Photophysical and photochemical processes of 2-methyl, 2-ethyl, and 2-tert-butylanthracenes on silica gel. A substituent effect study

Article abstract of DOI:10.1021/jp0022555

PAH are organic pollutants formed and released into the environment by the incomplete combustion of fossil fuel, coal, wood, gas, oil, garbage, tobacco, and charbroiled meat. The photophysics and photochemistry of 2-methylanthracene (2MA), 2-ethylanthracene (2EA), and 2-tert-butylanthracene (2TBA) adsorbed on silica were studied at a silica/air interface. 2MA, 2EA, and 2TBA showed no evidence of ground state pairing even at high surface coverages, and crystallized on the surface at higher surface coverages. The photolysis of 2MA, 2EA, and 2TBA at a silica/air interface proceeded more efficiently than the photolysis of anthracene, to produce the corresponding 9,10-endoperoxides formed by the addition of singlet molecular oxygen (type II) to the ground state molecules. At low surface coverages, no evidence of any dimer was observed in photolyzed 2MA, 2EA, and 2TBA samples. A small amount of dimers (isomeric) were observed at higher surface coverages, suggesting that the crystal forms of these molecules may be involved in the dimerization process. The photolysis rate decreased with increasing surface coverage, due to an inner effect induced by crystal formation. Photolysis rate decreased by an order of magnitude in the presence of < 1 monolayer of physisorbed water. Time-resolved transient studies of excited triplet states of 2MA, 2EA, and 2TBA showed that triplet lifetimes were shortened on wet silica. The efficiency of singlet molecular oxygen formation significantly decreased on wet silica. These results indicated that decreased photolysis rate was caused by reduced singlet quantum yield in the presence of physisorbed water.

Full text of DOI:10.1021/jp0022555

J. Phys. Chem. B 2000, 104, 10235-10241
10235
Photophysical and Photochemical Processes of 2-Methyl, 2-Ethyl, and
2-tert-Butylanthracenes on Silica Gel. A Substituent Effect Study
Reza Dabestani,* Jashua Higgin, Daniel Stephenson, Ilia N. Ivanov, and Michael E. Sigman
Chemical and Analytical Sciences DiVision, Oak Ridge National Laboratory, P.O. Box 2008,
MS-6100, Oak Ridge, Tennessee 37831-6100
ReceiVed: June 21, 2000; In Final Form: September 5, 2000
The photophysics and photochemistry of 2-methylanthracene (2MA), 2-ethylanthracene (2EA), and 2-tert-
butylanthracene (2TBA) adsorbed on silica was studied at a silica/air interface to determine the substituent
effect on the above processes. In contrast to anthracene (AN), which forms ground state stable pairs at surface
coverages as low as 1% of a monolayer, 2MA, 2EA, and 2TBA show no evidence of any ground state pairing
even at high surface coverages. Diffuse reflectance and fluorescence data indicate that 2MA, 2EA, and 2TBA
crystallize on the surface at higher surface coverages. Photolysis of 2MA, 2EA, and 2TBA at a silica/air
interface proceeds more efficiently than photolysis of AN to produce the corresponding 9,10-endoperoxides
formed by the addition of singlet molecular oxygen (type II) to the ground state molecules. Thermally unstable
endoperoxides slowly decompose on silica surface to give the corresponding 9,10-quinones and other
hydroxylated products of 2MA, 2EA, and 2TBA. Although photolyzed AN/silica samples show a significant
amount of dimer formation at low surface coverages, no evidence of any dimer is observed in photolyzed
samples of 2MA, 2EA, and 2TBA at low surface coverages. Small amount of dimers (isomeric), however,
are observed in the photolyzed samples of 2MA, 2EA, and 2TBA at higher surface coverages suggesting that
the crystal forms of these molecules may be involved in the dimerization process. The photolysis rate decreases
as the surface coverage is increased. This behavior can be attributed to the involvement of crystals (formed
at higher surface coverage) which act as a light sink to absorb the incident light without producing any net
chemistry. The photolysis rate decreases by an order of magnitude in the presence of e1 monolayer of
physisorbed water. Time-resolved transient studies of excited triplet states of 2MA, 2EA, and 2TBA have
revealed that triplet lifetimes are shortened on wet silica. Furthermore, the efficiency of singlet molecular
oxygen formation drops significantly on wet silica. These results suggest that the decrease in photolysis rate
is due to lowering of the singlet oxygen quantum yield in the presence of physisorbed water.
