BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention is a process for the dehalogenation of 9α- halo steroids (I) to produce the corresponding 11-oxygenated steroids (II) which are known to be useful as pharmaceuticals. The process can also be used to dehalogenate other steroids and non-steroids to produce compounds which are useful in the production of pharmaceuti¬ cals. 2. Description of the Related Art

11-Oxygenated steroids, especially corticoids, are very impor¬ tant commercial steroids. These can generally be made by two methods; first by fermentation with a microorganism which will hydroxylate at the 11 position in the β configuration and second by starting with a £? s ^L . ⁾ _ _or Δ^-steroid, forming the halohydrin and then dehalogenating. Processes for dehalogenation of 9α-halo-ll- oxygenated steroids and 12α-halo-11-oxygenated steroids are known.

US Patent 3,480,622, Tetrahedron Letters 3151 (1964) and J. Am

Chem. Soc, 88, 3016 (1966) describe a process for the debromination of a 9α-bromo-11-oxygenated steroid by reaction of the brominated steroid with several equivalents each of a chromium (II) salt and a hydrogen atom donor such as n-butane thiol. This process has several disadvantages. One is that it generates a i t- ¹ -)-steroid and a 5,9- cyclosteroid as by-products, see US 3,480,622, column 3. The Δ ⁹ ^ ) compound produced is difficult to separate from the desired 11- oxygenated product (II). Another is that the chromium (II) salt and the chromium (III) salt by-product are toxic, so the process creates a toxic waste disposal problem. Another disadvantage is that the process requires a polar solvent such as DMF, DMSO and DMAC which are expensive.

US Patents 4,304,727 and 4,325,881 improved upon the process of US Patent 3,480,622 by using thioglycolic acid as the hydrogen atom donor.

US Patent application Serial No 88/02430 filed July 22, 1988 improves upon these processes by (1) adding the 9α-halo steroid to the chromium ion rather than adding the chromium ion to the 9α-halo steroid and (2) using < 1 equivalent of soluble chromium ion in the presence of a reagent capable of converting chromium (III) to

chromium (II) . This process offers significant advantages over the original process. However, it still generates chromium (III) as a waste product (although a lesser amount) and requires the use of expensive polar solvents. It is known that the following reagents also dehalogenate alkyl halides: Raney nickel, palladium and hydrogen, platinum and hydrog¬ en, zinc/acetic acid, cobalt phthalocyanine/sodium borohydride, lithium trimethoxyaluminum hydride/cuprous iodide, vanadium di- chloride, lithium di-n-butyl-9-BBN and sodium cyanoborohydride/zinc chloride. However, none of the above agents which have been tried effectuated conversion of the 9σ-halosteroids (I) to the correspond¬ ing 11-oxygenated steroids (II) cleanly. For example, treatment of 9α-bromohydrocortisone 21-acetate (I) with zinc/ethanol/water; pall¬ adium/carbon, hydrogen and Raney nickel/diox./ether each produce the Δ ⁹ ^H) compound as the major product, see J. Am. Chem. Soc, 79, 1130 (1957). Treatment of 9α-bromoprednisolone 21-acetate (I) with chromous chloride/acetone produces a mixture of the Δ*'^' compound and the 5,9-cyclosteroid, see Proc. Chem. Soc. 207 (1961).

It has been known for some time that organotin hydrides such as tri-n-butyltin hydride dehalogenate 9α-halo steroids (I) to the corresponding 11-oxygenated steroids cleanly without formation of any Δ'C ¹¹ or 5,9-cyclosteroid by-product, see J. Org. Chem., 44, 151 (1979), footnote 25 in J. Am. Chem. Soc, 88, 3016 (1966) and US Patent 3,894,063. However, organotin hydrides are both expensive to prepare and are themselves toxic, see Kirk-Othmer Encyclopedia Chem. Tech. 23, 69 (1983). These disadvantages have precluded these reagents from being employed commercially.

The expense can be reduced by using a catalytic amount of tri-n- butyl tin chloride together with a stoichiometric amount of a cheap hydride_donor (lithium aluminum hydride, sodium borohydride or sodium cyanoborohydride) , see J. Org. Chem., 28, 265 (1963), J. Org. Chem., 40, 2554 (1975), J. Am. Chem. Soc, 108, 303 (1986) and J. Org. Chem., 52, 472 (1987). However, the problem of toxicity is not eliminated simply because a lesser amount of the organotin reagent is used. Also, the lithium aluminum hydride and sodium borohydride variants of this process are inoperable in the case of 11-oxygenated steroids (II) because these metal hydrides rapidly reduce the carbonyl group at C3, C20 and at C^ (if present).

Another dehalogenation strategy that has been extensively inves¬ tigated is that of using an organotin hydride that is attached to a polymer. Polymeric organotin hydrides have been described as useful for dehalogenating alkyl halides, see J. Org. Chem., 40, 1966 _^ (1975), US Patent 3,975,334 and J. Chem. Soc. Chem. Comm. , 798 (19 ^* 854) . These polymeric organotin hydrides can be easily separated from the desired 11-oxygenated steroid (II) products by filtration so their toxicity is irrelevant. However, these polymeric reagents are prepared by multi-step sequences and thus are too expensive to have bmmercial utility. In the case of the polymer described in J. Org. Ghem., 40,

1966 (1975) , an additional disadvantage is that regeneration ^s of the hydride form from the halide form cannot be accomplished effiςiently.

