Horner-Wadsworth-emmons reactions in THF: Effect of hydroperoxide species

Article abstract of DOI:10.1055/s-0033-1339200

Horner-Wadsworth-Emmons (HWE) reactions routinely employ tetrahydrofuran (THF) as the reaction solvent. In this paper we show that THF adducts (derived from THF-hydroperoxide species) of HWE phosphonate ester compounds are formed under microwave irradiation (i.e., under pressure), or in the presence of a reductant [e.g., P(OEt)₃] in a conventionally heated reaction. Georg Thieme Verlag Stuttgart · New York.

Full text of DOI:10.1055/s-0033-1339200

LETTER
▌1493
letter
Horner–Wadsworth–Emmons Reactions in THF: Effect of Hydroperoxide
Species
THF Hydroperoxide Effects in HWE Reactions
Amanda G. Jarvis, Elizabeth R. Wells, Ian J. S. Fairlamb*
Department of Chemistry, University of York, York YO10 5DD, UK
Fax +44(1904)322516; E-mail: ian.fairlamb@york.ac.uk
Received: 16.04.2013; Accepted after revision: 20.05.2013
tained a tetrahydrofuran ring, indicating 4 as a plausible
Abstract: Horner–Wadsworth–Emmons (HWE) reactions routine-
structure.
ly employ tetrahydrofuran (THF) as the reaction solvent. In this pa-
per we show that THF adducts (derived from THF–hydroperoxide
species) of HWE phosphonate ester compounds are formed under
PPh2
microwave irradiation (i.e., under pressure), or in the presence of a
reductant [e.g., P(OEt)3] in a conventionally heated reaction.
CHO
Ph2P
S
S
O
PPh2
O
Key words: alkenes, phosphorus, peroxides, furan, reduction
3 (2 equiv)
EtO2(O)P
P(O)OEt2
NaOH (4 equiv)
THF–H2O
The Horner–Wadsworth–Emmons (HWE) reaction (also
known as Horner–Wittig) of stabilized phosphonate an-
ions with aldehydes or ketones in the presence of a suit-
able base is widely employed in synthetic chemistry, with
many variants appearing in natural product syntheses.1
The reaction has been optimised over the years, allowing
either Z- or E-alkenes to be obtained with high stereose-
lectivity. We recently identified2 dbathiophos ligand 1 as
a flexible multidentate ligand for CuI. Further optimisa-
tion studies concerning the synthesis of 1 have revealed a
competing reaction path when conducted with microwave
assistance and using THF as the solvent, the findings of
which are detailed herein.
2 (1 equiv)
dbathiophos (1) 84%
reflux, 48 h
Scheme 1 HWE reaction of 2 and 3 to give dbathiophos 1
Compound 1 is readily accessed by conventional heating
methods (reaction of 2 and 3 in the presence of NaOH in
THF–H2O) in a closed system (Schlenk tube) in 84%
yield after 48 hours at reflux (Scheme 1).2a On conducting
the same reaction in a microwave reactor3 it was discov-
ered that the reaction time could be shortened to ca. one
hour, which gave 1 in 67% yield. Increasing the equiva-
lents of 2 from 1.0 to 1.2 resulted in a higher yield (80%).
It was during this optimisation that in one reaction4 the
yield of 1 was lowered dramatically (to 25%), and a sec-
ond unknown product was formed, also in 25% yield. The
1H NMR spectrum of the unknown product is shown in
Figure 1. ESI-MS showed a mass ion of m/z = 433.1380,
which corresponds to C26H26O2PS [MH]+. This suggests
the presence of an oxygen-containing cyclic ring (either
tetrahydrofuran or tetrahydropyran – presumably derived
from the THF solvent vide infra). The 1H NMR spectrum
of the side product showed that one C=C bond had formed
(not shown), but relatively complex aliphatic signals were
observed between δ = 1.2–4.1 ppm. Comparison of related
literature compounds5 indicated that the side product con-
Figure 1 1H NMR spectrum (expanded) of the unexpected side
product 4 in the HWE reaction of 2 and 3 to give 1 (* = toluene;
# = H2O)
The chemical structure of the side product was unambig-
uously confirmed by single-crystal X-ray diffraction (Fig-
ure 2). Interestingly, the crystal packing forces 4 to adopt
a conformation (in the solid state) where the enone moiety
is s-trans. A 2D NOESY spectrum of 4 in CDCl3 shows a
strong NOE between the α-olefin proton and the ortho
proton on the aromatic ring, showing that this conforma-
tion is also adopted in solution.
To fully understand how 4 had formed, a series of reac-
tions were carried out in the absence of aldehyde (Scheme
2). Compound 2 was converted in low yield into bis(tetra-
hydrofuran)acetone (5) in 8% yield. Benzylidene acetone
phosphonate ester (6),6 subjected to same reaction condi-
tions, gave compound 7 in 19% yield. Altering the base
from NaOH to NaOMe did not appreciably alter the yield
of 7 (21%). In the absence of water, however, this latter
reaction did not give 7. Phosphonate ester hydrolysis was
SYNLETT 2013, 24, 1493–1496
Advanced online publication: 26.06.2013
0
9
3
6
-
5
2
1
4
1
4
3
7
-
2
0
9
6
DOI: 10.1055/s-0033-1339200; Art ID: ST-2013-D0344-L
© Georg Thieme Verlag Stuttgart · New York

1494
A. G Jarvis et al.
LETTER
Davies and Pérez both described transition-metal-cata-
lysed carbene reactions,9 whereas Zhang and co-workers
functionalised a variety of ring-containing ethers using a
radical methodology.10 In the reaction described here, it is
not immediately obvious by which mechanism side prod-
uct 4 is being formed, as no intentionally added transition
metals or radical initiators are present. However, the reac-
tion conditions can be considered ‘forcing’, even for a mi-
crowave reaction (1 h at 110 °C). Crucially, compounds
such as 4 were not found in the conventionally heated re-
action above trace levels (>1%). We suspect that the
microwave-heated reaction generates higher pressures
than the conventionally heated reaction.
An intermediate acylcarbene (formed by α-elimination),11
followed by a C–H insertion reaction with THF, could po-
tentially explain the formation of the THF adducts. How-
ever, on the addition of carbene traps, for example,
cyclohexene, norbornene, and 1,2-tetramethylethene, to
either 2 or 6 no cyclopropanation formation was observed
by NMR spectroscopic or MS analysis. Changing the
solvent to 1,4-dioxane gave the same outcome.
Figure 2 X-ray structure of compound 4. Thermal ellipsoids shown
at 50% probability; disorder is observed around the THF moiety.
Selected bond angles (°): O(1)–C(21)–C(20): 119.69(13), C(18)–
C(13)–P(1): 120.25(10)
Tetrahydrofuran-2-hydroperoxide (THF-OOH) was iden-
tified as an alternative intermediate, formed by the reac-
tion of THF with O2 (air) in the presence of light.12 THF-OOH
could either generate a THF radical species, or be reduced
under the forcing microwave conditions to 2-hydroxytet-
rahydrofuran (THF-OH). Whilst no precautions were tak-
en to distil THF or use degassed water in either the HWE
or control reactions (vide supra), regular testing of the lab-
oratory-grade THF using peroxidase test strips13 showed
low but variable quantities of peroxide (typically ca. 1–5
ppm), which quantitatively could not fully account for
amounts of THF adducts observed in the above reactions.
A more quantifiable test for THF-OOH was clearly re-
quired (special precaution: if THF is suspected of contain-
ing THF-OOH it should not be distilled!).
Triethyl phosphite [P(OEt)3] reacts14 smoothly and effi-
ciently with hydroperoxides to yield the triethyl phos-
phate [PO(OEt)3] and the corresponding alcohol, THF-
OH. It was recognised that this would provide a useful
tool for quantification of the total amount of THF-OOH in
solution (monitoring reactions by 31P NMR spectrosco-
py). Upon addition of P(OEt)3 to the HWE reaction of 2
and 3, the formation of PO(OEt)3 (after 1 h) was observed
and quantified by 31P NMR spectroscopy (Table 1).
6 M NaOH (3.3 equiv)
O
O
O
O
O
O
(EtO)2P
P(OEt)2
THF–H2O, MW
110 °C, 1 h
2
5
(8%)
6 M NaOH (3.3 equiv)
THF–H2O, MW
O
O
O
OEt
P
Ph
Ph
OEt
110 °C, 1 h
6
8
O
O
Ph
O
P
as above
7 19%
OH
OH
Scheme 2 Reactions in the absence of aldehyde
considered as a possibility. However, benzylidene ace-
tone phosphonic acid (8)7 did not give 7, ruling it out as an
intermediate.
The unexpected introduction of a THF motif has previous-
ly been observed for other reaction types 8 and their for-
mation attributed to either carbene or radical mechanisms.
