Infrared spectra and ultraviolet-tunable laser induced photochemistryof matrix-isolated phenol and phenol-d5Barbara Michela Giuliano, Igor Reva, Leszek Lapinski, and Rui Fausto Citation: J. Chem. Phys. 136, 024505 (2012); doi: 10.1063/1.3666018 View online: http://dx.doi.org/10.1063/1.3666018 View Table of Contents: http://jcp.aip.org/resource/1/JCPSA6/v136/i2 Published by the American Institute of Physics. Related ArticlesCommunication: Manipulating the singlet-triplet equilibrium in organic biradical materials JCP: BioChem. Phys. 5, 12B401 (2011) Communication: Manipulating the singlet-triplet equilibrium in organic biradical materials J. Chem. Phys. 135, 241101 (2011) Orbital alignment in photodissociation probed using strong field ionization J. Chem. Phys. 135, 234311 (2011) Hot hole-induced dissociation of NO dimers on a copper surface J. Chem. Phys. 135, 224708 (2011) Reactions between chlorine atom and acetylene in solid para-hydrogen: Infrared spectrum of the 1-chloroethylradical J. Chem. Phys. 135, 174302 (2011) Additional information on J. Chem. Phys.Journal Homepage: http://jcp.aip.org/ Journal Information: http://jcp.aip.org/about/about_the_journal Top downloads: http://jcp.aip.org/features/most_downloaded Information for Authors: http://jcp.aip.org/authors

THE JOURNAL OF CHEMICAL PHYSICS 136, 024505 (2012)

Infrared spectra and ultraviolet-tunable laser induced photochemistryof matrix-isolated phenol and phenol-d5

Barbara Michela Giuliano,1 Igor Reva,1,a) Leszek Lapinski,1,2 and Rui Fausto1

1Department of Chemistry, University of Coimbra, 3004-535 Coimbra, Portugal2Institute of Physics, Polish Academy of Sciences, Al. Lotnikow 32/46, 02-668 Warsaw, Poland

(Received 3 October 2011; accepted 16 November 2011; published online 12 January 2012)

Monomers of phenol and its ring-perdeuterated isotopologue phenol-d5 were isolated in argon ma-trices at 15 K. The infrared (IR) spectra of these species were recorded and analyzed. In situ pho-tochemical transformations of phenol and phenol-d5 were induced by tunable UV laser light. Thephotoproducts have been characterized by IR spectroscopy supported by theoretical calculations ofthe infrared spectra. The primary product photogenerated from phenol was shown to be the phe-noxyl radical. The analysis of the progress of the observed phototransformations led to identificationof 2,5-cyclohexadienone as one of the secondary photoproducts. Spectral indications of other sec-ondary products, such as the Dewar isomer and the open-ring ketene, were also detected. Identifica-tion of the photoproducts provided a guide for the interpretation of the mechanisms of the observedphotoreactions. © 2012 American Institute of Physics. [doi:10.1063/1.3666018]

INTRODUCTION

Phenol is a fundamental building block of aromaticbiomolecules such as amino acids and neurotransmitters.Because of its function as chromophore in many systems(tyrosine and tyramine among others), the reactions ofphenol upon UV irradiation are of great importance ininvestigating molecular damage mechanisms.1 However,even the photochemistry of monomeric phenol is still acontroversial topic, as shown by the debates concerning thegeometry of excited states,2 favored photoprocesses and theirmechanisms,3–6 as well as characterization of intermediatesand photoproducts.7, 8 These issues have attracted great in-terest and were intensively investigated using computationaland experimental methods.

Various studies have been published about H-atomdetachment upon UV photolysis of phenol molecules inthe gas phase.3, 4, 9 The three lowest singlet electronic statesinvolved in the UV photochemistry of phenol are the groundand first two excited electronic states, S0(ππ ), S1(ππ*), andS2(πσ*). The profiles of the potential energy curves of thesestates result in two conical intersections, S0/S2 and S1/S2,due to the dissociative shape of the potential energy curve ofS2(πσ*) along the OH stretching coordinate.6 Two channelsof phenol phototransformations are predicted to competewith one another. One of them is associated with the presenceof a so-called prefulvene isomer, while the other is associatedwith the photodetachment of the hydrogen atom from thehydroxyl group. The existence of the prefulvenic reactionpath has already been predicted10 and observed11 in benzene,while in phenol it has not been confirmed yet.

In this paper, we present an investigation of the photo-chemistry of phenol and its isotopologue phenol-d5 isolatedin argon matrices. The use of narrowband tunable laser

a)Author to whom correspondence should be addressed. Electronic mail:reva@qui.uc.pt.

source of UV light allowed the exploration of differentreaction pathways. The characterization of intermediates andfinal products was undertaken by infrared (IR) spectroscopysupported by density functional theoretical calculations.When comparing our vibrational data to those available inthe literature, it turned out that assignment of the vibrationalspectrum of monomeric phenol is still controversial.12, 13

Moreover, the vibrational spectroscopy of phenol was re-cently reviewed14 and the experimental data on vibrationalspectroscopy of matrix-isolated phenol are still lacking.15, 16

Then, the additional purpose of the present study is to closethis gap and provide a detailed analysis of the experimentalinfrared spectrum of phenol and phenol-d5.

EXPERIMENTAL PROCEDURES

Crystalline phenol (99%) and phenol-2,3,4,5,6-d5

(98% D) were commercial products supplied by Panreac andAldrich, respectively. The sample was placed in an evacuatedglass tube connected through a needle valve to the vacuumchamber of a helium-cooled cryostat (APD Cryogenicsclosed-cycle refrigerator with DE-202A expander) as de-scribed in Ref. 17. Prior to deposition of the matrices, phenol(or phenol-d5) was purified from the volatile impurities bypumping through the cryostat at room temperature. Duringdeposition of the matrices, the sample tube with phenol (orphenol-d5) was immersed in ice-water mixture (at 0 ◦C),to reduce the saturated pressure of phenol and improve themetering function of the needle valve. In order to prepare thelow-temperature matrices, the vapors over solid phenol weredeposited together with a large excess of argon (purity N60,supplied by Air Liquide) onto a CsI window cooled to 15 K.The IR spectra were recorded in the 4000–400 cm−1 range,with 0.5 cm−1 resolution, using a Thermo Nicolet 6700Fourier-transform infrared (FTIR) spectrometer equippedwith a KBr beam splitter and a DTGS detector. Matrices

0021-9606/2012/136(2)/024505/11/$30.00 © 2012 American Institute of Physics136, 024505-1

024505-2 Giuliano et al. J. Chem. Phys. 136, 024505 (2012)

were irradiated through the outer quartz window of thecryostat with monochromatic UV light from the 345–220 nmrange. Frequency-doubled signal beam of the Quanta-RayMOPO-SL pulsed (10 ns) optical parametric oscillator (fullwidth at half-maximum (FWHM) ∼ 0.2 cm−1, repetition rate10 Hz, pulse energy ∼2.5 mJ) was applied for this purpose.The optical parametric oscillator was pumped with a pulsedNd:YAG laser.

