Synthesis and reactivity of dioxazirconacyclohexenes: Development of a zirconium-oxo-mediated alkyne-aldehyde coupling reaction

Article abstract of DOI:10.1021/om5008727

The zirconium-oxo-mediated coupling of an alkyne and an aldehyde for the synthesis of α,β-unsaturated ketones is presented. Each intermediate along the reaction pathway has been fully characterized, and the scope of the alkynes and aldehydes has been explored.

Full text of DOI:10.1021/om5008727

Communication
pubs.acs.org/Organometallics
Synthesis and Reactivity of Dioxazirconacyclohexenes: Development
of a Zirconium−Oxo-Mediated Alkyne−Aldehyde Coupling Reaction
Gregory D. Kortman, Meghan J. Orr, and Kami L. Hull*
Department of Chemistry, University of Illinois at Urbana-Champaign, 600 South Mathews Avenue, Urbana, Illinois 61801, United
States
S
* Supporting Information
ABSTRACT: The zirconium−oxo-mediated coupling of an
alkyne and an aldehyde for the synthesis of α,β-unsaturated
ketones is presented. Each intermediate along the reaction
pathway has been fully characterized, and the scope of the
alkynes and aldehydes has been explored.
risubstituted α,β-unsaturated ketones are a prominent
functional group which act as versatile synthetic
the [4 + 2] retrocycloaddition, with the long-term goal of
identifying a catalytically active transition-metal complex.
A modified literature procedure was used to synthesize the
dioxazirconacyclohexenes in very good yields (Scheme 2).8,9
T
intermediates;1 however, their regio- and stereoselective
formation can often be challenging.2−4 Inspired by the group
4 metal−imido-mediated alkyne−imine coupling reactions
reported by Bergman5 and Mindiola,6 we were interested in
developing a MO-mediated coupling of an aldehyde and an
alkyne for the synthesis of enones.7 Both are easily accessible
functional groups and, as they are at the same oxidation state as
the α,β-unsaturated carbonyl, would allow for synthesis of the
desired functionality without the requirement of a stoichio-
metric reducing agent.
Scheme 2. Synthesis of 3a
The proposed reaction pathway involves (1) [2 + 2]
cycloaddition between [M]O and the alkyne, followed by
(2) aldehyde insertion, and finally (3) a [4 + 2]
retrocycloaddition to generate the enone and regenerate the
[M]O catalyst (Scheme 1). Although examples are limited,
Cp*2Zr(OH)Cl (1) reacts with AgOTf to generate Cp*2Zr-
(OH)OTf. CsHMDS and diphenylacetylene are then added;
Cp*2ZrO is generated in situ and is subsequently trapped by
a cycloaddition with diphenylacetylene to form oxazirconacy-
clobutene 2. It was found that using CsHMDS as the base
greatly accelerated the rate of the reaction, relative to other
alkali-metal cations.12 One possible explanation for the
significant counterion effect is that the rate of elimination
from Cp*2Zr(O−X+)OTf is dependent on the counterion, with
the weaker ionic bond between [Zr]−O− and Cs+ leading to
faster reaction rates.13 Alternatively, as the elimination reaction
is heterogeneous, the Cs cation may have increased solubility
and therefore faster reaction rates.14 The subsequent carbonyl
insertion into p-trifluoromethylbenzaldehyde occurs cleanly at
room temperature to afford the dioxazirconacyclohexene 3a
Scheme 1. Proposed [M]O-Mediated Alkyne−Aldehyde
Coupling Reaction
Bergman has demonstrated that in situ generated Cp*2ZrO
undergoes the [2 + 2]-cycloaddition with simple alkynes.8,9
Hillhouse has shown that the Zr−C bond can insert into p-
tolualdehyde, n-heptanal, and formaldehyde.10 However, the [4
+ 2]-retrocycloaddition from dioxazirconacyclohexenes has not
been reported. Promisingly, a related reaction is known for the
generation of α,β-unsaturated imines and Cp2ZrO/Cp2Zr
NR from azaoxazirconacyclohexenes/diazazirconacyclohex-
enes.5 The desired retrocycloaddition occurred upon heating
the zirconacycles to 45−135 °C. However, as a [Zr]−O bond is
significantly stronger than a [Zr]−N bond,11 our initial efforts
have focused on studying the reactivity of dioxazirconacyclo-
hexene complexes and identifying conditions which promote
1
quantitatively, as determined by H NMR, and was isolated in
99% yield as a yellow solid. The structure of 3a was confirmed
by X-ray crystallography (Figure 1).
