Time-resolved spectroscopic and density functional theory study of the photochemistry of Irgacure-2959 in an aqueous solution

Article abstract of DOI:10.1021/jp506099n

The photocleavage reaction mechanism of 2-hydroxy-4′-(2-hydroxyethoxy)-2-methylpropiophenone (Irgacure-2959) was investigated using femtosecond (fs) and nanosecond (ns) transient absorption (-TA) spectroscopy and also picosecond (ps) and nanosecond (ns) time-resolved resonance Raman (-TR³) spectroscopy experiments in a water-rich (volume ratio of acetonitrile/water = 3:7) solution. TA spectroscopy was used to study the dynamics of the benzoyl radical growth and decay as well as to investigate the radical quenching process by the radical scavenger methyl acrylate. ps- and ns-TR³spectroscopies were employed to monitor the formation of the benzoyl radical and also to characterize its electronic and structural properties. The fs-TA experiments results indicate that the Irgacure-2959 lowest lying excited singlet state S₁underwent efficient intersystem crossing (ISC) to convert into its triplet state with a time constant of 4 ps. Subsequently, this triplet species dissociated into the benzoyl and alkyl radicals with a corresponding maximum absorption band at 415 nm. The TR³results in conjunction with results from DFT calculations confirmed that Irgacure-2959 cleaved into the benzoyl and alkyl radicals at a fast rate on the tens of picosecond time scale.

Full text of DOI:10.1021/jp506099n

Subscriber access provided by UNIV OF YORK
Article
Time-Resolved Spectroscopic and Density Functional Theory Study
of the Photochemistry of Irgacure-2959 in an Aqueous Solution
Mingyue Liu, Ming-De Li, Jiadan Xue, and David Lee Phillips
J. Phys. Chem. A, Just Accepted Manuscript • DOI: 10.1021/jp506099n • Publication Date (Web): 18 Aug 2014
Downloaded from http://pubs.acs.org on August 20, 2014
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
The Journal of Physical Chemistry A is published by the American Chemical Society.
1155 Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 8
The Journal of Physical Chemistry
1
2
3
4
5
6
7
Time-Resolved Spectroscopic and Density Functional Theory
Study of the Photochemistry of Irgacure-2959 in an Aqueous
Solution
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Mingyue Liu†, Ming-De Li*,†, Jiadan Xue‡, and David Lee Phillips*, †
†
Department of Chemistry, The University of Hong Kong, Pokfulam Road, Hong Kong S.A.R., P. R. China
‡
Department of Chemistry, Zhejiang Sci-Tech University, Hangzhou 310018, China
KEYWORDS: Norrish Type I Reaction, Irgacure-2959, Photoinitiator, Time-resolved, Raman, DFT.
ABSTRACT: The photocleavage reaction mechanism of 2-hydroxy-4´-(2-hydroxyethoxy)-2-methylpropiophenone (Irgacure-
2959) was investigated using femtosecond (fs) and nanosecond (ns) transient absorption (-TA) spectroscopy and also picosecond
(ps) and nanosecond (ns) time-resolved resonance Raman (-TR3) spectroscopy experiments in a water-rich (volume ratio of acetoni-
trile:water=3:7) solution. TA spectroscopy was used to study the dynamics of the benzoyl radical growth and decay as well as to
investigate the radical quenching process by the radical scavenger methyl acrylate. Ps- and ns-TR3 spectroscopies were employed to
monitor the formation of the benzoyl radical and also to characterize its electronic and structural properties. The fs-TA experiments
results indicate that the Irgacure-2959 lowest lying excited singlet state S1 underwent efficient intersystem crossing (ISC) to convert
into its triplet state with a time constant of 4 ps. Subsequently, this triplet species dissociated into the benzoyl and alkyl radicals
with a corresponding maximum absorption band at 415 nm. The TR3 results in conjunction with results from DFT calculations con-
firmed that Irgacure-2959 cleaved into the benzoyl and alkyl radicals at a fast rate on the tens of picosecond timescale.
the quantum yield of compounds with a para substituent of elec-
INTRODUCTION
tron withdrawing groups of n,π* and electron donating groups of
π,π* triplet characters. More recent work by Scaiano and co-
workers found that Irgacure-2959 is a good photocaged radical
source to reduce silver and gold nanoparticles in aqueous solu-
tions. It was further confirmed that Irgacure-2959 underwent a
cleavage reaction to produce benzoyl and ketyl radicals as shown
in Scheme 1,6 whose detailed mechanism of the cleavage process
still remains open for investigation. It is thus important to assign
the excited states and further work can be carried out to clearly
identify the intermediate species and to elucidate an overall reac-
tion pathway for Irgacure-2959 in an aqueous solution.
The Norrish type I reaction has been known for over 70 years,
and it can occur from either singlet or triplet states.1-3 In recent
years, 2-hydroxy-4’-(2-hydroxyethoxy)-2-methylpropiophenone
(Irgacure-2959) has received attention due to its applications in
using its Norrish type I reaction as a convenient photocaged free
4 - 7
radicals source for noble metal nanoparticles synthesis.
