Selective Fabrication of Nanowires with High Aspect Ratios Using a

May 13, 2019 - Metropolitan University, Minamiohsawa, Hachioji, Tokyo 192-0397. ‡. Department of Chemistry, Beijing Key Laboratory of Microanalytica...
1 downloads 0 Views 631KB Size
Subscriber access provided by SUNY PLATTSBURGH

Article

Selective Fabrication of Nanowires with High Aspect-Ratios using a Diffusion Mixing Reaction System for Applications in Temperature Sensing Haifeng Lin, Sifeng Mao, Hulie Zeng, Yong Zhang, Masato Kawaguchi, Yumi Tanaka, Jin-Ming Lin, and Katsumi Uchiyama Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.9b01122 • Publication Date (Web): 13 May 2019 Downloaded from http://pubs.acs.org on May 13, 2019

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 7 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

Analytical Chemistry

Selective Fabrication of Nanowires with High Aspect-Ratios using a Diffusion Mixing Reaction System for Applications in Temperature Sensing Haifeng Lin†, Sifeng Mao‡, Hulie Zeng§, Yong Zhang†, Masato Kawaguchi†, Yumi Tanaka†, Jin-Ming Lin*, ‡, Katsumi Uchiyama*, † Department of Applied Chemistry, Graduate School of Urban Environmental Sciences, Tokyo Metropolitan University, Minamiohsawa, Hachioji, Tokyo 192-0397 ‡ Department of Chemistry, Beijing Key Laboratory of Microanalytical Methods and Instrumentation, MOE Key Laboratory of Bioorganic Phosphorus Chemistry & Chemical Biology, Tsinghua University, Beijing 100084, China § School of Pharmacy, Fudan University, 826 Zhangheng Road, Shanghai 201203, China †

ABSTRACT: The selective fabrication of highly ordered nanowires with high aspect ratios was of low reproducibility, which remains a challenge for laboratory research. In this paper, we report a novel approach for selective fabrication of conductive nano-wires on a solid surface via diffusion mixing reaction system formed by a chemical pen. The nanoscale-mixing region was achieved by appropriately adjusting the viscosity of the solution and other parameters with the aid of dyes functioned as a flow boundary indicator. Finite element simulations and analysis were performed to understand the generation of mixing regions and guide the improvement of the chemical pen design. Under the optimal parameters, high aspect ratio silver nanowires (aspect ratio≈1800) were obtained. Silver nanowire arrays with uniform width, gradient width and complex patterns were successfully fabricated. The theoretical value of the temperature coefficient of resistance (TCR) for silver was 0.0038 Ω/℃. A single silver wire temperature sensor with 7-fold increase in temperature coefficient resistance (0.0261 Ω/℃) was fabricated to show the advantages of the chemical pen in the fabrication of nano-sensors. With the freedom of the region, simple operability and applicability, the chemical pen was expected to a potential and advanced method for selective nano-modification and processing on subcellular interfaces.

Nanofabrication has greatly promoted the miniaturization and integration of devices, and has led to extensive research in this area1-4. Benefiting from the large surface-to-volume ratio of nanostructures, nanostructured sensors have been utilized for the highly sensitive and specific detection of gases5-7, chemicals8-11 and biological specimens12-17. To this end, various chemical and physical methods, such as: projection lithography18, scanning beam19, thin film deposition20, selfassembly21,22 have been developed. Nanowires, onedimensional nanostructures, have attracted considerable attention for their outstanding electrical, magnetic, and optical properties23. However, the reproducible production of highly ordered and high aspect ratio nanowires that is essential for developing reliable nano-sensors using the above methods remain a challenge. In this case, advances in techniques that are intended to address the above limitations are vital and timely issues that need to be addressed. Substantial efforts have been made to develop methods for fabricating highly ordered and high length-width ratio nanowires. It has been reported that dip-pen nanolithography (DPN) can be used to deposit alkane thiols on a gold film with a line width of 30 nm24. Wang, F., etc. utilized direct laser lithography to pattern GaAs nanowires with a width of 800 nm on a transparent quartz substrate25. Interfacial reactions of two phase laminar flow in a microchip can be used to generate wires with feature sizes of less than 5 µm26. Hierarchical

hydrodynamic flow confinement is capable of patterning an antibody array with a resolution of 5 μm 27. Those approaches show outstanding performance in the area of micro- and nanofabrication, but they still face shortcomings, such as the high demand for precision equipment and strict environmental conditions, material limitations and limited spatial resolution. Chemical pen, developed by our research group, represents a versatile tool for nanofabrication or modification on complex surfaces28. Its large-scope has been confirmed in micromodification28,29, the direct writing of nanowires29,30 and single cell adhesion assay31-33. Although previous studies have reported that a chemical pen was capable of patterning nanowire29, stable fabrication of high aspect-ratio metal nanowires with a long length on a solid surface is still a huge challenge. Included among these challenges are vulnerability to changes in environment, where a small disturbance could destroy the generation of nanowires. The lengths of previously reported nanowires have not exceeded 20 μm, 29 thus making them difficult to use in nano-sensors. In addition, for a nanowire to be used in a sensor, long nanowires would be more competitive, since the sensitivity increases with the nanowire length available for interaction with the analytes. However, such structures are currently difficult to fabricate. Temperature sensors are widely used in our daily lives and in various areas of research34. For example, they play an important role in maintaining a precise temperature that is vital for

ACS Paragon Plus Environment

Analytical Chemistry 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

chemical or biological reactions. There are various types of temperature sensors, including thermostats, thermistors, resistive temperature detectors (RTD) and thermocouples. Among them, the resistive temperature detector (RTD) is attractive for its simplicity and low cost. Compared to a film resistive temperature detector, the nanowire resistive temperature detector has the advantage in thermal sensitivity and accuracy35. However, nanowire resistive temperature detectors produce a small output change in temperature which is a challenge and an indication that there is a great need for the development nanowire sensors with a high temperature coefficient. Herein, we reported on a micro-chemical pen (MCP) with a diffusion mixing reaction system (DMRS) that provided an efficient means of fabricating well-ordered and high aspectratio metallic wires at the micro or nanometer scale with a high degree of spatial resolution. The proof-of-concept of the MCP was demonstrated by patterning micro or nanometer silver wires via a silver mirror reaction. The nanoscale-mixing region was achieved by appropriately adjusting the viscosity of the solution and other parameters with the aid of dyes functioned as a flow boundary indicator. Using this procedure, it was possible to control the widths of silver wires from 40 μm to 167 nm by appropriately adjusting the parameters. Single silver wire temperature sensors with a 7-fold increase in the temperature coefficient of resistance (0.0261 Ω/℃) were manufactured as a demonstration of the advantages of chemical pen in fabrication nano-sensor. EXPERIMENTAL SECTION Chemicals and Materials. Unless otherwise specified, all the reagents was dissolved in water purified by water Milli-Q system system (Sitaba Scientific Technology, Itd., Japan ). Silver nitrate (AgNO3, Sigma. Aldrich, USA), ammonium hydroxide (NH3·H2O, 28%, Tokyo Chemical Industry Co., Ltd., Japan ), glucose (KANTO Chemical Co., INC, Japan), tin(Ⅱ)chloride (Wako Pure Chemical industries, Ltd., Japan), formaldehyde (36-38%, Wako Pure Chemical industries, Ltd., Japan), sodium hydroxide (Wako Pure Chemical industries, Ltd., Japan), hydrochloric acid (Wako Pure Chemical industries, Ltd., Japan ), rhodamine B (Tokyo Chemical Industry Co., Ltd., Japan) and sodium fluorescein (Tokyo Chemical Industry Co., Ltd., Japan) were used as received. Micro slide glasses were purchased from Matsunami (S7111, Tokyo, Japan). Glass capillary with filament (i.d. 500 μm, o.d. 1000 μm) were purchased from Narishige Scientific Instrument Lab (Tokyo, Japan). Heat-shrinkable tube were purchased from Trusco Co., Ltd. (Tokyo, Japan). Capillary tubes (i.d. 60 μm, o.d. 220 μm) were purchased from GL Sciences Inc. (Tokyo, Japan). PEEK tube (i.d. 350 μm, o.d. 500 μm) were purchased from IDEX Health & Science LLC (Tokyo, Japan). Conductive adhesive were purchased from Elephantech CO., Ltd (Tokyo, Japan). Manufacture of the chemical pen. Micro chemical pen was fabricated by the thermal stretching technology. Three glass capillaries (o.d. 500 μm, i.d. 350 μm, length 100 mm) were assembled through heat-shrink tubes. After the heat-shrinkable tubes shrunk, the three glass capillaries are stably fixed to each other. Then, the three glass-capillary bundles were stretched by a distance of 3 cm via a heat puller (PB-7, NARISHIGE, Tokyo, Japan), in which a sharp contraction of the capillary diameter occurred. The three glass-capillary bundles were then cut in the intermediate section into two sections using a glass cutter.

Page 2 of 7

Three capillary tubes (o.d. 220 μm, i.d. 60 μm, length: 300 mm) were used to connect the prepared three glass capillary tubes and three PEEK tubes (o.d. 500 μm, i.d. 350 μm, length: 50 mm). Finally, epoxy resin adhesive (AR-R30, NICHIBAN, Japan) was used to seal the seams and bond the preparing chemical pen on the holder. MCP of different inner diameters can be made by different heating times. The heating time is inversely proportional to the inner diameter of the MCP. MCP with a smaller inner diameter is needed for the experiment. Setup of diffusion mixing reaction systems (DMRS). We chose mirror reaction to generate silver wires on glass substrate, a silver wire produced accompanying with the movement of the glass substrate which is set on a XY stage (sigma KOKI Co., Ltd, Tokyo, Japan), installed on the inverted microscope (Olympus IX71). Tollesns reagent used contained silver nitrate (AgNO3, 5 mM), ammonia water (50 mM), sodium hydroxide (0.1 M) and rhodamine B (10 μM). The reducing agent was prepared from glucose solution (5 mM), formaldehyde (5 mM), sodium hydroxide (0.1 M) and sodium fluorescein solution(10 μM). sodium fluorescein and rhodamine B served as indicators for displaying the flow profiles. MCP was held on a manual XYZ stage (sigma KOKI Co., Ltd, Tokyo, Japan). The tip of MCP was precisely adjusted to keep a certain gap with substrate and maintained its position constant during patterning. The injection stream was aspirated by the negative pressure of the absorption holes to form a convective diffusion layer. A mixing reaction region was generated in the middle of the convection diffusion layer. Accompanying with the movement of the XY stage, controlled by an Opt Mike Controller (OMEC-2BF,Sigma KOKI Co., Ltd, Tokyo, Japan) , glass substrate moved to generate silver wire. By optimizing the experimental parameters, metallic wires of different widths and even nanometers can be obtained. The flow rate of injection and aspiration was controlled by the syringes pump (NE-100-SPR, USA). Simulation by Comsol Multiphysics software. For numerical simulation, using a laminar flow and diffusion module of Comsol Multiphysics (COMSOL 5.3, MA), we constructed a model with incompressible fluids, open boundaries, and surface slip conditions. The model combines the solutions of the Navier-Stokes equation and the convectiondiffusion equation. Both the injection liquid and the immersion liquid were selected as water (uncompressible Newtonian fluid, density 998 kg·m-3, dynamic viscosity 0.001 N sm-2), and all boundary liquid/suction channels outside the flowing liquid were defined as open (under atmospheric pressure). Anti-slip conditions are defined on the surface of the substrate and the MCP head, and the simulation is performed at a steady and transient state. The diffusion coefficients for silver ion and glucose and formaldehyde mixed solution were 2.65×10-11 m2 s-1 and 1.36×10-11 m2 s-1, respectively. Additional rules for other conditions were set same to experimental data, such as QI = 10 μl h-1, QA = 300 μl h-1, Gap = 20 μm and ID = 50 μm, to ensure reliable simulation results consistent with actual data. Fabrications of silver wires under different parameters. In order to firmly fix the silver wire on the surface of the glass, the glass substrate was previously immersed in a saturated aqueous solution of tin chloride for surface treatment. A thin fixing the silver wire. The width of the patterning silver wire is affected by various parameters, and we had investigated several parameters, including flow ratio Q (the ratio of the aspiration

ACS Paragon Plus Environment

Page 3 of 7 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

Analytical Chemistry flow rate QA to the injection flow rate QI), Gap (the distance layer of tin ions was adsorbed on the surface of the glass for between the glass substrate and the tip of the chemical pan), diameter of MCP (ID), move speed of glass substrate (υ), concentration of reagent, and the viscosity of reagent. Characterizations of the silver patterns. The Ag wires were charactered by an SEM system (JSM-7500, JEOL Co., Tokyo, Japan) and an AFM system (AFM5010, Hitachi Co., Tokyo, Japan). SEM images were performed on SEM system with 5 kV accelerating voltage, 67 μA emission current and a 6.9 μm working distance. Fabrication of the single metallic wire temperature sensor. The silver wire was supported with glass substrate and placed in a transparent chamber with inlet and outlet for measurements of voltage and current. We glued the seam of copper wire and patterning Ag wire using conductive adhesive. After 24 h at room temperature, the conductive adhesive solidified, and the sensor was ready to use. The temperature gradually increased using a thermal control system (OCE-TCR12075WL, OHM ElECTIRC Co., Ltd., Shizuoka, Japan) at a rate of 0.2 ℃/s. We used a conductive chamber to form a electrostatic shield for eliminating interference of ambient electro-magnetic noise and the temperature sensor and detection devices were contained by the conductive chamber. In order to prevent the breakage of the silver wire by resistance heat during measuring, small voltage generator was used to apply voltage. By measurement of current, we could successfully measure the resistance of silver nanowire without breakage. RESULTS AND DISCUSSION Setup and operations of the system. The setup of the diffusion mixing reaction system (DMRS) is illustrated in Figure 1 and Figure S1. The system includes a micro chemical pen (MCP), an inverted microscope system, a micro XY stage and an injection-suction system. The MCP was made from three heat-stretched glass capillaries with a tapered and attached end (Figure 1a), in which each pole has an inner diameter (i.d.) of 50 μm and an outer diameter (o.d.) of 100 μm (Figure 1d and

Figure 1. Schematic of the diffusion mixing reaction system (DMRS). (a) Construction of the DMRS. (b) Bottom view of the end of the chemical pen. (c) Illustration of silver patterning process. (d) The dimension of the end of micro chemical pen (MCP), and its fluorescence profile of hydrodynamic flow. (e) Scanning electron microscope (SEM) photograph of MCP.

Figure 2. Numeric simulation of the diffusion mixing reaction system. (a) Fluorescence assisted displaying fluid profiles using two dyes (rhodamine B and sodium fluorescein). (b) Simulating the concentration distribution at the surface of the substrate. (c) Profile of the fluorescence intensity along “line A” in the experiments. (d) Profile of the fluorescence intensity along the green dash line B in the simulations. (e) Fluorescent labeled fluid profiles under different Q and gaps. (f) The calibration of simulation values of the effective mixing region corresponding to the experimentally measured ones. Figure 1e). The inlets of the MCP were sealed with three capillaries (60 μm i.d., 220 μm o.d.) that were connected to three micro syringe pumps by PEEK tubes (350 μm i.d., 500 μm o.d.). Two of those apertures were used for injections and another one for aspiration. The tip of the MCP was placed close to the glass substrate by manually moving an XYZ stage that served as its holder. The XYZ stage was capable of adjusting the position at a gap (G) between the tip of the MCP and glass substrate with an accuracy of 1 μm. Here, the glass substrate in a petri dish filled with the surrounding media (e.g. water) was fixed on the electromotive XY stage, the movement of which could be controlled arbitrarily and automatically by a computer. In the experiments, the substrate fixed on the XY stage was moved while the position of the MCP was maintained steady. An inverted microscope was used to display the flow profile and survey the convection-diffusion conditions. The operation system is shown in Figure S2. Based on the concept of hydrodynamic flow confinement (HFC), a radial flow field formed between the injection aperture and the aspiration aperture (Figure 1b), when the gap (G) is sufficiently small and the aspiration flow rate (QA) is sufficiently larger than the injection flow rate (QI = Q1= Q2). The mixing region formed at the interface of the two radials flows underneath the end of the MCP when the parameters were properly adjusted. In the mixing region, different species mixed to generate new

ACS Paragon Plus Environment

Analytical Chemistry 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

substances that resulted in the formation of a precipitate on the glass substrate (Figure 1c). We observed the fluorescent fluid profile (Figure 1d) via an inverted microscope. Simulation and analysis of the mixing region. Numeric simulations were performed using the Comsol Multiphysics software to fully understand the generation of the diffusion mixing region (Figure 2 and Figure S3). Here, the 3D model of MCP was established, as shown in Figure S3a, whose geometry was the same as that in the experiment. Navier-Stoke equations and the convection-diffusion equation were coupled in the simulation. Tollens’ reagent (10 mM) and a glucose solution (10 mM) were set as the two individual injections. Gap = 20 μm, QA = 500 μL h-1 and Q1 = Q2 = 10 μL h-1 were used in the simulations. The concentration of the substance gradually decreased as the distance of diffusion increased. To determine the effective boundary of the mixing region, we defined 10% of the original concentration as the effective reaction concentration. In this case, we obtained the simulated mixing region contour (Figure 2b). The concentration in the contour was higher than 10% of the original concentration. Through a horizontal and vertical concentration distribution analysis of the effective mixed region, we obtained the concentration distribution for the two chemical substances in the mixing region along the lines B (Figure 2d) and line C (Figure S3e) and the geometrical size of the effective mixing region. The width of the mixing region was determined to be 900 nm, and the length was 5.8 μm. In order to characterize and measure the mixing region, we used rhodamine B and sodium fluorescein as indicators for displaying the fluidic boundary, as shown in Figure 2a. Rhodamine B emits a 610 nm wavelength red light under a 552 nm excitation light and sodium fluorescein emits a 515 nm wavelength light at 495 nm excitation light (Figure S4). To simultaneously obtain the fluorescence images of rhodamine B and sodium fluorescein, we selected an excitation light of 480550 nm (Figure 2a). We obtained the profile of the fluorescence intensity along “line A” (Figure 2c) using the “Image J” software, which showed vertical distribution across the mixing region. Due to the presence of ambient light, there were interference signals in the spectrum of the fluorescence intensity. Therefore, we defined the boundary conditions of the mixing region with a normalized fluorescence intensity that was three times larger than the noise intensity. In this case, it should be noted that the mixing region had a width of 4 μm. With different gaps and flow ratios (Q=QA/QI), mixing region presented different shapes and sizes. When the flow ratio increased and the distance reduced, the size of the mixing area would become smaller (Figure 2e). This trend was consistent with our simulation results and other reported findings36. The linearity of the simulation values of the effective mixing region corresponded to the experimentally measured values (Figure 2f) showing the relevance of our simulation model. Optimizations on the fabrications of micro/nanowires. We utilized the silver mirror reaction to carry out the chemical patterning of silver wires on a glass surface. Here, we used a MCP (75 μm i.d.), Tollens reagent (5 mM), a reducing solution (glucose 10 mM, formaldehyde 5 mM), a flow ratio of 35 (Q, Q=QA/Q1), moving speed of the substrate (υ= 0.17 μm s-1) and gap (G = 20 μm) as a normal mode for silver wire patterning. The injection flow rate (Q1 = 10 μL h-1) was maintained constant. To further understand the effects of dominant parameters, we explored different parameters. Firstly, different

flow ratios (20, 40, 60, 80, and 100) were examined by maintaining G at constant values (Figure 3a). The results indicated that the width of the patterned silver wires decreased

Figure 3. Dominant parameters for the preparation of the sliver wire. (a) Silver wire width for different flow ratios Q. (b) Silver wire width for a MCP with different inner diameters. (c) Silver wire width for different Gaps. (d) A model that explains the effect of gap on the width of patterning silver wires.

with an increase of the flow ratio Q, because of the shortened diffusion time resulting from the increasing flow ratio. In the same manner, MCPs with different inner diameters (118 μm, 98 μm, 81 μm, 61 μm and 40 μm) were examined by maintaining the other parameters the same as the normal mode. The results suggested that a MCP with a larger inner diameter usually resulted in a patterned silver wire with a larger width (Figure 3b). The moving speed of the substrate (υ) and the concentration of the reagents (c) were investigated as shown in Figure S5a and S5b. The concentration of formaldehyde remained positively linear with the width of the patterned silver wires, while the move speed has a negative correlation to width of preparing silver wires. An interesting phenomenon was that the gap (G) had a nonmonotonic relationship with the width of the silver wire (Figure 3c). The width of the patterning first decreased and then increased with increasing gap, and the minimum width of silver wire was obtained when the gap was 20 μm. A theoretical model was established to explain the above phenomenon (Figure 3d). We assumed the mixing region to be a spatial model with a three-dimensional dimensions, with its own intrinsic geometry. We defined the dimension along the gap direction (the green arrows) as its height (H, in the case of the normal mode, H = 20 μm), and the dimension in the horizontal direction (red arrows) as its width. When the gap was smaller than H, the substrate would be in contact with the mixing region. In this case, the contact area would increase with decreasing gap, and the width of the silver wire generated by the reaction would increase. When the gap was larger than H, the substrate would not be in contact with the mixing region, and the silver wire would be formed by the deposition of silver particles generated in the mixing region. Based on the rule of free fall accumulation, in this case, the width of the silver wire would increase with increasing gap. Therefore, the trend for the width of silver wires would be to retain a “V” shape when the size of the gap changed

ACS Paragon Plus Environment

Page 4 of 7

Page 5 of 7 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

Analytical Chemistry from small to large. The simulation results of the geometric dimensions of the mixing region under different slice (Figure S7) shown the simulation contour of mixing region was consistent with our theoretical model for explaining nonmonotonic change of nanowire width as a function of the gap, which revealed the credibility of our theoretical model. An interesting phenomenon was found that the roughness of surface of silver wire (G>H) is larger than that of G