A biomimetic synthesis of (-)-ascorbyl phloroglucinol and studies toward the construction of ascorbyl-modified catechin natural products and analogues

Article abstract of DOI:10.1016/j.tet.2011.09.102

A method for appending the ascorbyl moiety onto the framework of phenolic natural products has been developed. This reaction proceeds in two steps from l-ascorbic acid and employs acetic acid catalysis. Excellent stereoselectivity is observed during C-C b

Full text of DOI:10.1016/j.tet.2011.09.102

Tetrahedron 67 (2011) 9265e9272
Contents lists available at SciVerse ScienceDirect
Tetrahedron
journal homepage: www.elsevier.com/locate/tet
A biomimetic synthesis of (ꢀ)-ascorbyl phloroglucinol and studies toward the
construction of ascorbyl-modified catechin natural products and analogues
*
Sneha A. Belapure, Zachary G. Beamer, John E. Bartmess, Shawn R. Campagna
Department of Chemistry, University of Tennessee, Knoxville, TN 37996-1600, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 3 August 2011
Received in revised form 20 September 2011
Accepted 21 September 2011
Available online 28 September 2011
A method for appending the ascorbyl moiety onto the framework of phenolic natural products has been
developed. This reaction proceeds in two steps from L-ascorbic acid and employs acetic acid catalysis.
Excellent stereoselectivity is observed during CeC bond formation between the phenolic compound and
dehydroascorbic acid, and the process is also chemoselective for phenol derivatives bearing electron-
donating substituents in each of the 1, 3, and 5 positions. Further, good regioselectivity was also ob-
served when phenols lacking an axis of C2 symmetry were employed. This method has led to the syn-
thesis of (ꢀ)-ascorbyl phloroglucinol as well as the tetracyclic core of ascorbyl-modified catechin natural
products.
Keywords:
(ꢀ)-Ascorbyl phloroglucinol
Ascorbyl epigallocatechin
Ascorbylation reaction
Stereoselective reaction
Chemoselective reaction
Ó 2011 Elsevier Ltd. All rights reserved.
1. Introduction
Polyphenols produced by tea plants are well known to have
a range of interesting and medicinally relevant bioactivities.
Members of this class of molecules have been reported to be potent
antiobesigens,1 antivirals,2 antioxidants,3e5 and anticarcinogens,6
and also display other desirable properties.6e9 Many of these ac-
tivities have been ascribed to catechin-derived molecules, and
a series of ascorbyl-modified catechins have been reported. These
natural products display a range of complexity from the simple
ascorbylated phloroglucinol, (ꢀ)-ascorbyl phloroglucinol (1a), to
more complex ascorbylated catechins (Fig. 1).10 While the full
spectrum of biological activities displayed by molecules of this class
have not been fully explored, 8-C-ascorbyl-(ꢀ)-epigallocatechin 2a
has been shown to be effective in stopping the replication of HIV
in vivo with an EC50 of 8.33
in vitro with an IC50 of 0.646
m
M and for inhibiting pancreatic lipases
M.
m
(ꢀ)-Ascorbyl phloroglucinol 1a has previously only been iso-
lated as the corresponding peracetate derivative 3. Prior work has
also not been able to unambiguously assign the stereochemical
configuration at the tertiary and hemiacetal carbons (C20 and C30)
for any ascorbylated phenols, although modeling has suggested
Fig. 1. Natural product structures.
that the SSRS configuration for C20e50, respectively, is the most
stable for 3.11 These intriguing biological data coupled with the
interesting structures and open questions concerning the stability
and stereochemistry of these molecules led us to pursue the de-
velopment of a synthetic strategy to construct the tricyclic core of
this set of related compounds.
* Corresponding author. Tel.: þ1 865 974 7337; fax: þ1 865 974 9332; e-mail
address: campagna@ion.chem.utk.edu (S.R. Campagna).
0040-4020/$ e see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2011.09.102

9266
S.A. Belapure et al. / Tetrahedron 67 (2011) 9265e9272
2. Results and discussion
reactivity of several hydroxyarenes with 5 using AcOH as both the
solvent and the catalyst at 60 ꢁC (Table 1). As expected, the reaction
rate rose with increasing electron density of the arene. Nucleo-
philes containing methyl ethers were also able to be successfully
used in this reaction. Further, a high yield of product was obtained
even when a 1:1 ratio of electrophile and nucleophile was used.
2.1. Retrosynthetic analysis and design rationale
We reasoned that a late-stage introduction of the ascorbyl
moiety onto the trioxyarene would be desirable for the construc-
tion of these natural products and analogous structures. After ob-
serving the structural similarity for the ascorbylated natural
products, it became obvious that all of these compounds could be
constructed by a similar route in which the CeC bond between the
ascorbyl group and the arene were formed if a suitable electrophilic
version of ascorbic acid was used (Scheme 1). The natural nucleo-
philicity of the arene would then provide the other reacting partner.
The biosynthesis of these molecules has been proposed to proceed
via a similar strategy.12
Table 1
Reaction of 5 with hydroxyarenesa
Entry
Arene
Conditions
Temp (ꢁC)
t (h)
Yield (%)
1
2
3
4
6a
6b
6c
6d
AcOH (neat)
AcOH (neat)
AcOH (neat)
AcOH (neat)
60
60
60
60
2
1.5
2
85
86
80
68
3
a
The product structures denote relative, not absolute, stereochemistry.
2.3. Optimization of conditions for the addition of 6a to 4
Scheme 1. Retrosynthetic analysis.
Dehydroascorbic acid 4 can be easily generated by the oxidation
of ascorbic acid, and this molecule has the desired electrophilic
character.13 This is evidenced by the fact that the carbonyl groups at
both C2 and C3 are almost exclusively found to be the hemiacetal or
hydrate, respectively (Scheme 2). The observation that the bicyclic
hemiacetal predominates also led to the hypothesis that the CeC
bond would be formed preferentially from the opposite side of the
dihydroxyethyl group of 4 yielding products with the relative ste-
reochemistry shown in Scheme 1. Similar stereoselection has been
Once the reactivity of the arenes toward addition to 5 had been
determined, attention was placed on finding the optimal acid cat-
alyst for the reaction of 1 equiv phloroglucinol with 1 equiv of
dehydroascorbic acid, 4, to give (ꢀ)-ascorbyl phloroglucinol 1a.
