- Email: [email protected]
Preliminary study of hydroxyapatite coatings synthesis using solution precursor plasma spraying
Surface & Coatings Technology 277 (2015) 242–250
Contents lists available at ScienceDirect
Surface & Coatings Technology journal homepage: www.elsevier.com/locate/surfcoat
Preliminary study of hydroxyapatite coatings synthesis using solution precursor plasma spraying Rolando T. Candidato Jr. a, Paweł Sokołowski a,b, Lech Pawłowski a,⁎, Alain Denoirjean a a b
SPCTS, UMR 7315, 12, rue Atlantis, 87068 Limoges, France Faculty of Mechanics, Wrocław University of Technology, 50-371 Wrocław, Poland
a r t i c l e
i n f o
Article history: Received 17 April 2015 Revised 17 July 2015 Accepted in revised form 21 July 2015 Available online 27 July 2015 Keywords: Solution precursor plasma spray Hydroxyapatite Biomedical coatings
a b s t r a c t The liquid precursors of calcium hydroxide and diammonium hydrogen phosphate were injected into the plasma jet generated by the SG-100 torch to synthesize the hydroxyapatite (HA) coatings. Three operational deposition process parameters, namely: (i) electric power, (ii) spray distance, and (iii) scan speed were varied. The process enabled synthesization of coatings with HA as major phase onto stainless steel substrates. The coatings were deposited to reach the thickness of about 50 μm with high rate ranging from 3 μm to nearly 7 μm by pass of torch. The X-ray diffraction (XRD) analysis of coatings enabled finding of HA accompanied by calcium phosphates, calcium oxide, and calcium carbonate. The presence of the carbonates was confirmed by Fourier Transform Infrared (FTIR) spectroscopy and by elemental mapping made with the use of Electron Dispersive X-ray Spectroscopy (EDS). The morphology of coatings, observed using scanning electron microscope (SEM), revealed fine-grained microstructure and porosity in the range of 1.3 to 5.1%. The adhesion of coatings obtained using scratch test characterized by critical force was in the range of 2.5 to 3.6 N. © 2015 Elsevier B.V. All rights reserved.
1. Introduction Solution precursor thermal spraying is a relatively new deposition technique used for the first time in the end of the 90s of the last century to synthesize alumina coating with the use of aluminum nitrate solution in isopropanol injected into a combustion flame [1]. The application of plasma jet instead of combustion flame enabled the development of the solution precursor plasma spraying (SPPS) processes. The process has been used in recent years to synthesize such oxides as e.g. yttria stabilized zirconia [2], yttrium iron garnet [3] or cobalt ferrite [4]. The important issues related to the synthesis using solution precursor thermal spraying process were reviewed recently [5,6]. One of the particularly interesting oxides is hydroxyapatite (HA, Ca10(PO4)6(OH)2) which is used to obtain the bioactive coatings onto metallic implants for orthopedic and dental applications. The development of the SPPS method is an important step forward in the development of hydroxyapatite coatings. It is well known that the industrial method for the development of HA is via conventional powder plasma spraying which uses powders having 100 μm in size [7]. Another way of HA coating deposition started with milling of coarse HA powder to obtain fine particles with a few micrometer mean size. Such fine powder was used to formulate a water suspension, which was applied to a process called suspension plasma spraying (SPS) to obtain coating [8]. The coatings obtained using APS and SPS processes included some phases of the HA decomposition, ⁎ Corresponding author. E-mail address: [email protected] (L. Pawłowski).
http://dx.doi.org/10.1016/j.surfcoat.2015.07.046 0257-8972/© 2015 Elsevier B.V. All rights reserved.
namely tricalcium phosphate (TCP,Ca3(PO4)2), tetracalcium phosphate (TTCP, Ca4P2O9) and calcium oxide (CaO). The phases of decomposition are less bioactive that HA and it is interesting to find a process enabling deposition of coatings composed of crystalline HA. The SPPS may be such a process. The major motivation behind the development of the SPPS is the reduction of the tedious process of powder feedstock preparation and/or suspension formulation. These processes can be eliminated since the chemically precipitated HA aqueous solution is fed directly to the plasma. Another advantage of this method is that during wet chemical precipitation, morphology and size of the resulting precipitates can be controlled and the homogeneity at molecular level can be exploited. The synthesis of HA coatings was tested with the use of SPPS method by Garcia et al. [9] and Huang et al. [10]. Garcia and his co-workers used the organic route to prepare Ca-P sol–gel solution according to the method proposed by Liu et al. [11] in which calcium nitrate tetrahydrate and triethyl phosphite were used as calcium and phosphorus precursors. Huang and his colleagues [10] used calcium nitrate Ca(NO3)2 and diammonium hydrogen phosphate (NH4)2HPO4 to obtain HA following the reaction: 10CaðNO3 Þ2 þ 6ðNH4 Þ2 HPO4 þ 8NH4 OH→Ca10 ðPO4 Þ6 ðOHÞ2 þ 20NH4 NO3 þ 6H2 O:
ð1Þ The present paper is a continuation of the previous studies of our research group in which a powder synthesized with the use of reaction shown by Eq. (1) was used to formulate a suspension, applied to plasma
R.T. Candidato Jr. et al. / Surface & Coatings Technology 277 (2015) 242–250
2.2. Plasma spraying of coatings
Table 1 Solution precursor plasma spray parameters. Working gas composition and flow rate, sLpm
Ar + H2 and 45 + 5 sLpm
Solution feed rate, mL/min Type of injection
35 mL/min Nozzle inside torch having 0.3 mm internal diameter oriented radially at 90° relative to the plasma jet axis. 0.5 2 passes per shot with 9 s interruption after each pass. Each shot is hold to allow the samples to cool down to ~50 °C before performing the next shots of spraying. Rectangular patterns with offset distance of 3 mm after each torch run.
