Capacitive manganese oxide thin films deposited by reactive direct current sputtering

Article abstract of DOI:10.1149/1.3414823

Manganese oxide thin films are deposited on graphite foils by a simpler one-step reactive dc sputtering dry process with different volume flow rates of oxygen, sputtering time, pressure, and power. Maximum mass specific capacitance of 344 F/g is obtained in 0.5 M LiCl as well as with optimum sputtering conditions (volume flow rate of oxygen=10 sccm, sputtering pressure=20 mTorr, sputtering power=60 W, and sputtering time=60 min), and this demonstrates its good mass specific capacitance at a higher sweep rate of 100 mV/s. Furthermore, the geometric specific capacitance increases with increasing sputtering pressure, power, and time. Moreover, the electrochemical stability of the electrode decreases with increasing sputtering pressure.

Full text of DOI:10.1149/1.3414823

Journal of The Electrochemical Society, 157 ͑7͒ A753-A759 ͑2010͒
A753
0013-4651/2010/157͑7͒/A753/7/$28.00 © The Electrochemical Society
Capacitive Manganese Oxide Thin Films Deposited by Reactive
Direct Current Sputtering
,z
*
Chuen-Chang Lin and Peng-Yu Lin
Department of Chemical and Materials Engineering, National Yunlin University of Science and
Technology, Yunlin 64002, Taiwan
Manganese oxide thin films are deposited on graphite foils by a simpler one-step reactive dc sputtering dry process with different
volume flow rates of oxygen, sputtering time, pressure, and power. Maximum mass specific capacitance of 344 F/g is obtained in
0.5 M LiCl as well as with optimum sputtering conditions ͑volume flow rate of oxygen = 10 sccm, sputtering pressure
= 20 mTorr, sputtering power = 60 W, and sputtering time = 60 min͒, and this demonstrates its good mass specific capacitance
at a higher sweep rate of 100 mV/s. Furthermore, the geometric specific capacitance increases with increasing sputtering pressure,
power, and time. Moreover, the electrochemical stability of the electrode decreases with increasing sputtering pressure.
© 2010 The Electrochemical Society. ͓DOI: 10.1149/1.3414823͔ All rights reserved.
Manuscript submitted January 19, 2010; revised manuscript received February 26, 2010. Published May 5, 2010.
Electrochemical capacitors are charge-storage devices which
possess a higher power density and a longer cycle life than
batteries.1-3 Their applications include stop/go power sources for
hybrid vehicles, starting power for fuel cells, and burst-power gen-
eration in electronic devices, etc.4-9 Electrochemical capacitors are
classified into two types, electric double-layer capacitors ͑EDLC͒
and pseudocapacitors according to their energy-storage mechanisms.
The capacitance of an EDLC arises from the separation of charge at
the interface between the electrolyte and the electrode with a high
surface area inert material such as activated carbon.9,10 By contrast,
pseudocapacitance arises from fast and reversible redox reactions of
electroactive materials with several oxidation states.1,9-15
Ruthenium is not suitable for a wider commercialization due to
being relatively expensive. Hence, the more economical manganese
oxide with a variety of stable valence states appears to be a substi-
tute electrode material for electrochemical capacitors. Therefore,
manganese oxides were prepared by different wet processes for their
temperature and bias voltage͒, etc., have an effect on adhesion,22-27
which influences operational stability of electrochemical capacitors.
Enhanced adhesion was observed at a lower sputtering pressure and
for thicker films with rough surfaces.22 Sputter cleaning seemed to
be one of the most effective methods in the physical vapor deposi-
tion process to improve adhesion and sputter cleaning with argon
showed the best adhesion after 30 min of argon bombardment by
simple mechanical test,23 thus, sputter cleaning with argon for 30
min was applied for the substrate in this paper. In earlier work,
ruthenium oxide films were deposited on Si and Ti substrates by
reactive radio-frequency magnetron sputtering and the total anodic
voltammetric charges increased with an increasing roughness
factor.28 Because the highly rough surface should possess the high
surface area, the mass specific capacitance also increased with in-
creasing the roughness factor for manganese oxide.29 Furthermore,
the sputtering yield is reduced so that the deposition rate is reduced
also when the oxygen concentration in the sputtering gas ͑argon and
oxygen͒ is increased, because metal–oxygen reaction occurs on the
target surface.30 Thus, the oxygen concentration changes the surface
roughness, crystallization, and oxidation state, which influence spe-
cific capacitance and operational stability of electrochemical capaci-
tors. Moreover, sputtering also offers the opportunity to control
structural ͑such as surface roughness and crystallization͒ and stoi-
chiometric ͑oxidation state͒ development of mixing metal–oxygen
materials.