Irgacure-2959 is a typical member of highly efficient photoinitia-
tors for photocleavage reactions that involves its excited states
and dissociates into free radicals upon UV light irradiation
through an α-cleavage reaction. A number of efforts have been
devoted to metal nanoparticles synthesis by reducing metal ions
using the radicals from the photocleavage reaction of Irgacure-
2959. Yet the study of Irgacure-2959 itself and its related benzoyl
radicals photochemistical properties remains limited with some
laser flash photolysis and diode laser-based time-resolved infrared
spectroscopy (TRIR) experiments reported for its photochemistry
in acetonitrile solvent. 8-11
Direct information of reactive intermediates and their corre-
sponding kinetics are of importance to better understand the relat-
ed processes associated with the photochemical cleavage, which
requires time-resolved techniques. Most of the previous time-
resolved research work on this compound relied on laser flash
photolysis (LFP) with a time resolution on the nanosecond time-
scale.12,13 It would be useful to examine the reaction over a broad-
er time scale and in particular the femtosecond to picosecond time
scales to directly probe early time events in the reaction processes.
Time-resolved Resonance Raman (TR3) may obtain clear struc-
tural information from resonance Raman spectra that are convinc-
ing enough to provide direct pieces of evidence to distinguish
among temporal and structurally similar reactive intermediates,
especially, those that have overlapped absorption spectra in LFP
measurements. Therefore, TR3 is a powerful method to identify
the intermediates generated and to monitor their corresponding
Based on transient optical absorption experiments, Turro and
co-workers reported that different substituents on the para posi-
tion of the benzene ring play different roles on nature of the excit-
ed triplet state and the quantum yield of the α-cleavage reactions.8
In their investigation, they employed 355 nm as the pump laser
wavelength to excite the sample and pointed out that Irgacure-
2959 with a HO(CH2)O- substituent, has its triplet state involving
a mixed n,π* and π,π* triplet character that leads to an intermedi-
ate level of photocleavage quantum yield of 0.29 that lies between
ACS Paragon Plus Environment

The Journal of Physical Chemistry
Page 2 of 8
1.0 ps
25 ps
399 ps
2649 ps
a
kinetics of the short-lived species produced in the different time-
b
1.5 ps
1.8 ps
2.3 ps
3.0 ps
3.5 ps
4.5 ps
5.5 ps
7.0 ps
10 ps
25 ps
415 nm
415 nm
scale processes.14-20
= 3.8 0.2 ps
1
2
3
4
5
6
7
*
*
415 nm
Scheme 1. Proposed photochemistry pathway of the benzoin
Irgacure-2959 by Scaiano and co-workers
0
5
10 15 20 25 30 35 40
delay time/ps
417 nm
O
O
OH
OH
hν
+
OH
OH
O
O
350
400
450
500
550
600
benzoyl
radical
ketyl
radical
350
400
450
500
550
600
wavelenth/nm
wavelenth/nm
Irgacure-2959
8
9
Figure 2a,b. Shown are fs-TA spectra of Irgacure-2959 at early
time to later delay time recorded with 267 nm excitation in a
MeCN:H2O=3:7 mixed solution. The asterisks (*) mark subtrac-
tion artifacts.
Time-resolved resonance Raman spectroscopy in combination
with femtosecond to nanosecond transient absorption has been
successfully used to study processes from a very early timescale
to final products so as to analyze and characterize the nature and
properties of the intermediates generated after irradiation.21-25 In
this paper, Irgacure-2959 was studied using fs-TA, ns-TA and
time-resolved resonance Raman spectroscopy in an aqueous solu-
tion. These results found that the compound evolves from S1 to T1
through intersystem crossing and subsequently undergoes the
photocleavage of the triplet state that is responsible for the gen-
eration of the benzoyl radical. In combination with results from
density functional theory (DFT) calculations, the key species in-
volved in the photocleavage of Irgacure-2959 are assigned using
TR3 spectra and a clear overall reaction mechanism is proposed.
To our knowledge, this is the first time that time-resolved reso-
nance Raman spectroscopy has been used to study the photo-
cleavage reaction of Irgacure-2959.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 2a and 2b display the fs-TA spectra of Irgacure-2959
obtained in a MeCN:H2O=3:7 mixed solvent from 0 to 2649 ps
delay times. To better present the spectra with different time
scales of these experimental results, two separated spectra with
different time scales have been shown in Figure 2a (1-25 ps) and
b (25-2649 ps) respectively. Figure 2a shows the temporal evolu-
tion of the early time spectra scale has a major strong absorption
band growing in from 1 ps to 25 ps. At the beginning stage of 1 ps
a broad absorption band at 410 nm evolves at a very fast rate up to
5 ps. Then, this absorption band grows relatively slowly and stops
growing at about 14 ps. During this growth process, the maximum
of the absorption band slightly shifts from 410 to 415 nm. The
inset plot displays the kinetics of the transient absorption at 415
nm. The growth process is assigned to the S1 to T1 intersystem
crossing process with a time constant of ~4 ps, which is assigned
to the Irgacure-2959 triplet state growth for two reasons. First, an
intermediate of Irgacure-2959 in neat acetonitrile with an absorp-
tion band at 390 nm was suggested to be assigned as the triplet in
Turro and co-workers’ publication,8 although a slight redshift
occurred from 390 nm to 410 nm which is affected by different
polarity of the solvents. Secondly, a similar molecule (p-
hydroxyphenacyl diethyl phosphate) also was observed to under-
go an ISC process transforming its singlet state into its triplet
state, where the transient absorptions of the singlet excited state
and triplet state appeared around 320 nm and 400 nm respectively
with an isobestic point around 330 nm.26 Unfortunately, our cur-
rent lowest wavelength of the white continued light (WCL) only
could reach to 330 nm. Thus we did not obtain the transient ab-
sorption of Irgacure-2959 at wavelengths ≤330 nm. Even though,
there is no obvious isobestic point in the spectra of Figure 2b, the
spectral evolution is very similar with that reported for p-
hydroxyphenacyl diethyl phosphate (HPDP)26 and it is reasonable
to attribute the growth of the 415 nm band to the formation of the
triplet state.
