Acceleration of Enantioselective Cycloadditions Catalyzed by Second-Generation Chiral Oxazaborolidinium Triflimidates by Biscoordinating Lewis Acids

Article abstract of DOI:10.1021/jacs.6b08018

The activation of second-generation fluorinated oxazaborolidines by the strong acid triflimide (Tf₂NH) in CH₂Cl₂ solution leads to highly active chiral Lewis acids that are very effective catalysts for (4 + 2) cycloaddition. We report herein that this catalytic activity can be further enhanced by the use of Tf₂NH in combination with the biscoordinating Lewis acid TiCl₄ or SnCl₄ as a coactivator. The effective increase in acidity of an exceedingly strong protic acid is greater for biscoordinating TiCl₄ and SnCl₄ than for monocoordinating salts, even the strong Lewis acids AlBr₃ and BBr₃ in CH₂Cl₂ or CH₂Cl₂/toluene. The increase in the effective acidity of Tf₂NH can be understood in terms of a stabilized cyclic anionic complex of Tf₂N^- and TiCl₄, which implies a broader utility than that described here. The utility of Tf₂NH-TiCl₄ activation of fluorinated oxazaborolidines is documented by examples including the first enantioselective (4 + 2) cycloaddition to α,β-unsaturated acid chlorides.

Full text of DOI:10.1021/jacs.6b08018

Subscriber access provided by Northern Illinois University
Communication
Acceleration of Enantioselective Cycloadditions Catalyzed
by Second-Generation Chiral Oxazaborolidinium
Triflimidates by Bis-Coordinating Lewis Acids.
Barla Thirupathi, Simon Breitler, Karla Mahender Reddy, and E. J. Corey
J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.6b08018 • Publication Date (Web): 16 Aug 2016
Downloaded from http://pubs.acs.org on August 17, 2016
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of the American Chemical Society is published by the American Chemical
Society. 1155 Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 5
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
Acceleration of Enantioselective Cycloadditions Catalyzed by Se-
cond-Generation Chiral Oxazaborolidinium Triflimidates by Bis-
Coordinating Lewis Acids.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Barla Thirupathi,‡ Simon Breitler,‡ Karla Mahender Reddy,‡ and E. J. Corey*
Department of Chemistry and Chemical Biology, Harvard University, Cambridge, Massachusetts 02138, United
States
Supporting Information Placeholder
the idea that the reactive species in the case of aluminum
bromide activation is a cationic complex with dienophile in
ABSTRACT: The activation of second-generation fluorinat-
ed oxazaborolidines by the strong acid triflimide in CH2Cl2
solution leads to highly active chiral Lewis acids which are
very effective catalysts for (4+2)-cycloaddition. We report
herein that this catalytic activity can be further enhanced by
the use of triflimide (Tf2NH) in combination with the bis-
coordinating Lewis acid TiCl4 or SnCl4 as co-activator. The
effective increase in acidity of an exceedingly strong protic
acid is greater for the bis-coordinating TiCl4 and SnCl4 than
for mono-coordinating salts, even the strong Lewis acids
AlBr3 or BBr3 in CH2Cl2 or CH2Cl2–toluene. The increase in
effective acidity of triflimide can be understood in terms of a
stabilized cyclic anionic complex of Tf2N– and TiCl4, which
implies a broader utility than that described here. The utility
of Tf2NH–TiCl4 activation of fluorinated oxazaborolidines is
documented by examples including the first enantioselective
(4+2)-cycloaddition to α,β-unsaturated acid chlorides.
which the coordinating group on nitrogen is actually the
equivalent of the unknown mono-coordinated cation AlBr2
for instance structure 5 (Figure 2) in the case of an acrylate
ester as dienophile.
+
–
Figure 2. Possible Reactive Complexes between an AlBr3-
activated Catalyst 1 and an α,β-Unsaturated Dienophile
Strong evidence in favor of this possibility was obtained
from experiments which demonstrated that the rates of the
AlBr3-oxazaborolidine catalyzed (4+2)-cycloaddition could be
considerably enhanced by the further addition of one equiva-
lent of AgSbF6 per AlBr3 which led to the precipitation of
The introduction of chiral oxazaborolidinium ions has
greatly broadened the scope and utility of catalytic enanti-
oselective (4+2)-cycloaddition reactions,1 especially with the
advent of second generation fluorine-enhanced oxazaboroli-
dines such as 1–4 (Figure 1).2
+
AgBr and generation of AlBr2 -coordinated oxazaborolidine,
which in turn can complex with, e.g. an acrylate ester, to
generate the super-reactive species 6.2
Although a reaction pathway via cationic species such as 6
can explain the effectiveness of AlBr3 activation, it remained
unclear why protic acid activation (which also could generate
a cation) was less effective than the AlBr3 or AlBr3/AgSbF6
process. This concern led us to the surmise that the proton-
activated oxazaborolidine catalyst might be either a hydro-
gen-bonded complex of triflimide and 1,2,3 or 4 or a contact
ion pair. That possibility seemed not unreasonable because
the reaction medium for catalytic cycloadditions is generally
non-polar – specifically CH2Cl2 or CH2Cl2–toluene mixtures.
It was thought that a bis-coordinating Lewis acid such as
TiCl4 or SnCl4 might be especially effective in separating tri-
flimidate ion from a protonated oxazaborolidine because of
coordination to form the complex 7 (Figure 3), in which the
negative charge is far more delocalized than in the triflimi-
date ion.
Figure 1. Second Generation Oxazaborolidines 1–4
Very powerful cationic catalysts are available from 1–4 by
activation using one equivalent of the strong protic acid tri-
flimide (Tf2NH) but not from weaker acids such as CH3SO3H.
Aluminum tribromide is also an effective activator of 1–4, but
other (weaker) Lewis acids are not. The fluorine substituents
in the pre-catalysts 1–4 strongly decrease the electron-
density of the donor nitrogen and enhance the electrophilic
power of the activated catalysts. Triflic acid activation is not
quite as effective as triflimide activation, whereas AlBr3-
coordination to 1–3 leads to distinctly higher catalytic power
than protonation by Tf2NH.2 These facts led us to entertain
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 2 of 5
co-activator. There is only a modest further acceleration of
1
2
these cycloadditions by increasing the TiCl4/Tf2NH ratio
from 1 to 2.
3
4
5
6
7
8
9
In this reaction, TiCl4 alone is not an effective activator of
oxazaborolidines such as 8–12. The combination of Tf2NH
and TiCl4 is especially beneficial with the F10/F0 catalyst 12.
The weaker Lewis acid SnCl4 can also accelerate the reaction,
although not as effective as TiCl4. On the other hand, it is
notworthy that the mono-coordinating Lewis acids AlBr3 and
SbCl5 are considerably less effective coactivators, although
they are very strong Lewis acids.3
Figure 3. Possible Tf2N/TiCl4-ate complex 7
We were gratified to find by experiment that triflimide ac-
tivation of various oxazaborolidines could in fact be marked-
ly enhanced simply by the use of TiCl4 as co-activator (1:1
Tf2NH/TiCl4 ratio, 0.7 equiv based on oxazaborolidine). The
experimental results for the (4+2)-cycloaddition of cyclopen-
tadiene to the relatively unreactive dienophile ethyl croto-
nate using five different oxazaborolidines are summarized in
Table 1. All experiments were conducted under exactly the
same reaction conditions and so the % conversion to product
is an unambiguous indicator of the potency of the catalytic
mixture. It is clear from Table 1 that TiCl4 is a very useful
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The interaction between oxazaborolidinium triflimidates
1
and TiCl4 could also be detected by H-NMR spectroscopy.
1
Shown in Tables 2A and 2B are H-NMR chemical shift data
for the F0/F3 (8) and the F10/F0 (12) oxazaborolidines alone
and in the presence of one equivalent of Tf2NH or one equiv-
alent each of Tf2NH and TiCl4. As expected, the protons at-
tached alpha to the pyrrolidine nitrogen, which are shifted
downfield by one equivalent of Tf2NH, are further shifted
downfield by the 1:1 Tf2NH/TiCl4 combination. As a working
hypothesis, we suggest that the oxazaborolidinium triflimi-
date can be thought of as a contact ion pair and the TiCl4-
enhanced species as a solvent-separated oxazaborolidiniuim
triflimidate-TiCl4 complex.
Table 1. Acceleration of (4+2)-Cycloaddition Using
Various Oxazaborolidines with Tf2NH and TiCl4 or
SnCl4 as Co-activators
1
Table 2A and 2B. H-NMR Data for Two Oxazaboroli-
dines without and with Tf2NH or Tf2NH/TiCl4
ACS Paragon Plus Environment

Page 3 of 5
Journal of the American Chemical Society
The augmented catalytic potency shown by 1:1 mixtures of
Scheme 3. Enantioselective (4+2) Cycloaddition of
1
2
3
4
triflimide and TiCl4 in (4+2)-cycloadditions mediated by chi-
ral oxazaborolidines could also be observed in entirely differ-
ent chemical reactions, for example the Tf2NH-catalyzed
cyclization of 13 to 14 (Scheme 1) which proceeds very rapidly
in CH2Cl2 solution (0.05 M) at –40 °C using 1 mol% of Tf2NH.
Cyclopentadiene to β,β-Dimethyl acrolein
5
6
Scheme 1. Proton-Catalyzed Cyclization of 13 to 14
7
8
9
The rapid enantioselective construction of a product with
seven contiguous stereocenters was also accomplished effi-
ciently using a polysubstituted diene, as shown in Scheme 4.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 4. Enantioselective (4+2) Cycloaddition with
a Heavily Substituted Diene
The rate of the strictly pseudo-first order reaction was fol-
lowed by gas chromatography.4 The first order rate constant
was determined to be 7.1×10–4 s–1 for Tf2NH alone versus
1.2×10–2 s–1 for 1:1 Tf2NH/TiCl4, which corresponds to a 17-fold
acceleration due to TiCl4. There is no cyclization under these
conditions with TiCl4. In a similar experiment comparing
Tf2NH and 1:1 Tf2NH/SnCl4, an acceleration of 13-fold was
measured for the latter cyclization reaction. It seems reason-
able to think of the accelerated cyclization of 13 to 14 as in-
volving: (1) H-bonded π-complexation of Tf2NH and the ole-
finic linkage of 13 and (2) reaction of this olefin–Tf2NH com-
plex with TiCl4 (or SnCl4) which leads to accelerated genera-
tion of the product 14.
Direct comparison of the F10/F0 catalyst 12 with the F2/F2
catalyst 10 showed these two oxazaborolidines to be compa-
rably effective, e.g. in the addition of the less reactive 1,3-
cyclohexadiene to trifluoroethyl fumarate, as shown in
Scheme 5. It is important to note that the F10/F0 oxazaborol-
idine is simpler to make than the F2/F2 compound.
We propose that many reactions can be effected by using 1
mol% if a 1:1 mixture of Tf2NH/TiCl4 which normally require
much larger amounts of more conventional Brønsted acids
such as H2SO4, HCl, CH3SO3H or CF3CO2H, with obvious
advantages operationally.5
Scheme 5. Comparison of F10/F0 and F2/F2 in the
(4+2)-Cycloaddition of 1,3-Cyclohexadiene to Trifluo-
roethyl fumarate
The beneficial effect of TiCl4 as co-activator together with
Tf2NH is especially pronounced with F10-oxazaborolidines
(4, Figure 1). These are easy to make,2 use and recover for
reuse. Furthermore, their use can be effective even at the 1–2
mol% level, as illustrated by the following examples (Scheme
2).
Scheme 2. Diels–Alder Cycloaddition using 1–2 mol%
of F10/F0·Tf2NH/TiCl4 Catalyst
In earlier research we found that the use of the first-
generation oxazaborolidines for (4+2)-cycloaddition of acry-
loyl chloride or other acid chlorides to dienes proceeded with
poor (or no) enantioselectivity due to two unfavorable fac-
tors: (1) the intrinsically high reactivity and non-catalyzed
cycloaddition rates and (2) the lower carbonyl electron-
density which translated into poorer binding to the catalyst.7
Recently, we have discovered a dramatic effectiveness of the
F10/F0·Tf2NH/TiCl4 catalytic system with α,β-unsaturated
acid chlorides that has allowed access to the highly reactive
and useful products shown in Figure 4 with the use of just 5
mol% of catalyst in CH2Cl2 at –78 °C.8
β,β-Disubstitution in an α,β-unsaturated carbonyl sub-
strate is well-known to prevent (4+2)-cycloaddition reactions
with 1,3-dienes. Indeed, β,β-dimethyl acrolein was found to
be unreactive even with cyclopentadiene using first-
generation oxazaborolidine catalysts,1b and even several of
the second-generation catalysts. Nonetheless, smooth (4+2)-
cycloaddition of cyclopentadiene to β,β-dimethyl acrolein
was observed using just 5 mol% of F10/F0 catalyst 12 togeth-
er with Tf2NH/TiCl4 for activation at –78 °C, allowing an effi-
cient route to the bicyclic alcohol 15 (Scheme 3).6
Figure 4. Acid chlorides available using 5 mol% of the
F10/F0·Tf2NH/TiCl4 catalyst
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 4 of 5
We also found that these same products could be prepared
Supporting Information
1
2
3
4
5
6
with approximately the same yields and enantioselectivities
using the F2/F2 catalyst 10 and the Tf2NH/TiCl4 activator
combination. This combination appeared to be superior to
the F10/F0 precatalyst 12 for the Tf2NH/TiCl4-promoted reac-
tions of acid chlorides and acyclic dienes (see Supporting
Information).
Experimental procedures and characterization data for novel
1
reactions and products including copies of H- and 13C-NMR
spectra and chiral HPLC traces. This material is available free
of charge via the In-ternet at http://pubs.acs.org.
AUTHOR INFORMATION
7
8
9
Acid fluorides are much less reactive than acid chlorides in
oxazaborolidinium-catalyzed (4+2)-cycloadditions and, in
fact, seem to be surprisingly inert. A 1:1 mixture of acryloyl
chloride and fluoride with the F10/F0 catalyst and excess
cyclopentadiene at –78 °C formed exclusively the acid chlo-
Corresponding Author
*corey@chemistry.harvard.edu
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Author Contributions
1
ride adduct as determined by H-NMR and 19F-NMR analysis
‡These authors contributed equally.
of the total reaction product. Moreover, no catalytic reaction
occurred between cyclopentadiene and acryloyl fluoride even
at 0 °C over several hours. It is possible that acryloyl fluoride
does not coordinate to the catalyst.
Notes
The authors declare no competing financial interests.
ACKNOWLEDGMENT
α,β-Unsaturated acid chlorides were found to be the most
reactive dienophiles that we have studied using the
F10/F0·Tf2NH/TiCl4 catalytic system. A comparison of the
catalytic (4+2)-cycloaddition of cyclopentadiene with several
crotonate derivatives (CH3CH=CHCOX) showed the follow-
ing times to completion at –78 °C: X = Cl: 0.7 h; X =
CF3CH2O, 2 h; X = ClCH2O, 3 h; X = NCCH2O, 4.5 h; X =
C2H5O, ca. 8 h at –20 °C.9
We are grateful to the following for research grants: Pfizer,
Gilead and Bristol-Myers Squibb. S. B. acknowledges a post-
doctoral fellowship from the Swiss National Science Founda-
tion.
REFERENCES
(1) (a) Corey, E. J. Angew. Chem. In. Ed. 2002, 41, 1650-1667. (b) Co-
rey, E. J. Angew. Chem. In. Ed. 2009, 48, 2100-2117.
(2) Reddy, K. M.; Bhimireddy, E.; Thirupathi, B.; Breitler, S.; Yu, S.;
Corey, E. J. J. Am. Chem. Soc. 2016, 138, 2443-2453.
(3) For a complete list of results in Table 1 see Supporting Infor-
mation.
(4) For full details see Supporting Information.
One advantage to the use of acid chlorides as dienophiles
is the possibility of using the reactive Diels–Alder adducts
directly for conversion into a wide range of useful com-
pounds, e.g. amides, Wolff-rearrangement products via di-
azoketone or Curtius rearrangement products via acyl azides
(Scheme 6). Such compounds cannot be obtained directly
from Diels–Alder reactions of the corresponding dienophiles.
(5) A reviewer has suggested the citation of a paper on accelera-
tion of Mukayama–aldol reactions by reagents such as
Me3SiOTf/Et2AlCl, see Oishi, M.; Aratake, S.; Yamamoto, H. J. Am.
Chem. Soc. 1998, 120, 8271-8272.
(6) The bicyclic alcohol 15 was prepared for comparison from
camphene by hydroboration-oxidation, see Biellmann, J.-F.;
d’Orchymont, H. J. Org. Chem. 1982, 47, 2882-2886.
(7) Ryu, D. H.; Zhou, G.; Corey, E. J. Org. Lett. 2005, 7, 1633-1636.
(8) For determination of yields and enantioselectivities after in
situ quenching with various reagents, see Supporting Information.
(9) The reactivity of the various esters in the series
CH3CH=CHCOX with dienes and Tf2NH/TiCl4-acitvated oxazaborol-
idines parallels the infrared stretching frequencies (cm-1) of the C=O
group (COCl = 1758, 1740; COOCH2CF3 = 1736; COOCH2Cl = 1737;
COOCH2CN = 1727; COOCH2CH3 = 1716). On the other hand, it was
observed that the hexafluoroisopropyl and 2,6-dichlorophenyl esters
were quite unreactive, possibly because of steric deceleration. For
details see Supporting Information.
Scheme 6. Derivatizations of Acid Chloride Diels–
Alder Adducts
In summary, this research has demonstrated an unprece-
dented increase in the activation of chiral oxazaborolidines
by triflimide combined with an equimolar amount of TiCl4.
The Tf2NH/TiCl4 reagent would appear to have considerable
potential for other applications in CH2Cl2 or CH2Cl2–toluene
requiring strong protic acid catalysis. We suggest that the
effectiveness of the Tf2NH/TiCl4 combination is driven by the
greater stability of complex 7 in comparison with Tf2N–.
ASSOCIATED CONTENT
ACS Paragon Plus Environment

Page 5 of 5
Journal of the American Chemical Society
1
2
For Table of Contents Only
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
5
ACS Paragon Plus Environment