An unexpected base-induced [1,4]-phospho-Fries rearrangement

Article abstract of DOI:10.1039/c0dt00880j

An unexpected [1,4]-phospho-Fries rearrangement that gives rise to the formation of a O,O,O,O-tetraethyl methylenebis(thiophosphonate) derivative is reported. The regioselectivity of the metallation with n-BuLi or t-BuLi is the key factor that explains ei

Full text of DOI:10.1039/c0dt00880j

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/dalton | Dalton Transactions
An unexpected base-induced [1,4]-phospho-Fries rearrangement
Mathieu Berchel, Jean-Yves Salau¨n, He´le`ne Couthon-Gourve`s, Jean-Pierre Haelters and Paul-Alain Jaffre`s*
Received 20th July 2010, Accepted 23rd September 2010
DOI: 10.1039/c0dt00880j
An unexpected [1,4]-phospho-Fries rearrangement that
gives rise to the formation of
a O,O,O,O-tetraethyl
methylenebis(thiophosphonate) derivative is reported. The
regioselectivity of the metallation with n-BuLi or t-BuLi is the
key factor that explains either the [1,4] or [1,3] rearrangement
observed.
Scheme 1 Synthesis of an arylthiophosphonate via a [1,3]-phospho-Fries
rearrangement.
The phospho-Fries rearrangement, which was first reported
by Melvin,1 consists in the transposition of an aryl phos-
phate into an aryl phosphonate. This reaction requires as an
initial step: the formation of an ortho-metallated intermedi-
ate that, via a [1,3]-rearrangement, produces an aryl phos-
phonate. The ortho-metallation can be achieved by either de-
protonation with LDA or n-BuLi or by halogen–metal ex-
change. The good to excellent yields associated with this
phospho-Fries rearrangement explained its use for the syn-
thesis of o-hydroxyarylphosphonates.2 This rearrangement has
been extended to the synthesis of other aryl-phosphorylated
compounds including 2-hydroxyarylphosphonodiamidates,3 2-
hydroxyarylphosphinoxides,4 2-mercaptoarylphosphonates5 and
2-mercaptoarylphosphonodiamidates6 For the five-membered
pyrrole ring, the introduction of phosphorylated species in
the a-position to the nitrogen atom has been also reported.4
The [1,4]-phospho-Fries rearrangement has been attempted
by Buono and coworkers.7 who synthesized an O-(2,6-
dimethyphenyl)phosphoramide derivative. This compound in
the presence of an alkyllithium did not produce the [1,4]-
rearrangement that would have placed the phosphorus moiety
on one of the methyl groups localized in position 2 and 6 of
the phenyl ring. In that case only by-products resulting from
nucleophilic attack of the base at the phosphorus atom were
observed.7 Recently we explored [1,2]- and [1,3]-rearrangement for
the synthesis of O,O-dialkylarylthiophosphonates. When pyrrole
is used as a substrate, O,O-dialkyl-2-pyrrolthiophosphonates can
be obtained in good yields by a two-step process that makes
use of the [1,2]-phospho-Fries rearrangement.8 This procedure
can be repeated to yield pyrrol-bis-thiophosphonate, which is a
pincer ligand that has been engaged for the synthesis of palladium
and silver coordination complexes.9 Aryl-thiophosphonates can be
also obtained via a [1,3]-phospho-Fries rearrangement that makes
use of phenol derivatives as initial substrates (Scheme 1).10 In that
case the efficiency of the rearrangement is strongly dependent on
the nature of the precursor and of the base used. Indeed, with
an ortho-bromo derivative as a substrate (Scheme 1, X = Br),
the metallation is efficiently achieved by using n-BuLi and the
rearranged product is isolated in high yield. Interestingly even a
simultaneous double rearrangement can be achieved in high yield
(92%) on dibromophenol derivatives.10 On the other hand, when
a non-halogenated substrate is employed (Scheme 1, X = H), the
reaction requires a stronger base (t-BuLi). However a complete
conversion is usually not observed and thiophosphonates are
obtained in modest yields (around 50%).10
With the aim of synthesizing compound 5 via a double
[1,3]-phospho-Fries rearrangement on substrate 2 (Scheme 2),
we first synthesized the bis-thiophosphate 2 by reaction of
2-methylresorcinol 1 with O,O-diethylchlorothiophosphate and
triethylamine following a reported procedure10 (yield 40%)†. Then,
t-BuLi (3 equivalents), which is the base required when non-
halogenated substrates are used, was added to a solution of 2
in THF at –78 ◦C; after warming up the reaction media to
20 ◦C and a hydrolysis step, the crude product was purified
on silica gel to produce compound 3‡ in 29% yield instead of
the expected compound 5. Neither in the crude product nor
in a fraction of chromatography, was compound 5 detected by
Scheme
2 Synthesis of O,O,O,O-tetraethyl(2,6-dihydroxyphenyl)-
Universite´ Europe´enne de Bretagne, Universite´ de Brest, CNRS UMR 6521,
CEMCA, IFR 148 ScInBios, 6 Avenue Le Gorgeu, 29238, Brest, France.
