Letter pubs.acs.org/macroletters
Signal Enhanced FTIR Analysis of Alignment in NAFION Thin Films at SiO2 and Au Interfaces Tawanda J. Zimudzi and Michael A. Hickner* Department of Material Science and Engineering, The Pennsylvania State University, University Park, Pennsylvania 16802, United States S Supporting Information *
ABSTRACT: Spin-cast NAFION samples were prepared on silicon native oxide and gold substrates with film thicknesses ranging from 5 to 250 nm. The influence of NAFION film thickness on the infrared spectrum of the polymer was investigated in substrate overlayer attenuated total reflection (SO-ATR) geometry at incident angles between 60° and 65°. In the grazing angle SO-ATR geometry, the thickness of the film significantly affected the position and absorbance of characteristic peaks in the FTIR spectrum of NAFION. Two major peaks in the NAFION spectrum at 1220 cm−1 (predominantly vas(CF2) and vas(SO3−)) and 1150 cm−1 (predominantly vas(CF2)) appeared to systematically blueshift to 1256 and 1170 cm−1, respectively, as the thickness of the film decreased from 250 to 5 nm. The changes in the NAFION thin film FTIR spectrum can be attributed to two factors; (1) ordering of NAFION at the interface during spin coating and film formation and (2) the increase in the ppolarization character of the infrared evanescent wave as the polymer film became thinner between the internal reflection element and the film substrate overlayer. The increase in p-polarization resulted in an increase in characteristic peak absorbances of dipoles aligned normal to the substrate due to the overlayer enhancement of the electric field with NAFION films on Si or Au film substrates. These results show that the specific thin film sampling geometry, especially in internal reflection experiments, must be considered to rationally quantify changes in NAFION thin film infrared spectra.
I
in addition to effects of the adsorbed sulfonate groups. FTIR experiments conducted using a Pt electrode coated with a thin NAFION film assembled into a spectroelectrochemical cell11 showed that the doubly degenerate vas(SO3−) band at 1250 cm−1 dominates the spectrum of NAFION on Pt. This 1250 cm−1 peak combined with the observations that vs(SO3−) at 1060 cm−1 and v(−SO3H) at 1400 cm−1 changed significantly with potential indicated that only the −SO3− group participated in the adsorption process. NAFION adopts a multilamellar structure on silicon oxide surfaces and the morphology of the thin films can vary with thickness and processing conditions.12,13 On native silicon oxide surfaces, NAFION may interact with the surface by hydrogen bonding between its sulfonate groups and the surface silanol groups, but chemical evidence of this bond has not been unambiguously established. Several methods have been used to study ionomer/surface interactions, including vibrational techniques such as surface enhanced Raman and FTIR spectroscopies.14,15 FTIR has become a popular method of analyzing thin films due to the ease of sample preparation and speed of analysis. The study of thin films using vibrational spectroscopy remains difficult, as few sampling geometries provide adequate sensitivity or high
on-containing polymers like NAFION play an important role in applications including water purification, as separators and electrode binders in electrochemical cells, in sensors, and as solid-phase catalysts in the production of chemicals.1−3 These materials receive particular attention in the field of energy storage and conversion in devices such as solarfuel generators,4 batteries,5 and fuel cells6 where a thin ionomer film is responsible for the transport of ions to and from active electrocatalytic sites. In fuel cells, the ion-conducting polymer in the catalytic layer provides mechanical strength, can influence transport in the porous electrode and binding of catalyst particles, and may play a significant role in determining the activity of the catalyst.7 NAFION interacts with metallic surfaces anchored by oxygen atoms of the sulfonate (−SO3−) group. The interaction of sulfonate with metallic surfaces appears to be similar to that of sulfate groups although there may be additional acidity effects.8 Polarization modulation-infrared reflection absorption spectroscopy (PM-IRRAS) has been used by Kendrick et al.9 to study the interface between NAFION and a Pt catalyst in operando. Based on changes in the 1260 cm−1 peak intensity, the authors concluded that vas(CF3) mode is mechanically coupled to the sulfonate pure mode vas(SO3−) and −CF3 groups on the side chain coadsorb with the sulfonate groups.9,10 Kendrick’s observations show that the adsorption of NAFION on surfaces may involve reordering of the hydrophobic backbone, which can sterically block catalytically active sites © 2015 American Chemical Society
Received: November 9, 2015 Accepted: December 18, 2015 Published: December 22, 2015 83
DOI: 10.1021/acsmacrolett.5b00800 ACS Macro Lett. 2016, 5, 83−87
Letter
ACS Macro Letters
Figure 1. (a) Experimental geometry for SO-ATR measurements where the evanescent wave is defined in three components: one in the x direction, one in the y direction, and one in the z direction. (b) Propagation of evanescent wave in a three-layer system of a typical SO-ATR experiment showing the reflected and transmitted Fresnel factors.
Figure 2. (a) Enhancement of p-polarized electric field in SO-ATR shown by the higher p-polarized p-factor. (b) Polymer thickness dependence of enhancement for a NAFION film with a silicon substrate in SO-ATR. (c) Comparison of SO-ATR absorbance values with those in transmission on a silicon substrate, and (d) comparison of SO-ATR and IRRAS absorbance values on a gold substrate.
signal-to-noise ratio. Infrared reflection absorption spectroscopy (IRRAS) at near normal incident angles has thus far remained the most feasible method for studying films down to monolayer thicknesses. IRRAS, however, is limited to metallic substrates. Surface overlayer ATR (SO-ATR) not only provides sensitivity rivaling or exceeding IRRAS but also provides a superior signal-to-noise ratio even on nonmetallic substrates. SO-ATR is an internal reflection FTIR technique carried out at grazing incident angles with high refractive index ATR crystals in a crystal/sample/substrate geometry which yields sensitivity significantly greater than in external reflection and transmission geometries. Here, an optical model similar to that established by Milosevic16,17 was employed to demonstrate the increase in electric field strength and p-polarization for a system comprising a Ge ATR crystal, a spin-coated NAFION thin film and Au or native oxide Si film substrate as the overlayer.
The results of these calculations were then used to explain the peak shifts observed at 1220 and 1150 cm−1 in SO-ATR and to develop a detailed picture of NAFION adsorption to native oxide Si and Au surfaces. In SO-ATR, the experimental geometry consists of a germanium hemisphere, the sample, and a high refractive index overlayer with the infrared beam at an incident angle of 60−65°, Figure 1a. An evanescent wave propagates through the sample and can be divided into x, y, and z components (Figure 1a), with the y component constituting s-polarized electric field and the sum of x and z components constituting the ppolarized electric field. Unlike conventional ATR, however, the z component of the electric field is significantly enhanced by what may be thought of as constructive interference between the incident wave and reflected waves from the sample overlayer interface, Figure 1b. A three-layer optical model can 84
DOI: 10.1021/acsmacrolett.5b00800 ACS Macro Lett. 2016, 5, 83−87
Letter
ACS Macro Letters
Figure 3. (a) Spectra of bulk NAFION membrane several microns thick in ATR geometry showing little difference in s- and p-polarized spectra. (b) 200 nm thick film in SO-ATR geometry showing little difference in s- and p-polarized spectra. (c) 50 nm thick film in SO-ATR geometry showing peak shifts due to molecular ordering. (d) Comparison of normalized SO-ATR spectra of 50 and 5 nm thick films in absence of external polarizer showing peak shifts due to intrinsic enhancement of p-polarization and molecular orientation.
transmission or reflection experiments. For films with thicknesses greater than 150 nm, SO-ATR does not offer any advantage in terms of signal over transmission or IRRAS. However, for very thin films of 50 nm and below, SO-ATR provides a significant signal enhancement enabling detailed spectral analysis, which would otherwise be problematic. From Figure 2c,d it can be concluded that quantitative analysis with SO-ATR and comparison of the spectra with other IR sampling geometries such as transmission or IRRAS are challenging without understanding the ramifications of sampling geometry, enhancement, and polarization in these different techniques. The intrinsic p-polarization in SO-ATR will be referred to as z-polarization to distinguish it from p-polarization that results from the addition of a physical polarizer in the experimental setup. The implications of the z-polarization and signal enhancement on structure determination are demonstrated here with the use of thin NAFION films. For unoriented films, the ATR spectrum will be nearly identical in all polarizations in terms of peak positions and ratios.20 This is demonstrated in Figure 3a for an isotropic bulk NAFION membrane where all the peak positions are identical, as are the peak ratios. The difference in absorbance values can easily be explained by the lower electric field intensity under s-polarization. The SO-ATR spectrum of thick films is remarkably similar to the ATR spectra of isotropic films, Figure 3b, due to the rapid decay of the z electric field intensity and the low proportion of ordered molecules relative to unordered molecules. For thin films there
be used to describe the strength of each of these components based on electromagnetic theory and the Fresnel equations.17−21 These derivations for the SO-ATR geometry are described in detail in the Supporting Information and are used to determine p-factors that are directly related to the electric field intensity for each component of the field. Figure 2a shows the p-factors for p-polarized and s-polarized electric fields in a 50 nm thick NAFION film sampled in SO-ATR geometry with a silicon substrate as the overlayer. This plot demonstrates that, in this geometry, the p-polarized electric field is almost 10× higher than s-polarized field, even in the absence of a polarizer. The enhancement is highly dependent on the thickness of the polymer film, and as the film becomes thicker, the enhancement decreases due to the exponential decay of the evanescent wave through the film. This effect is demonstrated in Figure 2b by calculating the absorbance as a function of film thickness for a NAFION film at 1050 cm−1 with a silicon wafer overlayer. It is apparent from the calculated curve in Figure 2b that the absorbance increases with thickness until the sample is 25 nm thick due to a large enhancement factor, after which for thicker films the absorbance begins to decrease. This trend is confirmed with experimental measurements that deviate from the calculated values, possibly due to the complexity of the system and alternative enhancement mechanisms, such as phonon resonance on silicon overlayers.22,23 Similar trends in enhancement in the SO-ATR geometry for thin films on Si, Figure 2c, and gold, Figure 2d, are observed compared to 85
DOI: 10.1021/acsmacrolett.5b00800 ACS Macro Lett. 2016, 5, 83−87
Letter
ACS Macro Letters
thickness of the polymer layer decreases. The increased absorbance of the 1220 cm−1 band as p-polarization increases suggests that −CF2 groups become locked in a conformation where the vas(CF2) dipole is in the plane of the p-polarized electric field, which is consistent with the observations of Kendrick et al. in PM-IRRAS experiments. No difference was observed in this trend on both gold and silicon substrates, suggesting that the spin-casting process had more influence on orientation than the substrate, Figure S4. Above 50 nm, the unpolarized peak position corresponds to bulk isotropic films,26 which could be interpreted as evidence for a morphological transition at this thickness, as reported in the literature.27 While this is a likely factor in the peak position the changes in contribution of the x component in the unpolarized beam cannot be ruled out. The appearance of the thin film SO-ATR spectra can be explained by considering the orientation of NAFION chains, Figure 5. With backbone chains oriented
is significant z-polarization enhancement and a higher proportion of ordered molecules in the film resulting in peak position and ratio differences when comparing s- and ppolarized spectra, Figure 3c. The s-polarized peak at 1090 cm−1 originates from the vas (Si−O−Si) of the silicon substrate.24,25 It is interesting to note that the peak positions observed in ppolarized spectra for films with thicknesses of less than 20 nm are remarkably similar to what is observed in PM-IRRAS experiments.9 Unpolarized SO-ATR thin film spectra closely resemble thin film PM-IRRAS spectra, while thicker films resemble conventional ATR or transmission spectra, Figure 3d. For thin films below 20 nm on silicon, which is often employed as a model substrate, SO-ATR provides a routine method of obtaining IR spectra that would otherwise be challenging in other experimental geometries. The changes in the film structure of NAFION on silicon can be monitored by observing the change in peak position of the broad envelope between 1203 and 1257 cm−1, shown in Figure 4. As the film becomes thinner, the 1203−1257 cm−1, peak
Figure 5. Ordered conformation of NAFION on surface for thin films, where the backbone vas(CF2), side chain vs(SO3−), and vas(CF3) dipole changes are in the plane of the p-polarized z component of the evanescent wave.
parallel to the substrate, the backbone vas(CF2) is in the plane of the z-polarized electric field. Computational studies suggest that when the backbone is parallel to the surface the CF3 groups are oriented such that the vas(CF3) is in the z-polarized electric field plane which explains the 1256 cm−1 peak position.28 Vibrational studies have also demonstrated the SO3− group adsorbs to gold surfaces, which is consistent with the SO-ATR spectra reported here as the vas(SO3) is out of the z-polarized electric field plane.9,29 In conclusion, we demonstrate that the film thickness and polarization of SO-ATR experiments influence the observed spectrum of NAFION thin films and that with an understanding of these effects SO-ATR is a powerful method to obtain orientation information for thin films. From SO-ATR we demonstrate that silicon and gold films with thicknesses less than 50 nm orient with the backbone parallel to the surface and the SO3− group c3 axis perpendicular to the surface. The utility of SO-ATR extends beyond the system described in this report and will potentially simplify structural elucidation for a variety of polymer thin film systems.
Figure 4. Position of CF2/SO3− peak maximum as a function of film thickness for grazing angle ATR with s-polarized, p-polarized, and unpolarized light for a film on a silicon substrate.
position trends to higher frequencies in both p-polarized and unpolarized experiments due two factors: (1) the increase in zpolarization as film thickness decreases in both cases and (2) the increase in film ordering as thickness decreases. The spectral envelope from 1203−1257 cm−1 is composed of primarily −CF2 symmetric and asymmetric stretches, −CF3 asymmetric stretches and −SO3− asymmetric stretch bands. The position of the peak between 1203 and 1257 cm−1 depends on which vibrational modes are in plane or out of plane with respect to the incident electric field. As the chains align parallel to the surface, the backbone vas(CF2) band will be enhanced in p-polarized light. As a result of this intensity enhancement, the region between 1203 and 1257 cm−1 becomes dominated by the higher frequency vas(CF2) bands at 1232 and 1242 cm−1 resulting in an absorbance maximum close to 1257 cm−1. With s-polarized incident light in SO-ATR, there is an enhancement of lower frequency vs(CF2) bands at 1124 and 1191 cm−1 resulting in an absorbance maximum close to 1220 cm−1. As the thickness of the polymer layer decreases, the intensity of the spolarized electric field of the evanescent wave changes slightly. However, as the polymer film becomes thinner, the intensity of the z-polarized electric field increases dramatically. As a result, there is an increase of surface normal mode intensities and an apparent blueshift in the 1200−1250 cm−1 region as the
■
ASSOCIATED CONTENT
S Supporting Information *
The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acsmacrolett.5b00800. Detailed description of experimental methods and theoretical derivations (PDF).
■
AUTHOR INFORMATION
Corresponding Author
*E-mail:
[email protected]. 86
DOI: 10.1021/acsmacrolett.5b00800 ACS Macro Lett. 2016, 5, 83−87
Letter
ACS Macro Letters Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS The authors acknowledge the support of the U.S. Department of Energy, the Office of Energy Efficiency and Renewable Energy, the Fuel Cells Technology Program through a subcontract from General Motors Corporation under Grant DE-EE0000470.
