On the isolation of neat allylic fluorides

Article abstract of DOI:10.1016/j.jfluchem.2009.02.012

Neat cinnamyl fluoride, geranyl fluoride, 5-carbomethoxy-3-fluorocyclohexene and parent 3-fluorocyclohexene undergo spontaneous decomposition on contact with borosilicate glass or catalytic quantities of moderately strong acids that consists in polycondensation with elimination of HF initiated by electrophilic abstraction of F^-. Acid sensitivity of these allylic fluorides correlates with stability of respective allylic cations, exceeds that of parent benzyl fluoride, yet can be mitigated by the use of Teflon and PFA containers that permit their isolation and handling in the neat state.

Full text of DOI:10.1016/j.jfluchem.2009.02.012

Journal of Fluorine Chemistry 130 (2009) 474–483
Contents lists available at ScienceDirect
Journal of Fluorine Chemistry
journal homepage: www.elsevier.com/locate/fluor
On the isolation of neat allylic fluorides
*
Eunsung Lee, Dmitry V. Yandulov
Department of Chemistry, Stanford University, Stanford, CA 94305-5080, United States
A R T I C L E I N F O
A B S T R A C T
Article history:
Neat cinnamyl fluoride, geranyl fluoride, 5-carbomethoxy-3-fluorocyclohexene and parent 3-
fluorocyclohexene undergo spontaneous decomposition on contact with borosilicate glass or catalytic
quantities of moderately strong acids that consists in polycondensation with elimination of HF initiated
by electrophilic abstraction of Fꢀ. Acid sensitivity of these allylic fluorides correlates with stability of
respective allylic cations, exceeds that of parent benzyl fluoride, yet can be mitigated by the use of Teflon
and PFA containers that permit their isolation and handling in the neat state.
Received 23 January 2009
Received in revised form 12 February 2009
Accepted 13 February 2009
Available online 3 March 2009
Keywords:
ß 2009 Elsevier B.V. All rights reserved.
Cinnamyl fluoride
Geranyl fluoride
Cyclohexenyl fluoride
Allylic fluorides
Glass-mediated decomposition
1. Introduction
converted to gelatinous forms of variable coloration, with complete
transformation taking seconds after a sudden onset and resulting in
Organofluorine compounds continue to enjoy diverse applica-
tions in areas across the board from materials science to
biomedical fields and provide a constant need for the development
and improvement of methods of organofluorine synthesis [1].
Allylic fluorides constitute an important class of fluorinated
intermediates due to the versatility with which allylic fluoride
moiety can be constructed and elaborated on, including asym-
metric transformations [2]. Against this background, we were
struck to find the essential representatives of allylic fluorides
family, viz. cinnamyl fluoride and geranyl fluoride, to suffer
spontaneous decomposition in minutes after neat materials are
isolated in common laboratory glass containers. We report here
such decomposition to bear limiting stoichiometry of dehydro-
fluorination, present evidence of acid-catalyzed polycondensation
and offer simple solutions to isolation and handling of small
quantities of such materials, including parent and substituted
cyclohexenyl fluorides.
etching of the borosilicate glass container. We were fortunate to find
containers made of soda lime glass to impart sufficient stability to
neat 1 and this allowed us to sample common laboratory materials
for compatibility with 1, attempt to identify the cause of the
observed decomposition and seek means to prevent it. Fig. 1
summarizes the results of materials compatibility screen (Table 1).
Without exception, neat 1 decomposed on contact with all
borosilicate glass containers examined at RT under N2, albeit over
irreproducible times in any given material (average of six runs
exceeded three standard deviations in only one case). Contrasting
behavior of soda lime glass, which contains minimal amount of B2O3
and has not once been seen to cause decomposition of 1 at RT under
N2, suggested Lewis-acidic surface sites, such as three-coordinate
boron [5], to be responsible for the unexpected reactivity. According
to 1H and 19F NMR (Fig. 2), the gelatinous decomposition residues
formed from neat 1 on contact with borosilicate glass at RT under N2
contained significant amounts of organofluorine compounds devoid
of 1 and dominated by dimers 3, 30 (Chart 1, dr = 80:20, >40% of
organic F). The latter were isolated in 38% combined yield from
similar mixtures formed in two analogous runs from neat 1 in a PFA
container over several days at RT following a brief exposure to air.
With connectivities assigned from extensive JFH data, relative
configurations of 3 and 30 were assigned by correlating greater
relative shielding of C(3)H2 and C(F)H2 protons observed in the 1H
NMR spectra of 3 and 30, respectively, to the proximity of C(5)Ph ring
in the dominant staggered conformation around C4–C5 bond,
identified with reasonable consistency by molecular mechanics and
semiempirical methods (Chart 1).
2. Results and discussion
Notwithstanding several reported preparations of cinnamyl
fluoride (1) [3,21] and geranyl fluoride (2) [3b,4] listing isolated
yields, in our hands colorless, mobile liquids obtained initially upon
distillation of these allylic fluorides within an hour inevitably
* Corresponding author. Tel.: +1 650 725 9651; fax: +1 650 725 0259.
E-mail address: yandulov@stanford.edu (D.V. Yandulov).
0022-1139/$ – see front matter ß 2009 Elsevier B.V. All rights reserved.
doi:10.1016/j.jfluchem.2009.02.012

E. Lee, D.V. Yandulov / Journal of Fluorine Chemistry 130 (2009) 474–483
475
Table 1
Nominal composition of glasses (w/w, %) tested in Fig. 1.
Soda limea
Kimax1 N-51A
Pyrex1 7740
SiO2
B2O3
Na2O
Al2O3
K2O
68.7
<0.02
13.3
2.7
72
12
6
81
13
4
7
2
3.0
2
CaO
10.1b
2.0
1
BaO
a
Composition quoted for the specific material tested.
CaO + MgO.
b
The reactivity of 1 was examined more closely in solution
(Table 2). Thermally stable and resistant to bases, 1 reacted with
catalytic quantities of various Brønsted and Lewis acids to yield
product mixtures resembling those obtained from neat 1 by 1H and
19F NMR (Fig. 2) and featuring 3 in 40:60–100:0 dr [6] and up to
37% yield at full conversion. Activity of Brønsted acids, gauged
qualitatively by the time to complete conversion, generally
Fig. 1. Stability of neat 1 in eight different glass (Pyrex1, N-51A, soda lime: Table 1)
and one PFA container at RT under N2, unless otherwise indicated. Each horizontal
bar spans the range of times to decomposition onset observed in six individual runs
with the same material (3 for 2 in PFA) and marked with vertical bars.
Chart 1.
Fig. 2. (a) 1H (400 MHz, 22 8C) and (b) 19F (376 MHz, 22 8C) NMR spectra of the products of acid-catalyzed decomposition of 1.

476
E. Lee, D.V. Yandulov / Journal of Fluorine Chemistry 130 (2009) 474–483
Table 2
NMR-tube reactivity of 1a.
Entry
Conditionsb
Added reagent, T (8C)
Time
44 h
Conv. %
Inorg. Fꢀc
Org. Fꢀd
Other products identifiede
f
1
2
B, C, 30, P
B, C, 30, P
B, C, 50, T
S, 30, P
S, 45, P
S, 15, P
D, 80, T
D, 50, T
D, 60, T
D, 10, T
D, 50, T
D, 60, T
D, 60, P
D, 50, T
D, 50, P
D, 50, T
D, 50, T
D, 50, T
D, 60, T
B, 100, T
D, 50, T
D, 50, T
S, 90, T
B, 50, P
D, 50, T
D, 50, T
B, 25, P
B, 50, T
–
<10
100
<10
<10
70
–
–
–
–, +60
24 h
60 h
ꢀ157, ꢀ163
–
–
3
–, +60
–
–
–
–
¯
–
–
4
–, +145
18 h
–
–
5
4 CsF, +100
4 KOAc, +145
1.7 {HF}h, +45
24 h
–
H/D exch. (70)g
6
38 h
49
–
4 (37)
¯
7
98 h
52
HF + HF2 (168), BF4 (179)
ꢀ71, ꢀ73, ꢀ169, HF (28)
HF (33)
78:22 (29); (9)
82:18 (12); (14)
83:17 (23); (10)
100:0 (12)
100:0 (12); (9)
100:0 (15); (15)
64:36 (7); (6)
–
–
i
8
0.1 H3PO4
51 h
100
96
9
0.1 TFA
1.3 TFA
1.2 TFA
26 h
5 (11)
5 (66)
5 (37)
6 (3)
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
90 min
15 min
20 h
94
HF (83)
100
87
HF (81)
j
0.1 HNO3
HF (42)
0.1 HNO3
1.3 HNO3
1.1 HNO3
15 min
6 h
100
95
ꢀ149, ꢀ150, ꢀ161 (47)
HF (91)
6 (tr.)
