HFIP Solvent Enables Alcohols to Act as Alkylating Agents in Stereoselective Heterocyclization

Article abstract of DOI:10.1021/jacs.9b02198

A new method for the stereoselective synthesis of highly functionalized oxygen heterocycles using allyl or benzyl alcohols as alkylating agents is presented. The process is efficient and atom economic, generating water as the only stoichiometric byproduct. Substoichiometric amounts of Ti(OiPr)₄ in HFIP solvent are key to this reactivity, and the method tolerates a broad substitution pattern on both the alcohol initiator and homoallylic alcohol substrate. Preliminary mechanistic studies reveal in situ formation of a titanium complex with HFIP which may initiate the cyclization reaction. Further stereoselective functionalization of the products allows access to a diverse range of interesting heterocyclic structures.

Full text of DOI:10.1021/jacs.9b02198

Communication
pubs.acs.org/JACS
Cite This: J. Am. Chem. Soc. XXXX, XXX, XXX−XXX
HFIP Solvent Enables Alcohols To Act as Alkylating Agents in
Stereoselective Heterocyclization
Yuxiang Zhu, Ignacio Colomer, Amber L. Thompson, and Timothy J. Donohoe*
Department of Chemistry, Chemistry Research Laboratory, University of Oxford, Mansfield Road, Oxford OX1 3TA, United
Kingdom
S
* Supporting Information
Scheme 1. Stereoselective Alkene Functionalization in HFIP
ABSTRACT: A new method for the stereoselective
synthesis of highly functionalized oxygen heterocycles
using allyl or benzyl alcohols as alkylating agents is
presented. The process is efficient and atom economic,
generating water as the only stoichiometric byproduct.
Substoichiometric amounts of Ti(OiPr)4 in HFIP solvent
are key to this reactivity, and the method tolerates a broad
substitution pattern on both the alcohol initiator and
homoallylic alcohol substrate. Preliminary mechanistic
studies reveal in situ formation of a titanium complex with
HFIP which may initiate the cyclization reaction. Further
stereoselective functionalization of the products allows
access to a diverse range of interesting heterocyclic
structures.
eveloping new synthetic methods for the stereo-
Dcontrolled synthesis of complex and highly function-
alized molecules from simple and readily available starting
HFIP solvent both allylic and benzylic alcohols can be used as
precursors to potent electrophiles11,12 which initiate cycliza-
materials is of value for applications in many disciplines such as
tion.13−17
medicinal chemistry or materials science.
In this regard, we have investigated the unique role that the
polar and hydrogen bonding solvent hexafluoroisopropanol
(HFIP) can play in facilitating stereoselective functionalization
reactions of alkenes. Our work has uncovered the stereo-
selective synthesis of cyclobutanes III from alkenes I and II in
a formal [2+2] cycloaddition, facilitated via a single electron
oxidative pathway (Scheme 1).1,2 Moreover, treating alkenes I
with a nucleophile (R−OH) and TEMPO in HFIP enabled us
to develop a new metal-free syn-dihydroxylation reaction (see
IV in Scheme 1), which proceeds via formation of an
oxoammonium cation intermediate.3
Herein we report the formation of both a C−C and a C−O
bond during heterocyclization to form saturated oxygen
containing heterocycles from homoallylic alcohols V (Scheme
1). The reaction is notable for its use of an alcohol (VI) as an
initiator for cyclization and for the high levels of stereo-
selectivity that ensue. HFIP is essential for this reaction to
proceed and this protocol is capable of assembling complex
heterocycles in one step. The overall process involves the
stereoselective creation of two new bonds in a single step, via a
formal 5-endo-trig cyclization,4,5 and produces water as the
only stoichiometric byproduct. While related approaches to
form tetrahydrofurans by cyclization have been reported,6−8
the use of carbon-based electrophiles to trigger cyclization is
rare and under-exploited.9,10 In this work, we show that in
We performed our initial reaction screen using homoallyl
alcohol 1a as the THF forming component (the alcohol
substrate which acts as a nucleophile) and allyl alcohol 2a as
the partner (this alcohol acts as the electrophile). Using HFIP
at room temperature gave tetrahydrofuran 3a in low yield, but
as a single trans diastereoisomer (Table 1, entry 1). Extensive
screening revealed Ti(Oi-Pr)4 as a promising additive, which
improved the yield with 10 mol % loading (Table 1, entry 2).
