Confined Gelatin Dehydration as a Viable Route To Go Beyond

May 3, 2016 - Nowadays, microfluidic channels of a few tens of micrometers are required and widely used in many fields, especially for surface-process...
0 downloads 6 Views 8MB Size
Download PDF
Research Article www.acsami.org

Confined Gelatin Dehydration as a Viable Route To Go Beyond Micromilling Resolution and Miniaturize Biological Assays Raffaele Vecchione,*,†,§ Gabriele Pitingolo,†,§ Andrea P. Falanga,†,§ Daniela Guarnieri,†,§ and Paolo A. Netti†,‡,§ †

Center for Advanced Biomaterials for Healthcare, Istituto Italiano di Tecnologia (IIT@CRIB), Largo Barsanti e Matteucci, 53, Naples 80125, Italy ‡ Department of Chemical, Materials and Industrial Production Engineering, University of Naples Federico II, Naples 80125, Italy § Interdisciplinary Research Center on Biomaterials (CRIB), University of Naples Federico II, Naples 80125, Italy S Supporting Information *

ABSTRACT: Nowadays, microfluidic channels of a few tens of micrometers are required and widely used in many fields, especially for surface-processing applications and miniaturization of biological assays. Herein, we selected micromilling as a low-cost technology and proposed an approach capable of overcoming its limitations; in fact, microstructures below 20− 30 μm in depth are difficult to obtain, and the manufacturing error is rather high, as it is inversely proportional to the depth. Indeed, the proposed method uses a confined dehydration process of a patterned gelatin substrate fabricated via replica molding onto a micromilled poly(methyl methacrylate) substrate to produce a gelatin master with demonstrated final micrometric features down to 3 μm for the channel depth and, in specific configurations, down to 5 μm for the channel width. Finally, we demonstrated the ability to flux liquids in miniaturized microfluidic devices and fabricated and testedas an examplemicrometric microstructures arrays connected via microchannels for biological assays. KEYWORDS: gelatin dehydration, microfluidics, micromilling, miniaturization, brain endothelial cells

1. INTRODUCTION The microfabrication of polymer microfluidic chips is of growing interest in several application fields, from chemical processing to biotechnology, pharmaceutics, and food manufacturing because of the benefits offered by miniaturized platforms.1−3 In the past decade many research groups have investigated integrated microsystems, the so-called micrototal analysis systems (μ-TAS) and lab on a chip (LOC).4−6 There is a strong need for microfluidic channels less deep than 50 μm in several fields such as synthesis, diagnostics, and biology.7 For example, Hwang et al. used a 16 μm deep microchannel to synthesize nonspherical magnetic hydrogel microparticles.8 Bousse et al. developed a microfabricated analytical device on a glass chip (13 μm deep and 36 μm wide channel) that performs a protein sizing assay.9 In the “organ on a chip” field, very small microchannels are used to emulate microcapillary systems.10 Small microchannels are useful in microfluidics for high-throughput screening applications, like the screening station from Caliper Life Sciences.11 In addition, recently, many research groups have developed microfluidics platforms for manipulating and analyzing single cells, owing to their ability to confine cells within microstructures with comparable dimensions to those of the cells.12,13 Photolitography has been widely exploited in the production of microfluidic systems with © XXXX American Chemical Society

controlled features at the microscale but it requires costly and complex equipment.14 The photolithography requires masks and it becomes convenient only in the case of multiple reproduction of the same device configuration. Alternatively, some works on shrinking approaches to scale down the size of microstructures have been recently reported.15−17 For example, Focaroli et al. described a not confined shrinking approach based on the agarose shrinkage of microchannels of 0.8 mm in width, focusing on the lateral reduction of the microchannel.15 Das et al. described a method that provides a tridimensional size reduction by successive miniaturizations through the shrinkage of hydrogels.16 In principle, both methods could be exploited for the reduction of microchannel depths; however, not confined shrinkage typically promotes global deformation and limited reduction in the depth of the final structures due to the lateral shrinkage contribution. To the best of our knowledge, no strategy for a controlled and confined shrinkage of very small microchannels has so far been proposed. Micromilling is a low-cost and out of clean room manufacturing method for the preparation of polymeric microfluidic devices Received: April 7, 2016 Accepted: May 3, 2016

A

DOI: 10.1021/acsami.6b04128 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces

Clone S.p.A. (Milan, Italy). Red fluorescent hypo needles (WGA) and Alexa488-phalloidin were from Gibco Invitrogen. 2.2. Fabrication Methodology. The microfabrication steps to reduce the depth of the microchannels are outlined in Figure 1. The

and is particularly indicated for prototyping.18 One relevant limitation of micromilling is its inability to scale the microstructure depths down to a few micrometers.19 In addition, already under 50 μm the reproducibility of the micromilling process is quite low, as described in the Results and Discussion section. Therefore, the need for an easy, effective, reproducible fabrication technology of micrometric channel depths, not requiring complex and costly equipment, is still unfulfilled. To this aim, in this article, we present the confined gelatin dehydration (CGD) method that combines micromilling, as low-cost equipment, with gelatin dehydration in a laterally confined bowl to reduce the initial depth of microchannels, ensuring the preservation of the final device planarity. We compared tens of micrometers-sized microchannels (e.g., 20 μm) with same size microchannels directly fabricated by micromilling and analyzed the capability to scale down the depth to a few micrometers by modulating initial PMMA microchannel depth and gelatin concentration. We also adapted the proposed CGD method to a designed array of pillars to scale down (to 5 μm) the width of miniaturized channels interconnecting chambers. To demonstrate the efficacy of our method, the fabricated microfluidic channels were tested in terms of leakage by letting a red fluorescent liquid flow inside the bonded device; no leakage occurred. In addition, we tested peculiar patterns characterized by arrays of areas interconnected by miniaturized microchannelsobtained with the proposed methodfor confined cell culture. The geometry and quality of the polydimethylsiloxane (PDMS) microchannels fabricated from the master obtained by the dehydration process were characterized using optical microscopy and scanning electron microscopy (SEM), while a profilometer was used to characterize microchannel depth and roughness.

Figure 1. Schematic illustration of gelatin dehydration process involved in the preparation of dehydrated negative master with reduced depths. microchannels were fabricated by micromachining PMMA substrates with the above-mentioned micromilling machine. Micromilling is a subtractive microfabricating process that uses rotating cutting tools to remove polymer from a substrate.18,20 Such technology has two main functions: the fabrication of masters used in subsequent microfabrication steps and the direct fabrication of the microstructures of the final device; in this paper we explored the former. To correlate initial and final depths, we micromilled micrometer-sized microchannels at different width/depth ratios: 380:200, 380:100, 380:50, and 380:25 μm. Moreover, to show the applicability of the described approach in terms of aspect ratios, we micromilled microchannels 100:100 width/depth using a 100 μm microtool and micropillars, 350:100 μm (width/depth) using a 250 μm microtool. A PMMA bowl (30 × 30 mm and 2 mm depth) was fabricated and then used to prepare the hydrogel replica needed to form the final master. To design a draft of the microstructures, we made a layout by Draftsight (Cad Software). During micromilling, spindle speed, feed speed, and plunge rate per pass were set at 10 000 rpm, 15 mm s−1, and 20, respectively. From PMMA masters with negative features, PDMS positive replica masters were rapidly prepared using the PDMS thermal curing technique.21,22 A typical process involved pouring PDMS mixed at a ratio of 10:1 with curing agent onto the PMMA microchannel master (Figure 2A). Alternatively, a hydrophilic PDMS was used to facilitate PDMS prepolymer penetration in the case of more complex structures, such as the array of micropillars.23 In this case the PDMS mixture was then mixed with Silwet L-77 at 4 wt %. In both cases, PDMS precursor had been previously exposed to vacuum to eliminate air bubbles for at least 30 min. The PMMA master with PDMS poured was baked in the oven (1 h, 75 °C), to finalize the PDMS curing process (Figure 2B). Afterward, the PMMA bowl was filled with liquid gelatin at defined concentrations (Figure 2C), previously degassed with nitrogen for 10 min to eliminate bubbles. The hydrophilic or hydrophobic PDMS positive replica was placed

