mTorr Pressure Range Research Articles

A high-power impulse magnetron sputtering global model for argon plasma–chromium target interactions

A high power impulse magnetron sputtering (HiPIMS) discharge process is analyzed numerically and experimentally so that one may better understand and explain the effect of the pressure and pulse width on HiPIMS discharges and the deposited thin films. For this reason, a time-dependent global plasma model is developed for the ionization region in a HiPIMS discharge of a Cr target in Ar gas. It is based on the solving of a nonlinear equation system composed of the continuity equations of neutral and charged species in the ionization region considered in the reaction scheme. The pulse widths are about tens of microseconds for a frequency cycle of 1 KHz. The simulations are performed for a 150 W average power and 5–30 mTorr pressure range. In these average power and pressure ranges, a double peak of the electron temperature is observed. The high first peak is due to a high rapid increase of negative voltage during the rising time of pulses, while the second one is due to a diminution of the injected gas Ar in the ionization region. On the other hand, gas rarefaction characterized by the diminution of the Ar density during time-on is due to the high ionization degree and the sputtering wind effect. During plasma off, the density of Ar returns to the stationary state corresponding to the initial pressure. In addition, the simulations reveal that the Cr+ population is more important than that of Ar+ caused by the high ionization degree of Cr compared with Ar. This confirms the high ionization degree during time-on, leading to a high efficiency of sputtering of Cr material. However, Cr2+ is still weak. In addition, a good agreement is shown between the calculated time current evolution and the measured one.

Determination of recombination coefficients for hydrogen, oxygen, and nitrogen gasses via in situ radical probe system

The determination of the recombination coefficients of gases on solid surfaces depends on the plasma processing environment including factors such as temperature, surface morphology, impurities, and chamber geometry that play a role in energy transfer mechanisms of association, dissociation, and collisional nature of gases in low pressure plasmas. To determine those recombination coefficients, a zero-dimensional plasma model was created to predict radical and ion densities of hydrogen, oxygen, and nitrogen using experimental data, with electron temperatures and densities as inputs. The model inputs (electron density, electron temperature, and plasma gas temperature) were experimentally obtained by a Langmuir probe and a thermocouple. Each radical density measurement requires two radical probes with different catalytic coatings, which yield different temperatures due to different recombination coefficients of the coatings. The measurements are compared with the radical density obtained from a plasma model in order to determine the value of recombination coefficient. Recombination coefficient of hydrogen radicals on the gold surface is found to be 0.115 ± 0.018. Recombination coefficients of oxygen and nitrogen on copper are found to be 0.31 ± 0.063 and 0.18 ± 0.034, respectively. Ion densities vary from 109 to 1011 cm−3 s, over 10–100 mTorr pressure range and power range between 300 and 900 W. Radical densities are in the order of 1013 cm−3 to 1015 cm−3. Simultaneously with this article, a parallel study is published explaining in situ measurements of the radical probe system for single and mixed gases.

On uniformity and non-local transport in low pressure capacitively coupled plasmas

Low pressure (<30 mTorr) capacitively coupled plasmas (CCP) are important for many plasma processing applications. Due to high energy of electrons in these plasmas and long electron mean free path, kinetic phenomena dominate the electron behavior. A multi-dimensional particle-in-cell (PIC) model is utilized to study the physics of sub-100 mTorr Ar CCPs in this article. A one-dimensional model for helium plasma in the 30–300 mTorr pressure range is first used to benchmark the PIC code against published results by Turner et al [2013 Phys. Plasmas 013507]. The article then focuses on two-dimensional modeling of the gaseous electronics conference (GEC) reference cell in cylindrical geometry. Computed ion density and DC bias voltage at 100 mTorr are compared to available experimental data for a range of radio frequency voltages, demonstrating good agreement. The validated model is used to examine Ar plasmas in the GEC reference cell in the 12.5–100 mTorr pressure range. It is found that the peak in ion and electron densities is near the electrode edge at 100 mTorr due to strong electric field there. As the gas pressure decreases, the peak in plasma density moves to the chamber center. This change in plasma spatial profile is linked to kinetic behavior of electrons, both regarding plasma production and transport in the plasma. Time-dependent results are used to illustrate that electron current at 25 mTorr is mainly kinetic and carried by beam electrons generated by expanding sheaths.

Chemical Vapor Deposition Growth of Silicon Nanowires with Diameter Smaller Than 5 nm.

