Neuro-Symbolic Artificial Intelligence for Carbon-Constrained Chemical Process Synthesis: A Review of Advances, Challenges, and Future Directions
Keywords:
Neuro-symbolic , artificial intelligence, Carbon-constrained process optimization, Sustainable chemical engineeringAbstract
In recent years, the application of artificial intelligence (AI) in chemical process engineering has gained momentum, driven by the rising demand for sustainable transformation in the industry, especially in the field of carbon constrained-constrained process optimization and low emission-constrained-emission industrial processes. The review delves into the evolving importance of the role of neuro-symbolic artificial intelligence (AI) in enhancing sustainable chemical processes by combining data-driven learning and first-principles scientific reasoning. The study investigates carbon constrained-emission-constrained optimization methods such as energy minimization strategies, emission-aware process synthesis methods, hybrid thermodynamic–AI modeling, and multi-objective optimization methods,-constrainedmethods, which are instrumental for net-zero industrial transitions. Additionally, the review addresses various uses of neuro-symbolic AI for reaction pathway optimization, catalysis design, intelligent process monitoring, process fault diagnosis, and emission reduction strategies for industrial processes. A conceptual model is presented that describes the relationship between process data, neural architectures, symbolic reasoning engines, thermodynamic constraints constraints,and optimized low carbonconstraints,-carbon outputs. The review also sheds light on the increasing significance of explainable AI systems that preserve physical consistency while enhancing the predictive capabilities in complex industrial setups. However, the current state of progress comes with a number of drawbacks, such as computational requirements, limited application in the industry, availability of data, the absence of a uniform set of carbon-aware AI frameworks, and difficulties in integrating with existing process infrastructure. Overall, the study shows that this neuro-symbolic AI is a promising direction in the quest for intelligent, explainable, and sustainable chemical manufacturing systems to help meet the future decarbonization targets and circular industrial targets.
References
Akmal M, Foo DCY, Ooi J, Lim CH. A novel hybrid method combining mixed-integer linear programming and pinch analysis for technology selection. Process Integr Optim Sustain. 2026. doi:10.1007/s41660-025-00661-1.
Biegler LT. Multi-level optimization strategies for large-scale nonlinear process systems. Comput Chem Eng. 2024;185:108657. doi:10.1016/j.compchemeng.2024.108657.
Biswas S. Artificial intelligence in chemistry for future sustainability: An in-depth scientific review. Int J Innov Sci Res Technol. 2026:712. doi:10.38124/ijisrt/26mar158.
Chew JW, Andersson R, Bierweiler T, Ryttestal P, Wik T. AI in chemical engineering: From promise to practice. AIChE J. 2026. doi:10.1002/aic.70358.
Chowdhury SH, Ghosh A, Acharya S, Saha SC, Pal PK, Roy S, et al. Transforming chemical process engineering: The role of AI and machine learning in revolutionizing process systems. Can J Chem Eng. 2025;104(2):699-731. doi:10.1002/cjce.70032.
Decardi-Nelson B, Alshehri AS, Ajagekar A, You F. Generative AI and process systems engineering: The next frontier. Comput Chem Eng. 2024;187:108723. doi:10.1016/j.compchemeng.2024.108723.
Fózer D, Nimmegeers P, Tóth AJ, Varbanov PS, Klemeš JJ, Mizsey P, et al. Hybrid prediction-driven high-throughput sustainability screening for advancing waste-to-dimethyl ether valorization. Environ Sci Technol. 2023;57(36):13449-13462. doi:10.1021/acs.est.3c01892.
Galitsky B, Rybalov A. Neuro-symbolic verification for preventing LLM hallucinations in process control. Processes. 2026;14(2):322. doi:10.3390/pr14020322.
Hajireza K, Eklund D, Andersson R. Enhancing generalization and training efficiency of neural ordinary differential equations for chemical reactor modeling. Chem Eng J. 2025;519:165183. doi:10.1016/j.cej.2025.165183.
Ji W, Deng S. Autonomous discovery of unknown reaction pathways from data by chemical reaction neural network. J Phys Chem A. 2021;125(4):1082-1092. doi:10.1021/acs.jpca.0c09316.
Kiss AA, Grievink J. Perspectives on the essential role of first-principles modeling in the age of AI. Ind Eng Chem Res. 2026. doi:10.1021/acs.iecr.5c04156.
Kumar D, Dixit V, Ramteke M, Kodamana H. Learning system physics using symbolic neural integration (SyNISM) with applications to chemical processes. Ind Eng Chem Res. 2025;64(23):11441-11458. doi:10.1021/acs.iecr.5c00107.
Lye YX, Chew YE, Foo DCY, How BS, Andiappan V. Carbon emission reduction strategy planning and scheduling for transitioning process plants towards net-zero emissions. RSC Sustain. 2024;3(2):929-945. doi:10.1039/d4su00636d.
Mann V, Lu J, Venkatasubramanian V, Gani R. A perspective on artificial intelligence for process manufacturing. Engineering. 2025;52:60-67. doi:10.1016/j.eng.2025.01.014.
N KR. Decarbonizing industrial operations: A multi-objective optimization approach for balancing energy efficiency and production throughput in carbon-constrained environments. Zenodo. 2026. doi:10.5281/zenodo.18710893.
Neveux T, Commenge J, Vermeire FH. Editorial: Artificial intelligence-assisted design of sustainable processes. Front Chem Eng. 2024;6. doi:10.3389/fceng.2024.1391673.
Perumal D, Vaithiyalingam M, YS P, Dharanitharan K, Dinesh B, Kumar MS, et al. Neuro-symbolic AI in drug discovery. In: Advances in Computational Intelligence and Robotics. Hershey (PA): IGI Global; 2026. p. 307-332. doi:10.4018/979-8-3373-6400-1.ch011.
Pistikopoulos EN, Tian Y. Advanced modeling and optimization strategies for process synthesis. Annu Rev Chem Biomol Eng. 2024;15(1):81-103. doi:10.1146/annurev-chembioeng-100522-112139.
Rupprecht S, Gao Q, Karia T, Schweidtmann AM. Multi-agent systems for chemical engineering: A review and perspective. Curr Opin Chem Eng. 2025;51:101209. doi:10.1016/j.coche.2025.101209.
Sachin J. Design of sustainable chemical processes and supply chains via mathematical programming and hybrid modeling [dissertation]. Zürich: ETH Zürich Research Collection; 2025. doi:10.3929/ethz-c-000785582.
Schweidtmann AM, Esche E, Fischer A, Kloft M, Repke J, Säger S, et al. Machine learning in chemical engineering: A perspective. Chem Ing Tech. 2021;93(12):2029-2039. doi:10.1002/cite.202100083.
Shyalika C, Sharma A, Kalach FE, Jaimini U, Henson C, Harik R, et al. CausalTrace: A neurosymbolic causal analysis agent for smart manufacturing. Proc AAAI Conf Artif Intell. 2026;40(47):40432-40439. doi:10.1609/aaai.v40i47.41486.
Sivaram A, Venkatasubramanian V. XAI-MEG: Combining symbolic AI and machine learning to generate first-principles models and causal explanations. AIChE J. 2022;68(6). doi:10.1002/aic.17687.
Srinivasan KK, Puliyanda A, Thosar D, Bhakte A, Singh K, Addo P, et al. Artificial intelligence and machine learning at various stages and scales of process systems engineering. Can J Chem Eng. 2024;103(3):1004-1035. doi:10.1002/cjce.25525.
Sun X, Durkin A, Guo M. Life cycle optimisation tool development for process systems and centralised supply chain design. RSC Sustain. 2023;1(9):2224-2240. doi:10.1039/d3su00251a.
Tayane S, Gaber J. SymChemAI: A symbolically informed neural network for physicochemical modeling of reaction systems. ChemRxiv. 2025. doi:10.26434/chemrxiv-2025-2hp4c-v2.
Thon C, Finke B, Kwade A, Schilde C. Artificial intelligence in process engineering. Adv Intell Syst. 2021;3(6). doi:10.1002/aisy.202000261.
Venkatasubramanian V, Chakraborty A. Quo vadis ChatGPT? From large language models to large knowledge models. Comput Chem Eng. 2024;192:108895. doi:10.1016/j.compchemeng.2024.108895.
Venkatasubramanian V, Chakraborty A. Quo vadis ChatGPT? From large language models to large knowledge models. Comput Chem Eng. 2024;192:108895. doi:10.1016/j.compchemeng.2024.108895.
Weber JM, Guo Z, Zhang C, Schweidtmann AM, Lapkin AA. Chemical data intelligence for sustainable chemistry. Chem Soc Rev. 2021;50(21):12013-12036. doi:10.1039/d1cs00477h.
Wehinger GD, Ambrosetti M, Cheula R, Ding Z, Isoz M, Kreitz B, et al. Quo vadis multiscale modeling in reaction engineering? A perspective. Process Saf Environ Prot. 2022;184:39-58. doi:10.1016/j.cherd.2022.05.030.
Yin M, Wang Y, Yu W, Wilson D, Ma X. Industrial processes fault diagnosis method based on expert system-guided neural network decision-space sparsification. IEEE Trans Syst Man Cybern Syst. 2026:1-13. doi:10.1109/TSMC.2026.3673583.
