Generation and Mesolysis of PhSeSiR3.-: Mechanistic Studies by Laser Flash Photolysis and Application for Bimolecular Group Transfer Radical Reactions

Article abstract of DOI:10.1021/jo972345e

The investigation presented in this paper explores the mechanistic aspects and synthetic potentials of PET promoted reductive activation of selenosilane 1a to its radical anion 1a^.-. PET activation of 1a is achieved through a photosystem comprising a light-absorbing electron-rich aromatic (ERA), such as DMN or DMA, as an electron donor and ascorbic acid as a co-oxidant. The evidence for the ET from excited singlet states of DMN as well as DMA to 1a is suggested by estimating negative ΔG_et (-51 and -43.46 kcal mol^-1, respectively) values and nearly diffusion- controlled fluorescence quenching rate constants (k_qTR) 0.36 × 10¹⁰ M^-1 s^-1 and 0.28 × 10¹⁰ M^-1 s^-1, respectively, from time-resolved fluorescence quenching study. The transient absorption spectra of DMN^.+, DMA^.+, and 1a^.- are obtained initially by pulse radiolysis in order to correlate the time- resolved absorption spectral data. Laser flash photolysis studies in the nanosecond time domain have confirmed the generation of 1a^.-, DMN^.+, and DMA^.+, supporting the participation of the triplet state of DMN or DMA in the ET reaction. Mesolytic cleavage of 1a^.- produced a silyl radical and a phenyl selenide anion. The preparative PET activation of 1a in acetonitrile in the presence of DMN or DMA leads to the formation of 5 and 6, confirming the fragmentation pattern of la^.-. The overall ET rate constants (k_r(DMN) = 0.99 × 10¹⁰ M^-1 s^-1 and k_r(DMA) = 1.62 × 10¹⁰ M^-1 s^-1) and limiting quantum yields (φ_lim(DMN) = 0.034 and φ_lim(DMA) = 0.12) are estimated from the inverse plot (1/[1a] vs 1/φ_dis) obtained by measuring the dependence of photodissociation quantum yields of 1a at its maximum concentration in the presence of DMN or DMA. Silicon-centered radical species generated from the mesolysis of 1a^.- are utilized for initiating a radical reaction by the abstraction of halogen atom from -C-X (X = Cl, Br) bonds, while PhSe^-1 terminates the radical sequences via PhSeSePh. This concept is successfully applied for the bimolecular group transfer (BMGT) radical reactions and intermolecular radical chain addition reactions.

Full text of DOI:10.1021/jo972345e

3968
J . Org. Chem. 1998, 63, 3968-3978
Gen er a tion a n d Mesolysis of P h SeSiR3]•-: Mech a n istic Stu d ies by
La ser F la sh P h otolysis a n d Ap p lica tion for Bim olecu la r Gr ou p
Tr a n sfer Ra d ica l Rea ction s†
Ganesh Pandey,*,‡ K. S. Sesha Poleswara Rao,‡ D. K. Palit,§ and J . P. Mittal*,§,
Division of Organic Chemistry (Synthesis), National Chemical Laboratory, Pune-411 008, India, and
Chemistry Division, Bhabha Atomic Research Center, Mumbai-400 085, India
Received December 31, 1997
Abstract: The investigation presented in this paper explores the mechanistic aspects and synthetic
potentials of PET promoted reductive activation of selenosilane 1a to its radical anion 1a •-. PET
activation of 1a is achieved through a photosystem comprising a light-absorbing electron-rich
aromatic (ERA), such as DMN or DMA, as an electron donor and ascorbic acid as a co-oxidant.
The evidence for the ET from excited singlet states of DMN as well as DMA to 1a is suggested by
estimating negative ∆Get (-51 and -43.46 kcal mol-1, respectively) values and nearly diffusion-
controlled fluorescence quenching rate constants (kqTR) 0.36 × 1010 M-1 s-1 and 0.28 × 1010 M-1
s-1, respectively, from time-resolved fluorescence quenching study. The transient absorption spectra
of DMN•+, DMA•+, and 1a •- are obtained initially by pulse radiolysis in order to correlate the time-
resolved absorption spectral data. Laser flash photolysis studies in the nanosecond time domain
have confirmed the generation of 1a •-, DMN•+, and DMA•+, supporting the participation of the
triplet state of DMN or DMA in the ET reaction. Mesolytic cleavage of 1a •- produced a silyl radical
and a phenyl selenide anion. The preparative PET activation of 1a in acetonitrile in the presence
of DMN or DMA leads to the formation of 5 and 6, confirming the fragmentation pattern of 1a •-
.
The overall ET rate constants (kr(DMN) ) 0.99 × 1010 M-1 s-1 and kr(DMA) ) 1.62 × 1010 M-1 s-1
)
and limiting quantum yields (φlim(DMN) ) 0.034 and φlim(DMA) ) 0.12) are estimated from the
inverse plot (1/[1a ] vs 1/φdis) obtained by measuring the dependence of photodissociation quantum
yields of 1a at its maximum concentration in the presence of DMN or DMA. Silicon-centered radical
species generated from the mesolysis of 1a •- are utilized for initiating a radical reaction by the
abstraction of halogen atom from -C-X (X ) Cl, Br) bonds, while PhSe- terminates the radical
sequences via PhSeSePh. This concept is successfully applied for the bimolecular group transfer
(BMGT) radical reactions and intermolecular radical chain addition reactions.
In tr od u ction
exchange. The exchange of electron follows the free-
energy gap law,6 and the generation of free-radical ion
pairs (FRIP) occurs via geminate radical ion pairs.7 The
mesolysis of these high-energy odd-electron species have
afforded the means to apply PET reactions to drive
energetically uphill processes in chemical synthesis8-10
and energy conservation.7a,b,11
Interesting and synthetically useful chemistry has
developed from the mesolysis of PET-generated radical
cations.8-10 Fragmentation of radical cations is generally
associated with the deprotonation,12-14 carbon-car-
bon,15,16 and carbon-heteroatom17,18 bond scission, rear-
rangements,19 and addition to unsaturated bonds20,21 in
the primary event of PET processes. Mesolysis of radical
cations from group IVA organometallics (-CMR3; M )
Pb, Sn, Si, Ge) is generally accompanied by the loss of
Radical ions have emerged as critical intermediates in
the development of modern organic reactivity concept.1,2
Fragmentation of radical ions (mesolysis3) plays a major
role in governing the chemoselectivity and efficiency of
a wide variety of redox processes. Therefore, knowledge
regarding the dynamics of mesolytic reactions is crucial
to the design of an efficient and useful organic synthetic
reaction. Generation of radical ions by photosensitized
electron transfer (PET) processes has acquired promi-
nence in the past decade,4,5 as photoexcitation easily
renders well-defined redox potential differences between
interacting substratessa prerequisite for an electron to
* To whom the correspondence should be addressed. Fax: 91-212-
335153. E-mail: pandey@ems.ncl.res.in.
† Dedicated with respect to Dr. S. Rajappa on the occasion of his
65th birthday.
(5) Mattes, S. L.; Farid, S. In Organic Photochemistry; Padwa, A.,
Ed.; Marcel Dekker: New York, 1983; Vol. 6, p 233. (b) Davidson, R.
S. Adv. Phys. Org. Chem. 1983, 19, 1. (c) Kavarnos, G. J .; Turro, N. J .
Chem. Rev. 1986, 86, 401. (d) J ulliard, M.; Chanon, M. Ibid. 1983,
83, 425.
(6) Weller, A. Z. Phys. Chem. (Wiesbaden) 1983, 133, 93. (b) Weller,
A. Ibid. 1982, 130, 129. (c) Rhem, D.; Weller, A. Isr. J . Chem. 1970,
8, 259.
‡ National Chemical Laboratory.
§ Bhabha Atomic Research Center.
Also affiliated with J awaharlal Nehru Center for Advanced
Scientific Research, Banglore, as Honorary Professor.
(1) Kochi, J . K. Angew. Chem., Int. Ed. Engl. 1988, 21, 1227.
(2) Pross, A. Adv. Phys. Org. Chem. 1985, 21, 99. (b) Pross. A.;
Shaik, S. S. Acc. Chem. Res. 1983, 16, 363.
(3) (a) Maslak, P.; Chapman, W. H.; Vallombroso, T. M., J r.; Watson,
B. A. J . Am. Chem. Soc. 1995, 117, 12380. (b) Maslak, P.; Narvaez, J .
N.; Vallombroso, T. M., J r. ibid. 1995, 117, 12373.
(4) Fox, M. A.; Chanon, M. D. Photoinduced Electron Transfer;
Elsevier: Amsterdam, 1988; Parts A-D. (b) Fox, M. A. Adv. Photo-
chem. 1986, 13, 295.
(7) (a) Gould, I. R.; Moser, J . E.; Armitage, B.; Farid, S. Res. Chem.
Intermed. 1995, 21, 793. (b) Mattes, S. L.; Farid, S. Science 1984, 226,
917. (c) Gould, I. R.; Ege, D.; Moser, J . E.; Farid, S. J . Am. Chem.
Soc. 1990, 112, 4290. (d) Mataga, N.; Okada, T.; Kanda, Y. Shioyama,
H. Tetrahedron 1986, 42, 6143. (e) Masuhara, H.; Mataga, N. Acc.
Chem. Res. 1981, 14, 312.
