Article pubs.acs.org/cm
General Method for Forming Micrometer-Scale Lateral Chemical Gradients in Polymer Brushes Hyung-Jun Koo, Kristopher V. Waynant, Chunjie Zhang, Richard T. Haasch, and Paul V. Braun* Department of Materials Science and Engineering, Beckman Institute, and Frederick Seitz Materials Research Laboratory, University of Illinois at Urbana−Champaign, Urbana, Illinois 61801, United States S Supporting Information *
ABSTRACT: We report a general diffusion based method to form micrometer-scale lateral chemical gradients in polymer brushes via selective alkylation. A quaternized brush gradient is derived from a concentration gradient of alkylating agent formed by diffusion in permeable media around a microchannel carrying the alkylating agent. Polymer brushes containing both charge and aromatic gradients are formed using the alkylating agents, methyl iodide and benzyl bromide, respectively. The gradients are quantitatively characterized by confocal Raman spectroscopy and qualitatively by fluorescence microscopy. The length and gradient strength can be controlled by varying the diffusion time, concentrations, and solvents of the alkylating agent solutions. This microfluidic brush gradient generation method enables formation of 2-D chemical potential gradients with a diversity of shapes.
■
INTRODUCTION Surfaces presenting chemical gradients, two-dimensional continuous variations of chemical properties, are of interest to provide a sample library with an extremely dense variable set, as a smart platform to direct transport of molecules, biomaterials, and even large objects such as water droplets, and as a substrate for directing cellular differentiation.1−9 Narrow chemical gradients with length scales of micrometers rather than millimeters are of interest for many of these applications; however, formation of microscale gradients, in particular in polymer brushes, remains challenging. Methods to form a surface chemical gradient have been developed based on liquid or vapor diffusion of self-assembling organosilane molecules, spatially varied surface reaction or post treatments, and microfluidic or electrochemical means.10−14 However, the gradients formed by these methods generally ranged from the millimeter to centimeter. Shorter period gradients have required sophisticated patterning methods which either are appropriate for only small areas (e.g., scanning probe) or require top-down patterning methods which are generally not compatible with polymer brushes (e.g., photolithography followed by reactive ion etching).15−18 Here, we report a simple, reliable, and general procedure to form lateral chemical gradients with lengths down to at least 100 μm, and perhaps shorter, in a polymer brush, via selective quaternization. A solution of an alkylating agent, e.g., methyl iodide, is injected into a microchannel formed from an alkylating agent permeable material, polydimethylsiloxane (PDMS) for the work reported here, in contact with a polymer brush coated substrate. The solution diffuses into the permeable channel wall forming a spatially and temporally varying concentration gradient of the alkylating agent above the © 2014 American Chemical Society
polymer brush, resulting in the gradient of the degree of the quaternization in the brush. Quantitative analysis by confocal Raman spectroscopy is used to investigate the effect of time, supporting solvent, and concentration of the alkylating agent solution on the gradient characteristics. Using a noncontinuous brush pattern, we validate that molecular diffusion into the permeable PDMS channel wall is the dominant mechanism for gradient formation, not diffusion within the brush or at the brush−PDMS interface. Along with methyl iodide, benzyl bromide is used confirming the generality of forming gradients by diffusion based alkylation reactions. Finally, patterning of the quaternized brush is demonstrated by using the microchannels with various shapes. Such complex patterns are particularly difficult to form via most gradient generation methods.
