Conversion of a zinc disilazide to a zinc hydride mediated by LiCl

Article abstract of DOI:10.1021/ja102323g

An unusual β-elimination reaction involving zinc(II) and LiCl is reported. LiCl and a coordinatively saturated disilazido zinc compound form an adduct that contains activated SiH moieties. In THF/toluene mixtures, this adduct is transformed into a zinc hydride and 0.5 equiv. cyclodisilazane. The Li⁺ and Cl^- ions apparently affect the reaction pathway of the disilazido zinc in a synergistic fashion. Thus, the zinc hydride and cyclodisilazane products of formal β-elimination are not observed upon treatment of the zinc disilazide with Cl^- or Li⁺ separately.

Full text of DOI:10.1021/ja102323g

Published on Web 05/14/2010
Conversion of a Zinc Disilazide to a Zinc Hydride Mediated by LiCl
Debabrata Mukherjee, Arkady Ellern, and Aaron D. Sadow*
Department of Chemistry and U.S. Department of Energy Ames Laboratory, Iowa State UniVersity,
Ames, Iowa 50011
Received March 19, 2010; E-mail: sadow@iastate.edu
Organozinc compounds are valuable in synthetic chemistry as
Me3SiCl) in hexane give cyclodisilazanes, and silaimines are
suggested as intermediates in one of the two proposed mecha-
nisms.12 These literature transformations require nonpolar media,
and THF solvent gives substitution rather than elimination. Our
zinc system contrasts with that of Me3SiCl, with nonpolar solvents
giving substitution and THF favoring elimination. Neither HN(Si-
HMe2)2 and LiCl nor mixtures of 1, HN(SiHMe2)2, and LiCl afford
the cyclodisilazane.
alkyl, aryl, and hydride transfer agents that complement organo-
lithium and organomagnesium reagents.1 Importantly, alkali metal
and alkaline earth metal salt adducts of zinc reagents give selective
group transfer chemistry that is distinct from monometallic main
group reagents, and related adducts facilitate selective arene,
hydrocarbon, and alkylether metalations.2 Zn(II) centers also
mediate physiological processes involving group transfer as in liver
alcohol dehydrogenase (LADH), where hydride transfer from a zinc
alkoxide to NAD+ is proposed.3 A connection between synthetic
and physiological zinc chemistry is provided by molecular coor-
dination complexes such as tris(pyrazolyl)borato zinc alkoxides and
hydroxide that model LADH,4 carbonic anhydrase,5 and phos-
phatase.6 Likewise, hydrolase-like transesterifications are catalyzed
by tris-1,1,1-(oxazolinyl)ethane zinc dicarboxylato and ditriflato
compounds for kinetic resolution of chiral esters.7
The identity of zinc hydride 3 is provided by its independent
preparation in a two-step sequence. Reaction of 1 and KOt-Bu
provides ToMZnOt-Bu (4). As shown in Scheme 1, PhMeSiH2
Scheme 1. LiCl Adduct Formation, ꢀ-Elimination and Independent
Synthesis of ToMZnH
Given the importance of zinc-mediated group transfer chemistry,
it is interesting that ꢀ-hydrogen elimination is not a common
pathway for organozinc compounds. For example, ZnEt2 undergoes
ꢀ-elimination only upon IR laser pyrolysis at 600-650 °C, whereas
thermal treatment results in Zn-C bond homolysis.8 Few solution
phase ꢀ-eliminations have been suggested, most notably in the
thermolysis of [NaZnEt3] under reducing conditions.9 A three-
coordinate diketiminato zinc hydride is prepared from the corre-
sponding zinc chloride and KNHi-PrBH3 via an unobserved amido-
BH3 intermediate.10 ꢀ-Elimination was also proposed as an initiation
step in a zinc-catalyzed ketone hydrosilylation.11 Identification of
conditions that favor or disfavor ꢀ-H elimination in zinc(II)
compounds may have important implications in group transfer
reactions in synthetic and enzymatic chemistry. Here, we report a
coordinatively saturated oxazolinylborato disilazidozinc(II) com-
pound that undergoes a formal ꢀ-H elimination at room temperature
facilitated by LiCl.
