Preparation of a stable trifluoroborate salt for the synthesis of 1-aryl-2,2-difluoro-enolethers and/or 2,2-difluoro-1-aryl-ketones via palladium-mediated cross-coupling

Article abstract of DOI:10.1021/jo901789b

(Chemical Equation Presented) A bench-stable potassium trifluoroborate enol ether reagent has been prepared. This reagent is suitable for the incorporation of 2,2-difluoroenolethers into aryl and heteroaryl systems via palladium-mediated cross-coupling wi

Full text of DOI:10.1021/jo901789b

pubs.acs.org/joc
Preparation of a Stable Trifluoroborate Salt for the
Synthesis of 1-Aryl-2,2-difluoro-enolethers and/or
2,2-Difluoro-1-aryl-ketones via Palladium-Mediated
Cross-Coupling
FIGURE 1. Enol boron analogues of 1.
Jason D. Katz,* Blair T. Lapointe, and
Christopher J. Dinsmore
(Figure 1, 4a and 4b respectively) in a cross-coupling using
the common Suzuki-Miyaura conditions. Surprisingly,
given the extensive use of alkenyl boronates in modern
organic synthesis, there have only been a few reports of
1-alkoxyalkenyl boronic esters or acids in the literature.
Most reported examples focus on the preparation of the
3,4-dihydro-2H-pyran-6yl-boronic esters, where the enol
ether is contained in a ring system.4 However, encouraging
precedent was established by both Matteson’s preparation of
a (1-methoxyvinyl) boronic ester5 and the preparation of
1-ethoxy-butadienyl boronic esters by Venturello and Prandi
and their subsequent cross-coupling with aryl halides.6
Adapting the procedure described by Percy for the for-
mation of 1, 2-(methoxy)ethoxymethyl (MEM) protected
trifluoroethanol 5 was treated with 2 equiv of LDA at -78 °C
and quenched with triisopropyl borate (Scheme 2).2,7
Department of Chemistry, Merck and Co., Inc., 33 Avenue
Louis Pasteur, Boston, Massachusetts 02115
jason_katz2@merck.com
Received August 18, 2009
A bench-stable potassium trifluoroborate enol ether re-
agent has been prepared. This reagent is suitable for the
incorporation of 2,2-difluoroenolethers into aryl and
heteroaryl systems via palladium-mediated cross-cou-
pling with suitable halide coupling partners.
SCHEME 2. Preparation of Boronic Ester 4b
Following workup, the boronic acid 4a was isolated from
the crude reaction mixture, along with varying amounts of
enol ether 4d (typically ∼7%) and diisopropyl amine. How-
ever, boronic acid 4a proved relatively unstable and there-
fore was not an effective partner in our initial cross-coupling
reactions.8 Consequently, 4a was immediately converted to
the more stable boronic ester 4b, which was isolated as an oil.
Although 4b was normally used within 24 h for subsequent
cross-couplings, a toluene solution of 4b could be stored at
4 °C for >6 months, with no decomposition observed.
Gratifyingly, initial experiments with the key boronic ester
4b proved it to be an effective cross-coupling partner for the
Suzuki-Miyaura protocol. The desired coupling product 2
could be obtained in excellent yield (91%) under conditions
employing 2 equiv of 4b, 10% Pd2dba3, 20% P(o-tol)3, and 1
Introduction of fluorine into organic molecules is often
used in medicinal chemistry to affect a wide range of proper-
ties, such as desirable pKa, drug metabolism, and distribu-
tion in vivo.1 In a recent medicinal chemistry program, the
synthesis of a targeted compound required an R-gem-
difluoro ketone as a key intermediate. The initial approach
utilized chemistry described by Percy for the preparation
of the difluoroenolstannane reagent 1, followed by a cross-
coupling with aryl iodides under Stille conditions (Scheme 1,
1 f 2).2,3
SCHEME 1. Initial Approach to Difluoroketone 3
(4) (a) Bower, J. F.; Guillaneux, D.; Nguyen, T.; Wong, P. L.; Snieckus,
V. J. Org. Chem. 1998, 63, 1514. (b) Beckett, R. P.; Foster, R.; Henault, C.;
Ralbovsky, J. L. Gauss, C. M.; Gustafson, G. R.; Luo, Z.; Campbell, A.;
Shelekhin, T. E. WO 2008/157751 A2, 2008. (c) Allen, J. R.; Biwas, K.; Cao,
G.-Q.; Golden, J. E.; Mercede, S.; Peterkin, T.; Reed, A.; Tegley, C. M. WO 2008/
076425 A1, 2008. (d) Kikuchi, T.; Takagi, J.; Ishiyama, T.; Miyaura, N. Chem.
Lett. 2008, 37, 664.
(5) Matteson, D. S.; Beedle, E. C. Heteroat. Chem. 1990, 1, 135.
(6) Tivola, P. B.; Deagostino, A.; Prandi, C.; Venturello, P. Org. Lett.
2002, 4, 1275.
(7) (a) For a general review of R-metalated vinyl ethers, see: Friesen, R. W.
J. Chem. Soc., Perkins Trans. 1 2001, 1969. (b) After this work was complete, a
report was published describing the condensation of tosyloxydifluorovinyl-
lithium reagent with diisopropyl iodomethylboronate: Ramachandran, P. V.;
Chatterjee, A. Org. Lett. 2008, 10, 1195.
To enhance the practical value of this precedent, a side-
project was initiated to explore variations of this chemistry
that avoid problems arising from the use of tin reagents, such
as removal of the tin byproducts from the reaction mixture
and potential toxicity associated with exposure to tin. The
major focus of this effort was to test the feasibility of
employing the corresponding boronic acid or ester
(1) Hagmann, W. K. J. Med. Chem. 2008, 51, 4359.
(2) Patel, S. T.; Percy, J. M.; Wilkes, R. D. Tetrahedron 1995, 51, 9201.
(3) DeBoos, G. A.; Fullbrook, J. J.; Percy, J. M. Org. Lett. 2001, 3, 2859.
(8) The presence of 4d does not appear to interfere with subsequent cross-
couplings.
8866 J. Org. Chem. 2009, 74, 8866–8869
Published on Web 10/21/2009
DOI: 10.1021/jo901789b
r
2009 American Chemical Society

Katz et al.
JOCNote
M K2CO3(aq) in 1,4-dioxane at ambient temperature for
4 h.6 While not extensively optimized, the use of either the
corresponding aryl bromide, higher reaction temperatures,
lower catalyst loadings, or fewer equivalents of the boronic
ester 4b resulted in lower conversions and isolated yields of 2.
Furthermore, it appeared that boronic ester 4b was not
particularly stable under these reaction conditions
(temperature and/or base). Indeed, a toluene solution of
the boronic ester 4b stored over granular K2CO3 at 4 °C
decomposed overnight.
Whereas the cross-coupling with boronic ester 4b proved
adequate for the delivery of material for our internal pro-
gram, there were several characteristics of this reagent that
limited its’ broader utility. In particular, physical properties
(i.e., oil) made it inconvenient to prepare large quantities and
dispense aliquots for intermittent use.9 In addition, the
instability of the reagent under a variety of reaction condi-
tions was seen as a potential limitation to the substrate scope.
Molander and co-workers have demonstrated that alkyl and
aryltrifluoroborate salts often have superior reactivity com-
pared to that of the corresponding boronic acids and esters
and have the additional advantage of often being bench-
stable solids.10 Therefore, the preparation of the related
trifluoroborate salt 4c was investigated.
TABLE 1. Optimization of Reaction Conditions for Cross-Coupling of
4c with 1-Naphthyl Halides
isolated
yield (%)
entry
X
catalyst system
Base
NEt3
NEt3
1
2
Br
Br
Br
Br
Br
Br
Br
Br
Br
Br
Br
Br
Br
Cl
I
PdCl2(dppf) CH2Cl2
3
70
39
86
91
86
69
0
PdCl2/PPh3 (1:2)
3
4
5
6
7
8
9
10
11
12
13
14
15
PdCl2/X-Phos (1:3)
PdCl2/Ru-Phos (1:3)
Pd(OAc)2/Ru-Phos (1:3)
Pd2dba3/Ru-Phos (1:3)
PdCl2/Ru-Phos (1:3)
PdCl2/Ru-Phos (1:3)
PdCl2/Ru-Phos (1:3)
PdCl2/Ru-Phos (1:3)
PdCl2/Ru-Phos (1:3)
PdCl2/Ru-Phos (1:3)
PdCl2/Ru-Phos (1:3)
PdCl2/Ru-Phos (1:3)
PdCl2/Ru-Phos (1:3)
NEt3
NEt3
NEt3
NEt3
Cs2CO3
KOAc
K3PO4
K2HPO4
KH2PO4
NEt3
NEt3
NEt3
NEt3
36
53
64
6
25a
66b
78
48
Initially, with ample quantities of the boronic ester in
hand, trifluoroborate salt 4c was prepared by treating the
boronic ester 4b with KHF2 (Scheme 3). Workup and
aExperiment was conducted using boronic ester 4b. bExperiment was
initiated with 1.1 equiv of boronic ester 4b. After 2 h, an additional 1.1
equiv of boronic ester 4b was added.
as a ligand resulted in a much lower yield (entry 2); however,
using either the X-Phos or Ru-Phos13 ligands delivered the
coupled product 7 with a substantial increase in yield (entries
3 and 4).14 Given the success with Ru-Phos, it was decided to
use this ligand for future optimization of the cross-coupling
reaction. We next evaluated the source of palladium, as up to
this point, only PdCl2 had been used as a precatalyst.
However, screening other palladium sources did not improve
the yield relative to PdCl2 (entries 5 and 6).
Originally, NEt3 was selected as the base of choice, given
the potential instability issues of the boronic ester in the
presence of an inorganic base, as seen with potassium
carbonate (vide supra). Similarly, the coupling with Cs2CO3
proceeded poorly, with no desired product isolated from the
reaction mixture (entry 7). However, switching to the less
basic KOAc delivered the coupled product in 36% yield
(entry 8). A further examination using tri-, di-, and mono-
basic potassium phosphates (entries 9-11) revealed that
choosing an inorganic base with an appropriate pKa leads
to the formation of 7 in a moderate 64% yield.15
As a final comparison between the reactivity of the boronic
ester 4b and the trifluoroborate salt 4c, the cross-coupling of 4b
under the optimal conditions was examined in the context of
the 1-naphthyl system (entries 12 and 13). The reaction of 6b
with 1.1 equiv of boronic ester 4b resulted in a low 25% yield of
coupling product (entry 12). Given the propensity for the ester
reagent to decompose at these elevated temperatures, a sepa-
rate experiment was conducted where an additional 1.1 equiv
of 4b was added to the reaction mixture after 2 h (entry 13).
This modification resulted in a much improved yield (66%) of
SCHEME 3. Formation of Trifluoroborate Salt 4c
recrystallization from acetonitrile/ether delivered 4c in
91% yield (50-69% over 3 steps from 5). Similarly, the salt
could be generated more directly from the boronic acid 4a,
delivering 4c in a comparable 53% yield over two steps from
5. Pleasingly, trifluoroborate salt 4c was a solid that has
proven to be stable when stored for >6 months in a vial on
the benchtop (ambient light and temperature).
Turning to the Suzuki-Miyaura cross-coupling, optimi-
zation of the reaction conditions was explored using the
1-halo-naphthyl system (Table 1). Typical conditions for the
Suzuki-Miyaura cross-coupling of potassium trifluorobo-
rate salts (entry 1) delivered the desired product 7 in a
moderate 70% isolated yield.10 However, unlike the coup-
lings with boronic ester 4b, only 1.1 equiv of 4c was required
for complete conversion.11 In addition, the coupling with
trifluoroborate salt 4c required heating at 90 °C for at least
6 h to go to completion.11,12
Although the PdCl2(dppf) system proved adequate, addi-
tional catalyst-ligand combinations were screened to iden-
tify a more effective system. Not surprisingly, the use of PPh3
(9) Although we have prepared large batches of the toluene solution,
syringed out aliquots as needed, and removed the toluene in vacuo, this is not
ideal.
(10) Molander, G. A.; Ellis, N. Acc. Chem. Res. 2007, 40, 275 and
references therein.
(11) For a full list of these results, please refer to Table S1 in Supporting
Information. In addition, microwave conditions were briefly explored using
the optimal conditions. The Supporting Information contains a brief account
of these efforts.
(13) RuPhos=2-dicyclohexylphosphino-20,60-diisopropoxybiphenyl; X-
Phos=2-dicyclohexylphosphino-20,40,60-triisopropylbiphenyl.
(12) Although the reactions were mostly complete at 8 h, they were
typically allowed to go for 23 h for convenience.
(14) Molander, G. A.; Brown, A. R. J. Org. Chem. 2006, 71, 9681.
(15) These results suggest that pKa ≈ 7-10 could be an optimal range.
J. Org. Chem. Vol. 74, No. 22, 2009 8867

Katz et al.
JOCNote
the coupling product, suggesting that the rate of decomposi-
tion of 4b at the optimized coupling conditions is competitive
with the rate of the cross-coupling reaction.
Finally, the coupling of the trifluoroborate salt 4c with the
1-chloro- and 1-iodo-naphthyl analogues revealed that both
aryl chlorides and aryl iodides are competent coupling
partners (entries 14 and 15). The chloro analogue delivered
the product in an acceptable 78% yield, while the iodo
analogue proceeded in a lower 48% yield.
alkyl-substituted aryl bromides reveals that the efficiency of
the cross-coupling is not dependent on the degree of substitu-
tion about the halide (entries 1-3). Additionally, the condi-
tions are tolerant to various electron-donating and electron-
withdrawing substituents on the aryl halides (entries 4-7).
However, the coupling of 4c with a nitro-substituted analogue
did not afford any of the desired product (entry 8). The cross-
coupling was also shown to work with several heteroaryl
bromides, such as the 3-pyridyl (entry 9) and 7-indolyl
(entry 10) analogues. Similar to the observations in the aryl
series, not all heteroaryl bromides were competent coupling
partners in the Suzuki-Miyaura reaction (entry 11).
The results of the cross-coupling of trifluoroborate salt 4c
with a number of aryl and heteroaryl bromides is presen-
ted in Table 2. The cross-coupling with a variety of ortho
It is interesting to note that in the case of the keto
analogues (Table 2, entries 7 and 11), the trifluoromethyl
MEM ether side products 10 and 11 were also isolated from
the reaction mixture (Figure 2).16 One report in the patent
TABLE 2. Cross-Coupling of 4c with Aryl and Heteroaryl Bromides
FIGURE 2. Side-products isolated from coupling reactions.
literature describes a similar addition of HF to a related gem-
difluoroenolether in the presence of fluoride salts in protic
media.17,18 Therefore, it was postulated that 10 and 11
similarly arose as a result of the addition of fluoride into
the enolether, followed by protonation of the intermediate
from the n-propanol solvent. Indeed, when boronic ester 4b
was used in the cross-coupling with benzophenone (entry 7),
where no fluoride is present, only the desired enol ether was
isolated, albeit in a 29% yield.19 In the thiophene example
(entry 11), use of the boronic ester 4b did not result in the
formation of any desired product.
Building upon the result from entry 7, when the 3-nitro-
bromobenzene analogue was reacted with the boronic ester
4b, the desired product was isolated in 41% yield (entry 8).
Thus, although the trifluoroborate salt 4c proved effective
for most of the attempted cross-couplings, the use of the
boronic ester 4b can be explored in some cases as a secondary
strategy.
Whereas the 2,2-difluoroenolethers 9 are potentially
synthetically useful intermediates for more complex coup-
lings and rearrangements,18 simple deprotection of the
MEM protecting group reveals the originally desired gem-
difluoroketone (Scheme 1, 3). As a representative example,
the naphthyl derived enolether 7 was treated with TMSCl in
MeOH (Scheme 4), to reveal the gem-difluoroketone 12 in
91% isolated yield.3
In conclusion, the bench-stable potassium trifluoroborate
salt 4c has been prepared to enable practical incorporation of
2,2-difluoroenolethers and difluoroketone functionality.
(16) A similar product was also observed in the 4-acetophenone deriva-
tive as well (data not shown).
(17) Ohtsuka, T. Y.; Kuroki, Y.; Suzuki, A.; Sugiyama, A. WO2007/
142110, 2007.
(18) In addition, difluoroenolethers have previously been shown to be
electrophiles at the gem-difluoro position for a variety of reactions, such as
(a) oxidative cross coupling: Uneyama, K.; Tanaka, H.; Kobayashi, S.; Shioyama,
M.; Amii, H. Org. Lett. 2004, 6, 2733 and (b) 2,3-Wittig rearrangements: Miles,
J. A. L.; Mitchell, L.; Percy, J. M.; Singh, K.; Uneyama, E. J. Org. Chem. 2007,
72.
aReported yields are an average of two runs. bReaction was con-
ducted using the boronic ester 4b instead of the trifluoroborate salt 4c.
(19) As with the naphthyl system described in Table 1, entry 13, the yield
likely could be increased by the addition of extra 4b 2 h into the reaction.
8868 J. Org. Chem. Vol. 74, No. 22, 2009

Katz et al.
JOCNote
SCHEME 4. Removal of MEM Protecting Group
Synthesis of Potassium Trifluoborate Salt 4c from Ester 4b.
Potassium hydrogen fluoride (10.1 g, 129 mmol) was added to a
solution of 4b (6.03 g, 21.5 mmol) in water (70 mL)/acetone
(175 mL), and the mixture was allowed to stir for 2 h. The
solvents were removed in vacuo, and then excess water was
removed by azeotroping three times with toluene and once with
acetone to give a colorless solid. The solid was suspended in
boiling acetone, filtered, and concentrated. The resulting solid
was dissolved in acetonitrile (40 mL) and Et2O (100 mL) was
slowly added to precipitate a colorless solid. The slurry was then
cooled to 0 °C, and the solid was collected by filtration and
rinsed with cold diethyl ether to give 5.34 g of 4c (19.5 mmol,
This reagent is suitable for cross-coupling reactions with a
range of aryl and heteroaryl halides. In certain cases where
side reactions were observed, the use of the related boronic
ester reagent 4b allows for isolation of the desired products.
Current work is focused on expanding the scope of this
reagent into other types of chemical transformations
common to potassium trifluoroborate salts.
1
91%) as a colorless solid. H NMR (500 MHz, acetone-d6) δ
4.87 (s, 2H), 3.65-3.75 (m, 2H), 3.45-3.55 (m, 2H), 3.30 (s, 3H);
19F NMR (470 MHz, acetone-d6) -99.5 (br d, 1F, J=74.7 Hz),
-115.3 (dq, 1F, J=9.0, 74.7 Hz), -141.1 (3F, 1:1:1:1 q, J=33.0
Hz); 13C NMR (125 MHz, acetone-d6) δ 159.2, 119.0, 94.8, 71.7,
67.1, 58.0. LRMS calcd for C6H9BF5O3 [M - K]- 235.1, found,
235.1.
Experimental Section
Synthesis of Boronic Ester 4b. Butyllithium (2.5 M in hexanes,
179 mL, 446 mmol) was added dropwise via addition funnel to a
solution of diisopropylamine (63.6 mL, 446 mmol) in THF
(240 mL) at -78 °C, such that the internal temperature remained
below -65 °C. After the addition was complete, the funnel was
rinsed with 10 mL of THF, and the solution was allowed to stir at
-78 °C for 10 min. A solution of 1,1,1-trifluoro-2-[(2-methoxye-
thoxy)methoxy]ethane (40.0 g, 213 mmol) in 40 mL of THF was
then added dropwise via addition funnel so that the internal
temperature remained below -65 °C (ca. 2 h). The reaction
mixture initially formed a yellow-green solution and then turned
to a brownish-orange slurry by the end of the addition. After the
addition was complete, the reaction mixture was held at -78 °C
for 30 min, followed by the addition of triisopropyl borate
(99 mL, 425 mmol) dropwise via addition funnel at a rate to
keep the internal temperature below -65 °C (ca. 2 h). The
resulting dark red-brown solution was allowed to warm
to -30 °C over 2 h, then was quenched by the addition of
200 mL of saturated NH4Cl (aq), and allowed to warm to ambient
temperature. The reaction mixture was partitioned between Et2O
(800 mL) and water (400 mL), and the layers were separated. The
organic layer was washed with 200 mL of water, and the combined
aqueous layers were adjusted to pH 5 with concentrated HCl. The
aqueous layer was then extracted with Et2O (3 ꢀ 300 mL),
and these combined Et2O layers were dried over Na2SO4, filtered,
and concentrated to give 63.9 g of an orange oil containing {2,2-
difluoro-1-[(2-methoxyethoxy)methoxy]ethenyl}boronic acid 4a.
The crude reaction mixture was then dissolved in toluene (1 L)
and THF (100 mL), 2,2-dimethyl-1,3-propanediol (22.1 g, 213
mmol) was added, and the mixture was allowed to stir at
ambient temperature overnight. The reaction mixture was then
concentrated to approximately 200 mL, diluted in Et2O (1.2 L),
and washed with water (200 mL) and brine (300 mL). The
aqueous layer was back extracted with Et2O (300 mL), and the
combined organic layers were dried over Na2SO4, filtered, and
concentrated to give 42.6 g of 4b (152 mmol, 71% yield) as a
colorless oil contaminated with the protonated byproduct 4d (7%
Representative Coupling Procedure: 1-(2,2-Difluoro-1-((2-
methoxyethoxy)methoxy)vinyl)naphthalene (7). Trifluoroborate
salt 4c (150.0 mg, 0.550 mmol), palladium(II) chloride (4.4 mg,
0.03 mmol), and 2-dicyclohexylphosphino-20,60-diisopropoxy-
biphenyl (35.0 mg, 0.075 mmol) were added to a flask, and the
vessel was sparged with Ar for 5 min. 1-Bromo-naphthalene
(70.0 μL, 0.500 mmol), triethylamine (209 μL, 1.50 mmol), and
n-PrOH (2.5 mL) were added, and the mixture was sparged with
Ar for 10 min. The reaction mixture was heated to 90 °C and
allowed to stir for 23 h. The reaction mixture was diluted in
EtOAc, washed with saturated aqueous NaHCO3 and brine,
dried over Na2SO4, filtered, and concentrated. The residue was
purified by flash column chromatography on silica gel (EtOAc/
hexanes gradient, 0-14%) to give 133.6 mg (0.454 mmol, 91%)
of 7 as a colorless oil. Rf=0.18 (7.5% EtOAc/92.5% hexanes);
1H NMR (500 MHz, CDCl3) δ 8.10 (d, J = 8.0 Hz, 1H),
7.94-7.84 (m, 2H), 7.60-7.44 (m, 4H), 4.71 (d, J = 1.0 Hz,
2H), 3.83-3.75 (m, 2H), 3.53-3.48 (m, 2H), 3.36 (s, 3H); 13C
NMR (126 MHz, CDCl3) δ 155.3 (dd, J = 290.3, 281.0 Hz),
133.9, 132.1 (app. d, J=3.3 Hz), 130.4, 129.8, 128.6, 127.0, 126.5,
126.1 (dd, J=4.2, 1.9 Hz), 125.5, 125.4, 114.3 (dd, J=40.9, 18.6
Hz), 94.1-93.7 (m), 71.8, 68.1, 59.2 (d, J=2.8 Hz); 19F NMR
(470 MHz, CDCl3) δ -101.2 (d, J=61.3 Hz, 1F), -108.7 (d, J=
61.3 Hz, 1F); HRMS calcd for C16H17F2O3 [M þ H]þ 295.1146,
found 295.1158.
Acknowledgment. Part of this work was completed as part
of the Co-Op Education Program with Northeastern Uni-
versity, Boston, MA (B.T.L). We thank Bruce Adams and
Bridget Becker for assistance with NMR spectra and Adam
Beard for obtaining HRMS data.
Supporting Information Available: Supplemental Table S1,
exploration of microwave experiments, experimental proce-
dures (4c from 4a, 12, microwave conditions) and characteriza-
tion data for all compounds. This material is available free of
charge via the Internet at http://pubs.acs.org.
1
1
by H NMR): H NMR (500 MHz, DMSO-d6) δ 4.78 (s, 2H),
3.62-3.68 (m, 6H), 3.40-3.45 (m, 2H), 3.18 (s, 3H), 0.90 (s, 6H).
J. Org. Chem. Vol. 74, No. 22, 2009 8869