RESULTS AND DISCUSSION
UV/Vis and fs-TA spectra of Irgacure-2959 in a water-rich
solution. Figure 1 shows the UV/Vis spectra of Irgacure-2959
measured in pure MeCN and MeCN: H2O=3:7 solutions. The
most intense absorption band is located at 273 nm in MeCN
which is associated with π,π* transition and this band is gradually
red shifted to a longer absorption wavelength region at 279 nm as
the concentration of water is increased in mixed aqueous solu-
tions. Polarity of the solvent affects the absorption band position.
In general, absorption bands associated with a π,π* transition shift
to a longer absorption wavelength (redshift) when the solvent
polarity is increased.
Irgacure-2959 in MeCN
Irgacure-2959 in MeCN:H2O=3:7
279 nm
273 nm
218 nm
Figure 2b displays the later time scale spectral changes of the
fs-TA of Irgacure-2959. From 25 ps, it decays at a relatively fast
rate to 400 ps and then it decays relatively slowly up to 2649 ps,
with the maximum absorption slightly shifting to a longer wave-
length absorption band at 417 nm. This may be associated with
the cleavage of the triplet state of Irgacure-2959 into the benzoyl
and ketyl radicals. It should be noted that the absorption band of
the triplet species has a strongly overlapping absorption band with
that of the benzoyl radical due to the triplet state of Irgacure-2959
and its benzoyl radical having the same basic chromophore of
benzene and the HO(CH2)O- substituent group. This makes it
difficult to distinguish any reasonably distinct absorption bands
between the Irgacure-2959’s triplet and the benzyl radical from
the TA spectra due to these two transient species being strongly
overlapped with each other over an almost identical absorption
region. The kinetics of the decay process is displayed in Figure 3.
200
225
250
275
300
325
350
375
400
wavelength/nm
Figure 1. UV/Vis absorption spectra of Irgacure-2959 in MeCN
and MeCN:H2O=3:7 solvents with labels on the top right corner.
TD-DFT calculations were employed to simulate the UV/Vis
absorption spectra (see results depicted in Figure 1S) and the fron-
tier orbitals were also calculated from the B3LYP DFT calcula-
tions using a 6-311G** basis set (see results shown in Figure 2S)
to examine the characteristics of the absorption transitions. The
calculation results show that the absorption band at 273 nm arises
mainly from the transition from the HOMO to the LUMO, and the
band at 218 nm can be attributed to the HOMO to LUMO+1 tran-
sition. These two absorption bands are characteristic of π,π* ab-
sorption transitions. The pump laser wavelength employed in our
fs-TA and ps-TR3 experiments is 267 nm derived from the third
harmonic of the fundamental 800 nm from the regenerative ampli-
fier laser system.
ACS Paragon Plus Environment

Page 3 of 8
The Journal of Physical Chemistry
readily dissociates to produce vibrationally excited benzoyl radi-
1
2
3
4
5
6
7
8
9
cals. This is consistent with the behavior of the early time ps-TR3
spectra observed in Figure 4a where no obvious band are attribut-
ed to the T1 state and the initially produced benzoyl radical Ra-
man bands are weak and very broad at early times but sharpen up
on the tens of ps time scale typical of vibrationally cooling..
To better understand the later time scale reaction, ns-TR3 ex-
periments were also carried out in an aqueous solution. Figure 4b
displays the ns-TR3 spectra of 1mM Irgacure-2959 in open air in
an acetonitrile:water=3:7 mixed solution from 0 to 1000 ns. One
species is observed in the spectra. It decays gradually from the
beginning spectrum with a strong vibrational bands located at
941, 1162, 1329 and 1590 cm-1. Figure 3S shows the kinetics
fitted by the integrated area of the strongest Raman band at 1590
cm-1 vs the delay time. This fit results in a decay time constant of
79 ns. The ns-TR3 spectra clearly show that the species decays
gradually until several hundreds of nanoseconds. Figure 4S dis-
plays the comparison of the resonance Raman spectrum of
Irgacure-2959 obtained at a delay time of 10 ns with that from the
ps-TR3 spectra obtained with a delay time of 50 ps. Obviously,
they have good agreement with each other. Hence, it is reasonable
to deduce that these two spectra are contributed from the same
species. The view of the TR3 spectra from ps to ns timescale indi-
cates that the benzoyl radical is generated from the triplet state of
Irgacure-2959 and grows in at several hundred ps and decays
gradually to 1000 ns with a time constant of 79 ns in an open air
condition.
415 nm
500 1000 1500 2000 2500 3000
delay time/ps
0
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 3. Temporal dependence of transient absorption spectrum
at 415 nm for Irgacure-2959 recorded at later picosecond times.
Ps-TR3 and ns-TR3 experiments and results from DFT calcu-
lations.
