ArticlePDF Available

Abstract and Figures

More and more attention has recently been paid to the electrochemical treatment of wastewater for the degradation of refractory organics, such as phenol and its derivatives. The electrodeposition of different types of manganese oxides (MnOx) over two substrates, namely metallic titanium and titania nanotubes (TiO2-NTs), is reported herein. X-Ray Diffraction (XRD) and X-Ray Photoelectron Spectroscopy (XPS) analyses have confirmed the formation of different oxidation states of the manganese, while Field Emission Scanning Electronic Microscopy (FESEM) analysis has helped to point out the evolutions in the morphology of the samples, which depends on the electrodeposition parameters and calcination conditions. Moreover, cross section FESEM images have demonstrated the penetration of manganese oxides inside the NTs for anodically deposited samples. The electrochemical properties of the electrodes have been investigated by means of cyclic voltammetry (CV) and linear sweep voltammetry (LSV), both of which have shown that both calcination and electrodeposition over TiO2-NTs lead to more stable electrodes, which exhibited a marked increase in the current density. The activity of the proposed nanostructured samples toward phenol degradation has been investigated. The cathodically electrodeposited manganese oxides (α-MnO2) have been found to be the most active phase, with a phenol conversion of 26.8%. The anodically electrodeposited manganese oxides (α-Mn2O3), instead, have shown higher stability, with a final working potential of 2.9 V vs. RHE. The TiO2-NTs interlayer has contributed, in all cases, to a decrease of about 1 − 1.5 V in the final (reached) potential, after a reaction time of 5 h. Electrochemical impedance spectroscopy (EIS) and accelerated life time tests have confirmed the beneficial effect of TiO2-NTs, which contributes by improving both the charge transfer properties (kinetics of reaction) and the adhesion of MnOx films.
Content may be subject to copyright.
Applied
Catalysis
B:
Environmental
203
(2017)
270–281
Contents lists available at ScienceDirect
Applied
Catalysis
B:
Environmental
journal homepage: www.elsevier.com/locate/apcatb
Electro-oxidation
of
phenol
over
electrodeposited
MnOx
nanostructures
and
the
role
of
a
TiO2nanotubes
interlayer
Andrea
Massa,
Simelys
Hernández,
Andrea
Lamberti,
Camilla
Galletti,
Nunzio
Russo,
Debora
Fino
Department
of
Applied
Science
and
Technology
(DISAT),
Politecnico
di
Torino,
Corso
Duca
degli
Abruzzi
24,
10129
Torino,
Italy
a
r
t
i
c
l
e
i
n
f
o
Article
history:
Received
27
May
2016
Received
in
revised
form
15
September
2016
Accepted
8
October
2016
Available
online
11
October
2016
Keywords:
Manganese
oxides
Electrodeposition
TiO2nanotubes
Phenol
degradation
Electro-oxidation
a
b
s
t
r
a
c
t
More
and
more
attention
has
recently
been
paid
to
the
electrochemical
treatment
of
wastewater
for
the
degradation
of
refractory
organics,
such
as
phenol
and
its
derivatives.
The
electrodeposition
of
different
types
of
manganese
oxides
(MnOx)
over
two
substrates,
namely
metallic
titanium
and
titania
nanotubes
(TiO2-NTs),
is
reported
herein.
X-Ray
Diffraction
(XRD)
and
X-Ray
Photoelectron
Spectroscopy
(XPS)
anal-
yses
have
confirmed
the
formation
of
different
oxidation
states
of
the
manganese,
while
Field
Emission
Scanning
Electronic
Microscopy
(FESEM)
analysis
has
helped
to
point
out
the
evolutions
in
the
mor-
phology
of
the
samples,
which
depends
on
the
electrodeposition
parameters
and
calcination
conditions.
Moreover,
cross
section
FESEM
images
have
demonstrated
the
penetration
of
manganese
oxides
inside
the
NTs
for
anodically
deposited
samples.
The
electrochemical
properties
of
the
electrodes
have
been
investigated
by
means
of
cyclic
voltammetry
(CV)
and
linear
sweep
voltammetry
(LSV),
both
of
which
have
shown
that
both
calcination
and
electrodeposition
over
TiO2-NTs
lead
to
more
stable
electrodes
that
exhibited
a
marked
increase
in
the
current
density.
The
activity
of
the
proposed
nanostructured
samples
toward
phenol
degradation
has
been
investigated.
The
cathodically
electrodeposited
manganese
oxides
(-MnO2)
have
been
found
to
be
the
most
active
phase,
with
a
phenol
conversion
of
26.8%.
The
anodi-
cally
electrodeposited
manganese
oxides
(-Mn2O3),
instead,
have
shown
higher
stability,
with
a
final
working
potential
of
2.9
V
vs.
RHE.
The
TiO2-NTs
interlayer
has
contributed,
in
all
cases,
to
a
decrease
of
about
1–1.5
V
in
the
final
(reached)
potential,
after
a
reaction
time
of
5
h.
Electrochemical
impedance
spectroscopy
(EIS)
and
accelerated
life
time
tests
have
confirmed
the
beneficial
effect
of
TiO2-NTs,
which
contributes
by
improving
both
the
charge
transfer
properties
(kinetics
of
reaction)
and
the
adhesion
of
MnOxfilms.
©
2016
The
Authors.
Published
by
Elsevier
B.V.
This
is
an
open
access
article
under
the
CC
BY-NC-ND
license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
1.
Introduction
The
biological
treatment
of
wastewater
is
one
of
the
most
com-
mon
processes
throughout
the
world
for
the
abatement
of
a
large
variety
of
compounds.
However,
some
organic
pollutants,
such
as
aromatic
compounds,
are
refractory
and
toxic
to
traditional
biolog-
ical
methods.
Benzene,
phenol
and
its
derivatives
(chloro-phenols
and
nitro-phenols),
anilines,
benzoquinone
and
hydroquinone
are
some
of
the
most
common
pollutants
contained
in
industrial
wastewater
[1,2].
Abbreviations:
MnOx,
manganese
oxides;
Ti,
titanium;
TiO2-NTs,
titania
nano-
tubes;
CV,
cyclic
voltammetries;
LSV,
linear
sweep
voltammetries;
RHE,
reversible
hydrogen
electrode.
Corresponding
author.
E-mail
address:
simelys.hernandez@polito.it
(S.
Hernández).
