Quantitative comparison of the mesitoyl vs the benzoyl fragment in photoinitiation: A question of origin

Article abstract of DOI:10.1021/ma2001977

Photolytically generated radicals (at a wavelength of 351 nm) derived from the acetophenone-type photoinitiators benzoin (2-hydroxy-1,2-diphenylethanone) and 2,4,6-trimethylbenzoin (2-hydroxy-1-mesityl-2-phenylethanone, TMB) (specifically the benzoyl and

Full text of DOI:10.1021/ma2001977

ARTICLE
pubs.acs.org/Macromolecules
Quantitative Comparison of the Mesitoyl vs the Benzoyl Fragment in
Photoinitiation: A Question of Origin
Dominik Voll,† Thomas Junkers,‡ and Christopher Barner-Kowollik*,†
†Preparative Macromolecular Chemistry, Institut f€ur Technische Chemie und Polymerchemie, Karlsruhe Institute of Technology (KIT),
Engesserstr. 18, 76128 Karlsruhe, Germany
‡Institute for Materials Research (IMO), Polymer Reaction Design Group, Universiteit Hasselt, Agoralaan, Gebouw D, BE-3590
Diepenbeek, Belgium
S Supporting Information
b
ABSTRACT:
Photolytically generated radicals (at a wavelength of 351 nm) derived from the acetophenone-type photoinitiators benzoin (2-
hydroxy-1,2-diphenylethanone) and 2,4,6-trimethylbenzoin (2-hydroxy-1-mesityl-2-phenylethanone, TMB) (specifically the
benzoyl and mesitoyl radical) are quantified in their ability to serve as initiating species in methyl methacrylate (MMA), ethyl
methacrylate (EMA), and butyl methacrylate (BMA) bulk free radical polymerizations under optimized conditions. Herein, 2,4,6-
trimethylbenzoin is employed for the first time as photoinitiator in pulsed laser polymerizations (PLP) employing a high-frequency
excimer laser, constituting a new source for mesitoyl radicals. The current work presents an improved method for quantifying radical
efficiency of photoinitiation processes using coupled online size exclusion chromatographyꢀelectrospray ionization mass
spectrometry (SEC/ESI-MS) to analyze the obtained polymers. Because of the occurrence of side reactions during the benzoin-
initiated MMA polymerization, reduced laser energies (∼0.35 mJ/pulse) as well as low polymerization temperatures (∼ꢀ5 ꢀC)
were employed, which avoids side product formation. A plot of the ratio of benzoyl to mesitoyl (derived from 2,4,6-
trimethylbenzoin) end groups vs the ratio of both initiators in the reaction mixture indicates that the benzoin-derived benzoyl
radical is 3.0 (2.6, 2.4) times more likely to initiate the polymerization process of MMA (EMA, BMA) than the TMB-derived
mesitoyl fragment. This observation is in sharp contrast to the case when mesitil is employed as a source of mesitoyl radicals (8.6
times higher likelihood of benzoyl incorporation). These results clearly support the notion that the origin of a radical species
significantly determines its propensity to be incorporated at a polymer chain’s terminus. The cause of such an origin dependence is
tentatively assigned—at least in part—to different triplet lifetimes or intersystem crossing efficiencies (ΦISC) or both of TMB and
mesitil.
’ INTRODUCTION
ends are radicals,7 cations,8 or—in some cases—anions or weak
bases.9 Photoinitiators for radical polymerizations are classified
as cleavage (type I, for example benzoin and derivatives) and
H-abstraction type (type II, for example thioxanthone and
derivatives) initiators.10ꢀ12 The majority of type I photoinitiators
Photoinitiators are used for the polymerization of functional
monomers for the generation of variable oligomers and
polymers.1 They are employed in industrial applications includ-
ing the UV-curing of coatings and inks as well as for more
specialized applications, such as dental restorative materials,2,3
biomaterials,4,5 and the fabrication of 3-dimensional objects.6
Photoinitiation is usually applied to commence a chain growth
process where both the initiating species and the growing chain
Received: January 28, 2011
Revised:
March 13, 2011
Published: March 30, 2011
r
2011 American Chemical Society
2542
dx.doi.org/10.1021/ma2001977 Macromolecules 2011, 44, 2542–2551
|

Macromolecules
ARTICLE
Scheme 1. Photolytic Decomposition Pattern of 2,4,6-Tri-
methylbenzoin (A), Benzoin (B), and Mesitil (C)a
investigation employing ESI-MS to quantitatively study photo-
initiation processes, we have mapped in great detail the reaction
products of two specific photolytically generated radical frag-
ments during the bulk free radical polymerization of methyl
methacrylate (MMA), namely the mesitoyl and the benzoyl
fragments (see Scheme 1 for the structures).31 A large disparity
in net-initiation efficiency between both fragments with respect
to their ability to start chain growth was found. The initiator
fragments in the previous study were generated from the photo-
initiators benzoin and mesitil. The observed difference in effi-
ciency can be assigned to three effects, i.e., the actual initiation
rate and/or the overall quantum yield and/or the ability of the
molecule to absorb light. All three combined result in the above-
noted net-initiation efficiency that can be compared between
different photoinitiator systems. Thus, even though the rate of
initiation of a certain radical fragment toward the same monomer
should be universal, one may anticipate an origin dependence of
the net-initiation efficiency, and it is therefore mandatory to
study different sources of such fragments with respect to their
abundance in the polymeric product. In here, we employ the
above-mentioned PLP/SEC/ESI-MS method to quantitatively
study the reaction behavior of a related couple of initiators,
bearing the same initiator fragments as in our previous investiga-
tion (i.e., mesitoyl and benzoyl),31 yet originating from different
source molecules. PLPs were performed at low laser energy on a
homologous series of alkyl methacrylates, differing in chain
lengths of the ester side group (i.e., methyl (methyl methacrylate,
MMA); ethyl (ethyl methacrylate, EMA), and butyl (butyl
methacrylate, BMA)). Methacrylates are purposefully chosen
as they predominantly terminate via disproportionation, provid-
ing certainty that the terminal group in the chain has truly
initiated the macromolecular growth process.
The reason for employing online SEC/ESI-MS for quantifying
photolytically generated radicals in their ability to serve as
initiating species in the free radical bulk polymerization of
variable methacrylates is due to the fact that ESI-MS spectra,
which are measured via direct infusion of the polymer samples
can—under certain circumstances—be less accurate (for details
see the Results and Discussion section). Compared to our
previous study,31 the laser energy as well the temperature has
been reduced to minimize side product formation during the free
radical polymerization using the PLP technique and benzoin (B)
as initiator. The below-described observation of unassigned
signals in the ESI-MS spectra of macromolecules initiated by
benzoin presented herein, may be associated with the improve-
ment in the employed ESI-MS system in the current study (for
details see the Experimental section), which is up to 10 times
more sensitive compared to the ESI-MS system previously
employed.31 For quantitative evaluation of the SEC/ESI-MS
spectra it is of primary importance to assign or—better
eliminate—all of the side products, as long as their formation
process remains unresolved. In the present study, two benzoin-
type photoinitiators, i.e., benzoin (B) and 2,4,6-trimethylbenzoin
(A), were employed to study the (net) efficiency of the benzoyl
fragment (2) to initiate the free radical bulk polymerization of
methyl methacrylate (MMA), ethyl methacrylate (EMA), and
butyl methacrylate (BMA), which is quantitatively compared to
the initiation efficiency of the mesitoyl fragment (1).32
a The photoinitiators A and B are used in the current study for
quantifying the initiation efficiency of the radical fragments (1, mesitoyl)
and (2, benzoyl). To compare the latest results with our former study
between B and C, the decomposition pattern of C is also depicted. Note
that the combination of compound C with B was the subject of a
previous study.31
are aromatic carbonyl compounds with appropriate substitution.
