FIELD OF THE INVENTION
The present invention relates to processes especially suitable for the large scale production of tetrapyrrolic compounds, such as meso-formyl porphyrins, meso-acrylate porphyrins, purpurins and benzochlorins. In particular, tin ethyl etiopurpurin (SnET2), sometimes called rostaporfin, and the intermediates necessary for its production without chromatography are disclosed. In addition, much of the chemistry disclosed is applicable to the large scale manufacturing of benzochlorins. Purpurins, benzochlorins and several of the intermediates in the synthesis may be useful as photosensitizers in photodynamic therapy, or as porphyrin building blocks in the synthesis of other porphyrinic materials.
BACKGROUND OF THE INVENTION
Photodynamic therapy is a procedure that uses photoactive (light-activated) drugs to target and destroy diseased cells. Photoactive drugs transform light energy into chemical energy in a manner similar to the action of chlorophyll in green plants. The photoactive drugs are inactive until irradiated by light of a specific wavelength, thereby enabling physicians to target specific groups of cells and control the timing and selectivity of treatment. The result of this process is that diseased or unwanted cells are destroyed with less damage to surrounding normal tissues. For a more detailed description of photodynamic therapy, see U.S. Pat. Nos. 5,225,433, 5,198,460, 5,171,749, 4,649,151, 5,399,583, 5,459,159, and 5,489,590, the disclosures of which are incorporated herein by reference.
A large number of naturally occurring and synthetic dyes are currently being evaluated as potential photoselective compounds in the field of photodynamic therapy. Perhaps the most widely studied class of photoselective dyes in this field are the tetrapyrrolic macrocyclic compounds generally called tetrapyrroles, some of which are shown below.
In particular, and relevant to this invention, are the chlorin ring systems called purpurins and benzochlorins. Purpurins are a class of chlorin in which an annelated five membered cyclopentenyl ring is directly attached to the reduced pyrrole ring. A notable example of a metallo-purpurin that is showing great promise in the field of photodynamic therapy is tin dichloride ethyl etiopurpurin I (7) (currently prepared by Scheme 1). An older method for synthesis of (7) was outlined in U.S. Pat. No. 5,051,415, the disclosure of which is incorporated herein by reference.
Benzochlorins on the other hand have an annelated benzene ring directly attached to the reduced pyrrole ring. A notable example of a benzochlorin is octaethylbenzochlorin (13) (prepared by Scheme 2) which serves as a starting chlorin for many promising photosensitizers (see U.S. Pat. Nos. 5,552,134; 5,438,051; 5,250,668; 5,109,129; 4,988,808; 5,514,669; 6,008,211; 5,856,515; 5,744,598; 5,512,559; and 5,424,305, the disclosures of which are incorporated herein by reference). To date, very inefficient routes to the synthesis of purpurins and benzochlorin ring systems have been reported and there exists no reported satisfactory method of manufacturing these materials on a large scale.
As a result, a method that enables the synthesis of compounds having these two ring systems, the purpurins and the benzochlorins and their intermediates, on a large scale is of immense value. The present invention provides processes for the large scale preparation of meso-formyl porphyrins, meso-acrylate porphyrins, metal free meso-acrylate porphyrins, metallo-porphyrins, purpurins, metallated purpurins and benzochlorin compounds, the purification steps of which are achieved simply by fractional crystallizations.
It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present invention as claimed.
DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
The present invention relates to processes for synthesizing meso-formyl porphyrins, β-formyl porphyrins, metallo meso-acrylate porphyrins, metal free meso-acrylate porphyrins, purpurins, metallo-purpurins, metallo-porphyrins and benzochlorins and isolating these compounds simply by fractional crystallization techniques that are amenable to large scale syntheses of such compounds. Processes for the isolation of such compounds will be set forth in the following detailed descriptions of each of the steps of schemes 1 and 2. These processes require no chromatographic separations or additional chemical reactions and are therefore suitable for use on a large scale. The present invention is also particularly relevant to the formylation of metallo-tetrapyrrolic molecules, the reaction of formyl tetrapyrrolic compounds with Wittig reagents, the demetallation of metallo-tetrapyrrolic compounds, the cyclization of meso-acrylate tetrapyrroles to purpurin compounds, the reduction of meso-acrylate tetrapyrrolic compounds, the cyclization of meso-(3-hydroxypropenyl) tetrapyrrolic compounds to benzochlorins and the tin metallation of tetrapyrrolic molecules, all on large scale.
As used herein, the term “tetrapyrrole” or “tetrapyrrolic compound” is intended to encompass a large number of compounds with at least three joined pyrrolic rings having widely differing functionality as described in the literature (for example, see “Porphyrins and Metalloporphyrins” Ed. K. Smith, Elsevier, 1975, N.Y.; “The Porphyrins”, Ed. D. Dolphin, Vol. I-V, Academic Press, 1978; and “The Porphyrin Handbook”, Ed. K. Kadish, K. M. Smith, R. Guilard, Academic Press, 1999). These compounds contain various and ranging substituents on the β-pyrrole positions or meso-positions of the tetrapyrrolic rings, either symmetrically or asymmetrically substituted on the tetrapyrrolic macrocycle. Simple tetrapyrrolic ring systems include porphyrins, chlorins, iso-bacteriochlorins and bacteriochlorins. Additionally, molecules resembling porphyrins such as corroles, porphodimethenes, phthalocyanines, naphthalocyanines, azoporphyrins, phlorins, texaphyrins, porphyrin “isomers” (such as porphycenes, porphacyanine, homoporphyrins, corrphycenes, vinylogous corroles, vinylogous porphyrins, sapphyrins, pyriporphyrins, smaragdyrins, isosmaragdyrins, ozaphyrins, pentaphyrins, heteropentaphyrins, orangarins, dehydropentaphyrins, rubyrins, bronzaphyrins, octaphyrins, and the like ) have been developed with a wide range of functionality both at the peripheral positions and at the internal heterocyclic “core” of these molecules. All of these compounds are considered to be within the scope of the term “tetrapyrrole” or “tetrapyrrolic compound” as used herein.
In many of these macrocycles the inner heteroatoms have been replaced by O, S, Se, Te forming new macrocycles with interesting properties. Many of these materials are capable of coordinating metals and will undoubtedly find commercial uses in the fields of medicine and industry and thus are applicable to the inventions set forth in the specification, particularly with regard to formylation, Wittig reactions, demetallation and metallation with tin. Accordingly, there will be a need for highly pure material, especially in pharmaceuticals, which can be made on a large scale.
