Acyl palladium species in synthesis: Single-step synthesis of α,β-unsaturated ketones from acid chlorides

Article abstract of DOI:10.1039/b616582f

Conditions are reported for the facile, one-pot synthesis of α,β-unsaturated ketones via the palladium-catalysed cross-coupling of acyl chlorides with hydrozirconated acetylenes, and its use in the 2-step synthesis of d-5-O-benzyl deoxyxylulose. The Royal Society of Chemistry.

Full text of DOI:10.1039/b616582f

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/obc | Organic & Biomolecular Chemistry
Acyl palladium species in synthesis: single-step synthesis of a,b-unsaturated
ketones from acid chlorides†
Russell J. Cox* and Andrew S. Evitt
Received 13th November 2006, Accepted 24th November 2006
First published as an Advance Article on the web 5th December 2006
DOI: 10.1039/b616582f
Conditions are reported for the facile, one-pot synthesis of
a,b-unsaturated ketones via the palladium-catalysed cross-
coupling of acyl chlorides with hydrozirconated acetylenes,
and its use in the 2-step synthesis of D-5-O-benzyl deoxy-
xylulose.
The oxidative addition of palladium(0) complexes and acyl chlo-
rides leads to the formation of acyl palladium chloride species.1
These have been crystallised for X-ray structure elucidation,2
but have found limited use in catalytic cross-coupling reactions.
However, we,3 and others,4–7 have recently shown that acyl
Scheme 1 Regio- and stereo-selectivity observed in hydrozirconation.
palladium complexes can be useful synthetic tools. For example,
we have recently shown that acyl chlorides can be coupled with
through transmetallation of the hindered vinyl zirconocene to zinc.
terminal acetylenes, in a palladium-dependent reaction, to form
The reaction then proceeds in high yield.11
acetylenic ketones in a single step.3
The use of acyl halides in reactions with zirconium reagents
Cross-coupling reactions using acyl chlorides provide useful
has received less attention. It is known that catalysis is not
products which are difficult to form by more traditional methods.
required for the reaction of alkyl zirconocene complexes with
acyl chlorides as long as steric bulk is minimal. Vinyl zirconocene
complexes, however, fail to react under these conditions.12 Wipf
For example, addition of classical nucleophiles such as Grig-
nard or Reformatsky reagents to acyl chlorides usually leads
to multiple additions. Cross-coupling reactions involving acyl
chlorides, however, can be difficult to perform because of their
has shown that CuI·SMe2 catalysis can improve the yield of
products formed by the reaction of alkyl zirconium species with
incompatibility with base. This is especially relevant where the
acyl chlorides. Wipf, and more recently Huang, have also reported
nucleophilic component is generated in a base-dependent reaction.
the application of this catalyst to the reaction of vinyl zirconocenes
We have encountered this drawback in modified Sonogashira
with acyl chlorides, however few examples have been explored.13
reactions using acyl chloride substrates, and have previously
Both Negishi10 and Lipshutz14 have reported the transmetallation
reported conditions for improving yields through reducing the
of vinyl zirconium species to aluminium, prior to coupling to acyl
equivalents of triethylamine to a stoichiometric level.3
chlorides, but the reaction of vinyl zirconium reagents with acyl
Compatibility of cross-coupling reactions toward acyl chloride
palladium reagents has not been reported.
electrophiles would be significantly improved if the nucleophilic
We investigated the possibility of reacting vinyl zirconium
component was generated by a base-free technique. Reductive
species with acyl palladium species under catalytic conditions.
hydrometallation is one such technique, and when applied to
An initial reaction was conducted by first generating the vinyl
acetylenes generates nucleophilic vinyl–metal complexes. Of the
hydrometallation techniques available, hydrozirconation using
zirconocene complex 2,9 and then adding p-nitrobenzoyl chloride
and 5 mol% Pd(PPh3)2Cl2 (Scheme 2). The reaction yielded the
desired product 3, which was isolated as a single isomer in 79%
yield. A control reaction lacking Pd(PPh3)2Cl2 yielded no product,
Schwartz reagent 1 (zirconocene chloride hydride) is reported to
be the most chemo-, stereo- and regio-selective (Scheme 1).8
Vinyl zirconocene species have previously been reported as
supporting our hypothesis that acyl palladium species are key
nucleophiles in cross-coupling reactions. For example, Negishi
intermediates.
has demonstrated that vinyl zirconocene reagents react with
Two byproducts were observed (Scheme 2). The first of these,
aryl and vinyl halide electrophiles under palladium or nickel
diene 4, presumably occurs as a byproduct of activation of
catalysis.9,10 Direct cross-coupling is possible when using relatively
less hindered vinyl groups. However, more hindered substrates
the Pd(PPh3)2Cl2 precatalyst. The palladium chloride ligands
probably transmetallate with two equivalents of the vinyl zir-
react in low yield. Negishi showed that low yields can be overcome
conocene, and reductive elimination would then yield 4 and the
active catalyst, Pd0(PPh3)2 (Scheme 3). An additional 10 mol% of
School of Chemistry, University of Bristol, Cantock’s Close, Bristol, BS8
1TS, UK. E-mail: r.j.cox@bris.ac.uk; Fax: +44 (0)117 9298611
† Electronic supplementary information (ESI) available: Experimental
procedures, data for all synthetic compounds, and crystal structure data
acetylene and Schwartz reagent were used in further reactions to
compensate for this loss. The second byproduct 5 is the result of
addition of vinyl zirconocene to the target enone. Formation of 5
for compound 22. See DOI: 10.1039/b616582f
reduced the overall reaction yield but not by a significant amount.
This journal is
The Royal Society of Chemistry 2007
Org. Biomol. Chem., 2007, 5, 229–232 | 229
©

