Probing M-O bond cleavage in silicon and titanium bisenolate radical cations

Article abstract of DOI:10.1021/jo981793z

In this study, several novel sterically congested silicon and titanium bisenolates of exceptional hydrolytic stability have been synthesized and characterized. The structure of the titanium bisenolate T2 could be determined by X-ray structure analysis. Pr

Full text of DOI:10.1021/jo981793z

J . Org. Chem. 1999, 64, 3077-3085
3077
P r obin g M-O Bon d Clea va ge in Silicon a n d Tita n iu m Bisen ola te
Ra d ica l Ca tion s
,
†
†
‡
‡
Michael Schmittel,* Armin Burghart, Helmut Werner, Matthias Laubender, and
†
Rolf S o¨ llner
Institut f u¨ r Organische Chemie and Institut f u¨ r Anorganische Chemie der Universit a¨ t W u¨ rzburg, Am
Hubland, D-97074 W u¨ rzburg, Germany
Received September 3, 1998
In this study, several novel sterically congested silicon and titanium bisenolates of exceptional
hydrolytic stability have been synthesized and characterized. The structure of the titanium
bisenolate T2 could be determined by X-ray structure analysis. Preparative one-electron oxidation
of the sterically shielded metal bisenolates S1-S4 and T1-T2 led to formation of the benzofurans
B1 and B2. As the benzofuran formation is strong evidence for an initial M-O bond cleavage process
at the stage of the initial radical cation we were able to measure the kinetics of this scission process
-
1
using cyclic voltammetry techniques (first-order rate constants range at k ) 0.1-90 s ). These
f
data indicate that both silicon and titanium bisenolates are interesting candidates for oxidative
intramolecular carbon-carbon bond formation processes.
In tr od u ction
strategy based on the intramolecular oxidative coupling
9
of metal (M ) Si) bisenolates (Scheme 1) that not only
Transformations based on single electron transfer
SET) have become a valuable tool in modern synthetic
allows for a clean cross coupling (starting from unsym-
(
10
metrically substituted silyl bisenol ethers) but as an
1
chemistry. In particular, oxidative cyclization reactions
leading to C-C bond formation have received ample
attention.1 A great number of these processes are based
on the enoxy functionality (CdC-O) due to its low
additional benefit provides the products in high dia-
stereomeric excess (de ) 97%).
h,2
Since the initial C-C bond formation in Scheme 1
could equally proceed via other intermediates (e.g., via
an R-carbonyl radical or a bisenol dication), we wanted
to examine such systems in more detail. Obviously, for
the feasibility of the above strategy, M-O bond cleavage
has to be suppressed. Hence, in the following we have
investigated the oxidation chemistry of silyl and titanium
bisenol ethers using sterically bulky enolates with a
special emphasis on the analysis of the M-O bond
cleavage process.
3
oxidation potential and the ease of synthetic accessibil-
ity. In this context, enol esters, alkyl enol ethers,5
4
6
7
stannyl, and silyl enol ethers have been investigated
in a broader way.
A particular challenge in this field is to effect not only
intramolecular but also equally intermolecular C-C bond
formation reactions with an emphasis on the oxidative
cross coupling of two π-nucleophiles. For a reaction to
occur efficiently, several requirements for the substrates
8
(
oxidation potentials, steric blocking) have to be met in
Resu lts
order to have one of them selectively oxidized and the
other operating exclusively as the nucleophile. To cir-
cumvent these problems, we have recently developed a
Syn th esis a n d Str u ctu r e. Synthesis of the silicon bis-
and trisenolates S1-S4 was accomplished along stan-
dard procedures (Scheme 2). The stable enols E1 and E2
were deprotonated with stoichiometric amounts of tri-
ethylamine and subsequently silylated by iodosilane
generated in situ at room temperature. During workup
†
Institut f u¨ r Organische Chemie. New address: OC1-IB8 (Chemie
1
1
&
Biologie), Universit a¨ t Siegeu, Adolf-Reichwein-Strasse, D-57068
Siegeu, Germany.
‡
Institut f u¨ r Anorganische Chemie.
(
1) (a) Chanon, M.; Fox, M.-A. Photoinduced Electron Transfer;
Elsevier Science Publishers B.V.: Amsterdam, 1981; Parts A-D. (b)
Eberson, L. Electron-Transfer Reactions in Organic Chemistry;
Springer: Berlin, 1987. (c) Mattay, J . Synthesis 1989, 233. (d) Sch a¨ fer,
H. J . Organic Electrochemistry, 3rd ed.; Lund, H., Baizer, M. M., Eds.;
Marcel Dekker: New York, 1991. (e) Shono, T. Electroorganic Syn-
thesis; Academic Press: New York, 1991. (f) Volke, J .; Li sˇ ka, F.
Electrochemistry in Organic Synthesis; Springer: Berlin, 1994. (g)
Kyriacou, D. Modern Electroorganic Chemistry; Springer: Berlin, 1994.
(7) (a) Gassman, P. G.; Bottorff, J . J . Org. Chem. 1988, 53, 1097.
(b) Moeller, K. D.; Marzabadi, M. R.; New, D. G.; Chiang, M. Y.; Keith,
S. J . Am. Chem. Soc. 1990, 112, 6123. (c) Snider, B. B.; Kwon, T. J .
Org. Chem. 1990, 55, 4786. (d) Hudson, C. M.; Marzabadi, M. R.;
Moeller, K. D.; New, D. G. J . Am. Chem. Soc. 1991, 113, 7372. (e)
Snider, B. B.; Kwon, T. J . Org. Chem. 1992, 57, 2399. (f) Heidbreder,
A.; Mattay, J . Tetrahedron Lett. 1992, 33, 1973. (g) Fujii, T.; Hirao,
T.; Oshiro, Y. Tetrahedron Lett. 1992, 33, 5823. (h) Moeller, K. D.;
Tinao, L. V. J . Am. Chem. Soc. 1992, 114, 1033. (i) Rathore, R.; Kochi,
J . K. Tetrahedron Lett. 1994, 35, 8577. (j) Kohno, Y.; Narasaka, K.
Bull. Chem. Soc. J pn. 1995, 68, 322. (k) Snider, B. B. Chem. Rev. 1996,
96, 339. (l) Bockman, T. M.; Shukla, D.; Kochi, J . K. J . Chem. Soc.,
Perkin Trans. 2 1996, 1623. (m) Bockman, T. M.; Kochi, J . K. J . Chem.
Soc., Perkin Trans. 2 1996, 1633.
(
3
h) Schmittel, M.; Burghart, A. Angew. Chem., Int. Ed. Engl. 1997,
6, 2550. (i) Linker, T.; Schmittel, M. Radikale und Radikalionen in
der Organischen Synthese; Wiley-VCH: Weinheim, 1998.
2) (a) Snider, B. B. Chem. Rev. 1996, 96, 339. (b) Moeller, K. D.
Top. Curr. Chem. 1997, 185, 49.
(
(3) Schmittel, M. Top. Curr. Chem. 1994, 169, 183.