Details of the full screen of alternative solvents and additives
are given in the Supporting Information. Increasing the
reaction temperature to 40 and 70 °C also had a beneficial
impact with an encouraging 31% isolated yield (Table 1,
entries 3 and 4). Increasing the Ti(Oi-Pr)4 loading to 20 and
30 mol % notably improved the results with a respectable 52%
yield of 3a (Table 1, entries 5 and 6). Other Lewis acids, such
as Cu(OTf)2, Sc(OTf)2 or Zn(OAc)2 did not perform better
(Table 1, entries 7−9). Ionic additives such as Ph4P BF4,
designed to increase the cation stabilizing ability of the
solvent,18 were also screened with promising results (Table 1,
entry 10) although the starting homoallylic alcohol 1a was
completely consumed under these conditions; this was not the
case with the Ti(Oi-Pr)4 conditions. Notably, the use of other
Received: February 28, 2019
© XXXX American Chemical Society
A
DOI: 10.1021/jacs.9b02198
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
optimization using these model substrates were unfruitful, so
we opted to evaluate the scope of the method under the
optimal conditions. It is important to emphasize that there is
no need to use an excess of either component, leading to a very
efficient cyclization where every carbon on both components is
incorporated into the product in one step.
Table 1. Screening of Reaction Conditions
Entry
Additive
Temperature
Yield 3a (%)
The scope of this new transformation was first evaluated by
varying the nature of alcohol 2 (the electrophilic partner).
Using the more electron-rich 4-methoxycinnamyl alcohol
allowed the reaction to be carried out at room temperature
without harming the reactivity (Table 2, 3b). Benzylic
alcohols, such as 4-methoxybenzyl alcohol (Table 2, 3c) or
bis-benzylic, such as benzhydrol (Table 2, 3d) were used, with
a notable increase in yield. Expanding the scope to less
activated acyclic allylic alcohols, such as prenyl alcohol (Table
2, 3e) or diallyl alcohol (Table 2, 3f) was also possible,
allowing the incorporation of aliphatic (nonaromatic) sub-
stitution at C-3 in the tetrahydrofuran product, although with
diminished yield. The latter product 3f, was obtained as a
single regioisomeric conjugated diene product, in a net SN2′
fashion. In each case the product was isolated as a single, trans,
diastereoisomer.
1
2
3
4
5
6
7
8
9
−
RT
RT
3
11
15
31
38
52
0
10 mol % Ti(Oi-Pr)4
10 mol % Ti(Oi-Pr)4
10 mol % Ti(Oi-Pr)4
20 mol % Ti(Oi-Pr)4
30 mol % Ti(Oi-Pr)4
30 mol % Cu(Otf)2
30 mol % Sc(Otf)2
30 mol % Zn(OAc)2
40 °C
70 °C
70 °C
70 °C
70 °C
70 °C
70 °C
70 °C
70 °C
0
46
47
0
30 mol % Ph4P⊕BF4
30 mol % Ti(Oi-Pr)4
⊖
10
11 (DCM)
a
PMP: para-methyoxyphenyl.
solvents, for example dichloromethane, shut the reaction down
completely (Table 1, entry 11). Further attempts at the
Table 2. Scope of the Cyclization Reaction
a
b
c
Using 2.0 equiv of electrophilic partner 2. Performed at room temperature (RT). Performed at 0 °C.