E-mail: pjaffres@univ-brest.fr; Fax: +33 298 017 001; Tel: +33 298 016
153
methylene-dithiophosphoante 3: (i) Cl-P(S)(OEt)2 (2.2 equivalents), NEt3
(2.4 equivalents), DMAP (10 mol%), THF, 20 ◦C; (ii) t-BuLi, THF,
-78 ◦C then water.
11314 | Dalton Trans., 2010, 39, 11314–11316
This journal is
The Royal Society of Chemistry 2010
©

View Article Online
NMR. Spectroscopic data attest to the formation of compound
3. According to H NMR, a characteristic triplet, observed at
presence of t-BuLi. After the first [1,4]-rearrangement affording
II, the mesomeric effects favour the deprotonation on the benzylic
position of this intermediate to induce the formation of III which
then evolves into 3 after hydrolysis. This reaction constitutes
an interesting alternative way to the synthesis and the study of
methylenebis(thiophosphonate), which has been much less studied
than its diphosphonate analogue methylenediphosphonate. 11
With the aim of achieving a double [1,3]-phospho-Fries re-
arrangement in place of the [1,4] previously observed, we used
the dibromobis-thiophosphate 4 (Scheme 4) as substrate with
the objective of changing the regioselectivity of the metallation
and consequently the type of phospho-Fries rearrangement.
Dibromo-bisthiophosphate 4§ was synthesized following a two-
step sequence that starts by the bromination of 2-methylresorcinol
1,12 followed by the introduction of the two thiophosphate func-
tional groups according to a reported procedure.10 The reaction
of this compound with n-BuLi gives rise to the double [1,3]-
rearrangement indicating that a full regiocontrol of the metallation
governs the regioselectivity of the phospho-Fries rearrangement.
After purification compound 5 was isolated in 97% yield.¶
1
5.10 ppm with 3JHP = 26 Hz, is ascribed to the CH of the P–CH–
P moiety. 3 protons are observed in the aromatic resonance area
(6.45, 6.60 and 7.07 ppm). The 31P chemical shift for compound 3 is
84.8 ppm while the signal of the starting compound 2 is observed
at 62.9 ppm. 13C NMR spectroscopy corroborates the previous
characterizations, since a characteristic triplet attributed to the C–
H pattern (1JCP = 102.8 Hz) is observed at 49.5 ppm. Interestingly
the carbon atoms of the aromatic ring generate 6 peaks. The
presence of these 6 signals is explained by the absence of a plane
of symmetry due to the presence of the two thiophosphonate
functional groups on a pro-chiral carbon atom. Moreover the
peak observed at m/z 429.072 by MALDI-TOF spectrometry is
in agreement with the formula C15H27O6P2S2 (M + 1).†
With the aim of improving the yield of formation of compound
3, a screening of the required quantities of t-BuLi was achieved.