Phenol,
in
particular,
is
one
of
the
most
studied
molecules
in
the
sector
of
the
removal
of
recalcitrant
organic
compounds
from
water,
due
to
its
high
refractoriness
and
stability,
and
to
its
extensive
presence
in
several
industrial
plants,
such
as
petroleum
refineries,
and
plastics,
pesticides
and
pharmaceutical
factories.
Advanced
oxidation
processes
(AOPs),
such
as
Fenton
and
photo-
Fenton
reactions
[3],
ozonation
[4],
wet
air
oxidation
(WAO)
[5],
catalytic
wet
hydrogen
peroxide
oxidation
(CWHPO)
[6]
and
cat-
alytic
wet
air
oxidation
(CWAO)
[7]
are
all
effective
methods
that
can
be
used
to
treat
this
substance.
However,
electrochemical
degradation
is
another
attractive
procedure
that
is
being
con-
sidered
to
remove
recalcitrant
organics
from
water,
especially
in
low-volume
applications,
due
to
its
intrinsic
high
efficiency
[8,9].
According
to
the
above
mentioned
classification,
electro-
oxidation
is
considered
to
belong
to
the
AOP
family,
since
OH
radicals,
some
one
of
the
most
powerful
oxidizing
agents,
are
produced
on
the
surface
of
the
electrocatalytic
materials.
Many
http://dx.doi.org/10.1016/j.apcatb.2016.10.025
0926-3373/©
2016
The
Authors.
Published
by
Elsevier
B.V.
This
is
an
open
access
article
under
the
CC
BY-NC-ND
license
(http://creativecommons.org/licenses/by-nc-nd/4.
0/).
A.
Massa
et
al.
/
Applied
Catalysis
B:
Environmental
203
(2017)
270–281
271
catalysts
have
been
developed
for
this
particular
application
throughout
the
years:
Pt
[10,11],
IrO2[12,13],
RuO2[14,15],
PbO2
[11,16],
SnO2[17,18]
and
Boron
Doped
Diamond
(BDD)
[19,20].
However,
only
a
limited
amount
of
literature
has
reported
the
use
of
manganese
oxides
(MnOx)
as
catalysts
for
the
electro-
oxidation
of
refractory
organics
[21–26].
Furthermore,
in
many
of
these
works,
the
role
of
manganese
oxides
has
not
been
investi-
gated
in
depth
because
the
MnOxcoating
had
either
been
deposited
onto
highly
active
intermediate
substrates,
such
as
Sb-SnO2[21]
or
RuO2[22,23],
or
it
had
been
doped
with
ions
(Fe2+),
which
can
influence
the
performance
of
the
electrode
[24].
MnOxhave
been
employed
extensively
in
electrochemistry
as
cathodes
in
alkaline
batteries,
in
lithium-ion
batteries
[27]
and
as
pseudo-capacitive
electrodes
into
supercapacitors
[28,29],
or
as
photo-anodes
for
the
water
splitting
reaction
[30,31].
The
main
advantages
of
manganese
oxides
is
their
low
cost,
if
compared
to
Pt,
IrO2,
RuO2and
BDD,
and
lower
toxicity
than
Sb-SnO2and
PbO2.
This
work
describes
the
fabrication
and
characterization
of
electrochemically-deposited
manganese
oxides
over
a
titanium
support
for
the
electrochemical
oxidation
of
phenol
molecules
in
wastewater.
The
electrodeposition
parameters,
such
as
current
density,
deposition
time
and
precursors,
were
tuned
in
order
to
investigate
their
effect
on
the
activity.
Moreover,
the
effect
of
the
temperature
treatment
on
the
electrodeposited
MnOxwas
inves-
tigated
by
means
of
calcination
in
air.
Additionally,
the
role
of
an
interlayer
between
the
MnOxnanostructures
and
the
Ti
substrate
was
considered,
in
order
to
evaluate
its
effectiveness
in
prevent-
ing
the
passivation
of
the
Ti
substrate
and
possibly
increasing
the
electrode
surface
area
[32–36].
For
these
reasons,
the
optimized
parameters
described
above
were
used
for
deposition
over
a
TiO2
nanotubes
(TiO2-NTs)
array,
grown
directly
on
Ti
foil
by
means
of
the
anodic
oxidation
method.
TiO2-NTs
obtained
from
anodization
have
attracted
consider-
able
interest
in
the
last
few
decades,
since
their
unique
properties
make
them
useful
as
active
elements
for
several
applications,
rang-
ing
from
energy
production
(dye-sensitized
solar
cells,
[37,38]
water
splitting,
[39,40])
and
storage
(Li-ions
batteries,
[41,42]
supercapacitors
[43])
to
sensing
devices,
such
as
gas
sensing,
[44]
and
molecular
sensors
[45].
The
main
advantages
of
TiO2-NTs
concern
their
quasi
one-dimensional
arrangement,
which
leads
to
a
good
compromise
between
the
exposed
surface
area
(about
40
m2/g)
and
superior
electron
transport
properties,
and
results
in
a
performance
enhancement
in
all
the
different
fields
of
application
[46].
Moreover,
compared
to
other
synthesis
approaches,
electro-
chemical
anodization
is
a
simple,
convenient
and
“green”
technique
to
fabricate
uniform
layers
of
vertically
self-oriented
nanostruc-
tures,
which,
furthermore,
is
easy
to
be
scale
up
for
large-scale
industrial
productions.
2.
Experimental
2.1.
Ti
substrate
preparation
Several
1
×
2
cm2titanium
(Ti)
foils
(Sigma-Aldrich,
99.7%,
0.25
mm
thick)
were
mechanically
polished
with
320-grit
sandpa-
per
to
obtain
a
mirror
finish,
and
were
then
ultrasonically
washed
in
2-propanol.
After
rinsing
with
DI
water,
the
foils
were
degreased
in
a
40%
NaOH
solution
at
50 C
for
20
min,
and
were
then
rinsed
again
and
left
to
dry
in
air.
Shortly,
before
the
electrochemical
process,
the
titanium
was
etched
in
an
HF
(Carlo
Erba,
40%
w/w)
aqueous
solution
1.2%
w/w,
at
room
temperature
for
1
min,
in
order
to
obtain
a
fresh
metal
surface
for
NTs
growth.
