- Email: [email protected]
Electrophoretic deposition of hydroxyapatite nanostructured coatings with controlled porosity
Available online at www.sciencedirect.com
Journal of the European Ceramic Society 34 (2014) 97–106
Electrophoretic deposition of hydroxyapatite nanostructured coatings with controlled porosity Morteza Farrokhi-Rad, Saeede Kuche Loghmani, Taghi Shahrabi ∗ , Shirin Khanmohammadi Department of Materials Science & Engineering, Tarbiat Modares University, P.O. Box 14115-143, Tehran, Iran Received 10 June 2013; received in revised form 17 July 2013; accepted 21 July 2013 Available online 17 August 2013
Abstract Hydroxyapatite (HA) coatings with controlled porosity were prepared by electrophoretic deposition (EPD) method. Carbon black (CB) particles were used as the sacrificial template (porogen agent). Two component suspensions containing different concentrations of HA and CB particles were prepared in isopropanol. It was found that the finer and positively charged HA nanoparticles are heterocoagulated on the coarser and negatively charged CB particles to form CB–HA composite particles with net positive charge. The deposition rate from the suspensions with WR (CCB /CHA ratio) of 0.25 was faster than that of those with WR: 0.5 at initial times of EPD. However the situation was reversed at longer EPD times. It was also found that the amount of porosity in the coatings increases as the CB concentration in the suspension increases (15%, 24%, 31%, 43% for the coatings deposited from the suspensions with 20 g/L HA nanoparticles and 0, 5, 10 and 20 g/L CB particles, respectively). © 2013 Elsevier Ltd. All rights reserved. Keywords: Carbon black (CB) particles; Electrophoretic deposition (EPD); Hydroxyapatite (HA) nanoparticles; Isopropanol; Porosity
1. Introduction Porous ceramics have found applications in several fields such as biomaterials, filters, and optical materials.1 Controlling the pore characterizations such as size, size distribution and shape is of great importance in these applications. Different processing techniques have been used to fabricate porous ceramic materials.2 Colloidal processing has been used extensively to fabricate porous ceramics with highly controlled porous microstructures using polymeric or silica spherical particles as the sacrificial template (porogen agent) which are removed by either heat or chemical treatment.3–5 Electrophoretic deposition (EPD) is a colloidal processing method which has been used to fabricate porous coatings on the metallic substrates using polymeric particles as the sacrificial template.6–11 EPD is a two step process: in the first step charged particles are dispersed in a suitable liquid and migrate toward the electrode with opposite charge under the application of electric field; in the second step they deposit on the electrode and form a relatively dense layer of particles on it.12
Hydroxyapatite (HA) is the main inorganic part of human bone.13 HA has high bioactivity, biocompatibility, biodegradability and osteoconductvity14–16 making it an appropriate material for using in biomedical applications. However, HA has poor mechanical properties (for example low fracture toughness) limiting its usage in high load bearing applications. So usually HA is used in the form of coating on the biocompatible metals such as titanium and 316 L stainless steel and their alloys. Electrophoretic deposition (EPD) has been widely used to deposit HA coatings on the metallic substrates.17–21 It has been reported in the literature that the porous structure of the HA coatings is necessary to implant fixation by bone ingrowths into their pores.22–25 In the present work the economical fabrication of HA coatings with controlled porosity by EPD method has been reported. Carbon black (CB) particles have been used as the cost effective sacrificial template and the effect of their concentration on the EPD process as well as the porosity of obtained coatings have been discussed. 2. Experimental 2.1. Suspensions preparation
∗
Corresponding author. Tel.: +98 21 82883378. E-mail addresses: morteza farrokhi [email protected] (M. Farrokhi-Rad), [email protected] (T. Shahrabi). 0955-2219/$ – see front matter © 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.jeurceramsoc.2013.07.022
Hydroxyapatite (HA) nanoparticles were synthesized by metathesis method.26 Carbon black (CB) powder was used as
98
M. Farrokhi-Rad et al. / Journal of the European Ceramic Society 34 (2014) 97–106
the sacrificial template. The zeta potential and particles size distribution of HA and CB particles were measured in isopropanol (Malvern instrument, Worcestershire, UK). The suspensions of HA nanoparticles (10 and 20 g/L) with different concentrations of CB particles (CCB /CHA ratio (WR): 0, 0.25, 0.5 and 1) were prepared in isopropanol. The suspensions were magnetically stirred and ultrasonically dispersed (Sonopuls HD 3200, 20 kHz; Bandelin Co., Berlin, Germany) for 24 h and 10 min, respectively. The electrical conductivity of the suspensions was measured against CB concentration (Cond 720, WTW series; Inolab, Weilheim, Germany). The zeta potential of particles was measured in two component (HA + CB) suspension (Malvern instrument, Worcestershire, UK).
before and after sintering. The amount of porosity in the coatings was determined by image analyzing (MIP software).
2.2. Electrophoretic deposition The plates of 316 L stainless steel with the dimension of 40 mm × 20 mm × 1 mm were used as the substrate as well as counter electrodes (only 20 mm × 20 mm of substrate was exposed to deposition and remainder of their surface insulated). The distance between two electrodes was 1 cm in EPD cell. EPD was performed at 60 and 200 V from the suspensions with different concentrations of HA and CB particles using a laboratory D.C. power supply (HY30002E; Huayi Electronics Industry Co., Hangzhou, Zhejiang, China). The current density during EPD was recorded using a computer connected digital multimeter (289 True RMS; Fluke, Everett, WA). The immersion weight (Wimm ) of deposits was measured in situ according to the method described in Ref.27 The wet weight (Wwet ) of deposits prepared at 60 and 200 V for 15, 30, 60, 120, 240 and 360 s was measured immediately after EPD (GR-200 (0.1 mg accuracy); A & D Co., Tokyo, Japan). The wet density of deposits was calculated according to the Archimedes’ principle: ρwet =
Wwet Volwet
and
Volwet =
Fig. 1. Particles size distribution for HA and CB particles in isopropanol; the inset figures are the SEM images of HA and CB particles.
Wwet − Wimm ρISP
(1)
where Volwet is the wet volume of deposit and ρISP is the density of isopropanol (0.78 g/cm3 ). Thermogravimetric (TG) analysis was used to determine the temperature range where CB particles burn out completely as well as their amount in the deposits prepared from the suspensions with different concentrations of CB particles. The powders obtained from scratching the dried HA–CB composite coatings from the substrates were used as the samples for TG analysis. TG analysis was performed at the temperature range of 25–800 ◦ C (heating rate: 5 ◦ C/min). The deposits were dried at room temperature in air overnight and then sintered. The temperature program for sintering process was determined using the data obtained from TG analysis. The temperature was raised from room temperature (25 ◦ C) to 480 ◦ C at the heating rate of 10 ◦ C/min; then it was increased to 650 ◦ C at the heating rate of 1 ◦ C/min and was kept at this temperature for 30 min to complete the burning of CB particles; then temperature was raised to 700 ◦ C at the heating rate of 5 ◦ C/min and sintering was performed at this temperature for 1 h. The microstructure of deposits was observed by scanning electron microscope (SEM)
3. Results and discussion 3.1. Suspensions properties The zeta potential of HA and CB particles in isopropanol was +37.4 and −15.23 mV, respectively. The results for particles size distribution of HA and CB particles are shown in Fig. 1. The SEM images of HA and CB particles are also shown as the insets in Fig. 1. As can be seen the majority of CB particles are larger than HA particles. The mean agglomerate size and the range of particles size distribution are 159.7 nm and 37.84–531.2 nm for HA nanoparticles, respectively; while these parameters are 364.3 nm and 122.4–955.4 nm for CB particles, respectively. The zeta potential of particles against CB concentration in two component suspension is shown in Fig. 2. As can be seen particles are positively charged and their surface charge decreases continuously with CB concentration. When CB particles are added into the suspension of HA nanoparticles in isopropanol, the composite particles of CB–HA with a net positive surface charge are generated by the heterocoagulation of finer and positively charged HA nanoparticles (with the higher
Fig. 2. Zeta potential of particles in two component (HA + CB) suspension as a function of CB concentration.
