FlatChem 1 (2016) 1119

Contents lists available at ScienceDirect

FlatChem
j o u r n a l h o m e p a g e : w w w . e l s e v i e r . c o m / l o c a t e / fl a t c

Functionalised graphene as a barrier against corrosion

Karanveer S. Aneja a, H.L. Mallika Bhm c, A.S. Khanna a, Siva Bhm a,b,
a
Department of Metallurgical Engineering and Materials Science, IIT Bombay, Powai, Mumbai, India
b
Talga Advance Materials GmbH, Prof.-Hermann-Klare-Str. 25, 07407 Rudolstadt, Germany
c
Warwick Technology Center, Voyager Building, University of Warwick Science Park, Sir William Lyons Road, Coventry CV4 9XF, UK

a r t i c l e i n f o a b s t r a c t

Article history: The present work investigates the barrier protection mechanism of functionalised graphene when coated
Received 8 April 2016 on mild steel substrate. Electrochemical parameters of the coating such as pore resistance, coating capac-
Revised 26 July 2016 itance and water uptake characteristics were evaluated using electrochemical impedance spectroscopy.
Accepted 15 August 2016
Additionally, breakpoint frequencies were deduced to evaluate the electrochemical activity at the surface
over time. These electrochemical results in combination with the structural analysis of the coating help in
developing a systematic understanding of the barrier properties rendered by graphene.
Keywords:
2016 Elsevier BV. This is an open access article under the CC BY-NC-ND license (http://creativecommons.
Corrosion
Graphene
org/licenses/by-nc-nd/4.0/).
Protective coatings
Water uptake
Breakpoint frequency
Barrier properties
Electrochemical impedance spectroscopy

1. Introduction
The use of graphene against corrosion has been widely reported
in the past [19] but studies involving mechanistic understanding
are limited. Bhm [1] describes the plausible reasons for anti-
corrosive applications of graphene are its high surface area, its abil-
ity to provide a barrier against corrosive species and its excellent
electrical conductivity. A systematic understanding of each of
these parameters will lead to a better understanding of the corro-
sion protection mechanism of graphene.
A corrosion (anodic) reaction occurs in an electrolytic medium
along with a cathodic reaction which balances the net charge.
Water molecules that permeate through the coating provide an
electrolytic medium for the corrosion reaction to take place.
Further, it can induce internal stresses in the coating and cause
physical swelling [10].
Water uptake can be calculated by either measuring the weight
change on exposure to atmosphere (gravimetric technique) or by
measuring the changes in coatings electrical capacitance
(capacitive technique) over long exposure to aqueous environ-
ments using electrochemical impedance spectroscopy (EIS).
Gravimetric technique can incorporate experimental errors when
thin coatings are to be investigated [11]. Capacitive technique
Corresponding author at: Department of Metallurgical Engineering and
Materials Science, IIT Bombay, Powai, Mumbai, India.
E-mail address: siva@talgagmbh.com (S. Bhm).
provides a better estimate for thin coating systems and is based
on the principle that water permeation increases the capacitance
of the coating [10]. Further, by using capacitance measurements,
coating properties such as diffusivity and equilibrium water con-
tent can also be obtained [12]. Other possible techniques include
differential scanning calorimetry (DSC) [13], which only detects
water in clustered form and FTIR [14], which can only provide a
qualitative assessment.
The onset and propagation of coating damage are important
parameters in determining the effectiveness of a particular coating
system. Experimentally, it is quite difficult to deduce these param-
eters. Breakpoint frequency method is a simple yet powerful
technique to determine the onset and propagation of coating
degradation using empirically derived equations without the need
to physically measure the coating damage [15]. Studies proposed
by Hirayama and Haruyama [16] and Scully et al. [1719] allow
determination of damaged area as small as 0.01%. This technique
has been widely used in the past [1523] to evaluate the effect
of coating constituents on the performance of coating systems.
The present study combines water uptake and breakpoint fre-
quency techniques to evaluate the effect of graphene as a barrier
against corrosion. Although other plate like materials such as, exfo-
liated clays, glass flakes and micaceous iron oxide (MIO) have been
traditionally used as pigments that render similar barrier proper-
ties to the primer coating, the advantage of using graphene and
other 2D materials lie in their higher specific surface area, owing

http://dx.doi.org/10.1016/j.flatc.2016.08.003
2452-2627/ 2016 Elsevier BV.
This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
12 K.S. Aneja et al. / FlatChem 1 (2016) 1119

to their two dimensional planar structure, which provides better matched with those of bare mild steel, indicating complete delam-
coverage for the same amount of material used [1,2]. ination of the coating.
Time dependent Bode and Nyquist impedance plots are shown
2. Experimental in Fig. 3. For the sake of simplicity, time dependent plots of only
functionalised graphene (5% graphene) are shown. The impedance
2.1. Materials and coating values gradually decrease with increasing exposure times till about
480 h beyond which the impedance values match those of bare
Electrochemically extracted graphene nano platelets (Talga mild steel indicating complete delamination of the coating.
Resources Limited Vittangi Graphite Deposit) were used as raw A comparison between functionalised graphene (5% graphene)
material in this study [24]. High quality few layer graphene (13 and hydrolysed APTES (0% graphene) at the beginning of exposure
layers) was prepared by high shear liquid exfoliation (8000 rpm (t = 0 h) is shown in Fig. 4. These plots clearly depict an order of
for 6 h) in a solution containing deionised water (Millipore) and magnitude difference in impedance values at the onset of
surfactants (Fig. 1). Characterisation of produced graphene has exposure.
been discussed in an earlier work [2] of the authors. The data obtained were simulated using an analogue electrical
3-(Aminopropyl)triethoxysilane (APTES), purchased from Alfa circuit, shown in Fig. 5(a), to obtain coating parameters (plotted
Aesar, was hydrolysed with acidified demineralised water in Fig. 6) at regular intervals. The difference between modelled
(pH = 5), and used to chemically modify (functionalise) different and experimental data was less than 5%. The modelled and
concentrations of graphene. The exact formulations are provided experimental data are plotted in the form of Nyquist plot in
in Table 1. Fig. 5(b).
The functionalised graphene formulations were applied on Pore resistance (Rp) for all the formulations is initially high cor-
ultrasonically cleaned mild steel panels using a bar applicator responding to intact barrier coating resistance but as water perme-
(20 micron wet film thickness) (Fig. 2). Hydrolysed APTES ates, Rp approaches a saturated value. This effect can be explained
(0% graphene) was used as a benchmark. Curing of these coatings by considering the growth of pathways through the coating as
was performed at 120 C for 90 s however as explained in an earlier water permeates. Increase in Rp with increasing graphene concen-
work [2], in order to evaporate unwanted additives, these coatings tration at any given time, as seen in Fig. 6(a), suggests a strong
were further heat treated at a temperature of 200 C in a vacuum interlinked structure, arising due to the functionalisation of gra-
oven. phene, that resists migration paths for corrosive species.
The values of constant phase element (CPE) are converted to
2.2. Electrochemical Impedance measurements effective capacitance (Ceff), using Eq. (1) [25], in order to determine
the water absorption parameters.
Acrylic tubes were fixed onto the coated panels so that a panel 1n=n
C eff Q 1=n RP 1
area of 3.14 cm2 could be exposed to the test solution i.e. 3.5 wt.%
NaCl solution (freely exposed to air, pH 5.5 at 25 C), for long n is the CPE parameter and is equal to 1 for pure capacitance.
durations. A standard three-electrode configuration was used with Rapid increase in Ceff values followed by a plateau correspond-
the substrate acting as a working electrode, saturated calomel elec- ing to saturation level is observed, as shown in Fig. 6(b). Ceff
trode as a reference, and platinum mesh as the counter electrode. depends on the dielectric constant of the coating and hence on
The impedance measurements were performed using Biologic the water absorbed [11].
SP300 potentiostat driven by EC-Lab software. The open circuit Since water absorption in coatings is assumed to be mainly due
potential was considered to be stable when its fluctuation was less to diffusion [12], it is possible to obtain the diffusion coefficient
than 5 mV for a period of 1000 s. using a particular solution of Ficks second law [26]
!
2.3. Scanning electron microscopy Mt 8 X
1
1 2n 12 Dp2
1 2 exp t 2
Ms p n0 2n 12
L2
Coating cross sections of hydrolysed APTES and functionalised
Mt is the amount of water absorbed at any time t(s), Ms the
graphene coated mild steel were analysed using environmental
maximum amount of water absorbed, L the coating thickness
scanning electron microscopy (FEI, Quanta 200 D 7548) for
(cm) and D is the diffusion coefficient (cm2 s1).
elemental mapping.
When considering short periods of time, Eq. (2) can be simpli-
fied to:
3. Results and analysis p
M t 4 D p
p t 3
Electrochemical impedance spectroscopy was carried out peri- Ms L p
odically; by applying a sinusoidal voltage of amplitude 10 mV at
the corrosion potential (Ecorr) and responses were measured over The capacitance method of evaluating water transport into
frequencies 10 mHz and 100 kHz, until the impedance values coatings is based on the assumptions that the change in capaci-
tance is entirely due to the permeation of water into the film, film
swelling is negligible, and permittivity of absorbed water is
Table 1 constant [11]. Dependence of coating capacitance on water perme-
Coating formulations. ation can be understood by considering a simple capacitor whose
Formulation APTES Water Hydrolysed Graphene capacitance is directly proportional to the relative dielectric
APTES constant e of the coating and to the area A of the capacitor, and
(% by (% by (% by weight) (% by inversely proportional to the coating thickness d:
weight) weight) weight)
eeo A
1 39.96 59.94 99.9 0.1 C 4
2 39.8 59.7 99.5 0.5 d
3 39.6 59.4 99 1
4 38 57 95 5
eo the dielectric constant of free space (8.854  1014 F/cm).
is
Since the dielectric constant of water is high (78.3 at 25 C) as
 K.S. Aneja et al. / FlatChem 1 (2016) 1119
(a)
Fig. 1. (a) Talga Vittangi graphite deposit (b) few layer graphene produced.
Fig. 2. Schematic diagram of functionalised graphene coating.
compared to coatings (typically 315), absorption of water causes
an increase in mixed dielectric constant, resulting in a higher
capacitance [27].
In early 1950s, Brasher and Kingsbury [28] derived an expres-
sion for quantification of water uptake by a coating, which
correlates water uptake and coating capacitance:
log CCot
X v by % volume  100
log H2 O
Xv denotes the percentage of water absorbed, Co denotes the mea-
sured coating capacitance at the time of immersion (t = 0), Ct
denotes the measured coating capacitance at any time t and eH2O
is the relative dielectric constant of water. Lindqvist [29] compared
the BrasherKingsbury expression with other subsequently derived
ones and concluded that Eq. (4) gives the best agreement with the
gravimetric method.
Using Eq. (5), it is possible to obtain the amount of water
absorbed at any time t by measuring coating capacitances. The
amount of water absorbed follows a similar trend as Ceff. Addition
of graphene decreases the water uptake drastically, as evident in
Fig. 7. The decreasing amount of water absorbed, at any given time,
with increasing graphene content is an indication of its ability to
provide a tortuous path, making it difficult for water ions to
permeate through the coatings. The amount of water absorbed at
saturation decreases with increasing graphene concentration up
to 5% loading (74% reduction at 288 hours exposure). It may be
noted that beyond 5% graphene loading, the APTES solid content
reduces appreciably. Hence in order to avoid discrepancies, these
results are not reported.
Assuming that the amount of water absorbed (Vt)  volume of
coating (Vc)
13
(b)
Vt Vt
 6
Vt Vc Vc
Hence,
Vt
Xv 7
Vc
Combining Eqs. (5) and (7) and simplifying we get,
log CCot V t Mt
8
log CCos V s Ms
Cs, Vs and Ms are coating parameters (capacitance, volume and mass
respectively) when saturated stage is achieved. Combining this
expression with Eq. (3),
p
log CCot 4 D p
p t 9
log CCos L p
From the above expression, it can be deduced that if the L.H.S.,
also known as dimensionless film capacitance (DFC), is varying
linearly with t0.5, the value of D (diffusion coefficient of water in
coating) is constant and the water diffusion approaches ideal
Fickian behaviour. The value of D, in such a case, can be obtained
from the slope of this linear equation [26].
Fitting the initial period to the power law equation, y = axn,
yields exponent values (n) of nearly 0.5 for all the systems. Accord-
5 ing to the diffusion theory, if n = 0.5, the water transport through
the coating approaches ideal diffusion behaviour and diffusion
coefficient is constant [26]. Hence Eq. (9) can be used to calculate
the diffusion coefficient by calculating the slope of DFC vs. t1/2 plot.
Fig. 8(a) shows the DFC vs. t1/2 plots for all the systems. Line fitting
and slope calculations were performed using Origin Pro software.
The non-graphene fitted line does not start from theoretical zero
because of experimental error. The values of diffusion coefficient
calculated using slope values are plotted in Fig. 8(b) as a function
of graphene loading and show a sharp decrease of almost an order
in magnitude on addition of just 0.1% graphene, establishing its
ability to inhibit water diffusion even at very low concentrations.
Arrhenius law gives the relationship between diffusion coeffi-
cient and activation energy [27]
 
EA
D A exp  10
RT
A is the Arrhenius constant, EA is the activation energy for the
process (diffusion), R is the gas constant and T is the temperature.
Since the testing temperature (T) is the same for all coating
systems, the incorporation of graphene, from a thermodynamic
perspective increases the activation energy of water diffusion pro-
cess thereby acting as a restricting force against water permeation
and onset of corrosion.
14 K.S. Aneja et al. / FlatChem 1 (2016) 1119

Fig. 3. Bode and Nyquist plots of functionalised graphene (5% graphene) over long exposure time.

Breakpoint frequencies, i.e. fh (high breakpoint frequency) and fl
(low breakpoint frequency), are defined as boundaries at which a
capacitive region transitions to a resistive region [17,30]. At the
transition, the impedance of the capacitance is equal to the plateau
resistance and the phase angle is 45. Hence a convenient route to
determine fh and fl is to deduce the point where bode phase angle
curve shifts below 45. Using empirically derived expressions,
breakpoint frequencies allow determination of electrochemically
active (or defect) area of the sample as a function of time.
Breakpoint frequencies can be correlated with the active (or
defect) area ratio by making use of the following relationships
[15]:
1 A
fh
2peeo qo Ao
d A
fl
2peeo r d Ao
where Ao and A are the total exposed coating area and defect (or
electrochemically active) area respectively, (A/Ao) is the active area
ratio, qo is the resistivity of in pores, d is the coating thickness, rd is
the unit area defective resistance and e and eo are the dielectric
constant for the coating saturated with water and permittivity of
free space, respectively.
An important feature to note here is that analysis based on fh is
preferred since it is independent of the substrate whereas fl is
dependent on the area normalised polarisation resistance at the
metal/electrolyte interface and hence the substrate characteristics
[17]. Additionally, evaluation of fh involves analysis of high fre-
quency data rather than the complicated low frequency region.
Variation of high breakpoint frequency (fh) is shown as a
function of time in Fig. 9. The onset of fh, or the initiation of elec-
trochemical activity at the surface, can be considered as a criterion
to evaluate the susceptibility to coating degradation [30]. Scully
[18] states that the minimum measurable value of fh is 100 Hz. fh
occurs immediately for hydrolysed APTES coating while it gets
delayed as the concentration of graphene increases. For 5%
graphene coatings, fh appears after 48 h of exposure making it least
11
susceptible to coating degradation.
With increasing time, the breakpoint frequency increases for all
the coatings, indicating propagation of delamination [15]. At any
12
time, fb for hydrolysed APTES is much more than that of graphene
based systems. Further, rapid increase in fb values for hydrolysed
APTES indicates faster propagation of corrosion attack. The results
are in accordance with water permeation results, underlying the
correlation between the water permeation, corrosion attack and
its propagation.
The surfaces of 5% functionalised graphene and hydrolysed
APTES coated systems, before and after the completion of test,
 K.S. Aneja et al. / FlatChem 1 (2016) 1119 15

Fig. 4. Bode and Nyquist plots of functionalised graphene (5% graphene) and hydrolysed APTES (0% graphene) at the onset on exposure (t = 0 h).

(a) (b)
Fig. 5. Electrochemical equivalent circuit used to simulate the experimental data.

are shown in Fig. 10. The surface is severely damaged in case of
hydrolysed APTES coated mild steel after 288 h of exposure. The
damage is far less in case of 5% functionalised graphene coated
system, even after 480 h of exposure. This shows that even though
the functionalised graphene coating delaminates beyond a certain
point, majority of the surface remains unaffected.
Fig. 11 shows the SEM cross sectional image of functionalised
graphene (with 5% graphene) coating. The thickness of this coating
was found to be about 3 microns. Further characterisation was
performed using elemental area mapping.
Figs. 12 and 13 show elemental mapping in different coating
systems. Fig. 12 depicts distribution of all elements of interest
for hydrolysed APTES coating (without any graphene) and
functionalised graphene coating (5% graphene). The elemental
distribution reveals a dense silica network throughout the coating
in both cases.
16 K.S. Aneja et al. / FlatChem 1 (2016) 1119

Fig. 6. Variation of (a) Rpore and (b) Ceff.

Fig. 7. Variation of water absorbed (%) with time. Fig. 9. Variation of breakpoint frequency (fb) as a function of time.

(a) (b)
Fig. 8. Variation of (a) dimensionless film capacitance (DFC) vs. t0.5 and linear fitting of initial points to calculate the slope (b) diffusion coefficient with graphene
concentration.
 K.S. Aneja et al. / FlatChem 1 (2016) 1119
Fig. 10. Functionalised graphene (5%) coated mild steel (a) before and (b) after
exposure for 480 h and hydrolysed APTES coated mild steel (c) before and (d) after
exposure for 288 h.
Fig. 11. (a) Cross-section of functionalised graphene coating on mild steel.
Fig. 12. Elemental mapping in (a) APTES coating (b) functionalised graphene coating.
17
Fig. 13 depicts mapping of elemental carbon alone in different
functionalised graphene formulations. It may be noted that the car-
bon concentration are from a combination of graphene and those
from the propyl groups in APTES. Significant increase in carbon
concentrations is solely due to increasing amount of graphene
since there is no appreciable change in APTES concentrations as
can be seen from Table 1. A standout feature of these maps is the
uniform dispersion of carbon and hence graphene concentrations
throughout the coating. This can be attributed to the formation
of three-dimensional network as a result of functionalisation of
graphene.
With increasing concentrations of graphene, the density and
network of carbon atoms intensify, thereby increasing the tortuous
path to be travelled by an ion in reaching the surface. This is
reflected in the results discussed above where the increase in
percentage of graphene improves the durability and hence the
performance of the coating increases. The 5% graphene system
shows a rich concentration of graphene in the coating and hence
performs better than others in terms of water absorption and
breakpoint frequency results.
4. Discussion
Graphene based anticorrosive coatings have been well reported
in the past [19] but the performances have varied significantly. It
may be noted that most of these studies have used the CVD route
of graphene deposition. Kirkland et.al [5] compared the corrosion
protection mechanism between CVD graphene coated Ni and Cu.
The performance of graphene coated Cu was reported to be more
or less similar to that of bare Cu but Prasai et al. [3] reported an
improvement of seven times while Raman et.al [6] reported
an improvement of two orders of magnitude. Prasai et al. [3]
demonstrated that multilayer graphene layers deposited on Ni
using CVD route lowers the corrosion rate by 20 times, much more
than that reported by Kirkland et al. Further, Schriver et al. [7] have
reported long-term deleterious role of the graphene coatings and
have attributed it to poor coverage at the boundaries of graphene
domains.
The reason for these inconsistent results arise from the fact that
graphene produced from CVD technique contains intrinsic defects,
which allow transportation of some molecules degrading the bar-
rier properties of graphene [9]. Further adhesion to the surface
plays an important role in long-term sustainability. Unless the
metal surface is completely covered by an adherent layer of
18 K.S. Aneja et al. / FlatChem 1 (2016) 1119

Fig. 13. (a) Carbon elemental mapping of (a) hydrolysed APTES coating (b) 0.1% graphene (c) 0.5% graphene (d) 1% graphene and (e) 5% graphene.

graphene, the potential of graphene as an effective barrier cannot
be achieved.
Dennis et al. [8] have shown that liquid exfoliated graphene
based nanocomposite coating systems show more than three
orders of improvement in corrosion resistance. Further, authors
[2] have demonstrated in an earlier work that the use of function-
alised (liquid exfoliated) graphene can overcome the shortcomings
of CVD graphene coatings by creating a uniform adherent barrier
over the metal substrate. This coating was shown to sustain Salt
fog test (ASTM B117) for 500 h without any significant damage
[2]. The long-term exposure test carried out in this study to
evaluate water uptake characteristics further reiterates the barrier
protection behaviour of functionalised graphene coatings and
provides a mechanistic understanding into the protective role of
graphene in coatings when incorporated effectively.
5. Conclusions
Water permeation is an important parameter for evaluating
protective coatings. Substantial decrease in the water uptake and
diffusion coefficient, as a function of graphene concentration, con-
firms the ability of graphene to provide barrier protection. Delay in
onset and propagation of corrosion attack on addition of graphene,
deduced using breakpoint frequency technique, further justifies
 K.S. Aneja et al. / FlatChem 1 (2016) 1119
the barrier characteristics of graphene. From a thermodynamic
point of view, the incorporation of graphene increased the activa-
tion energy peak for the water diffusion process, making it difficult
for ions to permeate through the coating. Structural analysis of the
functionalised graphene revealed a dense network of graphene lay-
ers dispersed throughout the coating, which explains the tremen-
dous improvement in water permeation and corrosion attack.
Acknowledgement
Authors would like to thank Talga Resources Ltd for providing
graphene nano platelet samples.
References
[1] S. Bhm, Graphene against corrosion, Nat. Nanotechnol. 9 (10) (2014) 741
742.
[2] K.S. Aneja et al., Graphene based anticorrosive coatings for Cr (VI) replacement,
Nanoscale 7 (42) (2015) 1787917888.
[3] D. Prasai et al., Graphene: corrosion-inhibiting coating, ACS Nano 6 (2) (2012)
11021108.
[4] S. Chen et al., Oxidation resistance of graphene-coated Cu and Cu/Ni alloy, ACS
Nano 5 (2) (2011) 13211327.
[5] N. Kirkland et al., Exploring graphene as a corrosion protection barrier, Corros.
Sci. 56 (2012) 14.
[6] R.S. Raman et al., Protecting copper from electrochemical degradation by
graphene coating, Carbon 50 (11) (2012) 40404045.
[7] M. Schriver et al., Graphene as a long-term metal oxidation barrier: worse than
nothing, ACS Nano 7 (7) (2013) 57635768.
[8] R.V. Dennis et al., Graphene nanocomposite coatings for protecting low-alloy
steels from corrosion, Am. Ceram. Soc. Bull. 92 (4) (2013) 1824.
[9] S.C. OHern et al., Selective molecular transport through intrinsic defects in a
single layer of CVD graphene, ACS Nano 6 (11) (2012) 1013010138.
[10] A. Castela, A. Simoes, Assessment of water uptake in coil coatings by
capacitance measurements, Prog. Org. Coat. 46 (1) (2003) 5561.
[11] F. Bellucci, L. Nicodemo, Water transport in organic coatings, Corrosion 49 (3)
(1993) 235247.
[12] G. Walter, The application of impedance spectroscopy to study the uptake of
sodium chloride solution in painted metals, Corros. Sci. 32 (10) (1991) 1041
1058.
[13] J. Castle, J. Watts, H. Leidheiser, Corrosion Control by Organic Coatings, NACE,
Houston, TX, 1981, pp. 7786.
19
[14] T. Nguyen et al., In situ measurement of water at the organic coating/substrate
interface, Prog. Org. Coat. 27 (1) (1996) 181193.
[15] Z. Feng, G. Frankel, Evaluation of coated Al alloy using the breakpoint
frequency method, Electrochim. Acta 187 (2016) 605615.
[16] R. Hirayama, S. Haruyama, Electrochemical impedance for degraded coated
steel having pores, Corrosion 47 (12) (1991) 952958.
[17] H.P. Hack, J.R. Scully, Defect area determination of organic coated steels in
seawater using the breakpoint frequency method, J. Electrochem. Soc. 138 (1)
(1991) 3340.
[18] J.R. Scully, Electrochemical impedance of organic-coated steel: correlation of
impedance parameters with long-term coating deterioration, J. Electrochem.
Soc. 136 (4) (1989) 979990.
[19] J. Scully, S. Hensley, Lifetime prediction for organic coatings on steel and a
magnesium alloy using electrochemical impedance methods, Corrosion 50 (9)
(1994) 705716.
[20] F. Mansfeld, Electrochemical impedance spectroscopy (EIS) as a new tool for
investigating methods of corrosion protection, Electrochim. Acta 35 (10)
(1990) 15331544.
[21] F. Mansfeld, Recording and analysis of AC impedance data for corrosion
studies, Corrosion 37 (5) (1981) 301307.
[22] C. Tsai, F. Mansfeld, Determination of coating deterioration with EIS: part II.
Development of a method for field testing of protective coatings, Corrosion 49
(9) (1993) 726737.
[23] F. Mansfeld, Use of electrochemical impedance spectroscopy for the study of
corrosion protection by polymer coatings, J. Appl. Electrochem. 25 (3) (1995)
187202.
[24] M. Thompson, Talga innovation reduces graphite processing steps
for Li-ion Batteries 2016, 2016, http://www.talgaresources.com/IRM/PDF/
1742/TalgaReducesGraphiteProcessingStepsForLiionBatteries.
[25] B. Hirschorn et al., Determination of effective capacitance and film thickness
from constant-phase-element parameters, Electrochim. Acta 55 (21) (2010)
62186227.
[26] C. Perez et al., Characterisation of the barrier properties of different paint
systems: part I. Experimental set-up and ideal Fickian diffusion, Prog. Org.
Coat. 36 (1) (1999) 102108.
[27] S. Downey, O. Devereux, The use of impedance spectroscopy in evaluating
moisture-caused failure of adhesives and paints, Corrosion 45 (8) (1989) 675
684.
[28] D. Brasher, A. Kingsbury, Electrical measurements in the study of immersed
paint coatings on metal. I. Comparison between capacitance and gravimetric
methods of estimating water-uptake, J. Appl. Chem. 4 (2) (1954) 6272.
[29] S. Lindqvist, Theory of dielectric properties of heterogeneous substances
applied to water in a paint film, Corrosion 41 (2) (1985) 6975.
[30] H. Hack, J. Scully, Galvanic corrosion prediction using long-and short-term
polarization curves, Corrosion 42 (2) (1986) 7990.