Aziridination of alkenes and amidation of alkanes by bis(tosylimido)ruthenium(VI) porphyrins. A mechanistic study

Article abstract of DOI:10.1021/ja9913481

Bis(tosylimido)ruthenium(VI) porphyrins, [Ru(VI)(Por)(NTs)₂] (Por = TPP, TTP, 4-C1-TPP, 4-MeOTPP, OEP), were prepared in 60-74% yields by treatment of [Ru(II)(Por)(CO)(MeOH)] with (N-(p-tolylsulfonyl)- imino)phenyliodinane (PhI=NTS) in dichloromethane. In dichloromethane containing pyrazole, they reacted with alkenes or alkanes to give tosylamidoruthenium(IV) porphyrins, [Ru(IV)(Por)(NHTs)(pz)], in about 75% yields. The reactions of [Ru(VI)(TPP)(NTs)₂] and [Ru(VI)(OEP)(NTs)₂] with styrene, para-substituted styrenes, norbornene, cyclooctene, and β- methylstyrene afforded the corresponding N-tosylaziridines in 66-85% yields. The aziridination of cis-stilbene and cis-β-methylstyrene by [Ru(VI)(Por)(NTs)₂] is nonstereospecific with a partial loss of the alkene stereochemistry. Kinetic studies on the reactions between [Ru(VI)(TPP)(NTs)₂] and 16 alkenes (cyclooctene, norbornene, 2,3-dimethyl-2- butene, styrene, para-substituted styrenes, α and β-methylstyrene, and α- and β-deuteriostyrene) gave the second-order rate constants (k₂) ranging from (1.60 ± 0.06) x 10^-3 to (90 ± 4) x 10^-3 dm³ mo1^-1 s^-1 at 298 K. The slope of the linear plot of log k₂ vs E(l/2) for eight representative alkenes was found to be -1.7 V^-1. In the case of para-substituted styrenes, linear correlation between log k(R) (k(R) = relative rate) and σ⁺ gives a ρ⁺ value as small as -1.1. However, the effect of para substituents on k(R) can be best accounted for by considering both the polar and spin delocalization effect. Measurements on the secondary deuterium isotope effect revealed that only the β-carbon atom of styrene experienced a significant change in its hybridization in reaching the transition state. All these are consistent with rate-determining formation of a carboradical intermediate. The reactions of [Ru(VI)(TPP)(NTs)₂] and [Ru(VI-) (OEP)(NTs)₂] with adamantane, cyclohexene, ethylbenzene, and cumene resulted in tosylamidation of these hydrocarbons and afforded the corresponding amides in 52-88% yields. For cyclohexane and toluene, the tosylamidation products were formed in poor yields (ca. 10%). Kinetic studies on the reactions between [Ru(VI-) (TPP)(NTs)₂] and nine hydrocarbons (cumene, ethylbenzene, cyclohexene, and para-substituted ethylbenzenes) gave the second-order rate constants (k₂) in the range of (0.330 ± 0.008) x 10^-3 to (16.5 ± 0.3) x 10^-3 dm³ mol^-1 s^-1. These reactions exhibit a large primary deuterium isotope effect, with a k(H)/k(D) ratio of 11 for the tosylamidation of ethylbenzene. In the case of para-substituted ethylbenzenes, both electron-donating and -withdrawing substituents moderately promote the reaction. There is an excellent linear correlation between log k(R) and a related carboradical parameter. On the basis of these observations, a mechanism involving the rate-limiting formation of a carboradical intermediate is postulated.

Full text of DOI:10.1021/ja9913481

9120
J. Am. Chem. Soc. 1999, 121, 9120-9132
Aziridination of Alkenes and Amidation of Alkanes by
Bis(tosylimido)ruthenium(VI) Porphyrins. A Mechanistic Study
Sze-Man Au, Jie-Sheng Huang, Wing-Yiu Yu, Wai-Hong Fung, and Chi-Ming Che*
Contribution from the Department of Chemistry, The UniVersity of Hong Kong,
Pokfulam Road, Hong Kong, China
ReceiVed April 26, 1999
Abstract: Bis(tosylimido)ruthenium(VI) porphyrins, [RuVI(Por)(NTs)2] (Por ) TPP, TTP, 4-Cl-TPP, 4-MeO-
TPP, OEP), were prepared in 60-74% yields by treatment of [RuII(Por)(CO)(MeOH)] with (N-(p-tolylsulfonyl)-
imino)phenyliodinane (PhIdNTs) in dichloromethane. In dichloromethane containing pyrazole, they reacted
with alkenes or alkanes to give tosylamidoruthenium(IV) porphyrins, [RuIV(Por)(NHTs)(pz)], in about 75%
yields. The reactions of [RuVI(TPP)(NTs)2] and [RuVI(OEP)(NTs)2] with styrene, para-substituted styrenes,
norbornene, cyclooctene, and â-methylstyrene afforded the corresponding N-tosylaziridines in 66-85% yields.
The aziridination of cis-stilbene and cis-â-methylstyrene by [RuVI(Por)(NTs)2] is nonstereospecific with a
partial loss of the alkene stereochemistry. Kinetic studies on the reactions between [RuVI(TPP)(NTs)2] and 16
alkenes (cyclooctene, norbornene, 2,3-dimethyl-2-butene, styrene, para-substituted styrenes, R- and â-meth-
ylstyrene, and R- and â-deuteriostyrene) gave the second-order rate constants (k2) ranging from (1.60 ( 0.06)
× 10-3 to (90 ( 4) × 10-3 dm3 mol-1 s-1 at 298 K. The slope of the linear plot of log k2 vs E1/2 for eight
representative alkenes was found to be -1.7 V-1. In the case of para-substituted styrenes, linear correlation
between log kR (kR ) relative rate) and σ+ gives a F+ value as small as -1.1. However, the effect of para
substituents on kR can be best accounted for by considering both the polar and spin delocalization effect.
Measurements on the secondary deuterium isotope effect revealed that only the â-carbon atom of styrene
experienced a significant change in its hybridization in reaching the transition state. All these are consistent
with rate-determining formation of a carboradical intermediate. The reactions of [RuVI(TPP)(NTs)2] and [RuVI-
(OEP)(NTs)2] with adamantane, cyclohexene, ethylbenzene, and cumene resulted in tosylamidation of these
hydrocarbons and afforded the corresponding amides in 52-88% yields. For cyclohexane and toluene, the
tosylamidation products were formed in poor yields (ca. 10%). Kinetic studies on the reactions between [RuVI-
(TPP)(NTs)2] and nine hydrocarbons (cumene, ethylbenzene, cyclohexene, and para-substituted ethylbenzenes)
gave the second-order rate constants (k2) in the range of (0.330 ( 0.008) × 10-3 to (16.5 ( 0.3) × 10-3 dm3
mol-1 s-1. These reactions exhibit a large primary deuterium isotope effect, with a kH/kD ratio of 11 for the
tosylamidation of ethylbenzene. In the case of para-substituted ethylbenzenes, both electron-donating and
-withdrawing substituents moderately promote the reaction. There is an excellent linear correlation between
log kR and a related carboradical parameter. On the basis of these observations, a mechanism involving the
rate-limiting formation of a carboradical intermediate is postulated.
The development of metal-mediated nitrogen atom transfer
alkanes,11 and allylic amination of alkenes12-16 has received
considerable attention in recent years. The best explored systems
to date are those employing (N-(p-tolylsulfonyl)imino)phenyl-
iodinane (PhIdNTs) as the nitrogen source. Although in most
cases it is proposed that metal imido species (MdNTs) are
crucial intermediates in these reactions, their mechanisms are
still poorly understood and remain an open issue. In contrast,
there are extensive mechanistic studies in the literature on metal-
mediated oxygen atom transfer reactions such as alkene epoxi-
dation.17
By analogy with a mechanism postulated for epoxidation by
PhIdO, Mansuy and co-workers proposed that the aziridination
(11) Breslow, R.; Gellman, S. H. J. Chem. Soc., Chem. Commun. 1982,
1400.
(12) (a) Breslow, R.; Kluttz, R. Q.; Khanna, P. L. Tetrahedron Lett. 1979,
3273. (b) Breslow, R.; Gellman, S. H. J. Am. Chem. Soc. 1983, 105, 6728.
(13) Mahy, J. P.; Bedi, G.; Battioni, P.; Mansuy, D. Tetrahedron Lett.
1988, 29, 1927.
(14) Albone, D. P.; Aujla, P. S.; Taylor, P. C. J. Org. Chem. 1998, 63,
9569.
reactions such as aziridination of alkenes,1-10 amination of
(1) Groves, J. T.; Takahashi, T. J. Am. Chem. Soc. 1983, 105, 2073.
(2) (a) Mansuy, D.; Mahy, J.-P.; Dureault, A.; Bedi, G.; Battioni, P. J.
Chem. Soc., Chem. Commun. 1984, 1161. (b) Mahy, J.-P.; Bedi, G.; Battioni,
P.; Mansuy, D. J. Chem. Soc., Perkin Trans. 2 1988, 1517.
(3) (a) Evans, D. A.; Faul, M. M.; Bilodeau, M. T. J. Org. Chem. 1991,
56, 6744. (b) Evans, D. A.; Woerpel, K. A.; Hinman, M. M.; Faul, M. M.
J. Am. Chem. Soc. 1991, 113, 726. (c) Evans, D. A.; Faul, M. M.; Bilodeau,
M. T.; Anderson, B. A.; Barnes, D. M. J. Am. Chem. Soc. 1993, 115, 5328.
(d) Evans, D. A.; Faul, M. M.; Bilodeau, M. T. J. Am. Chem. Soc. 1994,
116, 2742.
(4) (a) Li, Z.; Conser, K. R.; Jacobsen, E. N. J. Am. Chem. Soc. 1993,
115, 5326. (b) Li, Z.; Quan, R. W.; Jacobsen, E. N. J. Am. Chem. Soc.
1995, 117, 5889.
(5) O’Connor, K. J.; Wey, S.-J.; Burrows, C. J. Tetrahedron Lett. 1992,
33, 1001.
(6) Ando, T.; Minakata, S.; Ryu, I.; Komatsu, M. Tetrahedron Lett. 1998,
39, 309.
(7) Lai, T.-S.; Kwong, H.-L.; Che, C.-M.; Peng, S.-M. Chem. Commun.
1997, 2373.
(8) Noda, K.; Hosoya, N.; Irie, R.; Ito, Y.; Katsuki, T. Synlett 1993,
469.
(9) Mu¨ller, P.; Baud, C.; Jacquier, Y. Tetrahedron 1996, 52, 1543.
(10) Au, S.-M.; Fung, W.-H.; Cheng, M.-C.; Che, C.-M.; Peng, S.-M.
Chem. Commun. 1997, 1655.
(15) Au, S.-M.; Zhang, S.-B.; Fung, W.-H.; Yu, W.-Y.; Che, C.-M.;
Cheung, K.-K. Chem. Commun. 1998, 2677.
(16) Na¨geli, I.; Baud, C.; Bernardinelli, G.; Jacquier, Y.; Moran, M.;
Mu¨ller, P. HelV. Chim. Acta 1997, 80, 1087.
10.1021/ja9913481 CCC: $18.00 © 1999 American Chemical Society
Published on Web 09/18/1999

