(Re)Investigation of the reactivity of uranium hexafluoride toward several organic functions at room temperature

Article abstract of DOI:10.1016/j.jfluchem.2014.06.007

The annual worldwide production of UF₆ is very large and this compound is not used. Consequently, it could be interesting to find some applications as organic reagent. UF₆ could be considered as an oxidizer of various functions. However, it seems also present some possibilities as a fluorinating reagent in mild conditions.

Full text of DOI:10.1016/j.jfluchem.2014.06.007

Journal of Fluorine Chemistry 167 (2014) 74–78
Contents lists available at ScienceDirect
Journal of Fluorine Chemistry
journal homepage: www.elsevier.com/locate/fluor
(Re)Investigation of the reactivity of uranium hexafluoride
toward several organic functions at room temperature
Olivier Roy a, Bernard Marquet b, Jean-Paul Alric d, Alex Jourdan e, Bertrand Morel d,
b,c,
b,**
*
Bernard R. Langlois
, Thierry Billard
a InstitutdeChimiedeClermont-Ferrand(ICCF)–UMR6296,Universite´ BlaisePascal,CampusdesCe´zeaux,24avenuedesLandais,BP8002663171Aubie`re,France
b
´
´
InstituteofChemistryandBiochemistry(ICBMS–UMRCNRS5246),UniversitedeLyon, UniversiteLyon1,CNRS,43Bddu11novembre1918, 69622Lyon,France
c CERMEP-In vivo Imaging, Groupement Hospitalier Est, 59 Bd Pinel, 69003 Lyon, France
d Usine de Pierrelatte, service R&D, Comurhex Co., BP 29, 26701 Pierrelatte, France
e BUC/DRD, Etudes et Projets, Areva NC, BP 44, 26701 Pierrelatte, France
A R T I C L E I N F O
A B S T R A C T
Article history:
The annual worldwide production of UF6 is very large and this compound is not used. Consequently, it
could be interesting to find some applications as organic reagent. UF6 could be considered as an oxidizer
of various functions. However, it seems also present some possibilities as a fluorinating reagent in mild
conditions.
Received 19 March 2014
Received in revised form 6 June 2014
Accepted 8 June 2014
Available online 19 June 2014
ß 2014 Elsevier B.V. All rights reserved.
Keywords:
UF6
Oxidizer
Fluorinating reagent
Acid fluoride
1. Introduction
As a Lewis acid, UF6 deprotects hydrazones and oximes into
ketones [7–12], acylhydrazides into acids [7–12], and benzyl or
Uranium hexafluoride is one of the most important fluorinated
compounds, the annual worldwide production of which was
around 60,000 tons in 2010 [1]. Because of the unique properties of
fluorine, and despite its high molecular weight, UF6 sublimes at
56.2 8C [2]. In the gas phase, its main use is the isotopic enrichment,
benzhydryl ethers into alcohols [10]. It also causes the self-
condensation of acetonitrile [12] and cyclohexanone under heating
[4]. As an oxidizer, it transforms, at 0 8C, methyl or benzyl ethers
into an oxonium salt which, after hydrolysis, leads to carbonyl
derivatives [7–11] or, in the presence of
a,v-dithiols or butyl
by gaseous diffusion or centrifugation, of natural uranium into 235
enriched uranium, used as fuel in nuclear plants. Depleted UF6
238U enriched UF6), which is useless until now, is then stored as
U
lithium, provides dithioketals or tertiary alcohols, respectively
[10,11]. It oxidizes benzyl alcohols [7–11] and benzyl bromides
[10] into ketones and tertiary N,N-dimethylamines into quaternary
iminium fluorides [7–11]. UF6 is also able to polymerize
cyclohexane with partial oxidation (even in liquid nitrogen) and
(
such or hydrolyzed into depleted uranium oxide. Storage, as well as
hydrolysis, is costly. Thus, in an economical and sustainable
context, it would be interesting to find a more noble use of
depleted UF6, i.e. as a chemical reagent. Nevertheless, to the best of
our knowledge, few things have been reported about that [2]. From
the literature, it appears that, at room temperature or even below,
UF6 mainly acts as a Lewis acid [2–5] or an oxidizer [6] but behaves
as a fluorinating agent at higher temperatures [2].
to deprotect ketals into ketones with partial oxidation into
v-
hydroxyesters [4]. Interestingly, it transforms, at 0 8C, alkyl or aryl
aldehydes into acyl fluorides [4,7–11] without intermediate
oxidation to acids, since the latter do not react with UF6 under
these conditions [10]. This was the only fluorination observed at
low temperature since the fluorination of adamantone into
difluoroadamantane, firstly reported by Olah et al. [10,11] has
been refuted later [4].
