Comparison study of a multispectral zoom lens using standard and novel optical materials

Abstract

We demonstrate two broadband multispectral infrared (3.5–11.5 μm), zoom ( 3 × ) systems with focal lengths adjustable from 50 mm to 150 mm. Both systems are successful in meeting the modulation transfer function (MTF) requirement of 20 lp/mm. The difference between the two designs is that one employs novel infrared-transparent glasses that permit the designer to achieve an improved system performance with dramatically fewer lens elements. The impact of these materials on the design performance is discussed in terms of MTF and chromatic focal shift as a function of temperature, and we conclude with a brief description of these new glasses and their optical functionality.

Full text

Comparison study of a multispectral zoom lens
using standard and novel optical materials
JAMIE RAMSEY,* GEORGE LINDBERG,PETER WACHTEL,J.DAVID MUSGRAVES,AND JOHN DEEGAN
Rochester Precision Optics, 850 John Street, West Henrietta, New York 14586, USA
*Corresponding author: jramsey@rpoptics.com
Received 3 April 2019; revised 30 May 2019; accepted 30 May 2019; posted 31 May 2019 (Doc. ID 364140); published 19 June 2019
We demonstrate two broadband multispectral infrared (3.5–11.5 μm), zoom (3×) systems with focal lengths
adjustable from 50 mm to 150 mm. Both systems are successful in meeting the modulation transfer function
(MTF) requirement of 20 lp/mm. The difference between the two designs is that one employs novel infra-
red-transparent glasses that permit the designer to achieve an improved system performance with dramatically
fewer lens elements. The impact of these materials on the design performance is discussed in terms of MTF and
chromatic focal shift as a function of temperature, and we conclude with a brief description of these new glasses
and their optical functionality. © 2019 Optical Society of America
https://doi.org/10.1364/AO.58.005045
1. INTRODUCTION
Traditional optical design has focused on the visible portion of
the spectrum since the advent of optics [1]. Because of this
design focus on a comparatively narrow spectral region
(400–800 nm), there has been little need to examine the behav-
ior of optical materials across additional spectral regimes.
However, with the increasing availability of infrared (IR) opti-
cal systems in research, industrial, and commercial applications,
there is a commensurately increasing need to address the gaps
in our knowledge of how optical materials operate across multi-
ple IR bands [2–7]. In general, the IR spectrum is divided into
three windows: short-wave IR (SWIR) extending from 1.4 μm
to 3 μm, mid-wave IR (MWIR) from 3 μmto5μm, and long-
wave IR (LWIR) from 8 μmto15μm. Each of these IR bands
is valuable in probing or imaging under different conditions:
the SWIR is excellent for imaging through rain or fog, the
MWIR for hot objects such as jet plumes, and the LWIR
for ambient thermal imaging. One challenge for the IR optical
designer is to create a system that is capable of functioning
across multiple bands, which can encompass thousands of
nanometers of optical bandwidth, while maintaining manufac-
turability and performance requirements. The optical design
goal is the elimination of all thermal and optical aberrations,
required for high-end sensing and detecting applications
[8–13].
One limitation in the field of IR system design is the com-
parative lack of optical materials that transmit in this spectral
region. The visible system designer has a catalogue of hundreds
of glasses and crystals with which to design [14]. Each of these
materials has a specific refractive index, index dispersion,
thermo-optic coefficient, and more, that can be used to balance
the other materials in the system when eliminating optical aber-
rations. In contrast to this case, the IR system designer has a
suite of between six and 10 materials with which to design
[15,16]. This lack of materials forces designers to use many
more diffractive elements and many more total lenses in order
to achieve similar dispersion control and aberration reductions
as those achieved in visible system designs. In addition, there
are proven “rules of thumb”for optical designers using a “map”
of visible glass types for balancing aberrations and thermal
differences. Because the IR spans such a vast wavelength region,
some materials significantly change their position on the glass
map from the MWIR to the LWIR, moving from flint to crown
positions. This factor leads to the requirement of many more
optical elements in our Baseline system (defined below) because
some of the lens materials that have powerful impact on
dispersion in the MWIR are ineffective in the LWIR (and vice
versa) meaning that “extra”elements are required in order for
the system to operate in both bands. In addition, we find that
the inclusion of more negative dn∕dT materials into the de-
signer’s toolbox allows athermal designs to be created with
many fewer lens elements.
We present here a comparison study between two multi-
spectral IR zoom lens systems that were designed using (case
1) the standard commercially available materials, and (case
2) newer glass compositions that can aid in reducing the com-
plexity of the optical design process. Using the new glasses
allows the designer to create a system with a greatly reduced
size, weight, and power (SWaP) of the optical system.
