Written by Jonathan Moseley – July 4th, 2024.
A recent review on the “Diversification of Pharmaceutical Manufacturing Processes: Taking the Plunge into the Non-PGM Catalyst Pool” in ACS Catalysis by a combined academic/pharma/catalyst supplier group caught my eye this week (ACS Catal. 2024, 14, 9708-9733; https://doi.org/10.1021/acscatal.4c01809) (PGM = platinum group metal, specifically Ru, Rh, Pd, Os, Ir and Pt). As the title suggests, it is predominantly focused on pharma and scale-up/manufacturing processes, particularly in the last 20 or so years, which should be of interest to readers of these blog pages. However, it does also list some discovery chemistry relevant examples including screening studies, on the justifiable grounds that academic developments in these areas are likely to produce scalable economic processes in due course; and it also includes some informative agrochemical examples.
Before getting into the detail however, there is a brief survey on the use of non-PGM catalysis in some well-known industrial processes conducted on huge scale (Haber-Bosch process (Fe); steam reforming (Ni); Fischer-Tropsch synthesis (Co/Fe); DuPont adiponitrile process (Ni) amongst others) – a potent reminder that other industries have been using non-PGM catalysis economically for decades, albeit for different types of chemistry (see Figure 6 for a collation).
For organic pharmaceutical type chemistry, this review is naturally focused heavily on C-C and C-N (C-X) cross-coupling reactions, although other reactions are also covered. There is a brief philosophical digression on why Pd rather than Ni came to prominence in this field in the 1970s. Despite some impressively efficient early examples (Kumada couplings with Ni), non-PGMs generally require higher catalyst loadings and are less selective and less stable to reaction conditions, have less well understood mechanisms and are more prone to poisoning or deactivation. The advantages of course are much reduced cost, greater global availability, reduced toxicity (not in every case), and usually easier removal from waste streams and products (APIs).
The authors speculate that the recent spike in PGM prices has re-focused interest back onto non-PGM catalysis (see below), such that in 2019 there were over 16,000 references to non-PGM chemistry, although not all relevant to catalysis of course. Aside from research funding for PGM catalysis possibly drying up, an unspoken consideration might also be that PGMs themselves generally come from geopolitically less stable countries (see their Table 1). This factor that doesn’t really affect Ni and Cu even if required in millions of tonnes (Co is more of a concern) – and certainly not at the quantities that high-value pharmaceutical and other manufacturing would require.
Given these obvious advantages, why has there not been greater take up of non-PGM catalysis by the pharmaceutical and related industries? The authors suggest that the much more limited range of commercially available ligands for say Ni versus Pd, coupled with greater mechanistic complexity and catalyst sensitivity has deterred academic research until more recently. And secondly by the longer development time and generally lower robustness of potential processes in an industrial setting – the value lost by being 6 months later to the market versus a competitor drug or having 6 month’s less patent exclusivity versus generic competition on a blockbuster drug may never be recovered compared to the relatively modest manufacturing cost savings when using non-PGM catalysis.
The non-PGM base metal catalysts specifically considered here are Ni, Cu, Co, Fe and Mn (also known as earth abundant metals, EAMs); Ca, Mg and Zn get a brief final mention. Sections on each metal start with a useful collation of recent reviews for those who want more detail. Indeed, this review serves as something of a “review of reviews” for recent developments in use of these base metal catalysts. A handful of examples will be illustrative:
Nickel: in the kilo-scale synthesis by Suzuki-Miyaura coupling of anti-cancer agent GDC-0941, the Ni-catalysed coupling was higher yielding and gave the final API directly compared to both of the previously developed Pd-catalysed methods. Inexpensive and available Ni(NO3)2.6H2O was used “at an impressive 0.03 mol%” with equally cheap Ph3P ligand (0.06 mol%). Furthermore, expensive scavengers were required to remove residual Pd, whereas residual Ni was easily removed with aqueous NH3 washes (Org. Process Res. Dev. 2013, 17, 97-107; https://doi.org/10.1021/op3002992).
Copper: illustrating other reasons for using non-PGMs, even at higher loadings (12 mol%), the CuI catalysed C-N coupling with a cheap diamine ligand to facilitate synthesis of the aminopyrazole moiety in two alternative steps avoided issues with the standard two-step nitro reduction sequence of energetic chemistry, need for hydrogenation equipment and potential genotoxic impurities (PGIs) – conducted on >50 kg scale. Residual Cu levels, much less toxic in mammals than PGMs, were readily reduced to <10 ppm (Org. Process Res. Dev. 2021, 25, 1065-1073; https://doi.org/10.1021/acs.oprd.1c00066).
Cobalt: is much less used than Ni or Cu, possibly due to a lack of specialised ligands. Several examples are supplied, including one recently very relevant, an academic study rapidly use-tested on >200 kg for the key cyclopropanation reaction in the synthesis of Nirmatrelvir (for SARS-CoV-2). This reaction does require stoichiometric Zn (another non-PGM) and Co at 15 mol% for success, so there is scope for development, but in a global pandemic, one wouldn’t complain (J. Am. Chem. Soc. 2023, 145, 9441-9447; https://doi.org/10.1021/jacs.3c01949; Org. Process Res. Dev. 2023, 27, 2260-2270; https://doi.org/10.1021/acs.oprd.3c00249).
Manganese: lastly, a mixed Mn/Cu catalysis from Syngenta illustrates an agrochemical example. Again, catalyst loadings are high at 10-20 mol%, which for a non-PGM can often be prohibitive for agrochemical economics, but one must also consider what chemistry is avoided by using such – in this case aryl lithium chemistry at -78 °C and a very toxic aryl lead intermediate. Synthesis was reported on “kilo scale”, so clearly the economics were at least worth considering (Org. Process Res. Dev. 2017, 21, 1625-1632; https://doi.org/10.1021/acs.oprd.7b00241).
In summary, this review does not claim to be exhaustive, but rather is illustrative of recent developments in non-PGM catalysis, and is perhaps prophetic in the short term – it is certainly a good starting point for reviewing any of the non-PGMs mentioned, and an easy read overall.
Training courses mentioned can be found at: https://www.scientificupdate.com/training/courses/