Bis(tosylimido)ruthenium(VI) Porphyrins
J. Am. Chem. Soc., Vol. 121, No. 39, 1999 9121
of alkenes by PhIdNTs catalyzed by iron porphyrins involves
addition of an FeVdNTs (or FeIVsNTs) species to the alkene
double bond, to form a carboradical intermediate, which yields
N-tosylaziridines via direct oxidative transfer of the NTs ligand
to the intermediate carbon-centered radical or via a four-
membered Fe-C-C-N metallocycle.2b However, since the
metal imido (FeVdNTs) species could not be isolated, it is hard
to assess this mechanism by conventional kinetic methods.
Scheme 1
The same problem is also encountered for the aziridination
of alkenes by PhIdNTs catalyzed by copper complexes,3,4 the
most efficient catalysts for the aziridination reactions which in
some cases also have high enantioselectivity.3c,4a It appears that
a concerted mechanism is operative, as proposed by Evans and
co-workers on the basis of product analysis and solvent effects.3d
Although the work by Jacobsen and co-workers strongly argues
for the intermediacy of a discrete CudNTs intermediate in such
systems,4b in no cases are such intermediates observed and
characterized.
Experimental Section
Materials. Dichloromethane (AR, Ajax) was sequentially washed
with concentrated sulfuric acid, deionized water, saturated sodium
bicarbonate, and again deionized water and then refluxed over calcium
chloride followed by distillation over calcium hydride. Methanol (GR,
Merck) and n-pentane (AR) were used as received. The following
reagents purchased from commercial vendors were further purified
before use: norbornene (purified by sublimation); trans-stilbene and
adamantane (both recrystallized from ethanol); cyclohexene, cyclo-
hexane, cis-cyclooctene, and styrene (all distilled from calcium hydride);
cis-stilbene, 4-methoxystyrene, 4-methylstyrene, 4-chlorostyrene, 4-flu-
orostyrene, 4-cyanostyrene, 4-acetoxystyrene, 4-trifluoromethylstyrene,
2,3-dimethyl-2-butene, ethylbenzene, cumene, R-methylstyrene, and
trans-â-methylstyrene (all purified by passing through a column of
activated alumina). The purity of the alkenes was checked by GC or
1H NMR analysis. cis-â-Methylstyrene,20 cis-â-deuteriostyrene,20 â-d2-
styrene,21a,b and R-deuteriostyrene21a,c were prepared according to the
literature procedures. Cyclohexene-d10 and ethylbenzene-d10 (both
purchased from Aldrich) and all the other commercially available
reagents or chemicals were used as received. The carbonylruthenium-
(II) porphyrins, [RuII(Por)(CO)(MeOH)] (1) (Por ) TPP, 1a; TTP, 1b;
4-Cl-TPP, 1c; 4-MeO-TPP, 1d; OEP, 1e),22 were prepared by the
standard method.23
While the putative metal imido intermediates in all the known
aziridination or amination reactions described above have never
been isolated, there are a plethora of metal imido complexes
which have been obtained in pure form.18 However, none of
these complexes are found to undergo aziridination reactions
with an alkene. To better understand the mechanism of alkene
aziridination involving metal imido intermediates, the quest for
a metal imido complex that is stable for isolation and yet
sufficiently reactive toward aziridination of alkenes is of
particular importance.
Recently, we preliminarily reported that treatment of carbo-
nylruthenium(II) porphyrins, [RuII(Por)(CO)(MeOH)], with
PhIdNTs readily affords bis(tosylimido)ruthenium(VI) porphy-
rins, [RuVI(Por)(NTs)2], and, strikingly, this type of high-valent
ruthenium imido complexes is reactive toward alkenes to give
aziridines in good yields.10 Herein we present a mechanistic
study on such reactions by employing conventional kinetic
methods. By analogy with the mechanisms proposed for the
alkene epoxidations by high-valent metal oxo species,17 a total
of five different pathways involving the intermediates I to V
shown in Scheme 1 should be considered in the case of alkene
aziridination. The ultimate goal of this work is to ascertain the
most probable pathway on the basis of kinetic results. In
addition, we have also found that the tosylimidoruthenium(VI)
porphyrins are reactive toward alkanes, leading to insertion of
the imido group into the C-H bonds. Mechanistic studies on
such amidation processes are also described. After completion
of this work, we were informed of the synthesis of [RuVI(TMP)-
(NTs)2] complex (TMP ) meso-tetrakis(2,4,6-trimethylphenyl)-
porphyrinato dianion) by Groves and co-workers.19
General Procedure for the Preparation of [RuVI(Por)(NTs)2] (2)
(Por ) TPP, 2a; TTP, 2b; 4-Cl-TPP, 2c; 4-MeO-TPP, 2d; OEP,
2e). A mixture of [RuII(Por)(MeOH)(CO)] (1) (0.05 mmol) and PhId
NTs (0.2 mmol) in dichloromethane (10 mL) was stirred for 5 min at
room temperature under a nitrogen atmosphere. Removal of the solvent
in vacuo followed by washing the residue with methanol gave a reddish-
violet solid. The solid was redissolved in dichloromethane (2 mL) and
chromatographed on a short column of neutral alumina with the same
solvent as the eluant. The first band was collected and the solvent was
removed in vacuo, affording complexes 2a-e in 60-74% yields.
[RuVI(TPP)(NTs)2] (2a): Anal. Calcd for C58H42N6O4RuS2: C,
66.21; H, 4.02; N, 7.99. Found: C, 66.11; H, 4.05; N, 8.02. UV/vis
(1.43 × 10-6 M, CH2Cl2): λmax/nm (log ꢀ) 421 (5.14), 536 (4.18), 568
(3.77). IR (Nujol, cm-1): “oxidation state marker” band 1016. 1H NMR
(300 MHz, CD2Cl2): Hâ 8.82 (8H, s), Ho(eq) 8.18 (8H, m), Hm(eq),
Hp(eq) 7.79 (12H, m); Hm(ax) 6.45 (4H, d, J ) 8.1 Hz), Ho(ax) 4.87
(4H, d, J ) 8.3 Hz), CH3(ax) 2.17 (6H, s). FAB mass spectrum: m/z
1051 [M]+, 882 [M - NTs]+, 727 [M - NTs - Ts]+, 713 [M - NTs
- NTs]+.
(17) (a) Samsel, E. G.; Srinivasan, K.; Kochi, J. K. J. Am. Chem. Soc.
1985, 107, 7606. (b) Groves, J. T.; Watanabe, Y. J. Am. Chem. Soc. 1986,
108, 507. (c) Castellino, A. J.; Bruice, T. C. J. Am. Chem. Soc. 1988, 110,
158. (d) Garrison, J. M.; Bruice, T. C. J. Am. Chem. Soc. 1989, 111, 191.
(e) Garrison, J. M.; Ostovic, D.; Bruice, T. C. J. Am. Chem. Soc. 1989,
111, 4960. (f) Ostovic, D.; Bruice, T. C. J. Am. Chem. Soc. 1989, 111,
6511. (g) Ostovic, D.; Bruice, T, C. Acc. Chem. Res. 1992, 25, 314. (h)
Arasasingham, R. D.; He, G. X.; Bruice, T. C. J. Am. Chem. Soc. 1993,
115, 7985. (i) Groves, J. T.; Gross, Z.; Stern, M. K. Inorg. Chem. 1994,
33, 5065. (j) Finney, N. S.; Pospisil, P. J.; Chang, S.; Palucki, M.; Konsler,
R. G.; Hansen, K. B.; Jacobsen, E. N. Angew. Chem., Int. Ed. Engl. 1997,
36, 1720. (k) Linde, C.; Arnold, M.; Norrby, P.; Åkermark, B. Angew.
Chem., Int. Ed. Engl. 1997, 36, 1723. (l) Linker, T. Angew. Chem., Int. Ed.
Engl. 1997, 36, 2060. (m) Palucki, M.; Finney, N. S.; Pospisil, P. J.; Gu¨ler,
M. L.; Ishida, T.; Jacobsen, E. N. J. Am. Chem. Soc. 1998, 120, 948.
(18) Wigley, D. E. Prog. Inorg. Chem. 1994, 42, 239.
(20) Lindler, H.; Dubuis, R. Org. Synth. 1973, 5, 880.
(21) (a) Overberger, G. H.; Saunders, J. H. Org. Synth. 1955, 3, 204. (b)
Hosokawa, T.; Ohta, T.; Kanagama, S.; Murahashi, S.-I. J. Org. Chem.
1987, 52, 1758. (c) Wesener, J. R.; Maskau, D.; Gu¨nther, H. J. Am. Chem.
Soc. 1985, 107, 7307.
(22) Abbreviations: TPP ) meso-tetraphenylporphyrinato dianion; TTP
) meso-tetrakis(p-tolyl)porphyrinato dianion; 4-Cl-TPP ) meso-tetrakis-
(p-chlorophenyl)porphyrinato dianion; 4-MeO-TPP ) meso-tetrakis(p-
methoxyphenyl)porphyrinato dianion; OEP ) octaethylporphyrinato dianion.
(23) Rillema, D. P.; Nagle, J. K.; Barringer, L. F., Jr.; Meyer, T. J. J.
Am. Chem. Soc. 1981, 103, 56.
(19) Groves, J. T. Personal communication.