Inorganic tin(II) salts are both less expensive and less toxic than organotin hydrides (see Kirk-Othmer Encyclopedia Chem. Tech. , 23 (1981) in particular pages 28 and 51). These salts are capable of effecting dehalogenation of certain "activated" alkyl halides (i.e., those in which the halogen is unusually susceptible to reduction) such as α-chloroacetophenone. The ease of dechlorination of a- chloroacetophenone is well documented; fro example, see J ^; . Am. Chem. Soc, 85, 2768 (1963). The following are examples of inorganic tin reagents that dechlorinate α-chloroacetophenone: stannous chloride/- diisobutylaluminum hydride [Chem. Lett. 2069 (1984)], stannous chloride/sodium borohydride [Z. Naturforsch. B., 416, 156 ^' 8 (19860], stannous chloride/sodium bromide [Synthesis, 570 (1986)], 'stannous chloride/sodium thiocyanate [Synthesis, 406 (1987)] and stannous chloride/sodium hydrosulfide [Syn. Communications, 16, 653 (1986)]. However, inorganic tin(II) salts do not dehalogenate simple, un- activated alkyl halides such as n-octyl bromide or chloride: see J. Organometallic Chem., 97, 167 (1975), where it is stated that "... reactions of stannous bromide or chloride with alkyl bromides or chlorides do not proceed." None of the above-listed inorganic tin (II) salts dehalogenate 9α-halosteroids (I) . ^J

Other functional groups that inorganic tin(II) salts are capable of reducing include organic azides (by -stannous chloride), see Tetrahedron Lett., 28, 5941 (1987) and nitro compounds (by stannous chloride hydrochloric acid), see Org. Synthesis Coll. Vol. II, 130 (1943). However, the fact that these functional groups are highly susceptible to reduction is well documented. For example, even

ammonium sulfide and sodium hydrosulfide, which do not reduce unactivated alkyl halides, reduce organic azides and nitro compounds at room temperature, see Tetrahedron Lett., 2031, (1974) for azides and Org. Reactions, 20, 455 (1973) for nitro compounds. Thus, the utility of inorganic tin (II) salts is limited by their low reducing power to reduction of easily reducible functional groups.

Changing the ligands on the inorganic tin(II) salts from chlorine to alklythio in order to enhance its reactivity is known. For example, see Tetrahedron Lett., 28, 5941 (1987), where it is reported that triethylammonium tris(phenylthio)stannite [ €2^ 2^ SN(S^)3 ^* ] is a more powerful reducing agent than is stannous chloride toward organic azides. However, even this reagent is not useful for dehalogenation of 9α-halosteroids (I) or other unactivated alkyl halides. Thus simply increasing the reducing power of the inorganic tin(II) salts alone does not solve the problem of dehalogenation of unactivated alkyl halides.

This invention consists of a process for dehalogenation of 9a- halo steroids (I) to the corresponding 11-oxygenated steroids (II) by treatment with either tin(0) , a tin(II) salt, lead(0) or a lead(II) salt combined with a hydrogen atom donor and initiator. Inclusion of a weak base is preferred in some cases, but is not required for the process to be operable. Combining the metals or metal(II) salts with the hydrogen atom donor and initiator creates new reagents that are capable of reducing relatively unreactive functional groups such as unactivated alkyl halides.

SUMMARY OF INVENTION Disclosed is a process for the preparation of an 11-oxygenated steroid of formula (II) where

Rg is o-Rg.^:3-R .2, where one of Rg_^ and Rg.2 is -H and the other is -H, -F or -CH ₃ ;

R _n is -0, α-H:0-0H or α-0H:0-H; which comprises

(1) contacting a 9σ-halo steroid of formula (I) where Rg is -Cl, -Br or -I and where Rg and R _j ^ are as defined above with a metal selected from the group consisting of tin (0) , a tin (II) salt, lead (0) or a lead (II) salt, a hydrogen atom donor and an initiator.

Also disclosed is a process for the preparation of a dehalo-

genated compound selected from the group consisting of compounds of the formulas (IV), (VI), (VIII), (X), (XII), (XIV), (XVI), (XVIII), (XX) and (XXII) where .... is a single or double bond; R3 is -H or -CO-R3. where R3-.1 is C -C or -φ ; Rιι is -0 or ^α -Ril-l ^: /9-Rιi-2 where one of Rχχ.l and Rχχ_2 s -H and the other is -H or -OH;

R 3 is is -H, C -C alkyl or -CO-Rχ3_ι where 13.1 is -H, C1-C3 alkyl or -φ ;

(D-I) R^ is β-Rι -i:3- ι .2 where one of Rχg.χ and χ .2 ^is " ^H and the other is -H, -OH or -CH3 and R ₁₇ is

-0, α-H:0-C0-CH ₃ , where Ri7_χ is -H or -CO-Rχ7_2 ^where R17-2 ^is C -C3 alkyl or φ and where 2I-I is -H or -CO-R2χ_2. where 21-2 ^ ^{s c} l"^3 a k l or φ optionally substituted with -Cl or - O2, ^α*OR 17-3 ^:9 " ^CN wh re ^R 17-3 ^is -H, THP, -CH ₂ -OCH ₃ ,

-CHR^7-.3 -0-R^7_32 where R17.31 is C1-C3 alkyl and R17.32 is C -C alkyl or φ and

