Lab on a Chip
Published on 14 April 2014. Downloaded by North Dakota State University on 22/06/2014 18:41:55.
TECHNICAL INNOVATION
View Article Online View Journal
Print your own membrane: direct rapid prototyping of polydimethylsiloxane Cite this: DOI: 10.1039/c4lc00320a
Tim Femmer,ab Alexander J. C. Kuehne*b and Matthias Wessling*ab Polydimethylsiloxane is a translucent and biologically inert silicone material used in sealants, biomedical implants and microscale lab-on-a-chip devices. Furthermore, in membrane technology, polydimethylsiloxane represents a material for separation barriers as it has high permeabilities for various gases. The facile handling of two component formulations with a silicone base material, a catalyst and a small molecular weight crosslinker makes it widely applicable for soft-lithographic replication of two-dimensional device geometries, such as microfluidic chips or micro-contact stamps. Here, we develop a new technique to Received 13th March 2014, Accepted 14th April 2014 DOI: 10.1039/c4lc00320a www.rsc.org/loc
directly print polydimethylsiloxane in a rapid prototyping device, circumventing the need for masks or sacrificial mold production. We create a three-dimensional polydimethylsiloxane membrane for gas–liquid-contacting based on a Schwarz-P triple-periodic minimal-surface, which is inaccessible with common machining techniques. Direct 3D-printing of polydimethylsiloxane enables rapid production of novel chip geometries for a manifold of lab-on-a-chip applications.
1. Introduction Polydimethylsiloxane (PDMS) is easy to handle and there exist several two-component formulations on the market allowing for facile soft-lithographic replication and production of lab-on-a-chip devices. The hydrophobic silicone rubber is flexible, translucent and chemically inert as well as electrically insulating. Moreover, PDMS is permeable to gases, facilitating high transport rates, thus making PDMS a suitable material for various membrane applications.1 Twocomponent PDMS formulations are usually composed of a base silicone material, which is mixed with a small molecular crosslinker and a catalyst. For device fabrication via softlithography, the PDMS formulation is cast onto a mold and subsequently crosslinked thermally. After curing, the replica is simply peeled off the mold to recover the desired structure within the slab of PDMS.2,3 The mold represents the negative of the desired shape. Hence, the geometry and the fidelity of the replica depend on the quality of the mold and the removal technique. The described process limits the geometry of the desired structure to the dimensionality of the mold. As the PDMS structure needs to be recovered after replication, one is usually limited to two-dimensional designs with
a
Chemical Process Engineering, RWTH Aachen University, Turmstraße 46, 52064 Aachen, Germany. E-mail: [email protected]; Fax: +49 241 8092252; Tel: +49 241 8095488 b DWI – Leibniz Institute for Interactive Materials, RWTH Aachen University, Forckenbeckstraße 50, 52074 Aachen, Germany. E-mail: [email protected]
This journal is © The Royal Society of Chemistry 2014
low degrees of complexity. Three-dimensional devices have previously been fabricated by aligning multiple layers of two-dimensional channel geometries on top of each other.4 However, a direct and fast one-step fabrication method for complex and three-dimensional PDMS geometries remains inaccessible to date.5–7 The need for such three-dimensional PDMS devices has recently been identified and discussed for microfluidic devices.8 Photoinitiated rapid prototyping systems can be applied to fabricate three-dimensional structures from photoresist materials. A broad range of photoresist materials have been applied for rapid prototyping; these include acrylates,9 ABS-type polymers,10 ceramics filled polymers,11 hydrogels,12,13 and biodegradable and biocompatible resists.14,15 These materials have been prototyped into structures for applications in tissue engineering,13,16 chemical reaction compartments,17,18 medical implants19 and for jewelry.20,21 Despite several reports on photo-crosslinkable PDMS materials, direct rapid prototyping has not been presented, due to problems associated with very long curing times, low spatial resolution and dark-polymerization resulting from the low glass transition temperature of PDMS.22–24 Here, we report a fast one-step fabrication approach to process PDMS by digital light processing rapid prototyping. We formulate a PDMS photoresist material, which allows for printing of silicone structures of nearly any desired geometry. We demonstrate our approach by fabricating a threedimensional gas–liquid-contactor with a free-standing PDMS membrane enabling gas transport into the liquid phase. The geometry of a membrane is a crucial factor for its performance25 and membranes based on sheet-like ‘triple-periodic
Lab Chip
View Article Online
Technical Innovation
minimal surfaces’ exhibit superior transport in gas–liquidcontactors over other state-of-the-art membrane geometries.26
Published on 14 April 2014. Downloaded by North Dakota State University on 22/06/2014 18:41:55.
