Wafer-Scale Multilayer Fabrication for Silk Fibroin-Based

Subscriber access provided by TULANE UNIVERSITY

Biological and Medical Applications of Materials and Interfaces

Wafer-Scale Multilayer Fabrication for Silk Fibroin-Based Microelectronics Geon Kook, Sohyeon Jeong, So Hyun Kim, Mi Kyung Kim, Sungwoo Lee, Il-Joo Cho, Nakwon Choi, and Hyunjoo Jenny Lee ACS Appl. Mater. Interfaces, Just Accepted Manuscript • DOI: 10.1021/acsami.8b13170 • Publication Date (Web): 27 Nov 2018 Downloaded from http://pubs.acs.org on December 4, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 42 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

Wafer-Scale Multilayer Fabrication for Silk FibroinBased Microelectronics Geon Kook,†,◆ Sohyeon Jeong,‡,§,◆ So Hyun Kim,‡,∥ Mikyung Kim,† Sungwoo Lee,† Il-Joo Cho,‡,§ Nakwon Choi,*,‡,§ and Hyunjoo J. Lee*,† † School of Electrical Engineering, Korea Advanced Institute of Science and Technology (KAIST), 291 Daehak-ro, Yuseong-gu, Daejeon 34141, Republic of Korea ‡ Center for BioMicrosystems, Brain Science Institute, Korea Institute of Science and Technology (KIST), 5 Hwarang-ro 14 gil, Seongbuk-gu, Seoul 02792, Republic of Korea § Division of Bio-Medical Science & Technology, KIST School, Korea University of Science and Technology (UST), 5 Hwarang-ro 14 gil, Seongbuk-gu, Seoul 02792, Republic of Korea ∥ SK Biopharmaceuticals Co., Ltd., 221 Pangyoyeok-ro, Bundang-gu, Seongnam-si, Gyeonggido, 13494, Republic of Korea KEYWORDS: silk fibroin, multilayer patterning, UV photolithography, wafer-scale fabrication, silk fibroin electronics

ABSTRACT: Silk fibroin is an excellent candidate for biomedical implantable devices because of its biocompatibility, controllable biodegradability, solution processability, flexibility, and transparency. Thus, fibroin has been widely explored in biomedical applications as biodegradable

ACS Paragon Plus Environment

1

ACS Applied Materials & Interfaces 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 42

films as well as functional microstructures. Although there exists a large number of patterning methods for fibroin thin films, multilayer micropatterning of fibroin films interleaved with metal layers still remains a challenge. Herein, we report a new wafer-scale multilayer microfabrication process named AMoS (Aluminum hard mask on silk fibroin) which is capable of micropatterning multiple layers composed of both fibroin and inorganic materials (e.g., metal, and dielectrics) with high-precision microscale alignment. To the best of our knowledge, our AMoS process is the first demonstration of wafer-scale multilayer processing of both silk fibroin and metal micropatterns. In the AMoS process, aluminum deposited on fibroin is first micropatterned using the conventional ultraviolet (UV) photolithography, and the patterned aluminum layer is then used as a mask to pattern fibroin underneath. We demonstrate the versatility of our fabrication process by fabricating fibroin microstructures with different dimensions, passive electronic components composed of both fibroin and metal layers, and functional fibroin microstructures for drug delivery. Furthermore, since one of the crucial advantages of fibroin is biocompatibility, we assess the biocompatibility of our fabrication process through the culture of highly-susceptible primary neurons. Because the AMoS process utilizes the conventional UV photolithography, the principal advantages of our process are multilayer fabrication with high-precision alignment, highresolution, wafer-scale large area processing, no requirement for chemical modification of the protein, and high throughput and thus low cost, all of which have not been feasible with silk fibroin. Therefore, the proposed fabrication method is a promising candidate for batch fabrication of functional fibroin microelectronics (e.g., memristors and organic thin film transistors) for nextgeneration implantable biomedical applications.

INTRODUCTION


2

Page 3 of 42 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Biodegradable electronics is an emerging technology which is expected to foster a wide range of new biomedical implantable devices.1 As transient electronics disintegrate and resorb into our body after a certain period of operation, every component of a device including substrates, insulators, conductors, and semiconductors must be biodegradable. A few conventional electronic systems have been redesigned to be fully biodegradable, such as CMOS logic circuit2-4, physiological sensor5-6, wireless drug delivery system,7 and memristor,8 by using biodegradable substrates (e.g,. silk fibroin, poly(lactic-co-glycolic acid))1, biodegradable metals (e.g., magnesium, tungsten)9, and conventional electronic materials in minute amount (e.g., silicon, silicon dioxide)10-11. Among biodegradable polymers, silk fibroin is one of the most extensively exploited materials for biodegradable electronics. Fibroin is a structural protein in Bombyx mori silkworm cocoons, which can be extracted into solution and cast as a polymer substrate.12-13 Fibroin is not only biodegradable, and U. S. Food and Drug Administration (FDA)-approved, but also exhibits advantageous properties such as controllable rate of biodegradation, tunable drugloading capability, optical transparency, and unique mechanical property (e.g., tunable elastic modulus from 2 GPa14 to 17 GPa15)16-21. Therefore, a large number of fibroin-based biomedical devices, such as implantable electronics,22-24 tissue scaffolds,25 optical biosensors26-29, and drugdelivery microneedles30-31 have been introduced that explore these properties. Due to these versatile applications of silk fibroin, simple fabrication methods such as cast molding,32-34 nanoimprinting,35-36 and inkjet printing37-39 have been applied to pattern fibroin films (Table S1). Advanced methods based on photolithography have also been developed to achieve microscale fibroin patterns: UV photolithography of chemically modified fibroin,40 multiphoton lithography,41 e-beam lithography,42 deep UV (DUV) photolithography,43 and hard-mask mediated patterning.44 However, to fabricate fibroin-based electronic devices, fibroin patterning alone is not


3


Page 4 of 42

sufficient; it is essential to pattern metal layers on top or below the fibroin layer (i.e., in a multilayer fashion) with high precision. Previously, the metal layers have been patterned on the fibroin films through shadow mask45 and transfer technique.46-47 However, these methods do not support multilayer microfabrication due to their inherently substantial misalignment. Thus, developing biomedical electronics composed of fibroin used as either a dielectric or intermediate layer still remains a challenge. Specifically, currently available methods do not allow fabrication of advanced fibroin-based components suitable for transient electronics, such as a microscale capacitor composed of fibroin as a dielectric, a fibroin-passivated electronic device with vertical interconnect accesses (vias) and interconnects, and a drug delivery system with fibroin microstructures on top of a microheater. One possible solution to achieve multilayer processing of fibroin film interleaved with metal layers is to apply conventional UV photolithography. In the conventional planar device fabrication process, patterns in two layers are aligned precisely in a micrometer scale by using a photosensitive mask (i.e., photoresist) and an optical mask aligner. The patterns on the second layer are formed by either removal or addition through the micropatterned mask layer. However, despite the advantage of multilayer fabrication with high-precision alignment, the conventional UV photolithography using photoresist has not been applied to pattern fibroin mainly because of two potential concerns: chemical damage of the fibroin film and endowed toxicity on the fibroin film from commonly used solvents in the UV photolithography. Fibroin films do not withstand some of the chemicals used in the conventional UV photolithography processes.44 Besides, infiltration of toxic solvents such as etchants, photoresists, and developers could jeopardize the biocompatibility of the fibroin.


