Cobalt-entrenched N-, O-, and S-tridoped carbons as efficient multifunctional sustainable catalysts for base-free selective oxidative esterification of alcohols

Article abstract of DOI:10.1039/c8gc01333k

We report the synthesis of sustainable and reusable non-noble transition-metal (cobalt) nanocatalysts containing N-, O-, and S-tridoped carbon nanotube (Co@NOSC) composites. The expensive and benign carrageenan served as the source of carbon, oxygen, and sulfur, whereas urea served as the nitrogen source. The material was prepared via direct mixing of precursors and freeze-drying followed by carbonization under nitrogen at 900 °C. Co@NOSC catalysts comprising a Co inner core and outer electron-rich heteroatom-doped carbon shell were thoroughly characterized using various techniques, namely, TEM, HRTEM, STEM elemental mapping, XPS, BET, and ICP-MS. The utility of the Co@NOSC catalyst was explored for base-free selective oxidative esterification of alcohols to the corresponding esters under mild reaction conditions; excellent conversions (up to 97%) and selectivities (up to 99%) were discerned. Furthermore, the substrate scope was explored for the cross-esterification of benzyl alcohol with long-chain alcohols (up to 98%) and lactonization of diols (up to 68%). The heterogeneous nature and stability of the catalyst facilitated by its ease of separation for long-term performance and recycling studies showed that the catalyst was robust and remained active even after six recycling experiments. EPR measurements were performed to deduce the reaction mechanism in the presence of POBN (α-(4-pyridyl-1-oxide)-N-tert-butylnitrone) as a spin-trapping agent, which confirmed the formation of CH₂OH radicals and H radicals, wherein the solvent plays an active role in a nonconventional manner. A plausible mechanism was proposed for the oxidative esterification of alcohols on the basis of EPR findings. The presence of a cobalt core along with cobalt oxide and the electron-rich N-, O-, and S-doped carbon shell displayed synergistic effects to afford good to excellent yields of products.

Full text of DOI:10.1039/c8gc01333k

View Article Online
View Journal
Green
Chemistry
Accepted Manuscript
This article can be cited before page numbers have been issued, to do this please use: D. Nandan, G.
Zoppellaro, I. Medrik, C. Aparicio, P. Kumar, M. Petr, O. Tomanec, M. B. Gawande, R. S. Varma and R.
Zboril, Green Chem., 2018, DOI: 10.1039/C8GC01333K.
This is an Accepted Manuscript, which has been through the
Volume 18 Number
7 7 April 2016 Pages 1821–2242
Royal Society of Chemistry peer review process and has been
accepted for publication.
Green
Chemistry
Accepted Manuscripts are published online shortly after
Cutting-edge research for a greener sustainable future
www.rsc.org/greenchem
acceptance, before technical editing, formatting and proof reading.
Using this free service, authors can make their results available
to the community, in citable form, before we publish the edited
article. We will replace this Accepted Manuscript with the edited
and formatted Advance Article as soon as it is available.
You can find more information about Accepted Manuscripts in the
author guidelines.
Please note that technical editing may introduce minor changes
to the text and/or graphics, which may alter content. The journal’s
standard Terms & Conditions and the ethical guidelines, outlined
in our author and reviewer resource centre, still apply. In no
event shall the Royal Society of Chemistry be held responsible
for any errors or omissions in this Accepted Manuscript or any
consequences arising from the use of any information it contains.
ISSN 1463-9262
CRITICAL REVIEW
G. Chatel et al.
Heterogeneous catalytic oxidation for lignin valorization into valuable
chemicals: what results? What limitations? What trends?
rsc.li/green-chem

Page 1 of 38
Green Chemistry
View Article Online
DOI: 10.1039/C8GC01333K
Cobalt entrenched Nꢀ, Oꢀ, and Sꢀtridoped carbons
as efficient multifunctional sustainable catalysts
for baseꢀfree selective oxidative esterification of
alcohols
Devaki Nandan, Giorgio Zoppellaro, Ivo Medřík, Claudia Aparicio, Pawan Kumar, Martin
Petr, Ondřej Tomanec, Manoj B. Gawande,* Rajender S. Varma and Radek Zbořil*
Regional Centre of Advanced Technologies and Materials, Department of Physical
Chemistry, Faculty of Science, Palacký University Olomouc, Šlechtitelů 27, 783 71
Olomouc, Czech Republic
*Corresponding authors: manoj.gawande@upol.cz (M.B.G.), radek.zboril@upol.cz (R.Z.)
ABSTRACT
We report the synthesis of sustainable and reusable nonꢀnoble transitionꢀmetal (cobalt)
nanocatalysts containing Nꢀ, Oꢀ, and Sꢀtridoped carbon nanotubes (Co@NOSC) composites.
The inexpensive and benign carrageenan served as carbon, oxygen, and sulfur source, while
urea supplied the nitrogen component. The material was prepared via direct mixing of
precursors, freezeꢀdrying followed by its carbonization under nitrogen at 900 °C. The
Co@NOSC catalysts comprising Co inner core and outer electronꢀrich heteroatomꢀdoped
carbon shell were thoroughly characterized with various techniques namely TEM, HRTEM,
1

Green Chemistry
Page 2 of 38
View Article Online
DOI: 10.1039/C8GC01333K
STEM elemental mapping, XPS, BET, ICPꢀMS, etc. The utility of the Co@NOSC catalyst
was explored for baseꢀfree selective oxidative esterification of alcohols to the corresponding
esters under mild reaction conditions; excellent conversions (up to 97%) and selectivities (up
to 99%) were discerned. Further substrates scope explored for the crossꢀesterification of
benzyl alcohol with long chain alcohols (up to 98%) and lactonization of diols (up to 68%).
The heterogeneous nature and stability of the catalyst facilitated by its ease of separation for
longꢀterm performance and recycling studies showed that catalyst was robust and remained
active even after six recycling experiments. EPR measurement was performed, to deduce the
reaction mechanism in the presence of POBN (αꢀ(4ꢀpyridylꢀ1ꢀoxide)ꢀNꢀtertꢀbutylnitrone) as
a spin trap agent, which confirmed the formation of •CH2OH radicals and H• radicals,
wherein solvent plays an active role in a nonconventional manner; plausible mechanism is
proposed for oxidative esterification of alcohols on the basis of EPR findings. The presence
of elemental cobalt core along with cobalt oxide and electronꢀrich Nꢀ, Oꢀ, Sꢀdoped carbon
shell displays synergistic effect to afford good to excellent yields of products.
Introduction
Nanomaterials have garnered incredible attention due to their unique and versatile
properties and numerous applications in various fields including energy conversion, chemical
engineering, environmental and biological applications, and sustainable catalysis.1–6 The
advancement in assembly of nanostructured materials has excelled due to modern
nanotechnological developments, predominantly not only because of nanoscale features but
also tunable properties such as compositions, shapes, size, surface area, porosity, and
accessible active sites.7–9 In recent years, a variety of nobleꢀmetal nanoparticlesꢀbased
catalysts have provided active sites for efficient catalysis and have been widely
investigated.10,11 The progress in sustainable development of heteroatomꢀdoped carbonꢀbased
2

Page 3 of 38
Green Chemistry
View Article Online
DOI: 10.1039/C8GC01333K
supports has been outstanding due to their semiconductive nature, in which electronic density
can be tuned by the amount of heteroatom doping.12–15 Due to the presence of lone pairs,
heteroatoms contribute additional electrons to the πꢀconjugated system, which in turn
increases charge density on carbon sheets. Moreover, such doping with heteroatoms is
reported to transform carbon sheets to semiconductors by shifting Fermi level from Dirac
point.16 The presence of varieties of heteroatoms provides more accessible active sites on the
surface of carbon materials and controls the overall hydrophobicity of the carbon sheets.17,18
The presence of electronꢀrich heteroatoms and associated functional groups such as
carboxylic, and hydroxyl groups on the surface of doped carbon species facilitate strong
interaction between nanosheets and metal nanoparticles to realize metal functionalized
catalysts.19 Furthermore, heteroatom doping has been shown to ensure prominent advantages
in terms of stabilizing the metal nanoparticles.20,21
Oxidation of alcohols to aldehydes, acids, esters, and acetals is common and widely
used reaction in organic transformations with catalysis.22–28 Direct synthesis of esters via
selective oxidation of alcohols is a challenging and useful reaction because of widespread
application of esters in the fields of fine chemicals, cosmetics and polymer industries.20,26,29,30
For this purpose various nobleꢀmetalꢀbased homogeneous catalysts have been
identified;28,31,32 however, the use of expensive noble metals, homogeneous nature of
reaction, and tedious separation process renders them as impractical choices. Sustainability
tenets dictate the use of inexpensive, nonꢀnoble metalꢀbased, and reusable heterogeneous
catalysts for the selective conversion of alcohols to esters.20,33 Transition metal nanoparticles
supported on active supports are becoming favorable choices to replace nobleꢀmetalꢀbased
catalysts; syntheses of esters from alcohols by using supported metal nanoparticles have been
attempted.20,33 The multistep preparation, poor stability, and leaching from the surface of
support materials are some of the aspects that need to be addressed besides the use of bases,
3

which is not environmentally friendly.26,33 In recent years, electronꢀrich surfaces of nitrogenꢀ
doped carbons have emerged as active supporting materials as well as promoters that catalyze
specific reactions;34–40 cobaltꢀbased catalysts have made inroads as inexpensive, nobleꢀmetalꢀ
free, and recyclable sustainable catalysts for aerobic oxidations of alcohols and selective
oxidation of alcohols to esters.20,33,35,41 Most of these procedures require the use of
environmentally compromising additives and bases for the aerobic oxidation of
alcohols.17,19,21,33 Metallic cobalt nanoparticles activated by nitrogen doping have displayed
improved activity under baseꢀfree conditions,20,41 but the multistep and wearying procedure
with expensive reagents limit their wider applications. Consequently, there is a persuasive
need to design a concise and allꢀencompassing multifunctional benign catalyst comprising
basic as well as metallic sites in a single step. Herein, we demonstrate the oneꢀstep synthesis
of efficient and multifunctional sustainable catalytic system capitalizing on nitrogenꢀ,
oxygenꢀ, and sulfurꢀrich carbonꢀnanotubesꢀcoated cobalt nanocomposites (Co@NOSC). The
assembly requires inexpensive and renewable carrageenan that provides carbon, oxygen, and
sulfur entities and urea as a nitrogen source. Co@NOSC catalyst has been prepared via the
direct pyrolysis of cobalt salt, urea, and bioꢀsource carrageenan mixture (Scheme 1) and
effectively deployed for the baseꢀfree selective oxidation of alcohols to esters under mild
conditions.
4

Page 5 of 38
Green Chemistry
View Article Online
DOI: 10.1039/C8GC01333K
The simple method entails biodegradable carrageenan biopolymer, which serves as a
precursor for carbon, oxygen as well as sulfur and possesses gelling properties, which
facilitates the entrapment of nitrogenꢀrich urea in the ensuing gel; distribution of urea in gel
network empowers flexible amount of nitrogen doping. The presence of diverse types of
basic nitrogen (pyridinic) along with the oxygen and sulfur on carbon shell enables the
promotion of baseꢀfree reactions. Further, the presence of hydrophilic oxygen, sulfur, and
nitrogen on hydrophobic carbon smooths the interaction between the catalysts and the
reactants.
Experimental
Synthesis of Co@NOSC. The homogeneous cobalt containing precursor solution was
prepared by dissolving carrageenan (3 g) in 80 mL water at 80 °C. Urea (16 g) was dissolved
separately in 40 mL water and slowly added to carrageenan solution under vigorous stirring,
followed by the addition of 2 g cobalt nitrate hexahydrate (Co(NO3)2·6H2O) in 20 mL water.
This whole mixture was stirred at 80 °C until the volume was reduced to 100 mL. This
ensuing mixture was immediately cooled in a deep freezer to obtain a solid gel, which was
then freezeꢀdried to obtain a pink powder. The material was transferred into a crucible and
heated up to 500 °C at a rate of 1 °C/min under N2 atmosphere and maintained this
temperature for 2 h. The temperature was then increased up to 900 °C at a rate of 2 °C/min
under N2 atmosphere and maintained for 2 h. The resulting black solid sample was washed
with boiling water at 100 °C (1500 mg catalyst in 45 mL water in closed 2ꢀdram vial) for 3 h
followed by methanol washing at 65 °C for 3 h to remove the weakly bound moieties. Owing
to the ease of oxidation of surfaceꢀbound metallic cobalt,36 the samples were heated at 400 °C
under H2 for 1 h and the material was vacuum preserved under N2 prior to their use. The
5

Green Chemistry
Page 6 of 38
View Article Online
DOI: 10.1039/C8GC01333K
synthesized magnetic material was identified as cobalt at nitrogenꢀ, oxygenꢀ, and sulfurꢀ
doped carbon nanotube composite (Co@NOSC); cobalt content of Co@NOSC, as confirmed
by ICPꢀMS analysis, was found to be 20.89 wt%. The corresponding reference sample NOSC
was prepared by following the aforementioned procedure without addition of cobalt salt.
Synthesis of urea–cobalt (Co@NC). Urea cobalt composite was synthesized by a similar
procedure except using carrageenanꢀincorporated gel (only urea and cobalt salt were used as
precursors). In a typical synthesis, 19 g urea was dissolved in 40 mL water followed by the
addition of 2 g Co(NO3)2·6H2O in 20 mL water. This mixture was stirred at 80 °C until the
volume was reduced to 20 mL. Although urea does not undergo gelling but for comparison
purposes 20 mL volume was cooled down in deep freeze to obtain a solidified sample, which
was then freezeꢀdried. The ensuing pink powder was transferred to a crucible, which was
heated to 500 °C at a rate of 1 °C/min under N2 atmosphere and maintained the temperature
for 2 h. Subsequently, the temperature was increased up to 900 °C at a rate of 2 °C/min under
N2 atmosphere and was maintained for 2 h. The synthesized magnetic material was identified
as cobalt at nitrogenꢀdoped carbon nanotube composite (Co@NC); cobalt content of
Co@NC, as confirmed by ICPꢀMS analysis, was found to be 60.59 wt%.
Synthesis of carrageenan–cobalt (Co@OSC). For the synthesis of carrageenan cobalt
composite similar approach was used by maintaining stoichiometric ratio of precursors
except urea was not added in the synthesis step. Briefly, 1.5 g carrageenan was dissolved in
40 mL water followed by the addition of 1 g Co(NO3)2·6H2O in 5 mL water. This mixture
was stirred at 80 °C until the volume was reduced to 35 mL. This mixture was immediately
cooled to initiate the gel formation. The solid gel was then freezeꢀdried and the pink powder
was transferred to a crucible, which was heated to 500 °C at a rate of 1 °C/min under N2
atmosphere and maintained at this temperature for 2 h. This was followed by increasing the
temperature up to 900 °C at a rate of 2 °C/min under N2 atmosphere. This temperature was
6

Page 7 of 38
Green Chemistry
View Article Online
DOI: 10.1039/C8GC01333K
maintained for 2 h. The synthesized magnetic material was identified as cobalt at oxygenꢀ
and sulfurꢀdoped carbon composite (Co@OSC); cobalt content of Co@OSC, as confirmed by
ICPꢀMS analysis, was found to be 37.39 wt%.
Catalytic reactions. The oxidative esterification of various types of alcohols was carried
out in 10 mL 2ꢀdram vial sealed with polytetrafluoroethylene/silicon cap (PTFE). The
reaction vessel was charged with 0.5 mmol of alcohol and 60 mg of the catalyst. The 2ꢀdram
vial was degassed and purged several times with pure oxygen and the balloon filled with pure
O2 was inserted with needle after the addition of alcohol and methanol. The 2ꢀdram vial was
transferred into a preheated oil bath at 60 °C and then heated for desired temperature with
continued stirring. After completion of the reaction, the catalyst was separated by using an
external magnet followed by washing and centrifugation. The conversion and yield of
products were determined by FIDꢀbased GC. The recyclability of the Co@NOSC was further
investigated under identical reaction conditions (0.5 mmol of benzyl alcohol, 60 mg of the
catalyst, 5 mL of methanol, 24 h, 60 °C). For the recycling studies, after the separation of
reaction catalyst via an external magnet/centrifuge, it was washed several times with water
and ethanol. The reusability was assessed by reducing the catalyst by H2 flow at 400 °C for 1
h and then using it in the next run (at least up to 6th cycle).
Results and discussion
Synthesis and structural analysis of Co@NOSC, Co@NC, and Co@OSC catalysts.
The possible formation of nitrogenꢀ, oxygenꢀ, and sulfurꢀrich (up to 1.7% N, 21.9% O, and
1.1% S based on XPS) carbon formed cageꢀlike organization, wherein interacted cobalt ion
with urea molecules are trapped in carrageenan gel (Scheme 1). The freeze drying of the
mixture ensures that cobaltꢀinteracted urea gets entombed in carrageenan gel. During the
carbonization process the urea molecules start condensing with the evolution of ammonia,
which is restricted due to the presence of more hydrophilic groups in carrageenan; such
7

Green Chemistry
Page 8 of 38
View Article Online
DOI: 10.1039/C8GC01333K
phenomenon has been observed by Shi et al.,42 wherein higher yields of carbon nitride were
attainable in presence of silica because of its hydrophilic nature. The high temperature
carbonization at 900 °C facilitates the formation of cobalt nanotube/nanoparticles core having
nitrogenꢀ, oxygenꢀ, and sulfurꢀrich carbon shell. For comparison purposes, we have prepared
the control materials by using the combination of carrageenan cobalt (Co@OSC), urea cobalt
(Co@NC), and urea carrageenan (NOSC); these materials have been synthesized by a similar
procedure as that of Co@NOSC.
The synthesized materials are well characterized using various analytical techniques. The
phase composition and crystalline structure of the sample were analyzed with powder Xꢀray
diffraction pattern (PXRD) (Fig. 1A). The PXRD pattern of a freshly synthesized Co@NOSC
catalyst shows sharp diffraction lines (2θ) at 51.81° and 60.59°, corresponding to interlayer
spacing (d) of 0.2048 and0.1770, respectively, thus indicating the presence of metallic cobalt
with cubic structure (space group Fm3m).37,43 The additional diffraction lines at 17.97°,
34.77°, 61.26° with dꢀspacing of 0.5743, 0.2993, and 0.1756 nm, respectively correspond to
crystal planes (111), (311), and (440) and they can be ascribed to the presence of cubic cobalt
sulfide (Co9S8, space group Fm3m).44 The presence of Co9S8 is due to transformation of some
cobalt into cobalt sulfide by reaction with sulfur present in carrageenan, during heating at
elevated temperature. Careful refinement of the XRD patterns allowed calculation of the
crystalline weight percentage of individual cobalt domain as metallic cobalt (~76.1%), Co9S8
(23.2) and CoO (0.7%) (Fig. S1†).
8

Page 9 of 38
Green Chemistry
View Article Online
DOI: 10.1039/C8GC01333K
Fig. 1 (A) Powder Xꢀray diffraction (PXRD) patterns of (a) fresh, (b) H2ꢀtreated, and (c)
recycled Co@NOSC catalyst samples. (B) Raman spectrum of Co@NOSC.
The presence of graphitic Nꢀ, Oꢀ, Sꢀtridoped carbon (with hexagonal space group
P63/mmc) was confirmed by the diffraction line at 2θ = 30.6° with dꢀspacing 0.3384 nm. The
average crystallite size, calculated with Scherrer's equation, was found to be 42.46 nm for
elemental cobalt and 40 nm for Co9S8. The recycled Co@NOSC catalyst also showed similar
pattern and revealed that there were no changes in phase structure during the reaction and the
washing step (Fig. 1A(b) and (c)). Further, the presence of specific D and G band in Raman
spectra of Co@NOSC catalyst proved the graphitic nature of carbon material (Fig. 1B). The
G band was originated due to inꢀplane vibration of sp2 carbons, while D band stems from outꢀ
plane vibration of sp3 carbons thus representing defect in graphenic structure. The intensity
ratio of D and G band (ID/IG) was found to be 1.21, which represents plenty of defects due to
doping of heteroatoms in conjugated network.45,46
The morphology of the Co@NOSC was thoroughly investigated using TEM and HRTEM.
The TEM image of Co@NOSC material shows irregular nanotube and nanospheres of
different shapes with clearly observable Co core surrounded by graphenic Nꢀ, Oꢀ, and Sꢀ
tridoped carbon shell (Fig. 2a,b, Fig. S2†). The average diameter of the core was found to be
9

Green Chemistry
Page 10 of 38
View Article Online
DOI: 10.1039/C8GC01333K
in the range of ~10 to 15 nm, while thickness of the shell was ~12 to 15 nm (Fig. 2c). The
fine structure of material analyzed by HRTEM shows that the material contains highly
integrated nanostructure of the layered Nꢀ, Oꢀ, Sꢀtridoped carbonꢀcoated cobalt nanotubes.
The HRTEM image at 20 nm scale shows various crystallite fringes, such as at 2 nm it clearly
shows interlayer spacing of 0.29 nm, 0.20 nm, and 0.33 nm ascribed to (311) plane of Co9S8,
(111) plane metallic cobalt, and (002) plane of NOSꢀdoped carbon, respectively (Fig. 2d,e,f).
The presence of cobalt in inner core and heteroatomꢀdoped carbon outer shell is further
confirmed by STEM elemental mapping of Co@NOSC catalyst (Fig. 3 and Fig. S3†).
Interestingly, it can be observed that the heteroatoms (N, O, and S) were uniformly
distributed in the carbon shell confirming the efficiency and even doping throughout the
material. The XRD pattern of controlled samples Co@OSC and Co@NC can be compared
with Co@NOSC (Fig. S4b†). XRD pattern of Co@OSC (Fig. S4 and Fig. S5†) showing
similar pattern as that of Co@NOSC having higher amount of Co9S8, along with metallic
cobalt and CoS but in case of Co@NC (Fig. S4 and Fig. S6†), in addition to metallic cobalt
with cubic structure (space group Fm3m) some other phases like cobalt oxide (space group
Fm3m), cobalt nitride (space group Fꢀ43m) and cobalt with hexagonal structure (P63/mmc)
also observed. The TEM observation of Co@OSC (Fig. S7a–f†) shows that cobalt inner core
is surrounded by carbon shell, but in Co@NC (Fig. S7gꢀl†) less coating was observed due to
more loss of ammonia during condensation of urea; this supports that the special interaction
(Scheme 1) operated in case of Co@NOSC, which is further supported by its catalytic
activity and stability (ICPꢀMS analysis).
10

Page 11 of 38
Green Chemistry
View Article Online
DOI: 10.1039/C8GC01333K
Fig. 2 TEM images of Co@NOSC catalyst sample at (a) 500 nm, (b) 100 nm, (c) 20 nm, (e)
20 nm scale, and (d) and (f) selected area at 2 nm scale bar showing dꢀspacing.
11

Green Chemistry
Page 12 of 38
View Article Online
DOI: 10.1039/C8GC01333K
Fig. 3 (a) HRTEM image, (b) HAADF image, (c–l) STEM elemental mapping of Co@NOSC
showing C, N, O, S, and Co.
The Xꢀray photoelectron spectroscopy (XPS) revealed the presence of oxygen, nitrogen,
and sulfur along with the carbon in the material (Fig. S8†). The elemental composition of
nitrogen, oxygen, sulfur, carbon, and cobalt in the material was found to be 1.7, 21.9, 1.1,
12

Page 13 of 38
Green Chemistry
View Article Online
DOI: 10.1039/C8GC01333K
71.6, and 3.7 at%, respectively, which shows significant doping of different heteroatoms in
the carbon sheets.
Fig. 4 XPS spectra of Co@NOSC catalyst (a) N1s, (b) O1s, (c) S2p, and (d) Co2p region.
The high resolution XPS spectra of the catalyst in C1s region can be deconvoluted in four
peak components with binding energy 284.6, 285.5, 286.9, and 288.5 eV corresponding to
C=C–C, C–O/C–N, C=O, and O=C–O type of carbons, thus showing the graphitic nature of
carbon material34 (Fig. S9a†). The appearance of specific N1s peak component at 398.5,
400.7, and 401.6 eV confirms the presence of pyridinic, pyrrolic, and graphitic nitrogen in
NOSCꢀdoped materials (Fig. 4a). The three peak components at 529.7, 531.6, and 533.11 eV
13

Green Chemistry
Page 14 of 38
View Article Online
DOI: 10.1039/C8GC01333K
in O1s spectra of material, originated due to presence of oxygen on the surface of cobalt,
C=O and C–O, reveal that graphitic carbon is doped with oxygen (Fig. 4b). The high
resolution S2p peaks are also deconvoluted, mainly into two peak components associated
with C−S−C (164.6/163.0) in case of a fresh sample, while for a reused sample slight shifting
of this peak component can be seen (163.6/162.4 eV) and C−SOx−C (169.8/168.7eV) species
in both the samples (Fig. 4c and ESI Fig. S7).34 The HRꢀXPS of catalyst in Co2p region
shows the presence of cobalt oxide (3.7 wt%) (Fig. 4d) on the surface of materials. Overall,
all the heteroatom doping along with cobalt is confirmed by XPS; elemental mapping shows
the presence of multifunctionalities of the Co@NOSC catalyst. Further, all these
functionalities are intact in the reused catalyst at least up to 6th cycle after reaction with
elemental compositions of nitrogen, oxygen, sulfur, carbon, and cobalt in the material being
3.1, 25.9, 1.0, 62.7, and 7.4 at%, respectively, according to XPS analysis (Fig. S8, S9b and
Fig. S10†). The BET surface area analysis of Co@NOSC catalyst displays high surface area
of 383.09 m2/g when compared to Co@NC (45.90 m2/g) and Co@OSC (369.90 m2/g). The
type IV nitrogen adsorption–desorption isotherm with a H4ꢀshaped hysteresis loop observed
for Co@NOSC and Co@OSC confirmed the microꢀmesoporous nature of these materials,
while H2 type hysteresis loop observed for Co@NC confirms the mesoporous nature of the
material (Fig. S11A†).47 The higher uptake at low pressure is associated with the filling of
micropore in case of Co@NOSC and Co@OSC as compare to Co@NC.47 This is further
supported by micropore surface area calculated by Tꢀplot method (Table S1), which confirms
the purely mesoporous nature of Co@NC, while Co@NOSC (micropore surface area 347.82
m2/g) and Co@OSC (micropore surface area 274.80 m2/g) are hierarchical porous materials.
Furthermore, it is also supported by respective pore size distribution displays (Fig S11B†);
the contribution of mesopore is higher for Co@NC as compared to Co@NOSC and
Co@OSC. Overall, from the above discussion it is apparent that materials, Co@NOSC and
14

Page 15 of 38
Green Chemistry
View Article Online
DOI: 10.1039/C8GC01333K
Co@OSC are hierarchical porous in nature (micro as well as mesoporous, while Co@NC is
mesoporous in nature). The presence of mesoporous nature of the Co@NOSC, Co@OSC,
Co@NC materials was further evidenced by low angle XRD pattern (Fig. S4A†). One broad
peak observed for all three materials signifying the average poreꢀcenter to poreꢀcenter
correlation length confirms the mesoporous nature of materials having hierarchical pores.48,49
Oxidative esterification of alcohols over the Co@NOSC catalyst. The synthesized and
wellꢀcharacterized Co@NOSC catalyst was tested for the selective oneꢀstep oxidation of
alcohols to esters without employing any base. Benzyl alcohol was chosen as a model
substrate for the reaction with methanol, which serves the dual role of solvent as well as
reagent, using O2 gas as terminal oxidant under mild and baseꢀfree conditions. As can be seen
from Table 1, the oxidative esterification did not proceed in the absence of catalyst or O2, or
in the presence of Co@NOSC sample heated at 550 °C (Table 1, entries 1 to 4). To compare
the catalytic activity, we have tested all three catalytic components namely Co@NC,
Co@OSC, and Co@NOSC having cobalt content 60.59 wt%, 37.39 wt%, and 20.89 wt%
respectively. Less amount of benzyl alcohol was oxidized to methyl benzoate (conversion
15%, selectivity 16%) over Co@OSC and Co@NC (conversion 50%, selectivity 74%), even
with a higher amount of cobalt content (Table 1, entries 5 and 6). These results indicate that
Co nanoparticlesꢀbased catalyst without doped carbon could not promote the oxidative
esterification of alcohol in the absence of base at low temperature. However, Co@NOSC
catalyst (having nitrogen, oxygen, sulfur, and carbon along with cobalt) gave excellent
conversion (97%) of benzyl alcohol under similar conditions with more than 98% selectivity
for methyl benzoate (Table 1, entry 17). These findings clearly show that synthesized
Co@NOSC catalyst considerably promotes such reaction to yield methyl benzoate in the
absence of any base. The catalytic activity of the present catalyst can be attributed to its
specific structure having inner metallic core and outer electronꢀrich heteroatomꢀdoped carbon
15

Green Chemistry
Page 16 of 38
View Article Online
DOI: 10.1039/C8GC01333K
(nitrogen, oxygen, and sulfur). To optimize the ideal reaction conditions, initially various
reaction parameters such as amount of catalyst and reaction time, amount of methanol that is
playing active role (acts as a coꢀcatalyst; confirmed by EPR, discussed in mechanism), etc.,
were appraised to achieve conversion as high as 97% (Table 1, entries 7 to 17). It was found
that for 25 mg of catalyst, up to 95% conversion can be achieved in 48 h, whereas 60 mg
catalyst afforded 97% conversion only in 24 h (Table 1, entries 8 and 17).
Table 1 Optimization of oxidative esterification of benzyl alcohol.a
Cobalt mmol/
Benzyl alcohol
mmol (molar
ratio)
Conversion Selectivity
Entry
Catalyst (mg)
Time (h)
(%)b
(%)b
0/0.50
0.21/0.50
00/0.50
ꢀ
1
2
3
NOSC (60)c
Co@NOSC (60)d
No catalyst
24
24
24
2
ꢀ
10
ꢀ
ꢀ
ꢀ
Co@NOSCꢀ550 °C
calcined (80)
4
24
NR
ꢀ
0.38/0.50
5
6
7
Co@OSC (60)
Co@NC (60)
24
24
24
15
50
72
16
74
93
0.61/0.50
0.08/0.50
Co@NOSC (25)
0.08/0.50
0.16/0.50
0.16/0.50
0.21/0.50
8
Co@NOSC (25)
Co@NOSC (50)
48
24
95
93
97
97
10
11
12
Co@NOSC (50)
Co@NOSC (60)
48
24
99
97
99
>98
16

Page 17 of 38
Green Chemistry
View Article Online
DOI: 10.1039/C8GC01333K
Co@NOSC (60) 2
mL MeOH
0.21/0.50
13
14
24
24
70
90
87
98
Co@NOSC (60) 4
mL MeOH
0.21/0.50
15
16
17
Co@NOSC (60)
Co@NOSC (60)
Co@NOSC (60)
0.21/0.50
0.21/0.50
0.21/0.50
7
60
91
97
87
93
17
24
>98
b
aReaction conditions: benzyl alcohol (0.5 mmol), methanol (5 mL), O2 balloon. Determined
c
by GC. NR = No reaction. Catalyst: prepared by carrageenan and urea calcined at 900 °C
and dreaction condition (a) without O2 under N2.
For the first 17 h, the reaction proceeds very fast with less selectivity (87–93%) but after this
period the reaction became slow and selectivity increased to more than 98% (Table 1, entries
15 to 17). Further, increase in the amount of catalyst weight up to 60 mg, 97% conversion,
and >98% selectivity of methyl benzoate was obtained after 24 h. The initial fast reaction
rate may be because of more freely available (pyridinic nitrogen and metallic cobalt along
with oxygen and sulfur) active sites. It is also discussed in mechanism part, both metallic
cobalt and heteroatoms, specially nitrogen, are synergistically playing active role for higher
conversion. We observed that both hydrogen free radical and H2O2 ensued during reaction;
because of this pyridinic nitrogen get protonated and possibly the oxidation of surface coated
cobalt. This could be the reason for slow reaction rate after 7h. The basic nature of nitrogen
also plays an active role in increasing the selectivity of methyl benzoate. From the entry 5, we
observe that the Co@OSC catalyst is having a lower amount of metallic cobalt (Fig. S5) and
the absence of nitrogen yields much less selectivity (~ 16%) for methyl benzoate. Moreover
Co@NC having higher amount of metallic cobalt and (Fig. S6) having less nitrogen doping
gave 50% conversion with 74% selectivity towards methyl benzoate.
From these optimization parameters, it became apparent that 60 mg (0.21 mmol cobalt) of
17

Green Chemistry
Page 18 of 38
View Article Online
DOI: 10.1039/C8GC01333K
catalyst with 5 mL methanol and O2 balloon is sufficient to achieve conversion of benzyl
alcohol as high as 97% with more than 98% selectivity in 24 h at 60 °C. The strong electronic
interactions and synergistic effect between Nꢀ, Oꢀ, Sꢀtridoped carbon and cobalt certainly
helps to improve the catalytic activity compared to their individual components such as
Co@OSC, Co@NC, NOSC, and Co@NOSC calcined at 550 °C.
With the optimized conditions in hand, the versatility of the catalyst was ascertained by using
a variety of substituted benzyl alcohols, biꢀbenzyl alcohols, Nꢀheterocyclic alcohols, Sꢀ
heterocyclic alcohols, and aliphatic alcohol substrates were investigated for the aerobic
oxidative esterification (Table 2). Table 2 shows that for 4ꢀhalo (–Cl, –Br) and 3ꢀhalo (–F)
substituted benzyl alcohols, more than 97% conversion and up to 98% selectivity could be
achieved (Table 2, entries 2 to 4); other substituted (–CH3, –OMe, –CF3, and –NO2, etc.)
benzyl alcohol derivatives were also explored (Table 2, entries 5 to 11). Notably, in case of –
CH3 and OMe, excellent performance of the catalyst with the high conversion (97%) and
selectivity (>98%) towards corresponding methyl ester was observed. Importantly, in case of
–NO2 substituted benzyl alcohol, excellent activity (>90% conversion and selectivity) was
noted, compared to previously reported nobleꢀmetalꢀbased homogeneous catalyst;22 this may
be due to the presence of more hydrophilic heteroatoms on the surface of the catalyst.
However, –CF3 substituted benzyl alcohol could afford only 58% conversion, which may be
ascribed to the presence of more electronegative fluorine atoms, which may restrict the
interaction of alcoholic group with the catalyst surface (Table 2, entry 10). We have chosen
paraꢀ, metaꢀ, and orthoꢀ isomers of methoxy benzyl alcohols to see the effect of steric
hinderance on reaction; the reactivity order was found to be para > meta > ortho (Table 2,
entries 6 to 8). The moderate conversion (78%) and selectivity (64%) for ortho –OMe
substituted methyl benzoate demonstrates the blocking effect of these functional groups on
the esterification reaction of the corresponding alcohols (entry 8).20 In contrast to poor yields
18

Page 19 of 38
Green Chemistry
View Article Online
DOI: 10.1039/C8GC01333K
(20%) and selectivity (~21%) reported in previous reports even with Pdꢀbased catalyst,22 4ꢀ
methoxy benzyl alcohol gave an excellent conversion (97%) and selectivity (>98%) in our
case. This is further supported by ortho –CH3 substituted benzyl alcohol, where 73%
conversion with 61% selectivity (Table 2, entry 9) was discerned. Heterocyclic alcohols
namely furfuryl alcohol and 2–pyridine methanol could be oxidized to corresponding esters
with good conversion (78–89%) and selectivity 95–98% (Table 2, entries 12 and 13); for
attaining maximum conversions in some of these reactants, the reaction time was increased
up to 48 h (Table 2, entries 12 and 13). Methyl cinnamate and dimethyl terephthalate are two
important chemicals widely used in polymer industries,50,51 and they could be prepared from
cinnamyl alcohol (80% conversion with 86% selectivity) and 1,4ꢀbenzenedimethanol (90%
conversion with 98% selectivity), respectively (Table 2, entries 14 and 15). Furthermore, the
ligninꢀderived bioꢀmolecule namely vanillyl alcohol could be used as a substrate to acquire
significant products such as vanillin (48% selectivity) and methyl vanillate (5% selectivity)
(Table 2, entry 16), which find application in foods, beverages, cosmetics, and drugs.52
Overall, the better use of novel and inexpensive carbon, oxygen, sulfur (carrageenan), and
nitrogen (urea) precursors for the efficient catalyst preparation is supported by its
superior/comparative performance, when compared to previous documentations (Table S2†).
19

Green Chemistry
Page 20 of 38
View Article Online
DOI: 10.1039/C8GC01333K
Table 2. Substrate scope of oxidative esterification reactions with methanol.a
Conversion
Entry
1
Substrate
Product
Selectivity (%)bc
(%)b
97
>98
99
2
3
4
5
6
97
97
97
96
96
98
99
93
98
7
8
84/95d
78/90d
98/98d
64/95d
20

Page 21 of 38
Green Chemistry
View Article Online
DOI: 10.1039/C8GC01333K
9
62/73d
58/60d
58/61d
10
>98/98d
>98
11
90
12
13
89/98d
78/98
>98/99d
95/98
14
15
66/80d
78/86d
90
98
16
36/44d
2/5d
aReaction conditions: benzyl alcohol (0.5 mmol), methanol (5 mL), O2 balloon, catalyst 60
b
c
mg (0.21 mmol cobalt). Determined by GC. Side product is only corresponding aldehyde.
dReaction time 48 h.
The aerobic esterifications of benzyl alcohol with other aliphatic alcohols, including
ethanol, propanol, butanol, and pentanol were also performed (Table 3). It was observed that
21

Green Chemistry
Page 22 of 38
View Article Online
DOI: 10.1039/C8GC01333K
linear alcohols gave good conversion relative to branched chain alcohols, which was
attributed to the +I effect of the methyl group thus decreasing the –C–H cleavage.
Furthermore, we explored this Co@NOSC catalytic system to the oxidative lactonization of
diols (Table 4); lactonization of diols is an important reaction because of the potential for
lactones in biological applications as well as in polymer industries.53 The versatility of our
catalytic system became apparent in the lactonization of diols with good conversion (66–
68%) and selectivity (91–92%).
Table 3 Co@NOSC catalyzed crossꢀesterification of benzyl alcohol with long chain
alcohols.a
Entry
1c
ROH
Product
Conversion (%)b
98
Selectivity (%)b
85
2c
40
50
3d
30
85
75
55
4e
22

Page 23 of 38
Green Chemistry
View Article Online
DOI: 10.1039/C8GC01333K
aReaction conditions: benzyl alcohol (0.05 mmol), ROH (2 mL), 35 mg (0.11 mmol cobalt)
b
c
catalyst, O2 balloon. Determined by GC. temperature 70 °C, dtemperature 100 °C,
etemperature 110 °C, reaction time 22 h.
Table 4 Co@NOSC catalyzed lactonization of diols.a
Entry
1
ROH
Product
Conversion (%)b
68
Selectivity (%)b
92
2
3
65
66
91
92
aReaction conditions: diols (0.1 mmol), toluene 3 mL, 80 mg (0.28 mmol cobalt) catalyst, O2
balloon, temperature 100 °C, reaction time 22 h. bDetermined by GC.
The reusability of a catalyst is a key factor for its sustainable applicability in heterogeneous
catalysis. We examined the stability and the reusability of the Co@NOSC catalyst by
recovering the used catalyst from reaction and their compositional integrity via
characterization. The catalyst recycling was accomplished by washing with water and ethanol
and heat treatment at 400 °C under hydrogen for 1 h. Heating of the catalyst at 400 °C under
H2 for 1 h after each step showed no significant loss in catalytic performance up to reuse
experiments (Fig. 5). The ICPꢀMS result of the filtrate shows no significant leaching (0.015
wt%) confirm the stable nature of the catalyst. The XRD spectra of Co@NOSC catalyst after
23

Green Chemistry
Page 24 of 38
View Article Online
DOI: 10.1039/C8GC01333K
six recycling did not show any significant visible change, which revealed that the catalyst
retained its crystalline nature even after sixth reuse (Fig. 1c). Owing to the oxidizing nature
of cobalt nanoparticles, cobalt oxide can be seen from XPS analysis of reused catalyst, where
we can see the slight shifting of Co 2p3/2 peak from binding energy value 778 eV to 779 eV
(Fig. S10d†). No detectable change in the binding energy of C1s and N1s was discerned
except slight shifting of pyridine nitrogen from 401.0 eV to 398.8 eV thus reaffirming no
discernible change in chemical nature of carbon shell (Fig. 9 B and Fig. S10a,c†). However,
slight shifting of pyridine nitrogen peak can be explained due to protonation of the pyridinic
nitrogen during the reaction. From the TEM images of reused catalyst after 6th cycle, we
observed that cobalt metal is still intact in inner core and surrounded by heteroatomꢀrich
carbon shell (Fig. S12†).
Fig. 5 Recyclability of the catalyst by treatment of the catalyst with hydrogen at 400 °C for 1
h.
24

Page 25 of 38
Green Chemistry
View Article Online
DOI: 10.1039/C8GC01333K
The oxidation reaction of benzyl alcohol in presence of the cobaltꢀcatalyst was furthermore
probed by EPR technique, in conjunction with spinꢀtrapping experiments, using αꢀ(4ꢀpyridylꢀ
1ꢀoxide)ꢀNꢀtertꢀbutylnitrone (POBN) as spin trapping agent. POBN was initially dissolved in
MeOH in presence of the substrate (benzyl alcohol) and the solution was warmed for 30 min
at 60 °C. The experiment revealed that clear degradation of the POBN radical precursor into a
nitroxide spinꢀactive species does not occur in these solvents, during the time frame and
temperature examined. In presence of the catalyst, and after stirring the reaction mixture for
30 min at 60 °C under air, the solution EPR spectrum (T = 218 K) revealed on the other hand
the clear signature of trapped radical species (Fig. 6a and Fig. S13†), evidence for the
formation of radical intermediates evolving within the oxidation reaction of benzyl alcohol.
Simulation of the observed EPR resonance (Fig. 6a, red line, Sim) indicates that the
resonance signal may arise from formation of the POBNꢀCH2OH radical adduct.54 Fig. 6c,
shows the optimized structure of the of POBNꢀCH2OH radical obtained by DFT/UBP86/6ꢀ
31G* (gasꢀphase) with the drawn spin density distribution. The observed solution EPR
resonance is well reproduced under the spinꢀHamiltonian framework by thirdꢀorder
perturbation theory (WinEPR SimFonia v. 1.25), upon using the following spinꢀHamiltonian
parameters for the nitroxide radical: hyperfine coupling constants (hfcc) aN of 15.00 G, aH of
3.30 G for the βꢀhydrogen, giso of 2.0052(2), Lorentzian/Gaussian ratio of 0.20 with lineꢀ
width function (a+bm+cm2) that included tumbling effects on the nitrogen (NꢀO● moiety in
Fig. 6c) with parameters a = 1.99, b = –0.24 and c = 0.20. These estimated hfcc parameters
are found in fair agreement with the isotropic Fermi contact coupling values obtained in gas
phase (DFT/UB3LYP/6ꢀ31G*(d,p)) for the POBNꢀCH2OH radical adduct, which are 13.06 G
for aN and 2.86 G for the
βꢀhydrogen aH (Table S4†). Similar EPR spectrum and spinꢀ
Hamiltonian parameters (aN of 15.25 G, aH of 3.75 G) were obtained by Halpern and coꢀ
25

Green Chemistry
Page 26 of 38
View Article Online
DOI: 10.1039/C8GC01333K
authors in the PBNꢀCH2OH radical adduct (see in addition Fig. S14†).55 Pou et al. obtained
also similar EPR spectrum by dissolving 4ꢀPOBN in oxygen saturated waterꢀbuffer solution,
in presence of Me2SO4 and H2O2, which gave the correspondent POBNꢀCH2OH radical
adduct with aN of 14.45 G and aH of 2.25 G.56 Therefore, the MeOH molecules play an active
role in controlling the fate of the radical species generated during catalysis. When H2O2 is
formed and converted to 2OH● radicals, OH● interact fast with the solvent molecules
●
(MeOH), giving rise to CH2OH and H2O. To further validate this hypothesis, spinꢀtrapping
experiments were performed with the sole presence of (i) the catalyst, (ii) the spin trapping
agent POBN, and (iii) the substrate (benzyl alcohol), following the same experimental
conditions as reported above, but without addition of MeOH. No radical species were
successfully observed/trapped by POBN under these conditions, as judged from the
featureless solution EPR spectrum recorded (Fig. S15†) We also tested the alternative
scenario in which only a small amount of MeOH was added to the mixture containing the
catalytic system and the aromatic substrate, hence by considering it as a limiting reactant.
Upon addition of few microliters of MeOH (20 ꢁL) to the mixture composed by catalyst,
POBN, and benzyl alcohol (2 mL of benzyl alcohol), and after 30 min of heating at 60 °C
under air, POBN trapped again radical species in the solution. However, the observed EPR
signal was found much more complex compared to the previous experimental setꢀup, namely
when the presence of a large excess of MeOH has been added to the reaction mixture. The
observed EPR signal is shown in Fig. 6b (T = 243 K, see also Fig. S14†) together with the
EPR spectrum simulation (Fig. 6b, red line, Sim). Two radical species contribute to the
●
observed signal, (1) POBNꢀCH2OH radical (entrapment of CH2OH) and (2) POBNꢀH
radical (entrapment of ●H). The simulated EPR spectrum by thirdꢀorder perturbation theory is
obtained by using the following spinꢀHamiltonian parameters: (2) (POBNꢀH) (aN) of 16.20
G, 2H (aH) of 8.10 G, Lorentzian/Gaussian ratio of 0.22 and lineꢀwidth function (a+bm+cm2)
26

Page 27 of 38
Green Chemistry
View Article Online
DOI: 10.1039/C8GC01333K
with parameters a = 1.99, b = –0.20 and c = 0.18) and giso of 2.0052(2). (1) (POBNꢀCH2OH)
(aN) of 14.35 G, 1H (aH) of 3.0 G, Lorentzian/Gaussian ratio of 0.23 and coefficients a =
1.30, b = –0.25, and c = 0.20, and giso of 2.0052(3). Similar spinꢀHamiltonian parameters for
the POBNꢀH radical were obtained by AurianꢀBlajeni et al. (upon irradiation of aqueous
suspensions of tungsten oxide in the presence of MeOH using POBN as a trapping agent
(POBNꢀH radical, aN of 16.6 G and aH of 10.25 G).57 From the spectrum simulation shown in
Fig. 6c, the relative weight of the two trapped spin components (●H, ●CH2OH), calculated as
percentage with respect to the total EPR doubleꢀintegrated signal, have been estimated as
45% and 55%, respectively; thus, these radicals appear to form together and close to 1:1 ratio
●
●
from the catalytic decomposition of CH3OH→ H + CH2OH. The contributing resonances
(POBNꢀCH2OH and POBNꢀH) in the spectrum simulation shown in Fig. 6b are explicitly
given in ESI† material as Fig. S16†. Fig. 6d shows the optimized structure obtained by
DFT/UBP86/6ꢀ31G* (with the spin density distribution, gasꢀphase) of the POBNꢀH radical
species together with the calculated isotropic Fermi contact coupling values
(DFT/UB3LYP/6ꢀ31G*(d,p)). It is noted here that the theoretical calculation (gasꢀphase)
poorly reproduces the experimentally determined hfcc, thus the interaction between POBNꢀH
with the substrate/solvent benzyl alcohol in the medium seems strong enough to produce a
large modulation in the hyperfine terms of the radical molecule.
27

Green Chemistry
Page 28 of 38
View Article Online
DOI: 10.1039/C8GC01333K
Fig. 6 Xꢀband (9.17 GHz) EPR spectrum of the trapped radical species. Panel (a) shows the
solution spectrum (T = 218 K) of the supernatant collected from the catalyst/benzyl
alcohol/POBN/MeOH mixture together with its EPR simulation (red line). Simulation
parameters are given in the main text. Panel (b) shows the solution spectrum (T = 243 K) of
the supernatant collected from the catalyst/benzyl alcohol/POBN/MeOH mixture in which
MeOH has been added as the limiting agent, together with its EPR simulation (red line). The
spinꢀHamiltonian parameters are given in the main text. Experimental parameters: 100 kHz
modulation frequency, 30 ms time constant, 2.0 G modulationꢀwidth, 5.0–6.0 mW of applied
microwave power, 1 min sweepꢀtime. Scans accumulated and averaged were in (a) 21 and in
(b) 11. Panel (c) shows the optimized structure (gasꢀphase) of POBNꢀCH2OH radical and (d)
the POBNꢀH radical (DFT/UBP86/6ꢀ31G*, energy –764.43954 a.u. for (c) and energy –
28

Page 29 of 38
Green Chemistry
View Article Online
DOI: 10.1039/C8GC01333K
649.91719 a.u. for (d)) where the spinꢀdensity isosurface was drawn at 0.002 IsoVal. Fermi
contact terms (hffc) obtained on the optimized structures (<S2> = 0.7500, c; <S2> = 0.7500,
d) were calculated by singleꢀpoint DFT theory (UB3LYP/6ꢀ31G*(d,p), gasꢀphase).
Therefore, on the basis of EPR results, from product analysis and taking into account
previous works from the literature, we propose a possible mechanism for the oxidative
esterification of benzyl alcohols by Co@NOSC catalyst (Fig. 7).10,20,58–60 The metallic cobalt
core of the Co@NOSC catalyst serves as a source of electrons, while the carbon shell
provides active sites that aid binding of the substrate on the surface and facilitates the
electron transfer reactions. At the onset of catalysis, it is presumed that molecular triplet
oxygen is formed quickly, after taking one electron from the metal catalyst, it is converted
●ꢀ
into superoxide anion radical (O2 ) (Step I). In the next step, this reactive superoxide anion
radical abstracts a hydrogen radical (●H) from surfaceꢀadsorbed molecules of methanol and
●
transforms into hydroperoxides radical (●HO2), while methanol forms a CH2OH radical
●
intermediate (Step II). These CH2OH radicals were effectively trapped by POBN, forming
●
the POBNꢀCH2OH radical adduct as detected in the EPR analysis. The HO2 radical further
●
●
abstracts one H from CH2OH and forms hydrogen peroxide together with an aldehyde
(CH2=O) (Step III). The H2O2, in the presence of catalyst, can be dissociated again into ●OH
●
(step IV). In the subsequent step OH radicals attack benzyl alcohol to form benzaldehyde
(step V). The formation of benzaldehyde as a stable molecule is supported by our GC
analysis as the sole side product. This can be seen from Table 1 (entries 15–17), where we
observed that initial lower selectivity for ester (87% in 7 h) increased with time to more than
98%. In the subsequent step, because of the presence of more heteroatoms on the catalyst’s
surface, the benzaldehyde reacts with methanol (present in excess) to a reactive hemiketal
29