J. Phys. Chem. 1989, 93, 1184-1187
1184 c- taken from Irish et aL3, x
3. The present analysis supports this approach as illustrated in Figure 1 where the fig from the neutron diffraction experiments are compared with (4) which was evaluated using several different models for c-. The three sets of Ki used to obtain the curves based on the present analysis are given, on the molarity scale, by K1 = K2 = 0.32, K3 = 10, K4 = 0.1 (set A); K, = 0.48, K2 = 1.86, K3 = 0.56, K4 = 1.41 (set B); and K1 = 2.0, K2 = 1.0, K3 = 10, K4 = 0.1 (set C). Set A has previously been used to interpret 67ZnN M R chemical shift data,I7 set B has been compared with dielectric spectroscopy data with some measure of success,8 and set C has been recommended by Smith and MartellI5 for ZnC12 solutions of ionic strength Z = 4. The R value for each Ki set used in (4) was fixed at the corresponding (2)value given in Table 11. The agreement between the measured and calculated fig is much better when c- is taken (17) Marciel, G. E.; Simeral, L.; Ackerman, J. J. H. J. Phys. Chem. 1977, 81, 263.
from the present model than when, like Powell et aI.,l c- is taken from Figure 9b in the paper of Irish et aL3 The agreement is, however, improved by taking the tabulated values given by Irish et aL3 (see Figure 1). The implication of 3 values that are independent of concentration and equal to a constant value (x) is that the individual xi (i = 1 , 2 , 3 , 4 ) are all equal to the same constant. This follows from (10) since IC1 # 0 for any set of Ki used in the present work and G C i j = 1. Thus the mean number of water molecules displaced from free chloride ions on complexing with Zn2+ is, within the experimental uncertainty on the neutron diffraction data, independent of the type of complex species. It would be of interest to test this result by repeating the neutron diffraction experiments at the same values of c but in solutions of constant ionic strength.
Acknowledgment. I thank Hugh Powell and Roderick Cannon for several useful discussions. The financial assistance of the UK Science and Engineering Research Council is gratefully acknowledged.
First Observation of Carbon Aggregate Ions >C,,,+ Transform Mass Spectrometry
by Laser Desorption Fourier
Hun Young So and Charles L. Wilkins* Department of Chemistry, University of California, Riverside, Riverside, California 92521 (Received: September 22, 1988)
Pulsed carbon dioxide laser desorption Fourier transform mass spectra (LD/FTMS) of benzene soot samples, deposited upon a stainless steel probe tip and a potassium chloride coated stainless steel probe tip, are reported. Under both conditions, polyaromatic hydrocarbon ions with m / z below 1000 and a series of even-numbered carbon aggregate positive ions extending above m / z 7200 are observed. Enhanced relative abundance of Cso+is found for soot deposited upon stainless steel, but not on KC1-coated stainless steel. In the high mass region above m / z 1O00, it appears that two distinct distributions of carbon aggregate positive ions can be detected. One distribution appears to be centered around c154+and a second, much broader, distribution extends from ca. C300+to above Cm+.
Formation of neutral and ionic aggregates exclusively containing carbon, via laser ablation processes from graphite,’-I0 has inspired a good deal of dialogue regarding the structures of these species. Because no direct structural information is available, various structural proposals including both spherical3 and linear crosslinked polymeric speciesI0 are rationalized in terms of mass spectrometric ion abundance measurements. One particular aggregate, Ca+, and its metal-attach,ed derivatives have been the subject of much discussion, primarily because of Smalley’s im(1) Rohlfing, E. A.; Cox, D. M.; Kaldor, A. J. J. Chem. Phys. 1984,81, 3322-3330. (2) Heath, J. R.; O’Brien, S.C.; Zhang, Q.; Liu, Y.;Curl,R. F.; Kroto, H. W.; Tittel, F. K.; Smalley, R. E. J. Am. Chem. Soc. 1985.107, 7719-7780. (3) Kroto, H. W.; Heath, J. R.; O’Brien, S.C.; Curl, R. F.; Smalley, R. E. Nature 1985, 162-163. (4) Bloomfield, L. A.; Geusic, M. E.; Freeman, R. R.; Brown, W. L. Chem. Phys. Lett. 1985, 121, 33-37. ( 5 ) OKeefe, A.; Ross, M. M.; Baronavski, A. P. Chem. Phys. Lett. 1986, 130, 17-19. (6) OBrien, S.C.;Heath, J. R.; Kroto, H. W.; Curl, R. F.; Smalley, R. E. Chem. Phys. Lett. 1986, 132,99-102. (7) Hahn, M. Y.; Honea, E. C.; Paguia, A. J.; Schriver, K. E.; Camarena, A. M.; Whetten, R. L. Chem. Phys. Leu. 1986, 130, 12-16. (8) McElvaney, S.W.; Nelson, H. H.; Baronavski, A. P.; Watson, C. H.; Eyler, J. R. Chem. Phys. Lett. 1987, 134, 214-219. (9) Knight, R. D.; Walch, R. A.; Foster, S.C.; Miller, T. A.; Mullen, S. L.; Marshall, A. G. Chem. Phys. Lett. 1986, 129, 331-335. (10) Cox,D.M.; Reichmann, K. C.; Kaldor, A. J. Chem. Phys. 1988,88, 1588-1597.
