Letter Cite This: ACS Sens. XXXX, XXX, XXX−XXX
pubs.acs.org/acssensors
Nanoliter Sensing for Infrared Bioanalytics Christoph Kratz,*,† Andreas Furchner,† Thomas W. H. Oates,† Dirk Janasek,‡ and Karsten Hinrichs† †
Leibniz-Institut für Analytische Wissenschaften − ISAS − e.V., Schwarzschildstr. 8, 12489 Berlin, Germany Leibniz-Institut für Analytische Wissenschaften − ISAS − e.V., Otto-Hahn-Str. 6b, 44227 Dortmund, Germany
‡
ABSTRACT: Nondestructive label-free bioanalytics of microliter to nanoliter sample volumes with low analyte concentrations requires novel analytic approaches. For this purpose, we present an optofluidic platform that combines surfaceenhanced in situ infrared spectroscopy with microfluidics for sensing of surface-immobilized ultrathin biomolecular films in liquid analytes. Submonolayer sensitivity down to surface densities of few ng/cm2 is demonstrated for the adsorption of the thiolate tripeptide glutathione and for the recognition of streptavidin on a biotinylated enhancement substrate. Nonfunctionalized and functionalized metal island films on planar oxidized silicon substrates are used for signal enhancement with quantifiable enhancement properties. A single-reflection geometry at an incidence angle below the attenuated-total-reflection (ATR) regime is used with ordinary planar, IR-transparent windows. The geometry circumvents the strong IR absorption of common polymer materials and of aqueous environments in the IR fingerprint region. This practice enables straightforward quantitative analyses of, e.g., adsorption kinetics as well as chemical and structural properties in dependence of external stimuli. KEYWORDS: surface enhanced infrared absorption, in situ IR spectroscopy, microfluidics, vibrational sensing, molecular sensing, absorption kinetic
O
(Figure 1b) exploit the effect of surface enhanced infrared absorption (SEIRA) providing an average signal enhancement of 10−100 over the full mid-IR range and over the total measurement area.14,15 Recently, the application of tailored plasmonic nanoantenna substrates for in situ studies with enhancement factors of up to 107 has been shown.16−18 These high factors, however, are obtained at the expense of spatially and spectrally highly localized enhancement, potentially limiting the applicability in sensor applications.19,20 The integration of enhancement substrates as presented in this work allows one to obtain high sensitivity down to surface densities in the nanogram per square centimeter range. With island film thicknesses in the nanometer range, the substrates are compatible with thin-film functionalization techniques, offer an a priori quantifiable enhancement, and can be used for quantitative analysis via optical thin film models.10,21 Our optofluidic platform’s single-reflection geometry under non-ATR conditions allows the use of simple planar IR substrates, circumventing the high absorption of common microfluidic chip materials like polymers or glass. In consequence, the platform is compatible with commercially available microfluidic chip materials. Limits in the length of the light path due to the strong absorption of water are avoided, thereby eliminating needs for fabricating chips from IR-
ptical detection techniques in combination with microfluidics have promoted the development of a plethora of novel sensor concepts for lab-on-chip, organ-on-chip, bioanalytics, as well as diagnostics applications with increasing impact over the past decade.1−3 Label-based nondestructive techniques such as fluorescence spectroscopy may offer sensitivities down to the single-molecule level4,5 at the cost of being limited either to self-fluorescent target molecules or to requiring the development of specialized labeling strategies of target molecules. Vibrational spectroscopy and in situ IR spectroscopy on the other hand are label-free techniques that allow molecular identification and quantification directly via the specific vibrational fingerprint of the analyte.6 Contrary to other label-free methods like quartz crystal microbalance (QCM) or surface plasmon resonance (SPR) based detection, the high material and structural contrast in the IR enables spectral vibrational analyses that provide direct chemical identification with additional structural information.7−10 Although label-free in situ IR spectroscopy can detect both analyte chemistry and structure, the analysis of nanoliter sample volumes with low analyte concentration remains a challenge. Different concepts for IR-spectroscopy based detection on microfluidic devices have been demonstrated,11−13 but their application in sensing concepts is generally limited by low sensitivity and incompatibility of transmission measurements with polymeric microfluidic chips.13 The developed optofluidic platform in this work is shown in Figure 1a. Detection windows covered with metal island films © XXXX American Chemical Society
