Acta Materialia 53 (2005) 4367–4377 www.actamat-journals.com
Transmission electron microscopy and atomic force microscopy characterisation of titanium-base alloys nitrided under glow discharge A. Czyrska-Filemonowicz a,*, P.A. Buffat b, M. Łucki a, T. Moskalewicz a, W. Rakowski a, J. Lekki c, T. Wierzchon´ d a
AGH University of Science and Technology, Faculty of Metallurgy and Materials Science, Al. Mickiewicza 30, PL-30059 Krako´w, Poland b Ecole Polytechnique Fe´de´rale de Lausanne, CH-1015 Lausanne, Switzerland c Institute of Nuclear Physics of the Polish Academy of Sciences, ul. Radzikowskiego 152, PL-31342 Krako´w, Poland d Warsaw University of Technology, ul. Wołoska 141, PL-02507 Warszawa, Poland Received 21 September 2004; received in revised form 19 May 2005; accepted 23 May 2005 Available online 19 July 2005
Abstract The microstructure of titanium alloys (Ti–1Al–1Mn, Ti–6Al–4V) nitrided under glow discharge was characterised using light microscopy, X-ray diffraction and analytical electron microscopy. The investigations revealed a complex multilayered microstructure produced by diffusion at 850 and 900 C at the surface. The layers formed on the Ti–1Al–1Mn alloy were identified as d-TiN (face-centred cubic NaCl type), d 0 -Ti2N (tetragonal body-centred), e-Ti2N (tetragonal primitive) phases and a (N) solid solution containing up to 19 at.% nitrogen. The layers formed on the Ti–6Al–4V alloy also contained, besides the phases mentioned above, Ti2AlN and other minor Ti–Al–N ternary phases. The layer topography of both nitrided alloys, examined by atomic force microscopy, exhibited pronounced surface development. A relationship between the microstructure of the alloys and their micromechanical and tribological properties has been established. It shows that the improved properties are related to the presence of nanocrystalline d-TiN in the outermost sublayer on the top of the layers composed of d 0 + e-Ti2N and Ti2AlN crystals. 2005 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved. Keywords: Biomaterials; Titanium alloys; Wear; TEM; AFM
1. Introduction Titanium alloys have excellent strength to weight ratio and good corrosion resistance. They are potentially applicable in the aerospace and automotive industries, in energy systems as well as in medicine, e.g., as surgical, dental and orthopaedic materials, because of their high biocompatibility. However, the application
*
Corresponding author. Tel.: +48 126 17 2929; fax: +48 126 17 3190. E-mail address:
[email protected] (A. Czyrska-Filemonowicz).
of titanium alloys, especially as implant materials, is limited by their relatively low wear resistance and the release of the elements into the cells or tissues in biological environments [1–3]. Therefore, a surface treatment is required to improve the quality of implants, in particular their surface properties. Several surface treatment methods have been used, such as thermal spraying, anodic oxidation, laser treatment, magnetron reactive sputtering technique, ion implantation including plasma-base ion implantation (PBII) and glow discharge assisted surface processing [1–20]. Of these, the last method has the major advantage of producing layers with controlled microstructure
1359-6454/$30.00 2005 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.actamat.2005.05.035
4368
A. Czyrska-Filemonowicz et al. / Acta Materialia 53 (2005) 4367–4377
even on surfaces with sophisticated shapes, as implants require [19,20]. Consequently, the functional properties of the implant materials can be tailored by producing a suitable microstructure. For the last two decades, titanium-base implants (especially the Ti–6Al–4V alloy for endoprostheses) have been treated by nitridation, with or without plasma assistance, and have been widely investigated [1–7, 11–18]. Several studies provided information about hard TiN layers (coatings) formed during surface treatment emphasising the important role of experimental conditions on their properties. However, little information has been given about the layer microstructure and phase composition (as seen by transmission electron microscopy, TEM), which has a significant effect on the functional properties and performance of nitrided materials [5,11,15]. In multicomponent systems, such as titanium-base commercial alloys, it is not easy to predict the layer composition and microstructure that form during surface treatment, especially under non-equilibrium conditions as in plasma nitridation. Therefore, characterisation of surface layers by advanced analytical techniques, especially by cross-section TEM, is mandatory. The present study is related to the development of graded titanium alloys with improved mechanical and tribological properties, especially for medical application. It is focused on a detailed characterisation of microstructure as well as the surface topography of the multilayers formed on the titanium-base alloys nitrided under glow discharge. We correlate the microscopy findings with the micromechanical and tribological properties of the layers, with the intention of better understanding the structural features influencing the quality of metallic implants.
