Concerted and stepwise Grignard additions, probed with a chiral Grignard reagent

Article abstract of DOI:10.1039/b009678o

The Grignard reagent 6, in which the magnesium-bearing carbon atom is the sole stereogenic centre has been added to CO₂, PhNCO, PhNCS and certain aldehydes with full retention of configuration. Reaction with benzophenone, electron-deficient aldehydes and several allyl halides proceeded with partial or complete racemization. The findings are discussed with respect to a dichotomy between concerted polar and stepwise SET reaction pathways.

Full text of DOI:10.1039/b009678o

Concerted and stepwise Grignard additions, probed with a chiral Grignard
reagent
Reinhard W. Hoffmann* and Bettina Hölzer
Fachbereich Chemie, Philipps-Universität Marburg, Hans-Meerwein-Strasse, D-35032 Marburg, Germany.
E-mail: rwho@chemie.uni-marburg.de; Fax: +49 6421 2828917; Tel: +49 6421 2825571
Received (in Cambridge, UK) 4th December 2000, Accepted 30th January 2001
First published as an Advance Article on the web 20th February 2001
The Grignard reagent 6, in which the magnesium-bearing
carbon atom is the sole stereogenic centre has been added to
CO2, PhNCO, PhNCS and certain aldehydes with full
retention of configuration. Reaction with benzophenone,
electron-deficient aldehydes and several allyl halides pro-
ceeded with partial or complete racemization. The findings
are discussed with respect to a dichotomy between concerted
polar and stepwise SET reaction pathways.
of conversion (k3) of R· into RA· is similar to or larger than the
rate of collapse (k2) of the postulated radical pair 3. Radical
rearrangements of known k3 (radical clocks) have been studied
in this context.5 It is clear, that an ultrafast radical reorganiza-
tion reaction would enlarge the scope of this approach.10 This
would hold e.g. for the conversion of a chiral carbon radical R·
to its enantiomer RA·. The barrier to the inversion of the tert-
butyl radical has been experimentally bracketed to be < 0.5 kcal
mol21 11
For most practical purposes alkyl radicals can be
.
Grignard reagents are among the oldest organometallic reagents
known.1 Their chemistry has evolved as the prototype of polar
organometallic compounds. Yet, despite the enormous body of
polar addition reactions recorded,2 the mechanism of these
additions cannot be considered as settled.3 Most textbooks
describe the addition of Grignard reagents to aldehydes—being
representative of carbonyl compounds—as a simple addition,
cf. 1. Evidence has however been provided4 that it is a Grignard
dimer (halogen-bridged or alkyl-bridged) that enters into the
reaction with the carbonyl group, cf. 2.
considered as being planar, that is prochiral. Therefore, the
stereochemical probes such as 412,13 and 514 have been used to
probe the mechanism of Grignard additions.15,16
Yet in the case of 4 and 5 it is a moot point, to what extent the
stereochemical outcome is influenced by the presence of the
additional stereogenic centres (only one case of epimerisation13
has been so far observed). The ideal probe would be a Grignard
reagent such as 6, in which the magnesium-bearing carbon atom
is the sole stereogenic centre. We have recently described an
access to such a species of ca. 90% ee by asymmetric
synthesis.17 We report here on the use of this reagent as a
mechanistic probe in Grignard additions to carbonyl com-
pounds and in Grignard-substitution reactions.
But still, this says little about the nature of the C–C bond
forming step. At least in some cases it has been shown that
electron transfer precedes the C–C bond formation, especially
in addition reactions to carbonyl compounds with a low
reduction potential, such as benzophenone.5,6 While it is
tempting to formulate5,7,8 all polar additions as being initiated
by an electron-transfer step, there is at the moment no
meaningful way to address the question to what extent electron
motion precedes nuclear motion in the formation of the new C–
C bond. Rather we have to be content with the heuristic
approach that a two-step process can be considered as
established, if it is possible to prove the existence of an
intermediate (likely the radical R· or radical pair 3).9
The reagent 6 is generated in ca. 90% ee from the
enantiomerically and diastereomerically pure sulfoxide 7.17
Due to the mode of generation, the solution of 6 contains ca. 2
equiv. of EtMgCl, one equiv. of p-Cl-C6H4-MgCl and one
equiv. of diethyl sulfoxide. The reagent 6 is configurationally
stable in this cocktail in THF solution up to 230 °C;
racemization proceeds with t1/2 = 5 h at 210 °C.
Therefore those polar additions can be investigated that
proceed readily at 230 °C or below. This holds e.g. for the
addition of 6 to CO2, PhNCO, or PhNCS which provides the
adducts 8, 9, and 10. The same level of enantiomeric purity of
these adducts suggests that the value of 90 ± 2% ee represents
the enantiomeric purity of 6 and that the addition proceeds
without racemization (Table 1).
This is usually done by diverting the radical R· to give
another radical RA·, a process that becomes manifest if the rate
The absolute configuration of compounds 818 and 1017 is
known, that of 9 has been established by chemical correlation
DOI: 10.1039/b009678o
Chem. Commun., 2001, 491–492
This journal is © The Royal Society of Chemistry 2001
491

View Article Online
Table 1 Trapping of the Grignard reagent 6 (ca. 90% ee) with various
electrophiles
were allowed to react with 6. We tend to interprete this as being
caused by a competition between a polar SN2-reaction and an
SET process. This is in line with the ordering of the reduction
potentials recorded for allyl chloride, bromide and iodide
(21.91; 21.29; 20.23 V vs. Hg).20 The SET process should be
faster with ethyl a-bromomethylacrylate and indeed, this gave
rise to 79% of racemic product, 13 (R = COOEt).
Thus, with the chiral Grignard reagent 6 it was possible to
probe the mechanism of Grignard addition to carbonyl com-
pounds and Grignard substitution reactions with respect to the
competition between polar concerted and stepwise SET path-
ways.
We are grateful to the Deutsche Forschungsgemeinschaft
(SFB 260 and Graduiertenkolleg Metallorganische Chemie) as
well as the Fonds der Chemischen Industrie for support of this
study. We thank Dr O. Knopff for preliminary experiments in
the racemic series.
Configura-
Electrophile Product(s)
tiona
Yield (%) ee (%)
CO2
HOOC-CH(Et)Bn 8
S
S
S
n.d.
80
60
56
41
92
89
91
PhNCO
PhNCS
ArCHOb
PhNHCO-CH(Et)Bn 9
PhNHCS-CH(Et)Bn 10
ArCHOH-CH(Et)Bn
D1: 89
D2: 84
D1: 88
D2: 84
D1: 43
D2: 47
12
PhCHO
PhCHOH-CH(Et)Bn
S
42
45
85
C6F5CHO 11 C6F5CHOH-CH(Et)Bn
Ph2CO 12 Ph2COH-CH(Et)Bn
n.d.
—
a At the former Grignard C-atom. b Ar
determined.
= p-MeO-C6H42. n.d.: not
with compound 8. This establishes that the addition reactions
proceeded with retention of configuration. This is in line with
the finding for the carboxylation of 5.14 As the addition of
formaldehyde to 5 proceeded as well without epimerisation,15
we looked at the addition of 6 to aromatic aldehydes, where the
intervention of electron transfer steps is more likely. Addition of
6 to aldehydes generates two diastereomeric adducts (D1; D2),
which were derivatized with Mosher’s reagent and analysed by
1H-NMR spectroscopy. Both for addition to benzaldehyde and
p-methoxybenzaldehyde the formation of the major diaster-
eomer proceeded without loss in enantiomeric purity. There is a
slight decrease in enantiomeric purity of the minor diaster-
eomer, a fact of uncertain significance. Addition to the more
electron deficient pentafluorobenzaldehyde clearly led to par-
tially racemized adducts. On addition to benzophenone, which
has a reduction potential that is by +0.20 V more positive than
that of benzaldehyde,19 racemization is extensive but not
complete.
The partial racemization observed in the addition to 11 and 12
can be interpreted in terms of a competition between a concerted
polar addition and a SET initiated process. This would imply
that even benzophenone undergoes a polar addition to the extent
of 12%. One could also argue that all of these reactions proceed
by SET5,8 and that rotation of R· within the radical pair 3 is only
in few instances faster than the collapse of the radical pair.
We then turned our attention to the reaction of Grignard
reagent 6 with allylic halides, which proceeds in high yield even
at 290 °C and therefore intuitively suggests an SET process.
While allylation with allyl iodide led indeed to racemic
product 13 (R = H) we were surprised to find sizeable
enantiomeric enrichment when allyl bromide or allyl chloride
Notes and references
1 J. Cologne, Bull. Chem. Soc. Fr., 1950, 17, 910.
2 (a) M. S. Kharash and O. Reinmuth, Grignard Reactions of Nonmetallic
Substances, Prentice Hall, New York, 1954; (b) B. J. Wakefield,
Organomagnesium Methods in Organic Synthesis, Academic Press,
1995.
3 T. Holm and I. Crossland, in Grignard Reagents: New Developments,
ed. H. G. Richey, Jr., J. Wiley & Sons Ltd., New York, 2000, pp.
1–26.
4 (a) C. G. Swain and H. B. Boyles, J. Am. Chem. Soc., 1951, 73, 870; (b)
E. C. Ashby, R. B. Duke and H. M. Neumann, J. Am. Chem. Soc., 1967,
89, 1964.
5 E. C. Ashby, Pure Appl. Chem., 1980, 52, 545.
6 E. C. Ashby, J. Laemmle and H. M. Neumann, Acc. Chem. Res., 1974,
7, 272.
7 C. Walling, J. Am. Chem. Soc., 1988, 110, 6846.
8 H. Yamataka, T. Matsuyama and T. Hanafusa, J. Am. Chem. Soc., 1989,
111, 4912.
9 Other approaches have used e.g. isotope effects or linear free energy
relationships to demonstrate that the addition of certain alkyl Grignard
reagents to benzophenone is not a one-step process: (a) T. Holm and I.
Crossland, Acta Chem. Scand., 1971, 25, 59; (b) T. Holm, Acta Chem.
Scand., 1973, 27, 1552; (c) J. J. Gajewski, W. Bocian, N. J. Harris, L. P.
Olson and J. P. Gajewski, J. Am. Chem. Soc., 1999, 121, 326; and ref.
8.
10 J. M. Tanko and L. E. Brammer, Jr., J. Chem. Soc., Chem. Commun.,
1994, 1165.
11 D. Griller, K. U. Ingold, P. J. Krusic and H. Fischer, J. Am. Chem. Soc.,
1978, 100, 6750.
12 (a) M. Tanaka and I. Ogata, Bull. Chem. Soc. Jpn., 1975, 48, 1094; (b)
H. Schumann, B. C. Wassermann and F. E. Hahn, Organometallics,
1992, 11, 2803.
13 D. Dakternieks, K. Dunn, D. J. Henry, C. H. Schiesser and E. R.
Tiekink, Organometallics, 1999, 18, 3342.
14 F. R. Jensen and K. L. Nakamaye, J. Am. Chem. Soc., 1966, 88,
3437.
15 J. S. Filippo and J. W. Nicoletti, J. Org. Chem., 1977, 42, 1940.
16 D. E. Bergbreiter and O. M. Reichert, J. Organomet. Chem., 1977, 125,
119.
17 R. W. Hoffmann, B. Hölzer, O. Knopff and K. Harms, Angew. Chem.,
2000, 112, 3206; Angew. Chem., Int. Ed., 2000, 39, 3072.
18 W. Kirmse, P. Feyen, W. Gruber and W. Kapmeyer, Chem. Ber., 1975,
108, 1839.
19 H.-W. Buckel and F. Wasgestian, Ber. Bunsen-Ges. Phys. Chem., 1983,
87, 154.
20 M. v. Stackelberg and W. Stracke, Z. Elektrochem. Angew. Phys. Chem.,
1949, 53, 118.
492
Chem. Commun., 2001, 491–492