Artificial intelligence (AI) is taking on the world to be the answer to everything. With its considerable noise, I thought it would be interesting to understand its knowledge about chemistry, equipment, process centricities, process development and design methodologies. This note is not pouring water on every chemist and/or chemical engineer’s efforts past, present and future or an effort that is being made but sharing what our band of chemists and chemical engineers did and still practice. “Hats off to all of my mentors and colleagues as we had the opportunity to explore and exploit our imagination and creativity”. Only by review we can understand and/or improve what is being attempted. Perspective presented is not influenced by “for or non-profit” organizations.
Since “Artificial Intelligence (AI)” is being talked about as the “go to” for all the answers, I thought it would be just wonderful to get its perspective and opinion about the three (chemistry, equipment and process) centricities and also it would be interesting to understand its views on chemistry, equipment and process centricities about process development and design methodologies. I am comparing AI’s perspective to my learnings and experiences.
Initial responses from two sites (not referenced) were long (more than three pages). I was not looking for a treatise or a dissertation but simple answers. By re-asking the question I received the following truncated responses. Longer responses might be acceptable to many but most trained chemists and chemical engineers would find them long and daunting. At least I did. I re-asked each site to simplify their responses and they are follows.
Centricities to me describe different aspects of process design and manufacture of a chemical product. Each prioritizes a different phase of the overall process. Site “A” compared to Site “B” had some details.
Site “A”:
· Chemical Centricity: The primary focus is on the molecule or formulation itself and its desired function or performance, often driven by a chemist in the lab. The goal is to identify a specific chemical or mixture that satisfies complex customer requirements, with less initial emphasis on the manufacturing process or equipment.
· Equipment Centricity: The approach is driven by the existing or available equipment/hardware. The product and process must be designed to work within the constraints and capabilities of the current machinery and facilities, which can be common in contract manufacturing or leveraging existing infrastructure.
· Process Centricity: The emphasis is on developing the most efficient, safe, and cost-effective manufacturing process to produce high volumes of a given chemical. This is typical for large-scale commodity chemicals where process optimization (yield, cost, safety, sustainability) is the key to profitability, and the product itself is already well-defined.
Site “B”:
· Chemistry-centric: Focus on reactions & molecules.
· Equipment-centric: Focus on machines & tools.
· Process-centric: Focus on the whole flow & control.
Each site itemizes university teachings and suggests consideration of chemistry recipe, equipment hardware and process steps individually and sequentially to create an economic, safe process that complies with environmental regulations. There is nothing wrong about the methods taught in our universities. They are science based. Each site has excellent capability to scout the landscape and what has been done but does not have the capability to differentiate between a complex or simple economic process. It just refers them.
Published literature and patents referenced by the two sites are references the published lab work/curiosity and most likely will never be commercialized. This observation is based on synthesis routes published for the products that are/were commercial before the articles were published. Logical question is why was the work not done before the products were commercialized. Seems significant dollars were wasted on the work of no value.
I was banking on AI sites to reveal something very different and enlightening about the chemistries and processes I was exploring but THAT did not happen. I guess I was expecting too much. May be AI sites are not ready to advise on process development options/routes etc. at least for now. I don’t think AI platform will be ready for the near future. May be tomorrow but as we knowTomorrow never comes. It might come tomorrow.
Methods Used and Some Examples:
We have to recognize that in fine/specialty chemical hemisphere there are two class of chemicals. One that “improves lifestyle” and one “extends life”. The later class are called pharmaceuticals. Their molecules, their chemistries and their manufacturing practices and methods are same as “improve lifestyle chemicals”. The ONLY difference is product quality and use. They are produced under strict guidelines. They are covered under FDA’s cGMP guidelines (2). Pharmaceutical quality cannot be compromised and has to be repeatable for time immemorable. One subtle fact we overlook is that the manufacturing processes of products that improve lifestyle can be continuously improved whereas the processes for pharma products generally require FDA reapproval (3), an expensive process. Following examples illustrate how and what we did and would do to create excellent, economic and environmentally friendly processes from the onset. They can be used by anyone.
My Experiences and Perspectives:
We did not call classify our product and process development methods in centricities as we had ONE TASK and it was to create and commercialize the best economic process. Yes we did start with round bottom flask with its agitator, piloted the process but simultaneously worked on what it will take for a commercial process design.
For an excellent, economic and environmentally friendly chemical process and blending operations there are some basic guidelines and rules that every chemical engineers and chemists follows. They have been established since principles of chemistry and chemical engineering were recognized almost a century ago. These was instilled in me in my sophomore year of our chemical engineering curriculum. Initially I did not understand them but went along as everyone else had. Later, when I started practicing them I realized their value.
Our teachers taught us fundamentals of chemistry and chemical reactions, reaction kinetics, thermodynamics, unit operations (4)(mixing, heat exchange, filtration, distillation, pumping, crystallization etc.], unit processes (5) [amination, nitration, hydrogenation, reduction, sulfation, chlorination etc.. McGraw Hills and other publishers have published series of books (6, 7) on the subject and they were and still are a godsend.
They also taught us all about physical and chemical properties (8). It is expected that we understand and figure out their mutual interaction, how to manipulate and marry them to create excellent and economic processes. We were/are exposed to all these but have no hands on experience. Thus, we carry our teachings as a myth. These parables become a reality when we get involved in process development, commercialization and manufacture of, petrochemicals, fine specialty chemicals, flavors and fragrances, additives, coatings and pharmaceuticals. It has been an exhilarating experience.
Laboratory process development starts with lab supply houses. There is a difference between lab processes and commercial processes. As soon as chemistry seemed feasible we switched to commercially available materials. This allowed us to build supply chain relationships and familiarity with their material. This influences commercial processes. Imagination and creativity in creating excellent processes is impeded if we do not work from the onset with commercially available materials. In our efforts we did not realize that we were unknowingly following some principles mentioned in “The Art of War” (9) in our effort to commercialize unique processes.
Our methods were/are a “GEMISH” of all centricities suggested by AI sites applied simultaneously. They worked well. In our product, process development and manufacturing Village (chemists, chemical engineers, accounting, manufacturing, quality control and purchasing) would be involved as soon as the synthesis chemistry of the chemical was written (10, 11, 12). That meant lab experiments, raw material sourcing, pilot plant testing and commercial process. Our mission always was how along with unit processes, unit operations, reaction kinetics, thermodynamics etc. we can apply physical properties, synthesis steps and process equipment (i.e. Sociochemicology) (10,11,12, 13,14, 15) to commercialize a product.
All of the above [chemistry, equipment and process] have been and are the three centricities that simplify and make manufacturing processes very economical and environmentally friendly and are THE fundamentals of chemical process design. How they are applied is company’s choice.
Examples:
Through the following examples I am sharing my experiences and perspectives about the synthesis of certain fine/specialty chemicals and active pharmaceutical ingredients (API). Some of the processes have been commercial for over sixty years. They still can be improved. Discussion can also be applied to products being developed and/or are commercial. For the existing products FDA approval (3) might be needed. Imagination, Creativity and Fundamentals of Chemical Engineering and Chemistry have to be exploited and applied.
Each written reaction stoichiometry tells us its reaction mechanism. It is up to us to understand it, test it and execute it commercially as intended. We prove the reaction feasibility in a round bottom flask and/or alike equipment. However, we do not or are not able commercially execute it. This is due to forms of raw materials available, tradition and the equipment used. In most of the cases the lab chemistry cannot be scaled up to have an economic commercial process. Effort is needed.
Lab processes using available raw materials are modified to fit in the existing and/or new equipment. This is an age old practice. Other options like modular plants are a possibility to have a hand and glove process. However, companies are have not considered and/or decided this route due to many reasons e.g. trained and experienced personnel and capital investment. Most commercial processes are not designed for the chemistry. Chemistries are designed and fitted in the equipment. Unless a judicious attempt is made, we might not able to capitalize on the unit processes and sociochemicology (5, 6, 7, 9, 10,11, 12, 13,14, 15)of the reactants in a commercial process. Process can be simplified if companies include something similar to the Village (10,11, 12)mentioned earlier. A concerted effort has to be made to incorporate them in the commercial process. Imagination and creativity is needed. Following are few examples.
Diazotization:
Aromatic amines used to produce respective diazo compound generally use hydrochloric acid and sodium nitrite. It is a very straight forward reaction as illustrated. “R” is the desired hydrocarbon amine.
RNH2 + 2 HCl + NaNO2 ----> RN2Cl + 2 H2O + NaCl (1) |
It is well known that diazo formation is a two-step reaction. As illustrated hydrochloride produced from amine reaction with hydrochloric acid reacts with sodium nitrite to produce the diazonium salt that is used for further reaction. (2,3)
RNH2 + 2 HCl ----> RNH2Cl (2)
RNH2HCl + HCl + NaNO2 ---> RN2Cl + 2 H2O +NaCl (3)
Due to exothermic diazo reactions it is recommended they be generally conducted at/or below zero degree centigrade temperatures. Solubility of amine hydrochlorides chlorides in water is well known. This can be used to advantage. Since the reactions are exothermic, value of heat of reaction can also be capitalized on to speed up the reaction. Proper precautions have to be taken. By capitalizing on solubility of amine hydrochlorides and further reaction with sodium nitrite produces the respective diazonium salt in water that is water soluble. This allows the reactions to be easily done at room or higher temperatures.
It is not done as it cannot be easily demonstrated in the laboratory. Concept of hydrochloride formation followed by diazonium salt can be easily done in a properly designed process. Higher temperature speeds the reaction rate and this can be used to advantage. Imagination and creativity along with proper process equipment design makes this feasible. Since diazonium salts are water soluble this can eliminate use of solvents that is suggested in most of the published information. A water based continuous commercial process has been operating since 1970 (15).
Diazo chemistry has been commercial for over ONE hundred years and the solubility of its intermediates in water is well known. In many of the published papers, use of very expensive flow reactors and as if flow chemistries, (is it a new science as no reaction takes place without flow?) along with solvent use is suggested as a way for continuous diazo process. They are not at all necessary as commercially available equipment costing about 10-25% of the flow reactors is commercially available. Chemical engineers have not recognized value of the existing equipment. Since they cannot be tested in the lab process, value of such equipment has been ignored.
We have to acknowledge that laboratories prove reaction feasibility but imagination and creativity of chemical engineers and chemists i.e. the experience and knowledge of the Village (10,11, 12) needs to be harnessed. That comes from familiarity with processing equipment and process chemistry and is extremely helpful. Experience of having a continuous diazo process i.e. practice of where and how reactants were added has been successfully tested in production of acrylic latex polymerization.
Synthesis of Phthalimide was documented since 1919 (16). Imagination and creativity led to different manufacturing processes.
Phthalic Anhydride + Ammonia -----> Phthalimide + H2O (4)
The described process (16) suggests adding ammonia gas from cylinders in molten phthalic anhydride (PA) in cast iron pans heated to about 240 °C. It is presumed that the pans had no lids. Product was cooled, crushed and ground. Other references suggest different ways to produce phthalimide e.g. reacting phthalic anhydride or phthalic acid with urea (17), using a bubble cap tray column with top feed of liquid phthalic anhydride and bottom feed of ammonia (18), feeding liquid ammonia in molten phthalic anhydride (19) in an agitated tank to produce phthalimide in batch operation.
Imagination, creativity and sociochemicology (1,13) was applied to continuously produce phthalimide by simultaneously feeding molten phthalic anhydride and liquid ammonia (around 1965) in stochiometric control to a properly designed small reactor in a reactor. In this process as liquid ammonia vaporizes in molten phthalic anhydride it agitates the melt to produce phthalimide that overflows in the larger reactor. Heat of reaction and liquid ammonia, as it vaporizes, assures complete mixing, a very ingenious design. No solvent is used. Produced water vapor is easily quenched.
Learning can be translated to other design applications. In this process designers exploited and capitalized physical properties and state of the raw materials to commercialize a continuous process. Simplicity of the process allowed it to meet variable product demands. It is interesting that it took about 50 years to have a simple solventless commercial process.
Metformin Hydrochloride:
Metformin hydrochloride, used for diabetics, is a very high volume product. Its global demand is expected to reach about 156 thousand metric tons by 2035 (20). Its chemistry and chemical structure suggests dicyandiamide, dimethyl amine and hydrochloric acid are reacted to produce 1,1-dimethylbiguanide hydrochloride i.e. metformin (12).
Dimethyl amine + Dicyandiamide + Hydrochloric acid ----> Metformin Hydrochloride (5)
Different commercial processes (12) use solvent. In one of the routes use of wiped film evaporator is used to produce the API. In the other alternate routes Oslo type crystallizer (21) or other crystallizer use is suggested to produce the API Fig. 1.
Figure 1: Metformin Hydrochloride Manufacture
A solventless chemistry (Figure 2) similar to phthalimide process discussed above is very possible for metformin hydrochloride production where molten dicyanamide and liquid dimethyl amine in stoichiometric ratio are fed to a reactor to produce liquid metformin. Liquid metformin would then be converted to metformin hydrochloride by quenching in commercial hydrochloric acid (37%). Quench would be further processed either using an Oslo type crystallizer or a wiped film evaporator. Direct acid quench route similar to phthalimide process discussed earlier can be easily designed and commercialized. It has not been tested.
Dicyandiamide + Dimethyl amine Metformin + Hydrochloric acid Metformin hydrochloride | ||||
Dicyandiamide | Dimethyl amine | Metformin | Hydrochloric acid | Metformin hydrochloride |
MW 84 | MW 45 | MW 129 | MW 36.5 | 165.6 |
MP °C 208 - 211 | BP°C 7-9 | MP°C 205.07 | BP°C 48 | MP°C MW 165.5 |
CAS 461-58-5 | CAS 124-40-3 | CAS 657-24-9 | CAS 7647-01-0 | CAS 1115-70-4 |
Figure 2: Solventless Metformin process
Compared to the existing processes this process has major advantage. Commercial grade raw materials are used. There is no solvent use or any recycling of materials is involved. Process designers will apply existing chemical engineering fundamentals and process control technology for the process. Based on the commercial success of phthalimide process metformin process would be a commercial success. FDA should not have to approval issues if the process is diligently designed. This product might need bioequivalence testing.
Dicyandiamide + Dimethyl amine ------> Metformin + Hydrochloric acid ------> Metformin Hydrochloride
Figure 3: Metformin hydrochloride Process schematic
Commercial chemistries use solvents. Chemistry outlined in Figure 2 & 3 is a solventless process. It is an extremely simple process and the product quality will depend on the process designer’s imagination and creativity to have an excellent process.
Described processes suggest that if we understand the chemistry of any reaction, understand the physical and chemical properties of the chemicals used and produced, can harness the equipment available used in any industry and apply our imagination and creativity, excellent processes can be commercialized. We can reduce the amount of solvents used and improve productivity of the processes thereby lowering costs. Since the annual volume of most of the APIs is not high cross fertilization and use of modular plants can be very helpful and needs to be considered. For that each company might have to review its staffing.
If process developers are constrained by of the available processing equipment, it is very likely their imagination and creativity is/will also be stifled. We have all the skills and knowledge to accomplish what is necessary. Onus is on us to exploit our learnings and have to figure out how to get there and it is not an easy task. Unless we are challenged, process developers/designers basically are comfortable with following established traditions and by fitting the processes in commercially available equipment following what we consider is easily doable.
AI in future might be able to make some humanoid suggestions but the responsibility still will be on us humans to understand the recommendations, test and implement them to simplify manufacturing. Human imagination and creativity still will be paramount.
Girish Malhotra, PE
EPCOT International
References:
1. Malhotra, Girish: Sociochemicology: Redefining Chemical Process Design for Efficiency and Sustainability, Profitability through Simplicity, February 26, 2025
2. FDA cGMP
3. FDA: Development & Approval Process
4. McCabe W. L & Smith J. M. Unit Operations of Chemical Engineering McGraw-Hill Book Company Second Edition 1967
5. Shreve, R. N. Unit Processes in Chemical Engineering, Industrial and Engineering Chemistry, 1956
6. McGraw Hill Chemical engineering Series
7. Perry's Chemical Engineers' Handbook, 9th Edition
8. Malhotra, Girish: Focus on Physical Properties To Improve Processes: Chemical Engineering, Vol. 119 No. 4 April 2012, pgs. 63-66
9. Sun Tzu: The Art of War, Simon and Schuster
10. Malhotra, Girish: Chemical Process Simplification: Improving Productivity and Sustainability, John Wiley & Sons, February 2011
11. Malhotra, Girish: Chapter 4: Simplified Process Development and Commercialization in Quality by Design—Putting Theory into Practice, co-published by Parenteral Drug Association and DHI Publishing, February 2011,
12. Malhotra, Girish: Active Pharmaceutical Ingredient Manufacturing: Nondestructive Creation, De Gruyter, April 2022
13. Malhotra, Girish: Sociochemicology, May 30, 2013
14. Malhotra, Girish: Profitability through Simplicity
15. Malhotra, Girish: Capitalizing on Mutual Behavior and Chemical Reactivity of Chemicals, Profitability through Simplicity,May 29, 2023
16. Cain, John C. The manufacture of dyes, 1922
17. Boehme W. R. et. al. Production of Phthalimide USP 3,819,648 Dawe’s Laboratories.
18. Hetzel E. et. al. Continuous Manufacture of Phthalimide USP 4,001,273 BASF Aktiengesellschaft
19. Schlaudecker G. F. Process for the Production of Phthalimide USP 2,668,326 Maumee Development Company
20. Metformin HCL Market - Global Industry Size, Share, Trends, Opportunity, and Forecast, 2020-2035
