Chinese Journal of Polar Since  2018, Vol. 31 Issue (2): 211-215

#### The article information

Tanveer Iqbal, Saima Yasin, Ahmad Shakeel, Hamayoun Mahmood, Fahad Nazir, Paul F. Luckham
Tanveer Iqbal, Saima Yasin, Ahmad Shakeel, Hamayoun Mahmood, Fahad Nazir, Paul F. Luckham
Analysis of Solvent Effect on Mechanical Properties of Poly(ether ether ketone) Using Nano-indentation

Chinese Journal of Polar Since, 2018, 31(2): 211-215

http://dx.doi.org/10.1063/1674-0068/31/cjcp1709173

### Article history

Accepted on: December 20, 2017
Analysis of Solvent Effect on Mechanical Properties of Poly(ether ether ketone) Using Nano-indentation
Tanveer Iqbala, Saima Yasina, Ahmad Shakeela, Hamayoun Mahmooda, Fahad Nazira, Paul F. Luckhamb
Dated: Received on September 18, 2017; Accepted on December 20, 2017
a. Department of Chemical, Polymer & Composite Materials Engineering, University of Engineering & Technology, Lahore, KSK Campus, 54890, Pakistan;
b. Department of Chemical Engineering & Technology, Imperial College London, South Kensington, SW7 2 AZ, UK
*Author to whom correspondence should be addressed. Tanveer Iqbal, E-mail:tanveer@uet.edu.pk, Tel.:+92-300-8430776
Abstract: Poly (ether ether ketone)(PEEK) is a high-performance semi-crystalline thermoplastic polymer.Exposure of the polymeric surface to solvents can have a strong effect like softening/swelling of polymeric network or dissolution.In this study, nano-indentation analysis was performed to study the effect of acetone on the surface mechanical properties of PEEK using different exposure time.The experiments were performed with a constant loading rate (10 nm/s) to a maximum indentation displacement (1000 nm).A 30-second hold segment was included at the maximum load to account for any creep effects followed by an unloading segment to 80% unloading.The indentation hardness and the elastic modulus were computed as a continuous function of the penetration displacement in the continuous stiffness mode (CSM) indentation.The experimental data showed that the peak load decreased from ~5.2 mN to ~1.7 mN as exposure time in solvent environment increased from 0 to 18 days.The elastic modulus and the hardness of PEEK samples also displayed a decreasing trend as a function of exposure time in the solvent environment.Two empirical models were used to fit the experimental data of hardness as a function of exposure time which showed a good agreement with the experimental values.
Ⅰ. INTRODUCTION

Poly(ether ether ketone) (PEEK) is a semi-crystalline thermoplastic polymer. It is usually composed of molecular chains of aromatic backbone and the chains are interconnected through ketones group [1]. It is an odorless semi-crystalline polymer and gray tan in color [2]. The melting point ($T_\textrm{m}$) of PEEK is 334 ℃ and glass transition temperature is 149 ℃. Due to the high melting point and glass transition temperatures, it can be used in high-temperature applications [3]. PEEK has a lot of applications in different fields like orthopedic surgery, the metallic surgical implants are being replaced by the polymer due to its high mechanical strength [1, 4]. It has a very high strength-to-weight ratio as compared to any other thermoplastic, which permits the aerospace industry to replace heavy metallic units by plastic parts to fulfill arduous demands of industry [5]. It is also used in insulation of high voltage wires to avoid heat dissipation of wires [1]. PEEK has also been widely used as a polymer matrix material for high-performance composites due to excellent mechanical characteristics, chemical inactivity, higher wear resistance, and ease in industrial processing [6-8].

Nano-indentation or depth sensing indentation is an interesting tool for analyzing nano-scale mechanical properties of different materials like metals, polymers, and biological tissues. This technique is very useful because several non-destructive measurements can be performed on the same sample without the requirement of optical imaging as used in conventional indentation experiments [9]. Penetration depth and indentation load are two key parameters which can be measured using nano-indenter. Different mechanical properties, like hardness and Young's modulus, can be extracted from the numerical analysis of the obtained load-displacement curve [10]. The load is applied through an indenter for a specific period of time and the subsequent residual indentation size is measured based on a calibration measurements from silica. The contact compliance method has been recently adopted for hardness and modulus computation; this method is based on the imposed displacement/reaction force response [11-13].