Received: August 26, 2014
© XXXX American Chemical Society
A
DOI: 10.1021/om5008727
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Communication
myriad of unidentified side products was observed (Table 1,
entry 6). Finally, Cp2ZrMe2, which is known to trap Cp2ZrO,
afforded only a 4% yield of 4a (Table 1, entry 7).5a,16
One possible explanation for the low conversion to 4a is that
the [4 + 2] retrocycloaddition reaction is reversible, with the
equilibrium favoring dioxazirconacyclohexene 3a. Accordingly,
the product that is formed may have resulted from
decomposition of Cp*2ZrO at the high reaction temperature.
If this were the case, exchange should be observed between the
dioxazirconacyclohexene and added α,β-unsaturated ketone.
When 3a and 4b are heated to 150 °C in toluene-d8, a
1:0.7:1:1 3a:3b:4a:4b mixture is produced. Approaching the
equilibrium from the other direction affords a nearly identical
mixture of products, suggesting that the reaction has reached
equilibrium (Scheme 3).12 The equilibration results indicate
Figure 1. Crystal structure of 3a. The hydrogen atoms and a solvent
molecule were omitted for clarity, and ellipsoids are shown at the 50%
probability level.
X-ray-quality crystals of 3a were obtained from the slow
evaporation of a concentrated pentane and benzene solution.
As depicted in Figure 1, the Zr is in a pseudotetrahedral
geometry, with the metallacyclohexene in a boat conformation;
C21, C22, C23, and O1 are coplanar and Zr1 and O2 lie above
the plane. To minimize steric repulsion, the three aromatic
rings in the backbone are rotated out of conjugation. The Zr−
enolate (Zr1−O1) bond length of 1.9892(13) Å is shorter than
the Zr−alkoxide bond length (Zr1−O2) of 2.0107(13) Å, as
expected.
Scheme 3. Equilibration Experiments
The key step for the proposed alkyne−aldehyde coupling
reaction is the [4 + 2] retrocycloaddition from the
dioxazirconacyclohexene. After the temperature was raised to
150 °C for 24 h, a 4% yield of 4 was observed by 19F NMR, as
determined by comparison to an internal standard (C6F6); no
oligomerized [Zr]O was observed (Table 1, entry 1).
that the Cp*2ZrO, once formed, is preferentially trapped by
α,β-unsaturated ketones over other ligands (Table 1 and
Scheme 3). Therefore, the addition of a sacrificial α,β-
unsaturated ketone should allow for the formation of our
desired product isomers.17 Furthermore, Cp*2Zr(O)pyridine
is an olefin isomerization catalyst; when pure (E)-4a (>20:1 E/
Z) is heated to 140 °C in the presence of Cp*2Zr(
O)pyridine, a 9:1 E/Z ratio is observed after 48 h.
Table 1. Study of the [4 + 2] Retrocycloaddition
Chalcone (5) is an inexpensive natural product that is
unsubstituted at the α position. When 2.5 equiv of 5 is added to
the reaction mixture at 140 °C, 4a and the [Zr]-chalcone
adduct 6 are observed in 64% and 32% yields, respectively
(Scheme 4). The thermodynamic bias for trapping with
a
yield (%)
entry additive (L) remaining 3a (%)
4a (dr)
Cp*2Zr(O)L
1
2
3
4
5
6
7
none
95
91
64
96
96
65
65
4 (1.6:1)
7 (1.5:1)
33 (1.6:1)
4 (1.6:1)
4 (1.5:1)
35 (1.5:1)
4 (1.6:1)
b
c
Scheme 4. Trapping of Cp*2ZrO with Chalcone (5)
pyridine
DMAP
PPh3
<5
c
<5
d
c
PhCCPh
PhCCPh
<5
e
c
<5
f
Cp2ZrMe2
d
a
In situ yields determined by comparison of the product to an internal
b
standard in the 19F NMR. White solid resulting from ZrO
c
oligomerization was not observed. Not observed by 1H NMR in
d
comparison to an authentic sample of Cp2Zr(O)L. An authentic
chalcone is consistent with the reduced A1,3 strain in 6, relative
to 3a. Unfortunately, the diastereomeric ratio of trans to cis
double bonds was relatively low, though consistent with a
thermodynamic mixture of olefin isomers.12
The reaction between 3a and 5 to form 6 could occur either
through an associative or dissociative mechanism. In the first,
chalcone (5) would coordinate to the Zr and promote the loss
of 4a; free Cp*2ZrO would not be generated. This
mechanism would be first order in both 3a and 5. Alternatively,
if the reaction were to proceed through a dissociative
mechanism, a [4 + 2] retrocycloaddition would generate the
Cp*2ZrO complex, which would undergo a rapid [4 + 2]
cycloaddition with chalcone (5); a rate-determining [4 + 2]
retrocycloaddition would be zero order in 5. Kinetic studies
e
sample could not be generated. 20 equiv of PhCCPh was added.
f
Decomposition of 3a was observed.
Addition of a ligand (L), including pyridine, 4-dimethylamino-
pyridine, or PPh3, which could trap the presumed [Zr]O
intermediate as Cp*2Zr(O)L (9), did not significantly
improve the reaction, affording 4a in 4, 7, and 33% yields,
respectively (Table 1, entries 2−4).15 Likewise, trace amounts
(4%) of 4a were observed with the addition of 2.0 equiv of
diphenylacetylene; increasing to 20 equiv increased the yield of
4a to 35%. However, oxazirconacyclobutene 2 was not formed
1
in either reaction (as compared to an authentic sample by H
NMR spectroscopy); rather, significant decomposition to a
B
DOI: 10.1021/om5008727
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Communication
were undertaken to distinguish between the two mechanisms.
Using initial rate kinetics, it was determined that the reaction
was first order in 3a and zero order in 5, consistent with a
dissociative mechanism.12 Further, there is a linear relationship
between 1/kobs and [6]/[5] and a nonzero intercept. Together,
these data suggest that the dissociative mechanism is occurring.
Once the conditions were identified for the formation of the
α,β-unsaturated ketone 4a and trapping of the Cp*2ZrO
intermediate with 5, the scope of both the aldehyde and the
alkyne coupling partners was examined. A variety of aldehydes
were subjected to the insertion reaction and then heated to 140
°C with 5 in a NMR tube sealed with a Teflon lined cap. Para-
substituted benzaldehyde derivatives with both electron-with-
drawing and electron-donating groups undergo the reaction
(Table 2, entries 1−4). The conditions are tolerant of fluoro,
Finally, the scope of oxazirconacyclobutenes, and therefore
the alkynes that can undergo the coupling reaction, was
examined. As seen in Table 3, entries 1 and 2, diary-
Table 3. Scope of Oxazirconacyclobutenes
a
yield (%)
entry
R1, R2
8
9 (dr)
6
1
2
3
4
5
R1, R2 = p-MeOC6H4 (a)
R1, R2 = p-CF3 C6H4 (b)
methyl, Ph (c)
>99
>99
>99
>99
73
95 (1.7:1)
49 (1.1:1)
0
55
b
12
0
30
65
n-butyl, Ph (d)
53 (1.2:1)
95 (5.2:1)
Table 2. Scope of Aldehydes in the Coupling Reaction
isopropyl, Ph (e)
a
b
See Table 2. 48 h.
loxazirconacyclobutenes 7a,b readily undergo insertion into 4-
methoxybenzaldehyde, which was chosen due to the lack of
side products in both the insertion and retrocycloaddition steps
(Table 3, entry 5). The subsequent [4 + 2] retrocycloaddition
of the electron-rich 8a affords the corresponding α,β-
unsaturated ketone 9a in 95% yield and 1.7:1 dr (Table 3,
entry 1). Conversely, the di-p-CF3 8b adduct was significantly
slower at undergoing the [4 + 2] retrocycloaddition, as only
36% of 9b was observed at 24 h along with 49% remaining
starting material; after an additional 24 h 9b was generated in
49% yield (67% brsm) (Table 3, entry 2). Extending the
reaction time did not improve the yield.
There are two possible constitutional isomers that could be
formed when unsymmetrical alkynes are used; the selectivity of
the reaction is dependent upon the [2 + 2] cycloaddition. The
reaction between 1-phenyl-1-propyne and Cp*2ZrO is
known to incorporate the phenyl group exclusively α to the
[Zr] and the methyl α to the oxygen atom (7c);6 therefore, it
was expected that the α,β-unsaturated ketone would form as a
single regioisomer.19 Likewise, when 7c was subjected to the
carbonyl insertion conditions, 8c was observed in 99% yield as
a single regioisomer, where R1 = Me and R2 = Ph (Table 3,
entry 3). However, 9c was not observed after heating at 140 °C
for 24 h; rather, side products consistent with deprotonation of
9c with the basic [Zr]O to form the enolate were observed.6
When R1 is n-butyl (7d) or isopropyl (7e), the retrocycloaddi-
tion reaction does proceed, affording 9d,e in 53% and 95%
yields, respectively, as single constitutional isomers (Table 3,
entries 4 and 5). The increased steric hindrance of 9d,e, relative
to 9c, likely slows the deprotonation reaction relative to the [4
+ 2] cycloaddition with chalcone (5). Increasing the size of the
groups on the alkene increases the dr; when R1 = iPr, the dr is
5.2:1.
a
yield (%)
entry
R
3
4 (dr)
6
1
2
p-F-C6H4 (b)
p-MeO2C-C6H4 (c)
p-Me-C6H4 (d)
p-MeO-C6H4 (e)
o-Me-C6H4 (f)
mesityl (g)
>99
>99
>99
>99
56 (1.4:1)
34
28
57
52
40
64 (1.8:1)
b
3
82 (nd)
4
>99 (1.1:1)
91 (1.4:1)
c
5
>99
d
6
0
e
bc
,
7
n-pentyl (h)
>99
40 (1.4:1)
86 (1.3:1)
18
59
f
8
cyclohexyl (i)
tert-butyl (j)
H (k)
83
f
9
0
10
>99
0
0
a
In situ yields determined by comparison of the product to an internal
b
1
standard in the H NMR. The dr was not determined (nd) due to
c
coincidental chemical shifts in the 1H NMR. 1.6 equiv of 2-
d
tolaldehyde was required for complete conversion. No reaction was
e
observed. 7a was used, rather than 3, for ease of calculating the in situ
f
yield. Formation of Cp*2Zr[η2-OCH(tBu)OCH(tBu)O] and free
PhCCPh was observed.
ester, and ether functional groups. Electron-rich aldehydes are
especially efficient; namely, when p-methoxybenzaldehyde was
used, the α,β-unsaturated ketone 4e was formed in 99% yield
and 1.1:1 dr. Ortho substitution is tolerated (Table 2, entry 5),
as the reaction between 2 and o-tolualdehyde afforded a 91%
yield and 1.4:1 dr of 4f. Additional steric hindrance was not
tolerated, as mesitylaldehyde did not undergo the insertion
reaction (Table 1, entry 6).
Aliphatic n-hexanal and cyclohexanal undergo the reaction in
40% and 86% yields, respectively (Table 2, entries 7 and 8).
Carbonyl insertion does not occur with the more hindered
pivalaldehyde (4j; Table 2, entry 9); rather, the known
metallacycle Cp*2Zr[η2-OCH(tBu)OCH(tBu)O] was gener-
ated. This undesired side reaction presumably occurs through
[2 + 2] retrocycloaddition to generate Cp*2ZrO, which
undergoes a subsequent reaction with pivaldehyde.18 Finally,
relatively unhindered formaldehyde readily undergoes the
insertion reaction to generate 3k quantitatively. However,
neither 4k nor any decomposition of 3k was observed after 24
h at 140 °C.
The yields of the organic and organometallic products in
Tables 2 and 3 were determined by comparison to an internal
1
standard (1,3,5-trimethoxybenzene) in the H NMR spectrum.
This was necessary, as the α,β-unsaturated ketone is one of the
decomposition products upon exposure of the dioxazirconacy-
clohexene to a GC injection port, proton source, or standard
purification techniques such as silica gel chromatography, and,
therefore, leads to artificially high yields of α,β-unsaturated
ketones.12 To demonstrate the isolation of the organic
products, the [4 + 2] retrocycloaddition of 3d, which is nearly
C
DOI: 10.1021/om5008727
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Communication
1
(8) Carney, M. J.; Walsh, P. J.; Hollander, F. J.; Bergman, R. G.
Organometallics 1992, 11, 761.
quantitative by H NMR, was scaled up to 0.33 mmol and 4e
was isolated in a 75% yield of a 1.1:1 mixture of diastereomers.
In conclusion, dioxazirconacyclohexene complexes are able
to undergo a reversible [4 + 2] retrocycloaddition to afford α,β-
unsaturated ketones and Cp*2ZrO and the equilibrium
favors the dioxazirconacyclohexene. Using chalcone (5) to trap
the Cp*2ZrO generates the α,β-unsaturated carbonyls in
good to excellent yields as a thermodynamic mixture of olefin
isomers. Moreover, each step of the proposed catalytic
cycle[2 + 2] cycloaddition, carbonyl insertion, and [4 + 2]
retrocycloadditionhas been demonstrated to occur under
similar reaction conditions with a variety of alkynes and
aldehydes. Future investigations will focus on identifying a M
O complex that can catalyze the alkyne−aldehyde coupling
reaction.
(9) For alternative oxozirconacyclobutene syntheses see: (a) Vaugh-
an, G. A.; Hillhouse, G. L.; Lum, R. T.; Buchwald, S. L.; Rheingold, A.
L. J. Am. Chem. Soc. 1988, 110, 7215. (b) Vaughan, G. A.; Sofield, C.
D.; Hillhouse, G. L. J. Am. Chem. Soc. 1989, 111, 5491. (c) Carney, M.
J.; Walsh, P. J.; Hollander, F. J.; Bergman, R. G. J. Am. Chem. Soc.
1989, 111, 8751. (d) List, A. K.; Koo, K.; Rheingold, A. L.; Hillhouse,
G. L. Inorg. Chim. Acta 1998, 270, 399.
(10) Vaughan, G.; Hillhouse, G.; Rheingold, A. J. Am. Chem. Soc.
1990, 112, 7994.
(11) Schock, L. E.; Marks, T. J. J. Am. Chem. Soc. 1988, 110, 7701.
(12) See the Supporting Information.
(13) Nunes, P.; Leal, J. P.; Chachata, V.; Raminhos, H.; Minas de
Piedade, M. E. Chem. Eur. J. 2003, 9, 2095.
(14) Dijkstra, G.; Kruizinga, W. H.; Kellogg, R. M. J. Org. Chem.
1987, 52, 4230.
(15) When DMAP was used as a trap for other complexes, 4c−e
were obtained in 11, 18, and 6% yields, respectively, along with 88, 81,
and 88% yields of 3c−e.
(16) Cp*2Zr(O)pyr (9) does not react with Cp2ZrMe2 when
combined directly and Cp*2Zr(Me)−O−(Me)ZrCp2 was not
observed, suggesting that the Cp* ligands hinder its formation.