and 4 react to give 3. Notably, 2 and PhMeSiH2 do not readily
provide 3, presumably due to the hindered, non-nucleophilic
nature of the zinc disilazide. The IR spectrum of 3 contains a
νZnH (1745 cm-1, KBr), and the ZnH resonance appears at 4.29
ppm in the 1H NMR spectrum (cf. HB(3-tBupz)3ZnH, δZnH 5.36;
νZnH 1770 cm-1).13 A single crystal X-ray diffraction study
reveals that 3 is monomeric and contains a terminal zinc hydride,
of which there are relatively few crystallographically studied
examples including the four-coordinate TptBuZnH and TpPh,MeZnH
(for which the ZnH are not located)13 and a three-coordinate
diketiminate ZnH (Zn-H 1.46(2) Å).10 The four-coordinate ZnH
in 3 (1.52(2) Å) is longer by 0.06 Å.
LiN(SiHMe2)2 and 1 react in THF-d8 to provide possible
intermediates in the apparent ꢀ-elimination process. Two Cs-
symmetric compounds are detected after 10 min, rather than C3V-
symmetric 1, 2, and 3. After 12 h at room temperature, the minor
species is partly converted into 3 and (Me2HSiN-SiMe2)2. Attempts
to isolate these intermediates from toluene/THF solvent mixtures
(crystallization conditions) afford crystals of 3.
Treatment of ToMZnCl (1) (ToM ) tris(4,4-dimethyl-2-oxazoli-
nyl)phenylborate) with LiN(SiHMe2)2 in benzene readily provides
1
ToMZnN(SiHMe2)2 (2), and no other products are detected by H
NMR spectroscopy before or after workup. The spectroscopic
features of the SiH, including its downfield chemical shift (5.26
1
ppm), high JSiH (185 Hz), and νSiH (2110 cm-1), are consistent
with a normal disilazido ligand. An X-ray crystal structure (see
Supporting Information) contains Zn · · · H and Zn · · · Si distances
(2.98 Å and 3.06 Å) that are longer than the sums of van der Waals
radii.
Although 2 is formed quantitatively in benzene, in a benzene
(10 mL) and THF (2 mL) mixture, the compounds 2, 1, and
ToMZnH (3) (identified later) are present in a ratio of 20:1:8 upon
workup after 24 h. Additionally, 1,3-diaza-2,4-disilacyclobutane
(Me2HSiN-SiMe2)2,12 is observed. This cyclodisilazane is the head-
to-tail dimer of silaimine Me2HSiNdSiMe2; its formation and the
presence of zinc hydride 3 suggest a ꢀ-elimination reaction.
Reactions of lithium hydrosilazides and group 14 electrophiles (e.g.,
We suspected that the intermediates formed from 1 and LiN-
(SiHMe2)2 in THF were LiCl adducts. Therefore, LiCl and zinc
disilazide 2 were allowed to interact. A crystallized sample of the
1:1 LiCl/2 adduct (5) has the same 1H NMR spectrum as 1:1
9
7582 J. AM. CHEM. SOC. 2010, 132, 7582–7583
10.1021/ja102323g  2010 American Chemical Society

C O M M U N I C A T I O N S
LiN(SiHMe2)2/1 (major isomer). The νSiH of this material is lower
saturation in both 2 and 5. It is tempting to suggest that Cl-
dissociation from 5 gives a three-coordinate zinc center that
undergoes ꢀ-elimination. However, such a mechanism requires an
unlikely 2× repetition of a Cl- coordination and dissociation
sequence since the final product, ToMZnH, does not form a
detectable adduct with LiCl, and Cl- appears to be necessary to
inhibit oxazoline ring-opening. Cl- also does not appear to bind to
silicon, as H-transfer is not observed in the absence of Li+.