Examples of the various substituents that can be present on the β-pyrrole or meso-positions of the tetrapyrrolic compounds of the invention include functional groups having a molecular weight less than about 100,000 daltons and can be a biologically active group or organic. Examples are, but are not limited to: (1) hydrogen; (2) halogen, such as fluoro, chloro, iodo and bromo (3) lower alkyl, such as methyl, ethyl, n-propyl, butyl, hexyl, heptyl, octyl, isopropyl, t-butyl, n-pentyl and the like groups; (4) lower alkoxy, such as methoxy, ethoxy, isopropoxy, n-butoxy, t-pentoxy and the like; (5) hydroxy; (6) carboxylic acid or acid salts, such as —CH2COOH, —CH2COONa,—CH2CH2COOH, —CH2CH2COONa, —CH2CH2CH(Br)COOH, —CH2CH2CH(CH3)COOH, —CH2CH(Br)COOH, —CH2CH(CH3)COOH, —CH(Cl)CH2CH(CH3)COOH, —CH2CH2C(CH3)2COOH, —CH2CH2C(CH3)2COOK, —CH2CH2CH2CH2COOH, C(CH3)2COOH, CH(Cl)2COOH and the like; (7) carboxylic acid esters, such as —CH2CH2COOCH3, —CH2CH2COOCH2CH3,—CH2CH(CH3)COOCH2CH3, —CH2CH2CH2COOCH2CH2CH3, —CH2CH(CH3)2COOCH2CH3, —CH2CH2COOCH2CH2OH, —CH2CH2COOCH2CH2 N(CH3)2 and the like; (8) sulfonic acid or acid salts, for example, group I and group II salts, ammonium salts, and organic cation salts such as alkyl and quaternary ammonium salts; (9) sulfonylamides such as —SO2NH(alkyl), —SO2N(alkyl)2, —SO2NH(alkyl-OH), —SO2N(alkyl-OH)2, —SO2NH(alkyl)-N(alkyl)2, —SO2N(alkyl-N(alkyl)2)2, SO2NH(alkyl)-N(alkyl)3 +Z−) and the like, wherein Z31 is a counterion), —SO2NHCH2CO2H, substituted and unsubstituted benzene sulfonamides and sulfonylamides of aminoacids and the like; (10) sulfonic acid esters, such as SO3(alkyl), SO3(alkyl-OH), SO3(alkyl-N(alkyl)2), SO3(alkyl-N(alkyl)3 +Z31 ) and the like, wherein Z 31 is a counterion), SO3CH2CO2H, and the like (11) amino, such as unsubstituted or substituted primary amino, methylamino, ethylamino, n-propylamino, isopropylamino, butylamino, sec-butylamino, dimethylamino, trimethylamino, diethylamino, triethylamino, di-n-propylamino, methylethylamino, dimethyl-sec-butylamino, 2-aminoethoxy, ethylenediamino, cyclohexylamino, benzylamino, phenylethylamino, anilino, N-methylanilino, N,N-dimethylanilino, N-methyl-N-ethylanilino, 3,5-dibromo-4-anilino, p-toluidino, diphenylamino, 4,4′-dinitrodiphenylamino and the like; (12) cyano; (13) nitro; or (14) a biologically active group; (15) amides, such as —CH2CH2CONHCH3,—CH2CH2CONHCH2CH3, —CH2CH2CON(CH3)2, —CH2CH2CON(CH2CH3) 2, —CH2CONHCH3, —CH2CONHCH2CH3, —CH2CON(CH3)2, —CH2CON(CH2CH3)2 and amides of amino acids and the like; or (16) iminium salts for example CH═N(CH3)2 +Z31; and the like, wherein Z31 is a counterion), (17) Boron containing complexes, (18) carbon cage complexes (e.g., C60 and the like), (19) metal cluster complexes, for example derivatives of EDTA, crown ethers, cyclams, cyclens, (20) other porphyrin, chlorin, bacteriochlorin, isobacteriochlorin, azoporphyrin, tetraazoporphyrin, phthalocyanine, naphthalocyanine, texaphyrins, tetrapyrrolic macrocycles or dye class and the like (21) alkynyl, including alkyl, aryl, acid and heteroatom substituted alkynes, and (22) haloalkyl where one or more halogens are substituted onto the alkyl carbon chain, the length of the carbon chain being from C1 to C20; and (23) any other substituent that increases the hydrophilic, amphiphilic or lipophilic nature or stability of the compounds.
The term “biologically active group” as used herein can be any group that selectively promotes the accumulation, elimination, binding rate, or tightness of binding in a particular biological environment. For example, one category of biologically active groups is the substituents derived from sugars, specifically, (1) aldoses such as glyceraldehyde, erythrose, threose, ribose, arabinose, xylose, lyxose, allose, altrose, glucose, mannose, gulose, idose, galactose, and talose; (2) ketoses such as hydroxyacetone, erythrulose, rebulose, xylulose, psicose, fructose, sorbose, and tagatose; (3) pyranoses such as glucopyranose; (4) furanoses such as fructo-furanose; (5) O-acyl derivatives such as penta-O-acetyl-α-glucose; (6) O-methyl derivatives such as methyl α-glucoside, methyl β-glucoside, methyl α-glucopyranoside, and methyl-2,3,4,6-tetra-O-methyl-glucopyranoside; (7) phenylosazones such as glucose phenylosazone; (8) sugar alcohols such as sorbitol, mannitol, glycerol, and myo-inositol; (9) sugar acids such as gluconic acid, glucaric acid and glucuronic acid, δ-gluconolactone, δ-glucuronolactone, ascorbic acid, and dehydroascorbic acid; (10) phosphoric acid esters such as α-glucose 1-phosphoric acid, α-glucose 6-phosphoric acid, α-fructose 1,6-diphosphoric acid, and α-fructose 6-phosphoric acid; (11) deoxy sugars such as 2-deoxy-ribose, rhammose (deoxy-mannose), and fructose (6-deoxy-galactose); (12) amino sugars such as glucosamine and galactosamine; muramic acid and neurarninic acid; (13) disaccharides such as maltose, sucrose and trehalose; (14) trisaccharides such as raffinose (fructose, glucose, galactose) and melezitose (glucose, fructose, glucose); (15) polysaccharides (glycans) such as glucans and mannans; and (16) storage polysaccharides such as α-amylose, amylopectin, dextrins, and dextrans.
Amino acid derivatives are also useful biologically active substituents, such as those derived from valine, leucine, isoleucine, threonine, methionine, phenylalanine, tryptophan, alanine, arginine, aspartic acid, cystine, cysteine, glutamic acid, glycine, histidine, proline, serine, tyrosine, asparagine and glutamine. Also useful are peptides, particularly those known to have affinity for specific receptors, for example, oxytocin, vasopressin, bradykinin, LHRH, thrombin and the like.
Another useful group of biologically active substituents are those derived from nucleosides, for example, ribonucleosides such as adenosine, guanosine, cytidine, and uridine; and 2′-deoxyribonucleosides, such as 2′-deoxyadenosine, 2′-deoxyguanosine, 2′-deoxycytidine, and 2′-deoxythymidine.
Another category of biologically active groups that is particularly useful is any ligand that is specific for a particular biological receptor. The term “ligand specific for a receptor” refers to a moiety that binds a receptor at cell surfaces, and thus contains contours and charge patterns that are complementary to those of the biological receptor. The ligand is not the receptor itself, but a substance complementary to it. It is well understood that a wide variety of cell types have specific receptors designed to bind hormones, growth factors, or neurotransmitters. However, while these embodiments of ligands specific for receptors are known and understood, the phrase “ligand specific for a receptor”, as used herein, refers to any substance, natural or synthetic, that binds specifically to a receptor.
Examples of such ligands include: (1) the steroid hormones, such as progesterone, estrogens, androgens, and the adrenal cortical hormones; (2) growth factors, such as epidermal growth factor, nerve growth factor, fibroblast growth factor, and the like; (3) other protein hormones, such as human growth hormone, parathyroid hormone, and the like; (4) neurotransmitters, such as acetylcholine, serotonin, dopamine, and the like; and (5) antibodies. Any analog of these substances that also succeeds in binding to a biological receptor is also included.
Particularly useful examples of substituents tending to increase the amphiphilic nature of the compounds include: (1) short or long chain alcohols, for example,—C12H24—OH where —C12H24 is hydrophobic; (2) fatty acids and their salts, such as the sodium salt of the long-chain fatty acid oleic acid; (3) phosphoglycerides, such as phosphatidic acid, phosphatidyl ethanolamine, phosphatidyl choline, phosphatidyl serine, phosphatidyl inositol, phosphatidyl glycerol, phosphatidyl 3′-O-alanyl glycerol, cardiolipin, or phosphatidal choline; (4) sphingolipids, such as sphingomyelin; and (5) glycolipids, such as glycosyldiacylglycerols, cerebrosides, sulfate esters of cerebrosides or gangliosides. It would be apparent to those skilled in the art what other groups or combinations of the groups described would be suitable to the invention.
Meso-Formylation of Metallo-Tetrapyrroles
It is well established in the literature that many metallated tetrapyrrolic compounds such as Nickel octaethylporphyrin (OEP) and Nickel etioporphyrin I (NiEtio I) undergo electrophilic substitution at the meso position with a Vilsmeier reagent according to schemes 1 and 2 to give compounds (2) and (9). In compounds where all the meso-positions are substituted and free β-pyrrolic positions exist, for example Nickel or Copper tetraphenyl porphyrin derivatives, Vilsmeier formylations occur at the β-pyrrolic positions. Generally on the laboratory scale (grams), the reactions are carried out in hot 1,2-dichloroethane (55-60° C.) and the resulting iminium salt intermediate is hydrolyzed with saturated sodium acetate solution at this elevated temperature. The solution is cooled to room temperature and the organic layer is separated from the aqueous layer. The organic layer is then dried, filtered and evaporated to dryness. The crude residue is best purified by chromatography on silica.
In addition to the mono-meso-formylated products, significant amounts of di-formylated products (Scheme 3) are produced in the reaction in yields between 9-20%. Longer reaction times produce more di-formylated products.