Titania
nanotubes
were
synthesized
on
the
pretreated
Ti
foils,
by
dipping
1
×
1
cm2into
a
solution
of
ethylene
glycol,
NH4F
(0.5%
w/w)
and
DI
water
(2.5%
w/w).
The
Ti
was
used
as
an
anode,
while
Pt
foil
was
used
as
both
a
cathode
and
a
reference
electrode.
A
con-
stant
voltage
of
60
V
was
applied
to
the
cell
for
10
min
and,
the
TiO2-NTs
were
calcined
at
450 C
for
30
min
after
a
long
rinsing
in
water.
Further
details
on
the
NTs
growth
and
characterizations
can
be
found
elsewhere
[47].
2.2.
MnOxelectrodeposition
To
the
best
of
the
authors’
knowledge,
no
literature
reports
are
available
on
the
direct
electrodeposition
of
manganese
oxides
onto
Ti.
In
this
work,
a
base-case
electrode
was
anodically
deposited
by
inmersing
1
×
1
cm2of
the
Ti
foil
or
the
TiO2-NTs
onto
Ti,
in
an
unstirred
and
undivided
cell
containing
15
ml
of
a
0.1
M
Mn(CH3COO)2and
0.1
M
Na2SO4aqueous
solution,
by
applying
a
current
density
of
0.25
mA/cm2for
10
min
(i.e.
sample
0.25
MnOx).
The
electrodeposition
parameters,
such
as
current
density,
time
and
Mn-precursor
concentration
were
then
varied
to
investigate
their
effects
on
the
performances,
as
shown
in
Table
1.
The
sam-
ples
were
named:
ic
MnOx/Ti
and
ic
MnOx/TiO2-NTs,
where
i
is
the
electrodeposition
current
density
and
c
indicates
the
calcined
samples.
As
the
current
density
was
found
to
be
the
factor
with
most
influence
on
the
morphology
and
electrochemical
behaviour,
fur-
ther
syntheses
were
conducted
at
2.5
mA/cm2,
with
a
deposition
time
of
10
min
and
an
Mn(CH3COO)2concentration
of
0.1
M
(i.e.
sample
2.5
MnOx).
In
order
to
investigate
the
effects
of
different
electrodeposi-
tion
techniques,
the
type
of
precursor
was
also
changed,
once
the
most
suitable
current
density
and
deposition
time
had
been
tuned.
Therefore,
cathodic
depositions
of
the
manganese
oxide
over
Ti
and
TiO2-NTs
were
then
carried
out
by
dipping
1
×
1
cm2of
Ti
foil
into
an
unstirred
and
undivided
cell
containing
15
ml
of
a
0.01
M
KMnO4
and
0.1
M
Na2SO4aqueous
solution,
applying
a
current
density
of
2.5
mA/cm2(i.e.
sample
-2.5
MnOx).
Fig.
1
presents
a
schematic
view
of
all
the
syntheses
carried
out
in
this
work.
Some
of
the
electrodes
were
calcined
at
500 C
for
1
h
in
air
to
obtain
different
manganese
oxidation
states,
and
the
temperature
ramp
was
also
varied
to
investigate
its
effect
on
the
possible
passivation
of
the
Ti
substrate:
2C/min
(slow)
and
20 C/min
(fast).
Fig.
1.
3D
scheme
representing
the
manganese
oxide
electrodeposition
processes
over
the
Ti
and
TiO2NTs
substrates.
272
A.
Massa
et
al.
/
Applied
Catalysis
B:
Environmental
203
(2017)
270–281
Table
1
Manganese
oxides
electrodes,
synthesized
by
means
of
electrodeposition
over
Ti
and
TiO2-NTs.
Sample
name
ED
current
(mA/cm2)
ED
time
(min)
CMn (M)
Tcalc (C)
Ramp
(C/min)
Precursor
Substrate
0.25
MnOx/Ti
0.25
5–10
0.01–0.1
M
Mn(CH3COO)2Ti
0.25 MnOx/TiO2-NTs 0.25
10
0.1
M
Mn(CH3COO)2TiO2-NTs
0.25c
MnOx/Ti(s)
0.25
10
0.1
M
500 C
2
Mn(CH3COO)2Ti
0.25c
MnOx/Ti
0.25
5–10
0.01–0.1
M
500 C
20
Mn(CH3COO)2Ti
0.25c
MnOx/TiO2-NTs
0.25
10
0.1
M
500 C
20
Mn(CH3COO)2TiO2-NTs
2.5
MnOx/Ti
2.5
10
0.1
M
Mn(CH3COO)2Ti
2.5
MnOx/TiO2-NTs
2.5
10
0.1
M
Mn(CH3COO)2TiO2-NTs
2.5c
MnOx/Ti
2.5
10
0.1
M
500 C
20
Mn(CH3COO)2Ti
2.5c MnOx/TiO2-NTs 2.5
10
0.1
M 500 C20
Mn(CH3COO)2TiO2-NTs
-2.5c
MnOx/Ti
2.5
10
0.01
M
500 C
20
KMnO4Ti
-2.5c
MnOx/TiO2-NTs
2.5
10
0.01
M
500 C
20
KMnO4TiO2-NTs
A
BIOLOGIC
VMP-300
potentiostat
was
used
for
the
electrode-
positions.
Ti
was
set
as
the
anode,
Pt
wire
was
used
as
the
cathode
and
Ag/AgCl
3
M
KCl
(+
0.209
V
vs
NHE)
was
used
as
the
reference
electrode.
All
the
potentials
reported
in
this
work
should
be
intended
vs.
RHE
(ERHE in
V
vs.
RHE),
and
calculated
according
to
Nernst’s
equa-
tion
(Eq.(1)):
ERHE =
EAg/AgCl +
0.209V
+
0.059
·
pH
(1)
2.3.
Characterization
The
morphology,
structure
and
physic-chemical
parameters
of
the
electrodes
were
evaluated
by
means
of
X-Ray
Diffraction
(XRD,
X’Pert
PRO
diffractomer,
Cu
K
radiation
=
1.54
Å),
X-
Ray
Photoelectron
Spectroscopy
(XPS,
PHI5000
VersaProbe)
and
Field
Emission
Scanning
Electronic
Microscopy
(FESEM,
Zeiss
Mer-
lin).
The
semi-quantitative
surface
composition
was
estimated
by
means
of
Electron
Energy-Dispersive
X-ray
spectroscopy
(EDX,
Oxford
X-Act).
