Laser flash photolysis and CIDNP studies of steric effects on coupling rate constants of imidazolidine nitroxide with carbon-centered radicals, methyl isobutyrate-2-yl and tert-butyl propionate-2-yl

Article abstract of DOI:10.1021/jo060787x

Time-resolved chemically induced dynamic nuclear polarization (TR-CIDNP) and laser flash photolysis (LFP) techniques have been used to measure rate constants for coupling between acrylate-type radicals and a series of newly synthesized stable imidazolidin

Full text of DOI:10.1021/jo060787x

Laser Flash Photolysis and CIDNP Studies of Steric Effects on
Coupling Rate Constants of Imidazolidine Nitroxide with
Carbon-Centered Radicals, Methyl Isobutyrate-2-yl and
tert-Butyl Propionate-2-yl
Dmitry Zubenko,† Yuri Tsentalovich,† Nataly Lebedeva,† Igor Kirilyuk,‡
Galina Roshchupkina,‡ Irina Zhurko,‡ Vladimir Reznikov,‡ Sylvain R. A. Marque,§ and
Elena Bagryanskaya*,†
International Tomography Center SB RAS, NoVosibirsk, Russia, Institute of Organic Chemistry SB RAS,
NoVosibirsk, Russia, and UniVersite´ de ProVence case 542, AVenue Escadrille Normandie-Niemen13397,
Marseille Cedex 20, France
elena@tomo.nsc.ru
ReceiVed April 13, 2006
Time-resolved chemically induced dynamic nuclear polarization (TR-CIDNP) and laser flash photolysis
(LFP) techniques have been used to measure rate constants for coupling between acrylate-type radicals
and a series of newly synthesized stable imidazolidine N-oxyl radicals. The carbon-centered radicals
under investigation were generated by photolysis of their corresponding ketone precursors RC(O)R (R
) C(CH3)2-C(O)OCH3 and CH(CH3)-C(O)-OtBu) in the presence of stable nitroxides. The coupling
rate constants kc for modeling studies of nitroxide-mediated polymerization (NMP) experiments were
determined, and the influence of steric and electronic factors on kc values was addressed by using a
Hammett linear free energy relationship. The systematic changes in kc due to the varied steric (Es,n) and
electronic (σL,n) characters of the substituents are well-described by the biparameter equation log(kc/M-
1s-1) ) 3.52σL,n + 0.47Es,n + 10.62. Hence, kc decreases with the increasing steric demand and increases
with the increasing electron-withdrawing character of the substituents on the nitroxide.
Introduction
reversible traps for other free-radical species,3 etc. Recently,
the synthesis of a series of imidazolidine nitroxides has been
reported which have different bulky substituents at positions 2
and 5 of the imidazolidine ring (1-4 in Table 1).4 It was shown
that they can be successfully applied as pH sensitive spin probes
in biological systems.5 The most important finding was the fact
that these new nitroxides were 20-30 times more stable in the
presence of ascorbate and had significantly longer half-lives in
rat blood as compared to 2,2,5,5-tetra-Me analogues. But the
Chemical reactions with participation of stable nitroxides have
attracted great attention of scientists during the past decades.
Nitroxides are widely used as spin labels in biology1 and as pH
probes in biological systems,2 as mediators in nitroxide mediated
polymerization (NMP) due to their possibility of acting as
* Address correspondence to this author.
In memory of Hanns Fischer.
† International Tomography Center.
‡ Institute of Organic Chemistry SB RAS.
§ Universite´ de Provence case 542.
(3) Solomon, D. H. J. Polym. Sci. Part A: Polym. Chem. 2005, 43, 5748-
5764.
(4) Kirilyuk, I. A.; Bobko, A. A.; Grigor’ev, I. A.; Khramtsov, V. V.
Org. Biomol. Chem. 2004, 2, 1025-1030.
(5) Bobko, A. A.; Khramtsov, V. V.; Grigor’ev, I. A. Org. Biomol. Chem.
2005, 3, 1269-1274.
(1) Volodarskii, L. B. Imidazoline Nitroxides; CRC Press: Boca Raton,
FL, 1988. Kocherginsky, N.; Swarts, H. M. Nitroxide Spin Labels; Reactions
in Biology and Chemistry; CRC Press: Boca Raton, FL, 1995.
(2) Khramtsov, V. V.; Grigor’ev, I. A.; Foster, M. A.; Lurie, D. J.;
Nicholson, I. Cell. Mol. Biol. (Paris) 2000, 46, 1361-1374.
10.1021/jo060787x CCC: $33.50 © 2006 American Chemical Society
Published on Web 07/08/2006
6044
J. Org. Chem. 2006, 71, 6044-6052

Steric Effects on Coupling Rate Constants
TABLE 1. Recombination Rate Constants kc for Nitroxides 1-4, 8a-g, and 12 and Radicals MiB and TBP at Room Temperature, with Their
Electrical Hammett Constants σL,n and Steric Constants Es,n
c
a In benzene unless otherwise mentioned. b Given by eq 2 in ref 29. σL,Me ) -0.01, σL,Et ) -0.01, σL,C(CH ) ) 2σL,Et ) -0.02, σL,C(CH ) ) 2σL,Et
)
2
4
3 4
-0.02, σL,Ph ) 0.12, σL,NMe ) 0.17, σL,CMeNMe ) -0.01, see refs 30 and 31. d r(Me) ) 0.0, r(Et) ) -0.38, see ref 29. e Given by eqs 3 and 5, see text.
f Given by eqs 4 and 5, see 2text. g aN for nitrog2en hyperfine constant in mT in CHCl3. h In acetonitrile. Thus a correcting factor of 2 times was applied to
obtain kc in benzene, see ref 14. i For r(C(CH2)4), r(C(CH2)5), r(CH(CH2)4CH), r(n-Bu), and r(Ph), see text. j Not determined.
reactions of these nitroxides with alkyl radicals as well as overall
suitability of these nitroxides for NMP has not yet been
determined.
NMP6 is one of the techniques for the synthesis of new block
copolymers.7 These materials have many applications in chem-
istry, physics, biology, and material science, and function as
J. Org. Chem, Vol. 71, No. 16, 2006 6045

Zubenko et al.
SCHEME 1
SCHEME 2
key building blocks in many areas of nanotechnology.8 The
simplest mechanism of NMP (Scheme 1) involves the reversible
dissociation of a dormant nitroxyl end-capped polymer chain
with rate constant kd, the propagation of carbon-centered radicals
kp, the cross-coupling reaction between the nitroxide and the
growing polymer chain with rate constant kc, and the irreversible
self-termination reactions of the carbon-centered radicals with
rate constant kt.
extensive studies on the polar and the steric effects for the
recombination between acrylate- and methacrylate-type radicals
and nitroxide.
