KiloMentor Stresses the Importance of the Integrity of the Reactor at Scale

Laboratory equipment costs just a minuscule fraction of that of process equipment. For that reason, scientists can perform a reaction that requires strong aqueous alkali in a glass round-bottomed flask even though one knows that at the end of the reaction the flask will be opaque and etched by the dissolution of a portion of the glass itself. One the other hand, precautions must be taken that a large scale reactor, that is expected to have a long useful life should nor e partially dissolved or pitted or weakened by any reactor contents. A process development chemist must never put a large scale reactor at risk. Consideration should be paid early on that reaction conditions are not incompatible with the materials of construction. Engineers are particularly knowledgable in this area and can provide an early warning that particular conditions must be vetted. This is normally done in the laboratory by placing weighed tiles of reactor surface material into the laboratory reactor throughout the process step of concern and at its conclusion, these tiles are fished out and carefully reweighed. Any experimentally significant difference between before and after weighings is suggestive that teethe reaction conditions are eroding the reactor surface material, 

At the same time, the experiment will detect any unexpected effect of the reactor’s material on the course of the process's reaction.

Loss of the surface of the reactor can also be caused by abrasion. The surface is simply rubbed off and probably remains as fine insoluble particles inside the reactor. Very little can be done about this except togged away from the abrasive reagent. Sometimes this problem can be solved by packing the abrasive agent tightly into a special column-shaped reactor tube and rapidly circulating the reaction mixture solution through the column past the insoluble abrasive agent.

Loss of the reactor surface may simply be caused by excessive pH and this cane controlled by an adjustment in the reactor material itself.

Another cause is the use of or the creation of a very strong chelating agent which simply rips metal ions out of the reactor surface. I have encountered such a situation. I was able to overcome the corrosion simply by adding a stoichiometric quantity of an inorganic iron salt into the reactor with the rest of the reagents. As the chelator formed it complexed the iron cations and left the reactor alone!

Identifying Chemical Process Stopping Points for Working in the Kilolab or Pilot Plant

It is not as if there is no planning in the laboratory. If a synthetic lab procedure is so long that the reaction and workup cannot be completed in a single day, chemists can use their experience to extrapolate from similar procedures and guess at what points manipulations can be stopped and under what conditions intermediate solutions or crude solids can be stored without damage. Occasionally there are misjudgments and surprises and a product will be prepared in lower than expected yield or poorer purity. But then even in the worst situation what is lost is no more than a couple of man-days of labor and the price of the starting materials consumed. Also, in the laboratory, because the capacities of refrigerators, freezers, and evaporators are so much greater than the quantities of material being transformed, there are do-able fixes at almost any stage for the situation where a stoppage is forced.

There is no room for such risk-taking on-scale. For advanced intermediates that are themselves the product of a series of sequential steps, one misstep can be economically disastrous. The more points in the process that have been verified as safe-to-stop, by actual test results, the more confidently the process team can be. Moreover, to be a safe stopping point it must be proven safe not just for the quantity and quality of the product but also for the protection of the processing equipment.

As a general rule once a reaction has been initiated, the kinetics must be allowed to run undisturbed to the proper end-point according to the batch sheet. The dynamic transformations cannot be expected to respond to any speed up or slow down without some quantity or quality deviation. After the endpoint condition has been reached and the reaction quenched then the mixture is likely more stable and various stopping points during the work-up can be tested by holding portions of a process mixture for given periods under controlled conditions and examining the mixtures and isolating the product to see whether an unacceptable deviation has occurred or not.

It is more difficult to demonstrate a good stopping point where the mixture in the process equipment is heterogeneous. The difficulty is taking a representative sample for analysis out of a heterogeneous mixture to show that no change affecting quantity, quality, or the protection of the reactor has occurred.  Since one cannot easily take a precise fraction of a heterogeneous mixture, working up that fraction after a pause will not accurately tell you whether the yield would have been different.

Finally, to fairly test the stability of an aliquot at a proposed stopping point the aliquot must be left in contact with a sample of the reactor material. In my experience, this is rarely ever done. At the very least it should be kept in mind where an aliquot might be corrosive to the reactor material.

