Synthetic and mechanistic aspects of halo-F-methylphosphonates

Article abstract of DOI:10.1016/j.jfluchem.2011.05.034

The synthesis of a variety of new halo-F-methylphosphonates has been achieved by a Michaelis-Arbuzov type reaction between a halo-F-methane and a trialkyl phosphite. This synthesis has proved to be of wide scope and utility for the high yield preparation of a number of heretofore unknown compounds. The ¹H, ¹⁹F, ¹³C and ³¹P NMR spectroscopic properties are reported in detail. The mechanism for the formation of bromodifluoromethylphosphonates has been shown to proceed through the intermediacy of difluorocarbene:CF₂. The phosphonate products have been shown to react with a wide variety of reagents. Fluoride and alkoxide ions react by attack at phosphorus with cleavage of the carbon-phosphorus bond and formation of [:CF₂] from the bromodifluoromethylphosphonates and the CFBr₂^- anion from the dibromofluoromethylphosphonates. Iodide ion and tertiary phosphines react by attack at the ester carbon to give stable phosphonate salts. Hydrolysis of the phosphonate esters with 50% aqueous HCl gives the expected phosphonic acids. Trimethylsilyl bromide attacks phosphoryl oxygen to afford the bis(trimethylsilyl) esters.

Full text of DOI:10.1016/j.jfluchem.2011.05.034

Journal of Fluorine Chemistry 132 (2011) 815–828
Contents lists available at ScienceDirect
Journal of Fluorine Chemistry
journal homepage: www.elsevier.com/locate/fluor
Synthetic and mechanistic aspects of halo-F-methylphosphonates
*
Richard M. Flynn, Donald J. Burton
Department of Chemistry, University of Iowa, Iowa City, IA 52242, USA
A R T I C L E I N F O
A B S T R A C T
Article history:
The synthesis of a variety of new halo-F-methylphosphonates has been achieved by a Michaelis–Arbuzov
type reaction between a halo-F-methane and a trialkyl phosphite. This synthesis has proved to be of wide
scope and utility for the high yield preparation of a number of heretofore unknown compounds. The 1H,
19F, 13C and 31P NMR spectroscopic properties are reported in detail. The mechanism for the formation of
bromodifluoromethylphosphonates has been shown to proceed through the intermediacy of
difluorocarbene:CF2. The phosphonate products have been shown to react with a wide variety of
reagents. Fluoride and alkoxide ions react by attack at phosphorus with cleavage of the carbon–
Received 7 February 2011
Received in revised form 23 May 2011
Accepted 30 May 2011
Available online 6 June 2011
Keywords:
Bromodifluoromethylphosphonates
Dibromofluoromethylphosphonates
Difluorocarbene
ꢀ
phosphorus bond and formation of [:CF2] from the bromodifluoromethylphosphonates and the CFBr2
anion from the dibromofluoromethylphosphonates. Iodide ion and tertiary phosphines react by attack at
the ester carbon to give stable phosphonate salts. Hydrolysis of the phosphonate esters with 50%
aqueous HCl gives the expected phosphonic acids. Trimethylsilyl bromide attacks phosphoryl oxygen to
afford the bis(trimethylsilyl) esters.
Michaelis–Arbuzov
Halo-F-methylphosphonates
ß 2011 Elsevier B.V. All rights reserved.
1. Introduction
later reported by McIvor [5] with no preparative details given.
Fluoromethyl and difluoromethylphosphonates have been pre-
The synthesis and chemistry of polyfluoroalkyl phosphines and
phosphoranes has been extensively reported in the chemical
literature. For example, polyfluorinated phosphonium salts have
been synthesized and found to be effective ylide precursors for the
preparation of a variety of fluoromethylene olefins as well as
precursors to halo-F-carbenes and halo-F-methyl carbanions [1]. In
contrast to this chemistry, there has been relatively little work
done in the investigation of fluoroalkyl phosphonates, especially
the halo-F-methyl derivatives even considering the fact that
phosphonic acids and their derivatives are more numerous than
any of the other compounds containing the carbon–phosphorus
bond [2]. We have previously reported our preliminary work on
the synthesis of these halo-F-methyl phosphonates [3] and with
this paper will discuss the complete details of their synthesis, the
mechanism of their formation as well as the multi-centered
reactivity of this interesting class of compounds.
Prior to this work there were only a few reports of any halo-F-
methyl phosphonic acid derivatives. Emeleus and co-workers [4]
had reported the difficult preparation of the trifluoromethyl
phosphonic, phosphorous and phosphinous acids by an oxidative
hydrolysis of trifluoromethylphosphine derivatives. Little further
chemistry of these materials was reported at that time though the
infrared spectrum of diethyl trifluoromethylphosphonate was
pared by the reaction of a sodium dialkyphosphite with CH2BrF or
CF2HCl respectively [6,7].
The Michaelis–Arbuzov reaction is by far the most widely
employed method for the synthesis of phosphonate esters [8].
CF3Cl and CF3I were found to be inert to the reaction with either
triethyl phosphite or sodium diethyl phosphite [9]. CFCl3 was
reported to react with triethyl phosphite under rigorous conditions
(elevated temperature in a sealed tube) to afford the fluorodi-
chloromethyl phosphonate although no analytical data was
presented in proof of the structure [10]. The early failures or
difficulty in the synthesis of these phosphonates probably
discouraged further attempts at this chemistry. We now wish to
expand on our earlier report on the facile Michaelis–Arbuzov
reaction that the halo-F-methanes CF2Br2 (1) and CFBr3 (2)
undergo with alkyl phosphites. The halo-F-methyl phosphonate
esters can be obtained in good to excellent yields in a very simple
synthetic procedure and have been found to undergo a variety of
reactions.
2. Results and discussion
2.1. Synthesis
When an equimolar amount of 1 is added to an ice cold solution
of triethyl phosphite in the absence of a solvent, a short induction
period is followed by an extremely vigorous reaction. However
when the reaction is carried out in a solvent such as triethylene
* Corresponding author. Tel.: +1 319 335 1363; fax: +1 319 335 1270.
E-mail address: donald-burton@uiowa.edu (D.J. Burton).
0022-1139/$ – see front matter ß 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.jfluchem.2011.05.034

816
R.M. Flynn, D.J. Burton / Journal of Fluorine Chemistry 132 (2011) 815–828
glycol dimethyl ether (triglyme, TG) or diethyl ether (EE) as a
diluent the Michaelis–Arbuzov reaction occurs smoothly and
readily at either room temperature or with diethyl ether, at reflux
temperature and requires no special precautions, as shown in
Eq. (1).
fluorodichloromethylphosphonate 15, there were also three other
components which were identified by NMR: (EtO)2P(O)H, (EtO)2-
P(O)C2H5 and (EtO)3P(O). The diethyl ethylphosphonate probably
arises directly from the reaction of the co-product C2H5Cl with
triethyl phosphite in a Michaelis–Arbuzov reaction. In a similar
reaction, it was found that although CF2BrCl did not react with
triethyl phosphite at room temperature in triglyme or under the
action of ultraviolet light, it did react in a similar manner to CFCl3 in
an autoclave (Entry 17). The only fluorinated phosphonate formed
in this reaction was 4. No (EtO)2P(O)CF2Cl was observed.
In contrast to these unreactive halo-F-methanes, CF2BrI reacted
rapidly with triethyl phosphite at room temperature in triglyme
(Entry 15). Two products were formed in a 7.5/1 ratio by NMR. The
minor product was identified as 4 by comparison with an authentic
sample. Distillation of the product, after removal of the more
volatile components (found to be ethyl bromide and ethyl iodide in
a 7.5/1 ratio) afforded diethyl iododifluoromethylphosphonate 14.
In diethyl ether as solvent, <1% of 4 was formed. This result will be
discussed in detail in the mechanism.
In addition to the trialkyl phosphites listed in Table 1, a number
of other phosphites were investigated. Triphenyl phosphite, tris-2-
chloroethyl phosphite and sodium diphenyl phosphite did not
react with 1 at either ambient temperature or at 80 8C to yield any
phosphonate products. Somewhat surprisingly, ethyl diphenyl
phosphite also did not react with 1 even at 170 8C in a pressure
vessel. Apparently the substitution of a single phenyl group with
an ethyl group did not give a phosphite which was appreciably
more reactive than triphenyl phosphite itself. However, benzyl
diethyl phosphite reacted readily with 1 to give 4 as the only
product. No trace of the other possible Michaelis–Arbuzov product
(PhCH2O)(EtO)P(O)CF2Br was observed. An alternative preparation
of 4 has been reported via the reaction of 1 with trialkyl phosphites
in the absence of solvent. For safety purposes we prefer that the
reaction be carried out in a solvent as diluent since the reaction can
be highly exothermic [11].
ðEtOÞ3P þ CF2Br2ð1Þ ! ðEtOÞ2PðOÞCF2Brð4Þ þ EtBr
(1)
Details of the reactions and workup procedures can be found in
Section 4. In addition to 1 and 2, other halo-F-methanes including
CF2BrI and CF2BrCl also have been found to undergo this reaction
and yield phosphonates as the final products. Table 1 is a
compilation of the compounds prepared and their yields as well
as some solvent comparisons.
The yields of phosphonate products generally range from good to
excellent. The lower yields observed in the two autoclave reactions
with CF2BrCl (Entry 17 in Table 1) and, as previously reported, CFCl3
(Entry 16) can be ascribed to the significantly reduced reactivity of
these two halo-F-methanes vs. the highly reactive 1 and 2. Diethyl
ether is the most convenient solvent for these reactions since it is
easily dried and easy to separate from the product (e.g. Entry 2). The
use of triglyme as solvent was found to give somewhat lower yields
(see Entry 3), probably due to losses during the workup procedures.
In addition, triglyme cannot typically be completely removed from
the product phosphonates by distillation and is removed by brine
washes prior to final purification by distillation. One important
exception to note isthat the reaction of trimethyl phosphite with1 is
very slow in diethyl ether while reaction in triglyme solvent is
complete in 24 h (Entry 1).
On the other hand it was found that the reaction between CFBr3
2 and trialkylphosphites was most conveniently carried out neat
with no additional solvent (Entries 9, 13, and 14). For these
reactions, it is possible to simply mix together the phosphite and 2
at room temperature. After
a short induction period, the
exothermic reaction will occur with a rise in temperature to
about 65 8C. After the exotherm is over the reaction is essentially
complete and the product can be isolated by vacuum distillation.
As noted above the only halo-F-methane previously reported to
react with triethyl phosphite was CFCl3. In an attempt to duplicate
this report, the reaction between CFCl3 and triethyl phosphite was
carried out in a 125 mL Hastelloy C autoclave at 120 8C and
autogenous pressure (Entry 16). After workup and distillation of
the products, it was found that in addition to the expected diethyl
As further examples of the generality of this reaction, it was
found that diethyl phenyl phosphonite and ethyl diphenyl
phosphinite both reacted readily with 1 giving the expected
products ethyl bromodifluoromethylphenylphosphinate 16 and
bromodifluoromethyldiphenylphosphine oxide 17 respectively as
illustrated in Eq. (2).
PhPðOEtÞ2 þ CF2Br2 ! PhPðOÞðOEtÞCF2Br
16
(2)
Ph2POEt þ CF2Br2 ! Ph2PðOÞCF2Br
17
Table 1
Halo-F-methylphosphonates. (RO3)P + halo-F-methane ! (RO)2P(O)CFXYa + RZ.a
A
photochemical modification of the Michaelis–Arbuzov
Entry
No.
R
Halo-F-methane
CFXY
Solvent
Yield (%)b
reaction to prepare halo-F-methylphosphonates which are not
amenable to the reaction scheme described here has been reported
previously, as shown below in Eq. (3) [12].
1.
3
CH3
CF2Br2
CF2Br2
CF2Br2
CF2Br2
CF2Br2
CF2Br2
CF2Br2
CF2Br2
CFBr3
CF2Br
CF2Br
CF2Br
CF2Br
CF2Br
CF2Br
CF2Br
CF2Br
CFBr2
CFBr2
CFBr2
CFBr2
CFBr2
CFBr2
CF2Ic
TG
55
95
55
55
42
75
65
81
46
78
60
22
63
53
60c
27
20
2.
4
C2H5
EE
3.
4
C2H5
TG
4.
5
n-C3H7
i-C3H7
i-C3H7
n-C4H9
i-C4H9
CH3
EE
ðEtOÞ P þ CF3Iuvꢀl!ightðEtOÞ PðOÞCF3 þ EtBr
(3)
3
2
5.
6
TG
6.
6
EE
7.
7
EE
2.2. Spectroscopic properties of Halo-F-methylphosphonates
8.
8
Neat
Neat
EE
9.
9
The 1H, 19F, 13C and 31P NMR data for the phosphonates
prepared are detailed in Section 4. The 19F NMR spectrum of the
phosphonates exhibits in all cases a doublet due to the fluorine
10.
11.
12.
13.
14.
15.
16.
17.
10
10
11
12
13
14
C2H5
CFBr3
C2H5
CFBr3
TG
i-C3H7
n-C4H9
i-C4H9
C2H5
CFBr3
EE
CFBr3
Neat
Neat
TG
Neatd
Neatd
2
CFBr3
phosphorus coupling with JPF coupling constants in the range of
CF2BrI
CFCl3
77 up to 93 Hz for the halofluoromethyl phosphonates up to
124 Hz for the previously reported (EtO)2P(O)CF3 [12]. These
coupling constants display a moderate correlation with the Pauling
electronegativity of the halo-F-methyl group attached to the
phosphorous. The electronegativity values were calculated by
summation of the electronegativities of the halo-F-methyl sub-
stituents, excluding phosphorus, using the Pauling elemental
15
e
C2H5
CFCl2
CF2Br
C2H5
CF2BrCl
a
X, Y =F, Cl, Br, I; Z =Br, Cl, I.
b
c
Isolated yield based on phosphite.
(EtO)2P(O)CF2I 14 and 4 in 7/1 ratio.
d
e
Reaction carried out in 125 ml Hastelloy C bomb under autogenous pressure.
4 only product formed.

