Polymeric Infrared Compatible Microfluidic Devices for

Letter pubs.acs.org/ac

Polymeric Infrared Compatible Microfluidic Devices for Spectrochemical Analysis Michael V. Barich and Amber T. Krummel* Chemistry Department, Colorado State University, Fort Collins, Colorado 80523-1872, United States ABSTRACT: An innovative fabrication method is presented that affords the combination of polydimethyl-siloxane (PDMS) microfluidic technology with vibrational spectroscopy. PDMS devices are produced with uniform thicknesses ranging from 25 to 400 μm. The optical characteristics of the microfluidic devices in the mid-infrared are reported. The broad utility of this approach is demonstrated through IR imaging of flows in functional gradient generators and flow-focusing devices.

M

device, optical densities at a level similar to typical solvents used in IR absorption spectroscopy are readily achieved. In addition to fabrication details, the optical characteristics of the devices in the mid-IR and a demonstration of the utility of these devices for IR spectrochemical analysis are provided. The IR compatible microfluidic devices, shown in Figure 1a, combine four separate layers, each prepared prior to the assembly of the complete microchip. In Figure 1b, the four layers necessary for construction are shown: a top support layer, a microchannel layer, an adhesion layer, and an IR transparent backbone such as CaF2 used here. The support layer is created via standard soft lithography replica molding,24−26 as is illustrated in Figure 1c. The mold is designed to guide the placement of fluid ports. Polydimethylsiloxane (PDMS) is removed from this layer to produce direct fluid ports to the microfluidic channels and the desired IR measurement window. The second layer is the microchannel layer, an optically thin layer containing the microfluidic features. The features in the microchannel layer are molded using a technique similar to exclusion molding (Figure 1d).27,28 Hexamethyldisilazane (HMDS Sigma Aldrich) is used to modify the surface of the mold in order to facilitate the transfer of the microchannel layer. The microchannel feature heights define the optical path length used in IR transmission experiments. Path lengths ranging from 20 to 150 μm were fabricated and exhibited identical characteristics to traditionally fabricated microchannels; most notably, the microchannels were measured to be uniform across each chip and experience no measurable deformation under flow conditions. Channel deformation is a concern due to the overall thickness of the PDMS; however, pressures measured at constant flow rates

icrofluidic device technology has been broadly applied to drive developments in many biological and chemical arenas including diagnostics,1−3 chemical dynamics,4 polymer and nanomaterial synthesis,5−8 and reaction monitoring,9,10 to name a few. While the advancement of microfluidic device technology has been rapid, the combination of microfluidic device technology with a chemical selective read out such as Raman or infrared (IR) spectroscopy has been limited.11 Linear and nonlinear IR spectroscopies are ideal for monitoring reaction kinetics,12 molecular conformational dynamics,13,14 and solvent dynamics15,16 with chemical specificity. In order to take advantage of these spectroscopies for on-chip measurements, an IR transmissive microfluidic platform is required. This is a challenging problem because standard materials used for microfluidics strongly absorb mid-IR light. Infrared compatible microfluidic devices have previously been reported. Devices have been produced via etching IR transparent calcium fluoride (CaF2) windows,17,18 by patterning photoresist onto CaF2 windows,17,19−21 and by printing parafin wax on CaF2 windows to pattern the microchannels.22 These approaches are unfortunately limited in their use due to the difficulties in fabrication, limited device functionality, and higher fabrication costs. Recently, polymer based devices have also been developed for interfacing with IR light, but these approaches are generally restricted to attenuated total reflection (ATR) geometries rather than transmission geometries.22,23 Thus, the overall utility of microfluidic devices to the vibrational spectroscopy community is diminished. Herein, we report a robust fabrication approach to polymeric IR compatible microfluidic devices using standard soft lithography that, for the first time, allow for transmission IR spectroscopy and IR microscopy experiments to be performed. The optical density (OD) of a device is directly related to the amount of material the IR light must travel through; therefore, by precisely controlling the thickness of material for each © 2013 American Chemical Society

Received: August 15, 2013 Accepted: October 7, 2013 Published: October 7, 2013 10000

dx.doi.org/10.1021/ac4026016 | Anal. Chem. 2013, 85, 10000−10003

Analytical Chemistry

Letter

Figure 2. Comparison IR spectra of PDMS and 2H2O (a). Optical density of PDMS as a function of thickness at five vibrational frequencies of interest (b).