(
4) Shono, T.; Nishiguchi, I.; Kashimura, S.; Okawa, M. Bull. Chem.
(8) (a) Baciocchi, E.; Casu, A.; Ruzziconi, R. Tetrahedron Lett. 1989,
30, 3707. (b) Paolobelli, A. B.; Latini, D.; Ruzziconi, R. Tetrahedron
Lett. 1993, 34, 721.
Soc. J pn. 1978, 51, 2181.
5) (a) Moeller, K. D.; Marzabadi, M. R.; New, D. G.; Chiang, M. Y.;
(
Keith, S. J . Am. Chem. Soc. 1990, 113, 6123. (b) Moeller, K. D.; Hudson,
C. M.; Tinao-Wooldridge, L. V. J . Org. Chem. 1993, 58, 3478. (c)
Hudson, C. M.; Moeller, K. D. J . Am. Chem. Soc. 1994, 116, 3347. (d)
New, D. G.; Tesfai, Z.; Moeller, K. D. J . Org. Chem. 1996, 61, 1578.
(9) Schmittel, M.; Burghart, A.; Malisch, W.; Reising, J .; S o¨ llner,
R. J . Org. Chem. 1998, 63, 396.
(10) Haeuseler, A. Diplomarbeit, W u¨ rzburg, 1997.
(11) Cazeau, P.; Duboudin, F.; Moulines, F.; Babot, O.; Dunogues,
J . Tetrahedron 1987, 43, 2076.
(6) Kohno, Y.; Narasaka, K. Bull. Chem. Soc. J pn. 1995, 68, 322.
1
0.1021/jo981793z CCC: $18.00 © 1999 American Chemical Society
Published on Web 04/15/1999

3
078 J . Org. Chem., Vol. 64, No. 9, 1999
Schmittel et al.
Sch em e 1
Sch em e 2
ity toward hydrolysis that is quite similar to that of
structurally related titanium monoenolates.12 Both types
of compounds can be stored under air for months without
any significant decomposition taking place.
The structure of T2 exhibits a three-blade propeller
conformation for both enolato groups in which the three
aryl groups are twisted from planarity with the double
bond (Figure 1). All three aryl groups in one enolato
moiety are twisted in the same direction, and both
enolates exhibit the same propeller conformation. The
enolato moieties in T2 show three different dihedral
angles with 43.5(8)° and 44.5(9)° for the two phenyl
groups, 55.4(7)° and 56.7(8)° for the two mesityl rings
cis to phenyl, and 62.4(8)° and 64.3(8)° for the two mesityl
rings trans to phenyl (Table 1). This contrasts the
corresponding enol E213 where the two mesityl rings
possess a very similar dihedral angle of 62.43° and 65.69°,
respectively, whereas for the phenyl group it is much
smaller: 33.34°. The CdC bond compared to E2 [1.339(4)
Å] is elongated whereas the C-O bond is shortened
[
E2: 1.394(4) Å]. The C-C double bond itself is twisted
by about 14°. The Ti-O bond is roughly 0.05 Å longer
than in the structurally similar titanium monoenolate
2 2
Mes CdC(H)OTiCp
Cl,12 an effect that was already
recognized in the case of acetophenone-based titanocene
bisenolates and explained by the decreasing π-donation
of the oxygen to titanium.14 The angle between the two
substituents attached to the titanocene moiety opens up
from 96.5(1)° in the case of the monoenolate Mes
2
CdC(H)-
Cl (T3)1 to 99.5(2)° for T2, a phenomenon that
2
2b
OTiCp
1
4
also could be observed by Floriani and can easily be
explained by the increased steric demand of the attached
substituents.
water was strictly excluded due to the instability of the
model compounds toward hydrolysis. All silicon enolates
were purified by chromatography on silica gel.
The titanium bisenolates T1 and T2 as well were
generated starting from enols E1 and E2, which were
deprotonated quantitatively by sodium hydride in THF
at room temperature. The resultant enolates were treated
with 0.5 equiv of dichlorotitanocene. In case of T1 the
reaction was readily accomplished at room temperature,
while for T2 the reaction mixture had to be refluxed for
P r ep a r a tive Oxid a tion s. In the preparative oxida-
tion of model compounds S1-S4, T1, and T2 with
F eP h en [tris(1,10-phenanthrolino)iron(III)hexafluoro-
phosphate] as a well-defined outer-sphere one-electron
1
5
oxidant (E1/2 ) +0.69 VFc
)
the benzofuran derivatives
B1 and B2 were furnished in mostly good yields even
when the yield was based on the enolato groups available
(Table 2, Scheme 3). Thus, in the bisenolates S1-S3, T1,
6
h. The necessity to increase the reaction temperature
and T2 both enolato moieties and in the trisenolate S4
all three enoxy groups were converted to benzofurans.
Characteristically, if the oxidations were stopped after
can be rationalized by the fact that it is quite difficult to
attach the second bulky enolate ligand in case of T2. The
dark red titanium bisenolate T1 could be purified by
column chromatography on silica gel, whereas T2 was
crystallized from a dichloromethane/acetonitrile mixture
at -40 °C, affording almost black crystals suitable for
X-ray structure analysis. The sterically congested tita-
nium bisenolates T1 and T2 exhibit an enormous stabil-
(12) (a) Schmittel, M.; S o¨ llner, R. Angew. Chem., Int. Ed. Engl. 1996,
35, 2107. (b) Schmittel, M.; Werner, H.; Gevert, O.; S o¨ llner, R. Chem.
Ber./ Recueil 1997, 130, 195.
(
(
13) Nadler, E. B.; Rappoport, Z. J . Am. Chem. Soc. 1987, 109, 2112.
14) Veya, P.; Floriani, C.; Chiesi-Villa, A.; Rizzoli, C. Organome-
tallics 1993, 12, 4892.

Sequential M-O Bond Cleavage
J . Org. Chem., Vol. 64, No. 9, 1999 3079
F igu r e 1. ORTEP view of titanocene bisenolate T2 (50% probability ellipsoids).
Ta ble 1. Selected Bon d Dista n ces a n d An gles for Tita n ocen e Bisen ola te T2
bond distance (Å) atoms angle (deg) atoms
1.905(3) C12-C11-O1 121.0(5) C14-C13-C30-C35
atoms
torsion angles (deg)
Ti-O1
Ti-O2
43.5(8)
55.4(7)
62.4(8)
44.5(9)
56.7(8)
64.3(8)
14.6(9)
13.3(9)
161.4(7)
1.902(5)
1.348(6)
1.346(7)
1.357(7)
1.372(8)
2.110(7)
2.108(7)
C14-C13-O2
Ti-O1-C11
Ti-O2-C13
O1-Ti-O2
121.5(6)
156.5(3)
157.8(3)
99.5(2)
C13-C14-C60-C61
C13-C14-C70-C71
C12-C11-C20-C25
C11-C12-C50-C51
C11-C12-C40-C45
O1-C11-C12-C40
C20-C11-C12-C50
Ti-O1-C11-C12
C11-O1
C13-O2
C11-C12
C13-C14
Cp1-Ti-Cp2
128.4(5)
a
Ti-Cp1
Ti-Cp2
b
a
Cp1: C1-C2-C3-C4-C5. b Cp2: C6-C7-C8-C9-C10.