R.M. Flynn, D.J. Burton / Journal of Fluorine Chemistry 132 (2011) 815–828
817
130
120
110
100
90
290
CF3
CF3
270
250
CF2Br
CFCl2
230
210
190
170
150
CF2H
CF2Br
CF2I
CF2I
CFCl2
CFBrH
CFBrH
CFBr2
CF2H
CCl3
80
CFBr2
70
9
10
11
12
Pauling Electronegativity
2
Fig. 1. JPF vs. Pauling electronegativity of halo-F-methyl group.
9
10
11
12
Pauling Electronegatiivity
electronegativities as calculated by Allred [13]. This method was
proved to be of value previously in the correlation of the 19F NMR
chemical shift and 2JPF coupling constants of a series of fluorinated
quaternary phosphonium salts [14].
1
Fig. 3. JCP vs. Pauling electronegativity of halo-F-methyl group.
any carbons further away were always singlets. Several com-
pounds showed more complex coupling patterns. For example, the
isopropyl groups of compounds 6 and 11 show non-equivalent
carbon atoms in the 13C NMR spectra and thus show two doublets
in the spectra instead of a simple doublet. In these cases, the
phosphorus atom acts as a center for magnetic asymmetry as
shown in the following projections along the C–O–P bonds (which
is shown as linear for simplicity). It is obvious that the two methyl
groups are non-equivalent in any rotamer. This non-equivalence
was not observed at the relatively low 60 MHz resolution NMR but
was easy to detect in the carbon NMR spectrum.
The 31P NMR chemical shifts range from ꢀ2.6 to 7.5 and exhibit
the expected multiplicity for coupling to fluorine. No correlation
was observed for these chemical shifts with either the electro-
negativity of the halo-F-methyl group or with the variation of the
substituents in the ester portion of the phosphonates (Fig. 1).
The interpretation of the 1H NMR spectra is straightforward
although the similarity of the 3JHH and 3JHP coupling constants led
to multiplets which were not completely resolved and usually gave
a multiplicity less than what would otherwise be expected. For
example, in most cases, the methylene protons in the CH3CH2OP
grouping appear to be a pentet, the methylene hydrogens in
CH2CH2OP a quartet and so on. These couplings are described in
Section 4 as ‘‘apparent’’ pentets, or quartets and the like and the
‘‘apparent’’ coupling constants given without regard to the nuclei
which are coupled.
The 13C NMR chemical shifts for the halo-F-methyl carbon
atoms range from 84 to 120 ppm and are roughly correlated to the
Pauling electronegativity of the halo-F-methyl carbon in the same
manner as described above for the 2JPF coupling constants (Fig. 2).
Diethyl bromofluoromethylphosphonate 20 whose preparation
is described in Section 4 also exhibited non-equivalency of the two
methylene carbons of the ester group. This compound is the only
material prepared which has an asymmetric carbon contained
within the molecule giving rise to this non-equivalency. A view
along the phosphorus–bromofluoromethyl bond makes this non-
equivalence clear.
1
No correlation was found to exist for the JCF with the alkyl
substituents at phosphorus or with the Pauling electronegativity of
the halo-F-methyl group. On the other hand the 1JCP values, while
independent of the alkyl groups in the ester portion of the
molecule, do correlate nicely with the Pauling electronegativities
1
as shown in Fig. 3. The sensitivity of the JCP values to changes in
the electronegativity at carbon has been previously observed and a
rationale has been outlined to account for it [15,16]. The
magnitude of this coupling constant is directly proportional to
the degree of ‘‘s’’ character in the phosphorus–carbon bond and
increases as the ‘‘s’’ character of the bond increases.
The carbons alpha and beta to oxygen in the ester portion of the
phosphonates were also observed to couple to phosphorus while
As shown in the three rotamers, the methylene carbons are
never in identical environments and hence are magnetically non-
equivalent. The ethyl bromodifluoromethyl phenyl phosphinate 16
was also spectroscopically of interest. The four substitutents on
phosphorus render the molecule asymmetric which should lead to
non-equivalence of the CF2 fluorines. Although this non-equiva-
lence was slight (the 31P NMR spectrum of 16 showed only a
triplet), the 19F NMR spectrum was an ABX pattern as would be
expected for non-equivalent coupled fluorine atoms (i.e. a doublet
of doublets – the outer weak lines were not observed). The
observed spectrum matched the calculated spectrum very well
[17].
130
CF3
120
CF2Br
CFCl2
CF2H
110
100
CF2I
90
80
CFBr2
10
CCl3
CFBrH
2.3. Reactions of halo-F-methyl phosphonates
9
11
12
Pauling Electronegativity
From a consideration of the general structure of a halo-F-
methyl phosphonate, one can see that there are at least five
Fig. 2. 13C NMR chemical shift vs. Pauling electronegativity of halo-F-methyl group.

818
R.M. Flynn, D.J. Burton / Journal of Fluorine Chemistry 132 (2011) 815–828
Table 2
potential reactive sites within the molecule: (1) the ester carbon,
(2) phosphorus, (3) phosphoryl oxygen, (4) halo-F-methyl carbon
and (5) halogen. Reactions have been observed at all of these sites
with the exception of 4.
ROH
Difluoromethyl ethers from 4 + alkoxide, ðEtOÞ2PðOÞCF2Br þ RONaaꢀ! ROCF2H.
R
Solvent
Product
Yield (%)b
CH3
CH3OH
C2H5OH
TG
CH3OCF2H
C2H5OCF2H
CF3CH2OCF2H
CH3OCF2H
PhOCF2H
52
56
58
46
11
C2H5
CF3CH2
Ph
2.3.1. Reaction at phosphorus
CH3OH
Both fluoride anion as well as alkoxide anion were found to be
effective in attack at the phosphorus atom. The driving force for
these reactions is presumably the strength of the phosphorus–
fluorine bond [18] (117 kcal/mol) and the phosphorus–oxygen
bond (97 kcal/mol) as opposed to the phosphorus–carbon bond
(63 kcal/mol).
aSodium salt formed by reaction of an excess of alcohol in the given solvent with
sodium metal except for sodium phenoxide which was formed by the reaction of
phenol with NaOH in methanol.
b19F NMR yield vs. PhCF3 internal standard; chemical shifts of the product ethers
agree well with reported values.
Potassium fluoride reacts with
4 to cleave the carbon–
phosphorus bond and generate difluorocarbene which was readily
trapped with 2,3-dimethyl-2-butene to afford the corresponding
difluoromethylenecyclopropane in an 88% (19F NMR) yield. When
repeated on a larger scale, this cyclopropane was isolated in 63%
yield as shown in Eq. (4).
that fluoride ion would attack silicon. When this reaction was thus
carried out in the presence of 2-methyl-2-butene, the major
product as characterized by 19F NMR was Me3SiF. Only a trace of
the cyclopropane and phosphorofluoridate were observed indicat-
ing that fluoride ion did indeed predominately attack silicon. The
structure of the product silane was further confirmed by repeating
this reaction on a larger scale with isolation of the silane by glpc on
Column B followed by GC–MS.
4 + KF + (CH3)2C=C(CH3)2
+ (EtO)2POF
F
F
Alkoxide ion has also been observed to attack phosphorus in 4
with the formation of [:CF2] carbene which can be trapped in
moderate yield by the alkoxide in the presence of excess alcohol to
give the corresponding difluoromethyl ethers. Table 2 tabulates
the ethers prepared in this reaction.
81 %
88 %
(4)
The phosphorofluoridate (EtO)2POF was also observed in 81%
yield but in general these materials were not isolated since this
particular class of compounds is exceedingly toxic [19]. In a similar
manner, 2-methyl-2-butene gave the cyclopropane in a 66% glpc
Note that the primary product from the reaction of phenoxide
with
4 is difluoromethyl methyl ether rather than phenyl
difluoromethyl ether. The large excess of the solvent methanol,
which is an excellent trapping agent for [:CF2], served to scavenge
much of the difluorocarbene generated in the reaction. The co-
product in these reactions is the trialkyl phosphate. Recently
Zafrani et al. have shown that the reaction of 4 with phenols and
yield and cyclohexene yielded 7,7-difluoronorcarane in a 30% (19
NMR) yield.
F
The reaction of 10 with KF in the presence of 2,3-dimethyl-2-
butene gave the corresponding bromofluoromethylenecyclopro-
10 + KF + (CH3)2C=C(CH3)2
+ (EtO)2POF + CFBr2H
(5)
F
Br
45 %
93 %
8 %
pane in a 45% (19F NMR) yield as well as CFBr2H in an 8% (19F NMR)
yield, as illustrated in Eq. (5).
thioethers leads to good to excellent yields of the corresponding
difluoromethyl ethers or thioethers. The reaction with phenol is
shown below in Eq. (6) [20].
The reaction of 10 with KF in the presence of ethanol afforded
CFBr2H (67% yield, 19F NMR) and a 72% yield of the phosphoro-
fluoridate. However the similar reaction of 4 with KFꢁ2H2O gave
only the phosphorofluoridate with none of the anticipated
CF2BrH. These results suggest that, in the reaction of 10 with
KF, the initial step is the attack by fluoride ion on phosphorus with
MOH
(6)
4
+
ꢀ
M = Li, K
the ejection of the CFBr2 anion. In the case of 4, the attack by
fluoride does not generate the CF2Brꢀ anion as a discreet entity,
since it would be expected that this anion would be trapped by the
proton source present. It is also possible that the lifetime of the
CF2Brꢀ anion in this system is very short and that it decomposes
into bromide and difluorocarbene which then undergoes decom-
position but there is no experimental evidence to support this
hypothesis.
The reaction of fluoride ion with bis(trimethysilyl) bromodi-
fluoromethylphosphonate (prepared as discussed below) was
carried out to determine which site in the molecule would
undergo attack by fluoride. Since the typical fluorine–silicon bond
has a bond strength of about 129 kcal/mol and the typical
phosphorus–fluorine bond is 117 kcal/mol, it might be expected
2.3.2. Reaction at oxygen
Trimethylsilyl halides are known to react with alkyl phospho-
nates by attack at the phosphoryl oxygen to give good yields of the
bis(trimethylsilyl) esters with the alkyl halide as co-product
[21,22]. In phosphonates of general structure (RO)2P(O)R0, when R0
is an alkyl group and bromotrimethylsilane (Me3SiBr) is the
silylating agent, the reaction is rapid at room temperature. With a
more electron withdrawing group such as trichloromethyl, the
reaction proceeds slowly at room temperature and requires
elevated temperatures to achieve goods yields in a reasonable
time (Scheme 1).

