Santos D. Mayo

Intermittent-contact scanning capacitance microscope for lithographic overlay measurement

A scanning-capacitance-microscope (SCM) mode was implemented by using an atomic force microscope (AFM) operated in intermittent contact and by measuring the tip-to-sample capacitance change at the tip vibration frequency. The intermittent-contact-mode SCM was able to image and determine the overlay separation of metal lines buried under a 1-μm-thick, planarized dielectric layer. Modeling of the intermittent-contact SCM signal across buried metal lines was consistent with the experimental results. This hybrid intermittent-contact AFM and SCM has the potential to measure the lithographic overlay between metal lines located at consecutive levels beneath dielectric layers in an integrated circuit.

Detection of sodium trace contamination in furnace atmospheres at 1000 °C

Free sodium atoms were detected by resonance fluorescence in an open contaminated quartz tube heated to 1000 °C. The quartz tube and furnace were similar to those used in semiconductor device processing. Fluorescence was excited by a cw dye laser tuned to the sodium D1 or D2 transition and directed along the axis of the furnace. Fluorescence from the sodium D2 line emitted in the axial direction was collected by a telescopic system and focused onto a photomultiplier tube. The estimated minimum detectable sodium density in the furnace is 5×105 atoms/cm3. No free sodium was detectable in a processing tube that had not been intentionally contaminated.

Breakdown mechanism in buried silicon oxide films

Charge injection leading to catastrophic breakdown has been used to study the dielectric properties of the buried oxide layer in silicon implanted with high‐energy oxygen ions. Current versus gate bias, current versus time, and capacitance versus gate bias were used to characterize, at various temperatures, MOS metal‐oxide‐semiconductor capacitors with areas in the 1×10−4–1×10−2 cm2 range fabricated with commercially available single‐ or triple‐implant separation by implanted oxygen silicon wafers. The data show that injected charge accumulates in the buried oxide at donorlike oxide traps ultimately leading to catastrophic breakdown. Both Poole–Frenkel and Fowler–Nordheim conduction, as well as impact‐ionization mechanisms, have been identified in the oxide. The charge and field to breakdown in the best buried oxides are, respectively, near 1 C cm−2 and 10 MV cm−1, similar to the thermally grown oxide parameters. Cumulative distributions of these parameters measured over a large number of capacitors show ...

Persistent photoconductivity in SIMOX film structures

Photoinduced transient spectroscopy (PITS) was used to measure the persistent photoconductive (PPC) response in n‐type separation by implanted oxygen (SIMOX) film resistors. A broadband, single‐shot, flashlamp‐pumped dye laser pulse was used to photoexcite interband electrons in the film, and the excess carrier population decay was measured at temperatures in the 60–220‐K range. The PPC signals exhibit nonexponential character and the conductivity transients are recorded as a function of temperature for variable periods up to 30 s. The photoconductive data are analyzed by using the Queisser and Theodorou potential barrier model, and a logarithmic time‐decay dependence is confirmed for the first time in SIMOX material. The hole‐trap density at the conductive‐film–buried‐silica interface is calculated to be in the high 1015 cm−3 to low 1016 cm−3 range. The sensitivity of PITS is demonstrated to be appropriate for characterization of the SIMOX interface structure and for material qualification.

Photoionization cross‐section studies of the platinum‐donor center in silicon

The relative photoionization cross section of the platinum donor center in silicon was measured over the wavelength range of 2.4 to 3.9 μm by electrical deep‐level optical spectroscopy on an n+p junction at 80 K. The data were analyzed in terms of the lattice‐coupling model proposed by Ridley and Amato which was modified for valence‐band nonparabolicity. Good agreement was obtained between the experimental results and the model calculations of the cross section with the energy level of the donor at 0.320±0.005 eV above the valence‐band edge and a Huang–Rhys factor S of approximately 1.4. This S value corresponds to a Franck–Condon energy shift of 70 meV with a phonon energy of 50 meV. Previously reported photoionization data of the gold donor were also fit by the same model, yielding S≂0.4, a surprisingly small value. Estimates were made of the majority‐carrier capture cross section for these two levels and for the platinum acceptor center in silicon, which was measured previously. These estimates, based ...

Measurement of patterned film linewidth for interconnect characterization

Test results from high-quality electrical and physical measurements on the same cross-bridge resistor test structure with approximately vertical sidewalls have shown differences in linewidth as great as 90 nm for selected conductive films. These differences were independent of design linewidth. As dimensions become smaller, the accurate measurement of the patterned conductor width is necessary to assure predictable timing performance of the interconnect system as well as control of critical device parameters.

Radiation dose at the silicon‐sapphire interface due to electron‐beam aluminization

Recently, the process of metallization in an electron‐beam evaporator has been shown to result in a buildup of trapped positive charge at the silicon‐sapphire interface during the fabrication of silicon‐on‐sapphire devices. This charge buildup can be attributed to radiation damage produced by x rays generated by electron impact on the aluminum to be evaporated. This paper gives the results of calculations of the radiation dose in the sapphire near the silicon‐sapphire interface due to the electron‐beam‐metallization process.

Analysis of persistent photoconductivity due to potential barriers

Persistent photoconductivity has been seen in thin silicon resistors fabricated with SIMOX material at temperatures between 60 and 220 K. This effect has been attributed to the depletion of carriers near the interface between the top silicon layer and the buried oxide, which is due to the large number of surface traps at this interface. The depletion of carriers is accompanied by a built-in field on the order of 10,000 V/cm, which causes a potential barrier that is nearly a quarter of the energy gap of silicon. The theory of the recombination kinetics of majority carriers with minority carriers trapped at the interface on the other side of a potential barrier is studied. Both the possibilities of tunneling and thermal activation have been considered. The results show that thermal activation dominates at the temperatures of our measurements in SIMOX material, while at lower temperatures tunneling would dominate.

Effect of spatial averaging on the compositional analysis of crystals by absorption spectroscopy

Calculations of optical absorption based on a model of a single crystal containing spatially periodic compositional variations are presented. These variations can contribute a significant soure of systematic error in the analysis of composition by optical or surface techniques. The model is most appropriate for melt‐grown crystals and in particular for striated semiconductor crystals, and the surface concentration profile which it predicts is confined by comparison with a published x‐ray topographic study of silicon. Implications of the results for optical absorption studies of impurities in silicon crystals are discussed and it is shown that significant measurement errors may occur.

An electrical method for determining the thickness of metal films and the cross-sectional area of metal lines

The electrical thickness of an aluminum-alloy metallization can be determined from resistance measurements of a van der Pauw cross structure at two temperatures, with corrections for the deviation from Matthiessens rule and for thermal expansion. Thickness determinations, made in this way, agree with those made with a calibrated scanning electron microscope (SEM) to within the uncertainty of the instrument. The electrical cross-sectional area of metal lines can be determined by making resistance measurements at two temperatures.