Asymmetric Hydrogenation of PZQ-enamide

A possible solution to the large-scale preparation of enantiopure PZQ is the approach shown below, originally suggested by Craig Williams.
 

 
First step:
We have a decent approach to this involving heating in a sulfur melt. This needs improvement, but we can generate gram quantities of the intermediate PZQ-enamide easily.
 
Second step:
This is the key. A number of different catalysts for this reaction have been tried [data coming soon, sorry] but no conversion has been observed. Reduction with rhodium on carbon quantitatively generates rac PZQ, so the enamide is reactive. The geometry of the enamide here is awkward in that the exocyclic carbonyl is not able to direct a metal-based asymmetric catalyst to the double bond, as in the classical model. It's a challenging reduction.
 
This second part of the synthetic route needs people who have expertise in screening hydrogenation catalysts, or people who have in their posession unusual catalysts for asymmetric hydrogenation that might be appropriate for this kind of reaction. The Todd group can mail people samples of the PZQ-enamide for this screening (plus samples of PZQ for the assay). Most any column will separate PZQ enantiomers. If you are interested in helping with this part of the project, please reply below.
 
Relevant groups/papers to consider (please fee free to suggest others here by editing the page):
Bernhard Breit
 
August 2010: We received our first offer of help with the asymmetric hydrogenation. Laurent Lefort from DSM in the Netherlands, has offered to include our substrate in future catalyst screening that they are undertaking. This is a great offer. Laurent has a great deal of relevant experience in this area. Aug 31st: The Todd lab has mailed a sample of the enamide plus a reference sample of PZQ to DSM. October 20: Second set of DSM results posted here. January 25 2011: Third, larger screen from DSM posted in the same place as the others.

Characterization of PZQ-enamide

The PZQ-enamide is shown below. This is an intermediate in the stereoablative route to enantiopure PZQ.

HPLC trace for PZQ-enamide (using ChiralcelOD-H, solvents: Hex:IPA:TEA 60:40:0.1, Flow Rate: 0.7 mL/min) gives retention time: 15.772 mins. Original HPLC trace attached below.
 
(Note comparison HPLC trace for PZQ itself is here.)
 
Other data: RF =0.35(1:1 Hexane: EtOAc); m.p. 131-134 ºC; 1H NMR (300 MHz, CDCl3): δ 1.15-1.85 (10H, m, cy), 2.35-2.65 (1H, m, cy ), 2.84 (t, 2H, J 5.2, H7), 3.82 (t, 2H, J 5.5, H6), 4.35 (1H, s, H4), 6.70 (1H, s, H2), 7.10-7.51 (4H, m, Ar); (Proton NMR spectrum attached below) 13C NMR (75 MHz, CDCl3): δ 26.1, 29.4, 38.8, 41.4, 45.9, 48.7, 106.0, 106.0, 123.1, 123.6, 126.9, 127.7, 128.2, 128.4, 128.6, 128.8, 129.0, 134.5, 164.3, 174.5; IR (CHCl3): 2952-2870 cm-1, 1652 cm-1, 1463 cm-1; MS (ESI) m/z: 311.1 [(MH)+, 100%], 333.3 [(MNa)+, 55%] HRMS (ESI) Calcd. for C19H22N2NaO2 (MNa+): 333.15735. Found: 333.15765.

Failed Attempts at Asymmetric Hydrogenation of PZQ-enamide

AttachmentSize
DSM 48 vial screen.pdf1.26 MB

This page will contain all examples of failed reactions in the attempted asymmetric hydrogenation of the PZQ enamide.

 

We are very grateful to Sigma-Aldrich for an initial donation of the catalysts employed below. Please note that we do not hold a library of other hydrogenation catalysts. Any suggestions for alternative catalysts are welcome, but in the interests of speed, the best suggestions here are catalysts + who could run the reactions (other than the Todd group) and to whom we could send PZQ enamide. This will accelerate the research.

 

In the Todd lab, we used a bespoke hydrogenator vessel of 15 mL capacity able to take pressures of 2000 psi and where the addition of reagents can be performed under Ar/H2. To verify that this set-up was able to reproduce literature results, we carried out the following asymmetric hydrogenation:

 

 

See a review of phospholane ligands for asymmetric catalysis for more, as well as the original paper [need ref] that the above reaction is taken from. While the results don't exactly match, the experimental set-up is clearly adequate for carrying out asymmetric hydrogenation reactions.

 

Under the same conditions as above, using the PZQ enamide, no reaction was observed, even after heating at 50 oC for 24 hours or increasing the pressure to 600 psi. Starting material is quantitatively recovered in each case.

 

Other conditions tried (no conversion in any case. Unless otherwise stated, the reactions were carried out in 5 mL solvent on 100 mg PZQ enamide) using the same conditions as for the test reaction above.

2 mol% (R)-RuCl[(p-cymene)(BINAP)]Cl.

(Formed in situ from 5 mol% Ru2(C6H6)2Cl4 and 10 mol% ligand)

PipPHOS (2 mol%) and Rh(COD)2BF4 (4 mol%)

[Details coming]

 

Let's assume that catalysts like those above do actually function by delivery of the metal centre (and bound H) to the olefin by intramolecular coordination of e.g. a neighbouring carbonyl:

(taken from this PNAS paper)

 

...then the geometry of the PZQ enamide prevents such delivery. This is the basic problem here.

 

Update September 20th - First results from DSM (results will periodically be posted as screens are carried out):

Two conclusions from this:

1) The method of preparation of the enamide (using a sulfur melt) does not in this case lead to contamination with sulfur that deactivates catalyst.

2) MonoPhos does not give conversion under these conditions.

 

Update October 20th 2010 - Second update from DSM

Conditions were: cat = 0.01mmol, substrate PZQ = 0.2mmol, 5mL solvent, 25 bar H2, 16h

Catalyst structures are:

Update Jan 25th 2011 (data from DSM received December 21st 2010)

 

New attempts with Rh. Preparation of the 48 catalysts: The catalyst is preformed by stirring Rh(COD)2BF4 with 1.1 eq ligand for 1 h in DCM at rt. [Rh] = 0.042 M. See table below for the list of the 48 ligands used.

 

Preparation of the 48 reaction mixtures: The catalyst Rh/L (0.0042 mmol Rh in 0.5 mL DCM) in DCM solution is transferred to hydrogenation vials with the liquid handling robot. The substrate in MeOH (0.035 mmol in 2.2 mL MeOH) is added. S/C = 8.

 

Hydrogenation conditions: 25 bar H2, 60°C, 18 h, stirring = 300 rpm

 

Results

No product was obtained for almost all catalysts - except for traces (a few %) of product for vials 29, 40, 42 corresponding to ligands CTH-PhanePhos (E4), JosiPhos-2-1 (H5), JosiPhos-212-1 (B6). These results are rather consistent. PhanePhos was already identified as a good substrate for a related substrate. The 2 JosiPhos present similar structural features: aryl-aliphatic phosphines with in both cases t-Bu groups.

 

 

See attached file “DSM 48 vial screen” for raw data for this screen, including the structures of the ligands employed.

 

Update July 3 2012 (using results obtained from DSM 22/09/2011)

Some promising results - the best to date. In the first run, below, there is full conversion for the first two entries, with some enantiomeric excess. The solvent change to TFE was probably important.

Conditions: cat = 0.01mmol, substrate PZQ = 0.2mmol, 5mL solvent, 25 bar H2, 16h. SL-M004-1 is the MandyPhos ligand of Solvias - structure may be seen in their catalog.

The second run used TFE. There is a lot of variation depending on ligand, but again some promising results here, with JosiPhos and PipPhos both doing well but wth low conversion.

Note added July 3 2012 - all results to date have been collated in a Google Spreadsheet for easier viewing.