4

Page 5 of 42 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Here, we propose a new process named AMoS (aluminum hard mask on silk fibroin) which overcomes the aforementioned concerns of the UV photolithography so that both fibroin and thin metal films can be micropatterned using the conventional planar process. Specifically, to prevent chemical and structural damage to the fibroin, we introduce an additional mask layer (i.e., hard mask) between the fibroin film and the photoresist to protect the fibroin film from harsh chemicals (Figure 1). While protecting the fibroin film, this additional hard mask is micropatterned with highprecision alignment to the underlying layer using the conventional UV photolithography which allows for multilayer processing (Figure 1a). We first investigate various materials as a candidate for the masking layer and optimize the process conditions to achieve high-yield processing on wafer scale. Also, we characterize our process to confirm the intactness of the chemical structure of fibroin and the biocompatibility of the produced patterns through the culture of highlysusceptible primary neurons. The micropatterned hard mask on a fibroin layer can also be used as a lift-off mask to pattern a metal layer on top of the fibroin film as well as as an etch mask for the underlying fibroin film (Figure 1b). Thus, the AMoS process enables micropatterning of both fibroin and thin metal films in wafer scale with high-throughput for the first time. Examples of next-generation biodegradable systems enabled by the AMoS process include biodegradable memristors, patterned shuttle for neural probes, and position-selective drug delivery using microheaters.


5


Page 6 of 42

Figure 1. a) Schematics of fabrication process of an aluminum hard mask on a silk fibroin film (AMoS). The fibroin film is decoupled from the harsh chemicals used in UV photolithography. b) Schematics of subsequent processes utilizing the AMoS mask and scanning electron microscopy (SEM) images of final patterns (scale bars: 20 µm): patterning of the fibroin film using oxygen plasma (upper) and patterning of another metal on a fibroin film using the lift-off process (lower). The insets show the cross-sectional schematics of the structures along the dotted line.

EXPERIMENTAL SECTION Extraction of Silk Fibroin. An aqueous silk fibroin solution was extracted from Bombyx mori silkworm cocoons based on the previously established protocol in the literature.13 After cutting


6

Page 7 of 42 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


cocoons into pieces, 5 g of cut cocoons were weighed and boiled in 2 L of 0.02 M Na2CO3 (aq) (Alfa Aesar) for 30 min to remove sericin. After drying overnight, 3 g of degummed fibroin was dissolved in 12 mL of 9.3 M LiBr (aq) (Alfa Aesar) for 4 h at 60C. Dissolved fibroin solution was injected in a 3,500-molecular-weight-cut-off (MWCO) dialysis cassette (Thermo Fisher Scientific) and then dialyzed against the deionized (DI) water for 2 days with occasional change of the DI water. Finally, the dialyzed fibroin solution was filtered through a 5-μm-pore-size syringe filter (Pall Corporation) to obtain the final clear aqueous solution of ~8% (w/v) fibroin. Aluminum hard mask process on silk fibroin films. Aluminum hard mask is patterned on a fibroin film as follows (Figure 1a). ~8% (w/v) aqueous fibroin solution was spin-coated on a silicon wafer and dried overnight to form a 1.2-μm-thick fibroin film. The film was treated with 70% (v/v) methanol for 2 min so that the film became insoluble in water. A 15-nm-thick Al layer was deposited on the fibroin film by thermal evaporation. This Al layer was patterned through the conventional photolithography using a positive PR (AZ HKT-501, Microchemicals) and a semiautomated mask aligner (MA/BA6 mask and bond aligner, SUSS MicroTec). The Al layer was then etched in buffered hydrofluoric acid (BHF) with a volume ratio of 6:1 of 1.3% (w/w) NH4F (aq) and 1.6% (w/w) HF (aq). Through this AMoS mask, the fibroin film was etched with oxygen plasma at a power of 200 W and a flow rate of 50 sccm (Figure 1b, upper) using an inductively coupled plasma (ICP) asher. After the fibroin patterning, the remaining AMoS mask was removed with the BHF. In addition to fibroin patterning, a metal layer can be patterned on the fibroin film using a lift-off process (Figure 1b, lower). For the lift-off process, inverse of the intended metal shapes was patterned on an Al layer as an AMoS mask. Without removing the PR on the AMoS mask, a Cr/Au (5/50 nm) layer was thermally evaporated on top. Then, the wafer was sonicated


7


Page 8 of 42

for a few minutes in acetone to remove the PR. During this process, the Cr/Au layer on top of the PR was removed along with the PR. Finally, the remaining Al mask was removed in the BHF. Curve fitting of Fourier self-deconvolution (FSD) spectra of the Fourier-transform Infrared (FTIR) spectra. Original FTIR spectra of an amorphous fibroin film and a fibroin film before and after the AMoS process were acquired using an FTIR Spectrometer (Nicolet iS50, Thermo Fisher Scientific Instrument). The FSD of the spectra was performed using a graphing and analysis tool (Origin 2017, OriginLab) with a gamma of 12 and a smoothing factor of 0.1. With the obtained FSD spectra, we conducted curve-fitting using the same software. After baseline correction, the number and positions of peaks in the spectra were decided from the second derivatives of the spectra. With Gaussian functions as fitting functions, the band positions were optimized. Initially, the band positions were fixed while full width at half maximum (FWHM) and heights were varied. Next, the band positions were varied while FWHM and heights were fixed. Isolation and culture of primary cortical neurons. Pregnant Sprague-Dawley rats (E17; Samtako Inc.) were purchased from Samtako Inc. for primary culture of cortical neurons, following a reported protocol.48 All procedures were performed according to the animal welfare guidelines approved by the Institutional Animal Care and Use Committee of Korea Institute of Science and Technology (KIST). Briefly, the entire cortices were dissected out as an intact form from the brain of decapitated embryos and then treated with the enzyme mixture from neural tissue dissociation kits (Miltenyi Biotec) and simultaneously triturated by running the gentlMACS dissociator (Miltenyi Biotec). Trypan blue (GIBCO) was applied to manually count live cortical neurons on a hematocytometer (INCYTO). Then, the isolated neurons were plated at a density of 400 cell mm-2 on various substrates coated with 100 μg mL-1 of poly-D-lysine (Sigma-Aldrich). The primary cortical neurons were cultured in medium consisting of neurobasal media (NBM;


8

Page 9 of 42 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


GIBCO) supplemented with 2% (v/v) B27-supplement (Invitrogen), 2 mM Glutamax-I (GIBCO) and 1% (v/v) penicillin-streptomycin (GIBCO), and maintained at 37°C in a 5% CO2 humidified incubator. Live/dead cell viability assay. The viability of the cortical neurons cultured for 3-7 days was assessed by incubating samples (i.e., cells on non-transparent Si-based substrates) in phosphatebuffered saline (PBS) containing 1 μM CellTraceTM Calcein Green, AM (calcein-AM; Thermo Fisher Scientific) and 15 μM propidium iodide (PI; Sigma-Aldrich) for 30 min at 37°C and 5% CO2. After washing with PBS, confocal laser scanning microscopy was performed to acquire fluorescence micrographs that displayed live (green) and dead (red) cells by flipping the nontransparent sample over. Cell proliferation assay. As an additional assessment to quantify the cell viability, we performed the cell proliferation assay. Primary cortical neurons cultured on various substrates were treated with water-soluble tetrazolium salt (WST) to allow for the production of formazan by the mitochondrial dehydrogenase of live cells using EZ-CYTOX (Daeillab). Briefly, 1000 cells/mm2 of cortical neurons were plated on four substrates: polystyrene plate (control), bare fibroin, fibroinpatterned Si, and Au-patterned fibroin. On day 7 of culture, 100 μL of the WST solution was applied to culture media (1 mL) and then incubated for 4 hr at 37℃ in a humidified 5% CO2 incubator. ‘Blank’ samples were prepared by mixing the WST solution with fresh culture media. After the incubation, mixtures of the WST solution and the culture media were transferred to a 96well plate. Finally, absorbance at 450 nm (A450) of formazan from each well was measured in a microplate reader (Tecan). To quantify the cell viability on day 7, we normalized backgroundsubtracted absorbance of each case (A450 – A450,blank) by background-subtracted absorbance of the control (A450,control – A450,blank).