Quantum confinement effects in silicon nanowires (SiNWs) are expected when their diameter is less than the size of the free exciton (with a Bohr radius ∼5 nm) in bulk silicon. However, their synthesis represents a considerable technological challenge. The vapor–liquid–solid (VLS) mechanism, mediated by metallic nanoclusters brought to the eutectic liquid state, is most widely used for its simplicity and control on the SiNWs size, shape, orientation, density, and surface smoothness. VLS growth is often performed within high-vacuum physical vapor deposition systems, where the eutectic composition and the pressure conditions define the minimum diameter of the final nanowire usually around 100 nm. In this article, we present and discuss the SiNWs’ growth by the VLS method in a plasma-based chemical vapor deposition system, working in the mTorr pressure range. The purpose is to demonstrate that it is possible to obtain nanostructures with sizes well beyond the observed limit by modulating the deposition parameters, like chamber pressure and plasma power, to find the proper thermodynamic conditions for nucleation. The formation of SiNWs with sub-5 nm diameter is demonstrated.

High resolution magnetic field measurements in hydrogen and helium plasmas using active laser spectroscopy.

Passive spectroscopic measurements of Zeeman splitting have been used reliably to measure magnetic fields in plasmas for decades. However, a requirement is that the field magnitude must be sufficiently strong to be resolved over Doppler and instrument broadening (typically >10 000 G). A diagnostic for measuring magnetic fields spectroscopically well below this limit (>20 G) with high sensitivity has been developed at the Oak Ridge National Laboratory. The diagnostic relies on measuring a high resolution spectral profile using Doppler-free saturation spectroscopy (DFSS) and then fitting the spectrum to a quantum mechanical model. DFSS is an active, laser based technique that greatly reduces the influence of Doppler broadening and eliminates instrument broadening. To date, the diagnostic has been successfully employed to measure the magnetic field in magnetized (550-900 G), low-temperature (5-10 eV), low-density (1010-1012 cm-3), hydrogen and helium plasmas in the 5-200 mTorr pressure range using a low power (25 mW) diode laser. Implementing an approximate crossover resonance model, the measurements are shown to be accurate within 5 G for helium and 83 G for hydrogen. The accuracy in hydrogen can be improved to 39 G if the crossover resonances are neglected. A more robust crossover model can decrease this error to <1 G.

Structure of the Clean and Oxygen-Covered Cu(100) Surface at Room Temperature in the Presence of Methanol Vapor in the 10-200 mTorr Pressure Range.

Using ambient pressure X-ray photoelectron spectroscopy (APXPS) and high pressure scanning tunneling microscopy (HPSTM), we show that in equilibrium with 0.01-0.2 Torr of methanol vapor, at room temperature, the Cu(100) surface is covered with methoxy species forming a c(2 × 2) overlayer structure. In contrast, no methoxy is formed if the surface is saturated with an ordered oxygen layer, even when the methanol pressure is 0.2 Torr. At oxygen coverages below saturation, methanol dissociates and reacts with the atomic oxygen, producing methoxy and formate on the surface, and formaldehyde that desorbs to the gas phase. Unlike the case of pure carbon monoxide and carbon dioxide, methanol does not induce the restructuring of the Cu(100) surface. These results provide insight into catalytic anhydrous production of aldehydes.

Evidence of local power deposition and electron heating by a standing electromagnetic wave in electron-cyclotron-resonance plasma.

Microwave plasmas excited at electron-cyclotron resonance were studied in the 0.5-15 mTorr pressure range. In contrast with low-limit pressure conditions where the plasma emission highlights a fairly homogeneous spatial structure, a periodic spatial modulation (period ∼6.2 cm) appeared as pressure increased. This feature is ascribed to a local power deposition (related to the electron density) due to the presence of a standing electromagnetic wave created by the feed electromagnetic field (2.45 GHz) in the cavity formed by the reactor walls. Analysis of the electron energy probability function by Langmuir probe and optical emission spectroscopy further revealed the presence of a high-energy tail that showed strong periodic spatial modulation at higher pressure. The spatial evolution of the electron density and of the characteristic temperature of these high-energy electrons coincides with the nodes (maximum) and antinodes (minimum) of the standing wave. These spatially-modulated power deposition and electron heating mechanisms are then discussed.