Initial attempts to perform this transformation in neat AcOH at
elevated temperature led to less than desirable yields, and de-
composition products were observed. Therefore, this reaction was
then attempted using 0.1 equiv of a Lewis acid, either AlCl3 or
BF3$OEt2, as the catalyst in THF. No conversion was observed under
these reaction conditions, even at elevated temperature (Table 2
entries 1 and 2). Fortunately, 1a could be obtained in moderate
yield when AcOH was used as a co-solvent in THF at either 65 ꢁC for
6 h or at room temperature (rt) for 12 h, although decomposition
was still observed at elevated temperature (Table 2 entries 3 and 4).
Following from these results, we then found that using a 0.5 equiv
excess of dehydroascorbic acid lowered the yield; however, use of
a 0.5 equiv excess of arene gave the desired product in 69% yield
based on 4 (Table 2 entries 5 and 6). Further, the arene was re-
coverable in these reactions.
observed in the reduction of
L-ascorbic acid to the corresponding L-
gulonic acid
g
-lactone.14
Scheme 2. Reactivity of 4.
Table 2
Reaction of dehydroascorbic acid with hydroxyarenes
2.2. Reaction of hydroxyarenes with indane 1,2,3-trione
The base-catalyzed addition of phloroglucinol 6a to 4 has been
reported, although the stereochemistry of the product remained
ambiguous.12 In our hands, we were unable to obtain good results
for the reaction of 6a with 4 under basic conditions using NaHCO3,
NEt3, or K2CO3. Due to this, we chose to investigate the use of acid
catalysis for the addition of hydroxyarenes to 4. The first step in this
process was to determine the relative reactivity of mono-, 1,3-di-,
and 1,3,5-trihydroxyarenes [phenol (6d), resorcinol (6c), and
phloroglucinol (6a)] with a model 1,2,3-triketone (indane 1,2,3-
trione 5). These reactions have been reported using acetic acid
(AcOH) as a catalyst, although a consistent set of conditions had not
been evaluated for each molecule.15,16 To this end, we explored the
Entry Arene
4
Conditions
Temp (ꢁC) t (h) Yield (%)
(equiv) (equiv)
1
2
3
4
5
6
1
1
1
1
1
1.5
1
1
1
1
1.5
1
AlCl3 (0.1 equiv), THF
BF3$OEt2 (0.1 equiv), THF Reflux
AcOH/THF (2:5 v/v)
AcOH/THF (2:5 v/v)
AcOH/THF (2:5 v/v)
AcOH/THF (2:5 v/v)
Reflux
6
6
NR
NR
42a
40
Reflux
6
rt
rt
rt
12
12
12
26b
69b
a
Decomposition was observed.
Yield based on 4. The arene was recoverable in all cases.
b

S.A. Belapure et al. / Tetrahedron 67 (2011) 9265e9272
9267
2.4. Probing the arene substrate scope
With optimized conditions for the synthesis of naturally oc-
curring 1a in hand, we set out to probe the range of hydroxyarenes
that could be employed to synthesize other ascorbylated natural
products and analogues. On one hand, a high reactivity of dehy-
droascorbic acid with hydroxyarenes is desirable for the construc-
tion of analogues; however, a lack of reactivity toward the
pyrogallyl (1,2,3-trihydroxybenzyl) and gallyl (3,4,5-trihydrox-
ybenzoyl) moieties is desirable for the chemoselective synthesis of
natural products, such as 2a,b, if strategies employing protecting
groups are to be avoided.
To begin this investigation, a series of phenol, resorcinol, and
phloroglucinol and their derivatives were reacted with dehy-
droascorbic acid 4 under the optimized conditions with AcOH as
the catalyst (Table 3). Although arenes with a range of electron
donating and withdrawing groups were studied, it was found that
three electron donating groups were necessary for reaction to
proceed with this substrate series (Table 3 entry 1). This is in
contrast to the reactions with 5 where even phenol was capable of
acting as a suitable nucleophile (Table 1). It was also noted that
arenes containing benzyl ethers decomposed under the reaction
conditions, and this has implications for further syntheses as dis-
cussed below in Section 2.7.
Table 3
Scheme 3. Chemoselectivity for reaction with 4.
Reaction of 4 with hydroxyarenes
was found to be amenable to standard purification techniques,
including flash chromatography on silica, as long as alcohols were
not used as the solvent in the presence of an acid source. If MeOH
was used to elute the column, the presence of an extra methoxy
group could be observed in the 1H NMR spectrum, presumably
arising from the formation of an acetal at C30 (data not shown). To
confirm that the material obtained from the reaction of 4 with
phloroglucinol was the structurally identical to naturally occurring
(ꢀ)-ascorbyl phloroglucinol, product 1a was peracetylated and
both the 1H and 13C NMR spectra were compared to those reported
for the derivatized natural product, and it was found that all res-
onances matched the reported values.11
Entry
Arene
X
Y
Yield (%)
1
2
3
4
5
6
7
8
6b
6c
6d
6e
6f
6g
6h
6i
OMe
H
H
H
H
H
OBn
OH
OMe
OH
H
OBn
OAc
OMe
OBn
OAc
72a
NR
NR
NR
NR
NR
0b
NR
The reactions leading to both (ꢀ)-ascorbyl phloroglucinol 1a
and the methylated analogue 1b proceeded to yield a single
product, although the exact stereochemical configurations at C20
and 30 were still ambiguous. The stereochemistry for carbons 20, 30,
40, and 50 had previously been hypothesized to be SSRS based on
NOE data and structural modeling of 3 using the SYBYL software
package.11 To obtain more robust evidence for the stereochemistry,
the crystallization of both 1a and 1b was attempted. These efforts
were unsuccessful for 1a; however, upon standing in hexanes/ethyl
acetate colorless needle-shaped crystals of 1b were obtained. The
configuration of 1b was determined to be SRRS from crystallo-
graphic data (Fig. 2). Based on these data and precedent from the
a
Yield based on 4. The arene was recoverable.
