Note: Descriptions are shown in the official language in which they were submitted.
~ I3~8107
PROCESS OF PREPARATION OF NEW CATALYSTS
CONTAINING RHODIUM AND THEIR APPLICATION
The present invention refer~ to the compound~ hydrazine-NI:N2)-
bi~[bis(triphenylphosphine)-chlororhodium (I)] and di(~-hydrazine-N~:N2)-
bis[bis(triphenylphosphine)rhodium (I)]dichloride, which are h- -gen-ous
hydrogenation catalyst~ and their application in the hydrogenation of the
exocyclic methylene group of acid addition ~alts of 6 ~ hyl-6-deoxy-6-
methylene-5-hydroxytetracycline (methacycline) to prepare a-6-deoxy-5-
hydroxytetracycline (doxy~-ycline).
Doxy~ycline is a wide-spectrum antibacterial agent, with widespread
application the treatment of numerous infections in humans and in animals.
The hydrogenation of the exocyclic methylene group of methacycline can
produce two epimers. The a-6-epimer i8 doxycycline, whilst the B-6-epimer,
called 6-epi-doxycycline, is devoid of clinical utility. Thus, it is
important that the hydrogenation doe~ not co-produce this B-6-epimer. In
fact, the British phA -copoeia 1980 established a limit for the content of
6-epi-doxy~ycline in doxycycline of 2%.
In the prior art, doxy~ycline was first described in 1960 in Blackwood
U.S. Patent N 3,200,149 (1965). Since that time many methods have been
described for its preparation, in all of which modification of the
catalytic system has been described as producing ; p,oved yields or a purer
product. In the field of heterogeneous catalysis, U.S. Patent Ns
3,444,198 (Korst, 1969), 3,849,491 (Villax, 1974), 3,954,862 (Morris, Jr.,
1976) and 4,597,904 (Page, 1986) and the report in Chemical Abstracts 86,
89476 f (1977) of Bodnar et al. Hungarian Patent 12,042 (1976) have all
taught ; _oved methods for the preparation of doxy~ycline and its
analogues.
The first use of h- -ge-ecus catalysis was described in Broggi et al.
U.S. Patent N 4,207,258 (1980 based on Italian priority 1972), wherein the
catalyst was a complex of rhodium with tertiary phosphine, arsine and
stibine ligands. Cotti U.S. Patent N 3,962,331 (1976) extended the above
process to the simultaneous reductive dehalogenation and hydrogenation of
an lla-halomethacycline. Brennen et al French Patent N 2,216,268 (1978)
later disclosed the use of the same cataly~t.
Since that time, other patents have appeared such as U.S. Patent Ns
3,907,890 (Scanio, 1975), 4,001,321 (Faubl, 1977) and 3,962,131 (Faubl et
al., 1976) all describing variations in the catalytic syetem and claiming
; oved yields and stereospecificity.
The first h~ -ganeous hydrogenation catalysts of the type of tertiary
phosphine-hydrazine-rhodium complexes were described in Page et al. U.S.
Patent N 4,550,096 (1985). These were prepared either by reacting a
rhodium salt, specifically
i''''
I 3381 07
rhodium trlchloride, with a tertiary phosphine and a hydraz,ne~ or by
reacting a rhodium complex, such as tris(triphenylphosphine)chlororhodium,
with a hydrazine. These complexes allowed the preparation of doxycycline,
containing less than 1% of the undesired 6-epi-doxycycline, in high yield
using considerably less rhodium than had been taught in the prior art.
These complexes have proved to be very satisfactory catalysts for the
hydrogenation of methacycline especially if an excess of a tertiary phos-
phine is included in the hydrogenation mixture as a promoter.
The exact chemical formulae and structures of the catalysts of this
U.S. Patent were not disclosed in the patent, ~ut reported elemental
analyses showed some significant variability in elemental composition
indicating variations in constitution.
It has now been found that, by changing the process conditions used in
the U.S. Patent, very satisfactory new catalysts can be made which have well
defined structures. It is advantageous, from general considerations, to be
able to use catalysts of precisely known formula and structure, and further-
more the new catalysts are very effective for the hydrogenation of metha-
cycline without the need to add any excess tertiary phosphine.
