Acceleration of Petasis reactions of salicylaldehyde derivatives with molecular sieves

Article abstract of DOI:10.1021/jo202117u

Mild reaction conditions for Petasis reactions of substituted salicylaldehydes with various amines and arylboronic acids in the presence of molecular sieves were developed. Molecular sieves (MS) significantly accelerated the reaction rates and drove the r

Full text of DOI:10.1021/jo202117u

Note
pubs.acs.org/joc
Acceleration of Petasis Reactions of Salicylaldehyde Derivatives with
Molecular Sieves
Xianglin Shi,*,† Dominique Hebrault,‡ Michael Humora,† William F. Kiesman,† Hairuo Peng,†
Tina Talreja,† Zezhou Wang,§ and Zhili Xin†
†Biogen Idec, 14 Cambridge Center, Cambridge, Massachusetts 02142, United States
‡Mettler Toledo Autochem, 95 Gould Street, Wakefield, Massachusetts, United States
§Department of Chemistry, Michigan Technological University, 1400 Townsend Drive, Houghton, Michigan 49930, United States
S
* Supporting Information
ABSTRACT: Mild reaction conditions for Petasis reactions of
substituted salicylaldehydes with various amines and arylbor-
onic acids in the presence of molecular sieves were developed.
Molecular sieves (MS) significantly accelerated the reaction
rates and drove the reactions to high conversions. The
conditions were applied to the synthesis of the core structure of BIIB042, a γ-secretase modulator, without stereochemical
erosion of a stereogenic center in the salicylaldehyde intermediate.
hree-component Petasis reactions of boronic acids,
amines, and aldehydes yield several types of important
the stereogenic center and to achieve an acceptable conversion
and yield for the reaction.
T
structures and have received significant attention recently,
largely due to the efficiency in synthesizing these structures
from readily available building blocks.1,2 Applications of Petasis
reactions in the synthesis of highly functionalized natural and
unnatural products of importance for medicinal research and
Petasis reactions that proceed with high stereoselectivity and
minimal protection of functional groups were exemplified
recently by Petasis,3 Wong,4 and Pyne.5 Also, a Petasis reaction
was successfully used in a concise synthesis of FTY720, a novel
S1P1 agonist (Gilenya) recently approved for the treatment of
multiple sclerosis.6
Besides actual structural effects of the reactants, the aspects
of the reaction conditions including solvents,7 microwave
irradiation,8 temperature, and catalysts9 can alter performance
of a Petasis reaction significantly. Often, specific conditions
need to be developed to achieve satisfactory results for a
desired set of reactants.1 The subject of this article is the study
of general effects of molecular sieves on Petasis reaction rates
and extends of conversions10,11 and a Petasis reaction
developed to synthesize BIIB042, a γ-secretase modulator
that was pursued for the treatment of Alzheimer’s Disease
(Scheme 1).12
The Petasis reaction employed to synthesize 2 involved
reaction of the slightly electron deficient 4-fluorophenylboronic
acid. The reaction was found to be sluggish, and heating was
necessary under typical literature conditions.3 The reaction only
proceeded to 70−80% conversion after 2−3 days. The
enantiomeric purities (by HPLC) in both recovered starting
material 1 and isolated product 2 were significantly lower than
that in the starting material 1 when the reaction was carried out
at ≥45 °C.13 Therefore, identification of milder Petasis reaction
conditions became an essential goal to preserve the integrity of
The first parameter explored was solvents because they can
play a critical role in Petasis reactions and sometimes allow the
reactions to proceed well at ambient temperature.3,7,14 For
preparation of 2 without racemization, however, in all solvents
screened the reaction was slow and stalled at low conversions of
1 at ambient temperature (Table 1).
According to the proposed mechanism,1 a Petasis reaction
occurs via an iminium intermediate, which is generated from
dehydration of a hemiaminal formed by reaction of an aldehyde
with an amine (see discussion below). We suspected the rate
and yield of the Petasis reaction of 1 would increase if
formation of the iminium intermediate was enhanced, and
removal of H2O from the reaction system could provide this
outcome. The effect of drying agents on the reaction was thus
studied, and the results are shown in Table 2. Molecular sieves
(MS) and MgCl2 dramatically improved the reaction rate and
the extent of reaction conversion at ambient temperature. The
conversion of 1 in the presence of MS reached 64% in just 4 h
(Table 2) and >98% in 2 days. MgCl2 gave comparable results
in the early stages of the reaction; however, the reaction
mixture formed a gel and the reaction rate slowed significantly.
To further investigate the effect of the MS, the kinetic
profiles of a series of reactions of salicylaldehyde, a model
compound for 1, with 4-fluorophenylboronic acid and
piperidine in several solvents were monitored. Figure 1 shows
the conversion data obtained with toluene, CH2Cl2, THF, and
EtOH as the reaction solvents. In the absence of the MS (data
in broken lines), the reactions were sluggish and stalled at low
conversions, particularly in EtOH and THF. Even with the
Received: October 17, 2011
Published: December 15, 2011
© 2011 American Chemical Society
1154
dx.doi.org/10.1021/jo202117u | J. Org. Chem. 2012, 77, 1154−1160

The Journal of Organic Chemistry
Note
Scheme 1. Proposed Synthesis of BIIB042 Using a Petasis Reaction
Table 1. Extent of Conversion in Different Solvents at Ambient Temperature
solvents
PhMe
25
PhCF3
38
CH2Cl2
43
(ClCH2)2
33
dioxane
20
CH3CN
16
THF
13
EtOH
10
a
conv (%)
a
Conversion of 1 to 2 determined by LC−MS detection at 214 nm.
Table 2. Effect of Drying Agents on the Reaction Rates
a
drying agents
none
28
MgSO4
28
MgCl2
62
molecular sieves
64
b
conv (%)
a
b
Aldrich 4 Å, activated, powdered, 2.5 μm molecular sieves were used for the experiments described in this paper. Conversion of 1 to 2 determined
by LC−MS detection at 214 nm.
Figure 1. Effect of molecular sieves on Petasis reaction rates and conversions of salicylaldehyde in various solvents.
optimal reaction solvent, CH2Cl2, conversion of salicylaldehyde
into 3a reached only ∼55 and 80% after 24 and 100 h,
respectively. However, when the MS were added, the reaction
rates accelerated significantly (data in solid lines). Moreover,
addition of the MS essentially drove reactions in CH2Cl2 (46
h), toluene (46 h), and THF (72 h) to completion. The effect
of the MS to the reaction in EtOH did not follow this trend.
Initially the reaction rate was similar to the reaction without the
1155
dx.doi.org/10.1021/jo202117u | J. Org. Chem. 2012, 77, 1154−1160

The Journal of Organic Chemistry
Note
Table 3. Petasis Reactions of Salicylaldehyde Derivatives with Arylboronic Acids and Secondary Amines in the Presence of
a
Molecular Sieves
a
All yields are isolated yields, and typical purity of the isolated products was >97% by HPLC (area %).
Scheme 2. Synthesis of the Core Structure 2 of BIIB042 via Petasis Reaction
MS. However, the extent of conversion eventually surpassed
the conversion reached by the reaction without MS.15
individual reaction. Structures and yields of the isolated
products are shown in Table 3.
As the data in Table 3 showed, these mild reaction
conditions have broad substrate generality.17 Most reactions
using secondary amines gave products in high yields. Regarding
the use of arylboronic acids, only the extremely electron
deficient 3-nitrophenylboronic acids failed to participate in the
reaction to form product 3o. This result is consistent with the
few literature examples using highly electron-deficient boronic
acids and is likely because these groups do not migrate well
from boron to the iminium carbon (see Scheme 3).18 3-
Nitrophenylboronic acid was reported to participate in a Petasis
reaction under harsher solvent-free conditions, i.e., by micro-
wave irradiation at 120 °C.8 The less electron-deficient 4-
cyanophenylboronic acid and 4-chlorophenylboronic acid
successfully reacted to produce the desired products 3e and
3m. For salicylaldehyde derivatives, observations similar to
those reported in the literature, substituents on salicylaldehyde
were generally well tolerated, except for an additional hydroxyl
group, which inhibited the formation of product 3p.19
Having shown that addition of the MS was beneficial, effects
of reaction concentrations and reagent stoichiometries of
piperidine and 4-fluorophenylboronic acid were then studied
in reactions with salicylaldehyde to prepare 3a. Higher reaction
rates at higher concentrations were observed.16 For studies on
the scope of the reactions discussed below, the reaction
concentrations (typically 5 vol relative to salicylaldehyde
derivatives) chosen were based on handling convenience.
Also, the reaction rates slowed near reaction completion. An
excess of the amines and arylbronic acids were used to limit the
reaction times to 24 h. These conditions were then applied to a
variety of reactants to determine the scope of the reactions
under these mild conditions (Table 3). Typically, 1.2−1.5 equiv
of the amines and boronic acids were used; reactions were
worked up after 24 h, and the products were isolated. In some
instances, on the basis of HPLC analysis, the reaction times
were extended to 72 h to achieve higher conversions. In
general, however, the conditions were not optimized for each
1156
dx.doi.org/10.1021/jo202117u | J. Org. Chem. 2012, 77, 1154−1160

The Journal of Organic Chemistry
Note
Scheme 3. Possible Mechanism for the Molecular Sieves Effect
The reactions produced few impurities. Product purifications
were accomplished by sequential washing of the reaction
mixtures with a slight excess of aqueous HCl and NaOH to
remove remaining amines and boronic acids, respectively,
followed by washing with brine, drying, and concentration.
Purities of the compounds obtained were typically >97% on the
basis of HPLC (area %) analysis and acquired NMR spectra.
Finally, these new mild conditions were applied to the
synthesis of 2. In the presence of the MS, a ∼97% conversion of
1 at ambient temperature was observed within 48 h (Scheme
2), and the reaction was very clean. With simple aqueous acid
and base washings to remove excess of the starting amine and
arylboronic acid, 2 was obtained in 85−90% yield as a 1:1
mixture of diastereomers (2a and 2b) and ∼99% purity (HPLC
area %). It is important to note that when enantioenriched (R)-
1 (∼80% ee) was used, product 2 was obtained as a 1:1
diastereomeric mixture with each diastereomer having the same
enantiomeric purity as 1 at the α-methyl stereogenic center.
Intermediate 2 was then converted into BIIB042 by a Suzuki
coupling reaction of the triflate intermediate20 following an
established methodology in our laboratory that preserved the
integrity of the α-methyl stereogenic center.12
The generally accepted mechanism for the Petasis reaction is
outlined in Scheme 3.1 All steps from hemiaminal (4)
formation through the iminium borate ester (6) are reversible,
followed by the irreversible and likely rate-limiting aryl
migration (step 4). The accelerating effect of MS on the
reaction may occur in step 2 of the proposed mechanism
(Scheme 3).1 By removing water from the reaction matrix,
added MS should drive the equilibrium from hemiaminal 4 to
iminium 5. Shift of the equilibrium should increase the
concentration of 5 and accelerate the overall reaction rate.
Furthermore, removal of water from the reaction matrix should
reduce the chance for hydrolysis of 6 before the desired
rearrangement to 7. Observations of a similar effect on the
reaction rate by drying agent MgCl2 and inhibition of the
reaction by H2O support this explanation.21
intermediate rapidly diminished, and formation of a new
species (presumably iminium 5) was observed. Observation of
an infrared absorption band at 1650 cm−1 for CN bond
(characteristic CN stretch of iminium ions is 1640−1700
cm−1)22 using the iC IR software for data interpretation
supported formation of 5.
After addition of 4-fluorophenylboronic acid, the IR data
showed disappearance of intermediate 5 and formation of a
new intermediate (6, probably) at a low concentration and a
product (7). The intermediate reached maximum concen-
tration 5 h after addition of the boronic acid and also displayed
an infrared absorption band at 1650 cm−1 corresponding to a
CN bond, which is present in an intermediate like 6. The low
concentration of 6 may indicate it exists mostly not as an
iminium ion but as a species more stable, for example, a
hemiaminal, which might be formed by reversible migration of
OH− from the B atom to the iminium ion C atom. Formation
of the presumed terminal compound 7 continued for 22 h after
addition of the boronic acid.
In summary, 4 Å activated MS significantly improved the
reaction rates and the degrees of conversions of Petasis
reactions of salicylaldehyde derivatives at ambient temperature.
Use of the MS allowed the reactions to be carried out under
mild conditions, which eliminated side reactions and permitted
facile isolation and purification of products by simple aqueous
workup.
EXPERIMENTAL SECTION
■
General Procedures. All reagents and solvents were purchased
from commercial sources and used as received. Both racemic and
enantio-enriched 2-(4-formyl-3-hydroxyphenyl)propanoic acid were
purchased from Adesis and used as received. Molecular sieves (4 Å,
activated, 2.5 μm, powdered) were purchased from Aldrich and used as
1
received. NMR spectra were recorded for H NMR at 400 MHz, for
13C NMR at 100 MHz, and for HSQC at 500 MHz (128 MHz for C),
and data were processed using ACDLABSv12 software. Chemical
shifts are expressed as δ (ppm) values using TMS or the residual
signals of the solvents as the internal standard. HPLC data were
collected with UV detection at 235 nm or as specified using Sunfire
column (C18, 3.6 μm, 150 × 4.6 mm) and mobile phase A (water with
0.1% TFA) and B (acetonitrile with 0.1% TFA) with a linear gradient
from 30 to 70% B in 15 min and a flow rate of 1 mL per minute. Chiral
HPLC data were collected with UV detection at 285 nm using Diacel
Chiralpak AD-H column (5 μm, 25 cm × 4.6 mm) and mobile phase
n-hexane/isopropanol/ethanol/trifluoroacetic acid/triethylamine
(250:50:5:0.5:0.25, v/v/v/v/v), with a flow rate of 0.3 mL per minute.
Real-time and postprocessing reaction IR data analysis was done using
the software iC IR 4.1 and ConcIRT.
Additional characterization data from in situ ATR-FTIR
spectroscopy monitoring of the reaction supported the role of
MS in the reaction mechanism described in Scheme 3. A
ReactIR probe was placed into the reaction vessel, and
monitoring was started when CH2Cl2 was added to the
reaction vessel, continued, and extended as salicylaldehyde,
piperidine, molecular sieves, and 4-fluorophenylboronic acid
were added, respectively. The reaction of salicylaldehyde and
piperidine occurred quickly to produce an intermediate, likely
hemiaminal 4, which persisted in solution in the absence of MS.
Upon addition of the MS, the concentration of this
Representative Procedure for Data Presented in Figure 1.
To a 10 mL vial were added 2-hydroxybenzaldehyde (0.5 g, 4.0
1157
dx.doi.org/10.1021/jo202117u | J. Org. Chem. 2012, 77, 1154−1160

The Journal of Organic Chemistry
Note
mmol), 4-fluorophenylboronic acid (0.618 g, 4.4 mmol, 1.1 equiv),
molecular sieves (Aldrich, 4 Å, activated, powdered, 2.5 μm, 0.45 g),
CH2Cl2 (2.0 mL), and piperidine (0.45 mL, 4.6 mmol, 1.15 equiv).
The mixture was stirred at 22−25 °C, and reaction conversions (based
on area percent) were determined by HPLC analysis at the indicated
times.
4H), 1.43−1.72 (br, m, 4H), 1.19−1.42 (br, 2H); 13C NMR (100
MHz, CDCl3) δ 162.2 (d, J = 245 Hz), 160.1, 158.3, 135.4 (d, J = 3
Hz), 130.4, 129.7, 117.7, 115.5 (d, J = 21 Hz), 105.3, 102.2, 74.9, 55.1,
52.3 (br, w), 26.1, 24.1; ESI-HRMS calcd for C19H23 FNO2 (M + H)+
316.17128, found 316.17078.
4-Bromo-2-((4-fluorophenyl)(piperidin-1-yl)methyl)phenol
(3c). Reaction of 2-hydroxy-4-bromobenzaldehyde (100 mg, 0.50
mmol), 4-fluorophenylboronic acid (76 mg, 0.55 mmol, 1.1 equiv),
molecular sieves (41 mg), and piperidine (54 μL, 0.55 mmol, 1.1
equiv) in CH2Cl2 (0.5 mL), 22 °C for 24 h gave product 3c (152 mg,
Representative Procedure for the Petasis Reactions in the
Presence of Molecular Sieves: Preparation of 2-((4-
Fluorophenyl)(piperidin-1-yl)methyl)-4-nitrophenol (3n). To a
magnetically stirred mixture of 2-hydroxy-5-nitrobenzaldehyde (200.0
mg, 1.20 mmol), 4-fluorophenylboronic acid (200.9 mg, 1.43 mmol,
1.2 equiv), molecular sieves (160 mg), and CH2Cl2 (2 mL) was added
piperidine (177 μL, 1.80 mmol, 1.5 equiv). The mixture was stirred at
22 °C for 24 h and filtered through a Celite bed. The filter cake was
washed with EtOAc (3 × 5 mL). The combined filtrates were washed
with 0.15 N HCl (5 mL, 1.5 equiv of the excess piperidine), 0.1 N
NaOH (3 mL, 1.5 equiv of the excess 4-fluorophenylboronic acid),
and brine, dried over anhydrous Na2SO4, filtered, and concentrated on
a rotary evaporator. The residue was dried under high vacuum at 22
°C overnight to give 3n as yellow oil, which slowly turned into solid
(0.36 g, 90%): 1H NMR (400 MHz, CDCl3) δ 14.2 (br, 1H), 7.96 (dd,
J = 2.76, 9.04 Hz, 1H), 7.74 (d, J = 2.51 Hz, 1H), 7.26 (br, s, 2H), 6.97
(dd, JH−H = 8.66 Hz, JH−F = 8.66 Hz, 2H), 6.81 (d, J = 9.04 Hz, 1H),
4.55 (s, 1H), 2.00−2.97 (br, 4H), 1.40−1.80 (br, m, 6H); 13C NMR
(100 MHz, CDCl3) δ 164.6, 162.7 (d, J = 247 Hz), 139.9, 133.0 (d, J =
4 Hz), 130.6, 125.5, 125.1 (2C), 117.5, 116.1 (d, J = 19 Hz), 74.8, 52.0
(br, w), 25.8, 23.7; HRMS calcd for C18H20FN2O3 (M + H+)
331.14580, found 331.14534.
