SNAr-derived decomposition by-products involving pentafluorophenyl triazolium carbenes

Article abstract of DOI:10.1055/s-0033-1338842

Pentafluorophenyl triazolium carbenes, widely used in NHC catalysis, can decompose by several mechanisms. Under high concentration conditions, the azolium may undergo a pentafluorophenyl exchange by a proposed S_NAr mechanism to give an inactive salt. In the presence of appropriate substrates, cyclization on the ortho-position of the arene can occur, also by S _NAr. These adducts provide a potential pathway for catalyst decomposition and serve as a caveat to the development of new reactions and catalysts. Georg Thieme Verlag Stuttgart. New York.

Full text of DOI:10.1055/s-0033-1338842

CLUSTER
▌1229
cluster
SNAr-Derived Decomposition By-products Involving Pentafluorophenyl
Triazolium Carbenes
Pentafluorophenyl Triazolium Carbenes
Xiaodan Zhao, Garrett S. Glover, Kevin M. Oberg, Derek M. Dalton, Tomislav Rovis*
Department of Chemistry, Colorado State University, Fort Collins, CO 80523, USA
Fax +1(970)4911801; E-mail: rovis@lamar.colostate.edu
Received: 19.04.2013; Accepted after revision: 30.04.2013
as the base in wet MeOH at room temperature for 16 hours
(Scheme 1), the hydrate 2 and anion-exchanged triazoli-
um salt 3 were observed.9 The hydrate 2 was obtained in
30% isolated yield.10 The hydrate is not particularly stable
Abstract: Pentafluorophenyl triazolium carbenes, widely used in
NHC catalysis, can decompose by several mechanisms. Under high
concentration conditions, the azolium may undergo a pentafluoro-
phenyl exchange by a proposed SNAr mechanism to give an inactive
salt. In the presence of appropriate substrates, cyclization on the and slowly decomposed to the salt 3 on standing. We test-
ortho-position of the arene can occur, also by SNAr. These adducts
provide a potential pathway for catalyst decomposition and serve as
a caveat to the development of new reactions and catalysts.
ed the catalytic activity of 3 with aldehyde 5 (Scheme 2).
Interestingly, the desired product 6 was produced in
around 20% conversion in the presence of 20 mol% acetic
acid but was not formed in its absence. This suggests that
Key words: pentafluorophenyl N-heterocyclic carbene, inner salt,
spirocyclic oxindole, decomposition
the salt 3 might convert into the corresponding carbene as
the catalytic species under acidic conditions.
N-Heterocyclic carbene (NHC) catalysis is dominated by
O
O
3
the triazolium scaffold, with the huge majority of such
catalysts bearing at least one aryl group on the azolium
ring.1 The nature of the aromatic ring has been document-
ed to have a profound impact on both reactivity and selec-
(20 mol%)
H
COOEt
MeOH
23 °C, 15 h
AcOH (20 mol%)
O
COOEt
O
5
6
tivity across
a
range of transformations.2 The
ca. 20% conversion
pentafluorophenyl substituent on the triazolium salt, first
introduced by us in 2004,3 exhibits high activity and selec-
tivity across a range of NHC-catalyzed transformations
and has become a substituent of choice across a range of
reactions, including the intermolecular Stetter,4 redox,5
cascade,6 cycloadditions7 and others.8 Catalyst loadings as
low as 0.1 mol% may be used for some reactions (intra-
molecular Stetter, e.g.4d) but this is an exception. Given
our long history with NHC catalysis and interest in ex-
tending the utility of these systems, we expended some ef-
fort at identifying decomposition pathways for these
carbenes with an eye at the development of the next gen-
eration of catalysts. Herein, we disclose our results.
