MEASURING THICKNESS USING ELLIPSOMETER


IMPORTANT: Latex gloves MUST be worn at all times when handling wafers.




A. Turn on Ellipsometer by fully turning key clockwise (key is on back right hand side of unit). The unit really isn't stable until it has warmed up, so it's good to have it on for half an hour before taking critical data. For most things in our lab, it begins to work adequately right away.
B. Place sample to be measured on the measurement platform (its a vacuum chuck but we typically don't use the vacuum). On the back of the output detector, there is a switch with M, A/M and A. Set it to the middle position, A/M Also, open the shutter at the laser output point.
C. Now we must level the sample. Begin by looking through the eyepiece. You will see two cross hairs, a broad set and narrow set. Adjust eyepiece in/out (it just slides) to focus narrower cross hairs. In general you won't need to do this as they will be in focus from the last user.
D. Below the sample chuck are a locking screw to hold the chuck fixed, two leveling screws toward the front and right, and a height adjustment screw toward the back. Loosen the locking screw. Adjust the leveling screws to center narrow cross hairs within wider cross hairs.
E. There is a signal monitor just below the eyepiece (a bar output). Rotate the output polarizer (its just in front of the detector) until the signal is either maximum, or if the output overloads, until the signal is about 2/3 of full scale. Now use the height adjustment screw to find the height where this signal is maximized. Tighten the locking screw on the sample chuck. Your sample should now be aligned at the right height and with the right (and equal) angles of incidence and reflection.
F. Switch to "A" setting on the back of the output detector. As you proceed at some point you will see the analyzer (output polarizer) begin to rotate.
G. On computer open Ellipsometer file
H. Run Standard program from computer (STD)
I. Using the PF keys, choose "SPECIP", then your desired wavelength (6328 is a good start), then pick the setting for the kind of film you would like to tell the computer you have. The "OXIDE" choice assumes oxide on a silicon substrate. It calculates thickness and index of refraction, but, if the index is uncertain, will assume the index of silicon dioxide. There is also a silicon nitride setting, a setting where the instrument must calculate the index as an unknown, and a setting which forces the instrument to assume an oxide or nitride index even if it feels it can calculate a better one. This program always assumes a silicon substrate. After you choose oxide or another setting, the instrument will make a measurement and output thickness and index.
J. When you have finished your measurements the PF10 key will take you to the initial STD program screen, and then allow you to quit. Turn off the computer monitor but leave the computer running.
K. Remove your sample, turn off the ellipsometer at the key, put on the instrument cover, and sign the logbook noting in particular any difficulties.

Some additional details that may be useful.

-To verify the ellipsometer is working correctly you can measure the test sample.
A. Place the pre-calibrated wafer on sample chuck. (Be very careful handling it. It can be damaged by fingerprints, dust, and especially by dropping.)
B. Follow the procedure above to measure the bluish region in the center of the test wafer using the "OXIDE" setting.
C.Sample should be about 900 angstroms.
D. Carefully return test wafer to its wafer container.

-The ellipsometer is very good at measuring index, but only gets film thickness to within a multiple of a quantity related to the film thickness, optical wavelength, and sample index of refraction.
A. If you get an answer where you know the thickness is wrong, but where the measurement seems to be stable and should work, try hitting the PF key that gives a list. This is a list of possible thicknesses that could go with your measurement. Your correct thickness is probably in the list.
B. Two other modes of operation are available with the standard program that can help decide which thickness is right. The two wavelength option essentially gives two lists, one for each wavelength, and chooses the thicknesses from the two lists which are in closest agreement. The two angle measurement holds the wavelength constant but makes measurements at both 70 and 50 degrees, again giving two lists and matching them. The two angle approach is really quite good and typically spot on. Before using either the two wavelength, or two angle methods, you MUST be trained and approved for these two modes. In particular you should not attempt to modify the system setup from the basic one angle measurement described above without being approved to do so. See the lab coordinator to obtain training and approval.