Anal. Chem. 2001, 73, 5596-5606
Temporal and Spatial Profiling of the Modification of an Electroactive Polymeric Interface Using Neutron Reflectivity Andrew Glidle,† Lee Bailey,‡ Charlotte S. Hadyoon,† A. Robert Hillman,*,‡ Angela Jackson,‡ Karl S. Ryder,†,§ Paul M. Saville,‡ Marcus J. Swann,† John R. P. Webster,⊥ Robert W. Wilson,‡,| and Jon M. Cooper†
Bioelectronics Research Group, Department of Electronics, Glasgow University, Glasgow, G12 8LT, U.K., Department of Chemistry, University of Leicester, Leicester, LE1 7RH, U.K., Department of Chemistry, De Montfort University, Leicester, LE1 9BH, U.K, ISIS Facility, Rutherford Appleton Laboratory, Didcot, Oxfordshire, OX11 0QX, U.K., and Institut Laue-Langevin, 38042 Grenoble Cedex 9, France
Electropolymerized films of the functionalized pyrrole, pentafluorophenyl-3-(pyrrol-1-yl)propionate (PFP), were reacted with a solution-phase nucleophile, ferrocene ethylamine. This reaction was chosen as a model representative of a postdeposition modification of the polymer membrane’s properties. For the first time, a nondestructive method for direct chemical analysis of the reaction profile within the electrodeposited polymer membrane after nucleophilic substitution is presented. This was achieved through the application of in situ neutron reflectivity with supplementary analytical information concerning the film’s chemical composition obtained from XPS, FT-IR, and electrochemical measurements. The results presented illustrate how, for a partially reacted film resulting from a short reaction time, the extent of reaction with ferrocene ethylamine is not homogeneous throughout the thickness of the film, but occurs predominantly at the polymer/solution interface. We show that the progress of the reaction within the polymer film is limited by the transport of reacting species in the dense regions of the membrane that are furthest from the solution interface. The data do not fit an alternative model in which there is spatially homogeneous progression of the reaction front throughout the bulk of the thin film polymer. Guided by the neutron reflectivity measurements, suitable modifications were made to the electrodeposition method to prepare films whose architecture resulted in faster rates of reaction. Spatial distribution of electroactive or molecular recognition sites is central to the dynamics and operational performance of modified electrodes, but their “vertical” distribution with respect to the interface has received surprisingly little attention. In this paper, we show how neutron reflectivity can be used to provide * To whom correspondence should be addressed. E-mail:
[email protected]. † Glasgow University. ‡ University of Leicester. § De Montfort University. ⊥ Rutherford Appleton Laboratory. | Institut Laue-Langevin.
5596 Analytical Chemistry, Vol. 73, No. 22, November 15, 2001
the spatial distribution of such sites introduced into a surfacebound polymer film by postdeposition substitution chemistry. The method has high vertical resolution, is nondestructive of the sample, and can be applied in situ. The technique has previously been used to determine spatial distributions of two (necessarily static) polymer components within a film1-3 and the film’s solvation profile.5-6 Now, for the first time, we use the technique to probe the propagation dynamics of a buried reaction profile front during polymer film modification by permeation of a mobile reactive entity. Although ex situ methods, for example, XPS, have been usefully employed to profile electroactive polymer film composition,7 we suggest that the nondestructive and in situ approach to analyzing film composition and internal microstructure presented here, “dynamic interfacial analysis”, will be of general importance in assessing the practical consequences of interfacial design strategies. Modification of electrode surfaces using an electroactive polymer film facilitates fabrication of sensor and transducing devices in which systematic variation of their chemical, optical, and electrical characteristics may be achieved.8 Alterations to the film structure can also be used to tune device performance, and consequently, such polymers, reviewed extensively elsewhere,9 have found application as solution