The benzylated arene decomposed.
b
Once it was determined that the arene substrate scope was
limited for the series of phenol, resorcinol, and phloroglucinol de-
rived molecules, the reactivities of 1,3,5-trimethoxybenzene (8),
pyrogallol (9), and methyl gallate (10) with 4 were explored. None
of these arenes displayed any reactivity toward AcOH catalyzed
addition to dehydroascorbic acid 4 (Scheme 3). The data obtained
from the reactions using 9 and 10 further indicate that the elec-
tronic properties of the arene are critical for successful reaction. It is
likely that three electron donating groups which all direct addition
to the same position on the arene ring are needed and that the
presence of any electron withdrawing group hinders reaction.
These conclusions are further supported by the data obtained using
the bicyclic analogues discussed in Section 2.7. The lack of reactivity
between 4 and 8 is also suggestive that hemiacetal formation is
necessary to bring the reactants into close proximity before CeC
bond formation can occur.
reduction of ascorbic acid to gulonic acid-g-lactone, we hypothe-
size that the addition of the arene to the side of 4 opposite the
dihydroxyethyl side chain will be a general feature of these
reactions.
2.6. Synthesis of the ascorbylcatechin tetracyclic core
Once the substrate scope and stereochemical outcome had been
determined for the reactions leading to tricyclic products, such as
phloroglucinol 1a, reactions using bicyclic arenes were investigated
to determine the constraints on the construction of the tetracyclic
core of the ascorbylated catechins. A set of four hydroxylated
chroman and chroman-4-ones was used, and only one of these
2.5. Structural characterization of (L)-ascorbyl
phloroglucinol 1a
Previous attempts to isolate 1a from natural sources employed
peracetylation, presumably to aid in purification.11 In our hands, 1a

9268
S.A. Belapure et al. / Tetrahedron 67 (2011) 9265e9272
Table 5
Selected COSY and HMBC data for 12c
Atom
dH, mult, J¼Hz
dC
COSY
HMBC
H7
H8
H9
H2
H6
C5
4.20, m
1.98, m
2.54, m
6.13, s
d
d
H8
C8, C9, C5
C7, C9, C4
C7, C8, C3, C4, C5
C1, C3, C4, C6
d
d
H7, H9
H8
d
d
Fig. 2. ORTEP for 1b.
d
d
d
d
153.7
d
H7, H9
molecules, chroman-5,7-diol (11c), was able to successfully react
with 4 (Table 4). These bicyclic substrates fit the constraints ob-
served for reactions with the substituted arenes discussed pre-
viously in that three activating groups that all direct to the same
position are needed for reaction and no deactivating groups can be
tolerated.
Table 6
Selected COSY and HMBC data for 13c or 13c0
Table 4
Reaction of 4 with bicyclic hydroxyarenes
Atom
dH, mult, J¼Hz
dC
COSY
HMBC
H7
H8
H9
H2
H6
C5
4.17, t, J¼5.2
1.94, m
2.54, t, J¼6.3
d
d
H8
C8, C9, C5
C7, C9, C4
C7, C8, C3, C4, C5
d
d
H7, H9
H8
d
d
d
6.03, s
d
d
d
C1, C2, C4, C5
H7, H9, H6
158.2
d
Entry
Arene
X
Y
Product Ratio
Yield (%)
(HMBC) were the most useful, and it was concluded that 12c and
13c were the major and minor component, respectively (Tables 5
and 6 and associated Figs.). The 1H peak shifts and correlations in
the COSY experiment allowed H7, H8, and H9 to be identified for
each molecule. With these data in hand, the peak for C5 could be
assigned from the HMBC experiment based on chemical shift and
coupling with H7 and H9. The connectivity of 12c and 13c (or 13c0)
could then be determined from the HMBC data, and the molecules
could be readily differentiated. For the skeleton of 12c, only a pair of
two or three bond 1He13C couplings involving C5 are possible be-
tween this atom and both H7 or H9. In contrast, a trio of two or
1
2
3
4
11a
11b
11c
11d
CH2
C]O
CH2
H
H
OH
OH
NA
NA
NR
NR
72
1.88:1 (12c/13c)
NA
C]O
NR
Regioisomers can be generated upon reaction of 4 with 11c. One
of theseisomers,12c, corresponds tothedesiredtetracyliccore of the
natural products 2a,b, while the other ring system, as in 13c, has not
been shown to be a naturally occurring compound. (Note, if Y is
a hydroxyl, e.g., as in13c and13c0 inTable6, two possible hemiacetals
could be generated and may interconvert after rotation around the
CeC bond formed during the reaction of the arene with 4.) In our
hands, an w2:1 mixture of two molecules was obtained when 11c
was reacted with 4 using the conditions optimized for the reaction of
1a. These molecules were separable by flash chromatography on
silica gel, and a battery of 1D and 2D NMR experiments allowed the
structure of both to be elucidated (see Supplementary data).
1
three He13C couplings should be observed for C5 of either 13c or
the isomer 13c0 since the aromatic proton H6 is also able to couple
with this key carbon atom along with both H7 and H9. Indeed, the
HMBC data were consistent with the expected proton-C5 coupling
patterns for both 12c and 13c, although 13c and 13c0 could not be
distinguished via NMR spectroscopy.
Both molecular mechanics (MMX17) and density functional
theory (DFT) calculations were performed to determine the relative
conformational stabilities of 12c, 13c, and 13c0. Solvation was not
considered in either case. PCModel v. 4.0 (Serena Software) was
used for the MMX calculations, and intramolecular hydrogen
bonding was maximized among all hydroxyl groups. MMX pre-
dicted 13c to be 1.1 kcal/mol more stable than either 12c or 13c0,
2.7. Structural characterization of 12c and 13c
Of the experiments, 1He1H correlation spectroscopy (COSY) and
1He13C heteronuclear multiple-bond correlation spectroscopy

S.A. Belapure et al. / Tetrahedron 67 (2011) 9265e9272
9269
which were within 0.2 kcal/mol of each other. The DFT calculations
were performed at the B3LYP/6-31þG* level of theory18 with
Gaussian 03,19 and the structures determined by MMX were used as
starting geometries. Again,13c was found to be more favorable than
13c0 by 1.0 kcal/mol, and both were found to be lower in energy
than 12c by 7.5 and 6.5 kcal/mol for 13c and 13c0, respectively.
These results suggest that the reaction may be under kinetic control
and that 13c is the most likely structure for the minor isomer, al-
though this has not been confirmed.