■
EXPERIMENTAL SECTION
Preparation of Polymer Brushes. A brush of poly(2(dimethylamino)ethylmethacrylate) (pDMAEMA), a polymer containing tertiary amine repeating units, was grown on a Si wafer substrate by surface-initiated atom transfer radical polymerization (ATRP). Glass slides or silicon wafers were first cleaned by a 30 min bath in Nano-Strip (Cyantek) at 75 °C; the substrates were then washed with deionized water produced by a Milli-Q Biocell System (R ≥ 18 MΩ cm) and dried in a nitrogen flow. The initiator, (11-(2bromo-2-methyl) propionyloxy) undecyltrichlorosilane, was synthesized by following a previously reported procedure.19 Following the printing of the initiator via μ-contact of a patterned or nonpatterned PDMS stamp, the substrates were rinsed with ethanol and dried in a nitrogen flow. The substrates with surface initiators now attached were Received: February 6, 2014 Revised: March 25, 2014 Published: March 27, 2014 2678
dx.doi.org/10.1021/cm5004388 | Chem. Mater. 2014, 26, 2678−2683
Chemistry of Materials
Article
Fluorescence images were obtained using an inverted fluorescence light microscope (Axiovert 200M, Carl Zeiss, Inc.). The brush thickness was measured by ellipsometry (632.8 nm single wavelength, Gaertner L116C, Gaertner Scientific Corp.), atomic force microscopy (MFP-3D, Asylum Research), and a profilometer (P-6, KLA Tencor Corp.). The quaternization ratio of the polymer brush was determined by X-ray photoelectron spectroscopy (XPS, Axis Ultra, Kratos Analytical).
placed in a reaction vessel and purged. Separately, 2-(dimethylamino) ethylmethacrylate (DMAEMA, 5.00 g, 0.03 mol, Sigma Aldrich), removed of inhibitor, was diluted with 20 mL of deionized water and 10 mL of isopropyl alcohol and degassed for at least 30 min. The solution was then added to a Schlenk flask under a positive flow of nitrogen. 1,1,4,7,10,10-Hexamethyltriethylene-tetramine as the ligand (90 μl, 0.33 mmol, Sigma Aldrich), CuBr (43 mg, 0.3 mmol, Sigma Aldrich), and CuBr2 (13 mg, 0.06 mmol, Sigma Aldrich) were added to the DMAEMA solution, and the mixture was sealed. The solution of the monomer, catalyst, and ligand was transferred via cannula into the reaction vessel containing the substrates with a preprinted initiator. After 10 h, the substrates were removed and rinsed with ethanol and deionized water and dried in a nitrogen flow. The thickness of the grown pDMAEMA brush measured by ellipsometry and atomic force microscopy varied between 70 and 100 nm. To compare the Raman spectra of the pDMAEMA quaternized by methyl iodide to an assynthesized quaternized brush, [2-(methacryloyloxy) ethyl] trimethylammonium iodide monomer was prepared according to the previous literature20 and polymerized in a similar fashion as above using a 9:1 MeOH/H2O solvent at 1 M concentration monomer, CuBr, CuBr2, and 2,2-bipyridine as a ligand in a 100:1:0.25:2 molar ratio. Gradient Formation in a Polymer Brush. Figure 1a illustrates the formation of a charge gradient in a polymer brush using what we
■
RESULT AND DISCUSSION Characterization of a Quaternization Gradient. The lateral distribution of the quaternization in the brush can be determined using confocal Raman spectroscopy. First, the Raman spectra of the pristine and quaternized pDMAEMA brushes were compared, to identify the characteristic peaks of each brush (Figure 2a). Pristine pDMAEMA has strong peaks
Figure 2. (a) Confocal Raman spectra of the pristine pDMAEMA, QpDMAEMA formed by quaternization with MeI, and Q-pDMAEMA synthesized from a quaternized monomer. (b) Confocal Raman spectra collected across the polymer brush gradient formed by MeIbased DRAPE for 10 min (see Supporting Information Figure S2 for high resolution spectra) using a ∼500-μm-wide channel. (c) Raman intensity profiles of the peaks at 2773 and 3450 cm−1 of the spectra of b as a function of lateral position. The pink band indicates the channel location.