In this work, we present the synthesis of new imidazolidine
nitroxides 8a-g and 12 (Table 1). The recognized preconditions
for testing these particular nitroxides for NMP processes are
that (i) there should exist low oxidation potentials for the
imidazolidine nitroxides (in general they should be much lower
than those of imidazolidine-3-ones)19 and (ii) there should exist
convenient methods for the synthesis of imidazolidine nitroxides
with bulky substituents at positions 2 and 5 of the imidazole
ring.4 We applied the time-resolved chemically induced dynamic
nuclear polarization (TR CIDNP) to determine the fate of the
alkyl radicals R•, (R ) C(CH3)2-C(O)OCH3 and CH(CH3)-
C(O)-OtBu) in the absence and in the presence of imidazolidine
nitroxides. Then, the coupling rate constants between these two
species were measured by laser flash photolysis (LFP) experi-
ments. It is our goal to derive kinetic parameters for modeling
NMP reactions, and to reveal the influence of steric and
electronic factors upon these reactions. Within the series, the
nitroxides we have studied differ in the steric and/or the
electronic character of the functional groups surrounding the
nitroxide moiety. The carbon-centered radicals were generated
in the presence of the stable nitroxides by laser flash photolysis
of the corresponding ketone precursors RC(O)R, whose pho-
tochemistry is well understood.20 The rate constants of decom-
position of alkoxyamines based on these series of nitroxides as
well as polymerization results will be published elsewhere.
Several authors9-12 have shown that for NMP to be efficient,
the equilibrium constant K ) kd/kc should fall in the range from
10-7 to 10-11 mol L-1. The values of kd and kc are sensitive
functions of the alkoxyamine structure,13,14 and knowledge of
both the kd and kc values is crucial for successful control of
polymer molecular weight. Recent efforts to improve the NMP
technique have focused on the precise tuning of the reactivity
of the corresponding radical species. It has been demonstrated
that the introduction of bulky substituents on the nitroxide
R-carbons may improve their suitability for NMP reactions.
Thus, the development of new nitroxide mediators with higher
kd and not too low kc values is an important endeavor. Such
new mediators could be used not only in styrene polymerization,
but also in the polymerization of methacrylates, acrylamides,
and dienes.15 In a recent paper, Siegenthaler and Studer16 showed
the importance of kc on the efficiency of the polymerization.
Thus, it is important to understand clearly the different processes
and effects involved during the reformation of the alkoxyamines,
and to have tools for predicting kc values as for the homolysis
of alkoxyamines. To date, there have only been two reports on
the steric effects of the aminoxyl radical structure on the values
of kc.17,18 Fischer et al.17 observed a correlation between the
energy of steric hindrance and coupling rate constants for the
cross-coupling of the six-membered-ring piperazinone N-oxyl
type radical and styryl radical. As far as we know, there are no
Results
Synthesis of Nitroxides. The nitroxides 8a-g were synthe-
sized in four steps, according to Scheme 2. The nitroxide 12
was prepared in four steps, according to the Scheme 3.
(6) U.S. Patent 4,581,429. Solomon, D. H.; Rizzardo, E.; Cacioli, P.
Chem. Abstr. 1985, 102, 221335q.
Time-Resolved CIDNP Experiments. The carbon-centered
radicals MiB (methyl isobutyrate-2-yl) and tBP (tert-butyl
propionate-2-yl) were generated by laser pulse photolysis of the
appropriate symmetric ketones R-CO-R, as a propagating
chain radical model formed for methyl methacrylate and tert-
butyl acrylate polymerizations, respectively. The photochemistry
and resulting ketone reaction paths are shown in Scheme 4.14,21
After excitation of the n-π* transition and intersystem crossing,
the ketones undergo R-cleavage from the triplet state to yield
geminate (in-cage) acyl-alkyl radical pairs (reaction 1 in
Scheme 4). The geminate reactions (reaction 2 in Scheme 4) of
recombination and elimination of vicinal hydrogen result in the
restoration of the initial ketone R-CO-R (I in Figure 1) and
the formation of aldehyde R-CHO (V in Figure 1)) and alkene
R(-H) (IV in Figure 1). The escaped acyl radicals undergo fast
decarbonylation with the rate constant kCO ) 2 × 107 s-1 for
(7) Georges, M. K.; Veregin, R. P. N.; Kazmaier, P. M.; Hamer, G. K.
Macromolecules 1993, 26, 2987-2988.
(8) Nanoscale Science and Technology; Kelsall, R. W., Hamley, I. W.,
Geoghegan, M., Eds.; John Wiley & Sons: Chichester, UK, 2005.
(9) Souaille, M.; Fischer, H. Macromolecules 2000, 33, 7378-7394 and
references therein.
(10) Fischer, H.; Souaille, M. Chimia 2001, 55, 109-113 and references
therein.
(11) Goto, A.; Fukuda, T. Prog. Polym. Sci. 2004, 29, 329-385.
(12) Greszta, D.; Matyjaszewski, K. Macromolecules 1996, 29, 7661-
7670.
(13) Bertin, D.; Gigmes, D.; Marque, S.; Tordo, P. Macromolecules 2005,
38, 2638-2650.
(14) Beckwith, A. L. J.; Bowry, V. W.; Ingold, K. U. J. Am. Chem. Soc.
1992, 114, 4983-4992. Boury, V. W.; Ingold, K. U. J. Am. Chem. Soc.
1992, 114, 4992-4997. Chateauneuf, J.; Lusztyk, J.; Ingold, K. U. J. Org.
Chem. 1988, 53, 1629-1630.
(15) Hawker, C. J.; Bosman, A. W.; Harth, E. Chem. ReV. 2001, 101,
3661-3688 and references therein.
(16) Siegenthaler, K. O.; Studer, A. Macromolecules 2006, 39, 1347-
1352.
(17) Marque, S.; Sobek, J.; Fischer, H.; Kramer, A.; Nesvadba, P.;
Wunderlich, W. Macromolecules 2003, 36, 3440-3442.
(18) Iwao, K.; Sakakibara, K.; Hirota, M. J. Comput. Chem. 1998, 19,
215-221.
(19) Chong, Y. K.; Ercole, F.; Moad, G.; Rizzardo, E.; Thang, S. H.
Macromolecules 1999, 32, 6895.
(20) Turro, N. J. Molecular Photochemistry, 3rd ed.; Benjamin: Menlo
Park, CA, 1974.
6046 J. Org. Chem., Vol. 71, No. 16, 2006

Steric Effects on Coupling Rate Constants
SCHEME 3
SCHEME 4
intensive CIDNP. The CIDNP spectrum obtained 100 µs after
photolysis of Pest (see Experimental Section) in the absence of
nitroxide is presented in Figure 1a. The phases of the CIDNP
signals are the absorption for the initial compound R-CO-R
(I in Figure 1) and the emission for the escape products R-R
and RH (II and III in Figure 1, respectively), as predicted by
Kaptein’s rules27 (∆g > 0, A(CH3) > 0). The alkene R(-H)
can be formed both in geminate reactions and in the bulk (IV
in Figure 1). Therefore, immediately after the laser pulse the
signal from the CH3 group of R(-H) is in emission, which then
decays on the microsecond time scale due to formation of
absorptively polarized molecules in the bulk (this is the CIDNP
cancellation effect28). In Figure 1, inserts above each signal show
the kinetic behavior of the signals on the microsecond time scale.