(17) The calculated relative energy is ΔG = 0.1 kcal/mol (as
determined using Gaussian ’03, B3LYP/6-31g*); the calculated
thermodynamic ratio is 1.6:1.
(18) (a) Howard, W. A.; Waters, M.; Parkin, G. J. Am. Chem. Soc.
1993, 115, 4917. (b) Howard, W. A.; Trnka, T. M.; Waters, M.;
Parkin, G. J. Organomet. Chem. 1997, 528, 95.
(19) The observed constitional isomer is the opposite of that formed
in alkyne−aldehyde coupling reactions which proceed through
oxetene/carbocation intermediates.5
ASSOCIATED CONTENT
* Supporting Information
Text, figures, and CIF files giving experimental procedures,
characterization data, and crystallographic data. This material is
available free of charge via the Internet at http://pubs.acs.org.
■
S
AUTHOR INFORMATION
Corresponding Author
*E-mail for K.L.H.: kamihull@illinois.edu.
Notes
■
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
The authors thank the University of Illinois, Urbana
Champaign, and the Petroleum Research Fund for their
generous support. Additionally, they thank Boulder Scientific
for the generous donation of Cp*−H. Finally, they thank A. K.
Gupta for assistance with the calculations.
REFERENCES
■
(1) (a) Carey, F. A.; Sundberg, R. J. Advanced Organic Chemistry Part
B: Reaction and Synthesis, 4th ed.; Kluwer Academic/Plenum
Publishers: New York, 2001. (b) Smith, M. B. March’s Advanced
Organic Chemistry: Reactions, Mechanisms, and Structure, 7th ed.; Wiley:
Hoboken, NJ, 2013.
(2) Takeda, T. Modern Carbonyl Olefination; Wiley-VCH: Weinheim,
Germany, 2004.
(3) Nielsen, A. T.; Houlihan, W. J. The Aldol Condensation; Wiley-
VCH: Weinheim, Germany, 2004.
(4) Willis, M. C. Chem. Rev. 2010, 110, 725.
(5) (a) Hanna, T. A.; Baranger, A. M.; Walsh, P. J.; Bergman, R. G. J.
Am. Chem. Soc. 1995, 117, 3292. (b) Hanna, T. A.; Baranger, A. M.;
Bergman, R. G. J. Org. Chem. 1996, 61, 4532. (c) Ruck, R. T.;
Bergman, R. G. Organometallics 2004, 23, 2231. (d) Ruck, R. T.;
Zuckerman, R. L.; Krska, S. W.; Bergman, R. G. Angew. Chem., Int. Ed.
2004, 43, 5372.
(6) (a) Basuli, F.; Aneetha, H.; Huffman, J. C.; Mindiola, D. J. J. Am.
Chem. Soc. 2005, 127, 17992. (b) Aneetha, H.; Basuli, F.; Bollinger, J.;
Huffman, J. C.; Mindiola, D. J. Organometallics 2006, 25, 2402.
(7) For related reactions catalyzed by strong acids, silver, and gold
complexes via an oxetene and/or carbocation intermediates see:
(a) Hayashi, A.; Yamaguchi, M.; Hirama, M. Synlett 1995, 195.
(b) Viswanathan, G. S.; Li, C.-J. Tetrahedron Lett. 2002, 43, 1613.
(c) Curini, M.; Epifano, F.; Maltese, F.; Rosati, O. Synlett 2003, 552.
(d) Rhee, J. U.; Krische, M. J. Org. Lett. 2005, 7, 2493. (e) Miranda, P.
́
O.; Díaz, D.; Padron, J. I.; Ramírez, M. A.; Martín, V. S. J. Org. Chem.
2005, 70, 57. (f) Tanaka, K.; Sasaki, K.; Takeishi, K.; Sugishima, K.
Chem. Commun. 2005, 4711. (g) Jin, T.; Yamamoto, Y. Org. Lett. 2007,
9, 5259.
D
DOI: 10.1021/om5008727
Organometallics XXXX, XXX, XXX−XXX