2. Experimental details The investigation was performed on the two commercial titanium-base alloys, Ti–1Al–1Mn (near-a alloy) and Ti–6Al–4V (a + b alloy). Their chemical compositions are given in Table 1. They were delivered as annealed (700 C/2 h) and then subsequently nitrided under glow discharge at temperatures of 850 C (Ti– 1Al–1Mn) or 900 C (Ti–6Al–4V) in a nitrogen atmosphere at a pressure of 4 hPa for 4 h. The microstructure analyses were performed using light microscopy (LM), X-ray diffraction (XRD), scanning (SEM) and transmission (TEM) electron microscopy. The microstructure of the bulk material (substrate) was examined by LM and by TEM as received and after the nitridation process. Thin foils of the bulk materials were prepared by conventional double-jet electropolishing with a Tenupol and A3 electrolyte
Table 1 Chemical composition of the investigated alloys (wt%); manufacturerÕs data Alloy element
Ti–1Al–1Mn
Ti–6Al–4V
Aluminium Vanadium Iron Oxygen Nitrogen Hydrogen Carbon Manganese Chromium Silicon Titanium
1.0 – 0.3 – – – – 0.9 0.3 0.12 Bal.
6.0 4.0 0.25 0.2 0.05 0.015 0.08 – – – Bal.
from Struers cooled down to 5 C and under 25 V polarisation. The thickness of the surface layers was determined on the cross-sectional specimens using LM. The surface layer microstructure was investigated by TEM on cross-sectional thin foils as well as on plan-view thin foils. The thin foils from the layers were prepared by dimpling and subsequently ion-beam thinning using PIPS of Gatan. The details of thin foil preparation are given in [21]. The TEM investigations were carried out using a JEOL JEM-2010 ARP, a Philips CM20T and a Philips CM300 UT/FEG equipped with scanning transmission electron microscopy attachment (STEM), energy dispersive X-ray spectrometer (EDS) and electron energy-loss spectrometer (EELS). Phase identification was performed by means of XRD, electron diffraction and EDS. Nitrogen concentration was estimated by EELS. The selected area electron diffraction (SAED) and nanodiffraction patterns were interpreted with the JEMS software [22]. The crystallographic data on the identified phases are given in Appendix 2. Image analysis and quantitative TEM metallography were performed using ‘‘Aphelion’’ and ‘‘AnalySIS’’ programmes. The topography of the surface layers was examined in air by an atomic force microscope (AFM) working in a contact mode, designed in the Institute of Nuclear Physics of the Polish Academy of Sciences [23]. Standard Si3N4 cantilevers (Vecco Instruments), characterised by a spring constant of 0.03 and 0.1 N/m and tip radius curvature of 30 nm, were used. A Micro-Combi Tester (MCT) of CSEM Instruments [24] was used to measure the micromechanical properties (microhardness and YoungÕs modulus) and to perform the scratch tests. Friction wear resistance was tested by means of the ‘‘three rollers + taper’’ (load up to 400 MPa) method according to the Polish standard [25], as well as by means of the ‘‘ball-on-disc’’ method using a CSEM tribometer [24], where an Al2O3 ball of 4 mm in diameter was rotated under a load of 10 N
A. Czyrska-Filemonowicz et al. / Acta Materialia 53 (2005) 4367–4377
4369
(according to ASMT G99-90 and DIN 5032E standards). The experimental details are given in [20,21].
3. Results and discussion 3.1. Microstructure 3.1.1. Near-a alloy: Ti–1Al–1Mn As received alloy. The microstructure of the as received alloy (Figs. 1(a) and (b)) consists of a phase platelets (hexagonal close-packed (hcp); thickness of about 2.2 lm) within the b phase (body-centred cubic (bcc)) grains (size of 50–140 lm). The amount of the b phase measured by XRD is low (3.4%) as expected for a ‘‘near a alloy’’, where the b phase volume fraction must not exceed 8%. Element distribution maps, recorded using the STEM-EDS method, shows an enrichment of Al in a phase and of Mn in the b phase (Fig. 1(c)). Nitrided alloy. The mean thickness of the surface layer on the nitrided alloy (Fig. 2(a)) was estimated on the LM cross-section specimens as 32 lm. Analytical TEM investigation performed on crosssectional thin foils (Fig. 2(b)) revealed that the nitriding under glow discharge led to the formation of a complex diffusion layer on the top of the bulk material (substrate). It consists of four sublayers (from the surface to the substrate) containing mainly: nanocrystalline d-TiN (face centred cubic NaCl type [26]) labelled ‘‘1’’ on the Fig. 2(b), d 0 -Ti2N (body-centred tetragonal) labelled ‘‘2’’, e-Ti2N (tetragonal primitive) labelled ‘‘3’’ and a (N) (hcp) solid solution (labelled ‘‘4’’) enriched by nitrogen up to 19 at.% (Fig. 2(c)). The microstructure of the underlying substrate after nitridation consists of a phase platelets (thickness of about 1.0 lm) within large grains of the b phase (grain size of about 1.5 mm). The amount of the b phase was measured as 8% (volume). 3.1.2. a + b alloy: Ti–6Al–4V As received alloy. The microstructure of the as received alloy consists of the a phase (hcp) with 12 vol.% of the b phase (bcc). The size of the equiaxed a phase grain was estimated to be about 5 lm. Element distribution maps, recorded using the STEM-EDS method, revealed Al enrichment of a phase and V and Fe enrichment of b phase. Nitrided alloy. Nitridation modified the original microstructure down to some 130–200 lm (Fig. 3(a)– (d)). LM shows that a continuous surface layer about 30 lm thick was formed above an intermediate layer containing large a-grains enriched with N grown penetrating into a + b fine microstructure (Fig. 3(a)). No sharp transition between this intermediate nitrided layer and the substrate was observed.