In the presence of nitroxides, both the CIDNP spectra and
kinetics change significantly. Figure 1b shows the spectrum
obtained 100 µs after laser flash photolysis of a 9 mM Pest
(see Experimental Section) solution in the presence of 8 mM
nitroxide 8f. Strong emissive signals (VI in Figure 1b) are
observed at 1.4-1.5 ppm, which corresponds to the CH3 group
of the coupling product R-Y resulting from the reaction of
nitroxide and alkyl radicals. These signals show fast initial
growth, but after 1-2 µs they practically do not change. The
presence of the nitroxide inhibits the formation of bulk products
R-R, RH, and R(-H), and this is indicated by the absence of
any emissive bulk polarization in these products (Figure 1b).
Generally, the reaction between alkyl radical R and nitroxide
Y can proceed either via coupling (reaction 5, two diastereo-
mers) or elimination of vicinal hydrogen (reaction 6). The
absence of the bulk polarization at the products of reaction 6,
R(-H) and YH, demonstrates that kc . kdis, and the main
channel of the reaction between R• and Y• is the coupling.
Similar results were obtained in CIDNP measurements per-
formed with other nitroxides, as well as in the photolysis of
tBest (see Experimental Section) with nitroxides.
MiB22 and kCO ) 8 × 106 s-1 for tBP23 (reaction 3 in Scheme
4). The consequent bulk reactions (reaction 4 in Scheme 4) of
radical termination give rise to the formation of products R-R,
RH, and R(-H) (II, III, and IV in Figure 1, respectively). In
the presence of nitroxide these reactions are partly or entirely
replaced by the coupling of carbon-centered radicals with the
nitroxide and elimination of vicinal hydrogen from the carbon-
centered radicals with the nitroxide (reactions 5 and 6 forming
either hydroxylamine YH and alkene R(-H), or the stable
adduct R-Y).24 In general, this system is well suited for deriving
coupling rate constants of alkyl radicals with nitroxides by
means of either flash photolysis21 or the switched external
magnetic field CIDNP technique.25,26
In the high magnetic field of the NMR spectrometer the
photolysis of either ketone leads to the formation of highly
The observed structure of CIDNP signals of RY (Figure 1b)
indicates the existence of two diastereoisomers of the resulting
coupling species with two unequivalent groups of methyl protons
in each of them. This fact is confirmed by NMR spectra of
corresponding alkoxyamines. Thus, the CIDNP study shows the
(21) Sobek, J.; Martschke, R.; Fischer, H. J. Am. Chem. Soc. 2001, 123,
2849-2857.
(22) Tsentalovich, Y. P.; Fischer, H. J. Chem. Soc., Perkin Trans. 2 1994,
4, 729-733.
(23) Schuh, H.; Fischer, H. HelV. Chim. Acta 1978, 61, 2463.
(24) Ananchenko, G.; Fischer, H. J. Chem. Soc., Perkin Trans. 2 2001,
1887-1889.
(25) Bagryanskaya, E. G.; Gorelik, V. R.; Sagdeev, R. Z. Chem. Phys.
Lett. 1997, 264, 655.
(26) Lebedeva, N. V.; Zubenko, D. P.; Bagryanskaya, E. G.; Ananchenko,
G. S.; Marque, S.; Tordo, P.; Sagdeev, R. Z. Phys. Chem. Chem. Phys.
2004, 6, 2254-2259.
FIGURE 1. CIDNP spectra, obtained upon the photolysis of MiB in
the absence (a) and in the presence (b) of nitroxide 8f. Inserts: CIDNP
kinetics of corresponding NMR lines in microsecond time scale. Peaks
attribution: (I) MeOOC-C(CH3)2-C(O)-C(CH3)2-COOMe, (II)
MeOOC-C(CH3)2-C(CH3)2-COOMe, (III) MeOOC-CH(CH3)2, (IV)
MeOOC-C(CH3)dCH2, (V) MeOOC-C(CH3)2-CHO, and (VI)
MeOOC-C(CH3)2-Y.
(27) Salikhov, K. M.; Molin, Y. N.; Sagdeev, R. Z.; Buchachenko, A.
L. Spin polarization and magnetic effects in radical reactions; Elsevier:
Amsterdam, The Netherlands, 1984.
(28) Vollenweider, J.-K.; Fischer, H.; Hennig, J.; Leuschner, R. Chem.
Phys. 1985, 97, 217-234.
J. Org. Chem, Vol. 71, No. 16, 2006 6047

Zubenko et al.
FIGURE 2. Time dependence of the methyl isobutyrate radical
absorption at 325 nm in benzene (room temperature, various concentra-
tions) and in the presence of 8f.
FIGURE 4. Temperature dependence of the cross-coupling rate
constant of the carbon-centered radical MiB with nitroxides 1 (O) and
3 (b).
et al.17 showed that the steric effect of substituents on the R
positions of the piperazinone N-oxyl moiety is expressed by eq
1 with Es, the steric constant, and ri, the individual steric
constants of each substituent. In the same paper, it was shown
that kc is linearly correlated to Es,n (eq 2), the steric constant of
the whole molecule.
Es(R1R2R3C) ) -2.104 + 3.429r1(R1) + 1.978r2(R2) +
0.649r3(R3) (1)
log(kc/M-1 s-1) ) log(kc,0/M-1 s-1) + δEs,n
(2)
FIGURE 3. Concentration dependence of the apparent rate constant
km for the cross-coupling of MiB with nitroxides in benzene at room
temperature: (2) for 1, (3) for 2, (0) for 4, (f) for 8b, and (O) for
8f.
It has to be mentioned that the larger the group, the more
negative ri, thus r1 is for the small group, r2 for the medium
group, and r3 for the large group. In his paper, Fischer did not
determine the cyclic steric constant, and thus assumed all the
possibilities for the size of the ring.17 Our series is too small,
and the alkyl radicals studied are very different from the styryl
radical studied by Fischer et al.,17 therefore the cyclic strain
for the five-membered ring is not determined. The cyclic steric
constant r(C5) is set in the first position when assumed larger
effective coupling reaction of a series of imidazolidine nitroxides
with alkyl radicals.