As discussed in the preceding section, the TA spectra cannot
clearly identify the triplet state of Irgacure-2959 and the benzyl
radical owing to their very similar transient absorption. To inves-
tigate the nature and properties of the crucial intermediates gener-
ated from Irgacure-2959, time-resolved resonance Raman experi-
ments were done in aqueous solutions. Figure 4a shows the over-
view of ps-TR3 spectra of Irgacure-2959 in acetonitrile:water=3:7
mixed solution. One species has been observed from 0 ps to 4000
ps. This species appears at approximately 4 ps with the most in-
tense bands at 934, 1157, 1319, 1447 and 1585 cm-1. Initially, it
grows in within 14 ps; later it grows more intense from 14 to 200
ps, and then slightly decays over the rest of delay time. According
to the fs-TA spectra, the growing species obtained in the ps-TR3
spectra after 14 ps can be assigned to the radical species of
Irgacure-2959 due to the timescale of its growth after the observa-
tion of the formation of T1 from S1 by ISC with a 3.8 ps time con-
stant. The triplet species is the precursor of its radical species. The
growth kinetics of this species was fitted by the plotting the inte-
grated area of the resonance Raman spectra at 1590 cm-1 versus
the corresponding delay time as shown in Figure 5. Compared
with kinetics of the fs-TA spectra in the inset plot of Figure 2a,
the growth time constant is 21.7 ps, which is different from the
one (3.8 ps) obtained by fs-TA experiment for the formation of
the T1 species of Irgacure-2959. From the temporal sequence, we
propose that irradiation of Irgacure-2959 first converts fast into its
triplet state through ISC with the time constant of 3.8 ps and then
dissociates into two types of radicals with the time constant of
21.7 ps. Meanwhile, it is noted that in the early time scale (less
than 14 ps) triplet state of Irgacure-2959 has not been easily ob-
served by ps-TR3 spectra, even though, triplet species and benzoyl
radical have similar absorption bands centered around 415 nm.
The 400 nm probe wavelength used in our ps-TR3 experiments
appears to be more sensitive to the benzoyl radical than the triplet
state of Irgacure-2959 and suggests that the resonant enhancement
at this wavelength is stronger for the benzoyl radical than for the
T1 species that is the precursor for the 21.7 ps time constant of
formation of the benzoyl radical. Close inspection of the 0 ps, 10
ps and 14 ps TR3 spectra in Figure 4a shows that the Raman
bands are significantly broader and appear to be sharper at longer
delay times. This is likely due to vibrational cooling which typi-
cally occurs on the tens of ps time-scale. This suggests that the
initial benzoyl radicals are formed vibrationally hot from the dis-
sociation of the T1 state of Irgacure 2959. Since the benzoyl radi-
cal and the T1 state of Irgacure 2959 have the same basic chromo-
phore of benzene and the HO(CH2)O- substituent group and the
vibrationally excited benzoyl radical at 0 ps and 10 ps have weak
resonance Raman signals, a vibrationally excited T1 state of
Irgacure 2959 would also likely be very weakly resonantly en-
hanced using a 400 nm probe wavelength in the ps-TR3 experi-
ments. This suggests that the S1 to T1 ISC observed with a 3.8 ps
time constant produces a vibrationally excited T1 state that then
a
1000ns
500ns
300ns
b
*
*
*
4000 ps
180ns
160ns
140ns
120ns
200 ps
100 ps
80ns
70ns
50ns
50 ps
30ns
20ns
20 ps
14 ps
10ns
0ns
1590
10 ps
0 ps
1298
1326
1162
941
1000
800
1000
1200
1400
1600
1800
800
1200
1400
1600
1800
Raman shift (cm-1
)
Raman Shift (cm-1
)
Figure 4a, b. ps-TR3 and ns-TR3 spectra of Irgacure-2959 in
MeCN:H2O=3:7 solvent obtained with 266 nm excitation wave-
length and 400 nm (ps) and 416 nm (ns) probe wavelength at
various delay times that inserted next to the spectra.
=21.7 7.9 ps
0
50
100
150
200
time delay/ps
Figure 5. Time dependence of the resonance Raman band at 1585
cm-1 for Irgacure-2959 (open circle) in water:acetonitrile=3:7
mixed solution fit by an one exponential function with a ca. 21.7
ps growth time constant. The solid line indicates the kinetics fit-
ting to the experimental data points.
To confirm the proposal that the major growth transient species
in Figure 4a is the benzoyl radical, DFT calculations have been
ACS Paragon Plus Environment

The Journal of Physical Chemistry
Page 4 of 8
carried out to predict the Raman spectra of likely intermediates.
quenched by the radical-quencher under the argon saturated con-
1
2
3
4
5
6
7
8
Figure 6 displays a comparison of the experimental resonance
Raman spectrum of Irgacure-2959 obtained at 50 ps and the DFT
calculated Raman spectrum of the Irgacure-2959 benzoyl radical.
The DFT calculation of the Raman spectrum vibrational bands of
the benzoyl radical exhibits a good agreement with the experi-
mental resonance Raman spectrum and this is the second piece of
evidence that the species growing in after 14 ps can be reasonably
assigned to the benzoyl radical of Irgacure-2959. Inspection of
Figure 6 shows that most of the benzoyl radical Irgacure-2959
Raman bands observed in the ps-TR3 are due to vibrations associ-
ated with the carbonyl C=O stretching vibration modes and ring
breathing and stretching of the aromatic ring. According to the
DFT results, the 941 cm-1 band is mainly due to the C-H bending
modes. The 1157 cm-1 band is associated with the carbonyl C-C
stretching mode located on C4 and C11 (see Figure 7). The 1326
cm-1 band is contributed to by the ring C-C stretching mode. The
most intense vibrational Raman band at 1585 cm-1 comes from the
aromatic ring C-C stretching and carbonyl C=O stretching mo-
tions. Figure 7 shows the optimized structure of the Irgacure-2959
benzoyl radical. The corresponding bond lengths are labeled near
the sides of each bond. The calculated benzoyl radical’s Cartesian
coordinates, sum of the electronic and thermal free energies for
the optimized geometry obtained from the UB3LYP/6-311G**
calculations are displayed in Figure 5S.
dition in the sample cell..