Optical diagnostics for characterization of electron energy distributions: argon inductively coupled plasmas

The assumption of thermodynamic equilibrium in low-temperature plasmas can lead to errors in determination of electron energy distribution functions (EEDFs) using optical diagnostics. Significant improvements in the accuracy of EEDFs determined using optical diagnostics on argon-containing inductively coupled plasmas (ICPs) are possible by fitting an analytic representation of the EEDF with two adjustable parameters (x and Tx) to recorded optical emission spectra. This so-called x-form of the EEDF captures the well-known suppression of the high-energy portion of the EEDF (compared with the Maxwellian form) observed in ICPs. A review of electron kinetics summarizes the physical mechanisms that lead to this form, and explains the weak dependence of x on operating pressure. For the 1–25 mTorr pressure range of argon-containing mixtures in the system examined, the fixed value x = 1.2 is found to represent the EEDF very well, followed by a near linear increase in x with pressure from x = 1.2 to 1.6 in the 25–50 mTorr range. Significantly improved agreement between predictions of effective electron temperature with a global model and both probe and optical measurements when a Maxwellian EEDF is replaced by the x-form further highlights the benefit of the x-form in improved accuracy in determining the EEDF with optical diagnostics.

Application of Doppler spectroscopy inH2to the prediction of experimentalD(d,n)3Hereaction rates in an inertial electrostatic confinement device

In our previous paper, we developed an optical measurement for absolute densities of fast (tens of keV) H ions in an abnormal hollow cathode discharge of hydrogen in the units to tens of mTorr pressure range. We apply this method to a cylindrically symmetric inertial electrostatic confinement (IEC) discharge of hydrogen using the Doppler spectrum of Halpha. We predict neutron production rates for an equivalent discharge of deuterium and compare these with experimental values under the separate conditions of constant deuterium gas pressure (approximately 6 mTorr) and voltage (30 kV). Our predictions capture the variation of production rates with dc current (10-50 mA) and agree with experiment to within an order of magnitude. The applicability of this diagnostic to the cylindrical IEC discharge is thus demonstrated, with our results supporting the theory that the discharge is dominated by energetic neutrals emerging from the cathode apertures.

Absolute densities of energetic hydrogen ion species in an abnormal hollow cathode discharge

We develop an optical measurement for densities of fast (units to tens of keV) hydrogen ions in an abnormal hollow cathode discharge in units to tens of mTorr pressure range. This method combines results from previous collisional-radiative models, comparing the intensity of Balmer H_{alpha} due to dissociative excitation of H2 by fast electrons to Doppler-shifted emission arising from charge exchange of energetic ions. The method requires only two inputs: the current density due to fast electrons and a single H_{alpha} spectrum of the characteristic emission channel at the anode. We model in particular the cylindrical interelectrode discharge of an inertial electrostatic confinement device. Experimentally, we find that the density of fast ions emerging from the cathode (bias -5 kV at 20 mTorr ) is in the order 10;{14}m;{-3} , increasing approximately linearly with current in the 10-30 mA range. Calculated densities agree with values obtained in similar apparatus using Langmuir probes and analysis of dust particle motion.

Global model simulations of low-pressure oxygen discharges

We use a global model (volume averaged) to study plasma discharges in molecular oxygen gas in the 1-100 mTorr pressure range. This model determines densities of positive ions O2+ and O+ , negative ion O- , electrons, ground state O2 and O atoms, and metastables O2(a¹deltag) and O(¹D), and electron temperature as function of gas pressure and input power, for a cylindrical discharge. We apply the model to O2 discharges and the results are compared to a particle-in-cell simulation (PIC), experimental data and a volume-averaged global model developed at the University of California at Berkeley. We find that the total positive ion density increases with pressure at low pressures (up to approximately 30 mTorr), and decreases at higher pressures. The electronegativity decreases with increased power and increased pressure as predicted by the global models presented in the literature. The predictions for electron temperature are also in agreement with these models. However, there is a discrepancy betweeen these global models and PIC simulations and experimental data, for 20 and 40 mTorr cases, concerning electronegativity calculations. PIC simulations yield much higher electronegativities. There are strong indications that this is due to the assumption of Maxwellian electron energy distribution functions in the global model, while in the PIC simulations this is clearly not the case.

Hydrogen dissociation on high-temperature tungsten

H 2 dissociation on polycrystalline tungsten is measured from 1700 to 3000 K using the filament temperature ( T) change and a normalized H-atom density at the chamber surface. The dissociation probability per H 2 filament collision ( P diss) saturates at 0.40 at high T and has a 2.25 ± 0.05 eV apparent activation energy when P diss ≪ 1. This activation energy is consistent with prior data and models, but the H 2 pressure dependence is not. P diss is independent of the H 2 pressure for this entire T range and the 1–85 mTorr pressure range studied, contradicting the primary model that has been used to explain H 2 dissociation on tungsten and other metals. We show that some apparently contradictory prior measurements are actually consistent with our observations and with each other, once this pressure dependence of P diss is recognized.

Relative densities of hydrogen ion species in a hollow cathode glow discharge

Highly resolved Doppler shifted peaks of the hydrogen Balmer lines, resulting from charge exchange of H+, H2 +, and H3 + with an H2 gas target, were obtained without the need for a deconvolution procedure. This enabled the unambiguous determination of the chemistry of these hydrogen species in the units of mTorr pressure range. This was obtained from a gaseous discharge using a biconical hollow cathode that yielded a well collimated and monoenergetic beam.

High Pressure Scanning Tunneling Microscopy Study of CO Poisoning of Ethylene Hydrogenation on Pt(111) and Rh(111) Single Crystals

Using a high-pressure scanning tunneling microscope we monitored the coadsorption of hydrogen, ethylene, and carbon dioxide on platinum(111) and rhodium(111) crystal faces in the mtorr pressure range at 300 K in equilibrium with the gas phase. During the catalytic hydrogenation of ethylene to ethane in the absence of CO the metal surfaces are covered by an adsorbate layer that is very mobile on the time scale of STM imaging. We found that the addition of CO poisons the hydrogenation reaction and induces ordered structures on the single-crystal surfaces. Several ordered structures were observed upon CO addition to the surfaces precovered with hydrogen and ethylene: a rotated ( × )R23.4° on Pt(111), and domains of c(4 × 2)-CO + C2H3, previously unobserved (4 × 2)-CO + 3C2H3, and (2 × 2)-3CO on Rh(111). A mechanism for CO poisoning of ethylene hydrogenation on the metal single crystals was proposed, in which CO blocks surface metal sites and reduces adsorbate mobility to limit adsorption and reaction rate of ethylene and hydrogen.

Performance of radio frequency hollow cathodes at low gas pressures

Typical hollow cathodes (HC) have distances between opposite inner walls in a millimeter range and can operate at gas pressures of approximately 0.1–10 Torr (13.3–1333 Pa). This is because the pressure and the cathode geometry must fulfill the condition for the hollow cathode effect, based on electron exchange between opposite walls. In this paper, we report on the performance of the radio frequency powered cylindrical and linear HC, with the gas flowing through the cathode, at the mTorr (0.1 Pa) pressure range typical for magnetron sputters or arc evaporators. Operation of simple cylindrical HC with inner diameters exceeding 10 mm was possible at pressures down to 0.9 mTorr (0.12 Pa) in argon. The plasma ion density did not exceed magnitude of 109 cm−3. However, a plasma density as high as 5×1011 cm−3 was measured 15 cm below 16 cm long linear Magnets-in-Motion (M-M) HC in mTorr pressure range. The M-M cathodes can be used as an efficient sputtering/evaporation source for large area PVD of films or incorporated into different hybrid PVD and PE CVD plasma systems with conventional sources for production of new quality coatings.

A combined linear ion trap time-of-flight system with improved performance and MS(n) capabilities.

A detailed description of a linear ion trap time-of-flight (TOF) mass spectrometer system, capable of sequential mass spectrometry (MS(n)), is given. Many improvements have been incorporated since the initial description of this system (Rapid Commun. Mass Spectrom. 1998; 12: 1463-1474). The pressure in the trap has been lowered from 7.0 to 1.8 mTorr, resulting in an increase in the mass resolution of ion excitation from 75 to 240. Use of the system for MS(3) is demonstrated. Dipole excitation of the n = 1 harmonic, instead of the n = 0 fundamental frequency of ion motion, is shown to have a higher frequency resolution, f/Deltaf, but lower mass resolution, m/Deltam. Both experiments and modeling demonstrate that at the lower pressure there is less collisional cooling of ions in the axial and radial directions of the trap. The efficiency of trapping is shown to be nearly 100% for periods up to 5 s. The demonstrated mass range for mass analysis has been extended to greater than m/z 16 250. To avoid the formation of adduct ions when trapping protein ions for extended times requires ultra-high vacuum cleanliness conditions, even though the trap operates in the mTorr-pressure range. Upgrading the TOF to a reflectron with higher quality ion optics results in an increase in the mass resolution of the TOF mass spectrometer to about 5000 at m/z 750.

Diagnostics of inductively coupled chlorine plasmas: Measurement of Cl2+ and Cl+ densities

The absolute densities of positive ions (Cl2+ and Cl+) are obtained over a 2–20 mTorr pressure range and 5–1000 W input radio-frequency rf power range in a transformer-coupled Cl2 plasma. The relative number density of Cl2+ is measured by laser-induced fluorescence. These laser-induced fluorescence data are calibrated by Langmuir-probe measurements of total positive-ion density at low powers to yield absolute values for nCl2+ and are corrected for changes in rotational temperature with rf power. In turn, the nCl2+ data are used to determine the effective-mass correction for refined Langmuir-probe measurements of the total positive-ion density. The density of Cl+ is then the difference between the total positive-ion and Cl2+ densities. For all the pressures, Cl2+ is found to be the dominant ion in the capacitively coupled regime (input powers below 100 W), while Cl+ is the dominant ion at higher powers (&amp;gt;300 W) of the inductively coupled regime. Experimental results are compared to those from a simple global model. This work is a continuation of a study that provides a complete set of experimentally measured plasma parameters over a broad range of conditions in a chlorine plasma, produced with a commercial, inductively coupled rf source.

S–T conversion induced by external magnetic field in gaseous oxalylfluoride excited to the 7151-level of the à 1Au state

Fluorescence decay of gaseous oxalylfluoride (COF)2 excited to the à 1Au(7151) level was measured as a function of gas pressure in absence and in presence of magnetic field, B=0.3 T. On excitation to this level, the dynamics in both zero and nonzero fields may be described in the intermediate-molecule limit, with the fluorescence exhibiting biexponential and triexponential decay, respectively. The fast component decay rate constant kf=(5.31±0.22)⋅107 s−1 is independent on the (COF)2 gas pressure or magnetic field strength, while that of the first slow component depends on both. The second slow component lifetime is independent of the field strength, while dependent on the gas pressure. The Stern–Volmer dependence, measured with a field in the 0.3–12 mTorr pressure range, is nonlinear. The results obtained were explained by the indirect (electronic- and nuclear-spin-decoupling) mechanism, proposed earlier.

SiC xN y thin films alloys prepared by pulsed excimer laser deposition

In this work, thin films of SiCN have been deposited by pulsed laser deposition on silicon substrates by KrF (248 nm) excimer laser ablation of a SiC sintered target in a vacuum system at room temperature. To obtain various stoichiometries, molecular nitrogen is introduced in the deposition chamber in the 5 to 500 mTorr pressure range. The resultant SiC x N y films are compared to the one prepared in a high vacuum environment (no N 2 gas). The film growth was monitored by real-time kinetic ellipsometry at a single photon-energy (2.5 eV). The film was analyzed by spectro-ellipsometry in the photon-energy range of 1.5< hv<5.0 eV at the end of the deposition process. Tauc's plots are used to estimate the optical band-gap of the films as a function of the N 2 gas pressure. High resolution in situ X-ray photoelectron spectroscopy characterization was performed on every film. The bonding character of the elements in the films is obtained by deconvoluting the XPS peaks. The ellipsometric and XPS results suggest that a new phase alloy is present in the SiC x N y films.

Optical diagnostics of d.c. and r.f. argon magnetron discharges

The plasmas of d.c. and r.f. magnetron discharges have been analysed by emission spectroscopy for argon and metal atom radiative states and by optical absorption for the Ar( 3P 2, 3P 0) metastable and Ar( 3P 1, 1P 1) resonant atoms. In the 6–200 mTorr pressure range for W = 1–50 W in r.f. and I = 2–50 mA in d.c. the metastable and resonant atoms are three to five times more populous in r.f. than in d.c. The densities are saturated for I > 20 mA in d.c. and W > 5 W in r.f., reaching a maximum value n Ar( 3P 2) ≈ 5 × 10 10 cm −3 . The emission line intensities I( λ) vary as I( λ) = W β in r.f., with β ≈ 0.5 for neutral Ar lines and β ≈ 1 for Ar ion and metal (Al, Mg, Mn) atom lines. In d.c. discharges I( λ) = I β , with β ≈ 2 for Ar ion and metal atom lines. These results suggest an electron density variation as n e ∞ W 0.5 in r.f. and an excitation mechanism proportional to n e 2 for Ar ions and metal atoms in both d.c. and r.f. discharges.