An ionizer (640, Fig.10) for forming a gas-cluster ion beam is disclosed (Fig. 10) including inlet and outlet ends partially defining an ionization region traversed by a gas-cluster jet and one or more plasma electron source(s) (642) for providing electrons (644) to the ionizing region for ionizing at least a portion of the gas-clusters to form a gas-cluster ion beam. One or more sets of substantially linear rod electrodes (452) may be disposed substantially parallel to and in one or more corresponding partial, substantially cylindrical pattern(s) about die gas-cluster jet axis, wherein some sets are arranged in substantially concentric patterns with differing radii (458, 464). In certain embodiments, the ionizer includes one or more substantially linear thermionic filaments disposed substantially parallel to the gas-cluster jet axis, heating means, electrical biasing means to judiciously bias sets of the linear rod electrodes with respect to the thermionic filaments to achieve electron repulsion.

Ionizer and method for gas-cluster ion-beam formation

Electrical charge neutralization effects are known to be factors that affect the dose or concentrations of beam treatment by high current implanters. Raising beam energies to 1 MeV and beyond requires an understanding of the effects of both charge stripping and charge neutralization as well as a numerically efficient model compensating for these effects. Charge stripping generates ions of a higher charge state and may cause the measured electronic current from a Faraday cup to overestimate the true particle current. An analysis is based on the concept of an effective charge state for an ion beam and results in a more general interpretation that covers both the charge stripping effect as well as ion neutralization. Dose control using the techniques requires two adjustable parameters: an apparent cross section of interaction between the beam and particles in the beam path and the ratio of the final steady charge state to the initial charge state.

Dose control for use in an ion implanter

Photoresist (PR) outgassing during ion implantation can alter the real-time measurements of implanted dose, limit the maximum allowed dose rate, and introduce unwanted changes to as-implanted depth profiles. Equipment manufacturers assume that the instantaneous gas load resulting from PR outgassing is almost exclusively hydrogen, However, we show that this is not the case during a significant portion of many production implant processes. In particular, we show that high energy implants are typically characterized by constant gas loads in which up to 50% of the evolved gas consists of species other than hydrogen, while in high current implants, the peak pressure excursions often occur during the first few per cent of the implant, when the effluents are largely carbon monoxide (CO) and hydrocarbons (HC). These differences are shown to be quantitatively related to both the instantaneous dose rate, and to the proximity of the dose endpoint to the critical dose. For example, we show how the gas composition from a 150 kV P source/drain implant begins to abruptly change from approximately 40% H/sub 2/, 60% (CO+CO/sub 2/+HC) at a critical dose of 4.5E14, to saturate at 95% H/sub 2/ at a dose of 1E15. We believe this transition to be a consequence of PR amorphization, since the critical dose is in good agreement with the PR outgassing model of T.C. Smith.

Photoresist outgassing in high energy and high current ion implantation

It is well known in ion implantation onto photo resist (PR) coated wafers that there occurs a dose shift due to PR outgassing which encounters a charge exchange reaction to ion beams. It is proven that the Pressure Compensation (Pcomp) process can work well to avoid the dose shift even for multiple exchange processes. The Pcomp necessarily requires to determine two non-linear factors for each recipe. Once the correct factors are obtained, the Pcomp system works reliably for implantation even in a relatively high pressure with a beam of very high power. The process of determination of the factors, however, is very complicated and takes time. An automatic generation of the factors with software has been developed in order to reduce the number of these steps and time for users' convenience and more productivity. This software collects data of the beam current and the system pressure in real time for a batch and calculates the optimized factors immediately after the batch, fitting the data using the least square method for non-linear equations. Users can check if the procedure is successfully done because the result is shown on the operator interface screen with graphically fitted data.

https://ieeexplore.ieee.org/iel5/6566/17540/00812178.pdf

Automatic generation of pressure compensation factors

It is well known that charge exchange phenomena are great issues on ion implanters, because they cause discrepancies such as dose shifts. To reduce the dose shifts, we introduced the system, so-called pressure compensation (Pcomp). On the Pcomp it is assumed that there is a simple relation among the true beam current (I/sub 0/), the detected beam current (I/sub m/) and the beam line pressure (P). From I/sub m/ and P, we can calculate I/sub 0/, which is used to control the dosimetry during implantation. However, for example, when the vacuum condition of the implanter system changes, the relation among I/sub 0/, I/sub m/ and P might change. Therefore, we developed a monitoring system to check how well the Pcomp works at every actual implantation batch.

Monitoring system of reliability and repeatability of pressure compensation

Thick photoresists, typically 3 microns or more in thickness, necessary for high energy implantation present some unique problems. The outgassing of thick photoresist in high energy applications varies from that of thinner resist and lower energies. It requires appropriate processing to cure without reticulation of field regions, deformity of features, or blistering during subsequent processing. This paper examines different resist treatments and their effects on implanter pressure during processing. Data on outgassing of thick photoresist, outgassing effects on absolute dose and dose uniformity as measured by sheet resistance contour maps, and the variation in gas composition are presented.

UV curing and photoresist outgassing in high energy implantation

Electrical charge neutralization effects are well known as an important factor in dosimetry of high current implanters. Extending this to energies of 1 MeV and beyond requires an understanding and a numerically efficient model for the effects of charge stripping as well as neutralization. This stripping effect generates ions of a higher charge state and may cause the measured electronic current from a faraday cup to overestimate the true particle current. We have developed an analysis based on the concept of an effective charge state for an ion beam to develop a more general formalism that can cover the stripping effect as well as neutralization effect. This patent pending formulation requires only two adjustable parameters: an apparent cross section of interaction and the ratio of the final steady charge state to the initial charge state. We present experimental measurements of the observed electrical currents and show that this new numerical model Is a sufficiently accurate representation of the observations to allow excellent dose uniformity and reproducibility.

Charge exchange effects in production high energy implanter dosimetry

Reduction of gate oxide thickness with device design rule dictates increasingly precise compensation of ion beam positive space charge with electrons and reduction of beam potential. In the present work we investigate charging damage to thin gate oxides due to high current ion implants performed on implanters equipped with different types of hardware for beam space charge neutralization. We are comparing GSD-200/E/sup 2/ high current implanters equipped with either Secondary Electron Flood (SEF) or Plasma Electron Flood (PEF) and the ULE2 ultra-low energy high current implanter, which employs the latest charge control technology (Plasma Cell) introduced by Eaton. Gate-antenna type capacitors and SPIDER structures were used to evaluate beam neutralization. We performed charge-to-breakdown (Q/sub BD/) measurements on MOS capacitors to evaluate gate oxide wearout, and V/sub t/ and charge pumping measurements on SPIDER structures to measure concentration of charge traps generated within the oxide and at the Si/SiO/sub 2/ interface. Good correlation of the data collected from Q/sub BD/ charge pumping, device parametrics (V/sub t/) measurements, and readings of in situ beam potential monitors is demonstrated. Experimental results demonstrate that Plasma Cell technology employed in the ULE2 family of ultra-low energy implanters under optimum conditions provides the best charging performance and allows achievement of minimal gate oxide wearout and lowest concentration of charge traps at the gate stack SiSiO/sub 2/ interface in the SPIDER structures.

http://ieeexplore.ieee.org/iel5/6566/17540/00812147.pdf

Comparison of charge control technologies for advanced high current ion implantation systems

We investigated the dependence of channeling suppression on wafer tilt angles for sub-5 keV B/sup +/ and sub-15 keV BF/sub 2/ implantations using a ultra-low energy ion implanter. A noticeable reduction of channeling tail for 1 keV B/sup +/ and 4.5 keV BF/sub 2/ implants was demonstrated to require tilt angles greater than 8/spl deg/. These results are compared with recently published Monte-Carlo simulation data which indicate that ion redistribution at tail region is dominated by channeling along direction with estimated critical angle of 9/spl deg/ for 1 keV B/sup +/. For boron energy at 3.3 keV, wafer tilt of 8/spl deg/ remarkably suppressed the channeling. We also observed that 15 keV BF/sub 2/ implantation, as an equivalent energy of 3.3 keV B/sup +/, with tilt angle of 4/spl deg/ was sufficient for suppressing the channeling tail in dopant distribution due to additional suppression effect of channeling by implant-induced surface amorphization.

Y. Erokhin

Papers

UV curing and photoresist outgassing in high energy implantation

Charge exchange effects in production high energy implanter dosimetry

Comparison of charge control technologies for advanced high current ion implantation systems

Critical angle of channeling for low energy ion implantation