A new efficient photoiniferter for living radical photopolymerization

Article abstract of DOI:10.1021/ma0615093

A new photoiniferter structure, an asymmetric sulfide based on thio-1-phenyl-1H-tetrazole and dithiocarbamate moieties with the thiotetrazole and the dithiocarbamate derived radicals correspond to the initiating and terminal radicals, was investigated. The polymerization of methyl methacrylate (MMA) obtained from Aldrich and 1,6-hexanediol diacrylate (HDDA), was carried out in a sealed glass tube. The monomer conversion was directly followed by real time Fourier Transform Infrared (FTIR) spectroscopy in the near-infrared spectral region. The laminated films of HDDA deposited on a BaF₂ pellet were irradiated with a Xe-Hg lamp. The evolution of the double bond content was followed by real time FTIR spectroscopy and the rates of polymerization were calculated from the linear part of the % conversion vs time curves. The results show that the photoiniferter is very powerful to control the final properties of the formed polymer.

Full text of DOI:10.1021/ma0615093

8216
Macromolecules 2006, 39, 8216-8218
Scheme 1. The New Photoiniferter Proposed (I) and the
Mechanism Associated with the Photopolymerization Process
A New Efficient Photoiniferter for LiWing Radical
Photopolymerization
J. Laleve´e,* X. Allonas, and J. P. Fouassier
Department of Photochemistry, UMR 7525 CNRS,
UniVersity of haute Alsace, ENSCMu, 3 rue Alfred Werner,
68093 Mulhouse Cedex France
ReceiVed July 6, 2006
ReVised Manuscript ReceiVed October 10, 2006
1. Introduction. The radical generation in conventional
radical polymerization is irreversible without any possibility to
control the final properties of the polymer formed such as the
number-average molecular weight (Mn) and the polydispersity
PDI. In controlled or living radical polymerization reactions,
the radicals are reversibly generated,1,2 this idea is already
largely used in thermal polymerizations.1
It is of great interest to apply this approach to light induced
polymerization reactions as shown, e.g., in refs 3-8 because
of the large possibilities of selected applications in photografting
reactions, production of block copolymers, manufacture of
multilayer photomaterials, photolithography, and spatial pho-
tomodification of polymer networks through the development
of 2D and 3D patterned and surface modified polymers with
end functional groups. In these reactions, a photoiniferter can
be used at once as a photoinitiating system, a transfer agent,
and a terminating radical.
Very few structures however have been proposed, the most
efficient of them leading to relatively broad Mn distributions.7-9
There is still a need for new photoiniferter structures providing
high rates of polymerization, high Mn, and narrow PDI and
allowing one to initiate the polymerization of a large range of
monomers for copolymers synthesis.
The new photoiniferter structure proposed in this paper
(Scheme 1) is an asymmetric disulfide based on thio-1-phenyl-
1H-tetrazole and dithiocarbamate moieties (I) where the thiotet-
razole (TZ•) and the dithiocarbamate (DC•) derived radicals
correspond to the initiating and terminating radicals, respec-
tively. Its ability to control a (meth)acrylate photopolymerization
will be compared to those of benzyl dimethyldithiocarbamate
(II) and two other symmetric derivatives: 5,5′-dithiobis (1-
phenyl-1H-tetrazole) (III) and tetramethyl thiuramdisulfide (IV).
Compounds II and IV are already known as photoiniferters.7-9
2. Experimental Part. Samples. Compounds II, III, and
IV (Aldrich) were used without further purification (purity
higher than 98%). I was prepared by a classical method
developed by Alam et al.: IV was added to a solution of III in
benzene and photolyzed for 1 h with a UV lamp (Hamamatsu
L2852) leading to I formation. The final purity of I was higher
than 95%.
Photopolymerization Procedure. The photopolymerization
of MMA was carried out in a sealed glass tube (sample
thickness: 1 mm) after Ar bubbling under irradiation with the
UV-light of a Xe-Hg lamp (Hamamatsu, L2852, 200 W). The
monomer conversion was directly followed by real time FTIR
spectroscopy in the near-infrared spectral region as usually done
for thick samples (4700-4800 cm-1).12
For cross-linking experiments, the laminated films of HDDA
deposited on a BaF2 pellet (25 µm thick; optical density < 0.15
at 366 nm) were irradiated with the same Xe-Hg lamp. The
evolution of the double bond content was continuously followed
by real time FTIR spectroscopy (with a Nicolet, Nexus 870).
