Polyisobutylene-based biomaterials

Authors


Abstract

This article highlights the biomaterial-related research of the Macromolecular Engineering Research Centre (MERC). The MERC group concentrated on polyisobutylene (PIB)-based biomaterials. In this article, first the unique properties of PIB are discussed, followed by a review of PIB-based potential biomaterials. MERC's systematic research program aimed to develop novel PIB-based biomaterials is then highlighted, including surface modification and biocompatibility studies. © 2004 Wiley Periodicals, Inc. J Polym Sci Part A: Polym Chem 42: 3091–3109, 2004

INTRODUCTION

Biomaterials have become a very attractive research area in polymer science. The importance of this area cannot be emphasized enough; the aging population in developed countries and the spread of devastating illnesses such as acquired immunodeficiency syndrome in the developing world necessitate the development of improved materials for medical applications. The use of polymers in medicine ranges from devices such as medical tubing and blood bags, to carriers for controlled drug release and artificial organs. Accordingly, the requirements for material properties can range from only mechanical and chemical performance to a combination of those with biocompatibility and biostability for long-term implant application. In the past, medical devices were developed using commercially available polymers. These were not specifically designed for the intended use. For instance, vascular grafts (artificial arteries) are made of polyester (Dacron) or fluoropolymer (Gore-Tex), which are not a real match for the soft arterial wall.1 To date, there is no satisfactory synthetic small-diameter vascular prosthesis.2, 3 In addition, during long-term use of certain polymers in implants, undesirable effects surfaced rather unexpectedly. Examples are plasticized polyvinyl chloride medical tubing4 or breast implants.5 This highlights the importance of designing polymeric materials specifically for the intended use, and thorough testing with the intended environment in mind. Unfortunately, this is easier said than done. Polymer scientists are expert in testing the chemical, physical, and mechanical properties of polymers; however, biocompatibility and biostability are elusive concepts. The European Society for Biomaterials arrived at the following definition: “Biocompatibility: the ability of a material to perform with an appropriate host response in a specific application.”6 Similarly, a material is considered biostable if “it does not degrade in the human body.”6 Material scientists and engineers are not used to such nonspecific definitions, as we pointed out in a recent review of polymers currently used in soft tissue replacement.1 Testing methods of the biocompatibility of a material have evolved over approximately the last 50 years. The first systematic approach was developed by Autian in 1961,7 and the scheme for testing plastics was codified by the U.S. Pharmacopoeia (USP) in 1965 and accepted as the norm by the Food, Drug, and Cosmetic Act in 1976.8 All new materials intended for biomedical use must pass the appropriate tests. This seems to be the most difficult challenge for polymer scientists; the tests, especially the in vivo tests, are very elaborate, expensive, and inherently involve risks. Thus, we must learn to select only the most promising materials. It has been recognized that, in addition to “bulk material” properties, surface properties are of critical importance in biomedical application. With the development of surface testing methods such as atomic force microscopy (AFM) or photoelectron spectroscopy (XPS), together with healthcare professionals, we must develop testing protocols that include both bulk and surface testing.

This article discusses a very promising class of biomaterials based on polyisobutylene (PIB). Several research groups reported the synthesis and characterization of potential PIB-based biomaterials. Our research group has also been active in this area. This article highlights our activities, with a brief review of publications related to PIB-based biomaterials. We also report our work on surface modification and surface testing of potential biopolymers.

PIB

PIB is a specialty polymer that can only be obtained by the cationic polymerization of isobutylene (IB) (Scheme 1). The polymerization of IB by strong acids at room temperature was first reported in 1873.9 The reaction yielded a “sticky liquid” (mainly dimers and trimers). PIB has unique properties: very low permeability, good thermal and oxidative stability, ozone resistance, chemical resistance, high hysteresis (mechanical dampening), and tack. Low- and medium-molecular-weight PIBs are used as viscosity modifiers, fuel and lubricating oil additives, tack improvers in adhesive formulations, and primary binders in caulking and sealing compounds. The discovery of low permeability led to the development of butyl rubber. Butyl elastomer is a random copolymer of IB and a small amount (1–4 mol %) of isoprene (Scheme 1). This copolymer was the first example of low unsaturation elastomers. The starting point of the development of butyl rubber was a collaborative research effort conducted by Otto (I. G. Farbenindustrie) and Thomas (Standard Oil Development Co.) in 1933–1935. The goal was the production of elastomers that could be cured to yield crosslinked PIB-based rubbers.10–13 The first butyl plants (Baytown and Baton Rouge, LA, and Sarnia, Canada) were constructed as part of the Synthetic Rubber Procurement Program during World War II, which was the second most important war program next to the Manhattan Project. The production of butyl rubber is still based on the original technology,14, 15 but a solution process called the “Russian process” was developed as well in the former Soviet Union.16 The most important physical properties of butyl rubber are essentially the same as those of PIB: low permeability, good chemical and thermal stability due to the low unsaturation content, and high damping. Butyl rubber is crosslinked (cured or vulcanized) with sulfur-based chemistry and mostly used as inner liner or tube in tires to prevent air leakage. Without butyl rubber, we would have to pump up our tires daily. Other applications include tank liners, vibration dampeners, protective clothing, tire curing bladders, railway pads, wire and cable coating, belting, and hoses. Both PIB and non-crosslinked butyl elastomers are approved by the Food and Drug Administration for chewing-gum base and other food-related applications.8 Crosslinked butyl rubber is also used for pharmaceutical stoppers and blood bags, because of its excellent barrier properties.

