Christer Hedlund

Role of fluorocarbon film formation in the etching of silicon, silicon dioxide, silicon nitride, and amorphous hydrogenated silicon carbide

The etching of Si, SiO2, Si3N4, and SiCH in fluorocarbon plasmas is accompanied by the formation of a thin steady-state fluorocarbon film at the substrate surface. The thickness of this film and the substrate etch rate have often been related. In the present work, this film has been characterized for a wide range of processing conditions in a high-density plasma reactor. It was found that the thickness of this fluorocarbon film is not necessarily the main parameter controlling the substrate etch rate. When varying the self-bias voltage, for example, we found a weak correlation between the etch rate of the substrate and the fluorocarbon film thickness. Instead, for a wide range of processing conditions, it was found that ion-induced defluorination of the fluorocarbon film plays a major role in the etching process. We therefore suggest that the fluorocarbon film can be an important source of fluorine and is not necessarily an etch-inhibiting film.

Microloading effect in reactive ion etching

The etch rate of silicon, during reactive ion etching (RIE), depends on the total exposed area. This is called the loading effect. However, local variations in the pattern density will, in a similar way, cause local variations in the etch rate. This effect is caused by a local depletion of reactive species and is called the microloading effect. Silicon wafers patterned with silicon dioxide have been etched in order to study the microloading effect. The pattern consists of a large exposed area and narrow lines at different distances from the edge of the large area. This arrangement makes it possible to study how the distance from the large area, which depletes the etchants, influences the etch rate. The influence of different processing parameters like, e.g., pressure, gas flow rate, and flow direction on the microloading effect have been investigated. It has been found that the microloading effect is small (<10%) compared to other pattern dependent nonuniformities. It is also shown that the nonuniformitie...

Oxidation and Induced Damage in Oxygen Plasma In Situ Wafer Bonding

In this paper we present our in situ, oxygen plasma‐activated wafer bonding process. By keeping one wafer on the anode and the other on the cathode, we have an asymmetric plasma load on the wafers, making our bonding process interesting for low‐temperature applications where damage or defect‐sensitive active layers are bonded to less sensitive carrier wafers. As a step in optimizing the discharge parameters for plasma bonding applications, the effect of the self‐bias voltage on surface energy, oxidation rates, and damage is investigated. An optimum in surface energy was found at moderate self‐bias voltages, both at room temperature bonding and after low‐temperature annealing at 200°C. This is explained by the fact that at these voltages there is a minimum oxide thickness, which promotes the diffusion of water from the bond interface, and also by the fact that at these voltages we have the best surface cleaning conditions. Also, the surface oxide generated by the oxygen plasma seems to be reactive. With our in situ oxygen‐plasma‐activated wafer bonding process there was a major increase in surface energy for wafers bonded at moderate self‐bias volt‐ages compared to conventional wafer bonding performed in ambient air.

SELECTIVE SIO2-TO-SI3N4 ETCHING IN INDUCTIVELY COUPLED FLUOROCARBON PLASMAS : ANGULAR DEPENDENCE OF SIO2 AND SI3N4 ETCHING RATES

In the fabrication of microstructures in SiO2, etch selectivity of SiO2 to masking, etch stop, and underlayer materials need to be maintained at corners and inclined surfaces. The angular dependence of the SiO2-to-Si3N4 etch selectivity mechanism in a high density fluorocarbon plasma has been studied using V-groove structures. The SiO2 etch rate on 54.7° inclined surfaces is lower than on flat surfaces, while the SiO2 etch yield (atoms/ion) is a factor of 1.33 higher. The results are consistent with a chemical sputtering mechanism. The Si3N4 etch yield is greater by a factor of 2.8 for 54.7° inclined surfaces than for flat surfaces. This large enhancement is explained by a fluorocarbon surface passivation mechanism that controls Si3N4 etching. The fluorocarbon deposition is decreased at 54.7° whereas the fluorocarbon etching rate is increased at 54.7°. This produces a thinner steady-state fluorocarbon film on the inclined Si3N4 surface, and results in a large enhancement of the Si3N4 etch yield.

Surface energy as a function of self-bias voltage in oxygen plasma wafer bonding

Abstract A limitation in the use of wafer bonding has been the necessity for high-temperature annealing after contacting the wafers at room temperature. In this paper, we try to find the highest surface energy as a function of self-bias voltage in oxygen plasma-activated wafer bonding, in order to achieve a low-temperature bonding process. The bonding was performed in situ the vacuum chamber. It was found that oxygen plasma has a smoothing effect on the surface roughness, rather independent of the plasma self-bias. However, a moderate self-bias voltage proved to give the highest surface energy for the bonded wafers, both at room-temperature and after annealing at 200°C. We believe that this is due to the fact that a moderate self-bias is the most efficient in removing surface contaminants, like water and hydrocarbons. It was also found that even after annealing at higher temperatures, 480°C and 720°C, the plasma-bonded wafers showed higher surface energy values than wafers bonded in ambient air. This investigation was focused on low-effect plasmas,

Anisotropic etching of Z-cut quartz

The etch rate in monocrystalline quartz depends on the crystalline orientation. Etch-rate diagrams for micromachining of monocrystalline quartz in, for instance, hydrofluoride-based etchants, are a necessity if one requires the best manufacturing conditions for an etched structure. In this paper we use the development of side-wall profiles in etched grooves, on a Z-cut quartz wafer, to produce two-dimensional etch diagrams. The etch conditions are eight combinations of temperature, from 22 degrees C to 80 degrees C, and etchant mixtures of HF and NH4F diluted in water.

Etch rates of crystallographic planes in Z -cut quartz—experiments and simulation

The anisotropic etching behaviour of monocrystalline quartz is studied both experimentally and with computer simulations. The etch rate minima were identified as the crystal planes m, r, r(2), s an ...

Compositional variations of sputter deposited Ti/W barrier layers on substrates with pronounced surface topography

Sputter deposited Ti/W barrier layers have been found to be Ti deficient with respect to the target composition, which is attributed to the preferential resputtering of Ti from the deposited films by energetic neutrals or ions from the discharge. On the other hand, the sputtering yield of most materials is known to be strongly dependent on the angle of incidence of the bombarding species. Due to this angular dependence the resputtering rate of Ti/W films will also be a function of the local orientation of the surface. The substrates of the IC circuits onto which the Ti/W barrier layers are deposited normally possess a pronounced surface topography. The purpose of this work is to study the deviation of the expected Ti/W concentrations at sloped surfaces as a function of the energy of the bombarding species as well as the atom to ion flux arrival ratio. It is shown that the Ti concentration in the films as a result of the preferential sputtering does exhibit substantial concentration variations across sloped surfaces at constant discharge parameters. The experiments are done in an Ar rf discharge, using a Ti/W target with 50 at.% Ti and 50 at.% W. The films have been analysed by Rutherford Backscattering Spectrometry (RBS) and energy-dispersive X-ray Spectroscopy (EDS). The experimental results are discussed and compared with dynamic simulations using the Transport of Ions in Matter (TRIM) code, which indicate that the loss of Ti in the deposited films is primarily due to ballistic preferential sputtering effects.

Angular dependence of the polysilicon etch rate during dry etching in SF6 and Cl2

The angular dependence of the etch rate in reactive ion etching (RIE) and inductively coupled plasma (ICP) systems for polysilicon etching with SF6 and Cl2 is determined using a recently developed direct measurement method. The latter utilizes specially patterned silicon groove structures consisting of 7–10 μm wide planar surfaces which form various angles with respect to the wafer normal. The structures are produced by highly anisotropic wet chemical etching of Si through a gratinglike mask pattern aligned along specific crystallographic orientations of the wafer which results in the development of planar surfaces of various orientations. These surfaces are then coated with the materials to be studied—polysilicon in this case. The deposited polysilicon is then etched under a variety of conditions in a RIE and an ICP reactor and the etch rates determined by interferometric measurements. Since only standard Si wafers are used and the size of the pattern is only a few μm the method is fully IC production co...

Membrane covered electrically isolated through-wafer via holes

An electrically isolated through-wafer via hole, covered by a membrane, has been realized by standard integrated circuits processes together with deep silicon etching. The deep silicon etching was performed by the Bosch silicon etch process in a Plasma-Therm etch system. The via structures were made in double-sided polished 300 µm thick silicon wafers and had widths down to 20 µm, which correspond to an aspect ratio of 15. The fact that the via structures are electrically isolated makes them a suitable start-point to realize interconnections through a wafer. Furthermore, the application field of the via structure is broadened by the flexibility in the design of the structure, which makes it possible to apply an almost arbitrary membrane material onto the via structure at the end of the process.