Injection pressure, bar Number of scans
Scan pattern
243
spray of coatings [12,13]. However, it was decided to use calcium hydroxide instead of calcium nitrate to simplify the reaction which does not need any pH control. Moreover, the liquid precursor was, after aging, injected directly to plasma jet to form coatings. The present study has a preliminary character and should show a feasibility of synthesis of HA coatings of acceptable quality from the used precursors. 2. Experimental methods 2.1. Preparation of liquid precursor The wet chemical precipitation method was employed for the preparation of the hydroxyapatite liquid precursor using calcium hydroxide (ACS Reagent 95%, Sigma Aldrich) and diammonium hydrogen phosphate (ACS Reagent 98%, Sigma Aldrich) as calcium and phosphorous ions sources respectively. Consideration of the above-mentioned precursors was based on the fact that, calcium hydroxide is relatively cheap and, moreover, the addition of ammonium hydroxide to control pH is not necessary for such preparation. In order to achieve the stoichiometric Ca/P ratio of 1.67, 150 mL of 0.3 M of diammonium hydrogen phosphate aqueous solution was added drop-wise into a stirred 250 mL of 0.5 M calcium hydroxide aqueous solution (pH = 12.4). For comparison, pH = 11 in the paper of Huang et al. [10] to obtain HA coatings. The mixed solution precursor was then magnetically stirred and heated at bath temperature of 70 °C for 3 h in a reaction vessel. The prepared liquid precursor was then aged for at least 24 h in ambient temperature and was submitted to an ultrasonic treatment prior to injection into plasma jet. During the wet chemical precipitation of HA solution, the particles that started to be formed were amorphous and aging of this solution induced crystallization of the precipitates because of the chemically active environment of the solution. The solution precursors were supposed to react in the plasma jet as follows: 10CaðOHÞ2 þ 6ðNH4 Þ2 HPO4 →Ca10 ðPO4 Þ6 ðOHÞ2 þ 12NH3 þ 18H2 O ð2Þ
The plasma torch SG-100 (Praxair S.T., Indianapolis, IN, USA) mounted on 5-axis IRB-6 robot (ABB, Zürich, Switzerland) was employed for the deposition of HA liquid precursors. The injection of HA liquid precursor was realized using a mechanical injector installed inside the anode-nozzle of the torch. The liquid precursor was submitted to a magnetic stirring while delivering to plasma jet from a solution container. The operational plasma spray parameters, kept constant during all the coating experiments, are listed on Table 1. The variable parameters used in the experimental runs are shown in Table 2. The coatings were sprayed onto stainless steel 316 having diameter 25 mm and thickness 5 mm and Ti plates having size 20 × 20 × 1 mm. The temperature profile during each spraying run was monitored using Impac IN 5 Pyrometer (LumaSense Technologies, Santa Clara, CA, USA). The substrates were prepared using sand alumina blasting prior to spraying. 2.3. Characterizations of liquid precursor and of coatings The precursor was characterized with Fourier Transform Infrared Spectroscopy (FTIR) with the use of spectrum one set-up of Perkin Elmer (Waltham, MA, USA). The precursor was dried, before testing, at 100 °C to obtain fine powder and grinded together with 0.1 g of potassium bromide. Functional groups of the obtained coatings were also determined by means of Attenuated Total Reflectance-Fourier Transform Infrared Spectroscopy (ATR-FTIR) using Nicolet 6700 (Thermofisher) spectrometer at scan range of 400–4000 cm−1 with 4 cm−1 resolution over 32 scans. The thermal behavior of the precursor was tested by thermogravimetric analysis and differential scanning calorimetry (TG-DSC) with the use of Labsys set-up of Setaram (Caluire, France). The fine powder formed in liquid precursor prior to spraying was tested using light particle size analyzer type Zetasizer Nano ZS of Malvern Instruments .(Malvern, UK). Scanning electron microscope (SEM) Philips XL30 (Eindhoven, Netherlands) coupled with electron dispersion spectrometer (EDS) was used to characterize the splats. The structural characteristics of fine powder and coatings were determined using X-ray diffraction (XRD) using Bruker D8 Advance diffractometer (Billerica, MA, USA) under Bragg-Brentano configuration with Cu Kα radiation. The resulting diffractogram was analyzed using Diffrac + EVA Software equipped with JCPDS-ICDD database. The following standards were used for phases' identification:
• • • • • •
Hydroxyapatite (HA) — JCPDS 00-009-0432; Calcium phosphate (α-TCP) — JCPDS 00-009-0348; Calcium phosphate (β-TCP) — JCPDS 00-009-0169; Tetracalcium phosphate (TTCP) — JCPDS 00-25-1137; Calcium oxide (CaO) — JCPDS 01-082-1690; Calcium carbonate (CaCO3) — JCPDS 00-017-0763.
Table 2 Experimental design of SPPS experiments. Run no.
Power input (kW)
Spray distance (mm)
Scan speed (mm/s)
No of torch passes over the substrate
Totala thickness, μm
Thickness in one pass, μm/pass
Maximum temperature of coatings, °C
Coating porosity, %
Critical load, N
1 2 3 4 5 6 7 8 9
36 36 36 36 40 40 40 40 38
60 60 80 80 60 60 80 80 70
400 600 400 600 400 600 400 600 500
22 17 10 16 14 12 10 12 12
63 60 66 58 65 54 51 51 59
2.9 3.5 6.6 3.6 4.6 4.5 5.1 4.3 4.9
671 670 617 692 617 564 836 465 897
1.6 1.3 2.2 1.6 2.1 2.3 5.1 4.7 4.6
3.3 2.5 3.6 3.0 2.7 2.7 2.9 2.5 2.8
a
The smallest value taken from 5 measurements made using an optical microscope on the coatings' cross-sections.
244
R.T. Candidato Jr. et al. / Surface & Coatings Technology 277 (2015) 242–250
Fig. 1. Thermogravimetric analysis (TGA) and differential scanning calorimetry (DSC) diagrams vs. temperature of liquid precursor used to spray.
The method, described in details by Prevey [14], was used to carry out the quantitative estimation of phases' content. The method uses external standard (corundum) and Diffrac Eva software to find the relative intensity ratio between the standard and calcium compounds present in coatings. The porosities of the sprayed coatings were determined using their metallographic cross-sections using the software ImageJ version 1.49 [15]. Finally, the adhesion of coatings was tested using scratch test described in details elsewhere [16]. The scratch test was performed using Revetest Scratch Tester of CSM Instruments (Peseux, Switzerland). The scratches were made with a 200 μm radius Rockwell diamond indenter. The scratch length was 10 mm using an applied load from 1 N linearly increased to 10 N with a speed of 1 N/min. The measurements were checked with the optical microscope which
enabled to find out the length of the scratch. The test was conducted on the as-sprayed coatings. The critical load was obtained basing on the first crack observed on the optical microscope image and was directly measured by looking at the length where the first crack was observed to the corresponding applied load. 3. Results 3.1. Liquid precursor characterization Thermal analysis of liquid precursor is shown in Fig. 1. A small exothermic peak starting at about 250 °C can be attributed to the dehydration of the samples and the peak at about 500 °C can be explained by
Fig. 2. Fourier transform infrared spectrum (FTIR) of liquid precursor dried at 100 °C and ground together with a small quantity of KBr.
R.T. Candidato Jr. et al. / Surface & Coatings Technology 277 (2015) 242–250
245
Fig. 3. Granulometry of fine HA particles included in the liquid precursor.
Fig. 4. SEM image (secondary electrons) of the microstructural features, i.e. cracked shells and small spheres observed after one pass of plasma torch working with parameters corresponding to run # 1, a); cracked shells observed at higher magnification b).
the crystallization of HA. Finally, the loss of mass of about 9% can be observed at the temperature rise from room temperature to RT to 1 200 °C. The XRD analysis of dried at 100 °C precursor indicated that it is composed of mainly crystalline hydroxyapatite and the FTIR analysis confirms this finding (Fig. 2). The observed peaks in FTIR spectrum were attributed to as follows:
presence of calcium hydroxide, which could have reacted with CO2 present in air. The liquid precursor includes already solid particles which can be attributed to HA.1 The granulometry of these particles, shown in Fig. 3, indicates that the sizes of fine HA are centered on 360 nm and on 2.4 μm.
• at 567 cm−1 to bending vibrational mode of phosphate group PO−3 4 ; • at 963 and 1 036 cm−1 to symmetric and antisymmetric vibration of phosphate group PO−3 4 ; • at 637 and 3 570 cm−1 to group OH−; • 1 400 to 1 450 cm−1 are attributed to the carbonate group CO−2 3 .
3.2. Observation of splats
The presence of PO−3 and OH− groups confirms the formation of 4 crystalline HA. The presence of carbonates may have resulted from the
The morphology of the splats sprayed in one pass of the torch over the substrate is shown in Fig. 4. The visible features are the small spherical particles and some cracked shells. The presence of these features is 1 Our coating process was in fact partly solution precursor plasma spraying (SPPS) and partly suspension plasma spraying (SPS). As the solution liquid was composed of precursor we preferred to stay with the name of SPPS.
246
R.T. Candidato Jr. et al. / Surface & Coatings Technology 277 (2015) 242–250
in which process variables noted in Table 2 correspond to X1 — electric power, X2 — spray distance, and X3 — scan speed: P ¼ 2:6 þ 0:94X 1 þ 0:79X 2 0:14X 3 þ 0:57X 1 X 2 þ 0:09X 1 X 3 −0:12X 2 X 3 0:04X 1 X 2 X 3
Fig. 5. SEM image (secondary electrons) of the surface of coating sprayed in run # 3.