In Figure 2a, the IR absorption spectrum of a traditionally fabricated PDMS chip (red ∼2 mm thick) is compared to two IR transparent PDMS chips that are 240 μm (blue) and 125 μm (magenta) thick, respectively. In addition, two IR absorption spectra of bulk 2H2O in CaF2 sample cells with Teflon spacers of 50 μm (green) and 25 μm (cyan) are provided for reference. The background optical density of the IR transparent microfluidic devices is similar to that of 2H2O in the spectral range of 1450−2800 cm−1, a frequency range applicable to chemical and biological systems. For example, nitriles, azides, and metal carbonyls all have characteristic vibrational frequencies that lie between 2000 and 2800 cm−1. These particular molecular species are important due to their recent use as site-specific vibrational labels.29−33 The optical density of IR compatible microchips on CaF2 as a function of the total chip thickness is presented in Figure 2b. Total chip thickness is the combined thickness of the PDMS microchannel and adhesion layers. This relationship is plotted for five different frequencies often utilized in linear and nonlinear IR spectroscopy. Each frequency is associated with a vibrational mode of chemical interest. For example, the red data corresponds to 1650 cm−1; natural vibrational frequencies of nucleic acids, peptides, and lipids exist in the frequency range of 1500−1800 cm−1. Structure and dynamics of these biomolecules are routinely probed using their natural vibrational modes. Thus, it is imperative that microfluidic devices can be transmissive in this region. The blue data correspond to the vibrational frequencies of azido groups found between 2000 and 2100 cm−1. Azido groups are common reactive moieties in chemical synthesis and can be a good vibrational marker for

Figure 1. (a) Complete IR compatible microchip. (b) Overview of complete microchip assembly. The mask and fabrication method (c) for support layer and (d) for microchannel layer.

exhibited no fluctuations. Therefore, no local chip deformations are apparent. The examples shown contain an adhesion layer of PDMS, applied to a calcium fluoride (CaF2) window. The layers are plasma sealed to facilitate PDMS−PDMS covalent bonding in the order presented, from bottom to top, in Figure 1b. PDMS−PDMS bonding affords homogeneous wetting properties in the device channels and allows for surface modification within the chip. 10001

dx.doi.org/10.1021/ac4026016 | Anal. Chem. 2013, 85, 10000−10003

Analytical Chemistry

Letter

the outer edge of the focused stream. Infrared peak intensities scale with the path length and the square of the oscillator strength of the vibrational mode of interest. The path length tunability of these devices is discussed above. The model compounds NaN3 and NMA are used here for demonstration purposes; NaN3 is representative of site-specific vibrational labels, and NMA is representative of the amide I mode monitored during protein investigations. Typical concentrations of site-specific vibrational labels are 5−10 mM in the label.33,36−38 On the basis of the ΔOD detected in the NaN3 experiment, we fully expect these devices to perform well with NaN3 concentrations in the range typically used in labeling experiments. Typical protein concentrations used in infrared spectroscopy studies range from 0.5 to 10 mM often depending on the size of the protein. For example, several linear IR and 2D IR absorption experiments have been performed on ubiquitin, a protein with 76 amino acids, at concentrations ranging from 1 to 6 mM ubiquitin in buffer which thus corresponds to 76−456 mM in peptide backbone carbonyls.3,39−41 The concentration of NMA used in the experiments reported here is well within this range. Therefore, the IR compatible devices described here will be useful for probing peptide and protein conformational dynamics and kinetics with precision control over reagent mixing and reaction initiations in kinetics experiments. The IR images are collected in transmission mode, a capability only made possible by the IR translucence of these devices. The polymeric IR compatible microfluidic platform presented here affords the advantages of PDMS microfluidic technology to be combined with vibrational spectroscopy and microscopy techniques. In addition, the rapid fabrication and flexibility of these devices allows the channels to be designed to accommodate specific laser beam diameters and geometries. This technology will be particularly important when tailoring microfluidic devices that can be interfaced with pulsed, ultrafast IR laser experiments in a straightforward manner.

monitoring the completion of a chemical reaction. The black data correspond to 2575 cm−1, which is a good reference wavelength for these devices. This wavelength lies at the midpoint of a uniform absorption feature in the spectrum of the device. The optical densities of these devices at 2575 cm−1 correlates strongly to the total thickness of the devices as measured by optical profilometry. In general, a total chip thickness less than 200 μm is sufficient to attain IR compatibility for important carbonyl, azide, and nitrile vibrational modes; each of these chemical moieities has an oscillator strength that allows detection at low optical densities. Functioning IR compatible gradient generators (Figure 3a) and flow-focusing devices (Figure 3b) were fabricated in order

Figure 3. (a) Photograph of an IR compatible gradient generator. (b) Bright-field image of an IR compatible flow-focusing device. (c) IR chemical map of focused streams of 50 mM NaN3 in 2H2O (left) and 100 mM NMA in 2H2O (right).