2.4.
Electrochemical
characterization
Cyclic
Voltammetries
(CV)
and
Linear
Sweep
Voltammetries
(LSV)
were
carried
out
in
an
unstirred
and
undivided
3-electrode
cell
system,
containing
15
ml
of
a
0.1
M
Na2SO4solution.
The
scan
limits
were
fixed
between
0.5
and
2
V
(vs.
Ag/AgCl)
and
the
scan
rates
were
20
mV/s
for
the
CV
and
5
mV/s
for
the
LSV,
respectively.
The
electrode
active
area
was
1
×
1
cm2;
a
Pt
wire
was
employed
as
the
counter
electrode
and
an
Ag/AgCl
3
M
KCl
(+
0.209
V
vs.
NHE)
was
used
as
the
reference
electrode.
2.5.
Electro-oxidation
tests
Electro-oxidation
tests
were
carried
out
in
an
unstirred
and
undivided
glass
cell
containing
15
ml
of
phenol
(C0=
100
mg/l)
and
0.1
M
Na2SO4as
the
supporting
electrolyte.
The
prepared
samples
(active
area
of
1
×
1
cm2)
were
used
as
the
anode,
while
Pt
wire
was
set
as
the
cathode
and
Ag/AgCl,
3
M
KCl
(+
0.209
V
vs
NHE)
as
the
reference
electrode.
Constant
currents
were
chosen
in
order
to
keep
the
working
potential
of
the
electrode
under
the
limit
of
the
instrument
(10
V)
over
a
reaction
time
of
5
h,
i.e.
0.25
mA/cm2for
electrodes
synthesized
at
low
current
densities
and
0.75
mA/cm2
for
samples
deposited
at
higher
current
densities.
The
solution
was
then
analyzed
by
High
Performance
Liquid
Chromatography
(Shi-
madzu
Prominence
HPLC)
with
a
Diode
Array
Detector
(DAD)
set
at
269
nm.
The
column
was
a
Rezex
ROA
(300
×
7.8
mm).
The
mobile
phase
was
5
mM
H2SO4and
the
flow
rate
was
0.5
ml/min.
Chemical
Oxygen
Demand
(COD)
analyses
were
carried
out
by
means
of
UV
spectroscopy,
using
a
HACH
LANGE
COD
cuvette
test
(LCI
400)
and
a
HACH
LANGE
DR5000
spectrophotometer.
Fig.
2.
XRD
patterns
of
MnOxelectrodeposited
at
0.25
mA/cm2,
10
min,
0.1
M
Mn2+:
a)
0.25
MnOx/Ti;
b)
0.25
MnOx/TiO2-NTs;
c)
0.25c
MnOx/Ti
(s);
d)
0.25c
MnOx/Ti;
e)
0.25c
MnOx/TiO2-NTs.
2.6.
Accelerated
lifetime
tests
In
order
to
asses
the
durability
of
the
synthesized
electrodes,
accelerated
lifetime
tests
were
carried
out
in
an
unstirred
and
undi-
vided
glass
cell
containing
15
ml
of
a
1
M
Na2SO4aqueous
solution.
Pt
wire
was
used
as
the
cathode
and
Ag/AgCl,
3
M
KCl
(+
0.209
V
vs
NHE)
as
the
reference
electrode.
A
current
density
of
100
mA/cm2
was
applied
to
the
cell.
The
electrode
was
considered
deactivated
when
the
measured
potential
reached
10
V.
3.
Results
and
discussion
3.1.
XRD
analysis
Fig.
2
shows
the
XRD
pattern
of
the
manganese
oxides
elec-
trodeposited
at
0.25
mA/cm2for
10
min
with
a
0.1
M
of
an
Mn2+
ion
concentration,
in
order
to
obtain
a
thicker
film
and
to
allow
easy
detection
of
the
MnOxpeaks.
As
expected,
non-calcined
0.25
MnOx/Ti
showed
no
peaks
that
could
be
attributed
to
any
crystalline
phase
of
MnOx(Fig.
2a),
thus
demonstrating
that
electrodeposition
alone
resulted
in
a
non-
crystalline
material;
only
Ti
peaks
were
visible
[40,48–50].
The
0.25
MnOx/TiO2-NTs
sample
(Fig.
2b)
showed
similar
results:
the
non-calcined
manganese
oxides
did
not
show
any
crystalline
phase,
while
TiO2peaks
were
clearly
visible
and
attributable
to
a
tetragonal
anatase
phase
(JCPDS
21-1272,
I41/amd,
a
=
b
=
0.379
nm,
c
=
0.951
nm),
as
expected,
due
to
the
calcination
of
TiO2-NTs
at
450 C.
Instead,
the
MnOxelectrodes
calcined
at
500 C
in
air
for
1
h,
for
both
the
slow
(Fig.
2c)
and
fast
ramps
(Fig.
2d),
showed
the
partic-
ular
peaks
of
a
cubic
bixbyite
crystalline
phase,
-Mn2O3(JCPDS
A.
Massa
et
al.
/
Applied
Catalysis
B:
Environmental
203
(2017)
270–281
273
Fig.
3.
XRD
patterns
of
MnOxelectrodeposited
at
2.5
mA/cm2,
10
min,
0.1
M
Mn2+
(or
0.01
M
Mn7+):
a)
2.5
MnOx/Ti;
b)
2.5
MnOx/TiO2-NTs;
c)
2.5c
MnOx/Ti;
d)
2.5c
MnOx/TiO2-NTs;
e)
2.5c
MnOx/Ti;
f)
2.5c
MnOx/TiO2-NTs.
41-1442,
Ia–3,
a
=
0.941
nm).
The
same
oxide
was
obtained
over
TiO2-NTs
(Fig.
2e).
Moreover,
it
can
be
noticed
that
two
peaks
appeared,
at
27.6
and
36.2,
on
the
MnOxcalcined
samples,
on
both
Ti
and
TiO2-NTs.
These
peaks
corresponded
to
the
formation
of
a
TiO2rutile
phase,
whose
transition
from
anatase
can
start
to
occur
at
about
500 C
in
air
[51–55].
This
observation
confirmed
the
hypothesis
of
the
formation
of
a
titanium
oxide
layer
between
the
Ti
substrate
and
MnOxcoating
during
the
heat
treatment.