Ta ble 2. Oxid a tion of Mod el Com p ou n d s S1-S4, T1, a n d
Ta ble 3. Oxid a tion P oten tia ls of th e Silicon Bis- a n d
Tr isen ola tes a n d Tita n iu m Bisen ola tes
a
T2 w ith F eP h en
equiv of
F eP h en
compd
Epa [VFc]
+0.90a
E1/2ox [VFc]
compd
reaction time
yield of benzofuran
b
S1
S2
S3
S4
T1
T2
+0.88
a
S1
S2
S2
S3
S4
S4
T1
T1
T2
T2
4
4
4
4
6
6
2
4
2
4
14 h
0.5 min
14 h
14 h
14 h
0.5 min
1 min
1 min
1 min
1 min
69% of B1
38% of B2
61% of B2
77% of B2
72% of B2
63% of B2
51% of B1
94% of B1
44% of B2
88% of B2
+0.70
n.d.
+0.81a
+0.79a
+0.27a
+0.27a
+0.74
b
b
+0.75
c
+0.27
d
+0.19
a
In acetonitrile at v ) 100 mV‚s-1. b In dichloromethane at v
1000 V‚s
dichloromethane at v ) 200 V‚s
-
1
c
-1 d
)
.
In dichloromethane at v ) 100 mV‚s
.
In
-
1
.
peak potentials of the silicon bis- and trisenolates S1-
a
Yields refer to conversion of both enolato electrophores.
1
5
S4 are in the range of 0.70-0.90 VFc (VFc ) V vs Fc ),
those of the titanium bisenolates T1 and T2 are both Epa
Sch em e 3
1
/2
)
+0.27 VFc. A decreasing ratio Ipa/v with increasing
scan rate is monitored, and Ipc/Ipa approximates 1.0 with
increasing scan rate v. Thus, for each system the diag-
nostic criteria16 point to a reversible ET followed by a
fast chemical reaction. In dichloromethane T1 exhibits
partially reversible waves already at scan rates v g 20
-1
mV‚s . At higher scan rates fully reversible waves could
be monitored as well for T2 and the silicon bis- and
trisenolates S1, S3, and S4 in dichloromethane. This
represents the first direct observation of silicon biseno-
0
.5 min the yields in the case of the silyl bisenol ethers
are diminished compared to entries where the reaction
mixture was allowed to stir overnight (e.g., for S2 and
S4).
(15) All potentials are referenced to their ferrocene/ferrocenium (Fc)
Electr oa n a lytica l Stu d ies. All model compounds
redox couple unless otherwise noted. To obtain values vs SCE, simply
add 0.39 V.
(16) Nicholson, R. S.; Shain, I. Anal. Chem. 1964, 36, 706.
-
1
showed irreversible oxidation waves at v ) 100 mV‚s
in acetonitrile at room temperature (see Table 3). The

3
080 J . Org. Chem., Vol. 64, No. 9, 1999
Schmittel et al.
late, silicon trisenolate, and titanium bisenolate radical
cations in solution and allowed for the determination of
the thermodynamically relevant redox potentials E1/2ox
.
The silicon bisenolates S2 and S3 exhibit higher
potentials in acetonitrile compared to the corresponding
ox
22
silicon monoenolate S5 with E1/2 ) +0.68 VFc
.
As
expected, the potential of S1 exceeds that of the systems
S2-S5, which carry an additional phenyl group. In
contrast, the titanocene bisenolates exhibit lower oxida-
tion potentials than their corresponding chlorotitanocene
ox
monoenolates Mes
2
CdC(H)OTiCp
CdC(Ph)OTiCp
Fc). This can be rationalized by the fact that in the
2
Cl (T3, E1/2 ) +0.50
ox
V
V
Fc) and Mes
2
2
Cl (T4, E1/2 ) +0.31
1
7
case of the silicon bisenolates an electron-donating meth-
yl group is replaced by an enolate and in the case of the
titanocene bisenolates an electron-withdrawing chloride
is substituted by the second enolate ligand.
The diarylethenoxy system S1 displays a second,
anodically shifted oxidation wave with Epa ) +1.06 VFc
in dichloromethane (Figure 2a) and Epa ) +1.03 VFc in
acetonitrile. This wave is irreversible as well and can be
ascribed to the oxidation of 3-mesityl-4,6,7-trimethylben-
ox
18
zo[b]furan (B1), E1/2 ) +1.06 VFc), a product that is
furnished through oxidation of sterically shielded poly-
aryl enols18 or derivatives thereof.
22-26
Analogously, for the silicon enolates S2-S4 a second,
but this time partially reversible oxidation wave (see
ox
Figure 2a,b), is monitored at E1/2 ) +0.85 VFc. This is
again due to the oxidation of the corresponding benzo-
ox
18
[
b]furan B2 (E1/2 ) +0.87 VFc).
In case of the titanocene bisenolate T1 the situation
is more complicated (see Figure 3). In acetonitrile, three
additional irreversible waves could be detected. The first
followup wave appeared at Epa ) +0.37 VFc (wave II) and
was assigned to the oxidation of the R-carbonyl radical
R1, which is formed by the mesolytic19 Ti-O bond
cleavage of a titanocene bisenolate radical cation. At Epa
)
+0.72 VFc (wave IV) another wave appeared that
18,20
represents the oxidation of the corresponding enol E1.
The third follow-up wavesoxidation of the benzofuran
B11 sappears at Epa ) + 1.08 VFc (wave V). This
assignment was supported by cv measurements after
addition of 2 equiv of trifluoromethanesulfonic acid where
the first two waves vanished completely but the last two
waves (oxidation of E1 and B1) increased in height, a
phenomenon that is already known from oxidation stud-
ies on titanocene monoenolates.17
8,20
F igu r e 2. Cyclic voltammetry investigation of (a) S1 in
dichloromethane, (b) S2 in acetonitrile (multisweep cv), and
(
c) S4 in acetonitrile (multisweep cv); all spectra recorded at
-
1
v ) 100 mV‚s
.
In dichloromethane T1 reveals a similar behavior
(Figure 4a,b). The wave of T1 is now followed by four
follow-up waves, the first of which can only be monitored
(
(
17) Schmittel, M.; S o¨ llner, R. Chem. Ber./ Recueil 1997, 130, 771.
18) (a) R o¨ ck, M.; Schmittel, M. J . Prakt. Chem. 1994, 336, 325. (b)
R o¨ ck, M. Ph.D. Thesis, Freiburg, 1994.
19) The expression mesolytic was coined by Maslak to describe bond
(
cleavage of radical ions to yield radical and ionic products: Maslak,
P.; Narvaez, J . N. Angew. Chem., Int. Ed. Engl. 1990, 29, 283.
