kilomentor

Kilomentor has already written about this subject. See the blog, Making a Good Recrystallization Better.

Kilomentor consistently prefers processe that utilize iintermediates that can be protonated or deprotonated at pHs accessible in water and which therefore can undergo phase switches as part of purifications. Such routes are more rugged to the extent that they are less depend upon purification by crystallization of intermediates whose physical properties at the time the route is designed, are unknown. Nevertheless, even with ones best efforts to use routes where the intermediates are acidic or basic, once the molecular weight of intermediates exceeds what can be practically distilled, crystallization is the predominant isolation and purification method for neutral-unionizable intermediates.

Crystallization is of such importance that it is taught early on in (what there is of) laboratory training in universities. The disadvantage of this is that treatment is elementary and laboratory scientists never seem to get around to more sophisticated discussions about it, but learn what more they can by experience, both good and bad.

The crystallization / recrystallization procedures are probably the most important in the laboratory, yet the basic principles that underlie the operations are not widely understood by the average practitioner and rule-of-thumb insights are not passed on.

The most common mistake relating to crystallization is try to crystallize without first applying other methods of purification. The deleterious effect by impurities upon the rate and completeness of crystal formation for the main product is so pronounced that crystallization of crude products should never be attempted until other methods of purification have been applied This is particularly true because these pre-treatments are often very simple. Although crystallization has a high probability for successfully removing impurities, many synthetic chemists do not recognize that they pay an unnecessarily heavy penalty by recrystallizing crude products because the loss from incomplete sluggish crystallization is avoidable. Consideration should be given to first:

· distilling in vacuum

· co-distilling with another solvent

· steam distillation with or without the presence of salt

· superheated vacuum steam distillation;

· exhaustive digestion with a poor solvent

· extraction in a continuous extraction apparatus;

· passage through a plug of a solid adsorbant;

· or acid base extraction (to remove acidic or basic impurities even when the major compound is neutral)

· treatment with derivatizing reagent to trap identified (or guessed impurities)

· charcoaling

· treatment with ionic exchange resins

· treatment with scavenger resins

Water and other solvent residues count as impurities that reduce the rate and completeness of crystallizations. Of course these solvents also play a part by affecting the actual crystalline substance that is obtained since they form solvates and hydrates. Thus adding water during the work up means that a thorough drying and pumping down needs to be done before crystallization. This in turn leads to the question whether the traditional pouring into water is always a good idea.

Aldehydes and Ketones are among the most common and best understood functional groups in organic chemistry; however, they can be problematic as intermediates in large scale process chemistry because they are neither markedly acidic nor basic and so cannot be extracted as salts into aqueous solution to achieve purification by phase shifting. During the route planning stage, the synthesis creator cannot readily guess whether these carbonyl intermediates with be crystalline or not. Aldehydes and ketones that are neutral and have no other extraction handle thus are potentially isolation and purification problems. They can turn out to be oils or low melting solids.

Low molecular weight aldehydes and ketones are most often purified by fractional distillation either at atmospheric pressure or under reduced pressure. When they have boiling points in the neighbourhood of 200 C, steam distillation can provide a partial fractionation, but steam distillation is almost always unacceptable because of the very high large point of maximum volume inherent in the procedure. Compounds such as 7-tridecanone [m.p. 30-32.5 C; b.p. 264 C]; 2-pentadecanone [m.p. 7-41 C; b.p. 293 C]; or 2-heptadecanone [m.p. 47-51 C] are representative of these in-between type substances. So although one cannot say with certainty that an intermediate molecular weigh,t neutral carbonyl compounds is going to difficult to separate/purify it is good to have some precautionary potential patches in mind.

Although oximes derivatives of carbonyl compounds are not completely dependably solids, the likelihood that the oxime is a recrystallizable is more than for the carbonyl itself and increases as the number of carbons increases. Shriner, Fuson and Curtin in their classic manual, The Systematic Identification of Organic Compounds, A Laboratory Manual Wiley 1964 report that out of 63 liquid ketones, 44 had solid oximes. Out of 44 liquid aldehydes, 34 had solid oximes, some of which were separated into syn and anti forms.

If the oxime is not a solid, then the possibility for making the oxime hydrochloride adds an additional opportunity to get one’s hands on a crystallizable solid that can be easily converted back into the original carbonyl. It is not routine for synthetic chemists to think of oximes as substances that can be converted into addition salts, because we more typically think of oximes as being reactive with acids to give Beckmann rearrangement products, but in fact the oxime nitrogen is reasonably basic and can produce acid addition salts with mineral and other strong organic acids. These salts can solidify and provide a means of phase shifting (from liquid or solution) to solid that can provide a basis for purification. In Organic Syntheses Coll. Vo. V pg. 266 2-chloro-cyclooctanone oxime in trichloroethylene solution was converted into an oxime hydrochloride by blowing in hydrogen chloride gas. When the solvent was removed the oil solidified to give oxime hydrochloride in 100% crude yield. It seems likely that all that is required to provide an isolable salt is the addition of a strong acid to an anhydrous medium with the oxime. Another non-solid strong acid that could be considered is the liquid acid, dichloroacetic acid.

Whether it is the oxime or the oxime addition salt that is isolated, whether it is as a crystallized or precipitated solid; obtaining the solid provides the opportunity foradequate purification. Then, either the oxime or oxime addition salt can be converted back to the carbonyl in high yield by a variety of well documented treatments. Successful application of this strategy would be one more demonstration of the concept that it is not always the protocol with the fewest identifiably steps, or the fewest chemical reagents, but the one that is simplest to execute and most rugged that is best suited for scale-up and cost minimization.

