Diisopropylethylamine mono(hydrogen fluoride) for nucleophilic fluorination of sensitive substrates: Synthesis of sevoflurane

Article abstract of DOI:10.1016/S0022-1139(01)00396-7

Diisopropylethylamine mono(hydrogen fluoride) (8), which can be prepared by addition of two equivalents of diisopropylethylamine to the complex diisopropylethylamine tris(hydrogen fluoride) (7), is shown to be an effective and selective nucleophilic fluorinating reagent when applied to halogen-exchange reactions of chloromethyl ethers, in particular the conversion of 1,1,1,3,3,3-hexafluoroisopropyl chloromethyl ether (3) to the volatile anesthetic sevoflurane (1,1,1,3,3,3-hexafluoroisopropyl fluoromethyl ether). The amine portion of the reagent does not react with the starting material to give a troublesome quaternary ammonium salt, as is the case for the species formed by addition of two equivalents of triethylamine to triethylamine tris(hydrogen fluoride). These particular chloromethyl ether substrates require a 1:1 stoichiometry of diisopropylethylamine to hydrogen fluoride to give useful rates of reaction and yields when solventless conditions are desired. Two other complexes, diisopropylethylamine bis(hydrogen fluoride) and 7, are ineffective for conversion of 3 to sevoflurane.

Full text of DOI:10.1016/S0022-1139(01)00396-7

Journal of Fluorine Chemistry 111 ꢀ2001) 11±16
Diisopropylethylamine monoꢀhydrogen ¯uoride) for nucleophilic
¯uorination of sensitive substrates: synthesis of sevo¯urane
Linas V. Kudzmaa, Chialang G. Huanga, Ralph A. Lessora, Leonid A. Rozova,
Syeda Afrinb, Florida Kallashib, Conor McCutcheonb, Keith Ramigb,*
aBaxter Pharmaceutical Products Inc., 95 Spring Street, New Providence, NJ 07974, USA
bDepartment of Natural Sciences, Baruch College, City University of New York, 17 Lexington Ave., New York, NY 10010, USA
Received 22 December 2000; accepted 22 March 2001
Abstract
Diisopropylethylamine monoꢀhydrogen ¯uoride) ꢀ8), which can be prepared by addition of two equivalents of diisopropylethylamine to
the complex diisopropylethylamine trisꢀhydrogen ¯uoride) ꢀ7), is shown to be an effective and selective nucleophilic ¯uorinating reagent
when applied to halogen-exchange reactions of chloromethyl ethers, in particular the conversion of 1,1,1,3,3,3-hexa¯uoroisopropyl
chloromethyl ether ꢀ3) to the volatile anesthetic sevo¯urane ꢀ1,1,1,3,3,3-hexa¯uoroisopropyl ¯uoromethyl ether). The amine portion of the
reagent does not react with the starting material to give a troublesome quaternary ammonium salt, as is the case for the species formed by
addition of two equivalents of triethylamine to triethylamine trisꢀhydrogen ¯uoride). These particular chloromethyl ether substrates require
a 1:1 stoichiometry of diisopropylethylamine to hydrogen ¯uoride to give useful rates of reaction and yields when solventless conditions are
desired. Two other complexes, diisopropylethylamine bisꢀhydrogen ¯uoride) and 7, are ineffective for conversion of 3 to sevo¯urane.
# 2001 Elsevier Science B.V. All rights reserved.
Keywords: Sevo¯urane; Synthesis; Fluorination
1. Introduction
conversions and use of excess reagents that result in large
amounts of chemical waste. Other methods employ
1,1,1,3,3,3-hexa¯uoroisopropyl methyl ether ꢀ2), conveni-
ently prepared by methylation of alcohol 1 [9]. Direct
mono¯uorination of the methyl group of methyl ether 2 is
possible using either ¯uorine or bromine tri¯ouride
[1,2,10,11], but these methods aren't practical on a large
scale due to the expense of and hazards associated with the
reagents. A practical way to convert methyl ether 2 to
sevo¯urane proceeds through 1,1,1,3,3,3-hexa¯uoroisopro-
pyl chloromethyl ether ꢀ3). The problem then becomes
selection of an inexpensive and ef®cient nucleophilic ¯uor-
inating reagent for the ®nal halex reaction. Bromine tri-
¯uoride has been used [1,2], but is not a viable reagent due
its aforementioned limitations. Alkali metal ¯uoride salts
can be used, either in a high-boiling solvent or neat
[1,7,8,12]. The high temperatures and pressures involved
and/or use of large amounts of solvent limit the usefulness of
these methods. Finally, while this work, which has culmi-
nated in a practical industrial-scale synthesis of sevo¯urane
[13], was in progress, a method independently appeared
employing various amine±hydrogen ¯uoride complexes to
convert chloromethyl ether 3 to sevo¯urane [14]. However,
Sevo¯urane ꢀ1,1,1,3,3,3-hexa¯uoroisopropyl ¯uoromethyl
ether) is a recently introduced inhaled anesthetic. It is now
one of the most widely used anesthetics because, unlike
other ¯uorinated inhaled anesthetics, it lacks a pungent odor,
and thus, can be used for induction of anesthesia. We report
herein a high-yielding and industrially practical synthesis of
sevo¯urane which showcases diisopropylethylamine monoꢀ-
hydrogen ¯uoride) as a halogen-exchange ꢀhalex) reagent.