2. Results The digital light processing rapid prototyper (Perfactory Minimultilense from EnvisionTEC) produces a three-dimensional structure in a layer-by-layer fashion by photo-crosslinking the resist in defined areas. By this technique, each layer contributes a fixed voxel-height to the three-dimensional structure. We formulate the photoresist to suit the requirements of fast curing and the detailed resolution to obtain small voxel base areas (x and y-axis), low voxel-heights (z-axis) and acceptable line edge roughness (see Fig. 1a). The resist consists of 97.95 wt% (7–9 mol% methacryloxypropyl methylsiloxane)dimethylsiloxane copolymer, 2 wt% ethyl(2,4,6-trimethylbenzoyl)phenyl phosphinate (TPO-L) as a photoinitiator and 0.05 wt% Orasol Orange as a dye for increased resolution. The photo-initiated crosslinking reaction of the silicone is illustrated in Fig. 1b. The UV-vis absorption spectrum exhibits the characteristic peak of the photoinitator at 370 nm (see Fig. 2a). This band disappears upon exposure and crosslinking of the resist. The dye is added to prevent UV-light leaking out of the desired voxel size. UV-penetration deep into the photoresist layer would diminish resolution in all directions. For optimal compatibilization, the silicone and photoinitiator are mixed in tetrahydrofuran, which is subsequently removed in vacuo. The rapid prototyper is equipped with a 180 W Hg vapour lamp, which is reflected off a micromirror array, illuminating 1400 × 1050 pixels.16 The exposure pixel size is 30 × 30 μm, and we apply a cure time of 12 s to cure a 100 μm thick voxel. The broadband light source emits light between 400 and 550 nm with a maximum at 440 nm (ref. 14) and a brightness of 7 mW cm−2. After printing, the three-dimensional structure is developed by removal of
Fig. 1 (a) SEM image of a 3D-printed (gold sputtered) PDMS specimen. The specimen is cured layer-by-layer in the z-direction. Scale bar = 1 mm. (b) Molecular schematic for the photo-initiated crosslinking reaction of methacrylate functionalized PDMS.
Lab Chip
Lab on a Chip
the uncured resist material using 2-propanol and ultrasonication. The crosslinking reaction of the methacrylate functionalized PDMS depends on the concentration of the TPO-L photoinitiator and the excitation dose. Higher concentrations of TPO-L allow fast curing kinetics, however we are limited by the solubility of TPO-L in silicone. We apply the highest possible concentration of TPO-L in the present PDMS resist at 2 wt%. To determine the optimal exposure parameters for this PDMS resist, we perform dose-curve experiments with and without the Orasol Orange dye. For reasons discussed above, Orasol Orange clearly increases the resolution of the PDMS photoresist as can be seen in Fig. 2b.27 The voxel curing can be described as: Cd = Dp × ln(E/Ec) with Cd as the resulting curing depth, Dp as the penetration depth, both in μm, Ec as the critical exposure energy and E as the exposure energy, both in (mJ cm−2). The penetration depth (Dp) decreases from 593 to 293 μm when the dye is added, while the critical exposure energy increases slightly from 123 to 141 mJ cm−2 (Fig. 2b). The drop in Dp and increase in Ec allow for improved resolution during device fabrication. To investigate the limitations of our PDMS resist towards the smallest possible feature size, we expose lines with varying thickness l and height with an aspect ratio of 1 : 1, as shown in Fig. 2c. Features below 100 μm in size remain inaccessible, while features with sizes between 150 and 300 μm are well resolved, however at a reduced size. This process producing smaller than expected features is reproducible, so that it is possible to compensate for the size difference. Structures of 400 μm and greater show almost no difference between the desired ld and obtained lo line width. As described above, Dp is approximately three times the height of the exposed voxel, which means that during exposure of a voxel in one layer, over- and underlying voxels receive leaked exposure light. To obtain high resolution during multi-layer device fabrication, we expose at only 60% of Ec, which translates into an energy density of 84 mJ cm−2. This way, leaked exposure radiation will not lead to undesired features. By contrast, desired features receive the full gelation-dose through light leaking from the exposure of surrounding voxels. The resolution could be further improved by optimizing the exposure dose while reducing the step-size in the z-direction. To increase the sensitivity of the resist and further improve resolution, a dye/photoinitiator system with matching absorption profiles could be used. The dye would act as a photosensitizer, which provides energy transfer to the photoinitiator, facilitating a more efficient initiation of the crosslinking reaction. To investigate the performance of rapidly prototyped PDMS membranes, we fabricate a cross-flow gas–liquidcontactor based on a Schwarz-P triple-periodic minimalsurface structure.26 2 × 2 × 1 sheet-like TPMS cubic unit cells
This journal is © The Royal Society of Chemistry 2014
View Article Online
Published on 14 April 2014. Downloaded by North Dakota State University on 22/06/2014 18:41:55.
Lab on a Chip
Technical Innovation
Fig. 2 (a) UV/vis absorption spectra of thin films of the resist with dye (grey) and without dye (black), uncured (dashed lines) and cured (solid lines). The characteristic peak of the photoinitator at 370 nm vanishes upon exposure while the addition of dye results in an additional absorption A band between λ= 500 and 700 nm. (b) Dose curves for curing of PDMS with and without dye. (c) Line width resolution experiments with an aspect ratio of 1 : 1. The insets show micrographs of printed lines (lo) with parameters for desired widths of ld 100, 250 and 400 μm.
are produced by geometrical inflation. In this process, the interpenetrating channels are sealed and opened, respectively, on the liquid and gas-sides, as shown in Fig. 3. Chipto-world connections are introduced during the printing process, so that tubing for gas and liquid inlets and outlets can be simply coupled to the contactor device. Finally, a glass slide is used to seal the printed chip. To investigate the contacting performance, we observe the acidification of water using CO2. In the contactor device, CO2 needs to diffuse through the printed, 1 mm thick PDMS membrane into the liquid. To observe the change in pH, we add a bromothymol blue indicator to the aqueous phase. The pH indicator displays a blue color at pH values above 7.6 and changes to yellow when the CO2 reduces the pH-value to below 6.0. We determine pure gas permeabilities using time-lag measurements28 of 3D-printed flat-sheet membranes with a thickness of 840 μm. The permeabilities of nitrogen, oxygen and carbon dioxide are investigated and shown in Table 1. From the permeabilities, we derive the ideal selectivities against N2 and compare our results with the literature values of standard
Table 1 Permeability and selectivity data of different gases in our PDMS resist and PDMS from the literature29,30
Gas
Permeabilitya
Ideal selectivity over N2
Young's modulusb
N2 O2 CO2
235 (280) 523 (600) 2611 (3200)
— (—) 2.3 (2) 11.1 (12)
11.45 ± 2.20 (1.44 ± 0.20)
a Time-lag permeabilities measured at 35 °C, values in parentheses are taken from ref. 29 for PDMS RTV 615 measured at 35 °C, in barrer (1 barrer = 1 × 10−10 cm3STP cm−1 s−1 cm Hg−1). b Young's modulus in MPa, numbers in parenthesis represent values for Sylgard 184 with a ratio of 10 : 1, base to crosslinker.
PDMS membranes.29,30 Our printed PDMS membranes have similar selectivities but approximately 15% lower permeabilities compared to the standard PDMS membranes. This might be due to a higher crosslinking density in the present material, which is necessary to provide sufficient mechanical strength during and after the rapid prototyping process. This difference in crosslinking is shown also in the increased Young's modulus over the standard PDMS material (see Table 1). Despite the reduced performance of the material, the membrane device can be improved over current state-ofthe-art contactor membranes because our approach allows full freedom in the design of the contactor interface geometry.
3. Conclusions Fig. 3 3D-printed PDMS membrane contactor and artistic rendering of the respective TPMS structure (inset right). Gas flows horizontally from left to right while the bromothymol blue pH indicator is pumped vertically from the bottom to the top through the contactor. The color change of the pH indicator from blue to yellow indicates the CO2 transport over the membrane.
This journal is © The Royal Society of Chemistry 2014
We report the direct fabrication of 3D printed PDMSmembranes using a common digital light rapid prototyper. Our PDMS resist has properties comparable with the standard PDMS, although the PDMS photoresist appears coloured to increase detail resolution of the printed parts. We present the first direct rapid prototyped three-dimensional PDMS
Lab Chip
View Article Online
Technical Innovation
Published on 14 April 2014. Downloaded by North Dakota State University on 22/06/2014 18:41:55.
membrane with effective gas transfer properties. With the technique at hand, we can improve membrane devices and develop novel microfluidic chip geometries. This direct method allows the production of three-dimensional PDMS within the timeframe of only a few hours. Our approach offers new device geometries for various unprecedented applications.
Acknowledgements The authors thank M. Mueller and A.J. Benitez for support with elasticity measurements. M.W. acknowledges the support through an Alexander-von-Humboldt Professorship. This work was performed in part at the Center for Chemical Polymer Technology CPT, which is supported by the EU and the federal state of North Rhine-Westphalia (grant no. EFRE 30 00 883 02). The authors also acknowledge funding by the Excellence Initiative of the German federal and state governments through the I3TM Seed Fund Program and the cooperation with EnvisionTEC GmbH. We also thank BASF for donating TPO-L and Orasol Orange.