The rate constants kc of radical cross-termination (reaction
4) were determined by using laser flash photolysis. Alkyl
radicals R were generated by 308 nm pulse photolysis of the
corresponding ketones, and their decays were monitored at the
absorption maximum of 325 nm. The acrylate radical absorption
spectrum has been measured by Sobek et al.21 In the absence
of nitroxides, the carbon-centered radicals decay by second-
order kinetics with kt/ꢀ ) 4.7 × 106 cm/s for MiB and kt/ꢀ )
1.9 × 106 cm/s for tBP, where ꢀ is the extinction coefficient of
the corresponding alkyl radical. With added nitroxide, their
decay accelerates and becomes pseudo-first-order decay (Figure
2) according to step 5 in Scheme 4. Typically, our measurements
were performed with at least four different concentrations of
nitroxides. To separate the exponential contribution into signal
decay, the kinetic traces were obtained with 4-5 different
intensities of the laser pulse for each solution. The observed
pseudo-first-order rate constant km was found to be proportional
to the nitroxide concentration, km ) k0 + kc[Y] (k0 for the side
reactions). Some typical kinetics obtained from these measure-
ments are shown in Figure 3, and the calculated values of kc
are listed in Table 1.
than the methyl group, that is, E1 is given by eq 3. When the
s
cyclic steric constant is assumed between that of the methyl
and the ethyl groups, the steric constant is also given by E1 .
s
The r(C5) is set in the third position when it is assumed smaller
than r(Et), that is, E3 is given by eq 4.
s
E1 ) -2.104 + 3.429r(C5) + 1.978r2(R2) + 0.649r3(R3)
s
(3)
E3 ) -2.104 + 3.429r1(R1) + 1.978r2(R2) + 0.649r(C5)
s
(4)
It was also assumed that the steric constant of the molecule
E
s,n is given by the sum of the steric constant of substituent in
positions 2 and 5 (eq 5).29
Es,n ) Es(R1R2R3C2) + Es(R1R2R3C5)
(5)
The values of kc for 2 and 8e are close to kc of 8d and indicate
that the steric demands of the spiro-cyclohexyl moiety and
geminal n-butyl groups are close to the steric demand of the
geminal ethyl groups. This is also observed for 8a and 3. The
same remarks hold for the spiro-cyclopentyl moiety in the
For cross-reaction of nitroxides 1 and 3 with MiB, rate
constants were measured over a considerable temperature range
(300-370 K) and data are shown in Figure 4.
The Multiparameter Approach. As we mentioned in the
Introduction, to date, there have only been two reports on steric
effects of the nitroxide structure on the values of kc.17,18 Fischer
(29) Marque, S. J. Org. Chem. 2003, 68, 7582-7590.
6048 J. Org. Chem., Vol. 71, No. 16, 2006

Steric Effects on Coupling Rate Constants
FIGURE 5. (a) Log(kc/M-1 s-1) vs Es,n estimated by using Es1 for 1-3 and 8 nitroxides: (9, s) for MiB radical, (b, - - - ) for tBP radical. (b)
Log(kc/M-1 s-1) vs Linear combination of molecular descriptors (σL,n and Es,n) for 1-4, 8a-g, and 12 nitroxides and the MiB alkyl radical.
TABLE 2. Coefficients for Eq 2 Depending on Both Alkyl
SCHEME 5
Radicals MiB and TBP and the Steric Constants E1 and E3
s
s
alkyl
radicals eqs
a
log kc,0
δ
Nb
sc
R2d
te
tBP
6f
7g
8f
10.64 ( 0.29 0.41 ( 0.06
9.80 ( 0.17 0.20 ( 0.03
7
7
0.10 0.91 99.91
0.10 0.91 99.92
MiB
11.09 ( 0.17 0.49 (0.07 10 0.08 0.96 99.99
nitroxide moiety would hamper the approach of the alkyl radical
and thus destabilize the transition state, i.e., a positive δ
coefficient should be observed in eq 10.
9g 10.06 ( 0.11 0.24 ( 0.02 10 0.08 0.96 99.99
a Value for the y-intercept and corresponding to the nitroxides exhibiting
no steric effect (Es ) 0.0). b Number of data. c Standard deviation. d Square
of the regression coefficient. e Student t-test given in percent. f For Es,n given
by eqs 3 and 5. g For Es,n given by eqs 4 and 5.
log(kc/M-1 s-1) ) FσL,n + δEs,n + log(kc0/M-1 s-1)
(10)
The statistical outputs for biparametric regression are very
good (Student t-test above 99.99%), and the coefficients are
positive (eq 11) as expected for the coupling of nitroxides 1-4,
8a-g, and 12 and the MiB radical when E1s is used to estimate
Es,n.
nitroxide 8f and geminal methyl groups in nitroxide 1. The lack
of steric effect of the linear alkyl chain has already been
reported.29 The six-membered ring formed by junction between
positions 4 and 5 of the imidazolidin N-oxyl radical is assumed
to exhibit the same steric demand as the ethyl group. Then, the
steric effect on the coupling reaction of 1-4 and 8a-g with
tBP and MiB was correlated to kc via eqs 2 and 5, and good
correlations are observed for both possibilities E1 and E3 (Table
log(kc/M-1 s-1) ) 3.52((0.43)σL,n + 0.47((0.03)Es,n
10.62((0.12) (11)
+
s
s
1, Figure 5a, eqs 6-9 in Table 2). The statistical outputs are
good for both possibilities and confirm the monotonic increase
of kc with the steric demand around the nitroxyl moiety.
The steric constants E1s and E3s for 4 and 12 were estimated
by assuming that the steric effect of the â substituents (two
methyl groups in position 4 of the imidazolidine ring) is the
same as for the series 1-4 and 8a-g (one methyl group), and
assuming a value of -1.40 for the steric constant ri of the phenyl
group.26 It has been shown26 that a biparametric equationspolar
effect and stabilization of the nitroxide accounted for by the
electrical Hammett constant σL, and the steric effect of the
substituents around the nitroxide moiety accounted for by the
steric constant Essdescribed the effects influencing the C-ON
bond homolysis rate constant kd. Therefore, it was assumed that
the same effectsspolar and steric effects of the substituents
around the nitroxide moiety, and the stabilization of the
nitroxidesinfluence increases the coupling rate constants kc. The
electron-withdrawing effect of the substituents should increase
the spin density on the nitroxide oxygen atom and, thereby,
should enhance the reactivity of nitroxide in the coupling
reaction by increasing the contribution of resonance form A
(Scheme 5).
for R2 ) 0.98, N ) 12, s ) 0.07, and F11,99.99% ) 191.
The statistical outputs are as good when E3 is applied,
s
assuming r(C5) is smaller than r(Ph), which is highly improb-
able. Other possible combinations afforded poorer statistical
outputs.
Discussion
The values of cross-coupling rate constants of different
nitroxides kc summarized in Table 1 are markedly below the
diffusion-controlled limit of about 5 × 109 M-1 s-1, and depend
very slightly on temperature, showing very small, or even
negative, activation energy. The same dependence previously
has been observed for the series of stable nitroxides by several
authors.18,14 In general, radical coupling reactions exhibit very
low Ea values, i.e., below 10 kJ/mol. Therefore, the variation
of kc with T is weak (see Arrhenius equation). This indicates
the entropy control of the reaction, i.e., hampered approach of
the alkyl radical to the oxygen centered radical.14,18
In general, coupling rate constants decrease with increasing
steric congestion of the reactive center of the nitroxides. As
mentioned above, Fischer et al.17 observed a correlation between
Es,n and log kc for the cross-coupling of six-membered-ring
piperazinone N-oxyl type radical and the 1-phenylethyl alkyl
radical. With the series 1-3 and 8a-g, which exhibits the same
polar effect, the linear relationship between Es,n and log kc
Therefore, such an effect accounted for by the electrical
Hammett constant σL would increase kc, i.e., a positive F
coefficient should be observed in eq 10. It was expected that
the steric effect due to the substituents in position R,R′ of the
J. Org. Chem, Vol. 71, No. 16, 2006 6049

Zubenko et al.
on an FT-IR spectrometer in KBr pellets (the concentration was
0.25%; the pellet thickness was 1 mm). The UV spectra were
measured in EtOH. The H NMR spectra were recorded on a 300
(eq 7 in Table 1) for the coupling of MiB and tBP alkyl radicals
exemplifies the generality of this approach, that is, the kc values
decrease with increasing congestion around the nitroxyl moiety.
Interestingly, as the steric effect is mainly contained in the
activation entropy ∆Sq (vide supra), it means that ∆Sq increases
monotonically with the size of the substituents in the position
R,R′. It is worthy to mention, as the slopes of eqs 6-9 (Table
2, Figure 5a) are very close, that the values of kc depend only
slightly on the type of alkyl radicals, propanoate-2-yl (acrylate
type) and 2-methylpropanoate-2-yl (methacrylate type) alkyl
radicals.
Very good biparametric correlation with positive slope F
(Figure 5b, eq 11) was observed when molecules 4 and 12,
exhibiting different σL values,30,31 were included in the series,
meaning that increasing polar/destabilization effect increased
kc. Assuming negligible polar effect in the transition state,
electron-withdrawing groups favor the form A over the form B
in the initial statesthe form B is destabilized for electron-
withdrawing groups attached to the nitrogen atomsthus the
relocation of the odd-electron on the oxygen atom is easier,
and hence, the activation barrier (∆Hq or Ea) is lower. This
destabilization is denoted by changes in activation enthalpy ∆Hq,
which decreases with increasing σL values. Therefore, it is
remarkable that eq 11 takes into account the effects occurring
both in the transition state and in the initial state, and involving
simultaneous changes in ∆Sq and ∆Hq. The increasing kc with
increasing σL values means the coupling reaction is facilitated
by the presence of electron-withdrawing groups close to the
nitroxyl function. It is worthwhile to mention that the polar effect
is rather important (35%) to the steric effect (65%).32
1
MHz NMR spectrometer for 5% solutions, using the signal of the
solvent as the standard. The 13C NMR spectra were recorded on
75 and 100 MHz NMR spectrometers for 5-10% solutions at 300
K, using the signal of the solvent as the standard. The assignment
of the signals in the 13C NMR spectra was made based on analysis
of intensities, on the spectra measured in J-modulation mode, and
using the data reported previously.4 High-resolution mass spectra
were recorded and element analyses were performed.
The nitroxides 1-4 (see Table 1) were prepared according to
the methods described in refs 4 and 35-37, respectively. The
nitroxides 7f and 7g were prepared according to the methods
described in refs 38 and 39, respectively.
2,2-Diethyl-3-methyl-1,4-diazaspiro[4,5]dec-3-en-1-ol (6a). A
suspension of 3-hydroxyamino-3-ethylpentan-2-one hydrochloride
5a4 (5 g, 27.5 mmol) and ammonium acetate (7.7 g, 100 mmol) in
a mixture of cyclohexanone (10 mL, 93 mmol) and methanol (8
mL) was stirred under argon for 24 h. The resulting solution was
poured into water (20 mL), basified with NaHCO3, and extracted
with diethyl ether (3 × 10 mL). The ether layers were thoroughly
washed with water (5 × 10 mL) and dried over Na2CO3. The ether
was removed under reduced pressure, the residue was triturated
with hexane at -5 °C, and the crystalline precipitate of 6a was
filtered off and washed with hexane, yielding 3.8 g, 60%, of
colorless crystals, mp 129-133° (hexane) (Found: C, 69.54; H,
10.48; N, 12.44. Calcd for C13H24N2O: C, 69.60; H, 10.78; N,
12.49); νmax (KBr)/cm-1 2924, 2870, 1655, 1472, 1451, 1428, 1384,
1282, 1266, 1175, 1034, 962, 934, 900, and 851; δH (200 MHz;
CDCl3) 0.85 (6 H, br t, J ) 7 Hz, 2 × CH3, Et), 1.23-1.87 (14 H,
br m, 7CH2), 1.90 (3 H, s, CH3CdN), and 6.00 (1 H, br s, OH); δC
(50 MHz; CDCl3) 8.8 (CH3, Et), 27.1 (CH2, Et), 25.0, 22.8, and
35.6 ((CH2)5), 16.2 (CH3-CdN), 78.2 (C5), 91.4 (C2), and 171.2
(CdN).
Conclusions
2,5-Dihydroimidazol-1-ols 6b-e were prepared in a similar
manner from 3-hydroxyamino-3-methylbutane-2-one hydrochloride
5b,40 3-hydroxyamino-3-ethylpentan-2-one hydrochloride 5a, and
2-hydroxyamino-2-methylcyclohexanone hydrochloride 5c40 and the
corresponding ketones (acetone, diethyl ketone, cyclohexanone, and
dibutyl ketone) at 25 °C or under reflux (for 6c) for 3 h to 7 days,
until hydroxyaminoketone was completely reacted (the sample was
basified with 25% NH3 and analyzed with use of TLC Silufol;
eluent: diethyl ether). The reaction mixtures were cooled to 5 °C,
and the precipitates of 6b-e were filtered off and washed with
water.
2,2,4-Trimethyl-5,5-diethyl-2,5-dihydroimidazol-1-ol (6b). Yield
80%, colorless crystals, mp 114-118 (hexane) (Found: C, 65.04;
H, 10.80; N, 15.37. Calcd for C10H20N2O: C, 65.18; H, 10.94; N,
15.20); νmax (KBr)/cm-1 3172 (br), 2977, 2925, 2880, 1661, 1465
(br), 1434, 1373, 1263, 1232, 1200, 1036, and 877; δH (400 MHz;
CDCl3) 0.85 (6 H, t, J ) 7 Hz, 2 × CH3, Et), 1.37 (6H, s, 2Me),
1.58 and 1.70 (each 2 H, m, 2CH2), 1.86 (3 H, s, CH3CdN), and
6.65 (1 H, br s, OH); δC (100 MHz; CDCl3) 5.3 (CH3, Et), 12.8
(CH3-CdN), 22.5 (CH2, Et), 22.6 (CH3), 75.5 (C5), 86.5 (C2),
and 168.0 (CdN).
All the results concur with the exclusive coupling of acrylate-
and methacrylate-type radicals with the imidazolidine nitroxides.
Furthermore, we showed that the approach proposed by Fischer
et al.17 to account for the steric effect is versatile enough to be
applied to other alkyl radicals such as MiB and tBP radicals.
