Nano-scale wear characterization of CoCrMo biomedical alloys

Tribology International ∎ (∎∎∎∎) ∎∎∎–∎∎∎

Contents lists available at ScienceDirect

Tribology International journal homepage: www.elsevier.com/locate/triboint

Nano-scale wear characterization of CoCrMo biomedical alloys V. Martinez-Nogues a,n, J.M. Nesbitt b, R.J.K. Wood a, R.B. Cook a a National Centre for Advanced Tribology, nCATS, Faculty of Engineering and the Environment, University of Southampton, University Road, Southampton SO17 1BJ, UK b RedLux Ltd, 3/7 Avenger Close, Chandlers Ford, Southampton SO53 4DQ, UK

art ic l e i nf o

a b s t r a c t

Article history: Received 6 October 2014 Received in revised form 12 February 2015 Accepted 26 March 2015

Low amplitude motions at the micro and the nano-scale at the femoral stem–cement interface under physiological loads can result in fretting and nano-wear on the stem surface. These are important wear processes in cemented total hip replacements as the release of metal debris and ions can trigger adverse local tissue reactions within the body, bone resorption and subsequent aseptic loosening of the femoral component resulting in the implant failure. However, the influence of the microstructure and manufacturing processes on the nano-wear behaviour of different cobalt chromium molybdenum (CoCrMo) alloys has not been studied extensively. Four CoCrMo alloys were tested under reciprocating wear conditions at the nano-scale level. Tangential friction forces, coefficient of friction and plastic deformation values were recorded. A new white-light-interferometer system was validated against atomic force microscopy and Nano Vantage Test System measurements to analyse the permanent plastic deformation caused in each of the samples. Significant differences were found in the total plastic deformation achieved by the as cast alloy compared to the forged, as cast single thermal treated and as cast double thermal treated samples. In addition thermal treated samples presented a tendency to produce a higher quantity of wear debris around the nano-wear scars. These findings indicate a possible relation between the wear resistance at the nano-scale and the manufacturing and thermal processes applied on the CoCrMo biomedical alloys. Crown Copyright & 2015 Published by Elsevier Ltd. All rights reserved.

Keywords: Nano-wear Single asperity contact Femoral stem–cement interface White-light-interferometry

1. Introduction In vivo wear of metallic materials have been demonstrated to result in the generation of metal particles and metallic ions which can cause adverse tissue reactions or bone resorption [1] and therefore the loosening of the implant and its revision. Wear particles are generated at the articular surface of total hip arthroplasties (THA), [2], but they can also be generated at other interfaces, in particular the femoral stem–cement or the taper– trunnion interfaces [3–5]. In cemented total hip replacements, (THR), the difference between the mechanical properties of the cement and the metallic femoral stem under loading, results in motion between the two surfaces. This motion generates cement or metallic particles which can act as single asperities increasing the maximum contact pressure applied that combined with the micromotions can accelerate the nano-scale wear happening at the femoral–cement interface [6]. Retrieval studies [2,6–8] have shown that the most

n

Corresponding author. E-mail address: [email protected] (V. Martinez-Nogues).

affected zones are mainly localised in the posterior and medial regions in the femoral stem. The level of damage has been found been dependent of the final surface roughness of the stem, the pores produced after the polymerization of the cement in contact with the metallic surface [9,10], the type of cement [11,12] and/or the presence of hard radiopacifier particles in it which can produce ploughing at the CoCrMo stem surface accelerating the fretting wear processes [13]. Different fretting and nano-scale wear mechanisms affecting both polished and matte femoral stems have been observed previously, [2]. Ductile wear together with pitting of the surface was found in polished stems, while abrasive wear processes were found in the matte ones. Recently, Bryant et al. quantified the tribochemical reactions occurring at the stem–cement interface and their influence on the stem degradation and the debris released into the biological environment [8,14]. They proposed a mechanism for the formation and transfer of the oxidised metallic debris from the metallic stem to the bone cement. The motion between the two contacting surfaces remove the passive film created on the CoCrMo alloy exposing a new reactive surface in contact with the liquid environment causing the oxidation of the metallic substrate and metallic debris. Due to the cyclic loading and the contact between the surfaces these particles

http://dx.doi.org/10.1016/j.triboint.2015.03.037 0301-679X/Crown Copyright & 2015 Published by Elsevier Ltd. All rights reserved.

Please cite this article as: Martinez-Nogues V, et al. Nano-scale wear characterization of CoCrMo biomedical alloys. Tribology International (2015), http://dx.doi.org/10.1016/j.triboint.2015.03.037i

V. Martinez-Nogues et al. / Tribology International ∎ (∎∎∎∎) ∎∎∎–∎∎∎

2

get compacted creating Cr2O3 layers which remain between both surfaces. This type of mechanism was also observed at the taper– trunnion interface as reported by Gilbert et al. [15]. The findings by Cook et al. [13] of hard zirconium dioxide radiopacifier particles between the femoral stem and the cement surfaces provides a field of investigation on the plastic deformation produced by these hard asperities in the metallic surface at the nanoscale. The existence of different CoCrMo biomedical alloys due to manufacturing processes and thermal treatments history provide an extensive range of microstructures and mechanical properties which can affect the deformation behaviour and therefore the nano-wear-corrosion processes. The role of the size grain, the presence or absence of the carbides in addition to their morphology has been reported playing an important role in the sliding or abrasive wear of CoCrMo at the articular surfaces and also in their corrosion processes [16–20]. Nano-wear testing of different biomedical materials, accelerated micro-wear of silicon or nano-scratch testing of different Stellite-21 alloys have been studied previously [21–23], proving the capability of nanomechanical testing to optimise biomaterials and compare the different mechanical properties, wear resistance and wear processes at the micro and nano-scale. This present study aims to establish a better understanding of the wear mechanisms at the nano-scale by comparing four CoCrMo alloys commonly used as implantable materials in total hip replacements. Investigations include mechanical properties measurements, coefficient of friction (COF) evaluation during the running-in and steady state periods, dispersed energy value calculation, plastic deformation measurements validated by the following methods: atomic force microscopy (AFM), (MFP-3D, Asylum Research, USA), white-light-interferometry (OptiLux, RedLux Ltd, UK) and by a mechanical system (NanoTest Vantage System, MicroMaterials Ltd, UK) and nano-wear scar morphology observation by optical microscopy (Alicona Infinite Focus, Alicona, Austria) and scanning electron microscopy (SEM), (JSM 6500F, JEOL, Japan). Knowing the processes which affect the different microstructure of CoCrMo alloys can be essential for the correct choice of the material for future cemented total hip replacement.