Use of free radical inhibitors or antioxidants to increase the overall yield of organic synthesis steps

The use of radical inhibitors or antioxidants to improve yields does not appear to have many precedents in organic synthesis. A keyword search in 2011 provided only two references- both related to the stabilization of m-chloroperbenzoic acid towards thermal degradation during the epoxidation of resistant olefins.

Y. Kishi, M. Aratani, H. Tanino, T. Fukuyama and T. Goto, J.C.S. Chem. Comm. 1972  64 and 

D.M. Tal, Steroids (1989),  54(1), 113-22.

The best inhibitor found by Kishi for stabilizing m-chloroperbenzoic acid was 4,4’-thiobis-(6-t-butyl-3-methyl-phenol) that allowed 100% of an m-chloroperenzoic acid charge to be retained after 3 hours heating at 90 C in ethylene dichloride. Octene-1, dodecene-1 and methyl methacrylate were quantitatively epoxidized using such stabilized oxidant.

Synthetic chemists apparently assume that free radical reactions do not occur unless free radical initiators are present in the reaction mixture or unless the reaction mixture is irradiated. It might seem they think it can’t happen unless they are intending it to happen. Obviously, this is not true! Free radical reactions can take place not just during the contemplated reaction phase but during the work-up of the reaction when we might think that all the reacting is stopped. Actually, the opportunity is greater in the work-up phase; this phase usually takes more time, particularly when the process is being scaled up.

Are free-radical reactions inhibited by particular pH ranges of the solvent medium? No, they are not. The most frequent type of free radical reaction is oxidation and only the relative amounts of different species that can be oxidized are affected by pH not particularly the oxidation rates.

Oxidation often produces coloured products when it can introduce new unsaturation into molecules. The presence of unexpected colour in a reaction is suggestive of unanticipated oxidation. I recall that in the preparation of some aniline compounds the procedure teaches the addition of hydrogen sulfide to the aqueous phase during isolation to prevent colour development from exposure to air during workup and crystallization. The usual response to a colored product is to use charcoal in the recrystallization rather than trying to prevent colored by-products in the first place.

If you are performing a distillation and the contents of the still pot are darkening why wouldn't you add an antioxidant? Answer- I've never thought of it.

The Potential Use of Acetic Anhydride/Acetic Acid for Enabling Solvent Switches during Work-Ups

Each reaction in a chemical process has solvents in which the conversions works better and the preferred solvents for consecutive reactions in a scheme are usually different. As a consequence, performing solvent switches is essential for telescoping process steps thereby avoiding unnecessary intermediate isolations.

The boiling points of acetic acid and acetic anhydride are respectively 117 and 140 C. Both acetic acid and acetic anhydride are quite inexpensive and they are biologically trouble-free.

Acetic acid is infinitely miscible with water and is an excellent solvent for broad classes of substrates. Mixed solutes dissolved in acetic acid lead upon water addition to decreasing solubility of most organic compounds.

Acetic anhydride is a solvent that reacts with solute molecules that have nucleophilic functionalities and particularly those with what is termed 'active hydrogens'. Because of its even higher boiling point, acetic anhydride can chase many lower boiling solvents during distillation. It can then be, itself, converted by hydrolysis to acetic acid, optionally neutralized with aqueous alkali, and washed away from lipophilic materials. Heating a solvent mixture in which acetic anhydride is a constituent dries it. Only enough acetic anhydride needs to be added to a crude product to provide liquidity, then distillation instituted until all the first reaction solvent has been removed. Even if an acetate ester or amide is formed during isolation, that can be reversed by alkaline hydrolysis after the solvent of the first reaction is removed.