Introduction
studied the photochemical behavior of several PAHs14-22 on
this list.14,15,17,20-22 The main emphasis of these studies has been
Polycyclic aromatic hydrocarbons (PAH) comprise a large
class of organic pollutants that are formed and released into
the environment by incomplete combustion of fossil fuel, coal,
wood, gas, oil, garbage, tobacco, and charbroiled meat. Because
of their resistance to biodegradation (especially PAHs with four
and more condensed rings) they are very persistent in the
environment. Since the environmental fate of PAHs (when
present in water, adsorbed on soil or air particulates and exposed
to sunlight over an extended period of time) could have a
significant impact on their remediation processes, our research
in the past few years has focused on the photophysical and
photochemical behavior of these compounds under environ-
mentally relevant conditions. Light-induced processes have been
claimed to enhance transformation of these anthropogenic
pollutants in the environment.1,2 Although the photophysics and
to some degree the photochemistry of a few PAHs adsorbed
on silica3-7 or other inorganic metal oxides8 have been reported,
detailed studies on the nature and identities of photoproducts
formed during phototransformation of these compounds on solid
surfaces is scarce.9-12 Seventeen PAHs are on the Environmental
Protection Agency’s priority pollutant list.13 To date, we have
on photoproduct identification and the mechanism of their for-
mation. However, we have also investigated the role of surface
adsorption and other factors that can impact the adsorption and
phototransformation of these molecules at a solid/air interface.
In a previous paper on the photophysics and photochemistry of
anthracene adsorbed on silica, alumina, and cab-o-sil,15 we re-
ported the observation of ground-state stable pairs by anthracene
molecules on silica at surface coverages as low as 1% of a
monolayer. The stable pairs were shown to be responsible for
the observed dimerization of anthracene on silica.15 A similar
phenomenon has been observed for other PAHs that either lead
to dimer formation14,16 or exhibit the so-called excimer like
emission.17-19,21,22 To demonstrate the role of substituents on
ground-state pairing and phototransformation of anthracene on
silica, we examined the photophysics and photochemistry of
2MA, 2EA, and 2TBA. In this paper we report our steady state
and transient studies of these molecules at a silica solid/air
interface. A detailed account of the photoproducts and the mech-
anism of their formation upon photolysis of these compounds
is also given in this paper.
Materials and Methods
* To whom correspondence should be addressed. Address: Chemical
and Analytical Sciences Division, Oak Ridge National Laboratory, P.O.
Box 2008, MS-6100, Oak Ridge, TN 37831-6100. Phone: (423)576-7325.
Fax: (423)574-4939. E-mail: dabestanir@ornl.gov.
2-Methylanthracene, 2-ethylanthracene, and 2-tert-butylan-
thracene (Aldrich Chemical Co., Inc., Milwaukee, WI) were
10.1021/jp0022555 CCC: $19.00 © 2000 American Chemical Society
Published on Web 10/19/2000

10236 J. Phys. Chem. B, Vol. 104, No. 44, 2000
Dabestani et al.
SCHEME 1 . Schematic Diagrams of the Apparatus Used for the Diffuse Reflectance Laser Flash Photolysis and Singlet
Molecular Oxygen Lifetime Measurements
chromatographed on silica gel (Baker Analyzed, Phillipsburg,
NJ, 60-200 mesh), using cyclohexane as eluent, before use.
Purity was determined by gas chromatography (GC) to be
>99.8%. Cyclohexane, acetonitrile, methanol, methylene chlo-
ride, and toluene were all HPLC grade (Baker Analyzed) and
were used as received. Methylene blue (Baker Analyzed) was
used as received. 2,5-dimethylfuran (Aldrich, purity 99%) was
chromatographed on silica gel before use. Silica-60 (Baker
analyzed; 60-200 mesh) was activated at 200 °C in air for at
least 24 h before use. It had a surface area of 274 m2/g with an
average pore radius of ca. 60 Å19 as described previously.14-22
Percent monolayer surface coverage was calculated by fractal
theory reported previously15,16 using geometrical areas of 60.7
× 10-20 m2/molecule for 2MA, 64.0 × 10-20 m2/molecule for
2EA, and 67.4 × 10-20 m2/molecule for 2TBA, respectively.
Adsorption of 2MA, 2EA, and 2TBA onto silica surface was
achieved by a cyclohexane slurry method described pre-
viously.14-22 Aerated silica samples containing adsorbed 2MA,
2EA, and 2TBA (2 g) were photolyzed at 350 nm in uranium
glass tubes (330 nm cutoff, outer diameter ) 1 in) using a
Rayonet RPR-208 photoreactor equipped with a merry-go-round
attached in a horizontal configuration as reported before.14-22
Photolysis products and the unreacted starting material were
extracted from the surface by multiple washing with methylene
chloride, methanol, and acetonitrile to ensure complete removal
of all materials from the surface. A final wash of the silica
surface with cyclohexane and toluene was performed to remove
dimer(s) which are normally insoluble in methylene chloride,
methanol, and acetonitrile. Solvents were then removed under
vacuum at room temperature and the residue was dissolved in
a known volume of acetonitrile or cyclohexane (for dimers only)
for quantitative HPLC analysis. Product analysis was carried
out by reverse phase HPLC (Hewlett-Packard model 1090) as
described previously.15 Calibration for quantitative HPLC
analysis of the photoproducts was performed with authentic
samples, which were either commercially available or synthe-
sized. Material balance was better than 80% of total 2MA, 2EA,
and 2TBA loss during photolysis.
