On the modes of reaction of xenon difluoride with organic substrates: The influence of solvent and vessel

Full text of DOI:10.1021/ja9804316

6842
J. Am. Chem. Soc. 1998, 120, 6842-6843
On the Modes of Reaction of Xenon Difluoride with
Organic Substrates: The Influence of Solvent and
Vessel
occurred and no signals due to un-ionized XeF
with the important exception of CH CN solution in which the
un-ionized XeF
is quite stable and shows a clean 19F spectrum.
Clearly, the glass surface catalyzes the rapid decomposition of
XeF and we presume that the weakly basic solvent CH CN
2
could be detected
3
8
2
2
3
Christopher A. Ramsden* and Rachel G. Smith
effectively neutralizes the acidic surface of the glass.
2
Department of Chemistry, Keele UniVersity
All our previous studies have been carried out in glass vessels.
We therefore investigated the fluorodesilylation of p-tolyltrimeth-
ylsilane (4, R ) Me) under identical conditions but in an FEP
flask and found that no reaction took place. We repeated the
reaction in a dry glass flask that had been prewashed with alkali
(NaOH) and again no reaction was observed. It is clear that the
acidic surface of the glass is necessary for the fluorodesilylation
Keele, Staffordshire, UK ST5 5BG
ReceiVed February 9, 1998
2
Xenon difluoride (XeF ) is a stable, commercially available
solid that is attracting increasing attention as an “electrophilic”
fluorinating agent.1 We wish to report that the choice of solvent
and reaction vessel profoundly influences the mode of reaction
of this reagent with organic substrates. Our own studies with
+
reaction (4 f 6), presumably to generate the electrophile FXe .
3
When this cannot occur, i.e., in FEP, alkali washed glass, or CH -
XeF
2
have focused on reactions with trimethylsilyl derivatives:2
CN solution, the XeF
an electrophile.
2
remains un-ionized and does not react as
this strategy avoids formation of hydrogen fluoride during reaction
and has enabled us to demonstrate conditions for two discrete
modes of reaction which we now describe.
The reaction conditions upon which two distinct modes of
reaction of XeF with organic substrates critically depend can
2
Our work with divalent xenon is part of a comparative study
of the chemistry of hypervalent derivatives of the Periodic Table
triad Te, I, and Xe that is ultimately directed to the application
now be recognized. Under protic conditions ionization occurs
and the reacting species is the electrophilic fluoroxenonium cation
+
(
FXe ), which in the case of fluorodesilylation (4 f 6) leads to
2-4
of novel reactions to biological problems.
In this context XeF
2
the observed product by a mechanism that we have discussed
provides access to fluoro derivatives of biological interest and to
methods for the rapid introduction of the isotope 18F (t1/2 110
min) into biological ligands for positron emission tomography
2,6
previously. The ligand coupling of the xenon intermediate (5
f 6) probably occurs via homolytic cleavage of the carbon-
xenon bond and subsequent in-cage coupling to fluorine in accord
(
PET).5 We are particularly interested in (i) how the nonbonding
9
with the principles recently described by Frohn and Bardin. Some
molecular orbitals of the hypervalent three-center, four-electron
escape of the aryl radical then satisfactorily accounts for the minor
IV III
II
[
3c-4e] bonds in these species (Te , I , and Xe ) influence their
2a
radical derived products that we observed. This contrasts with
mode of reaction and (ii) the ability of these reagents to function
10
the observation of Bardin and co-workers, who reported that in
+
+
+
2
first as electrophiles (e.g. An Te OH, PhI OAc, or Xe F) and
CH
products are hydrocarbons (e.g. 1) formed via the aryl radicals
. They rationalize the formation of these radicals in terms of a
3
CN together with a trace of fluoride catalyst (CsF) the
2
subsequently as exceptionally good leaving groups (i.e. An Te,
PhI, or Xe).6 These considerations led us to predict and
2
demonstrate the fluorodesilylation of aryltrimethylsilanes (4 f
10
one-electron oxidation of an intermediate anion 3. This mech-
6
) (Scheme 1) using xenon difluoride.2a
We have previously shown that the XeF
anism is entirely consistent with the presence of only un-ionized
2
-mediated fluorode-
Cl , CHCl
6 6
F solution and observed that no reaction takes place
XeF
reaction of trimethylsilyl benzoates 7 with ionized XeF
the rearranged products 8 whereas un-ionized XeF acts as a one-
2
under their conditions. Similarly, we now recognize that
silylation of aryltrimethylsilanes (4 f 6) occurs in CH
CFCl , or C
in CH
CN solution.2 In another study we showed that trimeth-
ylsilyl benzoates 7 react with XeF in CH Cl or C to give
rearranged aryl fluoroformates 8 whereas in CH CN the products
are derived from aryl radicals 2.2b These results prompted us to
make further studies of the influence of solvent on XeF reactions.
It has previously been shown that XeF is moderately stable in
CH Cl and CHCl and more stable in CH
confirmed that the F NMR spectra of XeF
CN, and CH CN/D O solution with fluorinated ethylene
propylene (FEP) liners show that the solute is stable over hours
and the characteristic “triplet”, made up of a singlet (δ -179)
plus a 1 Xe doublet (J 5642 Hz), confirms that it is un-ionized
i.e. F-Xe-F). In contrast, when we attempted the determination
2
2
3
,
2
leads to
3
2
a
3
2b
electron oxidizing agent giving radical-derived products (e.g. 1).
2
2
2
6 6
F
On the basis of the mechanistic studies described above we
3
2
anticipated that if reaction of XeF with trimethylsilyl enol
ethers1 occurs via a SET mechanism then MeCN/glass would
provide convenient and optimal reaction conditions. To test this
hypothesis 1-((trimethylsilyl)oxy)cyclohexene (10) was reacted
1a
2
2
CN.7 We have
,8
2
2
3
3
2
with XeF under these conditions and this did indeed result in a
1
9
2
in CH Cl , CHCl
2
2
3
,
very clean and quantitative transformation to 2-fluorocyclohex-
anone (9). Subsequently we have studied a series of trimethylsilyl
enol ethers and obtained similar results: details of the scope of
this convenient R-fluorination procedure together with further
evidence for a SET mechanism occurring via a radical cation
(Scheme 2) will be published elsewhere.1 In contrast, use of
C
6
F6, CH
3
3
2
F
29
(
1b
19
of the F spectra in glass NMR tubes very rapid decomposition
6 6
C F as solvent in a glass flask results in a much more complex
(
1) Tius, M. A. Tetrahedron 1995, 51, 6605.
reaction and the composition of the product mixture is dependent
upon reaction time and workup conditions. In addition to
(
2) (a) Lothian, A. P.; Ramsden, C. A. Synlett 1993, 753. (b) Nongkunsarn,
P.; Ramsden, C. A. J. Chem. Soc., Perkin Trans. 1 1996, 121.
+
compound 9, probably formed via electrophilic addition of FXe
(3) (a) Ramsden, C. A.; Rose, H. L. Synlett 1997, 27. (b) Bodajla, M.;
Jones, G.; Ramsden C. A. Tetrahedron Lett. 1997, 38, 2573. (c) Ramsden, C.
A.; Rose, H. L. J. Chem. Soc., Perkin Trans. 1 1997, 2319.
to the enol ether, cyclohexanone (11) and rearrangement products
including ꢀ-caprolactone (12) are formed (Scheme 1). Rear-
rangement to oxepane derivatives is characteristic of electron-
deficient intermediates2 and probably occurs by a mechanism
of the type shown in Scheme 3.
(4) (a) Clews, J.; Cooksey, C. J.; Garratt, P. J.; Land, E. J.; Ramsden, C.
A.; Riley, P. A. J. Chem. Soc., Chem. Commun. 1998, 77. (b) Cooksey, C. J.;
Garratt, P. J.; Land, E. J.; Pavel, S.; Ramsden, C. A.; Riley P. A.; Smit, N. J.
Biol. Chem. 1997, 272, 26226. (c) Clews, J.; Land, E. J.; Ramsden, C. A.;
Riley, P. A. J. Chem. Soc., Perkin Trans. 1 1998, 1009.
b,12
(
(
(
5) Pike, V. W. J. Psychopharmacol. 1993, 7, 139.
6) Ramsden, C. A. Chem. Soc. ReV. 1994, 23, 111.
(9) Frohn, H. J.; Bardin, V. V. Z. Anorg. Allg. Chem. 1996, 622, 2031.
(10) Bardin, V. V.; Stennikova, I. V.; Furin, G. G.; Leshina, T. V.;
Yakobson, G. G. Zh. Obshch. Khim. 1988, 58, 2580.
(11) (a) Filler, R.; Cantrell, G. J. Fluorine Chem. 1985, 27, 35. (b) Ramsden,
C. A.; Smith, R. G. Manuscript in preparation.
7) Dukat, W. W.; Holloway, J. H.; Hope, E. G.; Townson, P. J.; Powell,
R. L. J. Fluorine Chem. 1993, 62, 293.
(8) (a) Meinert, H.; R u¨ diger, S. Z. Chem. 1967, 7, 239. (b) Ingman, L. P.;
Jokisaari, J.; Oikarinen, K.; Seydoux, R. J. Magn. Reson. 1994, 111A, 155.
