G Model

JECS-10940; No. of Pages 9 ARTICLE IN PRESS

Journal of the European Ceramic Society xxx (2016) xxx–xxx

Contents lists available at www.sciencedirect.com

Journal of the European Ceramic Society

journal homepage: www.elsevier.com/locate/jeurceramsoc

Feature article

Dense nanostructured calcium phosphate coating on titanium by cold

spray
A.M. Vilardell a , N. Cinca a,∗ , I.G. Cano a , A. Concustell a , S. Dosta a , J.M. Guilemany a ,
S. Estradé b , A. Ruiz-Caridad b , F. Peiró b
a
Centre de Projecció Tèrmica (CPT), Dpt. Ciència dels Materials i Enginyeria Metal.lúrgica, Universitat de Barcelona, Martí i Franquès 1, E-08028, Barcelona,
Spain
b
LENS-MIND/IN2UB, Dept. d’Electrònica, Universitat de Barcelona, c/Martí i Franquès 1, 08028 Barcelona, Spain

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

Article history: This article deals with the understanding of building-up mechanisms of bioactive nanocrystalline hydrox-
Received 27 July 2016 yapatite coatings by Cold Spray, revealing very promising results in contrast to more conventional
Received in revised form techniques such as Plasma Spray. A full characterization of feedstock and coatings is provided. The
18 November 2016
agglomerated structure of the powder proved to be suitable to obtain successfully thick hydroxyap-
Accepted 25 November 2016
atite coatings. A crystallite size below ∼20 nm in the powder and the as-sprayed coatings is calculated
Available online xxx
by the Rietveld X-ray refinement method and agreed by Transmission Electron Microscopy. Some wipe
tests were carried out on Ti6Al4V substrates in order to study the deposition of single particles and
Keywords:
Coating
the nanoscale features were evaluated. The resulting structure indicates that there is no delimitation
Nanocrystalline calcium phosphate of particle boundaries and the overall coating has been formed by effective compaction of the original
Cold spray nanocrystallites, leading to consistent and consolidated layers.
Electron microscopy © 2016 Elsevier Ltd. All rights reserved.

1. Introduction preferred value of crystallinity. Actually, much interest has arisen

on nanocrystalline coatings due to the small crystal size which
Hydroxyapatite plasma sprayed (PS) coatings have been widely improves cell proliferation and, therefore, the osteoconductive
used in orthopedic surgery to reconstruct hip and knee joints. Stud- properties of the coatings coming from their higher degradability
ies reveal that once HA is implanted, it has the ability to bond [4]. Mainly due to the possibility of controlling coating crystallinity,
directly to the bone, to achieve earlier and greater fixation and low temperature depositing processes such as biomimetic depo-
to reduce the healing time. Moreover, these coatings can success- sition, electrochemical deposition and solution deposition, are of
fully reduce the incidence of thigh pain and femoral osteolysis [1]. interest. However, some are time-consuming and their mechani-
Despite the existence of controversy between the use of hydrox- cal properties are poor. Others such as Cold Spray (CS) [5], Aerosol
yapatite (HA) coatings and porous metallic coatings, HA coatings Deposition (AD) [6–8] and nano-particle deposition system (NPDS)
have shown to have faster bone growth [2]. [9,10] have become of interest during the last decade.
From all methods to produce HA coatings: dip coating, elec- HA layers have already been successfully achieved by AD
trophoretic deposition, hot isostatic pressing, ion-beam sputtering, [11–13]. AD mainly consists on a carrier gas supply system with
among others, PS appears to be the most favorable mainly because mass flow control, a powder chamber containing feedstock pow-
of the high deposition rates at low cost. Nevertheless, the high der, and a deposition chamber with motored X–Y stage and a nozzle
temperatures and high cooling rates from the PS technique lead to evacuated by a rotary vacuum pump; in addition, high vacuum
HA decomposition with the formation of a large content of amor- conditions are necessary. Dense structures occur by the reduc-
phous phases. The ideal HA coating for orthopedic implants would tion of crystallite size by fracture or plastic deformation; however,
be one with strong cohesive strength, good adhesion to the sub- the bonding mechanism between the fine particles themselves has
strate and low porosity [3], while few works attempt to report a not been clarified yet [8]. Dense nanostructured HA coatings on
titanium were achieved by fracturing HA particles into nanoscale
fragments to form highly dense coatings with a theoretical den-
sity of 98.5%. TEM micrographs reveal an average grain size of
∗ Corresponding author. 16.2 nm and amorphous phases, nonetheless with a good bond-
E-mail address: [email protected] (N. Cinca).

http://dx.doi.org/10.1016/j.jeurceramsoc.2016.11.040
0955-2219/© 2016 Elsevier Ltd. All rights reserved.