The best result was obtained by using 6.2 equivalents of t-BuLi. In
that case a conversion of 58% was observed and compound 3 was
isolated after chromatography in 40% yield. The generalization of
this rearrangement was investigated by using O-(2-methylphenyl)-
O,O-diethylthiophosphate as substrate. Surprisingly in that case
the addition of 2.2 equivalents of t-BuLi on this substrate was
unsuccessful and 90% of the starting compound was recovered.
This result indicates that the [1,4]-phospho-Fries rearrangement
can not be generalized by the application of this experimental
procedure. The formation of compound 3 is explained by a double
[1,4]-phospho-Fries rearrangement which is only observed when
the two thiophosphate groups are present in o-positions of the
methyl group. These groups might favour the proximity of t-
BuLi since the two thiophosphosphate moieties can act as a
pincer towards t-BuLi via Li ◊ ◊ ◊ O interactions. Consequently the
deprotonation on the methyl group might be favoured due to
this pre-organization. As soon as the metallation takes place, the
[1,4]-rearrangement can occur leading to the formation of the
likely intermediate II (Scheme 3). The fact that the protonated
form of this intermediate was never observed either in the crude
media or in the fractions of chromatography indicates that this
intermediate is more reactive than compound 2 itself in the
Spectroscopic data confirm the formation of compound 5. Its 31
P
NMR shows a peak at 78.1 ppm that contrasts with the 31P NMR
resonance for compound 4 (62.5 ppm). According to its 1H NMR
spectrum, the methyl group linked to the aromatic ring is observed
at 2.10 ppm. At 7.70 ppm the signal of the aromatic proton exhibits
a characteristic triplet which is ascribed to its coupling with two
phosphorus atoms (3JHP = 14.9 Hz). In 13C NMR 5 also exhibits
a characteristic doublet of doublet at 107.2 ppm (1JCP = 156.0 Hz,
3JCP = 13.0 Hz) that can be attributed to the sp2 carbon atoms
bound to a phosphorus atom. Two triplets are observed at 114.7
(2JCP = 9.5 Hz) and 132.5 ppm (3JCP = 8 Hz) which are attributed
respectively to the sp2 carbon atom linked to the methyl group
and to the Csp2–H. Finally the signal at 162.4 is ascribed to the
aromatic carbon atoms bearing hydroxyl groups.
Scheme 4 Synthesis of the aryl bis(thiophosphonate) 5 following a double
[1,3]-phospho-Fries rearrangement. (i) Br2, CH3COOH; (ii) Cl-P(S)(OEt)2
(2.2 equivalents), NEt3 (2.4 equivalents), DMAP (10 mol%), THF, 20 ◦C;
(iii) n-BuLi (2.4 equivalents), THF, -78 ◦C then water.
To the best of our knowledge, we report herein the first example
of a [1,4]-phospho-Fries rearrangement which is observed when
the bisthiophosphate 2, synthesized from 2-methylresorcinol, is
treated with t-BuLi. It is suggested that this [1,4]-rearrangement
is due to the presence of the two thiophosphate groups
that might localize t-BuLi in the proximity of the methyl
group thus helping the first metallation. This original reaction
Scheme 3 Proposed mechanism for the formation of compound 3.
This journal is
The Royal Society of Chemistry 2010
Dalton Trans., 2010, 39, 11314–11316 | 11315
©

View Article Online
offers
a
straightforward way to synthesize an O,O,O,O-
¶ Spectroscopic data for compound 5: 1H-NMR (400 MHz, CDCl3): 1.33
3
(t, JHH = 7.1 Hz, 12 H, CH3–CH2–O), 2.10 (s, 3H, CH3–), 4.13 (m, 8H,
tetraalkylmethylenebis(thiophosphonate) derivative. On the other
hand, when the metallation is imposed on the aromatic ring, by
using a metal–halogen exchange procedure, a more classic [1,3]-
phospho-Fries rearrangement is observed leading to the formation
of the arylbis(thiophosphonate) 5. The coordination behaviour of
compounds 3 and 5 will be investigated in our next study.
