We report a Pd-catalyzed route to heterocycles bearing a tetrasubstituted alkene fragment. Our approach merges the intramolecular carbopalladation of tethered alkynes with an alkylation step produced by the C-C cleavage of cyclobutanol derivatives. An alkenyl-Pd(II) intermediate has been isolated and characterized by X-ray diffraction studies. Interestingly, the nature of the tethering alkynyl chain influences the E/Z stereochemistry of the alkenyl fragment in the functionalized heterocycles.
We report a Pd-catalyzed route to heterocycles bearing a tetrasubstituted alkene fragment. Our approach merges the intramolecular carbopalladation of tethered alkynes with an alkylation step produced by the C-C cleavage of cyclobutanol derivatives. An alkenyl-Pd(II) intermediate has been isolated and characterized by X-ray diffraction studies. Interestingly, the nature of the tethering alkynyl chain influences the E/Z stereochemistry of the alkenyl fragment in the functionalized heterocycles.
The development of
Pd-catalyzed cascade reactions based on the
carbopalladation of alkynes has become a direct entry to the synthesis
of substituted alkenes.[1−9] Such reactions have been performed in either intra- or intermolecular
fashion, with the resulting alkenyl-Pd intermediate being coupled
afterward with different species, such as boronic acids,[10−12] organotin reagents,[13−18] and C-,[19]N-,[20,21] and O-nucleophiles,[22] among many others (a, Scheme ).[23−28]
Scheme 1
Merger of Carbopalladation of Alkynes and C–C Cleavage of
Cycloalkanols
Parallel studies have
demonstrated the ability of Pd to perform
the opening of strained cycloalkanols through β-carbon elimination
(b, Scheme ).[29,30] This process leads to a σ-alkyl-Pd(II)
intermediate, which can evolve in different manners, depending on
the substitution pattern of the cycloalkanol.[31−37] For instance, they can participate in further intramolecular steps,
or be cross-coupled with aryl-,[38−42] alkenyl-,[43,44] and alkynylhalides,[45] or propargylcarbonates,[46] among others.[29,47,48] Therefore, cyclopropyl- or cyclobutyl alcohols can behave as alkylating
reagents under the appropriate conditions.The merging of both
aspects of palladium chemistry (carbopalladation/alkylation
via opening of cycloalkanols) has rarely been reported in the literature.
Werz et al. disclosed an interesting cascade reaction relying on the
formal anti-carbopalladation of an internal alkyne,
evolving through further intramolecular trapping of the alkenyl-Pd(II)
intermediate by a tethered cyclopropanol moiety (c, Scheme ).[49] Very recently, Murakami, Chen, and co-workers reported
the synthesis of 2,3-dihydrobenzofurans through the use of alkenyl-tethered
aryliodides and benzocyclobutanols (d, Scheme ).[50,51]With these precedents in mind, and given our interest in the
topics
of Pd chemistry and the processes related to C–C cleavage,[52−57] we aimed to extend the applicability of these types of cascades
to the synthesis of heterocycles bearing an alkylated olefine moiety
(Scheme ).
Results
and Discussion
We studied the feasibility to perform the
envisioned carbopalladation/alkylation
cascade reaction employing the 2-bromoarylether 1a and
the cyclobutanol derivative 2a (Table ). Initial screening of experimental conditions
revealed the formation of some amounts of the byproduct 4a, likely arising from the protodepalladation of the plausible alkenyl-Pd(II)
intermediate generated upon the carbopalladation of the internal alkyne
moiety. The use of 10 mol% of [Pd(dba)2] along with 20
mol% of PPh3 showed good selectivity to give the desired
compound 3a in THF or toluene as solvents (entries 3
and 4, Table ). Replacing
PPh3 by other ligands such as JohnPhos, PCy3, or Xantphos did not improve the yields of 3a (entries
5–7, Table ). The increase of the amount of Cs2CO3 in
the reaction mixture could not suppress the protodepalladation process
leading to the byproduct 4a, and other organic bases
like NEt3 precluded the formation of 3a. We
tested Pd sources like Pd(OAc)2, [PdCl2(PPh3)2], and [Pd(PPh3)4]. While
the first two were not effective for this transformation, [Pd(PPh3)4] showed a comparable activity to [Pd(dba)2], reaching a 70% yield of the desired product.