(
20) Schmittel, M.; Baumann, U. Angew. Chem., Int. Ed. Engl. 1990,
9, 541. (b) Schmittel, M.; R o¨ ck, M. Chem. Ber. 1992, 125, 1611. (c)
Schmittel, M.; R o¨ ck, M. J . Chem. Soc., Chem. Commun. 1993, 1739.
21) Evans, D. H. J Electroanal. Chem. Interfacial. Electrochem.
995, 385, 201.
22) Schmittel, M.; Keller, M.; Burghart, A. J . Chem. Soc., Perkin
Trans. 2 1995, 2327.
23) Schmittel, M.; Steffen, J .-P.; Burghart, A. J . Chem. Soc., Chem.
Commun. 1996, 2349.
2
(
1
(
F igu r e 3. Cyclic voltammogram of T1 in acetonitrile at v )
-1
1
00 mV‚s .
(
(
24) Schmittel, M.; S o¨ llner, R. Chem. Commun. 1998, 565.
in multisweep experiments or at very slow scan rates (see
Figure 4b). This wave emerges at Epa ) +0.38 VFc (wave
II) and is again assigned to the oxidation of the R-car-
(25) Schmittel, M.; Heinze, J .; Trenkle, H. J . Org. Chem. 1995, 60,
2
726.
26) Schmittel, M.; Trenkle, H. Chem. Lett. 1997, 299.
(

Sequential M-O Bond Cleavage
J . Org. Chem., Vol. 64, No. 9, 1999 3081
-
1
F igu r e 4. (a) Cyclic voltammogram of T1 in dichloromethane at v ) 100 mV‚s . (b) Multisweep cyclic voltammogram of T1 in
-
1
-1
dichloromethane at v ) 100 mV‚s . (c) Cyclic voltammogram of T2 in acetonitrile at v ) 50 mV‚s . (d) Multisweep cyclic
-
1
voltammogram of T2 in dichloromethane at v ) 200 mV‚s
.
Sch em e 4
V
Fc) is quite similar to that of T1. Again, the oxidation
of the formed R-carbonyl radical can be observed at Epa
+0.27 VFc. Due to the fact that the oxidation potentials
)
of T2 and the carbonyl radical are very close, the
oxidation wave of T2 can only be observed as a shoulder
in the cathodic part of the wave of the radical (Figure
-
1
4
d). If the scan rate is increased to >20 V‚s only fully
reversible waves of T2 can be monitored and no wave
for the oxidation of the R-carbonyl radical appears. Under
•
+
these conditions T2 is stable on the time scale of the
experiment and cannot decompose to R2.
In an experiment devised to determine the amount of
electrons consumed in the course of substrate oxidation,
the anodic peak current Ipa of silicon monoenolate S5 and
silicon polyenolates S1, S3, and S4 was compared. To
address this issue, the anodic peak current of each
compound was referenced to E2. A comparison of peak
currents is useful only if the diffusion constants of the
compounds to be compared are similar.21 It is known that
the initial oxidation wave for monosilyl enol ether S5
contains two electrons due to (i) substrate oxidation and
bonyl radical R1 (Scheme 4). The second one appears at
E
pa ) +0.70 VFc (wave III) and is tentatively assigned to
the oxidation of a cationic titanocene monoenolate formed
in the same cleavage process as the R-carbonyl radicals.
The third wave shows up at Epa ) +0.85 VFc (wave IV;
oxidation of E1 in dichloromethane occurs at 0.82 VFc1
)
8b
and the fourth at Epa ) +1.08 VFc (wave V; oxidation of
1
8b
(ii) oxidation of the intermediate R-carbonyl radical
B1 occurs at 1.05 VFc ).
8,22
(
Scheme 5).1
It turned out that in the case of silyl
Cyclic voltammograms of titanium bisenolate T2 in
acetonitrile reveal only two followup waves (Figure 4c).
In this case, the oxidation of the R-carbonyl radical R2
cannot be detected due to the fact that the oxidation
bisenol ether S3 Ipa does not exceed Ipa of monosilyl enol
ether S5. Thus, for S3 two electrons are involved in
substrate oxidation. For S1 the oxidation wave contains
2.74 electrons. The reversible waves of S3 and S4 in
2
0c
potential of this species (Epa ) +0.24 VFc
)
is below that
-
1
of T2. The first follow-up wave appears at Epa ) +0.55
dichloromethane at 1000 V‚s contain about one elec-
tron. In the case of T1 the reversible oxidation wave
contains also one electron as could be demonstrated by
direct comparison of its oxidation current with that of
the reversible wave of the corresponding titanium mo-
8,20c
V
Fc and can be assigned to the oxidation of enol E2.1
The second follow-up wave is the partially reversible
ox
18,20c
wave of benzofuran B2 at E1/2 ) +0.87 VFc
.
ox
In dichloromethane, the situation for T2 (E1/2 ) +0.19

3
082 J . Org. Chem., Vol. 64, No. 9, 1999
Schmittel et al.
Sch em e 5
Sch em e 7
Sch em e 6
dissociative electron-transfer process; (3) bond cleavage
at the stage of the dication generated by the dispropor-
tionation of two metal bisenolate radical cations can be
ruled out since the kinetics of the follow-up reaction of
the radical cations does not show a concentration depen-
dence; and (4) a direct cyclization on stage of the radical
cations followed by liberation of the metal fragment (Ti,
Si) can be ruled out for steric reasons.
noenolate Mes
2 2
CdC(H)OTiCp Cl, which is known to
1
7
contain only a single electron.
As for the above systems two different mesolytic
cleavage pathways (homo- vs heterolytic) are possible
-
1
If recorded at adequate scan rates (v ) 10-1000 V‚s
for silicon bisenolates; v ) 20-500 mV‚s-1 for T1) the
oxidation waves of the model compounds in dichlo-
romethane become increasingly reversible. The ratio of
the reductive to oxidative peak current Ipc/Ipa is indicative
(
Scheme 7). To differentiate between the two reaction
pathways data are needed on the oxidation potentials of
the two cleavage parts, i.e., the R-carbonyl radicals and
metal(III)monoenolate. The preferred pathway will afford
the more readily oxidizable cleavage product as a cation.
f
of the rate constant k of the follow-up reaction (see
Scheme 4). Ipc/Ipa can be computed as function of the
While for the monoenolate radical cations of T3, T4,
mechanism (EC vs ECE) through digital simulation as
S5, and S61
2,17,22
the heterolytic cleavage mode was
described in earlier publications.17,22 For all compounds
derived from oxidation potential considerations, we have
obtained direct evidence for the heterolytic route in the
titanium bisenolate radical cation cleavage as the oxida-
tion wave of the R-carbonyl radical in the cyclic voltam-
mogram of T1 and T2 could be detected. In addition, the
heterolytic cleavage mode could equally be rationalized
by redox potential considerations.27 Similar consider-
ations apply for the silicon bis(tris)enolate radical cation
bond cleavage reactions. Consequently, the heterolytic
pathway operates for all the systems under investigation,
indicating that in sterically unshielded systems (cf.
Scheme 1) indeed a rapid M-O bond cleavage would be
disastrous as 1,4-dicarbonyl substrates would form via
nonstereoselective radical dimerization.
f
studied, k is not dependent on the substrate concentra-
tion (Table 5).
