Lewis acid triggered reactivity of a lewis base stabilized scandium-terminal imido complex: C-H bond activation, cycloaddition, and dehydrofluorination

Article abstract of DOI:10.1021/ja5061559

A stable scandium-terminal imido complex is activated by borane to form an unsaturated terminal imido complex by removing the coordinated Lewis base, 4-(dimethylamino)pyridine, from the metal center. The ensuing terminal imido intermediate can exist as a

Full text of DOI:10.1021/ja5061559

Communication
pubs.acs.org/JACS
Lewis Acid Triggered Reactivity of a Lewis Base Stabilized Scandium-
Terminal Imido Complex: C−H Bond Activation, Cycloaddition, and
Dehydrofluorination
†
†
‡
†
,‡
†
Jiaxiang Chu, Xianghao Han, Christos E. Kefalidis, Jiliang Zhou, Laurent Maron,* Xuebing Leng,
,
†
and Yaofeng Chen*
†State Key Laboratory of Organometallic Chemistry, Shanghai Institute of Organic Chemistry, Chinese Academy of Sciences, 345
Lingling Road, Shanghai 200032, P. R. China
‡LPCNO, CNRS & INSA, Universite
́
Paul Sabatier, 135 Avenue de Rangueil, 31077 Toulouse, France
*
S Supporting Information
Scheme 1. Generation and Reactivity of a Coordinatively
Unsaturated Scandium-Terminal Imido Intermediate (I)
ABSTRACT: A stable scandium-terminal imido complex
is activated by borane to form an unsaturated terminal
imido complex by removing the coordinated Lewis base, 4-
(
dimethylamino)pyridine, from the metal center. The
ensuing terminal imido intermediate can exist as a THF
adduct and/or undergo cycloaddition reaction with an
internal alkyne, C−H activation of a terminal alkene, and
dehydrofluorination of fluoro-substituted benzenes or
alkanes at room temperature. DFT investigations further
highlight the ease of C−H activation for terminal alkene
and fluoroarene. They also shed light on the mechanistic
aspects of these two reactions.
erminal imido complexes of early transition metals have
T
attracted intense interest and been extensively studied in
1
the past two decades. The research on such complexes has
revealed rich reactivity and applications in group-transfer and
catalytic reactions. One exception is those complexes with rare-
2
earth-metal ions. Due to the highly polarized nature of the rare-
earth-metal−nitrogen double bond, the rare-earth-metal termi-
nal imido species once formed can easily assemble into more
stable μ- or μ (n = 3, 4)-bridged bimetallic or multimetallic
n
3
4
species, or undergo C−H bond activation to give the amides.
Recently, we synthesized and structurally characterized the first
rare-earth-metal terminal imido complexes, a DMAP (4-
(
dimethylamino)pyridine)-containing scandium-terminal imido
complex supported by a tridentate ligand and a scandium5-
terminal imido complex supported by a tetradentate ligand.
After our works, several DMAP-containing scandium-terminal
imido complexes were prepared by other groups via DMAP-
promoted alkane elimination of scandium anilido methyl
to provide a coordinatively unsaturated scandium-terminal imido
intermediate (I), which shows an interesting reactivity toward
various substrates. DFT mechanistic studies on the C−H bond
activation of an α-olefin and the dehydrofluorination of
pentafluorobenzene by the imido intermediate I are also
presented to give further insight into this reactivity.
6
complexes. During the reactivity study of scandium-terminal
imido complexes, we found that coordination of Lewis base
7
The scandium-terminal imido complex [LSc
NDIPP(DMAP)] (1; L = [MeC(N(DIPP))CHC(Me)(NCH -
DMAP (or a pendant arm of the supporting ligand) to scandium
ion stabilizes the complexes but also reduces their reactivity due
to the congested environment and coordinative saturation of the
scandium center. Therefore, the complexes do not exhibit the
high reactivity one would expect from the highly polar ScN
bond. For example, the complexes are unable to activate two
2
−
i
5a
CH NMe)] , DIPP = 2,6- Pr C H ), 9-BBN (9-borabicyclo-
2
2
6
3
nonane), and 1-phenylpropyne were mixed in C D at room
6
6
1
temperature. Monitoring the reaction by H NMR spectroscopy
revealed the formation of a DMAP→BBN adduct and a new
8
important substrates, alkene and internal alkyne. Herein we
report that the coordinated DMAP can be removed by a borane,
Received: June 19, 2014
Published: July 18, 2014
©
2014 American Chemical Society
10894
dx.doi.org/10.1021/ja5061559 | J. Am. Chem. Soc. 2014, 136, 10894−10897

Journal of the American Chemical Society
Communication
9
complex (2). A subsequent scaled-up reaction in toluene
distance in the scandium anilido methyl complex [L(CH )Sc−
3
5a
provided 2 in 68% yield. Complex 2 was characterized by NMR
spectroscopy, elemental analysis, and X-ray crystallography,
confirming that 2 is a scandium azascandacyclobutene complex,
as shown in Scheme 1. To the best of our knowledge, complex 2
is the first azametallacyclobutene derivative of a rare-earth metal.