However, it should be noted that the energies calculated for all
molecules by MMX could be altered by up to 8 kcal/mol by breaking
hydrogen bonding with negligible geometry changes. This high-
lights the fact that core of these structures are similar in energy.
Further, intramolecular hydrogen bonding networks are typically
disrupted in hydrogen bonding solvents, such as the 5:2 AcOH/THF
(v/v) solution used for this reaction, and it is possible that solvation
of either the C1 or C3 hydroxyl in the intermediate(s) increases
steric interactions during the formation of either 12c or 13c, as
exemplified for C1 in Scheme 4. Such solvation effects, which were
ignored during the calculations described above, could explain the
formation of the product with the lowest gas phase stability under
the reaction conditions employed in this study.
the molecules reported herein and to expand the scope of this
methodology to the construction of more complex ascorbylated
natural products and analogues.
4. Experimental
4.1. Materials and general methods
All reagents and solvents were purchased from Fisher Scientific
ꢀ
and used without further purification unless noted. Silica gel (60 A,
40e63
m
m, 230ꢂ400 mesh) and polyester backed thin-layer chro-
matography (TLC) plates (Silica G w/UV, 200
m
M) were purchased
from Sorbent Technologies. All reactions were carried out under
nitrogen unless otherwise stated. All 1D 1H and 13C spectra were
recorded using a 300 MHz Varian Mercury or 500 MHz Varian
INOVA spectrometer. Deuterated solvents used for NMR analyses
were purchased from Cambridge Isotope Laboratories. The NMR
chemical shifts are reported in parts per million (ppm) and refer-
enced to the solvent peak. All 2D NMR spectra were recorded on
a 600 MHz Varian INOVA. IR spectra were obtained on a Varian
4000 FT-IR instrument. High-resolution mass spectra (HRMS) were
obtained with an Applied Biosystems MDS Sciex Qstar Elite time-
of-flight mass spectrometer (MS), and optical rotations were
recorded using a PerkineElmer 241 polarimeter. Data for the X-ray
crystal structure of 1b were collected on a Bruker-AXS Smart APEX
II diffractometer fitted with a Nicolet LT-2 low temperature device
ꢀ
and a graphite-monochromated Mo K
a
radiation at 0.71073 A.
4.1.1. General procedure for the reaction of hydroxyarenes with 5
(7aed). These reactions were carried out via slight modifications of
previously reported procedures.15,16 Ninhydrin (1.00 mmol) and the
appropriate hydroxyarene (1.00 mmol, 6aed) were placed in
a round-bottom flask and dissolved in 3 mL of AcOH. If the reagents
were not completely soluble, THF was added dropwise until the
mixture was homogeneous. The solution was then refluxed for
2e3 h. At this time, the reaction mixture was cooled to rt and di-
luted with 10 mL H2O. The product was isolated after extractions of
the aqueous layer with ethyl acetate (EtOAc, 3ꢂ15 mL). The com-
bined organic layers were dried with MgSO4, filtered, and then
concentrated in vacuo. The resulting crude product was purified by
column chromatography on silica gel using 20% acetone in CH2Cl2
as the eluent.
Scheme 4. Solvation effect during the reaction of 4 with 11c.
While the reaction of 11c with 4 yielded 12c with the naturally
occurring tetracyclic core as the major product, the fact that this
molecule was only 65% of the isolated material led us to investigate
whether protection at the hydroxyl on the carbon that would be-
come C3 would shift the selectivity further toward 12c due to
sterics. Unfortunately, the construction of either C3 benzylated or
silylated hydroxylated chromans proved difficult due to lability of
the protecting group (data not shown). These observations are
consistent with the decomposition of dibenzylated arene 6b under
the reaction conditions, and they highlight the good leaving group
ability of trioxyarenes.
4.1.1.1. 4b,7,9,9b-Tetrahydroxy-4bH-indeno[1,2-b]benzofuran-
10(9bH)-one (7a) and 4b,9b-dihydroxy-4bH-indeno[1,2-b]benzofu-
ran-10(9bH)-one (7d). These molecules have been reported pre-
viously,15,16 and all analytical data collected for these studies
matched those in the literature.
4.1.1.2. 4b,7,9b-Trihydroxy-4bH-indeno[1,2-b]benzofuran-
10(9bH)-one (7c). Rf (30% acetone/CH2Cl2) 0.68; mp¼226 ꢁC; IR
(KBr, thin film) nmax (cmꢀ1): 1150, 1466, 1499, 1626, 1731, 2957,
3280e3580 (br); 1H NMR (500 MHz, acetone-d6,
d): 8.62 (s, 1H),
7.98 (d, J¼7.8 Hz, 1H), 7.88 (t, J¼7.5 Hz, 1H), 7.74 (d, J¼7.7 Hz, 1H),
7.69e7.58 (m, 1H), 7.26 (d, J¼8.3 Hz, 1H), 6.55 (s, 1H), 6.44 (dd,
J¼8.3, 2.2 Hz, 1H), 6.25 (d, J¼2.2 Hz, 1H), 5.65 (s, 1H); 13C NMR
3. Conclusion
A method utilizing readily available 4 for the biomimetic
ascorbylation of phenolic natural products has been reported
herein. This reaction was found to be stereo-, chemo-, and regio-
selective, and a synthesis of the natural product, (ꢀ)-ascorbyl
phloroglucinol (1a), was accomplished. Data obtained from the
crystal structure of 1b corroborate previous theoretical assignment
of the configuration for the stereocenters at the ascorbylephenol
junction in compounds of this type. Further, the tetracyclic core of
ascorbylated catechins, such as 8-C-ascorbyl-(ꢀ)-epigallocatechin,
was constructed. Efforts are underway to evaluate the bioactivity of
(126 MHz, acetone-d6, d): 161.0, 158.7, 148.9, 136.3, 134.5, 130.7,
126.2, 124.9, 122.8, 116.3, 109.1, 97.3, 82.2; HRMS-ESI (m/z): [MꢀH]ꢀ
calcd for C15H10O5, 269.0450; found, 269.0454.
4.1.1.3. 4b,9b-Dihydroxy-7,9-dimethoxy-4bH-indeno[1,2-b]ben-
zofuran-10(9bH)-one (7b). Rf (30% acetone/CH2Cl2) 0.79;
mp¼204 ꢁC; IR (KBr, thin film) nmax (cmꢀ1): 1152, 1466, 1497, 1623,
1734, 2956, 3419; 1H NMR (500 MHz, acetone-d6,
d): 7.93 (d,
J¼7.2 Hz, 1H), 7.84 (t, J¼7.0 Hz, 1H), 7.73 (d, J¼7.3 Hz, 1H), 7.63 (d,
J¼7.0 Hz, 1H), 6.45 (s, 1H), 6.06 (s, 1H), 6.00 (s, 1H), 5.52 (s, 1H), 3.77

9270
S.A. Belapure et al. / Tetrahedron 67 (2011) 9265e9272
(s, 3H), 3.72 (s, 3H); 13C NMR (126 MHz, acetone-d6,
159.8, 159.2, 148.1, 135.9, 131.0, 124.7, 122.8, 110.5, 104.6, 93.9, 92.0,
88.2, 83.3, 55.0, 54.9; HRMS-ESI (m/z): [MꢀH]ꢀ calcd for C17H13O6,
313.0712; found, 313.072.
d
): 196.9, 164.3,
This material was purified by flash chromatography using silica
gel eluted with 20% acetone in CH2Cl2 to give 146 mg 11d as an
amorphous, pale-yellow solid (11%). Partial characterization of
this material has been reported previously.21 Rf (30% acetone/
CH2Cl2) 0.76; 1H NMR (300 MHz, DMSO-d6,
d
): 9.67 (s, 1H), 9.39
4.1.2. Preparation of -dehydroascorbic acid (4). This material was
L
(s, 1H), 6.11 (d, J¼2.3 Hz, 1H), 5.90 (d, J¼2.3 Hz, 1H), 2.73e2.62 (m,
prepared via a slight modification of a known procedure.13 Briefly,
4H); 13C NMR (75 MHz, DMSO-d6,
d): 196.7, 166.9, 164.0, 163.5,
iodine (1.40 g, 5.68 mmol) was added in one portion to a stirring
102.5, 96.1, 95.2, 66.7, 36.3; HRMS-ESI (m/z): [MꢀH]ꢀ calcd for
solution of
L-ascorbic acid (1.00 g, 5.68 mmol) dissolved in metha-
C9H7O4, 179.0344; found, 179.0349.
nol (10 mL), and the reaction mixture was allowed to stir for 10 min
at rt. At this time, lead carbonate (3.40e3.65 g) was added slowly to
the brown reaction mixture until it turned colorless. The resulting
mixture was filtered through Celite to remove the bulk of the Pb,
and the filtrates were then treated with H2S to precipitate the
remaining metal. The resulting solids were removed from the fil-
trate by a second filtration through Celite, and air was blown
through the crude solution containing 4 to remove excess H2S.
Concentration in vacuo gave the desired product as a white, sticky
solid in w25% yield and sufficient purity to be used in further re-
actions. Partial characterization of this material has been reported
4.1.5. General procedure for the ascorbylation of hydroxyarenes
(1aei, 12aed, and 13aed). Dehydroascorbic acid
4 (0.250 g,
1.43 mmol) was placed in a round-bottom flask and dissolved in
THF (5 mL). The phloroglucinol derivative (2.14 mmol, 6aei) was
then added to the reaction in one portion followed by glacial AcOH
(2 mL), and the resulting solution was stirred for 8e10 h. At this
time, the reaction mixture was concentrated in vacuo, and the
resulting material was purified via flash silica chromatography
using 70% EtOAc in hexanes as the eluent.
previously.13 1H NMR (300 MHz, D2O,
d
): 4.55 (ddd, J¼5.2, 2.7,
4.1.5.1. (3R,3aR,8bS)-3-((S)-1,2-Dihydroxyethyl)-3a,6,8,8b-tetra-
hydroxy-3,3a-dihydrofuro[3,4-b]benzofuran-1(8bH)-one (1a). This
material was obtained in 69% yield as a white, amorphous solid. Rf
0.8 Hz, 1H), 4.24 (dd, J¼10.4, 2.6 Hz, 1H), 4.12 (dd, J¼10.4, 2.6 Hz,
1H), 3.70e3.74 (m, 1H).
(50% EtOAc/hexanes) 0.56; ½a D25
ꢀ46 (c 0.01, H2O); IR (KBr, thin
ꢃ
4.1.3. Chroman-5,7-diol
(11c). Phloroglucinol
6a
(1.00 g,
film) nmax (cmꢀ1): 1150, 1472, 1633, 1773, 2962, 3256e3430 (br); 1H
7.93 mmol) was placed in a round-bottom flask and dissolved in
10 mL of aqueous 2 N NaOH at rt. An ethanolic solution of 1,3-
dibromopropane (0.881 mL, 8.72 mmol dissolved in 10 mL) was
then added slowly, and the reaction mixture was stirred for 10 h at
rt. At this time, the reaction was cooled to 0 ꢁC with an ice bath and
then acidified using 1 N HCl to pH 2. The aqueous layer was
extracted with EtOAc (3ꢂ20 mL). The combined organic layers
were then dried with MgSO4 and filtered. The resultant superna-
tants were concentrated under vacuum to give a crude liquid,
which was then purified via flash silica chromatography using 20%
acetone in CH2Cl2 as the eluent. This procedure yielded 0.450 g of
11c as a pure, amorphous white solid (34%). Rf (30% acetone/CH2Cl2)
0.75; mp¼176 ꢁC; IR (KBr, thin film) nmax (cmꢀ1): 949, 1189, 1278,
1475, 1521, 1621, 2975, 3100e3480 (br); 1H NMR (500 MHz, ace-
NMR (600 MHz, D2O,
d
): 6.08 (d, J¼1.1 Hz, 1H, AreH), 6.06 (d,
J¼1.8 Hz, 1H, AreH), 4.48 (d, J¼6.1 Hz, 1H, eOCHCHOHCH2OH), 4.21
(dd, J¼10.7, 5.3 Hz,1H, eOCHCHOHCH2OH), 3.85 (dd, J¼11.8, 4.0 Hz,
1H, eOCHCHOHCH2OH), 3.74 (dd, J¼11.7, 6.4 Hz, 1H,
eOCHCHOHCH2OH); 1H NMR (600 MHz, DMSO-d6,
d): 9.55 (s, 1H,
OH), 9.46 (s, 1H, OH), 7.77 (s, 1H, OH), 6.03 (s, 1H, OH), 5.88 (s, 1H,
AreH), 5.72 (s, 1H, AreH), 4.96 (s, 1H, OH), 4.70 (s, 1H, OH), 4.12 (d,
J¼4.7 Hz, 1H, eOCHCHOHCH2OH), 3.82 (m, 1H, eOCHCHO
HCH2OH), 3.52 (m, 1H, eOCHCHOHCH2OH), 3.39 (m, 1H,
eOCHCHOHCH2OH); 13C NMR (151 MHz, D2O,
d): 174.4, 161.2, 159.1,
156.1, 110.9, 101.4, 97.3, 90.9, 83.3, 78.7, 69.4, 62.1; HRMS-ESI (m/z):
[MꢀH]ꢀ calcd for C12H11O9, 299.0403; found, 299.0403.
4.1.5.2. (3R,3aR,8bS)-3-((S)-1,2-Dihydroxyethyl)-3a,8b-dihy-
droxy-6,8-dimethoxy-3,3a-dihydrofuro[3,4-b]benzofuran-1(8bH)-
one (1b). This material was generated in 72% yield as a white,
tone-d6,
d
): 8.02 (s, 1H, OH), 7.83 (s, 1H, OH), 5.96 (d, J¼2.3 Hz, 1H,
AreH), 5.79 (d, J¼2.3 Hz, 1H, AreH), 4.07e3.96 (m, 2H,
eOCH2CH2CH2e), 2.52 (t, J¼6.6 Hz, 2H, eOCH2CH2CH2e), 1.94e1.77
crystalline solid. Rf (70% EtOAc/hexanes) 0.43; ½a D25
ꢀ82.6 (c 0.01,
ꢃ
(m, 2H, eOCH2CH2CH2e); 13C NMR (126 MHz, CD3OD,
d): 156.2,
H2O); mp¼181 ꢁC; IR (KBr, thin film) nmax (cmꢀ1): 800, 1107, 1152,
156.0, 155.9, 101.3, 94.5, 94.4, 94.4, 65.8, 22.0, 18.5; HRMS-ESI (m/z):
1596, 1789, 2362, 2459, 2568, 2962, 3168, 3305, 3469; 1H NMR
[MꢀH]ꢀ calcd for C9H9O3, 165.0552; found, 165.0557.
(300 MHz, D2O,
d
): 6.20 (d, J¼1.9 Hz, 1H, AreH), 6.18 (d, J¼1.2 Hz,
1H, AreH), 4.46 (d, J¼6.4 Hz, 1H, eOCHCHOHCH2OH), 4.19 (dd,
J¼10.1, 5.7 Hz, 1H, eOCHCHOHCH2OH), 3.88e3.82 (m, 1H,
eOCHCHOHCH2OH), 3.81 (s, 3H, eOCH3), 3.78 (s, 3H, eOCH3), 3.72
4.1.4. 5,7-Dihydroxychroman-4-one (11d). This preparation fol-
lows a known procedure that employed resorcinol 6c.20 Phlor-
oglucinol 6a (0.88 g, 7.00 mmol) and 3-chloropropionic acid
(0.83 g, 7.70 mmol) were dissolved in 2.7 mL of triflic acid. The
reaction mixture was then heated at 80e90 ꢁC for 1 h. At this
time, the reaction was cooled to rt and diluted with EtOAc
(20 mL). The resulting mixture was then poured onto ice
(w60e70 g), and a precipitate formed. The entire mixture was
extracted with EtOAc (3ꢂ20 mL). The organic layers were then
combined, dried over MgSO4, filtered, and then concentrated
in vacuo to give 3-chloro-1-(2,4,6-trihydroxyphenyl)propan-1-
one as an orange oil that was of sufficient purity to be used di-
rectly in the next step. This intermediate was then placed in
a round-bottom flask and cooled to 0 ꢁC in an ice bath. 2 N NaOH
(50 mL) was then added slowly to the reaction vessel, and the
resulting mixture was stirred at rt for 2 h. At this time, the re-
action was cooled to 0 ꢁC and acidified to pH 2 with 6 N H2SO4.
The crude product was first purified by extraction with EtOAc
(3ꢂ25 mL); and the combined organic layers were dried over
MgSO4, filtered, and concentrated in vacuo to give a yellow oil.
(m, 1H, eOCHCHOHCH2OH); 1H NMR (500 MHz, DMSO-d6,
d): 8.00
(s, 1H, OH), 6.15 (s, 1H, AreH), 6.14 (s, 1H, AreH), 5.09 (s, 1H, OH),
4.74 (s, 1H, eOCHCHOHCH2OH), 4.17 (s, 1H, eOCHCHOHCH2OH),
4.02 (s, 1H, OH), 3.86 (s, 1H, OH), 3.74 (s, 6H, eOCH3), 3.53 (s, 1H,
eOCHCHOHCH2OH), 3.37 (s, 1H, eOCHCHOHCH2OH); 13C NMR
(151 MHz, D2O, d): 174.4, 164.8, 159.0, 158.7, 110.9, 102.3, 93.5, 89.4,
83.4, 78.9, 69.4, 62.0, 55.9, 55.8; HRMS-ESI (m/z): [MꢀH]ꢀ calcd for
C14H16O9, 327.0713; found, 327.0713.
4.1.5.3. Procedure for X-ray analysis of 1b. To obtain crystallo-
graphic data (Table 7), a suitable crystal was selected and mounted
using glue, and the data were collected at 23 ꢁC. The structures
were solved by direct methods, and non-hydrogen atoms were
anisotropically refined by treating all hydrogen atoms as idealized
contributions. Empirical absorption correction was performed with
the SADABS software package from Bruker. In addition, global re-
finements for the unit cell and data reduction of the structure were
performed using Saint version 6.02 (Bruker), and all calculations

S.A. Belapure et al. / Tetrahedron 67 (2011) 9265e9272
9271
were performed using the SHELXTL version 5.1 proprietary soft-
ware package from Bruker. Crystallographic data for 1b have been
deposited in the Cambridge Crystallographic Data Centre (CCDC) as
supplementary publication numbers CCDC 827129. These data can
be freely obtained from the CCDC by sending an application by
email to deposit@ccdc.cam.ac.uk.
4.1.6. Preparation of (3R,3aS,8bS)-3-((S)-1,2-diacetoxyethyl)-1-oxo-
1,3,3a,8b-tetrahydrofuro[3,4-b]benzofuran-3a,6,8,8b-tetrayl tetraa-
cetate (3). (ꢀ)-Ascorbyl phloroglucinol 1a (100 mg, 0.33 mmol)
was placed in a round-bottom flask contained distilled pyridine
(215 mL), and acetic anhydride (215 mL) was then added. The re-
action mixture was stirred for 14 h at rt. At this time, the contents of
the reaction vessel were poured into ice water (50 mL), and the
resulting solution was extracted with EtOAc (3ꢂ10 mL). The organic
layers were collected, combined, dried with MgSO4, and concen-
trated in vacuo to give an oil, which was purified via flash silica
chromatography using 10% acetone in CH2Cl2 as the solvent
providing 3 as a white, amorphous solid in 10% yield. Character-
ization of this material has been reported previously.11 1H NMR
Table 7
X-ray crystallographic data for 1b
Identification code
Empirical formula
Formula weight
Temperature
1b
C14H16O9
328.27
296(2) K
ꢀ
Wavelength
0.71073 A
(600 MHz, CDCl3,
d
): 6.71 (d, J¼1.6 Hz, 1H, AreH), 6.67 (d, J¼1.5 Hz,
Crystal system
Space group
Unit cell dimensions
Orthorhombic
P2(1)2(1)2(1)
1H, AreH), 5.78e5.59 (m, 1H, eOCHCHOHCH2OH), 4.97 (d,
J¼4.5 Hz, 1H, OCHCHOHCH2OH), 4.41 (dd, J¼11.8, 4.5 Hz, 1H, OCH-
CHOHCH2OH), 4.26 (dd, J¼11.8, 6.3 Hz, 1H, OCHCHOHCH2OH), 2.33
(s, 3H, eOAc), 2.28 (s, 3H, eOAc), 2.16 (s, 6H, eOAc), 2.11 (s, 3H,
a¼6.743(5) A
a
¼90ꢁ
ꢀ
¼90ꢁ
ꢀ
b¼13.829(10) A
b
c¼14.625(11) A
g
¼90ꢁ
ꢀ
3
ꢀ
Volume
Z
Density (calculated)
Absorption coefficient
F (000)
Crystal size
Theta range for data collection
Index ranges
1363.7(17) A
4
eOAc), 2.09 (s, 3H, eOAc); 13C NMR (151 MHz, CDCl3,
d): 170.5,
1.584 mg/m3
0.135 mmꢀ1
676
169.9,168.2,166.5,166.1,154.7,148.4,168.4,158.2,110.2,109.6,112.1,
83.9, 82.1, 68.0, 62.7, 21.3, 21.0, 21.1, 21.0, 20.1, 20.0.
0.05ꢂ0.05ꢂ0.4 mm3
2.03e28.46ꢁ
Acknowledgements
ꢀ8ꢄhꢄ8, ꢀ18ꢄkꢄ18,
ꢀ19ꢄIꢄ19
The authors would like to acknowledge several colleagues from
the University of Tennessee, Knoxville. We thank Drs. David C. Baker
and Michael D. Best for helpful discussion, as well as Jessica R.
Gooding and Amanda L. May (University of Tennessee, Knoxville) for
assistance in preparing this manuscript. We are also grateful to Julia
K.C. Abbott and Dr. Zheng Lu for assistance in obtaining crystallo-
graphic data. Support for this work was provided by start-up funds
from the University of Tennessee, Knoxville.
Reflections collected
15,452
Independent reflections
Completeness to theta¼28.46ꢁ
Refinement method
3262 [R(int)¼0.0374]
96.7%
Full-matrix least-squares on F2
3262/0/214
Data/restraints/parameters
Goodness-of-fit on F2
1.019
Final R indices [I>2
R indices (all data)
Absolute structure parameter
Largest diff. peak and hole
s
(I)]
R1¼0.0351, wR2¼0.0780
R1¼0.0492, wR2¼0.0848
ꢀ0.1(9)
ꢀꢀ3
0.188 and ꢀ0.185 e A
Supplementary data
4.1.5.4. (7aR,8R,10aS)-8-((S)-1,2-Dihydroxyethyl)-5,7a,10a-tri-
hydroxy-3,4,7a,8-tetrahydro-2H-furo[30,40:4,5]furo[2,3-h]chromen-
10(10aH)-one (12c). This material was the major product from re-
action of 4 with 11c and was generated in 47% yield as a white,
This material includes 1D and 2D NMR spectroscopic data for all
new compounds as well as X-ray crystallographic data for 1b.
Supplementary data associated with this article can be found, in the
online version, at doi:10.1016/j.tet.2011.09.102.
amorphous solid. Rf (60% acetone/CH2Cl2) 0.76;½a D25
ꢀ14 (c 0.01,
ꢃ
acetone); IR (KBr, thin film) nmax (cmꢀ1): 1155, 1455, 1511, 1638,
References and notes
1775, 2959, 3000e3620 (br); 1H NMR (500 MHz, D2O,
d): 6.13 (s,
1H, AreH), 4.46 (d, J¼5.7 Hz, 1H, eOCHCHOHCH2OH), 4.20 (m, 3H,
eOCH2CH2CH2e and eOCHCHOHCH2OH), 3.87 (dd, J¼11.9, 4.3 Hz,
1H, eOCHCHOHCH2OH), 3.75 (dd, J¼11.8, 6.5 Hz, 1H,
eOCHCHOHCH2OH), 2.61e2.47 (m, 2H, eOCH2CH2CH2e),
1. Murase, T.; Nagasawa, A.; Suzuki, J.; Hase, T.; Tokimitsu, I. Int. J. Obes. 2002, 26,
1459e1464.
2. Hashimoto, F.; Kashiwada, Y.; Nonaka, G.; Nishioka, I.; Nohara, T.; Cosentino, L.
M.; Lee, K. H. Bioorg. Med. Chem. Lett. 1996, 6, 695e700.
3. Benzie, I. F. F.; Szeto, Y. T. J. Agric. Food Chem. 1999, 47, 633e636.
4. Miura, S.; Watanabe, J.; Tomita, T.; Sano, M.; Tomita, I. Biol. Pharm. Bull. 1994, 17,
1567e1572.
5. Yen, G. C.; Chen, H. Y. J. Agric. Food Chem. 1995, 43, 27e32.
6. Fujiki, H.; Suganuma, M.; Okabe, S.; Sueoka, N.; Komori, A.; Sueoka, E.; Kozu, T.;
Tada, Y.; Suga, K.; Imai, K.; Nakachi, K. Mutat. Res. 1998, 402, 307e310.
7. Paschka, A. G.; Butler, R.; Young, C. Y. F. Cancer Lett. 1998, 130, 1e7.
8. Suzuki, K.; Yahara, S.; Hashimoto, F.; Uyeda, M. Biol. Pharm. Bull. 2001, 24,
1088e1090.
9. Valcic, S.; Timmermann, B. N.; Alberts, D. S.; Wachter, G. A.; Krutzsch, M.;
Wymer, J.; Guillen, J. M. Anti-Cancer Drugs 1996, 7, 461e468.
10. Nakai, M.; Fukui, Y.; Asami, S.; Toyoda-Ono, Y.; Iwashita, T.; Shibata, H.; Mit-
sunaga, T.; Hashimoto, F.; Kiso, Y. J. Agric. Food Chem. 2005, 53, 4593e4598.
11. Keusgen, M.; Falk, M.; Walter, J. A.; Glombitza, K. W. Phytochemistry 1997, 46,
341e345.
12. Hashimoto, F.; Nonaka, G.; Nishioka, I. Chem. Pharm. Bull. 1989, 37, 3255e3263.
13. Kenyon, J.; Munro, N. J. Am. Chem. Soc. 1948, 158e161.
14. Andrews, G. C.; Crawford, T. C.; Bacon, B. E. J. Org. Chem. 1981, 46, 2976e2977.
15. Na, J. E.; GowriSankar, S.; Lee, S.; Kim, J. N. Bull. Korean Chem. Soc. 2004, 25,
569e572.
2.03e1.92 (m, 2H, eOCH2CH2CH2e); 13C NMR (126 MHz, D2O,
d):
174.3, 158.3, 156.1, 153.7, 110.5, 105.3, 101.0, 90.4, 87.3, 82.9, 78.8,
76.0, 72.5, 69.2, 66.8, 61.9, 20.8, 18.4; HRMS-ESI (m/z): [MꢀH]ꢀ
calcd for C9H9O3, 339.0733; found, 339.0732.
4.1.5.5. (6bS,9R,9aR)-9-((S)-1,2-Dihydroxyethyl)-6,6b,9a-tri-
hydroxy-2,3,9,9a-tetrahydro-1H-furo[30,40:4,5]furo[2,3-f]chromen-
7(6bH)-one (13c). This material was the minor product from re-
action of 4 with 11c and was obtained in 25% yield as a white,
amorphous solid. Rf (70% acetone/CH2Cl2) 0.75; ½a D25
ꢃ
þ10.8 (c 0.01,
nmax
acetone); IR (KBr, thin film)
(cmꢀ1): 1155.16, 1239.66, 1455,
1511,1638,1775, 2959, 2990e3620 (br); 1H NMR (500 MHz, D2O,
d):
6.03 (s, 1H, AreH), 4.46 (d, J¼5.4 Hz, 1H, eOCHCHOHCH2OH), 4.24
(dd, J¼9.8, 6.7 Hz, 1H, eOCHCHOHCH2OH), 4.17 (t, J¼5.2 Hz, 2H,
eOCH2CH2CH2e), 3.87 (dd, J¼11.8, 4.4 Hz, 1H, eOCHCHOHCH2OH),
3.74 (dd, J¼11.8, 6.8 Hz, 1H, eOCHCHOHCH2OH), 2.54 (t, J¼6.3 Hz,
2H, eOCH2CH2CH2e), 1.94 (m, 2H, eOCH2CH2CH2e); 13C NMR
16. Na, J. E.; Lee, K. Y.; Seo, J.; Kim, J. N. Tetrahedron Lett. 2005, 46, 4505e4508.
17. Gajewski, J. J.; Gilbert, K. E.; McKelvey, J. In Advances in Molecular Modeling;
Liotta, D., Ed.; JAI: Greenwich, CT, 1990; Vol. 2.
18. Becke, A. D. J. Chem. Phys. 1993, 98, 5648e5652.
(126 MHz, D2O, d): 174.1, 158.3, 156.2, 153.0, 110.6, 101.2, 99.6, 97.3,
19. Frisch, M. J. T.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.;
Cheeseman, J. R.; Montgomery, J. A., Jr.; Vreven, T.; Kudin, K. N.; Burant, J. C.;
Millam, J. M.; Iyengar, S. S.; Tomasi, J.; Barone, V.; Mennucci, B.; Cossi, M.;
83.2, 78.8, 69.1, 67.1, 61.8, 20.7, 17.7; HRMS-ESI (m/z): [MꢀH]ꢀ calcd
for C15H16O9, 339.0733; found, 339.0731.

9272
S.A. Belapure et al. / Tetrahedron 67 (2011) 9265e9272
Scalmani, G.; Rega, N.; Petersson, G. A.; Nakatsuji, H.; Hada, M.; Ehara, M.;
Foresman, J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A. G.; Clifford, S.; Cioslowski, J.;
Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.; Martin, R. L.; Fox,
D. J.; Keith, T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara, A.; Challacombe, M.;
Gill, P. M. W.; Johnson, B.; Chen, W.; Wong, M. W.; Gonzalez, C.; Pople, J. A.;
Gaussian,: Wallingford CT, 2004.
Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao,
O.; Nakai, H.; Klene, M.; Li, X.; Knox, J. E.; Hratchian, H. P.; Cross, J. B.; Bakken,
V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A.
J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Ayala, P. Y.; Morokuma, K.; Voth, G.
A.; Salvador, P.; Dannenberg, J. J.; Zakrzewski, V. G.; Dapprich, S.; Daniels, A. D.;
Strain, M. C.; Farkas, O.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.;
20. Koch, K.; Biggers, M. S. J. Org. Chem. 1994, 59, 1216e1218.
21. Farmer, V. C.; Hayes, N. F.; Thomson, R. H. J. Chem. Soc. 1956, 3600e3607.