View Article Online
Table 1 Coupling reactions of phenylacetylene
Entry
Rꢀ
Product
Conditionsa
Yield (%)b
1
2
3
4
5
6
7
8
p-NO2Ph
Ph
3
a
a
a
a
a
b
a
a
79
79
75
73
91
90
88
67
6
p-BrPh
p-MeOPh
Cyclohexyl
tBu
7
8
9
10
11
12
nPr
Me
a Reagents and conditions: (a) i. phenylacetylene, zirconocene hydride
chloride, toluene, 1 h, 40 ◦C; ii. 5 mol% Pd(PPh3)2Cl2, acyl chloride, 3 h, r.t.;
(b) same as (a), but with stage ii being conducted for 24 h at r.t. b Isolated
purified yield.
Scheme 2 Reaction between acyl chloride and vinyl zirconocene.
Table 2 Coupling reactions of p-nitrobenzoyl chloride
Entry
R
Rꢀ
Product Conditionsa
Yield (%)b
1
2
3
4
5
6
7
8
9
10
11
12
13
14
H
H
H
H
H
H
H
H
H
H
Et
Me
Me
Me
(CH2)3Me
Cyclohexyl
tBu
13
14
15
15
16
17
18
19
20
21
22
23
23
24
a
a
a
b
a
a
a
a
a
a
a
a
c
89
71
14
63
43
79
65
27
0
tBu
SiMe3
Bn
CH2OBn
CH2OAc
Ac
CH2N(Me)2
Et
0
73
40
88
16
nPr
nPr
Ph
c
a Reagents and conditions: (a) i. acetylene, zirconocene hydride chloride,
toluene, 1 h, 40 ◦C; ii. 5 mol% Pd(PPh3)2Cl2, p-nitrobenzoyl chloride, 3 h,
r.t.; (b) same as (a), but using 15 mol% Pd(PPh3)2Cl2; (c) same as (a), but
with stage ii being conducted for 24 h at 40 ◦C. b Isolated purified yield.
Scheme 3 Proposed activation of the precatalyst and catalytic cycle.
Yields of the bis-addition product were reduced further in cases
where the substrate lacked a p-nitro substituent.
The observation that no reaction occurs between the vinyl zir-
conium species and the acyl chlorides in the absence of palladium
suggests that the reaction proceeds by the generally accepted
catalytic mechanism of oxidative addition, transmetallation, and
reductive elimination (Scheme 3). The stereochemistry achieved in
the initial hydrozirconation of the acetylene appears to be carried
through the catalytic cycle.
The reaction conditions were then applied to a wider range
of starting materials (Table 1 and Table 2). Typical reaction
conditions employed were addition of Schwartz reagent (1.12 eq.)
to the acetylene (1.10 eq.) in anhydrous toluene, under nitrogen
and protected from light. The mixture was warmed to 40 ◦C with
stirring for 1 h. The reaction was cooled to room temperature, and
then acyl chloride (1.0 eq.) and Pd(PPh3)2Cl2 (5 mol%) were added.
Completion of the reaction was often achieved within 3 h (TLC,
1H NMR). The use of more hindered reagents required heating or
extended reaction times. The completion of reaction was often, but
not always, marked by the precipitation of zirconocene dichloride.
Substituted benzoyl chlorides (Table 1, entries 2–4), and alka-
noyl chlorides (Table 1, entries 5–8) were successfully used as
the electrophilic component. p-Bromobenzoyl chloride (Table 1,
entry 3) reacted in good yield without any evidence of the aryl
bromide taking part in cross-coupling. No significant difference
was observed between strongly electron-donating (Table 1, entry 4)
or electron-withdrawing (Table 1, entry 1) substituents on the
benzoyl chloride in terms of yield or required reaction conditions.
The range of acetylenes tested included both terminal (Table 2,
entries 1–10) and internal examples (Table 2, entries 11–14).
230 | Org. Biomol. Chem., 2007, 5, 229–232
This journal is
The Royal Society of Chemistry 2007
©