Scheme 2
When the salt 1 was scaled up to 800 mg, the hydrate 2
and a significant amount of 3 were observed. Surprisingly,
the interesting inner salt 4 was generated in trace
amount.11 The structure of the inner salt 4 was confirmed
by X-ray crystallographic analysis (Figure 1). The two
fluorinated aryl groups are twisted relative to the triazoli-
um plane. It is clear that one fluoro group on the phenyl
ring has been replaced by oxygen.12
To account for the formation of 4, we propose the mecha-
nism shown below (Scheme 3). The triazolium salt 1 is
deprotonated by KOAc to give the free carbene A. Trace
water in the reaction results in the formation of triazolium
When morpholine-based scaffold pentafluorophenyl tri-
azolium salt 1 was treated with one equivalent of KOAc
O
O
O
O
F
F
BF4
N
N
N
N
F
F
F
N
N
N
N
N
N
F
N
N
KOAc (1 equiv)
C6F5
+
+
F
F
F
MeOH
23 °C, 16 h
F
F
OH
OH
F
F
F
F
F
F
F
2
1
3
F
O
4, trace
Scheme 1
SYNLETT 2013, 24, 1229–1232
Advanced online publication: 17.05.2013
0
9
3
6
-
5
2
1
4
1
4
3
7
-
2
0
9
6
DOI: 10.1055/s-0033-1338842; Art ID: ST-2013-S0364-C
© Georg Thieme Verlag Stuttgart · New York

1230
X. Zhao et al.
CLUSTER
react with a Michael acceptor to give a new side product.13
The treatment of Boc-protected oxindole 7 with achiral
pentafluorophenyl carbene 8 at room temperature for 48
hours in THF in the presence of potassium carbonate
forms the spirocyclic adduct 9 in good yield (75%,
Scheme 4).14,15 We have shown previously that carbenes
can form adducts with electrophiles and carbenes can be
released from the adducts.16 In this case, however, the pro-
cess appears irreversible and carbenes cannot be regener-
ated. When the reaction time was decreased to 16 hours,
the yield of the adduct diminished.17
An X-ray quality single crystal of adduct 9 was obtained
by slow volatilization of its solution in hexane and dichlo-
romethane. The structure was confirmed by X-ray crystal-
lography (Figure 2). Clearly, the fluoro-substituted
phenyl group connects to the spirocenter.
Figure 1 X-ray structure of the triazolium inner salt 4; thermal ellip-
soids are shown at the 50% probability level
salt 3 with hydroxide as the anion. The hydroxide group
can attack the triazolium cation to give the hydrate 2. Ad-
ditionally, the carbene A nucleophilically attacks the tri-
azolium cation B by SNAr to produce the inner salt D and
the neutral compound C. Hydroxide anion present in
solution then likely displaces the fluoride on the more
activated pentafluorophenyl group to give 4.
O
O
N
N
KOAc
O
H2O
N
N
N
N
C6F5
OH
C6F5
1
H
A
3
O
F
N
N
F
N
N
N
N
F
F
Figure 2 X-ray structure of spirocyclic oxindole 9; thermal ellip-
soids are shown at the 50% probability level
F
C
B
O
O
N
N
N
N
N
N
OH
C6F5
C6F5
The mechanism is proposed as shown below (Scheme 5).
The triazolium salt 8 first reacts with potassium carbonate
to generate the free carbene E. Then E attacks the double
bond on the substrate oxindole to give intermediate F. The
enolate anion displaces fluoro group by SNAr to form tri-
azolium salt G. Finally, base-induced deprotonation re-
sults in sequestration of HF to deliver adduct 9.
4
OH
F
F
F
F
D
2
F
Scheme 3 Proposed mechanism for the formation of 4
In conclusion, we have reported two decomposition path-
ways involving pentafluorophenyl-derived triazolium salt
organocatalysts. The structures of these by-products were
confirmed by X-ray crystallographic analysis. Taken to-
The SNAr mechanism is also at play in the decomposition
of carbenes covalently bound to some substrates. For ex-
ample, we have found that pentafluorophenyl carbene can
N
N
MeOOC
N
F
F
BF4
N
K2CO3
MeOOC
+
N
N
23 °C, 48 h
THF
F
C6F5
N
O
O
N
Boc
7
75%
F
Boc
8
9
Scheme 4
Synlett 2013, 24, 1229–1232
© Georg Thieme Verlag Stuttgart · New York

CLUSTER
Pentafluorophenyl Triazolium Carbenes
1231
Eur. J. Org. Chem. 2011, 4298. (e) Jacobsen, C. B.;
C6F5
MeOOC
Albrecht, L.; Udmark, J.; Jørgensen, K. A. Org. Lett. 2012,
14, 5526. (f) Liu, Y.; Nappi, M.; Escudero-Adán, E. C.;
Melchiorre, P. Org. Lett. 2012, 14, 1310. (g) Grossmann, A.;
Enders, D. Angew. Chem. Int. Ed. 2012, 51, 314.