S0022-3263(97)02345-1 CCC: $15.00 © 1998 American Chemical Society
Published on Web 05/15/1998

Bimolecular Group Transfer Radical Reactions
J . Org. Chem., Vol. 63, No. 12, 1998 3969
metal cations (MR3+) generating carbon-centered radi-
Sch em e 1
•+
cals,22 with the exception of -C-Se- •+ and -C-S-
,
(1)
which produced carbocation.23 Mesolyses of Me3Sn-
•+
•+
MMe3
(M ) Sn, Ge, Si) and Me3Si-SiR3
are also
reported.24-26 However, to the best of our knowledge, no
report is available on the generation and fragmentation
of radical anions from M-M bonds. It may be mentioned
that the chemistry from the mesolysis of radical anions
in general has remained unexplored, despite its great
mechanistic and synthetic significance in organic
chemistry.27-29
(2)
Sch em e 2
We have recently studied the PET generation and
•-
mesolysis of -C-Se-
and reported30 that the frag-
mentation of such species leads to the generation of
(8) (a) Mattay, J . Synthesis 1989, 233. (b) Mattay, J . Angew. Chem.,
Int. Ed. Engl. 1987, 26, 825.
carbon-centered radicals (Scheme 1, eq 1). Since carbon
and silicon are isoelectronic and often predictions on the
physical and chemical properties of silicon compounds are
made by the direct analogy with what occurs in carbon
chemistry, though the similarity between carbon and
silicon to a great extent is only formal,31-33 we became
(9) Mariano, P. S.; Stavinoha, J . L. In Synthetic Organic Photo-
chemistry; Horspool, W. M., Ed.; Plenum Press: London, 1983; p 145.
(10) (a) Pandey, G. In Photoinduced Redox Reactions in Organic
Synthesis in Molecular and Supramolecular Chemistry; Ramamurthy,
V., Ed.; Marcel Dekker, Inc.: New York, 1997; Vol. 2, Chapter 7, p
245. (b) Pandey, G. Top. Curr. Chem. 1993, 168, 175. (c) Pandey, G.
Synlett 1992, 546.
(11) (a) Willner, I.; Willner, B. Top. Curr. Chem. 1981, 159, 157 and
references therein. (b) Webber, S. E. Chem. Rev. 1990, 90, 1469. (c)
Gust, D.; Moore, T. A. Top. Curr. Chem. 1990, 159, 103. (d) Gust, D.;
Moore, T. A.; Moore, A. L. Acc. Chem. Res. 1993, 26, 198.
(12) (a) Baciocchi, E.; Del Giacco, T.; Elisei, F. J . Am. Chem. Soc.
1993, 115, 5, 12290. (b) Parker, V. D.; Tilset, M. J . Am. Chem. Soc.
1991, 113, 8778. (c) Parker, V. D.; Chao, Y.; Reitstoen, B. Ibid. 1991,
113, 2336. (d) Reitstoen, B.; Parker, V. D. Ibid. 1990, 112, 4968.
(13) Schlesener, C. J .; Amatore, C.; Kochi, J . K. J . Am. Chem. Soc.
1984, 106, 6, 7472.
(14) Dinnocenzo, J . P.; Banach, T. E. J . Am. Chem. Soc. 1989, 111,
8646.
(15) (a) Maslak, P.; Narvaez, J . N.; Vallombroso, T. M., J r. Angew.
Chem., Int. Ed. Engl. 1994, 33, 73. (b) Maslak, P. Top. Curr. Chem.
1993, 168, 1.
(16) (a) Sankararaman, S.; Perrier, S.; Kochi, J . K. J . Chem. Soc.,
Perkin Trans. 2 1993, 825. (b) Sankararaman, S.; Perrier, S.; Kochi,
J . K. J . Am. Chem. Soc. 1989, 111, 6448.
•-
interested in exploring the chemistry of -Si-Se-
species (Scheme 1, eq 2). We disclose herein the full
details34 of the PET reductive activation of selenosilanes
(1), delineation of the mechanistic features of primary
electron transfer (ET) processes by laser flash photolysis
experiments, mesolytic pattern of -Si-Se- •-, and the
application of this chemistry for bimolecular group
transfer (BMGT)35 radical reactions and intermolecular
radical chain transfer addition reactions. To the best of
our knowledge, this study is the first of its own kind in
the area of PET reductive activations and mesolysis of
M-M bond compounds.
(17) (a) Zhang, X.; Yeh, S. R.; Hong, S.; Freccero, M.; Albini, A.;
Mariano, P. S. J . Am. Chem. Soc. 1994, 116, 4211. (b) Yoon, U. C.;
Mariano, P. S. Acc. Chem. Res. 1992, 25, 233.
(18) J ohnston, L. J .; Kwong, P.; Shelemay, A.; Lee-Ruff, E. J . Am.
Chem. Soc. 1993, 115, 1664.
(19) (a) Weng, X.; Sheik, Q.; Roth, H. D. J . Am. Chem. Soc. 1995,
117, 10655. (b) Weng, X.; Du, X.-M.; Roth, H. D. Ibid. 1995, 117, 135.
(c) Ishiguro, K.; Turro, N. J .; Roth, H. D. Ibid. 1994, 116, 6933. (d)
Roth, H. D.; Herbertz, T. Ibid. 1993, 115, 9801.
(20) Horner, J . H.; Martinez, F. N.; Musa, O. M.; Newcomb, M.;
Shain, H. E. J . Am. Chem. Soc. 1995, 117, 11124.
(21) Mirafzal, G. A.; Kim, T.; Liu, J .; Bauld, N. L. J . Am. Chem.
Soc. 1992, 114, 10968.
(22) (a) Kochi, J . K. Organometallic Mechanisms and Catalysis;
Academic Press: New York, 1978; pp 445-499. (b) Symons, M. C. R.
Chem. Soc. Rev. 1984, 13, 393. (c) Eaton, D. F. J . Am. Chem. Soc.
1981, 103, 7235. (d) Pandey, G.; Kumaraswamy, G.; Bhalerao, U. T.
Tetrahedron Lett. 1989, 30, 6059.
(23) (a) Pandey. G.; Soma Sekhar, B. B. V. J . Org. Chem. 1994, 59,
7367. (b) Pandey, G.; Soma Sekhar, B. B. V.; Bhalerao, U. T. J . Am.
Chem. Soc. 1990, 112, 5650. (c) Platen, M.; Steckhan, E. Liebigs. Ann.
Chem. 1984, 1563.
(24) (a) Fukuzumi, S.; Kitano, T. J . Am. Chem. Soc. 1990, 112, 3246.
(b) Fukuzumi, S.; Kitano, T.; Mochida, K. J . Chem. Soc., Chem.
Commun. 1990, 1236.
Resu lts a n d Discu ssion
For studying the mesolysis of -Si-Se- •-, selenosi-
lanes of type 1 were selected. These compounds (1a -d )
were prepared36 in 70-77% yield by the nucleophilic
displacement of chloride ion of the corresponding R3SiCl
by in situ generated phenylselenide anion (Scheme 2).
The rate of chloride ion displacement by phenyl se-
lenide anion decreases as the steric bulk around silicon
increases.33,36 Among the compounds 1a -d , tert-butyl
diphenyl(phenylseleno)silane (1a ) was found to be the
most stable and could be easily handled in aqueous
solvents. Selenosilanes 1b-d were highly susceptible to
air and hydrolyzed to give the corresponding disiloxane
(2) and phenylselenol (3) as shown in Scheme 3. There-
(30) (a) Pandey, G.; Sesha Poleswara Rao, K. S.; Nageshwar Rao,
K. V. J . Org. Chem. 1996, 61, 6799. (b) Pandey, G.; Sesha Poleswara
Rao, K. S.; Soma Sekhar, B. B. V. J . Chem. Soc., Chem. Commun. 1993,
1636.
(25) (a) Mochida, K.; Itani, A.; Yokayama, M.; Tsuchiya, T.; Worley,
S. D.; Kochi, J . K. Bull. Chem. Soc. J pn. 1985, 58, 2149. (b) Mochida,
K.; Worley, S. D.; Kochi, J . K. Ibid. 1985, 58, 3389. (c) Walther, B.
W.; Williams, F.; Lau, W.; Kochi, J . K. Organometallics 1983, 2, 688.
(26) Wang, J . T.; Williams, F. J . Chem. Soc., Chem. Commun. 1981,
666.
(27) (a) Guthrie, R. D.; Patwardhan, M.; Chateauneuf, J . E. J . Phys.
Org. Chem. 1994, 7, 147. (b) Wu, F.; Guarr, T. F.; Guthrie, R. D. Ibid.
1992, 5, 7.
(28) (a) Masnovi, J . J . Am. Chem. Soc. 1989, 111, 9081. (b) Saeva,
F. D. Tetrahedron 1986, 42, 6132. (c) Koppang, M.; Woolsey, N. F.;
Bartak, D. E. J . Am. Chem. Soc. 1984, 106, 2799.
(29) Andrieux, C. P.; Gorande, A. L.; Saveant, J .-M. J . Am. Chem.
