NHS Activation Mechanisms ... - ACS Publications

Aug 19, 2011 - dx.doi.org/10.1021/la202267p |Langmuir 2011, 27, 12058-12068 ... Cuie Wang, Qin Yan, Hong-Bo Liu, Xiao-Hui Zhou, and Shou-Jun Xiao*...
0 downloads 0 Views 4MB Size
ARTICLE pubs.acs.org/Langmuir

Different EDC/NHS Activation Mechanisms between PAA and PMAA Brushes and the Following Amidation Reactions Cuie Wang, Qin Yan, Hong-Bo Liu, Xiao-Hui Zhou, and Shou-Jun Xiao* State Key Laboratory of Coordination Chemistry, School of Chemistry and Chemical Engineering, Nanjing National Laboratory of Microstructures, Nanjing University, Nanjing 210093, Jiangsu, China

bS Supporting Information ABSTRACT: Infrared spectroscopy was applied to investigate the well-known EDC/NHS (N-ethyl-N0 -(3-(dimethylamino)propyl)carbodiimide/N-hydroxysuccinimide) activation details of poly(acrylic acid) (PAA) and poly(methacrylic acid) (PMAA) brushes grafted on porous silicon. Succinimidyl ester (NHSester) is generally believed to be the dominant intermediate product, conveniently used to immobilize biomolecules containing free primary amino groups via amide linkage. To our surprise, the infrared spectral details revealed that the EDC/NHS activation of PMAA generated anhydride (estimated at around 76% yield and 70% composition), but not NHS-ester (around 5% yield and 11% composition) under the well-documented reaction conditions, as the predominant intermediate product. In contrast, EDC/NHS activation of PAA still follows the general rule, i.e., the expected NHS-ester is the dominant intermediate product (around 45% yield and 57% composition), anhydride the side product (40% yield and 28% composition), under the optimum reaction conditions. The following amidation on PAA-based NHS-esters with a model amine-containing compound, L-leucine methyl ester, generated approximately 70% amides and 30% carboxylates. In contrast, amidation of PAA- or PMAA-based anhydrides with L-leucine methyl ester only produced less than 30% amides but more than 70% carboxylates. The above reaction yields and percentage compositions were estimated by fitting the carbonyl stretching region with 5 possible species, NHS-ester, anhydride, N-acylurea, unreacted acid, unhydrolyzed tert-butyl ester, and using the BeerLambert law. The different surface chemistry mechanisms will bring significant effects on the performance of surface chemistry-derived devices such as biochips, biosensors, and biomaterials.

1. INTRODUCTION Cross-linking and conjugation of biomolecules to different kinds of substrates is widely used in biomaterials, biosensors, biochips, chromatography, and lab-on-a-chip.114 A standard method for immobilization of NH2-containing biomolecules onto carboxyl-containing substrates via covalent amide bond is using N-ethyl-N0 -(3-(dimethylamino)propyl)carbodiimide (EDC) and N-hydroxysuccinimide (NHS) (molecular structures of EDC and NHS, please see Scheme 1), which is well-known in peptide synthesis.15 The EDC/NHS activation approach possesses many merits: high conversion efficiency, mild reaction conditions, excellent biocompatibility with little influence on the bioactivity of target molecules, and much cleaner products than other crosslinking reagents such as glutaraldehyde and formaldehyde. Due to these advantages, EDC/NHS activation of carboxylic acids and the following amidation reaction have been widely applied in biomolecular conjugation and immobilization of proteins,5,1622 peptides,23 DNAs,24 and so forth to many kinds of substrates of polymers,16,17 noble metals,2527 silicon,18,28,29 nanoparticles,16,30,31 nanotubes,32,33 and so forth. The molecular activation mechanisms of EDC/NHS34 are generally believed to be as shown in Scheme 1: first an O-acylurea intermediate is formed by addition of carboxylic acid to carbodiimide of EDC ((5) in Scheme 1), then the main product of NHS-ester is formed by r 2011 American Chemical Society

