a neon matrix after a 25-h deposition of 20-nA CF3+ generated from electron impact of CF3C1. The final temperatures of the argon and neon matrices were -40 and -20 K, respectively.
of both sign may indicate that the counterions are being formed in discrete time periods, rather than continuously, suggesting some sort of layering in the matrix. When an argon matrix containing CF3+was subjected to 2 h of 40-nA, 10-eV electron bombardment (the CF3+ absorption was not reduced after this electron bombardment), the transient signal shown in trace a of Figure 3 was measured as the substrate was warmed. In this case, desorbed negative charge carriers clearly dominate. This experiment suggests that the matrix can tolerate some charge imbalance. Perhaps during positive ion deposition counterions are formed only after the potential in the matrix is sufficiently high to activate the counterion generation process. When argon was employed as the host, we also observed new FTIR bands at 938 and 933 cm-l, comparable in intensity to the 1670 and 1651-cm-l peaks, when CFj+ was deposited. These bands correspond closely to those previously assigned to CF3Clin argon by Prochaska and A n d r e w ~ .Presumably ~ this species is formed by electron attachment to CF3C1, which is present in the matrix in abundance. Dissociativeelectron attachment could also produce C1-, which is not detectable by FTIR. The origin of the electrons, however, is not clear. They may be extracted from the grounded window holder, or they could be ejected from the various grounded components of the cryostat upon impact of the ion beam and subsequently deposited into the matrix.'*
Acknowledgment. The authors thank the MSU Chemistry Department machine shop and electronic shop for their help in constructing the experimental apparatus. This work was supported in part by National Science Foundation grant C H E 9204136. References and Notes (1) Methods for producing ions for matrix-isolation spectroscopy are reviewed in the following: (a) Chemistry and Physics of Matrix-Isolated Species; Andrews, L., Moskovits, M., Eds.; Elsevier Science Publishing: Amsterdam, 1989. (b) Ion and Cluster Ion Spectroscopy and Structure; Maier, J. P., Ed.; Elsevier: Amsterdam, 1989. (c) Radical Ionic Systems; Lund, A., Shiotani, M., Eds.; Kluwer: Dordrecht, TheNetherlands, 1991. (d) Jacox, M. E.; Thompson, W . E. Res. Chem. Intermed. 1989, 12, 33. More recent publications on this subject include the following: (e) [microwave discharge] Forney, D.; Thompson, W. E.; Jacox, M. E. J . Chem. Phys. 1993, 99, 7393. (f) [X-irradiation] Knight, L. B., Jr.; Tyler, D. J. Kudelko, P.; Lyon, J. B.;McKinley, A. J. J. Chem. Phys. 1993,99,7384. (9) [vapor-phase electron impact] Szczepanski, J.; Vala, M.; Talbi, D.; Parisel, 0.;Ellinger, Y .J. Chem. Phys. 1993,98,4494. (h) [corona discharge] Bai, H.; Auk, B. S. Chem. Phys. 1993,169,3 17. (i) [chemical ionizationdischarge] Hacaloglu, J.; Suzer, S.; Andrews, L. J. Phys. Chem. 1990, 94, 1759. (j)[laser vaporization] Vala, M.; Chandrasekhar, T. M.; Szczepanski, J.; Pellow, R. J. Mol. Struct. 1990, 222, 209. (k) [xenon and mercury arc photolysis] Risinen, M.; Seetula, J.; Kunttu, H. J . Chem. Phys. 1993, 98, 3914. (2) Sabo, M. S.; Allison, J.; Gilbert, J. R.; Leroi, G. E. Appl. Spectrosc. 1991, 45, 535. (3) (a) Forney, D.; Jakobi, M.; Maier, J. P. J . Chem. Phys. 1989,90,600. (b) Maier, J. P. Mass Spectrom. Reu. 1992, 11, 119. (4) Gilbert, J. R.; Leroi, G. E.; Allison, J. In?. J . Mass Spectrom. Ion Processes 1991, 107, 247. (5) vssing, F. P. Bull. Soc. Chim. Belges 1972, 81, 125. (6) Birkinshaw,K.; Hirst, D. M.; Jarrold, M. F. J . Phys. E Sci. Instrum. 1978,11, 1037. (7) Biirger, H.; Burczyk, K.; Bielefeldt, D. Willner, H.; Ruoff, A, Molt K. Spectrochim. Acta 1979, 35A, 875. (8) Kelsall, B. J.; Andrews, L. J . Phys. Chem. 1981,85, 2938. (9) Prochaska, F. T.; Andrews, L. J . Am. Chem. Soc. 1978,100,2102. (10) Jacox, M. E. Chem. Phys. 1984,83, 171. (11) Milligan, D. E.; Jacox, M. E. J. Chem. Phys. 1968,48, 2265. (12) Hagstrum, H. D.; Becker, G. E. Phys. Rev. 1967, 159, 572.
