Mechanical Properties of Ultrahigh Molecular Weight PHEMA

Jul 21, 2010 - E-mail: [email protected]. ... For a more comprehensive list of citations to this article, users are .... Initiated-chemical vapor deposi...
0 downloads 0 Views 243KB Size
2116

Biomacromolecules 2010, 11, 2116–2122

Mechanical Properties of Ultrahigh Molecular Weight PHEMA Hydrogels Synthesized Using Initiated Chemical Vapor Deposition Ranjita K. Bose and Kenneth K. S. Lau* Department of Chemical and Biological Engineering, Drexel University, Philadelphia, Pennsylvania 19104 Received May 6, 2010; Revised Manuscript Received July 2, 2010

In this work, poly(2-hydroxyethyl methacrylate) (PHEMA), a widely used hydrogel, is synthesized using initiated chemical vapor deposition (iCVD), a one-step surface polymerization that does not use any solvents. iCVD synthesis is capable of producing linear stoichiometric polymers that are free from entrained unreacted monomer or solvent and, thus, do not require additional purification steps. The resulting films, therefore, are found to be noncytotoxic and also have low nonspecific protein adsorption. The kinetics of iCVD polymerization are tuned so as to achieve rapid deposition rates (∼1.5 µm/min), which in turn yield ultrahigh molecular weight polymer films that are mechanically robust with good water transport and swellability. The films have an extremely high degree of physical chain entanglement giving rise to high tensile modulus and storage modulus without the need for chemical cross-linking that compromises hydrophilicity.

1. Introduction Hydrogels are a class of materials that have found widespread acceptance in tissue engineering, primarily owing to their structural similarity to macromolecular-based components in the body. Poly(2-hydroxyethyl methacrylate) (PHEMA) is arguably one of the most versatile, widely studied and used representative of synthetic hydrogels ever since the pioneering work of Wichterle and Lim with cross-linked PHEMA hydrogels in 1960.1 Their similarities to soft living tissue, including high water content,2 soft rubbery consistency,3 porous structure,2 and, in some cases, biodegradability,4,5 as well as their ease of fabrication of various architectures,6 make them suitable for biomedical applications. PHEMA has been used extensively in applications such as controlled drug release,7-10 cell adhesion,11,12 cell growth, protein adsorption,13,14 separation devices,15,16 and biosensors.17 For use as any type of scaffold, the polymer must be biocompatible, have good transport properties for nutrient exchange, and be mechanically durable to support growth of cells. Lack of mechanical stability and poor transport are some of the biggest challenges faced by current scaffolds.18 PHEMA has been synthesized by a variety of techniques including conventional solvent-based polymerization,1,19 plasma polymerization,13,20 and more recently initiated chemical vapor deposition (iCVD).21,22 Importantly, iCVD is a one-step surface polymerization technique without the use of solvents. The elimination of solvents is especially beneficial in biomedical applications because residual organic solvents in solvent-processed matrices decrease the activity of incorporated biological factors and promote inflammatory responses in vivo.23,24 Additionally, unlike in plasma CVD, due to the absence of UV radiation, ion bombardment, and high energy electrons, iCVD yields polymer films with no uncontrolled chain cross-linking, undesired dangling bonds, or loss of chemical functionality of the monomer compared to films grown using plasma excitation.21,25 In this work, we synthesized PHEMA using iCVD, a novel vapor deposition technique, and evaluated properties of iCVD * To whom correspondence should be addressed. E-mail: klau@ drexel.edu.