According to the present invertion, there is provided a process for
the preparation of a complex of rhodium and hydrazine, containing triphenyl-
phosphine and chlorine, useful as a homogeneous hydrogenation catalyst,
which comprises reacting tris(triphenylphosphine)chlororhodium with
hydrazine or hydrazine hydrate in methanol under an inert atmosphere,
stirring the reaction mixture at room temperature, or refluxing it and then
recovering the solid complex from the mixture, characterised in that the
reaction is conducted in the absence of oxygen using degassed methanol, and
wherein either (a) a complex of formula I:-
PPh3 ~PPh3
Cl - Rh - NH2 - NH2 - Rh - Cl
PPh3 PPh3
( I )
wherein Ph his phenyl, is obtained when each mole of tris(triphenyl-
phosphine)chorJ~ ~ um ~ qeacted with at least one half of a mole of
~_ 1338107
-- 3 --
hydrazine with stlrring at room temperature until precipitation thereof from
~he mixture, or (b) a complex of formula II:-
C1 H2 H2
Ph3P \ ~ / N N \ ~ / PPh3
Rh Rh
Ph3P \ N ,~ / ~ PPh
H2 H2 Cl
( II )
wherein Ph is phenyl, is obtained when for each mole of tris(triphenyl-
phosphine)ch10rorhodium at least one mole of hydrazine is used, and the
reaction mixture is stirred at room temperature for a prolonged period or
refluxed, followed by standing at room temperature for at least 12 hours in
order to form crystals of the complex.
The invention includes the new catalyst compounds of formu1a I and
formula II per se, and also a process for the catalytic stereospecific hy-
drogenation of an acid addition salt of 6-demethyl-5-deoxy-6-methylene-5-
hydroxytetracycline to prepare -6-deoxy-5-hydroxytetracycline, wherein the
hydrogenation is carried out at a temperature between 60C and 100CC, at a
pressure of 1 to 10 kg/cm2 until the reaction is complete, followed by
isolation of the thus formed compound by known processes, characterised by
using a catalyst prepared by the process of the invention.
In the process of the invention for making the ne~ catalysts, the
tris(triphenylphos?hine)chlororhodium must be freshly prepared, and stored
and manipulated under an inert atmosphere. The preparation and isolation of
the complexes must be carried out under an inert atmosphere with complete
exclusion of air and in degassed reaction media, followed by drying under an
inert atmosphere or in vacuum. After eventual purification, the complexes
obtained are of a uniform composition and well defined formulae, being novel
compounds, never previously described.
According to the present invention, a catalyst of the formu~a I is
obtained by reacting, under an inert atmosphere at room temperature for up
to four hours, one mole of tris(triphenylphosphine)chlororhodium with at
least half a mole of hydrazine in degassed methanol, followed by isolation
1338107
and drying under an inert a; -2~h_re or vacuum.
The same reaction can be carried out, with the same precautions
as to the exclusion of air, at room temperature for a prolonged period or
at reflux, followed by st~n~ing at room temperature for at least 12 hours,
for example one to two day~, with at lea~t one mole of hydrazine, giving
crystals of the catalyst of the formula II.
Both fc l~e have been unequivocally established by elemental
analysis, as well as by infra-red and nuclear magnetic resonance
spectrQsccpy~ and in the seccnd case by X-ray cryatallography.
It i~ to be noted that both of these structures fall within the
general structure given in said U.S. Patent N 4,550,096.
The catalysts when prepared according to the conditions
de~cribed above, are fully active in the hyd~o~nation of methacycline to
doxycycline. Furthermore, it is not nece~sary to add excess
triphenylphosphine to ensure a near ~toichiometric yield of the required a-
epimer.
The complex bis(triphenylphosphine)hydrazinomethoxyrhodium
Ph3P ~~~~~ OCH3
~ Rh
Ph3P ~ ~ NH2NH2 ( III )
has been disclosed in Hovione Inter Ltd's European Patent Application N 85
305 045.8 (Publ. N 0 187 436, 1986). Thi~ compound is the result of a
side-reaction of the hydrazine and the
tris(triphenylpho~phine)chlo~o.hodium, with the methanol solvent taking
direct part in the reaction.