This reagent not only gives an excellent yield of sevo¯urane,
but when applied to the synthesis of other ¯uoromethyl
ethers, gives higher yields when compared with the more
commonly used triethylamine±hydrogen ¯uoride complex.
The shortest syntheses of sevo¯urane start from
1,1,1,3,3,3-hexa¯uoroisopropyl alcohol ꢀ1), which is treated
with either bisꢀ¯uoromethyl) ether, or a hydrogen halide
and formaldehyde, or aluminum trichloride and a formal-
dehyde equivalent, to give sevo¯urane directly [1±5] or in
two steps [6±8]. These are very practical routes, despite low
* Corresponding author. Fax: ‡1-212-802-3082.
E-mail address: keith_ramig@baruch.cuny.edu ꢀK. Ramig).
0022-1139/01/$ ± see front matter # 2001 Elsevier Science B.V. All rights reserved.
PII: S 0 0 2 2 - 1 1 3 9 ꢀ 0 1 ) 0 0 3 9 6 - 7

12
L.V. Kudzma et al. / Journal of Fluorine Chemistry 111 32001) 11±16
in the latter case, the percent conversions were in general
poor, and the yields were less than ideal.
2. Results and discussion
Our initial attempts at a sevo¯urane synthesis involved
treatment of the intermediate chloromethyl ether 3 with one
equivalent of the complex prepared from one-third of an
equivalent of triethylamine trisꢀhydrogen ¯uoride) ꢀ4) [15]
and two-thirds of an equivalent of triethylamine, which
presumably forms one equivalent of triethylamine monoꢀ-
hydrogen ¯uoride) ꢀ5)1 in situ ꢀScheme 1). The added amine
was not only necessary for neutralization of the hydrogen
chloride by-product, but also enhanced the reactivity of the
reagent; see below for further discussion of the latter aspect.
This particular reagent combination has been used pre-
viously for halex reactions of some ¯uorinated chloromethyl
ethers [16]. We expected that this reagent combination
would have the advantage of generating only a small amount
of chemical waste because the amine is in principle recycl-
able after reaction of the by-product hydrochloride salt with
base.
After heating at re¯ux overnight, sevo¯urane was distilled
from the reaction mixture and isolated in 79% yield, which is
in the range of typical yields found by Terrell [14]. While
this yield is good, it was not clear what was happening to the
rest of the product, for all of the starting material was being
consumed. A careful check of the reaction mixture revealed
the presence of quaternary ammonium salt 6, which we were
unable to convert to sevo¯urane. This suggested that the
reagent formed by addition of triethylamine to complex 4
was not simply an amine salt, but contained free amine as
well; this free amine was perhaps the culprit which caused
quaternization ꢀScheme 2). We hypothesized that the free
amine was in equilibrium with triethylamine monoꢀhydro-
gen ¯uoride) ꢀ5), which suggested that use of a more
sterically hindered amine would cause complete suppression
of by-product salt 6 and allow a high-yielding practical
synthesis of sevo¯urane.
Scheme 1. Synthesis of sevoflurane.