PHEMA as potential soft tissue substitutes. iCVD uses a chemical vapor deposition environment to first activate an initiator in the gas phase by a resistively heated filament array to form active free radicals. In the second step, monomer units and the generated radicals adsorb onto a substrate, and in the final step, the adsorbed initiator radicals initiate surface polymerization by linking monomer units. To promote surface adsorption of reactive species, iCVD relies on a cooled substrate that is at a temperature (∼25 °C) much lower than the initiator activation temperature (250-300 °C). Also, by using controlled polymerization chemistries typical of liquid phase polymerization, iCVD produces exceptionally clean polymers with stoichiometric compositions, tunable molecular weights, and with no residual solvents, additives or plasticizers.26-28 In previous work, we have shown that iCVD can be successfully used to produce clean and stoichiometric PHEMA hydrogel thin films.22 Differential scanning calorimetry (DSC) measured a high glass transition temperature for iCVD PHEMA. Importantly, cell adhesion and direct contact cytotoxicity studies on iCVD PHEMA surfaces showed that the films were noncytotoxic to adult human dermal fibroblast (HDF) cells, a representative cell line. The absence of solvents and the lack of entrained monomer in iCVD synthesis, together with the stoichiometric nature of the synthesized films make these materials noncytotoxic to living cells, enabling cell adhesion and cell growth.22 In this work, mechanical properties, thermal transitions and water uptake of these iCVD PHEMA hydrogels are studied to evaluate their use as biomaterials and give us an insight on their suitability in biomedical applications. iCVD reaction parameters were tuned to achieve suitably thick freestanding robust polymer films. Additionally, a set of cross-linked PHEMA films were synthesized by iCVD to compare their properties to the linear un-cross-linked PHEMA films.

2. Materials and Methods 2.1. Polymer Synthesis. For the synthesis of PHEMA, the monomer 2-hydroxyethyl methacrylate (HEMA; 97% Aldrich) and the initiator di-tert-amyl peroxide (TAPO; 97% Aldrich), shown in Figure 1a and

10.1021/bm100498a  2010 American Chemical Society Published on Web 07/21/2010

Synthesized PHEMA Hydrogels

Figure 1. (a) Monomer 2-hydoxyethyl methacrylate (HEMA), (b) initiator di-tert-amyl peroxide (TAPO), and (c) polymer poly(2-hydroxyethyl methacrylate) (PHEMA).

b, respectively, were used without further purification. Depositions were carried out in a stainless steel, custom-built vacuum reactor, similar to those described previously.23,28 The monomer HEMA was heated to 70 °C to achieve sufficient vapor pressure, while TAPO was kept at room temperature. Vapors of HEMA and TAPO were metered into the reactor using needle valves. The optimized flow rates for HEMA and TAPO were 2 and 1 sccm, respectively, for the homopolymer synthesis. The backside cooled stage was maintained at 25 °C, and the filament wire array was heated to 300 °C. Depositions were carried out under a reactor pressure of 125 mTorr controlled using a downstream throttle valve and pressure controller (MKS Instruments) together with a dry vacuum pump (iH80, Edwards Vacuum). The deposition rate was monitored in situ using an interferometry system equipped with a 633 nm HeNe laser (JDS Uniphase) passing through a glass window positioned on top of the reactor. 2.2. Cytotoxicity and Protein Adsorption Studies. Previous results from direct contact cytotoxicity on iCVD PHEMA showed good cell survival.22 In this work, indirect contact cell cytotoxicity studies were performed in which iCVD PHEMA was soaked in the media and cell growth in the polymer leachate was monitored. iCVD PHEMA films were soaked in fibroblast culture media (Promocell) and the resulting leachate was further used for the cytotoxicity study. Adult human dermal fibroblast (HDF) cells (Promocell) were cultured in the leachate for up to 7 days, and the standard MTT assay29 was used to quantify cell viability. For the polymer leaching, 50 mg of the iCVD PHEMA film were sterilized for 30 min using UV radiation, after which 100 mL of HDF media was added to it. The films were soaked at 37 °C for 72 h and were then removed and discarded, while the liquid leachate was used for the MTT assay. For the cell culture, 1 mL of the leachate was added to each well of a 24-well plate and a 1 mL cell suspension with a cell density of 5000 cells per mL was added to each well. The well plates were maintained at 37 °C under 5% CO2 before being assayed using an MTT kit (Trevigen) after days 1, 3, 5, and 7. Absorbance at the end of the MTT assay was measured on a microplate reader (Biotek ELx800) at 530 nm with subtraction of background at 670 nm. As a comparison with as-deposited iCVD PHEMA films, iCVD PHEMA films, which had been washed in DI water and dried overnight prior to leaching, were also studied. For further comparison, unwashed and washed standard liquid synthesized PHEMA films (mol. wt. 300000 g/mol, Scientific Polymer Products, solvent cast from methanol) were also tested. As positive controls for cytotoxicity, HEMA monomer was used in a concentration of 10 µL/mL of media. Negative controls in the form of pure media wells were also employed. Therefore, in all, the samples assayed were the as-deposited iCVD PHEMA films, washed iCVD PHEMA films, as-cast and washed standard PHEMA films, and positive as well as negative controls. As with the as-deposited iCVD PHEMA films, all the other samples were tested using the same protocol, and cell viability was determined after 1, 3, 5, and 7 days. Each sample was tested in triplicate. Protein adsorption onto PHEMA films was quantified using the standard bicinchoninic acid (BCA) assay (Pierce, U.S.A.) by applying a model protein, bovine serum albumin (BSA, Pierce), directly onto the surface of the films. Films with a surface area of 2.5 cm2 were incubated in 24-well plates at room temperature for 12 h in 1 mL of buffer containing 350 µg BSA. Silicon wafers cut to the same surface area were used as positive controls because silicon surfaces have been

Biomacromolecules, Vol. 11, No. 8, 2010

2117

shown to lead to high nonspecific protein adsorption.30,31 The protein remaining in the supernatant after the incubation period was determined by first incubating 25 µL of the supernatant with 200 µL of BCA reagent solution at 37 °C for 30 min and then measuring the absorbance at 570 nm with a microplate reader (Biotek ELx800). Absorbance values were converted to concentration with respect to a standard calibration curve of known BSA concentrations. All samples including the calibration samples were assayed in triplicate. 2.3. Tensile Testing. Tensile testing was performed on iCVD PHEMA films preconditioned at two relative humidity levels, 98% and 16% RH, reflecting the difference between humid and dry environments. These tests were performed on an Instron universal testing machine series 3340 with a load cell of 50 N at a constant strain rate of 1 cm/ min. Uniaxial tensile specimens were cut to size using a razor blade. Typical length and width of the specimens tested were 20 and 6 mm, respectively, and film thicknesses were 100-200 µm. To preserve the humidity of the films during the entire experiment, a controlled humidity chamber was created around the tensioning arms and held at constant RH. 2.4. Dynamic Mechanical Analysis. Dynamic mechanical analysis (DMA) was performed on a TA Instruments DMA Q800 using a tension film clamp. The oscillation amplitude was 15 µm, and a static force of 0.01 N at a constant strain of 125% was maintained on the film throughout the experiments. For the temperature sweeps, a ramp rate of 3 °C/min was used and samples were heated to a maximum temperature of 170 °C at a frequency of 1 Hz. Such elevated temperatures do not degrade PHEMA as it has a reported thermal degradation temperature of >300 °C.32 Frequency sweep experiments were carried out at a constant temperature of 150 °C between 0.1 and 50 Hz. Sample dimensions were similar to those used in tensile testing. The specimen was inserted in the tension film clamps and the gauge length of the sample clamped in the grips was typically 20-25 mm. A preload of 0.01 N (corresponding to about 1 kPa of stress) was applied to the sample to ensure that the film remained vertically taut without inducing creep. For the temperature sweep experiments, each sample was cycled three times from 30-170 °C to ensure any thermal history of the polymers was erased. Prior to running the frequency sweep experiments, a temperature sweep was performed to again erase any thermal history. In addition to iCVD PHEMA, two PHEMA standard films (300000 and 1000000 mol wt, Scientific Polymer Products, solvent cast from methanol) were also studied for comparison. 2.5. Water Vapor Uptake Studies. Water sorption by iCVD PHEMA films was studied using gravimetric measurement of water vapor uptake from a constant humidity environment. Aqueous solutions of glycerol were used to maintain 98% RH at 25 °C in 50 mL sealed glass jars. Dry polymer films of about 200 µm in thickness and 2 cm2 of surface area were placed in these jars on a floating weigh boat and their water vapor uptake was measured gravimetrically at different times of 15 min, 30 min, 1, 2, 3, 5, 7, 9, 12, 18, and 24 h. Prior to these measurements, the polymer films were thoroughly dried in vacuum overnight and their initial weight was recorded. 2.6. Comparison with Cross-Linked Films. Cross-linking a polymer is a well-known way to improve its mechanical properties. The experiments described in this section were performed to compare linear un-cross-linked PHEMA with cross-linked PHEMA. Ethylene glycol dimethacrylate (EGDMA) was used as the cross-linking agent, similar to previous work done with cross-linking HEMA using iCVD.21 To find sufficiently cross-linked polymers, an initial study was done in which the flow rate of HEMA was fixed at 1 sccm, while flow rates of EGDMA were varied from 0.5 to 2 sccm in increments of 0.5 sccm. EGDMA flow rate of 2 sccm was found to give a sufficient degree of cross-linking, and hence, this flow rate was selected for the remainder of the experiments. A homopolymer of EGDMA was synthesized using iCVD for comparison as well. The composition and structure of both linear and cross-linked iCVD PHEMA films were characterized using FTIR spectroscopy. Spectra were acquired on a Thermo Scientific Nicolet 6700 spectrometer using a diamond ATR crystal and an MCT/A