Intuitively it seems sensible that to obtain the highest melting derivative it might be sensible to derivatize with the highest melting derivatizing agent. Although there is no firm basis for such a prediction, there is some enhanced likelihood of this being so. The derivative will share some partial structure with the derivatizing agent and if features of that agent contribute significantly to the high melting property and if those features are preserved in the derivative’s structure, some of the high melting characteristic could be anticipated to be retained.

The highest melting pharmaceutical salt former for reaction with basic substances is orotic acid. Its melting point is around 325 C. Its structure contains multiple hydrogen bond donors and acceptors that most likely contribute the melting character and these hydrogen-bonding sites are retained in the pharmaceutical salt formed.

Despite this the orotic acid salts are rarely used.

The alkali salts of orotic acid are poorly soluble. Orotic acid is also poorly soluble in water. Thus the sodium salts can be precipitated by adding a solution of the N,N-dimethylethanol salt of orotic acid in 80% aqueous ethanol. This might be tried as a method for removing an alkali salt from an alcoholic solution of an organic compound. The alkaline cation as an alkaline salt would be precipitated with orotic acid. The residual orotic acid would also be essentially insoluble. N,N-dimethylethanol would be left in solution but this could be removed by extraction with a neutral organic solvent.

Orotic acid or a salt of orotic acid might be expected to form a complex with triphenyl phosphine oxide which is a good hydrogen bond acceptor. Orotic acid has an imide NH which is a good hydrogen bond donor.

Orotic acid may be a good substance to neutralizes aqueous base since both the alkaline salt and the free acid are essentially insoluble.

The unusual solubility properties of orotic acid and its metal salts make it worth bearing in mind when trying to isolate organic bases particularly because, since it is pharmaceutically acceptable, trace residues are not too critical.

It is the flavour f the month for Organic Research & Development Process Chemists to modify laboratory synthesis steps, which are environmentally troublesome, with green equivalents. Although this is often a priority pressure in the opposite direction can arise because, to meet deadlines for specified quantities, it is more straight forward to keep the original chemistry. This problem can come up rather frequently when the reaction ‘sinning’ environmentally is a classic such as the chromic acid oxidation.

Classic status

Chromium(VI) has been a popular oxidant in the laboratory for 70 years. It has been used in many complex reagent combinations. It has appeared in many of the volumes of Fiesers Reagents for Organic Synthesis. Twenty three procedures using a chromium (VI) appear in Coll. Vol. I-IV of Organic Syntheses.

Unfortunately all the variants of this oxidant present problems for a scaled up procedure. Some: (Sarrett and Snatze) can lead to fire when the oxidant is being prepared . Some variants use undesirable solvents (benzene, diethyl ether, HMPT, pyridine, acetic acid and methylene chloride). What all share is that chromium is a heavy metal that cannot be discarded in waste-water. Compounding these difficulties, the work-up of many of these oxidations produce inorganic gases, and large volumes of waste extraction solvents. Traces of residual chromium not removed quickly in the isolation are known to lead to by-products during the solvent concentration steps in f the isolation protocols causing products which ought to be white, to be yellowish.

Chromium (VI) in many variants have trivial names

Jones Reagent H2CrO4/ sulfuric acid/ sodium dichromate/ acetone

Brown-Garg H2CrO4/(ether or benzene)/water

Kiliani Reagent H2CrO4 /H2SO4/ water/ acetic acid

Chromic anhydride CrO3/ water/ acetic acid

Fieser Reagent CrO3/ acetic acid

Sarrett/Ratcliffe Reagent CrO3/ pyridine

Cornforth Reagent CrO3/ pyridine/ water

Thiele Reagent CrO3 /acetic anhydride/ H2SO4

Snatze Reagent CrO3/ DMF

and with no special name

CrO3/ acetic anhydride/ acetic acid

CrO3 /HMPT

The reagents can be used with co-catalysts like mercuric acetate, ceric ammonium nitrate, manganous nitrate, oxalic acid (F.& F. Vol. 9 pg. 114) and special effects come with controlled amounts of water in the reagent.

Compounds that contain double bonds often suffer from the presence of isomers in which the double bond is in a different location. These contaminants are usually very difficult to remove. Hexachlorocyclopentadiene is a reagent that reacts readily with a double bond that is not conjugated to any electron withdrawing group. This is precisely the situation that is most likely to be difficult to separate. Hexachlorocyclopentadiene is reactive enough that it adds twice to naphthalene molecules once at each of the two double bonds in an unsubstituted ring. This use in fact is its best known application where it acts as a protecting group for an unsubstituted ring in a naphthalene.

There is also a literature report that ordinary alkenes react at different rates with hexachlorocyclopentadiene. 1-Octene has an activation energy of 20 kcal and 4-methyl-1-cyclohexene has an activation energy of 24 kcal. These exemplify Diels-Alder reaction occuring with unactivated alkenes. Although hexachlorocyclopentadiene is expensive when obtained from Aldrich, it is still probably available somewhere in bulk inexpensively. The disposal of the chemical must be handled carefully since it is an environmental toxin.

The adducts of this reactive dienophile are reported to be split back into their precursors by heating with phosphorus pentoxide and distilling the non-chlorinated adduct. This is an inconvenient, inefficient and damaging procedure that cannot promote widespread use. Using the adducts to remove an unwanted contaminant by reaction however would not have this disadvantage since reaction with an impurity would not need to be reversed. An excess of the hexachlorocyclopentadiene could be used and when the reaction had proceeded to the most beneficial extent, as judged by in process analysis, the excess could be quenched with maleic anhydride followed by mild base extraction of the hydrolyzed adduct as sodium salt. This would leave the hexachlorocyclopentadiene adduct of a more reactive alkene impurity and the desired less reactive product. This should prove a much simpler separation.

The titration of a mixture of reactive alkene, hexachlorocyclopentadiene and adduct with bromine gives the residual amount of the olefin because neither reagent or the adduct react under ice temperature conditions.