Received: December 4, 2017 Accepted: February 6, 2018 Published: February 6, 2018 A
DOI: 10.1021/acssensors.7b00902 ACS Sens. XXXX, XXX, XXX−XXX
Letter
ACS Sensors
detection (LOD) in aqueous environment. The first experiment examined the immobilization of a small thiolate molecule, the tripeptide glutathione (GSH), which is the most abundant nonprotein bound thiol in cells.24 A 6 mM solution of glutathione (Sigma-Aldrich, L-glutathione reduced ≥98%) in deionized water (Millipore SAS, Direct Q 3UV) was pumped through the microfluidic channel with a volume of 155 nL at a volume flow rate of 0.2 mL/h. Consecutive spectra were recorded under continuous flow for a sample volume of approximately 2.5 nL over 6.5 h at 2.5 min per measurement. GSH binding was evaluated by determining the peak amplitude of a fitted Gaussian curve to the amide II band of GSH. Similar results were obtained for all observable vibrational bands. An initial spectrum recorded with the solution in the channel at the beginning of the experiment served as a reference spectrum. Figure 2a shows the time dependence of the fitted amide II peak amplitude during GSH adsorption on the enhancement
Figure 1. Optofluidic platform for in situ nanoliter IR spectroscopy. (a) Schematic cross-sectional image of the microfluidic flow cell for IR microscopy. Inset: Zoom of the detection layer with metal-islands enhancement substrate. (b) Exemplary SEM image of gold island film on silicon substrate.
transparent materials, as required for transmission cells.13 This simplification will facilitate a broad range of applications and potentially enables direct probes everywhere on the microfluidic chip. In addition, the presented optofluidic platform does not involve additional optics, as, e.g., prisms or structured surfaces of the substrate, as required when ATR conditions at the solid−liquid interface have to be obtained. The developed optofluidic platform can be used for example in an IR microscope in conventional reflection geometry. Similar to ATR, for quantitative evaluation the specular reflection can be analyzed by appropriate optical-layer-model simulations.10 In the experiments, a PDMS microfluidic chip fabricated by means of soft lithography22 with a straight channel (width = 100 μm, depth = 37 μm, length = 42 mm) was used for the handling of the fluid sample. A syringe pump (World Precision Instruments Ltd., AL1000-220) was used for sample delivery to the microfluidic chip. The employed enhancement substrates were gold island films prepared on IR-transparent silicon substrates (74 mm × 20 mm × 0.8 mm, MateK GmbH). The substrates’ enhancement properties were quantified a priori by an enhancement evaluation of the native oxide vibrational band. Details of substrate preparation, optimization, and enhancement quantification can be found elsewhere.23 An IR microscope (Hyperion 3000, Vertex 70 spectrometer, MCT detector, Bruker Corp.) with a Cassegrain objective (NA = 0.4; spot diameter = 160 μm) and corresponding software (OPUS 7.0, Bruker Corp.) was used for measurements at 2 cm−1 spectral resolution. The incidence angle on the substrate was 4.6°. Noise levels were defined by the 3σ criterion. We present two experiments with the platform system in order to evaluate the sensitivity and to define the limit of
Figure 2. In situ IR monitoring of GSH monolayer formation in microfluidic flow cell. (a) Peak amplitude of amide II vibrational band during monolayer formation and fit of an exponential function corresponding to Langmuir-type adsorption (red line). (b) Exemplary in situ IR spectra corresponding to marked points in (a). Spectra are referenced to initial spectrum recorded directly after the introduction of the solution and shifted for better comparison.
