Water contact angle
Figure 2 shows the results of the WCA measurements as a function of treatment time for different jet speeds and a fixed discharge power of 1.6 W. The WCA of untreated UHMWPE is 87° and decreases exponentially with increasing plasma treatment time for each of the 3 plasma jet velocities, until a plateau value of approximately 47° is reached. This plateau value is similar to a previously obtained value (49°) after Ar plasma treatment in a DBD-reactor at 5.0 kPa4. However, the energy density needed to obtain this plateau value in the DBD-reactor was 1.34 J/cm2 while the energy density needed with the plasma jet is calculated to be 564.14 J/cm². This difference in energy density can probably be explained by the manner of plasma exposure. In a DBD-reactor, the sample is in direct contact with the active plasma zone, while in the case of a plasma jet the sample is exposed indirectly to plasma as only the plasma afterglow reaches the surface. The decrease in WCA with treatment time can be attributed to the incorporation of oxygen-containing functional groups into the sample’s surface due to the treatment. This will be discussed in detail in the next section. Additionally, the results also show that the jet speed has no significant effect on the WCA, as using the same treatment time (by changing the number of repetitions) results in similar WCA values. For the same treatment time, using a higher jet speed with a high number of repetitions will therefore result in the same WCA value as using a lower jet speed with a low number of repetitions.
Based on these results, some specific plasma operational conditions were chosen to further investigate the plasma-induced effects on the surface chemistry, cell proliferation, adhesion to bone cement and CaP deposition. The chosen conditions are represented in Table 1 with their respective WCA values. Three different saturation conditions (S1 to S3) are first chosen to investigate whether the jet velocity has any effect at all. Condition B is a sample exposed to the plasma for a very short treatment time and is chosen to see the difference between saturated samples and samples that show a small treatment effect. Condition F is a combination of process parameters that is not represented in Fig. 2. The goal here is to see if an even more pronounced plasma treatment effect (high energy density resulting in a very low WCA value) has an influence on the results of the different tests. These parameters are based on previously obtained results27.
The chemical composition of samples treated with the different process conditions are analyzed using XPS and compared with untreated samples. The results are presented in Table 2. The untreated UHMWPE samples did not contain any oxygen, however after plasma treatment the oxygen concentration is considerably increased. This explains the decrease in contact angle, as the incorporation of oxygen leads to a more hydrophilic surface. Conditions S1, S2 and S3 lead to a similar O/C-ratio of 0.43, showing that the jet velocity has no influence on the overall incorporated oxygen concentration. Using a lower treatment time (condition B) leads to a lower O/C-ratio. Although the WCA for condition F was significantly lower than for the other conditions, the measured O/C-ratio is similar to conditions S1, S2 and S3. This discrepancy can most likely be explained by the depth of analysis in the two techniques. WCA measurements analyze approximately the top 1 nm of the surface, while XPS analysis depends on the used take-off angle relative to the sample surface, which is 45° in this case and which results in an analyzing depth of approximately 10 nm. This seems to indicate that the oxygen concentration is not homogeneous over these 10 nm and that there probably is a gradual decrease of the oxygen content from the surface into the sample. Comparing these results to the O/C-ratio (0.27) obtained at medium pressure4, it is clear that condition B leads to similar results while the saturation conditions lead to significantly higher oxygen concentrations, which could not be reached making use of the medium pressure DBD.
To investigate the chemical groups present on the treated samples, curve fitting of detailed C1s peaks is performed. Based on literature, the obtained C1s envelopes can be decomposed into 4 distinct components: a peak at 285.0 eV corresponding to C-C/C-H bonds, a peak at 286.5 eV attributed to C-O bonds present in alcohols and esters, a peak at 287.6 eV due to C = O bonds in aldehydes and ketones and a peak at 288.9 eV corresponding to O = C-O bonds in carboxylic acids and esters28. The results in Table 2 show that in terms of the type of incorporated oxygen-containing functional groups, the jet velocity does have an effect. Increasing the jet velocity seems to lead to higher carbonyl and carboxyl concentrations and a lower amount of C-O functionalities. This corresponds with other results found in literature29. Carton et al. hypothesized that the lower retention of carboxyls and carbonyls at lower jet velocities might be due to heating of the surface, which causes some of the carboxylic groups to be converted into CO2 or other volatile compounds. Table 2 also reveals that condition B leads to a lower amount of oxygen-containing functionalities, especially less carboxylic moieties. The relative percentages of oxygen-containing functional groups for condition F are quite similar to condition S3, showing that increasing the discharge power (i.e. energy density) does not have a significant effect on the way oxygen is incorporated into the surface.
The surface morphology is also researched using AFM and the acquired surface roughness (Rq) values for the different conditions are presented in Table 1. No significant plasma-induced effect on the surface roughness is found for all examined conditions, as the root mean square values for the treated samples are in the same order as for the untreated UHMWPE (Rq = 139 ± 8 nm). This corresponds to literature as it has already been reported that Ar plasma enhances radical reactions and restrains electron and ion etching effects30.
Figure 3 shows the evolution of the WCA as a function of storage time after plasma treatment. It is clear that for most samples the ageing effect stabilizes after 7 days of storage in ambient air. As ageing results in a loss of the treatment effect, the remaining treatment efficiency R (%) can be calculated using the following equation31:
where θsat is the saturation value of the WCA after plasma treatment, θageing is the value after 7 days of storage and θuntreated is the WCA value of the untreated material, which is equal to 87° in this case. These calculated R values are presented in Table 3. As could be deduced from the WCA results, samples treated with condition B show little to no ageing and keep most of their plasma treatment effect as an R value equal to 96% is found in this case. After 7 days in ambient air, the remaining treatment effect for condition F is lower (80%). For the S1, S2 and S3 conditions, this remaining treatment effect is even lower and drops to 77%, 76% and 73% respectively. Overall, the preservation of the treatment effect is much more pronounced after treatment with the plasma jet than with a medium pressure DBD-reactor, where R values of only 50% were found32, which is an additional advantage of the APPJ.
Table 3 shows the influence of the ageing effect on the chemical composition of the plasma-activated samples after storage for 7 days. For all conditions, a significant decrease (40–50%) of the O/C-ratio is observed, which correlates with the increasing WCA values. Comparing the percentages of the different carbon-containing functional groups immediately after treatment and after a storage time of 7 days shows that the amount of C-C/C-H bonds has increased and the percentage of C-O and O = C-O bonds has decreased for all conditions, while the amount of C = O bonds has remained more or less the same. These XPS results thus indicate a reordening of the functional groups during the ageing process and can be attributed to the tendency of a surface to restore its original surface energy18.
In order to assess the effect of the plasma treatment on the adhesion between MMA-based bone cement and UHMWPE, samples are fixed in bone cement and subsequently pulled out by a universal testing machine. The force that is needed to remove the UHMWPE samples out of the bone cement can be used as a measure for the adhesion between the two components. Figure 4 shows the results of the pull-out tests depicting the pull-out stress for different samples. It is clear that all plasma treatment conditions lead to a significantly improved adhesion between UHMWPE and bone cement. Even condition B, which caused a relatively minimal surface modification, results in a statistically significant difference (p < 0.05) in pull-out stress. This increase in adhesion can be correlated with the increase in oxygen concentration on the sample’s surface. More oxygen-containing functional groups result in more interaction with the functional groups present in the bone cement and therefore lead to a stronger adhesion. This can be seen in the difference between samples treated with condition B and the other conditions (p < 0.01), as all the other conditions result in a significantly higher oxygen concentration and as a result a significantly higher pull-out stress. Additionally, the type of oxygen functionality also seems to have an effect on the pull-out stress as there is a statistically significant difference between the samples S1 and S3 but no significant difference in O/C-ratio. The XPS-results in Table 2 however revealed that the S1 samples have a higher concentration of carboxyls. As the bone cement is based on MMA, it will contain a significant amount of ester functionalities. Therefore, the more carboxylic functionalities are present on the treated samples, the stronger the Van der Waals interactions and subsequently the stronger the adhesion between the two components will be. Hence, based on these results it can be concluded that plasma activation of UHMWPE with an APPJ results in a better adhesion between bone cement and UHMWPE, and that it is beneficial to use higher jet velocities during plasma activation as it leads to the incorporation of a higher amount of carboxylic groups.
Cell tests can be used to assess the viability of cells on a surface for different treatment conditions. Research has shown that multiple factors such as surface chemistry, topography and wettability influence the proliferation of cells on a substrate33,34,35,36,37,38. Earlier results have already showed that DBD-plasma treatment can significantly improve the proliferation of MC3T3 osteoblast cells on UHMWPE substrates5. Figure 5 shows the percentage of viable attached MC3T3 cells for different treatment conditions compared to a positive control culture for a culture time of 1 day and 7 days. These MTS assay results seem to indicate that (1) the S1 and S3 conditions lead to a significantly higher MC3T3 proliferation compared to the untreated sample 7 days after culturing and (2) the F and B conditions lead to a similar MC3T3 proliferation as the untreated sample 7 days after culturing. However, the fluorescence images depicted in Fig. 6 show that there is a significant difference in MC3T3 proliferation between the untreated sample on the one hand and all plasma-treated samples on the other hand. Additionally, there seems to be no large differences in MC3T3 proliferation amongst the different treatment conditions. The discrepancy between the fluorescence images and the results of the MTS assay can most likely be attributed to the low sensitivity of the MTS assay for this specific cell type. In this work, it can thus be stated that the conducted plasma treatments have a clear and significant effect on osteoblast cell viability, which can be attributed to the incorporation of oxygen in the surface and the subsequent change in surface wettability5,33,34,35,36,37. Furthermore, the fluorescence images shown in Fig. 6 demonstrate a difference in cell morphology between untreated and plasma-treated samples. While most of the cells on the treated samples have an elongated or triangular shape, which indicates attachment to the surface, the untreated samples contain considerably more cells with a round shape, which means these cells are not well attached to the surface. Therefore, it can be stated that plasma treatment not only improves MC3T3 proliferation, it also greatly enhances cell morphology and consequently cell adhesion on UHMWPE samples.
As mentioned before, the bioactivity of an uncemented sample can be examined by immersing it in SBF and studying the CaP formation on its surface. Figure 7 shows the FTIR-spectra of treated (condition S1) and untreated samples after immersion in 2.0 SBF for 14 days. It is clear that there is no significant difference between both spectra and therefore no difference in chemical composition of the CaP layer deposited on both sample types. The absorption bands at 560 cm−1, 600 cm−1 and around 1050 cm−1 show the presence of PO43− groups39. The absorption band from 2600 cm−1 to 3600 cm−1 can be attributed to absorbed H2O40. The absorption bands at 875 cm−1, 1420 cm−1, 1450 cm−1 and 1650 cm−1 arise due to the presence of CO32− groups, which substitute the phosphate ions40,41. It can therefore be concluded that B-type apatite is formed on both the untreated and plasma-treated samples40. Although there is no apparent difference in chemical composition, there is a significant difference in the homogeneity of the CaP layer on the different samples. As can be seen in the micrograph shown in Fig. 8, the CaP deposition does not cover the entire surface for both samples but is limited to different ‘islands’. On untreated UHMWPE however the deposition is localized near the edges of the sample, while on the plasma-treated sample the islands are distributed homogeneously over the entire surface. The homogeneity of the CaP deposition on the plasma-treated sample can be explained by the activation energy barrier that must be exceeded for nucleation to occur42. The homogeneous formation of functional groups such as –COOH and -COH on the surface due to the plasma treatment results in a decrease of the activation energy as the surface is saturated with nucleation sites. As such, a more homogeneous CaP deposition can be obtained on the plasma-treated sample. The SEM-images shown in Figs 9 and 10 confirm the formation of the CaP-islands and the obtained deposition seems to have a cauliflower shape with leaf-like shaped crystals for both samples. In conclusion, it can be stated that plasma treatment enhances the nucleation of a CaP layer on UHMWPE samples compared to untreated samples.