Section 2discusses the designs of the two multispectral zoom
lens systems using different materials to reduce SWaP.
Section 3compares the optical performance of the two system
Research Article Vol. 58, No. 18 / 20 June 2019 / Applied Optics 5045
1559-128X/19/185045-05 Journal © 2019 Optical Society of America

designs. The new glass compositions are discussed briefly in
Section 4, but the goal of the present work is to demonstrate
the improvements in the optical design that can be achieved by
adding more materials to the designer’s toolbox.
2. SYSTEM DESIGN
This section presents a comparison study between two different
multispectral IR zoom lens systems designed using the same
basic parameters for field of view (FOV), wavelength range,
modulation transfer function (MTF) performance, and f/#.
The system designed using the standard set of materials is
designated the Baseline system, and the one designed using
the new glasses, the SWaP system. The minimum required
value for MTF performance was set to 20 lp/mm. Parameters
such as system length, number of aspheres, presence of dou-
blets, and total lens count were considered secondary and their
values were driven by the goal of achieving the best perfor-
mance possible. The lens systems were deigned to be parfocal
and athermal over all zoom positions. The moving groups were
not allowed to cross the image plane, thus eliminating any de-
signs where performance was lost for this reason, and for ease in
manufacturing. Another constraint on the Baseline system lens
design was that materials used needed to be manufacturable:
this constraint removed materials such as KRS5, CdTe, and
BaF from the design toolbox because most optics manufactures
will not grind, polish, or mold these materials for imaging op-
tics. The details of the two lenses are outlined in Table 1, details
of the materials used in each of the lens elements are presented
in Table 2, and some of the physical properties of these
materials are shown in Table 3.
Both the Baseline and SwaP-reduced systems were designed
in CodeV, an optical design software from Synopsys. The start-
ing parameters for both of the lens designs were the magnifi-
cation of the relay, stop diameter, and the back focal distance
from the image plane to the cold stop. The full system was
broken into two separate subsystems: the relay and the objec-
tive. The relay subsystem was designed first; it is this section
where the magnification comes into play and also where the
back-focus camera’s parameters are included. For the objective
subsystem, the image height is one of the known parameters
and is defined through the relay. The FOV is also taken into
account in this subsystem as well. The change in the FOV is
handled by moving two different groups of lenses to different
positions.
The following two subsections describe the Baseline and
SWaP system designs in greater detail.
A. Baseline Multispectral Design
The Baseline multispectral zoom lens system was designed us-
ing the standard IR-transparent materials that are commercially
available to the optical designer; the full package of IR-
transparent materials available in CodeV were considered for
this design. Upon completion of the design, this lens system
has 21 total elements in with eight elements in the relay
and 13 elements in the objective. There are two moving groups,
illustrated in Fig. 1, for the zoom movement, which changes
the focal length from 50 mm to 150 mm. The first moving
Table 1. Characteristics of the Baseline and
SWaP-Reduced Optical Systems
Characteristics Baseline System SWaP System
Zoom ratio 3×3×
f/# 3 3
Focal length range 50–150 mm 50–150 mm
Spectrum 3.5–11.5 μm 3.5–11.5 μm
Semi image diagonal 8.8 mm 8.8 mm
Length 254 mm 187 mm
Lens count 21 12
MTF performance @
20 lp/mm
>20% (nominal) >30% (nominal)
Aspheres 2 2
Mass—optics only (g) 416 211
Avg. Transmission (%) 53% 71%
Table 2. Materials for Each of the Lens Elements Shown
in Figs. 1and 2a
Lens
Number
Baseline
Lens Material
SWaP-Reduced
Lens Material
1 GaAs IRG27
2 IRG26 NRL4
3 IRG26 NRL6
4 IRG24 IRG24
5 Ge ZnS
6 ZnS NRL4
7 IRG27 NRL8
8 ZnS IRG24
9 Ge ZnS
10 IRG24 NRL8
11 ZnS IRG22
12 IRG22 NRL6
13 ZnS
14 IRG23
15 IRG22
16 IRG24
17 IRG27
18 Ge
19 IRG24
20 ZnS
21 IRG24
aLenses are numbered L to R in the figures.
