Anal. Chem. 2008, 80, 5246–5250
Quartz Crystal Microbalance with Integrated Surface Plasmon Grating Coupler Yun Zong,† Fei Xu,† Xiaodi Su,† and Wolfgang Knoll*,†,‡ Institute of Materials Research and Engineering, A*STAR (Agency for Science, Technology and Research), 3 Research Link, Singapore 117602, and Max-Planck Institute for Polymer Research, Ackermannweg 10, 55128 Mainz, Germany We have integrated a surface plasmon grating coupler into a quartz crystal microbalance (QCM) for studying surface association/dissociation reactions. In the integrated system only QCM measurement is needed to record both the optical and the acoustic signals in the same association/ dissociation reaction. This integration considerably simplifies a conventional combination instrument of a gratingcoupled surface plasmon resonance (SPR) spectrometer and a quartz crystal microbalance by eliminating a number of SPR components. Moreover, in the integrated system detection of the light reflections is not needed by which one bypasses the interference problem caused by two coherent light reflections off the glass window used to seal the fluid sample and off the sensor surface. The utility of the integrated system is demonstrated using a layer-bylayer polyelectrolyte multilayer deposition protocol, in which the complete features of a conventional gratingcoupled SPR/QCM combination instrument are retained, including detection of optical and acoustic changes, as well as monitoring of adsorption kinetics. Surface plasmon resonance (SPR) spectroscopy is a surface analytical technique that measures the optical masses (molecular masses) of thin films deposited onto noble metal coated substrate.1 Quartz crystal microbalance (QCM) is a mechanical oscillator that records the acoustic masses, i.e., the sum of the masses of the deposited molecules and the masses of the solvent molecules entrapped in the matrix of these deposited molecules.2–6 For a given surface adsorption process, the comparison between SPR and QCM masses can give, among other parameters, the degree of hydration (for aqueous solutions) of the adsorbed layer that often relates well to the conformational properties of the molecules * Corresponding author. E-mail:
[email protected]. Fax: ++49-6131379-360. † Agency for Science, Technology and Research. ‡ Max-Planck Institute for Polymer Research. (1) Knoll, W. Annu. Rev. Phys. Chem. 1998, 49, 569–638. (2) Odonnell, J.; Honeybourne, C. L. J. Phys.: Condens. Matter 1991, 3, S337– S346. (3) Caruso, F.; Serizawa, T.; Furlong, D. N.; Okahata, Y. Langmuir 1995, 11, 1546–1552. (4) Kosslinger, C.; Uttenthaler, E.; Drost, S.; Aberl, F.; Wolf, H.; Brink, G.; Stanglmaier, A.; Sackmann, E. Sens. Actuators, B 1995, 24, 107–112. (5) Vikinge, T. P.; Hansson, K. M.; Sandstrom, P.; Liedberg, B.; Lindahl, T. L.; Lundstrom, I.; Tengvall, P.; Ho ¨o ¨k, F. Biosens. Bioelectron. 2000, 15, 605– 613. (6) Gabai, R.; Sallacan, N.; Chegel, V.; Bourenko, T.; Katz, E.; Willner, I. J. Phys. Chem. B 2001, 105, 8196–8202.