0022-3654/89/2093-1184$01.50/0
aginative nomenclature (which describes the parent species as “buckminsterfullerene” or “ b ~ c k y b a l l ” and ) ~ ~the ~ ~ ingenious experimental evidence he and his co-workers adduce for their thesis regarding its special geometry and ~ t a b i l i t y . ~ - ’However, ~ ~ ’ ~ it is by no means established that the claimed special stability exi s t ~ ~ , ’or, ~ ,indeed, ’~ that the stability, if it does exist, cannot be explained equally well by structural alternatives, including those advanced by Kaldorlvloand others.15 Ultraviolet lasers were used in most prior studies of carbon aggregate formation, both in the aggregate formation step and to photoionize the resulting species for subsequent mass spectral analysis. Here, we report a number of interesting new observations regarding direct observations of carbon aggregates formed from benzene soot, using a pulsed infrared (C02) laser, with Fourier transform mass spectrometric detection (LD/FTMS).16 Although these new data do not resolve the structural questions, they may assist in evaluating the validity (11) O’Brien, S. C.; Heath, J. R.; Curl,R. F.; Smalley, R. E. J. Chem. Phys. 1988,88, 220-230. (12) Weiss, F. D.; Elkind, J. L.; O’Brien, S. C.; Curl, R. F.; Smalley, R. E. J. Am. Chem. SOC.1988, 110, 4464-4465. (13) O’Brien, S. C.; Heath, J. R.;Kroto, H. W.; Curl, R. F.; Smalley, R. E. Chem. Phys. Lett. 1986, 132, 99-102. (14) Cox,D. M.; Trevor, D. J.; Reichmann, K. C.; Kaldor, A. J. Am. Chem. SOC.1986, 108, 2457-2458. ( 1 5 ) Frenklach, M.; Ebert, L. B. J. Phys. Chem. 1988, 92, 561-563. (16) Ijames, C. F.; Wilkins, C. L. J. Am. Chem. SOC.1988, 110, 2681-2688.
0 1989 American Chemical Society
Letters
The Journal of Physical Chemistry. Vol. 93. No. 4, 1989 1185
350
400
700
750
450
500
550
600
650
700
850
900
950
1000
1050
1800 2000 2200 2400 A.M.U. Figure 1. L a w de3orption Fourier transform mars ,pcctrum ofsaat dcpmitcd an B stainless sieel probe tip. Thirty timc-domain rpcctra. each of a
1200
fresh sample. wcrc co-added. A 1051.
1400
15-3
16b0 MASS
IN
delay follouing laser desorption uas used prior to excitation using a I-MH, bandwidth (lower mars cutoff m / z
of the various structural proposals extant Experimental Section Experiments were carried out using a dual-cell differentially pumped 7-T Nicolet FTMS-2000 mass spectrometer. A Tachisto 215G pulsed CO, laser, providing between IO6 and IO8 Wlcm' per pulse (40-100 ns), was used for laser desorption. The CO, laser was directed through a window in the source flange and focused upon the tip of a computer-controlled solids probe by an off-axis paraboloid mirror mounted on the source cell assembly. Soot samples were prepared by deposition upon stainless steel probe tips positioned abwe a burner. For all spectra, source cell detection was employed after trapping laser-desorbed ions using a 1.6-V trap voltage for between 5 and 15 s, following laser desorption, Excitation was by a 200-V, sweep at a rate of 300 Hzlps, using a I-mHz bandwidth with a lower frequency corresponding to m / r 105 for the spectra in Figures 1 and 2 and a 500-kHz bandwidth after ejection of ions with frequencies up to 96.8 kHz ( m / z 1087) for the spectrum in Figure 3. Twenty to thirty spectra were co-added to obtain the spectra shown in Figures 1-3. Discussion Although others have observed formation of large carbon aggregates by infrared laser desorption of aromatic and heteroaromatic polymers, no enhanced abundance of Cm+, was found1' Both negative and positive ion species with even numbers of carbons in 24-Da increments, ranging from 40 to 120 carbons (negative ions) and to 400carbons (positive ions) are formed from both polymer and asphaltene samples;ls because hydrogen is (17) Brown, C. E.: Kovacic, P.; Welch, K. J.: Cody, R. B.: Hein. R. E.; Kinsmger, J. A. 1.Polymn. ScC Pan A Polym. Chem. 1988,26. 131-148. (18) Nuwaysir, L. M.: Wilkins, C. L. Unpublished research, 1988.