nucleophilic attack of NHS to O-acylurea ((6) in Scheme 1) with a high yield under optimized reaction conditions (up to 100%, as claimed in ref 34). On the other side, competitive paths are formation of anhydride by dehydration of O-acylurea with a neighboring carboxylic acid ((7) in Scheme 1), which may further evolve to NHS-ester depending on its structural stability, and formation of stable N-acylurea via an intramolecular acyl rearrangement ((8) in Scheme 1), which is negligible at the wellaccepted reaction conditions and only becomes serious at extreme conditions such as high temperatures and high concentrations.35 Anhydride is also a linking moiety and can react directly with primary amine, resulting in equal quantities of two products, amide and carboxylic acid. NHS-ester, however, reacts with primary amine to produce only a single product of amide. From the product viewpoints, obviously the cross-linking efficiency of anhydride is only half that of NHS-ester. In addition, different structural anhydrides have much different hydrolysis rates, not like NHS-esters which have similar hydrolytic behavior during the amidation reaction. Therefore, NHS-ester is preferred in Received: June 16, 2011 Revised: August 18, 2011 Published: August 19, 2011 12058

dx.doi.org/10.1021/la202267p | Langmuir 2011, 27, 12058–12068

Langmuir

ARTICLE

Scheme 1. Chemistry Approaches on Hydride-Terminated Porous Silicon Surface to Couple NH2-Containing Biomoleculesa

a (1) ω-Undecylenyl alcohol is grafted on porous silicon under microwave irradiation by surface hydrosilylation,52 (2) a surface-bound initiator (2-bromoisobutyryl group) is prepared by acylation,52 (3) PAA and PMAA brushes are grown via surface-initiated atom transfer radical polymerization (SI-ATRP) and (4) post-treatment such as hydrolysis,5,17 (5) the carboxylic acid activation mechanism by EDC/NHS is generally accepted as that O-acylurea is the first intermediate by reaction of EDC with carboxylic acid, (6) O-acylurea is converted to NHS-ester by reaction with NHS, (7) O-acylurea is converted to anhydride by coupling a neighboring carboxylic acid, (8) O-acylurea can also be converted to N-acylurea via an intramolecular acyl transfer as a side product, (9) both NHS-ester and anhydride can couple with NH2-containing biomolecules to generate the biomoleculeimmobilized surface product. Our data show that, with RdH, NHS-ester is the dominant intermediate on PAA brushes, in contrast with RdCH3, anhydride is the dominant intermediate on PMAA brushes. EDC is represented in neutral form for conciseness, although it is in cationic form under our reaction conditions.

amidation even though anhydride can be easily obtained by using only EDC for dehydration of carboxylic acids. Although the EDC/NHS system has been widely used, the details of EDC/NHS activation of molecular monolayers on solid supports, e.g., gold and silicon,34,3640 and of the following amidation of NHS-ester for biomolecular immobilization have been reported only recently. Sam et al. investigated the in situ EDC/NHS activation of end carboxylic acids of an undecylenic acid (UA) monolayer on porous silicon34,36 and provided a phase diagram, where the pure NHS-ester can be obtained under the optimal conditions of 5 mM < [EDC] ≈ [NHS] < 10 mM at room temperature. Side products of anhydride and N-acylurea were also observed only with neglectable numbers. Other research groups, including G. J. Vancso et al. and L. Smith et al., reported the kinetic amidation and hydrolysis details of NHS-ester end groups in molecular monolayers.4143 The competition between amidation and hydrolysis, which influences the biomolecular conjugation efficiency, has been emphasized because NHS-ester is highly prone to base hydrolysis both in solution and on surface.

Recently, polymer brushes have attracted much attention for biomedical applications because they offer certain advantages: the hydrogel brushes are similar to biomembranes and thus provide biologically simulated environments to enable the attached biomolecules to maintain bioactivity; they have periodic functional groups on their backbones and therefore provide multiple sites for binding biomolecules and increase the surface loading capability of solid supports. Of these functional polymer brushes, poly(acrylic acid) (PAA) and polymethacrylate (PMAA) are water-soluble and conveniently achievable. Their side-chain carboxylic acid termini are widely used in anchoring biomolecules through EDC/NHS activation. For example, polymer brushes of PMAA were used to immobilize bovine serum albumin on hairy diamond nanoparticles,16 to covalently insert peptide-functionalized segments into the brush chain to influence the cellular morphologies.23 Polymer brushes of PAA were applied to immobilize as many as 80 monolayers of proteins for decreasing the detection limit of antibodies on microarrays,5 to graft 30 times more ribonuclease A than on self-assembled monolayers for 12059

dx.doi.org/10.1021/la202267p |Langmuir 2011, 27, 12058–12068

Langmuir increasing the loading of enzymes per area,20 to anchor DNAs for DNA chip fabrication,24 and so forth. In a reverse way, PAA polymers in aqueous media were grafted onto a nylon-6,6 film carrying primary amines to improve the biocompatibility of nylon-6,6 by means of EDC/NHS activation.44 Although EDC/NHS activation of PAA or PMAA and the following amidation have been widely used as mentioned above, to our knowledge, the details have not yet been described. Taking it for granted, nearly all researchers just accepted the general rule, EDC/NHS activation of carboxylic acids to NHS-esters, followed with coupling amine-containing biomolecules to amide products,5,16,23,4548 without questioning the truth. In our effort to activate both PAA and PMAA brushes by EDC/NHS for coupling biomolecules (Scheme 1), to our surprise, the infrared spectral details revealed that the EDC/NHS activation of PMAA generated anhydride as the predominant intermediate product under the well-documented activation conditions. In contrast, EDC/NHS activation of PAA brushes still followed the general rule, i.e., NHS-ester was the predominant product under the wellaccepted reaction conditions.5,17 We proposed a molecular mechanism for anhydride formation on PMAA: a six-member ring of cyclic anhydride with chair conformation and two methyl groups in axial positions should be a very stable intermediate product and it would not be converted to NHS-ester any more under the well-documented reaction conditions. The following amidation on PAA-based NHS-esters has much higher efficiency than on PAA- or PMAA-based anhydrides. We have focused on immobilization of biomolecules to planar and porous silicon surfaces for biomedical applications such as protein chips and biosensors.49,50a,51,52 Scheme 1 presents a synthesis strategy for such purposes using PAA or PMAA brushes: (1) ω-undecylenyl alcohol (UO) was grafted under microwave irradiation to porous silicon by surface hydrosilylation,52 (2) a surface-bound initiator (2-bromoisobutyryl group) was prepared by acylation,52 (3) PAA or PMAA brushes were grown via surface-initiated atom transfer radical polymerization (SI-ATRP) and (4) post-treatment such as hydrolysis,5,17 finally PAA or PMAA brushes were activated by EDC/NHS (5,6,7,8) followed by coupling NH2-containing biomolecules to product (9). Since the dominant EDC/NHS activation product on PAA is NHSester (6), while on PMAA it is anhydride (7), the amidation reactions with NH2-containing molecules should be much different. Therefore, we further studied their amidation kinetics and efficiencies with a model NH2-containing molecule, L-leucine methyl ester.

2. EXPERIMENTAL SECTION 2.1. Materials. Single-side polished silicon wafers (Æ100æ, p-type, boron-doped, 5.08.0 Ω.cm, 500 μm thick) were purchased from Hefei Kejing Materials Technology Co. Ltd. N-Ethyl-N0 -(3-(dimethylamino)propyl)carbodiimide hydrochloride (EDC) (98%) were from Aladdin. N-Hydroxysuccinimide (NHS) (98%), sodium methacrylate, bipyridine, tert-butyl acrylate, tert-butyl methacrylate, ω-undecylenyl alcohol (98%), 2-bromoisobutyryl bromide (98%), 4-morpholine ethanesulfonic acid hydrate (MES) and L-leucine methyl ester hydrochloride were from Alfa Aesar. Copper(I) bromide (CuBr, 98%) was from Aldrich. Water (18 MΩ.cm) was from a Milli-Q Ultrapure water purification system. 2.2. Fourier Transform Infrared Spectrometry. A Bruker V80 spectrometer equipped with a deuterated triglycine sulfate (DTGS) detector was used to illuminate the stepwise reactions. In all cases,

ARTICLE

samples were mounted in a vacuum chamber and thus the interference of CO2 and water vapor from air was greatly attenuated. Spectra were recorded with 32 scans at a 7.5 kHz velocity and a resolution of 4 cm1, using a cleaned planar Si (100) chip as reference. All spectra were disposed with the OPUS software and spectral windows were zoomed to highlight the significant changes. 2.3. Preparation of Porous Silicon. The wafer was cut into 15  15  0.5 mm3 pieces. Silicon chips were sonicated in ethanol, acetone, and deionized water, each for 15 min. Then, they were immersed in piranha solution (H2SO4/H2O2 (3/1 in v/v)) at the boiling temperature for 2 h to remove any organic residues. The cleaned chips were rinsed extensively with water and ethanol and subsequently dried in a stream of nitrogen (unless claimed, all the following chip cleaning and drying procedure adopted the same steps). We used the metal-assisted stain etching to produce porous silicon.50 The chip was immersed in 1% HF aqueous solution for 15 min to remove the surface silicon oxide film. After cleaning, the hydrided chip was deposited with a thin metallic film of Pt (12.8 nm thick) using a sputter-coater (SCD 500) at 15 mA for 200 s. The Pt-coated chip was etched in a mixture of 40% HF, 30% H2O2, and anhydrous ethanol with 2:2:1 (v/v/v) ratios for 4 min in dark and sequentially immersed in 1% HF aqueous solution for 1 min. After rinsing with copious amounts of water and ethanol and drying with N2, hydrogen-terminated porous silicon chips were gained.

