Design of orientation stages for step and flash imprint lithography

This paper presents the design of orientation stages for high-resolution imprint lithography machines. These machines implement a new lithography process known as Step and Flash Imprint Lithography (SFIL) and are intended for 1) sub 100 nm imprint demonstrations on flat substrates and 2) investigation of potential defect propagation during step and repeat imprinting. SFIL is an imprint lithography process that is a combination of chemical and mechanical steps and its implementation at room temperature and low pressure makes it an attractive process as compared to other imprint techniques. A critical component of an imprint machine is the orientation stage that is required to provide uniform intimate contact between the template and substrate surfaces. The orientation stage requirements are distinct from those used in photolithography since the depth of focus of projection optics allows for larger errors in the orientation alignment. Also, due to contact between the template and substrate surfaces in imprint lithography, the separation kinematics must be carefully controlled in the SFIL process. Two different orientation stages are designed for single- and multi-imprint machines. In order to eliminate the particle contamination due to frictional contacts, all joints are made with flexure joints. Imprint experiments have been performed to demonstrate sub 100 nm imprints.     

下一代光刻技术——压印光刻

通过分析集成电路制造工艺中的核心环节--光刻技术,采用机械微复形原理的压印光刻技术可避免传统光学光刻的光学衍射限制,能够对最小到6 nm特征尺寸的图形进行复制.通过研究压印光刻原理,详细分析对比热压印及常温压印的工艺特点、复形面积、模具结构和阻蚀胶固化过程等,揭示出常温压印更适用于多层套刻的图形制作.针对基于紫外光固化的常温压印光刻工艺在压印过程中的关键技术:阻蚀胶成膜控制、对准、套刻、精确加载、留膜厚度控制、阻蚀胶固化控制等进行深入研究,提出释放保形软压印工艺,能够实现基于常温软模具压印的大面积亚100 nm级特征尺寸图形的复制.最后,以SIL-Ⅰ型分步式常温紫外光固化压印光刻机的研究为例,对其结构特性做出深入分析,估算出系统的精度等级;通过试验,对系统的压印复形精度做出评估.     

步进扫描光刻机硅片台连续扫描时间优化算法

为减少步进扫描光刻机连续曝光扫描运动的扫描过渡时间(扫描时间和步进时间),将硅片台扫描方向曝光扫描运动分解为逻辑步进运动和逻辑扫描运动,提出扫描和步进运动时间重叠的轨迹规划算法(时间重叠算法)处理连续曝光扫描运动。在不破坏运动约束条件下(恒速运动要求以及速度、加速度边界限制),根据时间重叠算法重新计算并推导不同扫描运动路径下连续扫描运动过程的步进和扫描运动转折点。经理论分析和硅片曝光场连续曝光实例计算表明：基于时间重叠算法的轨迹规划可获得时间优化的硅片台连续曝光扫描,比传统扫描方法减少曝光扫描运动过程中的无效率时间,提高步进扫描光刻机生产率,为步进扫描光刻机工程实践提供理论基础。     

光成型技术及其发展概况

介绍了光成型技术的成型方法和现阶段的研究概况，重点介绍了提高加工精度和加工效率的技术手段，并重要介绍了微光成型技术的发展和有待研究解决的技术。     

Flow behavior control in immersion lithography

Over the past decade, immersion lithography has been the primary technology for exposure process in semiconductor manufacturing. Compared with traditional dry lithography method, this technology improves the exposure resolution greatly by inserting a high index liquid into the gap between lens and wafer surface. Keeping the immersion liquid pure and uniform as well as avoiding residual droplets during high speed scanning motion are two challenges faced by the development of immersion lithography. Contaminations, particles, bubbles, heating and stress in the liquid will destroy the continuity of the refractive index. High speed motion of the wafer during scanning may break the meniscus stability on the interface between the liquid and the gas surrounding it, and then generate residual droplets on the wafer. All above phenomena will affect the exposure performance of immersion lithography and corresponding flow behavior control methods are required to solve the problems. This paper gives a review of the studies on flow behavior control in immersion lithography, which contains two parts: studies for liquid purity and uniformity and studies for meniscus stability. In each part, the mechanism and character of the flow behaviors as well as their effects to exposure performance are presented firstly. Then, control approaches adopted by now, including material and surface modifications as well as immersion head utilization, are introduced. In addition, the challenges faced in future studies are also pointed out. The purpose of this review paper is to help the researchers to understand the flow control problems in immersion lithography and to provide better control approaches for exposure performance improvement.     

