We report measurements of the bulk, neutral gas temperature in a chlorine transformer-coupled plasma. A trace amount (2%–5%) of N2 was added to the discharge and the rotational temperature of the C3Πu state was determined from the C3Πu→B3Πg emission in the ultraviolet. This temperature has been shown by others to be equal to the rotational temperature of ground-state N2, which is the thermally equilibrated (translational and rotational) gas temperature (Tg). The gas temperature 3 cm above the wafer is equal to, or only slightly above the wall temperature (300 K) throughout the low-power, capacitively coupled regime (<60 W, 0.024 W/cm3). Between the lowest (130 W, 0.053 W/cm3) and highest (900 W, 0.36 W/cm3) inductively coupled mode powers investigated, Tg increases sublinearly with power (and electron density). The high-power (900 W) Tg increases with increasing pressure (650, 750, 900, and 1250 K at 2, 5, 10, and 20 mTorr, respectively). Mechanisms of neutral gas heating are discussed. The energy released in dissociation of Cl2 appears to be the dominant heating mechanism.

Trace rare gases-optical emission spectroscopy (TRG-OES) and Langmuir probe analysis have been used to measure the electron temperature, T e , in a high-density inductively (transformer) coupled (TCP) 10 mTorr oxygen plasma as a function of the 13.56 MHz radio frequency (rf) power. The oxygen atomic densities were estimated by O-atom optical emission (8446 Å), and rare gas actinometry (Ar, 7504 Å). In the H-(inductive)-mode, T e increases from 2.6 to 3.4 eV for the low-energy electrons sampled by the Langmuir probe and from ∼3.5 to 6.0 eV for the high-energy electrons sensed by TRG-OES as the rf power is increased from 120 to 1046 W. In the E-(capacitive)-mode, below 50 W, T e measured by TRG-OES increases with rf power from ∼4 eV at very low power (∼7 W) to ∼6.1 eV at 45 W. Between the highest E-mode power (∼50 W) and lowest H-mode power (∼120 W), the T e measured by TRG-OES drops from 6.1 to 3.5 eV, while T e derived from Langmuir probe measurements drops only slightly from 3.0 to 2.6 eV. In the H-mode, the electron energy distribution function (EEDF) is bi-Maxwellian from ∼120 to 1046 W. In the E-mode, the EEDF changes from nearly Maxwellian (possibly Druyvesteyn) at low rf powers (∼7 W) to bi-Maxwellian at the higher E-mode powers (∼45 W). O 2 dissociation is low (∼2%) at the maximum rf power density of 5.7 W cm −2 (1046 W), and this low value is attributed to the high rate of O-atom recombination on the mostly stainless-steel walls. A detailed accounting of the sources of O (8446 Å) emission revealed significant contributions from electron impact excitation from O( 1 S) and dissociative excitation of O 2 .

Trace rare gases optical emission spectroscopy (TRG-OES) is a new, nonintrusive method for determining electron temperatures (T(e)) and, under some conditions, estimating electron densities (n(e)) in low-temperature, low-pressure plasmas. The method is based on a comparison of atomic emission intensities from trace amounts of rare gases (an equimixture of He, Ne, Ar, Kr, and Xe) added to the plasma, with intensities calculated from a model. For Maxwellian electron energy distribution functions (EEDFs), T(e) is determined from the best fit of theory to the experimental measurements. For non-Maxwellian EEDFs, T(e) derived from the best fit describes the high-energy tail of the EEDF. This method was reported previously, and was further developed and successfully applied to several laboratory and commercial plasma reactors. It has also been used in investigations of correlations between high-T(e) and plasma-induced damage to thin gate oxide layers. In this paper, we provide a refined mechanism for the method and include a detailed description of the generation of emission from the Paschen 2p manifold of rare gases both from the ground state and through metastable states, a theoretical model to calculate the number density of metastables (n(m)) of the rare gases, a practical procedure to compute T(e) from the ratios of experimental-to-theoretical intensity ratios, a way to determine the electron density (n(e)), a discussion of the range of sensitivity of TRG-OES to the EEDF, and an estimate of the accuracy of T(e). The values of T(e) obtained by TRG-OES in a transformer-coupled plasma reactor are compared with those obtained with a Langmuir probe for a wide range of pressures and powers. The differences in T(e) from the two methods are explained in terms of the EEDF dependence on pressure.

This article presents measurements of absolute Cl2 and Cl number densities in a chlorine transformer-coupled plasma. It is part of a series of reports on measurements of densities and energy distributions of all charged and neutral species in the same plasma system over an extensive range of pressure and power. Cl2 and Cl number densities were determined from optical emission spectroscopy and advanced actinometry. Number densities relative to the Xe actinometry gas are reported as a function of pressure (1–20 mTorr) and power (10–1000 W) during slow etching of SiO2-covered Si wafers. A detailed treatment of the effects of gas temperature on the conversion of these ratios into absolute number densities is also included. Cl2 is largely (∼90%) dissociated at the highest powers, with a somewhat higher degree of dissociation at low pressure. The Cl number density becomes nearly independent of power at high powers (especially at lower pressure) due to the combination of a higher degree of dissociation of Cl2 and an overall drop in number density due to heating of the gas. A zero-dimensional (global) model is used to compute Cl2 and Cl number densities. It gives a Cl wall recombination coefficient of 0.04 on the plasma-seasoned stainless steel walls.

Characteristics of chlorine, transformer-coupled pulsed plasmas are reported. Time dependencies of electron (ne), positive ion (ni+), and negative ion (ni−) densities and electron temperatures (Te) were measured with a Langmuir probe and microwave interferometry at 240 and 500 W input powers, and pressures between 3 and 20 mTorr. During the OFF portion of the power modulation, ne decreases rapidly as Cl− is formed by dissociative attachment of Cl2. The formation of Cl− is accelerated at high Cl2 densities (at high pressures and low powers). At 10 mTorr and higher pressures, an ion–ion plasma forms near the end of the OFF portion of the cycle, the sheath collapses, and Cl− reaches the wafer. Te decays rapidly in the OFF period and increases with a similar time constant at the beginning of the ON cycle if electrons are present at a sufficiently high level. If ne is very low at the beginning of the ON cycle, such as at high pressure (10 mTorr), then Te exhibits a spike at the beginning of the ON period. In a comparison study, plasma induced damage is reduced when aluminum is etched under similar source power modulation conditions.