Please cite this article in press as: A.M. Vilardell, et al., Dense nanostructured calcium phosphate coating on titanium by cold spray, J
Eur Ceram Soc (2016), http://dx.doi.org/10.1016/j.jeurceramsoc.2016.11.040
G Model
JECS-10940; No. of Pages 9 ARTICLE IN PRESS
2 A.M. Vilardell et al. / Journal of the European Ceramic Society xxx (2016) xxx–xxx

ing strength of 30.5 ± 1.2 MPa [11]. Further, a heat treatment at Total Reflection (ATR) diamond corrector and a DTG detector in
400 ◦ C was required to avoid amorphous HA phases, leading to an a reflectance mode from 4000 to 675 cm-1 with a resolution of
increase in average grain size up to 29.3 nm; heat treatments up 4 cm−1 . All Fourier transform infrared spectroscopy FTIR analy-
to 500 ◦ C further lead to a grain size of 99 nm, showing a decrease sis were carried out in combination with ATR, providing excellent
in biological properties. It was concluded that grain size signifi- quality data and the best possible reproducibility. This is a sam-
cantly affects biocompatibility. Webster et al. [14] reported that pling method that enables direct examination of samples without
osteoblast functions on ceramics with grains smaller than 100 nm further preparation of liquids, gases or solids.
were significantly enhanced compared with larger grains. So, the The morphological characterization of feedstock powder and
production of nanocrystalline HA coatings on titanium seems desir- HA coatings was performed by Scanning Electron Microscopy SEM
able for implants if it can be accomplished at low temperature. (ProX Phenom) equipped with Energy Dispersive Spectroscopy
A difference between AD and CS is that AD uses submicrom- (EDS) for microanalysis. The phase identification of the powder and
eter particles as feedstock while CS was originally more suitable coating were analyzed by a X’Pert PRO MPD diffractometer (PAN-
for metallic feedstock within the 5–50 ␮m sized range and does alytical). A Rietveld analysis, using the FullProf software [27], was
not use vacuum conditions. CS is a solid-state coating deposition carried out to refine the lattice parameters, determine the crystal-
method where feedstock powders are propelled toward a gas jet at lite size and to find the percentage of the crystalline and amorphous
supersonic velocities up to 1100 m/s and lower temperatures than phase [28]. The Thomson-Cox-Hastings pseudo-Voigt profile func-
the melting point of the material, with minimum thermal input. tion was used for the refinement and the instrumental resolution
Metallic particles undergo plastic deformation and adhere to the was also introduced. In order to achieve a good correlation between
substrate. Temperature- and oxygen-sensitive materials such as the calculated and the experimental diffractograms, the minimum
polymer and titanium are successfully densely deposited [15–17]. value of the usual fit indicators of the FullProf software were moni-
Due to the inherent brittleness, ceramic deposition by CS is still a tored (i.e. the refinement should give a reduced chi-squared nearly
challenge and ceramic particles are deposited as cermets [18] or equal to 1). The increase of peak half width respect the instrumen-
co-deposited with other ductile materials (e.g titanium) [19,20]. tal contribution is assumed to mainly come from the grain size and
Lately, the research of HA CS coatings has been carried out strain effects; the different proper parameters were fitted through-
with a lot of interest within the biomedical field. Some attempts out the refinement. At each step, iteration cycles were conducted
of spraying HA were performed on ductile substrate materials until convergence was reached. In order to evaluate the amorphous
with successful results. Lee et al. [21] deposited a HA coating phase content, HA coating was detached from the substrate and
homogeneously onto polyetheretherketone (PEEK) disk implants crushed and mixed with a known weight of crystalline phase pat-
enhancing in vitro biocompatibility and promoting in vivo osseoin- tern, in our case alumina. It is possible to calculate directly the
tegration. Noorakma et al. [22] improved the biodegradability of weight fractions of the crystalline phases WJ and the weight frac-
magnesium alloy by the re-precipitation of an apatite layer as a con- tion WAM of the amorphous phases just knowing the fraction in
sequence of HA coating onto a magnesium alloy and HA-graphene weight WP of the pattern (experimental weighted alumina frac-
composites were deposited as well [23]. Recently, we focused on tion) and the weight fractions of the phases, WJO , quantified from
the deposition mechanisms of HA by CS and reported the feasibility the Rietveld refinement (including that of the pattern WPO ), with
of the deposition of sintered HA powders by dynamic fragmenta- the following equations [28]:
tion due to pore collapse + cracking and crushing mechanisms of
the HA particles, with a considerable decrease in crystal size [24]. WJ0 − WP
Apart from this, only numerical investigations have faced the issue WJ = (2)
WP0 (1 − Wp)
of HA deposition by CS, primarily analyzing the influence of nozzle
and particle geometry in particle velocity [25,26].
The present article is aimed at studying the build-up of CS W0P − WP
WAM = (3)
coatings from an agglomerated nano-crystalline HA powder, and WP0 (1 − WP )
compare it with the previous studies for a porous sintered crys-
®
talline HA powder [24]. The Cold Gas Spray equipment used is a CGS KINETICS 4000
(Cold Gas Technology, Ampfing, Germany) with a maximum oper-
2. Materials and methods ating pressure of 40 bar, temperature of 800 ◦ C and operated with
nitrogen as the propellant gas.
An agglomerated powder from Medicoat (France) was used as For the evaluation of single particle deposition, the so-called
feedstock. The particle size has been measured using a Laser Diffrac- “wipe test” was performed, which consisted in particle deposition
tion Particle Size Analyser Beckman Coulter LS 13320. The density onto a mirror-like polished Ti6Al4V alloy substrate at high gun
of the powder was measured through the standard specification traverse speed.
ASTM B-962-08; the density is evaluated introducing in a flask of The HA coated sample from previous studies [29] was used for
25 ml a known mass of the powder in consideration and filling with the study of particle-particle interaction. In order to examine those
a high wettable liquid with a known density, here cyclohexanone. features, a thin lamella from HA particle, single splat and coating
Knowing the density of the dissolvent, the powder density can be were obtained by Focused Ion Beam (FIB) lift-out technique and
calculated the following Eq. (1) the real density: compared to the initial feedstock structure. A FEI Strata Dual Beam
235 system was used. Finally, transmission electron microscopy
mp
␳=   (1) examinations were carried out using a JEM 2100 microscope, oper-
mt −mp
25 − ␳liq
ated at 200 kV (with current density of 80–250pA/cm2 ). Much care
was taken when irradiating the samples since some damage could
where mp is the mass powder introduced, mt is the total mass be noticed through the observation; however, the use of a spread
(powder + cyclohexanone) and ␳liq is the cyclohexanone density at beam to acquire HRTEM images produces a decrease of the elec-
working temperature. Thermogravimetry (TG) was carried out by tron dose on the sample, which gave us the possibility to obtain
SDT 2960 (TA instruments) while FTIR by Thermo SCIENTIFIC NICO- such information without being significantly degraded.
LET In10 MX of reflection with MCT detector. FTIR measurements In order to evaluate the adhesion, a Revetest (CSM Instrument)
were carried out in a Scientific Nicolet iZ10 MX, using Attenuated scratch equipment with a Rockwell C diamond indenter with a cone