Scheme 2. Proposed equilibrium and by-product formation.
of concentrated hydro¯uoric acid [19]. We have used the
former procedure in this work, and have found that no
puri®cation is necessary. The presumed ¯uorinating reagent,
diisopropylethylamine monoꢀhydrogen ¯uoride) ꢀ8), was
generated by addition of two equivalents of diisopropylethy-
lamine to tris complex 7 in situ. Alternatively, mono com-
plex 8 can be prepared directly by using a 1:1 molar ratio of
hydrogen ¯uoride and amine. Data which we will present
later will show that mono complex 8 is an ephemeral species,
making tris complex 7/two equivalents amine the reagent of
choice for work in the laboratory; this reagent combination
has been used as a source of nucleophilic ¯uoride in the
opening of epoxides [18,19]. On the other hand, for pre-
paration of multi-kilogram quantities of sevo¯urane, we
simply adjusted the molar ratio of amine to hydrogen
¯uoride to 1:1 in situ.
The reaction of chloroether 3 with mono complex 8 gave a
95% yield of sevo¯urane after 16 h at re¯ux ꢀScheme 3);
none of the quaternary salt analogous to salt 6 or any by-
products other than the expected diisopropylethylamine
hydrochloride were detected. The reaction can be run with
no solvent present, which allows simple distillation of the
product from the reaction mixture. And very little waste is
generated because we have been able to recover the amine
by treatment of the by-product hydrochloride salt with
aqueous base followed by phase separation.
Diisopropylethylamine, a much-used sterically hindered
amine, appeared to be a good candidate for replacement of
triethylamine. The complex diisopropylethylamine trisꢀhy-
drogen ¯uoride) ꢀ7) has been prepared by treatment of
diisopropylethylamine with three equivalents of either liquid
hydrogen ¯uoride [17,18] or hydrogen ¯uoride in the form
The following observations support our contention that
mono complex 8 is not a hydro¯uoride salt, but is possibly a
weakly coordinated complex2 as depicted: ®rst, we have
1 We prefer this formulation rather than that of a salt. Picq and
coworkers [20] has shown that addition of two equivalents of triethylamine
to tris complex 4 does not give a hydrofluoride salt, but instead gives
triethylamine bisꢀhydrogen fluoride), which can be isolated as a solid
in the presence of the excess amine. Based on this fact and some data
we will present pertaining to the closely related diisopropylethylamine±
hydrogen fluoride complexes, we view the 1:1 complex in solution of
either of these two amines and hydrogen fluoride as R3NÁHF in
equilibrium with other complexes and free amine, with the possibility
that a small amount of the ionic hydrofluoride salt is present.
2 All the data we present here are also consistent with the notion that
mono complex 8 is actually a mixture of species, including a
hydrofluoride salt, free amine, and bis and tris complexes, in
equilibrium with each other. In this case, it may be that the fluoride
ion of the hydrofluoride salt is acting as the nucleophile.

L.V. Kudzma et al. / Journal of Fluorine Chemistry 111 32001) 11±16
13
2:1 and 3:1 ratios of hydrogen ¯uoride to diisopropylethy-
lamine against the 1:1 ratio ꢀi.e. mono complex 8) and have
found that mono complex 8 and the 2:1 complex have
approximately the same rate of reaction in acetonitrile:
when chloroether 3 was treated with three equivalents of
either complex, each resulted in >90% conversion to sevo-
¯urane after 4 h at re¯ux, while three equivalents of complex
7 itself gave <20% completion after the same time period.
We have also tested the 2:1 and 3:1 complexes versus mono
complex 8 in the absence of solvent, and have found that
mono complex 8 gives by far a faster rate: after 2 h at re¯ux,
the reaction using three equivalents of mono complex 8 was
79% complete, while the reaction using the 2:1 complex was
only 22% complete and had appeared to stop converting
chloroether 3. Complex 7 was the least ef®cacious in the
absence of solvent: after 2 h, <10% conversion to sevo¯ur-
ane was observed, and the reaction had appeared to stop.
Whether neat or in acetonitrile, the 2:1 complex and tris
complex 7 both etch glassware during the reaction, while
mono complex 8 does not. So from a practical standpoint,
mono complex 8 is the preferred reagent for this particular
reaction.
In an attempt to gain some insight into the reactivity of
mono complex 8 with ¯uorinated chloromethyl ethers, we
have studied the kinetics of the reaction of the complex with
chloroether 3. Values of the rate constant k were calculated,
assuming either ®rst- or second-order kinetics [21], for the
reaction of mono complex 8 with chloroether 3 at 888C
ꢀTable 1). The homogeneous reaction mixture at various
times was monitored directly by 1H NMR analysis, using the
integration values of the ±OCH2Cl singlet versus the ±
OCH2F doublet to calculate the rate of disappearance of
chloromethyl ether 3 over time. Making the assumption that
the reaction can be only ®rst- or second-order overall, the
second-order k values were reasonably constant, indicating
an SN2 mechanism. Because the reaction is bimolecular in
nature, a classical SN1 mechanism, where a free carbocation
forms in a slow ®rst step can be ruled out. The powerfully
electron-withdrawing tri¯uoromethyl groups of chloroether
3 apparently do not allow the stabilization of a carbocation
by the neighboring oxygen atom. However, the reaction is
accelerated by presence of the oxygen atom: for example, a
Scheme 3. Synthesis of sevoflurane using mono complex 8.
