Asymmetric sulfoxidation by C1-symmetric V(IV)O(ONO) (S)-NOBIN Schiff-base vanadyl complexes

Article abstract of DOI:10.1139/cjc-2019-0017

C₁-symmetric vanadyl Schiff-base complexes were synthesized by reacting vanadium(IV) acetylacetonate with (S)-3-[(1-(2-hydroxynaphthalen-1-yl)naphthalen-2-ylimino]methyl]-phenanthrene-4-ol and (S)-2-{[1-(2-hydroxynaphthalen-1-yl)naphthalen-2-ylimino]methyl}tetraphene-1-ol. The complexes were characterized by MALDI-TOF-MS, UV-vis, and circular dichroism (CD) spectroscopy. The catalysts showed moderate activity for the oxidation of thioanisole to methyphenylsulfoxide with hydrogen peroxide, tert-butyl hydroperoxide, and cumene hydroperoxide as the oxidants.

Full text of DOI:10.1139/cjc-2019-0017

Page 1 of 14
Asymmetric Sulfoxidation by C1-symmetric V(IV)O(ONO) (S)-
NOBIN Schiff-base Vanadyl Complexes
Sanmitra Barmana, Christopher J. Levyb*
Department of Chemistry, BML Munjal University, 67 Milestone, NH-8, HR-122413, India a
Department of Chemistry, The University of Western Ontario, 1151 Richmond Street, London,
Ontario, Canada, N6A 5B7 b
E-mail: clevy9@uwo.ca & sanmitra.barman@bmu.edu.in
1

Page 2 of 14
ABSTRACT
C1-symmetric vanadyl Schiff-base complexes were synthesized by reacting vanadium(IV)
acetylacetonate
with
(S)-3-[(1-(2-hydroxynaphthalen-1-yl)naphthalen-2-ylimino]methyl]-
and (S)-2-{[1-(2-hydroxynaphthalen-1-yl)naphthalen-2-
phenanthrene-4-ol
ylimino]methyl}tetraphene-1-ol. The complexes were characterized by MALDI-TOF-MS, UV-
Vis, and circular dichroism (CD) spectroscopy. The catalysts showed moderate activity for the
oxidation of thioanisole to methyphenylsulfoxide with hydrogen peroxide, tert-butyl
hydroperoxide and cumene hydroperoxide as the oxidants.
Keywords: C1-symmetric, Schiff-base, Asymmetric Sulfoxidation, Vanadyl, NOBIN
2

Page 3 of 14
Introduction
Chiral sulfoxides are important building blocks in pharmaceutical and fine chemical industries.
The asymmetric catalytic oxidation of prochiral sulfides is an attractive and efficient pathway to
chiral sulfoxides that has garnered significant interest.1-4 There majority of studies have focused
on titanium and vanadium catalysts to affect the oxidation, with hydrogen peroxide or molecular
oxygen as preferred benign oxidants.5 Vanadium complexes have commonly had chiral tridentate
or tetradentate ligands in these systems.6 Previously we have described monohelical tetradentate
(ONNO) vanadyl salen complexes as catalysts for asymmetric sulfoxidations with hydrogen
peroxide.7 In keeping with other mechanistic proposals for non-radical oxidation, dissociation of
one ligand donor was deemed necessary to accommodate peroxide and substrate addition to the
vanadium center during catalysis. In these systems it has been proposed that the sulfide undergoes
direct nucleophilic attack by the hydroperoxo ligand (Fig. 1).6
Fig. 1. Direct HOO– nucleophilic attack (left, O to si face) and S-V precoordination (right, O to
re face) reaction models for vanadyl complex catalyzed thioanisole oxidation with hydrogen
peroxide.
Discussion
Tridentate ligands offer a more open coordination sphere and opportunity for direct sulfide
coordination to the metal center, as described by Romanowski (Fig. 1).8 This promotes stronger
3

Page 4 of 14
and more predictable interaction with the ligand framework. Notably, the dominant enantiomer
produced is reversed for the two mechanisms given similar steric topographies.
The majority of vanadium complexes with tridentate ligands that have been reported
incorporate Schiff-base and phenoxide/alkoxide donors in an ONO pattern.5 In this article, we
describe the syntheses of vanadyl complexes whose ligands are based on Schiff-base condensation
of a polyaromatic aldehyde to (S)-2-amino-2'-hydroxy-1,1'-binaphthyl [(S)-NOBIN]: these are the
first vanadium complexes based on NOBIN to be reported. Previously, we have described
titanium(IV) complexes of these ligands.9 Thus, the synthesis and characterization of (S)-1 and
(S)-2 have been reported.8
The new vanadyl (S)-NOBIN derived complexes (S)-3 and (S)-4 were produced from
VO(acac)2 under forcing conditions. Refluxing with the corresponding ligands (S)-1 or (S)-2 in the
presence of excess base (10 eq. NaOMe) were required to produce adequate yields. The best results
were obtained with 6 h of reflux in a 2:1 mixture of dry ethanol and dichloromethane. The
complexes were isolated as brown powders after Soxhlet extraction and solvent removal in vacuo
(Scheme 1). The compounds were air sensitive: solutions exposed to air darkened rapidly. The
presumed oxidation products are µ-oxo bridged OV(V)–O–V(V)O dimers, such as those
crystallographically-characterized by Correia et al. for related complexes under similar
conditions.10
4

Page 5 of 14
Scheme 1. Synthesis of V(IV) complexes (S)-3 and (S)-4
1
The vanadyl complexes showed broad H NMR spectra, consistent with the presence of
paramagnetic V(IV) centers rather than d0 V(V) complexes. Despite this, the chemical shifts are
comparable to those for diamagnetic complexes, consistent with previous observations.11 The
MALDI-TOF-MS spectra show the [M]+ ion peaks with the expected isotopic mass distributions.
No dinuclear species were detected in the MS, pointing to mononuclear complexes. The ligands
and complexes showed electronic absorption and CD spectra characteristic of 2,2́ substituted 1,1́
binaphthyls (Fig. 2).12 Three absorption bands are observed, which corresponds to three basic
absorption bands of β-mono-substituted naphthalene, and they can be conventionally assigned to
1
1
1
three transitions of naphthalene: Bb, La, and Lb
The intensities
in order of decreasing energy.
and signs of the absorptions in the CD spectra follow the pattern expected for chromophores in the
1
(S)-NOBIN unit. The is significant red-shifting (ca. 20 nm) of the Bb band for the complexes
compared to the ligands: (S)-1  (S)-3, 232  253 nm; (S)-2  (S)-4, 236  253 nm; negative
1La 282  289 nm. This reflects delocalization of frontier orbitals to both the metal center and
the polyaromatic unit. This latter effect is the result of a reduction in the naphthyl to
phenanthryl/benz[a]anthryl angle upon ligand coordination. The 267 nm bands of (S)-1 and (S)-2
5

Page 6 of 14
1
are due to the phenanthryl/benz[a]anthryl units. These become shoulders (253 nm) on the Bb
bands in the complexes. 9
Fig. 2. Electronic CD spectra from THF solutions of the ligands (red) and vanadyl complexes
(black).
Preliminary studies were carried out with (S)-3 and (S)-4 as catalysts for the oxidation of
thioanisole with three oxidants: hydrogen peroxide, tert-butyl hydroperoxide and cumene
hydroperoxide. The reaction conditions were optimized for solvent, temperature, oxidant and
reaction time. Selectivity was solvent dependent, with improved selectivity in methylene chloride.
Reactions were homogeneous, except with hydrogen peroxide. In this case the reaction takes place
under biphasic conditions: CH2Cl2 and H2O2 (aq). The reaction temperature was also an important
factor, with 0˚ C producing the best selectivity observed. Conversion of thioanisole to sulfoxide
reached 70% after 4 h and continued to increase up to ca. 12 h. Yields of sulfoxide decreased at
longer reaction times due to over-oxidation to sulfone. However, the selectivity (%ee) did not
change over the course of the reaction, suggesting that the stereochemical outcome is unaffected
6

Page 7 of 14
by any changes in catalyst structures, such as the formation of V(V)–O–V(V) dimers. The substrate
to oxidant ratio was varied to establish the effect of oxidant on conversion and selectivity. There
was a distinct increase in the sulfone production and a decrease in selectivity when excess oxidant
was used.
Scheme 2. Asymmetric sulfoxidation of thioanisole with vanadyl catalysts (major configuration
R-sulfoxide).
The results of catalytic oxidation of thioanisole are presented in Table 1. In all cases the major
enantiomer obtained from (S)-3 and (S)-4 was (R)-methylphenylsulfoxide. Moderate
enantioselectivity (ca. 30%ee) was achieved with H2O2 as the oxidant. In comparison, tert-butyl
hydroperoxide (~15%ee) and cumene hydroperoxide (~20%ee) are less selective oxidants under
the conditions examined. This observation can be attributed to the homogeneous reaction
conditions for TBHP and CHP: oxidant concentrations are higher and non-catalytic oxidation can
become competitive. This phenomenon is supported by the observation that the conversions are
higher for these oxidants in comparison to H2O2.
7

Page 8 of 14
Table 1. Sulfoxidation of thioanisole catalyzed by (S)-3 and (S)-4a
catalyst
oxidant
H2O2
TBHP
CHP
H2O2
TBHP
CHP
conv./%
84
90
92
82
SO/%
81
84
86
82
ee/%
29
16
14
18
SO2/%
19
16
14
18
3
3
3
4
4
4
95
94
84
85
16
15
16
15
a1% cat., 12 h at 0 ˚C in CH2Cl2. bSO: PhSOMe. cEnantiomeric excess measured by HPLC with
DIACEL OD-H column. dSO2: PhSO2Me
The proposed mechanism for catalytic sulfoxidation is consistent with that previously
presented for other vanadyl complexes with tridentate ligands.8 A key sequence is the coordination
of sulfide to the vanadyl center followed by oxygen transfer from the hydroperoxo group. The
stereochemical outcome is determined by whether the re or si prochiral face of the sulfide receives
the oxygen (Fig. 3). Selectivity for the R-sulfoxide is consistent with preferred delivery of oxygen
to the si face due to reduced steric repulsion between the sulfide and the polyaromatic ligand unit.
The moderate enantioselection suggests only modest facial discrimination. This finding is
supported molecular modelling (MM3) of the two possible intermediates, which indicated that
binding to the si face of thioanisole is favored by 1.38 kJ/mol. This translates to a maximum ee%
of 30%, without considering possible kinetic resolution effects often observed in similar systems.13
8

Page 9 of 14
Fig. 3. Proposed intermediates for thioanisole oxidation by [(R)-3(OOH)] (top) and MM3-
generated space-filling models (bottom).
The MM3 models indicate that si coordination is favored since the S-Me group interacts with
the large phenanthryl group, while the S-Ph group interacts with the smaller naphthyl group. The
larger S-Ph group has closer aromatic centroid-centroid contacts (3.40 & 3.87 Å) for re
coordination, than for si coordination (3.63 Å).
The sulfoxidation results are similar to those for vanadyl salen catalysts we have previously
examined.7 This indicates that the tridentate (ONO) ligand system employed is comparable to the
tetradentate (ONNO) systems. This supports the idea that the second ligand arm of the C2-
symmetrical complexes does not impact selectivity in these systems. One notable difference,
however, is the higher degree of overoxidation with the C1-symmetrical vanadyl NOBIN catalysts.
This can be partially explained by the formation of oxo-bridged dimers in solution due to the open
nature of the coordination sphere. This effectively removes catalyst from the system and allows
nonselective uncatalyzed oxidation of sulfide to compete. By comparison, the bulkier tetradentate
9

Page 10 of 14
ligand systems do not allow for rapid dimer formation, giving the catalytic pathways a chance to
dominate.
Conclusions
The general conclusion of the study is that the complexes did catalyze asymmetric
sulfoxidation, however the anticipated increase in activity compared to previous systems was not
realized. The observed enantioselectivity in this study supports mechanism involving coordination
of sulfide to the metal center before oxygen transfer to sulfur (Figure 4): the opposite selectivity
is expected from direct nucleophilic attack of sulfur by a hydroperoxo ligand. The modeling study
suggest that the maximum possible selectivity (30%ee) has been achieved for the system. The
planar nature of the phenyl group in thioanisole stacks with the planar naphthyl and
phenanthryl/benz[a]anthryl ligand groups in both diastereomeric coordination modes, and neither
is strongly favored. This result points to the possibility of higher enantioselectivity for substrates
with more sterically demanding groups, especially those that are non-planar. Further studies with
a wider array of sulfide substrates along with computational work are planned.
10

Page 11 of 14
H3C
S
O
N
O
+H2O2
– H2O
O
V
OH
O
hydroperoxo
product release
complex
formation
H3C
O
N
O
O
N
O
S
V
V
O
O
O
O
O
O
H
H
RDS
+ PhSMe
oxygen transfer
substrate
binding
H3C
S
O
N
V
O
O
O
O
H
Fig. 4. Proposed mechanism for sulfoxidation reactions catalyzed by (S)-3.
Experimental procedure
General experimental methods
All the reactions were performed under inert atmosphere of nitrogen or argon unless otherwise
mentioned. Solvents used were stored under calcium hydride or sodium benzophenone ketyl and
degassed before using via Schlenk line technique. Synthesis and work up of the vanadyl complexes
were done under rigorous inert atmosphere to avoid unwanted oxidation or decomposition. All
mass spectra are recorded in Shimadzu MALDI TOF-TOF mass spectrometer. IR spectra are taken
on neat samples using a Nicolet 380 FT-IR spectrometer with ZnSe ATR attachment. 1H and 13C
NMR spectra were obtained on a Varian Unity 400 MHz spectrometer with residual solvent
protons or TMS as an internal standard.
Synthesis of complexes
11

Page 12 of 14
(S)-3: To the solution (composed of a mixture of CH2Cl2 (15 mL) and EtOH (7.5 mL)) of (S)-1
(0.100 g, 0.205 mmol) was added vanadyl acetylacetonate (0.054 g, 0.205 mmol) and sodium
methoxide (0.100 g, 2.04 mmol). The solution was refluxed for 6 h followed by filtration to obtain
a brown colored precipitate. The compound was further purified by soxhlet extraction in CH2Cl2.
Pure brown colored compound was obtained after solvent removal under reduced pressure (0.093
1
g, 72% yield). H NMR (CDCl3, 400 MHz): broad peaks suggest paramagnetic compound. MS:
[M+OMe]+1 (C36H24N1O4V1)+1 586.261 (64%), 587.265 (24%), 588.272 (8%), 589.277 (4%). IR:
νV=O 818.87 cm-1, νC=N 1612.72 cm-1. UV: 245 nm sharp high intensity peak is due to the absorption
of the NOBIN backbone and the 275 nm peak is due to the interaction between the naphthyl ring
of the NOBIN backbone and the phenanthryl side arm. L →M charge transfer at 365 nm (broad
peak). CD: The chiral information in the complex is similar to the that of the ligand with CD peaks
have shifted to longer wavelength.
(S)-4: To the solution (composed of a mixture of CH2Cl2 (8 mL) and EtOH (4 mL)) of (S)-2 (0.063
g, 0.116 mmol) was added vanadyl acetyl acetonate (0.031g, 0.116 mmol) and sodium methoxide
(0.063 g, 1.16 mmol). The solution was refluxed for 6 h followed by filtration to obtain a deep
brown colored precipitate. The compound was further purified by soxhlet extraction in CH2Cl2.
Deep brown colored compound was obtained after solvent removal under reduced pressure (0.059
1
g, 74% yield). H NMR (CDCl3, 400 MHz): broad peaks in the aromatic region suggest
paramagnetic compound. MS: [M+OMe]+1 (C40H26N1O4V1)+1 636.169 (66%), 637.174 (26%),
638.183 (7%), 639.189 (1%). IR: νV=O 816.91 cm-1, νC=N 1613.33 cm-1. UV: The two sharp peaks
for the ligand at 225 nm and 275 nm has become broad after complex formation. This shows higher
interaction between the naphthyl rings in the NOBIN backbone and the benz[a]anthryl ring on the
side arm. Characteristic broad peak at 365 nm for more delocalization of the Π electrons from
12

Page 13 of 14
ligand to metal. CD: The chiral information for the complex is similar to that of the ligand with
red shift of the CD peaks.
General procedure for sulfoxidations
All sulfoxidation reactions are carried out by 1 mol% of catalyst. The catalyst was dissolved
in CH2Cl2 (3 mL) inside a two-neck round bottom flask under argon. After stirring for 10 minutes,
thioanisole was added and then the reaction mixture was stirred for 15 minutes. The system was
cooled to 0 ˚C in a regulated bath and 1.1 equivalents of oxidant (hydrogen peroxide, tert-butyl
hydroperoxide, cumene hydroperoxide) was added for a period of 2 h. The mixture was
additionally stirred for 10 h at 0 ℃ after the addition of the oxidant was complete. A saturated
solution of sodium sulfite was added to quench the reaction. The reaction mixture was extracted
with CH2Cl2 (3 × 10 mL) and then combined extracts were dried over sodium sulfate. The solution
was filtered, and the solvent evaporated. The product was purified by flash column
chromatography in 10% ethyl acetate in hexane. Thioanisole, methylphenylsulfoxide and
methylphenylsulfone can be identified from their signature peaks in 1H NMR. The enantiomeric
excess (ee) of the methylphenylsulfoxide was determined by HPLC analysis with a Daicel
Chiralcel OD-H column using hexane and isopropanol mixture (90:10) as the eluent. The R isomer
and the S isomer elutes at 17.2 and 20.8 minutes, respectively.
Acknowledgement
This work was supported by the National Science Foundation (US) (No. CHE-0349258),
Kansas State University and The University of Western Ontario (Canada). CJL would like to thank
Riley Hooper for his assistance with the computational work.
13

Page 14 of 14
Supporting Information is available on http://dx.doi.org/
References and Notes
(1) O’ Mahony, G. E.; Kelly, P.; Lawrence, S. E.; Maguire, A. R. Arkivoc, 2011, 1-110.
(2) Han, J.; Soloshonok, V. A.; Klika, K. D.; Drabowicz, J.; Wzorek, A. Chem. Soc. Rev.,
2018, 47, 1307-1350.
(3) Volcho, K. P.; Salakhutdinov, N. F. Russ. Chem. Rev. 2009, 78, 457.
(4) Wojaczynska, E.; Wojaczynska, J.; Chem. Rev. 2010, 110, 4303.
(5) Bryliakov, K. P. Chem. Rev. 2017, 117, 11406.
(6) Langeslay, R. R.; Kaphan, D. M.; Marshall, C. L.; Stair, P. C.; Sattelberger, A. P.; Delferro,
M. Chem. Rev., 2019, 119, 2128.
(7) Barman, S.; Patil, S.; Desper, J.; Aikens, C. M.; Levy, C. J. Eur. J. Inorg. Chem. 2013,
5708.
(8) Romanowski, G. J. Mol. Cat A 2013, 368-369, 137.
(9) Barman, S.; Desper, J.; Levy, C. J. Polyhedron, 2014, 84, 168.
(10) Adao, P; Pessoa, J.; Henriques, R. T.; Kuznetsov, M. L.; Avecilla, F.; Maurya, M. R.;
Kumar, U.; Correia, I. Inorg. Chem. 2009, 48, 3542.
(11) Pokutsa, A.; Kubaj, Y.; Zaborovskyi, A.; Maksym, D.; Paczesniak, T.; Mysliwiec, B.;
Bidzinska, E.; Muzart, J.; Sobkowiak, A. Mol. Cat. 2017, 434, 194.
(12) Studentsov, E. P.; Piskunova, O. V.; Skvortsov, N. K. Russ. J. Gen. Chem. 2009, 79, 962.
(13) Legros, J.; Dehli, J. R.; Bolm, C. Adv. Synth. Catal. 2005, 347, 19.
(14) NMR, IR, UV-Vis spectra and MS data for the complexes is provided as Supporting
Information which is available electronically.
14