On the chain length effect: Comparative reactivity of a macroradical and a radical

Article abstract of DOI:10.1016/j.cplett.2008.12.016

The reactivity of two aminoalkyl radicals formed from ethyldimethylaminobenzoate (EDB) and a polymeric derivative (PoEDB) toward different additives such as oxygen, TEMPO, methylacrylate MA, phenols, an onium salt, CBr₄ is investigated through

Full text of DOI:10.1016/j.cplett.2008.12.016

Chemical Physics Letters 468 (2009) 227–230
Contents lists available at ScienceDirect
Chemical Physics Letters
journal homepage: www.elsevier.com/locate/cplett
On the chain length effect: Comparative reactivity of a macroradical and a radical
*
J. Lalevée , X. Allonas, J.P. Fouassier
Department of Photochemistry, UMR 7525 CNRS, University of Haute Alsace, ENSCMu, 3 Rue Alfred Werner, 68093 Mulhouse Cedex, France
a r t i c l e i n f o
a b s t r a c t
Article history:
The reactivity of two aminoalkyl radicals formed from ethyldimethylaminobenzoate (EDB) and a poly-
meric derivative (PoEDB) toward different additives such as oxygen, TEMPO, methylacrylate MA, phenols,
an onium salt, CBr is investigated through Laser Flash Photolysis. The transient absorption spectra of the
4
radicals are in a good agreement with the spectra calculated by using molecular orbitals calculations. The
interaction rate constants with the different additives are measured. The reactivity of EDB and PoEDB
radicals appears as quite similar. Quantum mechanical calculations support the experimental results.
Ó 2008 Elsevier B.V. All rights reserved.
Received 29 October 2008
In final form 6 December 2008
Available online 14 December 2008
1
. Introduction
dine N-Oxyl radical (TEMPO), phenols, diphenyliodonium hexa-
+
fluorophosphate (
U
2
I ) and CBr
4
(Aldrich) have been received
In recent works, we have been concerned with the addition
and used without further purification. For Methyl Acrylate (MA),
the stabilizer (hydroquinone methylether) was removed by col-
umn purification (Aldrich AL-154).
reactions of radicals to double bonds or various additives and have
shown that it is possible to generate and directly observe suitable
radicals involved in photopolymerization reactions and to follow
their interaction with an added compound [1]. For example, this
The spectra and the interaction rate constants of the aminoalkyl
radicals were determined by nanosecond Laser Flash Photolysis
(see details on the apparatus in [10]). A known procedure was used
for the production of the aminoalkyl radicals [11]: it consists in the
Å
has been applied to the addition of an aminoalkyl radical (A ) to
an acrylate (M) double bond and to the further addition of the
Å
Å
A–M radical to M.
generation of a tert-butoxyl radical RO through the photochemical
As the chain length effect is the subject of debates, e.g. in the
polymer science area (see below) [2–9], the aim of the present
work is to directly investigate the comparative reactivity of two se-
lected radicals that differ by the length of a grafted chain. The
corresponding aminoalkyl radicals of two amines (that can be eas-
ily generated), one of them being grafted onto a polymeric chain,
are selected (Scheme 1). Using laser flash photolysis and quantum
mechanical calculations, the reactivity of these two radicals toward
several additives: oxygen, a radical trap (TEMPO: 2,2,6,6-tetra-
methylpiperidine N-Oxyl radical), an acrylate monomer (methyl
acrylate), phenols (Vitamin E, 4-methoxyphenol-4MeOP), an
decomposition of di-tert-butylperoxide, followed by a hydrogen
Å
abstraction reaction between RO and the amine (AH). Then, the
Å
reaction of the aminoalkyl radical (A ) with various additives can
be followed in a way similar to that recently described by us in var-
ious experimental conditions [11c,12,13].
Molecular orbital calculations were carried out with the GAUSSIAN
03 suite of programs [14a]. The absorption spectra were calculated
using the time dependent density functional theory (TDDFT) at TD/
*
*
MPW1PW91/6-311++G level on the basis of the frequency
*
checked geometries calculated at the UB3LYP/6-31G level. Struc-
tural and energetic factors governing the different reactions were
determined by using the Density Functional Theory (DFT), as pre-
viously described [1b,1c]. Calculations were carried out on fully
oxidizing compound (diphenyl iodonium hexafluorophosphate:
+
U
2
I ) and
a
halogen donating compound (CBr
4
)
will be
*
investigated.
optimized structures at UB3LYP/6-31G level – the geometries ob-
tained were frequency checked. The different addition reaction
enthalpies (DHr) were calculated as the energy difference between
2
. Experimental part
the products and the reactants.