^"SiR 17-33 ^R 17-34 ^R 17-35 ^{where R} 17-33- ^R 17-34 ^{and R} 17-35 ^are the same or different and are selected from the group consisting of C1-C alkyl, C1-C4 alkoxy, C1-C monohaloalkyl where halo is -Br or -Cl, φ optionally substituted with 1 or 2 -OCH3 or - H2; c--ORχ ₇ _4: -CO-CH3 where R17. is -H, -CO-R17. 1 where R17. 1 is C2-C4 alkyl or φ optionally substituted with 1 or 2 -OCH3, (D-II) the 16,17-acetonide of a compound where ι .χ is -OH, and where R ₇ is o-ORχ7.χ:^-CO-CH2-OR2χ.χ where R17.1 is -H where 21-I is as defined above; where n^ is 9-20; which comprises (1) contacting the corresponding halogenated c π ound of the formula (III), (V), (VII), (IX), (XI), (XIII), (XV), (XVII), (XIX) and (XXI) where X^ is -Cl, -Br or -I and where R3, RJ , R ₁₃ ,

^R 16» ^R 17 ^{and n} l ^{are as} defined above with a metal selected from the

group consisting of tin (0), a tin (II) salt, lead (0) or a lead (II) salt, a hydrogen atom donor and an initiator.

DETAILED DESCRIPTION OF THE INVENTION The present invention is a process for the preparation of 11- oxygenated steroids of formula (II) which comprises contacting 9α- halo steroids of formula (I) with (1) tin (0) , a tin (II) salt, lead (0) or a lead (II) salt, (2) a hydrogen atom donor and (3) an initiator. It is preferred that a weak base also be present during the contacting. The 9α-halo steroidal (I) starting materials are either known to those skilled in the art or can readily be prepared from known com¬ pounds by means known to those skilled in the art.

The tin (0) or lead (0) are added as a finely divided powder; tin (0) is preferred. Tin (II) and lead (II) salts are also oper- able; virtually any such salt is operable. For example, preferred stannous salts include, fluoride, chloride, bromide, iodide, oxalate, oxide, sulfide, methoxide, ethoxide, phenoxide, catecholate, ethylene glyoxide, 1,3-propylene glyoxide, acetate, propionate, 2-ethylhexano- ate, pyrophosphate, sulf te, tetrafluoroborate as well as dimethyl stannane and dicyclopentandienyl stannane. More preferred stannous salts are chloride, 2-ethylhexanoate and ethylene glyoxide. Examples of preferred lead (II) salts include fluoride, chloride, bromide, iodide, oxide, acetate, carbonate, sulfate, perchlorate, thiocyanate, 2-ethylhexanoate, methylmercaptide, orthosilicate, orthophosphate, 2,4-pentanedionate, sulfide, stannate, tetrafluoroborate, tungstate, zirconate, trifluoroacetate and titanate. More preferred are 2-eth- ylhexanoate, acetate and chloride.

If less than two equivalents of tin (0) or lead (0) are used, the dehalogenation is incomplete. The amount required for reaction Is about 2 equivalents in the case of 325 mesh tin (0) powder. With other forms of tin metal such as foil, wire, shot, stick, bar, granu¬ lation and mossy tin the prosess is operable but a larger excess is required because their surface areas are smaller. The 325 mesh tin powder is the preferred metal (0) . The amount of tin (II) or lead (II) salt needed for the reaction to go to completion is about 2 equivalents. With less the reaction produces dehalogenated product but does not go to completion. Use of greater than about two equiv¬ alents of tin (II) or lead (II) salt is wasteful but not deleterious.

The hydrogen atom donor is any compound which is capable of donating a hydrogen atom to the halogen-bearing carbon atom of the 9α-halo steroid (I). As used in this patent a hydrogen atom donor is defined as any compound which when contacted with 9α-bromopr^dnisol- one 21-acetate (I) and tin (0) in sec-butanol for 1 hour at aboust 75 to about 80° produces prednisolone 21-acetate (II) in a higher yield that that which is obtained using hypophosphorus acid, a known hydro¬ gen atom donor. Suitable hydrogen atom donors include hydrogen brom¬ ide, hydrogen iodide, hypophosphorous acid, 1,2- and 1,4-dihydrobenzene, 1,2 and 1,4-dihydrotoluene,

1,2- and 1,4-dihydro-(ortho, eta, or para)-xylene, 1,4-dihydronaρhthalene, 9,10-dihydroanthracene, cyclopentadiene,

1-benzyl-1,4-dihydronicotina ide, 3,5-di-t-butyl- -hydroxytoluene, H-Si-(X ₂ )3, H-Sn-(X ₂ )3, H-Ge-(X ₂ ) ₃ , H-P-(X ₂ )2- H-Se-X ₂ , H-B- (X2 3 or H-S-X2 where, when more than one 2 is present the X2*s are the same or different and are -H, ^c l" ^c 10 ^a lkyl, ^c 5" ^c 10 cycloalkyl, α-napththyl, -naphthyl,

-φ optionally substituted with 1 or 2 X3, where X3 is ^' -F, -Cl, _ -Br,

-I, -Φ , alkyl, C5-C7 cycloalkyl, -OX4 where X4 is -H, Cχ-C5 alkyl or C5-C7 cycloalkyl,

-COOX4 where X4 is as defined above,

-N(X5)2 where the X5's are the same or different and are -H, C -C5, and where the X5's are taken' together with the

attached nitrogen atom and optionally another heteroatom, to form a heterocyclic ring selected from the group consisting of pyrrolidine, piperidine, morpholine, piperazine or N-methylpiperazine, 3-mercaptoρropionic acid, mercaptoacetic acid,

2-mercaptoρropionic acid, ethane dithiol, propane 1,3-dithiol. The preferred hydrogen atom donors are H- S-X ₂ . The amount of the hydrogen atom donor required is from about 2 to about 10 equivalents, preferrably about 2 to about 5 equivalents, more preferrably about 2 to about 4 equivalents. If less than two equivalents are used, dehalogenated product is produced but undesire- able side reactions occur. In the case of 3-mercaptopropionic acid, tin (0) and 9α-bromoprednisolone 21-acetate, 2-4 equivalents are required to suppress these side reactions. Initiators are selected from the group consisting of AIBN, oxygen, triethylborane optionally combined with oxygen, 1,1'-azobis(cyclohexane carbonitrile) , diben- zoylperoxide, lauroyl peroxide, succinic acid peroxide, di- -butyl- peroxyoxalate, di(2-ethylhexyl)peroxydicarbonate, t-butylperbenzoate, t-amylperoxypivalate, di-t-butylperoxide, dicumylperoxide, light of about 250 to about 350 μ optionally combined with benzophenone, t-butylhydroperoxide optionally combined with acetic acid, hydrogen peroxide combined with ferrous perchlorate, t-butylperbenzoate com- bined with cuprous bromide, chloride or iodide, methylethylketone peroxide combined with cobalt naphthenate or cobalt octoate, potas¬ sium persulfate and potassium nitrosodisulfonate.

Air can fulfill the function of an initiator. For example, it has been found that when argon was bubbled through a reaction mixture containing a 9α-bromoprednisolone 21-acetate (I), 6.03 equivalents of tin (0) powder, and 10.02 equivalents of 3-mercaptopropionic acid in THF to purge all traces of dissolved oxygen, only about 15% of the 9o-bromoprednisolone 21-acetate (I) had been debrominated after 63 hr at 20-25*. When the reaction mixture was then exposed to air, conversion to prednisolone 21-acetate (II) speeded up dramatically and was complete within 45 hr. Air is not a preferred initiator because it oxidizes the tin (0) and thiol (if present) competitively. Preferred initiators are AIBN and dibenzoylperoxide; most preferred

is AIBN .

The amount of initiator is not critical. The process operates with 0.1 mole % or lower, but is slow. Other factors being equal, the greater the amount of initiator, the faster the reaction rate. The amount that results in the most convenient rate is about 1 to about 10 mole %.

It is preferred to include in the reaction mixture approximately one molar equivalent of a weak base. A weak base is a base whose conjugate acid has a pK _a of about 2 to about 12. Suitable weak bases include compounds selected from the group consisting of 3-mercaptopropionate;

(Xg)3-N where the Xg's are the same or different and are -H,

Cx-Cxo alkyl, C5-C10 cycloalkyl, α-naphth l, y8-naρhthyl,

-φ optionally substituted with 1 or 2 X7, where X7 is -F, -Cl,

-Br, -I, -Φ ,

Cχ-C ₅ alkyl, ^c 5 ^_c 7 cycloalkyl,

-OXβ where X ₈ is -H, Cχ-C ₅ alkyl or C5-C7 cycloalkyl, -COOX8 where Xg is as defined above,

-N(Xg)2 where the Xg's are the same or different and are -H, C^-C^ , and where the Xg's are taken together with the attached nitrogen atom and optionally another heteroatom, to form a heterocyclic ring selected from the group consisting of pyrrolidine, piperidine, morpholine, piperazine or N-methylpiperazine, Xg-COO ^* where Xg is as defined above; citrate; oxalate; tartrate; imidazole; N-methylimidazole;

2-me hylimidazole; pyridine;

4-dimethylaminopyridine; (any isomer of) lutidine; collidine,

N,N' -dimethylpiperazine; 1,4-diazabicyclo[2.2.2]octane; 1,8-diazabicyclo[5.4.0]undec-7-ene or 1,5-diazabicyclo[4.3.0]non-5-ene. Without the weak base the reaction generates minor amounts of less polar impurities, which makes purification of the final product more difficult. A strong base can also be used if the hydrogen atom donor is 3-mercaptoproρionic acid, since the strong base will convert the 3-mercaptopropionic acid to 3-mercaptopripionate salt, which serves as the weak base, see, for example, EXAMPLE 1 in which potassium tertiary butoxide is used.

The preferred reaction temperature is determined by the partic¬ ular initiator used. Acceptable and preferred temperature ranges for various initiators are set forth in TABLE 1, see CHART C. The reaction is performed either neat (in which case the hydrogen atom donor is the solvent) or in an inert solvent. Suitable inert solvents include water, alcohols, carboxylic acids, substituted amides, aliphatic or aromatic hydrocarbons, halogenated hydrocarbons, cyclic ethers, esters, ketones, nitriles and mixtures thereof. The 9α-halo steroids (I) need not even be soluble for the reaction to take place. In fact, superior yields are obtained in solvents in which the 9α-halo steroid (I) is not soluble. Thus the more pre¬ ferred solvents are 1,2-dichloroethane, sec-butanol and isopropanol. Even more preferred are sec-butanol and isopropanol because they are water-miscible so that the 11-oxygenated steroid (II) product can be readily obtained by water knock-out. The preferred amount of solvent is 2-8 ml/g of 9o-halo steroid (I), although the process operates efficiently with greater or lesser amounts.

The 9α-halo steroid (I), metal (0) or metal (II) salts, hydrogen atom donor, initiator, weak base and solvent can be mixed in any order. After mixing, the temperature is raised to the temperature at which the reaction occurs. While the reaction is occuring the mix¬ ture is stirred or agitated vigorously so that the metal (0) or metal

-11-

(II) salt is dispersed evenly throughout the reaction mixture.

When the reaction is complete, the 11-oxygenated steroids (II) are isolated and purified by known means.

Besides being useful for the conversion of the 9α-halo steroids (I) to the corresponding 11-oxygenated steroids (II) , the process of the present invention is also useful in the same way for the conver- sion of the particular halogenated compounds of formulas (III, V, VII, IX, XI, XIII, XV, XVII, XIX and XXI) to the corresponding dehalogenated compounds of formulas (IV, VI, VIII, X, XII, XIV, XVI, XVIII, XX and XXII). The information regarding how to perform the process of the present invention with the 9α-halo sterolαs (I) is generally the same, with non-critical variations, for the halogenated

Λ compounds of formulas (III, V, VII, IX, XI, XIII, XV and XVII) .

The product 11-oxygenated steroids (II) and the dehalogenated compounds are either useful pharmaceutical agents or intermediates in the preparation of useful pharmaceutical agents.

DEFINITIONS AND CONVENTIONS The definitions and explanations below are for the terms as used throughout this entire document including both the specification and the claims.

I. CONVENTIONS FOR FORMULAS AND DEFINITIONS OF VARIABLES The chemical formulas representing various compounds or molecu¬ lar fragments in the specification and claims may contain variable substituents in addition to expressly defined structural features. These variable substituents are identified by a letter or a letter followed by a numerical subscript, for example, "Z^" or "R^ ⁿ where

A,

"ι" is an integer. These variable substituents are either monovalent or bivalent, that is, they represent a group attached to the formula by one or two chemical bonds. For example, a group Z± would repres- ent a bivalent variable if attached to the formula CH3-C(-Z )H. Groups R^ and Rj would represent monovalent variable substituents if attached to the formula C^-C^-C^)(Rj)^. When chemical formulas are drawn in a linear fashion, such as those above, variable sub¬ stituents contained in parentheses are bonded to the atom immediately to the left of the variable substituent enclosed in parenthesis. When two or more consecutive variable substituents are enclosed in parentheses, each of the consecutive variable substituents is bonded to the immediately preceding atom to the left which is not ^' enclosed

in parentheses. Thus, in the formula above, both ^ and Rj are bonded to the preceding carbon atom. Also, for any molecule with an established system of carbon atom numbering, such as steroids, these carbon atoms are designated as C^, where "i" is the integer corre- sponding to the carbon atom number. For example, Cg represents the 6 position or carbon atom number in the steroid nucleus as tradition¬ ally designated by those skilled in the art of steroid chemistry. Likewise the term "Rg" represents a variable substituent (either monovalent or bivalent) at the Cg position. Chemical formulas or portions thereof drawn in a linear fashion represent atoms in a linear chain. The symbol "-" in general repre¬ sents a bond between two atoms in the chain. Thus CH3-0-CH2-CH(R )- CH3 represents a 2-substituted-l-methoxypropane compound. In a similar fashion, the symbol "-" represents a double bond, e.g., CH2-C(R£)-0-CH3, and the symbol "■" represents a triple bond, e.g., HC"C-CH(R^)-CH2-CH3. Carbonyl groups are represented in either one of two ways: -CO- or -C(-O)-, with the former being preferred for simplicity.

Chemical formulas of cyclic (ring) compounds or molecular frag- ments can be represented in a linear fashion. Thus, the compound 4- chloro-2-methylpyridine can be represented in linear fashion by N*-C(CH ₃ )-CH-CC1-CH-C*H with the convention that the atoms marked with an asterisk (*) are bonded to each other resulting in the form¬ ation of a ring. Likewise, the cyclic molecular fragment, 4-(ethyl)- 1-piperazinyl can be represented by - *- CH2)2-N(C2H5 -CH2-C*H2.

A rigid cyclic (ring) structure for any compounds herein defines an orientation with respect to the plane of the ring for substituents attached to each carbon atom of the rigid cyclic compound. For saturated compounds which have two substituents attached to a carbon atom which is part of a cyclic system, -C(X ) X2)- the two substitu¬ ents may be in either an axial or equatorial position relative to the ring and may change between axial/equatorial. However, the position of the two substituents relative to the ring and each other remains fixed. While either substituent at times may lie in the plane of the ring (equatorial) rather than above or below the plane (axial) , one substituent is always above the other. In formulas depicting such compounds, a substituent (X^) which is "below" another substituent (X2) will be identified as being in the alpha (α) configuration and

is identified by a broken, dashed or dotted line attachment to the carbon atom, i.e., by the symbol "- - -" or "...". The corresponding substituent attached "above" (X2) the other (X _j _) is identified as being in the beta (β) configuration and is indicated by an unbroken line attachment to the carbon atom.