S0002-7863(98)00431-4 CCC: $15.00 © 1998 American Chemical Society
Published on Web 06/24/1998

Communications to the Editor
J. Am. Chem. Soc., Vol. 120, No. 27, 1998 6843
Scheme 1
Scheme 2
with the glass flasks catalyze the reactions. In a recent review
Tius makes the following introductory statement: “Most reac-
1
2
tions with XeF can be conducted in glass apparatus with no
apparent diminution in yield.” This view must now be modi-
fied: the choice of flask and solvent is critical but has received
little attention. We have analyzed the primary literature for 54
discrete types of reaction reviewed by Tius.1 For this reaction
set the types of vessel used were as follows: FEP or similar flask
(
(
7), glass flask (12); both FEP and glass (3); metal containers
2); and nature of the flask not specified (30). The choice of
Scheme 3
2 2 3 3 6 6
solvent [CH Cl (57%), CHCl (2%), CH CN (5%), C F (2%),
other solvents or mixtures (34%)] was also variable and there
was no systematic selection or correlation with the choice of
vessel.
Finally, to exploit the catalytic effect of glass, we have
investigated a series of ten XeF
2
reactions that have been described
The reaction of XeF
or an electrophile (FXe ) has been discussed,
2
as either a one-electron oxidant (XeF
2
)
1,12
as requiring HF catalysis.
Our preliminary studies now show
+
1,13-15
and it is well-
2 2
that when these reactions are carried out in dry CH Cl solution
known that HF catalyzes the latter mode of reaction. However,
the influence of solvent and flask has not previously been
recognized. Reutov and co-workers16 have suggested that trace
amounts of “active electrophilic fluorinated silicon derivatives”
catalyze the fluorination of cyclophanes: Shellhamer and co-
in a glass flask the formation of the relevant products does occur
without the need for addition of HF. This is a much more
convenient procedure for performing electrophilic reactions of
XeF and should encourage wider usage of this reagent.
2
In summary, we have described the previously unrecognized
influence of solvent and vessel on the mode and quality of reaction
17
workers have suggested that “adventitious HF” formation occurs
as a result of traces of “moisture on the surface of solid XeF ”.
We now believe that in both these cases acidic protons associated
2
2
of XeF and have demonstrated examples of how an understanding
of this influence leads to improved synthetic procedures for both
the one-electron oxidation (e.g. R-fluorination) and electrophilic
addition (e.g. glass catalysis) modes.
(12) (a) Stavber, S.; Koren, Z.; Zupan, M. Synlett 1994, 265. (b) Zajc, B.;
Zupan, M. J. Org. Chem. 1990, 55, 1099. (c) Stavber, S.; Sket, B.; Zajc, B.;
Zupan, M. Tetrahedron 1989, 45, 6003. (d) Zajc, B.; Zupan, M. J. Org. Chem.
1
982, 47, 573.
Acknowledgment. We thank the EPSRC for the award of a project
studentship (to R.G.S.) and Drs. V. W. Pike and F. I. Aigbirhio (MRC
Cyclotron Unit, Hammersmith Hospital, UK) for valuable discussions.
(
13) Filler, R. Isr. J. Chem., 1978, 17, 71.
(
14) Zupan, M. In The Chemistry of Functional Groups, Supplement D;
Patai, S., Rappoport, Z., Eds.; John Wiley & Sons Ltd.: Chichester, 1983; pp
6
57-679.
15) Bardin, V. V.; Yagupolskii, Y. L. In New Fluorinating Agents in
Organic Synthesis; German, L., Zemskov, S., Eds.; Springer-Verlag: Berlin,
989.
(
JA9804316
1
(17) (a) Shellhamer, D. F.; Chiaco, M. C.; Gallego, K. M.; Low, W. S. C.;
Carter, B.; Heasley, V. L.; Chapman, R. D. J. Fluorine Chem. 1995, 72, 83.
(b) Druelinger, M. L.; Shellhamer, D. F.; Chapman, R. D.; Shackelford, S.
A.; Riner, M. E.; Carter, S. L.; Callahan, R. P.; Youngstrom, C. R. J. Chem.
Soc., Perkin Trans. 2 1997, 787.
(
16) Nikanorov, V. A.; Rozenberg, V. I.; Kharitonov, V. G.; Antonov, D.
Yu.; Mikul’shina, V. V.; Galakhov, M. V.; Krut’ko, D. P.; Kopelev, N. S.;
Guryshev, V. N.; Yur’ev, V. P.; Reutov, O. A. Dokl. Akad. Nauk SSSR (Engl.)
990, 315, 370.
1