COMPUTATIONAL METHODS

The geometries of phenol and the structures of possibleproducts photogenerated from phenol were fully optimizedusing the density functional theory method (DFT) withBecke’s three-parameter exchange functional,18 the gradient-corrected functional of Lee, Yang, and Parr,19 and the Vosko,Wilk, and Nusair correlation functional20 (B3LYP). Thestandard 6-311++G(d,p) basis set was used in these calcula-tions. At the optimized geometries, the harmonic vibrationalfrequencies and IR intensities were calculated at the sameDFT(B3LYP)/6-311++G(d,p) level. All the calculationswere performed with the GAUSSIAN 03 program.21

The theoretical normal modes were analyzed by carryingout potential energy distribution (PED) calculations. Trans-formations of the force constants with respect to the Carte-sian coordinates to the force constants with respect to themolecule fixed internal coordinates allowed the PED analy-sis to be carried out as described by Schachtschneider andMortimer.22 The internal symmetry coordinates used in thisanalysis (listed in Table S1 of the supplementary material)47

were defined as recommended by Pulay et al.23 The elementsof the computed PED matrices greater than 10% are presentedin Tables I and II.

The theoretically predicted spectra were used to assist theanalysis of the experimental spectra. The calculated harmonicfrequencies were scaled according to the empirical equationν(scaled) = 0.943 × ν(calculated) + 42.6, resulting from theleast squares linear fitting of the calculated frequencies to theexperimentally observed frequencies (for both isotopologuesof phenol). This scaling approach was then used throughoutthe study, and was also applied for scaling the predictedspectra of the photoproducts. For the purpose of modeling IRspectra, the scaled calculated frequencies, together with thecalculated intensities served to simulate the spectra shownin the figures by convoluting each peak with a Lorentzianfunction with a full width at half-maximum of 2 cm−1, sothat the integral band intensities correspond to the calculatedinfrared absolute intensity.24 Note that the peak intensities inthe simulated spectra are several times lower (in the arbitraryunits of “relative intensity”) than the calculated intensity (inkm mol−1).

RESULTS AND DISCUSSION

As-deposited matrix

The infrared spectra of phenol and phenol-d5 isolated inargon matrices are presented in Figures 1 and 2. The main ab-sorption features found in the spectrum of phenol correspond

well to the IR bands previously reported by Plokhotnichenkoet al.25 Due to the limitations of a grating infrared spectrom-eter, with resolutions of 3 cm−1 (OH stretching range) and 1cm−1 (fingerprint range), only selected spectral ranges werestudied in the previous work.25 Usage of a Fourier-transforminfrared spectrometer in the current work allowed registrationof the complete mid-infrared spectrum of phenol and phenol-d5. The better resolution of the FTIR spectrometer allowedobservation of some more details in the spectra. This concernssplitting of some IR bands that can be attributed to matrix-site effects and anharmonic resonances. Previously, matrix-isolated phenol has also been obtained as a result of the pho-tolysis of benzene/ozone mixtures in an argon matrix at 12 Kwith UV light of λ ≥ 280 nm.26 However, the infrared spec-trum of phenol isolated in argon in the presence of O2 and O3

molecules differs slightly from the present spectrum wherephenol is isolated in neat argon.

The infrared spectra of matrix-isolated phenol andphenol-d5 are compared (in Figures 1 and 2) with thetheoretical spectra calculated for both isotopologues at theDFT(B3LYP)/6-311++G(d,p) level. The good agreement be-tween the experimental and theoretical spectra allowed a re-liable assignment of the observed IR bands. The assignment,together with the PED analysis, is provided for phenol andphenol-d5 in Tables I and II.

A particular attention should be called here to the rangeof CH and CD stretching vibrations (Figure 2). In the ex-periments conducted with phenol-d5, no absorptions appearin the 3100–3000 cm−1 range (Figure 2(c)), where the CHstretching frequencies are expected, thus implying a completeperdeuteration of the aromatic ring. This detail will deemimportant in the interpretation of the photochemical experi-ments.

Irradiation at λ = 275 nm

Monomers of phenol isolated in Ar matrices were irra-diated with narrowband UV laser light. For the first irradi-ation, UV light with λ = 345 nm was chosen. Graduallyshorter wavelengths of UV light were used for subsequentirradiations of the matrix. After each irradiation, the matrixwas monitored by taking its IR spectrum. The cycles of UVirradiation and FTIR monitoring continued, with UV wave-length reduced every time by 1 nm. Irradiations with UVλ ≥ 276 nm did not induce any transformations in matrix-isolated phenol. By irradiation at 275 nm, the IR bands as-signed to phenol monomer started to decrease, while newbands due to photogenerated species appeared in the IR spec-trum. The wavelength 275 nm corresponds to the origin ofthe S1 ← S0 (π* ← π ) transition in phenol, as reported inseveral works.4, 27 The irradiation at λ = 275 nm continued(with periodical monitoring of the progress of the UV-inducedtransformations) for a total time of 90 min. After this period,the amount of phenol in the matrix decreased to ∼5% of theinitial abundance (Figure 3). The new IR bands due to pho-toproducts can be grouped in several sets, each one showinga different pattern of intensity changes during the progress of

024505-3 IR spectra and photochemistry of phenol J. Chem. Phys. 136, 024505 (2012)

TABLE I. Experimental wavenumbers (v /cm−1) and relative integral intensities (I) of the absorption bands in the infrared spectrum of phenol isolated in anAr matrix, compared with wavenumbers (v /cm−1), absolute intensities (Ath / km mol−1) and potential energy distribution (PED, %) calculated for phenol atthe B3LYP/6–311++G(d,p) level.

Experimental Ar matrix Calculated B3LYP/6–311++G(d,p)

v a I b v c Ath Sym. PEDd (%)

3638.3; 3633.9 90.8 3660.0 61.5 A′ ν(O7-H8) (100)3490.1e 13.73460.0e 20.4

43.8

⎧⎪⎪⎪⎪⎨⎪⎪⎪⎪⎩

3116.2; 3108.6; 3096.0; 3057.0 4.4 A′ νa(CH) (90)3080.4; 3074.1; 3050.8 16.3 A′ νb(CH) (69), νd(CH) (21)3060.2; 3053.0; 3048.6; 3047.0; 3037.6 16.8 A′ νd(CH) (76), νb(CH) (15)