■
REFERENCES
(1) Yamato, T.; Hideshima, C.; Prakash, G. K. S.; Olah, G. A. J. Org. Chem. 1991, 56 (6), 2089. (2) van der Wal, P. D.; de Rooij, N. F.; Koudelka-Hep, M. Sens. Actuators, B 1996, 35 (1−3), 119. (3) Hickner, M. A. Mater. Today 2010, 13 (5), 34. (4) Modestino, M. A.; Walczak, K. A.; Berger, A.; Evans, C. M.; Haussener, S.; Koval, C.; Newman, J. S.; Ager, J. W.; Segalman, R. A. Energy Environ. Sci. 2014, 7 (1), 297. (5) Xi, J.; Wu, Z.; Qiu, X.; Chen, L. J. Power Sources 2007, 166 (2), 531. (6) Smitha, B.; Sridhar, S.; Khan, A. A. J. Membr. Sci. 2005, 259 (1− 2), 10. (7) Kocha, S. S.; Zack, J. W. J.; Alia, S. M. S.; Neyerlin, K. C.; Pivovar, B. S. ECS Trans. 2013, 50 (2), 1475. (8) Yeh, K.-Y.; Restaino, N. a.; Esopi, M. R.; Maranas, J. K.; Janik, M. J. Catal. Today 2013, 202 (0), 20. (9) Kendrick, I.; Kumari, D.; Yakaboski, A.; Dimakis, N.; Smotkin, E. S. J. Am. Chem. Soc. 2010, 132 (49), 17611. (10) Masuda, T.; Fukumitsu, H.; Kondo, T.; Naohara, H.; Tamura, K.; Sakata, O.; Uosaki, K. J. Phys. Chem. C 2013, 117 (23), 12168. (11) Malevich, D.; Zamlynny, V.; Sun, S.-G.; Lipkowski, J. Z. Phys. Chem. 2003, 217 (5/2003), 513. (12) Dura, J. A.; Murthi, V. S.; Hartman, M.; Satija, S. K.; Majkrzak, C. F. Macromolecules 2009, 42 (13), 4769. (13) Dishari, S. K.; Hickner, M. A. ACS Macro Lett. 2012, 1 (2), 291. (14) Zeng, J.; Jean, D.; Ji, C.; Zou, S. Langmuir 2012, 28 (1), 957. (15) Nagao, Y. J. Phys. Chem. C 2013, 117 (7), 3294. (16) Milosevic, M.; Berets, S. L.; Corporation, H. S. Appl. Spectrosc. 1993, 47 (5), 566. (17) Milosevic, M.; Milosevic, V.; Berets, S. L. Appl. Spectrosc. 2007, 61 (5), 530. (18) Heavens, O. S. Optical Properties of Thin Films; Dover Publications, Inc.: New York, 1965. (19) Hansen, W. N. J. Opt. Soc. Am. 1968, 58 (3), 380. (20) Valeri, P.; Irina, V.; Valeri, A.; Tolstoy, V. P.; Chernyshova, I. V; Skryshevsky, V. A. Handbook of Infrared Spectroscopy of Ultrathin Films; Wiley Interscience: NJ, 2003. (21) Harrick, N. J. Internal Reflection Spectroscopy; John Wiley & Sons: New York, 1967. (22) Anderson, M. S. Appl. Phys. Lett. 2003, 83 (14), 2964. (23) Aroca, R. F. R.; Ross, D. D. J.; Domingo, C. Appl. Spectrosc. 2004, 58 (11), 324. (24) Nagasawa, Y.; Yoshii, I.; Naruke, K.; Yamamoto, K.; Ishida, H.; Ishitani, A. J. Appl. Phys. 1990, 68 (4), 1429. (25) Socrates, G. Infrared and Raman Characteristic Group Frequencies: Tables and Charts; Wiley, 2004; Vol. 283. (26) Korzeniewski, C.; Snow, D. E.; Basnayake, R. Appl. Spectrosc. 2006, 60 (6), 599. (27) Paul, D. D. K.; Karan, K.; Docoslis, A.; Giorgi, J. B.; Pearce, J. Macromolecules 2013, 46 (9), 3461. (28) Danilczuk, M.; Lin, L.; Schlick, S.; Hamrock, S. J.; Schaberg, M. S. J. Power Sources 2011, 196 (20), 8216. (29) Yagi, I.; Inokuma, K.; Kimijima, K.; Notsu, H. J. Phys. Chem. C 2014, 118 (45), 26182.
87
DOI: 10.1021/acsmacrolett.5b00800 ACS Macro Lett. 2016, 5, 83−87
