Optimization of variables for screening solid-supported metal complexes as oxidation catalysts

Article abstract of DOI:10.1016/S0040-4039(00)00850-9

Evaluation of polymer-supported metal complexes as oxidation catalysts revealed that product distribution is not dependent on metal stoichiometry, or resin but is dependent on the oxidant, oxidant stoichiometry and the metal incubation times. The model sy

Full text of DOI:10.1016/S0040-4039(00)00850-9

Tetrahedron Letters 41 (2000) 5783±5787
Optimization of variables for screening solid-supported metal
complexes as oxidation catalysts
Amarnath Natarajan and Jose S. Madalengoitia*
Department of Chemistry, University of Vermont, Burlington, VT 05405, USA
Received 24 March 2000; accepted 25 May 2000
Abstract
Evaluation of polymer-supported metal complexes as oxidation catalysts revealed that product distribution
is not dependent on metal stoichiometry, or resin but is dependent on the oxidant, oxidant stoichiometry
and the metal incubation times. The model systems investigated under the optimized conditions a€orded
highly reproducible results, indicating that this methodology can be used to reliably screen libraries of
metal complexes. # 2000 Elsevier Science Ltd. All rights reserved.
Keywords: supported reagents/reactions.
Since the introduction of combinatorial chemistry as a tool for the generation and identi®ca-
tion of lead compounds, the use of molecular diversity techniques in the discovery and optimization
of novel catalysts has held much interest. Accordingly, a number of reports on combinatorial
catalysis have appeared in the literature.1 5 We were particularly intrigued in the synthesis of
peptide libraries as ligands for redox active metals. In discovering new systems for catalysis using
combinatorial chemistry we envisioned a two-step process: (1) optimization of conditions to
establish a highly reproducible screening method; and (2) library evaluation using the established
methodology. In this ®rst paper we address the ®rst step: how experimental variables a€ect the
reliability of an `on bead assay'. Speci®cally, we explore the e€ects of metal incubation time,
metal/ligand stoichiometry, and substrate/oxidant stoichiometry on oxidation reactions using
four model ligands 1, 2, 3 and 4 (Fig. 1).
Ligands 1, 2, 3 and 4 were synthesized on aminomethyl polystyrene (AMP) resin using Fmoc
chemistry and standard peptide coupling conditions.6 10 The ®rst objective was to identify an active
catalyst system which could then be used to optimize reaction conditions. To this end, a small
.
library was generated by incubating ligands 1 and 2 with di€erent metal salts (Fe(ClO4)3 6H2O,
.
.
.
.
RuCl3 H2O, Ni(OAc)2 4H2O, CoCl2 6H2O, Mn(OAc)2 and Mn(OAc)3 6H2O) in degassed DMF
under N2 for 1 h. The resin-bound metal complexes were then assayed for their ability to catalyze
* Corresponding author.
0040-4039/00/$ - see front matter # 2000 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(00)00850-9

5784
Figure 1. Ligand structures synthesized on polystyrene support (0.5 mmol/g loading, 100±200 mesh particle size)
the oxidation of cis-stilbene (300 equiv.) with t-BHP (100 equiv.) in degassed CH2Cl2 at rt for 5 h.
Results from these experiments are summarized in Fig. 2. The turnover number for these com-
plexes was not particularly impressive, however, the study indicated that the Ru complex with
ligands 1 and 2 a€orded signi®cantly higher amounts of cis-epoxide than either the other metal
complexes or the control (PhIO/cis-stilbene). Although this study identi®ed rutheniun as a
potentially active metal for these systems we did not pursue the use of t-BHP as an oxidant
because the observed gas evolution (most likely O2) and the formation of trans-stilbeneoxide
suggested radical processes.
.
Figure 2. Evaluation of ligands 1 and 2 with metal salts [Fe(ClO4)3 6H2O, RuCl3 H2O, Ni(OAc)2 4H2O, CoCl2 6H2O,
.
.
.
.
Mn(OAc)2 and Mn(OAc)3 6H2O]
Identi®cation of ruthenium as an active metal allowed us to optimize the time dependence of
.
metal complexation (Fig. 3). In this series of experiments, ligand 3 was incubated with RuCl3 H2O
in degassed DMF under N2 over di€erent times. The resin bound material was then assayed for
its ability to catalyze the oxidation of cis-stillbene with iodosobenzene (oxidant) in degassed
CH2Cl2 at rt for 48 h. The results from these experiments are summarized in Fig. 2. This study
showed that cis-stilbene was not completely consumed with catalyst generated from less than a
24 h metal incubation. Likewise, the amounts of cis-stilbeneoxide and benzaldehyde products

5785
increases and plateaus between 24 and 72 h. Interestingly, the amount of trans-stilbeneoxide
formation goes from ꢀ19% to background level when the metal incubation time is increased
from 1 h to 72 h. These data suggest that with a 1 h incubation not only is there little metal
complexation but also the complex(es) formed have not fully equilibrated and might represent
weakly bound metal. Clearly Ru(III) complexation will be slower than other metals because
Ru(III) has a slow ligand exchange rate; however, these results indicate that longer incubation
times should minimize missing potentially active complexes.
Figure 3. Evaluation of ligand 3 at di€erent metal complexation times. A=recovered cis-stillbene; B=cis-stillbene-
oxide; C=trans-stilbeneoxide; D=benzaldehyde
We then studied if the metal to ligand stoichiometry has an e€ect on the product distribution.
This could arise if, for example, di€erent complexes (with di€erential reactivity) were formed
from alternate metal/ligand (M/L) ratios. The ligands were then incubated with 1, 0.5 and 1.5
.
equiv. of RuCl3 H2O in degassed DMF under N2 for 3 days. The complexes were then assayed
for their ability to catalyze the oxidation of cis-stillbene with PhIO in degassed CH2Cl2 at rt for
48 h. Table 1 summarizes the results from these experiments. Entries 1±3 (Table 1), in which
resin-bound ligand is incubated with variable amounts of Ru, exhibit essentially the same product
distribution (ꢀ2:1 oxidative cleavage/epoxidation). These data are indicative that irrespective of
the M/L stoichiometry, a unique ruthenium complex is formed which gives rise to a unique pro-
duct distribution. This information illustrates that the results of a high throughput screen will not
be subject to small errors in ligand weighting or metal delivery.
We next turned our attention to the e€ects of the oxidant/substrate ratio on product distribu-
tion. Entries 4±6 (Table 1) show the results from experiments in which the quantity of iodoso-
benzene is varied. Interestingly, the product distribution is sensitive to the amount of PhIO. At a
lower loading of PhIO, epoxide formation is favored (entry 4) while at a higher loading, oxidative
cleavage is favored (entry 6). The substrate is completely consumed at a 4:1 oxidant/substrate
ratio. These results indicate that to obtain the most consistent results in a high throughput screen,
the catalyst evaluation is best performed at a high PhIO loading. We next investigated if the
reaction performed at a low PhIO loading could be driven to completion by addition of PhIO
aliquots at 12 h intervals. Entry 7 now shows that indeed the reaction may be driven to completion
by addition of PhIO portions and that epoxide formation is favored by the addition of PhIO in
small amounts (60%). The isolated yield (entry 8) is consistent with the product distribution
obtained by NMR and validates the observed results. Evaluation of the Ru-complex formed with
ligand 4 reveals that this complex now strongly favors oxidative cleavage a€ording an 87% yield

5786
Table 1
Evaluation of ruthenium complexes as heterogeneous catalysts
of benzaldehyde and a 9% yield of epoxide. Since another complex exhibits a di€erent product
pro®le the oxidation is not due to some non-speci®c e€ect.
We next investigated di€erent solid supports. Ligand 3 was synthesized on four di€erent resins
(ClTrt, Wang, AMP, MBHA). Metal complexation and complex evaluation was carried out with
the optimized conditions. Results from these studies are summarized in Table 2. Examination of
the data indicates that the epoxide yield is essentially not a€ected by the steric crowding around
the linking site. Furthermore, upon treatment of the AMP resin±RuCl3 complex (Entry 5) 1%
TFA/CH2Cl2 (ClTrt resin cleavage conditions) the catalytic activity of the complex remained
una€ected. This experiment shows that after identi®cation of an active metal complex in a high
throughput screen, the complex may be synthesized on ClTrt resin and liberated under conditions
that do not a€ect the catalytic activity of the complex.
Table 2
E€ect of solid support on product distribution
In conclusion, we have investigated the potential sources of error in evaluating peptide-derived
metal complexes as oxidation catalysts. Our results indicate that at least in the systems investigated,
catalyst evaluation is not dependent on metal stoichiometry, or resin but is dependent on the
oxidant, oxidant stoichiometry and metal incubation time. Finally, it should be noted that our

5787
results are reproducible within þ5% of the values reported, thus indicating that this method of
catalyst evaluation should not miss potentially active catalysts or give rise to false-positives. The
following paper describes the results of evaluating a small combinatorial library using the methods
described.
General procedure for the synthesis and evaluation of the Ru complexes: Resin bound ligand (0.1
g, 0.05 mL) was placed in a reaction vial equipped with a fritted stopcock. Resin swelling (DMF)
.
was followed addition of RuCl3 H2O (0.02 M in DMF) and agitation under N2 for 48 h. The
polymer-supported Ru complexes were isolated via ®ltration. The resulting resin was rinsed with
9:1 DMF:CH2Cl2 (3Â5 min). The ®nal rinsing solutions were colorless. Evaluation: resin-bound
complex (10 mg, 5 mmol), PhIO (0.2 mmol), and cis-stilbene (1 mL 0.05 M in CH2Cl2) in CH2Cl2
were stirred for 48 h. An aliquot of the resulting solution was evaporated by blowing N2 over the
sample. Conversion and yield were determined by NMR.
Acknowledgements
This work was supported by the American Chemical Society Petroleum Research fund PRF
5-26407.
References
1. Francis, M. B.; Jacobsen, E. N. Angew. Chem., Int. Ed. Engl. 1999, 38, 937.
2. Sigman, M. S.; Jacobsen, E. N. J. Am. Chem. Soc. 1998, 120, 4901.
3. Shimizu, K. D.; Cole, B. M.; Krueger, C. A.; Kuntz, K. W.; Snapper, M. L.; Hoveyda, A. H. Angew. Chem., Int.
Ed. Engl. 1997, 36, 1704.
4. Gilbertson, S. R.; Wang, X. Tetrahedron Lett. 1996, 37, 6475.
5. Porte, A. M.; Reibenspies, J.; Burgess, K. J. Am. Chem. Soc. 1998, 120, 9180.
6. Corey, E. J.; Borror, A. L.; Foglia, T. J. Org. Chem. 1965, 30, 288.
7. Huang, L.; Quada Jr., J. C.; Lown, J. W. Bioorg. Med. Chem. 1995, 6, 647.
8. Huang, L.; Quada Jr., J. C.; Lown, J. W. Bioconjugate Chem. 1995, 6, 21.
9. Thompson, P. E.; Keah, H. H.; Gomme, P. T.; Stanton P. G.; Hearn, M. T. W. Int. J. Pept. Protein Res. 1995, 46,
174.
10. Robles, J.; Pedroso, E.; Grandas, A. J. Org. Chem. 1994, 59, 2482.