Optical Design Applying The Fundamentals

Tutorial Texts Series
 Optical Design Applying the Fundamentals, Max J. Riedl, Vol. TT84

 Infrared Optics and Zoom Lenses, Second Edition, Allen Mann, Vol. TT83
 Optical Engineering Fundamentals, Second Edition, Bruce H. Walker, Vol. TT82
 Fundamentals of Polarimetric Remote Sensing, John Schott, Vol. TT81
 The Design of Plastic Optical Systems, Michael Schaub, Vol. TT80
 Radiation Thermometry Fundamentals and Applications in the Petrochemical Industry, Peter Saunders,
Vol. TT78
 Matrix Methods for Optical Layout, Gerhard Kloos, Vol. TT77
 Fundamentals of Infrared Detector Materials, Michael A. Kinch, Vol. TT76
 Practical Applications of Infrared Thermal Sensing and Imaging Equipment, Third Edition, Herbert Kaplan,
Vol. TT75
 Bioluminescence for Food and Environmental Microbiological Safety, Lubov Y. Brovko, Vol. TT74
 Introduction to Image Stabilization, Scott W. Teare, Sergio R. Restaino, Vol. TT73
 Logic-based Nonlinear Image Processing, Stephen Marshall, Vol. TT72
 The Physics and Engineering of Solid State Lasers, Yehoshua Kalisky, Vol. TT71
 Thermal Infrared Characterization of Ground Targets and Backgrounds, Second Edition, Pieter A. Jacobs,
Vol. TT70
 Introduction to Confocal Fluorescence Microscopy, Michiel Müller, Vol. TT69
 Artificial Neural Networks An Introduction, Kevin L. Priddy and Paul E. Keller, Vol. TT68
 Basics of Code Division Multiple Access (CDMA), Raghuveer Rao and Sohail Dianat, Vol. TT67
 Optical Imaging in Projection Microlithography, Alfred Kwok-Kit Wong, Vol. TT66
 Metrics for High-Quality Specular Surfaces, Lionel R. Baker, Vol. TT65
 Field Mathematics for Electromagnetics, Photonics, and Materials Science, Bernard Maxum, Vol. TT64
 High-Fidelity Medical Imaging Displays, Aldo Badano, Michael J. Flynn, and Jerzy Kanicki, Vol. TT63
 Diffractive Optics–Design, Fabrication, and Test, Donald C. O’Shea, Thomas J. Suleski, Alan D. Kathman,
and Dennis W. Prather, Vol. TT62
 Fourier-Transform Spectroscopy Instrumentation Engineering, Vidi Saptari, Vol. TT61
 The Power- and Energy-Handling Capability of Optical Materials, Components, and Systems, Roger M.
Wood, Vol. TT60
 Hands-on Morphological Image Processing, Edward R. Dougherty, Roberto A. Lotufo, Vol. TT59
 Integrated Optomechanical Analysis, Keith B. Doyle, Victor L. Genberg, Gregory J. Michels, Vol. TT58
 Thin-Film Design Modulated Thickness and Other Stopband Design Methods, Bruce Perilloux, Vol. TT57
 Optische Grundlagen für Infrarotsysteme, Max J. Riedl, Vol. TT56
 An Engineering Introduction to Biotechnology, J. Patrick Fitch, Vol. TT55
 Image Performance in CRT Displays, Kenneth Compton, Vol. TT54
 Introduction to Laser Diode-Pumped Solid State Lasers, Richard Scheps, Vol. TT53
 Modulation Transfer Function in Optical and Electro-Optical Systems, Glenn D. Boreman, Vol. TT52
 Uncooled Thermal Imaging Arrays, Systems, and Applications, Paul W. Kruse, Vol. TT51
 Fundamentals of Antennas, Christos G. Christodoulou and Parveen Wahid, Vol. TT50
 Basics of Spectroscopy, David W. Ball, Vol. TT49
 Optical Design Fundamentals for Infrared Systems, Second Edition, Max J. Riedl, Vol. TT48
 Resolution Enhancement Techniques in Optical Lithography, Alfred Kwok-Kit Wong, Vol. TT47
 Copper Interconnect Technology, Christoph Steinbrüchel and Barry L. Chin, Vol. TT46
 Optical Design for Visual Systems, Bruce H. Walker, Vol. TT45
For a complete listing of Tutorial Texts, visit http://spie.org/tutorialtexts.xml

Tutorial Texts in Optical Engineering
Volume TT84
Bellingham, Washington USA

Library of Congress Cataloging-in-Publication Data
Riedl, Max J.
Optical design : applying the fundamentals / Max J. Riedl.
p. cm. -- (Tutorial texts series)
Includes bibliographical references and index.
ISBN 978-0-8194-7799-6
1. Optical instruments--Design and construction. I. Title.
QC372.2.D4R54 2009
681'.4--dc22
2009022688
Published by
SPIE
P.O. Box 10
Bellingham, Washington 98227-0010 USA
Phone: 360.676.3290
Fax: 360.647.1445
Email: Books@spie.org
www.spie.org
Copyright © 2009 Society of Photo-Optical Instrumentation Engineers
All rights reserved. No part of this publication may be reproduced or distributed

in any form or by any means without written permission of the publisher.
The content of this book reflects the thought of the author(s). Every effort has been made
to publish reliable and accurate information herein, but the publisher is not responsible
for the validity of the information or for any outcomes resulting from reliance thereon.
Printed in the United States of America.

I dedicate this book to Warren J. Smith, my friend and colleague
for more than 50 years. He has influenced my life more than anyone else.
Introduction to the Series
Since its inception in 1989, the Tutorial Texts (TT) series has grown to more than 80
titles covering many diverse fields of science and engineering. The initial idea for the
series was to make material presented in SPIE short courses available to those who
could not attend and to provide a reference text for those who could. Thus, many of
the texts in this series are generated by augmenting course notes with descriptive text
that further illuminates the subject. In this way, the TT becomes an excellent stand-
alone reference that finds a much wider audience than only short course attendees.
Tutorial Texts have grown in popularity and in the scope of material covered
since 1989. They no longer necessarily stem from short courses; rather, they are
often generated by experts in the field. They are popular because they provide a
ready reference to those wishing to learn about emerging technologies or the latest
information within their field. The topics within the series have grown from the
initial areas of geometrical optics, optical detectors, and image processing to include
the emerging fields of nanotechnology, biomedical optics, fiber optics, and laser
technologies. Authors contributing to the TT series are instructed to provide
introductory material so that those new to the field may use the book as a starting
point to get a basic grasp of the material. It is hoped that some readers may develop
sufficient interest to take a short course by the author or pursue further research in
more advanced books to delve deeper into the subject.
The books in this series are distinguished from other technical monographs and
textbooks in the way in which the material is presented. In keeping with the tutorial
nature of the series, there is an emphasis on the use of graphical and illustrative
material to better elucidate basic and advanced concepts. There is also heavy use of
tabular reference data and numerous examples to further explain the concepts
presented. The publishing time for the books is kept to a minimum so that the books
will be as timely and up-to-date as possible. Furthermore, these introductory books
are competitively priced compared to more traditional books on the same subject.
When a proposal for a text is received, each proposal is evaluated to determine

the relevance of the proposed topic. This initial reviewing process has been very
helpful to authors in identifying, early in the writing process, the need for additional
material or other changes in approach that would serve to strengthen the text. Once a
manuscript is completed, it is peer reviewed to ensure that chapters communicate
accurately the essential ingredients of the science and technologies under discussion.
It is my goal to maintain the style and quality of books in the series and to
further expand the topic areas to include new emerging fields as they become of
interest to our reading audience.
James A. Harrington
Rutgers University
vii
Contents
Preface…………….. .......................................................................... xv
Chapter 1 Law of Refraction:

The Foundation of Geometrical Optics .................... 1
1.1 Introduction ............................................................................ 1

1.2 Fermat’s Principle .................................................................. 1
1.2.1 Historic remarks ......................................................... 1
1.2.2 Derivation ................................................................... 1
1.3 Snell and the Lens .................................................................. 3
1.4 Graphical Ray Tracing ........................................................... 4
1.5 Paraxial Ray Tracing.............................................................. 5
1.5.1 Equations, symbols, and sign conventions .................. 6
Chapter 2 Best Shape for a Thin Lens ....................................... 11
2.1 Concept of Thin Lens............................................................. 11

2.2 Object at Any Position ........................................................... 12
2.3 Object at Infinity with Added Field of View ......................... 16
2.3.1 Spherical aberration .................................................... 16
2.3.2 Chromatic aberration .................................................. 18
2.3.3 Coma .......................................................................... 19
2.3.4 Astigmatism................................................................ 19
2.3.5 Total blur-spot size ..................................................... 21
2.4 Special Case ........................................................................... 21
Chapter 3 Best Shapes for Multiple Thin Lenses,

Aspherizing, and the Natural Stop Position ............. 23
3.1 Introduction ............................................................................ 23

3.2 Best Shapes for Minimum Spherical Aberration ................... 23
3.3 Aspherizing a Singlet to Eliminate Spherical Aberration ...... 26
3.4 Correcting Coma and Spherical Aberration ........................... 28
3.4.1 Eliminating coma........................................................ 28
3.4.2 Eliminating spherical aberration................................. 29
3.5 Natural Stop Position ............................................................. 31
ix
Chapter 4 Transition from a Thin Lens to a Thick Lens ........... 33
4.1 Introduction ............................................................................ 33

4.2 Adding a Thickness and Changing
the Second Surface Radius..................................................... 33
4.2.1 VIS singlet with n = 1.5.............................................. 34
4.2.2 MWIR singlet with n = 3.4 ......................................... 35
4.2.3 LWIR singlet with n = 4 ............................................. 35
4.3 Change of Spherical Aberration with Added Thickness ........ 35
Chapter 5 Achromats .................................................................. 39
5.1 Introduction ............................................................................ 39

5.2 Thin Achromat for the VIS Spectrum,
On-Axis Performance ............................................................ 39
5.2.1 Adding a field to the on-axis corrected achromat....... 43
5.2.2 Optimizing .................................................................. 45
5.3 Smith’s Method of Determining the Surface Shapes ............. 45
5.3.1 Curvatures and inverse object distances ..................... 45
5.3.2 The K-functions .......................................................... 46
5.3.3 The calculations .......................................................... 47
5.4 Achromat for the MWIR Region ........................................... 48
5.5 Achromat for the LWIR Region ............................................ 50
5.6 Diamond-Turned Hybrid ....................................................... 50
5.6.1 Hybrid for the MWIR region ...................................... 51
5.6.1.1 Basic lens shape ..................................................... 51
5.6.1.2 Aspherizing............................................................ 52
5.6.1.3 The diffractive phase profile.................................. 52
5.6.1.4 Comments .............................................................. 53
5.6.2 Useful nomograms...................................................... 55
Chapter 6 Systems with Two Separated Components ............. 57
6.1 Introduction ............................................................................ 57

6.2 Dialyte—An Air-Spaced Achromat ....................................... 57
6.2.1 Example for the MWIR region .................................. 59
6.2.1.1 Aspheric deformation
coefficients of surface 3.............................................. 61
6.2.1.2 Phase coefficients .................................................. 61
6.2.1.3 Step height at the zone transition ........................... 62
6.3 Telephoto and Reversed Telephoto ....................................... 62
6.3.1 Examples for the MWIR and LWIR regions .............. 63
x
Chapter 7 From an Air-Spaced Doublet to a Triplet .................. 65
7.1 Introduction ............................................................................ 65

7.2 Chromatic Aberration ............................................................ 66
7.3 Example, a Conventional Triplet ........................................... 66
7.4 Hybrid Petzval Objective with a Cold Stop ........................... 67
Chapter 8 A Hybrid for Two Wavelengths ................................. 69
8.1 Introduction ............................................................................ 69

8.2 Basic Lens Shape for the Long Wavelength .......................... 69
8.3 Required Diffractive Profile .................................................. 70
Chapter 9 Athermats ................................................................... 73
9.1 Introduction ............................................................................ 73

9.2 Focus Shift of a Refractive Element ...................................... 73
9.3 Athermalization with a Doublet ............................................. 74
9.4 Focus Shift of a Diffractive Lens ........................................... 74
9.5 Design Examples ................................................................... 78
9.5.1 Athermat with two elements
in an aluminum housing ............................................. 78
9.5.2 Hybrid athermat in an aluminum housing .................. 79
9.6 Impact of Housing Material .................................................. 80
9.7 Athermat for the CO2 Laser Line ........................................... 81
9.8 Athermalized Achromat ......................................................... 84
9.8.1 Three-element athermat in an aluminum housing ...... 85
9.8.2 Two-element athermat in an aluminum housing ........ 86
9.9 Effect of Quarter-Wave Limit without Athermalization........ 87
Chapter 10 The Ball Lens.............................................................. 89
10.1 Introduction ............................................................................ 89

10.2 Spherical Aberration .............................................................. 90
10.3 Coma 91
10.4 Astigmatism ........................................................................... 92
Chapter 11 Seidel and the Pegel Diagrams ................................. 95
11.1 Introduction ............................................................................ 95

11.2 Triplet for the LWIR Region ................................................. 95
11.2.1 Surface contributions ................................................ 96
11.2.2 Pegel diagram for the triplet ..................................... 96
11.2.3 Remarks to field curvature ........................................ 97
xi
11.3 Cassegrain Version with a Maksutov-Mangin Mirror
Combination for the LWIR Region ....................................... 98
11.3.1 Pegel diagram ........................................................... 98
Chapter 12 The Single-Imaging Mirror .........................................101
12.1 Introduction ............................................................................ 101

12.2 Spherical Mirror ..................................................................... 101
12.3 Toroidal Mirrors .................................................................... 104
12.4 Examples ................................................................................ 104
12.4.1 Spherical mirror ........................................................ 106
12.4.1.1 Additional comments ............................................. 106
12.4.2 Toroidal mirror ......................................................... 107
12.5 Parabolic Mirror ..................................................................... 107
12.6 Manufacturing Remarks......................................................... 109
12.7 Mangin Mirror ....................................................................... 109
Chapter 13 Eight Single Optical Elements

as Imaging Objectives ............................................... 113
13.1 Introduction ............................................................................ 113

13.2 Diffraction Limit .................................................................... 113
13.3 Eight Chosen Configurations ................................................. 113
13.4 Shapes of the Elements .......................................................... 115
13.5 Aberrations............................................................................. 116
13.6 Examples ................................................................................ 117
Chapter 14 A Progression of Performance with

an Increase in Lens Complexity ................................ 119
14.1 Objectives .............................................................................. 119
Chapter 15 Two-Mirror Systems as Telescope

and Microscope Objectives .......................................121
15.1 Introduction ............................................................................ 121

15.2 Basic Cassegrain Telescope Layout....................................... 122
15.2.1 Equations ................................................................. 123
15.3 Cassegrain with Two Spherical Mirrors ................................ 123
15.4 Classic Cassegrain System..................................................... 124
15.5 Dall-Kirkham Arrangement ................................................... 124
15.6 Ritchey–Chretien Configuration ............................................ 125
15.7 Examples ................................................................................ 125
15.8 Cassegrain with Mangin as a Secondary Reflector ................ 126
15.9 Gregorian Telescope .............................................................. 127
xii
15.10 Gregorian Microscope Objective ........................................... 130
15.11 Two Schwarzschild Objectives .............................................. 131
15.11.1 Prescription for the classic configuration
for an object at infinity .......................................... 133
15.11.1.1 Brief historic remark about the golden ratio ........ 134
15.11.2 Prescription for the inverted configuration
for an object at infinity .......................................... 134
15.12 Solid Microscope Objective ................................................... 136
Chapter 16 The Plane-Parallel Plate .............................................137
16.1 Introduction ............................................................................ 137

16.2 Aberrations............................................................................. 138
16.2.1 Examples .................................................................... 140
16.3 Shift of Image ........................................................................ 141
16.4 Tilted Plate ............................................................................. 143
16.4.1 Lateral displacement ................................................... 144
16.5 Two Tilted Plates ................................................................... 144
Chapter 17 MTF, Limits, and Pixel Sizes ..................................... 147
17.1 Introduction ............................................................................ 147

17.2 Optical Modulation Transfer Function .................................. 147
17.3 Focal Plane Array .................................................................. 148
Chapter 18 Details of a Hybrid Lens ............................................151
18.1 Introduction ............................................................................ 151

18.2 Hybrid .................................................................................... 151
18.3 Coordinates of the Combined Surface ................................... 153
Chapter 19 From the Höegh Meniscus

to Double Anastigmats ..............................................157
19.1 Introduction ............................................................................ 157

19.2 Höegh Meniscus .................................................................... 157
19.2.1 Approach and design method.................................... 158
19.3 Hypergon Lens ....................................................................... 160
19.4 Achromatic Double Lens ....................................................... 161
19.5 Double Anastigmats ............................................................... 165
Index………………… ........................................................................169
xiii
Preface
This book is written for engineers and scientists who have some experience in the
field of optics and want to know more about the details and derivations of
equations used in optical design. Such knowledge is especially valuable in the
layout stages of an optical system, when the question is “Where shall I begin?”
The other question may be “How come?” instead of just using a given equation.
The book begins with the derivation of the fundamental law of geometrical
optics, Snell’s law of refraction, and states the paraxial ray trace equations. The
following discussions are organized by subjects, starting with a thin lens and
progressing to increasingly more sophisticated components and multi-element
systems. Each subject is covered in depth to provide a good understanding for
performance and limitations. The often ignored effects of plane-parallel plates
are included as a separate subject.
While the text is based on general optical laws, special emphasis has been
placed on the two major infrared regions, the mid-wave (MWIR) and the long-
wave (LWIR). This is particularly apparent in the discussion about diffractive
hybrids, which have found their place in these long wavelengths areas for the
correction of chromatic aberrations and athermalization. Comments relating to
single-point diamond turning have been included because this process is
predominantly used to produce optical elements for the infrared regions.
The final subject gradually leads the reader from a single element as an
imaging objective for the visible spectrum, the historic Höegh, to a four-element
anastigmat. This is done by successively adding and shaping elements and
selecting suitable glasses for aberration reduction.
Finally, I thank Tim Lamkins, Scott Schrum, and Gwen Weerts of SPIE for
their special support and editorial assistance.
Max J. Riedl
July 2009
xv
Chapter 1
Law of Refraction: The Foundation of
Geometrical Optics
1.1 Introduction
Snell’s law of refraction is the fundamental law that governs geometrical optics.
We begin, therefore, with the proof of this basic rule, as it has been verified by
Fermat. We then demonstrate how this surprisingly simple law can be applied to
graphical ray tracing. With the equations for paraxial ray tracing, we provide the
tool required for the initial optical design phase. These equations are sufficient to
determine the third-order aberrations, which will be used throughout the book.
1.2 Fermat’s Principle1

1.2.1 Historic remarks
Pierre de Fermat was a jurist and mathematician. He pursued his mathematical
avocation mostly for his own enjoyment. He formulated his famous theorem in
1657. It declares that light takes the path that requires the least time. His
reasoning led to the proof of the law of refraction, which Wilibrord Snel van
Royen found experimentally some 20 years earlier. This law of refraction is the
foundation of geometrical optics and is stated by
n' sin i'  n sin i , (1.1)
where n and n′ are the indices of refraction of the media before and after
refraction. The angle of incidence is i, and the angle of the ray is i′, relative to the
normal, after refraction.
1.2.2 Derivation
Travel time with reference to Fig. 1.1 is
PO OP'
t 
v1 v2
and
b2   a  x 
2
h2  x 2
t  ,
v1 v2
1
2 Chapter 1
Figure 1.1 Schematic for Fermat’s theory.
where v1 is the speed of light in medium 1 before refraction, and v2 is the speed
in medium 2 after refraction. We differentiate this expression with respect to x
and set the result equal to zero.
dt x a  x
   0,
dx v1 h 2  x 2 v b 2   a  x 2
2
which means
x

a  x .
v1 PO v2 PO'
Law of Refraction: The Foundation of Geometric Optics 3
With
x a  x sin i sin i'

 sin i, and  sin i' , we get  .
PO OP' v1 v2
Since the index of refraction is defined as
vvacuum v1
n'   ,
vmedium v2
we finally obtain
1
sin i'  sin i ,
n'
which is, of course, Snell’s law with n = 1 (air).
1.3 Snell and the Lens

To apply Snell’s law to the refraction on a lens surface, we simply place a
tangent at point P where the ray meets the first lens surface. The angles before
and after refraction are then i and i′ as indicated in Fig. 1.2.
This simple procedure covers flat, spherical, and aspheric surfaces. It also
serves for calculations of chromatic aberrations because the index of refraction is
a function of wavelength.
Figure 1.2 Snell’s law applied to a curved lens surface.

4 Chapter 1
1.4 Graphical Ray Tracing

The procedure for graphical ray tracing is demonstrated in Fig. 1.3, where a
meridional ray is traced through a planoconvex lens with a refractive index of
1.5. The object is located at infinity. Therefore, the ray entering the lens is
parallel to the optical axis.
Draw line 1 from center C (center of the first surface radius) through point
P1, the point where the entering ray meets the first surface. Draw arcs 2 and 3
around point P1. The scales of the radii are chosen in proportion to the media
indices before and after the refraction. Our lens is surrounded by air with n1  1 ,
and the index of the lens material is n'1  n2  1.5 . The ratio between the radii 2
and 3 is therefore 1 to 1.5. Draw line 4 parallel to line 1 through the point where
arc 2 intersects the entering ray. From the intersection of arc 3 and line 4, draw a
line through point P1 and extend it until it crosses the second lens surface. This
point is marked P2 in the figure. Repeat the steps for the second lens surface to
find the direction of the exiting ray. The focus location F of the lens can be
established by tracing a ray near the optical axis, the paraxial region.
The lateral distance from the optical axis and the point where the marginal
ray passes through the focal plane is the transverse spherical aberration. The
distance from the crossover of the ray on the optical axis is the longitudinal
spherical aberration. While this approach may not appear to be too practical, it
gives a good insight and demonstrates the principle of refraction at lens surfaces.
Figure 1.4 serves for additional explanation of the relationships.
Figure 1.3 Procedure of graphical ray tracing.