ð3Þ
Consequently, porosity depends strongly on electric power and spray distance and on their interaction effects. Inversely, the plasma torch scan speed does not influence this coatings' property. Finally, the 2-D distribution of the elements obtained thanks to EDS such as O, P, Ca and C is shown in Fig. 7. The elements are distributed in a homogeneous way inside of coating. The high concentration of carbon present outside the coating's surface can be attributed to the epoxy resin in which the sample was embedded prior to metallographic preparation. The results of the scratch test are shown in Table 2. The values of critical load varied from 2.7 to 3.5 N. The best adherence had the samples sprayed in the runs # 3 and in run # 4. Total delamination of the coating was observed as the load increases. 3.4. X-ray diffraction and FTIR analyses
characteristic for the coatings thermally sprayed using solutions. The cracked shells were visible in many other places on the sprayed substrate. Their elemental composition made using EDS technique indicate the presence in at. % of: 57.1 of O; 21 of Au (sputtered at sample preparation stage); 11.3 of Ca; 6.4 of P; and 4.2 of C. The Ca/P ratio is about 1.36 being clearly smaller than the stoichiometric ratio of 1.67 for HA. 3.3. Coating morphology, element distribution and adhesion to substrate Typical surface of coatings indicates the finely grained microstructure (Fig. 5). Some small spherical particles and cracked shells similar to that observed in Fig. 4 are also present. The observations of the cross-sections of the sprayed coatings reveal dense structure with some pores and some vertical cracks more visible in samples sprayed in run # 7 than in run # 3 (Fig. 6). The cracks could have resulted from high temperature of 836 °C occurred at spraying in run # 7 (see Table 2). The porosities of all sprayed coatings were relatively low. The lowest one, of about 1.3% was found in coatings sprayed in run # 2 and the greatest one, of about 5% in the coatings sprayed in run # 7. The experiments were made in a statistical way using full factorial 23 design following the method described elsewhere, see e.g. [17]. Porosity (P) of the coatings can be expressed by a following regression equation
The X-ray diffraction diagram of dried precursor shows the presence of HA as the main crystal phase (Fig. 8a). The as-sprayed coating includes HA, the calcium phosphates as α — TCP, β — TCP and TTCP and calcium carbonate. The peaks in as-sprayed coatings are slightly shifted towards lower 2θ angles which indicate the presence of tensile stresses in the coatings. These stresses could have resulted in the cracking in the coatings observed in the coatings (Fig. 6b). The post spray heating at 600 °C carried out for resulted in an increase of the content of crystalline HA and in formation of CaO (Fig. 8c). The estimation of quantitative phases' content in samples sprayed in different experimental runs in presented in Table 3. The greatest content of crystalline HA of about 70 wt.% had the sample sprayed in run # 3 and quite similar value of 69 wt.% had the sample sprayed in run # 4. Thus, the optimal spray parameters enabling to reach high HA content included electric power input of 36 kW and spray distance of 80 mm. These spray conditions enabled also to preserve low content of calcium carbonate being about 5 wt.%. The samples sprayed in other experimental runs had the content of the carbonate of 20 wt.% and more. The XRD analysis was confirmed by the FTIR one shown for the samples sprayed with the use of 36 kW in Fig. 9. Consequently, the samples sprayed from the distance of 80 mm ion. have much smaller content of CO−2 3
Fig. 6. SEM image (secondary electrons) of the cross-sections of samples: sprayed in run # 3, a); and in run # 7, b).
R.T. Candidato Jr. et al. / Surface & Coatings Technology 277 (2015) 242–250
247
Fig. 7. EDS mapping of element 2-D distribution in the coatings sprayed in run #7 (shown in Fig. 6b): O, a); P, b); Ca, c); and C, d).
Fig. 8. X-ray diffraction diagrams of dried precursor (dotted line); sample as-sprayed in run # 3 (dashed line); and the same sample heat treated at temperature 600 °C for 1 h in open air (continuous line).
248
R.T. Candidato Jr. et al. / Surface & Coatings Technology 277 (2015) 242–250
Table 3 Estimation of crystal phases' content in wt. % in the samples sprayed in different experimental runs. Run no.
HA
α-TCP
β-TCP
TTCP
CaO
CaCO3
As-sprayed 1 2 3 4 5 6 7 8
62.1 59.5 70.9 69.4 58.5 59.5 57.0 58.2
5.7 5.6 6.7 7.4 6.4 5.6 6.8 7.2
4.7 3.2 8.9 9.0 4.6 3.2 4.4 4.8
7.0 4.4 8.5 9.1 4.5 4.4 4.1 4.2
0.5 0.3 0.2 0.2 0.5 0.3 0.4 0.3
20.0 27.0 4.8 4.9 25.5 27.0 27.3 25.3
3.7 6.2
2.6 4.9
4.6 4.2
7.2 0.8
14.1 3.4
Post-spray heat treated 1 67.8 3 80.5
4. Discussion The liquid precursor used to spray contained already some small solid HA particles with the size of a few micrometers (Fig. 3). The liquid was injected into plasma torch and sprayed onto the substrate. The observation of the morphology of the particles impacting the substrate enabled to visualize the presence of small spheres and cracked shells. The possible routes leading to the formation of these features is shown in Fig. 10. The physical phenomena associated with routes shown in the figure can be described as follows [18–20]: a) Water evaporates from the surface of liquid droplet and the solution gets more concentrated. The super-saturation precedes the homogenous nucleation and formation of solid shell of HA around the droplet. The shell may crack when the pressure of vapors becomes too high. The cracked shell may fall directly on the substrate or get molten and takes a spherical shape before impact. The molten particles could evaporate from their surface what reduced their size. b) Similar phenomena could have happened when liquid droplet contains a small solid precursor. Additionally, the heterogeneous nucleation was possible on the surface of small solid inside the droplet surrounded by the liquid inside solid shell. This small solid might get molten before the contact with substrate. c) Droplet having small volume of liquid around the solid particle would be, probably, submitted to evaporation of liquid. The remaining solid
would get molten before the contact with the substrate. Some evaporation from the molten particle could have also occurred. The chemical phenomena occurring in liquid precursor include the formation of HA following Eq. (2). HA is formed when precursor reaches temperature of about 500 °C (see Fig. 1). At lower temperatures the liquid calcium hydroxide may however react also with CO2 contained in open atmosphere to give the calcium carbonate: CaðOHÞ2ðlÞ þ CO2ðgÞ →CaCO3 ðsÞ þ H2 OðlÞ
ð4Þ
anions was indeed observed in liquid precurThe presence of CO−2 3 sor (see Fig. 2) and in all sprayed coating (sees Fig. 9). The coatings sprayed with the use of electric power of 36 kW and spray distance of 80 mm showed the lowest content of these ions and, consequently, the lowest content of carbonate. The carbonate decomposes at the temperature 500–600 °C to give solid calcium oxide (visible in Fig. 8c in heat treated coatings) and carbon dioxide in gaseous phase following the equation: CaCO3ðsÞ
500−600 C
→
CaOðsÞ þ CO2ðgÞ
ð5Þ
The operational deposition process parameters were varied to find their influence on morphology and chemistry of obtained coatings. An important characteristics of process was very high growth rate of coating which reached nearly 7 μm/pass in run # 3 (see Table 1). Such high growth rate renders the process interesting from the point of view of process application in industry. Before such application will take place, it is necessary to understand the origins and, eventually, to reduce the content of calcium carbonate in the coating. Knowing that the formation of the carbonate takes place when calcium hydroxide is liquid (see Eq. (3)) and in contact with gaseous carbon dioxide being present in air, the trajectory of liquid precursor being close to the jet periphery could have been be favorable to this reaction (see Fig. 11). Possibly, the solution injected to the plasma generated by the torch alimented with 40 kW was having such a trajectory (runs # 5 to 8) and the power input of 36 kW seems to result in the trajectory closer to jet axis. On the other hand, at the latter input power, the spray distance of 80 mm and low torch linear speed of 400 mm/s could have resulted in better evaporation of remaining calcium hydroxide and reduction of calcium carbonate content (see Fig. 9). Consequently, the
Fig. 9. Fourier transform infrared spectra of samples sprayed using electric power of 36 kW with different spray distances and different linear torch velocities.
