Curing Temperature Effects on the Tensile Properties and Hardness of γ− Fe2O3 Reinforced PDMS Nanocomposites

,e mechanical properties of plain polydimethylsiloxane (PDMS) and its nanocomposites have been exploited for various theranostic biomedical applications. Although several research groups have investigated the effects of preparation conditions—especially curing temperature and time—on bulk mechanical properties of plain PDMS, there are no reported similar studies for its nanocomposites. In this study, mechanical properties of PDMS reinforced by different volume fractions (φmnp � 0–2 vol. %) of c-Fe2O3 nanoparticles (NPs) were investigated and quantitative data presented for different curing temperatures (25, 100, and 150°C). To a large extent, c-Fe2O3 NPs were uniformly dispersed in the PDMS matrix with no primary chemical bonds formed. For the temperatures tested, the data showed an increase for Young’s modulus (E) of about 170% (1.36–3.71MPa) and a decrease of the ultimate tensile strength (UTS) of about 65% (6.48–2.93MPa) with increasing concentration of the NPs. Furthermore, hardness (Shore A) (H) increased with curing temperature but decreased with concentration. Based on the findings, we conclude that the linear relationship between the calculated mechanical properties (E, UTS, H) and small φmnp is independent of the curing temperature. ,e experimental data provide useful background information for the selection of processing parameters for PDMS nanocomposite fabrication.


