J . Phys. Chem. 1989, 93, 3223-3226
3223
some photoionization also occurs via an excited singlet state. In the case of blue-light irradiation, the excited-singlet pathway looks probable. It is also shown that the photoionization yield is controllable by changing the intersystem crossing rate. This may be a new and efficient way to control the charge-separation efficiency in vesicles and micelles.
crossing rate of Chl a was significantly enhanced by NaI addition by the external heavy-atom effect. As a result, the triplet yield and the photoionization yield increased. This gives evidence that the photoionization of Chl a in frozen EPC vesicles can go through the triplet state. This also suggests a new way to control the photoionization efficiency. Another photoionization pathway via an excited singlet state may also be possible.
Conclusions
Acknowledgment. This research was supported by the Division of Chemical Sciences, Office of Basic Energy Sciences, Office of Energy Research, U S . Department of Energy. Registry No. Chl a , 479-61-8; NaI, 7681-82-5.
The triplet state and the photoionization of Chl a in frozen EPC vesicles were observed by ESR in the presence of added K3Fe(CN),, PBQ, NaCI, NaI, and p-carotene. The intersystem
Electron Spin Resonance and Electron Spin Echo Modulation Studies of Ni' on Silica: Complexes with Hydrogen, Oxygen, and Ethylene W. Bogust and Larry Kevan* Department of Chemistry, University of Houston, Houston, Texas 77204-5641 (Received: May 20, 1988; In Final Form: September 12, 1988)
The electron spin resonance (ESR) signal Ni' produced by ultraviolet photoreduction in H, at -196 OC depends on whether the Ni+/Si02catalyst is activated prior to photoreduction near 650 OC or near 750 OC. ESR studies indicate the formation of one species, Ni+(A), at g l A = 2.66, g2* = 2.31, and ggA= 2.006 for activation at 673 OC and two species, Ni+(A) and Ni+(C), at g l c = 2.54, g2c = 2.31, and g3c = 2.036 for activation at 750 OC. When the H, is evacuated, a new Ni+(B) species is formed. After adsorption of O2a new Ni+(D) species is generated with glD= 2.25, gZD = 2.146, and ggD= 2.073 which is assigned to [Ni(O,)H-H]+. The inclusion of hydrogen in this complex is indicated by electron spin echo modulation (ESEM) data. Ethylene and trans-2-butene adsorption of Ni+/SiO, have also been studied, and the Ni+ complexes formed in each case appear to have quite different ESR spectra. ESEM data with deuterated ethylene indicate the formation of intermediate complexes of ethylene or butene with Ni+ on SiO, containing the equivalent of two or three ethylenes depending on the catalyst activation temperature.
Introduction
Nickel(1) ions formed by reduction of nickel(I1) can be stabilized in zeolite lattices and on aluminum oxide and silica surfaces.14 These Ni+ species form catalytically active sites for olefin oligomerization and acetylene cyclotrimerization Three different methods have been used to reduce NiZ+ions in these matrices: (a) thermal reduction in H 2 at 170-300 OC, (b) ultraviolet photoreduction in H z or C O at -196 OC, and (c) reduction by a hydrogen atom beam at 0 OC. When reduction of Ni2+ ions supported on silica is carried out by ultraviolet photoreduction, the number of Ni(1) species and the catalytic activity are several times higher than for thermal reduction." No metallic Ni formation is observed by ultraviolet photoreduction in contrast to thermal reduction. In preliminary work,' it was shown that the Ni+ species could be observed by electron spin resonance (ESR) spectroscopy and was a precursor of active intermediates for catalytic olefin dimerization. Subsequent work has abundantly confirmed However, the structure of the Ni+ complexes formed with olefins and detected by ESR remains unclear. In this work we use electron spin echo modulation (ESEM) spectroscopy to supplement ESR. By this technique it is possible to directly determine adsorbate coordination to a paramagnetic species such as Ni+. In past work in this laboratory we have attempted to study Ni+ produced by thermal reduction by ESEM methods.', Occasional ESEM signals were seen, but they were difficult to reproduce and no quantitative analysis was possible. This was thought to be due to varying amounts of cross-relaxation from Nio species also formed by the thermal reduction procedure. On leave from the Institute of Applied Radiation Chemistry, 93-590, L6dz,
Poland.
