On Surface-Initiated Atom Transfer Radical Polymerization Using

Dec 21, 2010 - Mie Lillethorup , Kyoko Shimizu , Nicolas Plumeré , Steen U. Pedersen , and .... Sampa Saha , Davor Copic , Srijanani Bhaskar , Nichol...
0 downloads 0 Views 1MB Size
pubs.acs.org/Langmuir © 2010 American Chemical Society

On Surface-Initiated Atom Transfer Radical Polymerization Using Diazonium Chemistry To Introduce the Initiator Layer Joseph Iruthayaraj,†,‡ Sergey Chernyy,† Mie Lillethorup,† Marcel Ceccato,† Troels Røn,† Mogens Hinge,†,‡ Peter Kingshott,‡ Flemming Besenbacher,‡ Steen Uttrup Pedersen,*,†,‡ and Kim Daasbjerg*,†,‡ ‡

† Department of Chemistry, Aarhus University, Langelandsgade 140, DK-8000 Aarhus, Denmark, and Interdisciplinary Nanoscience Center (iNANO), Department of Physics and Astronomy, Aarhus University, Ny Munkegade 120, DK-8000 Aarhus C, Denmark

Received October 13, 2010. Revised Manuscript Received November 26, 2010 This work features the controllability of surface-initiated atom transfer radical polymerization (SI-ATRP) of methyl methacrylate, initiated by a multilayered 2-bromoisobutyryl moiety formed via diazonium chemistry. The thickness as a function of polymerization time has been studied by varying different parameters such as the bromine content of the initiator layer, polarity of reaction medium, ligand type (L), and the ratio of activator (CuI) to deactivator (CuII) in order to ascertain the controllability of the SI-ATRP process. The variation of thickness versus surface concentration of bromine shows a gradual transition from mushroom to brush-type conformation of the surface anchored chains in both polar and nonpolar reaction medium. Interestingly, it is revealed that very thick polymer brushes, on the order of 1 μm, can be obtained at high bromine content of the initiator layer in toluene. The initial polymerization rate and the overall final thickness are higher in the case of nonpolar solvent (toluene) compared to polar medium (acetonitrile or N,N-dimethylformamide). The ligand affects the initial rate of polymerization, which correlates with the redox potentials of the pertinent CuII/CuI complexes (L=Me6TREN, PMDETA, and BIPY). It is also observed that the ability of polymer brushes to reinitiate depends on the initial thickness and the solvent used for generating it.

Introduction Surface-initiated polymerization is a “grafting-from” technique which involves in-situ polymerization from a suitable initiator covalently attached to the surface. Among other techniques, living radical polymerization has gained much popularity for providing simple and robust synthetic routes to form well-defined polymer brushes.1 The basic requirement of this kind of radical polymerization is the reversible activation-deactivation step that is established between the growing polymer chain end and a corresponding dormant species. Atom transfer is one of the ways to establish this equilibrium.2 In Scheme 1 the essential steps of an atom transfer radical polymerization (ATRP) is illustrated. A one-electron redox process between an alkyl halide, R-X, and a transition metal complex, Mz-Y/L, generates the alkyl radical, R•, and the oxidized metal complex, X-Mzþ1-Y/L, with a rate constant ka. In these structures, Y denotes a halide and L is an appropriate ligand. The generated radical can subsequently add to the vinyl monomer (M) with a propagation rate constant kp to form the growing polymer chain radical, Pn•, until it is reversibly deactivated by X-Mzþ1-Y/L to afford Pn-X and Mz-Y/L with a rate constant kd. The propagating radical can also undergo irreversible deactivation through coupling and/or disproportionation reactions with a rate constant kt. However, these reactions only happen seldom which can be explained by the fact that the surface concentration of radicals is inherently low owing to the favorable *Corresponding authors: Tel þ45 8942 3908, Fax þ45 8619 6199, e-mail [email protected] (S.U.P.); Tel þ45 8942 3922, Fax þ45 8619 6199, e-mail [email protected] (K.D.). (1) Tsujii, Y.; Ohno, K.; Yamamoto, S.; Goto, A.; Fukuda, T. Adv. Polym. Sci. 2006, 197, 1–45. (2) Mueller, L.; Golas, P.; Matyjaszewski, K. In New Smart Materials via Metal Mediated Macromolecular Engineering; Khosravi, E., Yagci, Y., Savelyev, Y., Eds.; Springer: Berlin, 2009; pp 3-16. (3) Wang, J. S.; Matyjaszewski, K. J. Am. Chem. Soc. 1995, 117, 5614–15.