3029.4 0.1 A′ νc(CH) (83), νb(CH) (11)3040.0; 3027.9; 3019.2 3012.1 13.7 A′ ν(C2-H9) (89)2065.1 1.21943.5 0.91921.8 3.41854.5 1.01836.5; 1831.8 2.11778.8; 1767.8 1.41703.2; 1696.8; 1691.6 3.41660.6 3.01610.3 41.1 1594.7 39.3 A′ νc(CC) (66), δd(CH) (12)1601.4; 1600.0 40.9 1584.8 49.1 A′ νb(CC) (68)1526.6 1.71504.5; 1501.0 67.2 1482.4 59.3 A′ δb(CH) (53), νd(CC) (18), νe(CC) (17)1470.9 36.2 1456.5 22.9 A′ δc(CH) (47), νd(CC) (18), νe(CC) (18)1342.5 22.4 1332.9 27.5 A′ δa(CH) (63), δ(O7-H8) (20)1331.1; 1330.1; 1327.1 6.01309.8 2.6 1314.2 5.6 A′ νf(CC) (71), δe(CH) (17)1264.3; 1262.8; 1260.7; 1259.5; 1256.3; 1255.7; 1255.1; 1253.7 62.2 1244.9 91.0 A′ ν(C1-O7) (50), δb(CH) (16), νa(CC) (14), δ(R1) (10)1202.2; 1199.6; 1196.8 14.91175.9; 1174.4 76.5 1166.6 128.1 A′ δ(O7-H8) (43), δa(CH) (16)1168.6 8.2 1165.2 2.0 A′ δd(CH) (68), νb(CC) (11), νc(CC) (10)1161.6; 1160.2; 1158.3 13.51155.9; 1153.1 3.11149.9 17.5 1152.1 26.0 A′ δe(CH) (67), νf(CC) (18)1070.9 6.5 1073.3 14.5 A′ νe(CC) (42), δc(CH) (37)1061.5 5.71025.8 3.6 1026.2 5.8 A′ νd(CC) (42), δb(CH) (25), νa(CC) (19)1000.5 5.2 997.4 2.4 A′ δ(R1) (63), νa(CC) (35)n.o.f 959.0 0.1 A′′ γ (C5-H12) (72), γ b(CH) (32), γ a(CH) (16)n.o. 940.2 0.2 A′′ γ c(CH) (98), γ b(CH) (10)883.7; 882.6; 880.7 11.4 870.6 4.7 A′′ γ d(CH) (102)830.1; 828.0; 826.0; 823.8 7.2812.3 20.5 822.8 22.8 A′ νa(CC) (28), ν(C1-O7) (27), δ(R1) (18), δ(R2) (15)n.o. 815.4 0.1 A′′ γ b(CH) (66), γ (C5-H12) (20), γ c(CH) (13)754.1; 753.3; 752.1 83.8 749.4 84.9 A′′ γ a(CH)(66), γ (C1-O7) (20), γ (C5-H12)(10)689.2; 687.1 43.0 673.0 9.8 A′′ τ (R1) (103)n.o. 639.1 0.4 A′ δ(R3) (83)591.2; 585.1 8.6 2× τ (O7-H8)526.2 1.5 548.0 2.1 A′ δ(R2) (76), ν(C1-O7) (12)509.0; 507.7; 506.1; 504.6; 503.8 21.7 522.4 14.2 A′′ γ (C1-O7) (46), τ (R2) (38), τ (R1) (12), γ a(CH) (10)n.o. 435.6 1.0 A′′ τ (R3) (116)402.1 9.7 424.6 10.5 A′ δ(C1-O7) (81)n.i.g 90.8 338.3 110.7 A′′ τ (O7-H8) (99)n.i. 257.4 1.4 A′′ τ (R2) (75), γ (C1-O7) (17), τ (R1) (12)

aWavenumbers of the strongest components of split bands are given in bold.bRelative integrated experimental intensities, normalized in such a way that their sum is equal to the sum of the corresponding calculated intensities.cTheoretical positions of absorption bands were scaled using the empirical formula vscaled = 0.943 · vunscaled + 42.6.dPEDs lower than 10% are not included. Definition of internal coordinates is given in Table S1.47

eDimer.fn.o.: not observed.gn.i.: not investigated.

024505-4 Giuliano et al. J. Chem. Phys. 136, 024505 (2012)

TABLE II. Experimental wavenumbers (v /cm−1) and relative integral intensities (I) of the absorption bands in the infrared spectrum of phenol-d5 isolated inan Ar matrix, compared with wavenumbers (v /cm−1), absolute intensities (Ath / km mol−1) and potential energy distribution (PED, %) calculated for phenol-d5

at the B3LYP/6–311++G(d,p) level.

Experimental Ar matrix Calculated B3LYP/6–311++G(d,p)

v a I b v c Ath Sym. PEDd (%)

3639.1; 3635.2; 3632.6 95.7 3660.0 61.2 A′ ν(O7-H8) (100)3490.5e 13.53460.3e 10.4

23.1

⎧⎪⎪⎪⎪⎨⎪⎪⎪⎪⎩

2291.5; 2277.2 2.4 A′ νa(CD) (93)2287.2; 2269.0 8.1 A′ νb(CD) (79)2282.9; 2281.2; 2277.1; 2256.5 9.4 A′ νd(CD) (80), ν(C2-D9) (11)

2246.0 0.2 A′ νc(CD) (71), νb(CD) (13), ν(C2-D9) (11)2263.9 2234.4 7.4 A′ ν(C2-D9) (71), νc(CD) (17)1592.5; 1591.2 22.0 1566.7 28.0 A′ νc(CC) (66)1574.6; 1573.0; 1572.1; 1570.9; 1567.5 71.0 1554.5 78.0 A′ νb(CC) (68)1471.7; 1473.0 0.91405.0; 1403.4; 1402.5 48.7 1386.3 54.0 A′ νe(CC) (30), δb(CD) (17), νd(CC) (17), ν(C1–O7) (16)1370.6; 1369.7 63.9 1357.7 75.0 A′ νd(CC) (35), δc(CD) (15), νe(CC) (15), δ(O7–H8) (14)1326.0 2.21304.2; 1302.6 20.6 1307.6 15.4 A′ νf(CC) (87)1202.8; 1198.0 6.61191.4; 1188.8; 1187.6 1188.1 2.0 A′ δ(O7-H8) (31), ν(C1-O7) (16), νe(CC) (15)1178.0; 1176.2; 1174.4; 1172.0; 1170.8; 1170.0 156.2 1170.7 187.6 A′ ν(C1-O7) (26), δ(O7-H8) (25)1119.3 3.91019.9 19.2 1018.1 16.0 A′ δa(CD) (70)960.9; 959.7 3.3 958.0 2.7 A′ δ(R1) (59), νa(CC) (38)n.o.f 877.9 0.1 A′ δd(CD) (40), νa(CC) (19), δb(CD) (10)840.0 0.9 848.5 0.8 A′ δe(CD) (48), δc(CD) (17), δa(CD) (14)833.3 1.9 840.8 2.9 A′ δd(CD) (35), δb(CD) (24)814.0; 813.2; 811.0 4.6 822.7 4.7 A′ δc(CD) (45), δe(CD) (30)n.o. 782.7 0.2 A′′ γ (C5-D12) (69), γ b(CD) (22), γ a(CD) (19)n.o. 766.1 1.0 A′′ γ c(CD) (90), γ b(CD) (19)755.5; 754.5 27.9 764.1 18.1 A′ δb(CD) (25), νd(CC) (18), ν(C1-O7) (16), νa(CC) (13)n.o. 737.9 6.2 A′′ γ d(CD) (61), γ (C1-O7) (33)n.o. 646.3 0.4 A′′ γ b(CD) (65), γ (C5-D12) (16), γ c(CD) (13)627.9; 626.5; 625.8 6.5 632.9 7.5 A′′ γ d(CD) (43), τ (R1) (25), γ (C1-O7) (16), γ (C5–D12) (11)n.o. 614.8 0.5 A′ δ(R3) (81)595.5 1.2587.8 10.2 2× τ (O7-H8)551.6; 550.7; 550.0 35.5 565.6 24.2 A′′ τ (R1) (62), γ a(CD) (37)512.3 1.9 535.2 1.7 A′ δ(R2) (76), ν(C1-O7) (10)437.1; 435.4; 433.9; 431.4 24.7 453.3 23.4 A′′ γ a(CD) (31), γ (C1-O7) (30), τ (R2) (27)n.o. 408.5 9.6 A′ δ(C1-O7) (78)n.i.g 387.4 4.4 A′′ τ (R3) (114)n.i. 336.5 107.5 A′′ τ (O7-H8) (96)n.i. 244.8 1.0 A′′ τ (R2) (80), γ (C1-O7) (14), τ (R1) (12)

aWavenumbers of the strongest components of split bands are given in bold.bRelative integrated experimental intensities, normalized in such a way that their sum is equal to the sum of the corresponding calculated intensities.cTheoretical positions of absorption bands were scaled using the empirical formula vscaled = 0.943 · vunscaled + 42.6.dPEDs lower than 10% are not included. Definition of internal coordinates is given in Table S1.47

eDimer.fn.o.: not observed.gn.i.: not investigated.

irradiation at 275 nm (see Figure 3). This provides an evi-dence of generation of several photoproducts.

