Solvents dominate mass input, energy demand, and environmental impact in pharmaceutical manufacturing, yet solvent selection and recovery are often evaluated using fragmented or non-comparable metrics. Here, we present a solvent-centric sustainability framework that integrates mass-based indicators with life-cycle and energy metrics to enable transparent comparison of conventional and redesigned solvent systems. The framework harmonizes Process Mass Intensity (PMI), circular PMI (cPMI), Global Warming Potential (GWP), and Cumulative Energy Demand (CED) within consistent cradle-to-gate system boundaries, supported by literature-derived data, machine-learning (ML) models, and digital-twin–based sustainability assessment tools. The methodology is demonstrated using Sertraline as a representative solvent-intensive active pharmaceutical ingredient (API). A simplified, literature-based synthesis route contextualizes solvent use across key reaction and isolation steps. Targeted solvent substitutions—most notably replacement of tetrahydrofuran, chlorinated solvents, and dipolar aprotic media with 2-methyltetrahydrofuran and ethanol-based systems—are evaluated alongside enhanced solvent recovery and catalytic hydrogenation. Relative to the solvent-dominant subsequence of the synthesis (PMI ≈ 78 kg·kg−1 API), for which detailed solvent mass-balance data are available, the redesigned solvent strategy reduces PMI to approximately 45 kg·kg−1 API, achieves a cPMI of 6–10 at ≥80% solvent recovery, and consistently decreases GWP and CED. By explicitly mapping solvent redesign outcomes to the 12 Principles of Green Chemistry, this study demonstrates how solvent-focused interventions, supported by predictive digital tools with excellent agreement between modelled and empirical trends, can deliver substantial sustainability improvements without modifying the underlying synthetic route or relying on proprietary process data. While not intended as an industrial benchmark, the Sertraline case study illustrates how harmonized metrics, life-cycle thinking, and AI-enabled digital assessment can support evidence-based solvent selection and sustainability-oriented process development in API manufacturing.
Isak et al. (Thu,) studied this question.