Emission Spectroscopic Investigation of Triplet Diarylcarbene

Oct 8, 2005 - Chemistry Department and Advanced Engineering Courses, Gunma College of Technology, Maebashi-shi, Gunma 371-8530, Japan. Wolfgang ...
0 downloads 0 Views 169KB Size
Download PDF
J. Phys. Chem. B 2005, 109, 20407-20414

20407

Emission Spectroscopic Investigation of Triplet Diarylcarbene Generated in Molecular Sieve VPI-5 Shuichi Hashimoto* Department of Ecosystem Engineering, Graduate School of Engineering, The UniVersity of Tokushima, Tokushima-shi, Tokushima 770-8506, Japan

Masashi Saitoh and Nobuyuki Taira Chemistry Department and AdVanced Engineering Courses, Gunma College of Technology, Maebashi-shi, Gunma 371-8530, Japan

Wolfgang Schmidt Max-Plank-Institute fu¨r Kohlenforschung, Kaiser-Wilhelm-Platz, 1 Mu¨lheim an der Ruhr D-45470, Germany

Katsuyuki Hirai Instrumental Analysis Facilities, Life Science Research Center, Mie UniVersity, Tsu-shi, Mie 514-8507, Japan

Hideo Tomioka* Chemistry Department for Materials, Faculty of Engineering, Mie UniVersity, Tsu-shi, Mie 514-8507,and Department of Applied Chemistry, Aichi Institute of Technology, Toyota-shi, Aichi 470-0392 Japan ReceiVed: August 30, 2005

We report an attempt to generate and characterize a triplet carbene, bis(2,4,6-trichlorophenyl)carbene (31), in zeolites Y and L and in a molecular sieve VPI-5 in which a possible dimerization and reaction with the precursor of carbene are significantly retarded, thus making the triplet carbene longer lived than in solution at room temperature. The adsorption of a corresponding diazomethane (1-N2), the precursor of 1, was carefully examined by comparing the absorption spectrum after adsorption with that of 1-N2 in n-pentane, which revealed that 1-N2 was adsorbed with the diazo group intact only in VPI-5, while in other zeolites 1-N2 was found to be decomposed upon adsorption. This difference in reactivity of the hosts was ascribed to the absence of Brønsted-acid sites in VPI-5. The photoirradiation of 1-N2 in VPI-5 at 77 K was monitored by emission spectroscopy, which revealed that bis(2,4,6-trichlorophenyl)methyl radical (1-H) was produced as the only detectable species under these conditions. This is interpreted as indicating that nascent 31 may undergo efficient hydrogen abstraction as a result of multiple excitation by repeated refraction and reflection of the light in a light-scattering medium. In accord with this interpretation, the emission due to 31 was observed when irradiation was carried out on a translucent glassy sample prepared by submerging VPI-5 incorporating 1-N2 in a refractiveindex-matching fluid such as propylene glycol or glycerol. ESR signals ascribable to 31 were also observed under these conditions. Laser photolysis of 1-N2 in VPI-5 at room temperature with fast detection of both emission and absorption showed that the bands due to 1-H were detected in the nanosecond time regime probably because of the extremely fast H abstraction by 31. However, a variable-temperature ESR study showed that the signals due to 31 survive up to 220 K in VPI-5 while the signals disappear at 120 K in 2-methyltetrahydrofuran, suggesting that triplet carbene is stabilized in VPI-5. Thus, a triplet carbene was generated and characterized in a zeolite for the first time and shown to be stabilized extrinsically. The present study also proposes a solution to the issues of acidic sites and multiple excitation often observed in zeolites.

Introduction For a long time carbenes were believed to be too unstable to be isolated in macroscopic amounts at room temperature.1 However, the recent syntheses of stable singlet carbenes, i.e., phosphinocarbenes2 and imidazol-2-ylidenes,3 have upset this extreme view. These carbenes are obviously thermodynamically stabilized by the π-donating ability of the heteroatom substituent directly attached to the carbenic carbon.4 Stabilization of carbenes with another electronic state, a triplet state, then

emerges as a challenging target. Attempts have been made to stabilize triplet carbenes, which revealed that they are more difficult to obtain since triplet carbenes are less susceptible to conjugative stabilization.5 Steric protection (or kinetic stabilization) together with electron delocalization led to a fairly stable triplet carbene,6 which was still not completely stable. Another method to stabilize carbenes at ambient temperature is to encapsulate them into a suitable host.7 This is termed as extrinsic stabilization as opposed to intrinsic stabilization mentioned above. In this approach the precursor of the carbene

10.1021/jp0548974 CCC: $30.25 © 2005 American Chemical Society Published on Web 10/08/2005