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected].

to demonstrate the broad applicability of this fabrication approach. The flow-focusing devices double as drop-makers if two immiscible fluids are chosen. Each of these chips has a uniform channel height of 33.4 μm measured with an optical profilometer. In Figure 3c, IR images34,35 of flow-focusing devices with 82 μm uniform channel heights are shown. The IR images were collected under a 15× objective on a Bruker Hyperion 3000 IR microscope outfitted with a 64 × 64 element focal plane array detector. The device on the left focuses an inner stream of 50 mM sodium azide (NaN3) in 2H2O with an outer stream of pure 2H2O. On the right, an inner stream of 100 mM N-methylacetamide (NMA) in 2H2O is focused with an outer stream of pure 2H2O. The relative flow rates control the focused width of the inner stream. A flow rate ratio of 8:1 was employed in the azide experiment (left) and a ratio of 6:1 was used in the NMA experiment (right). Chemical maps are produced from the IR images; the intensities in the chemical maps are generated from integrating the total area under the peak in the IR spectra associated with the azide antisymmetric stretch centered at 2040 cm−1 (left) and the amide I vibration in NMA centered at 1625 cm−1 (right). The ΔOD in the center of the azide and NMA focused streams are 0.4 and 0.8 OD, respectively. The intermediate intensities represent ΔOD equal to 0.25 and 0.4 OD for the azide and NMA experiments, respectively. The intensities of each chemical map fall to zero at

Author Contributions

The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS A. T. Krummel is grateful for generous funding from Colorado State University. M. V. Barich acknowledges the generous support of the NSF IGERT Fellowship Program: Multidisciplinary Approaches to Sustainable Bioenergy (NSF grant, 0801707). The authors thank Professor Charles Henry for allowing us open access to the fabrication laboratory.

■

REFERENCES

(1) Seong, G. H.; Heo, J.; Crooks, R. M. Anal. Chem. 2003, 75, 3161−3167. (2) Chen, D.; Love, K. T.; Chen, Y.; Eltoukhy, A. A; Kastrup, C.; Sahay, G.; Jeon, A.; Dong, Y.; Whitehead, K. A; Anderson, D. G. J. Am. Chem. Soc. 2012, 134, 6948−6951. (3) Toepke, M. W.; Brewer, S. H.; Vu, D. M.; Rector, K. D.; Morgan, J. E.; Gennis, R. B.; Kenis, P. J. A.; Dyer, R. B. Anal. Chem. 2007, 79, 122−128.