2. EXPERIMENTAL SECTION 2.1. Materials and Equipment. The poly(methyl methacrylate) (PMMA) substrates used in this study, thickness 1.2 mm, were purchased from Goodfellow Cambridge Ltd., (Cambridge, UK) and belonged to the same batch of supplied PMMA substrates. PDMS prepolymer and curing agent (Sylgard 184 elastomer kit) were purchased from Dow Corning Corporation (Midland, MI, USA). Silwet L77 was purchased from Helena Chemical Company (Collierville, TN, USA). Gelatin from porcine skin type A was obtained from Sigma-Aldrich (St. Louis, USA). The microtools used in the microfabrication process were three-flute end mills with different diameters: 380, 250, and 100 μm purchased from Performance Micro Tool (Janesville, WI, USA). Sulforhodamine B sodium salt was obtained from Sigma-Aldrich (USA). Silicone peroxide tubing/60 Shore i.d. 0.75 mm were obtained from IDEX Health & Science Gmbh (Oak Harbor, WA, USA). Hypo Needles 18 AWG were purchased from Warner Instruments LLC (Hamden, CT, USA). Vacucenter VC20/50 were obtained from Salvislab (Rotkreuz, Switzerland). A micromilling machine was purchased from Minitech Machinery Corporation (Norcross, GA, USA). Syringe pump 11 Helite Series was purchased from Harvard Apparatus (Holliston, MA, USA). Fabricated microchannels were inspected with a scanning electron microscope (SEM) (Ultraplus Zeiss) and a profilometer (Veeco Dektak 150). Mouse cerebral endothelial cells, bEnd.3 cells, were purchased from American Type Culture Collection (Manassas, VA). Dulbecco’s modified Eagle’s medium (DMEM) with high glucose, trypsin, Dulbecco’s phosphate-buffered saline (DPBS), fetal bovine serum (FBS), penicillin, and streptomycin were purchased from Gibco Invitrogen (Grand Island, NY, USA). Sodium pyruvate and 4′,6diamidino-2-phenylindole (DAPI) were purchased from Sigma-Aldrich (USA). Glutamine was purchased from Lonza Group Ltd. (Basel, Switzerland). Nonessential amino acids were purchased from EuroB

DOI: 10.1021/acsami.6b04128 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces

and vacuum for 1 h at 37 °C. Afterward, the microchannels were washed with PBS to remove fibronectin excess. The glass coverslip was removed under sterile conditions and 100 μL of brain endothelial bEnd.3 cell suspension, at a density of 2 × 105 cells/mL in a cell culture medium (DMEM + 10% FBS), were seeded on the array. After incubation for cell attachment in a 5% humidified CO2 incubator at 37 °C, nonadhered cells were removed 2 h after seeding by washing them twice with fresh cell culture medium. Afterward, cells were allowed to grow for 3 d within the device. At this time point, cells were fixed with 4% paraformaldehyde for 10 min and actin microfilaments and nuclei were stained with Alexa488-phalloidin (Molecular Probes) and DAPI (Sigma), respectively. Samples were observed by a multiphoton confocal microscope (Leica TCS SP5MP) equipped with a 20× dry objective.

3. RESULTS AND DISCUSSION 3.1. General Observations. Motivated by the strong need in μ-TAS and LOC for microchannels thinner than 50 μm, we developed an innovative shrinking approach to obtain microchannels of micrometric depths. Micromilling is one of the oldest microfabrication techniques and to date one of the most used in laboratories around the world.27,28 At the same time several shrinking methods proposed by different research groups have attempted to improve existing microfabrication technologies.15,16 In this respect, the main goal of our study was to develop an easy and reproducible method to reduce the manufactured depth associated with micromilling from tens of micrometers to a few micrometers. As discussed in section 3.3, we also studied the capability of the proposed method to reduce the fabrication error in the case of depths that can be directly fabricated by micromilling. The investigated method uses mechanical micromachining and soft lithography techniques that are easy, reproducible, and inexpensive; it is based on the dehydration of a gelatin master inside a laterally confined bowl. This confinement is the reason for the vertical shrinkage of the hydrogel, including the microchannel depth and the preservation of the planarity of the final device; also, it avoids any undesired global deformation typical of traditional 3D shrinkage techniques (Figure S2). During the experimental study, we observed that the microchannels obtained using the process, described in Figure 1, show an enlargement in width of the final microchannel, though with acceptable aspect ratios. Conversely, by dehydrating a positive gelatin master, it was possible to obtain microchannels with a shrunk width and a lower dimensional aspect ratio (width/depth). The first advantage of our gelatin dehydration process is its simplicity. The photolithographic process comprises spin-coating, prebaking, UV exposure, postbaking, and developing.29 In contrast, the proposed method requires only a micromilling machine, or other simple mill technologies; the process is finalized within an oven with controlled temperature (28 °C) and vacuum. The oven step can also be skipped, affecting only the dehydration time, but not the final result (36 vs 14 h). Gelatin masters could be stored for months and after reusing tens of units of time, as long as we needed, it was still intact for further uses. 3.2. Depth Shrinking by Dehydration Process. As a first experimental study, a gelatin solution (12% w/v) was used to obtain a gelatin negative master with the imprint of the initial PMMA master at different width/depth ratios. The initial depths of the microchannels after micromilling fabrication were, respectively, 240, 160, 55, and 30 μm, as reported in Figure 3. To characterize these depths, PDMS replicas fabricated directly from the initial PMMA masters were measured by using a profilometer. For the profilometer