6 (75)
6 (18)
70 min
25 min
5 min
1 h
99
ꢀ128, ꢀ149, ꢀ156, ꢀ161 (63)
HF (54)
100:0 (3); (12)
82:18 (19); (14)
–; (9)
k
0.1 H2SO4
98
0.4 H2SO4
100
100
100
100
100
100
<10
<10
100
100
<10
<10
ꢀ171 (4), HF (58)
ꢀ130 (10), HF (41)
HF (77)
0.3 BHT, 0.1 H2SO4, air
0.1 HOTf
100:0 (13); (19)
–; (12)
3 min
10 min
8 min
8 min
24 h
0.05 B2O3
HF (13)
69:31 (37); (36)
62:38 (10); (10)
47:53 (9); (9)
–
0.1 B2O3
ꢀ148, ꢀ151 (7), HF (46)
ꢀ148, ꢀ151 (11), HF (43)
–
0.1 B2O3, air
0.1 B2O3, +60
B2O3ꢁnNa2Ol
0.1 TMSOTf
0.1 BF3ꢁOEt2
0.1 AgOTf
–
–
91 h
–
100:0 (2)
78:22 (18); (15)
59:41 (26); (20)
10 min
5 min
96 h
HF (37), ꢀ157 (2)
¯
HF (51), BF4 (17)
–
–
1.0 6 or 5
21 h
–
a
All reactions at 22 8C under N2 unless otherwise noted (4: cinnamyl acetate, 5: cinnamyl trifluoroacetate, 6: cinnamyl nitrate).
Solvent (B = C6D6, C = CDCl3, D = CD2Cl2, S = d6-Me2SO), mM 1, NMR tube (P = Pyrex, T = Teflon insert).
b
c
19F NMR chemical shifts of unidentified inorganic fluoride products observed between
d
ꢀ71 and ꢀ195 ppm and assigned collectively to BF3, SiF4, PF5 and Fꢀ derivatives
such as BF3ꢁH2O, BF4ꢀ, SiF4(H2O)n, SiF62ꢀ, HPO3Fꢀ, HF2ꢀ, HFꢁnH2O and FꢀꢁnH2O, with individual or combined 19F NMR integrals in percent of that of initial 1 shown in
parentheses (mol% yield of fluorine-containing products of 1). HF was identified as a singlet at 19F NMR
d
ꢀ185 to ꢀ195 ppm and BF4ꢀ was identified by the presence of the
minor 10B isotopomer peak 0.05 ppm downfield from the 19F NMR singlet at
d
ꢂꢀ151 ppm.
d
3 dr (mol% yield from initial 1 by 19F NMR integration); (all other 19F signals, mol% yield from initial 1 by 19F NMR integration).
e
f
mol% yield from initial 1 by 1H NMR integration.
None added/observed.
g
h
i
E-PhCH = CHC{HnD2ꢀn}, n = 1 (50%), n = 0 (20%).
1.8[Bu4N]HF2 + 1.7BF3ꢁOEt2.
85 wt.% aq.
j
70 wt.% aq.
k
l
96 wt.% aq.
0.4 NaBO2ꢁ4H2O or 0.5 Na2B4O7ꢁ10H2O.
correlated with their aqueous acidity and concentration. Inactiva-
tion of B2O3, exceedingly effective in non-coordinating media, by
d6-Me2SO confirmed assignment of the acid function mediating
decomposition of neat 1 to three-coordinate boron oxo sites on the
surface of borosilicate glass [5,7]. Lack of inhibition by atmospheric
oxygen and BHT [8] provided evidence against participation of
radical intermediates in the acid-mediated solution decomposition
of 1. Dioxygen-triggered decomposition of neat 1 in soda lime glass
(Fig. 1) and PFA may thus signify formation of acidic initiators by
oxidative processes. All solution reactions of 1 with acids formed
HF [9], in 13–91% yield with the total 19F NMR mass balance of 54–
102% [10] (Table 2). Accelerated decomposition of 1 by acids in
solutions contained in glass tubes compared to Teflon inserts
(entries 12–15) therefore stems from autocatalytic formation of
H2SiF6 and BF3ꢁnOH2, both stronger acids than dilute HF [11], in
reaction of HF with borosilicate glass [12]. Substitution of Fꢀ with
TFA and NO3ꢀ, effected in good yields with stoichiometric TFA and
HNO3, constitutes a side reactivity in the acid-catalyzed decom-
position of 1, as the yields of cinnamyl trifluoroacetate (5) and
cinnamyl nitrate (6) did not correlate with rates of conversion of 1
and both were inert to 1 and B2O3 (Table 3).
progressive loss of structured signals from 1H NMR spectra. The
aging occurred faster with use of stronger acids and heat and
apparently precluded observation of 3 at complete conversion of 1
in the few rapid reactions with strong Brønsted acids (Table 2).
Treatment of neat 1 with 0.1 equiv. BF3ꢁOEt2 at >100 8C followed
by drying gave a yellow solid devoid of sharp 1H NMR features and
19F NMR signals (Fig. 2) that analyzed for elemental composition
{C9H8}m, was obtained in 89% yield and showed oligomeric peaks
of four to ten C9H8 unit masses in ESI-MS (Fig. 4a). These results
establish polycondensation with loss of HF as the limiting
stoichiometry of acid-catalyzed decomposition of 1.
Problems of instability of allylic fluorides have a sparse
literature record [2e,g,13-15]. Evidently, only Hammond et al.
observed 1 isolated in a round bottom flask to undergo ‘‘. . .gradual
decomposition to a gelatinous solid. . . upon standing at room
temperature’’ [3c]. Closest analogies to the observed behavior of 1
are found in the early studies of benzylic fluorides. The original
report of synthesis and characterization of parent benzyl fluoride
by Ingold and Ingold describes its autocatalytic dehydrofluorina-
tion to an opaque white glass of high molecular weight and
composition {C7H6}n that can be initiated by concentrated H2SO4,
HF, surface of only certain glass and ‘‘. . .may develop with almost
explosive violence’’ [16a]. Subsequent studies by Miller and Swain
revealed stabilizing influence of electron-withdrawing ring sub-
stituents in benzylic fluorides, utility of basic additives as chemical
Dimers 3, observed almost invariably at complete conversion of
1 with catalytic acids in yields of up to 37% (Table 2), reacted
further in the original reaction mixtures with formation and decay
of other aliphatic fluoride intermediates in 19F NMR (Fig. 3) and

E. Lee, D.V. Yandulov / Journal of Fluorine Chemistry 130 (2009) 474–483
477
Table 3
NMR-tube reactivity of other allylic derivativesa.
Entry
Allyl-X
Conditionsb
Added reagent, T (8C)
Time
58 h
Conv. %
Org. & inorg. Fꢀ (% yield)c
Other products (% yield)
d
1
2
3
4
5
6
6
5
2
B, 20, P
B, 30, P
C, 20, P
B, 20, P
S, 20, P
C, 50, T
0.1 B2O3
0.1 B2O3
–
<10
<10
100
100
<10
90
–
–
–
58 h
30 min
24 h
–
ꢀ147, ꢀ149, ꢀ161
12,
a
-terpinene, -terpinene, terpinolene
g
–, +60
–, +60
–, +60
ꢀ161, ꢀ166
a-terpinene (42), g-terpinene (12), terpinolene (tr.)
36 h
–
–
195 h
445 h
24 h
11 (60), 12 (12)
12 (60), -terpinene (9), 4-fluoro-p-menth-1-ene (2)
100
100
<10
100
<10
<10
<10
<10
100
<10
a
7
8
B, 60, T
B, 40, T
C, 70, P
S, 40, P
B, C, 20, T
B, 80, P
C, 90, P
–, +60
ꢀ170 (3), HF (5)
–
11 (10), 12 (59), a-terpinene (10), terpinolene (6)
–
0.1 CsF, +60
24 h
9
7
–
20 min
26 h
ꢀ150 to ꢀ162 (23)
14 (65)
10
11
12
13
–
–, +60
24 h
–
–
–
–
0.15 2,6-lutidine
30 min
25 h
–
13
–
–
–, +50
–, +60
4 h
ꢀ152, ꢀ156, ꢀ161 (15)
16 (21%), 15 (67)
14
B, 70, T
59 h
–
–
a
All reactions at 22 8C under N2 unless otherwise noted. Numbered structures are shown in Schemes 1 and 2.
Solvent (B = C6D6, C = CDCl3, D = CD2Cl2, S = d6-Me2SO), mM 1, NMR tube (P = Pyrex, T = Teflon insert).
b
c
19F NMR chemical shifts of unidentified organic and inorganic fluoride products, latter observed between
d
ꢀ129 and ꢀ195 ppm and assigned collectively to BF3, SiF4 and
Fꢀ derivatives such as BF3ꢁH2O, BF4ꢀ, SiF4(H2O)n, SiF62ꢀ, HF2ꢀ, HFꢁnH2O and FꢀꢁnH2O, with individual or combined 19F NMR integrals in percent of that of initial 1 shown in
parentheses (mol% yield of fluorine-containing products of 1). HF was identified as a singlet at 19F NMR
d
ꢀ185 to ꢀ195 ppm.
d
None added/observed.
stabilizers and positive effects of strictly anhydrous conditions and
scratch-free, fire-polished glassware in handling such compounds
[16b,c]. The sudden nature of glass-mediated decomposition of
BnF on storage was highlighted more recently in a safety alert of an
explosive incident [16d] and decomposition residue of neat BnF
was likened to hyperbranched polymers well known to result from
Friedel–Crafts condensation of heavier benzylic halides or alcohol
[17].
