Pattern Placement Accuracy Research Articles

Nanoscale Patterning of Colloidal Nanocrystal Films for Nanophotonic Applications Using Direct Write Electron Beam Lithography.

The small size of colloidal nanocrystal quantum dots (QDs) leads to a variety of unique optical properties that are well-suited to nanophotonics, including bright, tunable photoluminescence (PL). However, exploring the properties of solid QD assemblies at the nanoscale has proven challenging because of the limitations in the nanoscale QD patterning methods. Generally, the precise placement of QD solids is difficult to achieve, especially for tall structures with multiple QD layers, and when it is achieved the patterns often cannot withstand the further processing steps required for final device construction. Direct electron beam lithography of QDs has emerged as a straightforward patterning process that does not require ligand exchange and results in structures that retain bright PL. Here, we demonstrate that direct patterning QD films on substrates treated with a self-assembled monolayer of octadecyltrichlorosilane allows us to create feature sizes as thin as 30 nm with heights of multiple layers and characterize the pattern resolution, robustness, and placement accuracy. These structures withstand sonication in a variety of solvents, and the structures are placed within 20 nm of their intended location nearly 100% of the time. We further show how this patterning method can be applied to nanophotonics by measuring the complex refractive index of the QD materials to model the absorption and scattering cross sections of QD structures of various sizes and shapes. These simulations reveal that edge effects arising from the finite shape of the QD nanostructure lead to substantial absorption enhancement when compared to an equivalent volume region taken from a continuous QD film. Finally, we explore more complex structures by patterning QD arrays, multilayer QD structures, and QD disks inside plasmonic resonators.

Pattern Placement Accuracy in Block Copolymer Directed Self-Assembly Based on Chemical Epitaxy

The realization of viable designs for circuit patterns using the dense features formed by block copolymer directed self-assembly (DSA) will require a precise and quantitative understanding of self-assembled feature registration to guiding templates or chemical prepatterns. Here we report measurements of DSA placement error for lamellar block copolymer domains indexed to specific lines in the surface chemical prepattern for spatial frequency tripling and quadrupling. These measurements are made possible by the use of an inorganic domain-selective prepattern material that may be imaged upon polymer removal after DSA and a prepattern design incorporating a single feature serving as an in situ registration mark that is identifiable by pattern symmetry in both the prepattern and resulting self-assembled pattern. The results indicate that DSA placement error is correlated with average prepattern line width as well as prepattern pitch uniformity. Finally, the magnitude of DSA placement error anticipated for a uniform, optimized prepattern is estimated.

Atomic-force lithography with interferometric tip-to-substrate position metrology

Inadequacies in lateral tip stability and registration presently limit lithographic applications of scanning probes. The authors describe a tool they constructed to write sub-10-nm features with 1nm pattern registration. The tool utilizes an interferometric metrology technique called interferometric-spatial-phase imaging (ISPI) to continuously measure tip position relative to a substrate. Direct tip-to-substrate position measurement permits correction for the multitude of error sources encountered in the long mechanical path between a tip and a substrate. Experimental results indicate that the lateral tip position is stabilized by ISPI to 3σ=0.3nm, and pattern placement accuracy in a two-dimensional (2D) grid array is 3σ=0.2nm. According to ISPI measurements, 2D closed figures written in a polymer are overlaid to <1nm. Analysis of patterns written by the tip while under ISPI control provides an error bound that is in good agreement with the ISPI measurements.

Partial blanking of an electron beam using a quadrupole lens

Inadequate pattern-placement accuracy is becoming a limiting factor for conventional scanning-electron-beam lithography (SEBL) in many applications, such as mask writing and fabrication of integrated optical devices. Spatial-phase-locked electron-beam lithography (SPLEBL) with continuous feedback promises sub-1-nm placement accuracy by monitoring and correcting the beam position in real-time during exposure [J. T. Hastings, F. Zhang, and H. I. Smith, J. Vac. Sci. Technol. B 21, 2650 (2003)]. Ideally, SPLEBL provides a continuous reference signal to infer and correct beam position. Hence, the beam current must be reduced by a factor of 10 or more when traversing areas that are not to be exposed. To accomplish this we have investigated a partial-beam-blanker which consists of a quadrupole lens and a current-limiting aperture. This design has several desirable features. First, it does not shift the center of mass of the beam, so that spurious deflection can be avoided. Second, the operating voltage of the device is very low, which makes it suitable for high modulation rate. Third, this device can be easily incorporated into existing SEBL system design. In this article, the theoretical background will be provided, and the effect of the quadrupole lens on the beam probe size evaluated.