Because acetic anhydride has a bp of 140 C, it can chase many different first solvents. Just considering those that boil above 60 C they include diisopropyl ether, pet. ether, carbon tetrachloride, butyl chloride, methyl ethyl ketone, benzene, cyclohexane, chlorobenzene, acetonitrile, methyl chloroacetate, 2-nitropropane, MIBK, nitroethane, toluene, 1,1,2-trichloroethane, trifluorotoluene, 1,4-dioxane, nitromethane,  methylcyclohexane, heptane, propionitrile, cyclohexene, 1,2-dichloroethane,  fluorobenzene, 1,2-dimethoxyethane, 1,1-diethoxymethane, trichloroethylene, tetrachloroethylene, dimethylcarbonate, and diethylcarbonate.

 

Consider for example acetic anhydride’s potential for changing from the high boiling solvent chlorobenzene to ethyl acetate. In such a scenario, a mixture of chlorobenzene and acetic anhydride could be distilled to remove chlorobenzene and some acetic anhydride. The still-pot residue would comprise acetic anhydride and non-volatile reaction mixture components. This residue does not solidify because of the presence of the acetic anhydride. The minimum stirrable volume is maintained. Water is added along with the new second solvent which must be water-immiscible, in this case, ethyl acetate. Dilute mineral acid or base may be added to accelerate hydrolysis of the acetic anhydride. The acetic acid or acetate anion dissolves in the aqueous phase and is cut away. The reaction mixture is left dissolved in ethyl acetate.

In a different scenario, if the first solvents are low enough boiling, acetic acid itself can serve as the chase liquid for distilling away the first solvent. The product may not be particularly soluble anhydrous acetic acid or the acetic acid can be subsequently diluted with water used as an anti-solvent to cause precipitation or the acetic acid can be optionally neutralized and washed away with water after adding the new water-immiscible second solvent.

Acetic acid itself forms azeotropes with many common solvents that reduce the temperature at which they can be removed: butyl ether, chlorobenzene, cyclohexane, cyclohexane, tetrachloroethylene, trichloroethylene, toluene and xylene are among these.

Continuous Chemical Flow Reactors that Scale Well are not New

Even back in 2013 when this blog was first written, continuous flow reactors were increasingly popular. They have been available commercially for many years. They have become mechanically sophisticated in their pumping and controls. But even in Organic Synthesis Coll. Vol. III pg. 172 the synthesis of Carboxymethoxyamine Hydrochloride is described and it uses a continuous flow reactor in the first step.

The reactor works by gravity flow and is made from simple glassware and operates at 100 C using steam heating.

The procedure can be expected to work for reactions that are slow at room temperature or below but procedure rapidly at 100 C. The Organic Synthesis procedure combines acetone oxime with bromoacetic acid using an aqueous base:

“A mixture of 612 g. (4.4 moles) of bromoacetic acid and 500 g. of crushed ice is chilled in an ice-salt bath and made distinctly alkaline to litmus with sodium hydroxide ( about 440 g. of a 40% solution). During the neutralization, an additional 500 g. of ice is added. To the solution are then added 292 g. (4.0 moles) of acetoxime and 440 g. of 40% sodium hydroxide (4.4 moles), the temperature being held below 20 C during the addition of the alkali. The mixture is then allowed to flow dropwise, during 3-4 hours, through the inner tube of a steam-heated Liebig condenser (jacket 75 cm. long; inner tube 10-mm diameter; angle of inclination about 20 degrees) into a 5-l. round-bottomed flask cooled with running water (Note 2).”

Note 2 says that “[b]y this procedure, the reaction takes place in a few seconds, and the formation of by-products is minimized. If the solution of the reactants is heated in bulk, the reaction temperature cannot be controlled and a lower yield is obtained of a dark product which, however, can be purified by distillation under reduced pressure.”

The total throughput can be calculated to be 2784 g of solution which passes, in we can approximate, about 3.5 hours. That is 13.3 g. per minute. The actual duration that material is heated within the steam-heated 100 C zone is determined by the angle of declination of the condenser tube. One can imagine that using instead of a Liebig condenser an Allihn condenser,  that has a series of bulbs through which the liquid must pass, would imitate the effect of a series of continuously stirred tank reactors and the condenser would not need to be so long to have the heat contact time.