Table 1
Optimization of the Carbopalladation/Alkylation
Cascadea
entrya
Pd source (10 mol %)
ligand (20 mol %)
solvent
yield 3ab
1
[Pd(dba)2]
PPh3
1,2-DCE
traces
2
[Pd(dba)2]
PPh3
1,4- dioxane
traces
3
[Pd(dba)2]
PPh3
THF
62
4
[Pd(dba)2]
PPh3
toluene
68
5
[Pd(dba)2]
JohnPhos
toluene
–
6
[Pd(dba)2]
PCy3
toluene
60
7
[Pd(dba)2]
Xantphos
toluene
32
8
[Pd(OAc)2]
PPh3
toluene
traces
9
[PdCl2(PPh3)2]
–
toluene
traces
10
[Pd(PPh3)4]
–
toluene
70 (67)c
The reactions were
carried out using
0.14 mmol of 1-bromo-2-((3-phenylprop-2-yn-1-yl)oxy)benzene (1a), 1.2 equiv of 3-methyl,-1,3-diphenylcyclobutan-1-ol (2a), and 1.2 equiv of Cs2CO3 in 4 mL
of dry solvent, under nitrogen atmosphere at 100 °C, in a Carius
tube for 16 h.
NMR yields
using trimethylbenzene-1,3,5-tricarboxylate
as standard.
Isolated yield.
The reactions were
carried out using
0.14 mmol of 1-bromo-2-((3-phenylprop-2-yn-1-yl)oxy)benzene (1a), 1.2 equiv of 3-methyl,-1,3-diphenylcyclobutan-1-ol (2a), and 1.2 equiv of Cs2CO3 in 4 mL
of dry solvent, under nitrogen atmosphere at 100 °C, in a Carius
tube for 16 h.NMR yields
using trimethylbenzene-1,3,5-tricarboxylate
as standard.Isolated yield.With the optimized conditions
in hand, we proceeded to study the
scope and limitations of the reaction. Several aspects were assessed:
the presence of electron-donating/withdrawing groups in the haloaryl
moiety, the nature and length of the chain tethering the internal
alkyne, and the use of different substituted cyclobutanols.The reactions of haloaryl ethers bearing methyl, methoxy, fluoro,
or trifluoromethyl substituents with the 3,3-substituted cyclobutanol 2a afforded good yields of the expected dihydrobenzofuran
derivatives 3b–3e (Scheme ). The pyridine derivative 1g gave rise to the heterocycle 3f, albeit in
moderate yield, perhaps due to competing coordination of the pyridine
moiety to Pd(II). C3-unsubstituted cyclobutanol derivatives 2 were also productive in the cascade reaction, giving the
functionalized dihydrobenzofuran derivatives 3g–j in comparable yields to those obtained with 2a (Scheme ); therefore,
the possible byproduct formation arising from β-H elimination
processes seem to be overridden. The cyclobutanol derivative bearing
a mesityl group in α-position led to mixtures where the desired
compound 3k could not be identified. The compound 3l could be isolated in 44% yield from the reaction carried
out employing the tertiary cyclobutanol bearing an i-Pr group.
Scheme 2
Scope of the Carbopalladation/Alkylation Cascade for
the Synthesis
of Dihydrobenzofurane Derivatives
Finally, the cross-coupling reactions of 2b and Me-
or TMS-substituted alkynyl substrates were tested. We observed that
among such substrates, only the silylated alkyne was competent to
deliver the desired product 3m in 56% yield (Scheme ). Possibly, the
substrate leading to 3n could experience a β-H
elimination upon the carbopalladation step to render an allenyl moiety,
as described in other Pd-catalyzed reactions dealing with alkyl-substituted
alkynes.[58,59]In order to assess the stereochemistry
of the exocyclic double
bond present in the dihydrobenzofuran cores, a NOESY NMR experiment
was carried out for compound 3d. The NOE contacts between
the methylene group CH2c and the o-H atoms
from the Ph ring, as well as the Ha of the heterocycle
with the CH2b group of the aliphatic chain, pointed out
the Z-stereochemistry for these compounds (Scheme ).
Scheme 3
Selected NOE Contacts
Observed for Dihydrobenzofurane and Oxindole
Derivatives
As a general feature of compounds 3a–3m, we observed their relative sensitivity
to chromatography
purification in either silica gel or alumina. The decomposition of
the compounds could be minored by using silica gel previously deactivated
with Et3N, and Et3N/hexane/EtOAc mixtures as
eluents. Solutions of these compounds in CDCl3 also evolved
to more complex mixtures over time (see the Supporting Information). The instability of these compounds might be due
to the migration of the exocyclic double bond to form benzofuran derivatives,
a process that could be catalyzed by Lewis acids.[60]We examined the influence of the length and nature
of the chain
linking the 2-haloryl and alkyne fragments. The alkenylated indoline
derivative 3o was obtained in good yield from the corresponding
amine precursor (Scheme ). Nevertheless, no desired product 3p was produced
from the related ester starting material. Substrates with one extra
carbon atom in the chain reacted smoothly under the optimized conditions
to produce the six-membered heterocycles 3q and 3r. The 1H NMR of the crude reaction mixture arising
from N-(2-bromo-phenyl)-N-methyl-3-phenylpropiolamide
showed the formation of the corresponding coupling product 3s as the main component, which could be isolated in 58% yield (Scheme ). Similarly, the
oxindole derivatives 3t and 3u could be
isolated in moderate yields from the reactions of the corresponding
propiolamides and the C3-unsubstituted cyclobutanol 2b. The 1H NMR spectra of compounds 3s–u showed an aromatic signal belonging to the oxindole core
at a relatively low chemical shift (5.8–6.0 ppm). This shielding
on Ha (compound 3u, Scheme ) is provoked by the phenyl ring on the exocyclic
olefine moiety, as observed in related structures reported in the
literature.[23,61,62] In addition, the NOESY NMR analysis of 3u also confirmed
the E-stereochemistry of the exocyclic double bond.
The presence of minor Z-stereoisomers in the reaction
mixtures leading to 3s–u cannot be
discarded; however, we were unable to isolate such minor components
of the crude mixtures and identify their nature unambiguously.
Scheme 4
Scope of the Carbopalladation/Alkylation Cascade Varying the Nature
of the Linking Chain
Scheme 5
Use of Propiolamide
Substrates
The plausible mechanistic pathway
for this reaction is depicted
in Scheme . The aryl-Pd
species A would form upon oxidative addition of the C–Br
bond present in the starting material 1a to Pd(0) (Chart ). Next, the intramolecular syn carbopalladation of the tethered alkyne would render
the intermediate B. At this stage, Cs2CO3 would assist the deprotonation of the cycloalkanol, along
with the removal of the halogen ligand from the coordination sphere,
allowing the formation of the alkoxide complex C. The
opening of the strained cycloalkanol through β-C cleavage would
render the σ-alkyl-Pd(II) intermediate D, from
which reductive elimination could take place to deliver the substituted
olefin 3a upon C(sp2)–C(sp3) bond formation.
Scheme 6
Proposed Reaction Mechanism
Chart 1
Structure and Numbering of the Staring Materials 1
The fact that propiolamide substrates afford
the E-alkenylated oxindoles 3s–u as main
coupling products reveals that in those cases the alkenyl-Pd(II) intermediate,
arising from the syn carbopalladation step, could
undergo an isomerization process. There are several precedents in
the literature of related Pd-catalyzed cascade reactions involving
the syn carbopalladation of alkynes and subsequent
isomerization prior to the final C–Pd bond functionalization.[14,22,25,63−67] Generally, the isomerization of the alkenyl-Pd intermediates is
driven by steric factors. Nevertheless, α-alkyl-substituted
alkynyl substrates, such as 1a, require the use of bulky
phosphine ligands (Q-Phos, X-Phos, or PBu3 among others) to increase the steric hindrance around
the Pd center and therefore promote the isomerization.[25,63,64] In the case of α-acyl-substituted
alkynyl substrates, such as propiolamides 1m–o, the isomerization is a frequent feature in a range of different
conditions, probably due to the conjugation of the alkenyl-Pd moiety
and the carbonyl group, which might lower the energy barrier for the
C–C rotation process (Scheme ).[28,62,68,69] Likely the coordination of the carbonyl
moiety might facilitate such processes. Nevertheless, the opposite
isomerization has been observed in related systems (that is, the steric
factors seemed to predominate over the possible coordination of the
carbonyl group in intermediates such as E).[68,69]We carried out the reaction of substrate 1b with
a
stoichiometric amount of [Pd(PPh3)4] in CH2Cl2 at 50 °C for 18 h under N2 atmosphere
(Scheme ). From the
reaction mixture, the vinyl-Pd(II) complex 4 (analogous
to the intermediate B) could be isolated in 84% yield.
The complex 4 was subsequently heated in toluene at 100
°C in the presence of cyclobutanol 2a and Cs2CO3. The 1H NMR spectra of the crude
reaction mixture confirmed the formation of the functionalized dihydrobenzofuran 3a in 70% yield.
Scheme 7
Synthesis of Intermediate B
The crystal structure of complex 4 was solved by X-ray
diffraction studies (Figure , Chart ).
The PPh3 ligands adopted a trans disposition.
The palladium atom was in a slightly distorted square-planar environment,
with a mean deviation of the Pd(II) coordination plane of 0.088 Å.
The exocyclic double bond exhibited a E geometry,
with the phenyl ring located cis to the methylene
group of the dihydrobenzofuran ring. The heterocyclic nucleus formed
angles of 38.1° and 77.1° with the phenyl substituent at
the double bond and the Pd(II) coordination plane, respectively. This
way, the phenyl ring was rotated 23.3° with respect to the exocyclic
double bond plane.
Figure 1
Thermal ellipsoid plot (50% probability)
of complex 4 along with the labeling scheme. The hydrogen
atoms have been omitted
for clarity. Selected bond lengths (Å) and angles (deg): Pd(1)–I(1)
= 2.6995(4), Pd(1)–P(1) = 2.3376(8), Pd(1)–P(2) = 2.3501(9),
Pd(1)–C(1) = 2.051(4), C(1)–C(2) = 1.339(5), C(1)–C(11)
= 1.505(5); I(1)–Pd(1)–P(1) = 90.85(2), P(1)–Pd(1)–C(1)
= 89.59(10), C(1)–Pd(1)–P(2) = 89.91(10), P(2)–Pd(1)–I(1)
= 90.15(2), C(2)–C(1)–Pd(1) = 123.4(3), C(2)–C(1)–C(11)
= 122.9(3), C(11)–C(1)–Pd(1) = 113.7(2).
Chart 2
Structure and Numbering of the Intermediate Complex 4
Thermal ellipsoid plot (50% probability)
of complex 4 along with the labeling scheme. The hydrogen
atoms have been omitted
for clarity. Selected bond lengths (Å) and angles (deg): Pd(1)–I(1)
= 2.6995(4), Pd(1)–P(1) = 2.3376(8), Pd(1)–P(2) = 2.3501(9),
Pd(1)–C(1) = 2.051(4), C(1)–C(2) = 1.339(5), C(1)–C(11)
= 1.505(5); I(1)–Pd(1)–P(1) = 90.85(2), P(1)–Pd(1)–C(1)
= 89.59(10), C(1)–Pd(1)–P(2) = 89.91(10), P(2)–Pd(1)–I(1)
= 90.15(2), C(2)–C(1)–Pd(1) = 123.4(3), C(2)–C(1)–C(11)
= 122.9(3), C(11)–C(1)–Pd(1) = 113.7(2).
Conclusion
In summary, we have expanded the versatility
of Pd cascades relying
on intramolecular carbopalladation processes through its merging with
the opening of strained cycloalkanols. Thus, the carbopalladation
of tethered alkynes followed by an alkylation process delivers interesting O- and N-heterocyclic cores bearing a fully
substituted exocyclic double bond. In addition, we observed a different
behavior of haloarylether and propiolamide substrates, being the last
ones prone to afford the coupling products arising from isomerization
of the alkenyl-Pd(II) intermediate.
Experimental
Section
General Remarks
Infrared spectra were recorded on a
PerkinElmer spectrum 100 spectrophotometer. High-resolution ESI mass
spectra were recorded on an Agilent 6220 Accurate Mass TOF LC-MS spectrometer.
Melting points were determined using a Reichert apparatus and are
uncorrected. Nuclear magnetic resonance (NMR) spectra were recorded
on a 300, 400, or 600 MHz Bruker NMR spectrometers in CDCl3 at 298 K (unless stated otherwise). All chemical shift values are
reported in parts per million (ppm) with coupling constant (J) values reported in Hz. All spectra were referenced to
TMS for 1H NMR and the CDCl3 solvent peak for 13C{1H} NMR. The anhydrous solvents were purchased
from commercial sources and used as received. TLC tests were run on
TLC Alugram Sil G plates and visualized under UV light at 254 nm.
Chromatography: Separations were carried out on silica gel. The general
procedures and characterization for the substrates 1a–o are included in the Supporting Information.
Representative Procedure A for the Synthesis
of the Carbopalladation/Alkylation
Cascade Products 3
A Carius tube equipped with
a magnetic stirrer was charged with [Pd(PPh3)4] (16 mg, 10 mol %), Cs2CO3 (51 mg, 0.17 mmol,
1.2 equiv), 3-methyl,-1,3-diphenylcyclobutan-1-ol (40 mg, 0.17 mmol,
1.2 equiv), and the corresponding substrate (1a) (40
mg, 0.14 mmol). The tube was set under a nitrogen atmosphere, and
dry toluene (4 mL) was added. The tube was sealed, and the reaction
mixture was stirred for 16 h at 100 °C. After cooling the tube,
the crude was diluted with CH2Cl2 (50 mL) and
filtered through a plug of Celite. The filtrate was concentrated under
a vacuum, and the crude mixture was purified by column chromatography
to afford the desired cascade product (3a). Compounds 3a–o are sensitive to purification in
silica gel chromatography; therefore, the silica gel was previously
deactivated with Et3N. In addition, n-hexane
containing 1% Et3N and EtOAc mixtures were used as eluents.
Prepared according to the representative
procedure A from 0.14 mmol of substrate 1o and 0.17 mmol
of 1-phenylcyclobutan-1-ol (2b). The crude was purified
by column chromatography over silica gel using 0 to 25% gradient EtOAc
in n-hexane containing 1% Et3N to afford
the heterocycle 3u as a yellow oil (25 mg, 0.06 mmol,
43%). IR (cm–1) ν̅ 1685 (s), 1602 (s),
1498 (s), 1338 (m), 1098 (m), 778 (s), 697 (s). 1H NMR
(300 MHz, CDCl3) δ 7.93 (m, 2 H), 7.55–7.48
(m, 4 H), 7.45–7.41 (m, 2 H), 7.27–7.25 (m, 3 H), 6.66
(d, J = 8.4 Hz, 1 H), 5.98 (d, J = 2 Hz, 1 H), 3.47–3.39 (m, 2 H), 3.24 (s, 3 H), 3.11 (t, J = 7.6 Hz, 2 H), 1.96 (q, J = 7.8 Hz,
2 H). 13C NMR (100.1 MHz, CDCl3) δ 199.8
(s, Cq), 167.4 (s, Cq), 160.3 (s, Cq), 140. Eight (s, Cq), 140.5 (s, Cq), 136.9
(s, Cq), 132.9 (s, CH), 129.3 (s, CH), 128.9 (s, CH), 128.5
(s, CH), 128.0 (s, CH), 127.9 (s, CH), 126.74 (s, Cq),
126.70 (s, CH), 123.9 (s, Cq), 123.24 (s, CH), 108.2 (s,
CH), 38.3 (s, CH2), 34.3 (s, CH3), 25.8 (s,
CH2), 22.4 (s, CH2). Some signals are overlapped.
HRMS (+ESI) m/z calculated for C26H23ClNO2 [M + H]+ 416.1412,
found 416.1421.
Synthesis of Complex 4
A Carius tube was
charged with the substrate 1b (100 mg, 0.30 mmol), [Pd(PPh3)4] (350 mg, 0.30 mmol), and a magnetic stirrer.
The tube was set under a nitrogen atmosphere, and dry CH2Cl2 was added (7 mL). The tube was sealed, and the mixture
was stirred at 50 °C for 18 h. After the tube was cooled, the
solution was filtered through a Celite plug. The filtrate was concentrated
to ca. 2 mL, and n-pentane (15 mL) was added. The
suspension was filtered, and the solid was washed with n-pentane (2 × 3 mL) and air-dried to give crude 4 as a bright yellow solid. Yield: 243 mg, 0.25 mmol, 84%. Crude complex 4 was recrystallized from CH2Cl2/Et2O to give analytically pure 4. mp 204 °C
(dec). 1H NMR (400.9 MHz, CDCl3) δ 9.24
(d, 3JHH = 7.2 Hz, 1 H, H6,
C6H4), 7.52–7.42 (m, 12 H, o-H, PPh3), 7.37–7.30 (m, 6 H, p-H, PPh3), 7.25–7.18 (m, 12 H, m-H, PPh3), 7.03 (td, 3JHH = 7.8, 4JHH = 1.2
Hz, 1 H, H4, C6H4), 6.98 (“t”, 3JHH = 7.3 Hz, 1 H, p-H, Ph), 6.87 (td, 3JHH =
7.4, 4JHH = 0.8 Hz, 1 H, H5,
C6H4), 6.84 (t, 3JHH = 7.7 Hz, 2 H, m-H, Ph), 6.51 (d, 3JHH = 7.3 Hz, 2 H, o-H, Ph), 6.45 (“d″, 3JHH = 7.7 Hz, 1 H, H3, C6H4), 4.35 (“t”, 2JHH = 3.2 Hz, 2 H, CH2). 13C NMR (100.8 MHz, CDCl3) δ 163.3
(s, C2), 155.5 (t, JPH = 2.1 Hz, Cq), 143.9 (t, JPH = 2.9 Hz, i-C, Ph), 135.2 (t, JPH = 5.9
Hz, o-CH, PPh3), 134.4 (t, JPH = 5.1 Hz, Cq), 131.9 (t, JPH = 22.9 Hz, i-C, PPh3),
130.2 (s, C1), 130.0 (s, p-CH, PPh3),
129.0 (s, o-CH, Ph), 128.7 (s, CH4, C6H4), 127.4 (t, JPH = 5.0 Hz, m-CH, PPh3), 126.9 (s, m-CH,
Ph), 125.6 (s, p-CH, Ph), 121.9 (s, CH6, C6H4), 119.4 (s, CH5, C6H4), 109.1
(s, CH3), 77.1 (s, CH2). IR (Nujol, cm–1) ν̅ 1590 (w), 1231 (m), 1093 (m), 742 (s), 691 (s),
520 (s), 509 (s), 494 (m). Anal. Calcd for C51H41IOP2Pd: C, 63.47; H, 4.28. Found: C, 63.55; H, 4.33.
Single-Crystal X-ray Structure Determination
Single
crystals of complex 4, suitable for an X-ray diffraction
study, were obtained by slow diffusion of n-pentane
into a solution of 4 in CHCl3.
Data Collection
A crystal suitable for X-ray diffraction
was mounted in inert oil on a glass fiber and transferred to a Bruker
diffractometer. Data were recorded at 100(2) K, using graphite-monochromated
Mo Kα radiation (λ = 0.71073 Å) and omega and phi
scan mode. Multiscan absorption correction was applied.
Structure
Solution and Refinements
The crystal structure
was solved by dual method, and all non-hydrogen atoms were refined
anisotropically on F2 using the program
SHELXL-2018/3.[70] Hydrogen atoms were refined
using the riding model.