Biomacromolecules 2003, 4, 145-148
145
Silicone Elastomer Surface Functionalized with Primary Amines and Subsequently Coupled with Heparin Bjo ¨ rn Olander, Anders Wirse´ n, and Ann-Christine Albertsson* Department of Polymer Technology, Royal Institute of Technology, SE-100 44 Stockholm, Sweden Received August 29, 2002; Revised Manuscript Received October 28, 2002
Primary amines covalently bonded to the surface of poly(dimethylsiloxane) were obtained by hydrosilylation grafting of aminopropyl vinyl ether to Si-H groups formed during argon plasma treatment. The amine groups were derivatized using pentafluorobenzaldehyde and characterized by X-ray photoelectron spectroscopy. The graft yield was about 3% grafted molecules within the depth of the analysis. The terminal aldehyde groups of diazotized heparin was also coupled to the primary amines. This led to a silicone elastomer with covalently bonded heparin which was expected to be hydrolytically stable. This method of bonding primary amines to the surface of silicone elastomers and the subsequent coupling of aldehyde-containing molecules is a promising way of obtaining novel biomaterials. Introduction Silicone elastomers have been used as biomaterials since the 1950s with clinical applications ranging from breast implants and intraocular lenses to catheters. The commercial success of these materials is due to the suitable combination of bulk and surface properties which they offer, their convenient processability, and the acceptable costs of the products. However, there is still an ongoing debate over the biocompatibility of the biomaterials used today.1 This originates in the lack of consensus as to how the term “biocompatibility” should be defined. The vast majority of implants are eventually encapsulated and thus sealed off from the host. This can lead to implications regarding their longterm stability. For example, the most commonly reported failure of silicone breast implants is capsular contraction.2 Regardless of the definition adopted, the key to developing improved biomaterials probably lies in the ability to tailor surfaces that can interact with the host in a specific and desired way. Our goal is to tailor polymer surfaces to achieve stable structures with suitable properties as biomaterials. We have shown that graft polymerization of acrylamide initiated by electron beam irradiation can be Hofmann-degraded to a structure with a surface containing primary amines available for further coupling reactions. This has been achieved on the surface of the two degradable materials poly(-caprolactone)3 and poly(1,5-dioxepan-2-one),4 as well as on linear low-densisty polyethylene (LLDPE).5 To create a material with the desired biological performance, the outermost layer of the surface should be recognized and accepted by the host. Heparin is a polysaccharide of complex structure that is naturally present in mammalian tissue. Diazotization of heparin introduces terminal aldehydes which can be coupled to primary amino * To whom correspondence may be addressed. E-mail: polymer.kth.se.
aila@
groups present at surfaces.6 This leads to a nonthrombogenic surface with properties suitable for biomaterials.7 We have demonstrated that the primary amino groups obtained at the surface of LLDPE can be heparin-functionalized in a similar way. This resulted in a bioactive surface with promising properties according to in vitro experiments.8 Our current interest is in modifying the surfaces of silicone elastomers. However, the methods we have used in the earlier modifications are not expected to be successful for silicone elastomers. First, the electron beam irradiation may affect the elasticity of the material due to the formation of crosslinks by the combination of radicals in the bulk phase of the material, and second, the radicals introduced for monomer initiation can be prematurely terminated due to the pronounced chain mobility in the flexible siloxane main chain. The strategy we have chosen is to treat the silicone elastomer with argon plasma. This introduces Si-H groups at the outermost surface layer.9 The grafted structure should be linked to the material via