*
Corresponding author at: Institute of Chemistry and Biochemistry (ICBMS –
Under heating, UF6 becomes a fluorinating agent but the
reactions are not usually clean. For example, in the vapor phase at
UMR CNRS 5246), Universite´ de Lyon, Universite´ Lyon 1, CNRS, 43 Bd du 11
novembre 1918, 69622 Lyon, France. Tel.: +33 0472448129.
**
80 8C, simple alcohols are transformed into
a mixture of
Corresponding author.
fluoroalcanes, olefins (or ethers), the ratio of which depends on
E-mail address: thierry.billard@univ-lyon1.fr (T. Billard).
http://dx.doi.org/10.1016/j.jfluchem.2014.06.007
0022-1139/ß 2014 Elsevier B.V. All rights reserved.

O. Roy et al. / Journal of Fluorine Chemistry 167 (2014) 74–78
75
Scheme 1. Fluorination of undec-1-ene by UF6.
Scheme 2. Cation-mediated fluorination of undec-1-ene.
the substrate [13] whereas, at reflux of ClF2CCFCl2, 1-adamantol
delivers a mixture of 1-fluoroadamantane (55%) and difluoroada-
mantanes (45%) [14]. Under the same conditions, cyclohexene
epoxide gives 2-fluorocyclohexanol in a poor yield (20%), along
with 25 undetermined products, whereas chloro- or bromo-
cyclohexane deliver a mixture of more than 40 non-fluorinated
products [14]. Fluorination becomes more efficient, though not
totally selective, at higher temperatures: for example, above 150 8C
and sometimes up to 600 8C, tetrachloromethane was transformed
into perchlorofluoromethanes, formal addition of F2 (eventually
followed by chlorine/fluorine exchange) occurs from various
chlorinated ethylenes [15–17] and, between 350 and 450 8C,
tetrafluoroethanes lead to pentafluoroethane [18]. Also, in contrast
to results reported by Olah et al. at 0 8C [10], acetic and
trifluoroacetic acids partially yield acyl fluorides at 40–60 8C
[19,20]. A more selective fluorination can be obtained when pellets
of NaF, KF or CaF2 are added to the system. For example, between
100 8C and 400 8C in the presence of such fluorides, simple alkanes
are highly fluorinated by UF6 in excess [21]. In the same way, when
opposed to UF6/CaF2 at 100 8C, alcohols, diols, epoxides and
cycloalkenes lead to polyfluorinated alkanes whereas acid, ester
and aldehyde moieties are transformed into trifluoromethyl
groups [22]. Such an enhancement of reactivity could be due to
the fact that, because of itÀs strong Lewis acidity, UF6 can add
As the reaction was carried out under anhydrous conditions
and the results were the same whatever the sample of UF6, a
simple hydrofluorination by HF (arising from partial hydrolysis
of UF6) can be ruled out. Nevertheless, the migration of fluorine
along the carbon chain, as well as the predominance of internal
fluorides, constitutes a strong argument for the occurrence of a
cation or a radical-cation. However, as no intermediate could be
quenched by alkylsilanes or phenylselenenyl chloride, the
neutralization of the cations by FÀ could be a quasi-concerted
process. Finally, as the results were similar with 1 or 0.5
equivalent of UF6, intermediate R-UF5 is suspected to react with
olefins as UF6 does. These observations lead us to the mechanism
proposed in Scheme 2.
Nevertheless, as UF6 is a strong oxidizer, the formation of a
radical-cation cannot be ruled out. Such a mechanism could
explain that several olefins (and eventually dienes) were detected
by gas chromatography and that the yield of fluoralkanes never
exceeded 66% (Scheme 3).