Rates of polymerization (Rp) were calculated from the linear
part of the % conversion vs time curves. The compounds studied
were compared to 2,2-dimethoxy-2-phenylacetophenone (DMPA),
this compound being a classical UV photoinitiator.1
Molecular Weight Determination. The number-average
molecular weights Mn and PDI were determined by gel
permeation chromatography (GPC) using a Waters 2690 ap-
paratus equipped with Styragel columns (HR0.5, HR1, HR4)
and a refractive index detector (Waters 410). The molecular
weight resolving range of the columns is between 500 and
600 000 g mol-1. Polystyrene standards were used in order to
generate a universal calibration curve. Tetrahydrofuran (THF)
was used as an eluant.
3. Results and Discussion. The Living Polymerization
Mechanism. Scheme 1 recalls the mechanism of a classical
living radical photopolymerization process as originally pro-
posed by Otsu et al.9 The fourth reaction is photochemically
reversible (through a photodissociation of the dithiocarbamate
end groups) leading to the living character of the polymerization,
which is achieved provided that the exchange between the
reactive and the dormant species is fast in comparison with the
propagation.
For the monomers studied, methyl methacrylate (MMA) was
obtained from Aldrich and 1,6-hexanediol diacrylate (HDDA)
from Cray Valley.
Laser Flash Photolysis (LFP) Experiments. For nanosecond
laser flash photolysis LFP, the setup is based on a pulsed Nd
:Yag laser (Powerlite 9010, Continuum) and a transient absorp-
tion analysis system (LP900, Edinburgh Instruments). The
instrument response was equal to about 10 ns.11
The Photodissociation Process of the Iniferters. Laser
excitation of I, II, and IV at 355 nm leads to a strong absorption
at 600 nm of the DC• radicals (already characterized by Ito et
al.) which rises within the resolution time of the experimental
setup (<10 ns) (see Supporting Information). Interestingly, this
ultrafast cleavage process in II can also mimic the photodis-
sociation of the C-S single bond for the dormant species
* Corresponding author: E-mail: j.lalevee@uha.fr. Telephone: 33 (0)-
389336837. Fax: 33 (0)389336895.
10.1021/ma0615093 CCC: $33.50 © 2006 American Chemical Society
Published on Web 10/26/2006

Macromolecules, Vol. 39, No. 24, 2006
Communications to the Editor 8217
Table 1. Polymerization Rates of Methyl Methacrylate (bulk) and
•
RMn Concentrations in the Presence of the Different Photoiniferters
(1% w/w) and Photopolymerization Rates of a 1,6-Hexanediol
Diacrylate Film (50µM), Where DMPA
(2,2-Dimethoxy-2-phenylacetophenone) Is Used as a Reference UV
Photoinitiator
Rp (M s-1
)
[RMn ] 10-6 mol/L
•
monomer
photoiniferters
MMA
MMA
MMA
MMA
MMA
HDDA
HDDA
HDDA
HDDA
HDDA
I
II
III
IV
0.0027
0.0031
0.018
<0.00045
0.0017
3.8
0.13
0.14
0.89
<0.01
1.9
2.2
13.2
<0.33
1.2
I + IV (1%/1%)
DMPA
I
II
III
IV
(Scheme 1) demonstrating the fast reversibility of the termination
reaction under light irradiation.
For I, in addition to the absorption band ascribed to DC•, a
second absorption centered at 430 nm is also noted; as the same
band is observed for III (in a time scale lower than 10 ns) in
agreement with a previous report,14 it can be safely ascribed to
the TZ• radical generated after the S-S bond dissociation.
The Reactivity of the New TZ• Radical. The polymerization
abilities of the different structures, expressed by the correspond-
ing rates of polymerization Rp of MMA, are gathered in Table
1. Compound III is highly efficient. Therefore, TZ• is presum-
ably an excellent initiating radical. This is also supported by
the very high addition rate constants of TZ• to methyl acrylate
(ka is 5 × 107 M-1 s-1 for TZ• 15 and only 4 × 103 M-1 s-1for
the benzyl radical16).
Figure 1. Effect of the photoiniferters. Number-average molecular
weight (Mn) vs conversion curves (photopolymerization of MMA in
bulk in a sealed glass tube after Ar bubbling and irradiation with the
UV light of a Xe-Hg lamp). Inset: Conversion of MMA in bulk vs
time. Key: square, II, 1%; circle, I, 1% and up; triangle, I 1% + IV
1%.
The two basic photoiniferter structures I and II, used at a
fixed concentration, exhibit almost the same efficiency. The
higher reactivity of TZ• (compared to the benzyl radical) toward
the initiation process is probably counterbalanced by (i) the
lower absorption of I (at 366 nm, the molar extinction
coefficients are 35 and 40 M-1 cm-1 for I and II, respectively)
and (ii) the possibility of an in-cage recombination of the thiyl
radicals for I, a process already encountered in other disulfides.17
As expected, IV is unefficient (it contains two terminating DC•
radicals).
