Precious metal catalysts used in production of APIs can be costly and difficult to retrieve. Dr Pete Jackson, ceo of Reaxa, looks at ways to cut those costs and reduce residue levels
Homogeneous catalysis is certainly well established, with numerous examples of great chemistry being carried out at large scale, including Pd cross-coupling reactions, chiral chemocatalysis, hydroformylation and selective hydrogenation. Emerging technologies, such as metathesis, are also starting to make an impact.
However, as these precious metal technologies are becoming widely used - and they now feature in the production of almost 30% of new drugs moving into clinical trials - a number of corresponding and significant challenges need addressing.
Focused on cost and compliance, these typically lie in preventing metal contamination of end products; in low product yield; and in loss of expensive catalyst as waste. During process, work-up and purification, as much as 15% of the precious metal catalyst may be lost.
Pharmaceutical regulatory authorities are reducing the acceptable levels of precious metal contamination in active pharmaceutical ingredients (APIs). They aim to protect patient safety through ever more stringent regulations on Permitted Daily Exposure (PDE) to metal residues in medicines. This now means that total residual levels below 5ppm are routinely required. For complex processes involving several metal-catalysed steps, individual catalyst metal residues may therefore need to be 1ppm or lower.
Far upstream from material waste, another important aspect is that over the past few years precious metal demand has outstripped supply, leading to substantial rises in the cost of catalysts, and significant volatility in the month-to-month pricing. In addition, the cost of complex ligands, e.g. for chiral hydrogenation catalysts, may be higher than the cost of the metal. This is increasing demand for effective catalyst and metal re-use, recovery and recycling, for example using Umicore's Ecolyst technology to recover precious metal value from solvent waste streams.
Existing homogeneous processes often require significant process work-up and re-crystallisation to separate effectively the products from the catalyst and remove catalyst residues from APIs. For instance, one process using a Rhodium homogeneous catalyst system generates in excess of 100t of process effluent for each tonne of API manufactured. The impact and cost of an environmental burden on this scale clearly can't be sustained.
Reaxa has developed a twin-track, "immobilisation and scavenging" strategy to help customers address some of these challenges. Such strategies offer a new approach to immobilising precious metal catalysts, catalyst recovery and re-use - and then combine them with more efficient recovery of metal contaminants from products and process waste.
Immobilisation - Reaxa's EnCat microporous polymer encapsulation technology effectively traps precious metals and ligands, such as phosphines, resulting in an immobilised homogeneous catalyst.
Pd EnCat catalysts can be substituted into existing processes, such as Suzuki couplings of boronic acids to aryl halides, resulting in very low levels of contamination from both metal and phosphine. In typical examples, resulting crude API has less than 10ppm Pd contamination, removing the need for complex work-up and re-crystallisation. To date, the company has introduced seven EnCat product variants.
Scavenging - The company has also developed a range of precious metal scavengers called QuadraPure. These can be used in homogeneous processes where EnCat immobilisation is not possible, or additionally to remove residual metal contamination from heterogeneous catalytic processes, eg, hydrogenations using Pd or Pt supported on carbon. Counting the most recent extensions, the QuadraPure range is now a family of seven macroporous and three microporous resins with a wide range of functionalities.
scavenger screening
A growing trend within the pharmaceutical industry is the establishment of scavenger screening facilities, reflecting the increasing importance of this approach in drug synthesis.
Reaxa has established a screening capability to speed up the selection of the optimum QuadraPure scavenger system for client processes. This looks at the important parameters, such as loading, solvent selection and temperature, as well as offering the capability to screen in batch or flow-through mode. A screening guide is also available for customers.
New approaches to the challenges of limiting and managing contamination have also led to the application of "crossover" technology. For Reaxa, applying the benefits of the auto industry's latest emission control technologies to pharma production led, in late 2005, to a new licensing agreement with Japan's Hokko Chemical Industry Co. This opened the way for a new generation of palladium containing perovskite (LaPCat) catalysts, originally developed for use in catalytic converters, to be adapted and evolved in organic chemistry applications.
One of the advantages of perovskites as catalysts is the exceptionally low level of precious metal contamination. Simple removal of the bulk catalyst by filtration has led to palladium levels of below 2ppm being found in crude Suzuki reaction products. With current European specifications limits for metal catalysts at 5ppm, this would mean a significant saving in clean-up costs. Furthermore, in laboratory experiments, turnover numbers as high as 400,000 have been seen, meaning that very low amounts of LaPCat may be needed compared with conventional homogeneous catalysts.
The key parameter for process scientists is total process cost. This equates to the overall cost effectiveness of the manufacturing process taking into account raw materials costs, yields, cycle time, operability (e.g. manual operations, cleaning, etc), work-up, separations, crystallisation, re-crystallisation, isolation, drying and waste disposal, as well as any capital expenditure costs for plant construction or modification.
Re-engineering processes to implement EnCat technology can result in substantial reduction in total process cost. Work at Reaxa has shown that switching one customer's process from the previous existing homogeneous Pd process to use Pd EnCat resulted in crude product meeting the Pd specification - leading to a 50% reduction in solvent required, a 20% improved overall cycle time and giving a pure API yield increase of 10%.
QuadraPure scavenging can also lead to process cost-saving. One homogeneous process is operated at tonnes scale to generate two process waste streams, containing spent Pd catalyst at 500ppm (process waste) and 5ppm (crystallisation liquors & washings).
Scavenging can be used to recover each of these streams, but further process re-engineering to incorporate scavenging following reaction may be developed to eliminate the need for re-crystallisation and give even higher yields of pure API. In the example given in table 1, the cost of the scavenger is less than the direct process savings achievable and further savings may also be gained from the reduction in the cost of solvents, elimination of waste processing and more effective metal recovery.
Immobilised homogeneous catalysts also offer the potential for process intensification by use in flow reactors. Reaxa has demonstrated continuous flow processing at lab scale using model Suzuki coupling reactions over 10 days, generating 175g of product from 1g Pd EnCat (0.05g Pd).
Other experimental results show that EnCat polymer beads act as "micro-reactors" under microwave processing, due to preferential heating of the polymer matrix. This can lead to accelerated reaction rates and very clean products, with minimal impurities.
Flow microwave chemistry is becoming increasingly significant as a tool to speed up drug development and process efficiency. By replacing traditional glassware with columns and reactors that can be loaded with immobilised reagents, catalysts or scavengers, flow microwave set-ups facilitate effective removal of contaminants in a continuous and dynamic fashion (see figures 2 & 3).
Microwave chemistry is ideal for adoption into a continuous flow process, since the reaction times (and, therefore, residency times in a reactor) are short. Such methods open up the possibility of carrying out reaction sequences in an automated flow system, with additional in-line clean-up and purification.
Typical results of flow processing are shown in figure 4 for two examples of Suzuki couplings using Pd EnCat 30 catalyst. Product yields are >99%, purities are >98% and there is less than 10ppm of palladium contamination. In each case, the crude product was pure enough for use without additional washing and re-crystallisation.
Work is ongoing to scale up immobilised processes using conventional and microwave heating by a factor of 10 in the laboratory to allow production of kilos of product in flow reactors.
Beyond that, the challenge is to scale up further into pilot plant projects and production scale, and to extend the immobilisation concept to chiral cata-lysis - all developments that address the "easier, faster, cleaner" agenda. The main challenge is to gain acceptance for new, re-engineered processes at commercial scale production in cGMP environments using immobilised catalysts and scavengers to reap the rewards of the total cost reduction approach.
Whether homogeneous, immobilised or scavenging approaches are used, it is clear that precious metal catalysis will continue to grow in importance. Three of the drug industry's key drivers - cost, complexity and compliance - seem certain to sustain a high level of interest in technologies able to manage the contamination challenges that come with this territory.