treated tris complex 7 with two equivalents of diisopropy-
lethylamine in an attempt to prepare a salt, but have failed
under a variety of conditions. Mixing the two in ether/
methylene chloride to form a homogeneous solution at room
temperature for 1 h and removal of the solvent in vacuo
resulted in isolation of a bi-phasic mixture of the two
unchanged reactants. Similar results were obtained at ele-
vated temperature, either in acetonitrile at 828Cor neat at
958C. Not even the bis complex, diisopropylethylamine
bisꢀhydrogen ¯uoride) was isolated. In contrast, triethyla-
mine bisꢀhydrogen ¯uoride) is an isolable solid [20]. Sec-
ond, when tris complex 7 was dissolved in chloroform-d and
portions of diisopropylethylamine were added up to two
equivalents of amine, the chemical shifts of the methylene
and methine protons in the nuclear magnetic resonance
ꢀNMR) spectrum of the amine±hydrogen ¯uoride complex
gradually moved up®eld several tenths of a ppm, indicating
that these protons were becoming more shielded as amine
was added. The opposite trend would be expected if the
amine became a true salt ꢀi.e. the nitrogen atom acquired a
full positive charge) at the end of the addition of amine when
the stoichiometry was 1:1.
The 1:1 stoichiometry of amine to hydrogen ¯uoride is
important for the reaction to proceed. The results of Picq and
cowokers [20] indicated that a molar ratio of 2:1 of hydrogen
¯uoride:amine was most effective for nucleophilic displace-
ment reactions when triethylamine was used and the reaction
was run in acetonitrile as solvent, while the complex of 1:1
stoichiometry ꢀi.e. mono complex 5) was least effective, as
judged by the rate of the reaction. In contrast, we have tested
Table 1
First- and second-order rate constants for reaction of mono complex 8 with chloromethyl ether 3 at 888C
Time ꢀmin)a
Percent completion ꢀ%)
k  10À4 ꢀsÀ1) ꢀfirst-order)b
k  10À4 ꢀl mol sÀ1) ꢀsecond-order)c
10
20
24
35
42
49
53
58
72
3.0
2.9
2.6
2.5
2.2
2.2
1.7
1.2
1.2
1.1
1.1
1.0
1.1
1.1
30
40
50
60
115
a The 0 time point was when the mixture had reached 888C ; [3]o was calculated at this point.
b The formula k ˆ ꢀ2:303=t† ꢀlogꢀ[3]o/[3]t)) was used.
c The formula k ˆ ꢀ1=‰3Š À 1=‰3Š †=t was used, where t is the time in seconds, [3]o the original concentration of 3, and [3]t the concentration of 3 at time t.
o
t

14
L.V. Kudzma et al. / Journal of Fluorine Chemistry 111 32001) 11±16
simple primary alkyl mesylate does not react to a signi®cant
extent with complex 5 [20]. It is most likely that the reaction
of mono complex 8 with chloroether 3 occurs in one step via
a simple nucleophilic substitution reaction, with the oxygen
atom of chloroether 3 exerting an accelerating in¯uence. But
formation of a carbocation in a reversible fast ®rst step,
followed by rate-determining capture by ¯uoride ion cannot
be ruled out. It appears that even though the acid±base
interaction of diisopropylethylamine and hydrogen ¯uoride
in mono complex 8 is relatively weak, it is strong enough to
allow the ¯uorine atom to behave as a nucleophile, bearing
enough of a negative charge to displace chloride in this
specially activated substrate.
When applied to other chloromethyl ethers, mono com-
plex 8 gives superior yields when compared to mono com-
plex 5 ꢀTable 2, entries 1±6). In all cases, the cause of the
lower yield when mono complex 5 was used was traced to
quaternization. This suggests that quaternization of highly
reactive chloromethyl ethers may be a general phenomenon
when triethylamine is used as base in this type of halex
reaction. All reactions employed one equivalent of mono
complex 8 and used no solvent; thus, the products could be
distilled directly from the reaction mixture with only small
amounts of starting material as contaminant in some cases.