Scheme 1.

Polymerization of isobutylene (IB) to form polyisobutylene (PIB); copolymerization with isoprene (IP, 1–4 mol %) to form butyl elastomer.

Several researchers produced PIB-based structures intended for biomedical applications. First we review the development of PIB-based biomaterials, followed by a review of our work in this area.

Development of PIB-Based Biomaterials

PIB-based potential biomaterials have been developed using both conventional and living carbocationic polymerizations. Both the conventional and living carbocationic polymerization of IB have been thoroughly reviewed; some examples are given.17–20 Living carbocationic polymerization gives unprecedented control over the polymerization process. This said, the control achieved in polymerizations currently termed “living” or “controlled,” is a far cry from the precision controlled really living synthesis of proteins or other biological macromolecules. Percec et al.21 recently published a concept that could approach this control; with a rate constant of propagation (kp) controlled by polymer chain length, in principle nearly monodisperse polymers could be prepared. However, this concept has not yet been applied to IB polymerizations. Nevertheless, the control provided by current living IB polymerizations allowed the precision synthesis of an inventory of PIB structures, shown in Figure 1. Several academic research groups reported the synthesis and characterization of these structures. First, we will briefly review work related to potential PIB-based biomaterials.

Figure 1.

PIB homo- and block copolymers, and APN structures.

Potential PIB-Based Biomaterials

PIB has been combined with materials known to be biocompatible (polyacrylates and -methacrylates, polysiloxanes, polylactones, polyurethanes, poly(ethylene oxide), and poly(vinyl alcohol) (PVA). Numerous articles reported the synthesis and characterization of PIB combined with polyacrylates, polymethacrylates, and their derivatives; these are included in several reviews.22–24 Various amphiphilic blocks and networks containing PIB and poly(vinyl ether) derivatives were also reviewed.25, 26

PIB–polysiloxane block copolymers were synthesized via the anionic ring-opening polymerization of hexamethylcyclotrisiloxane, initiated by PIB-OH with n-BuLi27 (Scheme 2).

Scheme 2.

Synthesis of PIB–polysiloxane block copolymers.

Di-, tri-, multi-, and star-block copolymers comprising of poly(ethylene glycol) (PEG) and PIB blocks were produced by two approaches. One approach was to couple hydroxyl-terminated PEGs with isocyanated-terminated PIBs as shown in Scheme 3.28, 29 Another approach demonstrated in Scheme 4 was the hydrosilylation of allyl-terminated PEGs by Si(CH3)2H-terminated PIBs.30 It was found that PIB–PEG block copolymers containing 50–70 wt % PIB produced hydrogels, the integrity of which was maintained by physical crosslinks by PIB segments.30

Scheme 3.

Synthesis of PIB–PEG block copolymers and networks.

Scheme 4.

Hydrosilation of allyl-terminated PEGs by Si(CH3)2H-terminated PIBs.

The synthesis of amphiphilic PIB–PVA oligomers was also reported. Aldehyde-terminated PIB was chain-extended with t-butyl-dimethylsilyl vinyl ether through silyl aldol condensation.31 After monomer consumption, the t-butyl-dimethylsilyl groups were then hydrolyzed with fluoride in methanol to yield PIB–PVA (Scheme 5). The synthesis of PVA segments by silyl aldol polymerization showed some living characteristics, including stoichiometric number-average molecular weight (Mn) control. However, an attempt to reach higher molecular weight was not reproducible and often proceeded with low monomer conversion.31

Scheme 5.

Synthesis of PIB–PVA block copolymers.

PIB-b-poly(pivalolactone) (PPVL) diblock and PIB-PPVL-PIB triblock copolymers were synthesized by the anionic ring-opening polymerization of PVL using PIB macroinitiators with carboxylate potassium salt end groups (Scheme 6).32 The macroinitiators were obtained by quenching living diphenylethylene-capped PIB carbocations with 1-methoxy-1-trimethylsiloxypropene, followed by hydrolysis of the ω-methyoxycarbonyl end groups.

Scheme 6.

Synthesis of PIB–PPVL block copolymers.

Polyurethane-segmented copolymers are well known for a variety of biomedical applications.33 Polyurethaneurea-polyether (PEUU) is most prominently used in blood sacs in ventricular assist devices and artificial heart valves. It was thought that the incorporation of IB into PEUU would reduce the relatively rapid permeation of air and water vapor through PEUU membranes.34 Unfortunately, only modest reduction in water vapor and oxygen permeability was achieved in PEUU–PIB comb-like copolymers, which was comparable to that observed in polyurea-polyurethane containing polycarbonate soft segments.35

The materials discussed above are all potential biomaterials, but specific biocompatibility and biostability testing of these has not been reported. The next section discusses PIB-based biomaterials that have been tested in vitro or in vivo for biomedical applications.

PIB-Based Biomaterials Tested for Biomedical Applications

PIB–Poly(Methyl methacrylates) (PMMA) for Toughening of Bone Cements

Contemporary bone cements are essentially glassy PMMA which are inherently brittle materials [glass-transition temperature (Tg) ∼105 °C]. This is a major shortcoming in this application. To toughen bone cements, the research group from the Kennedy school at the University of Akron developed a PMMA–PIB molecular composite by simultaneously copolymerizing or crosslinking methyl methacrylate with methacrylate-tri-telechelic PIB (Scheme 7).36

Scheme 7.

Preparation of PIB–PMMA networks.

These novel cements combined the state-of-the-art bone material (PMMA) with carefully tailored PIB chains in a precisely designed two-phase molecular composite that showed very good impact- and fatigue-resistance. It was found that bone cements containing 9.2% PIB (Mn = 18,000 g/mol) exhibited superior overall properties in comparison with commercial bone cements.37

PIB–Cyanoacrylate (CA) for Intervertebral Disk Replacement

The chemistry of CAs combined with the advantage of viscoelastic properties of PIB provided the background for the design of a new synthetic material for intervertebral disk replacement. Kennedy's group has synthesized linear and three-arm star CA-terminated prepolymers (CA–PIBs) with various molecular weights (Scheme 8). The liquid, injectable CA–PIBs acted like “superglue” and polymerized upon exposure to moisture.24

Scheme 8.

Preparation of PIB-based CA networks.

To render it nontoxic and biocompatible, sufficiently high molecular weight CA–PIB needed to be used to protect the CA groups in the disk prosthesis from enzymatic attack by the flexible hydrophobic PIB coils. Indeed, CA–PIB of Mn = 1500–2000 g/mol was synthesized and readily started the polymerization upon contact with moisture.7 For the clinical applications, the CA–PIBs would be syringed by the surgeon into the cavity of the annulus fibrosa so that the CA–PIB would be polymerized in situ to give rise to a synthetic disk.

PIB-Based Amphiphilic Networks (APNs) for Immunoisolatory Membranes

Kennedy's group has developed PIB-based APNs containing water-soluble acrylates or PEG as hydrophilic strands, and PIB or polydimethylsiloxane as hydrophobic components (Fig. 1).22, 24 By selecting proper hydrophilic monomers and composition, PIB-based APNs have shown interesting properties for biomedical applications, for example, control of swellability by composition (see Fig. 1), pH-response of swelling, controlled-release of drugs, and bio- and blood compatibility.38–41 APNs with equal hydrophilic/hydrophobic content (50:50) were found to be biocompatible in vivo. The materials were fashioned into unique rubbery, slippery, robust, sterilizable, optically transparent semipermeable membranes with controlled pore dimensions (1.5–3.6 nm). These membranes were shown to adsorb less fibrinogen, albumin, and Hageman factor (factor XII) than glass, polyethylene, or silicone rubber. Reduced protein adsorption and cell adhesion also indicated hemocompatibility at blood-contacting surfaces. PIB–poly(dimethyl acrylamide) hydrogels were used to build water-swollen tubules that showed sufficient burst strength for implantation and immunoisolatory applications.38 These membranes were then used to envelope pig β cells. The pores of the membrane allowed the in-diffusion of glucose and nutrients, and out-diffusion of insulin and wastes, but prevented the entry of immunoproteins (immunoglobulin G). Implanted subcutaneously in rat, this revolutionary medical device corrected severe hyperglycemia.41

PIB-Based Thermoplastic Elastomers for Vascular Grafts

The breakthrough of producing high-molecular-weight PIB with narrow polydispersity by controlled polymerization42, 43 allowed the synthesis of linear and star-branched PIB–polystyrene (PS) block copolymers44 (see Fig. 1). These polymers with 10–40 wt % PS display thermoplastic elastomeric (TPE) properties, that is, they behave like crosslinked rubbers at room temperature, whereas melt at temperatures above the Tg of the glassy PS block and therefore can be processed like plastics.45 Because of the nanometer scale of the PS domains, these microphase-separated materials are macroscopically transparent. Transparency and processability are important considerations in medical-device manufacturing. In addition, PIB–PS TPEs have very low permeability, outstanding aging resistance, high damping, and excellent flex-fatigue properties.46, 47 The stability of PIB–PS is demonstrated in Figure 2, in comparison with crosslinked natural rubber; both samples were produced in 1991 and were kept at ambient atmosphere.

Figure 2.

Comparison of the oxidative stability of PIB–PS TPEs with crosslinked natural rubber.

The physical properties of PIB–PS depend on the composition and molecular weight; at Mn = 80,000 g/mol and 25 wt % PS, it is positioned between polyurethanes and silicone rubber, with about five times the tensile strength of the latter, and tensile strength about equivalent to the former when it is well hydrated in the body.48 The phase-separated morphology gives physical strength to the PIB–PS rubber without the need of chemical crosslinking. Figure 3 shows the morphology of a representative sample. In comparison, silicone rubber is produced by chemical crosslinking, as shown in Scheme 9. Thus, first the device needs to be manufactured to the desired shape, then crosslinked.

Figure 3.

AFM images of PIB–PS with 30 wt % PS.

Scheme 9.

Preparation of silicone rubber for medical applications.