Introduction
Polydimethylsiloxane (PDMS) is a silicone-based organic elastomer composed of a repeating n − C 2 H 6 OSi−n unit [1]. It has a glass transition temperature below −120°C and possesses an attractive combination of properties including inertness, nontoxicity, nonflammability, biocompatibility, optical transparency, and high elasticity [2][3][4]. Usually, PDMS comes as a two-part kit consisting of a prepolymer (monomer) and curing agent (cross-linker) that are combined in a 10 : 1 ratio during preparation [1]. e principal steps involved in the preparation of PDMS include (a) stirring of the monomer and cross-linker mixture for homogeneity, (b) degassing with vacuum pumps or centrifuges to remove bubbles, and (c) curing at different temperatures and times using ovens [1]. It can be prepared with a wide range of relatively easy techniques such as soft lithography and its derivatives, molding, dip casting, spin coating, and 3D printing [5][6][7][8]. Notably, the ability of PDMS to undergo relatively large elastic deformation has been exploited for the development of several theranostic biomedical applications such as lab/tissue/organs-on-chip devices, point-of-care devices, and two/three-dimensional cell culture [3,4,9]. Examples include micropumps employing elastomeric displacement amplification [10,11], cell-based biochips using elastomeric substrates [12,13], and microfluidic channel for pressure monitoring [14,15].
An increasing number of investigators have studied methods of customizing the elasticity of PDMS for different applications. Popular methods have focused on manipulating preparation and addition of relatively small amount of small particles. For the case of the former, parameters such as the monomer to cross-linker ratio as well as the curing temperature and time have been manipulated to alter the mechanical properties of PDMS. Liu and coworkers [16] investigated the influences of curing temperature (100, 150, 200, and 300°C) and time (between 30 min and 3 h) on Young's modulus and ultimate tensile strength and reported that they increased with temperature up to 200°C and then decreased. ey attributed the reduction to thermal decomposition. Johnston et al. [17] found that the compressive and tensile Young's modulus and the hardness (Shore A) increased by about 58% (117.8 to 186.9 MPa), 125% (1.32 to 2.97 MPa), and 22% (44 to 54 Sh A ), respectively, in the range of up to 40% strain when temperature was increased from 25 to 200°C. Campeau et al. [18] studied the effects of curing temperatures (50,80, 100, and 150°C) and sterilization techniques (UV, ethanol, and boiling water) on material properties of PDMS scaffold used in endothelial mechanobiological studies. eir results showed, amongst others, that there were significant differences in Young's modulus for the range of curing temperatures tested. Furthermore, they showed that sterilization by boiling affected the mechanical properties. e use of relatively small amounts of nanoparticles (NPs) fillers-any matter with dimensions between 1 and 100 nm [19]-to influence the mechanical properties of elastomers matrix such as PDMS has been known for decades. Notably, the incorporation of the NPs is known to increase stiffness, introduce the Mullins effect (an alteration of stiffness dependence on strain history), and change transient material behavior (e.g., stress relaxation and hysteresis). e mechanical properties of elastomeric nanocomposites have been attributed to how the fillers interact with the matrix as well as themselves: the former (filler-matrix interactions) has been shown to enhance the mechanical properties, whereas the latter (filler-filler interactions)-controlled by van der Waals and electrostatic forces-deteriorates them [20,21], which has been mainly attributed to the difficulty to uniformly disperse NPs within a polymer matrix due to their high surface area to volume ratio and the hydrophobicity of elastomeric polymer matrices. Several theoretical models are available to correlate mechanical properties such as Young's modulus and tensile strength with the volume fraction of nanoparticles [21][22][23][24]. e resulting composite material, so-called nanocomposite, has contributed significantly to the development of smart and novel applications in several fields of biomedical engineering including biosensing, thermotherapy, single molecule analysis, cell culturing, and biomimetics [25][26][27][28] due to their low cost, high strength, and simple manufacturing processes.
In recent years, magnetic ceramics-based NPs such as magnetite (Fe 3 O 4 ) and maghemite (c-Fe 2 O 3 ) have attracted a lot of interest in the fields of biomedical engineering because they exhibit exciting properties such as high electrical resistivity, ease of synthesis, resistance to corrosion/ wear, high stiffness, and biocompatibility [29][30][31]. e combination of these magnetic NPs with elastomeric polymers such as PDMS results in the formation of stimuli responsive magnetorheological nanocomposites. e possibility of manipulating the mechanical properties of these materials under the influence of magnetic fields is of interest in the field of biomedical engineering due to the opportunity it offers for the development of novel stimuli responsive and efficient biomedical engineering systems [32][33][34]. For instance, Kitano et al. exploited the alternating magnetic fieldcontrolled variable stiffness of a magnetorheological composite based on carbonyl iron particles and the polymer glycerol to develop a multijoint manipulator for laparoscopic surgery [33]. Antonel et al. showed that anisotropic properties were induced when CoFe 2 O 4 -PDMS composites were cured in the presence of an alternating magnetic field [34]. Furthermore, they reported a 4-point increase in Young's moduli along the direction parallel to the needles. Over the past few years, our group has been investigating the development of maghemite (c-Fe 2 O 3 ) NPs reinforced PDMS nanocomposite for the development of implantable devices for interstitial thermotherapy of tumors [35][36][37].
Although several investigators have separately studied the effects of curing temperature or nanoparticle concentration on the mechanical properties of PDMS under different preparation conditions, to the best of our knowledge, there has been no reported investigation of the effects of standard curing temperature conditions on bulk mechanical properties of PDMS nanocomposites. In this study, we investigated and presented quantitative data of the relationship between the bulk Young's modulus, E; ultimate tensile strength, UTS; and hardness (Shore A), H, of c-Fe 2 O 3 NPs reinforced PDMS that was prepared using soft lithography techniques and used to fabricate mechanical testing samples by a molding technique. e samples were prepared with varying volume fractions (0-2 vol. %) of the c-Fe 2 O 3 NPs and cured at three different curing temperatures (T c ) including 25, 100, and 150°C.

Materials.
Commercially available SYLGARD 184 Silicone Elastomer kit purchased from Dow Corning Corporation (Auburn, MI, USA) and 20 nm maghemite nanoparticles (99.5% purity) from US Research Nanomaterials Inc. (Houston, TX, USA) were used to fabricate all samples.