The exposure of solvent to the polymeric surface can have a strong effect which includes softening/swelling (volume increase) of polymeric network or dissolution in case of strong solvents [14]. The presence of physical entanglements in polymer chains resists the dissolution process, to achieve complete dissolution, these entanglements must be removed either by breaking or by sliding along their length [15]. Dissolution of PMMA has been studied under the influence of a number of solvents like tetrahydrofuran (THF), methyl acetate (MA) and methyl isobutyl ketone (MIBK). It was found that the diffusion rates and swelling power of MA and THF molecules were higher than that of MIBK [16]. Some non-solvents, like water, methanol, and ethanol, were also reported to be used in a lower concentration to enhance the dissolution rate. It was also noticed that higher alcohols diminished the dissolution rates of polymer and also the high concentration of non-solvents led towards the swelling of the polymeric film to a greater extent [17]. Shao and Vollrath [18] reported the effect of different solvents, such as water, urea solution, and different alcohols, on the contraction and mechanical properties of four different types of silk, which behave as a hard elastic polymer. Leach et al. [15] examined the modification of PMMA surface under the influence of different solvents like water, methanol, ethanol, and aqueous alcohol solutions using atomic force microscope. The results showed that the solvent exposure causes roughness and swelling of the polymeric surface. Brown et al. [19] also reported the fabrication of PMMA microfluidic chips after modifying the PMMA surface by hydrolysis and aminolysis processes. Nano-indentation tests were also performed to analyze the changes occurring at the surface of PMMA as a result of solvent treatment.

To date, the effect of solvents on the mechanical properties of PEEK has not been investigated in detail. Hence, in this work, a nano-indentation analysis was performed to study the effect of selected solvent on the mechanical properties (hardness and elastic modulus) of PEEK under different exposure time of solvent.

Ⅱ. EXPERIMENTS

Commercially available 1.2 mm thick sheets of semi-crystalline PEEK were purchased from Good fellows, UK for the experiments. The polymer sheets were used without any prior treatment. Analytical grade acetone (purity$>$99%) was utilized as a solvent for the plasticization study. PEEK samples were dipped into the selected solvent for different intervals of time (3 h, 3 days, 7 days, and 18 days).

A NANO INDENTER® IIs machine (provided by Nano Instruments Ltd., Tennessee, USA) was used for conducting the normal indentation experiments on the modified and un-modified PEEK samples. The machine uses a contact compliance indentation mode for analyzing surface mechanical properties of materials. Thus, the material hardness and the elastic modulus can be calculated without actually determining the area of indentation. In conventional indentation, the normal area of deformation is determined using optical methods (such as an optical microscopy or AFM method) after indentation. Whereas, in normal indentation, the area is determined using the geometry of the indenter tip and the depth of the indentation in contact compliance mode indentation. The depth of indentation could be determined by constantly monitoring the indenter position relative to the specimen surface. The standard tip area calibration indentations were performed against fused silica to determine the indenter tip geometry. A continuous stiffness mode indentation (CSM) was adopted for the present study. The indentation hardness and the elastic modulus are computed as a continuous function of the penetration displacement in the CSM indentation. A trigonal diamond pyramid tip, also known as a Berkovich Indenter, was mounted to the bottom of the indenter column. The Berkovich tip makes an angle of 65.3$^{\circ}$ with the normal to the base of indenter [20]. The normal force on the indenter was generated with the help of a magnetic field containing an electromagnetic aluminum coil placed at the top of inner indenter tube. The nano-indentation experiments were performed with a constant loading rate (10 nm/s) to a maximum indentation displacement (1000 nm). A 30-second hold segment was included at the maximum load to account for any creep effects followed by an unloading segment to 80% unloading; at this point, a final hold segment of 100 seconds was applied to account for any thermal drifts during the indentation experiments.