Features of the Living Radical Photopolymerization of
MMA. Figure 1 shows the conversion of MMA vs time and
Mn vs conversion plots for I, II and I + IV as photoiniferters.
The increase of the molecular weight with both the conversion
and the PDIs represent the best parameters to characterize the
living process. For I and II, the Mn vs conversion plots give
linear relationships (this behavior was already known for the
reference II7). This result demonstrates that the photodissociation
of the C-S bond at the chain end, the subsequent propagation
and recombination of the propagating radical with the DC•
radical also occur for I. The Mn evolution follows a linear trend.
For I and II, the deviation from the origin usually observed
(see Otsu et al.) indicates that a part of the carbon-carbon
termination mechanism still remains in the early stage of the
reaction, the linear trend being always ascribed to the carbon-
dithiocarbamyl radical recombination.
Figure 2. Effect of the photoiniferters. PDIs vs conversion. MMA in
bulk. Inset: PDIs vs Mn. Key: I (circle); II (square) and I + IV (up
triangle).
by a ∼ 2-fold factor (compared to I and II) in agreement with
the increase of the DC• production.
The initial slope of the conversion vs time plots is directly
related to the polymerization rate (Rp/[M0]100swhere M0 stands
for the initial monomer concentration). Using the classical
polymerization relationship (eq 1), the propagating radical
concentrations are evaluated in Table 1 (taking a propagation
rate constant of 143 M-1 s-1 for MMA18). For the (I + IV)
•
Rp ) kp [RMn ][M0]
(1)
The PDIs are lower than 2.0 at Mn < 8000 for I and II, albeit
slightly lower in the case of I for a given Mn value (Figure 2).
However, interestingly, the molecular weight for a given
conversion is found higher for I than for II. This corresponds
•
two- component system, the RMn concentration is decreased

8218 Communications to the Editor
Macromolecules, Vol. 39, No. 24, 2006
to a great advantage exhibited by I, compared to II, which can
be useful for an access to a larger Mn range. In the case of III,
higher PDIs are obtained (3 to 4): as expected, this compound
is not suitable for a photopolymerization control.
polymer. It leads to high Mn whereas a combination of I with
a tetramethyl thiuramdisulfide IV is better for obtaining both
low Mn and narrower PDI. Moreover, the control of the
polymerization of multifunctional monomers usable in the UV
Curing area also appears as feasible. Compound I can also create
a large variety of dormant species (R-Mn-DC) in a polymer
matrix: the formation of a PMMA-polystyrene copolymer
through a sequential approach was easily achieved (see Sup-
porting Information). Other copolymers can be probably formed
using the outstanding low selectivity property of TZ• toward
monomers (mentioned above). In fact, preliminary results have
underlined the high reactivity of TZ• radical i.e., the addition
rate constants are found higher than 107 M-1 s-1 for the addition
to electron deficient (methyl acrylate...) or electron rich (vinyl
acetate...) monomers.
As previously noted by Bertin et al. for the benzyl dithio-
carbamate derivatives, the PDIs can be greatly improved by an
increase of the terminator radicals concentration (DC•). In Figure
2, the PDIs obtained with I + IV are better with an average
value of 1.6 for conversions between 10 and 30%. The addition
of IV also leads to a decrease of the positive intercept in the
Mn vs conversion plots evidencing a decrease of the usual
bimolecular carbon-carbon termination reaction due to a better
control of the polymerization (in agreement with the lower PDIs
observed).
The III + IV systems also lead to a living character (linear
Mn/conversion relationship and lower PDI than with III alone).
The relative behavior of I vs III + IV is different since the
absorption and the photochemical and photophysical properties
of the systems are different. A large panel of situations can be
of course encountered: III alone should lead to a bad control
with a high Rp, the addition of IV should result in a better control
with a concomitant strong decrease of Rp. For applications which
do not require such conditions, a two component system (III
+ IV) can allow a good polymerization rate/control compromise.
Behavior of the Photoiniferters for Cross-Linking Experi-
ments of HDDA Films. The Role of the DC• Radicals. The
polymerization ability of the photoiniferters in a HDDA matrix
is quite similar to that noted for MMA in bulk (Table 1skinetics
are given in the Supporting Information). The polymerization
rates is about 20 times lower with I or II than with DMPA).
III exhibits a high reactivity but one-fourth than DMPA: this
lower efficiency of III is attributable to the lower UV absorption
of this compound compared to DMPA at 366 nm i.e., the
absorption coefficient at 366 nm (maximum emission of the
light source) is 127 M-1 cm-1 for DMPA and only 25 M-1
cm-1 for III. A control of the polymerization also occurs.