N
N
K2CO3
N
7
8
N
N
N
C6F5
N
O
(7) Zhao, X.; DiRocco, D. A.; Rovis, T. J. Am. Chem. Soc. 2011,
133, 12466.
E
Boc
F
(8) For representative examples, see: (a) Enders, D.;
Grossmann, A.; Fronert, J.; Raabe, G. Chem. Commun.
2010, 46, 6282. (b) Rong, Z.-Q.; Jia, M.-Q.; You, S.-L. Org.
Lett. 2011, 13, 4080. (c) Jia, M.-Q.; You, S.-L. ACS Catal.
2013, 3, 622.
F
N
N
N
N
(9) Typical Experimental Procedure for Decomposition of 1:
In a 5-mL vial were subsequently added triazolium salt 1
(0.2 mmol), KOAc (0.2 mmol) and MeOH (3 mL). The
solution was stirred under argon at 23 °C for 16 h, and then
concentrated. The resulting residue was purified by column
chromatography on silica gel [eluent: hexanes–EtOAc (1:1)
then 100% EtOAc then 100% MeOH] to give 2 and 3. The
salt 3 can be further purified by trituration with EtOAc.
Compound 2: yellow solid. 1H NMR (300 MHz, CDCl3): δ
= 8.26 (d, J = 12.5 Hz, 1 H), 7.30 (br s, 4 H), 7.00 (br s, 0.5
H), 6.31 (br s, 0.5 H), 4.71–4.74 (m, 1 H), 4.63 (t, J = 4.7 Hz,
1 H), 4.21–4.38 (m, 2 H), 3.26 (dd, J = 16.9, 5.1 Hz, 1 H),
3.12 (d, J = 16.9 Hz, 1 H). 13C{1H} NMR (75 MHz, CDCl3):
δ = 163.8, 158.5, 139.3, 128.7, 128.4, 127.4, 125.3, 123.3,
64.3, 57.2, 37.9, 29.3. 19F NMR (282 MHz, CDCl3): δ = –
142.8 (s, 0.5 F), –143.2 (s, 0.5 F), –145.8 (d, J = 19.8 Hz, 1
F), –152.6 (t, J = 21.4 Hz, 0.5 F), –153.4 (t, J = 21.4 Hz, 0.5
F), –160.8 (t, J = 19.5 Hz, 1 F), –161.5 (s, 1 F). LRMS (ES+):
m/z [M]+ calcd for C18H12F5N3O2: 397.09; found: 398.10.
Compound 3: slightly yellow solid. 1H NMR (400 MHz,
CDCl3): δ = 13.23 (s, 1 H), 7.71 (d, J = 7.5 Hz, 1 H), 7.23–
7.33 (m, 3 H), 6.50 (br s, 1 H), 4.98–5.05 (m, 3 H), 3.32 (dd,
J = 4.2, 17.1 Hz, 1 H), 3.19 (d, J = 17.1 Hz, 1 H). 13C{1H}
NMR (101 MHz, CDCl3): δ = 150.6, 148.1 (br), 139.8,
F
F
N
N
base
MeOOC
9
MeOOC
F
F
F
O
O
N
N
F
F
F
F
Boc
Boc
G
F
Scheme 5 Proposed mechanistic pathway
gether, these results shed light on potential pathways for
catalyst decomposition and may lead to the design of new,
more efficient catalysts for NHC-mediated transforma-
tions.
Acknowledgment
We are grateful to NIGMS (GM 72586) for generous support of this
research. T.R. thanks Amgen and Roche for support. We thank
Donald Gauthier and Greg Hughes (Merck) for a generous gift of
aminoindanol.