1,1-Disubstituted olefins, such as 2-methyl-undec-1-ene, or
internal olefins (7-tetradecene, (+)-p-menth-1-ene) were also
hydrofluorinated at room temperature but slower, so that 2
equivalents of UF6 were needed to complete the reaction within
16–19 h (respective yields: 35%, 51%, 23%). In this case, no
migration of fluoride was observed, that is consistent with the
higher stability of secondary and tertiary cations. Nevertheless, no
diastereomeric excess was obtained, that is also consistent with
the occurrence of a cation.
Then, reaction of UF6 with alcohols was examined at room
temperature. From the literature, no fluorination should be
expected since the only positive results were reported under
heating, eventually in the gas phase [13,14,22]. Indeed, in our
hands, aliphatic primary alcohols remained unaffected by UF6,
except 4-(biphenyl)methanol and 2-(phenyl)ethanol, which were
significantly converted but delivered a complex mixture among
which non-fluorinated products were largely predominant. In
these latter cases, oxidation probably matched extensively
fluorination, as reported by Olah et al. [7–11]. However, with a
secondary alcohol such as decan-2-ol, fluorination was predomi-
nant (46%) over some oxidation (10%) and dehydration (6%)
(Scheme 4).
À
fluoride to give UF7 or UF8 anions [23,24].
In order to rationalize these elements and to solve contra-
dictions between several reports, we reexamined the behavior of
depleted UF6 toward different functions, most often at room
temperature. As a large panel of functions was studied, the results
reported below were not fully optimized.
2. Results and discussion
As far as olefins are concerned, only the reaction of gaseous UF6
with polychlorinated olefins has been reported in the literature.
Thus, we studied the action of this reagent toward undec-1-ene in
CFCl2CF2Cl(F113)atroomtemperature. Itappearedthatamixtureof
1-, 2-, 3-, 4- and 5-fluoro-undecanes was obtained, among which 3-
fluoro and 4-fluoroundecanes were predominant (Scheme 1) [25].
Scheme 3. Radical-cation mediated fluorination of undec-1-ene.

76
O. Roy et al. / Journal of Fluorine Chemistry 167 (2014) 74–78
Scheme 4. Fluorination of decan-2-ol by UF6.
Scheme 6. Proposed mechanism for the reaction of aldehydes with UF6.
Scheme 5. Proposed mechanism for the fluorination of secondary alcohols. (UOF4 is
as reactive as UF6 and could react again [13]).
Scheme 7. Proposed mechanism for the reaction of dithioketals with UF6.
These reactions could occur as depicted in Scheme 5. 4-phenyl-
butan-2-ol behaved in a different way since the fluorinated
derivative was obtained in small amounts compared to non-
fluorinated products (probably ethers). As this behavior was
similar to that of 2-(phenyl)ethanol, the presence of an aromatic
ring may prevent fluorination because of the formation of a charge-
transfer complex between UF6 and the aromatic system [4]. Also,
2,2-dimethylbenzyl alcohol did not deliver any fluorinated
product: it was transformed into 1,1,3-trimethyl-dihydroindene,
as when treated with sulfuric acid or AlCl3 [26,27]. In this case, UF6
only acted as a Lewis acid and the intermediate cation was too
stable to react with hindered fluorides.
system (entries 4–6), its reactivity decreases dramatically but 3
becomes predominant over 2. A tentative mechanism is given in
Scheme 6.
Such a mechanism is consistent with our results. Firstly, if
cation 4 is not too conjugated, its reactivity increases and the
formation of 5 is favored. Secondly, route b (that is the formation of
3) would be favored if UF6 is present in excess and if cation 6 is
strongly conjugated. Another argument for the occurrence of
cation 6 is the fact that, when the reaction from aromatic
aldehydes was quenched with sodium methanethiolate, aryl
dimethyl thioketals were predominantly formed.
As described several times by Olah et al. [7–11] aldehydes are
rapidly transformed by UF6, at 0 8C, into acyl fluorides but the yield
did not exceed 50% [10]. However, when examining this reaction
upon dodecanal, we found that the yield was significantly
increased at room temperature, provided that reaction longer
times were applied. Other aldehydes were treated in the same way.