References 1 J. de Jong, R. G. H. Lammertink and M. Wessling, Lab Chip, 2006, 6, 1125–1139. 2 Y. Xia and G. M. Whitesides, Annu. Rev. Mater. Sci., 1998, 28, 153–184. 3 D. B. Wolfe, D. Qin and G. M. Whitesides, Microengineering in Biotechnology, Methods in Molecular Biology, ed. M. P. Hughes and K. F. Hoettges, Humana Press, New York, 2010, vol. 583, pp. 81–107. 4 A. Chen and T. Pan, Biomicrofluidics, 2011, 5, 046505. 5 J. C. Mcdonald, M. L. Chabinyc, S. J. Metallo, J. R. Anderson, A. D. Stroock and G. M. Whitesides, Anal. Chem., 2002, 74, 1537–1545. 6 A. Bonyár, H. Sántha, M. Varga, B. Ring, A. Vitéz and G. Harsányi, Journal of Material Forming, 2012, 1–8. 7 A. Bonyár, H. Sántha, B. Ring, M. Varga, J. Gábor Kovács and G. Harsányi, Procedia Eng., 2010, 5, 291–294. 8 A. Waldbaur, H. Rapp, K. Länge and B. E. Rapp, Anal. Methods, 2011, 3, 2681. 9 P. King and J. Covington, Procedia Chem., 2009, 1, 1163–1166.
Lab Chip
Lab on a Chip
10 M. S. Sulkin, E. Widder, C. Shao, K. M. Holzem, C. Gloschat, S. R. Gutbrod and I. R. Efimov, Am. J. Physiol., 2013, 305, H1569–H1573. 11 A. Butscher, M. Bohner, S. Hofmann, L. Gauckler and R. Müller, Acta Biomater., 2011, 7, 907–920. 12 J. Torgersen, X.-H. Qin, Z. Li, A. Ovsianikov, R. Liska and J. Stampfl, Adv. Funct. Mater., 2013, 23, 4542–4554. 13 T. Billiet, M. Vandenhaute, J. Schelfhout, S. Van Vlierberghe and P. Dubruel, Biomaterials, 2012, 33, 6020–6041. 14 T. M. Seck, F. P. W. Melchels, J. Feijen and D. W. Grijpma, J. Controlled Release, 2010, 148, 34–41. 15 F. P. W. Melchels, J. Feijen and D. W. Grijpma, Biomaterials, 2009, 30, 3801–3809. 16 F. P. W. Melchels, J. Feijen and D. W. Grijpma, Biomaterials, 2010, 31, 6121–6130. 17 A. J. Capel, S. Edmondson, S. D. R. Christie, R. D. Goodridge, R. J. Bibb and M. Thurstans, Lab Chip, 2013, 13, 4583–4590. 18 A. R. Romao Bineli, A. P. Gimenez Peres, A. L. Jardini, R. Maciel Filho, 6th Brazilian Conference on Manufacturing Engineering, ABCM, Caxias do Sul, 2011, entry 502. 19 A. D. Lantada and P. L. Morgado, Annu. Rev. Biomed. Eng., 2012, 14, 73–96. 20 C. Cheah, C. Chua, C. Lee, C. Feng and K. Totong, International Journal of Advanced Manufacturing Technology, 2004, 25, 308–320. 21 S. Kumar and S. Pityana, Adv. Mater. Res., 2011, 227, 92–95. 22 K. M. Choi and J. A. Rogers, J. Am. Chem. Soc., 2003, 125, 4060–4061. 23 K. Tsougeni, A. Tserepi and E. Gogolides, Microelectron. Eng., 2007, 84, 1104–1108. 24 W.-H. Chen, P.-C. Chen, S.-C. Wang, J.-T. Yeh, C.-Y. Huang and K.-N. Chen, J. Polym. Res., 2009, 16, 601–610. 25 J. M. Jani, M. Wessling and R. G. H. Lammertink, J. Membr. Sci., 2011, 378, 351–358. 26 A. Schoen, NASA Tech. Note, 1970. 27 D. W. Rosen, BioNanoFluidic MEMS, ed. P. J. Hesketh, Springer, New York, 2008, pp. 175–196. 28 M. Ajhar, Ph.D. thesis, RWTH Aachen University, 2011. 29 I. Blume, P. Schwering, M. Mulder and C. Smolders, J. Membr. Sci., 1991, 61, 85–97. 30 S. A. Stern, V. M. Shah and B. J. Hardy, J. Polym. Sci., Part B: Polym. Phys., 1987, 25, 1263–1298.