Complex 2 apparently originates from a [2+2] cycloaddition
between the ScN double bond and the CC triple bond. The
reaction is highly regioselective, and only the anti-Markovnikov
addition product was observed. The molecular structure of 2 is
shown in Figure 1. The C53−N4 bond (1.441(5) Å) is
N(H)DIPP] (2.047(3) Å). This demonstrates that the imido
ligand of 1 is transformed into an anilido in 2 and 3. The reaction
t
of complex 1 with 1.1 equiv of BBN−CH CH Bu, 20 equiv of
2
2
3,3-dimethyl-1-butene, and 20 equiv of deuterated 3,3-dimethyl-
1-butene [(CH ) CC(H)CD ] gave a substantial KIE value of
3
3
2
6.2, in line with a C−H bond activation. The kinetic study of the
formation of 3 by the reaction of complex 1 with 2 equiv of
t
BBN−CH CH Bu, and 50 equiv of 3,3-dimethyl-1-butene in the
2
2
t
presence of 10 equiv of DMAP→BBN−CH CH Bu adduct,
gave a pseudo-second order law, d[3]/dt = K [1][borane]. An
2
2
obs
⧧
Eyring analysis provided activation parameters of ΔH = 11.4(5)
⧧
⧧
kcal/mol, ΔS = −32(2) cal/mol, and ΔG = 20.9 kcal/mol for
2
98.15 K. The large negative entropy of activation observed
implies a highly ordered transition state.
Based on the above experimental observations, we proceeded
with the computational mechanistic investigation of the
formation of complex 3. DFT (B3PW91) calculations were
conducted (see the SI for computational details). Three possible
pathways for the protonation of the imido ligand using 3,3-
dimethyl-1-butene were considered, namely a direct vinyl
activation, 1,2-insertion followed by 1,3-hydride shift, and 2,1-
insertion with subsequent 1,2-hydride shift. The latter is unlikely
due to the energy-demanding 1,2-hydride shift step (Figure S34).
Starting from the intermediate I, an intermolecular C−H
Figure 1. Molecular structures of 2 and 3 with thermal ellipsoids at 30%
probability level. DIPP isopropyl groups and hydrogen atoms (except
the anilido hydrogen atom) and solvent molecules in the lattice have
been removed for clarity.
activation can take place passing through TS to afford the
II‑3
final anilido complex 3 in one step (Figure 2).
consistent with a single bond, whereas the C53−C54 bond
(
1.368(5) Å) reveals double-bond character. The C53−N4−Sc
and C54−C53−N4 angles are 87.0(2)° and 118.7(3)°,
respectively, the atoms C54, C53, N4, and Sc being coplanar.
Reaction of complex 1/9-BBN with 3,3-dimethyl-1-butene at
room temperature gave a scandium anilido alkenyl complex 3 in
8
3% yield. During the reaction, 9-BBN undergoes hydroboration
t
with 3,3-dimethyl-1-butene to give BBN−CH CH Bu first, and
2
2
t
the generated BBN−CH CH Bu reacts with complex 1 to
2
2
t
provide DMAP→BBN−CH CH Bu adduct and the scandium-
2
2
terminal imido intermediate I. Complex 3 was structurally
characterized by single-crystal X-ray diffraction (XRD). There
are two crystallographically independent molecules in the unit
cell of 3, and one of them is shown in Figure 1. Formation of
Figure 2. Plausible reaction profile for the intermolecular C−H
activation.