Practical Recyclable Chiral Acid Resolving Agents for Making Diastereomeric Salts: Lasalocid and (-)-DAG

When performing a chiral resolution at-scale it is important whether the resolving agent can be re-isolated, crystallized to a consistent purity, and thus practically reused. When the resolving agent is a carboxylic acid, this is simpler when the carboxylate salt of an alkali or alkaline earth metal is soluble in water while the free acid precipitates from water. Two common chiral acids have this characteristic: lasalocid and (-)DAG.

Lasalocid

Lasalocid sodium is a veterinary pharmaceutical available in large quantities. It is a chiral carboxylic acid that can be used to form diastereomeric salts with racemic amines. Based on tested examples it is predicted to work most dependably for primary amines that have their chiral center at the alpha or beta position as well as tertiary amines with a proximate chiral center with respect to the nitrogen atom. The ligand is capable of multipoint binding with the amine as it forms hydrogen bonds to many different oxygens. The ligand contains many different chiral centers. The molecule is made by fermentation. The acid is relatively inexpensive. It was covered by US 4,129,580 which expired in 1998.

(-)-2,3;4,6-di-O-isopropylidene-2-keto-L-gulonic acid hydrate also called (-)-DAG

(-)-DAG is also a water-insoluble chiral organic acid that can be used to resolve chiral asymmetric amines.

It is a relatively inexpensive compound that is used an intermediate in the synthesis of Vitamin C.  It was first prepared by Reichstein et al. Helv. 17, 311 (1934). Its use for resolution was taught in the expired US patent 3,682,925 (1972).

Reactor Cleaning: Where Organic Process Chemists Can Help Chemical Engineers in Process Development

For simplification in the operation of the plant, chemical engineers prefer a standard cleaning protocol no matter what process step has preceded it. This is often possible but for it to be workable without exception is wishful thinking. A standard protocol cannot take into account different substrates, different products, different processing conditions, different materials of construction, and the variety of different pieces of equipment in the reaction/isolation/purification train.

Because chemical engineers cannot as easily detect strongly adhering contamination in the larger equipment, they often learn about a problem far along in the development. The process chemists, in contrast, often working in transparent equipment that they clean themselves can be aware at an early stage when a cleaning difficulty is likely. Furthermore, so long as they know the standard cleaning protocol in the plant they are in a perfect position to know that it is likely to be seriously challenging.

Discovering an optimized reactor cleaning protocol can be regarded as unsophisticated stuff but it makes nonsense of our efforts to improve throughput with optimal processing conditions if, in fact, the reactor cleaning takes an order of magnitude more time to perform than the entire process! It very often can be easier and cheaper to improve throughput by reducing cleaning time by improving the cleaning protocol.

Reactor cleaning in API production is the most obvious situation where the process chemist can alert the engineers. It is in the reaction zone where highly insoluble, often polymeric, often baked or charred materials can become attached to the equipment. It is such impurities that provide the greatest challenge to cleaning methods because they cannot usually be treated by the physical abrasion of scrubbing. If an impurity can transfer either in solution or as a particulate downstream into the isolation/purification equipment that ability to migrate suggests an upper limit to the cleaning difficulty. Since it could be moved down the equipment chain it should be able to be moved out of the equipment entirely!

Neil G. Anderson in his monograph, Practical Process Research & Development, says nothing about reactor cleaning other than providing a reference to the article by I.I. Valvis, W.L. Champion Jr. “Cleaning and Decontamination of Potent Compounds in the Pharmaceutical Industry.”

Org. Process Res. Dev. 1999, 3, 44.  This latter article pertains to cleaning the residues from final products of known activity rather than unknown mixtures of compounds of unknown but probably low activity. Although the gunk that is tenaciously retained in the reactor zone is physically intractable it is likely not bioavailable.

The process chemist can do laboratory experiments in a fashion that will be more likely to show up such a gunking problem at an early stage. These difficult contaminants are often created when the reactor contents splash onto the vessel walls above the surface covered by solvent. This occurs in the plant because the entire wall of the reactor is heated not just up to the level of the reaction solvent. If in the laboratory the reaction flask is only lowered into the oil bath up to the solvent line, there will be no corresponding surface for this gunking to occur on and it might not be observed. To mimic more closely the process reactor both portions of the flask below and above the solvent line need to be heated.

