Synthesis of diaryl ethers, diaryl thioethers, and diarylamines mediated by potassium fluoride-alumina and 18-crown-6: Expansion of scope and utility

Article abstract of DOI:10.1021/jo980800g

An efficient alternative to the Ullmann ether synthesis of diaryl ethers, diaryl thioethers, and diarylamines involving the S(N)Ar addition of a phenol, thiophenol, or aniline to an appropriate aryl halide, mediated by potassium-fluoride alumina and 18-crown-6 in acetonitrile or DMSO, is described. Expansion of the reaction conditions to include DMSO as solvent has resulted in a far greater range of substitution patterns permitted on the electrophile. For example, it was found that electronically unfavorable 3- chlorobenzonitrile could be condensed with 3-methoxyphenol to form the corresponding diaryl ether in 66% yield, a combination not normally amenable to Ullmann coupling. Electron-withdrawing groups present on the electrophile may be as diverse as nitro, cyano, formyl, acetyl, ester, amide, and even aryl. The method features a simple reaction procedure that provides products in generally good to excellent purified yields.

Full text of DOI:10.1021/jo980800g

6338
J . Org. Chem. 1998, 63, 6338-6343
Syn th esis of Dia r yl Eth er s, Dia r yl Th ioeth er s, a n d Dia r yla m in es
Med ia ted by P ota ssiu m F lu or id e-Alu m in a a n d 18-Cr ow n -6:
Exp a n sion of Scop e a n d Utility1
J . Scott Sawyer,* Elisabeth A. Schmittling, J ayne A. Palkowitz, and William J . Smith, III
The Lilly Research Laboratories, Eli Lilly and Company, Indianapolis, Indiana 46285
Received April 28, 1998
An efficient alternative to the Ullmann ether synthesis of diaryl ethers, diaryl thioethers, and
diarylamines involving the SNAr addition of a phenol, thiophenol, or aniline to an appropriate aryl
halide, mediated by potassium-fluoride alumina and 18-crown-6 in acetonitrile or DMSO, is
described. Expansion of the reaction conditions to include DMSO as solvent has resulted in a far
greater range of substitution patterns permitted on the electrophile. For example, it was found
that electronically unfavorable 3-chlorobenzonitrile could be condensed with 3-methoxyphenol to
form the corresponding diaryl ether in 66% yield, a combination not normally amenable to Ullmann
coupling. Electron-withdrawing groups present on the electrophile may be as diverse as nitro,
cyano, formyl, acetyl, ester, amide, and even aryl. The method features a simple reaction procedure
that provides products in generally good to excellent purified yields.
Ch a r t 1
In tr od u ction
The presence of the diaryl ether subunit in a number
of synthetically challenging and medicinally important
natural products (Chart 1) such as combretastatin D-22
(1, antifungal), riccardin C3 (2, cytotoxin), piperazinomy-
cin4 (3, antifungal), and (-)-K-135 (4, ACE inhibitor), as
well as purely synthetic bioactive agents such as
LY2931116 (5, LTB4 receptor antagonist) and RH62017
(6, herbicidal), has renewed efforts in the development
of versatile methodology for its construction. Examina-
tion of the classic Ullmann ether synthesis2,8 has been
complemented by additional strategies including Pum-
merer-type rearrangements,9 inter- and intramolecular
SNAr reactions,10 arene-metal complexes,11 thallium-
promoted oxidative couplings,12 phenolic additions to
(1) Portions of this work have been presented at the following
conferences: (a) Schmittling, E. A.; Sawyer, J . S. Presented at the
205th National Meeting of the American Chemical Society, Denver,
CO, March 28-April 2, 1993. (b) Schmittling, E. A.; Palkowitz, J . A.;
Smith, W. J ., III.; Sawyer, J . S. Abstracts of Papers, 208th National
Meeting of the American Chemical Society, Washington, DC, August
21-26, 1994; American Chemical Society: Washington, DC, 1994;
ORGN 64. (c) Smith, W. J ., III.; Schmittling, E. A.; Sawyer, J . S.
Abstracts of Papers, 208th National Meeting of the American Chemical
Society, Washington, DC, August 21-26, 1994; American Chemical
Society: Washington, DC, 1994; ORGN 65.
(2) (a) Deshpande, V. E.; Gohkhale, N. J . Tetrahedron Lett. 1992,
33, 4213-4216. (b) Boger, D. L.; Sakaya, S. M.; Yohannes, D. J . Org.
Chem. 1991, 56, 4204-4207.
(3) Gottsegen, A.; Vermes, B.; Kajtarperedy, M.; Bihatsikarasai, E.;
Nogradi, M. Tetrahedron Lett. 1988, 29, 5039-5040.
cyclohexenone oxides,13 and Diels-Alder cyclizations.14
Related methods have been successfully employed in the
synthesis of important diaryl thioethers15 and diaryl-
(4) J ung, M. E.; Rohloff, J . C. J . Org. Chem. 1985, 50, 4909-4913.
(5) Boger, D. L.; Yohannes, D. J . Org. Chem. 1989, 54, 2498-2502.
(6) (a) Sawyer, J . S. Drugs Future 1996, 21, 610-614. (b) Sawyer,
J . S. Expert Opin. Invest. Drugs 1996, 5, 73-77.
(10) (a) Zhu, J . Synlett 1997, 133-144 and references therein. (b)
Boger, D. L.; Borzilleri, R. M.; Nukui, S.; Beresis, R. T. J . Org. Chem.
1997, 62, 4721-4736. (c) Boger, D. L.; Zhou, J . C.; Borzilleri, R. M.;
Nukui, S.; Castle, S. L. J . Org. Chem. 1997, 62, 4721-4736. (d)
Beugelmans, R.; Zamora, E. G.; Roussi, G. Tetrahedron Lett. 1997, 38,
8189-8192. (e) Rao, A. V. R.; Gurjar, M. K.; Lakshmipathi, P., Reddy,
M. M.; Nagarajan, M.; Pal, S.; Sarma, B. V. N. B. S.; Tripathy, N. K.
Tetrahedron Lett. 1997, 38, 7433-7436. (f) BoisChoussy, M.; Vergne,
C.; Neuville, L.; Beugelmans, R.; Zhu, J . Tetrahedron Lett. 1997, 38,
5795-5798. (g) Roussi, G.; Zamora, E. G.; Carbonnelle, A.-C., Beugel-
mans, R. Tetrahedron Lett. 1997, 38, 4401-4404. (h) Roussi, G.;
Zamora, E. G.; Carbonnelle, A.-C., Beugelmans, R. Tetrahedron Lett.
1997, 38, 4405-4406. (i) Vergne, C.; BoisChoussy, M.; Beugelmans,
R., Zhu, J . Tetrahedron Lett. 1997, 38, 1403-1406. (j) J anetka, J . W.;
Rich, D. H. J . Am. Chem. Soc. 1997, 119, 6488-6495. (k) Boger, D. L.;
Zhou, J . J . Org. Chem. 1996, 61, 3938-3939.
(7) J ohnson, W. O.; Kollman, G. E.; Swithenbank, C.; Yih, R. Y. J .
Agric. Food Chem. 1978, 26, 285-287.
(8) (a) Couladouros, E. A.; Soufli, I. C. Tetrahedron Lett. 1995, 36,
9369-9372. (b) Rao, A. V. R.; Chakraborty, T. K.; Reddy, K. L.; Rao,
A. S. Tetrahedron Lett. 1992, 33, 4799-4802. (c) Boger, D. L.;
Yohannes, D. J . Org. Chem. 1991, 56, 1763-1767. (d) Boger, D. L.;
Yohannes, D. J . Org. Chem. 1990, 55, 6000-6017. (e) J ung, M. E.;
J achiet, D.; Rohloff, J . C. Tetrahedron Lett. 1989, 30, 4211-4214. (f)
Evans, D. A.; Ellman, J . A. J . Am. Chem. Soc. 1989, 111, 1063-1072.
(g) Boger, D. L.; Yohannes, D. J . Org. Chem. 1988, 53, 487-499.
(9) J ung, M. E.; J achiet, D.; Khan, S. I.; Kim, C. Tetrahedron Lett.
1995, 36, 361-364.
S0022-3263(98)00800-7 CCC: $15.00 © 1998 American Chemical Society
Published on Web 08/14/1998

Synthesis of Diaryl Ethers, Diaryl Thioethers, and Diarylamines
J . Org. Chem., Vol. 63, No. 18, 1998 6339
amines.16 The genesis of the non-Ullmann methods listed
above is rooted in the insidious nature of the copper-
catalyzed addition of phenols or phenoxides to an ap-
propriate aryl halide. Frequently in the Ullmann sphere
one is confronted with elevated reaction temperatures,
formidable purification problems, and low yields, al-
though there are spectacular exceptions to this gener-
alization when considering select intermolecular versions.8f
With respect to the SNAr strategy for the formation of
simple diaryl ethers, a survey of the literature reveals a
plethora of variations on a very singular theme: the base-
catalyzed addition of a phenol to an electron-deficient
(usually nitro-containing) electrophile, preferably in a 1,2-
or 1,4-substitution pattern. Our interest in this field
grew out of a profound dissatisfaction with the Ullmann
paradigm and a desire to create a general SNAr method
for the synthesis of diaryl ethers, diaryl thioethers, and
diarylamines which would exhibit broad versatility with
maximum toleration for sensitive functionality. Origi-
nally, we communicated that stoichiometric quantities
of potassium fluoride-alumina, together with catalytic
quantities of 18-crown-6, effect coupling of phenols,
thiophenols, and anilines to 2- and 4-fluorobenzonitriles
in refluxing acetonitrile in good to excellent yields.17
Encouraged by the ease of construction of simple diaryl
ethers,18 we decided to further delineate the scope of our
general protocol, particularly in regard to the nature of
the solvent and reaction temperature. We now report a
full study of the original work together with an extension
of the method to encompass new nucleophiles and elec-
trophiles containing diverse functional groups, often in
electronically unfavorable patterns.
Ta ble 1. Cou p lin g of P h en ols, Th iop h en ols, a n d
An ilin es to F lu or oben zon itr iles in Aceton itr ile17
a
b
No reaction observed. Using 5 wt equiv (based on 4-nitro-
phenol) of KF‚Al2O3. Remainder was recovered starting materials.
2-fluorobenzonitrile. In the presence of 10 mol % tetra-
n-butylammonium bromide, this process resulted in a
94% chromatographed yield of 2-phenoxybenzonitrile
after a total reaction time of 7 days. While the rate of
the reaction was even slower in the absence of tetra-n-
butylammonium bromide, a brief survey of readily avail-
able reagents quickly led to the selection of 18-crown-6
as the phase transfer catalyst of choice. Under our now
standard acetonitrile conditions, the yield of 2-phenoxy-
benzonitrile was near quantitative after a reaction time
of 2 days (Table 1, entry 1). In the case of electronically
favorable 2- and 4-fluorobenzonitrile, substitution of the
mildly σ electron-withdrawing 3-methoxy group on the
nucleophile resulted in high yields of diaryl ether ac-
companied by somewhat lengthened reaction times (en-
tries 3 and 5). Unsurprisingly, 3-fluorobenzonitrile was
inert under these nonforcing conditions (entry 4).20 It is
also interesting to note that the yield of 2-(3-methoxy-
phenoxy)benzonitrile derived from the classic Ullmann
coupling of 3-methoxyphenol with 2-bromobenzonitrile
(43%) is substantially inferior to that illustrated in entry
3.21
In entries 6-11, 2-fluorobenzonitriles substituted with
a variety of electron-donating and/or electron-withdraw-
ing groups were reacted with 3-methoxyphenol over
varying (mostly moderate) reaction times. A remarkable
lack of deviation from the high yield observed for entry
3 is evident. The utility of the method is further
demonstrated by entry 12, where 4-fluorobenzonitrile was
successfully condensed with 2-tert-butylphenol, a steri-
cally crowded nucleophile notoriously uncooperative when
subjected to normal Ullmann conditions. Coupling of a
moderately electron-deficient phenol (methyl 4-hydroxy-
benzoate) with 2-fluorobenzonitrile (entry 13) also re-
Resu lts a n d Discu ssion
Acetonitrile was the solvent examined in our initial
foray. It was found that 1-2.5 wt equiv (based on the
nucleophile) of 37% w/w potassium fluoride supported on
basic alumina19 in refluxing acetonitrile is an efficient
mediator for promoting the SNAr addition of phenol to
(11) (a) Pearson, A. J .; Zhang, P. L.; Bignan, G. J . Org. Chem. 1997,
62, 4536-4538. (b) Pearson, A. J .; Zhang, P.; Lee, K. J . Org. Chem.
1996, 61, 6581-6586. (c) Pearson, A. J .; Bignan, G.; Zhang, P.;
Chelliah, M. J . Org. Chem. 1996, 61, 3940-3941. (d) Pearson, A. J .;
Bignan, G. Tetrahedron Lett. 1996, 37, 735-738. (e) Pearson, A. J .;
Park, J . G.; Zhu, P. Y. J . Org. Chem. 1992, 57, 3583-3589. (f) Pearson,
A. J .; Bruhn, P. R. J . Org. Chem. 1991, 56, 7092-7097.
(12) (a) Konishi, H.; Okuno, T.; Nishiyama, S.; Yamamura, S.;
Koyasu, K.; Terada, Y. Tetrahedron Lett. 1996, 37, 8791-8794. (b)
Inoue, T.; Sasaki, T.; Takayanagi, H.; Harigaya, Y.; Hoshino, O.; Hara,
H.; Inaba, T. J . Org. Chem. 1996, 61, 3936-3937. (c) Suzuki, Y.;
Nishiyama, S.; Yamamura, S. Tetrahedron Lett. 1989, 30, 6043-6046.
(d) Evans, D. A. Ellman, J . A.; DeVries, K. M. J . Am. Chem. Soc. 1989,
111, 8912-8914.
(13) (a) J ung, M. E.; Starkey, L. S. Tetrahedron 1997, 53, 8815-
8824. (b) J ung, M. E.; Starkey, L. S. Tetrahedron Lett. 1995, 36, 7363-
7366.
(14) (a) Olsen, R. K.; Feng, X.; Campbell, M.; Shao, R.; Math, S. K.
J . Org. Chem. 1995, 60, 6025-6031. (b) Feng, X.; Olsen, R. K. J . Org.
Chem. 1992, 57, 5811-5812.
(15) (a) Protiva, M. Drugs Future 1991, 16, 911-916. (b) Hobbs, D.
W.; Still, D. W. Tetrahderon Lett. 1987, 28, 2805-2808.
(16) (a) Gorvin, J . H. J . Chem. Soc., Perkin Trans. 1 1988, 1331-
1335. (b) Kulagowski, J . J .; Rees, C. W. Synthesis 1980, 215.
(17) Schmittling, E. A.; Sawyer, J . S. J . Org. Chem. 1993, 58, 3229-
3230.
(18) For pioneer reports detailing the condensation of phenols with
highly activated aryl chlorides mediated by potassium fluoride-
alumina, see: (a) Kang, R. H.; Zhu, J . S.; Li, W. Z.; Liu, S. R. Chin.
Chem. Lett. 1990, 1, 55-56. (b) Hwang, K.-J .; Park, S. K. Synth.
Commun. 1990, 20, 949-954.
(19) (a) Schmittling, E. A.; Sawyer, J . S. Tetrahedron Lett. 1991,
7207-7210. (b) For a review of organic reactions on alumina, see:
Kabalka, G. W.; Pagni, R. M. Tetrahedron 1997, 53, 7999-8065.
(20) Under the standard acetonitrile conditions, no reaction was also
observed when 3-methoxyphenol was reacted with either 2- or 4-chlo-
robenzonitrile, even after prolonged periods (9 days) and in the
presence of large quantities of potassium fluoride-alumina (2.5 wt
equiv based on 3-methoxyphenol).
(21) J ackson, W. T.; Froelich, L. L.; Gapinski, D. M.; Mallett, B. E.;
Sawyer, J . S. J . Med. Chem. 1993, 36, 1726-1734.

6340 J . Org. Chem., Vol. 63, No. 18, 1998
Sawyer et al.
Sch em e 1a
sulted in smooth product formation, although prolonged
reaction time was necessary to secure the maximum
yield. Unfortunately, nucleophiles containing strongly
electron-withdrawing groups, such as 4-nitrophenol,
proved to react sluggishly in acetonitrile even after
extended reaction time and utilization of 5 wt equiv of
potassium fluoride-alumina (entry 14). Since enhanced
phenoxide stabilization by the nitro group would nor-
mally be expected, it is unlikely that full deprotonation
by the potassium fluoride-alumina matrix22 is operative
in phenolic condensations. Rather, comparison of entries
1, 3, 13, and 14 suggests that, with phenols, nucleophi-
licity is decreased through the addition of electron-
withdrawing groups and that potassium fluoride-
alumina functions more as a proton sponge effecting only
partial deprotonation. These results are especially in-
teresting in light of the known preference for nonhin-
dered electron-rich phenols and electron-deficient aryl
halides as reactants in Ullmann-type chemistry.
a
Reaction conditions, reagents and yields: (a) KF‚Al2O3 (1 wt
equiv), 18-crown-6 (10 mol %), acetonitrile, reflux, 96 h, 82%; (b)
KOH, 30% H2O2, MeOH, EtOH, reflux, 18 h, 87%; (c) P2O5,
CH3SO3H, 18 h, 92%.
Ta ble 2. Cou p lin g of P h en ols to Nitr oben zen es in
Aceton itr ile
Evaluation of the reactivity of thiophenols and anilines
toward 2- or 4-flourobenzonitriles indicated some diver-
gence from purely phenolic couplings. In general, it was
found that thiophenols were more reactive as reflected
in lower overall reaction times (entries 15-19), a result
almost certainly due to the greater nucleophilicity of the
mercapto group. For example, unlike 3-methoxyphenol,
3-methoxythiophenol coupled quantitatively with 2-fluo-
robenzonitrile to provide the expected diaryl thioether
in only 18 h at reflux (entry 16, compare to entry 3); a
similar result was observed for coupling with 4-fluo-
robenzonitrile (entry 17, compare to entry 5). A longer
reaction time and slightly lower yield was realized with
2-chlorothiophenol, a nucleophile substituted with a
mildly electron-withdrawing group (entry 19). While
N-methylaniline could be successfully coupled to an
activated electrophile (entry 20), additions involving
unactivated electrophiles generally led to complex mix-
tures, as did the reaction of aniline with 2-fluorobenzoni-
trile. The addition of 4-nitroaniline to 2-fluorobenzoni-
trile (entry 21) did provide the expected diarylamine in
reasonable yield. Considering the reluctance of aniline
to add cleanly to 2-fluorobenzonitrile compared to phenol
(entry 1), it is interesting to note that the addition of
4-nitroaniline is a much more facile process than that of
4-nitrophenol (entry 14).
a
Major product derives from displacement of the 2-fluoro atom.
b
Remainder (19%) is the bis-substituted product. Electrophile is
p-dinitrobenzene. c Electrophile is m-dinitrobenzene. No reaction
d
observed.
While the examples shown in Table 1 are not exhaus-
tive, the nitrile-substituted products obtained represent
useful intermediates suitable for further synthetic ma-
nipulation. In a specific example, we have used our basic
acetonitrile method in an efficient construction of the key
diaryl ether fragment of LY293111 (5), a synthesis which
originally incorporated an obstreperous Ullmann ether
coupling.23 A further demonstration in the field of
xanthanoid synthesis is given in Scheme 1, where 3,4,5-
trimethoxyphenol was smoothly added to 2-fluoroben-
zonitrile to provide diaryl ether 7 in 82% yield. Hydrol-
ysis to acid 8 and subsequent Friedel-Crafts cyclization
provided 1,2,3-trimethoxyxanthone, a natural product
isolated from the stems and roots of Polygala arillata.24
The acetonitrile conditions have been conveniently
employed to mediate the coupling of phenols to nitroben-
zenes containing a variety of leaving groups (Table 2).
Reactions carried out with generally neutral phenols and
4-fluoronitrobenzene proceeded with reasonable reaction
times and in high yields (entries 22-24). The extreme
electrophilicity of 2- and 4-halonitrobenzenes is well
documented, and indeed there have been a number of
reagents of varying basicity applied to phenolic conden-
sations with great success.25 However, these examples
(22) The exact nature of the active basic species present in potassium
fluoride-alumina reagents has been the subject of some controversy,
see: (a) Ando, T.; Clark, J . H.; Cork, D. G.; Hanafusa, T.; Ichihara, J .;
Kimura, T. Tetrahedron Lett. 1987, 28, 1421-1424. (b) Weinstock, L.
M.; Stevenson, J . M.; Tomellini, S. A.; Pan, S.-H.; Utne, T.; J obson, R.
B.; Reinhold: D. F. Tetrahedron Lett. 1986, 27, 3845-3848.
(24) Ghosal, S.; Banerjee, S.; Chauhan, R. B. P. S.; Srivastava, R.
S. J . Chem. Soc., Perkin Trans. 1 1977, 740-743.
(23) Sawyer, J . S.; Bach, N. J .; Baker, S. R.; Baldwin, R. F.;
Borromeo, P. S.; Cockerham, S. L.; Fleisch, J . H.; Floreancig, P.;
Froelich, L. L.; J ackson, W. T.; Marder, P.; Palkowitz, J . A.; Roman,
C. R.; Saussy, D. L., J r.; Schmittling, E. A.; Silbaugh, S. A.; Spaethe,
S. M.; Stengel, P. W.; Sofia, M. J . J . Med. Chem. 1995, 38, 4411-4432.
(25) (a) Kajimoto, T.; Yahiro, K.; Nohara, T. Chem. Lett. 1988, 1113-
1114. (b) Fuchigami, T.; Awata, T.; Nonaka, T.; Baizer, M. M. Bull.
Soc. Chem. J pn. 1986, 59, 2873-2879. (c) Brunelle, D. J .; Singleton,
D. A. Tetrahedron Lett. 1984, 25, 3383-3386. (d) Wright, J .; J orgensen,
E. C. J . Org. Chem. 1968, 33, 1245-1246.

Synthesis of Diaryl Ethers, Diaryl Thioethers, and Diarylamines
J . Org. Chem., Vol. 63, No. 18, 1998 6341
Sch em e 2a
Ta ble 3. Cou p lin g of 4-Su bstitu ted Th iop h en ols to
Ha lon itr oben zen es in Aceton itr ile
a
Reaction conditions, reagents and yields: (a) KF‚Al2O3 (2 wt
equiv), 18-crown-6 (10 mol %), acetonitrile, reflux, 110 h, 10%; (b)
KF‚Al2O3 (2 wt equiv), 18-crown-6 (10 mol %), water (4 equiv),
acetonitrile, reflux, 144 h, 40%; (c) KF‚Al2O3 (2 wt equiv), 18-
crown-6 (10 mol %), acetonitrile, reflux, 40 h, 11%.
are usually conducted with neutral or electron-rich
phenols. In refluxing acetonitrile, potassium fluoride-
alumina was found to effectively mediate the addition of
4-nitrophenol, an electron-deficient nucleophile, to 4-fluo-
ronitrobenzene (entry 25). As with entry 13 (Table 1),
this is the exact reverse of the nucleophilic electronic
demand featured in conventional Ullmann ether synthe-
ses. Similar to fluorobenzonitrile additions, sterically
challenged nucleophiles couple well to halonitrobenzenes
provided sufficient reaction times are employed (entries
26-27). Entries 28-32 illustrate further examples
involving substitutions on the electrophile, including the
observation of partial bis coupling of 4-methoxyphenol
to 2,5-difluoronitrobenzene (entry 32). Unlike our expe-
rience with benzonitriles, nitrobenzenes may be aug-
mented with halogen leaving groups other than fluorine.
4-Chloro-, 4-bromo-, and 4-iodonitrobenzene all reacted
with 4-methoxyphenol to produce the expected diaryl
ether in good yields, although long reaction times were
needed (entries 33-35, compare to entry 23). While
addition of phenol to p-dinitrobenzene provided easy
access to 4-phenoxynitrobenzene (entry 36),26 the addition
of 3-methoxyphenol to electronically unfavorable m-
dinitrobenzene resulted in no reaction (entry 37). Many
different conditions have been used in such nitro dis-
placements for the preparation of diaryl ethers and
diarylamines.27
The exact mechanism operative in the case of electro-
philes such as 4-chloronitrobenzene is unclear. Clark et
al.28 have demonstrated that over a 24 h period at room
temperature 2-chloronitrobenzene undergoes a slow halo-
gen exchange when exposed to a dry fluoride source, in
marked contrast to the fast fluorodenitration process that
occurs with dinitrobenzenes. This result may explain the
lengthy reaction times observed for entry 33 (and possibly
entries 34-35), but considering the reaction conditions,
a slow, direct nucleophilic displacement of chloro by
4-methoxyphenol cannot be ruled out. Our own attempts
to examine potassium fluoride-alumina as a fluoridation
agent on 4-chloronitrobenzene were complicated by the
observation that 4-fluoronitrobenzene will react under
the standard acetonitrile conditions to produce 4-(4-
nitrophenoxy)nitrobenzene in low (no added water) to
moderate (with added water) yield (Scheme 2). Similar
results were observed with p-dinitrobenzene. The pro-
duction of diaryl ethers through the action of metal
Sch em e 3a
a
Reaction conditions, reagents and yield: (a) KF‚Al2O3 (2 wt
equiv), 18-crown-6 (10 mol %), DMSO, 60 °C, 5 h, 82%.
carbonates and phase transfer catalysts on activated
electrophiles such as 2-fluoronitrobenzene has been
reported previously.29 The oxygen source in that example
was derived from carbonate, whereas with the potassium
fluoride-alumina/acetonitrile conditions it probably origi-
nates from potassium hydroxide. While water may effect
direct nucleophilic displacement of fluoride from 4-fluo-
ronitrobenzene to form 4-nitrophenol, its role as a simple
potassium hydroxide solublizing agent cannot be dis-
counted. In the case of p-dinitrobenzene, another mech-
anism may be operative involving a single fluorodeni-
tration followed by attack of free nitrite anion and
subsequent disproportionation to give 4-nitrophenol.30
Examples of the addition of thiophenols to haloni-
trobenzenes in refluxing acetonitrile are presented in
Table 3. In general, the combination of superb nucleo-
philes (thiophenols) and electrophiles (halonitrobenzenes)
resulted in high yields over short reaction times, as is
evidenced by entries 38-40. Good yields could also be
realized when using less active electrophiles such as
4-bromo- and 4-iodonitrobenzene over longer reaction
periods (entries 41-43). Although several other bases
may be used to successfully prepare nitro-substituted
diaryl thioethers, few feature the simple reaction condi-
tions and workup inherent with potassium fluoride-
alumina in acetonitrile. However, in one subsequent
example (the addition of 4-methoxythiophenol to 2-chloro-
6-nitrobenzonitrile) it was apparent that use of refluxing
acetonitrile, while effective, resulted in a sluggish reac-
tion rate. We therefore turned to DMSO in an attempt
to shorten the reaction time of this particular addition
and possibly extend the scope of our general method.
Using 2 wt equiv of potassium fluoride-alumina in
DMSO at 60 °C, coupling smoothly proceeded to comple-
tion in only 5 h to give compound 10 (Scheme 3), an
intermediate used in the preparation of analogues of the
antischistosomal and antitumor agent hycanthone.31 The
preference for displacement of the nitro group is consis-
tent with studies involving the fluorodenitration of
(26) For
a discussion of the radical nature of the addition of
phenolate to p-dinitrobenzene, see: Mir, M.; Espin, M.; Marquet, J .;
Gallardo, I.; Tomasi, C. Tetrahedron Lett. 1994, 35, 9055-9058.
(27) (a) Shevelev, S. A.; Dutov, M. D.; Vatsadze, I. A.; Seruchkina,
O. V.; Korolev, M. A.; Rusanov, A. L. Russ. Chem. Bull. 1995, 44, 384-
385. (b) Sammes, P. G.; Thetford, D.; Voyle, M. J . Chem. Soc., Perkin
Trans. 1 1988, 3229-3231. (c) Clark, J . H.; Owen, N. D. S. Tetrahedron
Lett. 1987, 28, 3627-3630. (d) Gorvin, J . H. J . Chem. Soc., Chem.
Commun. 1985, 238-239. (e) Williams, F. J .; Donahue, P. E. J . Org.
Chem. 1977, 42, 3414-3419.
(29) Cella, J . A.; Bacon, S. W. J . Org. Chem. 1984, 49, 1122-1125.
(30) Maggini, M.; Passudetti, M.; Gonzales-Trueba, G.; Prato, M.;
Quintily, U.; Scorrano, G. J . Org. Chem. 1991, 56, 6406-6411.
(31) Archer, S.; Pica-Mattoccia, L.; Cioli, D.; Seyed-Mozaffari, A.;
Zayed, A.-H. J . Med. Chem. 1988, 31, 254-260.
(28) Clark, J . H.; Smith, D. K. Tetrahedron Lett. 1985, 26, 2233-
2236.

6342 J . Org. Chem., Vol. 63, No. 18, 1998
Sawyer et al.
Ta ble 4. Cou p lin g of P h en ols to Ha loben zen es in DMSO
Ta ble 5. Cou p lin g of Th iop h en ols to Ha loben zen es in
DMSO
Sch em e 4a
methyl 4-chloro-2-nitrobenzoate (electronically similar to
2-chloro-6-nitrobenzonitrile), where unsupported potas-
sium fluoride was used in DMSO at 130 °C.32
a
Reaction conditions, reagents and yield: (a) KF‚Al2O3 (2 wt
Further exploration of DMSO as a forcing solvent
quickly broadened the utility of potassium fluoride-
alumina as a mediator of diaryl ether synthesis. By
routinely running reactions in DMSO at 140 °C, we were
able to couple electrophiles containing electron-with-
drawing groups other than cyano or nitro (Table 4). The
reactivity of aryl fluorides substituted with formyl (entry
44) or acetyl (entry 45) groups is especially interesting
as these may be oxidized to ultimately produce phenols,
a transformation which represents a formal coupling of
a phenol with an oxygenated electrophile.33 Addition of
phenols to aryl fluorides substituted with esters or
amides was also feasible provided sufficient reaction
times were allowed and an ortho or para substitution
pattern employed (entries 46-48). Although these diaryl
ether products are accessible through the corresponding
nitriles, a more direct approach may be preferred syn-
thetically in some instances. We were surprised to find
that under the DMSO protocol electrophiles possessing
less favorable electron-withdrawing groups and substitu-
tion patterns still provided low yields of diaryl ethers
(entries 49-50), including the expected product from
addition of 3-methoxyphenol to 4-fluorobiphenyl (entry
51). Fluorobenzonitriles were effective electrophiles
when exhibiting both favorable (entry 52) and unfavor-
able (entry 53, compare to Table 1, entry 4) substitution
patterns, while 2- and 4-fluorobenzonitrile underwent
smooth coupling with electronically unfavorable 2-cy-
anophenol (entries 54-55). Additions of phenols to
bromo- and chlorobenzonitriles were equally effective
(entries 56-62), including that of 3-methoxyphenol to
3-chlorobenzonitrile, which produced the expected prod-
uct in 66% yield (entry 58). The only apparent downside
to the substitution of DMSO for acetonitrile in diaryl
ether or thioether formation is the higher temperature
equiv), 18-crown-6 (10 mol %), DMSO, 140 °C, 18 h, 82%.
range afforded by the former, a potential problem with
thermally sensitive substrates. This appears to be more
than offset by the broadened scope of the reaction.
Formation of diaryl thioethers was also found to be
significantly accelerated in DMSO (Table 5), where little
differentiation was seen between thiophenolic additions
to 2-, 3-, or 4-halobenzonitriles (entries 63-68). As in
the diaryl ether examples, ester and amide also served
as a functional electron-withdrawing group on the elec-
trophile (entries 69-72).
Attempts to extend the DMSO variation to anilines has
met with mixed success. While the N-arylation of indole
mediated by potassium fluoride-alumina has proven to
be a facile process,34 general diarylamine formation is still
problematical. However, it was possible to add indoline
to a highly activated electrophile using the DMSO
variation to provide the expected product in good yield
(Scheme 4). This result is consistent with entry 20 (Table
1), where an N-alkylaniline could be coupled in refluxing
acetonitrile to an activated electrophile in good yield.
Furthermore, we have found that primary anilines may
be efficiently added to activated compounds such as
2-chloro-5-nitropyridine under DMSO conditions (Scheme
5), a result that fully compliments the original synthesis
of 2-aryloxypyridines using potassium fluoride-alumina.18b
We conclude that the challenges posed by diarylamine
synthesis represent fertile ground ripe for further explo-
ration.
Con clu sion s
In this paper we have described a useful method for
the facile preparation of diaryl ethers, diaryl thioethers,
and diarylamines. The method originally developed
(32) Beaumont, A. J .; Clark, J . H.; Boechat, N. A. J . Fluorine Chem.
1993, 63, 25-30.
(33) Yeager, G. W.; Schissel, D. N. Synthesis 1991, 63-68.
(34) Smith, W. J ., III.; Sawyer, J . S. Tetrahedron Lett. 1996, 37,
299-302.

Synthesis of Diaryl Ethers, Diaryl Thioethers, and Diarylamines
J . Org. Chem., Vol. 63, No. 18, 1998 6343
Sch em e 5a
acetonitrile (25 mL) was refluxed with stirring for 96 h. The
mixture was cooled to room temperature, filtered, and diluted
with 1 N aqueous sodium hydroxide. The resulting mixture
was extracted once with diethyl ether. The organic layer was
washed once with saturated sodium chloride solution, dried
(sodium sulfate), filtered, and concentrated in vacuo to provide
3.00 g (82%) of the title product as tan needles: mp 103-105
°C. Anal. Calcd for C16H15NO4: C, 67.36; H, 5.30, N, 4.91.
Found: 67.10; H, 5.32; N, 4.85.
a
Reaction conditions, reagents and yield: (a) KF‚Al2O3 (1.0 wt
equiv), 18-crown-6 (10 mol %), DMSO, 140 °C, 18 h, 86%.
2-(3,4,5-Tr im eth oxyp h en oxy)ben zoic Acid (8). Com-
pound 7 (2.00 g, 7.01 mmol) was dissolved in a hot mixture of
methanol (5 mL) and ethanol (20 mL). Potassium hydroxide
(8.00 g, 143 mmol) and a 30% hydrogen peroxide solution (8
mL) was added, and the resulting mixture was refluxed with
stirring for 18 h. The mixture was concentrated in vacuo, and
the residue was dissolved in 5 N aqueous sodium hydroxide.
This solution was washed once with ether, acidified with 5 N
aqueous hydrochloric acid, and extracted with ethyl acetate.
The ethyl acetate layer was separated, dried (sodium sulfate),
filtered, and concentrated in vacuo to provide 1.86 g (87%) of
the title product as an off-white powder. Anal. Calcd for
16H16O6: C, 63.15; H, 5.30. Found: C, 63.40; H, 5.33.
1,2,3-Tr im eth oxyxa n th on e (9). Phosphorus pentoxide
(3.00 g, 20.8 mmol) was dissolved in methanesulfonic acid (20
mL). Compound 8 (1.60 g, 5.26 mmol) was added, and the
resulting solution was stirred at room temperature for 18 h.
The mixture was carefully poured over crushed ice, and the
resulting precipitate was (2.00 g) filtered. The filtrate was
extracted with methylene chloride. The organic layer was
separated, dried (sodium sulfate), and filtered to provide
additional crude product (0.65 g). The combined crude mate-
rial was recrystallized from acetone to give 1.38 g (92%) of the
title product as a tan solid: mp 117-120 °C: 1H NMR (CDCl3)
δ 8.31 (d, J ) 8 Hz, 1 H), 7.67 (t, J ) 7 Hz, 1 H), 7.40 (d, J )
8 Hz, 1 H), 7.36 (t, J ) 8 Hz, 1 H), 6.76 (s, 1 H), 4.05 (s, 3 H),
4.00 (s, 3 H), 3.93 (s, 3 H); MS-FD m/e 286 (p); IR (CHCl3,
cm-1) 1651, 1465, 1310. Anal. Calcd for C16H14O5: C, 67.13;
H, 4.93. Found: C, 67.36; H, 4.99.
involving the use of potassium fluoride-alumina, 18-
crown-6, and refluxing acetonitrile has proven to be
highly effective in mediating the condensation of phenols,
thiophenols, and certain anilines with fluorobenzonitriles
and halonitrobenzenes. The discovery that through the
use of DMSO the method may be extended to include
aromatic nucleophiles and electrophiles possessing un-
favorable functionality and substitution patterns proved
to be key in demonstrating its true scope. Furthermore,
a simple reaction procedure combined with easy workup
C
and generally good to excellent yields make for an
attractive alternative to other SNAr paradigms. Overall
the potassium fluoride-alumina method appears to be
of broad utility and may find particular application in
the construction of synthetically challenging and biologi-
cally important diaryl ethers.
Exp er im en ta l Section
Melting points were determined on a Thomas-Hoover ap-
paratus and are uncorrected. NMR spectra were determined
on a GE QE-300 spectrometer. Analytical data was deter-
mined by the Physical Chemistry Department (MC525) of the
Lilly Research Laboratories. All reactions were conducted
under nitrogen atmosphere with stirring unless otherwise
noted.
Gen er a l Cou p lin g P r oced u r e in Aceton itr ile. A mix-
ture of the phenol, thiophenol, or aniline (20 mmol), the
electrophile (20 mmol), 18-crown-6 (2.0 mmol), and 37% w/w
potassium fluoride-alumina19a (1-2.5 g per g of nucleophile)
in acetonitrile (35 mL) was refluxed and followed to reaction
completion via TLC. The reaction mixture was cooled to room
temperature, partitioned between equal parts ether and water,
and shaken vigorously. The aqeous layer and alumina sedi-
ments were drawn from the funnel, and the resulting organic
phase was washed once with a saturated potassium chloride
solution. The organic phase was dried (sodium sulfate),
filtered, and concentrated in vacuo to provide product. Ana-
lytically pure samples could usually be obtained through
recrystallization. In cases where further purification was
necessary, column chromatography on silica gel eluting with
ethyl acetate/hexane mixtures was used.
Gen er a l Cou p lin g P r oced u r e in DMSO. A mixture of
the phenol, thiophenol, or aniline (20 mmol), the electrophile
(20 mmol), 18-crown-6 (2.0 mmol), and 37% w/w potassium
fluoride-alumina19a (2.5 g per g of nucleophile) in DMSO (50
mL) was heated at 140 °C and followed to reaction completion
via TLC. The reaction mixture was cooled to room tempera-
ture, diluted with ether, and filtered. The resulting solution
was placed in a separatory funnel and washed once with water
and once with saturated potassium chloride solution. The
organic phase was dried (sodium sulfate), filtered, and con-
centrated in vacuo. Products were purified either through
recrystallization or by column chromatography on silica gel
eluting with ethyl acetate/hexane mixtures.
2-C h lo r o -6-(4-m e t h o x y t h io p h e n o x y )b e n zo n it r ile
1
(10): 82% yield, mp 114-116 °C; H NMR (CDCl3) δ 7.51 (d,
J ) 9 Hz, 2 H), 7.27 (d, J ) 7 Hz, 1 H), 7.22 (t, J ) 8 Hz, 1 H),
6.99 (d, J ) 9 Hz, 2 H), 6.74 (d, J ) 7 Hz, 1 H), 3.87 (s, 3 H);
MS-FD m/e 275 (p), 277; IR (CHCl3, cm-1) 2230, 1592, 1174.
Anal. Calcd for C14H10NClOS: C, 60.98; H, 3.65; N, 5.08; S,
11.63. Found: C, 61.11; H, 3.71; N, 5.09; S, 11.30.
1
2-(1-In d olin yl)-5-n itr oben zon itr ile (11): 82% yield; H
NMR (CDCl3) δ 8.51 (d, J ) 2 Hz, 1 H), 8.22 (dd, J ) 8, 2 Hz,
1 H), 7.65 (d, J ) 8 Hz, 1 H), 7.32 (d, J ) 7 Hz, 1 H), 7.18 (m,
2 H), 7.02 (t, J ) 8 Hz, 1 H), 4.39 (t, J ) 7 Hz, 2 H), 3.23 (t,
J ) 7 Hz, 2 H); MS-FD m/e 265 (p); IR (CHCl3, cm-1) 2228,
1571, 1338.
2-[N -(3-Tr iflu or om e t h ylp h e n yl)a m in o]-5-n it r op yr i-
1
d in e (12): 86% yield; H NMR (d6-DMSO) δ 10.40 (s, 1 H),
9.09 (d, J ) 2 Hz, 1 H), 8.36 (dd, J ) 9, 2 Hz, 1 H), 8.09 (s, 1
H), 7.99 (d, J ) 8 Hz, 1 H), 7.59 (t, J ) 8 Hz, 1 H), 6.98 (d, J
) 8 Hz, 1 H); MS-FD m/e 283 (p); IR (CHCl3, cm-1) 1603,
1330, 1292. Anal. Calcd for C12H8N3O2F3: C, 50.89; H, 2.85;
N, 14.84. Found: C, 50.70; H, 2.95; N, 14.89.
Ackn owledgm en t. The authors wish to thank mem-
bers of the Physical Chemistry Group, Lilly Research
Laboratories, for providing analytical data.
Su p p or tin g In for m a tion Ava ila ble: Spectroscopic data,
melting points, and elemental analyses for compounds listed
in the tables (12 pages). This material is contained in libraries
on microfiche, immediately follows this article in the microfilm
version of the journal, and can be ordered from the ACS; see
any current masthead page for ordering information.
2-(3,4,5-Tr im eth oxyp h en oxy)ben zon itr ile (7). A mix-
ture of 3,4,5-trimethoxyphenol (2.35 g, 12.8 mmol), 2-fluo-
robenzonitrile (1.55 g, 12.8 mmol), 37% potassium fluoride/
alumina (2.4 g), and 18-crown-6 (338 mg, 1.28 mmol) in
J O980800G