R.M. Flynn, D.J. Burton / Journal of Fluorine Chemistry 132 (2011) 815–828
819
R'
R'
O
P
-
(RO)2
R'
+ Me3SiX
R'
[(RO)2POSiMe3]X
+
RX + ROPOSiMe3
O
+
-
[ROPOSiMe3]X
(Me3SiO)2PR'
O
OSiMe3
Scheme 1. Reaction of phosphonate with trimethylsilyl halide.
Both 4 and 10 reacted with Me3SiBr at 60 8C to give the
corresponding bis(trimethylsilyl) esters which could be distilled at
reduced pressure. These esters were hydrolytically unstable to
solvolysis by water or methanol to give the phosphonic acids. This
exchange reaction can be monitored by 31P NMR which shows the
growth of a signal intermediate in chemical shift between the
starting material and the final bis(trimethylsilyl) phosphonate
ester. After the starting material phosphonate has been completely
consumed, this intermediate signal subsequently decreases as the
product bis(trimethylsilyl) ester signal continues to increase and
can confidently be ascribed to the intermediate alkyl trimethylsilyl
ester. It was found that 10 underwent this exchange reaction at a
much faster rate than 4, giving the desired bis(trimethylsilyl)
phosphonate after 2 h heating at 60 8C while 4 required the
addition of a larger excess of trimethylsilyl bromide and heating to
60 8C for at least 24 h. In view of the mechanism above, this is not
unexpected since the positive charge which develops in the
transition state would be expected to be destabilized to a greater
extent by the adjacent CF2Br group in 4 vs. the CFBr2 group in 10.
The primary synthetic utility for these bis(trimethylsilyl) esters
is their hydrolysis to the corresponding phosphonic acid. Solvolysis
occurs rapidly and quantitatively with water or methanol.
Methanol is particularly convenient since the co-product methyl
trimethylsilyl ether can be easily removed from the phosphonic
acid product by azeotropic distillation with the excess methanol. It
was also found that simply refluxing either 4 or 10 with 50%
aqueous HCl also sufficed to prepare the phosphonic acids and this
method was generally used to make larger quantities of these
acids.
phosphine and 3 was found to be rapid and gave a 50% isolated
yield of the corresponding phosphonium salt, illustrated by Eq. (9).
Ph3P þ ðMeOÞ2PðOÞCF2Brð3Þ ! ½Ph3PMeꢂþ½O2ꢀPðOMeÞCF2Brꢂ
(9)
2.3.4. Reaction at halogen
The reaction of the halo-F-methylphosphonates with phosphite
anion proceeds by attack on halogen and has been previously
reported [25]. For the bromodifluoromethylphosphonates, positive
halogen abstraction generates the F-methylene phosphonate ylide
which undergoes subsequent acylation with the phosphoryl halide
produced in the first step and provided the first useful route to F-
methylene bis(phosphonates). F-methylene bis(phosphonates)
have subsequently been reported by Ishihara and detailed below
in Eq. (10) [26].
ðBuOÞ2PðOÞCF2Br þ ðBuOÞ2PðOÞNa ! ðBuOÞ2PðOÞCFꢀ2 Naþ
þ ðBuOÞ2PðOÞBr ! ðBuOÞ2PðOÞCF2PðOÞðOBuÞ2
(10)
2.4. Mechanism of formation of halo-F-methyl phosphonates
The reaction between triethyl phosphite and 1 has been studied
in depth and various details of the mechanistic aspects of this
reaction have been determined. The Michaelis–Arbuzov reaction of
alkyl halides with trivalent phosphorus compounds generally
proceeds by an SN2 attack of phosphorus on the halogen bearing
carbon atom followed by a dealkylation step of the intermediate
phosphonium salt as shown in Eq. (11).
ðROÞ3P þ R0X ! ½ðROÞ3PR0ꢂþXꢀ ! ðROÞ2PðOÞR0 þ RX
(11)
2.3.3. Reaction at ester carbon
Iodide: Sodium iodide is a known useful reagent for the removal
of alkyl groups, particularly benzyl, from phosphorus esters [23].
For phosphonates, generally monodealkylation occurs [24]. It was
found that sodium iodide (or potassium iodide) in acetone reacts
rapidly and quantitatively with either 4 or 10 to give the
monodealkylated product as shown below in Eq. (7).
This reaction mechanism is not observed for the reaction
between triethyl phosphite and 1. In this reaction the initial attack
by phosphorus is on halogen with the concomitant formation of
difluorocarbene [:CF2]. The carbene is subsequently trapped by
additional phosphite which is transformed through a series of
steps into the final phosphonate product.
The SN2 reaction at carbon mechanismwas eliminated as a viable
mechanistic pathway by analysis of the reactions of (EtO)3P with
CF2BrI and CF2BrCl. If a simple SN2 mechanism were operative, the
reaction of (EtO)3P with CF2BrI would be expected to yield 4 as the
sole or at least major product. In a similar manner the reaction with
CF2BrCl would be expected to yield the chlorodifluoromethyl
phosphonate. In both cases the better leaving group would be
displaced in the formation of the intermediate phosphonium salt.
The halide anion of the salt would then dealkylate the salt to give the
expected phosphonate product and alkyl halide. Examples are
shown below in Eqs. (12) and (13).
ðEtOÞ2PðOÞR þ NaI
ꢀ!
NaþOꢀ2 PðOEtÞR þ C2H5I
(7)
R¼CF2Br 4;CFBr2 10
Even when an excess of sodium iodide was employed, only the
monosodium salt of the phosphonate was observed. The salts
prepared in this manner can be converted into the corresponding
acids by passage through an ion exchange column.
Phosphines: The reaction of 4 with triphenyl phosphine in
refluxing 1,2-dimethoxyethane was found to proceed relatively
slowly according to the following Eq. (8).
Ph3P þ 4 ! ½Ph3PEtꢂþ½O2ꢀPðOEtÞCF2Brꢂ
(8)
ðEtOÞ3P þ CF2BrI ! ðEtOÞ2PðOÞCF2Br þ EtI
ðEtOÞ3P þ CF2BrCl ! ðEtOÞ2PðOÞCF2Cl þ EtBr
(12)
(13)
Acidifying this reaction mixture with HCl/ether caused the
precipitation of a tan solid which was characterized as ethyl-
triphenylphosphonium chloride. The reaction between triphenyl

820
R.M. Flynn, D.J. Burton / Journal of Fluorine Chemistry 132 (2011) 815–828
When these reactions were carried out, however, the results
phosphines with polyfluorinated methanes [28,29]. This mecha-
nism is outlined in detail in Scheme 2.
were completely opposite to these expectations. The reaction of
triethyl phosphite with CF2BrI yielded the two products 4 and 14 in
a 1 to 7.5 ratio. This ratio was also observed for the co-products
ethyl bromide and ethyl iodide. With CF2BrCl, the only product
observed in the reaction with triethyl phosphite was 4 as detailed
in Eqs. (14) and (15).
This mechanism will now be analyzed in detail. First of all as
outlined in the partial mechanism (Scheme 3) for the reaction of
CF2BrI with triethyl phosphite, the halogen which is abstracted in
the first step is the halogen which is found in the final phosphonate
product. A similar mechanism can be inferred for the reaction of
the phosphite with CF2BrCl. The triethyl phosphite would
preferentially abstract the most polarizable halogen from each
of these methanes, iodine for CF2BrI and bromine for CF2BrCl,
leading to the observed products. For CF2BrI this preference is not
exclusive since some (EtO)2P(O)CF2Br which arises from abstrac-
tion of positive bromine in the first step was also observed, but it
was the minor product.
ðEtOÞ3P þ CF2BrI ! ðEtOÞ2PðOÞCF2I þ ðEtOÞ2PðOÞCF2Br þEtI
14ð7:5Þ
4ð1Þ
þ EtBr
(14)
(15)
ðEtOÞ3P þ CF2BrCl ! ðEtOÞ2PðOÞCF2Br þ EtBr
Paths B and C in Scheme 2 can be eliminated readily. These
alternate pathways arise from consideration of another possible
fate for the intermediate halophosphonium cation formed by the
initial abstraction of bromine from CF2Br2 1. This cation is unstable
and can decompose to diethyl phosphorobromidate and ethyl
bromide as shown. In Path B, the ylide generated from more
(EtO)3P and [:CF2] abstracts bromine from the cation to give the
intermediate trialkoxyphosphonium/phosphonate ion pair which
would then be expected to collapse to 4 and diethyl ethylpho-
sphonate. This path can at best account for only a very small part of
the reaction since no diethyl ethylphosphonate was observed as a
by-product in the reaction mixture as determined by the absence
of signals for this compound in H-1, P-31 and C-13 NMR. In
addition this path consumes an additional mole of (EtO)3P, but
since yields of 4 in excess of 95% have been observed using a 1:1
stoichiometry of phosphite:1, this alternate path cannot be
significant. If triethyl phosphite can abstract positive bromine
from the phosphorobromidate, Path C arises and gives the product
A radical pathway can be similarly eliminated since the
predicted product of reaction of CF2BrI with triethyl phosphite
would again yield the opposite product mixture to that which was
observed experimentally. In other words cleavage of CF2BrI would
yield the CF2Br and I radicals. The trapping of the CF2Br radical with
phosphite and reaction of the resulting phosphoranyl radical with I
radical would give the phosphonium salt [(EtO)3PCF2Br]+Iꢀ which
would further elaborate to a product mixture which was not
observed. Furthermore when the reaction between CF2Br2 and
triethyl phosphite was carried out in the presence of ethanol, no
CF2BrH, the expected product of the reaction of CF2Br radical with
ethanol, was observed.
With the elimination of either an SN2 or free radical mechanism
as a viable pathway, a mechanism which involves nucleophilic
attack by phosphorus on halogen remains to be considered.
Nucleophilic attack by phosphorous on halogen is well-known in
the literature and has been reviewed [27]. Several examples of
halogen attack are also known in the reactions of tertiary
-
+
[(EtO)3PCF2Br]Br + (EtO)3P
(EtO)2P(O)CF2Br + EtBr
+
-
-
+
[(EtO)3PBr]Br
(EtO)3PCF2
A
+
(EtO)3P
-
(EtO)3P + CF2Br2
(EtO)3PBr + :CF2 + Br
(EtO)2P(O)Br + EtBr
-
+
(EtO)3P
(EtO)3PCF2
C
B
-
+
-
+
[(EtO)3PBr][(EtO)2P(O)]
[(EtO)3PCF2Br][(EtO)2P(O)]
:CF2
-
+
[(EtO)3PBr][(EtO)2P(O)CF2]
(EtO)2P(O)Et + (EtO)2P(O)CF2Br
(EtO)2P(O)CF2Br
(EtO)3P +
Scheme 2. Mechanism of halophilic attack.

R.M. Flynn, D.J. Burton / Journal of Fluorine Chemistry 132 (2011) 815–828
821
-
+
(EtO)3PCF2
(EtO)3P
+
-
(EtO)3P + CF2BrI
(EtO)3PI + :CF2 + Br
-
+
[(EtO)3PI]Br
+
-
(EtO)3P + [(EtO)3PCF2I]Br
(EtO)2P(O)CF2I + EtBr
Scheme 3. Reaction of CF2BrI with triethyl phosphite.
4 as shown in Scheme 2. Triethyl phosphite is known to react with
(EtO)2P(O)Cl but the reaction only occurs at a temperature greater
than 130 8C [30]. Although the bromidate would be more reactive
than the chloridate it is doubtful that the reaction would proceed at
room temperature. In addition another objection to both Paths B
and C is that the diethyl phosphite anion which is produced in
these pathways is a highly reactive species and has been found to
react rapidly with the product phosphonate 4 at room temperature
to afford a difluoromethylene(bis)phosphonate as shown below in
Eq. (16) [25].
In an attempt to intercept the key difluorocarbene intermedi-
ate, the reaction between (EtO)3P and 1 was carried out in the
presence of 2,3-dimethyl-2-butene. This NMR reaction showed an
88% yield of 4 with only a trace of the cyclopropane which suggests
that the phosphite is a much better trapping agent for the carbene
than the butene. A similar result was observed for the reaction of
triethyl phosphite with CF2BrI though in this case only the two
expected phosphonate products were observed and none of the
cyclopropane was seen.
The trapping of difluorocarbene by triethyl phosphite is a
crucial requirement for this mechanism. Middleton [32] has
reported that hexafluorothioacetone reacts rapidly with trialkyl
phosphites to afford trialkoxybis(trifluoromethyl)methylenepho-
sphoranes in good yield. The postulated mechanism for this
reaction was attack by phosphorus on sulfur to give an
intermediate which collapsed to yield bis(trifluoromethyl)carbene
and triethyl thiophosphate. The carbene was subsequently trapped
by additional phosphite to give the isolated product as detailed in
Eq. (17).
ðEtOÞ2PONa þ ðEtOÞ2PðOÞCF2Br ! ðEtOÞ2PðOÞCF2PðOÞðOEtÞ2 (16)
If any dialkyl phosphite were generated as an intermediate in
this reaction, some (bis)phosphonate would likely have been
observed by F-19 NMR analysis of the reaction mixture; however,
none was detected. It is also known that dialkyl phosphonate
anion does react rapidly with (EtO)2P(O)Cl to afford tetraethyl
hypophosphate (EtO)2P(O)P(O)(OEt)2 in moderate yield [31]. Both
of these secondary reactions would reduce the yield of the
-
+
[(RO)3PSC(CF3)2]
(RO)3P=S + [(CF3)2C:]
(RO)3P + CF3C(S)CF3
(17)
(RO)3P=C(CF3)2
observed product 4 and hence can contribute very little to the
overall mechanism.
In other work, the irradiation of a mixture of trimethyl
phosphite and bis(trifluoromethyl)diazirine gave an 18% yield of
trimethyl bis(trifluoromethyl)methylenephosphorane [33]. It is
known that the diazirine readily affords [(CF3)2C:] carbene upon
irradiation [34] and it was shown that the diazirine
and trimethyl phosphite did not react in the absence of UV
light [33]. The observed formation of the phosphorane is thus
unequivocal evidence for the trapping of [(CF3)2C:] by the
phosphite.
Since it seemed unlikely that a better trapping agent than
triethyl phosphite could be used in some other competitive
reaction, a different approach to study the capture of [:CF2] by
triethyl phosphite was employed. A readily accessible source of
difluorocarbene is [Ph3PCF2Br]+Brꢀ. Under the action of KF, this
phosphonium salt generates difluorocarbene at low temperature
which can be readily trapped by for example 2,3-dimethyl-2-
butene to give the cyclopropane [35]. In our case, under the
required conditions of the reaction, it was found that triethyl
phosphite reacted with this salt to give 4 in good yield in an
As mentioned above, when the reaction between triethyl
phosphite and 1 was carried out in the presence of ethanol in
triglyme solvent, no CF2BrH was observed. This result strongly
implied that the abstraction of positive bromine from 1 produces
difluorocarbene [:CF2] and bromide directly without the intermedi-
acy of the CF2Brꢀ anionwhich ifformedwould berapidlyprotonated
by the ethanol. In Path A of Scheme 2, the difluorocarbene produced
in this initial step is intercepted by another mole of the nucleophile
triethyl phosphite to generate the difluoromethylene ylide inter-
mediate which rapidly would be expected to abstract bromine from
the bromophosphonium cation to generate the Michaelis–Arbuzov
phosphonium salt with the regeneration of triethyl phosphite. The
normal SN2 internal attack of bromide on this salt then produces the
observed phosphonate product and ethyl bromide. The overall
stoichiometry of this reaction scheme is thus 1:1 phosphite/1. In our
hands a 1:1 ratio of phosphite/1 has led to yields in excess of 95%
consistent with this mechanistic interpretation.