9122 J. Am. Chem. Soc., Vol. 121, No. 39, 1999
Au et al.
[RuVI(TTP)(NTs)2] (2b): Anal. Calcd for C62H50N6O4RuS2: C,
67.19; H, 4.55; N, 7.58. Found: C, 67.38; H, 4.61; N, 7.42. UV/vis
(1.45 × 10-5 M, CH2Cl2): λmax/nm (log ꢀ) 423 (5.21), 538 (4.29), 571
(3.88). IR (Nujol, cm-1): “oxidation state marker” band 1016. 1H NMR
(300 MHz, CD2Cl2): Hâ 8.86 (8H,s), Ho(eq) 8.03 (8H, d, J ) 8.10
Hz), Hm(eq) 7.58 (8H, d, J ) 8.09 Hz), CH3(eq) 2.73 (12H, s); Hm(ax)
6.48 (4H, d, J ) 8.01 Hz), Ho(ax) 4.84 (4H, d, J ) 8.25 Hz), CH3(ax)
2.18 (6H, s). FAB mass spectrum: m/z 1107 [M]+, 938 [M - NTs]+,
783 [M - NTs - Ts]+, 769 [M - NTs - NTs]+.
[RuVI(4-Cl-TPP)(NTs)2] (2c): Anal. Calcd for C58H38N6Cl4O4-
RuS2: C, 58.54; H, 3.22; N, 7.06. Found: C, 58.60; H, 3.17; N, 7.11.
UV/vis (1.40 × 10-5 M, CH2Cl2): λmax/nm (log ꢀ) 422 (5.09), 536
(4.20), 569 (3.77). IR (Nujol, cm-1): “oxidation state marker” band
1016. 1H NMR (300 MHz, CD2Cl2): Hâ(eq) 8.81 (8H, s), Ho(eq) 8.02
(8H, d, J ) 8.3 Hz), Hm(eq) 7.73 (8H, d, J ) 8.3 Hz); Hm(ax) 6.40
(4H, d, J ) 8.2 Hz); Ho(ax) 4.73 (4H, d, J ) 8.31 Hz), CH3(ax) 2.1
(6H, s). FAB mass spectrum: m/z 1189 [M]+, 1020 [M - NTs]+, 865
[M - NTs - Ts]+, 851 [M - NTs - NTs]+.
[RuVI(4-MeO-TPP)(NTs)2] (2d): Anal. Calcd for C62H50N6O8-
RuS2: C, 63.52; H, 4.30; N, 7.17. Found: C, 63.61; H, 4.40; N, 7.15.
UV/vis (1.47 × 10-5 M, CH2Cl2): λmax/nm (log ꢀ) 426 (5.10), 540
(4.19), 575 (3.91). IR (Nujol, cm-1): “oxidation state marker” band
1019. 1H NMR (300 MHz, CD2Cl2): Hâ 8.88 (8H, s), Ho(eq) 8.07 (8H,
d, J ) 8.46 Hz), Hm(eq), 7.34 (8H, d, J ) 8.49 Hz), MeO(eq) 4.11
(12H, s); Hm(ax) 6.49 (4H, d, J ) 8.13 Hz), Ho(ax) 4.84 (4H, d, J )
8.19 Hz), CH3(ax) 2.18 (6H, s). FAB mass spectrum: m/z 1171 [M]+,
1002 [M - NTs]+, 847 [M - NTs - Ts]+, 833 [M - NTs - NTs]+.
(1H); the Ho(ax′) resonance was not located. FAB mass spectrum: m/z
951 [M]+, 883 [M - pz]+, 781 [M - NHTs]+, 713 [M - NHTs -
pz]+.
[RuIV(OEP)(NHTs)(pz)] (3b): Anal. Calcd for C46H55N7O2RuS‚
C5H12: C, 64.94; H, 7.16; N, 10.39. Found: C, 64.85; H, 7.20; N,
10.33. UV/vis (1.97 × 10-6 M, CH2Cl2): λmax/nm (log ꢀ) 343 (4.19),
396 (4.86), 509 (br) (3.80). IR (Nujol, cm-1): “oxidation state marker”
1
band 1020. H NMR (500 MHz, CD2Cl2): Hmeso 10.3 (4H), Ha(eq)
15.5 (8H), Hb(eq) 11.4 (8H), CH3(eq) 1.3 (24H); NHTs: Hm(ax) 10.5
(2H), Ho(ax) 5.6 (2H), CH3(ax) 7.9 (3H); the NH resonance was not
located; pz: Hm(ax′), Hm′(ax′) -24.3 (1H), -28.4 (1H); the Ho(ax′)
resonance was not located. FAB mass spectrum: m/z 871 [M]+, 803
[M - pz]+, 701 [M - NHTs]+, 634 [M - NHTs - pz]+.
Kinetic Studies on the Reaction of [RuVI(TPP)(NTs)2] (2a) with
Alkenes. Kinetic measurements were performed on a Hewlett-Packard
8452A Diode Array spectrophotometer interfaced with an IBM-
compatible PC and equipped with a Lauda RM6 circulating water bath
by using standard 1.0-cm quartz cuvettes. The temperature of solutions
during kinetic experiments was maintained to within (0.2 °C.
The rates of imido group transfer from 2a to alkenes were measured
by monitoring the decrease of absorbance of 2a at 536 nm under the
condition that the concentration of alkenes was at least 100-fold in
excess of the ruthenium complex. Pseudo-first-order rate constants (kobs
)
were determined by nonlinear least-squares fits of (Af - At) to time (t)
according to the following equation:
(Af - At) ) (Af - Ai) exp(-kobst)
[RuVI(OEP)(NTs)2] (2e): Anal. Calcd for C50H58N6O4RuS2: C,
61.77; H, 6.01; N, 8.64. Found: C, 61.68; H, 6.12; N, 8.61. UV/vis
(1.71 × 10-6 M, CH2Cl2): λmax/nm (log ꢀ) 406 (5.07), 520 (4.21), 551
(4.16). IR (Nujol, cm-1): “oxidation state marker” band 1018. 1H NMR
(300 MHz, CD2Cl2): Hmeso 10.0 (4H, s), CH2 (eq) 4.07 (16H, q, J )
7.65 Hz), CH3 (eq) 1.99 (24H, t, J ) 7.64 Hz); Ho(ax) 6.48 (4H, d, J
) 8.01 Hz), Hm(ax) 4.63 (4H, d, J ) 8.25 Hz), CH3(ax) 2.29 (6H, s).
FAB mass spectrum: m/z 972 [M]+, 803 [M - NTs]+, 648 [M - NTs
- Ts]+, 634 [M - NTs - NTs]+.
where Af and Ai are the final and initial absorbance, respectively, and
At is the absorbance measured at time t. Kinetic data over 4 half-lives
(t1/2) were used for the least-squares fitting. Second-order rate constants,
k2, were obtained from the linear fit of the kobs values to the
concentration of alkenes.
Activation enthalpy (∆H*) and entropy (∆S*) were obtained from
the slope and the intercept, respectively, of the ln(k2/T) vs (1/T) plot
on the basis of the Eyring equation,
ln(k2/T) ) ln(R/Nh) + ∆S*/R - ∆H*/RT
Reactions of [RuVI(Por)(NTs)2] (2) with Alkenes and Alkanes.
General Procedures. All the stoichiometric reactions were carried out
under an argon atmosphere by employing standard Schlenk techniques.
In a typical run, to a solution of substrate (1-2 mmol) in dichlo-
romethane (2 mL) containing pyrazole (2% w/w) in a 10-mL Schlenk
tube was added [RuVI(Por)(NTs)2] (2) (0.05 mmol). The mixture was
stirred for 3 h. After removal of the solvent in vacuo, a mixture of
n-pentane and diethyl ether (5:1 v/v) was added, resulting in formation
of a reddish-purple precipitate. The precipitate, which was characterized
to be [RuIV(Por)(NHTs)(pz)] (pz ) pyrazolate anion) (3) (see below),
was collected by filtration and washed with n-pentane (yield: ca. 75%).
An aliquot of the filtrate, containing organic products, was taken out
and mixed with 1-bromo-4-chlorobenzene as internal standard and then
analyzed by GC. In the cases where authentic samples of aziridine or
amine products were unavailable, the solvent of the filtrate was removed
in vacuo, and the residue was purified by chromatography and then
analyzed by 1H NMR. The organic products, identified by comparison
of their spectral data with those reported in the literature if available3d
or otherwise independently characterized by 1H NMR and high-
resolution mass spectroscopy along with elemental analyses (see
Supporting Information), were quantified with the addition of 1,1-
diphenylethene as an internal standard. The cis-:trans-aziridine ratio
in the aziridination of cis-â-deuteriostyrene was determined from the
integration ratio of the appropriate proton resonance: δ ) 2.90 and
2.45 ppm for cis- and trans-â-deuteriostyrene aziridine, respectively.
where N is the Avogadro’s number, R is the universal gas constant, h
is Planck’s constant, and T is the temperature in Kelvin (K). The Eyring
plots were fit by unweighted linear least-squares methods by using the
software package Origin (Microcal Software, Inc.). Errors in the derived
activation parameters are the errors of the straight-line fit.
X-ray Crystal Determination of 3b‚C5H12. Crystals of 3b were
obtained by a slow diffusion of n-pentane into a solution of 3b in
chloroform. A purple crystal having dimensions of 0.30 × 0.20 × 0.15
mm mounted on a glass fiber was used for data collection at 301 K on
a Rigaku AFC7R diffractometer with graphite monochromatized Mo
KR radiation (λ ) 0.71073 Å) using ω-2θ scans with a ω-scan angle
(0.73 + 0.35 tan θ)° at a scan speed of 4.0 deg min-1 (up to 6 scans
for reflection with I < 15 σ(I)). Unit cell dimensions were determined
based on 25 reflections in the 2θ range of 20.35-25.7°. Intensity data
(in the range of 2θmax ) 50°; h 0 to 11; k 0 to 21; l -33 to 33; 3
standard reflections measured after every 300 reflections showed decay
of 6.45%) were corrected for decay and for Lorentz and polarization
effects, and empirical absorption corrections based on the Ψ-scan of
five strong reflections (minimum and maximum transmission factors:
0.913 and 1.000, respectively). The space group was determined
uniquely based on systematic absences and the structure was solved
by Patterson methods and expanded by Fourier methods (PATTY24)
and refined by full-matrix least-squares using the software package
TeXsan25 on a Silicon Graphics Indy computer. One crystallographic
asymmetric unit consists of one formula unit. In the least-squares
refinement, 57 non-H atoms of the complex molecule were refined
Characterization of [RuIV(Por)(NHTs)(pz)] (3). [RuIV(TPP)-
(NHTs)(pz)] (3a): Anal. Calcd for C54H39N7O2RuS: C, 68.20; H, 4.13;
N, 10.31. Found: C, 68.40; H, 3.92; N, 10.40. UV/vis (1.56 × 10-6
M, CH2Cl2): λmax/nm (log ꢀ) 412 (5.07), 530 (3.96), 562 (br) (3.65).
IR (Nujol, cm-1): “oxidation state marker” band 1012. 1H NMR (300
MHz, CD2Cl2): Hâ -17.3 (8H, s), Ho(eq), Hm(eq), Hp(eq) 7.3, 6.9
(20H); NHTs: Hm(ax) 11.8 (2H), Ho(ax) 5.0, CH3(ax) 8.9 (3H); the
NH resonance was not located; pz: Hm(ax′), Hm′(ax′) -27.5 (1H), -29.4
(24) PATTY: Beurskens, P. T.; Admiraal, G.; Beurskens, G.; Bosman,
W. P.; Garcia-Granda, S.; Gould, R. O.; Smits, J. M. M.; Smykalla, C. The
DIRDIF program system, Technical Report of the Crystallography Labora-
tory; University of Nijmegen: The Netherlands, 1992.
(25) TeXsan: Crystal Structure Analysis Package; Molecular Structure
Corporation: The Woodlands, Texas, 1985 & 1992.