With the series 1-4, 8a-g, and 12, we showed that the polar
effect exerts significant influence on kc. Consequently, for
developing a new nitroxide efficient in NMP it might be
interesting to change the scavenging properties by varying the
polarity of the substituents and keeping the same steric
hindrance. The suitability of the synthesized series of nitroxides
for NMP is under investigation, i.e., measurement of kd and
polymerization experiments.
Experimental Section
Materials. Ketone concentrations were 20 mM in the case of
2,4-dimethyl-3-oxopentanedioic acid di-tert-butyl ester (tBEst) and
2,2,4,4-tetramethyl-3-oxopentanedioic acid dimethyl ester (Pest).
The concentrations of nitroxides were varied from 2 to 10 mM.
The ketones were synthesized according to refs 33 and 34 and
purified by column chromatography. The solvents (benzene and
acetonitrile) were distilled prior to use. The IR spectra were recorded
(35) Volodarski, L. B.; Reznikov, V. A.; Kobrin, V. S. J. Org. Chem.
USSR (Engl. Transl.) 1979, 13, 364.
(36) Reznikov, V. A.; Volodarsky, V. A. IzV. Sib. Otd. Akad. Nauk SSR,
Ser. Khim. Nauk 1984, 8, 89-97 (in Russian); Chem. Abstr. 1985, 102,
6297y.
(30) Charton, M. Prog. Phys. Org. Chem. 1981, 13, 119-251.
(31) Charton, M. In AdVances in QuantitatiVes Structure-Property
Relationship; Charton, M., Ed.; Jai Press Inc.: Greenwich, CT, 1996; Vol.
1, pp 171-219.
(37) Martin, V. V.; Volodarsky, V. A.; Vishnivetskaya, L. A. IzV. Sib.
Otd. Akad. Nauk SSSR, Ser. Khim. Nauk 1981, 4, 94-103 (in Russian);
Chem. Abstr. 1982, 96, 6648y.
(38) Sevastjanova, T. K.; Volodarsky, L. B. Bull. Acad. Sci. USSR DiV.
Chem. Sci. (Engl. Transl.) 1972, 21, 2276.
(39) Volodarsky, L. B.; Kutikova, G. A. Bull. Acad. Sci. USSR DiV.
Chem. Sci. (Engl. Transl.) 1971, 20, 859-863.
(40) Volodarsky, L. B.; Sevastjanova, T. K. Zh. Org.Khim. 1971, 8, 1687
(in Russian); Chem. Abstr. 1971, 75, 140425r.
(32) The weigth of each effect was estimated with equations given in:
Shorter, J. In Correlation Analysis of Organic ReactiVity; Research Studies
Press: Chichester, UK, 1982; pp 73-126.
(33) Favorski, A. E.; Umnova, A. J. Prakt. Chem. II 1912, 88, 679.
(34) Knu¨hl, B.; Marque, S.; Fischer, H. HelV. Chim. Acta 2001, 84, 2290.
6050 J. Org. Chem., Vol. 71, No. 16, 2006

Steric Effects on Coupling Rate Constants
Methyl-5,6,7,7a-tetrahydrospiro[benzimidazole-2,1′-cyclo-
hexan]-1(4H)-ol (6c). Yield 32%, colorless crystals, mp 175-177
°C (ethyl acetate-heptane 1:1) (Found: C, 68.41; H, 9.97; N, 11.99.
Calcd for C13H22N2O‚1/3H2O: C, 68.38; H, 10.01; N, 12.27); νmax
(KBr)/cm-1 3171 (br), 2937, 2859, 1655, 1368, 1285, 1206, 1081,
1029, 983, 957, and 911; δH (200 MHz; CDCl3) 1.45 (3 H, s, Me),
1.60-2.80 (18 H, br m, 9 × CH2), and 6.45 (1 H, br s, OH); δC
(50 MHz; CDCl3) 15.0 (CH3), 88.9 (C1′), 67.2 (C5), 37.3 (C6),
34.2 (C2′), 30.4 (C6′), 26.1 (C9), 22.9 (C4′), 21.8 (C8), 19.9 (C7),
19.2 (C3′), 18.6 (C5′), and 172.0 (CdN).
2,2-Dibutyl-4,5,5-trimethyl-2,5-dihydroimidazol-1-oxyl (7e).
Yield 90%, orange oil; νmax (CHCl3, 2%)/cm-1 2960, 2934, 2873,
1643, 1466, 1428, 1380, 1238, and 1159; m/z, found 239.21217,
calcf for C14H27N2O 239.21233.
2,2-Diethyl-3,4-dimethyl-1,4-diazaspiro[4,5]decan-1-oxyl (8a).
Dimethyl sulfate (0.5 g, 4 mmol) was added to a solution of 7a
(0.5 g, 2.4 mmol) in dry diethyl ether (3 mL). The solution was
allowed to stand for 0.5 h at 25 °C and filtered, then diethyl ether
was removed under reduced pressure. The residue was heated to
50 °C for 30 min under reduced pressure and then triturated with
dry diethyl ether to form a highly hygroscopic yellow crystalline
precipitate of quaternary salt. The precipitate was filtered off
(without drying), washed with dry diethyl ether, and immediately
dissolved in ethanol (5 mL). NaBH4 (150 mg, 4 mmol) was added
to the solution and the mixture was stirred for 1 h. Ethanol was
removed under reduced pressure, and the residue was dissolved in
water (2 mL) and extracted with diethyl ether (3 × 2 mL). The
extract was dried over Na2CO3, then diethyl ether was removed
under reduced pressure to leave a yellow oil, which was purified
by column chromatography on silica (Kieselgel 60 Merck, eluent
diethyl ether:hexane 1:20) to yield the nitroxide 8a (480 mg, 90%)
as a yellow oil (Found: C, 70.11; H, 11.68; N, 11.49. Calcd for
C13H27N2O: C, 70.24; H, 11.37; N, 11.70); νmax (neat)/cm-1 2972,
2934, 2859, 2792, 1448, 1420, 1382, 1351, 1297, 1264, 1233, 1216,
1201, 1139, 1079, 1064, 1036, 965, 922, 907, 868, 846, and 788.
The nitroxides 8b-g and 12 were prepared by the same
procedure.
5,5-Diethyl-2,2,3,4-tetramethylimidazolidin-1-oxyl (8b). Yield
90%, yellow oil (Found: C, 66.64, H, 11.63; N, 13.98. Calcd for
C11H23N2O: C, 66.29; H, 11.63; N, 14.05); νmax (neat)/cm-1 2978,
2938, 2881, 2848, 2796, 1461, 1383, 1368, 1355, 1271, 1239, 1170,
965, 922, and 887; m/z, found 199.18070, calcd for C11H23N2O
199.18103.
3,7a-Dimethylhexahydrospiro[benzimidazole-2,1′-cyclohexan]-
1(3H)-oxyl (8c). Yield 30%, orange oil (Found: C, 71.07; H, 10.54;
N, 11.63. Calcd for C14H25N2O: C, 70.84; H, 10.62; N, 11.80);
νmax (neat)/cm-1 2977, 2960, 2939, 2860, 1463, and 1445.