The
XRD
patterns
of
the
manganese
oxides
electrodeposited
at
2.5
mA/cm2and
2.5
mA/cm2are
reported
in
Fig.
3.
Such
anodically
deposited
samples,
as
prepared
(i.e.
2.5
MnOx/Ti
and
2.5
MnOx/TiO2-NTs),
showed
only
Ti
(Fig.
3a)
and
TiO2anatase
(Fig.
3b)
XRD
patterns,
respectively,
thus
confirming
once
again
that
the
implemented
anodic
electrodeposition
led
to
a
non-crystalline
MnOxmaterial.
Instead,
after
annealing,
the
XRD
pattern
of
the
2.5c
MnOx/Ti
sample
(Fig.
3c)
showed
a
noticeable
difference,
com-
pared
to
the
electrode
deposited
at
0.25
mA/cm2.
In
this
case,
particular
peaks
of
two
MnOxcrystalline
phases
were
detected
in
the
spectrum:
a
cubic
bixbyite
-Mn2O3phase
and
a
tetrag-
onal
manganese
dioxide
-MnO2phase
(JCPDS
41-1442,
Ia–3,
a
=
0.941
nm;
JCPDS
044-014,
I4/m,
a
=
b
=
9.7847,
c
=
2.8630).
This
result
is
in
contrast
with
the
ones
obtained
for
the
same
sample
obtained
at
0.25
mA/cm2,
and
it
could
be
possibly
due
to
the
dif-
ferent
potentials
reached
during
electrodeposition,
because
of
the
higher
currents
provided
to
the
electrode.
In
fact,
a
change
in
either
the
potential
or
pH
at
a
constant
temperature
can
have
an
impor-
tant
effect
on
the
equilibrium
of
the
deposition,
and
can
thus
modify
the
structure
of
the
deposited
oxide
[56].
In
the
2.5c
MnOx/TiO2-NTs
sample
(Fig.
3d),
an
intense
peak
was
observed
at
32.9,
thus
confirming
the
presence
of
a
-Mn2O3
bixbyite
phase
in
the
catalyst
film,
while
-MnO2phase
peaks
were
absent.
The
XRD
pattern
of
the
tetragonal
-MnO2(JCPDS
044-014,
I4/m,
a
=
b
=
9.7847,
c
=
2.8630)
can
be
observed
for
the
cathodically
deposited
electrodes
(Fig
3e
and
f),
while
no
other
peaks,
apart
from
those
of
the
substrates,
were
detected.
In
general,
the
MnOxelectrodes
prepared
at
higher
electrodepo-
sition
current
densities
did
not
show
any
formation
of
TiO2rutile,
as
in
the
case
of
the
ones
prepared
at
lower
current
density
values.
This
could
be
due
to
the
higher
amount
of
MnOx,
which
covered
and
protected
the
Ti
substrate
and
TiO2-NTs
surface
from
further
oxidation.
In
order
to
investigate
the
surface
composition
of
the
elec-
trodeposited
MnOxand
to
support
the
XRD
analysis
outcomes,
XPS
analysis
was
carried
out
on
some
of
the
prepared
samples:
i.e.
2.5
MnOx/Ti,
2.5c
MnOx/Ti
and
-2.5c
MnOx/Ti.
The
results
of
the
Mn
3
s
XPS
spectra,
with
the
estimation
of
the
Average
Oxidation
State
(AOS)
[57]
of
the
surface
of
the
films
are
reported
in
the
Supplemen-
tary
information
(SI,
Fig.
S1
in
Supplementary
material).
It
is
easy
to
see
that
the
cathodic
sample
showed
the
highest
AOS
value
(i.e.
3.56),
which
was
close
to
the
Mn4+,
identified
by
the
XRD.
Instead,
the
AOS
value
for
the
2.5c
MnOx/Ti
was
3.11,
which
is
also
in
agree-
ment
with
the
presence
of
a
mix
of
Mn
oxides
with
oxidation
states
of
Mn3+ and
Mn4+,
as
revealed
by
the
XRD
measurement.
Finally,
the
non-calcined
MnOx,
whose
oxidation
state
could
not
be
identified
by
means
of
XRD,
gave
an
intermediate
AOS
value
(i.e.
3.25),
which
probably
means
that
also
a
mixed
oxide
Mn3+/Mn4+ could
have
been
formed
after
the
electrodeposition,
but
with
a
slightly
higher
presence
of
Mn4+ than
the
anodically
deposited
calcined
sample.
3.2.
FESEM
and
EDX
analyses
Fig.
4
reports
the
FESEM
images
of
the
manganese
oxides
elec-
trodeposited
at
0.25
mA/cm2.
The
FESEM
cross-section
images
of
the
TiO2nanotubes,
after
the
anodization
process,
are
shown
in
Fig.
4a.
The
NTs
were
5
m
long
and
vertically
aligned
with
respect
to
the
Ti
foil.
The
ordered
distribution
of
the
pores
can
be
appre-
ciated
in
the
top
FESEM
image
reported
as
an
inset
in
Fig.
4a:
the
inner
holes
had
an
average
dimension
of
about
70
nm
and
a
wall
thickness
of
around
20
nm.
After
the
MnOxelectrodeposition,
the
typical
formations
of
nanoflake
structures,
which
have
also
been
observed
in
other
works
[40,50,58],
were
formed
for
all
the
differ-
ent
types
of
electrodes.
The
thickness
of
the
manganese
oxide
layer
(
623
nm)
was
quite
uniform,
as
can
be
seen
in
the
inset
in
Fig.
4b.
A
slight
change
in
the
morphology
can
be
noticed
when
0.25
MnOx/Ti
(Fig.
4b)
is
compared
with
0.25c
MnOx/Ti,
which
was
calcined
with
the
fast
ramp
at
20 C/min
(Fig.
4e),
however
still
main-
taining
the
overall
nanoflake
disposition.
The
electrode
calcined
with
a
slow
ramp,
that
is,
0.25c
MnOx/Ti(s)
(Fig.
4d),
instead,
pre-
sented
an
almost
unchanged
nanoflakes
structure
with
respect
to
0.25
MnOx/Ti.
Both
samples
grown
on
NTs,
that
is,
the
not
calcined
0.25
MnOx/TiO2-NTs
(Fig.
4c),
and
the
calcined
0.25c
MnOx/TiO2-NTs
samples
(Fig.
4f),
showed
nanoflake
on
the
top
of
the
nanotube
layer.
The
deposition
of
nanoflakes
on
the
TiO2-NTs
substrate
was
less
homogeneous
than
the
deposition
on
metallic
titanium,
show-
ing
some
uncovered
areas,
probably
due
to
preferential
pathways
of
the
electrodeposition
currents,
where
the
electric
resistance
was
lower.
Fig.
5
shows
the
FESEM
images
of
the
manganese
oxide
electrodeposited
at
2.5
mA/cm2and
2.5
mA/cm2.
The
non
cal-
cined
anodically
deposited
samples,
i.e.
2.5
MnOx/Ti
(Fig.
5a)
and
2.5
MnOx/TiO2-NTs
(Fig.
5b),
still
showed
a
similar
nanoflake
struc-
ture
to
the
one
seen
for
the
sample
synthesized
at
lower
current
densities.
However,
the
nanoflakes
were
smaller
for
both
elec-
trodes,
due
to
the
higher
electrodeposition
current
density.
In
fact,
the
nucleation
rate
is
higher
and
the
critical
nucleation
radius
is
decreased,
as
the
potential
and
current
density
are
increased,
thus
the
formation
and
refinement
of
the
initial
grain
are
improved
[56].
The
inset
in
Fig.
5a
allows
the
thickness
of
the
MnOxfilm
on
the
2.5
MnOx/Ti
sample
(
3.7
m),
which
was
six
times
higher
than
the
one
on
the
samples
obtained
with
0.25
mA/cm2,
to
be
appre-
ciated.
The
calcined
2.5c
MnOx/Ti
(Fig.
5c)
and
2.5c
MnOx/TiO2-NTs
samples
(Fig.
5d),
instead,
showed
very
fine
nanoflakes,
that
coex-
isted
with
another
type
of
larger
crystalline
nanoparticles,
which
were
present
in
a
greater
amount
on
the
sample
grown
on
the
tita-
nium
substrate.
Such
a
difference
in
morphology
can
be
explained
by
considering
the
larger
amount
of
the
-MnO2phase
in
the
elec-
274
A.
Massa
et
al.
/
Applied
Catalysis
B:
Environmental
203
(2017)
270–281
Fig.
4.
FESEM
images
of
the
samples
electrodeposited
at
0.25
mA/cm2:
a)
TiO2-NTs;
b)
0.25
MnOx/Ti;
c)
0.25
MnOx/TiO2-NTs;
d)
0.25c
MnOx/Ti
(s);
e)
0.25c
MnOx/Ti;
f)
0.25c
MnOx/TiO2-NTs.
Fig.
5.
FESEM
images
of
the
samples
electrodeposited
at
2.5
mA/cm2:
a)
2.5
MnOx/Ti;
b)
2.5
MnOx/TiO2-NTs;
c)
2.5c
MnOx/Ti;
d)
2.5c
MnOx/TiO2-NTs;
e)
2.5c
MnOx/Ti;
f)
2.5c
MnOx/TiO2-NT.
trode
grown
on
Ti,
as
observed
in
the
XRD
analysis,
which
more
likely
composes
the
larger
nanoparticles.
Unlike
the
samples
electrodeposited
at
0.25
mA/cm2,
the
cross-
sectional
view
of
the
2.5
MnOx/TiO2-NTs
and
2.5c
MnOx/TiO2-NTs
samples
exhibited
evidence
of
the
penetration
of
the
manganese
inside
the
nanotubes
structure,
as
shown
in
the
inset
of
Fig.
5b
and
d,
similarly
to
the
penetration
of
the
electrodeposited
Sb-doped
SnO2films
into
TiO2-NTs
[32].
This
phenomenon
was
probably
due
to
the
high
electrodeposition
current
densities
which
led
to
smaller
nanoflakes
that
were
able
to
grow
inside
the
nanotube.
The
FESEM
images
of
the
MnOxsamples,
that
is,
-2.5c
MnOx/Ti
(Fig.
5e)
and
2.5c
MnOx/TiO2-NTs
(Fig.
5f),
prepared
by
cathodic
deposition,
pointed
out
a
totally
different
morphology
from
the
anodically
deposited
MnOx.
The
nanoflakes
formed
a
layer
directly
over
the
nanotubes,
from
which
polycrystalline
rod-like
structures
of
a
noticeable
thickness
(3.6
m)
grew.
In
this
case,
the
cross-
section
of
the
2.5c
MnOx/TiO2-NTs
sample,
shown
as
an
inset
in
Fig.
5f,
did
not
report
a
deep
penetration
of
manganese
oxide
inside
the
nanotubes.
This
could
be
due
to
either
a
different
wettability
of
the
nanotubes
by
the
KMnO4solution
from
the
Mn(CH3COO)2
used
in
the
anodic
deposition,
or
a
different
electric
field
induced
inside
and
on
the
top
of
the
nanotubes.
Tables
2
and
3
show
some
of
the
data
from
the
EDX
surface
composition
analyses
of
all
the
manganese
oxides
over
Ti
and
TiO2-NTs.
As
can
be
noticed,
for
all
the
samples
synthesized
at
0.25
mA/cm2,
the
atomic
ratio
(Mn/Ti)
is
comparable,
with
a
slightly
higher
manganese
content
for
the
non-calcined
electrode
grown
on
nanotubes
(i.e.
0.25
MnOx/TiO2-NTs).
For
samples
synthesized
at
2.5
or
2.5
mA/cm2,
the
atomic
ratio,
Mn/Ti,
is
up
to
10
or
even
100
times
higher
than
the
one
reported
for
the
electrodes
synthe-
sized
at
lower
current
densities,
as
a
greater
amount
of
manganese
oxide
was
deposited
on
these
samples.
Moreover,
the
Mn/Ti
val-
ues
measured
for
the
films
grown
on
Ti,
were
considerably
higher
than
the
ones
registered
for
those
grown
on
the
nanotubes,
and
this
could
be
due
to
the
higher
electric
resistance
of
the
TiO2-NTs
than
the
metallic
Ti,
which
limited
the
total
amount
of
manganese
oxide
deposited.
However,
because
of
the
higher
surface
area
of
the
TiO2-NTs,
it
is
more
likely
that
the
lower
Mn/Ti
ratios
observed
in
the
NTs
than
on
Ti
were
due
to
the
better
distribution
of
MnOxon
A.
Massa
et
al.
/
Applied
Catalysis
B:
Environmental
203
(2017)
270–281
275
Table
2
Electrooxidation
results
for
0.25
MnOxelectrodes.
El-Ox
current
density
0.25
mA/cm2.
Electrode
MnOxphaseaMn/Ti
ratiobFaradaic
efficiency
(%)
TON
(molPhenol/molMn )
Relative
conversion
(molPhenol/molMn ·
W
·
h)
0.25 MnOx/Ti Non-crystalline
0.12
7.8
0.25
37.2
0.25
MnOx/TiO2-NTs
Non-crystalline
0.16
33.2
2.36
469.0
0.25c
MnOx/Ti
-Mn2O30.11
10.7
0.86
212.6
0.25c
MnOx/TiO2-NTs
-Mn2O30.11
37.1
1.15
314.8
Ti
0.0
TiO2-NTs
22.4
aFrom
XRD
patterns.
bCalculated
from
EDX.
Table
3
Electrooxidation
results
for
2.5
MnOxelectrodes.
El-Ox
current
density
0.75
mA/cm2.
Electrode
MnOxphaseaMn/Ti
ratiobFaradaic
efficiency
(%)
TON
(molPhenol/molMn )
Relative
conversion
(molPhenol/molMn ·
W
·
h)
2.5
MnOx/Ti
Non-crystalline
15.2
43.3
0.51
17.1
2.5
MnOx/TiO2-NTs
Non-crystalline
2.11
42
0.48
20.1
2.5c
MnOx/Ti
-Mn2O3/
-MnO2
18.15
45.9
0.35
25.1
2.5c
MnOx/TiO2-NTs
-Mn2O31.59
26.7
0.26
25.7
-2.5c MnOx/Ti -MnO213.4
44.6
0.82
43.7
-2.5c
MnOx/TiO2-NTs
-MnO27.44
45.3
0.75
46.7
aFrom
XRD
patterns.
bCalculated
from
EDX.
Fig.
6.
LSV
in
0.1
M
Na2SO4of
the
samples
electrodeposited
at
0.25
mA/cm2.
the
top
of
the
TiO2-NTs,
as
well
as
inside
the
NTs
pores
(Fig.
5b
and
d).
3.3.
Electrochemical
behaviour
Figs.
6
and
7
show
the
LSV
of
all
the
manganese
oxide
samples
electrodeposited
at
0.25
mA/cm2and
2.5
mA/cm2,
respectively,
at
a
scan
rate
of
5
mV/s.
The
as-prepared
(non-calcined)
cathodic
sam-
ples
are
not
shown
because
they
demonstrated
very
poor
current
densities
and
became
detached
from
the
substrate
after
just
few
cycles.
In
both
graphs,
the
non-calcined
MnOx/Ti
(Figs.
6a,
7a)
and
all
the
samples
anodically
grown
on
the
NTs
(dashed
lines
in
Figs.
6b
and
e,
7b
and
d)
showed
pseudo-capacitive
behaviour,
which
is
characteristic
of
crystalline
MnO2,
or
non-crystalline
MnOxphases
[39]
and
which
could
also
be
due
to
titania
nanotubes
[39,41,42].
The
pseudo-capacitance
in
the
MnOxphase
was
caused
by
the
elec-
tron
transfer
at
the
Mn
surface
sites,
the
charge
transfer
being
balanced
by
either
the
chemisorption/desorption
of
the
electrolyte
cations
or
by
the
insertion/desinsertion
of
the
protons
[39,59].
In
the
case
of
TiO2,
due
to
its
semiconducting
properties,
electrons
Fig.
7.
LSV
in
0.1
M
Na2SO4of
the
samples
electrodeposited
at
2.5
and
2.5
mA/cm2.
are
accumulated
in
the
material
when
it
behave
like
a
cathode
(i.e.
presence
of
negative
current
values),
and
are
then
released
when
the
electrode
polarity
is
inverted
and
acts
as
an
anode
[39].
The
CVs
of
the
non-calcined
manganese
oxides,
for
the
depositions
at
both
0.25
mA/cm2and
2.5
mA/cm2(see
SI,
Figs.
S2–S7
in
Supplementary
material),
also
confirmed
the
high
capacitance
characteristic,
which
was
more
evidently
reported
for
the
electrodes
grown
on
the
nano-
tubes.
In
addition,
although
the
current
densities
were
similar,
the
manganese
oxides
grown
on
nanotubes
showed
a
slight
increment
in
stability
after
5
CV
cycles.
Instead,
the
LSVs
of
the
calcined
MnOx/Ti
samples
(Figs.
6c,
d
and
7c)
showed
a
diminuation
of
the
capacitive
properties
of
the
film,
due
to
the
change
in
the
crystalline
phase,
from
non-crystalline
or
-MnO2to
-Mn2O3[40],
or
because
of
the
variations
in
the
mor-
phology
(growth
in
the
particle
sizes)
of
the
deposited
film
after
the
thermal
treatment
[60].
On
the
other
hand,
the
calcined
MnOx/TiO2-
NTs
sample
containing
-Mn2O3,
still
showed
a
capacitive
trend,
probably
because
of
the
residual
capacitance
of
the
nanotubes
in
the
interlayer
[39].
276
A.
Massa
et
al.
/
Applied
Catalysis
B:
Environmental
203
(2017)
270–281
Fig.
8.
(A)
Electro-oxidation
curves
at
0.25
mA/cm2and
(B)
Nyquist
plots
of
the
EIS
measurements
at
3.3
V
vs.
RHE.
of
samples
synthesized
at
0.25
mA/cm2.
As
far
as
the
MnOxcathodically
deposited
on
both
Ti
and
TiO2-
NTs
is
concerned
(Fig.
7e
and
f),
although
the
prevalent
phase
formed
after
their
thermal
treatment
was
-MnO2,
they
showed
a
flat
LSV
and
low
current
densities
(<0.05
mA/cm2at
2.5
V
vs.
RHE)
towards
the
water
oxidation
reaction.
Such
phenomena
could
be
explained
by
considering
the
different
rod-like
morphologies
of
these
MnOxsamples,
which
have
less
exposed
surface
area
than
the
nanoflakes,
and
this
could
lead
to
a
lower
capacitive
effect.
It
has
been
reported
that
differences
in
MnO2morphology
under
differ-
ent
electrodeposition
conditions
and
the
post-thermal
treatments
could
contribute
to
the
differences
in
the
capacitive
behaviours
[61,62].
Indeed,
a
lower
capacitance
effect
has
been
observed
for
a
smaller
nanosheets
spacing
and
more
compactness
of
the
structure,
which
make
ion
diffusion
within
the
structure
difficult
[61].
More-
over,
the
interaction
with
the
substrate
seems
to
play
a
crucial
role
in
the
capacitive
behaviour
of
manganese
oxide
films,
and
also
leads
to
different
morphologies,
which
in
turn
affects
the
electrochemical
characteristics
of
the
electrode
[63,64].
In
general,
the
CVs
(see
SI,
Figs.
S2
to
S7
in
Supplementary
material)
showed
an
improved
stability
for
the
calcined
MnOxsam-
ples,
as
well
as
for
all
the
samples
deposited
on
TiO2-NTs,
possibly
because
the
passivating
TiO2layer
that
usually
formed
on
the
Ti
surface
under
oxidative
conditions
[33,36],
was
prevented
from
forming
by
the
presence
of
either
well
crystallized
manganese
oxides
structures
(formed
after
calcination)
or
crystalline
TiO2-NTs.
The
LSVs
of
the
as-prepared
and
calcined
electrodes
synthe-
sized
at
0.25
mA/cm2with
different
electrodeposition
times
(10
vs.
5
min)
and
precursor
concentrations
(0.1
vs.
0.01
M
Mn2+)
are
reported
in
the
SI
(Figs.
S8
and
S9
in
Supplementary
materail,
respectively).
As
can
be
noticed,
both
the
shape
of
the
curves
and
the
current
density
values
were
similar,
and
it
is
therefore
possible
to
state
that
these
two
parameters
did
not
affect
the
electrochem-
ical
behaviour
of
the
samples
to
any
great
extent.
For
this
reason,
10
min
and
0.1
M
Mn2+ were
selected
as
the
optimum
conditions
for
the
electrodepositions
at
higher
current
densities
(i.e.
2.5
mA/cm2).
The
2.5c
MnOx/TiO2-NTs
sample
produced
the
highest
final
cur-
rent
density
(0.4
mA/cm2at
2.5
V
vs.
RHE,
see
Fig.
7d),
which
was
one
order
of
magnitude
higher
than
the
similar
MnOxfilm
deposited
on
Ti
(2.5c
MnOx/Ti,),
and
than
the
cathodically
deposited
films
(-2.5c
MnOx/Ti
and
2.5c
MnOx/TiO2-NTs).
Moreover,
it
was
two
times
higher
than
the
current
densities
obtained
for
the
respec-
tive
non
calcined
samples,
on
both
Ti
and
TiO2-NTs.
This
trend
was
in
agreement
with
the
predominance
of
the
-Mn2O3phase
revealed
by
the
XRD
analysis
on
the
calcined
electrodes
containing
TiO2-NTs,
which
is
capable
of
higher
oxygen
evolution
rates
than
-MnO2[23,30,50],
thus
justifiyng
the
improved
behaviour
for
the
water
oxidation
reaction.
On
the
contrary,
the
presence
of
-MnO2
detected
on
the
sample
grown
on
Ti,
as
well
as
in
the
cathodically
deposited
MnOx,
led
to
a
decrease
in
the
water
oxidation
activity
[65,66],
which
instead
should
be
beneficial
for
the
degradation
of
organic
molecules
(e.g.
phenol)
with
a
higher
Nernst
potential
than
water
oxidation.
3.4.
Phenol
electro-oxidation
and
EIS
analysis
Figs.
8A
and
9A
show
the
chrono-potentiometric
curves
obtained
during
the
phenol
electro-oxidation
(El-Ox)
reaction
with
all
the
electrodes
synthesized
at
0.25
mA/cm2and
2.5
or
2.5
mA/cm2,
respectively.
Electro-oxidation
was
performed
for
a
total
time
of
5
h,
fixing
a
current
density
of
0.25
mA/cm2for
the
0.25
MnOx
samples,
while
the
2.5
MnOx
electrodes
were
tested
at
0.75
mA/cm2,
in
order
to
start
the
El-Ox
at
about
3
±
0.5
VRHE and
to
keep
the
final
working
potential
below
10
V.
In
addition,
electro-
chemical
impedance
spectroscopy
(EIS),
a
well-known
technique
that
is
often
employed
to
characterize
electrochemical
systems
with
the
aim
of
comparing
their
charge
transfer
and
transport
properties
[37,39,67],
was
also
performed
on
the
phenol
solution.
An
applied
DC
potential
of
3.3
VRHE,
which
was
the
average
initial
potential
achieved
during
the
El-Ox
process
for
all
the
samples,
was
employed
for
the
EIS
analyses.
The
results
are
reported
in
Figs.
8B
and
9B
(Nyquist
plots)
and
in
Fig.s
S10
and
S11
(Bode
plots:
phase
and
modulus
of
impedance
|Z|
vs.
frequency).
The
most
stable
electrodes
that
were
able
to
sustain
the
con-
stant
potential
(i.e.
for
which
the
total
increase
of
potential
over
time
was
lower
than
0.5
V)
were
found
to
be
the
calcined
anod-
ically
deposited
MnOxelectrodes
over
both
the
Ti
and
TiO2-NTs
subtrates
(see
Figs
8A-c,d,e
and
9A-c,d),
which
can
be
attributed
to
the
presence
of
highly
stable
-Mn2O3crystalline
phase.
Accord-
ingly,
the
Nyquist
plots
of
the
same
samples
in
Figs.
8B
and
9B
show
that
they
evidenced
lower
impedance
values
for
these
elec-
trodes
than
for
the
other
electrodes,
which
in
turn
indicated
an
increase
in
the
number
of
electrons
transferred
through
the
elec-
trode/electrolyte
interface
on
those
materials.
In
fact,
a
reduction
in
the
semicircles
diameter
in
the
Nyquist
plots
indicated
a
lower
resistance
to
the
charge
transfer
because
of
an
enhancement
of
the
reaction
kinetics,
which,
in
this
case,
depends
on
the
surface
properties
of
the
MnOx-based
electrode
materials
[40].
Neverthe-
less
it
is
not
possible
to
identify
which
is
the
prevalent
reaction
between
water
oxidation
and
phenol
degradation
from
these
elec-
trochemical
measurements,