Identification of one of the photoproducts (trace A inFigure 3), generated upon irradiation of matrix-isolatedphenol with UV λ = 275 nm light, was quite straightforward.Among the new observed IR absorptions, there appeared aset of bands (at 1550, 1515, 1481, 898, 784, and 635 cm−1,to list the strongest) that were previously reported28 as the

spectral signature of the phenoxyl radical (see Figure 4).Photoproduction of an analogous radical (with the strongestinfrared absorptions at 1509, 1467, 774, 660, and 488 cm−1)occurred upon UV λ = 275 nm irradiation of phenol-d5

(Figure 5). Upon 10–15 min of UV λ = 275 nm irradiation,the abundances of the phenoxyl radicals generated fromboth isotopologues of phenol reached their maxima (traceA in Figure 3). For longer irradiation times, the bands due

024505-5 IR spectra and photochemistry of phenol J. Chem. Phys. 136, 024505 (2012)

3700 3600 1600 1400 1200 1000 800 600 4000

50

100

Cal

c. In

tens

ity /

km m

ol−1

Wavenumber / cm −1

(b)

0.0

0.4

0.8 phenolA

bsor

banc

e(a)

3700 3600 1600 1400 1200 1000 800 600 4000

50

100

Cal

c. In

tens

ity /

km m

ol−1

Wavenumber / cm −1

(d)

\\

0.0

0.4

0.8

*

phenol -d5

Abs

orba

nce

(c)

*

FIG. 1. Experimental infrared spectra of (a) phenol and (c) phenol-d5 iso-lated in argon matrices at 15 K, compared with the theoretical spectra of(b) phenol and (d) phenol-d5 calculated at the B3LYP/6-311++G(d,p) level.The calculated wavenumbers were scaled using the empirical formula νscaled= 0.943 × νunscaled + 42.6 obtained from the least squares linear fit. Aster-isks in frame (c) designate absorptions due to traces of matrix-isolated H2Omonomers.

to phenoxyl and phenoxyl-d5 radicals (Tables S2 and S3)47

gradually decreased.Appearance of phenoxyl radical products clearly shows

that the UV-induced hydrogen-atom photodetachment pro-cess is effective for phenol molecules isolated in low-temperature matrices. The mechanism of the phenoxyl radi-cal formation, by photoinduced hydrogen-atom detachment,has recently been the subject of extensive state-of-the-artcalculations of the excited-state potential energy surfaces ofphenol.6, 29, 30 Repulsive πσ* excited states were postulated toplay a crucial role in the hydrogen-atom-detachment process.4

Photoinduced cleavage of the OH bond was observed in anumber of experimental studies5, 30, 31 on isolated moleculesof gaseous phenol. Hydrogen atoms with large kinetic energy,detached from the phenol precursors, were detected in thesestudies. For phenol molecules isolated in solid Ar, the largekinetic energy must allow some non-negligible part of pho-todetached hydrogen atoms to leave the matrix cage wherethey are formed.

Analysis of the IR spectra recorded after UV irradiationof matrix-isolated phenol suggests that the hydrogen-atomdetachment from the OH group is the primary act of allphotoinduced transformations observed for the compound.If the cleaved H-atom leaves the matrix cage, the phenoxylradical is produced. If the H-atom stays in the cage and

O

H

H

H

H

H

H

3100 2800 2500 22000

8

16

Cal

c. In

tens

ity /

km m

ol −

1

Wavenumber / cm −1

(b)

0.00

0.02

0.04

Abs

orba

nce

(a)phenol

O

D

D

D

D

D

H

3100 2800 2500 22000

8

16

Cal

c. In

tens

ity /

km m

ol −

1

Wavenumber / cm −1

(d)

0.00

0.01

0.02

* A

bsor

banc

e

(c)phenol -d5

FIG. 2. Experimental FTIR spectra, in the region of CH/CD stretching vi-brations, of (a) phenol and (c) phenol-d5 isolated in argon matrices at 15 Kcompared with the theoretical spectra of (b) phenol and (d) phenol-d5 calcu-lated at the B3LYP/6-311++G(d,p) level. The calculated wavenumbers werescaled as commented in Figure 1. Asterisk in frame (c) designates the absorp-tion due to traces of matrix-isolated CO2 monomers. Structures of phenol andphenol-d5 are shown in frames (b) and (d), respectively.

recombines with the remaining radical, either phenol canbe repopulated or photoproducts bearing a C=O groupmay be generated (Scheme 1). The carbonyl-containingstructures can be formed when the H-atom recombines withthe phenoxyl radical at ortho-, meta-, and para- positionsof the ring, which is equivalent to [1,3]-, [1,4]-, and [1,5]-hydrogen atom shifts, respectively (Scheme 1).

The spectral positions of IR bands due to the stretchingvibrations of C=O groups (νC=O), observed in the spectra ofthe products, allow a reasonable pre-selection of the putativephotoproduct structures. In the IR spectra recorded after UVλ = 275 nm irradiation, a doublet of intense absorptions wasfound at 1681.6 and 1677.6 cm−1. These bands (correspond-ing to the photoproduct designated by B in Figure 3) are most

024505-6 Giuliano et al. J. Chem. Phys. 136, 024505 (2012)

0 20 40 60 800

5

10

15

0 20 40 60 800

20

40

60

80

100

(C)

(B)

(A)

hν = 275 nm

Time / min

Abu

ndan

ce, %

(phenol)

FIG. 3. Main frame: Abundance of matrix-isolated phenol monomers as afunction of UV (λ = 275 nm) irradiation time. The integral intensities of threeabsorption bands (3638/3634, 1504/1501, and 753 cm−1) of phenol were: (i)divided by their calculated infrared intensities; (ii) averaged; (iii) normalizedto 100% for the spectrum prior to irradiation (0 min) in that order. Inset:Abundances of three photoproducts deduced from the experimental bands784 cm−1 (A), 1681.6/1677.6 cm−1 (B), and 1789 cm−1 (C) as functions ofUV (λ = 275 nm) irradiation time. Photoproducts (A, B, C) were assignedto phenoxyl radical (2), 2,5-cyclohexadienone (5), and Dewar isomer (7), re-spectively. The abundances of photoproducts (A, B, C) were normalized us-ing the same procedure and normalization factor as that obtained for phenol.

