Case study: Two-Steps Purification with E-PAK® Cartridges Following a Direct Pd-Catalyzed Borylation

> back to E-PAK Purification Cartridges

An R&D service performed for a mid-sized pharmaceutical company

> download the Case Study

PROJECT

This work was sponsored by a mid-sized pharmaceutical company that is a SiliCycle customer since 2014. The API involved herein is in Phase III trials.

In this project, after evaluating various & more traditional purification means, they opted for SiliCycle’s SiliaMetS Metal Scavengers for the efficient removal of Palladium which contaminated a synthetic intermediate following a direct Pd-catalyzed borylation reaction.

SiliCycle was able to successfully optimize the entire purification protocol through its new integrated solution: E-PAK® radial-flow cartridges, a product specifically developed for pharmaceutical purifications from R&D scale to full production. Advantages of this technology include full recovery of API, marginal leaching, solvent & extended pH range compatibility, straightforward scalability, environment friendly, sounds economics. The details of this optimization are described below.

As part of a large scale manufacturing process, the purification of a direct Pd-catalyzed borylation step was initially performed via two successive treatments: a two-pass filtration through activated carbon (using twice the same carbon filtration cartridge), followed by a Palladium scavenging step with SiliaMetS Thiol.

Experimental procedure

• In a 5 L three-neck-flask was introduced 1,93 L of methanol (HPLC grade), Pd(OAc)₂ (21.38 mmol), PPh3 (29.50 mmol) and Cs₂CO₃ (1,105 mmol). The reactor was purged with argon, while maintaining mechanical agitation. The iodobenzene compound (737.2 mmol) and the chiral diboron (884.7 mmol) were then added, while maintaining a slight argon stream.
• The mixture was stirred under inert atmosphere for 3 hours at 50°C, (starting iodide was totally consumed). Methanol was evaporated until approximately 1000 mL was left. A portion of 2,000 mL of ethyl acetate was added, and the organic phase washed with 1,100 mL of HClaq (4M).
• The obtained mixture was then used as is, to test different purification procedures / sorbents.

Albeit good recovery (95 %) and satisfying Palladium removal (< 20 ppm) obtained on a process scale of 50 - 100 Kg, one main issue was highlighted:

• The activated carbon filtration step showed to be inefficient, requiring high amount of Metal Scavenger in subsequent purification step, and thus increasing processing time. Improving the effectiveness of this step was required.

This one point was addressed as part of the optimization of this purification process on a 200 g scale.

FIRST, ROUGH PURIFICATION: ACTIVATED CARBON FILTRATION

Preliminary tests in standard filtration conditions

Out of the four available activated carbons in E-PAK Cartridges, three were tested in a standard filtration procedure following the borylation reaction.

Experimental procedure

• Following the borylation reaction work-up, 3 x 60 mL of the organic phase were filtered through 0.56 g of three different activated carbon (SiliaCarb CA, SiliaCarb HA and SiliaCarb VW) and rinsed with 12 mL of ethyl acetate.
• A 0.5 mL sample of each filtrate was collected, and analyzed by ICP-OES for Pd content.
• Each filtrate was washed again with 70 mL of aqueous HCl (4 M), and filtered through the same carbon filter cakes.
• A 0.5 mL sample of each filtrate was collected, and analyzed by ICP-OES for Pd content.

Conclusions

• Although all three carbons gave similar results, SiliaCarb HA showed slightly better scavenging performance.
• In all cases, the 2nd treatment improved the scavenging results to some extent.
• Hence, activated carbon SiliaCarb HA was selected for the subsequent tests with E-PAK cartridges.

Optimized purification through SiliaCarb HA Carbon E-PAK Cartridges

A 5 x 1 cm SiliaCarb HA (5 g) E-PAK cartridge was inserted in the appropriate housing (lab scale housing).

Experimental procedure

• 150 mL of ethyl acetate (HPLC grade) was first used to pre-condition the unit.
• 525 mL of the crude solution was then passed through the cartridge at a 12.5 mL/min flow rate.
• Lastly, the cartridge was eluted three times at the same flow.
• At each passage, a 0.5 mL sample was collected and analyzed by ICP-OES for Pd content.

Conclusion

Two consecutive passes through a 5 x 1 cm SiliaCarb HA E-PAK cartridge was shown to be sufficient to obviate a second HCl washing step and yield decreased Pd content.

SECOND, MORE TARGETED PURIFICATION: SiliaMetS METAL SCAVENGING

Preliminary tests in bulk conditions

Scavenging screenings are the best way to start off with SiliaMetS functionalized resins. This is because even for the same functionality or metal, a variation in the scavenging efficiency can be observed depending on the nature of the products present in the solution to be treated. For example, steric hindrance, electronic effects, H-bond, all are factors that can greatly influence the scavenging %.

In order to determine the most suitable Pd scavengers for the current system, a screening was realized using 8 of our most suitable grafted silicas. Influence of temperature and reaction time were evaluated at the same time.

Experimental procedure

• A 500 mL solution of 50 g of Intermediate S in ethyl acetate (HPLC grade) was made in a volumetric flask.
• In respect to Intermediate S: samples of 80 mg (10 % w/w), 160 mg (20 % w/w) and 240 mg (30 % w/w) of 8 different SiliaMetS were pre-weighed in 24 polypropylene tubes suited for SiliCycle’s MiniBlock® Platform.
• 8 mL of the above solution (containing 0.8 g of Intermediate S) was added to each tube.
• The MiniBlock Platform was orbitally shaken for 60 min. at room temperature, and sample of 0.5 mL of each solution were collected and filtered through 0.45 μm filters.
• The remaining solutions were shaken for 3 more hours and filtered off.
• All obtained samples (48) were analyzed by ICP-OES for Pd content.
• An identical manipulation was made at 50°C, generating 48 more samples to be analyzed.

Conclusions

• SiliaMetS DMT and SiliaMetS Thiourea were shown to be the best scavengers for efficient Pd removal from this process stream.
• 20 to 30 % w/w of grafted silica was necessary to obtain satisfactory Pd scavenging.
• Increasing temperature and reaction time slightly improved scavenging performances.
• For the conversion to E-PAK purification, SiliaMetS DMT was chosen over SiliaMetS Thiourea due to its ready availability on industrial scale.

Leaching investigation

In order to determine if any extractables had leached into the final product, the following experimentation was run.

Experimental procedure

• Samples of 0.3 g of SiliaMetS DMT was suspended into 5 mL of ethyl acetate (HPLC grade), and shaken for 4 hours at room temperature.
• The suspension was filtered through a 0.45 μm syringe filter, evaporated and reconstituted in 1 mL of methanol (HPLC grade).
• A blank sample was also assayed.
• LC-MS (ESI+) was run on both samples. No extractables were detected by LC/MS/MS.

Optimized purification through SiliaMetS DMT E-PAK Cartridges

Following previous results, E-PAK purification was conducted under 20 and 30 % w/w conditions.

Experimental procedure

• A 5 x 1 cm SiliaMetS DMT (8 g) E-PAK cartridge was inserted in the appropriate housing (laboratory scale).
• 150 mL of ethyl acetate (HPLC grade) was first used to pre-condition the unit.
• Solution S was prepared adding 40 g of Intermediate S to 200 mL of ethyl acetate in a volumetric flask. A 0.5 mL sample was retrieved for subsequent ICP-OES analysis.
• The entire solution was passed through the cartridge at a 12.5 mL/min flow rate.
• Lastly, the cartridge was eluted three times at the same flow.
• Samples of 0.5 mL were collected after each run, and analyzed by ICP-OES for Pd content.

A similar experiment was run following a 30 % w/w ratio (acc. to preconditionned E-PAK SiliaMetS DMT).

• Hence, 133 mL (26.6 g of Intermediate S) of Solution S was passed through a new E-PAK SiliaMetS DMT at 12.5 mL/min flow.
• Lastly, the cartridge was eluted 7 times at the same flow.
• Samples of 0.5 mL were collected after each passage, and analyzed by ICP-OES for Pd content.

Conclusions

20% w/w was sufficient to achieve maximal scavenging of Pd.

Side reaction investigation

Final products obtained above were analyzed by LC-MS (ESI-), along with a sample of intermediate S. Methanol (HPLC grade) as a blank sample was also tested.

Conclusion

No new product was detected.

Final Conclusions

• All tedious & time consuming steps of the initial traditional purification protocol could be avoided:
  • 1st filtration on activated carbon followed by washings with fresh ethyl acetate
  • 2nd washing of the obtained organic phase with HClaq (4 M)
  • 2nd filtration on activated carbon followed by washings with fresh ethyl acetate
  • Treatment with 30 % w/w of SiliaMetS Thiol at 60°C for 6 h
  • 3th filtration, followed by ethyl acetate evaporation
  • Recrystallisation from cold heptane
• Activated carbon SiliaCarb HA provided the best scavenging results among all three tested carbons.
• Two consecutive passes through an E-PAK SiliaCarb HA cartridge obviated a second HCl wash.
• Treatment of the Intermediate S with 20 – 30 % w/w of SiliaMetS DMT in bulk mode at room temperature provides a good way to get rid of over 98 % of Pd.
• Similar results were obtained when the compound was filtered through an E-PAK SiliaMetS DMT cartridge (20 – 30 % w/w of SiliaMetS DMT) at room temperature, but from an experimental point of view, the modus operandi with E-PAK cartridges may be considered as a much more user friendly procedure and great alternative to avoid the presence of insolubles in reactor.

SiliCycle can help you design, scale up & optimize metal removal processes.
Contact us at [email protected]

Learn more:

• E-PAK Cartridges Portfolio
• SiliaMetS Metal Scavengers