Reaction times at re¯ux were an overnight period or less,
except in the case of entry 9, where 35 h were required for a
reasonable level of completion. Entries 5 and 6 were per-
formed at room temperature and went to completion almost
instantaneously. Mono complex 8 shows some selectivity:
no exchange of the other chlorine atoms was seen in entries
7, 8, and 10. This new method complements antimony
pentachloride-catalyzed halex reactions, which are only
suitable in general for ethers more highly halogenated than
the ones shown in Table 2, due to the instability of ethers
containing a halomethyl group [22±24].
substrate CHClFCF2OCCl3 was treated with complex 8 at
re¯ux: the result of mono-reduction, CHClFCF2OCHCl2,
formed in 13% yield, was the predominant product. While
starting material made up most of this reaction mixture, also
seen was a 3% yield of the halogen-exchange product
CHClFCF2OCFCl2. It is likely that the reduction is occur-
ring via a radical-mediated process with the amine as
reductant. Possible initiating species include either adven-
titious metal ions [25] or oxygen, which could convert a
small amount of the amine into a radical-producing impur-
ity. It has been shown that mono-reductions of this type are
possible with alcohols or ethers as reductants [26,27], and in
some cases the reductions are thermally initiated [27]. In
support of the hypothesis of a radical-mediated process, we
have found that irradiation with UV light ꢀ254 nm high-
pressure Hg lamp shined through pyrex glass) of a solution
of CF3CCl2OCHF2 in excess complex 8 plus a trace of
acetone gives the product of mono-reduction,
CF3CHClOCHF2, in about 10% yield; no other products
were seen. This same conversion was obtained when diiso-
propylethylamine was substituted for complex 8. But on the
other hand, no reaction of any kind was seen when
CF3CCl2OCHF2 was treated with complex 8 at re¯ux in
the dark; in this case, the re¯ux temperature of the reaction
mixture was much lower than in the other cases, making a
thermally initiated free-radical reduction less likely. These
observations lead us to infer that diisopropylethylamine was
acting as the reductant in a free-radical reaction, with the
source of hydrogen being the C±H bond alpha to the nitrogen
atom. It has been shown [28] that amines can undergo
scission of this C±H bond to give a free radical, formation
of which is among the early steps in reductions of this type
[27].
Although the exact nature of the species formed by
addition of a two-fold excess of amine to a tertiary amine
trisꢀhydrogen ¯uoride) complex is far from being under-
stood, this study shows that the mono complex can be a
selective source of ¯uoride for bimolecular substitution
reactions. Further clari®cation of the structure and reactivity
An unexpected by-product was seen in the reaction
mixture of entry 7, Table 2. The product of both halogen-
exchange and reduction, CHClFCF2OCH2F, was obtained in
6% yield. This reduction process also occurred when the
Table 2
Preparation of fluoromethyl ethersa
Entry
Reagent
Starting material
Formula
Product
Formula
Yield ꢀ%)
Compound
Compound
1
2
8
5
8
5
8
5
8
8
8
8
CF3CHFOCH2Cl
CF3CHFOCH2Cl
CF3CHFCF2OCH2Cl
CF3CHFCF2OCH2Cl
CH3CH2OCH2Cl
CH3CH2OCH2Cl
CCl2FCF2OCH2Cl
CHClFCF2OCH2Cl
ꢀCF3)2CHOCHClF
CF3CHClOCHClF
9
9
CF3CHFOCH2F
CF3CHFOCH2F
CF3CHFCF2OCH2F
CF3CHFCF2OCH2F
CH3CH2OCH2F
CH3CH2OCH2F
CCl2FCF2OCH2F
CHClFCF2OCH2F
ꢀCF3)2CHOCHF2
CF3CHClOCHF2
10
10
12
12
14
14
16
18
20
22
78
52
81
3
11
11
13
13
15
17
19
21
4
35
5
65
Traceb
65
6
7
8
66
9
10
66
67
a Detailed preparations for sevoflurane, 10, and 12 can be found in [13].
b Quaternization of the starting material by the amine was the major outcome.