Table 1 Addition of P(OEt)3 to the Reaction 2 + 3 → 1 + 4a
P(OEt)3 (equiv)
P(OEt)3/PO(OEt)3 ratio
PO(OEt)3 (equiv formed)
Peroxide concn (mM)b
Yield (%) of 4 (1)
1
1:0.2
1:0.5
0.17
0.17
25
25
19 (33)
17 (22)
0.5
a Monitored by 31P NMR spectroscopy.
b The concentration of peroxide was obtained from relating the ratio of P(OEt)3/PO(OEt)3 by 31P NMR spectroscopic analysis to the equivalents,
and thus mmol of PO(OEt)3 formed, which was considered equal to the amount of THF-OOH in the solvent. Finally, this value was converted
into mM.
Synlett 2013, 24, 1493–1496
© Georg Thieme Verlag Stuttgart · New York

LETTER
THF Hydroperoxide Effects in HWE Reactions
1495
OH
reduction
O2, light
OH
O
MW or P(OEt)3
O
O
O
O
O
OH
ring opening
OH
O
O
OH
O
O
PO(OEt)2
HWE
HO
R
5-exo-trig
O
R
O
R
ring closure
O
Scheme 3 Formation and reduction of THF-OOH in HWE reactions
Regardless of the amount of P(OEt)3 added (0.5 or 1 water) and test rigorously for the presence of peroxides in
equiv) the same amount of PO(OEt)3 (0.17 equiv) was re- laboratory-grade THF. For the HWE reactions detailed
corded, which closely matches the observed yields of 4.15 herein, the conventionally heated reactions showed no
From the amount of PO(OEt)3 formed, the concentration side products containing a THF moiety. However, the em-
of THF-OOH was calculated as 25 mM.16 Finally, reac- ployment of a suitable reductant, for example, P(OEt)3,
tions of 2 and 3 conducted under a rigorous inert atmo- led to formation of the same side products. Thus, a sub-
sphere, using distilled and degassed THF and degassed strate containing a suitable functionality may play the
water (kept in the dark), no side product 4 was observed.
same role.
Using conventional heating, compound 4 was not previ-
ously observed, yet under microwave irradiation it was. It
was therefore hypothesised that the MW irradiation pro-
moted THF-OOH reduction to THF-OH. Therefore,
P(OEt)3 was added to the conventionally heated reaction,
which led to the formation of 4.
Acknowledgment
AGJ was funded by an EPSRC DTA PhD studentship. IJSF thanks
the Royal Society for support (University Research Fellow). We are
very grateful to Prof. G. C. Lloyd-Jones (University of Bristol, UK)
for mechanistic discussions relating to this study.
Finally, an independent reaction of THF-OH with com-
pound 2 in the presence of NaOH in THF–H2O gave bis-
THF product 5 in 56% yield. A competition experiment
(see the Supporting Information) showed that the THF-
OH, which is a masked aldehyde, is more reactive than
aldehyde 3.
Supporting Information for this article is available online at
http://www.thieme-connect.com/ejournals/toc/synlett.SnuIpofoiprmntgirStanoIuifpog
r
t
iornat
References and Notes
(1) Kelly, S. E. In Comprehensive Organic Synthesis; Vol. 1;
Trost, B. M.; Fleming, I., Eds.; Pergamon: Oxford, 1991,
729.
(2) (a) Jarvis, A. G.; Whitwood, A. C.; Fairlamb, I. J. S. Dalton
Trans. 2011, 40, 3695. (b) For a review on the use of dba-Z
ligands in palladium catalysis, see: Fairlamb, I. J. S. Org.
Biomol. Chem. 2008, 6, 3645.
(3) For other reported MW-assisted HWE reactions, see:
(a) Dakdouki, S. C.; Villemin, D.; Bar, N. Eur. J. Org.
Chem. 2010, 333. (b) Rossi, D.; Carnevale, Baraglia. A.;
Serra, M.; Azzolina, O.; Collina, S. Molecules 2010, 15,
5928. Microwave-assisted Wittig reactions (ref. 2c. in THF)
have been described, see: (c) Frattini, S.; Quai, M.; Cereda,
E. Tetrahedron Lett. 2001, 42, 6827. (d) Wu, J.; Wu, H.;
Weia, S.; Dai, W.-M. Tetrahedron Lett. 2004, 45, 4401.
(4) 1,3-Bis(phosphonato)acetone (2, 215 mg, 1.2 equiv, 0.65
mmol) was added to a stirring solution of compound 3 (350
mg, 2 equiv, 1.09 mmol) in THF (3 mL). To this NaOH (87
mg, 4 equiv, 2.17 mmol) dissolved in H2O (0.35 mL) was
added dropwise. The mixture was heated in a microwave for
ca. 1.5 h at 110 °C. After cooling, the solution was washed
with sat. NH4Cl (aq, 5 mL) and extracted with EtOAc (5 × 5
The involvement of THF-OOH in HWE reactions is
shown in a tentative mechanism given in Scheme 3. Once
THF-OOH is formed, a reduction step is required to give
THF-OH. This compound is in equilibrium with the ring-
opened aldehyde (given the basic conditions of the reac-
tion, we suspect that hydroxide deprotonates THF-OH,
with the equilibrium lying to the side of the ring-opened
aldehyde–alkoxide intermediate).