hydrolytically stable bonds. The introducing of peroxides at the silicone elastomer surface may lead to Si-O-C bonds which will eventually be hydrolyzed in the presence of water. We therefore react the introduced Si-H groups with molecules containing double bonds. This results in hydrolytically stable silicon-carbon bonds. We have previously shown that poly(dimethylsiloxane) (PDMS) can be surface modified by argon plasma treatment and subsequent hydrosilylation grafting of allyl tetrafluoroethyl ether (ATFEE).10 We determined the influence of the plasma parameters power, pressure, and treatment time on the graft yield of ATFEE. The main reason for using a fluorinated molecule in that study was its suitability for X-ray photoelectron spectroscopy (XPS) analysis, although fluorinated surfaces are also potentially interesting as biomaterial coatings. In this article we describe the introduction of primary amino groups by reacting aminopropyl vinyl ether (APVE) to the Si-H groups formed at the surface during argon
10.1021/bm025654+ CCC: $25.00 © 2003 American Chemical Society Published on Web 12/06/2002
146
Biomacromolecules, Vol. 4, No. 1, 2003
plasma treatment. The amino groups were derivatized with pentafluorobenzaldehyde (PFB) to obtain increased sensitivity in the XPS analysis of the surface. Finally, we obtained a heparin surface by reacting the primary amines with diazotized heparin. Experimental Section Materials. The silicone elastomers were prepared by mixing 9 g of (30-35%) methylhydro-(65-70%) dimethylsiloxane copolymer (PS123), 205 g of vinyldimethylterminated poly(dimethylsiloxane) (PS442), and 50 µL of platinum divinyltetramethyldisiloxane catalyst (PC072), all purchased from United Chemical Technologies, USA. This resulted in a SiH:vinyl ratio of approximately 2:1.11 The mixture was immediately cast in glass Petri dishes to a thickness of 3 mm and allowed to cure at room temperature for 7 days. The sheet obtained was Soxhlet extracted in analytical reagent grade hexane from Labscan Ltd, Ireland, for 7 days followed by slow deswelling and drying in air and vacuum. Plasma Treatment. The plasma treatment was carried out at 2.45 GHz in a microwave plasma system, model V15-G, from Plasma-Finish GmbH, Germany, connected to a Pfeiffer DUO 035 D C vacuum pump. The plasma treatment chamber was subjected to a cleaning process involving at least 20 min of oxygen plasma treatment before use. The samples were placed on a glass sample holder in the center of the chamber. Each sample was subjected to at least three degassing cycles where the pressure was decreased below 1 Pa followed by an argon flush resulting in a pressure exceeding 100 Pa. The plasma treatment was carried out at a predetermined power at a preset pressure and treatment time. In an attempt to terminate radicals after the plasma treatment, the chamber was immediately flushed with hydrogen for 300 s at a pressure of about 80 Pa. The chamber was ventilated to ambient pressure using nitrogen. The gases were purchased from AGA Gas AB, Sweden, and had a purity of at least 99.996%. Grafting of Aminopropyl Vinyl Ether (APVE). Immediately after the plasma treatment the silicone elastomers were put in a solution prepared by 0.5 mL of distilled aminopropyl vinyl ether from BASF, 19.5 mL of distilled analytical reagent grade toluene (Labscan), and 50 µL of PC072 catalyst. After 7 days with continuous stirring, the specimens were repeatedy washed with analytical reagent grade acetone (Labscan) and dried in a vacuum before being submersed in Milli-Q ultrapure water for 5 days followed by drying in a vacuum. Derivatization with Pentafluorobenzaldehyde (PFB). The specimens grafted with APVE were put in a hexane solution containing 2 vol % pentafluorobenzaldehyde (PFB), Acros, New Jersey, USA. After 48 h, the specimens were removed from the solution and Soxhlet extracted in hexane for 72 h, before deswelling and drying in a vacuum. Coupling of Heparin. Heparin with an activity of 185 IU/mg was purchased from Pharmacia, Stockholm, Sweden. The preparation of diazotized heparin led to terminal aldehyde groups and subsequent coupling to the surface
Olander et al.