probably due to the stretching vibrations of a C=O groupattached to a six-membered ring. Hence, cyclohexadienonespecies such as 2,4-cyclohexadienone (3 in Scheme 1) or2,5-cyclohexadienone (5), are good candidates for the photo-product B. In principle, the observed doublet can be a result ofmatrix splitting, anharmonic interactions, or an indication ofthe presence of both isomers, 2,4- and 2,5-cyclohexadienone.The unequivocal identification of the photoproduct B provedto be possible, based on its unique kinetics in the course of UVλ = 275 nm irradiations (Figure 3, trace B). When the phe-nol precursor was depleted in the matrix, product Bstopped growing; however, it was not consumed by the UVλ = 275 nm irradiation unlike the other products. At thelater stages of this experiment (vide infra), it was foundthat this product can be effectively destroyed by the UVλ = 245 nm irradiation. Thanks to these effects, separation ofits IR spectrum was possible (Figure 6(b)). This experimentalspectrum is well reproduced by the spectrum theoreti-cally predicted for 2,5-cyclohexadienone (5, Figure 6(c),Table S2 (Ref. 47)). On the contrary, no agreement was foundbetween the experimental spectrum of the photoproductand the spectrum calculated for 2,4-cyclohexadienone (3,Figure 6(a)), especially in the range below 1000 cm−1 (seeFigure 6).32 Therefore, the photoproduct characterized bythe pair of bands at 1681.6 and 1677.6 cm−1 (the doubletstructure is likely to appear due to matrix site-effect) can besafely assigned to 2,5-cyclohexadienone.33 This conclusionis further supported by an analogous analysis carried out for

1550 1500 1450 800 750 700 650

0

12

0.00

0.05

0.0

0.3

Inte

nsity

/ar

b. u

nits

(c)

Wavenumber / cm−1

Abs

orba

nce (b)

Abs

orba

nce (a)

FIG. 4. Selected regions of the IR spectrum of phenol in an argon matrix:(a) recorded before irradiation, (b) recorded after irradiation at 275 nm (withnew bands of the photoproducts) compared with (c) graphical presentation ofthe experimental IR bands of phenoxyl radical (2) isolated in an argon matrix,reported in Ref. 28, convoluted with Lorentzian profiles, (FWHM = 2 cm−1).

the corresponding photoproduct generated from phenol-d5

(Figures 6(d) and 6(f), Table S3 (Ref. 47)).Other characteristic band that emerged upon UV

λ = 275 nm irradiation of matrix-isolated phenol was foundat 1789.1 cm−1 (at 1787.8 cm−1 for phenol-d5). Bands atsuch frequencies usually indicate structures with a C=Ogroup attached to a four-membered ring.34, 35 In the currentlyconsidered case of photochemical transformations of phenol,

1500 1400 800 700 600 500

0

10

0.00

0.04

0.0

0.2

Inte

nsity

/ar

b. u

nits

Wavenumber / cm−1

Abs

orba

nce

Abs

orba

nce (a)

(b)

(c)

FIG. 5. Selected regions of the IR spectrum of phenol-d5 in an argon matrix:(a) recorded before irradiation, (b) recorded after irradiation at 275 nm (withnew bands of the photoproducts), compared with (c) graphical presentation ofthe experimental IR bands of phenoxyl-d5 radical isolated in an argon matrix,reported in Ref. 28, convoluted with Lorentzian profiles, (FWHM = 2 cm−1).

024505-7 IR spectra and photochemistry of phenol J. Chem. Phys. 136, 024505 (2012)

O H

O

HH

O

H H

]5,1[]3,1[ [1,4]

OH

H

O

−H

1 2

3 4 5

OHH

7

OHH

6

SCHEME 1. Pathways of isomerizations induced by UV (λ = 275 nm) irra-diation of matrix-isolated phenol considered in the current study. Structuresin brackets were not observed.

this band is most probably the spectral signature of the Dewarisomer 7 (Scheme 1). The band observed at 1789.1 cm−1

matches closely the calculated frequency of 7 (1788 cm−1,see Figure 7) and has a very high infrared intensity (calcu-lated value 384 km mol−1). Unfortunately, the vibrationalsignature of the Dewar form cannot be confirmed by otherbands. The only other characteristic intense IR band due tothe Dewar product, with a calculated frequency of 767 cm−1

(and infrared intensity of 76 km mol−1; see Tables S2 andS3 (Ref. 47) for full calculated IR spectra of relevant photo-produced species), is with all probability overlapped with thephenol band at 752 cm−1. However, non-observation of theother bands due to the Dewar isomer 7 is, in a certain sense,consistent with its predicted infrared spectrum (Tables S2,S3),47 where the νC=O band largely dominates and 28 outof 33 vibrational modes have calculated infrared intensities< 20 km mol−1 (i.e., almost 20 times less intense thanνC=O).

The most plausible precursor of 7 should be2,4-cyclohexadienone (3 in Scheme 1); however, thereis no direct experimental evidence for generation of 3 inthe present study. This is not unexpected, since α-pyrones(electronically equivalent with 2,4-cyclohexadienones) arereadily converted into their Dewar isomers or/and open-ringconjugated ketenes under similar UV-irradiation conditions,36

as shown in Scheme 2. Indeed, in the spectra of phenol andphenol-d5 isolated in Ar matrices and irradiated with UVλ = 275 nm light (Figure 7), IR bands characteristic of ketenespecies were also clearly seen in the 2100 cm−1 region. Suchbands correspond to the “antisymmetric” vibrations of theketene –C=C=O group.35, 37 The most probable carriers ofthese bands, split in many components, are the various possi-ble isomeric forms of the conjugated ketene 6 (Scheme 1, seealso Figure 8). Each of the components of the multiplet band

1600 1400 1200 1000 800 600

0

20

0.0

0.2

0

20

(5) (c)

Inte

nsity

Wavenumber / cm −1

(B) (b)

Abs

(x1

0)

(3)

Inte

nsity

(a)

1600 1400 1200 1000 800 600

0

20

0.0

0.2

(f)

(e)In

tens

ity

Wavenumber / cm −1

(B) -d5

∗

Abs

(x 1

0)

(d)

∗

0

20

Inte

nsity

(5) -d5

(3) -d5

FIG. 6. (a) Simulated IR spectrum of 2,4-cyclohexadienone (3); (b) extractedexperimental spectrum of the photoproduct (B); (c) simulated IR spectrumof 2,5-cyclohexadienone (5); (d)–(f) spectra for the pentadeuterated com-pounds analogous to (a, b, c), respectively. Asterisks in frame (e) indicatenon-compensated bands due to the presence of phenoxyl-d5 radical.

at 2150–2110 cm−1 shows a different behavior (Figure 9)at different stages of irradiation at λ = 275 nm. One of thecomponents (2136 cm−1) increases monotonously and can beassigned to the most stable rotamer E. Other components ofthe multiplet initially increase but then decrease during theirradiation, suggesting photoproduction of different keteneisomers and their conversion into the most stable form. Theseconformational transformations in the conjugated ketenephotoproduced from phenol correlate well with analogoustransformations of the conjugated ketene produced fromα-pyrone.35, 37

O

O

O

O

O

O

hν hν

SCHEME 2. Experimentally observed photochemical transformations ofα-pyrone, electronically equivalent with 2,4-cyclohexadienone.

024505-8 Giuliano et al. J. Chem. Phys. 136, 024505 (2012)

phenol

Abs

orba

nce

2200 2000 1800 1600

(1)

(2)(6)

(7) (5)

Rel

ativ

e In

tens

ity

Wavenumber / cm−1

(a)

(b)(6)

phenol -d5

*

Abs

orba

nce

2200 2000 1800 1600

Rel

ativ

e In

tens

ity

Wavenumber / cm−1

(c)

(d)

(1)

(2)(6)

(7) (5)(6)

FIG. 7. Difference experimental IR spectra, obtained by subtraction of thespectra recorded after the matrix deposition from the spectra recorded afterUV λ = 275 nm irradiation of matrix-isolated (a) phenol and (c) phenol-d5, compared with the spectra (b) and (d) simulated for the photoproductstructures [phenoxyl radical (2), 2,5-cyclohexadienone (5), open-ring ketene(6), and Dewar isomer (7)]. The simulated spectra of phenol and phenol-d5 reactants (1) are presented in frames (b) and (d) as negative, truncatedbands. In the theoretical spectra, the calculated intensities were weighted as[(2) × 50 + (5) × 15 + (7) × 15 + (6) × 20] − [(1) × 100]. The open-ringketene (6) was represented as a sum of its most stable four rotamers withequal weights (5 parts out of 20). The photoproduct band designated withasterisk in frame (c) is commented in the text.

Hydroxy-oxo equilibrium in phenol

The tautomeric equilibrium in phenol, between its moststable hydroxy form and the less stable oxo form, has been thesubject of many studies.7, 26, 38, 39 In the present work, the ob-servation of photoproduced ketene (6) and Dewar isomer (7)suggests that these two species are formed via intermediacy of

O

O

Aa (30.7) E (0.0b)

O O

B (14.0) F (7.0)

O

O

C (5.1) G (22.0)

OO

D (19.8) H (14.8)

FIG. 8. Structures and DFT(B3LYP)/6-311++G(d,p) calculated relative en-ergies (corrected for zero-point vibrational energy) of the open-ring ketenerotamers (6 in Scheme 1 and Figure 7). a: Chosen rotamer designation; b:relative energies, kJ mol−1. Absolute calculated energy of the most stablerotamer E is −307.398408 hartree.

2,4-cyclohexadienone (3). The latter product can be formedfrom phenol via intramolecular [1,3] hydrogen atom shift,and in principle, this reaction might be reversible. Experi-ments with pentadeuterated isotopomer phenol-d5 allowed usto verify this hypothesis. If the reaction between 1 and 3in the ring-perdeuterated compound was reversible, and the[1,3] migration of hydrogen was random, it would result inthe deuteration of the hydroxyl group (see Scheme 3), espe-cially taking into account that the deuteration of the phenolring in the present study was almost complete (see Figure 2).However, the irradiations of the matrix at λ = 275 nm didnot result in formation of any photoproduct absorbing around2660 cm−1, at a frequency where the OD stretch-ing vibration of phenol is expected.13 This means that2,4-cyclohexadienone (3), if formed from phenol under thepresent UV-irradiation conditions, is very reactive towards(6) or/and (7); it does not accumulate in the matrix and theputative channel leading to the back-formation to phenol isnot competitive. This result is in agreement with the heightsof barriers separating 2,4-cyclohexadienone from phenol and

024505-9 IR spectra and photochemistry of phenol J. Chem. Phys. 136, 024505 (2012)

2130 2120 2110 2100

0

200

400(A-H)

(6)

Inte

nsity

Wavenumber / cm −1

CE D

HFG

AB

(b)

52 min24 min10 min 4 min 0 min

2140 2120 2100

0.0

0.1

0.2

hν = 275 nmA

bsor

banc

e(a)

FIG. 9. (a) Evolution of the band profile due to the “antisymmetric” vibra-tions of the –C=C=O group of ketene (6) in the course of UV irradiation(λ = 275 nm) of matrix isolated phenol. Irradiation times are increasing frombottom (0 min) to top (52 min); (b) simulated spectra for the eight open-ringisomers of ketene (6) (see Figure 8).

ketene, which were theoretically calculated to be 214 and159 kJ mol−1, respectively.39

Irradiation at shorter wavelengths

After prolonged irradiation at λ = 275 nm, new absorp-tion features started to appear in the IR spectrum, indicat-ing generation of the next group of photoproducts. Amongthese photoproducts, carbon monoxide and cyclopentadienecan be reliably identified. The band of CO appears around2138 cm−1, while the most characteristic bands due to cy-clopentadiene (at 894 and 666/663 cm−1) could be directly as-signed by comparison with the positions of bands found in theoriginal experimental spectrum of cyclopentadiene isolatedin an Ar matrix.40 The formation of these two compoundsin the irradiated matrix (prolonged λ = 275 nm irradiation)can be a consequence of photofragmentation of the transientbicyclo[3.1.0]hex-2-en-6-one (structure (4) in Scheme 1).41

O

D

D

D

D

D

HO

D

D

D

D

D H

O

H

D

D

D

D

D

SCHEME 3. A hypothetical mechanism of deuteration of the hydroxyl groupin phenol-d5 via the hydroxy-oxo equilibrium (the OD-isomer was not gen-erated in the present study in the course of UV irradiations at λ = 275 nm).

O

H H

++++ CO ++++hν

245 nmH H

H H

H

H

SCHEME 4. UV-irradiation at λ = 245 nm results in decomposition of2,5-cyclohexadienone into allene, carbon monoxide and acetylene.

UV irradiation at wavelengths shorter than 275 nm re-sulted in decomposition of all the photogenerated C6H6O(C6D5HO) isomers into smaller species. Different decompo-sition characteristics were observed for different products.2,5-Cyclohexadienone (5) appeared to be remarkably stableagainst UV λ = 275 nm irradiation, but upon irradiation atλ = 245 nm it was readily consumed until total disappear-ance. The depletion of (5) at λ = 245 nm was accompanied byincrease of IR bands ascribable to carbon monoxide, allene,42

and acetylene43 (possibly in associated forms).44 Scheme 4summarizes the result of decomposition of (5).

A more energetic UV λ = 235 nm light had to be appliedto decompose the Dewar isomer (7). Among the photoprod-ucts generated upon UV λ = 275 nm irradiation, the phenoxylradical (2) was partially consumed at any of the irradiationsapplied; however, the residual IR bands due to this specieswere still present in the spectrum recorded after irradiation atλ = 220 nm (see Figure 10).

Some of the products appearing upon irradiation of theC6H6O system could be assigned only with the aid of theexperiments carried out on the C6D5HO counterpart. Thisconcerns, for example, the carrier of a relatively intenseband at 1538 cm−1 in phenol-d5 (designated with asterisk inFigure 7(c)). Most probably, this absorption corresponds tothe C=C stretching vibration of the perdeuterated cyclobu-tadiene (C4D4). Bands appearing at similar frequencies were

1800 1790 1780 1690 1680 1670

0.0

0.2

0.4

0.6

0.8

(5)

Wavenumber / cm −1

Abs

orba

nce

( x10

)

(7)

1550 1540

0.0

0.1

0.2(2)

(d)

(c)

(b)

(a)

FIG. 10. Evolution of bands due to the Dewar isomer (7),2,5-cyclohexadienone (5), and phenoxyl radical (2) in the course ofUV irradiations: after (a) 90 min of irradiation at λ = 275 nm, (b) plus 30min of irradiation at shorter wavelengths down to λ = 245 nm, (c) plus7 min of irradiation at shorter wavelengths down to λ = 235 nm, and (d)plus 11 min of irradiation at shorter wavelengths down to λ = 220 nm. Themost intense transitions showed are, respectively, the C=O stretching (1789cm−1) of Dewar isomer, the C=O stretching doublet (1682, 1678 cm−1)of 2,5-cyclohexadienone, and the CC stretching (1550 cm−1) of phenoxylradical.

