Note: Descriptions are shown in the official language in which they were submitted.
.il
;o ~ 1 330~0~
NP.52 603
CHE~ICAL PROCESS
This invention relates to a chemical process
and more particularly to a method for the preparation
of bromofluoromethane.
Bromofluoromethane is an important reagent
in the manufacture of intermediates, pharmaceuticals
and other chemicals.
ThiS compound has hitherto been prepared
~ 10 by three basic methods. In one method, bromofluoromethane
;~ is prepared from salts of fluoxoacetic acid using
a Hunsdiecker type of reaction as described by
Haszeldine in J. Chem. Soc., 1952, 4259-4268 and
USP 2716668. In another methv2, bromofluoromethane
is prepared from dibromofluoromethane by reductive
debromination with a Swarts reagent as described
in the early literature by Swarts et al., Bull.
Acad. Roy. Belg. l910, 113, 23. Other methods
prepare bromofluoromethane from a dihalomethane,
for example methylene bromide, by an halogen exchange
reaction or from a halomethane, for example bromomethane
or fluoromethane, by bromination or fluorination
over a catalyst such as alumina.
These earlier methods have generallv resulted
in poor yields or involved the use of dangerous
materials.
Seyferth et al. in Journal of Organic Chemistry
1963, 703-706, although presenting no supporting
data, describe the inciden~al preparation of bromo-
fluoromethane by a step-wise reduction of tribromo-
fluoromethane using tri-n-butyltin hydride. However,
the reaction conditions describêc in this paper
are such that either no bromofluoromethane will
result or the product will contain mixtures of
both hromofluoromethane and dibromofluoromethane.
Seyferth et al. also use a conventional distillation
, ,
.... .
~:
:~;
-i - 1 330~03
; 2 -
sy~tem under reduced pressure and this we helieve
will result in low yields of bromofluoromethane
since despite obtaining a 69% theoretical yield
of dibromofluoromethane from tribromofluoromethane
Seyferth et al. failed to obtain appreciable amounts
of bromofluoromethane.
Furthermore Seyferth et al. while giving
no supporting exDerimental data specifically for
reduction of tribromofluoromethane indicate that
0.02 moles of the hydride reaqent were reacted
with 0.04 moles of various halomethanes including
tribromofluoromethane, at 0C for about 1 hour,
with stirring at that temperature for 10 minutes ~
and then at room temperature for a further ten -
minutes. The description of the reaction products
is ambiguous in that the yield of dibromofluoromethane
is greater than the reactant ratios would suggest.
Although it is indicated that some hromofluoromethane
was also formed, the above discrepancy throws doub'
on the accuracy of this statement. If, however,
initially formed dibromofluoromethane was indeed
further reduced to bromofluoromethane even when
Z~ using the above apparently mild conditions there
is an implication that it would be difficult if
not impossible to use the same technique starting
wlth dibromofluoromethane to displace only a single
bromine atom to qive useful yields of bromofluoromethane.
Seyferth et al. thus do not suggest that organotin
hydride reduction of dibromo1uoromethane would
be a better route to bromofluoromethane than the
other methods disclosed above.
We have now found that bromofluoromethane
of good auality and purity can be prepared in good
yield and more efficiently from dibromofluoromethane
by the reductiee debromination of substantiallY
pure dibromofluoromethane using an organotin hydride
SUGh as tri-n-butYltin hydride.
; -. ~. .
~-
`~
`~
1 330808
3 20208-1370
According to the present lnventlon therefore we provlde
a process for the productlon o~ bromofluoromethane ln whlch
substantlally pure dlbromofluoromethane ls sub~ected to reductlon
wlth an organotln hydrlde, the molar ratlo of organotln hydrlde to
dlbromofluoromethane belng ln the range of 0.8:1 to 1.5:1.
We have surprlslngly found that the tendency, described
by Seyferth et al. of certaln polyhalomethanes to react too far
wlth such a hydrlde reagent so as to yleld doubly dehalogenated
products may be overcome by careful cholce of reactlon condltlons
so as to yleld a hlghly pure product. Under the preferred
reactlon condltlons the dlbromofluoromethane may be reacted wlth
at least a substantlally equal molar quantlty of the hydride
reagent so as to ensure substantially complete reactlon and
mlnlmlse contamlnatlon wlth unreacted startlng materlal. The term
"a substantlally e~ual molar quantlty" ls lntended to cover ratlos
of organotin hydrlde to dlbromofluoromethane between 0.8:1 and
1.5:1, preferably 0.9:1 to 1.3:1.
The good yleld and purlty of the reactlon may be
enhanced by use of a reflux condenser, e.g. a cold flnger or water
20 condenser, durlng the reductlon to return unreacted
dlbromofluoromethane to the reactlon vessel; where a water
condenser ls used thls may be connected to one or more cold traps
capable of condenslng bromofluoromethane, e.g. a cold flnger at
-196 to -40C. Thls system will permlt some vapourlsed
bromofluoromethane to dlstlll over durln~ the reductlon, thereby
removing lt from the reaction system. However, we have found that
bromofluoromethane whlch vaporlses durln~ the reduction carrles
with lt even at 0C to 5C, some dlbromofluoromethane reactant.
!B
.
. :'' : '
.
1 33080~
~! 3a 20208-1370
Consequently, where a water condenser ls used in thls way a
further cold finger trap, e.g. at -35C to +10C, ls deslrably
provlded before the -196 to -40C trap to condense and remove any
vapourlsed dibromofluorornethane.
V~
h `
S~
~ ~-
1 33030~
- 4 -
However, instead of allowing bromofluoromethane
to distill during the reduction, it may be preferable
to use a reflux condenser such as a cold finyer,
at much lower temperatures, e.g. -l96 to -40C,
to return both dibromofluoromethane and bromofluoro-
methane to the reaction, and to begin distillation
only when reduction is completed. In this case,
the reflux cold finqer will clearly have to be
removed prior to distillation.
Distillation is preferably effected at substant-
ially atmospheric pressure at relatively low temperatures,
for example 0 to 100C, preferably 20-45C. Such
distillation is slow, for example taking place
over a period of 0.5 to 15 hours, preferably 2-12
hours, for example 2-6 ho~rs.
A steady flow of an inert gas such as nitrogen
purging through the reaction mixture during distillation
;s particularly advantageous to increase the rate
of collection of bromofluoromethane. During distillation
a series of cold finger traps are advantageously
provided whereby unreacted dibromofluoromethane
is collecte~ in a first trap for example at -35
to +10C, e.g. an ice/salt trap, and the desired
bromofluoromethane in a second trap or traps at
much lower temperatures, e.g. at -196 to -40C.
Cold finqer traps which can be used are dry
ice (solid carbon dioxide)/acetone (ca. -78C),
I dry ice/carbon tetrachloride rca. -20C!, dry ice/-
methanol (ca. -70C) or liquid nitrogen (ca. -
196C). nry ice/acetone (ca. -78~) is preferred.
Thus, in general dibromofluoromethanP may
be contacted with the tri-n-butyltin hydride at
a temPerature in the ranqe of -20 r to +30C, preferably
-5C to + ]0C. The tri-n-butyl tin hydride is
preferably added slowly, e.g. over a period of
0.5 to 10 hours. The reaction vessel is ad~antageously
fitted with either a water condenser (5C to 50C)
, ~ .
.
1 330~0~
-- 5
or a cold finger (-196C to 5C) as explained above.
~;~ The reaction of the dibromofluoromethane
with the tri-n-butyltin hYdride is exothermic and
cooling is usually necessary to maintain the reaction
5 temperature at temperatures in the preferred range
t;;
-5 C to +10 C. We have found that the presence
of a free radical initiator such as ~, ~'-azoisobutyro-
nitrile in the reaction an~or additional illumination,
for example from a tungsten bulb, moderates the exothermic
10 effect. Such a free radical initiator may, for
example, be included in the reaction mixturP at
a level of up to 5% by weight relative to the dibromofluoro-
} methane, preferahly 0.1 - 1~ by weight.
After addition of the tri-n-butyl tin hydride,
l5 the mixture is advantageously further stirred at
-20C to +30C, preferably -5~ to +10C, for 0.1
to 10 hours to ensure completion of the reaction.
For distillation, the reflux cold finger
if used is removed and the mixture is warmed to
20 between 0C and 100c and stirred for a further
0.5 to 15 hours with a steady flow of nitroqen
purg ng through the reaction flask and trap(s),
advantageously at a rate of 5-500 ml per minute.
During the distillation period bromofluoromethane
25 may be collected in a cold f inger trap at -40C
to -196C. An add~ tiona7 cn7fl f-nger trap at -
35C to +10C can be inserted between the reaction
flask and the -40C to -196~ trap to reduce the
amount of dibromofluoromethane condensing in the
30 second trap.
The invention is illustrated by the following
examples. Dibromofluoromethane was obtained from
~luorochem Ltd., of Old Glossop, Derbyshire, England
and tri-n-butyltin hydride from Aldrich, ~illingham,
35 norset~ England.
k~:;
r. ~ ~ :
. .
6 - 1 33030~
xample 1
Preparation of bromofluoromethane from dibromofluoromethane
using tri-n-butyltin hydride
Tri-n-butyltin hydride ~25 g~ was added from a
dro~ping funnel over 2 hours to cold (5C~ dibromofluoro-
methane (16.5 g~ contained in a flask fitted with a
nitrogen inlet and a cold water reflux condenser
to return vapourised dibromofluoromethane to the
Il reaction. The reaction mixture was warmed to ~etween
j 25C and 30C and with a steady flow of nitroqen
purging through the reaction flask and traps the
reaction mixture was stirred for a further 2.5
hours. During the addition and warming up periods
bromofluoromethane (8.7 g) was taken off from the
reflux condenser and collected in a dry ice/acetone
cold finger trap (ca. -78~C). An ice/salt cold
finger trap (-10C) was present between the reaction
flask and the -73C trap to reduce the amount of
dibromofluoromethane condensinq in the second trap.