In most cases, the gem-difluoroderivatives were detected as side-
products and could sometimes become the major product
(Table 1).
From Table 1, it seems that when the electrophilicity of the
carbonyl moiety increases (entries 1 and 3), its reactivity also
increases and the yield of 2 largely exceeds that of 3. In contrast,
when the carbonyl moiety is strongly conjugated with an aromatic
As expected, ketones were far less reactive than aldehydes. For
example, under the conditions reported in Table 1 (entry 2),
benzophenone delivered difluorodiphenylmethane in a 3% yield
only. Thiones were not more reactive than ketones. Thus, the
fluorination of dithioketals was examined. In this case, satisfactory
results were obtained (Table 2).
The introduction of the substrate as a solution in AcOEt
(entries 3 and 6) or CF2ClCFCl2 (entries 11 and 12) could
sometimes improve the yield, probably because of solubility
reasons. The fact that better yields were usually obtained when
fluorination occurred in benzylic position (compare entries 2, 7
and 9) could suggest
intermediate cation should be not too stabilized by
a
cationic process. Maybe also, the
-effects
s
Table 1
Reaction of UF6 with aldehydes.
.
Entry
R
x
t (h)
1 (%)a
2 (%)a
3 (%)b
1
2
3
4
5
6
C11H23
2
2
2
2
4
2
67
17
19
22
66
72
10
54
20
92
52
89
79
41
78
1
2
4
Ph
4-NO2-C6H4
4-biphenyl
4-biphenyl
2-naphtyl
2
7
2
39
10
1
a
Determined by GC analysis.
Determined by 19F NMR titration.
b

O. Roy et al. / Journal of Fluorine Chemistry 167 (2014) 74–78
77
Table 2
Reaction of dithioketals with UF6.
.
Scheme 8. Reaction of acids with UF6 (Yields determined by 19F NMR titration).
Entry
R1, R2
x
8 (%)d
1
2
Ph, Ph 7a
1
2
1
2
2
2
2
4
2
4
2
4
40
50
79
16
13
30
20
26
18
37
38
47
Ph, Ph 7a
Ph, Ph 7a a
3
4
4-F-C6H4, Me 7b
4-biphenyl, Me 7c
4-biphenyl, H 7d a,b
C8H17, Me 7e
C8H17, Me 7e
C12H25, H 7f
C12H25, H 7f
C12H25, H 7f c
C12H25, H 7f c
5
6
7
8
Scheme 9. Proposed mechanism for the reaction of acids with UF6.
9
10
11
12
C4H9CH55CHCH2COOH) delivered a mixture of non-fluorinated
products. Again, disappointing results were observed when an aryl
a
Introduced as a solution in AcOEt.
Along with 23% of 4-Ph-C6H4-C(O)H.
Introduced as a solution in CF2ClCFCl2.
Determined by 19F NMR titration.
b
c
moiety is present in
v-position on the substrate (notice that
d
benzoic acid did not react at all). On the other hand, perfluor-
ooctanoic or camphosulfonic acid failed to react.
Aliphatic diacids were also treated by UF6 at room temperature.
In the presence of two equivalents of UF6, bis-acyl fluorides or
anhydrides were obtained, depending on the nature of the
substrate (Table 3). The selective formation of anhydrides was
probably favored by a Thorpe–Ingold effect (entries 2–4) and/or
the formation of 5- or 6-membered rings. It must be also noticed
that, despite the fact that conjugated unsaturated monoacids
and itaconic acid (9c) (Table 3, entry 4) behave as expected, maleic
and fumaric acids failed to react.
Simple esters did not react under these conditions, but dimethyl
malonate was fluorinated on the central carbon by UF6, though in a
low yield (15% after 20 h with 1 eq. of UF6). Probably, the enol form
of malonate was implied, as proposed in Scheme 10 (this
mechanism could suggest that 2 eq. of UF6 should be used).
Finally, we studied the behavior of UF6 toward (trichloro-
methyl)benzene. As depicted in Scheme 11, this substrate
delivered (chlorofluoromethyl)benzenes, and even (trifluoro-
methyl)benzene, under milder conditions than those reported
for the chlorine/fluorine exchange on CCl4 [17].