M. Farrokhi-Rad et al. / Journal of the European Ceramic Society 34 (2014) 97–106
Fig. 3. Electrical conductivity of HA nanoparticles suspensions (10 and 20 g/L) against CB concentration.
absolute value of surface charge, zeta potential: +37.4 mV) on the coarser and negatively charged CB particles (with the lower absolute value of surface charge, zeta potential: −15.23 mV). The electrical conductivity of the HA nanoparticles suspensions (10 and 20 g/L) in isopropanol against CB concentration are shown in Fig. 3. As can be seen the electrical conductivity of both the suspensions (10 and 20 g/L) increases with CB concentration. The conductivity initiates to increase more steeply at the CB concentrations higher than 10 and 20 g/L for the suspensions with 10 and 20 g/L HA nanoparticles, respectively. In contrast to non-conductive HA nanoparticles, CB particles are good electrical conductor. The schematics of the suspensions with the constant concentration of HA nanoparticles and the varied concentrations of CB particles (CB concentration continuously increases from (a) to (d)) are shown in Fig. 4. The surface of CB particles is completely covered with non-conductive HA nanoparticles when the concentrations of CB particles is low in the suspension (Fig. 4(a)) resulting in the non-conductive
99
CB–HA composite particles. There is large number of free HA nanoparticles (HA nanoparticles not heterocoagulated on the surface of CB particles) in the suspension for this case. The number of free HA nanoparticles in the suspension decreases when the CB concentration increases in it. There is a CB concentration ∗ ) where all the HA nanoparticles are heterocoagulated on (CCB the surface of CB particles and cover their surface with the close packed structure (Fig. 4(b)). The number of heterocoagulated HA nanoparticles on the surface of CB particles continuously ∗ decreases if the concentration of CB particles is higher than CCB ∗ (CCB > CCB ) (Fig. 4(c) and (d)). The large surface area of CB particles is free (bare) from HA nanoparticles in this case. The CB–HA composite particles are submicron in size resulting in their strong random Brownian motions in the suspension. They collide with each other at high frequencies due to the Brownian motions regarding to the relatively high concentration of particles in the suspension. The very large number of these collisions occurs between CB–HA particles through their HA free surface area. These collisions can create numerous short life time chains of HA–CB particles connected to each other through their HA free surface area. These chains act as the electrically conductive paths leading to the sharp increase in the conductivity at CB ∗ (C ∗ concentrations higher than CCB CB > CCB ). The following equation can be derived for determining the average number of HA nanoparticles (¯nHA ) heterocoagulated on the surface of each CB particles if it is assumed that the HA nanoparticles cover the surface of CB particles with the most closely packed structure:
n¯ HA
√ ¯ CB 2 2 3π R = × ¯ HA R 3
(2)
¯ HA and R ¯ CB are the mean radius size for HA (79.85 nm) where R and CB (182.15 nm) agglomerates, respectively. The value ¯ HA of n¯ HA ≈ 19 is obtained by substituting the values of R ∗ ¯ and RCB in Eq. (2). CCB can be obtained by the following
Fig. 4. Schematic representation of suspensions with the constant concentration of HA nanoparticles and the different concentrations of CB particles (the concentration of CB particles continuously increases from (a) to (d)).
100
M. Farrokhi-Rad et al. / Journal of the European Ceramic Society 34 (2014) 97–106
Fig. 5. Current density variations during EPD from the suspensions with 10 g/L (a–d) and 20 g/L HA (e–h) and different WR: 0 (a and e), 0.25 (b and f), 0.5 (c and g) and 1 (d and h) at 60 and 200 V.
M. Farrokhi-Rad et al. / Journal of the European Ceramic Society 34 (2014) 97–106
equation: CHA n¯ HA m CHA n¯ HA V¯ HA ρHA CHA ¯ HA = ⇒ = ⇒ ∗ ∗ ∗ ¯ CCB m ¯ CB CCB CCB VCB ρCB 4 ¯3 ¯ HA 3 ρHA n¯ HA πR ρHA CHA R = 4 3 3 HA ⇒ ∗ = n¯ HA × × ¯ CB ¯ R CCB ρCB 3 π RCB ρCB (3) where m ¯ HA , V¯ CB , V¯ HA , ρCB (≈2 g/cm3 ) and ρHA (≈3.15 g/cm3 ) are mean mass of CB and HA agglomerates, mean volume of CB and HA agglomerates and density of CB and ∗ ≈ 2.5 HA particles, respectively. The value of CCH /CCB ∗ ∗ (CCB /CHA (WR ) ≈ 0.4) is obtained by substituting the value ∗ are 4 and of known parameters in Eq. (2). So the values of CCB 8 g/L for the HA nanoparticles concentrations of 10 and 20 g/L, respectively. 3.2. Electrophoretic deposition (A) Current density: The variations of current density during EPD at 60 and 200 V from the suspensions with 10 and 20 g/L HA nanoparticles and the different concentrations of CB
101
particles are shown in Fig. 5. As can be seen the current density decreases at both voltages for the suspensions with 10 g/L HA plus 0 and 2.5 g/L CB as well as for those with 20 g/L HA plus 0 and 5 g/L CB (WR: 0 and 0.25 < WR* ≈ 0.4) due to the formation of ceramic layer with the higher resistivity than suspension from which it deposits.22 The interesting trend is observed for the current density variations during EPD from the suspensions with 10 g/L HA plus 5 and 10 g/L CB and 20 g/L HA plus 10 and 20 g/L CB (WR: 0.5 and 1 > WR* ≈ 0.4): the current density is nearly constant with time at 60 V while it increases with time at 200 V. The increase in current density during EPD is possible only when the resistivity of the coating deposited on the substrate electrode is less than that of the suspension from which it deposits. The high concentration of interconnected and conductive CB particles in the deposits prepared from the suspensions with WR: 0.5 and 1 results in the lower resistance of deposit than that of the corresponding suspension. (B) Wet density of deposits: The wet density of deposits against deposition time for those formed at 60 and 200 V from the suspensions with 10 and 20 g/L HA nanoparticles and the different concentrations of CB particles is shown in Fig. 6. As can be seen the wet density increases with deposition time at
Fig. 6. Wet density of deposits against time for those prepared at 60 (a and c) and 200 V (b and d) from the suspensions with 10 (a and b) and 20 g/L (c and d) of HA nanoparticles and different concentrations of CB particles.
102
M. Farrokhi-Rad et al. / Journal of the European Ceramic Society 34 (2014) 97–106
both voltages, due to the particles rearrangement within the deposit resulting from electro-osmotic flows around the particles in deposit.28–31 Also the wet density increases more steeply at 200 than 60 V due to the stronger electro-osmotic flow at higher voltages. The wet density of deposits increases as 2.5 and 5 g/L CB particles are added into the suspensions with 10 and 20 g/L HA nanoparticles, respectively (WR: 0.25 < WR* ≈ 0.4). However it decreases upon the further addition of CB particles (5 and 10 g/L in the suspensions with 10 and 20 g/L HA nanoparticles, respectively; WR: 0.5 > WR* ≈ 0.4). The schematic illustrations of the deposits formed at 60 and 200 V from the suspensions with WR: 0, 0.25 and 0.5 are shown in Fig. 7. As it can be seen the deposits are more closely packed at 200 than 60 V. As it was mentioned previously when WR < WR* , the surface of CB particles are completely covered with the HA nanoparticles in a closely packed structure and there is still large number of free HA nanoparticles in the suspension. Both the free HA nanoparticles and CB–HA composite particles move toward the substrate electrode (cathode) and deposit on it during EPD resulting in the structure schematically shown in Fig. 7(c) and (d). Although the density of CB (≈2 g/cm3 ) is less than HA (≈3.15 g/cm3 ) but the wet density increases when CB particles are incorporated in the deposit formed from the suspensions with WR: 0.25, due to the lower bulk density of HA nanoparticles with interparticles isopropanol between them than that of single CB particle. Also this may be due to the better packing of HA nanoparticles that strongly adhered to the CB particles. The wet density decreases for the deposits formed from the suspensions with WR: 0.5 > WR* ≈ 0.4. As it is schematically shown in Fig. 7(e) and (f) this can be due to the non-compact heterocoagulation of HA nanoparticles on the CB particles as well as the high concentration of lighter CB particles (CB density ≈ 2 g/cm3 and HA density ≈ 3.15 g/cm3 ) in the deposit. As mentioned previously the rearrangement of particles within the deposit can occur due to the intensive electro-osmotic flow at 200 V. Some CB particles can connect to each other as a result of this rearrangement and generate several electrical conductive paths; these paths result in the lower resistivity of the deposit than the corresponding suspension causing in the current density raising with EPD time at 200 V for the suspensions with WR: 0.5 > WR* ≈ 0.4. (C) Kinetics of deposition: The in situ weight of deposits against EPD time for those formed at 60 and 200 V from the suspensions with 10 and 20 g/L HA nanoparticles and the different concentrations of CB particles is shown in Fig. 8. As can be seen the deposition weight increases with increasing the applied voltage and particles concentration in the suspension in agreement with the Hamaker equation.32 The deposition rate increases when the CB particles are added into the 10 and 20 g/L suspension of HA nanoparticles with WR: 0.25 and 0.5 due to the increase in the effective particles concentration (HA + CB) in the suspension. The interesting observation is that the deposition rate from these suspensions (WR: 0.25 and 0.5) is nearly equal at 60 V; while it is different at 200 V. The addition of CB has dual effect on the EPD kinetics: on one hand it increases the deposition rate by enhancing the effective concentration of particles (HA + CB) in the suspension as well as decreasing the
Fig. 7. Schematic representation of coatings deposited at 60 (a, c and e) and 200 V (b, d and f) from the suspensions with different WR: 0 (a and b), 0.25 (c and d) and 0.5 (e and f). The dashed red lines in (f) show the electrical conducive paths generated by the connection of CB particles to each other. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
M. Farrokhi-Rad et al. / Journal of the European Ceramic Society 34 (2014) 97–106
103
Fig. 8. In situ weight of deposits against time for those prepared at 60 (a and b) and 200 V (c and d) from the suspensions with 10 (a and c) and 20 g/L (b and d) of HA nanoparticles and the different concentrations of CB particles.