This journal is © The Royal Society of Chemistry 2014
Published on 14 April 2014. Downloaded by North Dakota State University on 22/06/2014 18:41:55.
TECHNICAL INNOVATION
View Article Online View Journal
Print your own membrane: direct rapid prototyping of polydimethylsiloxane Cite this: DOI: 10.1039/c4lc00320a
Tim Femmer,ab Alexander J. C. Kuehne*b and Matthias Wessling*ab Polydimethylsiloxane is a translucent and biologically inert silicone material used in sealants, biomedical implants and microscale lab-on-a-chip devices. Furthermore, in membrane technology, polydimethylsiloxane represents a material for separation barriers as it has high permeabilities for various gases. The facile handling of two component formulations with a silicone base material, a catalyst and a small molecular weight crosslinker makes it widely applicable for soft-lithographic replication of two-dimensional device geometries, such as microfluidic chips or micro-contact stamps. Here, we develop a new technique to Received 13th March 2014, Accepted 14th April 2014 DOI: 10.1039/c4lc00320a www.rsc.org/loc
directly print polydimethylsiloxane in a rapid prototyping device, circumventing the need for masks or sacrificial mold production. We create a three-dimensional polydimethylsiloxane membrane for gas–liquid-contacting based on a Schwarz-P triple-periodic minimal-surface, which is inaccessible with common machining techniques. Direct 3D-printing of polydimethylsiloxane enables rapid production of novel chip geometries for a manifold of lab-on-a-chip applications.
1. Introduction Polydimethylsiloxane (PDMS) is easy to handle and there exist several two-component formulations on the market allowing for facile soft-lithographic replication and production of lab-on-a-chip devices. The hydrophobic silicone rubber is flexible, translucent and chemically inert as well as electrically insulating. Moreover, PDMS is permeable to gases, facilitating high transport rates, thus making PDMS a suitable material for various membrane applications.1 Twocomponent PDMS formulations are usually composed of a base silicone material, which is mixed with a small molecular crosslinker and a catalyst. For device fabrication via softlithography, the PDMS formulation is cast onto a mold and subsequently crosslinked thermally. After curing, the replica is simply peeled off the mold to recover the desired structure within the slab of PDMS.2,3 The mold represents the negative of the desired shape. Hence, the geometry and the fidelity of the replica depend on the quality of the mold and the removal technique. The described process limits the geometry of the desired structure to the dimensionality of the mold. As the PDMS structure needs to be recovered after replication, one is usually limited to two-dimensional designs with
a
Chemical Process Engineering, RWTH Aachen University, Turmstraße 46, 52064 Aachen, Germany. E-mail: [email protected]; Fax: +49 241 8092252; Tel: +49 241 8095488 b DWI – Leibniz Institute for Interactive Materials, RWTH Aachen University, Forckenbeckstraße 50, 52074 Aachen, Germany. E-mail: [email protected]
This journal is © The Royal Society of Chemistry 2014
low degrees of complexity. Three-dimensional devices have previously been fabricated by aligning multiple layers of two-dimensional channel geometries on top of each other.4 However, a direct and fast one-step fabrication method for complex and three-dimensional PDMS geometries remains inaccessible to date.5–7 The need for such three-dimensional PDMS devices has recently been identified and discussed for microfluidic devices.8 Photoinitiated rapid prototyping systems can be applied to fabricate three-dimensional structures from photoresist materials. A broad range of photoresist materials have been applied for rapid prototyping; these include acrylates,9 ABS-type polymers,10 ceramics filled polymers,11 hydrogels,12,13 and biodegradable and biocompatible resists.14,15 These materials have been prototyped into structures for applications in tissue engineering,13,16 chemical reaction compartments,17,18 medical implants19 and for jewelry.20,21 Despite several reports on photo-crosslinkable PDMS materials, direct rapid prototyping has not been presented, due to problems associated with very long curing times, low spatial resolution and dark-polymerization resulting from the low glass transition temperature of PDMS.22–24 Here, we report a fast one-step fabrication approach to process PDMS by digital light processing rapid prototyping. We formulate a PDMS photoresist material, which allows for printing of silicone structures of nearly any desired geometry. We demonstrate our approach by fabricating a threedimensional gas–liquid-contactor with a free-standing PDMS membrane enabling gas transport into the liquid phase. The geometry of a membrane is a crucial factor for its performance25 and membranes based on sheet-like ‘triple-periodic
Lab Chip
View Article Online
Technical Innovation
minimal surfaces’ exhibit superior transport in gas–liquidcontactors over other state-of-the-art membrane geometries.26
Published on 14 April 2014. Downloaded by North Dakota State University on 22/06/2014 18:41:55.