Soc. 1992, 114, 6892. (b) Mautner, M.; M.-N.; Neta, P.; Norris, R. K.;
Wilson, K. J . Phys. Org. Chem. 1986, 90, 168. (c) Ruhl, J . C.; Evans,
D. H.; Hapiot, P.; Neta, P. J . Am. Chem. Soc. 1991, 113, 5188.
(31) (a) West, R.; Carberry, E. Science 1975, 189, 179. (b) West, R.;
Carberry, E. Pure Appl. Chem. 1982, 54, 1041.
(32) Tandura, N.; Alekseev, N. V.; Voronkov, M. G. Top. Curr. Chem.
1986, 131, 101 and references therein.
(33) MacDiarmid, A. G. Advances in Inorganic Chemistry and
Radiochemistry; 1961; Vol. 3, p 207 and references therein.
(34) Pandey, G.; Sesha Poleswara Rao, K. S. Angew. Chem., Int. Ed.
Engl. 1995, 34, 2669.
(35) Analogous to BMGT, the term UMGT (unimolecular group
transfer) has been introduced recently by us.30a Curran et al. have
also used the term UMCT (unimolecular chain transfer) for radical
chain reactions. (a) Curran, D. P.; J inyou, Xu. J . Chem. Soc., Perkin
Trans. 1 1995, 24, 3061. (b) Curran, D. P.; J inyou, Xu.; Lazzarini, E.
Ibid. 1995, 24, 3049. (c) Curran, D. P.; J inyou, Xu.; Lazzarini, E. J .
Am. Chem. Soc. 1995, 117, 6603.
(36) Detty, M. R. J . Org. Chem. 1981, 46, 1283.

3970 J . Org. Chem., Vol. 63, No. 12, 1998
Pandey et al.
to be -3.82 kcal mol-1. The free energy change of ET
from triplet states of ERA [∆Get (tDMN) ) -29.90 kcal
mol-1; ∆Get (tDMA) ) -11.50 kcal mol-1] were also
estimated by taking the reported39 values of triplet
excitation energy and were found to be in the exoergic
region.
(ii) St ea d y-St a t e a n d Tim e-R esolved F lu or es-
cen ce Qu en ch in g Stu d ies. Preliminary information
about the interaction of the excited singlet (S1) state of
DMN/DMA with 1a was obtained from both steady-state
as well as time-resolved fluorescence-quenching studies.
Fluorescence lifetimes (τ) of DMN and DMA in deareated
acetonitrile solution were determined to be 16.67 and
12.27 ns, respectively.40 The fluorescence lifetimes of
both DMN and DMA were seen to be heavily quenched
by 1a in acetonitrile solution. The Stern-Volmer (S-
V) plots obtained from lifetime quenching studies were
perfectly linear (Figure 3) throughout the range of
concentrations of 1a , and the quenching rate constants
were determined to be kqTR (DMN) ) 0.36 × 1010 M-1
s-1 and kqTR (DMA) ) 0.28 × 1010 M-1 s-1, respectively.
Steady-state fluorescence quenching of DMN (λexi ) 300
nm, λemi ) 344 nm) or DMA (λex ) 410 nm; λem ) 440
nm) by 1a also obeyed S-V relation. From the slopes
(kqf × τ) of the straight lines (Figure 4) and singlet
lifetimes of DMN and DMA, respectively, the quenching
rate constants [kqfSS (DMN) ) 0.822 × 1010 M-1 s-1, kqfSS
(DMA) ) 1.46 × 1010 M-1 s-1] were calculated and found
to be near diffusion5a (kdiff ) 2.30 × 1010 M-1 s-1).
Emission and excitation spectra of DMN and DMA
remained unaffected in the presence of the maximum
concentration of 1a in acetonitrile. No exciplex emission
was noticed in either polar (CH3CN) or nonpolar (cyclo-
hexane) solvents. However, the absorption values at the
excitation wavelength were found to increase slightly
with the increase in the concentration of the quencher
(1a ). The fluorescence quenching rate constants deter-
mined from the time-resolved studies are found to be
lower than those obtained from steady state quenching
analysis. This discrepancy in the rates could possibly be
due to the static quenching by the absorption of excitation
light by the quencher. Since in both time resolved
(Figure 3) and steady state (Figure 4) fluorescence
quenching studies the S-V plots obtained are perfectly
linear, the possibility of exciplex formation or ground-
state complexation can be excluded. Quenching due to
heavy-atom-induced intersystem crossing could also be
suggested to be minimal on the basis of similar logic as
provided by Eaton et al.22c It is reasonable, therefore to
assume that fluorescence quenching in these cases takes
place due to the interaction of the excited singlet states
of DMN and DMA with 1a via a single electron-transfer
mechanism.
F igu r e 1. Photosystem for PET reductive activation.
Sch em e 3
fore, we elected to use 1a as a representative selenosilane
to study the PET reductive activation of -Si-Se- bonds.
(A) P ET Red u ctive Activa tion of 1a :
P h otop h ysica l Stu d ies
To generate radical anions from -Si-Se- bond com-
pounds (selenosilane 1a ), a photosystem comprising
electron-rich aromatics (ERA) such as 1,5-dimethox-
ynaphthalene (DMN) or 9,10-dimethoxyanthracene (DMA)
as the light-harvesting electron donor and ascorbic acid
as the sacrificial electron donor was utilized as shown in
Figure 1.
(i) Th er m od yn a m ic P a r a m eter s. The thermody-
namic feasibility of ET between ERA and 1a was first
deduced by estimating the Gibbs free energy change
(∆Get) for radical ion formation using the Weller equation6c
(eq 3), where E1/2ox.[D] is the oxidation potential of ERA,
∆Get ) E1/2ox.[D] - E1/2red.[A] - Eexc[D]
(3)
E1/2red.[A] is the reduction potential of 1a , and Eexc[D] is
the excitation energy of ERA. Substituting eq 3 with the
red.
appropriate values of E1/2
of 1a (0.344 eV/SCE),
obtained by cyclic voltammetry (for details see the
Experimental Section), E1/2 of DMN (1.28 eV),37 and
ox.
excitation energy of DMN (87.8 kcal mol-1),37 an exoergic
value of -51 kcal mol-1 was obtained. Similarly, ET
feasibility from ascorbic acid to DMN•+ was evaluated by
estimating ∆Get (-4.51 kcal mol-1) employing the equa-
ox.
ox.
tion ∆Get ) E1/2 - E1/2red.. The E1/2 value (1.084 eV/
SCE) for ascorbic acid was estimated by cyclic voltam-
metry. The electron-donating ability of ascorbate ion and
its transformation to dehydroascorbic acid and proton is
precedented in the literature.38 To achieve the PET
reduction of 1a by visible light irradiation, DMA (λmax
380 and 425 nm, Figure 2C) was also evaluated as a
potential electron donor. Accordingly, the ∆Get value
(-43.46 kcal mol-1) for the ET between DMA and 1a was
estimated utilizing oxidation potential and singlet excita-
(iii) La ser F la sh P h otolysis Stu d ies: Eva lu a tion
of th e Tr a n sien t Sp ecies. To gain direct evidence of
primary ET processes from the excited states of DMN or
DMA to 1a and also to observe the transient spectra of
the radical ions thus formed, a laser flash photolysis
study was undertaken. Since the absorption properties
of the DMN•+, DMA•+, and 1a •- were not known, pulse
tion energy values of DMA as 0.98 eV and 74 kcal mol-1
,
respectively.39 In a similar manner, the ∆Get value for
the ET from ascorbic acid to DMA•+ was also estimated
(37) Hamada, T.; Nishida, A.; Yonemitsu, O. J . Am. Chem. Soc.
1986, 108, 140.
(38) (a) Brown, G. M.; Brunschwig, B. S.; Creutz, C.; Endicott, J .
F.; Sutin, N. J . Am. Chem. Soc. 1979, 101, 1298. (b) Bielski, B. H. J .;
Comstock, D. A.; Bowen, R. A. Ibid. 1971, 93, 5624.
(39) (a) Zweig, A.; Maurer, A. H.; Roberts, B. G. J . Org. Chem. 1967,
32, 2, 1322. (b) Yeung, C. K.; J asse, B. Makromol. Chem. 1984, 185,
541. (c) Lee, G. A.; Israel, S. H. J . Org. Chem. 1983, 48, 4557.
(40) The lifetime value (τDMA ) obtained for DMA is in agreement
with the value reported in the literature,39 but the lifetime value (τDMN
)
for DMN is higher than the literature value (12.6 ns) as reported by
Hamada et al.37 Possibly, the lifetime value reported37 for DMN is
lower due to the presence of a trace amount of oxygen in the solution.

Bimolecular Group Transfer Radical Reactions
J . Org. Chem., Vol. 63, No. 12, 1998 3971
F igu r e 2. UV-absorption spectra of 1a (0.003 M, A), DMN (0.0025 M, B), and DMA (0.0018 M, C).
F igu r e 4. Stern-Volmer plots for steady-state fluorescence
quenching for DMN (A, 9) and DMA (B, 4) by 1a in CH3CN
solution.
Sch em e 4
F igu r e 3. Stern-Volmer plots for time resolved fluorescence
quenching for DMN (A) and DMA (B) by 1a in CH3CN solution.