sensors,10 gas sensors,11 energy (1) Hillman, A. R.; Saville, P. M.; Glidle, A.; Richardson, R. M.; Roser, S. J.; Swann, M. J.; Webster, J. R. P. J. Am. Chem. Soc. 1998, 120, 12882-12890. (2) Fernandez, M. L.; Higgins, J. S.; Penfold, J.; Ward, R. C.; Shackleton, C.; Walsh, D. J. Polymer 1988, 29, 1923-1928. (3) Fernandez, M. L.; Higgins, J. S.; Penfold, J.; Ward, R. C.; Shackleton, C.; Walsh, D. J. Polymer 1990, 31, 2146-2151. (4) Yim, H.; Kent, M.; McNamara, W. F.; Ivkov, R.; Satija, S.; Majewski, J. Macromolecules 1999, 32, 7932-7938. (5) Levicky, R.; Koneripalli, N.; Tirrell, M.; Ankner, J. F.; Kaiser, H.; Satija, S. K. Macromolecules 1998, 31, 4908-4914. (6) Kent, M. S. Macromol. Rapid Commun. 2000, 21, 243-270. (7) Abruna, H. D.; Denisevich, P.; Umana, M.; Meyer, T. J.; Murray, R. W. J. Am. Chem. Soc., 1981, 103, 1-5. (8) Osada, Y.; De Rossi, D. E. Polymer Sensors and Actuators; Springer-Verlag: London, 2000. Murray, R. W. Molecular Design of Electrode Surfaces; WileyInterscience: Toronto, 1992; Vol. 22. (9) Skotheim, T. A.; Elsenbaumer, R. L.; Reynolds, J. R. Handbook of Conducting Polymers, 2nd ed.; M. Dekker: New York, 1998. (10) Wang, J.; Jiang, M.; Mukherjee, B. Anal. Chem. 1999, 71, 4095-4099. (11) Slater, J. M.; Paynter, J.; Watt, E. J. Analyst 1993, 118, 379-384. 10.1021/ac0104882 CCC: $20.00
© 2001 American Chemical Society Published on Web 10/06/2001
Figure 1. Structure and reaction scheme for a poly(PFP) film reacting to form poly(Py-Fc).
storage devices,12 electrochromic devices,13 electronic muscles,14 bioelectronic interfaces,15 and biosensors.16 In all of these applications, film structure has a profound effect upon the properties of the modified electrode, particularly through an influence on the transport of species (molecules or electrons) into and within the film. An investigation into the structure-property relationship of these electrochemical systems permits a greater understanding of the factors that influence their behavior. In situ neutron reflectivity studies allow direct observation of the polymer’s structure (in terms of spatial distributions of the various components) in an environment appropriate to its ultimate use. Widespread applications of the technique, probing physical transformations of either buried or heterogeneous interfaces and thin films, have been recently reviewed.17,18 A notable advantage is that, in contrast to ellipsometric techniques, when using neutron reflectivity to probe solvated organic-based polymer systems, through judicious hydrogenation or deuteration of either the solvent or a component of the polymer, it is possible to alter dramatically the system’s reflectivity. Furthermore, as is demonstrated below, chemical replacement (through a designed reaction) of one moiety in a polymer by another with different neutron scattering properties can be readily probed in these reflectivity profiles. In this paper, the system under study is based on electropolymerization of the derivatizable pyrrole monomer, pentafluorophenyl-3-(pyrrol-1-yl)propionate (PFP)19 (Figure 1). PFP polymers and copolymers have the property that the initially prepared film may be transformed by reaction with solution-based (12) Otero, T. F.; Cantero, I.; Grande, H. Electrochim. Acta 1999, 44, 20532059. (13) Sapp, S. A.; Sotzing, G. A.; Reynolds, J. R. Chem. Mater. 1998, 10, 21012108. (14) Smela, E.; Inganas, O.; Lundstrom, I. Science 1995, 268, 1735-1738. (15) Schmidt, C. E.; Shastri, V. R.; Vacanti, J. P.; Langer, R. Proc. Natl. Acad. Sci. U.S.A. 1997, 94, 8948-8953. (16) Contractor, A. Q.; Sureshkumar, T. N.; Narayanan, R.; Sukeerthi, S.; Lal, R.; Srinivasa, R. S. Electrochim. Acta 1994, 39, 1321-1324. (17) Penfold, J.; Richardson, R. M.; Zarbakhsh, A.; Webster, J. R. P.; Bucknall, D. G.; Rennie, A. R.; Jones, R. A. L.; Cosgrove, T.; Thomas, R. K.; Higgins, J. S.; Fletcher, P. D. I.; Dickinson, E.; Roser, S. J.; McLure, I. A.; Hillman, A. R.; Richards, R. W.; Staples, E. J.; Burgess, A. N.; Simister, E. A.; White, J. W. J. Chem. Soc., Faraday Trans. 1997, 93, 3899-3917. (18) Sferrazza, M.; Jones, R. A. L.; Penfold, J.; Bucknall, D. B.; Webster, J. R. P. J. Mater. Chem. 2000, 10, 127-132. (19) Pickett, C. J.; Ryder, K. S. J. Chem. Soc., Dalton Trans. 1994, 2181-2189.
Figure 2. Illustrations depicting two possible modes for the progress of a reaction within a thin polymer film, dependent on the porosity of the film: left, a film with a homogeneous internal microstructure; right, a film whose internal microstructure comprises a network of pores and fissures that rapidly transport reacting materials throughout the film. Darker areas within the polymer correspond to regions of the film where reaction has occurred.
nucleophilic species under relatively mild conditions.19,20 The membrane’s physical integrity is not destroyed, and a new polymer functionalized with, for example, an electron transfer or bioactive motif can be formed. The PFP monomer is not unique in this respect, and other pyrroles containing activated ester groups have been employed, for example, the use of succinimide esters to graft oligonucleotide strands to polymer surfaces so as to form “gene” sensors.21,22 In this study, we are interested in discovering whether the nucleophilic substitution reaction progresses at a uniform rate throughout the film’s thickness or whether the layers of the film closest to the solution interface react faster than those layers closest to the substrate metal-electrode interface. Ferrocene ethylamine (FcNH2) is used as the probe nucleophile (leading to an amide coupled moiety) in this initial neutron reflectivity study because, in addition to the common use of the Fc+/0 couple as an electron-transfer mediator in (bio)catalytic systems, its immobilization can be characterized spectroscopically using FT-IR and XPS (via the Fe(2p) and F(1s) signals). It is anticipated that experience gained in this study will guide similar neutron reflectivity measurements to assist us in optimizing the covalent or affinity immobilization of biological macromolecules within pre(electro)deposited polymer films through modifying either the polymer deposition or solution reaction conditions. Notably, the slow transport of macromolecules within structurally dense immobilizing films can allow competitive side reactions with smaller species (e.g., water) to lead to less than optimally active sensors. Two possible simple models that describe the progress of the reaction within a polymer film are depicted in the illustration in Figure 2. In the first scenario, a reaction “front” starts at the outer (20) Hadyoon, C. S.; Glidle, A.; Morris, D. G.; Cooper, J. M. Chem. Commun. 1999, 1683-1684. (21) KorriYoussoufi, H.; Garnier, F.; Srivastava, P.; Godillot, P.; Yassar, A. J. Am. Chem. Soc. 1997, 119, 7388-7389. (22) Bidan, G.; Billon, M.; Livache, T.; Mathis, G.; Roget, A.; Torres-Rodriguez, L. M. Synth. Met. 1999, 102, 1363-1365.
Analytical Chemistry, Vol. 73, No. 22, November 15, 2001
5597
(solution) interface and progresses toward the substrate interface. In the second case, the polymer is composed of dense domains, uniformly distributed throughout the film thickness, interconnected by percolation pathways that facilitate the rapid transport of species from the bulk solution. This latter model permits the reaction to proceed at similar rates independent of the distance from the bulk solution interface. These two models represent two extremes (or limiting cases) of a broad spectrum of possible ways in which the spatial composition of the polymer film changes with increasing reaction time, and a more complex hybrid model may be applicable to a number of cases. One can draw the analogy between the present spatially variant chemical reaction and the spatially variant electrochemical processes involved in mediated charge transfer at modified electrodes. The latter was considered theoretically23,24 by exploring competing transport and kinetic processes via the dependence of electrode current-voltage responses on film and solution compositional parameters. The direct structural approach employed here has obvious attractions in terms of clarity. Thus, we report here, for the first time, how neutron reflectivity measurements can yield the depth to which the film has undergone a chemical reaction. The novel application of neutron reflectivity measurements in determining the reacted film’s compositional profile is underpinned by complementary quantitative spectroscopic and electroanalytical measurements, which explore different facets of the film’s microstructure and bulk or interfacial composition. The choice of FcNH2 as a probe facilitates assays using the quartz crystal microbalance, FT-IR, XPS and electrochemical measurements to give nonspatially resolved details of the amounts of different species within the film and interspecies bonding. XPS analyses of reacted polymer interfaces reveal the chemical nature of the species in the immediate vicinity of the exposed interface; although ex situ and destructive, they validate the neutron-derived interfacial composition. We conclude that neutron reflectivity measurements of multicomponent functionalized thin films represent a significant advance toward their nondestructive in situ characterization in terms of the spatial distributions of film components and locations of buried interfaces. EXPERIMENTAL METHODS The Neutron-Electrochemical Cell. The neutron-electrochemical cell was similar to that previously described.25 The cell incorporates a 20-nm thin-film Au working electrode prepared by metal evaporation onto a polished single-crystal quartz or silicon block (100 × 50 × 10 mm). For in situ grazing-angle neutron reflectivity measurements, the neutron beam passes through the quartz substrate before reflection occurs at the quartz/electrode/ polymer/solution interfaces. A large-area Pt gauze counterelectrode and Ag|AgCl (3 M NaCl) reference electrode are incorporated into the cell. Adherence of the Au electrode is promoted by the use of a “monolayer” surface modification of the substrate block using 3-mercapto-(propyl)trimethoxysilane.26 Polymer Electrode Fabrications and Treatments. Poly(PFP) was electropolymerized from 10 mM PFP + 0.2 M tetraethyl(23) Andrieux, C. P.; Dumas-Bouchiat, J. M.; Saveant, J. M. J. Electroanal. Chem. 1982, 131, 1-5. (24) Albery, W. J.; Hillman, A. R. J. Electroanal. Chem. 1984, 170, 27-49. (25) Wilson, R. W.; Cubitt, R.; Glidle, A.; Hillman, A. R.; Saville, P. M.; Vos, J. G. J. Electrochem. Soc. 1998, 145, 1454-1461. (26) Goss, C. A.; Charych, D. H.; Majda, M. Anal. Chem. 1991, 63, 85-88.
5598
Analytical Chemistry, Vol. 73, No. 22, November 15, 2001
ammonium perchlorate (TEAP) in dry acetonitrile (MeCN).19 The potential was cycled once from 0.00 V to +1.05 V (for nucleation) and then repeatedly (ca. 20 times) from 0.00 V to +1.00 V (for growth) at 20 mV s-1. Neutron reflectivity measurements of the unreacted films were performed in acetonitrile to characterize solvent volume fractions and determine polymer surface coverages without causing degradation of the film’s reactivity through solvolysis. Typically, a film was then immersed in a solution of 1 mM FcNH2 in dimethyl sulfoxide (DMSO) for a specified length of time so as to achieve a desired extent of polymer modification reaction (judged by trial RAIRS measurements). The reaction solution was then replaced with pure DMSO, and the filled cell was allowed to elute any unreacted species before final rinsing and filling with acetonitrile prior to further neutron reflectivity experiments. Neutron Instrumentation and Data Fitting. The principal neutron reflectivity experiments were carried out using the CRISP and SURF time-of-flight reflectometers at the ISIS spallation source of the Rutherford Appleton Laboratories, U.K. The chopperselected wavelength range was 0.05-0.65 nm and, by holding the sample at three different angles between 0.23 and 1.25° relative to the beam, an effective momentum transfer (q) range of 0.081.5 nm-1 was obtained. The collimation slits were set to define a beam footprint of 60 × 30 mm on the sample and q resolution of 3.5%. Reflectivity profiles were recorded for the bare electrode, the unreacted polymer, and the reacted polymer immersed in a number of neutron contrasts of both aqueous and nonaqueous solvents (in this paper, discussion is restricted to the measurements obtained when using acetonitrile as the solvent). Each reflectivity profile was corrected for transmission through the quartz or silicon block, and then the absolute reflectivity was calculated by scaling the critical edge (where present) to a reflectivity of 1. In the absence of critical reflectivity, the scale factor for reflectivity profiles was determined from measurements performed on the same polymer-coated substrate in which critical reflectivity had been observed during either previous or subsequent substrate characterization (e.g., when exposed to D2O or DMSO-d6 solutions). Fitting to the data was achieved using the Surface, JWMULF, and Parratt programs.27 Uncertainties in the fitted thickness and scattering-length density are estimated to be (1 nm and (0.2 × 1010 cm-2 for a particular model on the basis of the scatter in the parameters that produced equally good fits. The principal contributions to the uncertainties in any particular model arise from the statistical (counting) noise associated with data collection (0.1, coupled with agreement of the new model with other physical parameters of the polymer film that had been measured (e.g., using FT-IR, electrochemical, and XPS probes). The observed reflectivity fringes corresponding to both the polymer and gold layers mean that the goodness of fit parameter, χ2, can be very sensitive to variations of a few percent in some of the Nb, thickness, and roughness parameters of the layers used to describe a model of the film structure. Thus, although the range of model parameters describing the film structure is significantly limited, the quantitative precision of some aspects of the resulting description of the polymer structure is limited by uncertainties arising from the sample alignment and the model complexity available within the fitting software used. This leads to error bounds on the absolute position and size of the discrete inner layers of the model of