Mold Specifications.
In order to characterize the material properties of nanocomposite samples, two types of molds were manufactured in accordance with the respective American Society for Testing of Materials (ASTM) standards: ASTM D412-C and ASTM D2240-05 type A code for tensile and hardness testing, respectively.

Sample Preparation.
Nanocomposite samples were prepared by varying the volume fraction of c−Fe 2 O 3 NPs within a PDMS matrix. First, PDMS matrix was prepared by mixing a prepolymer base (monomer) and cross-linking agent (hardener) at the manufacturer's recommended ratio of 10 : 1. Second, the compound was stirred for 10 min with a spatula, before adding c-Fe 2 O 3 NPs, and stirred for another 5 min to ensure uniform distribution. ird, the resulting uncured mixture was degassed using a desiccator to remove bubbles, subsequently poured into respective molds, and degassed again to eliminate any remaining bubbles. Last, the samples were cured at temperatures recommended by Dow Corning: 25, 100, and 150°C for curing times of 48 h, 35 min, and 10 min, respectively [1]. e three types of samples that were prepared in the study were designated as MNP-0, MNP-1, and MNP-2 according to the volume fraction of nanoparticles. MNP-0, plain PDMS (c−Fe 2 O 3 -free nanocomposite), was studied as a control. e sample names and their compositions are summarized in Table 1.

Structural Characterization.
e crystal structure and phase purity of the c-Fe 2 O 3 NPs were substantiated using an X-ray diffractometer (D8 FOCUS X-ray, Bruker AXS Inc., Madison WI, USA). e experimental conditions used include Ni-filtered Cu-Kα radiation as the X-ray source at a power of 45 kV and 40 mA. Measurement was done at a temperature of 25°C, sampling angle of 0.08°, and 2θ scanning rate of 3 s step −1 . e morphology of the nanoparticles was studied with transmission electron microscope (Philips CM10, Philips Electron Optics, Eindhoven, e Netherlands). A Fourier transform infrared spectroscope (FTIR, Tensor 27, Bruker Inc., Madison WI, USA) was used to investigate the functional groups of the nanocomposite samples. A scanning electron microscope (SEM, Quanta 200 FEG MKII, FEI, Hillsboro OR, USA) was used to study the surface morphology of all samples (MNP-0, MNP-1, and MNP-2) as well as the micro distribution of the nanoparticles in the nanocomposites (MNP-1 and MNP-2). In order to prevent charging during scanning, the samples were coated with carbon. e photomicrographs were obtained at magnifications of 0.35 and 2 kx at an operating voltage of 10 kV.

Mechanical Characterization
2.5.1. Tensile Testing. Tensile testing of the samples was conducted on a Mark-10 ESM301 motorized test stand with force gauge model M5-500 (Mark-10 Corporation, Copiague, NY 11726, USA). Cross-head velocity was 25.4 cm min − 1 . e raw data was recorded as force and time, using the MESUR Lite software version 1.3.2. in sheets of Teflon were inserted between the samples and grips of the machine to ensure uniform pressure distribution across the sample. Samples were tested in triplicate, and the average ambient conditions during testing were 28°C and 59% relative humidity. e load, F (N), and time, t (s), data were transformed into engineering stress, σ e (Pa), and strain, ϵ e (1). σ was calculated as where A 0 (m 2 ) is the initial cross-sectional area of the test section of the sample. ϵ e was calculated as where △L (m) is the displacement and L 0 (m) is the initial length.
e stress-strain curve for all test results were plotted and fitted to a linear regression trendline in Microsoft Excel 2011 (Microsoft Corporation, Seattle, USA). According to Schneider et al. [38], Young's modulus of elastomers can be calculated via Hooke's law: using data in the linear region, which occurs at strain levels below 40%. ey also show that a geometric correction factor must be applied to accurately calculate the mechanical properties of elastomer-based materials. For the ASTM D412-C geometry used in this study, we used a correction factor of 0.50 38 . Furthermore, the UTS is calculated from the data as the stress at break point.

Hardness
Testing. H test was performed with a Sauter HBA 100-0 Shore A Durometer (Sauter GmbH, Balingen, Germany). A force of approximately 8 N was applied, and the corresponding Shore hardness value was read on the durometer. Measurements made were repeated three times and averaged across the surface of the sample to ensure the homogeneity of data. e average ambient conditions during testing were 28°C and 59% relative humidity. Figure 1(a) shows the X-ray diffraction spectra of the c−Fe 2 O 3 nanoparticles. e presence of peaks at 2θ � 30.28°, 35.72°, 43.24°, 57.24°, and 62.84°corresponding to (220), (311), (400), (511), and (440) diffraction planes confirmed the presence of c−Fe 2 O 3 . e average crystallite size of the powders was determined, using Scherrer's equation [39], to be about 12.29 nm. Figure 1 Figures 1(d) and 1(g)).

Structural Properties.
e fractured surfaces of sample MNP-1 were rougher than MNP-0 as the stress waves, cracks, and wrinkles increased with nanoparticle concentration (see Figures 1(e) and 1(h)). In addition, there was evidence of clustering of the nanoparticles and voids due to dislocation of the clusters because of the tensile load. It can be seen clearly from Figures 1(f ) and 1(i) that clustering and voids increased with nanoparticle concentration.

Mechanical Properties.
e mechanical test data of the three types of samples tested were analyzed. In total, 36 samples were tested and the values were determined by averaging the triplicates. e results are summarized in Table 2 and Figures 2-4.

Tensile Strength
(1) Young's Modulus. e results show that E increased with T c and ϕ mnp for the samples and temperatures tested in this study ( Table 2). E of MNP-0 increased by about 83% from    (Table 2). Similarly, the rate of change of E increased from ≈ 0.9 to ≈ 1.4 when ϕ mnp was increased from 0 to 2 vol. % (Table 2). Finally, the linear relationships between E and the independent variables T c and ϕ mnp are depicted in Figures 3(a) and 3(b), respectively.
(2) Ultimate Tensile Strength. e results show that UTS decreased with T c and ϕ mnp for the samples and temperatures tested ( Table 2). UTS of MNP-0 decreased by about 70% from 6.48 MPa to 4.51 MPa when T c was increased from 25°C to 150°C. For the case of MNP-1 and MNP-2, UTS decreased by about 69% (from 5.64 to 3.87 MPa) and 60% (from 4.88 to 3.71 MPa), respectively, for the same change in temperature. Furthermore, it can be seen from Figure 3(a) that UTS decreased linearly with T c for each sample. e rate of change of UTS as a function of ϕ mnp slightly varied from ≈ − 1 to ≈ − 0.9 when T c was increased from 25 to 150°C. e rate of change of UTS as a function of T c varied between 0.014 and 0.020 when ϕ mnp was increased from 0 to 2 vol. %. Finally, the linear relationships between UTS and the independent variables T c and ϕ mnp are depicted in Figures 3(a) and 3(b), respectively.

Hardness.
e results show that the H increased with T c but decreased with ϕ mnp for all samples and temperatures tested in this study (Table 2). When T c was increased from 25  of H as a function of ϕ mnp slightly varied between ≈ −2.12 to ≈ −2.62 when T c was increased from 25°C to 150°C. e rate of change of H as a function of T c varied between 0.073 and 0.077 when ϕ mnp was increased from 0 to 2 vol. %. Finally, the linear relationships between H and the independent variables T c and ϕ mnp are depicted in Figures 4(a) and 4(b), respectively.

FTIR.
According to the FTIR results, the characteristic peaks of PDMS were maintained after the addition of the nanoparticles and were independent of the concentration. is suggests that no primary chemical bonds were formed between the nanoparticle fillers and the polymer matrix during the material preparation process.