Upon absorption of light, benzoin and its derivatives sponta-
neously undergo R-cleavage and generate free radicals. Benzoin
and its derivatives are the most widely used photoinitiators for
free radical polymerization due to their high quantum efficiency
and reactivity of the generated radical fragments. Photoinitiation
processes via type II photoinitiators are based on the reaction of
their triplet excited states with a hydrogen donor, thereby
producing an initiating radical.12 For further details about
photoinitiated polymerizations and their opportunities, the
reader is referred to a recently published review by Yagci and
colleagues.13 Furthermore, photoinitiators are used in academia
for pulsed laser polymerization. Pulsed laser polymeriza-
tion coupled to size-exclusion chromatography (PLP-SEC),
pioneered by Olaj and co-workers in 1987,14 is the most
prominent and accurate technique for the determination of
the rate coefficient of propargation, kp, today.15ꢀ20 In a further
refinement of the technique, a recent study showed that PLP
in conjunction with online size exclusion chromatographyꢀ
electrospray ionization mass spectrometry (SEC/ESI-MS) can
successfully be applied for the determination of highly accurate
propagation rate coefficients in free radical polymerization.21
The potential of SEC/ESI-MS has been discussed in several
publications.22ꢀ24 For the PLP technique, the photoinitiators
employed must be able to initiate the polymerization of a wide
range of monomers, and the pathways of initiation are important
to know.25ꢀ27 Qualitative studies on photoinitiation processes
have been carried out in several publications using electrospray
ionization mass spectrometry (ESI-MS).28ꢀ30 While qualitative
investigations can reveal some important trends, it is of great
importance to quantify the initiation ability of the different
initiator fragments toward variable classes of monomers.
The initiator fragments are located at the R-end groups (and
if combination modes prevail also the ω-end groups) of
the polymeric material and should therefore be predestinated
for (quantitative) mass spectrometric analysis. In a previous
Scheme 1 depicts the chemical structure as well as the UV-
light-induced radical decomposition products of all the (in our
quantitative studies) applied photoinitiators, which have been
used to initiate macromolecular growth via PLP. Note that the
2543
dx.doi.org/10.1021/ma2001977 |Macromolecules 2011, 44, 2542–2551

Macromolecules
ARTICLE
disproportionation product having radical fragment 3 as an end
group is not the subject of the present study (for details see
below). The choice for studying the comparative efficiency of
benzoyl vs mesitoyl is driven by previous quantitative findings
that the mesitoyl fragment—when derived from mesitil—has an
8.6 times reduced propensity to be found as a polymer end group
in poly(methyl methacrylate) (pMMA) than that of the benzoyl
fragment (derived from benzoin).31 In the current study 2,4,6-
trimethylbenzoin was used for the first time in photoinitiation
processes of free radical bulk polymerizations of methacrylates.
The reason for employing 2,4,6-trimethylbenzoin instead of
mesitil is twofold: First—and most importantly—this initiator
has the same general structure as benzoin (benzoin-type photo-
initiator), thus potentially featuring a similar triplet lifetime and
intersystem crossing (ISC) behavior, therefore allowing a differ-
entiation of the origin of the sluggish incorporation of mesitoyl
radicals when they are derived from mesitil.27 We note that the
hypothesis of similar triplet lifetimes and ISC behavior for the
two initiators may not be tenable under all circumstances, as the
additional methyl groups on the phenyl ring in TMB may alter
the geometry of the molecule (out-of-plane twisted configura-
tions may be preferred), leading to an altered behavior (see ref 33
for some computational evidence for such an effect). Second, the
initiator comprises only one mesitoyl fragment. Therefore, the
ratio of both radical fragments (coming from TMB and benzoin,
respectively) as they are generated is 1:1. The primary aim of the
current study is thus to quantify the net-efficiency difference (i.e.,
their respective propensity to be found as the initiating group at
the chain end) between the benzoyl and the mesitoyl fragments,
derived from benzoin and 2,4,6-trimethylbenzoin, to establish
whether the propensity of the mesitoyl fragment to be found as
an initiating species is a function of its origin. The following
experimental strategy is employed: Mixtures of benzoin and
2,4,6-trimethylbenzoin (ranging from equimolar composition to
a large excess of A) are prepared and subjected to pulsed laser
polymerization as described in the Experimental section. The
generated polymer is isolated and subjected to online SEC/ESI-
MS. From the mass spectral evaluation procedure, the ratio of
benzoyl and mesitoyl radicals that have initiated macromolecular
growth is plotted as a function of the ratio of both initiators in the
reaction mixture. The evaluation procedure for such a plot, as
well as the mass spectral evaluation, is described elsewhere.34,35
Scheme 2 depicts the expected disproportionation and combina-
tion products, which are formed in the presence of MMA as
monomer (for schemes of EMA and BMA as monomer see
Scheme S1 and Scheme S2). Table 1 contains the corresponding
experimentally observed as well as the theoretically expected
masses for the identified disproportionation and combination
products in one repeat unit.
(THF, Scharlau, multisolvent GPC grade, 250 ppm BHT) and methanol
(VWR, chromanorm) were also employed as received. For UV/Vis
measurements methanol (VWR, chromanorm) was used.
Synthesis of Mesitylglyoxal. The following preparation proce-
dure of mesitylglyoxal is adapted from the literature.36ꢀ38 A solution of
6.48 g of selenium dioxide (58.4 mmol, 1.00 eq), 5.00 mL of water, and
9.56
g of 2,4,6-trimethylacetophenone (9.80 mL, 58.9 mmol,
1.00 eq) in 65.0 mL of dioxane was refluxed for 5 h at 105 ꢀC under
vigorous stirring. During the heating of the solution a gradual precipita-
tion of selenium takes place, and the color changes from colorless to
yellow. The solution was decanted from the free selenium, and the
dioxane was evaporated. The crude yellow product was purified via
vacuum distillation. The product was collected at T = 62 ꢀC and
p = 4.1 ꢁ 10ꢀ1 bar as a clear yellow oil which crystallized upon cooling
(yield: 2.95 g, 16.7 mmol, 28%). ESI-MS: [M þ Na]exp = 199.1 Da and
[M þ Na]calc = 199.07 Da. 1H NMR (400 MHz, CDCl3) [δ, ppm]: 9.53
(s, 1H, ꢀCHO), 6.89 (s, 2H, ꢀHarom), 2.31 (s, 3H, ꢀCH3), 2.18 (s, 6H,
2 ꢁ ꢀCH3). 13C NMR (100 MHz, CDCl3) [δ, ppm]: 196.79 (C1),
188.30 (C2), 141.37 (C3), 136.47 (C4), 131.18 (C5), 129.09 (C6), 21.28
(C7), 20.16 (C8). (For more details, see Figures S1 and S2 in the
Supporting Information.)
Synthesis of 2-Hydroxy-1-mesityl-2-phenylethanone
(2,4,6-Trimethylbenzoin, A). To a suspension of 1.53 g of anhy-
drous aluminum chloride (11.5 mmol, 2.01 eq) in 5.80 mL of benzene
at 10 ꢀC, 1.01 g of mesitylglyoxal (5.74 mmol, 1.00 eq) dissolved in
5.80 mL of benzene was added. The reaction was then vigorously stirred
over a period of 2 h at 10 ꢀC (cooled with ice). Stirring was continued
overnight at ambient temperature. The reaction mixture was subse-
quently poured onto ice and concentrated hydrochloric acid. The
organic layer was removed, and the aqueous layer was extracted twice
with a minimal amount of benzene. The combined organic layers were
dried over MgSO4, and benzene was evaporated under reduced pressure.