The barrier (EaTS) was evaluated by performing UB3LYP/6-
The two amines were: ethyldimethylaminobenzoate (EDB, from
Aldrich) and a polymeric derivative (Poly[oxy(methyl-1,2-etha-
nediyl)]- -(4-dimethylamino)benzoyl- -butoxy, PoEDB); average
number molecular weight: 740 – from Lambson Fine Chemicals
*
*
3
3
11++G single point energies on the corresponding UB3LYP/6-
*
**
*
1G structures (UB3LYP/6-311++G //UB3LYP/6-31G level) and
ZPE corrected at UB3LYP/6-31G level. The activation energy for
a
x
*
the addition reaction is obtained from the enthalpy changes be-
(
Speedcure PDA). Di-tert-butylperoxide, 2,2,6,6-tetramethylpiperi-
*
tween the reactants and the TS structure (
D
H ) as shown in Eq.
(
1) [14b,14c]:
*
Corresponding author. Fax: +33 (0)389336895.
E-mail address: j.lalevee@uha.fr (J. Lalevée).
ꢀ
nꢀÞRT
Ea ¼
D
H þ ð1 ꢁ
D
ð1Þ
0
009-2614/$ - see front matter Ó 2008 Elsevier B.V. All rights reserved.
doi:10.1016/j.cplett.2008.12.016

2
28
J. Lalevée et al. / Chemical Physics Letters 468 (2009) 227–230
O
O
C2H5
H C
O
O
Polymer Chain
H C
2
2
N
N
A1•
M = 192 g/mol
A2•
Mn = 740
Scheme 1.
where Dn* is the number of particle changes when going to the TS
structure. For these first calculations, the thermal energy correction
*
in
D
H has been neglected as in previous studies of radical addition
processes [14b].
The addition rate constants were also calculated; the determi-
nation of the preexponential factor in the Arrhenius equation (A)
is given by the activated complex theory (Eq. (2)) [14b,14c]:
ꢀ
Sꢀꢁ
j
h
T
ꢀ
D
R
ðR TÞꢁ expð1 ꢁ
0
D
n
D
nꢀÞ exp
ð2Þ
A ¼
v
where
D
S* is the entropy changes when going to the TS structure. K
is the Boltzmann constant, R the ideal gas constant, T the tempera-
ture in absolute unit and h the Planck constant, the transmission
coefficient (taken here equal to 1). The harmonic oscillator approx-
v
*
imation was adopted for the
DS calculations allowing the determi-
nation of A by Eq. (2). This treatment was found as accurate enough
to describe the addition of carbon centered radicals to double bonds
[
1g,11c]. From the calculated activation energy and the preexpo-
nential factor, the calculated addition rate constants were given
by the well known Arrhenius equation.
3
. Results and discussion
Å
Å
3
.1. Compared A and A reactivity
1 2
Fig. 1. Transient aminoalkyl radical absorption spectra (taken 500 ns after the laser
Å
Å
flash) in di-tert-butylperoxide. (A): A and (B): A . The quantity kq [Amine] is kept
1
2
The excitation of a di-tert-butylperoxide solution in the pres-
ence of amine leads to the formation of aminoalkyl radicals which
can be directly observed (Fig. 1). The rising time gives a direct
constant to 2 ꢂ 10
6
s .
ꢁ1
access to the hydrogen abstraction rate constants (k
H
) between
8
the tert-butoxyl radical and the amine: 2.3 ꢂ 10
and
ꢂ 10 mol l s for EDB and PoEDB, respectively. The reaction
exothermicities (
Hra) are ꢁ51.2 and ꢁ51.9 kJ/mol for EDB and
PoEDB as determined at UB3LYP/6-31G level. The
Table 1
Å
Å
8
ꢁ1
ꢁ1
Compared reactivity of the aminoalkyl radicals of EDB (A ) and PoEDB (A ) toward
different compounds (reactions at RT in di-tert-butylperoxide).