R.T. Candidato Jr. et al. / Surface & Coatings Technology 277 (2015) 242–250
249
Fig. 10. Possible routes for the initial droplets: entirely liquid, a); liquid with a small solid inside, b); solid with some liquid around, c).
coatings were relatively dense without large lamellas and some porosity. The adhesion of obtained coatings tested using scratch test was to be in the range of 2.5 to 3.6 N. The future works should lead to the reduction of the content of the calcium carbonate in the coatings. This reduction may be reached by further optimizing the operational process parameters or by changing the used precursor. Similarly the adhesion of synthesized coatings seems to be low and the way should be found to increase it. The optimized coatings will be submitted to the corrosion tests in the body simulated fluid. Finally, the numerical simulation of process should help in better understanding of physical and chemical phenomena occurring at deposition. Acknowledgments
Fig. 11. The possible influence of the trajectories of liquid precursor in plasma jet on their transformation.
run # 3 seems to be the closest to the optimal one. Finally, it is important to mention the role of the substrate temperature at spraying. The high temperature reached in run # 7 could have led to cracking of the coating visible in Fig. 6b. On the other hand, the crystal phase composition of analyzes coatings (see Table 3) was not correlated with this temperature although it is well known that decomposition of HA is promoted by high temperatures. 5. Conclusions The liquid precursor of calcium hydroxide and diammonium hydrogen phosphate was applied to synthesize the coatings with the use of jet generated by arc plasma torch. The operational parameters of deposition, namely electric power input to the torch, spray distance and scan speed were varied. The coatings grew up with high rate ranging from 3 μm to nearly 7 μm by pass of torch. The coatings had finely grained microstructure and the porosity varying from 1 to 5% and were composed of hydroxyapatite, calcium phosphates, calcium oxide and calcium carbonate. The presence of the carbonates was explained by the reaction of liquid calcium hydroxide with carbon dioxide contained in air. The
Drs. C. Goldner — Constantinescu and G.-L. Lecomte helped in experimental part of the study. Dr. B. Pateyron helped with a useful discussion at the redaction of this paper. Dr. Reynaldo M. Vequizo from MSU — Illigan Institute of Technology, Illigan City in Philippines and Ms. Ermie Bacarra from Philippine Council for Industry Energy and Emerging Technologies, Taguig City, Philippines helped in initiation of collaboration with The University of Limoges. References [1] J. Tikkanen, K.A. Gross, C.C. Berndt, V. Pitkänen, J. Keskinen, S. Raghu, M. Rajala, J. Karthikeyan, Characteristics of the liquid flame spray process, Surf. Coat. Technol. 90 (1997) 210–216. [2] S. Govindarajan, R.O. Dusan, S.V. Joshi, In situ particle generation and splat formation during solution precursor plasma spraying of yttria-stabilized zirconia coatings, J. Am. Ceram. Soc. 94 (12) (2011) 4191–4199. [3] X.Z. Guo, B.G. Ravi, P.S. Devi, J.C. Hanson, J. Margolies, R.J. Gambino, J.B. Parise, S. Sampath, Synthesis of yttrium iron garnet (YIG) by citrate–nitrate gel combustion and precursor plasma spray process, J. Magn. Magn. Mater. 295 (2005) 145–154. [4] N. Sanpo, C.C. Berndt, A.S.M. Ang, J. Wang, Effect of the chelating agent contents on the topography, composition and phase of SPPS-deposited cobalt ferrite splats, Surf. Coat. Technol. 232 (2013) 247–253. [5] L. Pawłowski, Suspension and solution thermal spray coatings, Surf. Coat. Technol. 203 (2009) 2807–2829. [6] L. Pawłowski, Application of solution precursor spray techniques to obtain ceramic films and coatings, in: N. Espallagas (Ed.), Future Development of Thermal Spray Coatings, Woodhead, Cambridge, UK 2015, pp. 123–141. [7] V. Deram, C. Minichiello, R.-N. Vannier, A. Le Maguer, L. Pawłowski, D. Murano, Microstructural characterizations of plasma sprayed hydroxyapatite coatings, Surf. Coat. Technol. 166 (2003) 153–159.
250
R.T. Candidato Jr. et al. / Surface & Coatings Technology 277 (2015) 242–250
[8] R. Tomaszek, L. Pawłowski, L. Gengembre, J. Laureyns, A. Le Maguer, Microstructure of suspension plasma sprayed multilayer coatings of hydroxyapatite and titanium oxide, Surf. Coat. Technol. 201 (2007) 7432–7440. [9] B. Garcia, Z.B. Zhang, T.W. Coyle, L. Ga, R. Pillar, Hydroxyapatite coatings produced by plasma spraying of organic based solution precursor, in: M. Mizuno (Ed.), Advances in Bioceramics and Biocomposites II, Wiley, Hoboken, N.J., USA 2007, pp. 103–111. [10] Y. Huang, L. Song, X. Liu, Y. Xiao, Y. Wu, J. Chen, F. Wu, Z. Gu, Hydroxyapatite coatings deposited by liquid precursor plasma spraying: controlled dense and porous microstructures and osteoblastic cell responses, Biofabrication 2 (2010) 1–10. [11] D.M. Liu, T. Troczynski, W.T. Tseng, Water based sol–gel synthesis of hydroxyapatite: process development, Biomaterials 22 (2001) 1721–1730. [12] R. Jaworski, C. Pierlot, L. Pawłowski, M. Bigan, M. Quivrin, Synthesis and preliminary tests of suspension plasma spraying of fine hydroxyapatite powder, J. Therm. Spray Technol. 17 (5–6) (2008) 679–684. [13] R. Jaworski, C. Pierlot, L. Pawłowski, M. Bigan, M. Martel, Design of the synthesis of fine HA powder for suspension plasma spraying, Surf. Coat. Technol. 203 (2009) 1092–1097.
[14] P.L. Prevey, X-ray diffraction characterization of crystallinity and phase composition in plasma sprayed hydroxyapatite coatings, J. Therm. Spray Technol. 9 (3) (2000) 369–376. [15] http://imagej.nih.gov/ij/index.html. [16] R. Jaworski, L. Pawłowski, F. Roudet, S. Kozerski, F. Petit, Characterization of mechanical properties of suspension plasma sprayed TiO2 coatings using scratch test, Surf. Coat. Technol. 202 (2008) 2644–2653. [17] C. Pierlot, L. Pawłowski, M. Bigan, P. Chagnon, Design of experiments in thermal spraying: a review, Surf. Coat. Technol. 202 (2008) 4483–4490. [18] S. Basu, B.M. Cetegen, Modeling of thermo-physical processes in liquid ceramic precursor droplets injected into a plasma jet, Int. J. Heat Mass Transf. 50 (2007) 3278–3290. [19] A. Saha, S. Seal, B. Cetegen, E. Jordan, A. Ozturk, S. Basu, Thermo-physical processes in cerium nitrate precursor droplets injected into high temperature plasma, Surf. Coat. Technol. 203 (2009) 2081–2091. [20] B. Pathak, S. Basu, R. Kumar, Heat and mass transfer and chemical transformation in a cerium nitrate droplet, Int. J. Heat Mass Transf. 63 (2013) 301–312.
