Distal-selective hydroformylation using scaffolding catalysis

Article abstract of DOI:10.1021/ja504247g

In hydroformylation, phosphorus-based directing groups have been consistently successful at placing the aldehyde on the carbon proximal to the directing group. The design and synthesis of a novel catalytic directing group are reported that promotes aldehyde formation on the carbon distal relative to the directing functionality. This scaffolding ligand, which operates through a reversible covalent bond to the substrate, has been applied to the diastereoselective hydroformylation of homoallylic alcohols to afford δ-lactones selectively. Altering the distance between the alcohol and the olefin revealed that homoallylic alcohols gives the distal lactone with the highest levels of regioselectivity.

Full text of DOI:10.1021/ja504247g

Communication
pubs.acs.org/JACS
Distal-Selective Hydroformylation using Scaffolding Catalysis
Candice L. Joe, Thomas P. Blaisdell, Allison F. Geoghan, and Kian L. Tan*,†
Department of Chemistry, Merkert Chemistry Center, Boston College, Chestnut Hill, Massachusetts 02467, United States
S
* Supporting Information
Table 1. Optimization of Ligand Structure
ABSTRACT: In hydroformylation, phosphorus-based
directing groups have been consistently successful at
placing the aldehyde on the carbon proximal to the
directing group. The design and synthesis of a novel
catalytic directing group are reported that promotes
aldehyde formation on the carbon distal relative to the
directing functionality. This scaffolding ligand, which
operates through a reversible covalent bond to the
substrate, has been applied to the diastereoselective
hydroformylation of homoallylic alcohols to afford δ-
lactones selectively. Altering the distance between the
alcohol and the olefin revealed that homoallylic alcohols
gives the distal lactone with the highest levels of
regioselectivity.
he functionalization of carbon−carbon double bonds
Tserves as a bedrock for synthetic chemistry. The ease of
Scheme 1. Scaffold Controlled Regioselective
a
b
c
entry
substrate
ligand
% conv
rs (p:d)
dr (a:s)
Hydroformylation
1
2
3
4
5
6
7
8
9
rac-1
rac-1
PPh3
2
30
60
65
87
55
92
63
88
44
95
46:54
19:81
28:72
9:91
53:47
76:24
67:33
88:12
74:26
87:13
84:16
85:15
58:42
87:13
(S)-1
(R)-1
(R)-1
(R)-1
(R)-1
(R)-1
(R)-1
(R)-1
3b
3b
3a
3c
3d
3e
4
24:76
9:91
14:86
10:90
44:56
9:91
d
10
3c
a
Determined by 1H NMR of the crude hydroformylation mixture
b
using mesitylene as an internal standard. Regioselectivity (prox-
imal:distal) determined by gas chromatography of the crude reaction
synthesis, ubiquity, and low cost of olefins makes them ideal
building blocks for organic synthesis. A critical challenge in
olefin functionalization reactions is the control of regioselec-
tivity. Often reactions have an inherent regio-preference that
can be exploited or amplified; however, in these cases, accessing
the inherently less favorable product is challenging. Olefins,
such as 1,2-disubstituted olefins, pose an alternative problem in
that there is often no electronic or steric differentiation of the
olefinic carbons; this presents the onerous task of synthesizing
either isomer of product selectively. To fully realize the
potential of olefin functionalization reactions, our goal is to
develop a general strategy that can overcome innate substrate
bias and, in the case of unbiased substrates, is able to
predictably synthesize either structural isomer of the product.
Hydroformylation is a classic example of an olefin
functionalization reaction where the control of regioselectivity
c
mixture after PCC oxidation. Diastereomer ratio (anti:syn) as
determined by 1H NMR (CD3OD) of the reaction mixture after
d
PCC oxidation. Reaction run using 0.10 mol % p-TsOH.
