An Interrupted Schmidt Reaction: C-C Bond Formation Arising from Nitrilium Ion Capture

Article abstract of DOI:10.1021/acs.orglett.8b02531

The rerouting of the nitrilium ion formed in the Schmidt reaction of ketones and TMSN₃ to encompass C-C bond formation with an electron-rich aromatic group is reported. Thus, when the reaction is carried out in HFIP using AlCl₃ or Al

Full text of DOI:10.1021/acs.orglett.8b02531

Letter
pubs.acs.org/OrgLett
Cite This: Org. Lett. XXXX, XXX, XXX−XXX
An Interrupted Schmidt Reaction: C−C Bond Formation Arising from
Nitrilium Ion Capture
Manwika Charaschanya, Kelin Li, Hashim F. Motiwala, and Jeffrey Aube*
†
‡
́
Division of Chemical Biology and Medicinal Chemistry, UNC Eshelman School of Pharmacy, University of North Carolina at
Chapel Hill, Chapel Hill, North Carolina 27599-7363, United States
S
* Supporting Information
ABSTRACT: The rerouting of the nitrilium ion formed in the
Schmidt reaction of ketones and TMSN3 to encompass C−C
bond formation with an electron-rich aromatic group is reported.
Thus, when the reaction is carried out in HFIP using AlCl3 or
AlBr3 as the promoter, imines, iminium ions, or enamide
derivatives are obtained through one-pot procedures. The scope
and possible mechanisms of these new transformations are
considered.
a
n its original and best-known form, the Schmidt reaction is
the acid-mediated conversion of a ketone to an amide using
Table 1. Survey of Reaction Conditions
I
hydrazoic acid or its equivalent.1 The most commonly
accepted mechanism of the reaction, as proposed by P. A. S.
Smith in 1948,2 involves addition of the azide component to
the carbonyl followed by dehydration to afford a diazoiminium
ion intermediate, which in turn undergoes loss of N2 along
with migration of an alkyl group to afford a key nitrilium ion
intermediate (Scheme 1). The “normal” Schmidt product is
Scheme 1. Generation and Trapping of Nitrilium Ions
through Schmidt Chemistry
b
c
d
entry
conditions
products (% conv. ); yield (%)
1
2
3
4
5
6
7
8
9
BF3·OEt2
TiCl4
TfOH
Al(OTf)3
AlCl3
AlCl3 (2 equiv)
AlCl3 (no basic workup)
AlBr3
AlBr3 (no basic workup)
1a (47) + 3a (46) + 4a (4)
3a (68)
1a (17) + 3a (60) + 4a (7)
1a (46) + 3a (38) + 4a (10)
3a (74); 54%
3a (88); 75%
3b (83); 78%
3a (95); 83%
3c (>100); 92%
e
10 AlCl3 (no slow addition)
11 AlBr3 (1:2 TFE/HFIP as the solvent)
1a (20) + 3a (54) + 4a (23)
1a (21) + 3a (42)
12 AlCl3 (1:2 toluene/HFIP as the solvent) 1a (90) + 3a (10)
a
Results of representative experiments are presented; see Tables S1
b
and S2 for full results. See the SI for detailed procedures; 1 equiv of
acid and NaHCO3 workup unless otherwise noted. % conversion as
determined by NMR analysis. Isolated yields. A single preparation
of this example was carried out on a 1.2 mmol scale; an 80% yield was
obtained in that reaction.
c
d
e
the amide obtained from the addition of water to the nitrilium
ion. As reported by K. F. Schmidt in the very first disclosure of
this chemistry,3 these reactions may also afford a tetrazole
product under some conditions (e.g., at high concentrations of
azide). In the intervening decades, very few examples of
Received: August 7, 2018
© XXXX American Chemical Society
A
DOI: 10.1021/acs.orglett.8b02531
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
Figure 1. Studies of the reaction scopes of (a) ketones and (b) nucleophiles. The reactions were carried out as described in Table 1, entry 9; see the
SI for reaction details. Isolated yields are shown, unless otherwise noted. Notes: aNMR yield. bAccompanied by a 2−19% yield of the corresponding
tetrazole.
intercepting the nitrilium ion with other oxygen or nitrogen
but not carbonnucleophiles have been noted.4 In contrast,
similar intermediates obtained through the Beckmann
rearrangement of oximes are known to react with allylsilanes,5
organometallics,6 and other reagents.7 Such “interrupted” reac-
tions, a class that also includes Nazarov,8 Pummerer,9 Fischer,10
and Bischler−Napieralski11 variations, enhance synthetic effi-
ciency by combining multiple bond formations in a single step.
We recently examined the Schmidt reaction of ketones with
trimethylsilyl azide to investigate the effect of the highly
hydrogen-bonding solvent hexafluoroisopropanol (HFIP) on
the reaction.4c In that work, we found that the main product of
the reaction carried out in HFIP was the tetrazole rather than
the lactam, with little dependence on the stoichiometry. These
results suggest that if an azide nucleophile more efficiently
reacts with a nitrilium ion in HFIP relative to other solvents,
other nucleophiles may do so as well, thus opening the door to
an interrupted Schmidt reaction accompanied by C−C bond
formation at the position of the nitrilium ion. Here we describe
the successful realization of this concept and its application to
the synthesis of imines, iminium ions, and enamides, all of
which are staples in preparative nitrogen chemistry.
We first sought appropriate mediators for the reaction of
4-phenylcyclohexanone (1a) with trimethylsilyl azide and
1,3,5-trimethoxybenzene (2a) in HFIP. The most promis-
ing results obtained from screening of 22 acids are shown
B
DOI: 10.1021/acs.orglett.8b02531
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
Scheme 2. Direct Reduction of Iminium Species
Scheme 3. One-Pot Conversion of Ketones to Enamides and
Related Compounds
in Table 1 (the full list is provided in Tables S1 and S2 in the
Supporting Information (SI)). These attempts clearly demon-
strate that the nitrilium ion presumably formed under these
conditions can lead to C−C bond formation, with particularly
good results obtained using AlCl3 and AlBr3 (Table 1, entries 5
and 8). Optimization experiments led us to identify slow
addition and 1 equiv of AlBr3 (as opposed to 2 equiv of AlCl3)
for these experiments (see Tables S3 and S4 for details). The
reactions afforded iminium ions 3b and 3c, respectively, as the
primary products, which could be isolated or converted to free
base 3a with a basic workup (cf. entries 5 and 7 and entries
8 and 9). An X-ray structure of compound 3c was obtained
(CCDC 1832152; also see the SI). We presume that the salts
arise from the reaction of the AlX3 promoter with HFIP sol-
vent to provide a source of HX. Diluting HFIP with either trifluo-
roethanol or toluene greatly diminished the reaction conversion,
as did omitting the slow addition of azide (entries 11 and 12).
Under some conditions, tetrazole product 4a was noted as a
minor byproduct, but the typical Schmidt product, 5-phenyl-
caprolactam (5a), was rarely observed (Tables S1 and S2).
With AlBr3 as the promoter, a range of cyclic and acyclic
substrates were successfully reacted to afford iminium salts
following nonbasic workup (Figure 1a). Reactions were
performed with slow addition of TMSN3 and stirring overnight
(some reactions also benefitted from heating to 35 °C; condi-
tions for specific compounds are available in the SI). In general,
the reaction was most useful for five- to seven-membered cyclic
ketones, although highly hindered examples and smaller-ring
ketones, many of which are known to be problematic in Schmidt
chemistry,1,12 gave poorer results. Among the heterocyclic
ketones tested, tetrahydropyran-4-one gave 3k in 63% yield,
whereas basic 1-methylpiperidine-4-one did not afford the
corresponding imine. Instead, the ketone reacted with 1,3,
5-trimethoxybenzene via an addition−elimination reaction to
afford 3l. Attempted reactions on two aromatic substrates,
flavanone and α-tetralone, led to quantitative recovery of the
starting ketones. Acyclic ketones also provided good conver-
sions, affording 3r−u. Unsurprisingly, the unsymmetrical sub-
strate 2-hexanone gave a mixture of constitutional isomers (3u +
3u′), as did the steroidal substrate 5α-cholestane (3v + 3v′).
In contrast to the wide scope of ketone reactants, the reac-
tion is currently limited to highly electron-rich aromatic nucle-
ophiles (Figure 1b). In addition to 1,3,5-trimethoxybenzene,
we were able to achieve good conversions using 1,3-dimethoxy-
benzene, pyrrole, and a number of substituted indoles. Unsuc-
cessful nucleophiles attempted under the present protocol
include 1,2,3-trimethoxybenzene, anisole, benzothiophene,
furan, and allyltrimethylsilane. Here we report 25 successful
examples of this reaction.
The iminium ions formed in this process are versatile
synthetic intermediates. For example, the iminium ion
obtained from the reaction of 1a with 2a is readily and
stereoselectively reduced with NaBH4 to afford 6a without
isolation or purification of the imine intermediate (Scheme 2).
This ring-expansion/arylation sequence was readily extrapo-
lated to enable an unprecedented one-pot conversion of
C
DOI: 10.1021/acs.orglett.8b02531
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
ketones to ring-expanded enamide derivatives (Scheme 3).
Enamides are useful synthetic intermediates with application to
natural product chemistry that warrant expanded synthetic
access.13 Thus, iminium ion 3c was generated from ketone 1a
by treatment with TMSN3 under AlBr3 promotion and
subsequently reacted with a series of electrophiles to afford
enamide, sulfonamide, or urea derivatives.
has been nearly exclusively used to react with water (to make
amides or lactams) or with azide species (to form tetrazoles,
either deliberately or as a nuisance byproduct). The present
demonstration that this intermediate, either directly or through
an iminyl halide species, can be leveraged for C−C bond
formation extends the utility of Schmidt chemistry to afford
synthetically useful iminium ions in addition to amides.
Our initial hypothesis rested on the idea that the nitrilium
ion was more persistent or more reactive in HFIP compared
with other solvents. We have now reconsidered this, especially
given the narrow scope of Lewis acids that can promote this
kind of interrupted Schmidt reaction. In particular, the success
of AlCl3 and especially AlBr3 suggests that these reagents,
which are used in stoichiometric amounts, may intercept the
nitrilium ion to generate an iminyl chloride or bromide. Either
the nitrilium ion or the iminyl halide could in principle react
with the nucleophilic aromatic reaction partner (Scheme 4a).
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.orglett.8b02531.
■
S
Experimental details and copies of spectra (PDF)
Accession Codes
CCDC 1832152 contains the supplementary crystallographic
data for this paper. These data can be obtained free of charge
via www.ccdc.cam.ac.uk/data_request/cif, or by e-mailing
data_request@ccdc.cam.ac.uk, or by contacting The Cam-
bridge Crystallographic Data Centre, 12 Union Road,
Cambridge CB2 1EZ, U.K.; fax: +44 1223 336033.
Scheme 4. Mechanistic Possibilities
AUTHOR INFORMATION
■
Corresponding Author
*jaube@unc.edu
ORCID
́
Jeffrey Aube: 0000-0003-1049-5767
Present Addresses
†MC: Chevron Science Center, Department of Chemistry,
University of Pittsburgh, 219 Parkman Avenue, Pittsburgh, PA
15260.
‡H.F.M.: Department of Chemistry, University of Michigan,
930 North University Avenue, Ann Arbor, MI 48109-1055.
Author Contributions
J.A. conceived the project and advised the experimentalists,
M.C. and K.L. performed experiments, H.F.M. performed
proof-of-concept experiments, J.A. and M.C. wrote the
manuscript, and all of the authors approved the final version
of the manuscript.
The iminium ion could also react, possibly reversibly, with
HFIP solvent (where X in Scheme 4 would be O(CF3)2),
although this was not observed and may be considered unlikely
given the low nucleophilicity of this solvent.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
The allowable intermediacy of the iminyl halide was demon-
strated by preparing a solution of iminyl chloride 10 from
4-phenylcaprolactam and triphosgene.14 A 23% yield of the
arylation product was obtained when 1,3,5-trimethoxybenzene
was added along with AlCl3, showing that the iminyl chloride
can react in this Friedel−Crafts (FC) process (Scheme 4b).
Interestingly, a higher yield of coupled product was obtained
when AlCl3 was omitted from this reaction, indicating that the
FC reaction between a suitably activated aromatic moiety and
iminyl chloride is possible in HFIP even without additional
promoters. Moreover, no arylation was observed when the ana-
logous reaction of anisolefound not to work using our inter-
rupted Schmidt/arylation protocolwas attempted. It is worth
noting that this type of iminyl chloride is generally prepared
and reacted in situ, underlining the attractiveness of making
compounds such as 3a through the presently reported process.15
The nitrilium ion was proposed as an intermediate in the
Schmidt reaction by P. A. S. Smith 70 years ago. Since then, it
We acknowledge financial support from the University of
Kansas (to M.C.) and the University of North Carolina at
Chapel Hill (to M.C. and K.L.). We thank Dr. Victor Day
(University of Kansas) for X-ray crystallography, Dr. Brandie
Ehrmann (UNC Chapel Hill) for HRMS analysis, and an
anonymous review of this manuscript for several helpful
suggestions.
REFERENCES
■
́
(1) Wrobleski, A.; Coombs, T. C.; Huh, C. W.; Li, S.-W.; Aube, J.
Org. React. 2012, 78, 1−320.
(2) Smith, P. A. S. J. Am. Chem. Soc. 1948, 70, 320−323.
(3) Schmidt, K. F. Ber. Dtsch. Chem. Ges. B 1924, 57, 704−706.
(4) (a) Litkei, G.; Patonay, T. Acta Chim. Hung. 1983, 114, 47−56.
(b) Zhang, F.-L.; Wang, Y.-F.; Lonca, G. H.; Zhu, X.; Chiba, S. Angew.
Chem., Int. Ed. 2014, 53, 4390−4394. (c) Motiwala, H. F.;
́
Charaschanya, M.; Day, V. W.; Aube, J. J. Org. Chem. 2016, 81,
1593−1609.
D
DOI: 10.1021/acs.orglett.8b02531
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
(5) (a) Schinzer, D.; Bo, Y. Angew. Chem., Int. Ed. Engl. 1991, 30,
687−688. (b) Schinzer, D.; Langkopf, E. Synlett 1994, 1994, 375−
377. (c) Schinzer, D.; Abel, U.; Jones, P. G. Synlett 1997, 1997, 632−
634. (d) Kang, K.-T.; Sung, T.-M.; Jung, H.-C.; Lee, J.-G. Bull. Korean
Chem. Soc. 2008, 29, 1669−1670.
(6) (a) Maruoka, K.; Miyazaki, T.; Ando, M.; Matsumura, Y.;
Sakane, S.; Hattori, K.; Yamamoto, H. J. Am. Chem. Soc. 1983, 105,
2831−2843. (b) Sakane, S.; Maruoka, K.; Yamamoto, H. Tetrahedron
Lett. 1983, 24, 943−946. (c) Sakane, S.; Matsumura, Y.; Yamamura,
Y.; Ishida, Y.; Maruoka, K.; Yamamoto, H. J. Am. Chem. Soc. 1983,
105, 672−674.
(7) Takuwa, T.; Minowa, T.; Onishi, J. Y.; Mukaiyama, T. Bull.
Chem. Soc. Jpn. 2004, 77, 1717−1725.
(8) (a) Grant, T. N.; Rieder, C. J.; West, F. G. Chem. Commun.
2009, 5676−5688. (b) Nakanishi, W.; West, F. G. Curr. Opin. Drug
Discovery Dev. 2009, 12, 732−751.
(9) (a) Akai, S.; Kita, Y. Top. Curr. Chem. 2007, 274, 35−76.
(b) Yorimitsu, H. Chem. Rec. 2017, 17, 1156−1167. (c) Meng, L.;
Zeng, J.; Wan, Q. Synlett 2018, 29, 148−156.
(10) (a) Mo, Y.; Zhao, J.; Chen, W.; Wang, Q. Res. Chem. Intermed.
2015, 41, 5869−5877. (b) James, M. J.; O’Brien, P.; Taylor, R. J. K.;
Unsworth, W. P. Chem. - Eur. J. 2016, 22, 2856−2881. (c) Susick, R.
B.; Morrill, L. A.; Picazo, E.; Garg, N. K. Synlett 2017, 28, 1−11.
(11) (a) Medley, J. W.; Movassaghi, M. Org. Lett. 2013, 15, 3614−
3617. (b) White, K. L.; Mewald, M.; Movassaghi, M. J. Org. Chem.
2015, 80, 7403−7411.
(12) Desai, P.; Schildknegt, K.; Agrios, K. A.; Mossman, C.; Milligan,
́
G. L.; Aube, J. J. Am. Chem. Soc. 2000, 122, 7226−7232.
(13) (a) Courant, T.; Dagousset, G.; Masson, G. Synthesis 2015, 47,
1799−1856. (b) Gopalaiah, K.; Kagan, H. B. Chem. Rev. 2011, 111,
4599−4657. (c) Carbery, D. R. Org. Biomol. Chem. 2008, 6, 3455−
3460.
(14) Jurczak, J.; Kozluk, T.; Kulicki, W.; Pietraszkiewicz, M.;
Szymanski, J. Synthesis 1983, 1983, 382.
(15) For recent alternative approaches to similarly reactive nitrilium
ions, see: (a) Ramanathan, M.; Wang, Y.-H.; Liu, Y.-H.; Peng, S.-M.;
Cheng, Y.-C.; Liu, S.-T. J. Org. Chem. 2018, 83, 6133−6141.
(b) Huang, P.-Q.; Huang, Y.-H.; Xiao, K.-J. J. Org. Chem. 2016, 81,
9020−9027.
E
DOI: 10.1021/acs.orglett.8b02531
Org. Lett. XXXX, XXX, XXX−XXX