Monomer- and polymer radicals of vinyl compounds: EPR and DFT studies of geometric and electronic structures in the adsorbed state

Article abstract of DOI:10.1016/j.saa.2012.08.054

Electron paramagnetic resonance (EPR) spectroscopy was applied to study the paramagnetic species formed from styrene, 1,1-diphenylethylene, α-methylstyrene, β-methylstyrene and methylmethacrylate adsorbed on amorphous silica gel after γ-irradiation at 77

Full text of DOI:10.1016/j.saa.2012.08.054

Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 98 (2012) 367–377
Contents lists available at SciVerse ScienceDirect
Spectrochimica Acta Part A: Molecular and
Biomolecular Spectroscopy
journal homepage: www.elsevier.com/locate/saa
Monomer- and polymer radicals of vinyl compounds: EPR and DFT studies
of geometric and electronic structures in the adsorbed state
Anders Lund , Marek Danilczuk b
a,
⇑
a
Department of Physics, Chemistry and Biology, IFM, Linköping University, S-581 83 Linköping, Sweden
Institute of Nuclear Chemistry and Technology, Dorodna 16, 03-195 Warsaw, Poland
b
h i g h l i g h t s
g r a p h i c a l a b s t r a c t
H atoms released by irradiation of surface silanol groups initiate radical formation in methylmethacry-
late, MMA. The radical attacks an MMA molecule at increased temperature forming a propagating radical
identified by EPR.
"
"
"
"
Radicals of vinyl compounds
adsorbed on silica gel were studied
by EPR.
Radicals were mainly formed by
addition of H atoms originating from
silanol groups.
The radical-surface complexes were
theoretically investigated by DFT
calculations.
The formation of a propagating
radical observed by EPR was
theoretically modeled.
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 1 June 2012
Received in revised form 10 August 2012
Accepted 20 August 2012
Available online 1 September 2012
Electron paramagnetic resonance (EPR) spectroscopy was applied to study the paramagnetic species
formed from styrene, 1,1-diphenylethylene,
a-methylstyrene, b-methylstyrene and methylmethacrylate
adsorbed on amorphous silica gel after -irradiation at 77 K. Radicals formed by the hydrogen atom addi-
c
tion at the vinyl group of the monomers were observed in all samples. The hydrogen atoms were shown
to originate to a large extent from the adsorbent by using silica gel with deuterated silanol groups. An EPR
spectrum assigned to a propagating radical was observed at increased temperature for samples contain-
ing methylmethacrylate (MMA). The structures of the adsorption complexes, the respective hyperfine
splitting constants and the adsorption energies were calculated by applying DFT quantum chemical
methods. The reaction between an MMA molecule and the MMA radical and the structure of the propa-
gating radical was modeled. The calculated hyperfine splitting constants for all radicals confirmed the
assignment of the experimental spectra.
Keywords:
Vinyl radicals
Adsorption
EPR
DFT modeling
Propagating radicals
Spectrum simulation
Ó 2012 Elsevier B.V. All rights reserved.
Introduction
e.g. zeolites, alumina, aluminosilicalites or higher-order oxides are
attractive materials in numerous applications in catalysis. Previous
Free radical reactions of organic substances adsorbed on porous
solids are of particular interest from the perspective of both mate-
rials science and heterogeneous catalysis. High surface area oxides
research has shown that those materials play a significant role in
the separation and stabilization of various reactive intermediates
as atoms, radicals, radical ions, and clusters [1–7]. Studies of such
intermediates are very important for a better understanding of the
mechanisms of chemical reactions in polymerization, heteroge-
neous catalysis, photochemistry, and radiation chemistry and
⇑
Corresponding author.
E-mail address: alund@ifm.liu.se (A. Lund).
1
386-1425/$ - see front matter Ó 2012 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.saa.2012.08.054

3
68
A. Lund, M. Danilczuk / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 98 (2012) 367–377
biology. However, in comparison with the reactions in the gas or
the liquid phase the corresponding processes in such systems are
less well understood.
A number of experimental techniques has been applied to the
characterization of the bonding interactions between the adsorbed
vinyl monomers and the surface of catalysts. The early studies of
radiation induced radical polymerization have focused on the graft
polymerization on the solid supports. These early results showed
that only free radicals in the track of fast electrons were able to ini-
tiate polymerization process [8–10].
understanding of the experimental results was supported by DFT
calculations of hyperfine splitting constants of radicals formed
from the vinyl monomers adsorbed on the silica surface, adapting
a previous procedure [29]. The formation of a propagating radical
in methylmetacrylate was investigated theoretically as well as
experimentally.
Experimental
Sample preparation
Due to the well defined structure, acidic and ion-exchange
properties, thermal and chemical stability, microporous natural
and synthetic zeolites have been considered more recently as cat-
alytic systems for controlled polymerization. Studies of vinyl
ethers adsorbed on acidic mordenite [11] and faujasite (H-Y) [12]
have shown that adsorbed monomers react near room temperature
to produce low molecular weight polymers. It has also been found
that monomers migrate to the catalyst–polymer interface and
polymerize there. Thermogravimetric analyses of zeolites and
polymer composites showed that only a very small fraction of
the polymer can be found inside the zeolite channels of mordenite
while in faujasite with larger pore size the amount of polymer is
significant [11,12].
Electron paramagnetic resonance (EPR) and nuclear magnetic
resonance (NMR) studies of radiation induced postpolymerization
of acronitrile (AN) and methyl methacrylate (MMA) adsorbed on
zeolite 13X show that overall postpolymerization kinetics may be
accounted for by assuming an exponential decay of radical propa-
gation and recombination reactions [13]. Studies of methylacety-
lene (MA) gas reaction in acid forms of the mordenite, omega, L,
Y, beta, ZSM-5 and SAPO-5 molecular sieves indicate that MA re-
acts with acid sites of zeolites and forms large, conjugated poly-
meric species within the zeolites channels [14].
2
Silica gel with a specific surface area of 667 m /g determined by
the BET method was obtained commercially (AB Kebo, Stockholm).
The main impurities determined by spectral analysis were Al
(350 ppm) and Fe (136 ppm). Further treatment by reflux boiling
2
in D O was employed in some experiments in order to replace
the hydrogen atoms of surface silanol groups with deuterium
[30]. Methyl methacrylate (MMA) and styrene (ST) were distilled
after removing the stabilizer just before use. All other chemicals,
obtained from Fluka AG, were used as received. The silica gel sam-
ples were placed in 4 mm EPR sample tubes made of Suprasil
quartz and dehydrated at 500 °C for 17 h in vacuo to a final pres-
sure below 0.01 Pa. Additives of previously degassed vinyl mono-
mers in the range 1–5 wt.% were then vacuum-distilled onto the
silica gel as described previously [28]. The samples were sealed un-
6
0
der vacuum and irradiated with Co
77 K.
c-rays to a dose of 34 kGy at
EPR measurements
The EPR spectra were recorded at X-band, using a solution of
Fremy’s salt for magnetic field calibration [31]. The g-factors were
determined relative to g = 2.0036 of DPPH kept in a double cavity
together with the sample under investigation. Measurements at
higher temperatures were made by warming the samples in a var-
iable temperature accessory attached to the EPR spectrometer.
Since the MCM-41 mesoporous silica (with
a hexagonal
arrangement of the pores with 2–5 nm diameter) was invented
at the beginning of the 1990’s [15–17], applications of mesoporous
molecular sieves in controlled polymerization have attracted much
attention [18,19]. In contrast to zeolites these materials have the
ability to encapsulate relatively large molecules.
Simulations
Radical polymerization of MMA with benzoyl peroxide within
the uniform channels of the mesoporous silica MCM-41 gives
poly(methyl methacrylate) (PMMA) with a much higher molecular
weight compared to those obtained in solution under identical
conditions [20]. Polymerization of MMA catalyzed by samarocene
complex supported on a series of mesoporous molecular sieves
with various pores size provide highly syndiotactic PMMA’s. Com-
pared to polymerization in solution, polymerization of MMA in
mesoporous silica provides PMAA with higher molecular weight
The experimental spectra were analyzed by previously de-
scribed software [32,33] modeled after the classical program by
Lefebvre and Maruani for the calculation of EPR spectra of free rad-
icals in amorphous systems [34]. The hyperfine splittings em-
ployed in the simulations were estimated by a combination of
the data obtained from the DFT calculations reported here, and a
manual stick-plot analysis.
Quantum chemistry calculations
[
21].
MCM-41 along with HY zeolite has been widely used in cationic
Geometry optimization calculations for minimum energy and
transition-state structures were performed using DFT theory. Iso-
tropic and anisotropic hyperfine splittings of the radicals formed
from the vinyl monomers were calculated taking into account
the interaction with the silica surface, similar to the procedure
used in [29]. Details regarding the used theoretical methods are
presented in the Computations section. The anisotropic splittings
are given in the Supplemental information.
polymerization of vinyl monomers. Polymerization of vinyl mono-
mers can be initiated by acidic protons in the H-form of the zeolite
or by co-adsorbed arylmethylium ions in MCM-4 silica [22]. In or-
der to control polymerization reactions, the surface of the MCM-41
mesoporous silica can be modified (functionalized) with different
organic groups [23].
Amorphous silica is one of the most common supports in catal-
ysis and was the subject of several experimental studies during the
past decades [4,5,24]. Silica gel support systems are widely used in
studies of transition metal ions or complexes [25] and are success-
fully applied in controlled polymerization [26] and in studies of io-
nic radicals [27,28]. High resolution of the EPR spectra indicated a
high degree of mobility for benzene cation radicals in an adsorbed
layer.
Results and discussion
EPR
Radicals formed by hydrogen addition to the vinyl monomers
were the main species observed by EPR after
c-irradiation at
Here we present EPR studies of radicals formed by H-atom addi-
tion to vinyl monomers adsorbed on amorphous silica surface. The
77 K of the vinyl monomers adsorbed on silica gel. A correlation
between the position of H atom addition and the free valence index

A. Lund, M. Danilczuk / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 98 (2012) 367–377
369
[
35] was observed as illustrated in Scheme 1 for styrene, using the
expression due to Burkitt et al. [36] to obtain the numerical values.
The isotropic hyperfine splitting constants of the spectra are gi-
ven in Table 1. The g-factors were in the range 2.0023–2.0027 with
an estimated uncertainty of ±0.0003. The anisotropic components
used in the simulations to support the assignments were taken
from the DFT calculations presented in the Supplemental
information.
An EPR spectrum assigned to a propagating radical was ob-
served at increased temperature for samples containing methyl
methacrylate (MMA), but not for the other monomers. Several of
the samples were colored after irradiation. It is possible that the
color is caused by ionic radicals. Radical cations of adsorbed unsat-
urated compounds have been observed previously [5, 27, 28, 37–
3
9]. This possibility was only briefly considered here, however. De-
tailed interpretations are presented below.
2
Styrene/silica gel (ST/SiO )
Fig. 1. (a) experimental EPR spectrum from a
styrene adsorbed on silica gel
spectra for the hydrogen atom addition radical CH
c
-irradiated sample containing 5%
The EPR spectrum in Fig. 1a recorded at 77 K from a
sample containing 5 wt.% of styrene adsorbed on silica gel is simi-
lar to that obtained by the -radiolysis of -phenylethylchloride in
-methylpentane glass [40]. The latter spectrum was assigned to
the radical CH HC . The same radical is formed here by hydro-
C˙
gen atom addition to the vinyl group as shown in Scheme 1. The
hyperfine structure is attributed to the H and methyl protons.
Smaller splittings of ca. 5 G are assigned to the ring protons at posi-
tions 1, 3 and 5. The splittings at positions 2 and 4 (Scheme 1) were
not resolved.
c
-irradiated
c
-irradiated and recorded at 77 K, (b–d) simulated
C˙
HC with
3
a
6
H ,
5
a
CH3
¼
14:5; 15:5 and 16:5 G, respectively.
c
a
3
3
a
6 5
H
Hb2 and Hb3 in Table 2 yielded theoretical values in the range
16.6–18.9 G. Similar values were derived from McConnell’s rela-
a
2
tion [41] a
tings (a
assuming an averaging due to the free rotation usually occurring
along the >C –CH bond in this type of radicals [42]. The anisot-
b
¼ A þ B cos h for b-protons using the theoretical split-
b
) and dihedral angles (h) of the methyl protons, and
Theoretical values of the anisotropic (dipolar) hyperfine split-
tings obtained from the DFT calculations were employed for the
simulations in Fig. 1b–d, while the isotropic values were slightly
adjusted to obtain a satisfactory fit to the experiments. For the
a
3
ropy of the methyl splitting was somewhat averaged by the rota-
tion to become axially symmetric. The isotropic splittings due to
the ring protons were fixed in the simulations with the values
CH
3
protons values in the range 14.5–16.5 G were used in the sim-
a
H1 = aH3 = aH5 = (ꢀ)5 G. For the H
a 3
and CH protons uncertainties
ulations, while an arithmetic average of the splittings due to Hb1
of the order 1–2 G of the isotropic values may occur.
0
.3947
0.4432
0.8207
H
C
2
4
1
5
C
H
H
H
0
.4148
3
C
Hα
0.4148
pffiffiffi
P
Scheme 1. Formation of CH
3
C˙
i j
the bond order for the C –C bond in Hückel molecular orbital theory.
a
HC
H
6 5
radical by hydrogen atom addition to styrene at the position of highest free valence index, defined by F
i
¼
3 ꢀ jPij, [36] where Pij is
Table 1
Isotropic hyperfine splitting constants (G, 1 G = 0.1 mT) of radicals in c-irradiated vinyl monomers adsorbed on the surface of silica gel.
Adsorbate
Styrene
Amount (wt.%)
5
Temperature (K)
77
Assignment
Hyperfine splitting (G)
CH
3
C˙
a
a
HC
6
H
5
a
a
a
a
a
a
a
a
a
a
a
a
a
a
= 16.5
CH3 ¼ 16:5
ring = 5.0 (3H)
CH3 ¼ 16:0
1
a
,1-Diphenyl–ethylene
-Methyl-styrene
3
3
77
77
CH
3
C˙
(C
C˙
6
H
5
)
2
(CH
3
)
2
a
C
6
H
5
CH3 = 15.0
ring = 5.0 (3H)
a
= 15.6
b1 = 31.7, ab2 = 15.6
ring = 4.0 (3H)
CH3 = 22.0
CH3 = 22.2
b1 = 14.7
b2 = 7.5
b-Methyl-styrene
3
77
CH
3
C
b
H
2
C˙
a
HC
6
H
5
Methyl-methacrylate
Methyl-methacrylate
3
3
77
123
(CH
3
)
2
C˙
C˙
a
COOCH
(CH
3
(R)
)COOCH
RC
b
H
2
a
3
3

3
70
A. Lund, M. Danilczuk / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 98 (2012) 367–377
Table 2
DPE cation was considered, but was not supported by the DFT cal-
culations in the Supplemental information. The signals decayed
around 250 K without changing the line shape.
Calculated isotropic hyperfine splitting constants (G) of the H atom addition radical of
styrene adsorbed on the 8T silica model cluster.
M05-2X
-31 + G(d)
B3LYP
6
EPR-III
6-31 + G(d)
EPR-III
a-Methyl styrene (AMST) and b-methyl styrene (BMST)
H
H
H
H
H
H
H
H
H
a
ꢀ18.40
15.62
3.51
33.06
ꢀ5.30
2.31
ꢀ12.50
14.86
2.98
31.82
ꢀ3.58
1.01
ꢀ18.04
16.52
3.61
34.55
ꢀ4.94
1.88
ꢀ15.49
17.07
3.47
36.08
ꢀ4.13
1.16
The resolution of the spectra from samples of -methyl styrene
a
b1
b2
b3
1
(AMST) and b-methyl styrene (BMST) shown in the Supplemental
information (Fig. 1S) was relatively poor, in part attributed to the
overlap with radiation induced EPR signals in silica gel [43].
