The importance of nitrogen substituents in chiral amino thiol ligands for the asymmetric addition of diethylzinc to aromatic aldehydes

Article abstract of DOI:10.1039/a707937k

A new series of N,S-chelate ligands derived from (S)-valine, which possess the capability of a stereogenic nitrogen donor atom, are catalysts for the addition of diethylzinc to aromatic aldehydes and gave the product secondary alcohols in up to 82% ee.

Full text of DOI:10.1039/a707937k

View Article Online / Journal Homepage / Table of Contents for this issue
The importance of nitrogen substituents in chiral amino thiol ligands for the
asymmetric addition of diethylzinc to aromatic aldehydes
James C. Anderson*† and Michael Harding
Department of Chemistry, University of Sheffield, Sheffield, UK S3 7HF
A new series of N,S-chelate ligands derived from (S)-valine,
which possess the capability of a stereogenic nitrogen donor
atom, are catalysts for the addition of diethylzinc to
aromatic aldehydes and gave the product secondary alcohols
in up to 82% ee.
orientation of this phenyl ring will affect that of the a-phenyl
ring and render the nitrogen atom stereogenic.§ In ligands of
type 4 the nitrogen atom is substituted with a large and small
group. Upon coordination to a metal atom the N-methyl
substituent should occupy the same face as the chiral centre on
the backbone of the chelate, forcing the larger phenyl
substituent to the underface [Fig. 1(b)], thus again rendering the
nitrogen atom stereogenic.¶ Here we report our initial studies
concerning these ligands in the 1,2-addition of diethylzinc to
aromatic aldehydes, in order to assay their enantioselectivity
and catalytic reactivity.
The syntheses of our desired ligands were accomplished in
high yield starting from either (S)-valine methyl ester or
(S)-valinol derived from the reduction of (S)-valine (Scheme
1).5 Diphenylation of (S)-valine methyl ester or (S)-valine was
not possible using palladium catalysed methods recently
reported.6 Instead monophenylation using triphenylbismuth
diacetate, promoted or catalysed by copper diacetate,7 gave 7 in
85% yield. This material could then be phenylated again, under
similar conditions, to give 8 in 69% yield. Reduction of the
diphenylamine 8 with LAH was followed by substitution of the
hydroxy function with sulfur under Mitsunobu-type conditions
The successful use of metal complexes for enantioselective
catalysis is largely dependant upon the structure and electronic
properties of chiral ligands. We set out to develop a series of
new chiral N,S-chelates, derived from amino acids, that would
help us understand the origins of chirality in a variety of
asymmetric catalytic processes. Very recently enantiomerically
pure amino thiols have been shown to catalyse the addition of
diethylzinc to benzaldehyde in high enantiomeric excess. Most