CH3–CH2–O), 7.70 (t, 3JHP = 14.9 Hz, 1H, C–H), 9.67 (s, 2H, O–H); 31P:
(161.9 MHz, CDCl3): 78.1; 13C NMR (75.5 MHz, CDCl3): 8.12 (t, 4JCP
=
1.5 Hz, Ar–CH3), 15.94 and 16.00 (2d, 3JCP = 4.0 Hz; O–CH2–CH3), 63.11
and 63.15 (2d, 2JCP = 2.7 Hz; O–CH2–CH3), 107.22 (dd, 1JCP = 156.0 Hz,
3JCP = 13.0 Hz, C–P); 114.70 (t, 2JCP = 9.5 Hz, C–H); 132.48 (t, 3JCP = 8 Hz,
2
4
C–CH3); 162.36 (~dd, JCP = 11 Hz, JCP = 3.5 Hz, C–OH); Anal, Calcd
for C15H26O6P2S2: C,42.05; H,6.12; S,14.97, Found: C, 41.70; H,6.32; S,
14.84; IR (ATR): 777 (52); 795 (53); 959 (O–C, 46); 1011 (P–O, 48); 1108
(67); 1388 (79); 1586 (74); 1599 (73); 2981 (86); 3131 (O–H . . . S; 79); Exact
Mass, Calcd for C15H27O6P2S2 (M+1): 429.0724; found: 429.0711; m/z
(MSMS m/z = 429; ESI): 429 (M + 1, 63%); 401 (M + 1 - CH2 CH2;
100%); 383 (31%); 373 (M + 1 - 2 ¥ CH2 CH2; 10%); 355 (34%).
Acknowledgements
We acknowledge CNRS for funding and the “Service Communs
de l’UBO: RMN-RPE et spectrome´trie de masse”.
1 L. S. Melvin, Tetrahedron Lett., 1981, 22, 3375–3376.
2 (a) B. Dhawan and D. Redmore, J. Org. Chem., 1984, 49, 4018–4021;
(b) C. M. Taylor and A. J. Watson, Curr. Org. Chem., 2004, 8, 623–636;
(c) K. P. Jayasundera, A. J. Watson and C. M. Taylor, Tetrahedron Lett.,
2005, 46, 4311–4313.
3 O. Legrand, J. M. Brunel, T. Constantieux and G. Buono, Chem.–
Eur. J., 1998, 4, 1061–1067.
4 T. L. Yeung, K. Y. Chan, R. K. Haynes, I. D. Williams and L. L. Yueng,
Tetrahedron Lett., 2001, 42, 457–460.
5 (a) S. Masson, J. F. Saint-Clair and M. Saquet, Synthesis, 1993, 485–
486; (b) B. F. Bonini, C. Femoni, M. Fochi, M. Gulea, S. Masson and
A. Ricci, Tetrahedron: Asymmetry, 2005, 16, 3003–3010.
6 C. Mauger, M. Vazeux and S. Masson, Tetrahedron Lett., 2004, 45,
3855–3859.
Notes and references
† Spectroscopic data for compound 2: 1H-NMR (400 MHz, CDCl3): 1.36
(t, 3JHH = 7 Hz, 12H, 4 ¥ O–CH2–CH3); 2.25 (s, 3H, C2–CH3); 4.24 (m, 8H,
4 ¥ O–CH2–CH3); 7.12 (s, 2H, C4–H et C6–H); 7.26 (s, 1H, C5–H); 31P-
NMR (161.9 MHz, CDCl3): 62.9; 13C-NMR (75.5 MHz, CDCl3): 10.83
3
2
(s, C2–CH3); 16.16 (d, JCP = 8 Hz, O–CH2–CH3); 65.30 (d, JCP = 6 Hz,
O–CH2–CH3); 117.55 and 117.58 (2d, 3JCP = 2 Hz, C4 and C6); 123.44 (t,
3JCP = 6 Hz, C2); 126.28 (t, 4JCP = 2 Hz, C5); 150.07 (d, 2JCP = 7 Hz, C1 and
C3); IR: 950 (C–O); 1015 (P–O); 1464; 1583 (C C); 2982 (C–H).
‡ Spectroscopic data for compound 3: 1H-NMR (400 MHz, CDCl3): 1.24
et 1.31 (2t, 3JHH = 7 Hz, 12H, 4 ¥ O–CH2–CH3); 4.16 (m, 8H, 4 ¥ O–CH2–
7 O. Legrand, J. M. Brunel and G. Buono, Tetrahedron, 2000, 56, 595–
603.
8 S. Marie, M. Lutz, A. L. Spek, R. J. M. Klein Gebbink, G. van Koten,
2
3
CH3); 5.10 (t, JHP = 26 Hz, 1H, CH–P); 6.45 et 6.60 (2d, JHH = 8 Hz,
2H, 2 ¥ CH); 7.07 (t, 3JHH = 8 Hz, 1H, CH); 7.42 (s, 1H, OH); 31P-NMR
(161.9 MHz, CDCl3): 84.8; 13C-NMR (75.5 MHz, CDCl3): 15.98 (m, O–
CH2–CH3); 49.46 (t, 1JCP = 102.8 Hz, CH–P); 63.80, 63.82, 64.12 and 64.15
(4d, 2JCP = 3.4 Hz, O–CH2–CH3); 106.28 (t, JCP = 5.6 Hz, Car); 107.94 (s,
Car); 111.40 (t, JCP = 2.4 Hz, Car); 129.63 (t, JCP = 2.5 Hz, Car); 154.70 (t,
JCP = 6.9 Hz, Car); 157.02 (t, JCP = 4.5 Hz, Car). MALDI-TOF: m/z calc.
for C15H27O6P2S2 [M + 1]: 429.0724, Found: 429.072.
N. Kervarec, F. Michaud, J. Y. Salaun and P. A. Jaffres, J. Organomet.
¨
`
Chem., 2009, 694, 4001–4007.
9 A. Fraix, M. Lutz, A. L. Spek, R. J. M. Klein Gebbink, G. van Koten,
J. Y. Salaun and P. A. Jaffres, Dalton Trans., 2010, 39, 2942–2946.
¨
`
10 T. Dieng, A. Fraix, J. Y. Salaun, I. Dez, R. J. M. Klein Gebbink, G. van
¨
Koten and P. A. Jaffres, Synlett, 2008, 3121–3124.
`
11 (a) J. P. Haelters, H. Couthon-Gourves, A. Le Goff, G. Simon, B.
`
§ Spectroscopic data for compound 4: 1H-NMR (400 MHz, CDCl3): 1.36
`
Corbel and P. A. Jaffres, Tetrahedron, 2008, 64, 6537–6543; (b) E.
3
(t, JHH = 7.1 Hz, 12 H, CH3–CH2–O), 2.41 (s, 3H, CH3–), 4.27 (m, 8H,
´
Migianu-Griffoni, C. Mbemba, R. Burgada, D. Lecercle, F. Taran
and M. Lecouvey, Tetrahedron, 2009, 65, 1517–1523; (c) G. Sturtz,
H. Couthon, O. Fabulet, M. Mian and S. Rosini, Eur. J. Med. Chem.,
1993, 28, 899–903.
CH3–CH2–O), 7.68 (s, 1H, C–H); 31P-NMR: (161.9 MHz, CDCl3): 62.5;
13C (75.5 MHz, CDCl3): 14.16 (s, Ar–CH3), 15.95 (d, JCP = 8.0 Hz; O–
3
CH2–CH3), 65.44 (m,; O–CH2–CH3), 113.39 and 113.46 (2d, 3JCP = 4.0 Hz,
C–Br); 129.15 (t, 3JCP = 3.7 Hz, C–CH3); 133.42 (s, C–H); 147.68 (m, Car–
O–P)
12 M. Vetrichelvan and S. Valiyaveettil, Chem.–Eur. J., 2005, 11, 5889–
5898.
11316 | Dalton Trans., 2010, 39, 11314–11316
This journal is
The Royal Society of Chemistry 2010
©