present in the starting materials, significant numbers of polyaromatic hydrocarbons are also formed in the process. Furthermore, in positive ion spectra, the Cbospecies is significantly enhanced for asphaltene spectra and even-numbered carbon aggregate ions containing as many as 500 carbons are observed.l8 In the present study, soot, formed by combustion of benzene, was analyzed by LD/FTMS. Following its deposition upon either a stainless steel probe tip or on a potassium chloride coated stainless steel probe tip, the soot was analyzed, with the results shown in Figures 1-3. Several significant observations can be made. First, two distinct categories of ions are present: ( I ) PAH ions, together with carbon aggregate ions, with m / z below 1000 are readily identified; (2) even-numbered carbon aggregate ions, separated by 24-Da intervals, extending from Ca to above C, are evident, although significant abundances of C, or larger ions are observed only when laser desorption from the KCI-coated probe tip is employed. Unambiguous confirmation of identities of PAH ions was provided by use of C6D6to produce the soot, with the expected isotopic shifts for the hydrogen-containing ions. Resolution is sufficiently good that comparisons of experimental and theoretical isotopic distributions of the carbon aggregates up to Cia is possible; those comparisons for the even-numbered carbon species C,, through C,,, unequivocally confirm the absence of hydrogen. A second significant observation is that, in the mass range above m / z 1000, two distinct distributions of even-numbered carbon aggregates are apparent. There is one distribution peaking a t about C,,,+ and a second, apparently incomplete, distribution beginning at about m / z 3000 and extending beyond m / z 7200. Both resemble the types of distribution expected for polymers analyzed by LD/FTMSi6.'9.m and suggest the possibility that two (19) Brown,R. S.: Weil, D. A,: Wilkins, C. L. Moeromolenrles 1986.19, 1255-1260.
1186 The Journal of Physical Chemistry, Val. 93, No. 4, 1989
350
400
450
7d0
750
800
,
Letters
500
550
600
650
700
850
900
940
1000
1050
1600 1800 2000 2200 2400 MASS I N A . M . U . Figure 1. Laser desorption Fourier transform mass spectrum of mot depasited on a stainless steel probe tip previously coated with a layer of KCl by application of a solution of KCI in methanol and evaporation of the methanol. Spectral measurement conditions were the same as for Figure 1. 1400
t ‘154
1
I
I
I
I
1000
2000
3000
4000
I
5000
60bO
70b0
MASS I N A . M . U . Figure 3. L a x r desorption Fourier transfarm mass SpKlNm of soot depo,ited on a KCI-coated stainless steel probe tip. After a 5 4 delay. following laser desorption. ions Hith m / z below 1087 were ejected. and a 500-kIlz bandwidth excitalion was employed. Twenty time-domain spectra. each of a fresh sample. were madded. Inset Thaws base-line noise. in the absence of laser dcsorption. distinct types of carbon aggregates may
he present.
Interestingly,
fhe relafiue ahundances of many of rhe carbon aggregafe ions,
including all fhme berween C I Mand C,80.are greafer than rhar of Cm+. This can bc s u n by comparing the relative abundance
J. Phys. Chem. 1989, 93, 1187-1189 of the CIs4+ion with that of Cm+ in Figure 2 and the abundance of the CIM+ion with those of the other ions in Figure 3. Although lower mass ions were ejected in order to allow enhanced mass resolution measurement of the Figure 3 spectrum, measurements under the same conditions, without ejection, showed that the C154/m ion ratio is similar to that in the Figure 2 spectrum. In view of these results, under conditions which are not expected to discriminate against detection of any ionic species (other than those specifically excluded), it seems likely that relative abundances are governed by kinetic effects in the soot formation process, rather than any “magic” stability of particular carbon aggregates. Finally, in accord with previous observations that mass spectral abundances of carbon aggregate ions are critically dependent upon formation conditions,I4 Cm+appears with enhanced abundance when desorption from uncoated stainless steel is employed (Figure (20) Nuwaysir, L. M.; Wilkins, C. L. Anal. Chem. 1988, 60, 279-282.