Furthermore, addition of the Lewis acids BPh3 or B(C6F5)3 to 6
does not provide cyclodisilazane, suggesting that Li+ is not acting
as a Lewis acid in 5 to mediate hydride transfer.
Lithium chloride also affects the electronic properties of the
disilazide ligand, as shown by the spectroscopy of the ꢀ-SiH moiety.
This electronic effect may be more significant than a low coordina-
tion number for zinc because the dicoordinate Zn(N(SiHMe2)2)2 is
not reported to undergo ꢀ-elimination.14 Therefore, we favor a
mechanism in which the zinc hydride is formed from the four-
coordinate [(κ2-ToM)ZnClN(SiHMe2)2]-. Given the importance of
Zn-mediated reactions in synthetic, catalytic, and enzymatic
chemistry, we are currently investigating related zinc amido, alkyl,
and alkoxide compounds in ꢀ-H and group transfer reactions.
1
(2061 cm-1) than in the case of 2, and the JSiH (102 Hz) is
significantly lower. Compound 5 is fluxional, as it crystallizes at
-80 °C from THF with a C1-symmetric structure (Figure 1).
Although spectroscopic features suggest [M]-SiH interactions,
there are no close contacts between the SiH moieties and the Zn or
Li centers in 5. Additionally, this interesting structure contains an
unusual O-Li-,N-Zn-coordinated bridging oxazoline group. The
phenyl group on boron and the chloride on zinc are disposed syn,
as are the N-lithiated oxazoline and N(SiHMe2)2 groups. Because
Li+ and Cl- are separate in 5, we investigated these ions
independently to determine their role in the formal ꢀ-elimination.
Acknowledgment. Dr. Bruce Fulton is thanked for valuable
NMR assistance. The U.S. DOE Office of Basic Energy Science
(DE-AC02-07CH11358) and the ACS Green Chemistry Institute-
PRF provided financial support. A.D.S. is an Alfred P. Sloan Fellow.
Supporting Information Available: Experimental procedures and
crystallographic data for all new compounds. This material is available
free of charge via the Internet at http://pubs.acs.org.
Figure 1. ORTEP diagram of 5 drawn at 35% probability.
Treatment of 2 with [n-Bu4N]Cl in a mixture of benzene-d6 and
THF-d8 also gives two Cs-symmetric species. One of the isomers
crystallizes and was structurally characterized as [n-Bu4N][(κ2-ToM)-
ZnClN(SiHMe2)2] (6). The IR spectrum (KBr) of 6 shows a broad,
intense νSiH at 2036 cm-1 which is notably lower energy than in
References
(1) Knochel, P.; Perrone, S.; Grenouillat, N. In Applications I: Main Group
Compounds in Organic Synthesis; Knochel, P., Ed.; Elsevier: Amsterdam,
2007; Vol. 9, pp 81-144.
(2) (a) Kennedy, A. R.; Klett, J.; Mulvey, R. E.; Wright, D. S. Science 2009,
326, 706–708. (b) Wunderlich, S. H.; Knochel, P. Angew. Chem., Int. Ed.
2007, 46, 7685–7688.
(3) Lipscomb, W. N.; Strater, N. Chem. ReV. 1996, 96, 2375–2434.
(4) (a) Parkin, G. Chem. ReV. 2004, 104, 699–767. (b) Bergquist, C.; Storrie,
H.; Koutcher, L.; Bridgewater, B. M.; Friesner, R. A.; Parkin, G. J. Am.
Chem. Soc. 2000, 122, 12651–12658. (c) Bergquist, C.; Parkin, G. Inorg.
Chem. 1999, 38, 422–423.
(5) (a) Alsfasser, R.; Trofimenko, S.; Looney, A.; Parkin, G.; Vahrenkamp,
H. Inorg. Chem. 1991, 30, 4098–100. (b) Alsfasser, R.; Powell, A. K.;
Vahrenkamp, H. Angew. Chem., Int. Ed. Engl. 1990, 29, 898–899. (c) Walz,
R.; Weis, K.; Ruf, M.; Vahrenkamp, H. Chem. Ber./Recl. 1997, 130, 975–
980.
1
the case of 2 (2110 cm-1) and 5 (2061 cm-1). The JSiH values in
6 (178 Hz) are slightly lower than in the case of 2 (185 Hz). After
1 week, neither 3 nor (Me2HSiN-SiMe2)2 is observed, and no change
1
is detected in the H NMR spectrum. Thus, although addition of
Cl- affects the νSiH of the disilazide, it does not promote ꢀ-elimina-
tion from the Zn(II) amide.
Addition of [Li(Et2O)n][B(C6F5)4] to 2 in benzene-d6/THF-d8
mixtures results in oxazoline ring-opening giving O-Si bond
formation and formal transfer of hydrogen from silicon to the
(former) imidine carbon (eq 1).
(6) Hikichi, S.; Tanaka, M.; Moro-oka, Y.; Kitajima, N. J. Chem. Soc., Chem.
Commun. 1992, 814–815.
(7) Dro, C.; Bellemin-Laponnaz, S.; Welter, R.; Gade, L. H. Angew. Chem.,
Int. Ed. 2004, 43, 4479–4482.
(8) (a) Kim, Y. S.; Won, Y. S.; Hagelin-Weaver, H.; Omenetto, N.; Anderson,
T. J. Phys. Chem. A 2008, 112, 4246–4253. (b) Linney, R. E.; Russell,
D. K. J. Mater. Chem. 1993, 3, 587–590.
(9) Lennartson, A.; Håkansson, M.; Jagner, S. Angew. Chem., Int. Ed. 2007,
46, 6678–6680.
(10) Spielmann, J.; Piesik, D.; Wittkamp, B.; Jansen, G.; Harder, S. Chem.
Commun. 2009, 3455–3456.
A bimolecular transformation, in which a Zn-N bond of 5 reacts
with a Si-H bond of a second molecule, might also explain the
ꢀ-H elimination chemistry. However, THF-d8 solutions of 5 and
Et3SiH (as a competitive tertiary SiH group) give ToMZnH and
cyclodisilazane, while the Et3SiH is unreacted. Also, only starting
materials are observed upon treatment of 2 with Et3SiH, ruling out
an intermolecular dehydrocoupling-type mechanism.
(11) Ge´rard, S.; Pressel, Y.; Riant, O. Tetrahedron: Asymmetry 2005, 16, 1889–
1891.
(12) (a) Wiseman, G. H.; Wheeler, D. R.; Seyferth, D. Organometallics 1986,
5, 146–152. (b) Kosse, P.; Popowski, E. Z. Anorg. Allg. Chem. 1993, 619,
613–616.
(13) (a) Han, R.; Gorrell, I. B.; Looney, A. G.; Parkin, G. J. Chem. Soc., Chem.
Commun. 1991, 717–719. (b) Looney, A.; Han, R.; Gorrell, I. B.; Cornebise,
M.; Yoon, K.; Parkin, G.; Rheingold, A. L. Organometallics 1995, 14,
274–288. (c) Rombach, M.; Brombacher, H.; Vahrenkamp, H. Eur. J. Inorg.
Chem. 2002, 153–159.
Clearly, Li+ and Cl- have a synergistic effect in this ꢀ-elimina-
tion reaction through the formation of the adduct 5, and this
requirement is surprising given the coordinative and electronic
(14) Liang, Y.; Anwander, R. Dalton Trans. 2006, 1909–1918.
JA102323G
9
J. AM. CHEM. SOC. VOL. 132, NO. 22, 2010 7583