L.V. Kudzma et al. / Journal of Fluorine Chemistry 111 32001) 11±16
15
of these complexes may allow prediction of which kind and
what proportion of amine would work best for a particular
class of substrates.
removed under vacuum into a dry ice-cooled trap, giving
7.55 g of liquid. GCanalysis indicated that this liquid was
28.5% 15 and 64.0% 16, equivalent to 4.83 g 16 ꢀ65% yield
based on consumed starting material). 1H NMR d 5.63
ꢀd, J ˆ 51:3 Hz). 19F NMR ꢀH-coupled) d À75.9 ꢀt, 1F,
J ˆ 8:2 Hz), À88.3 ꢀt, 2F, J ˆ 8:2 Hz), À154.8 ꢀtt, 1F,
J ˆ 51:2, 8.2 Hz).
3. Experimental
3.1. General methods and starting materials
3.4. 2-Chloro-1,1,2-trifluoroethyl fluoromethyl
ether 318) [33]
1H NMR spectra were recorded in CDCl3 solvent at either
300 MHz with a Bruker WH-300 instrument, or, for routine
analyses, at 60 MHz with a Hitachi R-1200 instrument;
chemical shifts ꢀd) are reported in ppm down®eld from
the standard tetramethylsilane, d ˆ 0. 19F NMR spectra
were recorded in CDCl3 solvent at 282 MHz with a Bruker
WH-300 instrument; chemical shifts are reported in ppm
up®eld from the standard ¯uorotrichloromethane, d ˆ 0.
Gas chromatographic analyses were performed on a Hew-
lett-Packard 5890 gas chromatograph ®tted with a packed
Supelco 10% igepalco-880/50% lb-550X on 80/60 Chro-
masorb PAW column with TCD detection.
Detailed preparations for sevo¯urane, 1,2,2,2-tetra¯uor-
oethyl ¯uoromethyl ether ꢀ10) [16], and 1,1,2,3,3,3-hexa-
¯uoropropyl ¯uoromethyl ether ꢀ12) [16] can be found in
[13]. The starting materials 1,2,2,2-tetra¯uoroethyl chloro-
methyl ether ꢀ9) [16], 1,1,2,3,3,3-hexa¯uoropropyl chloro-
methyl ether ꢀ11) [16], 2,2-dichloro-1,1,2-tri¯uoroethyl
chloromethyl ether ꢀ15) [29], 2-chloro-1,1,2-tri¯uoroethyl
chloromethyl ether ꢀ17) [29], 1,1,1,3,3,3-hexa¯uoroisopro-
pyl chloro¯uoromethyl ether ꢀ19) [30], and 1-chloro-2,2,2-
tri¯uoroethyl chloro¯uoromethyl ether ꢀ21) [31] were pre-
pared as previously reported. Chloromethyl ethyl ether ꢀ13),
diisopropylethylamine, triethylamine, triethylamine trihy-
dro¯uoride ꢀtriethylamine trisꢀhydrogen ¯uoride), 4) were
purchased from Aldrich Chemical Co.
A mixture of complex 7 ꢀ17.2 g, 91 mmol), diisopropy-
lethylamine ꢀ23.5 g, 182 mmol), and chloroether 17 ꢀ50.0 g,
273 mmol) was heated at re¯ux under nitrogen for 18 h. The
volatile components were transferred in vacuo to a dry-ice-
cooled receiver, giving 46.4 g liquid. This was washed
with 10% aqueous HCl and water to remove traces of
diisopropylethylamine. Drying over CaCl2 gave 39.8 g
liquid which showed by GCanalysis a 2.0:1.0 mixture of
18 and 17. Fractional distillation using a 2 ft  1=2 in.
vacuum-jacketed column packed with glass beads gave
21.7 g 18 ꢀ66% yield based on consumed starting material),
bp ˆ 828Cꢀlit : ˆ 83:58C[33]). 1H NMR d 6.14 ꢀdt, 1H
J ˆ 48:0, 4.28 Hz), 5.62 ꢀdt, 1H, J ˆ 54:2, 3.33 Hz). 19F
NMR ꢀH-decoupled) d À82 ꢀm, 2F), À154.5 ꢀt, 1F,
J ˆ 9:0 Hz), À154.8 ꢀbr t, 1F, J ˆ 11:6 Hz).
3.5. 1,1,1,3,3,3-Hexafluoroisopropyl difluoromethyl
ether 320) [30]
A mixture of complex 7 ꢀ3.1 ml, 16 mmol), diisopropy-
lethylamine ꢀ5.7 ml, 33 mmol), and chloroether 19 ꢀ10.8 g,
46.0 mmol) was heated at re¯ux under nitrogen for 35 h.
1
Periodic monitoring by H NMR of aliquots dissolved in
CDCl3 indicated the reaction was 77% complete after this
time. Simple distillation gave 7.27 g distillate, bp 52±568C
ꢀlit. bp of pure 20 ˆ 41:58C[30]). 1H NMR analysis indi-
cated that the distillate was composed of 91% ether 20, 6.5%
starting ether 19, and 2.5% diisopropylethylamine, equiva-
3.2. Fluoromethyl ethyl ether 314) [32]
Diisopropylethylamine ꢀ6.63 ml, 38.1 mmol) was added
to complex 7 under nitrogen, giving a biphasic mixture. With
rapid stirring, chloroether 13 ꢀ5.32 ml, 57.3 mmol) was
added in one portion, giving a mild exotherm. Upon cooling,
the mixture solidi®ed. Simple distillation under nitrogen
gave 2.9 g ꢀ65% yield) 14 as a colorless liquid, bp 368C. 1H
NMR d 5.27 ꢀd, 2H, J ˆ 57 Hz), 3.79 ꢀqd, 2H, J ˆ 7:2 and
1.2 Hz), 1.29 ꢀt, 3H J ˆ 7:2 Hz) [WARNING: Compound
14 is highly unstable to moisture. Upon exposure to the
atmosphere, it decomposes rapidly, producing HF fumes].
1
lent to 6.6 g ꢀ66% yield) of 20. H NMR d 6.44 ꢀt, 1H
J ˆ 70:5 Hz), 4.84 ꢀsept, 1H, J ˆ 5:66 Hz). 19F NMR ꢀH-
decoupled) d À74.6 ꢀt, 6F, J ˆ 4:40 Hz), À85.6 ꢀsept, 2F,
J ˆ 4:40 Hz).
3.6. 1-Chloro-2,2,2-trifluoroethyl difluoromethyl
ether 3isoflurane, 22) [29]
A mixture of complex 7 ꢀ7.00 ml, 36.8 mmol), diisopro-
pylethylamine ꢀ12.8 ml, 73.4 mmol), and chloroether 21
ꢀ20.0 g, 99.6 mmol) was heated at re¯ux under nitrogen
for 21 h. Periodic monitoring by 1H NMR of aliquots
dissolved in CDCl3 indicated the reaction was 77% complete
after 4 h. Simple distillation gave 16.4 g distillate, bp 55±
878Cꢀlit. bp of pure 22 ˆ 48:58C[29]). 1H NMR analysis
indicated that the distillate was composed of 86% ether 22,
and 14% diisopropylethylamine, equivalent to 16.4 g ꢀ77%)
3.3. 2,2-Dichloro-1,1,2-trifluoroethyl fluoromethyl
ether 316) [33]
A mixture of complex 7 ꢀ2.79 g, 14.7 mmol), diisopro-
pylethylamine ꢀ4.07 g, 31.5 mmol), and chloroether 15
ꢀ10.15 g, 46.69 mmol) was heated at re¯ux under nitrogen
for 16 h. After cooling to RT, the volatile component was

16
L.V. Kudzma et al. / Journal of Fluorine Chemistry 111 32001) 11±16
[10] M.T. Baker, J.H. Tinker, J.A. Ruzicka, US Patent 5,705,710 ꢀ1998);
Chem. Abstr. 128 ꢀ1998) 101825.
of product 22. The diisopropylethylamine was removed
by washing with cold 1 M aqueous HCl. Drying over CaCl2
[11] J.P. Russell, A.J. Szur, R.C. Terrell, US Patent 3,897,502 ꢀ1975);
Chem. Abstr. 83 ꢀ1976) 178306.
1
gave 12.3 g ꢀ67% yield) pure product 22. H NMR d 6.43
ꢀdd, 1H J ˆ 72:0, 69.9 Hz), 6.04 ꢀq, 1H, J ˆ 4:05 Hz). 19F
NMR ꢀH-decoupled) d À80.8 ꢀs, 3F), À87.6 ꢀd, 1F,
J ˆ 161 Hz), À88.6 ꢀd, 1F, J ˆ 161 Hz).
[12] D.F. Halpern, M.L. Robin, US Patent 4,874,901 ꢀ1989); Chem. Abstr.
112 ꢀ1990) 157680.
[13] L.V. Kudzma, R.A. Lessor, L.A. Rozov, K. Ramig, US Patent
5,886,239 ꢀ1999); Chem. Abstr. 130 ꢀ1999) 237281.
[14] R.C. Terrell, US Patent 5,969,193 ꢀ1999); Chem. Abstr. 130 ꢀ1999)
237274.
Acknowledgements
[15] M.A. McClinton, Aldrichim. Acta 28 ꢀ1995) 31 ꢀreview).
[16] H. Muffler, R. Franz, Ger. Offen. 2,823,969 ꢀ1979); Chem. Abstr. 92
ꢀ1980) 197322.
K.R. and F.K. thank the Eugene Lang Foundation for a
Faculty-Student Fellowship award. K.R. thanks the Weiss-
man School of Arts and Sciences of Baruch College for
research start-up funds. Dr. Lee Sprague is thanked for a
helpful discussion.
[17] R. Franz, J. Fluorine Chem. 15 ꢀ1980) 423.
[18] H. Suga, T. Hamatani, M. Schlosser, Tetrahedron 46 ꢀ1990) 4247.
[19] Z. Du, M.J. Haglund, L.A. Pratt, K.L. Erickson, J. Org. Chem. 63
ꢀ1998) 8880.
[20] M.B. Guidicelli, D. Picq, B. Veyron, Tetrahedron Lett. 31 ꢀ1990)
6527.
[21] C.W. Keenan, D.C. Kleinfelter, J.H. Wood, General College
Chemistry, 6th Edition, Harper & Row, 1980, pp. 404±434.
[22] D.C. England, J. Org. Chem. 49 ꢀ1984) 4007.
[23] M. Hudlicky, J. Fluorine Chem. 29 ꢀ1985) 349.
[24] K. Ramig, L.V. Kudzma, R.A. Lessor, L.A. Rozov, J. Fluorine Chem.
94 ꢀ1999) 1 and references therein.
References
[1] B.M. Regan, J.C. Longstreet, US Patent 3,683,092 ꢀ1972); Chem.
Abstr. 77 ꢀ1973) 156346.
[25] R.C. Terrell, K. Hansen, US Patent 4,334,105 ꢀ1982); Chem. Abstr.
94 ꢀ1982) 120858.
[2] B.M. Regan, J.C. Longstreet, US Patent 3,689,571 ꢀ1972); Chem.
Abstr. 77 ꢀ1973) 156346.
[26] O. Paleta, V. Dadak, V. Dedek, H.-J. Timpe, J. Fluorine Chem. 39
ꢀ1988) 397.
[3] C.L. Coon, R.L. Simon, US Patent 4,250,334 ꢀ1981); Chem. Abstr.
94 ꢀ1981) 208358.
[27] K. Ramig, F. Quiroz, G.G. Vernice, L.A. Rozov, J. Fluorine Chem. 80
ꢀ1996) 101 and references therein.
[4] C.L. Coon, R. L. Simon, US Patent 4,469,898 ꢀ1984); Chem. Abstr.
94 ꢀ1981) 208358.
[28] M. Barta, F. Hampl, F. Liska, V. Dedek, Collect Czech. Chem.
Comm. 59 ꢀ1994) 1820.
[5] T.A. Ryan, L. Burgess, WO Patent 97/25,303 ꢀ1997); Chem. Abstr.
127 ꢀ1998) 135551.
[29] R.C. Terrell, L. Speers, A.J. Szur, J. Treadwell, T.R. Ucciardi, J. Med.
Chem. 14 ꢀ1971) 517.
[6] T. Kawai, M. Watanabe, US Patent 5,811,596 ꢀ1998); Chem. Abstr.
127 ꢀ1998) 190459.
[30] L. Speers, A.J. Szur, R.C. Terrell, J. Treadwell, T.U. Ucciardi, J. Med.
Chem. 14 ꢀ1971) 593.
[7] C. Bieniarz, C. Behme, K. Ramakrishna, J. Fluorine Chem. 106
ꢀ2000) 99.
[31] K. Ramig, A. Krishnaswami, L.A. Rozov, Tetrahedron 52 ꢀ1996) 319.
[32] M. Hiyashi, H. Kato, M. Oyamada, J. Mol. Spectrosc. 83 ꢀ1980) 408.
[33] R.C. Terrell, L. Speers, A.J. Szur, T. Ucciardi, J.F. Vitcha, J. Med.
Chem. 15 ꢀ1972) 604.
[8] C. Bieniarz, K. Ramakrishna, C. Behme, US Patent 6,100,434
ꢀ2000).
[9] L.S. Croix, US Patent 3,911,024 ꢀ1975); Chem. Abstr. 83 ꢀ1976)
058108.