Please cite this article in press as: A.M. Vilardell, et al., Dense nanostructured calcium phosphate coating on titanium by cold spray, J
Eur Ceram Soc (2016), http://dx.doi.org/10.1016/j.jeurceramsoc.2016.11.040
G Model
JECS-10940; No. of Pages 9 ARTICLE IN PRESS
A.M. Vilardell et al. / Journal of the European Ceramic Society xxx (2016) xxx–xxx 3

Fig. 1. HA powder morphology at different magnifications.

angle 120◦ and tip radius 200 ␮m was used. The scratch speed was 10−3 A, c = 6892 ± 0.096 10−3 A and, (ii) monetite a = 6905 ± 0.656
from 0.4 to 600 mm/min until a maximum load of 200 N. 10−3 A, b = 6.642 ± 0.551 10−3 A, c = 7.001 ± 1.193 10−3 A,
␣= 96.423 ± 0.014◦ , ␤= 103.873 ± 0.015◦ , ␥= 88.514 ± 0.011◦ .
3. Results and discussion The minimum value of the usual fit indicator Chi2 that was
monitored by the FullProf software along the fitting was 5.45. A
3.1. Characterization of feedstock powder final crystallite value of 13.8 nm was found for the hydroxyapatite
phase. According to [14], this grain size should be suitable for
Fig. 1a displays the spherical morphology of HA particles with biological properties because is less than 100 nm
a particle size distribution of −63 + 15 ␮m and a mean diameter of An energy dispersive spectroscopy (EDS) analysis for the HA
32.9 ␮m. The HA powder with a density of 0.55 g/cm3 appears to powder indicated a Ca/P ratio of 1.76. Considering that a stoichio-
be composed of very fine agglomerates (Fig. 1b). TEM micrographs metric HA has a Ca/P ratio of 1.67, this indicates that they could
actually reveal a structure consisting of a needle-like morphology directly bond to bone in contrast to calcium phosphates that have
(Fig. 2a) whose electron diffraction exhibits a ring pattern denoting lower ratio of Ca/P than 1.4 [31].
its nanocrystalline nature. At higher magnification, nanoparticle FTIR spectrum of HA powder contains characteristic phosphate,
superposition becomes apparent, leading to the observation of carbonate and hydroxyl bands (Fig. 4) as presented in Table 1.
Moiré fringes (Fig. 2b). It is, however, quite obvious the analogy Theoretically, there are four vibrational modes present for phos-
with other works reporting these morphologies as a result of the phate ions, 1 , 2 , 3 and 4 . The phosphate 1 band is present
reaction between calcium hydroxide and ortho-phosphoric acid: at 962 cm−1 corresponding to non-degenerate symmetric P O
5Ca(OH)2 + 3H3 PO4 → Ca5 (PO4 )3 (OH) + 9H2 O [30]. stretching, whereas the 3 region between 1190 and 976 cm−1 cor-
Fig. 3a shows the X-Ray Dif;1;fraction (XRD) pattern of the pow- responds to anti-symmetric P O stretching and the 4 bands at
der with broad peaks more likely as a consequence of the very 560,600 and 629 cm−1 corresponding to triply degenerated O P O
small crystallite size. The identification revealed the presence of bending. 2 is not found in FTIR spectrum because of its wave
the 9–432 JCPDS HA syn. pattern corresponding to a hexagonal number (cm−1 ) is below the minimum of data spectrum [32]. The-
P63m/lattice, as well as the 9–080 JCPDS pattern, which corre- oretically, carbonate ions have four vibrational modes, in which
sponds to the monetite phase, a calcium hydroxide phosphate with three of them can be observed in the infrared spectrum and two
triclinic lattice. in Raman spectrum. Carbonate bands 3 and 2 are located in the
By using the FullProf software, the XRD pattern was refined region of 1650–1300 cm−1 and at 875 cm−1 respectively. Despite
(Fig. 3b) and the content of HA phase was found to be not distinguished by XRD, the presence of carbonate ions may come
88.65 ± 0.28%, the monetite phase 3.78 ± 0.14% and the amor- from a reaction between carbon dioxide in air and the high pH pre-
phous phase 7.56 ± 0.15% [28]. The refined lattice dimensions cursor solution in powder production [33]. Finally, in comparison
were found to be the following: (i) HA a = b = 9429 ± 0.134 with other HA phases such as ␤-TCP, the OH group two peaks are

Fig. 2. (a–b) TEM micrographs and SAED pattern of an agglomerated HA particle.