9


Page 10 of 42

Statistical analysis. Statistical analysis was performed with Prism (GraphPad Software). Oneway ANOVA with the Tukey test for multiple comparisons was used to compare relative absorbance at 450 nm from the cell proliferation assays. Quantitative study of small molecule diffusion from a fibroin film. Rhodamine B was loaded into a 1.2-µm-thick fibroin film by soaking the film in 1 mM aqueous solution of Rhodamine B for 1 h. The rate of diffusion was measured by immersing fibroin films loaded with Rhodamine B in the DI water. Specifically, the fibroin film on a silicon wafer was diced into 15 pieces (12 × 12 mm2) and each piece was immersed into a vial containing 2 mL of the DI water. The solution from each vial was collected at a different time to observe the change in the concentration of diffused Rhodamine B. The concentration of the Rhodamine B in the solutions was investigated using an UV-vis spectrophotometer (Lambda 1050, Perkin Elmer). In addition, to reduce the release rate, an extra fibroin layer of the same thickness was coated on the drug-loaded film. The release rate of the dye from this structure was investigated with the same protocol to study controllability of the release rate by immobilization. Drug delivery experiment using Rhodamine B through micropatterned fibroin on polyimide-based neural interfaces. We fabricated a polyimide-based electrocorticography (ECoG) electrodes and neural probes, and micropatterned fibroin through high precision alignment to the microelectrode array in wafer scale using our AMoS process (Detailed fabrication flow is in Figure S10). Then, we loaded Rhodamine B into the micropatterned fibroin using the method above. We placed the drug-loaded fibroin film on a mouse brain, which was extracted after the paraformaldehye perfusion following the standard protocol.49 After 30 min, we observed the successful targeted delivery of the dye with high spatial resolution (50 µm) at the specific area where the microelectrode arrays were positioned. All experiments were performed according to


10

Page 11 of 42 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


protocols approved by the Institutional Animal Care and Use Committee at the Korea Advanced Institute of Science and Technology (KA2017-28).

RESULTS AND DISCUSSION The selection of the material for the hard mask that decouples the fibroin from the harsh chemical is the most critical design consideration for the AMoS process. While this mask layer must be compatible with UV photolithography (i.e., ability to form a uniform thin layer), it must also protect the fibroin from harsh chemicals. Thus, we focused our choice on metals and dielectric materials as potential candidates for the mask layer. Also, the etchant that is used to pattern the mask layer must be compatible with the fibroin film because a portion of the fibroin film is exposed to the etchant during the AMoS process (Figure 1a). Therefore, we first searched for a suitable etchant chemical that would be compatible with fibroin. Although most etchants used in device fabrication contain either highly acidic or basic chemicals that are known to damage proteins, there are few candidate chemicals that do not damage fibroin films but etch some of the inorganic materials. For example, hydrofluoric acid (HF), a common etchant used in the device fabrication, does not alter the molecular structures of several proteins including fibroin.50 First, we exposed fibroin films in concentrated HF and confirmed that the films were structurally intact. Next, an appropriate material for the hard mask was determined by investigating etching properties of various materials in pure HF. Among several candidates, aluminum (Al) exhibited the most favorable characteristics as a hard mask such as controllable etch rate, excellent etch uniformity and outstanding adhesion with fibroin film (Table S2). For example, silicon dioxide (SiO2) exhibited poor adhesion with the underlying fibroin (i.e., delaminated during etching in HF), while titanium (Ti), nickel (Ni), silver (Ag), and other metals exhibited either nonuniform or impractical


11


Page 12 of 42

(i.e., too slow or too fast) etch rate in HF. Thus, Al and HF were chosen as the primary mask/etchant pair for our AMoS process as explained in Experimental Section (Figure 1b). We evaluated and confirmed the feasibility of our AMoS process by fabricating various fibroin microstructures and metal patterns on a fibroin film in wafer scale. 1.2-µm-thick fibroin films spincoated on a silicon (Si) wafer were patterned in several different shapes and scales using the AMoS process (Figure 2a-d). We demonstrated a linewidth resolution of fibroin down to 2 µm, which was limited by the resolution of an exposure system in our fabrication facility. Because we used O2 plasma to etch fibroin chemically, the sidewall slope of the patterned fibroin showed isotropic profile (Figure 2c,d). The close-in photo of the sidewall shows a line edge roughness of a few nanometers. Potential causes for this roughness include non-uniform etching during the patterning of the hard mask and spatial inhomogeneity of fibroin polymer chains (Figure 2d). In addition to fibroin patterning, we also successfully patterned a 50-nm-thick gold film on top of a 1.2-µm-thick fibroin layer (Figure 2e,f) with a high yield (~99%) across a 100-mm silicon wafer (Figure S1). We used the additive lift-off process; the metal on top of the Al/PR patterns was lifted-off as PR was removed, which left the inverse metal patterns on the fibroin film. Here, adhesion of the metal directly in contact with the fibroin layer must be strong to endure the lift-off process. Because of the moderate adhesion strength between fibroin and gold46, gold was successfully lifted-off without an adhesion layer. However, to accelerate the lift-off process by using sonication, we used a 5-nm-thick chromium layer to enhance the adhesion further.


12

Page 13 of 42 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 2. Fibroin microstructures and metal micropatterns on a fibroin film produced by using the AMoS mask. SEM images of a) 10-µm-wide (scale bars: 30 µm) and b) 2-µm-wide line patterns (scale bars: 10 µm) of 1.2-µm-thick fibroin coated on a silicon wafer. c) Cross-sectional SEM image of 10-µm-wide fibroin line patterns on silicon (scale bar: 20 µm). d) SEM image of the edge of the patterned fibroin (scale bar: 1 µm). e-f) Optical images of gold lines defined on fibroin films (scale bars: 200 µm, 20 µm).

We examined the effects of the AMoS process on fibroin through physical and chemical characterizations of the film before and after the process. First, surface profiles of 1.2-µm-thick fibroin patterns were analyzed with atomic force microscope (AFM) (Figure 3a,b). The pattern showed lateral over-etching around 1 μm from each side due to isotropic O2 plasma etching. We also noticed that the edge of the pattern (< 5 µm from the edge) was raised by approximately 200 nm. This bulging at the edge of the pattern is mostly attributed to the compressional stress incurred due to the decrease in water content during fibroin crystallization in methanol.51 Surface root mean


13


Page 14 of 42

square (RMS) roughness of the fibroin film increased from 3.4 to 4.3 nm after the process (Figure 3b,c), which was still sufficiently small for subsequent planar fabrication processes. The increase in the surface roughness is mainly attributed to the thermal deposition and removal of Al on the fibroin surface (Figure S2). Using AMoS mask, the maximum resolution of the hard mask is limited by the photolithography equipment, which is much smaller than that offered by the commonly used shadow mask. For our process, there are three possible factors that could deteriorate the resolution: fibroin etching method, crystal domain inhomogeneity in fibroin films, and stress in thick fibroin films. For further evaluation of our AMoS process, fibroin films of different thickness were tested. Since we used oxygen plasma for etching which is isotropic, the patterns were etched out at the same rate in all direction. (This implies that there is a lateral etching underneath the hard mask.) Since a thicker film requires a longer etching time, etching in the lateral direction should increase for a standard isotropic etching process. As predicted, for thicker fibroin films, smaller linewidths and shapes (circles, triangles, and pentagons) were observed compared to the initial mask patterns (Figure 3do, Figure S5). Moreover, we observed that the corners of the patterns (yellow dotted lines in Figure 3f,g,j,k,n,o) were sharper for thicker fibroin films (i.e., as the etching time in O2 plasma increases), which is also an expected characteristic of isotropic etching.52 Here, we intentionally designed the corners to be rounded during the mask design to observe the effects of over-etching. When such properties are not desirable, anisotropic etching methods such as reactive ion etching53 could be used. Also, we observed some irregular edges of the patterns (Figure 3i). These are possibly attributed to fibroin’s inhomogeneity that causes dissimilar etch rate.53 Furthermore, we noticed that the surface of the patterned fibroin films was less flat for thick fibroin films (Figure 3l-o). This


14

Page 15 of 42 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


non-flat surface is probably due to the volume decrease during the crystallization of the film, which results in more considerable film stress for a thicker fibroin film.

Figure 3. a) Atomic force microscopy (AFM) surface profile of a 30 µm × 30 µm fibroin structure patterned using the AMoS process on a silicon wafer. b) Close-in profile of the 5 µm × 5 µm dotted square in a) (top) and linear surface profile along the dotted line in a) (bottom). c) Comparison of root-mean-square (RMS) surface roughness of fibroin over the area of 5 µm × 5 µm of three cases: water-insoluble fibroin film as a control, the same film on which aluminum layer was deposited


15


Page 16 of 42

and stripped by BHF without patterning, and the patterned film in a). d-o) SEM images of various shapes patterned on fibroin films with different thickness: 700 nm (d-g), 3 µm (h-k), and 15 µm (l-o) (scale bars: 50 µm). The yellow dotted lines indicate sharpening of the corners of the patterns.

Next, we evaluated if there was any change in molecular structures in fibroin from the AMoS process. First, we measured FTIR spectra of a fibroin film before and after the process (Figure S3). Individual vibrational peaks and their intensities in the Amide I region (Figure 4) were analyzed by processing the original FTIR spectra with FSD; the FSD spectra were fitted as in a previously demonstrated analysis.54 The spectra revealed no difference in both positions and intensities of vibrational modes for random coils (R), beta-sheets (B), alpha-helices (A), turns (T), and side chains (S) after the process. In contrary, a water-soluble amorphous fibroin film showed significantly larger intensities of random coils to beta sheets.54 Likewise, no difference of vibrational peaks appeared in the Amide II region before and after the process (Figure S4). These results show that the AMoS process causes no significant perturbation in molecular structures in the fibroin films.


16

Page 17 of 42 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 4. Peak analysis and assignment of Fourier self-deconvolution (FSD) of the Fouriertransform IR (FTIR) spectra in Amide I region of the fibroin film before and after the process. The measured and analyzed FTIR spectra of an amorphous fibroin film is also shown for comparison. Abbreviations for letters of vibrational assignments are random coils (R), beta-sheets (B), alphahelices (A), turns (T), and side chains (S).

Lastly, we evaluated the effects of the AMoS process on the biocompatibility of fibroin. Due to the frequent use of solvents, despite the sufficient washing during and after the process, a minute amount of residue could exist on the surface of the device. We addressed this potential concern by


17


Page 18 of 42

conducting cell viability assays both qualitatively and quantitatively. Specifically, we assessed the viability of primary cortical neurons grown on three substrates: 1) bare fibroin coated on silicon, 2) patterned fibroin on silicon, and 3) patterned gold on fibroin (Figure 5). As primary neurons are notoriously sensitive to minute external perturbations, we first examined cytotoxicity of bare fibroin. We observed that the cell viability on the fibroin remained unaffectedly high in all cases for up to 7 days (Figure 5a), compared with that on the polystyrene plate (Figure S6). Although various organic solvents such as HF and acetone intervened during the AMoS process of patterning both fibroin-on-silicon and gold-on-fibroin respectively, the viability of the primary cortical neurons was still high on the patterns (Figure 5b [middle and bottom panels] and Figure 5c [middle panel]. We confirmed that appropriate rinsing (see Experimental Section) guaranteed the viability of the neurons. Patterns produced by using the AMoS process can be versatilely biocompatible with other cell types. For example, when NIH3T3 cells, which are widely used fibroblasts for testing cell viability, were cultured on the fibroin patterns, viability remained as similarly high as the fibroblasts on a polystyrene plate and a bare fibroin film (Figure S7). As additional confirmation of the biocompatibility of the AMoS process, we also performed the cell proliferation assay to compare the mitochondrial activity of live and healthy cells quantitatively. The primary cortical neurons cultured on the three substrates (Figure 5a-5c) showed statistically indistinguishable differences in the mitochondrial activity, represented by absorbance of formazan at 450 nm, on day 7 compared with the cells cultured on the polystyrene plate (Figure 5d). Interestingly, neurite outgrowth on the fibroin appeared to be prominent (Figure 5b and Figure S8). This feature serves as a physical cue55 to pattern cells including neurons in a particular direction or shape by an effect of the surface topography of the fibroin. Moreover, using the AMoS process, the cellular patterning is now possible on wafer scale (e.g., 100 mm in diameter). As a


18

Page 19 of 42 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


representative instance, neurite outgrowth markedly followed the topography of the fibroin (i.e., 'on pattern'): lines (Figure 5b [middle panel] and Figure 5c [middle panel]) and honeycombs (Figure 5b [bottom panel]. In comparison, the neurons on a bare fibroin film grew randomly over all areas (Figure 5a [middle and bottom panels]). We note that somas (i.e., neuronal cell bodies including nuclei) of the primary cortical neurons did not follow the topography of the fibroin primarily because neurons are not motile cells. If motile cells such as connective tissue cells (e.g., fibroblasts) and muscle cells (e.g., myocytes) were cultured, the cell bodies of these cells could have been patterned. We also note that the primary cortical neurons did not grow or were dead ‘off the fibroin patterns’ after a few days of culture in many cases (Figure S8). Conceivably, this result could be attributed to a non-uniform coating of poly-D-lysine on the bare silicon surface (i.e., ‘off the fibroin patterns’) (Figure S9).


19


Page 20 of 42

Figure 5. Biocompatibility assessment of the AMoS process. a) Top: schematics of fibroin-coated silicon. Middle and bottom: representative confocal microscopic images of live (green) and dead (red) primary cortical neurons cultured on the fibroin-coated silicon (scale bars: 100 µm in middle, 500 µm in bottom). b) Top: schematics of fibroin-patterned silicon. Middle and bottom: representative confocal microscopic images of live (green) and dead (red) primary cortical neurons cultured on the fibroin-patterned silicon (scale bars: 100 µm in middle, 500 µm in bottom). One of line patterns is depicted as a thin white-shaded rectangle. Honeycomb patterns are depicted as


20

Page 21 of 42 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


hexagons in different sizes (dotted and white shade in bottom). c) Top: schematics of goldpatterned fibroin. Middle: representative confocal microscopic image of live (green) and dead (red) primary cortical neurons cultured on the gold-patterned fibroin (scale bar: 100 µm). Line patterns are depicted as thin white-shaded rectangles. d) Bar graph displaying mitochondrial activity of primary cortical neurons cultured on various substrates involved in the AMoS process. The mitochondrial activity represents normalized absorbance of formazan at 450 nm: backgroundsubtracted absorbance (A450 – A450,blank) normalized by background-subtracted absorbance of the polystyrene plate (control; A450,control – A450,blank). Error bars indicate standard deviation of mean (triplicate measurements for each case). ns denotes indistinguishable statistical differences from the multiple comparisons of one-way ANOVA.

Since the AMoS process utilizes the conventional photolithography, it enables not only patterning of fibroin and metal on fibroin in a micrometer scale but also allows for multilayer patterning with high alignment accuracy (e.g., <  2 µm). Thus, here, we present two possible applications of the AMoS process to demonstrate these distinctive advantages: multi-functional neural interface systems and fibroin-based electronic systems. The multi-functional neural interface systems consist of micropatterned fibroins placed in close proximity to the microelectrode array for drug loading. These systems were fabricated in wafer-scale for high throughput, and it was possible to micropattern fibroin within 20 µm of process margin next to the microelectrode array because of the AMoS process. Such systems that enable simultaneous drug delivery and neural recording can be used to study the pharmaceutical effects of drugs as well as to study the structural connectivity of our brain.56-59 First, polyimide-based electrocorticography (ECoG) electrodes and neural probes were fabricated (Figure S10). Fibroin was micropatterned