5246
Analytical Chemistry, Vol. 80, No. 13, July 1, 2008
in the layer formed.7–10 In cases when SPR and QCM experiments are conducted separately, the reliability of the extracted information (hydration degree or conformational information) heavily depends on the extent of the similarity of the experimental conditions (sample concentration, temperature, liquid flow, surface conditions, etc.) of the two sets of experiments.11 One way to improve the reliability of this combinatorial data collection and analysis is to couple SPR and QCM functions in one device so that both the optical and the acoustic signals can be acquired simultaneously from a single adsorption process at otherwise identical experimental conditions. This was realized by combining a grating-coupled surface plasmon resonance (G-SPR) spectrometer with a QCM.12 Typically, surface plasmons here are excited at the noble metal/dielectric interface via a shallow periodic surface corrugation (a grating) structure1 with little perturbation on the oscillation of a so-called AT-cut (a temperaturecompensated cut, by which the temperature effect of quartz crystal is minimized) quartz crystal wafer.12–15 Such a combination technique has been used for studying protein adsorption12 and interfacial behavior of thin films formed either by solution adsorption13,14 or via an electrochemical polymerization process.15 Although powerful and promising, this combination technique often suffers from severe interference of two coherent reflections off the glass window used to seal the fluid sample and off the sensor surface, which are recorded by the photodiode detector. In this note, we present a considerable simplification of G-SPR/ QCM combination instrument so that only QCM readout is needed to acquire both optical and acoustic signals. The simplification is based on our recent observation that the oscillation frequency of an AT-cut quartz crystal wafer linearly responds to the amount of (7) Francis, L. A.; Friedt, J. M.; Zhou, C.; Bertrand, P. Anal. Chem. 2006, 78, 4200–4209. (8) Geddes, N. J.; Urquhart, R. S.; Furlong, D. N.; Lawrence, C. R.; Tanaka, K.; Okahata, Y. J. Phys. Chem. 1993, 97, 13767–13772. (9) Ho ¨o ¨k, F.; Kasemo, B.; Nylander, T.; Fant, C.; Sott, K.; Elwing, H. Anal. Chem. 2001, 73, 5796–5804. (10) Su, X. D.; Lin, C. Y.; O’Shea, S. J.; Teh, H. F.; Peh, W. Y. X.; Thomsen, J. S. Anal. Chem. 2006, 78, 5552–5558. (11) Reimhult, E.; Larsson, C.; Kasemo, B.; Ho ¨o ¨k, F. Anal. Chem. 2004, 76, 7211–7220. (12) Laschitsch, A.; Menges, B.; Johannsmann, D. Appl. Phys. Lett. 2000, 77, 2252–2254. (13) Bailey, L. E.; Kambhampati, D.; Kanazawa, K. K.; Knoll, W.; Frank, C. W. Langmuir 2002, 18, 479–489. (14) Plunkett, M. A.; Wang, Z. H.; Rutland, M. W.; Johannsmann, D. Langmuir 2003, 19, 6837–6844. (15) Bund, A.; Baba, A.; Berg, S.; Johannsmann, D.; Lubben, J.; Wang, Z. H.; Knoll, W. J. Phys. Chem. B 2003, 107, 6743–6747. 10.1021/ac800393d CCC: $40.75 2008 American Chemical Society Published on Web 05/31/2008
the photon energy dissipated in its front gold electrode.16 Such irradiation effects were taken as artifacts in previous studies in which illumination was involved17,18 which, however, is barely visible in a conventional G-SPR/QCM combination study due to the relatively low laser beam intensity. Once the laser intensity is increased by an order of magnitude or more, interesting features will present in an angular scan upon surface plasmon excitation at the surface of the AT-cut sensor quartz crystal wafer (with a grating structure)sa quasi-mirror image19 of the SPR curve (optical signals) appears in the frequency responses (acoustic signals, details described below). This novel feature allows for the integration of a surface plasmon grating coupler into a QCM, by which one considerably simplifies the conventional G-SPR/ QCM combination instrument via the elimination of a number of SPR components (lock-in amplifier, mechanical chopper, polarizers, photodiode detector and detector motor). Moreover, the integration bypasses the interference problem caused by the two coherent light reflections off the glass window used to seal the fluid sample and off the sensor surface, since the detection of reflections is no longer required in the integrated