822
R.M. Flynn, D.J. Burton / Journal of Fluorine Chemistry 132 (2011) 815–828
-
Scheme 2 is protonated by ethanol to give the intermediate
+
[(Me2N)3PCF2Br]Br + KF
phosphonium salt with ethoxide as counter ion. This ethoxide
anion can then attack the bromophosphonium cation with the
formation of the very strong P–O bond with release of bromide. The
newly formed mixture of salts can then decompose to the observed
products (Scheme 5).
(Me2N)3PF2 + :CF2 + 2 KBr
+
-
(EtO)2P(O)CF2H
[(EtO)3PCF2]
:CF2 + (EtO)3P
In summary, the key aspects of this mechanistic scheme for the
reaction between (EtO)3P and 1 are as follows: (1) abstraction of
positive bromine by phosphite with the concomitant formation of
[:CF2]; (2) interception of [:CF2] by phosphite to give an ylide
intermediate; (3) abstraction of bromine from the halopho-
sphonium cation by the ylide and (4) collapse of the resultant
salt by internal SN2 attack to give the observed products.
The mechanistic scheme for the reaction between trialkyl
phosphites and CFBr3 differs from that for the reaction with CF2Br2
although the initial step is the same. When the reaction between
triethyl phosphite and CFBr3 was carried out in the presence of
ethanol in triglyme solvent, it was found that 10 and CFBr2H had
been formed in NMR yields of 20% and 57% respectively with 10%
unreacted CFBr3 remaining. None of the reduced phosphonate
(EtO)2P(O)CFBrH was observed. In a control reaction it was also
found that 10 did not react with ethanol, as detailed in Eq. (20).
Scheme 4. Reaction of [(Me2N)3PCF2Br]+Brꢀ with KF.
exchange reaction which we have previously reported Eq. (18)
[36,37].
½Ph3PCF2BrꢂþBrꢀ þ ðEtOÞ3P ! Ph3P þ EtBr
þ ðEtOÞ2PðOÞCF2Br
(18)
Another phosphonium salt which is also a source of difluor-
ocarbene is bromodifluoromethyl-tris-dimethylaminophospho-
nium bromide [(Me2N)3PCF2Br]+Brꢀ. When the reaction between
this phosphonium salt, triethyl phosphite and ethanol was carried
out in triglyme, no formation of the reduced phosphonium salt
[(Me2N)3PCF2H]+Brꢀ,which would be formed by abstraction of
positive bromide from the salt followed by protonation, was
observed. Also no exchange reaction occurred between the salt and
triethyl phosphite alone. The reaction between this salt, triethyl
phosphite and KF was then carried out in triglyme solvent. Analysis
of the reaction mixture showed the presence of (Me2N)3PF2 (about
15% by NMR) and a low yield of the phosphonate (EtO)2P(O)CF2H
(16% NMR yield). Most of the remaining solid residue from the
reaction was unreacted phosphonium salt. It thus appears that
triethyl phosphite can trap difluorocarbene to yield an ylide
intermediate which in this case abstracted a proton from one of the
reagents (Scheme 4).
ðEtOÞ3P þ CFBr3 þ EtOH ! CFBr2H þ ðEtOÞ2PðOÞCFBr2
(20)
The mechanism which best fits this data is as follows in Scheme
6.
In the absence of ethanol, the initially formed phosphonium salt
undergoes a recombination to a new Michaelis–Arbuzov type salt
which then dealkylates to the observed products. In the presence of
ethanol the CFBr2ꢀ anion is captured to give CFBr2H. Since some 10
is also formed in the presence of ethanol, one may conclude that
the recombination of the initially formed salt is rapid and can
compete with protonation.
The difluoromethylene ylide in Path A of Scheme 2 is a key
intermediate. This intermediate was successfully intercepted by
carrying out the reaction between (EtO)3P and 1 in the presence of
ethanol according to the stoichiometry shown in Eq. (19).
More recent work that utilizes halo-F-methylphosphonates
includes treatment of 10 with BuLi (1:1) at low temperature to
afford by self-trapping the lithiated derivative of tetraethyl
fluoromethylenediphosphonate, which can be derivatized with
alkylating and halogenating agents or converted with high
selectivity into (E)-diethyl fluorovinylphosphonates by reaction
with carbonyl compounds [39]. Halo-F-methylphosphonate, 10,
has also been converted into diethyl-1-fluoromethylphosphonate
via a halogen–metal exchange reaction between 10 and BuLi/
ClSiMe3 at ꢀ90 8C [40]. A one-pot conversion of trifluoroacetic
2ðEtOÞ3P þ CF2Br2 þ EtOH ! 2EtBr þ ðEtOÞ2PðOÞCF2H
þ ðEtOÞ3PðOÞ
(19)
A control reaction between triethyl phosphite and ethanol did
not yield any triethyl phosphate. The reaction between triethyl
phosphite, ethanol and
1 (ratio of 1:1:1) yielded reduced
difluoromethyl phosphonate and triethyl phosphate. The reduced
phosphonate and triethyl phosphate were isolated by vacuum
distillation as an inseparable 1:1 mixture [for a yield of 43%
(EtO)2P(O)CF2H 19 and 43% (EtO)3P(O)]. These products could be
readily separated by glpc and the peak assigned to triethyl
phosphate was collected by the glpc capillary technique [38] and
analyzed by mass spectrometry which gave a spectrum identical to
that previously reported for triethyl phosphate. The reduced
phosphonate was identified by comparison of its NMR spectral
properties and glpc retention time with those of an authentic
sample. In a similar reaction using a triethyl phosphite:ethanol:1
ratio of 2:1:1, a 78% yield (glpc) of ethyl bromide was found along
with 14% unreacted 1. The ylide intermediate formed in Path A of
esters to a-fluoro-b-trifluoromethyl-b-alkoxyvinylphosphonates
has been achieved via a halogen–metal exchange reaction between
10 and 2 BuLi/ClSiMe3 at ꢀ78 8C followed by addition of the
trifluoroacetic ester [41]. The 1-lithio-1-fluoro-1-(trimethylsilyl)
methylphosphonate, derived from 10, reacts with chloroformates
to give good yields of diethyl-1-fluoromethylphosphonocarbox-
ylates. Reaction with CO2 leads to a novel synthesis of diethyl-1-
fluoromethylphosphonocarboxylic acid [42].
When 4 is treated with isopropylmagnesium chloride in THF at
low temperature, (EtO)2P(O)CF2MgCl is formed, which undergoes
reaction with strong electrophiles, such as HCl, TMSCl, halogens,
aldehydes and ketones. Product formation is strongly dependent
on reaction conditions. 2-Hydroxyphosphonates formed from the
-
+
+
-
-
+
[(EtO)3PBr]Br
[(EtO)3PCF2] + EtOH
[(EtO)3PCF2H]OEt
+
-
+
-
[(EtO)3POEt]Br
[(EtO)3PCF2H]Br
+
(EtO)2P(O)CF2H + (EtO)3PO + EtBr
Scheme 5. Protonation of ylide intermediate.

R.M. Flynn, D.J. Burton / Journal of Fluorine Chemistry 132 (2011) 815–828
823
-
-
+
+
no EtOH
[(EtO)3PCFBr2]Br
[(EtO)3PBr]CFBr2
CFBr3 + (EtO)3P
EtOH
10 + EtBr
CFBr2H + (EtO)3PO + EtBr
Scheme 6. Reaction of CFBr3 with triethyl phosphite.
reaction with aldehydes and ketones, undergo rearrangement in
the presence of base (NaH, LDA) to afford 2,2-difluoroethylpho-
sphonates without concomitant formation of 1,1-difluoroolefins
[43].
were equipped with TCD. The columns utilized in this work were as
follows: Column A was a 100 ꢃ 1/400 column packed with 20% (w/w)
Carbowax 20 M on 80-100 mesh Chromosorb P. Column B was a
100 ꢃ 1/400 column packed with 20% (w/w) SE-30 on 80-100 mesh
Chromosorb P. Infrared spectra were obtained as thin films
between NaCl plates in the case of liquid samples or as KBr pellets
for solid samples on a Beckman IR-20A Infrared Spectrometer. C, H,
3. Conclusion
The synthesis of a variety of halo-F-methylphosphonates in a
Michaelis–Arbuzov type reaction has been shown to proceed in
good yield to afford a number of previously unknown phosphonate
esters. The mechanism of formation of 4 has been shown to
proceed through the intermediacy of [:CF2] while 10 is formed via
an intermediate CFBr2ꢀ anion. The phosphonates have been found
to undergo facile reaction with a variety of reagents at several
positions in the molecule. Reagents studied include potassium
fluoride, sodium alkoxides, trimethylsilyl bromide, sodium dialkyl
phosphites, tertiary phosphines and sodium iodide.
N
analyses were carried out by service personnel of this
department on Perkin Elmer 240 (automated) Elemental
a
Analyzer. All melting points were obtained in capillaries on a
Thomas-Hoover Unimelt apparatus and are uncorrected. All bp
were determined during fractional distillation and are uncorrect-
ed. Potassium fluoride (J.T. Baker, anhydrous) was dried at 250 8C/
0.5 mm Hg for 18 h and stored in a desiccator. Alternatively, KF was
placed in an evaporating dish and heated to ꢄ450 8C with a Bunsen
burner. After breaking up of the initial crust which formed, the
finely divided salt was heated for 1 h at 450 8C and transferred to a
weighed container while still hot. Triglyme (triethylene glycol
dimethyl ether) and other glymes were purified by distillation
from a sodium benzophenone ketyl and stored in brown bottles
4. Experimental
˚
4.1. General experimental procedures
over 4 A molecular sieves. Diethylether was dried and stored over
sodium wire in a brown bottle until used. Trichlorofluoromethane
(Freon 11) and dibromodifluoromethane (Freon 12B2) were
obtained commercially and used as obtained. Trifluoroethanol
The 19F NMR spectra were recorded on a Varian HA-100
Spectrometer operated at 94.075 MHz in the HR (non-lock) mode.
The 19F NMR spectra were typically obtained as 10–15% (w/v)
solutions of a pure compound in CDCl3 or as aliquots of reaction
mixtures vs. an internal standard of CFCl3 unless otherwise noted.
Chemical shifts are reported as negative in ppm upfield of CFCl3.
Spectra were internally calibrated by the audio side band
technique. Quantitative measurements were carried out by
integration relative to an internal standard (C6H5CF3). Routine
1H NMR spectra were obtained as 10–15% (w/v) solutions of a pure
compound in CDCl3 or, occasionally, as aliquots of reaction
mixtures vs. an internal standard of TMS on a Varian A-60
analytical NMR Spectrometer. Chemical shift values are reported in
ppm downfield of TMS. 13C NMR spectra were obtained as 15% (w/
v) solutions in a suitable deuterated solvent (usually CDCl3), which
also provided the lock signal, on a Bruker HX-90E Spectrometer
operated at 22.63 MHz vs. TMS as internal standard. In all cases,
broad band proton decoupling was employed to simplify the
spectra. Chemical shift values are reported in ppm downfield from
TMS. 31P NMR spectra were obtained as 15% (w/v) solutions of a
pure compound in a suitable deuterated solvent (usually CDCl3)
which also provided the lock signal. The 31P NMR spectra were
˚
was obtained commercially and stored over 4 A molecular sieves.
Trimethyl, triethyl, tributyl and triisobutyl phosphites were
obtained from commercial vendors, distilled from sodium metal
˚
under vacuum and stored in brown bottles over 4 A molecular
sieves. Triisopropyl, triphenyl and tris(2-chloroethyl) phosphites
were obtained commercially and used as obtained. Diethylphe-
nylphosphite was prepared from Et3N, EtOH, and phenyldichlor-
ophosphine
in
59%
yield
(bp
70–738/0.3 mm Hg).
Tribromofluoromethane was prepared by the method of Birchall
and Haszeldine [44] and stored in a refrigerator over copper wire.
The salt [(Me2N)3PCF2Br]+Brꢀ was prepared as previously de-
scribed from this laboratory [45]. All other reagents were obtained
from common commercial sources and used directly.
4.2. Preparation of halo-F-methylphosphonates in diethyl ether
4.2.1. Preparation of diethyl bromodifluoromethylphosphonate 4 in
diethyl ether
A three-neck 1 l flask, fitted with a septum, stir bar, and a water
condenser with a nitrogen inlet, was cooled in an ice bath. Dry
ether (300 ml) and 99 g (0.6 mol) of triethylphosphite was added
to the cooled flask, followed by addition (via syringe) of 134 g
(0.64 mol) of dibromodifluoromethane 1. The ice bath was
removed and the reaction mixture refluxed for 24 h, followed by
removal of the ether, excess 1 and ethyl bromide via rotary
evaporation at reduced pressure. The remaining clear liquid was
distilled under vacuum through a 6-in. Vigreaux column to give
150 g (95%) of diethyl bromodifluoromethylphosphonate (bp 99–
obtained on
a Bruker HX-90E Spectrometer operated at
36.435 MHz vs. 85% H3PO4 as the external standard. In most
cases, broad band proton decoupling was employed to simplify the
spectra. Chemical shift values, reported in ppm vs. external H3PO4,
are assigned negative shifts when upfield and positive when
downfield. Mass spectra were obtained on a Hitachi-Perkin Elmer
RMU-6E Mass Spectrometer operated at 70 eV. Analytical GLPC
were carried out on either an
F & M Dual Column Gas
Chromatograph Model 700, or when it was desired to collect a
sample for mass spectral or C, H, N analysis on an F & M Dual
Column Research Chromatograph Model 5750. Both instruments
1028/16 mm Hg). 19F NMR(CDCl3) (ppm):
2JPF = 92 Hz); 31P NMR (CDCl3) (ppm):
= ꢀ1.2 (t, JPF = 93 Hz);
13C NMR (CDCl3) (ppm): = 116.8 (td, 1JCF = 330 Hz, 1JCP = 238 Hz),
d
= ꢀ61.9 (d,
2
d
d