Indeed, the concentration of the DC• radical can be changed by
using a two component system (III + IV) in which the
concentration in IV is adjusted. A plot of Rp vs 1/[IV] yields:
Rp ) -0.046 + 0.102(1/[IV]) (r ) 0.995 and n ) 5; see
Supporting InformationsFigure 3). This evidences a direct
relationship between Rp and 1/[IV] and therefore 1/[DC•]. A
steady-state analysis of Scheme 1 leads to a polymerization rate
described by eq 2 where kp, kt, Φi, and Φr represent the
propagation and the termination rate constants, the initiation
and dissociation quantum yields of the dormant species,
respectively (Iabs1 and Iabs2 stand for the light intensity absorbed
by the photoiniferter and the dormant species). The agreement
between eq 2 and the above relationship supports the fact that
the termination process corresponds to the recombination of the
Acknowledgment. Thanks are due to Dr. C. Delaite and
Dr. K. Hariri for the Mn and PDI measurements.
Supporting Information Available: Figure 1, polymerization
kinetics of HDDA film (50 µm) in the presence of 2,2-dimethoxy-
2-phenylacetophenone (DMPA), I, II, III, or IV, Figure 2, decay
of the dimethyldithiocarbamyl radicals observed at 600 nm by LFP,
Figure 3, plot of Rp vs 1/[IV] for HDDA films, and text giving a
discussion of the copolymer synthesis. This material is available
free of charge via the Internet at http://pubs.acs.org.
References and Notes
(1) Matyjaszewski, K. AdVances in Controlled/LiVing Radical Polym-
erization; ACS Symposium Series; American Chemical Society:
Washington, DC, 2002.
(2) Otsu, T. J. Polym. Sci. Part. A.; Polym. Chem. 2000, 38, 2121-
2136.
(3) Luo, N.; Metters, A. T.; Hutchinson, J. B.; Bowman, C. N.; Anseth,
K. S. Macromolecules 2003, 36, 6739-6745.
(4) Reddy, S. K.; Sebra, R. P.; Anseth, K. S.; Bowman, C. N. J. Polym.
Sci. Part. A.; Polym. Chem. 2005, 43, 2134-2144.
(5) Otsu, T.; Matsunaga, T.; Doi, T.; Matsumoto, A. Eur. Polym. J. 1995,
31, 67-78.
(6) Kuriyama, A.; Otsu, T. Polym. J. 1984, 16, 511-516.
(7) Otsu, T.; Taraki, T. Polym. Bull. (Berlin) 1986, 16, 277-282.
(8) Bertin, D.; Boutevin, B.; Gramain, P.; Fabre, J. M.; Montginoul, C.
Eur. Polym. J. 1998, 34, 85-90.
(9) Otsu, T.; Yoshida, M. Makromol. Chem. Rapid Commun. 1982, 3,
127-132.
(10) Alam, M. M.; Watanabe, A.; Ito, O. J. Org. Chem. 1995, 60, 3440-
3444.
(11) Laleve´e, J.; Allonas, X.; Fouassier, J. P. J. Am. Chem. Soc. 2002,
124, 9613-9621.
(12) Stansbury, J. W.; Dickens, S. H. Dental Mater. 2001, 17, 71-86.
(13) Ito, O.; Matsuda, M. J. Phys. Chem. 1981, 85, 1365-1368.
(14) Alam, M. M.; Watanabe, A.; Ito, O. Int. J. Chem. Kinet. 1996, 28,
405-411.
(15) Laleve´e, J.; Allonas, X.; Fouassier, J. P. To be published.
(16) Fischer, H.; Radom, L. Angew. Chem., Int. Ed. 2001, 40, 1340-
1371.
(17) Lochschmidt, A.; Eilers-Konig, N.; Heineking, N.; Ernsting, N. P.
J. Phys. Chem. A 1999, 103, 1776-1784.
(18) Brandrup, J.; Immergut, E. H. Polymer Handbook, 3rd ed.; John
Wiley and Sons: New York, 1989.
(19) A previously published study by Kannurpatti et al. (Macromolecules
1996, 29, 7310) on the use of a disulfide as both a photoinitiator
and an iniferter agent has indicated a square root dependence between
Rp and the disulfide concentration (tetraethylthiuram disulfide). This
current difference of behavior with III + IV can be ascribed to the
termination mechanism. A direct comparison with this previous work
is not straightforward.
•
propagating radical (RMn ) with DC•, the usual bimolecular
termination (between two propagating radicals) being predomi-
nantly avoided.19
Rp ) kp[ΦiIabs1 + ΦrIabs2][M]/(kt[DC•])
Conclusion
(2)
The new asymmetric disulfide photoiniferter I proposed here
appears as powerful to control the final properties of the formed
MA0615093