When the oxidation was undertaken under conditions
that the substrate wave was reversible, a second, but
irreversible oxidation wave was registered at a anodically
shifted potential. This wave can be assigned to the
oxidation of the radical cation to the dication (Table 6).
Discu ssion
Mesolyt ic Bon d Clea va ge. Oxidative formation of
1
8
benzofurans B1 and B2 from simple, stable enols and
various other enol derivatives1
7,22-26
is known to proceed
via an intermediate R-carbonyl cation that undergoes a
Nazarov-type cyclization, [1,2]-methyl shift, and depro-
tonation (Scheme 6). In all these oxidations mesolytic
bond cleavage constitutes the key reaction.
The mesolytic bond cleavage can follow two distinct
mechanistic pathways: the homolytic or the heterolytic
variant.19 The heterolytic pathway has been observed for
the deprotonation of enol radical cations as well as in
Kin etics of th e Mesolytic Bon d Clea va ge. Because
the only possible follow-up reaction of the metal bis- and
trisenolate radical cations is the mesolytic Si-O and
Ti-O bond cleavage, respectively, this process can be
kinetically investigated by cv measurements. For that
purpose, the ratio of cathodic to anodic peak current Ipc
/
I
pa of partially reversible oxidation waves has to be
2
2
23
the cleavage of silyl enol ether, enol phosphite, enol
determined (Table 4). For a precise kinetic analysis the
exact mechanism of the follow-up reaction has to be
known and an appropriate working curve for this mech-
anism must be calculated. In the case of the silicon
enolates this process follows an ECirrE mechanism for
which the working curve is already available. For tita-
nium bisenolate T1 the situation is a bit more compli-
cated. On one hand the process obeys an ECirr mechanism
as the oxidation potential of T1 is below that of the
R-carbonyl radical R1. However, the difference between
2
3
17
phosphinate, titanium enolate, and zirconium enolate
radical cations.24 The homolytic variant is realized in the
2
5
26
fragmentation of enol ester, enol carbonate and enol
carbamate radical cations.26
In analogy to the bond cleavage of titanium17 and
silicon22 monoenolate radical cations the primary follow-
up reaction of the investigated silicon and titanium bis-
and trisenolate radical cations is equally mesolytic bond
cleavage of the Ti-O and Si-O bond, respectively. This
mechanistic assignment is based on several observa-
tions: (1) the formation of benzofurans is indicative of a
Ti-O and Si-O bond cleavage at some stage of the
reaction; (2) the direct observation of short-lived metal
bisenolate radical cations excludes the possibility of a
ox
ox
(27) Since the E1/2 of T2 is lower than that of T4 (E1/2 ) +0.31
Fc), one expects the potential of [Mes CdC(Ph)O]Ti(III)Cp to be
shifted cathodically to that of Ti(III)Cp Cl (Epa ) -0.68 VFc). Therefore,
the corresponding R-carbonyl radicals (Epa(R1) ) +0.36 VFc; Epa(R2)
V
2
2
2
2
0c
) +0.24 VFc
)
should possess a higher oxidation potential.

Sequential M-O Bond Cleavage
J . Org. Chem., Vol. 64, No. 9, 1999 3083
Ta ble 4. Com p a r ison of th e Rela tive P ea k Cu r r en ts Ip a
of Som e Su bstr a tes a t Equ im ola r Am ou n ts (Th e Va lu e
for En ol E2 (Mes2CdC(P h )OH) Wa s Ta k en P er
Ta ble 6. Oxid a tion P oten tia ls of Ir r ever sible F ollow -Up
Wa ves of Rever sible Su bstr a te Wa ves
S1
S2
S3
S4
S5
S6
Defin ition em a s 2.00 () a Tw o-Electr on Wa ve, See Text))
a
370b
1.04
∆
Ec (mV), CH2Cl2
Epa (VFc)
280
1.16
300
nd
300
1.04
320
1.07
450
1.14
enol E2a
S1a
S3a
S3b
S4b
S5a
c
Ipa ≡ 2.00 Ipa ) 2.74 Ipa ) 1.93 Ipa ) 1.15 Ipa ) 0.93 Ipa ) 1.92
a
b
∆
Ec ) Epa (follow-up wave) - E1/2 (substrate wave). 420 mV
a
-1
-1
c
Irreversible oxidation waves in acetonitrile at 100 mV‚s
.
.
in acetonitrile. In dichloromethane.
b
Reversible oxidation wave in dichloromethane at v ) 1000 V‚s
The mesolytic bond cleavage rate of T1•+ is the slowest
in this series. The corresponding titanium monoenolate
Ta ble 5. F ir st-Or d er Ra te Con sta n ts kf for Mesolytic
M-O Bon d Clea va ge in Dich lor om eth a n e (see Sch em e 4)
•+
radical cation T3 shows a somewhat faster fragmenta-
tion rate, which nevertheless is within the same order
of magnitude.28 The mesolytic bond cleavage of T2 is
S1a
S522
S622
compd
S2
S3
T1
•+
1
2
k
f
(s-1) 9 × 10 8 × 10-1 4 × 10-1 1.3 × 10 1.3 × 10-1 9 × 10-2
•
+
definitively faster than that of T1 ; however, no good
a
In this case kf is probably a rough approximation because the
kinetic values were available. Since fully reversible waves
mechanism differs from the other compounds since Ipa at 100
-1
could be obtained at scan rates v g 20 V s the rate
-
1
mV‚s contains more than two electrons, see Table.
constant for this process can be estimated to k
f
< 1 ×
2
-1
1
0 s . The reason for the faster Ti-O bond cleavage of
these potentials is very small; therefore, a part of the
anodic peak current of the investigated wave may result
from oxidation of R1. On the other hand, protons are split
off in the course of benzofuran formation, and they
hydrolyze neutral T1 to the corresponding enol E1. That
this process is operative can be seen by the fact that an
oxidation wave for E1 appears in cyclic voltammograms
of T1. However, this process is only effective when the
titanium bisenolate is oxidized partially reversibly and
irreversibly. This opens an additional reaction pathway
•+
•+
T2 than T1 is most probably the steric strain caused
•+
by the bulkier enolate substituents of T2 .
Oxid a tion of th e Secon d En ola te Electr op h or e.
So far, the above considerations were concerned with the
first M-O bond cleavage reaction in metal bisenolate
radical cation. However, the one-electron oxidation ex-
periments reveal that both enolate electrophores of the
investigated metal bisenolates can be oxidatively cyclized
to the corresponding benzofuran derivatives if 4 equiv of
the one-electron oxidant F eP h en are employed (Table
•+
•+
for T1 that is intermolecular electron transfer from T1
2
).
to neutral E1, complicating the mechanistic interpreta-
tion. For that reason, we have carried out digital simula-
tions using the experimentally determined parameters
Several mechanistic alternatives are possible for the
oxidative cyclization of the second enolate electrophore
during the course of the preparative one-electron oxida-
tion experiments: (1) After the cleavage of the metal
bisenolate radical cation the resultant metal(IV)mo-
noenolate cations are further oxidized, which induces a
second M-O cleavage process; (2) as a variant of (1) the
intermediate metal enolate cation reacts with a nucleo-
phile in the reaction mixture, thus leading to an metal
enolate derivative exhibiting an even lower oxidation
potential; (3) the acid liberated in the oxidative benzo-
furan formation (see Scheme 5) hydrolyzes part of the
metal bisenolate, thereby setting free the enols that are
oxidized to the benzofurans.
to quantify the deviation of an ECirr and an ECirr
E
mechanism. These studies28 revealed that the usage of a
working curve for the ECirrE mechanism results only in
a very small inaccuracy that lies within the borders of
experimental error of this method.