is the preeminent challenge. The hydroformylation of olefins
generally prefers to form the aldehyde on the least hindered
carbon.1 Over the last several decades, branch selective
hydroformylation has been developed in the context of
enantioselective hydroformylation by exploiting substrates
that have an electronic bias to form the more hindered
aldehyde.2 Landis3 and Zhang4 have shown that substrates
bearing weakly Lewis basic functional groups can produce
Received: April 28, 2014
© XXXX American Chemical Society
A
dx.doi.org/10.1021/ja504247g | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
Table 2. Substrate Scope
a
Regioselectivities (proximal:distal) and diastereomer ratio (anti:syn) were determined by GC or 1H NMR of the crude reaction mixtures after PCC
b
c
oxidation. Isolated yield of combined distal lactone products. (i) 10 mol % 3c, 0.10 mol % p-TsOH, 45 °C, benzene; (ii) 3 mol % Rh(acac)(CO)2,
55 °C, 400 psi H2/CO, benzene; (iii) PCC, NaOAc, 3 Å sieves, DCM. Standard conditions except 20 mol % 3c and 6 mol % Rh(acac)(CO)2 were
used. Standard conditions except 12 mol % 3c and 4 mol % Rh(acac)(CO)2 were used. Standard conditions except the hydroformylation was run at
35 °C using 5 mol % 3c and 2 mol % Rh(acac)(CO)2.
d
e
f
branched aldehydes with practical levels of regio- and
enantioselectivity. In a pioneering report, Clarke reported
significant branched and enantioselectivity for unactivated
terminal olefins (e.g., 1-hexene) using ligand control.5 In an
alternative strategy, Reek et al. demonstrated control of regio-
and enantioselectivity in unactivated disubstituted olefins using
supramolecular catalysis.6
Stoichiometric phosphorus-based directing groups are
effective at accessing the branched isomer for terminal alkenes7
as well as favoring aldehyde formation on the carbon proximal
to the directing group for more highly substituted olefins.8
Improving upon the overall atom efficiency of this strategy, the
Breit group9 and our group10 reported the first catalytic
phosphorus-based directing groups in 2008.11 Although a
reliable strategy, the use of directing groups has been limited to
placing the aldehyde in the proximal position relative to the
directing functionality.12 Recently, this deficiency has been
partially addressed by using ligands that can hydrogen bond to
olefin substrates bearing a carboxylic acid functional group,
enabling the aldehyde to be formed on the carbon distal to the
directing functionality.13 Here, we report a novel scaffolding
ligand that utilizes reversible covalent bonding to perform distal
and diastereoselective hydroformylation of homoallylic alcohols
toward the synthesis of δ-lactones (Scheme 1).
Combining the design elements from the Breit supra-
molecular catalyst13a and our original azaphosphole ligand
(scaffold A),10a we targeted a small collection of ligands with
the substructure of scaffold B. The critical features for these
ligands are an oxazolidine group that our group has previously
shown to bind to alcohols14 and a triaryl phosphine, which
serves as the metal binding site. This overall series has a
modular synthesis allowing for both the tuning of the steric and
electronic properties of the phosphine as well as the positioning
of the substrate and metal-binding sites.
We initially tested the new ligands in the hydroformylation of
homoallylic alcohols bearing an allylic substituent (Table 1).
Using PPh3 as a control ligand, a slight preference exists for the
formation of the distal product with minimal levels of
diastereocontrol and low conversion (Table 1, entry 1); these
results highlight the dual challenge of achieving high levels of
reactivity and selectivity for disubstituted olefins in hydro-
formylation. Employing scaffolding ligand 2 (entry 2) results in
a significant increase in conversion and regio- and diaster-
eoselectivity. Next, chiral scaffolding ligand 3b, which is derived
B
dx.doi.org/10.1021/ja504247g | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
formed selectively (20:80 rs), the syn-isomer is observed as the
major stereoisomer. Notably, the magnitude of diastereose-
lection for the E-olefin in favor of the syn product (32:68) is
lower than that of the corresponding Z-olefin substrate for the
formation of the anti isomer (87:13). This result is consistent
with decreased A1,3-strain between the respective hydrogen and
butyl substituents interacting with the allylic methyl group.