B
DOI: 10.1021/jacs.9b02198
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
In an effort to increase the product complexity, we next
sought to utilize a set of racemic allylic alcohols as the
electrophile. We focused on acyclic alcohols with aromatic
(Table 2, 3g) or aliphatic substitution (Table 2, 3h) to avoid
unselective regiocontrol on a presumed allyl cationic
intermediate which initiates cyclization. Both alcohols
performed well in the reaction, with moderate to good
diasterocontrol at the new allylic stereocenter. We then moved
to the challenging unsymmetrically substituted allyl alcohol
(Table 2, 3i), where two separable regioisomeric products
were obtained in a 3:1 ratio, both of them as a 4:1 mixture of
diastereoisomers (at the allylic stereocenter). In spite of this
modest regio- and diasterocontrol only the 2,3-trans isomer
was observed within the THF ring. We then decided to make
use of cyclic allyl alcohols, first using cyclohexenyl and
cyclopentenyl alcohols as precursors of symmetric intermedi-
ates, and were delighted to find diastereocontrol (Table 2, see
3j and 3k). When using tertiary cyclopentenyl allyl alcohols,
with aliphatic (Table 2, 3l) or aromatic substitution (Table 2,
3m) complete regio- and high diastereocontrol over the new
allylic stereogenic center was obtained.
After showing that a broad variety of allyl and benzyl
alcohols act as competent electrophilic partners, we moved to
explore the versatility of the nucleophilic homoallyl alcohol 1.
Other less electron-rich aromatic rings attached to the alkene
were explored with satisfactory results using cinnamyl alcohol
as a partner (Table 2 3n and 3o). Electron-deficient (Table 2,
3p) and neutral aromatic rings (Table 2, 3q,r) were also
compatible substrates. Challenging aliphatic substitution (ie a
nonactivated alkene) at the C-2 position was attempted with
very satisfying results. While disubstituted alkenes did not
react, trisubstituted olefins gave valuable spirocycles (Table 2,
3s,t). Retaining the PMP unit at C-2, substitution at other
positions was then investigated. Aliphatic cyclic substitution at
C-5 produced complementary spirocycles in quantitative yields
(Table 2, 3u,v), while an acyclic variant gave valuable tertiary
centers (Table 2, 3w). Interesting benzofuran motifs can also
be obtained by using phenol-bearing starting materials (Table
2, 3x,y). Using substrates containing substituents at both C-4
and C-5 positions allowed us to access fully substituted
tetrahydrofurans, with the relative stereochemistry being
controlled by the C-4 substituent alone (see Table 2, 3z,aa).
It is noted that the method is not limited to the formation of
5-membered heterocycles; tetrahydropyrans can also be
obtained as a single trans isomer (Table 2, 3ab). In an effort
to expand the method to the synthesis of diverse and
interesting heterocycles, replacing the alcohols by the
corresponding carboxylic acids led to the production of γ-
butyrolactone and δ-valerolactone (Table 2, 3ac−ae). These
are common structures widely found in natural products and of
high relevance for medicinal chemistry.
example in enantiopure homoallyl alcohol 1q (Table 3).
Pleasingly, employing benzyl (Table 3, 3af,ag) or allyl (Table
Table 3. Synthesis of Enantiopure Products
3, 3ah) alcohols as electrophiles gave the enantiopure products
in a stereocontrolled manner. Thus, products with up to three
new stereogenic centers can be accessed in a single step by
using enantiopure homoallylic alcohols (Table 3, 3ai).
Further experiments were designed to shed light on the
mechanism of this new transformation. For those examples
that involve allyl alcohols as alkylating agents, using two
different regioisomeric allyl alcohols, 2a and 2a′, meant that
the same product trans-3aj was obtained in similar yields,
under identical conditions (Scheme 2A). These results support
Scheme 2. Experiments To Probe the Mechanism
the formation of a common allyl cationic intermediate, whose
formation is encouraged by the polar HFIP solvent. In
addition, the influence of double bond geometry within
homoallyl alcohol 1 was investigated. While the E isomer gives
the trans diastereoisomer with high stereoselectivity, the Z
isomer leads to a major, but not exclusive, cis product (Scheme
2B) and reacts via a less stereoselective pathway.
Finally, our attempts to detect reactive complexes and
intermediates in situ led to the isolation of a relatively stable
titanium complex that has been characterized by NMR (See
Supporting Information for further details; monomeric
structures are shown for clarity), whereby two isopropoxy
ligands of Ti(Oi-Pr)4 have been replaced by two hexafluor-
oisopropoxy units (Scheme 2C, 4). This type of Ti complex
Note that NOE experiments revealed the stereochemical
relationship between the THF ring substituents; the relative
stereochemistry of the products bearing an exocyclic stereo-
genic center (cyclic or acyclic) was assigned by reference to X-
ray crystallographic structures obtained using 3v or 10c,19 see
Supporting Information for details. Interestingly, these two X-
ray structures reveal differing relative stereochemistry at the
exocyclic carbon, which likely has its origins in the differing
geometries of the acyclic versus cyclic cation precursors.
Encouraged by the high stereoselectivity and broad scope
shown, we were particularly interested in exploiting the transfer
of chirality from a previous established stereogenic center, for
C
DOI: 10.1021/jacs.9b02198
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
has some precedent in the literature, although its properties
and applications remain unexplored;20 a related ligand
exchange using MoCl5 has recently been reported.21 When
we mixed preprepared titanium complex 4 and allyl alcohol 2a
in HFIP, a new species 5, derived from the substitution of one
hexafluoroisopropoxy moiety for an allylalkoxy group, was
detected by NMR spectroscopy (Scheme 2C, see Supporting
Information for further details). We note that alkoxy ligand
exchange at titanium is likely to be easy to accomplish and
reversible.22 Finally, the addition of homoallyl alcohol 1a to
this reaction mixture led to the formation of product 3a,
presumably via an electrophile-triggered cyclization.
Finally, selective functionalization of the heterocyclic
products 3 offers an opportunity to install groups that can
be used to manipulate these highly functionalized heterocycles.
Our reaction toolbox includes oxidation protocols, for example
ozonolysis of the remaining alkene afforded an aldehyde
(Scheme 3a, 7). Moreover, the PMP group can be oxidized to
further functionalization allows us to install useful functional
groups.
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/jacs.9b02198.
■
S
Full experimental details (PDF)
Copies of spectral data (PDF)
Crystallographic data (CIF)
AUTHOR INFORMATION
Corresponding Author
*timothy.donohoe@chem.ox.ac.uk
■
ORCID
Ignacio Colomer: 0000-0001-5542-7034
Timothy J. Donohoe: 0000-0001-7088-6626
Notes
a
Scheme 3. Selective THF Product Functionalization
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We thank the European Union and the European Commission
- Seventh Framework People Programme (Marie Curie
Actions) for financial support (IC) and Diamond Light Source
for an award of beamtime (MT13639).
REFERENCES
■
(1) Colomer, I.; Coura Barcelos, R.; Donohoe, T. J. Angew. Chem.,
Int. Ed. 2016, 55, 4748−4752.
(2) Colomer, I.; Batchelor-McAuley, C.; Odell, B.; Donohoe, T. J.;
Compton, R. G. J. Am. Chem. Soc. 2016, 138, 8855−8861.
(3) Colomer, I.; Coura Barcelos, R.; Christensen, K.; Donohoe, T. J.
Org. Lett. 2016, 18, 5880−5883.
(4) Baldwin, J. E.; Cutting, J.; Dupont, W.; Kruse, L.; Silberman, L.;
Thomas, R. C. J. Chem. Soc., Chem. Commun. 1976, 736−738.
(5) Alabugin, I. V.; Gilmore, K. Chem. Commun. 2013, 49, 11246−
11250.
(6) Ishibashi, H.; Sato, T.; Ikeda, M. Synthesis 2002, 2002, 695−713.
(7) Ichikawa, J.; Lapointe, G.; Iwai, Y. Chem. Commun. 2007, 2698−
2700. and references therein.
(8) Knight, D. W. Electrophile-induced 5-endo cyclizations. In
Progress in Heterocyclic Chemistry; Gribble, G. W., Gilchrist, T. L.,
Eds.; Elsevier, 2002; Vol. 14, Chapter 2, pp 19−51.
a
(i) O3, CH2Cl2, −78 °C/DMS; (ii) RuCl3 (2.5−5 mol %), NaIO4,
Na2HPO4, MeCN:CCl4; (iii) H2, 10 wt % Pd/C, EtOAc, 0 °C; (iv)
1.1−1.7 equiv. LDA, 1−1.5 equiv. R2−X, THF, −78 °C.
(9) For alternative methods for making rings and a new C−C bond
at the same time, see: (a) Saito, A.; Matsushita; Kaneko, H. Chem.
Lett. 1983, 12, 729−732. (b) Bergeron, S.; Brigaud, T.; Foulard, G.;
Plantier-Royon, R.; Portella, C. Tetrahedron Lett. 1994, 35, 1985−
1988. (c) Protti, S.; Fagnoni, M.; Albini, A. J. Am. Chem. Soc. 2006,
a carboxylic acid when attached directly to the heterocycle
(Scheme 3a, 8), or when derived from the cation electrophile
(Scheme 3a, 9). Alternatively, hydrogenation of the double
bond in these products gave saturated derivatives (see 6a,b in
Scheme 3b). The diastereoselective transformation of 3ai to 6c
is noteworthy, creating a new stereogenic center during
reduction. Finally, stereoselective alkylation or acylation can
be performed on the γ-butyrolactones (Scheme 3c, 10a−c).
These straightforward and stereoselective reactions performed
using heterocycles 3 exemplify the rich potential of these
compounds for synthetic organic chemistry.
In conclusion, we have developed a new method for the
stereoselective preparation of valuable oxygen-containing
heterocyclic compounds triggered by the use of alcohols as
alkylating agents. HFIP solvent is unique in promoting the
heterocyclization and allowing formation of a C−C and C−O
bond at the same time. Moroever, enantioenriched products
can be obtained from simple chiral starting materials and
́
128, 10670−10671. (d) Mullen, C. A.; Gagne, M. R. J. Am. Chem. Soc.
2007, 129, 11880−11881. (e) Lin, R.; Sun, H.; Yang, C.; Shen, W.;
Xia, W. Chem. Commun. 2015, 51, 399−401. (f) Shimogaki, M.;
Fujita, M.; Sugimura, T. Angew. Chem., Int. Ed. 2016, 55, 15797−
15801. (g) Jian, W.; Ge, L.; Jiao, Y.; Qian, B.; Bao, H. Angew. Chem.,
Int. Ed. 2017, 56, 3650−3654.
(10) For Prins type processes, see: (a) Yadav, J. S.; Borkar, P.;
Chakravarthy, P. P.; Reddy, B. V. S.; Sarma, A. V. S.; Basha, S. J.;
Sridhar, B.; Gree, R. J. Org. Chem. 2010, 75, 2081−2084. (b) Reddy,
B. V. S.; Borkar, P.; Yadav, J. S.; Reddy, P. P.; Kunwar, A. C.; Sridhar,
B.; Gree, R. Org. Biomol. Chem. 2012, 10, 1349−1358. (c) Srivastava,
N.; Reddy, P. P.; Sridhar, B.; Reddy, B. V. S. Synlett 2013, 24, 1263−
1268. (d) Reddy, B. V. S.; Reddy, M. R.; Sridhar, B.; Reddy, C. S.
Tetrahedron Lett. 2014, 55, 4110−4113.