The 9,10-endoperoxides of 2MA, 2EA, and 2TBA were
prepared by photolysis (λex ) 650 nm) of an oxygenated solution
of each in a 9:1 methylene chloride/methanol mixture using
methylene blue as the singlet oxygen sensitizer.15
Diffuse reflectance spectra were recorded on a Cary 4E UV-
vis spectrophotometer (Varian) equipped with an integrating
sphere. Steady-state fluorescence spectra were obtained on a
Spex-Fluorolog-2 spectrometer equipped with double mono-
chromators on both excitation and emission sides. All spectra
were recorded from the front face of the cell containing the
solid sample and were corrected for silica background and
instrument response using correction factors provided by the
manufacturer. Degassed samples for fluorescence measurement
were prepared in tubes equipped with a 1 cm quartz sidearm
cell and were flame-sealed under vacuum (P g 1 × 10-6 Torr).
Diffuse Reflectance Laser Flash Photolysis. Scheme 1
shows the schematic for diffuse reflectance laser flash photolysis
(DRLFP) apparatus assembled in our laboratory to carry out
time-resolved experiments. The third harmonic (355 nm) of a
Minilite II Nd:YAG laser (Continuum) was passed through the
P1 prism and S1 shutter and delivered to the sample as the
excitation beam. The excitation beam was overlapped with the
probe beam delivered to the sample through S2 shutter from
either a high pressure 450-W xenon or 250-W mercury-xenon
(Oriel Corp.) lamp. The laser and shutters were triggered through
an AT-DIO-32F DAQ board (National Instruments). In general,
pulse energies of less than 9 mJ/pulse (5 ns pulse width) were
employed to irradiate 0.55 cm2 exposed area of the sample
giving a fluence of 16.4 mJ/cm2 per pulse. The laser energy
was attenuated using a digital delay generator (DG355, Stanford
Research System, Inc.) to delay the Q-switch trigger with respect
to flash lamp trigger. The energy of the laser pulse was measured
using a beam splitter to direct the beam into a J25LP-1 detector
(Molectron) connected to an EPM 1000 laser energy meter

Photophysical and -chemical Processes of Anthracene
J. Phys. Chem. B, Vol. 104, No. 44, 2000 10237
Figure 1. Changes in the absorption profile (obtained by diffuse reflectance) of 2-tert-butylanthracene (2TBA) adsorbed on silica as a function of
surface coverage for (curve A) 0.012%, (curve B) 0.54%, (curve C) 1.2%, (curve D) 23.7%, and (curve E) 53% of a monolayer. The inset shows
normalized diffuse reflectance spectra of anthracene adsorbed on silica for different surface coverages: A (0.5%), B (1.7%), C (3.2%), and D (8.8%
of a monolayer).
(Molectron) for read out. Two lenses were used to focus the
diffusely reflected light from the sample onto the entrance slit
(0.25 mm slit) of a DH10 monochromator (Instrument SA). The
exit slit of the monochromator was attached to a PMT (Oriel
77346). The output of PMT was amplified by a two stage
amplifier (SR445, Stanford Research Systems, Inc.), and the
signal was fed into a LeCroy 9354A transient digitizer which
was triggered for data acquisition by a photodiode signal
activated by the reflection of the laser beam through P1 prism.
For time-resolved absorption experiments, the observed emission
signals with probe light on (probe, background signal with laser
beam off), laser beam on (pump, probe beam off), and both
probe and laser beam on (probe + pump) were stored in three
memory slots of the LeCroy transient digitizer. The cuvette
containing the solid sample was agitated after each laser shot
to avoid possible photodecomposition. The transient reflectance
signal (∆R/R) was calculated using the following equation
the delay generator through a GPIB interface. The mercury lines
of a 250-W high-pressure xenon-mercury lamp (Oriel Corp.)
were used to align the detection system and calibrate the ICCD.
Emission traces collected by ICCD were exported as ASCII
files to Origin 4.0 software to be fitted using standard
procedures.
Results and Discussion
Photophysical Studies. (a) Steady State. In our previous
report on the photochemistry of AN adsorbed on silica, we
assigned the observed red-shifted band (centered around 390
nm) in the absorption profile to the ground-state stable pairs,
which exhibit a distinct emission spectra.15 In contrast to AN,
diffuse reflectance spectra of 2MA, 2EA, and 2TBA show no
evidence of a red-shifted band even at surface loadings as high
as 50% of a monolayer indicating that steric factors play an
important role in inhibiting the formation of ground-state pairs.