DMA•+, thus obtained are given in Figure 5A. DMA•+
has an absorption peak at 400 nm and shoulders at 450,
530, and 580 nm, whereas the absorption of DMN•+
extends from 400 to 700 nm without having any well-
defined peak.
radiolysis experiments were carried out at first
individually.
DMN•+ or DMA•+ was generated indirectly by ET from
DMN or DMA to the electron-deficient species •CCl3,
produced by the dissociation of CCl4•-, formed by the
capture of an electron from a high-energy electron pulse
(Scheme 4).41 The absorption spectra of DMN•+ and
(41) Shida, T.; Haselbach, E.; Bally, T. Acc. Chem. Res. 1984, 17,
180 and references therein.

3972 J . Org. Chem., Vol. 63, No. 12, 1998
Pandey et al.
F igu r e 6. Singlet-singlet absorption spectra of DMA in the
absence (a) and presence of 0.03 M (b) and 0.1 M (c) of 1a in
CH3CN solution.
F igu r e 5. (A) Transient absorption spectra of DMN•+ (curve
1, +) and DMA•+ (curve 2, 2) obtained by pulse radiolysis of
DMN and DMA in CCl4 solution. (B) Transient absorption
spectra obtained by pulse radiolysis (curve 1, b) and flash
photolysis (curve 2, [) of 1a in CH3CN solution.
Sch em e 6
Sch em e 5
Sch em e 7
The transient absorption spectrum (Figure 5B, curve
1) obtained due to pulse radiolysis of 1a in acetonitrile
has two absorption maxima at 440 and 490 nm, suggest-
ing that highly unstable 1a •- possibly dissociates ef-
ficiently to some other species (Scheme 5).
tempted to carry out a picosecond laser flash photolysis
experiment on a mixture of DMA (OD ) 0.6) and 1a (0.03
M) in CH3CN. The transient absorption spectrum of
DMA in the picosecond time domain was recorded in the
absence and presence of 1a . The singlet-singlet (S-S)
absorption spectrum (Figure 6, curve a) of DMA (in
absence of 1a ) showed two peaks at 550 and 930 nm.
Addition of 0.03 M of 1a to DMA solution resulted in
significant quenching (Figure 6, curve b) of the singlet
of DMA. This indicated the possibility of an interaction
between the singlet state of DMA and 1a . The addition
of a higher concentration (0.1 M) of 1a to DMA solution
resulted in the complete quenching of the excited singlet
state of DMA (Figure 6, curve c). However, transient
species absorption corresponding to either DMA•+ (460
nm) or 1a •- (490 nm) could not be observed due to the
low sensitivity of picosecond absorption technique below
the 500 nm region where the DMA•+ has significant
absorption and also possibly a very low quantum yield
of the cation radical. Identical experiments conducted
using DMN as sensitizer also could not show the DMN•+
transient absorption due to the same reason as men-
tioned above.
(v) Na n osecon d La ser F la sh P h otolysis. To gain
possible evidence of interaction between the triplet states
of DMN or DMA and 1a , we undertook laser flash
photolysis studies and lifetime measurements in the
nanosecond time domain.
The triplet-triplet (T-T) absorption spectra of DMA
(OD ) 0.6) in acetonitrile solution were recorded in the
nanosecond time domain. The absorption spectrum
displayed a well-defined peak (λmax) at 440 nm that
extended up to 700 nm (Figure 7A, curve 1). The lifetime
To gather evidence of 1a •- dissociating into the corre-
sponding R3Si• and PhSe-, a homolysis experiment of 1a
with higher concentration (0.09 M) in acetonitrile using
355 nm light from a Nd:YAG laser was undertaken. The
absorption spectrum (Figure 5B, curve 2) of the transient
species showed a well-defined peak at 440 nm with a
shoulder at 510 nm. Homolysis of 1a was expected to
produce R3Si• and PhSe• species. The transient absorp-
tion at 440 nm could be assigned to the silyl radical
species on the basis of literature data.42a Therefore, the
other transient absorption at 510 nm may be assigned
to the phenylselenyl radical (PhSe•). The transient
absorption of PhSe• at 510 nm has been further confirmed
by the homolysis experiment of PhSeSePh using 355 nm
light pulses.42b Since not much change in the UV
absorption spectrum (200-450 nm) of 1a before and after
irradiation at 355 nm was observed, the possibility of a
faster rate of geminate radical recombination compared
to their dissociation rate was obvious (Scheme 6).
On the basis of the above experiments, it may, thus,
be pointed out that 1a •-, produced by the pulse radiolysis,
dissociates into the corresponding silyl radical (440 nm)
and phenylselenide anion (490 nm) (Scheme 5).
(iv) P icosecon d La ser F la sh P h otolysis. To ob-
serve possible generation of transient species 1a •- and
DMA•+ or DMN•+ during PET reaction, first we at-
(42) (a) Sluggett, G. W.; Leigh, W. J . Organometallics 1992, 11, 3731.
(b) Palit, D. K.; Bhasikuttan, A. C.; Sapre, A. V. Unpublished results.

Bimolecular Group Transfer Radical Reactions
J . Org. Chem., Vol. 63, No. 12, 1998 3973
F igu r e 8. Transient absorption spectra obtained by laser
flash photolysis of DMN in CH3CN in the absence (curve 1,
T-T absorption spectra of DMN, 9) and presence of 0.03 M
(curve 2, b) and 0.09 M (curve 3, 2) of 1a . Inset shows the
transient decay monitored at 520 nm in the absence (curve 1,
τ ) 28 µs) and presence (curve 2, τ ) 14 µs) of 1a .
F igu r e 7. (A) Transient absorption spectra obtained by flash
photolysis of DMA in CH3CN in the absence (curve 1, T-T
absorption spectra of DMA, +) and presence of 0.03 M of 1a
(curve 2, 9). Inset shows the decay traces monitored at 550
nm for only DMA (curve 1, τ ) 33 µs) and DMA + 0.03 M of
1a (curve 2, τ ) 25 µs). (B) Transient absorption spectra
obtained by laser flash photolysis of DMA in the presence of
0.05 M (curve 1, b) and 0.09 M (curve 2, 2) of 1a in CH3CN.
absence and presence of 1a . Curve 1 in Figure 8 is the
T-T absorption spectrum of DMN in CH3CN. The
lifetime of the tDMN* was determined to be 28 µs.
Addition of 0.03 M of 1a resulted in the quenching of the
triplet state of DMN, and the lifetime of DMN was seen
to decrease to 14 µs (Figure 8, inset). Addition of a higher
concentration (0.09 M) of 1a resulted in increased
absorption in the wavelength region 500-650 nm (Figure
8, curve 3), providing an indication of the formation of
DMN•+ in solution due to interaction between the DMN
triplet state and 1a via ET phenomena.
From the above steady-state and time-resolved fluo-
rescence quenching studies and laser flash photolysis
experiments, it is confirmed that the ET from the excited
state of DMA or DMN (both singlet as well as triplet) to
1a leads to the formation of 1a •-. However, due to strong
overlapping of the absorption spectra of the transients
produced, it has not been possible to comment on the
relative yields of the two different channels (singlet and
triplet) of the ET reaction. To support above photophysi-
cal observations and to study the mesolytic pattern of
1a •-, a preparative PET activation of 1a , utilizing the
photosystem as shown in Figure 1, was initiated.
Sch em e 8
of the triplet state of DMA was determined to be 33 µs.
Addition of 1a (0.03 M) resulted in the quenching of the
triplet state of DMA, and the lifetime of DMA was seen
to decrease to 25 µs (Figure 7A, inset). This observation
suggested the interaction of 1a with the triplet excited
state of DMA too.43
The shape of the T-T absorption spectrum obtained
in the presence of 0.03 M concentration of 1a was quite
different from that of the original T-T absorption
spectrum (Figure 7A, curve 2) of DMA. Kinetic analysis
at 440 nm was complicated possibly due to the presence
of more than one transient species, i.e., the triplet state
of DMA, DMA•+, and 1a •-. However, kinetic analysis of
the transient decay monitored at 550 nm revealed that
the triplet lifetime of DMA had been quenched to 25 µs,
indicating the interaction of the triplet state with 1a .
Monitoring the transient species after the addition of a
higher concentration of 1a (0.05 and 0.09 M, Figure 7B)
in DMA (OD ) 0.6) solution indicated the formation of
DMA•+ (460 nm) and 1a •- (440 nm). An indirect ap-
proach was also made to characterize the transient
absorption species related to DMA•+ by the sensitized
generation of BP•- (750 nm) (Scheme 8); however, this
attempt failed, possibly due to very low concentration of
DMA•+ generated in the solution and, therefore, suggest-
(B) P ET Activa tion of 1a : A P r ep a r a tive Stu d y
A dilute solution of acetonitrile containing a mixture
of 1a (0.3 mmol), DMN (0.1 mmol), and ascorbic acid (0.3
mmol) was irradiated through Pyrex-filtered light (λ g
310 nm) using a 450 W Hanovia medium-pressure lamp
at room temperature. The progress of the reaction was
monitored by HPLC (C18 reversed-phase column: CH3CN/
H2O ) 70:30) analysis. After 5 h of irradiation, when
considerable consumption of 1a was noticed, irradiation
was discontinued and solvent was removed under vacuum.