(compare entries 7 and 11). Thus, by analogy with the reaction of
dithioketals with DBH and HF/pyridine (9:1) [28], a mechanism
is proposed in Scheme 7. Such a mechanism is consistent with
the fact that UF6 is a soft Lewis acid which interacts better with
sulfur than with oxygen. This strong interaction could also
explain the good results obtained from 7a since it should lower
the formation of a charge-transfer complex between UF6 and aryl
groups present on the substrate. Also, as, after hydrolysis, 8d is
produced along the corresponding aldehyde, it could be
suspected that intermediate A is rather stable.
Though it was reported that acids did not react with UF6 at 0 8C
[10], we were pleased to obtain acyl fluorides from acids in fair to
good yields at room temperature and within a relatively short
period (Scheme 8).
This reaction could follow the mechanism proposed in
Scheme 9.
Unexpectedly, conjugated alkenyl or alkynyl acids also
reacted cleanly without any alteration of the multiple bond
but non-conjugated unsaturated acids (CH255CH(CH2)8COOH,
Table 3
Reaction of diacids with UF6.
Scheme 10. Proposed mechanism for the reaction of dimethyl malonate with UF6.
.
Entry
9
x
t (h)
10 (%)b
11 (%)b
12 (%)a
1
2
3
4
9a
9a
9b
9c
1
2
2
2
4
17
17
17
80
0
0
0
20
99
0
0
99
0
0
63
a
Determined by GC analysis.
Determined by 19F NMR titration.
Scheme 11. Reaction of (trichloromethyl)benzene with UF6 (Yields determined by
19F NMR titration).
b

78
O. Roy et al. / Journal of Fluorine Chemistry 167 (2014) 74–78
However, Ph-CH2Cl and Ph-CHCl2 did not reacted under these
Acknowledgements
conditions. The same differences in reactivity have been reported
many times for the Cl/F exchange on
a-chlorinated toluenes
One of the authors (O.R.) greatly thanks the Comurhex Co. for a
financial grant during his post-doctoral fellowship. This company
is also acknowledged for a generous gift of depleted UF6. The
authors thank the CNRS for their financial support and the French
Fluorine Network is also thanked for its support.
with HF. We also observed that PhSCCl3 did not reacted as
expected with UF6, probably because this soft Lewis acid
was quenched by sulfur and did not interact with chlorine atoms.
3. Conclusion
References
In conclusion, we demonstrated that, at room temperature
(or, in some rare cases, under mild heating), UF6 behaves not
only as an oxidizer, as previously reported, but also as a
fluorinating agent of a various panel of organic substrates. This
preliminary study could open the way to a potential valorization
of depleted UF6 in accordance with a sustainable development
context.
[1] http://www.world-nuclear.org.
[2] J. Burdon, Fluorinating Agents Introduction of Fluorine with High-Valency Oxi-
dizing Metal Fluorides, 4th ed., Georg Thieme Verlag, Stuttgart, 1998.
[3] P.J. Vergamini, J. Chem. Soc., Chem. Commun. (1979) 54–55.
[4] A. Goosen, C.W. McCleland, P.J. Venter, M.W. Venter, S. Afr. J. Chem. 40 (1987) 30–34.
[5] G.A. Olah, Friedel-Crafts and Related Reactions, Interscience, New York, 1963.
[6] G.M. Anderson, J. Iqbal, D.W.A. Sharp, J.M. Winfield, J.H. Cameron, A.G. McLeod, J.
Fluor. Chem. 24 (1984) 303–317.
[7] G.A. Olah, J. Welch, T.-L. Ho, J. Am. Chem. Soc. 98 (1976) 6717–6718.
[8] G.A. Olah, J. Welch, G.K. Surya Prakash, T.-L. Ho, Synthesis (1976) 808–809.
[9] G.A. Olah, J. Welch, Synthesis (1976) 809–810.
4. Experimental
[10] G.A. Olah, J. Welch, J. Am. Chem. Soc. 100 (1978) 5396–5402.
[11] G.A. Olah, J.G. Shih, G.K.S. Prakash, J. Fluor. Chem. 33 (1986) 377–396.