Bis(tosylimido)ruthenium(VI) Porphyrins
J. Am. Chem. Soc., Vol. 121, No. 39, 1999 9123
Scheme 2
anisotropically, but the 5 C atoms of the n-pentane solvent molecule
were refined isotropically. The 66 H atoms at calculated positions with
thermal parameters equal to 1.3 times that of the attached C atoms
were not refined. The final difference Fourier map was featureless, with
maximum positive and negative peaks of 1.00 and 0.40 e Å,
respectively.
1
Other Physical Measurements. H NMR spectra were recorded
on a Bruker DPX 300 or DRX 500 FT NMR spectrometer (300 or 500
MHz, respectively), and chemical shifts (δ, ppm) were reported relative
to tetramethylsilane (TMS). Ultraviolet and visible spectra were run
on a Perkin-Elmer Lambda 19 spectrophotometer. Infrared spectra were
recorded on a Bio-Rad FT-IR spectrometer (Nujol mulls). GC analyses
were performed on a HP 5890 Series II system equipped with a HP
5890A flame ionization detector and a HP 3395 integrator. Elemental
analyses were performed by Butterworth Laboratories Ltd. (Teddington,
Middlesex, U.K.). Fast Atom Bombardment (FAB) mass spectra were
recorded on a Finnigan MAT 95 mass spectrometer using 3-nitrobenzyl
alcohol as matrix. Electrospray mass spectra were measured on a
Finnigan LCQ quadruple ion trap mass spectrometer. Magnetic
susceptibility was determined by the Evans method.26
Figure 1. 1H NMR spectra (300 MHz) in CD2Cl2 of (a) [RuVI(TPP)-
(NTs)2] (2a) and (b) [RuVI(OEP)(NTs)2] (2e).
both prepared by a three-step method (Scheme 2), which,
however, has been found to be applicable to the meso-
tetraarylporphyrinato ligands only, and has not led to the
isolation of bis(alkyl- or arylimido)ruthenium(VI) porphyrins.
ii. Characterization. As usual for d2 ruthenium species, all
the complexes 2a-e are diamagnetic, which is revealed by their
1H NMR spectra showing sharp peaks at normal fields. Two
typical spectra of 2a and 2e are given in Figures 1a and 1b,
respectively. The Hâ chemical shifts of 2a-d (8.81-8.88 ppm)
are close to those observed for oxo(imido)ruthenium(VI) (ca.
8.9 ppm)28 and dioxoruthenium(VI) (ca. 9.1 ppm)27e complexes
with the same tetraarylporphyrinato ligands. The Hmeso chemical
shift of 2e (10.0 ppm) is also comparable to that of its dioxo
analogue (10.58 ppm).27d The ortho and meta protons of the
axial tosyl groups of 2a-e appear as doublets at ca. 4.8 and
6.5 ppm, respectively, which are at considerably higher fields
than those of free PhIdNTs, probably due to the porphyrin ring
current effect. In all cases the integration ratios are consistent
with the bis(tosylimido) formulation of complexes 2. Further-
more, the whole spectrum of each of 2a-d is extremely similar
to that of [OsVI(Por)(NTs)2]30 with the same porphyrinato ligand
(the structure of [OsVI(TPP)(NTs)2] has been determined by
X-ray crystallographic study30).
Results and Discussion
Bis(tosylimido)ruthenium(VI) Porphyrins [RuVI(Por)-
(NTs)2] (2). i. Synthesis. Treatment of [RuII(Por)(CO)(MeOH)]
(1) with slightly excess PhIdNTs in dichloromethane readily
afforded the bis(tosylimido)ruthenium(VI) complexes 2 (Scheme
2). The reaction is rather analogous to the oxidation of 1 with
excess m-CPBA (m-chloroperoxybenzoic acid) to give the
dioxoruthenium(VI) complexes [RuVI(Por)O2].27 By employing
a series of sterically unencumbered meso-tetraarylporphyrinato
and octaethylporphyrinato ligands, complexes 2a-e were
isolated in about 65% yields, demonstrating a good generality
of this synthetic route. Very recently, we were informed that
Groves and co-workers had prepared, through similar reaction,
a bis(tosylimido)ruthenium(VI) complex with the TMP mac-
rocycle, a typical sterically encumbered porphyrinato ligand.19
In contrast to the “one-pot” synthesis of 2, the previously
reported alkyl-28 and arylimido29 ruthenium porphyrins were
The para substituents on the meso-phenyl groups have little
influence on the chemical shifts of the Hâ, Ho(eq), and the axial
tosyl protons among the imido complexes 2a-d. However, the
Hm(eq) chemical shifts are relatively sensitive to such para
substituents; for example, the Hm(eq) signal of the complex with
chloro substituents (2c: 7.73 ppm) is considerably downfield
from the corresponding signals of the complexes with electron-
donating methyl (2b: 7.58 ppm) or methoxy (2d: 7.34 ppm)
substituents.
(26) Evans, D. F. J. Chem. Soc. 1959, 2003.
(27) (a) Groves, J. T.; Quinn, R. Inorg. Chem. 1984, 23, 3844. (b) Groves,
J. T.; Quinn, R. J. Am. Chem. Soc. 1985, 107, 5790. (c) Groves, J. T.; Ahn,
K.-H. Inorg. Chem. 1987, 26, 3831. (d) Leung, W.-H.; Che, C.-M. J. Am.
Chem. Soc. 1989, 111, 8812. (e) Ho, C.; Leung, W.-H.; Che, C.-M. J. Chem.
Soc., Dalton Trans. 1991, 2933.
(28) Huang, J.-S.; Che, C.-M.; Poon, C.-K. J. Chem. Soc., Chem.
Commun. 1992, 161.
(29) Leung, W.-H.; Hun, T. S. M.; Hou, H.-W.; Wong, K.-Y. J. Chem.
Soc., Dalton Trans. 1997, 237.
(30) Au, S.-M.; Fung, W.-H.; Huang, J.-S.; Cheung, K.-K.; Che, C.-M.
Inorg. Chem. 1998, 37, 6564.

9124 J. Am. Chem. Soc., Vol. 121, No. 39, 1999
Au et al.
Figure 2. UV-visible spectra in dichloromethane of (a) [RuII(TPP)-
(CO)(MeOH)] (9.3 × 10-6 M) and (b) after its complete conversion to
[RuVI(TPP)(NTs)2] (2a).
Figure 4. Positive-ion electrospray mass spectrum of [RuVI(OEP)-
(NTs)2] (2e) in dichloromethane. In the insets are shown (a) the
observed and (b) theoretical isotope patterns of the parent ion and the
fragment [Ru(OEP)N]+.
Figure 3. UV-visible spectra in dichloromethane of (a) [RuII(OEP)-
(CO)(MeOH)] and (b) [RuVI(OEP)(NTs)2] (2e).
group. Even by employing electrospray mass spectrometry,
which utilizes a considerably softer ion source, such fragmenta-
tion was still significant, although its extent was greatly reduced.
Figure 4 shows the positive-ion electrospray mass spectrum of
2e as an example. As is obvious from this figure, the most
intense peak at m/z 972 corresponds to the parent ion of 2e.
The peak of the fragment [Ru(OEP)N]+ at m/z 648 becomes
weaker but still intense. There are also a few weak peaks in
Figure 4, one of which, with m/z 804, might be ascribed to the
monoimido species {[Ru(OEP)(NTs)] + H}+ or the amido
species [Ru(OEP)(NHTs)]+. The facile formation of nitrido-
ruthenium porphyrins [Ru(Por)N]+ in the gas phase should be
very interesting. This suggests a considerable thermal stability
of such species, although their isolation has not been realized
so far.
iii. Stability and Reactivity. The bis(tosylimido) complexes
2a-e are stable for a few days at ca. -15 °C in the solid state.
In highly purified dichloromethane solutions under an inert
atmosphere, they could exist for several hours without detectable
decomposition. However, upon prolonged standing, 2 became
unstable in the solution and gradually degraded into some
unknown products. Therefore, we are unable to obtain, despite
extensive attempts, diffraction-quality single crystals for 2.
Notably, all these complexes exhibit remarkable stability toward
moisture. For example, addition of two drops of water to a
solution of 2a in CD2Cl2 in an NMR tube caused no change in
the proton resonances of 2a within a period of ca. 5 h.
All the complexes 2 react rapidly with PPh3 in dichlo-
romethane at ambient conditions. However, conversion of 2 to
mono(tosylimido)ruthenium(IV) porphyrins, [RuIV(Por)(NTs)],
UV/vis measurements reveal that conversion of [RuII(Por)-
(CO)(MeOH)] (1) to [RuVI(Por)(NTs)2] (2) causes a red shift
of the Soret band, as is evident from Figures 2 and 3. This is
similar to the conversion from 1 to the dioxo complexes [RuVI-
(Por)O2].27 However, while the latter conversion appreciably
blue shifts the â band, an obvious red shift of this band occurs
in the case of the former conversion. On the other hand, the
spectra of 2 are considerably different from their osmium
analogues, [OsVI(Por)(NTs)2].30 This is not surprising, since there
is also a large difference between the spectra of [RuVI(Por)O2]
and [OsVI(Por)O2].31
The IR spectra of 2 each exhibit an intense band in the range
of 909-920 cm-1, which might correspond to the ruthenium-
imido stretching vibrations. These frequencies are lower than
those of the osmium-imido bands (950-972 cm-1) found for
[OsVI(Por)(NTs)2].30 The “oxidation state marker” bands for
2a-d appear in the range of 1016-1019 cm-1, similar to those
of other ruthenium(VI) porphyrins.28 The substituents (H, Me,
Cl, or MeO) on the meso-phenyl rings again have only minor
effects on the position of the “oxidation state marker” band.
The FAB+ mass spectra of 2a-e show the cluster peak
attributable to the parent ion, albeit in a weak intensity.
Surprisingly, the most prominent peak in each of the spectra
corresponds to the fragment [Ru(Por)N]+, identified on the basis
of the excellent matching of the observed and simulated isotope
distributions, apparently due to the loss of a tosylimido group
and the cleavage of the N-Ts bond in the remaining imido
(31) Che, C.-M.; Cheung, W.-C.; Lai, T.-F. Inorg. Chem. 1988, 27, 2801.