Contents lists available at ScienceDirect
Surface & Coatings Technology journal homepage: www.elsevier.com/locate/surfcoat
Preliminary study of hydroxyapatite coatings synthesis using solution precursor plasma spraying Rolando T. Candidato Jr. a, Paweł Sokołowski a,b, Lech Pawłowski a,⁎, Alain Denoirjean a a b
SPCTS, UMR 7315, 12, rue Atlantis, 87068 Limoges, France Faculty of Mechanics, Wrocław University of Technology, 50-371 Wrocław, Poland
a r t i c l e
i n f o
Article history: Received 17 April 2015 Revised 17 July 2015 Accepted in revised form 21 July 2015 Available online 27 July 2015 Keywords: Solution precursor plasma spray Hydroxyapatite Biomedical coatings
a b s t r a c t The liquid precursors of calcium hydroxide and diammonium hydrogen phosphate were injected into the plasma jet generated by the SG-100 torch to synthesize the hydroxyapatite (HA) coatings. Three operational deposition process parameters, namely: (i) electric power, (ii) spray distance, and (iii) scan speed were varied. The process enabled synthesization of coatings with HA as major phase onto stainless steel substrates. The coatings were deposited to reach the thickness of about 50 μm with high rate ranging from 3 μm to nearly 7 μm by pass of torch. The X-ray diffraction (XRD) analysis of coatings enabled finding of HA accompanied by calcium phosphates, calcium oxide, and calcium carbonate. The presence of the carbonates was confirmed by Fourier Transform Infrared (FTIR) spectroscopy and by elemental mapping made with the use of Electron Dispersive X-ray Spectroscopy (EDS). The morphology of coatings, observed using scanning electron microscope (SEM), revealed fine-grained microstructure and porosity in the range of 1.3 to 5.1%. The adhesion of coatings obtained using scratch test characterized by critical force was in the range of 2.5 to 3.6 N. © 2015 Elsevier B.V. All rights reserved.
1. Introduction Solution precursor thermal spraying is a relatively new deposition technique used for the first time in the end of the 90s of the last century to synthesize alumina coating with the use of aluminum nitrate solution in isopropanol injected into a combustion flame [1]. The application of plasma jet instead of combustion flame enabled the development of the solution precursor plasma spraying (SPPS) processes. The process has been used in recent years to synthesize such oxides as e.g. yttria stabilized zirconia [2], yttrium iron garnet [3] or cobalt ferrite [4]. The important issues related to the synthesis using solution precursor thermal spraying process were reviewed recently [5,6]. One of the particularly interesting oxides is hydroxyapatite (HA, Ca10(PO4)6(OH)2) which is used to obtain the bioactive coatings onto metallic implants for orthopedic and dental applications. The development of the SPPS method is an important step forward in the development of hydroxyapatite coatings. It is well known that the industrial method for the development of HA is via conventional powder plasma spraying which uses powders having 100 μm in size [7]. Another way of HA coating deposition started with milling of coarse HA powder to obtain fine particles with a few micrometer mean size. Such fine powder was used to formulate a water suspension, which was applied to a process called suspension plasma spraying (SPS) to obtain coating [8]. The coatings obtained using APS and SPS processes included some phases of the HA decomposition, ⁎ Corresponding author. E-mail address: [email protected] (L. Pawłowski).
http://dx.doi.org/10.1016/j.surfcoat.2015.07.046 0257-8972/© 2015 Elsevier B.V. All rights reserved.
namely tricalcium phosphate (TCP,Ca3(PO4)2), tetracalcium phosphate (TTCP, Ca4P2O9) and calcium oxide (CaO). The phases of decomposition are less bioactive that HA and it is interesting to find a process enabling deposition of coatings composed of crystalline HA. The SPPS may be such a process. The major motivation behind the development of the SPPS is the reduction of the tedious process of powder feedstock preparation and/or suspension formulation. These processes can be eliminated since the chemically precipitated HA aqueous solution is fed directly to the plasma. Another advantage of this method is that during wet chemical precipitation, morphology and size of the resulting precipitates can be controlled and the homogeneity at molecular level can be exploited. The synthesis of HA coatings was tested with the use of SPPS method by Garcia et al. [9] and Huang et al. [10]. Garcia and his co-workers used the organic route to prepare Ca-P sol–gel solution according to the method proposed by Liu et al. [11] in which calcium nitrate tetrahydrate and triethyl phosphite were used as calcium and phosphorus precursors. Huang and his colleagues [10] used calcium nitrate Ca(NO3)2 and diammonium hydrogen phosphate (NH4)2HPO4 to obtain HA following the reaction: 10CaðNO3 Þ2 þ 6ðNH4 Þ2 HPO4 þ 8NH4 OH→Ca10 ðPO4 Þ6 ðOHÞ2 þ 20NH4 NO3 þ 6H2 O:
ð1Þ The present paper is a continuation of the previous studies of our research group in which a powder synthesized with the use of reaction shown by Eq. (1) was used to formulate a suspension, applied to plasma
R.T. Candidato Jr. et al. / Surface & Coatings Technology 277 (2015) 242–250
2.2. Plasma spraying of coatings
Table 1 Solution precursor plasma spray parameters. Working gas composition and flow rate, sLpm
Ar + H2 and 45 + 5 sLpm
Solution feed rate, mL/min Type of injection
35 mL/min Nozzle inside torch having 0.3 mm internal diameter oriented radially at 90° relative to the plasma jet axis. 0.5 2 passes per shot with 9 s interruption after each pass. Each shot is hold to allow the samples to cool down to ~50 °C before performing the next shots of spraying. Rectangular patterns with offset distance of 3 mm after each torch run.
Injection pressure, bar Number of scans
Scan pattern
243
spray of coatings [12,13]. However, it was decided to use calcium hydroxide instead of calcium nitrate to simplify the reaction which does not need any pH control. Moreover, the liquid precursor was, after aging, injected directly to plasma jet to form coatings. The present study has a preliminary character and should show a feasibility of synthesis of HA coatings of acceptable quality from the used precursors. 2. Experimental methods 2.1. Preparation of liquid precursor The wet chemical precipitation method was employed for the preparation of the hydroxyapatite liquid precursor using calcium hydroxide (ACS Reagent 95%, Sigma Aldrich) and diammonium hydrogen phosphate (ACS Reagent 98%, Sigma Aldrich) as calcium and phosphorous ions sources respectively. Consideration of the above-mentioned precursors was based on the fact that, calcium hydroxide is relatively cheap and, moreover, the addition of ammonium hydroxide to control pH is not necessary for such preparation. In order to achieve the stoichiometric Ca/P ratio of 1.67, 150 mL of 0.3 M of diammonium hydrogen phosphate aqueous solution was added drop-wise into a stirred 250 mL of 0.5 M calcium hydroxide aqueous solution (pH = 12.4). For comparison, pH = 11 in the paper of Huang et al. [10] to obtain HA coatings. The mixed solution precursor was then magnetically stirred and heated at bath temperature of 70 °C for 3 h in a reaction vessel. The prepared liquid precursor was then aged for at least 24 h in ambient temperature and was submitted to an ultrasonic treatment prior to injection into plasma jet. During the wet chemical precipitation of HA solution, the particles that started to be formed were amorphous and aging of this solution induced crystallization of the precipitates because of the chemically active environment of the solution. The solution precursors were supposed to react in the plasma jet as follows: 10CaðOHÞ2 þ 6ðNH4 Þ2 HPO4 →Ca10 ðPO4 Þ6 ðOHÞ2 þ 12NH3 þ 18H2 O ð2Þ
The plasma torch SG-100 (Praxair S.T., Indianapolis, IN, USA) mounted on 5-axis IRB-6 robot (ABB, Zürich, Switzerland) was employed for the deposition of HA liquid precursors. The injection of HA liquid precursor was realized using a mechanical injector installed inside the anode-nozzle of the torch. The liquid precursor was submitted to a magnetic stirring while delivering to plasma jet from a solution container. The operational plasma spray parameters, kept constant during all the coating experiments, are listed on Table 1. The variable parameters used in the experimental runs are shown in Table 2. The coatings were sprayed onto stainless steel 316 having diameter 25 mm and thickness 5 mm and Ti plates having size 20 × 20 × 1 mm. The temperature profile during each spraying run was monitored using Impac IN 5 Pyrometer (LumaSense Technologies, Santa Clara, CA, USA). The substrates were prepared using sand alumina blasting prior to spraying. 2.3. Characterizations of liquid precursor and of coatings The precursor was characterized with Fourier Transform Infrared Spectroscopy (FTIR) with the use of spectrum one set-up of Perkin Elmer (Waltham, MA, USA). The precursor was dried, before testing, at 100 °C to obtain fine powder and grinded together with 0.1 g of potassium bromide. Functional groups of the obtained coatings were also determined by means of Attenuated Total Reflectance-Fourier Transform Infrared Spectroscopy (ATR-FTIR) using Nicolet 6700 (Thermofisher) spectrometer at scan range of 400–4000 cm−1 with 4 cm−1 resolution over 32 scans. The thermal behavior of the precursor was tested by thermogravimetric analysis and differential scanning calorimetry (TG-DSC) with the use of Labsys set-up of Setaram (Caluire, France). The fine powder formed in liquid precursor prior to spraying was tested using light particle size analyzer type Zetasizer Nano ZS of Malvern Instruments .(Malvern, UK). Scanning electron microscope (SEM) Philips XL30 (Eindhoven, Netherlands) coupled with electron dispersion spectrometer (EDS) was used to characterize the splats. The structural characteristics of fine powder and coatings were determined using X-ray diffraction (XRD) using Bruker D8 Advance diffractometer (Billerica, MA, USA) under Bragg-Brentano configuration with Cu Kα radiation. The resulting diffractogram was analyzed using Diffrac + EVA Software equipped with JCPDS-ICDD database. The following standards were used for phases' identification:
• • • • • •
Hydroxyapatite (HA) — JCPDS 00-009-0432; Calcium phosphate (α-TCP) — JCPDS 00-009-0348; Calcium phosphate (β-TCP) — JCPDS 00-009-0169; Tetracalcium phosphate (TTCP) — JCPDS 00-25-1137; Calcium oxide (CaO) — JCPDS 01-082-1690; Calcium carbonate (CaCO3) — JCPDS 00-017-0763.
Table 2 Experimental design of SPPS experiments. Run no.
Power input (kW)
Spray distance (mm)
Scan speed (mm/s)
No of torch passes over the substrate
Totala thickness, μm
Thickness in one pass, μm/pass
Maximum temperature of coatings, °C
Coating porosity, %
Critical load, N
1 2 3 4 5 6 7 8 9
36 36 36 36 40 40 40 40 38
60 60 80 80 60 60 80 80 70
400 600 400 600 400 600 400 600 500
22 17 10 16 14 12 10 12 12
63 60 66 58 65 54 51 51 59
2.9 3.5 6.6 3.6 4.6 4.5 5.1 4.3 4.9
671 670 617 692 617 564 836 465 897
1.6 1.3 2.2 1.6 2.1 2.3 5.1 4.7 4.6
3.3 2.5 3.6 3.0 2.7 2.7 2.9 2.5 2.8
a
The smallest value taken from 5 measurements made using an optical microscope on the coatings' cross-sections.
244
R.T. Candidato Jr. et al. / Surface & Coatings Technology 277 (2015) 242–250
Fig. 1. Thermogravimetric analysis (TGA) and differential scanning calorimetry (DSC) diagrams vs. temperature of liquid precursor used to spray.
The method, described in details by Prevey [14], was used to carry out the quantitative estimation of phases' content. The method uses external standard (corundum) and Diffrac Eva software to find the relative intensity ratio between the standard and calcium compounds present in coatings. The porosities of the sprayed coatings were determined using their metallographic cross-sections using the software ImageJ version 1.49 [15]. Finally, the adhesion of coatings was tested using scratch test described in details elsewhere [16]. The scratch test was performed using Revetest Scratch Tester of CSM Instruments (Peseux, Switzerland). The scratches were made with a 200 μm radius Rockwell diamond indenter. The scratch length was 10 mm using an applied load from 1 N linearly increased to 10 N with a speed of 1 N/min. The measurements were checked with the optical microscope which
enabled to find out the length of the scratch. The test was conducted on the as-sprayed coatings. The critical load was obtained basing on the first crack observed on the optical microscope image and was directly measured by looking at the length where the first crack was observed to the corresponding applied load. 3. Results 3.1. Liquid precursor characterization Thermal analysis of liquid precursor is shown in Fig. 1. A small exothermic peak starting at about 250 °C can be attributed to the dehydration of the samples and the peak at about 500 °C can be explained by
Fig. 2. Fourier transform infrared spectrum (FTIR) of liquid precursor dried at 100 °C and ground together with a small quantity of KBr.
R.T. Candidato Jr. et al. / Surface & Coatings Technology 277 (2015) 242–250
245
Fig. 3. Granulometry of fine HA particles included in the liquid precursor.
Fig. 4. SEM image (secondary electrons) of the microstructural features, i.e. cracked shells and small spheres observed after one pass of plasma torch working with parameters corresponding to run # 1, a); cracked shells observed at higher magnification b).
the crystallization of HA. Finally, the loss of mass of about 9% can be observed at the temperature rise from room temperature to RT to 1 200 °C. The XRD analysis of dried at 100 °C precursor indicated that it is composed of mainly crystalline hydroxyapatite and the FTIR analysis confirms this finding (Fig. 2). The observed peaks in FTIR spectrum were attributed to as follows:
presence of calcium hydroxide, which could have reacted with CO2 present in air. The liquid precursor includes already solid particles which can be attributed to HA.1 The granulometry of these particles, shown in Fig. 3, indicates that the sizes of fine HA are centered on 360 nm and on 2.4 μm.
• at 567 cm−1 to bending vibrational mode of phosphate group PO−3 4 ; • at 963 and 1 036 cm−1 to symmetric and antisymmetric vibration of phosphate group PO−3 4 ; • at 637 and 3 570 cm−1 to group OH−; • 1 400 to 1 450 cm−1 are attributed to the carbonate group CO−2 3 .
3.2. Observation of splats
The presence of PO−3 and OH− groups confirms the formation of 4 crystalline HA. The presence of carbonates may have resulted from the
The morphology of the splats sprayed in one pass of the torch over the substrate is shown in Fig. 4. The visible features are the small spherical particles and some cracked shells. The presence of these features is 1 Our coating process was in fact partly solution precursor plasma spraying (SPPS) and partly suspension plasma spraying (SPS). As the solution liquid was composed of precursor we preferred to stay with the name of SPPS.
246
R.T. Candidato Jr. et al. / Surface & Coatings Technology 277 (2015) 242–250
in which process variables noted in Table 2 correspond to X1 — electric power, X2 — spray distance, and X3 — scan speed: P ¼ 2:6 þ 0:94X 1 þ 0:79X 2 0:14X 3 þ 0:57X 1 X 2 þ 0:09X 1 X 3 −0:12X 2 X 3 0:04X 1 X 2 X 3
Fig. 5. SEM image (secondary electrons) of the surface of coating sprayed in run # 3.
ð3Þ
Consequently, porosity depends strongly on electric power and spray distance and on their interaction effects. Inversely, the plasma torch scan speed does not influence this coatings' property. Finally, the 2-D distribution of the elements obtained thanks to EDS such as O, P, Ca and C is shown in Fig. 7. The elements are distributed in a homogeneous way inside of coating. The high concentration of carbon present outside the coating's surface can be attributed to the epoxy resin in which the sample was embedded prior to metallographic preparation. The results of the scratch test are shown in Table 2. The values of critical load varied from 2.7 to 3.5 N. The best adherence had the samples sprayed in the runs # 3 and in run # 4. Total delamination of the coating was observed as the load increases. 3.4. X-ray diffraction and FTIR analyses
characteristic for the coatings thermally sprayed using solutions. The cracked shells were visible in many other places on the sprayed substrate. Their elemental composition made using EDS technique indicate the presence in at. % of: 57.1 of O; 21 of Au (sputtered at sample preparation stage); 11.3 of Ca; 6.4 of P; and 4.2 of C. The Ca/P ratio is about 1.36 being clearly smaller than the stoichiometric ratio of 1.67 for HA. 3.3. Coating morphology, element distribution and adhesion to substrate Typical surface of coatings indicates the finely grained microstructure (Fig. 5). Some small spherical particles and cracked shells similar to that observed in Fig. 4 are also present. The observations of the cross-sections of the sprayed coatings reveal dense structure with some pores and some vertical cracks more visible in samples sprayed in run # 7 than in run # 3 (Fig. 6). The cracks could have resulted from high temperature of 836 °C occurred at spraying in run # 7 (see Table 2). The porosities of all sprayed coatings were relatively low. The lowest one, of about 1.3% was found in coatings sprayed in run # 2 and the greatest one, of about 5% in the coatings sprayed in run # 7. The experiments were made in a statistical way using full factorial 23 design following the method described elsewhere, see e.g. [17]. Porosity (P) of the coatings can be expressed by a following regression equation
The X-ray diffraction diagram of dried precursor shows the presence of HA as the main crystal phase (Fig. 8a). The as-sprayed coating includes HA, the calcium phosphates as α — TCP, β — TCP and TTCP and calcium carbonate. The peaks in as-sprayed coatings are slightly shifted towards lower 2θ angles which indicate the presence of tensile stresses in the coatings. These stresses could have resulted in the cracking in the coatings observed in the coatings (Fig. 6b). The post spray heating at 600 °C carried out for resulted in an increase of the content of crystalline HA and in formation of CaO (Fig. 8c). The estimation of quantitative phases' content in samples sprayed in different experimental runs in presented in Table 3. The greatest content of crystalline HA of about 70 wt.% had the sample sprayed in run # 3 and quite similar value of 69 wt.% had the sample sprayed in run # 4. Thus, the optimal spray parameters enabling to reach high HA content included electric power input of 36 kW and spray distance of 80 mm. These spray conditions enabled also to preserve low content of calcium carbonate being about 5 wt.%. The samples sprayed in other experimental runs had the content of the carbonate of 20 wt.% and more. The XRD analysis was confirmed by the FTIR one shown for the samples sprayed with the use of 36 kW in Fig. 9. Consequently, the samples sprayed from the distance of 80 mm ion. have much smaller content of CO−2 3
Fig. 6. SEM image (secondary electrons) of the cross-sections of samples: sprayed in run # 3, a); and in run # 7, b).
R.T. Candidato Jr. et al. / Surface & Coatings Technology 277 (2015) 242–250
247
Fig. 7. EDS mapping of element 2-D distribution in the coatings sprayed in run #7 (shown in Fig. 6b): O, a); P, b); Ca, c); and C, d).
Fig. 8. X-ray diffraction diagrams of dried precursor (dotted line); sample as-sprayed in run # 3 (dashed line); and the same sample heat treated at temperature 600 °C for 1 h in open air (continuous line).
248
R.T. Candidato Jr. et al. / Surface & Coatings Technology 277 (2015) 242–250
Table 3 Estimation of crystal phases' content in wt. % in the samples sprayed in different experimental runs. Run no.
HA
α-TCP
β-TCP
TTCP
CaO
CaCO3
As-sprayed 1 2 3 4 5 6 7 8
62.1 59.5 70.9 69.4 58.5 59.5 57.0 58.2
5.7 5.6 6.7 7.4 6.4 5.6 6.8 7.2
4.7 3.2 8.9 9.0 4.6 3.2 4.4 4.8
7.0 4.4 8.5 9.1 4.5 4.4 4.1 4.2
0.5 0.3 0.2 0.2 0.5 0.3 0.4 0.3
20.0 27.0 4.8 4.9 25.5 27.0 27.3 25.3
3.7 6.2
2.6 4.9
4.6 4.2
7.2 0.8
14.1 3.4
Post-spray heat treated 1 67.8 3 80.5
4. Discussion The liquid precursor used to spray contained already some small solid HA particles with the size of a few micrometers (Fig. 3). The liquid was injected into plasma torch and sprayed onto the substrate. The observation of the morphology of the particles impacting the substrate enabled to visualize the presence of small spheres and cracked shells. The possible routes leading to the formation of these features is shown in Fig. 10. The physical phenomena associated with routes shown in the figure can be described as follows [18–20]: a) Water evaporates from the surface of liquid droplet and the solution gets more concentrated. The super-saturation precedes the homogenous nucleation and formation of solid shell of HA around the droplet. The shell may crack when the pressure of vapors becomes too high. The cracked shell may fall directly on the substrate or get molten and takes a spherical shape before impact. The molten particles could evaporate from their surface what reduced their size. b) Similar phenomena could have happened when liquid droplet contains a small solid precursor. Additionally, the heterogeneous nucleation was possible on the surface of small solid inside the droplet surrounded by the liquid inside solid shell. This small solid might get molten before the contact with substrate. c) Droplet having small volume of liquid around the solid particle would be, probably, submitted to evaporation of liquid. The remaining solid
would get molten before the contact with the substrate. Some evaporation from the molten particle could have also occurred. The chemical phenomena occurring in liquid precursor include the formation of HA following Eq. (2). HA is formed when precursor reaches temperature of about 500 °C (see Fig. 1). At lower temperatures the liquid calcium hydroxide may however react also with CO2 contained in open atmosphere to give the calcium carbonate: CaðOHÞ2ðlÞ þ CO2ðgÞ →CaCO3 ðsÞ þ H2 OðlÞ
ð4Þ
anions was indeed observed in liquid precurThe presence of CO−2 3 sor (see Fig. 2) and in all sprayed coating (sees Fig. 9). The coatings sprayed with the use of electric power of 36 kW and spray distance of 80 mm showed the lowest content of these ions and, consequently, the lowest content of carbonate. The carbonate decomposes at the temperature 500–600 °C to give solid calcium oxide (visible in Fig. 8c in heat treated coatings) and carbon dioxide in gaseous phase following the equation: CaCO3ðsÞ
500−600 C
→
CaOðsÞ þ CO2ðgÞ
ð5Þ
The operational deposition process parameters were varied to find their influence on morphology and chemistry of obtained coatings. An important characteristics of process was very high growth rate of coating which reached nearly 7 μm/pass in run # 3 (see Table 1). Such high growth rate renders the process interesting from the point of view of process application in industry. Before such application will take place, it is necessary to understand the origins and, eventually, to reduce the content of calcium carbonate in the coating. Knowing that the formation of the carbonate takes place when calcium hydroxide is liquid (see Eq. (3)) and in contact with gaseous carbon dioxide being present in air, the trajectory of liquid precursor being close to the jet periphery could have been be favorable to this reaction (see Fig. 11). Possibly, the solution injected to the plasma generated by the torch alimented with 40 kW was having such a trajectory (runs # 5 to 8) and the power input of 36 kW seems to result in the trajectory closer to jet axis. On the other hand, at the latter input power, the spray distance of 80 mm and low torch linear speed of 400 mm/s could have resulted in better evaporation of remaining calcium hydroxide and reduction of calcium carbonate content (see Fig. 9). Consequently, the
Fig. 9. Fourier transform infrared spectra of samples sprayed using electric power of 36 kW with different spray distances and different linear torch velocities.
R.T. Candidato Jr. et al. / Surface & Coatings Technology 277 (2015) 242–250
249
Fig. 10. Possible routes for the initial droplets: entirely liquid, a); liquid with a small solid inside, b); solid with some liquid around, c).
coatings were relatively dense without large lamellas and some porosity. The adhesion of obtained coatings tested using scratch test was to be in the range of 2.5 to 3.6 N. The future works should lead to the reduction of the content of the calcium carbonate in the coatings. This reduction may be reached by further optimizing the operational process parameters or by changing the used precursor. Similarly the adhesion of synthesized coatings seems to be low and the way should be found to increase it. The optimized coatings will be submitted to the corrosion tests in the body simulated fluid. Finally, the numerical simulation of process should help in better understanding of physical and chemical phenomena occurring at deposition. Acknowledgments
Fig. 11. The possible influence of the trajectories of liquid precursor in plasma jet on their transformation.
run # 3 seems to be the closest to the optimal one. Finally, it is important to mention the role of the substrate temperature at spraying. The high temperature reached in run # 7 could have led to cracking of the coating visible in Fig. 6b. On the other hand, the crystal phase composition of analyzes coatings (see Table 3) was not correlated with this temperature although it is well known that decomposition of HA is promoted by high temperatures. 5. Conclusions The liquid precursor of calcium hydroxide and diammonium hydrogen phosphate was applied to synthesize the coatings with the use of jet generated by arc plasma torch. The operational parameters of deposition, namely electric power input to the torch, spray distance and scan speed were varied. The coatings grew up with high rate ranging from 3 μm to nearly 7 μm by pass of torch. The coatings had finely grained microstructure and the porosity varying from 1 to 5% and were composed of hydroxyapatite, calcium phosphates, calcium oxide and calcium carbonate. The presence of the carbonates was explained by the reaction of liquid calcium hydroxide with carbon dioxide contained in air. The
Drs. C. Goldner — Constantinescu and G.-L. Lecomte helped in experimental part of the study. Dr. B. Pateyron helped with a useful discussion at the redaction of this paper. Dr. Reynaldo M. Vequizo from MSU — Illigan Institute of Technology, Illigan City in Philippines and Ms. Ermie Bacarra from Philippine Council for Industry Energy and Emerging Technologies, Taguig City, Philippines helped in initiation of collaboration with The University of Limoges. References [1] J. Tikkanen, K.A. Gross, C.C. Berndt, V. Pitkänen, J. Keskinen, S. Raghu, M. Rajala, J. Karthikeyan, Characteristics of the liquid flame spray process, Surf. Coat. Technol. 90 (1997) 210–216. [2] S. Govindarajan, R.O. Dusan, S.V. Joshi, In situ particle generation and splat formation during solution precursor plasma spraying of yttria-stabilized zirconia coatings, J. Am. Ceram. Soc. 94 (12) (2011) 4191–4199. [3] X.Z. Guo, B.G. Ravi, P.S. Devi, J.C. Hanson, J. Margolies, R.J. Gambino, J.B. Parise, S. Sampath, Synthesis of yttrium iron garnet (YIG) by citrate–nitrate gel combustion and precursor plasma spray process, J. Magn. Magn. Mater. 295 (2005) 145–154. [4] N. Sanpo, C.C. Berndt, A.S.M. Ang, J. Wang, Effect of the chelating agent contents on the topography, composition and phase of SPPS-deposited cobalt ferrite splats, Surf. Coat. Technol. 232 (2013) 247–253. [5] L. Pawłowski, Suspension and solution thermal spray coatings, Surf. Coat. Technol. 203 (2009) 2807–2829. [6] L. Pawłowski, Application of solution precursor spray techniques to obtain ceramic films and coatings, in: N. Espallagas (Ed.), Future Development of Thermal Spray Coatings, Woodhead, Cambridge, UK 2015, pp. 123–141. [7] V. Deram, C. Minichiello, R.-N. Vannier, A. Le Maguer, L. Pawłowski, D. Murano, Microstructural characterizations of plasma sprayed hydroxyapatite coatings, Surf. Coat. Technol. 166 (2003) 153–159.
250
R.T. Candidato Jr. et al. / Surface & Coatings Technology 277 (2015) 242–250
[8] R. Tomaszek, L. Pawłowski, L. Gengembre, J. Laureyns, A. Le Maguer, Microstructure of suspension plasma sprayed multilayer coatings of hydroxyapatite and titanium oxide, Surf. Coat. Technol. 201 (2007) 7432–7440. [9] B. Garcia, Z.B. Zhang, T.W. Coyle, L. Ga, R. Pillar, Hydroxyapatite coatings produced by plasma spraying of organic based solution precursor, in: M. Mizuno (Ed.), Advances in Bioceramics and Biocomposites II, Wiley, Hoboken, N.J., USA 2007, pp. 103–111. [10] Y. Huang, L. Song, X. Liu, Y. Xiao, Y. Wu, J. Chen, F. Wu, Z. Gu, Hydroxyapatite coatings deposited by liquid precursor plasma spraying: controlled dense and porous microstructures and osteoblastic cell responses, Biofabrication 2 (2010) 1–10. [11] D.M. Liu, T. Troczynski, W.T. Tseng, Water based sol–gel synthesis of hydroxyapatite: process development, Biomaterials 22 (2001) 1721–1730. [12] R. Jaworski, C. Pierlot, L. Pawłowski, M. Bigan, M. Quivrin, Synthesis and preliminary tests of suspension plasma spraying of fine hydroxyapatite powder, J. Therm. Spray Technol. 17 (5–6) (2008) 679–684. [13] R. Jaworski, C. Pierlot, L. Pawłowski, M. Bigan, M. Martel, Design of the synthesis of fine HA powder for suspension plasma spraying, Surf. Coat. Technol. 203 (2009) 1092–1097.
[14] P.L. Prevey, X-ray diffraction characterization of crystallinity and phase composition in plasma sprayed hydroxyapatite coatings, J. Therm. Spray Technol. 9 (3) (2000) 369–376. [15] http://imagej.nih.gov/ij/index.html. [16] R. Jaworski, L. Pawłowski, F. Roudet, S. Kozerski, F. Petit, Characterization of mechanical properties of suspension plasma sprayed TiO2 coatings using scratch test, Surf. Coat. Technol. 202 (2008) 2644–2653. [17] C. Pierlot, L. Pawłowski, M. Bigan, P. Chagnon, Design of experiments in thermal spraying: a review, Surf. Coat. Technol. 202 (2008) 4483–4490. [18] S. Basu, B.M. Cetegen, Modeling of thermo-physical processes in liquid ceramic precursor droplets injected into a plasma jet, Int. J. Heat Mass Transf. 50 (2007) 3278–3290. [19] A. Saha, S. Seal, B. Cetegen, E. Jordan, A. Ozturk, S. Basu, Thermo-physical processes in cerium nitrate precursor droplets injected into high temperature plasma, Surf. Coat. Technol. 203 (2009) 2081–2091. [20] B. Pathak, S. Basu, R. Kumar, Heat and mass transfer and chemical transformation in a cerium nitrate droplet, Int. J. Heat Mass Transf. 63 (2013) 301–312.