voltage drop over the deposit by decreasing its resistivity due to the incorporation of conductive CB particles in it; on the other hand it decreases the deposition rate by decreasing the surface charge of composite particles (Fig. 2). It can be said that the net effect of these parameters are nearly balanced during EPD at 60 V from the suspensions with WR: 0.25 (effective particle concentration: 12.5 and 25 g/L for 10 and 20 g/L suspensions of HA nanoparticles, respectively and zeta potential: +30.3 mV) and 0.5 (effective particle concentration: 15 and 30 g/L for 10 and 20 g/L suspensions of HA nanoparticles, respectively and zeta potential: +21.7 mV) so that the deposition rate is nearly same for them. It can be seen in Fig. 8(c) and (d) that the deposition rate is faster from the suspensions with WR: 0.25 than ones with WR: 0.5 at 200 V and initial times of EPD, while the situation is reversed at longer EPD times. The mobility of free HA nanoparticles as well as CB–HA composite particles in the suspensions with WR: 0.25 is higher than that of CB–HA composite particles in the suspensions with WR: 0.5. The higher mobility of CB–HA composite particles in the suspensions with WR: 0.25 is due to the larger number of HA nanoparticles (positively charged) heterocoagulated on the CB particles (negatively charged) resulting in the higher net positive charge (higher zeta potential, Fig. 2) and so higher mobility for them. As it was mentioned the deposits prepared from the suspensions with WR: 0.25 have high
resistivity due to the lower concentration of CB particles in them. The deposition rate deviates from linearity during EPD from these suspensions at 200 V due to the faster deposition and so faster increase in the resistance of deposit and so considerable voltage drop over it (due to the high concentration of suspensions the effect of their depletion from particles on the deviation of deposition rate from linearity is negligible). While, deposition rate does not deviate from linearity during EPD from the suspensions with WR: 0.5 at 200 V due to the lower resistivity of the deposits prepared from these suspensions and so lower voltage drop over them. As mentioned previously the resistivity of the deposits prepared from the suspensions with WR: 0.5 and 1 are low because of the high concentration of conductive CB particles in their structure. The high deviation of deposition rate from linearity during EPD at 200 V from the suspensions with WR: 0.25 and the negligible deviation from linearity during EPD from the suspensions with WR: 0.5 results in relatively large difference between the weight of deposits prepared from these suspensions at longer EPD times (Fig. 8(c) and (d)). The deposition rate decreases and several fluctuations occur in the curve of in situ deposition weight versus time upon the further addition of CB particles (WR: 1) into the suspension. These fluctuations prove the successive detachment of deposit from the substrate and its reformation during EPD. The deposit
104
M. Farrokhi-Rad et al. / Journal of the European Ceramic Society 34 (2014) 97–106
detachment during EPD from the suspensions with WR: 1 can be due to the very low mobility of CB–HA composite particles resulting from the small number of HA nanoparticles heterocoagulated on the CB particles (the zeta potential of composite particles is +9.2 mV in the suspension with WR: 1). Also the fluctuations in electrical conductivity of suspension due to Brownian motions (as explained previously) result in the fluctuations in current density during EPD which can cause in the detachment of deposit through generating the turbulence at the interface between deposit and suspension. (D) TG analysis: The results obtained from the TG analysis of the powders scratched from the coatings deposited at 60 V for 1 min from the suspensions with 10 and 20 g/L HA nanoparticles and the different concentrations of CB particles are shown in Fig. 9. The results show that CB particles completely burn out at temperature range of 480–650 ◦ C. The CB amounts in the coatings are 17.6, 32.7 and 49.1%wt for those deposited from the suspensions with 10 g/L HA and 2.5 (WR: 0.25), 5 (WR: 0.5) and 10 g/L (WR: 1) CB particles,
Fig. 9. Results obtained from the TG analysis of the powder scratched from the coatings deposited at 60 V and 1 min from the suspensions with 10 (a) and 20 g/L (b) of HA nanoparticles and the different concentrations of CB particles.
respectively. Also the CB amounts in the coatings are 17.3, 33.2 and 49.4%wt for those deposited from the suspensions with 20 g/L HA and 5 (WR: 0.25), 10 (WR: 0.5) and 20 g/L (WR: 1) CB particles, respectively. These results show that the CB weight percent amounts in the coatings deposited from the suspensions with the same values of WR are nearly identical. Also the results show that the weight percent amount of CB in the suspensions and in the coatings deposited from them is nearly identical for those with WR: 0.5 (CCB /(CCB + CHA ) = 33.33%) and WR: 1 (CCB /(CCB + CHA ) = 50%). However, the weight percent amount of CB in the suspensions with WR: 0.25 (CCB /(CCB + CHA ) = 20%) is higher than that of the coatings deposited from them (about 17.6 and 17.3%wt for suspensions with 10 g/L HA plus 2.5 g/L CB and 20 g/L HA plus 5 g/L CB, respectively). As explained previously there is a large number of free HA nanoparticles with higher mobility than CB–HA composite particles in the suspensions with WR: 0.25 < WR* ≈ 0.4; these free HA nanoparticles move faster than CB–HA composite particles toward the substrate electrode during EPD leading to the lower weight percent amount of CB in the coatings than in the suspensions. There are no free HA nanoparticles in the suspensions with WR: 0.5 > WR* ≈ 0.4 and WR: 1 > WR* ≈ 0.4 and all of them are heterocoagulated on the surface of CB particles resulting in the nearly same weight percent amount of CB in these suspensions as well as the coatings deposited from them. (E) Microstructure of deposits: The SEM images of the coatings are shown in Fig. 10. These coatings have been deposited at 60 V for 30 s from the suspensions with 20 g/L HA nanoparticles and 0 (Fig. 10(a)–(c)), 5 (Fig. 10(d)–(f)), 10 (Fig. 10(g)–(i)) and 20 g/L (Fig. 10(j)–(l)) CB particles. Fig. 10(a), (d), (g) and (j) shows the microstructure of coatings before sintering while Fig. 10(b), (e), (h) and (k) shows their microstructure after sintering at 700 ◦ C for 1 h. Fig. 10(c), (f), (i) and (l) shows the results obtained from the image analyzing of the SEM images shown in Fig. 10(b), (e), (h) and (k). The porosities are represented in red color. As can be seen in Fig. 10(b), (e), (h) and (k) the microstructures with controlled porosity are obtained by removing the CB particles through their burning out during sintering process. The amounts of porosity obtained by image analyzing are 15%, 24%, 31% and 43% for the coatings deposited from the suspensions with 0, 5, 10 and 20 g/L CB particles (HA: 20 g/L). As it is expected the amount of porosity in the coatings increases as the concentration of CB particles increases in the suspension. When the sacrificial template is not used to generate porosity in EPD coatings, the pore size distribution and the amount of porosity which can be obtained are limited; the mean pore size is usually smaller and the amount of porosity is lower for this case depending on the suspension properties (concentration, zeta potential of particles and their size distribution), EPD parameters (voltage and time) and sintering conditions (temperature and time).8,31–34 While when sacrificial template is used, the pore size distribution and the amount of porosity in EPD coatings can be controlled by controlling the particle size distribution of sacrificial template particles and their concentration in the suspension, respectively.
M. Farrokhi-Rad et al. / Journal of the European Ceramic Society 34 (2014) 97–106
105
Fig. 10. SEM images of the coatings deposited at 60 V and 30 s from the suspensions with 20 g/L HA and different concentration of CB particles: 0 g/L (a–c), 5 g/L (d–f), 10 g/L (g–i) and 20 g/L (j–l) before sintering (a, d, g and j) and after sintering at 700 ◦ C for 1 h (b, e, h and k). (c, f, i and l) are the results obtained from the image analyzing of (b, e, h and k), respectively.
4. Conclusions Hydroxyapatite (HA) coatings with controlled porosity were prepared by electrophoretic deposition (EPD) method. Carbon black (CB) particles were used as the porogen agent. The suspensions were prepared at different concentrations of HA (10 and 20 g/L) and CB particles (WR: 0, 0.25, 0.5 and 1) in isopropanol. It was found that the finer and positively charged HA nanoparticles are heterocoagulated on the coarser and negatively charged CB particles and form CB–HA composite particles with net positive charge. The electrical conductivity of the suspensions
increases with CB concentration. The conductivity increases more steeply at the CB concentrations higher than 10 and 20 g/L for the suspensions with 10 and 20 g/L HA nanoparticles, respectively. It was fund that there is a concentration of ∗ ) where all the HA nanoparticles are heteCB particles (CCB ∗ were 4 and rocoagulated on their surfaces. The values of CCB 8 g/L for the HA nanoparticles concentrations of 10 and 20 g/L, respectively. The current density during EPD increased at 200 V for the suspensions with WR: 0.5 due to the lower resistivity of the deposits than that of the suspensions. The curve of in situ deposition weight versus time showed good linearity in case of
106
M. Farrokhi-Rad et al. / Journal of the European Ceramic Society 34 (2014) 97–106
suspensions with WR: 0.5. The amounts of porosity were 15%, 24%, 31%, 43% for the coatings deposited from the suspensions with 0, 5, 10 and 20 g/L CB particles (HA: 20 g/L). Acknowledgment The authors would like to thank Ms. Najibeh Tolu Farrokh for her valuable helps on this work. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/ j.jeurceramsoc.2013.07.022. References 1. Velev OD, Jode TA, Lobo RF, Lenhoff AM. Porous silica via colloidal crystallization. Nature 1997;389:447–8. 2. Studart AR, Gonzenbach UT, Tervoort E, Gauckler LJ. Processing routes to macroporous ceramics: a review. J Am Ceram Soc 2006;89:1771–89. 3. Jiang P, Ostojic GN, Narat R, Mittleman DM, Covin VL. The fabrication and bandgap engineering of photonic multilayers. Adv Mater 2001;13:389–93. 4. Wijnhoven JEGJ, Zevenhuizen SJM, Hendriks MA, Livos V, Vanmaekelbergh D. Electrochemical assembly of ordered macropores in gold. Adv Mater 2000;12:888–90. 5. Yan H, Zhang K, Blanford CF, Francis LF, Stein A. In vitro hydroxycarbonate apatite mineralization of CaO–SiO2 sol–gel glasses with a three-dimensionally ordered macroporous structure. Chem Mater 2001;13:1374–82. 6. Hamagami Ji Ato Y, Kanamura K. Fabrication of highly ordered macroporous apatite coating onto titanium by electrophoretic deposition method. Solid State Ion 2004;172:331–4. 7. Yousefpour M, Afshar A, Chen J, Zhang X. Electrophoretic deposition of porous hydroxyapatite coatings using polytetrafluoroethylene particles as templates. Mater Sci Eng C 2007;27:1482–6. 8. Ma J, Wang C, Peng KW. Electrophoretic deposition of porous hydroxyapatite scaffold. Biomaterials 2003;24:3505–10. 9. Radice S, Dietsch H, Mischler S, Michler J. Electrophoretic deposition of functionalized polystyrene particles for TiO2 multi-scale structured surfaces. Surf Coat Technol 2010;204:1749–54. 10. Neirinck BB, Mattheys T, Braem A, Fransaer J, Van der Biest O, Vleugels J. Porous titanium coatings obtained by electrophoretic deposition. Adv Eng Mater 2008;10:246–9. 11. Neirinck B, Fransaer J, Van der Biest O, Vleugels J. A novel route to produce porous ceramics. J Eur Ceram Soc 2009;29:833–6. 12. Besra L, Liu M. A review on fundamental and applications of electrophoretic deposition. Prog Mater Sci 2007;52:1–61. 13. Narasaraju TSB, Phebe DE. Review: some physico-chemical aspects of hydroxylapatite. J Mater Sci 1996;31:1–21. 14. Gomez-Vega JM, Saiz E, Tomsia AP, Marshall GW, Marshall SJ. Bioactive glass coatings with hydroxyapatite and Bioglass® particles on Ti-based implants. 1. Processing. Biomaterials 2000;21:105–11.
15. Ducheyne P, Hench LL, Kagan II A, Martens M, Bursens A, Mulier JC. Effect of hydroxyapatite impregnation on skeletal bonding of porous coated implants. J Biomed Mater Res 1980;14:225–37. 16. Tampieri A, Celotti G, Sprio S, Mingazzini C. Characteristics of synthetic hydroxyapatites and attempts to improve their thermal stability. Mater Chem Phys 2000;64:54–61. 17. Kwok CT, Wong PK, Cheng FT, Man HC. Characterization and corrosion behavior of hydroxyapatite coatings on Ti6Al4V fabricated by electrophoretic deposition. Appl Surf Sci 2009;255: 6736–44. 18. Xiao XF, Liu RF. Effect of suspension stability on electrophoretic deposition of hydroxyapatite coatings. Mater Lett 2006;60:2627–32. 19. Wei M, Ruys AJ, Milthorpe BK, Sorrell CC. Solution ripening of hydroxyapatite nanoparticles: effects on electrophoretic deposition. J Biomed Mater Res 1999;45:11–9. 20. Farrokhi-Rad M, Shahrabi T. Effect of triethanolamine on the electrophoretic deposition of hydroxyapatite nanoparticles in isopropanol. Ceram Int 2013;39:7007–13. 21. Kuche Loghmani S, Farrokhi-Rad M, Shahrabi T. Effect of polyethylene glycol on the electrophoretic deposition of hydroxyapatite nanoparticles in isopropanol. Ceram Int 2013;39:7043–51. 22. Cook SD, Thomas KA, Delton JE, Volkman TK, Whitecloud III ThS, Key JF. Hydroxylapatite coating of porous implants improves bone ingrowth and interface attachment strength. J Biomed Mater Res 1992;26: 989–1001. 23. Kienapfel H, Sprey C, Wilke A, Griss P. Implant fixation by bone ingrowth. J Arthroplasty 1999;14:355–68. 24. Hench LL. Bioceramics. J Am Ceram Soc 1998;81:1705–28. 25. Nguyen HQ, Deporter DA, Pilliar RM, Valiquette N, Yakubovich R. The effect of sol–gel-formed calcium phosphate coatings on bone ingrowth and osteoconductivity of porous-surfaced Ti alloy implants. Biomaterials 2004;25:865–76. 26. Wei M, Ruys AJ, Milthorpe BK, Sorrell CC. Precipitation of hydroxyapatite nanoparticles: effects of precipitation method on electrophoretic deposition. J Mater Sci: Mater Med 2005;16:319–24. 27. Farrokhi-Rad M, Shahrabi T. Electrophoretic deposition of titania nanoparticles: sticking parameter determination by an in-situ study of the EPD kinetics. J Am Ceram Soc 2012;95:3434–40. 28. Bohmer M. In situ observation of two-dimensional clustering during electrophoretic deposition. Langmuir 1996;12:5747–50. 29. Solomentsev Y, Bohmer M, Anderson JL. Particle clustering and pattern formation during electrophoretic deposition: a hydrodynamic model. Langmuir 1997;13:6058–68. 30. Yeh SR, Seul M, Shraiman BI. Assembly of ordered colloidal aggregates by electric-field-induced fluid flow. Nature 1997;386: 57–9. 31. Ji C, Lan W, Xiao P. Fabrication of yttria-stabilized zirconia coatings using electrophoretic deposition: packing mechanism during deposition. J Am Ceram Soc 2008;91:1102–9. 32. Hamaker HC. Formation of a deposit by electrophoresis. Trans Faraday Soc 1940;35:279–87. 33. Tabellion J, Clasen R. Electrophoretic deposition from aqueous suspensions for near-shape manufacturing of advanced ceramics and glasses-applications. J Mater Sci 2004;39:803–11. 34. Chen CY, Chen SY, Liu DM. Electrophoretic deposition forming of porous alumina membranes. Acta Mater 1999;47:2717–26.
Journal of the European Ceramic Society 34 (2014) 97–106
Electrophoretic deposition of hydroxyapatite nanostructured coatings with controlled porosity Morteza Farrokhi-Rad, Saeede Kuche Loghmani, Taghi Shahrabi ∗ , Shirin Khanmohammadi Department of Materials Science & Engineering, Tarbiat Modares University, P.O. Box 14115-143, Tehran, Iran Received 10 June 2013; received in revised form 17 July 2013; accepted 21 July 2013 Available online 17 August 2013
Abstract Hydroxyapatite (HA) coatings with controlled porosity were prepared by electrophoretic deposition (EPD) method. Carbon black (CB) particles were used as the sacrificial template (porogen agent). Two component suspensions containing different concentrations of HA and CB particles were prepared in isopropanol. It was found that the finer and positively charged HA nanoparticles are heterocoagulated on the coarser and negatively charged CB particles to form CB–HA composite particles with net positive charge. The deposition rate from the suspensions with WR (CCB /CHA ratio) of 0.25 was faster than that of those with WR: 0.5 at initial times of EPD. However the situation was reversed at longer EPD times. It was also found that the amount of porosity in the coatings increases as the CB concentration in the suspension increases (15%, 24%, 31%, 43% for the coatings deposited from the suspensions with 20 g/L HA nanoparticles and 0, 5, 10 and 20 g/L CB particles, respectively). © 2013 Elsevier Ltd. All rights reserved. Keywords: Carbon black (CB) particles; Electrophoretic deposition (EPD); Hydroxyapatite (HA) nanoparticles; Isopropanol; Porosity
1. Introduction Porous ceramics have found applications in several fields such as biomaterials, filters, and optical materials.1 Controlling the pore characterizations such as size, size distribution and shape is of great importance in these applications. Different processing techniques have been used to fabricate porous ceramic materials.2 Colloidal processing has been used extensively to fabricate porous ceramics with highly controlled porous microstructures using polymeric or silica spherical particles as the sacrificial template (porogen agent) which are removed by either heat or chemical treatment.3–5 Electrophoretic deposition (EPD) is a colloidal processing method which has been used to fabricate porous coatings on the metallic substrates using polymeric particles as the sacrificial template.6–11 EPD is a two step process: in the first step charged particles are dispersed in a suitable liquid and migrate toward the electrode with opposite charge under the application of electric field; in the second step they deposit on the electrode and form a relatively dense layer of particles on it.12
Hydroxyapatite (HA) is the main inorganic part of human bone.13 HA has high bioactivity, biocompatibility, biodegradability and osteoconductvity14–16 making it an appropriate material for using in biomedical applications. However, HA has poor mechanical properties (for example low fracture toughness) limiting its usage in high load bearing applications. So usually HA is used in the form of coating on the biocompatible metals such as titanium and 316 L stainless steel and their alloys. Electrophoretic deposition (EPD) has been widely used to deposit HA coatings on the metallic substrates.17–21 It has been reported in the literature that the porous structure of the HA coatings is necessary to implant fixation by bone ingrowths into their pores.22–25 In the present work the economical fabrication of HA coatings with controlled porosity by EPD method has been reported. Carbon black (CB) particles have been used as the cost effective sacrificial template and the effect of their concentration on the EPD process as well as the porosity of obtained coatings have been discussed. 2. Experimental 2.1. Suspensions preparation
∗
Corresponding author. Tel.: +98 21 82883378. E-mail addresses: morteza farrokhi [email protected] (M. Farrokhi-Rad), [email protected] (T. Shahrabi). 0955-2219/$ – see front matter © 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.jeurceramsoc.2013.07.022
Hydroxyapatite (HA) nanoparticles were synthesized by metathesis method.26 Carbon black (CB) powder was used as
98
M. Farrokhi-Rad et al. / Journal of the European Ceramic Society 34 (2014) 97–106
the sacrificial template. The zeta potential and particles size distribution of HA and CB particles were measured in isopropanol (Malvern instrument, Worcestershire, UK). The suspensions of HA nanoparticles (10 and 20 g/L) with different concentrations of CB particles (CCB /CHA ratio (WR): 0, 0.25, 0.5 and 1) were prepared in isopropanol. The suspensions were magnetically stirred and ultrasonically dispersed (Sonopuls HD 3200, 20 kHz; Bandelin Co., Berlin, Germany) for 24 h and 10 min, respectively. The electrical conductivity of the suspensions was measured against CB concentration (Cond 720, WTW series; Inolab, Weilheim, Germany). The zeta potential of particles was measured in two component (HA + CB) suspension (Malvern instrument, Worcestershire, UK).
before and after sintering. The amount of porosity in the coatings was determined by image analyzing (MIP software).
2.2. Electrophoretic deposition The plates of 316 L stainless steel with the dimension of 40 mm × 20 mm × 1 mm were used as the substrate as well as counter electrodes (only 20 mm × 20 mm of substrate was exposed to deposition and remainder of their surface insulated). The distance between two electrodes was 1 cm in EPD cell. EPD was performed at 60 and 200 V from the suspensions with different concentrations of HA and CB particles using a laboratory D.C. power supply (HY30002E; Huayi Electronics Industry Co., Hangzhou, Zhejiang, China). The current density during EPD was recorded using a computer connected digital multimeter (289 True RMS; Fluke, Everett, WA). The immersion weight (Wimm ) of deposits was measured in situ according to the method described in Ref.27 The wet weight (Wwet ) of deposits prepared at 60 and 200 V for 15, 30, 60, 120, 240 and 360 s was measured immediately after EPD (GR-200 (0.1 mg accuracy); A & D Co., Tokyo, Japan). The wet density of deposits was calculated according to the Archimedes’ principle: ρwet =
Wwet Volwet
and
Volwet =
Fig. 1. Particles size distribution for HA and CB particles in isopropanol; the inset figures are the SEM images of HA and CB particles.
Wwet − Wimm ρISP
(1)
where Volwet is the wet volume of deposit and ρISP is the density of isopropanol (0.78 g/cm3 ). Thermogravimetric (TG) analysis was used to determine the temperature range where CB particles burn out completely as well as their amount in the deposits prepared from the suspensions with different concentrations of CB particles. The powders obtained from scratching the dried HA–CB composite coatings from the substrates were used as the samples for TG analysis. TG analysis was performed at the temperature range of 25–800 ◦ C (heating rate: 5 ◦ C/min). The deposits were dried at room temperature in air overnight and then sintered. The temperature program for sintering process was determined using the data obtained from TG analysis. The temperature was raised from room temperature (25 ◦ C) to 480 ◦ C at the heating rate of 10 ◦ C/min; then it was increased to 650 ◦ C at the heating rate of 1 ◦ C/min and was kept at this temperature for 30 min to complete the burning of CB particles; then temperature was raised to 700 ◦ C at the heating rate of 5 ◦ C/min and sintering was performed at this temperature for 1 h. The microstructure of deposits was observed by scanning electron microscope (SEM)
3. Results and discussion 3.1. Suspensions properties The zeta potential of HA and CB particles in isopropanol was +37.4 and −15.23 mV, respectively. The results for particles size distribution of HA and CB particles are shown in Fig. 1. The SEM images of HA and CB particles are also shown as the insets in Fig. 1. As can be seen the majority of CB particles are larger than HA particles. The mean agglomerate size and the range of particles size distribution are 159.7 nm and 37.84–531.2 nm for HA nanoparticles, respectively; while these parameters are 364.3 nm and 122.4–955.4 nm for CB particles, respectively. The zeta potential of particles against CB concentration in two component suspension is shown in Fig. 2. As can be seen particles are positively charged and their surface charge decreases continuously with CB concentration. When CB particles are added into the suspension of HA nanoparticles in isopropanol, the composite particles of CB–HA with a net positive surface charge are generated by the heterocoagulation of finer and positively charged HA nanoparticles (with the higher
Fig. 2. Zeta potential of particles in two component (HA + CB) suspension as a function of CB concentration.
M. Farrokhi-Rad et al. / Journal of the European Ceramic Society 34 (2014) 97–106
Fig. 3. Electrical conductivity of HA nanoparticles suspensions (10 and 20 g/L) against CB concentration.
absolute value of surface charge, zeta potential: +37.4 mV) on the coarser and negatively charged CB particles (with the lower absolute value of surface charge, zeta potential: −15.23 mV). The electrical conductivity of the HA nanoparticles suspensions (10 and 20 g/L) in isopropanol against CB concentration are shown in Fig. 3. As can be seen the electrical conductivity of both the suspensions (10 and 20 g/L) increases with CB concentration. The conductivity initiates to increase more steeply at the CB concentrations higher than 10 and 20 g/L for the suspensions with 10 and 20 g/L HA nanoparticles, respectively. In contrast to non-conductive HA nanoparticles, CB particles are good electrical conductor. The schematics of the suspensions with the constant concentration of HA nanoparticles and the varied concentrations of CB particles (CB concentration continuously increases from (a) to (d)) are shown in Fig. 4. The surface of CB particles is completely covered with non-conductive HA nanoparticles when the concentrations of CB particles is low in the suspension (Fig. 4(a)) resulting in the non-conductive
99
CB–HA composite particles. There is large number of free HA nanoparticles (HA nanoparticles not heterocoagulated on the surface of CB particles) in the suspension for this case. The number of free HA nanoparticles in the suspension decreases when the CB concentration increases in it. There is a CB concentration ∗ ) where all the HA nanoparticles are heterocoagulated on (CCB the surface of CB particles and cover their surface with the close packed structure (Fig. 4(b)). The number of heterocoagulated HA nanoparticles on the surface of CB particles continuously ∗ decreases if the concentration of CB particles is higher than CCB ∗ (CCB > CCB ) (Fig. 4(c) and (d)). The large surface area of CB particles is free (bare) from HA nanoparticles in this case. The CB–HA composite particles are submicron in size resulting in their strong random Brownian motions in the suspension. They collide with each other at high frequencies due to the Brownian motions regarding to the relatively high concentration of particles in the suspension. The very large number of these collisions occurs between CB–HA particles through their HA free surface area. These collisions can create numerous short life time chains of HA–CB particles connected to each other through their HA free surface area. These chains act as the electrically conductive paths leading to the sharp increase in the conductivity at CB ∗ (C ∗ concentrations higher than CCB CB > CCB ). The following equation can be derived for determining the average number of HA nanoparticles (¯nHA ) heterocoagulated on the surface of each CB particles if it is assumed that the HA nanoparticles cover the surface of CB particles with the most closely packed structure:
n¯ HA
√ ¯ CB 2 2 3π R = × ¯ HA R 3
(2)
¯ HA and R ¯ CB are the mean radius size for HA (79.85 nm) where R and CB (182.15 nm) agglomerates, respectively. The value ¯ HA of n¯ HA ≈ 19 is obtained by substituting the values of R ∗ ¯ and RCB in Eq. (2). CCB can be obtained by the following
Fig. 4. Schematic representation of suspensions with the constant concentration of HA nanoparticles and the different concentrations of CB particles (the concentration of CB particles continuously increases from (a) to (d)).
100
M. Farrokhi-Rad et al. / Journal of the European Ceramic Society 34 (2014) 97–106
Fig. 5. Current density variations during EPD from the suspensions with 10 g/L (a–d) and 20 g/L HA (e–h) and different WR: 0 (a and e), 0.25 (b and f), 0.5 (c and g) and 1 (d and h) at 60 and 200 V.
M. Farrokhi-Rad et al. / Journal of the European Ceramic Society 34 (2014) 97–106
equation: CHA n¯ HA m CHA n¯ HA V¯ HA ρHA CHA ¯ HA = ⇒ = ⇒ ∗ ∗ ∗ ¯ CCB m ¯ CB CCB CCB VCB ρCB 4 ¯3 ¯ HA 3 ρHA n¯ HA πR ρHA CHA R = 4 3 3 HA ⇒ ∗ = n¯ HA × × ¯ CB ¯ R CCB ρCB 3 π RCB ρCB (3) where m ¯ HA , V¯ CB , V¯ HA , ρCB (≈2 g/cm3 ) and ρHA (≈3.15 g/cm3 ) are mean mass of CB and HA agglomerates, mean volume of CB and HA agglomerates and density of CB and ∗ ≈ 2.5 HA particles, respectively. The value of CCH /CCB ∗ ∗ (CCB /CHA (WR ) ≈ 0.4) is obtained by substituting the value ∗ are 4 and of known parameters in Eq. (2). So the values of CCB 8 g/L for the HA nanoparticles concentrations of 10 and 20 g/L, respectively. 3.2. Electrophoretic deposition (A) Current density: The variations of current density during EPD at 60 and 200 V from the suspensions with 10 and 20 g/L HA nanoparticles and the different concentrations of CB
101
particles are shown in Fig. 5. As can be seen the current density decreases at both voltages for the suspensions with 10 g/L HA plus 0 and 2.5 g/L CB as well as for those with 20 g/L HA plus 0 and 5 g/L CB (WR: 0 and 0.25 < WR* ≈ 0.4) due to the formation of ceramic layer with the higher resistivity than suspension from which it deposits.22 The interesting trend is observed for the current density variations during EPD from the suspensions with 10 g/L HA plus 5 and 10 g/L CB and 20 g/L HA plus 10 and 20 g/L CB (WR: 0.5 and 1 > WR* ≈ 0.4): the current density is nearly constant with time at 60 V while it increases with time at 200 V. The increase in current density during EPD is possible only when the resistivity of the coating deposited on the substrate electrode is less than that of the suspension from which it deposits. The high concentration of interconnected and conductive CB particles in the deposits prepared from the suspensions with WR: 0.5 and 1 results in the lower resistance of deposit than that of the corresponding suspension. (B) Wet density of deposits: The wet density of deposits against deposition time for those formed at 60 and 200 V from the suspensions with 10 and 20 g/L HA nanoparticles and the different concentrations of CB particles is shown in Fig. 6. As can be seen the wet density increases with deposition time at
Fig. 6. Wet density of deposits against time for those prepared at 60 (a and c) and 200 V (b and d) from the suspensions with 10 (a and b) and 20 g/L (c and d) of HA nanoparticles and different concentrations of CB particles.
102
M. Farrokhi-Rad et al. / Journal of the European Ceramic Society 34 (2014) 97–106
both voltages, due to the particles rearrangement within the deposit resulting from electro-osmotic flows around the particles in deposit.28–31 Also the wet density increases more steeply at 200 than 60 V due to the stronger electro-osmotic flow at higher voltages. The wet density of deposits increases as 2.5 and 5 g/L CB particles are added into the suspensions with 10 and 20 g/L HA nanoparticles, respectively (WR: 0.25 < WR* ≈ 0.4). However it decreases upon the further addition of CB particles (5 and 10 g/L in the suspensions with 10 and 20 g/L HA nanoparticles, respectively; WR: 0.5 > WR* ≈ 0.4). The schematic illustrations of the deposits formed at 60 and 200 V from the suspensions with WR: 0, 0.25 and 0.5 are shown in Fig. 7. As it can be seen the deposits are more closely packed at 200 than 60 V. As it was mentioned previously when WR < WR* , the surface of CB particles are completely covered with the HA nanoparticles in a closely packed structure and there is still large number of free HA nanoparticles in the suspension. Both the free HA nanoparticles and CB–HA composite particles move toward the substrate electrode (cathode) and deposit on it during EPD resulting in the structure schematically shown in Fig. 7(c) and (d). Although the density of CB (≈2 g/cm3 ) is less than HA (≈3.15 g/cm3 ) but the wet density increases when CB particles are incorporated in the deposit formed from the suspensions with WR: 0.25, due to the lower bulk density of HA nanoparticles with interparticles isopropanol between them than that of single CB particle. Also this may be due to the better packing of HA nanoparticles that strongly adhered to the CB particles. The wet density decreases for the deposits formed from the suspensions with WR: 0.5 > WR* ≈ 0.4. As it is schematically shown in Fig. 7(e) and (f) this can be due to the non-compact heterocoagulation of HA nanoparticles on the CB particles as well as the high concentration of lighter CB particles (CB density ≈ 2 g/cm3 and HA density ≈ 3.15 g/cm3 ) in the deposit. As mentioned previously the rearrangement of particles within the deposit can occur due to the intensive electro-osmotic flow at 200 V. Some CB particles can connect to each other as a result of this rearrangement and generate several electrical conductive paths; these paths result in the lower resistivity of the deposit than the corresponding suspension causing in the current density raising with EPD time at 200 V for the suspensions with WR: 0.5 > WR* ≈ 0.4. (C) Kinetics of deposition: The in situ weight of deposits against EPD time for those formed at 60 and 200 V from the suspensions with 10 and 20 g/L HA nanoparticles and the different concentrations of CB particles is shown in Fig. 8. As can be seen the deposition weight increases with increasing the applied voltage and particles concentration in the suspension in agreement with the Hamaker equation.32 The deposition rate increases when the CB particles are added into the 10 and 20 g/L suspension of HA nanoparticles with WR: 0.25 and 0.5 due to the increase in the effective particles concentration (HA + CB) in the suspension. The interesting observation is that the deposition rate from these suspensions (WR: 0.25 and 0.5) is nearly equal at 60 V; while it is different at 200 V. The addition of CB has dual effect on the EPD kinetics: on one hand it increases the deposition rate by enhancing the effective concentration of particles (HA + CB) in the suspension as well as decreasing the
Fig. 7. Schematic representation of coatings deposited at 60 (a, c and e) and 200 V (b, d and f) from the suspensions with different WR: 0 (a and b), 0.25 (c and d) and 0.5 (e and f). The dashed red lines in (f) show the electrical conducive paths generated by the connection of CB particles to each other. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
M. Farrokhi-Rad et al. / Journal of the European Ceramic Society 34 (2014) 97–106
103
Fig. 8. In situ weight of deposits against time for those prepared at 60 (a and b) and 200 V (c and d) from the suspensions with 10 (a and c) and 20 g/L (b and d) of HA nanoparticles and the different concentrations of CB particles.
voltage drop over the deposit by decreasing its resistivity due to the incorporation of conductive CB particles in it; on the other hand it decreases the deposition rate by decreasing the surface charge of composite particles (Fig. 2). It can be said that the net effect of these parameters are nearly balanced during EPD at 60 V from the suspensions with WR: 0.25 (effective particle concentration: 12.5 and 25 g/L for 10 and 20 g/L suspensions of HA nanoparticles, respectively and zeta potential: +30.3 mV) and 0.5 (effective particle concentration: 15 and 30 g/L for 10 and 20 g/L suspensions of HA nanoparticles, respectively and zeta potential: +21.7 mV) so that the deposition rate is nearly same for them. It can be seen in Fig. 8(c) and (d) that the deposition rate is faster from the suspensions with WR: 0.25 than ones with WR: 0.5 at 200 V and initial times of EPD, while the situation is reversed at longer EPD times. The mobility of free HA nanoparticles as well as CB–HA composite particles in the suspensions with WR: 0.25 is higher than that of CB–HA composite particles in the suspensions with WR: 0.5. The higher mobility of CB–HA composite particles in the suspensions with WR: 0.25 is due to the larger number of HA nanoparticles (positively charged) heterocoagulated on the CB particles (negatively charged) resulting in the higher net positive charge (higher zeta potential, Fig. 2) and so higher mobility for them. As it was mentioned the deposits prepared from the suspensions with WR: 0.25 have high
resistivity due to the lower concentration of CB particles in them. The deposition rate deviates from linearity during EPD from these suspensions at 200 V due to the faster deposition and so faster increase in the resistance of deposit and so considerable voltage drop over it (due to the high concentration of suspensions the effect of their depletion from particles on the deviation of deposition rate from linearity is negligible). While, deposition rate does not deviate from linearity during EPD from the suspensions with WR: 0.5 at 200 V due to the lower resistivity of the deposits prepared from these suspensions and so lower voltage drop over them. As mentioned previously the resistivity of the deposits prepared from the suspensions with WR: 0.5 and 1 are low because of the high concentration of conductive CB particles in their structure. The high deviation of deposition rate from linearity during EPD at 200 V from the suspensions with WR: 0.25 and the negligible deviation from linearity during EPD from the suspensions with WR: 0.5 results in relatively large difference between the weight of deposits prepared from these suspensions at longer EPD times (Fig. 8(c) and (d)). The deposition rate decreases and several fluctuations occur in the curve of in situ deposition weight versus time upon the further addition of CB particles (WR: 1) into the suspension. These fluctuations prove the successive detachment of deposit from the substrate and its reformation during EPD. The deposit
104
M. Farrokhi-Rad et al. / Journal of the European Ceramic Society 34 (2014) 97–106
detachment during EPD from the suspensions with WR: 1 can be due to the very low mobility of CB–HA composite particles resulting from the small number of HA nanoparticles heterocoagulated on the CB particles (the zeta potential of composite particles is +9.2 mV in the suspension with WR: 1). Also the fluctuations in electrical conductivity of suspension due to Brownian motions (as explained previously) result in the fluctuations in current density during EPD which can cause in the detachment of deposit through generating the turbulence at the interface between deposit and suspension. (D) TG analysis: The results obtained from the TG analysis of the powders scratched from the coatings deposited at 60 V for 1 min from the suspensions with 10 and 20 g/L HA nanoparticles and the different concentrations of CB particles are shown in Fig. 9. The results show that CB particles completely burn out at temperature range of 480–650 ◦ C. The CB amounts in the coatings are 17.6, 32.7 and 49.1%wt for those deposited from the suspensions with 10 g/L HA and 2.5 (WR: 0.25), 5 (WR: 0.5) and 10 g/L (WR: 1) CB particles,
Fig. 9. Results obtained from the TG analysis of the powder scratched from the coatings deposited at 60 V and 1 min from the suspensions with 10 (a) and 20 g/L (b) of HA nanoparticles and the different concentrations of CB particles.
respectively. Also the CB amounts in the coatings are 17.3, 33.2 and 49.4%wt for those deposited from the suspensions with 20 g/L HA and 5 (WR: 0.25), 10 (WR: 0.5) and 20 g/L (WR: 1) CB particles, respectively. These results show that the CB weight percent amounts in the coatings deposited from the suspensions with the same values of WR are nearly identical. Also the results show that the weight percent amount of CB in the suspensions and in the coatings deposited from them is nearly identical for those with WR: 0.5 (CCB /(CCB + CHA ) = 33.33%) and WR: 1 (CCB /(CCB + CHA ) = 50%). However, the weight percent amount of CB in the suspensions with WR: 0.25 (CCB /(CCB + CHA ) = 20%) is higher than that of the coatings deposited from them (about 17.6 and 17.3%wt for suspensions with 10 g/L HA plus 2.5 g/L CB and 20 g/L HA plus 5 g/L CB, respectively). As explained previously there is a large number of free HA nanoparticles with higher mobility than CB–HA composite particles in the suspensions with WR: 0.25 < WR* ≈ 0.4; these free HA nanoparticles move faster than CB–HA composite particles toward the substrate electrode during EPD leading to the lower weight percent amount of CB in the coatings than in the suspensions. There are no free HA nanoparticles in the suspensions with WR: 0.5 > WR* ≈ 0.4 and WR: 1 > WR* ≈ 0.4 and all of them are heterocoagulated on the surface of CB particles resulting in the nearly same weight percent amount of CB in these suspensions as well as the coatings deposited from them. (E) Microstructure of deposits: The SEM images of the coatings are shown in Fig. 10. These coatings have been deposited at 60 V for 30 s from the suspensions with 20 g/L HA nanoparticles and 0 (Fig. 10(a)–(c)), 5 (Fig. 10(d)–(f)), 10 (Fig. 10(g)–(i)) and 20 g/L (Fig. 10(j)–(l)) CB particles. Fig. 10(a), (d), (g) and (j) shows the microstructure of coatings before sintering while Fig. 10(b), (e), (h) and (k) shows their microstructure after sintering at 700 ◦ C for 1 h. Fig. 10(c), (f), (i) and (l) shows the results obtained from the image analyzing of the SEM images shown in Fig. 10(b), (e), (h) and (k). The porosities are represented in red color. As can be seen in Fig. 10(b), (e), (h) and (k) the microstructures with controlled porosity are obtained by removing the CB particles through their burning out during sintering process. The amounts of porosity obtained by image analyzing are 15%, 24%, 31% and 43% for the coatings deposited from the suspensions with 0, 5, 10 and 20 g/L CB particles (HA: 20 g/L). As it is expected the amount of porosity in the coatings increases as the concentration of CB particles increases in the suspension. When the sacrificial template is not used to generate porosity in EPD coatings, the pore size distribution and the amount of porosity which can be obtained are limited; the mean pore size is usually smaller and the amount of porosity is lower for this case depending on the suspension properties (concentration, zeta potential of particles and their size distribution), EPD parameters (voltage and time) and sintering conditions (temperature and time).8,31–34 While when sacrificial template is used, the pore size distribution and the amount of porosity in EPD coatings can be controlled by controlling the particle size distribution of sacrificial template particles and their concentration in the suspension, respectively.
M. Farrokhi-Rad et al. / Journal of the European Ceramic Society 34 (2014) 97–106
105
Fig. 10. SEM images of the coatings deposited at 60 V and 30 s from the suspensions with 20 g/L HA and different concentration of CB particles: 0 g/L (a–c), 5 g/L (d–f), 10 g/L (g–i) and 20 g/L (j–l) before sintering (a, d, g and j) and after sintering at 700 ◦ C for 1 h (b, e, h and k). (c, f, i and l) are the results obtained from the image analyzing of (b, e, h and k), respectively.
4. Conclusions Hydroxyapatite (HA) coatings with controlled porosity were prepared by electrophoretic deposition (EPD) method. Carbon black (CB) particles were used as the porogen agent. The suspensions were prepared at different concentrations of HA (10 and 20 g/L) and CB particles (WR: 0, 0.25, 0.5 and 1) in isopropanol. It was found that the finer and positively charged HA nanoparticles are heterocoagulated on the coarser and negatively charged CB particles and form CB–HA composite particles with net positive charge. The electrical conductivity of the suspensions
increases with CB concentration. The conductivity increases more steeply at the CB concentrations higher than 10 and 20 g/L for the suspensions with 10 and 20 g/L HA nanoparticles, respectively. It was fund that there is a concentration of ∗ ) where all the HA nanoparticles are heteCB particles (CCB ∗ were 4 and rocoagulated on their surfaces. The values of CCB 8 g/L for the HA nanoparticles concentrations of 10 and 20 g/L, respectively. The current density during EPD increased at 200 V for the suspensions with WR: 0.5 due to the lower resistivity of the deposits than that of the suspensions. The curve of in situ deposition weight versus time showed good linearity in case of
106
M. Farrokhi-Rad et al. / Journal of the European Ceramic Society 34 (2014) 97–106
suspensions with WR: 0.5. The amounts of porosity were 15%, 24%, 31%, 43% for the coatings deposited from the suspensions with 0, 5, 10 and 20 g/L CB particles (HA: 20 g/L). Acknowledgment The authors would like to thank Ms. Najibeh Tolu Farrokh for her valuable helps on this work. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/ j.jeurceramsoc.2013.07.022. References 1. Velev OD, Jode TA, Lobo RF, Lenhoff AM. Porous silica via colloidal crystallization. Nature 1997;389:447–8. 2. Studart AR, Gonzenbach UT, Tervoort E, Gauckler LJ. Processing routes to macroporous ceramics: a review. J Am Ceram Soc 2006;89:1771–89. 3. Jiang P, Ostojic GN, Narat R, Mittleman DM, Covin VL. The fabrication and bandgap engineering of photonic multilayers. Adv Mater 2001;13:389–93. 4. Wijnhoven JEGJ, Zevenhuizen SJM, Hendriks MA, Livos V, Vanmaekelbergh D. Electrochemical assembly of ordered macropores in gold. Adv Mater 2000;12:888–90. 5. Yan H, Zhang K, Blanford CF, Francis LF, Stein A. In vitro hydroxycarbonate apatite mineralization of CaO–SiO2 sol–gel glasses with a three-dimensionally ordered macroporous structure. Chem Mater 2001;13:1374–82. 6. Hamagami Ji Ato Y, Kanamura K. Fabrication of highly ordered macroporous apatite coating onto titanium by electrophoretic deposition method. Solid State Ion 2004;172:331–4. 7. Yousefpour M, Afshar A, Chen J, Zhang X. Electrophoretic deposition of porous hydroxyapatite coatings using polytetrafluoroethylene particles as templates. Mater Sci Eng C 2007;27:1482–6. 8. Ma J, Wang C, Peng KW. Electrophoretic deposition of porous hydroxyapatite scaffold. Biomaterials 2003;24:3505–10. 9. Radice S, Dietsch H, Mischler S, Michler J. Electrophoretic deposition of functionalized polystyrene particles for TiO2 multi-scale structured surfaces. Surf Coat Technol 2010;204:1749–54. 10. Neirinck BB, Mattheys T, Braem A, Fransaer J, Van der Biest O, Vleugels J. Porous titanium coatings obtained by electrophoretic deposition. Adv Eng Mater 2008;10:246–9. 11. Neirinck B, Fransaer J, Van der Biest O, Vleugels J. A novel route to produce porous ceramics. J Eur Ceram Soc 2009;29:833–6. 12. Besra L, Liu M. A review on fundamental and applications of electrophoretic deposition. Prog Mater Sci 2007;52:1–61. 13. Narasaraju TSB, Phebe DE. Review: some physico-chemical aspects of hydroxylapatite. J Mater Sci 1996;31:1–21. 14. Gomez-Vega JM, Saiz E, Tomsia AP, Marshall GW, Marshall SJ. Bioactive glass coatings with hydroxyapatite and Bioglass® particles on Ti-based implants. 1. Processing. Biomaterials 2000;21:105–11.
15. Ducheyne P, Hench LL, Kagan II A, Martens M, Bursens A, Mulier JC. Effect of hydroxyapatite impregnation on skeletal bonding of porous coated implants. J Biomed Mater Res 1980;14:225–37. 16. Tampieri A, Celotti G, Sprio S, Mingazzini C. Characteristics of synthetic hydroxyapatites and attempts to improve their thermal stability. Mater Chem Phys 2000;64:54–61. 17. Kwok CT, Wong PK, Cheng FT, Man HC. Characterization and corrosion behavior of hydroxyapatite coatings on Ti6Al4V fabricated by electrophoretic deposition. Appl Surf Sci 2009;255: 6736–44. 18. Xiao XF, Liu RF. Effect of suspension stability on electrophoretic deposition of hydroxyapatite coatings. Mater Lett 2006;60:2627–32. 19. Wei M, Ruys AJ, Milthorpe BK, Sorrell CC. Solution ripening of hydroxyapatite nanoparticles: effects on electrophoretic deposition. J Biomed Mater Res 1999;45:11–9. 20. Farrokhi-Rad M, Shahrabi T. Effect of triethanolamine on the electrophoretic deposition of hydroxyapatite nanoparticles in isopropanol. Ceram Int 2013;39:7007–13. 21. Kuche Loghmani S, Farrokhi-Rad M, Shahrabi T. Effect of polyethylene glycol on the electrophoretic deposition of hydroxyapatite nanoparticles in isopropanol. Ceram Int 2013;39:7043–51. 22. Cook SD, Thomas KA, Delton JE, Volkman TK, Whitecloud III ThS, Key JF. Hydroxylapatite coating of porous implants improves bone ingrowth and interface attachment strength. J Biomed Mater Res 1992;26: 989–1001. 23. Kienapfel H, Sprey C, Wilke A, Griss P. Implant fixation by bone ingrowth. J Arthroplasty 1999;14:355–68. 24. Hench LL. Bioceramics. J Am Ceram Soc 1998;81:1705–28. 25. Nguyen HQ, Deporter DA, Pilliar RM, Valiquette N, Yakubovich R. The effect of sol–gel-formed calcium phosphate coatings on bone ingrowth and osteoconductivity of porous-surfaced Ti alloy implants. Biomaterials 2004;25:865–76. 26. Wei M, Ruys AJ, Milthorpe BK, Sorrell CC. Precipitation of hydroxyapatite nanoparticles: effects of precipitation method on electrophoretic deposition. J Mater Sci: Mater Med 2005;16:319–24. 27. Farrokhi-Rad M, Shahrabi T. Electrophoretic deposition of titania nanoparticles: sticking parameter determination by an in-situ study of the EPD kinetics. J Am Ceram Soc 2012;95:3434–40. 28. Bohmer M. In situ observation of two-dimensional clustering during electrophoretic deposition. Langmuir 1996;12:5747–50. 29. Solomentsev Y, Bohmer M, Anderson JL. Particle clustering and pattern formation during electrophoretic deposition: a hydrodynamic model. Langmuir 1997;13:6058–68. 30. Yeh SR, Seul M, Shraiman BI. Assembly of ordered colloidal aggregates by electric-field-induced fluid flow. Nature 1997;386: 57–9. 31. Ji C, Lan W, Xiao P. Fabrication of yttria-stabilized zirconia coatings using electrophoretic deposition: packing mechanism during deposition. J Am Ceram Soc 2008;91:1102–9. 32. Hamaker HC. Formation of a deposit by electrophoresis. Trans Faraday Soc 1940;35:279–87. 33. Tabellion J, Clasen R. Electrophoretic deposition from aqueous suspensions for near-shape manufacturing of advanced ceramics and glasses-applications. J Mater Sci 2004;39:803–11. 34. Chen CY, Chen SY, Liu DM. Electrophoretic deposition forming of porous alumina membranes. Acta Mater 1999;47:2717–26.