Dear regular readers and occasional browsers, my company, ratiopharm, has been taken over by the largest generic company in the world, Teva. The regulatory approvals and other closing issues still need to be dealt with, but sometime around December of this year the senior management of Teva will take over. The result is an uncertainty which, frankly speaking, I am unaccustomed to in my professional career. Scientists who continually work at improving themselves, expanding and upgrading their skills, are usually the master of their own fate. I expect if you have been reading my blogs, you are also in this category. Put bluntly we have never had any anxiety about finding work because we are the most sincere, most desirable of high skill employees. But with a takeover like this, by a company that is known for its fair treatment of its own long-time employees, there are more considerations than just technical excellence and the ability to function in a team being considered; costs must be cut and old loyalties respected. I am at the age where others may think that the best solution is to retire me and keep the younger guy or gal. The problem is chemistry, formulation and intellectual property are more interesting than anything I can contemplate in retirement. Besides, I do not think of my pay as a cost of doing business but as an investment in growing business. I do not think of my employment as taking a job someone younger could have, but of creating values that will allow my company to grow and hire several more people.

So, if I want to keep doing what I love, I had better do my part and take some proactive steps to enable it. Hoping for the best while preparing for the worst is always good advice.

Acting on this principle, you will read below a resume from Dr. Clarke Slemon, alias Kilomentor. The blog shall then return, for as long as it lasts delivering what you came to this space to find: interesting considerations about chemistry.

CLARKE SLEMON Ph.D.
#43 4635 Regents Terrace,
Mississauga, ON, Canada, 75R 1X5slemon.khomasurya@sympatico.ca
(905)-755-1296; (home)

OBJECTIVE
Team with self-confident people, as a Chemical Process Development Specialist, Pharmaceutical Formulator or Patent & Legal Specialist.

MY SKILLS SUMMARY

Patent & Legal Specialist; Harvard Ph.D.; API Process Development Chemist; Solid and Liquid Dosage Formulator; Database Search Specialist; New Product Development; cGMP; Research Manager-Director; Synthetic Organic Chemistry; Medicinal Chemistry

EDUCATION

HARVARD UNIVERSITY, Cambridge MA
Ph.D. Organic Chemical Synthesis, Harvard University
Prof. R.B. Woodward (Nobel laureate) supervisor

CARLETON UNIVERSITY, Ottawa, ON
B.Sc. (Hon.) Chemistry
Prof. John ApSimon supervisor

EMPLOYMENT HISTORY

ratiopharm Inc., Mississauga ON
April 2005-present

Patent Manager

A member of a Patent & Legal team regarded by many law firms as best-in-class. We created a unique style characterized by intense, iterative, interplay between our technical professionals in ratiopharm and our lawyer representatives that operated throughout the process leading to an NOC.

As a person evaluating all the IP relevant to developing formulations, both internally and through external contractors, I became intimately familiar with all patents and patent applications relating to solid dosage pharmaceutical formulation. The Patent & Legal Department prepares a patent landscape report relating to formulation for every project that goes through development in Mirabel Quebec. More recently our reports are prepared for projects world-wide. The formulation department communicates regularly with our Patent Group to assess difficulties and plan new work. I am the only professional in ratiopharm with experience and training in each of API synthesis, Galenic development, Analytical development and Intellectual property. It has been my personal goal to read and compile everything pertinent to the development of generic bioequivalent formulations in a cost effective manner. This has been greatly assisted by the cooperation of the Director of Clinical Development.

I have experience evaluating the synthesis processes of API contractors and providing patent clearance for products and processes. This the second major aspect of my work.

Also, as the Patent & Legal representative on 1/3 of the company’s product development core teams, I am fully conversant with the entire product development process.

LLOYD, INC., Shenandoah IA
September 2003-April 2005

Lloyd, Inc. creates, manufactures and markets pharmaceutical products for companion animals.

Director of Research & Development

Within 4 months I successfully scaled a wet granulation tablet formulation by a factor of 1000 that included a change in reactor geometry and stirring mode. I produced with this formulation a prototype for a new chewable pet nutraceutical; designed two novel scalable routes to an animal injectable product; filed provisional patents; prepared and filed another provisional patent for a new product to treat an unmet medical need for canines. The company academic consultant, described the concept as ingenious. I worked with marketing and sales to revitalize moribund new products development collaboration. As a member of the New Products Development team, I investigated new licensable technologies. Intensive on-line searching, brought to light technology directly applicable in our own ongoing research. I heightened departmental focus on goals and improved reporting mechanisms; obtained hands-on and textbook experience with blending, granulation, milling and tableting.

QUEBEPHARMA RECHERCHE INC., Montreal PQ
June 1999-December 2002

QuebePharma Recherche Inc. was a Canadian controlled and Canadian managed small business corporation engaged in organic synthesis and process development related to drug development. I founded the business in 1999. In 2002, I sold my part ownership in this company.

President and Research Director

I steered research, building IP for shareholders and managed contract research for select clients; arranged office support so that 90% of my time was dedicated to accelerating the laboratory research; conceived the research program; designed the laboratory/offices and hired collaborators. Intensive on-line searching, patent planning and an emphasis on extensive bench experience for all associates were hallmarks of my program.

I strategized to design new substances from lead structures related either to promising nutraceuticals or active natural products. From this work came experience related to peptide and peptidomimetic; scaffolds for combinatorial libraries; advancing hits to leads and managing outside collaborations. The opportunities refined my good grounding in biochemistry and molecular biology. Research over the period resulted in three patents and eight trade secrets.

TORCAN CHEMICAL LTD., Aurora, ON
December 1987-June 1999

Torcan Chemical Ltd. was a research, development and manufacturing organization based on synthetic organic chemistry that was founded in 1980. The facility has successfully passed FDA inspection and follows cGMP and GLP practices. The company has earned industry wide recognition for its process innovation and supply of quality products for both clinical and commercial applications. It has become part of Piramel Healthcare.

Research Manager

I reported to the president. My research created intellectual property that s broke through technical logjams and solve business problems. I gained a thorough knowledge of patents, patent infringement and other intellectual property issues as the company patent liaison. At this time. I studied for the patent agent examinations; performed complicated literature searches for the entire company, including substructure searching, patent searching, citation searching and taught scientists search methods, while continuing my own research projects and directing my group.

Senior Research Scientist

Developed new chemistry circumventing patent difficulties. With a process development team of two co-workers, we developed and implemented in the plant, a process for the product with the second largest sales in the company. My group completed many projects that had resisted the efforts of others. I introduced statistical strategies of experimentation that efficiently optimize reaction conditions while simultaneously identifying sensitive variables. These methodologies allowed rapid reaction optimization.

OUTSIDE INTERESTS

KILOMENTOR
(December 2006-present)

The KILOMENTOR blog was set up to provide for free, chemical process development and synthetic organic chemistry training to anyone anywhere in the world with access to the world wide web. The project has been cleared by ratiopharm of any conflict of interests. The project aspires to level the playing field for chemists, world-wide, irrespective of their personal circumstances. By laying open my ideas and experience in synthetic chemistry, I try to exemplify the ideal of independent generous partners in a meritocracy.

The blog content is at the post-graduate level. By April 2010, the number of my blog articles was more than 80 and the total visits to the site had surpassed 76,000. This resource complements process development textbooks and Scientific Update conferences.

In one year time, I think it is likely that I will be retiring from my present position at ratiopharm. For one thing, I will be 65 years old, and for another, ratiopharm, by that time, will have been sold off to some other organization. Most likely ratiopharm will be part of either Teva, Pfizer or Actavis. Anyway, those, Reuters reports are the front-running contenders. Since the ratiopharm organization did not itself produce any active pharmaceutical ingredients but sourced everything outside, the organization didn’t mind me writing a blog whose purpose was to mentor process development chemists anywhere in the world. I do not suppose these new people will feel the same way.

I am still in good health and cannot imagine doing anything else as enjoyable as chemistry. The action to-day is in the developing world, so I wouldn’t mind getting back into process development /research management / mentoring there with a vigorous team of young scientists determined to efficiently deliver significant projects. My wife, who is Chinese, favors China which clearly moving the fastest. As a Canadian with British roots, I may be inclined somewhat more towards India, where I have visited and appreciated what I saw. Of course there are other potential destinations as well. Any suggestions from out there in cyberspace?

Regards,

Dr. Clarke Slemon

kilomentor@sympatico.ca

Kilomentor takes the position that, in the present state of the chemical art, electronic database searching has enabled chemists of ordinary skill to design ingenious reactions schemes by little more than electronically searching for reactions to string together. It is isolation and purification procedures where there is the least technical support to support individual ingenuity. Isolation methods cannot be searched because the search terms are the solution to the problem not the starting point.

Therefore Kilomentor wants to emphasize where inexpensive transition metal complexes such as those with Chromium (III) and Cobalt (III) can simplify the work-up of chemical process steps.

Chromium (III) is the most stable and important oxidation state of the element in general and particularly in the aqueous chemistry. Advanced Inorganic Chemistry 1966 pg. 823 states, “The foremost characteristic of this state is the formation of a large number of relatively kinetically inert complexes. Ligand displacement reactions of Cr (III) complexes are only about 10 times faster than those of Co(III), with half-times in the range of several hours. It is largely because of this chemical inertness that so many complex species can be isolated as solids and that they persist for relatively long periods of time in solution, even under conditions where they are thermodynamically quite unstable.” Note that it is the kinetically inert property of chromium complexes that makes them valuable. What this is saying is that complexes that are not the thermodynamically most stable nevertheless can be isolated. This means that many more compounds are in principle accessible.

Advanced Inorganic Chemistry 1966 pg. 873 says about cobalt chemistry that “The complexes of Cobalt (III) are exceedingly numerous. Because they generally undergo ligand exchange reactions slowly, but not too slowly, they have, from the days of Werner and Jørgensen, been extensively studied and a large fraction of our knowledge of the isomerism, modes of reaction and general properties of octahedral complexes as a class is based upon studies of Co (III) complexes.” What I take this to be saying is that many different complexes of cobalt would also be readily accessible in principle.

Iron also appears to be promising in terms of offering multiple potential complexes. Iron (III) forms a large number of complexes, mostly octahedral ones, and octahedron may be considered its characteristic coordination polyhedron. The affinity of iron (III) for amine ligands is very low. No simple amine complexes exist in aqueous solution; addition of aqueous ammonia only precipitates the hydrous oxide. Chelating amines, for example, EDTA, do form some definite complexes among which is the 7 coordinate [Fe(EDTA)H2O]ion. Also, those amines such as 2,2’-dipyridyl and 1,10-phenanthroline which produce ligand fields strong enough to cause spin-pairing form fairly stable complexes, isolable in crystalline form with large anions such as perchlorate.

Transition metals now have an extensive application as catalysts in organic chemistry. Nickel, palladium and platinum complexes are today extensively used to catalyze reactions for which there is no uncatalyzed equivalent.

An extensive chemistry has also been established centering on the practical question of the recovery and recycling of the noble metal catalysts, mainly palladium and platinum, since these represent expensive inputs into a process.

From the Kilomentor perspective of using of transition metal complexes for isolations the complexes of the wide variety of less expensive chromium, cobalt and iron complexes would seem most promising.