Current literature (K. M. Smith et al., J. Chem. Soc Perk I, p.581, 1982; K. M. Smith et al, Teft. Lett., vol.21, p.3747,1980; G. V. Ponomarev et al, Kimiya Geterotsiklicheskikh Soedinenii, vol.6, p.776, 1979; G. V. Ponomarev et al, Kimiya Geterotsiklicheskikh Soedinenii, vol.6, p.767, 1979; E. Watanabe et al, Tetrahedron, vol.31, p.1385, 1975; K. M. Smith et al, J. Chem. Soc. Perk I, p.439, 1983; G. V. Ponomarev et al, Kimiya Geterotsiklicheskikh Soedinenii, vol.9, p.1215, 1970; A. W. Nichol, J. Chem. Soc. (C) p.903, 1970; G. V. Ponomarev et al, Kimiya Geterotsiklicheskikh Soedinenii, vol.4, p.479, 1984; G. V. Ponomarev et al, Kimiya Geterotsiklicheskikh Soedinenii, vol.11, p.1507, 1982; J. W. Buchler, Leibigs Annalen Chemie, p.43, 1988; A. W. Johnson, J. Chem. Soc., p.4, 1966; H. Brockmann, Leibigs Annalen Chem, vol.148, p.718,1968; R. Grigg, J. Chem. Soc., p.1789,1972;; R. Grigg, J. Chem. Soc. CC1, p.557, 1979) and patent disclosures (U.S. Pat. Nos. 5,051,415; 5,216,012; 4,877,872; and 5,534,506, the disclosures of which are incorporated herein by reference) in general use chromatography or precipitation from methanol to isolate compounds (2) and (9) in particular from their di-formylated counterparts.
As chromatography on a large scale is an expensive alternative for purification of tetrapyrrolic molecules, the present inventors studied the precipitation of the crude formylated tetrapyrrolic compounds from methanol, ethanol and acetic acid. In this instance, the inventors chose to study compounds (2) and (9). Precipitation of the crude formylated reaction mixture of (2) from dichloromethane/methanol or dichloromethane/ethanol via distillation of the lower boiling solvent was found to be inefficient at removing the di-formylated impurity products from the precipitated product. Approximately the same ratio of diformylated products to mono-formylated products (Table 1 for etioporphyrin series) remained using this procedure.
However, the inventors have found that when alkylcarboxylic acids, such as acetic acid, are used as a precipitating solvent, and the dichloromethane distilled as before, the di-formyl by-products were successfully reduced by approximately 10% per precipitation. This is due to the fact that the di-formylated products possessed greater solubility in the alkylcarboxylic acids than in methanol or ethanol.
The use of alkylcarboxylic acid solvents like acetic acid has several additional advantages. Aqueous residues, which may contain sodium acetate from the hydrolysis step, that have inadvertently found their way into the organic layer during the extraction and separation process are soluble in alkylcarboxylic acids. This is particularly useful in large scale manufacturing where separation of the organic and aqueous phases is sometimes difficult. Additionally, metal free porphyrins, for example etioporphyrin I which is observed as a minor impurity with Nickel etioporphyrin I (due to incomplete Nickel insertion), has solubility in alkylcarboxylic acids and hence may be reduced in percentage in the final isolated Nickel meso-formyl etioporphyrin I The fact that many metal-free tetrapyrrolic compounds are soluble in alkylcarboxylic acids enables an efficient means of reducing this impurity in metallated porphyrin derivatives, which in general are less soluble in alkylcarboxylic acids. Additionally, polymeric material produced in the reaction is also soluble in alkylcarboxylic acids.
|TABLE 1 |
|HPLC of the crude reaction mixture and precipitations (single) |
| ||Crude ||Methanol ||Ethanol ||Acetic acid |
| ||Mixture ||pptn ||pptn ||pptn |
| || |
|Ni meso-formyl ||79.6% || 80% || 76% ||90.1% |
|Ni meso-diformyl || 16% || 15% || 16% ||6.8% |
|Etioporphyrin ||2.4% ||3.0% ||2.6% ||1.6% |
|Ni etioporphyrin ||0.9% ||1.3% ||1.2% ||1.2% |
Yet another major advantage lies with the boiling point of alkylcarboxylic acids. Once the 1,2-dichloromethane has been separated from the aqueous phase, it may be effectively distilled from a mixture of alkylcarboxylic acid/dichloromethane, affecting the precipitation of the desired meso-formyl porphyrin.
Similar observations were made for the formylation of Nickel octaethylporphyrin. The di-formylated Nickel octaethylporphyrin impurities were reduced by levels of 10-15% by a single precipitation from dichloromethane/acetic acid.
Similar results have been obtained for a large number of meso-formylporphyrins, including Nickel meso-formyl etioporphyrin I and II Copper meso-formyl etioporphyrin I and II Nickel meso-formyl coproporphyrin I and II tetra-alkyl esters, Copper meso-formyl coproporphyrin I and II tetraalkyl esters, Nickel β-formyl tetraphenyl porphyrin, Nickel β-formyl tetrakis((4′-methyl)phenyl))porphyrin, Nickel β-formyl tetrakis((4′-carbomethoxy)phenyl) porphyrin, Copper (II) octaethylporphyrin, Nickel (II) octaethylporphyrin, Copper mesoporphyrin dialkyl ester, and Nickel mesoporphyrin dialkyl ester. It is envisaged that this procedure is generally applicable to the purification of any formylated metallo-tetrapyrrolic compound on a large scale that is not soluble, or has limited solubility in alkylcarboxylic acids. The scope of the invention is not limited to the examples provided herein.
In accordance with another embodiment of the present invention, as embodied and broadly described herein, we have found that the formylation of metallo tetrapyrrolic compounds may be undertaken in a solvent other than 1,2-dichloroethane. As 1,2-dichloroethane is listed as a class 1 solvent by the International Conference on Harmonization (ICH), it would be advantageous on a large-scale to use a less toxic solvent for the formylation of metallo-tetrapyrroles. Toward this goal we used dichloromethane.
Formylation of Nickel etioporphyrin or Nickel octaethylporphyrin with Vilsmeier reagent occurs only slowly at room temperature. In refluxing dichloromethane at atmospheric pressure, the formation of the desired intermediate iminium salt progresses slowly, being complete after approximately 13-24 hours. If the same reaction is undertaken in dichloromethane in a glass lined metal reactor at an elevated temperature such as 35-60° C., preferably 50-60° C., under pressurized conditions, the reaction proceeds smoothly in 3-6 hours. Hence it is possible to replace 1,2-dichloroethane with the less toxic dichloromethane in standard large-scale reactor equipment that is capable of sustaining pressure. Metal glass lined reactor vessels are well suited and designed for such reactions. This procedure is generally applicable to the synthesis of any formylated metallo-tetrapyrrolic compound on a large scale, including but not limited to porphyrins, azoporphyrins, chlorins, iso-bacteriochlorins and bacteriochlorins. The applicability of this process to other formylated metallo-tetrapyrrolic compounds would be within the knowledge of those skilled in the art.
Wittig Reactions on Meso-Formyl Tetrapyrroles
It is well established in the literature that the formyl group on metallated meso-formyl porphyrins such as meso-formyl Nickel etioporphyrin I (2) and meso-formyl Nickel octaethylporphyrin (9) or on β-formylated porphyrins undergoes Wittig reactions with a large number of Wittig reagents such as, for example, (ethoxycarbonylmethylene)triphenylphosphorane or (methoxycarbonylmethylene)-triphenylphosphorane or the like, to produce the corresponding Wittig addition product porphyrins like meso-acrylate porphyrins (Scheme 4). It should be noted that the term meso-acrylate as used herein is broadly defined as including the following groups: —CH═CHCO2Et; —CH═CHCO2Me; —CH═CH(ester); —CH═CH(amide); —CH═CHCHO; —CH═CHCH(Oalkyl)2; —CH═CHCH(Ocyclicalkyl)2.; —(CH═CH)n(ester) where n =2, 3; —(CH═CH)n(CHO) where n=2, 3;—=CHCN; and —═CHCO2H. There are a number of “stabilized” Wittig reagents that can be isolated as powders with defined melting points. Such Wittig reagents are useful in the present invention. Also useful are stabilized Wittig reagents that are liquids.
In literature preparations, the Wittig reaction of (ethoxycarbonylmethylene)-triphenylphosphorane or (methoxycarbonylmethylene)triphenyl phosphorane or the like on meso-formyltetrapyrroles is carried out in refluxing xylenes (bp 138-145° C.) under atmospheric conditions (generally overnight) where upon completion of the reaction, the xylene is removed by rotary evaporation and the viscous gummy residue dissolved in dichloromethane. The solution is then chromatographed on silica to remove the desired meso-acrylate porphyrin from the tars and triphenylphosphine oxide produced in the reaction as outlined in the literature (D. P. Arnold, J. Chem. Soc. Perk. I, p.1660, 1978; H. Callott, Bull. Soc. Chim. France. p.3413, 1973; A. R. Morgan et al, J. Org. Chem., vol.51, p.1347, 1986; A. R. Morgan et al, J. Med. Chem., vol.34, p.2126, 1991) and in several patents (U.S. Pat. Nos. 5,051,415 and 5,534,506).
The present inventors have discovered that the reported procedures suffer from a number of disadvantages when going from bench scale (1-5 g) to larger scale. First, it was found that when the reaction was performed in xylene, a large excess of Wittig reagent (typically more than 5 equivalents) was always required to complete the reaction. Even at small scales the reaction was generally slow, requiring at least 15-18 hours of reflux. It was also noted that amounts as high as 10% of Nickel etioporphyrin (or Nickel octaethylporphyrin) were produced during the reaction via deformylation of the starting material. Chromatography of the reaction often yielded the metallated meso-acrylate porphyrin contaminated with triphenyl phosphine oxide due to tailing of the latter through the column. It should be noted that metal-free formyl tetrapyrrolic compounds can also be reacted with Wittig reagents under the same conditions, thereby producing the metal free meso-acrylate compounds directly. These compounds also suffer the disadvantages on purification described above.
As chromatography on a large scale is an expensive alternative for purification of such molecules and because of the problems described, the reaction conditions were modified to best optimize for time, purity, amount of Wittig reagent used, and ease of isolation of the final product. We investigated the reaction in a different solvent, dimethylformamide (DMF), under argon or another inert atmosphere. Under similar conditions as those described by the prior art, it was found that no appreciable amounts of Nickel etioporphyrin I were produced in the reaction. It was also discovered that reducing the volume of DMF in the reaction dramatically decreased the time taken to complete the reaction. In fact, it was found that the reaction could be performed as a melt in which a very small volume of the solvent was present and the temperature was slowly increased to 135° C. over an hour or so with adequate stirring. In this case the reaction was complete in 3-6 hours. Under these conditions, a 2 molar equivalent of phosphorane was used, dramatically reducing the amount of Wittig reagent required from the reported procedure. It was found optimal, however, to have a small amount of solvent (DMF) in the reaction, specifically within about 10% of the weight of the starting meso-formyl tetrapyrrole. The solvent aids in the initial dissolution of the melt and assists in bringing powdered starting material on the sides of the reactor flask into the reaction melt. It is envisaged that this procedure is generally applicable to the Wittig reaction of any formylated tetrapyrrolic compound on a large scale. It is further envisaged that this procedure is particularly applicable to any Wittig reaction where the Wittig reagent is a stable solid or liquid at room temperature. The scope of the invention is not limited to the examples provided herein, but is realized to be generally applicable to metallo or metal free formyl tetrapyrrolic molecules possessing at the β-pyrrole positions, or meso- positions, the functionality described at pages 8-12 herein.
The next challenge was to purify the Nickel meso-acrylate porphyrin from the reaction melt. This was achieved very successfully simply by optionally removing as much of the DMF as possible, allowing the solution to cool, dissolving the crude residue in a solvent and adding a precipitating solvent. Distillation of the solvent with stirring resulted in the precipitation of the desired Nickel meso-acrylate etioporphyrin. The porphyrin can then be filtered directly and optionally washed with the precipitating solvent. The yields from such a procedure are generally about 90% and the purity greater than 98%. Comparable results are obtained with the octaethylporphyrin series and the following meso-formyl tetrapyrrolic compounds: meso-formyl etioporphyrin I; Nickel meso-formyl etioporphyrin I and II Copper meso-formyl etioporphyrin I and II Nickel meso-formyl coproporphyrin I and II; Copper meso-formyl coproporphyrin I and II tetramethyl esters; Nickel β-formyl tetraphenyl porphyrin; Nickel β-formyl tetrakis((4′-methyl)phenyl))porphyrin; Nickel β-formyl tetrakis((4′-carbomethoxy)phenyl) porphyrin; and the purification of their acrylate analogues. The precipitation technique described is believed to be applicable to any meso-acrylate (or similar) tetrapyrrolic compound on a large scale that is not soluble, or has limited solubility in alcohols with the functionality described at pages 7-12.
In accordance with the invention, the solvent in the above described precipitation procedure can be halogenated or non-halogenated and is preferably selected from dichloroethane, dichloromethane, ethyl acetate, tetrahydrofuran, acetonitrile, acetone, benzene, toluene, and ethers. The precipitating solvent can also be halogenated or non-halogenated and is preferably selected from acetic acid, propionic acid, methanol, ethanol, 1-propanol, 2-propanol, 1-butanol, hexanes, acetonitrile, ethyl acetate, and iso-octane.
Of particular interest is the composition of the mother liquors of the Wittig reaction on Nickel meso-formyl etioporphyrin I They are largely enriched in Nickel etioporphyrin, etioporphyrin, and the Nickel meso-diacrylate etioporphyrins (produced in the reaction from the Nickel di-formyl etioporphyrins). Thus, the alcoholic mother liquors, rich in triphenylphosphine oxide, effectively remove up to 20% of di-formyl etioporphyrin impurities (as their di-acrylate derivatives) carried over in the meso-formyl porphyrin solid of the formylation step, as well as other impurities.
It should be noted that pure di-formyl porphyrins undergo conversion to the di-acrylate porphyrins using the melt conditions described. Additionally, they may be precipitated and purified according to the precipitation technique outlined above. It should be noted that this precipitation technique works especially well for 5, 15 di-acrylate porphyrins as they are generally less soluble in organic solvents. The corresponding 5, 10- diacrylate porphyrins are somewhat more soluble and suffer larger losses to the precipitation technique.
Demetallation of Tetrapyrrolic Compounds on a Large Scale
It is well known in the literature that the centrally coordinated metal of many metallated porphyrins can be removed by treating the complex with strong acids. Nickel, copper or cobalt porphyrins generally require strong acids, such as sulfuric acid to liberate the metal from the complex. Organic acids, like trifluoroacetic acid, are seldom strong enough to demetallate these complexes, or are very slow at demetallating the complex. Other metals such Zn, In, Ga, Ge, and TI, for example, are rapidly removed from the tetrapyrrolic macrocycles using sulfuric acid, hydrochloric acid, methane sulfonic acid, trifluoroacetic acid, and the like.
In literature preparations, demetallation reactions of nickel or copper porphyrins or other tetrapyrrolic complexes are usually carried out using a large excess of neat sulfuric acid. The acid is either added to the porphyrin or the porphyrin added to the acid. Table 2 gives some literature examples. The solution is then generally added to a neutralizing solution (for example a NaHCO3
solution) and the porphyrin extracted with a solvent and purified. Such procedures are routinely used to demetallate all tetrapyrrolic classes, including porphyrins, chlorins, isobacteriochlorins and bacteriochlorins.
|TABLE 2 |
|Literature examples of molar ratios of acid/porphyrin |
| ||Acid amount || |
|Compound/amount ||(molar equivalents) ||Reference |
|Meso-acrylate NiOEP ||10 mL H2SO4 ( ˜92 eq) ||U.S. Pat. No. |
|(Et ester) (621 mg) || ||4,877,872 |
|Meso-acrylate NiOEP ||5 mL H2SO4 ( ˜306 eq) ||D. P. Arnold et al, |
|(Me ester) (100 mg) || ||J. Chem. Soc, 1660, |
| || ||1978 |
|Meso-acrylate NiEtio ||3 mL H2SO4 ( ˜334 eq) ||D. P. Arnold et al, |
|(Me ester) (50 mg) || ||J. Chem. Soc, 1660, |
| || ||1978 |
|5,10-Diacrylate ||2 mL H2SO4 ( ˜579 eq) ||A. R. Morgan et al, |
|NiOEP(Et ester) (25 || ||J. Med. Chem., 34 (7), |
|mg) || ||2126, 1991 |
|5,15-Diacrylate ||2 mL H2SO4 ( ˜579 eq) ||A. R. Morgan et al, |
|NiOEP(Et ester) (25 || ||J. Med. Chem., 34 (7), |
|mg) || ||2126, 1991 |
We have discovered that the reported procedures suffer from a number of disadvantages when going from bench scale (1 g) to larger scale. In the literature procedures, even on small scale, a large excess of acid is generally required to demetallate the porphyrinic compounds (Table 2). Besides the expense related to using large volumes of acid on large scale, the hazards of handling and neutralizing acid wastes becomes a crucial issue. It would be advantageous to use the minimum amount of acid to effect demetallation so that safety and disposal become more manageable.