1187
1, but its relative abundance drops appreciably and those of aggregates containing more than 100 carbons are enhanced, when the soot is desorbed from the potassium chloride coated probe tip (Figures 2 and 3). This suggests factors other than the thermodynamic stability of Cs0+ are critical in determining its abundance (e.g., KC1 assisted nucleation of carbon aggregates during s&t formation in the low-temperature flame). Unfortunately, in common with all previous studies, the present observations neither refute nor confirm any particular structure hypothesis. However, comparison of our observations with those results, obtained under significantly different experimental constraints, could allow interpretation of the previous data with increased precision.
Acknowledgment. This work was supported by N I H Grant GM-30604 and by a contribution from the Shell Development Company, both of which are gratefully acknowledged.
Preparation of Highly Dispersed Anchored Vanadium Oxides by Photochemical Vapor Deposition Method and Their Photocatalytic Activity for Isomerization of trans-2-Butene Masakazu Anpo,* Masatoshi Sunamoto, Department of Applied Chemistry, College of Engineering, University of Osaka Prefecture, Sakai, Osaka 591, Japan
and Michel Che Laboratoire de RZactivitZ de Surface et Structure, UniversitZ P. et M. Curie, UA 1106-CNRS, 4 Place Jussieu, Tour 54, 75252 Paris Cedex 05, France (Received: October 13, 1988)
Highly dispersed anchored vanadium oxide catalysts have been first prepared by using facile reaction of the photochemically activated VOClo molecules with surface OH groups on Vycor glass. Anchored vanadium oxides prepared by photochemical method (photo-CVD) exhibit much higher yields in the phosphorescence with its long lifetimes (450 rs). These anchored vanadium oxides exhibit a higher activity for the photocatalytic isomerization reactions of 2-butene, as compared with those obtained for the oxides prepared by conventional impregnation method. A good linear relationship between the yields of photocatalytic isomerization and the yields of phosphorescence from the charge-transfer excited triplet state of the surface vanadyl species in tetrahedral coordination is obtained, indicating that not only the surface vanadyl species play a significant role in the photocatalytic reaction on vanadium oxide catalysts, but also a photochemical vapor deposition method leads to the higher dispersion of vanadium than impregnation method.
Introduction For the complete understanding of photocatalysis, it seems inevitable to elucidate the chemical nature and reactivity of photoproduced electrons and holes, as well as the excited state of catalysts, including dynamics.’ With well-defined chemically supported “anchored” catalysts, one can expect not only to achieve more active and selective catalytic and photocatalytic systems but also to obtain the detailed information on the true nature of active surface sites. Chemical vapor deposition (CVD) method by using various reactive transition metal compounds has been extensively studied to prepare the anchored catalystse2 Photochemical activation of various transition-metal compounds adsorbed on support surfaces is also expected to be a promising method to prepare highly dispersed well-defined catalysts; however, there are rather few
In this paper we report that not only the advantage of the method of the photochemical activation of adsorbed VOC13 molecules to prepare the highly dispersed anchored vanadium oxides but also the significant role of the charge-transfer excited triplet state of vanadyl groups in the photocatalytic isomerization reaction of 2-butene on supported vanadium oxide catalysts. Experimental Section Preparation of Catalysts by Photochemical Vapor Deposition Method. Anchoring of VOC13 onto porous Vycor glass (Corning code, 7930; BET surface area, 155 m2/g; major composition, SOl = 96% and B203 + 3%) was performed using a facile reaction of photoactivated VOC13 with surface O H groups of the dehydrated Vycor glass specimen at 273 K. Photoactivation of adsorbed V0Cl3 molecules was carried out by UV excitation of the
report^.^ (1) Anpo, M.; Kubokawa, Y. Reu. Chem. Intermed. 1987, 5 , 105. (2) Iwasawa, Y . Adu. Catal. 1987, 35, 187, and references therein.
0022-3654/89/2093-1187$01.50/0
(3) Gafney, H. D. In Photochemistry on Solid Surfaces; Anpo, M., Matsuura, T., Eds.; Elsevier: Amsterdam, 1989; p 272. Wada, Y.; Nakaoka, C.; Morikawa, A. Chem. Lett. 1988.25. h i , Y . ;Yano, K. Inorg. Chim. Acta 1983, 76, 17 1 .
0 1989 American Chemical Society