2
1
2
D
*
Å
Å
a
(C-H) Bond
A
1
A
2
ꢁ
1
ꢁ1
ꢁ1 ꢁ1
k (M
s
)
k (M
s
)
Dissociation Energies (BDE) are 92.8 and 92.1 kcal/mol: as already
stated elsewhere [10], the amine structure has a little influence on
the BDEs.
The transient spectra of the aminoalkyl radicals are quite simi-
lar (Fig. 1) with experimental absorption maxima at 360 and
5
5
MA
5 ꢂ 10
6.2 ꢂ 10
a
a
(
ꢁ166.1,12.5)
(ꢁ170.5,12.2)
8
8
O
2
5 ꢂ 10
8.2 ꢂ 10
8
8
TEMPO
MeOP
Vitamin E
1 ꢂ 10
0.8 ꢂ 10
6
6
4
<2 ꢂ 10
<2 ꢂ 10
8
8
5
30 nm. Accordingly, the predicted maxima for carbon centered
6 ꢂ 10
5.5 ꢂ 10
radicals A and AÅ located at the
Å
+
9
8
8
U I
2
1.6 ꢂ 10
6.7 ꢂ 10
8.5 ꢂ 10
a position of the nitrogen atom
1
2
8
CBr
4
4.5 ꢂ 10
(
Scheme 1) are also almost similar (the oscillator strengths are gi-
ven in brackets): 467.9 nm (0.076), 360.7 nm (0.024), 351.6 nm
0.259) for EDB and 470.9 nm (0.086), 365.2 nm (0.028),
56.2 nm (0.278) for PoEDB. There is an excellent agreement with
the experimental values. Another type of radical could be expected
in PoEDB from a hydrogen abstraction at the position of an oxy-
a
D *
ꢁ1 ꢁ1
S
(J mol
K ) and the barrier (kJ/mol) are given (see text).
(
3
8
rate constants ðk Þ of these radicals with oxygen: 5 ꢂ 10 and
O
2
ꢁ1
8
ꢁ1
a
8 ꢂ 10 mol l s for EDB and PoEDB. The corresponding reaction
exothermicities (
Hrb) are ꢁ115.7 (EDB) and ꢁ116.4 kJ/mol
(PoEDB). The reactivity toward TEMPO is also similar (interaction
gen atom in the polymer chain. The absorption of such a kind of
radical is known to arise in the UV range: for example, the 2-iso-
propyl ketyl radical absorbs below 300 nm [15]. The similar tran-
sient spectra observed in the visible wavelength range for EPD
and PoEDB rule out this possibility and confirm our assignment.
The interaction rate constants of the aminoalkyl radicals with
various additives are reported in Table 1. The kinetics observed
for the decay of the aminoalkyl radical absorption in argon satu-
rated and in aerated solution give a direct access to the interaction
D
8
8
ꢁ1
ꢁ1
rate constants (kTEMPO): 1 ꢂ 10 (EDB) and 0.8 ꢂ 10 mol l s
(PoEDB), reaction exothermicities (
D
Hrc): ꢁ101.8 (EDB) and
ꢁ102.5 kJ/mol (PoEDB). For the reaction with phenols (Vitamin E,
4-MeOP) and the halogen abstraction with CBr , very similar reac-
4
tivities are also observed. Some polar effects can be expected for
the halogen abstraction reaction in agreement with recent results
for aminoalkyl radicals [17]. The same is true for the electron

J. Lalevée et al. / Chemical Physics Letters 468 (2009) 227–230
229
data might be ascribed to (i) the choice of the polymer radicals
(
ethylenic vs. aminoalkyl), (ii) the different possible conformations
Å
in larger systems (such as in A ) which can reduce the contribution
2
of the mechanical effect on the chain length and (iii) the retained
hypothesis for the preexponential factor calculations.
4
. Conclusion
From the whole set of experimental data, it clearly appears that
the reactivity of AÅ and AÅ are quite similar for a large variety of
1
2
chemical reaction (addition to double bond, halogen abstraction,
hydrogen transfer, combination with TEMPO, oxidation process).
Chain length effects for the selected compounds might exist but
contrary to ethylene they remain within the uncertainty of the
measured rate constants (20%).