Figure 2. Preparation of dehydrated gelatin negative master. Pouring of the liquid PDMS prepolymer onto the master (A) to fabricate a PDMS positive replica (B); pouring of the liquid gelatin into a homemade PMMA bowl (30 × 30 × 2 mm) (C); gelatin negative master preparation by positioning PDMS positive replica onto gelatinfilled bowl before and after gelling (D, E); final dehydrated gelatin negative master inside (F) and outside (G,H) the PMMA bowl. onto the gelatin fluid bed to obtain a gelatin negative master (Figure 2D). After a few minutes to stabilize the liquid gelatin the system was put into the fridge for 20 min at 4 °C, to finalize the gelling process. After gelling, the PDMS was peeled off from the negative master ready for the dehydration process (Figure 2E). The gelatin master with the imprint of the PDMS master was then dried in a controlled environment (temperature 28 °C for 14 h at 200 mbar) (Figure 2G) in which the gel sample dried slowly but uniformly without waves and deformations on the surface (Figure 2H). Alternatively, in the absence of an oven, the samples can be left at room temperature for 36 h. The entire double replica21,24,25 procedure, using hydrophilic PDMS in the case of micropillars, was repeated to prepare final PDMS microchannels and microstructures arrays with reproducible and down to 3 μm depth, limiting the curing temperature of the PDMS at 37 °C to preserve dehydrated gelatin negative master in the first replica. Curing time was longer in this case; it took overnight. To obtain microchannels with aspect ratios (width/depth) down to 3, we explored the reverse process, where the only difference was to imprint gelatin directly with negative PMMA master to prepare a positive master. To demonstrate the applicability of the PDMS microchannels replicated onto dehydrated negative gelatin master, these were bonded through oxygen plasma activation to a glass slide using a plasma chamber (Plasma prep II, SPI) for 1 min at a pressure of 0.3 mbar and power of 37 W.26 2.3. Brain Endothelial bEnd.3 Cells Culture. To allow cell adhesion selectively, the PDMS array was clamped with a glass coverslip by using reversible magnetic bonding (Figure S1 of the Supporting Information). The fabricated microstructure was sterilized with 10% penicillin−streptomycin solution in PBS before cell seeding for 24 h at 4 °C to prevent contamination. The rhomboidal areas of the structure within the array were selectively coated with 10 μg/mL fibronectin (Sigma-Aldrich) through capillarity-induced penetration C

DOI: 10.1021/acsami.6b04128 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces

final depth both in the case of 12% and 18% w/v of gelatin concentration. A robust predictive model, taking into account relevant parameters including gelatin concentration and interdistance between multiple structures, is under investigation. In addition, as expected, we observed that when higher gelatin concentrations (from 12% to 18% w/v) were used, the standard deviation value decreased; indeed, less water had to be removed during the process, providing a higher level of control of the process. We also showed the capability to fabricate final microchannels with aspect ratios down to 3:1, dehydrating a gelatin master with a protruding microchannel 100:100 μm (width/depth). SEM images show microchannels made of PDMS and obtained respectively by replica from a micromilled PMMA master (100 × 100 μm) (Figure 4 A,C) and from a

Figure 3. Effect of initial depth on shrunk depth at different gelatin concentrations. Profilometer values plotted in the graph showing the shrunk depths obtained from 240, 160, 55, and 30 μm microchannel depths, respectively. Shrunk depths and related initial depths are reported in Table 1. Trend lines are displayed in the graph.