Please cite this article in press as: A.M. Vilardell, et al., Dense nanostructured calcium phosphate coating on titanium by cold spray, J
Eur Ceram Soc (2016), http://dx.doi.org/10.1016/j.jeurceramsoc.2016.11.040
G Model
JECS-10940; No. of Pages 9 ARTICLE IN PRESS
4 A.M. Vilardell et al. / Journal of the European Ceramic Society xxx (2016) xxx–xxx

Fig. 3. (a)XRD and (b) Rietveld analysis of the HA powder.

Fig. 4. FTIR spectrum of the HA powder. Fig. 5. TG curve of HA powder.

Table 1 droxylation of HA powder. This can be explained by the following

Infrared band positions for hydroxyapatite powder and coating. reaction: (Eq. (4)) [35].
HA POWDER HA COATING
Ca10 (PO4 )6 (OH)2 → Ca10 (PO4 )6 (OH)2−2x Ox 䊐x + xH2 O (4)
Hydroxyl stretch 5 3569 3570
635 635
Carbonate 3 1650–1300 1650–1300
3.2. Characterization of coating build-up
- (m) 1641 1643
- (m) 1448 1446 3.2.1. Single particle study
- (m) 1417 1417 CS bonding mechanism in metals can be attributed to adiabatic
- (m) – –
shear instability which occurs at the particle/substrate or parti-
Phosphate 3 1190–976 1190–976 cle/deposited material interfaces at high impact particle velocities.
- (vs) 1088 1087 (debil) However, the deposition of brittle or hard materials, such as ceram-
- (vs) 1023 1022
ics, is still under research. Fig. 6a shows a general view of the
Phosphate 1 (m) 962 962 wipe test on the Ti6Al4V substrate and Fig. 6b is a magnification
Carbonate 3 (ms) 875 875 of a single deposited particle. A quite high particle deposition is
Phosphate 4 660–520 660–520
- (m) 629 622
observed and the absence of craters is an indication that upon the
- (vs) 600 599 high shock pressure of the process and, given the hardness of the
- (vs) 560 560 substrate and the agglomerated nature of the feedstock, the parti-
cles are more prone to spread over the surface rather that bouncing
off leaving craters. This behavior is somehow different from what
observed at 3569 cm−1 and a weaker one at 635 cm−1 , which corre- was observed within the previous study with the sintered feed-
spond respectively to the stretching and libration mode of hydroxyl stock powder where, when the particles did not properly bond,
vibrations typical of HA [34]. Fig. 5 shows the TG curve for hydrox- they left ring-shape morphologies with the central part of the splat
yapatite powder. The curve decreases progressively from 200 to bounced off. That performance was well in agreement to the stress
1300 ◦ C. No significant peaks were observed, except for a weight distribution simulations usually carried out in CS which indicate
loss of ∼4.5%, which is assumed to be the result of gradual dehy- that upon impact, tensile stresses may be developed at the center

Please cite this article in press as: A.M. Vilardell, et al., Dense nanostructured calcium phosphate coating on titanium by cold spray, J
Eur Ceram Soc (2016), http://dx.doi.org/10.1016/j.jeurceramsoc.2016.11.040
G Model
JECS-10940; No. of Pages 9 ARTICLE IN PRESS
A.M. Vilardell et al. / Journal of the European Ceramic Society xxx (2016) xxx–xxx 5

Fig. 6. Wipe test of HA on the Ti6Al4V substrate at different magnifications.