The crude reaction mixture was purified by column
chromatography to give 5 and 6. The products (5 and 6)
t
ing a very low quantum yield of the DMA* electron-
transfer channel.
A similar experiment was also carried out using DMN
to monitor the corresponding transient species in the
1
were characterized by H NMR and mass spectral data.
(43) Absence of curvature in the S-V plots (Figures 3 and 4) may
possibly be due to very low quantum yield of electron-transfer channel
available from the triplet state of the excited donor.
DMN was recovered unchanged after the reaction in
about 98% yield.

3974 J . Org. Chem., Vol. 63, No. 12, 1998
Pandey et al.
Sch em e 10
with DMA was found45 to be 0.223. Similar correlation
studies, as described for DMN, gave kr(DMA) as 1.62 ×
1010 M-1 s-1 and φlim(DMA) as 0.12, respectively. The
overall reaction rate (kr(DMA); calculated from inverse
plot, Figure 9B) was found to be in good agreement with
the rate constant obtained from fluorescence quenching
[kqfSS(DMA) ) 1.46 × 1010 M-1 s-1].
On the basis of the above observations, the present
reaction could be summarized by considering single-
electron transfer (SET) from the excited state of DMN*
and DMA* (singlet as well as triplet) to 1a via charge-
transfer-stabilized exciplex intermediate (A). The inter-
mediate A dissociates into a free-radical ion pair (FRIP,
C) via the solvent-separated ion pair (SSIP, B) resulting
in 1a •- and ERA•+ as anticipated in solvents of high
dielectric constant. Fast disproportionation of 1a •- leads
to the formation of R3Si• and PhSe- (Scheme 10).
F igu r e 9. Inverse plot of [φdis] vs [1a ] with DMN (A, 0) and
DMA (B, 4).
Sch em e 9
(C) Syn th etic P er sp ectives
It is evident from the steady-state and laser flash
photolysis studies that photoexcitation of the DMN-1a
pair facilitates one-electron exchange to produce 1a •- as
well as DMN•+, and 1a •- on mesolysis generates R3Si• and
PhSe- species. This efficient dissociation pattern of 1a •-
led us to envisage the utilization of -Si-Se- bonds in
general as silyl radical equivalents, and hence, we
decided to explore the synthetic potentials of such
fragmentations for BMGT35 radical reactions, as Ph-
SeSePh, a good radical trapping agent,46 is produced in
the process. Most of the existing methodologies for
radical reactions are based on tin hydride,47,48 despite its
inherent limitations such as toxicity and difficulty in
removing tin byproducts. The serious disadvantage of
the tin hydride system lies in the loss of functionality
due to the termination of radical sequence by hydrogen
abstraction and reduction of the starting radical under
normal experimental conditions. Although some partial
solutions to these problems have been addressed by
introducing rather expensive reagents,49-51 atom transfer
approach,52 or utilizing modified tin-based reagents,53,54
a practical solution to this problem has remained elusive.
Mechanistically, the formation of 5 and 6 could be
rationalized by implicating the mesolysis of the primary
photoproduct 1a •- to R3Si• and PhSe-
. Subsequent
dimerization of the silyl radical led to the formation of 5
while the diphenyl diselenide (6) is obtained either by
the oxidative dimerization of the resultant phenyl se-
lenide anion directly or via the phenylselenol produced
by the protonation of PhSe-. Thus, it appears that the
electronegativity difference between selenium and silicon
directs the fragmentation of 1a •- as shown in Scheme 9.
The quantum yield (φdis(DMN)) for the disappearance of
1a in the above reaction (Scheme 9) was estimated by
photolyzing 1a and DMN using 310 nm monochromatic
light.44 The value of φdis(DMN) was found to be 0.042.
To implicate steady-state photochemical reaction and
photophysical observations through a common intermedi-
ate, a quantitative description was also derived by
correlating the rates of fluorescence quenching (kqfSS)
and overall ET quenching rate constant (kr) through a
reciprocal plot (Figure 9A) of 1/φdis vs 1/C. The overall
reaction rate constant kr(DMN) ) 0.991 × 1010 M-1 s-1
obtained from the reciprocal plot (Figure 9A) is in good
correlation with the Stern-Volmer fluorescence quench-
ing rate constant [kqfSS(DMN) ) 0.82 × 1010 M-1 s-1].
The limiting quantum yield (φlim(DMN)) for overall reac-
tion was found to be 0.034.
(45) The corresponding quantum yield value for the disappearance
of 1a with DMN is observed to be less when compared with DMA. This
difference could be possibly due to the competitive ground-state light
absorption of 1a at 310 nm.
(46) (a) Newcomb, M.; Marquardt, D. J .; Kumar, M. U. Tetrahedron
1990, 46, 2345. (b) Byers, J . H.; Gleason, T. G.; Knight, K. S. J . Chem.
Soc., Chem. Commun. 1991, 354.
(47) (a) Giese, B. Radicals in Organic Synthesis: Formation of
Carbon-Carbon Bonds; Pergamon Press: Oxford, 1986. (b) Beckwith,
A. L. J .; Westwood, S. W. Tetrahedron 1989, 45, 5269. (c) Ramaiah,
M. Tetrahedron 1987, 43, 3541. (d) Pereyer, M.; Quintard, J .-P.; Rahm,
A. Tin in Organic Synthesis; Butterworths: London, 1987; Part 2.
(48) (a) Curran, D. P.; Fevig, T. L.; J asperse, C. P. Chem. Rev. 1991,
91, 1237. (b) Curran, D. P. Synthesis 1988, 417, 489.
(49) Clark, K. B.; Griller, D. Organometallics 1991, 10, 746 and
references therein.
PET activation of 1a by visible light (410 nm)44 using
DMA as the electron donor also led to the formation of 5
and 6 in more or less the same proportions. The
quantum yield (φdis(DMA)) for the disappearance of 1a
(44) Murov, S. L. Hand Book of Photochemistry; Marcel Dekker:
New York, 1973; p 124.
(50) Giese, B. Angew. Chem., Int. Ed. Engl. 1985, 24, 553 and
references therein.

Bimolecular Group Transfer Radical Reactions
J . Org. Chem., Vol. 63, No. 12, 1998 3975
F igu r e 11. Radical sequences for group-transfer cyclization
reactions.
F igu r e 10. Complete photocycle for PET-initiated BMGT
radical reactions.
Sch em e 12
Sch em e 11
Since our study, as described above (Scheme 9), has
suggested the easy generation of silyl radicals and
PhSeSePh-, a good radical terminating agent, (formed
by the oxidative dimerization55 of PhSe-) in one pot by
the PET reductive activation of 1a , it was envisaged that
this chemistry could be utilized for the group-transfer
radical reactions as depicted in Figure 10.
Initially, substrate 9 was chosen to evaluate the
applicability of 1a as a reagent for BMGT radical
reactions. 9 was prepared in 79% yield by the sequential
alkylation of in situ generated diethyl malonate anion,
with allyl bromide (21 mmol) and 1,2-dibromoethane (15
mmol), respectively (Scheme 11).
PET cyclization of 9 was achieved by irradiating a
dilute solution containing a mixture of 9 (1.4 mmol), 1a
(1.4 mmol), DMN (0.55 mmol), and ascorbic acid (1.4
mmol) through Pyrex-filtered light (λ g 310 nm) using a
450 W Honovia medium-pressure mercury vapor lamp
at room temperature without removing dissolved oxygen
from the solution. The progress of the reaction was
monitored by both TLC and HPLC analysis. After 9 h
of irradiation, when quantitative consumption of 9 was
noticed, irradiation was discontinued. Solvent was re-
moved under vacuum, and the crude photolyzate was
purified by silica gel chromatography to give a viscous
yellow oily product (10a ) in 75% yield. Product 10a was
1
characterized by H NMR, 13C NMR, and MS spectral
analysis (for details see the Experimental Section). DMN
was recovered unchanged in about 98% yield.
Formation of another minor product (<12% yield,
estimated by HPLC analysis) was also observed during
the cyclization of 9; however, this could not be isolated
in pure form for complete spectral characterization, but
1H NMR of the crude mixture indicated the possibility
of this compound being an endo-trig cyclization product
(10b).
Generally, the carbon-centered radical generation by
Bu3SnH is performed by the reductive cleavage of the
-C-Br bond.47d The reaction using the -C-Cl bond has
been found to be significantly sluggish. Therefore, to
compare the -C-Cl bond cleavage efficiency by silyl
radical species, compound 11 was also included in our
study. Preparative PET reaction of 11, using an experi-
mental setup identical to that described for 9, led to the
formation of cyclized product 10a in 63% yield (Figure
11). A comparative rate study for the cyclizations of 9
and 11 suggested both to be comparable. Therefore, it
may be suggested that radical chain initiation through
this approach could be achieved by employing -C-Cl
bonds as well.
(51) (a) Chatgilialoglu, C. Acc. Chem. Res. 1992, 25, 188 and
references therein. (b) Chatgilialoglu, C.; Ferreri, C. Res. Chem.
Intermed. 1993, 19, 755. (c) Chatgilialoglu, C.; Ingold, K. U.; Scaiano,
J . C. J . Am. Chem. Soc. 1982, 104, 5123.
(52) (a) Curran, D. P.; Chang, C.-T. Tetrahedron Lett. 1987, 28, 2477.