Figure 1 shows representative diffuse reflectance spectra for
2TBA adsorbed on silica at various surface loadings. The slight
red-shifting observed at high surface coverages (curves D and
E, Figure 1 for 23.7 and 53% of a monolayer, respectively) can
be attributed to microcrystal formation by 2TBA at these
loadings. This notion is further supported by our fluorescence
data. Figure 2 shows the changes in emission spectra of 2TBA
as a function of surface coverage. As the surface loading is
increased from 0.012% to 53% of a monolayer, the emission
bands at 385 and 400 nm are displaced with a new band
appearing around 410 nm while the intensity of the shoulder at
430 nm increases (curves D and E, Figure 2). The observed
red-shifting of emission spectrum as the surface loading is
increased can be attributed to the formation of 2TBA micro-
crystals, which exhibit a different emission spectrum (curve F
in Figure 2 shows the emission spectrum of pure 2TBA crystals).
Thus, the overall emission at high surface loadings (e.g., 23.7
and 53% of a monolayer, curves D and E, Figure 2) is the sum
of contributions from monomeric and microcrystal forms of
2TBA. Similar results have been observed for 2MA and 2EA
(not shown). Our steady-state absorption and emission data
∆R/R ) [probe - ((probe + pump) - pump)]/probe
in conjunction with the math functions provided by LeCroy data
acquisition software that could be accessed on the transient
digitizer. All digitized kinetic signals from the transient digitizer
were fed to a 488.2 GPIB board (National Instruments) and
stored as ASCII files. The data were then transformed to time-
resolved absorption spectra using Origin 4.0 software (Microcal).
Time-resolved diffusely reflected emission (TRDRE) signals
from solid samples were collected using an Intensified Charge
Coupled detector (ICCD) with a slight modification to Scheme
1 for the DRLFP setup. Diffusely reflected emission from the
sample was focused, using a combination of lenses, into the
entrance slit of a polychromator (EG&G PARC-1226 with 150
groove/mm holographic grating). The light from the exit port
of polychromator was then focused on the front face of a gated
ICCD (InstaSpec IV, Andor Technology) operated at -5 °C.
Data acquisition was performed using the software (Insta Spec
1.11 version) provided by the manufacturer which was integrated
with a DG355 digital delay generator (Stanford Research
Systems, Inc.) to control the gate width and the delay timing of

10238 J. Phys. Chem. B, Vol. 104, No. 44, 2000
Dabestani et al.
Figure 4. Time-resolved transient difference spectra of 2TBA adsorbed
on silica (4.3 × 10-7 mol/g) showing reflectance changes following
laser excitation at 355 nm (laser energy: 5 mJ/pulse) after 19, 88, and
140 us. The inset shows the decay of transient at 420 nm and the
biexponential fit of the trace.
Figure 2. Changes in the emission (λex ) 350 nm) spectra of degassed
2TBA adsorbed on silica as a function of surface coverage for (curve
A) 0.012%, (curve B) 0.54%, (curve C) 1.2%, (curve D) 23.7%, and
(curve E) 53% of a monolayer. Curve F corresponds to the emission
of crystalline form of 2TBA. The inset shows variation in the excitation
(λem ) 400 nm for A and B and λem ) 420 nm for C) and emission
(λex ) 360 nm) spectra (normalized relative to total area) of anthracene
adsorbed on silica (degassed) as a function of surface coverage: A
(0.5%), B (3.2%), and C (6.4% of a monolayer).
g-1) gave lifetimes in the range 6 and 30 ns, respectively. These
lifetimes are in good agreement with the reported values for
anthracene on silica24 and suggest that alkyl substituents do not
alter the excited singlet state lifetime significantly. Addition of
water to silica (3 × 10-3 mol g-1) does not affect the short-
lived component but shortens the lifetime of the longer lived
component to about 20 ns.