substrate. Associated IR spectra are shown in Figure 2b. Successful adsorption is confirmed by the presence of characteristic vibrational bands of GSH.24 With a throughput volume of 8.3 μL of analyte solution, a first IR spectrum was obtained after merely 2.5 min. In order to estimate a LOD, the number of molecules contributing to the signal was estimated. The surface area of a metal island film increases compared to a flat gold film in B
DOI: 10.1021/acssensors.7b00902 ACS Sens. XXXX, XXX, XXX−XXX
Letter
ACS Sensors
for the ≈1.2 nm thick GSH monolayer, showing how its carboxyl groups are affected by environmental stimuli like different acidities.27 The molecules in the monolayer’s initial state at pH 5 (growth pH) show partially dissociated carboxyl groups. Further dissociation was obtained in an alkaline solution (NaOH (Carl Roth GmbH + Co. KG, sodium hydroxide ≥99%), pH 10), followed by protonation in an acidic solution (HCl (Merck, Suprapur, 30%), pH 2). Figure 3b shows corresponding referenced in situ IR microscopy spectra of the monolayer after equilibration at the respective pH. Upward- and downward-pointing bands amount, respectively, to a decrease and an increase in absorption. In alkaline environment (blue), an upward-pointing band of carboxyl groups around 1725 cm−1 and two downward-pointing bands of dissociated carboxyl groups around 1580 and 1395 cm−1 are monitored, whereas the reverse trend is observed in acidic environment (red). In agreement with Aureau et al.,28 the microfluidic study shows that the monolayer’s surface equilibrium of carboxyl and carboxylate groups can be reversibly tuned by proper adjustment of environmental pH. In the second experiment, recognition of streptavidin on a biotinylated substrate was investigated (Figure 4a). The streptavidin−biotin system serves as a model for the detection of molecules specifically adsorbed to linker molecules assembled on the surface of the enhancement substrate. The biotinylated enhancement substrate was prepared onchip using a biotin SAM formation reagent kit (Dojindo Molecular Technologies Inc.). The biotinylated substrate was subsequently exposed to a 0.6 μM solution of streptavidin (Sigma-Aldrich, streptavidin from Streptomyces avidinii) in PBS buffer (Sigma-Aldrich, PBS tablet in deionized water, pH 7) at a volume flow rate of 0.15 mL/h. Adsorption of streptavidin was monitored over 6 h (1024 scans, tmeas = 5 min) by integrating the area of the amide I vibrational band (Figure 4a). An initial spectrum recorded on the channel filled with PBS buffer was used as a reference. Exemplary spectra shown in Figure 4b demonstrate that characteristic vibrational bands in the amide I and amide II regions are obtained already after 5 min, confirming the presence of streptavidin.29 Taking the amide I band area for the complete streptavidin monolayer as a reference for full surface coverage with a surface density of (3.85 ± 0.15) pmol/cm230 and an effective surface density of (4.16 ± 0.16) pmol/cm2, the noise level corresponds to an estimated LOD of (0.12 ± 0.01) pmol/cm2 or (7.1 ± 0.3) ng/cm2 for streptavidin in an aqueous environment. The amide I band is centered at 1637 cm−1, which is characteristic of β-sheet secondary structure of proteins.31 This finding is in agreement with the known β-barrel structure of streptavidin in its native state.32 The visible changes in amide I and amide II band shapes during adsorption are indicative of structural changes of the protein as a consequence of biotin binding.32 With access to this additional information, detailed studies of structural effects within the streptavidin macromolecules upon biotin binding become feasible. A potentially occurring band due to the bending mode of water is not strong enough to be directly identified, a weaker broad band might however be overlapping with the protein’s amide I band. Consequently, the amide I band may appear weaker than it actually is, and the shape of the observed band may be potentially altered. In order to investigate the protein structure, a more in-depth analysis is required.10
consequence of the islands’ shape. With typical dimensions of a = 22 nm and c = 4 nm of the oblate spheroidal particles determined by AFM measurements, the surface is increased by approximately 8%. Taking the surface density of ΓGSH = 1.2 × 10−10 mol/cm2 on a flat gold electrode reported by Fang and Zhou,25 this results in an effective surface density of ΓGSH,eff = 1.3 × 10−10 mol/cm2. Under the simplifying assumption of dilution of molecular groups in the surface layer contributing to the adsorption band, a linear correlation between band amplitude (or area) and the number of molecules was assumed. The resulting detection limit defined by the noise level corresponds to a surface density of (3.4 ± 0.2) × 10−11 mol/ cm2 or (10.5 ± 0.6) ng/cm2. In addition to the low LOD, the platform also enables the detailed monitoring of adsorption kinetics. The GSH monolayer growth can be described by a first-order Langmuir-type adsorption model26 for monolayer formation. Assuming no desorption (kdes ≈ 0) for the thiol adsorption on gold, the measured peak amplitudes are described by an exponential function of the form A(t ) = A 0[1 − exp(−kadsct )]