[12] J.A. Berry, R.T. Poole, A. Prescott, D.W.A. Sharp, J.M. Winfield, J. Chem. Soc., Dalton
Trans. (1976) 272–274.
All the experiments were carried out in a glove bag. UF6 was
provided by the Comurhex Co. as a solution in CF2ClCFCl2.
[13] N.G. Schnautz, P.J. Venter, S. Afr. J. Chem. 45 (1992) 59–62.
[14] N.G. Schnautz, S. Afr. J. Chem. 46 (1993) 26–29.
[15] V.V. Shatalov, E.A. Filippov, V.T. Orekhov, V.F. Bagryantsev, A.V. Smol’nyi, B.V.
Lebedev, At. Energy 72 (1992) 196–199.
4.1. General procedure
The substrate, dissolved in CF2ClCFCl2, was placed under dry
nitrogen in a Nalgene1 vessel or in an autoclave. Then, a solution of
UF6 in CF2ClCFCl2 was introduced into the reactor. The mixture was
stirred at room temperature (or, in some cases, at higher
temperature) during the time indicated above.
[16] A.K. Denisov, V.Y. Zakharov, A.N. Golubev, A.P. Shishkanov, V.G. Tsar’kov, V.V.
Shatalov, V.T. Orekhov, A.E. Goncharov, A.E. Goncharov, Vserossiiskii Nauchno-
Issledovatelskii Institut Khimicheskoi Tekhnologii, Russia; Kirovo-Chepetskii
Khimicheskii Kombinat, RU2030380C1 (1995) (Chem. Abstr. 124:29251).
[17] V.T. Orekhov, Russ. Chem. Rev. 46 (1977) 420–435.
[18] Vserossiyskii Nauchno-Issledovatel’skii Institut Khimicheskoi Tekhnologii, Rus-
sia; OAO Kirovo-Chepetskii Khimicheskii Kombinat im. B. P. Konstantinova
RU2165919C1 (2001) (Chem. Abstr. 137:80608).
[19] N.S. Nikolaev, Y.Y. Kharitonov, A.T. Sadikova, T.A. Rasskazova, A.Z. Kozorezov, Izv.
Akad. Nauk SSSR, Ser. Khim. (1972) 757–764.
[20] A.T. Sadikova, N.S. Nikolaev, T.A. Rasskazova, Zh. Neorg. Khim. 15 (1970) 2012–2013.
[21] H.H. Gibbs, E. I. du Pont de Nemours & Co., US3235608 (1966) (Chem. Abstr.
64:84112).
[22] J. Hewes, A. Robertson, M. Kocur, AlliedSignal Inc., USA, WO9616002A1 (1996)
(Chem. Abstr. 125:85836).
[23] S. Katz, Inorg. Chem. 3 (1964) 1598–1600.
[24] S. Katz, Inorg. Chem. 5 (1966) 666–668.
[25] R.D. Chambers, A.M. Kenwright, M. Parsons, G. Sandford, J.S. Moilliet, J. Chem. Soc.,
Perkin Trans. 1 (2002) 2190–2197.
The reaction was monitored by 19F NMR and Gas Chroma-
tography.
After reaction, the reaction medium was carefully hydrolyzed
with H2O (2 mL) and Et2O (2 mL) then poured into a mixture of H2O
(5 mL) and Et2O (5 mL). An emulsion often resulted. It can be
homogenized by treatment with a saturated solution of ammoni-
um oxalate, eventually followed by filtration on Celite. The two
phases were separated. The aqueous phase was extracted twice
with Et2O (2 Â 10 mL) and the gathered organic phases were dried
over MgSO4 then evaporated under vacuum. The residue was
finally analyzed by 19F and 1H NMR and Gas Chromatography. After
hydrolysis, the metal was, in general, recovered as uranium oxide.
[26] K.T. Serijan, P.H. Wise, J. Am. Chem. Soc. 73 (1951) 4766–4769.
[27] M. Hojo, T. Ichi, K. Shibato, J. Org. Chem. 50 (1985) 1478–1482.
[28] S.C. Sondej, J.A. Katzenellenbogen, J. Org. Chem. 51 (1986) 3508–3513.