2,2-Diethyl-3,4,5,5-tetramethylimidazolidin-1-oxyl (8d). Yield
60%, orange oil (Found: C, 66.34; H, 11.38; N, 14.08. Calcd for
C11H23N2O: C, 66.29; H, 11.63; N, 14.05); νmax (neat)/cm-1 2978,
2935, 2879, 2851, 2795, 1452, 1376, 1361, 1337, 1226, 1203, and
969; m/z, found 199.18070, calcd for C11H23N2O 199.18103.
2,2-Dibutyl-3,4,5,5-tetramethylimidazolidin-1-oxyl (8e). Yield
70%, orange oil, νmax (neat)/cm-1 2956, 2971, 2870, 2791, 1473,
1451, 1375, 1361 and 1254; m/z, found: 255.24315, calc. for
C15H31N2O: 255.24362.
2,2,3,4-Tetramethyl-1,4-diazaspiro[4.4]nonan-1-oxyl (8f). Yield
48%, orange oil; νmax (neat)/cm-1 2956, 2871, 2794, 1468, 1453,
1378, 1360, 1336, 1263, 1207, 1191, 1144, 1067, and 1036; m/z,
found 197.16529, calcd for C11H21N2O 197.16538.
2,2,3,7a-Tetramethyloctahydro-1H-benzimidazol-1-oxyl (8g).
Yield 60%, orange oil (Found: C, 66.24; H, 10.42; N, 14.63. Calcd
for C11H21N2O: C, 66.29; H, 10.73; N, 14.20); νmax (neat)/cm-1
2976, 2935, 2859, 2796, 2738, 1463, 1445, 1369, 1274, 1242, 1208,
1174, and 993.
2,2-Diethyl-5,5-dimethyl-4-phenyl-2,5-dihydro-1H-imida-
zole 3-oxide (10). A solution of 2-amino-2-methyl-1-phenylpropan-
1-one oxime (9)41 (3.56.g, 20 mmol) and p-toluenesulfonic acid
(86 mg, 0.5 mmol) in pentane-3-one (5 mL, 60 mmol) was stirred
at 90 °C for 120 h. The mixture was separated by column
chromatography on silica (Kieselgel 60, Merck, chloroform as
eluent) to yield 1.4.g (28%) of 10, as colorless crystals, mp 77-79
(hexane) (Found: C, 73.03; H, 9.45; N, 11.43. Calcd for
C15H22N2O: C, 73.13; H, 9.00; N, 11.37); νmax (KBr)/cm-1 3257,
2977, 2961, 2934, 2876, 1587, 1569, 1497, 1462, 1358, 1266, 1222,
2,2-Diethyl-4,5,5-trimethyl-2,5-dihydro-1H-imidazol-1-ol (6d).
Reaction conditions: 40 °C for 3 h under argon. Yield 65%,
colorless crystals, mp 130-132 °C (hexane) (Found: C, 64.94; H,
10.87; N, 14.85. Calcd for C10H20N2O: C, 65.18; H, 10.94; N,
15.20); νmax (KBr)/cm-1 3166 (br), 2966, 2934, 2877, 1656, 1464,
1432, 1378, 1224, 1035, 957, and 931; δH (400 MHz; CDCl3) 0.87
(6 H, br t, J ) 7 Hz, 2 × CH3, Et), 1.23 (6H, s, 2Me,), 1.63 and
1.84 (each 2 H, br m, 2 × CH2, Et), 1.95 (3 H, s, CH3CdN), and
6.60 (1 H, br s, OH); δC (100 MHz; CDCl3) 6.8 (CH3, Et), 14.1
(CH3-CdN), 21.8 (CH3), 27.4 (CH2, Et), 70.2 (C5), 92.8 (C2),
and 173.0 (CdN).
2,2-Dibutyl-4,5,5-trimethyl-2,5-dihydro-1H-imidazol-1-ol (6e).
Yield 30%, colorless crystals, mp 115-117 °C (hexane) (Found:
C, 70.11; H, 11.83; N, 11.65. Calcd for C14H28N2O: C, 69.95; H,
11.74; N, 11.65); νmax (KBr)/cm-1 3169, 2960, 2932, 2871, 1655,
1467, 1226 1032, 1016, 982, and 956; δH (400 MHz; CDCl3) 0.87
(6 H, br t, J ) 7 Hz, 2 × CH3, n-Bu), 1.25 (6H, s, 2Me), 1.29 (m,
8H, 4 × CH2), 1.59 and 1.80 (each 2 H, br m, 2 × CH2, C2-
CH2), 1.96 (3 H, s, CH3CdN), and 6.50 (1 H, br s, OH); δC (100
MHz; CDCl3) 12.6 (CH3, Bu), 14.4 (CH3-CdN), 22.1 (CH3),
21.85, 24.9, and 35.5 (CH2, n-Bu), 70.5 (C5), 92.7 (C2), and 172.8
(CdN).
2,2-Diethyl-3-methyl-1,4-diazaspiro[4,5]dec-3-en-1-oxyl (7a).
Manganese dioxide (2 g, 23 mmol) was added to a stirred solution
of 6a (1 g, 4.5 mmol) in chloroform (20 mL). The suspension was
stirred for 0.5 h, manganese oxides were filtered off, and filtrate
was evaporated under reduced pressure to leave orange crystalline
solid, which was purified by column chromatography on silica
(Kieselgel 60, Merck; eluent: diethyl ether:hexane 1:20) to yield
nitroxide 7a (0.95 g, 95%), orange crystals, mp 86-88 (hexane)
(Found: C, 69.89, H, 10.32; N, 12.41. Calcd for C13H23N2O: C,
69.9; H, 10.4; N, 12.5); νmax (KBr)/cm-1 2972, 2961, 2935, 2853,
1637, 1452, 1423, 1386, 1376, 1356, 1325, 1294, 1263, 1210, 1172,
1142, 1109, 962, 933, 912, 845, and 813.
The nitroxides 7b-e were prepared with the same procedure.
2,2,4-Trimethyl-5,5-diethyl-2,5-dihydroimidazol-1-oxyl (7b).
Yield 90%, orange oil (Found: C, 65.92, H, 10.73; N, 14.93. Calcd
for C13H21N2: C, 65.54; H, 10.45; N, 15.29); νmax (KBr)/cm-1 2979,
2930, 2880, 1638, 1458, 1429, 1381, 1357, 1266, 1249, 1199, 1186,
953, 935, 869, and 843; m/z, found 183.14892, calcd for C10H19N2O
183.14973.
7a-Methyl-5,6,7,7a-tetrahydrospiro[benzimidazole-2,1′-cyclo-
hexan]-1(4H)-oxyl (7c). Yield 36%, orange crystals, mp 95-97
°C (hexane) (Found: C, 70.19, H, 9.72; N, 12.44. Calcd for
C13H21N2O: C, 70.55; H, 9.56; N, 12.66); νmax (KBr)/cm-1 2980,
2940, 2856, 1648, 1461, 1448, 1372, 1288, 1160, 962, and 907;
m/z, found 222.17289, calcd for C13H21N2O 222.17320.