Time-resolved transient absorption spectra of degassed 2TBA
adsorbed on silica (4.3 × 10-7 mol g-1) obtained at 19, 88,
and 140 µs after the laser pulse (λex ) 355 nm) are shown in
Figure 4. The transient exhibits three peaks at 390, 410, 420
(major), and a shoulder around 435 nm. No evidence of any
cation radical formation, which absorbs in the 700-750 nm
region, was observed under these conditions. The formation of
anthracene cation radical on silica by a biphotonic process has
been reported.25 The low energy of excitation light (3-5 mJ/
pulse laser power) used in our experiments accounts for the
lack of any cation radical formation under these monophotonic
excitation conditions. The fact that this transient was completely
quenched in the presence of oxygen further supports the above
notion and confirms the absence of any cation radical contribu-
tion to the observed transient. Thus, we assign this transient to
triplet-triplet absorption, T1 f Tn, of the first excited triplet
state of 2TBA. The biexponential fit of the decay curve at 420
nm gives a short-lived component with a lifetime of about 5.5
µs and a long-lived component with a lifetime of 0.24 ms,
respectively (inset, Figure 4). Similar transient absorption spectra
were also obtained for 2MA and 2EA adsorbed on silica
indicating that alkyl substituents did not greatly alter the shape
of triplet-triplet absorption relative to that of anthracene. The
lifetime of the transients determined for 2MA (9 × 10-7 mol
g-1) and for 2EA (6.9 × 10-7 mol g-1) were 3.8 µs and 0.15
ms and 5.2 µs and 0.12 ms, respectively. We assign the longer-
lived component to the triplet state of 2MA, 2EA, and 2TBA.
Because the origin of shorter-lived component is not clear to
us, we cannot assign it to any particular state.
Figure 3. Time-resolved emission spectra of 2TBA adsorbed on silica
(8.5 × 10-7 mol/g) following laser excitation at 355 nm. A 2 ns gated
ICCD camera was used to record the transient emission. Spectra were
obtained at 20, 25, 30, and 35 ns time intervals. The inset shows an
enlarged portion of the emission spectra taken at 35, 40, and 45 ns
after the laser pulse.
clearly show that introduction of bulky alkyl groups into AN
moiety inhibits the formation of ground-state pairs at low surface
coverages. At high surface loadings these alkyl substituted AN
molecules crystallize on silica.
(b) Transient. Excited-state dynamics of 2-alkyl substituted
anthracences adsorbed on silica were studied using the time-
resolved diffuse reflectance laser flash photolysis technique.23
All samples were degassed for several days under high vacuum
and flame-sealed before conducting transient studies. Time-
resolved transient emission of 2TBA adsorbed on silica (2.0 ×
10-6 mol g-1) using 355 nm excitation is shown in Figure 3.
The emission shows spectral features (bands at 388, 406, and
432 nm in Figure 3) similar to those obtained under steady-
state irradiation conditions (Figure 2). The emission decay for
2TBA at 406 nm was best fitted to a biexponential decay curve
showing a short-lived component with a lifetime of about 5.0
( 0.2 ns and a longer lived second component with a lifetime
of 33.0 ( 0.3 ns. A similar treatment of the emission decay
profiles for 2MA and 2EA adsorbed on silica (2 × 10-6 mol
The triplet lifetime was sensitive to the concentration of the
substrate on silica and decreased as the surface loading was
increased. We attribute this decrease in lifetime to the quenching
of triplet excited state by microcrystals formed at higher surface
loadings (i.e., self-quenching by neighboring molecules in close
proximity of the excited molecule). The rate constant for
quenching of 2TBA triplet state by ground-state molecules was
determined to be 4.3 × 105 g s-1 mol-1 from a plot of inverse
of the triplet lifetime vs concentration of 2TBA on silica (not
shown).

Photophysical and -chemical Processes of Anthracene
J. Phys. Chem. B, Vol. 104, No. 44, 2000 10239
Figure 5. Changes in the decay profile of excited-state triplet (420
nm) of 2TBA (4.3 × 10-7 mol/g) adsorbed on silica as a function of
concentration of physisorbed water (355 nm laser excitation). The inset
shows plot of triplet decay rate constant Vs concentration of physisorbed
water.
Figure 6. Effect of physisorbed water on the near-IR emission (1270
nm) of singlet molecular oxygen produced by 355 nm laser excitation
of an aerated sample of 2TBA adsorbed on silica (3.4 × 10-6 mol/g)
for (curve a) dry, (curve b) 72.6 wt % water/g of silica, and (curve c)
112 wt % water/g of silica. Curve d shows the observed emission from
a Suprasil quartz tube. A similar emission curve as d is observed with
silica present in the quartz tube.
The triplet lifetime was also very sensitive to the presence
of physisorbed water on silica and decreased as the water
concentration was increased. Figure 5 shows the variation in
triplet decay profile for 2TBA adsorbed on silica as the
concentration of physisorbed water is increased. From the slope
of a plot of 1/τT vs concentration of water (inset, Figure 5) on
silica surface, a bimolecular rate constant of 8.9 × 105 s-1 mol-1
g was determined for this process. The bimolecular rate
constants for 2EA and 2MA were determined to be 4.1 × 105
and 2.4 × 107 s-1 mol-1 g, respectively. A plausible explanation
for the observed decrease in triplet lifetime in the presence of
physisorbed water can be provided as follows. Polar water
molecules can easily displace (via hydrogen bonding with silanol
groups) loosely held 2-alkylanthracenes from the adsorbed sites
allowing the desorbed molecules to be more mobile (diffuse
more freely) on the surface. Since solubility of 2-alkylan-
thracenes in water is very low, the desorbed molecules will tend
to aggregate and form microcrystals as evidenced by diffuse
reflectance absorption and emission spectroscopy (not shown).