0022-3654/89/2093-3223$01.50/0
In this work, an ultraviolet photoreduction method at -196 OC is used to minimize possible complications from Nio formation. Experimental Section
The catalysts were prepared according to the Bonneviot procedure.I3 Silica gel (2 g) from Fisher Scientific Co. (grade 950, 60-200 mesh, 700 m2/g) was stirred at room temperature with 160 mL of an aqueous solution of ammonia at pH 11.6 for 24 h. The pH was maintained above 11.3. After filtration and washing with hot water, these samples were stirred for 2 days at room temperature with a solution containing 96 mL of H 2 0 , 3.5 mL of 0.1 M Ni(N03)2,0.14 g of N H 4 N 0 3 ,and 1 mL of 29% N H 4 0 H . This produced a Ni/SiOz catalyst with a maximum of 1 wt % Ni2+. The filtered and washed material was air-dried (1) Rabo, J. A.; Angell, C. L.; Kasai, P. H.; Shoemaker, V. Discuss. Faraday SOC.1966, 41, 328. (2) Garbowski, E.; Verdine, J. C. Chem. Phys. Lett. 1977, 48, 550. (3) Oliver. D.; Richard. M.; Che, M.; Bozou-Verdunaz, F.; Clarkson, R. B. J . Phys. Chem. 1980, 84,420. (4) Oliver, D.; Richard, M.; Che, M. Chem. Phys. Lett. 1978, 60, 77. (5) Elev, I. V.; Penshin, N.; Shelimov, B. N.; Kazansky, V. B. Kinet. Katal. 1982, 23, 936. (6) Garbowski, E. D.; Mauthieu, M. V.; Primet, M. Chem. Phys. Lett. 1977, 49,
247.
(7) Bonneviot, L.; Oliver, D.; Che, M.; Cottin, M. Proc. In?.ConJ Zeolites, 5th 1980, 108. (8) Bonneviot. L.: Che. M.: Dvrek. K.; Schollner. R.; Wendt, G. J . Phys. Chem. 1986, 90,2379. (9) Bonneviot, L.; Oliver, D.; Che, M. J . Mol. Catal. 1983, 21, 415. (10) Moller, B. W.; Kemball, C.; Leach, H. F. J . Chem. SOC.,Faraday Trans. 1 1983, 79, 453. (11) Kazansky, V. B.; Elev, I. V.; Shelimov, B. N. J . Mol. Catal. 1983, 21, 265. (12) Narayana, M.; Michalik, J.; Kevan, L., unpublished results. (13) Bonneviot, L.; Oliver, D.; Che, M. French Patent 8204888, 1982.
0 1989 American Chemical Society
3224
The Journal of Physical Chemistry, Vol. 93, No. 8, 1989 Ni+/Si02
\
Bogus and Kevan Ni+/Si02
0
x 50
x 50
-+
d
gy=2.07
Variation of ESR spectra at -196 OC of Ni/Si02 catalyst after (a) activation under vacuum to 673 OC,(b) adsorption of 350 Torr of H2 at 22 OC,(c) ultraviolet photoreduction at -196 "C for 2.25 h, (d) evacuation at 22 OC for 2.5 h, and (e) adsorption of 300 Torr of H2 at 22 o c . Figure 1.
for 1 day at room temperature. The activation procedure consisted of pretreatment in a flow of oxygen to 300 O C for 5 h and evacuation in an ESR tube at 300 O C for 15 h followed by heating for 2 h to 600-750 O C . The photoreduction was performed in the presence of 350 Torr of ultrahigh-purity H2 (Linde 99.999%) or D2 that was passed through a liquid nitrogen trap, using a high-pressure mercury lamp (Universal Light Source type BH-6) with a water filter. In order to prepare Ni+ complexes with various adsorbates, the reduced Ni/Si02 samples were evacuated at room temperature and then 02,C2H4, or trans-2-butene from Linde and CzD4 or D2 from MSD Isotopes were adsorbed at room temperature. ESR spectra were recorded at -196 OC or room temperature with a Varian E-4 spectrometer equipped with a Hewlett-Packard Model 5342A microwave frequency counter. ESE spectra were recorded at -269 O C with a home-built ~pectr0meter.I~Simul a t i o n ~of ' ~the normalized threepulse echo modulation were made as a function of an average number of nearest interacting deuterium nuclei, n, a t a distance r and with an isotropic hyperfine coupling uiw. The normalized modulation was multiplied by a polynomial decay function obtained from the experimental data to compare directly with experiment.
Results Figure 1 shows changes in the ESR spectra at -196 OC of Ni/Si02 catalyst after activation, reduction, and reversible adsorption of H2. The sample was activated by heating under vacuum up to 673 O C . A weak ferromagnetic ESR signal about 250 G wide at g = 2.2 is seen at both room temperature and -196 OC (Figure la). After adsorption of 350 Torr of H2 at room temperature, this ESR signal remained unchanged at room temperature but disappeared at -196 OC (Figure lb). Ultraviolet photoreduction in the presence of H2at -196 O C for 2.25 h resulted in the formation of a strong, anisotropic ESR ~ i g n a l ~ ~ *denoted '~.'' as signal A with orthorhombic g parameters of glA = 2.66, gZA = 2.31, and g3A= 2.006. Evacuation at room temperature changed the ESR signal at -196 "C into a very weak signal B (14) Ichikawa, T.; Kevan, L.; Narayana, P. A. J . Phys. Chem. 1979,83, 3378. ( 1 5 ) Ichikawa, T.; Kevan, L.; Bowman, K.; Dikanov, S. A,; Tsvetkov, Yu. D. J . Chem. Phys. 1979, 71, 1167. (16) Bonneviot, L.; Cai, F. X.;Che, M.; Kermerac, M.; Legendre, 0.; Lepetit, C.; Oliver, D. J . Phys. Chem. 1987, 91, 5912. (17) Michalik, J.; Narayana, M.; Kevan, L. J . Phys. Chem. 1984,88,5236.
y 93
gi.2 146
f g=:
2 073
Figure 2. ESR spectra at 22 OC (a, d) and at -196 O C (b, c, e, f) of Ni/SiOz catalyst after (a, b) activation under vacuum to 750 OC,(c, d) adsorption of 350 Torr of Hz,(e) photoreduction at -196 ' C and (f) subsequent evacuation at 22 OC for 25 h. Ni+/Si02 H--t
100 G
Figure 3. ESR spectra at -196 "C of Ni/Si02 catalyst activated to 600 "C after (a) photoreduction at -196 "C in 350 Torr of H2, (b) evacuation at 22 OC for 5 h, (c) adsorption of 40 Torr of 02,(d) readsorption of 350 Torr of H2,and (e) evacuation at 22 "C for 2 h.
with apparent axial g parameters of glls= 2.71 and gLB= 2.07 (Figure Id). Readsorption of H2 at -196 "C results in the reappearance of signal A (Figure le). ESR recorded at room temperature showed the same spectrum as in Figure l a for all sample treatments; thus, at room temperature the initial spectrum remained unchanged for the various sample treatments. Samples that were activated under vacuum up to 750 O C gave an ESR spectrum that had a more intense ferromagnetic signal than for activation at 673 O C . Adsorption of H2 at room temperature had little effect on this ferromagnetic signal (compare spectra a and d, Figure 2), but H2 adsorption at -196 "C caused the ferromagnetic signal to disappear (compare spectra b and c, Figure 2). After photoreduction in H2 at -196 "C species A is observed at the 673 O C activation temperature, but also a new species C appeared at giC = 2.54, gZc = 2.31, and g3c = 2.036. Signals A and C disappeared after 5 min of evacuation at room temperature. Signal B is probably formed but is obscured by a stronger signal D at room temperature with glD= 2.25, gZD= 2.146, and g3D= 2.073. Signal D persisted even after 25 h of evacuation. Signal D is due to O2contamination of the vacuum system as shown by separate experiments; signal D is formed by
The-Journal of Physical Chemistry, Vol. 93, No. 8, 1989 3225
Ni+ Complexes on Silica IO-
n = I, r = 032 nm I MHZ
5 oav,
a,,,=o
3
m
% 06>-
t (0
6 04z
I-
I
0
*
I
I.o
0.5
T, p
I.5
I
I
2.0
2.5
and simulated (- - -) three-pulse ESEM patterns at -269 OC of a Ni+/02(D2)complex on silica activated to 750 OC. The simulation parameters for the deuterium modulation are shown. The arrow on the inset ESR spectrum at -196 OC shows the field where the echo was obtained. Figure 4. Experimental (-)
"0
0.5
1.0
T ps
15
2.0
2.5
Figure 6. Experimental (-) and simulated (- - -) three-pulse ESEM patterns at -269 OC for Ni+/C2D4/Si02. The Ni/Si02 sample was activated to 673 "C, and C2H4was adsorbed at -78 "C. The simulation parameters for the deuterium modulation are shown. The inset ESR spectrum shows various g values and the field at which the echo was obtained.
Nit/Si02
n =8, r = 0.35 nm ais0=0.05MHz
ESE
1
n 1
A 2 47
=g:
2
4
7
w
t $I
I 98
Figure 5. ESR spectra of Ni+/Si02catalyst at -196 OC after (a) photoreduction in H2 at -196 OC, (b) subsequent evacuation of H2 and adsorption of 150 Torr of ethylene at -78 OC, and (c) subsequent warming to 0 OC for 30 s.
even trace amounts of 02.The adsorption of O2 on a sample previously activated at 600 "C, reduced, and evacuated also led to the formation of signal D (Figure 3c). The subsequent addition of 300 Torr of H 2 increased the intensity of signal D by 2-fold (Figure 3d). Evacuation of the sample for 2 h at rmm temperature reduced the intensity of signal D by 2-fold (Figure 3e; note the change in field sweep scale). Thus, the intensity enhancement due to the presence of H 2 seems reversible. A narrow ESR signal was usually observed near g = 2.002 for samples activated at temperatures no higher than 600 "C (see Figure 3a,b). N o ESE signals were observed for Ni+/Si02 species A, B, or C. But ESE signals were observed for Ni+ species D after ad2.08. When Ni' species D is generated sorption of O2 at g by photoreduction with D2 followed by adsorption of -40 Torr of 02,the ESE shows deuterium modulation indicating weak dipolar hyperfine interaction between Ni+ and deuterium (Figure 4). This modulation can be best simulated with one deuterium interacting at 0.32 nm with an isotropic hyperfine coupling (aiso) of 0.1 MHz. The adsorption of ethylene at -78 "C on a previously reduced and evacuated sample leads to the formation of the Ni+ ESR spectrum shown in Figure 5b. This spectrum seems to be composed of at least two signals. The spectrum near g = 2.059 is assigned to species F. On warming to 0 OC part of the spectrum near g = 2.059 in Figure 5b disappears, leaving the spectrum in Figure 5c which is assigned to species E. Only species E and not species F, as shown in Figure 5c, is formed when ethylene is adsorbed at room temperature instead of at -78 "C.