Å
The A to M addition reaction measured in solution (rate con-
stant: k) truly refers to the initiation step of a radical polymeriza-
ꢃ
tion (rate constant: k
i
) and A–M to M might correspond to the
Å
Fig. 2. Decay traces for
A
at 360 nm for [MA] = 0 M and 0.72 M in di-tert-
Å
Å
2
Å
early propagation (rate constant: k
be used to tentatively mimic the effect of the chain length on k
p
1 2
). The A and A radicals could
butylperoxide. A decrease of the A lifetime is observed.
2
i
for
the addition to a monomer unit. It is usually considered that a pen-
ultimate unit effect could exist in free radical polymerization as of-
ten discussed in pulsed laser polymerization PLP experiments [2–
transfer reaction with
U
2
I+. This reaction corresponds to the oxida-
tion of aminoalkyl radicals i.e. this process has been used to initiate
cationic polymerization processes [11f]. This is the first reported
oxidation rate constants by diaryliodonium salts for this important
class of radicals. The interaction rate constants (kMA) between the
two considered radicals and methylacrylate lead to values of
5
] and theoretically supported (ethylene [7] and acrylonitrile [9]
polymerization). Our experimental data unambiguously demon-
strate the relatively weak influence of the chain length on an addi-
tion reaction to a double bond when considering the behavior of
5
5
ꢁ1
ꢁ1
the available aminoalkyl macroradical AÅ in solution. In the poly-
5
.0 ꢂ 10 and 6.2 ꢂ 10 mol l s
for EDB and PoEDB, respec-
2
tively (Fig. 2) and the reaction exothermicities
DHrd are almost
merization area, the remaining objection that cannot be obviously
Å
identical: ꢁ91 (EDB) and ꢁ92.1 kJ/mol (PoEDB).
solved is the acceptance of A as a representative of a growing polymer
1
radical.
3.2. Chain length effect in the addition of a radical to an acrylate
monomer
Acknowledgment
According to the Arrhenius law, the rate constant of the addi-
tion reaction of A to M is dependent on the energy barrier and
The authors are grateful to Dr D. Anderson from Lambson Fine
Chemicals for the gift of the Speedcure compound.
Å
the preexponentional factor. The largest electronic effects influenc-
ing the barrier are caused by the radical center and the substitu-
ents in the vicinity of the radical center i.e. the polymer chain
does not affect the barrier [6–9,16] and it is assumed that the bar-
rier for the addition of small radicals are representative of the mac-
roradicals barriers. This is also exemplified by the barriers
calculated here for the addition to MA (through a DFT approach
as presented in detail in [11c]) which are found quite similar for
both radicals (12.5 and 12.2 kJ/mol for AÅ and A , respectively).
References
[1] (a) J. Lalevée, X. Allonas, J.P. Fouassier, J. Am. Chem. Soc. 125 (2003) 9377;
(
(
(
b) J. Lalevée, X. Allonas, J.P. Fouassier, J. Phys. Chem. A. 108 (2004) 4326;
c) J. Lalevée, X. Allonas, J.P. Fouassier, J. Org. Chem. 70 (2005) 814;
d) J. Lalevée, X. Allonas, J.P. Fouassier, Macromolecules 38 (2005) 4521;
(e) J. Lalevée, X. Allonas, J.P. Fouassier, Chem. Phys. Lett. 445 (2007) 62;
f) J. Lalevée, X. Allonas, J.P. Fouassier, K.U. Ingold, J. Org. Chem. 73 (2008)
6489;
g) J. Lalevée, F. Morlet-Savary, X. Allonas, J.P. Fouassier, J. Phys. Chem. A. 110
(
Å
1
2
(
The effect of the polymer chain on the frequency factor is more
complex [6–9,16]. For a given conformation, it has a mechanical
origin related to the mass and the size of the chain and also de-
(2006) 11605.
[2] (a) J.P.A. Heuts, G.T. Russell, Eur. Polym. J. 42 (2006) 3;
(
(
(
b) J.P.A. Heuts, G.T. Russell, G.B. Smith, Aust. J. Chem. 60 (2007) 754;
c) J.P.A. Heuts, G.T. Russell, G.B. Smith, A.M. van Herk, Macromol. Symp. 248
2007) 12.
pends on the involved vibrations. In our case, the activation entro-
*
pies for the addition to MA (
D
S ) were evaluated here (within the
[3] J.M. Asua et al., Macromol. Chem. Phys. 205 (2004) 2151.
4] M. Buback, M. Egorov, T. Junkers, E. Panchenko, Macromol. Rapid Commun. 25
2004) 1004.