824
R.M. Flynn, D.J. Burton / Journal of Fluorine Chemistry 132 (2011) 815–828
d
= 66.3 (2JCP = 7.4 Hz),
d
= 16.4 (d, 3JCP = 5.88 Hz); 1H NMR (CDCl3)
isolated yield (bp 85 8C/12 mm Hg). 19F NMR (CDCl3) (ppm):
3
3
2
(ppm):
d
= 4.41 (qd, JHH = 7.0 Hz, JHP = 8.5 Hz),
d
= 1.42 (t,
d
= ꢀ61.3 (d, JPF = 94 Hz); 31P NMR (CDCl3) (ppm): 1.5 (t,
3JHH = 7.0 Hz). GC–MS, m/z (relative intensity): 268 (0.2), 266
(0.2), 137 (83), 109 (100), 93 (35), 91 (27), 81 (86), 65 (41), 29 (66):
IR (P55O) 1285 cmꢀ1: Anal. Calcd. for C5H10O3PCF2Br: C, 22.47%; H,
3.75%. Found C, 22.34%; H, 3.80%.
2JPF = 94 Hz); 13C NMR (CDCl3) (ppm):
d
= 116.3 (td, JCF = 329 Hz,
1
1JCP = 240 Hz),
= 4.01 (d, 3JHP = 11 Hz). GC–MS, m/z (relative intensity): 159 (30),
109 (100), 93 (58), 81 (32), 79 (49), 59 (14), 47 (43), 15 (30); no
molecular ion observed: IR (P55O) 1285 cmꢀ1
d
= 56.1 (2JCP = 7.4 Hz); 1H NMR (CDCl3) (ppm):
d
.
4.2.2. Preparation of di-n-propyl bromodifluoromethylphosphonate 5
in diethyl ether
4.3.3. Preparation of diisopropyl bromodifluoromethylphosphonate 6
in triglyme and diethyl ether
Similar to Section 4.2.1, di-n-propyl bromodifluoromethylpho-
sphonate 5 was prepared from tri-n-propyl phosphite and 1 in 55%
isolated yield (bp 798/12 mm Hg). 19F NMR(CDCl3) (ppm):
Similar to Section 4.3.1, diisopropyl bromodifluoromethylpho-
sphonate was prepared from triisopropyl phosphite and 1 in 42%
isolated yield (bp 98 8C/10 mm Hg). 19F NMR (CDCl3) (ppm):
2
d
= ꢀ61.4 (d, JPF = 92 Hz); 31P NMR (CDCl3) (ppm):
d
ꢀ0.6 (t,
2JPF = 93 Hz); 13C NMR (CDCl3) (ppm):
d
d
d
= 116.9 (td, JCF = 329 Hz,
d
= ꢀ62.5 (d, JPF = 93 Hz); 31P NMR (CDCl3) (ppm): ꢀ2.6 (t,
1
2
1JCP = 239 Hz),
d
= 71.6 (2JCP = 7.4 Hz),
= 23.8 (d, JCP = 5.88 Hz),
2JPF = 93 Hz); 13C NMR (CDCl3) (ppm):
d
= 117.8 (td, JCF = 329 Hz,
3
1
3
d
= 9.9 (s); 1H NMR (CDCl3) (ppm):
= 4.23 (td, JHH = 6.6 Hz,
1JCP = 237 Hz), = 76.2 (2JCP = 7.4 Hz), = 24.3 (3JCP = 5.9 Hz),
d d
3JHP = 7.6 Hz),
d
= 1.73 (apparent sextet, J = 7.0 Hz),
d
= 0.97 (t,
d
= 23.7 (4JCP = 2.9 Hz); 1H NMR (CDCl3) (ppm):
d
= 4.9 (unresolved
3JHH = 7.0): IR (P55O) 1280 cmꢀ1: Anal. Calcd. for C7H14O3PCF2Br: C,
sextet, J = 6.5 Hz),
d
= 1.41 (d, JHH = 6 Hz). GC–MS, m/z (relative
4
28.47%; H,4.75%. Found C, 28.78%; H, 4.95%.
intensity): 123 (24), 109 (11), 83 (10), 45 (31), 43 (54), 42 (89), 41
(100), 40 (39), 39 (72); no molecular ion observed: IR (P55O)
1279 cmꢀ1: Anal. Calcd. for C7H14O3PCF2Br: C, 28.47%; H, 4.75%.
Found C, 29.51%; H, 5.09%.
4.2.3. Preparation of di-n-butyl bromodifluoromethylphosphonate 7
in diethyl ether
Similar to Section 4.2.1, di-n-butyl bromodifluoromethylpho-
sphonate 7 was prepared from tri-n-butyl phosphite and 1 in 65%
Similar to Section 4.2.1, diisopropyl bromodifluoromethylpho-
sphonate was prepared from triisopropyl phosphite and 1 in 75%
isolated yield using diethyl ether as the solvent with properties
identical in all respects to the phosphonate prepared in triglyme.
isolated yield (bp 107 8C/1 mm Hg). 19F NMR (CDCl3) (ppm):
d = –
2
61.5 (d, JPF = 93 Hz); 31P NMR (CDCl3) (ppm):
d
ꢀ0.6 (t,
2JPF = 93 Hz); 13C NMR (CDCl3) (ppm):
1JCP = 238 Hz), = 69.9 (2JCP = 7.4 Hz),
= 18.6 (s),
quartet, J = 7.5 Hz),
d
d
= 117.0 (td, JCF = 329 Hz,
1
3
d
= 32.5 (d, JCP = 5.89 Hz),
4.3.4. Preparation of diethyl dibromofluoromethylphosphonate 10 in
triglyme and diethyl ether
d
d
= 13.5 (s); 1H NMR (CDCl3) (ppm):
d
= 4.25 (apparent
= 0.8–1.1 (multi-
plet): IR (P55O) 1283 cmꢀ1: Anal. Calcd. for C8H18O3PCF2Br: C,
d
= 1.20–1.90 (multiplet),
d
Similar to Section 4.3.1, diethyl dibromofluoromethylpho-
sphonate was prepared from triethyl phosphite and 2 in 60%
2
33.44%; H, 5.57%. Found C, 33.64%; H, 5.93%.
isolated yield 19F NMR (CDCl3) (ppm):
d
= ꢀ76.5 (d, JPF = 77 Hz);
31P NMR (CDCl3) (ppm): 1.7 (t, JPF = 77 Hz); 13C NMR (CDCl3)
2
4.2.4. Preparation of di-isopropyl dibromofluoromethylphosphonate
11 in diethyl ether
(ppm):
(2JCP = 7.4 Hz),
= 4.40 (unresolved pentet, J = 7.0 Hz), d = 1.41 (t, JHH = 7.0 Hz).
d
= 88.4 (dd, 1JCF = 334 Hz, 1JCP = 202 Hz),
= 16.4 (3JCP = 5.2 Hz); 1H NMR (CDCl3) (ppm):
d = 66.8
d
3
Similar to Section 4.2.1, di-isopropyl dibromofluoromethylpho-
sphonate 11 was prepared from tri-isopropyl phosphite and 2 in
d
GC–MS, m/z (relative intensity): 193 (11), 191 (16), 138 (11), 137
(100), 110 (15), 109 (94), 93 (18), 91 (31), 81 (71), 65 (44), 59 (10),
47 (17), 45 (21), 29 (12); no molecular ion observed: IR (P55O)
1279 cmꢀ1
.
Similar to Section 4.2.1, diethyl dibromofluoromethylpho-
sphonate was prepared from triethyl phosphite and 2 in 78%
isolated yield using diethyl ether as the solvent with properties
identical in all respects to the phosphonate prepared in triglyme
(bp 79 8C/<1 mm Hg).
22% isolated yield (bp 818/<1 mm Hg). 19F NMR (CDCl3) (ppm):
2
d
= –77.2 (d, JPF = 80 Hz); 31P NMR (CDCl3) (ppm):
d
ꢀ0.2 (d,
2JPF = 78 Hz); 13C NMR (CDCl3) (ppm):
1JCP = 204 Hz),
d
= 89.2 (dd, JCF = 334 Hz,
1
d
= 76.1 (2JCP = 7.4 Hz),
d
= 24.2, 23.5 (signal
appeared as two doublets due to magnetically non-equivalent
3
geminal methyl groups, JCP = 5.9, 2.9); 1H NMR (CDCl3) (ppm):
3
d
= 4.95 (apparent sextet, J = 6.5 Hz),
d
= 1.44 (d, JHH = 6.0 Hz): IR
(P55O) 1271 cmꢀ1
.
4.3. Preparation of halo-F-methylphosphonates in triglyme
4.3.5. Preparation of 14 and 4 by reaction of triethyl phosphite with
CF2BrI
4.3.1. Preparation of 4 in triglyme
Similar to Section 4.3.1, diethyl iododifluoromethylphospho-
nate 14 and 4 were prepared from triethyl phosphite and CF2BrI in
60% isolated yield in a 7.5/1 ratio. The minor product was
confirmed to be 4 by spiking of the NMR sample with authentic 4
prepared as described previously. Flash distillation of the crude
reaction mixture at 10 mm Hg into a dry ice/isopropanol cooled
trap gave a liquid which was shown to be a mixture of ethyl
bromide and ethyl iodide in a 7.5/1 ratio, identified by comparison
of their glpc retention times with authentic samples. The residue
was vacuum distilled and the fraction at bp = 64–66 8C/0.5 mm Hg
Similar to Section 4.2.1 (similar apparatus), 99 g (0.6 mol) of
triethylphosphite and 134 g (0.64 mol) of 1 were stirred in
triglyme (300 ml) for 24 h at RT. Flash distillation (RT, 15 mm Hg)
removed most of the volatiles. The remaining solution was poured
into an equal volume of ice water; the upper aqueous layer
extracted with ether (2 ꢃ 100 ml) and the ether extracts combined
with the lower layer. The combined layers were washed with a
satd. NaCl solution (2 ꢃ 100 ml) to remove the triglyme. The
remaining ether solution was dried over CaSO4, the ether removed
by rotary evaporation at reduced pressure and the residue distilled
to give 88 g (55%) of 4 (bp 99–1028/16 mm Hg) – identical in all
respects to 4 prepared in diethyl ether.
identified as pure 14. 19F NMR (CDCl3) (ppm):
d
= ꢀ59.3 (d,
2
2JPF = 85 Hz); 31P NMR (CDCl3) (ppm): ꢀ2.7 (t, JPF = 86 Hz); 13C
1
1
NMR (CDCl3) (ppm):
d = 97.4 (td, JCF = 332 Hz, JCP = 219 Hz),
d
= 66.3 (2JCP = 7.4 Hz), = 16.4 (3JCP = 5.9 Hz); 1H NMR (CDCl3)
d
4.3.2. Preparation of dimethyl bromodifluoromethylphosphonate 3 in
triglyme
(ppm): d = 4.37 (unresolved pentet, J = 7.0 Hz), d = 1.41 (t,
3JHH = 7.0 Hz); GC–MS, m/z (relative intensity): 187 (29), 156
Similar to Section 4.3.1, dimethylbromodifluoromethylpho-
sphonate 3 was prepared from trimethylphosphite and 1 in 55%
(26), 137 (53), 109 (100), 93 (50), 81 (89), 65 (58), 45 (51), 29 (100);
molecular ion observed at 314: IR (P55O) 1282 cmꢀ1
.