Ⅲ. RESULTS AND DISCUSSION A. Effect of solvent on crystallinity of PEEK

FIG. 1 represents the loading and unloading cycle curves obtained from the nano-indentation analysis performed in five different regions (Indent 1-5) on the surface of virgin semi crystalline PEEK. The region in which the indentation load approaches to zero at very low indentation displacement during loading section is known as induction phase of indentation experiments [21]. It may arise due to the surface determination error, unreliable calibration procedure or unexplainable physical imperfection of indenter tip. The surface determination errors usually resulted from the roughness of surface at nano-matric scale. It can be seen from the figure that a harder behavior was evident for Indents 1 and 2 which requires 5.11-5.23 mN load to impose a penetration depth of 1000 nm whereas a decreased value of load 4.69-4.80 mN was observed for Indents 3, 4 and 5 which represents the softer region. This may be linked with the existence of crystalline and amorphous regions within the PEEK sample; in which the crystalline part shows a harder response than the amorphous one.

 FIG. 1 Load-displacement curves from five different positions (Indent 1-5) of virgin PEEK sample.

Similarly, FIG. 2 represents the load-displacement curves of nano-indentation experiments performed in five different regions of PEEK sample immersed in acetone for 18 days. It can be easily seen from the figure that all the curves follow more or less the same pattern and require $\sim$1.7 mN load to produce a penetration depth of 1000 nm. This result may represent the existence of only the softer/amorphous regions of polymer and the elimination of crystalline part from PEEK sample after having contact with the acetone for 18 days.

 FIG. 2 Load-displacement curves from five different positions (Indent 1-5) of PEEK sample having 18 days of exposure in acetone.
B. Effect of solvent on peak load

FIG. 3 represents the load-displacement curves of PEEK samples as a function of different exposure time in acetone environment. PEEK samples were dipped in acetone for the indicated times and it can be seen from the figure that the value of peak load is greatly affected by the exposure time in acetone. The value of peak load was maximum $\sim$5.2 mN for virgin PEEK while the minimum value of peak load $\sim$1.7 mN was observed for the exposure time of 18 days in acetone. Polymer segmental motion becomes enhanced due to the plasticizing/swelling effect of the solvent after entering into the polymer network and led to the further solvent intake [15]. The observed low values of peak load may be due to this softening or swelling caused by the increased rate of solvent intake.

 FIG. 3 Load-displacement curves of PEEK samples as a function of different exposure time in acetone.
C. Effect of solvent on hardness

FIG. 4 presents the hardness of semi crystalline PEEK as a function of indentation depth for different exposure time in acetone. Remarkable fluctuations were seen for hardness values at a very low penetration depth. These uncertainties were thought to be because of defects in the geometry of tip, errors in the determination of surface or effects of indentation size [20, 22]. At very low penetration depths (0-150 nm), a harder and more changing behavior was observed. This can be attributed to the confined changes in the physical properties of the material under the influence of harsh environment sometimes before the experimentation or during the fabrication process of polymers. These modifications near to surface may have occurred due to the localized oxidation of polymeric surface or due to the segregation of modifiers or other impurities near to the surface.

 FIG. 4 Hardness of PEEK samples as a function of indentation depth for different exposure time in acetone.

The response of hardness close to the surface is analogous to that observed by Briscoe and co-workers [20] in which they have studied a large number of vulnerabilities with respect to the hardness of polymer without applying continues stiffness mode. It can be seen that the highest value of hardness was observed for unmodified PEEK which remains more or less constant ($\sim$0.28 GPa) after the indentation depth of 300 nm. The values of hardness displayed a decreasing trend as a function of indentation depth for PEEK samples immersed in acetone environment, which becomes more pronounced for longer exposure times. This result can be linked with the softening/swelling of the polymer samples not only near the surface but further inside the sample as well due to the faster solvent intake. The hardness response to the corresponding indentation depth with different exposure times resembles the one investigated by Iqbal and co-workers for different polymers [9, 23].

D. Empirical model fitting

An empirical model was proposed to predict the hardness of PEEK sample as a function of exposure time in acetone environment, given as follows:

 $\begin{eqnarray} \textrm{ln}H=a+b\sqrt{t_\textrm{e}} \end{eqnarray}$ (1)

where $H$ and $t_\textrm{e}$ represent hardness and exposure time respectively, while $a$ and $b$ are fitting parameters having values of -1.244 and -0.055, respectively. FIG. 5 represents the fitting of experimental data with the proposed model and shows a good agreement ($r^2$$>0.99).  FIG. 5 Hardness as a function of exposure time for indentation displacement of 800 nm. The solid line represents model fitting of Eq.(1). Another empirical model was also proposed, similar to the Maxwell model given for the relaxation modulus as a function of time [25], as follows:  \begin{eqnarray} H(t)=\alpha \textrm{exp}\left(\frac{-t_\textrm{e}}{\beta}\right) \end{eqnarray} (2) where H(t) and t_e represent hardness and exposure time respectively, while \alpha has the units of hardness and its value (0.27 GPa) was more or less equal to the value of hardness of virgin PEEK, so this parameter can be represented as H_0 (hardness of pure polymer) as also represented in Maxwell model (E_0). The value of parameter \beta was 305 h and as this parameter has the dimensions of time, it may represent the value of exposure time in considered solvent after which the changes in mechanical properties of studied polymer will be very small. FIG. 6 represents the fitting of experimental data with this proposed model and shows a quite good agreement (r^2$$>$0.90).

 FIG. 6 Hardness as a function of exposure time for indentation displacement of 800 nm. The solid line represents model fitting of Eq.(2).
E. Effect of solvent on modulus

FIG. 7 represents the modulus data as a function of indentation depth obtained for semi-crystalline PEEK after immersing it in acetone environment for indicated time. It can be easily seen from the figure that the highest modulus was observed for virgin PEEK sample (4.1$\pm$1) GPa, comparable to the literature value of 4 GPa [24].

 FIG. 7 Modulus of PEEK samples as a function of indentation depth for different exposure time in acetone.

The modulus values of PEEK samples decreased with increasing indentation depths and this behavior becomes more prominent for larger exposure time. At very small indentation depth, higher values of modulus were observed as already explained for hardness behavior. Another interesting thing about modulus behavior can also be seen from the graph that after 3 dayss of exposure to acetone, modulus response showed a remarkable change as a function of indentation depth. Thus, a decrease in elasticity of PEEK samples was observed after the solvent exposure and this may again be attributed to the softening of the polymer network.

Ⅳ. CONCLUSION

The effect of solvent on the mechanical properties of PEEK was studied by adopting a normal mode of nano-indentation. A constant loading rate of 10 nm/s and a maximum indentation displacement of 1000 nm were used to perform nano-indentation experiments. A 30 s hold segment was included at the maximum load to account for any creep effects followed by an unloading segment to 80% unloading. The experimental data revealed that the crystallinity, peak load, elastic modulus, and hardness of PEEK samples decreased as exposure time in solvent environment increased, due to the softening/swelling of polymeric network. Uncertainties were observed in the data of hardness and elastic modulus at a very low penetration depth near to the surface. This was linked with the change in the physical and mechanical properties of polymers during their manufacturing or aging. However, the most prominent reason for this effect was the imperfection presented in the calibration of indenter tip. Two empirical models were used to fit the experimental data of hardness as a function of exposure time, which showed good agreement with the experimental values. Furthermore, the applicability of these models on the results of nano-indentation of different polymeric systems can be further investigated to incorporate more parameters into the models related to the characteristics of considered systems.