134.7, 129.7, 128.0, 125.4, 124.2, 77.5, 62.8, 60.3, 37.5. 19
F
Supporting Information for this article is available online at
http://www.thieme-connect.com/ejournals/toc/synlett.SnoIufproig
NMR (282 MHz, CDCl3): δ = –143.7 (d, J = 18.4 Hz, 2 F),
–145.7 (t, J = 21.8 Hz, 1 F), –157.9 to –158.1 (m, 2 F).
HRMS (ESI): m/z [M – OH]+ calcd for C18H11F5N3O:
380.0817; found: 380.0816.
m
iotSrat
ungIifoop
r
t
References and Notes
(10) Salt 3 proved difficult to purify to analytically pure material,
resulting in significant loss of material. Thus, we do not
report an isolated yield here.
(1) (a) Moore, J. L.; Rovis, T. Top. Curr. Chem. 2010, 291, 77.
(b) Vora, H. U.; Rovis, T. Aldrichimica Acta 2011, 44, 3.
(c) Bugaut, X.; Glorius, F. Chem. Soc. Rev. 2012, 41, 3511.
(2) (a) Rovis, T. Chem. Lett. 2008, 37, 2. (b) Mahatthananchai,
J.; Bode, J. W. Chem. Sci. 2012, 3, 192. (c) Schedler, M.;
Fröhlich, R.; Daniliuc, C.-G.; Glorius, F. Eur. J. Org. Chem.
2012, 4164.
(3) Kerr, M. S.; Rovis, T. J. Am. Chem. Soc. 2004, 126, 8876.
(4) (a) Liu, Q.; Perreault, S.; Rovis, T. J. Am. Chem. Soc. 2008,
130, 14066. (b) Liu, Q.; Rovis, T. Org. Lett. 2009, 11, 2856.
(c) DiRocco, D. A.; Oberg, K. M.; Dalton, D. M.; Rovis, T.
J. Am. Chem. Soc. 2009, 131, 10872. (d) DiRocco, D. A.;
Rovis, T. J. Am. Chem. Soc. 2011, 133, 10402. (e) Sánchez-
Larios, E.; Thai, K.; Bilodeau, F.; Gravel, M. Org. Lett.
2011, 13, 4942.
(5) For representative examples, see: (a) Reynolds, N. T.;
Rovis, T. J. Am. Chem. Soc. 2005, 127, 16406. (b) Vora, H.
U.; Rovis, T. J. Am. Chem. Soc. 2007, 129, 13796. (c) Vora,
H. U.; Moncecchi, J. R.; Epstein, O.; Rovis, T. J. Org. Chem.
2008, 73, 9727. (d) Vora, H. U.; Rovis, T. J. Am. Chem. Soc.
2010, 132, 2860.
(6) For representative examples, see: (a) Lathrop, S. P.; Rovis,
T. J. Am. Chem. Soc. 2009, 131, 13628. (b) Filloux, C. M.;
Lathrop, S. P.; Rovis, T. Proc. Natl. Acad. Sci. U.S.A. 2010,
107, 20666. (c) Ozboya, K. E.; Rovis, T. Chem. Sci. 2011, 2,
1835. (d) Enders, D.; Grossmann, A.; Huang, H.; Raabe, G.
(11) Compound 4: orange solid. 1H NMR (300 MHz, CDCl3): δ
= 7.35 (d, J = 4.1 Hz, 2 H), 7.13–7.17 (m, 1 H), 6.71 (d, J =
7.8 Hz, 1 H), 6.01 (t, J = 3.5 Hz, 1 H), 5.05–5.17 (m, 3 H),
3.43 (dd, J = 17.0, 4.5 Hz, 1 H), 3.34 (d, J = 17.0 Hz, 1 H),
1.86–1.62 (br s, 2 H from H2O). 13C{1H} NMR (100 MHz,
CDCl3): δ = 150.7, 139.7, 134.4, 130.3, 128.6, 126.0, 123.8,
98.7, 77.8, 62.9, 62.8, 60.1, 36.9. 19F NMR (376 MHz,
CDCl3): δ = –142.3 to –142.4 (m, 1 F), –144.1 (tt, J = 21.6,
4.4 Hz, 1 F), –144.2 to –144.3 (m, 1 F), –144.6 to –144.7 (m,
1 F), –147.3 to –147.4 (m, 1 F), –156.5 (td, J = 21.8, 6.8 Hz,
1 F), –156.8 (td, J = 21.8, 6.8 Hz, 1 F), –163.2 (t, J = 23.0
Hz, 1 F), –163.5 (t, J = 23.0 Hz, 1 F). HRMS (ESI): m/z [M
+ H]+ calcd for C24H11F9N3O2: 544.0708; found: 544.0712.
