Recent advances in computer-aided retrosynthesis (CAR), flow chemistry, and continuous manufacturing collectively offer new opportunities to enable environmentally sustainable development and manufacturing practices across the pharmaceutical development and manufacturing value chain. However, the implementation of these methods and technologies remains scattered and fragmented, preventing full realization of their potential to address one of the most urgent needs in the pharmaceutical and related sectors. This work introduces a holistic digital framework for the design and optimization of an end-to-end manufacturing process for paracetamol (acetaminophen). The framework integrates Green-by-Design synthetic and purification routes of the active pharmaceutical ingredient (API) aims to deliver cost efficiency and robust quality, safety, and environmental sustainability assurance. The approach integrates AI-driven CAR with plant wide modelling, Techno-Economic Analysis (TEA), and prospective cradle-to-gate prospective Life Cycle Assessment (LCA) to evaluate designs options and greener, chemically feasible, and more selective synthetic pathways. An end-to-end mathematical model is implemented in gPROMS Formulated Products® to optimize operating and design parameters and generate inventory data for TEA/LCA. The framework also embeds Quality by Digital Design (QbDD) to identify CPPs and CMAs that influence CQAs, enabling definition of a robust design space. Finally, a multicriteria decision-aiding approach is implemented to determine the best trade-offs among yield, productivity, resource efficiency, cost, and environmental performance, providing a transferable Safe-and-Sustainable-by-Design (SSbD) methodology for developing greener and more resilient pharmaceutical manufacturing systems.

Reliable multimodal monitoring in crystallization processes remains challenging due to heterogeneous PAT signal quality, sensor drift, asynchronous sampling and nonstationary noise. This work presents a machine-learning-assisted fusion framework that integrates multimodal PAT alignment, estimation and physics-guided regularisation to generate coherent concentration and particle-size trajectories. A mechanistically informed simulation platform is developed to produce synthetic Raman, FTIR, FBRM and image-based crystal size data with realistically simulated drift, heteroscedastic noise, dropouts and distortion patterns. Sensor reliability is inferred through a Random Forest model trained on variance-normalised discrepancies and quality metrics, which allows the dynamic adjustment of channel contributions. Across modalities, the Random Forest achieves MAE values of 0.03-0.20 for probability-type indicators and shows stable explanatory power for variance-inflation factors on particle-size channels (R2 = 0.81-0.93). Two representative PAT scenarios illustrate the performance of the proposed framework under different cross-sensor discrepancy structures, demonstrating improved robustness, reduced local variance and enhanced physical coherence compared with individual signals. Overall, the results highlight the potential of multimodal information fusion as a foundation for trustworthy online monitoring and modelling in crystallization.

This review explores sustainable synthetic methodologies, highlighting advances in solvent use, catalysis, raw materials, and flow chemistry for pharmaceutical and agrochemical production.

This review highlights systemic innovations, such as digital retrosynthesis, AI-guided design, smart manufacture, modular plants and 3D printing, as levers for scalable, low-impact fine chemical production.

Crystallization plays a critical role across multiple industries, determining key particulate product attributes such as purity, particle size distribution, morphology, and polymorphic form. The multiscale nature of this process, encompassing molecular interactions, phase transitions, and transport phenomena, imposes high levels complexity on the design and control of systems in which particle characteristics govern performance. Digital strategies, including emerging AI-driven approaches, are increasingly recognized as powerful tools for managing multiscale complexity, reducing inherent uncertainties, and enhancing process development. This review aims to explore recent advances and critically analyze how digital methods can be applied at each stage of process development. The discussion begins with data acquisition and augmentation, including synthetic data generation, model-based experimental design, and rigorous data preprocessing and validation. This is followed by the modeling strategies tailored to specific design and operation objectives, including mechanistic, data-driven, and hybrid approaches for predicting crystallization dynamics and particulate properties. Finally, recent control and optimization solutions are discussed, focusing on model-based and adaptive algorithms for open and closed-loop strategies. The review concludes with a forward-looking perspective on emerging trends, highlighting the integration of digital twins, real-time optimization, and sustainability metrics which together are expected to enable intelligent, resilient, and sustainability-aligned crystallization systems capable of meeting future industrial and regulatory requirements.

Environmental sustainability is increasingly recognized as a critical consideration in pharmaceutical development, yet it is rarely incorporated at the scale of molecular-level design. This study introduces a strategy to predict cradle-to-gate indicators that can be flexibly incorporated into multiple early-stage molecular prioritization scenarios. A dataset of 150 pharmaceutical-relevant molecules was compiled, with each molecule described by structural descriptors, thermophysical properties, and ReCiPe endpoint indicators representing human health, ecosystem quality, and resource scarcity. A dual-branch multi-task model combining graph-based and descriptor-based representations was trained to predict these three endpoint indicators. Model performance was evaluated through validation metrics, local sensitivity analysis, and SHAP-based interpretability. A case study with solubility-based feasibility constraints was then used to illustrate how different sustainability weighting schemes affect molecular ranking and to demonstrate the potential for incorporating sustainability assessment into early-stage molecular prioritization. The results indicate that sustainability preferences can lead to distinct prioritization patterns, while some candidates remain comparatively favourable across scenarios.

This paper presents a novel framework for the systematic evaluation, discrimination, and calibration of mathematical models. A diverse set of model candidates is first constructed to capture a broad range of underlying physical and kinetic phenomena. Structural identifiability analysis is then employed to determine whether unique parameter estimates can be derived from ideal, noise-free data, enabling early-stage model screening. A single model is then selected based on a rigorous model discrimination criterion. To calibrate and refine the selected model, parameter estimability analysis is integrated with Model-Based Design of Experiments (MBDoE), ensuring optimal sampling strategies, data-rich experiments, and reduced prediction uncertainty. To maximize the effectiveness of MBDoE, a novel temperature cycling protocol is introduced, enabling a single, well-designed, information-rich experiment that minimizes experimental effort. This methodology is demonstrated through a cooling crystallization case study using paracetamol. The results show that the proposed framework significantly enhances model discrimination, parameter precision, and estimability resulting in a model with superior predictive performance, as proven by the additional post-MBDoE validation experiment operated under different conditions and designed independently from the MBDoE framework. The proposed framework bridges the gaps between the different modeling pillars, reduces experimental efforts, and lays the foundation for more effective modeling and optimal design of experiments for the crystallization systems and beyond.

This study presents a data‐driven modeling and multi‐objective optimization framework for an integrated section of continuous pharmaceutical manufacturing, focusing on flow synthesis and continuous crystallization. To address data scarcity and trade‐offs among product quality, efficiency, and environmental impact, the framework combines generative adversarial networks (GANs), artificial neural networks (ANNs), and genetic algorithms (GAs). An integrated dual‐GAN (ID‐GAN) generates data under physicochemical constraints, which are merged with real data to train an ANN with 15%–20% mean absolute errors for particle size, productivity, and a sustainability throughput index. The ANN is then coupled with a GA to identify Pareto‐optimal solutions based on user‐defined objectives and constraints. Case studies validate the framework's capability to facilitate process design decisions by systematically exploring trade‐offs among competing objectives, underscoring its potential utility in the digitalization of critical units within continuous manufacturing systems.

Solvent selection in pharmaceutical crystallization plays a pivotal role in determining overall manufacturing efficiency while also significantly impacting environmental performance and regulatory compliance. A data-driven solution for sustainable solvent selection, applicable to both single and binary solvent systems, was developed and integrated into SolECOs (Solution ECOsystems), a modular and user-friendly platform for Sustainable-by-Design solvent selection in pharmaceutical manufacturing. A comprehensive solubility database containing 1186 active pharmaceutical ingredients (APIs) and 30 solvents was constructed and used in conjunction with thermodynamically informed machine learning models, including the Polynomial Regression Model-based Multi-Task Learning Network (PRMMT), the Point-Adjusted Prediction Network (PAPN), and the Modified Jouyban-Acree-based Neural Network (MJANN), to predict solubility profiles along with associated uncertainties. Sustainability assessment was performed using both midpoint and endpoint life cycle impact indicators (ReCiPe 2016) and industrial benchmarks such as the GSK sustainable solvent framework, enabling a multidimensional ranking of solvent candidates. Experimentally validated case studies involving APIs such as paracetamol, meloxicam, piroxicam, and cytarabine confirmed the approach's robustness, adaptability to various crystallization conditions, and effectiveness in supporting single and binary solvent screening and design.

This study presents a comprehensive comparison of the three alternative downstream manufacturing technologies for pharmaceuticals: i) Dry Granulation (DG) through roller compaction, ii) Direct Compaction (DC), and iii) Wet Granulation (WG) based on the economic, environmental and product quality performances. Firstly, the integrated dynamic mathematical models of the different downstream (drug product) processes were developed using gPROMS formulated products based on data from the literature or/and our recent experimental work. The process models were developed and simulated to reliably capture the impact of the different design options, process parameters, and material attributes. Uncertainty analysis was conducted using global sensitivity analysis to identify the set of critical process parameters (CPP) and critical material attributes (CMA) that mostly influence the quality and performance of the final pharmaceutical tablets in each case, captured by the critical quality attributes (CQAs). Based on the set of CPP and CMA, the combined design spaces, which guarantee the attainment of the targeted CQA, were identified and compared. Additionally, based on the process simulations results and inventory data, the techno-economic Analysis was performed alongside life cycle assessment (LCA). The LCA provided an in-depth evaluation of the environmental impacts associated with each manufacturing method, considering aspects such as energy consumption, raw material usage, emissions, and waste generation based on a cradle-to-gate approach. By integrating the CQAs and critical emission categories within a Quality and and Sustainability by Digital Design (QSbDD) paradigm, this study offers a holistic analysis that captures both the environmental and product quality performance.

This study investigates the synergistic integration of Computer-Aided Retrosynthesis (CAR) and continuous flow chemistry to identify and optimise shared synthetic pathways for multiple active pharmaceutical ingredients (APIs). CAR was employed to identify shared synthetic routes across 11 different APIs, leveraging a Hantzsch thiazole synthesis as a shared reaction step for all investigated targets. The results showed that transitioning from traditional batch synthesis to continuous flow led to significant enhancements, including a 95 % isolated yield under optimized conditions at 50 degrees C and a residence time of only 10 minutes. The optimized reaction recipes and conditions also enhanced the environmental footprint of the process, improving the overall GreenMotion score by 25 % and nearly doubling the 'Process' category score. Additionally, the study introduced a pH-induced crystallization method for purification, which streamlined the process and reduced resource intensity. The combined CAR and flow chemistry approach demonstrated enhanced flexibility and scalability, and reduced environmental impact, underlining its potential to transform API production through more holistic Green-byDesign strategies.

We present a shared industry-academic perspective on the principles and opportunities for Quality by Digital Design (QbDD) as a framework to accelerate medicines development and enable regulatory innovation for new medicines approvals. This approach exploits emerging capabilities in industrial digital technologies to achieve robust control strategies assuring product quality and patient safety whilst reducing development time/costs, improving research and development efficiency, embedding sustainability into new products and processes, and promoting supply chain resilience. Key QbDD drivers include the opportunity for new scientific understanding and advanced simulation and model-driven, automated experimental approaches. QbDD accelerates the identification and exploration of more robust design spaces. Opportunities to optimise multiple objectives emerge in route selection, manufacturability and sustainability whilst assuring product quality. Challenges to QbDD adoption include siloed data and information sources across development stages, gaps in predictive capabilities, and the current extensive reliance on empirical knowledge and judgement. These challenges can be addressed via QbDD workflows; model-driven experimental design to collect and structure findable, accessible, interoperable and reusable (FAIR) data; and chemistry, manufacturing and control ontologies for shareable and reusable knowledge. Additionally, improved product, process, and performance predictive tools must be developed and exploited to provide a holistic end-to-end development approach.

The synthesis of active pharmaceutical ingredients (APIs) is commonly perceived as more efficient when performed using continuous-flow methods, whereas batch processes are often seen as less favorable due to their limitations in yield, heat and mass transfer, and safety. This perception largely stems from existing studies that focus on green metrics such as the E-factor and yield. However, a comprehensive comparison of batch and flow processes through full techno-economic analyses (TEA) and life-cycle assessments (LCA) remains underexplored, leaving key aspects of their environmental and economic impacts inadequately assessed. This work addresses this gap by presenting a detailed comparison of batch and flow syntheses of seven industrially relevant APIs, including amitriptyline hydrochloride, tamoxifen, zolpidem, rufinamide, artesunate, ibuprofen, and phenibut. Eleven environmental impact categories within the framework of nine planetary boundaries were assessed, and the study also included an evaluation of capital and operating costs for both production methods. The results demonstrated that, on average, continuous-flow processes are significantly more sustainable with improvements in energy efficiency, water consumption, and waste reduction. Flow processes also show a marked reduction in carbon emissions and up to a 97% reduction in energy consumption, highlighting their potential for greener API manufacturing. Despite these advantages, the study identified areas where the continuous-flow technology requires further development. Specifically, manufacturing certain APIs in flow show lower-than-average improvements in operating expenditure and land system changes, the latter being directly correlated with the consumption of organic solvents, that can be comparable to or even higher than in batch. These challenges highlight the need for further optimization of flow processes to fully realize their potential in API production.

The development of effective synthetic pathways is critical in many industrial sectors. The growing adoption of flow chemistry has opened new opportunities for more cost-effective and environmentally friendly manufacturing technologies. However, the development of effective flow chemistry processes is still hampered by labor- and experiment-intensive methodologies and poor or suboptimal performance. In this context, integrating advanced machine learning strategies into chemical process optimization can significantly reduce experimental burdens and enhance overall efficiency. This paper demonstrates the capabilities of deep reinforcement learning (DRL) as an effective self-optimization strategy for imine synthesis in flow, a key building block in many compounds such as pharmaceuticals and heterocyclic products. A deep deterministic policy gradient (DDPG) agent was designed to iteratively interact with the environment, the flow reactor, and learn how to deliver optimal operating conditions. A mathematical model of the reactor was developed based on new experimental data to train the agent and evaluate alternative self-optimization strategies. To optimize the DDPG agent's training performance, different hyperparameter tuning methods were investigated and compared, including trial-and-error and Bayesian optimization. Most importantly, a novel adaptive dynamic hyperparameter tuning was implemented to further enhance the training performance and optimization outcome of the agent. The performance of the proposed DRL strategy was compared against state-of-the-art gradient-free methods, namely SnobFit and Nelder-Mead. Finally, the outcomes of the different self-optimization strategies were tested experimentally. It was shown that the proposed DDPG agent has superior performance compared to its self-optimization counterparts. It offered better tracking of the global solution and reduced the number of required experiments by approximately 50 and 75% compared to Nelder-Mead and SnobFit, respectively. These findings hold significant promise for the chemical engineering community, offering a robust, efficient, and sustainable approach to optimizing flow chemistry processes and paving the way for broader integration of data-driven methods in process design and operation.

This study explores the application of computer-aided retrosynthesis (CAR) for developing greener and more efficient synthetic routes for the active pharmaceutical ingredient (API) IM-204, a helicase primase inhibitor with potential against Herpes simplex virus (HSV) infections. Using various CAR tools, several total synthetic routes were identified, evaluated, and experimentally validated, with the goal to maximize selectivity and yield and minimize the environmental impact. The selected route achieved a significant improvement in the overall yield of IM-204 synthesis from 8% to 35% while also enhancing GreenMotion metrics from 0 to 18 overall and reducing the cost of building blocks by 300-fold. This work demonstrates the potential of CAR in drug development, highlighting its capacity to streamline synthesis processes, reduce environmental footprint, and lower production costs, thereby advancing the field towards more efficient and sustainable practices.

This paper exemplifies a Quality by Digital Design (QbDD) workflow for the crystallisation and isolation of active pharmaceutical ingredients (APIs). QbDD uses a digital first approach to improve manufacturability and sustainability whilst assuring product quality within practical constraints. This study uses three exemplar compounds (ibuprofen, lamivudine, and AZD0837), each of which presents a different challenge for crystallisation; these include agglomeration, solid state form, and slow growth rates, respectively. These cases are used to evaluate the benefits of the QbDD approach and identify gaps for future research. Results of this work show that the QbDD workflow reduces the number of physical experiments by 28% and the API material usage by 52-65% when compared to comparable API development processes not using this approach. This approach provides a route to practically implement and exploit the benefits of digital tools and overcome digital skill shortages. By exploiting digital tools for process simulation and optimisation, the workflow improves efficiency, even in complex cases where multiple workflow iterations are required. This workflow, therefore, paves the way for more sustainable and cost-effective API production and it promotes future standardisation of digital design in pharmaceutical development.

One of the most important challenges in the pharmaceutical industry is to produce crystals with desired size and shape distributions, to enhance the critical quality attributes of the drug product, such as efficacy, and to improve manufacturability during downstream processing, such as filtration, drying and granulation. The paper provides a framework for effective crystal shape and size tuning, based on a systematic exploration of standard techniques, such as the linear cooling and supersaturation control (SSC), and novel methods based on the systematic combination of several techniques, namely direct nucleation control (DNC), wet milling, SSC and shape modification additives. The crystallization of lovastatin, which is notorious for its challenging needle-shaped crystals, with an extremely high aspect ratio, was used as a case study, and polypropylene glycol (PPG-4000), at different concentrations, was used as an effective shape modifier from small-scale tests studied previously. The proposed techniques were implemented in the case of seeded and unseeded systems. It was demonstrated that the combination of temperature cycling and polymer additive enhances greatly the control over the aspect ratio and crystal size distribution, compared to conventional linear cooling and SSC strategies. The implementation of wet milling at the beginning of the process, or the introduction of seeds, enhances even further the control of the critical quality attributes of the crystalline product.