The EPR spectrum from a sample containing 3% AMST was
2
3
ꢀ5.72
ꢀ4.16
ꢀ5.77
ꢀ5.12
3 2 6 5
attributed to the radical (CH ) H superimposed on a broad
C˙ C
4
2.33
1.02
1.90
1.17
5
ꢀ5.39
ꢀ3.60
ꢀ5.21
ꢀ4.35
background. The isotropic splitting, aCH3 ¼ 15:0 G, estimated from
a stickplot analysis deviated only slightly from the values 15.8–
1
7.9 G derived from the DFT calculations in Table 4. Rapid rotation
about the > C˙ –CH
bond was assumed. The isotropic splittings of
1
,1-Diphenyl ethylene (DPE)
3
the phenyl proton of (ꢀ)5 G also agreed quite well with the corre-
sponding calculated values. The agreement thus supports the
The analysis of the spectrum in Fig. 2a from a sample containing
wt.% of 1,1-diphenylethylene adsorbed on silica gel -irradiated
assignment to the (CH
to the vinyl group of
3
)
a
2
C˙ C
-methyl styrene. The obtained constants
6 5
H radical formed by H atom addition
5
c
and recorded at 77 K was less straight-forward. A quartet of lines
was observed as expected for the radical (C formed by
C˙ CH
might, however, be less accurate than for the styrene case due to
the influence of the background signal on the shape of the spec-
trum. A simulation is shown in the Supplemental Information.
The EPR spectrum from a sample containing 3% BMST shown in
the Supplemental Information was attributed to the radical CH3-
6
H
)
5 2
3
hydrogen atom addition to the methylene group. The observed
four lines were not equidistant, however, and the relative intensity
of the lines deviated from the 1:3:3:1 ratio due to the hyperfine
structure of the three equivalent protons of the methyl group. At-
tempts to simulate the spectrum taking into account the hyperfine
anisotropy of those protons and of the additional structure due to
the ring protons were not successful. It seemed reasonable, how-
ever, that the two lines on the wings were due to the outer hyper-
fine lines of the methyl group of this radical, while the two inner
lines overlapped with those of a second species. The isotropic split-
CH
the
2
C˙
c
a
HC
6 5
H radical superimposed on the EPR signals present in
-irradiated silica gel. According to this interpretation the ob-
served splittings of 4 G is attributed to the ring protons, while the
overall width of ca. 80 G is accounted for by the hyperfine split-
tings with one H
analysis one of the b-protons has a hyperfine splitting twice as
large as that of the other b-proton and the -proton [44]. The good
a b
and two H protons. According to a previous
6 5 2 3
ting of the methyl group of the (C H ) radical was accord-
C˙ CH
a
agreement between the hyperfine splittings obtained from the
stickplot analysis reproduced in the Supplemental information
with those obtained by DFT calculations supports the assignment.
Attempts to obtain more precise values for the splittings by simu-
lations were unsuccessful. Unequal b-proton splittings may also
ingly adjusted to aCH3 ¼ 16:0 G to fit the two outer lines of the
experimental spectrum. The anisotropy of this splitting and the
phenyl protons were estimated by the DFT calculations presented
in the Supplemental information and taken into account in the
simulations, even though the splittings of the latter protons were
not resolved experimentally. The spectrum in Fig. 2c was addition-
ally employed to obtain a fitting (dotted line in Fig. 2a) to the
experimental spectrum. The assignment of this component to the
3 a 2 6 5
occur in the radical CH HCH C H . This alternative with hydro-
C˙
gen atom addition at the b-position of the vinyl group was consid-
ered less likely, however, since the DFT calculations of this species
could not account for the observed 4 G splitting attributed to the
ring protons. It is therefore concluded that free radical formation
occurs by hydrogen atom addition at the position of highest free
valence index in this case as well.
Methyl methacrylate (MMA)
The EPR spectrum from a sample containing 3 wt.% MMA ad-
sorbed on silica gel
seven equidistant lines split by 22 G. The spectrum was assigned
to the radical (CH . Similar spectra have been observed
C˙ COOCH
c-irradiated and recorded at 77 K, Fig. 3a, has
3
)
2
3
in pure irradiated MMA and PMMA [45–47]. The radiation induced
reactions are similar, the main net reaction at 77 K being a hydro-
gen atom addition to the vinyl group. It was observed, however,
that the atoms that take part in the reaction originate mainly from
the hydroxyl groups on the surface of the silica gel. This was veri-
fied by an experiment in which a sample of MMA adsorbed on deu-
terated silica gel was irradiated. Fig. 3b shows the spectrum
obtained from this system. It is composed of the ordinary 7-line
spectrum due to (CH
splitting constant of 22 G. The 6-line spectrum is assigned to the
radical CDH )COOCH . The five methyl protons give rise to
C˙ (CH
the sextet, while the hyperfine structure of the added deuterium
atom is too small to be resolved. The sextet spectrum (b–a) in
Fig. 3 was obtained by subtracting (a) from (b). The D atoms used
in the formation of this radical originate from deuterated –OD
3 2 3
) and a 6-line structure with a
C˙ COOCH
2
3
3
Fig. 2. (a) Experimental EPR spectrum (solid line) from a sample containing 5% of
1
,1-diphenyl ethylene (DPE), adsorbed on silica gel
c-irradiated and recorded at
7
7 K, and fitted spectrum (dotted line) by superposition of (b) and (c), (b) simulated
spectrum due to DPE H-addition radical, (C
synthetic spectrum of another species assumed in the simulation of (a).
6
H
)
5 2
C
˙
3
CH , with a
CH3
¼ 16:0 G, (c)

A. Lund, M. Danilczuk / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 98 (2012) 367–377
371
Fig. 5. (a) Experimental EPR spectrum of a sample
c
-irradiated at 77 K containing
3
% MMA adsorbed on silica gel, recorded at 123 K, (b) simulated spectrum with
CH3 ¼ 22:2 G, ab1 = 14.7 G, ab2 = 7.5 G at an exchange rate of 32 MHz for the b-
protons of the CH group in the radical RC (CH )COOCH R = (CH
C˙ C˙
COOCH , (c) simulated rigid limit spectrum.
a
2
b
H
2
a
3
3
,
3 2 a
)
3
Fig. 3. EPR spectra from samples
MMA adsorbed (a) on ordinary (b) on partly deuterated silica gel. The difference
spectrum (b–a) was attributed to the radical CDH )COOCH obtained by D
C˙ (CH
c-irradiated and recorded at 77 K containing 3%
2
3
3
atom addition to MMA. Weak lines attributed to unreacted D are marked by arrows.
indicated a conformational change with the chain length. Spectra
attributed to a very short PMMA chain probably consisting of the
initiating radical and one monomer unit were observed [51].
The spectrum observed in this work at increased temperature
silanol groups on the surface of the adsorbent. There might be two
reasons for the observation of the 7-line spectrum at the same
time. Firstly, the deuteration of the silica gel might not be complete
and secondly H atoms can also be formed by the radiolysis of the
monomer as indicated by the observation of the 7-line spectrum
in irradiated pure MMA [45–47].
3 2-
can be accounted for by the attachment of the initial radical (CH )
CCOOCH
˙
3
, (R), to the vinyl group of MMA forming a secondary rad-
ical with the structure RC
b
2 a 3 3
H C (CH )COOCH . EPR spectra of this
˙
type of radicals observed in the solid phase with hyperfine split-
tings aCH3 ¼ 22:2 G to the methyl protons and ab1 = 14.7 G,
ab2 = 7.5 G of the methylene group [50] are very similar to that in
Fig. 4b. The theoretical values, obtained by the DFT calculations
in the Computation section (Table 6) also support the assignment.
Thus, values in the range 19.0–21.9 G were predicted for the split-
Radical reactions
The MMA spectrum observed at 77 K changed to the 9 line pat-
tern in Fig. 4 at 123 K. Similar irreversible changes have been ob-
served for example in the benzoyl peroxide (BPO)-initiated
polymerization of MMA in frozen aromatic solvents [48]. The nine
lines, designated as the 5 + 4 lines were assigned to the propagat-
ting of the methyl protons, assuming free rotation about the C
a
–
CH bond in this type of radicals [45]. Different splittings of the
3
two b-protons were also predicted, in agreement with the
experiments.
According to Fig. 5c a rigid structure of the radical cannot ac-
count for the selective broadening of certain lines in the experi-
mental spectrum. Better agreement was obtained by considering
an internal motion that exchanged the ab1 and ab2 splittings be-
tween 7.5 and 14.7 G at a rate of ca. 30 MHz. A simulation taking
account of the chemical exchange with the program ESREXN [52]
is shown in Fig. 5b. Rate constants of the same order were obtained
from the observed EPR spectra of methacrylic acid radicals in fro-
zen solution [50]. Good agreement between experimental and sim-
ulated spectra has also been achieved by assuming a distribution of
conformations of the methylene group [45]. Irrespective of the de-
tailed mechanism of the broadening the 9 line spectrum in Fig. 5a
2 3
ing radical –CH )R. Details about the radical were obtained
C˙ (CH
from the shape of the spectrum, especially the relative intensity
of the lines which has been interpreted in terms of various factors
such as that there is more than one conformation [49], there is a
distribution of dihedral angles for a single conformation [45] or a
site exchange motion around the –CH
EPR spectra of PMMA radicals measured at different conversion
– C˙ < bond [50]. Analysis of
2
can safely be attributed to a radical of the type R-CH
COOCH , i.e. either the dimeric species predicted by the DFT calcu-
2 3
C˙ (CH
)-
3
lations given below or a polymer radical with a similar local
geometry about the radical site.
Computations
Geometry optimization
Geometry optimization calculations for minimum energy and
transition-state structures were performed using DFT theory. Cal-
culations were carried out using the hybrid meta GGA exchange–
correlation functional M05-2X with a double fraction of Hartree–
Fock exchange developed by Zhao and Truhlar [53]. According to
literature data this functional performs very well in DFT studies
Fig. 4. EPR spectra of a sample
silica gel, recorded (a) at 77 K, (b) at 123 K.
c-irradiated at 77 K containing 3% MMA adsorbed on

3
72
A. Lund, M. Danilczuk / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 98 (2012) 367–377
of the adsorption of hydrocarbons on a large 16T zeolite model
cluster [54,55] or modeling of water-zeolite interaction [56].
All geometries presented here were optimized with the
LANL2DZ [57–59] basis set for silicon atoms and the 6-31G(d)
hfsc’s are listed in the tables. The anisotropic values of the hfsc’s
obtained in the calculations are presented in the Supplemental
Information, however.
[
60] basis set for carbon, oxygen and hydrogen atoms. Effects of
hybrid functional and basis sets (including 6-31 + G(d,p) were ini-
tially tested on the AMST/SiO model. We found no significant dif-
Radical structures and hyperfine splitting constants
2
ST/SiO
2
ference in structural parameters between M05-2X/6-31G(d) and
M05-2X/6-31 + G(d,p), however. The C–H bond lengths were for
example1.08201 Å for 6-31G(d) and 1.08254 Å for 6-31 + G(d,p),
while the the O–H bonds in the silica cluster were 0.96393 Å and
Fig. 7 shows the optimized structure of the styrene (ST) radical
adsorbed on the hydroxylated surface of silica gel represented here
by the 8T molecular cluster. Selected structural parameters are also
shown in Fig. 7. The most stable conformation of the styrene rad-
0
.96073 Å, respectively.
ical adsorbed on the silica surface interacts via H
oxygen atom and via C of the aromatic ring with the H atom of a
silanol group. The calculated distance C ÆÆÆH–O– was found to be
.41 Å and the hydrogen-bond distance between H and the near-
est oxygen atom of the silica cluster was 2.83 Å. Such long dis-
tances between silica and the ST radical indicate very weak
a
with the nearest
Silica model
6
6
The adsorption site of a hydroxylated amorphous silica surface
was represented by a molecular cluster terminated with –OH sila-
nol groups. As shown in Fig. 6 the molecular cluster of SiO consists
of 8T (tetrahedral) sites. The cluster was fully optimized (without
any restrictions) with the M05-2X functional and LANL2DZ basis
set for silicon atoms and 6-31G(d) basis set for hydrogen and oxy-
gen atoms. The calculated O–H distances are 0.96 Å which is in a
very good agreement with literature data [61]. The Si–O bond
lengths are 1.64 Å and Si–O–Si angles are in the range of 143–150°.
2
a
2
2
interaction of the adsorbate molecule with the SiO surface. The
calculated adsorption energy (total energy relative to the isolated
ꢀ1
8
T cluster and the ST radical) was found to be 6.8 kcal mol
.
The interaction of the ST radical with a silanol group results in
slight O–H bond elongation to 0.97 Å compared to 0.96 Å in the iso-
lated cluster. The bond lengths of the adsorbed radical remained
unchanged compared to the isolated radical (gas phase). The radi-
cal optimized with a silica cluster is slightly twisted (1°) along the
Hyperfine splitting constants
C
a
–C
values of 1.08–1.10 Å, and C
.49 and 1.42 Å, respectively. All structural parameters of the aro-
6
bond, however. The calculated C–H bonds distances have
The proton hyperfine splitting constants (hfsc’s) were calcu-
lated for the radicals formed by the addition of a hydrogen atom
to the adsorbed vinyl monomers. The atom originates from the
a
–C and C –C bond distances are
b
a
6
1
matic ring are in a very good agreement with the geometry of ben-
zene adsorbed on various types of zeolites [67].
ꢀ
cleavage of a silanol bond. A hydrogen-loss O–Si„ radical center
is then also formed in the 8T cluster. Calculations of hfsc’s were
carried out in single-point runs with a combination of M05-2X
and B3LYP [62] functionals and with 6-31 + G(d) or EPR-III [63] ba-
sis sets. The best overall agreement between experimental and cal-
culated values was obtained for the B3LYP hybrid functional and
Calculated isotropic hyperfine splitting constants for the ST rad-
ical adsorbed on the SiO
H
2
surface are listed in Table 2. The isotropic
a
hfsc’s calculated with the M05-2X/6-31 + G(d) and B3LYP/6-
3
1 + G(d) methods are ꢀ18.40 and ꢀ18.04 G, respectively. Lower
values were obtained with M05-2X/EPR-III and B3LYP/EPR-III
EPR-III basis sets combination. This approach has been proven to
+
give reliable results for the Ag(C
H
2 4
)
2
or Na ÆÆÆCH
3
complexes stabi-
lized in SAPO-11 molecular sieve and 4A zeolite, respectively
64,65].
All calculations presented here were performed using the
[
Gaussian 09 package [66]. Since most of the experimentally ob-
served EPR spectra had an isotropic appearance, only isotropic
Fig. 6. Optimized structure of 8T cluster as a model of SiO
2
.
Fig. 7. Optimized styrene radical structure adsorbed on the 8T silica model cluster.

A. Lund, M. Danilczuk / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 98 (2012) 367–377
373
methods and are ꢀ12.50 and ꢀ15.49 G. Compared to the value
16.5 G used in the simulation of the experimental spectrum all
used methods give reasonable H splittings. However, the closest
value was obtained with the B3LYP/EPR-III method. The calculated
average (aHb1 + aHb2 + aHb3)/3 of the isotropic H hyperfine splitting
Table 3
Calculated isotropic hyperfine splitting constants (G) of the the H atom addition
radical of DPE adsorbed on the 8T silica model cluster.
ꢀ
a
M05-2X
B3LYP
b
6-31 + G(d)
EPR-III
6-31 + G(d)
EPR-III
constants varied from 16.55 G for M05-2X/EPR-III to 18.87 G for
the B3LYP/EPR-III method. Also the hyperfine splitting constants
H
Hb2
b1
28.14
4.23
29.94
1.72
29.07
2.25
30.40
2.11
of the aromatic protons of H
the method and were in the range of ꢀ3.58–(ꢀ)5.77 G. Much lower
splittings, 1.01–2.33 G were obtained for the aromatic protons H
and H , supporting the experimental assignment. The best agree-
ment with experiments was obtained with the B3LYP/EPR-III
method. The M05-2X/EPR-III method gives slightly underesti-
mated values of the calculated hyperfine splittings.
1
, H
3
and H
5
strongly depended on
H
H
H
H
H
b3
17.38
14.58
15.44
16.00
1
2
3
4
a/b
ꢀ1.73/ꢀ4.08
2.29/1.91
ꢀ2.97/ꢀ4.18
1.47/2.73
ꢀ3.14/ꢀ4.23
ꢀ2.22/ꢀ2.88
0.93/1.04
ꢀ2.11/ꢀ3.01
0.84/0.92
ꢀ2.14/ꢀ2.87
ꢀ2.88/ꢀ3.80
1.52/1.78
ꢀ3.12/ꢀ4.31
1.42/1.65
ꢀ2.96/ꢀ2.26
ꢀ2.37/ꢀ3.12
1.10/1.18
ꢀ2.67/ꢀ3.72
0.98/1.03
ꢀ2.41/ꢀ3.40
a/b
a/b
a/b
2
4
H5 a/b
ical. The calculated hyperfine splitting constants of the phenyl pro-
tons, not resolved in the experimental EPR spectrum of DPE, are
DPE/SiO
2
ꢀ
1.73–(ꢀ)4.23 G for H
1 3 5 2 4
, H , H and 0.84–2.73 G for H , H .
The optimized geometry of the DPE radical adsorbed on the SiO
surface is shown in Fig. 8. Compared to the isolated DPE radical,
only small changes in the \(C –C –C –C2a) and \(C –C –C
dihedral angles were observed. Similar to the other radicals in this
study the DPE radical interacts with the SiO surface via a C atom of
one of the aromatic rings and the H atom. The calculated distances
2
AMST/SiO and BMST/SiO
2
2
b
a
1
b
a
1
–C2b
)
Fig 9a and b illustrate the M05-2X/6-31G(d) optimized geome-
tries of the AMST and BMST adsorption complexes, respectively.
Similar to the ST radical both AMST and BMST interact with the sil-
ica surface through the C atom located in the aromatic ring and
through the hydrogen bond between one of the H
Si„ radical center. The calculated distance between C
radical and the silanol proton was found to be 2.40 Å compared to
2.41 Å calculated for the ST radical. The H-bond distance between
oxygen and H
pared to the isolated AMST radical, the –CH
sorbed on the silica surface are about 6° out of plane of the
2
b
are 2.54 and 2.56 Å respectively. The calculated adsorption energy
was found to be slightly higher compared to the smaller radicals
ꢀ1
and is 11.36 kcal mol
Calculated hyperfine splitting constants for the DPE radical are
listed in Table 3. As expected the CH group exhibits the largest
hyperfine splittings. The average isotropic constants (aHb1 + aHb2 + -
Hb3)/3 depended on the calculation method and were in the range
of 15.41–16.58 G compared to the experimental 16.0 G. All H pro-
.
b
and the ÆÆÆO–
of the AMST
1
3
a
b
of the CH
3
group („Si–OÆÆÆH–C) was 2.79 Å. Com-
groups of a radical ad-
b
3
tons also exhibit small anisotropy and this was taken into account
in the simulation of the experimental EPR spectrum of the DPE rad-
o
b a 1 2
molecule (dihedral angle (C –C –C –C ) = 6.2 ). The bond lengths
in the aromatic (phenyl) ring are in the same range as in the ST
radical.
The interaction of the AMST radical with the silica surface leads
to slight bond elongations in the 8T cluster. The calculated O–Si„
bond length is 1.66 Å and the O–H bond in the silanol group is
0
2
.97 Å. The calculated adsorption energy for the AMST/SiO system
ꢀ1
was found to be 8.7 kcal mol indicating a very weak interaction
of the AMST radical with the silica surface like for the ST radical.
Similar to the AMST/SiO
distance in the BMST/SiO system was found between the H atom
of silanol and C . The calculated distance was 2.41 Å. The hydro-
2
system the shortest radical to cluster
2
4
gen-bond distance between oxygen and Hb2 of the methylene
group („Si–OÆÆÆH–C) is shorter compared to the AMST and ST sys-
tems and was found to be 2.64 Å. As seen in Fig. 9b the BMST rad-
ical is twisted along C
a b c b
–C . The calculated dihedral angle (C –C –
C
a
1
–C ) is 14.1°. Adsorption on the silica surface does not affect
the geometry of the phenyl ring and all bond distances are in the
same range as in the ST and AMST radicals. This radical also weakly
2
interacts with the SiO surface. The calculated adsorption energy
ꢀ1
was 10.9 kcal mol
Calculated hyperfine splitting constants for both the AMST/
SiO and the BMST/SiO systems are listed in Table 4. The isotro-
pic hyperfine splitting constant of the six equivalent H protons
.
2
2
b
in the AMST radical is very close to the experimental value. The
values depend on the method, however, and vary from 15.8 G for
M05-2X/EPR-III, 16.55 G for M05-2X/6-31 + G(d), 17.49 G for
B3LYP/6-31 + G(d) and 17.95 G for B3LYP/EPR-III. The largest
hyperfine splitting constants of the phenyl protons were ob-
tained for H
1
, H
3
and H
5
and are in the range of ꢀ3.24–
and H are in the range of
.94–2.11 G. All these values are in very good agreement with
(ꢀ5.37) G, while the splittings of H
2
4
0
the experimental ones and support the interpretation of the
experimental spectrum.
Fig. 8. Optimized structure of DPE radical adsorbed on the 8T silica model cluster.

3
74
A. Lund, M. Danilczuk / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 98 (2012) 367–377
Table 4
Calculated hyperfine splitting constants (G) of of the H atom addition radicals of
AMST
a
-methyl styrene (AMST) and b-methyl styrene (BMST) adsorbed on the 8T silica model cluster.
BMST
M05-2X
-31 + G(d)
B3LYP
M05-2X
B3LYP
6
EPR-III
6-31 + G(d)
EPR-III
6-31 + G(d)
EPR-III
6-31 + G(d)
EPR-III
H
H
H
H
H
H
H
H
H
H
H
b1
b2
b3
b4
b5
b6
1
19.25
1.45
28.78
23.65
0.58
18.41
1.10
27.75
22.75
0.25
24.61
ꢀ3.27
0.94
20.33
1.70
30.27
25.02
0.77
26.88
ꢀ4.97
1.96
20.87
1.40
31.38
25.86
0.43
27.77
ꢀ4.02
1.10
H
H
H
H
H
H
H
H
H
H
H
a
ꢀ17.83
ꢀ7.99
14.74
ꢀ0.34
ꢀ0.36
ꢀ0.34
ꢀ5.32
2.14
ꢀ12.06
29.03
14.10
ꢀ0.24
ꢀ0.22
ꢀ0.28
ꢀ3.64
1.02
ꢀ17.96
31.69
15.84
ꢀ0.33
ꢀ0.35
ꢀ0.42
ꢀ5.36
2.06
ꢀ15.21
33.04
16.38
ꢀ0.28
ꢀ0.29
ꢀ0.38
ꢀ4.37
1.18
b1
b2
c
c
c
1
2
3
4
5
1
2
3
25.6
ꢀ4.94
2
2.11
3
ꢀ5.22
2.11
ꢀ4.90
ꢀ3.82
ꢀ5.37
ꢀ4.69
ꢀ5.80
ꢀ4.25
ꢀ5.96
ꢀ5.21
4
0.95
1.97
1.10
2.12
1.02
2.04
1.18
5
ꢀ3.24
ꢀ4.93
ꢀ3.97
ꢀ5.21
ꢀ3.59
ꢀ5.08
ꢀ4.13
MMA/SiO
2
Table 5
Calculated isotropic hyperfine splitting constants (G) of the of the H atom addition
radical of methyl methacrylate adsorbed on the 8T silica model cluster.
The most stable geometry of the MMA radical adsorbed on the
SiO surface is illustrated in Fig. 10. The radical is adsorbed via
interactions of the C@O group with the hydrogen atom of a silanol
group and between one of the H atoms and oxygen on the silica
surface. The calculated distances C@OÆÆÆHO–Si„ and H ÆÆÆO–Si„
2
M05–2X
–31 + G(d)
B3LYP
6
EPR-III
6–31 + G(d)
EPR-III
b
H
b1
H
b2
H
b3
H
b4
H
b5
H
b6
15.58
35.90
4.25
32.78
20.39
2.10
15.33
35.86
4.09
32.56
20.12
1.98
16.48
38.14
4.45
34.82
21.66
2.12
17.45
40.71
4.60
37.08
22.99
2.11
b
are 1.79 and 2.57 Å, respectively. The adsorption energy is slightly
higher compared to the other radicals and was found to be
ꢀ1
1
2.9 kcal mol , still indicating a weak interaction with the surface
of silica, however.
The calculated hyperfine splitting constants for the MMA radi-
cal are listed in Table 5. Because only H protons contribute to
b
the experimental EPR spectrum of MMA, only these values are
listed in the Table.
Table 6
Calculated hyperfine splitting constants (G) of the PMMA radical RC
COOCH , R = (CH COOCH
C˙
b
2 a 3
H (CH )-
C˙
3
3
)
2
a
3
.
M05-2X
-31 + G(d)
B3LYP
MMA polymerization
6
EPR-III
6-31 + G(d)
EPR-III
The free radical mechanism for the polymerization of vinyl
monomers has recently been intensively studied theoretically.
Density functional calculations of the Mayo and Flory mechanisms
for the self-initiation of styrene polymerization showed that
hydrogen atom transfer generates two radical species. This was
predicted to be a reasonable process that initiates MMA polymer-
a
a
a
a
a
a
CH3 ð1Þ
10.62
37.97
8.62
19.07
16.60
2.74
10.48
37.94
8.53
18.98
16.54
2.59
11.40
40.58
9.32
20.43
18.24
3.07
12.12
43.50
9.93
21.85
19.42
3.14
CH
3
ð2Þ
CH3 ð3Þ
CH3 ða
v
Þ
b1
b2
Fig. 9. Optimized structures of AMST (a) and BMST (b) radicals adsorbed on the 8T silica model cluster.

A. Lund, M. Danilczuk / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 98 (2012) 367–377
375
reaction paths were calculated at the B3LYP/6-31G level. They
found that the most favorable paths supported the Flory mecha-
nism [69]. Degirmenci et al. have applied various functionals such
as BMK, BB1K, MPW1B95, MPW1K, and MPWB1K in studies of
propagation kinetics [70] during free radical polymerization of a-
substituted acrylates and in the modeling of solvent effects for
the polymerization of MMA [71].
In our studies the interaction of the MMA radical with the sur-
face of SiO is rather weak and for computational reason and for
2
simplicity the polymerization reaction of MMA was therefore
investigated in the gas phase only. The transition state for the free
radical polymerization of MMA was localized by employing the lin-
ear synchronous transit followed by quadratic synchronous transit
algorithm [72,73]. Additionally, a transition structure (TS) was
checked by IRC (intrinsic reaction coordinate) calculations [74].
Single point energies were calculated at M05-2X/6-311 + G(2df,p)
level of theory. The transition state structure was characterized
by only one imaginary frequency.
The polymerization reaction is initiated by the hydrogen atom
transfer from the silanol group to the adsorbed MMA molecule.
The formed radical can then attack the double bond of the mono-
mer. The formation of the dimeric radical species depends on the
relative orientations of the monomer radical and the MMA mono-
mer [71]. Fig. 11 shows the potential energy surface and geome-
tries of MMA, the monomer radical, (CH
dimer radical, RC (CH )COOCH , computed at the M05-2X/6-
C˙
11 + G(2df,2p)//M05-2X/6-31G(d) level of theory. In the transi-
3 2 3
) (R), and the
C˙ COOCH
H
b 2
a
3
3
3
tion structure the CÆÆÆC distance between the radical and the mono-
mer unit is 2.27 Å in good agreement with the value calculated at
the MPWB1K/6-311 + G(3df,2p)//B3LYP/6-31 + G(d) level in
a
study of the polymerization of MMA in methanol [71]. The calcu-
ꢀ1
lated activation energy of 8.6 kcal mol is slightly higher, how-
ever, compared to the values obtained in the presence of solvent
molecules. The activation energy was unfortunately not deter-
mined experimentally.
Fig. 10. Optimized structure of the MMA radical adsorbed on the 8T silica model
cluster.
The calculated hyperfine splitting constants for the PMMA
radical are listed in Table 6. Because of the small anisotropy
only isotropic splittings are presented. The average value
ization [68]. Zhang et al. have investigated the mechanism of the
spontaneous initiation of the MMA polymerization. Six possible
2 3 3 b 2 a 3 3 3 2 3
Fig. 11. Molecular structures and energy profile for the reaction R + CH @C(CH )COOCH ? RC H (CH )COOCH , R = (CH ) , at the M05-2X/6-311 + G(2df,2p)//
C˙ C˙ COOCH
M05-2X/6-31G(d) level. (ZPE corrected energies in kcal/mol).

3
76
A. Lund, M. Danilczuk / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 98 (2012) 367–377
[13] J.P. Quaegebeur, T. Seguchi, H.L. Bail, C. Chachaty, ESR and NMR study of the c-
a
CH3ða
v
Þ
¼ 21:85 G, calculated by the B3LYP/EPR-III method agrees
ray-induced postpolymerization of vinyl monomers adsorbed on zeolite, J.
Polym. Sci. Polym. Chem. Ed. 14 (1976) 2703–2724.
well with the experimental splitting, supporting the assignment,
while the other methods predict slightly too low splitting. The val-
ues for ab1 and ab2 deviate somewhat from the splittings 14.7 G and
[
[
14] S.D. Cox, G.D. Stucky, Polymerization of methylacetylene in hydrogen zeolites,
J. Phys. Chem. 95 (1991) 710–720.
15] J.S. Beck, J.C. Vartuli, W.J. Roth, M.E. Leonowicz, C.T. Kresge, K.D. Schmitt,
C.T.W. Chu, D.H. Olson, E.W. Sheppard, A new family of mesoporous molecular
sieves prepared with liquid crystal templates, J. Am. Chem. Soc. 114 (1992)
10834–10843.
7
.5 G obtained experimentally. This may be due to difficulties in
estimating those values in the presence of chemical exchange
Fig. 5b), inaccuracies in the theory or a combination of both.
(
[
[
16] C.T. Kresge, J.C. Vartuli, W.J. Roth, M.E. Leonowicz, The discovery of
ExxonMobil’s M41S family of mesoporous molecular sieves, Studies Surf. Sci.
Catal. 148 (2004) 53–72.
17] C.J. Vartuli, S.S. Shih, C.T. Kresge, J.S. Beck, Potential applications for M41S type
mesoporous molecular sieves, Stud. Surf. Sci. Catal. 117 (1998) 13–21.
Conclusions
Radicals formed by hydrogen atom addition, preferably to the
carbon atom with highest free valence index, were the main spe-
[18] Y.S. Ko, T.K. Han, J.W. Park, S.I. Woo, Propene polymerization catalyzed over
MCM-41 and VPI-5-supported Et(ind)
Commun. 17 (1996) 749–758.
2 2
ZrCl catalysts, Macromol. Rapid
cies identified by EPR after
c-irradiation at 77 K of several vinyl
[
19] Y. Oumi, A. Hanai, T. Miyazaki, H. Nakajima, S. Hosoda, T. Teranishi, T. Sano,
Studies in surface science and catalysis nanotechnology in mesostructured
materials, in: Y. Oumi (Ed.), Proceedings of the 3rd International Materials
Symposium, Elsevier, 2003, pp. 753–756.
monomers adsorbed on silica gel. Experiments with deuterated sil-
ica showed that the hydrogen atoms originated from the surface
silanol groups of the silica gel.
The structures of the adsorption complexes were modeled and
the adsorption energies were calculated applying DFT quantum
chemical methods. Calculated hyperfine splitting constants at var-
ious levels of theory verified the assignment of the experimental
EPR spectra. The best overall agreement with experimental values
was obtained with a combination of the B3LYP hybrid functional
and the EPR-III basis set.
[
[
20] S.M. Ng, S.I. Ogino, T. Aida, K.A. Koyano, T. Tatsumi, Free radical polymerization
within mesoporous zeolite channels, Macromol. Rapid Commun. 18 (1997)
991–996.
21] X.F. Ni, Z.Q. Shen, H. Yasuda, Polymerization of methyl methacrylate with
samarocene complex supported on mesoporous silica, Chin. Chem. Lett. 12
(
2001) 821–822.
[22] S. Spange, A. Graser, H. Muller, Y. Zimmermann, P. Rehak, C. Jager, H. Fuess, C.
Baehtz, Synthesis of inorganic/organic host-guest hybrid materials by cationic
vinyl polymerization within Y zeolites and MCM-41, Chem. Mater. 13 (2001)
3698–3708.
A polymerization reaction was observed in the MMA/silica gel
system. Based on the experimental EPR spectra alone, it is impos-
sible to determine the chain length of the formed PMMA radical.
Gas phase DFT calculations at the M05-2X/6-311 + G(2df,2p)//
[23] K. Ikeda, M. Kida, K. Endo, Polymerization of methyl methacrylate with radical
initiator immobilized on the inside wall of mesoporous silica, Polym. J. 41
(
2009) 672–678.
[
[
24] T. Shiga, H. Yamada, A. Lund, Conformation and mobility of radicals in
heterogeneous systems ESR spectra at 4 K of Methyl, Ethyl, n-propyl radicals
adsorbed on silica gel, Z. Naturforsch. 29A (1974) 653–659.
ꢀ1
M05-2X/6-31G(d) level gives an activation barrier of 8.6 kcal mol
25] Y. Ben Taarit, J.H. Lunsford, Chem. EPR evidence for 17O on molybdenum oxide
supported by silica-gel, Phys. Lett. 19 (1973) 348–350.
for the first step in the free radical polymerization mechanism of
MMA. The close agreement between the calculated hyperfine cou-
plings of the formed dimeric radical and the experimental ones
[26] J. Stejskal, M. Trchova, S. Fedorova, I. Sapurina, J. Zemek, Surface
polymerization of aniline on silica gel, Langmuir 19 (2003) 3013–3018.
[
27] O. Edlund, P.O. Kinell, A. Lund, A. Shimizu, Electron spin resonance spectra of
monomeric and dimeric cations of benzene, J. Chem. Phys. 46 (1967) 3679–
3680.
may indicate
conditions.
a short chain length under the experimental
[
28] O. Edlund, P.O. Kinell, A. Lund, A. Shimizu, in: R.F. Gould (Ed.), Advances in
Chemistry Series, Radiation Chemistry 22 (1968) 311–326.
Appendix A. Supplementary data
[29] T. Takada, H. Tachikawa, Hybrid DFT study of the hyperfine coupling constants
of methyl radicals in model matrix lattices, Int. J. Quantum Chem. 105 (2005)
7
9–83.
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.saa.2012.08.054.
[
[
30] H.W. Kohn, Deuterium exchange on silica gel initiated by cobalt-60 irradiation,
J. Phys. Chem. 66 (1962) 1017–1021.
31] B.L. Bales, E. Wajnberg, O.R. Nascimento, Temperature-dependent hyperfine
coupling constant of the dianion radical of Fremy’s salt, a convenient internal
thermometer for EPR spectroscopy, J. Magn. Reson. A 118 (1996) 227–233.
32] A. Lund, K.A. Thuomas, J. Maruani, Calculation of powder ESR spectra of
radicals with hyperfine and quadrupolar interactions. Application to mono-
and dichloroalkyl radicals, J. Magn. Reson. 30 (1978) 505–514.
References
[
[
[
[
[
[
[
1] P.H. Kasai, Electron spin resonance studies of vinyl, propargyl, and butatrienyl
radicals isolated in argon matrices, J. Am. Chem. Soc. 94 (1972) 5950–5956.
2] M. Danilczuk, A. Lund, Adsorption of NO in Li-exchanged zeolite LTA. A density
functional theory study, Chem. Phys. Lett. 490 (2010) 205–209.
33] A. Lund, R. Erickson, EPR and ENDOR Simulations for disordered systems: the
balance between efficiency and accuracy, Acta Chem. Scand. 52 (1998) 261–
274.
2
3] M. Danilczuk, D. Pogocki, A. Lund, Interaction of CH OH with silver cation in
34] R. Lefebvre, J. Maruani, Use of computer programs in the interpretation of
electron paramagnetic resonance spectra of dilute radicals in amorphous solid
samples. I. High-field treatment. X-band spectra of pi-electron unconjugated
hydrocarbon radicals, J. Chem. Phys. 42 (1965) 1480–1496.
35] C.A. Coulson, Valence, 2nd ed., Oxford University Press, US, 1961.
36] F.H. Burkitt, C.A. Coulson, H.C. Longuet-Higgins, Free valence in unsaturated
hydrocarbons, Trans. Faraday Soc. 47 (1951) 553–564.
Ag-A/CH OH zeolite: a DFT study, Chem. Phys. Lett. 469 (2009) 153–156.
3
4] T. Shiga, A. Lund, g-Factor and hyperfine coupling anisotropy in the electron
spin resonance spectra of methyl-, ethyl, and allyl-type radicals adsorbed on
silica gel, J. Phys. Chem. 77 (1973) 453–455.
[
[
[
5] T. Komatsu, A. Lund, P.-O. Kinell, Electron spin resonance studies on irradiated
heterogeneous system VIII Radical cation formation from toluene, J. Phys.
Chem. 76 (1972) 1721–1726.
[
37] V. Fornes, H. Garcia, S. Jovanovic, V. Marti, 1,1-Diphenylethylene adsorbed
onto acid zeolites: Nature of the blue (605-nm) species, Tetrahedron 53 (1997)
[
[
[
6] T. Archipov, S. Santra, A.B. Ene, H. Stoll, G. Rauhut, E. Roduner, Adsorption of
benzene to copper in CuHY zeolite, J. Phys. Chem. C 113 (2009) 4107–4116.
7] G. Hübner, E. Roduner, ESR observation of redox processes on benzene-loaded
NaPdY, NaPtY and NaCeY zeolites, Magn. Reson. Chem. 37 (1999) 23–26.
8] J. Fock, A. Henglein, W. Schnabel, The effects of linear energy transfer in the
radiation-induced polymerization of several vinyl compounds, J. Phys. Chem.
4715–4726.
[
[
[
38] J.J. Rooney, R.C. Pink, Formation and stability of hydrocarbon radical-ions on a
silica-alumina surface, Trans. Faraday Soc. 58 (1962) 1632–1641.
39] E. Roduner, R. Crockett, L.M. Wu, Hyperfine anisotropy and mobility of organic
radical cations in zeolites, J. Chem. Soc. Faraday Trans. 89 (1993) 2101–2105.
40] D.W. Skelly, R.G. Hayes, W.H. Hamill, Paramagnetic resonance of alkyl radicals
6
8 (1964) 310–314.
[
9] K. Fukano, E. Kageyama, Study on radiation-induced polymerization of vinyl
monomers adsorbed on inorganic substances. I. Radiation-induced
polymerization of styrene adsorbed on several inorganic substances, J.
Polym. Sci. Polym. Chem. Ed. 13 (1975) 1309–1324.
from dissociative electron attachment in
Chem. 43 (1965) 2795–2798.
c-irradiated organic glass, J. Phys.
[
[
[
41] C. Heller, H.M. McConnell, Radiation damage in organic crystals. II. Electron
spin resonance of (CO H)CH CH(CO H) in beta-succinic acid, J. Chem. Phys. 32
1960) 1535–1539.
2
2
2
[
10] K. Fukano, E. Kageyama, Radiation-induced polymerization of vinyl monomers
adsorbed on inorganic substances. VI. Temperature dependence and effects of
additives on methyl methacrylate-silica gel system, J. Polym. Sci. Polym. Chem.
Ed. 14 (1976) 1031–1041.
(
42] M. Iwasaki, Y. Sakai, ESR spectra of poly(methacrylic acid) and poly(methyl
methacrylate): reinterpretation of the 9-line spectrum, J. Polym. Sci. A Polym.
Chem. 7 (1969) 1537–1547.
43] P.-O. Kinell, T. Komatsu, A. Lund, T. Shiga, A. Shimizu, Electron spin resonance
studies on irradiated heterogeneous systems. VI. The structure and reactivity
of paramagnetic centers in silica gel, Acta Chem. Scand. 24 (1970) 3265–3275.
[
[
11] R.M. Barrer, A.T.T. Oei, Polymerization with zeolite catalysts: I. Polymerization
of n-butylvinylether by H-mordenite, J. Catal. 30 (1973) 460–466.
12] R.M. Barrer, A.T.T. Oei, Polymerization with zeolite catalysts: II. Vinyl ethers
over H-mordenite and H–Y, J. Catal. 34 (1974) 19–28.

A. Lund, M. Danilczuk / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 98 (2012) 367–377
377
[
44] O. Edlund, P-O. Kinell, A. Lund, A. Shimizu, Electron spin resonance of
irradiated heterogeneous systems. Free radical formation in vinyl monomers.
Report LF-83 (1978) Swedish Research Councils’ Laboratory, Studsvik.
45] C.H. Bamford, A.D. Jenkins, M.C.R. Symons, M.G. Townsend, Trapped radicals in
heterogeneous vinyl polymerization, J. Polym. Sci. 34 (1959) 181–198.
46] S. Zhu, Y. Tian, A.E. Hamielec, D.R. Eaton, Radical trapping and termination in
free-radical polymerization of methyl methacrylate, Macromolecules 23
[62] A.D. Becke, Density-functional thermochemistry. III. The role of exact
exchange, J. Chem. Phys. 98 (1993) 5648–5652.
[63] N. Rega, M. Cossi, V. Barone, Development and validation of reliable quantum
mechanical approaches for the study of free radicals in solution, J. Chem. Phys.
105 (1996) 11060–11067.
[64] M. Danilczuk, D. Pogocki, A. Lund, J. Michalik, EPR and DFT study on the
stabilization of radiation-generated methyl radicals in dehydrated Na–A
zeolite, J. Phys. Chem. B 110 (2006) 24492–24497.
[65] M. Danilczuk, D. Pogocki, A. Lund, J. Michalik, Interaction of silver atoms with
ethylene in Ag-SAPO-11 molecular sieve: an EPR and DFT study, Phys. Chem.
Chem. Phys. 6 (2004) 1165–1168.
[
[
(
1990) 1144–1150.
47] T. Gillbro, P.-O. Kinell, A. Lund, ESR single crystal study of radical and radical-
pair formation in some -irradiated vinyl monomers, J. Polym. Sci. A 2 (1971)
495–1507.
[
c
1
[
48] M. Kamachi, M. Kohno, D. Liaw, S. Katsuki, ESR study of the propagating radical
in the polymerization of MMA in frozen state, Polym. J. 10 (1978) 69–75.
49] M. Kamachi, Y. Kuwae, S. Nozakura, K. Hatada, H. Yuki, Conformation of the
propagating radical of triphenylmethyl methacrylate as studied by ESR, Polym.
J. 13 (1981) 919–925.
50] Y. Sakai, M. Iwasaki, Change with temperature of the ESR spectra of
methacrylic acid radicals, J. Polym. Sci. A Polym. Chem. 7 (1969) 1749–1764.
51] M.P. Tonge, R.J. Pace, R.G. Gilbert, Electron paramagnetic resonance
investigation of the nature of the propagating species in methyl
methacrylate polymerization, Macromol. Chem. Phys. 195 (1994) 3159–3172.
52] J. Heinzer, EXREXN: simulation of exchange-broadened isotropic ESR spectra,
QCPE 11 (1972) 209.
53] Y. Zhao, N.E. Schultz, D.G. Truhlar, Design of density functionals by combining
the method of constraint satisfaction with parametrization for
thermochemistry, thermochemical kinetics, and noncovalent interactions, J.
Chem. Theory Comput. 2 (2006) 364–382.
54] Y. Zhao, D.G. Truhlar, Applications and validations of the Minnesota density
functional, Chem. Phys. Lett. 502 (2011) 1–13.
[66] Gaussian 09, Revision B.01, M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E.
Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, B. Mennucci, G. A.
Petersson, H. Nakatsuji, M. Caricato, X. Li, H. P. Hratchian, A. F. Izmaylov, J.
Bloino, G. Zheng, J. L. Sonnenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J.
Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, T. Vreven, J. A.
Montgomery, Jr., J. E. Peralta, F. Ogliaro, M. Bearpark, J. J. Heyd, E. Brothers, K.
N. Kudin, V. N. Staroverov, T. Keith, R. Kobayashi, J. Normand, K. Raghavachari,
A. Rendell, J. C. Burant, S. S. Iyengar, J. Tomasi, M. Cossi, N. Rega, J. M. Millam,
M. Klene, J. E. Knox, J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R.
E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, R. L.
Martin, K. Morokuma, V. G. Zakrzewski, G. A. Voth, P. Salvador, J. J.
Dannenberg, S. Dapprich, A. D. Daniels, O. Farkas, J. B. Foresman, J. V. Ortiz, J.
Cioslowski, D. J. Fox, Gaussian, Inc., Wallingford CT, 2010.
[67] R. Rungsirisakun, B. Jansang, P. Pantu, J. Limtrakul, The adsorption of benzene
on industrially important nanostructured catalysts (H-BEA, H-ZSM-5, and H-
FAU): confinement effects, J. Mol. Structure 733 (2005) 239–246.
[68] K.S. Khuong, W.H. Jones, W.A. Pryor, K.N. Houk, The mechanism of the self-
initiated thermal polymerization of styrene. Theoretical solution of a classic
problem, J. Am. Chem. Soc. 127 (2005) 1265–1277.
[
[
[
[
[
[
[
55] Y. Zhao, D.G. Truhlar, Benchmark data for interactions in zeolite model
complexes and their use for assessment and validation of electronic structure
methods, J. Phys. Chem. C 112 (2008) 6860–6868.
56] F. Labat, A.H. Fuchs, C. Adamo, Toward an accurate modeling of the water-
zeolite interaction: calibrating the DFT approach, J. Phys. Chem. Lett. 1 (2010)
[69] C. Zhang, X. Wang, L. Liu, Y. Wang, X. Peng, Modeling the spontaneous
initiation of the polymerization of methyl methacrylate, J. Mol. Model. 14
(2008) 1053–1064.
[
[70] I. Degirmenci, V. Aviyente, V. van Speybroeck, M. Waroquier, DFT study on the
7
63–768.
propagation kinetics of free-radical polymerization of
Macromolecules 42 (2009) 3033–3041.
[71] I. Degirmenci, E. Sukru, V. Aviyente, B. De Sterck, K. Hemelsoet, V. van
Speybroeck, M. Waroquier, Modeling the solvent effect on the tacticity in the
free radical polymerization of methyl methacrylate, Macromolecules 43
(2010) 5602–5610.
[72] C. Peng, H.B. Schlegel, Combining synchronous transit and quasi-Newton
methods to find transition states, Israel J. Chem. 33 (1993) 449–454.
[73] C. Peng, P.Y. Ayala, H.B. Schlegel, M.J. Frisch, Using redundant internal
coordinates to optimize geometries and transition states, J. Comp. Chem. 17
(1996) 49–56.
a-substituted acrylates,
[
[
57] T.H. Dunning, P.J. Hay, Modern theoretical chemistry, in: H.F. Schaefer III (Ed.),
Modern Theoretical Chemistry, Plenum, New York, 1976, pp. 1–28.
58] P.J. Hay, W.R. Wadt, Ab initio effective core potentials for molecular
calculations. Potentials for main group elements Na to Bi, J. Chem. Phys. 82
(
1985) 284–298.
59] P.J. Hay, W.R. Wadt, Ab initio effective core potentials for molecular
calculations. Potentials for to Au including outermost core orbitals, J.
Chem. Phys. 82 (1985) 299–310.
[
[
[
K
60] W.J. Hehre, R. Ditchfield, J.A. Pople, Self-consistent molecular orbital methods.
xii. further extensions of gaussian-type basis sets for use in molecular orbital
studies of organic molecules, J. Chem. Phys. 56 (1972) 2257–2261.
61] T. Takada, H. Tachikawa, DFT and direct ab-initio MD study on hyperfine
coupling constants of methyl radicals adsorbed on model surface of silica gel, J.
Mol. Catal. A: Chem 311 (2009) 54–60.
[74] C. Gonzalez, H.B. Schlegel, Reaction path following in mass-weighted internal
coordinates, J. Phys. Chem. 94 (1990) 5523–5527.