The crude, pale, yellowish solid was recrystallized from ethanol to give
2,4,6-trimethylbenzoin (A) as a white powder (yield: 0.58 g, 2.28 mmol,
40%). ESI-MS: [M þ Na]exp = 277.2 Da and [M þ Na]calc = 277.12 Da.
1H NMR (400 MHz, CDCl3) [δ, ppm]: 7.25ꢀ7.19 (m, 3H,
ꢀHph), 7.15ꢀ7.08 (m, 2H, ꢀHph), 6.74 (s, 2H, ꢀHarom), 5.59 (d,
3J = 5.2 Hz, 1H, ꢀHtert), 4.50 (d, 3J = 5.5 Hz, 1H, ꢀOH), 2.26 (s, 3H,
ꢀCH3), 1.82 (s, 6H, 2 ꢁ ꢀCH3). 13C NMR (100 MHz, CDCl3) [δ,
ppm]: 208.21 (C1), 139.55 (C2), 136.58 (C3), 134.59 (C4), 134.12
(C5), 128.37 (C6), 128.31 (C7), 126.77 (C8), 80.54 (C9), 21.11 (C10),
18.90 (C11). (For more details see Figures S1 and S2 in the Supporting
Information.)
Polymerizations. All MMA, EMA, and BMA samples consisted of
monomer (sample volume ∼0.5 mL) with a mixture of both photo-
initiators with an overall concentration of cPI = 5 ꢁ 10ꢀ3 mol Lꢀ1. The
samples were carefully freed from oxygen prior to laser irradiation by
purging the reaction mixture with high-purity nitrogen for 2 min. The
sample vial was subsequently placed into a sample holder, which was
held at a constant temperature of ꢀ5 ꢀC by a thermostat (model:
1196D, VWR, Darmstadt, Germany). The temperature was indepen-
dently measured at the sample holder. The sample temperature was
allowed to equilibrate for 5 min before starting the polymerization.
Photoinitiation was achieved by an excimer laser system (Coherent
Xantos XS-500, XeF, frequency variable from 1 to 500 Hz, operating at a
wavelength of 351 nm (pulse width ∼20 ns) and an laser energy of
2ꢀ6 mJ) at 100 Hz for an overall polymerization time of 90 000 laser
pulses (∼15 min). A self-made metal filter (fine-mesh metal grid) was
implemented next to the radiation exit window to obtain a reduced laser
energy of ∼0.35 mJ/pulse. The laser beam, which was adjusted to an
energy of close to 0.35 mJ/pulse hitting the sample, was redirected to
illuminate the vial from the bottom.
’ EXPERIMENTAL SECTION
Materials. Methyl methacrylate (MMA, Sigma-Aldrich, 99%, stabi-
lized), ethyl methacrylate (EMA, Acros, 99%, stabilized), and butyl
methacrylate (BMA, Acros, 99%, stabilized) were freed from inhibitor by
passing through a column of activated basic alumina (Merck). Benzoin
(Sigma-Aldrich, 98%) was recrystallized twice in ethanol prior to use.
Benzene (Sigma-Aldrich, purris. p.a.) was dried over molecular sieves
(4 Å) for 4 h. Dioxane (Acros), selenium dioxide (Acros, 99.8%),
aluminum chloride (ABCR, anhydrous), and 2,4,6-trimethylacetophe-
none (Sigma-Aldrich, purum) were used as received. For SEC/ESI-MS
measurements sodium iodide (Fluka, purris. p.a.), tetrahydrofuran
UV/Vis Spectra. UV/Vis spectra were recorded at ambient tem-
peratures using a Varian Cary 300 Bio photospectrometer.
2544
dx.doi.org/10.1021/ma2001977 |Macromolecules 2011, 44, 2542–2551

Macromolecules
ARTICLE
Scheme 2. Expected Polymeric Disproportionation and Combination Peaks in the Photochemically Initiated Bulk Free Radical
Polymerization of MMA in the Presence of a Cocktail of the Benzoin and 2,4,6-Trimethylbenzoin Photoinitiators Depicted in
Scheme 1a
a
Note that the position of the double bond in the unsaturated disproportionation product may also be at the site of the former R-methyl group. See
Table 1 for a collation of the masses of the individual radical fragments. Furthermore, the combination products are not of interest but are mentioned
because of completeness of assignment (C1,2,MMA, C2,1,MMA; C1,3,MMA, C3,1,MMA and C2,3,MMA, C3,2,MMA cannot be distinguished in the mass spectra). A
detailed description of the nomenclature is given in the Results and Discussion section.
NMR. The structure of the synthesized photoinitiator was confirmed
by 1H-NMR spectroscopy on a Bruker AM 400 spectrometer at
400 MHz for hydrogen nuclei and 100 MHz for carbon nuclei. All
samples were dissolved in CDCl3.
SEC/ESI-MS. Spectra were recorded on an LXQ mass spectrometer
(ThermoFisher Scientific, San Jose, CA) equipped with an atmospheric
pressure ionization source operating in the nebulizer-assisted electro-
spray mode. The instrument was calibrated in the m/z range
195ꢀ1822 amu using a standard containing caffeine, Met-Arg-Phe-Ala
acetate (MRFA), and a mixture of fluorinated phosphazenes (Ultramark
1621) (all from Aldrich). A constant spray voltage of 4.5 kV was used,
3 μm) with precolumn (Mesopore 50 ꢁ 4.6 mm) operating at 30 ꢀC.
THF at a flow rate of 0.30 mL minꢀ1 was used as eluent. The mass
spectrometer was coupled to the column in parallel to an RI-detector
(G1362A with SS420x A/D) in a setup described earlier.23 An aliquot
of 0.27 mL minꢀ1 of the eluent was directed through the RI detector
and 30 μL minꢀ1 infused into the electrospray source after post-
column addition of a 100 μM solution of sodium iodide in methanol
at 20 μL minꢀ1 by a microflow HPLC syringe pump (Teledyne
ISCO, model 100DM). Flow rates, instrument settings, and salt
concentrations were optimized to yield maximum ionization effi-
ciency while keeping salt cluster formation to a minimum.39 50 μL of
a polymer solution with a concentration of 5.5 mg mLꢀ1 was injected
onto the HPLC system.
and nitrogen at a dimensionless sweep gas flow rate of 2 (∼3 L minꢀ1
)
and a dimensionless sheath gas flow rate of 12 (∼1 L minꢀ1) were
applied. The capillary voltage, the tube lens offset voltage, and the
capillary temperature were set to 60 V, 110 V, and 275 ꢀC, respectively.
The LXQ was coupled to a Series 1200 HPLC-system (Agilent, Santa
Clara, CA) consisting of a solvent degasser (G1322A), a binary pump
(G1312A), and a high-performance autosampler (G1367B), followed by
a thermostated column compartment (G1316A). Separation was per-
formed on two mixed bed size exclusion chromatography columns
(Polymer Laboratories, Mesopore 250 ꢁ 4.6 mm, particle diameter
’ RESULTS AND DISCUSSION
In the current work we wish to establish whether the origin of a
specific photolytically generated radical (i.e., mesitoyl derived
from mesitil vs mesitoyl derived from 2,4,6-trimethylbenzoin)
has an influence on its ability to serve as an initiating species in
free radical polymerizations. In our previous study,31 the
2545
dx.doi.org/10.1021/ma2001977 |Macromolecules 2011, 44, 2542–2551

Macromolecules
ARTICLE
Table 1. Collation of the Polymeric Product Signals Ob-
served during SEC/ESI-MS of Poly(methyl methacrylate)
Samples Generated during the Pulsed Laser-Initiated Bulk
Free Radical Polymerization of MMA at 100 Hz and ꢀ5 ꢀCa
species
ionization
(m/z)theo/Da
(m/z)exp/Da
Δ(m/z)
D1,H,MMA
Dl,=,MMA
D2,H,MMA
D2,=,MMA
D3,=,MMA
D3,H,MMA
C1,1,MMA
C2,2,MMA
C3,3,MMA
Cl,2,MMA
C1,3,MMA
C2,3,MMA
Naþ
Naþ
Naþ
Naþ
Naþ
Naþ
Naþ
Naþ
Naþ
Naþ
Naþ
Naþ
1171.6 (n = 10)
1169.6 (n = 10)
1129.6 (n = 10)
1127.5 (n = 10)
1131.6 (n = 10)
1129.6 (n = 10)
1117.6 (n = 8)
1133.5 (n = 9)
1137.6 (n = 9)
1175.6 (n = 9)
1177.6 (n = 9)
1135.5 (n = 9)
1171.7
1169.7
1129.7
1127.6
1131.7
1129.7
1117.4
1133.7
1137.5
1175.6
1177.5
1135.4
0.1
0.1
0.1
0.1
0.1
0.1
0.2
0.2
0.1
0
0.1
0.1
a The table provides experimentally observed as well as theoretically
expected masses for the found disproportionation (D) and combination
(C) products (consisting 8 monomer units, n is the number of repeat
units). The structures corresponding to the individual peaks are depicted
in Scheme 2. The tabulated values correspond to the peaks displayed in
Figure 2.
Scheme 3. Synthesis of 2,4,6-Trimethylbenzoin (A); 2,4,6-
Trimethylacetophenone, Mesitylglyoxal
Figure 1. (a) ESI-MS spectra (one repeat unit) of benzoin-initiated
pMMA obtained at variable laser pulse energies (∼0.35 to 6 mJ/pulse;
measured in DCM/MeOH (3:1) via direct infusion) synthesized via
pulsed laser polymerization (PLP) at 100 Hz, ꢀ5 ꢀC, and cPI = 5 ꢁ 10ꢀ3
mol Lꢀ1. (b) ESI-MS spectra (one repeat unit) of 2,4,6-trimethylben-
zoin-initiated pMMA obtained at variable laser pulse energies (∼0.35 to
6 mJ/pulse; measured in DCM/MeOH (3:1) via direct infusion)
synthesized via pulsed laser polymerization (PLP) at 100 Hz, ꢀ5 ꢀC,
and cPI = 5 ꢁ 10ꢀ3 mol Lꢀ1. The nomenclature employed to identify the
individual disproportionation and combination products is provided in
Scheme 2.
effectiveness of mesitoyl (derived from mesitil) vs benzoyl
(derived from benzoin) radicals for initiating MMA polymeriza-
tion was established. The study concluded that it is unlikely that a
potential (relative) net-efficiency difference of up to a factor of
8.6 (or potentially up to 86 when additionally considering the
very different (factor of ∼10) UV absorptivities of mesitil and
benzoin at 351 nm, see also below) is solely caused by a
difference in the intersystem crossing ability of mesitil or its
triplet lifetime (or a combination of both).31 Employing an
alternative photoinitiator to mesitil, which also bears the mesitoyl
fragment, should aid in solving the question of origin depen-
dence. Such an initiator is 2,4,6-trimethylbenzoin (A, TMB),
bearing a mesitoyl fragment (1) and a hydroxyl fragment (3) (see
Scheme 1). TMB is a completely different type of initiator
compared to mesitil (i.e., TMB being of the benzoin type) and
should thus be ideally suited for a comparison with mesitil.
Concomitantly, we are interested in quantifying the net-initiation
efficiency of the benzoyl fragment (derived from benzoin) and
the mesitoyl fragment (derived from 2,4,6-trimethylbenzoin)
in photochemically induced PLP experiments with MMA, EMA,
and BMA as monomer to assess a potential change in relative
reactivity with a change in monomer. 2,4,6-Trimethylbenzoin
was synthesized according to a literature procedure (see Experi-
mental section) and is used for the first time as a photoinitiator in
PLP experiments with the above-mentioned monomers. To
ensure that TMB absorbs UV light at 351 nm, a UV/Vis
spectrum was recorded (see Figure 5). Within the current study
all assigned signals obtained in ESI mass spectra were labeled
according to the following nomenclature: Disproportionation
peaks occur in pairs removed by 2 amu from each other. Thus,
each peak is labeled with Dv,=,x or v,H,x, where v denotes the radical
fragment that has initiated the polymerization ((1), (2) or (3),
2546
dx.doi.org/10.1021/ma2001977 |Macromolecules 2011, 44, 2542–2551

Macromolecules
ARTICLE
see Scheme 1) and x denotes the monomer (MMA, EMA, or
BMA). Combination products are mentioned for the sake of
completeness of the assignments, yet they are of limited
use in the present study as it is not possible to differentiate
whether they have been generated by initiation or termination of
macromolecular growth. All combination peaks are labeled with
C
v,w,x where v and w denotes the radical fragment as end group
(see Scheme 1) and x denotes the monomer (MMA, EMA, or
BMA).
Reducing the Laser Energy. During the work with freshly
recrystallized benzoin as photoinitiator in photochemically in-
itiated free radical bulk polymerizations of MMA, EMA, and
BMA using the PLP method a surprising and unexpected
observation was made. The polymeric material, synthesized via
the PLP method up to 3% conversion, was analyzed with high
sensitivity ESI-MS where each repeat unit of the polymer should
only show four disproportionation products (in case of MMA as
monomer: D2,H,MMA, D2,=,MMA, D3,H,MMA, D3,=,MMA; see
Scheme 2) as well as three combination products (in case of
MMA as monomer: C2,2,MMA, C3,3,MMA, C2,3,MMA; see
Scheme 2). Optimization of the reaction temperature (ꢀ5 ꢀC)
leads to a minimization of the combination products; thus,
mainly disproportionation products were obtained in the ESI-
MS spectra. However, the ESI-MS spectra in the present work
show a new set of signals, which cannot be assigned to any known
disproportionation or combination product (see Figure 1a, peak
series denoted with a question mark). Neither can new radical
fragments derived from benzoin nor salt adducts be assigned to
the unexpected signals (denoted with a question mark in
Figure 1a). The only successful option to eliminate the side
products from the polymer was to minimize the laser energy. The
employed laser system provides stable laser energies down to
2 mJ/pulse. Figure 1a depicts the ESI-MS spectra of benzoin-
initiated pMMA synthesized at different laser energies via the
PLP method. The energy series indicates that the side products
are energy dependent (less laser energy leads to less side
products); however, at the lowest stable laser energy of 2 mJ
the ESI-MS spectra are not yet free of side products as evident
from Figure 1a (see peak ensemble at 1307.6 m/z). Only through
the subsequent installation of a fine-mesh metal grid as beam
attenuator next to the beam output window was it possible to
reduce the laser energy to ∼0.35 mJ/pulse. With a significant
reduction in laser energy the ideally expected scenario was
achieved, i.e., the disappearance of the unassigned signals.
Clearly, a large energy input into benzoin opens—primary or
secondary—reaction channels that lead to unexpected products.
This alone is a significant observation since such fragments can
have disturbing effects on PLP or any other technique relying on
photoinitiation and 2 mJ/pulse is usually considered a moderate
(or even low) energy input for a PLP experiment. Regardless, it
has thus become feasible to obtain side product free benzoin-
initiated pMMA, pEMA, and pBMA samples via a reduction of
the laser energy. A photoinitiator concentration or temperature
dependence of the side product formation was not detected.
To establish the behavior of the new initiator during PLP, a
second set of polymerizations under variation of the incident
energy was carried out for 2,4,6-trimethylbenzoin-initiated MMA
polymerization. The corresponding mass spectra are depicted in
Figure 1b. Pleasingly, no effect of the laser energy on the product
distribution can be observed; only the four expected dispropor-
Figure 2. (a) SEC/ESI-MS overview spectra of polymeric material
obtained from five different ratios of benzoin/2,4,6-trimethylbenzoin-
initiated (cPI,0 = 5 ꢁ 10ꢀ3 mol Lꢀ1, 1:1, 1:3, 1:6, 1:9, 1:18) PLP of
MMAbulk at ∼0.35 mJ laser energy, 100 Hz, ꢀ5 ꢀC. The figure depicts
the singly charged products ionized with sodium iodide at a retention
time between 16.86 and 19.93 min. (b) Zoom into one repeat unit of the
SEC/ESI-MS overview spectra of polymeric material obtained from five
different ratios of benzoin/2,4,6-trimethylbenzoin-initiated (cPI,0 = 5 ꢁ
10ꢀ3 mol Lꢀ1, 1:1, 1:3, 1:6, 1:9, 1:18) PLP of MMAbulk at ∼0.35 mJ laser
energy, 100 Hz, ꢀ5 ꢀC. For the masses of the assigned products see
Table 1. The nomenclature employed to identify the individual dis-
proportionation and combination products is provided in Scheme 2.
tion products (in case of MMA as monomer: C1,1,MMA, C3,3,MMA
C
,
1,3,MMA; see Scheme 2) are identified in the ESI-MS spectra.
A collation of the polymeric product signals observed during
ESI-MS via direct infusion of benzoin as well as 2,4,6-trimethyl-
benzoin initiated pMMA samples is shown in Table 1. For further
quantification studies of these two photoinitiators the following
pulsed laser polymerization conditions were used: f = 100 Hz,
T = ꢀ5 ꢀC, E ∼ 0.35 mJ/pulse and an overall concentration of
the photoinitiators of cPI = 5 ꢁ 10ꢀ3 mol Lꢀ1
.
Initiator Cocktails of Benzoin and 2,4,6-Trimethylben-
zoin. Figure 2a provides a depiction of all SEC/ESI mass spectra
recorded of pMMA, generated via the pulsed laser polymeriza-
tion of MMA in the presence of different mixtures of benzoin and
2,4,6-trimethylbenzoin (1:1, 1:3, 1:6, 1:9, 1:18). Furthermore,
tionation products (in case of MMA as monomer: D1,H,MMA
1,=,MMA, D3,H,MMA, D3,=,MMA; see Scheme 2) and three combina-
,
D
2547
dx.doi.org/10.1021/ma2001977 |Macromolecules 2011, 44, 2542–2551

Macromolecules
ARTICLE
any other peaks. The disproportionation peak ensembles of all
occurring radicals (1, 2, and 3) is depicted in Figure 3.
The enlarged disproportionation peak ensemble shown in
Figure 3 demonstrates that only the peaks labeled with D2,=,MMA
and D1,=,MMA are free of any isobaric interference. The isotopic
pattern of the disproportionation products associated with D1,=,MMA
and D1,H,MMA overlap with each other; for this reason, it is only
possible to use D1,=,MMA for quantitative evaluations. In the case of
the isotopic pattern associated with D2,=,MMA, D2,H,MMA, D3,=,MMA
,
and D3,H,MMA the situation is more complex. Radical fragments 2
and 3 differ in molecular weight by 2 Da, leading to an unavoid-
able overlap of both disproportionation signals. The signals for
D2,H,MMA and D3,=,MMA appear at m/z = 1129.6, making their
quantitative evaluation nonfeasible. In addition, the isobaric
overlap does not allow obtaining quantitative information about
the initiation ability of 3. It is also important to compare
disproportionation products that carry an identical end group
(unsaturated vs unsaturated or saturated vs saturated). Thus, the
only choice for quantitatively evaluating the SEC/ESI mass
spectra is to compare D2,=,MMA with D1,=,MMA. For the other
investigated polymers, (pEMA and pBMA) the same considera-
tions apply (D2,=,EMA vs D1,=,MMA and D2,=,BMA vs D1,=,BMA (for
further information see Figure S5 and Scheme S1 (for pEMA);
Figure S8 and Scheme S2 (for pBMA) in the Supporting
Information). Figure 3 additionally indicates how the heights
ΔhD2,=,MMA and ΔhD1,=,MMA are defined, which are employed for
the mass spectrometric evaluation procedure.
Data Analysis and Evaluation of the SEC/ESI-MS Spectra.
First of all, it should be mentioned why online SEC/ESI-MS is
used instead of direct infusion ESI-MS for the determination of
the disproportionation signals. Direct infusion ESI mass spectra
in the range of 200ꢀ2000 m/z show a slightly rising baseline,
caused by the presence of multiply charged ions from higher mass
ranges. Such multiply charged ions appear in the lower m/z range
(up to 2000 m/z) and lead to a mild, yet constant, baseline shift.
For our evaluation method of the mass spectra an accurate
measure of the height of the disproportionation products (as
mentioned above) is required. Therefore, it is of importance to
have a constant baseline as otherwise the derived initiation
efficiency of the different radicals is beset with an error. The
advantage of online SEC/ESI-MS is the possibility of choosing
only the singly charged ion region (defined range in the RI trace)
of an entire mass spectrum. By employing SEC/ESI-MS for
determining D1,=,MMA and D2,=,MMA, it is, however, only possible
to evaluate 8 or 9 repeat units within one mass spectrum (with
ESI-MS direct infusion it is possible to evaluate in the range of up
to 15 repeat units). Yet, it makes practically no difference for the
final initiation ratios when 7 repeat units less are used for the
evaluation of the mass spectra. In the most simple and straight-
forward approach for evaluating the mass data in the current
work, the height of the only two peaks that do not show isobaric
overlap (D1,=,MMA and D2,=,MMA), ΔhDv,=,x is evaluated in each
repeat unit. Figure 3 depicts a graphical representation on how
ΔhDv,=,x is determined. Furthermore, the height of the individual
peaks can be employed to arrive at the mole fraction of the
disproportionation product D1,=,MMA, F(i) (see eq 1). Never-
theless, it is important to establish if there exists any potential
chain length dependent mass bias. For this purpose we have
taken a modified method from G€unzler et al. established in
previous quantitative mass spectrometric evaluations.34,35 The
modified approach we have taken in here is equivalent to that
employed in our previous study on the mesitil/benzoin pairing.31
Figure 3. Enlarged section of the two disproportionation peak ensem-
bles associated with polymer chains (pMMA) initiated by radical
fragments 1, 2, and 3 (initiator ratio B:A = 1:3). The first signal
employed for evaluation purposes is D2,=,MMA as it is the only iso-
topically nonoverlapping signal derived from benzoin. The second signal
employed for quantitative evaluation purposes is D1,=,MMA derived from
2,4,6-trimethylbenzoin. The height of both signals in every repeat unit is
employed to calculate the quantitative ratio between these two initiator
fragments at the polymer chain terminus.
Figure 2b shows a representative repeat unit of the above mass
spectra, with every peak labeled according to the above-noted
nomenclature. The expected and found masses for the depicted
representative unit are collated in Table 1. The additionally—to a
much less part—occurring combination products are of no
interest for deducing initiation efficiencies as explained above.
Inspection of Table 1 indicates that the theoretically and
experimentally expected masses agree well within the accuracy
of the mass spectrometric analysis ((0.2 Da). Such an agreement
is found in every repeat unit and is representative for any initiator
mixture composition employed within the current study (data for
pEMA, pBMA and tables for the expected masses can be found in
Figures S4 and S5 as well as Table S1 (for pEMA); Figures S7 and
S8 as well as Table S2 (for pBMA) in the Supporting In-
formation). Before proceeding to a quantitative analysis of the
individual disproportionation peaks, it is worthwhile to briefly
provide a qualitative description of how the SEC/ESI mass
spectra changes with increasing amounts of 2,4,6-trimethylben-
zoin in the initiation cocktail. For this purpose a reconsideration
of Figure 2b is necessary. Inspection of this figure demonstrates
that with increasing amounts of 2,4,6-trimethylbenzoin in the
reaction mixture the number of chains carrying a mesitoyl end
group increases (D1,=,MMA) and the number of chains carrying a
benzoyl group decreases (D2,=,MMA). Qualitatively, one can also
note that a relatively low amount of 2,4,6-trimethylbenzoin in the
initiation cocktail is needed to achieve a significant signal of D1,=,
MMA. Compared to our previous study employing benzoin/
mesitil mixtures,31 a smaller amount of the mesitoyl bearing
photoinitiator is necessary to achieve a significant signal. To place
this qualitative observation onto a quantitative basis, the dis-
proportionation peaks corresponding to the chains initiated with
mesitoyl radicals (1) depicted in Figures 2b and 3 are placed into
relation to the peaks associated with chains initiated by benzoyl
radicals (2). The signals which are directly and quantitatively
compared to each other need to be free of isobaric overlap with
2548
dx.doi.org/10.1021/ma2001977 |Macromolecules 2011, 44, 2542–2551

Macromolecules
ARTICLE
Figure 5. UV spectra of the employed photoinitiators, dissolved in
MeOH at a concentration of cPI = 1 ꢁ 10ꢀ3 mol Lꢀ1 to compare the
absorbance at the laser wavelength of 351 nm with mesitil (C) (set to
100%).
Figure 4. Plotted is the ratio of the intensities of the disproportionation
products corresponding to polymer chains (pMMA, pEMA, and pBMA)
initiated with a benzoyl radical (derived from benzoin) and a mesitoyl
radical (derived from 2,4,6-trimethylbenzoin) vs the ratio of both
initiators in the reaction mixture (ratio = nTMB:nbenzoin). An evaluation
procedure has been adopted, which allows for the minimization of mass
bias and yields the mass bias free intensity ratio ÆG0æΔm/z,0. The figure
also presents the initiation parity between the benzoyl and the mesitoyl
radical at ÆG0æΔm/z,0 = 1 from pMMA, pEMA, and pBMA. The
calculation procedure for the error bars is given in the Supporting
Information.
implying that mass bias effects are not pronounced. (ii) Nearly 3
times excess of 2,4,6-trimethylbenzoin over benzoin is required
(in the case of MMA) to achieve an identical number of chains
initiated with mesitoyl and benzoyl radical fragments. In the case
of EMA and BMA the situation is similar (see below).
The mass bias corrected plot of G(i), ÆGæΔm/z,0, is used to
obtain the direct ratio of chains initiated with mesitoyl fragments
over those initiated with benzoyl radicals. This allows direct and
quantitative comparison of the initiation efficiency of benzoin
and 2,4,6-trimethylbenzoin with respect to the benzoyl and
mesitoyl radicals. Inspection of Figure 4 shows a linear correla-
tion between the ratio of benzoin to 2,4,6-trimethylbenzoin in
the reaction mixture and ÆGæΔm/z,0 of the benzoyl and mesitoyl
radicals that have initiated the polymerization process. Parity in
the initiation between 2,4,6-trimethylbenzoin and benzoin de-
rived fragments (ÆGæΔm/z,0 = 1) occurs at an initiator mixture
composition of 3.0 (MMA), 2.6 (EMA), and 2.4 (BMA).
Considering the error bars given in Figure 4, one can note that
there is a clear difference in the initiation parity between MMA
and EMA/BMA, while the difference between pEMA and pBMA
is not as pronounced.
The reason for the slight disparity between the three mono-
mers may be caused by their different densities and viscosities as
well as the mobility of the ester side chains. All these factors can
have an impact on the polymerization behavior and especially on
the solvent cage around the fragmenting photoinitiator, thus
changing its net-efficiency. Via inspection of Figure 4, one can
come to the conclusion that benzoyl radicals are a factor 3.0 (2.6,
2.4) more efficient in reacting with MMA (EMA, BMA) in the
initiation of bulk free radical polymerizations than mesitoyl
radicals derived from 2,4,6-trimethylbenzoin. This is a surprising
quantitative experimental finding because our former study with
mesitil as source for mesitoyl radicals demonstrated that benzoyl
radicals are a factor 8.6 more likely in being found as a pMMA
chain terminus.31 These results support our hypothesis that not
only the reactivity of specific radicals in relation to each other is
an important parameter, but furthermore the origin of a specific
radical plays a key role in determining its availability to initiate
macromolecular growth. As noted above, upon inspection of
1
,
¼
,
MMA
ΔhD
ðiÞ
2, ¼ ,
FðiÞ ¼
ð1Þ
ð2Þ
ð3Þ
ð4Þ
1
,
¼
,
MMA
MMA
ΔhD
ðiÞ þ ΔhD
ðiÞ
1
,
,
¼
,
,
MMA
ΔhD
ΔhD
ðiÞ
ðiÞ
GðiÞ ¼
2
¼
MMA
D1
,
,
¼
,
MMA
Δh
ΔhD
ðiÞ
G0ði, i ꢀ 1Þ ¼
2
,
¼
MMA
ði ꢀ 1Þ
1
,
,
¼
,
MMA
ΔhD
ΔhD
ðiÞ
G00ði, i þ 1Þ ¼
2,
¼
MMA
ði þ 1Þ
The basic equations for this approach are shown in eqs 2ꢀ4,
where G(i) is the ratio of the peak heights of D1,=,MMA and
D2,=,MMA " i (where i is the present repeat unit, (i ꢀ 1) on repeat
unit lower and (i þ 1) one repeat unit higher). After averaging,
one receives ÆGæ, ÆG0æ, and ÆG00æ, which can be plotted as
ÆGæΔm/z,0 against the used initiator ratios, yielding a mass bias
free ratio of the two disproportionation products in the polymer
sample (the same procedure is used for the evaluation of the
pEMA and pBMA samples).
For the mole fraction diagrams as a function of i of all
monomers see Figure S3 (for pMMA), Figure S6 (for pEMA),
and Figure S9 (for pBMA) in the Supporting Information. As
mentioned above, they depict the mesitoyl radicals that have
initiated the bulk free radical polymerization process towards the
different monomers as a function of the analyzed chain length,
i. Note that equal initiation ability is observed at F(i) = 0.5. Two
important observations can be made: (i) F(i) = F; i.e., there is no
dependence of F on the mass range that has been analyzed,
2549
dx.doi.org/10.1021/ma2001977 |Macromolecules 2011, 44, 2542–2551

Macromolecules
ARTICLE
Figure 4, small differences between pMMA, pEMA and pBMA
can be observed. However, the difference is small at low ratios of
B:A and significantly at higher ratios of B:A.
initiator is employed that produces both fragments at the same
time. From such an initiator, i.e., 2,4,6-trimethylphenylbenzil (1-
mesityl-2-phenylethane-1,2-dione), interferences from energy
transitions or cage effects on the initiation process can be excluded.
Thus, such an initiator was consequently synthesized40 and sub-
jected to PLP as well as the above product evaluation procedure.
Unfortunately, when initiated with low-energy laser pulses (∼0.35
mJ/pulse), no polymer was produced as was already observed with
mesitil (see the discussion above). At higher incident laser energies
(6 mJ/pulse), significant side products were observed making a
reliable evaluation of the disproportionation peaks impossible.