A theoretically possible solution to this problem is to “increase the loading” or decrease the equivalents of acid to the starting metal complex. Unfortunately, this approach does not work well with the demetallation of tetrapyrrolic compounds. Regardless of whether the tetrapyrrolic powder is added to the acid or vice versa, severe clumping of the powder occurs. In fact, the powder forms solid clumps that are difficult to disperse under rapid stirring and often stick to the sides of the reactor vessel. This dramatically impacts the amount of demetallation that takes place in the reaction and isolated “demetallated” tetrapyrrolic compound product is invariably contaminated with large amounts of the metallated starting material. The clumping problem makes it virtually impossible to predict when the demetallation reaction is complete.
Another serious limitation in the reported processes involves the neutralization of the acidic solution. The addition of sodium bicarbonate to a highly acidic solution on large scale would be a hazardous undertaking, as a large amount of carbon dioxide is released in the neutralization process (which must be processed accordingly). In addition, extensive frothing and foaming of the solution occurs, which even on a small scale is difficult to control. Total or over neutralization of the solution with NaOH, for example, may cause ester cleavage of tetrapyrroles with ester groups.
We have been able to successfully overcome these problems by utilizing a simple process involving the pre-dissolution of the metallo-tetrapyrrolic compound in a non-water-soluble solvent. Examples of non-water-soluble solvents include 1,2-dichloroethane, 1,1- dichloroethane, dichloromethane, chloroform, benzene, toluene, ether, hexane, xylene, and the like. Initial dissolution of Nickel meso-acrylate porphyrin, for example, in a halogenated solvent like dichloromethane occurs readily. The temperature of the reactor is lowered to approximately 0° C., and the slow addition of an acid with vigorous stirring or agitation results in the acid being dispersed onto the sides of the reaction vessel (as it is not soluble in dichloromethane). The lowering of the temperature avoids any exotherm due to the demetallation reaction. The metallated porphyrin dissolved in the dichloromethane passes over the acid layer and demetallates. It is immediately drawn into the acid layer as its tetraprotonated species. Over a matter of 0.5-1 hour, the metallated porphyrin is continually drawn out of the dichloromethane layer and demetallated.
The completion of the demetallation reaction is easily visualized when the dichloromethane layer is essentially colorless. At this point, water is added to the solution which enables the protonated porphyrin to enter the organic layer (probably as its diprotonated species) and the solution is at least partially neutralized with sodium hydroxide. The organic layer is separated from the at least partially neutralized aqueous layer and reduced in volume by distillation. A precipitating solvent such as, for example, ethanol or methanol is added and the remaining dichloromethane removed by distillation. The precipitating solvent in this instance also acts as a proton sponge, efficiently deprotonating the porphyrin. The thick precipitate is collected by filtration and washed with ethanol. The metal-free meso-acrylate porphyrin isolated is greater than 99% pure. This procedure works equally well with Nickel meso-acrylate octaethylporphyrin.
Using this procedure, it is possible to demetallate a large amount of metallo-tetrapyrrole with very small amounts of acid. Over 240 grams of Nickel meso-acrylate etioporphyrin 1 (3) can be demetallated with only 250 ml of H2SO4. Here, approximately 5 equivalents of acid are required to effect demetallation in approximately 1 hour. Identical results are achieved with a large number of different porphyrins. The process is believed to be generally applicable to the demetallation of any metallo-porphyrin, metallo-chlorin, metallo-isobacteriochlorin, metallo-bacteriochlorin, or other metallo-tetrapyrrolic compound that has a co-ordinated metal able to be removed with sulfuric acid. It is also applicable to metallo-tetrapyrroles that are capable of being demetallated with hydrochloric acid or phosphoric acid (for example zinc, indium, gallium, thallium, germanium). Included among the tetrapyrrolic metal complexes suitable for demetallation via this process are those outlined by Johann Walter Buchler in “The Porphyrins”, Ed. D. Dolphin, Volume I, Chapter 10, p.389-483, Academic Press, New York, 1978. It is envisaged that such tetrapyrrolic molecules may possess on the β-pyrrolic or meso positions the functionality or combination thereof discussed earlier herein at pages 7-12.
It would be within the general skill and knowledge of those of ordinary skill in the art as to what other functional groups would be amenable to the inventive process or what modifications to the disclosed procedure could be made without departing from the scope of the invention.
Cyclization of Meso-Acrylate Tetrapyrroles to Purpurins
Historically in the literature, the cyclization reaction of meso-acrylate porphyrins to give purpurins has been performed using two different methods. Meso-acrylate octaalkylporphyrins, as shown in Scheme 6, have historically been cyclized to purpurins under acidic conditions, using acetic acid under an inert atmosphere. The cyclization process is slow, requiring typically 24 hours of reflux to attain an equilibrium where approximately 5-10% of starting material is present in the final product. The reaction is also highly sensitive to the presence of oxygen, which causes the formation of other purpurin types that are difficult to remove from the desired product. A significant amount of decomposition also occurs such that the yield of desired product in the presence of oxygen is halved.
Alternatively, 5,15-Bis aryl 10-acrylate porphyrins (shown in Scheme 7) undergo cyclization to give 5,15-Bis aryl purpurins under basic conditions (Et3
N, KSCN or NaOH). Surprisingly, they do not cyclize under acidic conditions. The cyclization reaction is not sensitive to oxygen and there does not appear to be an equilibrium established between starting material and product. Meso-acrylate octaalkylporphyrins (Scheme 6) do not cyclize using triethylamine or sodium hydroxide as bases.
Clearly, the chemistry related to the large scale manufacturing of purpurins from meso-acrylate octaalkylporphyrins needs significant improvement to make it commercially feasible. In addition, meso-acrylate octaalkylporphyrins, such as meso-acrylate etioporphyrin I (or meso-acrylate coproporphyrin I), which bear different alkyl substituents on either side of the meso-acrylate group (scheme 1 at page 4 of the specification) suffer from the formation of cyclization isomers. An example of this is shown in Scheme I with the cyclization of meso-acrylate etioporphyrin I (4). Here, cyclization occurs toward an ethyl group on a pyrrole ring to give ethyl etiopurpurin I (5) or toward a methyl group on a pyrrole ring to give methyl etiopurpurin I (6). The product obtained from the acetic acid cyclization route of meso-acrylate etioporphyrin I consists of a mixture of (4), (5) and (6) in a ratio of 3: 9: 8 and each of (4) and (6) must be separated from (5) for the production of SnET2 (7). See, U.S. Pat. No. 5,051,415. U.S. Pat. No. 5,051,415 does not indicate that (6) is produced in the synthesis, nor does it describe how (4) and (6) are removed from (5).
Over the course of the development of the cyclization process, we investigated base catalysis as an alternative to the acetic acid conditions for cyclizing mono-meso-acrylate tetrapyrrolic compounds to the corresponding purpurin. We investigated a large number of bases and solvents to find the optimal conditions necessary to generate maximal amounts of (5). Table 3 outlines the conditions used. MAE in Table 3 is meso-acrylate etioporphyrin I (4), ET2 is ethyl etiopurpurin (5), and MET2 is methyl etiopurpurin (6).
While we found many of the bases were effective or partially effective at producing purpurin formation, the most effective bases of those we tested were the non-nucleophilic bases 1,5-diazabicyclo[4.3.0]non-5-ene (DBN), 1,5-diazabicyclo[5.4.0]non-5-ene (DBU), tetramethyl guanidine, and pyrrolidine, generally in higher boiling solvents like toluene. The reactions were not sensitive to air (oxygen), and after short reflux times of 4-6 hours, levels of starting material MAE (4) present in the mixture were generally equal to or less than those seen in the long acetic acid reflux cyclization reaction. DBU appeared to give the greatest ratio of ET2 (5):MET2(6):MAE (4), about 72:21:7. In addition to 1,5-diazabicyclo[4.3.0]non-5-ene (DBN), 1,5-diazabicyclo[5.4.0]non-5-ene (DBU), tetramethyl guanidine, and pyrrolidine, the non-nucleophilic base piperidine would also be expected to give similar favorable results.
|TABLE 3 |
|Base catalysized Cyclization of meso-acrylate etioporphyrin I (MAE) |
|Base ||MAE Wt ||Solvent ||Reflux time ||MAE (%) ||ET2 (%) ||MET2 (%) |
|DBU ||100 mg ||Toluene ||4 ||hr || ˜7% ||72% ||21% |
|DBN ||100 mg ||Toluene ||4 ||hr ||<10% ||55% ||35% |
|Pyrrolidine ||100 mg ||Toluene ||4 ||hrs || <7% ||71% ||22% |
|Et3N ||100 mg ||Toluene ||24 ||hr ||100% ||NR ||NR |
|DMAP ||100 mg ||Toluene ||24 ||hrs || 94% || 4% || 2% |
|Pyridine ||100 mg ||neat ||24 ||hrs ||100% ||NR ||NR |
|4-DMAP ||100 mg ||Toluene ||24 ||hrs || 75% ||15% || 5% |
|Tetra- ||100 mg ||Toluene ||6 ||hrs || 7% ||70% ||23% |
The challenge then came to separate the two isomeric purpurins ET2 (5) and MET2 (6) without chromatography. Our investigations into the solubility of both compounds led to the discovery that MET2 (6) was much more soluble in acetonitrile and acetone, than was ET2 (5). In fact, we found that ET2 (5) has only limited solubility in hot acetonitrile. Additionally, the starting porphyrin meso-acrylate etioporphyrin I (4) was found to have solubility in acetonitrile and acetone. These discoveries enabled us to develop a simple precipitation process to effectively separate ET2 (5) from MAE (4) and MET2 (6). This process is suitable for use in both traditional acetic acid cyclization reactions as well as the base catalyzed cyclization reactions disclosed herein.
In accordance with the invention, the crude reaction mixture from the base catalyzed or acetic acid cyclization reaction can be evaporated to dryness or near dryness and the residue dissolved in a solvent or mixture of solvents such as dichloromethane. A precipitating solvent such as acetonitrile or acetone can then be introduced, and the solvents(s) removed by distillation. The precipitated product is then filtered rapidly from the warm solution. The mother liquors are rich in small amounts of MAE (4) and mostly MET2 (6). Preferably, the process is repeated until the solid ET2 (5) is sufficiently pure by TLC (0.5% ethylacetate/dichloromethane) or HPLC to proceed to the metallation step and the production of SnET2 (7). In general, three precipitations in this manner has been found to be sufficient to obtain pure product. It would be understood by those skilled in the art that this process may be carried out with a variety of solvents and precipitation solvents in the same or similar manner. Both the solvent and the precipitating solvent can be halogenated or non-halogenated. Preferably, the solvent is dichloromethane, ether, 1,2-dichloroethane, chloroform, toluene, acetone, methanol, ethanol, tetrahydrofuran, ethyl acetate, benzene, or mixtures thereof. The precipitating solvent is preferably methanol, ethanol, 1-propanol, 2-propanol, 1-butanol, t-butanol, acetone, acetonitrile, hexane, heptane, isooctane, cyclohexane, or isopropyl ether.
The isolation of MET2 (6) from the acetonitrile or acetone mother liquors above is a relatively straightforward process. The biggest challenge to overcome lies with the separation of MAE (4) from the MET2 (6). In our tin metallation reactions described earlier, we have observed that free base porphyrins can be metallated effectively at temperatures where the chlorin cannot be metallated (typically 60-80° C.) (see following tin metallation section) in solvents like 1,2-dichloroethane or acetic acid. If the mother liquors of the ET2 precipitation technique described above are evaporated and dissolved in either 1,2-dichloroethane or acetic acid containing a tin salt (preferably SnCl2 predissolved in dimethylformamide) and sodium acetate, any metal-free porphyrins in the crude mixture can be metallated at 30-80° C. with little or no metallation of the chlorin. The crude reaction mixture can then be evaporated to dryness and redissolved in a solvent suitable for chromatography of the material. This method is preferable if the metallation is undertaken in acetic acid or the like.
Alternatively, if the metallation reaction is carried out in 1,2-dichloroethane the solution may be either reduced in volume and chromatographed or directly chromatographed. As tin compounds bind particularly tightly to either silica or alumina, the crude chlorin solution from the metallation reaction can be passed over a small pad of silica or alumina. The tin porphyrins remain on the silica while the chlorin fraction may be eluted from the column under most chromatographic conditions. Such a selective metallation and purification process enables the separation of porphyrinic impurities from chlorins or bacteriochlorins in general. Alternatively, other metals that increase the polarity of the porphyrinic compound on silica or alumina may be used. The metal salts most preferable are those that are readily incorporated into porphyrins at or between 60-80° C. including Sn2+In3+, Ga3+, TI3+, etc., because these metal salts, when complexed to a tetrapyrrolic molecule, possess an axial ligand that enhances polarity on silica or alumina and enhances the ability to separate the chlorin or bacteriochlorin from the porphyrin impurity. MET2 can then be isolated by chromatography and the resulting MET2 purified by precipitation or crystallization from a solvent such as dichloromethane or a mixture of solvents, and a precipitating solvent in which MET2 is not soluble. In this instance, the solvent can be removed by slow or rotary evaporation resulting in the precipitation of MET2.
The base catalyzed cyclization of meso-acrylate tetrapyrrolic compounds to give purpurins or purpurin type compounds is believed to be applicable to any meso-acrylate (or similar) tetrapyrrolic compound on a large scale. The functionality on the periphery of the meso-acrylate tetrapyrrolic molecule, either at the meso or β-pyrrolic positions may be varied widely as outlined earlier herein at pages 7-12. It would be within the knowledge of those skilled in the art what other functional groups or modifications to this procedure could be made to utilize the invention described.
Large Scale Synthesis of Octaethylbenzochlorin.
A variety of methods have been reported for the synthesis of octaethylbenzochlorin. U.S. Pat. No. 5,552,134 and literature references (Arnold, D. P., Gaete-Holmes, R., Johnson, A. W., Smith, R. P., Williams, G. A., J. Chem. Soc. Perkin I, 1660-1670, 1978; Morgan, A. R., Skalkos, D., Maguire, G., Rampersaud, A., Garbo, G., Keck, R., Selman, S. H., Photochem. Photobiol. Vol. 55, No.1, 133-136, 1992) outline the synthesis of octaethylbenzochlorin from Nickel meso-(β-formylvinyl)-octaethylporphyrin (14) or from the metal free meso-(β-hydroxymethylvinyl)porphyrin (12) (Scheme 8).
Historically, the formation of octaethylbenzochlorin (OEBC) has been via two routes. First, the metallated meso-acrolein porphyrins (14) and (15) have been cyclized under acidic conditions to give Nickel OEBC (17) (or copper benzochlorin if the copper meso-acrolein porphyrin is used). Demetallation of (17) (or Copper OEBC) is difficult and demetallation with sulfuric acid generally produces OEBC (13) and its sulfonic acid derivative (18) after 3 hours at room temperature.
An alternative route outlined by Morgan et al (Morgan, A. R., Skalkos, D., Maguire, G., Rampersaud, A., Garbo, G., Keck, R., Selman, S. H., Photochem. Photobiol. Vol. 55, No.1, 133-136, 1992) involves the reduction of the meso-acrylate porphyrin (11) with diisobutyl aluminium hydride in tetrahydrofuran (THF) at low temperature to give the meso-(β-hydroxymethylvinyl)porphyrin (12). This compound is then treated with sulfuric acid for 5 minutes to effect the cyclization and give OEBC (13). Longer reaction times in sulfuric acid leads to significant production of the sulfonated derivative (18).
Neither of these two routes is suitable for manufacturing OEBC on a large scale. The first route gives low yields of the Nickel or copper benzochlorin, generally not greater that 50%, and demetallation of the strongly bound metals (Nickel and Copper) has historically used sulfuric acid. As sulfonation of OEBC occurs rapidly in sulfuric acid (within 3 hours), demetallation of (17) invariably results in the formation of the sulfonated analog, which must be separated by chromatography. The yields thus are disappointing.