The interaction of these molecules with the excited triplet state
(a self-quenching process) could result in the observed shorten-
ing of the triplet lifetime in the presence of physisorbed water.
We believe the difference in bimolecular quenching rate
constants calculated for 2MA, 2EA, and 2TBA results from the
difference in their water solubility. The bimolecular rate constant
for the self-quenching of anthracene in ethanol has been reported
for the lifetime of the emitting impurity (curve d) present in
the Suprasil quartz tube.27 These lifetime data cannot explain
the observed decrease in photodecomposition rates for 2MA,
2EA, and 2TBA in the presence of physisorbed water.
The quantum efficiency of singlet oxygen production de-
creased as the concentration of water on silica was increased.
This conclusion was drawn by comparing the intensity of singlet
oxygen emission for dry and wet silica extrapolated to t ) 0
time. The observed emission intensities in the presence of
physisorbed water and deuterated water 28 were 0.46 and 0.22,
respectively, relative to the intensity from dry silica. Since
singlet oxygen is produced upon interaction of molecular oxygen
with the excited state of 2MA, 2EA, and 2TBA (type II), a
shortening of the excited-state lifetime in the presence of
physisorbed water can lead to lower efficiency for its formation.
Our transient data indicates that excited singlet states of 2MA,
2EA, and 2TBA are unaffected by the presence of physisorbed
water (not shown) while excited triplet states experience a
lifetime shortening. Thus, the observed decrease in photolysis
rate in the presence of physisorbed water can be attributed to
the shortening of 2MA, 2EA, and 2TBA excited-state triplet
lifetimes which in turn causes a decrease in the efficiency of
singlet molecular oxygen production.
Photochemical Studies. Photolysis of 2MA, 2EA, and 2TBA
at a silica/air was carried out in a Rayonet RPR-208 photoreactor
equipped with a merry-go-round using 350 nm excitation light
source as described previously.15 The photolysis rate was
dependent on the surface coverage and decreased as the surface
coverage was increased. We attribute this decrease in photolysis
rate at higher surface loadings to an inner filter effect induced
by crystal formation. These crystals absorb the incident light
(competing with 2MA, 2EA, and 2TBA monomeric forms)
without producing any net chemistry (light sinks). Figure 7
compares the first-order rate constant for photolysis of AN,
2MA, 2EA, and 2TBA. The rate constant decreases in the
following order 2TBA > 2EA > 2MA > AN. The data shows
that for the same surface coverage, 2TBA reacts almost twice
faster than AN. This difference in reactivity can be explained
on the basis of our spectroscopic findings. Steady state absorp-
tion and emission results indicate that at this surface coverage
a significant portion of AN molecules are present in stable pair
form while 2TBA exists mainly in the monomeric form.
Absorption of incident light by the AN stable pairs does not
lead to a photochemical reaction while the monomeric form of
to be 5.19 × 105 M-1 s-1 26
.
The decrease in the lifetime of excited triplet state will affect
the rate of bimolecular processes involving this state (e.g.,
energy transfer to molecular oxygen to form singlet oxygen).
To explain the observed decrease in photodecomposition rate
of 2-alkyl substituted anthracenes on silica in the presence of
physisorbed water (Vida infra), we carried out a time-resolved
near-IR study of singlet molecular oxygen emission (1270 nm)
using an apparatus similar to that reported by Bilski et al.27 The
purpose of this study was to examine the effect of physisorbed
water (if any) on the lifetime and/or quantum efficiency of
singlet molecular oxygen produced via energy transfer from
excited 2MA, 2EA, or 2TBA to molecular oxygen. Figure 6
shows the decay of singlet molecular oxygen emission (1270
nm) on silica, produced by 2TBA via energy transfer (type II),
in the presence of physisorbed water. A fit of the decay curves
(monoexponential) gives singlet oxygen lifetime values of 3.9
( 0.2 µs for dry silica (curve a), 7.6 ( 1.9 µs in the presence
of water (curve b), and 10.5 ( 0.6 us in the presence of
deuterated water (curve c). A value of 0.9 us was determined

10240 J. Phys. Chem. B, Vol. 104, No. 44, 2000
Dabestani et al.
Figure 7. Plot of first-order rate constant for the loss of anthracene, (b) AN, (1) 2MA, ([) 2EA, and (9) 2TBA at silica/air interface. First-order
rate constants calculated from the slopes are: kAN ) 3.1 ( 0.1 × 10-3 s-1, k2MA ) 4.0 ( 0.3 × 10-3 s-1, k2EA ) 5.5 ( 0.1 × 10-3 s-1, and k2TBA
) 5.8 ( 0.2 × 10-3 s-1
.
SCHEME 2 . Observed Photoproducts from Photolysis (λEx ) 350 Nm) of 2MA, 2EA, and 2TBA at a Silica/Air Interface
2TBA undergoes photochemical conversion to products (Vida
infra). As the surface coverage is increased, the decrease in
photolysis rate for 2MA, 2EA, and 2TBA is consistent with
the formation of crystals at higher surface loadings. A similar
behavior by other PAHs has been reported previously.14-22 The
photolysis rate was also very sensitive to the presence of
physisorbed water on the silica surface. Time-resolved transient
studies carried out on AN, 2MA, 2EA, and 2TBA adsorbed on
dry and wet silica have revealed that the observed decrease in
photolysis rate, in the presence of physisorbed water, results
from shortening of the excited-state triplet lifetime (Vida infra).
Scheme 2 shows the photoproducts formed upon photolysis
of 2MA, 2EA, and 2TBA at silica/air interface. Our data
indicates that at low surface coverages no photodimer is formed
and photooxidation products account for almost 100% of the
total products. Although small amounts of photodimers are
observed at higher surface coverages, they only account for
e5% of the total products. The observation that photodimers
form at higher surface coverages, when the molecules start to
crystallize, indicates that crystal forms of 2MA, 2EA, and 2TBA
are most likely responsible for the formation of photodimers.
The assigned structures of 9,10-photodimers shown in Scheme
2 is based on our solution studies where we observed (by LC-
MS) three dimeric products upon photolysis of degassed samples
of 2-substituted anthracenes in cyclohexane.
The role of singlet molecular oxygen in the oxidation of
2-substituted anthracenes on silica was examined by conducting
the photolysis in the presence of coadsorbed: (a) 2,5-dimeth-
ylfuran, a well-known singlet oxygen trap,29 and (b) methylene
blue, a good generator of singlet oxygen on the silica surface.30
When a 10:1 mixtures of methylene blue/silica (1 × 10-5 mol/
g) and 2-substituted anthracenes/silica (2 × 10-5 mol/g) was
photolyzed at 650 nm, where the substituted anthracenes do not
absorb, only the oxidation products shown in Scheme 2 and no
photodimers were observed. Direct photolysis (λex ) 350 nm)
of 2-substituted anthracenes/silica (2 × 10-5 mol/g) in the
presence of 2,5-dimethylfuran/silica (2 × 10-4 mol/g) in a 1:1
ratio produced a small amount of dimers but no oxidation
products. These results unequivocally implicate the role of
singlet molecular oxygen as the intermediate in inducing the
oxidation of 2-substituted anthracenes on silica.
Conclusions
Photophysical studies of 2-methyl, 2-ethyl, and 2-tert-
butylanthracenes indicate that in contrast to anthracene, which
forms stable pairs at very low surface coverages (1% of a
monolayer) when adsorbed on silica, these substituted an-
thracenes show no evidence of pairing even at surface coverages
as high as 53% of a monolayer. At higher surface coverages,

Photophysical and -chemical Processes of Anthracene
J. Phys. Chem. B, Vol. 104, No. 44, 2000 10241
(10) Bauer, R. K.; Borenstein, R.; de Mayo, P.; Okada, K.; Rafalska,
M.; Ware, W. R.; Wu, K. C. J. Am. Chem. Soc. 1982, 104, 4635.
(11) Hasegawa, K.; Murase, M.; Kuboshita, M.; Saida, H.; Shinoda, M.;
Miyamoto, M.; Shimasaka, C.; Yoshimura, T.; Tsukurimichi, E. EnViron.
Sci. Technol. 1993, 27, 1819.
(12) Mao, Y.; Thomas, J. K. Langmuir 1992, 8, 2501.
(13) Toxicological Profile for Polycyclic Aromatic Hydrocarbons Report;
U. S. Department of Health and Human Services, Public Health Service:
Washington, DC, 1993.
(14) Barbas, J. T.; Dabestani, R.; Sigman, M. E. J. Photochem.
Photobiol. A: Chem. 1994, 80, 103.
(15) Dabestani, R.; Ellis, K. J.; Sigman, M. E. J. Photochem. Phtotobiol.
A: Chem. 1995, 86, 231.
(16) Dabestani, R.; Nelson, M.; Sigman, M. E. Photochem. Photobiol.
these molecules tend to crystallize on silica surfaces. The
presence of bulky alkyl groups on the anthracene moiety is most
likely responsible for the lack of pairing by these substituted
anthracenes. Photodecomposition of 2-substituted anthracenes
on silica proceeds more efficiently than anthracene to produce
mainly oxidation products and small amounts of isomeric
dimers. This enhanced reactivity is attributable to the absence
of ground-state pairs, which can act as a light sink without
producing any net chemistry. The oxidation products are formed
by the addition of singlet molecular oxygen (sensitized by the
interaction of excited state of 2-substituted anthracenes with
molecular oxygen) to the ground state of 2-substituted an-
thracenes. The precursor to photodimers appears to be the
crystalline forms of these substituted anthracenes.