[
harmonic oscillator approximation solely available in our calcula-
(
ꢁ
1
ꢁ1
for AÅ and A ,
Å
tion procedure) as ꢁ166.1 and ꢁ170.5 J mol
K
1
2
[5] R.X.E. Willemse, A.M. van Herk, E. Panchenko, T. Junkers, M. Buback,
Macromolecules 38 (2005) 5098.
[6] J.P.A. Heuts, in: K. Matyjaszewski, T.P. Davis (Eds.), Handbook of Radical
Polymerization, John Wiley & Sons, 2002.
7] J.P.A. Heuts, R.G. Gilbert, L. Radom, Macromolecules 28 (1995) 8771.
[8] D.M. Huang, M.J. Monteiro, R.G. Gilbert, Macromolecules 31 (1998) 5175.
9] M.L. Coote, T.P. Davis, L. Radom, Macromolecules 32 (1999) 5270.
10] J. Lalevée, X. Allonas, J.P. Fouassier, J. Am. Chem. Soc. 124 (2002) 9613.
11] (a) M. Finn, R. Friedline, N.K. Suleman, C.J. Wohl, J.M. Tanko, J. Am. Chem. Soc.
126 (2004) 7578;
respectively. This leads to a 1.6 factor for the ratio of the preexpo-
nential factors ðAÅ =AÅ Þ. The barriers being similar, the addition
A1
A2
rate constants to MA should be in the same ratio: this confirms
[
Å
the relatively low experimental chain length effect in A compared
to A . These calculations are based on different approximations;
2
Å
[
1
[
[
absolute values cannot be expected.
In an interesting computational approach on the consecutive
addition of an ethylenic type radical to a further ethylene unit
(
(
(
b) J.M. Tanko, R. Friedline, N.K. Suleman, N.J. Castagnoli, J. Am. Chem. Soc. 123
2001) 5808;
c) J. Lalevée, B. Graff, X. Allonas, J.P. Fouassier, J. Phys. Chem. A. 111 (2007)
[
7], a noticeable chain length effect was evidenced through a de-
crease of the rate constant by a factor of about 10 when going from
the monomeric to the corresponding macro radical. The behavior
difference observed within this approach (where the preexponen-
tial factors for the propagation step were calculated using the
decomposition of the partition functions) and in our experimental
6991;
(d) J.C. Scaiano, J. Phys. Chem. 85 (1981) 2851;
(
(
(
e) D. Griller, J.A. Howard, P.R. Marriott, J.C. Scaiano, J. Am. Chem. Soc. 103
1981) 619;
f) J. Lalevée, M. El-Roz, X. Allonas, J.P. Fouassier, J. Polym. Sci. Part A: Polym.
Chem. 46 (2008) 2008.

2
30
J. Lalevée et al. / Chemical Physics Letters 468 (2009) 227–230
[
[
12] J. Lalevée, X. Allonas, J.P. Fouassier, Chem. Phys. Lett. 415 (2005) 287.
13] J. Lalevée, X. Allonas, J.P. Fouassier, J. Polym. Sci. Part A: Polym. Chem. 44
(c) P.D. Pacey, J. Chem. Edu. 58 (1981) 612; J.I. Steinfeld, J.S. Francisco, W.L.
Hase, Chemical Kinetics and Dynamics, second edn., Prentice Hall, New Jersey,
1998.
(
2006) 3477.
14] (a) M.J. Frisch et al., GAUSSIAN 98, Revision A.11, Gaussian, Inc., Pittsburgh PA,
001;
b) R. Arnaud, R. Subra, V. Barone, F. Lelj, S. Olivella, A. Solé, N. Russo, J. Chem.
[
[15] S. Jockusch, N.J. Turro, J. Am. Chem. Soc. 121 (1999) 3921.
[16] O.F. Olaj, M. Zoder, P. Vana, A. Kornherr, I. Schnoll-Bitai, G. Zifferer,
Macromolecules 38 (2005) 1944.
2
(
Soc. Perkin Trans. II (1986) 1517; R. Arnaud, N. Bugaud, V. Vetere, V. Barone, J.
Am. Chem. Soc. 120 (1998) 5733.
[17] J. Lalevée, X. Allonas, J.P. Fouassier, Chem. Phys. Lett. 454 (2008) 415.