2. Results The digital light processing rapid prototyper (Perfactory Minimultilense from EnvisionTEC) produces a three-dimensional structure in a layer-by-layer fashion by photo-crosslinking the resist in defined areas. By this technique, each layer contributes a fixed voxel-height to the three-dimensional structure. We formulate the photoresist to suit the requirements of fast curing and the detailed resolution to obtain small voxel base areas (x and y-axis), low voxel-heights (z-axis) and acceptable line edge roughness (see Fig. 1a). The resist consists of 97.95 wt% (7–9 mol% methacryloxypropyl methylsiloxane)dimethylsiloxane copolymer, 2 wt% ethyl(2,4,6-trimethylbenzoyl)phenyl phosphinate (TPO-L) as a photoinitiator and 0.05 wt% Orasol Orange as a dye for increased resolution. The photo-initiated crosslinking reaction of the silicone is illustrated in Fig. 1b. The UV-vis absorption spectrum exhibits the characteristic peak of the photoinitator at 370 nm (see Fig. 2a). This band disappears upon exposure and crosslinking of the resist. The dye is added to prevent UV-light leaking out of the desired voxel size. UV-penetration deep into the photoresist layer would diminish resolution in all directions. For optimal compatibilization, the silicone and photoinitiator are mixed in tetrahydrofuran, which is subsequently removed in vacuo. The rapid prototyper is equipped with a 180 W Hg vapour lamp, which is reflected off a micromirror array, illuminating 1400 × 1050 pixels.16 The exposure pixel size is 30 × 30 μm, and we apply a cure time of 12 s to cure a 100 μm thick voxel. The broadband light source emits light between 400 and 550 nm with a maximum at 440 nm (ref. 14) and a brightness of 7 mW cm−2. After printing, the three-dimensional structure is developed by removal of
Fig. 1 (a) SEM image of a 3D-printed (gold sputtered) PDMS specimen. The specimen is cured layer-by-layer in the z-direction. Scale bar = 1 mm. (b) Molecular schematic for the photo-initiated crosslinking reaction of methacrylate functionalized PDMS.
Lab Chip
Lab on a Chip
the uncured resist material using 2-propanol and ultrasonication. The crosslinking reaction of the methacrylate functionalized PDMS depends on the concentration of the TPO-L photoinitiator and the excitation dose. Higher concentrations of TPO-L allow fast curing kinetics, however we are limited by the solubility of TPO-L in silicone. We apply the highest possible concentration of TPO-L in the present PDMS resist at 2 wt%. To determine the optimal exposure parameters for this PDMS resist, we perform dose-curve experiments with and without the Orasol Orange dye. For reasons discussed above, Orasol Orange clearly increases the resolution of the PDMS photoresist as can be seen in Fig. 2b.27 The voxel curing can be described as: Cd = Dp × ln(E/Ec) with Cd as the resulting curing depth, Dp as the penetration depth, both in μm, Ec as the critical exposure energy and E as the exposure energy, both in (mJ cm−2). The penetration depth (Dp) decreases from 593 to 293 μm when the dye is added, while the critical exposure energy increases slightly from 123 to 141 mJ cm−2 (Fig. 2b). The drop in Dp and increase in Ec allow for improved resolution during device fabrication. To investigate the limitations of our PDMS resist towards the smallest possible feature size, we expose lines with varying thickness l and height with an aspect ratio of 1 : 1, as shown in Fig. 2c. Features below 100 μm in size remain inaccessible, while features with sizes between 150 and 300 μm are well resolved, however at a reduced size. This process producing smaller than expected features is reproducible, so that it is possible to compensate for the size difference. Structures of 400 μm and greater show almost no difference between the desired ld and obtained lo line width. As described above, Dp is approximately three times the height of the exposed voxel, which means that during exposure of a voxel in one layer, over- and underlying voxels receive leaked exposure light. To obtain high resolution during multi-layer device fabrication, we expose at only 60% of Ec, which translates into an energy density of 84 mJ cm−2. This way, leaked exposure radiation will not lead to undesired features. By contrast, desired features receive the full gelation-dose through light leaking from the exposure of surrounding voxels. The resolution could be further improved by optimizing the exposure dose while reducing the step-size in the z-direction. To increase the sensitivity of the resist and further improve resolution, a dye/photoinitiator system with matching absorption profiles could be used. The dye would act as a photosensitizer, which provides energy transfer to the photoinitiator, facilitating a more efficient initiation of the crosslinking reaction. To investigate the performance of rapidly prototyped PDMS membranes, we fabricate a cross-flow gas–liquidcontactor based on a Schwarz-P triple-periodic minimalsurface structure.26 2 × 2 × 1 sheet-like TPMS cubic unit cells
This journal is © The Royal Society of Chemistry 2014
View Article Online
Published on 14 April 2014. Downloaded by North Dakota State University on 22/06/2014 18:41:55.