024505-10 Giuliano et al. J. Chem. Phys. 136, 024505 (2012)

O

H D

OHD ++++

hν

D

D

D

D

++++

O

H

DO

D

H

C6D5HOC4D4

SCHEME 5. Decomposition mechanism of the Dewar isomer, deduced fromthe experiments with the C6D5HO isotopologue. The Dewar isomer yieldscyclobutadiene C4D4 and half-deuterated ketene (OCCHD), with subsequentdecomposition of C4D4 into two acetylenes and a possible isomerization ofOCCHD to ethynol via H- or D-shifts.

reported for partially deuterated cyclobutadienes isolated inAr matrices.45 The photoproduct bands at 883, 860, and734 cm−1 should also be assigned to C4D4. Upon irradiationat λ = 245 nm, these bands were disappearing at the samerate as the band at 1538 cm−1. The likely precursor of C4D4

should be the C6D5HO Dewar isomer, which can decom-pose with formation of the parent ketene (O=C=CHD) as aby-product (Scheme 5). A similar reaction generating C4H4

and CO2 was experimentally observed for matrix-isolatedα-pyrone.35, 45 In the present study, formation of the parentketene would be difficult to prove because of the overlapof the most intense band due to the C=C=O antisymmetricstretching vibration with the similar absorptions of (6). How-ever, appearance of the new absorptions around 640/629 cm−1

in the experiments with C6H6O may be treated as an evidenceof formation of associated O=C=CH2.

As the energy of the incident UV light increases, moreand more reagent molecules become depleted, while thehydrogen-shifts and bond cleavages between heavy atoms inthe photoproducts start to kick in. The matrices become over-populated with fragmented species, in the associated form,and their assignments turn ever harder. In spite of this, anew absorption that starts to develop around 2612 cm−1 uponirradiations at λ = 245 nm and lower must have an obvi-ous assignment, as the OD stretching vibration. In the frag-mented deuterated photoproducts, a possible carrier of thisband can be ethynol (HCCOD) isomeric of half-deuteratedketene (Scheme 5).46

CONCLUSIONS

The present work reports detailed infrared spectra of phe-nol and its perdeuterated phenol-d5 analogue isolated in argonmatrices at 15 K.

The narrowband tunable laser technique has been used tostudy the photochemistry of monomers of phenol and phenol-d5 isolated in low-temperature Ar matrices. Products gener-ated upon UV irradiation of matrix-isolated phenol (1) havebeen characterized by IR spectroscopy. On the basis of theanalysis of the collected data, H-atom detachment from the

OH group of the compound was shown to be the primary pho-tochemical act following the UV excitation. Phenoxyl radi-cal (2) and 2,5-cyclohexadienone (5) have been identified asproducts of unimolecular λ = 275 nm photochemistry of phe-nol. Spectral indications of photogeneration of an open-ringconjugated ketene (6) as well as Dewar isomer (7) were alsofound in the spectra recorded after UV λ = 275 nm irradi-ation. The photoproduced open-ring conjugated ketene ap-peared in the matrix in a variety of conformational structures.Upon shorter-wavelength UV excitation, further photoprod-ucts such as cyclopentadiene, carbon monoxide, allene, acety-lene, cyclobutadiene, and parent ketene were generated. Noindication of photochemical channels leading to a pre-fulveneisomer was detected.

The above conclusions were supported by the in-vestigation on UV-induced transformations of matrix-isolated pentadeuterated phenol-d5. Experiments on phenol-d5 showed also that ketonization of phenol (1) into its 2,4-cyclohexadienone isomer (3), if it occurs under the appliedirradiation conditions, is accompanied by the simultaneoustransformation of (3) into ketene (6) and Dewar (7) species.

ACKNOWLEDGMENTS

This work was supported by the Portuguese “Fun-dação para a Ciência e a Tecnologia” (FCT), ResearchProjects PTDC/QUI-QUI/111879/2009 and PTDC/QUI-QUI/118078/2010. B.M.G. acknowledges FCT for the post-doctoral Grant No. SFRH/BPD/44689/2008.

1A. D. McLaren and D. Shugar, Photochemistry of Proteins and Nu-cleic Acids (Pergamon, London, 1964); R. P. Rastogi, Richa, A. Kumar,M. B. Tyagi, and R. P. Sinha, J. Nucleic Acids 2010, 592980 (2010).

2C. Ratzer, J. Küpper, D. Spangenberg, and M. Schmitt, Chem. Phys. 283,153 (2002).

3M. G. D. Nix, A. L. Devine, B. Cronin, R. N. Dixon, and M. N. R. Ashfold,J. Chem. Phys. 125, 133318 (2006).

4M. N. R. Ashfold, A. L. Devine, R. N. Dixon, G. A. King, M. G. D. Nix,and T. A. A. Oliver, Proc. Natl. Acad. Sci. U.S.A. 105, 12701 (2008).

5C.-M. Tseng, Y. T. Lee, M.-F. Lin, C.-K. Ni, S.-Y. Liu, Y.-P. Lee, Z. F. Xu,and M. C. Lin, J. Phys. Chem. A 111, 9463 (2007).

6O. P. J. Vieuxmaire, Z. Lan, A. L. Sobolewski, and W. Domcke, J. Chem.Phys. 129, 224307 (2008).

7L. Zhu and J. W. Bozzelli, J. Phys. Chem. A 107, 3696 (2003).8Z. F. Xu and M. C. Lin, J. Phys. Chem. A 110, 1672 (2006).9A. Iqbal, L.-J. Pegg, and V. G. Stavros, J. Phys. Chem. A 112, 9531 (2008).

10S. Kato, J. Chem. Phys. 88, 3045 (1988); A. L. Sobolewski, C.Woywod, and W. Domcke, ibid. 98, 5627 (1993); I. J. Palmer, I. N. Raga-zos, F. Bernardi, M. Olivucci, and M. A. Robb, J. Am. Chem. Soc. 115, 673(1993); W. Domcke, A. L. Sobolewski, and C. Woywod, Chem. Phys. Lett.203, 220 (1993); B. Lasorne, M. J. Bearpark, M. A. Robb, and G. A. Worth,J. Phys. Chem. A 112, 13017 (2008).

11J. Dreyer and M. Klessinger, Chem.-Eur. J. 2, 335 (1996).12H. D. Bist, J. C. D. Brand, and D. R. Williams, J. Mol. Spectrosc. 24, 402

(1967); D. Michalska, W. Zierkiewicz, D. C. Bienko, W. Wojciechowski,and T. Zeegers-Huyskens, J. Phys. Chem. A 105, 8734 (2001); H. Lampert,W. Mikenda, and A. Karpfen, ibid. 101, 2254 (1997); W. Roth, P. Imhof,M. Gerhards, S. Schumm, and K. Kleinermanns, Chem. Phys. 252, 247(2000).