20408 J. Phys. Chem. B, Vol. 109, No. 43, 2005 is introduced into the host and the host-guest complex is then photoirradiated. The carbenes thus generated are protected from dimerization, from reaction with the precursor, and from reaction with solvent molecules that are too large to enter the inner phase. The extrinsic approach has been applied to stabilize singlet carbenes and most recently been shown to be very successful in isolating the singlet state. Thus, fluoro(phenoxy)carbene possessing a singlet ground state generated in a hemicarcerand host was shown to persist for days at 25 °C.8 On the other hand, phenylcarbene, a triplet ground state carbene, generated under essentially identical conditions, has undergone insertion into the C-H bonds and aryl groups of the host carcerands.9 This is somewhat expected since triplet carbenes are known to be efficient acceptors of a hydrogen atom.10 These observations clearly indicate that the extrinsic approach is also a very attractive way to stabilize triplet carbenes and that a host without a hydrogen source is probably preferable especially for stabilizing triplet carbenes. Inorganic hosts such as zeolites are especially intriguing in this respect. Thus, we decided to generate and study triplet carbenes in the cavity of a zeolite.11 Ample studies have been carried out on the generation and confinement of reactive intermediates in the cavities and channels of zeolites because the zeolites often allow guest species to have remarkably long lifetimes which normal solution cannot offer at room temperature. For instance, the strongest reducing agents, electrons, are known to be trapped inside the sodalite cages of zeolites A, X, and Y in the form of sodium ion clusters such as Na43+.12 The species is stable and survives even at room temperature in dehydrated zeolites. Stabilization at room temperature was also found for electronically excited states, radicals, and radical ions.13-15 For instance, roomtemperature phosphorescence with remarkable lifetimes similar to that at 77 K was observed for aromatic species such as naphthalene and 9-ethylcarbazole confined in a channel-type zeolite L.13 Presumably the rigid zeolite medium restricts vibrational and rotational motion of the tightly fitted guest species, leading to a fast intramolecular relaxation process. Spontaneous radical cation formation in acidic zeolites and photochemical generation of radical cations in zeolites with an electron-accepting nature as well as that of radical anions in zeolites with an electron-donating nature have been investigated.13,14 It was found that the rigid framework of the zeolite hosts serves as an excellent matrix, stabilizing otherwise reactive or unstable radical ions presumably by immobilization in the inert container, which also protects these species from external attacks. Furthermore, radical persistence was demonstrated for radicals such as benzyl and diphenylmethyl generated in zeolite MFI (ZSM-5).15 Here we report the first generation, observation, and characterization of relatively persistent triplet diarylcarbene in the cavity of a zeolite. We note that a less reactive singlet carbene, chloro(phenyl)carbene, has been studied recently by a diffuse reflectance laser photolysis technique and found to have a lifetime of less than a few microseconds within the cavities of potassium-exchanged zeolite Y (K+-Y).11a Experimental Section The precursor of triplet carbene, bis(2,4,6-trichlorophenyl)diazomethane (1-N2), was prepared by the method described in the literature.16 Bis(2,4,6-trichlorophenyl)methyl chloride (1HCl) was also prepared for the control experiment.16 The sodium form of zeolite Y (Na+-Y; Si/Al ) 2.75, Na/Al ) 1.0, Fe 120 ppm by ICP) and the potassium form of zeolite L (K+-L; Si/

Hashimoto et al. Al ) 3.0, K/Al ) 0.98, Na/Al ) 0.02, Fe 100 ppm by ICP) were provided by Nanyo Research Laboratory, Tosoh. Zeolites Na+-Y and K+-L were calcined in air at 450 °C for 8 h just before sample preparation. Molecular sieve VPI-5 was synthesized according to the literature method and characterized by powder X-ray diffraction.17 As-synthesized VPI-5 was calcined in flowing O2 at 600 °C for 12 h to remove the organic structuredirecting agent. As-synthesized VPI-5 was first evacuated at room temperature for 2 days before gradually heating to 300 °C under vacuum (10-2 Pa). The exhaustive dehydration at room temperature is necessary to minimize the phase transformation of VPI-5 to AlPO4-8 upon heating over 60 °C.18 The calcined VPI-5 was stored in an evacuated container before use. Complete isolation from atmospheric moisture as well as from oxygen is important for the preparation of 1-N2-zeolite complexes. In addition, the thermal decomposition of 1-N2 during the adsorption process should be avoided. Thus, intercalation of the precursor molecule 1-N2 into the zeolite hosts was carried out by a solution method at ambient temperature in a container attached to a vacuum line and equipped with an optical or ESR cell. Briefly, the dehydrated zeolite powder was stirred in a n-heptane or benzene solution of 1-N2 for 1 h, and the solid was dried by evacuation treatment (1-12 h). The control experiments in a glovebox charged with N2 confirmed by spectrophotometric assay of supernatant solution that over 95% of the guest precursor molecules are adsorbed in the hosts. Index-matching fluids were introduced to the zeolite samples in vacuo. The ground-state reflectance spectra were collected on a Shimazu UV-3101PC double-monochromator spectrophotometer equipped with an integrating sphere coated with BaSO4. The reference was BaSO4 (Kodak white reflectance standard). Absorption spectra of ground-state species were obtained using the Kubelka-Munk function. Fluorescence spectra were recorded on a Hitachi F-3010 spectrofluorometer in a front-face arrangement. The spectral response of the fluorometer was not corrected. ESR spectra were recorded on a JEOL JES-TE200D ESR spectrometer. Steady-state photolysis was carried out at liquid nitrogen temperature for the samples immersed in a Dewar flask with quartz windows by a 500 W ultra-high-pressure mercury arc (Ushio USH-500) with a combination of appropriate filters. Transient absorption and emission spectra excited at 266 nm were measured on a diffuse reflectance laser photolysis setup described previously at 23-25 °C.19 Results and Discussion Adsorption of Precursor Molecules in Zeolites. To generate and observe a triplet carbene in a zeolite host one needs to encapsulate a precursor molecule in the host. A diazo compound is the most efficient and clean and hence the most frequently employed precursor for carbene formation since it releases molecular nitrogen upon irradiation while leaving the carbene species. This reaction potentially takes place regardless of the reaction phase at any temperature. Thus, we chose bis(2,4,6trichlorophenyl)diazomethane (1-N2), which is shown to generate a fairly persistent triplet state bis(2,4,6-trichlorophenyl)carbene (31) upon UV irradiation.16 Carbenes that are generated from diazo compounds have not been studied in zeolite, possibly for the following reasons. First, diazo functional groups are not very stable and usually sensitive not only to heat and light but also to acid and metals.20 Second (and coincidentally), zeolites inevitably have Brønsted-acid sites even though an intentional proton exchange is not carried out.21