(12) This structure is somewhat disordered with what appears to
be partial protonation of the phenolic oxygen. There is a
corresponding counterion, also disordered, whose identity
could not be determined (hydroxide vs. fluoride). Both
structures (4 and 9) have been submitted to the Cambridge
Crystallographic Data Centre as CCDC 935024 and CCDC
935025.
(13) (a) Enders, D.; Breuer, K.; Raabe, G.; Runsink, J.; Teles, J.
H.; Melder, J.-P.; Ebel, K.; Brode, S. Angew. Chem. Int. Ed.
1995, 34, 1021. (b) Enders, D.; Breuer, K.; Kallfass, U.;
© Georg Thieme Verlag Stuttgart · New York
Synlett 2013, 24, 1229–1232

1232
X. Zhao et al.
CLUSTER
Balensiefer, T. Synthesis 2003, 1292. (c) Fischer, C.; Smith,
S. W.; Powell, D. A.; Fu, G. C. J. Am. Chem. Soc. 2006, 128,
1472.
Hz, 2 H), 1.64 (s, 9 H). 13C{1H} NMR (101 MHz, CDCl3): δ
= 177.2, 164.6, 158.8, 158.7, 149.5, 145.5, 138.7, 134.5,
128.7, 124.7, 123.8, 114.5, 109.9, 84.0, 50.2, 49.3, 28.1,
26.2, 21.2. 19F NMR (282 MHz, CDCl3): δ = –134.9 (ddd, J
= 5.3, 8.8, 22.5 Hz, 1 F), –150.7 (dd, J = 9.0, 20.5 Hz, 1 F),
–153.87 (dt, J = 5.1, 21.0 Hz, 1 F), –160.43 (t, J = 22.0 Hz,
1 F). HRMS (ESI): m/z [M + H]+ calcd for C27H23F4N4O5:
559.1599; found: 559.1603.
(14) Yang, L.; Wang, F.; Chua, P. J.; Lv, Y.; Zhong, L.-J.; Zhong,
G. Org. Lett. 2012, 14, 2894.
(15) Experimental Procedure for the Formation of 9: In a 5-
mL vial were subsequently added triazolium salt 8 (0.05
mmol), 7 (0.05 mmol), K2CO3 (0.05 mmol) and THF (4 mL).
The vial was capped. The mixture was stirred at 23 °C for 48
h, and then concentrated. The resulting residue was purified
by preparative TLC [eluent: hexanes–EtOAc (1:1)] to give
the adduct 9 as a solid. Compound 9: slightly yellow solid.
1H NMR (400 MHz, CDCl3): δ = 7.81 (d, J = 8.2 Hz, 1 H),
7.25 (td, J = 2.4, 8.3 Hz, 1 H), 7.01–7.07 (m, 2 H), 4.22–4.33
(m, 2 H), 3.34 (s, 3 H), 2.88–2.92 (m, 2 H), 2.54 (p, J = 7.5
(16) DiRocco, D. A.; Oberg, K. M.; Rovis, T. J. Am. Chem. Soc.
2012, 134, 6143.
(17) A catalyst deactivation pathway has been observed for N-
phenyl triazolium precursors, see: Hao, L.; Du, Y.; Lv, H.;
Chen, X.; Jiang, H.; Shao, Y.; Chi, Y. R. Org. Lett. 2012, 14,
2154.
Synlett 2013, 24, 1229–1232
© Georg Thieme Verlag Stuttgart · New York