application
as
electrode
materials
in
electrochemical
capacitors.10,16-18 For example, colloidal manganese oxide was pre-
pared by reducing KMnO4 with MnCl2. The sol–gel-derived nano-
particulate MnO2 thin film then formed on nickel foils by dip-
coating exhibits a high mass specific capacitance of 698 F/g and
long cycle life. However, it is a two-step process and the discharge
time achieved is only about 1.1 s at a constant current of 0.1 mA due
to the MnO2 thin film.18 However, some residual impurities could be
carried over from these wet processes, and these might affect elec-
trochemical properties ͑such as conductance͒ of electrodes. There-
fore, a simpler one-step reactive dc sputtering dry process was ap-
plied to deposit manganese oxide films on graphite foils in this
research. Mn films were deposited by sputtering and then manga-
nese oxide films were synthesized by anodic oxidation; the mass
specific capacitance of the films yielded 400–450 F/g at a potential
scan rate of 5 mV/s.9 Increasing the thickness of the dense Mn layer
fabricated in a sputter deposition system results in increasing the
thickness of the porous Mn oxide surface layer produced during the
electrochemical oxidation, but with a constant amount of hydration
and little change in the mass specific capacitance.19 The electro-
chemical oxidation of the Mn layer fabricated in a sputter deposition
system occurs through the following sequence of hydrated oxides:
Mn to MnO to Mn3O4 to Mn2O3 to MnO2.20 In earlier work, a
ruthenium oxide thin film was deposited on a Pt/Ti/Si substrate by a
specially designed dc reactive sputtering method: The capacitance
per volume of the thin-film supercapacitor at the first cycle was
38 mF/cm2 ␮m.21
This paper examines the effects of sputtering conditions ͑pres-
sure, power, time, and volume flow rates of oxygen͒ on specific
capacitance and sputtering pressure as well as time on the opera-
tional stability of electrochemical capacitors.
Experimental
A graphite foil was used as a substrate due to its lower specific
resistance ͑lower than that of activated carbon and silicon͒.31 In
mechanical polishing, the graphite foil ͑1 ϫ 1 or 1 ϫ 2 cm2͒ was
abraded with SiC paper and then rinsed ultrasonically with deion-
ized water for 10 min. To increase its surface roughness, for the
chemical acid wash, it was then etched in 6 M aqueous HCl at room
temperature for 30 min and subsequently rinsed ultrasonically with
deionized water for 10 min. Next, it was degreased ultrasonically in
acetone until any surface grease was completely eliminated, rinsed
with pure deionized water, and subsequently oven-dried in air
͑50°C͒ to constant weight. Finally, sputter cleaning with argon for
30 min was applied for the substrate and then it was reweighed for
sputtering.
The manganese oxide film was deposited on the pretreated and
grounded graphite foil substrate by reactive dc magnetron sputtering
from a 3 in. disk Mn ͑purity: 99.9%, melting point: about 535°C,
purchased from SCM, INC͒ target in a vacuum chamber with a
background pressure of 7 ϫ 10−6 Torr. The distance between the
target and the substrate was 8 cm. During deposition, the substrate
Contamination, the roughness of the substrate, film thickness,
some sputtering parameters ͑such as sputtering pressure, substrate
*
Electrochemical Society Active Member.
z E-mail: linchuen@yuntech.edu.tw
Downloaded on 2015-03-04 to IP 169.230.243.252 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

A754
Journal of The Electrochemical Society, 157 ͑7͒ A753-A759 ͑2010͒
(a)
(b)
100
60
100
60
40 W
60 W
80 W
100 W
10mTorr
20mTorr
30mTorr
20
20
-20
-20
-60
-100
Figure 1. CVs ͑the 100th charge–
discharge cycle͒ at a potential scan rate of
100 mV/s for the manganese oxide elec-
trode prepared at ͑a͒ different pressure
͑with optimum sputtering conditions: 60
min, 10 sccm oxygen, and 60 W͒, ͑b͒ dif-
ferent sputtering power ͑with optimum
sputtering conditions: 60 min, 10 sccm
oxygen, and 20 mTorr͒, ͑c͒ different vol-
ume flow rates of oxygen ͑with optimum
sputtering conditions: 60 min, 20 mTorr,
and 60 W͒, and ͑d͒ different sputtering
time ͑with optimum sputtering conditions:
10 sccm oxygen, 20 mTorr, and 60 W͒.