Figure 1.4 Principle of graphical ray tracing, which is also suitable for aspheric
surfaces.
1.5 Paraxial Ray Tracing

Snell’s law was stated with Eq. (1.1) as n 'sin i '  n sin i . The sine of an angle can
be expressed by
i3 i5 i7
sin i  i     ...... . (1.2)
3! 5! 7!
For small angles,
sin i  i , (1.3)
and Snell’s law reduces to
n'i'  ni . (1.4)
This is accurate to the sixth decimal place if the angles are not larger than about
41 minutes of arc. Ray tracing based on, but not limited to, this narrow region is
referred to as paraxial ray tracing. The simple paraxial ray-trace equations are
exceedingly useful in lens design. The image location is determined with these
equations. Deviations from the paraxial image position are the measures for
aberrations. By including the second term of the sine series, the so-called third-
order aberrations, known as Seidel aberrations, can be determined. This will be
demonstrated throughout the following chapters.
6 Chapter 1
Figure 1.5 Symbols and sign conventions for paraxial ray-trace equations: (a) an
axial ray and (b) an oblique ray.
1.5.1 Equations, symbols, and sign conventions

The symbols and sign convention used in the equations are identified in Figs.
1.5(a) and 1.5(b). The sign for a surface radius is positive if its center lies to the
right. If it lies to the left, it is negative.
Opening
The distance of an object from the first lens surface is l1. For an axial ray,
y1
u1   , (1.5)
l1
for an oblique ray,

y1  h  l1u1 , (1.6)
and for an object located at infinity,
u1  0, and y1  h . (1.7)
Axial ray [Figure 1.5(a)]

If the object is located 100 units to the left of the lens, l1  100 . With a ray
height of 10 units at the first surface of the lens, y1  10 , according to Eq. (1.5),
the slope of the axial ray is
y1 10
u1     0.1 .
l1  100 
Oblique ray [Figure 1.5(b)]

If the object height h  15 units, and the slope of this oblique ray u1  0.075 ,
the ray height at the first surface of the lens is
y1  h  l1u1  15  (100)  (0.075)  7.5 units.
Transfer to the next surface

1
n'i u'i  ni ui  yi ci  n'i  ni  with ci  (1.8)
Ri
yi1  yi  t 'i u 'n (1.9)
Equations (1.8) and (1.9) are repeatedly applied for the i surfaces in the
optical train.
Closing
For the axial ray,
yi
l'i   , (1.10)
u 'i
for the oblique ray,
h'  yi  l'i u'i , (1.11)
and for magnification,

8 Chapter 1
u1 h'
m  . (1.12)
u'i h
For mirror surfaces n  1 (reversal of ray direction).
Example
Given: A lens with the surface radii R1 = 50 (c1 =1/50 = 0.02) and R2 = −50 (c2 =
−0.02). t'1  10 .
n1  n2'  1 (air), n'1  n2  1.5, and l  200
Trace an axial ray with y1  20, and an oblique ray with h  12 and y1  10 .
1. Axial ray
y1 20
u1     0.1
l1 200
n'1u'1  n1u1  y1c1  n'1  n1   1  0.1  20  0.02  1.5  1  0.1
0.1
u'1   0.066667
1.5
y2  y1  t'1u'1  20  10   0.066667   19.333333
n'2u'2  n2u2  y2 c2  n'2  n2 

 1.5   0.066667   19.333333   0.02   1  1.5   0.293333  u'2
 y2 19.333333
l'2    65.909091
u'2 0.293333
u1 0.1
m   0.340909
u'2 0.293333
2. Oblique ray
h  y1 25  10
u1    0.075
l1 200
n'1u'1  n1u1  y1c1  n'1  n1   1   0.075   10  0.02  1.5  1  0.175

0.175
u'1   0.116667
1.5
y2  y1  t'1u'1  10  10   0.116667   8.833333
n'2u'2  n2u2  y2 c2  n'2  n2 

 1.5   0.116667   8.833333   0.02   1  1.5   0.263334  u'2
h'  mh  0.340909  25  8.522727
An oblique ray from the edge of the object through the center of the aperture
stop is called the chief, or principle, ray. An axial ray from the axial point of the
object through the edge of the aperture stop is the marginal axial ray. To
determine the third-order aberration contributions, only these two paraxial rays
have to be traced. Ray-trace equations in addition to the ones presented here can
be found in Ref. 2.
References
1. J. R. Meyer-Arendt, Introduction to Classical and Mirror Optics, Prentice
Hall, Inc., Englewood Cliffs, New Jersey (1984).
2. W. J. Smith, Modern Optical Engineering, 4th Ed., McGraw-Hill, New York
(2008).
Chapter 2
Best Shape for a Thin Lens
2.1 Concept of Thin Lens
The “thin lens” concept is an extremely useful one for quick preliminary layout
calculations and initial performance analyses. “Thin” means that the thickness of
the lens is ignored in the calculations.
The power  of a lens is the reciprocal of its focal length f. The reciprocal of
a surface radius R is called the surface curvature c. With t being the lens
thickness and n the index of refraction of the lens material, we can state
1 
    n  1  c1  c2 
 n  1 c c t  . (2.1)
1 2 
f  n 
Neglecting the thickness t yields the power of the thin lens
   n  1 c1  c2    n  1 c . (2.2)
In this equation, c is called the net curvature and is the difference between the
front and rear surface curvatures, i.e., c  c1  c2 . This indicates that
1
Figure 2.1 Bending a lens affects its spherical aberration but not its focal length.
Shown is the effect for a germanium lens with an index of n = 4.
11
12 Chapter 2
the power of a thin lens remains the same as long as the net curvature is kept
constant. Changing the curvatures but maintaining their difference is called lens
bending. This is a powerful tool because the shape of a lens primarily influences
its spherical aberration, as indicated in Fig. 2.1. The relation between lens shape
and spherical aberration of a thin germanium lens for an object located at infinity
is plotted. The ratio of the first curvature c1 and the net curvature c is known as
the shape factor K. We plot spherical aberration as a function of the lens shape K
and identify the surface radii of the lens. Because germanium has an index of
refraction of n = 4 at a wavelength  = 10 µm, the result for the best shape is
amazingly simple and easy to remember. The front surface radius is equal to the
focal length of the lens, and the radius of the rear surface is 1.5 times the focal
length.
2.2 Object at Any Position

Figure 2.2 shows the case where the object is located at a finite distance from the
lens and identifies object and image distances, l and l′, the semi-aperture y of the
lens, and the slope u′ of the exiting ray.
Figure 2.2 Thin lens with object located at a finite distance.
Figure 2.3 The marginal ray penetrates the image plane a distance away from
the optical axis. This lateral distance is called the transverse spherical aberration
TSC. See also Figure 2.4.
Best Shape for a Thin Lens 13
To derive the general equations for the best shape surface radii, when the
object is located at any position, we begin the derivation with the expression for
the third-order thin-lens transverse spherical-aberration contribution for any
conjugate2:
y4
TSC 
u'k
 G1c 3  G2c 2c1  G3c 2v  G4cc12  G5cc1v  G6cv 2  . (2.3)
In Eq. (2.3), the G-sums are functions of the optical material’s index of refraction
n. They are
n 2  n  1  2n  1 n  1 ,  3n  1 n  1
G1  , G2  G3 
2 2 2
G4 
 n  2  n  1 , G5 
2  n  1 n  1
, G6 
 3n  2  n  1
2n n 2n
G7 
 2n  1 n  1 , and G8 
n  n  1
.
2n 2
The reciprocal object distance v is defined in Eq. (2.3) as v  (1 / l ) , where l is

the object’s distance from the lens.
Differentiating Eq. (2.3) with respect to the first curvature c1 yields
dTSC y 4
dc1

u'k
 G2c 2  2G4cc1  G5cv  .
To find the optimum, we set this expression to zero and solve for c1,
G2c  G5v
c1  . (2.4)
2G4
The net curvature c  c1  c2    n  1 , which is derived from the thin lens

power equation:    n  1 c1  c2    n  1 c . Inserting this and the G-sums into
Eq. (2.4) leads to
n  2n  1 2  n  1
c1   v, (2.5)
2  n  2  n  1  n  2
14 Chapter 2
where c1  1 / R1 , the reciprocal first surface radius;   1 / f , the power or

reciprocal focal length of the lens; and v   1 / l  , the reciprocal object distance
from the lens, as stated above.
It is interesting to observe that the first term of Eq. (2.5) states the surface
curvature c1 for an object at infinity because, for that case, v = 0.
To find the curvature for the second surface, we use the relation
c2  c1    n  1 .
 n  2n  1   2  n  1    
c2    v   ,
 2  n  2  n  1    n  2     n  1 
which reduces to
 n  2n  1  4   2  n  1 
c2     v. (2.6)
 2  n  2  n  1    n  2  
Here, too, the first term is for an object at infinity. Considering that the inverse
object distance can be stated by    m / (m  1)  , where m is the magnification,
we insert this relation into Eqs. (2.5) and (2.6), to form
 n  2n  1 2  n  1 m 
c1     (2.7)
 2  n  2  n  1  n  2   m  1 
and
 n  2n  1  4 2  n  1 m 
c2     . (2.8)
 2  n  2  n  1  n  2   m  1 
To demonstrate the usefulness of these expressions, we analyze a lens with a
focal length of 100 mm    1/f  0.01 mm -1  , made from germanium with
n = 4. We choose the magnification to be m = –2.
The first surface curvature is then

 4   2  4  1 2   4  1  2  
c1       0.01
 2   4  2    4  1  4  2   2  1 
and
c1  1  1.666667  0.666667   0.01  0.001111 .
The first surface radius is

1 1
R1    900 mm .
c1 0.001111
The second surface curvature, according to Eq. (2.8), is

 4   2  4  1  4 2   4  1  2  
c2       0.01
 2   4  2    4  1  4  2   2  1 
and
c2   0.666667  1.666667  0.666667   0.01  0.0044444 .
The second surface radius is
1 1
R2    225 mm .
c2 0.0044444
The G-sums for n = 4 are
G1  24.00, G2  13.50, G3  19.50, G4  2.25, G5  7.50, and G6  5.25.
The net curvature c  c1  c2  0.0011111   0.0044444   0.0033333 and the

inverse object distance are
 m   2 
v      0.01  0.0066667 .
 m 1   2  1 
For a relative aperture of f/2, the semi-aperture y = 25 mm. The final slope angle
u'k      v  y    0.01  0.0066667   25  0.0833333 .
Now, we can insert all of these numbers into Eq. (2.3) and find the minimum
transverse third-order spherical aberration for this thin lens to be
254
TSC  [24  0.0033333  13.5  0.0033332  (0.001111)
0.083333
19.5  0.0033332  0.006667  2.25  0.003333  (0.001111)2
7.5  0.003333  (0.001111)  0.006667  5.25  0.003333  0.006667 2 ]
=  0.994859.
TSC  1 mm .
16 Chapter 2
Expressing the transverse spherical aberration directly as a function of the

index n, the semiaperture y, the power of the lens , the inverse object distance v,
and the G-sums transforms Eq. (2.3) into
y 3n  4n  1 2  4  n  12    v  v  .
TSC   (2.9)
8  n  1  n  2    v   
2
For our case, this results in
 253  0.01  4 
TSC    
 8  4  1  4  2  0.01  0.006666  
2
  4  4  1 0.012  4  4  1  0.01  0.006666  0.006666  ,

2
 
and
TSC    434.02691   0.0015  0.0008  0.998 mm  1 mm.
2.3 Object at Infinity with Added Field of View

In this section, we apply the third-order thin-lens aberration equations to
determine the blur spots contributed by spherical and chromatic aberrations, as
well as coma and astigmatism. Comparing the results with an actual ray trace will
indicate how valuable these thin-lens aberration expressions are. Of great interest
in this discussion is also the best image (detector) location.
2.3.1 Spherical aberration

For an object located at infinity, the image is formed at the focal plane of the
lens. As mentioned above, the equations for a singlet, shaped for minimum
spherical aberration, become simpler.1 The first surface radius is
 2  n  2  n  1 
R1   f, (2.10)
 n  2n  1 
and the second surface radius is
 2  n  2  n  1 
R2   f. (2.11)
 n  2n  1  4 
Since v = 0, Eq. (2.9) reduces to

y 3 n  4n  1 2 n  4n  1 f
TSC    . (2.12)
8  n  1  n  2  64  n  1  n  2  f / # 
2 2 3
The blur spot is half the size of TSC after refocusing from the paraxial focus
by    3 / 2  TSC  f / #  . Applied to our f/2, 100-mm focal-length germanium
lens set for infinity, we obtain
 2  n  2  n  1  2  4  2  4  1 36
R1   f   100   100  100 mm
 n  2n  1  4  2  4  1 36
and
 2  n  2  n  1  2  4  2  4  1 36
R2   f   100  100  150 mm .*
 n  2n  1  4  4  2  4  1  4 24
n  4n  1 f 4  15  100
TSC     0.217014 mm .
64  n  1  n  2  f /#  64  9  6  23
2 3
The minimum blur spot size is Bmin  TSC 2  0.217 2  0.1085 mm when
refocused by    3 2  TSC  f /#   1.5   0.217   2  0.651 mm. The best
image location is, therefore, 100 – 0.651 = 99.349 mm from the last lens surface.
Details are shown in Fig. 2.4.
Figure 2.4 Location and size of the minimum blur spot due to spherical
aberration, which is called “the circle of least confusion.”
*
Also see Fig. 2.1.
18 Chapter 2
2.3.2 Chromatic aberration

The blur spot size due to axial chromatic aberration for a thin lens is expressed
by
f
Bchrom  , (2.13)
2V  f / # 
where V is the inverse relative dispersion of the lens material (also known as
Abbe number, named after Ernest Abbe, who introduced this concept). In the
visible spectrum, V   nd  1  nF  nC  , where the indices of refraction refer to
the Fraunhofer absorption lines. In the infrared regions, V   nM  1  nS  nL  ,
with reference to the middle, short, and long wavelengths of the covered spectral
band. For our germanium lens in the LWIR region (8–12 µm), V ≈ 800.
Therefore, Bchrom  100  2  800  2   0.03125 mm for our lens. Note how mild
the chromatic aberration is with such a large Abbe number. In the visible
spectrum, the V values are lower by more than a factor of 10. As will be
discussed in Chapter 4, this attribute is the main justification for correcting
chromatic aberration in the infrared regions with diffractive phase profiles.
The expected on-axis blur spot, caused by spherical and chromatic

aberrations, according to the thin-lens aberration equations, is
Baxial  Bspher  Bchrom  0.10851  0.03125  0.140 mm. Figure 2.5 shows the spot
diagram of a thick lens, which was obtained with a computer ray-trace program.
Figure 2.5 Computer ray trace confirms validity of thin-lens calculations for on-
axis aberrations (spherical and chromatic).
Figure 2.6 Thin lens with added field for object at infinity. The principal ray is an
oblique ray that goes through the center of the aperture stop, which is located
here at the lens.
2.3.3 Coma
Figure 2.6 shows the layout of a thin lens for an object located at infinity and
identifies the image height h′ and the half-field angle up.
The thin-lens expression for sagittal coma is2
CC   h' y 2 c  0.25G5c1  G8 c  . (2.14)
If we insert the curvatures c and c1 from the thin lens (shaped for minimum
spherical aberration) and the expressions for the G-sums G5 and G8, we get
up f
CC  . (2.15)
16  n  2  f/# 
2
With a half field angle of up = 0.1 radian, the sagittal coma blur for our lens
becomes
0.1 100
Bcoma  CC   0.026042 mm .
16   4  2   22
2.3.4 Astigmatism
Another off-axis aberration is astigmatism, which introduces a blur spot size of 1:
u 2p f
Bastig  2TAC  . (2.16)
2  f / #
20 Chapter 2
Figure 2.7 Comparison of predicted blur spots with third-order thin-lens

aberration equations and the actual sizes for the thick germanium lens.
Figure 2.8 Balanced image position for equal blur spot sizes of the f/2, 100-mm
focal-length germanium lens for the LWIR spectrum with a half field of 5.73 deg.
The location of this blur (circle of least confusion) is located halfway

between the sagittal and tangential image shells. This distance is 0.5u 2p f . The
blur spot size due to astigmatism for our example is
0.12 100
Bastig   0.250 mm.
2 2
2.3.5 Total blur-spot size

We add all four effects together and get the approximate size of the total blur at
the best location:
Btotal  Bspher  Bchrom  Bcoma  Bastig

Btotal  0.10851  0.03125  0.02604  0.25000  0.42 mm.
This does not take into account the effects of field curvature. This can be seen in
Fig. 2.7, which is from a real ray trace of the thick lens. The off-axis blur is
somewhat larger than the predicted 0.42 mm. By slightly refocusing, the total
blur size is about 0.460 mm in diameter. The choice of best position depends on
the application. A 0.460-mm square detector element is large enough to accept all
the energy from a point source within the stated field of view.
For completeness, the best focus position relative to the paraxial focal plane
for the example is illustrated in Fig. 2.8.
2.4 Special Case

A closer look at Eq. (2.11) reveals that the second surface of a lens becomes a
flat when n  2n  1  4 . This is the case when n = 1.686141. An optical material
with an index of refraction very close to this value is sapphire at a wavelength of
3.7 µm with n = 1.687. In the visible spectrum there are the glasses SF5 (with nd
= 1.673) and LAKN13 with (nd = 1.694).
References
1. M. J. Riedl, Optical Design Fundamentals for Infrared Systems, SPIE Press,
Bellingham, Washington (2001).
(2008).
Chapter 3
Best Shapes for Multiple Thin Lenses,
Aspherizing, and the Natural Stop
Position
3.1 Introduction
Expanding on the previous case for a single lens in Chapter 2, we develop
expressions for elements of a multiple lens arrangement, made from the same
material and individually bent for minimum spherical aberration.
We then demonstrate the benefit of an aspheric surface. This is especially of

interest for elements used in the infrared spectrum because they can be routinely
machined with the process of single-point diamond turning.
Coma can be eliminated by properly shaping the single element or by placing

the aperture stop at a specific position, which is referred to as the natural stop
position. Details will be covered under this Chapter.
3.2 Best Shapes for Minimum Spherical

Aberration
To expand on the general case for a single lens with the object located at any
position as discussed under Chapter 2, one can develop the following expressions
if all elements of a multiple lens arrangement are made from the same material
and individually bent for minimum spherical aberration. The surface curvatures
of the elements are as follows:
For the first surface (front surface) of element j, curvature is
n  2n  1  4  n 2  1  j  1 2  n  1
c j1   v. (3.1)
2i  n  2  n  1  n  2
In this equation and in the following, j is the element number, and i is the total
number of elements in the set.
For the second surface (rear surface) of element j, curvature is
23
24 Chapter 3
n  2n  1  4  4  n 2  1  j  1 2  n  1
c j2   v. (3.2)
2i  n  2  n  1  n  2
The minimum spherical aberration is
y 3n
TSCi  
8i 2  n  1  n  2    v 
2
 2   2 j i  
  4n  1 2  4i 2  n  1  v    v   3  j  j  1   . (3.3)
  i j 1  
These expressions reduce to the forms for a single lens if we set j = i = 1. With v
= 0, the lens will be focused for an object at infinity.
Table 3.1 lists the radii and the spherical aberrations for an f/2, 100-mm-
focal-length lens, made from germanium with 1, 2, and 3 elements for an object
at infinity. Notice the change in shapes with the increase of elements for an
object at infinity.
Table 3.1 Radii and resulting transverse spherical aberrations of configurations

for an object at infinity.
i R11 R12 R21 R22 R31 R32 TSC

1 100 150 -0.217
2 200 300 75 85.714 +0.076
3 300 450 112.5 128.571 69.231 75 +0.130
Best Shapes for Multiple Thin Lenses, Aspherizing, and the Natural Stop Position 25
From Table 3.1, it can be seen that the single element is afflicted with a
negative TSC, which means that it has undercorrected spherical aberration. (The
marginal ray crosses the optical axis ahead of the paraxial image point.) With two
elements, the sign changes and the lens has overcorrected spherical aberration.
(The marginal ray crosses the optical axis behind the paraxial image). Three
elements amplify this trend. This means that the high index of germanium
“overshoots” the correction with only two elements.
Let us find the index for which the two-element configuration is zero, when
the object is located at infinity. For those conditions, Eq. (3.3) reduces to
y 3 n2  4n  1  16  n  12    1  .
TSC2   (3.4)
32  n  1  n  2 
2   4 
Setting this expression to zero leads to 4n  1  4  n  1 , for which the solution is

2
n = 2.5. That is close to the index for amorphous material transmitting infrared
radiation (AMTIR-1) for the mid-wave infrared (MWIR) with n4  2.514 , and for
the long-wave infrared (LWIR) where n10  2.498 . Also close to the desired value
is zinc selenide with n4  2.433 in the MWIR region, and n10  2.406 for the
LWIR band.
To find a material for which the TSC is zero when the system consists of
three elements, we proceed as above and get  4n  1  36  n  1 8 / 27 . This
2
leads to n = 1.75. That is the index for the rare earth glass LAFN7, for example,
with nd  1.7495 . For the LWIR spectrum, cesium iodide (CsI) has n10  1.74 ,
which is close to the desired value.
Table 3.2 on the next page shows the radii and the spherical aberrations for
lenses with a magnification of m  2 . The negative sign indicates that the image
is reversed, as demonstrated in Fig. 1.5(b). Notice how the shapes change for an
object at a finite distance.
It is worthwhile to point out here the advantage of an aspheric surface,

especially for the infrared spectrum, where the surface can be shaped by diamond
turning. It is an easy task to find the conic constant for the surface to be used to
eliminate (third-order) spherical aberration. We shall demonstrate the process for
a thin singlet, by aspherizing first the front surface and then the rear surface.
26 Chapter 3
Table 3.2 Radii and resulting transverse spherical aberrations of configurations

for finite conjugates (magnification m = –2).
i R11 R12 R21 R22 R31 R32 TSC

1 -900 -225 -0.998
2 -163.636 -128.571 450 1,800 -0,119
3 -128.574 -112.5 -450 -300 300 450 -0.043
3.3 Aspherizing a Singlet to Eliminate

Spherical Aberration
With reference to Eq. (2.3), the transverse spherical aberration of a thin lens is
y 4c
TSC 
u'
 G1c 2  G2cc1  G4c12  , (3.5)
when the object is located at infinity. With
n  2n  1 
c1  ,
2  n  2  n  1

c ,
 n  1
u '   y ,
and G1 , G2 , G4 , as identified in Sec. 2.2, and the object is at any position, Eq.
(3.5) becomes
y 3 n  4n  1 2
TSCmin spher   . (3.6)
8  n  1  n  2 
2
To get rid of spherical aberration, we aspherize one surface of the lens and set the
contribution from the asphere equal to the spherical aberration contribution from
the lens with spherical surfaces, but opposite in sign. In other words, we write
TSCasphere  TSClens best shape .
If we limit the correction to changing one spherical surface to a conic

section, the spherical aberration contribution of such a surface can be expressed
by
TSCasphere 
 n ' n  fy 3 c3 , (3.7)
2
where  is the conic constant, n and n′ are the indices of the media before and
after refraction, f is the focal length of the element, y is the semi-diameter of the
lens, and c is the curvature of the surface to be aspherized. This leads to a conic
constant for the first surface of
2 R13TSC
1  . (3.8)
 n  1 f y13
If the second surface is to be aspherized, the conic constant for that surface is
2 R23TSC
2   . (3.9)
 n  1 f y23
Example
For a 100-mm-focal-length lens (made from silicon) with an f/2 speed, with n =
3.4,
y 3 n  4n  1 2 253  3.4   4  3.4  1  0.012