Bis(tosylimido)ruthenium(VI) Porphyrins
J. Am. Chem. Soc., Vol. 121, No. 39, 1999 9125
by treatment with an equimolar amount of PPh3 seems difficult.
In situ 1H NMR measurements revealed that reaction of 2a with
1 equiv of PPh3 afforded a mixture of several products within
5 min, and at least two such products are paramagnetic
ruthenium porphyrins (Hâ: ca. -16 and -24 ppm, respectively).
Attempts to isolate and characterize both species were unsuc-
cessful. Complexes 2 are also reactive toward alkenes and
alkanes in dichloromethane. But again, such reactions are rather
complicated, as indicated by the absence of clean isosbestic
points when monitored by UV/vis spectrophotometry, giving
rise to intractable ruthenium porphyrin products. In contrast, in
the presence of pyrazole (2% w/w), the elusive transient
ruthenium(IV) species could be effectively trapped by pyrazole
to form tosylamido complexes, [RuIV(Por)(NHTs)(pz)] (3),
which were isolated in about 75% yields. Especially, under these
conditions, the conversions from 2a to 3a by almost all the
alkenes and alkanes used in this work appear to be smooth,
without accumulation of any long-lived intermediates (see the
section on kinetic studies).
Charactrization of [RuIV(Por)(NHTs)(pz)] (3). i. Spectros-
copy. It has been reported that ruthenium(IV) porphyrins can
be either paramagnetic27c,d,29,32a,b or diamagnetic,32c-e depending
on the nature of the axial ligands. While the bis(diphenylamido)-
ruthenium(IV) porphyrin, [RuIV(3,4,5-MeO-TPP)(NPh2)2] (3,4,5-
MeO-TPP ) meso-tetrakis(3,4,5-trimethoxyphenyl)porphyrinato
dianion), is found to be diamagnetic,32e both the tosylamido
complexes 3 are paramagnetic. Magnetic susceptibility measure-
ments for 3 by the Evans method gave µeff of ca. 3.2 µB at 300
K, consistent with two unpaired electrons in the ground state
of the complexes.
The 1H NMR spectra of 3a and 3b measured at 293 K feature
paramagnetically shifted signals, as shown in Figures 5a and
5b, respectively. On the basis of integration ratio and by
comparison of Figure 5a with Figure 5b, almost all the signals
could be reasonably assigned. In the case of 3a with the TPP
macrocycle, the Hâ proton resonances appear as a very broad
peak at -17.3 ppm, which is dramatically different from that
(8.82 ppm) of the diamagnetic species 2a. In contrast, the proton
resonances of the phenyl groups, Ho(eq), Hm(eq), and Hp(eq),
on the porphyrin ring shift upfield by about 1 ppm from those
of 2a. Surprisingly, although the axial ligands in 3a are not
identical, there is no appreciable splitting of the Ho(eq) or
Hm(eq) signals. Perhaps the magnetic nonequivalence of
Ho(eq) or Hm(eq) protons caused by the unsymmetrical axial
ligation is rather small and is not readily observable in this case.
Interestingly, the behavior of the OEP macrocycle in 3b toward
the paramagnetic ruthenium(IV) center is very different from
that of TPP described above. First, the Hmeso proton resonances
are located as a broad singlet at 10.3 ppm, only very slightly
downfield from that of Hmeso of 2e (10.0 ppm).33 Second, the
resonances of the methylene protons on the porphyrin ring,
which become diastereotropic due to the lack of mirror
Figure 5. 1H NMR spectra (500 Hz) in CD2Cl2 at 293 K of (a) [RuIV-
(TPP)(NHTs)(pz)] (3a) and (b) [RuIV(OEP)(NHTs)(pz)] (3b).
symmetry in the ring plane,34 clearly split into two broad singlets
at 15.5 and 11.4 ppm in a 1:1 ratio, although the separation of
4.1 ppm between these two singlets is considerably smaller than
that (>10 ppm) found for the paramagnetic [RuII(OEP)]2.34
Despite the above differences between TPP and OEP mac-
rocycles, the proton resonances of the axial ligands in 3a and
3b are fairly similar in both their chemical shifts and line shapes.
For both complexes, the Hm(ax), CH3(ax), and Ho(ax) signals
of the tosylamido axial ligands appear at ca. 11, 8, and 5 ppm,
respectively, whereas the Hm(ax′) and Hm′(ax′) signals of
pyrazolate axial ligands appear at ca. -26 and -28 ppm. The
resonances of Ho(ax′) of pyrazolate axial ligands were not
located, which might be too broad to be observed owing to their
short distances to the paramagnetic center.
The “oxidation state marker” band in the IR spectrum of 3a
is located at 1012 cm-1, identical with that observed for [RuIV-
(3,4,5-MeO-TPP)(NPh2)2] (1012 cm-1).32e As compared with
the UV/vis spectrum of 2a, 3a exhibits appreciably blue-shifted
Soret and â bands. Such is also the case for 3b vs 2e. The Soret
band of 3a at 412 nm is similar to those reported for the
arylimidoruthenium(IV) porphyrins.29 The formulation of 3 as
a ruthenium porphyrin bearing both tosylamido and pyrazolato
axial ligands is also inferred by their FAB mass spectra, which
exhibit the cluster peak attributable to the desired parent ion,
and their elemental analyses. The structure of 3a has been
determined by X-ray crystal analysis and reported in the previous
communication.10
(32) (a) Ke, M.; Sishta, C.; James, B. R.; Dolphin, D.; Sparapany, J.
W.; Ibers, J. A. Inorg. Chem. 1991, 30, 4766. (b) Cheng, S. Y. S.; Rajapakse,
N.; Rettig, S. J.; James, B. R. J. Chem. Soc., Chem. Commun. 1994, 2669.
(c) Collman, J. P.; Brothers, P. J.; McElwee-White, L.; Rose, E. J. Am.
Chem. Soc. 1985, 107, 6110. (d) Collman, J. P.; Brothers, P. J.; McElwee-
White, L.; Rose, E.; Wright, L. J. J. Am. Chem. Soc. 1985, 107, 4570. (e)
Huang, J.-S.; Che, C.-M.; Li, Z.-Y.; Poon, C.-K. Inorg. Chem. 1992, 31,
1315.
(33) We noted that the Hmeso chemical shifts of paramagnetic ruthenium-
(IV) porphyrins could in some cases be very similar to those of related
diamagnetic complexes. For example, the Hmeso chemical shifts of the
paramagnetic [RuIV(OEP)X2] (X ) Br, Cl, F) were found to be 3.50, 8.89,
and 9.63 ppm, respectively,where the last is very similar to the Hmeso
chemical shift of the diamagnetic [RuIV(OEP)Ph2] (9.68 ppm). See: Sishta,
C.; Ke, M.; James, B. R.; Dolphin, D. J. Chem. Soc., Chem. Commun. 1986,
787.
(34) Collman, J. P.; Barnes, C. E.; Swepston, P. N.; Ibers, J. A. J. Am.
Chem. Soc. 1984, 106, 3500.

9126 J. Am. Chem. Soc., Vol. 121, No. 39, 1999
Au et al.
7a). The ruthenium(IV) ion is essentially in the porphyrin ring
plane, with Ru-N(Por) distances ranging from 2.035(6) to
2.056(5) Å. As is evident from Figure 7b, the S1, N5, Ru1,
N2, and N4 atoms are basically coplanar. However, the
pyrazolate ring rotates from this plane by ca. 17°. The Ru1-
N5 and Ru1-N6 distance of 2.020(6) and 2.085(6) Å, respec-
tively, and the Ru-N5-S1 angle of 135.6(3)° are all similar
to the corresponding distances or angle found in 3a.10
Kinetic Studies on the Reactions of [RuVI(TPP)(NTs)2] (2a)
with Alkenes. The UV/vis spectral changes for the reaction of
2a with excess styrene in dichloromethane containing pyrazole
(2% w/w) are shown in Figure 8, which reveals a decrease in
absorbance of the Soret band of 2a and a concomitant increase
in absorbance of a new Soret band at 412 nm. The final spectrum
is virtually identical with that of the amido complex 3a. The
appearance of isosbestic points at 446, 518, and 552 nm suggests
that the conversion of 2a to 3a is a simple kinetic process
without accumulation of any long-lived intermediates. A similar
phenomenon has also been observed for the reactions of 2a with
all the other alkenes, except cis-/trans-stilbenes, used in the
kinetic studies.
Figure 6. ORTEP drawing of [RuIV(OEP)(NHTs)(pz)] (3b)‚C5H12 with
the atom-numbering scheme. Hydrogen atoms and the solvent molecule
are omitted for clarity. Thermal ellipsoids are drawn at the 35%
probability level.
Table 1. Structure Determination Summary
empirical formula
formula weight
cryst syst
space group
a, Å
b, Å
c, Å
C46H55N7O2SRu‚C5H12
943.27
monoclinic
P21/c (No. 14)
9.847(4)
17.938(6)
28.405(6)
96.55(3)
4984(2)
4
Under similar conditions, the reaction of 2e, containing OEP
macrocycle, with styrene in the presence of pyrazole is
somewhat different. The UV/vis spectral changes show isos-
bestic points at 378, 438, 510, 527, 538, and 565 nm at the
early stage but become more complicated as the reaction
proceeds to completion. Again, the final spectrum is virtually
identical with that of 3b. It seems that the conversion of 2e to
3b involves at least two distinct steps.
â, deg
V, Å3
Z
F(000)
1992
The reaction of 2a with alkenes follows the first-order rate
law under the pseudo-first-order conditions: [alkene] (0.2-1.8
M) . [2a] (1.6 × 10-5 M). In all cases, kobs is linearly
dependent on [alkene], and the slopes of such plots give the
second-order rate constants (k2). The absence of kinetic satura-
tion (even at high concentration of alkenes) indicates that a prior
coordination of alkene to 2a, e.g. via a π-complex, is either
absent or unimportant. Values of k2, so determined, are listed
in Table 3. In most cases, the k2 values are larger than those
reported for the epoxidation of the same alkenes by [RuVI(TPP)-
O2].27e It should be noted that, in the absence of alkenes, 2a
exhibited no appreciable reaction with pyrazole under the same
conditions on the time scale of the kinetic studies.
T, K
301
1.257
4.01
h, k, (l
9664
9100
4908
534
0.054
0.081
2.64
density (calcd), g/cm3
abs coeff, cm-1
data collected
no. of reflns collected
no. of independent reflns
no. of obsd reflns (I > 3σ(I))
no. of params, P
Ra
b
Rw
goodness-of-fit
(∆/σ)max
0.05
a R ) ∑||Fo| - |Fc||/∑|Fo|. b Rw ) [∑w||Fo| - |Fc|| /∑w|Fo| ]1/2
.
2
2
Table 2. Selected Bond Distances (Å) and Angles (deg) for
[RuIV(OEP)(NHTs)(pz)] (3b)‚C5H12
Examination of Table 3 reveals that both electronic and steric
effects are operating in the imido group transfer reactions
between 2a and alkenes. For example, the k2 values observed
for methylstyrenes follow the order p-methylstyrene > styrene
> cis-â-methylstyrene > trans-â-methylstyrene > R-methyl-
styrene. This order cannot be rationalized on the basis of the
nucleophilicity of the alkenes alone. Among all these methyl-
styrenes, p-methylstyrene should have the least steric hindrance
because of the remote location of the methyl group from the
reaction site. Therefore, the larger rate of the imido group
transfer to this substrate than to styrene almost entirely results
from an electronic effect. This suggests that electron-donating
substituents on the alkenes favor such transfer processes, and
the attack by 2a toward alkenes is electrophilic. In the cases of
other methylstyrenes, the methyl group is directly or closely
linked to the reaction site, which renders them even more
nucleophilic, but also much more sterically demanding, than
p-methylstyrene. Here, the steric effect should be responsible
for the smaller k2 values observed for these substrates.
Ru1-N1
Ru1-N3
Ru1-N5
N5-S1
2.040(5)
2.035(6)
2.020(6)
1.587(6)
1.426(5)
Ru1-N2
Ru1-N4
Ru1-N6
S1-O1
2.056(5)
2.056(5)
2.085(6)
1.451(6)
1.802(8)
S1-O2
S1-C21
N5-Ru1-N6
N1-Ru1-N2
N3-Ru1-N2
N5-Ru1-N1
N5-Ru1-N3
N6-Ru1-N1
N6-Ru1-N3
173.1(2)
90.1(2)
90.1(2)
92.4(2)
91.3(2)
88.1(2)
88.2(2)
Ru1-N5-S1
N1-Ru1-N4
N3-Ru1-N4
N5-Ru1-N2
N5-Ru1-N4
N6-Ru1-N2
N6-Ru1-N4
135.6(3)
89.9(2)
89.8(2)
84.4(2)
97.1(2)
88.7(2)
89.7(2)
ii. X-ray Crystal Structure of 3b‚C5H12. The ORTEP
drawing of 3b, together with the atom-numbering scheme, is
depicted in Figure 6. Crystallographic data for the structure
determination and selected bond distances and angles are listed
in Tables 1 and 2, respectively. The porphyrin ring is basically
planar, whose component atoms have a mean deviation of
0.0277 Å from the least-squares plane. The eight ethyl groups
on the porphyrin ring assume an aaaabbbb conformation (Figure
Studies of the effect of temperature on the kobs values for the
imido group transfer reactions between 2a and two representa-
tive alkenes afforded linear Eyring plots over a temperature

Bis(tosylimido)ruthenium(VI) Porphyrins
J. Am. Chem. Soc., Vol. 121, No. 39, 1999 9127
Figure 7. Ball and stick drawing of 3b showing (a) the aaaabbbb conformation of the eight ethyl groups on the porphyrin ring and (b) the
orientations of tosylamido and pyrazolate axial ligands with respect to the porphyrinato ligand. The insets a and b depict the core structures in (a)
and (b), respectively.
Table 3. Second-Order Rate Constants (k2) for the Imido Group
Transfer Reactions between Alkenes and [RuVI(TPP)(NTs)2] (2a) in
Dichloromethane Containing Pyrazole (2% w/w) at 298 K
entry
alkene
103k2/dm3 mol-1 s-1
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
cis-cyclooctene
norbornene
2,3-dimethyl-2-butene
styrene
2.80 ( 0.07
5.1 ( 0.1
17.0 ( 0.4
9.0 ( 0.1
90 ( 4
p-methoxystyrene
p-methylstyrene
p-fluorostyrene
p-chlorostyrene
p-acetoxystyrene
p-trifluoromethylstyrene
p-cyanostyrene
R-methylstyrene
trans-â-methylstyrene
cis-â-methylstyrene
R-d1-styrene
16.0 ( 0.8
11 ( 1
8.8 ( 0.1
5.4 ( 0.1
3.2 ( 0.2
1.60 ( 0.06
4.8 ( 0.1
5.2 ( 0.2
6.8 ( 0.1
9.3 ( 0.2
10.6 ( 0.2
Figure 8. UV-visible spectral changes during the reaction of [RuVI-
(TPP)(NTs)2] (2a) (1.6 × 10-5 M) with styrene (0.19 M) in dichlo-
romethane containing pyrazole (2% w/w) at 298 K. Scans were taken
at 30-s intervals.
â-d2-styrene
Table 4. Activation Parameters for the Imido Group Transfer
Reactions between Alkenes and [RuVI(TPP)(NTs)2] (2a)
range of 288-318 K. The activation enthalpies ∆H* and
entropies ∆S* for cis-cyclooctene and styrene by 2a are listed
in Table 4. The large negative ∆S* values in both cases suggest
an association of reactants in the transition states.
alkene
∆H*/kcal mol-1
∆S*/eu
cis-cyclooctene
styrene
cumene
ethylbenzene
cyclohexene
10.0 ( 0.3
5.1 ( 0.2
10.2 ( 0.2
8.9 ( 0.2
7.2 ( 0.3
-(36.5 ( 1.7)
-(50.7 ( 3.2)
-(37.5 ( 1.6)
-(39.8 ( 2.0)
-(43.9 ( 2.3)
Stoichiometric Reactions of [RuVI(Por)(NTs)2] (2) with
Alkenes. i. Product Analysis. The stoichiometric reactions were
carried out by using higher concentrations of 2 and alkene than
those used for the kinetic studies to increase the concentrations
of the products. The product yields, based on the amount of 2
used, for different alkenes are summarized in Table 5. In the
cases of aziridination of styrene by 2a-d, the product yields
range from 75 to 80%, which are rather insensitive to the nature
of para substituents on the meso-phenyl groups of the porphy-
rinato ligand. In contrast, reaction of 2e with styrene afforded
N-(p-tolylsulfonyl)-2-phenylaziridine in 85% yield. The im-
proved yield in this case may be ascribed to the relatively open
reaction site in the complex bearing OEP ligand.
of aziridine and TsNH2 is close to 100%, a phenomenon similar
to that observed previously for the imido group transfer reactions
catalyzed by iron or manganese porphyrins.2 Moreover, we
found that the N-tosylaziridine:TsNH2 ratio can be markedly
increased in anhydrous conditions. Unlike the imido group
transfer of PhIdNTs to aromatic alkenes catalyzed by iron
porphyrins,2 only trace amounts of epoxide and benzaldehyde
were detected except for the aziridination of trans- and cis-
stilbene. The reaction of 2a with trans-stilbene gave trans-N-
(p-tolylsulfonyl)-2,3-diphenylaziridine in low yield. No cis-
aziridine was detected. The low product yields from the
It is noteworthy that the reactions of 2 with alkenes also
afforded tosylamide (TsNH2), and in most cases, a mass balance

9128 J. Am. Chem. Soc., Vol. 121, No. 39, 1999
Au et al.
Table 5. Stoichiometric Imido Group Transfer Reactions between
Alkenes and [RuVI(TPP)(NTs)2] (2a) or [RuVI(OEP)(NTs)2] (2e) in
Dichloromethane Containing Pyrazole (2% w/w)
Figure 9. Plot of the log of the second-order rate constant (k2) for the
reaction of [RuVI(TPP)(NTs)2] (2a) with a series of eight alkenes vs
the 1e oxidation potential (E1/2) of the alkenes.
remained unchanged and the corresponding trans-aziridine was
not detected.
Evidently, the aziridination of cis-alkenes by 2a results in
only a partial loss of the alkene stereochemistry, in contrast to
the iron and manganese porphyrin-catalyzed aziridination of cis-
stilbene by PhIdNTs, in which a complete loss of alkene
stereochemistry was observed.2 The imido group transfer
reactions between 2a and cis-alkenes may involve both con-
certed and stepwise pathways. However, the clean second-order
kinetics observed for the reactions between 2a and styrene or
cis-â-methylstyrene is incompatible with the operation of
multiple reaction pathways. Inasmuch as the loss of alkene
stereochemistry is hard to reconcile with a concerted mechanism,
a stepwise pathway analogous to that proposed by Mansuy and
co-workers2b should be operative for the aziridination reactions
by 2a. The cis:trans ratio 14:86 for the aziridination of cis-
stilbene is low, perhaps due to a severe steric interaction between
the cis-phenyl groups in the proposed acyclic intermedieate,
which will greatly favor the C-C bond rotation to form the
trans-aziridine.
The Nature of the Rate-Determining Step. i. The Cor-
relation between k2 and E1/2. To get further information on
the nature of the transition state at the rate-determining step,
we examined the correlation between the second-order rate
constant (k2) of the reaction of 2a with an alkene and the 1e
oxidation potential (E1/2) of the alkene for all the alkenes in
Table 3 whose E1/2 values are available.17e The log k2 vs E1/2
plot is shown in Figure 9, which exhibits a good linearity with
k2 and E1/2 ranging from 1.6 × 10-3 to 9.0 × 10-2 M-1 s-1
and 1.15-2.05 V, respectively. It has been established that, for
a reaction proceeding via an initial single electron transfer and
a Isolated yield of aziridine based on 2a and 2e used. b Ratios of
trans:cis N-tosylaziridine were determined by 300-MHz 1H NMR
spectroscopy.
aziridination of trans-and cis-stilbene by 2a probably reflect a
lower reactivity of 2a toward both of the alkenes. However, of
significance are notable amounts of trans-stilbene oxide found
in the reaction of 2a with either trans- or cis-stilbene.
ii. Reaction Stereospecificity. The stereospecificity in metal-
mediated aziridination of cis-alkenes provides important infor-
mation on the mechanism of such a process. Mansuy and co-
workers reported that aziridination of cis-stilbene by PhIdNTs
catalyzed by iron or manganese porphyrins exclusively afforded
the corresponding trans-aziridine, and the nonstereospecificity
of these reactions is consistent with a stepwise mechanism, i.e.,
the formation of an acyclic intermediate that is sufficiently long-
lived to undergo the cis-trans isomerization through rotation
about the related C-C single bond.2b In contrast, the stereospe-
cific CuClO4-catalyzed aziridination by PhIdNTs of unfunc-
tionalized alkyl-substituted alkenes such as cis-4-octene reported
by Evans and co-workers argued for a concerted mechanism.3d
We have found that aziridinations of cis-alkenes by 2a are
generally nonstereospecific, with the formation of a mixture of
cis- and trans-aziridines in ratios of 14:86, 25:75, and 50:50
showing a linear free-energy correlation between log k2 and E1/2
,
the slope of the log k2 vs E1/2 plot should be (i) zero if the
diffusion together of the reactants is rate limiting, (ii) -16.6
V-1 if the 1e transfer is extremely endothermic, and (iii) in the
range of 0 to -16.6 V-1 in other cases.35 The slope in Figure
9 was determined to be -1.7 V-1, corresponding to a slope of
R ) 0.1 in the conventional Bronsted plot (log k2 vs log Keq).
Obviously, there is little charge transfer from the alkenes to 2a
in the transition state. Therefore, a mechanism involving rate-
limiting 1e tranfer to form an alkene-derived π-cation radical
(III) can be dismissed. For comparison, the slope of the log k2
1
(based on H NMR analyses) for aziridination of cis-stilbene,
cis-â-methylstyrene, and cis-â-deuteriostyrene, respectively.
These ratios were reproducible by repeating the reactions for
at least three times. The purity of the cis-alkenes (>99%) had
been checked by GC prior to use, indicating the absence of
trans-alkenes. At the end of the reactions, trans-alkenes were
1
again not detected by either GC or H NMR. It is noteworthy
that such a cis-trans isomerization unlikely results from an
isomerization of aziridine catalyzed by the Ru(IV) product 3a.
For example, after stirring a mixture of pure cis-Ν-(p-tolylsul-
fonyl)-2-methyl-3-phenylaziridine and 3a in dichloromethane
for 3 h under identical conditions, the starting cis-aziridine
(35) (a) Rehm, D.; Weller, A. Isr. J. Chem. 1970, 8, 259. (b) Andrieux,
C. P.; Blocman, C.; Dumas-Bouchiat, J.-M.; Saveant, J.-M. J. Am. Chem.
Soc. 1979, 101, 3431.

Bis(tosylimido)ruthenium(VI) Porphyrins
J. Am. Chem. Soc., Vol. 121, No. 39, 1999 9129
•
Table 6. Variation of log kR with σp+, σmb, and σJJ Scales for the
Scheme 3
Aziridination of Para-Substituted Styrenes by [RuVI(TPP)(NTs)2]
(2a)
log kR(calcd)a
+
•
p-X
log k (exptl)
σp
σmb
σJJ
OMe
Me
F
1
0.25
0.087
0
-0.778 -0.77
-0.311 -0.29
0.233
0.15
0.928
0.382
0.241
-0.073 -0.24 -0.02
H
0
0
0
0
Cl
CF3
CN
-0.00976
-0.449
-0.75
0.114
0.53
0.66
0.11
0.49 -0.01
0.86 0.42
0.22
-0.00044
-0.519
-0.682
a log kR(calcd) ) -1.05σmb + 0.522σJJ
.
•
vs E1/2 plot was reported to be -2.99 V-1 for the epoxidation
of alkenes by [(Br8TPP)CrV(O)(ClO4)] involving the rate-
determining formation of a charge-transfer complex17e and
-0.89 V-1 for the epoxidation of alkenes by [(Cl8TPP)MnIV-
(O)] where neutral carbon-radical was formed.17h
The involvement of the azametallacyclobutane intermediate
(II), an analogue of metallaoxetane, in the rate-determining step
might be plausible. However, the following observations seem
to argue against the intermediacy of such species. Since the
UV/vis spectral changes for the reactions of 2a with all the
alkenes in Figure 9 exhibit clean isosbestic points, there should
be no accumulation of any long-lived intermediate during these
reactions. On the other hand, molecular modeling studies by
Bruice and co-workers revealed that, in the case of alkene
epoxidation by hypervalent metal-oxo porphyrins, the formation
of metallaoxetanes would result in severe steric interactions.17g
It is unlikely that replacing the oxo group by the bulkier
tosylimido group will ease the congested spatial environment.
Moreover, the formation of intermediate II hinges on a seven-
coordinate ruthenium center, which is almost unprecedented in
ruthenium porphyrin chemistry.
Figure 10. Linear-free-energy correlation of log kR vs σ+ for the
reactions of [RuVI(TPP)(NTs)2] (2a) with a series of para-subsituted
styrenes at 298 K.
with electron-donating substituents are more reactive than
styrene, whereas electron-withdrawing substituents retard the
imido group transfer process. Fitting (by the least-squares
method) the log kR data in Table 6 to σ+ scale results in good
linearity (R ) 0.98) (Figure 10) with a F+ value of -1.1,10
indicating a minor influence of the electronic effect of para
substituents on the k2 values. This again argues against the
involvement of the alkene-derived cation radical III. The
involvement of the carbocation intermediate IV is also unlikely,
since the rate-limiting formation of carbocations in the elec-
trophilic addition to the CdC bond has F+ values as negative
as -3.5 (hydration)37a and -4.1 (bromination).37b However, such
a small variation of the rate constants with para substituents
observed in this work might be expected for the rate-limiting
formation of the carboradical intermediate V.
ii. Isotope Effect. The measurement of secondary kinetic
deuterium isotope effect can provide information about the
hybridization changes occurring at an isotopically substituted
carbon atom as a set of reactants pass from their ground states
to the transition states.36 In the case of alkene aziridination, two
sp2 carbons become more sp3-like in the aziridine product. If
either intermediate I or II is involved in the reaction, then a
simultaneous rehybridization of both R and â carbon atoms of
alkenes, such as styrene, from sp2 to sp3 is required. However,
the isotope effect data obtained from the reactions of 2a with
styrene and deuteriostyrenes show a clear distinction between
the R and â carbons in the transition state: only the â-carbon
atom changes its hybridization from sp2 to sp3 (kH/kD ) 0.85),
while the R-carbon atom remains essentially sp2 hybridized (kH/
kD ) 0.97). The kH/kD value of 0.85 observed for the â-carbon
falls in the range of kH/kD values (0.82-0.95) found for the
same carbon atom from the competitive epoxidation of styrene
and cis-â-deuteriostyrene catalyzed by [Mn(salen)Cl].17m This
implies that in the transition state there is a substantial Câ-N
but negligible CR-N bond formation (Scheme 3), which serves
as an additional argument against both the concerted mechanism
involving intermediate I and the stepwise pathway that involves
intermediate II.
According to the above studies on isotope effect, for the
aziridination of para-substituted styrenes, the carboradical
intermediates V if really involved should contain a para-
substituted benzyl radical moiety, in which a significant spin-
delocalization might exist.38 Indeed, a careful examination of
the log kR vs σ+ plot in Figure 10 reveals some sizable deviations
of the log kR values (for the substituents OMe, Me, and CN)
from the single-parameter regression. After considering the spin-
•
delocalization effect by employing the σJJ scale recently
iii. Substitution Effect. The effect of para substituents on
the rates of the imido group transfer of 2a to styrene and its
derivatives has been investigated. The log’s of the relative rates
(kR) obtained for several para-substituted styrenes (p-Y-C6H4-
CHCH2, Y ) OMe, Me, F, H, Cl, CF3, and CN) are summarized
in Table 6. It is evident from this table that substituted styrenes
•
developed by Jiang and co-workers, the log kR vs (σmb, σJJ )
plot38 gives rise to an appreciably better linearity (R ) 0.99),
as shown in Figure 11. It is likely that the spin-delocalization
effect does exist. Multiple linear regression for the dual-
(37) (a) Schubert, W. M.; Keefe, J. R. J. Am. Chem. Soc. 1972, 94, 559.
(b) Yates, K.; McDonald, R. S.; Shapiro, S. A. J. Org. Chem. 1973, 38,
2460.
(36) Collin, C. J.; Bowman, N. S. Isotope Effects in Chemical Reactions;
Van Nostrand Reinhold: New York, NY, 1970.
(38) Jiang, X.-K. Acc. Chem. Res. 1997, 30, 283.

9130 J. Am. Chem. Soc., Vol. 121, No. 39, 1999
Au et al.
Table 7. Stoichiometric Tosylamidation of Hydrocarbons by
[RuVI(TPP)(NTs)2] (2a) and [RuVI(OEP)(NTs)2] (2e) in
Dichloromethane Containing Pyrazole (2% w/w)
•
Figure 11. Linear-free-energy correlation of log kR vs (σmb, σJJ ) for
the reactions of [RuVI(TPP)(NTs)2] (2a) with a series of para-subsituted
styrenes at 298 K.
Scheme 4
a Yield based on 2a and 2e used. b The sum of the yields of the
tosylamidation product and TsNH2 is about 100%.
adamantyl)-tosylamide in 52 and 60% yield, respectively. In
contrast, cyclohexane, which lacks a tertiary C-H bond, was
tosylamidated by 2a,e to give cyclohexylamide in poor yields
(about 10%), comparable to its tosylamidation by PhIdNTs
catalyzed by iron and manganese porphyrins.11 Interestingly,
the phenyl substitution dramatically improves the reactivity of
C-H bonds of an alkane toward complexes 2. Despite the lack
of a tertiary C-H bond, ethylbenzene was tosylamidated by
2a,e to give R-(N-tosylamino)ethylbenzene in yields as high as
80%. Listed in Table 7 are the product yields for the tosyl-
amidation of a variety of hydrocarbons by 2a,e under the same
conditions.
The UV/vis spectral change during the reaction of 2a with
ethylbenzene in dichloromethane containing pyrazole (2% w/w)
is depicted in Figure 12. The final product was identified to be
3a as aforementioned. Plots of kobs vs [hydrocarbon] are shown
in Figure 13. The second-order rate constants (k2) are sum-
marized in Table 8. The activation enthalpy ∆H* and entropy
∆S*, determined over the temperature range of 285-305 K,
for the tosylamidation of cumene, ethylbenzene, and cyclohex-
ene are listed in Table 4. The large negative ∆S* values suggest
an association of reactants in the transition states.
The Nature of the Rate-Determining Step. i. Selectivity
Ratio and Steric Effect. It can be seen from Table 8 that
tosylamidation of the tertiary C-H bond in cumene is consider-
ably slower than that of the secondary C-H bond in ethylben-
zene. On the basis of the ratio of the k2 values of cumene and
ethylbenzene (Table 8) and by taking the statistical factor into
account, a selectivity ratio, k[tertiary(C-H)]/k[secondary(C-
H)], of 0.84 was obtained, which is less than unity and
substantially smaller than that of 2.9 obtained for the oxidation
of the C-H bonds in the same substrates by [RuVI(TPP)O2].27e
Such an unusually small selectivity ratio might result from a
steric effect. Since the tosylimido group is much larger than
the oxo group, there might be a significant steric hindrance in
parameter equation38
•
•
log kR ) Fmbσmb + FJJ σJJ
•
gives the values of Fmb and FJJ of -1.05 and +0.52, respec-
tively. The positive FJJ• value reveals that all the para substituents
delocalize the spin in the transition state. On the other hand,
the negative Fmb suggests that the attacks toward the alkenes
•
are all electrophilic. In addition, the |Fmb/FJJ | ratio of 2.02, which
is considerably larger than unity, indicates that the polar effect
is dominant in the imido group transfer reactions.38
We, therefore, postulate that the alkene aziridination by 2a
should occur via the rate-limiting formation of the carboradical
intermediate V (Scheme 4). This type of intermediates could
undergo C-C bond rotation prior to the ring closure to produce
aziridines. Analogous carboradical intermediates have also been
proposed in the aziridination of alkenes by PhIdNTs catalyzed
by iron porphyrins albeit without kinetic evidence,2b and also
in the epoxidation of alkenes by d2 dioxoruthenium(VI)
porphyrins.27e Since the imido complex 2a exhibits no ap-
preciable reaction with water at room temperature within several
hours, as described above, the tosylamide byproduct in the
reactions between 2a and alkenes should not result from a
prehydrolysis of 2a but may arise from the reaction of the
carboradical species with protic reagents.
Tosylamidation of C-H Bonds. Imido group transfer
reactions of complexes 2 have been observed not only with
alkenes but also with alkanes, leading to formation of tosyla-
mides as a result of insertion of the tosylimido group into the
C-H bonds. In the case of a saturated alkane, tosylamidation
is selectively directed toward the tertiary C-H bonds. For
example, the reactions of 2a,e with adamantane afforded N-(1-

