Aluminum salen complexes and tetrabutylammonium salts: A binary catalytic system for production of polycarbonates from CO2 and cyclohexene oxide

Article abstract of DOI:10.1021/ic048508g

A series of complexes of the form (salen)AIZ, where H₂salen = N,N′-bis(salicylidene)-1,2-phenylenediimine and various other salen derivatives and Z = Et or Cl, have been synthesized. Several of these complexes have been characterized by X-ray crystallography. An investigation of the utilization of these aluminum derivatives along with both ionic and neutral bases as cocatalysts for the copolymerization of carbon dioxide and cyclohexene oxide has been conducted. By studying the reactivity of these complexes for this process as substituents on the diimine backbone and phenolate rings are altered, we have observed that aluminum prefers electron-withdrawing groups on the salen ligands, thereby producing an electrophilic metal center to be most active toward production of polycarbonates from CO₂ and cyclohexene oxide. For example, the complex derived from H₂salen = N,N′-bis-(3,5-di-tert-butylsalicylidene)-1,2-ethylenediimine is essentially inactive when compared to the analogous derivative containing nitro substituents in the 3-positions of the phenolate groups. This is to be contrasted with the catalytic activity observed for the (salen)CrX systems, where electron-donating salen ligands greatly enhanced the reactivity of these complexes for the coupling of CO₂ and epoxides. While (salen)AIZ complexes are capable of producing poly(cyclohexene oxide) carbonate with low amounts of polyether linkage along with small quantities of cyclic carbonate byproducts, their reactivities, covering a turnover frequency range of 5.2-35.4 mol of epoxide consumed/ (mol of Al·h), are greatly reduced when compared to their (salen)CrX analogues under identical reaction conditions.

Full text of DOI:10.1021/ic048508g

Inorg. Chem. 2005, 44, 1433−1442
Aluminum Salen Complexes and Tetrabutylammonium Salts: A Binary
Catalytic System for Production of Polycarbonates from CO2 and
Cyclohexene Oxide
Donald J. Darensbourg* and Damon R. Billodeaux
Department of Chemistry, Texas A&M UniVersity, College Station, Texas 77843
Received October 25, 2004
A series of complexes of the form (salen)AlZ, where H2salen
) N,N′-bis(salicylidene)-1,2-phenylenediimine and
various other salen derivatives and Z Et or Cl, have been synthesized. Several of these complexes have been
)
characterized by X-ray crystallography. An investigation of the utilization of these aluminum derivatives along with
both ionic and neutral bases as cocatalysts for the copolymerization of carbon dioxide and cyclohexene oxide has
been conducted. By studying the reactivity of these complexes for this process as substituents on the diimine
backbone and phenolate rings are altered, we have observed that aluminum prefers electron-withdrawing groups
on the salen ligands, thereby producing an electrophilic metal center to be most active toward production of
polycarbonates from CO2 and cyclohexene oxide. For example, the complex derived from H2salen
) N,N′-bis-
(3,5-di-tert-butylsalicylidene)-1,2-ethylenediimine is essentially inactive when compared to the analogous derivative
containing nitro substituents in the 3-positions of the phenolate groups. This is to be contrasted with the catalytic
activity observed for the (salen)CrX systems, where electron-donating salen ligands greatly enhanced the reactivity
of these complexes for the coupling of CO2 and epoxides. While (salen)AlZ complexes are capable of producing
poly(cyclohexene oxide) carbonate with low amounts of polyether linkage along with small quantities of cyclic
carbonate byproducts, their reactivities, covering a turnover frequency range of 5.2
−35.4 mol of epoxide consumed/
(mol of Al h), are greatly reduced when compared to their (salen)CrX analogues under identical reaction conditions.
‚
Introduction
which afford similar transition metal complexes which serve
as effective catalysts for this process.5 These latter ligands
represent less expensive, more easily derivatized alternatives
to porphyrin ligands.
Both transition and main group metal porphyrin derivatives
have been employed as catalysts for the coupling of carbon
dioxide and epoxides to provide polycarbonates and/or cyclic
carbonates (eq 1). Notably, these complexes are structurally
similar to the magnesium-containing porphyrin species used
in photosynthesis, chlorophyll. Prominent among these
reports are the pioneering investigations of Inoue and co-
workers1 and the more recent efforts of Kruper,2 Holmes,3
and Chisholm.4 During this same time period we and others
have utilized various salen (bis(salicylaldiimine)) ligands
We have recently been interested in examining the uti-
lization of salen complexes of main group metals as catalysts
for reaction 1. Our initial attempts at the use of gallium salen
complexes were unsuccessful in producing polycarbonate.6
Given the success demonstrated by aluminum porphryinates
at production of polycarbonates,1,4 we have undertaken in-
vestigations into the use of salen aluminum derivatives as
catalysts for this process. These efforts are designed to opti-
mize the effectiveness of these aluminum reagents by sys-
tematic variation of the electronics of the salen ligand and
* Author to whom correspondence should be addressed. E-mail:
djdarens@mail.chem.tamu.edu.
(1) (a) Aida, T.; Inoue, S. J. Am. Chem. Soc. 1983, 105, 1304-1309. (b)
Aida, T.; Inoue, S. J. Am. Chem. Soc. 1985, 107, 1358-1364. (c)
Aida, T.; Ishikawa, M.; Inoue, S. Macromolecules 1986, 19, 8-13.
(2) Kruper, W. J.; Dellar, D. V. J. Org. Chem. 1995, 60, 725-727.
(3) Mang, S.; Cooper, A. I.; Colclough, M. E.; Chauhan, N.; Holmes, A.
B. Macromolecules 2000, 33, 303-308.
(4) Chisholm, M. H.; Zhou, Z. J. Am. Chem. Soc. 2004, 126, 11030-
11039.
10.1021/ic048508g CCC: $30.25
Published on Web 02/03/2005
© 2005 American Chemical Society
Inorganic Chemistry, Vol. 44, No. 5, 2005 1433

Darensbourg and Billodeaux
N,N′-bis(3,5-dichlorosalicylidene)-1,2-phenylenediimine [(phen)-
(Cl2)salenH2],10a and N,N′-bis(3,5-di-tert-butylsalicylidene)-1,2-
phenylenediimine [(phen)(tBu)2salenH2]10f were prepared according
the nature of the cocatalyst, much as we have accomplished
with the chromium derivatives.7,8 Indeed, the latest elegant
mechanistic studies reported by Chisholm and Zhou attest
to the many similarities in the details of the reaction path-
ways for the CO2/epoxide coupling reaction catalyzed by
aluminum porphyrin and chromium salen derivatives.4 Of
some consideration is the fact that these aluminum complexes
are expected to be much less toxic than their chromium
analogues.
1
to literature procedures. H and 13C NMR spectra were recorded
on Unity+ 300 MHz and VXR 300 MHz superconducting NMR
spectrometers. The operating frequency for 13C NMR experiments
was 75.41 MHz. Infrared spectra were recorded on a Mattson 6021
FT-IR spectrometer with DTGS and MCT detectors. Analytical
elemental analysis was provided by Canadian Microanalytical
Services Ltd.
Synthesis of N,N′-Bis(3-tert-butyl-5-nitrosalicylidene)-1,2-eth-
ylenediamine [(tBu)(NO2)salenH2] (1). 3-tert-Butyl-5-nitrosalicyl-
aldehyde (2.0 g, 8.96 mmol) was dissolved in 40 mL of ethanol in
a 250 mL round-bottom flask fitted with a reflux condenser.
Ethylenediamine (0.360 g, 4.48 mmol) in 10 mL of ethanol was
added along with 1 drop of formic acid and the condenser washed
with 10 mL of ethanol. Immediate formation of a dark yellow
precipitate was observed. The mixture was allowed to reflux for 4
h, cooled, and placed in a freezer at -30 °C for 12 h. The resultant
yellow solid was collected by filtration, washed with cold ethanol
(2 × 20 mL), and dried under vacuum to yield 1.5 g (71%) of dark
yellow powder. X-ray-quality crystals were grown by slow evapo-
Experimental Section
Methods and Materials. All manipulations were performed
using a combination of standard Schlenk and drybox techniques.