system. The utility of the integrated system (in situ interference-free detection of the optical and the acoustic thicknesses of thin films adsorbed onto a quartz crystal wafer as well as the adsorption kinetics) is demonstrated using a well-documented polyelectrolyte layer-bylayer deposition protocol.20–22 EXPERIMENTAL SECTION Reagents. Poly(diallyl dimethyl ammonium chloride) (PDADMAC, MW: 200 000-350 000, 20 wt % aqueous solution), sodium polystyrene sulfonate (PSS, MW: 70 000, 30 wt % aqueous solution), 3-mercapto-1-propane sulfonic acid sodium salt (MPS, 90%), sulfuric acid (H2SO4, 99%), anhydrous ethanol (>99.5%), and phosphate-buffered saline (PBS) tablets used in the sensor substrate surface cleaning and the polyelectrolyte layer-by-layer deposition were all Sigma-Aldrich products. Hydroperoxide (H2O2, 31 wt % aqueous solution) was purchased from Santoku BASF (Japan). All the reagents were used as received without further purification. The PDADMAC and PSS aqueous solution samples were 0.01 M with an ionic strength of 0.18 M, prepared by diluting 1 portion of their respective 0.10 M aqueous solution with 9 portions (v/v) of 0.20 M PBS buffer. The molecular structures of PDADMAC and PSS are shown schematically in Figure 1. Sensor Crystal Surface Preparation. The gold-coated grating-bearing AT-cut quartz crystal wafers with a diameter of ∼14 mm and a fundamental resonant frequency of ∼5 MHz were purchased from Res-Tech (Framersheim, Germany), with the grating constant and the grating amplitude being Λ ∼ 520 nm and h ) 20-30 nm, respectively. These sensor crystals were cleaned by treating their grating-bearing front electrode in a (16) Zong, Y.; Xu, F.; Su, X. D.; Knoll, W. J. Appl. Phys., in press. (17) Etchenique, R.; Furman, M.; Olabe, J. A. J. Am. Chem. Soc. 2000, 122, 3967–3968. (18) Hidaka, H.; Honjo, H.; Horikoshi, S.; Serpone, N. New J. Chem. 2003, 27, 1371–1376. (19) The oscillation frequency of the quartz crystal wafer increases as the reflectivity in the SPR angular scan decreases. (20) Decher, G. Science 1997, 277, 1232–1237. (21) Lvov, Y.; Essler, F.; Decher, G. J. Phys. Chem. 1993, 97, 13773–13777. (22) Caruso, F.; Niikura, K.; Furlong, D. N.; Okahara, Y. Langmuir 1997, 13, 3422–3426.
Figure 1. Schematic depiction of polyelectrolyte molecular structure: (a) poly(diallyl dimethyl ammonium chloride) (PDADMAC); (b) sodium polystyrene sulfonate (PSS).
Figure 2. Sketch of the combination instrument which consists of a home-built surface plasmon resonance spectrometer and the quartz crystal microbalance with dissipation monitoring (Q-Sense, Sweden): 1, laser; 2, chopper; 3, polarizers; 4, QCM window cell with a gratingfeatured sensor quartz crystal wafer being mounted inside. No glass window is used.
UV-ozone cleaner (Jelight, U.S.A.) for 5 min, followed by a treatment in a mixture of H2SO4/H2O2 (3:1) (highly corrosive, handle with care) for 10 min, rinsing with sufficient amount of Milli-Q water (18.2 MΩ · cm), and drying in a stream of dry nitrogen. The clean front electrode was then exposed to a 0.01 M MPS ethanol solution for the introduction of negative charges to the electrode surface. After rinsing thoroughly with ethanol, followed by rinsing in Milli-Q water, the senor crystal was mounted in a QCM window cell for either measurement in air of the correlation between optical-acoustic signals or in solution for the measurement of the alternate deposition of the PDADMAC and PSS monolayers. Instrumentation. A combination instrument consisting of a home-built grating SPR spectrometer and a QCM with dissipation monitoring23,24 (QCM-D, Q-Sense, Sweden), as shown schematically in Figure 2, was used to explore the correlation between the acoustic responses and the reflectivity loss (optical signals) in an SPR angular scan. In experiments a p-polarized laser beam with desired intensity and chopped pulse frequency, obtained by passing the beam of a helium-neon laser with a wavelength of λ ) 633 nm (JDS Uniphase, 1125P) through a chopper and two successive polarizers, was incident on the grating-bearing gold (23) Rodahl, M.; Höök, F.; Krozer, A.; Brzezinski, P.; Kasemo, B. Rev. Sci. Instrum. 1995, 66, 3924–3930. (24) Rodahl, M.; Kasemo, B. Rev. Sci. Instrum. 1996, 67, 3238–3241.
Analytical Chemistry, Vol. 80, No. 13, July 1, 2008
5247
Figure 3. Frequency change (solid curve) of a quartz crystal wafer with a grating structure (Λ ∼ 520 nm) as a function of the angle of incidence in a surface plasmon resonance angular scan (open circles): λ ) 633 nm, p-polarized, I ) 0.14 mW/mm2, chopper frequency Fp ) 1200 Hz.
electrode of the sensor wafer mounted in the QCM-D window cell (Q-Sense, Sweden). The window cell was placed on a twocircle goniometer sample stage which can be operated in a θ/2θ rotational mode. The intensity of the light reflected off the electrode surface was monitored by a photodiode detector, and a lock-in amplifier was applied to modulate the chopper frequency and ensure that only light with the same modulated frequency was detected (in order to avoid disturbance of surrounding illuminations). The Q-Sense D300 electronics unit was used to drive the QCM oscillation and measure the frequency responses. This experiment was conducted in air without the use of a glass window. The integrated system developed in this study is a simplification of the G-SPR/QCM-D combination instrument via the elimination of a number of costly SPR components (for details of the system see the Results and Discussion). The new system was taken to study polyelectrolyte layer-by-layer deposition from the aqueous solutions of PDADMAC and PSS, in which a transparent glass window was used to seal the solution inside the window cell and the average power density of the red laser beam (λ ) 633 nm) was 0.50 mW/mm2. RESULTS AND DISCUSSION On a conventional G-SPR/QCM-D combination instrument an SPR angular scan was taken from an incident angle of θ ) -20° to θ ) 20°, and the QCM frequency response was recorded throughout the scan. There a p-polarized laser beam with an average power density of 0.14 mW/mm2, chopped at a pulse frequency of 1200 Hz, was incident on the front gold grating electrode of the quartz crystal wafer, and the QCM frequency response is shown together with the SPR angular scan curve in Figure 3. One can see in Figure 3 that the amplitude of the frequency changes correlates exactly to the surface plasmon reflectivity loss across the whole angular spectrum except for angles below the critical angle (θ < -13°) and its mirror images (θ > 13°). In these two exceptional regions, the incident laser beam was diffracted into the -1/+1 diffraction orders, respectively, with a concomitant loss in the specularly reflected intensity.25 This, however, did not result in a change in the amount of the deposited photon energy, (25) Grygier, R. K.; Knoll, W.; Coufal, H. Can. J. Phys. 1986, 64, 1067–1069.
5248
Analytical Chemistry, Vol. 80, No. 13, July 1, 2008
Figure 4. Sketch of an optics-integrated quartz crystal microbalance (O-QCM): 1, laser; 2, sample rotary stage; 3, sensor crystal with grating; 4, flow cell inlet; 5, flow cell outlet.
and hence, the change in the oscillation frequency remained at the level of a plain illumination. With the angle of incidence approaching the surface plasmon resonance, the oscillation frequency of the quartz crystal wafer began to increase visibly with its maximum being reached at the resonant angle (θ ) -9.7°). For the scanned region from an incident angle of θ ) -2° to θ ) 2° the light was blocked by the photodiode detector, and accordingly the frequency change dropped to 0. A symmetric scan in the positive direction (θ ) 2° to θ ) 20°) gave virtually identical results. The strict correlation between the amplitude of the frequency change and the reflectivity loss allows for recordings of the SPR using the QCM frequency response. The sensitivity and the S/N ratio of this new detection scheme inherit that of the conventional grating SPR measurement, since in such “optical” recording not the absolute peak value (frequency) but the peak position (time) matters. In this case both the acoustic and the optical signals are obtainable from a single piece of QCM equipment in a single experiment. Hence, we call it a quartz crystal microbalance with integrated surface plasmon grating coupler, or for short, an opticsintegrated quartz crystal microbalance (O-QCM). In brief, an O-QCM consists of a p-polarized laser, a motor-driven rotary sample stage, a window cell, a grating-bearing sensor quartz crystal wafer, a frequency readout device, and a PC steering the motor and collecting the measurement data, as schematically shown in Figure 4. The utility of the O-QCM is calibrated with a study on the layerby-layer buildup of PDADMAC/PSS polyelectrolyte multilayers. The alternate deposition of a charged monolayer onto an oppositely charged surface and their high reproducibility have been evidenced separately in SPR26 and QCM studies.27–30 In the present study, a clean grating-sensor crystal was mounted in the QCM window cell and PBS of an ionic strength of 0.18 M was rinsed through the cell. After the collection of a stable frequency baseline, a fine angular scan from θ ) 0° to θ ) 25° (step size: ∆θ ) 0.1°) was taken with both the starting and the ending time of the scan being noted. In the scan a frequency trace with a shape (26) Advincula, R.; Aust, E.; Meyer, W.; Knoll, W. Langmuir 1996, 12, 3536– 3540. (27) Ariga, Y.; Lvov, T.; Kunitake, J. J. Am. Chem. Soc. 1997, 119, 2224–2231. (28) Baba, A.; Kaneko, F.; Advincula, R. C. Colloids Surf., A 2000, 173, 39–49. (29) Lvov, Y.; Decher, G.; Mo ¨hwald, H. Langmuir 1993, 9, 481–486. (30) Ko ¨stler, S.; Delgado, A. V.; Ribitsch, V. J. Colloid Interface Sci. 2005, 286, 339–348.
Figure 5. Deposition trace of two cycles of alternate deposition of PDADMAC/PSS bilayers on a negatively charged surface recorded by the O-QCM. λ ) 633nm, p-polarized, I ) 0.50 mW/mm2. Inset: magnification of the angular scans prior to the first injection of the polyelectrolyte solution.
feature inverse to a regular SPR angular scan curve (Figure 5, region a) appeared. As the motor was quickly moved back to the position of θ ) 0°, a narrow and sharp peak with a smaller frequency response (inset in Figure 5) was seen, which was essentially the repetition of the angular scan but with a much faster scan rate. After the collection of another baseline, an aqueous solution of the positively charged PDADMAC (0.01 M in PBS) was rinsed through the flow cell via a peristaltic pump at a flow rate of 0.5 mL/min. An immediate decrease of the quartz frequency was seen (region “A” in Figure 5), indicating the adsorption of the positively charged PDADMAC to the MPScovered substrate surface. Shortly after an equilibrium of the adsorption was reached, PBS buffer with an ionic strength of 0.18 M was pumped through the cell to remove any excess amount of free PDADMAC while keeping the liquid environment of the sensor wafer essentially unchanged (in order to minimize the buffer exchange effects). Another angular scan was then performed (Figure 5, region b). Again the motor was moved back, before an aqueous solution of the negatively charged PSS (0.01 M in PBS) was introduced. The adsorption of PSS led to a further decrease of the frequency (region “B” in Figure 5). After the PSS deposition and the rinsing step with 0.18 M PBS buffer, the angular scan recording was repeated (Figure 5, region c). With the repetition of the alternate polyelectrolyte layer-by-layer deposition films with desired thickness can be prepared. With a time-to-angle conversion (cf., the Supporting Information) of the frequency “kinetic” scans presented in Figure 5, one can generate a series of angular scans that correspond to the reflectivity measurements in classical SPR experiment, however, with inverse features (Figure 6a). As the PDADMAC and PSS monolayers deposited from their respective solutions to the grating quartz substrate, the frequency peak position shifted to larger angles of incidence, with the shifts being more pronounced for PSS than for PDADMAC. If one extracts the data from Figure 6a and plots the shifts of the peak position correlated angle of incidence (∆θ) against the amplitudes of the frequency decreases (-∆f) incurred in the same deposition cycles, as shown in Figure 6b, it can be seen that the ∆θ (optical mass, mSPR) versus the –∆f (acoustic mass, mQCM) for the depositions of PDADMAC and PSS fit to two different slopes (mSPR/mQCM) with the slope of the PDADMAC deposition being much smaller (larger mQCM/mSPR)
Figure 6. (a) Optical data converted from the O-QCM kinetic data. Scans were taken before and after the deposition of each individual layer; (b) shifts in angle of incidence against amplitudes of decrease in oscillation frequency for the deposition of individual layers.