View Article Online
The linear aliphatic acetylene (Table 2, entry 1) and cyclohexyl
acetylene (Table 2, entry 2) reacted in high yield under standard
reaction conditions; however, the branched tert-butyl acetylene
(Table 2, entry 3) reacted in low yield unless the catalyst loading
was increased to 15 mol%. TMS acetylene (Table 2, entry 5) also
reacted in lower yield.
The reaction of the benzyl ether (Table 2, entry 7) proceeded in
high yield, but alternative oxygen functionalities reacted in lower
yields. For example, reaction of the ester (Table 2, entry 8) gave
a low yield, and reaction of the ketone (Table 2, entry 9) failed
entirely. Reaction of the tertiary amine (Table 2, entry 10) also
failed, which may indicate an incompatibility with base.
The reaction of 3-hexyne (Table 2, entry 11) gave the expected
target material in high yield. However, the reaction of 2-hexyne
proceeded in low yield, but allowing the second stage of the
reaction to stir for a period of 24 h at 40 ◦C improved the
yield significantly (Table 2, entry 13). Unfortunately, extended
reaction times did not improve the yield of the phenyl methyl
acetylene reaction (Table 2, entry 14).
Negishi observed that although hindered vinyl zirconocenes
could be formed by the reaction of Schwartz reagent with substi-
tuted acetylenes, they appeared not to undergo transmetallation
to the palladium(II) intermediates derived from aryl halides.9 For
example, the reaction of the vinyl zirconocene derived from 3-
hexyne, in a palladium-catalysed cross-coupling reaction with
phenyl iodide, did not proceed until a stoichiometric zinc species
was introduced. This contrasts with our results, where the reaction
of the vinyl zirconocene derived from 3-hexyne with the p-
nitrobenzoyl chloride (Table 2, entry 11) proceeds in high yield
without the requirement for a zinc mediator.
pathway to terpenes and because of DXP’s role as a biological
precursor of pyridoxal phosphate and thiamine pyrophosphate.16
For example, Welzel and co-workers have described a 5-step
synthesis of 25 from ethyl bromoacetate,17 while Fechter and
co-workers have described an alternative 5-step synthesis from
acrolein.18 We recently described a 4-step sequence to the related
protected DXP.19 An important feature of all these routes is their
ability to incorporate isotopic labels at strategic positions.
We reacted propargyl benzyl ether 26 with Schwartz reagent,
and the reaction mixture was then treated with acetyl chloride
and Pd(PPh3)2Cl2 to yield the expected unsaturated ketone 27 in
72% yield. The olefin 27 was treated with stoichiometric OsO4
in the presence of (DHQD)2PHAL to afford the syn-diol 25 in
59% yield and 93% ee (Scheme 4). These reactions are eminently
suitable for the incorporation of isotopic labels. For example,
acetyl chloride can be synthesised from acetic acid, which is
2
available with all combinations of 13C and H labels. Propargyl
benzyl ether can be simply deuterated at the acetylenic position
by base treatment followed by D2O quench, and deuterium could
also be incorporated stereospecifically at C-4 of 25 by the use
of deuterated Schwartz reagent. These reactions are currently
underway in our laboratories, and the use of isotopically labelled
deoxyxylulose as a probe to investigate the mechanism of non-
mevalonate pathway enzymes will be reported in due course.
Analysis of the enone products from all reactions described in
Tables 1 and 2 showed that only single isomers were obtained. 1H-
NMR confirmed the expected regiochemistry, and NOE studies
suggested the correct stereochemistry. The stereochemistry of
enone 22 was confirmed by X-ray crystallography (Fig. 1).
Scheme 4 Synthesis of D-5-O-benzyl deoxyxylulose. Reagents and con-
ditions: i) Schwartz reagent, toluene, then acetyl chloride, Pd(PPh3)2Cl2;
ii) OsO4, CH2Cl2, (DHQD)2PHAL, then MeOH, HCl.
Notes and references
1 P. Fitton, M. P. Johnson and J. E. McKeon, Chem. Commun. (London),
1968, 6.
2 E. D. Dobrzynski and R. E. Angelici, Inorg. Chem., 1975, 14, 59–63.
3 R. J. Cox, D. J. Ritson, T. A. Dane, J. Berge, J. P. H. Charmant and A.
Kantacha, Chem. Commun., 2005, 1037.
4 J. K. Stille, Angew. Chem., Int. Ed. Engl., 1986, 25, 508.
5 H. Chen and M. Deng, Org. Lett., 2000, 2, 1649.
6 E. Negishi, V. Bagheri, S. Chatterjee, F. Luo, J. A. Miller and A. T.
Stoll, Tetrahedron Lett., 1983, 24, 5181.
7 V. F. C. Cardellicchio, G. Marchese and L. Ronzini, Tetrahedron Lett.,
1987, 28, 2053.
8 J. Schwartz and J. A. Labinger, Angew. Chem., Int. Ed. Engl., 1976, 15,
333.
Fig. 1 X-Ray structure of enone 22 (ORTEP probability 50%).15
9 E. I. Negishi and D. E. Van Horn, J. Am. Chem. Soc., 1977, 99,
3168.
10 F. Zeng and E. Negishi, Org. Lett., 2002, 4, 703–706.
11 E. I. Negishi, N. Okukado, A. O. King, D. E. Van Horn and B. I.
Spiegel, J. Am. Chem. Soc., 1978, 100, 2254.
12 D. W. Hart and J. Schwartz, J. Am. Chem. Soc., 1974, 96, 8115; D. B.
Carr and J. Schwartz, J. Am. Chem. Soc., 1977, 99, 638.
13 P. Wipf and W. Xu, Synlett, 1992, 718; X. Meihua and X. Huang,
J. Chem. Res., Synop, 2003, 138–139.
The ability to couple simple commercially available precursors,
to form useful synthetic intermediates that are difficult to syn-
thesise by other methods, is a key advantage of the reaction.
We applied the reaction to the 2-step synthesis of protected
deoxyxylulose 25. Deoxyxylulose (DX), and its 5-O-phosphate
(DXP), have formed the target for several syntheses because of
the central position of DXP in the non-mevalonate biosynthetic
14 B. H. Lipshutz, B. Ullmann, C. Lindsley, S. Pecchi, D. J. Buzard and
D. Dickson, J. Org. Chem., 1998, 63, 6092–6093.
This journal is
The Royal Society of Chemistry 2007
Org. Biomol. Chem., 2007, 5, 229–232 | 231
©