TSCmin spher     0.269007.
8  n  1  n  2  8   3.4  1   3.4  2 
2 2
The curvature of the first surface is
n  2n  1  3.4   2  3.4  1  0.01

c1    0.010231.
2  n  2  n  1 2   3.4  2    3.4  1
28 Chapter 3
For the second surface,
 0.01
c2  c1   0.010231   0.006065 .
 n  1  3.4  1
The radii are therefore R1  1 / c1  1 / 0.010231  97.742156 mm and

R2  1 / c2  1 / 0.006065  164.885496 mm. Aspherizing the first surface leads to
a conic constant of
2 R13TSC 2  97.7421563   0.269007 

1    0.13397 .
 n  1 f y13  3.4  1 100  253
If we decide to leave the first surface a sphere and aspherize the second, we
obtain a conic constant for that surface of
2 R23TSC 2  164.8854963   0.269007 

2     0.643146 .
 n  1 f y23  3.4  1  100  253
In the first case, the aspheric surface is a prolate ellipsoid, in the second it is an
oblate ellipsoid.
3.4 Correcting Coma and Spherical Aberration

If spherical aberration and coma are to be eliminated, the procedure is to remove
coma first by selecting the proper curvatures of the lens, and then to aspherize
one of the two surfaces.
3.4.1 Eliminating coma

For an object located at infinity, sagittal coma of a thin lens is stated by Eq.
(2.14) and repeated here:
CC   h' y 2 c  0.25G5c1  G8 c  . (3.10)
Setting this expression to zero yields the curvature of the first surface as
4G8c
c1  . (3.11)
G5
The G-sums have been identified previously, and c    n  1 . After insertion,

Eq. (3.11) changes to
n2
c1  . (3.12)
n 2
 1
The radius is the reciprocal of the curvature, i.e.,
R1 
n 2
 1
f. (3.13)
n2
The second surface curvature is
 n  n  1  1
c2  c1  c    (3.14)
 n2  1
and
R2 
n 2
 1
f . (3.15)
 n  n  1  1
3.4.2 Eliminating spherical aberration

As we know from Eq. (3.5), the transverse spherical aberration of a thin lens for
an object located at infinity is
y 4c
TSC 
u'
 G1c 2  G2cc1  G4c12  . (3.16)
When c1  n 2   n 2  1 from Eq. (3.12), c    n  1 , u'   y , and the G-sums

are inserted, Eq. (3.16) turns into
y 3 n3
TSCcoma=0   . (3.17)
2  n 2  1 f 2
2
To apply this formula to our silicon lens with y = 25, n = 3.4, and f = 100, we get
253  3.43
TSCcoma=0    0.275359 .
2   3.42  1  1002
2
30 Chapter 3
We eliminate spherical aberration again with an asphere. This time we have to set
TSCasphere  TSCcoma=0  0.275359 .
The first surface radius is
R1 
n 2
 1
f 
 3.4 2
 1
 100  91.349481 ,
2
n 3.42
and the conic constant for the first surface is then
2  91.3494813   0.275359 
κ1   0.111948 .
 3.4  1  100  253
This equation indicates the shape of a prolate ellipsoid. The second surface radius
is
R2 
n 2
 1
f 
3.4 2
 1
100  147.486034 .
 n  n  1  1 3.4   3.4  1  1
If we desire to aspherize the second surface instead of the first, the conic constant
for that surface is
2  147.4860343  0.275359
κ2   0.471140 ,
 3.4  1  100  253
which identifies again the shape of an oblate ellipsoid.
Remark
The same exercises can be repeated for the other monochromatic aberrations and
extended to include the effects of the aperture stop location. Such exercises give
much insight into the behavior of optical elements. We limit ourselves to the case
of a single lens, shaped for minimum spherical aberration, and determine the stop
location for which coma vanishes. This stop location is called the natural stop
position.
3.5 Natural Stop Position

We begin with the so-called stop-shift equation, which leads to the elimination of
coma. If the aperture stop is not located at the lens, the thin-lens, third-order
sagittal coma is
yp
CC*  CC  TSC , (3.18)
y
whereby it is interesting to note that coma cannot be eliminated if there is no

spherical aberration.
As we know, the thin-lens, third-order sagittal coma is expressed by
CC   h' y 2 c  0.25G5c1  G8 c  , (3.19)
with h'  u p f , and c  1/  n  1 f  . The curvature of the first surface for a
best-shaped lens (one with minimum spherical aberration) is
n  2n  1
c1  .
2  n  2  n  1 f
Figure 3.1 Stop position.

32 Chapter 3
With that, the sagittal coma for an object located at infinity is
up f
CC  . (3.20)
16  n  2  f / # 
2
The transverse spherical aberration is for that case
n  4n  1 f
TSC   . (3.21)
64  n  1  n  2  f / # 
2 3
From Fig. 3.1, it can be seen that y p  u p lstop . (lstop is defined as negative
because it is measured from the first lens surface). Eliminating coma requires that
CC  y pTSC y  0 . If the lens has been shaped for minimum spherical
aberration, inserting the relations from above leads to the simple expression for
the natural stop position,
4  n  1 y  f / # 
2
lstop   . (3.22)
n  4n  1
This reduces for our three examples with y = 25 and f/# = 2 to
200  n  1
2
lstop   .
n  4n  1
Table 3.3 lists the results.
Table 3.3 Natural stop position for three thin lenses shaped for minimum
spherical aberration, at which third-order coma vanishes.
Index n 1.5 (VIS) 3.4 (MWIR) 4.0 (LWIR)

Stop position lstop -6.666 mm -26.891 mm -30.000 mm
Chapter 4
Transition from a Thin Lens
to a Thick Lens
4.1 Introduction
It was shown under Chapters 2 and 3 that the “thin lens” concept is an extremely
useful one for preliminary calculations and analysis as well as a general design
tool. Determining the difference in performance between that fictitious thin lens
and a realistic thick lens is an interesting exercise which confirms the statement
above.
We begin our exercise with a thin lens shaped for minimum spherical
aberration, add a reasonable thickness, and modify only the rear surface radius to
maintain the focal length of the lens. We then calculate the spherical aberration
for both cases and compare the results for three f/2, 100-mm-focal-length lenses.
The first lens is made from glass with an index of refraction of n = 1.5 for the
visible spectrum (VIS). The second is for the infrared mid-wave region (MWIR,
3–5 µm), made from silicon with n = 3.4. The third singlet is made from
germanium, n = 4, for the application in the infrared long-wave region (LWIR,
8–12 µm).
4.2 Adding a Thickness and Changing the

Second Surface Radius
The focal length for a thin lens (one for which its thickness is assumed to be
zero) is expressed by
1 1 1 
  n  1    . (4.1)
f  R1 R2 
For a thick lens, where the thickness of the lens is accounted for, the equation is
stated as
1 1 1  n  1 t 
  n  1    , (4.2)
f  R1 R2 nR1R2 
with t being the thickness of the lens. Solving for R2 results in
33
34 Chapter 4
 n  1  n  1 t  nR1  f
R2  . (4.3)
n  R1   n  1 f 
We recall that the radius of a thin lens, shaped for minimum spherical
aberration, is
 2  n  2  n  1 
R1   f. (4.4)
 n  2n  1 
Inserting Eq. (4.4) into Eq. (4.3) leads to the rear radius of a thick lens, shaped
for minimum spherical aberration:
 n  1  2  n  2  f   2n  1 t 
R2  . (4.5)
 n  2n  1  2  n  2  
4.2.1 VIS singlet with n = 1.5

The radii for the thin lens are
 2  n  2  n  1   2  1.5  2   1.5  1 
R1    f    100  58.3333 mm
 n  2n  1   1.5   2  1.5  1 
and
 2  n  2  n  1   2  1.5  2   1.5  1 
R2    f    100  350mm .
 n  2n  1  4   1.5   2  1.5  1  4 
We add a thickness of 10 mm and solve for the second surface radius using Eq.
(4.5):
 n  1  2  n  2  f   2n  1 t 
R2 
 n  2n  1  2  n  2  
1.5  1   2  1.5  2  100   2 1.5  1 10
  330.00 mm.
1.5   2 1.5  1  2  1.5  2  
These results can be confirmed by using the ray trace equations [Eqs. (1.7) and
(1.8)] and solving for the curvature c2 . We demonstrate this for the thick lens.
Transition from a Thin Lens to a Thick Lens 35
With reference to Fig. 1.5(a), u1  0 because the object is located at infinity,

and n1  n '2  1 because the lens is surrounded by air, for which the index of
refraction is 1. We assume a ray height at the first surface of 10 units, i.e.,
y1  10 . The index of the lens material is n '1  n2  1.5 , and the front surface
curvature is c1  1 R1  1 58.333.....  0.017142857 . With that,
n'1u'1  n1u1  y1c1  n'1  n1   10  0.017142857  1.5  1 =  0.085714286,
and the angle after refraction at the first surface is
0.085714286
u'1  u2   0.05714285 .
1.5
With the lens thickness of t '1  10 , the ray height at the second surface is
y2  y1  t'1u'1  10  10   0.057142857   9.428571429 .
Recognizing that u'2   y1 f  10 100  0.1 , we rearrange Eq. (1.7) and
solve for the curvature of the second surface:
n2u2  n'2 u'2 1.5   0.057142857   1   0.1

c2    0.003030303 .
y2  n'2  n2  9.428571429  1  1.5 
The radius R2  1/ c2  1/  0.003030303  330.00 , which confirms the earlier

obtained result.
4.2.2 MWIR singlet with n = 3.4

For the thin lens, R1 = 97.737557 mm and R2 = 164.885496 mm. Adding a 5-mm
thickness to the element changes the second surface radius to R2 = 158.931299
mm.
4.2.3 LWIR singlet with n = 4

The thin-lens front surface radius R1 = 100.00 mm, and the rear surface radius R2
= 150.00 mm. Again, adding a 5-mm thickness to the element changes the second
surface radius to R2 = 144.375 mm.
4.3 Change of Spherical Aberration with Added

Thickness
We know from Eq. (2.12) that the third-order transverse spherical aberration for a
thin lens is
36 Chapter 4
n  4n  1 f
TSCthin lens   . (4.6)
64  n  1  n  2  f / # 
2 3
The contributions for our three lenses are summarized in Table 4.1.
Determining the spherical aberration of a thick lens is more elaborate and

based on the individual surface contributions.1 We follow the described
procedure and begin with
n2
 TSC  B i h' ,
n 1
2
n n (4.7)
whereby
n  n'  1
B y  u'  i  . (4.8)
2n' Inv
The optical invariant Inv  n'2u'2 h'  0.25h' for our f/2 lenses. This changes Eq.
(4.8) to
2n  n'  1
B y  u'  i  .
n'
From the paraxial ray trace equations nn' un'  un nn  yn cc  nn'  nn  ,

yn 1  yn  tn' un' , and in  cn yn  un , we find the ray height y2 and the angles u′
and i. Table 4.2 lists these values and the resulting spherical surface aberration
contributions.
Table 4.1 Spherical aberrations for three thin singlets in different spectral
regions.
Spectral region Transverse spherical aberration TSC (mm)

VIS -1.6741
MWIR -0.2690
LWIR -0.2170
Table 4.2 Ray trace and third-order surface aberration contributions for the VIS-
lens with n = 1.5
Surface y t′ u′ i B TSC
1 25 10 -0.142857 0.428575 -4.761967 -0.8747
2 23.57143 -0.250000 -0.214278 -16.415545 -0.7537
Transition from a Thin Lens to a Thick Lens 37
Figure 4.1 Undercorrected and overcorrected spherical aberration.
The total transverse spherical aberration of the thick lens is therefore
 TSC TSC 1  TSC2  1.6284 mm .
We compare this result with the thin-lens aberration and find
TSC   TSCthick  TSCthin  1.6284   1.6741  0.0457 mm .
As expected, the spherical aberration for the thick lens is somewhat less
because the added thickness contributes positive spherical aberration. Positive, or
overcorrected spherical aberration, exists when the marginal ray intercepts the
optical axis beyond the paraxial focal plane. Spherical aberration is called
negative, or undercorrected, when the marginal ray crosses the optical axis ahead
of the paraxial focal plane. This occurrence is illustrated in Fig. 4.1.
We follow the same procedure for the two IR lenses. The results are listed in
Table 4.3. The errors for these lenses are even smaller. This indicates that with
the higher indices of refraction for silicon and germanium, and the longer
wavelengths, the thin-lens aberration expressions are quite accurate in the
infrared regions.
Table 4.3 Transverse spherical aberrations for two singlets in the infrared
regions.
Spectral TSC TSC TSC TSC

region Thin lens Thick lens (mm) (%)
MWIR -0.26901 -0.26557 0.00344 1.2
LWIR -0.21701 -0.21437 0.00264 1.2
Reference
(2008).
Chapter 5
Achromats
5.1 Introduction
The progression of an achromat from the basic conventional form to a diamond-
turned aspheric hybrid for the infrared spectrum will be demonstrated in this
Chapter. We start with a cemented doublet for the visible spectrum, then change
to the mid-wave and long-wave infrared regions, and proceed with the rest of the
improvements relating to aberration reductions in the long-wavelength spectral
band. For this exercise, we elect a 100-mm focal length with an f/4 relative
aperture and add a field of ±2 deg.
5.2 Thin Achromat for the VIS Spectrum, On-

Axis Performance
A conventional achromat consist of two elements. One has positive power and a
low relative dispersion (high Abbe number), the other has negative power and a
high relative dispersion (low Abbe number). The elements’ powers, required for
chromatic aberration correction, are
Va
a   for the front element, (5.1)
Va  Vb 
and
Vb
b   for the rear element. (5.2)
Vb  Va 
Va and Vb are the mentioned Abbe numbers for the two elements.
With a   na  1 ca and b   nb  1 cb , Eqs. (5.1) and (5.2) change to
Va Vb
 na  1 ca   and  nb  1 cb  ,
Va  Vb  Vb  Va 
from which we can extract the net curvature for the front element a:
39
40 Chapter 5
Va
ca  . (5.3)
Va  Vb  na  1
The net curvature of the rear element b is
Vb
cb  . (5.4)
Vb  Va  nb  1
Since the shapes of the elements in an achromat do not influence the
chromatic aberration, we make the first element equi-convex, i.e.,
ca1  ca 2  0.5ca . Further, we give the first surface of element b the same
curvature as the second surface of element a, i.e., cb1  ca 2 . With that,
Va
ca1    ca 2 . (5.5)
2 Va  Vb  na  1
Furthermore, since cb1  ca 2 and
Vb
cb  ,
Vb  Va  nb  1
Vb 
cb 2  ca 2  cb  ca1  . (5.6)
Vb  Va  nb  1
For the front element a (crown) we take glass BK7 with na  1.517 and
Va  64.17. The rear element b (flint) is made from glass F2 with
nb  1.620 and Vb  36.37 . With that we get
64.17
ca1   0.01  0.022324 ,
2  64.17  36.37 1.517  1
or Ra1  1 / ca1  44.795387 mm and Ra 2  44.795387 mm.
With cb1  ca 2  0.022324,
36.37  0.01
cb 2  0.022324   0.001223 ,
 36.37  64.17 1.62  1
and Rb 2  1 / cb 2  817.966784 mm.

Achromats 41
Figure 5.1 Cemented achromat for the VIS spectrum (corrected for on-axis only)
after adding thicknesses and optimizing.
Figure 5.2 Blur-spot size of the optimized achromat for the visible spectrum.
42 Chapter 5
Figure 5.3 The encircled energy with reference to the diffraction limit.
Figure 5.4 Modulation transfer function of the achromat.

Achromats 43
This choice of curvatures results in a respectable small spherical aberration of
 TSC  0.0225 mm .
1
We add reasonable thicknesses of 5 mm and 2.5 mm and optimize. The

chromatic aberration remains corrected and the transverse spherical aberration
reduces to –0.0156 mm.
We summarize our findings for the expected performance with a 4-plot

report shown in Figs. 5.1, 5.2, 5.3, and 5.4. These selected plots are from the
ZEMAX lens design program, which was used for the optimization of the
achromat.
5.2.1 Adding a field to the on-axis corrected achromat

Adding a ±2-deg field of view to the on-axis corrected lens shows that there is a
relatively large amount of coma present. This is indicated in Fig. 5.5. To also
correct for coma, the shapes of the elements need to be changed. We perform this
task first by an optimization process with the computer, and then show for
comparison Smith’s analytical method.1 This is done to provide more insight into
the method of aberration correction.
Figure 5.5 The presence of coma indicates the need for further correction. This
is also indicated in the modulation transfer function shown in Fig. 5.6.
44 Chapter 5
Figure 5.6 The modulation transfer function drops off drastically for the 2-deg off-
axis point.
Figure 5.7 Improvement of blur spot after optimization.

Achromats 45
5.2.2 Optimizing
We include the off-axis aberrations and computer-optimize the configuration
from above. The result is shown in Fig. 5.7.
5.3 Smith’s Method1 of Determining the Surface

Shapes
In addition to satisfying the previously stated requirements of Eqs. (5.1) and
(5.2), there are two expressions that lead to the elimination of coma and thin-lens,
third-order spherical aberration. These expressions are
K1ca1  K 2  K 3cb1  K 4  0 , (5.7)
and
K 7 ca21  K8ca1  K9  K10 cb21  K11cb1  K12  0 . (5.8)
As before, the curvatures are identified as follows: ca1 and ca 2 refer to the first
and second surfaces of the front element a, and cb1 and cb 2 classify the curvatures
of first and second surfaces of the rear element b.
With K5   K3 / K3 , and K 6    K 2  K 4  / K 3 , the first radius of element b

(rear element) can be found with
cb1  K 5ca1  K 6 . (5.9)
The Ks are functions of the element curvatures, the inverse object distances, and
the familiar G-sums, which are, in turn, functions of the index of refraction of the
optical materials used for the elements, as we already know.
5.3.1 Curvatures and inverse object distances

The net curvatures for the two elements are ca  a  na  1 and cb  b  nb  1 .
We assume in our example calculation that the object is located at infinity.

For this case, the inverse object distance va for element a = 0. For element b, the
inverse object distance vb  1 / f a  a .
46 Chapter 5
5.3.2 The K-functions

K1  0.25G5a ca K 7  G4a ca
K 2  G c 2
8a a K8  G2a ca2
K3  0.25G5b cb K9  G1ca3
K 4  G7b cb vb  G8b cb2 K10  G4b cb
K5   K1 / K3 K11  G2b cb2  G5b cb vb
K6    K2  K4  / K3 K12  G1b cb2 vb  G3b cb2vb  G6b vb2
Substituting Eq. (5.9) into Eq. (5.8) leads to the quadratic equation
K 7  K10 K 52  ca21   K 8  2 K10 K 5 K 6  K11 K 5  ca1

 ( K 9  K10 K 62  K11K 6  K12 )  0
or
Aca21  Bca1  C  0 .
To solve for the first curvature of element a, we apply the standard form for a
quadratic equation:
 B  B 2  4 AC
ca1  , (5.10)
2A
where
A  K 7  K10 K 52 , B  K 8  2 K10 K 5 K 6  K11K 5
and
C  K 9  K10 K 62  K11K 6  K12 .
We now relate all these expressions to our example, where the focal length is
100 mm, and the field is ±2 deg. The indices of refraction are
na  1.517 and nb  1.620 , and the Abbe numbers are Va  64.17 and
Vb  36.37 . Notice, the relative aperture (f/#) does not have to be known for these
calculations.
Achromats 47
5.3.3 The calculations

64.17
a   0.01  0.023083
 64.17  36.37 
36.37
b   0.01  0.013083
 36.37  64.17 
0.023083
ca   0.44647
1.517  1
0.013083
cb   0.021101
1.620  1
Element a Element b
G1 = 0.5945 G1 = 0.8137
G2 = 1.0425 G2 = 1.3145
G3 = 1.4342 G3 = 1.8168
G4 = 0.5991 G4 = 0.6928
G5 = 1.7150 G5 = 2.0055
G6 = 1.1159 G6 = 1.3128
G7 = 0.6872 G7 = 0.8114
G8 = 0.3919 G8 = 0.5022
K1 = 0.019149
K2 = –0.000782
K3 = –0.010578
K4 = 0.000172
K5 = 1.810242
K6 = –0.057667
K7 = 0.026758
K8 = –0.002079
K9 = 0.000053
K10 = –0.014615
K11 = 0.000392
K12 = –0.000003731
The coefficients for the quadratic Eq. (5.10) are A = –0.021136329, B =
0.001681137, and C = –0.000021951.
With that, ca1  0.016465574 . Inserting this into Eq. (5.9) results in
cb1  1.810267  0.016466  0.057667  0.027859 . The remaining two
curvatures are ca 2  ca1  ca  0.016466  0.04461  0.028195, and
cb 2  cb1  cb  0.027866  (0.021097)  0.006769.
Finally, we get from these curvatures the radii of the elements.

48 Chapter 5
Radii
Ra1 = 60.7691mm,
Ra2 = –35.4673mm,
Rb1 = –35.8950mm, and
Rb2 = –147.7323mm
We use this prescription, add thicknesses to the elements, and optimize. There is
no difference in comparison with the more direct method applied earlier. If one
prefers this approach using the K-factors, it is highly recommended to create a
program for the lengthy calculations.
5.4 Achromat for the MWIR Region

We repeat the calculation for an achromat used in the MWIR (3–5 µm) region,
and select silicon for the front element and germanium for the rear element.
The indices are na =3.4253 (Si) and nb = 4.0251 (Ge), with the Abbe numbers
Va = 236.5 and Vb = 107.3.
The powers for the achromat are

Va 236.5
a    0.01  0.018305,
Va  Vb   236.5  107.3
1 1
fa    54.6300 mm,
a 0.018305
Vb 107.3
b    0.01  0.008305,
Vb  Va  107.3  236.5
and
1 1
fb    120.4101 mm,
b 0.008305
and the thin-lens radii for the silicon element a are R1 = 264.971 mm, R2 = R3 =
–R1 = –264.971 mm, and R4 = –971.817 mm.
Since elements for the MWIR spectrum cannot be cemented (the cement
does not transmit in that region), we have two additional freedoms with the
option of making R3 different from R2 and choosing a suitable spacing between
the elements. (This freedom of enlarging the spacing between the elements will
be discussed later). The lens shown in Fig. 5.8 has been optimized after adding
appropriate thicknesses to the elements.
Achromats 49
Figure 5.8 Blur spots for the MWIR achromat with a ±2-deg field.
Figure 5.9 This LWIR achromat is diffraction limited. The circles indicate the size
of the Airy disk. This indicates that the f/# could be reduced without losing image
quality.
50 Chapter 5
5.5 Achromat for the LWIR Region

For completeness, we design an achromat using the same procedure with the
same focal length and field coverage for the LWIR (8–12 µm) region.
The front element is made from AMTIR-1, with na  2.4975 and Va  113.4 .
For the rear element we use zinc sulfide with nb  2.2003 and Vb  22.7. The
result is a thin lens with the following radii: Ra1 = 91.8029, Ra2 = 393.0097, Rb1 =
−780.7366, and Rb2 =1,243.5363. Adding thicknesses leads to a diffraction-
limited lens whose spot sizes are shown in Fig. 5.9.
5.6 Diamond-Turned Hybrid2

Combining refractive and diffractive powers in one optical element is referred to
as a hybrid lens. It is frequently used to correct chromatic aberration, and by
aspherizing at least one surface, one can also control the spherical aberration.
Because most of the infrared materials have a mild spectral dispersion (high
Abbe number), the required number of zones for the diffractive phase profile is
low. Both the aspheric shape and the diffractive structure can be conveniently
generated with the process of single-point diamond turning on an ultra-precision
lathe. For protective reasons, the preferred arrangement is to place the asphere
with the diffractive profile on the second surface. Figure 5.10 schematically
shows the cross section of a hybrid element with the diffractive phase profile,
highly exaggerated. We will design such hybrids for both IR regions.
Figure 5.10 Hybrid lens with refractive and diffractive powers.

Achromats 51
5.6.1 Hybrid for the MWIR region

A suitable material for this region is silicon with n  3.4253 . The refractive Abbe
number VR  236.5 . The diffractive Abbe number is
0 4
VD    2 .
S  L 3  5
This relation indicates that the diffractive phase profile is independent of the lens
material. It also indicates by the negative sign that the chromatic aberration is in
the opposite direction of the aberration from the refractive portion of the lens.
The required refractive focal length for the chromatic aberration correction is
 V 
f R  1  D  f , (5.11)
 VR 
and the diffractive focal length is
 V 
f D  1  R f . (5.12)
 VD 
The radius of the first diffractive zone is
r1  2λ 0 f D , (5.13)
and the total number of zones is
D
itotal  , (5.14)
2r1
where D is the lens diameter. The step height of the profile is
λ0
d max  . (5.15)
n0  1
5.6.1.1 Basic lens shape

According to Eq. (5.11), the required refractive focal length is
 V   2 
f R   1  D  f  1   100  100.845667 mm .
 VR   236.5 
52 Chapter 5
From Chapter 2 we apply Eqs. (2.10), (2.11), and (2.12) for a “best-shaped” lens
(minimum spherical aberration) and get for the surfaces the radii
2  3.4253  2  3.4253  1
R1   100.845666  98.690419 mm
3.4253  2  3.4253  1
and
2  3.4253  2  3.4253  1
R2   100.845667  165.451409 mm .
3.4253  2  3.4253  1  4
The transverse spherical aberration of this best-shaped lens, with a 100-mm focal
length and an f/4 relative aperture, is
3.4253   4  3.4253  1  100.845667

TSC    0.033565 mm .
64   3.4253  1   3.4253  2   43
2
5.6.1.2 Aspherizing
To eliminate this spherical aberration, we aspherize the second surface with the
method described in Sec. 3.1.2. The conic constant for the aspheric second
surface is
2 R23TSC 2  165.4514093   0.033565 

κ2      0.636465 .
 n  1 f y23  3.4253  1 100.845667 12.53
5.6.1.3 The diffractive phase profile

To correct for chromatic aberration, we design the necessary phase profile and
begin with the requirement for the diffractive focal length,
 V   236.5 
f D  1  R  f  1    100.845667  12, 025.84579 mm .
 VD   2 
The first zone radius r1  2λ 0 f D  2  0.004  12,025.84579  9.809 mm and

the total number of zones needed across the aspheric surface is
itotal  D  2r1   25  2  9.809   1.274 , which means one full and a second
partial zone. The step height d max  0.004  3.4253  1  0.00165 mm.
Achromats 53
5.6.1.4 Comments
The phase equation is expressed by
2π
r  
λ
 S 2 r 2  S 4 r 4  S6 r 6  ...Sn r n  , (5.16)
and the quadratic coefficient for the diffractive surface can be calculated from
1
S2   . (5.17)
2 fD
In the ZEMAX program, the quadratic phase coefficient is expressed as
2π 2
P2  R0 S 2 , (5.18)
λ0
where R0 is called the normalizing radius and is usually chosen to be 100.
The phase coefficient is
1 1
S2    4.158  105 .
2 f D 2  12025.8
When converted into ZEMAX format, it is
 2π 
  100   4.158  10   653.137 .
5
P2   2
 0.004 
We add a reasonable thickness of 2.5 mm and optimize this hybrid with a 4-deg
total field. The result is summarized in Fig. 5.6. As a reference, it is mentioned
that the diffraction limit for this lens is BAiry  2.44  f/#   2.44  4  4  39 µm.
The size of the square detector element is shown in Fig. 5.11 with 40 µm.
Besides using a diffractive profile for chromatic aberration correction, it can

also be applied for athermalization of a lens. This will be discussed in Chapter 9.
54 Chapter 5
Figure 5.11 Blur-spot sizes of this optimized MWIR hybrid are equal to the size
of the diffraction-limited blur (Airy disk), which is also the size shown for the
square detector element (pixel).
Figure 5.12 Nomogram for hybrids in the MWIR region (object at infinity).
Achromats 55
Figure 5.13 Nomogram for hybrids in the LWIR region (object at infinity).
5.6.2 Useful nomograms

Two nomograms are inserted here to provide quick answers for the zone numbers
required to correct chromatic aberration in the two infrared regions. Figure 5.12
refers to the MWIR region, and Fig. 5.13 to the LWIR band.
The example in Fig. 5.12 shows that the first zone radius of an f/1.5 lens with
f = 50 mm, made from zinc selenide, is 6 mm, and there are eight zones.
56 Chapter 5
In the LWIR region (8–12 µm), germanium is a superior material with its
high Abbe number of about 800. This is indicated by the large separation from
the other materials on the lens material scale in Fig. 5.13.
Example 1 shows an f/1 lens with a 50-mm focal length. The first zone radius
is found to be 6.7 mm, leading to a total of 14 zones across the lens diameter.
Example 2 refers to the 100 mm, f/1 germanium lens discussed above. The first
zone radius is already quite large with 28.3 mm. This germanium lens requires
only three steps (four zones) across its 100 mm diameter.
References
(2008).
Chapter 6
Systems with Two Separated
Components
6.1 Introduction
The telephoto, the reversed telephoto, and the dialyte (which is an air-spaced
achromat) belong in this category, and there are others. We limit our discussion
to the case where the object is located at infinity. The basic layout of such a
system is shown in Fig. 6.1.
6.2 Dialyte—an Air-Spaced Achromat

The relations for an air-spaced achromat are
the focal length of front element are
 V f 
f A  1  B  f , (6.1)
 VA b 
the focal length of rear element
 V b 
f B  1  A b , (6.2)
 VB f 
Figure 6.1 General layout of system with two separated components for an
object located at infinity.
57
58 Chapter 6
and the separation of the two elements
 b
d  1   f A . (6.3)
 f 
It can be seen from these expressions that the achromatic doublet is covered
as a special case with d = 0, which occurs when b = f. Equations (6.1) and (6.2)
then take on the form of the equations presented in Chapter 5, Achromats. There
is a maximum useful separation for the elements, and the separation is dependent
on the materials used. We can derive this limit by first inserting Eq. (6.1) into Eq.
(6.3) and then differentiating the new expression with respect to b. Setting the
result to zero brings us the maximum separation d for the dialyte under
investigation. Let us demonstrate:
 b  b  V f 
d  1   f A   1    1  B  f . (6.4)
 f   f  VA b 
We differentiate Eq. (6.4) and find

dd VB 2 2
 f b 1.
db VA
Setting this result to zero yields
VB
b f .
VA
With that result, Eq. (6.4) changes to the simple expression

2
 V 
d max  1  B  f . (6.5)
 VA 

At this maximum distance, the focal lengths are equal but, of course, opposite in
sign, i.e., f B   f A .
Typical values for the three spectral windows, VIS, MWIR, and LWIR are
listed in Table 6.1.
Table 6.1 Maximum possible spacings for dialytes.
Region Material Abbe number Spacing

Front lens Rear lens Front lens Rear lens dmax
VIS BK7 S2 64.17 36.37 0.06109 f
MWIR Silicon Germanium 236.5 107.3 0.10656 f
LWIR AMTIR-1 Zinc sulfide 113.4 22.7 0.30536 f
Systems with Two Separated Components 59
Figure 6.2 Ranges of element separations for dialytes with a 100-mm focal
length.
Figure 6.2 shows a plot of the element separation d as a function of the back
focal length b for dialytes with a 100-mm focal length. The materials of the
elements are called out in Table 6.1. It is interesting to observe how the spacing
limits increase with wavelength.
6.2.1 Example for the MWIR region

For the 100-mm-focal-length objective, we decide on a 10-mm separation of the
Si/Ge elements, which is a bit less than the limit called out in Table 6.1.
Rearranging Eq. (6.4) leads to a quadratic equation for the back focal length b:
  V  V
b2   d  f 1  B   b  B f 2  0 . (6.6)
  VA   VA
For our example, this becomes
  107.3   107.3
b 2  10  100 1   b   1002  0 ,
  236.5  236.5

b  0.5 135.369979  135.3699792  4  4,536.997886 , 
b1  74.3378 mm, and b2  61.0322 mm .†
We add reasonable thicknesses to the elements and optimize both configurations

with a ±2-deg field. The first form with b = 74.3378 mm leads to a slightly better
solution. The performance plots of this preferred configuration are shown in Fig.
6.3. It is preferred because of better performance and compactness.
†
Also see Fig. 6.2.
60 Chapter 6
Figure 6.3 Dialyte performance of a configuration, derived from thin-lens

expressions.
As an interesting exercise, we use silicon for both elements and make the rear
element an aspheric hybrid. The basic prescription for dialyte is listed in Table
6.2, and Fig. 6.4 summarizes the performance of this air-spaced achromat with a
hybrid element.
EFL = f = 99.998 mm, BFL = b = 94.188 mm, f/4.
Table 6.2 Basic prescription of dialyte with a hybrid element.
Surface # Radius Thickness Material Index n V-Number

1 154.2073
3 silicon 3.4253 236.5
2 478.3504
10 air
3 −25.7727
3 silicon 3.4253 236.5
4 −27.9023
93.99151 air
Figure 6.4 Performance of an air-spaced achromat with a hybrid element.
6.2.1.1 Aspheric deformation coefficients of surface 3

A4 = 4.44547×10–8
A6 = –2.62939×10–10
6.2.1.2 Phase coefficients

OSLO and CODE 5
S2 = –6.5155×10–5
ZEMAX with R0 = 100 mm
P2 = –1,023.4466
Table 6.3 Diffractive Zones on Surface 3 (for the first diffraction order)
Zone # 1 2 3
Radius r 7.8353 11.0808 13.5712
62 Chapter 6
6.2.1.3 Step height at the zone transition

λ0 4
d max    1.649 µm
 nb  1  3.4253  1
It can be seen that the performance is somewhat better. An advantage of this
design is that it avoids the higher-priced germanium as material for the second
element.
6.3 Telephoto and Reversed Telephoto

Figures 6.5 and 6.6 show the telephoto and reversed telephoto arrangements. The
focal lengths for the elements are the same for both configurations, namely
df
fA  (6.7)
 f  b
and
db
fB  . (6.8)
d  b  f 
Figure 6.5 The basic telephoto system.
Figure 6.6 The basic reversed telephoto system.

6.3.1 Examples for the MWIR and LWIR regions

We present a reversed-telephoto system that was derived from the basic layout
shown in Fig. 6.6. The curved surfaces have been strongly aspherized. For its
specific application, a cold stop and a spectral bandpass filter have been placed
inside a Dewar housing. The objective covers a large field of ±13.5 deg and has a
relative aperture of 1.5. The front element is made from zinc sulphide. The
material for the rear element and the Dewar window, as well as the filter
substrate, is silicon.
The next example of a reverse telephoto for the LWIR region, as shown in
Fig. 6.8, uses only germanium for all elements including the filter. It has a 50-
mm focal length and a relative aperture of f/1.5. The field coverage is ±10 deg.
The front element is a hybrid with the phase profile on the aspherized second
surface. The second surface of the rear element is also aspherized. The aperture
stop is located at the front surface of the first element.
Figure 6.7 Reversed telephoto for the MWIR window, after Aldrich.
64 Chapter 6
Figure 6.8 Reversed telephoto for the LWIR with a hybrid front element.
Chapter 7
From an Air-Spaced Doublet to a Triplet
7.1 Introduction
As an informative exercise, we begin with a suggested formulation1 for a Petzval
objective as shown in Fig. 7.1.
Figure 7.1 Basic Petzval objective as a convenient starting configuration.
Figure 7.2 The proposed triplet.
65
66 Chapter 7
To correct chromatic aberration, we split the front element to form an

achromatic triplet.
7.2 Chromatic Aberration

The thin-lens third-order transverse chromatic aberration of the system is
expressed by
c
y 2
TAchC   . (7.1)
a Vu 'k
For an object at infinity, u'k   y . With that, Eq. (7.1) changes to
1 c
y 2
TAchC  
ya total
a V . (7.2)
To eliminate chromatic aberration, we have to satisfy the condition
c
y 2 ya2 a yb2 b yc2c
a V  V  V  V  0. (7.3)
a b c
We know that a  b  0.5tot , and therefore b  0.5tot  a . Further, c   tot ,

ya  yb  y, and yc  0.5 y .
After substitution and solving for the power of the first element, we get
a 
Vc  Vb / 2 Va  . (7.4)
2 Va  Vb Vc
tot
By choosing the same material for elements a and c, Eq. (7.4) simplifies to read
a 
Va  Vb / 2   . (7.5)
2 Va  Vb 
tot
7.3 Example, a Conventional Triplet

We design an objective, again for the MWIR region, maintaining our 100-mm
focal length. However, we reduce the relative aperture from f/4 to f/2 to collect
more energy. We also expand the field of view and double it to ±4 deg.
From an Air-Spaced Doublet to a Triplet 67
Elements a and c are made from silicon with na = 3.4253 and Va = 236.5. The
material for element b is germanium with nb = 4.0251 and Vb = 107.3. The power
of the front element is therefore
a 
 236.5  107.3 / 2   0.01  0.011229, and f  1  89.057 mm .
2   236.5  107.3
a
a
b  0.5  0.01  0.011229  0.006229, and f b  160.539 mm.
The power of the rear element is c  tot  0.01 . Therefore, f c  f tot  100 mm.
Assuming the shapes of the positive elements a and c to be convex/plano,

and the negative element c plano/concave, we find the surface curvatures to be
a 0.011229
ca1    0.004630 ,
 na  1  3.4253  1
b 0.006229
cb 2    0.002059 ,
1  nb  1  4.0251
and
c 0.01
cc1    0.004123 .
 nc  1  3.4253  1
The radii are therefore: Ra1 = 215.987 mm, Ra2 = infinity, Rb1 = infinity, Rb2 =
485.673 mm, Rc1 = 242.542 mm, and Rc2 = infinity.
We add thicknesses to the elements and adjust the spacings before we

optimize. The final result is shown in Fig. 7.3.
7.4 Hybrid Petzval Objective with a Cold Stop2

We replace the front doublet of the conventional configuration shown in Fig. 7.3
with an aspheric hybrid element and position the aperture stop inside a Dewar
housing where the stop will be kept cold so that the detector will only receive
energy from the scene. Included in the optical design is also the Dewar window
and the spectral filter, which is also located inside the Dewar housing. This
configuration, as shown in Fig. 7.4, is frequently used for IR imaging systems.
68 Chapter 7
Figure 7.3 Conventional triplet, derived from the basic Petzval form.
Figure 7.4 Hybrid Petzval objective for MWIR with a cold stop.
Reference
(2008).
2. M. J. Riedl, “The design of an IR Petzval objective using an aspheric and
diffractive element: an exercise with MOE for beginners,” Proc. SPIE 5865,
586501 (2005).
Chapter 8
A Hybrid for Two Wavelengths
8.1 Introduction
A single element that is corrected for spherical aberration for two chosen
wavelengths can be advantageous for alignment purposes in the visible spectrum
when the application is in the infrared region. We design such an element for the
helium-neon and the carbon dioxide lines. The wavelength for a HeNe laser is
0.6328µm, and is 10.6 µm for a CO2 laser. The design principle is to determine
the shape of the lens for the longer wavelength first and then place a diffractive
structure on one side of the lens to add the required power for the shorter
wavelength to maintain focus position. Since the wavelength ratio is large,
10.6/0.6328 ≈ 17, the step height of the diffractive structure for the short
wavelength is
λ 0.6328
d   0.4 µm .
n  1 2.5907  1
This is in the domain of surface roughness for the infrared region.
First, we have to pick a material that is fitting for both wavelengths. We
choose ZnSe with an index of refraction of n0 6328  2.5907 and n10 6  2.4027 .
We elect for our design example 100 mm as a focal length and a relative aperture
of f/# = 2, and begin with a thin-lens pre-design.
8.2 Basic Lens Shape for the Long Wavelength

Considering that it is much easier to place a diffractive structure on a flat surface
when the number of zones is high, which will be the case here; we opt for an
aspheric plano-convex shape and aspherize the first surface to eliminate spherical
aberration.
The front radius is R1   n10 6  1 f CO2   2.4027  1  100  140.27 mm . This
shape leads to a focal length for the shorter wavelength of
R1 140.27
f HeNe    88.1054 mm .
n0 6328  1 2.59207  1
To maintain a common focal length for both wavelengths, a negative diffractive

power has to be added to the flat surface. That diffractive power is determined
with  D  CO2  HeNe . The required focal length for the diffractive optical
element (DOE) is, therefore,
69
70 Chapter 8
f HeNe f CO2 88.1054  100

fD    740.72 mm .
f HeNe  f CO2   88.1054  100 
8.3 Required Diffractive Profile

The first zone radius of the diffractive structure is1
r1  2λ HeNe f D  2  0.0006328  740.72  0.9682 mm ,
and the total number of zones required is
2 2
 D   50 
itotal      667 .
 2r1   2  0.968 
As hinted before, this is a large number of zones, which is most likely more
cost effective to etch than to diamond-turn. The individual zone radii can be
obtained by ri  r1 i  0.9682 i . From this, one can see that at the edge of the
lens the zones are only 19 µm apart. We add a thickness of 6 mm to the lens,
aspherize the front surface slightly in the optimization process, and obtain a lens
that is within the quarter-wave Raleigh limit for both wavelengths, as can be seen
from the plots in Fig. 8.1.
Remark
If we consider the zone step height as the P-V surface roughness for the long
wavelength application, and take a ratio of 2 between the P-V and RMS values,
we have δ  0.5  P  V   0.5  0.4  0.2 µm .
Figure 8.1 The quarter-wave Raleigh limit has been maintained for both cases.
A Hybrid for Two Wavelengths 71
The total integrated scatter of an optical surface is expressed by

TIS   4πδ λ  .1 For our case this amounts to TIS   4π  0.2 10.6   5.6% , a
2 2
relatively low amount of the energy that will become stray radiation and reduce
the contrast ratio (MTF) somewhat.
Reference
Chapter 9
Athermats
9.1 Introduction
Athermats are lenses that are designed to compensate for the focus shift that
occurs with temperature excursions. The changing parameters of an optical
element are the radius, the thickness, and the index of refraction. The spacing of
the lens from the detector also changes and is a function of the coefficient of
expansion of the housing material.1
9.2 Focus Shift of a Refractive Element

The power (reciprocal of the focal length f) of a thin lens is given by
 1 1 
   n  1    , (9.1)
 R1 R2 
where n = index of refraction, and R1 and R2 are the surface radii. Differentiation
of Eq. (9.1) with respect to temperature yields
d  1   dR   1   dR    1 1  dn
  n  1  2   1    2   2       . (9.2)
dt  R1   dt   R2   dt    R1 R2  dt
We recognize in this equation that 1 R1  dR1 dt   1 R2  dR2 dt   α L , the

thermal coefficient of the lens material. Therefore,
d  1 1   1 1  dn
  n  1   α L  α L      .
dt  R1 R2   R1 R2  dt
Rearranging and making use of Eq. (9.1) results in
d     dn
  . (9.3)
dt   n  1  dt
Since   1 f , and therefore d  dt   1 f 2   df dt  , for a finite temperature

change of t we get a change in focal length of
73
74 Chapter 9
 dn / dt 
f   f   α L  t . (9.4)
  n  1 
 dn / dt  /  n  1   
L is called the thermal glass constant TR. With that, the
focus shift of a thin lens is simply
f  TR f t . (9.5)
9.3 Athermalization with a Doublet

For two elements, defocusing due to a temperature change is given by
 T T  
f   f  f  a  b   α h   t . (9.6)
  f a f b  
Element a is the front element and b is the rear element. h is the thermal
coefficient of the housing in which the lens is mounted.
Setting f = 0 to eliminate any focal shift requires that the focal length of the
front element a is
 T T 
fa   b a  f . (9.7)
 Tb  α h 
The focal length of the rear element b is then
 T T  fa f
fb   a b  f , or . (9.8)
 Ta  α h   fa  f 
9.4 Focus Shift of a Diffractive Lens

Based on the quadratic term of the general phase equation, the first zone radius of
a purely diffractive element is
r1  2λ 0 f D . (9.10)
Therefore, the focal length is
Athermats 75
r12
fD  . (9.11)
2λ 0
The zone radius r1 as a function of the temperature change t is
r1 ( t )  r1 1  α L t  , (9.12)
where L is the thermal coefficient of expansion of the lens material and t is the
temperature change from the design temperature t0. The diffractive focal length at
the changed temperature t can be expressed by
r12 t  r12 
1  2α L  α 2L  t   ,
2
f D t   
2λ 0 2λ 0  
and by neglecting the term α 2L  t  ,