2
complex 3 indicates that one terminal sp C−H bond of 3,3-
This process is exoergic by 6.6 kcal/mol with respect to the
adduct II. The possibility of the C−H activation to proceed from
the third available hydrogen of the olefin was considered as well.
However, due to the steric hindrance induced by the bulky
environment of the tridentate ligand, this appears not to be
possible. Instead of direct C−H activation, a reaction sequence
consisting of a 1,2-insertion of the olefin and a 1,3-hydride shift
was considered. Even though the activation barrier of the first
dimethyl-1-butene is cleaved by the scandium-terminal imido
intermediate I during the reaction. Reaction of complex 1/9-
BBN with a less bulky terminal alkene, 1-pentene, was then
studied, which showed the same reaction pattern and gave a
scandium anilido alkenyl complex (4) in 58% yield. Complex 4
was also structurally characterized by single-crystal XRD, and its
molecular structure is given in the Supporting Information
⧧
(
Figure S1). Although complexes 3 and 4 may have cis and trans
step is very close in energy (ΔG = 26.8 kcal/mol) to the C−H
isomers, only the trans isomers were observed due to steric
congestion around the metal center (confirmed by the
calculations). In complexes 3 and 4, the C52−C53 bond lengths
activation barrier, the high activation barrier for the ensuing 1,3-
hydride shift, 50.2 kcal/mol, makes it kinetically unreachable.
The same energetic situation is revealed for the other possible
isomeric pathway leading to the same product, as depicted in
Figure S33.
(
1.317(6) (or 1.325(6)) and 1.312(6) Å, respectively) reveal
substantial double-bond character. The alkenyl ligands coor-
dinate to the scandium ion in E-configuration, with the Sc−C52
bond lengths being 2.221(4) (or 2.225(5)) and 2.223(4) Å,
respectively. The Sc−N4 bonds in 3 and 4, 2.046(3) (or
Interestingly, reaction of complex 1/9-BBN with 1,4-difluoro-
benzene at room temperature provided a scandium anilido
fluoride (5). The molecular structure of the latter is shown in
1
19
2
.035(3)) and 2.050(3) Å, are significantly longer than the Sc
Figure S1. H and F NMR monitoring showed that reactions
with fluorobenzene, 1,3,5-trifluorobenzene, and pentafluoro-
benzene at room temperature also produce complex 5. The
imido
N
double bond in the scandium-terminal imido complex 1
anilido
(
1.881(5) Å). This is, however, close to the Sc−N
bond
1
0895
dx.doi.org/10.1021/ja5061559 | J. Am. Chem. Soc. 2014, 136, 10894−10897

Journal of the American Chemical Society
Communication
reaction rate depends on the substrates as follows: fluorobenzene
A plausible reaction profile for the dehydrofluorination of
C F H was computed and is given in Figure 3. It involves first a
<
1,4-difluorobenzene ≈ 1,3,5-trifluorobenzene < pentafluoro-
6
5
10
benzene, which is inconsistent with the acidity of the substrate.
C−H activation step, which requires a relatively low energy
barrier (10.5 kcal/mol). The ensuing intermediate IV is
stabilized energetically by 9.6 kcal/mol with respect to the
reactants, in agreement with the experimental observations of the
formation of an anilido aryl complex (A). The subsequent C−F
activation, yielding complex 5 and the ethenonaphthalene
derivative, corresponds to the most energetically demanding
step in the whole reaction coordinate, being 27.7 kcal/mol. The
latter activation barrier is consistent with the experimental
conditions of a slow reaction, and also in line with previous
On the other hand, complex 1 only reacts with 1,3,5-
trifluorobenzene and pentafluorobenzene at room temperature,
with the reaction with 1,3,5-trifluorobenzene being very slow.
NMR monitoring of the reactions revealed that a scandium
anilido aryl complex A is first formed and converts eventually into
the complex 5 (Scheme 2).
Scheme 2. Pathway for Dehydrofluorination of Fluoro-
Substituted Benzenes by the Scandium-Terminal Imido
Intermediate I
12
studies. Overall, the reaction is exoergic by 40.9 kcal/mol, with
the driving force being the energy stabilization offered by the
reaction of the tetrafluorobenzene with durene. It should be
noted that, in order to pass from TS to TS , the side arm of the
V
VI
flexible amine ligand has to de-coordinate, in order to release a
site for the C−F activation to proceed. The latter costs only few
kcal/mol when the product of the TS
is stabilized by the
V−VI
formation of the adduct VI, in which the benzyne is formed a
bond with the nitrogen.
The scandium-terminal imido intermediate I can react with 9-
BBN in the absence of any substrate to provide a scandium
anilido borohydride 6. Reaction of complex 1 with 1 equiv of 9-
BBN gives a DMAP→9-BBN adduct, a scandium anilido
The reaction is thus defined as a sequence of an easy C−H
activation followed by the break of the very strong C−F bond.
The reaction with fluorobenzene is highly regioselective; only the
ortho C−H bond activation product was observed. Such regio-
selectivity was already found by Mindiola and co-workers for the
C−H bond activation of fluorobenzene with a transient titanium
1
borohydride 6, and some unreacted complex 1. H NMR
monitoring showed that complex 1 is almost completely
converted into complex 6 when 3 equiv of 9-BBN is used. A
scaled-up reaction in toluene gave complex 6 in 83% yield.
Therefore, addition of the B−H bond of 9-BBN to the ScN
double bond of I takes place, followed by the coordination of
another 9-BBN molecule to provide the scandium anilido
borohydride. The molecular structure of 6 was determined by
single-crystal XRD (Figure 4). In complex 6, the borohydrido
ligand coordinates to the scandium center through two hydrogen
atoms, with two different Sc−H bond distances (1.89 and 2.02
Å) and two different B−H bonds (1.20 and 1.15 Å). Due to steric
congestion, the amino side arm of the tridentate nitrogen ligand
(L) is no longer coordinated to the scandium center. The
scandium-terminal imido intermediate I may perform an
intramolecular C−H bond activation, as shown in Scheme 1.
11
alkylidyne species. Formation of complex 5 from the scandium
anilido aryl complexes A may imply occurrence of β-fluoride
elimination with subsequent benzyne formation, as previously
12
demonstrated by Maron et al. in cerium chemistry. The
trapping of the putative o-benzynes was achieved by performing
the reaction of complex 1 (or complex 1/9-BBN) with
pentafluorobenzene and durene. This gives rise to the formation
of the 5,6,7,8-tetrafluoro-1,4-dihydro-2,3,9,10-tetramethyl-1,4-
ethenonaphthalene, which is the product of the [2+4] cyclo-
1
1
1
addition of o-tetrafluorobenzyne with durene. Furthermore, H
19
and F NMR monitoring showed that reactions of complex 1/
t
BBN−CH CH Bu with fluoro-substituted alkanes, such as
2
2
fluorocyclohexane and 1-fluorotetradecane, at room temperature
gave complex 5 and olefins (cyclohexene and 1-tetradecene). To
avoid the hydroboration reaction of the generated olefins with 9-
t
Reaction of complex 1 with 1 equiv of BBN−CH CH Bu at
2
2
room temperature gave a scandium anilido complex (7) as the
main product. The reaction is accelerated at 50 °C, and complex
t
2
BBN, the BBN−CH CH Bu was used instead.
2
Figure 3. Plausible reaction profile for the C−H and C−F activation of pentafluorobenzene.
1
0896
dx.doi.org/10.1021/ja5061559 | J. Am. Chem. Soc. 2014, 136, 10894−10897

Journal of the American Chemical Society
Communication
Sciences. L.M. is a member of the Institut Universitaire de
France. The Humboldt foundation is also acknowledged.
REFERENCES
■
(
1) For reviews, see: (a) Wigley, D. E. Prog. Inorg. Chem. 1994, 42, 239.
b) Schrock, R. R. Acc. Chem. Res. 1990, 23, 158. (c) Duncan, A. P.;
(
Bergman, R. G. Chem. Rec. 2002, 2, 431. (d) Hazari, N.; Mountford, P.
Acc. Chem. Res. 2005, 38, 839. (e) Mindiola, D. J. Acc. Chem. Res. 2006,
3
(
9, 813.
2) (a) Giesbrecht, G. R.; Gordon, J. C. Dalton Trans. 2004, 2387.
b) Liu, Z. X.; Chen, Y. F. Sci. Sin. Chim. 2011, 41, 304.
(
(
c) Summerscales, O. T.; Gordon, J. C. RSC Adv. 2013, 3, 6682.
Figure 4. Molecular structures of 6 and 8 with thermal ellipsoids at the
(3) (a) Trifonov, A. A.; Bochkarev, M. N.; Schumann, H.; Loebel, J.
30% probability level. DIPP isopropyl groups, hydrogen atoms (except
Angew. Chem., Int. Ed. 1991, 30, 1149. (b) Xie, Z. W.; Wang, S. W.; Yang,
Q. C.; Mak, T. C. W. Organometallics 1999, 18, 1578. (c) Wang, S. W.;
Yang, Q. C.; Mak, T. C. W.; Xie, Z. W. Organometallics 1999, 18, 5511.
the hydrogen atom on the borane), and solvent molecules in the lattice
have been removed for clarity.
(d) Chan, H. S.; Li, H. W.; Xie, Z. W. Chem. Commun. 2002, 652.
(e) Gordon, J. C.; Giesbrecht, G. R.; Clark, D. L.; Hay, P. J.; Keogh, D.
1
mostly disappeared in 16 h. A scaled-up reaction at 50 °C gave
W.; Poli, R.; Scott, B. L.; Watkin, J. G. Organometallics 2002, 21, 4726.
(f) Beetstra, D. J.; Meetsma, A.; Hessen, B.; Teuben, J. H.
Organometallics 2003, 22, 4372. (g) Avent, A. G.; Hitchcock, P. B.;
Khvostov, A. V.; Lappert, M. F.; Protchenko, A. V. Dalton Trans. 2004,
complex 7 in 52% yield. The molecular structure of 7 is shown in
Figure S2. The scandium-terminal imido intermediate I can be
trapped as a THF adduct (8) (Figure 4). The reaction of complex
2272. (h) Cui, D. M.; Nishiura, M.; Hou, Z. M. Angew. Chem., Int. Ed.
2005, 44, 959. (i) Conroy, K. D.; Piers, W. E.; Parvez, M. Organometallics
2009, 28, 6228. (j) Pan, C. L.; Chen, W.; Song, S. Y.; Zhang, H. J.; Li, X.
1
with 1 equiv of 9-BBN in THF generates complex 8 almost
quantitatively. The isolated yield is 45%, due to some loss in the
washing procedure of the crude product by a toluene/hexane
mixture in order to remove the byproduct DMAP→9-BBN
adduct. XRD analysis shows that the ScN(imido) bond
W. Inorg. Chem. 2009, 48, 6344. (k) Pan, C. L.; Chen, W.; Song, J. F.
Organometallics 2011, 30, 2252. (l) Schadle, D.; Schadle, C.; Tornroos,
̈
̈
̈
K. W.; Anwander, R. Organometallics 2012, 31, 5101. (m) Hong, J. Q.;
Zhang, L. X.; Wang, K.; Weng, L. H.; Zhou, X. G. Chem.Eur. J. 2013,
19, 7865. (n) Hong, J. Q.; Zhang, L. X.; Wang, K.; Chen, Z. X.; Wu, L.
M.; Zhou, X. G. Organometallics 2013, 32, 7312.
imido
distance and Sc−N −C bond angle in 7 are 1.852(4) Å and
o
1
68.6(3) , respectively. These values are close to those found in
5a
complex 1 (1.881(5) Å and 169.6(5)°, respectively).
(4) (a) Scott, J.; Basuli, F.; Fout, A. R.; Huffman, J. C.; Mindiola, D. J.
In summary, a coordinatively unsaturated scandium-terminal
imido complex is generated by removing the coordinated Lewis
base, 4-(dimethylamino)pyridine, using a Lewis acid, such as a
borane. The reactivity of this base-free scandium-terminal imido
complex appears to be very high and especially for the activation
of C−H bond of terminal alkene. Also, it can afford a
cycloaddition reaction with an internal alkyne and trigger
dehydrofluorination of fluoro-substituted benzenes or alkanes
at room temperature. DFT mechanistic investigation of the
reactivity with the 3,3-dimethyl-1-butene indicates that the
reaction is a direct vinyl activation rather than a two-step reaction
involving first an insertion. Similarly, computational studies of
the dehydrofluorination reaction of C F H with complex 1
Angew. Chem., Int. Ed. 2008, 47, 8502. (b) Wicker, B. F.; Scott, J.; Fout,
A. R.; Pink, M.; Mindiola, D. J. Organometallics 2011, 30, 2453.
(
c) Wicker, B. F.; Fan, H. J.; Hickey, A. K.; Crestani, M. G.; Scott, J.;
Pink, M.; Mindiola, D. J. J. Am. Chem. Soc. 2012, 134, 20081.
