Scanning Electron Microscopy

Transcription

1 Scanning Electron Microscopy For the semiconductor industry A tutorial Titel Vorname Nachname Titel Jobtitle, Bereich/Abteilung

2 Overview Scanning Electron microscopy Scanning Electron Microscopy (SEM) is a measurement technique that inquires information about the topography of a surface by observing the scattering of electrons. The use of electrons allows to reach higher resolution then with optical techniques, but requires a more sophisticated setup that includes the use of a vacuum chamber and sometimes very time consuming sample preparation. 2

3 Introduction The working principle An SEM consists of an electron gun used to generate high energy electrons which are then focused on the sample. This electron beam is in the following scanned over the surface, which results in a number of radiation products, mostly formed by inelastic or elastic scattered electrons, but also x-rays or luminescence photons. Recording these byproducts as a function of the position of the scanning beam allows to produce a two-dimensional map of the surface, which can then be further analyzed. x-rays other light primary beam backscattered electrons secondary electrons 3

4 Introduction The assembly The setup of an SEM usually consists of the following components: Electron Beam The electrons needed are usually extracted from a thermal or tunneling cathode, accelerated by electric fields and then focused by using magnetic field lenses. Vacuum In order for the electrons to reach the sample and to increase the chance of detection a vacuum is required. This usually limits the size of samples to be analyzed due to the finite size apperatus. Detector To detect the secondary products a number of detectors can be used inside the vacuum chamber. The most common detector is an Everhart-Thornley detector to register electrons. vacuum pump electron source magnetic lenses deflection lenses sample & detection 4

5 Introduction Information depth & products Depending on the energy given to the incoming (primary) electrons, they will be able to penetrate up to a certain depth into the sample. Depending on this depth different byproducts will surface, therefore an analysis of the different secondary products can obtain information about layers beneath the surface (down to a few micrometer) as well as about the surface and its topography. The analysis of the surface topography relies on backscattered primary electrons, other byproducts can be used to analyze the chemical composition. electron beam sample surface Auger electrons secondary electrons backscattered electrons x-rays 5

6 Measurements Recording the topology When primary electrons hit the surface there is a certain chance for them to be scattered back from the surface, similar to visible photons in optical microscopy or to generate secondary electrons from the surface. In case of a fixed position for the detector, the number of electrons detected will depend on the inclination of the sample relative to incoming beam and detector, therefore the inclination of the sample can be detected by comparing the electrons measured for each point. This can then be translated into a 2D map of the surface. In contrast to optical microscopy there is never a virtual or real image formed, the image is always only calculated from the sensor input. processor interconnects [1] 6

7 Parameters Resolution, current and contrast Theoretically the resolution of a SEM will improve the higher the momentum of the electrons is, so it will increase with increasing acceleration voltage. Realistically though most SEMs are more or less limited by the quality of the magnetic lenses and there ability to maintain a clean focus. With modern SEMs resolutions of ~ 1-5nm can be reached. This limit corresponds to the spot size formed by the electron beam on the sample. In case of increasing current, the number of electrons in the beam is increased. As electrons are carrying a charge and therefore interact with each other higher currents will increase the spot size and therefore resolution becomes limited at higher currents. Gold surface with grain size measurements 7

8 Parameters Magnification In contrast to optical techniques higher magnification of an image in SEM can be realized by slowing down the scanning speed of the incident electron beam. This will result in a larger number of pixels being scanned and therefore in a larger image. The usefulness of this technique is of course at one point restricted by the resolution and therefore the acceleration voltage. The magnification is therefore also independent form the objective length used and can possibly start at a few ten times and go up to >500,000 times with high acceleration voltages. x50 x500 Processor chip side wallincreasing magnification [1] x2000 8

9 Sample Preperation What samples can be investigated? A drawback of electron microscopy is that samples often can not be investigated straightforwardly, but need to be specially prepared. The main issue here are non conducting samples or samples including non conducting parts. The reason is that although a fraction of the incoming electrons is scattered from the surface, the larger fraction is absorbed in the sample and will lead to charging effects. Usually this is circumvented by grounding the sample so that all charges can flow off. Non conducting samples need to be metallized first. If the connection to ground is not efficient, even metallic samples can charge up. An other problem is that when imaging polymers (such as photoresist) at a high resolution the energy contained in the electron beam will be enough to alter and melt the resist. Here a good thermal contact in addition to a proper metallization needs to be ensured. 9

10 Parameters Depth of focus The depth of focus (DoF) is the measure how far object can be apart in the vertical direction, so that they give a sharp image. The DoF is inversely related to the resolution as tighter focused beam will diverge quicker resulting in blurred images for objects further away from the focal point. By using an almost parallel electron beam it is therefore possible to enhance the DoF. A high DoF is one of the key features for the high quality SEM images can provide. In electron microscopy a rule of thumb for the depth of focus is DoF = 0.2 beam divergence magnification SEM image with high depth of focus [1] 10

11 Measurements Voltage contrast & charge collection microscopy Here two useful modes of operation for the study of active microelectric circuitry are shortly presented: Voltage-contrast microscopy Here the contrast differences resulting from voltage differences in the metallization are recorded for changes in the voltage level on the device. In order for voltage contrast microscopy to work, any passivation layer needs to be removed before the analysis. Charge collection microscopy Here the voltages (currents) that are produced by the electron beam when hitting the sample are recorded. This can be used to analyze the behavior of single p-n junctions or Schottky barriers very accurately. Again any passivation must be removed. 11

12 Image analysis Distortion & artefacts Due to the image formation process samples might look distorted in certain measurements: Projection Distortion When a planar sample is tilted in reference to the electron beam, the final depth of focus will result in images with varying magnification along the imaging axis. To solve this either the tilt needs to be reduced or the focus must be dynamically adjusted. Scan distortions In case the scan speed is chosen too high, hysteresis effects might appear in the magnetic coils scanning the electron beam or there might be a mismatch between the two scan directions. This results in images, which appear stretched on one axis. Moire effects: This type of distortion is found when imaging periodic structures. Similar to light on a grating this will result in additional fringes appearing on the image. 12

13 Analysis Image enhancement SEM data is usually analyzed in the form of digital images which are still mostly analyzed by eye, therefore it is often crucial to enhance the images digitally in order to improve the interpretation of the data. The following techniques are commonly used: Modification of the histogram Often also referred to as auto scaling, the distribution of gray scale values is analyzed and then the picture is rescaled so that only the most often used values are fitting into the gray scale interval. False colors Algorithms to detect connected areas in SEM images are commonly used to highlight features, which remain hidden otherwise. These areas are then often given different colors to help the identification. False color SEM image [2] 13

14 Applications Semiconductor industry Scanning Electron Microscopy is one of the most widely used techniques in the semiconductor industry. Concrete examples of the usage of SEMs are: Visualization of structuring Characterization of surface structure and roughness Structuring details integrated circuitry Contamination analysis Investigation of etching and metallization results for MEMS processes Analysis of p-n junctions and Schottky contacts 14

15 Specifications Limits & Possibilities Depending on the individual systems and samples the specifications of AFM measurements can differ, so the following values can only be seen as guidelines. Resolution Down to roughly 5nm depending on the sample and Field of view A few mm² scanning area is possible. The size of the sample is often limited due to the required vacuum conditions. Depth of view High depth of view possible even without large degrading of the resolution. Measurement time Depending on the magnification, resolution and scanned field from a few seconds up to a few minutes. 15

16 Comparison Topological surface metrology The following table compares three of the most commonly used methods for measuring surface topologies: AFM SEM optical Sample preparation none Conductivity necessary none resolution ~0.1 nm ~5 nm ~0.5 µm Sample environment any vacuum any Depth of view low good good Field of view ~100 µm ~1 mm ~1 cm Time of measurement 2-10 min min > 1s Topology 3D 2D 2D+ 16

17 Closing Comments Further Reading The following reading can be recommended for deeper understanding: J. Goldstein, Scanning Electron Microscopy and X-ray Microanalysis, Springer 2008 P. Echlin, Handbook of Sample Preparation for Scanning Electron Microscopy and X-Ray Microanalysis, Springer 2009 L. Reimer, Scanning Electron Microscopy: Physics of Image Formation and Microanalysis, Springer 1998 References [1] Alex Pisarski, The Institute of Optics, The University of Rochester [2] S. Deléglise et al. Nature 482, 63 67, 2012 This work was financed by the project SEA4KET co-financed by the European Union under the FP7-ICT programme. 17