An alternative possibility is that γ-butyrolactone is formed
during the decomposition of THF-OOH.17,18 However, an
alkene reduction step would then be necessary to explain
the formation of compounds such as 4.
In conclusion, the reduction of THF-OOH to give THF-OH
(a masked aldehyde) occurs under microwave-assisted
HWE reaction conditions. THF-OH effectively competes
with the intentionally added aldehyde (carbonyl) reactant.
HWE reactions routinely employ THF as the reaction sol-
vent. To circumvent this problem one should degas the
THF solvent at the very least (and any co-solvent, e.g.,
© Georg Thieme Verlag Stuttgart · New York
Synlett 2013, 24, 1493–1496

1496
A. G Jarvis et al.
LETTER
(8) (a) Mahlokozera, T.; Goods, J. B.; Childs, A. M.;
mL). The organic phases were combined and then dried over
Na2SO3 and filtered. After removing the solvent in vacuo the
product was purified by column chromatography on silica
gel eluting with EtOAc–toluene (5:95, v/v) to afford 1 (88
mg, 25%) and compound 4 as an off-white solid (71 mg,
25%). In a repeat reaction, 4 was isolated in 21% yield (after
1 h heating in the microwave).
Thamattoor, D. M. Org. Lett. 2009, 11, 5095. (b) For a
related reference on unexpected reactions of THF, see:
Wang, X.-S.; Zhou, J.; Yang, K.; Li, Y.-L. Tetrahedron Lett.
2011, 52, 612.
(9) (a) Davies, H. M. L.; Hansen, T.; Churchill, M. R. J. Am.
Chem. Soc. 2000, 122, 3063. (b) Diaz-Requejo, M. M.;
Belderraín, T. R.; Nicasio, C. M.; Trofimenko, S.; Pérez, P.
J. J. Am. Chem. Soc. 2002, 124, 896.
(10) Cheng, K.; Huang, L.; Zhang, Y. Org. Lett. 2009, 11, 2908.
(11) (a) Moody, C. J.; Whitham, G. H. Reactive Intermediates;
Oxford University Press: Oxford, 2006. (b) Anslyn, E. V.;
Dougherty, D. A. Modern Physical Organic Chemistry,
Chap. 1; University Science Books: Sausalito, 2006.
(c) Doyle, M. P.; Duffy, R.; Ratnikov, M.; Zhou, L. Chem.
Rev. 2010, 110, 704.
(12) (a) Hill, D. J. T.; Shao, L. Y.; Pomery, P. J.; Whittaker, A. K.
Polymer 2001, 42, 4791. (b) Parsons, A. F. An Introduction
to Free Radical Chemistry; Blackwell Science: Oxford,
2000. (c) Troisi, L.; Granito, C.; Ronzini, L.; Rosato, F.;
Videtta, V. Tetrahedron Lett. 2010, 51, 5980.
(13) (a) Armarego, W. L. F.; Li, L. C. Purification of Laboratory
Chemicals, 6th ed.; Elsevier: Oxford, 2009, 73. (b) Hill, R.
H.; Finster, D. Laboratory Safety for Chemistry Students;
Wiley: Hoboken, 2010, 5–56.