Reference
 [1] S. Kurtz, PEEK Biomaterials Handbook, 1st Edn., Oxford: William Andrew, (2011). [2] J. H. Wu, M. S. Yen, C. W. Chen, M. C. Kuo, F. K. Tsai, J. S. Kuo, L. H. Yang, and J. C. Huang, J. Appl. Polym. Sci. 125 , 494 (2012). DOI:10.1002/app.v125.1 [3] G. Zhang, W. Y. Li, M. Cherigui, C. Zhang, H. Liao, J. M. Bordes, and C. Coddet, Prog. Org. Coat. 60 , 39 (2007). DOI:10.1016/j.porgcoat.2007.06.004 [4] S. M. Kurtz, and J. N. Devine, Biomaterials 28 , 4845 (2007). DOI:10.1016/j.biomaterials.2007.07.013 [5] L. O. Dandy, G. Oliveux, J. Wood, M. J. Jenkins, and G. A. Leeke, Polym. Degrad. Stab. 112 , 52 (2015). [6] A. M. Díez-Pascual, M. Naffakh, J. M. González-Domínguez, A. Ansón, Y. Martínez-Rubi, - Rubi, M. T. Martínez, B. Simard, and M. A. Gomez, Carbon 48 , 3485 (2010). DOI:10.1016/j.carbon.2010.05.046 [7] X. LiuJie, J. P. Davim, and R. Cardoso, J. Mater. Process. Technol. 189 , 374 (2007). [8] V. S. Nisa, S. Rajesh, K. P. Murali, V. Priyadarsini, S. N. Potty, and R. Ratheesh, Compos. Sci. Technol. 68 , 106 (2008). DOI:10.1016/j.compscitech.2007.05.024 [9] T. Iqbal, B. J. Briscoe, S. Yasin, and P. F. Luckman, Chin. J. Polym. Sci. 31 , 1096 (2013). [10] Y. C. Lin, Y. J. Weng, D. J. Pen, and H. C. Li, Mater. Des. 30 , 1643 (2009). DOI:10.1016/j.matdes.2008.07.028 [11] I. M. Hutchings, J. Mater. Res. 24 , 581 (2009). DOI:10.1557/jmr.2009.0085 [12] T. Iqbal, B. J. Briscoe, S. Yasin, and P. F. Luckham, J. App. Polym. Sci. 130 , 4401 (2013). [13] T. Iqbal, B. J. Briscoe, and P. F. Luckham, Eur. Polym. J. 47 , 2244 (2011). DOI:10.1016/j.eurpolymj.2011.09.022 [14] H. L. Frisch, J. App. Polym. Sci. 6 , 1657 (1970). [15] R. N. Leach, F. Stevens, C. Seiler, S. C. Langford, and J. T. Dickinson, Langmuir 19 , 10225 (2003). [16] A. C. Ouano, and J. A. Carothers, Polym. Eng. Sci. 20 , 160 (1980). DOI:10.1002/(ISSN)1548-2634 [17] W. J. Cooper, P. D. Krasicky, and F. Rodriguez, J. Appl. Polym. Sci. 31 , 65 (1986). [18] Z. Shao, and F. Vollrath, Polymer 40 , 1799 (1999). DOI:10.1016/S0032-3861(98)00266-3 [19] L. Brown, T. Koerner, and Horton, Lab on a Chip 6 , 66 (2006). DOI:10.1039/B512179E [20] B. J. Briscoe, L. Fiori, and E. Pelillo, J. Phys. D:Appl. Phys. 31 , 2395 (1998). DOI:10.1088/0022-3727/31/19/006 [21] H. Mohammad, Ph. D. Dissertation, UK: Imperial College London, (2004). [22] M. Bonne, B. J. Briscoe, C. J. Lawrence, S. Manimaaran, D. Parsonage, and A. Allan, Tribology Lett. , 18125 (2005). [23] T. Iqbal, B. J. Briscoe, S. Yasin, and P. F. Luckham, J. Macromol. Sci. Phys. 53 , 1522 (2014). [24] K. Nakamae, T. Nishino, Y. Shimizu, and T. Matsumoto, Polym. J. 19 , 451 (1987). [25] S. P. C. Marques and G. J. Creus, Computational Viscoelasticity, 1st Edn., Berlin Heidelberg: SpringerVerlag, 11(2012).

Tanveer Iqbala, Saima Yasina, Ahmad Shakeela, Hamayoun Mahmooda, Fahad Nazira, Paul F. Luckhamb
a. 巴基斯坦拉合尔工程技术大学KSK校区, 化学聚合物与复合材料工程系 54890;
b. 伦敦帝国理工学院化学工程与技术系, 南肯辛顿, SW7 2 AZ