Fig. 1. Microstructure of the as received Ti–1Al–1Mn alloy: (a) as seen by LM; (b) TEM bright field image and the corresponding SAED patterns taken from a and b phases; (c) elemental distribution maps, STEM-EDS.
4370
A. Czyrska-Filemonowicz et al. / Acta Materialia 53 (2005) 4367–4377
(size of some 300–500 lm). The a platelets behave different orientations within the b grains and form a characteristic Widmansta¨tten structure, similar to that described in [27]. The volume fraction of the b phase measured by XRD was 18%. The TEM investigation of the cross-section thin foils revealed that the continuous surface layer seen by LM consists of four sublayers (Fig. 3(c)). Bright-field (BF) images and SAED patterns show that the outermost sublayer (thickness of 0.3–0.5 lm) is nanocrystalline (Fig. 3(d)) and consists of the d-TiN (face centred cubic NaCl type, also called ‘‘osbornite’’). Several reflections on ring patterns recorded in the nanocrystalline sublayer but along the interface with the underlying one could only be explained by the presence of Ti2AlN (H-phase) suggesting that small grains of this phase may also form close to this interface (Fig. 4). Sparse reflections on SAED patterns from this nanocrystalline layer suggested the presence of minor phases, and some fits were found with large cell Ti–Al–N ternaries, but the present information was insufficient to assess any of them with reliability or to exclude double diffraction effects (see comments on phase determination in Appendix 1). Below this nanocrystalline sublayer, crystals are larger. Mostly d 0 -Ti2N (body-centred tetragonal) and e-Ti2N (tetragonal primitive) with some Ti2AlN (hexagonal H-phase) were found in the second and third sublayers. The fourth sublayer is identified as a (N) solid solution enriched by nitrogen. A feasibility measurement by TEM-EELS showed that its nitrogen content varied from 24 at.% at a 9 lm depth under the specimen surface to 8 at.% at 20 lm. The results of microstructural investigations of other a + b alloys for implants (Ti–5Al–2.5Fe and Ti–6Al– 3Cr–2Mo alloys) nitrided under glow discharge also showed the presence of the TiN and Ti2N phases within the nitrided layers. Also an enrichment of a solid solution in nitrogen was observed; however, quantitative data were not provided [4,11,15,19]. 3.2. Properties and adhesion of the layers
Fig. 2. Microstructure of the nitrided Ti–1Al–1Mn alloy: (a) crosssection LM image of the nitrided layer on the underlying bulk material (substrate); (b) cross-section TEM BF image of the surface multilayer and corresponding SAED patterns taken from the sublayers labelled 1, 2, 3, 4; (c) EELS spectrum taken from the 4th sublayer a (N) solid solution enriched with nitrogen.
The bulk (substrate), presented in Fig. 3(b), consists of colonies of a phase platelets (a platelets thickness in a range of 0.5–2.0 lm) within the prior grains of b phase
The microstructure findings described above were correlated with the alloy micromechanical and tribological properties. We found in earlier studies [28–30] that the nitrided titanium alloys exhibited a higher hardness than the untreated alloys. TEM observations have now revealed that the nitrided layers have a more complex microstructure. This led to new systematic measurements of the micromechanical properties (microhardness, YoungÕs modulus) of the layers, performed on the cross-section samples at increasing depth from 2 to 1000 lm under the surface. The results presented in Table 2 and
A. Czyrska-Filemonowicz et al. / Acta Materialia 53 (2005) 4367–4377
4371
Fig. 3. Microstructure of the nitrided Ti–6Al–4V alloy: (a) LM of the bulk material (substrate); (b) LM cross-section image of the surface and intermediate layer; (c) TEM BF and SAED images of the surface multilayer; (d) TEM BF image of the outermost d-TiN nanocrystalline sublayer.
Fig. 5(a) and (b) show, despite a pronounced scatter of the measurements, that the outermost nanocrystalline dTiN on the top of the sublayers composed of Ti2N and Ti2AlN crystals had an even higher microhardness and YoungÕs modulus than the underlying coarse grain sublayers. Tribological tests, performed by the ‘‘three rollertaper’’ method (Fig. 6), showed that the nitrided multi-
layers, formed on both alloys by glow discharge at various conditions at temperature range of 850–1000 C, always exhibited a higher resistance to frictional wear (measured as the wear depth) in comparison to the untreated alloys [18,19,29]. The samples of the as received alloy (without surface treatment) suffered seizure after a few minutes testing; it means that the untreated alloy has poor resistance.