1
84%), purity ∼100% (by HPLC): H NMR (400 MHz, CDCl3) δ
12.79 (br, 1H), 7.24 (br, s., 2H), 6.83−6.98 (m, 3H), 6.73 (dd, J =
1.88, 8.16 Hz, 1H), 6.61 (d, J = 8.28 Hz, 1H), 4.37 (s, 1H), 1.23 (br,
4H), 1.47−1.70 (m, 4H), 1.38 (br, 2H); 13C NMR (100 MHz,
CDCl3) δ 162.4 (d, J = 246 Hz), 158.3, 134.5 (d, J = 3 Hz), 130.4,
130.3, 124.4, 122.1, 121.8, 120.2, 115.7 (d, J = 21 Hz), 75.1, 52.3 (br,
w), 26.0, 24.0; ESI-HRMS calcd for C18H20BrFNO (M + H)+
364.07123, found 364.07090.
2-((4-Methoxyphenyl)(piperidin-1-yl)methyl)phenol (3d).
Reaction of 2-hydroxybenzaldehyde (200 mg, 1.6 mmol), 4-
methoxyphenylboronic acid (292 mg, 1.9 mmol, 1.2 equiv), molecular
sieves (160 mg), and piperidine (0.24 mL, 2.4 mmol, 1.5 equiv) in
CH2Cl2 (2 mL), 22 °C for 24 h gave product 3d (435 mg, 91%),
1
purity ∼98% (by HPLC): H NMR (400 MHz, CDCl3) δ 12.64 (br,
1H), 7.23−7.44 (m, 2H), 7.05−7.18 (m, 1H), 6.80−6.96 (m, 4H),
6.73−6.74 (m, 1H), 4.48 (s, 1H), 3.79 (s, 3H), 2.43 (br, 4H), 1.56−
1.89 (m, 4H), 1.47 (br, 2H); 13C NMR (100 MHz, CDCl3) δ 159.2,
157.2, 131.4, 130.1, 129.2, 128.2, 125.8, 119.0, 116.8, 114.0, 75.6, 55.2,
52.4 (br, w), 26.1, 24.2; ESI-HRMS calcd for C19H23NO2 (M + H)+
298.18070, found 298.18019.
Signals for hydrogens on the piperidine ring of compounds 3a, 3b,
1
3c, 3d, 3e, 3j, 3k, 3m, and 3n in the H NMR spectra obtained in
CDCl3 are very broad, and the integrations show less than the required
number of hydrogens. However, the spectra of 3a, 3b, 3c, 3j, and 3n
acquired in DMSO-d6 are normal (the NMR experiments in DMSO-d6
were not repeated for 3d, 3e, 3k, and 3m). The broadening of the
signals in CDCl3 is probably due to restricted rotation of the
piperidine ring,23 likely because of intramolecular H-bonding between
the N in the piperidine ring and the OH group on the phenyl group.
In DMSO, the H-bonding is disrupted and the rotation is less
restricted. Therefore, the NMR spectra are normal. For the same
reason, the signals of CH2−N in 13C NMR (CDCl3) spectra of 3a, 3b,
3c, 3d, 3e, 3j, 3k, 3m, and 3n appear at ∼δ 52 ppm as weak and broad
peaks but normal (∼δ 52 ppm) in the spectra acquired in DMSO-d6.
Assignment of the CH2−N signal was confirmed by HSQC (DMSO-
d6) experiments on 2j and 2n. The NMR data discussed here indicate
DMSO is a more appropriate solvent for 1H and 13C NMR spectra of
4-((2-Hydroxyphenyl)(piperidin-1-yl)methyl)benzonitrile
(3e). Reaction of 2-hydroxybenzaldehyde (200 mg, 1.6 mmol), 4-
cyanophenylboronic acid (283 mg, 1.9 mmol, 1.2 equiv), molecular
sieves (160 mg), and piperidine (0.24 mL, 2.4 mmol, 1.5 equiv) in
CH2Cl2 (2 mL), 22 °C for 24 h gave product 3e (195 mg, 42%), purity
1
∼90% (by HPLC): H NMR (400 MHz, CDCl3) δ 12.00 (br, 1H),
7.41−7.53 (m, 4H), 6.99−7.09 (m, 1H), 6.74−6.83 (m, 2H), 6.63 (dt,
J = 0.88, 7.47 Hz, 1H), 4.38 (s, 1H), 2.09−2.83 (br, m, 4H), 1.49−
1.74 (m, 4H), 1.42 (br, 2H); 13C NMR (100 MHz, CDCl3) δ 156.6,
145.4, 132.6, 129.2, 128.9, 124.5, 119.5, 118.6, 117.3, 111.7, 76.3, 52.9
(br, w), 26.0, 24.0; ESI-HRMS calcd for C19H20N2O (M + H)+
298.18070, found 293.16486.
2-((4-Fluorophenyl)(morpholino)methyl)phenol (3f). Reac-
tion of 2-hydroxybenzaldehyde (200 mg, 1.6 mmol), 4-fluorophe-
nylboronic acid (269 mg, 1.9 mmol, 1.2 equiv), molecular sieves (160
mg), and morpholine (0.21 mL, 2.4 mmol, 1.5 equiv) in CH2Cl2 (2
mL), 22 °C for 72 h gave product 3f (436 mg, 95%), purity ∼100%
(by HPLC): 1H NMR (400 MHz, CDCl3) δ 11.64 (br. s., 1H), 7.33−
1
these compounds. H and 13C NMR spectra of 3a, 3b, 3c, 3j, and 3n
acquired in both solvents are provided in the Supporting Information
to demonstrate this point.
2-((4-Fluorophenyl)(piperidin-1-yl)methyl)phenol (3a). Reac-
tion of 2-hydroxybenzaldehyde (1.0 g, 8.0 mmol), 4-fluorophenylbor-
onic acid (1.36 g, 9.6 mmol, 1.2 equiv), molecular sieves (330 mg),
and piperidine (0.96 mL, 9.6 mmol, 1.2 equiv) in CH2Cl2 (25 mL), 22
°C for 48 h gave product 3a (2.0 g, 88%), purity ∼100% (by HPLC):
1H NMR (400 MHz, CDCl3) δ 12.44 (br, 1 H), 7.36 (br, 2 H), 7.03−
7.14 (m, 1 H), 6.98 (dd, JH−F = 8 Hz, JH−H = 8 Hz, 2 H), 6.84 (m, 2
H), 6.69 (dt, J = 1.3, 7.4 Hz, 1 H), 4.45 (s, 1 H), 2.16−2.51 (br, 4 H),
1.57−1.69 (m, 4 H), 1.40−1.53 (br, 2 H); 13C NMR (100 MHz,
CDCl3) δ 162.3 (d, J = 245 Hz), 156.9, 135.4 (d, J = 3 Hz), 130.4,
129.1, 128.5, 125.4, 119.1, 117.0, 115.6 (d, J = 22 Hz), 75.7, 52.5 (br,
w), 26.1, 24.1; ESI-HRMS calcd for C18H21FNO (M + H)+ 286.16072,
found 286.16021.
7.53 (m, 2H), 7.11−7.19 (m, 1H), 7.01 (dd, JH−H = 8.66 Hz, JH−F
=
8.66 Hz, 2H), 6.91−6.96 (m, 1H), 6.89 (d, J = 8.03 Hz, 1H), 6.72−
6.78 (m, 1H), 4.41 (s, 1H), 3.57−3.93 (m, 4H), 2.61 (br, 2H), 2.27−
2.50 (m, 2H); 13C NMR (100 MHz, CDCl3) δ 162.4 (d, J = 245 Hz),
156.0, 135.2 (d, J = 3 Hz), 130.2 (d, J = 7 Hz), 129.3, 128.9, 124.7,
119.7, 117.2, 115.8 (d, J = 21 Hz), 76.0, 66.9, 52.2; ESI-HRMS calcd
for C17H18FNO2 (M + H)+ 288.13998, found 288.13947.
2-((Diallylamino)(4-fluorophenyl)methyl)phenol (3g). Reac-
tion of 2-hydroxybenzaldehyde (200 mg, 1.6 mmol), 4-fluorophe-
nylboronic acid (269 mg, 1.9 mmol, 1.2 equiv), molecular sieves (160
mg), and N-2-propenyl-2-propen-1-amine (0.30 mL, 2.4 mmol, 1.5
equiv) in CH2Cl2 (2 mL), 22 °C for 24 h gave product 3g (455 mg,
95%), purity ∼97% (by HPLC): 1H NMR (400 MHz, CDCl3) δ 12.02
(br. s., 1H), 7.43 (dd, JH−F = 5.52, JH−H = 8.28 Hz, 2H), 7.12−7.21 (m,
1H), 7.07 (dd, JH−H = 8.66 Hz, JH−F = 8.66 Hz, 2H), 6.88−6.94 (m,
1H), 6.83 (d, J = 7.28 Hz, 1H), 6.68−6.76 (m, 1H), 5.92 (tdd, J =
6.78, 10.23, 16.88 Hz, 2H), 5.27 (d, J = 10.29 Hz, 2H), 5.17 (d, J =
17.07 Hz, 2H), 5.07 (s, 1H), 3.40 (dd, J = 5.77, 14.05 Hz, 2H), 3.06
(dd, J = 7.53, 14.05 Hz, 2H); 13C NMR (100 MHz, CDCl3) δ 162.3
(d, J = 246 Hz), 157.4, 133.5 (d, J = 3 Hz), 133.3, 131.1 (d, J = 8 Hz),
129.2, 128.8, 124.8, 119.7, 119.2, 117.0, 115.6 (d, J = 21 Hz), 69.2,
2-((4-Fluorophenyl)(piperidin-1-yl)methyl)-4-methoxyphe-
nol (3b). Reaction of 2-hydroxy-4-methoxybenzaldehyde (100 mg,
0.66 mmol), 4-fluorophenylboronic acid (101 mg, 0.72 mmol, 1.1
equiv), molecular sieves (54 mg), and piperidine (71 μL, 0.72 mmol,
1.1 equiv) in CH2Cl2 (0.5 mL), 22 °C for 24 h gave product 3b (165
1
mg, 80%), purity ∼97% (by HPLC): H NMR (400 MHz, CDCl3) δ
12.34 (br, 1H), 7.24 (br, s, 2H), 6.89 (dd, JH−H = 8.66 Hz, JH−F = 8.66
Hz, 2H), 6.63 (d, J = 8.53 Hz, 1H), 6.34 (d, J = 2.76 Hz, 1H), 6.17
(dd, J = 2.51, 8.28 Hz, 1H), 4.36 (s, 1H), 3.64 (s, 3H), 1.99−2.77 (br,
1158
dx.doi.org/10.1021/jo202117u | J. Org. Chem. 2012, 77, 1154−1160

The Journal of Organic Chemistry
Note
52.3; ESI-HRMS calcd for C19H20FNO (M + H)+ 298.16072, found
298.16020.
156.9, 138.1, 133.7, 130.1, 129.1, 128.9, 128.6, 125.2, 119.2, 117.1,
75.8, 52.5 (br, w), 26.1, 24.1; ESI-HRMS calcd for C18H20ClNO (M +
H)+ 302.13117, found 302.13070.
2-((Diethylamino)(4-fluorophenyl)methyl)phenol (3h). Reac-
tion of 2-hydroxybenzaldehyde (200 mg, 1.6 mmol), 4-fluorophe-
nylboronic acid (269 mg, 1.9 mmol, 1.2 equiv), molecular sieves (160
mg), and N-ethylethanamine (0.25 mL, 2.4 mmol, 1.5 equiv) in
CH2Cl2 (2 mL), 22 °C for 72 h gave product 3h (389 mg, 89%),
Procedure for in Situ ReactIR Spectroscopy Monitoring. To a
100 mL EasyMax glass reactor equipped with mechanical stirring,
temperature probe, and ReactIR probe was added CH2Cl2 (10 mL),
and the IR spectroscopy monitoring was started. The reaction
temperature was maintained at 22 °C through the reaction. To the
mixture was added salicylaldehyde (1.0 mL, 9.4 mmol), followed by
addition of piperidine (1.0 mL, 1.1 equiv). The reaction was allowed to
continue for ∼2 h, and molecular sieves (1.0 g) were added. After
about 1 h, 4-fluorophenylboronic acid (1.45 g, 1.1 equiv) was added to
the reaction mixture, and the reaction was continued for ∼24 h.
Reference IR spectra of salicylaldehyde, 4-fluorophenylboronic acid,
and piperidine in CH2Cl2 at roughly the same concentrations as in the
reaction were taken separately.
1
purity ∼98% (by HPLC): H NMR (400 MHz, CDCl3) δ 12.51 (br,
1H), 7.37−7.52 (m, 2H), 7.13 (t, J = 7.40 Hz, 1H), 7.03 (dd, JH−H
=
8.66 Hz, JH−F = 8.66 Hz, 2H), 6.79−6.92 (m, 2H), 6.64−6.75 (m,
1H), 4.91 (s, 1H), 2.77 (qd, J = 7.06, 13.71 Hz, 2H), 2.59 (qd, J =
6.91, 13.65 Hz, 2H), 1.09 (t, J = 7.03 Hz, 6H); 13C NMR (100 MHz,
CDCl3) δ 162.3 (d, J = 246 Hz), 157.4, 134.9 (d, J = 3 Hz), 130.6 (d, J
= 7 Hz), 128.9, 128.5, 125.6, 119.0, 117.0, 115.5 (d, J = 22 Hz), 70.3,
42.8, 11.0; ESI-HRMS calcd for C17H20FNO (M + H)+ 274.16072,
found 274.16022.
2-(4-((4-Fluorophenyl)(4-methylpiperidin-1-yl)methyl)-3-
hydroxyphenyl)propanoic acid (2) Disodium Salt. To a mixture
of 2-(4-formyl-3-hydroxyphenyl) propanoic acid (388 mg, 2.0 mmol,
purity ∼95%, 9:1(R:S) enantiomer ratio), 4-fluorophenylboronic acid
(420 mg, 3.0 mmol, 1.5 equiv), and molecular sieves (194 mg) in
CH2Cl2 (2 mL) was added 4-methyl-piperidine (355 μL, 3.0 mmol,
1.5 equiv). The mixture was stirred at ∼22 °C for 44 h (>99%
conversion by LC−MS, detection 214 nm). The reaction mixture was
diluted with isopropyl acetate and filtered through Celite to remove
the molecular sieves. The filtrate was washed with H2O, dried, and
concentrated. The residue was dissolved in isopropyl acetate, and the
solution was extracted with aq NaOH (2 N, excess). To the aqueous
layer was added NaCl, and the mixture was extracted with isopropyl
acetate (2 × 5 mL). The combined organic layer was dried over
Na2SO4, filtered, and concentrated to give a white solid (0.70 g, 84%).
This material was >98% pure by HPLC and used for the following
triflation and Suzuki coupling reactions without further purification.
Diastereomer ratio of the isomers [(R,R) + (R,S)] to [(S,R) + (S,S)]
was 9:1, determined by chiral HPLC. Spectral data: 1H NMR (DMSO-
d6) δ 11.15 (br, 1H), 7.37−7.46 (m, 2 H), 7.03−7.15 (dd, JH−F = 8.0
Hz, JH−H = 8.0 Hz, 2 H), 6.92 (d, J = 7.97 Hz, 1 H), 6.77 (dd, J = 4.33,
1.76 Hz, 1 H), 6.57−6.64 (m, 1 H), 4.58−4.59 (2 peaks, 1 H), 3.16 (q,
J = 7.13 Hz, 1 H), 2.94 (d, J = 8.0 Hz, 1 H), 2.63 (d, J = 10.85 Hz, 1
H), 1.81−1.92 (m, 2 H), 1.53−1.64 (m, 2 H), 1.35 (br, 1 H), 1.10−
1.22 (m, 5 H), 0.88 (d, J = 6.53 Hz, 3 H); 13C NMR (100 MHz,
DMSO-d6) δ 177.6, 161.3 (d, J = 241 Hz), 155.1 (2 peaks, 1C), 145.0,
138.2 (d, J = 4 Hz), 129.7 (d, J = 8 Hz, 2C), 127.1, 123.8, 118.3, 115.1,
115.0 (d, J = 20 Hz, 2C), 70.2 (2 peaks, 1C), 51.9, 50.9, 48.2, 33.9
(2C), 30.2, 21.5, 19.8 (2 peaks, 1C); ESI-HRMS calcd for
C22H27FNO3 (M + H)+ 372.19706, found 372.19695.
2-((4-Fluorophenyl)(pyrrolidin-1-yl)methyl)phenol (3i). Reac-
tion of 2-hydroxybenzaldehyde (200 mg, 1.6 mmol), 4-fluorophe-
nylboronic acid (269 mg, 1.9 mmol, 1.2 equiv), molecular sieves (160
mg), and pyrrolidine (0.20 mL, 2.4 mmol, 1.5 equiv) in CH2Cl2 (2
mL), 22 °C for 24 h gave product 3i (397 mg, 91%), purity ∼100%
(by HPLC): 1H NMR (400 MHz, CDCl3) δ 12.16 (br, 1H), 7.46 (dd,
JH−F = 5.40, JH−H = 8.41 Hz, 2H), 7.08−7.17 (m, 1H), 6.93−7.01 (m,
3H), 6.88 (d, J = 8.28 Hz, 1H), 6.73 (t, J = 7.40 Hz, 1H), 4.39 (s, 1H),
2.65 (br, 2H), 2.49 (br, m, 2H), 1.68−1.99 (m, 4H); 13C NMR (100
MHz, CDCl3) δ 162.2 (d, J = 245 Hz), 156.5, 138.1 (d, J = 3 Hz),
129.4 (d, J = 8 Hz), 128.5, 128.2, 126.5, 119.3, 117.0, 115.5 (d, J = 22
Hz), 74.9, 53.2, 23.5; ESI-HRMS calcd for C17H18FNO (M + H)+
272.14507, found 272.14454.
2-((2-Methoxyphenyl)(piperidin-1-yl)methyl)phenol (3j). Re-
action of 2-hydroxybenzaldehyde (200 mg, 1.6 mmol), 2-methox-
yphenylboronic acid (292 mg, 1.9 mmol, 1.2 equiv), molecular sieves
(160 mg), and piperidine (0.24 mL, 2.4 mmol, 1.5 equiv) in CH2Cl2
(2 mL), 22 °C for 24 h gave product 3j (310 mg, 65%), purity ∼100%
(by HPLC): 1H NMR (400 MHz, CDCl3) δ 12.68 (br, 1H), 7.47 (d, J
= 7.03 Hz, 1H), 7.20−7.29 (m, 1H), 7.07−7.18 (m, 1H), 6.81−7.00
(m, 4H), 6.70 (t, J = 7.28 Hz, 1H), 5.36 (br, s, 1H), 3.89 (s, 3H), 2.49
(br, 4H), 1.55−1.80 (br, m, 4H), 1.48 (br, 2H); 13C NMR (100 MHz,
CDCl3) δ 158.0, 157.1, 129.7, 129.1, 128.6, 128.1, 127.3, 125.9, 120.9,
118.8, 116.7, 110.8, 65.4, 55.5, 51.5 (br, w), 26.2, 24.2; ESI-HRMS
calcd for C19H23NO2 (M + H)+ 298.18070, found 298.18023.
2-((3-Ethoxyphenyl)(piperidin-1-yl)methyl)phenol (3k). Re-
action of 2-hydroxybenzaldehyde (200 mg, 1.6 mmol), 3-ethoxyphe-
nylboronic acid (532 mg, 3.2 mmol, 2.0 equiv), molecular sieves (160
mg), and piperidine (0.24 mL, 2.4 mmol, 1.5 equiv) in CH2Cl2 (2
mL), 22 °C for 72 h gave product 3k (362 mg, 72%), purity ∼100%
(by HPLC): 1H NMR (400 MHz, CDCl3) δ 12.51 (br, s, 1H), 7.18−
7.26 (m, 1H), 7.08−7.15 (m, 1H), 6.99 (br, s, 2H), 6.91 (d, J = 7.03
Hz, 1H), 6.86 (d, J = 8.03 Hz, 1H), 6.80 (dd, J = 1.88, 8.16 Hz, 1H),
6.70 (t, J = 7.40 Hz, 1H), 4.43 (s, 1H), 4.01 (q, J = 6.94 Hz, 2H), 2.44
(br, 4H), 1.57−1.79 (m, 4H), 1.48 (br, s, 2H), 1.41 (t, J = 7.03 Hz,
3H); 13C NMR (100 MHz, CDCl3) δ 159.1, 157.1, 141.1, 129.6,
129.2, 128.3, 125.5, 119.0, 116.8, 113.5, 76.5, 63.3, 52.6 (br, w), 26.1,
24.2, 14.8; ESI-HRMS calcd for C20H25NO2 (M + H)+ 312.19635,
found 312.19585.
ASSOCIATED CONTENT
■
S
* Supporting Information
Copies of 1H/13C NMR spectra of the new compounds, HSQC
1
spectra and H/13C NMR assignment of 3j, 3n, and 2, HPLC
and chiral HPLC chromatograms for compound 2, and in situ
ReactIR Data. This material is available free of charge via the
Internet at http://pubs.acs.org.
2-(Phenyl(piperidin-1-yl)methyl)phenol (3l). Reaction of 2-
hydroxybenzaldehyde (200 mg, 1.6 mmol), phenylboronic acid (235
mg, 1.9 mmol, 1.2 equiv), molecular sieves (160 mg), and piperidine
(0.24 mL, 2.4 mmol, 1.5 equiv) in CH2Cl2 (2 mL), 22 °C for 24 h
AUTHOR INFORMATION
Corresponding Author
*E-mail: Xianglin.shi@biogenidec.com.
■
gave product 3l (250 mg, 58%), purity ∼99% (by HPLC): 1H and 13
C
NMR data match those reported for this compound.8
ACKNOWLEDGMENTS
2-((4-Chlorophenyl)(piperidin-1-yl)methyl)phenol (3m). Re-
action of 2-hydroxybenzaldehyde (200 mg, 1.6 mmol), 4-chlorophe-
nylboronic acid (301 mg, 1.9 mmol, 1.2 equiv), molecular sieves (160
mg), and piperidine (0.24 mL, 2.4 mmol, 1.5 equiv) in CH2Cl2 (2
mL), 22 °C for 24 h gave product 3m (193 mg, 40%), purity ∼96%
■
The authors thank Dr. Xiaoping Hronowski, Biogen Idec, for
acquiring high resolution mass spectra of the new compounds
reported in this paper and Ms. Annie Raditsis, Biogen Idec, for
acquiring the 2D NMR spectra; Dr. John Guzowski, Biogen
Idec, for his help with chiral HPLC analysis; Dr. Gnanasam-
bandam Kumaravel and Dr. Xiongwei Cai, Biogen Idec, for
helpful discussions during the work on this project. The authors
1
(by HPLC): H NMR (400 MHz, CDCl3) δ 12.21 (br, 1H), 7.21−
7.32 (m, 2H), 7.11−7.19 (m, 2H), 6.96−7.06 (m, 1H), 6.70−6.82 (m,
2H), 6.59 (dt, J = 0.88, 7.47 Hz, 1H), 4.34 (s, 1H), 2.28 (br, 4H),
1.44−1.71 (m, 4H), 1.36 (br, 2H); 13C NMR (100 MHz, CDCl3) δ
1159
dx.doi.org/10.1021/jo202117u | J. Org. Chem. 2012, 77, 1154−1160

The Journal of Organic Chemistry
Note
(16) For example, in the presence of the MS in THF at ambient
temperature, conversion of salicylaldehyde to 3a reached 96 and 81%
after 72 h at 2 and 0.8 M reaction concentrations, respectively.
(17) Molecular sieves were also found to accelerate the Petasis
reaction of glyoxylic acid hydrate with piperidine and 4-fluorophe-
nylboronic acid at room temperature. In the presence of the MS, 50−
60% of 4-fluorophenylboronic acid was converted into the desired
product in 24 h. In the absence of the MS, however, there was no
product observed.
(18) Churches, Q. I.; Stewart, H. E.; Cohen, S. B.; Shroder, A.;
Turner, P.; Hutton, C. A. Pure Appl. Chem. 2008, 80, 687.
(19) Petasis, N. A.; Boral, S. Tetrahedron Lett. 2001, 42, 539.
(20) Frantz, D. E.; Weaver, D. G.; Carey, J. P.; Kress, M. H.; Dolling,
U. H. Org. Lett. 2002, 4, 4717.
(21) Detrimental effect of H2O in the reaction of salicylaldehyde with
piperidine and 4-flurophenylboronic acid in THF was observed. When
4 equiv of H2O was added, the conversion was <2% (72 h, HPLC area
%). Under the same condition without added H2O, 16% conversion
(72 h) was observed, and in the presence of the MS, the reaction
proceeded to 81% conversion (72 h).
(22) Tehrani, K. A.; Kimpe, N. D. In Science of Synthesis, Houben-Weyl
Methods of Molecular Transformations; Padwa, A., Ed.; Georg Thieme
Verlag: Stuttgart, Germany, 2004; Vol 27, p 313. Product Class 8:
Iminium Salts. Heteroatom Analogues of Aldehydes and Ketones.
(23) Sanders, J. K.; Hunter, B. K. Modern NMR Spectroscopy: A Guide
for Chemists, 2nd ed.; Oxford University Press: Oxford, 1994; pp 216−
221.
are also grateful for constructive discussions with Dr. Don
Walker of Biogen Idec.
REFERENCES
■
(1) For a recent review, see: Candeias, N. R.; Montalbano, F.; Cal, P.
M. S. D.; Gois, P. M. P. Chem. Rev. 2010, 110, 6160.
(2) (a) Danieli, F.; Trabocchi, A.; Menchi, G.; Guarana, A. Eur. J. Org.
Chem. 2007, 1659. (b) Klopfenstein, S. R.; Chen, J. J.; Golebrowski,
A.; Li, M.; Peng, S. X.; Shao, X. Tetrahedron Lett. 2000, 41, 4835.
(c) Schlienger, N.; Bryce, M. R.; Hansen, T. K. Tetrahedron 2000, 56,
10023. (d) Wang, Q.; Finn, M. G. Org. Lett. 2000, 2, 4063.
(e) Candeias, N. R.; Cal, P. M. S. D.; Andre, V.; Duarte, M. T.; Veiros,
L. F.; Gois, P. M. P. Tetrahedron 2010, 66, 2736.
(3) (a) Petasis, N. A.; Zavialov, I. A. J. Am. Chem. Soc. 1997, 119, 445.
(b) Petasis, N. A.; Zavialov, I. A. J. Am. Chem. Soc. 1998, 120, 11798.
(c) Petasis, N. A.; Goodman, A.; Zavialov, I. A. Tetrahedron Lett. 1997,
53, 16463.
(4) (a) Hong, Z.; Liu, L.; Sugiyama, M.; Fu, Y; Wong, C.-H. J. Am.
Chem. Soc. 2009, 131, 8352. (b) Hong, Z.; Liu, L.; Hsu, C.-C.; Wong,
C.-H. Angew. Chem., Int. Ed. 2006, 45, 7417.
(5) (a) Ritthiwigrom, T.; Willis, A.; Pyne, S. G. J. Org. Chem. 2010,
75, 815. (b) Ritthiwigrom, T.; Pyne, S. G. Org. Lett. 2008, 10, 2769.
(c) Davis, A. S.; Pyne, S. G.; Skelton, B. W.; White, A. H. J. Org. Chem.
2004, 69, 3139.
(6) Sugiyama, S.; Arai, S.; Kiriyama, M.; Ishii, K. Chem. Pharm. Bull.
2005, 53, 100.
(7) For examples of solvent effects on Petasis reactions, see:
(a) Mandai, H.; Murota, K.; Sakai, T. Tetrahedron Lett. 2010, 51, 4779.
(b) Jourdan, H.; Gouhier, G.; Van Hijfte, L.; Angibaud, P.; Piettre, S.
R. Tetrahedron Lett. 2005, 46, 8027. (c) Yadav, J. S.; Reddy, B. V. S.;
Lakshmi, P. N. J. Mol. Catal. A: Chem. 2007, 274, 101.
(8) (a) Nun, P.; Martinez, J.; Lamaty, F. Synthesis 2010, 2063.
(b) McLean, N. J.; Tye, H.; Whittaker, M. Tetrahedron Lett. 2004, 45,
993.
(9) (a) Muncipinto, G.; Moquist, P. N.; Schreiber, S. L.; Schaus, S. E.
Angew. Chem., Int. Ed. 2011, 50, 8172. (b) Lou, S.; Schaus, S. E. J. Am.
Chem. Soc. 2008, 130, 6922. (c) Nanda, K. K.; Trotter, B. W.
Tetrahedron Lett. 2005, 46, 2025. (d) Tremblay-Morin, J.-P.; Raeppel,
S.; Gaudette, F. Tetrahedron Lett. 2004, 45, 3471.
(10) Schaus used molecular sieves in Petasis reactions (see ref 9a,b);
however, neither the effect of molecular sieves nor mention of any
reaction rate enhancements were discussed.
(11) In some cases, the use of molecular sieves did not have a
positive effect on the reaction yields. See: (a) Churches, Q. I.; White, J.
M.; Hutton, C. A. Org. Lett. 2011, 13, 2900. (b) Southwood, T. J.;
Curry, M. C.; Hutton, C. A. Tetrahedron 2006, 62, 236.
(12) Peng, H.; Talreja, T.; Xin, Z.; Cuervo, J. H.; Kumaravel, G.;
Humora, M. J.; Xu, L.; Rohde, E.; Gan, L.; Mi-young Jung, M.-Y.;
Shackett, M. N.; Chollate, S.; Dunah, A. W.; Snodgrass-Belt, P. A.;
Arnold, H. M.; Taveras, A. G.; Rhodes, K. J.; Scannevin, R. H. ACS
Med. Chem. Lett. 2011, 2, 786.
(13) The racemization was probably due to enolization involving the
stereogenic center, particularly of 1 with a strong electron-withdrawing
group -CHO at the para-position, in the presence of an amine base at
elevated temperatures. At ambient temperature, the enolization,
however, occurred very slowly and was not observed under the new
conditions.
(14) (a) Petasis, N. A.; Boral, S. Tetrahedron Lett. 2001, 42, 539.
(b) Naskar, D.; Roy, A.; Seibel, W. L.; Portlock, D. E. Tetrahedron Lett.
2003, 44, 8865. (c) Portlock, D. E.; Naskar, D.; Wex, L.; Li, M.
Tetrahedron Lett. 2002, 43, 6845. (d) Muncipinto, G.; Kaya, T.;
Wilson, J. A.; Kumagai, N.; Clemons, P. A.; Schreiber, S. L. Org. Lett.
2010, 12, 5230.
(15) The fact that 4 Å molecular sieves did not increase the reaction
rate in EtOH may be due to their inability to effectively remove H2O
in this solvent and/or the ability of EtOH to trap iminium ion 5 to
form a mixed N,O-acetal in situ that may revert to 5 slowly.
1160
dx.doi.org/10.1021/jo202117u | J. Org. Chem. 2012, 77, 1154−1160