The product was identified by its boiling point
('7C) and n.m.r. sPectrUm (10~ v/v CDC13, delta =6.1,
doublet). G.l.c. showed 11% impurities (8% dibromo-
fluoromethanel giving a corrected theoretical yieldof 79.8%.
Example 2
Preparation of bromofluoromethane from dibromofluoro-
methane usinq tri-n-butyltin hvdride
Tri-n-butyltin hydride (181.8 g, 1.2 equivalents)
was a~ded from a dro~ping funnel over 1 hour to
cold t5C) dibromofluoromethane (100 g) in a reaction
flask fitted with a nitrogen inlet and a dry ice/acetone
cold finger reflux condenser (ca. -78C) to return
r;
~ ~ -
::~ 1 330~0~
- 7 -
vapourised bromofluoromethane and dibromofluoromethane
to the reaction, the condenser leading to a dry
ice/acetone cold finger trap ~-78C) for eventual
collection of bromofluoromethane. The temperature
was maintained between 0C and 5C during the addition.
The mixture was further stirred for 1 hour at 0C
to 5C. The dry ice/acetone cold finger reflux
condenser was removed, the mixture warmed to 40C
and with a steady flow of nitrogen purging through
~' 10 the reaction flask and traps, the reaction mixture,'!;i was stirred for a further 5 hours. During the
y warming up and stirring periods bromofluoromethane, (51 g) was collected in the dry ice/acetone cold
-~,,.
finger trap. The product was identified by its
15 boiling point (17C~ and n.m.r. spectrum ~10~ v/v
~- CDCl~, delta = 6.1, doublet). G.l.c. showed 5.7
~; impurities (0.8% dibromofluoromethane) giving a
corrected theoretical yield of 81.8%.
20 ExamPle 3
PreE~ of bromofluoromethane from dibromo_luoro-
methane using tri-n-butyltin hydride
25 Tri-n-butyltin hydride rl.89 kg, 1.2 equivalents)
was added from a drop~ing funnel over 1 hour 15
min to cold (5 C) dibromofluoromethane fl.04 Kg~
in a reaction flask fitted with a nitrogen inlet
and a dry ice/acetone cold finger reflux condenser
30 (ca -78 C~ to re~urn vapourised bromofluoromethane
and dibromofluoromethane to the reaction, the condense.
leading to a dry ice/acetone cold finger trap
j (-78 C) ~or eventual collection of ~romofluoromethane.
The temperature was maintained between 0 C and
5 C during the addition. The mixture was further
stirred for 1 hour at 0 C to 5 C. The dry ice/acetone
cold finger reflux condenser was removed, the mixture
' ':
'~
,~
.'~ :.~ ,
... .
1 330~01~
- 8 -
warmed to 40 C and with a steady flow of nitrogen
purging through the reaction flask and traps, the
reaction mixture was stirred for a further 3 hour6.
~; The reaction mixture was then warmed to 70 C and
then stirred for a further 9 hours. During the
warming up and stirring periods bromofluoromethane
(510 g) was collected in the dry ice/acetone cold
finger trap. The product was identified by its
boiling point (17 C) and nmr spectrum (10% v/v
CDC13, delta = 6.1, doublet). Glc showed 4.7~ impurities
(0.2% dibromofluoromethane) qiving a corrected
theoretical yield of 79.3%.
Example 4
Preearation of bromofluoromethane from di~romofluoro-
methane using tri-n-butyltin hydride
Tri-n-butyltin hydride (1.89 kg, 1.2 e~ui~alents)
was added from a dropping funnel over 1 hour 15
min to cold (5 C) dibromofluoromethane (1.04 kg)
and , ~'-azoisobutYronitrile (AIBN) (1.04 g) in
a reaction flask fitted with a nitrogen inlet and
a dry ice/acetone cold finger reflux condenser
(ca -78 C) to return vapourised bromofluoromethane
and dibromofluoromethane to the reaction, the condenser
leading to a dry ice~acetone cold finger trap
(-78 C) for eventual collection of bromofluoromethane.
The temperature was maintained hetween 0 C and
5 C durinq the addition. The mixture was further
stirred for 1 hour at 0 C to 5 C. The dry ice/acetone
cold finger reflux condenser was removed, the mixture
warmed to 40 C and with a steady flow of nitrogen
purging through the reaction flask and traps, the
reaction mixture was stirred for a further 3 hours.
The reaction mixture was then warmed to 70 C and
then stirred for a further 9 hours. During the
~"' ' , ,
'.'`
':
1 330~08
. ` .
:~ g
warming up and stirring periods bromofluoromethane
(470 g) was collected in the dry ice/acetone cold
finger trap. The product was identified by its
;' boiling point (17 C) and nmr spectrum (10~ v/v
, 5 CDC13, delta = 6.1, doublet). Glc showed 3.9% impurities
! (0.7~ dibromofluoromethane~ giving a corrected
.~ theoretical yield of 72.1%.
.
. .
,/ ~: ~
