"Intel Corp. has revised its lithography strategy for the second time in the recent months, disclosing it has dropped 157-nm tools from its roadmap and is not pursuing the (157-nm) scanner technology for IC production. The move is expected to impact fab-tool and material vendors developing 157-nm products for Intel.’’ Reported by Mark LaPedus in Semiconductor Business News on May 23, 2003." The impact is indeed significant. There are estimates that the scanner tool suppliers have spent in the hundreds of millions of dollars on 157-nm lithography. They are not the only ones left holding the bag. There are suppliers for optical materials, photoresists, antireflection coatings, ellipsometers, pellicles, mask blanks, mask inspection, and repair tools. All of a sudden, Intel is blamed for their losses in development costs and potential profits. Can an industrial giant like Intel bend the laws of physics or the laws of materials? Can we solve the 157-nm problems by spending another billion dollars on it? If we did succeed, would the technology still be economically viable?

With wireless communications becoming an important technology and growth engine for the semiconductor industry, many semiconductor companies are developing technologies to differentiate themselves in this area. One means of accomplishing this goal is to find a way to integrate passive components, which currently make up more than 70% of the discrete components in a wireless handset, directly on-chip thereby greatly simplifying handsets. While a number of technologies are being investigated to allow on-chip integration, microelectromechanical systems technologies are an important part of this development effort. They have been used to create switches, filters, local oscillators, variable capacitors, and high-quality inductors, to name a few examples. The lithography requirements for these devices are very different than those found in standard semiconductor fabrication with the most important involving patterning over extreme topography. We discuss some of the fabrication challenges for these devices as well as some approaches that have been demonstrated to satisfy them.

Microfluidics is emerging as one of the fastest growing segments of micro-electro-mechanical system (MEMS) technologies due to its potential applications in biotechnology, chemical microreactors, and drug discovery. Micromixing is one of the most challenging problems in microfluidic systems, since it is a diffusion-limited process and can be very inefficient. A micromixing device based on an acoustic microstreaming principle is developed to enhance micromixing. The micromixer uses air bubbles as actuators that can be set into vibration by a sound field. The vibration of the air bubbles generates steady circulatory flows, resulting in global convection flows and thus rapid mixing. The time to fully mix dyed solutions in a 50-μL shallow chamber using acoustic microstreaming is significantly reduced from hours (a pure diffusion-based mixing) to 6 s. We demonstrate the use of this micromixer to enhance the performance of conventional DNA microarray biochips that often suffer from lengthy hybridization and poor signal uniformity due to a diffusion-limited hybridization process. Experiments showed that the acoustic micromixer results in five-fold hybridization signal enhancement with significantly improved signal uniformity, as compared to conventional diffusion-based biochips. Acoustic microstreaming has many advantages over most existing micromixing techniques, including a simple apparatus, ease of implementation, low power consumption (~ 2 mW), and low cost.

We present a way to fabricate microgrippers that can meet the industry's needs well, i.e., low cost and large tip deflection, etc. The microgripper is fabricated by bonding two identical micro NiTi-Si cantilever beams together with a silicon spacer in between. It can be actuated by electrical current directly. We have tested the behavior of micro NiTi-Si cantilever beams of three different sizes, and compared that with our simulation results. According to our simulation, the maximum strain and the maximum stress in NiTi should enable the grippers to survive after 10⁶ cycles. Due to the simple fabrication process, this design is very suitable for batch production at low cost, which is a significant advantage in both medical and manufacturing industries.

A scheme of a novel hybrid integrated microspectrometer is proposed, which greatly reduces the number of optical units used in the system and easily realizes the integration of the optical unit and the detecting array. At the same time, the effective area of the grating increases. One model of such device is fabricated. Its volume is about 60×40×40 mm and the volume can be further reduced. In the experiments, the spectrum signal of a Hg lamp is obtained. The results show that a resolution of 7 nm is achieved when a single mode fiber is used as the light input device.

Synthetic silica photomask substrates are currently manufactured by cutting from glass boules, which are prepared using a flame hydrolysis process. An alternative technique based on sol-gel processing demonstrates several potential advantages in fabricating high-quality substrates. This new approach allows near net shape fabrication of synthetic silica photomask substrates, eliminating the need for cutting and grinding. The complex relationship between glass properties and process parameters in the formulation, drying, and sintering steps has been determined, and a repeatable process has been established. The resulting substrates meet all SEMI specifications for ultra-low thermal expansion (ULTE) photomasks for 248-nm lithography. The technology may also be extended to 193-nm and 157-nm photomask substrates. This sol-gel-based process may represent a unique and cost-effective alternative for manufacturing photomask substrates for deep UV lithography.

Theoretical and experimental Cr photomask etch studies are carried out using different resists [ZEP, chemically amplified resists (CAR), and optical resists] and different brand etch tools. The effects of chrome loading are analyzed, and theoretical equations are developed for etch time calculations and endpoint determinations of extremely low Cr load photomasks. It was found that these equations agreed well with experimental data. Etch critical dimension (CD) movement data are analyzed and calculated, showing agreement with experimental data. Metrology measurement and characterization tools include a profilometer, an optical film measurement system, and SEM and optical CD measurement systems. Significant etch performance differences are noted across etch tools, irrespective of the resist type used. An etch property number method is proposed, which is found to accurately describe the etch process analysis and the extent to which etch performance can be expected to be improved. Etch properties are focused on etch CD movement, isolated/dense etch CD bias, radial CD etch contribution, and Cr load effects.

Maskless lithography systems (MLS) presented feature the digital mirror device (DMD) as the pattern generator to replace photomasks. Here 1.5-μm, 10-μm, and 20-μm line/space MLSs are developed. In the MLS, an 848×600 microlens and spatial filter array (MLSFA) was used to focus the light and filter the noise. The MLSFA produces light pads smaller than the 17-μm×17-μm micro mirrors of the SVGA DMD, and filters the noise produced from the DMD, optical lens system, and microlens array. This MLSFA is one of the key components for the Maskless Lithography System, and determines the resolution and quality of maskless lithography. A novel design and fabrication process of a single-package MLSFA for the Maskless Lithography System is introduced. To avoid problems produced by misalignment between a two-piece spatial filter and microlens array, MEMS processes are used to integrate the microlens array with the spatial filter array. In this paper, a self-alignment method used to fabricate exactly matched MLSFA is presented.

As the dimensions of semiconductor devices continue to shrink, mask alignment becomes increasingly important and difficult. The required accuracy of alignment is only a few percent of the used wavelengths. In optical projection lithography, the position of a wafer with respect to a mask is determined by the analysis of the light that is diffracted from an alignment mark on the wafer. The profile shape, depth, and asymmetry of the alignment mark significantly affect the observed alignment signal. In order to predict the influence of these factors, the simulation of the alignment system becomes necessary. We compare rigorous methods for the simulation of light diffraction from the alignment mark, such as the finite-difference time-domain method, rigorous coupled wave analysis, and the waveguide method (WG) to the Fresnel method that has been used in the past. It is shown that the three rigorous methods used in this study demonstrate a good convergence. For the specific geometry and material parameters of the alignment mark used in our investigations, WG is most appropriate for fast work. The difference between Fresnel and rigorous methods becomes important when the height of the alignment mark exceeds 30% of the wavelength of the alignment system.

We developed a technique using electron beams for inspecting contact holes immediately after dry etching and detecting incomplete contact failures. Wafers with deep-submicron contact holes that had high aspect ratios of 10 could be detected during practical inspection time by controlling the charging effect on the wafer surfaces. Measurements of the energy distribution in the secondary electrons exhausted from the bottom of the holes indicated that they were accelerated by the charge-up voltage on the wafer surfaces. Our analysis showed that high-density electron beams must be used to charge the surfaces when the aspect ratio is high. The minimum thickness of the residual SiO₂ that could be detected at the bottom of the contact holes was 2 nm using an aspect ratio of 8. Applying this mechanism to optimize the dry etching process in semiconductor manufacturing showed that we could achieve reliable process control.

The semiconductor industry has pushed linewidths on integrated-circuit chips down to 100 nm. To pattern ever finer lines by use of photolithography, the industry is now preparing the transition to extreme ultraviolet lithography (EUVL) at 13 nm by 2007. As EUVL matures, the requirements for the accuracy of reflectivity and wavelength measurements are becoming tighter. A high absolute accuracy and worldwide traceability of reflectance measurements are mandatory for worldwide system development. A direct comparison of EUV reflectance measurements at the Advanced Light Source (ALS) Center for X-Ray Optics (CXRO) and Physikalisch-Technische Bundesanstalt (PTB) yield perfect agreement within the mutual relative uncertainties of 0.14% for reflectance and 0.014% for wavelength.