Spectroscopic Investigation of Triplet Diarylcarbene

J. Phys. Chem. B, Vol. 109, No. 43, 2005 20409 is remarkably distorted both in Na+-Y and K+-L. The result clearly suggests that the precursor 1-N2 is adsorbed in VPI-5 without appreciable decomposition while it may have undergone decomposition when adsorbed in Na+-Y and K+-L. This difference can be interpreted in terms of the chemical structure of the zeolite frameworks. The aluminosilicate zeolites, Na+-Y and K+-L, carry cations that are susceptible to exchange with protons. Thus, 1-N2 is relatively easily decomposed by the Brønsted-acid site present on the surface of the aluminosilicate zeolites, most probably giving rise to a carbocation (1H+).20 The absorption spectra of bis(2,4,6-trichlorophenyl)methyl alcohol (1H-OH) in Na+-Y is reminiscent of that of 1HOH, which can be formed from 1H+ according to the scheme

Here we chose three types of zeolites Na+-Y, K+-L, and VPI-5 to determine which can include 1-N2 without decomposition. Na+-Y is a FAU structure possessing nearly spherical supercages with a diameter of 1.3 nm interconnected tetrahedrally through four 0.74 nm windows and employed most frequently for the adsorption of large aromatic species.21 K+-L consisting of a LTL structure is a channel-type zeolite with main channels with 0.71 nm openings.21 These two types are made of aluminosilicate frameworks whose minus charge must be balanced by charge-compensating cations. The chemical property of K+-L is considered to be less acidic than that of Na+-Y because of the different charge-compensating cations.22 The aluminophosphate structure VPI-5, on the other hand, is a channel-type VFI structure characterized by a large channel opening of 1.2 nm.17 The neutral framework of VPI-5 does not require charge-compensating cations. Previously VPI-5 was used for incorporation of large molecules such as C60 with a diameter of 1.0 nm.23 The precursor 1-N2 was adsorbed by a procedure reported in the Experimental Section, and the diffuse reflectance absorption spectra of the zeolite powders after adsorption were recorded and compared with the absorption spectrum of 1-N2 in n-pentane solution to examine whether 1-N2 is still intact (see Figure 1).

Figure 1. Absorption spectra of bis(2,4,6-trichlorophenyl)diazomethane (1-N2) in various media: (A) 1-N2 in n-pentane solution and (B) 1-N2 adsorbed in (a) VPI-5, (b) K+-L, and (c) Na+-Y measured by the diffuse reflectance mode at the same loading.

Close inspection of the spectra reveals that the spectral envelope of 1-N2 in solution is almost reproduced in VPI-5 but

We actually found 1H-OH as the main product in the extract from the spent samples of 1-N2-aluminosilicate zeolites, Na+-Y and K+-L. On the other hand, the neutral aluminophosphate framework of VPI-5, in principle, possesses no Brønsted-acid site, and hence, the precursor is included without appreciable decomposition. Notably, the entry aperture of VPI-5 (1.2 nm in diameter) is larger than that of K+-L (0.71 nm) or Na+-Y (0.74 nm). The molecular dimension of 1-N2 is roughly estimated to be 1.2 × 0.9 × 0.6 nm by the sum of van der Waals radii of the atoms constituting the molecule.24 We noted that a prolonged evacuating treatment for drying purposes appreciably reduced the absorption signal of 1-N2 adsorbed in K+-L, while no reduction in the absorption signal was detected in VPI-5 for a vacuumdrying period of 1-12 h. This may mean that adsorption of 1-N2 takes place only on the exterior surfaces of K+-L while 1-N2 can be included in the main channel of VPI-5. We also noted that the absorption signal in Na+-Y is very weak and very distorted even without the vacuum-drying process, suggesting that a more acidic medium in Na+-Y than that in K+-L may facilitate decomposition of 1-N2. Thus, the channel-type VPI-5 host is advantageous in terms of both a pore diameter larger than the molecular size of 1-N2 and the lack of Brønsted acidity. Accordingly, we chose the molecular sieves VPI-5 to study the behavior of carbenes in a confined system. Generation and Characterization of Triplet Carbenes in Zeolite. Triplet diarylcarbenes are usually characterized by ESR in organic matrix at low temperatures.25 Additionally, triplet diarylcarbenes were found to exhibit strong visible emission in organic matrix at 77 K.26 We observed that irradiation of 1-N2 in organic glasses (both ethanol and 2-methyltetrahydrofuran) gave rise to a species showing a structured emission band with the main peak at 490 nm (Figure 2) which is very similar to that observed for unsubstituted and substituted diphenylcarbenes.26 Since strong ESR signals ascribable to triplet carbene

20410 J. Phys. Chem. B, Vol. 109, No. 43, 2005

Hashimoto et al.

