Computational analysis of the effect of etch-inhibitors on the surface roughness developed during plasma etching

Postgraduate Thesis uoadl:2779468 384 Read counter

Unit:
Κατεύθυνση Τεχνολογίας Ολοκληρωμένων Κυκλωμάτων
Πληροφορική
Deposit date:
2018-08-07
Year:
2018
Author:
Skourti Vasiliki
Supervisors info:
Γεώργιος Κόκκορης, Συνεργαζόμενος Ερευνητής, Ινστιτούτο Νανοεπιστήμης & Νανοτεχνολογίας, ΕΚΕΦΕ Δημόκριτος
Original Title:
Υπολογιστική ανάλυση του ρόλου των παρεμποδιστών εγχάραξης στην τραχύτητα επιφανειών που εγχαράσσονται με πλάσμα
Languages:
English
Greek
Translated title:
Computational analysis of the effect of etch-inhibitors on the surface roughness developed during plasma etching
Summary:
Plasma etching is one of the main processes for making micro- and nanodevices and modifying (initially flat) surfaces. Etching mechanisms are pure chemical etching, physical sputtering and ion-enhanced etching. Plasma etching inevitably induces roughness on the sidewalls of lines of holes during pattern transfer and on the initially flat surfaces. The origin of surface roughness is local nonuniformity of the etching rate on the etched surface. Surface roughness is critical as it affects the properties of the surface and the operation of the devices: For example, the roughness of initially flat surfaces affects their wetting properties and sidewall roughness of lines or holes affects the operation of the relevant devices (e.g. transistors).
The present study focuses on the roughness created during plasma etching of initially flat surfaces. One of the proposed mechanisms for the creation and development of roughness is the participation of the so-called etch-inhibitors or inhibitors in etching. The inhibitors are species deposited on the surface, act as local etch-masks, and do not allow for uniform etching. The aim of this work is to study the effect of etch-inhibitors on the roughness developed during plasma etching. To achieve the goal, a Monte Carlo computational framework is used to simulate the evolution of the morphology of etched surfaces and a mathematical characterization of surface morphologies is performed with statistical measures such as root mean square (rms) roughness, correlation length and power spectral density.
Two sources of etch-inhibitors are examined: The first source is the bulk of the reactor and the second is the etching products being produced locally on the etched surface and redeposited thereon. The study covers both cases of simultaneous etching and deposition processes as well as cases of serial repeating etching and deposition. The effect of the sticking coefficient and etching selectivity of the etch-inhibitors is investigated.
The results of the simulations show that only when the etch-inhibitors come from the bulk of the reactor, periodicity exists on the etched surface. Increasing the sticking coefficient of the inhibitors results in a decrease in rms roughness when the etch-inhibitors come from the surface and an increase when the etch-inhibitors come from the bulk of the reactor. The serial repeating process of etching and deposition does not induce the same roughness with simultaneous etching and deposition. As the number of etching - deposition cycles increases (the frequency increases), the statistical measures of the surface tend to those of the simultaneous etching and deposition, though rms roughness needs a much larger number of cycles for it. Finally, the results of the simulation are compared with experimental results of etched surfaces of different substrates with different plasmas.
Main subject category:
Technology - Computer science
Keywords:
etching, roughness, plasma, deposition, etch-inhibitors, simulation, Monte Carlo
Index:
Yes
Number of index pages:
11
Contains images:
Yes
Number of references:
29
Number of pages:
157
Βασιλική Σκούρτη - διπλωματική εργασία.pdf (13 MB) Open in new window