2
r12
f D t   1  2α 2L t  . (9.13)
2λ 0 
Since f D t   f D  f D , the change in focal length due to a change in

temperature of t is
f D  2α L f D t , (9.14)
which means the thermal glass constant of a diffractive lens is simply
TD  2α L . (9.15)
This indicates that the change of a diffractive element due to a change in

temperature is independent of the index of refraction of the element material.
Due to the different behavior of a diffractive element, the equations for the
elements’ focal lengths (to eliminate any focal shift caused by a change in
temperature) are
 T  TD 
fR   R f . (9.16)
 TD  α h 
fR is the refractive focal length of the element, and f D is the diffractive focal
length of the element.
76 Chapter 9
T T  fR f
f D   R D  f or f D  . (9.17)
 TR  α h   fR  f 
A positive temperature change of the refractive portion of the hybrid reduces
the focal length [see Eq. (9.5)]. As the zone radii r of the diffractive portion
(phase profile) increases with a temperature increase, the diffractive focal length
also increases [see Eq. (9.14)]. It is important to remember that the diffractive
focal length is positive when the steps at the zone transition are negative. The
focal length is negative when the steps are cut in the opposite direction. The
direction of the transition steps is illustrated in Fig. 9.1.
The sign of the focal length is indicated in the phase equation
2π
r  
λ0
 S 2 r 2  S4 r 4  S6 r 6  ... , (9.18)
Figure 9.1 Direction of phase profile and effect of sign on focal length.
Athermats 77
where the first coefficient
1
S2   . (9.19)
2 fD
The size of the steps at the zone transition is
λ0
d . (9.20)
 n0  1
The change in step height with a temperature change is minimal and results in a
very slight shift of the wavelength at which the theoretical diffraction efficiency
is 1. This change can be neglected and can be confirmed with
λ  λ 0  d 1  α L t   n   dn / dt  t  1 . (9.21)
The symbols used in this equation have been defined above.
Tables 9.1 and 9.2 list the properties of the most frequently selected materials
for optical elements for the infrared spectra.
Table 9.1 Selected materials for the 3–5 µm spectral band (MWIR).
Optical Refractive Abbe Index Coefficient of Thermal glass

material index number change / C expansion / C constant
–6
n (at 4 m) V (at 3–5 m) dn/dt (10–6) L (10–6) TR (10 )
Ge 4.025 103 396 5.7 125.2
Si 3.425 236 150 2.6 59.3
ZnS 2.253 110 43.3 6.6 28.0
ZnSe 2.433 179 60 7.1 34.8
AMTIR-1 2.514 198 72 12 35.6
MgF2 1.349 13.5 2 12 –6.3
CaF2 1.410 21.7 –7.5 19 –37.3
BaF2 1.457 45.1 –16 18 –53.0
As2S3 2.412 181 9.3 21 –14.4
Al2O3 1.675 7.65 3.4 6 –1.0
GASIR-1 2.510 196 67 17 27.4
GASIR-2 2.606 170 70 16 27.6
78 Chapter 9
Table 9.2 Selected materials for the 8–12 m spectral band (LWIR).
Optical Refractive Abbe Index Coefficient of Thermal glass

material index number change / C expansion / C constant
–6 –6
n(10 m) V(8–12 m) dn/dt (10 ) L (10–6) TR (10 )
Ge 4.003 1000 396 5.7 126.2
ZnS 2.200 23 39 6.6 25.9
ZnSe 2.406 58 61 7.1 36.3
AMTIR-1 2.498 110 72 12 36.1
GaAs 3.278 110 185 5 76.2
CaTe 2.672 170 97 5 53.0
NaCl 1.495 19 –25 44 –94.5
CsI 1.740 230 –89 50 –170.3
GASIR-1 2.496 120 67 17 27.8
GASIR-2 2.586 101 70 16 28.1
9.5 Design Examples

We examine two f/4 lenses with 100-mm focal lengths used in the LWIR region
at 10 µm. The first one consists of two elements, and the second one is a
diffractive hybrid. The thermal conditions are the same for both cases.
The design temperature is 20°C, and the temperature extremes are –20°C and
+60°C. This means t = ±40°C. Both lenses are mounted in aluminum housings.
9.5.1 Athermat with two elements in an aluminum

housing
For the front element we choose germanium with Ta = 126.2×10–6. The rear
element is made from zinc sulfide with Tb = 25.9×10–6 . The thermal coefficient
of expansion of the aluminum housing ah = 23×10–6.
Equation (9.7) yields
 25.9  126.2 
fa     100  205.112 mm .
 25.9  23 
This leads, according to Eq. (9.8), to
fa f 205.112  100
fb    67.225 mm .
 f a  f  205.112  100
For simplicity, we use a plano-concave and a plano-convex lens, for which
the radii are Ra1   na  1 f a   4.003  1   205.112   615.953 mm and
Rb1   nb  1 f b   2.200  1  67.225  80.670 mm. Ra2 and Rb2 are infinite.
Athermats 79
9.5.2 Hybrid athermat in an aluminum housing

We use zinc selenide as the lens material, with n = 2.406 at 10 µm. TR = 36.3 and
TD =14.2.
Using Eqs. (9.16) and (9.17), we get
 T  TD   36.3  14.2 
fR   R  f    100  573.86 mm
 TD   h   14.2  23 
and
fR f  573.86  100 85.16 mm .

fD   
 f R  f   573.86   100
For the plano-concave lens, the front surface radius is

R1   n  1 f R   2.406  1   573.86   806.472 mm. R2 is infinite.
To determine the details of the phase profile for the flat rear surface of the
lens, we begin by using Eq. (9.19) and get the required phase coefficient,
1 1
S2     0.0058713 .
2 fD 2  85.16
For reference, we translate S2 into ZEMAX format, which is
2πRn2 2π  1002
P2  S2    0.0058713  36,890.5 .
λ0 0.01
The first zone radius is found with Eq. (9.10):
r1  2λ 0 f D = 2  0.01 85.16 =1.305067 mm
and
r2  r1 2  1.305067 2  1.84564 mm.
The other zone radii n can be found with ri  r1 i . The total number of zones
required is i   D 2r1    25 2 1.305067   92 . D is the lens diameter. It is 25
2 2
80 Chapter 9
mm for the f/4, 100-mm focal-length objective. The step size at the zone
transition is d  λ 0  n0  1  10  2.406  1  7.11 µm.
9.6 Impact of Housing Material

We analyze how different housing materials impact the focus shift of a hybrid
lens. The design focal length is 100 mm, and the optical material is zinc selenide
with n = 2.406 and L = 7.1×10–6, which means that the thermal glass coefficients
are TR = 36.3×10–6 (see Table 9.2) and TD  2 L  2  7.1 106  14.2 106 . The
second lens surface is chosen to be a plano for all cases. h is the thermal
coefficient of the housing material, fR is the refractive focal length of the hybrid,
fD is the diffractive focal length, R1 is the first surface radius of the element, r1 the
first diffractive zone radius, and i the total number of required zones across the
f/4 lens diameter.
Table 9.3 Change of parameters for a hybrid lens as a function of housing

material.
–6
Material h(×10 ) fR fD R1 (mm) r1 (mm) i
Alum 23 –573.86 85.16 –806.847 1.30507 92
Steel 12 2,295.45 104.55 3,227.403 1.44603 75
Invar 1.3 391.47 134.09 550.407 1.63954 59
Remark
It can be seen from Eq. (9.16) that the lens will be a plane-parallel plate with a
diffractive structure on the second surface, if the thermal expansion coefficient of
the housing h is equal to the thermal glass coefficient TD, which is 2L, i.e.,
twice the thermal coefficient of expansion on the lens material. AMTIR-1 (L =
12 ×10–6) and aluminium 6061 (h = 23×10–6) are such a matching combination.
We investigate the lens at  =10 µm with f = 100 mm, and D = 25 mm (f/4):
T T   36.1  24 
fR   R D  f     100  infinite ,
 TD  α h   24  24 
T T   36.1  24 
fD   R D  f     100  100  f ,
 TR  α h   36.1  24 
and
1 1
S2     0.005 .
2 fD 2  100
Athermats 81
Figure 9.2 Flat AMTIR-1 LWIR athermat in an aluminium housing.
The ZEMAX phase coefficient is
2πRn2 2π  1002
P2  S2    0.005   31, 415.927 .
λ0 0.01
r1  2λ 0 f D  2  0.01 100  2  1.4142 .
The total number of zones across the 25-mm lens diameter is
2 2
 D   25 
i     79 .
 2r1   2  2 
Figure 9.2 shows the blur spots of this flat, purely diffractive element at
–20°C, + 20°C (the design temperature), and at + 60°C.
9.7 Athermat for the CO2 Laser Line

We analyze an f/4 lens with a 100-mm focal length made from ZnSe. The
thermal index variation for ZnSe is dn/dt = 60×10–6, the index of refraction at
10.6 µm (CO2-line) is n = 2.40266, and the thermal coefficient of the element is
L = 7.6×10–6. The diffractive glass constant TD  2  7.1 106  14.2 106 , and
the refractive one is
 60 
TR    7.1  106  35.676  106 .
  2.40266  1 
82 Chapter 9
With the refractive 100-mm focal length we get
 35.676 
fD     100  251.1239 .
 14.2 
From Eq. (9.9) we obtain the quadratic phase coefficient
1 1
S2     0.001991 .
2 fD 2  251.1239
The first zone radius
r1  2λ 0 f D  2  0.0106  251.1239  2.307342 mm .
For the full lens diameter of D = 100/4 = 25 mm, the total number of
required zones is itotal   D 2r1    25 /  2  2.307342    30 .
2 2
Remark
ZEMAX applies a normalized radius of R0 =100 mm for the phase coefficient P.
With that,
2πR02 2π1002
P2  S2    0.001991  11,801 .
λ0 0.0106
We put the numbers from above into the ZEMAX program, add a suitable
thickness to the lens, and optimize. The result is shown in Fig. 9.3.
The top row shows the increase of the blur spot for a purely refractive
singlet. The lower row indicates that the blur spot size stays within the diffraction
limit over the entire 120°C temperature excursion.
Danger!
To demonstrate the effect of machining the phase profile in the wrong direction
(which has happened), we show in Fig. 9.4 how the slopes of the exiting rays are
changed. The focal length of our example lens would increase from 100 mm to
485 mm at the design temperature of 20°C.
Athermats 83
Figure 9.3 Change of blur-spot size due to defocusing caused by a temperature

change of ±60°C from the 20°C design temperature without the compensating
diffractive structure.
Figure 9.4 Devastating effects with the wrong (reversed) phase-profile direction.
See Fig. 9.1 for direction definition.
84 Chapter 9
9.8 Athermalized Achromat

To achromatize and athermalize an objective, we apply the following equations:
total power,
j
   i , (9.22)
i1
athermalization,
  T    α  ,
i 1
i i h (9.23)
and a chromatization,
j
 i 
 V   0 . (9.24)
i 1  i 
As previously stated, V is the Abbe number. VR is for a refractive element,

and VD is for a diffractive one. For the infrared spectrum
nmiddle  1
VR 
nshort  nlong
and
λ middle
VD  .
λ short  λ long
The required powers (inverse focal lengths) for the three elements are
 1 / V1   T1   h 
  
1  1 / V3  1 / V1   T3  T1 
2   , (9.25)
f 2  1 / V1  1 / V2   T1  T2 
  
 1 / V3  1 / V1   T3  T1 
1 1 / V1  1 / V2 1 / V1
3   2  , (9.26)
f 3 1 / V3  1 / V1 1 / V3  1 / V1
Athermats 85
and
1
1      2  3 . (9.27)
f1
We demonstrate the use of these seemingly cumbersome equations with two

examples of an achromat for the MWIR region. First, we use three elements and
then two with an added diffractive phase profile on the rear surface of the second
element. Both objectives are mounted in aluminum housings.
9.8.1 Three-element athermat in an aluminum housing

As lens material, we elect for the elements of the 100-mm focal-length f/4 lens‡:
 Front element 1: Silicon with V1 = 236, and T1 = 59.3×10–6,
 Center element 2: Germanium with V2 = 103, and T2 = 125.2×10–6, and
 Rear element 3: Zinc sulphide with V3 = 110, and T3 = 28.0×10–6
The thermal coefficient of expansion for aluminium is h = 23×10–6. With
Eqs. (9.25), (9.26), and (9.27), we obtain the elements’ focal lengths f1 = 57.6334
mm, f2 = –92.3000 mm, and f3 = 287.0930 mm. After adding appropriate
thicknesses to the elements and maintaining the proper powers of the elements,
we optimize and get the configuration shown in Fig. 9.5. We also added an
angular field of ±2 deg.
Figure 9.5 Performance of an athermalized, 100-mm, three-element achromat in

an aluminium housing. There is no change over a temperature excursion of
±40°C.
‡
Also see Table 9.1
86 Chapter 9
9.8.2 Two-element athermat in an aluminum housing

We reduce the number of elements but add a diffractive profile on the second
surface of the rear element. The front element is again made from silicon, and for
the rear element we use germanium. The elements’ refractive parameters remain
V1 = 236, T1 = 59.3×10–6, and V2 = 103, T2 = 125.2×10–6. The Abbe number for
the diffractive surface is
λm 4
V3    2
λ s  λl 3  5
and
T3  2α 3l  2  5.7  106  11.4  106 .
We recall that the thermal coefficient of expansion for the aluminum housing
is α h  23  106 . Equations (9.25), (9.26), and (9.27) lead to the elements’ focal
lengths. They are: f1 = 44.291 mm, f2 = –79.832 mm, and the diffractive focal
length f3 = –19,273.3 mm. This long focal length indicates that the chromatic
aberration to be corrected with the phase profile is very small. Only two
diffractive zones are required. Figure 9.6 is included for comparison with the
three-element design. Note, there is very little difference. The advantage, of
course, is the savings of the cost of one element, weight, and space.
Figure 9.6 Performance of an athermalized 100-mm focal-length achromat with

two elements and a diffractive surface in an aluminum housing.
Athermats 87
9.9 Effect of Quarter-Wave Limit without

Athermalization
Before adding the cost of optical system athermalization, it is wise to determine
the temperature range over which the required performance is acceptable.
We demonstrate the procedure with a singlet mounted in a housing with the

same length as the focal length of the lens. For a single element, Eq. (9.6)
reduces to
f   f TR  ah  t . (9.28)
We assume that the lens is diffraction limited and allow defocusing of

f  2  f / #  , which represents the well-known Rayleigh quarter-wave
2
criterion. § This changes Eq. (9.28) to
2  f / # 
2
t   . (9.29)
f  TR   h 
Table 9.4 lists the allowable temperature depatures from the design
temperature for three diffraction-limited singlets mounted in an aluminium
housing. The chosen focal length is again 100 mm, and the relative aperture is
f/4.
Table 9.4 Allowable temperature excursions for quarter-wave defocusing.

Spectral (μm) n Lens TR Δt
–6
region material (x10 ) (°C)
VIS 0.5876 1.5168 BK7 –4.0137 ±9.4
MWIR 4 3.4253 Si 61.7189 ±15.1
LWIR 10 4.0043 Ge 131.7705 ±20.7
It is interesting to note that the focal length of the BK7 lens increases with
the allowable increased temperature of 9.4°C from 100 mm to 100.004 mm,
while the focal length of the silicon lens reduces from 100 mm to 99.908 mm.
For the germanium lens, the focal length shortens to 99.729 mm.
§
The Rayleigh criterion states that a wavefront error of a quarter wave does not
noticeably affect the image quality.
88 Chapter 9
Figure 9.7 Performance plots for a diffraction-limited germanium lens in an

aluminium housing at a design temperature of 20°C.
Figure 9.8. Performance plots for the lens at 40.7°C. The quarter wave optical
path difference (OPD) is indicated.
The performance plots of the germanium lens at the design temperature of

20°C and at the elevated temperature of 40.7°C are shown in Fig. 9.7 and Fig.
9.8, respectively.
Reference
Chapter 10
The Ball Lens
10.1 Introduction
A ball lens is an interesting optical element worth a closer look. Figure 10.1
indicates the optical behaviour of such a lens. It can be seen that the principal
planes fall together and are located at the center of the sphere. For the
aberrations, we only consider here the case where the object is located at infinity.
Some remarks will be made for other situations. The focal length is
Rn
f  , (10.1)
2  n  1
and the back focal length is
bfl  f  R 
2  n R , (10.2)
2  n  1
Figure 10.1 Ball lens. The white lines suggest that the lens could be made into a
cylinder for easier mounting.
89
90 Chapter 10
where R is the radius of the sphere, and n is the refractive index of the lens
material.
Deriving the third-order aberrations is straightforward and instructive. We

will do that here and limit ourselves to the discussion of the spherical aberration,
coma, and astigmatism.
10.2 Spherical Aberration

The contribution of the first surface is
n1  n'1  n1 
TSC1  y1  u'1  i1  i12 , (10.3)
2n'1n'2u'2
recognizing that n1  n'2  1, and calling n'1  n , we find by paraxial ray tracing,
u'1    n  1 y1  nR  , i1  y1 R , and u'2   2  n  1 y1  nR  . Inserting all these
relations into Eq. (10.3) results in the simple expression
y3
TSC1   , (10.4)
4nR 2
where we substituted y for y1 . For the second surface we go about it in the same
way. The surface contribution is
n2  n'2  n2 
TSC2  2 y2  u'2  i2  i22 . (10.5)
2n'2 n'2u'2
The substitutes are

n2  n and n'2  1 ,
y2 
2  n y ,
n
y
i2  ,
Rn
and as before,
2  n  1 y
u'2  .
nR
The Ball Lens 91
With that we get
TSC2 
 n  2  2n  1 y 3 . (10.6)
4n 2 R 2
Adding these two surface contributions together leads to the straightforward

expression for the third-order transverse spherical aberration of the ball lens. We
remember that y is the ray height at the first surface as indicated in Fig. 10.1.
Transverse spherical aberration is
 n  n  3  1 y 3
TSCball   . (10.7)
lens 2n 2 R 2
We recall that the blur spot size due to spherical aberration is Bspher  TSC 2 , or
angularly expressed
n  n  3  1
βspher  . (10.8)
128  n  1  f / #
2 3
ball
In this format, we recognize that the relative aperture  f/#   f 2y .
10.3 Coma
To get the expression for coma, all we have to do is to multiply the spherical
aberration surface contributions with the factor i p1 i1  i p 2 i2  u p R y . The
symbol u p is the half-field angle. This leads to
 n  n  3  1 y 2u p
CCball  . (10.9)
lens 2n 2 R
Angularly expressed, with Bcoma  CC , the extent is
 n  n  3  1 u p f
β coma   . (10.10)
16n  n  1 f / # 
2
ball
92 Chapter 10
10.4 Astigmatism
By multiplying the coma expression once more with the factor
i p1 i1  i p 2 i2  u p R y , we get the third-order astigmatic contribution. The result
is
 n  n  3  1 yu 2p
TAC   . (10.11)
2n 2
The minimum astigmatic blur size is Bastig  2TAC . The angular measure is then
ball
 n  n  3  1 u 2p
β astig   . (10.12)
ball 2n 2  f / # 
Example
After that somewhat laborious preparation, we demonstrate the use of these
equations by designing a ball lens, made from BK7, for use with a HeNe laser ( 
= 0.6328µm). The index of refraction of BK7 is, for that wavelength, n  1.5151 .
We employ a relative aperture of f/2.5, and a rather large half-field angle of
u p  0.05 rad  2.8645 deg. The ball radius R = 2.5 mm.
The results are:
Transverse spherical aberration TSC = −0.0173 mm

Blur-spot size, linear Bspher  0.0087 mm
Blur-spot size, angular βspher  2.35 mrad
Coma CC = −0.0029 mm
Blur-spot size, linear Bcoma  0.0029 mm
Blur-spot size, angular β coma  0.80 mrad
Transverse astigmatism TAC = −0.0005 mm

Blur-spot size, linear Bastig  0.001 mm
Blur-spot size, angular β astig  0.272 mrad
The focal length f = 3.6767 mm, and the back focal length bfl = 1.1767 mm.
The Ball Lens 93
Figure 10.2 Image locations inside a ball lens, a function of the refractive index.
From Eq. (10.2) it can be seen that if n > 2, the focus falls inside the lens.
Therefore, a ball lens is not suitable for normal use if it is made from zinc
selenide, silicon, or germanium when the object is located at infinity. This is
indicated in Fig. 10.2. The image locations can be derived by using Eqs. (1.7)
and (1.9) from Chapter 1. With n being the index of the ball material, and
choosing the ray height y = 1, Eq (1.7) changes to nu'   n  1 R , and from Eq.
(1.9) we get u'  1 / l' . This leads to l'  n /  n  1 R  .
Table 10.1 Image distances inside the ball lens (n>2) for an object at infinity.
Material Index of refraction n Image distance l′

LASF35 2 2R (rear surface)
Zinc Selenide 2.4 1.7143 R
Silicon 3.4 1.4167 R
Germanium 4.0 1.3333 R
For the image to fall on the backside of the ball lens, the object has to be
placed at a distance l in front of the lens. This distance is a function of the lens
material and is expressed by
2R
l . (10.13)
 n  2
Table 10.2 lists this distance for selected materials as a function of the ball
radius. Figure 10.3 shows the relative positions for the different materials.
94 Chapter 10
Table 10.2 Object distances for image to fall on the rear side of a ball lens.
Material Index of refraction n Object distance l

Glass 1.5 -4R (virtual)
Zinc Selenide 2.4 5R
Silicon 3.4 1.4286R
Germanium 4.0 R
Figure 10.3 Limits of object distances for different materials.

Chapter 11
Seidel and the Pegel Diagrams
11.1 Introduction
In 1856, Ludwig von Seidel published his famous formulae for the primary
aberrations. They are based on the truncation of the series expansion for the sine
of an angle to sin x  x  x3 / 6 . That is why these calculations are also referred to
as third-order aberration calculations. This method has been applied in evaluating
the performance of the presented examples in the previous chapters. Besides
providing relatively quick answers, one additional important feature of these
calculations is that they point out the individual surface aberration contributions.
This, in turn, indicates the sensitivity to radii, thickness, and spacing tolerances.
The desire is to have the “work” equally divided among the surfaces; meaning
one likes to avoid wild swings between under correction and overcorrection from
surface to surface. While the sheer numbers of these surface contributions give
the full information, a graphic depiction, which is called a Pegel diagram, is even
more telling (Pegel is the German word for level). Another name for this form of
representation may be “aberration distribution diagram.” Of course, the ideal case
would be if all surfaces would be free of aberrations. Since this is not the case,
aberration balancing rather than aberration correction more accurately describes
the effort to improve the image quality.
11.2 Triplet for the LWIR Region

We shall demonstrate aberration balancing by analyzing a triplet with a cold stop
for the LWIR. The layout of the objective is shown in Fig. 11.1.
Figure 11.1 An f/1.4 triplet, with a 100-mm focal length and a field of ±3 deg.
95
96 Chapter 11
11.2.1 Surface contributions

We limit the presentation to spherical aberration, but remember that the other
Seidel aberrations can be shown in the same manner. The surface contributions
for spherical aberration are listed in Table 11.1.
Table 11.1 Seidel surface contributions in mm.
Surface 1 2 3 4 5
TSC +0.008224 +0.149244 –0.629787 +0.088954 –0.004288
Surface 6 7 8 9 10
TSC +0.404949 –0.582071 +0.571206 –0.307748 +0.305024
The sum of the surface contributions is
10
TSCtotal   TSC  0.003707 mm .
1
11.2.2 Pegel diagram for the triplet

The diagram in Fig. 11.2 shows the summation of the spherical aberration from
surface to surface. After the last surface, the size of the remaining system
aberration is the above stated 0.003707 mm.
Figure 11.2 This diagram clearly shows the aberration contributions of the
individual lens surfaces.
Seidel and the Pegel Diagrams 97
11.2.3 Remarks to field curvature

Astigmatism is well corrected for this lens, so that the field curvature is basically
represented by the Petzval curvature.
Josef Max Petzval, an Hungarian mathematician, is known for this theorem,

which states that in absence of astigmatism there is a field curvature, expressed
by the radius
h' 2
ρ , (11.1)
2 PC
where h′ is the image height, and  PC is the sum of the third-order longitudinal
Petzval contribution. For our example lens with a 100-mm focal length and a
3-deg half-field angle, h'  f tan u p  100  tan 3deg  5.241 mm .  PC is
determined by the same method as the other third-order aberration contributions.
We find  PC to be –0.04357. This leads to a Petzval radius of
ρ  h' 2  2 PC   5.2412  2   0.04353   315.5 mm . The negative sign
indicates that the curvature is bent inwards, i.e., towards the lens. This inward
bending of the field curvature was especially pronounced in the earlier cameras
where the objectives were made up of but a few elements, or even just one. To
compensate for this defocusing effect at the edge of the image, photographers at
the time used the trick in group pictures to align the assembly in an arc. This is
demonstrated in Fig. 11.3.
Figure 11.3 Petzval curvature “applied” for the object. (Shown in the picture are
some of the author’s professors from the Academy of Applied Sciences in
Munich.)
98 Chapter 11
Figure 11.4 Maksutov-Mangin combination for the LWIR region.
11.3 Cassegrain Version with a Maksutov-

Mangin Mirror Combination for the LWIR
Region
A more complex design example is shown in Fig. 11.4. The focal length of the
f/2.15 system is 554 mm. The field coverage is ±0.5 deg. The track length
(overall length from the first surface to the focal plane) is 384 mm. All elements
are germanium with the exception of one, which is made from zinc selenide. The
center portion of the inside Maksutov window is the secondary mirror of the
basic Cassegrain configuration. There are no aspheric surfaces in the system.
11.3.1 Pegel diagram

We elected this more elaborate system to show how a Pegel diagram provides a
quick overview of the aberration contributions. We limit ourselves again to the
display of the spherical aberration. It can be seen at a glance how strong a
contributor the second surface is, and how the third one overshoots with its
compensation. At the end, the sum is exceedingly small, with TSC = +0.01616
mm. The plus sign indicates overcorrection.
The individual surface contributions are listed in Table 11.3, and the Pegel
diagram is shown in Fig. 11.5.
Seidel and the Pegel Diagrams 99
Table 11.3 Third-order transverse spherical aberration surface contributions.
Surface 1 2 3 4 5
TSC –0.040580 +2.838347 –4.117814 +0.858031 +0.417371
Surface 6 7 8 9 10
TSC –0.526075 –0.046702 +0.046419 +0.259682 –0.671524
Surface 11 12 13 14 15
TSC +0.784353 +0.001294 +0.127990 +0.026294 –0.103859
Surface 16 17 18 TSC
TSC +0.166109 –0.177425 +0.174250 +0.016161
Figure 11.5 Pegel diagram for the Cassegrain version depicted in Fig. 11.4.
Chapter 12
The Single-Imaging Mirror
12.1 Introduction
Included in this discussion are mirrors with spherical and aspheric surfaces,
which are single-surface reflectors, and the Mangin mirror, which is a
catadioptric element. The advantage of a pure reflector is that it is free from
chromatic aberration and therefore suitable for applications over the entire optical
spectrum. This property allows alignments of IR systems in the visible spectrum
because there is no focus shift.
12.2 Spherical Mirror

We apply again the third-order aberration equations. With the object at infinity
and the aperture stop at the mirror, the aberrations are
spherical aberration,
y3 f
TSC    , (12.1)
64  f / # 
2 2
8f
sagittal coma,
y 2u p up f
CC   , (12.2)
16  f / # 
2
4f
astigmatism,
yu 2p u 2p f
TAC    , (12.3)
2 4  f / #
Petzval curvature,
yu 2p u 2p f
TPC   , (12.4)
4 4  f / #
and distortion is zero.
101
102 Chapter 12
The blur-spot sizes are again expressed as before with
TSC f
Bspher   , (12.5)
128  f / # 
3
2
coma,
up f
Bcoma  CC  , (12.6)
16  f / # 
2
astigmatism,
u 2p f
Bastig  2TAC  , (12.7)
2  f / #
and Petzval radius,
h' 2 h' 2 2u 2p f 2
ρ   f . (12.8)
2 PC 2u 'TPC 2u 2p f
Figure 12.1 Sagittal focus moves along the tangent of the circle as a function of
the half-field angle up.
The Single-Imaging Mirror 103
For a better understanding of how astigmatism occurs at the spherical mirror

we derive the measure of the blur spot due to this aberration in another way.
Figure 12.1 identifies the locations of the sagittal and tangential foci. The
difference in their location is twice the longitudinal astigmatic contribution 2AC.1
One can see that
1 f  1 
AC   f s  ft     cos u p  .

(12.9)
2 2  cos u p 
The transverse astigmatic contribution TAC  ACu'K . With
1
u'  ,
2  f / #
and
AC
TAC  ,
2  f / #
therefore,
f  1 
TAC    cos u p  .
4  f / #   cos u p 

For small angles,
u 2p
cos u p  1  .
2
With this substitution,
f  2 2  u 2p  fu 2p
TAC      .
4  f / #   2  u 2p 2  4  f / #
104 Chapter 12
Figure 12.2 Three-dimensional view of the forming of astigmatism at a concave

spherical mirror for an object at infinity.
The smallest blur spot is Bastig  2TAC  fu 2p  2  f / #   , which is, of course,

the same as derived with the third-order aberrations equation. Figure 12.2 gives a
3D view of the condition.
12.3 Toroidal Mirrors

Figure 12.3 illustrates the generation of toroidal mirrors. It can be seen that a
spherical surface is a special case in which the sagittal and tangential radii are
equal. While the sketches indicate mirrors, the shapes are also suitable and used
for lenses.
12.4 Examples
We analyze the image from a spherical mirror, modify the shape of the mirror to
a toroid, and compare the results.
To keep the spherical aberration small, we use a slow system of f/10 with a
100-mm focal length. The half-field angle up = 0.05 rad, which is 2.864789 deg.
Figures 12.1 and 12.2 show that the tangential radius Rt has been shortened, and
Figure 12.3 Generation of toric surfaces.
the sagittal radius Rs has to be increased to make the two focal lengths fs and ft
equal to each other and to f. The relations are
2f
Rt  2 f s   (12.10)
cos u p
and
Rs  2 ft  2 f cos u p . (12.11)
106 Chapter 12
Figure 12.4 Geometry of the toroidal surface for our example.
Figure 12.4 shows the shape of a square toroidal mirror and indicates again the
generation of it.
12.4.1 Spherical mirror

f 100
Bspher    0.000781 mm
128  f / #  128  103
3
up f 0.05  100
Bcoma    0.003125 mm
16  f / #  16  102
2
u 2p f 0.052  100
Bastig    0.0125 mm
2  f / # 2  10
The Petzval radius ρ  f  100 mm .
12.4.1.1 Additional comments

To remove coma and astigmatism, one needs to place the aperture stop at the
center of curvature of the spherical mirror.1 Spherical aberration remains.
Figure 12.5 Coma, the remaining aberration.
12.4.2 Toroidal mirror

The tangential radius
2 f 2  100
Rt    200.250261 mm ,
cos u p cos(2.864789deg)
and the sagittal radius

Rs  2 f cos u p  2  100  cos(2.864789 deg)  199.750052 mm .
With astigmatism corrected (third-order), a very slow relative aperture (f/10),

and a small field, the remaining aberration is essentially only coma. According to
Eq. (12.6), it is
up f 0.05  100
Bcoma   0.003125 mm .
16  f / #  16  102
2
This is the size of the sagittal coma. The tangential coma blur is three times as
large, i.e., 0.009375 mm. This is confirmed in Fig. 12.5, which is the result of a
real ray trace.
12.5 Parabolic Mirror

A typical application for a single-imaging mirror is a collimator. Figure 12.6
shows an arrangement frequently found in an optical laboratory for testing the
108 Chapter 12
performance of an infrared system. The imaging mirror is a section of an

paraboloid. The only purpose of the flat-fold mirror is to achieve a more
convenient overall arrangement. In such a device, the focal length is made rather
long, the f/# large, and the pinhole small. Spherical aberration is corrected by the
parabolic shape of the imaging mirror. The off-axis aberrations can be ignored
due to the very narrow field. The pinhole should be as small as possible to
represent a point source, but sufficiently large enough to provide the required
energy from the black body for reliable performance measurements of the system
under test.
Figure 12.6 Typical collimator arrangement for infrared system measurements.
Figure 12.7 Oscillating Slow Tool Servo generating a saddle-shaped reflector on

a single-point diamond turning ultra-precision lathe. Movement of cutting tool
z = f(x,).
12.6 Manufacturing Remarks

Generating toroidal mirrors and off-axis conic sections is costly and difficult at
best. For many applications, such mirrors can be fabricated by single-point
diamond turning if the chosen substrate material is suitable for the process. A
typical candidate is electroless nickel-plated aluminium. With the introduction of
a special mechanism, called a Slow Tool Servo (STS), such free-form surfaces
can be readily machined on-axis. The trick is that the cutting tool moves in the z-
axis (axis of rotation) as a function of the lateral ordinate, the x direction, and
rotation angle  Figure 12.7 points out the meaning of these symbols and
illustrates the cutting of a saddle-shaped surface by this method.
12.7 Mangin Mirror

In 1876, the French Army officer A. Mangin invented this catadioptric element to
avoid the manufacturing difficulties associated with an aspheric surface. It can be
seen from Fig. 12.8 that the element is a negative meniscus with a reflecting
second surface. The principle of correcting spherical aberration is as follows: a
spherical mirror has undercorrected spherical aberration, and a negative lens is
overcorrected. By proper choice of the radii and the substrate material, spherical
aberration can be corrected. While the original application was intended for
illumination (search lights), the Mangin mirror has since been employed in other
areas. One example is the objective for a radiometer. It also can serve as a
secondary mirror in a modified Cassegrain telescope. This will be discussed as
the next chapter.
Figure 12.8 Mangin mirror, with the object at infinity.

110 Chapter 12
The power (reciprocal focal length) of the Mangin mirror is
1
   2  n  1 c1  2nc2 , (12.12)
f
with the curvatures (reciprocal radii) c1  1 / R1 and c2  1 / R2 . The refractive

index of the substrate material is n.
Expressing the second curvature as a function of the first surface, which is of

course also the third one, and assuming zero thickness for the element, leads to
2  n  1 c1  
c2  . (12.13)
2n
The second radius is then
1 2nR1 f
R2   . (12.14)
c2 2  n  1 f  R1
Finding the curvature c1 for which spherical aberration vanishes leads to the
cubic equation
 c1   n  3  c1   4n  5   c1   4n  3
3 2 2
          0. (12.15)
 2  4    8  n  1
The derivation of the aberration expressions and the analytical solution of Eq.
(12.5) are quite cumbersome2 but enlightening. Matters become much easier if
we set the curvature of the first surface equal, but opposite in sign, to the mirror’s
power, i.e., c1   , which means making the first surface radius R1   f . With
this postulation for the thin Mangin mirror, we get
2nR1 2nf
R2   . (12.16)
 2n  1  2n  1
The angular blur spots due to the different aberrations for this shape are
βspher 
 2n  3 , (12.17)
128n 2  f / # 
3
coma,
up
β coma  , (12.18)
16n 2  f / # 
2
astigmatism,
u 2p
β astig  , (12.19)
2  f / #
chromatic,
β chrom 
 n  1 , (12.20)
 2nV  f / #  
and Petzval radius,
n2 f
ρ . (12.21)
 2n 2
 1
V is the Abbe number, and up is the half-field angle in radians.
Remark
An interesting case exists when n = 1.5. From Eq. (12.17) we see that with n =
1.5, spherical aberration vanishes. This was recognized by Mangin.
Table 12.1 lists the radii required for different materials to eliminate
spherical aberration of the Mangin mirror when the object is located at infinity.
The results are obtained with Eqs. (12.14) and (12.15).
Table 12.1 Surface radii for Mangin mirrors, corrected for third-order spherical
aberration.
Material n R1 R2
Glass 1.5 −1.0 f −1.5 f
Zinc Selenide 2.4 −1.1506 f −1.3980 f
Silicon 3.4 −1.1790 f −1.3409 f
Germanium 4.0 −1.1823 f −1.3169 f
112 Chapter 12
References
(2008).
2. M. J. Riedl, “The Mangin mirror and its primary aberrations,” Appl. Opt.,
13(7), 1690–1694 (1974).
Chapter 13
Eight Single Optical Elements
as Imaging Objectives
13.1 Introduction
A single optical element is often adequate to meet the image quality expected,
especially from an infrared system, if the field to be covered is moderate. To
provide a reference for a quick evaluation and comparison, eight different singlet
configurations are presented in the form of a summary matrix of the previous
subjects covered. This matrix will indicate whether a single element is sufficient
for the task at hand or if a multi-element objective is required.
To make matters manageable and simple, it has been assumed that the object
is located at infinity and the aperture stop is placed at the objective. To avoid
unpleasant surprises, it is recommended to always check first the size of the
diffraction blur formed by the element. This is especially important for
applications in the infrared region where the wavelengths are five to ten times
longer than in the visible spectrum.
13.2 Diffraction Limit

The diameter of the diffraction blur, which contains 83.9% of the energy of an
imaged object point, is
Bdiffr  2.44 λ  f/#  , (13.1)
where  is the wavelength and f/# the relative aperture or “speed” of the
objective. This diffraction blur is also known as Airy disk, named after Lord
Airy, the British mathematician and astronomer.
13.3 Eight Chosen Configurations

The eight chosen single elements, as possible imaging objectives, are pictured in
Fig. 13.1. Aspheric lens #1 has a concentric second surface with respect to the
focal point. Therefore, rays entering the lens parallel to the optical axis are not
refracted at that surface. Aspheric lens #2 has been shaped to eliminate spherical
aberration and coma.
113
114 Chapter 13
Figure 13.1 Eight selected shapes for a single imaging element.
Figure 13.2 Basic relations for a thin lens (object at infinity).

Eight Single Optical Elements as Imaging Objectives 115
Before making any statement about aberrations, Fig. 13.2 identifies the
symbols and relations used in the expressions.
Table 13.1 Symbol definitions.
2y free aperture diameter

f focal length
f/(2y) f/# = relative aperture or f-number
n index of refraction of the lens material
V   nM  1 /  nS  nL  the inverse relative dispersion or Abbe number for the
infrared spectra (M = middle, S = short, and L = long
wavelength).
R1 and R2 first and second surface radii respectively
h′ image height
up half field angle
u′ final slope angle of axial ray
Table 13.2 Shapes of the elements.
Conf. Radius R1 Radius R2 Conic 1 Conic 2

1  n  1 f plano 0 0
2  n  2  n  1 2  n  2  n  1
2 f f 0 0
n  2n  1 n  2n  1  4
 n  1 f 
1
3 f n2
0
n
4
 n  1
2
f
n 2
 1
f 0
n3  n  1
 n  n  1  1
3
n2 n  n  1  1
5
n 2
 1
f
n 2
 1
f 0
n3  n  1
 n  n  1  1
3
n2 n  n  1  1
2n
6 f f 0 0
1  2n 
7 2 f - 0 -
8 2 f - –1 -
13.4 Shapes of the Elements

Table 13.2 lists the surface shapes of the eight elements. For the diffractive
hybrid (configuration 5), there is a phase profile superimposed on the second
surface. It consists of a number of diffraction zones (see Chapters 3 and 14). The
116 Chapter 13
radius of the first zone r1  2λf d , with  being the wavelength, and fd being the
focal length of the diffractive portion of the lens. The other zone radii of the
profile can be obtained with ri  r1 i . The step size at the transition from one
zone to the next is d max  λ /  n  1 .
13.5 Aberrations1
Remarks
1. The aberrations are stated as angular-blur spot sizes , in radians. The linear
blur-spot sizes are B   f .
2. The blur size, due to astigmatism, remains the same for all eight
configurations, namely β astig  u 2p  2  f / #   .
Table 13.3 Angular aberration blur spot sizes of the elements in radians.
Config. spher chrom coma

n  n  2  2
2
1  n  n  1  1 u p
1
32n  n  1 2
  f / # 3 2V  f / #  8n  n  1 f / # 
2
n  4n  1 1 up
2
128  n  1  n  2  f / # 
2 3 2V  f / #  16  n  2  f / # 
2
1 up
3 0
2V  f / #  8  n  1 f / # 
2
1
4 0 0
2V  f / # 
5 0 0 0
 2n  3   n  1 up
6
128n 2  f / #  2nV  f / #  16n 2  f / # 
3 2
1 up
7 0
128  f / #  16  f / # 
3 2
up
8 0 0
16  f / # 
2
Eight Single Optical Elements as Imaging Objectives 117
13.6 Examples
We determine the shapes and aberration blur-spot sizes for an element with f =
100 mm, f/# = 1.5, M = 4 m, S = 3 m, and L = 5 m. The half field angle
u p = 0.05 rad.
With silicon as the chosen lens material, the index of refraction n = 3.425,
and the Abbe number V = 236. For a point of reference, we determine first the
diffraction blur size.
Bdiff = 2.44  0.004  1.5  0.015 mm, which is angularly expressed as

diff  Bdiff / f = 0.15 mrad. Analyzing all eight configurations yields an
informative comparison matrix. For the definitions of R1, R2, and , see Tables
13.1 and 13.2. The angular blur-spot functions are stated in Table 13.3.
Table 13.4 Comparison Matrix. Linear dimensions are in mm. Angular

dimensions are in mrad.
# R1 R2  sph ch c a 
1 242.50  0 8.60 1.41 2.44 0.83 13.28
2 97.861 164.074 0 3.16 1.41 0.26 0.83 5.66
3 70.803 100.00 −0.08525 0 1.41 1.15 0.83 3.39
4 91.475 146.882 0.45595 0 1.41 0 0.83 2.24
5 91.475 146.882 0.45595 0 0 0 0.83 0.83
6 −100.0 −117.094 0 0.76 1.00 0.12 0.83 2.71
7 −200.0 - 0 2.32 0 1.39 0.83 4.54
8 −200.0 - -1 0 0 1.39 0.83 2.22

118 Chapter 13
There are 12 zones for the diffractive hybrid, Element 5. Their radii are listed
below.
Table 13.5 Radii for the 12 diffractive hybrid zones.
Radius # r (mm) Radius # r (mm)

1 9.757 7 25.815
2 13.799 8 27.597
3 16.900 9 29.271
4 19.514 10 30.855
5 21.817 11 32.360
6 23.900 12 33.799
The maximum step dimension at the transition radii is d max = 0.00165 mm.
Reference
Chapter 14
A Progression of Performance with an
Increase in Lens Complexity
14.1 Objectives
We apply the knowledge gained from the preceding sections by analyzing six
different objectives for the MWIR spectrum with the same focal length (50 mm),
the same f/# (1.5), and the same square field coverage (4 deg × 4 deg). We start
with the simplest form of a lens and reach (in five steps) an objective that is
frequently found in thermal imaging systems.
The six configurations are:
1. plano-convex singlet
2. best shaped singlet
3. aspheric singlet
4. aspheric hybrid
5. Petzval objective with one aspheric element
6. Petzval objective with one aspheric hybrid
The plot in Fig. 14.1 is based on using square sensor elements (pixels) that
contain 80% of the received energy from a point source at infinity. It is clear that
the simple singlets are not suitable for IR cameras, but are sufficient for
radiometers (pyrometers) if the target is large enough to resolve the projected
sensor element. The single aspheric hybrid with a 70-µm pixel size over a linear
field of view of 3.5 mm in the horizontal and vertical directions may begin to get
interesting for thermal imaging. The total number of pixels amounts
to (3.5 / 0.07)2  2500 , and the angular resolution is 0.07/50 =1.4 mrad. At the
end of the list is the frequently employed Petzval objective, which was discussed
in detail in Sec. 7.4. With a 20-µm pixel size, the focal plane array contains more
than 30,000 sensing elements. The angular resolution is 0.4 mrad.
Remarks
1. For the off-axis calculations, the field was increased by the factor 2 to
cover the pixels located at the extreme of the field’s diagonal.
2. All lenses are made from silicon with the exception of the aspheric hybrid,
which is made from GASIR-1, a material that lends itself to molding. Its
refractive index is 2.5116 at 4 µm, and its Abbe number in the MWIR region
is 197.3.
119
120 Chapter 14
3. For reference, notice that the Airy disk, containing 84% of the energy from
an object point, is 15 µm for the stated wavelength of 4 µm and the assumed
relative aperture of f/#  1.5 .
4. For an increased field of view and a lower f/#, elements have to be added to
these basic designs with a choice of making use of the properties of different
materials.
Figure 14.1 Performance improvement with increasing complexity of the

objective.
Chapter 15
Two-Mirror Systems as Telescope and
Microscope Objectives
15.1 Introduction
In 1663, the Scottish mathematician and astronomer James Gregory suggested a
two-mirror system with a parabolic primary and an elliptical secondary mirror.
The configuration is shown in Fig. 15.1. At that time, he was unable to find an
optician to grind and polish the elliptical mirror.
Nine years later, in 1672, the Frenchman Guillaume Cassegrain came up with
the idea to intercept the reflected rays from the primary mirror before they come
to a focus. For that purpose, a hyperbolic secondary mirror is needed.
Cassegrain’s system is much shorter and has found its place in many
applications—from astronomical telescopes to thermal imaging devices. It is
interesting to note that Isaac Newton, the great British scientist, never gave
Cassegrain any credit for his invention. He referred to it as “just an obvious
extension” of Gregory’s concept. If the field of view is narrow, only spherical
aberration has to be corrected. Dutchmen Horace Dall and Allan Kirkham
conceived of the idea to replace the hyperbolic secondary mirror with an easier-
to-manufacture convex spherical mirror. To keep the system free from spherical
aberration, the primary mirror becomes, in that case, an ellipsoid. This increases
coma somewhat, but that was not of much concern to the two astronomers. In the
1920s, the American George Ritchey and the Frenchman Henri Chretien
modified the basic Cassegrain arrangement so that it is corrected for spherical
aberration and third-order coma. This Ritchey-Chretien configuration is now the
most widely used combination. The famous Hubble Space Telescope is of this
Figure 15.1 Gregorian telescope.
121
122 Chapter 15
type. We shall discuss the details of all these configurations, including the use of
a Mangin mirror as a secondary reflector. Additionally, we will present the
classic Schwarzschild objective and a special configuration, the inverted
Schwarzschild, which is corrected for spherical aberration, coma, astigmatism,
and field curvature.
15.2 Basic Cassegrain Telescope Layout

A convenient way of identifying the layout of a Cassegrain system is by choosing
a desired focal length f, the spacing d of the two mirrors, and the back focal
distance b. The separation d and the back focal distance b depend on the
application. If, for example, a 25% obstruction ratio (in area) from the secondary
mirror is acceptable, d will be half of the primary focal length f or a quarter of the
primary mirror radius R1. Another factor to consider is whether there is a need to
insert, for example, filters or a folding reflector behind the primary mirror, which
would influence the length of the back focal distance b. Figure 15.2 shows the
typical Cassegrain arrangement and identifies the symbols used in the
subsequently presented equations. For reference and comparison, the location and
size of the secondary mirror of an equivalent Gregorian telescope is indicated
with reference to Newton’s remark.
Figure 15.2 Layout of a Cassegrain telescope.

Two-Mirror Systems as Telescope and Microscope Objectives 123
15.2.1 Equations1
From the geometry of the layout and a paraxial ray trace one finds the radii of the
mirrors. They are
2df
R1  (primary mirror) (15.1)
b  f 
and
2db
R2  (secondary mirror). (15.2)
b  d  f 
15.3 Cassegrain with Two Spherical Mirrors

With spherical surfaces for both mirrors, the angular third-order aberration blur
spots are
f  b  f   b  f  d  b  f  d  b 
3 2
βspher  , (15.3)
128 fd 3  f /# 
3
coma,
 2 f  b  f 2   f  d  b  f  d  b  d  f  b   u p
β coma   , (15.4)
32d 2 f  f /# 
2
and astigmatism,
 4bf  b  f    f  d  b  d  f  b 2  u p 2
β astig   . (15.5)
8 fbd  f /# 
The Petzval radius is
bfd
ρ Petzval  . (15.6)
fd   b  f 
2
124 Chapter 15
15.4 Classic Cassegrain System

To correct spherical aberration, the surface of the primary becomes a paraboloid,
for which the conic constant κ1  1 . The conic constant for the hyperbolic
secondary mirror is
2
 f  d b
κ2     . (15.7)
bd  f 
Coma for this configuration is expressed by
up
β coma  , (15.8)
16  f /# 
2
and astigmatism is
 d  f  u 2p
β astig  . (15.9)
2b  f /# 
The Petzval radius remains the same with
bfd
ρ Petzval  .
fd   b  f 
2
15.5 Dall-Kirkham Arrangement

We recall that this form has a spherical secondary mirror. Spherical aberration
correction is achieved with an aspheric primary mirror. The conic constant is
f  f  b   b  f  d  b  f  d  b 
3 2
κ1  . (15.10)
f b  f 
3
Coma is
u p  2 f  b  f    f  d  b  f  d  b  d  f  b  
2
β coma   , (15.11)
32d f  f /# 
2 2
and astigmatism is
u 2p  4bf  b  f    f  d  b  d  f  b  
2
β astig   . (15.12)
8dbf  f /# 
15.6 Ritchey–Chretien Configuration

As stated earlier, spherical aberration and third-order coma are corrected in the
Ritchey–Chretien configuration. The required conic constants are, for the primary
mirror
2bd 2
κ1  1 , (15.13)
b  f 
3
and for the secondary mirror
 2 f  b  f 2   f  d  b  f  d  b  d  f  b  
κ2   . (15.14)
b  d  f 
3
Left is astigmatism with
u 2p  2 f  d 
β astig  . (15.15)
4b  f /# 
15.7 Examples
For performance comparison, we analyze several different telescope systems
with the identical focal length of f = 500 mm, a separation of the elements of d =
100 mm, and back focal distance of b = 150 mm. The relative aperture is f/5, and
the half-field angle up = 0.01 radians. The mirror radii are, according to Eqs.
(15.1) and (15.2),
2df 2 100  500
R1    285.714286 mm
 b  f  150  500 
and
2db 2 100  150
R2    120.0000 mm .
 b  d  f  150  100  500 
The Petzval radius is the same for all configurations, namely
bfd 150  500  100
Petzval    103.448 mm.
fd  (b  f ) 2
500  100  (150  500) 2
126 Chapter 15
Table 15.1 Conic constants and angular blur-spot sizes for chosen two-mirror
systems.
Configuration   spher coma astig

Basic Cassegrain 0 0 1.73 0.152 0.01
Classic Cassegrain –1 –3.24 0 0.025 0.027
Dall-Kirkham –0.645775 0 0 0.152 0.01
Ritchy-Chretien –1.069971 –388 0 0 0.03
The angular blur spot sizes spher , coma , and astig are in milliradians.
We analyze now a Cassegrain system that uses a Mangin mirror as a

secondary reflector. The focal length, spacing, back focal distance, f/#, and the
covered field are the same as in the previous examples.
15.8 Cassegrain with Mangin as a Secondary

Reflector
This configuration uses only spherical surfaces. The third-order spherical
aberration is corrected. Due to the refractive secondary element, chromatic
aberration is introduced. An analytical approach to solve this task is not practical.
Therefore, a solution was created with ZEMAX. Figure 15.3 shows the on-axis
blur spot size, which is due to chromatic aberration. The system was designed for
the MWIR region with silicon as the material for the Mangin mirror. As
mentioned, spherical aberration has been corrected with the Mangin mirror. The
blur-spot shape on the right side clearly shows that the predominant contributor
to the blur is coma. The data are for a half-field angle of up = 0.01 radians.
Figure 15.3 Blur spots from a Cassegrain system with a Mangin mirror as a
secondary reflector.
One can see what improvement a Mangin mirror brings if the blur-spot sizes
are compared with the ones from a two-sphere mirror system, where the spherical
aberration blur alone is 0.865 mm when refocused to the best position. This can
be confirmed with Eqs. (15.3), (15.4), and (15.5).
15.9 Gregorian Telescope

All the equations presented for the Cassegrain and Ritchey–Chretien systems
remain valid for the Gregorien telescope, with the requirement that the focal
length f is inserted as a negative number. Maintaining the focal length f = 500
mm, the f/# with 5, and the image height h′ = 5 mm, we determine first with the
help of Fig. 15.4 the required spacing d of the two elements and the back focal
distance b. From the ratios, one finds d = 220 mm, b = 270 mm, and y2 = 27 mm.
We now can calculate the radii of the mirrors with Eqs. (15.1) and (15.2), and
the conic constants, for which we apply Eqs. (15.13) and (15.14) to correct
spherical aberration and coma. Recall that changing the sign of the focal length
for the equations leads to
2df 2  220   500 

R1    285.714286 mm
 b  f   270   500  
and
2db 2  220  270

R2    120.0 mm.
 b  d  f   270  220   500 
Figure 15.4 Change in d, b, and y as shown transforms a Cassegrain system to

Gregory system. The sketch is not to scale.
128 Chapter 15
The conic constants are
2  270  2202
κ1   1  0.942751 ,
 270  500 
3
and
κ2 
2   500  270   500   990   550   450  0.358524.
2
 270  220   500  

3
Astigmatism is
u 2p (2 f  d ) 0.012   2  (500)  200

astig     0.000022 rad.
4b( f /#) 4  270  5
While this astigmatic blur is very small, the Petzval curvature is steep with a
radius of
bfd 270   500   220

ρ Petzval    42.2535 mm .
fd   b  f   500   220  270   500 
2 2
For the classical Cassegrain and R/C systems, with the same focal length and f/#,
the Petzval radius is –103.4483 mm.
After refocusing by a small amount, the geometrical blur spots are still
smaller than the diffraction limit because of the relatively slow speed of f/5 and
the wavelength of 4 µm that we had assumed. The Airy disk is
BAiry  2.44λ  f /#   2.44  0.004  5  0.0488 mm .
We expand on this system by describing a Gregorian system that was

cleverly shortened by inserting a flat folding mirror to place the secondary mirror
at the location of the primary mirror. The system was developed by the fSONA
Systems Corporation in Canada. Figure 15.5 shows the arrangement. The primary
and secondary mirrors form a monolithic element, which was fabricated by
single-point diamond turning on an ultra-precision lathe. In Fig. 15.6 we show
the ray traces, based on the Gregorian system we designed above for the MWIR
region.
Figure 15.5 Folded Gregorian telescope.
Figure 15.6 Folded Gregorian system, designed with the provided equations. (f =
500 mm, d = 220 mm, b = 270, folded to d/2 = 110 mm.)
130 Chapter 15
15.10 Gregorian Microscope Objective

It is rewarding to use the same equations to design an inverted Gregorian system
that can serve as a microscope objective, for which as before, spherical aberration
and coma are corrected.
Let us demonstrate with f = 20 mm, d = 55 mm, and b = 60 mm.
Figure 15.7 Inverted Gregorian as a microscope objective.
Figure 15.8 Blur-spot sizes for the inverted 20-mm-focal-length, f/1.5 Gregorian
objective. The reference circle is the size of the Airy disk.
Figure 15.9 Compact folded Gregorian microscope objective.
The calculations yield surface radii of R1  27.5 mm and R2  48.888 mm .

The conic constants are κ1  0.29106 and κ 2  0.083473.
Figure 15.7 shows the layout, and Fig. 15.8 depicts the linear blur-spot sizes,
compared with the diffraction limit in the visible spectrum.
We now present a folded variation of this objective. For this case, the
effective focal length is 10 mm, and the relative aperture f/# =1.5. The folding
mirror has been slightly aspherized to aid in the correction of the off-axis
aberrations. A possible layout, as a so-called snap-together arrangement, suitable
for diamond turning, is shown in Fig. 15.9. Notice the large free-working
distance.
15.11 Two Schwarzschild Objectives

The classic Schwarzschild objective has been discussed in many textbooks and is
presented here for comparison with the lesser known second configuration. The
first kind uses only spherical surfaces, while for the second kind, two aspheres
are required. Until recently, aspheres have been very expensive to manufacture.
This may have been a reason to avoid this second configuration. The two types
shown in Fig. 15.10 have the same focal length and the same relative aperture.
132 Chapter 15
Figure 15.10 Two Schwarzschild configurations.

The first striking difference is the size. Another one that is not quite as
obvious is the fact that in the second form, the “inverted” configuration, there is
no need for a spider to hold the secondary mirror in place as is the case with the
classic one. The first kind has a large free-working distance which is frequently
required. On the other hand, the short free-working distance of the second kind is
no shortcoming for applications in, for example, photolithography.
15.11.1 Prescription for the classic configuration for an

object at infinity
The conditions required to correct spherical aberration, coma, and astigmatism
for this configuration are:
 d 2f
 b  52 f
 R1   
5 1 f
 R2   5 1 f 
f
 y1 
2  f /# 
 52 f 
 y2   
5  2 y1 
2  f /# 
 Fractional area obscuration ε  1 / 5  20%
Note that both mirror surfaces are spherical and concentric to each other. It is
an interesting exercise to demonstrate that the third-order spherical aberration,
coma, and astigmatism vanish for this configuration. To eliminate the Petzval
curvature and distortion, one can position a field flattener in the focal plane with
a front radius of
n2  1
R3  f
1  n  n  1
and a rear radius of R4   n 2  1 f . It is also remarkable to note that in this

configuration, the mirror radii can be derived from the “golden ratio.” Let us
demonstrate with the help of Fig. 15.11.
134 Chapter 15
Figure 15.11 Relations of the “golden ratio.”
The diagram shows the relations of the “golden ratio,” which says,
2f x
 .
x 2 f  x
The root of this equation is x1  f ( 5  1) , which is equal to the radius of the

first mirror. The radius of the second sphere can then be easily derived since it is
longer by 2f and concentric to R1.
Therefore,
R2  R1  d  R1  2 f   
5 1 f  2 f   
5  1 f . **
15.11.1.1 Brief historic remark about the golden ratio2

As early as 1202, the Italian mathematician Leonardo Fiboncci discussed in his
book Liber abacihis a series that leads to the ratio stated above. This ratio was
rediscovered toward the end of the 19th century and found especially pleasing for
its harmonic proportions.
Example
For a system with f = 10 mm and a relative aperture of f/1, we get d = 20 mm, b =
42.3607 mm, R1 = 12.3607 mm, R2 = 32.3607 mm, y1 = 5 mm, and y2 = 21.1803
mm.
15.11.2 Prescription for the inverted configuration for

an object at infinity
In addition to correcting spherical aberration, coma, and astigmatism, field
curvature is also eliminated in this configuration:
**
See the fourth equation on page 133, under 15.11.1.
 d  2 f (Same as above)
 
b  1 2 f 
 R1  R2  2 2 f  
 
2
 κ1  1  2 = 5.828427 (Conic constant for first surface)
 
2
 κ2  1 2  0.171573 (The reciprocal of κ1 )
f
 y1 
2  f /# 
 y2 
1  2  f
2  f /# 
 
y2  1  2 y1 
 
2
 Fractional area obscuration ε  2  1  1 / 5.8  17.3%
Note that both surfaces’ radii are equal. By aspherizing, both become oblate
ellipsoids.
Figure 15.12 Inverted Schwarzschild objective. The layout suggests two

diamond-turned components with a spidered disk inserted to block the central
energy from flooding the focal plane.
136 Chapter 15
Figure 15.13 Solid catadioptric microscope objective, suggested by D.D.

Maksutov. The chosen material is CaF2, which can be diamond turned.
Here is another example for the same focal length and f/#. The system is
shown in Fig 15.12. d = 20 mm as before. b = 24.1421 mm, and R1 = R2 =
28.2843 mm. The conic constants are independent of focal length and f/#. They
are 1 = 5.828427 and 2 = 1/1 = 0.171573.
15.12 Solid Microscope Objective

To close this chapter, we present a solid microscope objective as suggested by
Maksutov3 and shown in Fig. 15.13. With calcium fluoride as the material, the
objective can be used in the visible spectrum as well as in the MWIR region. The
large radius is aspherized with two higher-order deformation terms. The entrance
and exit surface radii are concentric to the axial object and image points to
eliminate refraction.
References
(2008).
2. Bibliographisches Institut & F. A. Brockhaus AG, Germany (2007).
3. K. Mütze, ABC der Optik, Verlag Werner Dausien, Hanau, Germany (1961).
Chapter 16
The Plane-Parallel Plate
16.1 Introduction
There are two aspects that have to be dealt with when using a plane-parallel plate
inserted in a converging or diverging beam. First, there are the aberrations, and
second there is a relocation of the image. This relocation is a simple longitudinal
shift if the plate is perpendicular to the optical axis. If the plate is tilted, as is the
case with most beam splitters, there is a longitudinal and a lateral displacement,
which is not too easy to assess because they interact. We shall address both cases.
It should also be clear that most prisms behave like thick plane-parallel plates.
This is indicated in Fig. 16.2 with a number of examples. The “unfolding” of the
prism into an equivalent plane parallel plate is called the tunnel diagram. The
first two examples are equivalent to plane-parallel plates inserted perpendicularly
to the optical axis. The third one, the Dove prism, corresponds to a tilted plate,
which of course introduces astigmatism and coma as will be pointed out further
down in the text.
For several reasons, such as cost and absorption, prisms are not much used in
the infrared spectrum. However, wedges, which are thin prisms with a small apex
angle, are employed in this region as beam-steering elements. The deviation of a
ray exiting such a wedge is    n  1  , where n is the index of refraction, and α
is the wedge angle. Details are shown in Fig. 16.1.
A pair of counter-rotating wedges, known as Risley prisms, is used as a

linear-scanning device.
Figure 16.1 Thin prism (wedge) and the deviation of an incoming ray.
137
138 Chapter 16
16.2 Aberrations1
It is frequently overlooked that a plane-parallel plate also contributes aberrations
when it is inserted in a converging or diverging beam. Using the third-order
surface-contribution equations again, we demonstrate step by step the derivation
for the spherical aberration. The symbols used in the derivation are identified in
Fig. 16.3
The transverse spherical aberration surface contribution is
TSC  Bi 2 h' , (16.1)
where
n  n'  n 
B y  u'  i  ,
2n'n'k Inv
where i  cy  u , and h′ is the image height.
The total transverse spherical aberration is the sum of the two surface
2
contributions  TSC  TSC
1
1  TSC2 . We make things easier if we set y1  0 ,
2
because then TSC1  0, and  TSC  TSC
1
2 . Equation (16.1) can now be restated
as
 n  1 y
B 2  u'2  i2  i22 h' .
2nInv
For the optical invariant Inv we set h'n'2u'2   h'  2  f /#   and

y2  y1  tu'1   t  2n  f /#   , with y1  0, and u'1  u1 n  1  2n  f /#   .
 
2
i22  1 /  2n  f /#    1 /  4n 2  f /#  
2
With and recalling that the
minimum blur size, when refocused, is half the transverse spherical aberration,
we finally get the blur size of a plane-parallel plate with thickness t and index n.
It is simply
Bspher 
n 2
 1 t
. (16.2)
32n3  f /# 
3
The Plane-Parallel Plate 139
Figure 16.2 Tunnel diagrams of three prisms show the relation to a plane-parallel
plate.
Figure 16.3 The plane-parallel plate.

140 Chapter 16
The blur-spot sizes due to coma, astigmatism, axial chromatic, and lateral color
can be derived by the same procedure. They are
coma,
Bcoma 
n 2
 1 tu p
, (16.3)
8n3  f /# 
2
astigmatism,
Bastig 
n 2
 1 tu 2p
, (16.4)
2n3  f /# 
axial chromatic,
Bchrom 
 n  1 t , (16.5)
2n 2V  f /# 
and lateral color,
 n  1 tu p
Blat color  , (16.6)
n 2V
where up is the half field angle, and V is the Abbe number.
These effects need to be taken into account, especially if filters with different
thicknesses and indices are inserted in the converging beam. The Dewar window
of a cooled detector package is another example of that.
The impact is amplified when the plane-parallel plate is used as a beam

splitter under 45 deg.
16.2.1 Examples
In Table 16.1, we compare the performance of three 2-mm-thick beam splitters
made from different materials, inserted in a converging f/2 cone. For the visible
spectrum we use BK7, for the MWIR range from 3 µm to 5 µm, we contrast
silicon against sapphire.
Figure 16.4 Blur spot sizes of a 2-mm beam splitter, tilted about 45 deg. The
circles indicate the calculated blur-spot sizes.
Table 16.1 Blur spot sizes as a function of material and spectral range.
Material Region Bspher Bcoma Bastig Bchrom Blat color Btotal

BK7 VIS 2.9 18.3 115.5 1.8 5.5 143.5 µm
Silicon MWIR 2.1 13.1 82.2 0.4 1.4 99.2 µm
Sapphire MWIR 3.0 18.9 118.5 15.8 49.7 205.9 µm
The calculated sizes, using the developed equations, are compared with the
results from a full computer analysis in Fig. 16.4.
The scaled blur-spot diagrams from the computer confirm the validity of the
simple third-order expressions. The plot for the sapphire substrate shows clearly
the impact of the very high dispersion for that material in the MWIR. Especially
pronounced is the lateral color, which is already indicated in the table.
16.3 Shift of Image

Figure 16.5 shows a plate perpendicular to the optical axis and indicates the
longitudinal displacement of the image. This shift is a function of the substrate
material and the thickness of the plate. The longitudinal displacement is
 1  sin 2 u 
L   1  2 t . (16.7)
 n  sin 2 u 
 
For small angles this reduces u to
l 
 n  1 t . (16.8)
n
142 Chapter 16
Figure 16.5 Longitudinal image shift caused by a plane-parallel plate.
To assess the error introduced by substituting Eq. (16.8) for (16.7), we

compare the results for different relative apertures (f/#).
From the relation tan u  D 2 f  1 2  f /#  , we get the slope angle u for the
insertion in Eq. (16.7). Table 16.2 lists the errors in percent for a glass plate with
a refractive index n = 1.5. Table 16.3 indicates that the difference is much smaller
for silicon, for which the index n = 3.4 in the MWIR region.
Table 16.2 Error resulting from substituting Eq. (16.8) for small angles u. n =1.5
(glass)
Relative Slope Exact image For small Error in %

aperture f/# angle u in displacement L angles l   L  l 
deg  100
L
4 7.125016 0.336208 t 0.333333 t 0.86
3 9.462322 0.338419 t 0.333333 t 1.50
2 14.036243 0.344614 t 0.333333 t 3.27
1.5 18.434949 0.353003 t 0.333333 t 5.57
1 26.565051 0.375305 t 0.333333 t 11.18
Table 16.3 Error resulting from substituting Eq. (16.8) for small angles u. n = 3.4
(silicon).
Relative Slope Exact image For small Error in %

aperture f/# angle u in displacement L angles l   L  l 
deg  100
L
4 7.125016 0.707959 t 0.705882 t 0.29
3 9.462322 0.709544 t 0.705882 t 0.52
2 14.036243 0.713935 t 0.705882 t 1.13
1.5 18.434949 0.719761 t 0.705882 t 1.93
1 26.565051 0.734628 t 0.705882 t 3.91
16.4 Tilted Plate

Figure 16.6 indicates the effect of a tilted plate. In addition to the longitudinal
shift, there is also a lateral displacement of the image.
Figure 16.6 Displacements caused by a tilted plate.

144 Chapter 16
Figure 16.7 Image displacements due to tilt and thickness of a plane-parallel

plate.
16.4.1 Lateral displacement

The lateral or transverse displacement is
 1  sin 2 ut 
T   1  2  t sin ut . (16.9)
 n  sin 2 ut 
 
ut is the tilt angle, as indicated in Fig 16.6.
The combined effect of transverse displacements and longitudinal shifts as a

function of the tilt angle is shown in Fig. 16.7. The adjustments for locating the
minimum blur spots are indicated as obtained from actual retracings. A 20-mm-
thick plate has been chosen to magnify the excursions.
16.5 Two Tilted Plates

To correct for the astigmatism introduced by a tilted plate, one can insert a
second plate, made from the same material, with the same thickness and the same
tilt but rotated by 90 deg, as shown in Fig. 16.8. The offset of the optical axis as
stated in Eq. (16.9) will then be the same in the x-axis as in the y-axis. We
demonstrate the situation with two germanium plates tilted by 45 deg. The first
serves as a beam splitter, and the second is inserted to correct astigmatism. The
Figure 16.8 The arrangement for correcting astigmatism.
Figure 16.9 Blur-spot comparison with one and two plates arranged as shown in
Fig. 16.8. Also indicated is the best longitudinal image position.
plates are 2 mm thick and are inserted in an f/2.5 cone. The remaining aberrations
are negligible, except coma, which is approximately 20 µm [according to Eq.
(16.3)] for the two plates. The results are compared against the performance of a
beam splitter without a compensator. Figure 16.9 compares the blur-spot shapes
and sizes of a single plate with a two-plate arrangement.
Reference
Chapter 17
MTF, Limits, and Pixel Sizes
17.1 Introduction
The optical modulation transfer function (MTF) is a well-known concept for
evaluating the performance of a lens.1 It contains valuable information about the
resolving power of the lens and the contrast in the image. Modern infrared
imaging systems use sensor arrays that are made up of small individual sensing
elements, the pixels. The practical minimum size of such a pixel depends on the
wavelength of the collected radiation and sets the spatial resolution limit of the
optical system, referred to as the Nyquist frequency.
17.2 Optical Modulation Transfer Function

The MTF for diffraction limited systems is expressed by 1,2
          
2
2
MTFdiffr ( )  arccos      1     , (17.1)
π  0   0    0  

where v is the spatial frequency of interest,  0 is the cut-off frequency, and

 /  0 is the normalized frequency. Figure 17.1 shows the graph of the
diffraction-limited MTF as a function of the normalized frequency  /  0 . In Fig.
17.1 there is also a remark about the Nyquist frequency, which will be discussed
in the following section.
The cut-off frequency is expressed by
1
0  (17.2)
λ  f /# 
This indicates that for the visible spectrum with  = 0.5 µm, the cut-off
frequency in line pair/mm v0VIS  2000 /  f /#  . In the MWIR region, with  = 4
µm,  0MWIR  250  f /#  . For the LWIR spectrum, with  = 10 µm,
 0LWIR  100  f / #  . These numbers are reminders of the sizeable differences
among the different spectral regions.
147
148 Chapter 17
Figure 17.1 Diffraction-limited MTF.
17.3 Focal Plane Array

Figure 17.2 schematically shows the geometry of a focal plane array (FPA) with
square pixels.
With a common postulation that the pixel size should be not smaller than the
Airy disk, we can state that p  2.44λ  f /#  . The Nyquist frequency is then
1
 Ny  . (17.3)
5λ  f /# 
Forming the normalized cut-off frequency, relating to Nyquist, we can write
 0Ny 1 / 5λ  f /#   1
  .
0 1 /  λ  f /#   5
In other words, the normalized Nyquist cut-off frequency is five times lower than
the optical one. This is marked in Fig. 17.1.
Referring to the different spectral regions, Eq. (17.3) states that the limiting
Nyquist frequency, expressed in line pairs per millimeter, is 50/(f/#) for the
MWIR region and 20/(f/#) for the LWIR window.
MTF, Limits, and Pixel Sizes 149
Figure 17.2 Focal plane array. Columns 3 and 4 form a line pair 2p wide. The
spatial frequency is the Nyquist frequency  Ny  1 / (2 p) .
References
(2008).
Chapter 18
Details of a Hybrid Lens
18.1 Introduction
Hybrid lenses are being frequently used in infrared systems, primarily for color
correction. We shall analyze one such lens, which is the third element of the
MWIR objective shown in Fig. 18.1.
18.2 Hybrid1
The prescription of the diffractive surface (front surface) is as follows:
1. Radius R1 = 21.246406 mm
2. Conic constant  = –0.151513 (prolate ellipse)
3. Phase coefficient P2 = −5975.55048.
λP2 0.004   5975.550548 

4. S2    3.804  104
2πR0 2
2π 1002
5. Free aperture D = 11.84 mm
For reference, we state the general phase equation of a diffractive surface as
Figure 18.1 25-mm focal-length wide-angle objective with a cold stop for the
MWIR region. The relative aperture is f/1.4, and the field coverage is ±12 deg.
The third lens is a germanium hybrid, which will be analyzed.
151
152 Chapter 18
2π
r  
λ0
 S 2 r 2  S4 r 4  S6 r 6  ... . (18.1)
Note that, for this design, only the quadratic term of the phase function has been
used, and the asphere is a conic section without additional higher deformation
coefficients. This has been done to simplify the analysis without impacting the
correctness of the general approach.
For this equation, r is the zone radius of the diffractive profile and λ 0 is the
design wavelength, which in our example is 4 µm.
The first zone radius can be obtained from the relation
λ0
r1  . (18.2)
S2
The other zone radii are found with
ri  r1 i . (18.3)
The total number of zones required over a free lens aperture is
2
 D
itotal   . (18.4)
 2r1 
The step height of the profile at the zone transition is
λ0
d max  . (18.5)
 n  1
We now apply all these relations to our example and find
0.004
r1   3.42657 mm .
3.804 104
The other radii are r2  r1 2  4.5858 mm , r3 = 5.6164 mm, r4 =6.4853 mm, and
r5 =7.2508 mm. Equation (18.4) states that
Details of a Hybrid Lens 153
2 2
 D  13.6 
itotal      4.397 ,
 2r1   2  3.24657 
which means the fifth zone is not a full zone. Only a fraction of it will be used.
The zone transition step size is
λ0 0.004
d max    0.001322 mm .
 n  1  4.02506  1
18.3 Coordinates of the Combined Surface

To generate the profile of the aspheric/diffractive surface with an ultra-precision
lathe, one needs to know the axial tool position z as a function of the lateral cross
position r. Details for this statement are given in Fig. 18.2. The conic asphere is
expressed by
cr 2
zasph  , (18.6)
1  1   κ  1 c 2 r 2 
 
with c = 1/R, R = vertex radius, and  = conic constant.
The diffractive profile is represented by
1  2 S2 2 
zdiff   S 2 r  λ 0 Int r  . (18.7)
 n0  1  λ0 
Figure 18.2 Coordinates for a diffractive profile on an aspheric surface.

154 Chapter 18
Let us apply this to a point at r = 5 mm:
zasph 
1/ 21.24606   52  0.000669 mm
1  1   0.151513  1  1/ 21.24606   52 
2
  
zdiff 
1 
  3.804 10   5  0.004  Int
4 2  3.804  104  2
5


 4.02506  1  0.004 

1
zdiff    0.009510  0.004  2   0.000499 mm
3.02506
ztotal  zasph  zdiff = 0.000669 - 0.000449 = 0.00022 mm
A good check is to insert these equations at a known transition radius. We

demonstrate with r4 = 6.4853.
1
zdiff  
 4.02506  1

 3.804  10   6.4853  0.004  Int
4 2
 3.804  104 
 6.48532


 0.004 
 
If we insert the third radius with 5.616400 mm in the integer (Int) bracket,
we get zdiff  1  4.02506  1   0.016  0.012   0.001322 mm . This is the
maximum step height as stated with Eq. (18.5). Figure 18.3 shows performance
plots of this objective.
Details of a Hybrid Lens 155
Figure 18.3 25-mm MWIR objective with germanium hybrid lens.
Reference
Chapter 19
From the Höegh Meniscus to
Double Anastigmats
19.1 Introduction
This chapter outlines the design steps for several photo objectives in an
evolutionary way. Beginning with a single element, the Höegh meniscus, we
progress by adding elements. Two elements lead to the Hypergon, a
symmetrically arranged doublet. In the next step, the two lenses are shaped into
cemented achromats. Finally, these cemented achromats are split into airspaced
achromats, which leads to an arrangement referred to as a double anastigmat. In
the final optimization, we break the symmetry and change the radii of the second
achromat. The configuration is an excellent general-purpose lens with a relative
wide angle coverage.
19.2 Höegh Meniscus

Emil von Höegh reasoned that if both radii of a meniscus are equal, the Petzval
curvature must be zero. Furthermore, if the stop is placed at the proper position,
astigmatism can be eliminated. This arrangement is shown in Figure 19.1.
Figure 19.1 Höegh Meniscus with aperture stop placed in front.
157
158 Chapter 19
19.2.1 Approach and design method

Third-order transverse Petzval surface contributions of the two meniscus surfaces
are
  n1  n1'  c1h'Inv
TPC1 
2n1n1'
and
  n2  n'2  c2 h'Inv
TPC2  .
2n2 n'2
With c2  c1  c, n1  n2'  1, and n2  n1'  n , we find
 1  n  ch'Inv
TPC1 
2n
and
  n  1 ch'Inv
TPC2   TPC1 .
2n
Therefore,  TPC  0. (All symbols have been identified in Chapter 4).

The focal length of this special meniscus can be easily derived from equation
(2.1), which is
1 
  n  1 c1  c2 
 n  1 c c t  .
1 2 
f  n 
By inserting c2 = c1, we get
2
 r  n
f   .
 n 1  t
For a given focal length, we solve for the thickness t and obtain
2
 r  n
t   .
 n 1  f
From the Höegh Meniscus to Double Anastigmats 159
We select for our example r = −10 mm, n = 1.5, and f = 100 mm. This results in a
thickness of t = 6 mm.
To correct astigmatism, the aperture stop has to be placed so that

 TAC  0. To find that proper place in front of the lens, as indicated in Fig.
19.1, is a somewhat cumbersome (but interesting) exercise.
Assuming a relative aperture of f/30, and a half-field angle of 40 deg, we

trace axial and principal rays through the lens with an estimate of the stop
position. We chose a first stop position at l p  5 mm and then one at
l p  6 mm . u p1  tan 40 deg  0.8391 . The third-order transverse astigmatism
TAC is 0.1842 mm if the stop is located at lstop  5 mm , and −0.1190 mm if it is
at lstop  6 mm .
By interpolation we determine that TAC is 0 when lstop  5.627 mm . All

third-order aberrations as a function of the stop position are summarized in Table
19.1. We add the plots showing the astigmatism for the three stop positions,
which are graphically demonstrated in Figure 19.2.
For a better understanding of the impact of the material selection, we repeat

the calculations with a glass that has a higher index of refraction and a higher
Abbe number. Our choice is Schott glass LAKN13 with an index of n = 1.6935
and an Abbe number of V = 53.33. The high index leads to a thinner lens
(reduced from 6 mm to 3.5 mm) and a lower lateral color as well, even though
the Abbe number is not as large as that used in the first calculation, where n was
assumed to be 1.5 and V was 62.5. With a somewhat lower Abbe number, the
axial chromatic aberration is slightly larger. All this can be seen from the data in
Table 19.2.
Table 19.1 The third-order aberrations.
lp TSC CC TAC TPC DC TAchC TchC

−5.000 −0.0998 −0.2759 0.1842 0 −7.7747 −0.0498 −0.3579
−6.000 −0.0998 −0.3262 −0.1190 0 −7.7192 −0.0498 −0.3829
−5.627 −0.0998 −0.3074 0 0 −7.6860 −0.0498 −0.3736
Table 19.2 Impact on third-order aberrations by changing the lens material.
Glass TSC CC TAC TPC DC TAchC TchC

5:625 −0.0998 −0.3074 0 0 −7.6860 −0.0498 −0.3736
LAKN13 −0.0840 −0.2133 0 0 −6.0809 −0.0542 −0.3595
160 Chapter 19
Figure 19.2 Astigmatism as a function of stop position.
19.3 Hypergon Lens

Two Höegh lenses mounted symmetrically about the aperture stop in a double
meniscus arrangement is known as the Hypergon lens, as shown in Fig. 19.3. For
comparison, we maintain the focal length, the f/#, and the field coverage for our
calculations.
The principle of symmetry results in coma, lateral color, and distortion being
corrected as well but only at unit magnification. Plots of the OPDs shown in
Figure 19.4 indicate that this objective performs rather well.
The third-order and chromatic aberrations are tabulated below. They indicate
that the f/# has to be kept large due to a large spherical aberration.
Table 19.3 Third-order and chromatic aberrations.
TSC CC TAC TPC DC TAchC Tch

−0.08026 0.00564 −0.00019 0 −0.01661 −0.04732 −0.01896
Figure 19.3 Hypergon lens (double meniscus).
Figure 19.4 Performance of a Hypergon lens.
19.4 Achromatic Double Lens

For an achromatic double lens, we maintain the symmetry but replace the singlets
with cemented achromats. For variety, we reduce the f/# from f/30 to f/20, and the
field from ±40 deg to ±20 deg.
162 Chapter 19
Figure 19.5 Symmetrically spaced achromats.

We keep the two achromatic lenses identical and mount them symmetrically
about the aperture stop as shown in Figure 19.5.
With fictitious indices and Abbe numbers, and chosen surface radii, lens
thicknesses, and stop position, the starting layout prescription is summarized in
Table 19.4. We use a substrate with a high index of refraction and a low
dispersion for the positive front element A, and a material with a high dispersion
and low index material for the negative rear element B. Such a combination is
known as the “new achromat”.
Table 19.4 Prescription of starting layout. Radii and thicknesses are given in
millimeters. Notice that the outside radii are all the same, namely 26 mm.
Surface Radius Thickness n V

1 26
2 1.9 40
2 37
2 1.6 35
3 26
6
4 (stop) 
6
5 −26
2 1.6 35
6 −37
2 1.9 40
7 −26
With the help of ZEMAX, we vary the radii and substitute real glasses. For
element A, we select Schott glass N-LASF31, with nA = 1.880609 and VA =
41.0098. For element B, we pick TIFN5, with nB = 1.593555 and VB = 35.5145.
Finally, we scale the lens so that we obtain the desired 100-mm focal length. The
result is summarized in Table 19.5.
Table 19.5 Radii and spacings of the scaled lens.
Surface Radius Thickness

1 23.4103
2.6617
2 45.9339
2.0709
3 21.0019
14.2406
4 (stop) 
14.2406
5 -21.0019
2.0709
6 -45.9339
2.6617
7 -23.4103
Optimizing the input data from Table 19.5 yields the configuration shown in
Figure 19.6.
Figure 19.6 Optimized achromatic double lens.

164 Chapter 19
The performance plots for the achromatic double lens are shown in Figure
19.7 in the form of OPDs.
Figure 19.7 Performance of an achromatic double lens.
Figure 19.8 Double Anastigmat (Celor).

19.5 Double Anastigmats

Splitting the cemented achromats and providing airspace between elements A
and B provides additional degrees of freedom for aberration correction. Such a
configuration goes back more than 100 years, when it was designed by Höegh at
the Goerz Company and offered under the name Celor.1
The lens consists of two positive outer elements and two negative inner
elements. The aperture stop is located between the negative elements, as shown
in Fig. 19.8.
For this exercise, we decide on a 50-mm focal length and a relative aperture
of f/5 with a half-field coverage of 20 deg.

We start with a system consisting of two thin achromats, each containing half of
the system’s power.
For the positive element A we pick SSKN5 as the material, and we select
LF5 for the negative element B. The indices and V values of the chosen materials
are nA = 1.658 and VA = 50.9, and nB = 1.581 and VB = 40.9, respectively. This
combination is again a “new achromat”.
Since the total system power is   1 / f  1 / 50  0.02 mm 1 , the power for

one achromat is  AB   / 2  0.02 / 2  0.01mm 1 .
Neglecting the spacing between element A and B, and assuming thin lenses,
the requirements for the achromat are
VA 50.9
A   AB   0.01  0.0509 mm 1
VA  VB   50.9  40.9 
and
VB 40.9
B   A    0.0509  0.0409 mm 1 .
VA 50.9
Therefore,
1 1
fA    19.646365 mm
 A 0.0509
166 Chapter 19
and
1 1
fB    24.449878 mm .
 B 0.0409
For an equi-convex element A, the radii are
RA1  2  nA  1 f A  2  1.658  1  19.646365  25.854616 mm
and RA2  25.854616 mm . Similarly, we calculate the radii for the equi-
concave element B and find RB1   RB 2  28.410758 mm .
We put this prescription into the computer and add a thickness of 2 mm for
the positive elements and 1 mm for the negative elements. The stop is placed
symmetrically between the two negative elements, which are 2 mm apart. After
setting the half-fields for 14 deg and 20 deg, we make the radii variables but
maintain symmetry (i.e., the radii of the elements after the stop are identified as
pick-ups). For the first optimization, we set the operand for the focal length and
use the default merit function. The result after this first optimization is indicated
in the graphs in Figures 19.9 and 19.10.
Figure 19.9 MTF after first optimization.

Figure 19.10 Ray intercept curves after the first optimization.
Figure 19.11 MTF after second optimization.

168 Chapter 19
Figure 19.12 Ray intercept curves after the second optimization
For the second optimization, we vary all radii, which breaks the symmetry.
This asymmetric configuration is now called a Dogmar.2 The improvement is
shown in Figures 19.11 and 19.12.
We end this chapter with the understanding that adding elements and varying
curvatures, thicknesses, and spacings, as well as lens element materials, improve
performance even more.
References
1. R. Kingslake, Lens Design Fundamentals, Academic Press, New York
(1978), page 236.
2. W. J. Smith, Modern Optical Engineering, 4th Edition , McGraw-Hill, New
York (2008), page 486.
Index
Abbe number, 18 focus shift
for the infrared spectra, 115 of a diffractive lens, 74
aberrations, 116 of a refractive element, 73
chromatic, 18 frequency
achromats, 39 cut-off, 147
air-spaced normalized, 147
doublet, 65 Nyquist, 149
triplet, 65 spatial, 147
Airy disk, 113
AMTIR-1, 58 Gregorian
aspherizing, 23, 52 microscope objective, 130
a singlet, 26 telescope, 127
astigmatism, 19, 104 G-sums, 15
athermats, 73
half field angle, 115
ball lens, 89 hybrid, 69
bending a lens, 11 lens, 151
best shape, 11 petzval objective, 67
for minimum spherical aberration,
23 image height, 115
impact of housing material, 80
Cassegrain
basic telescope layout, 122 limits, 147
classic system, 124 lens
with two spherical mirrors, 123 thick, 34
chromatic aberration, 18 thin, 11
cold stop, 67 LWIR region, 50
coma, 19
curvature, 11 manufacturing remarks, 109
materials for the 3–5 µm spectral
Dall-Kirkham arrangement, 124 band (MWIR), 77
detector, 16, 21 materials for the 8–12 m spectral
Dewar housing, 67 band (LWIR), 78
dialyte, 57 microscope objectives, 121
diamond-turned hybrid, 50 minimum blur, 17
diffraction limit, 113 multiple lens arrangement, 23
diffractive phase profile, 52 MWIR region, 48
f/#, 115 n, 12
focal plane array, 148 net curvature, 11
169
170 Index
Newton, Isaac, 121 Seidel

nomograms, 55 diagrams, 95
Nyquist frequency, 149 surface contributions, 96
shift of image, 141
optical modulation transfer function, silicon, 141
147 single optical elements, 113
single-imaging mirror, 101
parabolic mirror, 107 single-point diamond turning, 108,
paraxial focal plane, 21 109, 128
Pegel diagrams, 95, 98 slow tool servo, 108
performance plots, 59 Smith’s method, 45
Petzval spherical aberration, 16
curvature, 101
objective, 65 telephoto, 62
phase telescope objectives, 121
coefficients, 61 thermal glass constant, 74
equation, 76 thick lens, 34
pixel sizes, 147, 148 thin lens, 11
plane-parallel plate, 137 concept, 33
power, 11 toroidal mirrors, 104
prisms, 137 transverse spherical-aberration
contribution, 12
relative aperture, 113 third-order, 35
reversed telephoto, 62 tunnel diagrams, 139
Ritchey-Chretien configuration, 125 two Schwarzschild objectives, 131
two separated components, 57
sapphire, 141 two-mirror systems, 121
About the Author
Max J. Riedl was born in Kempten, Germany, where he
was educated at the Mathematisch Mechanisches
Institut. He also graduated from the Akademie für
angewandte Technik in Munich, where he studied
precision mechanics and optics. He has worked in the
field of mathematical and optical instruments for many
years and holds numerous patents for mechanical,
optical, and electro-optical devices. He held technical as
well as business leadership positions during the 45 years he lived in the United
States. He has published and presented many papers on fundamental lens design
as well as on replicating optical elements and diamond turning diffractive optical
components for applications in the infrared spectrum. He regularly teaches short
courses on these subjects. He is the author of the well-received tutorial text
Optical Design Fundamentals for Infrared Systems (SPIE Press, Vol. TT48),
which is now in its second edition. He has also translated that textbook into
German. He is a Fellow of SPIE. Presently, he resides in Bavaria, Germany.

Optical Design Applying The Fundamentals

Uploaded by

Copyright:

Available Formats

You might also like

Optical Design Applying The Fundamentals

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Optical Design Applying The Fundamentals

Uploaded by

Copyright:

Available Formats

Tutorial Texts Series

 Optical Design Applying the Fundamentals, Max J. Riedl, Vol. TT84

For a complete listing of Tutorial Texts, visit http://spie.org/tutorialtexts.xml

Bellingham, Washington USA

Copyright © 2009 Society of Photo-Optical Instrumentation Engineers

All rights reserved. No part of this publication may be reproduced or distributed

Printed in the United States of America.

When a proposal for a text is received, each proposal is evaluated to determine

Chapter 1 Law of Refraction:

1.1 Introduction ............................................................................ 1

Chapter 2 Best Shape for a Thin Lens ....................................... 11

2.1 Concept of Thin Lens............................................................. 11

Chapter 3 Best Shapes for Multiple Thin Lenses,

3.1 Introduction ............................................................................ 23

4.1 Introduction ............................................................................ 33

Chapter 5 Achromats .................................................................. 39

5.1 Introduction ............................................................................ 39

Chapter 6 Systems with Two Separated Components ............. 57

6.1 Introduction ............................................................................ 57

7.1 Introduction ............................................................................ 65

Chapter 8 A Hybrid for Two Wavelengths ................................. 69

8.1 Introduction ............................................................................ 69

Chapter 9 Athermats ................................................................... 73

9.1 Introduction ............................................................................ 73

Chapter 10 The Ball Lens.............................................................. 89

10.1 Introduction ............................................................................ 89

Chapter 11 Seidel and the Pegel Diagrams ................................. 95

11.1 Introduction ............................................................................ 95

Chapter 12 The Single-Imaging Mirror .........................................101

12.1 Introduction ............................................................................ 101

Chapter 13 Eight Single Optical Elements

13.1 Introduction ............................................................................ 113

Chapter 14 A Progression of Performance with

14.1 Objectives .............................................................................. 119

Chapter 15 Two-Mirror Systems as Telescope

15.1 Introduction ............................................................................ 121

Chapter 16 The Plane-Parallel Plate .............................................137

16.1 Introduction ............................................................................ 137

Chapter 17 MTF, Limits, and Pixel Sizes ..................................... 147

17.1 Introduction ............................................................................ 147

Chapter 18 Details of a Hybrid Lens ............................................151

18.1 Introduction ............................................................................ 151

Chapter 19 From the Höegh Meniscus

19.1 Introduction ............................................................................ 157

1.2 Fermat’s Principle1

n' sin i'  n sin i , (1.1)

Figure 1.1 Schematic for Fermat’s theory.

x a  x sin i sin i'

Since the index of refraction is defined as

which is, of course, Snell’s law with n = 1 (air).

1.3 Snell and the Lens

Figure 1.2 Snell’s law applied to a curved lens surface.

1.4 Graphical Ray Tracing

Figure 1.3 Procedure of graphical ray tracing.

1.5 Paraxial Ray Tracing

For small angles,

and Snell’s law reduces to

1.5.1 Equations, symbols, and sign conventions