0.9
0.059 mM
0.188 mM
0.421 mM
0.667 mM
0.800 mM
0.6
0.3
0.0
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
0
5
10
15
20
25
delay time/ microsecond
Figure 8. ns-TA kinetics of Irgacure-2959 at 410 nm by adding
quencher with different concentrations normalized quenching
delay time in an argon saturated condition.
12
10
8
6
4
2
0
radical quencher
linear fit
1585
0.0
0.2
0.4
0.6
0.8
[n-methylarylate]/mM
1326
1157
1447
941
Figure 9. Pseudo-first-order decay rate constants (kobs) of the
benzoyl radical versus n-methylacrylate concentration, measured
following laser flash photolysis (266 nm) of argon-saturated
MeCN:H2O=3:7 solution in the presence of different n-
methylacrylate concentrations.
O
OH
O
Benzoyl radical
DFT calculations for the activation energy barrier To fur-
ther verify that cleavage reaction takes place from the triplet state,
DFT calculations were done to examine the activation energy
barrier. The computations were performed at the (U)B3LYP/6-
311G** level of theory. The optimized structures of the reactant
complex (RC), the transition state (TS), and the product complex
(PC) are shown in Figure 10, in which the cleaved C-C bond
length is labeled. The reaction barrier was found to be 5.0
kcal/mol by the DFT calculations. The calculated relative energy
profile for the cleavage reaction is shown in Figure 11. The cleav-
age reaction is initiated through the C-C bond as indicated by
distance labeled in Figure 10. As the reaction goes from the TS to
the PC, the C-C bond is completely cleaved. The reaction energy
barrier is very small, which demonstrates that it is reasonable for
Irgacure-2959 to undergo cleavage through its triplet state. This is
consistent with the experimental results that the triplet state of
Irgacure-2959 undergoes the cleavage reaction efficiently. The
Cartesian coordinates, sum of the electronic and thermal free en-
ergies for the optimized geometry from the UB3LYP/6-311G**
calculations for Irgacure-2959’s transition state of interest in the
paper are given in Figure 6S.
800
1000
1200
1400
1600
1800
Raman Shift (cm-1
)
Figure 6. Comparison the experimental resonance Raman spec-
trum of Irgacure-2959 (top) obtained in the MeCN:H2O=3:7 sol-
vent at 50 ps delay time with the DFT calculated spectrum for the
benzoyl radical species. Dotted lines display the correlation be-
tween the experimental and calculated Raman bands.
1.188
1.402
C11
1.387
1.473
C4
1.401
1.404
1.360
1.403
1.383
1.521
1.438
1.421
Figure 7. Optimized structure of the benzoyl radical of Igacure-
2959 obtained from the UB3LYP/6-311G** calculations. Selected
bond lengths are displayed in the structure.
Radical quenching experiments were carried out to confirm the
hypothesis that the intermediate generated from the triplet is the
benzoyl radical. Figure 8 displays the kinetics versus an increas-
ing concentration of the carbon-centered radical quencher n-
methylacrylate obtained by ns-TA spectra with an argon saturated
solution condition in the sample cell. The results in Figure 8 clear-
ly shows that the life-time of the radical decreased dramatically
by increasing the concentration of the n-methylacrylate quencher.
The related pseudo-first-order decay rate constant (kobs) of the
benzoyl radical versus the n-methylacrylate concentration spec-
trum is displayed in Figure 9. The data can be well fitted linearly
which indicates that the transient species should be directly
RC
TS
PC
Figure 10. The optimized geometries of the RC, TS and PC cal-
culated by B3LYP/6-311G(d,p) for the cleavage reaction of
Irgacure-2959. Selected interatomic distances (in angstrom) are
labeled in the structures.
ACS Paragon Plus Environment

Page 5 of 8
The Journal of Physical Chemistry
sistent with experimental observations. The time-resolved spec-
5.0
TS
1
2
3
4
5
6
7
8
9
troscopic study for Irgacure-2959 reported here presents a clear
photochemical mechanism for Irgacure-2959 irradiated by UV
light in water-rich solutions.
0.0
RC
METHODS
Materials:
The
Irgacure-2959
(2-Hydroxy-4’-(2-
hydroxyethoxy)-2-methylpropiophenone) sample was commer-
cially available and purchased from Sigma-Aldrich Company with
98% purity. It was used as received with acetonitrile and water to
prepare a solution with a volume ratio of 3:7 for the experiments
reported here. The sample solutions of Irgacure-2959 employed in
the picosecond time-resolved resonance Raman (ps-TR3) experi-
ments were prepared at a concentration such that the absorbance
was 2 using a 1 mm thickness quartz cell at the excitation wave-
length of 266 nm.
-10.4
PC
Reaction coordination
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 11. The reactive energy profile obtained from the DFT
calculation for Irgacure-2959 photocleavage reaction.
For a clear view of the different species involved in the whole
photochemical reaction process, Scheme 2 presents an overview
of the mechanism that Irgacure-2959 evolves from ground state
through singlet and triplet to cleave into the radical species after
266 nm laser irradiation. As depicted in Scheme 2, Irgacure-2959
was excited from ground state to its excited singlet state then con-
verted into the triplet state which has an absorption band at 415
nm with a time constant of 4.0 ps. Then a homolysis cleavage
reaction occurred resulting in the generation of benzoyl and ketyl
radicals with a time constant of 27.1 ps.
Fs-TA experiments: The femtosecond time-resolved transient
absorption (fs-TA) experiments were performed using a commer-
cial femtosecond Ti: Sapphire regenerative amplified Ti: Sapphire
laser system from Spectra Physics, Spitfire-Pro and an automated
data acquisition system from Ultrafast System, Helios. The pump
laser pulse wavelength is 267 nm, the third harmonic of the fun-
damental 800 nm from the regenerative amplifier. While the
probe laser pulse was a white-light continuum from 350 nm to
800 nm generated in a CaF2 crystal by using approximately 5% of
the amplified 800 nm output from the Spitfire. The probe beam is
split into two parts before passing through the sample: one goes
through the sample, the other travels directly to the reference
spectrometer that monitors the fluctuations in the probe beam
intensity. The instrument response time is estimated to be 150 fs.
At each of the temporal delay time, the data were averaged for 0.5
s. The maximum extent of the temporal delay time was 3300 ps
for the optical stage used in the experiments. The detailed appa-
ratus and methods for the fs-TA experiments have been discussed
previously.21
Ps-TR3 experiments: Picosecond time-resolved resonance
Raman (ps-TR3) experiments were done using an ultrafast system
in our laboratory. The samples were pumped with a 267 nm laser
pulse and probed with a 400 nm laser pulse. The durations of the
pump and probe lasers were 1.5 ps and the pulse energies focused
onto the sample was 5-10 μJ. The time resolution of the system
was 2 ps and the time delay between the pump and probe pulses
were adjusted using an optical delay line. A thin film stream of
the sample was excited by the focused laser pulses of light and the
signal of the Raman light was collected using a back-scattering
configuration and detected by a liquid nitrogen cooled CCD. De-
tailed of the experimental apparatus and methods employed have
been described previously.26
Ns-TR3 experiments: Nanosecond time-resolved resonance
Raman (ns-TR3) measurements were performed using a pump-
probe apparatus and methods described previously.21 Concisely,
the pump laser pulse wavelength of 266 nm was generated from
the fourth harmonic of a Nd:YAG nanosecond pulsed laser. The
probe laser pulse wavelengths of 354.7 nm for Irgacure-2959 in
pure acetonitrile and 416 nm for Irgacure-2959 in acetoni-
trile:water with volume ration both 1:1 and 1:9 solutions were
both generated from the third harmonic of the Nd:YAG nanosec-
ond pulsed laser and the first Stokes hydrogen shifter laser line of
the third harmonic. The energy of the laser used in the experi-
ments was in a range from 2.5 to 3.5 mJ with a 10 Hz repetition
rate. These two lasers were synchronized by a pulse delay genera-
tor to electronically control the pump and probe lasers time delay
that was monitored by a fast photodiode and 500 MHz oscillo-
scope. The apparatus time resolution was about 10 ns. The pump
and probe laser beams were optically aligned and focused onto a
spot of a running sample solution stream so that these two laser
This is the first time the benzoyl radical and its dynamics in
water-rich solution was characterized by time resolved resonance
Raman vibrational spectroscopy. The results show that in water,
the benzoyl radicals are efficiently generated with a time constant
of 21.7 ps. This provides significant information for the reactive
intermediates with its corresponding kinetics involved in the pho-
tocleavage reaction of Irgacure-2959.
Scheme 2. Proposed Photocleavage Reaction Mechanism of I-
2959 in water:acetonitrile mixed solution.
CONCLUSION
The photocleavage reaction mechanism of Irgacure-2959 was
explored by using the fs-TA, ns-TA, ps-TR3 and ns-TR3 spectros-
copies. In acetonitrile-water mixed solution, the singlet excited
state converts into a triplet state through a highly efficient ISC
process with a time constant of 4.0 ps. The triplet state of
Irgacure-2959 has a strong absorption bands at 415 nm. The tran-
sient absorption band at 415 nm slightly decays and has a red-shift
to 417 nm, this process is confirmed to be the photocleavage of
the triplet state of Irgacure-2959 which results in the generation of
benzoyl radicals that appear to be vibrationally excited when they
are first produced. The comparison of the resonance Raman spec-
tra from ps-TR3 spectra and the DFT calculated Raman spectrum
of the benzoyl radical further supports the assignment of the ben-
zoyl radical in the aqueous solution. Radical quenching experi-
ments done using ns-TA detection were also carried and these
results also supported that Irgacure-2959 cleaves to produce a
benzoyl radical. DFT calculations indicated that the activation
energy barrier for the cleavage of Irgacure-2959 is 5.0 kcal/mol,
this is complementary evidence to support that the triplet state of
Irgacure-2959 undergoes the cleavage at an ultrafast rate con-
ACS Paragon Plus Environment

The Journal of Physical Chemistry
Page 6 of 8
pulses were spatially overlapped on the sample solution. The scat-
(2) Vink, C. B.; Woodward, J. R. Effect of a weak magnetic field
1
2
3
4
5
6
7
8
tered Raman light was collected by using reflective optics into a
spectrometer whose grating dispersed the light onto a charge-
coupled device (CCD). The Raman signal was accumulated for 30
s by the CCD before being read out and stored to a computer. The
spectra included here were obtained by the subtraction of a reso-
nance Raman spectrum with a negative time delay of the probe
before the pump -100 ns spectrum from the resonance Raman
spectrum obtained with a positive time delay of pump before
probe spectrum. The known acetonitrile solvent’s Raman bands
were used to calibrate Raman shifts with an approximate accuracy
of 5 cm-1.
on the reaction between neutral free radicals in isotropic solution.