PIB–PS raised the interest of biomaterial manufacturers. Biocompatibility and biostability testing results were mostly published in the patent literature. Superior oxidation and acid hydrolysis resistance were demonstrated in vitro. Under the same conditions, polyurethanes were destroyed, and silicon rubber became brittle and retained only 10% of its tensile strength. The excellent biostability of PIB–PS was further confirmed in vivo by 6-month and 2-year implantation/explantation and scanning electron microscopy analysis of a PIB–PS porous membrane attached to an Elgiloy braided wire stent. After explantation, no cracked microfibers were found in the membranes. Also, no encrustation or inflammation was found, but partial-to-complete endothelialization was observed.48, 49 The formation of a monolayer of endothelial cells on the surface of vascular grafts is critical to prevent thrombosis.1 The excellent biostability and biocompatibility of the PIB–PS was explained with the presence of alternating quaternary and secondary carbon atoms in the main chain.48 It was suggested that this structure would not be oxidized in the body, where biodegradation generally involves various combinations of oxidative, acidic hydrolytic, and enzymatic pathways. For instance, in the case of polyether-urethane implants, the main degradation site was found to be in the vicinity of the ether linkage.50

The patent literature suggests that PIB–PS is a potential biomaterial with mechanical properties bridging the range between polyurethane and silicone rubber, whereas much softer than either biomaterial, the need for which is evidenced by the literature. The use of PIB–PS in medical devices may become a reality in the 21st century. CPIB-PS has recently received FDA approval for use as the polymeric coating in a drug-eluting coronary stent.)

PIB-BASED BIOMATERIAL RESEARCH IN THE MACROMOLECULAR ENGINEERING RESEARCH CENTRE (MERC)

Since its foundation of 1996, the MERC group has been interested in developing and testing materials specifically for biomedical applications. We continued investigating structure–property relationships in PIB–PS block copolymers, and developed new carbocationic initiators, and new PIB-based structures.51–64 We consider linear triblock PS-PIB-PS (Fig. 1) the first generation of this new class of TPEs, with multiarm-star PIB–PS being the second generation and arborescent PIB–PS the third generation. Research related to PIB–PS blocks has been reviewed18, 65–68 —here we will highlight our contribution.

One-Step Hydroxyl-Functionalization of PIB by Living Initiation with Epoxides—A New Class of Initiators for Carbocationic Polymerization

We discovered a new class of epoxide initiators for carbocationic polymerization. Scheme 10 illustrates the epoxides used for initiation of IB polymerization. The substituted epoxides were commercially available or simply obtained by the reaction of their corresponding olefins with purified m-chloroperoxybenzoic acid.69 Clearly, the easy synthesis renders these initiators very attractive for carbocationic polymerizations.

Scheme 10.

Chemical structure of epoxides used for the initiation of IB polymerization.

The α-methylstyrene epoxide (MSE)/TiCl4 initiation system has been investigated in detail,69–72 revealing living carbocationic polymerization of IB. Linear Mn versus conversion plots together with first-order kinetics of monomer consumption, yielding narrow-molecular-weight distribution polymers (MWD = 1.05) in the −80 to −60 °C range, demonstrated living conditions. The comprehensive study of MSE-initiated IB polymerization demonstrated the production of PIB carrying one primary hydroxyl head group and one tertiary chloride end group. The presence of primary hydroxyl groups was verified by Fourier transform infrared (FTIR) and 1H NMR spectroscopy. Polymers with hydroxyl functional groups are of great interest in terms of their versatility for further chemical modification. It is important to consider the significant reactivity difference between tertiary and primary alcohols for subsequent chain-extension and functionalization reactions. Primary hydroxyl-functionalized PIBs were first synthesized by Kennedy's group.73 Hydroxyl-functionalization was conducted in three steps: tert-Cl-terminated PIB was obtained by the inifer technique, followed by the quantitatively regioselective dehydrochlorination of PIB-Cl into PIB-CH2C(CH3)[DOUBLE BOND]CH2 which was quantitatively turned into PIB-CH2[BOND]OH by hydroboration/alkaline oxidation. Later, the use of tert-Cl-terminated PIB prepared by living polymerization, and allyl-terminated linear and three-arm star PIBs obtained by one-pot living carbocationic polymerization simplified the reaction sequence.74 In comparison with these rather complicated and costly multistep methods, the attractiveness of the one-pot synthesis of primary hydroxyl-functionalized PIBs via epoxide living initiation is evident.

The mechanism of initiation by epoxide/TiCl4 systems was studied extensively, including in situ FTIR monitoring.69, 70, 75 The competitive reaction mechanism shown in Scheme 11 was proposed, according to which Ieff was defined by the competition between SN1 and SN2 mechanistic pathways. The carbenium ions forming in the SN1 pathway initiated the carbocationic polymerization of IB, whereas the initiator reacting in the SN2 pathway was “lost” in terms of carbocationic activity. Detailed mechanistic studies are still in progress in our laboratories.

Scheme 11.

Proposed mechanism of initiation by epoxide/TiCl4 systems.

The epoxide initiators investigated so far are shown in Scheme 10. In terms of Ieff, MSE and hexaepoxi squalene (HES) were the most effective initiators in conjunction with TiCl4 for living IB polymerization. Ieff in percent represents the fraction of initiator yielding carbocations via the SN1 pathway that will subsequently initiate IB polymerization, and is calculated according to eq 1:

equation image(1)

where Mn and Mn,theo are the measured and theoretically expected molecular weights. Ieff was 35% for MSE and 40% for HES, whereas TMPO-1 and TMPO-2 yielded low Ieff values at 3 and 10%, respectively. The low Ieff likely resulted from the loss of these epoxides due to side reactions during ring opening, such as the formation of dioxanes, aldehydes, and polyethers from competition between SN1 and SN2 pathways as shown in Scheme 11. These side reactions were evidenced by in situ FTIR monitoring.69, 70, 75 Because Ieff is defined as the percentage of the initiator that yields carbocations to induce IB polymerization, the relative proportion of the simultaneous reactions in the epoxy ring-opening reaction was apparently influenced by the epoxide structure. The effect of the chemical structure of styrene-based epoxides (SE) on the initiation was subsequently investigated. SE initiated the living carbocationic polymerization of IB, but with low efficiency (Ieff = 8%). The polymerizations initiated by pM-SE and pM-MSE were not living, although pM-MSE yielded a linear Mn-conversion plot.

Among the epoxides investigated, MSE was the best initiator to prepare primary hydroxyl-functionalized PIB on a pilot scale.54 In the temperature range of −60 to −70 °C, the TiCl4/MSE system yielded well-defined PIB carrying one primary hydroxyl head group and one tertiary chloride end group, low-molecular-weight, and narrow-molecular-weight distribution; 24 larger-scale experiments yielded polymers with Mn = 6100 ± 700; MWD = 1.09 ± 0.04 with quantitative functionality (Fn = 1.09 ± 0.16 and 1.30 ± 0.12 by NMR and FTIR, respectively). At −50 °C, the functionality was lower than 1.

One-Pot Synthesis of Multiarm-Star PIB and PIB–PS Block Copolymers

Based on our studies of the novel epoxide initiators discussed above, we developed synthetic routes for the one-pot synthesis of multiarm-star PIB and PIB–PS (Fig. 1), using HES/TiCl4 as a multifunctional initiating system.53, 76–79 The first-order kinetics of IB consumption and linear relationship between Mn and conversion indicated the absence of detectable termination and chain transfer, and attested to the livingness of the HES/TiCl4-initiated IB polymerization. Investigations revealed that initiator efficiency (both the external and the internal efficiency, Ieff,ext and Ieff,int, the former being the percentage of HES initiating IB polymerization, defined in eq 1, and the latter being the number of arms growing relative to the six initiating sites of each HES molecule) is strongly dependent on reaction conditions such as solvent polarity, the ratio of HES to IB, and the usage of electron pair donors (EDs). Ieff,ext varied between 10 and 90%, whereas Ieff,int varied between 50 to 160%, this latter corresponding to 5–10 PIB arms per HES molecule. Dimethyl acetamide (DMA), a strong ED displayed a dramatic effect on the initiation. In the absence of DMA, Ieff,ext was high (80–90%) and there was a considerable apparent intercept in the Mn-conversion plots. However, in the presence of DMA, this intercept disappeared and Ieff,ext decreased to 45%. This demonstrates that in carbocationic polymerization the effect of EDs is not merely proton trapping. The branched nature of the PIBs was proven directly by 1,1-diphenyl ethylene end capping, and indirectly by kinetic analysis and size exclusion chromatography (SEC). After 1,1-diphenyl ethylene end capping, integration of the aromatic protons in comparison with the aliphatic protons in the 1H NMR spectrum of a PIB with Mn = 21,000 g/mol yielded an average of 5.2 arms per molecule. The absolute molecular weights determined by multi-angle laser light scattering were found to be considerably higher than those obtained by applying the universal calibration principle. Because it is known that the universal calibration principle underestimates the molecular weight of branched polymers, this indirectly demonstrates the branched nature of HES-initiated PIBs.

Multiarm-star PIB–PS block copolymers were produced by adding styrene together with DMA and the proton trap (di-tert-butylpyridine) to the reactor after the living, narrow-molecular-weight distribution PIB [weight-average molecular weight (Mw)/Mn = 1.1–1.2] has reached the desired molecular weight. Multiarm-star block copolymers with 8.9–28.6 wt % PS content and Mn from 164,000 to 609,000 g/mol (Mw/Mn = 1.32–1.88) were synthesized and characterized by 1H NMR and SEC.

Although this synthetic route yielded multiarm-star PIBs and PIB–PS, controlling the initiating step was found critical. At this point, we turned our attention to the synthesis of novel arborescent PIB and PIB–PS (see Fig. 1). We consider this the third generation of PIB-based TPEs. With this, we demonstrated a conceptionally new way to achieve TPE properties in block copolymers.45

Aborescent PIB–PS Block Copolymers by Inimer-Type Living Polymerization

Dendritic polymers have gained much attention because of their unusual properties such as low viscosity and low shear stress compared with their linear counterparts.80–83 Gauthier and Möller84 introduced the term arborescent to describe randomly branched tree-like polymers, which combine the characteristics of dendrimers and hyperbranched polymers.

The approach we developed to successfully produce very high-molecular-weight aborescent PIBs is the one-pot inimer-type living polymerization.85, 86 An inimer is a compound carrying both an initiator and a monomer functionality. In conjunction with TiCl4, 4-(2-hydroxyisopropyl)styrene and 4-(2-methoxyisopropyl)styrene as inimers yielded the desired products. Scheme 12 depicts the formation of aborescent PIB initiated by 4-(2-methoxyisopropyl)styrene/TiCl4. PIBs with Mn up to 8 × 105 g/mol with narrow-molecular-weight distribution (MWD ∼ 1.2) were produced within 15–60 min reaction time. The incorporation of the inimer into the PIB chains was verified by 1H NMR and SEC. The branching frequency was determined by selective destruction of the branching points and taking the ratio of the molecular weights before and after link destruction; PIBs with branching frequencies ranging from 3 to 57 were prepared. The process was also scaled up to produce 400 g PIB per batch in order to obtain sufficient amount of material for meaningful physical characterization. The resulting polymers displayed an interesting balance of properties, summarized in Table 1; higher moduli and much higher elongation than a linear butyl polymer with comparable molecular weight, and loss factor (tan δ) values nearly independent of shear rate. This behavior is similar to crosslinked polymers. The linear and nonlinear viscoelastic properties of these polymers were investigated and revealed distinct differences between them and star-branched PIBs.58, 60, 61

Scheme 12.

Formation of aborescent PIB initiated by 4-(2-methoxyisopropyl)styrene/TiCl4.

Table 1. Tensile Properties of Linear and Aborescent PIBs
SampleYoung's Modulus (MPa)Elongation at Break (%)Tensile Strength at Break (MPa)
PIB10.152000.02
PIB20.484800.13
PIB30.603200.17
PIB40.643200.20
PIB50.572700.15
Linear0.471800.07

Subsequently, aborescent PIB–PS block copolymers with 7–34 wt % PS content were synthesized by the addition of styrene into living aborescent PIB charges with DMA and proton trap. The aborescent PIB–PS block copolymers produced were very soft with hardness of Shore A2 22-41 and showed 5–10 MPa tensile strength with 490–1800% elongation; Table 2 shows an inventory of arborescent PIB–PS block copolymers. The tensile strength values are lower than the 17–24 MPa with 600–800% elongation published for linear PS-PIB-PS triblocks52, 58 but comparable to the tensile strength of commercially produced, test-marketed PS-PIB-PS (σ = 10 MPa, ϵ = 600%).59 Importantly, all of these values exceed the strength of soft tissues.1 AFM studies of the arborescent blocks revealed irregular phase morphologies. Three-dimensional AFM image and section analysis indicated a surface topography of high-lying plateaus or hills and low-lying plateaus or valleys in the block copolymers, which become more prominent during annealing.56, 57

Table 2. Tensile Properties of Arborescent PIB–PS Block Copolymers
SampleHardness (Shore A2)Tensile Strength at Break (MPa)Elongation at Break (%)
  • a

    Commercially tested marketed sample of linear PIB–PS block copolymer.

TPE-1349.2820
TPE-2226.3900
TPE-3245.0920
TPE-4416.0490
TS30a9.4590

Arborescent PIB–PS samples are currently undergoing systematic biocompatibility and biostability studies in our laboratories. Preliminary results of an in vivo urinary tract encrustation study have been published.87, 88 Circular disks with 6-mm diameter, cut from 0.1-mm compression-molded aborescent PIB–PS sheets, and medical-grade silicon rubber (SIL-K™), were implanted into 20 female New Zealand White rabbits. After 14 days of the study period, 18 disks were recovered (two animals died from complications during surgery). Visual observation of the disks showed similar encrustation for (SIL-K™) and PIB–PS; Figure 4 shows a representative example for PIB–PS. Table 3 demonstrates the mean weights and the weight ratios of adsorbed calcium relative to the implanted polymer disk. Based on these results, PIB–PS appeared to encrust to a lesser degree, but the limited sample size did not allow any firm conclusion.

Figure 4.

Images of clear and encrusted PIB–PS disks.

Table 3. Adsorbed Calcium on Polymer Disks during In Vivo Urinary Tract Encrustation
PolymerPIB–PSSilicone SIL-KTM
Sample weight (g)0.388 ± 0.1390.467 ± 0.211
Mean weights of absorbed calcium (g)0.134 ± 0.0370.166 ± 0.042

The results of the encrustation studies underlined the utmost importance of surface properties in the application of polymers for medium- and long-term implants. Our studies have been extended to the investigation of surface properties in polymers.

Investigation of Polymer Surface Properties for Biomedical Application

The importance of surface properties in biomedical applications is well known,89 but it is unclear what should be the “ideal” surface. Recent publications pointed to positive host response of implants with rough surfaces. This effect was also highlighted in our investigations into the use of PIB–PS as a potential implant material in the urinary tract.87, 88 AFM investigations revealed that the attachment of a common uropathogenic species, Escherichia coli 67, in vitro in the presence of saline was greatly reduced on first-generation PIB–PS as compared with SIL-K™.87 When both PIB–PS and medical-grade silicone were coated with a recombinant form of a 29-kDa protein (p29), further significant reduction of uropathogen attachment was observed on both, with about 90% reduction on PIB–PS in comparison with 60% on SIL-K™. p29, isolated from the probiotic strain Lactobacillus fermentum RC-14 and expressed in E. coli, was shown to prevent inflammatory sepsis by Staphylococcus aureus without having the pathogens eradicated. The protein coat was shown to be present for at least 96 h. The difference between PIB–PS and SIL-K™ was explained with the greater surface roughness of PIB–PS.88

To get greater insight into the role of polymer surface properties in biocompatibility, we embarked on fundamental studies of protein adsorption onto polymers.

Protein-Adsorption Studies

We performed our protein-adsorption studies with model PS carrying thymine functionalities.90 The polymers were prepared via free-radical emulsion copolymerization of styrene and 1-(vinylbenzyl)thymine (1-VBT). XPS analysis showed that the thymine content was much higher on the surface (7–20 mol %) than in the bulk (1–4 mol %) of the PS–VBT polymers. Interestingly, the increased VBT content on the surface barely affected the water contact angle measured on these polymers (see Table 4).