When a variable substituent is bivalent, the valences ^" may be taken together or separately or both in the definition of the variable. For example, a variable R^ attached to a carbon atom as -C(- .£)- might be bivalent and be defined as oxo or keto (thus form- ing a carbonyl group (-CO-) or as two separately attached monovalent variable substituents α-Ri-j and J-Ri-fc- When a bivalent variable, R^, is defined to consist of two monovalent variable substituents, the convention used to define the bivalent variable is of the form and J-Ri-.^ are attached to the carbon atom to give

-C α-R _j L_j) - • For example, when the bivalent variable Rg, -C(-Rg)- is defined to consist of two monovalent variable stibstitu- ents, two monovalent variable substituents are α-Rg_ : ?-Rg_2 α-Rg_ ₉ : -Rg_ ₁₀ , etc, giving -C(α-Rg. ₁ )( -Rg_ ₂ )- -C(α-Rg.g) (/9-Rg.ιo ^_ _> ^etc - Likewise, for the bivalent variable R ₁₍ -C(-R ₁ ^)-, two monovalent variable substituents are α.-R _χ: ?-Rι _2. For a ring substituent for which separate a and β orientations do not exist (e.g. due to the presence of a carbon carbon double bond in the ring) , and for a substituent bonded to a carbon atom which is not part of a ring the above convention is still used, but the α and β designations are omitted.

Just as a bivalent variable may be defined as two separate mono¬ valent variable substituents, two separate monovalent variable sub¬ stituents may be defined to be taken together to form a bivalent variable. For example, in the formula -C _j (Ri)H-C2 R )H- (C^ and C ₂ efine arbitrarily a first and second carbon atom, respectively) R^ and Rj may be defined to be taken together to form (1) a second bond between 0 and C2 or (2) a bivalent group such as oxa (-0-) and the formula thereby describes an epoxide. When R^ and R* are taken together to form a more complex entity, such as the group -X-Y- , then the orientation of the entity is such that 0 in the above formula is bonded to X and C2 is bonded to Y. Thus, by convention the designa¬ tion "... R^ and Rj are taken together to form -C^-CHijr-O-CO- ..."

means a lactone in which the carbonyl is bonded to C2. However, when designated "... Rj and R^ are taken together to form -CH2-CH2-O-CO- the convention means a lactone in which the carbonyl is bonded to C _] _. The carbon atom content of variable substituents is indicated in one of two ways. The first method uses a prefix to the entire name of the variable such as "C1-C4", where both "1" and "4" are integers representing the minimum and maximum number of carbon atoms in the variable. The prefix is separated from the variable by a space. For example, "C1-C alkyl" represents alkyl of 1 through 4 carbon atoms, (including isomeric forms thereof unless an express indication to the contrary is given). Whenever this single prefix is given, the prefix indicates the entire carbon atom content of the variable being defined. Thus C2-C alkoxycarbonyl describes a group CH3-(CH2 _n -0- C0- where n is zero, one or 2. By the second method the carbon atom content of only each portion of the definition is indicated sepa¬ rately by enclosing the "C^-C^" designation in parentheses and placing it immediately (no intervening space) before the portion of the definition being defined. By this optional convention (C -C3) alkoxycarbonyl has the same meaning as C2-C alkoxycarbonyl because the "C -C3" refers only to the carbon atom content of the alkoxy group. Similarly while both C2"Cg alkoxyalkyl and (C -C3)alkoxy(C^- C3)alkyl define alkoxyalkyl groups containing from 2 to 6 carbon atoms, the two definitions differ since the former definition allows either the alkoxy or alkyl portion alone to contain 4 or 5 carbon atoms while the latter definition limits either of these groups to 3 carbon atoms.

II. DEFINITIONS All temperatures are in degrees Centigrade. TLC refers to thin-layer chromatography. THF refers to tetrahydrofuran.

DMF refers to dimethyIformamide. DMSO refers to dimethysulfoxide. DMAC refers to dimethylacetamlde. AIBN refers to 1,1'-azobis[isobutyronitrile] . φ refers to phenyl CgH5 .

When solvent pairs are used, the ratios of solvents used are volume/volume (v/v) .

Hydrocortisone refers to ll ,17α,21-trihydroxypregn-4-ene-3,20-

^* ; dione .

Prednisolone refers to ll^,17α,21-trihydroxypregna-l,4-diene-

3,20-dione.

_ indicates that there are 2 possible orientations for the attached group, (1) a or β when attached to the steroid ring ^" and (2) cis or trans when attached to a carbon atom of a double bond.

EXAMPLES

Without further elaboration, it is believed that one skilled in the art can, using the preceding description, practice the present invention to its fullest extent. The following detailed examples describe how to prepare the various compounds and/or perform the various processes of the invention and are to be construed as merely illustrative, and not limitations of the preceding disclosure in any way whatsoever. Those skilled in the art will promptly recognize appropriate variations from the procedures both as to reactants and as to reaction conditions and techniques.

PREPARATION 1 3α-Iodo-5α-cholestane (XIX)

Following the general procedure described in J. Chem. Soc.

Perkin I. 2866 (1980) a solution of dihydrocholesterol (5.0000 g) , imidazole (2.6272 g) and triphenylphosphine (6.7481 g) in toluene

(124 ml) is heated to reflux, then treated with iodine (9.7,970 g) .