考虑二元Copula统计量的晶圆制造叠加误差监测

目前,大多数晶圆制造研究集中在基于离散数据的缺陷模式识别上,而芯片的光刻制造是连续叠加过程,因此基于连续数据的晶片重叠误差监测具有挑战性和必要性。在数据监测过程中充分考虑数据的可解释性,同时结合晶圆数据特性及其物理意义加入新的惩罚项,改进LTS-SPCA降维模型,提出了灵活度较高的稳健稀疏主成分分析技术;然后基于Copula的置换对称、反射对称两种性质,考虑晶圆的几何特征,建立了最佳多元耦合统计量,用于监测晶圆制造的叠加过程异常。所提方法监测异常数据的准确率可达91.75%,具有较高的工程应用价值。     

Finite element analysis on global and local estimation for thermo-mechanical response of EPL wafer heating

In this paper, an approach is presented for the computation of the in-plane pattern placement error (PPE) caused by the wafer heating during exposure of electron projection lithography (EPL) using the finite element method (FEM), which is one of candidates in the next-generation lithography (NGL) exposure tools. The PPE is the global and local distortion, and is the thermo-mechanical response due to the thermal deformation of the wafer in the lithography process. The prediction of PPE requires high accuracy for NGL exposure tools. The simultaneous estimation of the global and local PPE to the whole wafer of a full three-dimensional FE model using a solid element requires excessive computation time. A novel technique of numerical simulation is developed and proposed, which is the employment of a shell element combined with previously proposed the dynamic meshing technique (DMT), being possible to predict PPE in realistic computation time with high accuracy. Simulations are performed for the wafer heating of EPL effectively using three techniques, that is the equivalent average heating technique and two proposed techniques. The simulation results agree closely with the result of a full three-dimensional FE analysis, and the required computation time becomes 1/16 or much less of that.     

Imprinting a needle array on a polycarbonate substrate

A micro-needle array was fabricated on a polycarbonate (PC) substrate using an electroformed-Ni mold with a conical concave pattern. The diameter and length of each needle were 50 μm and 135 μm, respectively. The needle array pattern of the electroformed-Ni mold was produced by combining a grayscale mask for X-ray lithography with Ni electroforming technology. The X-ray grayscale mask was composed of Si absorbers and a SU-8 membrane. Each Si absorber had a three-dimensional cone shape rather than a rectangular shape. Threedimensional Si structures were formed to etch an active Si layer in a silicon-on-insulator wafer using a taperedtrench etching technology. Beamline BL-4 in the TERAS synchrotron radiation facility at AIST was used for the Xray lithography experiments. X-rays that penetrated the X-ray grayscale mask irradiated a polymethylmethacrylate (PMMA) sheet. Pt was deposited on the PMMA structure after developing, and Ni was electroformed on it. The electroformed-Ni object was processed by grinding to complete a Ni mold. Finally, the micro-needle array was fabricated by thermal-imprinting on a 0.5-mm thick PC sheet with the electroformed Ni mold.     

浸没式光刻机浸没流场的仿真与试验

浸没式光刻是当前45 nm以下集成电路(Integrated circuit,IC)生产线上唯一实际应用的技术.它通过在最后一片投影物镜和硅片之间填充高折射率的浸没液体来提高光刻的分辨率.作为光刻系统中光路的一部分,浸没液体需要保持良好的均一性.然而,曝光过程中光刻胶泄漏污染和曝光温升的问题,会破坏流场的均一性,并最终...     

光刻机模拟隔振试验装置内部世界动力学模型分析

为给步进扫描光刻机的设计研究提供理论指导，解决光刻机掩模台、工件台由于高速运动而引起光刻机工作核心部分——工作平台的振动问题，根据工业应用中步进扫描光刻机的运动特点和工作要求，设计了一种内部世界和外部世界用精密6自由度减振器连接的步进扫描光刻机隔振试验装置；建立了内部世界6自由度的动力学模型，运用设计的参数计算出了内部世界的固有频率和对应的振型；同时运用ANSYS有限元软件进行了分析计算，两者的计算结果吻合很好，验证了动力学模型的正确性。     