Bis(tosylimido)ruthenium(VI) Porphyrins
J. Am. Chem. Soc., Vol. 121, No. 39, 1999 9131
Table 9. Variation of log kR with TE for the Tosylamidation of
Para-Substituted Ethylbenzenes by [RuVI(TPP)(NTs)2] (2a)
p-X
TE
log kR
OMe
Me
Cl
F
H
1.08
0.42
0.23
0.16
0.00
0.662
0.273
0.145
0.078
0.0
Figure 12. UV-visible spectral changes observed during the reaction
of [RuVI(TPP)(NTs)2] (2a) (1.3 × 10-5 M) with ethylbenzene (0.19
M) in dichloromethane containing pyrazole (2% w/w) at 298 K. Scans
were taken at 60-s intervals.
Figure 14. Linear-free-energy correlation of log kR vs TE for the
reactions of [RuVI(TPP)(NTs)2] (2a) with a series of para-substituted
ethylbenzenes at 298 K.
pyridylmethyl)propylenediamine) (kH/kD ) 12).39b The isotope
effect for the tosylamidation of cyclohexene by 2a was
somewhat smaller, with a kH/kD ratio of 6.1. All these kH/kD
values are typical for H-atom abstraction reactions to form
carboradical intermediates.39
iii. Substitution Effect. Studies of the effect of para sub-
stituents on the rate of tosylamidation of ethylbenzene by 2a
reveal that both electron-donating and -withdrawing substituents
promote the reaction. However, the overall influence of the para
substituents on the relative rates is minor, as shown in Table 9,
which is incompatible with the intermediacy of either alkene-
derived cation radical or carbocation intermediates in the rate-
determining step.
Figure 13. Plots of kobs vs [hydrocarbon] for the tosylamidation of (a)
cyclohexene, (b) ethylbenzene, and (c) cumene by [RuVI(TPP)(NTs)2]
(2a).
Table 8. Summary of Second-Order Rate Constants for the
Tosylamidation of Hydrocarbons by [RuVI(TPP)(NTs)2] (2a)
entry
hydrocarbon
cumene
ethylbenzene
ethylbenzene-d10
cyclohexene
cyclohexene-d10
p-methoxyethylbenzene
p-methylethylbenzene
p-chloroethylbenzene
p-fluoroethylbenzene
103k2/dm3 mol-1 s-1
It is likely that the rate-determining step of the tosylamidation
by 2a involves a carboradical intermediate. To obtain further
evidence for this, we examined the correlation between the
relative reactivity data (log kR) and TE, i.e., the substitution
effect on the C-H bond dissociation energy.40 The TE (total
effect) parameters have been calculated by Wu and co-workers
by the density functional method for a series of para-substituted
toluene, which are the differences between the substituent effects
on the stability of radical (RE) and ground state (GE).40 The
TE values used in this work are listed in Table 9. The log kR vs
TE plot depicted in Figure 14 exhibits very good linearity (R
1
2
3
4
5
6
7
8
9
1.5 ( 0.1
3.60 ( 0.07
0.330( 0.008
7.7 ( 0.3
1.30 ( 0.04
16.5 ( 0.3
6.8 ( 0.1
5.0 ( 0.1
4.3 ( 0.1
the transition state of the tosylamidation of cumene by 2a due
to the presence of a methyl group at the reaction site.
•
) 0.99) with a FTE value of +0.62 (the slope). This provides
ii. Isotope Effect. Reactions involving C-H bond cleavage
in the transition states usually exhibit a large primary kinetic
isotope effect.39 Indeed, for the tosylamidation of ethylbenzene
by 2a, a kH/kD of 11 was obtained, which is identical or very
close to the corresponding values found for the oxidation of
C-H bonds by cis-[RuVI(Tet-Me6)(O)2]2+ (Tet-Me6 ) N,N,N′,N′-
tetramethyl-3,6-diazaoctane-1,8-diamine) (kH/kD ) 11)39c and
trans-[RuVI(pytn)(O)2]2+ (pytn ) N,N′-dimethyl-N,N′-bis(2-
added support for the rate-limiting formation of a benzylic
radical intermediate in the tosylamidation of ethylbenzene.
On the basis of the above results, we propose a mechanism
for the tosylamidation reactions by complex 2a as depicted in
Scheme 5. Abstraction of a hydrogen atom by the ruthenium
imido complex should occur on the periphery of the complex,
since the imido moiety is bound to the coordinately and
electronically saturated ruthenium center. Subsequently, the
radicals produced are efficiently scavenged by the ruthenium
(39) (a) Che, C.-M.; Leung, W.-H.; Li, C.-K.; Poon, C.-K. J. Chem. Soc.,
Dalton Trans. 1991, 379. (b) Che, C.-M.; Tang, W.-T.; Wong, K.-Y.; Li,
C.-K. J. Chem. Soc., Dalton Trans. 1991, 3277. (c) Cheng, W.-C.; Yu,
W.-Y.; Li, C.-K.; Che, C.-M. J. Org. Chem. 1995, 60, 6840.
(40) Wu, Y.-D.; Wong, C.-L.; Chan, K. W.-K.; Ji, G.-Z.; Jiang, X.-K. J.
Org. Chem. 1996, 61, 746.

9132 J. Am. Chem. Soc., Vol. 121, No. 39, 1999
Au et al.
Scheme 5
rate-limiting formation of an alkene-derived π-cation radical.
(ii) In the case of para-substituted styrenes, linear correlation
between log kR (kR ) relative rate) and σ+ gives a F+ value as
small as -1.1, which is hard to reconcile with the rate-
determining formation of the π-cation radical or a carbocation
intermediate. The fact that the para-substitution effect can be
best accounted for by considering both polar and spin-delocal-
ization effects suggests the involvement of a carboradical at
the rate-determining step. (iii) The secondary deuterium isotope
effect kH/kD was found to be 0.85 and 0.97 for the â- and
R-carbon atom of styrene, respectively, implying that there is a
substantial Câ-N but negligible CR-N bond formation in the
transition state. This argues against both the concerted mech-
anism and the stepwise pathway involving an azametallacy-
clobutane intermediate.
(5) Reactions of 2a and 2e with adamantane, cyclohexene,
ethylbenzene, and cumene result in tosylamidation of these
hydrocarbons, affording the corresponding tosylamides in 52-
88% yields. For cyclohexane and toluene, the tosylamidation
products were formed in poor yields (ca. 10%). Kinetic studies
on the reactions between [RuVI(TPP)(NTs)2] and nine hydro-
carbons give the second-order rate constants (k2) in the range
of (0.330 ( 0.008) × 10-3 to (16.5 ( 0.3) × 10-3 dm3 mol-1
porphyrin. The large deuterium isotope effect supports the
symmetrical transition state for the hydrogen atom abstraction.39
Conclusion
(1) Treatment of [RuII(Por)(CO)(MeOH)] (1) with PhIdNTs
in dichloromethane provides a general, convenient route to the
preparation of bis(tosylimido)ruthenium(VI) porphyrins [RuVI-
(Por)(NTs)2] (2). Complexes 2a-e (Por ) TPP, TTP, 4-Cl-
TPP, 4-MeO-TPP, and OEP, respectively), isolated in 60-74%
yields, have been characterized by 1H NMR, UV/vis, IR, FAB,
and electrospray mass spectroscopy along with elemental
analyses.
(2) The isolated high-valent ruthenium tosylimido complexes
2a-e can undergo imido group transfer reactions with a series
of alkenes and alkanes in dichloromethane containing pyrazole.
After these reactions, 2a,e are converted into paramagnetic
tosylamidoruthenium(IV) porphyrins [RuIV(Por)(NHTs)(pz)] (3).
The structures of complexes 3a,b (Por ) TPP and OEP,
respectively) have been determined by X-ray crystallography.
(3) Reactions of 2a and 2e with 12 alkenes afford the
corresponding N-tosylaziridines in 66-85% yields. The aziri-
dination of cis-alkenes, such as cis-stilbene and cis-â-methyl-
styrene, is nonstereospecific with a partial loss of the alkene
stereochemistry. Kinetic studies reveal that the reactions between
2a and 16 alkenes have the second-order rate constants (k2)
ranging from (1.60 ( 0.06) × 10-3 to (90 ( 4) × 10-3 dm3
mol-1 s-1 at 298 K, which can be rationalized on the basis of
electronic and steric effect.
s-1
.
(6) The above tosylamidation reactions probably proceed via
a rate-determining formation of a carboradical intermediate.
First, these reactions exhibit a large primary deuterium isotope
effect (kH/kD ) 11) for the tosylamidation of ethylbenzene,
indicative of a substantial C-H bond cleavage at the rate-
limiting step. Second, in the case of para-substituted ethylben-
zenes, both electron-donating and -withdrawing substituents
promote the reaction; however, the overall influence is minor.
Furthermore, there is a very good linear correlation between
log kR and a related carboradical parameter.
Acknowledgment. This work was supported by the Hong
Kong University Foundation and the Hong Kong Research
Grants Council. We are grateful to Dr. K.-K. Cheung for the
X-ray structural determinations of complex 3b.
Supporting Information Available: UV-visible spectral
changes during the reaction of 2e with styrene in dichlo-
romethane containing pyrazole; plots of kobs vs [alkene] for
reactions of 2a with styrene, norbornene, cis-cyclooctene,
p-methylstyrene, cis-â-methylstyrene, trans-â-methylstyrene,
and R-methylstyrene; UV-visible spectrum of 3b in dichlo-
romethane; characterization of some of the organic aziridine
and amine products; and tables of final coordinates, bond
lengths, bond angles, and anisotropic displacement parameters
of 3b‚C5H12 (PDF). This material is available free of charge
via the Internet at http://pubs.acs.org.
(4) The above aziridination reactions probably proceed via a
stepwise mechanism involving the rate-determining formation
of a carboradical intermediate according to the following lines
of evidence: (i) The slope of the linear plot of log k2 vs E1/2
for eight representative alkenes was found to be -1.7 V-1
(corresponding to R ) 0.1 in the conventional Bronsted plot),
a value too small to be accommodated to a mechanism with
JA9913481