Transcript Precision Displacement Measurement via a Heterodyne Laser

Precision Displacement Measurement via a
Distance Measuring Interferometer (DMI)
Why DMI Is Needed
500
150
450
135
Node (DRAM 1/2 Pitch) (nm)
Overlay (nm)
Wafer Diameter (mm)
400
105
350
90
300
75
250
60
200
45
150
30
100
15
50
0
2000
Wafer Diameter (mm)
Node (nm)
120
0
2004
2008
2012
2016
Year
Important characteristics of ``International Technology Roadmap for Semiconductors: 2001''
published by the SIA.
General System
Laser
IF
Receiver
Electronics
DMI system is comprised of
three parts:
1. IF
2. Laser
3. Electronics
DMI History
Year
Description
DMI Resolution
1887
Michelson-Morley Experiment. Leads to
Michelson interferometer.
NA
1902
Pieter Zeeman wins Nobel Prize for effects of
magnetic fields on atomic spectra. Leads to
Zeeman split laser.
1960
Bell Laboratories develop first HeNe laser
1964
Airborne Instrument Labs, Division Cutler
Hammer, first commercial displacement IF
1965
Zeeman HeNe Laser (HP)
1968
Perkin-Elmer “Lasergage” homodyne IF
1970
Zeeman laser IF (HP)
l/16
1987
20 MHz Heterodyne, with 2-pass (Zygo)
l/512
~1996
Current electronics (HP/Zygo).
l/2048
Michelson Interferometer


Michelson-Morley experiment (c.a. 1887).
Typical use of the Michelson interferometer is to compare a test optical surface against
a known high quality reference flat. The output of the measurement is a light fringe
pattern viewed from a diffuse surface. These fringes are spatial fringes.
Michelson Interferometer,
Using Polarized Light
Laser
I
Photodetector
Heterodyne
Single Axis Interferometer
10706B Plane Mirror IF
Multiaxis IF
10735A
Design Considerations
CNC Protective Covers
Design Considerations
Split Frequency Limit on Velocity
f 
f split
|
0
c
f HeNe  4.74 1014 Hz
l
7.5 MHz: Agilent

 20 MHz: Zygo
|
5
Vmax 
|
|
|
10
15
20
Frequency (MHz)
lf split
4
|
25
For a four pass plane
mirror IF.
|
30
|
35
Design Considerations
Miscellaneous
•
•
•
•
•
•
•
Vacuum compatibility.
Low adjustability.
Beam size.
Metric vs. U.S. Customary.
CTE between parent structure and IF parts.
Peak-to-Valley (PV) wavefront per optic.
Remote Receiver fibers (bend radius).
DMI System Errors
Deadpath & Environment
Deadpath: Difference in physical optical path between Reference and Measure.
L0
Therefore, in this example the
deadpath is L=4L0. This is
assuming that the air space
between the PBS and the two
quarter wave plates are equal.
DMI System Errors
Deadpath & Environment

Edlen’s Equation
Metrologia, Vol. 2, No. 2, Pg. 71, 1966

Air Temperature, T (°C)

Barometric pressure, P (mmHg)

Relative humidity, H (%)

Gas composition. Typically not measured.
1  (0.817  0.0133T ) 106 
3
0.05762T
n  0.3836391P 
  3.033 10 He
1  0.0036610T



n
c0
L

c OPL
or
OPL 
L
n
, therefore

 n 
OPL  OPL 

 n  n 
DMI System Errors
Abbe Error
,L
 Abbe
 Abbe  L tan 
Abbe error can be eliminated
through with a  measurement,
and by knowing L.
 y1  y2 

 D 
  atan 
DMI System Errors
Cosine Error
•Reduced through proper alignment.
•Part of the accuracy budget, and not the repeatability budget.
•As an example, you can expect a 10706B to have a cosine error of 0.05 ppm
(50 nm for a 1 m travel).
DMI System Errors
Errors Summary
1.
2.
3.
Make the ambient
environment tightlycontrolled and stable,
and apply atmospheric
compensation tools.
Minimize deadpath
distances and Abbe
offsets, and subtract in
the processing.
Properly align the
optics.
Summary
• DMI is currently the most accurate and sensitive linear translation
measurement scheme. Additionally, it has a near limitless translation
measurement bandwidth.
• Relative, not absolute.
• Noncontact.
• Near coaxial measurement of translation axis.
•
•
•
•
•
•
Resolution (Agilent 10897B electronics and two-pass IF): 1.2 nm.
Accuracy: ~2-3 nm.
Max. range: > 10 meters.
All 6 DOFs of a rigid body, are indirectly measurable.
Max. velocity (two-pass IF): 2 m/s.
Typical beam diameters: 3, 6 and 9 mm. 9mm is preferred.