After 75 min, conversion to 3α-iodo-5α-cholestane is complete as measured by TLC (ethyl acetate/heptane, 1/1). The reaction mixture is then cooled to 20-25°, washed with saturated aqueous sodium bicarbonate (3 x 50 ml) followed by aqueous sodium thiosulfate (15%,

2 x 50 ml) , followed by water (50 ml) . The toluene phase is then dried over magnisium sulfate and concentrated to a solid. The solid is chromatographed on silica gel (330 g) eluting with ethyl acetate/ hexane (1/99) . The approprate fractions are pooled and concentrated to give the title compound.

EXAMPLE 1 Debromination of 9α-Bromoprednisolone 21 _; Acetate (I) to Prednisolone 21-Acetate (II) ' ^" Using Tin Powder'

A solution of potassium tertiary butoxide in THF (20% 10.QS79 g) is added dropwise to a solution of 3-mercaptoproρiσnic acid (6.3849 g) in secondary butanol (30 ml). The resulting mixture is treated with tin powder (325 mesh, 7.1251 g) followed ;by 9α*bromo- prednisolone 21-acetate [I, JACS 77, 4438 (1955), 7.2211 g) followed by AIBN (0.4944 g) . Secondary butanol (30 ml) is used for a rinse.

The resulting slurry is heated at 82° under an atmosphere of nitrogen with vigorous stirring until TLC (acetone/methylene chloride/cyclo- hexane; 2/5/3) indicates that conversion of the 9 -bromoprednisolone 21-acetate to prednisolone 21-acetate is complete (8 hr) . After 30 min the reaction mixture is added dropwise to water (600 ml) at 0°. The resulting mixture is stirred at 0* for 25 min. The solids are filtered off and dried by a stream of air. The solids are stirred in methylene chloride/methanol (1/1, 2 x 75 ml) and filtered. The filtrate is concentrated under reduced pressure to give a solid which is prednisolone 21-acetate in essentially pure form as measured by TLC.

The prednisolone 21-acetate is taken up in methylene chloride/ methanol (1/1, 54 ml) with a brief warming over a steam bath. This mixture is treated with magnesol (0.503 g) and activated charcoal (0.501 g), stirred at 30° for 15 min, filtered thru celite and the filter cake washed with methylene chloride/methanol (1/1, 5 x 5 ml). The filtrate is then concentrated by distillation thru a short-path (bath temperature is 65°) to a volume of 18 ml. The resulting slurry is then diluted with ethyl acetate (36 ml) and reconcentrated to 18 ml. Again the mixture is diluted with ethyl acetate (36 ml) and reconcentrated to 18 ml. The slurry is diluted with ethyl acetate (17 ml) and water (1 ml), reconcentrated to 18 ml, cooled to -20° for 15 min and filtered. The filter cake is washed with -20* ethyl acetate (3 x 5 ml). The cake is then dried by a stream of nitrogen to give prednisolone 21-acetate.

EXAMPLE 2 Debromination of 9α-Bromohydrocortisone 21-Acetate

(I) to Hydrocortisone 21-Acetate (II) Using Tin Powder

3-Mercaptopropionic acid (2.1312 g) is treated with solid sodium bicarbonate (0.4215 g) and stirred at 20-25* for 30 min. 1,2-Di- chloroethane (20 ml) is then added, followed by tin powder (325 mesh, 1.1907 g) , 9α-bromohydrocortisone 21-acetate [I, JACS 77, 4438 (1955), 2.4170 g] and AIBN (0.0828 g) . The resulting slurry is is heated at 80 ^* under an atmosphere of nitrogen until TLC analysis (acetone/methylene chloride/cyclohexane; 2/5/3) indicates that con- version to hydrocortisone 21-acetate is complete (4 hr) . The reac¬ tion mixture is diluted with acetic acid in 1 ml portions until all the precipitated solids dissolve (5 ml) . Next activated charcoal (0.204 g) is added and the mixture filtered thru celite. The

filtrate is then treated with a solution of potassium carbonate (89.1 mmoles) and potassium bicarbonate (49.9 mmoles) in water (42 ml). The slurry is then stirred at 0* for 1 hr and filtered. The cake is washed with water (2 x 20 ml) followed by methylene chloride' (10 ml). The solids are then air-dried to give hydrocortisone 21-acetate (II). EXAMPLE 3 Debro ination of 9α-Bromoprednisolone 21-Acetate (I) to Prednisolone 21-Acetate (II) Using Stannous 2- Ethylhexanoate A mixture of stannous 2-ethylhexanoate (4.0728 g) in THF (10 ml) is treated with dropwise with a mixture of 3-mercaptopropionIc acid (3.2043 g) in THF (10 ml). The resulting slurry is then treated with 9α-bromoprednisolone 21-acetate (I, 2.4083 g) followed by AIBN (0.1644 g) . The mixture is stirred vigourously at reflux under an atmosphere of argon for 100 min at which time conversion to pred- nisolone 21-acetate (II) is complete as measured by TLC.

EXAMPLE 4 Debromination of 9α-Bromoprednisolone 21-Acetate (I) to Prednisolone 21-Acetate (II) Using Tin Powder A mixture of potassium acetate (0.5903 g) In neat 3-mercapto- propionic acid (1.5921 g) is treated successively with 9α-bromo- prednisolone 21-acetate (I, 2.4080 g) , tin powder (325 mesh, 1.7828 g) , AIBN (0.1650 g) and isopropanol (5.0 ml). The resulting slurry is stirred vigorously at reflux under an argon atomosphere for 1.5 hr at which time conversion to prednisolone 21-acetate (II) is complete as measured by TLC. The reaction mixture is then poured into water (0°, 125 ml) using isopropanol (6.0 ml) for rinsing the reaction flask. After stirring for 1 hr at 0*, the solids which consist of a mixture of prednisolone 21-acetate and tin powder are filtered off and dried by a nitrogen stream. The prednisolone 21-acetate is purified as in EXAMPLE 1. EXAMPLE_5 Debromination of 9α-Bromoprednisolone 21-Acetate (I) to Prednisolone 21-Acetate (II) Using Lead Powder A mixture of lead powder (325 mesh, 0.2202 g) , 3-mercapto- propionic acid (0.5428 g) and 9α-bromoprednisolone 2i-acetate (I, 0.0825 g) in THF (3 ml) is refluxed under an atmosphere of nitrogen. After 38 hr conversion to prednisolone 21-acetate (II) is complete. The product is isolated and purified as in EXAMPLE 1. EXAMPLE 6 Deiodination of 3α-Iodo-5α-cholestane (XIX) to 5a- Cholestane (XX)

A mixture of 3α-iodo-5α-cholestane (XIX, PREPARATION 1, 0.4994 g), 3-mercaptopropionic acid (0.9697 g) and AIBN (0.0343 g) in THF (3.0 ml) is treated with stannous ethyleneglyoxide (0.5370 g) . The resulting slurry is stirred vigorously at reflux under an argon atmosphere for 15.5 hr. Additional stannous ethyleneglycoxide (0.3583 g), 3-mercaptopropionic acid (0.6435 g) and AIBN (0.0428 g) is added. The mixture is refluxed an additional 6 hr at which time the TLC indicates that conversion to 5α-cholestane is complete. The mixture is poured into aqueous hydrochloric acid (5%, 50 ml) and extracted with cyclohexane (2 x 20 ml) . The cyclohexane extracts are washed with a solution of sodium hydroxide (10%) and sodium sulfite (5%) in water and filtered thru celite. The cyclohexane phase is dried over magnesium sulfate and concentrated under reduced pressure to give 5 -cholestane which is identical with an authentic sample of 5α-cholestane by NMR, CMR and TLC (hexane, Rf - 0.85).

EXAMPLE 7 Deiodination of 1-Iodo-n-undecane (XXI) to n-Undecane (XXII) A mixture of 1-iodo-n-undecane (XXI, 1.403 g) , 3-mercaptopropio- nic acid (1.596 g) and AIBN (0.1641 g) in isopropanol (4.0 ml) is treated with tin powder (325 mesh, 1.7792 g). The resulting slurry is then stirred vigorously at reflux under an argon atmosphere for 4.5 hr. Additional tin powder (1.7597 g), 3-mercaptopropionic acid (1.596 g) and AIBN (0.1644 g) is then added and the resulting mixture is refluxed for another 1.5 hr at which time conversion to n-undecane (XXII) is complete. The mixture is poured into half saturated ammonium chloride (10 ml) and extracted with pentane. The pentane phase is washed with sodium thiosulfate (5%, 10 ml) followed by aqueous sodium bicarbonate (5%, 10 ml). The pentane phase is then washed with aqueous hydrochloric acid (10%, 10 ml), followed by a solution of sodium thiosulfate (5%) and sodium bicarbonate (5%) in water (10 ml), followed by water (10 ml). The pentane phase is dried over magnesium sulfate and concentrated under reduced pressure to give an oil and solids. The solids are filtered off and the filtrate is concentrated under reduced pressure to give an oil which is identical with an authentic sample of n-undecane by NMR, CMR and TLC (hexane, Rf - 0.75).

CHART A

20

25

CHART A - CoπfiτniP.ri

20

25

CHART B

CHART B - Continued

CHART B - Continued

X _r (CH ₂ ) _nι .CH ₃ H-(CH ₂ ) • CH«

(XXI) (XXII)

4

-24-

CHART C

Initiator Acceptable T Preferred T

Oxygen (0 ₂ ) 0 to 100 ^* 20 to 40*

Triethylborane optionally combine w/02 -78 to 20 ^* -78 to -20* Benzophenone and light of 250-350 mμ -78 to 80° 0 to 80*

Succinic acid peroxide 40 to 100° 60 to 90*

Dibenzoylperoxide 40 to 110* 80°

Lauroyl peroxide 40 to 90* 60 to 80°

Di-t-butylperoxyoxalate 0 to 60° 20° t-Butylperbenzoate 60 to 100° 80 to 100°

Di-t-butylperoxide 80 to 140° 110 to 140 ^c

Di-cumylperoxide 100 to 140° 120° t-Butylhydroperoxide and acetic acid 20 to 60° 40 to 60° t-Butylperbenzoate and cuprous bromide 60 to 100° 80° H2O2 and ferrous perchlorate 0 to 40° 20°

Potassium persulfate 40 to 80° 60°

Potassium nitroso disulfonate 40 to 80° 60°

AIBN 40 to 90* 60°

1,1' -Azobis(cyclohexanecarbonitrile) 40 to 90* 60° Di-(2-ethylhexyl)peroxydicarbonate 30 to 80* 50° t-Amyl peroxypivalate 30 to 80* 50° Methylethylketone peroxide combined with cobalt naphthenate or cobalt octoate 0 to 50* 20°