The role of ligand displacement in Sm(II)-HMPA-based reductions

Article abstract of DOI:10.1021/ja049161j

Addition of HMPA to [Sm{N(SiMe₃)₂}₂] produces a less reactive reductant in contrast to addition of HMPA to Sml ₂. While the [Sm{N(SiMe₃)₂}₂]-HMPA combination results in a more powerful reductant based on the redox potential, the observed decrease in reactivity is attributed to steric hindrance caused by the nonlabile ligand-N(SiMe₃)₂ and HMPA around the Sm metal. The importance of ligand displacement (exchange) in Sm(II)-HMPA-based reactions and insight into the mechanism of [Sm{N(SiMe₃) ₂}₂]-HMPA and Sml₂-HMPA reductions are presented.

Full text of DOI:10.1021/ja049161j

Published on Web 05/11/2004
The Role of Ligand Displacement in Sm(II)-HMPA-Based
Reductions
†
‡
,†
Edamana Prasad, Brian W. Knettle, and Robert A. Flowers, II*
Contribution from the Department of Chemistry, Lehigh UniVersity,
Bethlehem, PennsylVania 18015, and Department of Chemistry and Biochemistry,
Texas Tech UniVersity, Box 41061, Lubbock, Texas 79409-1061
Received February 16, 2004; E-mail: rof2@lehigh.edu
Abstract: Addition of HMPA to [Sm{N(SiMe
of HMPA to SmI . While the [Sm{N(SiMe
based on the redox potential, the observed decrease in reactivity is attributed to steric hindrance caused
by the nonlabile ligand -N(SiMe and HMPA around the Sm metal. The importance of ligand displacement
exchange) in Sm(II)-HMPA-based reactions and insight into the mechanism of [Sm{N(SiMe ]-HMPA
3 2
) }
2
] produces a less reactive reductant in contrast to addition
2
3 2 2
) } ]-HMPA combination results in a more powerful reductant
3 2
)
(
3 2 2
) }
and SmI -HMPA reductions are presented.
2
Introduction
examination of the reactivity of SmI -HMPA and Sm(OTf) -
2
2
HMPA shows that these reductant-HMPA combinations have
The addition of HMPA to reactions that employ SmI2
considerably accelerates the electron-transfer process. The
seminal work of Inanaga demonstrated that the rate of reduction
of organic halides by SmI2 was increased dramatically upon
5
similar reactivity patterns and hence may produce similar
-
intermediates (if OTf is displaced in a manner analogous to
-
I ). Conversely, hard, basic alkoxide and amide ligands should
have a higher affinity for Sm(II) and as a result would be less
likely to be displaced upon addition of HMPA or basic
substrates. Upon addition of HMPA to SmI2, the reducing power
1
addition of HMPA. Aside from favorable rate enhancements,
HMPA addition to SmI2 provides increased diastereoselectivities
2
and beneficial product distributions in a wide range of reactions.
4,6
of the complex is enhanced, and the rates of reaction increase
dramatically; that is, enhanced reducing power leads to faster
A study by Molander showed that the addition of HMPA
produces a sterically encumbered reductant that not only dictates
the diastereoselectivity of reductive coupling reactions, but also
stabilizes reactive intermediates (ketyls, radicals) in close
proximity to the Sm-HMPA complex, thereby preventing
1
,7
reaction rates. Based on the brief discussion above, one may
ask, how does HMPA influence the reducing power and rates
of reduction for a Sm(II) reductant containing a less labile
ligand? Is ligand displacement mechanistically important in the
reduction of various organic functional groups? To initially
address these questions, the impact of HMPA on the redox
potential and rates of reaction of a Sm(II) complex containing
the less labile -N(SiMe3)2 ligand was examined. The experi-
ments presented herein show that ligand displacement upon
HMPA addition plays a role in the reduction of organic
functional groups by Sm(II) reagents, presumably through
altering the steric environment of the reductant.
3
competing reaction processes. Although numerous other ad-
ditives have been used and alternative protocols have been
developed, none appear to approach the general utility of SmI2-
HMPA. Thus, despite its carcinogenicity, HMPA remains the
additive of choice for reactions of SmI2.
Mechanistic studies of Sm(II)-based reagents and the unusual
effects of HMPA on them are beginning to appear in the
literature. The elegant work of Daasbjerg and Skrydstrup has
shown that addition of 4 equiv of HMPA to SmI2 in THF likely
Experimental Section
produces [Sm(HMPA)4(THF)2]I2 and addition of greater than
4
1
0 equiv of HMPA produces [Sm(HMPA)6]I2. Because HMPA
Material and General Procedures. THF was distilled from sodium
benzophenone ketyl, under nitrogen atmosphere. HMPA was dried by
vacuum distillation from calcium oxide and degassed and backfilled
is an excellent donor ligand, displacement of the labile iodide
to the outer sphere of the complex is not surprising. Because it
-
is possible to displace I upon the addition of HMPA to SmI2,
with N
stored in an Innovative Technology, Inc. drybox containing a nitrogen
3 2 2
atmosphere and a platinum catalyst for drying. The [Sm{N(SiMe ) } ]
2
at least three times to remove all oxygen. Dried solvents were
does this displacement have mechanistic consequences? Limited
8
†
complex was prepared according to the procedure of Evans. The
Lehigh University.
Texas Tech University.
‡
(
1) Inanaga, J.; Ishikawa, M.; Yamaguchi, M. Chem. Lett. 1987, 1485.
2) (a) Molander, G. A. Chem ReV. 1992, 92, 29. (b) Molander, G. A.; Harris,
C. R. Chem. ReV. 1996, 96, 307. (c) Steel, P. G. J. Chem. Soc., Perkin
Trans. 1 2001, 2727.
(5) Fukuzawa, S.; Mutoh, K.; Tsuchimoto, T.; Hiyama, T. J. Org. Chem. 1996,
61, 5400.
(
(6) (a) Shabangi, M.; Flowers, R. A., II. Tetrahedron Lett. 1997, 38, 1137. (b)
Shabangi, M.; Kuhlman, M. L.; Flowers, R. A., II. Org. Lett. 1999, 1, 2133.
(c) Enemaerke, R. J.; Daasbjerg, K.; Skrydstrup, T. Chem. Commun. 1999,
343-344.
(3) Molander, G. A.; McKie, J. A. J. Org. Chem. 1992, 57, 3132.
4) Enemaerke, R. J.; Hertz, T.; Skrydstrup, T.; Daasbjerg, K. Chem.-Eur. J.
(
2
000, 6, 3747.
(7) Prasad, E.; Flowers, R. A., II. J. Am. Chem. Soc. 2002, 124, 6895.
10.1021/ja049161j CCC: $27.50 © 2004 American Chemical Society
J. AM. CHEM. SOC. 2004, 126, 6891-6894
9
6891

A R T I C L E S
Prasad et al.
concentration of the Sm complex was determined by iodometric
9
titration. 2-Butanone and 1-iodobutane were purchased from Aldrich
and distilled under vacuum from calcium oxide before use.
The redox potential of the [Sm{N(SiMe ) } ]-HMPA complex was
3 2 2
measured by cyclic voltammetry, employing a BAS 100B/W MF-9063
Electrochemical Workstation. The working electrode was a standard
glassy carbon electrode. The electrode was polished with polishing
alumina and cleansed in an ultrasonic bath. The auxiliary electrode
was a platinum wire, and the reference electrode was a saturated Ag/
3
AgNO electrode. The scan rate for all experiments was 100 mV/s.
The electrolyte used was tetrabutylammonium hexafluorophosphate.
The concentration of the Sm(II) species in each experiment was 0.50
mM. All solutions were prepared in the drybox and transferred to the
electrochemical analyzer for analysis.
UV-vis experiments were performed on a Shimadzu UV-1601 UV-
visible spectrophotometer controlled by UV Probe (version 1.11)
software. The VPO experiments were carried out on a Wescor 5500-
XR vapor pressure osmometer operating at 25 °C in a drybox.
Calibration curves of VPO reading versus molality were obtained using
Figure 1. Stopped-flow decay trace of Sm{N[Si(CH3)3]2}2 (0.02 M) in
the presence of 2-butanone (0.20 M) and HMPA (10 equiv to Sm(II)
complex concentration) in THF.
9
biphenyl as the calibration standard.