characterization we used a standard scan type on a hills profile, a stylus type radius of 2.5 μm, and a Meas range of 65.5 or 6.5 μm, depending on the final depth. To quantify the depth reduction obtained with the proposed approach, we profiled the PDMS replicas fabricated from the dehydrated gelatin masters (Figure S3). The gelatin master (Figure 2 F) obtained after dehydration showed about 88% of shrinkage in depth from the size of freshly fabricated gelatin master, as shown by the profile graphs (Figure 3). Table 1 reports the shrinking depth obtained Figure 4. SEM images of PDMS microchannels replicated directly from a micromilled microchannel master (100 × 100 μm) (A, C) and from a dehydrated replica obtained using the proposed method in the version of negative and positive gelatin masters (B, D).

Table 1. Shrunk Depth at 12% and 18% w/v Gelatin Concentration no.

initial depth (μm)

shrunk depth gel 18% (μm)

1 2 3 4 5 6 7 8 9 10 11 12

240 240 240 160 160 160 55 55 55 30 30 30

37 35 34 28 26 26 15 17 17 12 13 12

SD gel 18% (μm) 1.52

1.15

1.15

0.57

shrunk depth gel 12% (μm) 31 26 26 19 21 22 8.9 5.9 5.6 3.1 2.1 3.2

SD gel 12% (μm) 2.64

dehydrated gelatin master obtained from the same PMMA master using the two processes proposed (Figure 4B,D); the depth reduction and the preservation of the frame thanks to the hydrogel confinement in a PMMA bowl are clearly visible, different from the shrinkage methods reported in the literature. Final aspect ratios were 20:1 and 3:1, respectively, as indicated in Figure 4. In addition, our approach is effective in reducing the surface roughness and in manufacturing round profiles. Specifically, roughness profiles by profilometer showed an enhancement of the smoothness since roughness decreased from 156 to 24 nm (Figure S4). This significant reduction is highly desirable in microfluidics and it can also be advantageous from an optical standpoint since the final microchannels result in being much more transparent and internally inspectable. To show the versatility of the described innovative approach, we prepared different discrete microstructures, such as micropillars (750 μm in diameter), and we showed the possibility to reduce the initial depth using the proposed process (Figure 5A,B). Taking advantage of the enlargement of the pillars, we fabricated microchamber arrays connected via 5 μm wide microchannels (Figure 5D,F) starting from arrays of micropillars (350 μm in diameter) (Figure 5C,E). The procedure previously described can be further miniaturized by reducing the diameter of the pillars and the distance between each of them. As reported in the Supporting Information (Figure S5),

1.52

1.8

0.6

by the differently sized microchannels. At the end of the process, a depth reduction of ∼88% was confirmed for each microchannel. The results summarized in Table 1 and Figure 3 show an almost linear dependence of the shrunk depth on the initial depth; a minimum microchannel depth of ∼3 μm was obtained by a 30 μm depth microchannel. Then we tested the same microchannels using another gelatin concentration corresponding to 18% w/v. In Figure 3, shrunk depths are plotted as a function of the initial depths; also in these cases an almost linear trend was obtained with 85% shrinkage with similar slope. From the trend lines displayed in Figure 3 it is possible to estimate the initial depth needed to obtain a desired D

DOI: 10.1021/acsami.6b04128 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces

Table 2. Direct Micromilling Fabrication Compared to the Dehydration Process To Obtain a Microchannel 20 μm Depth

Figure 5. SEM images of PDMS replicas, before (A, C, E) and after (B, D, F) the dehydration process in microstructures like pillars.

micropillars present a shrinkage in the depth from 100 to 14 μm and after shrinkage the corners are rounded. 3.3. Micromilling versus Dehydration Process. In this section, we provide a comparison, in terms of final error in preparing the same target depth, between direct fabrication by micromilling and the proposed method based on fabrication by micromilling combined with confined gelatin dehydration. Several microfluidic applications require a high level of precision in the microfabrication process for a faithful reproduction of the initial CAD/CAM project. The error of the micromilling process, including the z axis, has been reduced in recent years thanks to a higher stage control. However, part of the error in the final depth is due to the manual alignment of the microtool in z of the workpiece, needed to define the coordinate origin.19,30,31 Despite some alignment techniques recently proposed, the fabrication is still affected by an error in z whose impact depends on the final depth: the lower the depth, the higher the relative percentage error. Here, we compared the direct fabrication by micromilling with the proposed method in terms of reproducibility in producing a depth target of 20 μm. The results are summarized in Table 2 and Figure 6. Variability of the results in the case of direct fabrication by micromilling is evident (SD = 11.42), confirming common knowledge. Conversely, depths of PDMS microchannels obtained with the proposed method results in a much higher reproducibility (SD = 2.5) since the starting masters were deeper than the final target depths and were therefore affected by a more negligible error of the micromilling. For completeness, we also tested the error only due to the shrinkage starting from a PMMA master with fixed depth andas expectedthe reproducibility was even better (SD = 1.6) since in this case the only source of error was in the starting concentration of gelatin in water. In addition, microchannels obtained by direct micromilling technology exhibited a high roughness profile, while in many applications a

no.

depth micromilling (μm)

depth dehydration A (μm)

depth dehydration B (μm)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

20 42 13 21 18 21 22 26 24 12 25 25 18 30 14 49 21 20 55 18

22 19 18 21 23 24 21 22 19 18 20 21 19 22 21 20 21 19 22 21

17 17 18 13 19 18 17 19 20 17 18 20 18 19 14 21 14 24 17 19

Figure 6. Fabrication error graph for microchannels produced by direct micromachining “Micromilling”, by the proposed method starting from the same master “Dehydration A” and from 20 different micromachined PMMA masters “Dehydration B”, (A) and table of experimental results on micromilling fabrication error (Table 2).

very smooth surface is required. Interestingly, from the SEM image it is possible to observe the absence of the engraving lines in the case of the mold prepared with the gelatin CGD approach (Figure 5B,D,F). 3.4. Microchannel Validation and Use in an Example of Cell Culture Localization. To validate the obtained microchannels, we bonded previously described microchannels (aspect ratio 1:20 and 1:3) using the plasma treatment according to the protocol reported in the Experimental Section. We fluxed a red aqueous solution containing sulforhodamine B sodium salt at a concentration of 0.1 mg/mL by syringe pump to test the sealing (Figure 7A). At the end of the experiment no leakage was observed (Figure 7B) and a final characterization on the bonded microchannels was performed using a confocal E

DOI: 10.1021/acsami.6b04128 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces

Figure 7. Photograph of the experimental setup (A) and microdevice detail (B); validation of the microchannel by flowing a red aqueous solution in the microfluidic channel (C, D) and into rhomboidal and hexagonal arrays by capillarity (E, F). SEM image of an entire array obtained by dehydration process (G) and localization of the bEnd.3 cells (H,I).

4. CONCLUSIONS To summarize, we demonstrated a novel, simple, and reproducible method that does not require advanced and expensive technologies to add value to a low-cost technology based on micromilling in terms of miniaturization capability. Indeed, the CGD method, here proposed, uses a confined dehydration process of patterned gelatin substrates fabricated via replica molding onto micromilled PMMA; the process yields gelatin masters with demonstrated final micrometric features down to 3 μm for the channel depth and, in specific configurations, down to 5 μm for the channel width. In addition, the results show that the proposed method is very advantageous in making the preparation of microchannels under 50 μm highly reproducible compared to direct fabrication by micromilling, as demonstrated when choosing 20 μm as target depth. We demonstrated that the extent of reduction can be finely controlled simply using defined initial depths and gelatin solutions of varying water content. Some tests on microfluidic devices of reduced sizeprepared with the proposed technologywere provided as a validation of the above. We also used an array of shaped grooves obtained by applying the proposed process on an array of pillars and demonstrated the capability to confine cells in the grooves thanks to a previous localization of the fibronectin by capillarity through the interconnecting microchannels.

microscope (Leica TCS SP5MP) with a 25× water-immersion objective (Figure 7C,D). Experimental results confirmed that the proposed method can be used for the fabrication of microchannels with a depth down to a few micrometers. As a further applicative example for the proposed method, rhomboid and hexagonal arrays with reduced depth and connected via 5 μm wide microchannels were fabricated and tested. Recently, some research groups developed similar geometries to reproduce microfluidic logic circuits or for the interrogation and control of cell signaling mechanisms. We were able to obtain the above-mentioned geometries by applying the proposed process onto an array of pillars properly located as shown in Figure 5D,F. We first tested bonded microstructures inserting by capillarity a red fluorescent liquid in a vacuum chamber (Figure 7E,F), taking advantage of the material surface modification.7,32 As reported in the Supporting Information(Figures S6 and S7), the microstructure arrays present red aqueous solution of sulforhodamine B within all microstructures (hexagonal and rhomboid), which confirms that our dehydration process is uniform. The confinement of fluids and molecules in a specific area of the microstructured arrays may provide several advantages for cell−material interaction studies, in particular, by selectively promoting cell adhesion and growth. To address this issue, we filled the microstructured array, reported in Figure 7G, with a fibronectin solution to obtain a selective localization of this adhesion protein only in the rhomboidal areas and observed brain endothelial cell adhesion and growth in the array. After 3 d of culture, bEnd.3 cells appeared selectively localized in the rhomboidal area (Figure 7H,I), thus indicating the potentiality of the microstructured array to act as a mean to guide cell behavior.

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acsami.6b04128. Picture of a device under a temporary magnetic bonding, comparison between confined and nonconfined gelatin dehydration, profiles, and roughness measurements, and F

DOI: 10.1021/acsami.6b04128 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces

■

(18) Guckenberger, D. J.; de Groot, T. E.; Wan, A. M. D.; Beebe, D. J.; Young, E. W. K. Micromilling: a Method for Ultra-Rapid Prototyping of Plastic Microfluidic Devices. Lab Chip 2015, 15, 2364−2378. (19) de Oliveira, F. B.; Rodrigues, A. R.; Coelho, R. T.; de Souza, A. F. Size Effect and Minimum Chip Thickness in Micromilling. Int. J. Mach Tool Manu 2015, 89, 39−54. (20) Chen, P.-C.; Pan, C.-W.; Lee, W.-C.; Li, K.-M. An Experimental Study of Micromilling Parameters to Manufacture Microchannels on a PMMA Substrate. Int. J. Adv. Manuf Tech 2014, 71, 1623−1630. (21) Xia, Y. N.; Whitesides, G. M. Soft Lithography. Angew. Chem., Int. Ed. 1998, 37, 550−575. (22) Brittain, S.; Paul, K.; Zhao, X. M.; Whitesides, G. Soft Lithography and Microfabrication. Phys. World 1998, 11, 31−36. (23) Yang, W.; Nam, Y. G.; Lee, B.-K.; Han, K.; Kwon, T. H.; Kim, D. S. Fabrication of a Hydrophilic Poly(dimethylsiloxane) Microporous Structure and Its Application to Portable Microfluidic Pump. Jpn. J. Appl. Phys. 2010, 49, 06GM01. (24) Zhao, X. M.; Xia, Y. N.; Whitesides, G. M. Soft Lithographic Methods for Nano-Fabrication. J. Mater. Chem. 1997, 7, 1069−1074. (25) Occhipinti, L. G. Surface Treatment of an Organic or Inorganic Substrate for Enhancing Stability of a Lithographically Defined Deposited Metal Layer, U.S. Patent Application (2010), 12/835. (26) Bhattacharya, S.; Datta, A.; Berg, J. M.; Gangopadhyay, S. Studies on Surface Wettability of Poly(Dimethyl) Siloxane (PDMS) and Glass Under Oxygen-Plasma Treatment and Correlation with Bond Strength. J. Microelectromech. Syst. 2005, 14, 590−597. (27) Spath, D., Tritschler, H., Bischoff, L., Schulz, W. Micromilling High Potential Technology for Micromechanical Parts. In AMST’02 Advanced Manufacturing Systems and Technology, Proceedings of the Sixth International Conference; Springer-Verlag: New York, 2002; Vol. 437, pp 859−864.10.1007/978-3-7091-2555-7_100 (28) Okano, K. Micromachining of Micromachine Parts. Int. J. Jpn. Soc. Precis. Eng. 1994, 28, 196−199. (29) Pimpin, A.; Srituravanich, W. Review on Micro- and Nanolithography Techniques and Their Applications. Eng. J. 2011, 16, 37−56. (30) Lee, K.; Dornfeld, D. A. An Experimental Study on Burr Formation in Micro Milling Aluminum and Copper. Namri/Sme 2002, 10, 255−262. (31) Chae, J.; Park, S. S.; Freiheit, T. Investigation of Micro-Cutting Operations. Int. J. Mach Tool Manu 2006, 46, 313−332. (32) Juncker, D.; Schmid, H.; Bernard, A.; Caelen, I.; Michel, B.; de Rooij, N.; Delamarche, E. Soft and Rigid Two-Level Microfluidic Networks for Patterning Surfaces. J. Micromech. Microeng. 2001, 11, 532−541.

SEM and optical images of some of the prepared microstructures (PDF)

AUTHOR INFORMATION

Corresponding Author

*E-mail: raff[email protected] (R.V.). 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 The authors would like to thank Roberta Infranca for the proofreading and Valentina La Tilla for the abstract graphic contribution.

■

REFERENCES

(1) Becker, H.; Gartner, C. Polymer Microfabrication Methods for Microfluidic Analytical Applications. Electrophoresis 2000, 21, 12−26. (2) Alrifaiy, A.; Lindahl, O. A.; Ramser, K. Polymer-Based Microfluidic Devices for Pharmacy, Biology and Tissue Engineering. Polymers 2012, 4, 1349−1398. (3) Salic, A.; Tusek, A.; Zelic, B. Application of Microreactors in Medicine and Biomedicine. J. Appl. Biomed. 2012, 10, 137−153. (4) Whitesides, G. M. The Origins and the Future of Microfluidics. Nature 2006, 442, 368−373. (5) Ziolkowska, K.; Kwapiszewski, R.; Brzozka, Z. Microfluidic Devices as Tools for Mimicking the in Vivo Environment. New J. Chem. 2011, 35, 979−990. (6) Haeberle, S.; Zengerle, R. Microfluidic Platforms for Lab-on-aChip Applications. Lab Chip 2007, 7, 1094−1110. (7) Delamarche, E.; Juncker, D.; Schmid, H. Microfluidics for Processing Surfaces and Miniaturizing Biological Assays. Adv. Mater. 2005, 17, 2911−2933. (8) Hwang, D. K.; Dendukuri, D.; Doyle, P. S. Microfluidic-Based Synthesis of Non-Spherical Magnetic Hydrogel Microparticles. Lab Chip 2008, 8, 1640−1647. (9) Bousse, L.; Mouradian, S.; Minalla, A.; Yee, H.; Williams, K.; Dubrow, R. Protein Sizing on a Microchip. Anal. Chem. 2001, 73, 1207−1212. (10) Bogorad, M. I.; DeStefano, J.; Karlsson, J.; Wong, A. D.; Gerecht, S.; Searson, P. C. Review: In Vitro Microvessel Models. Lab Chip 2015, 15, 4242−4255. (11) Sundberg, S. A. High-Throughput and Ultra-High-Throughput Screening: Solution- and Cell-Based Approaches. Curr. Opin. Biotechnol. 2000, 11, 47−53. (12) Bennett, M. R.; Hasty, J. Microfluidic Devices for Measuring Gene Network Dynamics in Single Cells. Nat. Rev. Genet. 2009, 10, 628−638. (13) Zare, R. N.; Kim, S. Microfluidic Platforms for Single-Cell Analysis. Annu. Rev. Biomed. Eng. 2010, 12, 187−201. (14) Cao, H.; Tegenfeldt, J. O.; Austin, R. H.; Chou, S. Y. Gradient Nanostructures for Interfacing Microfluidics and Nanofluidics. Appl. Phys. Lett. 2002, 81, 3058−3060. (15) Focaroli, S.; Mazzitelli, S.; Falconi, M.; Luca, G.; Nastruzzi, C. Preparation and Validation of Low Cost Microfluidic Chips Using a Shrinking Approach. Lab Chip 2014, 14, 4007−4016. (16) Das, A. L.; Mukherjee, R.; Katiyer, V.; Kulkarni, M.; Ghatak, A.; Sharma, A. Generation of Sub-Micrometer-Scale Patterns by Successive Miniaturization Using Hydrogels. Adv. Mater. 2007, 19, 1943−1946. (17) Chen, C. S.; Breslauer, D. N.; Luna, J. I.; Grimes, A.; Chin, W. C.; Lee, L. P.; Khine, M. Shrinky-Dink Microfluidics: 3D Polystyrene Chips. Lab Chip 2008, 8, 622−624. G

DOI: 10.1021/acsami.6b04128 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX