Natural zwitterionic organosulfurs as surface ligands for antifouling ...

3 downloads 0 Views 785KB Size Report
1J. E. Raynor, J. R. Capadona, D. M. Collard, T. A. Petrie, and A. J.. Garcia, Biointerphases 4, FA3 (2009). 2R. Vasita, K. Shanmugam, and D. S. Katti, Curr. Top.
Natural zwitterionic organosulfurs as surface ligands for antifouling and responsive properties Chun-Jen Huang,a) Lin-Chuan Wang, and Chia-Yu Liu Graduate Institute of Biomedical Engineering, National Central University, Jhong-Li, Taoyuan 320, Taiwan

Anthony S. T. Chiang Chemical and Materials Engineering Department, National Central University, Jhong-Li, Taoyuan 320, Taiwan

Ying-Chih Changb) Genomics Research Center, Academia Sinica, 128, Sec. 2 Academic Road, Taipei 115, Taiwan

(Received 22 January 2014; accepted 12 March 2014; published 25 March 2014) Natural sulfur-containing zwitterionic compounds, L-cysteine (Cys), L-methionine, and glutathionine (GSH), have been employed as surface ligands to prevent protein nonspecific adsorption on planar substrates. These organosulfur compounds form self-assembled monolayers (SAMs) on gold substrates by gold–sulfur interaction. The chemical elements of SAMs were confirmed using x-ray photoelectron spectroscopy. The surface wetting tests for SAMs show that films prepared from Cys and GSH exhibited super-hydrophilicity (contact angles of h ¼ 5 ) due to their high coverage and strong hydration via ionic solvation and formation of hydrogen bonding. Quartz crystal microbalance with dissipation sensor was used to quantitatively and qualitatively monitor the adsorption of bovine serum albumin (BSA) from buffer onto these SAMs. It was found that the GSH film enables the resistance of BSA adsorption to the best extent at a physiological pH. Moreover, the surface charges of modified substrates were modulated by varying the pH value to control BSA adsorption. The effect of electrostatic repulsion on the antifouling behavior becomes prominent at a pH where the protein and the surface carry same charges. Consolidating the BSA adsorption measurements at different pH values, the antifouling properties of GSH-modified Au should be attributed to prevention of entropy gain and enthalpy loss, making BSA adsorption energetically unfavorable. It is believed that the surface modification with natural organosulfur ligands holds great potential in improving the biocompatibility of medical devices and in offering intelligent C 2014 American Vacuum Society. biointerfaces in response to environmental stimuli. V [http://dx.doi.org/10.1116/1.4869300] I. INTRODUCTION Surface modification in biomedical devices is of tremendous importance in determining their biocompatibility and applicability. Enormous efforts were devoted to developing toolkits for engineering surfaces with an attempt to meet desirable interfacial properties, such as roughness, wettability, free energy, and chemical functionality.1–3 Despite that, some critical issues associated with surface chemistry such as blood clotting, bacterial infection, and thrombosis remain troublesome in clinic practices. In order to tackle these problems, the self-assembled monolayers (SAMs) at interfaces have attracted substantial attentions due to their unique properties, such as control on the molecular level, ordered structure, high packing, the ease of preparation, and functionalization as so on.4 The most studied and best understood system among ligands is the alkanethiol at Au(111) surfaces, which takes advantage of the environmental inertness of gold substrate, the high affinity of the sulfur–gold bonding, and the formation of ordered structure by the van der Waals forces between long carbon chains.4–6 One example in the a)

Electronic mail: [email protected] Electronic mail: [email protected]

b)

029010-1 Biointerphases 9(2), June 2014

wide spectrum of its biological applications is a tailor surface with SAMs tethered with various functional head groups for manipulating the adhesion of proteins, cells, bacteria, and even large marine organisms.7,8 Biocompatibility of biomaterials relies strongly on their surface properties. The bioinert or antifouling materials have been pursued to avoid adverse biological effects and to allow the full exploitation of medical devices. Poly(ethylene oxide) (PEO) is extensively applied for the preparation of fouling resistant surfaces. However, its limited applications in environments with high temperature and high ionic strength have been discussed.9 Zwitterionic materials have emerged as a good alternative due to its excellent antifouling property and stability in complex media such as 100% human blood.3,10–13 Zwitterionic polymers carrying ionic polar pendants of sulfobetaine (SB), phosphorylcholine (PC), or carboxybetaine (CB) have been investigated and shown a great reduction in nonspecific adsorption from complex media.3,10 For example, Yang et al. demonstrated a glucose sensor coated with crosslinked CB hydrogel having extraordinary performance in whole blood and long-term stability for up to 42 days.13 The effectiveness of zwitterionic materials comes mainly from their strong hydration and charge balancing ability, leading to a reduction in the entropy gain and

1934-8630/2014/9(2)/029010/6/$30.00

C 2014 American Vacuum Society V

029010-1

029010-2 Huang et al.: Natural zwitterionic organosulfurs as surface ligands for antifouling and responsive properties

the enthalpy loss against the protein adsorption. Whitesides’ group carried out some pioneer tests for zwitterionic SAMs, with a series of charged thiols in contact with protein solutions in a surface plasmon resonance (SPR) sensor.14 This study showed the advantages of zwitterionic SAMs over poly(ethylene) glycol-based SAMs in terms of stability to oxidation, and sensitivity to pH and ionic strength. In this study, a systematical investigation has been conducted on the SAMs of natural zwitterionic organosulfur compounds, i.e., L-cysteine (Cys), L-methionine (Met), and glutathionine (GSH) on gold surfaces, particularly on the interfacial physicochemical properties and the resistance to fouling (Scheme 1). Among these compounds, Cys is the most popular surface ligand for the preparation of biofunctional surfaces, as a model system for the investigation of organic–inorganic interfaces.15–24 A variety of techniques, including x-ray photon electron spectroscopy (XPS), ion mass spectroscopy, scanning tunneling microscopy, surfaceenhanced Raman scattering, and surface-enhanced second harmonic generation, have been applied to scrutinize Cys on metals at a molecular level.15–17,22–24 The ability of Cysmodified surfaces to repel protein adsorption was also well documented.15,16,19 Comparing to the intensive studies of Cys, the other natural organosulfur compounds, i.e., Met and GSH, are somewhat less reported.16,25 In this study, a comparative investigation of natural organosulfur molecules anchored on substrates was carried out to elucidate their correlation between surface characteristics and antifouling properties. XPS was utilized to identify the elemental composition of SAMs. The surface wetting was measured using a contact angle goniometer. The surface zeta potentials were measured by an electrokinetic analyzer in order to reveal the ionization of carboxyl and amine groups upon the pH changes. For the protein adsorption, bovine serum albumin (BSA), the most abundant plasma protein, was used to test the antifouling properties of modified surfaces with a quartz crystal microbalance with dissipation (QCM-D) sensor. We will demonstrate that the natural organosulfur molecules are excellent candidates for the preparation of biocompatible and responsive biointerfaces, thus opens up possibility for potential applications in nanomaterials and medical

029010-2

devices. Meanwhile, the fundamental understanding of protein–surface interactions gained through this study will also facilitate the development of antifouling surface coatings. II. EXPERIMENT A. Materials

The chemicals including reduced L-cysteine, L-methionine, glutathionine, absolute ethanol, sodium hydroxide, and BSA were purchased from Sigma-Aldrich (St. Louis, MO). Potassium hydrogen phthalate, potassium dihydrogen phosphate, sodium tetraborate, sodium chloride, hydrochloric acid, and sodium dodecyl sulfate (SDS) were obtained from Acros Organics (NJ). B. Preparation of SAMs by natural organosulfur compounds

A gold substrate was prepared by deposition onto a 20 mm  20 mm glass slide in a high-vacuum e-beam evaporator. A 5 nm thick chromium layer was first deposited as adhesive layer before the addition of a 50 nm thick layer of high purity gold. The gold substrates were cleaned by thoroughly washing with 1% SDS detergent, absolute ethanol, and then, dried in a stream of nitrogen. The clean substrates were exposed to O2 plasma twice using a plasma cleaner (PDC-32G, Harrick Plasma, Ithaca, NY) for 20 min each in order to remove final trace of organic contaminations from the surface. The substrates were immersed in an aqueous organosulfur solution of 1 mM for 15 min at a temperature of 50  C. They were washed with copious deionized water afterward and dried in a stream of nitrogen and stocked in the refrigerator (4  C) for further measurements. In this study, the time and temperature for the SAM formation are critical in terms of film stability. Therefore, the details of experiments and discussion are given in the supplementary material.26 C. Contact angle measurements

A contact angle goniometer (FDS-OCA15 plus, Dataphysics, Germany) was used to measure static water contact angles at solid–liquid interfaces. The droplets used were 3 ll with a microsyringe, and the measurements were performed at least three times at random locations on each sample. D. XPS for element analysis on surfaces

SCHEME 1. Natural organosulfur compounds used in this study and the surface charges of cysteine SAM in response to the pH of solutions. Biointerphases, Vol. 9, No. 2, June 2014

The chemical element spectra were detected by a PHI 5000 VersaProbe system (ULVAC-PHI, Chigasaki, Japan) with a microfocused, monochromatic Al KR x-ray (25 W, 100 lm). The takeoff angle (with respect to the surface) of the photoelectron was set at 45 . The pressure of the system is below 108 Pa using oil-less ultrahigh vacuum pumping systems. A dual beam charge neutralizer (Arþ gun and flooding electron beam) was employed to compensate for the charge-up effect. Spectra were collected with the pass energy set to 58.7 eV, while the binding energy measured was

029010-3 Huang et al.: Natural zwitterionic organosulfurs as surface ligands for antifouling and responsive properties

normalized against the Au 4f7/2 peak at 84.1 eV. The typical data acquisition time was 30 min. The ratio of peak intensity was converted to atomic percentage using the sensitivity factors built into MULTIPAK software package. E. Zeta potential measurements

The accumulated surface charges form the electrochemical double-layer, the immobile Stern layer and the mobile layer, that are separated by a plane of shear. The potential at the interface between the immobile and the mobile layer is known as the electrokinetic or zeta potential (f). The f potential of the solid substrate surface was detected by the streaming current of planar interfaces using a cell with a height-adjustable channel in an electrokinetic analyzer (SurPass, Anton Paar GmbH, Graz, Austria). Two 20 mm  10 mm planar samples with the surfaces of interest facing each other were aligned in parallel with a 100 m spacer to form a microchannel. Buffered solutions with pH values from 3 to 10 were prepared from salts of potassium hydrogen phthalate (pH ¼ 3–5), potassium dihydrogen phosphate (pH ¼ 6–8), and sodium tetraborate (pH ¼ 9–10) at concentrations of 0.05, 0.05, and 0.023 M, respectively. To increase the ionic strength, 0.15 M NaCl was added to the buffers as electrolyte. The pH of solutions was titrated by adding NaOH (0.1 M) or HCl (0.1 M). The buffered solution was pumped through the microchannel at 0–300 millibars with a syringe pump, and the current across the channel was measured using Ag/AgCl electrodes. The sample was rinsed thoroughly with the electrolyte between measurements so that the acid-base reaction between the surface functional group and the electrolyte reached equilibrium at a given pH. The results of the streaming current measurement were converted to f potentials using the Helmholtz–Smoluchowski equation and the Fairbrother–Mastin approach. Each f potential at a given pH value represents an average of at least three individual measurements. F. QCM-D for BSA fouling tests

The Au(111)-covered QCM crystal chips (AT-cut quartz crystals, f0 ¼ 5 MHz) (Q-Sense AB, Gothenburg, Sweden) were cleaned in 1% SDS detergent, followed by rinsing with deionized water, drying with nitrogen, and exposing to oxygen plasma irradiation for 10 min. Before the measurement, the chamber was rinsed with phosphate buffered saline (PBS) and temperature-stabilized at 25  C. A 1 mg ml1 BSA solution in PBS was brought in contact with the sensor chip at a flow rate of 1 ml min1 for 10 min, followed by rinsing with pure PBS twice. All measurements were recorded at the third overtone (15 MHz); the data shown here were normalized to fundamental frequency (5 MHz) by dividing the overtone number. Since BSA is a globular and relative rigid molecule, the increased mass on the chip is well related to the change in frequency of the oscillating crystal through Sauerbrey relationship DmSauerbrey ¼ ð CQCM •D f Þ=n; Biointerphases, Vol. 9, No. 2, June 2014

(1)

029010-3

where DmSauerbrey represents mass adsorbed on the quartz sensor, Df is resonance frequency, CQCM is the mass-sensitivity constant (¼17.7 ng cm2 Hz1 at f ¼ 5 MHz), and n is the overtone number (¼1, 3, 5, and 7).27–29 III. RESULTS AND DISCUSSION A. Formation and characterization of SAMs

The effects of formation time and temperature on the surface wettability of the natural organosulfur SAM and its long-term stability were first studied to identify the proper procedure. Three different protocols with experimental conditions of 50  C for 15 min, 50  C for 60 min, and 25  C for 60 min were used to prepare the SAMs. There was no significant difference in their initial wetting behavior. However, the contact angle changed dramatically, particularly Cys-Au and GSH-Au cases, after exposure to ambient light for 24 h (Fig. S1).26 This is an indication of the adlayer instability. The 50  C/15 min protocol showed the smallest change, thus the most stable SAM was thereby employed for the following study. As Fig. 1 shows, untreated Au substrate was more hydrophobic in comparison with freshly prepared SAM samples. The Cys-Au and GSH-Au samples were super-hydrophilic, with contact angles of around 5 . The results can be taken as an evidence of high coverage and strong interaction with water molecules by ionic solvation and formation of hydrogen bonding.16 However, the Met-Au sample exhibited low wettability, reflecting the fact that the methyl sulfide group (-SCH3) of Met interferes its affinity to gold and results in the low surface coverage.17 The chemical compositions of adsorbed organosulfur components on surfaces were confirmed using XPS. The representative XPS spectra of GSH-Au for elements C 1s, N 1s, O 1s, and S 2p were shown in supplementary material (Fig. S2).26 The binding energy scale was calibrated with respect to the Fermi level and the Au 4f7/2 peak centered at 84.1 eV. Table I summarizes the atomic concentration of all samples, as well as the experimental N/S element ratios. The N/S ratios were in fair consistence with the expectation. The high contents of C and O, on the other hand, should be

FIG. 1. Contact angle measurements for samples of bare Au, Cys-Au, MetAu, and GSH-Au.

029010-4 Huang et al.: Natural zwitterionic organosulfurs as surface ligands for antifouling and responsive properties

029010-4

TABLE I. Atomic concentrations of different elements detected at the organosulfur-modified surfaces. The experimental ratio of N and S is the area ratio of the N 1s and S 2p (corrected by a sensitivity factor) in the XPS spectra; the theoretical ratio of N and S is calculated by a stoichiometric ratio of SAMs on surfaces. Atomic concentration (%) Sample

C 1s

N 1s

O 1s

S 2p

N/S ratio theoretical

N/S ratio experimental

Cys-Au Met-Au GSH-Au

31.8 30.3 33.4

2.8 1.6 5.2

19.2 20.5 29.9

2.6 1.3 1.8

1 1 3

1.1 1.2 2.9

attributed to surface contaminations in hydrocarbon rich atmosphere and metallic oxide.30,31 Moreover, the incomplete coverage of Met is suggested from the lower surface concentration of N 1s and S 2p comparing to Cys and GSH, echoed with its lower surface wettability measured previously. The surface f potential on planar substrates was evaluated as a function of pH using electrokinetic analyzer. The f potential was determined by the streaming potential and the specific electrical conductivity of the electrolyte solution. Thus, the buffered solutions with distinct pH values from 3 to 10 but at identical ionic strength were prepared as described in Sec. II. As shown in Fig. 2, the f values for the bare gold and Met-Au samples were all negative over the pH range studied, whereas a switching of surface charges upon pH increase was observed for Cys-Au and GSH-Au. The isoelectric points (pI) were pI ¼ 5.2 and 3.9 for Cys-Au and GSH-Au, respectively. The pI value of Cys-Au obtained is comparable with that in the literature20 but significantly higher than the reported value of 4.1 for Cys on silver nanoparticles.21 At low pH, both carboxyl and amine groups are protonated leading to the net positive charge at surfaces, whereas deprotonation at high pH rendered the surfaces negatively charged. It should be noted that the pI values observed for adsorbed Cys and GSH are lower than those in

FIG. 2. (Color online) Zeta potential (f) of thiol-gold samples as a function of pH, ranging from 3 to 10. The zeta potential was tested on planar substrates in contact with solutions at controlled electrolyte concentration of 0.15 M using an electrokinetic analyzer. Biointerphases, Vol. 9, No. 2, June 2014

liquid phase (pI ¼ 6.25 and 5.93 for Cys and GSH, respectively), suggesting the effect of substrates on the surface net charge. Overall, the pH-dependent properties of Cys-Au and GSH-Au have been determined and could be employed as active biointerface for controlling the adsorption of charged molecules. B. Protein fouling resistance of SAMs

The QCM-D sensor was employed to quantify the protein adsorption on natural organosulfur SAMs in order to test their antifouling property. The QCM-D is a real-time screening tool for protein adsorption on Au sensor chips. An alternating voltage is applied to drive the oscillation of the quartz sensor at its resonance frequency. The mass change at the sensor surface will be proportional to a change in frequency (Df). When the adsorbed mass is sufficiently soft that it does not follow the sensor oscillation perfectly, this attributes to internal friction in the adlayer and thus to dissipation of oscillation energy. This mass is the dynamic mass incorporating associated water.27,28,32 The more flexible the adlayers, the more the oscillation will induce. Therefore, monitoring protein adsorption requires using the dissipation parameter to fully characterize the adsorption of a viscoelastic structure. The ratio of DD/Df for both protein layers on various surfaces can be important to illustrate the size and structural flexibility of the adsorbed molecules. In this study, the chips covered with pristine Au, Cys-, Met-, and GSH-SAMs were placed in contact with BSA solutions for 10 min, followed by washing with PBS as shown in Fig. 3. Clearly, these natural organosulfur ligands exhibited distinct resistance to protein adsorption. The GSH-Au appeared to be the best antifouling film by having the least adsorption amount (Df ¼ 2.11 Hz, corresponding to DmSauerbrey ¼ 37.26 ng cm2). This is an improvement of 20 times compared to that of the bare Au (Df ¼ 42.32 Hz, corresponding to DmSauerbrey ¼ 749.06 ng cm2). The highest fouling (Df ¼ 26.43 Hz, corresponding to DmSauerbrey ¼ 467.81 ng cm2) among the SAMs was found on the Met-Au sample, most likely due to its low surface coverage and wettability. For the Cys-Au sample which also showing super-hydrophilic characteristic, the BSA adsorption was higher than GSH-Au. This indicated that there are other factors beyond hydrophilicity when accounting for antifouling properties. Notice that the QCM-D measures the

029010-5 Huang et al.: Natural zwitterionic organosulfurs as surface ligands for antifouling and responsive properties

029010-5

FIG. 3. (Color online) QCM-D studies of BSA adsorption from PBS at pH 7.4 on samples of bare Au, Cys-Au, Met-Au, and GSH-Au. The frequency (a) and dissipation (b) changes as a function of time are present. The BSA solutions of 1 mg ml1 were flowed over the samples for 10 min, followed by washing with PBS twice.

mass of adsorbed proteins along with the trapped water, which varies significantly depending on the structure of the protein film.32,33 Therefore, the resultant D values provide an assessment on the viscoelasticity of the protein films.34 In this sense, the film with high DD/Df value, representing induced energy dissipation per coupled unit mass, is regarded as a viscous or hydrated adlayer. From Fig. 3, we find that the ratio of DD/Df of BSA on Cys-Au (DD/Df ¼ 7.69  108 Hz1) is higher than that on Met-Au (DD/ Df ¼ 4.81  108 Hz1) and bare Au (DD/Df ¼ 3.79  108 Hz1). In other words, a more rigid and flatter adsorbed BSA layer was found on the incompletely covered Met-Au and bare gold than that on Cys-Au. This strongly suggests the occurrence of tight binding and denaturation of BSA on the bare Au surfaces. Compared with the polymer brush systems of antifouling coatings, natural organosulfur SAMs remain insufficiently capable to resist protein fouling. After optimization of the film thickness35 and packing density,36 surface-grafted zwitterionic polymer brushes enable to repel the protein adsorption, particularly in 100% serum and plasma, to levels below 5 ng cm2. The excellent performance of polymer-based systems relies on not only the chemical composition of polymers but also the physical parameters, which are tunable by varying the reaction time, monomer concentration, initiator density, types of solvent, catalyst and ligand, and so on.37 In previous study, Yang et al. controlled the film thickness of the CB polymer by changing the mole ratio of catalysts and found that the lowest fouling level was achieved at the thickness of 21 nm.35 For SAM systems, their antifouling properties are strongly limited by absence of chain flexibility (steric barrier) and complete coverage on “nonideal” flat surfaces.3 However, the SAM coatings are still valuable due to their advantages of versatility, rapid, and facile preparation. The surface charge of GSH-Au could then be correlated with the BSA adsorption behavior. The pI values of BSA and adsorbed GSH are 4.7 and 3.9, respectively. Thus, the pH range studied was chosen from 3 to 6 to further evaluate the impact of electrostatic interactions on protein adsorption. Biointerphases, Vol. 9, No. 2, June 2014

Figure 4 shows that high fouling level (Df ¼ 47.50 Hz) occurred at pH 4 and relatively low fouling level (Df ¼ 8.72 Hz) was found at pH 5. At pH 3 and pH 6, the surface basically repels BSA as if it was in PBS (Fig. 3). The strong correlation between the surface charges and the adsorbed protein suggested that Coulombic attraction between BSA and GSH plays a role at pH 4. The protein adsorption occurs when it is energetically favorable or when the total free energy of the system (i.e., proteins, surface, and water molecules) experiences a negative change (DG < 0). To prevent the adsorption of proteins, enthalpy loss and entropy gain should be avoided. On hydrophobic surfaces, such as the Met-Au and bare Au, there is an entropy gain (DS > 0) associated with the disturbing of the interface water molecules and the replacement of them with denaturized protein. Consequently, high fouling level and strong adhesion of protein are observed (Fig. 3). In contrast, tightly bounded water layers are formed on the SAM of GSH via hydrogen bonds and ionic salvation. The bounded water layers prevent the entropy gaining process to happen.

FIG. 4. (Color online) pH effects on protein adsorption investigated by QCM-D. The BSA solutions at pH 3 to 6 were prepared and flowed over the GSH-Au surfaces for 10 min. Afterward, the buffer solutions at corresponding pH were introduced to remove unbound proteins.

029010-6 Huang et al.: Natural zwitterionic organosulfurs as surface ligands for antifouling and responsive properties

The charge interaction also plays an important role in this model system. From Figs. 2 and 4, we find that BSA and GSH-Au exhibit opposite net charges at pH 4. Thus, the protein fouling is energetically favorable due to the Coulombic interaction and charge-induced exothermic (DH < 0) reaction.38 However, the electrostatic repulsion becomes prominent in the antifouling behavior at a different pH (e.g., pH ¼ 3), where the protein and the surface carry the same charges. The surface potentials of GSH-Au can thus be modulated by varying the pH value to control molecular interactions. Overall, GSH shows the best effectiveness in antifouling properties under the physiological environment among the natural organosulfur molecules. IV. SUMMARY AND CONCLUSIONS In conclusion, we have found by the QCM-D study that the GSH exhibits the strongest repelling force to BSA, thereby the most effective antifouling ability among the three natural organosulfur ligands. The zwitterionic nature and a complete surface coverage have made the BSA adhesion an energetically unfavorable process. In addition, the surface potential of organosulfur-modified surfaces can be modulated by adjusting pH values. This unique feature allows controlling the adsorption behavior to become a responsive biointerface. This systematic investigation was carried out not only for demonstrating the BSA resistance of zwitterionic organosulfur compounds but for providing fundamental insight into designing the surface chemistry for biocompatible medical devices and nanomaterials. ACKNOWLEDGMENTS The authors acknowledge the (NSC-101-2218-E-008-009) for project. They would like to (Academia Sinica) for assistance measurements.

National Science Council financial support of this thank Jing-Jong Shyue of XPS and zeta potential

1

J. E. Raynor, J. R. Capadona, D. M. Collard, T. A. Petrie, and A. J. Garcia, Biointerphases 4, FA3 (2009). 2 R. Vasita, K. Shanmugam, and D. S. Katti, Curr. Top. Med. Chem. 8, 341 (2008). 3 S. F. Chen, L. Y. Li, C. Zhao, and J. Zheng, Polymer 51, 5283 (2010). 4 J. C. Love, L. A. Estroff, J. K. Kriebel, R. G. Nuzzo, and G. M. Whitesides, Chem. Rev. 105, 1103 (2005). 5 R. G. Nuzzo, L. H. Dubois, and D. L. Allara, J. Am. Chem. Soc. 112, 558 (1990).

Biointerphases, Vol. 9, No. 2, June 2014

6

029010-6

H. Kondoh, C. Kodama, and H. Nozoye, J. Phys. Chem. B 102, 2310 (1998). 7 C. S. Chen, M. Mrksich, S. Huang, G. M. Whitesides, and D. E. Ingber, Science 276, 1425 (1997). 8 I. Banerjee, R. C. Pangule, and R. S. Kane, Adv. Mater. 23, 690 (2011). 9 H. J. Milton, Poly(Ethylene Glycol) Chemistry: Biotechnical and Biomedical Applications (Plenum, New York, 1992). 10 S. Y. Jiang and Z. Q. Cao, Adv. Mater. 22, 920 (2010). 11 C. J. Huang, N. D. Brault, Y. T. Li, Q. M. Yu, and S. Y. Jiang, Adv. Mater. 24, 1834 (2012). 12 A. J. Keefe and S. Y. Jiang, Nat. Chem. 4, 59 (2012). 13 W. Yang et al., Biomaterials 33, 7945 (2012). 14 R. E. Holmlin, X. X. Chen, R. G. Chapman, S. Takayama, and G. M. Whitesides, Langmuir 17, 2841 (2001). 15 K. Uvdal, P. Bodo, and B. Liedberg, J. Colloid Interface Sci. 149, 162 (1992). 16 P. Tengvall, I. Lundstrom, and B. Liedberg, Biomaterials 19, 407 (1998). 17 J. W. Kim, Y. M. Lee, S. M. Lee, M. J. Son, H. Kang, and Y. Park, Langmuir 26, 5632 (2010). 18 J. E. Rosen and F. X. Gu, Langmuir 27, 10507 (2011). 19 W. H. Liu, H. S. Choi, J. P. Zimmer, E. Tanaka, J. V. Frangioni, and M. Bawendi, J. Am. Chem. Soc. 129, 14530 (2007). 20 W. Sanders and M. R. Anderson, J. Colloid Interface Sci. 331, 318 (2009). 21 E. Csapo, R. Patakfalvi, V. Hornok, L. T. Toth, A. Sipos, A. Szalai, M. Csete, and I. Dekany, Colloids Surf. B 98, 43 (2012). 22 G. Dodero, L. De Michieli, O. Cavalleri, R. Rolandi, L. Oliveri, A. Dacca, and R. Parodi, Colloids Surf., A 175, 121 (2000). 23 G. Gonella et al., J. Phys. Chem. B 109, 18003 (2005). 24 O. Cavalleri, G. Gonella, S. Terreni, M. Vignolo, L. Floreano, A. Morgante, M. Canepa, and R. Rolandi, Phys. Chem. Chem. Phys. 6, 4042 (2004). 25 M. Lestelius, B. Liedberg, I. Lundstrom, and P. Tengvall, J. Biomed. Mater. Res. 28, 871 (1994). 26 See supplementary material at http://dx.doi.org/10.1116/1.4869300 for the representative XPS spectra of GSH-Au for elements C 1s, N 1s, O 1s, and S 2p. 27 C. A. Keller and B. Kasemo, Biophys. J. 75, 1397 (1998). 28 C. A. Keller, K. Glasmastar, V. P. Zhdanov, and B. Kasemo, Phys. Rev. Lett. 84, 5443 (2000). 29 C. J. Huang, N. J. Cho, C. J. Hsu, P. Y. Tseng, C. W. Frank, and Y. C. Chang, Biomacromolecules 11, 1231 (2010). 30 Y. Q. Fu, H. J. Du, S. Zhang, and W. M. Huang, Mater. Sci. Eng., A 403, 25 (2005). 31 R. E. Thomas, M. J. Mantini, R. A. Rudder, D. P. Malta, S. V. Hattangady, and R. J. Markunas, J. Vac. Sci. Technol. A 10, 817 (1992). 32 F. Hook et al., Colloids Surf., B 24, 155 (2002). 33 C. Fant, H. Elwing, and F. Hook, Biomacromolecules 3, 732 (2002). 34 X. D. Su, Y. J. Wu, R. Robelek, and W. Knoll, Langmuir 21, 348 (2005). 35 W. Yang, H. Xue, W. Li, J. L. Zhang, and S. Y. Jiang, Langmuir 25, 11911 (2009). 36 C. J. Huang, Y. T. Li, J. B. Krause, N. D. Brault, and S. Y. Jiang, Macromol. Rapid Commun. 33, 1003 (2012). 37 S. Edmondson, V. L. Osborne, and W. T. S. Huck, Chem. Soc. Rev. 33, 14 (2004). 38 R. Kun, M. Szekeres, and I. Dekany, J. Therm. Anal. Calorim. 96, 1009 (2009).