Absolute rate coefficients for the addition of benzoyl and
mesitoyl radicals to n-butyl acrylate have been previously re-
ported using laser flash photolysis and fast time-resolved infrared
spectroscopy (TR-IR).41 Although the current study investigates
methacrylates, a comparison with the data reported in the above
study may nevertheless be instructive. The rate coefficients for
the addition to n-butyl acrylate read k = 2.7 ꢁ 105 Mꢀ1 sꢀ1 for
the benzoyl radical (derived from 1-hydroxycyclohexyl phenyl
ketone) and k = 1.8 ꢁ 105 Mꢀ1 sꢀ1 for the mesitoyl radical
(derived from (2,4,6-trimethylbenzoyl)diphenylphosphine oxi-
de). The rate coefficient for the addition of the benzoyl radical to
n-butyl acrylate is thus 1.5 times higher than that of the mesitoyl
radical, thus indeed suggesting a somewhat decreased reactivity
of the mesitoyl radical. If one entertains the possibility that a
similarly (small) difference in reactivity holds for the two radicals
toward addition to methacrylate monomers, the conclusion can
be drawn that the difference in net-initiation efficiency found
between mesitoyl and benzoyl (a factor of 8.6 when the mesitoyl
radical is derived from mesitil31 and a factor of close to 3.0 when
the mesitoyl radical is derived from TMB) is not, however,
primarily due to a difference in their reactivity toward the
monomer, but rather caused by variable triplet lifetimes and
ISC efficiencies (as well as potential cage effects). However, at
this stage no definite conclusion is possible, as this would require
detailed knowledge of the above parameters. Regarding the
question of the generally increased relative termination ability
of the mesitoyl radical (compared to the benzoyl radical) as
suggested earlier from qualitative data based on both radicals
being derived from benzoin and bis(2,4,6-trimethylbenzoyl)phe-
nylphosphine oxide,29 caution needs to be exercised: Origin
effects may also play a significant role in governing the relative
propensity of primary radicals to undergo termination and
initiation reactions.
To establish whether the observed difference in the propensity
of the mesitoyl vs the benzoyl fragment to initiate the polymer-
ization is potentially caused by a difference in the UV absorptiv-
ity, it is mandatory to establish whether the two initiators absorb
UV light of 351 nm equally well. Figure 5 depicts a comparative
set of UV spectra (mesitil, benzoin, and 2,4,6-trimethylbenzoin).
At a concentration of cPI = 1 ꢁ 10ꢀ3 mol Lꢀ1 (in methanol) and
an optical path length of 10 mm, mesitil (C), benzoin (B), and
2,4,6-trimethylbenzoin (A) have an absorption of 1.1 (set to
100%), 0.07 (6.9%), and 0.02 (2.0%), respectively. It is important
to note that mesitil features the highest absorbance, yet in PLP
experiments with MMA at low laser energy (conditions: f =
100 Hz, T = ꢀ5 ꢀC, E ∼ 0.35 mJ, cPI = 5 ꢁ 10ꢀ3 mol Lꢀ1) it was
not possible to obtain polymer. The cause for this observation
may be twofold: First, it implies that a certain (relatively high)
amount of energy (of photons featuring a 351 nm wavelength) is
required—in the case of mesitil—to cause a significant decay of
the photoinitiator (the generation of radicals occurs out of T1 via
intersystem crossing, ISC). The above experimental finding
supports our quantitative results that the mesitoyl radical—when
derived from mesitil—is found 8.6 times less as end group with
respect to the benzoyl radical in pMMA. Second, in addition to
the efficiency of intersystem crossing (ΦISC) the question of the
lifetime of the triplet state needs to be considered. The longer the
triplet lifetime, the higher the chances that deactivation
processes can reduce the quantum yield for the forming radical
(ΦR). In the current system (benzoin/2,4,6-trimethylbenzoin,
∼0.35 mJ/pulse) the UV absorbance of 2,4,6-trimethylbenzoin is
more than 3.5 times less with respect to benzoin. Our experi-
mental quantitative finding that the mesitoyl radical can be found
∼3 times less as initiating group than the benzoyl radical in bulk
free radical polymerizations of MMA, EMA, and BMA appears to
directly reflect the reduced ability of TMB to absorb the laser UV
light, thus potentially being the cause for the provision of less
mesitoyl radicals per single laser pulse. If this notion is accepted,
the conclusion may be drawn that both initiators TMB and
benzoin—although having different UV absorbtivities at
351 nm—feature similar ΦISC as well as triplet lifetimes.
Considering their structural similarities, such a hypothesis may
not be entirely untenable. However, it must be noted that the
differences in UV absorbance (at 351 nm) which correspond to
the difference in the net-initiation efficiencies may be purely
coincidental. Nevertheless, these observations may imply
that the previously observed 8.6 times difference31 (which rises
up to a factor of 86 when considering that mesitil absorbs
∼10 times more energy at 351 nm than benzoin and the
(possibly questionable, see above) assumption is made that the
difference in UV absorbance is directly translated into a reduced
ability to provide radicals) in the net-initiation efficiency between
mesitoyl and benzoyl radicals derived from mesitil and benzoin is
potentially caused by a significantly increased triplet lifetime of
mesitil (leading to more deactivation processes), a strongly
reduced ΦISC (leading to a smaller population of T1), or a
combination of both. Irrespective of the individual contributions
of the factors governing the net-efficiency, the present data
unambiguously evidence a strong origin dependence of the
propensity of photolytically generated radical fragments for
initiating polymerizations. In principle, a direct comparison of
the mesitoyl and the benzoyl fragment could be made if an
’ CONCLUSIONS
Within the current contribution we present an improved
method for determining initiator efficiencies in the system of
benzoin and 2,4,6-trimethylbenzoin toward MMA, EMA, and
BMA, via online SEC/ESI-MS, which enables collection of
baseline-shift free mass spectra for the precision evaluation of
photolytically generated end groups in macromolecules. In
addition, reducing the temperature to ꢀ5 ꢀC to further reduce
combination products and reducing the laser energy by employ-
ing a beam attenuator leads to the formation of side product free
polymers that only carry the characteristic benzoin-type frag-
ments as end groups. In addition, 2,4,6-trimethylbenzoin was
employed for the first time as a photoinitiator in the bulk free
radical polymerization using the PLP method. The key finding of
the current study is that the benzoyl radical (derived from
benzoin) is 3.0 times more likely to initiate the polymerization
2550
dx.doi.org/10.1021/ma2001977 |Macromolecules 2011, 44, 2542–2551

Macromolecules
ARTICLE
process of MMA than the mesitoyl radical (derived from 2,4,6-
trimethylbenzoin) (2.6 (EMA) and 2.4 (BMA)). These data—in
comparison with previously recorded initiation efficiency data of
mesitoyl and benzoyl fragments derived from mesitil and benzoin—
provide strong evidence that a significant origin dependence of the
initiators’ net-efficiency to commence macromolecular growth.
Further studies in our laboratories will focus on directly measuring
the triplet lifetimes of mesitil and TMB as well as establishing
whether radical cage effects can contribute to differences in initiation
ability.
(15) Buback, M.; Gilbert, R. G.; Hutchinson, R. A.; Klumperman, B.;
Kuchta, F.-D.; Manders, B. G.; O’Driscoll, K. F.; Russell, G. T.; Schweer,
J. Macromol. Chem. Phys. 1995, 196, 3267–3280.
(16) Beuermann, S.; Buback, M.; Davis, T. P.; Gilbert, R. G.;
Hutchinson, R. A.; Olaj, O. F.; Russell, G. T.; Schweer, J.; van Herk,
A. M. Macromol. Chem. Phys. 1997, 198, 1545–1560.
(17) Beuermann, S.; Buback, M.; Davis, T. P.; Gilbert, R. G.;
Hutchinson, R. A.; Kajiwara, A.; Klumperman, B.; Russell, G. T. Macro-
mol. Chem. Phys. 2000, 201, 1355–1364.
