Conventional lithium bases as unconventional sources of methyl anion: Facile Me-Si and Me-C bond cleavage in RLi, R2NLi, and BR4- by an electrophilic osmium dihydride

4030

Organometallics 2002, 21, 4030-4049

Articles

Con ven tion a l Lith iu m Ba ses a s Un con ven tion a l Sou r ces

of Meth yl An ion : F a cile Me-Si a n d Me-C Bon d

-

Clea va ge in RLi, R₂NLi, a n d BR₄by a n Electr op h ilic

Osm iu m Dih yd r id e

Dmitry V. Yandulov, J ohn C. Huffman, and Kenneth G. Caulton*

Department of Chemistry and Molecular Structure Center, Indiana University,

Bloomington, Indiana 47405-4001

Received J anuary 4, 2002

cis,trans-Os(H)₂(OTf)(NO)(PⁱPr₃)₂(1-OTf) and several other precursors (1-X) to Os(H)₂-

(NO)(PⁱPr₃)₂⁺(1⁺) react with (Me₃Si)₂NLi, (Me₃Si)₂CHLi, lithium 2,2,6,6-tetramethylpiperi-

dide (TMPLi), Me₃SiCH₂Li, and B(CH₂SiMe₃)₄^-by a highly unusual, facile â-Me^-transfer,

the exclusive reaction pathway for the first two in nonpolar solvents. A series of lithium

alkyls and alkylamides and organoborate reagents have been examined to reveal widespread

occurrence of the direct â-R′^-transfer (R′ ) H, Me) to the Os electrophile, being completely

selective for â-H^-over â-Me^-, with the sole (surprising) exception of NpLi. The â-R′

elimination was ruled out as the mechanism of the net â-R′^-transfer for two representative

RLi cases with R′ ) H, Me, and a single-electron-transfer mechanism was shown to be

inoperative for tetraalkylborates. The mechanistic studies also uncovered the important role

of Li in RLi and R₂NLi, which acts as a potent Lewis acid to abstract the halide/pseudohalide

X from 1-X in generating the unsaturated Os species. The proposed intimate mechanism of

Me-C and Me-Si bond cleavage is a direct S_E2 substitution at carbon with inversion of the

Me group, supported by DFT calculations. While the imines formed in the process of C-H

and C-Me cleavage are lithiated by, and compete for the Os with, the original base, the

unsaturated silicon species formed by Si-Me cleavage react with the remaining base by

1,2-addition of (N,C)-Li, forming intermediates that are also reactive by â-Me^-transfer. A

complex mixture of Os-free coproducts is obtained in both cases. The structural features of

1⁺responsible for its unusual reactivity are discussed.

In tr od u ction

chemistry,^4,6,7including anionic olefin polymerization

processes.^8,9Bulky amides, such as hexaalkyldisilazides,

are widely used to stabilize transition-metal com-

plexes^10,11and main-group compounds¹²with low coor-

dination numbers, and 2,2,6,6-tetramethylpiperidide

has found increasing use in a similar fashion.^11,13The

Lithium alkyls^1-3(RLi) and alkylamides⁴(R₂NLi) are

invaluable synthetic tools⁵in organic^1,2and inorganic

* To whom correspondence should be addressed. E-mail: caulton@

indiana.edu.

(1) Wakefield, B. J . The Chemistry of Organolithium Compounds;

Pergamon Press: New York, 1976.

(2) Wakefield, B. J . Organolithium Methods; Academic Press: San

Diego, 1988. Stowell, J . C. Carbanions in Organic Synthesis; Wiley:

New York, 1979. Comprehensive Carbanion Chemistry; Buncel, E.,

Durst, T., Eds.; Elsevier: New York, 1984.

(6) Fischer, E. O. Adv. Organomet. Chem. 1976, 14, 1.

(7) Reference 1, Part IV.

(8) Szwarc, M. Carbanions, Living Polymers and Electron-Transfer

Processes; Wiley: New York, 1968.

(9) Reference 1, Chapter 7. Reference 4, Chapter 2.

(3) Bates, R. B.; Ogle, C. A. Carbanion Chemistry; Springer-

Verlag: New York, 1983.

(10) Reference 4, Chapter 8. Wannagat, U.; Bu¨rger, H. Monatsh.

Chem. 1963, 94, 1007. Wannagat, U.; Bu¨rger, H. Monatsh. Chem. 1964,

95, 1099. Bradley, D. C.; Copperthwaite, R. G. Inorg. Synth. 1978, 18,

112. Andersen, R. A.; Templeton, D. H.; Zalkin, A Inorg. Chem. 1978,

17, 2317. Tilley, T. D.; Andersen, R. A.; Zalkin, A. Inorg. Chem. 1984,

23, 2271. Murray, B. D.; Power, P. P. Inorg. Chem. 1984, 23, 4584.

Murray, B. D.; Power, P. P. J . Am. Chem. Soc. 1984, 106, 7011.

Bartlett, R. A.; Power, P. P. J . Am. Chem. Soc. 1987, 109, 7563.

Bartlett, R. A.; Ellison, J . J .; Power, P. P.; Shoner, S. C. Inorg. Chem.

1991, 30, 2888. Ellison, J . J .; Ruhlandt-Senge, K.; Power, P. P. Angew.

Chem., Intl. Ed. Engl. 1994, 33, 1178. Stewart, J . L.; Andersen, R. A.

Polyhedron 1998, 17, 953.

(4) Lappert, M. F.; Power, P. P.; Sanger, A. R.; Srivastava, R. C.

Metal and Metalloid Amides; Wiley: New York, 1980.

(5) For extensive structural studies of organolithium compounds

see: Pauer, F.; Power, P. P. In Lithium Chemistry: Theoretical and

Experimental Overview; Sapse, A.-M., Schleyer, P. v. R., Eds.; Wiley:

New York, 1995; Chapter 9. Beswick, M. A.; Wright, D. S. In

Comprehensive Organometallic Chemistry II; Abel, E. W., Stone, F.

G. A., Wilkinson, G., Eds.; Elsevier: New York, 1995; Vol. 1, Chapter

1. Weiss, E. Angew. Chem., Int. Ed. Engl. 1993, 32, 1501. Setzer, W.

N.; Schleyer, P. v. R. Adv. Organomet. Chem. 1985, 24, 353. Gregory,

K.; Schleyer, P. v. R.; Snaith, R. Adv. Inorg. Chem. 1994, 37, 47.

Mulvey, R. E. Chem. Soc. Rev. 1991, 20, 167. Seebach, D. Angew.

Chem., Int. Ed. Engl. 1988, 27, 1624. Lucht, B. L.; Collum, D. B. Acc.

Chem. Res. 1999, 32, 1035.

(11) Olmstead, M. M.; Power, P. P.; Sigel, G. Inorg. Chem. 1986,

25, 1027. Chen, H.; Bartlett, R. A.; Dias, H. V. R.; Olmstead, M. M.;

Power, P. P. J . Am. Chem. Soc. 1989, 111, 4338. Olmstead, M. M.;

Power, P. P.; Shoner, S. C. Inorg. Chem. 1991, 30, 2547.

Publication on Web 08/29/2002

Lithium Bases as Sources of Methyl Anion

Organometallics, Vol. 21, No. 20, 2002 4031

Sch em e 1. Gen er a tion a n d Rea ctivity of 2

typical modes of reactivity of RLi and R₂NLi, as a

nucleophile or as a base, are often complicated by single-

electron-transfer reduction¹⁴and â-hydride-tansfer^15-17

of olefin polymerization.²⁵The direct transfer of the

nucleophilic R^-from BR₄^-can also proceed in competi-

tion with transfer of an R-¹⁹or â-substituent²⁶(to boron),

including hydride.

processes.

Tetraorganylborates (BR₄

- 18

)

are gaining increased

Full substitution of the â-positions in RLi and R₂NLi

solves the â-H^-transfer problem, and â-Me^-trans-

fer does not typically take place. We are aware of only

a few precedents of an organolithium base serving as

a source of MeLi: Shiner and co-workers obtained

the â-Me^-transfer product in 33% yield in the ther-

mal reaction of lithium 2,2,6,6-tetramethylpiperidide

(TMPLi) with a nonenolizable ketone,²⁷and Pines et al.

have implicated loss of RNa from a cyclohexadienyl

intermediate, driven by achieving aromaticity, in a

sodium/benzylsodium-catalyzed thermal dehydromethy-

lation of several cyclohexadienes.²⁸

attention as mild but selective synthetic alternatives to

organolithium reagents.^19-22In comparison to RLi and

R₂NLi, the SET reactivity is much more widespread

with BR₄,

^{- 23,24}commonly used as free-radical initiators

(12) Simpson, K. M.; Gedrige, R. W.; Higa, K. T. J . Organomet.

Chem. 1993, 456, 31. Fjeldberg, T.; Hope, H.; Lappert, M. F.; Power,

P. P.; Thorne, A. J . J . Chem. Soc., Chem. Commun. 1983, 639. Chorley,

R. W.; Hitchcock, P. B.; Lappert, M. F.; Leung, W.-P.; Power, P. P.;

Olmstead, M. M. Inorg. Chim. Acta 1992, 198-200, 203. Braunschweig,

H.; Chorley, R. W.; Hitchcock, P. B.; Lappert, M. F. J . Chem. Soc.,

Chem. Commun. 1992, 1311.

(13) Ruppa, K. B. P.; Feghali, K.; Kovacs, I.; Aparna, K.; Gam-

barotta, S.; Yap, G. P. A.; Bensimon, C. J . Chem. Soc., Dalton Trans.

1998, 1595. Lam, H.-W.; Wilkinson, G.; Hussain-Bates, B.; Hursthouse,

M. B. J . Chem. Soc., Dalton Trans. 1993, 1477. Linti, G.; Frey, R.;

Polborn, K. Chem. Ber. 1994, 127, 101. Linti, G.; Frey, R.; Polborn, K.

Chem. Ber. 1994, 127, 1387. Bradley, D. C.; Frigo, D. M.; Hursthouse,

M. B.; Hussain, B. Organometallics 1988, 7, 1112. Lappert, M. F.;

Slade, M. J .; Atwood, J . L.; Zaworotko, M. J . J . Chem. Soc., Chem.

Commun. 1980, 621. Krossing, I.; No¨th, H.; Tacke, C.; Schmidt, M.;

Schwenk, H. Chem. Ber. 1997, 130, 1047. Knabel, K.; Krossing, I.;

No¨th, H.; Schwenk-Kircher, H.; Schmidt-Amelunxen, M.; Seifert, T.

Eur. J . Inorg. Chem. 1998, 1095.

This study describes facile â-Me^-transfer from (Me₃-

Si)₂NLi, (Me₃Si)₂CHLi, Me₃SiCH₂Li, TMPLi, and B(CH₂-

-

SiMe₃)₄to an osmium electrophile under ambient

conditions, the exclusive reaction pathway for the

former two reagents in nonpolar solvents, which rep-

resents rare examples of intermolecular sp³C-Si^29-31

and C-C³²bond cleavage.

(14) Reference 8, Chapter VI. Guthrie, R. D. In Comprehensive

Carbanion Chemistry; Buncel, E., Durst, T., Eds.; Elsevier: New York,

1984; Part A, Chapter 5. McKinley, J .; Aponic, A.; Raber, J . C.; Fritz,

C.; Montgomery, D.; Wigal, C. T. J . Org. Chem. 1997, 62, 4874. Brown,

D. W.; Lindquist, M.; Mahon, M. F.; Malm, B.; Nilson, G. N.; Ninan,

A.; Sainsbury, M.; Westerlund, C. J . Chem. Soc., Perkin Trans. 1 1997,

2337. Renaud, P.; Fox. M. A. J . Am. Chem. Soc. 1988, 110, 5702. Ashby,

E. C.; Park, B.; Patil, G. S.; Gadru, K.; Gurumurthy, R. J . Org. Chem.

1993, 58, 424. Ashby, E. C.; Bowers, J . R., J r. J . Am. Chem. Soc. 1981,

103, 2242. Ashby, E. C.; Goel, A. B. J . Am. Chem. Soc. 1981, 103, 4983.

Scott, L. T.; Carlin, K. J .; Schultz, T. H. Tetrahedron Lett. 1978, 47,

4637.

Resu lts

I. Gen er a l Rem a r k s on th e “Deh yd r oh a logen a -

tion ” of Os(H)₂(OTf)(NO)L₂. The main focus of this

study is the mechanism of HOTf removal from cis,trans-

Os(H)₂(OTf)(NO)L₂(L ) PⁱPr₃; 1-OTf), accomplished

with a variety of organolithium reagents. As detailed

in a separate report,³³this reaction affords the highly

reactive 16-electron Os intermediate {OsH(NO)L₂} (2),

which, eluding direct observation even at low temper-

atures, scavenges an additional phosphine ligand and

reacts by intra- and intermolecular C-H activation with

phosphine alkyls and aromatic solvents, respectively,

with progressively larger ∆G^qand -∆G° values (Scheme

1). Mechanistic studies³³revealed the involvement of 2

as the true reaction intermediate and C-H bond-

(15) Bennet, G. B.; Houlihan, W. J .; Mason, R. B. J . Organomet.

Chem. 1975, 99, 185. Brown, H. C.; Kramer, G. W.; Hubbard, J . L.;

Krishnamurthy, S. J . Organomet. Chem. 1980, 188, 1. Reetz, M. T.;

Stephan, W. Angew. Chem., Int. Ed. Engl. 1977, 16, 44. Reetz, M. T.;

Schinzer, D. Angew. Chem., Int. Ed. Engl. 1977, 16, 44. Buhler, J . D.

J . Org. Chem. 1973, 38, 904. Kowalski, C.; Greary, X.; Rollin, A. J .;

Burke, M. C. J . Org. Chem. 1978, 43, 2601. Melamed, U.; Feit, B.-A.

J . Chem. Soc., Perkin Trans. 1 1980, 1267. Reference 3, p 59.

(16) Wittig, G.; Schmidt, H. J .; Remmer, H. Chem. Ber. 1962, 95,

2377. Newcomb, M.; Burchill, M. T. J . Am. Chem. Soc. 1984, 106, 8276.

(17) For an extensive overview of elimination reactions of carbanions

see ref 3, Chapter VI.

(18) Throughout the paper, BR₄^-refers to borates with at least one

alkyl substituent, borohydrides excluded.

(19) Negishi, E.-I. J . Organomet. Chem. 1976, 108, 281.

(23) Chatterjiee, S.; Davis, P. D.; Gottschalk, P.; Kurz. M. E.;

Sauerwein, B.; Yang, X.; Schuster, G. B. J . Am. Chem. Soc. 1990, 112,

6329.

(20) Zhu, D.; Kochi, J . K. Organometallics 1999, 18, 161.

(24) Sarker, A. M.; Lungu, A.; Mejiritski, A.; Kamenko, Y.; Neckers,

D. J . Chem. Soc., Perkin Trans. 2 1998, 2315. Schuster, G. B. Pure

Appl. Chem. 1990, 62, 1565.

(21) Zhu, D.; Kochi, J . K. J . Organomet. Chem. 1999, 580, 295.

(22) Miyaura, N.; Itoh, M.; Suzuki, A. Synthesis 1976, 9, 618.

Miyaura, N.; Sasaki, N.; Itoh, M.; Suzuki, A. Tetrahedron Lett. 1977,

38, 3369. Yamada, K.; Yano, T.; Miyaura, N.; Suzuki, A. Bull. Chem.

Soc. J pn. 1979, 52, 275. Hunter, R.; Michael, J . P.; Tomlison, G. D.

Tetrahedron 1994, 50, 871. Thaler, E. G.; Caulton, K. G. Organome-

tallics 1990, 9, 1871.

(25) Footnote 6 in ref 20.

(26) Haag, A.; Hesse, G. Liebigs Ann. Chem. 1971, 751, 95.

(27) Shiner, C. S.; Berks, A. H.; Fisher, A. M. J . Am. Chem. Soc.

1988, 110, 957.

(28) Pines, H.; Eschinazi, H. E. J . Am. Chem. Soc. 1956, 78, 5950.

4032 Organometallics, Vol. 21, No. 20, 2002

Yandulov et al.

Sch em e 2. Obser ved P a th w a ys of Rea ctivity of

breaking and -forming reactions proceeding by oxida-

tive-addition/reductive-elimination mechanisms. Gen-

eration of 2 under ambient conditions effectively leads

to the formation of the metalated isomers 1-P ∼C, which

in C₆D₆solvent convert to the products of aromatic C-D

oxidative addition Os(H)(D)(C₆D₅)(NO)(PⁱPr₃)₂(1-P h -

d ,d ₅) on the time scale of hours; integration of hydride

or ³¹P resonances of 1-P ∼C and 1-P h -d ,d ₅can be

reliably used to evaluate the extent of formation of 2.

Of additional importance to the mechanistic studies

presented here, generation of 2 often yields the trihy-

dride 1-H³⁴and free L, in a decomposition fashion, the

yields of which vary, although both typically constitute

at most a trace amount, with the exception of the

reaction with NpLi.³³

1-OTf w ith RLi a n d BR₄-

II. Reactivity of Or gan olith iu m Bases with 1-OTf.

(a ) RLi a n d R₂NLi. Originally, it was found that (Me₃-

Si)₂NLi instantaneously reacts with 1-OTf under ambi-

ent conditions in a variety of solvents via an unprec-

edented methyl anion transfer to quantitatively yield

1-Me, thus serving as a formal source of MeLi. A

number of other lithium alkyls and alkylamides were

then examined to determine the scope and generality

of such an unusual process. Indeed, in several other

cases, (Me₃Si)₂CHLi, Me₃SiCH₂Li, and lithium 2,2,6,6-

tetramethylpiperidide (TMPLi), â-Me^-transfer was

found to be at least a significant reaction pathway. To

the best of our knowledge, such reactivity is unprec-

edented for the above reagents, with the exception of

TMPLi.²⁷Distantly related examples of similar reactiv-

ity include â-Me^-abstraction from Cp′₂ZrCl(CH(SiMe₃)₂)

by AlCl₃,³⁰from (Cp*∼O)Ti(CH₂SiMe₃)₂by B(C₆F₅)₃,³¹

and from (η⁵-cyclohexadienyl)ruthenium complexes by

Brønsted acidic catalysts.³⁵

Ta ble 1. Rea ctivity of 1-OTf w ith RLi: Yield s of

th e Rea ction P r od u cts in C₆D₆a t 20 °C Obta in ed

by ¹H NMR In tegr a tion w ith in 15 Min , Accor d in g

to P a th w a ys A a n d B (Sch em e 2)

yield, %

R(Li)

pathway A (R′)

pathway B

(Me₃Si)₂N

(Me₃Si)₂CH

Me₃SiCH₂

TMP

100 (Me)

40 (Me)

50 (Me)

- (Me)

96 (H), 0 (Me)

100 (H)

53 (H)

60

13^a

100

Np

ⁱPr₂N

^tBu

ⁿBu

47

19

Et

81 (H)

a

Pathway B or C, see text.

The reactivity of 1-OTf with all organolithium re-

agents examined was found to generally conform to

three independent pathways, as shown in Scheme 2:

â-R′^-transfer (A), triflate

metathesis (B), and deprotonation (C), all of which

eventually lead to 2 (with the exception of A with R′ )

H, as the trihydride 1-H is stable to H₂reductive

elimination³⁴). The independence of pathways A and B

is discussed below.