2,2-Diethyl-4,5,5-trimethyl-2,5-dihydroimidazol-1-oxyl (7d).
Yield 90%, orange oil (Found: C, 65.86, H, 10.23; N, 15.12. Calcd
for C10H19N2O: C, 65.54; H, 10.45; N, 15.29); νmax (neat)/cm-1
2976, 2938, 2881, 1642, 1462, 1430, 1377, 1167, 953, and 928.
(41) Volodarskii, L. B.; Tormyshewa, N. Y. IzV. Sib. Otd. Akad. Nauk
SSSR. Ser. Khim. Nauk. 1976, 4, 136 (in Russian); Chem. Abstr. 1976, 85,
159592.
J. Org. Chem, Vol. 71, No. 16, 2006 6051

Zubenko et al.
1203, 1166, 1046, 1016, 807, and 759; λmax (EtOH)/nm 284 (log ꢀ
3.79); δH (400 MHz; CDCl3) 0.94 (6 H, t, J ) 7 Hz, 2CH3, Et),
1.55 (6 H, s, 2CH3), 1.84 and 1.97 (both 2 H, m, 2 × CH2), 7.0 (3
H, m, Ph-m,p), and 8.05 (2 H, m, Ph-o); δC (75 MHz; CDCl3) 7.6
(CH3, Et), 29.8 (5-CH3), 30.0 (CH2, Et), 62.0 (C5), 93.5 (C2), 142.5
(C4), Ph: 128.0 (C-i), 128.3 (C-o), 127.6 (C-m), and 129.5 (C-
p).
2977, 2941, 2879, 2812, 1599, 1491, 1466, 1444, 1388, 1330, 1262,
1231, 1144, 984, 924, 764, 747, and 705.
Time Resolved-CIDNP. A sample, purged with argon and sealed
in a standard NMR Pyrex ampule, was irradiated by a excimer laser
(wavelength 308 nm, pulse energy up to 150 mJ) in the probe of
a 200 MHz NMR spectrometer. TR-CIDNP experiments were
carried out with the usual pulse sequence: presaturation-laser
pulse-evolution time-detection pulse-free induction decay. Be-
cause the background signals in CIDNP spectra are suppressed by
the presaturation pulses, only signals of the polarized products
formed during the variable delay between the laser and NMR radio
frequency (rf) pulse appear in the CIDNP spectra. The rf pulse
used has a duration of 3 µs, which corresponds to a flip angle of
30°.
Laser Flash Photolysis. A detailed description of our LFP
experiments has been published.42 Solutions in a rectangular cell
with inner dimensions 10 mm × 10 mm were irradiated with a
excimer 308 nm laser with pulse energy up to 100 mJ and pulse
duration of 15-20 ns. The monitoring system includes a xenon
short-arc lamp connected to a high current pulser, a homemade
monochromator, a photomultiplier, and a digitizer. The monitoring
light, concentrated in a rectangular of 3 mm height and 1 mm width,
passed through the cell along the front of the laser irradiated
window. Thus, in all experiments the excitation optical length was
1 mm, and the monitoring optical length was 8 mm. To obtain one
kinetic trace, 15-20 signals were averaged. All solutions were
bubbled with argon for 15 min prior to and all the time during the
experiments.
2,2-Diethyl-1,5,5-trimethyl-4-phenyl-2,5-dihydro-1H-imida-
zole 3-Oxide (11). Formic acid, 90% (1.5 mL, 35 mmol), was added
portionwise to a solution of 10 (1.3 g, 5.2 mmol) in an aqueous
solution of formaldehyde 30% (3 mL, 32 mmol). The solution was
allowed to stand overnight, saturated with NaCl, and extracted with
chloroform (3 × 50 mL). The organic layer was washed with water
and dried over MgSO4. The solvent was removed in a vacuum and
the residue was purified by column chromatography on silica
(Kieselgel 60, Merck, chloroform as eluent) to yield 1 g (74%) of
11, as a colorless oil (Found: C, 73.55; H, 8.83; N, 10.88. Calcd
for C16H24N2O: C, 73.81; H, 9.29; N, 10.76); νmax (KBr)/cm-1 3057,
2974, 2936, 2878, 2809, 1681, 1576, 1540, 1495, 1460, 1444, 1362,
1335, 1264, 1247, 1170, 1142, 1101, 1040, 1022, 996, 948, 843,
814, and 764; λmax (EtOH)/nm 279 (log ꢀ 3.70); δH (400 MHz;
CDCl3) 0.85 (6 H, t, J ) 7 Hz, 2 × CH3, Et), 1.44 (6 H, s, 2CH3),
1.63 and 2.05 (both 2 H, dq, J1 ) 14 Hz, J2 ) 7 Hz, 2CH2), 2.41
(3 H, s, N-CH3), 7.40 (3 H, m, Ph-m,p), and 7.97 (2 H, m, Ph-o);
δC (75 MHz; CDCl3) 8.4 (CH3, Et), 25.8 (CH3), 26.0 (N-CH3),
28.5 (CH2), 64.5 (C5), 95.00 (C2), 144.5 (C4), Ph: 128.4 (C-i),
128.6 (C-o), 128.1 (C-m), and 129.5 (C-p).
2,2-Diethyl-5,5-diphenyl-3,4,4-trimethylimidazolidin-1-oxyl (12).
A solution of 11 (1 g, 3.8 mmol) in dry THF (20 mL) was added
dropwise to a 1 M solution of PhLi in diethyl ether (10 mL). The
reaction mixture was stirred under argon for 1 h, and then quenched
with a saturated solution of NH4Cl in water. The organic layer was
separated and water solution was extracted with ether (2 × 20 mL).
The combined organic phase was stirred with MgSO4 and MnO2
(2 g, 23 mmol) for 1 h. The resulting solution was filtered, the
solvent was removed in a vacuum, and the residue was separated
with column chromatography on Al2O3 to yield 12, 0.77 g (60%),
mp 91-93 °C (hexane) (Found: C, 78.18; H, 8.74; N, 8.23. Calcd
for C22H29N2O: C, 78.30; H, 8.66; N, 8.30); νmax (KBr)/cm-1 3050,
Acknowledgment. This work was supported by the Uni-
versity of Provence, CNRS, RFBR (grant Nos. 05-03-32370
and 04-03-32299), CRDF RUC1-2635-NO-05. D.Z. is very
grateful to grant INTAS YS N 05-109-5102 and the financial
support of the Zamaraev Scientific Foundation. We are thankful
to Prof. M. Forbes for useful discussion.
JO060787X
(42) Molokov, I. F.; Tsentalovich, Y. P.; Yurkovskaya, A. V.; Sagdeev,
R. Z. J. Photochem. Photobiol. A: Chem. 1997, 110, 159.
6052 J. Org. Chem., Vol. 71, No. 16, 2006