2.4. Surface Hydrosilylation and Introduction of Surface Initiator. The freshly etched silicon chips were transferred into a vial (80 mL) containing 10 mL neat UO. The bottle was purged with N2 for 15 min so as to vent the air. The reaction was performed in a CEM Discovery microwave reactor, controlled with a dynamic mode to reach 120 C in 10 min and held there for 20 min. After reaction, the chip was washed sonically with anhydrous alcohol for 3 min and water for 3 min, respectively. The UO-modified chip with hydroxy termini was put into a glass bottle containing 10 mL CH2Cl2 and 2 mL Et3N. The reaction vessel was cooled in an ice bath for 15 min, and then 2 mL 2-bromoisobutyryl bromide was dropped in slowly, and finally, the reaction solution stayed at room temperature for 12 h. After reaction, the chip with surface initiator of 2-bromoisobutyryl groups was rinsed with copious amounts of CH2Cl2 and ethanol, and dried with a mild nitrogen stream. 2.5. Preparation of PAA and PMAA Brushes. PAA and PMAA brushes covalently linked to porous silicon can be prepared through two approaches: (i) poly(tert-butyl acrylate) (PtBA) and poly(tert-butyl methacrylate) (PtMBA) brushes were first grown by ATRP directly from a 2-bromoisobutyryl pendant porous silicon chip, then followed by hydrolysis to generate PAA and PMAA polymer brushes, respectively;5,17 (ii) ATRP of sodium acrylate and sodium methacrylate, respectively, followed by acidification.53 Surface-initiated polymerization of tert-butyl acrylate (or tert-butyl methacrylate) was performed in a N2-filled vessel with 18 mg (0.125 mmol) CuBr and 40 mg (0.26 mmol) bipyridine (bipy) dissolved in a 3 mL solution of monomer (tert-butyl acrylate or tert-butyl methacrylate))/ CH3OH (1/2 in v/v) at 40 C for 12 h. Then, the chips were removed from the vials, washed with THF and ethanol sequentially, and dried under a stream of N2. PAA (or PMAA) brushes were obtained by hydrolysis of PtBA (or PtBMA) in a solution of methanesulfonic acid (0.3 mL) in dichloromethane (10 mL) for 3 h at room temperature. The thickness of PAA (or PMAA) brushes after staining with 1% acetate uranium for 5 min was measured to be at around 200 nm by SEM (Hitachi S-4800, Japan) cross section imaging (see Figure S3 in Supporting Information). ATRP of sodium acrylate or sodium methacrylate followed by acidification also provided PAA or PMAA brushes, which showed experimental data similar to hydrolyzed ones. We did not provide the detail and experimental data here for conciseness.

2.6. Surface EDC/NHS Activation of PAA and PMAA Brushes. The EDC/NHS solutions with different concentrations 12060

dx.doi.org/10.1021/la202267p |Langmuir 2011, 27, 12058–12068

Langmuir

Figure 1. (a) Transmission FTIR spectra of polymer brushes on porous silicon via SI-ATRP: (1) PAA obtained by hydrolysis of PtBA, (2) PtBA, (3) PMAA obtained by hydrolysis of PtMBA, (4) PtMBA. (b) Fitting spectra of υCdO in PAA. (c) Fitting spectra of υCdO in PMAA. and molar ratios were prepared in ionized water. The PAA or PMAA chip was immersed in an appropriate concentration mixture of EDC and NHS (2 mL) at 25 C for 1 h. After reaction, the chip was rinsed with water and dried under a stream of nitrogen.

2.7. Amidation of NHS-Ester and Anhydride in a L-Leucine Methyl Ester Carbonate Buffer. The NHS-esterified PAA chip was incubated in 0.1 mM L-leucine methyl ester carbonate buffer at pH 8.5 (carbonate buffer was prepared by addition of 0.1 M NaHCO3 to 0.1 M Na2CO3 and pH is adjusted to 8.5) and taken out at 1, 4, 10, 15, 20, 30, and 40 min for infrared measurements. The anhydride surface was immersed in the same amidation solution for only 1 min. While the anhydride of the PMAA chip was incubated in the same amidation solution and taken out at 30, 60, 90, 120, and 150 min for infrared measurements.

3. RESULTS AND DISCUSSION 3.1. Preparation of PAA and PMAA Brushes by SurfaceInitiated Atom Transfer Radical Polymerization. PAA and

PMAA brushes were grown on porous silicon from a surfacegrafted initiator, 2-bromoisobutyryl group, via SI-ATRP. Two routes were tried: SI-ATRP of tert-butylacrylate and tert-butyl(meth)acrylate, followed by hydrolysis of the tert-butyl ester protective groups;5,17,54 SI-ATRP of sodium acrylate and sodium (meth)acrylate, followed by acidification.53 Since both PAA and PMAA brushes obtained from the first route are more even and more controllable than the latter,5,17,54 for quantification purposes, we carried on our works mainly on PAA and PMAA brushes hydrolyzed from PtBA and PtBMA ones. However, the second route also has its own merits:53 the brush density is higher; the brush can grow thicker; ATRP is carried out in aqueous media; and most importantly, acidification can convert all sodium salts into acids, while hydrolysis of PtBA or PtBMA cannot reach 100% and it always leaves some tert-butyl ester residues behind. Figure 1 illuminates the infrared spectra of hydrolysis of PtBA ((2) in a) to PAA ((1) in a) and PtBMA ((4) in a) to PMAA ((3) in a) and the deconvolution spectra of υCdO in PAA (b) and in PMAA (c), respectively. Obviously, the presence or absence of the absorption of tert-butyl groups causes the spectral differences between PtBA and PAA (or PtBMA and PMAA). Main differences can be addressed as follows: a broad υCdO band centered at 1717 cm1 appears on PAA (or PMAA) against the narrow ester band at 1733 cm1 on PtBA (or 1729 cm1 on PtMBA), contributed mainly from the newly formed acid species and to a lesser extent from the remaining ester residues; the CH3 stretching bands at 2978 cm1 PtBA (or PtMBA) greatly decrease on PAA (or PMAA) due to the leaving

ARTICLE

Figure 2. Transmission IR spectra of PAA brushes after EDC/NHS activation: a, 5 mM EDC, 10 mM NHS; b, 50 mM EDC, 100 mM NHS; c, 0.1 M EDC, 0.2 M NHS; d, 0.5 M EDC, 1 M NHS.

of tert-butyl groups. To know the compositions of acid and ester, we fit the υCdO band with acid and ester. For PAA brushes in Figure 1b, the fitting results show that the acid composition is 70% and the ester composition is 30%, while for PMAA brushes in Figure 1c, acid is 72% and ester is 28%. 3.2. Activation of PAA Brushes by EDC/NHS. Influence of NHS and EDC Concentrations. Since a molar ratio of EDC/NHS = 1/2 for activation of acids is widely used in the literature,16,55 we adapted this ratio first and investigated the influence of different concentrations on the surface products. Figure 2 shows the infrared spectra of surface products after activation of PAA brushes in the EDC/NHS mixture at four different concentrations: a, 5 mM EDC, 10 mM NHS; b, 50 mM EDC, 100 mM NHS; c, 0.1 M EDC, 0.2 M NHS; d, 0.5 M EDC, 1 M NHS. The characteristic infrared bands are assigned in Table 1 and classified as NHS-ester with associated triplex bands of 1740, 1780, and 1815 cm1;34,49 anhydride with associated doublet bands at 1760 and 1804 cm1, acid with a single band at 1717 cm1, tert-butyl ester with a single band at 1733 cm1, N-acylurea (see Scheme 1 for its structure) with associated doublet bands of amide at 1550 and 1650 cm1 and an imide band at ∼1730 cm1. Except for the amide bands, carbonyl bands from the above 5 species overlap with each other. Obviously, at the lowest concentration of 5 mM EDC and 10 mM NHS, a small amount of anhydride and N-acylurea appears in Figure 2a, while most of the acids, indicated by the acid band at 1717 cm1, still remain unreacted. The carbonyl stretching band (υCdO) was broadened after activation, mainly due to the appearance of side products of anhydride and N-acylurea. By increasing the concentration to 50 mM EDC and 100 mM NHS, more anhydride appears in Figure 2b; NHS-ester does so but only the strongest band at 1740 cm1 exhibits an observable peak, and the other two bands, 1780 and 1815 cm1, are covered by the strong anhydride bands. Further increasing the concentration to 0.1 M EDC and 0.2 M NHS, the target product, NHS-ester, dominates the carbonyl stretching region with overwhelming typical triplex bands of 1740, 1780, and 1815 cm1 in Figure 2c. To confirm that the optimum concentration is 0.1 M EDC and 0.2 M NHS, we recorded a spectrum at a very high concentration of 0.5 M EDC and 1 M NHS in Figure 2d, where the spectrum does not show any significant improvement of NHS-ester bands but an increasing acid band at 1717 cm1 from Figure 2c, indicating more acid residues left unreacted. The decrease of activation efficiency at 12061

dx.doi.org/10.1021/la202267p |Langmuir 2011, 27, 12058–12068

Langmuir

ARTICLE

Table 1. Infrared Peak Assignments in the Carbonyl Stretching Region from 1500 to 2000 cm1

a

1740

νas (imidyl CdO)

1780

νs (imidyl CdO)

NHS-ester CdO symmetric stretch

1815

ν (ester CdO)

NHS-ester carbonyl stretch

NHS-ester CdO antisymmetric stretch

1760

νas (anhydride CdO)

anhydride antisymmetric CdO stretch

1804 1733

νs (anhydride CdO) ν (ester CdO)

anhydride symmetric CdO stretch ester CdO stretch of tert-butyl ester, N-acylurea; L-leucine methyl estera

1710

ν (acid CdO)

carboxylic acid CdO stretch

1660

ν (amide CdO)

peptide CdO stretch (amide I)

1550

ν (NH), ν (COO)

mostly peptide NH bend (amide II); COOa

These carbonyl stretches are not shown in Figure 2, but in Figures 7 and 8.

Figure 3. Deconvolution of the carbonyl stretching region of Figure 2c: red triplex bands, NHS-ester; blue doublet bands, anhydride; pink single band, tert-butyl ester; brown single band, acid.

higher concentrations such as 0.5 M EDC and 1 M NHS is most probably due to precipitation of reactants on the chip surface which somehow blocks the surface reaction.34,41 From the above data analysis, we chose 0.1 M EDC and 0.2 M NHS as the optimum concentration for EDC/NHS activation of PAA brushes. It is about 10 times the optimum concentration for EDC/NHS activation of carboxylic acid monolayers (5 mM < [EDC] ≈ [NHS] < 10 mM),34,36 most probably due to more surface carboxylic acid species. After determining the optimal activation concentration, we also ran two control experiments, activation of PAA with 0.1 M EDC and 0.2 M NHS separately. The IR spectrum recorded after exposure to NHS did not reveal any significant changes from PAA brushes except for limited surface oxidation.56,57 However, the spectrum recorded after exposure to EDC exhibited strong anhydride stretching bands at 1760 and 1804 cm1, indicating that anhydride can be easily obtained with treatment of EDC. From the above spectral qualitative analysis, obviously the υCdO band between 1670 and 1850 cm1 is the most attractive region. It includes contributions of 5 species: NHS-ester, anhydride, N-acylurea (appearing at some extreme conditions such as low and high reactant concentrations), unreacted carboxylic acid, and unhydrolyzed tert-butyl ester. To quantitatively analyze the activation efficiency, we fitted the υCdO band of Figure 2c with four species, NHS-ester, anhydride, acid, and ester, shown in Figure 3, where N-acylurea was not included because it did not appear here. The EDC/NHS activation efficiency can be judged from the following three criteria, high yield of NHS-ester, low yield of byproducts such as anhydride and N-acylurea, and high

conversion efficiency of acid. These parameters can be estimated from the fitting spectra of Figure 3. Among the 5 species contributed to the carbonyl region, the tert-butyl ester residue at 1733 cm1 does not participate in the EDC/NHS activation procedure and will be taken out and not be counted in the following calculations for activity efficiency assay. In addition, the N-acylurea byproducts can be judged from simultaneous occurrence of amide (∼1550 and ∼1650 cm1) and imide (∼1735 cm1) bands. N-Acylurea does not exist in Figure 2c according to amide bands not appearing, and so, it is excluded from fitting in Figure 3 too. The relative percentages of surface species can be derived from their corresponding absorption bands by means of the BeerLambert law: A = εbc (where A is the absorbance, ε the extinction coefficient, b the length of the sample layer (here, the thickness of polymer brushes), and c the concentration). The BeerLambert law for NHS-ester, anhydride, acid, and N-acylurea can be written individually as ANHS-ester ¼ εNHS-ester bcNHS-ester Aanhydride ¼ εanhydride bcanhydride AN-acylurea ¼ εN-acylurea bcN-acylurea Aacid ¼ εacid bcacid where b is the same thickness of polymer brushes for all 4 species. We used the largest peak of each species for calculation: 1760 cm1 for anhydride, 1740 cm1 for NHS-ester, ∼1735 cm1 for N-acylurea, and 1717 cm1 for carboxylic acid, which are very strong in infrared spectra. To estimate the reaction yields and the abundance of each species, the linear relationship among the individual extinction coefficients, εNHS-ester, εanhydride, εacid, εN-acylurea, should be known. We used four monomer model compounds: glutaric acid, glutaric anhydride, acetic acid N-hydroxysuccinimide ester, and 1-acetylurea to measure their extinction coefficients (see Supporting Information). We assume that the extinction coefficients from four monomers can represent the linear relationships among PAA-derived εNHS-ester, εanhydride, εacid, and εN-acylurea, as follows: εNHS-ester =εanhydride =εN-acylurea =εacid ¼ 2 : 1:5 : 1:5 : 1 We used the peak height to represent the absorbance A, assuming that all reactants of acids are converted to NHS-ester, anhydride, N-acylurea, acid, and not any other compounds, and then the product yields of NHS-ester, anhydride, N-acylurea, and 12062

dx.doi.org/10.1021/la202267p |Langmuir 2011, 27, 12058–12068

Langmuir

ARTICLE

Figure 4. (a) Transmission IR spectra of PAA brushes grafted on porous silicon after EDC/NHS activation. EDC and NHS concentrations are indicated on the right-hand side of each spectrum. (b) Reaction yields of NHS-ester, anhydride, N-acylurea, and the acid residue (%) are plotted against molar ratios of [NHS]/[EDC].

the residue carboxylic acid can be calculated as follows: yieldNHS-ester ¼ ðANHS-ester =εNHS-ester Þ=ðANHS-ester =εNHS-ester þ 2Aanhydride =εanhydride þ Aacid =εacid  ð1Þ þ AN-acylurea =εN-acylurea 100%  yieldanhydride ¼ 2Aanhydride =εanhydride =ðANHS-ester =εNHS-ester þ 2Aanhydride =εanhydride þ Aacid =εacid  þ AN-acylurea =εN-acylurea 100% ð2Þ  yieldN-acylurea ¼ AN-acylurea =εN-acylurea =ðANHS-ester =εNHS-ester þ 2Aanhydride =εanhydride þ Aacid =εacid  ð3Þ þ AN-acylurea =εN-acylurea 100% residueacid ¼ ðAacid =εacid Þ= ANHS-ester =εNHS-ester þ 2Aanhydride

=εanhydride þ Aacid =εacid þ AN-acylurea  =εN-acylurea 100%

ð4Þ

From eq 1, the reaction yield of NHS-ester is 45%; from eq 2, the yield of anhydride is 40%; N-acylurea does not exist in this case, and thus, its yield is zero; from eq 4, 15% acid residues remain unreacted. In another way, if we estimate the product percentage compositions, we use coefficient 1 instead of 2 for the item Aanhydride/εacid in the above equations. Therefore, the percentage composition of NHS-ester is 57%, anhydride 28%, and acid 15%. Effect of Molar Ratios of EDC/NHS on Activation. To further explore the influence of molar ratios of EDC/NHS on the reaction, we fixed the concentration of EDC to 0.1 M and changed the concentration of NHS to 0, 0.05, 0.1, 0.2, 0.5, and 1.0 M. Their corresponding infrared spectra are shown in Figure 4a. By fitting the υCdO bands between 1670 and 1850 cm1 of each spectrum in Figure 4a, and calculating the yields of NHS-ester, anhydride, N-acylurea, and the acid residue percentage according to eqs 14, we can plot 4 curves against molar ratios of NHS/EDC, representing activation efficiencies, in Figure 4b. If we increase the NHS concentration while keeping the EDC concentration at 0.1 M, the activation yield varies non-monotonically in Figure 4b. One can notice that the shape of the NHSester triplet band evolves differently from Figure 4a (1) to (6). The activation efficiency can be clearly seen in Figure 4b: The

reaction yield of NHS-ester increases with increasing molar ratio [NHS]/[EDC] from 0 to 2, then decreases with increasing molar ratio further; it reaches the highest (45%) at the molar ratio of [NHS]/[EDC] = 2, which suggests the optimum molar ratio region of NHS/EDC at around 2. The maximum yield of anhydride is 85% when only EDC exists at [EDC] = 0.1 M while [NHS] = 0; its yield decreases to 60%, 50%, 40%, 17%, and 2% with increasing [NHS] to 0.05, 0.1, 0.2, 0.5, and 1.0 M sequentially. The acid residue remains nearly the same (∼20%) in the first 4 points in the [NHS]/[EDC] molar ratio range 02; then it increases faster to 40% and 73% with increasing [NHS] to 0.5 and 1.0 M, respectively. The reaction yield of the side product, N-acylurea, is 0 at the first 4 points ([NHS]/[EDC] = 0, 0.5, 1, 2); then it increases to 23% and 27% at [NHS]/[EDC] = 5 and 10, respectively. Obviously, only at the [NHS]/[EDC] = ∼2 region can we obtain the highest NHS-ester yield (∼45%). This is why we determined the optimum EDC/NHS activation condition for PAA brushes at [EDC] = 0.1 M and [NHS] = 0.2 M. Influence of Temperature. Temperature is another important parameter to influence the chemical reactions. Figure 5 shows the infrared spectral evolution against temperature. The very good shape of NHS-ester at 4 and 20 C indicates the high performance of activation. However, with increasing temperature, the intensity of NHS-ester gradually decreases, most probably due to the hydrolysis of NHS-ester. Large amounts of unreacted acid residues and byproducts of N-acylurea can be found at high temperatures (50, 60, and 70 C), indicated by the acid stretching band at 1717 cm1 and the N-acylurea amide bands at 1650 (amide I) and 1550 cm1 (amide II) and the N-acylurea imide band at ∼1735 cm1, respectively. Formation of N-acylurea is via an intramolecular rearrangement of acyl transfer from O-acylurea. We suggest that, at higher temperatures, hydrolysis of EDC, NHS-ester, and acyl transfer from O-acylurea to stable N-acylurea be the main course for decreasing the activation efficiency. The spectral evolution in Figure 5 indicates that a room temperature (420 C) is most favorable for formation of NHS-ester. 3.3. Activation of PMAA by EDC/NHS. For PMAA, the EDC/ NHS activation is much different from PAA. Figure 6a shows infrared spectral traces at different concentrations of EDC and NHS. Unfortunately, whatever concentrations and molar ratios of EDC/NHS we used, we did not observe the obvious triplex bands of NHS-ester. Instead, we observed the dominant doublet bands of anhydride even at the optimum concentration of 0.1 M EDC and 0.2 M NHS, which was chosen for generating NHS12063

dx.doi.org/10.1021/la202267p |Langmuir 2011, 27, 12058–12068

Langmuir

ARTICLE

Figure 5. Transmission IR spectra of PAA brushes after activation treatment with [NHS]/[EDC] = 2 at different temperatures.

ester on PAA brushes. Does any NHS-ester exist on PMAA brushes? During fitting the spectrum of Figure 6a-(4), we needed the NHS-ester triplet bands to perform the deconvolution, as shown in Figure 6b. The fitting bands represent 4 species: anhydride, NHS-ester, tert-butyl ester, and acid. Similar to Figure 4b, via calculation from eqs 14 we plotted the reaction yields of anhydride, NHS-ester, N-acylurea, and the acid residue against molar ratios of [NHS]/[EDC], shown in Figure 6c. The product of anhydride dominates over the [NHS]/[EDC] ratios of 0, 0.5, 1, and 2. Like in Figure 4b, acid residues and N-acylurea have evolved into considerable components at higher [NHS]/ [EDC] ratios of 5 and 10. NHS-ester only occupies a small part of a yield of 70%) than that of anhydride (