| Zugriffsnummer | 20596 |
| Dokumenttyp | Konferenzartikel |
| Sprache | Englisch |
| Titel | Activities in SEM photomask metrology at PTB |
| Autor(in); Institution |
Frase, Carl Georg; 5.2, Längen- und Winkelteilungen, PTB-Braunschweig
|
| Quelle/Jahr | 8th Multinational Congress on Microscopy: Proceedings:(2007), 49 - 54 |
| Herausgeber(in) |
Nebesarova, Jana
|
| ISBN | 978-80-2399397-4 |
| Verlag | Ceske Budejovice: |
| Konferenzangaben | 8th Multinational Congress on Microscopy, Prague, 17-21, June, 2007, Czech Republic |
| Freie Schlagworte | Scanning Electron Microscopy ; Photomask ; Metrology ; Critical Dimension ; Monte Carlo |
| Zusammenfassung | Photomasks are a central part of the photolithographic fabrication process of microelectronic devices and integrated circuits. The photomasks contain the microstructure patterns to be projected onto a light-sensitive resist layer on top of a silicon wafer. Therefore, the quality of the patterns on the mask is of crucial importance for the function and properties of the integrated circuits. Important dimensional measurement parameters on the masks are the width of microstructures (critical dimension, CD) and their positions (Registration). Photomask standards which contain special test structures in different arrangements and for different nominal CD values were developed by the Physikalisch-Technische Bundesanstalt (PTB) together with industry partners. Standard measurement tools for photomask metrology are scanning electron microscopy (SEM), scanning force microscopy (SFM, AFM), and UV-microscopy. SEM combine high resolution imaging with a short image acquisition time and are especially useful in CD metrology. The SEM-based metrology system used at PTB is called Electron Optical Measuring System (EOMS). Amongst other tasks, it is applied for CD calibration of photomasks. It basically consists of a large vacuum chamber with an integrated one-level 2D stage which allows to load larger planar measurement objects and to travel over 300 mm in both directions. A low voltage (LV-) SEM with in-lens detection capability of secondary electrons for high resolution imaging is mounted on top of the chamber. The calibration of scan position as well as the detection (and correction) of scan field distortions is done by means of the instrument’s laser interferometer controlled 2D specimen stage. The image resolution of the SEM is controlled by 2D Fourier analysis at images of gold-on-carbon test samples. In 2006, the EOMS was upgraded with a new type of LV-SEM (Zeiss ULTRA SEM), which offers in addition to the conventional in-lens secondary electron detection an in-lens detection of higher energy backscattered electrons. Prerequisite for the use of scanning electron microscopy in CD metrology is the accurate physical modelling of the SEM image formation which correlates the SEM images with the specimen topography. Very useful tools in this image formation analysis are Monte Carlo simulations which simulate the diffusion of probe electrons in solid state and the excitation and emission of secondary electrons. We use the program package MOCASIM, developed at the university of Muenster. Furthermore, a new modular and expandable Monte Carlo platform is currently under development at PTB. Important aspects of this Monte Carlo project are modules for three-dimensional specimen structure definition with an interface to CAD programs, realistic BSE and SE detectors based on electron ray tracing, and a three-dimensional representation of the electron probe. Monte Carlo simulation programs are used to analyze image formation, to derive algorithms for CD evaluation, and to generate synthetical SEM images of exactly known specimen structures to test CD evaluation algorithms. The width or critical dimension of a structure is deduced from SEM images using CD evaluation algorithms. An algorithm which was developed and tested in PTB evaluates top CD (i.e. the width of the top plateau of a strucure) at photomask structures and other specimen (e.g. silicon structures) by approximating an exponential function to the inner, exponential growing flank of the edge peaks. The approximated function is extrapolated to a value of 100 % peak maximum and this position is defined as top edge position. The top CD operator was extensively tested in Monte Carlo simulations and shows good conformity with other measurement methods as scanning force microscopy and UV microscopy. Other, more complex CD evaluation algorithms have been developed and are currently in test state. Photomask standards of different material (chrome and MoSi on quartz substrate) were calibrated by SEM (using the top CD operator) and UV microscopy. Physical modeling was performed on the basis of supporting AFM measurements. The CD measurement results for SEM and UV microscopy are in good conformity with the model. |