Table 3. Selected Thermal and Optical Properties of the
Materials Used in the Two Systems
Material
CTE
(ppm/°C)
dn∕dT
(ppm/°C@3μm)
n
(@3 μm)
γ—Thermal
Glass Constant
GaAs 5.39 148 3.3128 58.6
Ge 5.9 424 4.0442 133.4
IRG22 12.5 68 2.5180 32.3
IRG23 13.4 106 2.8111 45.1
IRG24 20 23 2.6274 −5.9
IRG26 21.4 36 2.8015 −1.4
IRG27 22.5 −2 2.4218 −23.9
NRL4 30.2 −19 2.6485 −41.7
NRL6 18.3 164 3.1717 57.2
NRL8 25.7 0.3 2.6633 −25.5
ZnS 6.6 43 2.2581 27.6
ZnSe 7.6 62 2.4356 35.6
5046 Vol. 58, No. 18 / 20 June 2019 / Applied Optics Research Article

group has three elements, and the second moving group has
four elements. The lens system is 254 mm long, which includes
a cold stop distance of 25 mm and a window of ZnSe of 1 mm
thickness. The lens is optimized to operate from −40°Cto 85°C
with the design temperature being 20°C. This lens system is
illustrated in Fig. 1, where the top of the figure shows the
150 mm focal length, progressing downward through
the 100 mm zoom position to the 50 mm focal length at the
bottom of the figure.
Even though allowed by the design process, no doublets
were used in the Baseline design because of the mismatch in
the coefficient of thermal expansion (CTE) of the variety of
crystalline and glassy materials employed. Large CTE
differences can cause delamination of the doublets over the
broad temperature space we tested. As shown in Table 3, many
of the new Naval Research Laboratories (NRL) materials have
CTE values that are close enough to each other to survive this
delamination tendency and can successfully be employed in
bonded doublets.
B. SWaP-Reduced Design
The SWaP-reduced lens system was designed using the same
parameters as the Baseline system; however, it was designed us-
ing the new glasses discussed in Section 4. These materials aid
in the athermalization and achromatization of the lens system,
allowing for fewer elements to be utilized and delivering better
performance. In this embodiment of the design, there are six
lenses in the objective and six elements in the relay. There are
two moving groups to change the focus of the system, just as in
the Baseline design, but in this case, the first group has two
elements and the second group has three elements. This lens
system is 187 mm in length, 67 mm shorter than the
Baseline design that used standard materials.
Because we did not set length minimization as a primary
goal for this design, there is still sufficient length for the on-
axis performance to be corrected by the 12 lens elements.
Had we desired to minimize the length as well, the design likely
would have required more than the two aspheric surfaces used.
Future iterations of this design may strive to minimize the
length and improve the off-axis performance, in which case
more aspheric surfaces will be required to correct the full field
with shorter F/#.
3. SYSTEM PERFORMANCE
The aim of this effort was to design a lens system that met the
diffraction-limited MTF performance of 50% at 20 lp/mm
while minimizing the SWaP consumption of the full optical
system. This section lays out the SWaP reduction achieved
and discusses the optical performance (in terms of MTF and
chromaticity) of both lens systems.
A. SWaP Reduction
Size. The total length of the Baseline system is 254 mm for the
21 optical elements. This is an acceptable length based upon
the number of lenses and the element count while maintaining
performance across the zoom range. The total length of the
SWaP-reduced system is 187 mm for the 12 optical elements,
a reduction of 67 mm. Assuming that the diameters of the sys-
tems are identical, this represents a length (and thus volume)
decrease of 26%.
Weight. The mass of the Baseline system is 416 g for the 21
optical elements. The mass of the SWaP-reduced system is
211 g for the 12 optical elements. This represents a mass
reduction of 205 g, or 49% between the two lens systems.
Power. The principal sources of power consumption in the
lens system in Fig. 2are the piezoelectric motors of the zoom
and the camera that would be required to capture the image in
real-world systems. In the Baseline design (Fig. 1), the masses of
zoom groups 1 and 2 are 132 g and 64 g, respectively. In the
SWaP design (Fig. 2), the masses of these groups are 117 g and
11 g, respectively, and represent a weight reduction of 11% and
82%. The piezoelectric motors used to control the zoom
groups will experience commensurately lower power consump-
tion. The power consumption of the camera will also be im-
proved: the 18% increase in transmission (Table 1) will
improve the illumination of the focal plane and reduce the gain
level at which the camera needs to operate, minimizing power
consumption. The exact power consumption is extremely de-
pendent on the control motors and focal plane arrays chosen.
B. Modulation Transfer Functions
When comparing the performance of the two lens systems, the
Baseline lens, which has 21 elements, performs worse than the
Fig. 1. Layout of the Baseline system showing the three zoom
positions.
Fig. 2. Layout of the SWaP-reduced system showing the three zoom
positions.