21


Page 22 of 42

through high precision alignment to the microelectrode array in wafer scale using our AMoS process (Figure 6a,b, Figure S11). Since the neural signal recording must not be interfered by the drug delivery, it is essential that the fibroin patterns were placed such that they do not overlay with the microelectrode. Because of the high precision alignment capability up to 1 µm of our process, micropatterning fibroin patterns within 20 µm from multiple microelectrodes were readily achieved (Figure 6b, Figure S11b). We chose Rhodamine B as a model drug to visualize the selective drug loading and delivery. We observed the successful transfer of Rhodamine B from the fibroin patterns onto a mouse brain at the specific area where the microelectrode arrays were positioned (Figure 6c). Also, we observed diffusion of the Rhodamine B from the fibroin patterns on the neural probe to the surrounding 0.6% (w/w) agarose gel (Figure S11c). The diffusion rate of Rhodamine B from the fibroin film was investigated by immersing the drug-loaded fibroin film in the DI water and by observing the concentration at different time using a UV-vis spectrophotometer (Figure S12). Compared to other works that report on drug release from a fibroin film (Table S3), our results showed a similar release profile to that of small molecule drugs. Moreover, we reduced the release rate by applying an extra fibroin layer on top (Figure S12b). If a physical or chemical immobilization of the drug molecules is developed to control the rate of delivery, the AMoS process can be applied to previously reported epidermal and implantable electronic systems to achieve target-specific drug delivery.7, 23-24, 60 The second application is a fibroin-based electronic system which was enabled through the multilayer processing of metal interconnects through pad openings and vias. Using the multilayer processing, a parallel plate capacitor composed of fibroin as a dielectric can be fabricated where the bottom electrode is interfaced to the top layer through a vertical interconnect access (via). In addition, fabrication of passive fibroin-based electronic components such as resistors (R),


22

Page 23 of 42 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


capacitors (C), and inductors (L), where the fibroin layer is used as either a dielectric or a substrate, is possible (Figure 6d). Among various electronic components fabricated using the AMoS process, we measured the impedance of a resistor and a capacitor connected in series (RC) where fibroin was used as both a dielectric and a passivation layer (Figure 6e,f). The impedance spectra of the fibroin RC matched the typical characteristic curve where a pole at 1/RC was observed at 13.3 MHz as predicted (Figure S13). Compared to the electronic circuits demonstrated on a fibroin film in the previous work,2, 7, 22 our process achieved a wafer-scale fabrication of highly dense circuit components interconnected with vias, capacitors, and openings purely on silk fibroin films. In addition, as the fabricated devices were supported purely on a fibroin film, the overall system was flexible, transparent, and biodegradable (Figure 6f,g, Figure S14). Degradation of the fabricated device was observed in a protease solution where the transient decrease in the mass was quantitatively observed (Figure S14). Although the metals we integrated for demonstration (gold, silver, and chromium) are not known to be biodegradable, various biodegradable metals, such as tungsten, could be readily integrated for fully biodegradable electronics instead.9 Potential applications of our technology that enables fabrication of fibroin-based microscale devices include implantable microheaters and electrophysiological sensors which do not have to be removed after surgery.


23


Page 24 of 42

Figure 6. a) Fabricated polyimide-based flexible neural electrodes with fibroin patterns on a 100-mm silicon wafer (scale bar: 20 mm). The insets are microscope images of the fibroin patterns on the electrodes. Left: electrocorticography (ECoG) electrodes (scale bar: 400 µm).


24

Page 25 of 42 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Right: a neural probe (scale bar: 100 µm). b) Photograph of the fabricated ECoG electrodes (scale bar: 2 mm). Rhodamine B was loaded into fibroin to visualize the fibroin patterns. The inset is a close-up view of one of the microelectrode array (scale bar: 200 µm): alignment margin between the electrode opening and the fibroin pattern was below 50 µm, which was easily achievable with alignment limit of 1 µm of our process. c) Left: photographs showing the sequence of Rhodamine B delivery onto a mouse brain from the ECoG electrode through the micropatterned fibroin layers with high spatial resolution (scale bar: 5 mm). Right: close-up view of the sites showing localized drug delivery (scale bar: 1 mm). d) Wafer-scale fabrication process of µm-scale series RC components using the AMoS process by interleaving two fibroin layers and two metal layers. e) Optical microscope image of the fabricated device. f) Photograph of fabricated microscale RLC circuits on a fibroin film (scale bar: 3 mm). Gold (yellow) and silver (white) are patterned on different layers with a fibroin layer as a passivation layer in between. g) Photograph of metal patterns on a flexible fibroin film (scale bar: 5 mm).

CONCLUSIONS In summary, we have developed and demonstrated a new set of fabrication methods that enables multilayer processing of silk fibroin and metal films using an aluminum hard mask on silk fibroin, AMoS. By using the UV photolithography, the advantages of the planar UV photolithography directly apply to our AMoS process including wafer-scale mass production, high pattern resolution, and multilayer processing with high alignment accuracy. By confirming that this process did not alter the fibroin film chemically and physically and that the biocompatibility was not jeopardized, we demonstrated that our fabrication methods could be applied to implement useful fibroin-based devices in applications. Furthermore, since the AMoS process allows


25


Page 26 of 42

fabrication of metal interconnects and vias through fibroin layers for the first time, the proposed AMoS process would enable implementation of new fibroin-based electrical functional devices such as an active drug delivery system and biodegradable memristors.


26

Page 27 of 42 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


ASSOCIATED CONTENT Supporting Information. Supplementary figures and tables (PDF) AUTHOR INFORMATION Corresponding Author * Hyunjoo J. Lee | E-mail: [email protected]. Phone: +82423507536. * Nakwon Choi | E-mail: [email protected]. Phone: +8229586742. Author Contributions The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript. ◆These authors contributed equally. Funding Sources This research was supported by a grant of the Korea Health Technology R&D Project through the Korea Health Industry Development Institute (KHIDI), funded by the Ministry of Health & Welfare, Republic of Korea (grant number: HI15C3155) , by the Brain Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Science and ICT (2016M3C7A1904343) and by the Korea government (MSIT) (No. 2016R1C1B2009798). Notes The authors declare no competing financial interest. ACKNOWLEDGMENT


27


Page 28 of 42

This research was supported by a grant of the Korea Health Technology R&D Project through the Korea Health Industry Development Institute (KHIDI), funded by the Ministry of Health & Welfare, Republic of Korea (grant number: HI15C3155), by the Brain Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Science and ICT (2016M3C7A1904343) and by the Korea government (MSIT) (No. 2016R1C1B2009798). REFERENCES (1)

Feig, V. R.; Tran, H.; Bao, Z. N., Biodegradable Polymeric Materials in Degradable

Electronic Devices. Acs Central Sci 2018, 4 (3), 337-348. (2)

Hwang, S. W.; Tao, H.; Kim, D. H.; Cheng, H. Y.; Song, J. K.; Rill, E.; Brenckle, M. A.;

Panilaitis, B.; Won, S. M.; Kim, Y. S.; Song, Y. M.; Yu, K. J.; Ameen, A.; Li, R.; Su, Y. W.; Yang, M. M.; Kaplan, D. L.; Zakin, M. R.; Slepian, M. J.; Huang, Y. G.; Omenetto, F. G.; Rogers, J. A., A Physically Transient Form of Silicon Electronics. Sci 2012, 337 (6102), 1640-1644. (3)

Hwang, S. W.; Song, J. K.; Huang, X.; Cheng, H. Y.; Kang, S. K.; Kim, B. H.; Kim, J. H.;

Yu, S.; Huang, Y. G.; Rogers, J. A., High-Performance Biodegradable/Transient Electronics on Biodegradable Polymers. Adv. Mater. 2014, 26 (23), 3905-3911. (4)

Chang, J. K.; Fang, H.; Bower, C. A.; Song, E. M.; Yu, X. G.; Rogers, J. A., Materials and

Processing Approaches for Foundry-Compatible Transient Electronics. Proc. Natl. Acad. Sci. U. S. A. 2017, 114 (28), E5522-E5529. (5)

Yu, K. J.; Kuzum, D.; Hwang, S. W.; Kim, B. H.; Juul, H.; Kim, N. H.; Won, S. M.; Chiang,

K.; Trumpis, M.; Richardson, A. G.; Cheng, H. Y.; Fang, H.; Thompson, M.; Bink, H.; Talos, D.; Seo, K. J.; Lee, H. N.; Kang, S. K.; Kim, J. H.; Lee, J. Y.; Huang, Y. G.; Jensen, F. E.; Dichter, M. A.; Lucas, T. H.; Viventi, J.; Litt, B.; Rogers, J. A., Bioresorbable Silicon Electronics for


28

Page 29 of 42 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Transient Spatiotemporal Mapping of Electrical Activity from the Cerebral Cortex. Nat Mater 2016, 15 (7), 782-791. (6)

Hwang, S. W.; Lee, C. H.; Cheng, H. Y.; Jeong, J. W.; Kang, S. K.; Kim, J. H.; Shin, J.;

Yang, J.; Liu, Z. J.; Ameer, G. A.; Huang, Y. G.; Rogers, J. A., Biodegradable Elastomers and Silicon Nanomembranes/Nanoribbons for Stretchable, Transient Electronics, and Biosensors. Nano Lett. 2015, 15 (5), 2801-2808. (7)

Tao, H.; Hwang, S. W.; Marelli, B.; An, B.; Moreau, J. E.; Yang, M. M.; Brenckle, M. A.;

Kim, S.; Kaplan, D. L.; Rogers, J. A.; Omenetto, F. G., Silk-Based Resorbable Electronic Devices for Remotely Controlled Therapy and in vivo Infection Abatement. Proc. Natl. Acad. Sci. U. S. A. 2014, 111 (49), 17385-17389. (8)

Wang, H.; Zhu, B. W.; Ma, X. H.; Hao, Y.; Chen, X. D., Physically Transient Resistive

Switching Memory Based on Silk Protein. Small 2016, 12 (20), 2715-2719. (9)

Yin, L.; Cheng, H. Y.; Mao, S. M.; Haasch, R.; Liu, Y. H.; Xie, X.; Hwang, S. W.; Jain,

H.; Kang, S. K.; Su, Y. W.; Li, R.; Huang, Y. G.; Rogers, J. A., Dissolvable Metals for Transient Electronics. Adv. Funct. Mater. 2014, 24 (5), 645-658. (10)

Kang, S. K.; Hwang, S. W.; Cheng, H. Y.; Yu, S.; Kim, B. H.; Kim, J. H.; Huang, Y. G.;

Rogers, J. A., Dissolution Behaviors and Applications of Silicon Oxides and Nitrides in Transient Electronics. Adv. Funct. Mater. 2014, 24 (28), 4427-4434. (11)

Hwang, S. W.; Park, G.; Edwards, C.; Corbin, E. A.; Kang, S. K.; Cheng, H. Y.; Song, J.

K.; Kim, J. H.; Yu, S.; Ng, J.; Lee, J. E.; Kim, J.; Yee, C.; Bhaduri, B.; Su, Y.; Omennetto, F. G.; Huang, Y. G.; Bashir, R.; Goddard, L.; Popescu, G.; Lee, K. M.; Rogers, J. A., Dissolution Chemistry and Biocompatibility of Single-Crystalline Silicon Nanomembranes and Associated Materials for Transient Electronics. Acs Nano 2014, 8 (6), 5843-5851.


29


(12)

Page 30 of 42

Yamada, H.; Nakao, H.; Takasu, Y.; Tsubouchi, K., Preparation of Undegraded Native

Molecular Fibroin Solution from Silkworm Cocoons. Mat Sci Eng C-Bio S 2001, 14 (1-2), 41-46. (13)

Rockwood, D. N.; Preda, R. C.; Yucel, T.; Wang, X. Q.; Lovett, M. L.; Kaplan, D. L.,

Materials Fabrication from Bombyx mori Silk Fibroin. Nat. Protoc. 2011, 6 (10), 1612-1631. (14)

Tao, H.; Kaplan, D. L.; Omenetto, F. G., Silk Materials - A Road to Sustainable High

Technology. Adv. Mater. 2012, 24 (21), 2824-2837. (15)

Perez-Rigueiro, J.; Viney, C.; Llorca, J.; Elices, M., Mechanical Properties of Single-Brin

Silkworm Silk. J. Appl. Polym. Sci. 2000, 75 (10), 1270-1277. (16)

Romero, I. S.; Schurr, M. L.; Lally, J. V.; Kotlik, M. Z.; Murphy, A. R., Enhancing the

Interface in Silk-Polypyrrole Composites through Chemical Modification of Silk Fibroin. Acs Appl Mater Inter 2013, 5 (3), 553-564. (17)

Yamaura, K.; Kuranuki, N.; Suzuki, M.; Tanigami, T.; Matsuzawa, S., Properties of

Mixtures of Silk Fibroin Syndiotactic-Rich Poly(Vinyl Alcohol). J. Appl. Polym. Sci. 1990, 41 (910), 2409-2425. (18)

Liang, C. X.; Hirabayashi, K., Improvements of the Physical-Properties of Fibroin

Membranes with Sodium Alginate. J. Appl. Polym. Sci. 1992, 45 (11), 1937-1943. (19)

Kweon, H.; Ha, H. C.; Um, I. C.; Park, Y. H., Physical Properties of Silk Fibroin/Chitosan

Blend Films. J. Appl. Polym. Sci. 2001, 80 (7), 928-934. (20)

Seo, J. W.; Kim, H.; Kim, K.; Choi, S. Q.; Lee, H. J., Calcium‐Modified Silk as a

Biocompatible and Strong Adhesive for Epidermal Electronics. Adv. Funct. Mater. 2018, 1800802. (21)

Yucel, T.; Lovett, M. L.; Keplan, D. L., Silk-Based Biomaterials for Sustained Drug

Delivery. J. Controlled Release 2014, 190, 381-397.


30

Page 31 of 42 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(22)

Zhu, B. W.; Wang, H.; Leow, W. R.; Cai, Y. R.; Loh, X. J.; Han, M. Y.; Chen, X. D., Silk

Fibroin for Flexible Electronic Devices. Adv. Mater. 2016, 28 (22), 4250-4265. (23)

Kim, D. H.; Viventi, J.; Amsden, J. J.; Xiao, J. L.; Vigeland, L.; Kim, Y. S.; Blanco, J. A.;

Panilaitis, B.; Frechette, E. S.; Contreras, D.; Kaplan, D. L.; Omenetto, F. G.; Huang, Y. G.; Hwang, K. C.; Zakin, M. R.; Litt, B.; Rogers, J. A., Dissolvable Films of Silk Fibroin for Ultrathin Conformal Bio-Integrated Electronics. Nat Mater 2010, 9 (6), 511-517. (24)

Kim, D. H.; Ghaffari, R.; Lu, N. S.; Wang, S. D.; Lee, S. P.; Keum, H.; D'Angelo, R.;

Klinker, L.; Su, Y. W.; Lu, C. F.; Kim, Y. S.; Ameen, A.; Li, Y. H.; Zhang, Y. H.; de Graff, B.; Hsu, Y. Y.; Liu, Z. J.; Ruskin, J.; Xu, L. Z.; Lu, C.; Omenetto, F. G.; Huang, Y. G.; Mansour, M.; Slepian, M. J.; Rogers, J. A., Electronic Sensor and Actuator Webs for Large-Area Complex Geometry Cardiac Mapping and Therapy. Proc. Natl. Acad. Sci. U. S. A. 2012, 109 (49), 1991019915. (25)

Kim, U. J.; Park, J.; Kim, H. J.; Wada, M.; Kaplan, D. L., Three-Dimensional Aqueous-

Derived Biomaterial Scaffolds from Silk Fibroin. Biomaterials 2005, 26 (15), 2775-2785. (26)

Diao, Y. Y.; Liu, X. Y.; Toh, G. W.; Shi, L.; Zi, J., Multiple Structural Coloring of Silk-

Fibroin Photonic Crystals and Humidity-Responsive Color Sensing. Adv. Funct. Mater. 2013, 23 (43), 5373-5380. (27)

Murphy, A. R.; Kaplan, D. L., Biomedical Applications of Chemically-Modified Silk

Fibroin. JMCh 2009, 19 (36), 6443-6450. (28)

Burke, K. A.; Brenckle, M. A.; Kaplan, D. L.; Omenetto, F. G., Evaluation of the Spectral

Response of Functionalized Silk Inverse Opals as Colorimetric Immunosensors. Acs Appl Mater Inter 2016, 8 (25), 16218-16226.


31


(29)

Page 32 of 42

Zhou, Z. T.; Shi, Z. F.; Cai, X. Q.; Zhang, S. Q.; Corder, S. G.; Li, X. X.; Zhang, Y. S.;

Zhang, G. Z.; Chen, L.; Liu, M. K.; Kaplan, D. L.; Omenetto, F. G.; Mao, Y.; Tao, Z. D.; Tao, T. H., The Use of Functionalized Silk Fibroin Films as a Platform for Optical Diffraction-Based Sensing Applications. Adv. Mater. 2017, 29 (15), 1605471. (30)

Tsioris, K.; Raja, W. K.; Pritchard, E. M.; Panilaitis, B.; Kaplan, D. L.; Omenetto, F. G.,

Fabrication of Silk Microneedles for Controlled-Release Drug Delivery. Adv. Funct. Mater. 2012, 22 (2), 330-335. (31)

Lee, J.; Park, S. H.; Seo, I. H.; Lee, K. J.; Ryu, W., Rapid and Repeatable Fabrication of

High A/R Silk Fibroin Microneedles Using Thermally-Drawn Micromolds. Eur. J. Pharm. Biopharm. 2015, 94, 11-19. (32)

Kim, S.; Mitropoulos, A. N.; Spitzberg, J. D.; Tao, H.; Kaplan, D. L.; Omenetto, F. G., Silk

Inverse Opals. Nat Photonics 2012, 6 (12), 817-822. (33)

Li, C. M.; Hotz, B.; Ling, S. J.; Guo, J.; Haas, D. S.; Marelli, B.; Omenetto, F.; Lin, S. J.;

Kaplan, D. L., Regenerated Silk Materials for Functionalized Silk Orthopedic Devices by Mimicking Natural Processing. Biomaterials 2016, 110, 24-33. (34)

Wu, F.; Tien, L. W.; Chen, F. J.; Berke, J. D.; Kaplan, D. L.; Yoon, E., Silk-Backed

Structural Optimization of High-Density Flexible Intracortical Neural Probes. JMemS 2015, 24 (1), 62-69. (35)

Bettinger, C. J.; Cyr, K. M.; Matsumoto, A.; Langer, R.; Borenstein, J. T.; Kaplan, D. L.,

Silk Fibroin Microfluidic Devices. Adv. Mater. 2007, 19 (19), 2847-2850. (36)

Lawrence, B. D.; Marchant, J. K.; Pindrus, M. A.; Omenetto, F. G.; Kaplan, D. L., Silk

Film Biomaterials for Cornea Tissue Engineering. Biomaterials 2009, 30 (7), 1299-1308.


32

Page 33 of 42 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(37)

Perry, H.; Gopinath, A.; Kaplan, D. L.; Dal Negro, L.; Omenetto, F. G., Nano- and

Micropatterning of Optically Transparent, Mechanically Robust, Biocompatible Silk Fibroin Films. Adv. Mater. 2008, 20 (16), 3070-3072. (38)

Mondia, J. P.; Amsden, J. J.; Lin, D. M.; Dal Negro, L.; Kaplan, D. L.; Omenetto, F. G.,

Rapid Nanoimprinting of Doped Silk Films for Enhanced Fluorescent Emission. Adv. Mater. 2010, 22 (41), 4596-4599. (39)

Parker, S. T.; Domachuk, P.; Amsden, J.; Bressner, J.; Lewis, J. A.; Kaplan, D. L.;

Omenetto, F. G., Biocompatible Silk Printed Optical Waveguides. Adv. Mater. 2009, 21 (23), 2411-2415. (40)

Kurland, N. E.; Dey, T.; Kundu, S. C.; Yadavalli, V. K., Precise Patterning of Silk

Microstructures Using Photolithography. Adv. Mater. 2013, 25 (43), 6207-6212. (41)

Sun, Y. L.; Li, Q.; Sun, S. M.; Huang, J. C.; Zheng, B. Y.; Chen, Q. D.; Shao, Z. Z.; Sun,

H. B., Aqueous Multiphoton Lithography with Multifunctional Silk-Centred Bio-Resists. Nat Commun 2015, 6. (42)

Kim, S.; Marelli, B.; Brenckle, M. A.; Mitropoulos, A. N.; Gil, E. S.; Tsioris, K.; Tao, H.;

Kaplan, D. L.; Omenetto, F. G., All-Water-Based Electron-Beam Lithography Using Silk as a Resist. Nat Nanotechnol 2014, 9 (4), 306-310. (43)

Park, J.; Lee, S. G.; Marelli, B.; Lee, M.; Kim, T.; Oh, H. K.; Jeon, H.; Omenetto, F. G.;

Kim, S., Eco-Friendly Photolithography Using Water-Developable Pure Silk Fibroin. Rsc Adv 2016, 6 (45), 39330-39334. (44)

Joseph, J.; Singh, S. G.; Vanjari, S. R. K., Leveraging Innate Piezoelectricity of Ultra-

Smooth Silk Thin Films for Flexible and Wearable Sensor Applications. IEEE Sens. J. 2017, 17 (24), 8306-8313.


33


(45)

Page 34 of 42

Tao, H.; Amsden, J. J.; Strikwerda, A. C.; Fan, K. B.; Kaplan, D. L.; Zhang, X.; Averitt,

R. D.; Omenetto, F. G., Metamaterial Silk Composites at Terahertz Frequencies. Adv. Mater. 2010, 22 (32), 3527-3531. (46)

Tsioris, K.; Tao, H.; Liu, M. K.; Hopwood, J. A.; Kaplan, D. L.; Averitt, R. D.; Omenetto,

F. G., Rapid Transfer-Based Micropatterning and Dry Etching of Silk Microstructures. Adv. Mater. 2011, 23 (17), 2015-2019. (47)

Brenckle, M. A.; Kaplan, D. L.; Omenetto, F. G., Direct Transfer Printing of Water

Hydrolyzable Metals onto Silk Fibroin Substrates through Thermal-Reflow-Based Adhesion. Adv Mater Interfaces 2016, 3 (15), 1600094. (48)

Gitler, D.; Xu, Y. M.; Kao, H. T.; Lin, D. Y.; Lim, S. M.; Feng, J.; Greengard, P.;

Augustine, G. J., Molecular Determinants of Synapsin Targeting to Presynaptic Terminals. J. Neurosci. 2004, 24 (14), 3711-3720. (49)

Gage, G. J.; Kipke, D. R.; Shain, W., Whole Animal Perfusion Fixation for Rodents. Jove-J

Vis Exp 2012, (65). (50)

Katz, J. J., Anhydrous Hydrogen Fluoride as a Solvent for Proteins and Some Other

Biologically Important Substances. Arch. Biochem. Biophys. 1954, 51 (1), 293-305. (51)

Yazawa, K.; Ishida, K.; Masunaga, H.; Hikima, T.; Numata, K., Influence of Water Content

on the Beta-Sheet Formation, Thermal Stability, Water Removal, and Mechanical Properties of Silk Materials. Biomacromolecules 2016, 17 (3), 1057-1066. (52)

Koelmans, W. W.; Peters, T.; Berenschot, E.; de Boer, M. J.; Siekman, M. H.; Abelmann,

L., Cantilever Arrays with Self-Aligned Nanotips of Uniform Height. Nanot 2012, 23 (13), 135301.


34

Page 35 of 42 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(53)

Narayanamoorthy, J.; Tsioris, K.; Omenetto, F. G.; Hopwood, J., Plasma Etching of Silk

Fibroin: Experiments and Models. Plasma Process Polym 2013, 10 (5), 451-458. (54)

Hu, X.; Kaplan, D.; Cebe, P., Determining Beta-Sheet Crystallinity in Fibrous Proteins by

Thermal Analysis and Infrared Spectroscopy. Macromolecules 2006, 39 (18), 6161-6170. (55)

Li, W.; Tang, Q. Y.; Jadhav, A. D.; Narang, A.; Qian, W. X.; Shi, P.; Pang, S. W., Large-

Scale Topographical Screen for Investigation of Physical Neural-Guidance Cues. Sci. Rep. 2015, 5, 8644. (56)

Shin, H.; Lee, H. J.; Chae, U.; Kim, H.; Kim, J.; Choi, N.; Woo, J.; Cho, Y.; Lee, C. J.;

Yoon, E. S.; Cho, I. J., Neural Probes with Multi-Drug Delivery Capability. LChip 2015, 15 (18), 3730-3737. (57)

Lee, H. J.; Son, Y.; Kim, J.; Lee, C. J.; Yoon, E. S.; Cho, I. J., A Multichannel Neural Probe

with Embedded Microfluidic Channels for Simultaneous in vivo Neural Recording and Drug Delivery. LChip 2015, 15 (6), 1590-1597. (58)

Lee, H. J.; Son, Y.; Kim, D.; Kim, Y. K.; Choi, N.; Yoon, E. S.; Cho, I. J., A New Thin

Silicon Microneedle with an Embedded Microchannel for Deep Brain Drug Infusion. Sensor Actuat B-Chem 2015, 209, 413-422. (59)

Kook, G.; Lee, S. W.; Lee, H. C.; Cho, I. J.; Lee, H. J., Neural Probes for Chronic

Applications. Micromachines-Basel 2016, 7 (10). (60)

Tien, L. W.; Wu, F.; Tang-Schomer, M. D.; Yoon, E.; Omenetto, F. G.; Kaplan, D. L., Silk

as a Multifunctional Biomaterial Substrate for Reduced Glial Scarring around Brain-Penetrating Electrodes. Adv. Funct. Mater. 2013, 23 (25), 3185-3193.


35


Page 36 of 42

Table of Contents/Abstract Graphic


36

Page 37 of 42 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


a) Schematics of fabrication process of an aluminum hard mask on a silk fibroin film (AMoS). The fibroin film is decoupled from the harsh chemicals used in UV photolithography. b) Schematics of subsequent processes utilizing the AMoS mask and scanning electron microscopy (SEM) images of final patterns (scale bars: 20 µm): patterning of the fibroin film using oxygen plasma (upper) and patterning of another metal on a fibroin film using the lift-off process (lower). The insets show the cross-sectional schematics of the structures along the dotted line. 160x127mm (288 x 288 DPI)



Fibroin microstructures and metal micropatterns on a fibroin film produced by using the AMoS mask. SEM images of a) 10-µm-wide (scale bars: 30 µm) and b) 2-µm-wide line patterns (scale bars: 10 µm) of 1.2µm-thick fibroin coated on a silicon wafer. c) Cross-sectional SEM image of 10-µm-wide fibroin line patterns on silicon (scale bar: 20 µm). d) SEM image of the edge of the patterned fibroin (scale bar: 1 µm). e-f) Optical images of gold lines defined on fibroin films (scale bars: 200 µm, 20 µm). 160x83mm (288 x 288 DPI)


Page 38 of 42

Page 39 of 42 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


a) Atomic force microscopy (AFM) surface profile of a 30 µm × 30 µm fibroin structure patterned using the AMoS process on a silicon wafer. b) Close-in profile of the 5 µm × 5 µm dotted square in a) (top) and linear surface profile along the dotted line in a) (bottom). c) Comparison of root-mean-square (RMS) surface roughness of fibroin over the area of 5 µm × 5 µm of three cases: water-insoluble fibroin film as a control, the same film on which aluminum layer was deposited and stripped by BHF without patterning, and the patterned film in a). d-o) SEM images of various shapes patterned on fibroin films with different thickness: 700 nm (d-g), 3 µm (h-k), and 15 µm (l-o) (scale bars: 50 µm). The yellow dotted lines indicate sharpening of the corners of the patterns. Next, we evaluated if there was any change in molecular structures in fibroin from the AMoS process. First, we measured FTIR spectra of a fibroin film before and after the process (Figure S3). Individual vibrational peaks and their intensities in the Amide I region (Figure 4) were analyzed by processing the original FTIR spectra with FSD; the FSD spectra were fitted as in a previously demonstrated analysis.54 The spectra revealed no difference in both positions and intensities of vibrational modes for random coils (R), betasheets (B), alpha-helices (A), turns (T), and side chains (S) after the process. In contrary, a water-soluble amorphous fibroin film showed significantly larger intensities of random coils to beta sheets.54 Likewise, no difference of vibrational peaks appeared in the Amide II region before and after the process (Figure S4). These results show that the AMoS process causes no significant perturbation in molecular structures in the fibroin films. 160x146mm (288 x 288 DPI)



Figure 4. Peak analysis and assignment of Fourier self-deconvolution (FSD) of the Fourier-transform IR (FTIR) spectra in Amide I region of the fibroin film before and after the process. The measured and analyzed FTIR spectra of an amorphous fibroin film is also shown for comparison. Abbreviations for letters of vibrational assignments are random coils (R), beta-sheets (B), alpha-helices (A), turns (T), and side chains (S). 80x136mm (288 x 288 DPI)


Page 40 of 42

Page 41 of 42 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 5. Biocompatibility assessment of the AMoS process. a) Top: schematics of fibroin-coated silicon. Middle and bottom: representative confocal microscopic images of live (green) and dead (red) primary cortical neurons cultured on the fibroin-coated silicon (scale bars: 100 µm in middle, 500 µm in bottom). b) Top: schematics of fibroin-patterned silicon. Middle and bottom: representative confocal microscopic images of live (green) and dead (red) primary cortical neurons cultured on the fibroin-patterned silicon (scale bars: 100 µm in middle, 500 µm in bottom). One of line patterns is depicted as a thin white-shaded rectangle. Honeycomb patterns are depicted as hexagons in different sizes (dotted and white shade in bottom). c) Top: schematics of gold-patterned fibroin. Middle: representative confocal microscopic image of live (green) and dead (red) primary cortical neurons cultured on the gold-patterned fibroin (scale bar: 100 µm). Line patterns are depicted as thin white-shaded rectangles. d) Bar graph displaying mitochondrial activity of primary cortical neurons cultured on various substrates involved in the AMoS process. The mitochondrial activity represents normalized absorbance of formazan at 450 nm: background-subtracted absorbance (A450 – A450,blank) normalized by background-subtracted absorbance of the polystyrene plate (control; A450,control – A450,blank). Error bars indicate standard deviation of mean (triplicate measurements for each case). ns denotes indistinguishable statistical differences from the multiple comparisons of one-way ANOVA. 160x154mm (288 x 288 DPI)



Figure 6. a) Fabricated polyimide-based flexible neural electrodes with fibroin patterns on a 100-mm silicon wafer (scale bar: 20 mm). The insets are microscope images of the fibroin patterns on the electrodes. Left: electrocorticography (ECoG) electrodes (scale bar: 400 µm). Right: a neural probe (scale bar: 100 µm). b) Photograph of the fabricated ECoG electrodes (scale bar: 2 mm). Rhodamine B was loaded into fibroin to visualize the fibroin patterns. The inset is a close-up view of one of the microelectrode array (scale bar: 200 µm): alignment margin between the electrode opening and the fibroin pattern was below 50 µm, which was easily achievable with alignment limit of 1 µm of our process. c) Left: photographs showing the sequence of Rhodamine B delivery onto a mouse brain from the ECoG electrode through the micropatterned fibroin layers with high spatial resolution (scale bar: 5 mm). Right: close-up view of the sites showing localized drug delivery (scale bar: 1 mm). d) Wafer-scale fabrication process of µm-scale series RC components using the AMoS process by interleaving two fibroin layers and two metal layers. e) Optical microscope image of the fabricated device. f) Photograph of fabricated microscale RLC circuits on a fibroin film (scale bar: 3 mm). Gold (yellow) and silver (white) are patterned on different layers with a fibroin layer as a passivation layer in between. g) Photograph of metal patterns on a flexible fibroin film (scale bar: 5 mm).


Page 42 of 42

Page 43 of 42 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


160x190mm (288 x 288 DPI)


Wafer-Scale Multilayer Fabrication for Silk Fibroin-Based

Recommend Documents