- Email: [email protected]
A new type of anchored homogeneous catalyst
Studies in Surface Science and Catalysis 130 A. Corma, F.V. Melo, S. Mendioroz and J.L.G. Fierro (Editors) 9 2000 Elsevier Science B.V. All rights reserved.
3351
A New Type of Anchored Homogeneous Catalyst Robert L. Augustine, Setrak K. Tanielyan, Steven Anderson, Hong Yang and Yujing Gao Center for Applied Catalysis, Seton Hall University South Orange, NJ 07079, USA Almost all previous attempts to design a "heterogenized homogeneous catalyst" involved first attaching the ligand to a solid support material and then preparing the heterogeneous catalyst using this solid ligand. In contrast, we have succeeded in attaching preformed homogeneous complexes to a solid support using a heteropoly acid as the anchoring agent. In this way, the heteropoly acid is first attached to the support material and then interacts with a homogeneous catalyst through the metal atom of the complex.We have now used such anchored rhodium and ruthenium complexes for a number of chiral and achiral hydrogenations. We have established that the amount of metal lost in these reactions is less than 1 ppm. Details of these reactions will be presented along with the data used to determine the nature of the interaction between the complex and the heteropoly acid. 1. INTRODUCTION Interest in enantioselective synthesis has been increasing steadily over the past several years particularly because of the need for developing chiral drugs and agrochemicals. One of the most effective means of attaining such syntheses is to use an enantioselective catalyst and, thus, transfer the chirality of the catalyst to many thousands of product molecules. Almost all of the currently effective enantioselective catalysts presently used in organic synthesis are chiral, metal complexes which are soluble in the reaction medium. The most serious problems associated with the use of homogeneous catalysts lie in the difficulties encountered in separating the catalyst, both the metal and the accompanying ligands, from the product. Over the past twenty-five years or so a number of attempts have been made to 'heterogenize' the more versatile homogeneous catalysts. Reacting organometallic complexes with solids such as polymers or metal oxides which had been modified by the addition of phosphine or amine ligands to the surface of the support is the most general method for trying to accomplish the 'heterogenization'. Such attempts have been made for quite some time with varying degrees of success.[1-5] These 'heterogenized' catalysts are not widely used since their activity is generally lower than that of the corresponding homogeneous analog. Further, it has been reported that activity is frequently lost on attempted re-use. This approach, which requires the incorporation of the ligand onto a solid material, is particularly difficult to apply with most of the chiral ligands used in the highly efficient enantioselective homogeneous catalysts in use today. The ability to attach a preformed homogeneous catalyst to a support would have a distinct advantage. We have accomplished this by developing a method for "anchoring" preformed chiral homogeneous catalysts to a variety of support materials. In many instances the activity and selectivity observed with the supported species was greater than those found using the
3352
L..
L ........ M e ...........
9=i~L.....
Fig. 1. A cartoon showing how a homogeneous catalyst can be anchored to a support material using a heteropoly acid as the tethering agent. corresponding homogeneous catalysts. On re-use, the supported species usually exhibited an increase in activity and selectivity which was maintained on re-use in up to fifteen cycles. Catalyst leaching was not detected. A large number of different complexes have been anchored to various supports and used to catalyze a variety of reactions.[6]
2. NATURE OF THE CATALYST These catalysts are composed of three components; a soluble catalytically active organometallic complex, a support material and a heteropoly acid which is use to tether the complex to the support. Fig. 1 depicts how we perceive this arrangement. It has been reported that Rh and Ir complexes react with a heteropoly acid to give a material to which the metal atom is attached to the heteropoly acid through the oxygen atoms on its surface.[7] Attachment of the heteropoly acid to a support such as alumina takes place by interaction of the hydroxy groups of the acid with the support.[8] The 31p MAS nmr chemical shifts for the phosphorous atom in phosphotungstic acid (PTA), PTA on alumina and the Rh(DiPamp) supported on PTA/A120 3 both before and after use in a hydrogenation were all identical showing that the PTA remained intact throughout the entire preparation and reaction sequence.[9]
3. REACTION DATA
H2CyCO2Me H21 Atm.CatalystRH3C@ .T."-~ cO2Me +
H3C,~COzMe
NAc
NAc
NAc
1
(R)
(S)
(Eqn. 1)
We have used as the general probe reaction the enantioselective hydrogenation of methyl 2-acetamidoacrylate (1) (Eqn. 1). Fig. 2 shows the hydrogen uptake curves for the reaction run over Rh(DiPAMP) tethered to a PTA modified Montmorillonite K clay. In its first use, this catalyst was less active and selective than a homogeneous solution of Rh(DiPAMP). However, when this catalyst was re-used for this reaction it became more active and selective,
3353
t 4th Use ee=93% 0.5 _~ 9 3rd Use r ~ " ee=90% 0.4 I- ~' 2nd Use | ~ ' ee=93% 0.3 0.6
0
"10
r L_ 0 .Q
"10
l::
,/,, I
9
I
I
0.2
Soluble Catalyst 9 9 . ee=76% 9 9 1st Use 9 ee=67%
0.1 0.0 r r 0
9
i 60
i 120
i 180
i 240
i 300
i 360
J 420
Time, min
Fig. 2. Hydrogen uptake curves and product ee's for the hydrogenation of I over Rh(DiPAMP) supported on PTA modified Montmorillonite K. The catalyst was freshly prepared and re-used four times. retaining this activity and selectivity for up to fifteen cycles as shown by the data listed in Table 1. Analysis of the combined reaction mixtures from the last ten cycles showed that the rhodium loss in these reactions amounted to less than 1 ppm. This increase in activity and selectivity after what appears to be an induction period presents an interesting problem. While we have no clear rational for these observations, the 31p MAS nmr taken of a catalyst sample isolated after the first use, dried and kept in an inert
T a b l e 1. R e a c t i o n rate and p r o d u c t ee data f r o m the multiple h y d r o g e n a t i o n s of 4 over a Rh(DiPamp)/PTA/Montmorillonite catalyst, a
Use Number
Rate b
Product ee (%)
Homogeneous
0.25
76
1
0.18
67
2
1.20
92
3
1.26
94
6
1.49
96
9
1.29
97
15
c
97
a Hydrogenation run at 25~ under 1 atm H 2 using 20 ~tmoles of supported Rh(DiPamp) to saturate 0.8 mmole of I for each run. b moles of H 2 / mole Rh / min c Rate data for this run were lost due to a computer malfunction.
3354
Table 2. Reaction rate and product ee data obtained on hydrogenation of 1
over Rh(DiPAMP) tethered to PTA modified supports. Support
Use Number
Rate a
ee%
Montmorillonite K Clay
1 3 1 3 1 3 1 3
0.18 0.56 0.07 0.40 0.32 1.67 0.38 0.44
76 91 83 90 90 95 91 92
Carbon Alumina Lanthana a moles H 2 / mole Rh / min
atmosphere showed a small splitting of the PTA peak may be the result of a change in oxidation state of some of the tungsten ions surrounding the central phosphorous in the PTA. The fact that in some reactions the catalyst changed from yellow to a light gray-violet on use and back to yellow on exposure to air further suggests that this activation may be the result of some partial reduction of the tungsten in the PTA. Further work is planned to clarify this matter. One of the advantages of this approach to anchoring homogeneous catalysts is its apparent generality in that this procedure can be used to anchor a variety of pre-formed active homogeneous catalysts onto a number of different supports. Table 2 shows the reaction rates and product ee's obtained on hydrogenation of I over Rh(DiPAMP) on several PTA modified supports. In all cases except one, the activity and selectivity increase on re-use of the catalyst. With the basic lanthana support, however, activity and selectivity remained ;essentially constant. In Table 3 are listed the reaction rate and product ee data observed for the hydrogenation of 1 over several of Rh complexes anchored to PTA modified alumina along with the corresponding homogeneous catalyst data.
Table 3. Hydrogenation of 1 over alumina supported Rh complexes containing different chiral ligands. Ligand
Use #
Heterogeneous
Homogeneous
Rate a
ee
Rate a
ee
DiPAMP
1 3
0.32 1.67
90% 92%
0.25 na
76% na
PROPHOS
1 3 1 3 1 3
2.0 2.6 1.8 4.4 3.75 8.15
68% 63% 83% 95% 21 87
0.26 na 3.3 na 7.4 na
66% na 96% na 84 na
MeDUPHOS BPPM
a moles H2/mole metal/rain
3355
Scheme 1 Heteropoly
Supl
- +--..p..._j
We have also used these anchored homogeneous catalysts for the hydrogenation of a number of other prochiral substrates such as methyl 2-acetamidocinnamate, 2-actamidocinnamic acid, dimethyl itaconate and 2-methyl-2-hexenoic acid. A Ru(Binap) complex anchored to PTA treated alumina was used for successive hydrogenations of 1 and dimethyl itaconate with the reaction rate and product ee increasing after the first use in both cases. Analysis of the product mixtures from these reactions found no detectable ruthenium present. In every instance, the reaction rates and product ee were at least equal to and frequently better than the results obtained using the corresponding homogeneous catalyst especially when the catalysts were re-used. In some instances, even the first use of the catalyst gave superior results. Heteropoly acids other than PTA were also successful anchoring agents. This supported catalyst was also found to have a reasonable shelf life as shown by the fact that a sample of the catalyst which had been stored in a vial in air for eight months had the same activity and selectivity as that of a freshly prepared catalyst. This anchoring procedure has also been applied to achiral complexes with similar results. For instance, hydrogenation of 1-hexene over a Wilkinson's catalyst anchored to PTA treated alumina proceeded 2-3 times faster than the corresponding homogeneously catalyzed reaction, even on the first use of the heterogenized material. A Rh(dppb) complex anchored to a PTA modified alumina was used for several successive hydrogenations of 1-hexene with a combined substrate/catalyst ratio of about 8000/1. An analysis of the product mixtures from these reactions found no detectable rhodium present. While the general procedure for tethering a complex is to stir a solution of the complex in a slurry of a modified support, we recognized that there would be times when it would be more desirable to add a ligand to a tethered complex precursor. As depicted in Scheme 1 a tethered Rh(COD)2 can be prepared first and then treated with the ligand of choice before initiating the reaction. Table 4 lists the comparative reaction data for tethered Rh(DiPAMP) catalysts prepared by either the anchoring of the preformed catalyst or by reaction the tethered Rh(COD)2 precursor with the DiPAMP ligand. Similar results have been obtained using an
3356
Table 4. Reaction rates and product ee's obtained on hydrogenation of 1 over Rh(DiPAMP) catalysts prepared both by reaction with a supported precursor or by direct reaction with PTA modified alumina. Prepared from Precursor
Prepared from Complex
Use Number
Rate a
ee %
Rate a
ee %
1
0.284
79
0.389
84
2
0.478
85
0.487
87
3 4 5
0.491 0.579 0.535
87 87 85
0.538
87
a moles H 2 / mole Rh / min anchored Ru(p-cymene) precursor material to prepare a number of different anchored ruthenium complexes. 4. A C K N O W L E D G E M E N T S
The development of these catalysts was funded by grants CTS-9312533 and CTS-9708227 from the US National Science Foundation. 5. R E F E R E N C E S 1.
2. 3. 4. 5. 6. 7. 8.
J.P. Collman, L. S. Hegedus, M. P. Cooke, J. R. Norton, G. Dolcetti, D. N. Marquardt, J. Am. Chem. Soc., 94 (1972) 1789. W. Dumont, J.-C. Poulin, T. P. Daud and H. B. Kagan, J. Am. Chem. Soc., 95 (1973) 8295. L. L. Murrell, "Advanced Materials in Catalysis", Academic Press, New York, 1977, Chapter 8. V. Isaeva, A. Derouault and J. Barrault, Bull. Soc. Chim., Fr., 133 (1996) 351. U. Nagel and J. Leipold, Chem. Ber., 129 (1996) 815 S.K. Tanielyan and R.L. Augustine, U.S. Pat. Appl., 08/994,025; PCT Int. Appl., WO9828074; Chem. Abstr., 129 (1998) 109217. M. Pohl, D. K. Lyon, N. Mizuno, K. Nomiya, R. G. Finke, Inorg. Chem., 34 (1995) 1413. Y. Izumi, R. Hasere, K. Urabi, J. Catal. 84 (1983) 402
9. The MAS nmr spectra were obtained by Xiaolin Yang, Engelhard Corp., Iselin, NJ, USA.
3351
A New Type of Anchored Homogeneous Catalyst Robert L. Augustine, Setrak K. Tanielyan, Steven Anderson, Hong Yang and Yujing Gao Center for Applied Catalysis, Seton Hall University South Orange, NJ 07079, USA Almost all previous attempts to design a "heterogenized homogeneous catalyst" involved first attaching the ligand to a solid support material and then preparing the heterogeneous catalyst using this solid ligand. In contrast, we have succeeded in attaching preformed homogeneous complexes to a solid support using a heteropoly acid as the anchoring agent. In this way, the heteropoly acid is first attached to the support material and then interacts with a homogeneous catalyst through the metal atom of the complex.We have now used such anchored rhodium and ruthenium complexes for a number of chiral and achiral hydrogenations. We have established that the amount of metal lost in these reactions is less than 1 ppm. Details of these reactions will be presented along with the data used to determine the nature of the interaction between the complex and the heteropoly acid. 1. INTRODUCTION Interest in enantioselective synthesis has been increasing steadily over the past several years particularly because of the need for developing chiral drugs and agrochemicals. One of the most effective means of attaining such syntheses is to use an enantioselective catalyst and, thus, transfer the chirality of the catalyst to many thousands of product molecules. Almost all of the currently effective enantioselective catalysts presently used in organic synthesis are chiral, metal complexes which are soluble in the reaction medium. The most serious problems associated with the use of homogeneous catalysts lie in the difficulties encountered in separating the catalyst, both the metal and the accompanying ligands, from the product. Over the past twenty-five years or so a number of attempts have been made to 'heterogenize' the more versatile homogeneous catalysts. Reacting organometallic complexes with solids such as polymers or metal oxides which had been modified by the addition of phosphine or amine ligands to the surface of the support is the most general method for trying to accomplish the 'heterogenization'. Such attempts have been made for quite some time with varying degrees of success.[1-5] These 'heterogenized' catalysts are not widely used since their activity is generally lower than that of the corresponding homogeneous analog. Further, it has been reported that activity is frequently lost on attempted re-use. This approach, which requires the incorporation of the ligand onto a solid material, is particularly difficult to apply with most of the chiral ligands used in the highly efficient enantioselective homogeneous catalysts in use today. The ability to attach a preformed homogeneous catalyst to a support would have a distinct advantage. We have accomplished this by developing a method for "anchoring" preformed chiral homogeneous catalysts to a variety of support materials. In many instances the activity and selectivity observed with the supported species was greater than those found using the
3352
L..
L ........ M e ...........
9=i~L.....
Fig. 1. A cartoon showing how a homogeneous catalyst can be anchored to a support material using a heteropoly acid as the tethering agent. corresponding homogeneous catalysts. On re-use, the supported species usually exhibited an increase in activity and selectivity which was maintained on re-use in up to fifteen cycles. Catalyst leaching was not detected. A large number of different complexes have been anchored to various supports and used to catalyze a variety of reactions.[6]
2. NATURE OF THE CATALYST These catalysts are composed of three components; a soluble catalytically active organometallic complex, a support material and a heteropoly acid which is use to tether the complex to the support. Fig. 1 depicts how we perceive this arrangement. It has been reported that Rh and Ir complexes react with a heteropoly acid to give a material to which the metal atom is attached to the heteropoly acid through the oxygen atoms on its surface.[7] Attachment of the heteropoly acid to a support such as alumina takes place by interaction of the hydroxy groups of the acid with the support.[8] The 31p MAS nmr chemical shifts for the phosphorous atom in phosphotungstic acid (PTA), PTA on alumina and the Rh(DiPamp) supported on PTA/A120 3 both before and after use in a hydrogenation were all identical showing that the PTA remained intact throughout the entire preparation and reaction sequence.[9]
3. REACTION DATA
H2CyCO2Me H21 Atm.CatalystRH3C@ .T."-~ cO2Me +
H3C,~COzMe
NAc
NAc
NAc
1
(R)
(S)
(Eqn. 1)
We have used as the general probe reaction the enantioselective hydrogenation of methyl 2-acetamidoacrylate (1) (Eqn. 1). Fig. 2 shows the hydrogen uptake curves for the reaction run over Rh(DiPAMP) tethered to a PTA modified Montmorillonite K clay. In its first use, this catalyst was less active and selective than a homogeneous solution of Rh(DiPAMP). However, when this catalyst was re-used for this reaction it became more active and selective,
3353
t 4th Use ee=93% 0.5 _~ 9 3rd Use r ~ " ee=90% 0.4 I- ~' 2nd Use | ~ ' ee=93% 0.3 0.6
0
"10
r L_ 0 .Q
"10
l::
,/,, I
9
I
I
0.2
Soluble Catalyst 9 9 . ee=76% 9 9 1st Use 9 ee=67%
0.1 0.0 r r 0
9
i 60
i 120
i 180
i 240
i 300
i 360
J 420
Time, min
Fig. 2. Hydrogen uptake curves and product ee's for the hydrogenation of I over Rh(DiPAMP) supported on PTA modified Montmorillonite K. The catalyst was freshly prepared and re-used four times. retaining this activity and selectivity for up to fifteen cycles as shown by the data listed in Table 1. Analysis of the combined reaction mixtures from the last ten cycles showed that the rhodium loss in these reactions amounted to less than 1 ppm. This increase in activity and selectivity after what appears to be an induction period presents an interesting problem. While we have no clear rational for these observations, the 31p MAS nmr taken of a catalyst sample isolated after the first use, dried and kept in an inert
T a b l e 1. R e a c t i o n rate and p r o d u c t ee data f r o m the multiple h y d r o g e n a t i o n s of 4 over a Rh(DiPamp)/PTA/Montmorillonite catalyst, a
Use Number
Rate b
Product ee (%)
Homogeneous
0.25
76
1
0.18
67
2
1.20
92
3
1.26
94
6
1.49
96
9
1.29
97
15
c
97
a Hydrogenation run at 25~ under 1 atm H 2 using 20 ~tmoles of supported Rh(DiPamp) to saturate 0.8 mmole of I for each run. b moles of H 2 / mole Rh / min c Rate data for this run were lost due to a computer malfunction.
3354
Table 2. Reaction rate and product ee data obtained on hydrogenation of 1
over Rh(DiPAMP) tethered to PTA modified supports. Support
Use Number
Rate a
ee%
Montmorillonite K Clay
1 3 1 3 1 3 1 3
0.18 0.56 0.07 0.40 0.32 1.67 0.38 0.44
76 91 83 90 90 95 91 92
Carbon Alumina Lanthana a moles H 2 / mole Rh / min
atmosphere showed a small splitting of the PTA peak may be the result of a change in oxidation state of some of the tungsten ions surrounding the central phosphorous in the PTA. The fact that in some reactions the catalyst changed from yellow to a light gray-violet on use and back to yellow on exposure to air further suggests that this activation may be the result of some partial reduction of the tungsten in the PTA. Further work is planned to clarify this matter. One of the advantages of this approach to anchoring homogeneous catalysts is its apparent generality in that this procedure can be used to anchor a variety of pre-formed active homogeneous catalysts onto a number of different supports. Table 2 shows the reaction rates and product ee's obtained on hydrogenation of I over Rh(DiPAMP) on several PTA modified supports. In all cases except one, the activity and selectivity increase on re-use of the catalyst. With the basic lanthana support, however, activity and selectivity remained ;essentially constant. In Table 3 are listed the reaction rate and product ee data observed for the hydrogenation of 1 over several of Rh complexes anchored to PTA modified alumina along with the corresponding homogeneous catalyst data.
Table 3. Hydrogenation of 1 over alumina supported Rh complexes containing different chiral ligands. Ligand
Use #
Heterogeneous
Homogeneous
Rate a
ee
Rate a
ee
DiPAMP
1 3
0.32 1.67
90% 92%
0.25 na
76% na
PROPHOS
1 3 1 3 1 3
2.0 2.6 1.8 4.4 3.75 8.15
68% 63% 83% 95% 21 87
0.26 na 3.3 na 7.4 na
66% na 96% na 84 na
MeDUPHOS BPPM
a moles H2/mole metal/rain
3355
Scheme 1 Heteropoly
Supl
- +--..p..._j
We have also used these anchored homogeneous catalysts for the hydrogenation of a number of other prochiral substrates such as methyl 2-acetamidocinnamate, 2-actamidocinnamic acid, dimethyl itaconate and 2-methyl-2-hexenoic acid. A Ru(Binap) complex anchored to PTA treated alumina was used for successive hydrogenations of 1 and dimethyl itaconate with the reaction rate and product ee increasing after the first use in both cases. Analysis of the product mixtures from these reactions found no detectable ruthenium present. In every instance, the reaction rates and product ee were at least equal to and frequently better than the results obtained using the corresponding homogeneous catalyst especially when the catalysts were re-used. In some instances, even the first use of the catalyst gave superior results. Heteropoly acids other than PTA were also successful anchoring agents. This supported catalyst was also found to have a reasonable shelf life as shown by the fact that a sample of the catalyst which had been stored in a vial in air for eight months had the same activity and selectivity as that of a freshly prepared catalyst. This anchoring procedure has also been applied to achiral complexes with similar results. For instance, hydrogenation of 1-hexene over a Wilkinson's catalyst anchored to PTA treated alumina proceeded 2-3 times faster than the corresponding homogeneously catalyzed reaction, even on the first use of the heterogenized material. A Rh(dppb) complex anchored to a PTA modified alumina was used for several successive hydrogenations of 1-hexene with a combined substrate/catalyst ratio of about 8000/1. An analysis of the product mixtures from these reactions found no detectable rhodium present. While the general procedure for tethering a complex is to stir a solution of the complex in a slurry of a modified support, we recognized that there would be times when it would be more desirable to add a ligand to a tethered complex precursor. As depicted in Scheme 1 a tethered Rh(COD)2 can be prepared first and then treated with the ligand of choice before initiating the reaction. Table 4 lists the comparative reaction data for tethered Rh(DiPAMP) catalysts prepared by either the anchoring of the preformed catalyst or by reaction the tethered Rh(COD)2 precursor with the DiPAMP ligand. Similar results have been obtained using an
3356
Table 4. Reaction rates and product ee's obtained on hydrogenation of 1 over Rh(DiPAMP) catalysts prepared both by reaction with a supported precursor or by direct reaction with PTA modified alumina. Prepared from Precursor
Prepared from Complex
Use Number
Rate a
ee %
Rate a
ee %
1
0.284
79
0.389
84
2
0.478
85
0.487
87
3 4 5
0.491 0.579 0.535
87 87 85
0.538
87
a moles H 2 / mole Rh / min anchored Ru(p-cymene) precursor material to prepare a number of different anchored ruthenium complexes. 4. A C K N O W L E D G E M E N T S
The development of these catalysts was funded by grants CTS-9312533 and CTS-9708227 from the US National Science Foundation. 5. R E F E R E N C E S 1.
2. 3. 4. 5. 6. 7. 8.
J.P. Collman, L. S. Hegedus, M. P. Cooke, J. R. Norton, G. Dolcetti, D. N. Marquardt, J. Am. Chem. Soc., 94 (1972) 1789. W. Dumont, J.-C. Poulin, T. P. Daud and H. B. Kagan, J. Am. Chem. Soc., 95 (1973) 8295. L. L. Murrell, "Advanced Materials in Catalysis", Academic Press, New York, 1977, Chapter 8. V. Isaeva, A. Derouault and J. Barrault, Bull. Soc. Chim., Fr., 133 (1996) 351. U. Nagel and J. Leipold, Chem. Ber., 129 (1996) 815 S.K. Tanielyan and R.L. Augustine, U.S. Pat. Appl., 08/994,025; PCT Int. Appl., WO9828074; Chem. Abstr., 129 (1998) 109217. M. Pohl, D. K. Lyon, N. Mizuno, K. Nomiya, R. G. Finke, Inorg. Chem., 34 (1995) 1413. Y. Izumi, R. Hasere, K. Urabi, J. Catal. 84 (1983) 402
9. The MAS nmr spectra were obtained by Xiaolin Yang, Engelhard Corp., Iselin, NJ, USA.