Letter
NJC
balanced eqn (3) does not acknowledge the formation of 1, and
chlorine gas evolution has not been obviously observed.
UO2(NO3)2Á6H2O + 4Cl3C–C(Cl)QCCl2 - UCl4
+ 4Cl2CQC(Cl)COCl + N2O4 + 2Cl2 + 6H2O
(1)
In an age of growing environmental concerns, chlorinated
solvents are increasingly becoming restricted in their use and supply,
and in this regard HCP is no exception. The price of HCP has risen
dramatically in recent years and a 2.5 litre bottle of HCP can easily
cost ca. fifty times that of the equivalent volume of tetrahydrofuran.
Original preparations of UCl4 make no mention of recycling HCP,
and given that the commonly employed ratio of UO3 to HCP is 10 g
to 100 ml (a 1 : 23 molar ratio), it is not sustainable to use and
discard 100 ml of HCP per 10 g reaction. We therefore considered it
worthwhile to assess the recyclability of HCP since only a small
molar fraction should be consumed during the preparation of UCl4.
The HCP mother liquor is usually filtered away from the
UCl4 precipitate and collected in a flask. We have found that
fractional distillation of the HCP mother liquor at reduced
pressure affords three fractions. Distillation at 10 mbar and
35–36 1C affords the first low-boiling fraction, adjustment to
10 mbar and 46–48 1C affords the second fraction, and finally
distillation at 6 mbar and 69–71 1C affords the third fraction.
The latter fraction is found by 13C{1H} NMR spectroscopy to be
essentially pure HCP which can be recycled for future use
(90 ml recovered from 200 ml of HCP). Indeed, interestingly
we find that the use of recycled HCP results in reactions with
UO3 or UO2Cl2 with significantly reduced or absent exotherms.
This is certainly desirable from a safety perspective, though we
recommend that reactions are still closely monitored through
the induction period as a precaution. Importantly, there is no
reduction in the respective yields of UCl4 when the recycled
HCP is used, irrespective of whether UO3, UO2(NO3)2Á6H2O, or
UO2Cl2 are used and the HCP can be recycled several times.
NMR spectroscopic analysis of as-supplied HCP usually reveals,
in addition to the three resonances attributable to HCP, a fourth
resonance. We have been unable to identify this compound and
note that it is not separable by distillation in the first instance,
but we find that after reaction to produce UCl4 and distillation it
Encouraged by the fact that U3O8 can be converted to UCl4 by
HCP, following the original method,4 UO2(NO3)2Á6H2O and
HCP were mixed in a 1 litre round bottom flask equipped with
three condensers stacked in-line. The mixture was heated to
reflux at which point an exotherm was observed. Briefly removing
heat allowed the exotherm to subside then heating was resumed.
The exotherm was accompanied by the formation of a brown gas,
presumably N2O4, and it should be noted it is more vigorous than
the ‘normal’ exotherm with UO3. Apart from the vigorous
exotherm, there is no evidence to suggest any side reactions of
the N2O4 with the HCP, or the chlorine that is in principle
concomitantly formed, and as described below the yield remains
excellent. It should also be noted that although chlorine is
classically proposed as one of the reaction byproducts,4 we have
not observed any obvious chlorine gas evolution and the equation
as proposed in eqn (1), whilst balanced, does not acknowledge the
formation of 1 (see below). In terms of the exotherm, in this
regard the reaction offers no improvement over the use of UO3,
but where UO3 is not available UO2(NO3)2Á6H2O now provides an
alternative, more readily available precursor to UCl4. Soon after
the exotherm the formation of a yellow precipitate is observed,
which may well be UO2Cl2 en route to UCl4. As heating is
continued this yellow precipitate dissolves and is eventually
replaced by the formation of UCl4. Once the reaction is complete
(overnight reflux), the UCl4 can be isolated by the usual filtration
and washing steps with dichloromethane to afford UCl4 in
quantitative yield. This offers a modest increase in yield, since
we usually observe yields of ca. 95% when UO3 is used.‡
UO2(NO3)2Á6H2O + xs HCl(aq) - UO2Cl2 + 2HNO3(aq)
(2)
UO2Cl2 + 2Cl3C–C(Cl)QCCl2 - UCl4 + 2Cl2CQC(Cl)COCl + Cl2
(3)
Since the observation of a yellow precipitate suggested the is absent. It is tempting to speculate that this species may have a
formation of UO2Cl2, and the exotherm was found to be quitÀe role to play in the exotherm, but we cannot unequivocally state
violent, which may be associated with the liberation of NO3
this to be the case on the basis of the evidence to hand. The first
and its conversion to N2O4, we investigated the use of UO2Cl2 two fractions to distill out appear to be acid chlorides but we
directly. UO2Cl2 may be prepared from the reaction of aqueous have not pursued their characterisation any further.
HCl on UO3; however, to provide a route to UCl4 where UO3 may
Once the first three fractions of the HCP mixture are distilled
not be available but UO2(NO3)2Á6H2O is plentiful we focussed away, which represents the majority of the material, a small oily
on the use of the latter. Accordingly, we treated UO2(NO3)2Á6H2O residue remains. This material crystallises when stored at room
with an excess of 12 M HCl to give a clear yellow solution, temperature. Alternatively, standing of the HCP–dicloromethane
then removed all volatile materials under vacuum to give a washings from the work-up of the UCl4 affords colourless crystals
yellow solid, eqn (2). The large excess of HCl ensures complete of the same material. We analysed this material and found it to be
conversion of the nitrate to chloride. Treatment of the product the 2,5-dichlorohexachlorofulvene compound Cl2CQC(CCl2CCl)2 (1).
of eqn (2) with HCP following the standard protocol results in This compound formally arises from the combination of two
the formation of UCl4 in yields averaging 92% after work-up, HCP molecules that have each lost two Cl atoms, one from the
eqn (3). Although this route requires an extra step of HCl 1- and 3-positions and in the other from the 2- and 3-positions.
treatment, this can be done quickly, and an exotherm is still The identity of this compound was confirmed by X-ray crystallo-
observed, it is significantly less vigorous than for eqn (1). This graphy as 1 has previously been structurally characterised from a
may reflect the fact that two chlorides are already installed at radical-induced cyclisation reaction of perchloro-hexatriene.6
uranium by this method and we also note the absence of
The isolation of 1 is surprising, because the reaction of HCP
evolution of any brown gas by this route. As for eqn (1), whilst with uranium oxides is usually rationalised on the basis of
New J. Chem.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2015