(18) Beuermann, S.; Buback, M.; Davis, T. P.; García, N.; Gilbert,
R. G.; Hutchinson, R. A.; Kajiwara, A.; Kamachi, M.; Lacík, I.; Russell,
G. T. Macromol. Chem. Phys. 2003, 204, 1338–1350.
(19) Asua, J. M.; Beuermann, S.; Buback, M.; Castignolles, P.;
Charleux, B.; Gilbert, R. G.; Hutchinson, R. A.; Leiza, J. R.; Nikitin,
A. N.; Vairon, J.-P.; van Herk, A. M. Macromol. Chem. Phys. 2004,
205, 2151–2160.
(20) Beuermann, S.; Buback, M. Prog. Polym. Sci. 2002, 27, 191–254.
(21) Gruendling, T.; Voll, D.; Guilhaus, M.; Barner-Kowollik, C.
Macromol. Chem. Phys. 2010, 211, 80–90.
’ ASSOCIATED CONTENT
S
Supporting Information. NMR data, mole fraction plots
b
for all polymers, SEC/ESI-MS mass spectra of pEMA and pBMA,
tabulation of the observed signals in SEC/ESI-MS spectra, and
structural assignments for pEMA and pBMA as well as an
assessment of errors. This material is available free of charge
via the Internet at http://pubs.acs.org.
(22) Aaserud, D. J.; Prokai, L.; Simonsick, J. W. J. Anal. Chem. 1999,
71, 4793–4799.
(23) Gruendling, T.; Guilhaus, M.; Barner-Kowollik, C. Anal. Chem.
2008, 80, 6915–6927.
(24) Gruendling, T.; Guilhaus, M.; Barner-Kowollik, C. Macromole-
’ AUTHOR INFORMATION
Corresponding Author
*Tel (þ49) 721 608-45642; Fax (þ49) 721 608-45740; e-mail
cules 2009, 42, 6366–6374.
(25) Barner-Kowollik, C.; G€unzler, F.; Junkers, T. Macromolecules
2008, 41, 8971–8973.
christopher.barner-kowollik@kit.edu.
(26) Junkers, T.; Voll, D.; Barner-Kowollik, C. e-Polym. 2009, 076.
(27) Barner-Kowollik, C.; Bennet, F.; Schneider-Baumann, M.; Voll,
D.; R€olle, T.; Facke, T.; Weiser, M.-S.; Bruder, F.-K.; Junkers, T. Polym.
Chem. 2010, 1, 470–479.
(28) Szablan, Z.; Lovestead, T. M.; Davis, T. P.; Stenzel, M. H.;
Barner-Kowollik, C. Macromolecules 2006, 40, 26–39.
(29) Szablan, Z.; Junkers, T.; Koo, S. P. S.; Lovestead, T. M.; Davis,
T. P.; Stenzel, M. H.; Barner-Kowollik, C. Macromolecules 2007,
40, 6820–6833.
’ ACKNOWLEDGMENT
C.B.-K. acknowledges financial support for the current project
from the German Research Council (DFG) as well as from the
Karlsruhe Institute of Technology (KIT) in the context of the
Excellence Initiative for leading German universities and the
Ministry for Science and Arts of the state of Baden-W€urttemberg.
We thank Prof. P. Vana from the University of G€ottingen for the
initial provision of the beam attenuators.
(30) Vana, P.; Davis, T. P.; Barner-Kowollik, C. Aust. J. Chem. 2002,
55, 315–318.
(31) G€unzler, F.; Wong, E. H. H.; Koo, S. P. S.; Junkers, T.; Barner-
Kowollik, C. Macromolecules 2009, 42, 1488–1493.
(32) The term “(net) initiation efficiency” is employed to denote the
propensity of a radical to commence macromolecular growth from the
point of the laser pulse hitting the source molecule. It thus summarizes a
true initiator efficiency (net efficiency), including all effects from the
ability of the initiator to absorb light to the reactivity of the radical
towards the monomer units.
(33) Spichty, M.; Turro, N. J.; Rist, G.; Birbaum, J.-L.; Dietliker, K.;
Wolf, J.-P.; Gescheidt, G. J. Photochem. Photobiol., A 2001, 142, 209–213.
(34) G€unzler, F. PhD Thesis, G€ottingen, 2008.
(35) Buback, M.; G€unzler, F.; Russell, G. T.; Vana, P. Macromolecules
2009, 42, 652–662.
(36) Weinstock, H. H.; Fuson, R. C. J. Am. Chem. Soc. 1936,
58, 1986–1988.
(37) Fuson, R. C.; Weinstock, H. H.; Ullyot, G. E. J. Am. Chem. Soc.
1935, 57, 1803–1804.
(38) Gray, A. R.; Fuson, R. C. J. Am. Chem. Soc. 1934, 56, 739–741.
(39) Gruendling, T.; Guilhaus, M.; Barner-Kowollik, C. Macromol.
Rapid Commun. 2009, 30, 589–597.
(40) Weinstock, H. H.; Fuson, R. C. J. Am. Chem. Soc. 1936,
58, 1233–1236.
(41) 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. J. Am. Chem. Soc. 2002, 124, 14952–14958.
’ REFERENCES
(1) Kaur, M.; Srivastava, A. K. J. Macromol. Sci., Part C: Polym. Rev.
2002, 42, 481–512.
(2) Ge, J.; Trujillo, M.; Stansbury, J. W. J. Dent. Matter 2005,
21, 1163–1169.
(3) Anseth, K. S.; Newman, S. M.; Bowman, C. N. Adv. Polym. Sci.
1995, 122, 177–217.
(4) Fisher, J. P.; Dean, D.; Engel, P. S.; Mikos, A. G. Annu. Rev. Mater.
Res. 2001, 31, 171–181.
(5) Anseth, K. S.; Metters, A. T.; Bryant, S. J.; Martens, P. J.; Elisseeff,
J. H.; Bowman, C. N. J. Controlled Release 2002, 78, 199–209.
(6) Sun, H. B.; Kawata, S. Adv. Polym. Sci. 2004, 170, 169–273.
(7) Arsu, N.; Reetz, I.; Yagci, Y.; Mishra, M. K. Photoinitiated Radical
Vinyl Polymerization. In Handbook of Vinyl Polymers: Radical Polymeri-
zation, Process, and Technology; Mishra, M. K., Yagci, Y., Eds.; CRC Press:
Boca Raton, FL, 2009; Vol. 20, pp 141ꢀ204.
(8) Yagci, Y.; Reetz, I. Prog. Polym. Sci. 1998, 23, 1485–1538.
(9) Kutal, C.; Grutsch, P. A.; Yang, D. B. Macromolecules 1991,
24, 6872–6873.
(10) Gruber, H. F. Prog. Polym. Sci. 1992, 17, 953–1044.
(11) Hageman, H. J. Prog. Org. Coat. 1985, 13, 123–150.
(12) Arsu, N.; Aydin, M.; Yagci, Y.; Jockusch, S.; Turro, N. J. In
Photochemistry and UV Curing: New Trends; Fouassier, J. P., Ed.;
Research Signpost, Fort P.O.: Trivandrum, 2006; p 37/661.
(13) Yagci, Y.; Jockusch, S.; Turro, N. J. Macromolecules 2010,
43, 6245–6260.
(14) Olaj, O. F.; Bitai, I.; Hinkelmann, F. Makromol. Chem. 1987,
188, 1689–1702.
2551
dx.doi.org/10.1021/ma2001977 |Macromolecules 2011, 44, 2542–2551