Takehisa Yahiro

Nikon EB stepper: its system concept and countermeasures for critical issues

The imaging concept of electron projection lithography (EPL) with silicon stencil reticle is explained. A silicon membrane thickness of 1 - 4 micrometer is suitable for the reticle. A scattering contrast of greater than 99% is expected. Nikon EB steppers dynamic writing strategy of discrete exposure on a sub-field by sub-field basis with deflection control of the electron beam is explained. The basic system configuration of EB stepper is introduced. Examples of error budget for CD variation and Overlay/Stitching are shown. Nikons policy for countermeasures for critical issues such as proximity effect correction, sub-field/complementary stitching and wafer heating influence are explained. For extensibility down to 70 nm and below, both exposure tool and reticle should be improved.

Direct measurement of Coulomb effects in electron beam projection lithography

We describe a direct measurement of the image blur and defocus induced by Coulomb effects (stochastic and global space charge) in electron beam projection lithography. The Nikon 100 kV electron beam projection experimental column was used for the experiments. This column has similar values of electron optical parameters to those of electron beam projection lithography (EPL) systems. The Coulomb effect image blur and defocus were directly measured by a knife-edge method. The experimental data of Coulomb effect image blur and defocus as functions of beam current are shown. The experimental results show excellent agreement with our Monte Carlo simulation results.

Nikon EB Stepper: the latest development status

The latest development status of EB Stepper is reported. The experimental data include the latest resist image data exposed by 100keV electron beam, mask error factors and dosage margins at several backscattered electron levels, transmission data of continuous membrane reticles, and recommended structures for alignment marks, etc. The basic studies related to system design are also explained, those are the strategy for the management of reticle deformation and the stitching accuracy in overlaid layers, etc. Through these data, the resolution capability down to 50nm technology node is clearly shown and alignment/stitching capability is also described. The requirement to a continuous membrane reticle is indicated from experimental data.

High accuracy aerial image measurement for electron beam projection lithography

A direct means of measuring an image blur of electron beam projection lithography (EPL) tools is described. An aerial image sensor used for the image blur measurement was fabricated and evaluated. The signal to noise ratio (SNR) was very high and the signal contrast was 97%. The measured image blur, defined as the distance between 12% and 88% of the beam edge profile, under the optimum condition was 13 nm and the measurement repeatability was 3 nm (e sigma). The measurement error due to the sensor was extremely small, and a quantitative measurement of the image blur can be realized using this technique. The application of this technique to a system calibration is demonstrated. Focus and astigmatism were measured and the optimum settings of focus coils and stigmators were determined with an excellent repeatability. The potential for this technique to provide an automated self-calibration system on the EPL tools is clearly shown.

Full-field exposure performance of electron projection lithography tool

Electron projection lithography (EPL) is a realistic technology for the 65nm node and below, as a complementary technology of optical lithography especially for contacts and gate layers because of its high resolution and large process margin. Nikon has developed an EPL exposure tool as an electron-beam (EB) stepper and the first generation EB stepper; NSR-EB1A is now almost completed as an R&D tool for the 65nm technology node. Using a ϕ200mm reticle, a 20mm×25mm exposure field is realized. Full-field exposure performance of NSR-EB1A is shown. A 70nm isolated line and 1:1 nested lines are simultaneously resolved, as are 50nm 1:2 nested lines. 60nm contact holes are resolved with a depth of focus over a 10μm range and dosage window over ±6%. Stitching accuracy is about 20nm (3σ) and the single machine overlay is about 30nm (mean + 3σ). These data mean sufficient performance for device manufacturing of the 65nm technology node. The concept of a large subfield is one candidate for resolution and throughput e...

First dynamic exposure results from an electron projection lithography tool

Electron projection lithography (EPL) is one of the promising technologies below the 65 nm node, especially for contact hole and gate layers. Nikon is developing an EPL exposure tool as an electron beam (EB) stepper and the first generation EB stepper is now being manufactured. The voltage of 100 kV is adopted for electron beam acceleration. The subfield size is 0.25 mm×0.25 mm on the wafer and the deflection width of the electron beam is 5 mm on the wafer. The magnification of the projection optics is 1/4. A 5 mm×25 mm area from the φ200 mm reticle can be exposed by the combination of beam deflection and stage scanning motion (dynamic exposure). This area is called “a mechanical stripe.” After one mechanical stripe exposure, the reticle and wafer stages turn around and the next exposure of the adjacent mechanical stripe starts as a scan and stitch stage motion. Finally, a 20 mm×25 mm exposure field from the φ200 mm reticle is exposed. We report the first dynamic exposure in the history of EPL although on...

EPL electron optics performance on test stand: 1. resolution results

The performance of the electron optical (EO) column of Nikons EB-stepper mounted on its test stand is reported, mainly focusing on the subject of resolution as preliminary results. Stitching data will be presented in the next paper at MNE2002. Resolution performance data are presented with SEM photos of line/space patterns and contact hole features, printed in both positive and negative resists. EO adjustment techniques were established to obtain the maximum resolution with best trajectory optimisation so that both resolution and distortion can meet their specifications. It is essential to have a metrology capability to measure, distortion as well as resolution. Good resolution results have been achieved over the large deflection (5 mm) field of the EO column using the curvilinear variable axis lens (CVAL) concept, which has been demonstrated by IBM theoretically and experimentally. Current beam blur, depth of focus (DOF) and some characteristic features of the EO conditions are also demonstrated. Image blur at small-deflected positions has also been determined by a direct measurement method. The minimum feature size of 50 nm and a large depth of focus over 5 µm, have been obtained. Overall performance over the whole 5 mm deflected range has not been accomplished yet due to stability limitations of the test stand stage, but good 80-nm line and space resist images have been obtained for the whole deflected range.

Data of scattered electron characteristics in 100-kV EB stepper

Nikon is developing an Electron Beam (EB) stepper as one of the next-generation lithography systems for feature sizes of less than 100 nm. As a reticle for the EB stepper using a high power EB (acceleration voltage: 100 kV, current on reticle: 100 (mu) A), a scattering stencil reticle with a grid-grillage structure has been investigated, EB projection experimental column which operates a high power EB was constructed. Some experimental data of scattered electron characteristics using the EB projection experimental column are given as follows: (1) Scattering contrast of 99.9% can be obtained using 100 kV electron beam (membrane thickness; 2 micrometer, aperture half angle onto reticle; 2 mrad). (2) Changes of resist pattern width of 1:1 and 1:2 lines and spaces are around 40% and around 20% respectively due to the proximity effects by backscattered electrons form the silicon substrate. (3) Contrast of EB mark detection for the system calibration, the reticle alignment, and the wafer registration is obtained. Comparing with the values that be obtained by theoretical calculation, some of experimental data gave good agreement.

Direct measurement of chromatic aberrations induced by SiNx continuous membrane mask

In this article, we present experimental results that quantify the magnitude of chromatic aberrations induced by a SiNx continuous membrane mask. We used a continuous membrane mask, which consists of a TaSi scatterer layer on a SiNx membrane layer of 100 nm thickness, for the experiments. The beam edge profiles were measured directly by the knife-edge method, which can measure the image blur with the repeatability of 3 nm (3 sigma). The effect of electron energy loss induced by inelastic scattering is clearly shown in the measured beam edge profiles. The magnitude of chromatic aberrations is quantified from the measured beam edge profiles considering lens aberrations and measurement errors. The results show that lower atomic number Z and lower density materials and thinner membrane are required on the mask membrane for a high resolution lithography.

High-accuracy aerial image measurement for electron beam projection lithography

A direct means of measuring the image blur of electron beam projection lithography (EPL) tools is described. We developed an aerial image sensor using a Si membrane knife-edge and a transmitted electron detection technique. The aerial image sensor is designed to increase signal amplitude and signal contrast in order to yield a large signal to noise ratio even under a low beam current density condition. The image blur can be quantified accurate to a few nanometers because the measurement error due to the sensor is extremely small. The aerial image sensor was installed in Nikons electron beam projection experimental column and was evaluated. The measured image blur, defined as the distance between the 12% and 88% points of the beam edge profile, under the optimum condition was 13 nm, and the measurement repeatability was 3 nm (3 sigma). The application of this technique to a system calibration is demonstrated. Focus and astigmatism were measured and the optimum settings of focus coils and stigmators were determined with excellent repeatability. The potential for this technique to provide an automated self-calibration system on EPL tools is clearly shown.