SEM Micrographs.
e SEM micrographs of the fractured surfaces presented in Figure 1 reveal the presence of clusters, which are dislocated under tension leaving voids on the surface. It is well known that the main factors that affect the mechanical properties of nanocomposites are the filler-matrix and filler-filler interactions [20,21]. e former improves the mechanical properties, whereas the latter deteriorates it. e filler-filler interactions (manifested as clusters) can be attributed to factors such as the high surface area to volume ratio of nanoparticles, which makes it difficult to uniformly disperse them within a polymer matrix aiding the formation of clusters as well as their incompatibility with the hydrophobic polymer matrices due to their hydrophilicity. Although, thorough mixing can achieve uniform filler dispersion with minimal or no clusters, preprocessing techniques such as the use of saline-coupling agents [40,41] and grafting polymers [42,43] are sometimes used to engineer the interfacial region.

Mechanical Properties.
e mechanical measurements obtained in this study show that the mechanical behavior of c-Fe 2 O 3 -PDMS nanocomposites is a function of T c (°C) and ϕ (vol. %). erefore, here, we discuss the results within the context of factors that affect the mechanical behavior of nanocomposites at a given temperature by comparing our results to previously reported results in the literature and theoretical models. As stated earlier, a high degree of particle-particle interaction leads to a deterioration of the performance, which is significantly improved if particlematrix interactions are optimal.

Tensile
(1) Young's Modulus. Several investigators have studied the effect of processing parameters such as curing temperature and time as well as the nanoparticle concentration on Young's modulus of elastomers and their nanocomposites. Generally, these reports are similar to our results. Johnston et al. explored the effect of curing temperature on Young's modulus of plain PDMS samples manufactured by SYL-GARD 184 and reported that Young's modulus of plain PDMS increased with curing temperature in the range of 25°C and 200°C 17 . Campeau and coworkers [18] found that higher curing temperatures stiffen the plain PDMS and Young's modulus varied between 1.7 and 3.7 MPa when temperature was increased within the range 50 − 150°C. Wu et al. [44] reported an elastic modulus of 1.56 MPa for plain PDMS cured at 150°C for 15 min. Also, they found that this varied between 1.71 and 2.34 MPa when weight fraction of carbon nanotubes was increased from 1.0 to 4.0.
Increasing the curing temperature of PDMS yields a more cross-linked network. Also, the bond strength of the polymerized system increases with increasing temperature due to a characteristic change from adhesive to cohesive bonding [45]. e polymerization intensity therefore increases due to increasing molecular mobility. ese processes eventually result in a material with higher moduli. e lower modulus of samples that were cured at 25°C could be attributed to inconsistencies in the cross-linking process.
ese anomalies are associated with curing temperatures that are less than 60°C [46].
Several theoretical models are available to correlate Young's modulus with nanoparticle volume fraction in a manner similar to our results. e Guth model [23,47] is a modified version of the model proposed by Einstein [48,49] to estimate the relationship between Young's modulus and rigid spherical particles. It accounts for the effects of the particle-particle interaction (Figure 1), which is not considered in the Einstein model. e Guth model is defined by where E m is Young's modulus of the plain PDMS (sample MNP-0) and g f is a term that is typically between 4 and 10 and accounts for the clustering effects. Young's moduli predicted using equation (4) for lower (� 4) and upper (� 10) limits of the factor g f are compared with the experimental values at 25°C and 150°C as shown in Figures 5(a) and 5(b), respectively. Similar to our experimental values, the predicted values increased linearly with ϕ mnp . e predicted values were closer to the experimental values when g f � 10 at both temperatures; however, the relative error increased with T c and ϕ mnp . For instance, at T c � 25°C and ϕ mnp � 1, the relative error was about 8%; this increased to about 9% when the ϕ mnp was increased to 2% at the same temperature. For the case of T c � 150°C, the relative errors were estimated to be around 19% and 24%, for ϕ mnp of 1 vol. % and 2 vol. %, respectively. is might be due to factors such as the influence of clustering as well as the nature of the interfacial interactions between PDMS matrix and c−Fe 2 O 3 filler.
(2) Ultimate Tensile Strength. Similar to our results, Liu and coworkers [16] studied the effects of different curing temperature on UTS and also found that the UTS decreased with curing temperature. Kan-Dapaah et al. [35] studied the UTS for the same range of nanoparticle concentration at a curing temperature of 100°C and found that the UTS decreased Advances in Materials Science and Engineering from 5.60 MPa to 2.41 MPa when the nanoparticle concentration increased from 0 to 10 wt.%. Based on the hypothesis that the fillers are uniformly distributed within the matrix [50,51], the tensile strength of polymer nanocomposites has been shown to theoretically depend mainly on factors such as adhesion strength, layer thickness, and distribution profile of the fillers within the matrix. Equation (5)  Prediction, g f = 4 Prediction, g f = 10 Experiment Prediction, g f = 4 Prediction, g f = 10 Experiment  Advances in Materials Science and Engineering that relates the ultimate tensile strength to the particle concentration. e model, which assumes very strong bonding between the matrix and the fillers, is defined by Here UTS T and UTS m are the ultimate tensile strengths obtained using the Turcsanyi model and the matrix, respectively, and B is the interfacial interaction parameter, which is an empirical constant for characterizing the interfacial adhesion strength. e ultimate tensile strength, predicted with (5), as a function of ϕ mnp was compared to the experimental data obtained at 25°C (Figure 5(c)) and 150°C ( Figure 5(d)). e fit of (5) to the experimental data was made by varying the constant B. Similar to our results, the UTS decreased linearly with ϕ mnp at both temperatures. Furthermore, at T c � 25°C, the predicted values were closer to the experimental data when the constant B was −10 and −15, respectively. Again, these discrepancies between experiments and predictions can be the influence of clustering and the nature of the interfacial interactions between PDMS matrix and c−Fe 2 O 3 filler.