Related to the behavior of BnF, the overall acid-catalyzed
decomposition of 1 is proposed to follow a series of basic steps
shown in Scheme 1. The observed products of allylic substitution
(5, 6) and regioselectivity of dimerization to 3 signify electrophilic
activation of 1 at the allylic, as opposed to olefinic [18], position
and formation of anion-stabilized cinnamyl cation 8+ as a reactive
intermediate. Acid-mediated fluoride abstraction proposed for 1
finds ample precedent in solvolysis of 18-, 28-, 38-alkyl fluorides,
engenders fluorides with the highest Friedel–Crafts alkylation
reactivity of aliphatic halides in general [18]. Thus, 3 results from
regioselective addition of 8+ to 1 and trapping of the resulting
benzylic 9+ with Fꢀ. The allylic and benzylic carbocations so
formed can further lead to oligomeric products via a combination
of cationic addition polymerization (CA), and Friedel-Crafts
polycondensation pathways with aromatic (AC) or olefinic
substitution (OC). Because CA faces greater steric hindrance to
C–C bond formation [20] than do both polycondensation pathways
and since benzylic carbocations alkylate olefins with good
regioselectivity to aromatic substitution, e.g., in styrene [21,18],
we consider OC to be the favored pathway by which acid-catalyzed
dehydrofluorination of 1 takes place. This pathway is well suited
for rapid formation of higher oligomers, since acid lability of the
condensation dimer 10 is enhanced relative to that of its precursor
1 by electron-donating
b-substitution [21a].
Ph3CF, BnF, BzF,
a fluorinated allylic fluoride [19,16c], and
Fig. 3. 19F NMR spectra (376 MHz, 22 8C) showing aging of the initial products of
decomposition of 1 (30–50 mM, C6D6, Pyrex1 NMR tubes) formed in the following
reactions: 1 + 0.1 BF3ꢁOEt2, RT, 8 min (a), RT, 5 h 20 min (b), +100 8C, 24 h (c); 1 + 0.1
TMSOTf, RT, 15 min (d), RT, 7 h 30 min (e); 1 + 0.1 B2O3, RT, 8 min (f), RT, 9 h (g);
1 + 0.1 TFA, RT, 7 h 30 min (h), RT, 3d 19 h (i). Signals of 3 and 30 are marked with
arrows.
Fig. 4. ESI-MS spectra of the products of complete dehydrofluorination of (a) 1
({C9H8}m, THF:MeOH 1:1, satd. NH4Cl) and (b) 2 ({C10H16 m, THF:MeOH 1:2).
}

478
E. Lee, D.V. Yandulov / Journal of Fluorine Chemistry 130 (2009) 474–483
Scheme 1. Mechanism of acid-catalyzed decomposition of 1.
Acid lability of 1 is greater than that of BnF, which can be
showed oligomeric peaks of three to nine C10H16 unit masses in
ESI-MS (Fig. 4b). As for 1, these results establish polycondensation
with loss of HF as the limiting stoichiometry of acid-catalyzed
decomposition of 2, which we presume to stem from analogous
acid-mediated fluoride abstraction and involve several intramo-
lecular elimination and cyclization routes [22] in addition to
pathways outlined for 1 in Scheme 1. Without added reagents in
50 mM CDCl3 solution in a Teflon insert at +60 8C, 2 undergoes
nitrated [16a] (cf. Table 2). This enables 3 to persist past full
conversion of 1 in most cases before reacting more slowly at the
benzylic fluoride during the aging phase, via the intermediacy of 9+
[6]. Most resistant to acids, unactivated aliphatic fluorides (19
F
NMR
d
< ꢀ220 ppm) disappear last from 1/acid reaction mixtures
on extended aging (Fig. 3), via, e.g., Friedel–Crafts electrophilic
activation (Scheme 1). Although the structure of the final,
dehydrofluorinated oligomeric material {C9H8}m is not known,
pathways available for its formation, and subsequent cross-linking,
can in principle render each monomer unit connected at one or
more of all but the aromatic ipso positions, as schematically shown
in Scheme 1.
isomerization to linaloyl fluoride (11) followed by cyclization to a-
terpinyl fluoride (12) in good yields (Scheme 2). The isomerization
of 2 is analogous to that of respective alcohols [22b] and is
catalyzed by adventitious acid, as evidenced by inhibition by CsF
(Table 3).
3-Fluorocyclohexene (7) [15] and its novel derivative 13 were
prepared from respective chlorides by halide exchange with AgF
and isolated in good yields with help of Teflon/PFA equipment
(Scheme 2). In agreement with previous reports, 7 is remarkably
intolerant of borosilicate glass, however can be maintained
without significant decomposition in 20 mM C6D6 or CDCl3
solutions in a Teflon insert at +60 8C over 24 h, or 40 mM d6-
Me2SO solution in a borosilicate NMR tube at RT for 26 h (Table 3).
Consistent with electrophilic mode of activation of allylic
fluorides to decomposition, the less electron-rich 13 exhibit
increased stability to glass. Elimination and hydrolysis account for
up to 88% of products formed on decomposition of 7 and 13 on
contact with borosilicate glass (Table 3), which include oligomeric
cyclohexadienes {C6H8}5–11 and {C8H10O2}3–14 observed by
ESI-MS.
2.1. Other allylic fluorides
Geranyl fluoride (2) exhibits greater sensitivity than does 1 and
suffers decomposition (Fig. 1) not only in soda lime glass (6–29 h),
but also in PFA containers over 4–7 days at RT, possibly due to the
presence or accumulation of trace contaminants. It is nevertheless
possible to isolate 2 reproducibly in 86% yield (Scheme 2) by
distillation in a Teflon/PFA assembly and store it for months at
ꢀ35 8C under N2. Triggering instantaneous decomposition of 2 in
pentane solution or neat with 0.01 equiv. BF3ꢁOEt2 followed by
heating of the initial decomposition products at >100 8C and
vacuum-drying produced yellow to purple solids devoid of sharp
1H NMR features and 19F NMR signals that analyzed for elemental
composition {C10H16}m, were obtained in 71–77% yield and
Scheme 2. Synthesis and decomposition of allylic fluorides.

E. Lee, D.V. Yandulov / Journal of Fluorine Chemistry 130 (2009) 474–483
479
3. Conclusions
collected by filtration, washed with cold THF and dried at RT under
<10 mTorr vacuum to constant weight. Material purity and water
Some of the simplest known allylic fluorides, including cinnamyl
content were determined by integration of 1H (
d
ꢂ5.8 (H2O) and
fluoride which lacks aliphatic
b-hydrogens, are susceptible to acid-
3.47 (N(CH2C3H7)4+)) and 19F (
d
ꢂꢀ110 (F(H2O)nꢀ) and ꢀ152
catalyzed elimination of HF that is initiated by electrophilic
abstraction of the fluoride with moderately strong Brønsted or
Lewis acids and results from polycondensation. Acid sensitivity of
cinnamyl fluoride exceeds that of parent benzyl fluoride and is
elevated in the other allylic fluorides examined by structural
features that stabilize corresponding allylic cations. Lewis acidic
sites on the surface of borosilicate glass, such as three-coordinate
boron oxo, are particularly effective in promoting such reactivity in
neat liquids that therefore are best contained in soda lime glass,
Teflon or PFA vessels, refrigerated under N2, stabilized by mild bases
and/or stored in solutions. Allylic cations that result from the loss of
Fꢀ undergo addition, elimination, cyclization and hydrolysis by
water from etching of silica glass by the evolving HF, which also
makes decomposition of neat materials autocatalytic. Although
oligomeric materials ultimately dominate organic products of acid-
mediated decomposition of allylic fluorides studied here, isolation
of addition dimers of cinnamyl fluoride in modest yield suggests
that selectivity of electrophilic activation of allylic fluorides may be
optimized for olefin carbofluorination [23], as it has been in related
transformations [3f,13]. Conversely, mitigation of adventitious
acids by resorting to ordinary Teflon/PFA containers permits
isolation and handling of such electron-rich allylic fluorides as
geranyl fluoride and secondary 5-carbomethoxy-3-fluorocyclohex-
ene and 3-fluorocyclohexene in the neat state.
(HF2ꢀ)) NMR spectra recorded in C6D6, assuming the material to be
a mixture of the only species observed by NMR, [Bu4N]Fꢁ(H2O)n and
[Bu4N]HF2 [24]. Solid of a typical composition [Bu4N]Fꢁ(H2O)2
(91%, w/w) forms a homogeneous solution in C6D6 at 50 mM and
although NMR signals of Fꢀ(H2O)n and N(CH2CH2CH2CH3)4 may
+
appear as several closely lying peaks due to ion-pairing, depending
on the water content, same 1H and 19F NMR integration values are
obtained from spectra recorded in d8-THF that are free from ion-
pairing effects.
4.3. Tetrabutylammonium bifluoride [Bu4N]HF2
[Bu4N]Fꢁ(H2O)2.3 (91.3 wt.%, 503 mg, 1.52 mmol) was dried to
constant weight at 60 8C under <10 mTorr vacuum (38 h) in a glass
container. The resulting colorless oil crystallized on cooling to
ꢀ35 8C and remained solid at RT. Yield 264 mg (0.94 mmol, 103%).
Spectral data are in accord with those of a commercial material.
4.4. [(1E)-3-fluoroprop-1-en-1-yl]benzene (cinnamyl fluoride, 1)
Cinnamyl chloride (610 mg, 4.00 mmol) was added to Me2SO
(10 mL) solution of Bu4NFꢁ(H2O)1.7 (1.53 g, 93.0 wt.%, 4.87 mmol)
in a 20 mL scintillation vial in the glovebox. After stirring for 24 h at
RT, the solution was poured to 30 mL of CH2Cl2 in a 120 mL
separatory funnel, washed with 7ꢄ 10 mL of water, dried over
MgSO4 and filtered through Celite. The filtrate in a 100 mL RBF was
concentrated to ꢂ2 mL (regular borosilicate glassware can be used
up to this point) and the solution was transferred to a soda lime
glass bottle. The product 1 was distilled at ꢂ30 8C under <10 mTorr
vacuum in a soda lime assembly (Fig. 5) and collected as a colorless
oil at 0 8C, d 1.03 g/mL. Yield 453 mg (3.33 mmol, 83%; repeat
synthesis: 81%); this quantity of neat material can be stored over
extended periods in a soda lime container under N2 at ꢀ35 8C.
