kilomentor

Separation of Diethyl Malonate and Monosubstituted Diethyl Malonate: Another Phase Switching Hdrolysis

kilomentor | 29 April, 2012 20:04

In Organic Synthesis Coll. Vol. II pg 279 there is a procedure for monomethylation of ethyl malonate (diethyl methylmalonate). The crude, neat diester product is “shaken for exactly one minute with a cold solution of 10 g. of sodium hydroxide in 30 cc. of water.” The corresponding note reports that “[t]he ester is treated in this manner to remove any unchanged ethyl malonate [diethyl malonate]. Michael¹has shown that this treatment will completely remove unchanged ethyl malonate [diethyl malonate] while hardly attacking ethyl methylmalonate. No ethyl dimethyl-malonate is formed when methyl bromide is used as the methylating agent….A separation of the desired product from traces of unchanged starting material and from ethyl dimethylmalonate cannot be accomplished by distillation as the boiling points of the three esters lie within three and one-half degrees of one another.” The notes contain this additional information. “Michael found that unchanged malonic ester can be removed completely by taking advantage of the greater ease with which it is hydrolyzed by alkali ….”The references are Michael, J. Prakt. Chem. (2) 72, 537 (1905) and Fieser and Novello, J. Am. Chem. Soc. 62, 1856 (1940).
This appears to be an extraordinary example of separation by competitive reaction using heterogeneous media. For this to work the rate of hydrolysis of the starting material under these conditions must be at least one hundred times faster than that for the monomethyl product. Why would this be and how can other separations take advantage of this? It might be that the deprotonation and extraction into the aqueous basic phase of ethyl malonate in a vigorously agitated two phase mixture is much faster for an unsubstituted ethyl malonate than for one that is substituted, even with a very small group like methyl. It is very much less likely to relate to a difference in solubility between the ethyl and ethyl methyl malonates or even of their respective monoanions. Once ethyl malonate gets deprotonated and carried into the aqueous phase it is probably quickly further hydrolyzed to give carboxylates which keep it there. Kilomentor has given other examples of such ‘phase switching hydrolysis’ separating esters where there is a free phenol in only one component of a binary mixture and calls the method. http://kilomentor.chemicalblogs.com/55_kilomentor/archive/1011_extractive_and_phase_switching_hydrolysis_in_chemical_process_development.html

If the hypothesized explanation is correct this technique would be generally applicable in any monoalkylation of ethyl malonate with a hydrophobic side chain. It would mean that an excess of the inexpensive ethyl malonate could be used in such alkylation to speed up the overall rate in the knowledge that the excess ethyl malonate could be easily removed by alkaline hydrolysis and a water wash. This would also allow the amount of dialkylation to be reduced to essentially nothing.

About Extracurricular Training

kilomentor | 16 April, 2012 07:40

Vis-a-vis my last post, I am responding to Bender’s questions. Whether one gets an interview for a job depends both upon the degree of urgency that the employer faces and the quality of the competition. Does reading kilomentor prepare one for a process chem job better than actual experience in process chem? No. Definitely no. It can supplement but not replace. Does it prepare one better than having no process exposure at all? Definitely. Does it prepare a graduating chemist better than most of what is offered as process chemistry in academia? I think so. I’m trying hard. As it stands right now, I think kilomentor can replace academic training in its present form but can’t replace process development experience. Kilomentor can give new chemical graduates a leg up for an entry level position.

As far as the overall usefulness of organic synthesis as preparation for other chemical functions, I have seen throughout my career synthetic organic chemists switching with outstanding success into other jobs- analytical, regulatory affairs, and in my own case, solid dosage formulation. I have seen none move the other way. My conclusion is that thinking about any chemical problem from a base perspective of structure and reactivity has tremendous advantages. I entered solid dosage formulation with no academic training; just several weeks working with an old formulator, two days of those intensive classes, a textbook, and the chemical literature! A synthetic background is the best preparation for any of these jobs in my experience.

An Overview and Update of the Kilomentor Blog

kilomentor | 12 April, 2012 13:50

The Kilomentor educational blog is likely to have even more visitors starting next week. A letter from the author appearing in the letters section of the Journal of Chemical & Engineering News wil comment on the utility of blogs for educational purposes. Therefore, a review of the history, purpose, and perspective of the blog may be appropriate.

The Kilomentor blog was begun in 2006. The goal was to provide training and updating in the methods for chemical process development, emphasizing scale up of organic synthesis, particularly for high value pharmaceutical products. The author recognized that there were textbooks, symposia, and courses for this purpose but they were expensive and not equally available anywhere in the world. Moreover, in academia, the treatment of chemical process development was neither widespread nor generally thorough enough. The Kilomentor blog would be free, have post graduate content, and be available wherever access to the world wide web was possible.

The particular perspective of this blog was that designing synthetic routes selection has become easier than in in the past. Today, ever since electronic sub-structure searching became readily available, finding promising potential schemes has been simplified. What has not been made easier is choosing between these hypothetical process routes and finding rugged/fast methods of isolating process intermediates. Kilomentor has chosen to highlight techniques that both work on scale and make isolating clean intermediates easier. The articles encourage development chemists and chemical engineers to choose processes for scaling that comprise more of such easily purified intermediates. It has suggested that routes containing more of these rugged intermediates may even be preferable to processes with fewer apparent steps. Thus routes that can have higher throughput tend to win out over routes that just have fewer ‘ on paper’ transformations.

The Kilomentor blog has been well received. It has had more than 184,000 visits since its start and there are now more than 115 separate blog articles. Visits, in 2012, have reached about 200 per day. A search tool on the site will look for key words and identify relevant articles.

Thanks for participating; and if you are new, then welcome!

A Large Rigid Acid for Making Crystalline Salts : 4’’-n-pentoxy-[1,1’:4,1’’]-terphenyl-4-carboxylic acid

kilomentor | 31 March, 2012 14:41

I have talked about a number of recyclable reagents, such as lasalocid, in the previous blog. The title compound to-day can be used either as a salt forming reagent or as an appendage or protecting group.
It has a structure containing three phenyl groups strung together, end to end, through their para positions with an n-pentyloxy cap at one terminal and a carboxylic acid function at the other. All three rings are therefore para disubstituted. This acid, after coupling with the echinocandin B macrocycle, gives a product with excellent pharmaceutical properties. The substructure compound, as an amide derivative, is part of the drug anidulafungin. Since this linkage would be broken during the drug’s metabolism and since the drug has been found to be safe, this would seem to establish that this acid component is fairly nontoxic to the human metabolism.

This compound is a high melting solid and would be predicted to give high melting salts and other covalent derivatives. For example, the esters and amides are more likely to be solids using this rather than other carboxyl functionalities. The rigidity of the carbon skeleton might be expected to cause it to have a significant steric effect even remote to the point of attachment with another substructure. The expense of this potential protecting group would be ameliorated because the acid could be easily recovered and recrystallized to a dependable purity for recycling. Its use might be particularly promising where the process intermediates are expected to be liquids, oils, or waxes, such as in prostaglandin synthesis.

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

kilomentor | 22 March, 2012 14:46

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 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 centre at the alpha or beta position as well as tertiary amines with a proximate chiral centre 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 centres. 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).

More Discussion of Exhaustive Digestion in Chemical Process Development

kilomentor | 07 March, 2012 11:24

James Myslinski has asked me to elaborate on the technique of purification by ‘exhaustive digestion’. This is a phrase that I used in my blog entitled, Crystallization and Recrystallization Particularly On Scale. In this blog I make the point that chemists scaling up a laboratory process often fail to do adequate preliminary purification of a solid and as a consequence get a lower recovery than is possible if a preliminary treatment is done. I propose ‘exhaustive digestion’ as one pretreatment before recrystallization. I don’t know whether there are any widely accepted differences between digestion and trituration but, when used in my blogs, first, the noun trituration can be taken as applying equally to oils or solids while digestion will only apply to solids. Second, trituration can be taken to be done with ambient temperature or cold solvent while trituration and particularly exhaustive trituration is more often done with hot liquids, usually rather poor solvents for the desired product.

In the old literature, the progress of purification by digestion on the laboratory scale is followed by observing the change in boiling point in the digesting solvent or solvent mixture using a Beckmann thermometer, which is a thermometer that can measure differences as small as one tenth of one degree. I don’t know how long it has been since I have seen anyone use a Beckmann thermometer but, at least in Phys. Chem., it was done quite often in my younger days. The range of a Beckmann thermometer can be adjusted by removing or adding mercury to the column of mercury used for measurement by moving mercury back and forth from an attached mercury reservoir.

