Preparation of 3-hydroxy-2,3-dialkoxy-2-phenylchroman-4-ones and 3,3-dihydroxy-2-alkoxy-2-phenylchroman-4-ones by oxidation of 3-hydroxyflavone with copper(II) bromide: Structure, reactivity and characterization

Article abstract of DOI:10.1016/j.ica.2020.119855

The reaction of 3-hydroxyflavone with copper(II) bromide in various alcohols (ethanol, isopropanol, and methanol) yielded a mixture of their respective 3-hydroxy-2,3-dialkoxy-2-phenylchroman-4-ones (hemiacetals) and 3,3-dihydroxy-2-alkoxy-2-phenylchroman-4-ones (hydrates). Herein, we report the synthesis, reactivity and characterization of three hemiacetals (1, 3, and 5) and three hydrates (2, 4, and 6) as well as the single crystal X-ray structures of 1–4 & 6. While the hemiacetals were shown to be kinetically accessible, the hydrates were thermodynamically preferred. The single X-ray structures of compounds 1–4 & 6 reveals the alkoxy groups occupying the axial position of the coumarin ring, which is surmised to be the source of the instability of the hemiacetals. As a result of the axial strain, the hemiacetals (1, 3, and 5) readily convert in to their respective hydrates (2, 4, and 6) when heated in non-alcoholic solvents. The qualitative rate of these conversion was largely a function of the size of the alkoxy group (i-Pr > Et > Me). We propose a mechanism for the conversion of hemiacetals to hydrates that involves a chromane-3,4-dione intermediate, which was evidenced by trapping it as a diimine, quinoxaline heterocycle.

Full text of DOI:10.1016/j.ica.2020.119855

Journal Pre-proofs
Research paper
Preparation of 3-hydroxy-2,3-dialkoxy-2-phenylchroman-4-ones and 3,3-di‐
hydroxy-2-alkoxy-2-phenylchroman-4-ones by oxidation of 3-hydroxyflavone
with Copper(II) Bromide: Structure, reactivity and characterization
Elizabeth M. Beasley, Joseph G. Bazemore, Anthony Petrillo, Clifford W.
Padgett, Will E. Lynch, Brandon Quillian
PII:
DOI:
Reference:
S0020-1693(20)31054-9
https://doi.org/10.1016/j.ica.2020.119855
ICA 119855
To appear in:
Inorganica Chimica Acta
Received Date:
Revised Date:
Accepted Date:
25 March 2020
23 June 2020
26 June 2020
Please cite this article as: E.M. Beasley, J.G. Bazemore, A. Petrillo, C.W. Padgett, W.E. Lynch, B. Quillian,
Preparation of 3-hydroxy-2,3-dialkoxy-2-phenylchroman-4-ones and 3,3-dihydroxy-2-alkoxy-2-
phenylchroman-4-ones by oxidation of 3-hydroxyflavone with Copper(II) Bromide: Structure, reactivity and
characterization, Inorganica Chimica Acta (2020), doi: https://doi.org/10.1016/j.ica.2020.119855
This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover
page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version
will undergo additional copyediting, typesetting and review before it is published in its final form, but we are
providing this version to give early visibility of the article. Please note that, during the production process, errors
may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
© 2020 Elsevier B.V. All rights reserved.

Preparation of 3-hydroxy-2,3-dialkoxy-2-phenylchroman-4-ones and 3,3-dihydroxy-2-
alkoxy-2-phenylchroman-4-ones by oxidation of 3-hydroxyflavone with Copper(II)
Bromide: Structure, reactivity and characterization
Elizabeth M. Beasley, Joseph G. Bazemore, Anthony Petrillo, Clifford W. Padgett, Will E.
Lynch, Brandon Quillian*
Department of Chemistry and Biochemistry, Georgia Southern University, 11935 Abercorn St.,
Savannah, GA 31419, United States
Email: bquillian@georgiasouthern.edu
Phone: 1-912-344-2977
*Corresponding Author
Abstract: The reaction of 3-hydroxyflavone with copper(II) bromide in various alcohols (ethanol,
isopropanol, and methanol) yielded a mixture of their respective 3-hydroxy-2,3-dialkoxy-2-
phenylchroman-4-ones (hemiacetals) and 3,3-dihydroxy-2-alkoxy-2-phenylchroman-4-ones
(hydrates). Herein, we report the synthesis, reactivity and characterization of three hemiacetals (1,
3, and 5) and three hydrates (2, 4, and 6) as well as the single crystal X-ray structures of 1-4 & 6.
While the hemiacetals were shown to be kinetically accessible, the hydrates were
thermodynamically preferred. The single X-ray structures of compounds 1-4 & 6 reveals the
alkoxy groups occupying the axial position of the coumarin ring, which is surmised to be the source
of the instability of the hemiacetals. As a result of the axial strain, the hemiacetals (1, 3, and 5)
readily convert in to their respective hydrates (2, 4, and 6) when heated in non-alcoholic solvents.
The qualitative rate of these conversion was largely a function of the size of the alkoxy group (i-
Pr > Et > Me). We propose a mechanism for the conversion of hemiacetals to hydrates that involves
a chromane-3,4-dione intermediate, which was evidenced by trapping it as a diimine, quinoxaline
heterocycle.
Keywords: 3-hydroxyflavone, copper(II) bromide, hemiacetal, hydrate, oxidation, chromane-
3,4-dione
S1

1. Introduction:
Quercetin, a naturally occurring yellow flavan-based pigment commonly found in plants, can
be degraded by dioxygenation via the enzyme quercetinase using its Cu(II) bioactive site to form
depside with concomitant release of carbon monoxide.[1, 2] The active site of quercetinase has
been shown to have two geometric forms (tetrahedral and trigonal bipyramidal), supported by
histidine N-coordination and glutamate O-coordination.[3-5] Flavone‒the base structure of
quercetin‒forms similar chelates to copper(II), often as bis-flavonate copper complexes (Cu(Fla)2,
Fla = Flavonate).[6] Interest in the study of flavones has been spurred on by their antineoplastic
and cytotoxicity activity.[7] We have been interested in preparing biomimetic models of
quercetinase using tripodal chelate ligands such as bis(pyrazolyl)acetate[8] and tris-
imidazolylphosphine [9] to prepare 3-hydroxyflavone copper(II)-chelate complexes. The ultimate
goal of this project was to develop photonically driven carbon monoxide releasing molecules
(PhotoCORM) for therapeutic purposes.[10-12] Therapeutic applications of carbon monoxide
have included the treatment of ailments such as idiopathic pulmonary fibrosis, pulmonary
hypertension, chronic obstructive pulmonary disease and other respiratory diseases.[13]
In our ongoing study to prepare a copper-based PhotoCORM, it was found that the reaction
of 3-hydroxyflavone with copper(II) bromide (CuBr2) in alcoholic solvents led to the oxidation of
the flavone to form hemiacetals and hydrates (Figure 1).
S2

O4
HO2
6
9
R'O3
4
7
8
3
5
1
15
2
14
13
O1
10O5R
11
12
1
2
3
4
5
R, R' = : Me, Me; : Me, H; : Et, Et; : Et, H; : i-
6
Pr, i-Pr; :i-Pr, H
Figure 1. Structure of synthesized hemiacetals and hydrates 1-6.
A number of articles have disclosed the copper catalyzed oxidation of quercetin and
flavones using copper(II) and copper(I) salts,[14-17] which showed formation of hemiacetals and
hydrates when the reactions are performed in alcohols; however, conducting these reactions in
non-protic solvents such as acetonitrile led to the decomposition of flavones to generate a number
of fragmented species including carbon monoxide.[16, 17] In all cases, the authors suggested the
reactions proceeded through a chromane-3,4-dione cationic intermediate, where the 3-hydroxyl
group was oxidized and a carbocation was formed at the 2-position (Scheme 1). This leads to the
nucleophilic attack by alcohol at the cation and either alcoholysis or hydrolysis of the ketone to
form hemiacetals or hydrates, respectively. Mechanistic studies have shown the reaction to be
catalytic with one equivalent of exogenous oxygen consumed to oxidize inactive Cu(I) to active
Cu(II).
S3

O
O
O
O
OH
Ph
O
Cu(II)
HOR
Ph
O
O
O
O
OH
OR
HOR
Ph
O
Ph
OR
OR
Scheme 1. Proposed mechanism for the Cu(II) oxidation of flavones in alcohol.
The mechanism of this reaction is relevant to our understanding the mechanism of the
observed reactivity in this study involving the conversion of hemiacetals (1 & 3) in to hydrates (2
& 4). The qualitative rates of these conversions appear to correlate to the sterics of the alcohols.
Moreover, to our knowledge, the single crystal X-ray diffraction structural characterization of
compounds 1-4 & 6 has not been reported and have not been fully characterized spectroscopically.
Herein, we report the synthesis, characterization and single crystal X-ray structures of 3-hydroxy-
2,3-dimethoxy-2-phenylchroman-4-one (1), 3,3-dihydroxy-2-methoxy-2-phenylchroman-4-one
(2),
3-hydroxy-2,3-diethoxy-2-phenylchroman-4-one (3), 3,3-dihydroxy-2-ethoxy-2-
phenylchroman-4-one (4), and 3,3-dihydroxy-2-isopropoxy-2-phenylchroman-4-one (6). We also
1
reported the in situ H NMR spectroscopic characterization of 3-hydroxy-2,3-diisopropoxy-2-
phenylchroman-4-one (5), which was difficult to isolate from 6. Notably, these compounds are the
first structurally characterized hydrates and hemiacetals of 3-hydroxyflavone.
2. Materials and Methods
2.1 General:
S4

3-hydroxyflavone and copper(II) bromide were used as received without further purifications.
Ortho-phenylenediamine was purified by crystallization from chloroform. All alcohols were
1
stored over molecular sieves. H NMR spectra were recorded on a JEOL ECS300 FT 300 MHz
NMR spectrometer. All 1H spectra are referenced against residual proton signals (1H NMR) and
chemical shifts are recorded in ppm (δ). Coupling constants (JHH values) are reported in Hz.
Infrared spectra were recorded on a Perkin Elmer Spectrum-100 FTIR using ATR methods. Gas
chromatography–Mass Spectrometry (GC–MS) samples were recorded on an Agilent
Technologies 7890GC system using an Agilent 191015-44 column and mass spectra were recorded
on a 5975CVLMSD quadruple mass spectrometry system with a Triple-Axis detector. All GC–
MS samples were introduced via split injection with a two-minute hold time at 50°C and a
temperature ramp of 15°C/minute over a range of 50–250°C. Single crystal X-ray diffraction
measurements were made on a Rigaku XtaLAB mini diffractometer
2.2 Synthesis of 3,3-dihydroxy-2-alkoxy-2-phenylchroman-4-ones and 3-hydroxy-2,3-dialkoxy-2-
phenylchroman-4-ones
2.2.1 Method A. To improve the yield of 3-hydroxy-2,3-dialkoxy-2-phenylchroman-4-ones
(compounds 1, 3, and 5), the following procedure was performed. Copper(II) bromide (467 mg,
2.09 mmol) was dissolved in appropriate alcohol (50 mL) in a 100 mL round bottom flask, then 3-
hydroxyflavone (500 mg, 2.10 mmol) and water (30 µL, 2.10 mmol) were added. A rubber septum
was attached and an oxygen filled balloon equipped with a needle was inserted in the septa. The
reaction was allowed to stir at room temperature for two weeks. The solution was filtered over
Celite and washed with applicable alcohol. The solution was transferred to a round bottom flask
and all solvent was removed by rotary evaporation with heat. All solids were broken up with a
S5

spatula. Pentane and a magnetic stir bar were added and the slurry was stirred for 5 minutes. The
slurry was added to a separatory funnel and extracted three times with a saturated solution K2CO3.
The organic layer was collected, dried with MgSO4 filtered, transferred to a round bottom flask and
all solvent removed under reduced pressure with heat to give off-white solids.
2.2.2 Method B: To improve the yield of the 3,3-dihydroxy-2-alkoxy-2-phenylchroman-4-ones
(compounds 2, 4, and 6), the following procedure was performed. Copper(II) bromide (467 mg,
2.09 mmol) was dissolved in applicable alcohol (50 mL) in a 100 mL round bottom flask, then 3-
hydroxyflavone (500 mg, 2.10 mmol) and water (30 µL, 2.10 mmol) were added. The reaction was
refluxed overnight. The solution was then transferred to a separatory funnel and water (100 mL)
was added. The solution was extracted twice with diethyl ether (50 mL). The organic extracts were
combined and extracted with aqueous solution of K2CO3 (1 M). The organic solution was collected,
dried with MgSO4, filtered, and all solvent was removed by rotary evaporation to give a pale-yellow
solid. The solid was then added to a flask, equipped with a stir bar, and pentane (50 mL) was added.
The pentane solution was decanted into a round bottom and all solvent was removed in vacuo,
which contained the hemiacetals. The remaining insoluble solid was dissolved in diethyl ether (50
mL) and extracted three times with a saturated K2CO3 solution (1 M, 15 mL each) to remove
remaining unreacted 3-hydroxyflavone. The combined organic extracts were dried with MgSO4,
filtered, and all solvent was removed by rotary evaporation to give a pale-yellow solid of the
hydrates.
2.3 Spectroscopic data and yields of hemiacetals and hydrates (1-6)
2.3.1. 3-hydroxy-2,3-dimethoxy-2-phenylchroman-4-one (1). Yield: Method A: 0.178 g, 44%,
Method B: 0.047 g, 7.47%. 1H NMR (300 MHz, CDCl3) δ: 2.98 (s, 3H, CH3), 3.07 (s, 3H, CH3),
4.72 (bs, 1H, OH), 7.16 (t, 1H, 3JHH =8 Hz, Ar-H), 7.38-7.48 (m, 3H, Ar-H), 7.62 (t, 1H, 3JHH=8
S6

Hz, Ar-H), 7.7-7.8 (m, 2H, Ar-H), 7.92 (d, 1H, Ar-H, 3JHH=8 Hz). 13C NMR (75.57 MHz, CDCl3)
δ: 50.1 (CH3), 50.6 (CH3), 93.4, 106.5, 118.4, 118.5, 122.3, 127.4, 127.9, 128.7, 129.4, 133.2,
137.1, 157.2, 191.1 (C=O). IR ATR cm-1: 3431 (w, OH), 2963 and 2925, (w, C-H), 1708 (m, C=O),
1610 (m, C=C), 1085 and 1002 (vs, C-O).
2.3.2. 3,3-dihydroxy-2-methoxy-2-phenylchroman-4-one (2). Method A: 0.044 g, 27%. Method B:
0.401 g, 67%. 1H NMR (300 MHz, CDCl3) δ: 2.64 (bs, 1H, OH), 3.11 (s, 3H, CH3), 4.89 (s, 1H.
OH), 7.12-7.20 (m, 2H, Ar-H), 7.47-7.56 (m, 3H, Ar-H), 7.63 (dt, 1H, Ar-H, 3JHH =8 Hz, 4JHH = 1
Hz), 7.79-8.85 (dd, 2H, Ar-H, 3JHH =7 Hz, 3JHH = 5 Hz ), 7.96 (dd, 1H, Ar-H, 3JHH =7 Hz, 4JHH =2
Hz) 13C NMR (300 MHz, CDCl3) δ:51.09 (CH3), 90.17, 106.78, 118.44, 118.49,127.79, 128.41,
128.76. 129.99. 132.24, 137.23 157.17, 189. 95 (C=O). IR ATR cm-1: 3502 and 3426 (w, OH),
1682 (s, C=O), 1610 and 1465 (m, C=C), 999 (vs, C-O).
2.3.3. 3-hydroxy-2,3-diethoxy-2-phenylchroman-4-one (3). Method A. 0.253 g, 58%. Method B:
0.051 g, 7.4%. 1H NMR (300 MHz, CDCl3) δ: 0.86 (t, 3H, CH3, 3JHH =7 Hz,), 0.93 (t, 3H, CH3,
3
3JHH =7 Hz), 3.04 (dq, 1H,CH2, JHH=7 Hz, 2JHH=2 Hz), 3.29 (dq, 1H, CH2, 3JHH=7 Hz, 2JHH=2
Hz), 3.33-3.48 (m, 2H, CH2), 4.72 (bs, 1H, OH), 7.12 (t, 2H, Ar-H, 3JHH=8 Hz), 7.39-7.50 (m, 3H,
Ar-H), 7.60 (dt, 1H, Ar-H, 3JHH=8 Hz, 4JHH = 2 Hz), 7.71-7.80 (m, 2H, Ar-H), 7.91 (dd, 1H, Ar-H,
3JHH=6 Hz, 4JHH=2 Hz). 13C NMR (75.57 MHz, CDCl3) δ: 15.05 (CH3), 15.18 (CH3), 58.04 (CH2),
59.18 (CH2), 93.20, 106.67, 118.31, 118.72, 122.02, 127.29, 127.69, 128.65, 129.18, 134.18,
136.86, 157.50, 191.84 (C=O). IR ATR cm-1: 3495 and 3428 (w, OH), 2964 (w, C-H), 1694 (s,
C=O), 1610 and 1465 (s, C=C), 1078 and 1004 (vs, C-O).
2.3.4. 3,3-dihydroxy-2-ethoxy-2-phenylchroman-4-one (4): Method A: (N/A) Method B: 0.198 g,
31.43%. 1H NMR (300 MHz, CDCl3) δ: 0.95 (t, 3H, CH3, 3JHH=9 Hz), 2.55 (bs, 1H, OH), 3.40 (q,
3
3
3
2H, CH2, JHH=9 Hz), 4.88 (s, 2H, OH), 7.15 (t, Ar-H, JHH=9 Hz), 7.48 (d, Ar-H, JHH=3 Hz),
S7

7.50 (d, Ar-H, 3JHH=3 Hz), 7.62 (t, Ar-H, 3JHH=9 Hz), 7.80 (d, Ar-H, 3JHH=3 Hz), 7.82 (d, Ar-H,
3JHH=3 Hz), 7.95 (s, Ar-H), 7.97 (s, Ar-H). 13C NMR (75.57 MHz, CDCl3) δ: 15.06 (CH3), 59.73
(CH2), 90.24, 106.74, 118.45, 122.45, 127.76, 128.37, 128.66, 129.92, 133.09, 137.20, 190.08
(C=O). IR ATR cm-1: 3386 and 3496 (w, OH), 1695 (m, C=O), 1607 and 1464 (s, C=C), 1087
and 1002 (s, C-O).
2.3.5. 3-hydroxy-2,3-diisopropoxy-2-phenylchroman-4-one (5):Method A (crude yield determined
by 1H NMR, 15.13%). 1H NMR (300 MHz, CDCl3) δ: 0.64-074 (m, 9H, 3 overlapping CH3 of i-
Pr), 1.02 (d, 3H, CH(CH3), 3JHH= 6 Hz), 3.71 (sept., 1H, CH(CH3), 3JHH= 6 Hz), 3.88 (sept., 1H,
CH(CH3), 4.71 (bs, 1H, OH), 7.01-71.6 (m, 2H, Ar-H), 7.37-7.45 (m, 3H, Ar-H), 7.58 (t, 1 H, p-
Ph, 3JHH= 8 Hz), 7.79-7.97 (m, 3H, Ar-H).
2.3.6. 3,3-dihydroxy-2-isopropoxy-2-phenylchroman-4-one (6). Yield Method A: 0.059 g, 34%.
1
3
Method B: 0.346 g, 50%. H NMR (300 MHz, CDCl3) δ: 0.75 (d, 3H, CH(CH3), JHH=6 Hz), 1.03
(d, 3H, CH(CH3), 3JHH=6 Hz), 2.54 (bs, 1H, OH), 3.88 (sept, 1H, CH(CH3), 3JHH=6 Hz), 4.95 (s,
3
3
2H, OH), 7.12 (t, Ar-H, JHH=8 Hz), 7.41-7.52 (m, 3H, Ar-H), 7.61 (ddt, 1H, Ar-H, JHH=8 Hz,
4
4JHH=2 Hz, JHH=1 Hz), 7.86-7.94 (m, 2H, Ar-H).13C NMR (300 MHz, CDCl3) δ: 23.31
(CH(CH3)), 23.89 (CH(CH3)), 68.53 (CH(CH3)), 90.31, 107.19 , 118.12, 118.56, 122.18, 127.71,
128.15, 129.03, 129.92, 133.70, 137.13, 157.84, 190.03 (C=O). IR ATR cm-1: 3421 and 3369 (w,
OH), 2975 (C-H), 1708 (m, C=O), 1605 and 1462 (s, C=C), 1078 and 995 (vs, C-O).
2.4. 1H NMR monitoring of the conversion of 3-hydroxy-2,3-dialkoxy-2-phenylchroman-4-one to
3,3-dihydroxy-2-alkoxy-2-phenylchroman-4-one. The respective 3-hydroxy-2,3-dialkoxy-2-
phenylchroman-4-ones (1 or 3) (10 mg, 1 = 0.033 mmol; 3 = 0.030 mmol) and CuBr2 (< 1mg)
S8

were dissolved in CDCl3 and added to an NMR tube. A mesitylene sealed insert and H2O (1 molar
equivalent) were then added. The NMR tube was placed in an oil bath at 60 °C. A 1H NMR was
recorded every hour over five hours. In the case of 1, which showed significantly slower
conversion, a 1H NMR spectrum was recorded every day for five days.
2.5 Trapping of chromane-3,4-dione intermediate: Synthesis of 6-ethoxy-6-phenyl-5-oxa-7,12-
diaza-6H-tetraphene (7) (Equation 3): Ortho-phenylenediamine (8 mg, 0.0739 mmol) was
1
dissolved in CDCl3 and added to an NMR tube. An initial H NMR spectrum was recorded.
Compound 3 (11.69 mg, 0.0356 mmol) and a catalytic amount of CuBr2 (<1mg) was added and
1
1
another H NMR spectrum was recorded. The NMR tube was placed in a 60 °C oil bath and H
NMR spectra was recorded every hour for seven hours. The NMR tube was heated at 60 °C for a
total of 24 hours and another 1H NMR was recorded, which revealed the starting materials were
1
completely consumed and the formation of 7. H NMR (300 MHz, CDCl3) δ: 1.06 (t, 3H, CH3,
3JHH =7 Hz), 3.66 (m, 2H, CH2), 7.20-7.31 (m, 2H, Overlapping B-ring Ar-H 2 and p-Ph), 7.41-
3
7.55 (m, 4h, Overlapping B-ring Ar-H 3/4, and m-Ph), 7.62 (t, 1H, A-ring Ar-H9/10, JHH =7 Hz),
3
3
7.72 (t, 1H, A-ring Ar-H 9/10, JHH =7 Hz), 7.80 (d, 2H, o-Ph, JHH =8 Hz), 7.96 (d, 1H, A-ring
Ar-H8/11 , 3JHH =7 Hz), 8.11 (d, 1H, A-ring Ar-H8/11, 3JHH =7 Hz), 8.54 (d, 1H, B-ring Ar-H1, 3JHH
=7 Hz). IR ATR cm-1: 3065, 2975, 2924, 2850 (m, C-H), 1609, 1592, 1555 (m, C=C), 1489 and
1458 (C-N), 1070 1027, 989 (s, C-O). GCMS (70 eV, EI): m/Z(%) 355.15 (M+, 0.43) 281.15
(0.34), 221.15 (0.92), 206.10 (100), 179.10 (37.65) 50.00 (14.93), 76.05(27.39), 103.10 (16.41).
S9

2.6. Crystal Growth and X-ray diffraction:
Compounds 1-4 & 6 were crystalized from pentane by slow evaporation at room temperature over
several days to yield large colorless crystals. X-ray diffraction quality crystals of each compound
were mounted on a Mitigen micromount. Single crystal X-ray diffraction data were measured at
low temperature (173 K) for compounds 2, 3, and 4 and at room temperature (298K) for
compounds 1 and 6. All data were collected on a Rigaku XtaLAB mini CCD diffractometer using
a graphite monochromated Mo K radiation ( = 0.71075 Å). The crystal data are listed in Table
1.
The structures were solved using direct methods and refined using full matrix least squares
using SHELXT and SHELXL. [18, 19] Full-matrix least-squares/difference Fourier cycles were
performed that located the remaining non-hydrogen atoms. All non-hydrogen atoms were refined
with anisotropic displacement parameters. Hydrogen atoms were placed in ideal positions and
refined as riding atoms with relative isotropic displacement.
Crystallographic data of the structures for compounds 1-4 & 6 have been deposited with
the Cambridge Crystallographic Data Center with Deposition Numbers 1986286-1986290,
respectively. Copies of the data are available from CCDC, 12 Union Road, Cambridge, CB2 1EZ,
UK
(fax:
+44-1223-336033;
email:
deposit@ccdc.cam.ac.uc.uk
or
www:
http://www.ccdc.cam.ac.uk).
3. Results and Discussion:
3.1 Synthesis of compounds 1-6: Our initial attempts to oxidize 3-hydroxyflavone with copper(II)
bromide (CuBr2) in various alcohols (methanol, ethanol, isopropanol) at room temperature for 24
S10

hours produced a mixture of 3-hydroxy-2,3-dialkoxychromanones (hemiacetals 1, 3, and 5) or
3,3-dihydroxy-2-alkoxy-chromanones (hydrates 2, 4, and 6) with varying yields of either class of
compound depending on polarity of alcohol(eq 1). The yield of the products from reaction-to-
reaction were often inconsistent and low. On average, compound 1 was isolated in 27% yield,
while its hydrate 2 was isolate in only 5%. Reactions with ethanol produced a nearly 50/50 mixture
of the hemiacetal and hydrates (Compound 3 (17%), Compound 4 (15%)). With isopropanol, the
reaction yielded nearly exclusively the hydrate product (6) (34% yield). Compound 5 could not be
isolated but was detected in small amounts (less than 5%) by 1H NMR spectroscopy from crude
samples.
O
OH
OR
O
OR
O
1 3
5
( , , and )
OH
CuBr2, ROH
24h, rt
+
(eq 1)
O
O
OH
OH
O
OR
2 4
6
( , , and )
1
2
3
4, R
5 6
= Et; and , R = i-Pr
and , R = Me; and
The initial color of the reaction solutions varies with the polarity of the alcohol, which is likely the
result of the alcohol coordinating with CuBr2. The color of the solutions changes with alcohol
identity. Methanolic solutions appear amber in color, ethanol appearing umber, and isopropanol
was dark green. In all reactions conducted at room temperature, the starting 3-hydroxyflavone
remains, even after 21 days of reactivity, suggesting a thermodynamic equilibrium had been
S11

reached. The unreacted flavone can be removed from the hemiacetal/hydrate mixture by base
extraction with an aqueous potassium carbonate solution.
The yield of the hemiacetals, compounds 1, 3, and 5 (44%, 58%, 15%, respectively), is improved
by reaction of 3-hydroxyflavone with copper(II) bromide (CuBr2) in either methanol, ethanol, and
isopropanol at room temperature with exogenous oxygen for 14 days at room temperature
(Scheme. 2). The yield of hydrates, 2, 4, and 6 can be improved by refluxing the reaction overnight
(67%, 32%, 50%, respectively) (Scheme 2).
O
OH
OH
CuBr2, ROH,
H2O, reflux, 24 h
O
O
O
O
OR
OH
OH
OR
CuBr2, ROH, O2,
H2O, rt, 14 d
O
OR
Scheme 2. Different conditions to preferably synthesize hydrates (top) and hemiacetals (bottom).
While Takada previously reported the synthesis of 1, 2, 3 and 4;[14] studies with isopropanol were
derelict. Unlike the reactions with methanol and ethanol in this study, where both the hydrates and
hemiacetals were isolated, only the hydrate, 3,3-dihydroxy-2-isopropoxychromanone (6), could be
isolated when 3-hydroxyflavone was reacted with isopropanol and copper(II) bromide (CuBr2)
(yield) over a 24 hour period at room temperature. The yield of 6 can be improved by conducting
the reaction under reflux overnight. Its hemiacetal (5) could not be isolated but was detected by
1H NMR spectroscopy in crude product from reactions conducted at room temperature (vide infra).
S12

1
3.2 H NMR spectroscopic characterization of compounds 1-6: Clear indication of methoxy
1
hemiacetal (1) was determined with H NMR spectroscopy by the two singlets at 2.99 and 3.07
ppm for the two CH3 groups and the broad singlet at 4.73 ppm attributed to the OH group, whereas
the methoxy hydrate (2) is exemplified by a single singlet at 3.11 pm for the CH3 group and two
broad resonances at 2.64 and 4.89 ppm for the OH groups. Despite the geminal alcohols sharing
the same carbon, their chemical shifts differ by almost 2 ppm. This is likely due to the hydrogen
bonding interaction of the equatorial OH (see X-ray discussion) with the carbonyl, causing the OH
resonance to shift down field.[20]
1
The H NMR spectrum of the ethoxy hemiacetal (3) shows two triplets at 0.86 and 0.93
ppm for the CH3 groups. The two CH2 groups of the ethyl groups are diastereotopic and appear a
complex set of resonances in the range of 3.0-3.5 ppm, which appear as two well-separated
doublets of quartets at 2.89 ppm and 3.29 ppm (1H each) and overlapping resonances between
3.36-3.48 ppm (2H). The OH group resonance appears at 4.68 ppm. The 1H NMR spectrum of the
ethoxy hydrate (4) shows a distinctive triplet at 0.95 ppm for the CH3 group and the methylene
group is again diastereotopic, appearing as a complex multiplet at 3.25-3.45 ppm. The two OH
groups appear as broad resonances at 4.85 ppm and 2.55 ppm.
1
While we could not isolate and purify compound 5, it could be detected by H NMR
spectroscopy from the crude reaction mixture of reactions conducted under prolong times with
exogenous oxygen at room temperature. Clear evidence for the formation of the isopropoxy
hemiacetal (5) could be had by the two septets (1H) at 3.71 and 3.90 ppm for the methine CH
groups. The upfield section of the 1H NMR spectrum of 5 shows two sets of multiplets at 0.72 and
1.0 ppm that integrate for 9 and 3 protons, respectively. This clearly shows that one of the CH3
groups of the chemically distinct isopropyl groups is experiencing a unique deshielding
S13

environment. The OH group appears as a broad singlet at 4.71 ppm. (Supporting information,
Figure 11). Attempts to isolate 5 from 6 by crystallization failed. Even after serial recrystallizations
of the mixtures, only the hydrate (6) was isolated. We ascribed this phenomenon to the rapid
conversion of 5 to 6 in non-alcoholic solvents, which prevents its isolation by crystallization.
1
The isopropoxy hydrate (6) was characterized by H NMR spectroscopy and shows two
doublets (3H each) at 0.74 and 0.99 ppm for the two CH3 groups, a septet (1H) at 3.85 ppm for the
methine group and two broad singlets at 2.54 ppm and 4.94 ppm for the two OH groups.
3.3. Conversion of hemiacetals (1 and 3)to hydrates (2 and 4 ): It was serendipitously found that
the hemiacetals (compounds 1 & 3) gradually convert in to hydrates (compounds 2 & 4) when
stored in CDCl3 over several days. Compounds 1 and 3 were intentionally converted in to their
hydrates (2 and 4, respectively) by heating in CDCl3 at 60 °C with a catalytic amount of CuBr2 (eq
2).
O
O
O
OH
OR
OH
OR
OR
OH
CuBr2,
CDCl3, 60C
(eq 2)
O
1
3
2
4
&
&
1
2
3
4
& , R= Me; & , R= Et
1H NMR spectroscopic monitoring was used to determine the qualitative rates of their conversion
by observing the loss of the integration values for the two well-resolved CH3 resonances in 1 and
3 and the growth of the single CH3 resonance in 2 and 4. A series of representative 1H NMR spectra
are shown in Figure 2 for the conversion of compound 1 in to compound 2. Refer to supporting
1
information for representative H NMR spectra of the monitoring of the conversion of 3 in to 4
S14

(Figure S12). While the conversion of 1 in to 2 required more than five days for complete
reactivity (Figure 2), the conversion of 3 in to 4 required less than three hours. This suggests the
hemiacetals (compounds 1 & 3) are kinetically accessible products and the hydrates (compounds
2 & 4) are more thermodynamically favored. Takeda studies also showed that the oxidation of
quercetin and 3-hydroxyflavone followed a similar reaction pathway, wherein hemiacetals are
formed first, followed by the formation of the hydrates.[14] Further evidence of this assertion is
that 1 was formed in higher yield at room temperature, while 2 is better isolated under refluxing
conditions. The increased rate of conversion of 3 as compared to 1 is likely attributed to the steric
repulsion of the alkoxy groups residing in axial position. We discuss this in further detail in the X-
ray structure discussion section (3.2).
Figure 2. 1H NMR monitoring of reaction of 1 with catalytic CuBr2 in CDCl3 at 60 °C, showing
the conversion of 1 in to 2 over five days.
3.2 Single Crystal X-Ray structural analysis:
S15

Slow evaporation of pentane solutions of 1-4 and 6 affords large, colorless, X-ray quality crystals.
We could not crystalize 3-hydroxy-2,3-diisopropoxy-chromanone (5) (vide supra). While the
synthesis of compounds 1-4 have been reported,[14] to our knowledge their structural analysis by
single crystal X-ray crystallography has not been. Compound 6 has not been reported, neither its
synthesis nor characterization. The X-ray diffraction experimental details, including crystal
systems and space groups, is shown in Table 1. Each structure is discussed in detail, however due
to the similarities in the structures, only representative examples are shown for the hemiacetals, 3-
hydroxy-2,3-dialkoxychromanone (1) (Figure 3a), and the hydrates, 3,3-dihydroxy-2-
alkoxychromanone (4) (Figure 4a). These two examples not only show representative structures
for the two classes of compounds isolated in this study but they also provide a glimpse into their
different packing schemes. Images of the X-ray structures of 2, 3 and 6 are included in the
supporting information (Supporting information, Figures S13-S15).
Table 1. X-ray experimental details for compounds 1-4 & 6.
Crystal data
1
2
3
4
6
Chemical
formula
Mr
C17H16O5
C16H14O5
C19H20O5
C17H16O5
C18H18O5
300.3
286.27
328.35
300.3
314.32
Crystal
Monoclinic,
Monoclinic,
Monoclinic,
Monoclinic,
Triclinic,
system, space
group
P21/n
P21/c
C2/c
P21/n
P-1
a, b, c (Å)
α, β, γ (°)
13.405 (13),
6.910 (6),
15.761 (15)
7.822 (5),
16.605 (6),
10.284 (4)
16.3348 (9),
17.5697 (8),
12.1713 (7)
10.938 (8),
8.036 (6),
17.059 (12)
6.0672 (6),
11.4284 (19),
11.9424 (16)
90, 95.196 (9), 90, 90.32 (3), 90, 108.269 (6), 90, 95.165 (9), 79.03 (4), 78.74 (4),
90
90
90
90
78.90 (4)
V (Å3)
1454 (2)
4
1335.7 (11)
4
3317.1 (3)
8
1493.3 (18)
4
786.8 (3)
2
Z
S16

Rint
0.067
0.064
0.032
0.051
0.118
The hemiacetals 1 and 3 both crystalize in the monoclinic crystal system, but in different space
groups (compound 1 = P21/n; compound 3 = C2/C), and show a number of similarities. For
example, both show the two alkoxy groups occupying the axial position in a trans-configuration
with nearly a 180° dihedral angle. This configuration is obviously adopted to avoid steric
interactions between themselves, but it also avoids the bulkier phenyl group from being in the axial
position. Both compounds also crystalize as a pair of enantiomers, thus revealing the reaction is
not stereoselective. Moreover, both X-ray structures show the two enantiomers associate with
each other to form a dimeric complex using hydrogen bonding between the carbonyl (O4 donor)
and the HO group (H2 acceptor). Similarly, the structures of both compounds 1 and 3 show the
equatorial HO group (O2-H2) orienting itself in a nearly coplanar configuration with carbonyl
oxygen atom (O4) (θ°; O2-C3-C4-O4: 1 = 15.83°, 3 = 22.50°). While the dihedral angle is slightly
larger in 3, it is dissimilar to the expected dihedral angle (theoretical 60°).
S17

a)
b)
Figure 3. Displacement ellipsoid plot of the of 3-hydroxy-2,3-dimethoxy-2-phenylchroman-4-one
(1) (30% probability level) (a) and the dimeric association of the two enantiomers (b). Hydrogen
atoms on carbons have been removed for clarity.
Hydrates, 2, 4, and 6, were characterized by single crystal X-ray crystallography. While 2 and 4
crystalize in the monoclinic crystal system, they have different space groups (2 = P21/c; 4 = P21/n).
Compound 6 crystalizes in the triclinic crystal system in the P-1 space group. The single X-ray
structure of compound 4 was chosen as a representative example (Figure 4a) due to its unique
lattice packing. Compounds 2, 4, and 6 all show the alkoxy group in the axial position with the
phenyl group sitting in the equatorial position. The packing of 2 and 4 reveals a zig-zag network
of alternating enantiomers, orchestrated by hydrogen bonding interactions. For 2, these hydrogen
S18

bonding interactions occur between the carbonyl oxygen (O4) acceptor with one of the geminal
hydroxyl donor (H2-O2) (H∙∙∙∙A = 1.96(2) Å) and the other geminal hydroxyl group (H3-O3) donor
with the methoxy oxygen (O5) acceptor (H∙∙∙∙A = 1.96(2 ) Å) (See supporting information, Figure
S13). In 4, the hydrogen bonding interactions occur between the equatorial –OH (O2) group and
the carbonyl oxygen (O4) to make a dimeric pair, such as that observed in 1 (Figure 3b) and 3
(supporting information, Figure S14); however, it is also augmented with an intermolecular
hydrogen bonding interaction between the axial –OH groups (O3) of independent molecules to
form a zig-zag hydrogen bonding network. Importantly, the ethoxy oxygen (O5) in 4 does not
participate in hydrogen bonding, as was seen in its homologue (compound 2). In the significantly
bulkier 3,3-dihydroxy-2-isopropoxy-2-phenylchroman-4-one (6), the carbonyl is not involved in
hydrogen bonding (See supporting information, Figure S15). Instead, the two enantiomers of 6
form a dimeric complex held together by hydrogen bonding between the germinal alcohols, where
the axial OH group (O2) of one molecule interacts with the equatorial OH group (O3) of its
enantiomer and vice versa.
a)
S19

b)
Figure 4. Displacement ellipsoid plot of 3,3-dihydroxy-2-ethoxy-2-phenylchroman-4-one (4) (a)
(50% probability level), showing the zig-zag hydrogen bonding motif of the two enantiomers (b).
Hydrogen atoms have been removed for clarity.
3.3 Mechanism discussion of the conversion of hemiacetals in to hydrates
In the oxidation of flavone-based compounds, it has been intimated that the reaction proceeds
through a chromane-3,4-dione cationic species, where the alcohol attacks the cation followed by
nucleophilic acyl addition to the 3-ketone by either water or alcohol to form the hydrate or
hemiacetal, respectively (Scheme 1). It was observed in this study that the hemiacetals readily
convert in to the hydrates when stored in non-alcoholic solvents. It is likely the hydrates are
significantly more stable as compared the hemiacetals due to the axial alkoxy groups inducing
steric and torsional strain. We ascribe this reaction phenomenon to the reformation of the 3,4-dione
by elimination of the alcohol from the hemiacetal at the 3-position to form a ketone, which is then
hydrolyzed by residual water (Scheme 3) to form the hydrates. The benzylic ketone (4-ketone)
resists hydration due to resonance stabilization.
S20

O
O
O
O
O
O
OH
OH
H2O
OH
OR
O
-HOR
Ph
OR
Ph
Ph
OR
OR
Hydrate
Hemiacetal
3,4-dione
Scheme 3. Proposed Mechanism for the conversion of hemiacetals to hydrates.
It is well known that imines are readily formed when reacted with amines. In addition, it has been
shown that the reaction of ortho-phenylenediamines with α-diketones can lead to the formation of
quinoxalines.[21] Others have also shown that nickel catalyzed systems can produce a variety of
quinoxalines by dehydrogenative coupling of vicinal diols with ortho-diamines.[22] To test if our
mechanism was operable, it was thought the chromane-3,4-dione intermediate could be trapped by
reaction of ortho-phenylenediamine to form a stable quinoxaline heterocycle. In this aim, the
hemiacetal (3) was reacted with ortho-phenylenediamine in the presence of a catalytic amount of
CuBr2 in CDCl3 and monitored by 1H NMR spectroscopy at 60 °C (eq 3). After 24 hours, a fairly
clean spectrum was recorded showing significant changes that suggest the formation of 6-ethoxy-
1
6-phenyl-5-oxa-7,12-diaza-6H-tetraphene (7) (see Figure 5 for H NMR monitoring). The most
compelling 1H NMR spectroscopic evidence for the formation of 7 was the loss of the OH peak at
4.68 ppm, the generation of free ethanol (3.72 and 1.23 ppm) and the consumption of diamine
(broad singlet at 6.70 ppm). The 2-ethoxy group remains in 7 as identified by the triplet for the
CH3 group (1.09 ppm) and the diastereotopic CH2 group (3.66 ppm), which is well downfield of
those of its corresponding hydrate (4) (multiplet 3.25-3.45 ppm). The methyl group is similarly
1
shifted downfield as compared to 4. Due the complicated aromatic region, a gCOSY H NMR
experiment was conducted on 7, revealing coupling of the two doublets (1H each) at 8.10 ppm and
7.95 ppm to the two triplets (1H each) at 7.73 ppm 7.63 ppm, respectively. The two triplets are
S21

themselves are coupled to each other. These resonances were assigned to the fused aromatic ring
connected to the diimine (A-ring, see equation 4 for numbering and ring designation). The
downfield doublet at 8.53 ppm was assigned to the proton at the 1-position of the B-ring, which is
couple to the upfield resonance at 7.24 ppm (overlapping triplets). The remaining two protons
(protons 3 and 4) of the B-ring are buried in the complicated multiplet centered at 7.48 ppm. The
doublet at 7.81 ppm was assigned to the ortho protons (2H) of the 6-phenyl group. This signal is
coupled to a resonance at 7.45 (overlapped) for the meta-protons, which is coupled to a resonance
at 7.26 ppm for the para-proton (overlap). We also assessed 7 by infrared spectroscopy. For
comparison purposes, the infrared spectrum of compound 3 was recorded, which clearly shows
carbonyl and hydroxyl stretches at 1694 and 3495 cm-1, respectively. These stretching frequencies
are no longer present in the IR spectrum of 7. Furthermore, there are no NH2 stretches in the range
of 3100-3500 cm-1 in the IR spectrum of 7, suggesting both amines have reacted. Compound 7 is
unstable to GC-MS conditions, decomposing into multiple species. However, we were able to
observe a peak a 355.14 m/Z for the parent complex as well as and another larger fragment at 206
m/Z that compared well with 2-phenylquinoxaline in the NIST14 Library, which is also observed
1
in the fragmentation pattern for 355.14 m/Z. These GCMS spectrometric and IR and H NMR
spectroscopic data suggest the successful trapping of the 3,4-dione intermediate as a quinoxaline
heterocycle (7).
S22

10
11
9
A
12 N
O
8
OH
1
NH2
O
N 7
Ph
2
3
NH2
CuBr2, CDCl3,
60C, 24h
+
(eq 3)
B
6
O
Ph
O
O
4
O
5
3
7
Figure 5. 1H NMR spectroscopic monitoring of the reaction of 3 with o-phenylenediamine in the
presence of CuBr2 at 60 °C for 24 hours, observing the formation of the quinoxaline, 7.
3. Summary and Conclusion.
We have reported the synthesis and characterization of three hemiacetals (1, 3, and 5) and three
hydrates (2, 4, and 6) that were isolated from the reaction of 3-hydroxyflavones with CuBr2 in
alcohols. We have shown that the hemiacetals are kinetically accessible while the hydrates are
S23

thermodynamically preferred. Our single X-ray studies revealed the alkoxy groups in these
compounds occupy the axial position to prevent the phenyl group from inducing torsional and
steric strain. We surmise the axial orientation of the alkoxy groups is the source of the instability
of the hemiacetals. This instability allows for the ready conversion of hemiacetals (1, 3, and 5) in
to their respective hydrates (2, 4, and 6). It was observed that the comparable qualitative rates of
this conversion were largely based on the sterics (i-Pr> Et > Me). We suggest that the conversion
of hemiacetals to hydrates follows a mechanism that involves a chromane-3,4-dione intermediate,
which provides evidence trapping it as a diimine quinoxaline heterocycle (7).
Acknowledgements: We thank Georgia Southern University SEED program and Summer
Research Scholars program, the Georgia Power Research Scholar program and the American
Chemical Society Petroleum Research Fund (PRF#53848-UNI3) for partial support of this work.
Declaration of interests: The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influence the work reported in this
paper.
References
[1] S. Hattori, I. Noguchi, Nature, 184 (1959) 1145-1146.
[2] H.K. Hund, J. Breuer, F. Lingens, J. Huttermann, R. Kappl, S. Fetzner, Eur J Biochem, 263
(1999) 871-878.
[3] I.M. Kooter, R.A. Steiner, B.W. Dijkstra, P.I. van Noort, M.R. Egmond, M. Huber, Eur J
Biochem, 269 (2002) 2971-2979.
[4] G.W. Hay, D.W. Westlake, F.J. Simpson, Can J Microbiol, 7 (1961) 921-932.
S24

[5] F. Fusetti, K.H. Schroter, R.A. Steiner, P.I. van Noort, T. Pijning, H.J. Rozeboom, K.H. Kalk,
M.R. Egmond, B.W. Dijkstra, Structure, 10 (2002) 259-268.
[6] E. Balogh-Hergovich, J. Kaizer, G. Speier, G. Argay, L. Parkanyi, J Chem Soc Dalton, (1999)
3847-3854.
[7] J.M. Edwards, R.F. Raffauf, P.W. Lequesne, Lloydia, 42 (1979) 85-91.
[8] B. Quillian, W.E. Lynch, C.W. Padgett, A. Lorbecki, A. Petrillo, M. Tran, J Chem Crystallogr,
49 (2019) 1-7.
[9] W.E. Lynch, D. Nivens, B. Quillian, C.W. Padgett, A. Petrillo, N. Peek, J. Stone, J Mol Struct,
1185 (2019) 99-106.
[10] T. Soboleva, H.J. Esquer, A.D. Benninghoff, L.M. Berreau, J Am Chem Soc, 139 (2017)
9435-9438.
[11] S. Sorenson, M. Popova, A.M. Arif, L.M. Berreau, Acta Crystallogr C, 73 (2017) 703-709.
[12] M. Popova, T. Soboleva, A.M. Arif, L.M. Berreau, Rsc Adv, 7 (2017) 21997-22007.
[13] M. Knauert, S. Vangala, M. Haslip, P.J. Lee, Oxid Med Cell Longev, (2013).
[14] M. Utaka, A. Takeda, J Chem Soc Chem Comm, (1985) 1824-1826.
[15] J. Clark-Lewis, I. Dainis, E. McGarry, M. Baig, Aust J Chem, 25 (1972) 857-864.
[16] A. Gulsen, B. Turan, D.P. Makris, P. Kefalas, Eur Food Res Technol, 225 (2007) 435-441.
[17] E. Baloghhergovich, G. Speier, J Mol Catal, 71 (1992) 1-5.
[18] G.M. Sheldrick, Acta Crystallogr A, 71 (2015) 3-8.
[19] G.M. Sheldrick, Acta Crystallogr C, 71 (2015) 3-8.
[20] M.A. Smith, R.A. Webb, L.J. Cline, J Org Chem, 30 (1965) 995-+.
[21] C. Delpivo, G. Micheletti, C. Boga, Synthesis-Stuttgart, 45 (2013) 1546-1552.
[22] S. Shee, D. Panja, S. Kundu, J Org Chem, 85 (2020) 2775-2784.
S25

Supporting Information
Preparation of 3-hydroxy-2,3-dialkoxy-2-phenylchroman-4-one and 3,3-dihydroxy-2-alkoxy-2-
phenylchroman-4-one by oxidation of 3-hydroxyflavone with Copper(II) Bromide: Structure, reactivity and
characterization
Elizabeth M. Beasley, Joseph G. Bazemore, Anthony Petrillo, Clifford W. Padgett, Will E. Lynch, Brandon Quillian*
Department of Chemistry and Biochemistry, Georgia Southern University, 11935 Abercorn St., Savannah, GA 31419, United States
Email: bquillian@georgiasouthern.edu
Phone:1-912-344-2977
*Corresponding Author
S26

Figure S1. 1H NMR spectrum of 3 -hydroxy-2,3-dimethoxy-2-phenylchroman-4-one (1) in CDCl3 at 300 MHz.
27

S#524789
0.70
0.65
0.60
0.55
0.50
0.45
0.40
0.35
0.30
0.25
0.20
0.15
0.10
0.05
0.00
-0.05
single pulse decoupled gated NOE
220 210 200 190 180 170 160 150 140 130 120 110 100
90
80
70
60
50
40
30
20
10
0
-10
-20
f1 (ppm)
Figure S2. 13C NMR spectrum of 3 -hydroxy-2,3-dimethoxy-2-phenylchroman-4-one (1) in CDCl3 at 75 MHz.
28

Figure S3. 13C NMR spectrum of 3,3 -hydroxy-2-methoxy-2-phenylchroman-4-one (2) in CDCl3 at 300 MHz.
29