Phosphine oxide-catalysed chlorination reactions of alcohols under Appel conditions

Article abstract of DOI:10.1039/c002825h

A phosphine oxide-catalysed chlorination reaction of primary and secondary alcohols has been developed. This process represents the first triphenylphosphine oxide-catalysed alcohol chlorination under Appel conditions. The Royal Society of Chemistry 2010.

Full text of DOI:10.1039/c002825h

View Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
Phosphine oxide-catalysed chlorination reactions of alcohols under
Appel conditionsw
Ross M. Denton,* Jie An and Beatrice Adeniran
Received (in Cambridge, UK) 10th February 2010, Accepted 10th March 2010
First published as an Advance Article on the web 26th March 2010
DOI: 10.1039/c002825h
A phosphine oxide-catalysed chlorination reaction of primary
and secondary alcohols has been developed. This process
represents the first triphenylphosphine oxide-catalysed alcohol
chlorination under Appel conditions.
During the last five decades halophosphonium salts have
emerged as versatile reagents for the conversion of aliphatic
alcohols to alkyl halides.1 Whilst chlorophosphonium salts
and other phosphorus(V) halide reagents afford high yields
of halides from alcohols under mild reaction conditions
(e.g. eqn (1)) the concomitant generation of stoichiometric
phosphine oxide by-products impacts severely on the atom
efficiency and large-scale applicability of these reactions.
Furthermore, purification of the product is not always
straightforward.
Scheme 1 Proposed phosphine oxide-catalysed chlorination reaction.
very difficult. Harsh conditions are required for the reduction
of phosphine oxides to phosphines from which the chloro-
phosphonium reagents are conventionally derived. To over-
come these difficulties we sought a method to convert
phosphine oxides into chlorophosphonium salts under mild
conditions e.g. (1 - 3, Scheme 1),5 thereby bypassing the
difficult phosphine oxide reduction with a stoichiometric
reductant.
(1)
(2)
We set out to establish whether oxalyl chloride could be
used as indicated (1 - 3, Scheme 1)5 to transform triphenyl-
phosphine oxide into the required chlorophosphonium reagent
and close the catalytic cycle. We first confirmed that stoichio-
metric chlorotriphenylphosphonium chloride
3 could be
rapidly generated in situ from triphenylphosphine oxide
and oxalyl chloride5 and that this reagent was effective at
chlorinating decanol (eqn (3), entry 1, Table 1).
Whilst many creative strategies have been developed to
remove phosphine oxides from reaction mixtures2 aiding
purification the fundamental problem of phosphine oxide
generation in these and other reactions has remained
unsolved.3 To address this problem we are developing new
catalytic versions of the most important phosphorus-mediated
transformations.4 Herein we report the first triphenyl-
phosphine oxide-catalysed chlorination of alcohols under
Appel conditions (eqn (2) and Scheme 1).
Next we established the extent of the unwanted background
reaction between oxalyl chloride and decanol (entry 2). The
rapid (o10 min) consumption of starting material and
formation of chloroglyoxalate 7 and bisester 8 indicated a
significant side reaction which could compete with chlorina-
tion. However, an initial trial of the catalytic reaction using
30 mol% of triphenylphosphine oxide afforded a promising
46% of the desired chloride (entry 3) along with 7 and 8.
Optimisation experiments were then carried out to suppress
the formation of the esters. Addition of the alcohol to a
solution of oxalylchloride and the catalyst over 2 hours
afforded a 65% isolated yield of 6 (entry 4). A further
modification involving simultaneous addition of both oxalyl
chloride and decanol to a solution of 15 mol% catalyst and
oxalyl chloride over 7 hours afforded an excellent yield of 6
(entry 5). Repetition of this reaction without catalyst (entry 6)
resulted in exclusive formation of chloroglyoxalate 7.
Reducing catalyst loading to 10 mol% had a detrimental effect
on yield (entry 7). Finally, the addition time could be
The driving force for chlorination reactions under Appel
conditions, in which chlorophosphonium salts such as 3
react with alcohols to afford chlorides, is the formation of
the strong PQO bond in the phosphine oxide by-product 1.
However, it is the strength of this bond which makes catalysis
Room C24, School of Chemistry, University Park,
University of Nottingham, Nottingham NG7 2RD, UK.
E-mail: ross.denton@nottingham.ac.uk; Fax: +44 115 9513555;
Tel: +44 115 9514194
w Electronic supplementary information (ESI) available: General
experimental information, spectroscopic data for new compounds
and copies of relevant NMR spectra. See DOI: 10.1039/c002825h
ꢀc
This journal is The Royal Society of Chemistry 2010
Chem. Commun., 2010, 46, 3025–3027 | 3025