Estimation of Mask Placement Error Caused by Multilayer Stress Profile

In order to clarify the influence of multilayer stress distribution on the pattern placement accuracy of an extreme ultraviolet (EUV) mask, the in-plane displacement (IPD) of the mask was calculated using a simulation model based on two-dimensional plane stress theory. In order to deal with multilayer stress distribution, average stress, fluctuation stress and periodic length were defined as parameters. The parameters, which have a significant influence on the mask IPD, are the fluctuation stress and the periodic length. It was found that the average stress of a multilayer film did not affect the mask IPD. An EUV mask with pattern density of 100% makes IPD small by decreasing the fluctuation stress. For EUV masks of 150 mm and 100 mm squares, the periodic lengths that generate the peak IPD in the mask are 200 mm and 150 mm, respectively. In order to improve EUV mask pattern placement accuracy, it is important to achieve the development of uniform multilayer stress distribution.

Particle-induced distortion in extreme ultraviolet lithography reticles during exposure chucking

Since the goal of extreme ultraviolet lithography is to produce circuit patterns with critical dimensions less than 65 nm, a key to its success will be to identify and minimize the major sources of image placement (IP) error at the wafer. Two sources of IP error are in-plane distortion (IPD) and out-of-plane deformation (OPD) of the patterned reticle during chucking in the exposure tool. Among the many possible causes of IPD and OPD is particle contamination. Small pieces of debris lodged between the reticle and chuck have the potential to distort the pattern that is transferred to the device wafer. Such distortions may consume an unduly large portion of the error budget allotted to image placement. In order to limit these IP errors, it is first necessary to gain a thorough understanding of the behavior of a particle trapped during the chucking process. This article describes the techniques that were used to study these trapped particles and their potential effects on pattern placement accuracy.

Optical waveguides with apodized sidewall gratings via spatial-phase-locked electron-beam lithography

We describe a technique to fabricate Bragg gratings in the sides of optical waveguides using a single lithographic step. This technique is particularly suited to the apodized gratings required for add/drop filters in dense-wavelength-division multiplexing. Apodization minimizes cross talk between channels and improves the filter response. Silicon-on-insulator rib waveguides with both uniform and apodized gratings were fabricated using direct-write spatial-phase-locked electron-beam lithography (SPLEBL). This approach combines SPLEBL’s pattern-placement accuracy with the flexibility of direct-write device prototyping. The resulting grating-based devices exhibited substantially reduced side-lobe levels.

Prediction of placement error of extreme ultraviolet lithography mask by simulation model with equivalent layout pattern

An equivalent modeling technique for estimating the deformation of extreme ultraviolet lithography (EUVL) masks caused by a complex pattern layout has been developed. An equivalent layout pattern (ELP) model provides a means of investigating ways of improving pattern placement accuracy. The simulation model is based on the finite difference method. In order to investigate the possibility of making EUVL masks with a high pattern placement accuracy, the ELP model was used to simulate how pattern density and substrate thickness affect placement error. The results obtained agree well with those for an exact model, with the relative error being at most 4%. They suggest that, when the stresses of the multilayer and absorber films cannot be reduced, it may be useful to increase the substrate thickness to improve the pattern placement accuracy. In addition, when the exposure field covers two chips, the placement error should be reduced through proper control of the stress of the absorber film rather than by changing the absorber coverage.

Performance of the Raith 150 electron-beam lithography system

The performance of a Raith 150 electron-beam lithography system is reported. The system’s resolution, stability, intrafield distortion, stitching, and overlay performance are evaluated. Patterning at low- and high-acceleration voltages is compared. The system was used to pattern sub-20 nm features, and the largest intrafield distortion for a 100 μm field was measured to be 15 nm. Pattern-placement accuracy below 35 nm, mean plus twice the standard deviation, was demonstrated.

