Self-excited Electron Resonance Spectroscopy Research Articles

Implementation of a robust virtual metrology for plasma etching through effective variable selection and recursive update technology

Virtual metrology (VM) is attracting much interest from semiconductor manufacturers because of its potential advantages for quality control. Plasma etching equipment with state-of-the-art plasma sensors are attractive for implementing VM. However, the plasma sensors requiring physical understanding make it difficult to select input parameters for VM. In addition, those sensors with high sensitivity frequently cause several issues in terms of VM performance. This paper will address plasma sensor issues in implementing a robust VM, where self-excited electron resonance spectroscopy, optical emission spectroscopy, and VI-probe are utilized for critical dimension prediction in a plasma etching process. An optimum sensor selection technique which can give insight into effectiveness of plasma sensors is introduced. In this technique, a numerical criterion, integrated squared response, is proposed for effective selection of important sensors for particular manipulated variables. Sensor data shift across equipment preventive maintenance (PM) and its impact on VM performance are also addressed, where a recursive data centering technique is introduced to handle PM-to-PM sensor data drift in a cost-effective way. The application of the technique introduced in this paper is shown to be effective in dynamic random access memory manufacturing. Hopefully, these results will encourage further implementation of robust virtual metrology in plasma etching for semiconductor manufacturing.

Plasma parameter investigation during plasma-enhanced chemical vapor deposition of silicon-containing diamond-like carbon films

In this work, tetramethylsilane-based plasma-enhanced chemical vapor deposition (PECVD) processes were studied, and the deposited silicon-containing diamond-like carbon (DLC) films were analyzed. The main goal was to identify correlations between plasma parameters and the film structure and properties. The electron temperature, gas temperature, and hydrogen and silicon particle densities in these plasmas were calculated using optical emission spectroscopy measurements; the electron density and elastic electron collision rate were determined using self-excited electron resonance spectroscopy. The elemental composition of the films was determined by glow discharge optical emission spectroscopy, and the hardness and Young's modulus were characterized using nanoindentation. The plasma parameters of the gas temperature and electron temperature revealed stringent correlations with the film composition and properties and thus can already monitor the resulting properties during the deposition process. Increasing the gas temperature using power variation leads to reduced incorporation of silicon and hydrogen in the diamond-like carbon films with a simultaneous increase of the film hardness. However, a gas temperature increase using a higher gas flow rate results in a decrease in the film hardness and an increase in the silicon and hydrogen contents. These results are promising concerning the use of plasma parameters for process control of CVD processes.

Electron heating in capacitively coupled discharges and reactive gases

The effective collision frequency νeff of electrons in capacitively driven discharges of Ar∕Kr, Cl2 and BCl3 has been investigated using self-excited electron resonance spectroscopy. The most prominent features are the steep increase of νeff at low power inputs in all three gases and a slight but systematic decrease of νeff versus p for Ar∕Kr and BCl3 over the whole pressure range investigated. At medium pressures, the effective collision rate νeff in Cl2 increases by 2 orders of magnitude which is a clear manifestation for the transition from stochastic to ohmic heating. These features have been correlated with data gained with a V(I) probe. The dependence of the ohmic discharge resistance is mainly determined by the drastic change of νeff rather than by the variation of electron density ne.

Comprehensive analysis of chlorine-containing capacitively coupled plasmas

Capacitively coupled discharges of strongly reactive atmospheres containing mixtures of boron trichloride (BCl3) and chlorine (Cl2) are investigated employing spatially resolved Langmuir probe measurements, and three probes that are spatially integrating methods: optical emission spectroscopy (OES), self-excited electron resonance spectroscopy (SEERS), and impedance characteristics of the discharge. The analysis covers the pure gases including some mixtures, discharge pressure, and rf power over nearly two orders of magnitude, and their impact on important plasma parameters of “first order,” such as plasma density, plasma potential, electron temperature, temperature of the plasma bulk, electron collision rate with neutrals, and actual rf power coupled into the discharge. From these, other properties (electrical conductivity, capacitance, plasma bulk resistance, sheath resistance, and its electrically defined thickness) can be derived. Since the methods are partially complementary, a mutual control of the obtained data is made possible, and we finally obtain a self-consistent model for capacitive coupling connecting data obtained with electrical and optical probes. Compared to electropositive discharges of inert atomic gases (Ar) and molecular gases (H2), which are used as calibration standard for BCl3 and Cl2, the electron plasma density ne is definitely lower, whereas the electron temperature Te is significantly higher, which would be expected by electron attachment to the electronegative molecules—at least at higher discharge pressures. Furthermore, we compared values for Te and ne obtained with OES and SEERS, respectively, and with the Langmuir-probe system. The agreement in electron plasma density and electron temperature for Ar is surprisingly good, despite the fact that the electron energy distribution would be described with two temperatures. For argon plasma, the variation of the calculated dc conductivity for nearly pure capacitive coupling either from impedance measurements or SEERS is within 30%. This is a result of uncertainties in current path rather than principal faults of the various methods. For the reactive, molecular gases, however, the results vary significantly. These data serve to determine several derived properties. Among these, are the sheath thickness, which is compared with optical and electrical data, and the conductivity of the plasma bulk. As they are derived from simultaneous, but independent measurements, they confirm the relative simple model of an electropositive discharge (argon and argon/krypton), and stress the difficulty to describe plasmas consisting of electronegative constituents (Cl2, BCl3, and their mixtures) which is due mainly to a pressure-dependent transition from stochastic to ohmic heating and from electropositive to electronegative behavior.

Chamber maintenance and fault detection technique for a gate etch process via self-excited electron resonance spectroscopy

With the introduction of 300mm wafer and sub-100nm technology processes, semiconductor manufacturers are gradually paying attention to efficient methods for process and equipment control, which is conventionally called advanced process control (APC) and advanced equipment control (AEC). As a potential strategy, an APC∕AEC technique by self-excited electron resonance spectroscopy (SEERS) was evaluated in a dynamic random access memory gate etch process, in terms of chamber maintenance and process control. Small changes in the chamber conditions after wet cleaning, which could not be detected under conventional monitoring methods, were identified by analyzing the electron collision rate of plasma. This event justifies that plasma monitoring is inevitable in chamber maintenance, especially considering that process results gradually tend to be affected by even small chamber changes in sub-100nm technology process era. Also, the first wafer effect, one of the most serious process drifts in an etch process, could be clearly detected by comparing average electron collision rates of plasma during each wafer process. In addition, a strong correlation between average electron collision rate and remaining oxide thickness enables us to control the gate etch process more tightly. Consequently, the APC∕AEC technique by SEERS is expected to be a potent strategy for plasma etch processes in semiconductor manufacturing.

Analysis of chlorine-containing plasmas applied in III/V semiconductor processing

Capacitively coupled discharges of strongly reactive atmospheres containing mixtures of boron trichloride and chlorine are investigated with optical emission spectroscopy and self-excited electron resonance spectroscopy. This analyzes the whole area spanned by these gases and their impact on important plasma parameters like plasma density, electron temperature, and electron collision rate with neutrals. Using these data, roughly calculated cross sections for these gases are obtained in the low-energy region. Molecular chlorine ions, Cl2+, are evidently present to a preponderant amount as a main agent, which are accompanied by chlorine radicals, Cl(I), in mixtures with chlorine. Absolutely no chlorine ions could be found in the plasma which referred to the effective cooling of the Cl-containing species rather than the nonexistence of these species.

Influence of excitation frequency on plasma parameters and etching characteristics of radio-frequency discharges

Plasma parameters such as electron density, electron-collision rate, resonance frequency and the bias voltage were shown to be important for the characterization of radio-frequency (RF) plasmas. By means of self-excited electron resonance spectroscopy (SEERS), the plasma parameters were compared at two different frequencies, 13.56 MHz and 40.68 MHz, in argon, oxygen, fluorine and their mixtures at low pressure.Upon increasing the generator frequency, more RF power has to be applied to achieve the same bias voltage. At the lower pressure, the bias voltage indicates the mean energy of the ions impacting on the substrate and causing damage at the surface. The experimental results show that generally more power was dissipated in the plasma instead of the sheath if the frequency of the substrate generator was decreased. This results in a significantly higher electron density at the higher frequency. The increase of mean ion energy, owing to lower sheath thickness, has to be taken into account as a second-order effect.In order to show the influence of frequency for process applications, the etch rate of SiO2 on silicon was determined in CF4. For the same generator power, there is no well-pronounced dependence of the etch rate of SiO2 on silicon (CF4) on the frequency. For the same bias voltage, the etch rate of SiO2 increases roughly with the frequency.

Surface modification of powder particles by plasma deposition of thin metallic films

Abstract Iron powder particles with a mean diameter of about 2 μm have been charged and trapped in an r.f. plasma with an electron density in the order of 108–109 cm−3. The iron particles were coated with a surrounding aluminum layer by d.c. magnetron sputtering during their confinement. The plasma parameters at r.f. and magnetron operation were confirmed by Langmuir probes, self-excited electron resonance spectroscopy and energy-resolved mass spectrometry, respectively. The coated powder particles were characterized by scanning electron microscopy (SEM) and X-ray photoelectron spectroscopy (XPS). The results of the surface analysis indicate that coating iron powder particles with a close and compact aluminum film by d.c. magnetron sputtering while they are trapped in a weak r.f. plasma is an attractive procedure for powder modification.