View Online
Table 1 Development of a catalytic chlorination reaction
Entry
Ph3PO mol%
Addition protocola/addition time
Product (yield%)
1
2
3
4
5
6
7
8
100
0
30
25
15
0
5 added to (COCl)2 + Ph3PO
5 added to (COCl)2
5 added to (COCl)2 + Ph3PO
5 added to (COCl)2 + Ph3PO over 2 h
5 and (COCl)2 added to (COCl)2 + Ph3PO over 7 hd
5 and (COCl)2 added to (COCl)2 over 7 hd
5 and (COCl)2 added to (COCl)2 + Ph3PO over 7 hd
5 and (COCl)2 added to (COCl)2 + Ph3PO over 5 hd
6 (80)c
7 (78) + 8 (22)b
6 (46) + 7 (44) + 8 (10)b
6 (65)c
6 (83)b
7 (89)b
6 (75)b
6 (73)b
10
18
a
b
Unless otherwise indicated the reactants were added in one portion. Yield determined by 1H NMR spectroscopy using Cl4C2H2 as an internal
c
d
standard. Isolated yield. Chloroform solutions of (COCl)2 and 5 were added simultaneously to a solution of (COCl)2 and Ph3PO in chloroform
over the time indicated; the initial amount of (COCl)2 used was equimolar with the catalyst; 1 eq. of (COCl)2 was used in total.
decreased to 5 hours and a good yield of product obtained
with 18 mol% catalyst (entry 8).
as cyclohexanol were relatively unreactive with respect to
chlorination (entry 9).z
With a protocol for catalytic chlorination in hand we
examined a range of alcohol substrates under the optimal
conditions (eqn (4), Table 2). The new method is effective for
primary and secondary alcohols with yields ranging from
7–87%. The reaction is tolerant of aryl, alkene and alkyne
substituents. The data in Table 2 also indicate the scope of the
reaction at present; sterically more demanding substrates such
The corresponding bromination reaction using oxalyl
bromide is feasible; however, this, and further catalytic
halogenation reactions, awaits optimisation (entry 10).
A further two control experiments with no catalyst were
conducted with more activated substrates (entries 3 and 10).
No chloride products were formed in these reactions.
In a final study we examined the role of the phosphine oxide
in more detail. In particular, we sought to establish whether
glyoxalate esters such as 17 could decompose to chlorides via an
uncatalysed or catalysed process (e.g. 17 or 18 - 19, Scheme 2)
as part of a potential second catalytic cycle (cycle 2, Scheme 2).
To this end we prepared chloroglyoxalate 7 from decanol,
the key intermediate in cycle 2 (R = C9H19), by slow addition
of the latter to oxalyl chloride (Scheme 3).
Table 2 Catalytic chlorination of alcohols
Entry
Alcohol substrate
Product
Yielda (%)
1
2
3
4
5
6
5
83 (68)
80 (42)
0c
The stability of 7 was then investigated in the presence of
triphenylphosphine oxide. Thus, exposure of 7 to one equi-
valent of triphenylphosphine oxide for 16 hours afforded no
chlorinated product. The reaction mixture was then heated at
9
9
10
11
12
73
70
88 (70)
7
8
13
14
67 (52)
69 (54)
9
15
15
64
0c
10
11
cyclohexanol
16
17b
7
12
48 (30)
a
Yield determined by 1H NMR spectroscopy using Cl4C2H2 as an
Scheme 2 Possible catalytic cycles for chlorination reaction.
b
internal standard; isolated yields in brackets. Oxalyl bromide was
used in place of oxalyl chloride in this reaction which afforded
bromodecane. Chloroform solutions of the substrate (1.00 eq.) and
(COCl)2 (0.85 eq.) were added simultaneously to a solution of Ph3PO
(0.15 eq.) and (COCl)2 (0.15 eq.) in chloroform over 7 h at room
c
temperature. The above procedure was used without Ph3PO catalyst.
Scheme 3 Investigation of the reactivity of chloroglyoxalate 7.
ꢀc
This journal is The Royal Society of Chemistry 2010
3026 | Chem. Commun., 2010, 46, 3025–3027

