Extreme ultraviolet sources enable future chip manufacturing
One of the biggest hurdles facing the introduction of extreme ultraviolet (EUV) lithography is the development of reasonably-priced EUV sources that can accommodate a high wafer throughput. Overcoming this would require source power in the kilowatt range at the 13.5nm exposure wavelength, and an optics-limited emission volume of a few cubic millimeters. This has been our goal (see Figure 1).
Plasmas are known as efficient emitters at 13.5nm at temperatures of about 200,000°C: approximately 35 times higher than at the surface of the sun (see Figure 1). Plasmas can be generated either by an electrical discharge or by means of pulsed laser excitation. Both methods can generate the small plasma volumes demanded by photolithography's technical requirements.1,2
Currently, the highest EUV powers are achieved by so-called ‘pinch plasma’ or ‘Z-pinch’ sources, which achieve the necessary temperatures by compressing, or ‘pinching’, the cross sections of a highly-ionized plasma streams. The plasmas are generated by discharging several joules of energy from a capacitor bank in short pulses to an electrode system, which is located in a low-pressure gas ambient (xenon or tin). After ignition of the discharge, the current generates a magnetic field that compresses the plasma stream onto the axis between the electrodes, generating the high temperatures. This process can be repeated at rates of several kHz to achieve high average power.
While pinch plasma sources can generate sufficient power for the development of EUV lithography3,4 there is still a long way to go to meet the performance requirements for high volume manufacturing, even when switching from xenon to the more efficient tin fuel.5,6 The heat load on the static electrodes configuration can lead to fast erosion and even melting of surfaces, limiting the power increases that can be achieved with higher repetition rates. Alternative technologies with moving electrodes have been investigated that may solve the heat problems, including the use of rotating disk electrodes (RDE). Depending on the disk diameter and rotation speed, the heat-load dissipation of this arrangement can reduce electrode temperatures and extend their lifetimes.
Timing limitations of the pulse-to-pulse dose control impose an additional limit on source repetition rates, which should be kept below 10kHz. Thus, increasing the total source power requires an increase in the energy of each pulse. At XTREME technologies, we have developed a new laser-based excitation scheme that achieves 800W in 2π; steradians during burst operation, a world record in pulse energy. With this method, the repetition rate can be reduced to an acceptable level.7
Combining RDE and the laser excitation scheme, the new approach uses two tin-covered electrodes on a rotating axis. As a high-voltage pulse is applied to the electrodes, a pulsed laser evaporates tin from the electrodes and ignites the vacuum spark discharge. EUV radiation is emitted from the tin plasma (2).
We have investigated other excitation schemes as well, in order to generate higher pulse energies from RDE sources. The output pulse energy of the best approach was four times higher than a conventional setup can achieve. Also, the plasma size was very small, supporting high collection efficiency and etendue transmission through the collector and illuminator optics. With this technology it will be possible to match the power requirements for the final EUV source.
XTREME technologies has made significant gains toward an EUV source suitable for high volume manufacturing. We have developed RDE to address electrode heat dissipation and cooling capabilities, drawing on our extensive experience with high-power pinch-plasma EUV sources. Our new excitation approach has allowed us to increase pulse energy by a factor of four over the initial RDE approach, which should permit a reasonable repetition rate that can meet total power requirements. Further developments will be directed at combining this power scaling capability with reliability goals.
The work performed at XTREME technologies GmbH is partially funded by the German BMBF under contracts 13N8131 and 13N8866 and by the European Commission within the project ‘more Moore’, contract no. IST-1-507754-IP. The development is part of the European MEDEA+ projects T405 ‘EUV source development’ and 2T301 ‘EAGLE’.