-60
-100
-0.1 0.1 0.3 0.5 0.7 0.9 1.1
-0.1 0.1 0.3 0.5 0.7 0.9 1.1
E(Vvs. Ag/AgCl)
E(Vvs.Ag/AgCl)
(c)
(d)
100
100
60
60
20
5sccm
30 min
60 min
90 min
20
10sccm
15sccm
-20
-60
-100
-20
-60
-100
-0.1 0.1 0.3 0.5 0.7 0.9 1.1
-0.1 0.1 0.3 0.5 0.7 0.9 1.1
E( Vvs. Ag/AgCl)
E(Vvs.Ag/AgCl)
Results and Discussion
͑about 55°C͒ was not intentionally heated. The volume flow rate of
argon was maintained at 25 sccm. The volume flow rates ͑5, 10, and
15 sccm͒ of oxygen were varied, sputtering pressure ͑10, 20, and 30
mTorr͒ was varied, sputtering power ͑40, 60, 80, and 100 W͒ was
varied, and sputtering time ͑30, 60, and 90 min͒ was also varied.
Finally, the graphite foil with manganese oxide thin films was
weighed to five decimals ͑0.00001 g͒ and thus the mass of the film
could be determined with the small percentage errors.
The electrochemical measurements for the prepared graphite foil
electrodes were performed by an electrochemical analyzer ͑CH In-
struments CHI 608B, U.S.͒. The three-electrode cell consisted of
Ag/AgCl as the reference electrode, Pt as the counter electrode and
the prepared electrode as the working electrode. The electrolytes
were degassed with purified nitrogen gas before voltammetric mea-
surements, and nitrogen was passed over the solution during all the
measurements. The solution temperature was maintained at 25°C by
means of a circulating water thermostat ͑HAAKE DC3 and K20,
Germany͒. The cyclic voltammogram ͑CV͒ was taken with the po-
tential scan rate ͑100 mV/s͒ and a 0.5 M aqueous electrolyte ͑LiCl,
pH = 6.7͒. The potential window in the range of 0–1 V was used in
all measurements except where stated. Capacitance is normalized to
1 g of manganese oxide.
Surface morphology of the electrodes prepared at different con-
ditions was conducted by a field-emission-scanning electron micro-
scope ͑JEOL JSM-6700F, Japan͒. In addition, the chemical environ-
ment of the manganese oxide film deposited on the graphite foil
with the different volume flow rates of oxygen and the electrode
͑using optimum sputtering conditions: 10 sccm oxygen, 60 min, 20
mTorr, and 60 W͒ before as well as after the potential cycling was
explored by an X-ray photoemission spectroscope ͑XPS, Fison VG.
ESCA210, England͒. Furthermore, additional information on the
surface roughness of the electrodes prepared at different conditions
was obtained by atomic force microscope ͑AFM, Digital Instrument
NanoMan NS4 + D3100, U.S.͒. Moreover, X-ray diffraction ͑XRD,
MAC SCIENCE, Japan͒ with a low angle of incidence was used to
characterize the crystalline structure of the graphite foil and the
manganese oxide film deposited on the graphite foil using optimum
sputtering conditions.
CVs at a potential scan rate of 100 mV/s for the manganese
oxide electrode prepared at ͑a͒ different pressure, ͑b͒ different
sputtering power, ͑c͒ different volume flow rates of oxygen, and
͑d͒ different sputtering time are shown in Fig. 1. All CVs in
Fig. 1 for the manganese oxide electrode prepared at different
sputtering conditions are similar ͑no evidently observable redox
peaks͒ and show a mirror image with respect to the zero-current line
and a rapid current response on voltage reversal at potentials near
the two limits of the potential window. These features indicate that
hydrated cations ͑such as Li+ and proton͒ of the electrolyte ͑pH
= 6.7͒ have a rapid chemisorption/desorption reaction rate and are
involved in the charge-storage process within the very near surface
thin layer of manganese oxide electrodes.32,33 Electrochemical
quartz-crystal microbalance as well as X-ray photoelectron spectros-
copy data further indicate that proton plays the predominant role and
the charge–discharge reaction in the LiCl solution can be expressed
as:
Mn͑IV͒O2 + 0.767␭H+ + 0.233␭Li+ + ␭e− ⇔ H+0.767␭Li+0.233␭
͓Mn͑III͒ Mn͑IV͒ ͔O2.34 In addition, the capacitance of the pure
␭
1−␭
graphite foil ͑1 ϫ 2 cm2͒ is 0.00216 F, which is about 6.27% of the
total capacitance of the manganese oxide electrode using optimum
sputtering conditions. Therefore, if there is some contribution to the
total capacitance of the manganese oxide electrode from the graphite
foil which manganese oxide thin films were sputtered on, it seems
unimportant.
Figure 2 shows the effects of the sputtering pressure on the
mass specific capacitance. The mass specific capacitance reached a
maximum at 20 mTorr of sputtering pressure. The mass specific
capacitance increased at the range from 10 to 20 mTorr of sputter
ing pressure. This picture may be explained as follows. A higher
sputtering pressure leads to higher collision frequencies ͓Z
= ͑␲d2CrelP͒/͑kT͒, where Z is the collision frequency, d is the av-
erage diameter of particles, Crel is the relative mean speed, P is
pressure, k is Boltzmann’s constant, and T is temperature͔ and a
lower kinetic energy. A lower kinetic energy leads to the lower mo-
bility on the surface of the substrate for particles ͑Mn atoms or MnO
“clusters”͒. This would cause a higher surface roughness ͑see Fig. 3,
Downloaded on 2015-03-04 to IP 169.230.243.252 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

Journal of The Electrochemical Society, 157 ͑7͒ A753-A759 ͑2010͒
A755
ꢅꢀꢀ
ꢄꢁꢀ
ꢄꢀꢀ
ꢃꢁꢀ
ꢃꢀꢀ
ꢂꢁꢀ
ꢂꢀꢀ
ꢁꢀ
ꢀꢁꢂꢃ ꢄꢅꢆ ꢇ ꢈ ꢉꢅ
ꢀ
ꢀ
ꢂꢀ
ꢃꢀ
ꢄꢀ
ꢅꢀ
ꢆꢇ ꢈꢈꢉꢊꢋꢌꢍ ꢇꢊꢉꢎꢎ ꢊꢉ ꢏꢐꢑꢒꢊꢊꢓ
Figure 2. ͑Color online͒ The effects of sputtering pressure on the mass
specific capacitance ͑with optimum sputtering conditions: 60 min, 10 sccm
oxygen, and 60 W͒.
ꢀꢊꢂꢃ ꢄꢅꢆ ꢇ ꢋꢌꢍ ꢉꢅ
i.e., there is an increase in the height of the surviving particles in a
particle-migration coalescence process͒, thus leading to a higher
mass specific capacitance ͑a similar result has been published in
previously28,29͒. However, mass specific capacitance decreased at a
sputtering pressure in the range 20–30 mTorr. This picture may be
attributed to a too high sputtering pressure, leading to too many
particles, and thus causing a shadowing effect which enhances many
collision probability, and so declining much kinetic energy/mobility
on the surface of the substrate for particles. This leads to a decreas-
ing surface roughness ͑see Fig. 3͒ and mass specific capacitance.
Figure 4 shows the effects of sputtering power on the mass spe-
cific capacitance. The mass specific capacitance reached a maximum
at 60 W. The mass specific capacitance increased at sputtering power
in the range 40–60 W. This illustrates that the higher the sputtering
power, the higher the sputtering yield, thus enhancing the deposition
rate. Then particles falling on the substrate do not have enough time
to rearrange and resputter themselves, which leads to increasing
surface roughness ͑see Fig. 5͒, and so leading to a higher mass
specific capacitance. However, the mass specific capacitance de-
creased with sputtering power in the range 60–100 W. This picture
may be attributed to a too high sputtering power, leading to a too
strong ion as well as reflected neutral bombardment of the substrate
surface, then only causing more damages, thus decreasing surface
roughness ͑see Fig. 5͒ and mass specific capacitance.
ꢀꢎꢂꢃ ꢄꢅꢆ ꢇ ꢏꢌ ꢉꢅ
Figure 6 shows the effects of different volume flow rates of oxy-
gen on the mass specific capacitance. The mass specific capacitance
reached the maximum at 10 sccm oxygen. The mass specific capaci-
tance increased in the range 5–10 sccm of oxygen. The reason be-
hind this behavior may be that at higher volume flow rates of oxy-
gen, the more oxygen ions would lead to a greater oxygen ion
bombardment of the substrate surface due to a sheath voltage ͑the
difference between the plasma potential and the substrate potential͒,
enhancing surface roughness ͑see Fig. 7͒, and finally increasing the
mass specific capacitance. However, the mass specific capacitance
decreased in the range 10–15 sccm of oxygen. This may indicate
that with a too high volume flow rate of oxygen, too many oxygen
ions would cause a shadowing effect which prevented argon ions
from bombarding the target and even lead to lighter oxygen ions ͑as
opposed to heavier argon ions͒ bombarding the target because of the
sheath voltage as well as oxide formation on the target. Thus the
sputtering yield is reduced so that the deposition rate drops. Conse-
quently, particles falling on the substrate have sufficient time to
rearrange and resputter themselves, which leads to decreasing sur-
face roughness ͑see Fig. 7͒ and mass specific capacitance. To inves-
tigate the chemical environment of manganese oxide on the graphite
foil electrode, XPS was used to examine the binding energy of man-
ganese oxide thin films sputtered with optimum sputtering pressure
Figure 3. ͑Color online͒ AFM micrograph ͓͑a͒ 10, ͑b͒ 20, and ͑c͒ 30 mTorr͔
of the electrode prepared at different pressure ͑with optimum sputtering con-
ditions: 60 min, 10 sccm oxygen, and 60 W͒.
͑20 mTorr͒, power ͑60 W͒, as well as time ͑60 min͒ and different
volume flow rates of oxygen. XPS spectral data of Mn 2p3/2 and
Downloaded on 2015-03-04 to IP 169.230.243.252 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

A756
Journal of The Electrochemical Society, 157 ͑7͒ A753-A759 ͑2010͒
ꢅꢀꢀ
ꢄꢁꢀ
ꢄꢀꢀ
ꢃꢁꢀ
ꢃꢀꢀ
ꢂꢁꢀ
ꢂꢀꢀ
ꢁꢀ
ꢀꢁꢂꢃ ꢄꢅꢆ ꢇ ꢈꢉ ꢊꢅ
ꢀ
ꢃꢀ
ꢅꢀ
ꢆꢀ
ꢇꢀ
ꢂꢀꢀ
ꢂꢃꢀ
ꢈꢉꢊꢊꢋꢌꢍꢎꢏ ꢈꢐꢑꢋꢌ ꢒꢓꢔ
Figure 4. ͑Color online͒ The effects of sputtering power on the mass specific
capacitance ͑with optimum sputtering conditions: 60 min, 10 sccm oxygen,
and 20 mTorr͒.
ꢀꢋꢂꢃ ꢄꢅꢆ ꢇ ꢈꢌꢍ ꢊꢅ
ꢀꢁꢂꢃ ꢄꢅꢆ ꢇ ꢈꢉ ꢊꢅ
ꢀꢋꢂꢃ ꢄꢅꢆ ꢇ ꢌꢉꢍ ꢊꢅ
ꢀꢎꢂꢃ ꢄꢅꢆ ꢇ ꢏꢐ ꢊꢅ
ꢀꢑꢂꢃ ꢄꢅꢆ ꢇ ꢐꢒꢈ ꢊꢅ
ꢀꢎꢂꢃ ꢄꢅꢆ ꢇ ꢍꢌ ꢊꢅ
Figure 5. ͑Color online͒ AFM micrograph ͓͑a͒ 40, ͑b͒ 60, ͑c͒ 80, and ͑d͒ 100
W͔ of the electrode prepared at different sputtering power ͑with optimum
sputtering conditions: 60 min, 10 sccm oxygen, and 20 mTorr͒.
ꢅꢀꢀ
ꢄꢁꢀ
ꢄꢀꢀ
ꢃꢁꢀ
ꢃꢀꢀ
ꢂꢁꢀ
ꢂꢀꢀ
ꢁꢀ
ꢀ
ꢀ
ꢁ
ꢂꢀ
ꢂꢁ
ꢃꢀ
ꢆꢇ ꢈꢉꢊ ꢋ ꢇꢌ ꢍꢎꢏꢊ ꢇꢋ ꢇꢐꢑꢒꢊꢓ ꢔꢕꢖꢖꢉꢗ
Figure 7. ͑Color online͒ AFM micrograph ͓͑a͒ 5, ͑b͒ 10, and ͑c͒ 15 sccm͔ of
the electrode prepared at different volume flow rates of oxygen ͑with opti-
mum sputtering conditions: 60 min, 20 mTorr, and 60 W͒.
Figure 6. ͑Color online͒ The effects of different volume flow rates of oxy-
gen on the mass specific capacitance ͑with optimum sputtering conditions:
60 min, 20 mTorr, and 60 W͒.
Downloaded on 2015-03-04 to IP 169.230.243.252 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

Journal of The Electrochemical Society, 157 ͑7͒ A753-A759 ͑2010͒
A757
ꢄꢁꢁ
ꢃꢂꢁ
ꢃꢁꢁ
ꢀꢂꢁ
ꢀꢁꢁ
ꢁ
ꢃꢁ
ꢅꢁ
ꢆꢁ
ꢇꢀꢁ
ꢈꢉꢊꢊꢋꢌꢍꢎꢏ ꢊꢍꢐꢋ ꢑꢐꢍꢎꢒ
Figure 9. ͑Color online͒ The effects of sputtering time on the mass specific
capacitance ͑with optimum sputtering conditions: 10 sccm oxygen, 20
mTorr, and 60 W͒.
the amount of trivalent manganese oxide as well as the higher the
surface roughness, and the higher the mass specific capacitance ͑see
Table I and Fig. 6 and 7͒. A similar result has been published
previously.3,28,29 However, at volume flow rates of oxygen in the
range 10–15 sccm, the amounts of trivalent manganese oxide were
almost the same due to the excess oxygen, and the mass specific
capacitance decreased because of decreasing surface roughness ͑see
Table I and Fig. 6 and 7͒.
Figure 9 shows the effects of sputtering time on the mass specific
capacitance. The mass specific capacitance reached a maximum at
60 min. The mass specific capacitance increased with sputtering
time in the range 30–60 min. This picture may be that a higher
sputtering time led to a higher surface roughness ͑see Fig. 10͒, thus
leading to a higher mass specific capacitance. However, the mass
specific capacitance decreased for sputtering times in the range
60–90 min. This may indicate that higher sputtering times lead to a
lower surface roughness ͑see Fig. 10͒, and thus to a lower mass
specific capacitance. Furthermore, Fig. 11 shows the effects of sput-
tering time and different charge–discharge cycles on the mass spe-
cific capacitance of the electrode ͑fabricated by optimum sputtering
conditions: 10 sccm oxygen, 20 mTorr, and 60 W͒. The mass spe-
cific capacitance of the electrode increased gradually with increasing
number of charge–discharge cycles during the first 600 cycles of
potential cycling. The reason behind this behavior may be that at
more potential cycles, the percentage of incorporated water may
increase from the electrolyte ͑the 0th cycle of potential cycles: 19.20
atom % incorporated water, and the 100th cycle of potential cycles:
26.20 atom % incorporated water͒. This causes a higher mass spe-
cific capacitance because the combined water ͑incorporated water
and free water from the electrolyte͒ promotes the movement of the
protons in the solid phase,18 and thus electrolyte ions more easily
diffuse into the inner pores of the electrodes and consequently some
potentially electroactive sites on the inner pores are able to contrib-
ute to the mass specific capacitance with higher levels of combined
water.
Figure 8. ͑Color online͒ XPS spectral data of ͑a͒ Mn 2p3/2 and ͑b͒ Mn 3s
for manganese oxide sputtered with optimum dc sputtering pressure ͑20
mTorr͒, power ͑60 W͒, as well as time ͑60 min͒ and different volume flow
rates of oxygen.
Mn 3s ͑see Fig. 8͒ for manganese oxide thin films sputtered with
different volume flow rates of oxygen are listed in Table I. For
Mn 2p3/2, binding energy for manganese oxide increased from
641.96 to 642.06 and 642.30 eV following 10 and 15 sccm oxygen,
respectively. These results show that the oxidation state of manga-
nese oxide increased after increasing the volume flow rate of oxygen
from 5 to 10 or 15 sccm oxygen. The oxidation state of manganese
oxide was also determined by the differences in binding energies
between the splitting peaks of Mn 3s in Table I. The difference in
the binding energy for manganese oxide changed from 5.27 to 4.79
eV and 4.83 eV following 10 and 15 sccm oxygen, respectively.
Thus, manganese oxide thin films sputtered with 5 sccm oxygen
mainly consist of Mn3+ species and manganese oxide thin films
sputtered with 10 or 15 sccm oxygen almost consisted of Mn4+
species.33,35 Therefore, at volume flow rates of oxygen in the range
5–10 sccm, the higher the volume flow rates of oxygen, the larger
Table I. XPS Mn 2p3Õ2 and Mn 3s binding energies (eV) and
differences of binding energies (eV) for manganese oxide sput-
tered with optimum dc sputtering pressure (20 mTorr), power (60
W), as well as time (60 min) and different volume flow rates of
oxygen.
XRD patterns of the freshly polished graphite foil and the elec-
trode fabricated by optimum sputtering conditions ͑volume flow rate
of oxygen = 10 sccm, sputtering pressure = 20 mTorr, sputtering
power = 60 W, and sputtering time = 60 min͒ are shown in Fig.
12. The peaks in Fig. 12a are identified as coming from carbon
crystals. The major peaks in Fig. 12b is the same as those in Fig. 12a
except for the very low intensity peak from the MnO2 ͑2␪
= 37.237 from JCPDS͒ crystals in Fig. 12b due to a thin film of
poorly crystalline MnO2 deposited on the graphite foil.
Mn 2p3/2
Mn 3s
Volume flow rates of oxygen
͑sccm͒
Eb
͑eV͒
Eb͑1͒
͑eV͒
Eb͑2͒
͑eV͒
∆E
͑eV͒
The structure of the sputtered films depends on the substrate
temperature/target melting point and inert sputtering gas pressure.27
The substrate temperature ͑about 55°C͒/target melting point ͑about
5
10
15
641.96
642.06
642.30
83.01
83.44
84.62
88.28
88.23
89.45
5.27
4.79
4.83
Downloaded on 2015-03-04 to IP 169.230.243.252 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

A758
Journal of The Electrochemical Society, 157 ͑7͒ A753-A759 ͑2010͒
ꢅꢀꢀ
ꢀꢁꢂꢃ ꢄꢅꢆ ꢇ ꢈ ꢉꢅ
ꢄꢁꢀ
ꢄꢀꢀ
ꢃꢁꢀ
ꢃꢀꢀ
ꢂꢁꢀ
ꢂꢀꢀ
ꢁꢀ
ꢄꢀ ꢐꢑ
ꢒꢀ ꢐꢑ
ꢓꢀ ꢐꢑ
ꢀ
ꢀ
ꢁꢀꢀ
ꢂꢀꢀꢀ
ꢂꢁꢀꢀ ꢃꢀꢀꢀ ꢃꢁꢀꢀ
ꢆꢇ ꢈꢉꢊ ꢋꢌ ꢍꢎꢍꢏꢉ
Figure 11. The effects of sputtering time and different charge–discharge
cycles on the mass specific capacitance of the electrode ͑fabricated by opti-
mum sputtering conditions: 10 sccm oxygen, 20 mTorr, and 60 W͒.
tering pressure ͑10, 20, and 30 mTorr͒ in this research. Therefore,
according to the similar zone scheme ͑structure zone model͒ intro-
duced by Thornton with four zones ͑1, T, 2, 3͒, the manganese oxide
film on the pretreated graphite foil ͑substrate͒ belongs to the zone 1
structure ͑voided boundaries and fibrous grains͒, which appears in
amorphous and crystalline deposits, is the result of shadowing ef-
fects which overcome limited adatom surface diffusion.27 In addi-
tion, sputtering pressure affects film microstructure through several
indirect mechanisms. For example, if sputtering pressure is in-
creased to the point where the mean free path for elastic collisions
between the sputtered atoms and sputtering gas becomes of the order
of the target-substrate distance, the oblique component of the depo-
sition flux is increased because of gas scattering. This results in a
more open zone 1 structure.27 Moreover, the roughness and tempera-
ture of the substrate affect the growth of the manganese oxide film.27
To increase surface roughness of the graphite foil, it was then etched
in aqueous HCl and thus zone 1 was promoted by the roughness and
low temperature ͑about 55°C͒ of the substrate in this study.
ꢀꢊꢂꢃ ꢄꢅꢆ ꢇ ꢈꢋꢌ ꢉꢅ
Conclusions
A dc magnetron sputtering dry process was employed to deposit
manganese oxide onto graphite foil. Maximum mass specific capaci-
tance of 344 F/g was obtained with optimum sputtering conditions.
Furthermore, the mass specific capacitance increased at a lower
sputtering pressure/volume flow rates of oxygen, but decreased at a
higher sputtering pressure/volume flow rates of oxygen due to a
shadowing effect. The mass specific capacitance increased at a lower
sputtering power, but decreased at a higher sputtering power because
of a too strong ion and reflected neutral bombardment. Moreover,
the higher the volume flow rates of oxygen, the larger the amount of
trivalent manganese oxide and the higher the surface roughness, the
higher the mass specific capacitance at volume flow rates of oxygen
in the range 5–10 sccm, but the amounts of trivalent manganese
oxide were almost the same and the mass specific capacitance de-
creased at volume flow rates of oxygen in the range 10–15 sccm.
ꢀꢍꢂꢃ ꢄꢅꢆ ꢇ ꢎ ꢉꢅ
Figure 10. ͑Color online͒ AFM micrograph ͓͑a͒ 30, ͑b͒ 60, and ͑c͒ 90 min͔
of the electrode prepared at different sputtering time ͑with optimum sputter-
ing conditions: 10 sccm oxygen, 20 mTorr, and 60 W͒.
Figure 12. ͑Color online͒ XRD patterns of the freshly polished graphite foil
and the electrode fabricated by optimum sputtering conditions ͑volume flow
rate of oxygen = 10 sccm, sputtering pressure = 20 mTorr, sputtering
power = 60 W, and sputtering time = 60 min͒.
535°C͒ is about 0.1 and the inert sputtering gas pressure is about
0.83–3.34 Pa calculated from the volume flow rate ͑25 sccm͒ of
argon, volume flow rates ͑5, 10, and 15 sccm͒ of oxygen, and sput-
Downloaded on 2015-03-04 to IP 169.230.243.252 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

Journal of The Electrochemical Society, 157 ͑7͒ A753-A759 ͑2010͒
A759
17. C. C. Lin and H. W. Chen, Electrochim. Acta, 54, 3073 ͑2009͒.
18. S. C. Pang, M. A. Anderson, and W. C. Thomas, J. Electrochem. Soc., 147, 444
͑2000͒.
National Yunlin University of Science and Technology assisted in meeting
the publication costs of this article.
19. B. Djurfors, J. N. Broughton, M. J. Brett, and D. G. Ivey, J. Power Sources, 156,
741 ͑2006͒.
20. B. Djurfors, J. N. Broughton, M. J. Brett, and D. G. Ivey, J. Electrochem. Soc.,
153, A64 ͑2006͒.
References
1. B. E. Conway, Electrochemical Supercapacitors: Scientific Fundamentals and
Technological Applications, Kluwer Academic/Plenum Publishers, New York
͑1999͒.
2. R. Kötz and M. Carlen, Electrochim. Acta, 45, 2483 ͑2000͒.
3. J. K. Chang and W. T. Tsai, J. Electrochem. Soc., 150, A1333 ͑2003͒.
4. Y. S. Chen and C. C. Hu, Electrochem. Solid-State Lett., 6, A210 ͑2003͒.
5. Y. U. Jeong and A. Manthiram, J. Electrochem. Soc., 149, A1419 ͑2002͒.
6. C. C. Hu and C. C. Wang, J. Electrochem. Soc., 150, A1079 ͑2003͒.
7. H. P. Park, O. O. Park, K. H. Shin, C. S. Jin, and J. H. Kim, Electrochem. Solid-
State Lett., 5, H7 ͑2002͒.
21. J. H. Lim, D. J. Choi, H. K. Kim, W. H. Cho, and Y. S. Yoon, J. Electrochem. Soc.,
148, A275 ͑2001͒.
22. L. P. Ward and P. K. Datta, Thin Solid Films, 271, 101 ͑1995͒.
23. K. Koski, J. Holsa, J. Ernoult, and A. Rouzaud, Surf. Coat. Technol., 80, 195
͑1996͒.
24. G. Reiners and M. Griepentrog, Surf. Coat. Technol., 76–77, 809 ͑1995͒.
25. H. Liu and X. Zhang, Thin Solid Films, 240, 97 ͑1994͒.
26. X. Wang, G. Yang, X. Liu, Z. Zheng, W. Huang, and S. Zou, Mater. Sci. Eng., A,
156, 91 ͑1992͒.
27. M. Ohring, Materials Science of Thin Films, Academic, New York ͑2002͒.
28. G. Battaglin, V. Rigato, S. Zandolin, A. Benedetti, S. Ferro, L. Nanni, and A. De
Battisti, Chem. Mater., 16, 946 ͑2004͒.
8. R. N. Reddy and R. G. Reddy, J. Power Sources, 124, 330 ͑2003͒.
9. J. N. Broughton and M. J. Brett, Electrochim. Acta, 49, 4439 ͑2004͒.
10. M. Wu, G. A. Snook, G. Z. Chen, and D. J. Fray, Electrochem. Commun., 6, 499
͑2004͒.
11. A. Burke, J. Power Sources, 91, 37 ͑2000͒.
29. C. C. Hu and T. W. Tsou, J. Power Sources, 115, 179 ͑2003͒.
30. J. H. Huang and J. S. Chen, Thin Solid Films, 382, 139 ͑2001͒.
31. C. C. Lin and C. C. Yen, J. Appl. Electrochem., 37, 813 ͑2007͒.
32. H. Y. Lee and J. B. Goodenough, J. Solid State Chem., 144, 220 ͑1999͒.
33. M. Toupin, T. Brousse, and D. Bélanger, Chem. Mater., 16, 3184 ͑2004͒.
34. S. L. Kuo and N. L. Wu, J. Electrochem. Soc., 153, A1317 ͑2006͒.
35. M. Chigane and M. Ishikawa, J. Electrochem. Soc., 147, 2246 ͑2000͒.
12. J. K. Chang and W. T. Tsai, Electrochem. Commun., 6, 666 ͑2004͒.
13. M. S. Hong, S. H. Lee, and S. W. Kim, Electrochem. Solid-State Lett., 5, A227
͑2002͒.
14. J. H. Park, J. M. Ko, and O. O. Park, J. Electrochem. Soc., 150, A864 ͑2003͒.
15. J. R. Zhang, B. Chen, W. K. Li, J. J. Zhu, and L. P. Jiang, Int. J. Mod. Phys. B, 16,
4479 ͑2002͒.
16. C. C. Lin and C. C. Yen, J. Appl. Electrochem., 38, 1677 ͑2008͒.
Downloaded on 2015-03-04 to IP 169.230.243.252 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).