noteworthy are systems derived from ephedrine,1,2 (1R,
2S)-(2)-1,2-diphenyl-2-aminoethanol3 and van Koten’s N,S-
chelated bis{2-[(R)-1-)dimethylamino)ethyl]phenylthiolato}-
zinc complex.4 All of these ligands possess a symmetrical
nitrogen donor atom and we anticipated that a potentially
stereogenic nitrogen donor atom could have positive effects in
this reaction system and other metal catalysed processes. In the
design of our ligands careful attention was paid to the notion
that the chirality inherent to the backbone of the amino acid
could be transmitted closer towards the reaction centre by the
correct choice of substituents on nitrogen. Upon chelation, the
nitrogen atom could become stereogenic and reinforce the
stereoinducing power of the ligand. We have tackled this idea
by preparing ligands 1–5 that all possess a ligating sulfur atom
in place of the hydroxy group from the derived amino acid. We
hoped we could get good catalytic activity as sulfur would have
a high affinity towards most metals useful in catalytic reactions,
and has less tendency to diminish the Lewis acidity of the metal
compared to metal alcoholates.‡ For instance, when ligand 1
and gave ligand
1 in 75% yield. Formation of the
N-phenylformamide by warming 7 with acetic formic anhy-
dride8 in 96% yield was followed by treatment with LAH to
effect simultaneous reduction to the N-methyl substituent and
Pri
H2N
Pri
i
CO2Me
CO2Me
PhNH
7
ii
viii
vi
Pri
Ph2N
Pri
Pri
Pri
CO2Me
CO2Me
PhN
PhNH
OH
SR3
R1NR2
CHO
8
10
9
1 R1 = R2 = Ph, R3 = H
2 R1 = R2 = Bn, R3 = H
3 R1–R2 = –(CH2)5–, R3 = H
4 R1 = Ph, R2 = Me, R3 = H
5 R1 = R3 = Ph, R2 = H
vii
ix
iii–v
5
1
4
Pri
H2N
xi
x
2
3
OH
chelates to a metal atom the nitrogen donor with two phenyl
substituents will possess different orientations of its atomatic
rings [Fig. 1(a)]. The rotational freedom of the b-phenyl ring
will be restricted by the proximity of the chiral centre. The
Scheme 1 Reagents and conditions: i, Ph3Bi(OAc)2 (1.2 equiv.), Cu(OAc)2
(10 mol%), CH2Cl2, room temp., 24 h, 85%; ii, Ph3Bi(OAc)2 (1.5 equiv.)
Cu(OAc)2 (1 equiv.), CH2Cl2, room temp., 14 d, 69%; iii, LAH (4 equiv.),
Et2O–THF, room temp., 1 h; iv, diisopropyl azodicarboxylate (2 equiv.),
Ph3P (2 equiv.), AcSH (2 equiv.), room temp., 77% over two steps; v, LAH
(4 equiv.) Et2O–THF, 0 °C, 5 min, 98%; vi, AcOCHO (2.6 equiv.) HCO2H
(3.2 equiv.), THF, 70 °C, 2.5 h, 96%; vii, LAH (5 equiv.) Et2O, 60 °C, 0.5
h, then (iv) followed by (v), 84% over three steps; viii, LAH (2 equiv.),
Et2O–THF, room temp., 0.5 h, 93%; ix, PhSSPh (3 equiv.), Bu3P (4 equiv.),
THF, 80 °C, sealed tube, 24 h, 75%; x, BnBr (2.2 equiv.), K2CO3 (2.5
equiv.), EtOH, room temp. 24 h, 91%, then (iv) followed by (v), 81% over
two steps; xi, Br(CH2)5Br (2 equiv.), K2CO3 (5 equiv.), EtOH, 60 °C, 48 h,
78%, then (iv) followed by (v), 72% over two steps
(a)
(b)
Pri
Pri
Me
S
N
S
N
M
M
Fig. 1
Chem. Commun., 1998
393