Hardness.
Although hardness (Shore A) values increased with T c for all samples, they decreased with ϕ mnp at given T c . Schneider et al. [38] studied the effects of curing temperature on hardness of pure PDMS and also observed a similar behavior. Nobrega et al. [52] studied the effect of different oxide-based nanoparticles types (ZnO, BaSO 4 , and TiO 2 ) and suggested that the reason for the reduction could be related to the nature of the nanoparticle as well as the degree of particle-particle interaction within the polymer matrix.
ese interactions lead to the formation of agglomerates or clusters, which can be detrimental to the mechanical properties of the nanocomposites when their sizes exceed that of the polymer chain [52].

Conclusion
is study was performed to investigate and present quantitative data on the effects of curing temperature and nanoparticle volume fraction on the mechanical behavior of PDMS reinforced with cFe 2 O 3 nanoparticles. For the temperature range tested (25 − 150°C), Young's modulus increased by over 170%, from 1.36 MPa to 3.71 MPa, and the ultimate tensile strength decreased linearly from 6.48 MPa to 2.93 MPa when nanoparticle volume fraction was varied up to 2 vol. %. Hardness (Shore A) of samples MNP-0, MNP-1, and MNP-2 increased from 42.50 to 52.32, 39.67 to 48.88, and 37.38 to 47.08, respectively, when curing temperature was increased from 25°C to 150°C. On the other hand, the hardness (Shore A) obtained at curing temperatures 25°C, 100°C, and 150°C decreased from 42.50 to 37.38, 46.75 to 42.52, and 52.32 to 47.08, respectively, when the nanoparticle volume fraction was increased from 0 to 2%. In conclusion, we observed a linear relationship between the mechanical properties, and nanoparticle concentration was independent of the curing temperatures tested in this study.
Furthermore, the results obtained in this study provide useful background information for the selection of processing parameters for PDMS nanocomposite fabrication.

Data Availability
All data used to support this study are included in the article.

Conflicts of Interest
e authors declare that they have no conflicts of interest.