than that of the PSS deposition. This suggests that the PDADMAC film adapted a more flexible or less densely packed structure with larger content of solvent entrapment,31–33 which coincides with their molecular structure. The quantitative thickness evaluation conducted by taking the Pockrand approximation12,34 for the optical thickness and Sauerbrey equation35 for the acoustic thickness (for details see the Supporting Information) results in the value for each individual layer of the deposited polyelectrolyte film, as listed in Table 1. One can see from Table 1 that PSS gives a larger monolayer optical thickness than PDADMAC. For PSS the optical thicknesses (dSPR) and the acoustic thicknesses (dQCM) are close (12-30% deviation), whereas for PDADMAC the thicknesses deviation is as large as 100-200%. These results indicate that the acoustic thicknesses of the PSS monolayers mainly came from the contribution of the PSS polymer molecules, whereas that of the PDADMAC monolayers received a large contribution from the entrapped water among the polymer molecules. It is noteworthy that the acoustic thicknesses obtained here in PBS buffer for both PSS and PDADMAC are found larger than the reported values for the same monolayers deposited from their respective aqueous solutions prepared by dissolving the polyelectrolyte materials in deionized water.28 This was caused by the addition (31) Ho ¨o ¨k, F.; Voros, J.; Rodahl, M.; Kurrat, R.; Boni, P.; Ramsden, J. J.; Textor, M.; Spencer, N. D.; Tengvall, P.; Gold, J.; Kasemo, B. Colloids Surf., B 2002, 24, 155–170. (32) Su, X. D.; Wu, Y. J.; Knoll, W. Biosens. Bioelectron. 2005, 21, 719–726. (33) Su, X. D.; Wu, Y. J.; Robelek, R.; Knoll, W. Langmuir 2005, 21, 348–353. (34) Schweiss, R.; Lu ¨ bben, J. F.; Johannsmann, D.; Knoll, W. Electrochim. Acta 2005, 50, 2849–2856. (35) Sauerbrey, G. Z. Phys. 1959, 155, 206–222.
Analytical Chemistry, Vol. 80, No. 13, July 1, 2008
5249
Table 1. Calculated Optical/SPR and Acoustic/QCM Thicknesses of the Polyelectrolyte Layers Depositeda second first layer of first layer second layer layer of PDADMAC of PSS of PDADMAC PSS dtotal dSPR/nm dQCM/nm dQCM/dSPR
2.79 8.47 3.04
4.99 5.60 1.12
2.78 5.54 1.99
4.97 6.45 1.30
15.53 26.06 1.68
a The optical thickness is calculated by assuming dielectric constants of f ) 2.25 (for the PDADMAC layer) and f ) 2.56 (for the PSS layer). For the acoustic calculation, the film deposited is treated as rigid layers and a film density of 1.2 g/cm3 is used for both PDADMAC and PSS layers.
of salt to the polyelectrolyte solutions which results in a decrease of the radius of gyration of the polymer coil by a charge screening effect.29,30 If one considers a thin layer rich in entrapped water (for the PDADMAC monolayers), it is natural to argue that the dielectric constant of f ) 2.25 used above in its optical thickness calculation was an overestimation (the dielectric constants for common dry organic films and PBS at λ ) 633 nm are ) 2.25 and ) 1.778, respectively). Taking this into account one may take a lower dielectric constant, e.g., f ) 2.0, to recalculate the optical thickness of PDADMAC, which will lead to a more consistent picture in the molecular packing (i.e., similar thickness). Solvent entrapment increases the acoustic mass of the deposited layer, but not the layer thickness. This, again, shows the merit of additional acoustic signals that help correct the dielectric constants
5250
Analytical Chemistry, Vol. 80, No. 13, July 1, 2008
of the deposited layers in studies using optical methods. This merit can be easily achieved with the integrated system shown above. CONCLUSION We have integrated grating-coupled SPR detection into a QCM and used it for studying layer-by-layer polyelectrolyte multilayer assembly. The innovation of this integrated system is the elimination of the SPR detection system, using only a QCM detection module to gain the full features of SPR and QCM, including the detection of the optical thickness, the acoustic thickness, as well as the adsorption kinetics. The acquired optical and acoustic signals from the same adsorption process allow for a precise comparison of the optical and the acoustic thicknesses of the adsorbed films, from which the packing information of the films can be extracted. The integrated system here is demonstrated in a tandem detection mode; however, the simultaneous mode of detection with further simplified data treatment is doable (via further software automation). With its considerably reduced requirement for optical detection modules and, hence, much lower cost the integrated system will find wide applications. SUPPORTING INFORMATION AVAILABLE Additional information as noted in text. This material is available free of charge via the Internet at http://pubs.acs.org. Received for review February 26, 2008. Accepted April 12, 2008. AC800393D
