TITLE: Efficient Synthesis of the gallotannin penta-Ogalloyl--D-glucopyranose (-PGG)
AUTHOR NAMES Klaus Himmeldirk*,†, Yulin Ren‡, Xiaozhuo Chen‡
AUTHOR ADDRESS
Department of Chemistry and Biochemistry and Edison Biotechnology Institute, Ohio University,
Athens, Ohio 45701
AUTHOR EMAIL ADDRESS
Klaus.Himmeldirk.1@ohio.edu
RECEIVED DATE
CORRESPONDING AUTHOR FOOTNOTE
† Department
‡ Edison
of Chemistry and Biochemistry.
Biotechnology Institute.
ABSTRACT (Word Style “BD_Abstract”). All manuscripts must be accompanied by an abstract. The
abstract should briefly state the problem or purpose of the research, indicate the theoretical or
experimental plan used, summarize the principal findings, and point out major conclusions. The optimal
length is one paragraph.
MANUSCRIPT TEXT (Word Style “TA_Main_Text”).
Step 1: Methyl 3,4,5-tribenzyloxybenzoate = Methyl 3,4,5-tris(phenylmethoxy)benzoate (CA Reg.No.
70424-94-1)
1
HO
BnO
O
HO
O
Acetone
+
BnO
+ KI + K2CO3
OCH3
Cl
OCH3
HO
BnO
Literature:
Journal of Organic Chemistry 1991, 56, 7092-7097
Organic Process Research & Development 2001, 5, 54-60
Procedure: Adaptation of the literature procedures
A mixture of methyl 3,4,5-trihydroxybenzoate (10 g, 54.3 mmol), potassium iodide (4 g, 24 mmol), and
anhydrous powdered potassium carbonate (44 g, 318 mmol) in acetone (500 mL) was stirred at room
temperature for 20 min. Benzyl chloride (22 g, 174 mmol) dissolved in 100 mL acetone was added. The
suspension was refluxed for 18 h. At this point TLC shows full conversion of the starting material and a
single product spot at Rf = 0.30 (hexanes : ethyl acetate = 8 : 2, UV detection). The solid was filtered off
and the filtrate was evaporated. The residue was taken up in 400 mL dichloromethane. The suspension
was filtered through celite and the filtrate was evaporated. The residue was dried at room temperature in
an oil pump vacuum for 45 min.
Yield: 26.528 g (107%) (the product contains some benzyl chloride as an impurity)
Step 2: 3,4,5-Tribenzyloxybenzoic acid = 3,4,5-tris(phenylmethoxy)benzoic acid (CA Reg.No. 148648-2)
BnO
BnO
O
BnO
O
Ethanol
+
NaOH
BnO
OCH3
BnO
OH
BnO
Literature:
Organic Process Research & Development 2001, 5, 54-60
2
Procedure: Adaptation of the literature procedure
The ester (26.52 g, ~54.3 mmol, crude product of the previous step) was suspended in 95% ethanol (500
mL) and sodium hydroxide pellets (3.54 g, 88.5 mmol) were added. The mixture was heated under
reflux for 2 h. The hot solution was poured into a mix of concentrated hydrochloric acid (25 mL) and
water (500 mL). A thick, voluminous suspension formed. It was swirled for 10 min before the solid was
filtered off. The product was washed successively with a 1 : 1 mix of 95% ethanol and water (100 mL),
pure water (100 mL), 95% ethanol (100 mL), methanol (2 x 50 mL), and tert-butyl methyl ether (2 x 50
mL). The white solid was dried overnight at room temperature in an oil pump vacuum.
Yield: 22.63 g (94.6% over 2 steps)
Mp
191 - 195C
Step 3: 3,4,5-Tris(phenylmethoxy)benzoyl chloride (CA Reg.No. 1486-47-1)
O
O
O
BnO
OH
+
BnO
Cl
Cl
O
OBn
Toluene,
DMF
BnO
Cl
BnO
OBn
Assemble a 250 mL flask equipped with a magnetic stirring bar, a dropping funnel with pressure
equilibration, and a ”bubbler” on top of the dropping funnel. The flask is charged with 11.57 g of the
acid (26.26 mmol), 120 mL of toluene, and 0.2 mL DMF. A “thick” suspension forms that is hard to stir
magnetically. Via the dropping funnel, add a solution of 5 g oxalyl chloride (39.4 mmol) and 17 mL
toluene over a period of 10 minutes. After stirring at room temperature for an additional 20 min, the
mixture is heated to 50ºC. The ceasing gas evolution indicates the end of the reaction after about 1 hour.
A yellow solution has formed. The solvent is evaporated on a rotary evaporator. The residue is taken up
in 50 ml of toluene at 70ºC. A yellow, highly viscous,
sticky oil remains undissolved. The product solution is decanted. The chloride is precipitated by adding
55 mL of hexanes to the hot toluene solution. Crystallization starts within a minute. The reaction
mixture turns into a semisolid mass that is very difficult to stir. The white solid is filtered off, pressed
dry, washed with hexanes, and dried.
Yield: 9.46 g (78%)
M.p.: 118.1 – 120.5ºC
Remarks:
 A large scale set-up will need mechanical stirring.
 Adding more hexanes to the filtrate and collecting the precipitate that forms could
improve the yield.
3
-D-Glucopyranose pentakis [3,4,5-tris(phenylmethoxy)benzoate]
Step 4:
(CA Reg.No. 70424-95-2)
-D-Glucopyranose pentakis [3,4,5-tris(phenylmethoxy)benzoate]
(CA Reg.No. 122625-60-9)
Penta-O-(3,4,5-tri-O-benzylgalloyl)-D-glucopyranose
OH
O
O
HO
HO
GBn3
BnO
+
OH
OH
BnO
Cl
BnO
DMAP
CH3CN
O
Bn3G
Bn3G
GBn3 GBn3
 :  = 96 : 4
The acid chloride (459 mg, 1.0 mmol) and finely powdered D-glucose (36 mg, 0.2 mmol) are suspended
in 10 mL acetonitrile at room temperature. DMAP (128 mg, 1.05 mmol) is added last, and the mixture is
stirred at room temperature. The chloride dissolves within 5 to 10 minutes after the addition of DMAP.
After 18 hrs at room temperature, the solvent is evaporated. The residue is taken up in 5 mL of toluene
at 60ºC. After cooling to room temperature, the solution is filtered through a layer of silica gel (250 mg,
1.2 cm thick). The silica gel is washed with 5 mL of a mix of toluene and ethyl acetate (100 : 4) to elute
any product that may stick to the silica gel. The solvent is evaporated. A highly viscous oil is obtained
after drying in an oil pump vacuum for 5 hrs.
Yield: 455 mg (99%)
of the -isomer in CDCl3: 7.18-7.56 (68 H, m), 6.89 (1H, d, 3.0Hz), 6.37 (1H, t, 10.0 Hz),
5.74 (1H, t, 10.0 Hz), 5.64 (1H, dd, 3.0Hz, 10.0Hz), 4.88-5.24 (24H, m), 4.52 (1H, ddd, 2.0Hz, 4.5Hz,
10.0Hz), 3.84 (1H, dd, 2.0Hz, 12.5Hz), 3.75 (1H, dd, 4.5Hz, 12.5 Hz).
1H-NMR
of the -isomer in CDCl3: 7.18-7.51 (68 H, m), 6.23 (1H, d, 8.0Hz), 6.03 (1H, t, 10.0 Hz),
5.83 (1H, dd, 8.0Hz, 10.0 Hz), 5.72 (1H, t, 10.0Hz), 4.96-5.21 (24H, m), 4.31 (1H, ddd, 2Hz, 5.5Hz,
10Hz), 3.88 (1H, dd, 2Hz, 12.5Hz), 3.74 (1H, dd, 5.5Hz, 12.5 Hz).
1H-NMR
Remarks:
 The best ratio of reagents between chloride and glucose was found to be 5 : 1. If we use
a higher excess of the chloride, the anhydride of gallic acid is formed. (The anhydride
can be difficult to remove from the reaction mixture.) The : ratio is not influenced by
the ratio of the reagents.
Step 5: -D-Glucopyranose pentakis(3,4,5-trihydroxybenzoate) (CA Reg.No. 70470-10-9)
-D-Glucopyranose pentakis(3,4,5-trihydroxybenzoate)
(CA Reg.No. 14937-32-7)
Penta-O-galloyl--D-glucopyranose and Penta-O-galloyl--D-glucopyranose
4
GBn3
Bn3G
Bn3G
OG
O
THF
+ Pd/C + H2
GO
GO
GBn3 GBn3
O
OG
OG
The benzyl protected starting material (455 mg, 0.27 mmol) is dissolved in dry THF (20 mL). 10%
Palladium on charcoal (100 mg, 0.094 mmol) is added. The suspension is then stirred at high speed at
40C under a hydrogen gas atmosphere at normal pressure for 18 hrs. The mixture is cooled, filtered
through Celite, and the filtrate is evaporated. The residue is taken up in 5 mL water (it may take about 5
min until all material is dissolved). The mixture is evaporated at a temperature < 40ºC to remove all
residual organic solvent. The residue is dried in a oil pump vacuum at room temperature for 4 hrs.
A yield of 166 mg (88% over two steps) of crude -PGG is obtained.
HPLC analysis of the product shows a 96 : 4 ratio ( : ), and a purity of about 88% for -PGG.
Beckman Ultrasphere C-18 reversed phase column (4.6mm×25cm, 5µm). The flow rate is 1ml/min. The
detection wavelength is set at 260nm, and eluent A is water with 0.1% trifluoroacetic acid, eluent B is
acetonitrile with 0.1% trifluoroacetic acid.
1mg PGG sample is dissolved in 1ml water. The sample is filtered (0.45μm) before injection. Injection
volume is 5µl.
Method 18 1 50 5
#
Time (min)
Module
Function
Value
Duration
(min)
1
0.00
Pump
%B
18.00
1.00
2
0.00
Pump
Flow rate
1.00
0.00
3
0.00
Det 168-4
Autozero
4
0.00
Det 168-4
Relay on
2
0.01
5
0.00
Injector 508
Alarm
6
5.00
Pump
%B
18.00
15.00
7
25.00
Pump
%B
50.00
3.00
8
33.00
Pump
%B
18.00
2.00
9
35.00
Pump
Flow rate
1.00
5.00
10
40.00
Det 168-4
Stop data
11
0
0.10
NMR:
5
1H-NMR
(500 MHz, acetone-d6) δ 3.79 (dd, J = 12.5, 5.0 Hz, 1H), 3.87 (dd, J =, 12.5, 2.0Hz, 1H), 4.62
(ddd, J = 10.0, 5.0, 2.5 Hz, 1H), 5.45 (dd, J = 10.0, 4.0 Hz, 1H), 5.63 (t, J = 10.0 Hz, 1H), 6.12 (t, J =
10.0 Hz, 1H), 6.72 (d, J = 4.0Hz, 1H), 6.98 (s, 2H), 6.99 (s, 2H), 7.07 (s, 2H), one missing 7.26 (s, 2H),
7.8-8.5 (br s, 12H).
of the -isomer in acetone-d6: 7.9-8.5 (12H, broad m), 7.11 (2H, s), 7.04 (2H, s), 6.98 (2H, s),
6.95 (2H, s), 6.31 (1H, d, 8.0Hz), 5.96 (1H, t, 10.0 Hz), 5.57 (1H, dd, 8.0Hz, 10.0 Hz), 5.53 (1H, t,
10.0Hz), 4.50 (1H, ddd, 2.5Hz, 5.5Hz, 10.0Hz), 3.88 (1H, dd, 2.5Hz, 12.5Hz), 3.78 (1H, dd, 5.5Hz, 12.5
Hz).
1H-NMR
Recrystallization:
1 g of crude -PGG (purity 88 – 90%) is dissolved in 20 mL of water by heating the mixture for 5 min
at 70ºC. (It is important to keep the time of heating short since PGG decomposes at elevated temperature
in water). Remove undissolved particles (if necessary) by filtration. Leave the solution at room
temperature for 5 days (if you added seed crystals), or 7 days if you worked without seed crystals. Filter
off the product and repeat the recrystallization two more times.
1 g of starting material yields 0.5 to 0.55 g of pure -PGG (HPLC > 98%) after three recrystallizations.
Remarks:
 The second evaporation is crucial because it removes most of the remaining solvent
(THF and Toluene) that “sticks” to the product after the first evaporation. If too much
organic solvent remains, the following recrystallization might not work.
 We are using double distilled water with a pH of around 5.5 in our lab.
 At 6ºC (fridge) no crystals are obtained. Instead, a sticky amorphous oil forms.
 The dissolution of the starting material for the first recrystallization can be done at
room temperature. Since the product of the reaction is an amorphous material, it easily
dissolves in water at room temperature.
 The crystals are forming very slowly. They are extremely fine needles that look like
bundles of hair under the microscope.
 A test solution that was 73.5% pure (HPLC UV detector, 260 nm) in -PGG, showed
only a purity of 44.8% after keeping the solution (aqueous buffer, pH 5.5) for 1 day at
80ºC. Beside a number of unknown decomposition products that occur in small
quantities, gallic acid could be identified as the main side product (14.5%) after
heating.
On a large scale the evaporation of water with the help of a lyophilizer might be better
than the use of a rotary evaporator. It would avoid the prolonged heating of -PGG in
an aqueous solvent
6
ACKNOWLEDGMENT
We thank Ohio University (Grant No. OURC 2002-42) and MetaCor Pharmaceuticals USA, Inc. for
financial support of this research.
SUPPORTING INFORMATION PARAGRAPH (Word Style “TE_Supporting_Information”). A brief
statement in nonsentence format listing the contents of material placed here should be included at the
end of the manuscript as a separate section titled Supporting Information Available. The contents of
Supporting Information may include the following: (1) large tables, (2) extensive figures, (3) lengthy
experimental procedures, (4) mathematical derivations, (5) analytical and spectral characterization data,
(6) molecular modeling coordinates, (7) modeling programs, (8) crystallographic information files
(CIF), (9) instrument and circuit diagrams, (10) and expanded discussions of peripheral findings. For
complete instructions on how to prepare this material for publication, check the Guide, Notes, Notice, or
Instructions for Authors that appear in each publication’s first issue of the year and on the World Wide
Web at http://pubs.acs.org.
FIGURE CAPTIONS (Word Style “VA_Figure_Caption”). Each figure must have a caption that
includes the figure number and a brief description, preferably one or two sentences. The caption should
immediately follow the figure with the format “Figure X. Figure caption.”. All figures must be
mentioned in the text consecutively and numbered with Arabic numerals. The caption should be
understandable without reference to the text. It is preferable to place the keys to symbols used in the
figure in the caption, not in the artwork. Ensure that the symbols and abbreviations in the caption agree
with those in the figure itself and in the text and that the figure is already sized appropriately.
SCHEME TITLES (Word Style “VC_Scheme_Title”). Chemical reactions and flow diagrams may be
called schemes. Schemes may have brief titles describing their contents. The artwork for each scheme
should immediately follow the scheme title. The title should follow the format “Scheme X. Scheme
7
Title”. All schemes must be mentioned in the text consecutively and numbered with Arabic numerals.
Schemes may also have footnotes (use Word Style “FD_Scheme_Footnote”), inserted after the artwork.
CHART TITLES (Word Style “VB_Chart_Title”). Lists of structures may be called charts. Charts may
have brief titles describing their contents. The title should follow the format “Chart X. Chart Title”.
Charts may also have footnotes (use Word Style “FC_Chart_Footnote”). To insert the chart into the
template, be sure it is already sized appropriately and paste it immediately after the chart title.
TABLES.
Table 1: Approximate /-ratio of the products in acylation reactions of glucose (Glc).
/-ratio for
/-ratio for
Glc + Acid
chloride + DMAP
Glc + Acid
anhydride + DMAP
3.5
2
Ethyl Acetate
6
1.8
Dichloromethane
10
0.8
DMF
10
7
11.5
2
THF
15
-
1,4-Dioxane
26
2
Acetonitrile
>30
1.6
Solvent
Pyridine
Acetone
Each table must have a brief (one phrase or sentence) title that describes its contents. The title should
follow the format “Table X. Table Title” (Word Style “VD_Table_Title”). The title should be
understandable without reference to the text. Put details in footnotes, not in the title (use Word Style
“FE_Table_Footnote”). Define nonstandard abbreviations in footnotes.
8
Use tables (Word Style “TC_Table_Body”) when the data cannot be presented clearly as narrative, when
many precise numbers must be presented, or when more meaningful interrelationships can be conveyed
by the tabular format. Do not use Word Style “TC_Table_Body” for tables containing artwork.
Tables should supplement, not duplicate, text and figures. Tables should be simple and concise. It is
preferable to use the Table Tool in your word-processing package, placing one entry per cell, to generate
tables.
REFERENCES (Word Style “TF_References_Section”). In many journals, references are placed at the
end of the article, while in others they are treated as footnotes. In any case, place your list of references
at the end of the manuscript. The appropriate placement will be made as part of the editorial process. In
ACS publications, references are cited in three ways: superscript numbers, italic numbers on the line and
in parentheses, and by author name and year of publication in parentheses inside the punctuation. Please
use the appropriate style for the journal for which you are submitting your manuscript. Authors are
responsible for the accuracy and completeness of all references. Authors should check all parts of each
reference listing against the original document. Detailed information on reference style can be found in
The ACS Style Guide, 2nd ed., available from Oxford Press.
SYNOPSIS TOC (Word Style “SN_Synopsis_TOC”). If you are submitting your paper to a journal that
requires a synopsis graphic and/or synopsis paragraph, see the Guide, Notes, Notice, or Instructions for
Authors that appear in each publication’s first issue of the year and the journal’s homepage for a
description of what needs to be provided and for the size requirements of the artwork.
9