The recorded cyclic voltammograms were analyzed
according to the appropriate working curves. This led us
to the determination of the concentration-independent,
first-order rate constants k for the mesolytic M-O bond
f
cleavage of metal bis- and trisenolate radical cations in
dichloromethane at room temperature, which are de-
picted in Table 5.
As could be shown previously,22 the Si-O bond cleav-
age of silyl enol ether radical cations is induced by
nucleophiles. This has qualitatively been verified for the
silyl bisenolates as well through the addition of methanol.
In light of the enol oxidation wave that is visible in
the CV experiments at low scan rates for most systems,
mechanism 3 should dominate in the absence of a strong
base. However, the appearance of wave III in cyclic
voltammetry experiments of T1 in dichloromethane
(Figure 4), which was tentatively assigned to the oxida-
tion of the metal(IV)monoenolate cation, indicates that
route 1 may be equally viable to some extent.
f
As a consequence, the cleavage rate constant k should
depend on the steric shielding of the Si atom.22 Thus,
mainly for steric reasons and because of the nonnucleo-
•+
•+
2 2
philic solvent CH Cl , the rate constants of S2 and S3
-
1
-1
-1 -1
(k
f
) 8 × 10
s
and 4 × 10
s
) are quite low. In line
Con clu sion . The present investigation on metal bise-
nolates indicates that cleavage of the M-O bond is a slow
process (M ) Si, Ti), which takes place on the stage of
the radical cations and not of the dications. Earlier
studies concerning the intramolecular coupling of steri-
cally unencumbered metal bisenolates (cf. Scheme 1)
revealed that the primary follow-up reaction of such
with this reasoning, their rate constants compare better
to that of the sterically shielded silyl monoenolate radical
•
+
-1
-1
•+
cation S6 (k
f
) 1.3 × 10
s ) than of S5 . In
•
+
comparison, the sterically more shielded compound S3
•
+
undergoes a slower scission than S2 . Hence, even in
dichloromethane Si-O cleavage seems to be partly
induced in a nucleophile assisted process. On the other
-
1
9
radical cations is very fast (k > 10 s ). This rate
constant is much higher than that of the M-O bond
scission in the hindered bisenolate radical cations, sug-
gesting that the fast primary follow-up reaction of
sterically uncongested metal bisenolate radical cations
can be assigned to the proposed intramolecular C-C bond
formation depicted in Scheme 1.
•
+
hand, the bond cleavage in S1 is faster by about 2
•+
orders of magnitude, almost matching the rate S5 . The
higher rate is presumably due to the reduced shielding
and the higher oxidation potential.
(28) S o¨ llner, R. PhD Thesis, W u¨ rzburg, 1998.

3
084 J . Org. Chem., Vol. 64, No. 9, 1999
Schmittel et al.
In summary, we synthesized sterically congested sili-
acetonitrile, colorless crystals were obtained (0.16 g, 0.21
1
mmol, 38%): mp 102 °C; H NMR (200 MHz, CDCl
3
) δ -0.18
con and titanocene bisenolates, one of which was a
characterized by X-ray structure analysis demonstrating
the steric bulk about the O-M-O moiety and the
â-carbons of the enolate units. We have investigated and
characterized the bisenolate radical cations that undergo
a clean mesolytic M-O bond cleavage process. The first-
order kinetics of these fragmentations could be deter-
mined by cyclic voltammetry. Both electrophores present
in these systems can oxidatively be converted into the
benzofuran. Furthermore, we made a suggestion con-
cerning the mechanistic details of the oxidation of the
second enolato group.
(
br s, 6 H), 1.72 (br s, 6 H), 1.92 (s, 12 H), 2.18 (s, 6 H), 2.24
s, 6 H), 2.43 (br s, 6 H), 6.63 (s, 4 H), 6.78 (br s, 4 H), 6.80-
.10 (m, 8 H), 7.26-7.29 (m, 2 H); 13C NMR (50 MHz, CDCl
(
7
3
)
δ -2.8, 20.7, 20.8, 21.1, 21.4, 119.8, 127.2, 127.5, 129.1, 129.2,
29.3, 135.6, 135.7, 136.3, 137.8, 138.3, 149.3; IR (KBr) 2921,
1610, 1560, 1474, 1444, 1255, 1200, 1151, 1077, 1030, 853, 806
1
-
1
cm . Anal. Calcd for C54
C, 84.52; H, 7.97.
60 2
H O Si: C, 84.33; H, 7.86. Found:
Bis(2,2-d im esit yl-1-p h en ylet h en oxy)m et h ylp h en ylsi-
la n e (S3). S3 was prepared analogously to S1 with the
following reagents: 2,2-dimesityl-1-phenylethenol (E2) (0.57
g, 1.6 mmol), triethylamine (0.16 g, 0.22 mL, 1.6 mmol),
dichloromethylphenylsilane (0.15 g, 0.80 mmol), and NaI (0.24
g, 1.6 mmol). Chromatography was performed using dichlo-
romethane/n-hexane 1:1 (R ) 0.69). The product was obtained
f
Exp er im en ta l Section
as colorless crystals (0.19 g, 0.23 mmol, 29%) and was
1
All reactions were carried out under an atmosphere of dry
argon by using standard Schlenk tube techniques. Solvents
were purified by standard literature methods and distilled
directly from their drying agents under nitrogen: THF/
potassium, acetonitrile/calcium hydride, dichloromethane/
recrystallized from acetonitrile: mp 91-93 °C; H NMR (200
MHz, CDCl ) δ -0.10 (br s, 3 H), 1.54 (br s, 6 H), 1.87 (s, 6 H),
3
1.93 (s, 6 H), 2.14 (s, 6 H), 2.22 (br s, 12 H), 6.45 (s, 4 H), 6.59
1
3
(s, 4 H), 6.65-7.18 (m, 10 H), 7.25-7.38 (m, 5 H); C NMR
(50 MHz, CDCl ) δ -2.8, 20.7, 20.8, 21.0, 21.1, 21.2, 21.3, 120.0,
3
P
4
O
10. Solvents for CV measurements and one-electron oxi-
127.3, 127.6, 128.8, 129.3, 129.6, 129.8, 132.7, 133.3, 133.4,
133.5, 133.9, 134.2, 134.8, 135.4, 135.6, 136.4, 137.7, 138.0,
138.7, 148.8; IR (KBr) 2918, 1609, 1593, 1491, 1443, 1261,
dation experiments: acetonitrile was purchased in HPLC
quality from Riedel-de-Ha e¨ n, distilled from calcium hydride
and filtered through basic alumina (ICN); dichloromethane
was purchased in HPLC quality from Riedel-de-Ha e¨ n, distilled
-
1
1202, 1123, 1077, 1028, 829, 787, 735, 697 cm . Anal. Calcd
for C H O Si: C, 85.25; H, 7.52. Found: C, 84.89; H, 7.40.