Self-dressing effect using a fixed abrasive platen for single-sided lapping of sapphire substrate

Single-sided lapping is crucial in sapphire wafering processes for improving flatness and achieving the target wafer thickness using loose abrasives. In single-sided lapping process, the Material removal rate (MRR) is a key factor for reducing process time and cost. However, the MRR is limited when using loose abrasives because abrasives mostly act by rolling and sliding. Many researchers have studied fixed abrasives to increase the MRR, but the MRR decreases with time. To solve this problem, the self-dressing effect was studied with various pressures, velocities, cutting fluids and wafers. The MRR decreased due to the wear of abrasives, and the pressure and velocity have little effect on the self-dressing. Lapping experiments were done using cutting fluid with a lapped wafer and sawed wafer. The MRR, plate roughness and thickness were measured to study the wear of the abrasive and the self-dressing effect. The cutting fluid delayed the wear of the abrasives and thus improved the decrease in MRR, but it had little effect on the self-dressing effect, like in the case when water was used. When using cutting fluid and a sawed wafer, the MRR was high and did not decrease. A concentrated load on the plate caused by shape error and saw marks on the sawed wafer could produce the self-dressing effect. We verified that a sawed wafer could produce the self-dressing effect on even a worn plate.     

Modeling and investigation on wafer shape in wafer rotational grinding method

With the fast development of the integrated circuits, the size of the wafer is getting larger and the demand for the wafer shape and flatness has become higher and higher. There are many factors which could influence the wafer shape. So in this paper based on the rotational coordinates method, a theoretical model of the ground wafer shape in wafer rotational grinding is firstly developed, in which the main factors were considered, including parameters of the dressing vacuum chuck and the wafer grinding, etc. Secondly, the simulation system was developed by Visual C++ and OpenGL according to the model, by which the simulation results of wafer shape could be made and the relationship between the wafer shape and parameters can be gotten. Finally, the experiments of wafer grinding were carried out on an ultra-precision grinder (VG401MKII), and the developed model is verified to be correct by comparing the experiment results and simulation results. The research indicates the wafer shape and flatness can be predicted exactly, and it is quite significant to choose the reasonable parameters to effectively control the wafer shape in the wafer rotational grinding. The total thickness variation can be controlled within 1 nm if the parameters are chosen reasonably.     

SU-8胶光刻工艺研究

SU-8胶是一种基于环氧SU-8树脂的环氧型的、近紫外光、负光刻胶.其专门用于在非常厚的底层上需要高深宽比的应用.但是SU-8胶对工艺参数的改变非常敏感.本文对影响光刻后图形质量的主要工艺参数前烘温度和时间、中烘温度和时间、曝光时间及显影时间进行了研究,发现前烘时间和显影时间是影响图形分辨率及高深宽比的最主要的参数.随后给出了200μm厚SU-8光刻胶的建议工艺条件:200μm/s甩胶,1h的95°C前烘,近紫外光(400nm)接触式曝光,95°C的中烘 30min,PGMEA中显影20min.另外对实验中实现的主要问题基片弯曲和光刻胶的难以去除作了一定的探讨,给出了合理化建议:对于基片弯曲可采用以下四种措施来降低,降低中烘的温度同时增加中烘的时间、用厚硅片来代替薄硅片、对于薄硅片在前烘后可用金刚刀切成4～8小片、适当的设计掩模板;对于光刻胶的去除用热丙酮泡、超声清洗、反应离子刻蚀和高温灰化法相结合,能达到较好的效果.     

释放保型软压印光刻工艺

针对压印过程中模具与晶圆密贴度和压印面积的矛盾,提出释放保型软压印光刻工艺。此工艺包括五步加载过程,可充分挖掘压印图型转印误差根源以及图型转印保真度与阻蚀胶留膜厚度的内在矛盾。通过目标载荷量的调节与控制将模具弹性回弹调整到紫外光固化步骤前,消除压印过程的内在矛盾,实现压印面积由2 cm2向8 cm2的提升,同时保证阻蚀胶留膜厚度的要求。基于创新的释放保型软压印工艺,一套低成本及结构简单的压印光刻原型机被设计构建,一系列压印结果证明应用此工艺的压印光刻机具有同场压印不同面积、特征尺寸结构图型的多次复制能力。     

Effects of wafer curvature caused by film stress on the chemical mechanical polishing process

A theoretical model based on two-body contact theory is established to simulate the contact pressure distribution arising from wafer curvature which is caused by film stress during CMP process. Both wafer and pad deformations during the contact process are considered. The profiles of the contact pressure distribution for wafers with different curvature radius are simulated. The influences of wafer curvature on mean removal rate and within wafer removal rate nonuniformity (WIWNU) are simulated and compared with the experimental data. According to the two-body contact model, when the pad is in contact completely with the wafer, the profile of the contact pressure has almost the same trend whether the wafer has an upward or a downward curvature. The mean value of the contact pressure will increase with increasing of radius of downward curvature. WIWNU will decrease with increasing pre-polish wafer bow from concave (upward curvature) to convex (downward curvature). The results from the simulation correlated with the experimental data and demonstrated the validity of the model. The results are helpful for controlling and reducing the wafer to wafer removal rate nonuniformity and within wafer removal rate nonuniformity in CMP.     

A nano-scale alignment method for imprint lithography

A novel nano-scale alignment technique based on moiré signal for room-temperature imprint lithography in the submicron realm is proposed. The moiré signals generated by two pairs of quadruple gratings on mold and wafer are optically projected onto two photo-detector arrays, then the detected moiré signals are used to estimate the alignment errors in the x and y directions. The experiment results indicate that complex differential moiré signal is sensitive to relative displacement of the mold and wafer, and the alignment accuracy obtained in the x and y directions and in ϑ are ±20 nm, ±25 nm and ±1 μrad (3σ), respectively. They can meet the requirements of alignment accuracy for submicron imprint lithography.     

基于遗传算法的步进式扫描光刻机路径规划

根据步进式扫描光刻机硅片曝光过程的特点,建立了光刻机路径规划问题的数学模型。采用遗传算法对光刻机路径规划模型进行了合理的规划。规划结果表明,规划后的扫描路径能有效地提高光刻机的生产效率。     

Nanopositioning and Nonlinearity Compensation for Step Imprint Lithography Tool

In this paper, the motion mode and nanopositioning accuracy in the step imprinting lithography process are presented, and the positioning errors different from the traditional errors, such as the gap error existing in the hinges of the stage structure and the random error produced during the process of the stage position adjustment, are analyzed. To avoid and eliminate these nonlinearity errors, radial basis function-proportional integral derivative and position control algorithms are introduced into the macro- and microdriving processes, respectively. The innovation of this driving method is that the motion locus is monotone, nonoscillatory, and a multistep approaching target, which eliminates the root of the random error by single direction driving mode and avoids the backlash error through preloading function. Driving experiments of different motion ranges prove that this nonlinearity compensation is very effective and the positioning accuracy during the step imprinting process can be improved up to 10-nm.     

基于模糊Petri网推理的半导体生产线动态调度研究

针对动态调度的特点,为了综合优化半导体生产线性能指标,尝试使用基于模糊Petri网推理的方法进行动态调度。首先分析了影响动态调度决策的生产线状态信息;然后建立了模糊Petri网形式化推理机,继而构建了面向半导体生产线的模糊Petri网推理模型;最后使用实际半导体生产线模型,将提出的方法与FIFO、EDD和CR策略进行了仿真比较。结果表明基于模糊Petri网推理的动态调度方法能够改善半导体生产线多种性能指标。     

光刻机工件台宏动三自由度建模及自适应神经网络控制

提出了一种工件台宏动三自由度建模方法以解决光刻机工件台宏动部分在X和Y方向的运动耦合问题并实现它的超精密长行程微米精度的跟踪定位。该建模方法将X方向电机的偏转角度作为被控对象并且在模型中包含了耦合效应对X方向运动的影响。基于此模型提出了一种自适应神经网络控制策略,该策略采用径向基函数(Radial Basis Function,RBF)神经网络对模型参数信息及外界非线性扰动进行实时在线估计,以减小未建模动态、电机齿槽力波动、端部效应、摩擦等扰动对控制系统性能的影响。通过对控制策略的理论推导和稳定性分析,保证了闭环控制系统的收敛性。最后在光刻机工件台上进行了S曲线跟踪定位试验,验证了宏动三自由度建模方法和控制策略的效果。试验结果显示:X和Y方向的位置跟踪误差均小于3μm,X方向电机偏转角度小于1μrad,满足工件台宏动部分跟踪定位精度的要求。