(b) Curran, D. P.; Chen, M.-H.; Kim, D. J . Am. Chem. Soc. 1986, 108,
2489.
(53) Zard, S. Z.; Boivin, J .; Camara, J . J . Am Chem. Soc. 1992, 114,
7909.
(54) (a) Keck, G. E.; Yates, J . B. J . Am. Chem. Soc. 1982, 104, 5821.
(b) Keck, G. E.; Enholm, E. J .; Yates, J . B.; Wiley, M. R. Tetrahedron
1985, 41, 4079.
To evaluate the effectiveness of PET-generated silyl
radical for the generation of vinyl as well as arene
radicals, cyclizations of substrates 12 and 13 were also
initiated, which furnished 14 and 15 in 63% and 55%
yield, respectively (Scheme 12). These products were
1
characterized by H NMR, 13C NMR, and MS spectral
(55) Foster, D. G. Organic Synthesis; Wiley: New York, 1955;
Collect. Vol. 3, p 771.
data.

3976 J . Org. Chem., Vol. 63, No. 12, 1998
Pandey et al.
Ta ble 1. Bim olecu la r Gr ou p Tr a n sfer Ra d ica l
Rea ction sa
F igu r e 12. Radical chain sequences for intermolecular-chain
addition reactions.
Sch em e 13
chain addition reactions. It was envisioned that the
polymerization problem, normally observed during in-
termolecular radical additions using tributyltin hydride
reagent, could be avoided as the termination step in this
case (Figure 12) involves fast phenylselenyl group trans-
fer rather than H atom transfer. Toward this endeavor,
a
Key: (i) see the Supporting Information for the preparation
and characterization of 16-22; (ii) characterized by H NMR, 13
C
1
NMR, and mass spectral analyses; (iii) cis/trans and E/Z ratios
are confirmed by using HPLC; isomers were not separable by
column chromatography; (iv) in the case of 24 and 28 only the
major isomer was isolated; the minor isomer was not obtained in
enough quantity for spectral characterization; (v) isolated yields,
but not optimized.
•
we desired to study the addition of CH2CO2C2H5 [from
the precursor bromoethylacetate (30)] across the CdC
bond of 31. Substrate 31 was prepared by following the
simple steps as shown in Scheme 13.
Usual PET reaction of a dilute solution of acetonitrile
containing a mixture of 1a (1.6 mmol), DMN (0.5 mmol),
ascorbic acid (1 mmol), 31 (1.8 mmol), and 30 (2.3 mmol)
furnished a yellow oily product in 55% yield. The product
was characterized as 32b by 1H NMR, 13C NMR, and
mass spectral analysis (for details see the Experimental
Section).
Gen er a lity of BMGT Rea ction s. To test the gen-
erality and applicability of 1a as a bimolecular group
transfer reagent, preparative cyclizations of a series of
bromoallyl ethers 16-19 and bromopropargyl ethers 20-
22 were undertaken, and the results are presented in
Table 1. Spectral details of the cyclic products 23-29
are presented in the Experimental Section.
Con clu sion
In ter m olecu la r Ra d ica l Ad d ition Rea ction s. En-
couraged by the success in conducting BMGT radical
cyclization reactions using 1a as reagent, we diverted our
attention using this protocol toward intermolecular group
transfer radical chain addition reactions. Although much
attention have been focused on radical cyclization reac-
tions, intermolecular radical chain addition reactions
have remained unexplored due to the possible competing
hydrostannation reaction of electron-deficient alkenes
using conventional trialkyltin reagent as initiator.48b
Furthermore, the primary alkyl radical formed by the
cleavage of the -C-X (X ) halogen) bond has two
possible options for termination with similar rate con-
stants either by hydrogen atom abstraction from tin
hydride or by addition to an electron-deficient alkene.
Although the addition of alkyl radicals to an alkene can
be directed to some extent by using excess of olefin, the
coupling of adduct radicals with another molecule of
alkene leads to the formation of telomers and polymers.48b
Despite these limitations, the unique features of radical
additions make them useful, and some intermolecular
radical additions are precedented in the literature.56-58
In this context, we evaluated the expansion of the scope
of our methodology (Figure 10) for intermolecular radical-
In conclusion, we have developed a conceptually new
photosystem to drive PET reductive activation of -Si-
Se- bonds. Although preliminary studies such as steady-
state photolysis experiments indicated the ET processes
from DMN or DMA to -Si-Se- bonds, a keen insight
on the mechanistic understanding is obtained from
picosecond and nanosecond laser flash photolysis studies.
The synthetic potential of 1a •- mesolysis is explored in
BMGT radical reactions and intermolecular group trans-
fer chain addition reactions. This study adds a new
dimension in the radical chemistry from a conceptual and
ecological viewpoint.
Exp er im en ta l Section
Gen er a l Meth od s. DMN,37 DMA,39 and PhSeSePh59 were
synthesized and purified following the literature procedures.
Cyclohexane and acetonitrile that are used in fluorescence
(56) (a) Giese, B.; Gonzalez-Gomez, J . A.; Witzel, T. Angew. Chem.,
Int. Ed. Engl. 1984, 23, 69. (b) Giese, B. Ibid. 1983, 22, 753.
(57) (a) Curran, D. P.; Eichenberger, E.; Collis, M.; Roepel, M. G.;
Thoma, G. J . Am. Chem. Soc. 1994, 116, 4279. (b) Curran, D. P.;
Thoma, G. Ibid. 1992, 114, 4436.
(58) (a) Byer, J . H.; Lane, G. C. J . Org. Chem. 1993, 58, 3355. (b)
Byer, J . H.; Harper, B. C. Tetrahedron Lett. 1992, 33, 6953.

Bimolecular Group Transfer Radical Reactions
J . Org. Chem., Vol. 63, No. 12, 1998 3977
studies and HPLC analysis, respectively, were of spectroqual-
ity (E. Merck, India) grade. Silica gel for column chromatog-
raphy (finer than 200 mesh) was obtained from Acme, India.
All nuclear magnetic resonance spectra were recorded on
either Bruker AC 200 FT NMR or Brucker MSL 300 NMR
spectrometers using CDCl3 as solvent. All chemical shifts are
reported in parts per million downfield from TMS; coupling
constants are given in hertz. IR spectra were taken on a
Perkin-Elmer model 2830 spectrometer. Mass spectra were
obtained at a voltage of 70 eV on a Finnigan MAT-1020B
instrument.
Fluorescence lifetimes were measured using the time-
resolved fluorescence spectrometer (model 199 from Eding-
burgh Instruments, U.K.), which uses a hydrogen-discharge
flash lamp providing pulses of 1 ns duration for excitation and
single-photon-counting techinque for detection.
Fluorescence spectra were recorded on a Spex-Fluorolog 212
spectrofluorimeter. The excitation and emission slit widths
were maintained at 0.5 mm. The steady-state emission
spectral measurements were carried out using a 1 cm × 1 cm
quartz cell. A right angle configuration for the cell holder was
utilized during the measurement of excitation and emission
spectra.
third (355 nm, 5 mJ ) harmonic output pulses of 35 ps duration
from an active passive mode-locked Nd:YAG laser (Continuum,
Model 501-C-10) used for excitation and continuum probe
pulses in the 400-920 nm region were generated by focusing
the residual fundemental in the H2O/D2O mixture (50:50). The
probe was delayed with respect to the pump pulse using a 1
m long linear motion translation stage, and the transient
absorption spectra at different delay times (up to 6 ns) were
recorded by an optical multichannel analyzer (Spectroscopic
Instruments, Germany) interfaced to an IBM-PC. At the zero
delay position, the probe light reaches the sample just after
the end of the pump pulse. Transients surviving beyond 50
ns were studied by monitoring their absorption using a
tungsten filment lamp in combination with a Bausch and
Lomb monochromator (f/10, 350-800 nm), Hamamatsu R 928
PMT, and a 500 MHz digital oscilloscope (Tektronix, TDS-
540A) connected to a PC.
Qu a n tu m Yield Mea su r em en ts. Samples for the quan-
tum yield determinations were prepared by pipetting out a
noted volume from the stock solution of 1a into Pyrex tubes
and irradiated in a Quantum yield reactor (Model QYR-20)
using a 200 W mercury lamp. Irradiations were carried out
for a short interval of time to bring about only 8-10% of the
conversion. A uranyl oxalate actinometer was used to monitor
the intensity of light. The quantitative loss of 1a was
estimated by HPLC [column: C18 , eluent: CH3CN/H2O (3:1) ].
The limiting quantum yield (φlim) for 1a was obtained from
the inverse plot of the variable 1a concentrations vs quantum
yields i.e., [1a ]-1 vs φdis-1, by keeping the DMN/DMA concen-
tration fixed.