1070 DOI: 10.1021/la104125n

deactivation of the growing radical chain ends in the equilibrium step.3 The activation-deactivation cycle continues until all the monomers are consumed by the polymerization process. Through surface-initiated atom transfer radical polymerization (SI-ATRP) and similar methods based on this principle a much better control is exerted on the resulting polymer brush structure. Various strategies have been used to covalently anchor ATRP initiators, depending on the substrate. A few examples include self-assembled monolayer of thiols on gold,4-6 chlorosilanes on silicon oxide surfaces,7-9 and alkyl cations through ion exchange on mica10 and montmorillonite.11 Surface functional groups of polymeric materials such as cellulose,12 chitosan,13 poly(ether imide),14 and graphene oxide15 have also been chemically modified to introduce ATRP initiators. Recently, aryldiazonium-based compounds16-18 have been successfully employed to introduce initiators (4) Niwa, M.; Date, M.; Higashi, N. Macromolecules 1996, 29, 3681–85. (5) Jones, D. M.; Brown, A. A.; Huck, W. T. S. Langmuir 2002, 18, 1265–69. (6) Kim, J.-B.; Bruening, M. L.; Baker, G. L. J. Am. Chem. Soc. 2000, 122, 7616–17. (7) Ejaz, M.; Tsujii, Y.; Fukuda, T. Polymer 2001, 42, 6811–15. (8) Blomberg, S.; Ostberg, S.; Harth, E.; Bosman, A. W.; Horn, B. V.; Hawker, C. J. J. Polym. Sci., Polym. Chem. 2002, 40, 1309–20. (9) Xu, C.; Wu, T.; Mei, Y.; Drain, C. M.; Batteas, J. D.; Beers, K. L. Langmuir 2005, 21, 11136–40. (10) Velten, U.; Shelden, R. A.; Caseri, W. R.; Suter, U. W.; Li, Y. Colloids Surf., A 1999, 154, 87–96. (11) Behling, R. E.; Williams, B. A.; Staade, B. L.; Wolf, L. M.; Cochran, E. W. Macromolecules 2009, 42, 1867–72. (12) Carlmark, A.; Malmstrom, E. E. Biomacromolecules 2003, 4, 1740–45. (13) Tang, F.; Zhang, L.; Zhu, J.; Cheng, Z.; Zhu, X. Ind. Eng. Chem. Res. 2009, 48, 6216–23. (14) Li, L.; Yan, G.; Wu, J.; Yu, X.; Guo, Q. High Perform. Polym. 2009, 21, 455–67. (15) Lee, S. H.; Dreyer, D. R.; An, J.; Velamakanni, A.; Piner, R. D.; Park, S.; Zhu, Y.; Kim, S. O.; Bielawski, C. W.; Ruoff, R. S. Macromol. Rapid Commun. 2010, 31, 281–88. (16) Matrab, T.; Chehimi, M. M.; Perruchot, C.; Adenier, A.; Guillez, A.; Save, M.; Charleux, B.; Cabet-Deliry, E.; Pinson, J. Langmuir 2005, 21, 4686–94.

Published on Web 12/21/2010

Langmuir 2011, 27(3), 1070–1078

Iruthayaraj et al.

Article

Scheme 1. Overall ATRP Equilibriuma

a Involves dormant polymer chain (P-X), transition metal complex (Mz-Y/L) of oxidation state z stabilized by ligand (L), active polymer chain radical (Pn•) generated as a result of a one-electron redox process between P-X and Mz-Y/L with a rate constant ka, addition of vinyl monomer (M) to active polymer chain with a rate constant kp, deactivation of active polymer chain with a rate constant kd, and termination of the active chain with a rate contant kt.

on conducting substrates such as metals and glassy carbon (GC) through an electrochemical approach. In contrast to initiators formed by self-assembly process, the diazonium approach results in a randomly oriented multilayer structure with thickness ranging from a few nanometers19 to micrometers,20 depending on the choice of diazonium salt and experimental conditions. In earlier studies (1-bromoethyl)benzene layer was directly electrografted using the corresponding diazonium salt as precursor.16-18 The secondary benzyl bromide was initiated via ATRP process to form poly(methyl methacrylate), PMMA, brushes on GC and iron substrates using toluene as a nonpolar solvent at 75 °C. Since these first reports on the successful introduction of the ATRP process on diazonium generated initiator layers, no systematic study has addressed the issue as to which extent the rather inhomogeneous distribution of initiators in a multilayered film might affect the controllability of the SI-ATRP process. To shed light on this issue, we have carried out a thorough and comprehensive study on the effect various parameters, such as the substrate, surface density of a tertiary alkanoyl bromide initiator, polarity of the reaction medium, and the redox properties of ligand-stabilized CuI catalyst, exerts on the structure of PMMA brushes formed. It is revealed that although the general trend for brush formation is consistent with that observed in the case of other modification procedures, it is also so that the nature of the substrate [GC or stainless steel (SS)] plays an important role for the final outcome despite the surface being covered by the same kind of initiator-containing multilayer in all cases. Furthermore, it will be shown that polymer films with thicknesses even in the micrometer range may be constructed under appropriate conditions.

Experimental Section Materials. Methanol, toluene, dichloromethane, acetone, and hexane were all HPLC grade from Sigma-Aldrich. Acetonitrile (MeCN) (anhydrous, 99.9%) was purchased from Lab-Scan and used as received. 4-(2-Hydroxyethyl)aniline (Aldrich) was used to synthesize 4-(2-hydroxyethyl)benzenediazonium tetrafluoroborate (1) following the procedure outlined in the Supporting Information (SI 1.1). The monomer, methyl methacrylate (Aldrich), and the initiator precursor, 2-bromoisobutyryl bromide (Aldrich), were freshly distilled under reduced pressure. The ligands, N,N, N0 ,N00 ,N00 -pentamethyldiethylenetriamine (PMDETA) and 2,20 bipyridine (BIPY), were used as received from Aldrich without further purification. Tris[2-(dimethylamino)ethyl]amine (Me6TREN) (17) Matrab, T.; Save, M.; Charleux, B.; Pinson, J.; Cabet-deliry, E.; Adenier, A.; Chehimi, M. M.; Delamar, M. Surf. Sci. 2007, 601, 2357–66. (18) Nguyen, M. N.; Matrab, T.; Badre, C.; Turmine, M.; Chehimi, M. M. Surf. Interface Anal. 2008, 40, 412–17. (19) Bernard, M.-C.; Chausse, A.; Cabet-Deliry, E.; Chehimi, M. M.; Pinson, J.; Podvorica, F.; Vautrin-Ul, C. Chem. Mater. 2003, 15, 3450–62. (20) Adenier, A.; Combellas, C.; Kanoufi, F.; Pinson, J.; Podvorica, F. I. Chem. Mater. 2006, 18, 2021–29.

Langmuir 2011, 27(3), 1070–1078

was synthesized according to the Eschweiler-Clarke methylation procedure (see Supporting Information, SI 1.2). The catalysts, CuICl and CuIICl2, were used as received from Aldrich. The supporting electrolyte, tetrabutylammonium tetrafluoroborate (Bu4NBF4), was prepared using a standard procedure. Electrodes. Glassy carbon (GC) plates (Sigradur G, HTW, 10  10  1 mm) were cleaned by sonication in acetone and hexane (15 min in each solvent). Stainless steel (SS) plates (ASTM 316, Sanistaal, Denmark, 10  10  0.7 mm) were polished successively using sand paper (P180, P500, P1000, and P2000) and diamond suspensions (Struers; grain size: 9, 3, 1, and 0.25 μm). Afterward, the plates were washed thoroughly with water and ethanol and finally sonicated for 10 min in acetone and another 10 min in hexane. Electrochemical Setup. A standard three-electrode electrochemical setup (CH Instruments 660B or CH Instruments 601C) consisting of a plate (GC or SS) as working electrode, platinum wire as auxiliary electrode and a Ag/AgI pseudo-reference electrode (i.e., a silver wire immersed in an MeCN solution of 0.1 M Bu4NBF4 and 0.01 M Bu4NI) was used for electrografting 1 (Scheme 2). At the end of each experiment the standard potential 0 of the ferrocenium/ferrocene couple, E Fc þ, was measured, and all potentials were referenced against SCE using a previous determi0 21 nation of E Fc þ = 0.41 V vs SCE in MeCN. Ellipsometry. The thickness of the modified films (in the dry state) was measured using a rotating analyzer ellipsometer (Dre, Germany). The GC and the SS plates were measured at 65° and 75° angle of incidence. The ellipsometric parameters of the bare (Δs, Ψs) and the grafted (Δg, Ψg) substrates were measured in air at ambient temperature, where Δ is the phase shift and tan Ψ is the amplitude ratio upon reflection. The complex refractive index of the bare substrate was calculated from the measured Δs and Ψs values. A three-layer optical model consisting of a substrate with a complex refractive index, the grafted layer characterized by its refractive index and thickness, and the surrounding medium (air) was used to calculate the overall reflection coefficients for in-plane (Rp) and out-of-plane (Rs) polarized lights. Ellipsometric measurements were performed on the same area of the sample plates before and after electrografting. The real and the imaginary parts of the refractive index of the GC plates were obtained by measuring the clean GC substrate prior to modification. Unfortunately, a specific value for GC cannot be used due to surface heterogenities. The ranges of refractive index values we have obtained for various bare substrates were 2.06 to 2.46 (real part) and -0.93 to -1.8 (imaginary part). It is to be noted that the refractive index also varies from one batch of GC plates to another. Because measurements are carried out on dried and collapsed polymeric films, the refractive index of the layer is fixed at a constant value (1.55 and 0 for the real and imaginary parts, respectively), independent of the thickness. The average and the standard deviation values reported correspond to data points obtained with three spot measurements on each plate.

Infrared Reflection Absorption Spectroscopy (IRRAS). IRRAS spectra were recorded on a Bio-Rad FTS 65A (Randolph, MA) FTIR spectrometer equipped with an external experiment module with a narrow band mercury-cadmium-telluride (MCT) detector cooled in liquid nitrogen. The infrared beam was p-polarized by a gold wire polarizer. The spectral resolution and number of scans averaged were 2 cm-1 and 240 for the GC substrate and 4 cm-1 and 800 for the SS substrate. The GC and SS substrates were irradiated with an incident grazing angle of 60° and 80°, respectively. The p-polarized reflectivity of the film, Rp(d), was divided with the reflectivity of the bare substrate, Rp(0), and presented as IRRAS absorbance [-log(Rp(d)/Rp(0))] after baseline-correction using the facilities of the Digilab Resolution Pro 4.0 program. The area of the CdO stretching absorbance band was calculated by the program as well. All spectra were recorded at room temperature in dry atmosphere. (21) Bard, A. J.; Faulkner, L. R. Electrochemical Methods, 2nd ed.; Wiley: New York, 2001.

DOI: 10.1021/la104125n

1071

Article

Iruthayaraj et al.

Scheme 2. (a) Formation of a Covalently Attached Initiator Film through Electrografting of 4-(2-Hydroxyethyl)benzenediazonium Tetrafluoroborate Followed by a Chemical Reaction with 2-Bromoisobutyryl Bromide; (b) Compound 2 Is a Potential Grafting Agent Incorporating an Initiator Group in Terms of a Tertiary Alkanoyl Bromidea

a

Note that the multilayered structures drawn are just two of many possible scenarios.

X-ray Photoelectron Spectroscopy (XPS). A Kratos Axis Ultra-DLD instrument operated with a monochromatic Al KR X-ray source at a power of 75 W was employed for the XPS analysis. Survey scans were acquired by accumulating two sweeps in the 0-1400 eV range at a pass energy of 160 eV. High-resolution scans were acquired at a pass energy of 20 eV. The pressure in the main chamber during the analysis was in the range of 10-8 mbar. The generated XPS data were processed using the CasaXPS software. Atomic surface concentrations were determined by fitting the core spectra using Gaussian line shapes and a linear background and binding energies of the components in the spectra were determined by calibrating against the C-H/C-C peak in the C 1s spectra at 285.0 eV. The systematic error is estimated to be of the order of 5-10%. Electrochemical Grafting of 1. All electrochemical graftings were performed using 1 (Scheme 2) in 0.1 M Bu4NBF4/MeCN. First, one cyclic voltammogram was recorded at a sweep rate of 0.2 V s-1 to measure the peak potential (Ep) of the reduction wave pertaining to 1. Subsequently, the substrate was modified by carrying out potentiostatic electrolysis at a potential of Ep - 0.2 V for 300 s. After modification the substrate was rinsed thoroughly in MeCN followed by sonication in the same solvent for 10 min. Preparation of Initiator Layer. Hydroxyl-terminated substrates obtained from the electrografting of 1 on GC and SS were immersed in a dichloromethane solution containing different concentrations of 2-bromoisobutyryl bromide (0.007-0.5 M) and 0.05 M triethylamine (TEA) at room temperature for different reaction times (0.25-3 h). After the reaction was finished the substrates were rinsed thoroughly in dichloromethane followed by sonication in dichloromethane and acetone. The thickness of the initiator layer (in the dry state) was measured using an ellipsometer, the characteristic functional groups were detected using IRRAS, and the chemical composition of the layer was verified using XPS. SI-ATRP Procedure. A typical procedure for polymerization of methyl methacrylate (MMA) (entry 6, Table 3) includes the following steps. A 15 mL solution of MMA in MeCN (6.6 M) was added to a dried Schlenk flask together with the ligand, PMDETA (12.2 μL, 58.4 μmol). The mixture was degassed by only one freezepump-thaw cycle and backfilled with argon. The catalyst, CuICl (4.5 mg, 45.5 μmol), and the deactivator, CuIICl2 (1.8 mg, 13.4 μmol), were introduced followed by two additional freeze-pump-thaw cycles. After degassing, the mixture was sonicated for 5 min and stirred at 35 °C for half an hour to solubilize the catalyst. The 1072 DOI: 10.1021/la104125n

Table 1. Film Thicknesses Measured by Ellipsometry for the Initiator Layer on GC and SS substrate

d/nm

GC-1 GC-1Br SS-1 SS-1Br

4.4 (0.6) 6.8 (1.0) 5.5 (0.5) 6.5 (0.5)

plates grafted with initiator were immersed into the reaction mixture under argon flow. In each specific case a series of plates (typically 5-7) were exerted to varying polymerization times in the same reaction flask. Each sample extracted from the reaction medium at a given time was thoroughly rinsed and sonicated in the solvent for polymerization followed by acetone.

Results Initiator Layer. Creation of polymer brushes on surfaces through the SI-ATRP technique requires formation of a surface bound initiator layer. Our protocol for introducing covalently anchored initiators on conducting surfaces by means of the aryldiazonium approach is shown in Scheme 2. The first step is to carry out an electrografting of 1 under potentiostatic conditions (see Experimental Section and Supporting Information, SI 2.1) to form a covalently attached multilayered film. The covalent bond formation finds its origin in a surface directed attack of aryl radicals which are generated close to the surface upon reduction of 1 and expulsion of N2.22 In most cases the reactivity of the aryl radicals is so high that they also attack already grafted aryl groups in an uncontrolled process to form a multilayered aryl-based structure. These modified substrates are henceforth represented as GC-1 and SS-1 corresponding to GC and SS, respectively. Thicknesses of the films in the dry state, measured by means of ellipsometry, were found to be 4.4 and 5.5 nm for GC-1 and SS-1, respectively (Table 1). Using a length of 0.8 nm for phenylethanol (calculated from the ChemDraw Ultra 12.0 software; CambridgeSoft), these numbers would roughly reflect that 5-7 layers are present on the surfaces. The number of molecular layers may very well be underestimated, if as most likely the (22) Pinson, J.; Podvorica, F. Chem. Soc. Rev. 2005, 34, 429–39.

Langmuir 2011, 27(3), 1070–1078

Iruthayaraj et al.

Article

Table 2. Atomic Percentage, C/Br and O/Br Ratios, and Percentage of Conversion of Hydroxyl to Ester Groups for GC-1Br and SS-1Br Plates, % Conv, along with Atomic Percentage for PMMA-Modified Plates, GC-PMMA, at Varying Film Thicknesses entry

electroded

O 1s (atom %)

C 1s (atom %)

N 1s (atom %)

Br 3p (atom %)

O/Br ratio

C/Br ratio

% conv O/Bra

% conv C/Brb

c GC-1Br (0.25 h, 0.005 M) 7.3 86.3 5.7 0.04 182.5 2876 0.5 c GC-1Br (0.5 h, 0.05 M) 7.5 85.7 6 0.19 39.5 451.0 2.6 c GC-1Br (0.5 h, 0.07 M) 7.8 85.1 6.1 0.32 24.4 265.9 4.3 c GC-1Br (1 h, 0.3 M) 8.4 85.1 5.3 0.77 10.9 110.5 10.1 c GC-1Br (3 h, 0.5 M) 8.8 82.3 5.8 2.95 3.0 27.9 50 c SS-1Br (3 h, 0.5 M) 32.7 52.1 3.2 1.96 16.7 26.6 35 23.9 75.1 0.68 0.14 GC-PMMA9 nm e e GC-PMMA17 nm 25.4 74.6 e e GC-PMMA64 nm 25.6 74.4 a Calculated as % conv = 100%/(O/Br ratio - 1). b Calculated as % conv = 800%/(C/Br ratio - 4). c Not calculated for reasons given in the text. d Reaction time and concentration of 2-bromoisobutyryl bromide for the acylation in parentheses. e Not detectable.

1 2 3 4 5 6 7 8 9

attack of the aryl radicals takes place at the meta and ortho position of the already grafted aryl layers. To anchor the initiators in terms of tertiary alkanoyl bromides at the surface, acylation of the GC-1/SS-1 electrodes using 2-bromoisobutyryl bromide is carried out as shown in Scheme 2. We denote the resulting chemically modified substrates GC-1Br and SS-1Br. Previous studies have indicated that such acyl substitutions23 along with the somewhat related addition reactions involving isocyanates24 will not necessarily take place only in the outer part of the layer. In particular, the use of dichloromethane as solvent may swell the organic layers, hence letting the reactants penetrate further into the film. In fact, the thickness of GC-1Br and SS-1Br was measured to be 6.8 and 6.5 nm, respectively (Table 1), which corresponds to an increase of film thickness by 2.4 and 1.0 nm compared with GC-1 and SS-1, respectively. These values are much more than the expected value (∼0.4 nm) for an extension of the grafted layer with a 2-bromo-2-methylpropanoyl unit only. The large film expansion observed therefore indicates that quite an extensive number of the inner-layer hydroxyl groups have participated in the acylation reactions. The fact that the increase in film thickness is larger for GC-1Br than SS-1Br points at the same time to the presence of a substrate effect. The oxide layer on SS along with the high surface roughness presumably result in differences in the structure and porosity of the organic layer deposited on SS-1 and GC-1 during the electrografting procedure. To quantify the conversion percentage, % conv, of alcohol groups in the reaction with 2-bromoisobutyryl bromide, we measured the atomic distribution of the various elements present on GC-1Br and SS-1Br by means of X-ray photoelectron spectroscopy (XPS). It can easily be shown that % conv can be obtained from the O/Br ratio as % conv = 100%/(O/Br ratio - 1) and from C/Br as 800%/(C/Br ratio - 4). It is important to emphasize that these formulas assume that there are no contributions whatsoever from carbon and oxygen on the surfaces themselves. Obviously, this does not hold for the relatively thin layers we are studying, being the reason that the C/Br and O/Br ratios are not considered for GC and SS (protected by an oxide layer), respectively. XPS results for GC-1Br and SS-1Br are summarized in Table 2 along with the calculated conversion factors. (Additional data for the bare and the electrografted electrodes are available in Table S1 of the Supporting Information, SI 2.2.) Clearly, as the concentration of 2-bromoisobutyryl bromide is increased from 0.005 to 0.5 M and the reaction time from 0.25 to 3 h, the percentage of (23) Yu, S. S. C.; Tan, E. S. Q.; Jane, R. T.; Downard, A. J. Langmuir 2007, 23, 11074–82. (24) Ceccato, M.; Nielsen, L. T.; Iruthayaraj, J.; Hinge, M.; Pedersen, S. U.; Daasbjerg, K. Langmuir 2010, 26, 10812–21.

Langmuir 2011, 27(3), 1070–1078

Scheme 3. SI-ATRP of Methyl Methacrylate To Form Poly(methyl methacrylate) Using the CuICl/CuIICl2 Catalyst System and Either of the Three Ligands N,N,N0 ,N00 ,N00 -Pentamethyldiethylenetriamine (PMDETA), Tris[2-(dimethylamino)ethyl]amine (Me6TREN), or 2,20 -Bipyridine (BIPY)

conversion for GC-1Br increases from 0.5 to 50% (entries 1-5). It was also noted that reaction times above 3 h did not lead to further increase in the yield of bromine in the initiator layer. In the corresponding case of SS-1Br the maximal conversion factor was found to be close to 35%. These numbers support the conclusion from the thickness measurements, i.e., that a substantial part of the inner layer alcohol groups becomes acylated, taking into account that at least 5-7 molecular layers are present on the surface. On the assumption that the surface structure of GC is approximated with that of graphite the carbon surface density may be calculated as suggested by McCreery (Γgraphite = 7.3  10-9 mol cm-2).25 From this and the conversion factors listed in Table 2, an approximative value of the bromine surface density ensues (Γ = 3.6  10-10 mol cm-2), which is equal to or even below surface coverages typically obtained for self-assembled monolayers.4 It is worthwhile mentioning that an alternative and more direct approach for obtaining GC-1Br was attempted through a direct electrografting of compound 2, which in its structure incorporates both the diazonium and initiator functionalities (Scheme 2b). Somewhat surprisingly, the XPS spectrum of the grafted surface in this case showed no trace of bromine, and an attempted SIATRP did not succeed. We speculate that the lack of initiator is due to an efficient halogen abstraction reaction occurring between the tertiary alkanoyl bromide and the aryl radicals generated during the grafting process. In that respect, it is interesting to note that the bromine content also was found to be low (∼1 atom %) in the study of Matrab et al.,16 where a secondary benzylic bromide was used as initiator; the latter originated from the direct electrografting of the (1-bromoethyl)benzenediazonium salt. In comparison, ∼3 atom % surface concentration of bromine is achieved in the two-step approach used herein. SI-ATRP. With the initiator layer formed and characterized focus is now turned toward the SI-ATRP itself. The protocol for growing brushes on the surfaces is shown in Scheme 3. In all (25) Liu, Y.-C.; McCreery, R. L. J. Am. Chem. Soc. 1995, 117, 11254–59.

DOI: 10.1021/la104125n

1073

Article

Iruthayaraj et al.

Figure 1. IRRAS spectra recorded for (a) GC-PMMA (conditions: [MMA] = 6.6 M, [CuICl] = 0.003 M, [CuIICl2] = 0.0009 M, and [PMDETA] = 0.0039 M in MeCN at 35 °C) and (b) SS-PMMA (conditions: [MMA]=7.0 M, [CuICl] = 0.03 M, [CuIICl2] =0.003 M, and [PMDETA] = 0.033 M in toluene at 75 °C) corresponding to various polymerization times (20, 40, 80, 120, 160, and 200 min). The most intense spectra correspond to longest polymerization time. The 1275-1150 cm-1 band is assigned to C-O-C stretching in the ester,26 1480-1430 cm-1 to bending of -CH2- in the aliphatic chain, 3000-2900 cm-1 to the C-H symmetric and antisymmetric stretching of aliphatic -CH3 and -CH2, and finally 1740 cm-1 corresponds to the ester carbonyl stretch. Inset: linear correlation between the absorbance of the carbonyl group and ellipsometric thickness.

experiments the brush formation is accomplished by immersing GC-1Br or SS-1Br in the reaction medium containing MMA, solvent, CuICl/CuIICl2 catalyst system, and either of the three ligands PMDETA, Me6TREN, and BIPY. The PMMA-modified plates are denoted GC-PMMA and SS-PMMA, respectively. Spectroscopical Evidence. In Figure 1a,b IRRAS spectra for GC-PMMA and SS-PMMA recorded for samples at varying polymerization times are shown. The enhancement of the absorbance signal as a function of polymerization time on both substrates provides strong evidence to the occurrence of a controlled SI-ATRP reaction, the details of which will be better understood from a direct measurement of the thicknesses by means of ellipsometry (vide infra). Note that the inset in Figure 1 exhibits a linear correlation between the IRRAS absorbance of the ester carbonyl group and the film thickness obtained by ellipsometry. In Table 2, the atomic percentages obtained by means of XPS are included for three GC-PMMA electrodes at varying thicknesses (9, 17, and 64 nm). Of particular interest is the decrease seen in Br going from almost 3 atom % for GC-1Br (entry 5) before the polymerization to 0.14 atom % for GC-PMMA9 nm (entry 7) and 0 atom % for GC-PMMA17 nm (entry 8) and GC-PMMA64 nm (entry 9). Neither was chlorine detectable which would originate from the transfer of chlorine from CuCl2 to the polymer chain radical in Scheme 1. In principle, this lack of halogen could indicate that termination or transfer reactions had taken place in competition with the deactivation step to form the dormant Pn-X (Scheme 1). Although this may be true to some extent, a more likely explanation is that the long brushes which are straightened out in solution will fold during the drying process, offering the outer halogen group the possibility to be buried in the film to compensate for the loss of solvation energy. Still, it is surprising that no halogen can be detected at all, considering that the analysis depth of XPS is 10 nm. A simple but convincing argument for the still presence of halides is that film growth can be continued in further polymerization reactions, even if the film thickness exceeds the 17 nm, where no bromine is detectable by XPS. Moreover, reinitiation can be carried out on most films (vide infra). Effect of Initiator Density. The bromine density on GC-1Br was varied from 0.04 to 2.9 atom % by means of varying the 1074 DOI: 10.1021/la104125n

reaction time and the concentration of 2-bromoisobutyryl bromide (entries 1-5; Table 2). At this point this will allow us to investigate the effect the bromine density exerts on the resulting thickness of the polymer films. Two reaction media were selected, i.e., MeCN at 35 °C and toluene at 75 °C, using a polymerization time of 2 and 3.5 h, respectively. In Figure 2 the results obtained are shown for GC substrate. For both solvents at low bromine content (0.04-0.1 atom %), the polymer thickness remains almost constant. After reaching a critical bromine density, 0.16 atom % in the case of MeCN and 0.08 atom % in the case of toluene, the polymer thickness increases with an increase in bromine content. A peculiar behavior is noticed in the case of toluene, where an abrupt increase in the polymer thickness occurs at bromine content higher than 1.2 atom % to give a film as thick as 1 μm. Formation of thick films at high bromine content in toluene at 75 °C was likewise observed in the case of SS substrate. This should be compared to a corresponding thickness of 20 nm obtained in the case of MeCN. In fact, it can be seen from Figure 2c that the thickness ranges from 70 nm to 1.5 μm as the polymerization time goes from 1 to 5 h in toluene and, moreover, that a linear relationship between the two parameters is obtained. In all further studies described below, the highest possible initiator density, i.e. 2.9 atom % of bromine, was employed. Effect of Solvent. A number of solvents of varying polarity were included in the study such as toluene, dioxane, MeCN, and DMF. The purpose is to reveal the extent by which the polarity of the medium affects the overall reaction rate. In Table 3 all relevant experimental conditions are collected. Note that the first five entries of the reaction mixtures are heterogeneous because of a relatively low solubility of CuICl in these solvents. Parts a and b of Figure 3 show the film thickness versus polymerization time obtained in the four solvents on GC and SS, respectively. In all cases with the exception of toluene the tendency is that the thickness increases fast initially to attain an asymptotic value toward the end. This nonlinearity feature is consistent with a low controllability of the pertinent SI-ATRP processes which cease to exist because of the occurrence of termination reactions. It is also seen that the final thickness of the polymer film decreases with increase in solvent polarity going from toluene to dioxane Langmuir 2011, 27(3), 1070–1078

Iruthayaraj et al.

Article

Figure 2. Film thickness of dried films measured by ellipsometry versus atomic concentration of bromine on GC-1Br for the case of (a) MeCN at 35 °C (entry 6 in Table 3; vide infra), (b) toluene at 75 °C (entry 1 in Table 3; vide infra), and (c) thickness versus polymerization time in toluene at 75 °C for substrates containing 2.9 atom % of bromine (entry 1 in Table 3; vide infra). Table 3. Experimental Conditions Used for SI-ATRP of MMA on GC-1Br and SS-1Br entry

substrate

L

solvent

temp (°C)

[MMA] (M)

[CuICl] (M)

[CuIICl2] (M)

[L] (M)

1 2 3 4 5 6 7 8 9 10

GC-1Br SS-1Br GC-1Br GC-1Br SS-1Br GC-1Br GC-1Br GC-1Br GC-1Br GC-1Br

PMDETA PMDETA PMDETA PMDETA PMDETA PMDETA PMDETA PMDETA Me6TREN BIPY

toluene toluene dioxane MeCN MeCN MeCN DMF MeCN MeCN MeCN

75 75 40 40 40 35 35 35 35 35

7.0 7.0 6.3 6.3 6.3 6.6 7.5 7.5 7.5 7.5

0.03 0.03 0.03 0.03 0.03 0.003 0.03 0.003 0.003 0.003

0.003 0.003 0.003 0.003 0.003 0.0009 0.003 0.0009 0.0009 0.0009

0.033 0.033 0.033 0.033 0.033 0.0039 0.033 0.0039 0.0039 0.0078

and further to DMF and MeCN. This is so, even though the solubility of the catalyst/ligand system is lowest for the most nonpolar solvents such as toluene and dioxane. Scanning electron microscopy images of a few of these samples are collected in the Supporting Information (SI 2.3). Effect of Ligand. Figure 4 shows three plots of the polymer thickness against polymerization time corresponding to the three nitrogen-based ligands PMDETA, Me6TREN, and BIPY in MeCN on GC. A comparison of PMDETA with Me6TREN reveals that the initial growth rate is faster in the latter case, and as a consequence the overall growth exhibits much less controllability. In fact, the thickness quickly approaches an asymptotic value of (26) Socrates, G. Infrared and Raman Characteristic Group Frequencies Tables and Charts, 3rd ed.; Wiley: New York, 2004.

Langmuir 2011, 27(3), 1070–1078

32 nm due to the occurrence of chain termination processes. Use of PMDETA allows the formation of films of thicknesses larger than 60 nm. With BIPY as ligand no films are essentially produced at all, indicating that initiation hardly takes place in this case. Reinitiation of ATRP. For living radical processes reinitiation should be possible at least as long as the concentration of halide initiators at the surface is sufficiently large to support the generation of polymer brushes. As mentioned above, XPS is not an optimal technique for detecting the terminal halide groups in polymer brushes as they presumably to a large extent will be buried inside the layer in the dry state. Hence, it would be interesting to perform reinitiation experiments at various stages of the polymerization to detect indirectly the presence of halides but also to illustrate the possibility of making coblock polymers. DOI: 10.1021/la104125n

1075

Article

Iruthayaraj et al.

Figure 3. Plots of film thickness against polymerization time showing the effect of various solvents on the SI-ATRP process when using PMDETA as ligand for (a) GC and (b) SS substrates: toluene at 75 °C (b: entries 1 and 2 in Table 3), dioxane at 40 °C (O: entry 3 in Table 3), MeCN at 40 °C (1: entries 4 and 5 in Table 3), and DMF at 35 °C (Δ: entry 7 in Table 3).

Discussion

electrografting of an aryldiazonium salt is used as starting point for the growth of polymer brushes. Several parameters, such as initiator density, solvent polarity and ligand type have been varied to elucidate the controllability of the ATRP process under various conditions. Initiator Layer and Initiator Density. The effect of surface concentration of bromine on polymer thickness for the case of polar (MeCN) and nonpolar (toluene) solvent is shown in Figure 2. In both solvents the overall behavior is similar; i.e., at low bromine content the thickness values are constant, suggesting that the anchoring centers of the chains are far apart such that the chains grow in isolation. In other words, they do not overlap laterally, resulting in a mushroom/pancake-type growth. However, after reaching a certain critical bromine density (0.16 atom % in the case of MeCN and 0.08 atom % in the case of toluene), an increase in thickness is commenced. At this bromine content, we suggest that the anchoring points on the surface are close enough such that the growing chains start to stretch away from the interface to avoid overlapping, forming a brushlike conformation. Hence, the thickness becomes a function of the bromine content of the initiator layer. In the above discussion we have treated the bromine content of the initiator layer as equivalent to the graft density of brush. This assumption may not be fully correct owing to the plausible scenario that only a fraction of bromine atoms participate in surface initiation and/or a fraction of initiated bromine may be lost irreversibly during the polymerization process. However, qualitatively Figure 2 demonstrates that there is a change in the conformation of the polymer after a certain value of bromine concentration is reached, hence resembling the mushroom-to-brush crossover seen in the case of surface anchored polyacrylamide.27 Thick Film Regime. At high bromine density (1.2-2.9 atom %) very thick PMMA films, in the order of micrometers, are obtained using toluene as a solvent at 75 °C. Even in this thick film regime observed for polymerization times >1.2 h the polymer growth is well controlled as can be judged from the uniform growth rate of 450 nm h-1 (Figure 2c), which further is 4 times larger than the initial growth rate of 100 nm h-1 for polymerization times PMDETA > BIPY. The expected correlation for the SI-ATRP process is confirmed from the initial rates extracted from Figure 4 with the reactivity for Me6TREN, PMDETA, and BIPY in MeCN being estimated to be 60, 34, and 7 nm h-1, respectively. In fact, in the case of BIPY the surface initiation is so slow that it is useless for construction of large brushes. The higher reactivity induced by the PMDETA ligand is, on the other hand, sufficient to secure the formation of brushes, the lengths of which may vary from a few tens of nanometers until several micrometers, depending on the polymerization time. Going further to the Me6TREN ligand, even faster initial polymerization rates are observed because of the more negative oxidation potential of the catalyst CuICl/Me6TREN. The result is the creation of a high radical concentration at the surface with the thickness reaching quickly an asymptotic value because of the occurrence of termination processes. Hence, among the three ligands studied, PMDETA exhibits better control of SI-ATRP, illustrating that there exists a threshold radical concentration at the surface below which the initiation is too slow and above which the termination becomes prevalent. In addition to solvent and ligand the exact concentrations used of the copper salts are also very important for the polymerization rates and hence for exerting control on the ATRP process. One example illustrating this issue is provided by the plots for entries 4 (Figure 3) and 8 (Figure 4) in MeCN, in which the main difference is a 10 times lower concentration of Cu(I) in the latter case. At the same time the concentration of CuII is lower by a factor of 3. The overall result is a slower and a better controlled ATRP process giving thicker films in the end. Reinitiation. The negative effect of solvent polarity on thickness and the polymer growth rate is predominantly due to enhanced loss of halogen in polar solvents. This effect also comes fully through in the reinitiation experiments. Reinitiated thickness (29) Bergenudd, H.; Coullerez, G.; Jonsson, M.; Malmstrom, E. Macromolecules 2009, 42, 3302–08. (30) Qiu, J.; Matyjaszewski, K.; Thouin, L.; Amatore, C. Macromol. Chem. Phys. 2000, 201, 1625–31.

DOI: 10.1021/la104125n

1077

Article

Iruthayaraj et al.

was higher for those brush layers that were initially polymerized in a nonpolar medium. In the case of brushes synthesized in polar medium the ability to carry on the polymerization decreases as the initial polymer thickness is larger, and for thicknesses greater than 30 nm no reinitiation could be accomplished. It may seem peculiar that for polymerizations both initiated (