Sputtering

From Wiki on Atomic, Molecular, Plasma-Material Data for Fusion
Jump to: navigation, search

Contents

Introduction

Based on articles in Data Compendium for Plasma-Surface interactions, Nucl. Fusion, Special Issue, IAEA, Vienna (1984) [1]

An energetic particle penetrating a solid causes electronic excitations and nuclear collisions. The latter are responsible for the sputter processes and can be approximately described by binary collisions. Surface atoms are emitted if the energy transfer in such collisions is high enough to overcome surface potential. This model implies the existence of a threshold energy for the impinging particle to cause sputtering. Sputtering is measured by the average number of emitted target atoms per incident particle. This quantity is called the total sputtering yield Y. The definition has been chosen for single-element targets, but it also holds for multicomponent systems and is used if the number of sputtered atoms is of interest. The sputtered atoms show characteristic angular and energy distributions, which are determined by the differential sputtering yield. The differential yield is measured by the average number of sputtered atoms in a certain energy and angular interval per incident particle. Both the total and differential yields depend on the target material and the type of projectile. For a given target-projectile combination, the yield depends on the energy E0 and the incident angle alpha of the bombarding particle (alpha is measured with respect to the surface normal of the target). The differential yield is also a function of the Energy E and the emission angle of the sputtered atoms. The emission angle is characterized by the polar exit angle beta and the azimuthal angle phi. Like alpha, beta is measured with respect to the surface normal; phi is measured with respect to eh intersection line between the surface and the plane of incidence.

Chemical Effects in Sputtering

In physical sputtering, also called knockon sputtering, the sputtered particles receive enough energy from collisions with the incident particles to overcome the surface binding energy. The development of the collision cascade and the surface binding energy may be altered by changes in the chemical composition of the surface due to ion irradiation, but as long as the atoms are ejected via a collision cascade, the process is called chemically enhanced or chemically reduced physical sputtering and sometimes physico-chemical sputtering. If molecules are formed on the surface due to a chemical reaction between the incident particles and the target atoms with binding energies low enough to permit desorption at the temperature of the solid under investigation, the the process is called chemical sputtering .

The differences in the emission processes should lead to clearly different energy distributions for sputtered and thermally desorbed molecules. The borderline between chemical and physical sputtering, however, is not clearly drawn. Compound surface layers are formed in most ion-solid interactions, and the formation of compound molecules is often observed in sputtering. Molecules with low binding energy to the surface may undergo both sputtering and desorption, resulting in energy distributions that do not allow the identification of a single emission process. Chemical sputtering is trongly temperature dependent, as both compound formation and desorption of molecules depend on surface temperature. In some temperature ranges, both chemical and physical sputtering occur, and chemical sputtering may be observable only at particle energies close to or below the threshold energy for physical sputtering.


Physico-Chemical Sputtering

Chemical Sputtering

Physical Sputtering

Based on an article by E. Thomas et al. [2] and APID Volume 7 [3]

Physical Sputtering is defined as a kinetic process by which energy is transferred from an energetic incident atom or ion (projectile) on target atoms. The developing collision cascade leads to the emission of target atoms.

The characteristic value of importance is the sputter yield defined as the number of sputtered atoms divided by the number of projectiles. The sputter yield exhibits a threshold below which the amount of energy transferred to the target atoms is too small for them to overcome the surface barrier. With increasing energy of the projectiles the sputter yield increases, reaches a maximum and decreases again. This decrease at higher energies is caused by the increasing depth of the collision cascade, moving away from the surface. Whereas the collision kinetics is governed by the mass ratio of target atom mass to projectile mass, each element has its specific surface binding energy (usually the heat of sublimation is assumed).

Sputtered particles generally eject from the first one of two surface layers. For heavy projectiles ejection occurs primarily as a result of a series of collisions between recoiling target atoms in the collision cascade; this will be related to nuclear stopping . For light projectiles the backscattering of the projectile from some depth in the solid may be quite significant and as the reflected projectile returns to the surface it may eject a surface atom by a head-on collision. Thus for light projectiles the ejection is related not only to the energy deposited in the collision cascade close to the surface but also to the reflection of the projectile from the solid.

In addition to the dependence of the sputter yield on the collision partners and projectile energy, the yield depends also on the angle of incidence, measured from the surface normal. The sputter yield increases with increasing angle of incidence (as the collision cascade moves closer to the surface), reaches a maximum (typically between 55° and 80°, depending on the projectile target system), and decreases for glancing angles of incidence due to the increase of the particle reflection coefficient. This process will be largely independent of the target temperature, unlike the chemical sputtering where a projectile comes to (nominal) rest in the solid, combines with a target atom, and then is thermally released.

Most experimental data have been obtained with the weight loss method. The errors are typically in the 10% to 20% range, but sometimes the reproducibility can be as much as a factor of two, which is attributed to surface structure changes with bombarding fluence. In general, the surfaces in the experiments are not well characterized regarding surface roughness, and to a lesser extent, surface impurities (depending on residual gas pressure, flux and fluence of the incident beam).

Sputter yields calculated by computer simulation depend on mean repulsive interaction potentials which may be better known for some projectile-target combinations than for others. Uncertainties due to this effect should be less than a factor of two in most cases.

Nearly all simulations consider a flat surface (roughness of the order of half a monolayer thickness is often taken into account). The calculated values are valid for nearly flat surfaces. To check the sensitivity of plasma edge simulation results on surface roughness a sputter yield of twice the yield at normal incidence and independent on the angle of incidence can be tried (an assumption used in DIVIMP for rough surfaces).

Static programs, like TRIM.SP and ACAT for example, provide yields only at low fluence (‘zero’ fluence case). Sputtering of compounds or mixtures of elements usually leads to a preferential sputtering of the lighter species, and therefore to a composition change in the implantation range. Bombardment of targets with non-volatile species can also lead to composition changes and to deposited layers of this species on the substrate. These processes depend on the incident fluence and have to be determined in each case.

Fitting Expressions

  • Bohdansky formula
  • Revised Bohdansly formula
  • New fit formula[3]

Available Data Sets

Data Sets compiled in the APID Volume 7[3] are available at IAEA ALADDIN Database.

References

  1. R. A. Langley, J. Bohdansky, W. Eckstein, P. Mioduszewski, J. Roth, E. Taglauer, E. W. Thomas, H. Verbeek, K. L. Wilson, Data Compendium for Plasma-Surface interactions, Nucl. Fusion, Special Issue 1984, IAEA, Vienna (1984)
  2. E. W. Thomas, R.K. Janev and J. J. Smith, “Particle Reflection From Surfaces – A Recommended Data Base”,INDC(NDS)-287 report, IAEA, Vienna, Austria (1991)
  3. 3.0 3.1 3.2 W. Eckstein, J. A. Stephens, R. E. H. Clark, J. W. Davis, A. A. Haasz, E. Vietzke and Y. Hirooka Physical Sputtering and Radiation-Enhanced Sublimation , Atomic and Plasma-Material Interaction Data for Fusion, Vol. 7 Part B (2000)
Personal tools