P r ep a r a t ion of ter t-Bu t yld ip h en yl(p h en ylselen o)-
sila n e (1a ).36 A solution of diphenyl diselenide (15.6 g, 0.05
mol) in 100 mL of dry THF was added to freshly prepared
sodium sand (3 g, 0.14 mol), and the resulting mixture was
refluxed for 5 h. tert-Butyldiphenylsilyl chloride (27.5 g, 0.1
mol) was added to the stirring solution, and the resulting
mixture was allowed to reflux for 21 h. The reaction mixture
was cooled to 0 °C, and 5 mL of methanol was added to kill
the excess of sodium. The reaction mixture was extracted with
100 mL of ether, washed with several portions of cold water,
dried over sodium sulfate, and concentrated under reduced
pressure. Distillation of the crude reaction mixture gave 30
g (79%) of 1a as thick pale yellow liquid, which on long
standing (24 h) solidified: mp ) 57 °C (lit.36 mp 56-58 °C);
IR (KBr) 3038, 2910, 2840, 1563, 805, 720, 660 cm-1; 1H NMR
(200 MHz) δ 7.64 (m, 4H), 7.29 (m, 6H), 6.98 (m, 5H), 1.14 (s,
9H); MS 396 (M+).
The linear accelarator used in pulse radiolysis experiments
was obtained from M/S Viritech Ltd., U.K.
A mode-locked Nd:YAG laser was used for excitation pur-
porses in picosecond and nanosecond laser flash photolysis
experiments.
HPLC analysis was performed on a Perkin-Elmer (model
250 binary LC pump along with LC 135C diode array detector)
liquid chromotograph using reversed-phase C18 (bondapack 0.5
µm) column, eluting with CH3CN/H2O (75:25) solvent mixture
degassed by a freeze-thaw cycle procedure.
Cyclic Volta m m etr y. The cyclic voltammetry experiments
were carried out with a three-electrode assembly on a PAR
175 universal programmer and PAR RE0074 XY recorder. The
cell consisted of a Metro E410 hanging mercury drop electrode
(HMDE) and Pt wire (auxiliary electrode). The supporting
electrolyte was tetraethylammonium perchlorate, and poten-
tials are referred to SCE and are uncorrected for liquid
junction potentials.
Stea d y-Sta te F lu or escen ce Qu en ch in g. Quenching of
DMN or DMA fluorescence was carried out by using 1a as
quencher. For the determination of Stern-Volmer constants
(KqfSS), the intensity (I0) of steady-state fluoresence at the
maximum emission (λemi(DMN) ) 344 nm; λemi(DMA) ) 440
nm) was measured from DMN/DMA solution (3 × 10-4 M) in
CH3CN at 25 °C, keeping the excitation wavelength (λexi ) at
310 nm for DMN and 410 nm for DMA. Subsequently, the
fluoresence-quenching intensity (I) was measured as a function
of concentration [Q] of 1a . Linear plots were obtained on the
basis of the equation I0/I ) 1 + Kqfτ[Q], where I0 denotes the
fluoresence intensity in the absence of quencher. No curvature
was noticed in any system, and intercepts were 1.05 ( 0.007
in all cases. Slopes were determined by least-squares fit, and
coefficients were always >0.99.
P u lse Ra d iolysis. High-energy (7 meV) electron pulses of
50 ns duration generated from a linear electron accelerator
(LINAC) were used for pulse radiolysis experiments. The
details of the pulse radiolysis setup have been described
elsewhere.60 The transient species produced were detected by
monitoring the optical absorption. The absorbed dose was
determined by using the aerated KSCN solution Gꢀ ) 21 520
dm3 mol-1 cm-1 for 100 eV of absorbed dose (the G value is
the number of radicals or molecules produced per 100 eV of
P ET Activa tion of 1a Usin g DMN a s Sen sitizer .
A
dilute solution of acetonitrile (300 mL) containing a mixture
of 1a (0.120 g, 0.3 mmol), DMN (0.070 g, 0.1 mmol), and
ascorbic acid (0.05 g, 0.3 mmol) was irradiated through Pyrex-
filtered light (λ g 310 nm) using a 450 W Hanovia medium-
pressure lamp at room temperature. The progress of the
reaction was monitored by HPLC (C18 reversed-phase col-
umn: CH3CN/H2O ) 70:30) analysis. After 5 h of irradiation,
when considerable consumption of 1a was noticed, irradiation
was discontinued, and solvent was removed under vacuum.
The crude reaction mixture was purified by column chroma-
tography to give 5 and 6. The products (5 and 6) were
characterized by comparing their spectral data with the
reported values. DMN was recovered unchanged after the
reaction in about 98% yield.
Visib le-Ligh t P E T Act iva t ion of 1a Usin g DMA a s
Sen sitizer . PET activation of 1a by visible light (410 nm)
using DMA as the electron donor also led to the formation of
5 and 6 in more or less the same proportions. This reaction
was performed in a specially designed double-walled photo-
reactor. The photoreactor consisted of three chambers. The
first and outermost chamber contained irradiation solution,
while the second one was charged with a CuSO4‚5H2O/NH3
filter solution. This filter solution allowed only the 410 nm
wavelength light to pass through.31 A 450 W Hanovia medium-
pressure mercury lamp was used as the light source and was
housed into a water-circulated double-jacketed chamber im-
absorbed energy and ꢀ is the molar absorptivity at 500 nm for
•-
the transient (SCN)2
.
La ser F la sh P h otolysis. The laser flash photolysis ex-
perimental setup has been described elsewhere.61 In brief, the
(59) Hori, T.; Sharpless, K. B. J . Org. Chem. 1978, 43, 1689.
(60) Guha, S. N.; Moorthy, P. N.; Kishore, K.; Naik, D. B.; Rao, K.
N. Proc. Indian. Acad. Sci. (Chem. Sci.) 1989, 99, 261.
(61) Ghosh, H. N.; Sapre, A. V.; Palit, D. K.; Mittal, J . P. J . Phys.
Chem. B 1997, 101, 2315.

3978 J . Org. Chem., Vol. 63, No. 12, 1998
Pandey et al.
mersed in the second chamber, maintaining 1 cm path length
of the filter solution. The whole photoreactor was made of
Pyrex glass.
67 (100). Anal. Calcd for C14H18OSe: C, 59.79; H, 6.45.
Found: C, 59.55; H, 6.43.
.
Com p ou n d 26: IR (neat): 3080, 1615, 1125 cm-1 1H NMR
Gen er a l Ir r a d ia tion P r oced u r e for th e P ET-P r om oted
Ra d ica l Ch a in Gr ou p Tr a n sfer Rea ction s. Meth od A.
Usin g DMN a s Sen sitizer . This is illustrated by taking 9
as an example.
(200 MHz): δ 1.65-1.85 (m, 9H), 2.60 (m, 1H), 3.00 (m, 2H),
3.45 (m, 1H), 3.95 (m, 1H), 4.55 (m, 1H), 7.30 (m, 3H), 7.55
(m, 2H). 13C NMR (75.47 MHz): δ 133.18, 132.77, 130.55,
130.13, 128.88, 127.05, 126.77, 86.45, 85.78, 73.53, 72.34,
50.16, 48.15, 47.25, 43.58, 34.64, 34.18, 32.52, 30.81, 26.54,
26.14, 25.32, 24.25. MS m/e (relative intencity): 296 (M+, 10),
157 (15), 139 (29), 77 (52).
Com p ou n d 27: IR (neat) 3085, 1620, 1200, 1129 cm-1; 1H
NMR (200 MHz, CDCl3) δ 1.25 (t, 3H, J ) 8.1 Hz), 2.75 (m,
2H), 3.50 (m, 1H), 3.80 (m, 1H), 4.50 (m, 2H), 5.35 (m, 1H),
6.40 (m,1H), 7.25 (m, 3H), 7.55 (m, 2H); 13C NMR (50.32 MHz,
CDCl3) δ 145.24, 143.83, 131.55, 131.29, 131.16, 129.24, 127.73,
126.89, 126.80, 108.44, 107.59, 103.51, 69.66, 69.44, 62.79,
62.65, 40.73, 39.60, 15.26; MS m/e (rel intensity) 284 (M+, 1),
157 (15), 129 (100), 115 (23), 91 (16). Anal. Calcd for
A dilute solution containing a mixture of 9 (0.43 g, 1.4
mmol), 1a (0.55 g, 1.4 mmol), DMN (0.13 g, 0.55 mmol), and
ascorbic acid (0.25 g, 1.4 mmol) was irradiated through Pyrex-
filtered light (λ g 310 nm) using a 450 W Honovia medium-
pressure mercury vapor lamp at room temperature without
removing dissolved oxygen from the solution. The progress
of the reaction was monitored both by TLC and HPLC analysis.
After 9 h of irradiation, when quantitative consumption of 9
was noticed, irradiation was discontinued. Solvent was re-
moved under vacuum, and the crude photolyzate was purified
by silicagel chromatography to give 0.39 g of a viscous yellow
oily product (10a ) in 75% yield. The product 10a was
characterized by 1H NMR, 13C NMR, and mass spectral
analysis, and the data are as follows: IR (neat) 3070, 2860,
C
13H16O2Se: C, 55.13; H, 5.69. Found: C, 54.89; H, 5.71.