Research Article Vol. 58, No. 18 / 20 June 2019 / Applied Optics 5047

SWaP lens with only 12 elements. In the near FOV (NFOV)
this performance is approximately 5% worse for the on-axis
field and 10% across all other fields; see Fig. 3(a). This perfor-
mance decrease occurs across the operating temperature range
and is a direct result of residual axial color still exhibited by the
Baseline system. In the middle FOV (MFOV), Fig. 3(b), the
performance losses are approximately 10% across all fields. The
performance loss across the temperature range, however,
remains constant.
For the wide FOV (WFOV), Fig. 3(c), the Baseline lens’
performance is approximately 10% lower than that of the
SWaP lens; however, the full field is significantly lower
(15% or more) than the rest of the fields in the Baseline lens.
The performance over the temperature range varies slowly, as it
did in the MFOV as well. This difference in performance over
the zoom positions is due to the fact that the NFOV is com-
promised slightly because it has more lenses than is required for
the MFOV or the WFOV.
C. Chromaticity
The loss in performance of the Baseline system versus that of
the SWaP-reduced system is illustrated in greater detail in
Fig. 4, where the chromatic focal shift is compared between
the two systems. The change in the curvature of the SWaP sys-
tem focal shift is due to the change from the system being apo-
chromatic in the NFOV and changing to an achromatic one as
it zooms. To correct for color over the thermal range, a tradeoff
is made in correcting for axial color and correcting for focus
over the thermal range. This tradeoff results in a performance
that is compromised at all temperatures due to the movements
of the zoom groups that are being used to refocus the system
and to the mechanical constraints.
During the initial design stages, axial color must be mini-
mized to produce a lens that is well corrected. This axial color at
the design temperature is one of the aberrations that varies over
the zoom position and is minimized only in each zoom posi-
tion. This is a result of each different zoom position being in a
compromised state: being too long for the desired focal length,
but having too many lenses for the given zoom position. This
occurs because some of the zoom positions need more lenses
and need to be longer than is required for the other positions.
At the start of the athermalization process, the lens system is
well corrected for chromatic aberration; however, when the ther-
mal proprieties of the materials are taken into account at the
different temperature ranges, degradation in focal shift occurs.
Fig. 3. Modulation transfer functions for the Baseline (red) and
SWaP-reduced (black) systems. (a) Near field, (b) mid field, and
(c) wide field of view.
Fig. 4. Chromatic focal shift for the two lens systems at all three
fields of view.
5048 Vol. 58, No. 18 / 20 June 2019 / Applied Optics Research Article

This focal shift can be small, well within the depth of focus of the
lens, or large and needs to be corrected actively. For the case of
the two lens systems used in this comparison study, the systems
are corrected actively by adjusting the two moving groups.
For the Baseline design, the lens was well corrected for axial
color at the design temperature. However when the tempera-
ture variation was included and the zoom groups were adjusted
to accommodate for the thermal focal shift of the system, the
lens system had a chromatic focal shift of 0.4 mm. This large
focal shift is a result of the two moving lens groups tending to
crash into each other in the WFOV position; hence, a compro-
mise was required for all zoom groups to minimize the effects of
the mechanical constraints.
For the Baseline lens system, the CTEs of the materials are
fairly closer together (see Table 2); however, the values of the
thermo-optical coefficient, dn∕dT, vary widely. This broad
variation in dn∕dT requires a very large movement in the
two mobile lens groups, and in the case of the Baseline lens,
the tendency of moving groups is to crash into each other during
design. To alleviate this issue, a minimum distance between the
moving groups was set. This in turn results in degradation in the
performance of the system resulting from more axial color.
For the SWaP-reduced lens with the new glasses these issues
did not occur because some of the materials had negative or
very small dn∕dT values, as well as CTE values that balance
out better (compared to the Baseline design). These values per-
mitted the system design to minimize the travel distance of
mobile lens groups and removed any of the adverse effects (such
as lenses crashing into each other), which in turn produces a
lens system that is well corrected over the temperature range.
4. OPTICAL MATERIALS
The improvements in SWaP and system performance demon-
strated in Section 3are a result of the inclusion of three new
materials to the designer’s toolbox: NRL4, 6, and 8 glasses.
These three materials have been developed at the NRL over
the last few years, and are not broadly commercially available,
but a selection of their properties has recently been made public
[17]. A list of all of the materials used in both systems is listed in
Table 1, and some of their properties are listed in Table 3.
Because one of the design goals for this multispectral zoom
system was to athermalize its behavior as far as possible, it is
critically important to know how the glass and crystal elements
behave as a function of temperature so that they can be
balanced against each other. This is expressed in terms of
the change, as a function of temperature, in the lens shape
(CTE) and its refractive index (dn∕dT).
