Iodination and metal halogen exchange of aromatic compounds: An improved preparation of a key oxazolidinone antibiotic intermediate

Article abstract of DOI:10.1021/op0002949

A new method for the preparation of a key intermediate for oxazolidinone antibiotic analogues is discussed. We have found that iodine-magnesium exchange can be used to prepare a dianion if a fluorine is present ortho to the aryl anion and that exchange ta

Full text of DOI:10.1021/op0002949

Organic Process Research & Development 2001, 5, 80−83
Iodination and Metal Halogen Exchange of Aromatic Compounds: An Improved
Preparation of a Key Oxazolidinone Antibiotic Intermediate
Paul M. Herrinton,* Carolyn E. Owen, and James R. Gage
Pharmacia Corporation, Chemical Process Research and Preparations, Kalamazoo, Michigan 49001, U.S.A.
Abstract:
mediate, 5.3 This intermediate can then be converted to
oxazolidinones such as 2 using chloropropanediol and
alkoxide. This procedure was dilute and long, requiring the
in situ preparation of anhydrous magnesium salts. Therefore,
we sought to improve this conversion by a direct magnesium-
halogen exchange reaction on the aryl iodide. Direct mag-
nesium-halogen exchange on the bromide could not be
effected, however we felt that the aryl iodide could possibly
undergo exchange.
A new method for the preparation of a key intermediate for
oxazolidinone antibiotic analogues is discussed. We have found
that iodine-magnesium exchange can be used to prepare a
dianion if a fluorine is present ortho to the aryl anion and that
exchange takes place on iodoanisoles. In addition, we report
an improved method for iodination of aromatics using N-
chlorosuccinimide/hydriodic acid and potassium iodide. By
using these conditions, reactions are faster and higher-yielding
than using either N-iodosuccinimide or iodine monochloride.
Linezolid (1), the first representative of a new class of
oxazolidinone antibacterials, is indicated as a hospital-use
antibiotic for treatment of serious gram-positive pathogens
(Staphylococci, Streptococci, and Entercocci) implicated in
skin and soft tissue infections, pneumonias, and bacteremias.1
Oxazolidinones containing carbon-aryl bonds, such as
compound 2, in place of the nitrogen-aryl bond of linezolid
are also active as antibacterials.2 To examine the properties
of these products it is important to develop high-yielding
methods for their preparation. One of the critical synthetic
challenges involved in approaching these compounds is the
formation of the carbon-carbon bond linking the fluorinated
aromatic ring with the sulfur heterocycle.
Acylation of 3-fluoroaniline with isobutylchloroformate
under Schotten-Baumann conditions proceeded smoothly
to provide the carbamate 3 in 96% yield. Preparation of the
aryl iodide was achieved by electrophilic iodination of the
isobutyl carbamate.4 Reaction of the crystalline carbamate
with N-iodosuccinimide was slow and required more than
48 h to reach completion giving an 86:14 mixture of the
ortho and para-iodo products, 7 and 8.5 Crystallization of
the product from ethyl acetate or heptane provided only the
desired ortho isomer 7 in 57% isolated yield. If the crude
methylene chloride solution from acylation was used without
isolation, the iodination reaction was complete in less than
24 h. We suspected that carry-over of some inorganic
material was responsible for this and found that addition of
1-10 mol % of chloride ion accelerates the reaction when
using isolated carbamate. The ratio of ortho to para reaction
products was not affected by this catalysis, but the reaction
was complete in less than 6 h at 25 °C to give the desired
ortho product in 72% isolated yield. The catalytic effect is
In a previous report, lithium-bromine exchange on the
carbamate, 4, followed by addition of magnesium salts
provided an aryl-magnesium species, which was reacted
with tetrahydrothiopyran-4-one to provide the desired inter-
(3) Gage, J. R.; Perrault, W. R.; Poel, T. J.; Thomas, R. C. Tetrahedron Lett.
2000, 41, 4301-4305.
(1) Brickner, S. J.; Hutchinson. D. K.; Barbachyn, M. R.; Manninen, P. R.;
Ulanowicz, D. A.; Garmon, S. A.; Grega, K. C.; Hendges, S. K.; Toops,
D. S.; Ford, C. W.; Zurenko, G. E. J. Med. Chem. 1996, 39, 673-679.
(2) Poel, TJ.; Martin, J. P., Jr.; Barbachyn, M. R. PCT Int. Appl. WO Patent
9929688, 1999; Chem. Abstr. 1999, 131, 31934.
(4) For a review of iodination of aromatics, see: Merkushev, E. B. Synthesis
1988, 923.
(5) Olah, G. A.; Wang, Q.; Sandford, G.; Prakash, G. K. S. J. Org. Chem.
1993 58, 3194; Orazi, O. O.; Corral, R. A.; Bertarello, H. E. J. Org. Chem.
1965 30, 1101.
80
•
Vol. 5, No. 1, 2001 / Organic Process Research & Development
10.1021/op0002949 CCC: $20.00 © 2001 American Chemical Society and The Royal Society of Chemistry
Published on Web 11/23/2000

substantial early in the reaction, but diminishes as the reaction
approaches completion, suggesting the presence of a catalytic
species that is consumed or poisoned during the course of
the reaction. The counterion for the added halogen has a
limited effect. The group 1A metals are indistinguishable
from each other, but tetrabutylammonium and ammonium
salts accelerate the reaction to a much greater extent and
lead to over-reaction products in some cases. This may be
due to a phase-transfer effect in the methylene chloride/water
solvent system. When bromide ion was added, the bromide
product was formed in an amount corresponding to the
amount of bromide catalyst added, suggesting that a mixed
halogen species such as ICl or IBr was being generated in
situ.
Adapting the chemistry of Knochel et. al.,8 metalation of
7 using iso-propylmagnesium chloride was explored as an
alternate means for generating the aryl magnesium.
We found that metalation of 7 could be effected at -40
°C in 2-3 h with 2.1 equiv of iso-propylmagnesium chloride
in THF. Formation of the carbamate anion was fast at this
temperature, with iodine-magnesium exchange being the
rate-limiting reaction. The exchange was much faster (<1
h) at -25 °C. We observed no formation of benzyne-derived
products when the reaction temperature was kept below -15
°C.
If we were indeed generating mixed halogens, we
theorized we would be able to iodinate with N-chlorosuc-
cinimide (NCS) and iodide. The reaction of 3 with N-
chlorosuccinimide and potassium iodide provided only the
iodide in a 56% yield.6 Using hydriodic acid as the iodide
source, 7 was obtained in a 78% isolated yield after overnight
at room temperature. This result was comparable to the use
of N-iodosuccinimide and potassium chloride. The optimized
conditions were found to be 1.5 equiv of N-chlorosuccin-
imide, 0.5 equiv of HI (57% aqueous), and 1.0 equiv of KI
in methylene chloride and water at 20-35 °C. Under these
conditions we were able to convert 6 to 7 in an overall yield
of 79%. Due to the lower cost of positive chlorine (NCS)
species versus positive iodine (NIS, ICl) species, this
iodination procedure represents a less expensive way to
introduce iodine into aromatic nuclei. Attempts to replace
the NCS in this system with either NaClO2 or chloramine-B
hydrate were unsuccessful.
We applied the NCS/HI/KI iodination conditions to
several other substrates, and we found that these conditions
are general for electron-rich substrates, but that electron-
deficient rings do not react. Anisole, 9, p-xylene, 10, and
the des-fluoro carbamate, 11, were reacted under these
conditions to provide the monoiodination products, 12, 13,
and 14 in 72-94% yield. Reactions on these substrates under
the NCS/KI/HI conditions were complete 2-3 times faster
than reactions on the same substrates with N-iodosuccinimide
without catalyst. We also found that 3-fluoroaniline did not
give any of the desired iodinated products under these
reaction conditions. This was rather unexpected, as similar
methods are reported to work for anilines.7
Addition of a THF solution of the tetrahydrothiopyran-
4-one to the dianion resulted in a rapid addition reaction to
provide the desired tertiary alcohol, 5 in 63% isolated yield.
The major byproducts are 3, which results from proton
transfer from the ketone, and the aldol addition product of
the resulting ketone enolate with another molecule of ketone.
Use of additives and cosolvents were explored to help
minimize the proton transfer. Addition of copper(I) iodide
or cerium(III) chloride resulted in poor yields while adding
3 mol equiv of TMEDA increased the yields to 80-90%.
Use of more TMEDA did not show a continued increase in
yield. Use of less than 3 equiv of TMEDA gave yields in
the 65-70% range. We conducted experiments in which the
TMEDA was added before transmetalation, after transmeta-
lation but before thiopyranone addition, and without TME-
DA. There were no significant differences in the yield of
the first two experiments, indicating that the TMEDA does
not likely play a role in the iodine-magnesium transfer.
However, the experiment without TMEDA gave a lower
isolated yield (63% vs 84%), indicating that TMEDA has a
role in the addition reaction. Examination of the reaction
byproducts of the experiments shows that less proton transfer
takes place when TMEDA is present.
We have found that the addition is very rapid and that
warming the reaction before quenching is detrimental to the
yield and purity. We prefer to quench the addition reaction
within 30 min of the thiopyranone addition at -25 °C by
adding acetic acid and then warming the reaction mixture.
Warming the reaction mixture before quenching promotes
the aldol condensation of the excess tetrahydrothiopyran-4-
one and reduces the product purity.
(6) This reaction is known for radio labeling using I-125; (a) Wilson, A. A.;
Dannals, R. F.; Ravert, H. T.; Burns, H. D.; Wagner, H. N., Jr. J. Labeled
Comput. Radiopharm. 1986, 23, 83 (b) Youfeng, H.; Coenen, H. H.;
Petzold, G.; Stocklin, G. J. Labeled Comput. Radiopharm. 1982, 19, 807;
(c) Ponchant, M.; Koscielniak, T.; Hamon, M.; Gozlan, H. J. Labeled
Comput. Radiopharm. 1991, 29, 1147.
The tertiary alcohol product can be transformed into a
variety of oxazolidinones using the methodology described
in the previous report.3
(8) Roymond, L.; Rottlander, M.; Cahiez, G.; Knochel, P. Angew. Chem., Int.
Ed. 1998, 37, 1701-1703.
(7) Edgar, K. J.; Falling, S. N. J. Org. Chem. 1990, 55, 5287-5291.
Vol. 5, No. 1, 2001 / Organic Process Research & Development
•
81

The direct metalation of 7 was surprising because the
previously cited examples were all activated aromatics.8 The
transmetalation chemistry was also examined with p-io-
doanisole, 12, iodoxylene 13, and the des-fluoro compound
14. The resulting anions were trapped with either benzalde-
hyde or tetrahydothiopyran-4-one.
7.2 mL, 111 mmol) was added and the mixture stirred at
30-35 °C for 15 min. The mixture was cooled to 25 °C and
the pH adjusted from 8.7 to 1.9 with concentrated hydro-
chloric acid. The phases were separated, and the aqueous
was washed with methylene chloride (100 mL). The com-
bined organics were washed with water (200 mL), and the
water wash was back-extracted with methylene chloride (100
mL). [If desired, crystallization at this point from heptane
at -30 °C afforded 3 in 98.1% yield.] To a 1 L three-neck
round-bottom flask was charged potassium iodide (52.3 g,
315 mmol), N-chlorosuccinimide (102.2 g, 765 mmol), and
450 mL of water. The mixture was heated to 32 °C. A light
green suspension resulted. Hydriodic acid (47%, 29.7 mL,
225 mmol, 0.5 equiv) was added, which turned the mixture
purple, and the slurry nearly went into solution. A 5 °C
exotherm was observed. The solution of 3 in methylene
chloride was added and the mixture stirred at 32 °C until
the reaction was complete by HPLC. Reaction was complete
after 3 h. The solution turned from a dark purple to a light
orange throughout the reaction.
The mixture was cooled to 23 °C and quenched by adding
a solution of 57.0 g of sodium sulfite in 340 mL of water
over 5 min, keeping the temperature less than 33 °C. The
solution immediately turned to a dark purple at the start of
the quench, then to a light yellow by the end of the quench.
The layers were separated, and the upper aqueous layer was
extracted with 120 mL of methylene chloride. The combined
organics were washed with 120 mL of water. The layers were
separated, and the aqueous was washed with 50 mL of
methylene chloride. The combined organics were concen-
trated to a 250 mL volume (solids present) on the rotovap,
and 500 mL of octane was added (more crystals formed).
The slurry was concentrated to 350 mL on the rorovap. The
slurry was very thick. Octane (200 mL) was added to the
slurry and the mixture cooled to 0 °C. The slurry was filtered
on a 600 mL sintered glass funnel and washed with 2 ×
100 mL of octane pre-cooled to less than 10 °C. The filtration
rate was very fast. The solids were dried in a vacuum oven
overnight at 55 °C. This yielded 109.5 g of 7, as off-white
fluffy solids.
In the case of 12, the reaction proceeded using either
carbonyl compound; however, the temperature had to be
raised to greater than 0 °C for iodine-magnesium exchange
to occur. The isolated yields were lower in these reactions
than in the case of 7. This was attributable, in part, to
incomplete iodine-magnesium exchange on a less activated
ring system and in part to an increased level of side products
that were formed at higher temperatures. Attempts to react
compounds 13 and 14 were unsuccessful. Little iodine
exchange was observed, and quenching with benzaldehyde
did not give any of the desired secondary alcohol adduct.
In summary, we have found that iodine-magnesium
exchange can take place on aromatics without strong
activating groups and can be used to prepare a dianion. In
addition we report an improved method for iodination of
aromatics using NCS/HI/KI.
Experimental Section
General Procedures. All reagents were commercially
obtained and used as received unless otherwise noted. All
non-aqueous reactions were performed in dry glassware
under an atmosphere of dry nitrogen. Air- and moisture-
sensitive liquids or solutions were transferred via syringe or
polypropylene cannula. Organic solutions were concentrated
by rotary evaporation at ∼80 mmHg and less than 60 °C
except where noted. Chromatographic purification of prod-
ucts was accomplished using forced-flow chromatography
on EM Silica Gel 60. Thin-layer chromatography was
performed on Analtech Chromatography Products Uniplate
Silica Gel GF 0.25 mm plates. Visualization of the developed
chromatogram was performed by fluorescence quenching or
phosphomolybdic acid (PMA) stain or 50% sulfuric acid
char.
NMR spectra were measured on a Bruker AM-400
operated at 400 and 100 MHz, for 1H and 13C, respectively,
with data reported as follows: chemical shift (ppm),
multiplicity (s, singlet; d, doublet; t, triplet; q, quartet; m,
3: mp 62-63 °C; TLC Rf ) 0.34 (5% EtOAc/hexanes);
1H NMR (CDCl3, 300 MHz) δ 7.33 (d, 1H, J ) 11.0), 7.22
(q, 1H, J ) 8.4), 7.03 (d, 1H, J ) 8.1), 6.86 (bs, 1H), 6.74
(td, 1H, J ) 2.3, 8.3), 3.95 (d, 2H, J ) 6.7), 1.97 (nonet,
1H, J ) 6.7), 0.96 (d, 6H, J ) 6.7); 13C NMR (CDCl3, 75
MHz) δ 163.21 (ds, JC-F ) 244.6), 153.58 (s), 139.71 (ds,
JC-F ) 10.6), 130.10 (dd, JC-F ) 9.8), 113.90 (d), 109.96
(dd, JC-F ) 21.1), 106.05 (dd, JC-F ) 27.9), 71.59 (t), 27.95
(d), 19.03 (q); MS (CI, NH3) m/z (relative intensity) 230 (10),
229 (100), 213 (2.7), 212 (25), 211 (7.7). Anal. Calcd for
C11H14FNO2: C, 62.55; H, 6.68; N, 6.63. Found: C, 62.65;
H, 6.76; N, 6.67.
1
multiplet), integration and coupling constant (J, Hz). H-
13C Multiplicities are reported on the basis of DEPT data.
Where 13C-19F coupling occurs, the coupling constant (JC-F
)
is reported. Elemental analyses were obtained from Phar-
macia and Upjohn Physical and Analytical Chemistry. HPLC
analyses were carried out on a Dionex DX 500 Chromatog-
raphy System. GC analyses were carried out with a Hewlett-
Packard 5890a gas chromatograph.
Preparation of 7 from 6. To a solution of 3-fluoroaniline,
6, (50.0 g, 450 mmol) in methylene chloride (200 mL) was
added a solution of potassium carbonate (46.9 g, 339 mmol)
in water (200 mL) at room temperature. The mixture was
warmed to 32 °C, and isobutyl chloroformate (66.2 g, 485
mmol) was added over 13 min while maintaining 30-35
°C. The mixture was stirred at 30-35 °C for 2.5 h until
complete by GC analysis. Aqueous ammonia (29.3 wt %,
7: 1H NMR (CDCl3, 400 MHz) δ 7.41 (t, 2H, J ) 8.4),
6.96 (dd, 1H, J ) 2.0, 8.4), 6.87 (bs, 1H), 3.96 (d, 2H, J )
6.4), 1.97 (nonet, 1H, J ) 13.5), 0.96 (d, 6H, J ) 6.6); 13
C
NMR (CDCl3, 75 MHz) δ 159.24 (ds, JC-F ) 245.5), 153.39
(s), 138.98 (ds, JC-F ) 10.1), 133.33 (d), 115.10 (d), 107.08
(dd, JC-F ) 28.2), 102.03 (ds, JC-F ) 21.1), 71.78 (t), 27.92
82
•
Vol. 5, No. 1, 2001 / Organic Process Research & Development

(d), 19.01 (q). Anal. Calcd for C11H13FINO2: C, 39.16; H,
3.90; N, 4.15; I, 37.64. Found: C, 39.25; H, 3.86; N, 4.17;
I, 37.64.
product oiled out). The mixture was cooled to 23 °C (solids
formed on cooling) and the solids filtered on a 60 mL coarse
sintered glass funnel using house vacuum. The filtration rate
was very fast. The solids were washed with 2 × 25 mL
octane and pulled dry with vacuum. The product was then
dried in the vacuum oven overnight at 40 °C. This provided
9.4 g of 5 as off white solids, a 96% yield. The solids
contained about 10% of the thiopyranone dimer for a
corrected yield of 86%. Recrystallization of the material from
ethyl acetate provided an analytically pure sample.
Preparation of 5 from 7. To a 250 mL three-necked
round-bottom flask equipped with a mechanical stirrer,
thermocouple, and nitrogen purge was added 7 (10.0 g, 29.6
mmol), THF (60 mL), and TMEDA (10.3 g/13.4 mL, 88.9
mmol). The resulting solution was cooled to -25 °C, and
isopropylmagnesium chloride (31.4 mL, 2 M, 62.8 mmol)
was added via add funnel over 15 min, keeping the
temperature less than -25 °C. The solution turned from a
light yellow to a light brown color during the addition. The
solution was stirred at -25 °C for 20 min and then sampled
for HPLC. HPLC showed complete reaction of the idodide.
Tetrahydrothiopyran-4-one (4.8 g, 41.5 mmol) was dissolved
in 12 mL of THF and added dropwise to the reaction mixture
over 10 min, keeping the temperature less than -20 °C. A
7 °C exotherm was observed during the addition. The mixture
was stirred at -20 °C for 1 h. The reaction was quenched
by addition of the reaction mixture all at once into a mixture
of 13 mL of glacial acetic acid and 66 mL of water. The
quench was started at 0 °C, and the temperature rose to 27
°C. MTBE (25 mL) was added and the mixture stirred for
20 min. The layers were separated, and the lower aqueous
layer was washed with 25 mL of MTBE. The separation was
not very clear; thus, 15 mL of octane was added which gave
a clear phase break. The combined organics were washed
with 25 mL of saturated ammonium chloride. The layers
were separated, and the lower aqueous phase was washed
with 25 mL of MTBE. The combined organics were
concentrated to a 50 mL volume via rotovap. Octane (50
mL) was added and the solution concentrated to 50 mL (the
7: mp 148-151 °C; TLC Rf ) 0.35 (25% EtOAc/
1
hexanes); H NMR (CDCl3, 400 MHz) δ 7.38 (t, 1H, J )
8.8), 7.32 (d, 1H, J ) 14.4), 7.00 (dd, 1H, J ) 1.6, 8.4),
6.74 (bs, 1H), 3.96 (d, 2H, J ) 6.4), 3.23 (t, 2H, J ) 12.8),
2.44 (d, 2H, J ) 14.0), 2.37 (td, 2H, J ) 3.6, 13.6), 2.05 (d,
2H, J ) 14.4), 1.96 (nonet, 1H, J ) 6.8), 0.96 (d, 6H, J )
6.8); 13C NMR (CDCl3, 100 MHz) δ 160.28 (ds, JC-F
)
243.0), 153.49 (s), 138.71 (ds, JC-F )12.0), 129.93 (ds, JC-F
) 44), 126.67 (dd, JC-F ) 6), 113.85 (d), 107.03 (dd, JC-F
) 29.4), 71.65 (d), 71.24 (ds, JC-F ) 4.0), 37.67 (t), 37.64
(t), 27.95 (d), 23.90 (t), 19.03 (q); MS (CI, NH3) m/z (relative
intensity) 327 (7.0), 312 (6.4), 311 (17), 310 (100). Anal.
Calcd for C16H22FNO3S: C, 58.69; H, 6.77; N, 4.28.
Found: C, 58.39; H, 6.68; N, 4.27.
Acknowledgment
We thank William Perrault and Bruce Pearlman for their
assistance during this work.
Received for review September 6, 2000.
OP0002949
Vol. 5, No. 1, 2001 / Organic Process Research & Development
•
83