J. Am. Chem. Soc. 2004, 126, 16730-16731.
(3) Maliakal, A.; Weber, M.; Turro, N. J.; Green, M. M.; Yang, S.
Y.; Pearsall, S.; Lee, M.-J. Chemically Induced Dynamic Electron
Polarization Studies of a pH-Dependent Free Radical Cage
Formed in a Photoinitiator Labeled Poly(methacrylic acid)
Macromolecules 2002, 35, 9151-9155.
(4) McGilvray, K. L.; Decan, M. R.; Wang, D.; Scaiano, J. C.
Facile Photochemical Synthesis of Unprotected Aqueous Gold
Nanoparticles J. Am. Chem. Soc. 2006, 128, 15980-15981.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Density Function Theory (DFT) computations were em-
ployed to study the properties of Irgacure-2959 for comparison
purposes to make assignments of the intermediates generated in
the excited states in various solvents. The optimized geometries,
vibrational modes and the vibrational frequencies for the different
species were obtained from DFT calculations that employed a
(U)B3LYP/6-311G** basis set. No imaginary frequency modes
were observed at the stationary states of the optimized structures
presented here. The default G03 method was used to obtain the
Raman spectra to determine the Raman intensity and shifts. A
factor of 0.975 were used to scale the calculations of the Raman
frequencies by the DFT calculations with (U)B3LYP/6-311G**
basis set for comparison with the experimental Raman results in
order to achieve a better understanding of the experimental data
assignments. For the reaction energy profile for the Irgacure-2959
photocleavage reaction calculated from DFT computations, the
sum of the electronic and thermal free energies of the TS (transi-
tion state) was subtracted from that of the RC (reactant complex)
to obtain the energy barrier between the TS and the RC. We also
obtained the energy barrier between the PC (product complex)
and the RC in a similar manner. All of the calculations were done
using the Gaussian 03 program suite.27
(5) Marn, M. L.; McGilvray, K. L.; Scaiano, J. C. Photochemical
Strategies for the Synthesis of Gold Nanoparticles from Au(III)
and Au(I) Using Photoinduced Free Radical Generation. J. Am.
Chem. Soc. 2008, 130, 16572-16584.
(6) Scaiano, J. C.; Stamplecoskie, K. G.; Hallett-Tapley, G. L.
Photochemical Norrish type
metal nanoparticlesynthesis:
I
reaction as
a
tool for
importance
of proton coupled electron transfer Chem. Commun. 2012, 48,
4798-4808.
(7) Stamplecoskie, K. G.; Scaiano, J. C. Silver as an Example of
the Applications of Photochemistry to the Synthesis and Uses of
Nanomaterials. Photochemistry and Photobiology 2012, 88, 762-
768.
(8) Jockusch, S.; Landis, M. S.; Freiermuth, B.; Turro, N. J.
Photochemistry and Photophysics of R-Hydroxy Ketones.
Macromolecules 2001, 34, 1619-1626.
(9) Jockusch, S.; Turro, N. J. Radical Addition Rate Constants to
Acrylates and Oxygen: R-Hydroxy and R-Amino Radicals
Produced by Photolysis of Photoinitiators. J. Am. Chem. Soc.
1999, 121, 3921-3925
ASSOCIATED CONTENT
Supporting Information Available: The predicted absorption
spectrum of singlet state of Irgacure-2959 is given. The DFT cal-
culated frontier orbitals of Irgacure-2959 are provided. The Kinet-
ic of Irgacure-2959 in MeCN:H2O=3:7 solvent is given. Compari-
son the experimental resonance Raman spectrum of Irgacure-2959
obtained in the MeCN solvent with the experimental resonance
Raman spectrum of Irgacure-2959 in the MeCN:H2O=3:7 solvent
are given. The Cartesian coordinates are provided. The Irgacure-
2959 transition state Cartesian coordinates are given.
(10) Colley, C. S.; Grills, D. C.; Besley, N. A.; Jockusch, S.;
Matousek, P.; Parker, A. W.; Towrie, M.; Turro, N. J.; Gill, P. M.
W.; George, M. W. Probing the Reactivity of Photoinitiators for
Free
Radical
Polymerization:
Time-Resolved
Infrared
Spectroscopic Study of Benzoyl Radicals. J. Am. Chem. Soc.
2002, 124, 14952-14958.
(11) Jockusch, S.; Turro, N. J. Phosphinoyl Radicals: Structure
and Reactivity. A Laser Flash Photolysis and Time-Resolved ESR
Investigation. J. Am. Chem. Soc. 1998, 120, 11773-11777.
Corresponding Author
(12) Jockusch, S.; Koptyug, I. V.; McGarry, P. T.; Sluggett, G.
W.; Turro, N. J.; Watkins, D. M. A Steady-State and Picosecond
Pump-Probe Investigation of the Photophysics of an Acyl and a
Bis(acyl)phosphine Oxide. J. Am. Chem. Soc. 1997, 119, 11495-
11501.
*Phone: +852-2859-2160. Fax: +852-2857-1586.
E-mail: phillips@hku.hk; mdli@hku.hk
ACKNOWLEDGEMENT
This work was supported by a grant from the Research Grants
Council of Hong Kong (HKU 7048/11P) and the University
Grants Committee Special Equipment Grant (SEG-HKU-07 to
DLP.
(13) Fisher, H.; Baer, R.; Hany, R.; Verhoolen, I.; Walbiner, M.
2,2-Dimethoxy-2-phenylacetophenone: photochemistry and free
radical photofragmentation. J. Chem. Soc. Perkin Trans. 1990, 2,
787-798.
REFERENCES
(14) Sheehan, J. C.; Umezawa, K. Phenacyl photosensitive
blocking groups. J. Org. Chem. 1973, 38, 3771-3774.
(1) Diau, E. W.-G.; Kötting, C.; Zewail, A.H. Femtochemistry of
Norrish type-I reactions: II. The anomalous predissociation
dynamics of cyclobutanone on the S-1 surface. ChemPhysChem.
2001. 2, 273-393.
(15) Epstein, W. W.; Garrossian, M. p-Methoxyphenacyl esters as
photodeblockable protecting groups for phosphates. J. Chem.
Soc., Chem. Commun.1987, 532-533.
ACS Paragon Plus Environment

Page 7 of 8
The Journal of Physical Chemistry
(16) Baldwin, J. E.; McConnaughie, A. W.; Moloney, M. G.;
Pratt, A. J.; Shim, S. B. New photolabile phosphate protecting
group. Tetrahedron 1990, 46, 6879-6884.
1
2
3
4
5
6
7
8
(17) Li, M. D.; Yeung, C. S.; Guan, X.; Li, W.; Ma, J.; Ma, C.;
Phillips, D. L. Water- and Acid-Mediated Excited-State
Intramolecular Proton Transfer and Decarboxylation Reactions of
Ketoprofen in Water-Rich and Acidic Aqueous Solutions. Chem.
Eur. J. 2011, 17, 10935-10950.
(18) Kwok, W. M.; Ma, C.; Parker, A. W.; Phillips, D.; Towrie,
M.; Matousek, P.; Phillips, D. L. Picosecond time-resolved
resonance Raman observation of the iso-CH2I–I photoproduct
from the “photoisomerization” reaction of diiodomethane in the
solution phase. J. Chem. Phys. 2000, 113, 7471-7478.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(19) Kwok, W. M.; Ma, C.; Matousek, P.; Parker, A. W.; Phillips,
D.; Toner, W. T.; Towrie, M.; Zuo, P.; Phillips, D. L. Time-
resolved spectroscopy study of the triplet state of 4-
diethylaminobenzonitrile (DEABN). Phys. Chem. Chem. Phys.
2003, 5, 3643-3652.
(20) Ma, C.; Kwok, W. M.; Matousek, P.; Parker, A. W.; Phillips,
D.; Toner, W. T.; Towrie, M. Time-Resolved Study of the Triplet
State of 4-dimethylaminobenzonitrile (DMABN). J. Phys. Chem.
A 2001, 105, 4648-4652.
(21) Li, M.-D.; Ma, J.; Su, T.; Liu, M. Y.; Yu, L. H.; Phillips, D.
L. Direct Observation of Triplet State Mediated Decarboxylation
of the Neutral and Anion Forms of Ketoprofen in Water-Rich,
Acidic, and PBS Solutions. J. Phys. Chem. B 2012, 116,
5882−5887.
(22) Li, M.-D.; Su, T.; Ma, J.; Liu, M.; Liu, H.; Li, X.; Phillips, D.
L. Phototriggered Release of a Leaving Group in Ketoprofen
Derivatives via a Benzylic Carbanion Pathway, But not via a
Biradical Pathway. Chem. Eur. J. 2013, 19, 11241-11250.
(23) Li, M.-D.; Ma, J.; Su, T.; Liu, M.; Phillips, D. L. A time-
resolved spectroscopy and density functional theory study of the
solvent dependent photochemistry of fenofibric acid. Phys. Chem.
Chem. Phys. 2013, 15, 1557-1568.
(24) Ma, C.; Kwok, W. M.; Chan, W. S.; Du, Y.; Kan, J. T. W.;
Toy, P. H.; Phillips, D. L. Ultrafast Time-Resolved Transient
Absorption and Resonance Raman Spectroscopy Study of the
Photodeprotection and Rearrangement Reactions of p-
Hydroxyphenacyl Caged Phosphates. J. Am. Chem. Soc. 2006,
128, 2558-2570.
(25) An, H.-Y.; Kwok, W. M.; Ma, C.; Guan, X.; Kan, J. T. W.;
Toy, P. H.; Phillips, D. L. Photophysics and Photodeprotection
Reactions of p-Methoxyphenacyl Phototriggers: An Ultrafast and
Nanosecond Time-Resolved Spectroscopic and Density
Functional Theory Study. J. Org. Chem. 2010, 75, 5837-5851.
(26) Ma, C.; Chan, W. S,; Kwok, W. M.; Zuo, P.; Phillips, D. L.
Time-Resolved Resonance Raman Study of the Triplet State of
the p-Hydroxyphenacyl Acetate Model Phototrigger Compound. J
Phys. Chem.B 2004,108, 9264-9276.
(27) Frisch, J. N.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.;
Robb, M. A.; Cheeseman, J. R.; Montgomery, J. A.; Vreven, T.;
Kudin, K. N.; Burant, J. C.; et al. Gaussian 03; Gaussian, Inc.:
Wallingford, CT, 2004.
ACS Paragon Plus Environment

The Journal of Physical Chemistry
Page 8 of 8
1
2
Table of Contents
3
4
5
6
Time-Resolved Spectroscopic and Density Functional Theory Study of the Photochemistry of
Irgacure-2959 in an Aqueous Solution
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
266 nm
ACS Paragon Plus Environment