Lab on a Chip
Technical Innovation
Fig. 2 (a) UV/vis absorption spectra of thin films of the resist with dye (grey) and without dye (black), uncured (dashed lines) and cured (solid lines). The characteristic peak of the photoinitator at 370 nm vanishes upon exposure while the addition of dye results in an additional absorption A band between λ= 500 and 700 nm. (b) Dose curves for curing of PDMS with and without dye. (c) Line width resolution experiments with an aspect ratio of 1 : 1. The insets show micrographs of printed lines (lo) with parameters for desired widths of ld 100, 250 and 400 μm.
are produced by geometrical inflation. In this process, the interpenetrating channels are sealed and opened, respectively, on the liquid and gas-sides, as shown in Fig. 3. Chipto-world connections are introduced during the printing process, so that tubing for gas and liquid inlets and outlets can be simply coupled to the contactor device. Finally, a glass slide is used to seal the printed chip. To investigate the contacting performance, we observe the acidification of water using CO2. In the contactor device, CO2 needs to diffuse through the printed, 1 mm thick PDMS membrane into the liquid. To observe the change in pH, we add a bromothymol blue indicator to the aqueous phase. The pH indicator displays a blue color at pH values above 7.6 and changes to yellow when the CO2 reduces the pH-value to below 6.0. We determine pure gas permeabilities using time-lag measurements28 of 3D-printed flat-sheet membranes with a thickness of 840 μm. The permeabilities of nitrogen, oxygen and carbon dioxide are investigated and shown in Table 1. From the permeabilities, we derive the ideal selectivities against N2 and compare our results with the literature values of standard
Table 1 Permeability and selectivity data of different gases in our PDMS resist and PDMS from the literature29,30
Gas
Permeabilitya
Ideal selectivity over N2
Young's modulusb
N2 O2 CO2
235 (280) 523 (600) 2611 (3200)
— (—) 2.3 (2) 11.1 (12)
11.45 ± 2.20 (1.44 ± 0.20)
a Time-lag permeabilities measured at 35 °C, values in parentheses are taken from ref. 29 for PDMS RTV 615 measured at 35 °C, in barrer (1 barrer = 1 × 10−10 cm3STP cm−1 s−1 cm Hg−1). b Young's modulus in MPa, numbers in parenthesis represent values for Sylgard 184 with a ratio of 10 : 1, base to crosslinker.
PDMS membranes.29,30 Our printed PDMS membranes have similar selectivities but approximately 15% lower permeabilities compared to the standard PDMS membranes. This might be due to a higher crosslinking density in the present material, which is necessary to provide sufficient mechanical strength during and after the rapid prototyping process. This difference in crosslinking is shown also in the increased Young's modulus over the standard PDMS material (see Table 1). Despite the reduced performance of the material, the membrane device can be improved over current state-ofthe-art contactor membranes because our approach allows full freedom in the design of the contactor interface geometry.
3. Conclusions Fig. 3 3D-printed PDMS membrane contactor and artistic rendering of the respective TPMS structure (inset right). Gas flows horizontally from left to right while the bromothymol blue pH indicator is pumped vertically from the bottom to the top through the contactor. The color change of the pH indicator from blue to yellow indicates the CO2 transport over the membrane.
This journal is © The Royal Society of Chemistry 2014
We report the direct fabrication of 3D printed PDMSmembranes using a common digital light rapid prototyper. Our PDMS resist has properties comparable with the standard PDMS, although the PDMS photoresist appears coloured to increase detail resolution of the printed parts. We present the first direct rapid prototyped three-dimensional PDMS
Lab Chip
View Article Online
Technical Innovation
Published on 14 April 2014. Downloaded by North Dakota State University on 22/06/2014 18:41:55.
membrane with effective gas transfer properties. With the technique at hand, we can improve membrane devices and develop novel microfluidic chip geometries. This direct method allows the production of three-dimensional PDMS within the timeframe of only a few hours. Our approach offers new device geometries for various unprecedented applications.
Acknowledgements The authors thank M. Mueller and A.J. Benitez for support with elasticity measurements. M.W. acknowledges the support through an Alexander-von-Humboldt Professorship. This work was performed in part at the Center for Chemical Polymer Technology CPT, which is supported by the EU and the federal state of North Rhine-Westphalia (grant no. EFRE 30 00 883 02). The authors also acknowledge funding by the Excellence Initiative of the German federal and state governments through the I3TM Seed Fund Program and the cooperation with EnvisionTEC GmbH. We also thank BASF for donating TPO-L and Orasol Orange.
References 1 J. de Jong, R. G. H. Lammertink and M. Wessling, Lab Chip, 2006, 6, 1125–1139. 2 Y. Xia and G. M. Whitesides, Annu. Rev. Mater. Sci., 1998, 28, 153–184. 3 D. B. Wolfe, D. Qin and G. M. Whitesides, Microengineering in Biotechnology, Methods in Molecular Biology, ed. M. P. Hughes and K. F. Hoettges, Humana Press, New York, 2010, vol. 583, pp. 81–107. 4 A. Chen and T. Pan, Biomicrofluidics, 2011, 5, 046505. 5 J. C. Mcdonald, M. L. Chabinyc, S. J. Metallo, J. R. Anderson, A. D. Stroock and G. M. Whitesides, Anal. Chem., 2002, 74, 1537–1545. 6 A. Bonyár, H. Sántha, M. Varga, B. Ring, A. Vitéz and G. Harsányi, Journal of Material Forming, 2012, 1–8. 7 A. Bonyár, H. Sántha, B. Ring, M. Varga, J. Gábor Kovács and G. Harsányi, Procedia Eng., 2010, 5, 291–294. 8 A. Waldbaur, H. Rapp, K. Länge and B. E. Rapp, Anal. Methods, 2011, 3, 2681. 9 P. King and J. Covington, Procedia Chem., 2009, 1, 1163–1166.
Lab Chip
Lab on a Chip
10 M. S. Sulkin, E. Widder, C. Shao, K. M. Holzem, C. Gloschat, S. R. Gutbrod and I. R. Efimov, Am. J. Physiol., 2013, 305, H1569–H1573. 11 A. Butscher, M. Bohner, S. Hofmann, L. Gauckler and R. Müller, Acta Biomater., 2011, 7, 907–920. 12 J. Torgersen, X.-H. Qin, Z. Li, A. Ovsianikov, R. Liska and J. Stampfl, Adv. Funct. Mater., 2013, 23, 4542–4554. 13 T. Billiet, M. Vandenhaute, J. Schelfhout, S. Van Vlierberghe and P. Dubruel, Biomaterials, 2012, 33, 6020–6041. 14 T. M. Seck, F. P. W. Melchels, J. Feijen and D. W. Grijpma, J. Controlled Release, 2010, 148, 34–41. 15 F. P. W. Melchels, J. Feijen and D. W. Grijpma, Biomaterials, 2009, 30, 3801–3809. 16 F. P. W. Melchels, J. Feijen and D. W. Grijpma, Biomaterials, 2010, 31, 6121–6130. 17 A. J. Capel, S. Edmondson, S. D. R. Christie, R. D. Goodridge, R. J. Bibb and M. Thurstans, Lab Chip, 2013, 13, 4583–4590. 18 A. R. Romao Bineli, A. P. Gimenez Peres, A. L. Jardini, R. Maciel Filho, 6th Brazilian Conference on Manufacturing Engineering, ABCM, Caxias do Sul, 2011, entry 502. 19 A. D. Lantada and P. L. Morgado, Annu. Rev. Biomed. Eng., 2012, 14, 73–96. 20 C. Cheah, C. Chua, C. Lee, C. Feng and K. Totong, International Journal of Advanced Manufacturing Technology, 2004, 25, 308–320. 21 S. Kumar and S. Pityana, Adv. Mater. Res., 2011, 227, 92–95. 22 K. M. Choi and J. A. Rogers, J. Am. Chem. Soc., 2003, 125, 4060–4061. 23 K. Tsougeni, A. Tserepi and E. Gogolides, Microelectron. Eng., 2007, 84, 1104–1108. 24 W.-H. Chen, P.-C. Chen, S.-C. Wang, J.-T. Yeh, C.-Y. Huang and K.-N. Chen, J. Polym. Res., 2009, 16, 601–610. 25 J. M. Jani, M. Wessling and R. G. H. Lammertink, J. Membr. Sci., 2011, 378, 351–358. 26 A. Schoen, NASA Tech. Note, 1970. 27 D. W. Rosen, BioNanoFluidic MEMS, ed. P. J. Hesketh, Springer, New York, 2008, pp. 175–196. 28 M. Ajhar, Ph.D. thesis, RWTH Aachen University, 2011. 29 I. Blume, P. Schwering, M. Mulder and C. Smolders, J. Membr. Sci., 1991, 61, 85–97. 30 S. A. Stern, V. M. Shah and B. J. Hardy, J. Polym. Sci., Part B: Polym. Phys., 1987, 25, 1263–1298.
This journal is © The Royal Society of Chemistry 2014