10002

dx.doi.org/10.1021/ac4026016 | Anal. Chem. 2013, 85, 10000−10003

Analytical Chemistry

Letter

(4) Nichols, K. P.; Pompano, R. R.; Li, L.; Gelis, A. V.; Ismagilov, R. F. J. Am. Chem. Soc. 2011, 133, 15721−15729. (5) Wu, T.; Mei, Y.; Cabral, J. T.; Xu, C.; Beers, K. L. J. Am. Chem. Soc. 2004, 126, 9880−9881. (6) Li, W.; Greener, J.; Voicu, D.; Kumacheva, E. Lab Chip 2009, 9, 2715. (7) Puigmartí-Luis, J.; Rubio-Martínez, M.; Hartfelder, U.; Imaz, I.; Maspoch, D.; Dittrich, P. S. J. Am. Chem. Soc. 2011, 133, 4216−4219. (8) Martínez Arias, E. L.; Fazzio Martins, P.; Jardini Munhoz, A. L.; Gutierrez-Rivera, L.; Maciel Filho, R. Ind. Eng. Chem. Res. 2012, 51, 10755−10767. (9) Yue, J.; Schouten, J. C.; Nijhuis, T. A. Ind. Eng. Chem. Res. 2012, 51, 14583−14609. (10) Lubbeck, J. L.; Dean, K. M.; Ma, H.; Palmer, A. E.; Jimenez, R. Anal. Chem. 2012, 84, 3929−3937. (11) George, P. A.; Hui, W.; Rana, F.; Hawkins, B. G.; Smith, A. E.; Kirby, B. J. Opt. Express 2008, 16, 1577−1582. (12) Anna, J. M.; Baiz, C. R.; Ross, M. R.; McCanne, R.; Kubarych, K. J. Int. Rev. Phys. Chem. 2012, 31, 367−419. (13) Kolano, C.; Helbing, J.; Bucher, G.; Sander, W.; Hamm, P. J. Phys. Chem. B 2007, 111, 11297−11302. (14) Bredenbeck, J.; Helbing, J.; Sieg, A.; Schrader, T.; Zinth, W.; Renner, C.; Behrendt, R.; Moroder, L.; Wachtveitl, J.; Hamm, P. Proc. Natl. Acad. Sci. U. S. A. 2003, 100, 6452−6457. (15) Fayer, M. D. Annu. Rev. Phys. Chem. 2009, 60, 21−38. (16) Fenn, E. E.; Wong, D. B.; Fayer, M. D. J. Chem. Phys. 2011, 134, 54512. (17) Birarda, G.; Grenci, G.; Businaro, L.; Marmiroli, B.; Pacor, S.; Vaccari, L. Microelectron. Eng. 2010, 87, 806−809. (18) Pan, T.; Kelly, R. T.; Asplund, M. C.; Woolley, A. T. J. Chromatogr., A 2004, 1027, 231−235. (19) Hinsmann, P.; Frank, J.; Svasek, P.; Harasek, M.; Lendl, B. Lab Chip 2001, 1, 16−21. (20) Grenci, G.; Birarda, G.; Mitri, E.; Businaro, L.; Pacor, S.; Vaccari, L.; Tormen, M. Microelectron. Eng. 2012, 98, 698−702. (21) Kölhed, M.; Hinsmann, P.; Svasek, P.; Frank, J.; Karlberg, B.; Lendl, B. Anal. Chem. 2002, 74, 3843−3848. (22) Chan, K. L. A.; Kazarian, S. G. Anal. Chem. 2012, 84, 4052− 4056. (23) Chan, K. L. A.; Gulati, S.; Edel, J. B.; de Mello, A. J.; Kazarian, S. G. Lab Chip 2009, 9, 2909−2913. (24) Duffy, D. C.; McDonald, J. C.; Schueller, O. J.; Whitesides, G. M. Anal. Chem. 1998, 70, 4974−4984. (25) McDonald, J. C.; Duffy, D. C.; Anderson, J. R.; Chiu, D. T.; Wu, H.; Schueller, O. J. A.; Whitesides, G. M. Electrophoresis 2000, 21, 27− 40. (26) McDonald, J. C.; Whitesides, G. M. Acc. Chem. Res. 2002, 35, 491−499. (27) Jo, B.-H.; Beebe, D. J. SPIE Conf. Microfluid. Devices Syst. II 1999, 3877, 222−229. (28) Folch, A.; Jo, B. H.; Hurtado, O.; Beebe, D. J.; Toner, M. J. Biomed. Mater. Res. 2000, 52, 346−353. (29) King, J. T.; Kubarych, K. J. J. Am. Chem. Soc. 2012, 134, 18705− 18712. (30) Urbanek, D. C.; Vorobyev, D. Y.; Serrano, A. L.; Gai, F.; Hochstrasser, R. M. J. Phys. Chem. Lett. 2010, 1, 3311−3315. (31) Getahun, Z.; Huang, C.-Y.; Wang, T.; De León, B.; DeGrado, W. F.; Gai, F. J. Am. Chem. Soc. 2003, 125, 405−411. (32) Krummel, A. T.; Zanni, M. T. J. Phys. Chem. B 2008, 112, 1336− 1338. (33) Dutta, S.; Cook, R. J.; Houtman, J. C. D.; Kohen, A.; Cheatum, C. M. Anal. Biochem. 2010, 407, 241−246. (34) Koenig, J. L.; Wang, S. Q.; Bhargava, R. Anal. Chem. 2001, 73, 360A−369A. (35) Lewis, E. N.; Treado, P. J.; Reeder, R. C.; Story, G. M.; Dowrey, A. E.; Marcott, C.; Levin, I. W. Anal. Chem. 1995, 67, 3377−3381. (36) Watson, M. D.; Gai, X. S.; Gillies, A. T.; Brewer, S. H.; Fenlon, E. E. J. Phys. Chem. B 2008, 112, 13188−13192.

(37) Tucker, M. J.; Gai, X. S.; Fenlon, E. E.; Brewer, S. H.; Hochstrasser, R. M. Phys. Chem. Chem. Phys. 2011, 13, 2237−2241. (38) Londergan, C. H.; Wang, J.; Axelsen, P. H.; Hochstrasser, R. M. Biophys. J. 2006, 90, 4672−4685. (39) Wagner, C.; Buchegger, W.; Vellekoop, M.; Kraft, M.; Lendl, B. Anal. Bioanal. Chem. 2011, 400, 2487−2497. (40) Kakuta, M.; Hinsmann, P.; Manz, A.; Lendl, B. Lab Chip 2003, 3, 82−85. (41) Chung, H. S.; Ganim, Z.; Jones, K. C.; Tokmakoff, A. Proc. Natl. Acad. Sci. 2007, 104, 14237−14242.

10003

dx.doi.org/10.1021/ac4026016 | Anal. Chem. 2013, 85, 10000−10003