YANG Dong-fang，LU Zi-feng，LIU Hua，SHAN Gui-ye

（Center for Advanced Optoelectronic Functional Materials Research, and Key Laboratory for UV Emitting Materials and Technology of Ministry of Education, National Demonstration Center for Experimental Physics Education, Northeast Normal University, Changchun 130024, China）

Abstract:In order to improve the detection efficiency of Micro Array Electrodes (MAE) and reduce the production cost,a technology combining Digital Micromirror Device (DMD) maskless projection lithography with electrochemical deposition was proposed.Firstly,a user-defined micro array was fabricated by using the advantages of lithography system such as high-resolution PZS motion and imaging flexibility of DMD.And a uniform Au microarray electrode (Au/MAE) was fabricated after obtaining an Au conducting layer by electrodeposition.Then,the electrochemical properties of Au/MAE with different structures were compared by cyclic voltammetry,and the optimized structural parameters were obtained.Finally,the current response of optimized Au/MAE to the glucose with different concentrations and pH values was studied,and the anti-interference of Au/MAE in glucose detection was tested by chronoamperometry.The electrochemical analysis shows that the simple Au/MAE has a significant amperometric response,a strong anti-interference ability and a sensitivity of 101 μA·cm?2·mM?1 in the electrochemical detection of glucose.This method has the advantages of high resolution,high consistency,simple process and low cost,which provides a feasible operation scheme for the fabrication of biosensor array.

Key words:DMD lithography;electrochemical deposition;microstructure array;electrode

1 Introduction

Electrochemical Biosensor Array (EBA) has been widely used in food detection,environmental monitoring,health care and other fields due to its advantages such as high selectivity,high sensitivity and fast analysis speed[1-4].As a signal converter,Micro Array Electrode (MAE) is an important part of EBA,and its performance will directly affect the detection performance of EBA[5-8].However,the fabrication technique of MAE is the basis and key of array electrode research[9-12].The fabrication techniques of microarray electrodes include the following ones.The preparation process based on mask needs to prepare a large number of masks according to different microelectrode structures.In order to realize the optimal design of structural parameters of gold microarray,the design and preparation of the masks need a lot of time and thus have a high cost and poor flexibility[13].The lithography based on etching technology is currently the preferred method for preparing microarray electrodes due to the advantages of good controllability,high precision,good electrode surface flatness and high interbatch reproducibility in preparing planar ordered array electrodes[14].Femtosecond laser direct-writing technology can prepare microelectrode arrays with highresolution graphics,but cannot achieve large-area machining[15].The maskless lithography based on DMD has as many as 2 million micromirrors and each micromirror can be equivalent to an independent light source,so the exposure process is equivalent to the point-to-point exposure of multiple beams.At the same time,the switching frequency of DMD micromirrors matches the running speed of the sample stage,thus realizing the rolling exposure of graphics and the preparation of any periodic structural array.Compared with femtosecond laser direct-writing system,the maskless lithography technology based on DMD has much higher production efficiency and lower fabrication cost.

We combined the DMD maskless multi-step lithography system built by our research group with electrochemical deposition technology to prepare a user-defined micro-structure array[16-19].Indium Tin Oxides (ITO) conducting glass is widely used in the field of electrochemistry because of its excellent properties such as low resistivity,high conductivity and good machining performance.Therefore,with ITO conducting glass as substrate,many periodic and aperiodic MAEs with various shapes were prepared by using the combined machining techniques(including flexible DMD patterning,multi-step lithography and dose modulation technique),designing the mask patterns and controlling the lithography process parameters (exposure time and exposure energy).Then,electrochemical deposition technique was used to uniformly deposit Au onto the prepared microstructural array template to obtain various periodic and aperiodic Au microarray electrodes (Au/MAEs) with arbitrary shapes.The electrochemical properties of the MAEs with different structures were compared and analyzed,and the influence of shape,period and surface area on the REDOX peak current was systematically studied.The Au/MAE structure prepared with the total surface area unchanged and a cell surface area of 1.7×10?5cm2had a higher REDOX peak current.The structural parameters of Au/MAE were optimized.In addition,the distance between adjacent cell structures was minimized to prepare the Au/MAEs with densely arranged cells which could be used for glucose detection.The sensitivity of this Au/MAE was determined as 101 μA·cm?2·mM?1through testing its glucose-detecting performance and anti-interference performance,thereby demonstrating the feasibility of this MAE preparation method.

2 Principle and method of Au/MAE preparation

2.1 DMD projection lithography system and electrochemical deposition experiment setup

Fig.1(a) shows a maskless projection lithography system based on DMD.The DMD (Texas Instruments) consists of an array of 1 024×768 micromirrors.The side length of each micromirror is 1 pixel,that is 1.368 μm.The piezoelectric stage (PZS) can move slightly alongXandYdirections in horizontal plane.Its motion range is 100 μm×100 μm,and its displacement accuracy is up to 7 nm.After the beam expansion and collimation,the LED light source with a central wavelength of 385 nm is incident on DMD at a certain angle to the optical axis of the optical system.The pre-designed pattern is input into DMD chip by computer through the corresponding software.The light beam reflected by DMD carries the image information.Through the projection objective with a minification of 10,the beam is finally focused on the substrate (ITO),which is spin-coated with positive photoresist (S1318) and placed on the PZS.Finally,the loading of a digital mask onto DMD and the high-precision PZS motion is synchronously controlled by computer to realize multi-step lithography.

Fig.1 (a) Schematic diagram of maskless projection lithography system based on DMD and (b) experimental setup of electrochemical deposition圖1 (a) 基于DMD 的無掩模投影光刻系統(tǒng)示意圖；(b) 電化學沉積的實驗裝置圖

Fig.1(b) shows the experimental setup of electrochemical deposition,which mainly includes three-electrode system (working electrode,reference electrode and auxiliary electrode),electrolytic bath,electrolyte,potentiostat,etc.In this experiment,Ag/AgCl was used as the reference electrode,a platinum wire was used as the auxiliary electrode,and an ITO substrate engraved with a microstructural array was used as the working electrode.HAuCl4solution and KCl solution (volume ratio:4∶5) was used as electrolyte.Au nanoparticles were electrodeposited for a certain time under a certain voltage to prepare an Au conducting layer.

2.2 Preparation process of Au/MAE

We combined digital multi-step lithography(DMSL) with electrochemical deposition technique to prepare Au/MAE with ITO as the substrate.The specific experimental process is shown in Fig.2(coler online).Fig.2(a) shows the substrate pretreatment process,that is,washing the ITO glass substrate with acetone,anhydrous ethanol,distilled water and ultrasound successively and then drying it in a nitrogen flow.Fig.2(b) shows the spincoating and pre-drying process of photoresist,which is to apply positive photoresist evenly on the substrate with the aid of a pipette,spin-coat it on ITO glass at a certain speed to obtain a certain thickness of photoresist layer,and then bake the photoresist layer for 10 min on a 95℃ hot plate to remove residual solvent,improve the photoresist sensitivity and make it solidify.Fig.2(c) shows the maskless lithography and development process based on DMD,which is to expose photoresist to the pre-designed mask pattern through DMD lithography system,develop it in 5‰ NaOH solution for a certain time,and then remove the exposed photoresist.Fig.2(d) shows the electrochemical deposition process of Au nanoparticles,which is to use 0.1 M HAuCl4·4H2O and 0.1 M KCl solution(Volume ratio of HAuCl4∶KCl is about 1∶1) as electrolyte,deposit it for a certain time under a certain voltage,and then wash and remove the unexposed photoresist with acetone and distilled water successively to obtain the MAE structure as shown in Fig.2(e).

Fig.2 Flow chart of Au/MAE preparation process.(a) Substrate pretreatment;(b) photoresist spin-coating and pre drying;(c) exposure and development;(d) electrochemical deposition of Au nano layer;(e) photoresist removal圖2 Au/MAE 的制備流程圖。(a) 基片預處理；(b) 旋涂光刻膠和前烘；(c) 曝光并顯影；(d) 電化學沉積Au 納米層；(e) 去除光刻膠

2.3 Au/MAE preparation

A series of MAEs was prepared by DMSL (exposure time:3 s,exposure energy:145 mJ/cm2),Au/MAE was obtained by the electrochemical deposition of Au nano-conducting layer,and the images of MAE and Au/MAE were observed under Olympus microscope.Given the same total surface area (3.7×10?3cm2) and cell surface area (1.7×10?5cm2),the circular,hexagonal and triangular microarray structures with periodic arrangement were prepared as shown in Fig.3(a),(b) and (c).The corresponding microscopic images of Au/MAEs are shown in Fig.3(d),(e) and (f).It can be seen that both MAE and Au/MAE have regular shapes,clear edges,smooth lines and good morphology.Similarly,when the total surface area and cell surface area remained unchanged,we prepared the elliptical,hexagonal and five-pointed Au/MAE structures with aperiodic arrangement,as shown in Fig.4.By using the combination of dose modulation and DMSL,the user-defined MAEs can be fabricated.

Fig.3 Results of periodic structures.The actual exposure and electrodeposition results of MAE and Au/MAE structures of circular ((a) and (d));hexagonal ((b) and (e))and triangular ((c) and (f)) under optical microscope圖3 周期性結(jié)構(gòu)測試結(jié)果。圓形 ((a) 和 (d)) ；六邊形 ((b) 和 (e)) 以及三角形 ((c) 和 (f)) 的MAE 和Au/MAE 在光學顯微鏡下的實際曝光和電沉積結(jié)果

Fig.4 Aperiodic structures results.The actual exposure and electrodeposition results of MAE and Au/MAE of elliptical ((a)and (d))；hexagonal ((b) and (e)) and five-pointed ((c) and (f)) under optical microscope圖4 非周期性結(jié)構(gòu)結(jié)果。橢圓形((a)和(d))；六邊形((b)和(e))以及五角星((c)和(f))的MAE 和Au/MAE 在光學顯微鏡下的實際曝光和電沉積結(jié)果

3 Optimization of Au/MAE structure parameters

Since the height of the prepared MAE cell structure was negligible compared with its surface width,the influence of MAE cell area and density on the electrode performance was investigated.The electrochemical response of the Au/MAEs with different cell shapes in PBS buffer solution (pH 7.0)was investigated by cyclic voltammetry at a scanning rate of 100 mV/s within a potential range of 0?1.4 V.

First,the electrochemical performance of the Au/MAEs with different shapes was studied when the cell surface area was 5.9×10?6cm2,8.3×10?6cm2,1.7×10?5cm2or 3.3×10?5cm2,and the total surface area remained unchanged (3.7×10?3cm2).The cyclic voltammetry (CV) curves of circular,hexagonal and triangular structures with periodic arrangement are shown in Fig.5(a),(b) and (c).It can be seen from Fig.5 (color online) that the Au/MAEs with the same cell shape but different cell areas and with the same total surface area will have different REDOX peak currents.By comparing the three structures,it is found that although they have different cell shapes,their CV currents change in the same way,that is,to increase first and then decrease with the increase of cell structure area.This is because the REDOX peak current is closely related to the number of REDOX contact sites and the strength of signal transmission in microarray structure.When the surface area of a cell structure increases from 5.9×10?6cm2to 3.3×10?5cm2and the total surface area remains unchanged,the number of cell structures will decrease to so that the number of contact sites will decrease and the REDOX peak value will drop.On the other hand,the presence of fewer cell structures will result in the reduction of Au/MAE resistance and the increase of REDOX peak current.Under the competition of these two factors,the peak current will show a change trend of first increasing and then decreasing,instead of linear change.When the cell area is 1.7×10?5cm2,the peak current in both oxidation and reduction will be the highest.Then,with the total surface area unchanged,the Au/MAE structures with different shapes and random arrangement were prepared according to the above experimental parameters,and the electrochemical performance of each Au/MAE structure was tested.The CV curves of elliptical,hexagonal and five-pointed structures with aperiodic arrangement are shown in Fig.5(d),(e) and (f) respectively.It can be seen that for the same aperiodic structure,the REDOX peak current will also increase first and then decrease with the increase of cell surface area when the total surface area remains unchanged.This analysis shows that the total area and cell area are also the main factors affecting the electrochemical performance of an Au/MAE structure with random arrangement,which is consistent with the experimental conclusion obtained above.

Fig.5 Effect of cell surface area on REDOX peak current under the same total surface area.CV diagrams of (a) circular;(b)hexagonal and (c) triangular Au/MAE structures with periodic arrangement and of (d) elliptical;(e) hexagonal and (f)five-pointed Au/MAE structures with aperiodic arrangement圖5 總表面積不變，單元表面積對氧化還原峰電流的影響，周期性排列的(a)圓形；(b)六邊形；(c)三角形以及非周期性排列的(d)橢圓；(e)六邊形；(f)五角星結(jié)構(gòu)的Au/MAE 的CV 圖

It can also be seen from Fig.5 that when both the total surface area and the cell area remain unchanged,the CV current results of Au/MAE structures with periodic or aperiodic arrangement and different cell shapes will be basically the same.In summary,it can be concluded that the REDOX peak current value of Au/MAE structures is closely related to the total surface area and cell structure area,and has nothing to do with cell structure shape.

The Fig.6(a) shows the relationship between the value of oxidation peak current and the surface area of electrode cell structure when the total surface area is constant.As can be seen from the figure,with the increase of cell surface area,the oxidation peak will first increase and then decrease.The curves of the circular,square and hexagonal structures basically coincide,indicating that the oxidation peak current is independent of the cell shape when the total surface area and the cell surface area remain constant.The Fig.6(b) shows the oxidation peak current of a square Au/MAE structure with different cell surface areas as a function of the total surface area.It can be seen that when the cell surface area is constant,the value of oxidation peak current will increases linearly with the increase of the total surface area.

From the above analysis,it can be seen that the main factors affecting the electrochemical performance of Au/MAE are the total surface area and the cell surface area,rather than cell shape.Based on this,we optimized the structural parameters of Au/MAE,that is,the cell surface area (1.7×10?5cm2)producing the optimal REDOX peak current was selected when the total surface area was constant(3.7×10?3cm2).In addition,by minimizing the spacing between two adjacent cell structures,an Au/MAE with densely arranged square cells was prepared,and its performance was tested.

4 Evaluation of Au/MAE performance

4.1 Electrochemical performance of Au and Au/MAE electrodes

According to the optimized parameters obtained from the above study,we prepared the Au/MAE with densely arranged square cells.The cell surface area was 1.7×10?5cm2,and the actual spacing between adjacent cell structures was 1.26 μm.The Fig.7(a) (color online) and Fig.7(b)(color online) respectively show the optical microscope images of the MAE and Au/MAE with densely arranged cells,whose spacing is about 1.26 μm.When the spacing is reduced to 1 pixel (1.368 μm),the gold nanoparticles will spread outward during electrodeposition,resulting in a smaller spacing that cannot be distinguished.Therefore,when the spacing of the design pattern structures is 2 pixels,the microarray composed of square cell structures with this spacing is the densest.The Fig.7(c) (color online) and Fig.7(d) (color online) give the optical microscope images of Au/MAE in full exposure and of single Au electrode respectively.The electrochemical properties of single Au electrode and Au/MAE were compared by cyclic voltammetry.

Fig.7 (a) Microscope image of the MAE with the most densely arranged cells;(b) microscope image of Au/MAE after electrodeposition;(c) microscope image obtained after full exposure;(d) microscope image of single Au electrode after electrodeposition圖7 (a) 密集排列的MAE 的顯微鏡圖像；(b) 電沉積后的Au/MAE 的光學顯微鏡圖；(c)全部曝光后的顯微鏡圖像；(d)電沉積后的單個Au 電極光學顯微鏡圖

Fig.8 (color online) shows the CV curves of Au electrode and Au/MAE obtained before and after adding 2 mM glucose to 0.1 mM PBS buffer solution (pH 7.0).Before glucose was added (Fig.8(a)),Au electrode had basically no current response.In contrast,Au/MAE had an obvious REDOX peak.After 2 mM glucose was added (Fig.8 (b)),REDOX peaks appeared at the 0?1.4 V scanning potential for both Au electrode and Au/MAE.The peak pair appearing at the voltages of 0.97 V and 0.44 V was the REDOX peak of Au nanoparticles.Compared with Au electrode,Au/MAE showed good electrochemical response.The obtained oxidation peak current was about 6 times that of Au electrode,indicating that Au/MAE had a higher electrochemically-active specific surface area.

Fig.8 Comparison of electrochemical performance between Au electrode (a) and Au/MAE (b)圖8 Au 電極（a）與Au/MAE 電極（b）電化學性能比較

4.2 Glucose detection by Au/MAE

Compared with Au electrode,the optimized Au/MAE showed obvious current response with or without the presence of glucose.This confirmed that the prepared electrode had good electrochemical performance and played an important role in the electrochemical oxidation of glucose.Therefore,this electrode can be applied to the research of glucose sensing and the further electrochemical detection of glucose.The effect of Au/MAE on the oxidation of glucose (concentration:1 mM) at different scanning rates (10?1 000 mV/s) was studied to obtain the results as shown in Fig.9(a) (color online).The REDOX peak current increases with the scanning speed,the anodic peak moves slightly from 0.95 V to 1.04 V,and the cathodic peak moves from 0.48 V to 0.39 V.Therefore,the electrode is highly sensitive to glucose oxidation.In Fig.9(b) (color online),the obtained anodic and cathodic peak currents versus scanning speed were plotted and fitted.It can be seen from the figure that with the increase of scanning rate from 10 mV/s to 1 000 mV/s,the absolute value of REDOX peak current will increase linearly,and the linear fitting regression coefficients of cathodic and anodic peaks will be 0.997 and 0.994 respectively (ifR2is closer to 1,the fitting will be better).According to the electrochemical theory,the peak current is directly proportional to the scanning speed in the electrode reaction process controlled by surface adsorption.The ratio of anodic peak current to cathodic peak current is close to 1.These typical characteristics indicate that the electron transfer process on Au/MAE electrode is controlled by surface adsorption.

Fig.9 For Au/MAE (glucose concentration:1 mM) in 0.1 mM PBS (pH 7.0) buffer solution,(a) CV diagram at different scanning rates (10?1 000 mV/s);(b) fitting diagram of anodic and cathodic peak currents at different scanning rates圖9 Au/MAE 在含有1 mM 葡萄糖的0.1 mM PBS(pH 7.0)緩沖溶液中,(a)不同掃描速度下 (10?1 000 mV/s) 的CV 圖以及(b)不同掃描速率下的陽極和陰極峰值電流擬合圖

The above experiments verify the ability of Au/MAE to sense glucose,so the prepared Au/MAE can be used as a non-enzymatic glucose-sensing electrode.The electrochemical response of Au/MAE to different glucose concentrations was studied when the scanning rate was 100 mV/s.Fig.10(a)(color online) shows the linear voltammetry curves of Au/MAE in PBS (0.1 mM,pH 7.0) at different glucose concentrations.It can be seen that the Au/MAE oxidation/reduction peak currents increase with the increase of glucose concentration from 1 mM to 8 mM.The corresponding standard curve is shown in Fig.10(b).It can be seen that within the whole concentration range,the response of Au/MAE to glucose concentration satisfies a linear relationship.To determine the sensitivity and response time of Au/MAE as a glucose sensor,we measured the amperometric response of Au/MAE to the continuous addition of glucose to 0.1 mM PBS while keeping the potential at +0.5 V.It can be seen that the electrode has a fast current response to glucose and can reach a stable current density within 5 s.With the addition of the glucose with different concentrations,the baseline of the curve will drift slightly (as shown in Fig.10(c)),because the consumption rate of glucose on the electrode surface is faster than its diffusion rate,or because the midbody is adsorbed to the active site.The fitting curve of the glucose sensor is shown in Fig.10(d),indicating that Au/MAE peak current has a linear relationship with the glucose concentration in the range of 0.01 mM?1 mM (R2=0.986).The sensitivity calculated according to the standard curve is 101 μA·cm?2·mM?1.This proves that Au/MAE has excellent glucose-sensing performance.

Fig.10 (a) Cyclic voltammograms of Au/MAE electrode at different glucose concentrations in 0.1 mM PBS (pH 7.0) (scanning rate:100 mV/s) and (b) the corresponding calibration curve;(c) amperometric response of Au/MAE electrode to the continuous addition of glucose to 0.1 mM PBS at the voltage of 0.5 V and (d) the corresponding fitting curve圖10 0.1 mM PBS (pH 7.0) 條件下，(a) 葡萄糖濃度不同時Au/MAE 的循環(huán)伏安圖（掃描速率100 mV/s）及(b) 相應的校準曲線；(c) Au/MAE 在0.5 V 電壓下，0.1 mM PBS 中連續(xù)加入葡萄糖時的安培響應；(d) 對應的擬合曲線

The electrochemical catalytic behavior of Au/MAE on glucose is provided by Au nanoparticles,and the process of generating current response marks the formation of active gold oxide[20].This can be explained by the following equations:

Glucose reacts with the active gold oxide formed in Equation (1) and then is oxidized into glucolactone.

4.3 Anti-interference test of Au/MAE

By testing the anti-interference of Au/MAE against common substances in serum,such as urea and ascorbic acid,we can determine whether the prepared electrode can be used for the practical detection of glucose in serum.By using the chronoamperometry,we studied the current response of the electrode to glucose and other anti-interference substances at 0.5 V voltage,as shown in Fig.11(a).As can be seen from the figure,when 1.5 mM glucose is added,the current will respond rapidly and produce an obvious current signal (close to 0.1 mA).After the interfering substances (uric acid,ascorbic acid,lactose and NaCl) were added in sequence over time,the change of current signal was very small and almost negligible.With further addition of 6 mM glucose,the current signal changed significantly again (up to 0.5 mA).It can thus be seen that the current signal of the electrode is greatly influenced by glucose concentration and has nothing to do with other substances.This may be because Au/MAE surface has a good adsorption effect on glucose molecules,and the electrooxidation effect of glucose molecules is higher than that of lactose,ascorbic acid and other coexisting substances.The Fig.11(b) shows the percentage of the signals of other interfering substances compared to glucose.With glucose signal as reference,the signal percentage of other substances is as small as less than 8%.Obviously,the common substances in blood will not significantly affect the detection of glucose by the electrode,indicating that the electrode has a good selectivity for the amperometric detection of glucose.Therefore,it is proved that Au/MAE bioelectrode has strong anti-interference ability and can be used for the detection of glucose in actual serum.

Fig.11 (a) Amperometric response of the electrode to the continuous addition of 1.5 mM glucose,1 mM urea,1 mM AA,1 mM lactose,1 mM NaCl and 6 mM glucose to PBS (0.1 mM,pH 7.0) buffer solution at 0.5 V voltage;(b) the percentage of interfering signals compared with the target analyte圖11 (a) 0.5 V 電壓下，在PBS（濃度0.1 mM，pH 7.0）緩沖溶液中連續(xù)添加1.5 mM 葡萄糖、1 mM Urea、1 mM AA、1 mM 乳糖、1 mM NaCl 和6 mM 葡萄糖時，電極的安培響應；(b) 與目標分析物相比，相應的干擾信號的百分比

5 Conclusion

Any periodic or aperiodic Au/MAE can be successfully prepared on ITO substrate by combining DMD photolithography with electrochemical deposition.The cyclic voltammetry test shows that the REDOX performance of Au/MAE is related to the area of each structural cell.When the total surface area remains unchanged,the REDOX peak current of a periodic or aperiodic Au/MAE will first increase and then decrease with the increase of cell area,independent of the cell shape.The Au/MAE with dense cell arrangement was prepared by optimizing the structural parameters of the electrode,and then the electrode performance in glucose detection was evaluated.The experimental results show that the electrode has strong catalytic activity and interference resistance in the detection of glucose,with a sensitivity of 101 μA·cm?2·mM?1.It can be used as a non-enzymatic glucose sensor.With the advantages of simple and rapid operation,low cost and high repeatability,this MAE-preparing method can be widely used.

——中文對照版——

1 引言

陣列電化學生物傳感器 (Electrochemical Biosensor Array,EBA) 憑借其高選擇性、高靈敏度、分析速度快等優(yōu)點而被廣泛應用在食品檢測、環(huán)境監(jiān)測、醫(yī)療衛(wèi)生等領域[1-4]。作為信號轉(zhuǎn)換器的微陣列電極 (Microarray electrodes,MAE)是EBA的重要組成部分，其性能將直接影響EBA 的檢測性能[5-8]。而微陣列電極的制作技術是陣列電極研究的基礎和關鍵[9-12]。微陣列電極的制作技術包括以下幾種：基于掩模板的制備工藝因為需要依據(jù)不同的微電極結(jié)構(gòu)制備大量的掩模板，若要實現(xiàn)對金微陣列結(jié)構(gòu)參數(shù)的優(yōu)化設計，掩模板的設計和制備需耗費大量時間，因此成本高、靈活性差[13]；而基于刻蝕技術的光刻法，在制備平面有序陣列電極方面具有可控性好、精密度高、電極表面平整度好、批間重現(xiàn)性高等優(yōu)點，是目前制備微陣列電極的首選方法[14]；飛秒激光直寫技術雖然可以實現(xiàn)高分辨率圖形的微電極陣列的制備，但是無法實現(xiàn)大面積的加工[15]；基于數(shù)字微鏡器件(Digital-Micromirror-Device,DMD)的無掩模光刻技術中，DMD 的微反射鏡數(shù)目多達200 萬個，并且每一個微反射鏡都可以等效為一個獨立的光源，因而其曝光過程相當于多光束逐點曝光，同時DMD 微反射鏡的開關頻率與樣品臺的運行速度相匹配，可實現(xiàn)圖形的滾動曝光和任意周期與結(jié)構(gòu)陣列的制備。與飛秒激光直寫系統(tǒng)相比，大大提高了系統(tǒng)的生產(chǎn)效率，降低了制作成本。

利用本課題組自行搭建的DMD 無掩模多步光刻系統(tǒng)，結(jié)合電化學沉積技術，制備了用戶可自定義的微結(jié)構(gòu)陣列[16-19]。氧化銦錫(ITO)導電玻璃具有較低的電阻率、高的導電性能以及加工性能好等優(yōu)異特性，被廣泛應用于電化學領域。因此，以ITO 導電玻璃作為基底，利用復合加工技術（包括DMD 的靈活的圖案生成，多步光刻技術和劑量調(diào)制技術），通過設計掩模圖案并控制光刻工藝參數(shù)（曝光時間與曝光能量），制備了各種形狀、任意周期以及非周期的微結(jié)構(gòu)陣列(MAE)。再結(jié)合電化學沉積技術 (electrochemical deposition technique)，將Au 均勻沉積在所制備的微結(jié)構(gòu)陣列模板上，以獲得任意形狀、周期以及非周期的金微電極陣列 (Au/MAE) 。比較分析了不同結(jié)構(gòu)微電極陣列的電化學性能，系統(tǒng)研究了形狀、周期以及表面積等參數(shù)對氧化還原峰電流的影響，得知，在總表面積不變，單元結(jié)構(gòu)表面積為1.7×10?5cm2時制備的Au/MAE 結(jié)構(gòu)具有較高的氧化還原峰電流，接著，優(yōu)化了Au/MAE 的結(jié)構(gòu)參數(shù)，并將相鄰單元結(jié)構(gòu)間的距離降低為最小，制備了密集排列的可用于葡萄糖檢測的Au/MAE。通過該Au/MAE 對葡萄糖的檢測性能和抗干擾性能進行測試，得到的靈敏度為101 μA·cm?2·mM?1，驗證了這種陣列電極制備方法的可行性。

2 Au/MAE 的制備原理和方法

2.1 DMD 投影光刻系統(tǒng)與電化學沉積實驗裝置

圖1 (a) 所示是基于DMD 的無掩模投影光刻系統(tǒng)示意圖，所用DMD (Texas Instruments) 由1 024×768 個微鏡陣列構(gòu)成，微鏡單元的邊長為一個像素的大小，即1.368 μm。壓電平臺 (Piezoelectric stage,PZS) 可在水平面內(nèi)沿X、Y方向發(fā)生微移動，運動范圍為100 μm×100 μm，位移精度可達7 nm。中心波長為385 nm 的LED 光源經(jīng)過擴束與準直后，與光學系統(tǒng)的光軸成一定角度斜入射到DMD 上，預先設計好的圖案由計算機通過相應軟件輸入到DMD 芯片中，被DMD 反射后的光束攜帶有圖像信息，經(jīng)縮小倍率為10 的投影物鏡最后聚焦在基片上 (ITO)，該基片上旋涂有正性光刻膠 (S1318) 并放置于PZS 上。計算機同步控制DMD 加載數(shù)字掩模板與PZS 的高精度運動來實現(xiàn)多步光刻。

圖1 (b) 所示為電化學沉積實驗裝置示意圖，主要包括三電極系統(tǒng)（工作電極、參比電極、輔助電極）、電解槽、電解液、恒電位儀等。本實驗中，將Ag/AgCl 作為參比電極，鉑絲作為輔助電極，將刻有微結(jié)構(gòu)陣列的ITO 基板作為工作電極。用HAuCl4溶液和KCl 溶液 (體積比4∶5) 作為電解液，在一定的沉積電壓和電鍍時間下電沉積Au 納米粒子，制備Au 導電層。

2.2 Au/MAE 的制備流程

將多步光刻 (DMSL)和電化學沉積技術相結(jié)合，以ITO 為襯底，制備了Au/MAE，具體實驗過程如圖2（彩圖見期刊電子版）所示。圖2 (a) 所示為基片的預處理過程，即分別用丙酮、無水乙醇、蒸餾水超聲清洗ITO 玻璃基板，然后在氮氣流中干燥。圖2(b) 所示為光刻膠的旋涂與前烘過程，即用移液槍將正性光刻膠均勻涂覆于基板上后，再以一定的轉(zhuǎn)速旋涂于ITO 玻璃上，可得到一定厚度的光刻膠層，接著在95℃的熱板上烘烤10 min，以去除殘留溶劑，提高光刻膠的靈敏度并使其固化。圖2(c) 所示為基于DMD 的無掩模光刻和顯影過程，將預先設計好的掩模圖形通過DMD 光刻系統(tǒng)對光刻膠曝光后，置于5‰的Na-OH 溶液中顯影一定時間，再去除被曝光的光刻膠。圖2(d) 所示為Au 納米粒子的電化學沉積過程，以0.1 M 的HAuCl4·4H2O和0.1 M 的KCl 溶液 (HAuCl4∶KCl 體積比約1∶1) 作為電解液，在一定電壓下沉積一定時間后，分別用丙酮和蒸餾水清洗去除未曝光的光刻膠，就可獲得如圖2(e) 所示圖形的微陣列電極結(jié)構(gòu)。

2.3 Au/MAE 的制備

采用DMSL（曝光時間為3 s，曝光能量為145 mJ/cm2）制備了一系列的MAE，并結(jié)合電化學沉積金納米導電層，獲得了Au/MAE，在奧林巴斯顯微鏡下分別觀察了MAE 和Au/MAE 的圖像。在總表面積（均為3.7×10?3cm2）和單元結(jié)構(gòu)表面積（1.7×10?5cm2）相同的條件下，制備了如圖3 (a)、3(b)、3(c) 所示的周期性排列的圓形和六邊形以及三角形微陣列結(jié)構(gòu)；圖3(d)、3(e)、3(f) 分別為相應的Au/MAE 的顯微鏡圖像。從圖3 可見，MAE 和Au/MAE 形狀規(guī)則、邊緣清晰、刻線流暢、形貌良好。同理，在上述總表面積和單元結(jié)構(gòu)表面積不變的情況下，又制備了非周期排列的橢圓形和六邊形以及五角星結(jié)構(gòu)的Au/MAE，如圖4所示。利用劑量調(diào)制與DMSL 相結(jié)合的復合技術，可以實現(xiàn)用戶自定義微結(jié)構(gòu)陣列電極的制備。

3 Au/MAE 結(jié)構(gòu)參數(shù)的優(yōu)化

由于所制備的MAE 單元結(jié)構(gòu)的高度與其表面寬度相比很小，可忽略不計，因而探究了微結(jié)構(gòu)單元面積和排布密集度對電極性能的影響。利用循環(huán)伏安法，在掃描速率為100 mV/s 的條件下，電位范圍從0 至1.4 V，研究了不同形狀單元結(jié)構(gòu)的Au/MAE 在PBS (pH 7.0) 緩沖溶液中的電化學響應。

首先，保持總表面積不變（3.7×10?3cm2），研究了不同形狀的Au/MAE 在結(jié)構(gòu)單元表面積分別 為5.9×10?6cm2、8.3×10?6cm2和1.7×10?5cm2以及3.3×10?5cm2時的電化學性能。圖5(a)、5(b)、5(c) 分別顯示了周期性排列的圓形、六邊形和三角形3 種結(jié)構(gòu)的循環(huán)伏安曲線 (Cyclic Voltammogram，CV) 。由圖5（彩圖見期刊電子版）可知，同種單元形狀的Au/MAE，當總表面積不變時，單元結(jié)構(gòu)面積不同的Au/MAE 顯示出不同的氧化還原峰電流。對比3 種結(jié)構(gòu)發(fā)現(xiàn)，盡管單元結(jié)構(gòu)形狀不同，但循環(huán)伏安電流的變化規(guī)律相同，都隨著單元結(jié)構(gòu)面積的增加表現(xiàn)出先增大后減小的趨勢。這是因為氧化還原峰電流的大小與氧化還原反應的接觸位點的數(shù)量及微陣列結(jié)構(gòu)對信號傳輸?shù)膹娙趺芮邢嚓P。當單元結(jié)構(gòu)表面積從5.9×10?6cm2增加到3.3×10?5cm2，而總表面積不變時，單元結(jié)構(gòu)的數(shù)目將會變少，因而接觸位點會減少，氧化還原峰值會減弱。另一方面，單元結(jié)構(gòu)數(shù)目變少會使金微陣列結(jié)構(gòu)的電阻變小，氧化還原峰電流增大。在這兩種因素的競爭作用下，峰電流會表現(xiàn)出先增大后減小的趨勢，而不是線性變化。當單元面積為1.7×10?5cm2時，氧化和還原峰電流均最高。隨后，保持總表面積相同，按上述實驗參數(shù)制備了隨機排列的不同形狀的Au/MAE，并且測試了每種結(jié)構(gòu)Au/MAE 的電化學性能，如圖5(d)、5(e)、5(f) 分別顯示了非周期性排列的橢圓形、六邊形以及五角星結(jié)構(gòu)的循環(huán)伏安曲線?？梢妼τ谕环N非周期結(jié)構(gòu)，當總表面積相同時，隨著單元結(jié)構(gòu)表面積的增大，氧化還原峰電流亦呈現(xiàn)先增大后減小的趨勢。上述分析說明了對于隨機排列的Au/MAE，影響其電化學性能的主要因素也是總面積和單元面積，這同上面得到的實驗結(jié)論是一致的。

從圖5 還可得到，當總表面積相同、單元面積也相同時，不同單元形狀周期排列和非周期排列的Au/MAE 的循環(huán)伏安電流結(jié)果基本相同。綜上所述，可以得出Au/MAE 的氧化還原峰電流值與總表面積和單元結(jié)構(gòu)面積密切相關，與單元結(jié)構(gòu)形狀無關的結(jié)論。

圖6 (a) 給出了總表面積一定時，氧化峰電流值與電極單元結(jié)構(gòu)表面積的關系，從圖中可以看出，隨著單元結(jié)構(gòu)表面積的增大，氧化峰呈現(xiàn)先增大后減小的趨勢。且圓形、方形和六邊形3 種結(jié)構(gòu)的圖線基本重合，這說明當總表面積和單元結(jié)構(gòu)表面積保持一定時，氧化峰電流與單元形狀無關。圖6 (b) 給出了不同單元結(jié)構(gòu)表面積的方形結(jié)構(gòu)Au/MAE 的氧化峰電流隨總表面積的變化關系，可見當單元結(jié)構(gòu)表面積一定時，隨著總表面積的增大，氧化峰電流值呈線性增長。

通過上述分析，可見影響Au/MAE 電化學性能的主要因素是總面積和單元面積，與單元形狀無關。以此為依據(jù)，優(yōu)化了Au/MAE 的結(jié)構(gòu)參數(shù)，即總表面積一定的情況下(3.7×10?3cm2)，選擇使氧化還原峰值電流最優(yōu)的單元面積(1.7×10?5cm2)，并通過使相鄰兩個單元結(jié)構(gòu)的間距最小，制備了方形結(jié)構(gòu)且密集排列的Au/MAE，并測試了該電極相應的性能。

4 Au/MAE 性能的評價

4.1 Au 電極與Au/MAE 的電化學性能

根據(jù)上面得到的最優(yōu)化參數(shù)，制備了方形單元結(jié)構(gòu)且密集排列的Au/MAE，單元表面積為1.7×10?5cm2，相鄰單元結(jié)構(gòu)的實際間距為1.26 μm。圖7 (a)(彩圖見期刊電子版)和7(b)(彩圖見期刊電子版) 分別為密集排列的MAE 與Au/MAE 的光學顯微鏡圖像，相鄰單元結(jié)構(gòu)的間距約為1.26 μm，當將該間距降為1 個像素(1.368 μm) 時，由于電沉積時金納米粒子會向外延展，導致間距變小無法區(qū)分，因此當設計圖案結(jié)構(gòu)間距為2 個像素時，制備的方形單元結(jié)構(gòu)微陣列最為密集。圖7(c)(彩圖見期刊電子版)和7(d)分別是全曝光和單個Au 電極的光學顯微鏡圖像。利用循環(huán)伏安法，對比研究了單個Au 電極與Au/MAE 的電化學性能。

Au 電極和Au/MAE 在0.1 mM 的PBS 緩沖溶液（pH 7.0）中添加2 mM 葡萄糖前后的循環(huán)伏安曲線如圖8(彩圖見期刊電子版)所示，添加之前(圖8(a))，Au 電極基本無電流響應，相較之下，Au/MAE 具有明顯的氧化還原峰。添加2 mM 葡萄糖后(圖8 (b))，在掃描電位0?1.4 V 之間均產(chǎn)生了氧化還原峰，電壓為0.97 V 和0.44 V 出現(xiàn)的一對峰是金納米粒子的氧化還原峰。和Au 電極相比，Au/MAE 表現(xiàn)出良好的電化學響應，觀察到的氧化峰電流是Au 電極的6 倍左右，表明Au/MAE 電活性比表面積更高。

4.2 Au/MAE 對葡萄糖的檢測

前文得到經(jīng)優(yōu)化后的Au/MAE 相較Au 電極而言，在有無葡萄糖時均表現(xiàn)出很明顯的電流響應，證實了制備的電極具有良好的電化學性能，在葡萄糖電化學氧化過程中起到了重要作用。因此該電極可應用于葡萄糖傳感研究，進一步對葡萄糖進行相關的電化學檢測。本文研究了Au/MAE在不同掃描速率（10～1 000 mV/s）時對葡萄糖（濃度1 mM）氧化的影響，結(jié)果如圖9(a)(彩圖見期刊電子版)所示?？梢?，氧化還原峰電流值隨著掃描速度的增大而增大，陽極峰值從0.95 V 小幅移動到1.04 V，陰極峰值從0.48 V 移動到0.39 V，表明電極對葡萄糖氧化具有高靈敏度。圖9(b)(彩圖見期刊電子版)將其和陰極峰值電流對掃描速度作圖并進行擬合。由圖可知，隨著掃描速率從10 mV/s 增加到1 000 mV/s，氧化還原峰電流的絕對值均線性增大，陰極峰和陽極峰的線性擬合回歸系數(shù)分別為0.997 和0.994（R2越接近1，表明擬合情況越好）。根據(jù)電化學理論知，在表面吸附控制電極反應過程中，峰電流與掃描速度成正比。陽極峰電流與陰極峰電流比值接近于1，這些典型的特征表明Au/MAE 電極表面的電子傳遞過程屬于表面吸附控制過程。

上述實驗驗證了Au/MAE 對葡萄糖的傳感能力，因而制備的Au/MAE 可作為非酶葡萄糖傳感電極，在掃描速率為100 mV/s 時，研究了Au/MAE 對不同濃度葡萄糖的電化學響應。圖10(a)(彩圖見期刊電子版)顯示了不同葡萄糖濃度下Au/MAE 在PBS(0.1 mM，pH 7.0)中的線性伏安曲線?？梢娫谄咸烟菨舛葟? 到8 mM 逐漸增大的過程中，Au/MAE 的氧化和還原峰的電流均隨著葡萄糖濃度的增加而增大。相應的標準曲線如圖10(b) 所示，可見在整個濃度范圍內(nèi)，Au/MAE 對葡萄糖濃度的響應滿足線性關系。為了確定作為葡萄糖傳感器的靈敏度和響應時間，通過保持電位為+0.5 V，測量了Au/MAE 在0.1 mM PBS 中連續(xù)添加葡萄糖時的安培響應?？梢娫撾姌O對葡萄糖有快速的電流響應，可以在5 s 內(nèi)達到穩(wěn)定的電流密度。隨著不同濃度葡萄糖的加入，如圖10(c)，曲線的基線略有漂移，這是因為電極表面的葡萄糖消耗速率快于其擴散速率，或者是因為中間體吸附在活性位點上。葡萄糖傳感器的擬合曲線如圖10(d) 所示，其在0.01 mM 至1 mM 范圍內(nèi)與葡萄糖濃度呈線性關系 (R2=0.986)，根據(jù)標準曲線計算可得靈敏度為101 μA·cm?2·mM?1。這證明了Au/MAE 對葡萄糖有著優(yōu)良的傳感性能。

Au/MAE 對葡萄糖的電化學催化行為由金納米粒子提供，產(chǎn)生電流響應的過程標志著金活性氧化物的形成[20]?？梢酝ㄟ^下面的方程來解釋。

葡萄糖與 (1) 反應式中形成的金活性氧化物反應，將葡萄糖氧化為葡萄糖內(nèi)酯。

4.3 Au/MAE 抗干擾測試

通過檢測Au/MAE 對血清中常見物質(zhì)如尿素、抗壞血酸等碳水化合物的抗干擾能力，可以判斷制備的電極是否可以用于血清中葡萄糖的實際檢測。通過采用計時電流法，研究了在0.5 V電壓下電極對葡萄糖等抗干擾物質(zhì)的電流響應，如圖11 (a) 所示。由圖可知，當加入1.5 mM 葡萄糖時，電流反應迅速，產(chǎn)生了明顯的電流信號（接近0.1mA）。隨時間變化依次添加尿酸、抗壞血酸、乳糖和NaCl 這些干擾物質(zhì)之后，電流信號的變化非常微小，幾乎可以忽略不計，繼續(xù)加入6 mM 的葡萄糖，電流信號再一次發(fā)生明顯變化（達到0.5 mA），可見該電極的電流信號受葡萄糖濃度影響較大，而與其他物質(zhì)無關。這可能是由于Au/MAE 表面對葡萄糖分子有較好的吸附作用，與乳糖、抗壞血酸等共存物質(zhì)相比，葡萄糖分子的電氧化作用相對較高。圖11 (b) 給出了相較于葡萄糖的其他干擾物質(zhì)信號的百分比，以葡萄糖信號作為參考，其他物質(zhì)信號百分比小于8%，占比很小，顯然血液中常見的物質(zhì)并不會明顯影響電極對葡萄糖的檢測，表明該電極對葡萄糖的安培檢測具有良好的選擇性。因此，證明了Au/MAE 生物電極具有很強的抗干擾能力，可用于實際血清中葡萄糖的檢測。

5 結(jié)論

通過DMD 光刻技術和電化學沉積技術相結(jié)合的方法成功地在ITO 襯底上制備了任意周期或非周期的Au/MAE。循環(huán)伏安法測試表明，Au/MAE 的氧化還原性能與結(jié)構(gòu)單元面積有關，在總表面積相同的情況下，無論是周期性結(jié)構(gòu)還是非周期性結(jié)構(gòu)，隨著結(jié)構(gòu)單元面積的增大，氧化還原峰電流表現(xiàn)出先增大后減小的趨勢，而與單元結(jié)構(gòu)形狀無關。通過優(yōu)化微陣列電極的結(jié)構(gòu)參數(shù)，制備了密集排列的Au/MAE，并評價了該電極對葡萄糖的檢測性能。實驗結(jié)果表明：該電極在對葡萄糖檢測時表現(xiàn)出很強的催化活性，且具有較強的抗干擾能力，靈敏度為101 μA·cm?2·mM?1，可以作為一種非酶葡萄糖傳感器。這種制備微陣列電極的方法，操作簡單、快速且成本低，重復性高，可被廣泛使用。