Spectral data are in accord with literature values [3b,c,2l].
4. Experimental
4.1. General experimental procedures
All air- and moisture-sensitive manipulations were performed
using oven- and flame-dried glassware, standard Schlenk and
gloveboxtechniquesunderanatmosphereofnitrogen, withsolvents
dried and deoxygenated by appropriate methods. All starting
materials and reagents were obtained from commercial suppliers
and used as received unless otherwise stated. CAUTION: pure neat
allylic fluorides described here undergo spontaneous and potentially
exothermic polymerization with release of HF on contact with
borosilicate glass or traces of acids; stability is imparted under air
atmosphere. 1H, 13C{1H} and 19F NMR spectra were recorded on
Varian 400 and 500 MHz spectrometers and referenced to the
4.5. [(1E)-5-fluoro-4-(fluoromethyl)-5-phenylpent-1-en-1-
yl]benzene (3 and 30)
Cinnamyl fluoride (1, 70 mg, 514
mmol) in a 7 mL PFA vial
under N2 was exposed to air for 10 s and stored at RT for 3 days. The
resultant pale purple gel was chromatographed on silica gel (85%
residual protio solvent peaks (1H: CDCl3,
5.32; (CD3)2SO, 2.49; CD3CN,
1.93 relative to Me4Si), solvent 13
signals (CDCl3, 77.23; C6D6, 128.39; CD2Cl2, 54.00; (CD3)2SO,
d7.27; C6D6, d7.16; CD2Cl2,
d
d
d
C
hexanes–CH2Cl2) to afford 3 (96:4 dr, Rf 0.30, 22 mg, 81
mmol, 32%)
d
d
d
d
and 30 (97:3 dr, Rf 0.33, 4.7 mg, 17
m
mol, 7%) as colorless oils.
39.51; relative to Me4Si), dissolved or external neat PhF (19F,
ꢀ113.15 ppm relative to CFCl3) or dissolved C6F6 (19F, ꢀ164.9 ppm
relative to CFCl3). Signals are listed in ppm, coupling constants in Hz,
and multiplets identified as s = singlet, d = doublet, t = triplet,
q = quartet, qn = quintet, sx = sextet, sp = septet, m = multiplet,
br = broad. Where identified, heteronuclear couplings were con-
firmedby selectivehomo-and heteronuclear decoupling, which was
19
also used in 1H and F NMR signal assignments of 3, 30 and 13.
Complete assignments of 1H and 13C signals of 13 were additionally
based on 1H–1H COSY, 13C DEPT and 1H–13C HSQC methods. In situ
conversion was quantified by integration of fully relaxed 1H and 19
F
NMR spectra vs. internal standards; relative integral values are
accurate to ꢃ10% in the 2s sense.
4.2. Tetrabutylammonium fluoride, hydrate [Bu4N]Fꢁ(H2O)n
Material available commercially as a 1 M THF solution or neat
semi-solid was dried at RT to solid consistency and recrystallized
from dry THF at ꢀ35 8C. The resultant white, crystalline solid was
Fig. 5. Soda lime and Teflon/PFA vacuum-distillation assembly.

480
E. Lee, D.V. Yandulov / Journal of Fluorine Chemistry 130 (2009) 474–483
3, 1H NMR (400 MHz, CDCl3, 22 8C):
d
7.46–7.21 (m, 10H, C6H5),
157.6 (q, J = 42.5, C(O)O), 137.33, 137.32, 135.6, 129.0, 127.1, 120.3,
3
4
3
6.40 (dt, JH1-H2 = 15.6, JH1-H3,3 = 1.4, 1H, C(1)H), 6.10 (ddd, JH1-
114.7 (q, J = 285.8, CF3), 68.8. 19F NMR (376 MHz, CDCl3, 22 8C):
ꢀ74.8.
d
0
H2 = 15.6, 3JH2-H3,3 = 7.7, 6.7, 1H, C(2)H), 5.53 (dd, JF-H5 = 46.3, 3JH5-
2
0
2
2
H4 = 7.4, 1H, C(5)H), 4.75 (dddd, JHF = 47.4, JHH = 9.4, 3JH-H4 = 4.5,
2
2
3
JF5-H = 1, 1H, C(4)CHH’F), 4.58 (dddd, JHF = 46.7, JHH = 9.3, JH-
4.8. (2E)-3-phenylprop-2-en-1-yl nitrate (cinnamyl nitrate, 6)
3
3
0
H4 = 3.6, JF5-H = 1.1, 1H, C(4)CHH’F), 2.34 (m, JH4-H3 = 8.3, JH4-
2
0
H3 = 5.2, 1H, C(4)H), 2.28 (m, JH3-H3 = 13.7, 1H, C(3)H’), 2.20 (m,
Cinnamyl chloride (150 mg, 0.98 mmol) and AgNO3 (334 mg,
1.97 mmol) were stirred in 10 mL of CH3CN for 24 h. The solution
was filtered through Celite and filtrate distilled at +50 8C under
<10 mTorr vacuum to yield cinnamyl nitrate as a colorless liquid,
which was collected at +5 8C. The product crystallized at ꢀ35 8C
and remained solid at room temperature, mp 29–30 8C. Yield
80 mg (0.45 mmol, 46%). Spectral data are in accord with literature
values [26].
1H, C(3)H). 13C NMR (126 MHz, CDCl3, 22 8C):
d 138.1 (d, J = 20.0),
137.3, 133.2, 129.0 (d, J = 2.1), 128.8 (d, J = 8.4), 127.5, 126.6,
126.55, 126.53, 126.3, 93.6 (dd, J = 173.9, 5.1), 81.8 (dd, J = 168.6,
5.3), 45.9 (dd, J = 21.5, 18.4), 30.1 (dd, J = 5.2, 4.2). 19F NMR
2
3
(376 MHz, CDCl3, 22 8C):
d
ꢀ179.4 (dd, JF-H5 = 46.4, JF-H4 = 14.6,
1F, C(5)F), ꢀ232.6 (td, 2JHF = 47.0, 3JF-H4 = 26.1, 1F, C(4)CH2F). Anal.
Calcd. for C18H18F2: C, 79.39; H, 6.66. Found: C, 79.11; H, 6.74.
30, 1H NMR (400 MHz, CDCl3, 22 8C):
d 7.44–7.20 (m, 10H, C6H5),
3
4
3
0
6.45 (dt, JH1-H2 = 15.9, JH1-H3,3 = 1.5, 1H, C(1)H), 6.14 (ddd, JH1-
4.9. (2E)-1-fluoro-3,7-dimethylocta-2,6-diene (geranyl fluoride, 2)
H2 = 15.7, 3JH2-H3,3 = 8.4, 6.4, 1H, C(2)H), 5.65 (dd, JF-H5 = 46.8, 3JH5-
2
0
2
2
3
H4 = 5.9, 1H, C(5)H), 4.52 (dddd, JHF = 47.0, JHH = 9.4, JH-H4 = 5.9,
Geranyl chloride [27] (298 mg, 1.73 mmol) was added to Me2SO
(5 mL) solution of Bu4NFꢁ(H2O)1.7 (660 mg, 93.0 wt.%, 2.10 mmol)
in a scintillation vial in the glovebox. After stirring for 24 h at RT,
the solution was poured to 15 mL of CH2Cl2 in a 120 mL separatory
funnel, washed with 7ꢄ 7 mL of water, dried over MgSO4, and
filtered through Celite. The filtrate was collected in a 100 mL RBF,
concentrated and transferred to a 26 mL PFA tube. The solution
was concentrated further to 2 mL, the product 2 was vacuum-
distilled (<10 mTorr) in a Teflon/PFA assembly (Fig. 5) and
collected as a colorless oil at 0 8C, d 0.86 g/mL. Yield 235 mg
(1.50 mmol, 87%; repeat synthesis: 85%); this quantity of neat
material can be stored over extended periods in a PFA container
2
2
3
JF5-H = 1.7, 1H, C(4)CHH’F), 4.31 (ddd, JHF = 47.1, JHH = 9.4, JH-
2
3
3
0
H4 = 4.0, 1H, C(4)CHH’F), 2.57 (m, JH3-H3 = 14.3, JH4-H3 = 4.2, JF-
H3 = 2.3, 1H, C(3)H), 2.39 (m, 3JH4-H3 = 9.3, 1H, C(3)H’), 2.28 (m, 1H,
0
C(4)H). 13C NMR (126 MHz, CDCl3, 22 8C):
d 138.5 (d, J = 20.4),
137.4, 132.9, 128.8, 128.7 (d, J = 1.3), 127.5, 127.1, 126.3, 125.9 (d,
J = 7.6), 93.5 (dd, J = 174.5, 4.0), 82.5 (dd, J = 168.8, 6.1), 46.4 (dd,
J = 22.9, 18.1), 29.0 (dd, J = 5.5, 4.3), one signal obscured. 19F NMR
2
3
(376 MHz, CDCl3, 22 8C):
d
ꢀ189.7 (dd, JF-H5 = 46.8, JF-
2 3
H4 = 20.4 Hz, 1F, C(5)F), ꢀ229.1 (td, JHF = 46.9, JF-H4 = 23.9, 1F,
C(4)CH2F). Anal. Calcd. for C18H18F2: C, 79.39; H, 6.66. Found: C,
79.66; H, 6.41.
under N2 at ꢀ35 8C. 1H NMR (400 MHz, d6-Me2SO, 22 8C):
d 5.43
3
3
4
4.6. Preparative scale dehydrofluorination of 1
(d ꢄ t ꢄ sx, JFH = 9.3, JH1-H2 = 7.2, JH2-C(3)CH = 4JH2-C(4)H,H’ = 1.3,
3
JH6-H5,5 = 6.8, 4JH6-C(7)CH = 4JH6-
0
1H, C(2)H), 5.06 (t ꢄ sp,
2
3
Neat liquid 1 (35.9 mg, 264
m
mol) was treated with BF3ꢁEt2O
closed 7 mL PFA vial,
C(8)H = 1.4, 1H, C(6)H), 4.87 (dd, JHF = 47.8, JH1-H2 = 7.3, 2H,
C(1)H2), 2.09–2.00 (m, 4H, C(4)H2, C(5)H2), 1.67 (dd, JFH = 4.8,
(3.1 L, 24.5 mol, 0.1 equiv.) in a
m
m
4
immediately producing visible smoke and turning into a deep
purple solid. After standing for 5 min, the vial was vented to N2,
then heated closed under N2 at +100 8C for 13 h and dried at +100
to 130 8C for 14 h, producing a yellow solid. Yield 27.3 mg
4JH2-C(3)CH = 1.5, 3H, C(3)CH3), 1.63 (d, JH6-CH = 1.2, 3H, CH3), 1.56
(d, 4JH6-CH = 1.0, 3H, CH3). 13C NMR (101 MHz, d6-Me2SO, 22 8C):
d
143.4 (d, 3JCF = 11.2, C3), 131.1, 123.6, 119.0 (d, 2JCF = 17.2, C2), 78.9
(d, 1JCF = 154.4, C1), 38.9 (d, JCF = 2.6), 25.7 (d, JCF = 3.4), 25.5, 17.5,
4
(235
m
mol, 89%). 1H NMR (400 MHz, CDCl3, 22 8C):
d
8.5–5.5
16.1 (d, JCF = 3.4, C(3)CH3); data are in accord with literature
values [3b,4]. 19F NMR (376 MHz, d6-Me2SO, 22 8C):
(br), 4.5–0.0 (br). Anal. Calcd. for C9H8: C, 93.06; H, 6.94. Found: C,
92.89; H, 6.68. ESI-MS (THF:MeOH 1:1, satd. NH4Cl): 465.1 (M4H+),
581.2 (M5H+), 697.4 (M6H+), 813.4 (M7H+), 929.6 (M8H+), 1045.5
(M9H+), 1161.9 (M10H+); 482.2 [M4+18]+, 598.3 [M5+18]+, 714.5
[M6+18]+, 830.6 [M7+18]+, 946.8 [M8+18]+, 1062.9 [M9+18]+,
1179.0 [M10+18]+; M = C9H8 (Fig. 4a).
In another preparation, neat 1 contained in a PFA vial was
decomposed by contact with Pyrex glass over several minutes at RT
and the initial products subjected to aging under 1 atm N2 at +90 8C
for 3d. ESI-MS spectrum of the resulting pale brown solid showed
series of signals consistent with elemental composition of partly
dehydrofluorinated oligomers of 1, {C9H8}m(HF)nNa+, m = 5–16,
n = 1–5.
d
ꢀ205.5
2
3
0
(t ꢄ d ꢄ sx, JHF = 47.8, JF-H2 = 9.6, JF-C(3)CH = JF-H4,4 = 4.8).
4.10. Preparative scale dehydrofluorination of (2)
Neat liquid 2 (28.9 mg, 185
m
mol) in a 7 mL PFA vial was
mol, 0.01 equiv.), immedi-
treated with BF3ꢁEt2O (0.23
m
L, 1.8 m
ately producing visible smoke and turning into a deep purple solid.
The solid was evacuated at RT for 3 h, heated closed under N2
atmosphere at +90 8C for 25 h and dried at +100 to 125 8C for 3
days, becoming light purple. Yield 19.4 mg (142
NMR (400 MHz, CDCl3, 22 8C): 6.96 (br), 5.33 (br), 5.23 (br), 5.10
m
mol, 77%). 1H
d
(br), 4.69 (br), 3.17–0.5 (br). Anal. Calcd. for C10H16: C, 88.16; H,
11.84. Found: C, 88.03; H, 11.68. ESI-MS (THF:MeOH 1:2): 409.1
(M3H+), 545.3 (M4H+), 681.6 (M5H+), 817.8 (M6H+), 953.9 (M7H+),
1090.1 (M8H+), 1226.2 (M9H+); M = C10H16 (Fig. 4b). In a separate
4.7. (2E)-3-phenylprop-2-en-1-yl trifluoroacetate (cinnamyl
trifluoroacetate, 5)
reaction, 2 (28.7 mg, 184
Teflon insert was treated with BF3ꢁEt2O (0.23
m
mol) dissolved in 0.5 mL of pentane in a
L, 1.9 mol,
A
CH2Cl2 solution (5 mL) of cinnamyl alcohol (280 mg,
m
m
2.09 mmol) was treated at RT with trifluoroacetic anhydride
(657 mg, 3.13 mmol) and the mixture was stirred for 30 min.
Distillation at +70 8C under <10 mTorr vacuum afforded cinnamyl
trifluoroacetate as a colorless liquid, which was collected at 0 8C.
0.01 equiv.), immediately producing orange coloration and pre-
cipitation of an orange solid. After 30 h at RT 19F NMR showed near
complete conversion of organofluorine intermediates. The reaction
mixture was rinsed out to a 7 mL PFA vial with help of CH2Cl2,
evaporated to dryness and the residue dried under <10 mTorr
vacuum at +100 to 120 8C for 24 h, remaining yellow. Yield 17.9 mg
Yield 385 mg (1.67 mmol, 80%). 1H NMR (500 MHz, CDCl3, 22 8C):
d
7.45–7.43 (m, 2H, Ho), 7.39–7.31 (m, 3H, Hm, Hp), 6.78 (d, 1H, 3JH2-
3
3
H3 = 16.0, C(3)H), 6.31 (dt, 1H, JH2-H3 = 15.8, JH2-H1 = 6.9, C(2)H),
5.00 (dd, 2H, 3JH2-H1 = 6.9, 4JH3-H1 = 0.8, C(1)H2); data are in accord
(131
(br), 5.23 (br), 5.10 (br), 4.69 (br), 3.17–0.5 (br). Anal. Calcd. for
C10H16: C, 88.16; H, 11.84. Found: C, 87.88; H, 11.60.
m d 6.96 (br), 5.33
mol, 71%). 1H NMR (400 MHz, CDCl3, 22 8C):
with literature values [25]. 13C NMR (126 MHz, CDCl3, 22 8C):
d

E. Lee, D.V. Yandulov / Journal of Fluorine Chemistry 130 (2009) 474–483
481
4.11. Thermolysis of 2 in C6D6 in a Teflon NMR tube insert
the reaction was quenched with CsF (45.5 mg, 0.30 mg) to absorb
HF and distilled under <10 mTorr vacuum at +55 8C, to yield
dicyclohex-2-en-1-yl ether (14) as a colorless liquid, collected at
A CDCl3 solution of 2 (0.4 mL, 50 mM, Teflon insert capped with a
Teflon cap inside a J. Young tube) containing an internal 19F NMR
standard was heated to +60 8C. After 8d 3 h, 1H and 19F spectra
showed formation of 3-fluoro-3,7-dimethylocta-1,6-diene (linaloyl
+5 8C. Yield 10 mg (52:48 dr, 56
mmol, 37%). Diastereomeric ratio
was determined by 13C and 1H NMR integration. Spectral data are
in accord with literature values [28]. Another sample of 7 in C6D6
solution was decomposed on contact with borosilicate glass and
decomposition products aged as formed for a month at RT. ESI-MS
of the reaction mixture in C6D6 showed series of signals consistent
with elemental composition of cyclohexadiene oligomers,
{C6H8}n(H,OH)+, n = 5–11.
fluoride, 11) in 60% yield by 19F NMR integration. 11, 1H NMR
3
(400 MHz, CDCl3, Teflon insert, 22 8C):
d
5.89 (td, JH2-
3
3
H1cis = 3JHF = 17.8, JH2-H1trans = 11.0, 1H, C(2)H), 5.27 (dt, JH2-
H1cis = 17.4, 2JHH = 4JHF = 1.3, 1H, C(1)Hcis), 5.13(dt, 3JH2-H1trans = 10.9,
3
0
2
JHH = 4JHF = 1.0, 1H, C(1)Htrans), 5.10 (t, JH6-H5,5 = 7.0, 1H, C(6)H),
2.14-1.98 (m, 2H), 1.77-1.65 (m, 2H), 1.69 (s, 3H, CH3), 1.61(s, 3H,
3
CH3), 1.42 (d, JFH = 21.8, 3H, C(3)CH3). 13C NMR (101 MHz, C6D6,
4.14. cis/trans-Methyl 5-fluorocyclohex-3-ene-1-carboxylate (13)
Teflon insert, 22 8C):
d 141.6 (d, J = 22.5, C2), 132.0, 124.8, 113.4 (d,
J = 11.2, C1), 95.9 (d, J = 171.7, C3), 41.0 (d, J = 23.2, C4), 26.1, 25.9 (d,
J = 25.1, C(3)CH3), 23.1 (d, J = 4.9, C5), 17.9; data are in accord with
literature values [4]. 19F NMR (376 MHz, CDCl3, Teflon insert, 22 8C):
Under N2, a 26 mL PFA tube was charged with AgF (436 mg,
3.47 mmol), P(o-Tolyl)3 (35 mg, 0.12 mmol), C6H6 (6 mL) and
trans-methyl
5-chlorocyclohex-3-ene-1-carboxylate
[29]
3
3
3
3
0
d
ꢀ148.5 (m, JH4-F = 23.3, JC(3)CH-F = 21.1, JH2-F = 18.3, JH4 -
(202 mg, 1.16 mmol). The mixture was sonicated at +20 8C in
the dark for 44 h, at which time complete conversion of the
chloride was established by NMR of an aliquot. The solids were
centrifuged off and the clear supernatant solution was decanted to
a clean PFA tube in the glovebox. The product 13 was vacuum-
distilled (<10 mTorr) in a Teflon/PFA assembly described in Fig. 5
and collected as a colorless oil at +5 8C, d 1.05 g/mL (approxi-
F = 16.6). After 18d 13 h of thermolysis, 1H and 19F NMR spectra
showed formation of 4-(1-fluoro-1-methylethyl)-1-methylcyclo-
hexene(
12, 1H NMR (400 MHz, CDCl3, Teflon insert, 22 8C):
a
-terpinylfluoride, 12),in60%yieldby19FNMRintegration.
d
5.39 (br s, 1H,
3
C(2)H), 2.12–1.61 (m, 7H), 1.67 (s, 3H), 1.33 (d, JFH = 22.0, 3H,
C(F)CH3), 1.32 (d, JFH = 22.0, 3H, C(F)CH3). 19F NMR (376 MHz,
CDCl3, Teflon insert, 22 8C): ꢀ140.4 (sp ꢄ d, JF-C(F)CH3 = 22.0, JF-
H4 = 11.5); data are in accord with literature values [14]. 13C NMR
(101 MHz, C6D6, Teflon insert, 22 8C): 134.1, 121.1, 97.1 (d,
3
3
3
d
mately by weight of a 10
pure (1.7%, w/w methyl benzoate and 3.2% (w/w) 1-carbo-
methoxy-2,4-cyclohexadiene identified as impurities),
mL volume). Yield 158 mg (95%, w/w
d
J = 167.6, C(4)CF), 44.2 (d, J = 22.1, C4), 31.2 (C6), 27.3 (d, J = 7.1), 25.3
(d, J = 25.1, C(4)CCH3), 24.5 (d, J = 5.2), 24.2 (d, J = 25.4, C(4)CCH3),
23.9.
0.95 mmol, 82%; repeat synthesis: 74%); neat material can be
stored over extended periods in a PFA container under N2 at
ꢀ35 8C. The ratio cis:trans = 40:60 was determined by 1H NMR
integration. In a C6D6 solution (>0.1 M 13) contained in a
borosilicate glass NMR tube, elimination of HF occurs on the
timescale of hours at RT producing 1-carbomethoxy-2,4-cyclo-
hexadiene and lowering cis:trans ratio. 1H NMR (400 MHz, C6D6,
4.12. 3-Fluorocyclohexene (7)
A 26 mL PFA tube was loaded with AgF (653 mg, 5.15 mmol),
P(o-Tolyl)3 (52 mg, 0.17 mmol), Et2O (10 mL) and 3-chlorocyclo-
hexene (200 mg, 1.72 mmol) in the glovebox and the mixture was
sonicated at 20 8C in the dark for 4 days, after which time complete
conversion was established by NMR. The solids were centrifuged
and the colorless solution was decanted off to a clean PFA tube.
After removing most of the ether at 350 Torr/RT, the residue was
distilled at 10 Torr/RT in a Teflon/PFA assembly described in Fig. 5.
The distillate collected in a PFA vial at ꢀ40 8C was additionally
dried under <10 mTorr vacuum at ꢀ78 8C for 2–3 h and quickly
transferred to a ꢀ35 8C glovebox freezer under N2. Yield 106 mg
(93.5 wt.% pure (1.6 wt.% 1,4-cyclohexadiene, 1.3 wt.% Et2O, and
3.6 wt.% 4-chlorocyclohexene as impurities), 0.99 mmol, 58%;
repeat synthesis: 56%); neat material can be stored over a few
22 8C): d5.67-5.60 (m, 1H, cis-C(4)H (5.63) [30], trans-C(4)H (5.62)
3
3
[30]), 5.57 (dddd, JH3-H4 = 10.1, JHF = 3.5, JHH = 5.2, 2.8, 0.6H,
3
trans-C(3)H), 5.41 (ddddd, JH3-H4 = 10.2, JHH = 4.5, 2.8, 1.6,
3JHF = 2.2, cis-C(3)H), 4.77 (dm, JHF = 49.0, 0.4H, cis-C(5)H), 4.63
(dm, JHF = 48.2, 0.6H, trans-C(5)H), 3.30 (s, 1.2H, cis-COOCH3),
3.28 (s, 1.8H, trans-COOCH3), 2.73 (dddd, JH1-H6a = 12.7, JH1-
2
2
3
3
3
3
H2a = 11.1, JH1-H2e = 5.2, JH1-H6e = 3.1, 0.6H, trans-C(1)H), 2.28
(ddtt, JHF = 15.8, JHH = 14.5, JH6e-H1 = 3JH6e-H5 = 3.0, JHH = 1.3,
1.3, 0.6H, trans-C(6)He), 2.23-2.15 (m, 0.8H, cis-C(6)H (2.19) [30],
cis-C(1)H (2.18) [30]), 2.13–1.82 (m, 2.4H, cis-C(2)H+trans-C(2)H
(2.1) [30], trans-C(2)H (1.98) [30], cis-C(6)H (1.91), cis-C(2)H
3
2
3
4
3
2
3
(1.86) [30]), 1.53 (dddd, JHF = 36.4, JHH = 14.5, JH1-H6a = 12.8,
3JH6a-H5 = 3.9, 0.6H, trans-C(6)Ha). 13C NMR (101 MHz, C6D6,
175.0 (trans-C(O)OMe), 173.9 (d, JCF = 1.5, cis-
C(O)OMe), 132.9 (d, JCF = 10.1, trans-C3), 130.0 (d, JCF = 9.3,
cis-C3), 127.6 (d, 2JCF = 21.3, cis-C4), 124.9 (d, 2JCF = 16.5, trans-C4),
weeks in a PFA container under N2 at ꢀ35 8C. 1H NMR (500 MHz,
22 8C)):
d
4
3
3
3
C6D6, Teflon insert, 22 8C):
d
5.79–5.74 (m, 1H), 5.66 (dq, JH1-
2
H2 = 10.2, JHH = 3.4, 1H), 4.75 (dq, JHF = 49.5, JHH = 4.5, 1H, C(3)H),
1.83–1.64 (m, 2H), 1.58–1.42 (m, 3H), 1.22–1.16 (m, 1H). 19F NMR
1
1
87.4 (d, JCF = 164.9, cis-C5), 83.7 (d, JCF = 164.9, trans-C5), 51.7
(cis-C(O)OCH3), 51.6 (trans-C(O)OCH3), 37.6 (d, 3JCF = 8.2, cis-C1),
(376 MHz, C6D6, Teflon insert, 22 8C):
accord with literature values [15b]. 13C NMR (126 MHz, C6D6,
Teflon insert, 22 8C): 133.6 (d, J = 9.3, C1), 126.6 (d, J = 18.5, C2),
d
ꢀ165.9 (m); data are in
2
2
35.1, (trans-C1), 32.2 (d, JCF = 20.2, cis-C6), 32.1 (d, JCF = 21.7,
4
4
d
trans-C6), 28.1 (d, JCF = 3.7, trans-C2), 27.7 (d, JCF = 2.6, cis-C2).
85.6 (d, J = 164.2, C3), 29.8 (d, J = 20.0, C4), 25.4 (d, J = 9.3, C5), 18.7
(d, J = 3.5, C6). GC retention time, T = 2.50; HP-5MS column (30 m,
19F NMR (376 MHz, C6D6, 22 8C)):
d
ꢀ162.1 (m, 0.6F, trans-C(5)F),
2
ꢀ172.4 (dm, JHF = 48.3, 0.4F, cis-C(5)F). Anal. Calcd. for
ID 0.25 mm, film thickness 25
m
m); 508(50) ! 208/0 ! 2508(50);
(C8H11FO2)0.951ꢁ(C8H8O2)0.017ꢁ(C8H10O2)0.032 = C8H10.917O2F0.951
:
sample stabilized with 0.2 equiv. Et3N in pentane in a glass vial. EI-
MS: m/z (int) 100 (M+, 55), 99 (14), 85 (66), 80 (35), 79 (34), 77 (16),
72 (100), 67 (11), 59 (30), 54(32), 53 (11), 51 (16).
C, 61.14; H, 7.00. Found: C, 61.54; H, 6.71. Relative configurations
of isomers of 13 were assigned on the basis of JHH and JHF coupling
constants, measured by selective decoupling of signals assigned
with help of JCF constants, 1H–1H COSY, 13C DEPT and 1H–13C HSQC
spectra. Specifically, trans-configuration of the major isomer was
established from assignment of C(1)H as axial and C(5)H as
equatorial (Chart 2). These assignments were consistent with
available NOE data, limited in scope and magnitude by signal
overlap.
4.13. Preparative scale dehydrofluorination of 7
A C6D6 (0.5 mL) solution of 7 (30 mg, 0.30 mmol) was poured
from a Teflon NMR tube insert to a borosilicate glass vial, turning
from colorless to deep blue-purple within a minute. After 10 min,

482
E. Lee, D.V. Yandulov / Journal of Fluorine Chemistry 130 (2009) 474–483
4.17. NMR-tube reactivity studies of 1 (Table 2) and other allylic
derivatives (Table 3)
All experiments listed in Tables 2 and 3 were done with internal
standards using flame-dried J. Young tubes with and without
Teflon inserts under N2, unless otherwise stated. Nominal
quantities of liquid fluorides were measured approximately, by
volume in a soda lime glass capillary (1 only) or a length of a Teflon
cannula (24 AWG, thin wall) calibrated manually. In reactions with
accurately measured, dissolved internal standards, actual amount
of allylic fluoride present initially in solution was determined by 1H
and/or 19F NMR integration and was used to calculate its actual
concentration and stoichiometric quantities of reagents that were
added subsequently to the NMR tube solution. In general, the
nominal and actual, integration-based quantities agreed to within
10% of each other. The listed reaction times mark the end of the
acquisition of the last spectrum used in quantification.
Chart 2.
4.15. Preparative scale dehydrofluorination of 13
A sample of 13 (30 mg, 0.19 mmol) in a C6D6 solution (0.6 mL)
in a J. Young tube turned from colorless to light green-blue within a
day. The reaction mixture was transferred to a PFA vial and
distilled under <10 mTorr vacuum at +130 8C to yield dimethyl
5,50-oxybiscyclohex-3-ene-1-carboxylate (15) as a colorless liquid
Acknowledgments
collected at +18 8C. Yield 7.5 mg (25
mmol, 27%). Isolated material
We thank Stanford University for funding and L. Chen for initial
observations.
is a mixture of at least 6 diastereomers of 15 according to GC, 13
NMR and elemental analysis. 1H NMR (500 MHz, CDCl3, 22 8C):
C
d
References
5.94–5.66 (m, 4H,), 4.33–3.94 (m, 2H), 3.70 (s, 6H), 2.90–2.58 (m,
2H), 2.42–2.11 (m, 6H), 1.81–1.63 (m, 2H). 13C NMR (125 MHz,
[1] (a) M. Pagliaro, R.J. Ciriminna, Mater. Chem. 15 (2005) 4981–4991;
(b) P. Maienfisch, R.G. Hall, Chimia 58 (2004) 93–99;
(c) R.E. Banks, B.E. Smart, J.C. Tatlow (Eds.), Organofluorine Chemistry: Principles
and Commercial Applications, Plenum Press, New York, 1994;
(d) T. Hiyama, Organofluorine Compounds: Chemistry and Applications,
Springer, New York, 2000.
[2] (a) M.C. Pacheco, S. Purser, V. Gouverneur, Chem. Rev. 108 (2008) 1943–1981;
(b) L. Hunter, D. O’Hagan, A.M.Z. Slawin, J. Am. Chem. Soc. 128 (2006) 16422–
16423;
CDCl3, 22 8C):
d 176.38, 176.34, 176.32, 175.25, 175.23, 175.21,
129.9, 129.80, 129.78, 129.77, 129.7, 129.49, 129.48, 128.1, 127.85,
127.77, 127.76, 127.7, 127.1, 127.0, 126.9, 126.7, 72.62, 72.56, 72.4,
72.2, 68.8, 68.53, 68.51, 68.47, 52.05, 51.97, 51.96, 38.8, 38.65,
38.57, 35.04, 34.95, 34.90, 34.88, 32.61, 32.55, 32.19, 32.16, 31.5,
31.1, 31.0, 27.99, 27.98, 27.75, 27.74, 27.71. GC retention times, T
(FID % of total): 15.890 (15.6%), 15.910 (23.2%), 16.140 (21.6%), 16.180
(24.0%), 16.380 (6.2%), 16.410 (7.3%); HP-5MS column (30 m, ID
(c) M. Tredwell, V. Gouverneur, Org. Biomol. Chem. 4 (2006) 26–32;
(d) R. Roig, J.M. Percy, Sci. Synth. 34 (2005) 331–343;
(e) M. Prakesch, D. Gre
0.25 mm, film thickness 25
m
m); 508(50) ! 208/0 ! 2508(100). Anal.
´e, R. Gre´e, Tetrahedron 59 (2003) 8833–8841;
(f) M. Prakesch, D. Gre´e, R. Gre´e, Acc. Chem. Res. 35 (2002) 175–181;
Calcd. for C16H22O5: C, 65.29; H, 7.53. Found: C, 65.55; H, 7.39. The
yellow oil of the distillation residue showed series of signals in ESI-
MS (THF:MeOH 1:1) consistent with elemental composition of
oligomers of 16, {C8H10O2}nNa+, n = 3–14.
´
´
(g) M. Prakesch, D. Gree, R. Gree, J. Org. Chem. 66 (2001) 3146–3151;
(h) V. Madiot, D. Gre´e, R. Gre´e, Tetrahedron Lett. 40 (1999) 6403–6406;
´
´
(i) S. Legoupy, C. Crevisy, J.-C. Guillemin, R. Gree, J. Fluorine Chem. 93 (1999) 171–
173;
(j) R.L. Gre´e, J.-P. Lellouche, in: V.A. Soloshonok (Ed.), Enantiocontrolled Synthesis
of Fluoro-Organic Compounds, John Wiley & Sons Ltd., Chichester, 1999, pp. 63–
106;
4.16. Screen of compatibility of neat 1 and 2 with various materials
(k) F. Benayoud, L. Chen, G.A. Moniz, A.J. Zapata, G.B. Hammond, Tetrahedron 54
(1998) 15541–15554;
(Fig. 1, Table 1)
(l) A. Boukerb, D. Gre´e, M. Laabassi, R. Gre´e, J. Fluorine Chem. 88 (1998) 23–27;
(m) F.A. Davis, P.V.N. Kasu, G. Sundarababu, H. Qi, J. Org. Chem. 62 (1997) 7546–
7547;
(n) D.M. Gree, C.J.M. Kermarrec, J.T. Martelli, R. Gree, J.-P. Lellouche, L.J. Toupet, J.
Org. Chem. 61 (1996) 1918–1919;
(o) M. El-Khoury, Q. Wang, M. Schlosser, Tetrahedron Lett. 31 (1996) 9047–9048;
(p) J.M. Dolence, C.D. Poulter, Tetrahedron 52 (1996) 119–130;
(q) M. Ceruti, S. Amisano, P. Milla, F. Viola, F. Rocco, M. Jung, L. Cattel, J. Chem. Soc.
Perkin Trans. 1 (1995) 889–893.
Materials of eight common laboratory glass containers were
drawn into ꢂ10
mL capillaries, flame-sealed at one end, fitted with
a sleeve of a Teflon cannula (16 AWG, standard wall) over the open
end, washed with DI water and dried at +160 8C for overnight
together with plugs made of the same material. The capillaries
were loaded with ꢂ3–5
mL of neat 1 or 2, pipetted with a length of
a Teflon cannula (24 AWG, thin wall), plugged to prevent
evaporation and positioned in front of a Logitech QuickCam1
Pro 9000 web camera, all inside a glovebox. For runs under air or O2
(grade 2.6) atmosphere, capillaries loaded in the glovebox under
N2 were placed under appropriate atmosphere inside a secondary
glass container, plugged promptly in the air and returned for video
monitoring into the secondary glass container maintained under
air. Decomposition typically manifested itself by formation of gas
bubbles and change of liquid consistency to a gelatinous solid of
various coloration that evolved over ꢂ2 min. The time of
decomposition onset was recorded at the first sign of gas evolution
to within a minute. Stability of 1 in soda lime and PFA containers
was monitored periodically by visual inspection of appearance and
1H and 19F NMR spectra taken on the neat liquid contained in
plugged capillaries made of soda lime in triplicate; no decom-
position was observed over 6 (soda lime) and 3 (PFA) months at RT
under N2 atmosphere. Analogous tests of 2 in PFA produced semi-
solid white products at times recorded to within 24 h.
[3] (a) J. Ichikawa, K. Sugimoto, T. Sonoda, H. Kobayashi, Chem. Lett. (1987) 1985–
1988;
(b) O.S. Park, H.J. Son, W.Y. Lee, Arch. Pharm. Res. 10 (1987) 239–244;
(c) D.J. de Mendonca, C.A. Digits, E.W. Navin, T.C. Sanders, G.B. Hammond, J.
Chem. Ed. 72 (1995) 736–739;
(d) N. Yoneda, T. Fukuhara, Chem. Lett. (2001) 222–223;
(e) N. Yoneda, T. Fukuhara, K. Shimokawa, K. Adachi, S. Oishi, US Patent 6,784,327
(2004).
