Research On The Control System Of Adhesive-Stamp-based Transfer Technique For Micro-LEDs

In recent years,micro light-emitting diode(Micro-LED)has attracted a lot of attentions for its vast potential applications in various fields,such as micro-display,visible light communication,biomedical and wearable devices.Compared to LCD liquid crystal display(LCD)and organic light-emitting diode(OLED)display,Micro-LED display is advantageous for some unique features,including excellent light properties(e.g.,contrast,hue and brightness),fast response time,high power efficiency and long lifetime.However,the mass transfer process has become the bottleneck of the commercialization of the Micro-LED display,hindering the mass production of Micro-LEDs.The mass transfer of Micro-LED can be defined as a process,where large-scale micron-grade LED chips(tens or even hundreds of thousands)must be accurately and efficiently transferred from the donor substrate onto the receiving substrate at a time.The current critical technical difficulties for mass transfer lie in micron-grade precision,high yield,high efficiency and low cost.Adhesive stamp mass transfer can be described as a process where a stamp picks up the chips from the donor substrate and prints the chips into the receiving substrate.The adhesive stamp mass transfer printing process can be achieved by utilizing the viscoelastic rate-dependent adhesion at the stamp-ink interface to enable either retrieval or printing by controlling the separation velocity.Notably,the adhesive stamp mass transfer shows good reliability,high throughput,success probability,scalability of stamps and low cost,making it one of the mainstream transfer techniques.The complete transfer printing consists of the pick-up stage and the print stage.Most researches focus on the optimization of print stage.Whereas,the parameters that influence the results in the pick-up stage are rarely studied.In the paper,the control system of adhesive stamp mass-transfer of Micro-LEDs,including the hardware and the software,is established and a scheme for optimization on adhesive stamp mass-transfer of Micro-LEDs with support vector machine(SVM)model is proposed to classify and predict the feature signals and the corresponding pick-up results in the pick-up stage.The control system can achieve micron-grade location precision and the classification predictive accuracy of the SVM model is high.This paper is organized as follows:1.The influence of mass-transfer technology on the industrialization and commercialization of Micro-LED display is mainly introduced.A variety of mass-transfer techniques to assemble Micro-LEDs,such as electrostatic mass transfer,laser-assisted mass transfer,roll-based mass transfer,fluidic self-assembly and the adhesive stamp mass transfer adopted in this work,are analyzed and their mechanisms,advantages and disadvantages and limitations of application are commented.2.Given the stability of the system based on these factors,such as stable platform,device and positioning precision,real-time feedback,micro-force control,I/O operation and man-machine interaction,the control system of adhesive stamp mass-transfer of Micro-LEDs,including the hardware and the software,is established.Develop the reliable operation control software based on Delphi 7 and the hardware system consists of motion unit,machine vision unit and pressure sensor unit.The positioning resolution of the motion unit based on a XY-stage,a Zstage,an orientation and a motion controller is 0.625μm.An optical camera,a point light source,a constant current controller and telecentric lens make up of the machine vision unit.The precision of the pressure sensor unit based on pressure sensors and display instruments is 1mg.The operation control software is composed of three modules:motion control module,image processing module and pressure force acquisition module.The alignment between the stamp and Micro-LED chips and the picking up and the releasing of Micro-LED chips can be accurately accomplished with the collaboration between the hardware devices and the operation control software,achieving the high precision and automatic transfer of the whole system.The exact process of the picking up of the Micro-LED chips is introducted.3.A scheme for optimization on adhesive stamp mass-transfer of Micro-LEDs with SVM model is proposed to classify and predict the feature signals and the corresponding pick-up results in the pick-up stage.In the pick-up stage,the two feature signals,namely the separation speed and the force between stamp and donor substrate,are two critical parameters that determine the success or failure in picking up chips from the donor substrate.The pick-up experiments have been performed repeatedly for hundreds of times from which the separation speed and the force between the stamp and the donor substrate are extracted by the motion controller and the pressure sensor respectively.Random sets of vectors are employed as the training sets for constructing the SVM model,the other sets of vectors as the testing sets.Then the particle swarm optimization(PSO)algorithm was used to improve the classification,optimizing the cost parameter C and the Gaussian kernel function parameter γ,which are 27 and 1 respectively.The testing data set is brought into the SVM model and the resulting predictive accuracy is 85%,which implies that the transfer process can be optimized by adjusting the transfer parameters of the model and that both transfer yield and efficiency can be improved with the model.