́
(11) For recent reviews, see: (a) Baeza, A.; Najera, C. Synthesis
2013, 46, 25−34. (b) Dryzhakov, M.; Richmond, E.; Moran, J.
Synthesis 2016, 48, 935−959.
D
DOI: 10.1021/jacs.9b02198
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
(12) (a) Jin, J.; MacMillan, D. W. C. Nature 2015, 525, 87.
(b) Nacsa, E. D.; MacMillan, D. W. C. J. Am. Chem. Soc. 2018, 140,
3322−3330.
(13) Colomer, I.; Chamberlain, A.; Haughey, M.; Donohoe, T. J.
Nat. Rev. Chem. 2017, 1, 0088. and references therein.
́
(14) Trillo, P.; Baeza, A.; Najera, C. J. Org. Chem. 2012, 77, 7344−
7354.
(15) Muniz, K.; Barreiro, L.; Romero, R. M.; Martínez, C. J. Am.
̃
Chem. Soc. 2017, 139, 4354−4357.
(16) Qi, C.; Hasenmaile, F.; Gandon, V.; Lebœuf, D. ACS Catal.
2018, 8, 1734−1739.
́
(17) Roy, S.; Motiwala, H. F.; Koshlap, K. M.; Aube, J. Eur. J. Org.
Chem. 2018, 2018, 306−315.
(18) Tian, Y.; Xu, X.; Zhang, L.; Qu, J. Org. Lett. 2016, 18, 268−271.
(19) (a) Diffraction data were collected using a (Rigaku) Oxford
Diffraction SuperNova or I19-1 at Diamond Light Source. Structures
were solved using SuperFlip and refined using CRYSTALS as per the
CIF. Crystallographic data (in CIF format) are available as associated
content and have been deposited with Cambridge Crystallographic
Data Centre (reference codes CCDC 1895660−1); these data can
also be obtained free of charge via www.ccdc.cam.ac.uk/data_-
request/cif. (b) Allan, D. R.; Nowell, H.; Barnett, S. A.; Warren, M.
R.; Wilcox, A.; Christensen, J.; Saunders, L. K.; Peach, A.; Hooper, M.
T.; Zaja, L.; Patel, S.; Cahill, L.; Marshall, R.; Trimnell, S.; Foster, A.
J.; Bates, T.; Lay, S.; Williams, M. A.; Hathaway, P. V.; Winter, G.;
Gerstel, M.; Wooley, R. W. Crystals 2017, 7 (11), 336. (c) Palatinus,
L.; Chapuis, G. J. Appl. Crystallogr. 2007, 40, 786−790. (d) Betteridge,
P. W.; Carruthers, J. R.; Cooper, R. I.; Prout, K.; Watkin, D. J. J. Appl.
Crystallogr. 2003, 36, 1487. (e) Cooper, R. I.; Thompson, A. L.;
Watkin, D. J. J. Appl. Crystallogr. 2010, 43, 1100−1107.
(20) (a) Blackmore, T.; Bruce, M. I.; Davidson, P. J.; Iqbal, M. Z.;
Stone, F. G. A. J. Chem. Soc. A 1970, 3153−3158. (b) Nomura, K.;
Pracha, S.; Phomphrai, K.; Katao, S.; Kim, D.-H.; Kim, H. J.; Suzuki,
N. J. Mol. Catal. A: Chem. 2012, 365, 136−145. (c) Nomura, K.;
Tewasekson, U.; Takii, Y. Organometallics 2015, 34, 3272−3281.
(21) Schubert, M.; Leppin, J.; Wehming, K.; Schollmeyer, D.;
Heinze, K.; Waldvogel, S. R. Angew. Chem., Int. Ed. 2014, 53, 2494−
2497.
(22) Woodard, S. S.; Finn, M. G.; Sharpless, K. B. J. Am. Chem. Soc.
1991, 113, 106−113.
E
DOI: 10.1021/jacs.9b02198
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX