Benign-by-design preparation of humin-based iron oxide catalytic nanocomposites

Article abstract of DOI:10.1039/c7gc01405h

The current acid-catalyzed conversion of biomass feedstocks yields substantial quantities of undesired by-products called humins, for which applications are yet to be found. This work aims to provide a starting point for valorisation of humins via preparation of humin-based iron oxide catalytic nanocomposites from humins and thermally treated humins (foams) via solvent-free methodologies including ball milling and thermal degradation. The prepared materials were found to be active in the microwave-assisted selective oxidation of isoeugenol (conversions >87%) to vanillin, proving the feasibility to use humin by-products as template/composite materials.

Full text of DOI:10.1039/c7gc01405h

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: L. Filiciotto, A. M.
M. Balu, A. A. Romero, E. Rodriguez-Castellón, J. K. van der Waal and R. Luque, Green Chem., 2017, DOI:
1
0.1039/C7GC01405H.
Volume 18 Number
7
7
April 2016 Pages 1821–2242
This is an Accepted Manuscript, which has been through the
Royal Society of Chemistry peer review process and has been
accepted for publication.
Green
Chemistry
Cutting-edge research for a greener sustainable future
www.rsc.org/greenchem
Accepted Manuscripts are published online shortly after
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 31
Green Chemistry
View Article Online
DOI: 10.1039/C7GC01405H
Benignꢀbyꢀdesign preparation of iron oxides/humins catalytic nanocomposites
a
a
a
b
Layla Filiciotto , Alina M. Balu , Antonio A. Romero , Enrique RodríguezꢀCastellón , Jan C. van
c
a
der Waal , Rafael Luque *
a
Departamento de Química Orgánica, Universidad de Córdoba, Edificio Marie Curie (Cꢀ3), Ctra
Nnal IVꢀA, Km 396,E14014 Córdoba, España, eꢀmail: q62alsor@uco.es
b
Departamento de Química Inorgánica, Cristalografía y Mineralogía, Facultad de Ciencias,
Universidad de Málaga, 29071 Málaga, España.
c
Avantium Chemicals B.V., Zekeringstraat 29, 1014 BV Amsterdam, The Netherlands
Abstract
The current acidꢀcatalyzed conversion of biomass feedstocks yields substantial quantities of
undesired byꢀproducts called humins, for which applications has yet to be found. This work aims to
provide a starting point for humins valorisation via preparation of huminꢀbased iron oxide catalytic
nanocomposites from humins and thermally treated humins (foams) via solventꢀfree methodologies
including ball milling and thermal degradation. The prepared materials were found to be active in
the microwaveꢀassisted selective oxidation of isoeugenol (conversions > 87%) to vanillin, proving
the feasibility to use humin byꢀproducts as template/composite materials.
Introduction
World’s continuous scientific improvement, alongside the depletion of fossil feedstocks,
calls upon the investigation of new sustainable processes using renewable sources. The synthesis of
new active catalytic materials will play an important role achieving this. Recent progress in
biorefinery concepts leads the way in the production of platform chemicals from biomass as widely
1
ꢀ6
available renewable resource . As a matter of fact, Avantium’s YXY technology has announced
plams to build a 50 kta plant to convert biomassꢀderived sugars into platform chemicals (e.g.
furanics, levulinics) for the synthesis of bioꢀplastics and fuels. In particular, the polymerization of
YXY furanic building block (FuranDiCarboxylic Acid, FDCA) with monoꢀethylene glycol gives
1

Green Chemistry
Page 2 of 31
View Article Online
DOI: 10.1039/C7GC01405H
rise to PEF (polyethyene furanoate), a polyester with superior barrier and thermal properties
7
compared to extensively used PET (polyethylene terephthalate) . However, the current conversion
of plantꢀbased stocks yields parallel substantial quantities of insoluble and intractable, black, sticky
8
ꢀ10
substances usually referred to as humins . These humins are generally considered to comprise of
a series of oligomers (270ꢀ650 g/mol) containing furanic, sugars, and aromatic moieties linked by
mainly short aliphatic chains. Additionally, they include at a smaller extent acetal or ether bonds,
deriving from random condensation reactions of the main bioꢀderived molecules formed during the
11,12
acidꢀcatalysed dehydration of sugars (e.g. HMF, furfural, carbohydrates, levulinates, Scheme 1)
.
Functional groups including carboxylic, ketone, aldehyde and/or hydroxyl are shown by bulk
analysis in the final structure of humin byꢀproducts, thus the presence of superficial oxygen
13
functionalities are expected . Typical elemental composition of humins is found to be in the order
1
4
15
of ca. 65wt% carbon, 5wt% hydrogen, and 30wt% oxygen . Recently, Bruijnincx et al. reported
the quantification and classification of carbonyls groups for industrial humins and lignin, finding a
higher contents of both aliphatic and conjugated carbonyl groups in the humins sample (6.6 wt%)
compared to Alcell (3.3 wt%) and Indulin Kraft (1.7 wt%) lignins. The significant presence of
organic functional groups and the broad range of molecular weights suggest a wide applicability of
humin byꢀproducts as stabilizers, templates and combustible agents, to name a few. It could be
hypothesised that the residual oxygen functionalities could play an important role as catalytic active
16
sites for redox and oxidation reactions, in a similar fashion to graphene oxide .
Current research on humins is mainly focused on limiting the production of these
17
compounds, so far considered waste . Nonetheless, a few humin valorisation studies have been
1
8
carried out, e.g. gasification to syngas , hydrocracking and hydrodeoxygenation to hydrocarburic
19
molecules or as a resin additive as shown by the improved mechanical properties of
20
PFA(perfluoroalkoxy alkanes)/humins composite . The limited number of valorisation studies, so
far promted to research the possibilities of employing humins as a material additive or a platform
chemical. Alternatively, this study has been aimed to provide an alternative valorisation of humins
byꢀproducts as material composite/template in the synthesis of active catalytic materials.
2

Page 3 of 31
Green Chemistry
View Article Online
DOI: 10.1039/C7GC01405H
Scheme 1. Pathways formation of humin byꢀproducts.
Selective oxidative cleavage of hydrocarbons is one of the most relevant reactions in organic
synthesis, finding applications in the conversion of both fossil and renewable feedstocks to
21ꢀ25
industrial bulk and fine chemicals
. Traditional approaches often resort to toxic and/or
environmentally harmful oxidants, e.g. ozone, permanganate, dichromate, or hypervalent iodine
27ꢀ32
reagents
, which also often lead to extremely high Eꢀfactors (i.e. the ratio of mass of waste
26
products over the mass of the desired compound) . Thanks to increasingly stringent environmental
policies, but also an increased awareness of the scientific community in terms of sustainability and
renewability, research focuses have shifted towards the implementation of catalytic alternatives that
33
allow the use of greener oxidants, such as molecular oxygen and hydrogen peroxide .
Homogeneous and heterogeneous catalytic systems have been both investigated in the selective
oxidative cleavage of various olefins, where i) homogeneous catalysts are often of difficult
preparation and recoverability, and ii) heterogeneous catalysts are often nonꢀselective and prone to
deactivation. Nonetheless, major improvements in both fields of catalysis have been reported in
literature, as well as metalꢀfree conversions of alkenes, which are throughly discussed in the recent
3
4
and excellent review by Urgoitia et al. . Among the metalꢀbased catalytic systems, iron catalysts
show promising results in the oxidative cleavage of olefins.
In fact, as one of the most abundant, inexpensive and environmentally friendly metals on earth, iron
has attracted the attention of researchers as a candidate in the synthesis of catalytic materials,
finding applicability in a variety of emerging fields, ranging from energy storage to
35,36
biomedicine . Moreover, ironꢀbased catalysis has showed potential in the green and selective
oxidative cleavage of alkenyl aromatics (e.g. styrene), both in homogeneous and heterogeneous
systems. As homogeneous catalysis concerns, Pillai et al. have reported the innovative solventless
oxidation of a variety of cyclic olefins employing hydrated iron(III) nitrate and 3.4 bar of molecular
oxygen at 60ꢀ80 °C, obtaining high yields in the oxidation of styrene to benzaldehyde, while lower
conversion to the corresponding carbonyls were obtained for the other alkenes present in the
3
7
study . A ligand bearing homogeneous Fe(III)ꢀbased catalyst was instead investigated more
recently by Xiao and coworkers in the aerobic cleavage of a plethora of monoꢀ and disubstituted
styrenes, showing remarkable isolated yields of the corresponding aldehydes or ketones with only 1
38
2
bar of O at 70ꢀ78 °C .
Although homogeneous catalysis often results in exceptionally high conversions and selectivities,
the reusability and separation from the reactor mixture of the catalyst still cause major concerns in
the scientific community. Thus, the development of catalytic systems with facile recoverability
3

Green Chemistry
Page 4 of 31
View Article Online
DOI: 10.1039/C7GC01405H
often resorts to heterogenization of the active material. Along these lines, the high stability and the
magnetic properties of iron oxides nanoparticles make these materials attractive to use as catalytic
39,40
materials . Iron oxide is found as different polyforms including magnetite, i.e. Fe
4
O
3
(iron(II,III)
41
oxide); or αꢀFe O (hematite) and γꢀFe O (maghemite), i.e. iron(III) oxide phases . The magnetic
2 3 2 3
properties of magnetite and maghemite make these oxides highly attractive as a convenient
approach for separation and reusability of these materials. Furthermore, iron oxide has been proven
4
2ꢀ44
as an effective catalyst for various oxidation reactions, e.g.
. In the light of these properties, iron
oxideꢀbased catalytic systems have been investigated as well in the selective oxidation of vinyl
aromatics. For instance, Rak et al. compared hollow, etched and solid ferrite nanoshells in the
conversion of styrene to benzaldehyde under 4 bar of oxygen at 90 °C, obtaining satisfactory
4
5
results . Hong et al. reported the oxidation of styrene employing iron oxide coated platinum
nanowires finding acetonitrile as the best solvent and, even though the highest benzaldehyde yield
was obtained with the use of molecular oxygen (38.2% at 70 °C for 24 h under 1 atm of oxygen),
the highest selectivity to the aldehyde was achieved by employing hydrogen peroxide as an
4
6
oxidant .
Since the only theoretical byꢀproduct produced is water, hydrogen peroxide has also
attracted the attention of green researchers involved in the selective oxidation of alkenyl aromatics
47ꢀ49
with remarkable results and Feꢀbased catalysts have been found able to activate such oxidant
.
As for iron oxide concerns, Shi et al. investigated nanoꢀγꢀFe in the oxidation of alcohols and
2
O
3
50
olefins including styrene, obtaining higher yields using higher equivalents of hydrogen peroxide .
Exceptionally high isolated yields were reported by Rajabi et al. in the hydrogen peroxideꢀiron
51
oxide nanoparticles oxidation of styrene derivatives in aqueous media under reflux conditions .
Thus, iron oxide nanoparticles coupled to hydrogen peroxide have proved to be active in the
selective oxidation of alkenyl aromatics.
Vanillin is one of the most important flavouring employed in pharmaceutical, food,
5
2
beverages, and perfume industries ; however, natural extraction of vanillin from vanilla beans only
yields to a mere fraction (ca. 50 tons) of the total annual vanillin demand (>15000 tons), making
53,54
synthetic vanillin a must
. Current production of synthetic vanillin is mainly based on the
guacaiolꢀbased route using depleting fossil feedstocks as carbon source, thus raising an
55
environmental and sustainability concern , while bioconversion strategies of isoeugenol are often
5
6ꢀ58
complex processes that lead to low vanillin yields
routes need to be developed, comprising the use of bioꢀderived substrates (e.g. isoeugenol) and
green oxidants (e.g. H ).
. Hence, more sustainable, efficient and green
2
O
2
4

Page 5 of 31
Green Chemistry
View Article Online
DOI: 10.1039/C7GC01405H
Isoeugenol is an alkenyl aromatic present in certain essential oils, such as clove oil, tobacco,
59ꢀ61
dill seed, gardenia, petunia, to name a few
; moreover, isoeugenol can be also derived from the
extraction of sawdust and the controlled depolymerization of lignin, i.e. an important component of
62
inedible biomass whose conversion to highꢀend products still requires further research . For this
reason, isoeugenol has been generally viewed as a model compound, facilitating the identification
of the best catalyst to be further upgraded for the conversion of lignin residues. As a consequence,
the selective oxidative cleavage of the vinyl group of isoeugenol is a promising strategy towards a
ligninꢀbased production of synthetic vanillin, thanks to its possible derivation from inexpensive,
natural, and inedible sources. Focusing on nonꢀenzymatic isoeugenol conversion strategies,
Herrmann et al. investigated the use of methyltrioxorhenium catalyst in the oxidative cleavage of
isoeugenol and transꢀferulic acid to vanillin in the presence of anhydrous hydrogen peroxide which
was postulated to run via subsequent epoxidation and diolization steps followed by a final oxidation
63
to vanillin . Shul’pin et al. reported the oxidation of isoeugenol to vanillin by hydrogen peroxide
with a combinative nꢀBu
4
NVO
3
/pyrazineꢀrꢀcarboxylic acid (PCA) catalytic system under air
64
proposing a radical mechanism . The mechanistic differences given by a strict dependency
between oxidant/catalytic stystem further complicate the understanding and implementation of the
oxidative cleavage of isoeugenol. Moreover, these homogeneous systems present the usual
drawbacks of these types of catalysis (vide supra), thus incentivating research towards the use of
heterogeneous catalysts. Heterogenized coꢀporphyrin complexes intercalated into lithium taeniolite
clay were studied by Adilina et al., finding a vanillin yield of 72% when employing low isoeugenol
65
2
concentrations (0.3 mmol) and high 100% O pressure (3 bar) in acetonitrile . Benefits of a low
isoeugenol concentration were also observed by Bohre et al. in the aerobic oxidation of isoeugenol
over reduced graphene oxideꢀsupported copper oxide, but also a higher pressure and temperature (4
bar O
2
at 50 °C) were found to suppress byꢀproducts formation and improve overall vanillin yields.
Moreover, the authors suggest a synergistic effect of the electronꢀrich graphene oxide in stabilizing
16
the copper species, which could be expected from the humin byꢀproducts as well . Both of the
aforementioned heterogeneous isoeugenol conversion studies imply the formation of a
dioxygenated intermediate between the isoeugenol radical and the activated metal centre prior the
selective cleavage to vanillin, discarding the epoxidation route proposed by Herrmann et al. (vide
supra). A radicalꢀfree mechanism is more likely to favor the selectivity towards one particular
product (e.g. aldehyde, epoxide), and suppress the formation of high molecular weight coꢀproducts.
Thus, although an unequivocal mechanism for the selective oxidation of the vinyl group of
isoeugenol has yet to be established, the consistency in the mechanistic speculations further support
the investigation of heterogeneous catalytic systems. To the best of our knowledge, ironꢀbased
5

Green Chemistry
Page 6 of 31
View Article Online
DOI: 10.1039/C7GC01405H
heterogeneous catalysts have yet to be reported in the oxidative cleavage of isoeugenol to vanillin,
in particular, if combined to a carbonꢀbased sideꢀproduct such as humins.
66
One of the 12 principles of Green Chemistry developed by Anastas and Warner states the
importance of minimizing the use of solvents whenever possible. Keeping this principle in mind,
this work aimed to prove the possibility of synthesizing catalytic active materials from humins and
iron precursors without the aid of any solvent. In particular, mechanochemical synthesis has taken
hold as a promising route for the preparation of catalytic materials in solventꢀfree conditions. It was
speculated that the combination of the accumulated potential energy, shear and friction forces in the
milling process create a highly defective surface which could lead to an improved reactivity of the
67
material . Thus, ball milling (BM) as well as alternative thermal degradation (TD) techniques in
the absence of any solvent were investigated for the preparation of magnetic catalytic materials,
68
employing thermally treated (humins foams ) and fructoseꢀderived humins provided by Avantium
B.V. (Netherlands), respectively. The catalytic activity of these materials was subsequently
investigated in the microwaveꢀassisted selective oxidation of isoeugenol to the high added value
product, vanillin.
Experimental
Materials
69
Humins and humins foams were provided by Avantium B.V. (Netherlands) and used as
received. The catalysts precursors, ferrous chloride tetrahydrate (99%, FeCl ·4H O) and ferric
nitrate nonahydrate (≥98%, FeNO ·9H O) were purchased from Sigma Aldrich, as well as Fe
and Fe nanopowders (nanoꢀFe and nanoꢀFe , respectively) which were used as a
2
2
3
2
2 3
O
3
O
4
2
O
3
3 4
O
comparison to the synthesised nanocomposites in the catalytic testing. Concerning the materials for
the reaction mixture, hydrogen peroxide (50wt%) and isoeugenol (98%) were purchased from
Sigma Aldrich, whereas acetonitrile (99.9%) was purchased from Panreac. All of the materials were
used without further purification.
Catalyst preparation
Ferrous chloride tetrahydrate (FeCl ·4H O) and ferric nitrate nonahydrate (FeNO ·9H O)
2
2
3
2
were used for the synthesis of iron oxide composites. FꢀBMꢀCL and FꢀBMꢀNO catalysts were both
6

Page 7 of 31
Green Chemistry
View Article Online
DOI: 10.1039/C7GC01405H
prepared from humins foams (
g of ferrous chloride (CL) were employed for FꢀBMꢀCL, and the same amount in weight of ferric
nitrate (NO) for FꢀBMꢀNO; the iron precursors were respectively milled at 350 rpm for 45 minutes
2 minutes intervals) with 4 g of humins foams, in order to assure the complete grinding of the
carbonaceous material. Samples were subsequently dried overnight at 100 °C and calcined at 400
F) in a planetary ball mill (BM) Retsch, model PM 100. In particular,
2
(
°C for 4 hours (β=5 °C/min).
Hꢀcatalyst series were instead synthesised by alternative thermal degradation (TD) of
3 2
humins (H) and FeNO ·9H O. The intrinsic stickiness of humins (caramelꢀlike, i.e. highly viscous
black liquid) makes these compounds difficult to handle, therefore an uncertainty of ±0.3 g in the
quantities of humins used should be taken into consideration. HꢀTDꢀC and HꢀTDꢀW were
synthesised by liquefying the humins in an oven at 100 °C over the iron precursor. In particular, Hꢀ
TDꢀC was prepared by close contact (
in a crucible for ten minutes at 100 °C; for HꢀTDꢀW, instead, the humins (5 g) were liquefied at
00°C from a higher altitude compared to the layer of the iron precursor (2.5 g), by employing a
watchglass ( ) over the crucible. For both samples, an increase of the volume of the solid solution
C) melting of humins (4 g) over a layer of ferric nitrate (4 g)
1
W
has been observed, seemingly due to an incorporation of gases in the mix. A plausible explanation
would be that upon decomposition of the iron(III) precursor, the humins undergo partial
2
mineralization to CO , causing the visible increase of volume. Both of the samples were then
manually ground, and calcined with the same conditions of the ironꢀfoams materials (vide supra).
For the HꢀTDꢀL and HꢀTDꢀE catalysts, discs of humins were prepared by melting the caramelꢀlike
carbon material at 100 °C over a nonꢀsticky surface and subsequently solidified in a fridge. In
particular, a layer (
quantity that the surface of the layer could uptake) and calcined in a folded position, yielding to Hꢀ
TDꢀL; HꢀTDꢀE was synthesised by rolling a 2 g humins disc around 1 g of FeNO ·9H O, in a way
that all the iron(III) nitrate would be embraced ( ) by the carbon material. Both samples were then
L) of 3 g of humins was covered with 0.66 g of iron precursor (maximum
3
2
E
directly calcined at 400 °C, as all the other samples.
7

Green Chemistry
Page 8 of 31
View Article Online
DOI: 10.1039/C7GC01405H
A
B
1
4
2
5
3
6
Figure 1. Images of the starting materials (Aꢀhumins, Bꢀhumins foams), and synthesised
nanocomposites (1ꢀFꢀBMꢀCL, 2ꢀFꢀBMꢀNO, 3ꢀHꢀTDꢀE, 4ꢀHꢀTDꢀL, 5ꢀHꢀTDꢀW, 6ꢀHꢀTDꢀC).
Catalyst Characterization
Nitrogen physisorption measurements were performed at 77 K by using an ASAP 2000
volumetric adsorption analyzer from Micrometrics. Samples were degassed for 24 h at 130 °C
ꢀ2
under constant vacuum (p<10 Pa) before performing the measurements. Surface areas were
calculated according to the BET equation, whereas average pore diameter and pore volume were
2
obtained from the N desorption branch.
Xꢀray Powder Diffraction (XRD) experiments were recorder on a Bruker D8 Discover
diffractometer (40 kV, 40 mA), equipped with a goniometer Bragg Brentano θ/θ of high precision,
8

Page 9 of 31
Green Chemistry
View Article Online
DOI: 10.1039/C7GC01405H
using CuKα (λ=0.15406 nm) radiation. Scans were performed over a 2θ range from 10 to 80, at step
size of 0.05° with a counting time per step of 143.3 s.
The chemical composition of the surface and the study of the chemical state of the elements
constituent of solids were carried out by XPS. A Physical Electronics spectrometer (PHI Versa
Probe II Scanning XPS Microprobe) was used, with scanning monochromatic Xꢀray Al Kα
radiation (100 ꢁm, 100 W, 20 kV, 1,486.6 eV) as the excitation source, and a dual beam charge
neutralizer. Highꢀresolution spectra were recorded at a given takeꢀoff angle of 45° by a concentric
hemispherical analyzer operating in the constant pass energy mode at 23.5 eV, using a 1400 ꢁm line
(
with a 100 ꢁm diameter of the xꢀray highly focused beam) analysis area. The spectrometer energy
scale was calibrated using Cu 2p3/2, Ag 3d5/2, and Au 4f7/2 photoelectron lines at 932.7, 368.2 and
4,0 eV, respectively. Under a constant pass energy mode at 23.5 eV condition, the Au 4f7/2 line
8
was recorded with 0.73 eV FWHM at a binding energy (BE) of 84.0 eV. PHI Smart SoftꢀVP 2.6.3.4
software package was used for acquisition and data analysis. A Shirleyꢀtype background was
subtracted from the signals. Recorded spectra were always fitted using Gauss–Lorentz curves.
Atomic concentration percentages of the characteristic elements of the surfaces were determined
taking into account the corresponding area sensitivity factor for the different measured spectral
regions.
Scanning electron micrographs (SEM) and elemental composition of materials were
recorder using both the JEOL JSM 6300 and the JEOL JSM 7800F microscopes, the latter
employed for energyꢀdispersive Xꢀray microanalysis (EDX) Xꢀmax 250 (detector SiLi, ATW2), at
2
0 kV.
Transmission electron micrographs (TEM) were recorded on a JEOL JEM 1400 instrument
at 120 kv with a lattice resolution of ca. 0.4 nm. Samples were suspended in ethanol and deposited
straight away on a copper grid prior to analysis.
Magnetic susceptibility was measured at room temperature with the Bartington MSꢀ2
instrument at low frequency (470 Hz).
Catalytic testing
The microwaveꢀassisted oxidation of isoeugenol to vanillin was performed in a CEMꢀ
Discover microwave reactor equipped with a PCꢀcontrolled interface. In particular, the Discover
9

Green Chemistry
Page 10 of 31
View Article Online
DOI: 10.1039/C7GC01405H
method in a sealed vessel was employed, which allows the system to build up pressure with the
formation of gaseous products during the reaction. The observed pressure increase could however
not be ascribed to mineralization of isoeugenol to CO
ideal gas law calculations. In all experiments, 2.5 mg of catalyst, 0.2 mL of isoeugenol, 0.3 mL of
(50wt%), and 1 mL of acetonitrile were employed, under 300 W of irradiation. Different
reaction times were carried out for these experiments, due to the runaway character of the oxidation
rapid increase of pressure). Therefore, 3 minutes reactions with a maximum reachable temperature
2 2 2 2
, or decomposition of H O to O , based on
2 2
H O
(
of 140 °C under variable irradiation (0ꢀ300 W) were performed in order to have a better
understanding of the catalytic activities. Blank reactions were also carried out under these
conditions, in particular, were carried out: i) a reaction with solely the reagent mixture, ii) a reaction
with the reagent mixture in the presence of the humins foams (2.5 mg), iii) a reaction in the
presence of (commercial) nanoꢀFe O and iv) a reaction in the presence of presence of (commercial)
2
3
3 4
nanoꢀFe O (2.5 mg). Recyclability tests were run for all the magnetic nanocomposites up to the
first reuse with 2.5 mg of catalyst. Reusability was investigated for the most stable nanocomposites
with 50 mg of catalyst, in the same reaction conditions stated before. Results were analysed by GC
using an Agilent 7890A fitted with a Supelco column (100 mx0.25 nmx0.5ꢁm) and a flame
ionization detector (FID). Quantification of the products was obtained from the ratio of the
corresponding peak areas, as justified by a close agreement of some representative product/reagent
mixtures compared against a multicomponent calibration curve previously done by our research
group. Furthermore, no solid residues or increase in viscosity was observed at the end of the
reactions, thus indicating no unwanted polymerization of isoeugenol. This further was substantiated
by consistent total peak areas in the GC analysis (vide supra). The reaction products were
confirmed by HPLC and GCꢀMS analysis, carried out by the staff of the Central Service for
Research Support (SCAI) of the University of Córdoba, Spain.
Results and Discussion
The catalysts presented in this study have been synthesised via ball milling of humins foams
with two different iron precursors (FꢀBMꢀCL, iron(II) chloride; and FꢀBMꢀNO, iron(III) nitrate) or
by alternative thermal degradation of humins combined to Fe(III) nitrate in alternative ways,
resulting into the Hꢀcatalyst series (HꢀTDꢀE, HꢀTDꢀL, HꢀTDꢀC, HꢀTDꢀW). In particular, HꢀTDꢀE
and HꢀTDꢀL were synthesised by employing humins discs either embracing (HꢀTDꢀE) or being
covered (HꢀTDꢀL) by the iron precursor; HꢀTDꢀC and HꢀTDꢀW were instead prepared by the
1
0

Page 11 of 31
Green Chemistry
View Article Online
DOI: 10.1039/C7GC01405H
melting of humins either in close contact with the iron precursor (HꢀTDꢀC) or from a higher altitude
(HꢀTDꢀW) with the aid of a watchglass over the crucible.
The resulting textural properties determined by the BETꢀN physisorption method of the
2
composites are summarised in Table 1. The physical nature of humins implies no porosity, whereas
the BET analysis of humins foams (entry 2) shows generally low values of surface area and pore
volume. Interestingly, all of the composites, present systematically higher surface areas and pore
volumes, with pore diameters in the mesopore range (2ꢀ50 nm). Prominently, the use of the ferrous
2
chloride precursor (FꢀBMꢀCL, entry 3) led to a very low surface area solid (ca. 8 m /g), found to be
2
consistently higher when ferric nitrate was employed (S.A.FꢀBMꢀNO=34 m /g). For the Hꢀseries, Hꢀ
TDꢀL possessed the highest surface area with a smaller average pore diameter, given probably by
the layer disposition of the iron(III) precursor on the humins disc; the encapsulation of iron(III)
nitrate into a humins disc (HꢀTDꢀE, entry 5) led, as expected, to a lower surface area and pore
volume. Nonetheless, the average pore diameter of the HꢀTDꢀE composite is larger as compared to
other Hꢀseries materials, signalling the formation of shorter and wider pores, probably due to the
spatial constriction given by the humins embrace during preparation.
Table 1. Textural properties of humins and foams composites.
Surface area/ Pore volume/ Average pore diameter/
2
ꢀ1
3
ꢀ1
m g
cm g
Nm
ꢀ
a
1
2
3
4
5
6
7
8
Humins
ꢀ
ꢀ
b
Foams
<5
0.06
0.11
0.15
0.11
0.19
0.20
0.17
11.8
43.2
12.5
16.4
9.6
FꢀBMꢀCL <10
FꢀBMꢀNO 34
HꢀTDꢀE
HꢀTDꢀL
HꢀTDꢀW
HꢀTDꢀC
20
58
45
43
12.9
12.1
1
1

Green Chemistry
Page 12 of 31
View Article Online
DOI: 10.1039/C7GC01405H
a
Physical nature of humins (black liquid phase) does not allow BET analysis
b
Starting material of Fꢀcatalyst series
Powder Xꢀray Diffraction (XRD) analysis results of the materials are presented in Figure 2.
The high intensity of the peaks show a high degree of cristallinity for all the samples. In particular,
FꢀBMꢀCL presents an almost pure hematite phase, indicating the previously reported oxidation of
4
9,69
Fe(II) precursor to Fe(III) during the preparation under ambient atmosphere
. Mixtures of iron
oxide phases were observed, with a lower degree of cristallinity, when Fe(III) precursors were
employed probably due to impurities. In fact, the peak at 33.1° (*, Figure 2) is typical of hematite,
whereas the 2θ=35.5 (°, Figure 2) can be adscribed to either maghemite, magnetite, or hematite,
indistinguishable by XRD. As a first analysis, FꢀBMꢀNO and HꢀTDꢀL exhibited a higher hematite
content (HꢀTDꢀC has an almost 50:50 ratio of the iron oxides phases), while HꢀTDꢀW and HꢀTDꢀE
have a lower hematite content. XRD analysis of HꢀTDꢀE also shows the presence at 31.6° and 33.7°
(#, Figure 2), probably due to the presence of impurities including sodium sulfate or similar most
probably coming from the humins matrix.
1
2

Page 13 of 31
Green Chemistry
View Article Online
DOI: 10.1039/C7GC01405H
Figure 2. XRD patterns of the Fꢀ and Hꢀseries composites. (*,hematite; °, either maghemite,
magnetite, or hematite; #, sodium sulfate or similar, see text)
In fact, Xꢀray Photoelectron Spectroscopy (XPS) analysis showed for all the samples the presence
of Fe, C and O (Table 2), but also Cl (0.3ꢀ1.1 wt%), S (1.1ꢀ7.6 wt%, as sulfate) and Na (3.5ꢀ13.3
wt%), in good agreement with XRD results and supporting its procedence from intrinsic impurities
in humins and foams. In particular, HꢀTDꢀE showed the highest sodium and sulfur contents of
respectively 13.3 wt% and 7.6 wt%, which justifies the presence of the unidentified XRD peaks
(vide supra). Importantly, only Fe(III) species were found for all the samples (Figure 3), concluding
that the composites are mixtures of hematite and maghemite phases, except for FꢀBMꢀCL which is
purely hematite (see XRD).
1
3

Green Chemistry
Page 14 of 31
View Article Online
DOI: 10.1039/C7GC01405H
13000
12000
11000
10000
9000
8000
7000
6000
5000
4000
740
735
730
725
720
715
710
705
700
Binding Energy (eV)
Figure 3. XPS spectra of the Fe2p regions of FꢀBMꢀNO, representative of all the nanocomposites.
Detailed analysis of the C 1s spectra (Figure 3) showed the presence of CꢀC (48.7ꢀ70.2%), CꢀO
ꢀ
(
19.3ꢀ34%), C=O (3.8ꢀ12.4%), and –COO (4.8ꢀ8.6%) species. These contributions can be
adscribed to residual humins which are expected to have a higher degree of aromatization upon
9
14
heating as previously found by Hoang and van Zandvoort , which might lead to rapid electron
trasmission, thus higher catalytic activity.
Table 2 summarises the mass concentrations of Fe, C, and O obtained from XPS, alongside
the magnetic susceptibility of the composites. The use of Fe(II) chloride precursor led to a
nonꢀmagnetic material (FꢀBMꢀCL, entry 1), whereas the use of Fe(III) nitrate resulted in the
formation of promising magnetic nanocomposites.
Table 2. Mass concentrations of iron, carbon and oxygen deducted by XPS analysis, and magnetic
susceptibility.
Mass Concentrations (%)
Magnetic susceptibility/
ꢀ6
3
ꢀ1
1
0 m kg
Fe2p
C1s
O1s
a
1
2
3
FꢀBMꢀCL
FꢀBMꢀNO
HꢀTDꢀE
46.28
53.85
39.64
11.07
9.27
34.99
33.63
36.25
ꢀ
51.8
15.07
118.9
1
4

Page 15 of 31
Green Chemistry
View Article Online
DOI: 10.1039/C7GC01405H
4
5
6
HꢀTDꢀL
HꢀTDꢀW
HꢀTDꢀC
48.72
57.12
57.06
9.08
6.28
8.11
35.12
32.14
31.9
176.8
202.3
178.4
a
Nonꢀmagnetic material
For the Hꢀseries catalysts, it is possible to correlate a higher iron concentration to a greater magnetic
susceptibility, thus a lower hematite content, in line with the first interpretation by XRD analysis.
Nonetheless, the extremely high superficial oxygen content (>50 at.%, XPS) suggests a potential
use of these composites as oxidation catalysts. The order of iron content determined by XPS
decreased in the order: HꢀTDꢀW≈HꢀTDꢀC>FꢀBMꢀNO>HꢀTDꢀL>FꢀBMꢀCL>HꢀTDꢀE.
Morphology studies and SEMꢀEDX of the foams and humins composites show a highly defective
and nonꢀhomogeneous surface for most of the samples. Tendentiously, the composites form curious
agglomerates from which independet nanoparticles can be clearly visualised at higher
magnifications.
1
5

Green Chemistry
Page 16 of 31
View Article Online
DOI: 10.1039/C7GC01405H
1
6

Page 17 of 31
Green Chemistry
View Article Online
DOI: 10.1039/C7GC01405H
6
Figure 4. SEM micrographs of the iron oxide/humins composites. A1ꢀA2: FꢀBMꢀCL; B1ꢀB2: Fꢀ
BMꢀNO; C1ꢀC2: HꢀTDꢀE; D1ꢀD2: HꢀTDꢀL; E1ꢀE2: HꢀTDꢀW; F1ꢀF2: HꢀTDꢀC.
SEM micrographs of FꢀBMꢀCL (Figure 4ꢀA) shows the most regularity in shape, being
hexagonal particles of various sizes (100ꢀ500 nm), in good agreement with the crystal lattice shape
of iron oxide nanoparticles from TEM analysis (Figure 5ꢀA). The hexagonal particles of FꢀBMꢀCL
were found to contain ca. 56wt% Fe, 10wt% C and 33 wt% O, and the remaining 1wt% as
impurities (i.e. Na, S, and Cl) via SEMꢀEDX analysis independent from the area of sampling,
showing a fairly homogeneous sample. Similar values of weight percentages of the elements found
with XPS analysis (vide supra) suggests the presence of iron oxide both in the bulk and the surface
of the particles. The strucuture of FꢀBMꢀNO (Figure 4ꢀB1), instead, reminds the macroscopic
structure of untreated humins foams (Figure 1ꢀB), however presenting embedded spherical ironꢀrich
particles (SEMꢀEDX: Fe 82wt%, C 3wt%, O 12wt%) of the order of 1ꢀ2 ꢁm (Figure 4ꢀB2). This
could be given by the reduction by carbon of the iron(III) species towards metallic iron, creating an
ironꢀrich core. TEM analysis of FꢀBMꢀNO (Figure 5ꢀB) depicts irregular nanoparticles of 20 nm on
1
7

Green Chemistry
Page 18 of 31
View Article Online
DOI: 10.1039/C7GC01405H
average (extremes identified as between 6 and 32 nm) arranged to create a porous material,
suggesting that the ball milling of foams and iron(III) nitrate gave a similꢀtemplate synthesis.
The Hꢀseries of materials are even less homogeneous, even at higher magnification. SEM
micrographs show a more flakeꢀlike structure of iron oxide agglomerates in all samples, however
differing from the preparation method (Figures 4ꢀC,ꢀD,ꢀE,ꢀF). For instance, HꢀTDꢀL, where the iron
precursor was layered on a humins disc, shows areas where the iron oxide has directly deposited
onto the disc (Figure 4ꢀD1, SEMꢀEDX: Fe 65wt%, C 13wt%, O 20wt%), with the flakeꢀlike
structures having an elemental composition of similar order (Figure 4ꢀD2, SEMꢀEDX: Fe 54wt%, C
7
wt%, O 34wt%). On the other hand, for the carbonꢀembraced iron precursor preparation which
resulted in HꢀTDꢀE, SEMꢀEDX analysis shows blocks of high carbon content materials (Figure 4ꢀ
C1, SEMꢀEDX: Fe 0.4 wt%, C 83wt%, O 16wt%), probably derived from the outside shell of the
humins disc not in contact with the iron precursor, and at the same time flakeꢀlike structures of high
iron content (Figure 4ꢀC2, SEMꢀEDX: Fe 79wt%, C 5wt%, O 10wt%). Figure 3ꢀE and –F show
instead the SEM pictograms of the samples prepared by either close contact or higher altitude
melting of humins, thus HꢀTDꢀC and HꢀTDꢀW, respectively. SEMꢀEDX of both samples showed
high iron content of the flakeꢀlike structures in a similar fashion as HꢀTDꢀE, finding 76wt% of iron
for HꢀTDꢀC (Figure 4ꢀF1) and 79 wt% for HꢀTDꢀW (Figure 4ꢀE2). As for FꢀBMꢀNO (vide supra),
the presence of humins could reduce the iron species towards a zeroꢀvalent state. Figure 4ꢀF2,
instead, shows a highly defective iron oxide block in HꢀTDꢀC (SEMꢀEDX: Fe 69wt%, C 4wt%, O
2
6wt%) which resembles a carbonaceous agglomerate, suggesting again a templateꢀlike synthesis.
TEM images (Figure 4ꢀC, ꢀD, ꢀE, ꢀF) of the Hꢀseries show, where possible to determine,
nanoparticles of up to 300 nm, with the majority in the 4ꢀ30 nm range. In general, an amorphous
carbon can be identified in the samples, often making it difficult to determine the single
nanoparticles. Figures 5ꢀC1, ꢀD1, ꢀF1, and ꢀF2 are representative of the challenging determination
of the nanoparticles in the various samples, almost resembling high resolution pictograms of solid
solutions. Figure 5ꢀC2, in particular, shows what resembles a carbon block decorated with
nanoparticles in the range of 12ꢀ14 nm. This could be ascribed to the chosen synthesis of this
sample (i.e. HꢀTDꢀE), since the encapsulation of the iron precursor by the humins layer might have
dispersed the nanoparticles in the carbon matrix.
1
8

Page 19 of 31
Green Chemistry
View Article Online
DOI: 10.1039/C7GC01405H
1
9

Green Chemistry
Page 20 of 31
View Article Online
DOI: 10.1039/C7GC01405H
Figure 5. TEM micrographs of the iron oxide/humins composites. A: FꢀBMꢀCL; B: FꢀBMꢀNO; C1ꢀ
C2: HꢀTDꢀE; D1ꢀD2: HꢀTDꢀL; E1ꢀE2: HꢀTDꢀW; F1ꢀF2: HꢀTDꢀC.
Catalytic activity of all synthesized nanocomposites was subsequently investigated in the
microwaveꢀassisted selective oxidation of isoeugenol to vanillin (Scheme 2) at a fixed microwave
power (300 W) with a remarkably low catalyst loading of 2.5 mg. Results of conversion and
selectivity for the different catalytic systems have been summarised in Table 3, obtaining the
optimum results for FꢀBMꢀCL with selectivity over 60% to vanillin at ca. 90% conversion after 2
minutes reaction. In general, high conversions could be achieved (>87%), with selectivities to
vanillin ranging between 42 and 64%, although with each system requiring different reaction times
and reaching different temperatures, though of similar order.
Table 3. Catalytic activities of the humins/foams composites in the microwaveꢀassisted oxidation of
isoeugenol to vanillin.
Time
Catalyst
T (°C) P (PSI) Conversion (%) Selectivity (%)
(min)
1
2
3
4
5
FꢀBMꢀCL
2.2
4.8
4.2
3.5
5.5
169
152
152
155
156
241
246
246
244
247
89
88
64
54
63
63
50
FꢀBMꢀNO
HꢀTDꢀC
HꢀTDꢀW
HꢀTDꢀE
89
91
>99
2
0

Page 21 of 31
Green Chemistry
View Article Online
DOI: 10.1039/C7GC01405H
6
HꢀTDꢀL
4.7
149
248
>99
42
Reaction conditions: 2.5 mg of catalyst, 0.2 mL isoeugenol, 0.3 mL H O (50wt%), 1
2
2
mL acetonitrile, irradiation of 300 W until runaway reaction.
Conversion of isoeugenol with hydrogen peroxide by the huminsꢀiron oxide nanocomposites
leads to vanillin as main reaction product, alongside a plethora of byꢀproducts (Scheme 2),
comprising vanillin methyl ketone (1), isoeugenol diol (2), eugenol (3), and dihydrodiisoeugenol
(4), with 1 and 2 being the most prominent coꢀproducts, in order. Similar molecules were reported
16,63ꢀ65
in the past reports of said reactions
. In particular, the literature on the conversion of
6
4
isoeugenol to vanillin claimed maximum: a) 50% yield at 80 °C and 0.4 M of H
2
O
2
after 2 hours ;
1
6
b) 53% yield at 50 °C and 4 bar of O after 24 hours ; and c) 72% yield at 50 °C and 3 bar of O
2
2
6
5
after 24 hours ; the nanocomposites reported herein, instead, grant vanillin yields between 42ꢀ57%
in less than 5 minutes of reactions, although at higher operating temperatures (i.e. 149ꢀ169 °C).
Nonetheless, further optimization of the reaction conditions is needed, as well as a better
understanding of the mechanism of reaction.
Scheme 2. Reaction scheme of isoeugeneol conversion to vanillin and coꢀproducts: vanillin methyl
ketone (1), isoeugenol diol (2), eugenol (3), and dihydrodiisoeugenol (4), presented in order of
magnitude.
By analyzing the reaction products, a mechanistic evaluation might be drafted. Notably, the
absence of vanillic acid as coꢀproduct with our strong oxidizing conditions (isoeugenol/H
2 2
O ratio:
1
:5) and high temperature is rather surprising. Nonetheless, this particular compound was only
7
0,71
reported within biotransformations of isoeugenol to vanillin , suggesting that both homogeneous
and heterogeneous catalytic systems do not promote further oxidation of vanillin to its carboxylic
6
4
acid form. In fact, kinetic studies on vanadium complexes carried out by Shul’pin research group
suggested a faster consumption of isoeugenol compared to vanillin under oxidizing conditions,
therefore limiting the formation of vanillic acid. Thus, short reaction times, such as the ones
2
1

Green Chemistry
Page 22 of 31
View Article Online
DOI: 10.1039/C7GC01405H
reported herein, would minimize the formation of higher oxidized products, although a higher
temperature is required in order to activate the catalytic material in such small time frame.
Moreover, isoeugenol epoxide was not detected among the reaction products, in contrast to the
aforementioned reports; this difference with literature can be simply ascribed to the waterꢀrich
environment given by the H
observed corresponding diol (2), whereas the previous findings in literature operated in anhydrous
environments (O or anhydrous H ). Furthermore, the production of 4 suggests the formation of
radical intermediates of isoeugenol which undergo radical coupling, as supported by previous
2 2
O aqueous solution employed which favors the hydrolysis to the
2
2 2
O
1
6
reports (e.g. ). However, first row transition metalꢀcatalyzed oxidations with hydrogen peroxide
33
2 2
are believed to follow the nonꢀradical oxometal pathway , where H O oxidizes the metal centers
that eventually coordinate the reagent. Furthermore, the contribution from residual humins must be
taken into consideration. In fact, disordered and electronꢀrich carbon materials have been found to
72
be highly active in the decomposition of hydrogen peroxide into radicals . The synergy between
the metal and carbon active sites further complicates the understanding of the overall mechanism
governing the reaction; however, the absence of higher condensation or polymerization products
may suggest the prevalence of a nonꢀradical mechanism, which can be justified by the higher
concentration of iron species on the surface (see XPS). Nonetheless, further kinetic studies are
indeed required to further ascertain the mechanism of ironꢀcarbon nanocomposites. As the scope of
this work is to prove the employability of humins as carbon source for catalytic purposes, particular
focus is given to the selectivity to vanillin throughout this study.
Due to the fixed microwave power employed, resulting into different reaction temperatures
and time, the catalytic activity of the nanocomposites cannot be unequivocally ascribed. As an
attempt to better compare the catalytic activities, all reactions were performed for 3 minutes with a
maximum reachable temperature of 140 °C, with the same catalyst loading, and results are reported
in Table 4. A blank test of the reaction (no catalyst) was carried out under these conditions (Table 4,
entry 0), also in the solely presence of the humins foams (i.e. the support for the FꢀBMꢀCL and Fꢀ
2 3 3 4
BMꢀNO nanocomposites) (Table 4, entry 1), and commercial Fe O and Fe O nanopowders (Table
4
, entry 2 and 3, respectively).
2
2

Page 23 of 31
Green Chemistry
View Article Online
DOI: 10.1039/C7GC01405H
Table 4. Catalytic activities of the humins/foams composites in the oxidation of isoeugenol to
vanillin under microwave irradiation (300 W), with a maximum temperature of 140 °C.
Time
Catalyst
T (°C)
P (PSI)
Conversion (%)
Selectivity (%)
(min)
0
1
2
3
4
5
6
7
8
9
Blank
3
3
3
3
3
3
3
3
3
3
120*
140
140
140
140
140
140
140
140
140
67
120
N/A
N/A
96
36
91
25
28
76
66
66
75
53
60
<20
Foams
<
20
nanoꢀFe O
<20
<20
48
2
3
3
nanoꢀFe O
4
FꢀBMꢀCL
FꢀBMꢀNO
HꢀTDꢀE
103
112
120
107
95
26
32
HꢀTDꢀL
47
HꢀTDꢀW
HꢀTDꢀC
35
28
*Temperature results lower due to the absence of a catalytic material
In general, the results evidenced a moderate to high activity of all nanocomposites under the
investigated conditions, improved as compared to the blank reactions. Notably, the humins foams
resulted into an incredibly remarkable conversion of >90%, although with low selectivity to
1
6
vanillin. This is a clear signal that humins, similar to that reported for graphene oxide , possess
exceptional oxidation properties thanks to the high presence of electronꢀrich functionalities. The
2 2
low selectivity to vanillin might be ascribed to higher H O decomposition rates into unselective
radical species. However, an extreme loss of the support at the end of the reaction was observed,
suggesting that the humins themselves might undergo oxidation by OH• species. When the
commercial nanopowders were employed (entry 2 and 3), low conversion were obtained (<30%),
even lower as compared to the blank reaction in the absence of any catalytic material (36%, entry
0
). This could be ascribed to differences in the governing mechanisms, where i)the solely presence
of hydrogen peroxide under microwave irradiation leads to the formation of OH• species and, thus,
fair conversion although unselective; whereas ii)the commercial nanopowders promote the nonꢀ
radical oxometal pathway, although resulting ineffective activity under the same reaction
conditions. This strongly suggests that a positive synergistic effect must exist between the iron
oxide and carbon residues, given by the systematically higher conversions and selectivities of the
2
3

Green Chemistry
Page 24 of 31
View Article Online
DOI: 10.1039/C7GC01405H
nanocomposites (Table 4, entry 4ꢀ9). Surprisingly, vanillic acid was detected only when nanoꢀFe
3 4
O
(i.e. Fe(II) and Fe(III) mixed oxide, alias magnetite) was employed. A reason for the over oxidation
of vanillin to its carboxylic acid could be given by a possible higher oxidation activity of the Fe(II)
species present in the magnetite nanopowder, which further limit the selectivity to vanillin. This
speculation is in line with the proposed mechanism of styrene oxidation by iron(III) nitrate
37
species .
As for the nanocomposites synthesized for this study, FꢀBMꢀCL (entry 4) yielded to the
highest conversion and selectivity to vanillin. This result can be ascribed to the larger presence of
7
3
highly crystalline hematite in the catalyst, which has been found to be the active iron oxide phase .
Comparably, the order of reactivity (in terms of conversion) for the maghemiteꢀhematite mixed
phase catalysts was found to be as follows: HꢀTDꢀL>HꢀTDꢀE>FꢀBMꢀNO>HꢀTDꢀC>HꢀTDꢀW.
74
According to the findings of Gurol and Lin , hydrogen peroxide activation should be
directly proportional to the iron oxide concentration. However, by analyzing the conversion
efficiency of the Hꢀseries catalysts (Table 4, entry 6ꢀ9), the superficial iron concentration (see
XPS), and the relative hematite/maghemite concentration (see XRD), a similar trend is evidenced if
considering solely the hematite concentration. This suggests that hematite is thus the catalytically
active morph of iron oxide. Along these lines, the higher activity of HꢀTDꢀL could thus be
attributed to the higher hematite content. From a synthetic point of view, the use of a layer (disc) of
humins in the synthesis of HꢀTDꢀL and HꢀTDꢀE has led to better performing catalytic materials, as
compared to the use of native humins (namely in the preparation of HꢀTDꢀC and HꢀTDꢀW). The
process of melting and subsequent cooling the humins into a thin layer might induce the polymeric
chains to reꢀorientate in a more crystalline conformation, playing an overall positive effect in the
electron transfer ability of the byꢀproducts.
Recyclability tests were subsequently conducted for all catalysts, with those presenting
magnetic properties included in Figure 6. Substantial loss of activity (ꢀ15% ca.) is noted for the
HꢀTDꢀL and HꢀTDꢀC composites, whereas only a very small activity loss (<5%) could be observed
for reused HꢀTDꢀE material. Comparably, FꢀBMꢀNO and HꢀTDꢀW fully preserved their initial
activity after the first (Figure 6), illustrating their potentialꢀparticularly FꢀBMꢀNO and FꢀBMꢀCL as
highly stable and reusable huminꢀbased oxidation catalysts. We believe the stability of FꢀBMꢀCL
and FꢀBMꢀNO systems relates to its mechanochemical synthesis and thus inherent chemical
modification of the surface in the final material, in good agreement with previous studies of the
4
9
group . Additionally, no colour change was observed in the final solution/filtrate after recovery of
the catalyst, which may indicate a negligible Fe leaching. In contrast, the high stability of HꢀTDꢀW
2
4

Page 25 of 31
Green Chemistry
View Article Online
DOI: 10.1039/C7GC01405H
may be related to the observed phase transition from maghemite to hematite after reuses (results not
shown). The final solution after reaction was also yellowish, in good agremeent with similar results
observed for FꢀBMꢀNO. HꢀTDꢀW and FꢀBMꢀNO were therefore investigated further up to four
reuses, with a higher catalyst loading of 50 mg in the same reaction conditions.
8
7
6
5
4
3
2
1
0
0
0
0
0
0
0
0
0
Fresh Catalyst
st reuse
1
F-BM-NO H-TD-E
H-TD-L H-TD-W H-TD-C
Figure 6. Recyclability tests for the magnetic humins/foams composites.
Figure 7 shows conversion and selectivity of the two nanocomposites, FꢀBMꢀNO and
HꢀTDꢀW. To be noted, a higher catalyst loading (20 times higher) did improve the conversion of
isoeugenol, but at the expense of the selectivity, in particular for the sample HꢀTDꢀW. This could be
ascribed to a higher hydrogen peroxide decomposition rate which decreases the selectivity to the
aldehyde, while favoring the formation of the coꢀproducts, in particular the vanillin methyl ketone
(Scheme 3, 1). At a first glance, an incredible stability can be seen for both of the catalysts, however
th
a curious behaviour has been noted for the HꢀTDꢀW composite. In fact, the results of the 4 reuse
of HꢀTDꢀW are referred to a spontaneous room temperature reaction (see video, ESI) that occurred
when the catalyst entered in contact with the reaction medium. Fe content after reaction
(filtrate/reused catalyst) measured by ICP/MS showed negligible leaching (<2ppm Fe into solution,
cumulative after five use) after subsequent reuses, with investigated catalysts (i.e. FꢀBMꢀNO and Hꢀ
th
TDꢀW) exhibiting an almost Fe content in the 4 reused catalyst as compared to the fresh one (<5%
Fe loss after 5 uses).
2
5

Green Chemistry
Page 26 of 31
View Article Online
DOI: 10.1039/C7GC01405H
Overall, not only a satisfactory vanillin yield is obtained, but also the nanocomposites seems to
improve its activity with its use. This could be ascribed to a phase rearrangement of the
nanocomposites iron oxide phase towards a more active specie, i.e. hematite, arising from the
oxide/peroxide functionalities of the humins residue. Nonetheless, this behaviour could represent a
breakthrough in both the carbonꢀbased materials (e.g. humins) and the catalysis fields.
100
9
8
7
6
5
4
3
2
1
0
0
0
0
0
0
0
0
0
0
Conversion (%)
Selectivity (%)
Fresh
3rd
4th
Fresh
3rd
4th
F-BM-NO
H-TD-W
Figure 7. Recyclability runs for FꢀBMꢀNO and HꢀTDꢀW up to the fourth reuse with increased
catalyst loading (50 mg).
In any case, further investigations are already ongoing to fully understand and improve the
chemical structure, surface and stability of the synthesized humin/iron nanocomposites, alongside
the optimization of the reaction conditions for the production of vanillin from lignocellulosic
derivatives.
Conclusions
Humins have been generally regarded as undesired byꢀproducts in the acidꢀcatalyzed conversion of
biomass. This work shows the possibility of employing humins in the preparation of catalytic
nanocomposites. All of the composites were prepared via solventꢀfree direct methodologies, in
particular: FꢀBMꢀCL and FꢀBMꢀNO were mechanochemically synthesised via inexpensive ball
2
6

Page 27 of 31
Green Chemistry
View Article Online
DOI: 10.1039/C7GC01405H
milling, whereas HꢀTDꢀE, HꢀTDꢀL, HꢀTDꢀW, and HꢀTDꢀC via alternative thermal degradation. The
use of iron(III) nitrate precursor yielded magnetic composites with mixed iron oxide phases. All of
the composites were found to be highly active in the microwaveꢀassisted oxidation of isoeugenol to
vanillin, with a particularly high stability after reuses of mechanochemically synthesized
nanomaterials. The proposed approach proves the feasibility of producing catalytically active
composites from biorefinery waste and may pave the way of a more intensive utilisation of
biomassꢀderived waste feedstocks towards their valorisation to high added value products.
Acknowledgements
We kindly acknowledge the EU Framework Programm Horizon 2020 for financial support as in the
HUGS project (ID: 675325) and project CTQ2015ꢀ68951ꢀC3ꢀ3R (Ministerio de Economía y
Competitividad, Spain and FEDER Funds). The author would also like to thank the support center
SCAI of the University of Cordoba for the GCꢀMS analysis, Pierluigi Tosi for providing the humins
foams, and María Dolores Márquez Medina for her tremendous support during the completion of
this publication.
References
[
1]
L. Filiciotto, A.M. Balu, J.C. van der Waal, R. Luque, Catal. Today, 2017, DOI:
1
0.1016/j.cattod.2017.03.008.
[
[
[
[
[
[
[
[
2]
L. Wu, T. Moteki, A.A. Gokhale, D.W. Flaherty, F.D. Toste, Chem., 2016, 1, 32.
M.J. Climent, A. Corma, S. Iborra, Green Chem., 2014, 16, 516.
S. De, A.M. Balu, J.C. van der Waal, R. Luque, ChemCatChem., 2015, 7, 1608.
F.D. Pileidis, M.ꢀM. Titirici, ChemSusChem., 2016, 9, 562.
3]
4]
5]
6]
7]
8]
9]
M.ꢀF. Li, S. Yang, R.ꢀC. Sun, Bioresour. Technol., 2016, 200, 971.
Avantium, (n.d.). http://www.avantium.com (accessed February 15, 2017).
F.A.H. Rice, J. Org. Chem., 1958, 23, 465.
T.M.C. Hoang, L. Lefferts, K. Seshan, ChemSusChem., 2013, 6, 1651.
2
7

Green Chemistry
Page 28 of 31
View Article Online
DOI: 10.1039/C7GC01405H
[
[
10] I. V. Sumerskii, S.M. Krutov, M.Y. Zarubin, Russ. J. Appl. Chem., 2010, 83, 320.
11] I. van Zandvoort, Y. Wang, C. B. Rasrendra, E. R. H. van Eck, P. C. A. Bruijnincx, H. J.
Heeres, B. M. Weckhuysen, ChemSusChem., 2013, 6, 1745.
[
12] I. van Zandvoort, E.J. Koers, M. Weingarth, P.C.A. Bruijnincx, M. Baldus, B.M.
Weckhuysen, Green Chem., 2015, 17, 4383.
[
13] A. Mija, J.C. van der Waal, J.ꢀM. Pin, N. Guigo, E. de Jong, Constr. Build. Mater., 2017,
39, 549ꢀ601.
1
[
14] I. van Zandvoort, E. R. H. van Eck, P. de Peinder, H. J. Heeres, P. C. A. Bruijnincx, B. M.
Weckhuysen, ACS. Sustainable Chem. Eng., 2015, 3, 533ꢀ543.
[
15] S. Constant, C.S. Lancefield, B.M. Weckhuysen, P.C.A. Bruijnincx, ACS Sustainable Chem.
Eng., 2017, 5, 965ꢀ972.
[
16] A. Bohre, D. Gupta, M.I. Alam, R.K. Sharma, B. Saha, ChemistrySelect, 2017, 2, 3129ꢀ
136.
3
[
[
[
[
17] X. Tang, Y. Sun, X. Zeng, W. Hao, L. Lin, S. Liu, Energy Fuels, 2014, 28, 4251.
18] J.ꢀP. Lange, Angew. Chemie Int. Ed., 2015, 54, 13186.
19] T.M.C. Hoang, L. Lefferts, K. Seshan, ChemSusChem., 2013, 6, 1651.
20] Y. Wang, S. Agarwal, A. Kloekhorst, H.J. Heeres, ChemSusChem., 2016, 10, 951.
[
21] R. A. Sheldon and J. K. Kochi, Metal Catalyzed Oxidation of Organic Compounds,
Academic Press, New York, 1981.
[
[
[
[
[
[
22] J.E. Bäckvall, Modern Oxidation Methods, Wiley–VCH, 2004.
23] S.M. Roberts, Catalysts for Fine Chemical Synthesis, Wiley–VCH, 2007.
24] I. Ibushi and A. Yazaki, J. Am. Chem. Soc., 1981, 103, 7371ꢀ7373.
25] K. A. Jorgensen, Chem. Rev., 1989, 89, 431ꢀ458.
26] R. A. Sheldon, Green Chem., 2017, 19, 18ꢀ43.
27] D.C. Whitehead, B.R. Travis, B. Borhan, Tetrahedron Lett., 2006, 47, 3797−3800.
2
8

Page 29 of 31
Green Chemistry
View Article Online
DOI: 10.1039/C7GC01405H
[
[
[
[
[
[
[
[
[
[
28] S. Bhadra, L. Adak, Tetrahedron Lett., 2008, 49, 2588−2591.
29] V. Kumar, V.P. Reddy, R. Sridhar, S. Srinivas, K.R. Rao, Synlett, 2009, 2009, 739−742.
30] M. Shaikh, F.ꢀE. Hong, Adv. Synth. Catal. 2011, 353, 1491−1496.
31] K. Ghosh, P. Kumar, I. Goyal, Inorg. Chem. Commun., 2012, 24, 81−86.
32] M.R. Maurya, N. Kumar, J. Mol. Catal. A: Chem. 2014, 383−384, 172−181.
33] H. WójtowiczꢀMłochowska, Arkivoc, 2017, part ii, 12−58.
34] G. Urgoitia, R. SanMartin, M. T. Herrero, E. Domínguez, ACS Catal., 2017, 7, 3050ꢀ3060.
35] H. L. Liu, C. H. Soon, J. H. Wu, K. M. Lee, Y. K. Kin, Biomaterials, 2008, 29, 4003ꢀ4011.
36] M. Armand, J. M. Tarascon, Nature, 2008, 451, 652ꢀ657.
37] U. R. Pillai, E. SahleꢀDemessie, V. V. Namboodiri, R. S. Varma, Green Chem., 202, 4, 495ꢀ
4
97.
[
38] A. GonzalezꢀdeꢀCastro, J. Xiao, J. Am. Chem. Soc., 2015, 137, 8206ꢀ8218.
[
39] P. Tartaj, M.P. Morales, T. GonzalezꢀCarreño, S. VeintemillasꢀVerdaguer, C.J. Serna, Adv.
Mater., 2011, 23, 5243.
[
[
[
40] G.S. Parkinson, Surf. Sci. Rep., 2016, 71, 272ꢀ365.
41] Y.L. Pang, S. Lim, H.C. Ong, W.T. Chong, Ceramics Int., 2016, 42, 9.
42] F. Arena, G. Gatti, L. Stievano, G. Martra, S. Coluccia, F. Frustreri, L. Spadaro, A.
Parmaliana, Catal. Today, 2006, 117, 75ꢀ79.
[
43] F. Shi, M. K. Tse, M. M. Pohl, A. Bruckner, S. M. Zhang, M. Beller, Angew. Chem., Int.
Ed., 2007, 46, 8866–8868.
[
44] K. Schröder, K. Junge, B. Bitterlich and M. Beller, Top. Organomet. Chem., 2011, 33, 83–
09.
1
[
[
45] M. J. Rak, M. Lerro, A. Moores, Chem. Commun., 2014, 50, 12482ꢀ12485.
46] H. Hong, L. Hu, M. Li, J. Zheng, X. Sun, X. Lu, X. Cao, J. Lu, H. Gu, Chem. Eur. J., 2011,
1
7, 8726ꢀ8730.
2
9