I will illustrate exhaustive digestion by quoting from an experiment taken from Laboratory Technique in Organic Chemistry by Avery Adrian Morton, First Edition , McGraw-Hill, 1938, pg. 128-229.
“The apparatus consists of a tube or flask of such size that a large part of the bulb is covered by the solid material being extracted , a thermometer graduated in tenths of a degree, a reflux condenser, and a water or oil bath. Place the sample in the container, insert the thermometer so the bulb is immersed in or covered by the sample, and add solvent to cover the solid and the thermometer bulb. The solvent is usually one that will dissolve impurities without appreciable quantities of the desired material. Petroleum ether or ligroin is often suitable. A mixture of solvents may be employed without effecting their utility in this experiment, as long as the composition of the mixture is not varied . Reflux the solution [I think slurry is intended] using a water or oil bath as a source of heat. After about 20 minutes the temperature has reached a constant level. Usually no trouble will be experienced from superheating as long as solid phase is present, although a little agitation with the thermometer is sometimes needed. Filter either by pouring into a Buchner funnel or by using a filter stick. Add more of the same solvent, reflux once more, and observe the temperature after equilibrium conditions have been reached. Continue the operation until the temperature of two or more successive operations are identical. The more soluble impurities have now been removed, and the solution contains only the pure compound or the compound with [less soluble type] impurities that have not been entirely removed…. The operation may be continued, if desired, until all the product has been dissolved. A change of solvent [may be implemented at some point for further purification]….. Constancy in the boiling points of successive portions constitutes further evidence of the purity of purity of the material, whereas a drop in the boiling point is evidence of the exhaustion of still another component. Usually the final portions of exhaustive digestion are pure materials.”

In light of the current interest in PAT, I cannot resist including the short paragraph that followed the quote from this 1938 text.

“Use the same apparatus to follow the progress of a reaction in which a solid phase is present, e.g., conversion of a triarylcarbinol to the chloromethane compound while under a mixture of petroleum ether and acetyl chloride.”

By continuously measuring and recording the changes, in real time, of the boiling point without any sampling this is an early application of a PAT technology. With the improvements in the size and sensitivity of temperature measuring devices, this following of small changes in boiling point under conditions where superheating does not occur perhaps deserves more frequent consideration.

Exhaustive digestion if used at-scale would require some more rugged method for measuring small changes in the boiling point of the slurry than A Beckmann thermometer loaded with mercury. Stirring using the thermometer as described in the above laboratory procedure of course out of the question. In a small amount of a thick slurry, the volume could easily fall below the minimum stirrable volume in the reactor so a plow type stirrer might be needed.

Allophanates- Potentially Useful Reversible Highly Crystalline Solid Derivatives of Alcohols

kilomentor | 28 January, 2012 15:44

In scale-up during chemical process development, Kilomentor has proposed that it is not just minimizing the number of synthetic steps that leads most often to the most cost efficient process, but more often the simplicity and ruggedness of the steps. This is because it is the time and other resources that get invested in the separation and purification functions rather than the reaction per se that contributes most to the overall cost. It is quite possible that preparing a solid derivative and then converting it back to the original functionality may have cost advantages and purity advantages.

Alcohols that are liquids at ambient temperature in particular may be better isolated as a solid derivative rather than as selected fractions from a fractional distillation. Similarly even if an alcohol is a solid, if it is low melting and present in a product mixture, a better yield of pure product may be available by making a derivative and then converting it back to free alcohol. No alcohol derivative introduces more polarity in terms of hydrogen bond donors and acceptors for the same small increase in molecular weight than the allophanate derivative.

The allophanate derivative ( C-O-CO-NH -CO-NH2) is formed by condensation of the alcohol function with two equivalents of isocyanic acid, O=C=NH (or which can also be represented as its tautomer cyanic acid).

The formation of the allophanate can be expected to increase the water solubility relative to the parent alcohol compound and decrease its solubility in organic solvents. Residual cyanuric acid formed during the preparation of allophanates is somewhat soluble in cold water but particularly in hot water. It is insoluble in cold methanol, ether, acetone, benzene and chloroform. Because of its acidity it: Ka₁ 6.31X 10^-8 pKa₁ 7.20; Ka₂ 7.94 X 10^-12pKa₂ 11.14; it is quite soluble in alkaline media.

Except for those arising from lower alcohols, allophanate are well melting, highly crystalline compounds suitable for isolation . The compounds are easily recrystallized and the parent alcohol can be regenerated by warming the allophanate with methanoic alkali.

Depolymerization of cyanuric acid can be done at 360-400 C in a slow stream of carbon dioxide. The gas can be absorbed directly into the neat alcohol or the reagent can be absorbed in an organic solvent such as ether to create a 30-35% by weight solution.

Consider the hypothetical reduction of 4-methyl-3-penten-2-one, (mesityloxide), bp 129 C, by hydrogenation. There are three possible products that are alcohols: 4-methyl-2-pentanol, bp 132 C, (ketone and double bond reduced); 4-methyl-4-penten-2-ol, bp 131.7 C, (only ketone reduced): and 4-methyl-3-penten-2-ol, bp 132 C, (ketone reduced and double bond isomerized). It is likely that reaction conditions can be found that lead to substantially one desired substance but such an enriched mixture still could not be separated based on boiling points. A solution that could be considered is the formation of the allophanate derivatives and recrystallization to achieve the predominant compound pure followed by hydrolysis back to the parent alcohol.

The Prins reaction is expected to produce a 1,3-diol from formaldehyde and an olefin. This is not necessarily a clean reaction. In fact the infrequency of its application suggests that it may lead to multiple products. Formation of the bis-allophanates as a method to obtain a pure crystalline product seems to me to be interesting.

The literature suggests that allophanates are derivatives that can be expected to crystallize from even quite difficult mixtures. For example the method was useful in the isolation and purification of various vitamins from natural sources. Fieser & Fieser in Organic Chemistry, the Third Edition, Reinhold Publishing Company, 1956. “The isolation of two pure factors from wheat-germ oil concentrates in 1936 was simplified by the discovery of crystalline derivatives, allophanates, resulting from esterification of the factors with cyanic acid….. on hydrolysis of the derivatives, the two pure active factors were obtained as highly active pale yellow oils named alpha and beta tocopherol.”

Similarly the derivative was used by Windaus and coworkers in isolating Vitamin D₃ from an irradiation mixture. This is reported in Fieser & Fieser’s, Reagents for Organic Chemistry Vol. 1 pg. 171:

“Treatment of the crude, oily mixture with isocyanic acid afforded directly a solid product easily purified by recrystallization from acetone and converted into pure vitamin by hydrolysis.” Vitamin D₃ has mp 82-84 C while the allophanate had mp 173-174 C, so one can see the inherent advantage. The co-products of the hydrolysis are conveniently totally water soluble!

The Chemical Manufacturing Process Route Selection: The Inclusion of Phase Switching and its Relationship to Validation

kilomentor | 20 January, 2012 11:31

Neal G. Anderson in his monograph, Practical Process Research & Development, says nothing about validation but he does speak about the need “to freeze the final process early, that is, to identify early the most desirable final chemical transformation to prepare the drug substance.” He credits this concept to P. Shutts. [“Freeze the Commercial Process-Issues and Challenges” a talk contributed at the Third International Conference on Process Development Chemistry, Amelia Island, FA, March 26 1997]. Earlier in his book, Anderson says specifically: “In order to establish routine impurity profiles and levels, ideally the final isolation of drug substance should be optimized, and this process should be used for the preparation of material destined for toxicological studies and later Phase I studies. The types of impurities found in drug substance will be largely determined by the starting materials and reagents used in the final step to prepare the drug substance. Thus the ideal penultimate compound should be identified in investigations, and parallel development work should converge on this penultimate compound.” In this particular passage that the last step Anderson is talking about is not the formation of a pharmaceutical salt. Almost certainly he is talking about a covalent bond forming step that completes the final structure. Also Anderson is speaking about the ideal situation, but what this assumes is not just that the last step, from penultimate compound to actual API, does not change, but also that this final intermediate is so efficiently purified that the steps preceding do not contribute to its impurity profile other than as inconsequential trace substances such as would be below 0.1%. It would only be in such instance that different routes to this penultimate compound would not leave behind any of their own impurities in the drug substance. If this is translated into the vocabulary of process validation, this would declare that there should be no critical steps before the purification of the ultimate intermediate! This is indeed a truly ideal situation and would constitute an impractical goal for real-life penultimate compound purifications to routinely seek.

Even though Anderson does not actually use the word validation in his book, even if we accept that validation is the most significant economic outcome of a developed pharmaceutical process; Anderson does convey the idea that certain characteristics of the final process step that simplify validation, are very, very important.

To achieve this requires a very selective choice of the penultimate compound. It must be easily and efficiently isolated and purified. As Kilomentor has argued this suggests that the step in which it is prepared and worked up involve several phase switches, because it is phase switches that provide purification and itt is compounds containing acid and/or basic functionalities that provide the most common opportunities for simple phase switches.

A Question from the LinkedIn Organic Process Research & Development Group that Fosters Practicing Creativity

kilomentor | 14 January, 2012 15:24

A question was asked at the LinkedIn Organic Process Research & Development Networking Group http://www.linkedin.com/groupItem?view=&srchtype=discussedNews&gid=1902030&item=88459764&type=member&trk=eml-anet_dig-b_pd-ttl-cn&ut=07TQuSi7kXnl41 “How could one remove a small amount of ethoxytrialkylsilane from a solution of ethanol containing some water at a scale of a large commercial waste stream.

I do not know whether a good answer is possible for this question. What interests me is what instruction the question can provide to students of process chemistry about how one searches for possible constructive leads for research.What is the entire real question?

Larry Fertel who posed the problem says, “I've got a stream of ethanol contaminated with <0.09 wt. % of a low boiling, trialkylethoxysilane. Any ideas on how to remove this (adsorbtion, coordination with a flocculent, etc.)? Distillation is not feasible, as the b.p.s are similar. Note that this is on a commercial scale….There actually is 9% or so water in the stream, water content is not an issue….The problem/issue is that the silane has a b.p. within 2 deg C of the azeotrope, sorry for leaving that piece of info out….Let me say that the ethanol stream is many millions of gallons per month, so this is not a research project. The ethanol is a by-product of a commercial process, so replacing it is not an option. Any treatment must work within the economics of reselling the ethanol into the marketplace (i.e., can only cost so many cents/lb.) . One option we are looking at is to convert to ethyl acetate for reselling. We are considering carbon treatment, one can build large columns in parallel that can be used on a continuous basis and get switched around once the carbon gets saturated in the impurity (assuming it works)…. I work for a contract research company, we were approached by a customer with this issue (good for us ($$), bad for them!!) So we need to be the problem solvers. I worry about the kinetics of adding a small amount of a reagent to a small amount of impurity. Considering that it is a bimolecular reaction, wouldn't the rate be slow? Not sure of the spec needed for the final product, at this point we are at the "screening" level.”

The problem is that there is <0.09% of an unacceptable compound contaminating millions of gallons per month of ethanol/water which cannot be removed by distillation because the undesired trialkethoxysilane distils close to 78.2 C, the ethanol/water azeotrope.
The problem is not a solution of ethanol/water. It is the presence of the trialkylethoxysilane at the level of <0.09%. There must be some low level that is acceptable otherwise the problem is to reach an undetectable level at low cost and that is almost certainly impossible. We must assume that the actual level is around 0.09% and it needs to be substantially reduced to let us say 0.01%

The problem is that the process as designed gives this level. It is not that that the level is around 0.09% because of some deviation in the process. There was no time when the level was acceptable. The unacceptable level is always present in the exiting waste stream from the process.

To solve the problem we might argue that we cannot remove the ethanol/water from the stream. That is we might say we cannot act on the major content of the stream because that would be expensive but apparently that is not the case. Larry Fertel says that distillation is rejected because it does not work not because it is too expensive. This opens the possibility that if we could selectively convert the trialkylethoxysilane to something separable from ethanol/water it would constitute a possible solution. Similarly it would seem that somehow changing the distillation temperature of the silane would be a possible solution and it would be particularly advantageous if it were the impurity that vaporized at lower temperature. Adsorption or adsorbtion onto or into another phase would be advantageous and probably most preferred because the predominant ethanol-water would be unchanged.

The hydrolysis and adsorption ideas have been pretty well explored by Stone, Tyrell, and others in the LinkedIn discussion. Kilomentor would like to look at the distillation and reaction possibilities a bit more.

One way to increase the degree of separation between the points of distillation of the trialkylethoxysilane and the ethanol/water is to add something to the mixture which will enhance the rate of vaporization of the more lipophilic material. This is the idea of extractive distillation. At the same time one does not want to leave anything behind in the ethanol/water that you are trying to decontaminate. Suppose we were to add an essentially immiscible perfluorocompound (expensive) or a much more volatile pentane (cheap). Would the vaporizing added compound sweep with it the undesired silane? I am reminded that there is a patent that claims the removal of traces of DMSO by co-vapor entrainment with ??.
It has already been suggested that a hydrophobic styrene-divinylbenzene cross linked resin might work. What about just using beads of solid paraffin? These would be cheap, sturdy and insoluble. They could be created by dropping molten paraffin into the stirred bulf solvent. They would float on the surface of the ethanol/water and could be filtered off and melted to drive off the impurity.

Remember guys, during the idea generation stage of problem solving criticism is supposed to be put on hold!

Now what can we get by playing with the idea of reaction of the impurity? What distinguishes the undesired impurity from the bulk? Molecular weight. The presence of silicon. The presence of a silicon oxygen bond. The presence of carbon-silicon bonds. The presence of silicon in the impurity seems to jump out at me. What is the strongest bond that silicon can form because that is what I want to do if I am going to drive this product into some other substance? The strongest silicon bond is one to fluorine. So perhaps the ion exchange resins that are being contemplated should have fluoride ion loaded on them. The conversion R₃SiOEt + F^-+ H₂O going to R₃SiF + H-OEt + OH^- would have an enthalpy advantage of about 27 kcal/mole. Hydrolysis in contrast has no enthalpy advantage, I think.

Anyway, problem solving is a combination of a good statement of the problem, a non-critical atmosphere for generating ideas, and most important well loaded retrievable associative memory banks. Note that you cannot come up with good ideas based on stuff you can’t recall but need to look up. Looking stuff up can be used to refine ideas but it cannot help to create them. You need to know things well enough that they can be recalled by associative memory when presented with the clues offered by the problem itself.

The Strategy of using Split Runs in Chemical Process Development

kilomentor | 23 December, 2011 15:25

Split run does not appear to be a widely understood term; however, in my personal chemical process development experience it is a common term with a well defined meaning. Split runs are experiments done to test different work-up/isolation and/or purification procedures starting with equal portions from a single large reaction mixture. As a consequence, even if a reaction mixture is both complicated and uncharacterized, a chemist can compare different follow-on procedures for working it up without being concerned about reaction differences that may still exist from batch to batch because the reaction process itself has not yet been finalized and critic parameters completely controlled.

In the general case, the isolated yield fraction, (which when expressed as a percentage is often called simply the percentage yield), is the product of the reaction yield fraction (determined by an assay of the crude reaction mixture before work-up) multiplied by the isolation yield fraction (the fraction representing the effectiveness of the recovery of pure product from the crude reaction mixture). In the case of comparing split runs, the ratio of isolated split run yields is proportional to the ratio of the isolation yields because the pure reaction yield is the same for each, since they come from the same reaction batch. Thus, in comparing the recovered product from split runs, the difference represents only the effectiveness of the isolation protocols….it is independent of and not confounded by the exact details of the reaction conditions. In this way, even at an early stage in developing the particular reaction step, the development chemist can get some idea of the comparative effectiveness of different treatments of the crude reaction mixture.

Of course it is not just the best isolation yield that is of interest and of importance to the process development chemist but also the relative purities of the two products and the amounts of different impurities within each.

The benefit in using split runs results because the optimization of the reaction conditions and the screening of potential isolation methodologies can proceed in parallel. It is not necessary to wait until the optimized processing conditions have been selected to work on the best isolation/purification means. The work can proceed more quickly and an earlier completion date can be set for some project milestones.

This does not mean that reaction optimization and isolation optimization can be completely decoupled. This can be seen to be intuitively obvious if one realizes that in an extreme case, if the assay yield in the reaction becomes 100%, there will be no residual starting material and no by-products formed and so this could simplify the work-up beyond all normal proportion and perhaps allow the most mundane isolation to give a fine outcome.

How thinking about Separating Acetic acid and Acetic Anhydride can sharpen our Chemical Process Development Skills

kilomentor | 16 December, 2011 10:50

Ganesh Wagh in the Chemical Process Development Group on LinkedIn asked the question, “What is the best way to separate a mixture of 60% acetic anhydride and 40% acetic acid to recover the acetic anhydride?” This solicited quite a few responses. The particular forum is now closed and although Kilomentor did contribute one suggestion to the forum I think the topic deserves some further discussion that doesn’t fit well into the question-comments format.

A plurality of people suggested either distillation of distillation under reduced pressure. These solutions may very well be the best if one has the proper fractional distillation equipment and it is not busy performing more valuable separations, but there is the implication in the question itself that, at least in the questioners hands, there was some difficulty with these solutions, since they are the first thing to come to a process chemist’s mind. Indeed Ganesh later noted that he did encounter problems in his hands.