bottom of the splats resulting in detachment [36]. Similar ceramic using the FullProf software, the Chi2 value obtained at the final fit-
particle behavior was found in the AD method. The AD method is ting was 8.71. The XRD pattern was refined (Fig. 9b) and the content
based on shock-loading compaction due to the impact of ultrafine of the hydroxyapatite was found to be 86.58 ± 0.47%, the monetite
ceramic particles onto a surface. First, particles are mixed with a phase 2.07 ± 0.16% and the amorphous phase 11.34 ± 0.20%. It did
gas to generate an aerosol. During the impact onto the substrate, not result then in any microstructural and compositional change.
part of the particles kinetic energy is converted into thermal energy Also, a final value of 11 nm in crystallite size was found for the
causing increase in temperature at the point-of-impact and pro- hydroxyapatite phase. No significant changes were thus perceived
moting bonding between the substrate and the particles and also in comparison with feedstock powder; the HA powder preserved
between multiple particles. The result is a process that yields an the nanoscale nature structure after CS deposition. Fig. 10 shows the
acceptable consolidation at room temperature without the need FTIR spectrum of the HA coating that contains the same characteris-
for a thermal treatment. This process is called room temperature tic phosphate, carbonate and hydroxyl bands as the FTIR spectrum
impact consolidation (RTIC) [7]. of the HA powder (Fig. 4), previously discussed. The unique differ-
The morphology of CS HA bonded particles, either coming from ence between both spectrums is the intensity of the peaks, which
a sintered feedstock or an agglomerate seem to be quite similar are less intense in the HA coating because of the surface rough-
(Fig. 6) but further insight into the microstructure reveals the dif- ness. Actually, the main sampling problem in FTIR characterization
ferences [24]. The nature of the powder exhibited in Fig. 2a and b of biomaterials is that nearly all solid materials are too opaque
made already foresee that its deposition would be quite different in their normal forms for direct transmission analysis in the mid-
from that of sintered particles, i.e by pore collapsing and crush- infrared region. This problem can be solved by reducing the optical
ing. Therefore, the powder processing route is really important to density of samples to a suitable level by employing various sam-
be considered. A prepared FIB-splat lamella was prepared from HA pling techniques but, since these procedures can alter the nature of
splat (Fig. 7a). In general, at low magnification, a random orienta- the sample and are time consuming. For that, ATR was employed;
tion of the small crystallites was observed all over the splat (Fig. 7b). however, it is limited in some applications due to stringent surface
At certain zones, more in the upper side of the prepared FIB-splat requirements. This confirms from the compositional point of view,
lamella, the needle-like morphology can still be observed (Fig. 7c); that the feedstock composition is preserved after spraying.
some clearer areas are most likely artifacts related to the TEM The compacity of the coating and the interaction with the sub-
specimen observation [37]. In other more inner regions (Fig. 7d), strate were analyzed also by TEM. Several micrographs of different
however, Moiré fringes are also observed but without observing zones are shown in Figs. 11 and 12 showing the upper and bot-
such needle-like structure; rather, crystallite overlapping appears tom part of the section respectively. Moiré patterns were observed
to be more significant. Fig. 7e shows the region near the interface as in the splat but the upper side (Figs. 11 and 12) exhibiting an
with the substrate, where such overlapping is visible. It must be effective consolidation. This must be due to the tamping effect pro-
said that all along the interface, a good adhesion was observed, duced by continuous impact of incoming particles onto the already
similar to what was reported by Yao et al. for TiO2 [38]. It seems adhered ones. Particle-particle boundaries are difficult to be recog-
that the impact has produced compaction of nanocrystalline grains nized in Fig. 12. This might be explained by the shock compaction
within the particle, preserving the nanoscale nature. This is slightly of nano-sized particles [39]. Mamalis et al. [40] related this shock
different from the room-temperature impact consolidation (RTIC) compaction to the fracture of particles, filling the interstices, clean-
mechanism reported for AD [7] because it would first require frac- ing the surfaces and heating of the particle surfaces which lead to
turing of submicrometer particles. either partial surface melting and welding or solid-state diffusion
bonding.
3.2.2. Coating deposition This thermal effect is well included in the energy balance upon
Thick HA coatings were achieved by CS (Fig. 8a). The top sur- particle impact. The bonding mechanisms of metal particles within
face consists of a powdering texture, consistent with the observed a coating deposited by CS are known to be predominantly by plas-
agglomerated nature of the feedstock (Fig. 8b). tic deformation at high strain rates. The heat generated upon the
Fig. 9a shows the XRD pattern of the HA coating deposited by impact may promote dynamic recrystallization or melting in shear
CS. The same phases were identified; the monetite phase has been bands, which are occurring at the interfacial regions, i.e. particle-
found to be present already in the feedstock, so it is not produced substrate or particle-particle interfaces. Since it has been proved
as result of the thermal history of the powder in the gas stream, that not enough heat is produced to be able to melt the impinging
whose temperature is much lower than the typical temperatures of particle [41], the only heat considered for energetic considerations
conventional thermal spray processes that produce powder decom- is that of adiabatic heating, dissipated locally at the interface via
position. Both XRD patterns maintain similar peak broadness. By plastic deformation. A proper energy balance may be then the fol-

Please cite this article in press as: A.M. Vilardell, et al., Dense nanostructured calcium phosphate coating on titanium by cold spray, J
Eur Ceram Soc (2016), http://dx.doi.org/10.1016/j.jeurceramsoc.2016.11.040
G Model
JECS-10940; No. of Pages 9 ARTICLE IN PRESS
6 A.M. Vilardell et al. / Journal of the European Ceramic Society xxx (2016) xxx–xxx

Fig. 7. (a) FIB preparation of HA splat on Ti6Al4V alloy. (b) General TEM micrograph representative of the whole splat, (c) upper, (d)central and (e) bottom part of the HA
splat onto Ti6Al4V.

lowing: (Eq. (5)) Ekinetic = Eplastic + Ethermal , where Dewar et al. [42], through a failure model. The Johnson-Holmquist model con-
were able to calculate the particle-substrate interface temperature tains a strength model and a failure model that was used by
assuming that none of the kinetic energy that is converted to heat Chun et al. [39] to simulate the impact of an alumina particle
is lost to the environment, to the rest of the bulk splat, or to the by NPDS; this model uses the normalized stress and pressure
substrate. Obviously, this is an approximation since all the physical values from the strength and pressure components of the Hugo-
and thermal properties of the splat are assumed to be constant. niot elastic limit (␴HEL , PHEL ). In this case, the kinetic energy is
For the impact of a ceramic particle at low-temperature deposi- transformed into thermal and fragmentation energies: (Eq. (6))
tion systems, fragmentation phenomenon has also to be included Ekinetic = Efracture + Ethermal = 12 P(V00 -V0 ), where V00 and V0 are the

Please cite this article in press as: A.M. Vilardell, et al., Dense nanostructured calcium phosphate coating on titanium by cold spray, J
Eur Ceram Soc (2016), http://dx.doi.org/10.1016/j.jeurceramsoc.2016.11.040
G Model
JECS-10940; No. of Pages 9 ARTICLE IN PRESS
A.M. Vilardell et al. / Journal of the European Ceramic Society xxx (2016) xxx–xxx 7

Fig. 8. (a) Macrograph of HA coatings and (b) top surface SEM micrograph.

Fig. 9. (a) X-ray diffraction and (b) Rietveld analysis of the HA coating.

Fig. 11. TEM micrograph of the upper part of HA coating.

Fig. 10. FTIR spectrum of the HA coating.

In the present case, however, the feedstock is not sintered

specific volumes of the porous and solid material respectively. and, therefore, Efracture is assumed to be low and most of the
As already commented previously, the fracturing depends on the kinetic energy is transformed into thermal energy. The particle den-
shock intensity and therefore to the impact conditions and, on the sity is 3001 g/cm3 for the sintered particles in [24] and, as said
powder initial density. before, 0,55 g/cm3 for the agglomerated particles of the present
If we consider a sintered particle, it is obvious that Efracture can be work; therefore, for the same particle size, the lower mass of the
significant and the gaps of the initial pores can be more or less effec- agglomerated particles leads to lower kinetic energy. Then, the
tively filled depending on the impact conditions; this can also result temperature rise is obviously not high enough to produce sintering
in the presence of plastic deformation mechanism. According to the since the structure of Figs. 11 and 12 leads to a consolidation that
simulations of Akedo et al. [8], the maximum rising impact pres- keeps the nanocrystallites with good cohesion.
sure was almost the same as the experimental fracture toughness A first approach to calculate the mechanical properties was per-
of the starting particle and enough to destroy the particle. formed by the scratch test. A thick HA coating with approximately

Please cite this article in press as: A.M. Vilardell, et al., Dense nanostructured calcium phosphate coating on titanium by cold spray, J
Eur Ceram Soc (2016), http://dx.doi.org/10.1016/j.jeurceramsoc.2016.11.040
G Model
JECS-10940; No. of Pages 9 ARTICLE IN PRESS
8 A.M. Vilardell et al. / Journal of the European Ceramic Society xxx (2016) xxx–xxx

[6] J. Akedo, Aerosol deposition method for fabrication of nano crystal ceramic
layer, Mater. Sci. Forum. (2004) 43–48.
[7] J. Akedo, Room temperature impact consolidation (RTIC) of fine ceramic
powder by aerosol deposition method and applications to microdevices,
J.Therm. Spray Technol. 17 (2008) 181–198.
[8] J. Akedo, Aerosol deposition of ceramic thick films at room temperature:
densification mechanism of ceramic layers, J. Am. Ceram. Soc. 89 (2006)
1834–1839.
[9] M.-S. Kim, D.-M. Chun, J.-O. Choi, J.-C. Lee, K.-S. Kim, Y. Kim, et al., Room
temperature deposition of TiO2 using nano particle deposition system
(NPDS): application to dye-sensitized solar cell (DSSC), Int. J. Precis. Eng.
Manuf. 12 (2011) 749–752.
[10] D.-M. Chun, J.-O. Choi, C.S. Lee, S.-H. Ahn, Effect of stand-off distance for cold
gas spraying of fine ceramic particles (<5 (m) under low vacuum and room
temperature using nano-particle deposition system (NPDS), Surf. Coat. Tech.
206 (2012) 2125–2132.
[11] B.-D. Hahn, D.-S. Park, J.-J. Choi, J. Ryu, W.-H. Yoon, K.-H. Kim, et al., Dense
nanostructured hydroxyapatite coating on titanium by aerosol deposition, J.
Am. Ceram. Soc. 92 (2009) 683–687.
[12] B.-D. Hahn, J.-M. Lee, D.-S. Park, J. Choi, W.-H. Yoon, et al., Mechanical and
in vitro biological performances of hydroxyapatite–carbon nanotube
Fig. 12. TEM micrograph of the bottom part of HA coating. composite coatings deposited on Ti by aerosol deposition, Acta Biomater. 5
(2009) 3205–3214.
[13] B.-D. Hahn, D.-S. Park, J. Choi, W.-H. Yoon, J.-H. Choi, et al., Aerosol deposition
350 ␮m showed a critical load of 53 ± 10N, while a thin carbonated of hydroxyapatite–chitosan composite coatings on biodegradable magnesium
alloy, Surf. Coat. Tech. 205 (2011) 3112–3118.
biomimetic nanocrystalline appatite coating by CS with a thickness
[14] T.J. Webster, C. Ergun, R.H. Doremus, R.W. Siegel, R. Bizios, Enhanced functions
of about 60 ␮m showed a value of 7 ± 2 N [43]. On the other hand, of osteoblasts on nanophase ceramics, Biomater 21 (2000) 1803–1810.
HA coatings with a 150 ␮m thickness were obtained by PS onto tita- [15] V.K. Champagne, The Cold Spray Materials Deposition Process: Fundamentals
nium, indicating a critical load of 12.3 N, and 10.5 N onto aluminum and Applications, Woodhead, Cambridge, 2016.
[16] R. Ghelichi, M. Guagliano, Coating by the Cold Spray Process: a state o the art,
[44]. HA coatings also were obtained by microplasma spraying [45], Fratt. Integr. Strutt. 8 (2009) 30–44.
and it was observed that for a HA coating, with a thickness of [17] H.R. Singh, T.S. Sidhu, S.B.S. Kalsi, Cold spray technology: future of coating
232 ± 8.25 ␮m, 91.6% of crystallinity and a 19.17 ± 1.98% of poros- deposition processes, Fratt. Integr. Strutt. 22 (2012) 69–84.
[18] S. Dosta, M. Couto, J.M. Guilemany, Cold spray deposition of a WC-25Co
ity, the critical normal load that caused localized delamination cermet onto Al7075-T6 and carbon steel substrates, Acta Mater. 61 (2013)
through microcracking was ∼400 mN. 643–652.
[19] A. Choudhuri, P.S. Mohanty, J. Karthikeyan, Bio-ceramic composite coatings
by cold spray technology, Proceedings of the International Thermal Spray
4. Conclusions Conference (2009) 391–396.
[20] X. Zhou, Hydroxyapatite/titanium Composite Coating for Biomedical
Cold Spray has been shown to be a suitable technique to deposit Applications, University of Michigan, 2012 (PhD thesis).
[21] J.H. Lee, H.L. Jang, K.M. Lee, H.-R. Baek, K. Jin, K.S. Hong, et al., In vitro and
nanocrystalline hydroxiapatite particles. TEM micrographs actually in vivo evaluation of the bioactivity of hydroxyapatite-coated
reveal a structure consisting of a needle-like morphology. It seems polyetheretherketone biocomposites created by cold spray technology, Acta
that the impact has produced compaction of nanocrystalline grains Biomater. 9 (2013) 6177–6187.
[22] A.W. Noorakma, H. Zuhailawati, V. Aishvarya, B.K. Dhindaw,
within the particle. Moiré patterns were observed as in the splat but Hydroxyapatite-Coated magnesium-Based biodegradable alloy: cold spray
the upper side shows partial contacts between crystallites while deposition and simulated body fluid studies, J. Mater. Eng. Perform. 22 (2013)
the bottom side exhibits much more effective consolidation. This 2997–3004.
[23] Y. Liu, J. Huang, H. Li, Nanostructural characteristics of vacuum cold-Sprayed
must be due to the tamping effect produced by continuous impact
hydroxyapatite/graphene-nanosheet coatings for biomedical applications, J.
of incoming particles onto the already adhered ones. Powder and Therm. Spray Technol. 23 (2014) 1149–1156.
coating characterization confirm that there is no amorphization of [24] N. Cinca, A.M. Vilardell, S. Dosta, A. Concustell, I. Garcia Cano, J.M. Guilemany,
et al., A new alternative for obtaining nanocrystalline bioactive coatings:
HA and the microstructure is preserved. Moreover, no size changes
study of hydroxyapatite deposition mechanisms by cold gas spraying, J. Am.
were detected in nano-sized grains by TEM. Ceram. Soc. (2016) 1–9.
[25] L. Zhang, W.T. Zhang, Numerical investigation on particle velocity in cold
spraying of hydroxyapatite coating, Adv. Mater. Res. 18 (2011) 717–722.
Acknowledgements [26] R.P. Singh, B. Batra, Effect of cold spraying parameters and their interaction an
hydroxyapatite deposition, J. Appl. Fluid Mech. 6 (2013) 555–561.
This work was supported by the Generalitat de Catalunya for [27] J. Rodriguez-Carvajal, Introduction to the Program FULLPROF: Refinement of
Crystal and Magnetic Structures from Powder and Single Crystal Data, 2014.
the project [2014 SGR 1558]; the University of Barcelona for the
[28] X. Alcobé, El método de rietveld en el caso de muestras multifásicas, in:
award of APIF scholarship that has allowed the development of this Análisis Cuantitativo. Apectos Prácticos Y Ejemplos, XIII Escuela sobre el
project. Metodo Rietvld, Universidad Jaume I Castellón, 2013.
[29] AM. Vilardell, N. Cinca, S. Dosta, IG. Cano, JM. Guilemany, 2016 Mulifuctional
miro-nano structured hydroxyapatite Coatings on tianum Ti6Al4V for
References biomedical Applications by Cold Gas Spray, Núm. de solicitud: 809 País de
prioridad: ESPAÑA Entidad Titular: UBAR − Universitat de Barcelona.
[1] B. Chambers, S.F. St Clair, M.I. Froimson, Hydroxyapatite-Coated tapered Depósito Legal: 17/04/2015.
cementless femoral components in total hip arthroplasty, J. Arthroplasty 22 [30] R. Kumar, K.H. Prakash, P. Cheang, K.A. Khor, Temperature driven
(2007) 71–74. morphological changes of chemically precipitated hydroxyapatite
[2] Y.-L. Chen, T. Lin, A. Liu, M.-M. Shi, B. Hu, Z. Shi, et al., Does hydroxyapatite nanoparticles, Langmuir 20 (2004) 5196–5200.
coating have no advantage over porous coating in primary total hip [31] K. De Groot, C.P.A.T. Klein, J.D. Wolke, J.M.A. de Blieck-Hogervorst, Chemistry
arthroplasty? A meta-analysis, J. Orthop. Surg. Res. 10 (2015) 1–14. of calcium phosphate bioceramics Handbook of Bioactive Ceramics, vol. 2,
[3] Y.C. Tsui, C. Doyle, T.W. Clyne, Plasma sprayed hydroxyapatite coatings on 1990, pp. 3–16.
titanium substrates Part 1: Mechanical properties and residual stress levels, [32] I. Rehman, W. Bonfield, Characterization of hydroxyapatite and carbonated
Biomater 19 (1998) 2015–2029. apatite by photo acoustic FTIR spectroscopy, J. Mater. Sci. Mater. Med. 8
[4] A. Bernhardt, R. Dittrich, A. Lode, F. Despang, M. Gelinsky, Nanocrystalline (1997) 1–4.
spherical hydroxyapatite granules for bone repair: in vitro evaluation with [33] K. Venkateswarlu, A. Chandra Bose, N. Rameshbabu, X-ray peak broadening
osteoblast-like cells and osteoclasts, J. Mater. Sci.: Mater. Med. 24 (2013) studies of nanocrystalline hydroxyapatite by Williamson–Hall analysis,
1755–1766. Physica B 405 (2010) 4256–4261.
[5] A.M. Vilardell, N. Cinca, A. Concustell, S. Dosta, I.G. Cano, J.M. Guilemany, Cold [34] L. Berzina-Cimdina, N. Borodajenko, Research of calcium phosphates using
spray as an emerging technology for biocompatible and antibacterial fourier transform infrared spectroscopy, in: T. Theophile (Ed.), Infrared
coatings: state of art, J. Mater. Sci. 50 (2015) 4441–4462. Spectroscopy − Materials Science, Engineering and Technology, InTech, 2012,

Please cite this article in press as: A.M. Vilardell, et al., Dense nanostructured calcium phosphate coating on titanium by cold spray, J
Eur Ceram Soc (2016), http://dx.doi.org/10.1016/j.jeurceramsoc.2016.11.040
G Model
JECS-10940; No. of Pages 9 ARTICLE IN PRESS
A.M. Vilardell et al. / Journal of the European Ceramic Society xxx (2016) xxx–xxx 9

http://www.intechopen.com/books/infrared-spectroscopy-materials- [40] A. Mamalis, I. Vottea, D. Manolakos, On the modelling of the compaction

science-engineering-and-technology/research-of-calcium-phosphates-using- mechanism of shock compacted powders, J. Mater. Process Tech. 108 (2001)
fourier-transformation-infrared-spectroscopy (accessed November 14, 2016). 165–178.
[35] B.C. Idrissi, K. Yamni, A. Yacoubi, A. Massit, A novel method to synthesize [41] T. Schmidt, H. Assadi, F. Gärtner, H. Richter, T. Stoltenhoff, H. Kreye, et al.,
nanocrystalline hydroxyapatite: characterization with x-ray diffraction and From particle acceleration to impact and bonding in cold spraying, J. Therm.
infrared spectroscopy, IOSR J. Appl. Chem. 7 (2007) 107–112. Spray Technol. 18 (2009) 794–808.
[36] K. Kim, M. Watanabe, S. Kuroda, Jetting-out phenomenon associated with [42] M.P. Dewar, A.G. McDonald, A.P. Gerlich, Interfacial heating during
bonding of warm-sprayed titanium particles onto steel substrate, J. Therm. low-pressure cold-gas dynamic spraying of aluminum coatings, J. Mater. Sci.
Spray Technol. 18 (2009) 490–499. 47 (2011) 184–198.
[37] J. Reyes-Gasga, R. Garcia-Garcia, E. Brès, Electron beam interaction, damage [43] E. Kergourlay, D. Grossin, N. Cinca, First cold spraying of carbonated
and reconstruction of hydroxyapatite, Phys. B: Condens Matter 404 (2009) biomimetic nanocrystalline apatite on Ti6Al4V: physical-chemical,
1867–1873. microstructural, and preliminary mechanical characterizations, Adv. Eng.
[38] H.-L. Yao, G.-J. Yang, S.-Q. Fan, C.-X. Li, C.-J. Li, Ceramic Mater. 18 (2016) 496–500.
nano-particle/Substrate interface bonding formation derived from dynamic [44] S. Kozerski, L. Pawlowski, R. Jaworski, F. Roudet, F. Petit, Two zones
mechanical force at room temperature: HRTEM examination, J. Therm. Spray microstructure of suspension plasma sprayed hydroxyapatite coatings, Surf.
Technol. 24 (2015) 720–728. Coat. Tech. 204 (2010) 1380–1387.
[39] D.-M. Chun, S.-H. Ahn, Deposition mechanism of dry sprayed ceramic [45] A. Dey, A.K. Mukhopadhyay, S. Gangadharan, M.K. Sinha, D. Basu,
particles at room temperature using a nano-particle deposition system, Acta Characterization of microplasma sprayed hydroxyapatite coating, J. Therm.
Mater. 59 (2011) 2693–2703. Spray Technol. 18 (2009) 578–592.

Please cite this article in press as: A.M. Vilardell, et al., Dense nanostructured calcium phosphate coating on titanium by cold spray, J
Eur Ceram Soc (2016), http://dx.doi.org/10.1016/j.jeurceramsoc.2016.11.040