2118

Biomacromolecules, Vol. 11, No. 8, 2010

Bose and Lau

Figure 2. Although typical free-standing linear PHEMA films are 100-200 µm thick, much thicker films can be produced, in this case ∼350 µm.

detector over the range of 650-4000 cm-1 at a resolution of 4 cm-1 averaged over 64 scans. In addition, contact angle measurements were performed using a KSV Instruments CAM 200 optical tensiometer to measure the change in contact angle of water on the cross-linked films compared to their linear counterparts. Dynamic mechanical analysis (DMA) and water vapor uptake studies were also performed on the cross-linked polymers using the same experimental procedures as for the linear polymers.

3. Results and Discussion 3.1. Polymer Structure. Nuclear magnetic resonance (NMR) and FTIR spectroscopies (described in detail previously22) were used to ensure that the linear PHEMA synthesized by iCVD was stoichiometric, devoid of cross-linking, and free from entrained monomer, as shown in Figure 1c. Depositions were carried out to yield films ranging in thickness from 100-200 µm. The high deposition rates (>1.5 µm/min) yielded thick freestanding PHEMA films, as shown in Figure 2. For iCVD, the ratio of monomer partial pressure to the saturation vapor pressure of the monomer (Pm/Psat) is a measure of surface availability of the monomer for polymerization.23 For these experiments, the Pm/Psat ratio was close to unity, which provided maximum monomer availability and contributed toward a high deposition rate. Previous work studying kinetics in iCVD of acrylate polymers has shown that the molecular weight is a function of deposition rate.33 In this work, the extremely fast deposition rates, as compared to other vapor deposition techniques, therefore predict the presence of extremely long molecular chains. The presence of long molecular chains most likely imparts good mechanical properties to these polymers, as will be discussed. There have been previous reports that commercially available HEMA monomer is known to contain trace impurities of the cross-linker EGDMA.34 To ensure the absence of inadvertent cross-linking in linear PHEMA films during iCVD synthesis using the HEMA monomer, as-deposited films were dissolved in DMSO-d6 and 1H NMR was performed on the solution before and after filtering through a 0.45 µm PTFE filter. Because there was no discernible decrease in the intensity of spectral peaks of the filtered solution (data not shown), it can be assumed that no measurable cross-linking was present. This is also supported by our previous work where spectral peak ratios of iCVD PHEMA matched that of a linear PHEMA standard.22 3.2. Indirect Contact Cytotoxicity. Noncytotoxicity, an essential property for a biomaterial, was tested using HDF cells and an MTT assay. Our previous work has shown that by growing HDF cells directly on iCVD PHEMA surfaces (a direct contact cytotoxicity study) the polymer supported cell adhesion and proliferation.22 Figure 3 shows the percentage of cells that survived at days 1, 3, 5, and 7. For the positive control samples of the HEMA monomer (not shown in graph), complete cell

Figure 3. Indirect contact cytotoxicity results show cell survival for adult HDF cells grown in polymer leachate. Percent cell survival was normalized to the control on day 1. A student’s t-test (p < 0.001 as indicated by *) reveals that iCVD PHEMA gave similar cell survival as the control, while solvent cast PHEMA standards gave poorer cell survival.