The use of diisobutyl aluminium hydride (DIBALH) to reduce the ester functionality of the meso-acrylate porphyrin (11) following Morgan's exact methodology as reported, [(200 mg (11) in THF (100 mL; dry), −78° C./N2; add DIBALH in THF (20 mL of 1M solution (63 equivalents)), stir 1 hr at −78° C.; add water (100 mL) followed by 10% NaOH solution (100 mL) and water (200 mL)] does not work. When we repeated this reported protocol only starting material was isolated. Indeed, the starting material precipitates out of the THF on the addition of water and sodium hydroxide solution. There are no organic layers formed as reported and no appreciable amounts of (12) formed by NMR or TLC. We have observed that the reduction of the ester functionality in THF is an extremely slow reaction, requiring more that 2 days stirring (under Morgan's conditions) to see any appreciable amount of (12). Even under these conditions, the product is mostly (11). Reactions using LiAlH4 under a variety of conditions give unsatisfactory quantities of (12), which are invariably contaminated with other multiple products by TLC.
An alternative route to the synthesis of NiOEBC has been described by Arnold and co-workers (Scheme 9). In this methodology, the Nickel derivatives (10) or (14) are reduced to (19) with LiAlH4
respectively. Reduction of (10) produces (19) in 36% yield. Treatment of (19) with acid gives Nickel OEBC in 28% yield.
All of the known reported methods for the synthesis of NiOEBC or OEBC directly, suffer from low yields or products that require chromatography to purify. Indeed, the cyclization of the porphyrin precursors in sulfuric acid form sulfonated products like (18). As a result, none of the reported methods is suitable for manufacturing OEBC on a large scale. We have discovered methods that give excellent yields of OEBC from either (11) or NiOEBC, which are described below in detail.
Production of OEBC from meso-acrylate octaethylporphyrin (11)
The reduction of the ester group in (11) using DIBALH in THF has been shown to be prohibitively slow for use on a large scale (or even a small scale). We have discovered that the reduction reaction rate is entirely dependent on the solvents utilized in the reaction. If the reduction is undertaken in dichloromethane, using DIBALH in toluene (2.5 equivalents) as the reducing agent, large quantities (>100 g) of (11) are efficiently transformed to (12) in about 4 hours. If THF is used instead of dichloromethane to dissolve (11) and DIBALH in toluene is added to the reaction under the same conditions, the reaction is 1.5 to 2 times slower. Thus, it appears that the reduction of (11) with DIBALH is dependent on the solvent in which the DIBALH is dissolved. The inventors believe that this reaction proceeds most efficiently using a chlorinated solvent such as dichloromethane or 1,2-dichloroethane. The reduction is also dependant on the temperature of the reaction. If the addition of the DIBALH is not closely monitored and the temperature is allowed to rise, significant by-products occur in the reaction. In particular, metallated porphyrins are formed, presumably aluminium porphyrins, at higher temperatures (>−35° C.). It is preferred that the reaction be carried out between about −80° C. and about −-35° C.
Literature methods (Morgan et al) have reported isolating the alcohol (12) from the reduction reaction prior to cyclization to give OEBC. We have found that it is not necessary to isolate the alcohol (12) prior to cyclization. During the course of the reduction reaction, small aliquots are taken from the reaction mixture and neutralized with acetic acid/water or ethylacetate and ammonium hydrochloride solution. TLC indicates whether the reaction is complete or not. Once complete, excess DIBALH is quenched with isopropanol/methanol and an acid such as, for example, phosphoric acid (85%) is added. The organic volatiles (dichloromethane and toluene) are removed by distillation (rotoevaporation or other) and the phosphoric acid solution is heated at 60-130° C., preferably at about 100° C. for 3 hours to effect cyclization of (12) to OEBC. The OEBC is conveniently isolated from the phosphoric acid solution by precipitation with water (1.5×the H3PO4 volume). The OEBC is simply filtered from the acidic aqueous liquors. Any porphyrinic impurities remain protonated and soluble in the acidic aqueous mother liquors. The OEBC solid is then dissolved in dichloromethane and reprecipitated as before (using phosphoric acid/water) or from methanol or ethanol, via the distillation of the dichloromethane. The OEBC obtained in this manner is sufficiently pure to be used further (>97%) and is typically obtained in 70-75% yields. In addition to phosphoric acid, acids that can be preferably used in the above method include, for example, methane sulfonic acid and hydrochloric acid.
If the alcohol (12) needs to be isolated, we have found it convenient to quench the reaction with aqueous ammonia hydrochloride solution. Care must be taken during the quenching process to make sure that the quenching solution is acidic, as it appears that aluminium can be incorporated into the porphyrin. It should be noted also that TFA will not cyclize (12) to OEBC.
OEBC made in this way can be conveniently sulfonated at large scale to produce (18), by dissolving OEBC in sulfuric acid (with or without oleum). After the reaction is complete, (18) is conveniently isolated simply by adding the sulfuric acid to chilled water which precipitates the sulfonated product. It is then filtered and dried in a vacuum oven.
Both the reduction reaction and the acid catalyzed cyclization described in this section are believed to be applicable to any meso-acrylate (or similar) tetrapyrrolic compound on a large scale. The functionality on the periphery of the meso-acrylate tetrapyrrolic molecule, either at the meso or β-pyrrolic positions, can be varied widely as described earlier herein at pages 7-12. It would be within the knowledge of those skilled in the art what other functional groups are susceptible to modification via the reduction conditions or what modifications to this procedure could be made to utilize the invention described.
Demetallation of NiOEBC or CuOEBC without Sulfonation
The centrally coordinated metal (nickel or copper) of metallated benzochlorin or benzochlorin-type compounds requires strong acid conditions for removal (usually concentrated sulfuric acid) and long reaction times at room temperature (usually overnight). Unfortunately, sulfonation of the benzochlorin occurs readily in this solvent. It is thus very difficult to control the conditions necessary to obtain high yields of demetallated benzochlorins such as OEBC, without concomitant production of the corresponding sulfonated product (18). We have found that the central metal (nickel or copper) of OEBC can be efficiently removed by warming the compound to, for example, 80° C. in methane sulfonic acid. The reaction can be monitored by neutralizing small aliquots of the reaction, dissolving in dichloromethane and evaluating by TLC (30% hexane/dichloromethane). When deemed complete, the reaction can be diluted with ice water (equal volume) and the solid collected by filtration. The solid is washed with methanol or ethanol, redissolved in dichloromethane and precipitated from methanol, to give OEBC in about 80-90% yield, sufficiently pure to undergo further reactions (>97%).
The demetallation reaction described above is believed to be applicable to any benzochlorin or similar tetrapyrrolic compound on a large scale. The functionality on the periphery of the meso-acrylate tetrapyrrolic molecule, either at the meso or β-pyrrolic positions, can be varied widely as described earlier herein at pages 7-12. It would be within the knowledge of those skilled in the art what other functional groups could be used in order to use this demetallation reaction.
Tin Insertion into Tetrapyrrolic Compounds
The formation of metallated porphyrins, chlorins, bacteriochlorins and iso-bacteriochlorins is well established in the literature. Incorporation of metals into these tetrapyrrolic macrocycles can change the photophysical and pharmacokinetic attributes, distribution, metabolism and toxicology of the metallated compound from that of the parent metal-free molecule. In particular, one such metal, tin, has been incorporated into a number of tetrapyrrolic macrocycles that are of interest in medicine or phototherapy. Examples of these compounds include, for example, tin (dichloride) ethyl etiopurpurin I (SnET2; Rostaporfin), tin protoporphyrin (IX) [SnPP(IX)] and tin meso-porphyrin (IX) [SnMPP(IX)] (shown below). SnET2 is currently being evaluated as a photosensitizer in the treatment of age related macular degeneration. SnPP(IX) and SnMPP(IX), as their disodium salts, are currently being evaluated as heme oxygenase inhibitors that decrease the production of bilirubin in infants suffering from hyperbilirubinemia. While there have been few human studies thus far with SnPP(IX) in this role, promising results have been obtained. As the general applicability of photomedicine is realized in disease indications, there will be an increasing need for pure tin tetrapyrrolic macrocyclic compounds.
The invention described herein relates to the insertion of tin (II) complexes into tetrapyrrolic macrocycles to form tin (IV) metallo-tetrapyrrolic macrocycles. Such compounds, in addition to medicine or phototherapy, may also be useful as molecular wires or as templates for molecular or chiral recognition. In addition, these compounds may also be useful as pharmaceuticals, data storage devices, molecular switches or mimics of biosynthetic processes.
To achieve the advantages in accordance with the purpose of the invention, as embodied and broadly described therein, the inventors have found that a necessary component in the successful formation of highly pure tin (IV) metalloporphyrin complexes, is the abundance of molecular oxygen in the reaction mixture. In deed, tin insertion into a porphyrin molecule may be achieved at relatively low temperatures provided that an abundance of molecular oxygen is present in the reaction mixture.
Work in our laboratory has explored in detail the chemistry of tin insertion into tetrapyrrolic macrocycles. While there exists many reported methods for inserting tin into tetrapyrrolic macrocycles, we have discovered that even at relatively small scales (grams) the classical methods of inserting tin (II) into tetrapyrrolic macrocycles often lead to the formation of “reduced” tetrapyrrolic side products. Such products are often undesired and are extremely difficult to remove from the desired metallated products as purification of tin tetrapyrroles by chromatography on silica or alumina is extremely difficult. Scheme 10 represents a typical tin insertion reaction into a tetrapyrrolic macrocycle.
In the metal insertion process, Sn (II) is believed to be inserted into the tetrapyrrolic core as a Sn (II) cation, whereupon rapid oxidation occurs by oxygen or traces of oxidizing impurities to produce the isolated Sn (IV) tetrapyrrolic species. Evidence for the formation of Sn (II) tetrapyrrole complexes has been observed by the isolation of a Sn (II) phthalocyanine. Sn (II) complexes of porphyrins are relatively unknown.
In the metallation reaction, oxidation of Sn (II) to Sn (IV) occurs via oxygen present in the reaction, or other electron rich molecules according to the following equation:
Unfortunately, while the reaction sequence appears relatively straightforward, problems occur when tin metallation is attempted under normal atmospheric conditions in concentrations desirable for large scale manufacturing (or even manufacturing at small scale˜1 g). First, metal insertion is usually undertaken in solvents like glacial acetic acid, dimethylformamide or pyridine at, or close to, their boiling points in the presence of a proton scavenger that absorbs protons of metallation. A particularly preferred proton scavenger is sodium acetate, but others such as salts of other organic acids or amines could also be used effectively. The solvents quickly degas under elevated temperature conditions. In large-scale reactors, the volume between the reaction solvent and the top of the reactor vessel is called headspace. In order to maximize operating efficiency and lower the cost of plant production of the material, headspace is preferably kept to a minimum. Reactions are generally carried out at the highest concentrations of reactants possible in order to maximize efficiency. The combination of solvent degassing, solvent saturated headspace at the reflux temperature, and highly concentrated reaction solutions, leads to less than optimal results in the formation of pure Sn (IV) tetrapyrroles.
The driving force for tin insertion into the macrocycle is the reduction of Sn (II) to Sn (IV). As the metal desires to incorporate into the macrocycle, tetrapyrroles with reducible bonds undergo reductions. Such reductions typically occur on the ring of the macrocycle or at groups on the periphery of the molecule (such as vinyl groups etc.). The following are observed examples of unwanted reactions at the porphyrin periphery, which adequately illustrate the observed problem.
Tin metallation of Methyl pyrropheophorbide (20)
Attempts to insert SnCl2
into (20) under standard reaction conditions (500 mg of (20) in 70 mL of AcOH, 5 equivalents of NaOAc, 7 equivalents of SnCl2
, followed by reflux) results in the formation of the desired tin metallated pyrropheophorbide (21) and
appreciable amounts (>20%) of (22) in which the peripheral vinyl group has been reduced. If pyridine or dimethylformamide is used as solvent instead of acetic acid, the major product of the reaction is (22).
Tin metallation of Ethyl etiopurpurin (7)
Attempts to insert SnCl2 into (5) under standard reaction conditions (1 g of (5) in 100 mL of AcOH, 5 equivalents of NaOAc, 7 equivalents of SnCl2, followed by reflux) results in the formation of the desired tin metallated purpurin (SnET2) (7) and appreciable amounts (5-10%) of SnET2H2 (24) in which the peripheral vinyl group of the isocyclic ring has been reduced. If pyridine or dimethylformamide is used as solvent instead of acetic acid, the major product of the reaction is (24). Additionally, the reaction slows substantially if adequate oxygen is not present. We have found that the formation of the impurity (24) is reduced to levels below the limit of quantitation by HPLC if air (or other oxygen containing gas) is bubbled through the solution during the metallation reaction.
Tin metallation of meso-acrylate etioporphyrin I
Attempts to insert SnCl2 into (4) under standard reaction conditions (1 g of (4) in 100 mL of AcOH, 5 equivalents of NaOAc, 5 equivalents of SnCl2, followed by reflux) results in some formation of the desired tin metallated porphyrin (25), however the major product of the reaction is (26) in which the peripheral vinyl group of the acrylate has been reduced. If pyridine or dimethylformamide is used as the solvent instead of acetic acid, the major product of the reaction is (26). We have found that the meso-acrylate tin porphyrin (25) can only be formed in high purity when dichloroethane is used as a solvent and a vigorous stream of air (or other oxygen containing gas) is bubbled through the solution.
Tin metallation of protoporphyrin IX, dimethyl ester (27)
Attempts to insert SnCl2 into (27) under standard reaction conditions (1 g of (27) in 100 mL of AcOH, 5 equivalents of NaOAc, 7 equivalent of SnCl2, followed by reflux) results in the formation of the desired tin metallated porphyrin (28). However, by NMR at least 2 major impurities (ranging from 5-15%) are observed. These may be compounds (29) to (31). If pyridine or dimethylformamide is used as the solvent instead of acetic acid, the impurity products increase to 15-25%. The formation of these impurities is limited substantially (to undetectable levels by NMR (<0.4%)) if air (or another oxygen containing gas) is vigorously bubbled through the solution during the metallation reaction. As the solubility of tin salts in a number of common metallation solvents is low, the inventors have also found it advantageous to pre-dissolve the tin salt in a suitable solvent like dimethylformamide, prior to addition to the bulk reaction. This enables rapid incorporation of the tin salt into the macrocycle. Other examples of tin salts that can be used in the invention in addition to SnCl2 include, for example, Sn(Oac)2 and Sn(acac)2. In addition to dimethylformamide, predissolving solvents suitable for use in this process include, for example, acetic acid, propionic acid, or pyrridine.
Additionally, it is recognized in the art that some metallo-tetrapyrrolic macrocycles such as cadmium tetrapyrrolic will exchange the coordinated metal for a second metal. Such metal exchange reactions are encompassed within the present invention.
In accordance with the invention, as embodied and broadly described herein, the inventors have found that if adequate air is used in the reaction, high purity tin tetrapyrrolic compounds are formed from their corresponding non-metallated tetrapyrrolic compounds. It should be noted that a gas with an oxygen content of at least about 14% by weight is particularly advantageous. In general, many of these compounds crystallize or precipitate from the reaction mixture itself, enabling effective isolation by filtration.
The inventors have found that to obtain highly pure product (>99%), excess salts or impurities can be effectively removed from the reaction mixture on a large scale by precipitation techniques. In the case of tin ethyl etiopurpurin I (7), for example, the crude precipitate of (7) from the metallation reaction can be reprecipitated by first dissolving the purpurin in dichloromethane, adding acetic acid and removing the dichloromethane by distillation. In addition to dichloromethane, other solvents can be used to dissolve the purpurin, such as ether, dichloroethane, chloroform, toluene, or benzene. Other solvents in addition to acetic acid that have been found effective at precipitating the tin purpurin product include, for example, acetone, ethanol, methanol, dimethylformamide and acetonitrile. Of these, acetone and ethanol are preferred. This precipitation technique works similarly with almost all tin tetrapyrrolic compounds on a larger scale (>100 g). In particular, this precipitation technique works effectively with (21), (24), (25), (28) and (31) on scales greater than 50 grams. Of particular note is that the use of high quality solvents with low water contents (especially in the alcohols and acetone) greatly lowers the potential of ligand exchange on the centrally coordinated tin compounds.
It would be within the knowledge of those skilled in the art that other solvents can be used to effectively precipitate the tin tetrapyrrolic complexes. These would include hexanes and the like, ethers and the like, and other alcohols. Examples set forth in the experimental section highlight the general applicability of the precipitation technique.
Described above are general procedures for the large scale manufacturing of tetrapyrrolic compounds such as meso-formyl porphyrins, meso-acrylate porphyrins, purpurins, tin metallated tetrapyrroles and benzochlorins. Purification is readily achieved by a series of fractional crystallizations. Additional advantages of the invention will be set forth in the detailed examples that follow, and in part will be obvious from the description supplied or may be learned by practice of the invention. The advantages of the invention can be realized and attained by means of the elements and combinations particularly pointed out in the appended claims. The following examples are included to highlight the advantages over the existing methods and are in no way intended to limit the invention.