R.M. Flynn, D.J. Burton / Journal of Fluorine Chemistry 132 (2011) 815–828
825
4.4. Preparation of halo-F-methylphosphonates ‘‘neat’’
After cooling, the autoclave was opened and the residue
concentrated on a steam bath and distilled under vacuum. After
distillation of the unreacted triethyl phosphite, a distillate fraction
was obtained (bp = 96–99 8C/8 mm Hg) which contained 15 in
approximately 30% yield. Diethyl ethylphosphonate, diethyl
phosphite and triethyl phosphate were also identified as con-
4.4.1. Preparation of dibutyl dibromofluoromethylphosphonate 12
A 25 ml one-neck round bottom flask with septum port was
equipped with a magnetic stir bar and a water condensor topped
by a glass tee leading to a source of dry nitrogen and a mineral oil
bubbler. To the flask was added tri-n-butyl phosphite (6.25 g,
0.025 mol) and 2 (6.78 g, 0.025 mol). After an induction period of
about 10 min, the flask became warm as the reaction progressed.
After stirring overnight, the residue was distilled to yield 6 g (63%)
taminants by NMR spectral data. 19F NMR (CDCl3) (ppm):
d
= ꢀ73.4
2
2
(d, JPF = 88 Hz); 31P NMR (CDCl3) (ppm):
d = 2.4 (d, JPF = 88 Hz);
13C NMR (CDCl3) (ppm):
d
= 113.4 (dd, 1JCF = 316 Hz, 1JCP = 222 Hz),
d
= 66.6 (d, 2JCP = 7.3 Hz), = 16.4 (d, 3JCP = 4.4 Hz); 1H NMR (CDCl3)
d
of dibutyl dibromofluoromethylphosphonate (bp 117–125 8C/
(ppm): d = 4.40 (apparent pentet, J = 7.0 Hz), d = 1.42 (t,
2
0.6 mm Hg). 19F NMR (CDCl3) (ppm):
d
= ꢀ75.9 (d, JPF = 78 Hz);
3JHH = 7.0 Hz),: IR (P55O) 1277 cmꢀ1
.
2
31P NMR (CDCl3) (ppm):
d
1.5 (d, JPF = 77 Hz); 13C NMR (CDCl3)
1
1
(ppm):
2JCP = 7.4 Hz),
1H NMR (CDCl3) (ppm):
= 1.20–2.0 (multiplet),
d
= 88.4 (dd, JCF = 334 Hz, JCP = 202 Hz),
d
d
= 70.4 (d,
= 13.5 (s);
4.5.2. Reaction of triethyl phosphite with CF2BrCl to give 4
Similar to Section 4.5.1, the reaction of triethyl phosphite with
CF2BrCl was carried out in a 128 ml Hastelloy C autoclave at
autogenous pressure to yield 4 in 20% yield with properties
identical in all respects to 4 as previously prepared.
3
d
= 32.5 (d, JCP = 5.9 Hz),
d
= 18.6 (s),
d
= 4.34 (apparent quartet, J = 7.0 Hz),
= 0.95 (multiplet): IR (P55O) 1275 cmꢀ1
d
d
:
Anal. Calcd. for C9H18O3PCFBr2: C, 33.44%; H, 5.57%. Found C,
33.64%; H, 5.93%.
4.5.3. Preparation of ethyl bromodifluoromethylphenylphosphinate
16
4.4.2. Preparation of diisobutyl dibromofluoromethylphosphonate 13
‘‘neat’’
Diethyl phenylphosphonite (11.6 g, 0.059 mol) and diethyl
ether (125 ml) were combined in a 250 ml round bottom flask
equipped with a water condensor topped by a glass tee leading to a
source of dry nitrogen and a mineral oil bubbler. After cooling the
flask in an ice bath, 1 (15 g, 0.07 mol) was added. The reaction
mixture was stirred overnight at ambient temperature. The ether
was removed by rotary evaporation at aspirator pressure and the
residue distilled to give 16 in 58% yield (bp = 109–111 8C/
Similar to Section 4.4.1, diisobutyl dibromofluoromethylpho-
sphonate 13 was prepared from triisobutyl phosphite and 2 in 53%
yield (bp 145 8C/1 mm Hg). 19F NMR (CDCl3) (ppm):
d
= ꢀ75.9 (d,
2
2JPF = 77 Hz); 31P NMR (CDCl3) (ppm):
d
2.3 (d, JPF = 77 Hz); 13C
1
1
NMR (CDCl3) (ppm):
d
d
= 88.2 (dd, JCF = 334 Hz, JCP = 203 Hz),
= 29.3 (d, 2JCP = 5.9 Hz),
= 76.2 (d, 2JCP = 8.1 Hz),
d
d
= 18.5 (s); 1H
= 2.0
NMR (CDCl3) (ppm):
d
= 4.10 (apparent triplet, J = 6.5 Hz), d
3
(multiplet),
d
= 1.0 (d, JHH = 6.5 Hz): IR (P55O) 1275 cmꢀ1: Anal.
0.6 mm Hg). 19F NMR (CDCl3) (ppm): ABX pattern
d
2
Calcd. for C8H18O3PCF2Br: C, 28.13%; H, 4.69%. Found C, 29.56%; H,
4.95%.
(F1) = ꢀ62.5;
d
(F2) = ꢀ62.1 (2JPF1 = 2JPF2 = 82 Hz; JF1F2 = 180 Hz);
31P NMR (CDCl3) (ppm):
d
= 20.9 (t, 2JPF = 82 Hz); 13C NMR (CDCl3)
1
(ppm):
d
= 120.2 (td, JCF = 333 Hz, 1JCP = 150 Hz),
d
= 64.5
4.4.3. Preparation of diisobutyl bromodifluoromethylphosphonate 8
‘‘neat’’
(2JCP = 7.4 Hz),
d
= 16.5 (3JCP = 5.9 Hz) and for the ring carbons,
1
d
= 123.5 (d, JCP = 146 Hz,
a
-carbon of ring),
d
= 128.9 (d,
= 133.4 (d, JCP = 10.3 Hz, o-ring
= 134 (d, 1JCP = 2.9 Hz, p-ring carbon); 1H NMR (CDCl3)
= 1.45 (t, 3JHH = 7 Hz), = 4.45 (qd, 3JHH = 7 Hz,
3JHP = 7.5 Hz),
d = 7.5–8.0 (unresolved multiplet). GC–MS, m/z
1
Similar to Section 4.4.1, diisobutyl dibromofluoromethylpho-
sphonate 13 was prepared from triisobutyl phosphite and 2 in 53%
1JCP = 13.2 Hz, m-ring carbons),
d
carbons),
(ppm):
d
d
yield (bp 145 8C/1 mm Hg). 19F NMR (CDCl3) (ppm):
d
= ꢀ75.9 (d,
d
2
2JPF = 77 Hz); 31P NMR (CDCl3) (ppm):
d
2.3 (d, JPF = 77 Hz); 13C
1
1
NMR (CDCl3) (ppm):
d
d
= 88.2 (dd, JCF = 334 Hz, JCP = 203 Hz),
= 29.3 (d, 2JCP = 5.9 Hz),
(relative intensity): 169 (57), 141 (100), 77 (64), 51 (34), 47
(15), 31 (22), 29 (34); no molecular ion observed: IR (P55O)
1252 cmꢀ1: Anal. Calcd. for C8H10O2PCF2Br: C, 36.12%; H, 3.34%.
Found C, 36.13%; H, 3.25%.
= 76.2 (d, 2JCP = 8.1 Hz),
d
d
= 18.5 (s); 1H
= 2.0
NMR (CDCl3) (ppm):
(multiplet),
d
= 4.10 (apparent triplet, J = 6.5 Hz), d
3
d
= 1.0 (d, JHH = 6.5 Hz): IR (P55O) 1275 cmꢀ1: Anal.
Calcd. for C8H18O3PCF2Br: C, 28.13%; H, 4.69%. Found C, 29.56%; H,
4.95%.
4.5.4. Preparation of bromodifluoromethyldiphenylphosphine oxide
17
4.4.4. Preparation of dimethyl dibromofluoromethylphosphonate 9
‘‘neat’’
Ethyl diphenylphosphinite (5.7 g, 0.025 mol) and diethyl
ether (30 ml) were combined in a 100 ml round bottom flask
equipped with a water condensor topped by a glass tee leading
to a source of dry nitrogen and a mineral oil bubbler. After
cooling the flask in an ice bath, 1 (7.35 g, 0.035 mol) was added.
The reaction mixture was then heated at reflux for 18 h. The
ether was removed by rotary evaporation at aspirator pressure
to yield a clear oil which crystallized on the addition of fresh
ether. The crystals were filtered and washed once with a small
portion of ice cold ether to give 5.7 g (69%) 17 (mp = 66–69 8C).
After recrystallization from benzene and again from Skelly B, the
Similar to Section 4.4.1, dimethyl dibromofluoromethylpho-
sphonate 9 was prepared from trimethyl phosphite and 2 in 46%
yield (bp 85 8C/0.4 mm Hg). 19F NMR(CDCl3) (ppm):
d
= ꢀ76.4 (d,
2
2JPF = 77 Hz); 31P NMR (CDCl3) (ppm):
d
= 3.6 (d, JPF = 77 Hz); 13C
1
1
NMR (CDCl3) (ppm):
d
d
= 87.4 (dd, JCF = 334 Hz, JCP = 204 Hz),
2
= 56.8 (d, JCP = 7.4 Hz); 1H NMR (CDCl3) (ppm):
3JHP = 11.0 Hz): IR (P55O) 1275 cmꢀ1
Anal. Calcd. for
d = 4.04 (d,
:
C2H6O3PCFBr2: C, 12.00%; H, 2.00%. Found C, 12.14%; H, 2.21%.
4.5. Preparation of other derivatives
melting point was 68.5–69.5 8C. 19F NMR (CDCl3) (ppm):
2
d
= ꢀ59.2 (d, JPF = 71 Hz); 31P NMR (CDCl3) (ppm):
d
= 27.5 (t,
= 123.0 (td,
-carbon
= 132.4 (d,
= 133.8 (d, 4JCP = 2.9 Hz,
= 7.5–8.2 (multiplet).
GC–MS, m/z (relative intensity): 201 (15), 77 (13), 39 (100);
molecular ion 330, 332; intensity very low: IR (P55O) 1222 cmꢀ1
4.5.1. Preparation of diethyl dichlorofluoromethylphosphonate 15 by
reaction of triethyl phosphite with CFCl3
2JPF = 71 Hz); 13C NMR (CDCl3) (ppm):
d
1
1JCF = 336 Hz, 1JCP = 101 Hz),
d
= 125.5 (d, JCP = 106 Hz,
a
Triethyl phosphite (10.4 g, 0.06 mol) and CFCl3 (10.3 g,
0.075 mol) were placed into a 128 ml Hastelloy C autoclave
equipped with a glass liner and magnetic stir bar. The autoclave
was sealed and heated to 120 8C for 10 days during which time the
pressure in the autoclave remained nearly constant at 59 psig.
of ring),
JCP = 10.3 Hz) (o- and m-ring carbons)],
p-ring carbon); 1H NMR (CDCl3) (ppm):
[
d
= 129.0 (d, JCP = 13.2 Hz) and
d
d
d
:

826
R.M. Flynn, D.J. Burton / Journal of Fluorine Chemistry 132 (2011) 815–828
Anal. Calcd. for C12H10OPCF2Br: C, 47.13%; H, 3.02%. Found C,
47.18%; H, 3.02%.
3JHH = 7.0 Hz): IR (P55O) 1275 cmꢀ1: Anal. Calcd. for C4H10O3PCCl3: C,
23.48%; H, 3.91%. Found C, 23.05%; H, 3.93%.
4.5.5. Preparation of dibutyl difluoromethylphosphonate 18
4.6. Reactions of halo-F-methylphosphonates
Dibutyl difluoromethylphosphonate 18 was prepared accord-
ing to the literature method by reaction of the anion of dibutyl
phosphite, generated from dibutyl phosphite and sodium metal in
dry Skelly B, with CF2HCl (bp = 122–129 8C/10 mm Hg; lit. [7]
4.6.1. Reaction of 4 with KF in the presence of 2,3-dimethyl-2-butene
To a solution of 4 (9.35 g, 0.035 mol) and 2,3-dimethyl-2-
butene (2.1 g, 0.025 mol) in dry triglyme (40 ml) was added
anhydrous KF (2.9 g, 0.05 mol). The mixture was heated to 60 8C
overnight after which time the 19F NMR spectrum indicated the
bp = 124–125 8C/12 mm Hg). 19F NMR (CDCl3) (ppm):
d
= ꢀ135.4
2
2
(dd, JPF = 90 Hz, JHP = 50 Hz); 31P NMR (CDCl3) (ppm):
d = 4.9 (t,
2JPF = 91 Hz); 13C NMR (CDCl3) (ppm):
d
= 111.7 (td, JCF = 258 Hz,
complete consumption of 4 with the formation of (EtO)2P(O)F [19
F
1
1JCP = 213 Hz),
= 18.6 (s),
2JHF = 49.0 Hz, 2JHP = 27.2 Hz),
= 0.8–1.8 (multiplet): IR (P55O) 1265 cmꢀ1
d
= 68.2 (d, JCP = 6.6 Hz),
= 13.5 (s); 1H NMR (CDCl3) (ppm):
d
= 32.6 (d, JCP = 5.9 Hz),
= 5.94 (td,
= 4.22 (apparent pentet, J = 7.0 Hz),
NMR(CDCl3) (ppm):
d
= ꢀ81.6 (d, 1JPF = 981 Hz; Lit [47]:
d
= ꢀ81.5);
2
3
3
d
d
d
1H NMR (CDCl3) (ppm):
d
= 1.40 (t, JHH = 7 Hz),
d
= 4.28 (qd,
2
d
3JHH = 7.1 Hz, JHP = 9.0 Hz)] and of 1,1-difluorotetramethylcyclo-
propane. The reaction solution was then flash distilled at
<1 mm Hg into a trap cooled in dry ice/isopropanol to yield
2.1 g (63%) 1,1-difluorotetramethylcyclopropane. 19F NMR (CDCl3)
d
.
4.5.6. Preparation of diethyl difluoromethylphosphonate 19
Similar to Section 4.5.5, diethyl difluoromethylphosphonate 19
was prepared from diethyl phosphite, sodium metal and CF2HCl in
23% yield (bp = 66–71/0.2 mm Hg).
(ppm):
d
= ꢀ148.7 (m); 1H NMR (CDCl3) (ppm):
d = 1.10 (t,
4JHF = 2.1 Hz) [Lit. [48]:
d
= 1.08 (t, JHF = 2.0 Hz)].
4
19F NMR (CDCl3) (ppm):
d
= ꢀ136.0 (dd, 2JPF = 93 Hz,
4.6.2. Small scale reaction of 4, KF and cyclohexene or 2-methyl-2-
butene
2
2JHP = 49 Hz); 31P NMR (CDCl3) (ppm):
d
= 4.8 (d, JPF = 91 Hz);
13C NMR (CDCl3) (ppm):
d
d
= 111.7 (td, 1JCF = 258 Hz, 1JCP = 213 Hz),
= 16.5 (d, 3JCP = 5.9 Hz); 1H NMR (CDCl3)
To a solution of 2-methyl-2-butene (0.53 g, 0.0075 mol) and 4
(1.34 g, 0.005 mol) in triglyme (15 ml) was added anhydrous KF
(0.6 g, 0.01 mol) and the reaction mixture stirred overnight.
Benzotrifluoride (0.24 g, 0.0017 mol) was added as an internal glpc
standard and the solution analyzed (Column A) and found to contain
d
= 64.6 (d, 2JCP = 7.4 Hz),
2
(ppm):
d
= 5.92 (td, JHF = 49.0 Hz, 2JHP = 27.5 Hz),
d
= 4.32 (appar-
3
ent pentet, J = 7.0 Hz),
d
= 1.39 (t, JHH = 6.9 Hz): IR (P55O)
1265 cmꢀ1: GC–MS, m/z (relative intensity): 173 (2), 159 (8),
138 (48), 110 (100), 92 (20), 82 (98), 66 (30), 45 (32), 29 (86); no
molecular ion observed.
1,1-difluorotrimethylcyclopropane in 66% yield by NMR [19F NMR
2
(CDCl3) (ppm):
d
= ꢀ138.9 and
d
= ꢀ150.8 (dm, JFF = 154 Hz) [Lit
[48]:
d
= ꢀ139 and
d
= ꢀ150.8 (dm, 2JFF = 150 Hz)]].
4.5.7. Preparation of diethyl bromofluoromethylphosphonate 20
A 50 ml flask with septum port was equipped with a magnetic
stir bar and a condensor topped by a glass tee leading to a source of
dry nitrogen and a mineral oil bubbler. To the flask was added
diethyl phosphite (12.9 g, 0.093 mol) and sodium metal (0.6 g,
0.026 mol). The sodium metal was consumed in a vigorous
reaction in <5 min. After cooling the solution in an ice bath,
diethyl dibromofluoromethylphosphonate 10 was slowly added by
syringe and the resulting reaction mixture stirred overnight at
ambient temperature. The reaction mixture was poured into water,
a lower fluorochemical layer separated and the water layer
extracted twice with diethyl ether. The lower fluorochemical layer
and the ether extracts were combined, washed with water and
dried over anhydrous magnesium sulfate. The ether was distilled at
atmospheric pressure and the residue distilled to give 1.5 g (23%
yield) 20 (bp = 66–71 8C/0.2 mm Hg). 19F NMR (CDCl3) (ppm):
In a similar manner, cyclohexene (0.62 g, 0.0075 mol) afforded
a 24% NMR yield of 7,7-difluoronorcarane: [19F NMR(CDCl3) (ppm):
d
= ꢀ124.8 (dt, 2JFF = 154 Hz, 4JHF = 15 Hz (cis F)) and
d
= ꢀ149.8 (d,
2JFF = 154 Hz (trans F)) [Lit [49]:
4JHF = 14 Hz)]].
d
= ꢀ129.2 (dt, JFF = 163 Hz,
2
4.6.3. Small scale reaction of 4 with KFꢁ2H2O
KFꢁ2H2O (0.94 g, 0.01 mol) was added to a solution of 4 (1.34 g,
0.005 mol) and PhCF3 (0.5 g, 0.0033 mol) in triglyme (15 ml) and
the mixture stirred overnight at room temperature. 19F spectral
analysis showed that (EtO)2P(O)F had been formed in an 83% yield.
No CF2BrH was observed.
4.6.4. Small scale reaction of 10 with anhydrous KF and ethanol
KF (0.6 g, 0.01 mol) was added to a solution of 10 (1.64 g,
0.005 mol), PhCF3 (0.24 g, 0.0017 mol) and ethanol (0.46 g,
0.01 mol) in triglyme (15 ml) and the mixture stirred overnight
at room temperature. 19F spectral analysis showed that CFBr2H
d
= ꢀ166.9 (dd, 2JPF = 77 Hz, 2JHF = 50 Hz); 31P NMR (CDCl3) (ppm):
2
d
= 7.5 (d, JPF = 75 Hz); 13C NMR (CDCl3) (ppm):
d
= 84.1 (dd,
1JCF = 265 Hz, JCP = 191 Hz),
d
= 65.4 and 64.8 (signal appears as
was formed in 67% yield [19F NMR (CDCl3) (ppm):
d
= ꢀ83.8 (d,
1
2
two doublets with identical coupling constants, see discussion);
2JHF = 49.9 Hz) [Lit [50]:
d
= ꢀ84.6 (d, JHF = 51 Hz)]]. Diethyl
(2JCP = 7.4 Hz),
d
= 16.5 (3JCP = 5.2 Hz); 1H NMR (CDCl3) (ppm):
phosphorofluoridate was formed in 72% yield.
3
3
3
d
= 4.38 (qd, JHH = 7.0 Hz, JHP = 3.0 Hz),
d
= 1.40 (t, JHH = 7.0 Hz),
2
2
d
= 6.51 (dd, JHF = 47.5 Hz, JHP = 10.4 Hz): IR (P55O) 1255 cmꢀ1
:
4.6.5. Small scale reaction of 10 with anhydrous KF and 2,3-dimethyl-
2-butene
Anal. Calcd. for C4H10O3PCFBrH: C, 24.10%; H, 4.42%. Found C,
24.16%; H, 4.49%.
KF (2.6 g, 0.045 mol) was added to a solution of 10 (2.8 g,
0.0085 mol) and 2,3-dimethyl-2-butene (1.41 g, 0.017 mol) in
triglyme (18 ml) and the mixture stirred overnight at room
temperature. PhCF3 (0.41 g, 0.0028 mol) was added and 19F
spectral analysis showed that the solution contained (EtO)2P(O)F
(93%), CFBr2H (8%), 1-bromo-1-fluorotetramethylcyclopropane
4.5.8. Preparation of diethyl trichloromethylphosphonate 21
Diethyl trichloromethylphosphonate 21 was prepared via the
literature procedure [46] by reaction of carbon tetrachloride(175 ml)
with triethyl phosphite (33.0 g, 0.2 mol) at reflux over 24 h. The
excess carbon tetrachloride was removed by rotary evaporation at
aspirator pressure and the residue distilled to yield 47.2 g (92%)
product (bp = 135 8C/20 mm Hg: [46] 135–137 8C/16 mm Hg). 31P
(45%) and an unknown signal at
probably a ring-opened product of the cyclopropane.
d
= ꢀ112.7 (30%) which is
NMR (CDCl3) (ppm):
d
= 5.4 (s); 13C NMR (CDCl3) (ppm):
d
= 88.8 (d,
4.6.6. Reaction of 4 with sodium methoxide in methanol
Sodium metal (0.46 g, 0.02 mol) was added to anhydrous
methanol (20 ml) and allowed to dissolve completely. After cooling
1JCP = 197 Hz),
(CDCl3) (ppm):
d
= 67.0 (2JCP = 5.9 Hz),
d
= 16.4 (3JCP = 5.9); 1H NMR
3
3
d
= 4.40 (qd, JHH = 7.0 Hz, JHP = 8.5 Hz),
d
= 1.41 (t,

R.M. Flynn, D.J. Burton / Journal of Fluorine Chemistry 132 (2011) 815–828
827
the solution in an ice bath, 4 (2.67 g, 0.01 mol) was added dropwise
phosphonic acid as a clear, viscous oil. With the methanol
solvolysis, the silyl ester was treated with a large excess of
anhydrous methanol and the mixture stirred for an hour followed
by an atmospheric distillation to remove the Me3SiOMe until the
distillation head reached a temperature of 64 8C. The remaining
methanol was then removed by rotary evaporation.
by syringe. PhCF3 (0.58 g, 0.004 mol) was added and 19F spectral
analysis showed that the solution contained CH3OCF2H [19F NMR
2
(CDCl3) (ppm):
d
= ꢀ86.6 (d, JHF = 77 Hz) [Lit [51]:
d
= ꢀ88.2 (d,
2JHF = 74 Hz)]] in 52% yield.
4.6.7. Reaction of 4 with sodium ethoxide in ethanol
For the larger scale syntheses of (HO)2P(O)CF2Br and
(HO)2P(O)CFBr2, it was found that hydrolysis using aqueous HCl
was more efficient as described below.
In a similar manner ethyl difluoromethyl ether was prepared in
56% NMR yield. [19F NMR(CDCl3) (ppm):
d
= ꢀ83.9 (d, 2JHF = 76 Hz)
2
[Lit [28]:
d
= ꢀ83.5 (d, JHF = 74 Hz)]].
4.6.13. Preparation of (HO)2P(O)CF2Br
4.6.8. Reaction of 4 with sodium trifluoroethoxide in triglyme solvent
Sodium metal (0.23 g, 0.01 mol) was added to a solution of
trifluoroethanol (2.8 g, 0.028 mol) in triglyme (20 ml) and allowed
to dissolve completely. Subsequently 4 (1.27 g, 0.0048 mol) was
added dropwise by syringe. 19F spectral analysis as above showed
that the solution contained CF3CH2OCF2H in 58% yield [19F NMR
50% aqueous hydrochloric acid (70 ml) was added to 4 (4.5 g,
0.017 mol) and the mixture heated to reflux for about 16 h. The
bulk of the water was removed by rotary evaporation at reduced
pressure and the residue was subsequently dried under high
vacuum.
[
19F NMR (acetone-d6/water) (ppm):
d
= ꢀ61.0 (d,
2
2JPF = 87 Hz); 31P NMR (ppm):
d
= ꢀ3.1 (t, JPF = 89 Hz); 13C NMR
1
1
(CDCl3) (ppm):
d
= ꢀ74.5 (tt, 3JHF = 9.6 Hz, 5JFF = 2.4 Hz);
d
= ꢀ84.6
(ppm):
d
= 120.4 (dt, JCF = 328 Hz, JCP = 228 Hz)].
2
5
3
(dq, JHF = 75 Hz, JFF = 2.4 Hz) [Lit [51]:
d
= ꢀ75.6 (t, JHF = 8 Hz);
2
5
d
= ꢀ87.1 (dq, JHF = 72 Hz, JFF = 2 Hz)]].
4.6.14. Preparation of (HO)2P(O)CFBr2
50% aqueous hydrochloric acid (70 ml) was added to 10 (9.8 g,
0.03 mol) and the mixture heated to reflux for about 16 h. The bulk
of the water was removed by rotary evaporation at reduced
pressure and the residue was subsequently dried under high
vacuum whereupon the acid crystallized to give an off-white solid
4.6.9. Reaction of 4 with sodium phenoxide in methanol solvent
Sodium hydroxide (0.4 g, 0.01 mol) and phenol (0.94 g,
0.01 mol) were added to anhydrous methanol. After complete
solution had occurred, the solution was cooled in an ice bath and 4
(2.67 g, 0.01 mol) was added dropwise by syringe. PhCF3 (0.58 g,
0.004 mol) was added and 19F spectral analysis showed that the
solution contained a mixture of CH3OCF2H (46%) and PhOCF2H
which rapidly absorbed moisture from the air. 19F NMR (D2O)
2
(ppm):
d
= ꢀ73.7 (d, JPF = 70 Hz); 31P NMR (ppm):
d
= ꢀ2.5 (t,
2JPF = 71 Hz); 13C NMR (acetone-d6) (ppm):
1JCF = 331 Hz, JCP = 196 Hz); Anal. Calcd. for CH2Br2FO3P. H2O: C,
d = 92.5 (dd,
(11%) [19F NMR (CDCl3) (ppm):
2JHF = 73.9 Hz]].
d
= ꢀ80.9 (d, 2JHF = 75 Hz) [Lit [51]:
1
4.14%; H, 1.38%. Found C, 4.08%; H, 1.45%.
4.6.10. Reaction of 4 with bromotrimethylsilane
4.6.15. Reaction of 4 with sodium iodide
Bromotrimethylsilane (18.4 g, 0.12 mol) was added to 4 (13.6 g,
0.05 mol) and the mixture stirred under nitrogen at 60 8C for 18 h.
The reaction mixture was then distilled under vacuum to afford
Sodium iodide (7.5 g, 0.05 mol) was added to dry acetone
(30 ml) and the mixture stirred until the solution was homoge-
neous. To this mixture was then added 4 (13.3 g, 0.05 mol) and the
solution heated at reflux for 3 h. The acetone was removed by
rotary evaporation to afford 12.8 g (98%) of an off-white solid
identified as Na+ [O2P(OEt)CF2Br]ꢀ. The melting point of the salt,
following recrystallization from benzene/ethyl acetate was 227–
17 g (92% yield) (Me3SiO)2P(O)CF2Br (bp 82–84 8C/1.2 mm Hg).
2
[
19F NMR (CDCl3) (ppm):
d
= ꢀ63.0 (d, JPF = 96 Hz); 1H NMR
(CDCl3) (ppm):
d
= 0.38 (s); 31P NMR (CDCl3) (ppm):
d
= ꢀ16.1 (t,
2JPF = 96 Hz)]. In a separate experiment it was found that a signal
could be observed in the 31P NMR which was intermediate
between the starting material and the final silylated product at
230 8C with decomposition.
[
19F NMR (acetone-d6) (ppm):
2
d
= ꢀ56.6 (d, JPF = 76 Hz); 1H NMR (DMSO-d6) (ppm):
d = 1.16
d
= ꢀ7.2 ppm (t, 2JPF = 94 Hz). This was presumably the mixed ester
(t, 3JHH = 7 Hz),
d
= 3.92 (apparent pentet, J = 7 Hz); 31P NMR (D2O)
2
EtOP(O)(CF2Br)(OSiMe3).
(ppm):
d
d
= ꢀ1.88 (t, JPF = 82 Hz); 13C NMR (acetone-d6) (ppm):
= 124.2 (dt, 1JCF = 334 Hz, 1JCP = 201 Hz), = 17.0 (d 3JCP = 5.9 Hz),
d
2
4.6.11. Reaction of 10 with bromotrimethylsilane
d = 63.7 (d, JCP = 7.3 Hz)]; Anal. Calcd. for C3H5BrF2NaO3P: C,
Bromotrimethylsilane (0.76 g, 0.005 mol) was added to 10
(0.49 g, 0.0015 mol) in a 5 mm NMR tube equipped with a wired-
on septum cap and the tube heated to 60 8C for 90 min. At the end
of this time the excess silane and co-product ethyl bromide were
13.79%; H, 1.92%. Found C, 13.59%; H, 2.01%.
4.6.16. Reaction of 10 with sodium iodide
Sodium iodide (3.34 g, 0.022 mol) was added to dry acetone
(about 30 ml) and the mixture stirred until the solution was
homogeneous. To this mixture was then added 10 (7.3 g,
0.022 mol) and the solution heated at reflux for 3 h. The acetone
was removed by rotary evaporation to afford 4.6 g (64%) of Na
[O2P(OEt)CFBr2]ꢀ. The salt was recrystallized from acetone/
removed under vacuum. [19F NMR (CDCl3) (ppm):
d
= ꢀ76.3 (d,
2JPF = 81 Hz); 1H NMR (CDCl3) (ppm):
d
= 0.38 (s); 31P NMR (CDCl3)
(ppm):
d
= ꢀ13.2 (d, 2JPF = 79 Hz)]. In a separate experiment it was
found that a signal could be observed in the 31P NMR which was
intermediate between the starting material and the final silylated
+
2
product at
EtOP(O)(CFBr2)(OSiMe3).
d
= ꢀ4.5 ppm (d). This was presumably the mixed ester
benzene (1/5). [19F NMR (D2O) (ppm):
d
= ꢀ72.2 (d, JPF = 68 Hz);
3
1H NMR (D2O) (ppm):
d
= 2.14 (t, JHH = 7 Hz),
d
= 5.02 (apparent
pentet, J = 7 Hz); 31P NMR (D2O) (ppm):
NMR (acetone-d6/H2O) (ppm):
d C
= 2.4 (d, 2JPF = 69 Hz); 13
4.6.12. Hydrolysis reactions of (Me3SiO)2P(O)CF2Br
d
= 98.6 (dd, 1JCF = 337 Hz,
3
2
(Me3SiO)2P(O)CF2Br was readily converted to the phosphonic
acid (HO)2P(O)CF2Br as expected by using either aqueous
hydrolysis or an exchange reaction with anhydrous methanol. In
the aqueous hydrolysis method, the silyl ester was treated with a
large excess of water at room temperature for 24 h. The initial two
phase solution became homogeneous during this time. The
solution was then extracted with chloroform and the remaining
aqueous layer distilled by rotary evaporation to yield the
1JCP = 166 Hz),
d
= 17.1 (d, JCP = 7 Hz),
d
= 64.6 (d, JCP = 7 Hz)];
Anal. Calcd. for C3H5Br2FNaO3P: C, 11.18%; H, 1.55%. Found C,
11.17%; H, 1.10%.
4.6.17. Reaction of 4 with triphenylphosphine
Triphenylphosphine (10.5 g, 0.04 mol) and dry dimethox-
yethane (50 ml) were placed into a flask equipped with a magnetic
stirrer, septum port and condensor protected from moisture with a