4372
A. Czyrska-Filemonowicz et al. / Acta Materialia 53 (2005) 4367–4377
Fig. 4. Compound diffraction pattern obtained by superimposing two SAED patterns. The first one, highlighted by the 2-D lattice, was taken along the [1 2 1] zone axis on a large e-Ti2N grain found in the third sublayer. It is used for re-calibration of the diffraction camera length under actual conditions. The second one comes from the deepest part of the nanocrystalline sublayer and a small part of the underlying second layer. It contains reflections of several grains that best fit the position of the calculated rings for polycrystalline d-TiN and Ti2AlN (H-phase) among all phases considered so far.
Fig. 7 shows the results of the frictional wear resistance tests performed by the ‘‘ball-on-disc’’ method using a CSEM tribometer. This is a suitable tool for Table 2 Microhardness and YoungÕs modulus of the nitrided Ti–1Al–1Mn and Ti–6Al–4V alloys Alloy
Depth (lm)
Microhardness, HV0,005
YoungÕs modulus (·105 Mpa)
Ti–1Al–1Mn
2 3 10 20 30 40 1000 2 3 10 20 30 200 1000
1533 ± 152 1054 ± 74 706 ± 66 523 ± 71 255 ± 15 265 ± 18 230 ± 11 1579 ± 96 893 ± 53 696 ± 47 442 ± 17 419 ± 33 255 ± 17 239 ± 29
1.95 ± 0.16 1.60 ± 0.15 1.34 ± 0.10 1.23 ± 0.07 1.17 ± 0.11 1.05 ± 0.11 1.06 ± 0.11 1.96 ± 0.12 1.44 ± 0.12 1.30 ± 0.11 1.10 ± 0.04 1.09 ± 0.10 1.00 ± 0.06 0.98 ± 0.10
Ti–6Al–4V
determination of the wear of the hard layers by measuring the appropriate wear track profile. The results obtained confirmed the previous findings. Both nitrided alloys exhibited much better wear resistance than the untreated alloys. In particular, for the Ti–6Al–4V alloy, the track depth was measured as 9.0 lm in the case of the untreated alloy surface, while it was only 0.8 lm for the nitrided alloy. The coefficients of friction were measured as 0.4 and 0.13, respectively [31]. As shown previously, the multilayer on the nitrided Ti–6Al–4V alloy consists of several multiphase sublayers. The presence TiN and Ti2AlN phases contribute to high microhardness, while Ti2N layer is known for its excellent wear resistance [10,17]. One of the most important properties of the surface layers, especially for implants, is their adhesion to the underlying bulk (substrate). This property was investigated by scratch tests. Our previous investigation [19,28,29] showed that in the case of Ti–1Al–1Mn alloy nitrided at 730, 850 and 1000 C, the layers exhibit a good adhesion to the underlying bulk materials. The
A. Czyrska-Filemonowicz et al. / Acta Materialia 53 (2005) 4367–4377
4373
Fig. 7. Frictional wear resistance of the Ti–1Al–1Mn and Ti–6Al–4V alloys nitrided under glow discharge compared with the wear of untreated alloys.
Fig. 5. (a) Microhardness and (b) YoungÕs modulus of the Ti–1Al– 1Mn and Ti–6Al–4V alloys nitrided under glow discharge.
best adhesion during the test within the load range examined (up to 25 N) was observed for the layers produced at 1000 C. Also the results of the scratch test of the Ti–6Al–4V alloy showed that no flaking nor delamination occurred for both alloys investigated during testing under a load of 25 N [21,30]. The TEM microstructure investigation showed that the multilayers produced during nitriding on both alloys have a diffusion character and do not exhibit a sharp
Fig. 6. Linear wear resistance of the Ti–6Al–4V alloy nitrided at 850 and 1000 C.
interface with the underlying substrates. Especially in a case of Ti–6Al–4V alloy, a wide sublayer of nitrogenrich a phase (thickness of about 100–170 lm) adjacent to the substrate is clearly visible (Fig. 3(a)). This morphology may contribute to a good adhesion of the nitrided layers formed by diffusion under glow discharge on the underlying substrates, which is one of the most important features for application of coatings for service. The results presented above show that nitriding under glow discharge is an effective method for the improvement of the mechanical and tribological properties of titanium alloys by formation of diffusion hard surface layers. Although an improvement in the mechanical and tribological properties of Ti–6Al–4V alloy by various plasma-assisted nitridation methods is not a new finding, the multiphase layer composition and micro/nanostructure as seen by cross-sectional TEM is generally not precisely described in the literature. It was reported that plasma assisted nitridation produce the hard surface layers of TiN/Ti2N [1–3,5,11,12,16–18]. The layer microstructural investigations, published in the literature, were performed mainly by scanning electron microscopy and X-ray diffraction. Some phases were identified (d-TiN, e-Ti2N, Ti2AlN), but a lack of cross-sectional TEM investigations did not allow the identification of the phases within very thin layers or multilayers consisting of thin sublayers, e.g., for differentiation between a body-centred tetragonal Ti2N (also known as d 0 -Ti2N) and tetragonal primitive Ti2N (also known as e-Ti2N). Also unambiguous identification of nano-sized phases which may form during nitridation is not possible without applying advanced TEM. In the present study, analytical TEM and electron diffraction (SAED, nanodiffraction) supported by
4374
A. Czyrska-Filemonowicz et al. / Acta Materialia 53 (2005) 4367–4377
Fig. 8. Surface topography of the nitrided Ti–1Al–1Mn (a) and Ti–6Al–4V (b) alloys; AFM.
JEMS software allowed these difficulties to be overcome. However, EDS was not sufficient for phase identification, but it did lead to the definition of a set of more than 20 phases containing one or more elements from the Ti–N–Al–O system among all phases referenced in the PCPDF and ICSD bases. Several of them have close lattices or large unit cells and special care was required in the systematic phase identification using the electron diffraction pattern indexation by JEMS. These results should contribute to redressing the lack of information about phase composition and morphology of the nitrided layers described in the literature. Furthermore, the investigations of layers by crosssectional TEM supported by image analysis revealed a nanocrystalline morphology of the outermost d-TiN sublayer formed on the Ti–1Al–1Mn and Ti–6Al–4V alloys during nitridation. This is an important finding of this study. Consequently, the desired properties of these alloys, which are closely related to their micro/nanostructures, can be suitable tailored. 3.3. Surface topography Surface topography has a significant contribution to the mechanical performance of the surface treated alloys, especially in terms of their hardness and wear resistance. Surface topography is also one of the most important properties that determines implants quality. The femoral head of an implant should be as smooth as possible, whereas it could be advantageous to increase the roughness of the stem and other parts of the implant system that should adhere to bone tissue [32–34]. However, increased roughness has a deleterious effect on the corrosion resistance of implants [34]. Therefore, the investigation of surface topography of the nitrided alloys was performed by AFM. Fig. 8 shows typical images of surface topography of the nitrided
Table 3 Results of AFM measurements (Rq, Rt and SAF surface parameters) of the nitrided Ti–1Al–1Mn and Ti–6Al–4V alloys Surface parameters
Ti–1Al–1Mn
Ti–6Al–4V
Rq (lm) Rt (lm) SAF
0.26 ± 0.08 1.87 ± 0.45 1.07 ± 0.08
0.38 ± 0.13 2.37 ± 0.71 1.13 ± 0.04
Ti–1Al–1Mn and Ti–6Al–4V alloys. The values of Rq1 (RMS), Rt2 and SAF3 are given in the Table 3. The increased values of Rq and SAF parameters and the slope of the bearing area fraction curve (the so-called Abbott–Firestone curve) [30] indicate that both nitrided layers exhibited pronounced surface development. The SAF parameter, which is lower for the Ti–1Al–1Mn alloy, while the corresponding Rq and Rt are higher, suggests that the peaks and valleys on this surface were smaller, but their distribution was more dense than in the case of the nitrided Ti–6Al–4V alloy. A pronounced surface development plays an important role in increasing mechanical and tribological properties of the investigated alloys. It also has an influence on the cell proliferation and viability of human cells. The study of the cellular behaviour performed on the nitrided Ti–1Al–1Mn alloy [19] showed an improvement in cell proliferation and viability of human fibroblasts compared to the untreated alloy. This study also demonstrated a good corrosion resistance in a biological medium and no metal ion release into the cell culture environment. Improved corrosion resistance and biocompatibility with human cells of the Ti–1Al–1Mn and Ti–6Al–4V alloys nitrided under glow discharge, 1 Rq (lm) – the root mean square (RMS) of the height values of all points of the 3-D scan. 2 Rt (lm) – the maximum peak-to-valley height of the entire image. 3 SAF (surface area factor) – the ratio of the real 3-D surface area to that of its 2-D projection in a direction normal to the average sample surface.
A. Czyrska-Filemonowicz et al. / Acta Materialia 53 (2005) 4367–4377
established that in vitro and in vivo experiments [19,35,36] are of importance for future various clinical application of these materials.