Theoretical Analysis of Placement Error due to Absorber Pattern on Extreme Ultraviolet Lithography Mask

The pattern placement accuracy of an extreme ultraviolet lithography (EUVL) mask strongly depends on the stresses present in the multilayer and absorber films, since film stress causes both out-of-plane and in-plane distortions. To analyze this elastic deformation, we have developed simulation models that handle the interaction between the multilayer and absorber patterns. The models are based on two-dimensional theories of bending plates and plane stress. The numerical calculations employ the finite difference method and the successive over-relaxation method. To examine the validity and accuracy of the models, the deformation of EUVL masks using simple absorber patterns was calculated. For the calculations, we assumed a standard 6 inch quartz substrate with a reflective coating. The influence of the absorber pattern on placement error was investigated by simulations, and it was found that the absorber pattern is the main factor determining the pattern placement accuracy of an EUVL mask.

Evaluation of a 100 kV thermal field emission electron-beam nanolithography system

We report on the results of performance evaluation tests of a JEOL model JBX-9300FS electron- beam nanolithography system operating at 100 kV. The system was tested in the areas of lithographic resolution, stability, exposure uniformity, and pattern placement accuracy. A minimum spot size of 4 nm was measured at a current of 100 pA. The spot remained below 6.5 nm for beam currents up to 8 nA. Maximum digital to analog converter linearity and deflection amplifier errors were below 0.5 nm. Pattern placement accuracy in a 50 mm square area was found to be within ±16 nm and inside the 500 μm writing field within ±7 nm.

Two-dimensional spatial-phase-locked electron-beam lithography via sparse sampling

We report a new mode of spatial-phase-locked electron-beam lithography based on alignment of each e-beam deflection field to a fiducial grid on the substrate. Before exposing the pattern in a given field, the fiducial grid is sparsely sampled with the electron beam at a subexposure dose. These samples form a two-dimensional moiré pattern that is analyzed to calculate field shift, scale, rotation, nonorthogonality, and trapezoidal distortion. Experimental verification of the approach was carried out with a scintillating fiducial grid quenched by interference lithography. Despite a poor signal-to-noise ratio, we achieved sub-beamstep field-stitching and pattern-placement accuracy.

Relativistic corrections in displacement measuring interferometry

Displacement measuring interferometry is based on measuring the Doppler frequency shift that a beam of radiation undergoes upon reflection off a mirror connected to a moving stage. Usually the velocity of the reflecting stage is very small compared to the speed of light and is therefore deduced using the classical expression for the Doppler shift. We calculate relativistic corrections to the Doppler frequency shift, considering arbitrary stage motion in two dimensions and multiple passes through the moving interferometer arm. Changes in optical path lengths due to the varying stage displacement are explicitly taken into account. For stage velocities on the order of only 1 m/s the resulting corrections to the classically derived stage displacement can amount to nanometers. We discuss model velocity profiles similar to those currently employed in industrial step-and-scan systems for integrated circuit manufacturing, and for recently proposed scanning-beam interference lithography schemes. Expected future increases in stage speed and wafer sizes will necessitate the inclusion of relativistic corrections to the Doppler shift to maintain pattern placement accuracy at the nanometer level.

Improved pattern-placement accuracy in e-beam lithography via sparse-sample spatial-phase locking

A sparse-sampling algorithm has been developed for two-dimensional, spatial-phase-locked electron-beam lithography (SPLEBL). This algorithm corrects for field shift, scale and rotation errors. The algorithm can be used with a scintillating polymeric global-fiducial grid distributed over the substrate. An experimental evaluation of the sparse-sampling algorithm indicates that sub-10-nm pattern-placement accuracy can be obtained with this mode of SPLEBL.

Pattern displacement measurements for Si stencil reticles

We have fabricated Si stencil reticles that are employed by a new type of e-beam projection lithography system (EB stepper). We applied a stress reduction technique to the Si membrane to improve the pattern placement accuracy. The residual stress of Si membranes which were fabricated by anisotropic etching of B-doped Si wafers in KOH aqueous solution was reduced by annealing at 1150 °C. We carried out pattern-displacement measurements for a Si stencil reticle made of a Si membrane where the residual stress was reduced to 10 MPa, and we observed that the pattern displacement error was reduced to less than 20 nm. Furthermore, the pattern displacement in the stencil reticle had a high correlation with the displacement determined from a simulation based on a finite element model. However in the same reticle, we discovered additional, comparatively small displacements in random directions, which was not expected in a membrane that had a homogeneous tensile stress. As a cause of the pattern displacement in random directions, we identified a pattern-width broadening in the dry etching process.

An object pattern post-processor for validation of electron optical modelling

For quantifying the performance of electron beam lithography systems, such as SCALPEL ® , it is desirable to compare experimental measurements directly with the simulated performance predicted by CAD tools. For this purpose, a post-processing software package has been developed, that takes the raw aberrations data computed by our electron optical column design software, and plots current density contours and histograms in the target plane. The new software simulates any type of fabricated pattern, including crosses used for measuring pattern placement accuracy in an optical metrology tool, and high-resolution L-patterns used for measuring critical linewidths in a CD SEM. The method is illustrated for typical simulated patterns.

Modeling and simulation of membrane distortions in Next Generation Lithography (NGL) masks

Next Generation Lithographies will face major challenges to meet the allotted error budgets for sub-130 nm technologies. The development of low distortion masks will be essential. Using comprehensive finite element simulation models, mask distortions can be efficiently and accurately predicted. Pattern-specific distortions have been evaluated to investigate pattern placement accuracy and stability. For the three principal membrane mask technologies, typical pattern-transfer results are presented.

Overlay budget analysis for the 100 nm device generation

In order to improve overlay capabilities for future exposure tools, it is necessary to understand the contributions from various sources. The most important contributions come from well-known sources like mask pattern placement accuracy, alignment system accuracy and from exposure tool stage performance. As the allowances for overlay budget decrease, and improvements in mask fabrication and stage performance are made, a number of previously less significant contributions need to be considered. These contributions include resolution, optical interference, focusing accuracy of the alignment system, and wafer processing deviations. These contributions are characterized in detail in this paper. The investigation was focused on a proven optical alignment system (ALX^T^M) contributions from wafer processing (resist coating and metallization).

Overlay modeling for proximity x-ray lithography

Proximity x-ray lithography is one of the most promising manufacturing technologies for the fabrication of future electronics devices with ground rules of 100 nm and less because of its wide exposure latitude and processing simplicity. However, the ability to make 1× x-ray masks, with high pattern placement accuracy, has continued to be one of the main technical challenges preventing the wide use of this technology by the semiconductor industry. A model has recently been developed to examine the parameters affecting the overall lithographic system overlay performance, such as the mask pattern placement error as well as the error sources from the x-ray aligner. The model treats all of the error sources statistically. The major assumption of the model is that all the error sources are statistically independent. This is a fairly accurate assumption for the error sources under consideration. For example, the mask pattern placement error is and should be independent of the aligner wafer stage stepping errors. Furthermore, the model can simulate error sources that do not follow the normal statistics. The model has three major components: (1) the x-ray aligner machine model that simulates all the major aligner error sources such as mask and wafer alignments, (2) the application parameter component that considers any application-specific setup tool parameters such as the number of wafer alignment marks and the number and location of global alignment sites, and (3) the mask component that accounts for the effects of pattern placement accuracy on alignment marks and overlay metrology targets. A detailed description of the model and the experimental data used to verify its validity will be given. The relative impact of the various subcomponent error contributions on the overall overlay performance will be provided.

Optimization of the temperature distribution across stencil mask membranes under ion beam exposure

The pattern placement accuracy of a stencil mask is affected by thermomechanical distortion induced during ion beam exposure. Any temperature gradient across the mask membrane could result in thermal distortions which contribute to the stringent error budget. Therefore, a concept has been proposed to control the temperature across the mask membrane by using radiation heat transfer between exposure station components. This work extends previous efforts by including the effects of the backside radiation loss of the exposure area, the aperture, etc. Here the exposure station which has been proposed by IMS-Ion Microfabrication Systems is investigated and simulated with finite element models. Parametric studies have been performed in order to optimize design variables such as material properties, and the temperature settings of the cooled lens electrodes and aperture.