When at the end of the reaction period the reaction vessel is visibly contaminated to an extent where hot reaction solvent will not make it visually clean, a scale-up problem is possible and potential solutions need to be considered in advance.

At the very least the process chemist should record and retain information about what was tried and what seemed useful in removing the visible impurities. It would also be useful to know at what point the impurities became apparent, whether they were deposited above the solvent level, below it, or in both places. Sometimes the gunk is more concentrated near the point of addition of some reagent or it may accumulate on the stirring paddle or the stirring shaft to a greater extent. 

The chemist may be able to make some useful guesses about the mechanism for producing the impurities and whether, for example, the impurities derive from a co-product (which will not be reduced in the optimization) or from a byproduct that could be reduced by optimizing. Since very often these dark-colored, low solubility substances are polymeric, consideration might be given to how a radical chain inhibitor might change things.

Polymers can also often be reduced by technologies that create an environment of high dilution for one or more of the reactants.

Definitions

Full cleaning is the more thorough cleaning protocol used when a different process step or a different product is going to be produced next in the reactor being cleaned. This is also referred to as decommissioning cleaning.

Partial cleaning is the less thorough cleaning protocol that is applied when the same process step is to be repeated next in the equipment. Some residual detectable contaminants are acceptable since they are the same as will be produced by the repetition of the step.

Boil outs, rinses, and swabs are three different methods for obtaining a sample to analyze to determine the extent of the cleaning.

A boil out is performed by refluxing a solvent in a closed reaction system in order to clean its interior surfaces and provide a sample of the residues in solution. The cleaning effectiveness of a boil out is a function of dissolution, mixing shear, and vapor extraction all resulting in an exponential dilution cleaning profile.

A rinse sample is performed using spraying or misting nozzles to send solvent where boil out would typically be impossible as for example in piping or portable equipment.

A swab sample is obtained by wiping a surface with solvent-moistened cotton gauze and it is used to grossly quantify the presence or absence of a contaminant.

Since boil outs result in exponential dilution profiles, equal results from two consecutive boil outs are sufficient to validate cleanliness.

The most common solvent to use in boil outs is methanol. Because it is miscible with water it does not form two phases even if the reactor is a bit wet. Although it is a good cleaning solvent for drugs since to be bioavailable they must have some solubility in water and hence likely some in polar organics, it is not necessarily good for process intermediates that may be very hydrophobic.

Acetamide is a solid at normal pressure mp 81℃ but it is liquid under reduced pressures: bp₇₆₀ 222℃; bp₁₀₀ 158℃; bp₄₀ 136℃; bp₂₀ 120℃ ; bp₁₀ 105℃ ; or bp₅ 92℃ . According to the Merck Index, 1 gram of acetamide dissolves in 0.4 ml of water, 2 ml of alcohol, or 6 ml of pyridine. It is also soluble in chloroform, glycerol, and hot benzene. Merck reports molten acetamide is reported to be an excellent solvent for many organic and inorganic compounds. It has been reported to be the most universal of all solvents. The high temperature required for melting and vaporizing the material will increase the dissolution. Under vacuum, the conditions for a boil-out are obtained in the reactor. Molten acetamide or condensing acetamide vapor can be expected to dissolve both organic and inorganic compounds.

Another idea for removing gunk would be to reflux the azeotropic mixture of diisobutylketone (isovalerone) and water. The minimum azeotrope boils at 97.0 C. When the azeotropic composition condenses it splits into two immiscible phases: 53.4% relative volume of >99% diisobutylketone and 46.6% of >99% water. Thus it is possible to boil out with a constant boiling mixture that applies a pure organic liquid of low surface tension to all the equipment surfaces.

Alternatively, using the azeotropic composition of the diisobutylketone reduction product, 2,6-dimethyl-4-heptanol and water (29.6% alcohol and 70.4% water) a constant boiling azeotrope can be boiled out in the system that upon condensation returns to immiscible alcohol and water phases.