5
9
62
2
from P
4
O
10 and filtered through basic alumina (ICN). Dichlo-
Tr is(2,2-dim esityl-1-ph en yleth en oxy)m eth ylsilan e (S4).
S4 was prepared analogously to S1 with the following re-
agents: 2,2-dimesityl-1-phenylethenol (E2) (0.57 g, 1.6 mmol),
triethylamine (0.16 g, 0.22 mL, 1.6 mmol), NaI (0.24 g, 1.6
mmol), and trichloromethylsilane (80 mg, 0.53 mmol). Puri-
fication of the product was performed by chromatography with
n-hexane/diethyl ether 4:1 (Rf ) 0.72). The product was
obtained as a crystalline solid. After recrystallization from
n-hexane colorless crystals (0.16 g, 0.14 mmol, 28%) were
2
9
30
rotitanocene, 2,2-dimesitylethenol, and 2,2-dimesityl-1-
1
3
phenylethenol were prepared according to literature proce-
dures. The supporting electrolyte tetra(n-butyl)ammonium
hexafluorophosphate (Fluka) was of electrochemical grade and
1
13
used without further purification. H and C NMR spectra
were recorded on Bruker AM 200 and AM 250 spectrometers;
chemical shifts refer to tetramethylsilane; owing to the nature
1
3
of the â,â-dimesityl moiety signals in the C NMR spectra
often are superimposed. IR spectra were recorded on a Perkin-
Elmer FT-IR 1605. Elemental analyses were carried out on a
Carlo Erba elemental analyzer 1106. Melting points were
determined by using a differential scanning calorimeter Du-
pont 910.
1
obtained: mp 52 °C; H NMR (200 MHz, CDCl ) δ -0.52 (br
3
s, 3 H), 1.74 (br s, 9 H), 1.90 (s, 9 H), 1.93 (s, 9 H), 2.18 (s, 9
H), 2.23 (s, 9 H), 2.45 (br s, 9 H), 6.63 (s, 6 H), 6.79 (br s, 6 H),
7.00-7.12 (m, 9 H), 7.26-7.29 (m, 6 H); 13C NMR (50 MHz,
CDCl ) δ -5.9, 20.7, 20.8, 21.0, 21.3, 120.3, 127.3, 127.7, 128.7,
3
Bis(2,2-d im esityleth en oxy)m eth ylp h en ylsila n e (S1). A
solution of 2,2-dimesitylethenol (E1) (0.39 g, 1.4 mmol) in
acetonitrile was treated with triethylamine (0.14 g, 0.20 mL,
.4 mmol) and allowed to stir for 10 min. Then, dichloro-
methylphenylsilane (0.13 g, 0.70 mmol) and successively NaI
0.21 g, 1.4 mmol) were added. After the mixture was stirred
for 2 d, the solvent was removed under reduced pressure. The
residue was filtered through a short column of deactivated
10% water) silica gel with dichloromethane/n-hexane 1:1 as
129.2, 135.7, 135.9, 136.1, 137.8, 138.1, 138.3, 138.9, 148.4;
IR (KBr) 2920, 1654, 1610, 1560, 1491, 1443, 1241, 1200, 1151,
-
1
1078, 1030, 974, 944, 875, 825, 695 cm . Anal. Calcd for
Si: C, 85.51; H, 7.63. Found: C, 85.30; H, 7.78.
Bis(2,2-d im esityleth en oxy)tita n ocen e (T1). To a sus-
1
79 84 3
C H O
(
pension of sodium hydride (316 mg, 13.2 mmol) in THF (70
mL) was added solid 2,2-dimesitylethenol (E1) (1.01 g, 3.59
mmol) in small portions at room temperature. The mixture
was stirred for 1 h, and the surplus sodium hydride was
removed by filtration. Titanocene dichloride (448 mg, 1.80
mmol) was added to the filtrate and stirred overnight at room
temperature. The solvent was removed under reduced pres-
sure, and the residue was taken up in dichloromethane (70
mL) and filtered. Purification of the product was performed
(
eluent (R
f
) 0.65). The product was obtained as a colorless oil
that crystallized on standing. After recrystallization from
acetonitrile colorless crystals were obtained (0.18 g, 0.3 mmol,
1
3
3
2
7
2
1
1
1
9%): mp 68-69 °C; H NMR (200 MHz, CDCl
3
) δ 0.31 (br s,
H), 1.76 (br s, 6 H), 1.99 (s, 6 H), 2.03 (s, 6 H), 2.07 (s, 6 H),
.21 (s, 6 H), 2.25 (s, 6 H), 6.34 (s, 2 H), 6.76 (br s, 8 H), 7.24-
f
by chromatography on silica gel with dichloromethane (R )
1
3
.47 (m, 5 H); C NMR (50 MHz, CDCl
3
) δ -4.0, 20.5, 20.7,
0.95). The product was obtained as a dark red microcrystalline
1
0.9, 21.0, 21.1, 21.3, 120.5, 127.8, 128.6, 129.4, 129.9, 130.5,
30.7, 133.1, 133.2, 133.7, 134.1, 135.1, 135.5, 135.6, 137.6,
40.6; IR (KBr) 2958, 2918, 1625, 1560, 1478, 1439, 1375, 1261,
206, 1170, 1125, 1087, 850, 812 cm-1. Anal. Calcd for
solid (777 mg, 1.05 mmol, 59%): mp 211 °C dec; H NMR (250
MHz, CDCl
3
) δ 1.75 (br s, 6 H), 1.98 (s, 12 H), 2.23 (s, 12 H),
2.48 (br s, 6 H), 6.05 (s, 10 H), 6.70 (br s, 8 H), 6.91 (s, 2 H);
13
C NMR (63 MHz, CDCl ) δ 20.8, 20.9, 21.2, 111.2, 115.6,
3
C
47
H
54
O
2
Si: C, 83.14; H, 8.02. Found: C, 82.85; H, 8.10.
Bis(2,2-d im esit yl-1-p h en ylet h en oxy)d im et h ylsila n e
S2). S2 was prepared analogously to S1 with the following
128.4, 129.4, 134.6, 134.8, 135.3, 136.4, 137.6, 159.4; IR (KBr)
2955, 2915, 2849, 1609, 1578, 1556, 1477, 1441, 1372, 1246,
-1
(
1219, 1173, 1089, 1016, 904, 850, 807, 680, 627, 554, 487 cm .
reagents: 2,2-dimesityl-1-phenylethenol (E2) (0.39 g, 1.1
mmol), triethylamine (0.11 g, 0.15 mL, 1.1 mmol), dichlo-
rodimethylsilane (70 mg, 0.55 mmol), and NaI (0.16 g, 1.1
mmol). Chromatography was performed with n-hexane/diethyl
Anal. Calcd for C50
H, 7.83.