5) (a) Lu, E. L.; Li, Y. X.; Chen, Y. F. Chem. Commun. 2010, 46, 4469.
b) Lu, E. L.; Chu, J. X.; Borzov, M. V.; Chen, Y. F.; Li, G. Y. Chem.
(
(
Commun. 2011, 47, 743.
(6) (a) Jian, Z. B.; Rong, W. F.; Mou, Z. H.; Pan, Y. P.; Xie, H. Y.; Cui,
D. M. Chem. Commun. 2012, 48, 7516. (b) Chu, T.; Piers, W. E.; Dutton,
J. L.; Parvez, M. Organometallics 2013, 32, 1159. (c) Rong, W. F.; Cheng,
J. H.; Mou, Z. H.; Xie, Y.; Cui, D. M. Organometallics 2013, 32, 5523.
(7) (a) Chu, J. X.; Lu, E. L.; Liu, Z. X.; Chen, Y. F.; Leng, X. B.; Song, H.
B. Angew. Chem., Int. Ed. 2011, 50, 7677. (b) Lu, E. L.; Zhou, Q. H.; Li, Y.
X.; Chu, J. X.; Chen, Y. F.; Leng, X. B.; Sun, J. Chem. Commun. 2012, 48,
6
5
demonstrate that it involves first an easy C−H activation step
followed by the most energy demanding ortho C−F abstraction.
3
403. (c) Chu, J. X.; Lu, E. L.; Chen, Y. F.; Leng, X. B. Organometallics
2
013, 32, 1137. (d) Chu, J. X.; Kefalidis, C. E.; Maron, L.; Leng, X. B.;
Chen, Y. F. J. Am. Chem. Soc. 2013, 135, 8165.
ASSOCIATED CONTENT
Supporting Information
Experimental and computational details and a zip file containing
CIFs for 1−8 and DMAP→BBN adduct. This material is
available free of charge via the Internet at http://pubs.acs.org.
■
(8) DMAP-containing scandium-terminal imido complex can react
with phenylacetylene, which has the relatively acidic proton, to give a
deprotonated product of scandium anilido phenylacetylido complex.
*
S
The same observation was reported by Piers and co-workers in ref 6b.
1
(
9) Some featured signals for DMAP→BBN adduct in the H NMR
3
spectrum (300 MHz, C D , 25 °C): δ 8.01 (d, J = 7.2 Hz, 2H; ortho
H of DMAP), 5.55 (d, J = 7.2 Hz, 2H; meta H of DMAP), 1.94 (s,
6
6
HH
3
HH
AUTHOR INFORMATION
■
6
H; CNMe ). DMAP→BBN adduct was also structurally characterized
2
Corresponding Authors
laurent.maron@irsamc.ups-tlse.fr
yaofchen@mail.sioc.ac.cn
by single-crystal XRD, and its molecular structure is shown in Figure S3.
10) Theoretical prediction of pK values of aromatic heterocyclic
(
a
compounds in DMSO: Shen, K.; Fu, Y.; Li, J. N.; Liu, L.; Guo, Q. X.
Tetrahedron 2007, 63, 1568.
Notes
(
11) Fout, A. R.; Scott, J.; Miller, D. L.; Bailey, B. C.; Pink, M.;
Mindiola, D. J. Organometallics 2009, 28, 331.
12) Maron, L.; Werkema, E. L.; Perrin, L.; Eisenstein, O.; Andersen, R.
A. J. Am. Chem. Soc. 2004, 127, 279.
The authors declare no competing financial interest.
(
ACKNOWLEDGMENTS
This work was supported by the National Natural Science
Foundation of China (Grant Nos. 21325210, 21132002, and
■
2
1121062), the State Key Basic Research & Development
Program (Grant No. 2012CB821600), and Chinese Academy of
1
0897
dx.doi.org/10.1021/ja5061559 | J. Am. Chem. Soc. 2014, 136, 10894−10897