View Article Online
15 Crystal structure data for compound 22. C13H15NO3, M = 233.26,
16 O. Meyer, J.-F. Hoeffler, C. Grosdemange-Billiard and M. Rohmer,
Tetrahedron, 2004, 60, 12153.
17 U. Schuricht, K. Endler, L. Hennig, M. Findeisen and P. Welzel,
J. Prakt. Chem., 2000, 342, 761.
18 M. H. Fechter, R. Gaisberger and H. Griengl, J. Carbohydr. Chem.,
2001, 20, 833.
19 R. J. Cox, A. de Andres-Gomez and C. R. A. Godfrey, Org. Biomol.
Chem., 2003, 1, 3173.
˚
triclinic, a = 8.2761(17), b = 10.042(2), c = 15.189(3) A, a = 74.44(3),
◦
3
˚
b = 81.70(3), c = 82.47(3) , U = 1197.7(5) A , T = 100(2) K, space
group P1, Z = 4, l(Mo-Ka) = 0.092 mm−1, 13 744 reflections measured,
¯
5466 unique (Rint = 0.0676) which were used in all calculations. The
final R1 and wR2 were 0.1222 and 0.2105 (I > 2rI). CCDC reference
number 622616. For crystallographic data in CIF or other electronic
format see DOI: 10.1039/b616582f.
232 | Org. Biomol. Chem., 2007, 5, 229–232
This journal is
The Royal Society of Chemistry 2007
©