Compound 4: mp 152–154 °C. 1H NMR (400 MHz, CDCl3):
δ = 8.36 (d, J = 16.5 Hz, 1 H, Hh), 7.87–7.78 (m, 4 H, Ar),
7.73–7.68 (m, 1 H, Hj), 7.57–7.43 (m, 7 H, Hk, p-H and Ar),
7.29 (app. tdd, J = 7.5, 2.5, 1.5 Hz, 1 H, Hl), 7.03 (ddd,
J = 14.5, 8.0, 1.5 Hz, 1 H, Hm), 6.34 (d, J = 16.5 Hz, 1 H,
Hg), 4.09 (app. quin, J = 6.5 Hz, 1 H, Hd), 3.80 (ddd, J = 8.0,
7.0, 6.5 Hz, 1 H, Ha), 3.70–3.64 (m, 1 H, Ha), 2.60 (dd,
J = 15.5, 7.0 Hz, 1 H, He), 2.46 (dd, J = 15.5, 6.0 Hz, 1 H,
He), 2.04–1.94 (m, 1 H, Hc), 1.91–1.77 (m, 2 H, Hb), 1.42–
1.31 (m, 1 H, Hc). 31P NMR (162 MHz, CDCl3): δ = 42.23
(s). 13C NMR (100 MHz, CDCl3): δ = 199.5 (Cf), 143.3 (d,
J = 8 Hz, Hh), 138.5 (d, J = 8 Hz, Ci), 133.6 (d, J = 83 Hz,
ipso-C), 133.1 (d, J = 11 Hz, Cm), 132.60 (d, J = 11 Hz, Ar),
132.56 (d, J = 11 Hz, Ar), 132.2 (d, J = 3 Hz, Ck), 132.2 (d,
J = 84 Hz, ipso-C), 132.1 (d, J = 85 Hz, ipso-C), 132.1 (d,
J = 3 Hz, p-Ar), 132.0 (d, J = 3 Hz, p-Ar), 131. 7 (d, J = 10
Hz, Ci), 129.6 (Cg), 129.4 (d, J = 12 Hz), 128.9 (d, J = 13
Hz, Ar), 128.8 (d, J = 13 Hz, Ar), 128.6 (d, J = 9 Hz, Cj),
75.3 (Cd), 67.9 (Ca), 44.4 (Ce), 31.6 (Cb), 25.7 (Cc). ESI-
HRMS: m/z calcd for C26H26OPS: 433.1386; found:
433.1380 [MH]+. ESI-LRMS: m/z (rel. %): 455 (80) [MNa]+,
433 (100) [MH]+, 401 (4), 301 (4), 236 (7). IR (ATR): ν =
2966 (w), 2861 (w), 1658 (br, m), 1583 (w), 1478 (w), 1457
(w), 1435 (m), 1387 (w), 1311 (w), 1260 (w), 1184 (w), 1162
(m), 1120 (w), 1097 (m), 1059 (m), 1027 (m), 998 (m), 970
(m), 798 (br, m), 755 (m), 709 (s), 690 (s) cm–1. Crystals
suitable for study by X-ray diffraction were grown by slow
evaporation from 1,4-dioxane.
(14) Tsuji, S.; Kondo, M.; Ishiguro, K.; Sawaki, Y. J. Org. Chem.
1993, 58, 5055.
(15) A control reaction of P(OEt)3 and 6 M NaOH in dioxane
showed no evidence of oxidation of P(OEt)3 under the
reaction conditions by 31P NMR spectroscopic analysis.
(16) The concentration of THF-OOH was determined in
‘Suzuki–Miyaura’ reaction mixtures employing aq THF
(10:1) as ≥8.0 mM, see: Butters, M.; Harvey, J. N.; Jover, J.;
Lennox, A. J. J.; Lloyd-Jones, G. C.; Murray, P. M. Angew.
Chem. Int. Ed. 2010, 49, 5156.
(5) (a) Jung, H. H.; Floreancig, P. E. Org. Lett. 2006, 8, 1949.
(b) Gauthier, R.; Axiotis, G. P.; Chastrette, M. J.
Organomet. Chem. 1977, 140, 245.
(6) Kim, D. Y.; Kong, M. S.; Lee, K. J. Chem. Soc., Perkin
Trans. 1 1997, 1361.
(7) (a) Kumar, G. D. K.; Saenz, D.; Lokesh, G. L.; Natarajan, A.
Tetrahedron Lett. 2006, 47, 6281. (b) Xu, Y.; Qian, L.;
Prestwich, G. D. Org. Lett. 2003, 5, 2267. (c) Hawkins, M.
J.; Powell, E. T.; Leo, G. C.; Gauthier, D. A.; Greco, M. N.;
Maryanoff, B. Org. Lett. 2006, 8, 3429.
(17) Robertson, A. Nature (London) 1948, 162, 153.
(18) Due to health and safety restrictions we have been unable to
run direct control reactions with THF-OOH. The isolation
and purification of THF-OOH has been previously
attempted, see: (a) Nikishin, G. I.; Glukhovtsev, V. G.;
Peikova, M. A.; Ignatenko, A. V. Izv. Akad. Nauk SSSR Ser.
Khim. 1971, 10, 2323. (b) We do not recommend
purification of THF-OOH by distillation – this could be
extremely hazardous and dangerous!
Synlett 2013, 24, 1493–1496
© Georg Thieme Verlag Stuttgart · New York

Copyright of Synlett is the property of Georg Thieme Verlag Stuttgart and its content may not
be copied or emailed to multiple sites or posted to a listserv without the copyright holder's
express written permission. However, users may print, download, or email articles for
individual use.