where c is the concentration of the bulk solution. Fitting this function results in an adsorption rate constant of kads = (8.6 ± 0.3) L/(mol s). Access to the vibrational information provides further details on the analyte’s chemical state, which cannot be gained from other methods like QCM or SPR. This is illustrated in Figure 3
Figure 3. In situ monitoring of dissociation of 1.2 nm thin GSH monolayer. (a) Schematic representation of dissociation reaction. (b) In situ IR microscopy spectra at pH 5 (black), pH 10 (blue), and pH 2 (red). All spectra referenced to spectrum at pH 5. C
DOI: 10.1021/acssensors.7b00902 ACS Sens. XXXX, XXX, XXX−XXX
Letter
ACS Sensors
facilitate the screening for novel drugs and targets, also in combination with additional downstream analytics. For highthroughput drug screening further development of the optofluidic platform and surrounding optics would be required. In the future, we aim to combine the platform with the novel concept of IR laser spectroscopy for even improved sensitivity and reduced measurement time. The platform’s analytical potential could also be extended by combining it with other optical methods like Raman and UV−vis spectroscopy. This integration will enable multimodal analyses of nanoliter samples, may increase identification rates, and give access to further relevant analyte parameters.
■
AUTHOR INFORMATION
Corresponding Author
*E-mail:
[email protected]. ORCID
Christoph Kratz: 0000-0002-4046-0760 Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS The authors thank I. Engler, Ö . Savas, and the ISAS workshop team for technical support. Financial support by the Ministerium für Innovation, Wissenschaft und Forschung des Landes Nordrhein-Westfalen, the Regierende Bürgermeister von Berlin − Senatskanzlei Wissenschaft und Forschung, the Bundesministerium für Bildung und Forschung, and the European Union through the EFRE program EFRE 1.8/13 is gratefully acknowledged.
■
REFERENCES
(1) Wu, J.; Gu, M. Microfluidic sensing: state of the art fabrication and detection techniques. J. Biomed. Opt. 2011, 16, 080901−080901− 12. (2) Fan, X.; White, I. M. Optofluidic microsystems for chemical and biological analysis. Nat. Photonics 2011, 5, 591−597. (3) TermehYousefi, A.; Bagheri, S.; Adib, N. Integration of biosensors based on microfluidic: a review. Sens. Rev. 2015, 35, 190−199. (4) Yin, D.; Deamer, D. W.; Schmidt, H.; Barber, J. P.; Hawkins, A. R. Single-molecule detection sensitivity using planar integrated optics on a chip. Opt. Lett. 2006, 31, 2136−2138. (5) Kühn, S.; Phillips, B.; Lunt, E. J.; Hawkins, A.; Schmidt, H. Ultralow power trapping and fluorescence detection of single particles on an optofluidic chip. Lab Chip 2010, 10, 189−194. (6) Dukor, R.; Chalmers, J.; Griffiths, P. Handbook of Vibrational Spectroscopy; John Wiley and Sons: New York, 2001. (7) Hoy, O.; Zdyrko, B.; Lupitskyy, R.; Sheparovych, R.; Aulich, D.; Wang, J.; Bittrich, E.; Eichhorn, K. J.; Uhlmann, P.; Hinrichs, K.; et al. Synthetic hydrophilic materials with tunable strength and a range of hydrophobic interactions. Adv. Funct. Mater. 2010, 20, 2240−2247. (8) Kroning, A.; Furchner, A.; Aulich, D.; Bittrich, E.; Rauch, S.; Uhlmann, P.; Eichhorn, K.-J.; Seeber, M.; Luzinov, I.; Kilbey, S. M.; et al. In Situ Infrared Ellipsometry for Protein Adsorption Studies on Ultrathin Smart Polymer Brushes in Aqueous Environment. ACS Appl. Mater. Interfaces 2015, 7, 12430−12439. (9) Ayankojo, A. G.; Tretjakov, A.; Reut, J.; Boroznjak, R.; Ö pik, A.; Rappich, J.; Furchner, A.; Hinrichs, K.; Syritski, V. Molecularly Imprinted Polymer Integrated with a Surface Acoustic Wave Technique for Detection of Sulfamethizole. Anal. Chem. 2016, 88, 1476−1484. (10) Furchner, A.; Kroning, A.; Rauch, S.; Uhlmann, P.; Eichhorn, K.J.; Hinrichs, K. Molecular Interactions and Hydration States of
Figure 4. In situ IR monitoring of streptavidin binding to biotin monolayer in microfluidic flow cell. (a) Time dependence of integrated peak area of amide I vibrational band during streptavidin adsorption. (b) Exemplary in situ IR spectra at different time point. Displayed spectra are referenced to a background spectrum recorded in PBS buffer, shifted for better comparison and smoothed using cubic spline interpolation.
In summary, the developed optofluidic platform enables nondestructive label-free measurements of immobilized submonolayers in nanoliter volumes, demonstrating a variety of sensing possibilities for small biomolecules. Submonolayer sensitivity toward identification and molecular structure was obtained for the direct binding of GSH. Submonolayer recognition of the model protein streptavidin on a biotinylated substrate showed the compatibility of the platform to a variety of surface functionalization techniques for sensor applications. Monitoring characteristic vibrational bands allows the investigation of adsorption kinetics as well as the elucidation of interactions in, e.g., receptor−ligand binding. The estimated LOD for molecular recognition is on the order of ng/cm2, which is comparable to other label-free techniques such as SPR and QCM based methods. The proposed optofluidic platform combines molecular detection with access to structural information, which is of high interest for protein and drug analysis. It offers the potential to, e.g., identify and quantify proteins and to evaluate their response to potential drugs or treatments. This possibility may D
DOI: 10.1021/acssensors.7b00902 ACS Sens. XXXX, XXX, XXX−XXX
Letter
ACS Sensors Ultrathin Functional Films at the Solid−Liquid Interface. Anal. Chem. 2017, 89, 3240−3244. (11) Baena, J. R.; Kulka, S.; Kaun, N.; Frank, J.; Svasek, P.; Moss, D.; Vellekoop, M. J.; Lendl, B. Mid-IR synchrotron radiation for molecular specific detection in microchip-based analysis systems. Anal. Bioanal. Chem. 2004, 378, 1735−1740. (12) Chan, K. L. A.; Gulati, S.; Edel, J. B.; de Mello, A. J.; Kazarian, S. G. Chemical imaging of microfluidic flows using ATR-FTIR spectroscopy. Lab Chip 2009, 9, 2909−2913. (13) Perro, A.; Lebourdon, G.; Henry, S.; Lecomte, S.; Servant, L.; Marre, S. Combining microfluidics and FT-IR spectroscopy: towards spatially resolved information on chemical processes. Reac. Chem. Eng. 2016, 1, 577−594. (14) Osawa, M. Surface-enhanced infrared absorption. Near-Field Optics and Surface Plasmon Polaritons 2001, 81, 163−187. (15) Aroca, R. F.; Ross, D. J.; Domingo, C. Surface-enhanced infrared spectroscopy. Appl. Spectrosc. 2004, 58, 324A−338A. (16) Adato, R.; Altug, H. In-situ ultra-sensitive infrared absorption spectroscopy of biomolecule interactions in real time with plasmonic nanoantennas. Nat. Commun. 2013, 4, 2154. (17) Neubrech, F.; Pucci, A.; Cornelius, T. W.; Karim, S.; GarcíaEtxarri, A.; Aizpurua, J. Resonant plasmonic and vibrational coupling in a tailored nanoantenna for infrared detection. Phys. Rev. Lett. 2008, 101, 157403. (18) Dong, L.; Yang, X.; Zhang, C.; Cerjan, B.; Zhou, L.; Tseng, M. L.; Zhang, Y.; Alabastri, A.; Nordlander, P.; Halas, N. J. Nanogapped Au Antennas for Ultrasensitive Surface-Enhanced Infrared Absorption Spectroscopy. Nano Lett. 2017, 17, 5768−5774. (19) Neuman, T.; Huck, C.; Vogt, J.; Neubrech, F.; Hillenbrand, R.; Aizpurua, J.; Pucci, A. Importance of Plasmonic Scattering for an Optimal Enhancement of Vibrational Absorption in SEIRA with Linear Metallic Antennas. J. Phys. Chem. C 2015, 119, 26652−26662. (20) Adato, R.; Aksu, S.; Altug, H. Engineering mid-infrared nanoantennas for surface enhanced infrared absorption spectroscopy. Mater. Today 2015, 18, 436−446. (21) Hinrichs, K.; Röseler, A.; Roodenko, K.; Rappich, J. Surfaceenhanced infrared absorption: Infrared ellipsometry of au evaporated ultrathin organic films. Appl. Spectrosc. 2008, 62, 121−124. (22) Xia, Y.; Whitesides, G. M. Soft Lithography. Annu. Rev. Mater. Sci. 1998, 28, 153−184. (23) Kratz, C.; Oates, T.; Hinrichs, K. Optimization and quantification of surface enhanced infrared absorption using gradient gold island films. Thin Solid Films 2016, 617, 33−37. (24) Picquart, M.; Grajcar, L.; Baron, M. H.; Abedinzadeh, Z. Vibrational spectroscopic study of glutathione complexation in aqueous solutions. Biospectroscopy 1999, 5, 328−337. (25) Fang, C.; Zhou, X. Voltammetry and EQCM investigation of glutathione monolayer and its complexation with Cu2+. Electroanalysis 2003, 15, 1632−1638. (26) Bieri, M.; Bürgi, T. Adsorption kinetics of L-glutathione on gold and structural changes during self-assembly: an in situ ATR-IR and QCM study. Phys. Chem. Chem. Phys. 2006, 8, 513−520. (27) Kanagaraja, S.; Alaeddine, S.; Eriksson, C.; Lausmaa, J.; Tengvall, P.; Wennerberg, A.; Nygren, H. Surface characterization, protein adsorption, and initial cell−surface reactions on glutathione and 3-mercapto-1,2,-propanediol immobilized to gold. J. Biomed. Mater. Res. 1999, 46, 582−591. (28) Aureau, D.; Ozanam, F.; Allongue, P.; Chazalviel, J.-N. The titration of carboxyl-terminated monolayers revisited: In situ calibrated Fourier transform infrared study of well-defined monolayers on silicon. Langmuir 2008, 24, 9440−9448. (29) Jiang, X.; Zuber, A.; Heberle, J.; Ataka, K. In situ monitoring of the orientated assembly of strep-tagged membrane proteins on the gold surface by surface enhanced infrared absorption spectroscopy. Phys. Chem. Chem. Phys. 2008, 10, 6381−6387. (30) Peluso, P.; Wilson, D. S.; Do, D.; Tran, H.; Venkatasubbaiah, M.; Quincy, D.; Heidecker, B.; Poindexter, K.; Tolani, N.; Phelan, M.; et al. Optimizing antibody immobilization strategies for the
construction of protein microarrays. Anal. Biochem. 2003, 312, 113− 124. (31) Venyaminov, S. Y.; Kalnin, N. Quantitative IR spectrophotometry of peptide compounds in water (H2O) solutions. II. Amide absorption bands of polypeptides and fibrous proteins in α-, β-, and random coil conformations. Biopolymers 1990, 30, 1259−1271. (32) Weber, P. C.; Ohlendorf, D.; Wendoloski, J.; Salemme, F. Structural origins of high-affinity biotin binding to streptavidin. Science 1989, 243, 85−88.
E
DOI: 10.1021/acssensors.7b00902 ACS Sens. XXXX, XXX, XXX−XXX