1996, 64, 80.
(17) C. Reyes, C.; Sigman, M. E.; Arce, R.; Barbas, J. T.; Dabestani,
R. Photochem. Photobiol. A: Chem. 1998, 112, 277.
(18) Barbas, J. T.; Sigman, M. E.; Arce, R.; Dabestani, R. J. Photochem.
Photobiol. A: Chem. 1997, 109, 229.
Acknowledgment. This research was sponsored by the
Division of Chemical Sciences, Geosciences, and Biosciences,
Office of Basic Energy Sciences, U. S. Department of Energy
under contractor DE-AC05-96OR22464 with Oak Ridge Na-
tional Laboratory, managed by UT-Battelle for the Department
of Energy.
(19) Sigman, M. E.; Barbas, J. T.; Chevis, E. A.; Dabestani, R. New. J.
Chem. 1996, 20, 243.
(20) Barbas, J. T.; Sigman, M. E.; Dabestani, R. EnViron. Sci. Technol.
1996, 30, 1776.
(21) Barbas, J. T.; Sigman, M. E.; Buchanan, A. C.; Chevis, E. A.
Photochem. Photobiol. 1993, 58, 155.
(22) Reyes, C. A.; Medina, M.; Crespo-Hernandez, C.; Cedeno, M. Z.;
Arce, R.; Rosario, O.; Steffenson, D. M.; Ivanov, I. N.; Sigman, M. E.;
Dabestani, R. EnViron. Sci. Technol. 2000, 34, 415.
(23) Wilkinson, F.; Kelly, G. P. In Handbook of Organic Photochemistry;
Scaiano, J. C., Ed.; CRC Press: Boca Raton, FL, 1989; Vol. 1; pp 293-
314.
(24) Bauer, R. K.; Borenstein, R.; deMayo, P.; Ware, W. R.; Wu, K. C.
J. Phys. Chem. 1982, 86, 3781.
(25) Wilkinson, F.; Worrall, D. R.; Williams, S. L. J. Phys. Chem. 1995,
99, 6689.
(26) Saltiel, J.; Marchand, G. R.; Dabestani, R.; Pecha, J. M. Chem.
References and Notes
(1) Hutzinger, O., Ed. The Handbook of EnVironmental Chemistry;
Springer: Berlin, 1982; Vol. 2, Part B.
(2) Zika, R. G.; Cooper, W. J., Eds. Photochemistry of EnVironmental
Aquatic Systems; ACS Symposium Series 327; American Chemical
Society: Washington, DC, 1987.
(3) de Mayo, P.; Natarajan, L. V.; Ware, W. R. In Organic Phototrans-
formation in Nonhomogeneous Media; ACS Symposium Series 278 Fox,
M. A., Ed.; American Chemical Society: Washington, DC, 1984; pp 1-19.
(4) Bauer, R. K.; de Mayo, P.; Ware, W. R.; Wu, K. C. J. Phys. Chem.
1982, 86, 3781.
Phys. Lett. 1983, 100, 219.
(27) Bilski, P.; Chignell, C. F. J. Biochem. Biophys. Methods 1996, 33,
73.
(28) Deuterated water was employed as a means to see whether singlet
oxygen lifetime is increased relative to water indicating preferential diffusion
of this species from the surface into the liquid phase. Our data, however,
shows no significant change in the lifetime of singlet oxygen for physisorbed
water or deuterated water ruling out a substantial contribution from such
diffusion.
(5) Bauer, R. K.; de Mayo, P.; Okada, K.; Ware, W. R.; Wu, K. C. J.
Phys. Chem. 1983, 87, 460.
(6) de Mayo, P.; Natarjan, L. V.; Ware, W. R. Chem. Phys. Lett. 1984,
107, 187.
(7) Levitz, P.; Van Damme, H.; Keracis, P. J. Phys. Chem. 1984, 88,
2228.
(29) Gollnick, K.; Griesbeck, A. Photochem. Photobiol. 1985, 41, 2057.
(30) Brkic, D.; Frozatti, P.; Pasquoss, I.; Trifiro, F. J. Mol. Catal. 1977,
3, 173.
(8) Thomas, J. K. Chem. ReV. 1993, 93, 301.
(9) de Mayo, P.; Okada, K.; Rafalska, M.; Weedon, A. C.; Wong, G.
S. J. Chem. Soc., Chem. Commun. 1981, 820.