View Article Online
hydroxy function. Substitution by sulfur as before gave ligand 4
in 84% yield over three steps. Treatment of 7 with LAH gave
alcohol 10 in 93% yield which was converted to the phenylthiol
ligand 5 by heating in a sealed tube with diphenyl disulfide and
tri-n-butylphosphine in 75% yield.9
Dibenzylation of (S)-valinol was achieved by treatment with
benzyl bromide and potassium carbonate in 91% yield.
Dialkylation of the amine function of (S)-valinol with 1,5-di-
bromopentane, under similar conditions, proceeded in 78%
yield to form a piperidine. Substitution of the hydroxy function
in each of these compounds by the aforementioned method gave
ligands 2 and 3 in 81 and 72% yield, respectively (Scheme 1).
The enantiohomogeneity of these ligands has been verified by
racemic synthesis followed by NMR and HPLC comparison of
their Mosher derivatives with the enantiomerically pure mate-
rials.
With these ligands in hand we screened their effectiveness as
enantioselective catalysts in the addition of diethylzinc to
aromatic aldehydes. Our initial experiments in this area are
promising, as summarised in Table 1, and show good levels of
enantioinduction in the presence of catalytic amounts (10
mol%) of the ligands 1–5. Reactions were performed in toluene,
at room temperature, for 3–20 h using 2.2 equiv. of di-
ethylzinc.
These preliminary studies indicate that giving the nitrogen
atom the potential to become stereogenic leads to a better
system for enantioselection, as evidenced by the superior
enantioselection of ligands 1 and 4 over 2 and 3 respectively.
The best enantioselection (82% ee, entries 8 and 9) found with
ligand 4 suggests that the donating ability of the nitrogen lone
pair could be a factor in the efficiency of these ligands. This
necessitates the need for further studies to try and separate the
steric and electronic factors that are responsible for enhanced
enantioselection. Ligands having a labile proton on nitrogen
perform very poorly as catalysts in this particular reaction and
along with the desired addition product, benzyl alcohol (38%)
was formed from reduction of benzaldehyde.10 These results
show that non-symmetrical phenyl-substituted nitrogen donor
atoms have a positive effect on the efficiency of these particular
ligand systems. We believe we have a system with which we
can probe the origins of chiral induction by further manipulation
of the steric and electronic properties of these ligands. Further
systematic modifications, studies designed to describe a
transition state model and investigation of other metal catalysed
systems will be reported in due course.
We thank the EPSRC (M. H.), the University of Sheffield,
Pfizer and Zeneca for financial support. The loan of GC
equipment from Dr Varinder K. Aggarwal and the help of Dr
Elfyn Jones with GC is gratefully acknowledged.
Notes and References
† E-mail: j.anderson@sheffield.ac.uk
‡ These original assumptions have been supported by other workers (ref.
11).
§ A similar transmission of chirality occurs in the successful chiral
diphosphine ligand chiraphos in various asymmetric palladium catalysed
reactions (ref. 12).
¶ A similar conformational analysis has been invoked to explain the
excellent enantioselectivities obtained in Diels–Alder reactions using an
enantiomerically pure [N,NA-bis(trifluoromethylsulfonyl)-1,2-diphenyl-
ethane-1,2-diamine]aluminium complex as catalyst (ref. 13).
∑
In each case a positive rotation was obtained, indicating the
(R)-enantiomer (ref. 14).
1 R. P. Hof, M. A. Poelert, N. C. M. W. Peper and R. M. Kellog,
Tetrahedron: Asymmetry, 1994, 5, 31.
2 J. Kang, J. W. Lee and J. I. Kim, J. Chem. Soc., Chem. Commun., 1994,
2009.
Table 1 Diethylzinc additions to aromatic aldehydes catalysed by chelate
ligands 1–5
3 J. Kang, D. S. Kim and J. I. Kim, Synlett, 1994, 842.
4 E. Rijnberg, J. T. B. H. Jastrzebski, M. D. Janssen, J. Boersma and
G. van Koten, Tetrahedron Lett., 1994, 35, 6521.
5 M. J. McKennon, A. I. Meyers, K. Drauz and M. Schwarm, J. Org.
Chem., 1993, 58, 3568.
O
OH
Et
Et2Zn, 1–5 (10 mol%)
toluene, room temp.
Ar
H
Ar
6 A. S. Guram, R. A. Rennels and S. L. Buchwald, Angew. Chem., Int. Ed.
Engl., 1995, 34, 1348; J. Louie and J. F. Hartwig, Tetrahedron Lett.,
1995, 36, 3609; D.Ma and J. Yao, Tetrahedron: Asymmetry, 1996, 7,
3075.
7 D. H. R. Barton, J.-P. Finet and J. Khamsi, Tetrahedron Lett., 1989, 30,
937.
8 S. Krishnamurthy, Tetrahedron Lett., 1982, 23, 3315.
9 R. Siedlecka and J. Swarzewski, Synlett, 1996, 757.
10 B. Marx, E. Henry-Basch and P. Freon, C. R. Acad. Sci., Ser., 1967,
264.
11 J. Kang, J. B. Kim, J. W. Kim and D. Lee, J. Chem. Soc., Perkin Trans.
2, 1997, 189.
Entry
Ligand
Ar
Yield (%)a
Ee (%)b
1
2
3
4
5
6
7
8
9
1
1
1
2
3
3
3
4
4
4
5
Ph
85
88
91
91
78
92
100
80
74
52
62
58
66
65
62
82
82
78
0
o-MeOC6H4
p-MeOC6H4
Ph
Ph
o-MeOC6H4
p-MeOC6H4
Ph
o-MeOC6H4
p-MeOC6H4
Ph
83
90
35c
12 B. Bosnich and M. D. Fryzuch, in Topics in Stereochemistry, ed. G. L.
Geoffroy, Wiley, 1981, vol. 12, p. 119 (see pp. 121–125).
13 E. J. Corey, R. Imwinkelried, S. Pikul and Y. B. Xiang, J. Am. Chem.
Soc., 1989, 111, 5493.
10
11
Isolated yield. Determined by chiral GLC using a ChrompackTM CP-
Cyclodex B column or optical rotation (see note ∑). Reaction time 3
days.
a
b
14 R. H. Pickard and J. Kenyon, J. Chem. Soc., 1914, 1115.
c
Received in Cambridge, UK, 4th November 1997; 7/07937K
394
Chem. Commun., 1998