Figure 4. Emission (λex ) 340 nm) and excitation (λem ) 560 nm) spectra obtained after photolysis of bis(2,4,6-trichlorophenyl)methyl chloride (1H-Cl, 2.5 × 10-3 moL‚dm-3) in ethanol glass at 77 K. UV irradiation was carried out for 1 min at 77 K. 31

Figure 2. Emission and excitation spectra obtained after photolysis of 1-N2 (1.0 × 10-3 moL‚dm-3) in 2-methyltetrahydrofuran (A) and ethanol glasses (B) at 77 K: (a) emission spectra excited at 320 nm and (b) excitation spectra monitored at 490 nm. UV irradiation was carried out at 77 K for 1-10 min by an ultra-high-pressure mercury lamp with a Kenko U-330 filter and a K2Cr2O4 solution filter.

Figure 3. Emission spectra of 1-N2 (7.5 × 10-6 moL‚g-1) adsorbed in VPI-5 powders before (a) and after (b) UV irradiation for 10 min at 77 K excited at 350 nm. (c) Excitation spectrum of the sample after irradiation monitored at 560 nm.

were observed under identical conditions, the emission observed here is safely assigned to triplet 1 (31). Distinct visible emission from 31 will make identification of the species facile, especially confined in zeolites since ESR and absorption spectroscopic techniques routinely used in the solution phase are not easily applied in such a heterogeneous system. The VPI-5 powders containing 1-N2 were irradiated (λex ) 350 nm) for 10 min at 77 K, and emission spectra were measured immediately afterward at 77K. The sample after irradiation showed intense emission which was absent in the sample before irradiation and hence was ascribable to a photoproduct from 1-N2 (Figure 3). However, the observed emission (λmax ) 560 nm) and sharp excitation spectra are remarkably different from those observed for the triplet carbene

in organic glasses (Figure 2). Since it has been well documented25,26 that photolysis of diazo compounds generally results in generation of the corresponding carbenes regardless of the reaction phase and temperature, the product is likely to be formed from an initially produced carbene. It has been documented that diarylmethyl radicals also exhibit strong emission in the visible region.27,28 We confirmed that irradiation of bis(2,4,6-trichlorophenyl)methyl chloride (1HCl) in ethanol matrix at 77 K actually resulted in formation of a species showing identical emission and excitation spectra (Figure 4) to that observed in photolysis of 1-N2 in VPI-5 (Figure 3). Note that diarylmethyl halides undergo photochemical C-halogen bond cleavage to generate the corresponding diarylmethyl radicals.28

What is the origin of radical 1H then? 1H may arise from the corresponding alcohol (1H-OH) since alcohol is obtained as the main product (vide infra) and since diarylmethyl alcohols are shown to undergo homolysis of the C-O bond to generate the corresponding radical upon 248 nm laser excitation.29 However, this possibility is eliminated by the finding that no emission ascribable to 1H was detected upon irradiation of 1HOH in either ethanol glass or VPI-5 under essentially identical conditions where photolysis of 1-N2 and 1H-Cl generates emission due to 1H. Thus, we assign the species formed in the photolysis of 1-N2 in VPI-5 matrix to bis(2,4,6-trichlorophenyl)methyl radicals (1H), which can be produced as a result of hydrogen abstraction by 31. Triplet diarylcarbenes usually show high ability to abstract hydrogen to form the corresponding radicals.10 It is rather surprising to note here that triplet carbene 31 was not detected

Spectroscopic Investigation of Triplet Diarylcarbene

even at 77 K, which means that the carbenes generated in VPI-5 are quenched with unusual ease even at 77 K. This rather unusual observation is ascribable to the fact that in light-scattering media such as powders the light travels considerably long path lengths by repeated reflection and refraction at the interface between the void space and the solid particle before coming out. For instance, the diffusely reflected light remains more than 10-20 ps when a single 350 fs whitelight continuum pulse was introduced into a 2 mm thick sample of 1-100 µm PMMA (poly(methyl methacrylate)) latex particles, suggesting that the light travels at least a 30-60 times longer path length in a limited space than in transparent medium.30 This means that the light can visit the same place many times in the light-scattering medium. An effect similar to this may bring about multiple excitation of the carbenes once generated in the molecular sieve crystals. This will enhance the opportunity for H abstraction because the excited states of triplet carbenes are known to be more efficient at H abstraction than the ground-state species.31 This interpretation was supported by the following control experiment. Benzene forms a polycrystalline solid rather than a transparent glass when cooled to 77 K. When a polycrystalline sample of benzene containing 1-N2 was irradiated at 77 K, only the emission due to radical 1H was observed (Figure 5). This is in marked contrast with irradiation in glass matrixes such as ethanol and 2-methyltetrahydrofuran where the emission due to triplet carbene 31 was actually observed (Figure 2). Benzene is known as one of the poorest hydrogen donors and hence least reactive solvents toward triplet carbenes. For instance, triplet carbenes showed no sign of formation of the corresponding radicals when generated in benzene at room temperature.5 Thus, formation of radical 1H in benzene even at 77 K is evidence for the presence of the multiphotoexcitation process we assumed above. One way to avoid this undesirable effect is to immerse the zeolite sample in a fluid solution having similar refractive index to make an index-matched glassy solid. We chose propylene glycol (n ) 1.43) and/or glycerol (n ) 1.49) as such a medium because they have a refractive index similar to that of VPI-5 (n ≈ 1.5)32 and also form a transparent glassy matrix at 77 K. Thus, when the 1-N2-loaded molecular sieve VPI-5 is immersed in these fluids, the light-scattering effect of the powders is expected to be reduced. Indeed, irradiation of the sample resulted in formation of initial photoproduct exhibiting an emission essentially identical to that of 31 (Figure 6a). Additionally, ESR measurement showed triplet signals ascribable to 31 in the presence of the index-matching fluid, whereas only the signal due to organic radicals was detected in the absence of the fluids. Thus, the photoproduct formed under these conditions could safely be assigned to 31. Interestingly, as irradiation is continued, the emission due to 31 initially increased (Figure 6b) but decreased later with a concomitant increase in the emission ascribable to radical 1H (Figure 6c and d). This indicates that 31 undergoes a hydrogen-abstraction reaction upon photoexcitation even under these conditions. The results support the idea that the observation of radical 1H in the photolysis of 1-N2loaded VPI-5 at 77 K is due to the multiple excitation of 31.