13G. Keresztury, F. Billes, M. Kubinyi, and T. Sundius, J. Phys. Chem. A102, 1371 (1998).

14F. Billes and I. Mohammed-Ziegler, Appl. Spectrosc. Rev. 42, 369 (2007).15A few studies involving the matrix isolated phenol do exist, but the main

focus of these studies is devoted to the changes in the OH stretching fre-quency of phenol acting as a hydrogen bond donor/hydrogen bond acceptorin different complexes (see, for example, Ref. 16).

024505-11 IR spectra and photochemistry of phenol J. Chem. Phys. 136, 024505 (2012)

16K. Sundararajan and N. Ramanathan, J. Mol. Struct. 920, 369 (2009);J. Gebicki and A. Krantz, J. Am. Chem. Soc. 106, 8097 (1984); J. Gebickiand A. Krantz, ibid. 106, 8093 (1984).

17A. Andrzejewska, L. Lapinski, I. Reva, and R. Fausto, Phys. Chem. Chem.Phys. 4, 3289 (2002).

18A. D. Becke, J. Chem. Phys. 98, 5648 (1993).19C. T. Lee, W. T. Yang, and R. G. Parr, Phys. Rev. B 37, 785 (1988).20S. H. Vosko, L. Wilk, and M. Nusair, Can. J. Phys. 58, 1200 (1980).21M. J. Frisch, G. W. Trucks, H. B. Schlegel et al., GAUSSIAN 03, Revision

C.02„ Gaussian, Inc., Wallingford CT, 2004.22J. H. Schachtschneider and F. S. Mortimer, Vibrational Analysis of Poly-

atomic Molecules. VI. FORTRAN IV Programs for Solving the VibrationalSecular Equation and for the Least-Squares Refinement of Force Constants.Project No. 31450. Structural Interpretation of Spectra, Technical ReportNo. 57-65, Emeryville, CA, 1969.

23P. Pulay, G. Fogarasi, F. Pang, and J. E. Boggs, J. Am. Chem. Soc. 101,2550 (1979).

24K. K. Irikura, SYNSPEC, National Institute of Standards and Technology,Gaithersburg, MD, USA, 1995.

25A. M. Plokhotnichenko, E. D. Radchenko, Yu. P. Blagoi, andV. A. Karachevtsev, Low Temp. Phys. 27, 666 (2001).

26J. K. Parker and S. R. Davis, J. Am. Chem. Soc. 121, 4271 (1999).27G. Berden, W. L. Meerts, M. Schmitt, and K. Kleinermanns, J. Chem. Phys.

104, 972 (1996); B. Trost, J. Stutz, and U. Platt, Atmos. Environ. 31, 3999(1997).

28J. Spanget-Larsen, M. Gil, A. Gorski, D. M. Blake, J. Waluk, andJ. G. Radziszewski, J. Am. Chem. Soc. 123, 11253 (2001).

29A. L. Sobolewski, W. Domcke, C. Dedonder-Lardeux, and C. Jouvet, Phys.Chem. Chem. Phys. 4, 1093 (2002); M. Abe, Y. Ohtsuki, Y. Fujimura,Z. Lan, and W. Domcke, J. Chem. Phys. 124, 224316 (2006); R. N. Dixon,T. A. A. Oliver, and M. N. R. Ashfold, ibid. 134, 194303 (2011).

30M. N. R. Ashfold, G. A. King, D. Murdock, M. G. D. Nix, T. A. A. Oliver,and A. G. Sage, Phys. Chem. Chem. Phys. 12, 1218 (2010).

31C.-M. Tseng, Y. T. Lee, and C.-K. Ni, J. Chem. Phys. 121, 2459(2004).

32In the study of Parker and Davis (Ref. 26), the C=O stretching vibration of2,4-cyclohexadienone was assigned to a band appearing at 1728.6 cm−1.No photoproduct having absorption around this frequency was observed inthe present study.

33A doublet at similar frequencies, 1683.2/1679.0 cm−1 was observed in thereaction of oxygen atoms with benzene at λ > 280 nm, in the study ofParker and Davis (Ref. 26). In that study, the carrier of this doublet bandremained unassigned.

34S. Breda, L. Lapinski, I. Reva, and R. Fausto, J. Photochem. Photobiol. A162, 139 (2004).

35S. Breda, I. Reva, L. Lapinski, and R. Fausto, Phys. Chem. Chem. Phys. 6,929 (2004).

36I. D. Reva, M. J. Nowak, L. Lapinski, and R. Fausto, Chem. Phys. Lett.429, 382 (2006); I. Reva, M. J. Nowak, L. Lapinski, and R. Fausto, ibid.452, 20 (2008).

37S. Breda, I. Reva, L. Lapinski, M. L. S. Cristiano, L. Frija, and R. Fausto,J. Phys. Chem. A 110, 6415 (2006).

38P. Mulder, H.-G. Korth, D. A. Pratt, G. A. DiLabio, L. Valgimigli,G. F. Pedulli, and K. U. Ingold, J. Phys. Chem. A 109, 2647 (2005);I. Gómez, E. Rodríguez, and M. Reguero, J. Mol. Struct.: THEOCHEM767, 11 (2006); B. Osmialowski, E. D. Raczynska, and T. M. Krygowski,J. Org. Chem. 71, 3727 (2006); C. S. Shiner, P. E. Vorndam, and S. R. Kass,J. Am. Chem. Soc. 108, 5699 (1986).

39C. A. Taatjes, D. L. Osborn, T. M. Selby, G. Meloni, A. J. Trevitt, E.Epifanovsky, A. I. Krylov, B. Sirjean, E. Dames, and H. Wang, J. Phys.Chem. A 114, 3355 (2010).

40J. Miyazaki and Y. Yamada, J. Mol. Struct. 692, 145 (2004).41Other C6H5OH isomers do not yield cyclopentadiene upon decomposition,

as deduced from the experiments using irradiations at shorter wavelengthsof UV light (see below).

42For example, a very characteristic band around 1952 cm−1 due to the anti-symmetric C=C=C stretching vibration.

43For example, a very characteristic band around 3330 cm−1 due to the anti-symmetric CH stretching vibration.

44In particular, cyclopentadiene does not look to be among the products gen-erated from 2,5-cyclohexadienone (5). Decrease of population of (5), in-duced by irradiation at λ = 245 nm, was not accompanied by any increaseof IR bands ascribed to cyclopentadiene.

45B. R. Arnold and J. Michl, J. Phys. Chem. 97, 13348 (1993).46Interestingly, the present calculations predict the CD stretching frequency

of the other isotopomer of ethynol (DCCOH) to be only ∼22 cm−1 belowthe OD stretching frequency of HCCOD.

47See supplementary material at http://dx.doi.org/10.1063/1.3666018 for alist of internal coordinates used in the normal mode analysis for phenol andphenol-d5 (Table S1), for the observed bands of phenoxyl radical (2), 2,5-cyclohexadienone (5), Dewar isomer (7), and cyclopentadiene (8) and theirproposed assignments (Table S2); for the observed bands of d5-analogsof (2, 5, 7, 8) and their proposed assignments (Table S3). The originalexperimental spectra related to Figures 1(a), 1(c), 2(a), 2(c), 4(b), 5(b),6(b), and 6(e) are available in numeric format (*.SPA files for OMNIC

software).