Com p ou n d 28: IR (neat) 3085, 1625, 1200, 1129 cm-1; 1H
NMR (300 MHz, CDCl3) δ 1.65 (m, 3H), 2.30-2.55 (m, 2H),
3.50 (m, 1H), 3.80 (m, 1H), 4.50 (m, 2H), 5.35 (m, 1H), 6.40 (s,
1H), 7.25 (m, 3H), 7.55 (m, 2H); 13C NMR (50.32 MHz, CDCl3)
δ 144.97, 132.85, 129.03, 128.99, 127.33, 101.55, 69.95, 61.05,
41.21, 25.06, 23.12; MS m/e (rel intensity) 296 (M+, 7), 157
(18), 139 (100), 125 (22), 91 (28). Anal. Calcd for
1
1710 cm-1; H NMR (200 MHz) δ 1.25 (t, 6H, J ) 9.75 Hz),
1.45 (m, 1H), 1.90 (m, 2H), 2.25 (m, 3H), 2.55 (m, 1H), 2.95 (d,
2H, J ) 4.75 Hz), 4.20 (q, 4H, J ) 9.75 Hz), 7.25 (m, 3H), 7.55
(m, 2H); 13C NMR (75.47 MHz): δ 172.57, 132.96, 129.27,
127.06, 61.61, 60.41, 40.94, 40.24, 34.06, 33.48, 32.64, 14.27;
MS m/e (rel intensity) 384 (M+, 18), 339 (7), 227 (52), 119 (27),
153 (100), 79 (28). Anal. Calcd for C18H24O4Se: C, 56.39; H,
6.31. Found: C, 56.34; H, 6.29.
C
14H16O2Se: C, 56.96; H, 5.46. Found: C, 56.79; H, 5.44.
Com p ou n d 29: IR (neat) 3085, 1645, 1210 cm-1; 1H NMR
(200 MHz, CDCl3) δ 1.60-1.75 (m, 6H), 2.75 (m, 1H), 3.60 (m,
2H), 4.05 (t, 1H, J ) 4.3 Hz), 6.65 (m, 1H), 7.30 (m, 3H), 7.55
(m, 2H); 13C NMR (75.47 MHz, CDCl3) δ 145.75, 146.01,
133.12, 133.99, 129.75, 129.30, 129.04, 117.75, 117.23, 83.02,
82.95, 74.45, 73.82, 46.15, 45.99, 37.91, 37.72, 33.35, 32.47,
26.75, 25.21; MS m/e (rel intensity) 280 (M+, 3), 157 (16), 123
(100), 109 (20), 91 (15). Anal. Calcd for C14H16OSe: C, 60.22;
H, 5.78. Found: C, 59.99; H, 5.76.
Meth od B. Usin g DMA a s Sen sitizer . Identical irradia-
tions (for the substrates 9, 11-13, and 16-22) as described
in method A were carried out using DMA as sensitizer at 410
nm wavelength light, in a specially designed photoreactor (see
Visible Light PET Activation of 1a ). Similar products (10a ,
14-15, and 23-29) were obtained in more or less the same
proporations.
Identical irradiation of substrates 11-13 and 16-22 gave
10a , 14, 15, and 23-29, respectively. The spectral details of
14, 15, and 23-29 are as follows.
Com p ou n d 14: IR (neat) 3050, 1620, 1550, 1460, 1210
cm-1; 1H NMR (200 MHz) δ 1.80-1.60 (m, 4H), 2.25-2.35 (m,
2H), 2.55-2.60 (m, 1H), 3.00 (m, 1H), 3.55 (m, 2H), 4.30-4.15
(m, 1H), 5.80 (d, 2H, J ) 4.5 Hz), 7.25 (m, 3H), 7.55 (m, 2H);
13C NMR (75.47 MHz) δ 147.21, 136.81, 135.40, 130.29, 128.93,
109.26, 79.62, 71.97, 50.21, 43.83, 33.15, 30.97, 26.35; MS m/e
(rel intensity) 294 (M+, 10), 238 (5), 157 (8), 125 (75), 107 (60),
81 (100).
Com p ou n d 15: IR (neat) 1590, 1470, 1230 cm-1; 1H NMR
(200 MHz) δ 2.15 (s, 3H), 3.10 (m, 2H), 3.30 (m, 1H), 4.80 (d,
2H, J ) 7.5 Hz), 6.75 (t, 1H, J ) 7.5 Hz), 6.95 (dd, 2H, J )
7.5, 16.5 Hz), 7.25 (m, 3H), 7.55 (m, 2H); 13C NMR (50.32 MHz)
δ 157.77, 133.00, 129.65, 129.30, 129.10, 127.24, 125.42,
122.34, 120.48, 119.59, 81.53, 37.70, 32.70, 15.11; MS m/e (rel
intensity) 304 (M+, 25), 172 (10), 157 (10), 147 (100).
Com p ou n d 23: IR (neat) 3070, 1200, 1125 cm-1; 1H NMR
(200 MHz, CDCl3) δ 1.20 (t, 3H J ) 9.7 Hz), 1.75 (m, 1H), 2.20-
2.50 (m, 2H), 3.00 (m, 2H), 3.45 (m, 1H), 3.75 (m, 2H), 4.05 (t,
In ter m olecu la r Ra d ica l Ch a in Ad d ition Rea ction .
Identical PET reaction of a dilute solution of acetonitrile
containing a mixture of 1a (0.55 g, 1.4 mmol), DMN (0.1 g,
0.5 mmol), ascorbic acid (0.18 g, 1 mmol), 30 (0.38 g, 2.3 mmol),
and 31 (0.49 g, 1.8 mmol) furnished a yellow oily product in
1
55% yield (0.32 g). The product 32b was characterized by H
NMR, 13C NMR, and mass spectral analysis, and the data are
as follows.
1H, J ) 9.7 Hz), 5.15 (m, 1H), 7.25 (m, 3H), 7.55 (m, 2H); 13
C
NMR (50.32 MHz, CDCl3) δ 132.97, 131.60, 130.00, 129.26,
129.18, 127.79, 127.09, 104.20, 104.01, 72.04, 62.99, 62.78,
39.86, 39.64, 38.67, 37.91, 32.33, 31.54, 15.42, 15.33; MS m/e
(rel intensity) 286 (M+, 286), 240 (15), 157 (23), 91 (42), 83
Com p ou n d 32b: IR (neat) 1710, 1580, 1420 cm-1; 1H NMR
(200 MHz) δ 0.10 (s, 9H), 0.95 (t, 3H, J ) 7.5 Hz), 1.45 (m,
3H), 1.25 (m, 10H), 1.95 (dd, 2H, J ) 4.5, 8.5 Hz), 2.55 (1H),
3.15 (m, 1H), 4.10 (2H, J ) 7.5 Hz), 7.25 (m, 3H), 7.55 (m,
2H); 13C NMR (75.47 MHz) δ 173.05, 135.64-127.25, 70.5,
70.4, 60.16, 43.5, 42.32, 41.99, 37.75-29.33, 25.84, 25,78, 24.68,
24.62, 22.42, 17.93, 14.1, -4.1, -4.4; MS m/e (rel intensity)
514 (M+, 4), 244 (100), 457 (18), 171 (30), 157 (20).
(100). Anal. Calcd for
Found: C, 54.53; H, 6.38.
C13H18O2Se: C, 54.74; H, 6.36.
Com p ou n d 24: IR (neat) 3085, 1210, 1122 cm-1; 1H NMR
(200 MHz, CDCl3) δ 1.60 (m, 4H), 2.15 (m, 1H), 2.65 (m, 1H),
2.95 (m, 2H), 3.70 (m,3H), 4.05 (t, 1H, J ) 7.15 Hz), 5.25 (d,
1H, J ) 4.3 Hz), 7.30 (m, 3H), 7.55 (m, 2H); 13C NMR (75.47
MHz, CDCl3) δ 132.90, 129.33, 128.95, 127.04, 101.62, 69.92,
60.90, 40.93, 37.21, 25.60, 22.80, 19.05; MS m/e (rel intensity)
298 (M+, 6), 197 (20), 157 (12), 141 (51), 116 (42), 95 (33), 77
(55), 69 (88), 55 (100). Anal. Calcd for C14H18O2Se: C, 56.57;
H, 6.10. Found: C, 56.45; H, 6.11.
Ack n ow led gm en t. This work was partly supported
by BRNS, Bombay, India. One of us (K.S.S.P.R.) thanks
BARC, Bombay, India, for the award of the Dr. K. S.
Krishnan fellowship.
Com p ou n d 25 : IR (neat) 3080, 1615, 1125 cm-1; 1H NMR
(200 MHz, CDCl3) δ 1.65-1.85 (m, 4H), 2.10-2.25 (m, 3H),
2.60 (m, 1H), 3.00 (m, 2H), 3.45 (m, 1H), 3.95 (m, 1H), 4.55
(m, 1H), 7.30 (m, 3H), 7.55 (m, 2H); 13C NMR (75.47 MHz,
CDCl3) δ 132.88, 132.54, 130.25, 130.03, 129.02, 126.98, 126.84,
86.18, 85.15, 73.31, 72.17, 50.07, 47.97, 47.04, 43.37, 34.46,
34.01, 32.32, 30.73, 26.34, 25.93, 25.21, 24.05; MS m/e (rel
intensity) 282 (M+, 10), 157 (15), 124 (17), 107 (20), 95 (75),
Su p p or tin g In for m a tion Ava ila ble: Experimental pro-
cedures and spectral characterization of 9, 11-13, 16-22, and
31 (5 pages). This material is contained in libraries on
microfiche, immediately follows this article in the microfilm
version of the journal, and can be ordered from the ACS; see
any current masthead page for ordering information.
J O972345E