2. Material and methods Mirror polished biomedical CoCrMo samples, Ra ¼12 70.5 nm, were used for the present study. Two CoCrMo alloys with different thermal and manufacturing history were used: forged (ASTM F1537), as cast (ASTM F75), as cast single thermal treated (as cast-TT), and as cast double thermal treated (as cast-2TT). 2.1. Mechanical properties Micro Vickers hardness values (HV) were obtained using a square based pyramidal indenter holding a 1 N load constantly during 15 s, (Matzsuzawa Seiki Co. Ltd, Japan). HV values were converted into indentation hardness (HIT) values using Eq. (1) according to Annex F from ISO 14577-1:2002 [24]. Where HV is the Vickers hardness, HIT is the indentation hardness, Ap is the projected area of the indenter, As is the contact surface area and g is the acceleration due to gravity. HV ¼

H IT  AP ¼ 0:0945H IT ðGPaÞ g n  As

ð1Þ

Indentation maps of 20 by 20 indents, under depth control up to 500 nm, were performed on the samples using a diamond Berkovich tip to determine the hardness and elastic modulus. Loading and unloading rates were set at 2 mN/s and a dwell time of 60 s was established at the maximum depth. Thermal drift

values were recorded during 40 s before and after the experiment to correct the obtained values.

2.2. Nano wear experiments Nano-wear experiments were performed using the NanoTest Vantage System. Two different experimental conditions were used to assess the frictional behaviour and the nano-wear resistance of the CoCrMo alloys. To monitor the tangential friction forces a friction probe was attached to the rig during the experiments. A spherical diamond tip of 200 μm was used while the normal loads applied were 5, 10 and 20 mN, which corresponds to maximum contact pressures of 0.97, 1.2 and 1.5 GPa, respectively. COF values were calculated from the ratio of tangential force to the normal force at each load during running-in and steady state periods. Dispersed energy values were calculated from the hysteresis area of the force–displacement curves as a measure of the interfacial mechanical energy dissipated during the wear experiments. The reciprocating nano-wear resistance experiments were performed using a spherical 5 μm radius diamond tip, with a load range from 1 to 30 mN, providing contact pressures from 6.6 to 20.70 GPa, respectively. This range is representative of a single asperity contact between hard inert ceramic ZrO2 radiopacifier agglomerates present in the cement and the CoCrMo stem surface [25]. The radiopacifier agglomerates typically have a mean diameter between 5 and10 microns in size [26,27] although bigger size agglomerates have been found as well due to mixing inhomogeneity’s [28]. The diamond tips were observed under optical microscope before and after wear experiments without observing any geometry changes which might affect the results. Depth measurements were monitored throughout the duration of the experiments and pre and post thermal drift values were collected during 60 s to correct the obtained depth data. Displacement of the tip was measured by means of a two parallel capacitor plates situated behind the diamond tip. One plate is attached to the tip holder and moves with it while the second plate is fixed. When the tip is pressed into the test specimen, the spacing between the parallel plates changes, hence the capacitance changes. By measuring it using a capacitance bridge, the depth changes are measured. The system stiffness was calculated by indenting on a fused silica reference sample over a broad range of loads using the reciprocating nano-wear holder and using that values to correct the depth values measured by the capacitor plates. A fixed displacement amplitude of 10 μm was used on the basis of clinical studies which demonstrated that micro-motions between the femoral stem and the cement are typically below 40 μm [29,30]. To shorten the test duration, tests were carried out using a oscillation frequency of 7.5 Hz rather than a physiologically 1 Hz frequency. Each test ran for 3000 cycles. The instrument uses a multi-layer piezo-stack to generate the oscillating sample motion. The piezo movement is magnified by means of a lever arrangement to achieve larger amplitudes [31]. Track length values depend on the frequency and voltage applied to the piezo-stack inside the nano-wear stage. The piezo voltage is produced by a signal generatorþamplifierþtransformer combination. During the calibration, it is mounted perpendicularly to the experimental oscillation axis so that the magnitude of the oscillation (track length) can be measured by the depth sensor and the voltage applied can be correlated to the length of the wear track produced. As is shown in Fig. 1, an initial low contact load of 0.1 mN was applied for 15 s to record the initial contact depth. The load is then linearly increased over 10 s until the maximum test load was reached (1). This load was kept constant during the 3000 cycles (400 s), (2). Finally the samples were unloaded linearly in 10 s to 0.1 mN and held for the next 15 s to record the final non-loaded

Please cite this article as: Martinez-Nogues V, et al. Nano-scale wear characterization of CoCrMo biomedical alloys. Tribology International (2015), http://dx.doi.org/10.1016/j.triboint.2015.03.037i

V. Martinez-Nogues et al. / Tribology International ∎ (∎∎∎∎) ∎∎∎–∎∎∎

depth, to provide information on the permanent plastic deformation produced in the alloy (3). Sliding wear scar depth measurements were obtained from the mechanical system. To validate the results, AFM and white-lightinterferometry measurements were performed. A high performance white-light-interferometer with a lateral resolution of 0.8 μm and vertical resolution of 0.1 nm was used to perform depth measurements after the nano-wear experiments. AFM measurements using tapping mode were performed on the samples to validate the values obtained from the interferometer. An example of the 3D measurements taken by the white-lightinterferometer system and the depth profiles are shown in Fig. 2. Three cross section profiles in horizontal and transverse directions were used to evaluate the nano-wear scars depths by the two techniques.

1

2

20

400

10

3 -10 -20 0

Force (mN)

Depth (nm)

0 200

-30

Tangential friction force Normal applied load

-200

Depth

-40 -50

0

100

200

300

400

2.3. Microscopy analysis Post-test surface imaging of the wear scars was performed using an optical microscope (Alicona Infinite Focus) and a SEM microscope to investigate the possible wear mechanisms and the sliding wear scar morphologies. All experiments were performed at room temperature and dry conditions. Five samples per material were tested for each experiment. Two-tailed paired T-test statistical analysis was performed on the samples to establish differences and/or similarities in the behaviour between the four CoCrMo alloys. A p-value below 0.05 was considered significant for all statistical analyses.

3. Results 3.1. Hardness and reduced modulus

30

600

3

500

Time (s) Fig. 1. Load profile applied during the nano-wear experiments (red line), tangential friction force (blue line) and depth values recorded (black line). Note that the numbers, 1, 2 and 3 correspond to the initial depth, final depth and unloaded final depth, respectively. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Micro-hardness (HV and its transformation to HIT-μ), nanoindentation hardness (HIT-nano) and reduced modulus (Er) values are summarised in Table 1. Micro-hardness, HIT-μ, present the maximum values for the forged sample (5.8 70.5 GPa), followed by the as cast (4.747 0.5 GPa), the as cast single thermal treated (4.40 70.3 GPa) and the as cast double thermal treated (4.20 70.2 GPa). Thermal treatments provide a reduction of hardness of 7% and 11% for the single and the double thermal treatment respectively comparing with the non-thermal treated sample. The values are statically different between the alloys with a p-value below 0.05. Considering the hardness values obtained from the nano-indentation maps, the highest hardness and reduced modulus is presented by the as cast alloy with 9.5 71.1 GPa followed by the forged, 7.1 70.3 GPa, and the single and double thermal treated ones with non-significant differences between them. The same tendency is followed for the reduced modulus with values between 278 722 and 235718 GPa. Note that the HIT-nano values for the alloys with dispersed carbides, as

Fig. 2. Example of 3D surface measurement obtained with the white-light-interferometer system and its correspondent depth profile for one of the nano-wear wear scars used to validate the depth measurements done by the NanoTest Vantage system.

Please cite this article as: Martinez-Nogues V, et al. Nano-scale wear characterization of CoCrMo biomedical alloys. Tribology International (2015), http://dx.doi.org/10.1016/j.triboint.2015.03.037i

V. Martinez-Nogues et al. / Tribology International ∎ (∎∎∎∎) ∎∎∎–∎∎∎

4

0.4

Table 1 Summary of the micro-Vickers hardness, the nano-indentation hardness and the reduced modulus values for the alloy’s matrix and their correspondent carbides for the as cast and the single thermal treated as cast alloys. Nano-indentation

HV

HIT-μ (GPa)

HIT-nano (GPa)

Er (GPa)

548 749 4487 50 – 4157 29 – 398 7 22

5.8 70.5 4.747 0.5 – 4.40 70.3 – 4.20 70.2

7.17 0.3 9.5 7 1.1 147 2 6.7 7 0.5 9.4 7 1.5 6.6 7 0.3

2217 7 2787 22 303 7 20 2217 12 2357 18 2197 8

Coefficient of friction (µ)

Forged As cast As cast carbide As cast-TT As cast-TT carbide As cast-2TT

Vickers micro-indentation

0.3

0.0 -0.1 -0.2

100.2

100.3

100.4

100.5

50 0

0 -50 -100

-2

-150 -200

Tangential Friction force COF

-250 100.2

100.3

100.4

-4

100.5

Time (s) Fig. 3. Typical data collected for 0.5 s during nano-wear testing of CoCrMo alloy. The tangential friction force (red line), the depth mesurement (black line) and the calculated coefficient of friction (blue line) are shown for an applied normal load of 20 mN and a frequency of 7.5 Hz. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Fig. 4. Representative coefficient of friction for the forged CoCrMo-diamond contact at normal loads from 5, 10 and 20 mN.

0.350 0.325 0.300 0.275

Coefficient of friction (µ)

2

Tangential Friction Force (mN) COF (µ)

4

100

100.1

100.1

Time (s)

Depth

150

Depth (nm)

0.1

-0.4 100.0

200

-300 100.0

0.2

-0.3

300 250

5 mN 10 mN 20 mN

0.250 0.225 0.200 0.175 0.150 0.125 0.100 0.075 0.050 0.025

Forged AsCast AsCast-TT AsCast-2TT

0.000 5

cast and as cast single thermal treated, are reflecting a value which is a combination of the matrix and the carbides mechanical properties. Solid carbides present in the as cast alloy also show approximately 50% higher hardness and reduced modulus values compared to the “fragmented-like” carbides from the single thermal treated alloy.

An example of the mechanical data captured during 0.5 s of the sliding motion of the diamond tip against the CoCrMo surface is shown in Fig. 3. The tangential friction force and the calculated coefficient of friction change their values when the motion is reversed. The same occurs with the depth measurement. Fig. 4 shows a full set of COF plots for 0.5 s of motion obtained during testing. No change in the slope of the tangential forces as the motion direction reverses suggests that no-stick-slip is occurring, therefore a full sliding regime was assumed for the dispersed energy calculations for the three normal loads applied, 5, 10 and 20 mN. The plateau shape with a relatively constant value found for the coefficient of friction confirms a sliding regime with no decrease in the motion amplitude. Increasing the normal contact load resulted in a lower COF. A higher dispersion is also observed for the COF values at the lowest normal load of 5 mN. Running in and steady state COF, μ, for 5, 10 and 20 mN loads for the four alloys are shown in Fig. 5. Independently from the alloy, both regimes present a systematic variation in COF values with load where higher loads resulted in lower COF. For the running in period, it varies between 0.175 and 0.2, for a 5 mN applied load, and 0.150 for 20 mN. The range during the steady state period decreased from 0.2–0.275 to 0.150 for the 5 and

15

20

15

20

Load (mN)

0.350 0.325 0.300 0.275

Coefficient of friction (µ)

3.2. Coefficient of friction and dispersed energy values

10

0.250 0.225 0.200 0.175 0.150 0.125 0.100 0.075 0.050 0.025

Forged AsCast AsCast-TT AsCast-2TT

0.000 5

10

Load (mN) Fig. 5. (a) Running-in and (b) steady state coefficient of friction at 5, 10 and 20 mN for the four CoCrMo alloys.

20 mN loads, respectively. Comparing between the alloys shows there are no statistical differences during the running in period, however, slightly higher values were observed for the forged alloy at the steady state. Fig. 6 shows a representative friction force–displacement plot at three different normal loads, 5, 10 and 20 mN, for the forged sample. Due to the low contact pressures used, 0.97 to 1.5 GPa, the diamond tip achieves the maximum displacement of 10 μm, which

Please cite this article as: Martinez-Nogues V, et al. Nano-scale wear characterization of CoCrMo biomedical alloys. Tribology International (2015), http://dx.doi.org/10.1016/j.triboint.2015.03.037i

V. Martinez-Nogues et al. / Tribology International ∎ (∎∎∎∎) ∎∎∎–∎∎∎ 5

3

Forged

Nanovantage system AFM Interferometer

1600 1400

2

1200

Depth (nm)

Friction force (mN)

1800

5 mN 10 mN 20 mN

4

1 0 -1

1000 800 600

-2

400

-3

200

-4

0

-5 -15

-10

-5

0

5

10

0

15

2

4

6

8

10 12 14 16 18 20 22 24 26 28 30 32

Normal load (mN)

Displacement (µm) Fig. 6. Friction force–displacement plot for the forged alloy when normal loads of 5, 10 and 20 mN are applied.

Fig. 8. Comparison of the final depth values by the three different techniques used: NanoTest Vantage system, AFM and white light white-light-interferometer for the as cast sample.

2000

40

30

Depth measured by AFM (nm)

Etogether-forged Etogether-as cast Etogether-as cast-TT Etogether-ascast2TT

35

Energy / cycle (µJ)

5

25 20 15 10

Pearson's r = 0.94 R2 = 0.89 1500

1000

500

0

5 0

0 0

2

4

6

8

10

12

14

16

18

20

22

Load (mN) Fig. 7. Dispersed energy values per cycle at 5, 10 and 20 mN loads for the forged, as cast, as cast-TT and as cast-2TT.

implies the sliding regime was maintained during the test and the width of the rectangle remains constant during the cycles independently of the applied load for any of the CoCrMo samples tested. Although the tangential frictional forces change when the load is increased, the load levels are not enough to decrease the effective sliding distance due to increased sticking. The dispersed energy per cycle (μJ) was calculated from the area limited by the friction force (mN) and the displacement values (μm). The energy values are summarised in Fig. 7. Energy per cycle values increased linearly when the applied normal load was increased. The total dispersed energy variation was between 57 and 60%. Although the percentage variation were similar for the four alloys the average energy values shown by the forged sample were statistical different, p-value o0.05, from the as cast sample values at 10 and 20 mN loads, 21 and 31 μJ/cycle compared to the 17 and 27 μJ/ cycle, respectively. Energy values were similar for the 5 mN load (between 9.62 and 12.26 μJ/cycle). 3.3. Nano-wear scars depth measurement and white-lightinterferometer validation Depth values obtained from the NanoTest Vantage system were validated against AFM and white-light-interferometry measurements. Fig. 8 compares the depth values for the load range from

200

400

600

800

1000

Depth measured by Nanovantage system (nm) Fig. 9. Correlation between the depths measured by the NanoTest Vantage system and the AFM.

1 to 30 mN for the as cast CoCrMo sample. From 1 to 27 mN almost all the values obtained by the three techniques are coincident, however, for the highest loads the AFM triples the values measured by the NanoTest Vantage and the interferometer systems. Also the interferometer values show a bigger dispersion at some points compared to the AFM or the NanoTest Vantage system. Linear correlations are plotted from Figs. 9–11. AFM microscopy and NanoTest Vantage system measurements exhibit a positive Pearson’s correlation coefficient equal to 0.94, while r is equal to 0.89 while the white-light-interferometer and NanoTest Vantage systems correlation coefficient is 0.88 and 0.90 for the white-lightinterferometer-AFM correlation. Note that values obtained using the AFM microscope at higher loads have been excluded from the analysis because the AFM tip is not able to obtain representative results when pile up or sliding wear scars depth exceeds a limit over 950 nm. A comparison between the initial-loaded depth, the finalloaded depth and the final-unloaded depth are shown in Figs. 12–14, respectively. The initial depth values for the as cast samples were significantly higher than the other alloys for loads between 18 and 30 mN. No significant difference was seen in the depth, with all samples following an increasing linear trend when the load is increased. The maximum final loaded-depths, (elastic and plastic deformation combined) for the as cast and the as cast double thermal treated samples were significantly higher than

Please cite this article as: Martinez-Nogues V, et al. Nano-scale wear characterization of CoCrMo biomedical alloys. Tribology International (2015), http://dx.doi.org/10.1016/j.triboint.2015.03.037i

V. Martinez-Nogues et al. / Tribology International ∎ (∎∎∎∎) ∎∎∎–∎∎∎

6

800

Pearson's r = 0.88 R2 = 0.78 1500

600

1000

500

500 400 300 200 100

0

0 0

200

400

600

800

1000

0

Depth measured by Nanovantage system (nm)

5

10

15

20

25

30

Load (mN)

Fig. 10. Correlation between the depths measured by the NanoTest Vantage system and the white-light-interferometer.

Fig. 13. Comparison of the final depth at 400 s of the nano-wear experiments for the forged, as cast, as cast-TT and as cast-2TT alloys.

2000

800

Pearson's r = 0.90 R2 = 0.80

Forged As cast As cast-TT As cast-2TT

700

1500

Unloaded final depth (nm)

Depth measured by interferometer (nm)

Forged As cast As cast-TT As cast-2TT

700

Final depth (nm)

Depth measured by interferometer (nm)

2000

1000

500

0

0

200

400

600

800

1000

1200

1400

1600

1800

Depth measured by AFM (nm)

600 500 400 300 200 100 0 0

Fig. 11. Correlation between the depths measured by the AFM and the white-lightinterferometer. Note the final measurements, over 900 nm, measured by AFM are excluded from the correlation.

5

10

15

20

25

30

Load (mN) Fig. 14. Comparison of the unloaded depth at the end of the nano-wear experiments for the forged, as cast, as cast-TT and as cast-2TT alloys.

800

500

Figs. 13 and 14 show two separate regions: from 1 to 15 mN the deformation increased linearly and the alloys deformation level is similar. From 15 to 30 mN the increase in depth is exponential and the values became statistical different between the as cast and the rest of the alloys.

400

3.4. Wear scars morphology

Forged As cast As cast-TT As cast-2TT

700

Initial depth (nm)

600

300 200 100 0 0

5

10

15

20

25

30

Load (mN) Fig. 12. Comparison of the initial depth at 15 s of the nano-wear experiments for the forged, as cast, as cast-TT and as cast-2TT alloys.

those for the single thermal treated and the forged ones. Analysing the unloaded depth values, up to 15 mN the values are coincident for the four alloys. However, over 15 mN the as cast sample starts progressing deeper when the load is increased; reaching a maximum difference of 8% higher than the forged and the thermal treated as cast samples.

Table 2 shows a summary of the most significant wear scar morphologies for the forged, as cast, as cast single thermal treated and as cast double thermal treated for different group loads. At the lowest loads, 1–2 mN, sliding wear damage is only visible for the as cast sample. For the rest, no signs of deformation are distinguishable by optical microscopy. Severe plastic deformation is observed for the highest loads applied. Particle debris and pile up on the scar edges starts to be visible when the applied stresses passes 20 mN. All the samples show a predominantly ductile response to sliding wear with ploughing and pile-up and wear debris at the sides of the wear-track. However, the quantity and shape of the debris around the nano-wear scars in the double thermal treated as cast sample is larger than in the rest of materials. SEM images in Fig. 15 show the evolution of the wear scar while the load increases. At low loads, (Fig. 15a and b), a reduced material displacement is seen together with some powder-like debris, with typical particle sizes of around the micron at the top and bottom part of the scar. Once the load is

Please cite this article as: Martinez-Nogues V, et al. Nano-scale wear characterization of CoCrMo biomedical alloys. Tribology International (2015), http://dx.doi.org/10.1016/j.triboint.2015.03.037i

V. Martinez-Nogues et al. / Tribology International ∎ (∎∎∎∎) ∎∎∎–∎∎∎

7

Table 2 Optical images of the nano-wear scars under 1, 10, and 30 mN load for the forged, as cast, as cast-TT and the as cast-2TT CoCrMo alloys.

increased, the wear particles generated around the wear track increased in size forming aggregates and also some plate-like delamination debris, see Fig. 15c. At the highest loads the severe pile up is predominant over the particle debris formation, see Fig. 15d.

4. Discussion

some of the matrix measurements are influenced by the mechanical properties of the neighbouring or subsurface carbides, overestimating the as cast matrix nanohardness value which is apparently higher than the forged one. Carbide morphology, bulky for the as cast samples and fragmented-like for the single thermal treated one, have an influence. The more irregular shape in the thermal treated sample can lead to lower values especially if the indenter lands on the edge of the carbide or if the carbide is not well-supported by the underlying matrix.

4.1. Mechanical properties 4.2. Frictional behaviour and dispersed energy per cycle The observed differences in indentation response are in agreement with previous reports where forged alloys are harder than as cast alloys [32,33]. One of the reasons for the difference is related with their microstructure. The forged sample has a smaller grain size (around 20 μm) and a more homogeneous microstructure, compared to the as cast alloys which grain size might be close to 200 μm. Several of these studies have reported an indentation size effect (ISE) in hardness [34] and also the difference in carbide size in comparison to the indentation size. The indentation diagonal distance for the micro indentation experiments is around 20 μm which is bigger than the typical carbide size in the as cast and single thermal treated alloys. Therefore, indentations are not able to differentiate between the hardness contributions from carbide and metal matrix phases. In contrast, the nano-hardness test provides a comparison between the carbides and matrix. However,

In dry conditions, the COF showed a significant load and material dependency, see Fig. 5. Higher COF values during the steady state period for the lower loads, in agreement with [35], might be explained because of the low contact pressures applied and the stresses produced in the contact area. Initially, a low level of plastic deformation is induced, so friction is mainly due to the rubbing between the contacting surfaces. As the experiment continues the cycling loading produced the fracture of the asperities and the volume of debris increases resulting in the COF values rising. When the load passes a critical limit, the tip indents into the surface increasing the contact area. During the first cycles the asperities are immediately removed and the COF values remains constant during the test duration and running-in and steady state values are coincident. Previous studies have focused

Please cite this article as: Martinez-Nogues V, et al. Nano-scale wear characterization of CoCrMo biomedical alloys. Tribology International (2015), http://dx.doi.org/10.1016/j.triboint.2015.03.037i

8

V. Martinez-Nogues et al. / Tribology International ∎ (∎∎∎∎) ∎∎∎–∎∎∎

2 mN

10 mN

15 mN

30 mN

Fig. 15. SEM images of nano-wear scars morphologies for the as cast-2TT sample at four representative load levels.

on the evolution of friction and wear of metallic materials such as Ti, Cu or Ni during micro-scale sliding [36,37] affirming that yield stress plays the dominant role on the evolution of friction, independently of the grain size and decreasing with increasing hardness. In relation to hardness, the opposite was found in this study, with higher friction coefficient values for the forged alloy. This is however in agreement with the work by Huang et al. [38] were the tribological behaviour of wrought and as cast CoCrMo alloys were compared. However, the relative difference between samples and the higher dispersion of friction values monitored at low loads might be a limitation of the system which is not capable of recording the tangential friction forces applied below a specific load limit. The dispersed energy per cycle (Fig. 7) showed that the hardest alloy presented the higher dispersed energy, which means the work needed to deform the material is higher than for the as cast alloys. At this contact pressure level the thermal treatments applied on the as cast sample did not play a significant role in the dispersed energy. Full sliding was assumed for all the samples, which can be considered as a system limitation because the real displacement was not monitored during the experiments. However, no slope changes were observed in the tangential frictional forces recorded for any of the applied loads, which according to [39], indicates a full slip regime with no partial or full sticking occurring, validating the assumption made. 4.3. Deformation behaviour and depth measurement validation The deformation behaviour of the four alloys when the applied stress exceeded the stress yield was predominantly plastic, with the wear response was mainly dominated by the matrix properties. However, the wear scars which landed on a carbide showed a decrease in contact depth when compared to the matrix. The

smaller size carbides present in the single thermal treated alloy do not exhibit the same effect as the carbides in the as cast alloy. The as cast sample matrix deforms more initially when the sample is loaded (Fig. 12). Levels of final on load deformation, Fig. 13, are similar for the as cast and the as cast double thermal treated sample and 57% higher than the forged and the as cast single thermal treated. However, the unloaded depth at the end of the experiment showed the as cast sample to have a significantly higher level of plastic deformation. The increase of carbon content is known to increase the stacking fault energy [40], which produce narrower stacking faults and increasing the mobility of dislocations, making easier the deformation process. Ploughing and pile-up has been observed in previous studies at the nano and micro-scales [21,31] under sliding conditions. In agreement with the study it increased when increasing the applied load. The evidence of minor plastic deformation in the as cast alloy for the lower load levels, Table 2b1, agrees with the higher deformation level discussed in the deformation behaviour section comparing to the other alloys. The significant increase of wear debris in the as cast double thermal treated alloy, but also observed in the single thermal treated sample, suggests that besides the level of plastic deformation at intermediate stress, the thermal treatments plays a role in the production of wear debris. Is extensively known that the size and quantity of metallic particles disseminated in the human body can cause osteolysis, loosening of the implant and its revision [41–43]. As cast thermal treated samples might accelerate the nano-wear process by the production of wear debris which remains between the two contacting surfaces acting as third body particles. AFM is used extensively as a measuring system and although it provided high levels of detail about the wear scars in this study, it was time consuming due to low area covered and was unable to measure highly deformed scars, as the height range exceeded the range of

Please cite this article as: Martinez-Nogues V, et al. Nano-scale wear characterization of CoCrMo biomedical alloys. Tribology International (2015), http://dx.doi.org/10.1016/j.triboint.2015.03.037i

V. Martinez-Nogues et al. / Tribology International ∎ (∎∎∎∎) ∎∎∎–∎∎∎

motion of the tip. The combination of in-situ depth measures using the NanoTest Vantage system and the white-light-interferometer validation confirms the interferometry measurements are faster and reliable even for the high deformation levels. Despite the increasing use of non-cemented fixation techniques which covers a 42% of the total hip arthroplasties undertaken in the UK during 2013, 33% of hip replacement procedures still use cemented fixation, with the cement containing radio pacifier particles [44]. Wear at the nano-scale observed in retrieved implants might have its origin in the hard radio pacifier particles and despite the experimental limitations such as the lack of a closer physiological environment, the work introduces a controlled experimental procedure where the effect of several thermal treatments and manufacturing processing history are correlated with their wear resistance at the nanoscale, deformation process and wear debris production. In general only hardness and reduced modulus have been recognized as significant variables to explain the sliding wear processes between the femoral stem and the cement interface. However, it has been confirmed by this study that the manufacturing process and the thermal treatment which the CoCrMo are subjected to play a crucial role. Future work will introduce wet environment to study the synergy processes and the variables affecting corrosion and wear at the nano-scale.

5. Conclusions In the present study, four CoCrMo alloys were subjected to reciprocating wear experiments at the nano-scale level to compare their wear resistance and their frictional behaviour. A single ZrO2 radiopacifier asperity was considered acting between the metallic femoral stem and the cement interface to generate the nano-wear scars. The influence of the CoCrMo manufacturing and thermal processing together with the load dependence was investigated. In addition a new white-light-interferometry system was used and validated by AFM and NanoTest Vantage System depth measurements to obtain the permanent plastic deformation values produced in each of the alloys. The following conclusions could be drawn from this work:

 Reciprocating nano-wear COF values are load dependent. It





decreases when load is increased. A higher load reduces the number of asperities between the contact surfaces, therefore, the COF decreases between them when the maximum load is applied. Running in and steady state friction coefficients are coincident for the as cast CoCrMo alloys with no statistical differences found between them with an average value from 0.25 to 0.150. The forged alloy presented statistical differences between the initial running in period and the steady state one except when a 20 mN load is applied, with higher values after the initial period. Permanent plastic deformation on the samples depends on the specific microstructure, manufacturing and thermal history which the alloy was subjected to. The as cast sample underwent higher deformation levels with visible sliding wear scars under 1 mN load. Besides the plastic deformation, thermal treated samples generated a higher number of particles around the fretting scars compared to the forged and the non-thermal treated as cast samples. The white-light-interferometer system was validated by linear correlation against the depths obtained from the NanoTest Vantage system and the AFM measurements. The white-lightinterferometer has been proved to be able to measure the wear scars at the nano scale.

Understanding the nano-wear mechanisms for the different CoCrMo alloys can help in the correct choice of future implantable

9

biomaterials depending on their microstructural features, manufacturing processes and thermal history.

Acknowledgements The authors would like to acknowledge the University of Southampton for the founding received, to Christian Maul for allowing the use of the white-light-interferometer at RedLux Ltd. and to Ben Beake and Andrew Bird from Micro Materials Ltd. for their support with the NanoTest Vantage System. References [1] Fricka KB, Ho H, Peace WJ, Engh Jr CA. Metal-on-metal local tissue reaction is associated with corrosion of the head taper junction. J Arthroplasty 2012;27 (Suppl. 8):26–31 e1. [2] Howell JR, Blunt LA, Doyle C, Hooper RM, Lee AJC, Ling RSM. In vivo surface wear mechanisms of femoral components of cemented total hip arthroplasties: the influence of wear mechanism on clinical outcome. J Arthroplasty 2004;19(1):88–101. [3] Cook RB, Bolland BJRF, Wharton JA, Tilley S, Latham JM, Wood RJK. Pseudotumour formation due to tribocorrosion at the taper interface of large diameter metal on polymer modular total hip replacements. J Arthroplasty 2013;28(8):1430–6. [4] Cooper HJ, Della Valle CJ, Jacobs JJ. Biologic implications of taper corrosion in total hip arthroplasty. Sem Arthroplasty 2012;23(4):273–8. [5] Goldberg JR, Gilbert JL, Jacobs JJ, Bauer TW, Paprosky W, Leurgans S. A multicenter retrieval study of the taper interfaces of modular hip prostheses. Clin Orthop Relat Res 2002;401:149–61. [6] Blunt LA, Zhang H, Barrans SM, Jiang X, Brown LT. What results in fretting wear on polished femoral stems. Tribol Int 2009;42(11–12):1605–14. [7] Howell JR, Blunt L, Ling RSM. An analysis of fretting damage seen on explanted femoral stems. J Bone Jt Surg 1999;81(B):163. [8] Bryant M, Ward M, Farrar R, Freeman R, Brummitt K, Nolan J, et al. Characterisation of the surface topography, tomography and chemistry of fretting corrosion product found on retrieved polished femoral stems. J Mech Behav Biomed Mater 2014;32(0):321–34. [9] Zhang H, Brown LT, Blunt LA, Jiang X, Barrans SM. Understanding initiation and propagation of fretting wear on the femoral stem in total hip replacement. Wear 2009;266(5–6):566–9. [10] Zhang HY, Blunt L, Jiang XQ, Brown L, Barrans S, Zhao Y. Femoral stem wear in cemented total hip replacement. Proc Inst Mech Eng H 2008;222(5):583–92. [11] Bryant M, Farrar R, Brummitt K, Freeman R, Neville A. Fretting corrosion of fully cemented polished collarless tapered stems: the influence of PMMA bone cement. Wear 2013;301(1–2):290–9. [12] Zhang HY, Blunt LA, Jiang XQ, Fleming LT, Barrans SM. The influence of bone cement type on production of fretting wear on the femoral stem surface: a preliminary study. Clin Biomech 2012;27(7):666–72. [13] Cook RB, Shearwood-Porter NRM, Latham JM, Milton JA, Palmer MR, Wood RJK. The mechanism and magnitude of tribocorrosion damage at the cement– stem interface. In: Elsevier, editor. Second international conference on biotribology, Toronto, Canada; 2014. [14] Bryant M, Farrar R, Freeman R, Brummitt K, Nolan J, Neville A. Galvanically enhanced fretting-crevice corrosion of cemented femoral stems. J Mech Behav Biomed Mater 2014;40(0):275–86. [15] Gilbert JL, Buckley CA, Jacobs JJ. In vivo corrosion of modular hip prosthesis components in mixed and similar metal combinations. The effect of crevice, stress, motion, and alloy coupling. J Biomed Mater Res 1993;27(12):1533–44. [16] Büscher R, Fischer A. Metallurgical aspects of sliding wear of fcc materials for medical applications. Materialwiss Werkstofftech 2003;34(10-11):966–75. [17] Clemow AJT, Daniell BL. The influence of microstructure on the adhesive wear resistance of a CoCrMo alloy. Wear 1980;61(2):219–31. [18] Saldívar-García AJ, López HF. Microstructural effects on the wear resistance of wrought and as-cast Co–Cr–Mo–C implant alloys. J Biomed Mater Res Part A 2005;74A(2):269–74. [19] Varano R, Bobyn JD, Medley JB, Yue S. The effect of microstructure on the wear of cobalt-based alloys used in metal-on-metal hip implants. Proc Inst Mech Eng H 2006;220(2):145–59. [20] Hoeppner DW, Chandrasekaran V. Fretting in orthopaedic implants: a review. Wear 1994;173(1–2):189–97. [21] Beake BD, Liskiewicz TW. Comparison of nano-fretting and nano-scratch tests on biomedical materials. Tribol Int 2013;63(0):123–31. [22] Beake BD, Liskiewicz TW, Pickford NJ, Smith JF. Accelerated nano-fretting testing of Si(1 0 0). Tribol Int 2012;46(1):114–8. [23] Ahmed R, Ashraf A, Elameen M, Faisal NH, El-Sherik AM, Elakwah YO, et al. Single asperity nanoscratch behaviour of HIPed and cast Stellite 6 alloys. Wear 2014;312(1–2):70–82. [24] ISO, T.E.S.E. Metallic materials—instrumented indentation test for hardness and materials parameters. ISO 14577-1:2002. In: Part 1: Test method, Brussels; 2002.

Please cite this article as: Martinez-Nogues V, et al. Nano-scale wear characterization of CoCrMo biomedical alloys. Tribology International (2015), http://dx.doi.org/10.1016/j.triboint.2015.03.037i

10

V. Martinez-Nogues et al. / Tribology International ∎ (∎∎∎∎) ∎∎∎–∎∎∎

[25] Mitchell A, Shrotriya P. Onset of nanoscale wear of metallic implant materials: influence of surface residual stresses and contact loads. Wear 2007;263(7–12):1117–23. [26] Lewis G, van Hooy-Corstjens CS, Bhattaram A, Koole LH. Influence of the radiopacifier in an acrylic bone cement on its mechanical, thermal, and physical properties: barium sulfate-containing cement versus iodinecontaining cement. J Biomed Mater Res B Appl Biomater 2005;73(1):77–87. [27] Ginebra MP, Albuixech L, Fernández-Barragán E, Aparicio C, Gil FJ, San Román J, et al. Mechanical performance of acrylic bone cements containing different radiopacifying agents. Biomaterials 2002;23(8):1873–82. [28] Wang JS, Diaz J, Sabokbar A, Athanasou N, Kjellson F, Tanner KE, et al. In vitro and in vivo biological responses to a novel radiopacifying agent for bone cement. J R Soc Interface 2005;2(2):71–8. [29] Maloney WJ, Jasty M, Burke DW, O’Connor DO, Zalenski EB, Bragdon C, et al. Biomechanical and histologic investigation of cemented total hip arthroplasties. A study of autopsy-retrieved femurs after in vivo cycling. Clin Orthop Relat Res 1989;249:129–40. [30] Ebramzadeh E, Billi F, Sangiorgio SN, Mattes S, Schmoelz W, Dorr L. Simulation of fretting wear at orthopaedic implant interfaces. J Biomech Eng 2005;127(3):357–63. [31] Liskiewicz TW, Beake BD, Smith JF. In situ accelerated micro-wear—a new technique to fill the measurement gap. Surf Coat Technol 2010;205(5):1455–9. [32] Liao Y, Pourzal R, Stemmer P, Wimmer MA, Jacobs JJ, Fisher J, et al. New insights into hard phase of CoCrMo metal-on-metal hip replacements. J Mech Behav Biomed Mater 2012;12:39–49. [33] Stemmer P, Pourzal R, Liao Y, Marks LD, Morlock M, Jacobs, JJ. et al. Microstructure of retrievals made from standard cast HC-CoCrMo alloys. In: Metalon-metal total hip replacement devices, A.I.-S. 1560, editor; 2013.

[34] Voyiadjis G, Peters R. Size effects in nanoindentation: an experimental and analytical study. Acta Mech 2010;211(1-2):131–53. [35] Maruyama N, Kawasaki H, Yamamoto A, Hiromoto S, Imai H, Hanawa T. Friction-wear properties of nickel-free Co–Cr–Mo alloy in a simulated body fluid. Mater Trans 2005;46(7):1588–92. [36] Wang CT, Gao N, Gee MG, Wood RJK, Langdon TG. Effect of grain size on the micro-tribological behavior of pure titanium processed by high-pressure torsion. Wear 2012;280–281(0):28–35. [37] Hanlon T, Chokshi AH, Manoharan M, Suresh S. Effects of grain refinement and strength on friction and damage evolution under repeated sliding contact in nanostructured metals. Int J Fatigue 2005;27(10–12):1159–63. [38] Huang P, Salinas-Rodriguez A, López HF. Tribological behaviour of cast and wrought Co–Cr–Mo implant alloys. J Mater Sci Technol 1999;15(11):1324–30. [39] Swaminathan V, Gilbert JL. Fretting corrosion of CoCrMo and Ti6Al4V interfaces. Biomaterials 2012;33(22):5487–503. [40] Lee S-H, Takahashi E, Nomura N, Chiba A. Effect of carbon addition on microstructure and mechanical properties of a wrought CoCrMo implant alloy. Mater Trans, JIM 2006;47(2):287–90. [41] Ding J, Leen SB, Williams DH, Shipway PH. A multi-scale model for fretting wear with oxidation-debris effects. Proc Inst Mech Eng Part J J Eng Tribol 2009;223(7):1019–31. [42] Lombardi AV, Mallory TH, Vaughn BK, Drouillard P. Aseptic loosening in total hip arthroplasty secondary to osteolysis induced by wear debris from titanium-alloy modular femoral heads. J Bone Joint Surg 1989;71(9):1337–42. [43] Tipper JL, Firkings PJ, Ingham E, Fisher J, Stone MH, Farrar R. Quantitative analysis of the wear and wear debris from low and high carbon content CoCr alloys used in metal on metal total hip replacements. J Mater Sci—Mater Med 1999;10:353–62. [44] Registry, U.N.J. 2013 UK national joint registry annual report; 2013.

Please cite this article as: Martinez-Nogues V, et al. Nano-scale wear characterization of CoCrMo biomedical alloys. Tribology International (2015), http://dx.doi.org/10.1016/j.triboint.2015.03.037i