Anal. Chem. 2007, 79, 8803-8806
Combined in Situ Atomic Force MicroscopyInfrared-Attenuated Total Reflection Spectroscopy Martin Brucherseifer, Christine Kranz, and Boris Mizaikoff*
School of Chemistry and Biochemistry, Georgia Institute of Technology, Atlanta, Georgia 30032-0400
The combination of atomic force microscopy (AFM) with infrared attenuated total reflection (IR-ATR) spectroscopy for simultaneous spectroscopic evanescent field absorption and scanning probe measurements is presented. The capabilities of the combined setup are demonstrated by in situ AFM imaging of the dissolution process of urea in a cyclohexane/butanol solution with nanometer topographical resolution, while simultaneously recording the correlated bulk spectral changes by mid-infrared evanescent field absorption spectroscopy. Hence, surface modification processes such as dissolution or deposition can be simultaneously monitored by AFM imaging and IR spectroscopy in liquid environments, which has not been demonstrated to date. This combined technique will in the future enable kinetic studies on physical, chemical, and biological processes at a wide variety of surfaces providing chemical specificity via IR spectroscopy in addition to high-resolution imaging via AFM. Scanning probe microscopy (SPM) introduced by Binnig et al.1 with the development of scanning tunneling microscopy was the breakthrough for gaining topological information at sample surfaces at a nanometer scale. Besides AFM and STM, near-field scanning optical microscopy (NSOM) and derivative techniques based on the interaction of electromagnetic waves with the sample surface in the near-field regime are gaining importance providing additional optical and/or spectroscopic information. A typical approach in NSOM guides an optical beam through a subwavelength aperture such as, e.g., a tapered glass fiber. Thereby, optical resolution down to 20 nm has been achieved.2-4 This technique is well established in the visible to near-ultraviolet range5 with proven capabilities in many application areas including biochemical analysis. Especially for applications demanding high molecular selectivity, spectroscopic information in the mid-infrared spectral region is the frequency regime of choice for the identification of unknown molecules due to the excitation of fundamental vibrational modes.6,7 Specifically, studies at organic molecules benefit * Corresponding author. E-mail:
[email protected]. Phone: (404) 894-4030. Internet: http://asl.chemistry.gatech.edu. (1) Binnig, G.; Rohrer, H.; Gerber, Ch.; Weibel, E. Phys. Rev. Lett. 1982, 49 (1), 57-61. (2) Pohl, D. W.; Denk, W.; Lanz, M. Appl. Phys. Lett. 1984, 44 (7), 651-653. (3) Durig, U.; Pohl, D. W.; Rohner, F. J. Appl. Phys. 1986, 59 (10), 33183327. (4) Betzig, E.; Isaacson, M.; Lewis, A. Appl. Phys. Lett. 1987, 51 (25), 20882090. (5) Aoki, H.; Hamamatsu, T.; Ito, S. Appl. Phys. Lett. 2004, 84 (3), 356-358. 10.1021/ac071004q CCC: $37.00 Published on Web 10/16/2007
© 2007 American Chemical Society
from the so-called “fingerprint” region (>10 µm) providing distinct absorption patterns characteristic for almost any organic species.8,9 Only a few approaches have been reported using either a subwavelength aperture10,11 or an apertureless approach12,13 extending the conventional NSOM concept into the mid-infrared domain. However, high optical losses limit the application of nearfield IR spectroscopy in aqueous environments, which has not been demonstrated to date. Recently, the combination of AFM with other in-situ techniques such as scanning electrochemical microscopy14-16 and imaging amperometric biosensors17 has been demonstrated corroborating the increasing importance of combined analytical techniques for monitoring complex chemical and biological processes. This technical note describes the first instrumental combination of midinfrared (MIR) attenuated total reflection (IR-ATR) spectroscopy18,19 with AFM, thereby providing molecule specific spectroscopic information at the sample surface, while simultaneously performing high-resolution imaging of surface processes occurring at the ATR crystal interface in liquid environments. As a model example for demonstrating the successful combination AFM-IRATR, the dissolution of urea is spectroscopically monitored, while changes in surface topology resulting from the dissolution process are simultaneously imaged with AFM EXPERIMENTAL SECTION Figure 1 provides a schematic of the developed combined AFM-IR-ATR setup. For IR-ATR spectroscopy, IR radiation from (6) McKelvy, M. L.; Britt, T. R.; Davis, B. L.; Gillie, J. K.; Graves, F. B.; Lentz, L. A. Anal. Chem. 1998, 70 (12), 119R-177R. (7) Wetzel, D. L.; LeVine, S. M. Science 1999, 285 (5431), 1224-1225. (8) Subramaniam, V.; Kirsch, A. K.; Jovin, T. M. Cell. Mol. Biol. 1998, 44 (5), 689-700. (9) Delain, E.; Michel, D.; Le Grimellec, C. Morphologie: Bull. Assoc. Anatom. 