Electron-beam exposures

Unsaturated polyester finishes of this type do not need to be stoved to effect crosslinking, but will cure at room temperature once a suitable peroxide initiator cobalt salt activator are added. The system then has a finite pot life and needs to be applied soon after mixing. Such a system is an example of a two-pack system. That is the finish is supplied in two packages to be mixed shortly before use, with obvious limitations. However, polymerisation can also be induced by ultra violet radiation or electron beam exposure when polymerisation occurs almost instantaneously. These techniques are used widely in packaging, particularly cans, for which many other unsaturated polymers, such as unsaturated acrylic resins have been devised. 

Even though the vacuum-oriented surface techniques yield much useful information about the chemistry of a surface, their use is not totally without problems. Hydrated surfaces, for example, are susceptible to dehydration due to the vacuum and localized sample heating induced by x-ray and electron beams. Still, successful studies have been conducted on aquated inorganic salts (3), water on metals (3), and hydrated iron oxide minerals (4). Even aqueous solutions themselves have been studied by x-ray photoelectron spectroscopy (j>). The reader should also remember that even dry samples can sometimes undergo deterioration under the proper circumstances. In most cases, however, alterations in the sample surface can be detected by monitoring the spectra as a function of time of x-ray or electron beam exposure and by a careful, visual inspection of the sample. 

Figure 38. Schematic of an electron beam exposure system.

This problem leads to two undesirable phenomena first, since an electron beam exposure system requires 10 to 60 minutes in order to serially expose a wafer or mask, sections of the surface exposed last will show a degree of crosslinking less than sections exposed earlier. Since the linewidth of an image is proportional to the exposure dose, it is imperative that we allow the reaction to proceed to the same extent across the entire wafer if we are to control the linewidth across a wafer or mask. 

Figure 19. The fraction of film remaining after electron beam exposure and development as a function of vacuum curing time for three negative...

Resolution. The ultimate resolution capability of the materials was determined by electron beam exposure at 20kv and 5X10 C/cm. ... 

The authors wish to thank E. A. Chandross for many helpful discussions, M. Y. Heilman for the determination of polymer molecular weights, J. Frackoviak for the electron beam exposures and SEM micrographs, and A. C. Adams for the plasma etch data. 

The saturated hydrocarbons are very susceptible to electron beam damage, both in the monolayer and multilayer forms. While aromatic hydrocarbons and other conjugated systems exhibit minimal or no beam damage effects during the time necessary to carry out the LEED experiments, the ordered structures of paraffins disappear after 5 sec of electron beam exposure as a result of desorption or partial dissociation of the organic adsorbates. 

Figure 12 shows the more favorable case of electron beam exposure of the multilayer siloxane resist. Here, the exposure due to backscattered electrons is reduced from 86% to 50% of the primary exposure, and the imaging layer is very thin. Exposure in the 1 x 1 aperture exceeds 70% and is... 

For resist exposure, the resolution limit will be set by the range over which the ions interact with the resist. As with electron beam exposure, ions create secondary electrons up to several nanometers away from the beam, and these electrons can travel further before their energy is absorbed. Ultimate resolution will probably be about 10-20 nm, as it is with electrons. At present this limit is beyond the capabilities of the ion optical systems. 

Clearly, the sensitivity of a resist should be commensurate with machine design parameters to allow optimized throughput. For example, an electron beam exposure system writing at a modulation rate of 20 MHz (dwell time of 50 nsec), a beam current of 5 x 10-8 amps at 10 kV, and an address structure (spot size) of 0.25 2 would require a resist with a sensitivity of 10 6 C/cm2 (1 / 2) or better in order to write the maximum number of wafers per hour of which it is capable. The same argument also applies to other exposure tools. 

An electron beam exposure machine ELS-5000 (Elionix) operated at 20 kV was used for lithographic evaluation. Sensitivity curves to X-ray and deep UV (254 nm) were determined with a X-ray exposure machine SR-1 (Mo target), developed in our laboratory 12) and a Spectro Irradiator, CRM-FA (JASCO), respectively. After exposure, SNR was developed with (3/1) methylethylketone-isopropanol solvent mixture and rinsed with isopropanol. 

Figure 2. SNR sensitivity curves to electron beam exposure using an accelerated voltage of 20 kV.

Figure 4. Molecular weight dependence of contrast to electron beam exposure.

Formulation of Resist Solutions. Forty grams of a Novolak resin was mixed with 10 g of the photoactive compound, and dissolved in 100 g of bis-2-methoxy-ethylether. After wafers were spin-coated, the samples were immediately placed on a hot plate at 82 C for 14 min. The formulation procedure of a composite resist of poly (2-methyl-1-pentene sulfone) in the Novolak resin is as follows the polysulfone was mixed with the resin (13 wt% solid), and then dissolved in 2-methoxyethyl acetate the films were spin-coated onto silicon wafers, and then baked at 100°C for 20 min prior to electron beam exposure. 

Figure 1. Dissolution rates of a composite resist made of a diazonaphthoquinone sensitizer and o-chloro-m-cresol-formaldehyde Novolak resin after 5 /cm2 electron beam exposures. Note this kind of an induction period appeared only in the high-molecular-weight fraction resin.

---