Deoxygenation of polyhydroxybenzenes: An alternative strategy for the benzene-free synthesis of aromatic chemicals

Article abstract of DOI:10.1021/ja0176346

New synthetic connections have been established between glucose and aromatic chemicals such as pyrogallol, hydroquinone, and resorcinol. The centerpiece of this approach is the removal of one oxygen atom from 1,2,3,4-tetrahydroxybenzene, hydroxyhydroquinone, and phloroglucinol methyl ether to form pyrogallol, hydroquinone, and resorcinol, respectively. Deoxygenations are accomplished by Rh-catalyzed hydrogenation of the starting polyhydroxybenzenes followed by acid-catalyzed dehydration of putative dihydro intermediates. Pyrogallol synthesis consists of converting glucose into myo-inositol, oxidation to myo-2-inosose, dehydration to 1,2,3,4-tetrahydroxybenzene, and deoxygenation to form pyrogallol. Synthesis of pyrogallol via myo-2-inosose requires 4 enzyme-catalyzed and 2 chemical steps. For comparison, synthesis of pyrogallol from glucose via gallic acid intermediacy and the shikimate pathway requires at least 20 enzyme-catalyzed steps. A new benzene-free synthesis of hydroquinone employs conversion of glucose into 2-deoxy-scyllo-inosose, dehydration of this inosose to hydroxyhydroquinone, and subsequent deoxygenation to form hydroquinone. Synthesis of hydroquinone via 2-deoxy-scyllo-inosose requires 2 enzyme-catalyzed and 2 chemical steps. By contrast, synthesis of hydroquinone using the shikimate pathway and intermediacy of quinic acid requires 18 enzyme-catalyzed steps and 1 chemical step. Methylation of triacetic acid lactone, cyclization, and regioselective deoxygenation of phloroglucinol methyl ether affords resorcinol. Given the ability to synthesize triacetic acid lactone from glucose, this constitutes the first benzene-free route for the synthesis of resorcinol. Copyright

Full text of DOI:10.1021/ja0176346

Published on Web 05/02/2002
Deoxygenation of Polyhydroxybenzenes: An Alternative Strategy for the
Benzene-Free Synthesis of Aromatic Chemicals
Chad A. Hansen and J. W. Frost*
Department of Chemistry, Michigan State UniVersity, East Lansing, Michigan 48824
Received November 29, 2001
Benzene-free syntheses of mono-,1a di-,1b,c and trihydroxy
benzenes1d have primarily been based on the shikimate pathway.
Table 1. Products, Isolated Yields, and Reaction Conditions for
Deoxygenation of Polyhydroxybenzenes
A completely different route to aromatic chemicals involving
intermediacy of polyhydroxyaromatics derived from glucose is
described in this report. The centerpiece of this strategy is the
discovery that catalytic hydrogenation of polyhydroxybenzenes
followed by acid-catalyzed dehydration of the dihydro intermediates
results in loss of one oxygen atom from the starting polyhydroxy-
benzene. The result (Table 1) is a new synthesis of pyrogallol (4)
from glucose, a new benzene-free synthesis of hydroquinone (6),
and the first benzene-free synthesis of resorcinol (2).
Reduction of phloroglucinol (1a) with NaBH
4
followed by
azeotropic removal of H
2
O is known to lead to resorcinol (2).2a As
for catalytic reduction, formation of the sodium salt of dihydro-
phloroglucinol has been reported when phloroglucinol is hydroge-
2 3
nated using 5% Rh on Al O although dehydration of this
intermediate was not examined.2 Phloroglucinol was thus an ideal
substrate to test the viability of polyhydroxybenzene deoxygenation
as a route for synthesis of aromatic chemicals. Phloroglucinol (1a)
was hydrogenated over 5% Rh on Al O under basic conditions.
2 3
Subsequent acid-catalyzed dehydration led to the formation of
b
a
Polyhydroxybenzene (1 M) in degassed, aqueous 1 M NaOH was
reacted with 0.012 equiv of Rh, Pt, or Pd (5 wt %, metal/solid support) in
a Parr hydrogenation apparatus at room temperature under 50 psi H2 for 12
resorcinol (2, Table 1) in 82% isolated yield after distillation. Use
of Rh on Al O gave somewhat higher yields of resorcinol (2) than
2 3
use of activated C as the support. Successively lower yields of 2
h. After filtration, acidification to pH 6.0, and concentration, the residue
b
(0.25 M) was refluxed in 0.5 M H2SO4 under Ar for 9 h. Yields (mol/
mol) are of purified product after distillation.
were obtained for use of Rh, Pt, and Pd as hydrogenation catalysts
Scheme 1 a
(Table 1). 1,2,3,4-Tetrahydroxybenzene (3) and hydroxyhydro-
quinone (5) were then similarly reacted to test the generality of
the deoxygenation strategy for polyhydroxybenzenes that can be
synthesized from glucose.
In route to 1,2,3,4-tetrahydroxybenzene (3), myo-inositol (8a)
was microbially synthesized from glucose and then oxidized by a
second microbe (Scheme 1).3 As previously reported,3 acid-
catalyzed dehydration of the resulting myo-2-inosose (9) was used
to obtain 1,2,3,4-tetrahydroxybenzene (3, Scheme 1). Catalytic
hydrogenation of 1,2,3,4-tetrahydroxybenzene (3) using Rh on
Al
metal or its solid support had little impact on the yield of pyrogallol
4, Table 1). Synthesis of pyrogallol by way of myo-2-inosose (9)
required only 4 enzyme-catalyzed steps and 2 chemical steps
Scheme 1). By comparison, synthesis of pyrogallol (4) by way of
gallic acid (10) and the shikimate pathway (Scheme 1) required at
2 3
O led to a 44% yield of pyrogallol (4, Table 1). Changing the
(
(
1d
least 20 enzyme-catalyzed steps. Phytic acid (8b, Scheme 1),
which constitutes about 8% of the dry weight of corn steeping
liquor, is an abundant alternative source of myo-inositol.4
Established methods are available for the isolation of phytic acid
a,b
a
Key: (a) E. coli JWF1/pAD1.88A, 11%, see ref 3; (b) G. oxydans
ATCC 621, 95%, see ref 3; (c) H2SO4, H2O, reflux, 66%, see ref 3; (d)
44%, see Table 1; (e) phytase; (f) E. coli KL7/pSK6.161, 12%, see ref 1d;
g) E. coli RB791 serA::aroB/pSK6.234, 93%, see ref 1d.
(
8b) and its dephosphorylation.4c
(
A more direct derivation of an inosose is possible using 2-deoxy-
scyllo-inosose synthase, which catalyzes the conversion of glucose
6-phosphate (7b) into 2-deoxy-scyllo-inosose (11, Scheme 2).
*
Corresponding author. E-mail: frostjw@cem.msu.edu.
Reaction of inosose 11 with Ac O under acidic conditions produces
2
5
926 J. AM. CHEM. SOC. 2002, 124, 5926-5927
9
10.1021/ja0176346 CCC: $22.00 © 2002 American Chemical Society

C O MMU N I C A T I O N S
Scheme 2 a
3). Reaction of triacetic acid lactone methyl ether (13b) as a melt
in Na and MeOH gave an excellent yield of phloroglucinol methyl
ether (1b, Scheme 3).7 To circumvent the toxicity of (CH
a
O)
3
2
-
2
SO , triacetic acid lactone (13a) was converted to its methyl ether
1
3b by using (CH
3
O)
3
PO or by refluxing in MeOH with Dowex
+
50 (H ) (Scheme 3). Hydrolysis of the phloroglucinol methyl ether
(
1b) afforded phloroglucinol (1a) along with substantial amounts
7b
of 2,4,6,3′,5′-pentahydroxybiphenyl (15, Scheme 3). Fortuitously,
hydrogenation of phloroglucinol methyl ether (1b) followed by acid-
catalyzed dehydration of the intermediate dihydrophloroglucinol
proved to be highly regioselective with formation of resorcinol (2)
in 80% yield (Scheme 3). The syntheses of phloroglucinol (1) and
resorcinol (2) from glucose contrast with the synthesis (Scheme 3)
8
of these aromatics from toluene-derived TNT (16) and benzene-
derived 1,3-diisopropylbenzene (17), respectively.
a
9
Key: (a) i. hexokinase, ii. 2-deoxy-scyllo-inosose synthase, 38%, see
ref 5a; (b) 0.5 M H3PO4, reflux, 39%; (c) 53%, see Table 1; (d) E. coli
QP1.1/pKD12.138, 20% (from glucose), see ref 1b; (e) i. NaOCl, ii. reflux,
New synthetic connections have been made between glucose and
aromatic chemicals by recruiting biosynthetic pathways other than
the shikimate pathway. To fully exploit these syntheses in the future,
a microbe that directly synthesizes myo-2-inosose from glucose (7a)
or phytic acid (8b) remains to be constructed. A microbe capable
of synthesizing 2-deoxy-scyllo-inosose (11) from glucose is also
needed5 along with a microbe that synthesizes triacetic acid lactone
8
7%, see ref 1b.
Scheme 3 a
a
(13a) in high yield. Compelling reasons for surmounting these
biocatalytic challenges can be found in the drastic reductions in
the numbers of enzymes required to synthesize pyrogallol and
hydroquinone. The access gained to resorcinol is comparably
important. With polyhydroxybenzene deoxygenation, benzene-free
syntheses have now been established for all of the “building block”
1
a
1c,5a
hydroxybenzenes including phenol, resorcinol, catechol,
hydroquinone,1b pyrogallol, hydroxyhydroquinone, and phloro-
1d
glucinol.
Acknowledgment. Research was supported by the National
Science Foundation.
Supporting Information Available: Syntheses of 1a,b, 2, 4, 5, 6,
and 13b (PDF). This material is available free of charge via the Internet
at http://pubs.acs.org.
a
Key: (a) see ref 6; (b) (CH3O)2SO2, K2CO3, acetone, reflux, 85%; or
+
References
(
CH3O)3PO (neat), K2CO3, 79%; or Dowex 50 (H ), MeOH 43%; (c) Na,
MeOH, 185 °C, 85%; (d) 12 N HCl, 56% 1a, 15% 15; (e) 80%, see Table
(
1) (a) Gibson, J. M.; Thomas, P. S.; Thomas, J. D.; Barker, J. L.; Chandran,
1
.
S. S.; Harrup, M. K.; Draths, K. M.; Frost, J. W. Angew. Chem., Int. Ed.
2001, 40, 1945. (b) Ran, N.; Knop, D. R.; Draths, K. M.; Frost, J. W. J.
5
a 29% yield of hydroxyhydroquinone triacetate. We find that
Am. Chem. Soc. 2001, 123, 10927. (c) Draths, K. M.; Frost, J. W. J. Am.
Chem. Soc. 1995, 117, 2395. (d) Kambourakis, S.; Draths, K. M.; Frost,
J. W. J. Am. Chem. Soc. 2000, 122, 9042.
hydroxyhydroquinone (5) can be directly obtained (Scheme 2) by
acid-catalyzed dehydration of an aqueous solution of chemically
(2) (a) Fray, G. I. Tetrahedron 1958, 3, 316. (b) Smissman, E. E.; Voldeng,
synthesized5b 2-deoxy-scyllo-inosose (11). Hydrogenation of hy-
A. N.; McCarthy, J. F. J. Pharm. Sci. 1966, 55, 1101.
(3) Hansen, C. A.; Dean, A. B.; Draths, K. M.; Frost, J. W. J. Am. Chem.
Soc. 1999, 121, 3799.
2 3
droxyhydroquinone with Rh on Al O (Table 1) produced hydro-
(
4) (a) Sands, S. H.; Biskobing, S. J.; Olson, R. M. In Phytic Acid: Chemistry
and Applications; Graf, E., Ed.; Pilatus Press: Minneapolis, MN, 1986;
p 119. (b) Hull, S. R.; Montgomery, R. J. Agr. Food Chem. 1995, 43,
1516. (c) Wodzinski, R. J.; Ullah, A. H. J. AdV. Appl. Microbiol. 1996,
quinone (6) in 53% yield. While similar yields of 6 were obtained
with Rh and Pt as the hydrogenation catalysts, a substantially lower
yield was realized with Pd (Table 1). The 2 enzyme-catalyzed steps
and 2 chemical steps required for the conversion of glucose into
hydroquinone via 2-deoxy-scyllo-inosose (11) contrast with the 18
enzyme-catalyzed steps and 1 chemical step required for the
synthesis of hydroquinone (6) from glucose (7a) via quinic acid
42, 263.
(
5) (a) Kakinuma, K.; Nango, E.; Kudo, F.; Matsushima, Y.; Eguchi, T.
Tetrahedron Lett. 2000, 41, 1935. (b) Yamauchi, N.; Kakinuma, K. J.
Antibiot. 1992, 45, 756.
(
6) Richardson, M. T.; Pohl, N. L.; Kealey, J. T.; Khosla, C. Metab. Eng.
1999, 1, 180.
1
b
(7) (a) Huber, U. U.S. Patent 4,112,003, 1978. (b) McKillop, A.; Howarth,
(
12) and the shikimate pathway (Scheme 2).
B. D.; Kobylecki, R. J. Synth. Commun. 1974, 4, 35.
The report6 of microbe-catalyzed synthesis of triacetic acid
lactone (13a) and the reported conversion of 13b into phloroglu-
cinol (1a, Scheme 3) allowed advantage to be taken of the highest
(
8) Leston, G. In Kirk-Othmer Encyclopedia of Chemical Technology;
7
Kroschwitz, J. I., Howe-Grant, M., Eds.; Wiley: New York, 1996; Vol.
19, p 778.
(
9) Krumenacker, L.; Costantini, M.; Pontal, R.; Sentenac, J. In Kirk-Othmer
Encyclopedia of Chemical Technology; Kroschwitz, J. I., Howe-Grant,
M., Eds.; Wiley: New York, 1992; Vol. 13, p 996.
yielding deoxygenation of a polyhydroxybenzene (Table 1).
Literature-based7a alkylation of triacetic acid lactone (13a) by
(
CH
3
O)
2
SO
2
afforded the corresponding methyl ether (13b) (Scheme
JA0176346
J. AM. CHEM. SOC.
9
VOL. 124, NO. 21, 2002 5927