View Online
reflux for a further two hours. Again no conversion of 7 to 6
took place.6 That compound 7 is stable under these conditions
suggests that the catalytic chlorination reactions are
proceeding via the Appel-type pathway (cycle 1).
2 (a) For a recent review, see: S. Dandapani and D. Curran,
Chem.–Eur. J., 2004, 10, 3130; for separation via impurity annihila-
tion, see: (b) A. G. M. Barrett, R. S. Roberts and J. Schroder, Org.
¨
Lett., 2000, 2, 2999; for the use of tagged phosphines and phase-
switching, see: (c) C. D. Smith, I. R. Baxendale, G. K. Tranmer,
M. Baumann, S. C. Smith, R. A. Lewthwaite and S. V. Ley, Org.
Biomol. Chem., 2007, 5, 1562; for the Appel reaction using
polymeric triphenyl phosphine, see: (d) C. R. Harrison, P. Hodge,
B. J. Hunt, E. Khoshdel and G. Richardson, J. Org. Chem., 1983,
48, 3721; (e) H. M. Relles and R. W. Schluenz, J. Am. Chem. Soc.,
1974, 96, 6469.
3 For other recent chlorination reactions of alcohols not mediated by
phosphorus(V) reagents, see: (a) B. D. Kelly and T. H. Lambert,
J. Am. Chem. Soc., 2009, 131, 13930; (b) L. De Luca, G. Giacomelli
and A. Porcheddu, Org. Lett., 2002, 4, 553; (c) M. Yasuda,
S. Yamasaki, Y. Onishi and A. Baba, J. Am. Chem. Soc., 2004,
126, 7186; (d) M. Yasuda, S. Yamasaki, Y. Onishi and A. Baba,
Org. Synth., 2006, 83, 86; (e) L. Gomez, F. Gellibert, A. Wagner and
C. Mioskowski, Tetrahedron Lett., 2000, 41, 6049.
4 Other research groups are also active in developing catalytic
versions of phosphorus-mediated transformations. For the first
phosphine-catalysed Wittig reaction, see: (a) C. J. O’Brien,
J. L. Tellez, Z. S. Nixon, L. J. Kang, A. L. Carter, S. R. Kunkel,
K. C. Przeworski and C. G. Chass, Angew. Chem., Int. Ed., 2009, 48,
6836; for a creative phosphine oxide-catalysed aza-Wittig reaction,
see: (b) S. P. Marsden, A. E. McGonagle and B. McKeever-Abbas,
Org. Lett., 2008, 10, 2589; a phosphine oxide catalysed chlorination
of alcohols using thionyl chloride has been reported, see:
(c) T. Rhode, O. Huttenloch, F. Osswald and K. Wissel, US Pat.
Appl. 0228016 A2, 2008. This process does not proceed via a
chlorophosphonium species; (d) A. Dabee, P. Gauthier and
J. P. Senet, US Pat. 5 672 770, 1994.
In conclusion the first triphenylphosphine oxide-catalysed
Appel-type chlorination reaction has been developed. The
reaction is effective for acyclic primary and secondary alcohols
and generates only two gasses as byproducts. Given the
diverse chemistry associated with halophosphonium salts1c
this study provides a proof-of-concept for a range of redox
neutral phosphine oxide-catalysed transformations driven by
oxalyl halides and active oxalyl esters. The full details of these
ongoing studies will be reported in due course.
This work was financially supported by The University of
Nottingham (New Researcher’s Fund NF5012) and the Royal
Society (Research Grant RG090319).
Notes and references
z General procedure for chlorination reactions in Table 2: to a solution
of triphenylphosphine oxide (42 mg, 0.15 mmol) in CHCl3 (1.5 mL) was
added oxalyl chloride (0.012 mL, 0.142 mmol) and the reaction mixture
stirred for 5 min. The appropriate alcohol (1.00 eq.) and oxalyl chloride
(0.073 mL, 0.863 mmol) as solutions in CHCl3 (1.0 mL) were then
added simultaneously over 7 h via syringe pump. The solvent was
1
removed in vacuo and the yield determined via H NMR spectroscopy
using tetrachloroethane as an internal standard (see ESIw).
1 (a) R. Bohlmann, in Comprehensive Organic Transformations,
ed. R. C. Larock, Wiley-VCH, New York, 2nd edn, 1999,
pp. 689–702; (b) R. Appel, Angew. Chem., Int. Ed. Engl., 1975, 14,
801; (c) Organophosphorus Reagents in Organic Synthesis, ed. J. I. G.
Cadogan, Academic Press, London, 1979, ch. 9; (d) G. W. Brown, in
The Chemistry of Functional Groups, ed. S. Patai, Wiley-VCH,
New York, 1971, p. 593; (e) W. Pluempanupat, O. Chantarasriwong,
P. Taboonpong, D. O. Jang and W. Chavasiri, Tetrahedron Lett., 2007,
48, 223.
5 This reaction has been known for some time and proceeds in less
than one minute at room temperature in CHCl3, see: (a) M. Masaki
and K. Fukui, Chem. Lett., 1977, 151; for a very recent application
of this reaction, see: (b) T. Yano, M. Hoshino, M. Kuroboshi and
H. Tanaka, Synlett, 2010, 810.
6 These results are in agreement with a previous study on the
decomposition of chloroglyoxalates in the presence of pyridine,
see: S. J. Rhoads and R. E. Michel, J. Am. Chem. Soc., 1963, 85,
585.
ꢀc
This journal is The Royal Society of Chemistry 2010
Chem. Commun., 2010, 46, 3025–3027 | 3027