Customer
dsm-firmenich
Context
Hydroformylation is a cornerstone reaction in producing aldehydes used in fragrances, specialty chemicals, and pharmaceuticals. Traditional R&D approaches struggle to efficiently optimize these reactions because Rhodium-based catalysts are scarce and expensive, and the multi-parameter reaction space makes conventional experimentation slow, costly, and resource-intensive. Accelerating this process is critical to delivering sustainable, high-quality products at competitive cost.
Challenge
- Precious metal dependency: Hydroformylation reactions rely heavily on rhodium-based catalysts, one of the most scarce and expensive metals used in industrial chemistry.
- Economic and sustainability pressures: Need to dramatically reduce catalyst loading/usage to meet green chemistry goals and reduce manufacturing costs without compromising reaction performance.
- Complex optimization landscape: Traditional research methods (OFAT, DoE) are inefficient for navigating the vast search space of interdependent reaction parameters, including temperature, pressure, catalyst concentration, ligand selection, substrate ratios.
- High R&D workload: Scientists spend excessive time on trial-and-error experimentation, slowing innovation cycles in a competitive industry.
Solution
- Bayesian Optimization through SDLabs platform to intelligently navigate a 7-dimensional search space of ~2.9 billion possible combinations in just 88 experiments, significantly reducing experimental workload.
- Targeted maximization of conversion, selectivity, and catalyst performance while minimizing Rhodium usage and reaction time.
Business Impact
- Catalyst Efficiency: 10-30x reduction in Rhodium catalyst loading, sustaining high conversion/selectivity.
- Cost Reduction: 95-97% reduction in Rhodium cost contribution (from €102 to €5/kg and €127 to €4/kg).
- Time Efficiency: Reaction time cut by 50% (16 hours to 8 hours).
- Sustainability & Innovation: Fewer experiments, less Rhodium consumption, greener chemistry achieved.
- Strategic Impact: Demonstrates AI, and the SDLabs platform as an effective tool for accelerating R&D, reducing costs, and enabling greener industrial chemistry.
Resources
- Link to Publication: Saudan et al. Bayesian Optimization for Resource-Efficient Hydroformylation. ACS Catalysis. 2025. https://doi.org/10.1021/acscatal.5c06595
- Link to News Announcement on Atinary and dsm-firmenich collaboration
- More about dsm-firmenich