As a first example let us consider Reinicke and Rhodalilate Salts:

The Chromium Salt NH4[Cr(NH3)2 (SCN)4] is red in color. It is soluble in ethanol or hot water and is reported to dependably yield precipitates with primary and secondary amines. The implication of many reference books seems to be that the salt does not form precipitates with t-amines, but this is untrue. According to Cotton & Wilkinson’s Advanced Inorganic Chemistry Comprehensive Text, it can be used, in general, to precipitate large cations, either organic or inorganic. It seems likely however that although, thermodynamically, precipitation of Reinecke salts may not be as selective as has been publicized, fractional precipitation based on rates of precipitation can provide purification as suggested for the closely related Rhodanilate salts (see later for rhodanilate definition).

The Reinecke and the related Rhodanilate salt possibly could be used to precipitate particular amines in the presence of others. One idea is that because the Reinicke salt is soluble in alcohol alone, a useful separation could be done on a substrate, which is sensitive to water. An example of this is that the possible difference in rates of precipitation might be useful in the case of alkylating of an amine where an excess of the starting amine could perhaps be selectively precipitated.

Amines are frequently used as reagents to neutralize acidic co-products of a reaction and thereby drive any equilibrium towards completion in a particular direction. The most frequently used amine in this regard is triethylamine. Some advantages of triethylamine are that even if it is employed in excess any unused base is

i) volatile enough to be removed by vacuum

ii) water soluble enough to be carried away in a water wash

• iii) inexpensive enough to be discarded and
• iv) easily made anhydrous.

Disadvantages are that it is volatile enough to escape from reactions that require heating and nucleophilic enough to compete in some displacements and deprotonations. Employing the Reinicke or Rhodalinate salts in a work-up of mixtures containing more complex amines may make them recoverable and recyclable and so practical as traps for acidic co-products.

Another possibility is that initial formation of an easily isolable amine salt of a complex anion X could be followed by the switch from the amine salt to the inorganic metal salt via a Rienecke or or Rhodalinate reagent which could precipitate the intermediate amine.

For example, a salt of an amine with a complex anion X might be converted into the salt of a metallic cation MX by adding that M in the form of acetate and precipitating the amine (here R3N) as the Reinicke salt precipitate.

M^{+ -} OAc + R3NH⁺ X^- + NH4[Cr(NH3)2 (SCN)4] going to

R3NH [Cr(NH3)2 (SCN)4] (insoluble) + NH4 OAc + M X

I do not know of any experimental examples of this, however.

Amine Recovery from lithium amide reagents

The lithium salts of many sterically hindered secondary amines, lithium diisopropylamide for example, are used for quantitative deprotonation in chemical synthesis. Because they are sterically hindered the resulting secondary amine co-products do not interfere in subsequent reactions of the carbanions they helped create. These sterically hindered secondary amines may need to be separated from desired product in the reaction work-up and if they are expensive recovered for recycling.

Reinicke Salts, Rhodanilate salts or Trisoxalatochromate salts can potentially be used to precipitate these secondary amines and remove them as filterable solids. Diisopropylamine, dicyclohexyl amine, 2,2,6,6-tetramethylpiperidine, isopropyl-cyclohexylamine, and pentamethylpiperidine need to be examined to see whether they can be quantitatively or semi-quantitatively precipitated.

According to Max Bergmann’s article in J. Biol. Chem.109, 471 (1935) proline and hydroxyproline can be precipitated from gelatine hydrolysates using Reinecke’s salt and the amines liberated by forming a complex with N,N-dimethylaniline or pyridine. This liberation shows that all amines can react.

In this Bergmann article the formation of what he regards as more selective complexing agents can be achieved replacing the ammonia ligands with other amines. Displacing the two ammonia with aniline what is called ammonium rhodanilate is formed About this Bergmann says, “ Rhodanilic acid forms rose-coloured, well crystallized salts with basic nitrogen compounds, and in particular with alkaloids and with amino-acids. Although rhodanilic acid lacks definite specificity, the various rhodanilates differ greatly in their solubilities, crystalline form, and rate of crystallization. It is therefore often possible to separate from mixtures of amines, amino acids , or peptides, single homogeneous products by fractional precipitation with rhodanilic acid. In most cases where several rhodanilates form simultaneously, a separation by fractional crystallization is often possible.”

With regard to the amount of ammonium rhodanilate in the fractional precipitation Bergmann says that “The quantity necessary was determined by examining the precipitate under the microscope in the course of successive additions.” I interpret this to mean that the precipitation was controlled by the kinetics and the fastest precipitating compound came out first followed by other compounds and the precipitate was collected in fractions that were subsequently combined on the basis of their microscopic crystal shape.

In the case of preparing proline rhodanilate, the free amino acid was simply achieved using excess pyridine.

“In order to obtain the free amino acid from proline rhodanilate, advantage was taken of the fact that pyridine rhodanilate is very difficultly soluble in water. It is therefore sufficient to suspend the solid proline rhodanilate in water and to add a little pyridine in order to precipitate almost instantly the entire rhodanilic acid as pyridine salt. On filtration a faintly colored aqueous solution of l-proline is obtained.

The by-product of such a purification is pyridine rhodanilate. It may easily be recycled and reconverted into ammonium rhodanilate with ammonia and so recovered for further use.

Thus ammonium rhodanilate can be used to precipitate a complex amine, the amine rhodaniliate can be freed from the complex with pyridine to precipitate the very poorly soluble pyridine rhodanilate and then the ammonium rhodanilate can be reformed from the pyridine salt by treatment with excess ammonia.

In a finl use, of the Reinicke salt, if mercuric acetate bound to an ion exchange resin is used as a source of mercuric ions in a reaction. Water from such a reaction, that could contain small amount s of mercury ion, can be decontaminated with Reinicke Salt which precipitates the mercury ion.

