Electron multi-beam mask writer for next generation chips

Chip structure

The basic principle of this chip is conceptually relatively simple: In the electron multi-beam mask writer a fanned-out electron beam is split into 512 x 512 = 262,144 individual beams by a perforated silicon plate. Each electron beam has a diameter of only a few micrometers. The 262,144 individual beams then hit the beam deflection chip, which is located downstream from the sensor. Its integrated a CMOS chip, together with the MEMS processing performed at ISIT, enables the targeted deflection of individual electron beams. This allows precise and selective exposure of the mask substrate located at the end of the beam path, from which photomasks are subsequently produced.

The MEMS components of the chip consist of apertures and gold electrodes through which the passing electrons can be deflected. To illustrate, the principle can be thought of as similar to that of an inkjet printer. The apertures are channels a few micrometers in size etched through the silicon CMOS chip. These channels, also called apertures, are each surrounded by high deflection and shielding electrodes made of galvanic gold. Each of these deflection electrodes is addressable via the integrated CMOS logic and can be energized. When voltage is applied, the passing electron beam is slightly deflected, rendering it ineffective. Deflected electron beams are filtered out by the downstream electron projection optics (shown in Figure 1a by rectangles and horizontal structures along the electron beam path). The CMOS logic allows the voltage state of each electrode to be switched on and off in the nanosecond range.