Thus, the preparation of any one of these three c~ ~_nds is
solely dependent on the physical parameter~ of the reaction. On thi~
basis, the _ _ -'~ of structure~ I and II are in equilibrium in the
reaction mixture obtained from the tris(triphenylphosphine)chlororhodium
and hydrazine. The precise proportion of each _ ~nd is substantially
controlled by the physical parameters existing in the mixture.
It is believed that the -~ ic structure:-
- S - 1338107
Ph3P \ / Cl
Rh
Ph 3 p \ NH2NH2
( IV )
is also in equilibrium in the reaction mixture, although this has not been
isolated so far.
The conditions of preparation of the catalysts of the present inven-
tion are illustrated in ~xamples 1 and 2. The tris(triphenylphos-
1~ phine)chlororhodium and hydrazine can be reacted in the molecular proportioncorresponding to their respective formulae, but it is advantageous to use
hydrazine in excess so as to obtain the maximum yield in relation to the
expensive rhodium complex.
The hydrazine can be used as either the anhydrous base or as the mono-
hydrate. It has been verified that the anhydrous base allows shorter re-
ac~ion times.
To achieve the best results in prepa-ing the compound of formula I,
tris(triphenylphosphine)chlororhodium (1 mole) and hydrazine hydrate (3
moles) are mixed in degassed methanol under a nitrogen atmosphere. After
stirring for a few hours, the yellow solid precipitates and is filtered and
dried under vacuum.
When tris(triphenylphosphine)chlororhodium (1 mole) and hydrazine
hydrate (3 moles) are refluxed in degassed methanol under a nitrogen atmos-
phere, in adequate equipment, followed by standing at room temperature,
filtration and drying under vacuum, yellow crystals of the compound of
formula II are formed, In contrast, cooling, preferably after concentra-
tion, favours the isolation of a yellow solid of formula III.
The complexes of formulae I and II are stable for at least one month,
providing they are stored under nitrogen at reduced temperatures. After
this period, slightly diminished catalytic activity is sometimes observed.
Therefore, these complexes should be in preference freshly prepared to ob-
tain the best hydrogenation results. Al~ernatively, they can be prepared
immedlate,ly prior to use and then employe~ without isolation by addition to
roq ~ 0~
the h~rogcffltion reaction mixture, whereby equally good results can be
achieved.
As already indicated, the hydrazino-rhodium complexes of the present
S h om~gen~o~
nventlon are efficient homog~,eous stereos?ecific hydrogenation catalysts,
1338107
in general. The present invention, however, has specifically been directed
to their application in the hydrogenation of the exocyclic methylene group
of 6-demethyl 6 deo~y-6-methylene-s-hydLu~y~etracycline present in the
hydrogenation reaction mixture as an acid addition salt, so as to yield a-
6-deoxy-5-hydroxytetracycline in a near stoichiometric yield.
The starting methacycline can be prepared by any of the known
processes, such as that described in U.S. Patent N 3,849,491, but should
not contain impurities which may act as a catalyst inhibitor.
Although the new complexes will catalyse the hydrogenation of
metha-cycline base, the rate is 80 slow that the time of hydrogenation does
not permit the yields obtained when using an acid addition salt.
The rate of hydrogenation increases with the temperature.
Temperatures from 60C to 100C can be used, but to achieve the best yields
and stereospecificity, the optimum reaction temperature range should be
between 85C and about 90C. At 95C the yields are slightly lower than
for instance at 88C. Below 85C, the catalytic system starts to be
sensitive to the eventual presence of certain trace impurities which may
interfere with the rate of hydrogenation.
In the context of the hydrogenation of methacycline acid
addition salts for the preparation of doxycycline, the present invention
has several advantages when the temperature range during hydrogenation is
85C to about 90C.
First, there is no necessity for extremely high hydrogen
pressures. It has been found that from 1 kgtcm2 to 10 kg/cm2 will ensure
complete conve,sion of the methacycline substrate in between 6 to about 10
hours. Typically, the hydrogenation is carried out at 88-89C at a
hydrogen pressure of 7 to 9 kg/cm2 and is complete after 6 1/2 to 7 hours.
Second, the amount of rhodium necessary to obtain complete
conversion is of 'he order of 0.003 part by weight of rhodium in relation
to the methacycline acid addition salt substrate.
The painstaking preparation of the catalysts under strictly
inert conditions can be alleviated by their preparation in degassed
methanol under a nitrogen atmosphere immediately prior to use, followed by
addition to the hydrogenation reaction mixture, after which the actual
hydrogenation is carried out.
The transformation of the methycycline acid addition salt into
doxycycline using the catalysts of the present invention gives a purity above
~r
13~81o7
- 7 -
95% in the reaction mixture, as analysed by high performance liquid
chromatGgLaphy (h.p.l.c.).
In contrast to the catalysts of U.S. Patent N 4,550,096, the
cat-alysts of the prasent invention when prepared, dried, and stored under a
strictly inert al -_~hera, exert full activity without the necessity of
adding an excess of tertiary phosphine, more specifically
triphenylphosphine, to the hydrogenation mixture so as to achieve the best
yields .
An explar.ation for thi~ is tha. the catalysts prepared
according to the proce~s of U.S. Patent N 4,550,096 were believe to be
stable and, in fact, they exert a very high catalytic activity, even when
stored for long periods, becau~e they were subsequently employed in
presence of a controlled excess of tertiary phosphine. It is now believed
that the catalysts prepared according to the process described in U.S.
Patent 4,550,096 oxidise slowly, but the presence of the excess tertiary
phosphine in the hydrogenation reaction mixture allowed substitution of the
oYi~;~ed part of the tertiary phosphine, thereby regenerating the original
catalytic eystem.
As has been previously mentioned, the catalyst is most
conveniently prepared i -~ia~tely prior to use. Thus, hydrazine hydrate
(0.5 to 4 mole~) is added with stirring to
tris(triphenylphosphine)chlo,orllodium (1 mole) in degassed methanol in a
glass vessel, under a nitrogen at ~~phare. Upon addition of the hydrazine,
the initial red colour turns to yellow. The reaction mixture is stirred
for between a few minutes and two hours, and then transferred to the
hydrogenator containing the methacycline acid addition salt in methanol at
50C, under nitrogen.
Subsequently, the reaction vessel is purged again with
nitrogen, then with hydrogen, finally being pressurised to 8 kg/cm2 with
hydLogen. The reaction mixture is heated to 88C under stirring, and the
tF -_ature maintained at 88C ~ 2C until the velocity of consumption of
hydrogen slow~ down dra~tically, which occurs after about 6 to 7 hours. At
this time, the reaction mixture contains nearly exclusively a-6-deoxy-5-
I.yd-o~Letracycline.
The purity of the reaction mixture thus obtained is such that
the do~y~ycline can be directly crystallised from the reaction mixture by
adding exces~ p-toluenesulphonic acid, followed by cooling, yielding
do~y~ycline p-toluenesulphonate with a purity about 99%.
The new catalyst~ have been shown to be effectively superior to
the W~ n~~- catalytic system.
1338l 07
-- 8
So as to elucidate the behaviour of the triphenylphosphine-
hydrazino-chlororhodium catalysts, as well as the actual role of the
hydrazine present in the complex, an extensive study on this catalytic
system was carried out.
According to U.S. Patent No. 4,550,096, triphenylphosphine-
hydrazino-chlororhodium catalysts can be prepared "in situ" by the addition
of rhodium trichloride, triphenylphosphine and hydrazine hydrate to the
cold hydrogenation reaction mixture, followed by heating, after which the
actual hydrogenation is carried out.
Therefore, a 6eries of parallel hydrogenations was carried out,
using methacycline (MOT) hydrochloride in presence of the "in situ"
prepared catalyst, in the molar proportion of one mole of rhodium
trichloride trihydrate and one mole of triphenylphosphine, without addition
of hydrazine hydrate, and with the addition of one and two moles of
hydrazine hydrate. Whilst in the absence of hydrazine hydrate, only 60.87%
of the a-epimer was formed, together with 12.44% of ~-epimer and 14.95% of
degradation products, it was verified that in the presence of hydrazine
hydrate, the formation of the a-epimer increased drastically:
RhCl3.3H20 Ph3PNH2NH2-H2O a-epimer % ~-epimer ~o MOT % D~.~ldtio~
products 70
Expt. 1 1 mole1 mole 0 6087 12.44 1174 1495
Expt.21 mole 1 mole1 mole 82.86593 575 546
Expt. 3 1 mole1 mole 2 moles 92.91 2.2 456 033
As can be ascertained from these values, the
triphenylphosphine-hydrazino-chlororhodlum catalytic system contains an
active species of only one mole of triphenylphosphine and two moles of
hydrazine for each mole-atom of rhodium, which is in contrast with the
30 Wil ki n~on catalytic system, which must contain at least two moles of
triphenylphosphine for each mole-atom of rhodium in order to be efficient.
, ,
-
- 9 - 13381 07
In parallel hydrogenation experiment, with the rnodium-hydrazine cata-
lytic system and the well-known Wilkinson catalyst, using deuterium instead
of hydrogen, the products were analysed by mass spectrometry. It has been
found that the doxycycline obtained ~ith the hydrazine-containing system
contains significantly less deuterium than the doxycycline obtained using
the Wilkinson catalyst. This difference indic~tes that the hydrazine takes
an active part in the catalytic hydrogenation.
These resul's show the positive e~fect of the hydrazine ligand in the
catalytic system of the present invention.
The following examples serve to illustrate the present invention,
without in any way limiting the scope thereof.
EXAMPLES
1) Preparation of (~-hydrazine-N1:N2)-~is~bis(triphenylphosphine)-chloro-
rhodium (I)~
Tris(triphenylphosphine)chlororhodium (Q.50 9; 0.54 mmoles) was placed
in a two necked round bottom flas~. The solid was s~irred under vacuum for
20 30 minutes and tnen under an atmosphere of nitrogen. Dry, degassed methanol
(50 ml) was added and the mixture was stirred for 15 minutes. A methanolic
solution of hydrazine hydrate (15 ml of a methanolic solution of hydra~ine
hydrate containing 5.91 mg/ml; 1~77 mmoles) was added, and the mixture
stirred for 3 hours at room temperature. A yellow precipitate formed, which
25 was filtered off and dried under vacuum.
The proton nmr spectrum showed a complex signal centred on ~ 7~5~
(phenyl ring protons) and a broad peak at ~ 1.8 (hydrazine protons). The
infra-red spectrum shows a doublet at 3130 cm 1 (N - H stretching), a band
at 305 cm~1 (Rh - Cl stretching), as well as bands indicative of triphenyl-
30 phosphine.
Elemental analysis: C: 63~31~o H: 4.81% N: 2~73% P 9~15%
C72H64Cl2N2P4Rh2 requires: C: 63~68% H: 4~75% N: 2~06~ P 9.1~do
35 A repeat preparation using hydrazine hydrate (1 ml; 20~57 mmoles) gave the
same product after stirring at room temperature for 1 hour.
1338107
-- 10 --
2) P~epa,alion of di(~-hydrazine-N':N2)-bis rbis (triphenylphosphine)rhodium(I)~dichloride
Tris(l~ henylphosphine)chlororhodium (1.05 g; 1.13 mmoles) was placed in a tw
s necked round bottom flask. The solid was stirred under vacuum for 30 minutes and then
under an atmosphere of nillogen. Dry, deg~sed methanol (170 ml) was added and the
IlliAIUl~ was stirred for 15 minutes. A methanolic solution of hydrazine hydrate (30 ml of
a lu~ olic solution co~ inil-g 5.91 mg/ml; 3.54 mmoles) was added. The reaction
IniA~ure was refluxed for 2 hours. Upon st~n-ling overnight, yellow crystals were
10 deposited, which were filtered and dried under vacuum.
A single crystal of apploAimate ~imPnSions 0.3 mm x 0.15 mm x 0.1 mm was
sealed under argon in a thin walled glass capillary. Unit cell and intensity data were
obtained using a four circle X-ray diffractometer, following standard procedures. Details
of the e,.~lill,enlal fe&lul~s are as follows:-
Crystal data: [C72H68N4P4Rh2].[Cl]2.CH3OH, Mw = 1422.02, monoclinic, space groupP2l/n, a = 15.009(3)A, b = 13.294(2)A, c = 18.391(4)A"B = 108.9(1), V = 3471.9
A3, z = 2, Dc = 1.36 g.cm~3, ~(Mo - K~x) = 6.14 cm~l.
Data collection: Data were recovered for 1.5 < ~ < 21 at room te---~l~ture, 291K
and coll~;led for absorption empirically. 3716 intensities were measured, of which 1805
20 were considered observed [I ~ 1.5 ~(I)] and used in the analysis."
The structure was solved via the heavy atom method and refined by full matrix least
squa~es. In view of the sm.~Jl number of observeid data, the structure was refined in the
anisotropic approximation, but with the phenyl groups defined as rigid bodies. No
hydrogens were confidently located on the hydrazine nitrogen atoms and none were25 included. The final R value is 0.06.
The complex was found to contain a dimeric cation in which two (Ph3P)2Rh units
were linked together by two bridging hydrazine molecules, as shown in II:-
1338107
_ 2+
,H2 H2
Ph~P \ / N N \ / PPh3
Rh Rh
Ph3P \ N - N/ \ PPh3
H2 H2
The rhodium centres have the expected s~uare planar configuration, with the
Rh-P and Rh-N dis~ances being normal. The central Rh2N4 ring has a chalr
conforma+ion, compatible with its centrosymmetric nature.
3. Comparative hydrogenations in presence and absence of hydrazine
A. Molar ratio - Rh:PPh3 = 1:1
Methacycline hydrochloride (4.0 g; 8.35 mmoles) was suspended in
methanol (60 m1) and rhodium trich7Oride trihydrate (200 mg; 0.75 mmoles)
and triphenylphosphine (196.5 mg; 0.75 mmoles) added. The mix+!~re was then
hydrogenated for 6 hours at a hydrogen pressure of 8 kg/cm2 and a tempera-
ture of 80C in a convent~onal stainless steel high pressure reactor. At
the end of the reaction, a sample of the reaction mixture was analysed by
h.p.l.c. and gave:-
a-epimer = 60.87%
~-epimer = 12.44%
methacycline = 11.74%
others = 14.95
B. Molar ratio - Rh:PPh3:NH~NH2 = 1:1:1
The experiment as described in A above was repeated but wi'h the ad-
dition of 0.95 ml of a 0.814 M solution of hydrazine hydrate in methanol
(0.77 mmoles NH2NH2). H.p.l.c. of the crude reaction mixture ~ave:-
a-epimer = 82.86~
~-epimer = 5.93%
methacycline = 5.75%
others = 5.46%
~~ - 12 - 1 3381 0 7
C. Molar ratio - Rh:PPh3:NH2NH2 = 1:1:2
The experiment as described in A above was repeated but with the ad-
dition of 1.9 ml of 0.814 M solution of~hydrazine hydrate in methanol (1.55
mmoles NH2NH2). H.p.l.c. of the crude reaction mixture gave:-
a-epimer = 92.91%
Q-epimer = 2.2 %
methacycline = 4.56%
others = 0.33%
4. Hydrogenation of methacycline using a non-isolated catalyst
Tris(triphenylphos?hine)chlororhodium (15.0 mg; 0.016 mmoles) and a
methanolic solution of hydrazine (124 ~l of a 0.4M solution; 0.050 mmoles)
were added to methanol (60 ml) in a nitrogen atmosphere with stirring.
lS Methacycline hydrochloride (7.38 g; 15.41 mmoles) was added, and the mixturetransferred to a conventional stainless steel high pressure reactor, which
was charged to a pressure of 8 kg/cm2 with hydrogen, and then reacted at
88C for 6 1/2 hours.
p-Toluenesulphonic acid (3.3 g) was then added and the mixture stirred
at 0C for 2 hours. The doxycycline p-toluenesulphonate obtained by fil-
tration and drying at 35C weighed 8.84 9 and had a puri~y of 98.9%.
5. Hydrogenation of methacycline using (~-hydrazine-~l:NZ)-bis[bis
(triphenylphosphine)chlororhodium (I)]
(~-Hydrazine-Nl:N2)-bis[bis(triphenylphosphine)chlororhodium (I)] (11
mg; 0.0081 mmoles; 0.0162 mmoles of rhodium), as prepared in Example 1, was
added to a suspension of methacycline hydrochloride (7.25 g; 15.14 mmoles)
in methanol (59 ml) and the mixture hydrogenated at a pressure of 8 kg/cm2
for 6 1/2 hours at 88aC. Thereafter, p-toluenesulphonic acid (3.3 g) was
added and the mixture stirred at 0C for several hours. The yield of doxy-
cycline p-toluenesulphonate was 8.45 g, 90.5%, which had a purity of 99.0q.
