SPIE Membership Get updates from SPIE Newsroom
  • Newsroom Home
  • Astronomy
  • Biomedical Optics & Medical Imaging
  • Defense & Security
  • Electronic Imaging & Signal Processing
  • Illumination & Displays
  • Lasers & Sources
  • Micro/Nano Lithography
  • Nanotechnology
  • Optical Design & Engineering
  • Optoelectronics & Communications
  • Remote Sensing
  • Sensing & Measurement
  • Solar & Alternative Energy
  • Sign up for Newsroom E-Alerts
  • Information for:
SPIE Photonics Europe 2018 | Register Today!

2018 SPIE Optics + Photonics | Register Today




Print PageEmail PageView PDF

Micro/Nano Lithography

Projection maskless lithography for 22nm half-pitch and below

Successful testing of a pre-alpha setup promises an effective solution to the problem of masks for advanced CMOS technologies.
17 March 2009, SPIE Newsroom. DOI: 10.1117/2.1200902.1528

Electron-beam direct-write (EBDW) tools have been around for decades and are still successfully used in the semiconductor industry as mask writers and for rapid device development. Typically, EBDW devices operate at 50keV electron-beam energy and have excellent resolution capabilities. The only reason the tools are not widely used in production is their lack of throughput. Multibeam maskless lithography (ML2) offers all the benefits of EBDW while resolving the volume problem by implementing a massive parallelization of electron beams.

At IMS Nanofabrication Vienna (IMS-NANO), we have developed projection maskless lithography (PML2) based on so-called charged-particle large-field projection optics. PML2 is a major innovation on an evolutionary path that—to a large extent—capitalizes on existing e-beam technology parts (e.g., source, lenses, wafer stage, beam software, magnetic shielding, and tool software). The main advantage of PML2 is the realization of pattern transfer through an array of several hundreds of thousands of individually addressable electron beams, thereby pushing the potential productivity from hours per wafer into the wafer-per-hour (wph) regime. Single-axis PML2 has 0.5–1wph throughput potential for the 22nm half-pitch (hp) node, while multiaxis PML2 offers 5–10wph, with possible extension to smaller nodes. Furthermore, several PML2 tools can be clustered on the floor space allocated for an extreme-ultraviolet scanner, potentially providing 100wph for the 22nm hp node with the additional benefit of using the same common infrastructure.

Figure 1. Principles of projection maskless lithography (PML2).

The basic concept of PML2 is shown schematically in Figure 1. In an electron-optical PML2 column, electrons are generated by a flat emitter of high brightness. The condenser below the electron source is used to form a large-field electron beam of high telecentricity, which then impinges with a kinetic energy of 5keV onto the programmable aperture-plate system (APS). Here, the homogeneous electron beam is split into several one-hundred-thousand electron beamlets. These beamlets are accelerated to 50keV and projected with 200× reduction onto a resist-coated wafer. This way, 3.2μm square openings in the APS are imaged as 16nm spots, allowing for controlled 22nm hp nanolithography on the wafer. Further reduction of the APS apertures to below 2μm square will provide <10nm beamlets at the wafer and thus enable extension to smaller nodes. The first PML2 tool featuring all these properties is the PML2 pre-alpha setup (see Figure 2), which we and others developed within the European MAGIC project.1 This tool is capable of handling 300mm wafers and represents an upgrade of the RIMANA 200× test bench,2 which has demonstrated patterning capabilities down to 16nm hp at 15keV,3 to 50keV beam energy.

The central feature of all PLM2 tools, however, is the programmable APS, which constitutes the object in the imaging electron optics. The APS consists of two silicon plates, both of which exhibit a periodic array of apertures. Individual beamlets are formed by the aperture plate, while dynamic structuring is realized by the blanking plate below. Deflection electrodes at every aperture of the blanking plate allow for individual control of the beamlets, i.e., each can be switched off at will. This is enabled by microelectromechanical systems-fabricated deflection electrodes on a postprocessed CMOS chip with large openings. As shown in Figure 1, all deflected beams are blocked at a stopping plate close to the final crossover of the projection optics. Only undeflected beams make it to the wafer surface.

Figure 2. PML2 pre-alpha setup (left), and 300mm chuck (right).

Figure 3. CMOS-APS: programmable aperture plate system with integrated CMOS electronics.

In September 2008 the Fraunhofer Institute for Silicon Technology delivered the first fully processed and electrically functional blanking-plate chip with integrated CMOS electronics (CMOS-BLC). Characterization of this chip in an electron optical test bench with 1:1 projection (1×TB) showed 99.96% functionality, i.e., 42,991 out of 43,008 blankers were operational. In a final assembly step an aperture plate with 3.75μm square openings was mounted onto this blanking plate with ±0.5μm precision at IMS-NANO, resulting in the first completed CMOS-APS module (see Figure 3). The programmable deflection capabilities of this APS unit were successfully tested in the 1×TB and are shown in Figure 4.

Figure 4. Test-bench characterization (1×) of the CMOS-APS unit.

This CMOS-APS unit will be inserted into the PML2 pre-alpha setup early in 2009 to generate initial exposure results. Future CMOS-APS units, however, will use an aperture plate with 2.5×2.5μm openings, providing about 2300 programmable electron beams in the pre-alpha setup with 12.5nm spot size at the wafer. This should lead to convincing proof-of-concept lithography results that are key for further developing and commercializing the PML2 technology together with a strategic partner.

Christof Klein, Elmar Platzgummer, Hans Loeschner, Gerhard Gross 
Vienna, Austria

Christof Klein is the PML2 project manager. From 2004 to mid-2006, he was an Erwin Schroedinger Fellow at the National Center for Electron Microscopy/Lawrence Berkeley National Laboratory, in Berkeley, CA. Prior to that he was a postdoctoral scientist at the Institute for General Physics, Vienna University of Technology, where he had previously received his PhD in physics.