Dipolar aprotic solvents such as N-methylpyrollidone, dimethyl formamide, N-methyl formamide, dimethyl acetamide and dimethyl sulfoxide are often drowned out with water and then extracted to isolate organic products. No cheap and convenient method has been worked out to separate these polar organics from the bulk of the water and return them to an anhydrous condition suitable for reuse.

Kilomentor at his present employment does not have access to a chemical laboratory where experimentation could be done so the following idea has not been tested, but on the basis of the physical properties of the chemicals it might be workable.

Diisopropyl ether (DIPE) forms an azeotrope with water that is reported to boil at 62.2 C. This is a heterogeneous azeotrope that according to the Chemical Rubber Handbook splits into a water-poor DIPE upper phase and a water rich lower phase. Addition of DIPE therefore to one of these higher boiling solvents and water, and boiling of the ternary mixture under a Dean-Stark trap with continuous return of the top DIPE phase should gradually separate a lower water rich phase which could be periodically drained away. The high boiling dipolar aprotic solvent that is being dried should theoretically be confined to the still pot at the low azeotropic boiling point. In the real laboratory situation, however, a small amount of dipolar solvent vapour entrained in the reflux stream could be all that is needed to prevent the distillate from separating into two phases in the trap and this would scupper the procedure. It is crucial for a practical process that the DIPE be recycled since the distillate is 97% DIPE and only 3% water. Recycling is essential to be able to remove a large amount of water using only a small amount of DIPE.

Other solvents that boil above 100 C that can potentially be separated from water and dried are: nitromethane, acetic acid, dioxane, ethylene diamine, sulfolane and isoamyl alcohol.

After the water has been completely removed continued distillation will separate the DIPE. Even if small amounts of DIPE might remain they are usually unreactive. If particular importance they are inert towards organometallic reagents.

Although mixtures of carboxylic acids or mixtures of amines are each fit for separations based on extractions at controlled pHs, most functional groups are not so easy to isolate from each other. Kilomentor thinks a good deal about how to simplify separating molecules with the same functional group in slightly different structural environments. Although it has not been demonstrated in the literature yet, two different molecules each with a nitrile functionality but differing in the steric environments around them can likely be separated by selective reaction followed by a simple acid-base extraction.

Nitriles are known to react smoothly with azide in a 3+2 cycloaddition to give 1,2,3,4 tetrazoles. This is a ‘click chemistry’ reaction . Cycloadditions are typically quite sensitive to steric environment. Thus, although these reactions are generally fast, it is likely that conditions can easily optimized to get good selectivity between cyanide groups in different molecules using an insufficient amount of azide. The result will leave a nitrile in one substrate untouched and the nitrile in the other substrate converted essentially completely to tetrazole. The beauty of this is that these tetrazoles have the acidity of carboxylic acids and can be extracted into water with alkali. Thus the tetrazole derivative removes the reactive nitrile substrate from the organic phase leaving the unreactive nitrile substrate clean for a simple recovery.

Two references that I could loctate concerning the kinetics of the reaction of nitriles with azide are: Khimiya Geterotsiklicheskikh Soedinenii (1992) (9) 1214-17, which is in Russian [C.A. 11; 8945a, 8948a]; and Inorganica Chimica Acta (1985). 102(2), 157-62 that does the condensation with the nitrile coordinated in a Co(III)complex.

The synthesis of tetrazoles from nitriles and azide has been studied intensively because of its relevance to the preparation of the sartan family of drugs. Relevant patents are US5744612, US6040454, WO2005014602, WO2007054965 and CN1718574

Crystallization is a process step that has for a very long time generally defied attempts to scale it up. A solution is to crystallize by continuous micromixing small volumes and accumulating the resulting slurry in a large reservoir for isolation at scale.

One standard crystallization variant contacts a supersaturated solution of the substrate with an appropriate anti-solvent in a stirred vessel. The anti-solvent initiates primary nucleation as it mixes into the supersaturated solution of active and these seeds then grow. The process is adaptable to using preformed seed crystals and/or further aging of the solid, once formed, which digests the crystals to change their initial sizes and/or polymorphic forms. Using such a methodology, in order to get the smaller crystals preferred for enhanced bioavailability, the saturated solution needs to be added into the anti-solvent to get many tiny seeds forming rapidly. Using this reverse addition technology a concentration gradient cannot be avoided in a large reactor because the introduction of feed solution into the anti-solvent in the stirred vessel does not afford a thorough mixing of the two fluids prior to the initiation of crystallization. The presence both of the concentration gradients and a heterogeneous fluid environment interferes with optimal crystal structure creation and allows greater entrainment of impurities. On scale even the fastest bulk mixing cannot smooth out the microenvironments in which the seeds form. Furthermore, in a large bulk reactor the number of seeds present at the beginning of the nucleation process is very different from the seeds present in the bulk when the last of the supersaturated solution enters the tank. On scale stirring cannot handle the micromixing requirement.

Another standard crystallization procedure cools a solution of the desired product in order to bring the solution to its supersaturation point but cooling in batch processing is a slow process that becomes even slower as the batch size increases. Although the solvent gradient is abolished, it is replaced by a themal gradient. In any case the crystals are undesirably larger with the slower process. The characteristics of size, purity and stability are difficult to control with the technology.

Technology is now coming off patent that may solve this problem in may cases. Particle size is more important for drug substances because of te relationship with bioavailability. Most drugs are provided as pharmaceutically acceptable salts and consequently it is the particle size of these salts that is of most commercial importance.

CA2044706, Crystallization Method to Improve Crystal Structure and Size, expires June 14^th, 2011 in Canada and the corresponding US5314506 expires May 24^th 2011. The invention addresses the general problem, how to obtain a reproducible micronization of a pharmaceutical compound without milling. The patents are assigned to Merck.

The technology taught in CA2044706 teaches pumping both solution and antisolvent in impinging jets of fluid that because of their small volumes and high velocities create almost instantly, where they collide, a region of high intensity micromixing. Once the fast crystallization has occurred the mixture of solution and antisolvent can be accumulated and filtered when all the material has been processed or it can be collected after any other appropriate time.

This impinging jet technology removes the problem of scale. Larger scale just translates into a longer period pumping the same streams together. The heterogenous slurry in which the seed crystals form becomes a function of the pumping rates, the concentration of solute in solvent and anti-solvent and the radii of the columnar jets of colliding fluids. All these parameters are within engineering control. Because of this the surface area, crystallinity, stability and purity can be optimized. Because the equivalent of a milled material is available directly, a step is saved and the noise, dust, yield loss, equipment cost and exposure hazard of milling are all by-passed.

Preparation and Crystallization of Pharmaceutical Salts using Impinging Jet Micromixing Technology.

Some bimolecular reactions proceed in superior yield if both reactants are added simultaneously to a pot of solvent so that the concentrations of both reactants is kept low. Formation of pharmaceutical salts can benefit from such a process. Two impinging jets of solutions, one containing the substrate molecule and the other either the pharmaceutical acid or pharmaceutical base required to make a pharmaceutical salt can be implemented. As above, where the jets collide the intense micromixing both form the salt and cause nucleation into small seed crystals. This is taught in CA2349136. The advantages are the same. The process is completely under control of the engineering, product can be synthesized in micronized form in a single step with a single piece of equipment.. The US equivalent, US6558435B2, expired May 6^th 2003. The Canadian family member application died August 15^th 2007. They are now free to use.

Crystallization of micron sized controlled particle size may be becoming easier not harder on scale!

It has long been known that some chemical reactions that proceed slowly or not at all in solution can become predominant when the reactants are adsorbed on solid adsorbants. This can be dramatically demonstrated if we perform preparative chromatography and leave the material on the prep plate or in the column overnight. Very often you would discover the next day that the anticipated product had at least partially decomposed. Fieser & Fieser’s Reagents for Organic Chemistry has numerous entries for reactions on alumina or silica. For alumina, these are Vol.1 pg. 19-20; Vol. 2 pg. 17; Vol.3 pg. 6; Vol. 4 pg. 8;Vol. 6 pg. 16-17; Vol7, pg. 5-7; Vol. 8, pg. 9-13; Vol. 9, pg. 8-11; Vol.14, pg. 20-21;Vol. 16, pg. 16-17; Vol. 18 pg 16-17. For Silica, (also alternatively indexed as silic acid), the entries are Vol. 6, pg. 510; Vol. 9, pg 4`10; Vol. 15, pg 282; and Vol. 18, pg 319. Closer inspection of these references, however, would reveal that they deal mostly with reactions that don’t occur in the absence of the solid phase. This it seems to me is misleading for the true potential of the methodology. If presence of a catalytic surface can activate new reaction pathways, logically, it might improve the energetics for processes that already proceed to some extent. One can significantly ask whether a useful augmentation of the product can be delivered through this pathway. Adding alumina or silica gel to an otherwise homogeneous reaction system has the potential to selectively catalyze one out of several competing reactions and lead to an optimization of a process step. Even if the possibility is a long shot iit is easy to test this at the process investigation stage. Supposing that a required transformation is not proceeding either at all or fast enough and if raising the reaction temperature is not an option,( perhaps because the bp of the solvent is limiting,) rather than tossing the trial reaction into the waste, we could try adding alumina or silica and observe. There is no down-side because you are working with what will be discarded. In the test, add as much solid as possible consistent with maintaining stirring. The goal is to see something desirable happen. Economics and material handling problems, which are real issues for such a solid catalyzed protocol, are not an issue unless something promising happens. All that needs to occur is that the rate limiting mechanistic step have its transition state energy reduced by adsorbing an intermediary on the solid; but there must be enough solid to provide adequate sites. Another possible mechanism that would lead to a success is for a reversible reaction to be driven forward by the removal through adsorbtion of the product or a co-product. Filtration and washing are easy isolation steps that completely removed the added adsorbant. The ease of work-up in the event of success are the potential advantages.

Kilomentor tries to keep up with opinions expressed concerning chemical process development. Girish Malhotra recently wrote an article, Less is More in API Process Development published electronically at PharmaManufacturing.com. He asked the question, Why does U.S. pharmaceutical industry persist in using complex manufacturing processes to make active pharmaceutical ingredients? He cites as examples US6037483, US6245913, US6867306, US6835848 and US6331638. Am I missing something? None of these applications is from the laboratory of an American company. The authors are without exception Indians either working in India or in one case working in Ireland.

I would dispute many of the contentions made in the article; for example, I think the priorities of innovative manufacturing chemists and generic manufacturing chemists may explain an enormous part of the differences he tries to assert. However, with such examples, I think a discussion need go no further.

Organic synthesis activity devotes a substantial amount of time searching for reactions, which proceed selectively so that difficult separations are not needed to obtain pure products. At the same time it is known that acids and bases, nd most frequently represented as carboxylic acids or amines, are characheristically easier to separate from each other because of the sensitive effect of substituent patterns in their pK_as. It would be very useful if more functional groups could be similarly dependably purified.

Esters are common derivatives of carboxylic acids and esters can be hydrolyzed easily in high yield in moist solvents to the carboxylic acids. These free acids can be esterified in high yield. Is there a capacity to purify reaction product mixtures, comprising esters by selective hydrolysis? This is not a question typically asked by organic synthesis chemists. Yhe answer, although it was known to us during our undergraduate or graduate studies, has probably since faded away.

Esters are quite sensitive both to electronic and to steric factors in their relative rates of hydrolysis. Newman’s Rule of Six states that the rate of hydrolysis of an ester is substantially dependent upon the sum of the number of atoms bonded to the combination of those atoms, which are six bonds away from the nucleophile attacking at the carbonyl. From example, in the literature, it can be seen that isomeric compounds differing significantly (i.e. a difference of four substituents six atoms away) can have rates of hydrolysis under the same conditions differing by a factor of 50 or more.

W shall examine the case in which two esters in two molecules of a product mixture differ in reactivity by a factor of 50. As we shall see the difference in reactivity is variable. Additionally and for simplicity we shall assume that the conditions of hydrolysis are selected to give unimolecular reaction kinetics for both the major and minor isomer. In the case where the mechanism of hydrolysis of the esters are different the separation is often solved simply by the intuitive application of this information.

For the development of a general mathematical treatment,

Let [A]_t be the concentration at time t of the major substance

Let [B]_t be the concentration at time t of the minor substance

Let [A]_t^o be the concentration at time0 of the major substance

Let [B]_t^o be the concentration at time 0 of the minor substance

-d[A]_t/dt = k [A]_t

-d[B]_t/dt = 50 k [B]_t

This would be true as we have postulated if the rate of hydrolysis of B is 50 times that of A.

By integrating and subtracting these equations from each other we see how time is related to the ratios of esters at the beginning of hydrolysis and at any time during the hydrolysis

log {[A]_t/[B]_t} = (50-1) kt/2.303 + log {[A]_t^o/[B]_t^o}

49kt/2.303 = log {[A]_t/[B]_t} - log {[A]_t^o/[B]_t^o}

t = 2.303 {log ([A]_t[B]_t^o/[B]_t[A]_t^o)}/49 k

Thus we can calculate the time required to achieve any particular ratio of esters remaining.

Given that for a unimolecular reaction the rate is related to the half time by

k=0.693/ t_½[A] where t_½ is the half life of the major, slower hydrolyzing reaction constituent under the hydrolysis conditions we can substitute and get

t = 2.303 t_½[A] { log ([A]_t[B]_t^o /[B]_t[A]_t^o)}/( 49(0.693))

The half life of a hydrolysis can be approximated without knowing the structure or the concentration of an ester or even separating it from its mixture with the minor ester. The material is simply hydrolyzed until its residual spot is equivalent to a spot of one-half the original concentration when quantitatively equal volumes are spotted. This can be done simply by following the reaction by tlc. The original molar ratio of the mixture can be estimated either by tlc , by integration of appropriate signals in the nmr or by other means convenient for the particular case.

We know the ratio of esters present in the mixture before hydrolysis which is [B]_t^o / [A]_t^o Let it be 9/1 or 90% the major isomer for example. We can choose the desired enrichment or the ratio of major ester to minor ester at the end of enrichment, that is [A]t / [B]_e. Let us set it at 50/1 that is 98% pure. Now the log term becomes simple number and we can solve for time of hydrolysis required to reach that enrichment in units of the half time for the hydrolysis of the major isomer.

In the case of a 90 to 10 molar composition where the half time for the major A components hydrolysis was 300 minutes, the time for enrichment to 98% purity would be just under 54 minutes.

We can intuitively see that only a small amount of the desired A component would be destroyed in that time, so it would be a useful separation.

The degree of enrichment/purification that is needed before the main component of the mixture will purify itself further, (during crystallization for example), is estimated based on experience with similar compounds; then the time for the competitive hydrolysis can be calculated as a fraction of the hydrolysis half time of the major component. After this period, acid-base extraction is applied to separate the mixture of esters and acids and the enriched material recovered. The acid fraction of course will be enriched in the minor, more easily hydrolyzed material.

The same thing is stated in more qualitative terms in the statement that among aliphatic carboxylic acids those of primary structure are esterified readily with an alcohol and a mineral acid catalyst whereas those in which the carboxyl is joined to a quartenary carbon react sluggishly, probably because the alkyl groups dominate so much of the space in the neighbourhood of the carboxyl group that they block the formation of a protonated intermediate. The alkyl groups combine to break up the required solvation shell around the charged activation intermediate and raise its free energy slowing the reaction.

This is even more striking among benzoic acid or heterocyclic acid systems when there are two ortho substitutuents. Meyer in 1894 investigated the response of aromatic acids to attempted esterification under the conditions of refluxing 3-5 hours a solution of aromatic acids in methanol containing 3% hydrogen chloride or by saturating a methanol solution with HCl in the cold and allowing the solution to stand overnight. Doubly ortho substituted materials yielded little or no ester. Even a single ortho substitutent exerted a significant blocking effect compared to benzoic acid. The carboxyl of salicyclic acid which has an ortho hydroxyl must be performed five times as long to give methyl salicylate in reasonable quantity. This difference in reaction rates may be able to be increased further by using a larger alcohol in the esterification. This would add an unfavorable equilibrium to the already slow forward reaction rate. Fieser & Fieser’s Organic Chemistry Third Edition. pg.671-673.

Although one might think that such thinking is only applicable to simple aromatic substitution problems, when structures become even more complex carboxyl and amines can be hindered by even quite remote parts of a structure in terms of intervening bond distance and steric hindrance can come into play.

Of course, the best means for differentiating between esters is an enzyme. We are familiar with using enzymes to hydrolyze one enantiomer of a pair of mirror image compounds in a racemate. It follows from this, however, that enzymes should be able to even more easily distinguish esters of distinctly different structures. In the past this was not a productive path because the most likely result would have been that both esters were not substrates for the enzyme. Today many more esterases are available and it is quite likely that an appropriate one can be found to selectively hydrolyze one ester structure in the presence of another. Of course if the compound that hydrolyzes is chiral, the enzyme may only hydrolyze one of the chiral pair. Besides a separation one would achieve a resolution in the same pot!