TIO2 Shen2005
TIO2 Shen2005
TIO2 Shen2005
G.X. Shen,* R.G. Du,* Y.C. Chen,* C.J. Lin,‡,* and D. Scantlebury**
ABSTRACT INTRODUCTION
Using ethyl acetoacetate (EAcAc) as a chelating agent, tita- Type 316L (UNS S31603)(1) stainless steel is recog-
nium dioxide (TiO2) sol with ultra-fine particles has been pre- nized as a highly corrosion-resistance alloy, and is
pared and deposited on Type 316L (UNS S31603) stainless the most widely used material for medical and dental
steel to form a nano-TiO2 coating by the dip-coating. A hydro-
applications.1 However, its self-passivation properties
thermal posttreatment method has been applied to obtain
may decrease because of local corrosion in chloride
crack-free coatings and to optimize the surface structure and
ion-containing media, and it has been reported that
properties. A self-assembly of fluoroalkylsilane (denoted as
FAS-13) has been conducted to enhance the hydrophobic prop- stainless steel does corrode in the biological environ-
erty for the surface of the nano-TiO2 coatings. The particle siz- ments.2 During the corrosion process, the release of
es of TiO2 sol have been analyzed by ζ potential analysis, and metallic ions such as Ni2+, Cr2+, and Cr3+ is particu-
the surface morphology, structure, and properties have been larly important because it may produce something
characterized by contract angle, x-ray diffraction (XRD), and toxic or irritant to the body tissue.3 A recent study
scanning electron microscopy (SEM) measurements. The re- showed that Type 316L stainless steel produces corro-
sults have shown that the surface of coatings is porous, with sion products above certain nonlethal concentrations,
approximately 375 nm thickness; the diameter of the particles which may thereby disturb the proliferation/differ-
of anatase TiO2 is uniform, in the range from 15 nm to 18 nm.
entiation relationship of osteoblastic human alveolar
The electrochemical tests have indicated that the hydrophobic
bone cell cultures, in a dose-dependent manner.4 So it
coatings of nano-TiO2 exhibit an excellent corrosion resistance.
is very necessary to develop novel techniques of sur-
KEY WORDS: corrosion resistance, hydrophobic coatings, face modification to improve its corrosion resistance
nano-titanium dioxide, Type 316L stainless steel and biological properties as well.
It is possible to minimize the corrosion of the
surface modification of Type 316L stainless steel sig-
nificantly, arrest the release of metallic ions and cor-
Submitted for publication December 2003; in revised form, rosion products, and improve biocompatibility. Many
December 2004.
‡
Corresponding author. E-mail: cjlin@xmu.edu.cn. attempts have already been made in this direction. In
* State Key Laboratory of Physical Chemistry of Solid Surfaces, particular, ceramic coatings such as titania (TiO2),5
Department of Chemistry, Xiamen University, Xiamen 361005,
China.
alumina (Al2O3),6 and zirconia (ZrO2)7 play a critical
** Corrosion and Protection Centre, UMIST, PO Box 88, Manchester role in providing excellent biocompatibility as well as
M60 1QD, United Kingdom. corrosion resistance. Up to now, two kinds of coat-
(1)
UNS numbers are listed in Metals and Alloys in the Unified Num-
bering System, published by the Society of Automotive Engineers ing deposition methods have been reported, viz. from
(SAE International) and cosponsored by ASTM International. gaseous-phased (physical vapor deposition [PVD]8
0010-9312/05/000191/$5.00+$0.50/0
CORROSION—Vol. 61, No. 10 © 2005, NACE International 943
CORROSION SCIENCE SECTION
and chemical vapor deposition [CVD])9 methods and and 0.3 µm Al2O3 powder, and then ultrasonically
from the liquid phase (sol-gel10 and electrodeposi- cleaned in acetone (CH3COCH3), ethanol, and distilled
tion process11) methods. Most of the ceramic coatings water for 10 min, respectively. After hot-air drying, it
are formed by the sol-gel method, because the sol- was coated with nano-scale TiO2 coatings using a dip-
gel techniques require considerably less equipment, coating method. In this procedure, the substrate was
can be amenable to any accessories in a large and immersed in the sol solution for 5 min and withdrawn
complex shape, and are applicable to the substrates at a speed of 0.5 mm/s ~ 1 mm/s. After drying natu-
that cannot withstand high temperatures.12 However, rally under ambient conditions, the specimens were
the coatings fabricated by the sol-gel process usu- heated in an oven at 150°C for 30 min. The operation
ally encounter a technical problem; that is, there mentioned above was repeated four times to increase
always exist crevice and cracks in the coatings result- the coating thickness. After this, the samples were
ing from the sintering in a high temperature, which heat-treated at 450°C for 30 min to enable the oxide
may become a detrimental factor, causing corrosion conversion and removal of solvent and residual organ-
of the substrates and coating disbonding from the ics in the coating. However, during the procedure of
substrates, and, finally, may limit the application of densification, the crystallization of TiO2, and the re-
ceramic coatings for corrosion protection.13 moval of residual hydroxyl and organic groups, the
The purpose of this study was to prepare nano- nano coatings were usually prone to cracking. To
TiO2 coatings on the surface of the Type 316L stain- eliminate the defects in the coating and to optimize
less steel and to improve its corrosion resistance. A the coating structure and properties, the nano-TiO2-
hydrothermal posttreatment method was developed to coated samples were immersed in boiling water for
eliminate cracks and crevices in the coatings and to 10 min ~ 50 min, and then hydrothermally treated
optimize the coatings’ structure. To enhance the sur- again at 450°C for 10 min.
face hydrophobic property, fluoroalkylsilane (FAS-13) FAS-13 was used as a water-repellent agent. A
was applied on the surface of nano-TiO2 coating using methanol (CH3OH) solution of 1% FAS-13 was hydro-
a self-assembly method. The electrochemical perfor- lyzed by adding a three-fold-molar excess of water at
mances of the nano composite coatings in oxygen- room temperature and was continuously stirred for
saturated Ringer solution (8.6 g/dm3 sodium chloride 3 h. The nano-TiO2 coatings were immersed in the so-
[NaCl], 0.3 g/dm3 potassium chloride [KCl], and lution for 3 h and dried at 140°C for 1 h.14
0.48 g/dm3 calcium chloride [CaCl2]) were measured
using direct current (DC) and alternating current (AC) Electrochemical Measurements
techniques, and the mechanism of their corrosion re- The measurements of Tafel polarization curves
sistance is discussed in the paper. and electrochemical AC impedance spectra for the
prepared coatings in the oxygen-saturated Ringer
EXPERIMENTAL PROCEDURES solution were carried out using an electrochemical
measurement system. At the same time, the effects of
Preparation of Titanium Dioxide the hydrothermal posttreatment on the corrosion re-
Colloidal Solution sistance of the TiO2 coatings have also been evaluated
Using ethyl acetoacetate (EAcAc) as a chelating using these methods. A three-electrode cell, including
agent, TiO2 sol was prepared by hydrolyzing tetra-n- a saturated calomel reference electrode (SCE), a plati-
butyl titanate (Ti [O-n-Bu]4) according to the following num auxiliary electrode, and the samples as working
procedure: First, 20 mL of ethanol (C2H5OH) and 1 mL electrodes, was used in the present study. The back
of EAcAc without further dehydration were mixed, and sides of the working electrodes were enveloped by
then at room temperature 2 mL of Ti (O-n-Bu)4 was epoxy resin and there only remained an exposure area
added dropwise into the mixed solution under con- (0.5 cm2 by 0.5 cm2) for testing. The Tafel polarization
stant stirring. The mixture was vigorously stirred at curves were carried out after the samples were im-
room temperature for 1 h, followed by the addition of mersed in the solution for 10 min. The potential scan-
0.2 mL of distilled water in 30 min for hydrolysis and ning range was ±120 mV at the open-circuit potential,
then continuously stirred for 10 h. Finally, the trans- and the potential scanning speed was 0.167 mV/s.
parent TiO2 sol was aged for 24 h before being coated The electrochemical AC impedance measurements
on the metal surface. were performed at open-circuit potential with 10 mV
amplitude of the sinusoidal voltage signal. The applied
Fabrication of the Hydrophobic frequencies ranged from 105 Hz to 10–3 Hz using five
Nano-Titanium Dioxide Coatings points/decade. The resulting impedance spectra were
The pretreatment of the substrate surfaces was processed using the “Equivalent Circuit ver. 3.96”
an important step in the preparation of a uniform made by Bankamp.15 All electrochemical tests were
nano coating. The Type 316L stainless steel sheets executed in a Faraday cage to minimize the external
(0.5 cm2 by 0.5 cm2) were ground with no. 320 to electronic interference, and the ambient temperature
1500 emery papers gradually, and polished with 1 µm condition was selected for all the measurements.
( Ti(OBu)x
nCH3COCH2COOCH3 + m Ti(OBu)4
O O
CH3C C COCH3 ( y FIGURE 1. Particle diameters distribution of the TiO2 sol.
H2
2O
pH
{Ti(OH)4}z (1) use of ceramic coatings under certain conditions. To
eliminate the cracks, various methods have been re-
In Equation (1), a simple polymeric reaction equa- ported, for example, PEG,16 polyvinyl alcohol (PVA),17
tion is given, where n, m, and p delegate the Moore and hydroxypropyl-cellulose (HPC),18 etc., have been
number of molecules participating in the reaction, added into the colloidal solution to improve the dry-
and x, y, and z stand for the polymeric degree. The ing properties of the gel because these are thought to
titanium alkoxide species reacts first with EacAc and avoid aggregation of the particles, adjust the viscosity
produces the hex-cyclic molecules; then the hex-cyclic of the sol, and increase the strength of the material
molecules form a tridimensional network macromo- preventing crack formation.19 However, these methods
lecular in a polymerization process. When adding H2O have made a little progress in preventing cracks in the
into the solution, the hydrolysis rate of the macromol- coatings using such techniques, and the processes
ecules is slower than that of the titanium alkoxide, are more complicated as well. The hydrothermal post-
which may control effectively the size of the particles treatment method developed in this work is able not
of the sol. And the sol remains stable and transpar- only to eliminate the cracks in the coatings success-
ent for a long time. Figure 1 shows the distribution of fully, but also to make the morphology, structure, and
TiO2 sol particle diameters measured using ζ potential properties of the coatings more perfect.
analysis. The diameter of ultra-fine particles is mostly Effects of immersion time of the TiO2 coatings in
1 nm in scale and no aggregation is observed in the boiling water on their morphology, particle size, and
coating. hydrophobic property have been specifically investi-
gated, and the results are listed in Table 1. It can be
Morphology and Properties seen that when the coatings are kept in boiling water
of the Nano-TiO2 Coatings for 10 min, the cracks disappear, the particle size gets
The SEM photographs of the TiO2 coatings are smaller, in the range from 15 nm ~ 18 nm, and the
shown in Figure 2. It is indicated that the coating sur- pore sizes among the particles are about 4 nm (Fig-
face is smooth with pore sizes in the range from ure 2[b]). The surface is smooth, uniform, and has no
8 nm ~ 14 nm, and the diameter of clusters is 18 nm aggregation (Figures 2[b] and [c]). The XRD pattern
~ 31 nm. From the cross section of the SEM image (Figure 3) shows that the TiO2 coating is pure anatase.
in Figure 2(c), the thickness of the coatings is about However, when the immersion time increases to
375 nm. It is found that there exist some cracks in 20 min, the aggregation phenomenon of the particles
the coating due to the shrinkage occurring during the is observed and the surface becomes coarse. When the
thermal process in Figure 2(d). These cracks show immersion time is longer than 30 min, the morphol-
little influence on the mechanical and adhesive prop- ogy and size of the particles becomes similar to that
erties of the coatings, because they appear only on the for the sample immersed for 10 min. It is interesting
upper surface of the films, as shown in Figure 2(c). to note that these phenomena are present repeatedly
However, as a protective coating for the metals, these for nearly every 10-min immersion. We believe that
cracks may become a potential risk causing local deg- during the hydrothermal posttreatment, there exists
radation of the metals. Actually, cracks in the coating a dynamic and competitive process for the rearrange-
are one of the most important reasons preventing the ment and crystallization of titania, which may play an
(a) (b)
(c) (d)
FIGURE 2. SEM photograph of the nano-TiO2-coated Type 316L: (a) and (d) before hydrothermal posttreatment, and
(b) and (c) after hydrothermal posttreatment.
important role in the morphology, size, and hydropho- derstood up to now. It could not be simply explained
bic property of the films. The thickness of the coat- by removal of the hydroxide radical and dust contami-
ings does not change too much in the course of the nation on the coatings.20 Tadanaga, et al., prepared
whole experiment, implying that the surface reaction super-water-repellent Al2O3 coatings21 and considered
retains equilibrium states and bond between the coat- that the hydrophobic property could result from the
ings, and the substrates are strong. Furthermore, it is nano-sized particles and the existence of air in the
indicated that the hydrophobic property of the coat- pores on the coating surface.
ings are closely associated with the morphology of the
coating. The contact angle of a fresh TiO2 surface is Corrosion Resistance
nearly 0 ± 5°, and while the coatings are kept in of the Hydrophobic Coatings
the dark for one month, the contract angle becomes The corrosion protection of a coated metal de-
104.4 ± 5°. The mechanism of the hydrophobic prop- pends closely on the barrier properties of the coating
erty of the nano-TiO2 coatings has not been well un- specifically to water and may include:
TABLE 1
Effects of the Immersion Time of the TiO2 Coatings in Boiling Water
on Morphology, Particle Sizes, and Hydrophobic Property of the TiO2 Coatings
No. Time (min) Configuration Diameter (nm) Thick (nm) Contact Angle (±5°)
1 0 Pores, aggregation, cracks 30 ~ 21 375 41.5
2 10 Pores, evenly, no cracks 15 ~ 18 375 104.4
3 20 Pores, aggregation, no cracks 39 ~ 18 375 78.5
4 30 Pores, evenly, no cracks 18 ~ 21 375 90.5
5 40 Pores, aggregation, no cracks 18 ~ 25 372 64.8
6 50 Pores, evenly, no cracks 18 ~ 21 370 78.5
FIGURE 4. Tafel polarization curves for bare Type 316L and coated
FIGURE 3. XRD pattern of the TiO2 coated on Type 316L. Type 316L electrodes in oxygen-saturated Ringer solution: (a) bare
Type 316L; (b) TiO2-coated Type 316L and before hydrothermal
post-treatment; (c) TiO2 coated after treatment; (d) FSA/TiO2-coated
—water absorption of the coating Type 316L.
—transportation of water into the coating
—accumulation of water at the coating/substrate
interface trodes were placed in the dark for more than 1 month
Thus, the hydrophobic coatings with low wettability before measurements, and all experiments were car-
should be able to effectively prevent water penetra- ried out in a dark room to avoid photoelectrochemical
tion and diffusion into the substrate, and exhibit an influence. By using “Corrosion Rate” software of the
excellent corrosion resistance in the wet environment. Electrochemical Measurement System, the corrosion
Using self-assembly of FAS-13 on the nano-TiO2 coat- parameters from the Tafel curves for the bare and
ings, its contract angle increased to 129 ± 5°, indicat- coated Type 316L stainless steel systems have been
ing that the hydrophobic property of the coating is presented in Table 2. From data in Tables 2(c) and
further strengthened. (d), we can see that TiO2- and FAS/TiO2-coated Type
Figure 4 shows a comparison of the Tafel polar- 316L electrodes exhibit relatively lower current densi-
ization curves for bare Type 316L, TiO2-coated, and ties, icorr, and decrease by 2 ~ 3 orders of magnitudes
FAS/TiO2-coated Type 316L stainless steel electrodes in comparison with the bare Type 316L stainless steel
in the Ringer solution. The nano-TiO2-coated elec- electrode under the same experimental condition. The
TABLE 2
Corrosion Parameters from the Tafel Polarization Measurements Shown in Figure 4
Icorr βc βa Ecorr Rp Corrosion Rate
Electrodes (A/cm2) (V/decade) (V/decade) (V) (Ω·cm2) (mm/y)
(a) 1.61 × 10–7 0.152 0.266 –0.145 3.35 × 105 1.29 × 10–3
(b) 5.58 × 10–8 0.074 0.243 –0.051 1.15 × 106 9.03 × 10–4
(c) 1.65 × 10–9 0.126 0.035 0.143 2.63 × 107 4.40 × 10–5
(d) 1.63 × 10–10 0.087 0.023 0.152 1.25 × 108 2.73 × 10–6
(a)
(b)
(c)
(d)
FIGURE 5. EIS plots of the bare Type 316L and the coated Type 316L electrodes in oxygen-saturated Ringer solution, at
the open-circuit potential. The frequency range was 105 Hz ~ 10–3 Hz. (a) Bare Type 316L; (b) nano-TiO2-coated and before
hydrothermal treatment; (c) nano-TiO2-coated and after hydrothermal treatment; (d) FSA/TiO2-coated Type 316L.
(a) (b)
FIGURE 6. Equivalent circuit for the (a) bare Type 316L and (b) the coated Type 316L according to the EIS measurements
in Figure 5.
polarization resistances (Rp) of TiO2- and FAS/TiO2- hydrothermal treatment on the corrosion resistance of
coated Type 316L are nearly 102 ~ 103 times higher the nano-TiO2 coatings. After applying a hydrothermal
than that of bare Type 316L stainless steel. The corro- posttreatment, the corrosion resistance of the coated
sion rate of the nano-TiO2-coated Type 316L are, re- sample is greatly improved due to the tiny cracks
spectively, approximately 300 and 2,000 times smaller being removed from the coatings. Furthermore, the
than that of the bare stainless steel. In addition, the FAS/TiO2-coated Type 316L is able to make a further
effects of the hydrothermal posttreatment on the cor- improvement in the corrosion resistance for the sub-
rosion protection of nano-TiO2 coatings has also been strates, which may result from an enhancement of the
given in Figure 4 and Tables 2(b) and (c). After the hy- hydrophobic property by the FAS-13 self-assembly
drothermal posttreatment, Rp of the nano-TiO2-coated on the perfect TiO2 coating, and by preventing the
Type 316L electrode increases by more than 10 times, solution from reaching the metallic surface. It is well
and icorr decreases by nearly three times compared known that a TiO2 coating has a very low electronic
with the samples without hydrothermal treatment. It conductance and might provide a barrier action.22
is noted furthermore that the corrosion potential of Since the discovery of the photoelectrochemical prop-
the TiO2- and FAS/TiO2-coated Type 316L electrodes erty of TiO2 film, it has been proposed that a cathodic
are more noble than that of the Type 316L stainless photoprotection mechanism based on the fact that the
steel, the corrosion potential shifting positively from photopotential of a TiO2 film generated by ultraviolet
–0.145 V to 0.143 V and 0.152 VSCE, respectively, (UV) illumination is more negative than the corrosion
which may be explained by the fact that the coatings potential of the metallic substrate. However, the long-
become denser and the substrates lean to a more term stability of this system is limited.23
stable passive state. In addition, the dimensionless The EIS spectra are analyzed with the EQUI-
rate (r) of the anodic Tafel slope (βa) to the cathodic VERT† program (ver. 3.96), where Q stands for the
Tafel slope (βc) for the coated electrodes (c) and (d) is constant phase element (CPE) in the circuit descrip-
less than 1. Consequently, there is evidence that the tion code, which is commonly used to replace ca-
corrosion resistance of the nano-TiO2- and FAS/TiO2- pacitance, because it is hardly a pure capacitance in
coated stainless steels are significantly improved. the real electrochemical process and it is defined by
The EIS spectra have also been obtained at the admittance Y and power index number n, given by Y =
open-circuit potential from bare Type 316L, TiO2-, and Y0(jω)n. This is a general dispersion formula; it shows
FAS/TiO2-coated Type 316L stainless steel electrodes a pure resistance if n = 0, and it is a pure capacitance
in the Ringer solution (Figure 5). The Bode diagrams if n = 1. In all cases under study, n is close to 1, rep-
of the coated electrodes evidently show a difference resenting a capacitive characteristic of the interfaces.
in the corrosion resistance from the bare Type 316L The equivalent circuit model is proposed in Figure 6,
electrodes, containing two time constants in the series in which Rs represents the solution resistance and a
presented in Figures 5(b), (c), and (d) and the cor- pair of elements of Qc (CPE of coatings) and Rc in par-
responding equivalent circuit in Figure 6. In general, allel replaces the dielectric properties of the coatings.
the AC impedance data at high frequencies represent Qdl (CPE of double layer) and Rt (reaction resistance)
the responses of coatings in the solution, and at low are adopted to describe the charge-transfer process
frequencies reflect the Faradaic reaction resistance Rt happening at the solution/steel interface in the pin-
and double-layer capacitance of the metal substrates. holes of the coating. Using the equivalent circuits, the
The Nyquist plots illustrate that the TiO2- and FAS/ curve fitting performed for both TiO2 and FAS/TiO2
TiO2-coated Type 316L stainless steel have an excel- coating systems and for the bare Type 316L stainless
lent corrosion resistance in comparison with the bare steel yields the values of parameters for the circuit
Type 316L stainless steel, as can be seen from the elements (Table 3). From Table 3, it is interesting to
loops with increasing resistance values (real axis). Fig- note that the apparent capacitance of the coatings QC,
ures 5(b) and (c) have also shown the influence of the especially for the electrodes (c) and (d), is very low
(approximately 10–8 F ~ 10–9 F). This lower value is
†
Trade name. consistent with the insulating nature of the coatings.
TABLE 3
EIS Parameters Corresponding to the Equivalent Circuit (Figure 6) Calculated from the EIS Measurement Shown in Figure 5
Qc Qdl
Rs Y0 Rc Y0 Rt
Electrodes (Ω·cm2) (Ω–1·cm–2·sn) n (Ω·cm2) (Ω–1·cm–2·sn) n Ω·cm2)
(Ω
(a) 30.9 — — — 6.08 × 10–6 0.863 1.32 × 105
(b) 45.8 4.4 × 10–8 0.705 2.15 × 103 1.90 × 10–5 0.873 1.68 × 106
(c) 57.1 2.8 × 10–8 0.741 2.68 × 104 3.58 × 10–7 0.795 8.00 × 107
(d) 38.0 1.61 × 10–9 0.589 8.25 × 102 1.54 × 10–6 0.922 2.03 × 108