Table 4. Water Contact Angles Measured on PS–VBT Copolymers
Sample IDVBT (mol %)Contact Angle in Degreea
BulkbSurfacecStaticAdvancingReceding
  • a

    Contact Angle Goniometer (Rame-Hart model 100).

  • b

    1H NMR, FTIR, and elemental analysis.

  • c

    XPS.

Glass (reference)28.2 ± 0.839.0 ± 4.6<10
PS89.8 ± 1.395.2 ± 1.678.8 ± 1.6
PS-VBT1-81.267.6190.7 ± 0.694.0 ± 0.978.3 ± 2.8
PS-VBT2-122.1411.2791.5 ± 2.894.7 ± 0.678.2 ± 1.2
PS-VBT4-203.8519.5891.2 ± 2.196.3 ± 2.971.2 ± 2.1

FTIR and adsorption isotherm analysis of bovine serum albumin (BSA) and bovine hemoglobin model proteins demonstrated that while both proteins were physisorbed onto PS through nonspecific hydrophobic interactions, adsorption onto PS–VBT occurred by both physisorption and chemisorption via hydrogen bonding. FTIR analysis also indicated conformational changes in the secondary structure of BSA and bovine hemoglobin adsorbed onto PS, whereas little or no conformational change was seen in the case of adsorption onto PS–VBT. Similar findings earlier91 were explained by suggesting that BSA adsorbs by rearranging its hydrophobic domains toward the PS surface, thus the orientation of the protein molecule changes. The fact that practically no conformational changes were seen with PS–VBT underlines again the importance of surface modification of hydrophobic polymer surfaces for biomedical application. However, the insignificant effect of surface VBT on water contact angle demonstrates that this test alone may not be reliable for biomedical applications. AFM studies, consistent with the isotherm results, also demonstrated monolayer adsorption for both proteins and showed the effect of VBT content on the orientation of BSA on the surface. The AFM results agreed well with theoretically calculated and experimentally obtained adsorption capacities. Interestingly, AFM showed that PS had a smooth surface with almost no features, whereas PS–VBT displayed some surface roughness with a vertical height of ∼1 nm.92

We plan to extend protein-adsorption studies to surface-modified arborescent PIB–PS biomaterials.

Antibiotic Adsorption/Desorption Studies

We also demonstrated the effective adsorption of Ciprofloxacine, shown in Scheme 13, a wide-range antibiotic, to thymine-functionalized PS via hydrogen bonding (manuscript in preparation).

Scheme 13.

Chemical structure of Ciprofloxacine.

Figure 5 shows the high-resolution XPS spectra of the copolymers before and after adsorption of Ciprofloxacine. Figure 5(a) shows C 1s spectrum of a copolymer with 30 mol % thymylmethylstyrene (TMS) units in the bulk and 50 mol % on the surface. (PS–TMS was prepared by copolymerizing styrene with p-chloromethylstyrene, followed by converting the chloromethyl groups into thymine. Detailed synthesis and characterization data will be published elsewhere.) The C[DOUBLE BOND]C, C[BOND]C, and C[BOND]H carbons are assigned to the PS backbone, whereas the C[BOND]N and C[DOUBLE BOND]O carbons represent the thymine functionalities. The C[BOND]O carbons indicate hydrolyzed chloromethyl groups. The O 1s spectrum in Figure 5(b) also shows the C[DOUBLE BOND]O and C[BOND]O groups, whereas the single peak of the N 1s spectrum in Figure 5(c) is assigned to the amide nitrogen in the thymine functionalities. In Figure 5(d), the C[DOUBLE BOND]C, C[BOND]H, and C[BOND]C peaks appeared at slightly higher (0.2 eV) binding energies, indicating nonspecific physisorption of Ciprofloxacine onto non-functionalized PS.93 The C[BOND]N, C[BOND]O, and C[DOUBLE BOND]O peaks shifted by 0.4 eV to lower binding energies, with broadening of the C[BOND]O and C[BOND]N peaks, demonstrating hydrogen bonding of the Ciprofloxacine onto the thymine carbonyls and amides of the polymer.94 The two extra C 1s peaks represent the carbonyl and carboxylic carbons of the adsorbed Ciprofloxacine molecules. The O 1s spectrum [Fig. 5(e)] shows the C[DOUBLE BOND]O and C[BOND]O peaks shifting by 0.4 and 0.2 eV to lower binding energies, respectively, due to hydrogen bonding. The N 1s spectrum in Figure 5(f) shows three peaks. The largest represents the amide peak of the thymine units that did not adsorb Ciprofloxacine. The small peak on the left was assigned to the amine nitrogen of hydrogen-bonded Ciprofloxacine, whereas the other small peak on the right represents the hydrogen-bonded amide nitrogen of the thymine units. This clearly indicates that Ciprofloxacine can hydrogen bond to thymine on two sites: the carbonyl groups and to a small extent on the amide groups.

Figure 5.

High-resolution XPS spectra of PS–TMS (thymylmethylstyrene) copolymers (30 mol % TMS in the bulk and 50 mol % on the surface) before (a–c) and after (d–f) Ciprofloxacine adsorption.

Desorption of the Ciprofloxacine was performed under simulated physiological conditions (pH = 7.4, 37 °C). Close to 90% of the total load was released from the copolymer over a time period of 16 h.

Surface Modification of PIB–PS Biomaterials for Better Biocompatibility

As discussed above, E. coli 67 attachment to p29-coated PIB–PS in vitro was reduced by 90%. However, subsequent exposure of the samples to urine removed the protective coating.88 This clearly indicates the necessity to bond p29 covalently to the surface. The adsorption studies also pointed out that introducing functional groups such as thymine (a nucleic acid base) or amino acids such as alanine onto the surface of the polymer is very advantageous in terms of protein adsorption, and allows the attachment of antibiotics to the surface. We plan to combine the epoxide initiating chemistry and the one-pot inimer copolymerization to introduce hydroxyl functionalities onto the surface of PIB. Using the inimer shown in Scheme 14, OH groups would be introduced to the branching points of arborescent PIBs. Subsequent blocking with PS would yield the required biomaterials. AFM investigations of triblock TPEs with polybutadiene rubbery segments indicated that the rubber phase segregates to the surface, coating the embedded PS phases.95 This must be verified for the PIB–PS TPEs as well, but it is a feasible explanation why boiling in 65% nitric acid did not destroy the PS phases.49 If this were to be proven for PIB–PS, the hydroxyl functionalities of the PIB would then end up on the surface of the material, ready for binding.

Scheme 14.

Chemical structure of the designed epoxy inimer.

Preparatory to the PIB–PS surface modification outlined above, model studies were performed. First, we attached alanine to PIB-OH prepared by epoxide initiation, using the reaction sequence shown in Scheme 15.96N-(tert-butoxycarbonyl) alanine was used to protect the amine group. 1,3-Dicyclohexyl carbodiimide (DCC) and 4-dimethylamino-pyridine (DMAP) were used as catalysts for the esterification reaction. The attachment was demonstrated by NMR and FTIR (see Figs. 6 and 7, respectively). After the attachment, the tert-butoxycarbonyl group was removed by 10–40% trifluoroacetic acid (TFA) in dichloromethane, to yield the alanine-functionalized PIB.

Scheme 15.

Reaction route of the synthesis of PIB-alanine.

Figure 6.

1H NMR spectrum of PIB-O-Ala-tBoc.

Figure 7.

FTIR spectra of PIB-OH, tBoc-Ala, and PIB-O-Ala.

Preliminary results showed that the attachment of polyalanine was also successful (the details will be published shortly). Eventually, we plan the covalent attachment of Ciprofloxacine, p29, and RGD to further improve the biocompatibility of our arborescent PIB–PS biomaterials.

In summary, PIB–PS biomaterials show great promise in biomedical application. Research will also allow us to produce novel surface-modified polymers for biomedical application. For instance, these new biomaterials could be used as drug-carrying polymer coatings on medicated coronary stents (FDA approval has just been granted for such an application), or as silicone rubber replacement.

SYNOPSIS

In the United States, demand for cosmetic breast implants increased dramatically from 32,607 in 1992 to 167,318 in 1999, a 6-fold increase! Demand for post-mastectomy implants also increased dramatically, 29,607 in 1992 to 82,975 in 1999, which is a 2.5-fold increase, despite the controversy surrounding the silicone implants. The major problem with silicone implants is the permeability of the shell. PIB–PS with its exceptional barrier properties could easily be used as an inner liner/inner tube to prevent leakage of the silicon gel. This would solve a major problem for women with implants.

Acknowledgements

The collaboration of Michael Cunningham and Argyrios Margaritis in the surface adsorption studies is greatly appreciated. The contribution of Miroslawa El Fray (Technical University of Szeczin, Poland) and Gabor Kaszas (NewCo., a subsidiary of Bayer Inc., Canada) to the manuscript is greatly appreciated.

Biographical Information

original image

Judit E. Puskas

Dr. Puskas received a Ph.D. in plastics and rubber technology in 1985, and an M.E.Sc. in organic and biochemical engineering in 1977 from the Technical University of Budapest, Hungary. Her advisors were Professors Ferenc Tüdös and Tibor Kelen of Hungary, and Professor Joseph P. Kennedy at the University of Akron, Ohio, USA, in the framework of collaboration between the National Science Foundation of the USA and the Hungarian Academy of Sciences. She started her academic career in 1996. Before that she was involved in polymer research and development in the microelectronic, paint, and rubber industries. Her present interests include living carbocationic polymerization, polymerization mechanisms and kinetics, and polymer structure/property relationships, with focus on the biomedical application of polymers and the combination of biopolymers and synthetic polymers. She is a regional Editor and member of the Advisory Board, of the European Polymer Journal, and member of the IUPAC Working Party IV.2.1 “Structure-property relationships of commercial polymers.” Puskas has been published in more than 200 publications, including technical reports, and is an inventor or co-inventor of 16 U.S. patents and applications. She is the recipient of several awards, including the 1999 PEO (Professional Engineers of Ontario, Canada) Medal in Research & Development, and a 2000 Premier's Research Excellence Award. Puskas held the Bayer/NSERC (Natural Science and Engineering Research Council of Canada) Industrial Research Chair in Elastomer technology from 1998 to 2003, and was the Director of the Macromolecular Engineering Research Centre at the University of Western Ontario in Canada. Starting in August of 2004, she will join the Faculty in the Department of Polymer Science in the College of Polymer Science and Polymer Engineering of the University of Akron, where she will hold the Bayer Chair.

Ancillary