(f) K. Hirano, H. Yorimitsu, K. Oshima, Org. Lett. 6 (2004) 4873–4875;
(g) B.P. Bandgar, S.V. Bettigeri, Monatsh. Chem. 135 (2004) 1251–1255;
(h) B.P. Bandgar, V.T. Kamble, A.V. Biradar, Monatsh. Chem. 136 (2005) 1579–
1582.
[4] D.V. Banthorpe, M.J. Ireland, J. Prakt. Chem. 338 (1996) 279–282.
´
[5] (a) L.D. Pye, V.D. Frechette, N.J. Kreidl (Eds.), Borate Glasses: Structure, Properties,
and Application Materials Science Research, vol. 12, Plenum Press, New York,
1979;
(b) S. Prasad, T.M. Clark, T.H. Sefzik, H.-T. Kwak, Z. Gan, P.J. Grandinetti, J. Non-
Cryst. Solids 352 (2006) 2834–2840.
[6] The diastereomers of 3 formed in the reaction of 1 with 0.1 B2O3 in C6D6 in 10 min
were observed by NMR to undergo acid-catalyzed equilibration during the
following hour.
[7] Apparent lack of reactivity of the latter with solutions of 1 in non-basic solvents is
partly due to dilution of neat 1 (7.6 M) by a factor of ꢂ100, which alone decreases

E. Lee, D.V. Yandulov / Journal of Fluorine Chemistry 130 (2009) 474–483
483
the rate of a surface reaction by the same or greater factor; full decomposition was
indeed accomplished by heating 30 mM solutions of 1 in C6D6 or CDCl3 for 24 h in
borosilicate NMR tubes, while analogous runs in Teflon inserts resulted in no
reaction (Table 2).
[14] M. Wu¨ st, D.B. Little, M. Schalk, R. Croteau, Arch. Biochem. Biophys. 387 (2001)
125–136.
[15] (a) G.A. Olah, M. Nojima, I. Kerekes, J. Am. Chem. Soc. 96 (1974) 925–927;
(b) M. Hudlicky, J. Fluorine Chem. 32 (1986) 441–452.
[8] (a) D.B. Priddy, Styrene Plastics Kirk-Othmer Encyclopedia of Chemical Technol-
ogy (online), Wiley–Interscience, New York, vol. 23, 2006, pp. 358–416;
(b) K.S. Khuong, W.H. Jones, W.A. Pryor, K.N. Houk, J. Am. Chem. Soc. 127 (2005)
1265–1277;
[16] (a) C.K. Ingold, E.H. Ingold, J. Chem. Soc. (1928) 2249–2262;
(b) J. Bernstein, J.S. Roth, W.T. Miller Jr., J. Am. Chem. Soc. 70 (1948) 2310–2314;
(c) C.G. Swain, R.E.T. Spalding, J. Am. Chem. Soc. 82 (1960) 6104–6107;
(d) S.S. Szucs, Chem. Eng. News 68 (34) (1990) 4.
(c) The retardation seen with 0.3 equiv. of BHT was similar to the base effect of
0.3 equiv. of MeOH. Complete inhibition of decomposition of 1 at 0, 0 and 25%
conversion in reactions with 0.1 equiv. H2SO4, 0.1 B2O3 and 0.1 BF3ꢁOEt2 was
effected by 0.15, 0.25 and 0.15 equiv. TEMPO, respectively, as a result of its acid-
mediated disporportionation that consumes stoichiometric equivalent of the
acid: J.M. Bobbitt, C.L. Flores, Heterocycles 27 (1988) 509–533. Z. Ma, J.M. Bobbitt,
J. Org. Chem. 56 (1991) 6110–6114. S. Kishioka, T. Ohsaka, K. Tokuda, Electrochim.
Acta 48 (2003) 1589–1594. X. Wang, R. Liu, Y. Jin, X. Liang, Chem. Eur. J. 14 (2008)
2679–2685.
[17] Y.H. Kim, O. Webster, J. Macromol, Sci. Polym. Rev. C42 (2002) 55–89.
[18] (a) G.A. Olah, Friedel–Crafts Chemistry, Wiley, New York, 1973, pp. 40, 81, 432;
(b) G.A. Olah (Ed.), Friedel–Crafts and Related Reactions, vol. 2, Interscience, New
York, 1964, pp. 313, 428, 767.
[19] (a) W.T. Miller Jr., J. Bernstein, J. Am. Chem. Soc. 70 (1948) 3600–3604;
(b) N.B. Chapman, J.L. Levy, J. Chem. Soc. (1952) 1677–1682;
(c) C.W.L. Bevan, R.F. Hudson, J. Chem. Soc. (1953) 2187–2189;
(d) C.G. Swain, R.B. Mosley, J. Am. Chem. Soc. 77 (1955) 3727–3731;
(e) C.G. Swain, T.E.C. Knee, A. MacLachlan, J. Am. Chem. Soc. 82 (1960) 6101–6104;
(f) T.J. Dougherty, J. Am. Chem. Soc. 86 (1964) 2236–2240;
(g) M.M. Toteva, J.P. Richard, J. Am. Chem. Soc. 124 (2002) 9798–9805.
[20] (a) A. Mizote, T. Tanaka, T. Higashimura, S. Okamura, J. Polym. Sci., Part A: Polym.
Chem. 3 (1965) 2567–2578;
(b) V.C. Armstrong, Z. Katovic, A.M. Eastham, Can. J. Chem. 49 (1971) 2119–2124.
[21] (a) H. Mayr, W. Striepe, J. Org. Chem. 48 (1983) 1159–1165;
(b) G.A. Olah, S.J. Kuhn, D.G. Barnes, J. Org. Chem. 29 (1964) 2685–2687.
[22] (a) J.A. Miller, H.C.S. Wood, Angew. Chem., Int. Ed. 3 (1964) 310–311;
(b) O. Cori, L. Chayet, L.M. Perez, C.A. Bunton, D. Hachey, J. Org. Chem. 51 (1986)
1310–1316.
[9] Identified by 19F NMR spectra of reactions in Teflon inserts as a singlet at d ꢀ185 to
ꢀ195 ppm that decoalesced into a JFH = 441 Hz doublet in spectra of entry 19
reaction following dilution with ether and cooling below RT. D.K. Hindermann,
C.D. Cornwell, J. Chem. Phys. 48 (1968) 2017–2025. C. MacLean, E.L. Mackor, J.
Chem. Phys. 34 (1961) 2207–2208. G.J. Schrobilgen, J. Chem. Soc., Chem. Commun.
(1988) 863–865.
[10] With the median mass balance of 83%, we suspect the deficiencies to stem from
partial escape of HF to the gas phase and leaking out of Teflon insert over periods
of days, as well as formation of multiple organofluorine intermediates with low
intensity signals (Fig. 2b). Reactions in Pyrex tubes showed lower 19F NMR mass
balance, of 34–79% with a median of 60%, due to precipitation/adsorption of
inorganic fluoro anions as metal salts on the glass surface.
[23] P. Wolfrum, in: B. Baasner, H. Hagemann, J.C. Tatlow (Eds.), Organo-Fluorine
Compounds, Methods of Organic Chemistry (Houben-Weyl), vol. E10, E10b1,
Thieme Verlag, Stuttgart, 1999, p. 308.
[11] (a) E.T. Urbansky, Chem. Rev. 102 (2002) 2837–2854;
(b) D. Fa˘rcaS¸u, A. Ghenciu, J. Catal. 134 (1992) 126–133;
[24] 2Fꢀ + H2O = HF2ꢀ + OHꢀ acid–base equilibrium is strongly shifted to the left in
Me2SO: H. Sun, S.G. DiMagno, J. Am. Chem. Soc. 127 (2005) 2050–2051.
[25] J.J. Gajewski, K.R. Gee, J. Jurayj, J. Org. Chem. 55 (1990) 1813–1822.
[26] (a) M. Ochiai, E. Fujita, M. Arimoto, H. Yamaguchi, Chem. Pharm. Bull. 32 (1984)
5027–5030;
(c) K.W. Kolasinski, J. Electrochem. Soc. 152 (2005) J99–J104.
[12] (a) A.J. Muscat, A.G. Thorsness, G. Montan˜o-Miranda, J. Vac. Sci. Technol. A 19
(2001) 1854–1861;
(b) D.M. Knotter, J. Am. Chem. Soc. 122 (2000) 4345–4351;
(c) G.A.C.M. Spierings, J. Mater. Sci. 28 (1993) 6261–6273;
(b) P. Frøyen, Phosphorous, Sulfur Silicon Relat. Elem. 129 (1997) 89–97.
[27] S. Nowotny, C.E. Tucker, C. Jubert, P. Knochel, J. Org. Chem. 60 (1995) 2762–2772.
[28] A. Uzarewicz, R. Dresler, Pol. J. Chem. 71 (1997) 181–195.
[29] J.-E. Ba¨ckvall, K.L. Granberg, A. Heumann, Isr. J. Chem. 31 (1991) 17–24.
[30] Chemical shifts of overlapping resonances determined from 1H–13C HSQC spec-
trum.
(d) R.A. Smith, Fluorine Compounds, Inorganic, Hydrogen Kirk-Othmer Encyclo-
pedia of Chemical Technology (online), vol. 13, Wiley–Interscience, New York,
2005, pp. 1–26.
[13] K. Hirano, K. Fujita, H. Yorimitsu, H. Shinokubo, K. Oshima, Tetrahedron Lett. 45
(2004) 2555–2557.