Figure 1. (a) Schematic mechanism for polymer brush gradient formation by DRAPE. (b) Alkylation reactions of pDMAEMA brushes by the two alkylating agents, MeI and BnBr, during DRAPE. term “Diffusion of Reactive Agent into PErmeable media” (DRAPE). A PDMS slab containing an embedded microfluidic channel was fabricated by replica molding21 and placed on top of the pDMAEMA brush. The PDMS slab and the substrate were sandwiched between polycarbonate substrates and clamped. An alkylating agent solution was injected into the channel at 2 μL/min (a velocity within the channel of 0.95 mm/s) to maintain an effectively constant concentration of alkylating agent in the channel. The alkylating agent was diluted in isopropyl alcohol unless otherwise mentioned. The alkylating agent solution diffuses into the solvent permeable PDMS22 and forms a concentration gradient of the alkylating agent. The alkylating agents, methyl iodide (MeI), and benzyl bromide (BnBr), begin to quaternize the pDMAEMA brush upon contact, and the now quaternized brush (Q-pDMAEMA) is positively charged (Figure 1b). Since the quaternization rate is first order with respect to the quaternization agent concentration,23,24 the concentration gradient of the alkylating agent molecules formed by the diffusive permeation leads to a gradient in the quaternization, and thus a charge gradient in the polymer brush. After the gradient formation, the PDMS slab was peeled off, and the substrate with the quaternized brush gradient was rinsed consecutively with isopropyl alcohol and water. Instrumentation. All Raman measurements were collected using a Raman confocal imaging microscope (Horiba LabRAM HR 3D, Horiba). A 532 nm laser and a 50× 0.5 numerical aperture lens was used, providing a theoretical lateral resolution of 0.65 μm and vertical resolution of 4.3 μm (much larger than the polymer brush thickness).
at 2773, 2827, and 2950 cm−1, which are associated with C−H vibration modes of the tertiary amine groups (−CH2− N(CH3)2).25−27 After quaternization, the peak at 2773 cm−1 is reduced, and the new peaks at 2927, 3009, and 3450 cm−1 appear. The band at 3450 cm−1 is possibly associated with the O−H stretching vibration of water molecules adsorbed in the more polar quaternized brush.28 It is confirmed that the Raman spectrum of the pDMAEMA after the MeI quaternization is identical to that of the synthesized Q-pDMAEMA (Figure 2a). The quaternized brush gradient formed in the pDMAEMA brush by DRAPE for 10 min was characterized with scanning confocal Raman spectroscopy. Figure 2b shows the overlapped Raman spectra across the quaternized brush on pDMAEMA, perpendicular to the channel direction. A change in the peak intensities is observed around 1500 μm, which is where the ∼500-μm-wide channel was placed. The peak intensity at 2773 and 3450 cm−1 is plotted as a function of position (Figure 2c). Due to the quaternization of the pDMAEMA brush by the alkylating agent flowing in the channel, the intensity at 3450 cm−1 is higher and that at 2773 cm−1 is lower at the channel 2679
dx.doi.org/10.1021/cm5004388 | Chem. Mater. 2014, 26, 2678−2683
Chemistry of Materials
Article
location. The degree of the quaternization in the pDMAEMA brush can be estimated on the basis of the intensity change in these characteristic Raman peaks. Since the quaternization increases the molecular weight of the brush, the quaternization gradient also leads to a gradient in the thickness of the pDMAEMA brush (Figure S1). Estimation of the Quaternization of the Polymer Brush. The degree of quaternization of the pDMAEMA brush was estimated from the ratio of the Raman peaks at 2773 and 2950 cm−1 (Figure 3a) after calibration from the N1s XPS
Figure 4. N1s XPS spectra of pDMAEMA brushes as a function of the 0.1 M MeI treatment time: (a) 0 min (pristine), (b) 1 min, (c) 5 min, and (d) 30 min. The solid lines are the deconvoluted components for a tertiary amine (399 eV) and a quaternary ammonium compound (403 eV). The ratios of the quaternary ammonium peak area to the overall N1s peak area are (a) 12%, (b) 15%, (c) 26%, and (d) 53%.
net % of quaternized amines 53 − 12 = = ∼47% amine groups available 100 − 12
(1)
When the pDMAEMA was quaternized with a higher MeI concentration of 0.25 M for 60 min, the peak ratio in the XPS spectra remained almost identical with that in Figure 4d (Figure S3), indicating 47% is an effective upper bound to the degree of quaternization of the pDMAEMA brush. The degree of the quaternization (DoQ) of the MeI treated brush is estimated from the Raman data as
−1
Figure 3. (a) Raman intensity ratio of the peak at 2773 cm to that at 2950 cm−1 in Figure 2b. (b) Relative Raman spectra of the pDMAEMA brush as a function of alkylation reaction time with 0.25 M MeI. The red dotted line is at 2773 cm−1. (c) The relative Raman spectra intensity at 2773 cm−1 in b. (d) The calculated degree of quaternization in the brush treated by MeI based DRAPE for 10 min in Figure 2b. Pink bands in a and d indicate the channel location.
0.511 −
spectra (Figure 4). Even though the absolute peak intensity in the Raman spectra fluctuates, the ratio of the peaks at 2773 and 2950 cm−1 is almost constant in a nonreacted pDMAEMA brush (for example, 0∼500 μm in Figure 3a) and forms a smoothly varying curve in the gradient region. Raman spectra in Figure 3b were collected from a homogeneous pDMAEMA brush immersed in a 0.25 M MeI solution for varying times. The ratio of the peaks at 2773 cm−1 and 2950 cm−1 (Figure 3c) decreases for about 30 min and then becomes nearly constant. The peak intensity ratio is 0.042 at 60 min, which is assumed to be the maximum degree of quaternization. To convert this to a degree of quaternization, the Raman data were calibrated using the N1s XPS spectra. Figure 4 shows XPS spectra of the pDMAEMA brush treated with a 0.1 M MeI/IPA solution for different times. The peaks at 399 and 403 eV are assigned to tertiary amine (−N) and quaternary ammonium (−N+), respectively. The ratio of the −N+ peak area to the overall N1s peak area (−N + −N+ areas), is a direct measure of the fraction of nitrogen atoms in the pDMAEMA brush which are charged. However, the −N+/N1s ratio for nonquaternized pDMAEMA (Figure 4a) is 0.12, and so this ratio cannot directly be used as measure of quaternization (some amines are protonated under ambient conditions).29 Adjusted by the initial protonation, the effective degree of quaternization after MeI treatment for 30 min is
DoQ (%) =
I@2773 cm−1 I@2950 cm−1
0.511 − 0.042
× 47%
(2)
where 0.511 is the ratio of the Raman peaks at 2773 and 2950 cm−1 of the unquaternized brush, 0.042 is the ratio of the Raman peaks at 2773 and 2950 cm−1 of the maximally quaternized brush, and 47% is the maximum quaternization as determined by XPS. We know this equation is not rigorously true, as the intensity of the peak at 2773 cm−1 is similar for the 47% quaternized MeI treated Q-pDMAEMA and the synthesized Q-pDMAEMA, which we expect is 100% quaternized; however since the goal of this work is to demonstrate formation of a gradient and not a detailed study on polymer brush quaternization, we decided to not investigate this further. Figure 3d shows the estimated degree of the quaternization profile of the brush treated by MeI based DRAPE for 10 min in Figure 2b. The gradual change in the degree of the quaternization starting from both edges of the channel position and extending away from the channel indicates a ∼700 μm quaternization gradient in the pDMAEMA brush under this specific condition. The degree of quaternization also describes the charge gradient between the pDMAEMA brush (neutrally charged) and the quaternized pDMAEMA brush (positively charged). 2680
dx.doi.org/10.1021/cm5004388 | Chem. Mater. 2014, 26, 2678−2683
Chemistry of Materials
Article
Validation of DRAPE Mechanism. It is important to confirm that the alkylating agent is primarily permeating through the PDMS channel walls and not diffusing laterally in the brush or at the substrate-PDMS interface. We previously demonstrated that transport at a bare substrate-PDMS interface is greatly suppressed;30 however transport in the brush may still operate. To demonstrate that brush transport of the alkylating agent is not dominant, DRAPE was applied on a dot patterned pDMAEMA brush. If the alkylating agent primarily diffused through the brush, no gradient of the quaternized brush would be observed since the agent would not be able to travel between the isolated dots on the bare substrate.30 Figure 6a shows
Effect of DRAPE Treatment Time and Alkylating Agent Solution Concentration on Brush Gradients. Since the alkylating agent molecules diffuse further into the PDMS surrounding the channel with increasing time, the duration of DRAPE affects the gradient length. Figure 5a shows
Figure 5. Degree of quaternization as a function of (a) the DRAPE treatment time using 0.25 M MeI in isopropyl alcohol, (b) the MeI concentration in isopropyl alcohol using a DRAPE time of 10 min, and (c) alkylating solvents using 0.25 M MeI concentration after 10 min of DRAPE. The pink bands indicate the channel location.