Toluene and tetrahydrofuran (THF) were freshly distilled from
sodium benzophenone prior to use. Acetonitrile was first dried by
distillation from CaH2 onto P2O5 followed by distillation onto CaH2
and then freshly distilled from CaH2 prior to use. Cyclohexene oxide
(CHO) was distilled from CaH2 prior to use. CH2Cl2 was freshly
distilled from P2O5 prior to use. Methanol and ethanol were distilled
from the corresponding magnesium alkoxide prior to use. Bone
dry CO2 was purchased from Scott Specialty Gases. AlEt3 and
Et2AlCl were purchased from Aldrich as 1.9 M toluene solutions
and used as received. Salicylaldehyde, ethylenediamine, and 1,2-
phenylenediamine were purchased from Aldrich and used as re-
ceived. Cyclohexyldiamine was purchased as a mixture of trans
and cis isomers from Aldrich and used as received. 2,4,6-Trimeth-
ylphenol and phenol were purchased from Aldrich and sublimed
prior to use. 3,5-Di-tert-butylsalicylaldehyde and 3-tert-butyl-5-
nitrosalicylaldehyde were prepared according to literature proce-
dures.9 Synthesis of the ligands N,N′-bis(3,5-di-tert-butyl-salicylidene)-
1,2-ethylenediimine [(tBu)2salenH2],10d N,N′-bis(3,5-di-tert-butyl-
salicylidene)-R,R-cyclohexyldiimine [(cyc)(tBu)2salenH2],10c N,N′-
bis(salicylaldehyde)-1,2-phenylenediimine [(phen)salenH2],10b N,N′-
bis(3,5-dichlorosalicylidene)-1,2-ethylenediimine [(Cl2)salenH2],10e
1
ration of a concentrated CH2Cl2 solution. H NMR (C6D6): δ )
1.07 [s, 18H, C(CH3)3], 2.62 [s, 4H, NCH2CH2N], 6.89 [s, 2H,
phenyl-H], 7.55 [s, 2H, phenyl-H], 7.99 [s, 2H, phenyl-CHdN],
14.78 [s, 2H, -OH]. 13C NMR (C6D6): δ ) 28.83, 35.22, 57.68,
125.34, 126.30, 166.35. Anal. Calcd for C24H30N4O6: C, 60.26;
H, 6.43; N, 11.9. Found: C, 59.77; H, 6.41; N, 11.14.
Synthesis of N,N′-Bis(3-tert-butyl-5-nitrosalicylidene)-1,2-phen-
ylenediamine [(phen)(tBu)(NO2)salenH2] (2). A 250 mL round-
bottom flask was charged with 3-tert-butyl-5-nitrosalicylaldehyde
(3.2 g, 14.3 mmol) and 50 mL of dry ethanol. A 25 mL ethanol
slurry of 1,2-phenylenediamine (0.775 g, 7.15 mmol) and 2 drops
of formic acid were added. The immediate formation of an orange
precipitate is observed. The flask was fitted with a reflux condenser
and the mixture refluxed for 2 h. The mixture was cooled to -30
°C overnight. The orange precipitate was collected by filtration and
washed with cold ethanol (2 × 20 mL) to yield 4.78 g of orange
(5) (a) Paddock, R. L.; Nguyen, S. T. J. Am. Chem. Soc. 2001, 123,
11498-11499. (b) Darensbourg, D. J.; Yarbrough, J. C. J. Am. Chem.
Soc. 2002, 124, 6335-6342. (c) Eberhardt, R.; Allendinger, M.; Rieger,
B. Macromol. Rapid Commun. 2003, 24, 194-196. (d) Darensbourg,
D. J.; Yarbrough, J. C.; Ortiz, C.; Fang, C. C. J. Am. Chem. Soc.
2003, 125, 7586-7591. (e) Darensbourg, D. J.; Rodgers, J. L.; Fang,
C. C. Inorg. Chem. 2003, 42, 4498-4500. (f) Qin, Z.; Thomas, C.
M.; Lee, S.; Coates, G. W. Angew. Chem., Int. Ed. 2003, 42, 5484-
5487. (g) Shen, Y.-M.; Duan, W.-L.; Shi, M. J. Org. Chem. 2003, 68,
1559-1562. (h) Darensbourg, D. J.; Fang, C. C.; Rodgers, J. L.
Organometallics 2004, 23, 924-927. (i) Darensbourg, D. J.; Mack-
iewicz, R. M.; Rodgers, J. L.; Phelps, A. L. Inorg. Chem. 2004, 43,
1831-1833. (j) Lu, X.-B.; Wang, Y. Angew. Chem., Int. Ed. 2004,
43, 3574-3577. (k) Darensbourg, D. J.; Rodgers, J. L.; Mackiewicz,
R. M.; Phelps, A. L. Catal. Today 2004, 98, 485-492.
(6) Darensbourg, D. J.; Billodeaux, D. B. C. R. Chim. 2004, 7, 755-761.
(7) Darensbourg, D. J.; Mackiewicz, R. M.; Rodgers, J. L.; Fang, C. C.;
Billodeaux, D. B.; Reibenspies, J. H. Inorg. Chem. 2004, 43, 6024-
6034.
(8) Darensbourg, D. J.; Mackiewicz, R. M.; Phelps, A. L.; Billodeaux,
D. R. Acc. Chem. Res. 2004, 37, 836-844.
(9) (a) Casiraghi, G.; Casnati, G.; Puglia, G.; Sartori, G.; Terenghi, G. J.
Chem. Soc., Perkin Trans. 1 1980, 1862. (b) Cogan, D. A.; Liu, G.;
Kim, K.; Backes, B. J.; Ellman, J. A. J. Am. Chem. Soc. 1998, 120,
8011.
(10) (a) Chen, D.; Martell, A. E. Inorg. Chem. 1987, 26, 1026. (b)
Nishinaga, A.; Tsutsui, T.; Moriyama, H.; Wazaki, T.; Mashino, T.;
Fujii, Y. J. Mol. Catal. 1993, 83, 117. (c) Larrow, J. F.; Jacobsen, E.
N.; Gao, Y.; Hong, Y.; Nie, X.; Zepp, C. M. J. Org. Chem. 1994, 59,
1939. (d) Skarzewski, J.; Gupta, A.; Vogt, A. J. Mol. Catal. A 1995,
103, L63. (e) Bermejo, M. R.; Castineiras, A.; Garcia-Monteagudo,
J. C.; Rey, M.; Sousa, A.; Watkinson, M.; McAuliffe, C. A.; Pritchard,
R. G.; Beddoes, R. L. J. Chem. Soc., Dalton Trans. 1996, 14, 2935.
(f) Woltinger, J.; Backvall, J.-E.; Zsigmond, A. Chem.sEur. J. 1999,
5, 1460.
1
powder (60%). H NMR (C6D6): δ ) 1.40 [s, 18H, C(CH3)3],
6.62-6.65 [m, 2H, N-phenyl-H], 6.99-7.00 [m, 2H, N-phenyl-
H], 7.61 [s, 2H, phenyl-H], 7.78 [d (JH-JH, 2.4 Hz), 2H, phenyl-
H], 8.32 [d (JH-JH, 0.3 Hz), 2H, phenyl-CHdN], 14.80 [s, 2H,
-OH]. 13C NMR (C6D6): δ ) 1.26, 28.88, 119.66, 125.79, 127.08,
162.93. Anal. Calcd for C28H30N4O6: C, 64.85; H, 5.83; N, 10.8.
Found: C, 62.92; H, 4.32; N, 9.66.
Synthesis of N,N′-Bis(3-tert-butyl-5-nitrosalicylidene)-1,2-eth-
ylenediamine [(tBu)(NO2)salenH2] (1). 3-tert-Butyl-5-nitrosalicy-
laldehyde (2.0 g, 8.96 mmol) was dissolved in 40 mL of ethanol
in a 250 mL round-bottom flask fitted with a reflux condenser.
Ethylenediamine (0.360 g, 4.48 mmol) in 10 mL of ethanol was
added along with 1 drop of formic acid and the condenser washed
with 10 mL of ethanol. Immediate formation of a dark yellow
precipitate was observed. The mixture was allowed to reflux for 4
h, cooled, and placed in a freezer at -30 °C for 12 h. The resultant
yellow solid was collected by filtration, washed with cold ethanol
(2 × 20 mL), and dried under vacuum to yield 1.5 g (71%) of dark
yellow powder. X-ray-quality crystals were grown by slow evapo-
1
ration of a concentrated CH2Cl2 solution. H NMR (C6D6): δ )
1.07 [s, 18H, C(CH3)3], 2.62 [s, 4H, NCH2CH2N], 6.89 [s, 2H,
phenyl-H], 7.55 [s, 2H, phenyl-H], 7.99 [s, 2H, phenyl-CHdN],
14.78[s, 2H, -OH]. 13C NMR (C6D6): δ ) 28.83, 35.22, 57.68,
125.34, 126.30, 166.35. Anal. Calcd for C24H30N4O6: C, 60.26;
H, 6.43; N, 11.9. Found: C, 59.77; H, 6.41; N, 11.14.
1434 Inorganic Chemistry, Vol. 44, No. 5, 2005

Catalytic System for Production of Polycarbonates
Synthesis of N,N′-Bis(3-tert-butyl-5-nitrosalicylidene)-1,2-
phenylenediamine [(phen)(tBu)(NO2)salenH2] (2). A 250 mL
round-bottom flask was charged with 3-tert-butyl-5-nitrosalicyla-
ldehyde (3.2 g, 14.3 mmol) and 50 mL of dry ethanol. A 25 mL
ethanol slurry of 1,2-phenylenediamine (0.775 g, 7.15 mmol) and
2 drops of formic acid were added. The immediate formation of
an orange precipitate was observed. The flask was fitted with a
reflux condenser and the mixture refluxed for 2 h. The mixture
was cooled to -30 °C overnight. The orange precipitate was col-
lected by filtration and washed with cold ethanol (2 × 20 mL) to
yield 4.78 g of orange powder (60%). 1H NMR (C6D6): δ ) 1.40
[s, 18H, C(CH3)3], 6.62-6.65 [m, 2H, N-phenyl-H], 6.99-7.00 [m,
2H, N-phenyl-H], 7.61 [s, 2H, phenyl-H], 7.78 [d (JH-JH, 2.4 Hz),
2H, phenyl-H], 8.32 [d (JH-JH, 0.3 Hz), 2H, phenyl-CHdN], 14.80
[s, 2H, -OH]. 13C NMR (C6D6): δ ) 1.26, 28.88, 119.66, 125.79,
127.08, 162.93. Anal. Calcd for C28H30N4O6: C, 64.85; H, 5.83;
N, 10.8. Found: C, 62.92; H, 4.32; N, 9.66.
2H, phenyl-H], 9.35 [s, 2H, phenyl-H], 9.62 [s, 2H, phenyl-CHd
N]. Anal. Calcd for C30H33N4O6Al: C, 61.93; H, 5.81; N, 9.78.
Found: C, 60.76; H, 6.40; N, 9.09.
Synthesis of {N,N′-Bis(3,5-dichlorosalicylidene)-1,2-eth-
ylenediamine}aluminum(III) Ethyl [{(Cl2)salen}AlEt] (3). Method
A was followed with (Cl2)salenH2 (0.585 g, 1.4 mmol) and AlEt3
(0.737 mL of a 1.5 M toluene solution, 1.4 mmol). This yielded
1
0.588 g (90%) of a bright yellow powder. H NMR (DMSO-d6):
δ ) -0.58 [q (JH-JH, 7.8 Hz), 2H, AlCH2CH3], 0.63 [t (JH-JH,
8.1 Hz), 3H, AlCH2CH3], 3.91 [s, 2H, NCH2CH2N], 7.20-7.71
[m, 4H, phenyl-CH], 8.58 [s, 2H, phenyl-CHdN]. Anal. Calcd for
C18H15N2O2AlCl4: C, 46.99; H, 3.29; N,6.09. Found: C, 45.61;
H, 3.88; N, 5.33.
Synthesis of {N,N′-Bis(salicylidene)-1,2-phenylenediamine}-
aluminum(III) Ethyl [(phen)salenAlEt] (4). The procedure de-
scribed by Dzugan and Goedken was followed. Spectroscopic
measurements were in agreement with the literature.12
Synthesis of {N,N′-Bis(3,5-dichlorosalicylidene)-1,2-phen-
ylenediamine}aluminum(III) Ethyl [{(phen)Cl2salen}AlEt] (5).
Method B was followed with (phen)Cl2salenH2 (0.454 g, 1.0 mmol)
and 0.53 mL (1.0 mmol) of a 1.9 M toluene solution of AlEt3. The
immediate formation of a dark orange-red solution was observed.
The mixture was stirred at room temperature 2 h, and an orange
precipitate began to form. The solution was concentrated under
reduced pressure to ∼10 mL, and 30 mL of hexanes was added to
completely precipitate the product. The precipitate was collected
by filtration and dried under vacuum to yield 0.420 g (87%) of
orange powder. 1H NMR (DMSO-d6): δ ) -0.71 [q (JH-JH, 8.1
Hz), 2H, AlCH2CH3], 0.47 [t (JH-JH, 8.1 Hz), 3H, AlCH2CH3],
7.55-7.59 [m, 2H, N-phenyl-CH], 7.65 [d, 2H, phenyl-CH], 7.70
[s, 2H, phenyl-CH], 8.00-8.03 [m, 2H, N-phenyl-CH], 9.16 [s,
2H, phenyl-CHdN]. Anal. Calcd for C22H15N2O2AlCl4: C, 52.00;
H, 2.98; N, 5.51. Found: C, 51.92; H, 3.05; N, 6.12.
Synthesis of {salen}AlIIIEt Complexes. The synthesis of
{salen}AlEt complexes was performed via one of two methods:
A or B.
Method A. This procedure is an adaptation of procedures
described in the literature by Atwood et al.11 A 100 mL Schlenk
flask fitted with a reflux condenser was charged with salenH2 and
30 mL of toluene. A 50 mL Schlenk flask was charged with 1
equiv of a 1.9 M toluene solution of AlEt3. An additional 10 mL
of toluene was added, and the solution was cannulated onto the
ligand mixture. The reflux condenser was washed with 5 mL of
toluene. The mixture was stirred under reflux for 1 h, cooled, and
stirred at room temperature for 1 h. The solvent was removed under
reduced pressure.
Method B. This procedure is an adaptation of a procedure
described in the literature by Dzugan and Goedken.12 A 50 mL
Schlenk flask was charged with salenH2 in 30 mL of CH3CN. A
second 50 mL Schlenk flask was charged with 1 equiv of a 1.9 M
toluene solution of AlEt3. A 10 mL volume of CH3CN was added,
and the solution was cannulated onto the ligand mixture. The
mixture was stirred at room temperature 2 h and the solvent
removed under reduced pressure.
Synthesis of {N,N′-Bis(5-nitrosalicylidene)-1,2-ethylenedi-
amine}aluminum(III) Ethyl (6). Method A was followed with
(NO2)salenH2 (0.359 g, 1.0 mmol) and AlEt3 (0.55 mL of a 1.5 M
toluene solution, 1.0 mmol). This yielded 0.303 g (73%) of pale
1
yellow powder. H NMR (DMSO-d6): δ ) -0.599 [q (JH-JH,
Synthesis of {N,N′-Bis(3-tert-butyl-5-nitrosalicylidene)-1,2-
ethylenediamine}aluminum(III) Ethyl [{(tBu)(NO2)salen}AlEt]
(1b). Method A was followed with (tBu)(NO2)salenH2 (0.471 g,
1.0 mmol) and 0.53 mL (1.0 mmol) of a 1.9 M toluene solution of
AlEt3. The immediate formation of a dark orange-red solution was
observed. The solvent was removed under reduced pressure to yield
8.1 Hz), 2H, Al-CH2CH3], 0.614 [t (JH-JH, 8.1 Hz), 3H, Al-
CH2CH3], 3.81-4.02 [m, 4H, NCH2CH2N], 6.87-6.90 [d, 2H,
phenyl-H], 8.13-8.17 [m, 2H, phenyl-H], 8.42-8.43 [d, 2H,
phenyl-H], 8.70 [s, 2H, phenyl-CHdN].
Synthesis of {N,N′-Bis(3,5-di-tert-butylsalicylidene)-R,R-
cyclohexyldiimine}aluminum(III) Ethyl [{(cyc)(tBu)2salen}AlEt]
(7). This was prepared as described in the literature. Spectroscopic
data were in accordance with those reported.13
Synthesis of {salen}AlCl. The syntheses of {salen}AlCl com-
plexes are adaptations of the procedure described by Rutherford
and Atwood in the literature.14 Further experimental details are given
below.
1
0.448 g (85%) of orange powder. H NMR (CDCl3): δ ) -0.31
[q (JH-JH, 8.1 Hz), 2H, AlCH2CH3], 0.72 [t (JH-JH, 8.1 Hz), 3H,
AlCH2CH3], 1.51 [s, 18H, C(CH3)], 3.90-3.93 [m, 2H, NCH2-
CH2N], 4.08-4.17 [m, 2H, NCH2CH2N], 8.18 [s, 2H, phenyl-H],
8.35 [s, 2H, phenyl-H], 8.50 [s, 2H, phenyl-CHdN]. Anal. Calcd
for C24H33N4O6Al: C, 58.53; H, 6.34; N, 10.68. Found: C, 57.68;
H, 6.13; N, 9.84.
Synthesis of {N,N′-Bis(3-tert-butyl-5-nitrosalicylidene)-1,2-
ethylenediamine}aluminum(III) Chloride [{(tBu)(NO2)salen}-
AlCl] (1c). A 50 mL Schlenk flask was charged with (tBu)(NO2)-
salenH2 (0.471 g, 1.0 mmol), which was then dissolved in 40 mL
of toluene. A 100 mL Schlenk flask was charged with 0.53 mL
(1.0 mmol) of a 1.9 M toluene solution of Et2AlCl, and an additional
10 mL of toluene was added. The ligand was cannulated into the
flask containing the metal, and the mixture was stirred for 12 h at
room temperature. A yellow precipitate was observed. The reaction
Synthesis of {N,N′-Bis(3-tert-butyl-5-nitrosalicylidene)-1,2-
phenylenediamine}aluminum(III) Ethyl [{(phen)(tBu)(NO2)-
salen}AlEt] (2b). Method B was followed with (phen)(tBu)(NO2)-
salenH2 (0.250 g, 0.54 mmol) and AlEt3 (0.30 mL of a 1.9 M
toluene solution, 0.54 mmol). This yields 0.253 g of reddish-brown
powder (82%). H NMR (DMSO-d6): δ ) -0.53 [q (JH-JH, 8.1
Hz), 2H, AlCH2CH3], 0.48 [t (JH-JH, 8.1 Hz), 3H, AlCH2CH3],
1.53 [s, 18H, C(CH3)], 8.13-8.20 [m, 4H, N-phenyl-H], 8.60 [s,
1
(11) (a) Atwood, D. A.; Hill, M. S.; Jegier, J. A.; Rutherford, D.
Organometallics 1997, 16, 2659. (b) Sahasrabudhe, S.; Yearwood,
B. C.; Atwood, D. A. Inorg. Synth. 2004, 34, 14-20.
(13) Leung, W.-H.; Chan, E. Y. Y.; Chow, E. K. F.; Williams, I. D.; Peng,
S.-M. J. Chem. Soc., Dalton Trans. 1996, 1229.
(14) Rutherford, D.; Atwood, D. A. Organometallics 1996, 15, 4417.
(12) Dzugan, S. J.; Goedken, V. L. Inorg. Chem. 1986, 25, 2858.
Inorganic Chemistry, Vol. 44, No. 5, 2005 1435

Darensbourg and Billodeaux
mixture was concentrated to ∼10 mL, and 30 mL of hexanes was
added to precipitate the product. The precipitate was collected by
filtration and dried under vacuum to yield 0.480 g (91%) of yellow
AlCl] (12). This was prepared according to the literature procedure.
Yield and spectroscopic data were in accordance with those reported
in the literature.14
1
powder. H NMR(DMSO-d6): δ ) 1.53 [s, 18H, C(CH3)3], 3.96
Synthesis of {N,N′-Bis(3,5-di-tert-butylsalicylidene)-1,2-
[s, 4H, NCH2CH2N], 8.18-8.19 [d, 2H, phenyl-H], 8.45-8.46 [d,
2H, phenyl-H], 8.81 [s, 2H, phenyl-CHdN]. Anal. Calcd for
C24H28N4O6AlCl: C, 54.29; H, 5.32; N, 10.55. Found: C, 54.24;
H, 5.48; N, 10.16.
Synthesis of {N,N′-Bis(3,5-di-tert-butylsalicylidene)-1,2-eth-
ylenediamine}aluminum(III) Chloride [{(tBu)2salen}AlCl] (8).
This was prepared according to the literature procedure. Yield and
spectroscopic data were in accordance with those reported in the
literature.14
ethylenediamine}aluminum(III)
2,4,6-Trimethylphenoxide
[{(tBu)2salen}Al(OC6H2(CH3)3)] (13). A 100 mL Schlenk flask
fitted with a reflux condenser was charged with (tBu)2salenH2 (0.500
g, 1.0 mmol), which was then dissolved in 30 mL of toluene. A 50
mL Schlenk flask was charged with 0.53 mL (1.0 mmol) of a 1.9
M toluene solution of AlEt3. An additional 10 mL of toluene was
added, and the solution was cannulated onto the ligand mixture.
The reflux condenser was washed with 5 mL of toluene. The
mixture was stirred under reflux 1 h, cooled, and stirred at room
temperature 1 h to yield a bright yellow solution. A 10 mL toluene
solution of 2,4,6-trimethylphenol was added and the mixture stirred
1 h. The solution was concentrated to ∼10 mL, and 30 mL hexanes
was added to precipitate the product. The precipitate was collected
by filtration and dried under vacuum to yield 0.340 g (52%) of a
Synthesis of {N,N′-Bis(3,5-dichlorosalicylidene)-1,2-eth-
ylenediamine}aluminum(III) Chloride [{Cl2salen}AlCl] (9). A
50 mL Schlenk flask was charged with Cl2salenH2 (0.219 g, 0.54
mmol), which was then dissolved in 40 mL of toluene. A 100 mL
Schlenk flask was charged with 0.30 mL (1.0 mmol) of a 1.9 M
toluene solution of Et2AlCl, and an additional 10 mL of toluene
was added. The ligand was cannulated into the flask containing
the metal, and the mixture was stirred for 12 h at room temperature.
A yellow precipitate was observed. The reaction mixture was
concentrated to ∼10 mL, and 30 mL of hexanes was added to
precipitate the product. The precipitate was collected by filtration
and dried under vacuum to yield 0.224 g (89%) of yellow powder.
1H NMR (DMSO-d6): δ ) 3.91 [s, 4H, NCH2CH2N], 7.56-7.57
[m, 2H, phenyl-H], 7.69-7.72 [m, 2H, phenyl-H], 8.60 [s, 2H,
phenyl-CHdN].
Synthesis of {N,N′-Bis(salicylidene)-1,2-phenylenediamine}-
aluminum(III) Chloride [{(phen)salen}AlCl] (10). A 50 mL
Schlenk flask was charged with (phen)salenH2 (0.454 g, 1.0 mmol),
which was then dissolved in 30 mL of CH3CN. A 50 mL Schlenk
flask was charged with 0.53 mL (1.0 mmol) of a 1.9 M toluene
solution of Et2AlCl. A 10 mL volume of CH3CN was added, and
the solution was cannulated onto the ligand mixture. The immediate
formation of a dark orange-red solution was observed. The mixture
was stirred at room temperature 2 h, and an orange precipitate began
to form. The solution was concentrated under reduced pressure to
∼10 mL, and 30 mL of hexanes was added to completely precipitate
the product. The precipitate was collected by filtration and dried
under vacuum to yield 0.420 g (87%) of orange powder. 1H NMR
(DMSO-d6): δ ) 6.81 [t (JH-JH, 7.8 Hz), 2H, phenyl-H], 6.95 [d,
2H, phenyl-H], 7.49-7.54 [m, 4H, N-phenyl-H], 7.69 [d, 2H,
phenyl-H], 8.15-8.18 [m, 2H, phenyl-H], 9.34 [s, 2H, phenyl-CHd
N]. Anal. Calcd for C16H10N2O2AlCl5: C, 41.19; H, 2.16; N, 6.00.
Found: C, 40.29; H, 2.19; N, 5.40.
1
yellow powder. H NMR (C6D6): δ ) 1.37 [s, 18H, C(CH3)3],
1.79 [s, 18H, C(CH3)3], 2.02 [s, 6H, phenoxide-CH3], 2.17 [s, 3H,
phenoxide-CH3], 2.54-2.63 [m, 2H, NCH2CH2N], 3.22-3.29 [m,
2H, NCH2CH2N], 6.72 [s, 2H, phenoxide-CH2], 6.95-6.96 [d (JH-
JH, 2.4 Hz), 2H, phenyl-CH], 7.46 [s, 2H, phenyl-CH2], 7.78 [d
(JH-JH, 2.5 Hz), 2H, phenyl-CHdN]. Anal. Calcd for C41H57N2O3-
Al: C, 75.43; H, 8.80; N, 4.29. Found: C, 75.12; H, 8.94; N, 4.59.
Synthesis of {N,N′-Bis(3,5-di-tert-butylsalicylidene)-1,2-eth-
ylenediamine}aluminum(III) Phenoxide [{(tBu)2salen}Al(OPh)]
(14). The method described above was repeated using (tBu)2salen}-
H2 (0.374 g, 0.76 mmol), AlEt3 (0.400 mL, 0.76 mmol), and phenol
(0.072 g, 0.76 mmol). This yielded 0.328 g (71%) of a bright yellow
1
powder. H NMR (C6D6): δ ) 1.39 [s, 18H, C(CH3)3], 1.83 [s,
18H, C(CH3)3], 2.51-2.60 [m, 2H, NCH2CH2N], 3.15-3.24 [m,
2H, NCH2CH2N], 6.68-6.70 [m, 2H, phenoxide-H], 6.73-6.78 [m,
2H, phenoxide-H], 6.95-6.96 [m, 2H, phenyl-H], 7.48 [s, 2H,
phenyl-H], 7.78-7.79 [m, 2H, phenyl-CHdN].
Copolymerization of Epoxides and CO2. (salen)AlZ (50 mg)
and the corresponding cocatalyst were dissolved in 20 mL of neat
epoxide. The solution was added via injection port into a 300 mL
Parr autoclave previously dried in vacuo at 80 °C overnight. The
autoclave was charged with 500 psi CO2 and maintained at 80 °C
for a period of 8 h. The autoclave was then cooled to room
temperature and vented in a fume hood. The polymer was extracted
with dichloromethane and dried under vacuum at 100 °C for a
period of 6 h.
1
The collected polycarbonate was analyzed by H NMR, where
the amount of ether linkages was determined by integrating the
peaks corresponding to the methane protons of polyether at ∼3.45
ppm and polycarbonate at ∼4.6 ppm. Further analysis was done
by infrared spectroscopy to identify monomeric trans-cyclohexyl
carbonate with ν(CdO) at 1825 cm-1 or cis-cyclohexyl carbonate
with ν(CdO) at 1804 cm-1. The percentage of trans-cyclohexyl
carbonate relative to copolymer was determined by integrating the
1H NMR peaks corresponding to the methane protons of the
polymer at ∼4.6 ppm and trans-cyclohexyl carbonate at ∼4.0 ppm.
High-Pressure in Situ Kinetic Measurements. High-pressure
kinetic measurements were carried out using a stainless steel Parr
autoclave modified with a SiComp window to allow for attenuated
total reflectance spectroscopy using infrared radiation (ASI ReactIR
1000 in situ probe). After the autoclave was dried in vacuo overnight
at 80 °C, 10 mL of neat epoxide was loaded via injection port.
After the background solvent reached 80 °C, a single 128-scan
background was collected. The desired catalyst and tetrabutylam-
monium salt cocatalyst were dissolved in 10 mL of neat epoxide
Synthesis of {N,N′-Bis(3,5-dichlorosalicylidene)-1,2-phen-
ylenediamine}aluminum(III) Chloride [{(phen)Cl2salen}AlCl]
(11). A 50 mL Schlenk flask was charged with (phen)Cl2salenH2
(0.250 g, 0.55 mmol), which was then dissolved in 30 mL of
CH3CN. A 50 mL Schlenk flask was charged with 0.31 mL (0.55
mmol) of a 1.9 M toluene solution of Et2AlCl. A 10 mL volume
of CH3CN was added, and the solution was cannulated onto the
ligand mixture. The immediate formation of an orange-yellow pre-
cipitate was observed. The mixture was stirred at room temperature
2 h and concentrated under reduced pressure to ∼10 mL, and 30
mL of hexanes was added to completely precipitate the product.
The precipitate was collected by filtration and dried under vacuum
to yield 0.235 g (83%) of yellow powder. H NMR (DMSO-d6):
δ ) 7.57-7.72 [m, 4H, N-phenyl-H], 7.97 [s, 2H, phenyl-H], 8.29
[s, 2H, phenyl-H], 9.64 [s, 2H, phenyl-CHdN]. Anal.
Synthesis of {N,N′-Bis(3,5-di-tert-butylsalicylidene)-1,2-phen-
ylenediamine}aluminum(III) Chloride [{(phen)(tBu)2salen}-
1
1436 Inorganic Chemistry, Vol. 44, No. 5, 2005

Catalytic System for Production of Polycarbonates
and injected into the autoclave followed by immediate charging
with 50 bar of CO2 pressure. With maintenance of a reaction
temperature of 80 °C, a single 128-scan spectrum was collected
every 3 min for 10 h. Profiles of the absorbance at 1750 cm-1
(polycarbonate) and ∼1825 cm-1 (cyclic carbonate) versus time
were recorded after baseline correction. After the sample was cooled
and vented in a fume hood, the polymer was extracted as a
dichloromethane solution and dried under vacuum at 100 °C
overnight. The polymer was weighed, and the carbonate and cyclic
1
content were then analyzed by H NMR as described previously.
These weights and NMR spectra were used to report the catalytic
reactivity listed here after.
Figure 1. Thermal ellipsoid plot of 1. Ellipsoids are at the 50% level.
X-ray Structural Studies. A Bausch and Lomb 10× microscope
was used to identify suitable crystals from a representative sample
of crystals of the same habit. Crystals were coated with mineral
oil, placed on a glass fiber, and mounted on a Bruker SMART
1000 CCD diffractometer. X-ray data were collected covering more
than a hemisphere of reciprocal space by a combination of three
sets of exposures. Each exposure had a different æ angle for the
crystal orientation, and each exposure covered 0.3° in ω. The crystal
to detector distance was 4.9 cm. Decay was monitored by repeating
collection of the initial 50 frames collected and analyzing the
duplicate reflections. Crystal decay was negligible. The space group
was determined on the basis of systematic absences and intensity
statistics.13 The structure was solved by direct methods and refined
by full-matrix least squares on F2. All non-hydrogen atoms were
refined with anisotropic displacement parameters. All H atoms were
placed in idealized positions with fixed isotropic displacement
parameters equal to 1.5 times (1.2 for methyl protons) the equivalent
isotropic displacement parameters of the atom to which they are
attached.
phenolate ligand and an electron-withdrawing nitro group
at the 5-position. These two ligands (tBu)(NO2)salenH2 (1)
and (phen)(tBu)(NO2)salenH2 (2) were synthesized by fol-
lowing established procedures whereby 2 equiv of the
appropriate salicylaldehyde are reacted with 1 equiv of an
appropriate diamine in an alcohol solvent. Complete synthetic
and spectroscopic details are presented in the Experimental
Section. X-ray quality crystals of 1 were obtained from a
concentrated dichloromethane solution by slow evaporation.
A thermal ellipsoid representation is provided in Figure 1.
Crystal and data collection parameters are listed in Table 1.
Synthesis and Structures of Aluminum Salen Com-
plexes. The direct reaction of salenH2 with Et2AlZ (Z ) Et
or Cl) proceeds rapidly with evolution of 2 equiv of ethane
to give the target (salen)AlZ complexes in good yield as
indicated in Scheme 1. It is important to note that the reaction
may be readily conducted in toluene when the salen diimine
is aliphatic but requires acetonitrile as the reaction solvent
when the salen diimine is aromatic. When Z ) Et, the axial
ligand can be replaced by direct exchange with a phenol
yielding 13 and 14.
The following programs were used: data collection and cell
refinement, SMART;15 data reduction, SAINTPLUS(Bruker16);
programs used to solve structures, SHELXS-97 (Sheldrick17);
programs used to refine structures, SHELXL-99 (Sheldrick18);
molecular graphics and publication materials, SHELXTL-Plus
version 5.0 (Bruker19).
X-ray quality crystals of 4, 7, 13, and 14 were obtained.
The solid-state structures show that all four display the
distorted square pyramidal geometry typical of group 13
salen complexes.11,12,20-24 Crystal and data collection pa-
rameters are given in Table 1. Crystals of 4‚CH2Cl2 were
obtained by slow evaporation of a concentrated CH2Cl2
solution at 10 °C. 4‚CH2Cl2 crystallizes in the monoclinic
space group P21/c with two independent molecules and a
molecule of CH2Cl2 in the asymmetric unit. A thermal
ellipsoid plot is provided in Figure 2. Crystals of 7 were
obtained by slow evaporation of a concentrated CH2Cl2
solution at 10 °C. 7 crystallizes in the monoclinic space group
P21/c. The two carbons of the cyclohexyldiimine bound to
the nitrogen atoms are disordered with the major orientation
at 52% occupancy. A thermal ellipsoid plot is provided in
Figure 3. Selected bond lengths and angles of 4 and 7 are
Results and Discussion
Previously we have developed an extensive library of salen
ligands bearing electron-donating group for our investigations
of chromium(III) salen derivatives as catalysts for the
copolymerization of cyclohexene oxide and CO2, for these
were found to be most effective.7,8 During the focus of this
report we have established that more electron deficient salen
ligands provide better aluminum(III) salen complexes which
serve as catalysts for the epoxide/CO2 coupling process (vide
infra). Unfortunately, salen ligands containing electron-
withdrawing substituents on the phenolate groups (e.g.,
chloride) are often insoluble in common organic solvents,
in particular epoxides. To combat this we have synthesized
two new salen ligands possessing both a solubilizing,
electron-donating tert-butyl group at the 3-position of the
(20) Gurian, P. L.; Cheatham, L. K.; Ziller, J. W.; Barron, A. R. J. Chem.
Soc., Dalton Trans. 1991, 1449.
(21) Cameron, P. A.; Jhurry, D.; Gibson, V. C.; White, A. J. P.; Williams,
D. J.; Williams, S. Macromol. Rapid Commun. 1999, 20, 616.
(22) Duxbury, J. P.; Warne, J. N. D.; Mushtaq, R.; Ward, C.; Thornton-
Pett, M.; Jian, M.; Greatrex, R.; Kee, T. P. Organometallics 2000,
19, 4445.
(15) SMART 1000 CCD; Bruker Analytical X-ray Systems: Madison, WI,
1999.
(16) SAINT-Plus, version 6.02; Bruker: Madison, WI, 1999.
(17) Sheldrick, G. SHELXS-97: Program for Crystal Structure Solution;
Institut fur Anorganische Chemie der Universitat: Gottingen, Ger-
many, 1997.
(18) Sheldrick, G. SHELXL-99: Program for Crystal Structure Refinement;
Institut fur Anorganische Chemie der Universitat: Gottingen, Ger-
many, 1999.
(23) Mun˜oz-Hernandez, M.-A.; Keizer, T. S.; Parkin, S.; Zhang, Y.;
Atwood, D. A. J. Chem. Crystallogr. 2000, 30, 219.
(24) Mun˜oz-Hernandez, M.-A.; Keizer, T. S.; Wei, P.; Parkin, S.; Atwood,
D. A. Inorg. Chem. 2001, 40, 6782.
(19) SHELXTL, version 5.0; Bruker: Madison, WI, 1999.
Inorganic Chemistry, Vol. 44, No. 5, 2005 1437

Darensbourg and Billodeaux
Table 1. Crystal Data and Structure Refinement
1
4‚CH2Cl2
7
13
14
empirical formula
fw
C18H22.50N3O4.50
352.89
C22H18AlN2O2‚CH2Cl2
454.3
C38H55AlN2O2
598.82
C41H57AlN2O3
652.87
C38H51AlN2O3
610.79
temp (K)
cryst system
space group
a (Å)
b (Å)
c (Å)
110(2)
rhombohedral
R3h
24.557(4)
24.557(4)
10.193(2)
90
90
120
5323.7(17)
1.321
12
110(2)
monoclinic
P21/c
15.977(4)
28.209(7)
9.176(2)
90
106.069(5)
90
3974.1(16)
1.380
4
0.258
24 045
110(2)
monoclinic
P21/n
17.223(5)
10.372(3)
19.645(6)
90
99.558(7)
90
3460.4(17)
1.149
4
0.093
20 643
110(2)
triclinic
P1h
110(2)
triclinic
P1h
10.004(9)
14.185(12)
15.389(13)
115.006(14)
96.996(16)
96.653(17)
1930(3)
1.123
11.608(6)
11.909(6)
15.176(7)
111.451(10)
109.693(9)
95.296(9)
1781.9(15)
1.138
R (deg)
â (deg)
γ (deg)
V (Å3)
Dc (Mg/m3)
Z
2
2
µ (mm-1
)
0.096
10 985
0.090
8953
0.094
11 761
reflcns collcd
indpndt reflcns
2696
8791
7732
5491
7713
params
158
0.772
R1 ) 0.1012
Rw ) 0.2186
R1 ) 0.2232
Rw ) 0.2670
516
1.015
R1 ) 0.0865
Rw ) 0.1908
R1 ) 0.1658
Rw ) 0.2331
400
1.129
R1 ) 0.0990
Rw ) 0.2215
R1 ) 0.2713
Rw ) 0.2851
439
1.013
R1 ) 0.0513
Rw ) 0.1195
R1 ) 0.0750
Rw ) 0.1314
409
1.018
R1 ) 0.0531
Rw ) 0.1212
R1 ) 0.0904
Rw ) 0.1522
goodness-of-fit on F2
final R indices [I > 2σ(I)]
final R indices (all data)
Scheme 1
found in Tables 2 and 3. Despite the differences in electronics
provided by the salen ligands of these two complexes, there
are remarkable similarities in their geometries. The Al-C
bond lengths of the axial ethyl groups are in close agreement
(1.968(5) Å vs 1.960(9) Å), and the coordination sphere of
the salen ligands are nearly identical. Both complexes are
in concurrence with other published (salen)Al(alkyl) struc-
tures in terms of bond lengths and angles.11,12
Crystals of 13 and 14 were obtained by slow diffusion of
pentane into a concentrated THF solution at -30 °C. 13 and
14 crystallize in the triclinic space group P1h. As would be
expected, the geometries of these two complexes are very
nearly identical with the greatest deviation belonging to the
Al1-O2 bond length (1.792(2) Å vs 1.7717(17) Å). Ther-
mal ellipsoid plots are provided in Figures 4 and 5. Bond
lengths and angles in 13 are in agreement with those of
{N,N′-bis(salicylidene)diethylimine}aluminum(III) 2,4,6-tri-
methylphenoxide previously published and are given in
Tables 4 and 5.20
It is interesting to note that when Z ) Et, as in 4 and 7,
the basal plane formed from N1, N2, O1, and O2 is quite
planar with the rms deviation from planarity of the atoms
equal to 0.0669(16) and 0.0736(16) Å for the two indepen-
dent molecules of 4 and 0.0696(34) Å for 7. The Al atom
sits an average of 0.551(7) Å above the plane in the two
molecules of 4 and 0.5484(41) Å in 7. When Z ) phenoxide,
as in 13 and 14, the same plane is more distorted. The rms
deviation for 13 ) 0.2321(10) Å, but the Al atom is located
0.4915(13) Å from the plane and the corresponding measure-
ments are 0.2211(10) and 0.4626(10) Å for 14.
1438 Inorganic Chemistry, Vol. 44, No. 5, 2005

Catalytic System for Production of Polycarbonates
Figure 4. Thermal ellipsoid of 13. Ellipsoids are at the 50% level. H
atoms are omitted for clarity.
Figure 2. Thermal ellipsoid plot of 4‚CH2Cl2. Ellipsoids are at the 50%
probability level. H atoms and one solvent molecule are omitted for clarity.
Figure 5. Thermal ellipsoid of 14. Ellipsoids are at the 50% level. H
atoms are omitted for clarity.
Figure 3. Thermal ellipsoid plot of 7. Ellipsoids are at the 50% level.
Only the major orientation of the disordered carbons adjacent to N1 and
N2 is shown. H atoms are omitted for clarity.
Table 4. Selected Bond Lengths (Å) for 13 and 14
13
14
Table 2. Selected Bond Lengths (Å) for 4‚CH2Cl2 and 7
Al(1)-O(1)
Al(1)-O(2)
Al(1)-O(3)
Al(1)-N(1)
Al(1)-N(2)
1.800(2)
1.792(2)
1.750(2)
1.986(3)
2.001(2)
Al(1)-O(1)
Al(1)-O(2)
Al(1)-O(3)
Al(1)-N(1)
Al(1)-N(2)
1.8014(17)
1.7717(17)
1.7476(17)
1.985(2)
Al(1)-O(1A)
Al(1)-O(2A)
Al(1)-N(1A)
Al(1)-N(2A)
Al(1)-C(21A)
1.818(3)
1.814(3)
2.026(4)
2.036(4)
1.965(5)
Al(2)-O(1B)
Al(2)-O(2B)
Al(2)-N(1B)
Al(2)-N(2B)
Al(2)-C(21B)
1.808(3)
1.823(3)
2.051(4)
2.014(4)
1.968(5)
1.9931(19)
Al(1)-O(1)
Al(1)-O(2)
Al(1)-N(1)
1.799(6)
1.816(5)
2.046(7)
Al(1)-N(2)
Al(1)-C(1)
2.015(7)
1.960(9)
Table 5. Selected Bond Angles (deg) for 13 and 14
13
14
Table 3. Selected Bond Angles (deg) for 4‚CH2Cl2 and 7
O(2A)-Al(1)-O(1A) 86.80(14) O(2B)-Al(2)-O(1B)
O(3)-Al(1)-O(2)
O(3)-Al(1)-O(1)
O(2)-Al(1)-O1)
O(3)-Al(1)-N(1)
O(2)-Al(1)-N(1)
O(1)-Al(1)-N(1)
O(3)-Al(1)-N(2)
O(2)-Al(1)-N(2)
O(1)-Al(1)-N(2)
N(1)-Al(1)-N(2)
114.85(9)
103.83(10)
91.73(10)
109.33(9)
134.14(10)
89.20(10)
91.74(10)
88.01(11)
162.89(9)
78.84(11)
O(3)-Al(1)-O(2)
O(3)-Al(1)-O(1)
O(2)-Al(1)-O(1)
O(3)-Al(1)-N(1)
O(2)-Al(1)-N(1)
O(1)-Al(1)-N(1)
O(3)-Al(1)-N(2)
O(2)-Al(1)-N(2)
O(1)-Al(1)-N(2)
N(1)-Al(1)-N(2)
114.59(8)
103.79(8)
90.83(8)
107.33(8)
136.71(8)
89.33(8)
90.63(8)
89.24(8)
164.06(8)
79.77(8)
88.44(14)
O(2A)-Al(1)-C(21A) 110.35(17) O(2B)-Al(2)-C(21B) 113.69(19)
O(1A)-Al(1)-C(21A) 109.25(18) O(1B)-Al(2)-C(21B) 107.18(18)
O(2A)-Al(1)-N(1A) 141.75(15) O(2B)-Al(2)-N(1B) 141.24(16)
O(1A)-Al(1)-N(1A)
C(21A)-Al(1)-N(1A) 107.01(17) C(21B)-Al(2)-N(1B) 104.21(18)
87.47(14)
88.56(14) O(1B)-Al(2)-N(1B)
87.98(14)
O(2A)-Al(1)-N(2A)
88.03(14) O(2B)-Al(2)-N(2B)
O(1A)-Al(1)-N(2A) 151.03(15) O(1B)-Al(2)-N(2B) 149.97(16)
C(21A)-Al(1)-N(2A) 99.27(17) C(21B)-Al(2)-N(2B) 101.69(17)
N(1A)-Al(1)-N(2A)
O(2)-Al(1)-O(1)
O(2)-Al(1)-C(1)
O(1)-Al(1)-C(1)
O(2)-Al(1)-N(1)
O(1)-Al(1)-N(1)
78.12 (15) N(1B)-Al(2)-N(2B)
76.95 (14)
101.8(3)
87.9(3)
141.6(3)
106.8(4)
77.5(3)
Copolymerization of Cyclohexene Oxide and Carbon
Dioxide. Our initial studies utilizing aluminum salen deriva-
tives as potential catalysts for the coupling of cyclohexene
oxide and CO2 focused on salen ligand frameworks similar
to those found in our most active chromium catalysts for
this process.7 That is, we investigated complexes processing
electron-donating tert-butyl substituents in the 3- and 5-posi-
tions of the phenolate rings, 7 and 8, paired with the most
effective Lewis base cocatalysts (phosphines or PPNX salts).
The experiments involving phosphines (PPh3 or PCy3) as
cocatalysts for the copolymerization reaction proved to be
unsuccessful, producing low molecular weights oligomers
with ether linkages in excess of 50% and large amounts of
cyclic carbonate. Reactions carried out in the presence of a
89.6(2)
106.8(3)
110.6(4)
150.6(3)
86.5(3)
C(1)-Al(1)-N(1)
O(2)-Al(1)-N(2)
O(1)-Al(1)-N(2)
C(1)-Al(1)-N(2)
N(1)-Al(1)-N(2)
Thus far, all other published (salen)AlZ structures possess
the square pyramidal geometry displayed by the complexes
reported herein. Nevertheless, it is possible that complexes
such as complex 10 which contains a salen ligand devoid of
substituents in the 3- and 5-positions of the phenolate rings
might be dimeric, containing chloride bridging groups
between the two metal centers.23 Unfortunately, we were
unable to obtain suitable crystals of 10 for characterization
by X-ray analysis.
Inorganic Chemistry, Vol. 44, No. 5, 2005 1439

Darensbourg and Billodeaux
Table 6. Copolymerization of Cyclohexene Oxide and Carbon
[PPN][Cl] cocatalyst produced only cyclic carbonate. In an
attempt to increase the reactivity of these complexes, the
axial ligand was changed to a phenoxide, complexes 13 and
14. Nevertheless, these alterations did little to enhance the
reactivity of these complexes, producing copolymer detect-
able by infrared spectroscopy but not in isolable quantities
(<0.15 g).
Dioxidea
d
run
catalb
cocatalb
TOFc
% cyclicd
% CO2
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
1b
1b
1b
1c
1c
1c
1c
3
3
3
4
4
4
9
9
9
10
10
10
Bu4NCl
Bu4NN3
Bu4NOAc
Bu4NCl
Bu4NN3
Bu4NOAc
None
Bu4NCl
Bu4NN3
Bu4NOAc
Bu4NCl
Bu4NN3
Bu4NOAc
Bu4NCl
Bu4NN3
Bu4NOAc
Bu4NCl
Bu4NN3
Bu4NOAc
15.0
20.7
28.2
35.4
27.2
25.8
24.2
20.6
7.3
4
2
1
3
<1
2
10
5
2
3
4
1
3
5
4
98
95
98
96
>99
98
74
98
96
97
90
93
95
94
97
92
99
99
99
Recent studies by Gibson and co-workers have indicated
that increasing the electrophilicity of the metal center by
adding electron-withdrawing groups to the phenolate rings
of the salen ligands increases the complexes’ ability to
catalyze the polymerization of (D,L)- and (L)-lactide at
ambient temperatures.21 We have employed a similar strategy
by removing the tert-butyl groups from the phenolate ring,
utilizing complexes 4 and 10. Employing 4 as a catalyst with
PPNCl provided poly(cyclohexene carbonate) in good yield
but with a sizable quantity of cyclic carbonate. However,
attempts at using complex 10 with PPNCl as a cocatalyst
proved ineffective due to their insolubility in a common
organic solvent.25 In addition, efforts to increase the elec-
trophilicity of the metal center by employing highly electron
withdrawing chloro groups at the 3- and 5-positions of the
phenolate rings were accomplished with the synthesis of
complexes 3 and 11. However, to obtain highly soluble
electron-deficient metal salen complexes it was necessary
to introduce solubilizing electron-donating tert-butyl groups
in the 3-positions offset with strongly electron-withdrawing
nitro groups in the 5-position. Using these salen ligands,
complexes 1b,c and 2b,c were prepared.
26.0
5.2
15.3
17.7
23.2
18.0
12.7
11.5
23.3
28.7
7
3
4
2
a 500 psi of CO2 at 80 °C. b 50 mg of catalyst and 1 equiv of cocatalyst.
c mol of epoxide consumed/(mol of Al‚h) based on 8 h reaction times.
1
d Estimated via H NMR.
mol of epoxide consumed/(mol of Al‚h). The worst among
the active species, (phen)salenAlEt and Bu4NCl (Table 6,
entry 11), has a TOF ) 5.2 mol of epoxide consumed/
(mol of Al‚h). However, we could not establish any trends
for the series of catalysts or cocatalysts. For example, while
4 follows the trend Cl- < N3- < OAc- in terms of increasing
-
reactivity, 1c follows the trend N3 < OAc- < Cl-.
Nevertheless, there were not major differences in the
activities of these aluminum salen complexes with variation
in these Bu4N+ salts. The catalysts were however very
susceptible to changes in the electronic character of the salen
ligand. For example, the modification from (tBu)(NO2)-
salenAlEt (1b) to (NO2)salenAlEt (6) produced an inac-
tive species in combination with Bu4N+ salts as cocatalysts.
Alteration of the diimine backbone can also drastically
effect catalytic activity. That is, active catalyst/cocatalyst
combinations, (tBu)(NO2)salenAlEt/Bu4NN3 (Table 6, entry
2) and {(Cl)2salen}AlEt/Bu4NN3 (Table 6, entry 9), be-
come completely inactive by changing the diimine back-
bone from ethylenediamine to 1,2-phenylenediamine. Like-
wise, a change from {(tBu)2salen}AlCl (8) to {(phen)-
(tBu)2salen}AlCl (12) produces an active species in conjunc-
tion with Bu4NN3. This particular combination produces a
large amount of cyclic carbonate in addition to the copoly-
mer, but it is interesting to note the significant increase in
activity over that of complex 8 under similar reaction
conditions.
On the basis of previous studies which have shown that
aluminum porphyrinates and aluminum salen derivatives
produce polycarbonates and cyclic carbonates,1c,26 respec-
tively, in the presence of tetrabutylammonium salts
(Bu4N+X-), we have examined these salts as cocatalysts in
conjunction with our aluminum salen complexes for their
ability to copolymerize cyclohexene oxide and carbon
dioxide. These salts have the benefits over their PPN+
analogues of being relatively inexpensive and readily sol-
uble in epoxide. For example, the combination of any of
our aluminum salen complexes with a Bu4N+ salt pro-
duced a complex soluble in cyclohexene oxide. We have
conducted copolymerization studies with 1b,c, 3, 4, 9, and
10 in combination with Bu4N+X- (X ) N3, Cl, or OAc).
Results of these copolymerization studies are summarized
in Table 6.
These studies show that, in almost all cases, the (salen)-
AlCl complexes are more active than their (salen)AlEt
counterparts as displayed by the turnover frequency (TOF).
The most active catalyst/cocatalyst combination, (tBu)(NO2)-
salenAlCl and Bu4NCl (Table 6, entry 4), has a TOF ) 35.4
Since the TOF listed in Table 6 were determined from
isolated yields, in situ infrared monitoring of one catalyst
system, specifically complex 1c with the various Bu4N+X-
cocatalysts, was performed for comparative purposes. Fig-
ure 6 displays the time traces of the absorbances due to the
(25) It is necessary to first stir the solubilized metal salen complexes and
PPNCl salt in a common organic solvent (typically CH2Cl2 or benzene/
methanol), thereby forming the six-coordinate anionic complex which
is soluble in the epoxide. The thus formed metal derivative is dried
under vacuum prior to its use as a catalyst.
(26) (a) Lu, X.-B.; Feng, X.-J.; He, R. Appl. Catal., A 2002, 234, 25. (b)
Lu, X.-B.; Zhang, Y.; Bin, L.; Hui, W.; He, R. Chin. J. Catal. 2003,
24, 317. (c) Lu, X.-B.; Zhang, Y.; Liang, B.; Li, X.; Wang, H. J.
Mol. Catal., A 2004, 210, 31. (d) Lu, X.-B.; Zhang, Y.-J.; Jin, K.;
Luo, L.-M.; Wang, H. J. Catal. 2004, 227, 537-541.
νCO vibrational mode of the copolymer at 1750 cm-1 for
2
these processes. As can be seen from a comparison of the
data in Table 6 with Figure 6, there is little difference
noted for the three binary catalyst systems studied, with the
1c/Bu4N+Cl- combination being the most effective. Fur-
1440 Inorganic Chemistry, Vol. 44, No. 5, 2005

Catalytic System for Production of Polycarbonates
Scheme 2
interacting DMAP cocatalyst vs 1 equiv. The increase in the
CO2 content noted in the copolymer afforded upon in-
creasing the relative concentration of N-MeIm to 1c (run
2 vs run 1 in Table 7) is consistent with the CO2 inser-
tion process being greatly accelerated in the presence of a
good donating ligand trans to the growing polymer chain
(see eq 2). A similar explanation accounts for the lower
percentage of carbonate linkages in the copolymer produced
utilizing the poorer donating Lewis base, pyridine as co-
catalyst (runs 5 and 9 in Tables 7). These observations are
in complete agreement with the proposed mechanism,
supported by experimental data, described by Chisholm and
Zhou for the (TPP)AlX (TPP ) tetraphenylporphyrin)
catalyzed coupling of propylene oxide and CO2 in the
presence of DMAP.4
Figure 6. Peak profiles of the carbonate signal at 1750 cm-1 for in situ
infrared monitoring of the copolymerization reactions involving 1c and three
tetrabutylammonium salts.
Table 7. Copolymerization of Cyclohexene Oxide and Carbon
Dioxidea
d
run
catal
cocatalb
TOFc
% cyclicd
% CO2
1
2
3
4
5
6
7
8
9
1c
1c
1c
1c
1c
1b
10
10
10
4
N-MeIm (1)
N-MeIm (2.5)
DMAP (1)
DMAP (3)
pyridine (1)
DMAP (1)
N-MeIm (2.5)
DMAP (1)
pyridine (2.5)
N-MeIme
14.9
17.0
27.7
32.0
7.6
18.0
7.2
14.9
2.5
6
3
2
2
6
4
4
3
5
81
93
97
96
61
90
92
91
85
48
10
1.9
34
a 500 psi of CO2 at 80 °C. b Equivalents of cocatalyst show in
parentheses. c mol of epoxide consumed/(mol of Al‚h) based on 8 h re-
action times. d Estimated via 1H NMR. e In the presence of 1 equiv of
methanol.
Concomitantly, in the absence of a strong coordinating
Lewis base cocatalyst the active five-coordinate aluminum
species containing the growing polymer chain exists as the
alkoxide intermediate for an extended period of time, thereby
allowing for production of the cyclic carbonate via the back-
biting mechanism as depicted in eq 3.
thermore, it is of importance to note that the (salen)AlCl
catalyst (1c) in the absence of a Bu4N+X- cocatalyst affords
a copolymer with a substantial reduction in CO2 content
along with a significant rise in cyclic carbonate production.
As mentioned earlier, there are several reports of aluminum
porphyrin derivatives employed as catalysts in conjunction
with the Lewis base cocatalysts, 1-methylimidazole (N-
MeIm) and 4-(dimethylamino)pyridine (DMAP), for the
copolymerization of epoxides and carbon dioxide.1,4 Simi-
larly, we and others have carried out extensive studies of
this process utilizing a wide variety of (salen)CrX catalysts
in the presence of N-MeIm and DMAP.5,7 Hence, we have
examined herein the activity of our more effective (salen)-
AlZ catalysts in the presence of the neutral Lewis bases
N-MeIm, DMAP, and pyridine. The results of these studies
are summarized in Table 7.
The most striking example of this can be seen in run 7 of
Table 6 for the (salen)AlCl catalyst (1c) in the absence of a
cocatalyst. Similar trends in product selectivity are noted for
the (salen)CrCl-catalyzed coupling of cyclohexene oxide or
propylene oxide and CO2 as the carbon dioxide pressure is
reduced.5c,27 Finally, an attempt to activate the potentially
most active (salen)AlEt derivative (complex 4) with 1 equiv
of methanol in the presence of the cocatalyst N-MeIm was
unsuccessful (run 10 in Table 7).28 Consistent with the
detailed mechanistic studies of Chisholm and Zhou4 for the
(TPP)AlX/DMAP system, we propose the initiation steps to
It is evident from the results reported in Table 7 that for
both complex 1c and 10 the order of activity as a function
of cocatalyst is DMAP > N-MeIm > pyridine. Further-
more, as has been observed in the (salen)CrX-catalyzed
system, complex 1c in the presence of 1 equiv of N-MeIm
results in a lower TOF as well as a reduction in carbonate
linkages in the copolymer than in the presence of 3 equiv of
N-MeIm. A similar less dramatic increase in TOF is seen
for 1c in the presence of 3 equiv of the more strongly
(27) Darensbourg, D. J.; Mackiewicz, R. M.; Billodeaux, D. R. Organo-
metallics 2004, in press.
(28) Consistent with the lack of activity of complex 4 under these
circumstances it was found via 1H NMR to be unchanged after
extended reaction time in refluxing methanol. By way of contrast, the
aluminum salen derivative containing an electron-donating salen ligand,
complex 7, easily underwent alcoholysis at ambient temperature.
Inorganic Chemistry, Vol. 44, No. 5, 2005 1441

Darensbourg and Billodeaux
Concluding Remarks
We have synthesized and characterized an array of five-
coordinate aluminum salen complexes with various perturba-
tions of the salen ligand via modifications at the 3- and
5-positions of the phenolate ring as well as the diimine
backbone. These aluminum derivatives were employed as
catalysts in conjunction with a series of anionic and neutral
cocatalysts to affect the copolymerization of carbon dioxide
and cyclohexene oxide. Our studies have shown that, in
general, a more electron-withdrawing salen framework is
necessary to produce significant quantities of copolymer with
a high CO2 content and an absence of the concurrent syn-
thesis of sizable quantities of cyclic carbonate. In all vari-
ations of salen ligand, apical group, and cocatalyst, the alum-
inum salen catalysts are considerably less active than our
previously described chromium salen systems; that is, the
TOF obtained in this instance ranges from 5.2 to 35.4 h-1,
while chromium systems under similar conditions possess
TOF’s as high as 1150 h-1. Despite their lower activities
for the coupling of CO2 and epoxides these aluminum salen
complexes aid in supporting some of our earlier mechanistic
proposals put forth for the chromium salen derivatives. For
example, an increase in the concentration and/or donating
ability of the neutral Lewis base cocatalysts, or species de-
riVed therefrom, enhances both TOF and CO2 incorporation.
Figure 7. IR stack plot showing ring-opening of epoxide by (salen)AlEt
and Bu4NN3 in three steps. (a) Bu4NN3 is added to reactor at 80 °C in
13 mL of toluene giving νN ) 1999 cm-1. (b) At t ) 24 min, 1 equiv
of (salen)AlEt is added in3 5 mL of toluene with an immediate de-
crease of signal at 1999 cm-1 and νN shifts to 2073 cm-1 indi-
3
cating formation of six-coordinate aluminum species. (c) At t ) 81 min,
276 equiv (5 mL) of cyclohexene oxide is added and νN shifts to
3
2095 cm-1 indicating formation of ring-opened epoxide product. At t )
180 min the reactor was charged with 200 psi of CO2 and growth of signal
at 1750 cm-1 was observed indicating insertion of CO2 and copolymer
formation.
proceed via the epoxide ring-opening processes depicted in
Scheme 2. In situ infrared spectra of the ring-opening process
described in A involving a (salen)AlEt catalyst in the
presence of the Bu4NN3 cocatalyst are contained in Figure
7. The process described in run 10 of Table 7 would be
initiated via a species analogue to C, where chloride would
be replaced by methoxide. It is important to recall here
that the ring-opening of epoxides catalyzed by aluminum
tetraphenylporphyrin complexes has been shown to be first-
order in [Al].4 This observation is consistent with the fact
that when utilizing chiral aluminum salen derivatives as
catalysts, this ring-opening process occurs with high %
conversion but low % ee. On the other hand, the chiral
chromium salen analogue complexes, which have been
shown to ring open epoxides via a process second-order
in [Cr], proceeds with both high activity as well as high
% ee.29
Acknowledgment. Financial support from the National
Science Foundation (Grant CHE 02-34860) and the Robert
A. Welch Foundation is greatly appreciated.
Supporting Information Available: Complete details for the
crystallographic studies of complexes 1, 4‚CH2Cl, 7, 13, and 14 in
CIF format. This material is available free of change via the Internet
at http://pubs.acs.org.
IC048508G
(29) Jacobsen, E. N. Acc. Chem. Res. 2000, 33, 421-431.
1442 Inorganic Chemistry, Vol. 44, No. 5, 2005