56 2
H O Ti: C, 81.50; H, 7.66. Found: C, 81.53;
Bis[2,2-d im esityl-1-p h en yleth en oxy]tita n ocen e (T2).
To a suspension of sodium hydride (417 mg, 17.4 mmol) in THF
(70 mL) was added solid 2,2-dimesityl-1-phenylethenol (E2)
(713 mg, 2.00 mmol) in small portions at room temperature.
The mixture was stirred for 1 h, and the surplus sodium
hydride was removed by filtration. Titanocene dichloride (249
mg, 1.00 mmol) was added to the filtrate and the mixture
refluxed for 6 h. After cooling, the solvent was removed under
ether 4:1 (R
f
) 0.74). The product was obtained as a colorless
oil that crystallized on standing. After recrystallization from
(
29) Wilkinson, G.; Birmingham, J . B. J . Am. Chem. Soc. 1954, 76,
281.
30) Biali, S. E.; Rappoport, Z. J . Am. Chem. Soc. 1984, 106, 5641.
4
(

Sequential M-O Bond Cleavage
J . Org. Chem., Vol. 64, No. 9, 1999 3085
reduced pressure. The residue was taken up in dichlo-
romethane, filtered, and crystallized from dichloromethane/
acetonitrile. The product was obtained as black crystals (152
Da ta of 3-m esityl-2-p h en yl-4,6,7-tr im eth ylben zo[b]fu -
r a n (B2):1 H NMR (250 MHz, CDCl
8 1
) δ 1.89 (s, 3 H), 2.02 (s,
3
6 H), 2.39 (s, 3 H), 2.40 (s, 3 H), 2.54 (s, 3 H), 6.76 (s, 1 H),
6.98 (s, 2 H), 7.17-7.30 (m, 3 H), 7.48-7.55 (m, 2 H).
mg, 171 µmol, 17%) suitable for X-ray analysis: mp 164 °C
1
dec; H NMR (200 MHz, CDCl
3
) δ 2.04 (br s, 12 H), 2.12 (s, 6
Cyclic Volta m m etr y. In a glovebox tetra(n-butyl)ammo-
niumhexafluorophosphate (232 mg, 600 µmol) and the elec-
troactive species (6 µmol) were placed into a thoroughly dried
CV cell. At a high purity argon line acetonitrile or dichlo-
romethane (6.0 mL) was added through a gastight syringe.
Then a platinum disk working electrode (L ) 1 mm), a
platinum wire counter electrode, and a silver wire as pseudo
reference electrode were placed into the solution. The cyclic
voltammograms were recorded at various scan rates using
different starting and switching potentials. For determination
of the oxidation potentials, ferrocene (E1/2 ) +0.39 V vs SCE)
was added as the internal standard. Cyclic voltammograms
were recorded using a Princeton Applied Research model 362
potentiostat with a Philips model PM 8271 XYt-recorder for
H), 2.19 (s, 6 H), 2.28 (br s, 12 H), 5.60 (s, 10 H), 6.58 (s, 4 H),
6
.83 (s, 4 H), 7.05-7.15 (m, 6 H), 7.40-7.55 (m, 4 H); 13C NMR
50 MHz, CDCl ) δ 20.7, 20.8, 21.6, 22.2, 113.2, 116.7, 126.9,
(
3
1
1
1
1
27.5, 128.9, 129.1, 130.8, 134.8, 135.4, 137.9, 138.1, 138.2,
38.5, 141.3, 167.1; IR (KBr) 2989, 2960, 2916, 2853, 1608,
526, 1488, 1474, 1442, 1370, 1276, 1243, 1201, 1153, 1071,
017, 968, 912, 852, 815, 773, 747, 700, 661, 628, 595, 570 cm-1
.
Anal. Calcd for C62
H, 7.37.
64 2
H O Ti: C, 83.76; H, 7.26. Found: C, 83.46;
Gen er a l P r oced u r e for On e-Electr on Oxid a tion s. In an
argon-filled glovebox the desired amounts of the one-electron
oxidant and the metal bisenolates were placed into two
separate test tubes equipped with stirring rods. At a high
purity argon line, 3 mL of the appropriate solvent (acetonitrile
or dichloromethane) was added to each test tube to dissolve
the reactants. The solution of the metal bisenolates was added
via syringe to the solution of the one-electron oxidant. The
resulting mixture was stirred at room temperature for 1 min
-
1
scan rates <1 V s . For fast scan cyclic voltammetry, a
Hewlett-Packard model 3314A function generator was used
connected to a three-electrode potentiostat developed by Ama-
tore.32 The employed working electrodes were self-made gold
(
diameter: 25 µm) and platinum (diameter: 10 µm) ultrami-
croelectrodes. The ratios Ipc/Ipa were determined according to
(
in the case of the titanium bisenolates) or 14 h (in the case of
the silicon enolates), quenched with saturated aqueous NaH-
CO (10 mL), and diluted with dichloromethane (10 mL). The
the procedure of Nicholson.33
3
Ack n ow led gm en t. We gratefully acknowledge the
financial support by the Deutsche Forschungsgemein-
schaft (SFB 347: ”Selective Reactions of Metal Acti-
vated Molecules”). In addition, we are most indebted to
the Fonds der Chemischen Industrie for the ongoing
support of our research as well as to the Degussa for a
generous gift of electrode materials.
aqueous layer was extracted three times with dichloromethane,
and the combined organic layers were washed with water and
dried with Na
was filtered through silica gel in order to remove the red [Fe-
phen) ](PF . Removal of the solvent afforded the crude
2 4
SO . If F ep h en was used as oxidant the solution
(
3
6 2
)
1
product. Product analysis was performed by H NMR spec-
troscopy. All products were identified by comparison with data
of authentic samples. Yields were determined by adding
1
Su p p or tin g In for m a tion Ava ila ble: X-ray data for tita-
nium bisenolate T2 (CCDC-115647) are provided. This
material is available free of charge via the Internet at
http://pubs.acs.org.
m-nitroacetophenone as an internal H NMR standard.
Da ta of 3-m esityl-4,6,7-tr im eth ylben zo[b]fu r a n (B1):31
H NMR (250 MHz, CDCl ) δ 1.92 (s, 3 H), 2.06 (s, 6 H),
1
3
2
2
.33 (s, 3 H), 2.38 (s, 3 H), 2.43 (s, 3 H), 6.75 (s, 1 H), 6.94 (s,
H), 7.33 (s, 1 H).
J O981793Z
(
32) Amatore, C.; Lefrou, C.; Pfl u¨ gler, F. J . Electroanal. Chem.
(31) Schmittel, M.; Baumann, U. Angew. Chem., Int. Ed. Engl. 1990,
Interfacial Electrochem. 1989, 270, 43.
2
9, 541.
(33) Nicholson, R. S. Anal. Chem. 1966, 38, 1406.