In C₆D₆, the reactivity of 1-OTf with all organolithium

reagents examined predominantly follows pathways A

and B. The starting material 1-OTf is fully consumed

at 20 °C within 15 min of mixing the reagents³⁶

(essentially instantaneously), giving the products cor-

responding to pathways A and B in yields summarized

in Table 1. The surprising result is that only NpLi does

not react via pathway A, as inferred from the -40 °C

reaction in d₈-PhMe,³³while all other bases studied

transfer the â-R′^-substituent (Me or H) to a significant

extent. Although all 1-R (R * H) species formed via

either of the pathways A and B reductively eliminate

R-H under ambient conditions, these subsequent reac-

tions are only marginally developed after 15 min at 20

°C (except for 1-Np ). With the latter exception, a

significant amount of secondary products was observed

only in ⁿBuLi and EtLi reactions, in which case 2 was

trapped by the olefins released in the â-H^-transfer as

adducts 2-C₄H₈(6%) and 2-C₂H₄(5%), respectively. The

two reaction pathways, A and B, account for all of the

observed products in all but two cases, TMPLi and LDA,

in which case 1-OTf undergoes additional reactions with

the imines released in pathway A. Additionally, small

amounts of 1-H, 1-P h -d ,d ₅, and 1-P ∼C (13%), indicative

of the intermediate formation of 2, were invariably

observed in the case of TMPLi and, although consistent

with formation of a highly reactive 1-TMP intermediate

(not observed at 20 °C) via pathway B, these could also

result from the deprotonation pathway C.

The syntheses of the four RLi reagents exhibiting the

â-Me^-transfer did not employ MeLi, and therefore their

reactivity via pathway A (with R′ ) Me) involves

genuine cleavage of Me-Si and Me-C bonds by Os(H)₂-

(NO)(PⁱPr₃)₂⁺(1⁺) at 25 °C. While for bulky (Me₃Si)₂NLi,

(Me₃Si)₂CHLi, and TMPLi triflate metathesis (pathway

B) is likely severely impeded by the steric bulk of the

PⁱPr₃phosphines in 1⁺, the reactivity of Me₃SiCH₂Li

reveals a considerable kinetic preference for the â-Me^-

transfer (A) over transfer of Me₃SiCH₂(B) (the product

of B, 1-CH₂SiMe₃, is quite stable at T < 20 °C). In

comparison to the â-H^-transfer, however, â-Me^-trans-

fer is observed experimentally to be strongly disfavored,

(29) Hoffman, P.; Heiss, H.; Neiteler, P.; Mu¨ller, G.; Lachmann, J .

Angew. Chem., Int. Ed. Engl. 1990, 29, 880.

(30) Horton, A. D.; Orpen, A. G. Organometallics 1992, 11, 1193.

(31) Gielens, E. E. C. G.; Tiesnitsch, J . Y.; Hessen, B.; Teuben, J .

H. Organometallics 1998, 17, 1652.

(32) Liou, S.-Y.; Boom, M. E. v. d.; Milstein, D. J . Chem. Soc., Chem.

Commun. 1998, 687.

(33) Yandulov, D. V.; Caulton, K. G. Manuscript in preparation.

(34) Yandulov, D. V.; Huang, D.; Huffman, J . C.; Caulton, K. G.

Inorg. Chem. 2000, 39, 1919.

(35) Older, C. M.; Stryker, J . M. J . Am. Chem. Soc. 2000, 122, 2784.

(36) RLi and R₂NLi were typically used in 20-50% molar excess.

Lithium Bases as Sources of Methyl Anion

Organometallics, Vol. 21, No. 20, 2002 4033

Ta ble 2. Rea ctivity of 1-OTf w ith BR₄^-: Yield s of

th e Rea ction P r od u cts in C₆D₆a t 20 °C Obta in ed

by ¹H NMR In tegr a tion w ith in 15 Min , Accor d in g

to P a th w a ys A a n d B (Sch em e 2)

are the reagents of choice for the preparation of the

corresponding 1-Me and 1-Et, because the reactions

(quantitative for 1-Me) can be carried out in MeOH at

low temperature to yield crystalline but temperature-

sensitive products, isolated in moderate to good yields.

While isolation of 1-ⁿBu and 1-Np ³³was not attempted,

due to their limited thermal stability, 1-CH₂SiMe₃is

highly soluble and could be isolated only in low yields

and as a ca. 15:85 mixture with 1-Me.

yield, %

-

Q⁺BR₄

pathway A (R′)

pathway B (R)

LiBMe₄

100

[Bu₄N][Ph₃BMe]

100 (Me), 0 (Ph)

LiBEt₄

23 (H)

39 (H)

59 (Me)

77^a

[Li(TMEDA)₂][ⁱPr₃BMe]

[Li(THF)₄][B(CH₂SiMe₃)₄]

61 (Me), 0 (ⁱPr)

41

All dihydrido alkyl complexes 1-R exhibit NMR data

sufficient for unambiguous assignments of the cis,trans-

Os(H)₂(R)(NO)(PⁱPr₃)₂octahedral structures. The R-CH

¹H NMR resonances of the alkyl groups reveal moderate

couplings to P (4-5 Hz) and the cis-hydride (1-2 Hz),⁴¹

a

With 2-C₂H₄included.

at least kinetically, as the reaction with LDA reveals

no trace of 1-Me.

1

while the differences of the hydride H NMR chemical

shifts, ∆δ ) δ(H_A) - δ(H_M), for all 1-R are distinctly

lower than the values of halide/pseudohalide derivatives

1-X.³³

The main reactivity mode characteristic of all 1-R

compounds is the reductive elimination of the corre-

sponding RH to yield 2 and its subsequent products.

Scheme 3 shows the order of stability of the complexes

observed in aromatic solvents,³³which ranges from a

(b) BR₄^-. Several tetraorganylborates were examined

for reactivity with 1-OTf under analogous conditions,

and similar reactivity pathways, consisting of â-R′^-

transfer (i.e. â to boron) (pathway A) and triflate

metathesis with R^-(pathway B), were found, again

involving both H^-and Me^-transfer in A (Table 2,

Scheme 2). While cases of â-H^-transfer from BR₄^-have

been documented,²⁶â-Me^-transfer from B(CH₂SiMe₃)₄

-

t

_1/2(-20 °C) value of about 1 h for 1-Np to a t_1/2(+60 °C)

is unprecedented, to the best of our knowledge.³⁷On the

basis of the following observations, the products in Table

2 are attributed to the reactivity of intact LiBR₄rather

than the corresponding RLi, possibly formed from the

former in a preequilibrium fashion together with BR₃.

The solution ¹H NMR studies of LiBMe₄revealed no

evidence for such an equilibrium, as the B-H coupling

was resolved at temperatures of up to 100 °C in PhMe³⁸

and 55 °C in Et₂O.³⁹The LiBEt₄reaction gives a product

distribution (A vs B) considerably different from that

obtained with EtLi (Table 1), while the reaction of [Li-

(THF)₄][B(CH₂SiMe₃)₄] leads to organic products dra-

matically different from those obtained in the reaction

of Me₃SiCH₂Li in the presence of B(CH₂SiMe₃)₃and the

value of about 10 h for the most thermodynamically

stable 1-P h (t_1/2(25 °C) ≈ 24 h for 1-Me). Consideration

of the relative stability of the alkyl complexes (R ) Me,

Et, ⁿBu, Np), with reasonably similar electronic proper-

ties, clearly shows that increasing steric bulk of R⁴²

accelerates the reductive elimination of RH.

Sch em e 3. Rela tive Or d er of Sta bility of

1-R Com p lexes tow a r d Red u ctive

Elim in a tion of RH

R ) Np , ⁿBu ≈ CH₂SiMe₃≈ Et < Me , Ph

(d ) Evid en ce for th e Dep r oton a tion P a th w a y

(C). Under conditions that favor minimal aggregation

of organolithium bases, with strong solvation of Li, the

reactivity of 1-OTf with selected RLi can follow neither

of the pathways dominating the reactivity in aromatic

solvents and available spectroscopic data point to depro-

tonation mechanism C. Thus, reaction of 1-OTf with 5

equiv of Me₃SiCH₂Li in d₈-THF at -40 °C proceeds on

the time scale of minutes to yield a new major product;

only a trace of 1-CH₂SiMe₃was observed at -20 °C,

after complete conversion. The ¹H NMR spectrum of this

new species, tentatively assigned as the osmate OsH-

-

Bu₄N⁺salt of B(CH₂SiMe₃)₄exhibits reactivity very

similar to that of [Li(THF)₄][B(CH₂SiMe₃)₄], with a

related Os compound. The reactions of LiBEt₄with

1-OTf in d₄-MeOH and THF give very similar product

ratios, both with an increased proportion of B relative

to that obtained in C₆D₆, while pure LiBEt₄shows no

significant hydrolysis in d₄-MeOH on the time scale of

the reaction with 1-OTf (10-15 min at 20 °C). However,

the reaction of 1-OTf with a d₄-MeOH solution of

LiOCD₃, prepared by quenching with 1.2 equiv of EtLi

and containing 1.5 equiv of BEt₃, gave products very

similar to those obtained with LiBEt₄. In view of the

fact that BEt₃is inert to 1-OTf in d₄-MeOH, this result

suggests that BEt₃(OMe)^-may exhibit similar reactiv-

ity.⁴⁰

-

(CH₂SiMe₃)(NO)L₂(2-CH₂SiMe₃^-), formed via path-

way C as shown below, at -40 °C shows all of the

signals expected for a TBP structure with trans-L, in

the correct intensity ratio. The hydride appears as a 32.8

Hz triplet at -10.38 ppm, while the R-CH₂protons at

0.60 ppm exhibit a P-(C)H coupling of 5.4 Hz, and

(PCH)Me signals are diastereotopically shifted. Most

importantly, the selectively decoupled ³¹P{¹H} signal⁴³

is a doublet, confirming the presence of a single hydride

ligand.

(c) Ch a r a cter iza tion a n d In tr in sic Rea ctivity of

1-R. The identity of the â-Me^-transfer product, 1-Me,

was confirmed by independent synthesis from 1-OTf

-

and MeLi. However, the borates Ph₃BMe^-and BEt₄

(37) Bumagin, N. A.; Bykov, V. V.; Artamkina, G. A.; Beletskaya, I.

P. Dokl. Akad. Nauk 1995, 340, 493. Bumagin, N. A.; Bykov, V. V.;

Artamkina, G. A.; Beletskaya, I. P. Izv. Akad. Nauk SSSR, Ser. Khim.

1990, 11, 2664. Artamkina, G. A.; Tuzikov, A. B.; Beletskaya, I. P.;

Reutov, O. A. Izv. Akad. Nauk SSSR, Ser. Khim. 1980, 10, 2430.

(38) Groves, D.; Rhine, W.; Stucky, G. D. J . Am. Chem. Soc. 1971,

93, 1553.

(39) Williams, K. C.; Brown, T. L. J . Am. Chem. Soc. 1966, 88, 4134.

(40) Siegmann, K.; Pregosin, P. S.; Venanzi, L. M. Organometallics

1989, 8, 2659.

(41) Such coupling probably results from through-space interactions

and is not unusual: analogous observations were made for cis,trans-

OsH(Me)(CO)₂L₂complexes: Esteruelas, M. A.; Lahoz, F. J .; Lopez,

J . A.; Oro, L. A.; Schlu¨nken, C.; Valero, C.; Werner, H. Organometallics

1992, 11, 2034.

(42) Tolman, C. A. Chem. Rev. 1977, 77, 313.

(43) Hydride-only coupled ³¹P signal.

4034 Organometallics, Vol. 21, No. 20, 2002

Yandulov et al.

Since the product of pathway B, 1-CH₂SiMe₃, is

stable for hours at temperatures of up to 0 °C, formation

of 2-CH₂SiMe₃^-cannot be explained by the intermedi-

ate formation of 1-CH₂SiMe₃undergoing reductive

elimination to give 2, which then reacts with Me₃SiCH₂-

Li to form the observed Os product. Therefore, 2 is

formed via deprotonation of 1-OTf and is promptly

complexed by the remaining Me₃SiCH₂Li, with loss of

OTf^-. An alternative mechanism for the formation of

2-CH₂SiMe₃^-, deprotonation of 1-CH₂SiMe₃, is plau-

sible; however, the hydrides are expected to be consider-

ably more acidic in 1-OTf, due to the greater electrone-

gativity of OTf, compared to that of CH₂SiMe₃. Indeed,

using [cis,trans-Os(H)₂(THF)(NO)L₂][B(C₆H₃-3,5-(CF₃)₂)₄],

1-THF ⁺(BAr′₄^-), one of the most reactive forms of 1⁺,

under conditions analogous to those of the 1-OTf

reaction yields the identical product, 2-CH₂SiMe₃^-, but

much faster: the reaction is complete in minutes at -80

°C, consistent with greater hydride acidity in the

cationic species. Additionally, reaction of 1-OTf with 15

equiv of Me₃SiCH₂Li and 20 equiv of TMEDA in

d₈-PhMe at -60 °C gives predominantly 1-CH₂SiMe₃

and 22% of 2-CH₂SiMe₃^-, identified by a similar -10.05

occurrence of â-hydride elimination reactions of transi-

tion-metal alkyl⁴⁵and, to a lesser extent, alkylamide⁴⁶

complexes suggests that formation of the products of

pathway A, with R′ ) H, may be the result of â-hydride

elimination from the products of pathway B. While

numerous examples of analogous â-methyl elimination

reactions, involving C-C and C-Si bond cleavage, have

been documented,^47-49these are by far much less

common and often require fairly high activation ener-

gies.⁴⁷Nevertheless, several early-metal metallocene

complexes have been shown to undergo facile â-methyl

elimination,⁴⁸which demonstrates the possibility of such

a mechanism with R′ ) Me.

Since all possible products of pathway B are saturated

complexes, some transformation is necessary to generate

a vacant Os orbital, in order for the â-R′ elimination to

proceed.^50,51We have considered a nondissociative â-R′

elimination mechanism, such as bending of NO⁵²or

migration of a ligand to the NO nitrogen,⁵²and two

dissociative mechanisms, reversible loss of L and re-

versible loss of H₂. While reversible loss of NO is a

possibility,⁵²it was not considered, since 1-H, a repre-

sentative of 1-R complexes in this regard, is quite stable

at 90 °C in MeOH under H₂for 24 h.³⁴We have studied

two pairs of complexes, 1-H/1-Et and 1-Me/1-CH₂SiMe₃,

to evaluate the possibility of the â-R′ elimination mech-

anism for â-R′ ) H, Me, as opposed to a direct â-R′^-

transfer from RLi that does not involve intermediate

formation of the products of pathway B.

1

ppm (32 Hz) hydride triplet in the H NMR spectrum

and a ³¹P NMR chemical shift similar to that observed

in d₈-THF (also a doublet with selective decoupling).

However, both products evolve with the same rates and,

therefore, are being formed via independent processes

(B and C), which rules out the intermediacy of 1-CH₂-

SiMe₃in the formation of 2-CH₂SiMe₃^-. Notably, the

analogous reaction in d₈-PhMe in the absence of TMEDA

yields no observable trace of the osmate species, which

shows that solvation of Li is necessary for the operation

of the deprotonation mechanism.

(a ) â-R′ ) H. Monitoring the evolution of the rapidly

formed initial products of the reaction of 1-OTf with

1

LiBEt₄in C₆D₆at 28 °C by H NMR for 11 h showed

only a slight increase in the amount of 1-H, from 23%

of all observed Os products measured initially to a

maximum of 27%, comparable to the experimental

integration error. At the same time, the 1-Et complex

fully decayed in a first-order reaction via reductive

elimination of ethane, forming 1-P ∼C, which subse-

quently converted into 1-P h -d ,d ₅.³³The amount of the

2-C₂H₄adduct slightly increased, from 19 to 24%.

Following the evolution of pure 1-Et, devoid of an

observable trace of 1-H, under analogous conditions

revealed a ca. 1% trace of 1-H appearing after 95%

conversion of 1-Et. This amount of the trihydride is

attributable to the reactivity of 2 (Scheme 1), since decay

-

Further characterization of the osmate 2-CH₂SiMe₃

was hampered by its very limited thermal stability. In

all cases where its formation was established by low-

temperature NMR, the complex decomposed within

minutes at room temperature, producing mainly free L.

Attempts to convert it to 1-CH₂SiMe₃by protonation

failed: reacting 2-CH₂SiMe₃^-with excess MeOH even

at -80 °C gave several new species, but with no

appreciable amount of the desired product. In the case

of NpLi, a similar hydride signal was observed in the

1-OTf/NpLi reaction in d₈-PhMe,³³assignable to the

analogous 2-Np ^-osmate, although in trace yield, and

it also did not persist at ambient temperature.

In sharp contrast to the Me₃SiCH₂Li results, using a

much weaker base,⁴⁴(Me₃Si)₂NLi, in a reaction with

1-THF ⁺in d₈-THF at -60 °C, gave no observable

amount of 2 or related products but cleanly and quan-

titatively produced 1-Me on the time scale of hours, with

no intermediates. Additionally, the reaction of 1-OTf

with 1.5 equiv of Me₃SiCH₂Li in d₈-THF at room

temperature gave products of pathways A and B in ca.

60% total yield. Therefore, while pathway C in principle

can be operational, it requires a very strong base, such

as highly disaggregated alkyllithium, and yet still may

not be the kinetically preferred reactivity pathway,

apparently due to the high strength of the Os-H bonds

and limited steric accessibility of (Os)H.

(45) Collman, J . P.; Hegedus, L. S.; Norton, J . R.; Finke, R. G.

Principles and Applications of Organotransition Metal Chemistry;

University Science Books: Mill Valley, CA.; pp 95, 386.

(46) Reference 45, p 64. Tsai, Y. C.; J ohnson, M. J . A.; Mindiola, D.

J .; Cummins, C. C. J . Am. Chem. Soc. 1999, 121, 10426. Kuhlman, R.;

Streib, K.; Caulton, K. G. J . Am. Chem. Soc. 1993, 115, 5813.

(47) Ankianiec, B. C.; Christou, V.; Hardy, D. T.; Thomson, S. K.;

Young, G. B. J . Am. Chem. Soc. 1994, 116, 9963. Thomson, S. K.;

Young, G. B. Organometallics 1989, 8, 2068. McNeill, K.; Andersen,

R. A.; Bergman, R. G. J . Am. Chem. Soc. 1995, 117, 3625. McNeill, K.;

Andersen, R. A.; Bergman, R. G. J . Am. Chem. Soc. 1997, 119, 11244.

Dimauro, P. T.; Wolczanski, P. T. Polyhedron 1995, 14, 149.

(48) Hajela, S.; Bercaw, J . E. Organometallics 1994, 13, 1147.

Horton, A. D. Organometallics 1996, 15, 2675. Guo, Z.; Swenson, D.

C.; J ordan, R. F. Organometallics 1994, 13, 1424.

(49) Sini, G.; Macgregor, S. A.; Eisenstein, O.; Teuben, J . H.

Organometallics 1994, 13, 1049. Karrass, S.; Schwarz, H. Organome-

tallics 1990, 9, 2409. Hartwig, J . F.; Bergman, R. G.; Andersen, R. A.

J . Am. Chem. Soc. 1990, 112, 3234. For an extensive list of additional

references, see ref 35.

(50) Reference 45, p 95.

(51) Crabtree, R. H. The Organometallic Chemistry of the Transition

Metals, 2nd ed.; Wiley: New York, 1994; p 174.

III. In d ep en d en ce of P a th w a ys A a n d B: Dir ect

â-R′^-Tr a n sfer vs â-R′ Elim in a tion . The widespread

(52) Richter-Addo, G. B.; Legzdins, P. Metal Nitrosyls; Oxford

University Press: London, 1992; Chapter 7.

(44) Arnett, E. M.; Moe, K. D. J . Am. Chem. Soc. 1991, 113, 7068.

Lithium Bases as Sources of Methyl Anion

Organometallics, Vol. 21, No. 20, 2002 4035

of, for example, 1-Me often generates similar traces of

1-H. These results rule out a nondissociative â-hydride

elimination mechanism. However, a dissociative mech-

anism may be responsible for the suspected conversion

of 1-Et to 1-H, if the product is more prone to the loss

of either L or H₂and, due to some fortuitous combina-

tion of rates, the formation of the product effectively

stops the transformation. This possibility is clearly

inconsistent with the fact that pure 1-Et in solution

produces no observable trace of 1-H after 15 min.

Nevertheless, reaction of 1-OTf with a C₆D₆solution of

EtLi, also containing 10 equiv of PⁱPr₃, gave 79% of 1-H,

while the analogous reaction under an H₂atmosphere

produced 78% of 1-H (cf. Table 1). In the latter case,

the small amount of 2-C₂H₄typically formed in the

absence of H₂(ca. 5% yield) was absent, indicating that

2 that was produced was trapped by the added H₂as

1-H and therefore increasing the observed yield of the

latter. Such variations between the yields obtained in

the presence of H₂and L and in their absence (Table 1)

are insignificantly small. Thus, these results rule out a

â-hydride elimination mechanism as the source of 1-H

in the reaction of 1-OTf with EtLi, which therefore

proceeds via direct â-hydride transfer.

F igu r e 1. Time evolution of 1-Y (circles) and [1-Me]/[1-

CH₂SiMe₃] ratio (squares) in the reaction of 1-OTf with

15 equiv of Me₃SiCH₂Li, monitored by H NMR at -70 °C

in d₈-PhMe.

1

1-OTf and Me₃SiCH₂Li to give 1-H, bypassing the

1-CH₂SiMe₃intermediate (to give 1-H-d in the D₂

reaction),⁵³they also show that within 10 min up to 10%

of the latter can be converted to 1-H, via intermediate

2 (to give 1-H-d ₂in the D₂reaction). In the absence of

added H₂, this amount of 2 probably converts to 1-P ∼C,

(b) â-R′ ) Me. Having established the â-hydride

elimination mechanism inoperative in 1-Et suggests

that the analogous â-Me^-elimination is unlikely in

principle, in light of the much more frequent occurrence

of the former. Indeed, experiments analogous to those

performed for the â-H^-case lead to the same conclu-

sions. Monitoring the reaction of 1-OTf with 2 equiv of

1

not observed by H NMR presumably due to the broad-

ness of its hydride signals.³³Thus, the attempted

suppression of either dissociative mechanism gives a

negligible effect.

1

Finally, following the evolution of the products

1-CH₂SiMe₃and 1-Me by ¹H NMR in a reaction of

1-OTf with 15 equiv of Me₃SiCH₂Li in d₈-PhMe at -70

°C (Figure 1) provides direct evidence against the

intermediacy of 1-CH₂SiMe₃in the formation of 1-Me.⁵⁴

While the formation of 1-Me follows pseudo-first-order

kinetics, with k(1-Me) ) [3.8(1)] × 10^-4s^-1, the evolu-

tion of 1-CH₂SiMe₃has an induction period, effectively

leading to a smaller k(1-CH₂SiMe₃) ) [1.5(1)] × 10^-4

s^-1, as does the decay of 1-OTf, such that the plot of

ln([1-OTf]) vs time is curved upward and its loss is

Me₃SiCH₂Li in d₈-PhMe by H NMR at -40 °C shows

the reaction proceeding to completion within minutes

at this temperature. A 50:50 mixture of 1-CH₂SiMe₃

and 1-Me, generated in a separate experiment with 4

equiv of Me₃SiCH₂Li at the same temperature, re-

mained unchanged after several hours at that temper-

ature and temperatures of up to 0 °C. In the reaction of

1-OTf with 15 equiv of Me₃SiCH₂Li and 20 equiv of

TMEDA in d₈-PhMe at -60 °C, a 94:6 mixture of the

above products was produced on the time scale of hours

and again remained unchanged on warming to room

temperature. Under ambient conditions in C₆D₆both

1-CH₂SiMe₃and 1-Me reductively eliminate Me₄Si and

CH₄, respectively, with the former reaction proceeding

ca. 10 times faster. While it was impossible to isolate

pure 1-CH₂SiMe₃, the low-temperature generation

results clearly rule out a nondissociative â-Me^-elimina-

tion mechanism, since 1-CH₂SiMe₃does not convert to

1-Me once its formation is complete (i.e. once 1-OTf is

consumed). When the two dissociative mechanisms were

tested, 1-OTf was reacted with a C₆D₆solution of Me₃-

SiCH₂Li, also containing 10 equiv of PⁱPr₃, to give a 56:

44 mixture of 1-CH₂SiMe₃and 1-Me, while reaction

under an H₂atmosphere led to a 54:46 ratio of the

products. Notably, the latter reaction also gave 23% of

1-H. However, the analogous reaction under D₂gave

ca. 10% of each of the 1-H-d and 1-H-d ₂isotopomers,³⁴

together with a 55:45 ratio of 1-CH₂SiMe₃and 1-Me;

for comparison, exposing the reaction mixture generated

from 1-OTf and Me₃SiCH₂Li to an atmosphere of H₂2

min after mixing of the reagents showed only 10% of

1-H with a 60:40 ratio of the major products, after 10

min. While these results indicate that H₂can react with

described by a lowered k(1-OTf) ) [1.6(1)] × 10^-4s^-1

.

Since the growth of 1-Me levels off, rather than decay-

ing, if it somehow were an intermediate for the forma-

tion of 1-CH₂SiMe₃, the acceleration of pathway B is

apparently caused by the release of LiOTf, presumably

due to the formation of mixed aggregates of LiOTf with

Me₃SiCH₂Li that exhibit reactivity different from that

(53) Me₃SiCH₂Li possibly undergoes hydrogenolysis prior to the

reaction with 1-OTf, giving LiH, the source of 1-H-d in the D₂

reaction: Gilman, H.; J acoby, A. L.; Ludeman, H. J . Am. Chem. Soc.

1938, 60, 2336. Screttas, C. G.; Eastham, J . F.; Kamienski, C. W.

Chimia 1970, 24, 109. Screttas, C. G.; Eastham, J . F. J . Am. Chem.

Soc. 1966, 88, 5668.

(54) (a) The presence of a large excess (>5 equiv) of RLi or LiBR₄in

solutions of dihydrido alkyl/aryl complexes 1-R with poor solvation of

Li causes a considerable increase in the difference of the hydride

signals ∆δ ) δ(H_A) - δ(H_M) due to the formation of “lithium bonding”

interactions^54bwith the nitrosyl oxygen, and not due to the presence

of paramagnetic species. These intermolecular interactions are tem-

perature-dependent, and formation of soluble RLi/LiOTf aggregates

effects a greater increase in the ∆δ parameter. Strong acceleration of

the reductive elimination of RH results from such lithium bonding

interactions, as detailed elsewhere. Analogous lithium bonding interac-

tions with the X ligand of halide/pseudohalide complexes 1-X lead to

a

similar increase in the ∆δ value. (b) Scheiner, S. In Lithium

Chemistry: Theoretical and Experimental Overview; Sapse, A.-M.,

Schleyer, P. v. R., Eds.; Wiley: New York, 1995; Chapter 3.

4036 Organometallics, Vol. 21, No. 20, 2002

Yandulov et al.

Sch em e 4. Syn th esis of 3-H a n d 3-Li

of the pure Me₃SiCH₂Li. These results clearly demon-

strate that the ratio of 1-Me to 1-CH₂SiMe₃decreases

with time (Figure 1), thus ruling out the intermediacy

of the latter in the formation of the former.

Therefore, these results rule out â-Me^-elimination

from 1-CH₂SiMe₃as a mechanism for formation of

1-Me in the reaction of 1-OTf with Me₃SiCH₂Li, which

thus proceeds via direct â-Me^-transfer.

(c) Oth er RLi a n d BR₄^-. Despite the above conclu-

sions, one can envision a â-R′^-elimination mechanism

operating for the products of pathway B for the other

reagents: (Me₃Si)₂NLi, (Me₃Si)₂CHLi, TMPLi, LDA,

(3-H), and EA (3-Li).⁵⁶The formation of 3-Li in the

above reaction likely proceeds via 1,2-addition of (Me₃-

Si)₂N-Li to the transient Me₃SiNdSiMe₂, either as a

true intermediate or not (i.e., â-Me^-transfer may occur

concurrently with (Me₃Si)₂N^-attack at the Si that

donates Me^-). Silanimines are highly reactive spe-

cies,^57-60becoming isolable only with heavy substitution

at both Si and N,⁶¹that undergo a variety of transfor-

mations,^57-59,62such as adduct formation at Si, 1,2-

addition and cycloaddition of a number of reagents, and

oligomerization. Because 1,2-addition to a SidN double

bond is symmetry-forbidden, such reactions proceed via

a polar mechanism,^57,59in which a nucleophile (e.g., O

in ROH) first attacks Si, followed by migration of an

electrophile (H in ROH) to N. Although we are not

aware of studies of 1,2-addition of lithium alkylamides

to silanimines, such reactions are expected to be facile,

considering the high nucleophilicity of N in R₂N-Li and

the fact that the N-Li bond is easily ionizable, provided

the addition of the R₂N^-nucleophile is not sterically

prohibitive. In fact, Wiberg and co-workers have re-

ported on analogous 1,2-addition of alkyllithium re-

agents to silanimines⁶³and silenes.^64,65This reasoning

also suggests that 3-Li, always observed in the -60 °C

d₈-THF reaction above together with 3-H, is not formed

via 1,2-addition of (Me₃Si)₂N-H to Me₃SiNdSiMe₂and

subsequent lithiation by (Me₃Si)₂NLi but rather via 1,2-

addition of (Me₃Si)₂N-Li, due to the lower nucleophi-

licity of (Me₃Si)₂NH compared to that of (Me₃Si)₂NLi.

i

^tBuLi, and Pr₃BMe^-(R ) ⁱPr), since any such transient

intermediates, not observed at 20 °C and even at -60

°C in the case of (Me₃Si)₂NLi, very likely would experi-

ence a considerable steric repulsion with PⁱPr₃ligands,

thus encouraging their loss, or any other transformation

considered above, to generate an unsaturated species

necessary for the migration. However, such steric repul-

sion would also facilitate the reductive elimination of

RH/R₂NH, as is clearly evident from the relative stabil-

ity of 1-R alkyls (Scheme 3). Since only a small amount

of the products derived from 2 is observed, and only for

TMPLi, the products of B likely never form for all of

-

the other RLi and BR₄reagents (except possibly

TMPLi). Therefore, we conclude that pathway A, fol-

lowed with all reagents considered except NpLi and

possibly TMPLi, is independent of pathway B and the

â-R′^-transfer occurs in a direct fashion from RLi and

-

BR₄

.

IV. F a te of th e Or ga n ic P r od u cts in P a th w a y A.

The expected organic products were identified by ¹H

NMR (not quantified) in the C₆D₆reactions of 1-OTf

t

with BuLi (isobutylene), ⁿBuLi (1-butene), EtLi (eth-

ylene), LiBEt₄(ethylene and BEt₃as the only volatile

products), and [Li(TMEDA)₂][ⁱPr₃BMe] (propylene; BR₃

not analyzed). When â-R′^-transfer leads to the forma-

tion of unsaturated silicon compounds (formally), silenes

((Me₃Si)₂CHLi, Me₃SiCH₂Li, and [Li(THF)₄][B(CH₂-

SiMe₃)₄]) and silanimine ((Me₃Si)₂NLi), and imines

(TMPLi and LDA), the organic products of pathway A

underwent subsequent transformations, described in

detail below.

Notably, 3-Li also exhibits â-Me^-transfer reactivity,

quantitatively forming 1-Me in a C₆D₆reaction with

(a ) Sila n im in es a n d Silen es. The C₆D₆reactions of

1-OTf with (Me₃Si)₂NLi and Me₃SiCH₂Li produce highly

complex mixtures of organic products, as evidenced by

observation of numerous sharp signals in the (-0.5 to

0.8 ppm) (Si)Me region of ¹H NMR spectra, while several

broad features are observed in that region in the case

of (Me₃Si)₂CHLi. However, using 4 equiv of (Me₃Si)₂NLi

in a d₈-THF reaction with 1-THF ⁺at -60 °C (conditions

under which (Me₃Si)₂NLi exists as a solvated monomer

to a significant extent⁵⁵) gave very different results. On

the time scale of hours at -60 °C, this reaction produces

the trisilazide 3-Li (Scheme 4) in an equilibrium

mixture with trisilazane 3-H, (Me₃Si)₂NLi, and (Me₃-

Si)₂NH, due to hydrolysis by trace water, as essentially

the only organic products (>90% total yield of 3-Li +

3-H by ¹H NMR integration vs 1-Me) accompanying the

quantitative formation of 1-Me. The trisilazane deriva-

tives were characterized by independent synthesis

(56) Fink, W. Chem. Ber. 1963, 96, 1071.

(57) Wiberg, N. J . Organomet. Chem. 1984, 273, 141.

(58) Raabe, G.; Michl, J . In The Chemistry of Organic Silicon

Compounds; Patai, S., Rappoport, Z., Eds.; Wiley: New York, 1989;

Chapter 17.

(59) Raabe, G.; Michl, J . Chem. Rev. 1985, 85, 419.

(60) Hemme, I.; Klingebiel, U. Adv. Organomet. Chem. 1996, 39,

159.

(61) Wiberg, N.; Schurz, K.; Fischer, G. Angew. Chem., Int. Ed. Engl.

1985, 24, 1053. Wiberg, N.; Schurz, K.; Reber, G.; Mu¨ller, G. J . Chem.

Soc., Chem. Commun. 1986, 591. Hesse, M.; Klingebiel, U. Angew.

Chem., Int. Ed. Engl. 1986, 25, 649. Wiberg, N.; Schurz, K. Chem. Ber.

1988, 121, 581. Niesmann, J .; Klingebiel, U.; Scha¨fer, M.; Boese, R.

Organometallics 1998, 17, 947.

(62) Wiberg, N.; Preiner, G.; Karampatses, P.; Kim, C.-K.; Schurz,

K. Chem. Ber. 1987, 120, 1357. Wiberg, N.; Schurz, K. J . Organomet.

Chem. 1988, 341, 145. Zigler, S. S.; J ohnson, L. M.; West, R. J .

Organomet. Chem. 1988, 341, 187. Walter, S.; Klingebiel, U.; Schmidt-

Ba¨sse, D. J . Organomet. Chem. 1991, 412, 319. Niesmann, J .; Klinge-

biel, U.; Rudolph, S.; Herbst-Irmer, R.; Noltemeyer, M. J . Organomet.

Chem. 1996, 515, 43. Niesmann, J .; Klingebiel, U.; Noltemeyer, M. J .

Organomet. Chem. 1996, 521, 191.

(63) Wiberg, N. In Organosilicon Chemistry II. From Molecules to

Materials; Auner, N., Weis, J ., Eds.; VCH: New York, 1996; p 405.

(64) Wiberg, N. In ref 63, p 367.

1

(Scheme 4), H (3-H, 3-Li) and ²⁹Si NMR (3-H), EI-MS

(65) Wiberg, N.; Passler, T.; Wagner, S.; Polborn, K. J . Organomet.

Chem. 2000, 598, 292. Wiberg, N.; Passler, T.; Wagner, S. J . Orga-

nomet. Chem. 2000, 598, 304.

(55) Kimura, B. Y.; Brown, T. L. J . Organomet. Chem. 1971, 26,

227.

Lithium Bases as Sources of Methyl Anion

Organometallics, Vol. 21, No. 20, 2002 4037

Sch em e 5. P r op osed Mech a n ism for Oligom er iza tion of Sila za n es in th e Rea ction of 1-OTf w ith

(Me₃Si)₂NLi

Sch em e 6. P r op osed Mech a n ism for th e

1-OTf within 10 min at 20 °C, in addition to a mixture

of organic products even more intractable than that

F or m a tion of 5

obtained with (Me₃Si)₂NLi. In view of the above results

it appears likely that the complex mixture of organic

products formed in a 20 °C C₆D₆reaction of (Me₃Si)₂NLi

with 1-OTf is composed of branched oligo- and cyclosi-

lazanes. The latter are produced in a cascade fashion

via successive â-Me^-transfer from R₂NLi and 1,2-

addition of R₂N-Li to putative R₂SidNR intermediates;

cyclodimerization of these terminates the reaction se-

quence (Scheme 5). Although neither 3-Li nor 3-H is

observed in the C₆D₆reaction of 1-OTf with (Me₃-

Si)₂NLi, 1,2-addition of R₂N-Li is likely to compete with

dimerization of R₂SidNR in C₆D₆, at least at the initial

stages of the reaction, since the presence of only a 4-fold

excess of (Me₃Si)₂NLi in a d₈-THF reaction with 1-THF ⁺

at -60 °C already leads to a nearly quantitative yield

of 3-Li, which, in turn, reacts with 1-OTf in C₆D₆

quantitatively via â-Me^-transfer. The reaction sequence

proposed in Scheme 5 should yield branched products,

i.e., containing N(SiMe₂R)₃centers, that are not acces-

sible via oligomerization of Me₃SiNdSiMe₂. It is also

notable that ¹H NMR signals reported⁶⁶for cyclodisi-

lazane 4 (Scheme 5) were not observed in the reactions

of 1-OTf with either (Me₃Si)₂NLi or 3-Li, suggesting

that 4 is at least not the major reaction product in either

case. In contrast, the corresponding cyclodisilazane was

the predominant product in an analogous type of reac-

tion, formal transfer of â-H^-from lithium tetrameth-

yldisilazide to organic electrophiles, investigated by

Seyferth et al.⁶⁷

In close analogy to silanimines, silenes are also highly

reactive species, exhibiting similar reactivity.^57-59,68-69

Although Me₃SiCH₂SiMe₂CH₂Li, the analogue of 3-Li,

could not be clearly identified in the reaction of 1-OTf

with 15 equiv of Me₃SiCH₂Li in d₈-PhMe at -70 °C, 20

°C reactions of 1-OTf with Me₃SiCH₂Li invariably gave

small amounts of (Me₃Si)₂CH₂, the analogue of 3-H,

which suggests that a reaction sequence analogous to

that outlined for Me₃SiNdSiMe₂in Scheme 5 operates

in the case of H₂CdSiMe₂as well, and likely for (Me₃-

Si)HCdSiMe₂, the formal product of (Me₃Si)₂CHLi in

1

pathway A, also. Notably, all H and ³¹P NMR signals

of 1-CH₂SiMe₃, formed from 1-OTf and Me₃SiCH₂Li,

often overlapped a set of analogous, very closely situated

signals, possibly corresponding to the product of reaction

of 1-OTf and Me₃SiCH₂SiMe₂CH₂Li via pathway B: i.e.,

1-CH₂SiMe₂CH₂SiMe₃. The NMR signals of the latter

were integrated up to 10% of the main product,

1-CH₂SiMe₃, and both were always summed together

in reporting the yield in Table 1 according to pathway

B in the Me₃SiCH₂Li reactions.

In contrast to the complex mixtures of organic prod-

ucts obtained from the above RLi and R₂NLi reagents

in the reactions with 1-OTf in C₆D₆, â-Me^-transfer

from [Li(THF)₄][B(CH₂SiMe₃)₄] under analogous condi-

tions gave the novel borane (Me₃SiCH₂)₂B(CH₂SiMe₂-

CH₂SiMe₃) (5) together with B(CH₂SiMe₃)₃, the product

of B, as essentially the only organic products. The

1

formation of 5, characterized by H NMR⁷⁰and EI-MS

(M⁺at m/z 344), may proceed via â-Me^-transfer

concurrent with attack of another alkyl group at the Si

that donates Me^-(Scheme 6).²⁶The reaction of 1-OTf

with a C₆D₆solution containing Me₃SiCH₂Li and 1.6

equiv of B(CH₂SiMe₃)₃(which do not react to an observ-

able extent) gave no trace of 5. While this result shows

that reactivity of [Li(THF)₄][B(CH₂SiMe₃)₄] via pathway

A involves genuine borate rather than Me₃SiCH₂Li, it

(66) Wiberg, N.; Preiner, G.; Schieda, O. Chem. Ber. 1981, 114, 3518.

(67) Wiseman, G. H.; Wheeler, D. R.; Seyferth, D. Organometallics

1986, 5, 146.

(68) Wiberg, N.; Wagner, G. Chem. Ber. 1986, 119, 1467.

(69) Brook, A. G.; Baines, K. M. Adv. Organomet. Chem. 1986, 25,

1.

(70) A ¹¹B NMR spectrum recorded on a 89:11 mixture of 5 and

B(CH₂SiMe₃)₃in C₆D₆showed a several ppm broad signal at 78.6 ppm,

indistinguishable from that of pure B(CH₂SiMe₃)₃.

4038 Organometallics, Vol. 21, No. 20, 2002

Yandulov et al.

Sch em e 7. Rea ctivity of Im in es F or m ed in th e

Rea ction s of TMP Li a n d LDA w ith 1-OTf

does not rule out formation of free H₂CdSiMe₂in the

B(CH₂SiMe₃)₄^-reaction, which subsequently reacts with

B(CH₂SiMe₃)₃via 1,2-addition of a B-C bond, because

the silene may react much faster with Me₃SiCH₂Li.

However, it is notable that the reaction of BEt₄via A

does not follow the mechanism of Scheme 6, producing

exclusively ethylene and BEt₃instead.

F igu r e 2. ORTEP representation of cis,trans-Os(H)₂-

(C₈H₁₄N)(NO)(PⁱPr₃)₂(1-C₈H₁₄N) with phosphine methyl

groups and selected hydrogens omitted, showing the atom

labeling and a N(10)‚‚‚H-C(17) hydrogen bond. Thermal

ellipsoids are drawn at the 50% probability level.

-

Ta ble 3. Deta ils of X-r a y Str u ctu r e

Deter m in a tion s of

(b) Im in es. In contrast to the unsaturated silicon

transients, which react via 1,2-addition as discussed

above, the imines formed from TMPLi and LDA in

pathway A undergo deprotonation of the R-CH₃group

and the resulting Li azaallyls react with remaining

1-OTf via OTf^-metathesis, competing for 1-OTf with

the primary reaction pathways (Scheme 7).⁷¹Thus, as

described earlier,⁷²reaction of 1-OTf with a C₆D₆

solution of TMPLi gives 1-C₈H₁₄N in 33% yield, in

addition to 1-Me (50%), as well as products derived from

2 (13%) and 1-Cl (2%) and 1-OH (2%), which result from

LiX impurities in TMPLi. While using a solution of 6

equiv of TMPLi does not significantly decrease the ratio

1-C₈H₁₄N:1-Me, it can be increased essentially to the

theoretical maximum of 50:50 if the steady-state con-

centration of TMPLi is limited by reacting 1-OTf with

large crystals of TMPLi instead of a homogeneous

solution. Furthermore, the reaction of solid 1-OTf and

TMPLi with a C₆D₆solution of 1.4 equiv of the corre-

sponding imine, 1,2-dehydro-2,6,6-trimethylpiperidine,

gives 1-C₈H₁₄N in 90-95% yield, in addition to trace

1-H and free L. Formation of 1-Me is completely

suppressed in this reaction, demonstrating that imine

lithiation and subsequent OTf^-metathesis with 1-OTf

are both considerably faster than â-Me^-transfer from

TMPLi.

cis,tr a n s-Os(H)₂(C₈H₁₄N)(NO)(P ⁱP r ₃)₂(1-C₈H₁₄N)

a n d [cis,tr a n s-Os(H)₂((THF )(NO)(P ⁱP r ₃)₂]-

[B(C₆H₃-3,5-(CF ₃)₂)₄]‚THF (1-THF ⁺)

1-C₈H₁₄

N

1-THF ⁺

formula

fw

C

₂₆H₅₆N₂OOsP₂

C

₅₈H₇₀BF₂₄NO₃OsP₂

666.94

triclinic

P1h

1550.16

monoclinic

P2₁/c

cryst syst

space group

color of cryst

a, Å

pale yellow

8.8330(2)

11.5202(3)

16.0550(4)

103.704(1)

99.448(1)

98.890(1)

-160

yellow

15.183(2)

17.785(2)

24.578(4)

b, Å

c, Å

R, deg

â, deg

γ, deg

T, °C

97.41(1)

-160

4

Z

2

R(F_o)^a

0.0159

0.0134

0.0627

0.0487

1.082

0.03

R_w(F_o)^b

GOF for the last cycle 0.549

max ∆/σ for last cycle

0.002

a

b

2

R ) ∑||F_o| - |F_c||/∑|F_o|. R_w) [∑w(|F_o| - |F_c|)²/∑w|F_o| ]^1/2

,

where w ) 1/σ²(|F_o|).

distorted by compression of the cis-C-Os-H and cis-

H-Os-H angles to 80.7(8) and 72.9(11)°, respectively,

most reliably identified by the increase in the cis-C-

Os-N angle from 90 to 104.15(7)°. This type of distor-

tion, which allows the ligands with strong trans influ-

ence, hydride and alkyl, to avoid being perfectly trans

to any other ligand in the molecule, has been identified

previously³⁴in the structures of mer,trans-M(H)₃(NO)-

L₂(M ) Ru, Os, L ) PR₃) and 1-Cl and rationalized on

the grounds of strengthening M-H σ-bonds at the

expense of increased MfNO π-back-bonding. The dis-

tortion in 1-C₈H₁₄N explains the conformation of the

C₈H₁₄N ligand around the Os(1)-C(4) bond, in which

the C₅N ring is oriented away from the smaller H(1)

and toward the bulkier NO ligand (dihedral N(2)-Os-

(1)-C(4)-C(5) ) 26.5°) and also takes place for the

The dihydride 1-C₈H₁₄N, isolated in 69% yield from

the reaction of 1-OTf with TMPLi in the presence of

the imine, and fully characterized⁷²by NMR and X-ray

diffraction, reductively eliminates 1,2-dehydro-2,6,6-

trimethylpiperidine in C₆D₆on the time scale of hours

with a rate intermediate between that of 1-Me and 1-Et,

in close analogy to the other 1-R complexes. The crystal

structure determination of 1-C₈H₁₄N (Figure 2, Tables

3 and 4) revealed an octahedral geometry significantly

(71) Analogous subsequent transformations of these or closely

related imines takes place with organic electrophiles.^16,27

(72) Yandulov, D. V.; Huffman, J . C.; Caulton, K. G. New J . Chem.

2000, 24, 649.

Lithium Bases as Sources of Methyl Anion

Organometallics, Vol. 21, No. 20, 2002 4039

(BAr′₄^-) and no ligand (BAr′₄^-) were used to evaluate

their reactivity toward X-ligand substitution with se-

lected reagents in various solvents (the qualitative rate

dependence is shown in Table 5). Aside from the product

distribution, all of these reactions led to products (via

A and B) identical with those discussed for 1-OTf above.

Ta ble 4. Selected In ter a tom ic Dista n ces (Å) a n d

An gles (d eg) for

cis,tr a n s-Os(H)₂(C₈H₁₄N)(NO)(P ⁱP r ₃)₂(1-C₈H₁₄N)

Os(1)-P(13)

Os(1)-N(2)

Os(1)-H(1)

N(2)-O(3)

N(10)-C(9)

2.3673(5)

1.7748(16)

1.599(22)

1.1940(20)

1.4840(26)

Os(1)-P(23)

Os(1)-C(4)

Os(1)-H(2)

N(10)-C(5)

N(10)‚‚‚C(17)

2.3706(5)

2.2396(19)

1.668(25)

1.2804(25)

3.368(3)

The primary precursor to all of the 1-Y complexes,

the chloride 1-Cl, is very inert toward organolithium

P(13)-Os(1)-N(2) 101.98(5)

P(23)-Os(1)-N(2) 97.73(5)

P(23)-Os(1)-C(4) 92.46(5)

P(23)-Os(1)-H(1) 81.9(8)

P(23)-Os(1)-H(2) 82.7(9)

n

reagents in general. Reactions with MeLi, BuLi, and

P(13)-Os(1)-C(4)

P(13)-Os(1)-H(1)

P(13)-Os(1)-H(2)

92.57(5)

77.6(8)

83.2(9)

^tBuLi in C₆D₆or C₆D₁₂proceed slowly, on the time scale

of hours at 20 °C, producing considerable amounts of

free L in addition to the 1-R products formed in the

corresponding 1-OTf reactions (Table 1), while (Me₃-

Si)₂NLi‚OEt₂shows no reaction after 18 h (entry 1). The

formation of substantial amounts of free L in these

reactions is suggestive of a competitive operation of a

single-electron-transfer mechanism, although it could

also implicate the deprotonation mechanism C. The

unsolvated lithium borates (entries 2 and 4) are among

the few reagents that afford clean conversion of 1-Cl

on reasonable time scales.

P(13)-Os(1)-P(23) 157.782(17) H(1)-Os(1)-H(2)

72.9(11)

Os(1)-N(2)-O(3)

C(4)-Os(1)-H(1)

N(2)-Os(1)-H(1)

N(2)-Os(1)-C(4)

N(10)-C(5)-C(4)

177.15(15) Os(1)-C(4)-C(5) 118.25(13)

80.7(8)

175.2(8)

104.15(7)

C(4)-Os(1)-H(2) 153.5(9)

N(2)-Os(1)-H(2) 102.3(9)

C(5)-N(10)-C(9) 121.53(18)

118.93(18) N(10)-C(5)-C(6) 124.53(18)

phosphines,³⁴such that cis-P-Os-N angles, 101.98(5)

and 97.73(5)°, are considerably greater than 90°. All

hydrogens of the C₅N ring were located, and the CdN

double bond was unambiguously identified as N(10)-

C(5) (1.2804(25) Å, as compared to the single N(10)-

C(9) bond of 1.4840(26) Å). Notably, the imine nitrogen

points toward a phosphine methine C-H bond, possibly

forming a weak hydrogen bond (N(10)‚‚‚C(17) ) 3.368-

(3) Å) that may lock the alkyl group in this position,

preventing a disorder.

The fluoride derivative 1-F is substantially more

reactive than 1-Cl, affording 1-Me in a C₆D₆reaction

with (Me₃Si)₂NLi‚OEt₂on the time scale of minutes

(entries 1 and 5). The fluoride complex, generated in

situ from 1-OTf and [Bu₄N][Ph₃SiF₂], is unstable to-

ward H₂loss, producing OsF(NO)L₂in solution on the

time scale of hours, which, together with H₂, exists in a

kinetically slow equilibrium with 1-F , such that both

Os complexes are observed in a constant ratio after

prolonged periods of time in a closed system, and this

ratio can be reversibly perturbed by raising the tem-

perature to 90 °C.³³This inherent reactivity of 1-F leads

to the formation of a trace amount of 1-H in the entry

5 reaction and illustrates one of the pathways 1-H is

formed in the reactions generating 2. Adventitious

water or trace methanol of crystallization may afford

the corresponding 1-OR species by oxidative addition,

which, in close analogy to 1-F , slowly lose H₂(as

established for 1-OH³³) scavenged by the remaining 2

to give comparable amounts of 1-H and Os(OR)(NO)L₂.

The latter is not always observed by ³¹P NMR, indicat-

ing that hydride ligand scavenging is also operative in

the formation of 1-H.⁷³

The triflate derivative 1-OTf reacts with all reagents

listed in Tables 1 and 2 in C₆D₆essentially instanta-

neously at 20 °C. The reaction with (Me₃Si)₂NLi‚OEt₂

is also instantaneous in alkanes, Et₂O, and neat (Me₃-

Si)₂NH (entry 6) but proceeds slowly in d₈-THF, requir-

ing over 2 h for full conversion. This allows us to

compare the reactivity of 1-OTf and 1-THF ⁺, and the

latter reacts instantly at 20 °C, thus demonstrating that

OTf^-inhibits the â-Me^-transfer. Using the least

reactive reagent, B(CH₂SiMe₃)₄^-, allows evaluation of

the influence of the coordinated THF in 1-THF ⁺, and

entries 11 and 12 show that THF suppresses the

reactivity of 1⁺, the most reactive form of osmium

species abstracting â-Me^-. The reaction in entry 12

The imine released via pathway A of the 1-OTf

reaction with a C₆D₆solution of LiNⁱPr₂(LDA) under-

went subsequent transformation analogous to that

found in the case of TMPLi. Two products were observed

in the case of LDA, in unequal ratio: the minor product

with -7.21 and -8.35 ppm hydride signals, essentially

identical with those of 1-C₈H₁₄N (-7.23 and -8.37), and

the major product with hydride signals at -6.86 and

-8.24 ppm, both with similar ³¹P chemical shifts of 30.0

(30.1 for 1-C₈H₁₄N) and 28.3 ppm, respectively. On the

basis of the similarity of the spectroscopic data of the

dihydrides formed in the LDA reaction and those of

1-C₈H₁₄N, the former species likely correspond to the

isomers of cis,trans-Os(H)₂(CH₂C(Me)dNⁱPr)(NO)L₂, with

cis and trans configurations around the CdN double

bond. Indeed, subjecting the reaction mixture to a 60

°C thermolysis for 14 h quantitatively transformed both

compounds into 1-P h -d ,d ₅isomers via the intermediacy

of 2,³³while ¹H NMR analysis of the volatile components

revealed the presence of only HNⁱPr₂, ⁱPrNdCMe₂, and

trace PⁱPr₃. In contrast to the TMPLi reaction, the

i

metalated derivatives of PrNdCMe₂were formed in

only 4% yield, comprising all of the observed products

together with the product of A, 1-H (96%). The de-

creased yield of the metalated imine products formed

with LDA relative to that formed with TMPLi is

consistent with â-H^-transfer being inherently faster

than â-Me^-transfer, as determined from the strong

kinetic preference for the former in the LDA reaction

(Table 1). In analogy to TMPLi, the yield of metalated

imines also increases, up to 20%, if large crystals of LDA

are used in the reaction with 1-OTf instead of a

homogeneous solution.

V. Mech a n istic In sigh ts. (a ) In flu en ce of th e Os

Lea vin g Gr ou p . The rates of A and B in the reactions

of 1-X with organolithium bases and borates strongly

depend on the nature of X and the electrophilicity of

Li. A series of 1-X complexes with X ) Cl, F , OTf, THF ⁺

i

(73) Dehydrogenation of an Pr group of PⁱPr₃via â-hydride elimina-

tion from the metalated 1-P ∼C intermediate as a route to 1-H is

inconsistent with the ³¹P chemical shift being essentially identical with

that of the authentic 1-H; additionally, 1-H is often formed in the

analogous reactions with L ) P^tBu₂Me, also with a ³¹P chemical shift

identical with that of the authentic trihydride, in which case â-hydride

elimination from the metalated species is not possible.

4040 Organometallics, Vol. 21, No. 20, 2002

Yandulov et al.

-

Ta ble 5. Qu a lita tive Dep en d en ce of th e X-Su bstitu tion Ra tes in 1-X by Selected RLi, R₂NLi, a n d Q⁺BR₄on

th e Na tu r e of X a n d Solven t a t RT

-

entry

X

RLi/R₂NLi/Q⁺BR₄

solvent

time

conversn, %

1

2

3

4

5

6

7

8

9

Cl

F

OTf

Me₃SiCH₂Li; (Me₃Si)₂NLi‚OEt₂

LiBMe₄

C₆D₁₂

C₆D₆

Et₂O

C₆D₆

15 min; 18 h

20 min

15 min

35 min; 2 h

10 min; 1 h

10 min

10 min; 2 h

10 min

15 min

10 min

0

73

0

LiBEt₄

25; 41

10; 71

100

38; 77

100

73

100

14

100

(Me₃Si)₂NLi‚OEt₂

[Bu₄N][B(CH₂SiMe₃)₄]

[Li(THF)₄][B(CH₂SiMe₃)₄]

(Me₃Si)₂NLi‚OEt₂

[Li(THF)₄][B(CH₂SiMe₃)₄]

[Bu₄N][B(CH₂SiMe₃)₄]

Et₂O, C₆D₆, C₆D₁₂, (Me₃Si)₂NH

C₆D₆

d₈-THF

d₈-PhMe/PhF

OTf

10

11

12

THF⁺

Sch em e 8. Rela tive Or d er of th e Rea ctivity

of 1-X by X Su bstitu tion

proceeds via A (63%) and B (37%), similar to that of [Li-

(THF)₄][B(CH₂SiMe₃)₄] with 1-OTf in C₆D₆(Table 2).

The highly electrophilic five-coordinate [cis,trans-Os-

(H)₂(NO)L₂][B(C₆H₃-3,5-(CF₃)₂)₄] (1⁺(BAr′₄^-)), originally

prepared in PhF from 1-BF ₄and NaBAr′₄,⁷⁴can be

conveniently generated via hydride abstraction from

1-H by [Ph₃C][BAr′₄].³⁴While 1⁺invariably forms oils,

largely precluding its isolation, the THF adduct 1-THF ⁺

is crystalline and was isolated in 81% yield. The X-ray

structure determination of 1-THF ⁺(Table 3; Figure S2,

Supporting Information) revealed the presence of a

coordinated THF molecule; however, extensive disorder

precluded meaningful analysis of geometric parameters.

A second THF molecule is present in the lattice but is

not involved in hydrogen bonding to the hydride ligands

(not located), as the closest lattice THF oxygen is 7.13

Å from Os and THF is pointing away from the metal,

X ) Cl < F < OTf < THF⁺(BAr′₄^-) < - (BAr′₄

)

-

(b) Th e Role of Lith iu m . In addition to the X-ligand

dependence, the rates of X^-substitution are strongly

influenced by the extent of solvation of Li. Entries 7 and

+

8 in Table 5 show that Li(THF)₄imparts higher

reactivity to B(CH₂SiMe₃)₄than does Bu₄N⁺, while

-

entries 6 and 9 reveal strong suppression of the rate of

reaction with (Me₃Si)₂NLi‚OEt₂by THF. Analogous

suppression of the rate by solvation of Li is evident in

the LiBEt₄reactions in entries 3 and 4. Overall, the

unsolvated lithium borates are much more reactive

toward 1-Cl than any RLi or R₂NLi compound studied,

consistent with the minimal stabilization of Li⁺possible

in LiBR₄via interactions with the C-H bonds.^39,75These

results demonstrate the electrophilic role that lithium

plays in assisting the loss of the X^-in the reactions with

1-X studied here. Thus, the higher reactivity of 1-F (vs

1-Cl) is caused by the stronger Li‚‚‚X interaction with

the harder Lewis base, F.

(c) Lit h iu m Bon d in g In t er a ct ion s w it h X(Os).

While no strong evidence for intermediates in pathways

A and B have been obtained, monitoring slow reactions

of organolithium reagents with several 1-X compounds

by ¹H NMR revealed lithium bonding interactions with

the X ligand. Remarkably similar to the lithium bonding

interactions with NO oxygen⁵⁴in their effect on the

hydride chemical shifts, the interactions with the X

ligand also cause a substantial increase in the ∆δ

parameter (δ(H_A) - δ(H_M)), by effectively reducing the

trans influence of X. Thus, the ∆δ value of 1-Cl was

found to be 0.62 ppm greater than normal during the

reaction with LiBMe₄(Table 5, entry 2), while in the

case of LiBEt₄, this increase was 1.13 ppm at 25%

conversion and decreased to 0.90 ppm at 41% conver-

sion, due to the consumption of LiBEt₄and precipitation

of LiCl. In the case of 1-F (entry 5), the increase in ∆δ

amounted to 0.55 ppm after 10 min and also decreased

to 0.15 ppm at 71% conversion. In neither case were

the ∆δ values of the 1-R products significantly affected,

thus ruling out interactions with NO oxygen for 1-X.

The latter can be expected to be more pronounced for

1-R than for 1-X, since ν(N-O) of 1-Me is 26 cm^-1lower

than that of 1-Cl, for example. Although the trans J _F-H

coupling in 1-F of 69.3 Hz was not significantly affected

-

toward BAr′₄aromatic hydrogens. The absence of

hydrogen-bonding interactions is consistent with low

acidity of the hydrides in 1-THF ⁺. The coordinated THF

is labile, and a single, exchange-averaged set of THF

1

resonances is observed by H NMR in CD₂Cl₂solution

at 20 °C for the total of 1.9 equiv of THF (the lattice

THF was partly lost under vacuum), while the H_M

hydride signal is sharp (H_Ais obscured by (PC)Me

resonances). However, THF coordination is preferred to

+

that of CD₂Cl₂, as hydride signals of 1-CD₂Cl₂

-

- 74

(BAr′₄

)

are strongly exchange-broadened at 20 °C.

The weak coordination of THF allows for a significant

population of the five-coordinate form in a 50:50 solvent

mixture of d₈-PhMe and PhF (Table 5, entry 12), as

opposed to THF solvent (Table 5, entry 11), which

determines the substantial difference in reactivity

-

toward B(CH₂SiMe₃)₄between the two cases.

The above results, summarized in Scheme 8, show a

clear trend in that X-substitution reactivity of 1-X

follows the leaving group ability of X, with the exception

of F (rationalized below). Therefore, loss of X^-from 1-X

is a crucial step in the mechanisms of both A and B,

consistent with the five-coordinate cation 1⁺exhibiting

the highest reactivity. Additionally, this leaving-group

dependence explains the marked difference in product

yields in the reaction of 1-OTf with Me₃SiCH₂Li in d₈-

THF, which proceeds with over 90% preference via C

at -40 °C, but gives 60% yield of A and B products at

20 °C: the accessibility of 1⁺via dissociation of OTf^-

increases with temperature, making pathways A and

B operative.

(75) Rhine, W. E.; Stucky, G.; Peterson, S. W. J . Am. Chem. Soc.

1975, 97, 6401. Clegg, W.; Lamb, E.; Liddle, S. T.; Snaith, R. J .

Organomet. Chem. 1999, 573, 305.

(74) Yandulov D. V.; Streib, W. E.; Caulton K. G. Inorg. Chim. Acta

1998, 280, 125.

Lithium Bases as Sources of Methyl Anion

Organometallics, Vol. 21, No. 20, 2002 4041

Ta ble 6. Dep en d en ce of Yield s via A vs B on th e

Exten t of Solva tion of Li in th e Rea ction s of 1-OTf

by the presence of (Me₃Si)₂NLi‚OEt₂, the hydride signals

were strongly broadened, concealing J _P-Hand J _H-H

couplings. Similar broadening, concealing all couplings

and consistent with reversible formation of a lithium-

bonded adduct, was always observed in the low-tem-

perature (e-40 °C) reactions of 1-OTf in d₈-PhMe with

(Me₃Si)₂NLi‚OEt₂, NpLi, Me₃SiCH₂Li, and Me₃SiCH₂-

Li (15 equiv) in the presence of TMEDA (20 equiv), but

not with Me₃SiCH₂Li in d₈-THF at -80 °C, consistent

with solvation of Li precluding such interactions in the

latter case. While the fact that the hydride signals of

1-OTf are sufficiently sharp to resolve all couplings in

d₈-THF at -80 °C rules out hindered rotation of the

phosphines⁷⁶as the cause of the line broadening, the

hydride signals of 1-R complexes, comparable in size

to 1-OTf, were not broadened under identical conditions,

thus ruling out typical T₂broadening. Finally, an IR

spectrum of 1-Cl in heptane in the presence of 1 equiv

of (Me₃Si)₂NLi showed a new ν(N-O) band, blue-shifted

by 20 cm^-1, which, fully consistent with lithium bonding

to Cl reducing the back-bonding to NO, rules out an

interaction with NO oxygen; the latter causes a 60 cm^-1

red shift in the case of 1-R complexes,³³in close analogy

to the effect of hydrogen bonding^77,78and to formation

of a number of other Lewis acid adducts.⁵²The ∆δ value

of 1-Cl is increased by 0.79 ppm in C₆D₆in the presence

of 2 equiv of (Me₃Si)₂NLi. No low-frequency ν(N-O)

band was found for 1-Cl in the presence of 1 equiv of

(Me₃Si)₂NLi, which indicates that lithium bonding to

the Cl is stronger than that to the NO oxygen.

(d ) P ossible Mech a n ism s of Su bst it u t ion of X.

The Li‚‚‚X interactions are crucial to the thermodynamic

driving force for â-Me^-transfer from organolithium

reagents. An attempt to deprive the reactants of reason-

able nucleophiles, in the reaction of 1⁺with (Me₃Si)₂NLi

in a 50:50 solvent mixture of d₈-PhMe and PhF at 20

°C, resulted in a slow reaction proceeding on a time scale

of minutes (cf. entries 6 and 10) to yield none of the

products representative of either â-Me^-transfer or

deprotonation of 1⁺. No 1-Me, no 1-P ∼C, no products

of oxidative addition of aromatic solvent C-H bonds,

and no free L were observed. Thus, sequestering Li⁺

with nucleophiles is integral to pathway A being kineti-

cally facile, while entries 9 and 10 of Table 6 show that

solvent THF is as efficient as a coordinating X^-. Overall,

these results implicate the loss of nucleophile-stabilized

Li⁺as an important mechanistic feature of A, obviously

not essential to the reaction of [Bu₄N][B(CH₂SiMe₃)₄]

(entry 12), and reveal that the facility with which â-Me^-

transfer from organolithium reagents to 1-OTf occurs

stems from the combination of good leaving-group

properties of OTf^-and large interaction energy between

the hard Li⁺and OTf^-.

-

w ith Q⁺BR₄a n d Me₃SiCH₂Li^a

yield, %

pathway pathway

entry

Q⁺BR₄^-/RLi

LiBEt₄

solvent

C₆D₆

THF

d₄-MeOH

A

B

1

2

3

4

23

10

11

59

24

46

77

90

89

41

76

54

[Li(THF)₄][B(CH₂SiMe₃)] C₆D₆

5^b,c[Li(THF)₄][B(CH₂SiMe₃)] d₈-THF

6^d15 Me₃SiCH₂Li

d₈-PhMe

(-70 °C)

d₈-PhMe

(-60 °C)

C₆D₆

7^c,e15 Me₃SiCH₂Li +

5

73

54

20 TMEDA

2 Me₃SiCH₂Li +

20 TMEDA

8

46

9

Me₃SiCH₂Li

C₆D₁₂

d₈-THF

Et₂O

45

60

80

55

40

20

10^c,f3 Me₃SiCH₂Li

11^c,fMe₃SiCH₂Li

a

Unless otherwise indicated, all reactions were carried out at

20 °C with

a

slight excess of Q⁺BR₄^-/RLi, and yields were

determined by ¹H or ³¹P NMR integration within 10-15 min with

quantitative conversion of 1-OTf into the products of pathways A

and B. ^b18% conversion after 15 min. ^c1-Me and 1-CH₂SiMe₃are

major products. 100% conversion after 7 h. ^e100% conversion

d

after 4 h. ^fMe₃SiCH₂Li dissolved 30 s prior to the addition of

1-OTf.

quent abstraction of X^-by Li⁺is involved in the rate-

determining step. These observations are consistent

with (a) an associative mechanism, in which a nucleo-

phile (R or â-R′ of Scheme 2) coordinates to Os with

concomitant bending of NO,⁷⁹(b) an interchange mech-

anism, with the seven-coordinate intermediate of (a)

being a transition state, which additionally may or may

not require NO bending, or (c) a dissociative mechanism,

in which X^-first migrates onto Li to generate the five-

coordinate 1⁺and the resulting ion pair collapses via

either A or B and loss of LiX. These mechanisms do not

specify the sequence of R/â-R′ transfer and loss of free

LiX. In the absence of direct observation of either (a) or

(c) intermediates neither mechanism can be ruled out.

However, (c) appears to be most likely, since the

reactions that do not involve abstraction of X^-by Li⁺

(entries 12 and 11), or at least not to the extent possible

in noncoordinating solvents (entries 10 and 9), clearly

proceed via a dissociative mechanism, analogous to (c)

except for the formation of solid LiX, because excess X^-

(OTf^-or THF, respectively) suppresses the rate.

(e) Via bility of th e SET Mech a n ism . An important

mechanistic possibility within either of the above mech-

anisms involves a single-electron transfer to Os, fol-

lowed by R^•/â-R′^•abstraction, as opposed to transfer of

the anionic fragments. Kochi and co-workers have

-

thoroughly examined the reactivity of BR₄with or-

ganic²⁰and organometallic²¹electrophiles, demonstrat-

ing the viability of both SET and nucleophilic mecha-

nisms. The selectivity of R group transfer from mixed

R′BR′′₃^-borates was diagnostic of the specific pathway

followed.²⁰A more substituted R group is transferred

preferentially in a SET mechanism due to both the

weaker B-C bond and the greater stability of the R^•

radical, both of which correlate with the C-H bond

strength in the corresponding RH;^23,80a less substituted,

The results described in parts a-d of section V show

that reactions of 1-X (X ) F, Cl, OTf) with organolithium

reagents in noncoordinating solvents begin with forma-

tion of a lithium bond to the X ligand and the subse-

(76) Notheis, J . U.; Heyn, R. H.; Caulton, K. G Inorg. Chim. Acta

1995, 229, 187 and references therein.

(77) Shubina, E. S.; Belkova, N. V.; Krylov, A. N.; Vorontsov, E. V.;

Epstein, L. M.; Gusev, D. G.; Niedermann, M.; Berke, H. J . Am. Chem.

Soc. 1996, 118, 1105. Belkova, N. V.; Shubina, E. S.; Ionidis, A. V.;

Epstein, L. M.; J acobsen, H.; Messmer, A.; Berke, H. Inorg. Chem.

1997, 36, 1522.

(79) Song, J .; Hall, M. B. J . Am. Chem. Soc. 1993, 115, 327.

(80) March, J . Advanced Organic Chemistry, 3rd ed.; Wiley: New

York, 1985; p 166.

(78) Shubina, E. S.; Belkova, N. V.; Epstein, L. M. J . Organomet.

Chem. 1997, 536-537, 17.

4042 Organometallics, Vol. 21, No. 20, 2002

Yandulov et al.

in the case of MeBR₃^-the methyl group bears a portion

of the delocalized negative overall charge and is there-

fore nucleophilic, in the carbanionic fragments of RLi

and R₂NLi the negative charge is largely localized at

the lithiated center and only a small fraction of it is

delocalized over to the â-substituents, through meso-

meric forms. Nevertheless, this similarity suggests that

analogous S_E2 with Me^-inversion may be the intimate

mechanism of C-Si and C-C bond cleavage in the

reactions of 1-X with RLi and R₂NLi.

Sch em e 9. Rela tion sh ip a m on g R-, R₂N-, a n d

BR₄-

The DFT calculations (B3LYP/LANL2DZ) on the

simplified models [Os(H)₂(NO)(PH₃)₂][Me₃ECH_n] (E )

C, Si, n ) 2; E ) B, n ) 3) show that the Me^-transfer

from free anions via backside electrophilic attack by Os

is facile (Figure 3). Analogous to the structurally

characterized MeBR₃^-adducts, charge-assisted agostic

minima were located for Me₃ECH₂^-(E ) Si, C), bound

primarily through a CH₃hydrogen. The activation

energies (with ZPE) for the â-Me^-transfer with inver-

sion relative to these agostic structures are only 7.7 (C)

and 3.2 (Si) kcal/mol, which correlate with the E-CH₃

bond strengths⁸⁴and approach the inversion barrier of

more accessible fragment is predominantly transferred

via a nucleophilic attack at the substrate.

The reactivity of the mixed [Li(TMEDA)₂][ⁱPr₃BMe]

borate (Table 2) clearly shows that, at least with the

borate reagents, pathway B does not involve single-

electron transfer, since the Me group is transferred

i

exclusively.⁸¹The more substituted Pr group is not

transferred to an observable extent, and 1-H is formed

via direct â-H^-transfer from the (B)ⁱPr group. However,

even if all of the observed 1-H were produced via â-H

elimination from the unobserved product of ⁱPr^-trans-

fer, the yield of the Me^-transfer product would still be

greater, as the former would have been formed with

3-fold statistical preference.

-

free CH₃of 2.1 kcal/mol calculated at the same level.

Neither the agostic adduct nor the transition state for

the Me^-transfer could be located for BMe₄^-; the PES

for the [Os(H)₂(NO)(PH₃)₂][Me₃BCH₃] system is very flat

in the region of B-C and Os-C distances estimated

from the Me₃ECH₂^-results, which indicates that, if the

agostic minimum exists at the current level, the activa-

tion energy relative to it is below 2 kcal/mol. Although

a transition state for â-Me^-transfer with retention of

configuration, known to be relevant for relatively small

electrophiles,⁸⁵has not been located, such a mechanism

is highly unlikely here in view of the limited steric

accessibility of Os in [Os(H)₂(NO)(PⁱPr₃)₂]⁺.

Thermodynamically, for the reaction of [Os(H)₂(NO)-

(PH₃)₂]⁺with Me₃ECH₂^-, pathway B, giving Os(H)₂-

(CH₂EMe₃)(NO)(PH₃)₂, is favored over A, giving Os(H)₂-

(CH₃)(NO)(PH₃)₂and Me₂E ) CH₂, for both E ) Si (by

43.7 kcal/mol) and E ) C (by 7.2 kcal/mol). The large

difference in thermodynamics of A and B between Si

and C stems from the weakness of the SidC double

bond.^84,86The value calculated for Si, although likely

to be reduced in the experimental system due to

subsequent 1,2-addition of Me₃SiCH₂Li, is sufficiently

The trihydride 1-H undergoes instantaneous electron-

transfer reduction by [Li(TMEDA)₂][naphthalene]^•in

C₆D₆, resulting in the formation of free L and develop-

ment of a brown, from pale yellow, color. No other new

species are detected by either ¹H or ³¹P NMR, except

for, perhaps, a very broad background in the (PC)Me

region. Notably, reactions of 1-Cl with several RLi gave

very similar results, in that partial conversion into the

corresponding 1-R products was invariably accompanied

by formation of substantial amounts of free L and

development of a brown color. Therefore, it is possible

that the SET pathway is operative with 1-Cl, because

the substitution of Cl is the slowest; i.e., the reactivity

pathways available to 1-OTf are strongly inhibited, but

they do not necessarily lead to formation of 1-R but,

rather, to decomposition. All other combinations of 1-X

and RLi/R₂NLi did not give any strong evidence for the

involvement of SET, as, for example, 1-H formed in

C₆D₆, C₆D₁₂, or d₈-THF as a result of decomposition

(section I) was typically free of partially deuterated

1

isotopomers, easily distinguishable by H NMR.³⁴

(82) Chen, E. Y.-X.; Marks, T. J . Chem. Rev. 2000, 100, 1391. Deck,

P. A.; Beswick, C. L.; Marks, T. J . J . Am. Chem. Soc. 1998, 120, 1772,

12167. Braga, D.; Grepioni, F.; Tedesco, E.; Biradha, K.; Desiraju, G.

R. Organometallics 1997, 16, 1846. Braga, D.; Grepioni, F.; Tedesco,

E.; Calhorda, M. J . Z. Anorg. Allg. Chem. 2000, 626, 462. Kehr, G.;

Fro¨lich, R.; Wibbelig, B.; Ekert, G. Chem. Eur. J . 2000, 6, 258. Chan,

M. S. W.; Vanka, K.; Pye, C. C.; Ziegler, T. Organometallics 1999, 18,

4624. Lanza, G.; Fragala, I. L.; Marks, T. J . J . Am. Chem. Soc. 1998,

120, 8257. Chen, Y.-X.; Metz, M. V.; Stern, C. L.; Marks, T. J . J . Am.

Chem. Soc. 1998, 120, 6287. Deck, P. A.; Marks, T. J . J . Am. Chem.

Soc. 1995, 117, 6128. Yang, X.; Stern, C. L.; Marks, T. J . J . Am. Chem.

Soc. 1994, 116, 10015. Yang, X.; Stern, C. L.; Marks, T. J . J . Am. Chem.

Soc. 1991, 113, 3623. Bochmann, M.; Lancaster, S. J .; Hursthouse, M.

B.; Malik, K. M. A. Organometallics 1994, 13, 2235.

(f) P r op osed In tim a te Mech a n ism of th e â-Me^-

Tr a n sfer . The carbanionic fragments of RLi and R₂-

NLi containing â-Me^-are closely related to methyltri-

alkylborates in that both are adducts of unsaturated

compounds (olefins or imines and boranes) with Me^-

(Scheme 9). Such borates, most commonly exemplified

by MeB(C₆F₅)₃^-, are well-known to interact with elec-

trophilic transition-metal centers via the bridging meth-

yl group.⁸²The structures of such charge-assisted

adducts may be considered as snapshots along the Me^-

transfer pathway, from B to M, proceeding with inver-

sion of the methyl group in an S_E2 mechanism.⁸³While

(83) Reference 80, Chapter 12. Reference 51, Chapter 8.5. Shinozaki,

H.; Ogawa, H.; Tada, M. Bull. Chem. Soc. J pn. 1976, 49, 775. Fritz,

H. L.; Espenson, J . H.; Williams, D. A.; Molander, G. A. J . Am. Chem.

Soc. 1974, 96, 2378.

(84) Corey, J . C. In ref 58, Chapter 1.

(81) All attempts to prepare MeBⁿBu₃^-or MeBEt₃^-via reactions of

BR₃with MeLi invariably gave considerable amounts of the other four

Me_nBR_4-n^-isomers, in contrast to BⁱPr₃, as determined by observation

of closely situated (B)CH₃resonances in ¹H NMR spectra. Using such

mixtures gave very similar results in that only small amounts of 1-H

and 1-ⁿBu accompanied the formation of 1-Me in the C₆D₆reaction

with 1-OTf.

(85) Fukuto, J . M.; J ensen, F. R. Acc. Chem. Res. 1983, 16, 177.

Dong, D.; Slack, D. A.; Baird, M. C. Inorg. Chem. 1979, 18, 188. Dong,

D.; Baird, M. C. J . Organomet. Chem. 1979, 172, 467. Labinger, J . A.;

Hart, D. W.; Seibert, W. E., III; Schwartz, J . J . Am. Chem. Soc. 1975,

97, 3851.

(86) Gordon, M. S. In Molecular Structure and Energetics; VCH:

Deerfield Beach, FL, 1986; Chapter 4.

Lithium Bases as Sources of Methyl Anion

Organometallics, Vol. 21, No. 20, 2002 4043

F igu r e 3. Selected geometrical parameters (Å, deg) and relative electronic energies (kcal/mol; corrected with ZPE in

-

parentheses) of the structures involved in the intermolecular Me^-transfer from Me₃ECH₂(E ) C, Si) to Os(H)₂(NO)-

(PH₃)₂⁺, as computed at the B3LYP/LANL2DZ level. trans-PH₃groups were omitted for clarity. ∑° denotes the sum of

-

angles around the carbon of the transferred CH₃and around ECH₂

.

large to indicate that â-Me^-transfer from Me₃SiCH₂Li

(g) In flu en ce of Solva tion of Li on th e Distr ibu -

occurs under kinetic control. Conversely, pathway B

appears to be kinetically favored over A for NpLi, as

the yield via B increases from Me₃SiCH₂Li to NpLi

(Table 1) despite decreasing thermodynamic preference

and consistent with the calculated higher activation

energy for the â-Me^-transfer from C than from Si.

Thus, inhibiting pathway B sterically, in the case of

TMPLi, results in the predominant operation of A with

C-C bond cleavage (Table 1).

The model systems in Figure 3 closely represent the

experiment only for BMe₄^-but are oversimplified in the

case of Me₃ECH₂^-, as the loss of solvated Li⁺from C

may occur concomitantly with or following the â-Me^-

transfer and is not modeled here. Therefore, these model

studies are not intended to represent the actual mech-

anism of A with RLi and R₂NLi but, rather, to demon-

strate that a backside electrophilic attack at the â-Me

group by Os is a viable intimate mechanism of Si-Me

and C-Me bond cleavage established experimentally.

The use of such model systems for the pathway B of

BR₄^-reagents is justified by the rate-suppression data,

which demonstrate the involvement of the five-coordi-

nate 1⁺, and the S_E2 mechanism with inversion is

corroborated by calculations. A detailed description of

the sequence of Os-X, Li-XR′′_n(Scheme 2), and (Si,C)-

CH₃bond cleavage events in the reactions of 1-X with

RLi and R₂NLi, while certainly integral to the mecha-

nism, is beyond the scope of this study.

tion of P r od u cts via A vs B. Tight ion pairing of the

borate reagents enhances the â-R′^-transfer relative to

the R^-transfer to 1-OTf in the reactions with LiBEt₄

(Table 6, entries 1-3), and a similar effect is seen for

[Li(THF)₄][B(CH₂SiMe₃)₄] (entries 4 and 5). This is

consistent with the presence of the cation in a tight ion

pair inhibiting the access of the Os electrophile to the

B-C bonds. An analogous effect of aggregation on the

product distribution is observed for Me₃SiCH₂Li, such

that pathway B is accelerated when the nucleophilic

CH₂Li site is more accessible, in lesser-aggregated

structures. Thus, while reaction of 1-OTf with 15 equiv

of Me₃SiCH₂Li in d₈-PhMe at -70 °C gives a 46:54 ratio

of A and B at complete conversion (entry 6), similar to

that obtained with a slight excess of RLi in C₆D₆at 20

°C (Table 1), the analogous low-temperature reaction

in the presence of an additional 20 equiv of TMEDA

(entry 7) proceeds via B in an overwhelming preference

to A. The highly disaggregated form of Me₃SiCH₂Li in

fact reacts exclusively via C at low temperatures in d₈-

THF, and this reactivity preference is reflected in the

-

formation of 22% of 2-CH₂SiMe₃in the low-tempera-

ture reaction in d₈-PhMe in the presence of TMEDA. A

similar effect on the product distribution can be inferred

from the acceleration of pathway B by the formation of

LiOTf (Figure 1) in the -70 °C reaction of base-free Me₃-

SiCH₂Li: the nucleophilic CH₂Li site may be more

sterically accessible in the mixed aggregates⁸⁷of Me₃-

4044 Organometallics, Vol. 21, No. 20, 2002

Yandulov et al.

SiCH₂Li and LiOTf, which would facilitate the access

of the Os electrophile.

less stable toward reductive elimination of RH and is

greatly favored thermodynamically. This kinetic prefer-

ence, however, is insufficient to overcome a higher

barrier for Me-C bond cleavage, as compared to Me-

Si, in NpLi, such that only with additional steric

congestion of the nucleophilic site, in the case of TMPLi,

does the cleavage of Me-C bond become the major

pathway.

Discu ssion

(a ) Rea son s for th e Un u su a l Rea ctivity of 1⁺. The

organolithium reagents studied here have been used for

several decades in a variety of settings, and yet little to

no evidence has been reported for the â-Me^-transfer

reactivity, the exclusive reaction pathway of 1-OTf with

(Me₃Si)₂NLi and (Me₃Si)₂CHLi in nonpolar solvents.

Why?

While the absence of π-donor ligands in 1⁺is essential

to the high electrophilicity of Os, the presence of

hydrides would seem to suggest deprotonation (C) as a

favored reactivity pathway of 1⁺. Cationic metal hy-

drides are known to be hydrogen-bond donors⁷⁸and can

be Brønsted acidic.⁸⁹Although indeed operational, this

reaction mode is quite unfavorable, at least kinetically,

requiring very strong bases. The weak kinetic acidity

of the hydrides exemplifies the strength of Os-H bonds,

a structural feature shown to be central to the stabiliza-

tion of an isoelectronic ruthenium complex.⁸⁸Notably,

substitution of a Cl ligand for a hydride in 1⁺, while

predictably lowering the electrophilicity of Os,⁷⁴also

activates the deprotonation mechanism, and OsHCl(NO)-

L₂⁺(BAr′₄^-) is rapidly deprotonated by NEt₃at 20 °C to

yield the stable d⁸square-planar species OsCl(NO)L₂.⁹⁰

The analogous thermodynamic stability of the deproto-

nation product is lacking in the case of 1⁺, as 2 suffers

from the combination of high unsaturation and the

strong reducing power of Os(0), both increased relative

to OsCl(NO)L₂due to the absence of a π-donor, halide.

Stabilization of 2 via simple adduct formation, possible

in the case of the precursor 1-OTf, is also not very

favorable, as the presumed primary product of depro-

tonation, the adduct of 2 with OTf^-, is not observed, and

even the observed species, 2-CH₂SiMe₃^-, is only stable

at low temperatures. Thus, the absence of π-donors and

the presence of pure σ-donors, hydrides, in 1⁺makes

the complex quite resistant to deprotonation, both

kinetically, due to the strong σ-bonding of the hydrides,

and thermodynamically, as 2 is highly unstable.

The high degree of unsaturation of Os in 1⁺might

have triggered electron transfer to metal, but this

reaction mode was found to be at most a minor pathway.

The +2 oxidation state for Os is not characteristic of

particularly strong oxidative behavior, and while the

most electron-rich Os(II) species studied here, 1-H,

according to the lowest ν(N-O),^34,74does undergo facile

reduction with Li[naphthalene]^•, this pathway must be

slow relative to the two X-substitution pathways and

can only be activated, to a limited extent, by suppressing

the last two.

Finally, the product of Me^-transfer, 1-Me, is suf-

ficiently stable to be isolated and identified. The pro-

pensity of the Me^-transfer product for further reactivity

is increased by the presence of hydride ligands, essential

for high electrophilicity of the metal center, as the cis

arrangement of H and Me in late-transition-metal

complexes is typically unstable,⁹¹although several

examples stable to MeH reductive elimination are

The dihydride Os(H)₂(NO)(PⁱPr₃)₂⁺is a highly unsat-

urated complex. The high electrophilicity of 1⁺results

in reversible OTf^-abstraction from 1-OTf by NaBAr′₄,

unless the product is stabilized by coordination of CD₂-

Cl₂, and requires the use of a better leaving group (1-

BF ₄) or a very potent electrophile (Ph₃C⁺) for generation

in the five-coordinate form in solution and only with

PhF as a polar cosolvent.⁷⁴The absence of π-donor

ligands, the positive charge, and the presence of a very

strong π-acid, linear NO⁺, all contribute to maximize

the Lewis acidity of Os in 1⁺. The presence of the

π-acidic NO compatible with a high degree of unsatura-

tion is a rare combination that may be the key feature

of the electronic structure of 1⁺responsible for the facile

Me^-abstraction. Additionally, the presence of a 5d

metal Os is beneficial to the Me^-transfer via backside

electrophilic attack, as the LUMO of 1⁺, extended trans

to the apical hydride in the square-based pyramid,⁸⁸is

quite diffuse. It is the high electrophilicity of Os in 1⁺

that determines the facility with which the Me^-transfer

occurs, while several other factors serve to inhibit the

other reactivity pathways.

A typical reaction course of an M-X fragment with

an RLi reagent, halide metathesis to yield M-R and

LiX, is kinetically inhibited in the case of 1⁺by the two

bulky PⁱPr₃ligands. A PⁱPr₃adduct of 1⁺, cis,mer-Os-

(H)₂(NO)(PⁱPr₃)₃⁺(BAr′₄^-), appears to test the steric

limits of the vacant coordination site, as the ³¹P signals

are slightly exchange broadened at 20 °C. It is important

to note that the steric discrimination, imparted by the

bulky phosphines to the Os center, is primarily kinetic

in origin in the reactions with organolithium reagents.

In fact, the products of X-metathesis with a bulky Y can

be stabilized via either a reductive elimination of H-Y,

as in the case of Y ) Np,³³or a loss of free L. Therefore,

the role of steric factors in the reactivity of 1⁺is best

described as impeding the access of Os to the congested

nucleophilic sites in the aggregated structures of orga-

nolithium reagents, thus kinetically favoring abstraction

of an “outside” R′^-group. Using increasingly disaggre-

gated forms of RLi indeed accelerates the metathesis

pathway B under certain conditions. Nevertheless, the

sterically driven kinetic preference for A is quite pro-

nounced, as the reaction with Me₃SiCH₂Li with rare

exceptions yields substantial amounts of 1-Me (Table

6), yet the corresponding product of B is only slightly

(87) Novak, D. P.; Brown, T. L. J . Am. Chem. Soc. 1972, 94, 3739.

Henderson, K. W.; Dorigo, A. E.; Liu, Q.-Y.; Williard, P. G.; Schleyer,

P. v. R.; Bernstein, P. R. J . Am. Chem. Soc. 1996, 118, 1339.

Romesberg, F. E.; Collum, D. B. J . Am. Chem. Soc. 1994, 116, 9187.

Romesberg, F. E.; Collum, D. B. J . Am. Chem. Soc. 1994, 116, 9198.

Lewis, H. L.; Brown, T. L. J . Am. Chem. Soc. 1970, 92, 4664.

(88) Heyn, R. H.; MacGregor, S. A.; Nadasdi, T. T.; Ogasawara, M.;

Eisenstein, O.; Caulton, K. G. Inorg. Chim. Acta 1997, 259, 5.

(89) Reference 45, p 91.

(90) The reaction of [OsHCl(NO)(PⁱPr₃)₂][BAr′₄] with 2 equiv of NEt₃

in C₆D₆within seconds gave a dark brown-green solution containing

trans-OsCl(NO)(PⁱPr₃)₂as the only Os product, identified by ¹H and

³¹P NMR: Werner, H.; Flu¨gel, R.; Windmu¨ller, B.; Michenfelder, A.;

Wolf, J . Organometallics 1995, 14, 612.

(91) Reference 51, p 154.

Lithium Bases as Sources of Methyl Anion

Organometallics, Vol. 21, No. 20, 2002 4045

known.^41,92If the primary product of Me^-transfer is not

sufficiently stable to CH₄elimination, the metal-contain-

ing reaction products would be identical with those

resulting from a deprotonation reaction, while the

(diagnostic) organic products can be quite complex. This

rapid subsequent reactivity may therefore mask the

original reaction pathway.

nucleophilic center. This reactivity is also completely

selective for the transfer of â-H^-over that of â-Me^-, as

ⁱPr₂NLi transfers â-H^-exclusively.

Mechanistic studies reveal that substitution reactions

of halide/pseudohalide X at 1-X begin by coordination

of electrophilic Li in RLi/R₂NLi to the X ligand, forming

a “lithium bond”. The electrophilicity of Li is of consid-

erable importance to the substitution rates, which shows

that in the absence of external donor ligands Li acts as

a potent Lewis acid, abstracting X^-to produce the

unsaturated Os species. Coordination of intact R^-to Os

is a pathway unrelated to the transfer of â-R′^-via â-R′

elimination, as the resulting 1-R complexes react over-

whelmingly by reductive elimination of RH. Single-

electron-transfer reduction from a substrate to Os,

(b) Scop e of th e Rea ction . The Me^-transfer

reactivity investigated here requires the reagent to

contain a strongly nucleophilic center, as essentially no

reaction occurred between 1-OTf and SnMe₄in C₆D₆

even on prolonged heating (18 h at 60 °C). The isoelec-

tronic series of the organolithium species that exhibited

Me^-transfer reactivity, Me(ER₂)XR′′_n^-(Scheme 9) with

E ) C, Si and X ) C, N, could not be extended to

alkoxides (X ) O), as the increasing electronegativity

of X strongly diminished the nucleophilicity of the â-Me

groups. Neither of the Me₃EOM′ (E ) C, Si; M′ ) Li, K)

reagents reacted with 1-OTf or 1-THF ⁺via A, instead

giving complex mixtures of products of B and, possibly,

C; lithium reagents were particularly inert, requiring

long reaction times and/or thermal activation. In addi-

tion to the decreased nucleophilicity of the â-Me groups,

the alkoxides lack the steric hindrance of the nucleo-

philic site present in lithium alkyls and alkylamides.

However, the borate B(CH₂SiMe₃)₄^-, in which the Me₃-

SiCH₂group is substantially less nucleophilic than in

Me₃SiCH₂Li, was found to be reactive via A, although

less so than the amide (Me₃Si)₂NLi.

-

followed by abstraction of R′^•, was ruled out for BR₄

reagents as the mechanism of net â-R′^-transfer, on the

basis of the reactivity of mixed borate Pr₃BMe^-. The

i

proposed intimate mechanism of the intermolecular

Me-C/Si bond cleavage is a direct S_E2 substitution at

the methyl carbon with inversion of the Me group,

supported by DFT calculations. For the systems exhibit-

ing competitive transfer of â-R′^-and intact R^-, increas-

ingly disaggregated forms of RLi react preferentially by

the latter pathway, which establishes the transfer of

easily accessible, “outside” â-R′^-as being kinetically

driven. However, this dependence is not strongly pro-

nounced, likely due to indiscriminately high electrophi-

licity of Os in Os(H)₂(NO)(PⁱPr₃)₂

.

+

Overall, the â-Me^-transfer reactivity described here

requires primarily a highly electrophilic metal center

with appropriate steric protection, which is resistant to

deprotonation and SET reduction.

While imines formed in the â-R′^-transfer reaction

are lithiated by and compete with the original base for

the Os, unsaturated silicon species appear to react by

1,2-insertion with remaining RLi/R₂NLi, generating

species that are also active â-Me^-donors. In both cases,

a rather complex mixture of final organic products is

obtained. Of several factors identified as responsible for

the unusual reactivity of Os(H)₂(NO)(PⁱPr₃)₂⁺, high

electrophilicity, augmented by the presence of the very

strong π-acid NO, and steric protection of the metal

center are of utmost importance. In view of the common

occurrence of these conditions in transition-metal chem-

istry, it is reasonable to anticipate the â-Me^-transfer

reactivity to be rather widespread.

Con clu sion s

This study demonstrates how an attempted substitu-

tion at a sterically congested M-X with lithium alkyls

and alkylamides devoid of â-hydrogens can follow a

pathway dramatically different from that expected and

result in a net transfer of a â-Me^-group to M. The

electrophile studied here, Os(H)₂(NO)(PⁱPr₃)₂⁺, is highly

potent and abstracts â-Me^-not only from the very bulky

(Me₃Si)₂NLi, (Me₃Si)₂CHLi, and TMPLi but also from

the relatively unhindered Me₃SiCH₂Li, which transfers

Exp er im en ta l Section

-

the intact Me₃SiCH₂to Os at a rate competitive with

the â-Me^-transfer. The latter example reveals a

considerable kinetic preference for the abstraction of the

â-R′^-group, also evident from widespread abstraction

of â-hydrides from conventional unhindered nucleo-

philes, such as EtLi and ⁿBuLi. Out of an extensive

series of RLi and R₂NLi examined, only NpLi is com-

pletely resistant to the â-R′^-transfer; in comparison to

Me₃SiCH₂Li, this surprisingly different behavior is

suggested by DFT calculations to be due to kinetic

control. While this unusual reactivity of Os(H)₂(NO)-

Gen er a l Con sid er a tion s. All manipulations were carried

out using standard Schlenk, high-vacuum, and glovebox

techniques under argon, with flame- or oven-dried glassware.

Bulk solvents were purified by appropriate methods, distilled,

and stored under Ar in gastight solvent bulbs with Teflon

closures. Deuterated solvents were dried and deoxygenated

accordingly, vacuum-transferred, degassed, and stored in

Teflon-stoppered bulbs in an argon-filled glovebox. ¹H, ¹⁹F, ³¹P,

¹¹B, and ²⁹Si NMR spectra were recorded on a Varian Gemini

2000 (¹H, 300 MHz; ³¹P, 122 MHz; ¹⁹F, 282 MHz) or a Varian

Inova 400 (¹H, 400 MHz; ¹⁹F, 376 MHz; ³¹P, 162 MHz; ¹¹B,

128 MHz; ²⁹Si, 80 MHz) spectrometer and referenced to the

residual protio solvent peaks (¹H), internal Me₄Si standard

(²⁹Si), or external 85% H₃PO₄(³¹P), neat BF₃‚OEt₂(¹¹B), and

neat CF₃COOH (¹⁹F, -78.5 ppm relative to CFCl₃). Chemical

shifts are reported in ppm relative to tetramethylsilane (¹H,

²⁹Si), 85% H₃PO₄(³¹P), BF₃‚OEt₂(¹¹B), and CFCl₃(¹⁹F). NMR

probe temperatures were calibrated with a methanol standard.

During variable-temperature (VT) measurements, samples

were allowed at least 10 min to equilibrate at each tempera-

ture, which was maintained to (0.5 °C. Infrared spectra were

(PⁱPr₃)₂includes the abstraction of the â-Me^-from a

+

weakly nucleophilic R in B(CH₂SiMe₃)₄^-, neutral sub-

strates (SnMe₄) or even alkali-metal alkoxides (Me₃-

EOM′, E ) Si, C) were found to be inert to the transfer

of Me^-, evidently due to insufficient nucleophilicity of

the Me groups and/or lack of steric protection at the

(92) Norton, J . R. Acc. Chem. Res. 1979, 12, 139. Thorn, D. J . Am.

Chem. Soc. 1980, 102, 7109. Burk, M. J .; McGrath, M. P.; Wheeler,

R.; Crabtree, R. H. J . Am. Chem. Soc. 1988, 110, 5034.

4046 Organometallics, Vol. 21, No. 20, 2002

Yandulov et al.

recorded on a Nicolet 510P FT-IR spectrometer in a KBr

solution cell.

in a low-temperature bath, and the tube, cooled to a temper-

ature of 10-20 °C lower than the probe temperature, was

promptly transferred in the precooled NMR probe.

IR Mea su r em en ts. In an argon-filled glovebox an ap-

propriate amount of heptane, measured with a microliter

syringe, was added to the mixture of solid reagents in an oven-

dried vial equipped with a stirbar. IR spectra were recorded

on homogeneous solutions within 5 min.

[Li(TMEDA)₂][P h ₃BMe]. A solution of BPh₃(500 mg, 2.07

mmol) in 10 mL of Et₂O was treated with 0.9 equiv of MeLi

(Et₂O solution, titrated immediately prior to use¹⁰⁶). After the

mixture was stirred for 3 min, TMEDA (600 mg, 5.16 mmol)

was added, leading to immediate precipitation of a white solid,

which was filtered off, washed with 3 × 10 mL of Et₂O, and

dried in vacuo. Yield: 712 mg (1.43 mmol, 77%). ¹H NMR (d₅-

Mater ials. KOSiMe₃, LiOSiMe₃, KO^tBu, LiO^tBu, [Bu₄N][Ph₃-

SiF₂], Et₄NF‚nH₂O, BEt₃(1.0 M, hexanes), ^tBuLi (1.5 M,

n

pentane), BuLi (2.0 M, pentane) (Aldrich), and MeLi (1.6 M,

i

Et₂O) (Acros) were used as received. Pr₂NLi (Aldrich)⁹³and

BPh₃(Strem) were sublimed. 1-Cl, 1-OTf, 1-H,³⁴[Bu₄N][Ph₃-

BMe],^40,94(Me₃Si)₂NLi‚OEt₂,⁹⁵(Me₃Si)₂CHLi,⁹⁶NpLi⁹⁷(Me₃-

SiCH₂Li analogously), TMPLi,⁹⁸EtLi,⁹³LiBMe₄,⁹⁹LiBEt₄,¹⁰⁰

and [Ph₃C][B(C₆H₃-3,5-(CF₃)₂)₄]¹⁰¹were prepared according to

the published procedures or slight modifications thereof. 1,2-

Dehydro-2,6,6-trimethylpiperidine¹⁰²was isolated and used as

a mixture with 25% of 2,2,6,6-tetramethylpiperidine. The

boranes BⁱPr₃and B(CH₂SiMe₃)₃

were prepared by the

103

method of Brown.¹⁰⁴(Me₃Si)₂NLi was prepared from (Me₃-

Si)₂NH (Aldrich) and ⁿBuLi in pentane. [Li(THF)₄][B(CH₂-

SiMe₃)₄]¹⁰⁵was prepared from Me₃SiCH₂Li and B(CH₂SiMe₃)₃

in THF at -78 °C and isolated by layering the THF solution

with pentane at -20 °C; the Bu₄N⁺salt was prepared by

adding a solution of the Li⁺salt in a minimal amount of THF

to a water solution of 1.5 equiv of Bu₄NBr.⁹⁴All organolithium

and organoborate reagents were purified by crystallization

and/or sublimation; with the exception of TMPLi, which

invariably contained ca. 2% each of LiCl and LiOH, all were

free of lithium halides/hydroxide that instantly react with

1-OTf by halide metathesis. Preparation and characterization

of 1-P h , 1-Np , 1-P ∼C, and 2-L will be reported separately.³³

NMR Tu be Rea ction s. Unless otherwise specified in the

text, in a typical experiment 10 mg of solid 1-X and 1.2-1.5

equiv of an appropriate organolithium/organoborate reagent

were mixed in a flame-dried 5 mm NMR tube fitted with a

Teflon stopcock (Young tube), followed by the solvent (0.5-

0.6 mL), or a solution of RLi was added by syringe to a solution

of 1-X. The contents of the tube were vigorously shaken, the

precipitate was centrifuged to the top, and the ¹H NMR

spectrum was recorded within 10-15 min of the start of the

reaction. In the reaction carried out under a gas atmosphere,

C₆D₆was vacuum-transferred onto the solids, the tube was

pressurized while frozen, at ca. -20 °C, and the contents were

vigorously shaken while the solvent was thawing.

Low -Tem p er a t u r e NMR Tu b e R ea ct ion s. Solid re-

agents, ground into fine powders, were placed in either a 5

mm NMR tube annealed to or a Young tube connected through

a glass fitting to a short manifold. The flame-dried manifold

was capped with a Teflon valve and led directly to the outlet

of a 5 mL flask, capped with a Teflon valve, from which about

0.6 mL of an appropriate solvent was vacuum-transferred onto

the solids, while maintaining the NMR tube maximally

submerged in a -78 °C bath. The NMR tube was either sealed

off under vacuum, while keeping the contents frozen at -198

°C, or refilled with Ar (Young tube). The mixture was

homogenized at an appropriate temperature by applying an

NMR tube mixer to the top of the tube, maximally submerged

pyridine, 20 °C): δ 8.11 (br. s, 6H, B(o-C₆H₅)), 7.33 (t, J _H-H

7.4 Hz, 6H, B(m-C₆H₅)), 7.12 (t, J _H-H) 7.2 Hz, 3H, B(p-C₆H₅)),

2.36 (s, 8H, TMEDA), 2.15 (s, 24H, TMEDA), 1.21 (q, J _B-H

3.7 Hz, 3H, BCH₃). Anal. Found (calcd) for C₃₁H₅₀LiBN₄: C,

75.54 (74.99); H, 9.83 (10.15); N, 10.89 (11.28).

)

[Li(TMEDA)₂][ⁱP r ₃BMe]. A procedure analogous to that

used for the Ph₃BMe^-salt, using BⁱPr₃(150 mg, 1.07 mmol)

in 10 mL of Et₂O with 0.9 equiv of MeLi (Et₂O solution, titrated

immediately prior to use) and TMEDA (311 mg, 2.68 mmol)

gave a white solid, which was filtered off, washed with 3 × 5

mL of Et₂O, and dried in vacuo. Yield: 340 mg (0.86 mmol,

89%). ¹H NMR (d₄-MeOH, 20 °C): δ 2.46 (s, 8H, TMEDA), 2.25

(s, 24H, TMEDA), 0.69 (dq, J _H-H) 7.4 Hz, J _B-H) 2.5 Hz,

18H, B(CH(CH₃)₂)₃), 0.37 (m, 3H, B(CH(CH₃)₂)₃), -0.92 (q, J _B-H

) 3.6 Hz, 3H, BCH₃) (hydrolysis becomes apparent after

several hours at room temperature). Anal. Found (calcd) for

C

₂₂H₅₆LiBN₄: C, 66.64 (66.99); H, 13.75 (14.31); N, 14.08

(14.20).

cis,tr a n s-Os(H)₂Me(NO)(P ⁱP r ₃)₂(1-Me). A MeOH solution

(5 mL) of 1-OTf (300 mg, 434 µmol) was treated with a MeOH

solution (3 mL) of [Li(TMEDA)₂][Ph₃BMe] (237 mg, 477 µmol)

at room temperature, leading to precipitation of a yellow solid.

The resulting mixture was stirred for 3 min and cooled to -40

°C to yield yellow crystalline material after several days, which

was washed with 3 × 5 mL of MeOH at -40 °C, dried under

full vacuum at room temperature for 30 min and stored under

Ar at -20 °C. Yield: 206 mg (369 µmol, 85%) ¹H NMR (C₆D₁₂

,

20 °C): δ 2.36 (m, 6H, P(CH(CH₃)₂)₃), 1.22 (dvt, J _H-H) 6.9

Hz, N ) 13.8 Hz, 18H, P(CH(CH₃)₂)₃), 1.21 (dvt, J _H-H) 6.9

Hz, N ) 13.8 Hz, 18H, P(CH(CH₃)₂)₃), 0.67 (td, J _P-H) 4.4 Hz,

J _(C)H-(Os)H) 2.1 Hz, 3H, (OsC)H₃), -7.99 (tdq, J _P-H) 25.4 Hz,

J _H-H) 8.3 Hz, J _(C)H-(Os)H) 2.1 Hz, 1H, ON-OsH¹⁰⁷), -8.14

(td, J _P-H) 18.3 Hz, J _H-H) 8.3 Hz, 1H, Me-OsH). ³¹P{¹H}

NMR (C₆D₁₂, 20 °C): δ 30.7 (s). IR (C₆H₆): 1676 cm^-1(v_NO).

Anal. Found (calcd) for C₁₉H₄₇NOOsP₂: C, 40.54 (40.92); H,

8.11 (8.49); N, 2.53 (2.51).

cis,tr a n s-Os(H)₂Et(NO)(P ⁱP r ₃)₂(1-Et). A MeOH solution

(1.5 mL) of 1-OTf (20 mg, 28.9 µmol) was treated with a MeOH

solution (0.3 mL) of LiBEt₄(6.2 mg, 46.3 µmol) at -78 °C with

stirring and placed in a -40 °C freezer overnight. The yellow

needles that formed were washed with 3 × 5 mL of MeOH at

-40 °C and dried for 5 min under full vacuum at room

temperature (crystals notably darkened). Yield: 8 mg (14.0

µmol, 48%). ¹H NMR (C₆D₆, 20 °C): δ 2.31 (m, 8H, P(CH(CH₃)₂)₃

and OsCH₂CH₃), 2.11 (approximately t, J _H-H) 7.4 Hz, 3H,

OsCH₂CH₃), 1.18 (dvt, J _H-H) 7.0, N ) 14.0 Hz, 18H, P(CH-

(CH₃)₂)₃), 1.13 (dvt, J _H-H) 7.0 Hz, N ) 14.0 Hz, 18H, P(CH-

(93) Kim, Y.-J .; Bernstein, M. P.; Galiano Roth, A. S.; Romesberg,

F. E.; Williard, P. G.; Fuller, D. J .; Harrison, A. T.; Collum, D. B. J .

Org. Chem. 1991, 56, 4435.

(94) Damico, R. J . Org. Chem. 1964, 29, 1971.

(95) Lappert, M. F.; Slade, M. J .; Singh, A.; Atwood, J . L.; Rogers,

R. D.; Shakir, R. J . Am. Chem. Soc. 1983, 105, 1302.

(96) Beswick, C. L.; Marks, T. J . Organometallics 1999, 18, 2410.

(97) Schrock, R. R.; Fellmann, J . D. J . Am. Chem. Soc. 1978, 100,

3359.

(98) Hall, P. L.; Gilchrist, J . H.; Harrison, A. T.; Fuller, D. J .; Collum,

D. B. J . Am. Chem. Soc. 1991, 113, 9575.

(99) Hurd, T. J . Org. Chem. 1948, 13, 711.

(100) Fryzuk, M. D.; Lloyd, B. R. Clentsmith, G. K. B.; Rettig, S. J .

J . Am. Chem. Soc. 1994, 116, 3804.

(CH₃)₂)₃), -7.80 (approximately td, J _P-H) 26.6 Hz, J _H-H

)

8.0 Hz, 1H, ON-OsH), -7.94 (approximately td, J _P-H) 19.9

Hz, J _H-H) 8.0 Hz, 1H, Et-OsH). ³¹P{¹H} NMR (C₆D₆, 20 °C):

δ 29.1 (s).

(101) Bahr, S. R.; Boudjouk, P. J . Org. Chem. 1992, 57, 5545.

(102) Dneprovski, A. S.; Mil’tsov, S. A. J . Org. Chem. USSR (Engl.

Transl.) 1988, 24, 1836.

cis,tr a n s-Os(H)₂(CH₂SiMe₃)(NO)(P ⁱP r ₃)₂(1-CH₂SiMe₃).

A pentane (3 mL) suspension of finely ground 1-OTf (100 mg,

(103) Seyferth, D. J . Am. Chem. Soc. 1959, 81, 1844.

(104) Brown, H. C.; Racherla, U. S. J . Org. Chem. 1986, 51, 427.

(105) Artamkina, G. A.; Tuzikov, A. B.; Beletskaya, I. P.; Reutov,

O. A. Izv. Akad. Nauk SSSR, Ser. Khim. 1980, 10, 2429.

(106) Remenar, J . F.; Collum, D. B. J . Am. Chem. Soc. 1998, 120,

4081.

(107) Signifies the hydride site trans to the specific ligand.

Lithium Bases as Sources of Methyl Anion

Organometallics, Vol. 21, No. 20, 2002 4047

145 µmol) was treated with a pentane (1 mL) solution of Me₃-

SiCH₂Li (22 mg, 234 µmol), vigorously stirred for 5 min at room

temperature, and concentrated to ca. 2 mL at -13 °C.

Precooled methanol (10 mL) was added at -13 °C, and the

mixture was filtered at -13 °C through a small frit attached

to the end of a Teflon cannula and placed in a -40 °C freezer

for several days to yield yellow crystalline material. The solid

was washed with 3 × 5 mL of MeOH at -40 °C and dried for

20 min under full vacuum at room temperature. Yield: 18 mg

of a 85:15 mixture of 1-Me and 1-CH₂SiMe₃(31.7 µmol, 22%).

Selected ¹H NMR (C₆D₆, 20 °C): 0.94 (td, J _P-H) 5.8 Hz,

J _(C)H-(Os)H) 1.3 Hz, 2H, OsCH₂SiMe₃), 0.50 (s, 9H, OsCH₂Si-

(CH₃)₃), -7.39 (td, J _P-H) 27.0 Hz, J _H-H) 8.2 Hz, J _(C)H-(Os)H

not resolved, 1H, ON-OsH), -9.27 (td, J _P-H) 20.0 Hz, J _H-H

) 8.1 Hz, 1H, H₂C-OsH). ³¹P{¹H} NMR (C₆D₆, 20 °C): δ 28.0

(s).

Si)₂NLi‚OEt₂in C₆D₆under a C₂H₄atmosphere for several

days at 20 °C. 2-C₂H₄: ¹H NMR (CD₂Cl₂, 20 °C) δ 2.14 (m,

6H, P(CH(CH₃)₂)₃), 1.56, (br t, J _P-H) 5.3 Hz, 4H, (Os)C₂H₄),

1.25 (dvt, J _H-H) 6.5 Hz, N ) 13.0 Hz, 18H, P(CH(CH₃)₂)₃),

1.21 (dvt, J _H-H) 6.5 Hz, N ) 13.0 Hz, 18H, P(CH(CH₃)₂)₃),

-8.40 (t, J _P-H) 23.0 Hz, OsH); ³¹P{¹H} NMR (CD₂Cl₂, 20

°C): δ 17.0 (s).

NMR Tu be Rea ction of 1-OTf w ith ⁿBu Li. Following the

general procedure with C₆D₆solvent gave a mixture of 1-H

(53%), 1-ⁿBu (41%), and OsH(η²-C₄H₈)(NO)(PⁱPr₃)₂(2-C₄H₈;

6%) (Table 1, part a of section II). 1-ⁿBu : selected ¹H NMR

(C₆D₆, 20 °C) δ -7.82 (approximately td, J _P-H) 26.6 Hz, J _H-H

) 8.2 Hz, 1H, ON-OsH), -7.94 (approximately td, J _P-H) 19.8

Hz, J _H-H) 8.2 Hz, 1H, ⁿBu-OsH); ³¹P{¹H} NMR (C₆D₆, 20 °C)

δ 28.9 (s). 2-C₄H₈: selected ¹H NMR (C₆D₆, 20 °C) δ -10.06

(approximately t, J _P-H) 34 Hz, OsH); ³¹P{¹H} NMR (C₆D₆,

20 °C) AB, δ(A) 28.6, δ(B) 25.9, J _A-B) 62 Hz. After 23 h at

room temperature, both 1-ⁿBu and 2-C₄H₈have fully con-

verted into a mixture of 1-P h -d ,d ₅isomers.

NMR Tu be Rea ction of 1-OTf w ith LDA. Following the

general procedure with C₆D₆solvent gave a mixture of 1-H

and two new products, assigned as cis and trans isomers of

cis,trans-Os(H)₂(CH₂C(Me)dNⁱPr)(NO)(PⁱPr₃)₂with respect to

the CdN bond (Table 1, part b of section IV). Major product:

selected ¹H NMR (C₆D₆, 20 °C) δ -6.86 (br td, J _P-H) 26 Hz,

1H, ON-OsH), -8.24 (td, J _P-H) 21.0 Hz, J _H-H) 7.8 Hz, 1H,

H₂C-OsH); ³¹P{¹H} NMR (C₆D₆, 20 °C) δ 28.2 (s). Minor

product: selected ¹H NMR (C₆D₆, 20 °C) δ -7.21 (obscured),

-8.35 (td, J _P-H) 20.1 Hz, J _H-H) 8.1 Hz, 1H, H₂C-OsH);

³¹P{¹H} NMR (C₆D₆, 20 °C) δ 29.9 (s).

cis,tr a n s-Os(H)₂(C₈H₁₄N)(NO)(P ⁱP r ₃)₂(1-C₈H₁₄N). To a

solid mixture of 1-OTf (150 mg, 217 µmol) and TMPLi (63.8

mg, 434 µmol) was added a benzene solution (6 mL) of 1,2-

dehydro-2,6,6-trimethylpiperidine (477 µmol) with stirring.

The resulting dark brown homogeneous solution was stirred

for 30 min at room temperature and brought to oil in vacuo.

The residue was extracted with 4 × 5 mL of Me₄Si, and the

combined extracts were filtered through Celite and concen-

trated to 5 mL to yield a brown crystalline material after

standing at -40 °C for several days, which was washed with

3 × 5 mL of pentane at -40 °C, dried in vacuo, and stored

1

under Ar at -20 °C. Yield: 100 mg (150 µmol, 69%). H NMR

(C₆D₆, 20 °C): δ 3.17 (td, J _P-H) 4.5 Hz, J _(C)H-(Os)H) 1.3 Hz,

2H, OsCH₂), 2.49 (m, 6H, P(CH(CH₃)₂)₃), 2.37 (t, J _H-H) 6.7

Hz, 2H, (C³)H₂), 1.80 (m, 2H, (C⁴)H₂), 1.52 (m, 2H, (C⁵)H₂),

1.38 (s, 6H, (C⁶)(CH₃)₂), 1.15 (dvt, J _H-H) 7.0 Hz, N ) 14.0

Hz, 18H, P(CH(CH₃)₂)₃), 1.14 (dvt, J _H-H) 7.0 Hz, N ) 14.0

Low -Tem p er a tu r e NMR Tu be Rea ction of 1-OTf w ith

Me₃SiCH₂Li in d ₈-THF . Following the general procedure with

d₈-THF solvent and 5 equiv of Me₃SiCH₂Li, starting at -80

°C, gave a new major product after minutes at -40 °C,

Hz, 18H, P(CH(CH₃)₂)₃), -7.23 (td, J _P-H) 28.1 Hz, J _H-H

)

1

assigned as OsH(CH₂SiMe₃)(NO)(PⁱPr₃)₂^-(2-CH₂SiMe₃^-). H

7.7 Hz, J _(C)H-(Os)Hnot resolved, 1H, ON-OsH), -8.37 (td, J _P-H

) 20.0 Hz, J _H-H) 7.7 Hz, 1H, H₂C-OsH). ³¹P{¹H} NMR (C₆D₆,

20 °C): δ 30.1 (s). IR (C₆D₆): 1681 cm^-1(v_NO).

NMR (d₈-THF, -40 °C): δ 2.21 (m, 6H, P(CH(CH₃)₂)₃), 1.16

(dvt, J _H-H) 6.4 Hz, N ) 12.8 Hz, 18H, P(CH(CH₃)₂)₃), 1.11

(dvt, J _H-H) 6.4 Hz, N ) 12.8 Hz, 18H, P(CH(CH₃)₂)₃), 0.60 (t,

J _P-H) 5.4 Hz, 2H, OsCH₂SiMe₃), -0.10 (s, 9H, OsCH₂Si-

(CH₃)₃), -10.38 (t, J _P-H) 32.8 Hz, OsH). ³¹P{¹H} NMR (d₈-

THF, -40 °C): δ 37.3 (s), selectively decoupled doublet. Using

1-THF ⁺with 4 equiv of Me₃SiCH₂Li in an analogous reaction

gave the identical product within minutes at -80 °C.

[cis,tr a n s-Os(H)₂(THF)(NO)(P ⁱP r ₃)₂][B(C₆H₃-3,5-(CF₃)₂)₄]‚

THF (1-THF ⁺). THF (10 mL), precooled to -65 °C, was added

via cannula to a mixture of solid 1-H (221 mg, 407 µmol) and

[Ph₃C][B(C₆H₃-3,5-(CF₃)₂)₄] (441 mg, 399 µmol) at -65 °C. With

stirring, the mixture was warmed to -45 °C within 45 min,

as all [Ph₃C] salt had dissolved. The resulting deep red

homogeneous solution was concentrated to 5 mL at room

temperature and layered with pentane (15 mL) to give large

red crystals after standing at -20 °C overnight. The solid was

washed with 4 × 5 mL of a 4:1 mixture of pentane and THF

at -20 °C, briefly dried in vacuo (small crystals notably became

pink, due to the loss of lattice THF), and stored under Ar at

-20 °C. Yield: 508 mg (328 µmol, 82%). ¹H NMR revealed the

presence of only 1.9 equiv of THF, and the complex was

counted as a mono-THF adduct in all reactions, to ensure its

complete conversion. ¹H NMR (CD₂Cl₂, 20 °C): δ 3.76 (m,

THF), 2.42 (m, 6H, P(CH(CH₃)₂)₃), 1.88 (m, THF), 1.29 (dvt,

J _H-H) 7.7 Hz, N ) 15.4 Hz, 18H, P(CH(CH₃)₂)₃), 1.27 (dvt,

J _H-H) 7.7 Hz, N ) 15.4 Hz, 18H, P(CH(CH₃)₂)₃), -13.29 (td,

J _P-H) 16.6 Hz, J _H-H) 6.8 Hz, THF-OsH) (ON-Os-H is

completely obscured by the (PC)Me resonances; irradiation of

that region collapses the visible hydride signal into a triplet).

³¹P{¹H} NMR (CD₂Cl₂, 20 °C): δ 41.5 (s). ¹⁹F NMR (CD₂Cl₂,

20 °C): δ -63.6 (s). IR (CD₂Cl₂): 1761 cm^-1(v_NO). Anal. Found

(calcd) for C₅₈H₇₂BF₂₄NO₃OsP₂: C, 44.53 (44.94); H, 4.51 (4.68);

N, 0.93 (0.90).

Low -Tem p er a tu r e NMR Tu be Rea ction of 1-OTf w ith

Me₃SiCH₂Li in th e P r esen ce of TMEDA. Following the

general procedure with d₈-PhMe solvent, 15 equiv of Me₃-

SiCH₂Li and 20 equiv of TMEDA gave complete conversion of

1-OTf after several hours at -60 °C into 1-Me (5%),

-

1-CH₂SiMe₃(73%), and 2-CH₂SiMe₃(22%). 2-CH₂SiMe₃

:

selected ¹H NMR (d₈-PhMe, -60 °C) δ -10.05 (t, J _P-H) 32

Hz, OsH); ³¹P{¹H} NMR (d₈-PhMe, -60 °C): δ 34.4 (s),

selectively decoupled, doublet.

Gen er ation of [Os(H)₂(NO)(P ⁱP r ₃)₂][B(C₆H₃-3,5-(CF₃)₂)₄].

Following the general procedure with 1-H and 0.8 equiv of

[Ph₃C][B(C₆H₃-3,5-(CF₃)₂)₄] in a 50:50 solvent mixture of d₈-

PhMe and PhF gave a deep red solution of the title species, in

addition to Ph₃CH, with the ¹H and ³¹P NMR signals⁷⁴

considerably broadened due to the degenerate intermolecular

hydride site exchange with the remaining 1-H.

Gen er a tion of cis,tr a n s-Os(H)₂X(NO)(P ⁱP r ₃)₂(1-X). X )

F . Using 1-OTf and 1.2 equiv of either [Bu₄N][Ph₃SiF₂] or Et₄-

NF‚nH₂O provided 1-F quantitatively in C₆D₆after 30 min.

1

Selected H NMR (C₆D₆, 20 °C): δ -0.70 (tt, J _P-H) 22.4 Hz,

J

_F-H(cis) ) J _H-H) 8.9 Hz, 1H, ON-OsH), -11.52 (dtd,

_F-H(trans) ) 69.3 Hz, J _P-H) 12.8 Hz, J _H-H) 8.1 Hz, 1H,

NMR Tu be Rea ction of 1-OTf w ith EtLi. Following the

general procedure with C₆D₆solvent gave a mixture of 1-H

(81%), 1-Et (14%), and OsH(η²-C₂H₄)(NO)(PⁱPr₃)₂(2-C₂H₄; 5%)

F-Os-H). ³¹P{¹H} NMR (C₆D₆, 20 °C): δ 36.4 (d, J _P-F) 20.5

Hz), selectively decoupled, doublet of triplets. ¹⁹F NMR (C₆D₆,

20 °C): δ -345 (dtd, J _F-H(trans) ) 71.0 Hz, J _P-F) 20.6 Hz,

1

(Table 1, part a of section II). 2-C₂H₄: selected H NMR (C₆D₆,

20 °C) δ -8.17 (t, J _P-H) 23.0 Hz, OsH); ³¹P{¹H} NMR (C₆D₆,

20 °C): δ 17.0 (s), selectively decoupled, doublet. 2-C₂H₄is

stable for days at room temperature in C₆D₆and can be

quantitatively generated by the reaction of 1-OTf with (Me₃-

J

_F-H(cis) ) 9.6 Hz).

X ) OH. Using 1-OTf, 2 equiv of solid KOH (85%), and 1.5

equiv of 18-crown-6 in C₆D₆gave 1-OH after 29 h at room

4048 Organometallics, Vol. 21, No. 20, 2002

Yandulov et al.

temperature as the major hydride-containing species. Selected

mTorr to give 150 mg of analytically pure colorless, very

viscous liquid that partially crystallized under Ar at room

temperature after several weeks. At variance with a previous

report, 3-Li is quite stable thermally, showing no decomposi-

tion under 130 °C vacuum distillation. ¹H NMR (C₆D₆, 20 °C):

δ 0.41 (s, 6H, Si(CH₃)₂), 0.35 (s, 18H, N(Si(CH₃)₃)₂), 0.27 (s,

9H, N(Li)(Si(CH₃)₃)). ¹H NMR (d₈-THF, 20 °C): δ 0.21 (s, 18H,

N(Si(CH₃)₃)₂), 0.07 (s, 6H, Si(CH₃)₂), -0.1 (br s, 9H, N(Li)(Si-

(CH₃)₃)). Anal. Found (calcd) for C₁₁H₃₃LiN₂Si₄: C, 42.03

(42.25); H, 10.31 (10.64); N, 8.86 (8.96).

X-r a y Str u ctu r e Deter m in a tion s. Gen er a l Con sid er -

a tion s. The crystal was mounted under an inert atmosphere

on a glass fiber using silicone grease and transferred to the

goniostat, where it was cooled to -160 °C using a gas-flow

cooling system of local design. The Bruker-AXS SMART6000

system was used for data collection. Data were corrected for

Lorentz and polarization effects as well as absorption using

the Bruker SAINT software.

(a ) cis,tr a n s-Os(H)₂(C₈H₁₄N)(NO)(P ⁱP r ₃)₂(1-C₈H₁₄N).

The data were collected using 5 s frames with an ω scan of

0.30°. The structure was readily solved using SHELXTL and

Fourier techniques. All hydrogen atoms were visible in a

difference Fourier phased on the non-hydrogen atoms and were

refined isotropically in the final cycles of refinement. A final

difference Fourier was essentially featureless. There was one

peak of intensity 2.43 e/Å³at the metal site, and all other peaks

were less than 0.5 e/Å³.

(b ) [cis,tr a n s-Os(H )₂(TH F )(NO)(P ⁱP r ₃)₂][B(C₆H ₃-3,5-

(CF ₃)₂)₄]‚THF (1-THF ⁺). The data were collected using 30 s

frames with an ω scan of 0.30°. It was observed that the data

decreased in intensity rapidly with increasing (sin θ)/λ. The

structure was solved with some difficulty using SHELXTL and

Fourier techniques. It was discovered that the cation suffered

from disorder, the most serious being the presence of two

possible positions for the metal atom. There was no evidence

of a space group ambiguity. Many of the atoms in the THF

solvent and the anion have large anisotropic thermal param-

eters. The disorder is such that there are two independent sets

of Os-ligand distances present. The Os(91a)-O(24) distance

is 2.62 Å, compared to 2.24 Å for Os(1)-O(24), while the

remaining Os-ligand distances are similar for the two sites.

A final difference Fourier still had several peaks of intensity

up to 2.2 e/Å³, mostly in the vicinity of the disordered metal

atoms.

¹H NMR (C₆D₆, 20 °C): δ -1.96 (td, J _P-H) 22.2 Hz, J _H-H

8.4 Hz, 1H, ON-OsH), -10.38 (td, J _P-H) 14.0 Hz, J _H-H

)

8.4 Hz, 1H, HO-Os-H). ³¹P{¹H} NMR (C₆D₆, 20 °C): δ 32.8

(s).

X ) Br . Using 1-OTf and 3 equiv of Bu₄NBr in C₆D₆gave

1-Br quantitatively after 28 h at room temperature. Selected

¹H NMR (C₆D₆, 20 °C): δ -2.87 (td, J _P-H) 23.8 Hz, J _H-H

7.6 Hz, 1H, ON-OsH), -10.27 (td, J _P-H) 14.9 Hz, J _H-H

)

7.6 Hz, 1H, Br-Os-H). ³¹P{¹H} NMR (C₆D₆, 20 °C): δ 29.8

(s).

X ) I. Using 1-OTf and 3 equiv of Bu₄NI in C₆D₆gave 1-I

quantitatively after 1 h at room temperature. Selected ¹H NMR

(C₆D₆, 20 °C): δ -4.44 (td, J _P-H) 24.3 Hz, J _H-H) 7.6 Hz,

1H, ON-OsH), -10.07 (td, J _P-H) 15.9 Hz, J _H-H) 7.6 Hz,

1H, I-Os-H). ³¹P{¹H} NMR (C₆D₆, 20 °C): δ 26.7 (s).

Gen er a tion of [cis,m er -Os(H)₂(NO)(P ⁱP r ₃)₃][B(C₆H₃-3,5-

(CF ₃)₂)₄]. Using 1-THF ⁺and 1.5 equiv of PⁱPr₃in a 50:50

solvent mixture of d₈-PhMe and PhF afforded the title complex

quantitatively in seconds. Selected ¹H NMR (50:50 d₈-PhMe:

PhF, 20 °C): δ -7.04 (approximately qd, J _P-H) 24.4 Hz, J _H-H

) 6.8 Hz, 1H, ON-OsH), -7.21 (dtd, J _P-H(trans) ) 61.6 Hz,

J

_P-H(cis) ) 30.0 Hz, J _H-H) 6.8 Hz, 1H, P-Os-H). ³¹P{¹H}

NMR (50:50 d₈-PhMe:PhF, 20 °C): δ 19.6 (br. d, J _P-P) 7.3

Hz, 2P), 10.5 (br t, 1P).¹⁰⁸

Gen er a tion of (Me₃SiCH₂)₂B(CH₂SiMe₂CH₂SiMe₃) (5).

A mixture of 1-OTf (20 mg) and 1.1 equiv of [Li(THF)₄][B(CH₂-

SiMe₃)₄] in 0.5 mL of C₆H₆was stirred for 5 min at room

temperature and distilled onto a cold finger at 80 °C (bath)/

ca. 10 mTorr. The condensed liquid was rinsed down with C₆H₆

and distilled again, to give a 89:11 mixture of 5 and B(CH₂-

SiMe₃)₃, with a trace of 1-H. ¹H NMR (C₆D₆, 20 °C): δ 1.01 (s,

2H, BCH₂SiMe₂), 0.97 (s, 4H, B(CH₂SiMe₃)₂), 0.20 (s, 6H,

BCH₂Si(CH₃)₂), 0.15 (s, 18H, B(CH₂Si(CH₃)₃)₂), 0.12 (s, 9H,

SiMe₂CH₂Si(CH₃)₃), -0.18 (s, 2H, SiMe₂CH₂SiMe₃). ¹¹B NMR

(C₆D₆, 20 °C): δ 78.6 (br s), indistinguishable from the signal

of pure B(CH₂SiMe₃)₃. EI-MS: 344 (M⁺).

(Me₃Si)₂NSiMe₂N(H)SiMe₃(3-H). (Me₃Si)₂NSiMe₂Cl¹⁰⁹was

prepared from (Me₃Si)₂NLi and Me₂SiCl₂in THF for 90 min

at 70 °C,¹¹⁰isolated and converted into (Me₃Si)₂NSiMe₂NH₂,¹⁰⁹

which was also isolated, by passing NH₃through the solution

in pentane at -15 °C for 30 min. A pentane (20 mL) solution

of (Me₃Si)₂NSiMe₂NH₂(2.187 g, 9.32 mmol) and NEt₃(4.72 g,

46.64 mmol) was treated with Me₃SiOTf (2.28 g, 10.26 mmol)

with stirring. The resulting mixture was stirred for 15 min,

and the colorless solution was decanted off the heavy brown

oil and filtered through Celite. Distillation at 133°/5 Torr gave

3-H as a colorless liquid (bp 110 °C/9.5 Torr¹¹⁰). Yield: 2.432

Com p u ta tion a l Deta ils. All calculations were performed

with the Gaussian 98¹¹²suites of programs, using the hybrid

density functional method B3LYP,¹¹³with LANL2DZ,^114-116

a

valence double-ú basis set with relativistic effective core

potentials for Os,¹¹⁵Si,¹¹⁶and P¹¹⁶centers. Addition of polar-

ization functions to H (hydrides and hydrogens of the trans-

ferred Me group, exponent 1.0), C (0.75), N (0.80), O (0.85), Si

(0.284),¹¹⁷and P (0.387)¹¹⁷in the [Os(H)₂(NO)(PH₃)₂][Me₃-

SiCH₂] model system increased the activation energy (∆(E +

1

g (7.93 mmol, 85%). H NMR (C₆D₆, 20 °C): δ 0.31 (s, 6H, Si-

(CH₃)₂), 0.28 (s, 18H, N(Si(CH₃)₃)₂), 0.13 (s, 9H, N(H)(Si-

(CH₃)₃)). ¹H NMR (d₈-THF, 20 °C): δ 0.86 (br. s, 1H, NH), 0.24

(s, 6H, Si(CH₃)₂), 0.22 (s, 18H, N(Si(CH₃)₃)₂), 0.09 (s, 9H, N(H)-

(Si(CH₃)₃)). ²⁹Si (C₆D₆, 20 °C): δ 1.85 (m, 2Si, N(SiMe₃)₂), 1.50

(m, 1Si, N(H)(SiMe₃)), -6.20 (m, 1Si, SiMe₂).¹¹¹EI-MS: m/z

291.155 35 (M - Me)⁺(calcd. 291.156 44).

(Me₃Si)₂NSiMe₂N(Li)SiMe₃(3-Li). 3-H (1.0 g, 3.26 mmol)

in pentane (10 mL) was treated with 1.0 equiv of ⁿBuLi (2.0

M, pentane) dropwise at 20 °C. The resulting mixture was

filtered through Celite and brought to dryness in vacuo to give

a slightly off-white waxy solid after prolonged evacuation.

Yield: 0.76 g (2.43 mmol, 75%). A 200 mg amount of this

material was distilled on a cold finger at 130 °C (bath)/0.2

(112) Frisch, M. J .; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.;

Robb, M. A.; Cheeseman, J . R.; Zakrzewski, V. G.; Montgomery, J . A.,

J r.; Stratmann, R. E.; Burant, J . C.; Dapprich, S.; Millam, J . M.;

Daniels, A. D.; Kudin, K. N.; Strain, M. C.; Farkas, O.; Tomasi, J .;

Barone, V.; Cossi, M.; Cammi, R.; Mennucci, B.; Pomelli, C.; Adamo,

C.; Clifford, S.; Ochterski, J .; Petersson, G. A.; Ayala, P. Y.; Cui, Q.;

Morokuma, K.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.;

Foresman, J . B.; Cioslowski, J .; Ortiz, J . V.; Stefanov, B. B.; Liu, G.;

Liashenko, A.; Piskorz, P.; Komaromi, I.; Gomperts, R.; Martin, R. L.;

Fox, D. J .; Keith, T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara, A.;

Gonzalez, C.; Challacombe, M.; Gill, P. M. W.; J ohnson, B.; Chen, W.;

Wong, M. W.; Andres, J . L.; Gonzalez, C.; Head-Gordon, M.; Replogle,

E. S.; Pople, J . A. Gaussian 98, revision A.6; Gaussian, Inc.: Pitts-

burgh, PA, 1998.

(113) Becke, A. D. J . Chem. Phys. 1993, 98, 5648. Lee, C.; Yang,

W.; Parr, R. G. Phys. Rev. B 1988, 37, 785. Stephens, P. J .; Devlin, F.

J .; Chabalowski, C. F.; Frisch, M. J . J . Phys. Chem. 1994, 98, 11623.

(114) Dunning, T. H., J r.; Hay, P. J . In Modern Theoretical Chem-

istry; Schaefer, H. F., III., Ed.; Plenum: New York, 1976; pp 1-28.

(115) Hay, P. J .; Wadt, W. R. J . Chem. Phys. 1985, 82, 299.

(116) Wadt, W. R.; Hay, P. J . J . Chem. Phys. 1985, 82, 284.

(108) The NMR data are analogous to those of the known PPh₃

analog: Boyar, E. B.; Dobson, A.; Robinson, S. D.; Haymore, B. L.;

Huffman, J . C. J . Chem. Soc., Dalton Trans. 1985, 621.

(109) Wannagat, U.; Niederpru¨m, H. Z. Anorg. Allg. Chem. 1961,

308, 22. Mignani, G.; Seyferth, D. Synth. React. Inorg. Met.-Org. Chem.

1993, 23, 113.

(110) Bush, R. P.; Lloyd, N. C. J . Chem. Soc. A 1969, 257.

(111) Assignments were made by comparison to extensive ²⁹Si NMR

data of various silazanes: Kupcˇe, Eh, Lukevics, E.; Varezhkin, Y. M.;

Mikhailova, A. N.; Sheludyakov, V. D. Organometallics 1988, 7, 1649.

Lithium Bases as Sources of Methyl Anion

Organometallics, Vol. 21, No. 20, 2002 4049

ZPE)) for â-Me^-transfer by only 1.4 kcal/mol. All structures

were fully optimized with standard convergence criteria

without symmetry constraints, and all were confirmed to be

real minima or transition states via frequency analysis, which

was also used to calculate zero-point energies (ZPE) without

scaling. In view of the essentially C_s-symmetric nature of both

transition states in Figure 3, the PES for the [Os(H)₂(NO)-

(PH₃)₂][Me₃BCH₃] model system was searched with C_ssym-

metry. For all transition states, motion corresponding to the

imaginary frequency was visually checked and most structures

were additionally optimized to the minima they connected after

correspondingly perturbing the TS geometry.

Ack n ow led gm en t. This work was supported by the

NSF and a graduate fellowship to D.V.Y. from Indiana

University.

Su p p or tin g In for m a tion Ava ila ble: Full X-ray struc-

tural information on 1-C₈H₁₄N and 1-THF ⁺as an X-ray

crystallographic file in CIF format. This material is available

free of charge via the Internet at http://pubs.acs.org.

(117) Ho¨llwarth, A.; Bo¨hme, M.; Dapprich, S.; Ehlers, A. W.; Gobbi,

A.; J onas, V.; Ko¨hler, K. F.; Stegmann, R.; Veldkamp, A.; Frenking,

G. Chem. Phys. Lett. 1993, 208, 237.

OM0200104

Article Doi

Conventional lithium bases as unconventional sources of methyl anion: Facile Me-Si and Me-C bond cleavage in RLi, R2NLi, and BR4- by an electrophilic osmium dihydride

DOI: 10.1021/om0200104

Source and publish data:

Authors:

Article abstract of DOI:10.1021/om0200104

Full text of DOI:10.1021/om0200104

Products guided by the article

R&D Labs maybe for 474407-15-3

Relevant to this article

Hot Product