4. Conclusions Two titanium alloys, Ti–1Al–1Mn and Ti–6Al–4V, were nitrided under glow discharge and subsequently their microstructure and surface topography were characterised using various analytical methods, including TEM and AFM. Analytical TEM investigation revealed that diffusion under a nitrogen glow discharge forms a complex multilayered micro/nanostructure on both alloys. The multilayer formed on the Ti–1Al–1Mn alloy is composed of the d-TiN (osbornite), d 0 -Ti2N, e-Ti2N and nitrogen-rich a solid solution. The multilayer on the Ti–6Al–4V alloy contains, besides the phases mentioned above, Ti2AlN, Ti4AlN3 or Ti3Al2N2 and possibly Ti3AlN as minor phases. The fairly high number of possible Ti–Al–N–(O–V or –Mn) phases and the similarity of the crystal lattice of several of them required special care in phase identification by TEM. Ambiguities in assessing phases to diffraction patterns require an accuracy of the camera length better than 1% or consideration of the relative intensities between reflections as well as analytical information from EDS or EELS spectrometries. From the TEM results, it was concluded that the better mechanical and tribological properties of the nitrided alloys are related to the presence of the nanocrystalline d-TiN outermost sublayers on the top of the sublayers composed of Ti2N and some Ti2AlN phases. AFM investigation showed that the layers exhibit a pronounced surface development, which contribute to improved hardness, wear resistance and biocompatibility of the nitrided alloys. The results show that nitridation under glow discharge is an effective method for improvement of the mechanical and tribological properties of titanium-base alloys by formation of graded diffusion hard surface layers.
Acknowledgements The study was partially supported by the State Committee for Scientific Research (Projects No 4 T08C 02424 and PBZ-KBN 082/T08/2002). Valuable contributions of Prof. Stadelmann (EPFL), Dr. Spiradek-Hahn (ARCS), Mrs. Biel, Dr. Kot and Dr. Zimowski (AGH-UST) as well as the AGH-UST, ARCS and EPFL technical staff are kindly acknowledged.
4375
Appendix 1. Comment on phase determination The Ti–Al–N system contains numerous binary and ternary phases. The designation and existence of phases is not always clear in the literature and the following references may be useful [26,37–47]. Moreover, some oxygen may react at the surface during high temperature glow discharge treatment. Crystallographic data of the eight identified phases plus TiO (hongquiite) are summarised in Appendix 2. In total, more than 40 binary or ternary phases have been considered; some of them have very similar crystallographic structures. For instance considering only Fm–3m structures TiN (referred as d-TiN on phase diagrams, mineral name ‘‘osbornite’’ a = 0.4235– 0.424 nm) has nearly the same lattice parameter as Ti2AlN (a = 0.419 nm) and non-stoichiometry may shift its lattice parameters up to some 1%. The presence of aluminium in the alloy may also lead to formation of AlN (a = 0.437–0.431 nm) or to some Al substitution in TiN giving a Ti1xAlxN with a = ad-TiN-0.008 nm for x = 0.46 [43]. If some oxidation during nitriding is suspected, TiO (‘‘hongquiite’’ a = 0.417–0.429 nm) may also be considered. The diffraction patterns of these phases are hardly distinguishable within usual experimental errors and the diffraction camera has to be calibrated with an accuracy better than 1%. Moreover large cell hexagonal compounds as P6/ mmc Ti2AlN (H-phase a = 0.2994, c = 1.360 nm), Ti4AlN3 (a = 0.299, c = 2.340 nm) and P63mc Ti3Al2N2 (a = 0.299, c = 2.340 nm) give a high number of reflections that fit sparse reflections on ring patterns or zone axis spot patterns of many other structures within a few per cent accuracy on spot distances and angles. All polycrystalline (ring) and single crystal zone axis (spot) diffraction patterns were analysed with the JEMS software [22] by scrolling through all the candidate phases. But under such circumstances, even an accurate calibration of the electron diffraction camera was often insufficient to choose on the basis of geometrical arguments between several suggested solutions that have nearly identical rings or zone axis patterns. Semi-quantitative comparison of intensities on experimental diffraction patterns with the kinematical, or even dynamical, calculated ones allowed the rejection of some solutions by considering spot extinctions or relative intensities of reflections, being aware of thickness and double diffraction effects. EDS analysis provides some help in distinguishing between phases. However, it has to be considered as a semi-quantitative method at best. Some of the electrons scattered by a targeted grain hit bulk parts of the surrounding sample. In the present case, it leads to a spectrum that contains a significant contribution from the average composition of the bulk alloy in addition to that one from the aimed grain itself [48]. This explains the apparently abnormally high content of Ti and the pres-
4376
A. Czyrska-Filemonowicz et al. / Acta Materialia 53 (2005) 4367–4377
Appendix 2. Summary of the crystallographic data of the identified phases Phase name
Lattice (nm)
Space group
Pearson symbol
Wyckoff atom position and occupancy
d-TiN ICSD#64905 ‘‘Osbornite’’ PDF# 381420 (or ICSD#64907) (or ICSD#64908) TiO ICSD#40125 ‘‘Hongquiite’’ PDF#862352 TiAlN2 ICSD#58012 no PDF number
Cubic a = 0.4244 (a = 0.4320) (a = 0.4225) Cubic a = 0.4293 Cubic a = 0.419
Fm-3m 225
cF8
Ti, a, 0, 0, 0, 1 N, b, 0.5, 0.5, 0.5, 1
Fm-3m 225 Fm-3m 225
cF8
Ti3AlN ICSD#52642 PDF#371140 (a = 0.2988 in ICSD#52642 is a transcription error) Ti2AlN ICSD#52641 ‘‘H-phase’’ PDF#1870 Ti4AlN3 ICSD#91772 Ti4AlN2.89 PDF#00-053-0444
Cubic a = 0.4112
Pm-3m 221
cP5
Ti, a, 0, 0, 0, 1 O, b, 0.5, 0.5, 0.5, 1 Ti, a, 0, 0, 0, 0.5 Al, a, 0, 0, 0, 0.5 N, b, 0.5, 0.5, 0.5, 1 Ti, c, 0, 0.5, 0.5, 1 Al, a, 0, 0, 0, 1 N, b, 0.5, 0.5, 0.5, 1
Hexagonal a = 0.2994 c = 1.361 Hexagonal a = 0.2991 c = 2.340
P63/mmc 194
hP8
P63/mmc 194
hP16
Ti3Al2N2 ICSD#71097 PDF#371141 and 802286
Hexagonal a = 0.2988 c = 2.335
P63 mc 186
hP14
e-Ti2N ICSD#33715 PDF#760198
Tetragonal (primitive) a = 0.4945 c = 0.3034 Tetragonal (body-centred) a = 0.414 c = 0.8805 (c = 0.8786)
P42/mnm 136
tP6
I41/amd 141
tI12
d 0 -Ti2N ICSD#23403 PDF#730959 (PDF#771893)
ence of some Al, V or Mn in spectra of phases identified by electron diffraction as Ti–Al–N ternaries and TiN or Ti2N. Furthermore, some bending of the thin crosssection foils, due to strain relaxation during preparation, sometimes introduced unexpectedly low energy X-rays absorption, which led to an underestimation of the O, N and possibly Al contents.
cF8
Ti, f, 0.3333, 0.6667, 0.086, 1 Al, d, 0.3333, 0.6667, 0.75, 1 N, a, 0, 0, 0, 1 Ti, f, 0.3333, 0.6667, 0.0542, 1 Ti, e, 0, 0, 0.1547, 1 Al, c, 0.3333, 0.6667, 0.25, 1 N, a, 0, 0, 0, 1 N, f, 0.6667, 0.3333, 0.105, 1 N, b, 0.3333, 0.6667, 0.05, 0.9 Ti, b, 0.3333, 0.6667, 0.1, 0.1 Ti, b, 0.3333, 0.6667, 0.2, 0.9 N, b, 0.3333, 0.6667, 0.25, 0.1 Al, b, 0.3333, 0.6667, 0.4, 1 N, b, 0.3333, 0.6667, 0.55, 0.1 Ti, b, 0.3333, 0.6667, 0.6, 0.9 Ti, b, 0.3333, 0.6667, 0.7, 0.1 N, b, 0.3333, 0.6667, 0.75, 0.9 Ti, a, 0, 0, 0, 1 Al, a, 0, 0, 0.3, 1 Ti, f, 0.296, 0.296, 0, 1 N, a, 0, 0, 0, 1
Ti, e, 0, 0.25, 0.14, 1 N, e, 0, 0.25, 0.375, 1
References [1] Brunette DM, Tengvall P, Textor M, Thomsen P. Titanium in medicine. Berlin: Springer-Verlag; 2001. [2] Peters M, Leyens Ch. Titan und Titanlegierungen. Weinheim: Wiley-VCH Verlag; 2002. [3] Luetjering G, Albrecht J, editors. Ti-2003 science and technology. Weinheim: Wiley-VCH Verlag; 2004.
A. Czyrska-Filemonowicz et al. / Acta Materialia 53 (2005) 4367–4377 [4] Bordji K, Jouzeau JY, Mainard C, Payan E, Netter P, Rie KT, et al. Biomaterials 1996;9:929. [5] Raveh A, Hansen PL, Avni R, Grill A. Surf Coatings Technol 1989;38:339. [6] Long M, Rack HJ. Biomaterials 1998;19:1621. [7] Sitting C, Textor M, Spencer ND, Wieland M, Vallotton PH. J Mater Sci: Mater Med 1999;10:35. [8] Matsuura K, Kudoh M. Acta Mater 2002;50:2693. [9] Meltis EI. Surf Coatings Technol 2002;149:95. [10] Matthes B, Broszeit J, Aromaa J, Ronkainen H, Hannula SP, Leyland AMA. Surf Coatings Technol 1991;49:489. [11] Gołe˛biewski M, Kruzel G, Major R, Mroz B, Wierzchon´ T, Ebner J, et al. Mater Chem Phys 2003;81:315–8. [12] Rie KT, Stucky T, Silva RA, Letao E, Bordji K, Jouzeau JY, et al. Surf Coatings Technol 1995;74–75:973. [13] Pinkas M, Frage N, Froumin N, Pelleg J, Dariel MP. J Vac Sci Technol, A 2002;20:887. [14] Toth A, Mohai M, Ujavari T, Bell T, Dong H, Bertoti I. Surf Coatings Technol 2004;186:248. [15] Wan GJ, Huang N, Leng YX, Yang P, Chen JY, Wang J, et al. Surf Coatings Technol 2004;186:136. [16] Fouguet V, Pichon L, Straboni A, Drouet M. Surf Coatings Technol 2004;186:34. [17] Wei R, Booker T, Rincon Ch, Arps J. Surf Coatings Technol 2004;186:305. [18] Lanning BR, Wei R. Surf Coatings Technol 2004;186:314. [19] Czarnowska E, Wierzchon T, Maranda-Niedbala A, Kaczmarewicz E. J Mater Sci: Mater Med 2000;11:73. [20] Wierzchon T. Mater Sci Forum 2003;426–432:2563. [21] Łucki M. PhD thesis, AGH University of Science and Technology, Krako´w 2004 [in Polish]. [22] Stadelmann P. JEMS Java Electron Microscopy Software. Available from: http://cimewww.epfl.ch/people/stadelmann/jemsWebSite/jems.html. [23] Lekki J, Lekka M, Romano H, Cleff B, Stachura Z. Acta Phys Polonica A 1996;3:315. [24] Randal N, editor. Advances in surface mechanical properties characterization. CSEM Instruments, Applications Bulletin; 1996. Available from: http://www.csem.ch/instrum. [25] Polish Standard PN-83/H-D4302. Strength tests of metals. [26] Brager A. Acta Physicochim 1939;9:617.
4377
[27] Bhattacharyya D, Viswanathan GB, Denkenberger R, Furrer D, Fraser Hamish L. Acta Mater 2003;51:4679. [28] Czyrska-Filemonowicz A, Buffat PA, Moskalewicz T, SpiradekHahn K. In: Matsuda, K editor. Proceedings of the interanational seminar on nanotechnology for fabrication of hybrid materials, 6-9.11.2002, Toyama, Japan, 2002. p. 43. [29] Moskalewicz T, Rakowski W, Czyrska-Filemonowicz A. Eng Biomater 2002;23–25:33. [30] Czyrska-Filemonowicz A, Buffat PA, Łucki M, Moskalewicz T, Wierzchon´ T. In_zynieria Materiałowa 2004;3:335. [31] Rakowski W, Kot M, Moskalewicz T, Czyrska-Filemonowicz A. Wear [in preparation]. [32] Wennerberg A. Cells Mater 1999;9:1. [33] Ronold HJ, Ellingsen JE, Lyngstadaas SP. J Mater Sci: Mater Med 2003;14:10. [34] Ungersbock A, Rahn B. J Mater Sci: Mater Med 1994;5– 7:434. [35] Czyrska-Filemonowicz A, Buffat PA, Czarnowska E, Wierzchon´ T. Mater Sci Forum [in press]. [36] Cukrowska B, Sowin´ska A, Zaja˛czkowska A, Sobiecki JR, Wierzchon´ T, Czarnowska E. Ann Transplant 2004;9(1A):76. [37] Anderbouhr S, Gilles S, Blanquet E, Bernard C, Madar R. Chem Vapour Depos 1999;9:109. [38] Schuster JC, Bauer J. J Solid State Chem 1984;53:260. [39] Yue R, Wang Y, Chen Ch. Surf Interface Anal 1999;27:98. [40] Chen G, Sundman B. J Phase Equilib 1998;19:146. [41] Bahr JP, Etchessahar E, Debuigne J. J Less Common Met 1977;52:51. [42] Arbuzov MP, Golub SA, Khaenko BV. Izvestiya Akademii Nauk SSSR, Neorganiczeskie Materialy 1977;10:1779. [43] Knotek O, Leyendecker T. J Solid State Chem 1987;70:318. [44] Toth LE. Transition metal carbides and nitrides. New York (NY): Academic Press; 1971. [45] Lobier G, Marcon JP. Comptes Rendus Hebdomadaires des Se´ances de lÕAcademie des Sciences, Se´rie C 1969;268:1132. [46] Christensen AN, Alamo A, Landesman JP. Acta Crystallogr C 1985;45:1009. [47] Durlu N, Gruber U, Pietzka MA, Schmidt H, Schuster JC. Z Metallkd 1997;88:390. [48] Buffat PA. Yearbook 2004. European Microscopy Soc, 1609-1191 2004;78.