Figure 6. (a) Fluorescence images of continuous and noncontinuous pDMAEMA brushes after 30 min of DRAPE. The brush was treated with aqueous HPTS solution and washed with deionized water before fluorescence imaging. The pink bands indicate the channel position. (b) The degree of quaternization and the normalized fluorescence intensity as a function of 0.1 M MeI treatment time. The degree of quaternization was calculated both from the XPS data and by using eq 2. For the fluorescence intensity measurement, the brushes were treated with 0.01 mM HPTS dye solution. The images on the top of the graphs visualize the fluorescence from the dye adsorbed QpDMAEMA under UV light. The pDMAEMA brush was treated with 0.1 M MeI for (i) 0, (ii) 1, (iii) 5, and (iv) 30 min. The fluorescence intensity was obtained from the gray scaled images with a green channel filter using ImageJ software.
the effect of DRAPE time on the quaternization gradient in the pDMAEMA brush. The DRAPE time was varied from 1 to 60 min during continuous flow of the MeI solution into the channel. As expected, the gradient length increases with DRAPE time, ranging from 500 μm after 1 min to 2 mm after 60 min. The effect of the alkylating agent concentration on the gradient shape was also evaluated (Figure 5b). The pDMAEMA brush was treated by DRAPE for 10 min with different MeI concentrations (0.05, 0.25, 1, 3 M in isopropyl alcohol). In the gradient region (above 250 μm), the degree of the quaternization increases with the MeI concentration. In the channel (below 250 μm), the brush appears nearly fully quaternized at concentrations above 0.25 M. The gradient curve using the 3 M solution is convex due to the saturation of quaternization even outside the channel region, which is different from other gradient curves which are generally concave. Thus, both the length and the shape of the brush gradient formed by DRAPE can be controlled by varying the DRAPE time and alkylating agent concentration.
fluorescence images of the resulting brushes with and without the dot patterns after 30 min of DRAPE. To visualize the quaternized region, the brush film was treated with an aqueous solution of 8-hydroxypyrene-1,3,6-trisulfonic acid (HPTS), a negatively charged fluorescent dye which selectively adsorbs in the positively charged quaternized brush. The samples were treated with a pH 9 buffer solution to enhance its fluorescence intensity31 and improve the selectivity of the dye adsorption.32 Since the pKa of the pDMAEMA is 7.5, the pDMAEMA brush is essentially uncharged under this condition, and the difference of the charge density between the pristine pDMAEMA and the 2681
dx.doi.org/10.1021/cm5004388 | Chem. Mater. 2014, 26, 2678−2683
Chemistry of Materials
Article
The Raman peak intensity profile at 1601, 2770, and 3057 cm−1 is shown in Figure 7b. The gradient length is ∼500 μm, which is ∼200 μm shorter than the gradient by MeI under similar conditions. Since the alkylation reaction rates by BnBr and MeI are almost identical,33 the shorter gradient presumably results from a lower diffusion coefficient of the larger BnBr molecule relative to MeI in PDMS. It is also likely that the degree of the pDMAEMA brush quaternization by the bulkier BnBr may be lower than for MeI because of BnBr’s greater steric interactions with the dense brush. The brush gradient moving away from the channel after BnBr based DRAPE presents two overlapping gradients of charge (positive to neutral) and capacity for π−π interaction (strong to weak). Gradient Patterning. A benefit of the DRAPE technique relative to many gradient generation methods is that the quaternized brush region can be patterned into many shapes by using different shaped channels. Figure 8a and b display the
quaternized pDMAEMA is distinct. Both fluorescent brushes with and without the dot patterns are observed up to ∼1500 μm from the channel, which is about the same as the gradient length on a similarly treated uniform brush film in Figure 5a. The gradients are similar, indicating that the diffusion of the alkylating agent into the permeable PDMS is the dominant process for the brush gradient formation in DRAPE. It should be noted that the fluorescence intensity from the adsorbed HPTS dye does not quantitatively represent the gradient of quaternized brushes. The degree of the brush quaternization and the fluorescence intensity of the adsorbed HPTS dye as a function of MeI treatment time are compared in Figure 6b. At >5 min of MeI treatment time, the fluorescence intensity saturates, perhaps due to self-quenching, and therefore does not correlate with the degree of quaternization. It is reasonable to assume the concentration distribution of the alkylating agent molecules in the PDMS would be influenced by solvent, since the solvent also enters the elastomer22 and permeation of the alkylating agent presumably changes with the solvent. To investigate solvent effects, the gradient profiles formed by chloroform, isopropyl alcohol (standard solution in this study), and dimethyl sulfoxide (DMSO) alkylating agent solutions were compared. Compared to isopropyl alcohol, chloroform swells PDMS more, whereas DMSO swells PDMS less.22 We hypothesized that a solvent which swells PDMS more results in faster transport of the alkylating agent solute into the channel walls. As shown in Figure 5c, the alkylating agent in chloroform forms a longer gradient than that in isopropyl alcohol, while a significantly shorter gradient is formed by the alkylating agent in DMSO. The different degree of quaternization in the channel region (0∼250 μm) results from the solvent dependent alkylation rate.33 The alkylation reaction is faster in a more polar solvent (e.g., DMSO) than in a less polar one (e.g., chloroform). The fact that the permeability of the solvent into the PDMS affects the gradient length is another parameter for controlling the gradient shape and length. Quaternization Gradient Formation by a Different Alkylating Agent. Along with MeI, other alkylating agents could also form a quaternized brush gradient via DRAPE. BnBr was chosen as an example of an aromatic alkylating agent. Figure 7a displays confocal Raman spectra across the brush gradient generated by BnBr. Besides the changes in the C−H vibration peaks at 2770 (aliphatic) and 3057 cm−1 (aromatic),27 a change in the intensity of another characteristic peak at 1601 cm−1, associated with an aromatic ring vibration, confirming quaternization of the pDMAEMA brush by BnBr, is observed.25
Figure 8. (a) Fluorescence images of patterns of the quaternized pDMAEMA brush formed by DRAPE. The samples were quaternized (i) by 0.25 M MeI in DMSO for 1 min and (ii) by 0.25 M MeI in isopropyl alcohol for 30 min, respectively. (b) Fluorescence images of quaternized brush patterns formed by microfluidic channels with various shapes. Samples were quaternized by 0.25 M MeI in DMSO for 1 min. Samples in a and b were treated with a HPTS fluorescent dye solution after quaternization. (c.i,ii) 2-D mapping of the Raman peak intensity at 2773 cm−1 of the area enclosed by dotted lines in a.i and a.ii, correspondingly. (d) Relative Raman intensity profiles along the dotted lines in c and for a sample treated using 0.25 M MeI in DMSO for 30 min. See Supporting Information Figure S4c for the magnified plots of d.
fluorescence images of the quaternized brush patterns after the adsorption of the HPTS dye. As was the case for the line pattern (Figure 5), the gradient length formed along the patterns can also be controlled by varying the DRAPE time and solvent. Two fluorescence images shown in Figure 8a compare samples quaternized by 0.25 M MeI in DMSO for 1 min and by 0.25 M MeI in isopropyl alcohol for 30 min, clearly illustrating the longer gradient length formed under the latter condition. 2D mapping of the relative Raman peak intensity at the characteristic wavenumbers (2773, 3009, and 3450 cm−1) reveals the spatial distribution of the quaternized brush (Figure 8c and Figure S4a and b), confirming that the fluorescent region in Figure 8a corresponds to the quaternized region. Quantitative analysis of the gradient formation is performed by comparing the relative Raman peak intensity profiles at 2773 cm−1 across the circle pattern under three different solvents and quaternization time conditions (Figure 8d). The gradient
Figure 7. (a) Confocal Raman spectra across the polymer brush gradient formed by BnBr based DRAPE for 10 min. (b) Raman intensity profiles of the peaks at 1601, 2770, and 3057 cm−1 in the spectra of a. The pink band indicates the channel location. 2682
dx.doi.org/10.1021/cm5004388 | Chem. Mater. 2014, 26, 2678−2683
Chemistry of Materials
Article
length ranges from about 100 to 1000 μm. Although a channel cannot be truly a circle, effective radial gradients can be realized via the DRAPE technique through appropriate pattern design (Figure 8a.ii).
(5) Ross, A. M.; Lahann, J. J. Polym. Sci. Pol. Phys. 2013, 51, 775−794. (6) Meier, B.; Zielinski, A.; Weber, C.; Arcizet, D.; Youssef, S.; Franosch, T.; Rädler, J. O.; Heinrich, D. Proc. Natl. Acad. Sci. U. S. A. 2011, 108, 11417. (7) (a) Chaudhury, M. K.; Whitesides, G. M. Science 1992, 256, 1539−1541. (b) Hernández, S. C.; Bennett, C. J. C.; Junkermeier, C. E.; Tsoi, S. D.; Bezares, F. J.; Stine, R.; Robinson, J. T.; Lock, E. H.; Boris, D. R.; Pate, B. D.; Caldwell, J. D.; Reinecke, T. L.; Sheehan, P. E.; Walton, S. G. ACS Nano 2013, 7, 4746−4755. (8) Kapur, T. A.; Shoichet, M. S. J. Biomater. Sci., Polym. Ed. 2003, 14, 383−394. (9) (a) Montcouquiol, M.; Kelley, M. W. J. Neurosci. 2003, 23, 9469− 9478. (b) Sharma, R. I.; Snedeker, J. G. Biomaterials 2010, 31, 7695− 7704. (10) (a) Genzer, J.; Bhat, R. R. Langmuir 2008, 24, 2294−2317. (b) Wu, T.; Efimenko, K.; Genzer, J. J. Am. Chem. Soc. 2002, 124, 9394−9395. (11) (a) Vasilev, K.; Mierczynska, A.; Hook, A. L.; Chan, J.; Voelcker, N. H.; Short, R. D. Biomaterials 2010, 31, 392−397. (b) Harding, F.; Goreham, R.; Short, R.; Vasilev, K.; Voelcker, N. H. Adv. Healthcare Mater. 2013, 2, 585−590. (12) Jeon, N. L.; Dertinger, S. K. W.; Chiu, D. T.; Choi, I. S.; Stroock, A. D.; Whitesides, G. M. Langmuir 2000, 16, 8311−8316. (13) Krabbenborg, S. O.; Nicosia, C.; Chen, P.; Huskens, J. Nat. Commun. 2013, 4, 1667. (14) Hansen, T. S.; Lind, J. U.; Daugaard, A. E.; Hvilsted, S.; Andresen, T. L.; Larsen, N. B. Langmuir 2010, 26, 16171−16177. (15) Zamborini, F. P.; Crooks, R. M. J. Am. Chem. Soc. 1998, 120, 9700−9701. (16) Piner, R. D.; Zhu, J.; Xu, F.; Hong, S.; Mirkin, C. A. Science 1999, 283, 661−663. (17) Fuierer, R. R.; Carroll, R. L.; Feldheim, D. L.; Gorman, C. B. Adv. Mater. 2002, 14, 154−157. (18) Choi, D.-G.; Yu, H. K.; Jang, S. G.; Yang, S.-M. J. Am. Chem. Soc. 2004, 126, 7019−7025. (19) Matyjaszewski, K.; Miller, P. J.; Shukla, N.; Immaraporn, B.; Gelman, A.; Luokala, B. B.; Siclovan, T. M.; Kickelbick, G.; Valant, T.; Hoffmann, H.; Pakula, T. Macromolecules 1999, 32, 8716−8724. (20) Xu, X.; Wang, Y.; Liao, S.; Wen, Z. T.; Fan, Y. J. Biomed. Mater. Res., Part B 2012, 100B, 1151−1162. (21) Chang, S. T.; Uçar, A. B.; Swindlehurst, G. R.; Bradley, R. O.; Renk, F. J.; Velev, O. D. Adv. Mater. 2009, 21, 2803−2807. (22) Lee, J. N.; Park, C.; Whitesides, G. M. Anal. Chem. 2003, 75, 6544−6554. (23) Brown, H. C.; Cahn, A. J. Am. Chem. Soc. 1955, 77, 1715−1723. (24) Chovino, C.; Gramain, P. Macromolecules 1998, 31, 7111−7114. (25) Roy, D.; Guthrie, J. T.; Perrier, S. Soft Matter 2008, 4, 145−155. (26) Liu, G.; Wu, D.; Ma, C.; Zhang, G.; Wang, H.; Yang, S. ChemPhysChem 2007, 8, 2254−2259. (27) Bower, D. I.; Maddams., W. F. The Vibrational Spectroscopy of Polymers; Cambridge University Press: New York, 1989. (28) Jiang, Y.-X.; Li, J.-F.; Wu, D.-Y.; Yang, Z.-L.; Ren, B.; Hu, J.-W.; Chow, Y. L.; Tian, Z.-Q. Chem. Commun. 2007, 0, 4608−4610. (29) Karamdoust, S.; Yu, B.; Bonduelle, C. V.; Liu, Y.; Davidson, G.; Stojcevic, G.; Yang, J.; Lau, W. M.; Gillies, E. R. J. Mater. Chem. 2012, 22, 4881−4889. (30) Yonet-Tanyeri, N.; Evans, R. C.; Tu, H.; Braun, P. V. Adv. Mater. 2011, 23, 1739−1743. (31) Wolfbeis, O. S.; Herbert Kroneis, E. F.; Marsoner, H. Fresenius' Z. Anal. Chem. 1983, 314, 119−124. (32) Jeong, J. H.; Kim, S. W.; Park, T. G. Prog. Polym. Sci. 2007, 32, 1239−1274. (33) Friedli, F. E. Rate of Quaternization as a Function of the Alkylating Agent, Proc. World Conf. Oleochem; Applewhite, T. H., Ed.; American Oil Chemists’ Society: Champaign, IL, 1990; pp 296−297.
■
CONCLUSION In summary, we developed a facile and reliable procedure we term DRAPE for gradient formation in polymer brushes which can generate gradients as narrow as 100 μm. An alkylating agent solution, e.g., containing MeI, flowing in a microchannel diffuses into a permeable channel wall intrinsically forming a concentration gradient. This concentration gradient results in the formation of a gradient in the degree of quaternization in the underlying polymer brush. We characterize the charge gradient in a selectively quaternized polymer brush by using confocal Raman spectroscopy and XPS and confirm that the gradient length and gradient profile can be manipulated by controlling the DRAPE process time and the concentration of the alkylating agent solution. Diffusion of the alkylating agent solution into the permeable channel wall is found to be the dominant mechanism behind the DRAPE method, based on the results of the gradient formation in an unconnected brush pattern and the effect of the solvent interactions with the channel material on gradient length. Another alkylating agent, BnBr, is used to realize a gradient in both π−π interaction strength and charge in a brush. The DRAPE method is capable of patterning a quaternized brush by using the microchannels with various shapes, which is visualized by fluorescence microscopy and confocal Raman spectroscopy. The DRAPE procedure is simple and scalable, and the resulting gradient is reliably controllable. We expect DRAPE will be compatible with other chemical modifications, e.g., hydrolysis, and propargyl functionalization for click chemistry, as long as the reactive molecules are able to diffuse into the walls of the microchannel. The capabilities of DRAPE potentially enable elaborate patterns and precise control of molecular transport, e.g., directed translation, concentration, or separation, driven by the chemical potential gradients formed in polymer brushes.
■
ASSOCIATED CONTENT
S Supporting Information *
Thickness profile on a brush gradient and additional Raman and XPS spectra. This material is available free of charge via the Internet at http://pubs.acs.org.
■
AUTHOR INFORMATION
Corresponding Author
*E-mail:
[email protected] Notes
The authors declare no competing financial interests.
■
ACKNOWLEDGMENTS The work was supported by Defense Threat Reduction Agency under award number HDTRA 1-12-1-0035. The authors thank Dr. Brian Pate (DTRA) for helpful discussions.
■
REFERENCES
(1) Sheehan, P. E.; Whitman, L. J. Nano Lett. 2005, 5, 803−807. (2) Jayaraman, S.; Hillier, A. C. J. Comb. Chem. 2003, 6, 27−31. (3) Simon, C. G.; Lin-Gibson, S. Adv. Mater. 2011, 23, 369−387. (4) Perl, A.; Gomez-Casado, A.; Thompson, D.; Dam, H. H.; Jonkheijm, P.; Reinhoudt, D. N.; Huskens, J. Nat. Chem. 2011, 3, 317−322. 2683
dx.doi.org/10.1021/cm5004388 | Chem. Mater. 2014, 26, 2678−2683