amino groups.7 The deazotation was performed by dissolving 1 g of heparin in 300 mL of Milli-Q water at 0 °C followed by addition of 10 mg of sodium nitrite. The pH was adjusted to 2.7 using 1 M hydrochloric acid, and the mixture was stirred at 0 °C for 2 h followed addition of 1 M sodium hydroxide to pH 7.0. The reaction mixture was dialyzed against Milli-Q water for 24 h in a membrane and freezedried to constant weight. A water solution containing diazotized heparin (0.2 mg/mL), sodium cyanoborohydride (0.01 mg/mL) ,and 0.15 M sodium chloride at pH 3.4 was reacted with the introduced primary amino groups at the surface for 48 h at 50 °C. The specimens were washed 0.5 M NaCl solution for 24 h followed by repeated washing in Milli-Q water for 24 h before being dried in a vacuum. X-ray Photoelectron Spectroscopy (XPS). An AXIS-HS X-ray photoelectron spectrometer (Kratos Analytical, Manchester, U.K.) was used with a monochromatic Al KR X-ray source operated at 15 kV and 20 mA. The takeoff angle was 90° with respect to the sample surface, and the pass energy was 80 eV. The pressure was below 10-8 Torr during the measurements. Background subtractions were made by drawing a straight line between two suitable points,12 and the sensitivity factors were supplied by the manufacturer. Results and Discussions In the present study we have created a PDMS surface with amino groups linked via hydrolytically stable bonds obtained by hydrosilylation grafting of aminopropyl vinyl ether (APVE). These amino groups have then been for further coupling reactions with the aldehyde groups of pentafluorobenzaldehyde (PFB) as well as diazotized heparin. Hydrosilylation Grafting of Aminopropyl Vinyl Ether. Primary amino groups attached to a surface provide possibilities for the further coupling of interesting molecules using well-established methods. We chose allylpropyl vinyl ether (APVE) for the direct hydrosilylation grafting to the Si-H groups formed during argon plasma treatment of a silicone elastomer. To obtain increased sensitivity during the XPS analysis, the amino groups were derivatized using PFB according to Scheme 1. This way of derivatization of the amino groups originating from the grafted APVE leads to a higher sensitivity for XPS analysis, since the sensitivity factor of fluorine is defined as unity whereas the value for nitrogen is only 0.42. In addition, each PFB molecule contains five fluorine atoms compared to the single nitrogen atom in the original structure. A simple estimation of the increase in sensitivity by this derivatization is thus 1/0.42 due to the differences in sensitivity factor and 5/1 due to the stoichiometrical differences. This gives a combined effect of an almost 12 times higher than expected sensitivity in XPS analysis if complete derivatization is obtained. The derivatization is also expected to reduce the mobility in the surface layer during the analysis, due to the bulkiness of the grafted molecule. The graft yield was determined by measuring the fluorine concentration and relating the value to the expected concentration of a desired structure having this stoichiometric composition.
Biomacromolecules, Vol. 4, No. 1, 2003 147
Functionalized Elastomers for Biocompatibility
Scheme 1. Hydrosilylation Grafting of Aminopropyl Vinyl Ether to Si-H Groups Formed by Argon Plasma Treatment of PDMS and Subsequent Derivatization Using Pentafluorobenzaldehyde
Scheme 2. Hydrosilylation Grafting of Aminopropyl Vinyl Ether (APVE) to Si-H Groups Formed by Argon Plasma Treatment of Silicone Elastomer and Subsequent Coupling of Heparin
Table 1. Elemental Compositions of PDMS Grafted with APVE and Derivatized by PFB According to XPSa power (W)
time (s)
pressure (Pa)
50 50 50 50 150 150 150 150 100 100 100
10 10 30 30 10 10 30 30 20 20 20
20 60 20 60 20 60 20 60 40 40 40
a
C (%)
Si (%)
O (%)
F (%)
graft yield
52.7 51.9 46.9 49.5 52.6 49.5 49.5 52.0 49.2 49.1 48.9 53.7
23.0 23.6 26.0 23.5 22.9 24.9 23.7 21.8 24.5 24.3 25.2 23.2
24.1 23.9 26.9 26.1 23.6 25.1 25.7 24.6 25.9 26.2 25.6 22.9
0.3 0.6 0.2 0.9 0.9 0.6 1.1 1.6 0.5 0.4 0.3 0.2
1.2 2.5 0.8 3.9 3.8 2.6 4.8 7.0 2.2 1.9 1.4 0.8
The graft yield is calculated from the expected theoretical composition.
We assume that the desired product has the structure SiO(CH3)(CH2CH2OCH2CH2CH2NdCHC6F5). Since hydrogen cannot be detected by XPS, the structure can be summarized as SiC13O2NF5, i.e. a theoretical fluorine content Ftheoretical of 5/22. The graft yield as a percentage of the surface consisting of this structure is expressed by G (%) ) Fdetected (%)/Ftheoretical (%)
(1)
where Fdetected is the measured fluorine content. The elemental concentrations and graft yields of the specimens exposed to different plasma conditions are presented in Table 1. The reference specimen was not exposed to plasma but treated in the same way as the others. It has a minor amount of fluorine at the surface corresponding to a graft yield of 0.8%. This can be explained either as residual traces of unreacted species despite the extensive extractions or as being due to reaction with Si-H groups present in the material before plasma treatment. The three replicates of the PFBgrafted specimens exposed to argon plasma at 100 W, 20 s, and 40 Pa had graft yields between 1.4 and 2.2%. The average graft yield under the investigated conditions was
about 2.9%, which is significantly higher than the 0.8% found on the sample not exposed to plasma if the scatter of the replicate results is considered. The concentration of nitrogen was below the quantifiable level due to the low concentration and the relatively low sensitivity factor of the N(1s) in XPS. Heparin Coupling to Grafted Amino Groups. The derivatization of APVE-functionalized PDMS by PFB showed that a coupling between aldehyde groups and amino groups at the surface is possible. We therefore diazotized heparin to introduce terminal aldehyde groups. The route for heparin grafting to the primary amino groups is shown in Scheme 2. The heparin coupling follows the same reaction pathway as the PFB coupling except that the double bond formed is reduced using NaCNBH3 during the coupling reaction, to obtain a bond that is not susceptible to the equilibrium hydrolysis as shown in Scheme 3. The XPS data for the silicone elastomer grafted with APVE and subsequently coupled with heparin is presented in Table 2. The sulfur is difficult to quantify by XPS when only small amounts are present, due to the pronounced slope in the S(2p) region. To quantify the sulfur content at the surface of heparinized polyacrylamide-grafted linear low-density polyethylene (LLDPE),8 radicals were introduced by electron beam irradiation and used to initiate a radical polymerization of the acrylamide. The amide groups were Hofmanndegraded to amino groups that were used for subsequent coupling of heparin in a comblike configuration. This resulted in a relatively thick and dense layer of heparin with a sulfur surface concentration of about 2% according to XPS. In the present case, each anchoring site at the surface consists of a single amino group available for heparin coupling instead of a chain with a large number of pendant amino groups. We therefore expect a thinner and less dense heparin coverage of the surface and thus considerably lower sulfur concentrations that are difficult to quantify under the present analysis conditions. However, the changes in concentration
148
Biomacromolecules, Vol. 4, No. 1, 2003
Olander et al.
Scheme 3. Reduction of Double Bond by Sodium Cyanoborohydride to Form a Hydrolytically Stable Bond
Table 2. Elemental Compositions of PDMS Grafted with APVE and Coupled with Heparin According to XPS pressure (Pa)
power (W)
time (s)
20 20 60 20 60 20 60 40 40
50 150 150 50 50 150 150 100 100
10 10 10 30 30 30 30 20 20
a
C (%)
Si (%)
O (%)
N (%)
S (%)
54.5 57.0 56.2 54.1 53.5 52.6 53.6 55.9 53.1 55.8
20.6 18.2 18.3 20.7 21.6 20.9 21.3 19.1 21.3 20.5
24.2 23.4 24.3 24.1 24.3 26.0 24.4 23.8 24.6 22.8
0.7 1.5 1.1 1.2 0.6 0.6 0.7 1.2 1.0 0.9
n.d.a n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d.
Even though the sulfur could not be quantified in the relatively low concentrations expected from this grafting procedure, the combined indications of the elemental composition according to XPS support the presence of heparin at the surface. We therefore conclude that the introduction of primary amines followed by heparin coupling is successful, although XPS has limitations when such relatively thin layers are analyzed. Future investigations focused on the biological activity of related surfaces may provide additional insight into the in vitro performance of these surfaces, comparable with our heparin-functionalized LLDPE.8 Conclusions
n.d. ) not detected.
Figure 1. Data from Table 2 plotted as carbon concentration against silicon concentration. The least-squares-fitted line is presented as a guide for the eye.
Primary amino groups were introduced onto the surface of a silicone elastomer by argon plasma treatment and subsequent hydrosilylation grafting of aminopropyl vinyl ether (APVE). This led to a covalent silicon-carbon bond with the hydrolytic stability that is a necessary requirement for a stable biomaterial surface. The amines were derivatized using pentafluorobenzaldehyde to increase the sensitivity for XPS. The graft yield in terms of molecular percentage within the depth of analysis was about 3%, which should be sufficient for full surface coverage of larger molecules. Heparin surfaces were obtained by coupling diazotized heparin to the amino groups. The methods for coupling molecules to primary amines are well established, and they therefore open up possibilities for achieving other relevant surface structures. Although there are some limitations related to the XPS quantification of these relatively thin layers, we conclude that the heparin is covalently bonded to the silicone surface via hydrolytically stable silicon-carbon bonds. The method we describe is a promising way of obtaining biomaterials based on silicone elastomers with interesting surface functionalities. References and Notes
Figure 2. Data from Table 2 plotted as nitrogen concentration against silicon concentration. The least-squares-fitted line is presented as a guide for the eye.
of the other elements support the coupling of heparin. First, the nitrogen concentration that could not be detected from the PFB-coupled surfaces is on average almost 1% in this case. This nitrogen probably originates from heparin, and second, the average silicon concentration is about 20%, which is lower than that for untreated PDMS, plasma-treated PDMS, and the other grafted and derivatized surfaces. This suggests that the surface is covered with a layer as is expected in a heparin-coupled surface. There also appears to be a correlation between a low silicon concentration and high concentrations of nitrogen and carbon as shown in Figures 1 and 2, respectively.
(1) Castner, D. G.; Ratner, B. D. Surf. Sci. 2002, 500 (1-3), 28-60. (2) Gerszten, P. C. Biomaterials 1999, 20 (11), 1063-1069. (3) Ohrlander, M.; Lindberg, T.; Wirse´n, A.; Albertsson, A.-C. J. Polym. Sci., Part A: Polym. Chem. 1999, 37 (11), 1651-1657. (4) Ohrlander, M.; Palmgren, R.; Wirse´n, A.; Albertsson, A.-C. J. Polym. Sci., Part A: Polym. Chem. 1999, 37 (11), 1659-1663. (5) Wirse´n, A.; Albertsson, A.-C. J. Polym. Sci., Part A: Polym. Chem. 1995, 33 (12), 2039-2047. (6) Hoffman, J.; Larm, O.; Scholander, E. Carbohydr. Res. 1983, 117, 328-331. (7) Larm, O.; Larsson, R.; Olsson, P. Biomater., Med. DeVices, Artif. Organs 1983, 11 (2-3), 161-73. (8) Wirse´n, A.; Ohrlander, M.; Albertsson A.-C. Biomaterials 1996, 17, 7 (19), 1881-1889. (9) Gaboury, S. R.; Urban, M. W. Polymer 1992, 33 (23), 5085-5089. (10) Olander, B.; Wirse´n, A.; Albertsson, A.-C. Biomacromolecules 2002, 3 (3), 505-510. (11) Wrobel, D. In Organosilicon Chem. II.; Auner, N., Weis, J., Eds.; VCH: Weinheim, Germany, 1996; pp 633-648. (12) Seah, M. P. In Practical surface analysis; Briggs, D., Seah M. P., Eds.; John Wiley & Sons Ltd.: England, 1990; p 233.
BM025654+