2000, 84 (265), 25-30. (10) Michaels, C. A.; Stranick, S. J.; Richter, L. J.; Cavanagh, R. R. J. Appl. Phys. 2000, 88 (8), 4832-4839. (11) Talley, D. B.; Shaw, L. B.; Sanghera, J. S.; Aggarwal, I. D.; Cricenti, A.; Generosi, R.; Luce, M.; Margaritondo, G.; Gilligan, J. M.; Tolk, N. H. Mater. Lett. 2000, 42 (5), 339-344. (12) Knoll, B.; Keilmann, F. Nature 1999, 399 (6732), 134-137. (13) Akhremitchev, B. B.; Pollack, S.; Walker, G. C. Langmuir 2001, 17 (9), 2774-2781. (14) Kranz, C.; Friedbacher, G.; Mizaikoff, B.; Lugstein, A.; Smoliner, J.; Bertagnolli, E. Anal. Chem. 2001, 73 (11), 2491-2500. (15) Kueng, A.; Kranz, C.; Lugstein, A.; Bertagnolli, E.; Mizaikoff, B. Angew. Chem., Int.. Ed. 2003, 42 (28), 3238-3240. (16) Kueng, A.; Kranz, C.; Mizaikoff, B.; Lugstein, A.; Bertagnolli, E. Appl. Phys. Lett. 2003, 82 (10), 1592-1594. (17) Kueng, A.; Kranz, C.; Lugstein, A.; Bertagnolli, E.; Mizaikoff, B. Angew. Chem., Int. Ed. 2004, 44, 3419-3422. (18) Fahrenfort, J. Spectrochim. Acta 1961, 17, 698-709.
Analytical Chemistry, Vol. 79, No. 22, November 15, 2007 8803
Figure 1. Schematic setup of combined AFM-IR-ATR. The sample is deposited at the surface of a single-reflection hemispherical ATR crystal mounted at the bottom of an adapted AFM liquid cell and is penetrated by the exponentially decaying evanescent field. A topdown AFM scanner is aligned to the center of the ATR crystal on top of the sample mount. Components in schematic are not to scale.
a Fourier transform infrared (FT-IR) spectrometer emitted by a SiC filament is collimated and focused by an off-axis parabolic gold mirror onto the curvature of a single-reflection hemispherical ZnS ATR crystal (diameter 3 mm). The hemispherical shape of the ATR crystal additionally focuses the IR radiation to a spot size of approximately 250 µm in diameter at the flat surface of the hemispherical ATR element, thereby increasing the sensitivity of the obtained spectroscopic information. The reflected beam is collimated by a second off-axis parabolic gold mirror and detected by a liquid nitrogen cooled photoconductive mercury-cadmiumtelluride (MCT) semiconductor detector. The obtained signals are analyzed by the FT-IR after passing through a Michelson interferometer (Oriel 8000, L.O.T.-Oriel). A top-down AFM scanner with vertical sample approach (PicoPlus, Agilent Technologies) is aligned on top of the ATR unit coaxial with the ATR crystal center, thereby enabling the combination of spectroscopic evanescent field absorption measurements during atomic force microscopy imaging by integrating the ATR crystal directly into the sample holder of the AFM setup. For measurements, the sample of interest is located at the ATR crystal surface, which simultaneously serves as the bottom of a liquid sample cell. During the IR-ATR measurement, the spectroscopically probed volume of the sample is defined by the penetration depth dp of the exponentially decaying evanescent field strength following Ez ) E0 exp(-z/dp). Hence, spectroscopic information is only obtained from a thin film segment of the liquid cell adjacent to the ATR waveguide surface, thereby enabling direct measurements in liquid without the optical losses encountered in transmission absorption measurements or near-field measurements through the liquid.20 To estimate the penetration depth, the angle of incidence of the IR radiation is assumed at 45°. In this approximation we neglect that at the given optical conditions the angle of incidence actually comprises an angular range due to the focusing effect of the hemispherical ATR crystal. Consequently, the penetration depth of the evanescent field originating at the interface of a ZnS ATR crystal surface (n2 ) 2.2) covered with a thin urea film (n1 ) 1.48) is approximately 1-5.5 µm, in dependence of the observed wavelength window ranging from 3500-600 cm-1. (19) Harrick, N. J. Internal Reflectance Spectroscopy; Wiley-Interscience: New York, 1967. (20) Mizaikoff, B. Anal. Chem. 2003, 75, 258A-267A.
8804 Analytical Chemistry, Vol. 79, No. 22, November 15, 2007
Figure 2. Comparison of two absorption spectra recorded with and without urea precipitated onto the ATR crystal surface. A typical reference spectrum collected without urea is shown in the inset.
Effects of humidity and CO2 fluctuations within the optical path during IR measurements are largely eliminated by enclosing the system within a glove box purged with dry air. During evanescent field absorption measurements, the sample spectrum (I) is ratioed against the reference spectrum recorded with the bare ATR crystal (I0) following A ) -log (I/I0), thereby eliminating remaining atmospheric fluctuations. RESULTS AND DISCUSSION As a model system demonstrating the principal feasibility of combined AFM-IR-ATR measurements, the dissolution of urea in a solvent mixture of cyclohexane and n-butanol was simultaneously monitored providing high-resolution topography images along with IR spectroscopic information on the dissolution process. Prior to simultaneous spectroscopic-microscopic investigations, urea crystals were precipitated onto the ATR crystal surface from methanol solution saturated with urea. Slow evaporation of methanol facilitated the formation of amorphous urea structures at the ATR crystal surface covering an area of approximately 0.02 mm2 with a typical height of several micrometers. Characteristic absorption spectra with and without urea precipitated onto the ATR crystal surface are shown in Figure 2. Without urea precipitation, a spectrum similar to the reference spectrum (shown in the inset of Figure 2) is obtained, which provides information on the wavelength dependent noise of the system. In the spectral range between 600 and 3500 cm-1 where the reference signal strength is more than 10% of its maximum amplitude, a signal-to-noise ratio (SNR) of >50 is achieved in a single scan measurement. The absorption spectrum of urea shown in Figure 2 matches well with urea IR-ATR spectra reported in the literature.21 The IR spectrum of urea reveals vibrational absorption lines at 1466, 1595, 1624, and 1674 cm-1, along with absorption features in the range of 3300-3500 cm-1. To estimate the sensitivity of the ATR measurement, the effective probed volume of urea has to be considered, which depends on the penetration depth of the evanescent field as described above. Assuming a refractive index of amorphous urea similar to that of (21) Groen, H.; Roberts, K. J. Cryst. Growth Des. 2004, 4 (5), 929-936.
Figure 3. In a selection of a representative time series, a sequence of AFM deflection images (left panel) reveals changes of the topology during the dissolution of solid amorphous urea precipitate structures. Different lines emphasize the sequence of the dissolution process. The AFM images shown in the left panel were collected during the dissolution process of solid urea precipitate in a solution of n-butanol/cyclohexane ) 1/10. Correlated IR evanescent field absorption spectra (right panel) are simultaneously recorded and inherently time synchronized to each AFM image. The different line types of the absorption curves match the line types indicating the sequence of the dissolution process shown in the AFM images. The provided AFM images and correlated evanescent field absorption spectra represent 0, 6, 12, and 18 min of the dissolution experiment, with t ) 0 representing the start of a 20 min observation window during the continuous dissolution experiment.
crystalline urea (n ) 1.48), the penetration depth of the evanescent field is estimated at 2.1 µm with a maximum sensitivity at approximately 1600 cm-1. Considering the absorption amplitude shown in Figure 2, the minimum spectroscopically detectable change of the amount of urea deposited at the crystal surface in a single scan is determined at approximately 5.5 ng (9 × 10-11 mol). The dissolution experiments were carried out in a liquid flow cell arrangement of a modified AFM sample trough with the ATR crystal directly integrated into the bottom of the liquid cell. After formation of solid amorphous urea by deposition from methanol solution, the liquid cell is initially filled with cyclohexane (Snyder polarity index is 0.2), which ensures that urea is not yet dissolving. An IR reference spectrum is collected to account for the spectral background of the liquid environment and crystalline urea initially covering the ATR crystal surface. Slow dissolution of urea is then
initiated by gradually increasing the content of n-butanol (Snyder polarity index is 3.9) in cyclohexane from an initial ratio of 1:50 to a ratio of 1:10. IR evanescent field absorption spectra were continuously recorded at a rate of 60 spectra/min, while an area of 20 µm × 15 µm is simultaneously scanned in contact mode AFM, thereby providing in situ images of structural changes during urea dissolution, which are inherently correlated with the observed spectral changes. The simultaneously recorded IR-ATR absorption spectra provide bulk spectroscopic information on the dissolution of the precipitated urea reflecting continuously decreasing amounts of solid urea precipitate within the penetration depth of the evanescent field. Complementarily, AFM imaging provides laterally resolved information on the topographical changes during this dissolution process. Analytical Chemistry, Vol. 79, No. 22, November 15, 2007
8805
Figure 3 shows a representative series of AFM deflection images taken during a period of approximately 20 min along with simultaneously recorded correlated IR-ATR absorption spectra. A selected same sample area of 20 µm × 15 µm covered with urea was continuously imaged via AFM at a scan speed of 83 µm/s using a silicon nitride cantilever (length, 200 µm; nominal spring constant, 0.12 N/m) with a pyramidal tip (base, 4 µm × 4 µm; height, 2.86 µm). The deflection images clearly resolve the edges of the urea precipitate. In Figure 3, the different stages of dissolution are marked with different line types in a series of selected subsequent AFM images and correlated IR-ATR absorbance spectra recorded within an observation window of 20 min during the dissolution experiment. The AFM images document a slow dissolution process of solid urea in the cyclohexane/n-butanol mixture reflected in the changes of the associated topographical features. The images in Figure 3 were collected at a ratio of n-butanol/cyclohexane 1:10. The edges of the amorphous urea precipitate reveal a typical tetragonal structure. Correlated to the AFM images, simultaneously obtained evanescent field absorbance spectra are shown in Figure 3, which represent a spectroscopic snapshot at the end of each related AFM scan. The AFM images and correlated evanescent field absorption spectra provided in Figure 3 represent 0, 6, 12, and 18 min of the dissolution experiment, with t ) 0 representing the start of a 20 min observation window during the continuous dissolution experiment. The absorption peak of urea at 1595 cm-1 was selected for demonstrating spectroscopic monitoring of the urea dissolution process. The line type of each additional spectrum corresponds to the respective edge marked in the correlated AFM images. The dissolution of urea progressing with time induces a decrease of the characteristic absorption feature around 1595 and 3400 cm-1, as a result of the continuously decreasing amount of solid amorphous urea present within the analytical volume probed by the evanescent field. Once the thickness of the initial amorphous urea precipitate has decreased below the penetration depth of the evanescent field, the latter penetrates beyond the solid urea phase into the adjacent liquid. However, given the low concentration of urea in the liquid phase (total liquid cell volume is 0.6 mL) by slowly dissolving the urea precipitate, and the exponential decay of the evanescent field strength, dissolved urea is not detected at these experimental conditions. Hence, a decrease in absorbance at 1595 cm-1 is observed relative to the initial spectrum of the urea precipitate (22) Charlton, C.; Katzir, A.; Mizaikoff, B. Anal. Chem. 2005, 77, 4398-4403. (23) Charlton, C. M.; Giovannini, M.;; Faist, J.; Mizaikoff, B. Anal. Chem. 2006, 78, 4224-4227. (24) Shin, H.; Hesketh, P.; Mizaikoff, B.; Kranz, C. Anal. Chem. 2007, 79, 47694777.
8806
Analytical Chemistry, Vol. 79, No. 22, November 15, 2007
prior to initiating dissolution, which reflects steady dissolution of the amorphous structure. While the dissolution of urea is neither a process in doubt, nor of significant scientific relevance, Figure 3 impressively demonstrates the ability of the described AFMIR-ATR combination for this exemplary system by imaging changes of the topology at a sample surface with high-resolution AFM, while simultaneously determining changes of the bulk spectral properties of the same sample in the mid-infrared range. CONCLUSIONS In the present study, the first in situ instrumental combination of atomic force microscopy with infrared attenuated total reflection spectroscopy for synchronized AFM-IR-ATR studies in liquid environments is demonstrated. The analytical capabilities of this combination are verified by simultaneous spectroscopic (IR-ATR) and topographic (AFM) monitoring of the dissolution process at solid amorphous urea precipitates in a cyclohexane/n-butanol environment providing laterally resolved information via AFM imaging and corresponding bulk spectroscopic information via IRATR evanescent field absorption spectroscopy with temporal correlation in a single synchronized measurement. With the recently reported first combination of single-mode mid-infrared waveguides that are frequency-matched to distributed feedback quantum cascade lasers, the achievable sensitivity during IR evanescent field measurements will be significantly enhanced due to absolute optical mode control and thus enhanced intensity of the evanescent field22,23 and will further facilitate the integration of IR technology with AFM techniques. Consequently, is anticipated that the applicability of this combined analytical technique extends into multiparametric monitoring of a wide range of relevant chemical and biological processes involving topographical and correlated spectroscopic changes of the sample such as observed in protein folding, biomolecular interactions, and cellular signaling processes. Furthermore, our current research focus on the combination of the established AFM-IR-ATR platform with conductive AFM techniques using, e.g., atomic force scanning electrochemical microscopy (AFM-SECM) tips will enable in situ studies on the laterally resolved deposition of conducting polymers.14,24 ACKNOWLEDGMENT The authors gratefully acknowledge supported of this work by the National Institute of Health, NIH (Grant R01-EB000508).
Received for review May 16, 2007. Accepted August 17, 2007. AC071004Q