A substrate bearing a phenyl substituent at the allylic position
underwent the directed hydroformylation efficiently in high
levels of selectivity (entry 4). Various electron-deficient and
electron-rich para-substituted (entries 5−7) and meta-sub-
stituted (entries 8,9) aromatic substituents were well tolerated
and afforded the δ-lactones in a regio- and diastereoeselective
fashion. Highlighting the functional group tolerance of the
reaction, a substrate bearing a silyl ether substituent at the
allylic position gave the anti δ-lactone product in a selective
fashion (entry 10).17 Employing a terminal olefin substrate
(entry 11), the inherent regiochemical preference can be
amplified, in comparison to a control reaction where PPh3 is
used as ligand, to afford the six-membered lactone as the major
product.18
A critical question for this new ligand class is where the
regioselectivity in the hydroformylation reaction originates
from. In part, we hypothesize that the preference for the δ-
lactone vs the γ-lactone is the formation of a less strained, larger
metallacycle during the catalytic cycle. This is analogous to
carbocycles that exhibit a significant drop in ring strain moving
from 11- to 12-membered rings. To probe this question, we
investigated the hydroformylation of allylic, homoallylic, and
bis-homoallylic alcohols (Table 3). Allylic alcohols undergo
hydroformylation with a slight preference for the proximal
product in the presence of ligand 3c, which is similar to the
regiochemical preference employing PPh3 as ligand (Table 3,
entry 1). Relative to the control reaction with PPh3, homoallylic
alcohols show a significant preference for the δ-lactone product
(entry 2). Interestingly, under these conditions a low level of
enantioselectivity (19% ee) is observed. Decreasing the
pressure to 100 psi H2/CO increases the enantioselectivity to
37% ee with similar levels of regioselectivity for the δ-lactone.19
Notably, the reaction with PPh3 is sluggish (40% conversion),
which highlights the rate accelerating affect in the presence of
ligand 3c. Minimal levels of regiocontrol for the distal product
are observed in the case of a bis-homoallylic alcohol substrate
(entry 3), but enhanced conversion is observed. These results
are most consistent with a directed reaction that is nonselective,
but an undirected reaction cannot be ruled out at this time.
Overall these results demonstrate that the regiochemical
outcome of the reaction is highly dependent on the appropriate
choice of tether length between the olefin and the phosphine
on the ligand. This is in contrast to proximal selective directing
groups, which generally are more promiscuous with respect to
substrate class.
Table 3. Importance of Substrate Tether
a
Conversion based on remaining starting material after the hydro-
formylation reaction using mesitylene as an internal standard.
b
Regioselectivities (proximal:distal) were determined by 1H NMR.
c
Crude hydroformylation reaction was subjected to Pinnick oxidation
d
(see Supporting Information for experimental details). Crude
hydroformylation reaction was subjected to PCC oxidation (see
e
Supporting Information for details). No derivatization of the crude
f
hydroformylation reaction was carried out. Results in parentheses run
under identical conditions, except PPh3 was used rather than ligand 3c.
from L-valine, was synthesized. A matched/mis-matched
relationship is observed when employing enantiopure variants
of the substrate (entries 3 and 4), with the (R)-isomer being
matched. Reducing the size of the substituent on the
oxazolidine backbone to a methyl group (entry 5) results in
lower conversion and selectivities; however, ligand 3c bearing
an oxazolidine backbone derived from L-tert-leucine affords high
levels of regio- and diastereoselectivity (entry 6). Ligand 3c is
also the most active, giving 92% conversion in the hydro-
formylation reaction. With the tert-leucine backbone in hand,
the electronics on the phosphine were perturbed. Both
electron-rich (ligand 3d) and electron-deficient (ligand 3e)
phosphines gave inferior results when compared to electron
neutral ligand 3c. In particular, both the conversion and
diastereoselectivity during the hydroformylation reaction
suffered (entries 7 and 8). Using ligand 4, which contains no
substrate-binding site, the reaction proceeds sluggishly; more-
over, both the regio- and diastereoselectivities are comparable
to the results utilizing PPh3 as ligand (entry 9). This
demonstrates the importance of the substrate-binding site for
both achieving high levels of selectivity and obtaining rate
acceleration in the reaction. Employing a catalytic amount of p-
TsOH during the reaction afforded nearly identical results
(entry 10) but proved to be a necessary additive to obtain
reproducible conversions and selectivities.
In the context of hydroformylation of 1,2-disubstituted
olefins, the challenge of directing to the carbon distal to the
directing group has gone relatively unaddressed. We have
developed a highly tunable catalytic directing group, which can
carry out the distal and diastereoselective hydroformylation of
homoallylic alcohols. Together with previous examples of
proximal-selective hydroformylation, these results begin to
more fully address the challenge of generally controlling
regioselectivity in olefin functionalization reactions.
The selective formation of the anti diastereomer of the δ-
lactone (Table 2, entry 1) is attributed to the minimization of
A1,3-strain in the hydrometalation step of the reaction.15 In
support of this hypothesis, the diastereoselectivity is amplified
in favor of the anti diastereomer (91:9 dr) when a larger
isopropyl substituent is placed in the allylic position (Table 2,
entry 2).16 To further demonstrate the effect of A1,3-strain on
diastereoselectivity, an E-configured olefin substrate was
examined (Table 2, entry 3). While the δ-lactone is still
C
dx.doi.org/10.1021/ja504247g | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
J.; Park, J. W.; Jun, C. H. Acc. Chem. Res. 2008, 41, 222. (e) Bedford, R.
B.; Coles, S. J.; Hursthouse, M. B.; Limmert, M. E. Angew. Chem., Int.
Ed. 2003, 42, 112. (f) Bedford, R. B.; Limmert, M. E. J. Org. Chem.
2003, 68, 8669. (g) Oi, S.; Watanabe, S. I.; Fukita, S.; Inoue, Y.
Tetrahedron Lett. 2003, 44, 8665. (h) Carrion, M. C.; Cole-Hamilton,
́
D. J. Chem. Commun. 2006, 4527. (i) Lewis, J. C.; Wu, J.; Bergman, R.
G.; Ellman, J. A. Organometallics 2005, 24, 5737.
ASSOCIATED CONTENT
* Supporting Information
Ligand syntheses, experimental details, compound character-
ization, and determination of absolute configurations. This
material is available free of charge via the Internet at http://
pubs.acs.org.
■
S
(12) A notable example of distal selective hydroformylation using a
stoichiometric phosphorus-based directing group was reported by
Burke et al. in the total synthesis of (+)-phyllanthocin: Burke, S. D.;
Cobb, J. E. Tetrahedron Lett. 1986, 27, 4237.
AUTHOR INFORMATION
Corresponding Author
kian.tan@novartis.com
■
̌
(13) (a) Smejkal, T.; Breit, B. Angew. Chem., Int. Ed. 2008, 47, 311.
Present Address
(b) Dydio, P.; Reek, J. N. H. Angew. Chem., Int. Ed. 2013, 52, 3878.
†Novartis, 250 Massachusetts Ave., Cambridge, MA 02138.
̌
(c) Smejkal, T.; Gribkov, D.; Geier, J.; Keller, M.; Breit, B. Chem.
Eur. J. 2010, 16, 2470. (d) Dydio, P.; Dzik, W. I.; Lutz, M.; de Bruin,
B.; Reek, J. N. H. Angew. Chem., Int. Ed. 2011, 50, 396.
(14) (a) Sun, X.; Worthy, A. D.; Tan, K. L. Angew. Chem., Int. Ed.
2011, 50, 8167. (b) Worthy, A. D.; Sun, X.; Tan, K. L. J. Am. Chem.
Soc. 2012, 134, 7321. (c) Giustra, Z. X.; Tan, K. L. Chem. Commun.
2013, 49, 4370. (d) Sun, X.; Lee, H.; Lee, S.; Tan, K. L. Nat. Chem.
2013, 5, 790. (e) Blaisdell, T. P.; Lee, S.; Kasaplar, P.; Sun, X.; Tan, K.
L. Org. Lett. 2013, 15, 4710.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We thank the Alfred P. Sloan Foundation (K.L.T.), the John
LaMattina Graduate Fellowship (C.L.J.), and the NIGMS
(RO1GM087581) for funding this project. Mass spectrometry
instrumentation at Boston College is supported by funding
from the NSF (DBI-0619576).
(15) Absolute configuration of the major δ-lactone product was
further confirmed by 1D NOE experiments. See Supporting
Information for details.
(16) The low isolated yield can be attributed to incomplete
converson (72%) during the hydroformylation reaction. In the
presence of 12 mol % PPh3, 6 mol % Rh(acac)(CO)2, 75 °C, 400
psi H2/CO, benzene no appreciable conversion was observed.
(17) The low isolated yield can be attributed to incomplete
conversion (75%) during the course of the hydroformylation reaction.
(18) In the presence of PPh3 under identical conditions, 28:72 rs
(prox:dist) is obtained.
(19) Landis has also seen a similar pressure effect on
enantioselectivity in the asymmetric hydroformylation of styrene:
Watkins, A. L.; Landis, C. R. J. Am. Chem. Soc. 2010, 132, 14352. See
Supporting Information for further details on pressure effects on the
enantioselectivity.
REFERENCES
■
(1) Franke, R.; Selent, D.; Borner, A. Chem. Rev. 2012, 112, 5675.
(2) (a) Klosin, J.; Landis, C. R. Acc. Chem. Res. 2007, 40, 1251.
̈
(b) Gual, A.; Godard, C.; Castillon
Asymmetry 2010, 21, 1135. (c) Dieguez, M.; Pam
Tetrahedron: Asymmetry 2004, 15, 2113. (d) Agbossou, F.; Carpentier,
J. F.; Mortreux, A. Chem. Rev. 1995, 95, 2485. (e) Claver, C.; Dieguez,
M.; Pamies, O.; Castillon, S. Top. Organomet. Chem. 2006, 18, 35.
(f) Jouffroy, M.; Gramage-Doria, R.; Armspach, D.; Semeril, D.;
́
, S.; Claver, C. Tetrahedron:
́
̀
ies, O.; Claver, C.
́
̀
́
́
Oberhauser, W.; Matt, D.; Toupet, L. Angew. Chem., Int. Ed. 2014, 53,
3937.
(3) McDonald, R. I.; Wong, G. W.; Neupane, R. P.; Stahl, S. S.;
Landis, C. R. J. Am. Chem. Soc. 2010, 132, 14027.
(4) Zhang, X.; Cao, B.; Yu, S.; Zhang, X. Angew. Chem., Int. Ed. 2010,
49, 4047.
(5) Noonan, G. M.; Fuentes, J. A.; Cobley, C. J.; Clarke, M. L. Angew.
Chem., Int. Ed. 2012, 51, 2477.
(6) (a) Kuil, M.; Soltner, T.; van Leeuwen, P. W. N. M.; Reek, J. N.
H. J. Am. Chem. Soc. 2006, 128, 11344. (b) Gadzikwa, T.; Bellini, R.;
Dekker, H. L.; Reek, J. N. H. J. Am. Chem. Soc. 2012, 134, 2860.
(c) Bellini, R.; Reek, J. N. H. Chem.Eur. J. 2012, 18, 7091.
(7) (a) Jackson, W. R.; Perlmutter, P.; Suh, G.-H. J. Chem. Soc. Chem.
Commun. 1987, 21, 724. (b) Krauss, I. J.; Wang, C. C.-Y.; Leighton, J.
L. J. Am. Chem. Soc. 2001, 123, 11514.
(8) Breit, B.; Grunanger, C. U.; Abillard, O. Eur. J. Org. Chem. 2007,
̈
15, 2497.
(9) (a) Breit, B.; Demel, P.; Gebert, A. Chem. Commun. 2004, 24,
114. (b) Grunanger, C. U.; Breit, B. Angew. Chem., Int. Ed. 2008, 47,
̈
7346. (c) Usui, I.; Nomura, K.; Breit, B. Org. Lett. 2011, 13, 612.
(d) Grunanger, C. U.; Breit, B. Angew. Chem., Int. Ed. 2010, 49, 967.
̈
(e) Ueki, Y.; Ito, H.; Usui, I.; Breit, B. Chem.Eur. J. 2011, 17, 8555.
(10) (a) Lightburn, T. E.; Dombrowski, M. T.; Tan, K. L. J. Am.
Chem. Soc. 2008, 130, 9210. (b) Worthy, A. D.; Gagnon, M. M.;
Dombrowski, M. T.; Tan, K. L. Org. Lett. 2009, 11, 2764. (c) Sun, X.;
Frimpong, K.; Tan, K. L. J. Am. Chem. Soc. 2010, 132, 11841.
(d) Worthy, A. D.; Joe, C. L.; Lightburn, T. E.; Tan, K. L. J. Am. Chem.
Soc. 2010, 132, 14757. (e) Joe, C. L.; Tan, K. L. J. Org. Chem. 2011, 76,
7590. (f) Lightburn, T. E.; De Paolis, O. A.; Cheng, K. H.; Tan, K. L.
Org. Lett. 2011, 13, 2686.
(11) Examples of reversible covalent bonding in other metal-
catalyzed reactions: (a) Lewis, L. N. Inorg. Chem. 1985, 24, 4433.
(b) Lewis, L. N.; Smith, J. F. J. Am. Chem. Soc. 1986, 108, 2728.
(c) Preston, S. A.; Cupertino, D. C.; Palma-Ramirez, P.; Cole-
Hamilton, D. J. J. Chem. Soc., Chem. Commun. 1986, 977. (d) Park, Y.
D
dx.doi.org/10.1021/ja504247g | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX