Disclaimer

All opinions are my own.

Saturday, April 20, 2019

Considerations to Simplify Organic Molecule (API) Manufacturing Processes: My perspective

Organic chemical synthesis is a pathway chemists take/suggest to synthesize different organic chemicals. These are generally multi-step processes and are a road map of their thinking. Their processes are based on the laboratory equipment available to them. In the development of synthesis process many times tradition prevails. Chemical engineers use the laboratory road map to develop and commercialize an optimized process. 

I have randomly selected three patents and reviewed them. This should not be considered as criticism or questioning anyone’s knowledge. My discussion is based on what all I would consider and experiment to simplify and economize these processes. If anyone looks at the chemistries, they are like any other organic chemical but these create a product that cure illnesses and are called active pharmaceutical ingredient. 

Generally majority of these processes have to be reconfigured and economized before they can be commercialized. In certain cases, simplification can be a complex task and even can be out of reach. Still effort has to be made. 

There are certain process and design considerations that are necessary for an economic process. I mention few of these considerations but others might have additional considerations. Basic underlying thinking, at least mine, is how a process can be made economic, efficient and least burdensome on the environment. 

I have used the following considerations (1, 2) to simplify and commercialize chemical processes.
  1. How the chemical and physical properties can be exploited? 
  2. How the solvent use be minimized and/or eliminated? It impacts process productivity?
  3. Can the solvent use be limited to one in addition to water?
  4. How the conversion of each reaction step can be improved by applying chemical kinetics fundamentals?
Instead of reviewing each patent in totality, few steps are discussed to convey the thought process of how to simplify the whole process. 

PROCESS AND NOVEL POLYMORPHIC FORM OF VORTIOXETINE AND ITS PHARMACEUTICALLY ACCEPTABLE SALTS US 2019/0002421 A1, Applicant Apicore US LLC, Somerset, NJ

Step 1: Preparation of 1-(2-nitrophenyl) piperazine 
To a solution of piperazine (224.0 gm) dissolved in methanol (600.0 ml) under an atmosphere of nitrogen, 1-chloro-2-nitrobenzene (100.0 gm) was slowly added while the temperature was maintained at 30.degree. C. The temperature of the reaction mass was raised to 70.degree. C. (70-100.degree. C.) and the contents maintained at this temperature for 24 hours (20-24 hours). After the reaction was complete the reaction solution was distilled under vacuum below 50.degree. C. Charged water (600 ml) and toluene (600 ml) into the crude. The pH of the reaction mass was adjusted to below 1.0 with hydrochloric acid. The organic layer and aqueous layer were separated. The aqueous layer was taken and the reaction mass pH was adjusted to 7.0-8.0 with sodium carbonate solution. The obtained aqueous layers were extracted with dichloromethane (300 ml and 150 ml), and the organic layer dried with sodium sulphate. The organic layers were collected and concentrated under reduced pressure at 45.degree. C. to furnish the title compound (120.0 gm) as brown viscous oil which was used for the next step without further purification. 

            
                                                              Reaction Step 1 Chemistry

Step 1 Analysis:

Table 1 lists each raw material. 1-chloro-2-nitrobenzene is considered the key staring material and in my analysis is used as the basis of the reaction stoichiometry. Based on reaction mass balance one mole of piperazine would be theoretically required for the reaction. However, in the described reaction more than four moles of piperazine is being used. This is excessive. If it is not recovered, would be a waste and will add additional burden on the effluent system. Such an excess also increases the manufacturing cost. Since the reactants and the product from this reaction step have high boiling points, reaction should be tried at the highest possible temperature to increase the reaction rate i.e. reduce the process batch time. I would also consider an environmentally and reaction friendly high boiling point solvent.

At 100% conversion one mole of 1-(2-nitrophenyl) piperazine (207 grams) would be produced. However, based on patent description only 120 grams of the product is produced. This is about ~58% yield. Three different solvents are being used. I consider them to be excessive and would make every effort to economize. 

I have not explored the chemistry in the laboratory but based on my experiences, I would consider using a single solvent that will work for each reaction step. Every step has to be re-configured to maximize yield, reduce the reaction time and improve process productivity. Use of excessive amount of low boiling solvent (methanol ~64.7 oC) not only lowers the productivity but also prevents speedier reaction i.e. lower batch cycle time, as has been mentioned earlier. 

Step 2: Preparation of 1-benzyl-4-(2-nitrophenyl) piperazine 
1-(2-nitrophenyl) piperazine (100.0 gm) was dissolved in DMF (200.0 ml) and toluene (500 ml) at room temperature. The contents were cooled to 0.degree. C. followed by addition of potassium carbonate (160.0 gm) and benzyl chloride (73.0 gm) while maintaining the temperature at 0.degree. C. The temperature was raised to 100.degree. C. (100-120.degree. C.) and the reaction mixture maintained at this temperature for 2 hours (2-4 hours). On completion of the reaction (monitored by TLC), the contents were brought to room temperature and slowly poured into ice water (1000 ml) with vigorous stirring. The organic layer was washed successively with water (2.times.400 ml), brine (400 ml), dried over anhydrous sodium sulphate and solvent removed under reduced pressure to furnish the title compound (140 gm) as thick orange oil. The product was used for the next step without further purification. 


Reaction Step 2 Chemistry 

Step 2 Analysis:

In this step again, large quantities of solvent are being used. Reactants are less than 20% of the reactant mass. This is not an effective use of reactor. Yield is poor also. I would go back to the bench and create the best conditions to improve productivity and yield. Potassium carbonate is being used in situ to neutralize the generated HCl. In a commercial process HCl can be neutralized using a caustic scrubber. I would also minimize benzyl chloride use e.g. to 5% excess.

If the laboratory process is commercialized, my conclusion is more money will be spent in remediation of the effluent rather than earned from the process. Reaction yield of the two-process step is dismal ~27%, extremely poor for any organic synthesis. 

My conjecture is that if the remaining steps are also carried out as outlined in the patent, and if each has a 90% yield per step, overall process would be less than 20%, an extremely poor and an uneconomic process. I would review each step to determine how the yield per step can be improved to better than 95% for a respectable overall yield. 

Apicore LLC
USP 2019/0002421
First two steps for Vortioxetine process

Step 1



Step 2


Chemical name
1-Chloro-2-nitrobenzene
Piperazine
1-(2-Nitrophenyl) piperazine

1- (2-Nitrophenyl) piperazine
Benzyl Chloride
1-Benzyl-4-(2-nitrophenyl) piperazine
CAS #
88-73-3
110-85-0
59084-06-9
100-44-7
199105-16-3
Chemical formula
C6H4ClNO2
C4H10N2
C10H13N3O2

C10H13N3O2
C7H7Cl
C17H19N3O2
Mol. Wt. 
157.5
86
207

207
126.5
297
MP, o C
31
108
131




BP, C
246
146
392


179
441
Grams
100
224


100
73

Moles
0.63
2.60


0.48
0.58

Moles/ mole 1-Chloro-2-nitrobenzene
1.00
4.10

Moles/ mole 1-(2-Nitrophenyl) piperazine
1.00
1.19

Theoretical yield, gm


207.00



297
Actual exp. Yield


120.00



140
% yield


57.97%



47.14%
Table 1: Stoichiometry of First two steps Vortioxetine USP 2019/0002421

Process for the Preparation of Vortioxetine: US 10,227,317 Assigned to Lupin Limited, Mumbai (IN)

Step 1: Preparation of S-(2,4-dimethylphenyl) ethanethioate


Reaction Step 1 Chemistry 

100 g of 2,4 dimethyl thiophenol was suspended in 500 ml of dichloromethane and added 94 gm of pyridine at 25.degree. C. Reaction mixture was cooled to 5-10.degree. C. and then drop wise added acetyl chloride (64 gm, 0.815 mole). Resulting reaction mixture was heated at 5-15.degree. C. for 1 hr. RM was monitored by HPLC. When reaction was completed cooled to RT, added water and stirred for 10-20 mins. Desired product was extracted with ethyl acetate twice. Combined organic layer was washed with water, brine, dried over sodium sulfate and concentrated under reduced pressure to give oily mass as desired product (130 gm, 100%). 

Analysis Step 1:

Based on my training, using industrial stoichiometry fundamentals, the yield of ethanethioate is about 72% rather than claimed 100%. I am not sure of the basis for 100% yield. In addition, I would consider using a higher boiling solvent to speed the reaction. Generated HCl can be scrubbed and acetyl chloride condensed back to the reactor (3). In addition, I would consider significant reduction of amount of solvent used. I even question the use any solvent. Since the reaction product is a liquid, possibilities to lower or eliminate solvent do exist and should be explored. 

Step 2: Preparation of (2-bromophenyl)(2,4-dimethylphenyl) sulfane 


Reaction Step 2 Chemistry 

100 g of S-(2,4-dimethylphenyl) ethanethioate was suspended in 500 ml of degassed toluene. Then added Pd(dba).sub.2 (3.25 gm, 0.006 mole), rac-BINAP (6.4 gm, 0.01 mole), and sodium tert-butoxide (62.5 g, 0.65 mole) under nitrogen purging. Reaction mixture was heated to 45-55.degree. C. and then added 2-iodo-1-bromo benzene (165 gm, 0.583 mole). Reaction mixture was stirred at 90-100.degree. C. for 3 hrs. under nitrogen atmosphere. After completion of the reaction, it was cooled to room temperature and added 400 ml of water, and allowed to stir for 15-20 mins. Reaction mixture was allowed to settled and organic layer was separated and concentrated under reduced pressure to give thick viscous oil to give desired product (2-bromophenyl)(2,4-dimethylphenyl) sulfane (162 gm, 100%). 

Lupin Limited       USP 10,227,317
First two steps for Vortioxetine
Step 1
Step 2
Chemical name
2-4 dimethyl thiophenol
Acetyl Chloride
S-(2,4-Dimethylpheny) ethanethioate

S-(2,4-Dimethylphenyl) ethanethioate
2-iodo-1-Bromobenzne
(2-Bromophenyl)(2,4-dimethylphenyl)sulfane
CAS #
13616-82-5
75-35-6
94602-20-7
94602-20-7
583-55-1
960203-41-2
Chemical formula
C8H10S
C4H10N2
C10H12OS
C10H12OS
C6H4BrI
C14H13BrS
Mol. Wt.
138
78.5
180
180
283
293
MP, o C
9
BP, C
207
51.8
121
329
Grams
100
64
100
165
Moles
0.72
0.82
0.56
0.58
Moles/ mole 2-4 dimethyl thiophenol
1.00
1.13
Mole per
S- (2,4-Dimethylphenyl) ethanethioate
1.00
1.05
Theoretical yield, gm
180.00
293
Actual exp. Yield
130.00
162
% yield
72.22%
55.29%
Table 1: Stoichiometry of First two steps Vortioxetine USP 10,227,317
Analysis of Step 2:

This step using rationale mentioned earlier need to be optimized. Effort to reduce sodium tert-butoxide with less expensive base is necessary. I would review and explore alternate and simpler ways instead of the suggested expensive chemistry? In addition, stoichiometry needs careful consideration to minimize use of excessive amount of all chemicals. 

Process for preparation of levothyroxine and salts thereof: USP 9,932,295

This is an interesting patent. It uses oxidative properties of bleach in a reaction. Since the oxidative properties of bleach have been used in the past, I am wondering about the validity of this patent. Patent claims:

1.     A process for the preparation of Levothyroxine of formula (II) comprising iodination of compound of formula (III) with sodium iodide and sodium hypochlorite in the presence of an aliphatic amine. 

2.    The process according to claim 1, wherein said aliphatic amine is selected from the group consisting of methyl amine, ethyl amine, propyl amine, isopropyl amine, tert-butyl amine, diisopropyl amine, diisopropyl ethyl amine, n-hexyl amine, morpholine, triethylamine, and mixtures thereof. 

3.    The process according to claim 1, wherein said iodination reaction is carried out in suitable solvent comprising water, dioxane, methanol, ethanol, or mixtures thereof. 

In the suggested process two moles of sodium iodide per mole thyronine are needed. Bleach acts as the oxidizing agent. Some excess of sodium iodide is necessary but not the quantity mentioned in the patent. Quantity of bleach has to be optimized. For a commercial process, I would prefer water-based chemistry as they are commercial. This is suggested because at Sherwin Williams Chemicals in late sixties used a similar chemistry to produce isatoic anhydride from phthalic anhydride by a continuous process using bleach and sodium hydroxide after phthalic anhydride had been converted to phthalimide. Bayer AG was granted a USP 3, 687,951. 

Intent here is not to challenge the Lupin patent but present that substitution reactions using bleach and an alkali salt have been commercial for over fifty years. The outlined process can be economized and if executed properly Lupin using modular operation can produce global demand for Levothyroxine from a single plant that would require multiple plants to formulate. 

Syntheses proposed by chemists are a laboratory pathway to a product. However, optimization and careful pruning of the laboratory process is necessary to have an economic commercial process. There are many routes. Creativity and imagination are necessary in the simplification process. Careful exploitation of mutual social behavior of the chemicals and unit operations is necessary for creating an economic process. Underlying goal of the chemists and chemical engineers has to be that the commercial process they will create is their best ever and the most economic. It is a challenge because the product demand dictates the type of process they will create. 

Another factor which is latent is that how API manufacturing process impact our environment. We may not have the necessary empathy but we need to recognize that active pharmaceutical ingredients are organic chemicals that are toxins that kill disease causing bacteria, a living organism. We don’t know the toxicity of every chemical and other reaction byproducts. Thus, if we can minimize their presence in the environment, we will make a difference while curing illness. Their presence impacts mammal, aquatic and bird life.  

Girish Malhotra, PE
EPCOT International

Reference:

  1. Malhotra, Girish: Chemical Process Simplification: Improving Productivity and Sustainability John Wiley & Sons, February 2011
  2. Malhotra, Girish: Focus on Physical Properties To Improve Processes: Chemical Engineering, Vol. 119 No. 4 April 2012, pgs 63-66
  3. Communication with Dr. Charles Kausch, Senior Scientist, OMNOVA Solutions, Akron, April 16, 2019

No comments:

Post a Comment