death was observed at the day 1 time point. Significantly, the unwashed standard PHEMA samples showed poorer survival compared to the controls at all time points. Interestingly, samples of the washed standards fared slightly better with poorer survival appearing only at longer time periods (days 5 and 7). These results strongly indicate that entrained monomer from the liquid phase synthesis or entrained solvent (in this case, methanol) from film casting might have contributed to cell cytotoxicity. A washing of the as-cast samples most likely reduced the amount of entrained monomer/solvent, leading to lower cytotoxicity, but which was not completely removed, as indicated by the increased toxicity later on. Leachate of entrained monomer and solvents is a major contributing factor in the cytotoxicity of polymers obtained from liquid-based polymerizations.35-37 In a previous report on the cytotoxicity of conventionally synthesized PHEMA, the authors reported a 35-50% decline in cell population growth compared to the negative control,38 which we believe might be due to entrained monomer and solvent leaching out over time. Importantly, our work with iCVD PHEMA films yielded no significant change in cell survival and similar growth patterns compared to the negative control. Additionally, there was no significant difference between the as-deposited and washed iCVD samples. Therefore, these results strongly support the lack of monomer entrainment in the growing polymer film during iCVD synthesis. It should be noted since no solvents are used in iCVD synthesis, these films are inherently free from entrained solvent. Polymers made by conventional liquid-based synthesis will typically have entrained solvent that demand post-synthesis purification. 3.3. Protein Adsorption. Adsorption of BSA onto PHEMA surfaces and silicon wafer positive controls using the BCA assay shows that the as-deposited films of PHEMA have lower BSA binding compared to bare silicon. The iCVD synthesized PHEMA film had a protein adsorption of 24.6 ( 11 µg/mL which was 3 times less than the protein adsorption of bare silicon 72.2 ( 22 µg/mL. Similar values of low protein adsorption have been observed for PHEMA synthesized from the liquid phase.13,31,39 The low binding of BSA on iCVD PHEMA suggests it is a promising polymer for synthetic scaffolds because it is known that nonspecific protein binding typically reduces biomaterial viability. 3.4. Tensile Properties. Tensile testing performed in dry as well as humidified iCVD PHEMA yielded stress-strain behavior as plotted in Figure 4. The Young’s moduli obtained from the slope of the curves for the dry and humidified films

Synthesized PHEMA Hydrogels

Biomacromolecules, Vol. 11, No. 8, 2010

2119

have been reported in the literature, the values obtained using DSC (ranging from 85 to 105 °C),45-47 are lower than those obtained using DMA (100-120 °C).48,49 In addition, from the DMA temperature sweep experiments, the dynamic elastic modulus was found to be 2051, 1984, and 1229 MPa, respectively, for iCVD PHEMA, and 1000000 and 300000 g/mol PHEMA standard films. Dynamic modulus (E) was calculated from

E ) E′ + iE′′

Figure 4. Stress-strain data of linear iCVD PHEMA films obtained under humidity conditions of 98 and 16% RH. Modulus of elasticity decreases significantly when a tensile test is conducted under humidified ambient as opposed to dry conditions.

(1)

where E′ and E′′ are the storage and loss moduli, respectively, at the low temperature plateau. From the frequency sweep experiments performed on DMA (data not shown), chain entanglement can be determined using the low frequency plateau of a storage modulus curve. Theory of rubber elasticity of cross-linked polymers gives a relation50 between molecular weight between chemical cross-links, Mc, and storage modulus, E′, for any temperature T, where T > Tg

Mc )

3FRT E′

(2)

where F is the density of the polymer, R is the universal gas constant, and T is the temperature. Likewise, for linear polymers, molecular weight between physical chain entanglements can be calculated using a similar equation,50 where Mc is replaced by Me,

Me ) Figure 5. DMA temperature traces of tan δ. iCVD synthesized linear PHEMA has a Tg of 134.7 °C based on the tan δ curve. Tg of PHEMA standards of 1000000 and 300000 g/mol are 139.4 and 136.6 °C, respectively.

are 16.8 and 1.1 MPa, respectively. For application as scaffolds, the films would be subjected to humidified or wet conditions so it is essential that they show mechanical durability in humidified conditions as well. Human skin has a Young’s modulus of 0.4-0.8 MPa, depending on the age of the subject, skin thickness, and its location on the body.40,41 Healthy cartilage on the other hand has a more widespread distribution of Young’s modulus depending on cartilage location in the body and mode of testing, that is, tensile, compressive, or shear loads. Typically, femoral cartilage showed higher Young’s (>1.0 MPa) and dynamic (>8 MPa) moduli than tibial or patellar cartilage (Young’s modulus