Several people (Antonio Manka, Anton Frenkel, Marta Krawczyk, Pravinchandra Vankawala) suggested fractional distillation, either with some variation of the distillation condition for atmospheric pressure or under reduced pressure. I am not an expert on distillation. That is perhaps why it is not clear to me why distillation at reduced pressure would be expected to simplify separating these two liquids compared to an atmospheric pressure fractionation. Also, whatever the pressure under which the fractionation is done, these methods appear to require the complete removal of the acetic acid before the acetic anhydride distillation and that must leave some lost still bottoms and column residues that must subtract from the recovery. On top of this, I would imagine that fractional distillation must be slow and labour intensive. One suggestion, variation of the packing in the distillation column,which would change the height equivalent of a theoretical plate, is typically not readily accessible in a multipurpose plant particularly where the object is no more than reagent recycling.

Mark Bratt contributed the idea “Any thoughts as to a non-aqueous separation e.g. cyclohexane or methylcyclohexane with methanol or acetonitrile, with an organic base?” Although I don’t endorse trying a non-aqueous, two liquid extraction that uses methanol, which would be both miscible with and reactive towards acetic anhydride (this is almost certainly just a slip – it is true that methanol is very frequently a partner phase in such applications), the idea is suggestive of a type of strategy that is adapted and taken up by others later in the discussions. Perhaps we could simplify this idea. What would happen if we just added enough triethylamine to neutralize the acetic acid? Wouldn’t the triethylammonium acetate form a separate liquid phase aside from the acetic anhydride? Even if it didn’t, wouldn’t distillation of this mixture amount to an reactive distillation in which the acetic acid would be held back by its reversible interaction with the triethylamine allowing the acetic anhydride to distil over cleanly first? Acetic acid and triethylamine are reported to give a higher boiling azeotrope, with composition 69% acetic acid 31% triethylamine, that distils at 163 C. If triethylamine didn’t do the job what about neutralization with tributylamine?

Kees Hoek made a suggestionto add acetyl chloride that would, if it is worked, convert the acetic acid back to acetic anhydride thereby, in principle, multiplying the acetic anhydride recovery. My rough bond energy calculation for this reaction between acetyl chloride and acetic acid gives an enthalpy difference of close to zero so if the kinetics were practical it most likely would produce an equilibration of starting material and products. Driving out the hydrogen chloride would be expected to drive the equilibrium towards acetic anhydride according to LeChatelier. The usefulness might boil down ( pun intended) to how fast one can scrub hydrogen chloride. It might be attractive if one also had in the plant an alkaline waste that needed neutralizing.

Pedro Guivisdalsky, among other ideas, suggested adding toluene to form the binary lower boiling azeotrope with acetic acid that distils at 105.4 C. This is a variant on the extractive distillation idea whereby the acetic acid is made relatively more volatile. Waste toluene could be used and the toluene/acetic acid washed to recover the toluene.

John Knight, Scientific Director at Scientific Update, suggests neutralizing the acetic acid using the completely non-volatile basic ion-exchange resin. As he comments, this risks the exotherm that accompanies every mode of purification that features a neutralization. It would also involve bashing around a lot of insoluble resin that may be ground down and give some tedious still bottoms. Nevertheless, the core idea is that insoluble nonvolatile ionic exchange resins should come to mind when neutralization is considered in separations generally. Also, Dr. Knight again relates back to the idea of using some agent to convert the acetic acid to acetic anhydride. “What about phosphorus pentoxide”, I am musing to myself? The coproduct will be non-volatile and can be rinsed out of the still-pot after careful hydrolysis.

Many contributors to the discussion recommend looking at how acetic anhydride is made commercially as a pointer to an economic method. This is of course excellent advice since the cheapest starting material for acetic anhydride must be acetic acid. The methodology that is used for commercial product may of course be out of the question for recovering usable acetic anhydride in a multipurpose plant where the cost reduction for an accompanying process may be the actual motivation for the question.

Ilya Avrutov seems to have garnered the most endorsements for his suggestion that the acetic acid in the mixture could be made non-volatile by treatment with sodium carbonate, preferably anhydrous sodium carbonate. Others would include a drying agent. All understand the problem that the carbonate leads to water that must be taken into account. Ilya’s concept is another variant of a non-aqueous neutralization like that of Mark Bratt, above; it is simplified by not requiring further organic solvent additions but complicated because the coproduct of his neutralization can react with acetic anhydride to regenerate acetic acid.

This in turned stirs up another pertinent question. Are there cheap agents that can neutralize acetic acid without producing something that would react with acetic anhydride? Two chemicals that come to mind are calcium hydride and calcium carbide. These substances should come to mind whenever one wants to generate calcium salts of carboxylic acids for isolation or derivatization. In this instance the coproduct with calcium hydride would be hydrogen gas which would be innocuous but needs to be treated carefully. Using calcium carbide the byproduct would be acetylene gas that needs to be recognized and handled safely. In each case exactly how these solids are contacted with the acetic acid/acetic anhydride mixture would need to be worked out carefully. I suggest adding either as a slurry in more acetic anhydride.
I think I can still take credit for the most off-the-wall suggestion in the discussion. It was based on a simple question. What is the relation between the water hydrolysis of acetic anhydride and pH? Couldn't one stir the mixture of acetic anhydride and acetic acid with an aqueous salt solution buffered at the most unreactive pH and quickly extract the acetic acid into the aqueous phase? Stirring should be gentle to reduce interfacial contact and the temperature should be kept low to minimize reaction, and of course, separation of the layers should be done promptly. The idea here is that the difficulties that often stop us from reaching a simple solution simply may be mistaken. Here we assume that acetic anhydride will react with water and make a water wash unthinkable. But how well does this reaction actual take place and can it be slowed down enough to perform washing? Remember the Hinsberg reaction of a sulfonyl chloride is performed in contact with an aqueous alkaline solution!

Anyway, there were lots of ideas proffered. Thanks to everyone who contributed. My apologies if I created any distortion in presenting your contribution. The discussion was at http://www.linkedin.com/groupAnswers?viewQuestionAndAnswers=&discussionID=83697798&gid=1902030&commentID=60647184&goback=%2Egmr_1902030%2Eamf_1902030_76918230%2Eamf_1902030_49383814&trk=NUS_DISC_Q-subject#commentID_60647184
If you want to blast me at length please send e-mail to kilomentor@sympatico.ca. Short blasts can go in the comments here at kilomentor.

An Example of a Possible Kilomentor Type Separation at Scale

kilomentor | 22 November, 2011 10:22

Hi Marto,

Thanks for the problem. Although you do not provide enough information to exactly identify the compounds that is OK. I understand there may be intellectual property or privacy issues. I think I can make some suggestions that are at least mentally stimulating and which certainly exemplify the kilomentor approach.

Marto has asked for suggestions of how to achieve a particular separation.

He wrote: “I am having a problem in the isolation of mixture containing biphenyl benzylacetate compound along with diarylated benzylacetate copound. I have been carrying out this reaction on 100 g- 200 g scale, the ratio of biphenyl and diaryl compounds is 10:1, isolation by recrystallization was unsuccessful. Both have similar retention factor in TLC. Suggest me any purification method for the isolation of these two compounds.”

First, I would hydrolyze the acetates. When compounds are strongly hydrophobic, separation should most often be tried on the most polar functional groups available.

In the case where there is a difference between the number of substituents ortho to the primary alcohol function, the compounds can probably be separated by the difference in the rates they form the O-sulfate derivatives. Sulfonate the mixture with only enough sulfonating agent ( say chlorosulfonic acid in one of the solvents pyridine or dimethylaniline; see the kilomentor blog on making O-sulfonates). The sulfonated material and the non-sulfonated are easily separated by acid-base extraction and the sulfonation is easily reversedto give two separated alcohols.

In the situations where there is no difference in the number of ortho substituents between the alcohols one must depend upon a difference in inductive and electronic effects and the best chance comes when the primary alcohol mixture is converted into two carboxylic acids. The required hydrolysis and oxidations of the esters can be done together without isolation and the mixture of two carboxylic acids can be obtained cleanly and easily by simple acid-base extraction.

I think there will be a good chance that these acids can be separated by what is called extractive crystallization. (kilomentor has also written a blog about extractive crystallization). Extractive crystallization works by taking advantage of any small difference in pKas of the two acids augmented by any small difference in the solubilities of the two acids in an immiscible organic solvent selected to augment this pKa difference. The goal is to get one compound overwhelmingly as a salt in water and the other compound overwhelmingly as free acid in the organic solvent.

In any case, even if this cannot be made to work, as free acids there are many more salts, both organic and inorganic, that are accessible to find substances that do fractionally crystallize.

Also when one has a mixture of acids one can screen a portfolio of enzymes to find one that esterifies only one of these acids. Thus one uses the most discriminating of reagents, enzymes, catalytically in their favorite reversible reaction, ester formation/ hydrolysis.

Once separated the alcohol-acetates can easily be reformed if that is what is required. The extra steps in both directions are simply probably high yield and the separations are trivial acid/base extractions.

If you are reforming the acetates remember that the ester (non alcohol) can probably be easily separated from any residual alcohol by treatment with lithium bromide or calcium bromide or calcium chloride in hexanes where the alcohol is likely to form an insoluble inorganic complex leaving the ester in hexane solution.

What Might be the Best Solvent for Difficult Cleaning Jobs on the Reactor Walls of a Plant Reactor

kilomentor | 24 October, 2011 07:36

The walls of a large scale reactor can sometimes be difficult to clean. The problem is compounded because they are not easily accessible and cannot be inspected closely. Methods that can be applied in the laboratory for many reasons are off limits. Scrubbing is impractical, dangerous, and potentially damaging to the equipment. What is needed is a powerful but innocuous solvent that can work by vapor condensation not just below the surface of the refluxing liquid cleaner but above the surface and on the reactor walls where the reaction mixture may have splashed, caked and baked.

In the very old literature a common, inexpensive, and innocuous compound was claimed to be the best solvent known and one that would dissolve both organic and inorganic materials; salts as well as uncharged covalent molecules. It was molten boiling acetamide. Acetamide can be synthesized in situ in the reactor by heating ammonium carbonate and acetic acid and distilling out water. This in fact is the first preparation in the First Collective Volume of Organic Synthesis. Acetamide has bp₇₆₀222.0 C ; bp₁₀₀ 158 C; bp₄₀ 136 C; bp₂₀120 C; bp₁₀ 105 C; or bp₅ 92 C. As a white solid it has mp 82.3 C. The solubility is 2 grams per ml of water. Acetamide has been advocated as a “green” solvent [http://acs.confex.com/acs/green07/techprogram/S3384.HTM]The ninth edition of the Merck Index describes it as “Solvent; molten acetamide is an excellent solvent for many organic and inorganic compounds. Solubilizer; renders sparingly soluble substances more soluble in water by mere addition or by fusion.” Way back in 1933, Professor O.F. Stafford of the University of Oregon wrote that acetamide dissolved more different chemicals than any other known solvent. [J. Am. Chem. Soc., 1933, 55 (10), pp 3987–3988].

Process chemist sometimes forget that their responsibility is for the minimized overall cost of the process and this is much more than the chemicals only cost. The throughput per unit time is a major factor in the overall cost. That time includes the equipment cleaning time required between batch runs and between the end of one campaign and a new one for another product to be run in the same equipment. It makes little sense to invest extensive research efforts in reducing processing time when the same throughput efficiencies can be more easily achieved by reducing cleaning time between runs.

Many plants use a standard cleaning protocol implemented as an SOP. Only when it fails to remove all the contamination are special cleaning procedures resorted to. In some cases the standard cleaning or rinsing will even exaggerate the problem. For example, in the synthesis of adamantane described in Organic Synthesis Coll. Vol. III pg. 16-19, specific instructions are provided to avoid treating the vessel with water until acetone is used first to completely remove the tar.

The development chemists are the first to get an indication that unusual cleaning difficulties could arise after certain processing. Giving the plant scale-up people a heads-up and some suggestions will improve both teamwork and overall efficiency.

The Importance of Understanding Process Validation for Pharmaceutical Process Chemists

kilomentor | 17 October, 2011 05:15

Process validation is a legal requirement applied to active pharmaceutical ingredient manufacturers by governments and it is administered by their regulators. Process development chemists contribute to the documentation of process validation.

Philosophical Differences between Process Chemists and Regulators

Process chemists when faced with a synthesis problem ask themselves the question, “How might this be made to work?” We then seek the optimal operating parameters for a scheme that achieves the desired physical reality. Regulators look at what we have achieved and ask themselves, “How could this fail?” Instead of looking at the optimal conditions that have been found, they are looking at the limits for failure of every critical parameter of the system we have optimized and are asking how this could upset the output quality of the system? We, as creators, struggle to succeed at jigging the system to give the result we need. We struggle to understand the coded messages of our experiments to deduce what a practical process would be. The regulators analyze what we have created to see at what points it could break down. Since any process can be made to fail by deviating enough from its teaching, this thinking is alien and frustrating to us.

Drug Substance vs Drug Product

As Dr. Oljan Repic has pointed out in his book,[ Principles of Process Research and Chemical Development in the Pharmaceutical Industry, John Wiley & Sons, Inc. p. 179-194], a chemical process for making an active pharmaceutical ingredient (API) from commercial bulk chemicals is very, very different from making a final dosage form from its API. This is because, in the former, chemical steps result in the creation of new substances by first combining a great many undesired materials (substrates, substrate impurities, reagents, catalysts, solvents etc) and then, after exposure to some reacting conditions, separating these and other undesirable materials (co-products, by-products, unanticipated and unidentified impurities) from the desired reaction product. Contrastingly, making a drug product from an API does not involve chemical reactions or purifications (other than drying). The drug product is some kind of mechanical mixture or micro-engineered assemblage of the inputs and so it is a simple combination of what is put into the process modulated by the manipulations. This is not to say that how an active ingredient is formulated into a drug product doesn’t affect the medicine’s effectiveness or even safety. It does. What I (and I think, Dr. Repic) are maintaining is that formulating is an inherently simpler change than synthesis and consequently failing to control the process of the former within the limits established is much more certain to cause the product to fail its specifications. Since there is no purification aspect in formulating, if it is garbage in, it will most certainly be garbage out. There is no possibility for purification in formulation so there is no capacity to rectify any serious deviations.

The synthesis of a pure chemical substance such as an API is nowhere near so dependent upon the starting material purity and the nitty-gritty of the process. In fact in the limit (100% purity), the properties of a pure chemical substance are completely independent both of its route of synthesis and the details of the most immediate transformation by which it is prepared. Even in the real world of actual APIs, the only evidence of the route of synthesis that remains in ‘pure’ product is the faint fingerprint of trace impurities that are still carried in it. Effective purification decreases this below any level of practical significance related to efficacy or safety and even, in many cases, below the level of analytical detection.

Even if it might come as a surprise to beginning formulators that how an API and the other excipients in a drug product recipe are combined can significantly alter the bioavailability of a tablet; it would not surprise even one tyro chemist that deviation from the details of a chemical process would cause the quality and yield of such product to suffer.

It is the Purification that Must Work

In a synthesis it is really only the final purification that needs to be effective to deliver a drug substance of the required purity, with the proviso of course that the analytical methods used to assess that purity need be validated. The process by which the final product is produced needs to be known essentially only so that the analytical methods can be tested for their capacity to detect and quantify the authentic likely/expected impurities.

This however is not the law. Regulators and politicians err on the side of caution and the validation rules they approve are more rigid than they need to be, because the rules derive historically from, and still closely resemble, those for formulated drug products. Process validation was originally conceived and implemented as a method to assure that each unit of a drug product could be guaranteed safe even when it could not be analyzed, either because the analysis would destroy the product or because the proper functioning of the product could not be analyzed. Drug security by validating the process was initially directed to drugs formulated for injection into the blood stream. If these were contaminated by bacteria the patient would likely quickly die. Even if one took a proper statistical representative sample of the product there still could be some fatally toxic units of product undiscovered. The solution was to control the process used to make the units so strictly that it would be extremely unlikely to make even one contaminated sample. This was the only course of action that could be effective because the fatal contamination came from the process (from the bacteria carried in the air) not from any starting materials.

Process validation also turned out to be valuable for assuring the efficacy of drug dosage forms where the delivery of the active ingredient from the dosage form ( ie tablet, capsule) into the blood stream of the patient was challenging. In these cases, analysis of the dosage form was not showing reliably whether a particular lot of drug product could deliver the clinically required exposure to the medicine. Drug product prepared by a validated process was found much more likely to produce the desired clinical outcome.

Process validation differs significantly from chemical analysis in its ability to provide protection from different types of contamination problems. A drug substance is almost always a pure chemical compound. It is a single molecular entity and can be analyzed; precisely, to confirm its constitution, and sensitively, to show the presence of homogeneously dispersed contaminants. Drug substances are not mechanical mixtures like flakes of this and that stirred together as is often the case for tablets and capsules. Such mixtures are statistically homogeneous only at the scale of a dose but heterogeneous at the particle scale. This is what one finds in a mixture that is pressed into a pill. Process validation is better than analysis to guarantee that heterogeneous, most often particulate, contamination is not present. The reason for this is that chemical analysis of an active pharmaceutical ingredient (API) requires taking a representative sample from the entire batch to perform the analysis upon. It is quite likely that even with an approved statistically sound sampling procedure a heterogeneously contaminated batch would not get any contaminant into the sample being analyzed. Take for example the situation where glass particles from a chipped piece of equipment contaminated a batch. The dense glass would migrate to the bottom of the container and might escape the sampling. Thus it would not be detected. A validated process, however, could make sure that there was no conceivable condition where any glass could remain in a final product. The final filtration before the final crystallization of an API is part of a process for just such a validation purpose.

Process Validation: A Development Scientist’s Most Important Activity

Drug discovery chemists do not need to learn about process validation; process development chemists do. It is the latter’s work that is being examined. Nevertheless, scientists in general always like to know whatever they can about the synthetic routes they devise and discovery chemists may eventually transition into career development chemists. The additional experimentation that process validation requires does increase scientific knowledge about that process. It answers questions about the practical limits of those process conditions that most significantly impact final product quality. But aside from the scientific merits, it is important that pharmaceutical chemists, particularly process development chemists, understand process validation, since, because the pharmaceutical business is a highly government regulated activity, any API they might make cannot be included in a medicine for sale unless it is made by a validated process. This makes validation the most economically significant test to which a scaled up industrial process can be subjected and so all the chemist’s activities should be viewed from the perspective of its contribution towards success in validation.

Many and perhaps most companies that manufacture APIs exclude scientists below the managerial level from any contact with the government regulators who assess the quality and completeness of the validation process. Likewise, it is only senior technical people who sign off on or assemble the validation report, which is the core document of the validation exercise. Yet exposing employees, who have very specialized tasks, to the all encompassing validation program lets them see their precise role in it. Specifically, chemists, who are focused on process development, are likely to do much better in organizing their data, writing reports, etc. if they understand the larger purpose their data and conclusions will serve when process validation protocols are assembled and executed and the Development Report is folded into the Validation Report. Indeed it is highly recommended that general training in the concepts of process validation should be given to all individuals who are part of the development and manufacturing program not just to those who require such training as part of Good Manufacturing Practice (GMP).

The Corporate Validation Policy

Validation is so important to API producing companies that very often it is heralded in a written Corporate Validation Policy. Where this document exists it is the overarching document pertaining to the validation methodology. As such it sets down general principles that employees should adhere to, rther than the nitty-gritty details; nevertheless, it is a good place to start in examining that most important activity called process validation.

A lifecycle policy approach to the validation concept is regarded by many experts as preferable or even essential. Validation activities begin with the inception of the development program and logically should not end until the product is retired. A corporate validation policy should provide validation expectations over the lifetime of the process. It should state the company’s expectations for both the organization’s scientific and its documentation goals. Both of these classes of goals need to be kept in mind and adhered to from the first day of process development to the last. In other words the Product Development Report (PDR) should be a ‘living’ progress report that is continuously updated through the product lifecycle. Many firms make the mistake of creating a static, concrete, dated, PDR paper which is archived once that first regulatory approval is obtained. Because improvements are being made throughout the lifetime of the process, these may require additional validation work and regulatory filings. These changes should set in motion an update of the PDR.

Since every organization has somewhat different internal structure, any policy needs to allocate responsibilities for maintaining the product/process documentation including the mature product responsibilities relating to monitoring, trending, and change control.

The Validation Master Plan

The Validation Master Plan is the most senior planning document that pertains to a specific validation. It makes concrete and project specific what is set down in the Validation Policy only in a general way. It takes into account even the most recent changes in the organization’s structure, responsibilities and reporting chains. The Validation Master Plan assigns actual project milestones to named persons and groups. It sets dates for meeting those milestones. It sets out what activities are to be performed when and it decides the science that will take place at small and at large scales. If the validation is for a new drug substance the Validation Master Plan needs to be revisited and modified as the clinical phases advance.

All aspects of the validation documentation are important. The regulators look not for what the company considers most important; what is scientifically most significant; what has been presented most clearly and completely or where the maximum inherent risk is, but what has been overlooked because that is from their perspective where that process deviation leading to dangerous medicine could most likely arise.

Some Aspects of Process Validation that are not the Development Chemists’ Responsibility

Validation must demonstrate that the process being examined is performed under good manufacturing practice (GMP). Mercifully there are many aspects of GMP that are not the responsibility of development chemists. Every organization that manufactures product for sale, makes batches for clinical (human) tests, or to support regulatory filings must have a separate and independent quality control unit to test, approve, and reject products; set and approve specifications for raw materials and products; and write and approve these and other written procedures. Another requirement of GMP is that all personnel associated with production must be trained concerning GMP and the company must be able to prove that training with the signatures of those trained. You may have been asked to submit your resume to a company representative responsible for meeting GMP standards. The company is required to show that its scientists too are adequately credentialed for their work.

The facilities and equipment to be included in a process need to be qualified before they can be used in an acceptable validation exercise. These qualifications are usually unfamiliar to practicing chemists. They are matters that are implicit in our everyday activities and are completely taken for granted by us.

Facility qualification relates to the suitability of the physical site and its infrastructure and personnel to carry out the process that is to be validated. Most of these matters are strictly the domain of architects, contractors and engineers.

Drug substances in general have requirements for the facilities in which they are manufactured. The FDA will not allow someone to make a GMP drug substance in a bathtub in a garage! It insists upon written assurance that the manufacturing facility has adequate space, heating, ventilation, plumbing and maintenance.

The Development Report

The core document that development chemists do contribute for API validation is the Development Report. Process chemists recognize the Development Report as the history of the entire process development but it is at the same time a regulatory validation document of enormous importance. Validation would need to be done, if for nothing else, at least to convince regulators that the product is reproducibly safe, but API product safety is not the only goal of process development. Optimization aims to minimize cost, maximize throughput, achieve appropriate safety, reduce environmental burden and deliver consistent, convenient operation of the steps. Consequently some IPCs and some specifications are not critical in the regulatory sense of controlling final product quality. Deviations from these do not put the API quality at risk. Some of these aforementioned standard ranges alert batch sheet reviewers, when they are changing non-randomly, to potential creeping deviations (trends) that can have economic, safety, environmental or simply consistency consequences, without there being quality ones. This means that an API could fail a noncritical IPC and still be safe and effective to use. Failing a specification or IPC that correlates with a critical process parameter (CPP) and in turn a critical quality attribute (CQA), on the other hand, would mark a failed batch.

A synthesis can in strict theory be traced back to the most elementary of chemical building blocks- even atoms themselves. What needs to be decided for validation is how many of the steps in the retrosynthetic direction need to be validated to insure the required safety. This ‘depth’ of the validation study establishes what chemical transformations or treatments are included within the validation and what steps are accepted simply as the making of starting materials to be controlled by their material specifications independent of their particular synthesis.

Today, it is a good bet that most synthetic API processes are multi-step with at least one complex chemical transformation. Similarly it is unlikely to be disputed that the final process step that creates an API, and the subsequently entailed purification, needs to be validated. Beyond this there is no consensus answer as to when to apply process validation for intermediate steps in a multi-step process. Each process it seems must be evaluated as a separate case. So many factors are in play. This does not mean that a coherent policy or procedure cannot be established for a company describing how it should do that evaluation. In fact, putting such a decision tree in place can lead to better, faster, even less expensive, choices.

The process validation for an API is at least the sum of separate validations; one for each intermediate step that contains a critical parameter. Thus determination of what steps are to be included in the full validation amounts to identifying what steps contain at least one critical parameter. This is much the same as asking what the starting materials are for the steps that will be validated because steps for making an API starting material do not need validation. Terminology can be the source of as much disagreement as the chemistry in these discussions so definitions are important. A ‘step’ leads either to an isolated or non-isolated intermediate. Other important determinations to get straight are: what is(are) the ultimate intermediate(s) or final intermediate(s), and what overall scheme of reactions and purifications will be required to produce the final API.

Determining what constitutes a starting material is debatable. According to Roger W. Koops, [Process Validation of Synthetic Chemical Processes for the Production of Active Pharmaceutical Ingredients (APIs), Journal of Validation Technology, Volume 8 Number 2 January 2002], “a starting material should be a readily available item that has a standard grade (or grades) associated with it, and has been well characterized. If produced under contract by a vendor, it should be produced by a known and established process, and the end product should be, again, capable of achieving a standard grade. Second, the vendor should be qualified (under a vendor qualification program)to to ensure that a material meets consistent quality standards. A rudimentary quality agreement should be established to outline change notification and quality requirements. What must be avoided are unshared, unannounced process changes that can change the impurity profile of the downstream drug. If a material is manufactured by a contract party and the process was supplied by the innovator, the material is simply a third-party manufactured intermediate. The contractor now becomes an extension of the innovator, and the transferred process must be included in the full validation evaluation”.

Starting materials, generally speaking, are inputs to a synthesis whose atoms are partially incorporated into the drug substance. They do not need to have their own syntheses validated. Their quality is accepted based on their certificates of analysis. In contrast, intermediates are created directly or indirectly from starting materials. They arise downstream in the process with respect to some starting materials. Most intermediates arise in steps that need validation, but at least in principle there can be exceptions. A step starting with an intermediate and resulting in another intermediate does not need to be validated if the interconversion process step has no critical parameters. That is, the step is so rugged that no matter how it is executed it does not affect the purity of the final drug product. This would be a rare case indeed. The reaction would need to be ultra tolerant of reaction parameter deviations and/or the isolation would need to be exceptionally powerful at excluding impurities.

As the overall chemical development proceeds an impurity profile for the final product should emerge The validation policy should emphasize the importance of characterizing process impurities as early in the development process as possible. From this, each intermediate process step can be evaluated as to its influence on the purity of the end product. If an intermediate needs to be controlled because it could contribute to an adverse quality profile of the API, process validation should be applied to its related process step.

Each intermediate needs to be evaluated for how it contributes to the final API, in regard to the impurity profile and the specific process for that intermediate. Besides isolated intermediates, non-isolated intermediates should also be evaluated, since steps comprising these may be ones where even stricter controls may be required. For example, an intermediate may not be isolated due to structural instability when not in solution. In this case, concentration and/or potency of the intermediate may be a critical attribute that needs to be controlled.

Either the corporate validation policy or the particular validation plan needs to provide guidance whether the validation exercise at plant scale should be designed so as to “stress” the accepted parameters by operating at the maximum or minimum accepted level (the edge of failure), or to aim for the set point target of each parameter. Some companies have attempted to matrix the parameter limits in an effort to test the extreme limits of process parameters.

This author holds the view that the laboratory development phase is the place to establish the boundaries of the operating ranges, and how combinations between extremes of parameters affect the final outcome of the process. Process validation, by definition, is used to demonstrate the limits of consistency of the process as designed and thus should aim for the set points of operating parameters. From a manufacturing standpoint, it is advantageous to operate based upon the targeted parameter values for batch consistency.

Because the Process Development Report (PDR) is such a key element of validation documentation, it is important for the corporate validation policy to set out at what points interim reports need to be submitted during process development because interim reports make the compilation of the PDR easier to manage.

The Corporate Validation Policy may need to provide some guidance concerning choosing critical parameters since their number strongly affects the breadth and hence the expense of the experimentation.

Some organizations choose to call critical only parameters, that affect the impurity profile and the ability of a material to pass specifications. However, the ability of an output product to meet specification is not the sole indicator that its process is running smoothly. There are sometimes parameters that, for example, effect yield, without affecting the final product’s properties. Nonetheless, variation of yield beyond normal experimental error points towards a process out of control. Whatever the inclusion/exclusion rules, the basis of decision can be either a general policy prescription or something considered only on a case-by-case basis at the level of the Validation Master Plan (VMP).

Validation Protocols

Validation Master Plan amalgamates validation protocols. The hallmark of a process step that is ‘under control’ is that is predictable. The sequence of operations that constitute the step are unequivocally established in advance, the acceptable operating ranges of the critical parameters are set down in advance, and the critical specifications of the output material are set down in advance. So what sets a validation batch apart from any other pilot experiment is that there can be no recourse to any procedural or corrective activity that is not set out in advance in the protocol. The organization cannot redefine success in a process step being validated after the requisite batch is complete and so eliminate deviations between what was predicted and what transpired. The regulators are the referees who decide whether, when the organization executes the protocol, it follows its own procedures and meets its own standards.

Although some experts urge that the protocol should be drafted to be ‘as concise as possible’, protocols are normally known for their verbosity. A process validation protocol will describe in excruciating detail the ‘how to’ for the validation exercise. The protocol resembles a master record in many ways but provides a greater level of descriptive detail regarding the execution of the process, the monitoring of the steps containing critical parameters and sampling instructions for the in-process checks, particularly critical in-process checks. In every case, either a procedure should be described in detail in the protocol or a reference should be provided in the protocol to an already established procedure for handling the task (ie a standard operating procedure). The instructions should not just be sufficiently unambiguous that it would direct a trained operator to choose the correct action. It should be such that a regulatory reviewer can only conceive a single proper performance of the instruction. Certain styles of drafting can condense the presentation without losing explicitness. Where applicable, data should be entered directly into the protocol. If an exercise is particularly complex, a series of protocols may be preferable. Protocols should have places in them for reviewers’ initials or signatures.

One way of looking at the components of a validation protocol is to see that they reproduce many of the parts of a summary report about a set of batches, but unlike the report the protocol is written in advance (remember after developing a process you are supposed to be able to predict everything that is crucial to product safety in advance).

The protocol should include many sections that process chemists would likely consider report padding. There should be an introductory section defining the scope of the protocol relating back to the overall Validation Master Plan. The question where does this particular protocol fit into the overall validations should also be answered. The responsibilities for execution, testing, review and approval for the overall validation exercise need to be reproduced here even if they also appear elsewhere. Also, for each particular protocol the department and job responsibilities need to be restated and the entire protocol needs to be approved by the top technical officer and the top quality officer. The purpose of the sign-off is to make unequivocally clear that top management has seen and bought into the exact protocol exercise that is to be executed.

Because a description in words alone may be more difficult to follow than an illustration, a process flow diagram (PFD) for the particular step or portion of a process that is being validated should be provided annotated with starting materials, reagents, solvents, process conditions, transfers, by-product concentrates, waste streams, and purifications etc. If there are any recycle loops or patches prepared in advance these need to be presented here in advance. The operations should be assigned an identification code so that further elaborations can be related unequivocally to the correct portion of the PFD. If seed crystals are used where they come from, what quantity they are what specifications they meet need to be set down. It is generally important and advantageous to identify critical process steps containing one or more critical process parameters on the PFD document. The idea is to give the operators no ‘wiggle room’ as to how the execution of the exercise is to proceed. This in turn gives regulators confidence that you have confidence in your process understanding and control.

A complete protocol should provide a bill of materials for every consumable item used during the process. Particular attention should be given to be sure that even materials that are not part of the chemical transformations also be there, such as filters, filter aids, packaging for chemicals before and after the main chemical transformation, seed crystals etc..Some materials have both a name and an identification code. A reference to the specification or COA should be in the listing. If the PFD has the operations in the sequence labelled, this label should be here as a cross reference.

All the equipment used in the process needs to be referenced. The actual documentation does not need to be reproduced in the report but the IQ, OQ and PQ report numbers need to be given as well as the cleaning procedures and the cleaning validation reports and instructions. The ranges of parameters, particularly any critical parameters that will be used should be noted and highlighted. Of course, the required operating range needs to be within the validated range for each equipment. If there is any special procedures that involve the equipment during the process these procedures must be detailed here or referenced accurately.

Facility documentation only needs to be referenced along with the DQ, IQ, OQ and PQ materials. The only recapitulation of the facility validation would refer to the particular importance of any unusual facility capability for the protocol being validated. One example of such a special competence would be the inclusion in the facility of Class C clean rooms.

Critical Parameters

Only at this point do we get to elements that process chemist would consider the essence of the scaled up inventiveness. Specific page references should be made to the Development Report where the critical parameters are established and the data used to establish them is organized. Various ranges of the critical parameter need to be discussed. The master procedure range is the range that will be set out in the master batch sheet. Within that range will be the set point which is the numerical setting used on the regulating equipment. Under regular operating conditions the set point can only be transiently met because of the equipment’s inherent limitations to overshoot and undershoot. The master procedure range of course cannot be less than the range set by the equipment’s inherent limitations. The next wider range is the verified range which is the range actually experienced in the plant. The Development Report may make predictions of a wider verifiable range from laboratory experiments. Using development data, a prediction will also be made of the process limits for the parameter. This is the edge of failure exceeding which the API is expect to suffer quality repercussions. Another limit that may be mentioned is the practical range of the equipment with respect to the critic parameter. For example in a steam-heated reactor the practical limit for external heating would be 150 C. For each critical parameter the controls that will be applied to maintain it need to be listed.

In-Process Tests

Every batch sheet comprises in-process tests that are used to assess the progress of the processing. A list is required of all the in process tests within the protocol showing at what point the sample for testing is taken, what the test method will be, who will perform the test and how the results will be reported. If the test is a go/no go test which is repeated until the go signal is obtained this should be explained. Achieving a particular result may be a critical parameter. If this is the case that needs to be made clear. For example neither the reaction temperature nor the time may be critical but the outcome of an in-process test for reaction completion may be critical. Thus a lower temperature and longer time or higher temperature and shorter time may each achieve the completeness required to pass a critical IPC.

Sampling Plan

Process development chemists know that sampling heterogeneous reaction mixtures does not give results representative of the mixture in its entirety. Even testing of the separated phases individually is fraught with difficulties. Solids that are to be sampled are rarely homogeneous. Even homogeneous solid mixtures can demix with mild shaking. As a consequence if one is going to have a significant analytic result, the sampling must achieve a representative sample. Within the protocol, the plan should indicate the sampling location, the amount of sample to be taken, how it is to be collected, how it is to be stabilized if necessary, how stored, and how transported to the testing facility. Both operators and regulators need to know before the exercise exactly how the sampling is planned. Of course at what point the test will be done must be indicated unequivocally. If there is a sampling deviation, it is helpful if it is reported in the batch sheet before any result for that sample is available.

Acceptance Criteria

All criteria that must be within predefined ranges must be listed in advance both critical parameters and IPC results and testing on the final product.

Deviations and Investigations

The most desirable outcome from executing a validation protocol is that all critical parameters are held within the operating range and the resultant intermediate has test measurements within the range of intermediates that have provided safe and effective IPA. That is what we aim for. What if some critical parameters are measured outside the master procedure range but not outside the verified range? What if some critical parameters have strayed outside the verified range but the intermediate seems unchanged in properties from intermediate prepared without deviations? What if the migration out of bounds was brief or very small in magnitude? What is very important is that as much as possible the assessment of the impact of a possible deviation should be made before the validation is actually executed. Identifying a parameter as critical when it is not critical is unlikely after a thorough development but getting the limits of failure somewhat incorrect is rather likely. One does not want to abandon or fail a validation batch when in fact it will result in pure safe final IPA. Falling outside of most critical parameter ranges is easily avoided but sometimes the desired operating range falls close to a limit of failure because the manufacturing needs to achieve some objective not related to final product safety or effectiveness but to minimizing cost, improving safety or reducing waste. It is in these situations where preliminary considerations of less desirable outcomes can be invaluable.

Equipment and Facility Qualification

Equipment Design qualification (DQ) relates to equipment/machines. Design qualification is normally the responsibility of the equipment vendors. Successful Installation qualification (IQ) demonstrates that the equipment has been properly installed. Operational qualification (OQ) demonstrates that the equipment can achieve the operating parameters intended when exposed to typical substrates. An OQ does not not need to specifically test the conditions of a particular process. That is the function of performance qualification (PQ). In general, it is best to have all facility and equipment qualifications completed prior to process validation (certainly all IQ/OQ should be completed). Performance Qualification (PQ) can be performed during process validation, but it is generally advisable to have completed this testing prior to process validation since it would add additional risk to the exercise. However, there may be some instances where it is useful to concurrently perform an equipment or facility PQ during process validation. In those cases, a separate PQ protocol should be drafted and referenced in the process validation protocol.

Analytical Validation

The validation of the required analytical methods that are used to follow the trends of the critical parameters during the process experimentation may be regarded as a sub-program to the overall validation project. Secure conclusions can only be derived from reliable data collected using validated methods, so methods validation needs to proceed hand-in-hand with the examination of the chemical transformation processes themselves.

Cleaning Validation

Equipment cleaning methods need to be included as part of the Validation Report. Cleaning validation is also sufficiently complex that it is better treated as a sub-discipline with its own report sub-section. Some API producers prefer to have standard cleaning procedures for their equipment which are always executed irrespective of the process that is being cleaned up. These methods are part of the SOPs for operations. Where these methods alone cannot be expected to satisfy the cleaning requirements additional cleaning may be set in place as part of the Master Batch Record following a particular exposure. Another organization might specify the cleaning treatment based on what the chemical step has exposed equipment to. That decision can be part of the Corporate Validation strategy. Cleaning methods can be developed in parallel with the chemical process development and cleaning validation can conveniently occur either prior to or concurrent with process validation. These decisions can also be policy decisions or part of a Validation Master Plan for a particular validation.

Cleaning of reactors in which any broadly insoluble, polymeric, or tarry by-product is formed can expend valuable reactor time and penalize throughput. When process chemists include such steps in a process going to scale up they should take a special responsibility to work with production personnel to develop efficient cleaning or at least pass along a warning of the difficulty and whatever they already know about overcoming it.

Conclusion

In conclusion, although process chemists need not become validation experts and although they may not be personally enmeshed in the validation exercise, some overall familiarity with the subject, from their own perspective and emphasizing their unique role, can enhance their work and their enjoyment of it.

Is Glycerin a General Green Solution to Solvent Recovery and Replacement for Process Chemistry?

kilomentor | 13 June, 2011 07:15

At laboratory scale, when one solvent needs to be replaced with another, the solution contents are placed in a r.b. flask, set spinning on the vacuum rotary evaporator with appropriate heating and condensing facilities and when the first solvent has been completely evaporated then the required new replacement solvent is added and the solutes brought back into solution by swirling, scrapping and heating. There is no equivalent procedure for solvent exchanges at scale. In the plant situation, because stirring and heat transfer become ineffective below the minimum stirrable volume, the solvent and substrates combined volumes can never be taken below the minimum stirrable volume. Thus, solvent is never completely removed so no essentially solvent free state arises. When the solvent replacement is of a more volatile one with a less volatile one, the latter solvent can be used to chase the former in some form of distillation; but when the replacement solvent is of more volatile one, tricks, such as using azeotropes, come into play. Even where an azeotrope can work there are still fractions of mixed solvents that become environmentally troublesome waste. It is proposed that a general solution to this problem is to use the minimum stirrable volume of glycerin as a chaser for the first solvent. Glycerin is cheap, biodegradable and has a bp of 182C @ 10 mm Hg. It would be expected to remain behind in a standard distillation when combined with any of the common organic reaction solvents. Even DMSO (bp 189), DMF (bp153), NMP (bp 81-82 10mm hg) would be expected to be chased by glycerin. All the first solvent could be distilled because the volume in the reactor would not go below the minimum stirrable volume. The less volatile solutes from the starting solvent would be retained in the glycerol although not necessarily kept in solution. Advantageously glycerol is immiscible with many common organic solvents such as hexane, methylene chloride, acetone, chloroform, benzene, probably toluene. Thus the solutes can be extracted from the glycerin phase into the new lower boiling solvent. If the transfer is inhibited by partition coefficients, the glycerin can be diluted with some water to perhaps improve the transfer. Whatever glycerin is taken into the new lower boiling solvent can be removed using a drying agent that strongly binds it such as calcium chloride. The distilled first solvent contaminated with traces of glycerin upon simple treatment might be ready for reuse. Thus the first solvent is no longer a waste and there are no mixed fractions of solvents to dispose of. The waste glycerin is a biodegradable material and the quantity used is no more than the minimum stirrable volume of the reactor.

kilomentor

Separation of Diethyl Malonate and Monosubstituted Diethyl Malonate: Another Phase Switching Hdrolysis

About Extracurricular Training

An Overview and Update of the Kilomentor Blog

A Large Rigid Acid for Making Crystalline Salts : 4’’-n-pentoxy-[1,1’:4,1’’]-terphenyl-4-carboxylic acid

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

More Discussion of Exhaustive Digestion in Chemical Process Development

Allophanates- Potentially Useful Reversible Highly Crystalline Solid Derivatives of Alcohols

The Chemical Manufacturing Process Route Selection: The Inclusion of Phase Switching and its Relationship to Validation

A Question from the LinkedIn Organic Process Research & Development Group that Fosters Practicing Creativity

The Strategy of using Split Runs in Chemical Process Development

How thinking about Separating Acetic acid and Acetic Anhydride can sharpen our Chemical Process Development Skills

An Example of a Possible Kilomentor Type Separation at Scale

What Might be the Best Solvent for Difficult Cleaning Jobs on the Reactor Walls of a Plant Reactor

The Importance of Understanding Process Validation for Pharmaceutical Process Chemists

Is Glycerin a General Green Solution to Solvent Recovery and Replacement for Process Chemistry?

Navigation

Menu

Search

Calendar

Recently...

Categories

Syndicate

«	May 2012					»
Mo	Tu	We	Th	Fr	Sa	Su
	1	2	3	4	5	6
7	8	9	10	11	12	13
14	15	16	17	18	19	20
21	22	23	24	25	26	27
28	29	30	31