Stopped-Flow Studies. Kinetic experiments in THF were performed
with a computer-controlled SX-18 MV stopped-flow reaction spec-
trophotometer (Applied Photophysics Ltd. Surrey, UK). The
3 2 2
[Sm{N(SiMe ) } ]-HMPA complex and substrates were taken sepa-
rately in airtight Hamilton syringes from a drybox and injected into
the stopped-flow system. The cell box and the drive syringes of the
stopped-flow reaction analyzer were flushed a minimum of three times
with dry, degassed solvents to make the system oxygen-free. The
concentration of [Sm{N(SiMe
study was 20 mM. The concentration of the substrates was kept high
relative to [Sm{N(SiMe ]-HMPA (0.20-0.40 M) to maintain
pseudo-first-order conditions. The reaction rate constants were deter-
3 2 2
) } ]-HMPA complexes used for the
3 2 2
) }
1
0
mined using standard methods. Reaction rates were determined from
the decays of the [Sm{N(SiMe ]-HMPA complex at 555 nm. The
decay of the [Sm{N(SiMe ] or the HMPA complex displayed first-
order behavior over >4 half-lives for [Sm{N(SiMe ]-substrate
combinations. The kinetics for 2-butanone followed the rate law (1)
shown below.
3 2 2
) }
3 2 2
) }
3 2 2
) }
Figure 2. The UV-vis spectra of [Sm{N(SiMe3)2}2] (thick solid line) and
[
Sm{N(SiMe3)2}2] containing 2 and 10 equiv of HMPA in THF (dashed
and thin solid lines, respectively).
-
d[Sm(II) complex]/dt ) k[Sm(II) complex][ketone] (1)
other Sm-based reductants will provide the insight necessary
to design alternative protocols for mimicking the useful features
of SmI2-HMPA. The goal of this work was to establish the
importance of HMPA-induced ligand displacement in reactions
of Sm(II)-based reductants. Experiments described herein were
designed to examine the effect of HMPA addition on a Sm
reductant containing a nonlabile ligand. Beginning experiments
were initiated to establish the effect of HMPA on the reducing
power of the [Sm{N(SiMe3)2}2] complex in THF. Follow-up
experiments established the likely structure of the complex
formed between [Sm{N(SiMe3)2}2] and HMPA and examined
the rates for reduction of 1-iodobutane and 2-butanone. With
these data, comparisons of both structure and reactivity could
be made for the [Sm{N(SiMe3)2}2]-HMPA complex and the
one formed between SmI2 and HMPA.
The kinetics of 1-iodobutane followed the rate law (2) shown below.
-
d[Sm(II) complex]/dt ) 2k[Sm(II) complex][alkyl iodide] (2)
A representative stopped-flow trace for the reduction of 2-butanone
by [Sm{N(SiMe ]-HMPA (10 equiv) is shown in Figure 1. The
3 2 2
) }
temperature studies used to determine activation parameters were carried
out over a range of 30-50 °C using a Neslab circulator connected to
the sample-handling unit of the stopped-flow system. The step size
used for the temperature study was 5 °C, and each kinetic trace was
recorded at a known temperature that was monitored by a thermocouple
in the reaction cell. The temperature measurements are accurate to 0.01
°
C.
Results and Discussion
Because HMPA is a hazardous material, one of the goals of
numerous research programs is the design of protocols that
mimic the favorable features of SmI2-HMPA, but do not require
the presence of the carcinogenic additive. A detailed mechanistic
understanding of the role of HMPA in reactions of SmI2 and
Figure 2 shows the UV-vis spectrum for [Sm{N(SiMe3)2}2]
and the spectra after the addition of 1 and 10 equiv of HMPA.
Upon addition of 1 equiv of HMPA, there is a slight blue shift
of the absorbance at 555 nm. A broad shoulder begins to appear
at 670 nm. Also, the intensity of the broad absorption at 410
nm sharpens to some extent. Further addition of HMPA (up to
(
8) Evans, W. J.; Drummond, D. K.; Zhang, H.; Atwood, J. L. Inorg. Chem.
988, 27, 575.
9) Shotwell, J. B.; Sealy, J. M.; Flowers, R. A., II. J. Org. Chem. 1999, 64,
10 equiv) displayed no further change in the UV-vis spectrum.
1
(
The absorption bands in Sm(II) complexes have been attributed
5
251.
n
n-1
1
11
to f to d (4f to 4f 5d ) transitions. Because the d orbitals
are spatially extended towards the ligand field, it is reasonable
(
10) Pedersen, S. U.; Lund, T.; Daasbjerg, K.; Pop, M.; Fussing, I.; Lund, H.
Acta Chem. Scand. 1998, 52, 657.
6892 J. AM. CHEM. SOC.
9
VOL. 126, NO. 22, 2004

Ligand Displacement in Sm(II)−HMPA-Based Reductions
A R T I C L E S
Figure 3. Cyclic voltammogram of [Sm{N(SiMe3)2}2] and [Sm{N(SiMe3)2}2] containing 2 and 10 equiv of HMPA at a glassy carbon electrode in THF
versus Ag/AgNO3 (sat’d) reference electrode.
to assume that displacement of THF by HMPA would result in
a perturbation of d orbital energy and hence a change in the
UV-vis spectrum. Upon addition of HMPA to SmI2, there is a
pronounced blue shift in the UV-vis spectrum.4 Addition of
HMPA to [Sm{N(SiMe3)2}2] also exhibits similar (although less
pronounced) changes in the UV-vis spectrum. Regardless of
the exact nature of the electronic transitions, comparison of the
absorption data with that obtained for SmI2-HMPA clearly
shows that the complex obtained by addition of HMPA to
2 equiv of HMPA provided data consistent with the average of
the MW of [Sm{N(SiMe3)2}2(HMPA)] and HMPA. Calorimet-
ric experiments were performed to estimate the stoichiometry
of the complex by titration of excess HMPA into a solution of
a,9
9
[Sm{N(SiMe3)2}2] in THF. Although it was difficult to obtain
statistically stable fits of the data, repeated experiments provided
results consistent with coordination of 2 equiv of HMPA
(Supporting Information). If a second equivalent of HMPA is
coordinating to [Sm{N(SiMe3)2}2(HMPA)], it is likely to have
a lower affinity than the first ligand, presumably due to steric
crowding around Sm. In the absence of definitive crystal-
lographic data, the collective evidence is consistent with
[Sm{N(SiMe3)2}2(HMPA)2] as the active reductant formed
between [Sm{N(SiMe3)2}2] and 10 equiv of HMPA.
[Sm{N(SiMe3)2}2] is unique.
Next, cyclic voltammetry was utilized to estimate the
thermodynamic redox potential of [Sm{N(SiMe3)2}2] in THF
and the impact of HMPA on the reducing power of the complex.
The voltammogram of [Sm{N(SiMe3)2}2] was irreversible
(
Figure 3), and the redox potential was found to be -2.1 ( 0.1
Although the VPO and calorimetric data do not clearly discern
whether 1 or 2 equiv of HMPA is coordinated to the Sm-
silylamide complex, the electrochemical data show that a more
powerful reductant is formed upon addition of HMPA to
12
V (vs sat’d Ag/AgNO3). Addition of 2 equiv of HMPA to the
complex produced a reductant with a redox potential of -2.5
(
0.1 V. Addition of up to 10 equiv of HMPA showed no
further impact on the redox potential. Another notable feature
of these experiments was that, upon addition of HMPA, the
irreversible voltammogram became quasireversible. The cyclic
voltammetry experiment clearly shows that a more powerful
reductant (in thermodynamic terms) is produced upon addition
of HMPA to [Sm{N(SiMe3)2}2] in a manner analogous to SmI2.
The experiments described to this point indicate that addition
of HMPA to [Sm{N(SiMe3)2}2] produces a reductant that is
unique as compared to the complex formed upon the addition
of HMPA to SmI2. However, these studies do not address the
solution structure of the active reductant in solution. Examina-
tion of the [Sm{N(SiMe3)2}2(THF)2] complex prepared by
Evans and co-workers shows that it is highly unlikely that more
[
Sm{N(SiMe3)2}2]. Because the hallmark of the SmI2-HMPA
combination is enhanced rates of reduction of a wide range
of functional groups (as compared to SmI2), a series of stopped-
flow experiments were carried out to examine the impact
of HMPA on the rates of reduction of 1-iodobutane and
2
-butanone by [Sm{N(SiMe3)2}2]. Reduction of 1-iodobutane
produced butane, while reduction of 2-butanone produced the
pinacol product after workup. Figure 4 shows a plot of k versus
HMPA concentration for the reduction of 1-iodobutane by
[Sm{N(SiMe ) } ]. The experiment yielded interesting results.
Upon the addition of 1 equiv of HMPA, the rate of reduction
of 1-iodobutane was enhanced. However, upon further addition
of another equivalent of HMPA, the rate of reduction decreased.
Further addition of 6 and then 10 equiv of HMPA to
[Sm{N(SiMe ) } ] showed that the rate decreased and leveled
out at 10 equiv. Subsequent addition of up to 25 equiv of HMPA
provided no further decrease in the rate. The same experiment
was carried out using 2-butanone as the substrate and yielded
similar results (see Supporting Information). The rate data
described in Figure 4 show that an excess of HMPA is required
to bring the rate of reduction of 1-iodobutane to a steady value.
It is likely that addition of 1 equiv of HMPA to [Sm{N(SiMe3)2}2]
3
2 2
8
than 1 or 2 equiv of HMPA can be accommodated. To examine
this issue in more detail, vapor pressure osmometry (VPO) was
utilized to gain insight into the structure (MW) of the active
reductant in solution. Addition of 1 equiv of HMPA to a
standardized solution of [Sm{N(SiMe3)2}2] in THF provided a
MW corresponding to [Sm{N(SiMe3)2}2(HMPA)]. Addition of
3
2 2
(
11) Dorenbos, P. J. Phys.: Condens. Matter 2003, 15, 575.
(
12) Prasad, E.; Knettle, B. W.; Flowers, R. A., II. J. Am. Chem. Soc. 2002,
1
24, 14663.
J. AM. CHEM. SOC.
9
VOL. 126, NO. 22, 2004 6893

A R T I C L E S
Prasad et al.
of HMPA to [Sm{N(SiMe3)2}2] provides a reductant incapable
of interaction with substrates through inner-sphere interactions
with the metal center. Taken together, these results suggest that
addition of HMPA to SmI2 results in a reductant which provides
open coordination sites for substrate along the reaction coor-
dinate and addition of HMPA to [Sm{N(SiMe3)2}2] which
contains a nonlabile ligand leads to a less reactive reductant.
The evidence described above suggests that providing a more
powerful reductant through HMPA addition is not sufficient to
produce a more reactive species. Substrate access to the metal
center is of paramount importance as well. The role of ligand
displacement and the interplay between substrate access and
reducing power of Sm(II) complexes has not been addressed in
previous studies, but reasonable inferences can be drawn from
comparison to related systems. Recently, Hilmersson and co-
workers have developed a cosolvent mixture capable of ac-
celerating reactions of SmI2 utilizing addition of water and
Figure 4. A plot of k versus equivalents of HMPA in Sm{N[Si(CH3)3]2}2/
1
-iodobutane in THF.
Table 1. Rate Constants and Activation Parameters for the
Reduction of 2-Butanone and 1-Iodobutane by [Sm{N(SiMe3)2}2]
and [Sm{N(SiMe3)2}2] Containing 10 equiv of HMPA
b-d
q e
q e
q f
k
M
∆S ,
∆H ,
∆G ,
-1
-1
s-1
cal mol K-1 kcal mol
-1
-1
system
kcal mol
13
amines. This combination of reagents significantly accelerates
the rate of reduction of numerous functional groups and in some
cases reduces substrates more efficiently than SmI2-HMPA.
Although extensive mechanistic studies have not been carried
out on the SmI2/H2O/amine reducing system, there are two
important features of this combination worth mentioning. First,
it is believed that displacement of iodide through precipitation
of ammonium iodide provides the driving force for the reaction.
Second, the redox potential of the SmI2/H2O/amine system in
THF is nearly identical to that of SmI2 alone. Thus, coordina-
tion of a strong electron donor ligand such as HMPA and
production of a more powerful ground-state reductant is not a
precondition for accelerating reactions of SmI2.
2
[
Sm{N(SiMe
3
)
2
}
}
2
2
]- (1.7 ( 0.3) × 10 -43 ( 1 1.4 ( 0.3 14.8 ( 0.6
a
2
-butanone
[
Sm{N(SiMe
3
)
2
]-
16 ( 1
19 ( 1
-17 ( 1 11.6 ( 0.4 17.0 ( 0.5
-47 ( 1 1.7 ( 0.1 16.5 ( 0.3
2
1
-butanone-
0 equiv of HMPA
[
Sm{N(SiMe
3
)
2
}
2
]-
a
1
-iodobutane
[
Sm{N(SiMe
3
)
2
}
2
]- 1.8 ( 0.2
-20 ( 1
11 ( 1
17 ( 1
1
1
-iodobutane-
0 equiv of HMPA
a
b
Data initially reported in ref 7. All rate data are the average of at
least two independent runs. Experimental uncertainties were propagated
14
c
d
through these calculations, and all values are reported as (σ. All rate
e
studies were carried out at 25 °C. Eyring activation parameters were
obtained from ln(kobsh/kT) ) -∆H /RT + ∆S /R. Calculated from ∆Gq
q
q
f
q
q
)
∆H - T∆S .
Conclusions
produces a more powerful reductant still capable of interacting
with substrate while the decrease in rate of reduction of
The high reactivity of SmI2-HMPA toward ketones and alkyl
iodides is a consequence of HMPA-induced iodide displacement
which provides open coordination sites for substrate access to
the Sm reductant. Replacement of iodide with the less labile
-N(SiMe3)2 leads to a less reactive reagent upon addition of
HMPA through the creation of a sterically encumbered reductant
incapable of providing substrate access through ligand exchange.
Although addition of HMPA to SmI2 or [Sm{N(SiMe3)2}2]
provides a thermodynamically more powerful reductant, en-
hanced reducing power of the ground-state reductant is not a
requirement for increased reactivity. This facet of Sm(II)-
HMPA chemistry should be considered in the design of new
approaches to enhance the reactivity of Sm(II)-based reductants.
1
-iodobutane and 2-butanone upon further addition of HMPA
is presumably due to a higher degree of shielding of the Sm by
the bulky HMPA and -N(SiMe3)2 ligands.
To obtain more information on this behavior, the acti-
vation parameters for the reduction of both substrates by
[
Sm{N(SiMe3)2}2] containing 10 equiv of HMPA were de-
termined. The results and comparison to reduction by
[
Sm{N(SiMe3)2}2] alone12 are given in Table 1.
Examination of the rate constants shows that, in the reduction
of either 2-butanone or 1-iodobutane, the presence of HMPA
significantly decreases the rate of reduction of both substrates
by an order of magnitude even though [Sm{N(SiMe3)2}2]-
HMPA is a more powerful reductant (based on its thermody-
namic redox potential). The activation parameters are consistent
with less order and a higher degree of bond reorganization in
the activated complex in the presence of HMPA, findings
distinctly different for those determined for reduction of ketones
and alkyl iodides by [Sm(THF)2(HMPA)4]I2 and [Sm(HMPA)6]-
I2. For instance, reduction of 2-butanone and 1-iodobutane by
SmI2 was faster upon the addition of HMPA, and the presence
Acknowledgment. R.A.F. is grateful to the National Science
Foundation (CHE-0413845) for support of this work. We also
thank Drs. Rebecca S. Miller and Sudhadevi Paivallickal for
their useful comments on the manuscript.
Supporting Information Available: General experimental
details, decay traces, and plots of rate data. This material is
available free of charge via the Internet at http://pubs.acs.org.
of the sterically demanding ligand leads to negative entropies
JA049161J
of activation in the range of -40 cal mol-1
K
-1
with relatively
(
13) (a) Dahlen, A.; Hilmersson, G. Tetrahedron Lett. 2002, 43, 7197. (b) Dahlen,
A.; Hilmersson, G. Chem.-Eur. J. 2003, 9, 1123. (c) Dahlen, A.; Hilmersson,
G. Tetrahedron Lett. 2003, 44, 2661. (d) Dahlen, A.; Petersson, A.;
Hilmersson, G. Org. Biomol. Chem. 2003, 1, 2423. (e) Dahlen, A.;
Sundgren, A.; Lahmann, M.; Oscarson, S.; Hilmersson, G. Org. Lett. 2003,
5, 4085. (f) Kim, M.; Knettle, B. W.; Dahlen, A.; Hilmersson, G.; Flowers,
R. A., II. Tetrahedron 2003, 59, 10397.
low enthalpies of activation. These studies were interpreted to
be consistent with coordination between substrates and SmI2-
HMPA, presumably occurring through displacement of a ligand
by substrate. Conversely, the more positive entropies of activa-
tion and higher enthalpies of activation obtained upon the
addition of HMPA to [Sm{N(SiMe3)2}2] indicate that addition
(14) Dahlen, A.; Hilmersson, G.; Knettle, B. W.; Flowers, R. A., II. J. Org.
Chem. 2003, 68, 4870.
6894 J. AM. CHEM. SOC.
9
VOL. 126, NO. 22, 2004