-
I
I
05
10
I
I
15
20
I
25
T, PS
and simulated (- - -) three-pulse ESEM patterns at -269 OC for Ni+/C2D4/Si02. The Ni/Si02 sample was activated to 750 OC, and C2H4was adsorbed at -78 "C. The simulation parameters for the deuterium modulation are shown. The inset ESR spectrum shows various g values and the field at which the echo was obtained. Figure 7. Experimental (-)
When trans-2-butene is adsorbed at room temperature on Ni+/Si02 in place of ethylene, an ESR spectrum similar to that of species F is found. This suggests that species F may be a Ni+(C4H8)complex. When only species E was present, no ESE signal could be seen. But when species F was present, an ESE signal could be seen at g 2.06. However, the ESE phase memory time, T,, of species F was quite different when species F was formed from ethylene (T, = 0.47 ps) or from trans-Zbutene (T, = 0.13 ws). This may indicate different locations for species F formed in these two ways. To try to determine the structure of the ethylene or butene complex with Ni+, ESE was carried out with deuterated ethylene and the deuterium modulation was analyzed. Although the ESR spectra of Ni+/Si02 with adsorbed C2D4 were the same for prior activation of Ni+/Si02 at 673 or 750 "C, the ESEM spectra are different. Figure 6 shows ESEM for a Ni+/SiOz sample activated at 673 "C. The best fit simulation indicates 12 deuteriums at 0.34 nm. However, in Figure 7 for sample activation at 750 "C the best fit simulation of the ESEM occurs for eight deuteriums at 0.35 nm. The distance is essentially the same for the two activation temperatures, but the number of deuteriums is decidedly less for the higher activation temperature. Table I summarizes the g values of the various Ni+ species observed on silica with our assignments together with prior results of Ni' on silica.
-
Discussion Activation of a Ni2+/SiO2 catalyst at 600-750 OC under vacuum results in a broad ESR signal at 77 K and room tem-
3226 The Journal of Physical Chemistry, Vol. 93, No. 8, 1989 TABLE I: ESR g Values at -196 Ni+ species/support Ni+(H2)/Si02
code' A
Ni+(H2)/Si02
A
Nit(H2)/Si02
C A
Nit/Si02
B
Ni+/Si02 [Nit(O2)H-HIt/SiO2
B D
N i '(0,)/ S O 2
D
Nit[C2H4]/Si02
E
Nit(C2H4)/Si02
E
Ni+(C2H4)/Si02 Nit[C4H8]/Si02
E
Nit[C4H8]/Si02
F
Nit(t-2-C4H8)/Si02 Ni+(t-3-C6H,,)/SiO,
F
F
Bogus and Kevan
OC for Supported Nit Species
formation conditions activated under vacuum to 673 OC, UV photoreduced in H2 at -196 OC activated under vacuum to 750 OC, UV photoreduced in H 2 at -196 OC same as above activation, UV photoreduced in H, at -196 "C, evacuated and H, adsorption activated under vacuum at either 673 or 750 OC, UV photoreduced in H2 at -196 OC, evacuated at room temperature reduction in H 2 at 170 "C, evacuated at room temperature activation under vacuum at 673 or 750 "C, UV photoreduction in H 2 at -196 OC, evacuation and adsorption of 0, at room temperature activation, UV photoreduction in H2 at -196 OC, evacuation and 0, adsorption activation under vacuum at 673 or 750 "C, UV photoreduction in H, at -196 O C , evacuation and adsorption of 0, at room temperature activation, UV photoreduction in H2 at -196 "C, and olefin adsorption at room temperature same as above activation, UV photoreduction, evacuation, and adsorption of C2H4at -78 OC same as above but with adsorption of trans-2-butene at room temperature same as above same as above with adsorption of hexene at room temperature
2.66
2.3 1
g3 ref 2.006 this work
2.66
2.3 1
2.007 this work
2.54 2.64
2.31 2.30
2.036 2.015 11
gl
= 2.07
g2
= 2.71
this work
= 2.097 gll = 2.75 2.25 2.146
10 2.073 this work
g,
gli
g,
2.254
2.134
g, = 2.47
gll
= 1.98
g, = 2.469 gli = 1.975 g, = 2.45
2.059
2.037 11 this work 9
= 1.581 11 2.15 1 .99b this work
gil
g, = 2.058 gll = 2.15
this work
g, = 2.057 gli = 2.153
9 9
g, = 2.60
gll
= 2.013
'Our species designation. bThis g value may not go with the others in this row.
perature at g = 2.2 which has been assigned to Nio clusters.16 After H 2 adsorption at room temperature the ESR signal of the Nio clusters at room temperature is unaffected. However, if the sample temperature with adsorbed H2 is lowered to -196 OC, the ESR signal disappears. This may be due to stronger chemisorption of H 2 on the Nio clusters at the lower temperature. Subsequent ultraviolet photoreduction of Ni2+/Si02 in H2 at -196 OC generates an ESR signal A which has been assigned to [Ni(H2)J+ or [Ni(H2)]+complexes.Is This ESR signal is observed only at -196 "C in the presence of H2. This complex dissociates at room temperature. The new species C observed with species A in samples activated to 750 OC is also assigned to a Ni+(H2),, complex with n unspecified. A weak signal B at gllB= 2.71 and gLB = 2.07 is observed at -196 O C after evacuation of H2 from a reduced sample and has been assigned previously to a [Ni+(H2),] complex.'* We suggest that signal B may equally well be assigned to Ni+ on silica without coordination to molecular hydrogen. Signal D is assigned to an oxygen complex of Ni+. Its g factors agree reasonably well with a Ni+(02) complex assigned previously" (see Table I). The reason for its reversible intensity enhancement in the presence of H 2 is unclear. However, it is analogous to the intensity changes of signal A in the presence and absence of H2. It is also unclear why paramagnetic O2does not broaden the Ni+ signal more. ESEM results confirm the direct interaction of species D with only one atom of hydrogen where the hydrogen interaction arises in the photoreduction process. This suggests an end-on conformation with a hydrogen molecule. The other hydrogen atom of the H 2 molecule would then be too far from Ni+ to be detected by ESEM. The ESR spectra for N i + samples with adsorbed ethylene as shown in Figure 5 are somewhat similar to those described by Bonneviot et aL9 as shown in Table I. Species E and F are assigned in Table I to ethylene and butene complexes of Ni+, respectively. Bonneviot et observed only the ethylene complex, species E, 1 min after ethylene adsorption at room temperature. At longer times, ESR lines assigned to butene and hexene complexes were observed; their butene complex assignment is consistent with ours. (18) Bonneviot, L.; Oliver, D.; Che, M. J . Chem. SOC.,Chem. Commun. 1982, 952.
The large difference in the g factors for species E assigned to Ni+(C2H4)and species F assigned to Ni+(C4Hs) is puzzling and is not yet understood. At this point it can be said that the chemistry supports the assignments of species E and F. The ESEM data on the Ni+-olefin complexes demonstrate that Ni+ is coordinated to ethylene or to trans-2-butene as a reaction product. For sample activation at 750 OC the ESEM data indicating eight interacting deuteriums are most reasonably interpreted as one molecule of product 2-butene coordinated to Ni+. However, for sample activation at 673 " C 12 coordinated deuteriums are indicated which can be interpreted as coordination of one molecule of 2-butene and one molecule of ethylene to Ni+. The difference with sample activation temperature may be connected with previously described'"22 changes in the silica surface structure during evacuation in this temperature range. Silica heated under vacuum above 663 OC loses hydroxyl groups2' and reduces its effective surface area by sintering. These changes may affect the bonding of Ni+ to the silica surface and its coordination ability for olefins. The exact nature of this is unknown. Conclusions
Species Ni+(A) and Ni+(C) are formed in a silica surface depending on the temperature of activation of the Ni/Si02 catalyst. the adsorption of H2 on species A and C is fully reversible with temperature. During the initial stages of the Ni+-catalyzed dimerization of ethylene, intermediate complexes of ethylene or butene with Ni+ on S i 0 2 are formed involving two or three molecules of ethylene depending on the prior activation temperature. Acknowledgment. This research was supported by the National Science Foundation, the Robert A. Welch Foundation, and the Texas Advanced Research Technology Program. We thank Dr. M. Colaneri for his help in recording the ESE signals. Registry No. Ni, 7440-02-0; H2, 1333-74-0; O,, 7782-44-7; C2H4, 74-85-1; CdHg, 624-64-6. (19) Young, G. J. J . Colloid Sci. 1958, 13, 67. (20) Fripiat, J. J.; Uytterhoeven, J. J . Phys. Chem. 1962, 66, 800. (21) Anderson, J. R. Structure of Metallic Catalysts; Academic Press: New York, 1975. (22) Peri, J. B.; Hensley, A. L. J . Phys. Chem. 1968, 72, 2926.