828
R.M. Flynn, D.J. Burton / Journal of Fluorine Chemistry 132 (2011) 815–828
calcium chloride drying tube. Into this solution was syringed 4
(10.7 g, 0.04 mol) and the mixture heated to reflux for four days.
After the reaction mixture was cooled to room temperature, a
solution of HCl in dry diethyl ether (4.14 M, 9.7 ml, 0.04 mol) was
syringed into the flask. The tan solid which precipitated was
filtered, washed with ether and dried under vacuum to afford 2.5 g
(20%) of a solid identified as [Ph3PC2H5]+Clꢀ (mp = 234–236.5 8C,
Lit [52]: 234–236 8C).
peaks corresponding to 1 and EtBr in a 0.97:1 ratio. Correcting for
the difference in thermal conductivity, this mixture corresponded
to a yield of 78% EtBr with 14% unreacted 1.
Acknowledgement
RMF thanks the Graduate College of the University of Iowa for
financial support.
4.6.18. Reaction of 3 with triphenylphosphine
To a solution of triphenylphosphine (15.6 g, 0.025 mol) in dry
diethyl ether (25 ml) was added 3 (5.2 g, 0.025 mol) and the
solution brought to reflux. The white precipitate which formed
rapidly was filtered and washed with several portions of ether to
afford 10 g (50%) of a solid [Ph3PMe]+[O2P(OMe)CF2Br]ꢀ which was
recrystallized from benzene mp = 116–117 8C. [19F NMR (ppm):
References
[1] D.J. Burton, Z.Y. Yang, W. Qiu, Chem. Rev. 96 (1996) 1641–1715.
[2] L.D. Freedman, G.O. Doak, Chem. Rev. 57 (1957) 479–523.
[3] D.J. Burton, R.M. Flynn, J. Fluorine Chem. 10 (1977) 329–332.
[4] H.J. Emeleus, J. Chem. Soc. (1955) 563–574;
H.J. Emeleus, J. Chem. Soc. (1954) 3896–3904;
H.J. Emeleus, J. Chem. Soc. (1954) 3598–3603.
d
= ꢀ55.7 (d, 2JPF = 71 Hz); 1H NMR (ppm):
d
= 3.15 (d, 2JHP = 10 Hz,
= 7.5–8.0 (m); 31P NMR
= 9.5
= 199.3 (d,
d d
= 3.61 (d, 3JHP = 10 Hz, CH3O–P),
d
[5] R.A. McIvor, C.E. Hubley, Can. J. Chem. 37 (1959) 869–876.
[6] E. Gryszhiewicz-Trochimowski, Bull. Soc. Chim. Fr. (1967) 4289.
[7] L. Soborovskii, N. Baina, J. Gen. Chem. USSR 29 (1959) 1115.
[8] R.G. Harvey, E.R. De Sombre, Topics in Phosphorus Chemistry, vol. 1, Interscience,
New York, 1964, p. 57.
CH3–P),
(ppm):
= ꢀ0.7 (t, 2JPF = 70 Hz),
d d
= 22 (s); 13C NMR (ppm):
(d, 1JCP = 58 Hz),
d
= 54.0 (d, 2JCP = 7.35 Hz),
d
1JCP = 90 Hz),
= 135.1 (s),
d
= 130.5 (d, JCP = 13 Hz),
d
= 133.2 (d, JCP = 10 Hz),
3
2
[9] A. Isbell, CA 58 (1963) 11394f.
[10] G. Kamai, CA 46 (1952) 6081.
1
1
d
d = 124.7 (dt, JCF = 337 Hz, JCP = 187 Hz)].
[11] T. Mahmood, J.M. Shreeve, Synth. Commun. 17 (1987) 71–75.
[12] D.J. Burton, R.M. Flynn, Synthesis (1979) 615.
4.7. Reactions in support of the reaction mechanisms
[13] A.L. Allred, J. Inorg. Nucl. Chem. 17 (1961) 215–221.
[14] M.J. Van Hamme, D.J. Burton, P. Greenlimb, Org. Magn. Res. 11 (1978) 275.
[15] T.A. Albright, W.J. Freeman, E.E. Schweitzer, J. Am. Chem. Soc. 97 (1975) 2942–
2946.
[16] G.A. Gray, J. Am. Chem. Soc. 95 (1973) 7736–7742.
[17] The 19F NMR spectrum was calculated and plotted utilizing programs available at
the University of Iowa from Dr. W.E. Bennett.
4.7.1. Reaction of [(Me2N)3PCF2Br]+Brꢀ, KF and (EtO)3P
The salt [(Me2N)3PCF2Br]+Brꢀ (5.4 g, 0.014 mol) was suspended
in triglyme (30 ml) and (EtO)3P (2.4 g, 0.014 mol) syringed into the
mixture. To this mixture was added anhydrous KF (0.84 g,
0.014 mol) and the mixture stirred for 18 h at room temperature.
[18] J. Emsley, D. Hall, The Chemistry of Phosphorus, vol. 35, John Wiley and Sons, New
York, 1976.
[19] B.C. Saunders, Adv. Fluorine Chem. 2 (1961) 188.
1
19F NMR analysis showed
d
= ꢀ52.9 (d, JPF = 707 Hz) for the
aminophosphorane (Me2N)3PF2 and a doublet of doublets assigned
to diethyl difluoromethylphosphonate 19 and confirmed by the
addition of an authentic sample as prepared in Section 4.5.6. The
yield of 19 was 16% using benzotrifluoride as an internal NMR
standard.
[20] Y. Zafrani, G. Sod-Moriah, Y. Segali, Tetrahedron 65 (2009) 5278–5283.
[21] C.E. McKenna, M.T. Higa, N.H. Cheung, M. McKenna, Tetrahedron Lett. 18 (1977)
155–158.
[22] R. Rabinowitz, J. Org. Chem. 28 (1963) 2975–2978.
[23] V.M. Clark, A.R. Todd, J. Chem. Soc. (1950) 2030–2034.
[24] K.A. Petrov, N.K. Bliznyuk, T.N. Lysenko, J. Gen. Chem. USSR 30 (1960) 1943.
[25] D.J. Burton, R.M. Flynn, J. Fluorine Chem. 15 (1980) 263–266.
[26] D.J. Burton, T. Ishihara, R.M. Flynn, J. Fluorine Chem. 20 (1982) 121–126.
[27] B. Miller, Topics in Phosphorus Chemistry 2, Interscience, New York, 1965, p. 133.
[28] H.S. Kesling, Ph.D. Thesis, Univ. of Iowa, 1975.
[29] (a) R.N. Haszeldine, B.O. West, J. Chem. Soc. (1956) 3631–3637;
(b) M.J. Van Hamme, Ph.D. Thesis, Univ. of Iowa, 1974;
(c) H. Teichmann, Z. Chem. 14 (1974) 216.
[30] M. Baudler, W. Giese, Z. Anorg. Chem. 290 (1957) 258–269.
[31] J. Michalski, A. Zwierzak, Proc. Chem. Soc. (1964) 80.
[32] W.J. Middleton, W.H. Sharkey, J. Org. Chem. 30 (1965) 1384–1390.
[33] Y. Inouye, Ph.D. Thesis, Univ. of Iowa, 1978.
[34] D.M. Gale, W.J. Middleton, C.G. Krespan, J. Am. Chem. Soc. 88 (1966) 3617–3623.
[35] D.J. Burton, D.G. Naae, J. Am. Chem. Soc. 95 (1973) 8467–8468.
[36] D.J. Burton, D.G. Naae, R.M. Flynn, B.E. Smart, D.R. Brittelli, J. Org. Chem. 48 (1983)
3616–3618.
[37] R.M. Flynn, D.J. Burton, D.M. Wiemers, J. Fluorine Chem. 129 (2008) 583–589.
[38] K. Burson, C. Kenner, J. Chromatogr. Sci. 7 (1969) 63.
[39] B. Iorga, F. Eymery, P. Savignac, Tetrahedron Lett. 39 (1998) 4477–4480.
[40] R. Waschbusch, J. Carran, P. Savignac, C.R. Acad Sci. Paris, Ser. IIc 1 (1998) 49–52.
[41] Y. Shen, Y. Zhang, J. Fluorine Chem. 108 (2001) 69–71.
[42] R. Waschbusch, J. Carran, P. Savignac, Tetrahedron 53 (1997) 6391–6400.
[43] R. Waschbusch, M. Samadi, P. Savignac, J. Organomet. Chem. 529 (1997) 267–278.
[44] J.M. Birchall, R.N. Haszeldine, J. Chem. Soc. (1959) 13.
[45] D.G. Naae, Ph.D. Thesis, Univ. of Iowa, 1972.
4.7.2. Reaction between (EtO)3P, 1 and EtOH
Ethanol (2.3 g, 0.05 mol), 1 (12.6 g, 0.05 mol) and triethyl
phosphite (8.31 g, 0.05 mol) were combined with dimethox-
yethane solvent (20 ml) in a round bottom flask equipped with
a reflux condensor under dry nitrogen and the clear solution
brought to reflux for about 16 h. The low boiling materials were
removed by distillation at atmospheric pressure and the residue
distilled under vacuum to yield 7.9 g (86%) of a 1:1 inseparable
mixture of 19 and triethyl phosphate (bp = 93–99 8C/8 mm Hg). 19
was identified by comparison of its spectral data with an authentic
sample. The triethyl phosphate was separated by glpc (Column B)
and the peak assigned to the phosphate collected and analyzed by
GC–MS [GC–MS, m/z (relative intensity): 182(7, M+), 155 (71), 127
(45), 125 (18), 109 (47), 99 (100), 82 (37), 81 (62), 29 (56)] which
was in good agreement with the literature data [53].
In a similar reaction but using a stoichiometry of 2:1:1 for
(EtO)3P:1:EtOH, ethanol (1.15 g, 0.025 mol), 1 (6.3 g, 0.025 mol)
and triethyl phosphite (8.31 g, 0.05 mol) were combined with
[46] G. Kosolapoff, J. Am. Chem. Soc. 69 (1947) 1002–1003.
[47] C.H. Dungan, J.R. Van Wazer, Compilation of Reported 19F NMR Chemical Shifts,
Wiley-Interscience, New York, 1970.
[48] D. Seyferth, J.Y. Mui, M.E. Gordon, J.M. Burlitch, J. Am. Chem. Soc. 87 (1965) 681–
682.
[49] G.A. Wheaton, D.J. Burton, J. Fluorine Chem. 9 (1977) 25–44.
[50] R.W. Vander Haar, Ph.D. Thesis, Univ. of Iowa, 1973.
[51] R.A. Mitsch, J.E. Robertson, J. Heterocycl. Chem. 2 (1965) 152.
[52] H.J. Bestman, B. Arnason, Chem. Ber. 95 (1962) 1513.
[53] F.W. McLafferty, Anal. Chem. 28 (1956) 306–316.
dimethoxyethane solvent (20 ml) in
a round bottom flask
equipped with a reflux condensor under dry nitrogen and the
clear solution brought to reflux for about 16 h. The reaction
mixture was distilled at atmospheric pressure and the distillate
boiling up to 84 8C collected. Salt water (25 ml) was added to the
distillate and the lower layer separated and dried over anhydrous
sodium sulfate. Glpc analysis of the distillate (5.0 g) showed two