J. Phys. Chem. B, Vol. 109, No. 43, 2005 20411

Figure 5. Emission spectra (λex ) 340 nm) obtained before (a) and after (b) irradiation of 1-N2 (1.0 × 10-3 moL‚dm-3) in a polycrystalline solid of benzene at 77 K. UV irradiation was carried out for 30 min at 77 K.

Figure 6. Emission spectral change (λex ) 340 nm) at 77 K obtained after irradiation of 1-N2 (1.2 × 10-5 moL‚g-1) adsorbed in VPI-5 immersed in propylene glycol as a refractive-index-matching fluid: irradiation time (a) 1, (b) 2, (c) 12, and (d) 40 min.

It is possible that 1-N2 molecules once adsorbed into VPI-5 may come out of the matrix and eventually dissolve into the refractive-index-matching fluids employed. This concern was eliminated by the fact that neither the supernatant of the mixture of 1-N2-loaded VPI-5 and glycerol gave absorption signal of 1-N2 nor irradiation of supernatant at 77 K gave the emission or ESR signal of 31. This is consistent with the generally accepted observation in zeolite science that aromatic molecules once adsorbed do not come out unless they are replaced by residual water molecules in the employed solvent.33 These observations clearly show that 31 is indeed generated in the cavity of VPI-5. Thus, the next step is to examine how triplet carbene is stabilized inside the cavity. Stability of Triplet Carbenes in the Cavity of the Zeolite. Previous laser photolysis studies of diaryldiazomethanes in solution have shown that application of a single nanosecond laser pulse to the precursors is sufficient to generate diphenylcarbenes and simultaneously excite to the upper manifolds to allow for the observation of the emission spectrum.31b,34 The transient absorption spectra of diphenylcarbenes have also been measured previously in solution by the laser photolysis technique.35 Laser photolysis of 1-N2 incorporated in the molecular sieve VPI-5 was carried out at ambient temperature by irradiation with a 266 nm, 8 ns pulsed-laser light and with fast detection of both the emission and the transient absorption signals. Transient absorption at room temperature in the nanosecond time regime on excitation of 1-N2 in VPI-5 is shown in Figure 7A, which also includes absorption spectra of VPI-5 in the absence of 1-N2 (Figure 7A(b)) and the transient band due to 31 observed

20412 J. Phys. Chem. B, Vol. 109, No. 43, 2005

Figure 7. (A) Transient absorption spectra obtained after irradiation of 1-N2 with a pulsed laser light: (a) transient absorption of 1-N2 adsorbed in VPI-5 100 ns after a 266 nm laser pulse, (b) transient absorption of VPI-5 host in the absence of 1-N2 100 ns after a 266 nm laser pulse, and (c) transient absorption signal of 1-N2 in benzene solution 50 µs after the laser pulse (reproduced from ref 16b). (B) (a) Transient emission signal obtained after photolysis of 1-N2 adsorbed in VPI-5 100 ns after a laser pulse, and (b) emission signal of VPI-5 host in the absence of 1-N2 100 ns after a laser pulse.

in a similar laser photolysis of 1-N2 in benzene at room temperature (Figure 7A(c)). The emission spectrum given in Figure 7B(a) showed a rather sharp emission centered at 560 nm on a broad one centered at 500 nm. Since the emission spectrum of the host VPI-5 in the absence of 1-N2 exhibited a broad spectrum centered at 500 nm (Figure 7B(b)), which is similar to the broad one in Figure 7B(a), the sharp emission at 560 nm is assigned to a transient species formed from 1-N2. Essentially the same absorption and emission spectra were observed when laser photolysis was carried out for the 1-N2loaded molecular sieves VPI-5 immersed in glycerol as a refractive-index-matching fluid. The transient absorption and emission spectra are similar to that of radical 1H rather than triplet carbene 31.36 Thus, it is likely that only diarylmethyl radical 1H is detected in the nanosecond time regime, suggesting rapid hydrogen transfer to initially generated 31 from the medium. In light of the fact that 31 was shown to have a half-life of 18 ms in benzene at room temperature,16b,c the unusually short lifetime of 31 due to rapid hydrogen abstraction is probably caused by the multiple excitation effect in the zeolite medium. In other words, it is likely that a net stabilization effect by zeolite on triplet carbene cannot properly be estimated as long as the pumping light interferes. Potential hydrogen donors present in the VPI-5 environment deserve some comment at this time. First, the solvent molecules possibly remaining in the channels of VPI-5 after evacuation may participate in the reaction. In the usual sense, benzene does