A useful metric in the optical design process is the thermal
glass constant, γ:
γ
dn∕dT
n−1
−α,
where nis the refractive index, and αis the CTE of the glass.
This constant embodies the competition between the change in
the lens shape and the lens index as a function of temperature
for a given material; a positive value of γindicates that the focal
shift due to temperature outweighs the expansion of the glass, a
negative value indicates the opposite, and a value of γ0
would mean that these two were perfectly offset. Table 3shows
that there are only a few materials for which γis negative: IRG
24 (−5.9), IRG27 (−23.9), NRL4 (−41.7), and NRL8 (−25.5).
By being able to include these two additional materials into the
design suite, we have doubled the number of options available
for optical designers to achieve their goals.
5. CONCLUSION
We have demonstrated two lens systems designed to operate
over the 3–11.5 μm range (MWIR to LWIR), which have been
color corrected and athermalized, and can zoom from 50 mm
to 150 mm focal length.
With the new NRL materials being added to the list of avail-
able glasses that work over the MWIR to the LWIR, designs
that were traditionally longer and contained more elements
can be realized with fewer elements and be shorter in length.
This new SWaP design will also have better transmission and
better performance over the thermal operating range.
REFERENCES
1. W. Smith, Modern Optical Engineering (SPIE, 2008).
2. J. L. Rayces and M. Rosete-Aguilar, “Differential equation for normal
glass dispersion and evaluation of the secondary spectrum,”Appl.
Opt. 38, 2028–2039 (1999).
3. J. L. Rayces and L. Lebich, “Thermal compensation of infrared ach-
romatic objectives with three optical materials,”presented at the
International Lens Design Conference, Monterey, CA, 1990.
4. J. N. Vizgaitis, “Optical concepts for dual band infrared continuous zoom
lenses,”Proc. SPIE 7652, 76522E (2010).
5. N. A. Carlie, “Instantaneous dispersion: a window into property rela-
tionships for optical glass,”Int. J. Appl. Glass Sci. 6, 364–374 (2015).
6. B. E. Catanzaro, M. Dombrowski, J. Hendrixson, and E. Hillenbrand,
“Design of dual-band SWIR/MWIR and MWIR/LWIR imagers,”Proc.
SPIE 5406, 543875 (2004).
7. C.-W. Kuo, “Achromatic triplet and athermalized lens assembly for both
midwave and longwave infrared spectra,”Opt. Eng. 53, 021102 (2013).
8. H. Kohler and F. Strahle, “Design of athermal lens systems,”in Space
Optics, B. J. Thompson and R. R. Shannon, eds. (International
Commision for Optics, 1974) p. 116.
9. J. L. Rayces and M. Rosete-Aguilar, “Selection of glasses for achro-
matic doublets with reduced secondary spectrum. I. Tolerance condi-
tions for secondary spectrum, spherochromatism, and fifth-order
spherical aberration,”Appl. Opt. 40, 5663–5676 (2001).
10. M. Rosete-Aguilar and J. L. Rayces, “Selection of glasses for achro-
matic doublets with reduced secondary spectrum. II. Application of the
method for selecting pairs of glasses with reduced secondary spec-
trum,”Appl. Opt. 40, 5677–5692 (2001).
11. N. A. Thompson, “Optical design of common aperture, common focal
plane, multispectral optics for military applications,”Opt. Eng. 52,
061301 (2013).
12. J. N. Vizgaitis, K. Witte, R. Littleton, and P. Perconti, “Compact dual
field of view SWIR/MWIR optical system,”presented at the Infrared
Technology and Applications XXXVII, 2011.
13. R. Li, “Passively athermalized broadband optical design using doublet
combinations,”Appl. Opt. 53, 3903–3907 (2014).
14. SCHOTT, Optical Glass Catalog 2018.
15. P. Klocek, Handbook of Infrared Optical Materials (CRC Press, 1991).
16. S. Bayya, J. Sanghera, W. Kim, D. Gibson, E. Fleet, B. Shaw, M. Hunt,
and I. Aggarwal, “New multiband IR imaging optics,”Proc. SPIE 8704,
870428 (2013).
17. S. Bayya, D. Gibson, V. Nuygen, J. Sanghera, M. Kotov, G. Drake, J.
Deegan, and G. P. Lindberg, “Design and fabrication of multispectral
optics using expanded glass map,”B. F. Andresen, G. F. Fulop, C. M.
Hanson, P. R. Norton, eds., Proc. SPIE9451, 94511N (2015).
Research Article Vol. 58, No. 18 / 20 June 2019 / Applied Optics 5049