Hashimoto et al. not serve as a H donor whereas n-pentane can be a donor. However, one should take into account the observation that benzene may act as a hydrogen donor under conditions where the multiple-excitation mechanism is operating (vide supra). Although we could not obtain solid evidence that the solvent molecules can act as a H donor, this possibility cannot be ruled out. Second, one may suspect H2O, which may remain in the cavities even after thorough dehydration treatments, is responsible. Although we tried the air- and moisture-free sample preparation method, the possibility of a small quantity of water molecules adsorbed in the host still remains. It is generally accepted that complete dehydration of zeolites is difficult to attain even after a prolonged period of evacuation under high vacuum at temperatures as high as ∼800 °C.21 However, water is usually not regarded as a good hydrogen donor unless the electron transfer leading to ion radical species is involved. Instead, water reacts with the singlet state of carbenes giving rise to alcohols as a result of O-H insertion.1 It has been reported that large amounts of product arising from reaction of singlet carbene with water were produced in the photolysis of diazirines entrapped in zeolite.11 We also isolated a hydroxide compound, bis(2,4,6-trichlorophenyl)methyl alcohol (1H-OH), after irradiation of 1-N2 incorporated in VPI-5, which can be indicative of OH insertion. Third, we examine the possibility that hydroxyl groups may exist in the pores of VPI-5 and participate in the H abstraction. In general, the external surfaces of zeolites are covered by terminal OH groups.21 However, in principle, the neutral aluminophosphate framework of VPI-5 does not allow such OH groups to exist inside the pores because [AlO4]- and [PO4]+ tetrahedra with opposite charges are found in a strictly alternating arrangement.17 Such OH groups may exist as a result of structural defects; however, the heating treatment at 500-600 °C should result in condensation of such bonds according to

Al-OH + HO-P f Al-O-P + H2O Since we calcined the materials in our preparation, the number of terminal OH groups within the pores should be low and the role of the terminal OH groups as H donors should be minor. The stability of triplet carbenes can also be assessed by a variable-temperature ESR study. In this case, triplet carbenes are generated by photolyzing the precursor compounds in a matrix at low temperature, usually at 77 K. Then, matrix containing the triplet carbenes is gradually warmed to the desired temperature, followed by standing at this temperature for 5 min and then recooling to 77 K to measure the signal. The measurement is carried out until the signals disappear completely. This procedure can compensate for weakening of signals due to Curies law.37 In this procedure an accelerated decay of the species by a multiple excitation mechanism is not expected. Photolysis of 1-N2 in VPI-5 immersed in glycerol by UV irradiation at 77 K gave ESR signals with typical fine structure patterns for unoriented triplet species, i.e., triplet carbene 31, along with strong signals centered at ca. 330 mT ascribable to a radical species (Figure 8). Signals at 66, 464, and 516 mT are assigned to a low-field z and a set of high-field x and y transitions, respectively, from which the zero-field splitting parameters were obtained as D ) 0.370 cm-1 and E ) 0.012 cm-1. The variable-temperature measurements were made until the signals disappeared irreversibly. The highest survival temperature observed was found to be 220 K. This is significantly higher than that observed in 2-methyltetrahydrofuran in which the highest temperature is estimated to be 120 K.

Spectroscopic Investigation of Triplet Diarylcarbene

Figure 8. ESR spectra obtained at 77 K by irradiation of 1-N2 (a) dissolved in MTHF and (b) adsorbed in VPI-5 immersed in glycerol as a refractive-index-matching fluid.

Conclusion We were able to encapsulate a diazo compound into a zeolite without any appreciable decomposition and generate, directly observe, and characterize triplet diarylcarbene in the cavity of the zeolite by photolyzing the complex for the first time. Due to an unexpected event of multiexcitation in zeolites and the role of the VPI-5 host acting as a hydrogen donor, we were unable to show direct evidence that the lifetime of triplet carbenes at room temperature is increased by retarding the bimolecular combination process (the main decay channel of the carbenes in solution). Nevertheless, confinement by zeolite was demonstrated to have a significant effect on the stability of this extremely active species. Finally, it is interesting to note here that laser excitation of “parent” diphenyldiazomethane in hydrocarbon solvent at room temperature gives 9,10-diphenylanthracene, 9,10-diphenylphenanthrene, and fluorene in addition to tetraphenylethylene, while conventional irradiation results in formation of benzophenone azine, and hence the products obtained from laser excitation are assumed to be formed as a result of re-excitation of triplet diphenylcarbene.38 Our product analysis and spectroscopic studies indicated alcohol 1H-OH is the only final product in the present system. This is not surprising since triplet carbene16 employed in this study shows very different reactivity from “parent” diphenylcarbene and hence is not likely to undergo similar reaction even upon further excitation. In this respect it may be interesting to study the reactivity of parent diphenylcarbene in zeolites. Acknowledgment. S.H. is grateful to the Japanese Ministry of Education, Culture, Sports, Science and Technology, Grantin-Aid for Scientific Research on Priority Areas 417, #17029067, and the Japan Society for the Promotion of Science, Grant-inAid for Scientific Research, #16550024, for support of this work. References and Notes (1) For reviews of general reactions of carbenes, see: (a) Kirmse, W. Carbene Chemistry, 2nd ed.; Academic Press: New York, 1971. (b) Carbenes; Moss, R. A., Jones, M., Jr., Eds.; Wiley: New York, 1973 and 1975; Vols. 1 and 2. (c) Carbene(oide), Carbine; Regitz, M., Ed.; Thieme: Stuttgart, 1989. (d) Wentrup, C. ReactiVe Intermediates; Wiley: New York, 1984; pp 162-264. (2) Igau, A.; Gru¨tzmacher, H.; Baceiredo, A.; Bertrand, G. J. Am. Chem. Soc. 1988, 110, 6463-6466. See, for review: Bourissou, D.; Guerret, O.; Gabba¨i, F. P.; Bertrand, G. Chem. ReV. 2000, 100, 39-92. (3) Aruduengo, A. J., III; Harlow, R. L.; Kline, M. J. Am. Chem. Soc. 1991, 113, 361-363. See, for reviews: Aruduengo, A. J., III Acc. Chem. Res. 1999, 32, 913-921.

J. Phys. Chem. B, Vol. 109, No. 43, 2005 20413 (4) (a) Kirmse, W. Angew. Chem., Int. Ed. 2003, 42, 2117-2119. (b) Regitz, M. Angew. Chem., Int. Ed. Engl. 1991, 30, 674-676. (c) Dagani, R. Chem. Eng. News 1991, Jan 28, 19. Dagani, R. Chem. Eng. News 1994, May 2, 20. (5) See, for reviews on persistent triplet carbenes: (a) Tomioka, H. Acc. Chem. Res. 1997, 30, 315-321. (b) Tomioka, H. In AdVances in Carbene Chemistry; Brinker, U., Ed.; JAI Press: Greenwich, CT, 1998; Vol. 2, pp 175-214. (c) Tomioka, H. In Carbene Chemistry; Bertrand, G., Ed.; Fontis Media S. A.: Lausanne, 2002; pp 103-152. (6) Iwamoto, E.; Hirai, K.; Tomioka, H. J. Am. Chem. Soc. 2003, 125, 14664-14665. (7) Kirmse, W. Angew. Chem., Int. Ed. 2005, 44, 2476-2479. (8) Liu, X.; Chu, G.-S.; Moss, R. A.; Sauers, R.; Warmuth, R. Angew. Chem., Int. Ed. 2005, 44, 1994-1997. (9) (a) Warmuth, R.; Marvel, M. A. Angew. Chem., Int. Ed. 2000, 39, 1117-1119. (b) Warmuth, R.; Marvel, M. A. Chem. Eur. J. 2001, 7, 12091220. (c) Warmuth, R. J. Am. Chem. Soc. 2001, 123, 6955-6956. (d) Kerdelhue, J.-L.; Langenwalter, K. J.; Warmuth, R. J. Am. Chem. Soc. 2003, 125, 973-986. (10) Tomioka, H. In ReactiVe Intermediate Chemistry; Moss, R. A., Platz, M. S., Jones, M., Jr., Eds.: Wiley: New York, 2004; pp 375-461. (11) Reactions of carbenes with a singlet ground state in a zeolite have been studied. See, for instance: (a) Moya-Barrios, R.; Cozens, F. Org. Lett. 2004, 6, 881-884. (b) Kupfer, R.; Poliks, M. D.; Brinker, U. H. J. Am. Chem. Soc. 1994, 116, 7393-7398. (12) Zhang, G.; Liu, X.; Thomas, J. K. Radiat. Phys. Chem. 1998, 51, 135-152. (b) Yoon, K. B. In Handbook of Zeolite Science and Technology; Auerbach, M., Carrado, K. A., Dutta, P. K., Eds.; Marcel Dekker: New York, 2003; pp 591-720. (13) Hashimoto, S. J. Photochem. Photobiol. C: ReV. 2003, 4, 19-49. (14) Garcı´a, H.; Roth, H. D. Chem. ReV. 2002, 102, 3947-4007. (15) Turro, N. J. Acc. Chem. Res. 2000, 33, 637-646. (16) (a) Zimmerman, H. E.; Paskovich, D. H. J. Am. Chem. Soc. 1964, 86, 2149-2160. (b) Tomioka, H.; Hirai, K.; Nakayama, T. J. Am. Chem. Soc. 1993, 115, 1285-1289. (c) Tomioka, H.; Hirai, K.; Fujii, C. Acta Chem. Scand. 1992, 46, 680-682. (17) (a) Annen, M. J.; Young, D.; Davis, M. E.; Cavin, O. B.; Hubbard, C. R. J. Phys. Chem. 1991, 95, 1380-1383. (b) Schmidt, W.; Schu¨th, F.; Reichert, H.; Unger, K.; Zibrowius, B. Zeolites 1992, 12, 2-8. (c) He, H.; Klinowski, J. J. Phys. Chem. 1994, 98, 1192-1197. (18) Yu, J.; Xu, R. Acc. Chem. Res. 2003, 36, 481-490. (19) Hashimoto, S. Solid State and Surface Photochemistry; Ramamurthy, V., Schanze, K. S., Eds.; Marcel Dekker: New York, 2000; pp 253294. (20) Regitz, M.; Maas, G. Diazo Compounds-Properties and Synthesis; Academic Press: Orlando, FL, 1986. (21) Breck; D. W. Zeolite Molecular SieVes; Krieger: Malabar, 1984. (22) Ramamurthy, V.; Robbins, R. J.; Thomas, K. J.; Lakshiminarasimhan, P. H. In Organized Molecular Assemblies in the Solid State; Whitesell, J. K., Ed.; Wiley: Chichester, 1999; pp 63-140. (23) (a) Gu¨gel, A.; Mu¨llen, K.; Reichert, H.; Schmidt, W.; Scho¨n, G.; Schu¨th, F. Angew. Chem., Int. Ed. 1993, 32, 556-557. (b) Anderson, M. W.; Shi, J.; Leigh, D. A.; Moody, A. E.; Wade, F. A.; Hamilton, B.; Carr, S. W. J. Chem. Soc., Chem. Commun. 1993, 533-536. (c) Kwon, O.-H.; Park, K.; Jang, D.-J. Chem. Phys. Lett. 2001, 346, 195-200. (24) The geometry of 1-N2 was optimized with a molecular orbital calculation program “Spartan” (Wavefunction, Inc.) using the STO-3G basis set. (25) (a) Sander, W.; Bucher, G.; Wierlacher, S. Chem. ReV. 1993, 93, 1583. (b) Trozzolo, A. M.; Wasserman, E. In Carbenes; Jones, M., Jr., Moss, R. A., Eds.; Wiley: New York, 1975; Vol. 2, pp 185-206. (c) Tomioka, H. In AdVances in Strained and Interesting Organic Molecules; Halton, B., Ed.; JAI Press: Greenwich, CT, 2000; Vol. 8, pp 83-112. (26) (a) Trozzolo, A. M. Acc. Chem. Res. 1968, 1, 329-335 and references therein. (b) Ono, Y.; Ware, W. R. J. Phys. Chem. 1983, 87, 4426-4433. (27) (a) Favaro, G. Spectrochim. Acta 1970, 26A, 2327-2332. (b) Scaiano, J. C.; Weir, D. Can. J. Chem. 1988, 66, 491-494. (28) (a) Bromberg, A.; Schmidt, K. H.; Meisel, D. J. Am. Chem. Soc. 1984, 106, 3056-3057. (b) Meisel, D.; Das, P. K.; Hug, G. L.; Bhattacharyya, K.; Fessenden, R. W. J. Am. Chem. Soc. 1986, 108, 4706-4710. (29) (a) Faria, J. L.; Steenken, S. J. Phys. Chem. 1993, 97, 1924-1930. (b) We thank a referee for pointing out this possibility. Faria and Steenken (ref 29a) observed formation of diarylmethyl radicals by 248 nm laser photolysis of Ar2CH-OH in acetonitrile at room temperature. However, the quantum yield for the photohomolysis of Ar2CH-OH seems to be very low compared with that of Ar2CH-Cl, although not stated accurately in the literature. Besides the difference in the excitation wavelength from our experiments, we also suspect that a biphotonic process might be involved in formation of diarylmethyl radicals in the laser photolysis. The two factors can mainly be responsible for our failure to observe 1H by (monophotonic) photolysis of 1H-OH in addition to the difference in reaction medium.

20414 J. Phys. Chem. B, Vol. 109, No. 43, 2005 (30) Furube, A.; Asahi, T.; Masuhara, H. Jpn. J. Appl. Phys. 1999, 69, 4236-4243. (31) (a) Leyva, E.; Barcus, R. L.; Platz, M. S. J. Am. Chem. Soc. 1986, 108, 7786-7788. (b) Scaiano, J. C. In Kinetics and Spectroscopy of Carbenes and Biradicals; Platz, M. S., Ed.; Plenum Press: New York, 1990; pp 353-368. (32) Schneider, J.; Fanter, D.; Bauer, M.; Schomburg, C.; Wo¨hrle, D.; Schults-Ekloff, G. Microporous Mesoporous Mater. 2000, 39, 257-263. (33) (a) Hashimoto, S.; Ikuta, S.; Asahi, T.; Masuhara, H. Langmuir 1998, 14, 4284-4291. (b) Pauchard, M.; Devaux, A.; Calzaferri, G. Chem. Eur. J. 2000, 6, 3456-3470. (34) Dupuy, C.; Korenowski, G. M.; McAuliffe, M.; Hetherington, W. M., III; Eisenthal, K. B. Chem. Phys. Lett. 1981, 77, 272-274.

Hashimoto et al. (35) (a) Horn, K A.; Allison, B. D. Chem. Phys. Lett. 1985, 116, 114118. (b) Johoston, L. J.; Scaiano, J. C. Chem. Phys. Lett. 1985, 116, 109113. (36) For transient absorption spectrum of diphenylmethyl radicals, see, for instance: (a) Hadel, L. M.; Platz, M. S.; Sciano, J. C. J. Am. Chem. Soc. 1984, 106, 283-287. (b) Sciano, J. C.; Tanner, M.; Weir, D. J. Am. Chem. Soc. 1985, 107, 4396-4403. (37) (a) Platz, M. S. In Diradicals; Borden, W. T., Ed.; Wiley: New York, 1982; pp 195-258. (b) Wasserman, E.; Hutton, R. S. Acc. Chem. Res. 1977, 10, 27-32. (c) Breslow, R.; Chang, H. W.; Wasserman, E. J. Am. Chem. Soc. 1967, 89, 1112-1119. (38) Turro, N. J.; Aikawa, M.; Butcher, J. A., Jr.; Griffin, G. W. J. Am. Chem. Soc. 1980, 102, 5127-5128.