ArticlePDF Available

Systematic design of microscope objectives. Part I: System review and analysis

Authors:

Abstract and Figures

In the three connected papers, a systematic analysis and synthesis approach for microscope objectives is introduced. To subtract off the hidden assumptions in the historical development of the microscope objective lenses and extract the intrinsic lens modules, a large objective database is implemented including most of the patented systems for standardised applications. Based on the systematic analysis of the database, in Part I, a general review of the development history is given. A systematic classification method is proposed with respect to the five most significant parameters. According to the review and classification, the impacts of applications, manufacture and technology considerations are systematically analysed and summarised. Details of the lens modules will be discussed in Part II, and the synthesis approach utilising the lens modules will be introduced in Part III.
Content may be subject to copyright.
Adv. Opt. Techn. 2019; aop
Review Article
Yueqian Zhang* and Herbert Gross
Systematic design of microscope objectives.
Part I: System review and analysis
https://doi.org/10.1515/aot-2019-0002
Received January 4, 2019; accepted March 25, 2019
Abstract: In the three connected papers, a systematic anal-
ysis and synthesis approach for microscope objectives is
introduced. To subtract off the hidden assumptions in the
historical development of the microscope objective lenses
and extract the intrinsic lens modules, a large objective
database is implemented including most of the patented
systems for standardised applications. Based on the
systematic analysis of the database, in Part I, a general
review of the development history is given. A systematic
classification method is proposed with respect to the five
most significant parameters. According to the review and
classification, the impacts of applications, manufacture
and technology considerations are systematically ana-
lysed and summarised. Details of the lens modules will be
discussed in Part II, and the synthesis approach utilising
the lens modules will be introduced in Part III.
Keywords: aberration correction; microscope objective;
microscopy; optical design.
1 Introduction
The microscope objective lens is the most typical high
numerical aperture (NA) system, which provides a high-
contrast image with diffraction-limited resolution for
various microscopy applications. The development of
microscope objective has a long history over hundreds
of years. However, because conventional microscope
systems are highly standardised and objective devel-
opment is strongly application-oriented, the lenses
were designed with traditional approach following the
technology roadmap of a specific vendor. Therefore, the
systems were developed with accumulated tremendous
complexity. A systematic analysis and design approach,
which decouples the impact of high NA physical nature,
application impact and system assembly consideration,
has been rarely reported.
Although the early age of microscope objective devel-
opment has been well described in various literatures
[1–5], there are only two articles reviewing modern objec-
tives after 1970s, which were reported by Riesenberg
in 1988 [6] and Broome in 1992 [7]. From the mid-1990s,
the growing demand from fluorescence microscopy and
the semiconductor industry significantly influenced the
development of microscope objective. A clear review of
the latest arts is still missing. In addition, design princi-
ples were seldom discussed [8], and illustrative systematic
synthesis approach cannot be found for high NA cases
[9–11]. Last year, we have proposed a new systematic
approach for microscope objective analysis and synthesis
[12], which is based on 116systems from patents. However,
it only focused on the functionality of lens modules for
aberration correction, without discussing the impact of
application and manufacturing consideration.
‘Zu den Sachen selbst’ (back to the things themselves)
is a key concept of phenomenological Epoché in Husserl’s
phenomenology, where ‘Sachen’ or ‘thing’ in English refers
to any phenomenon that may confront the ego in con-
sciousness [13], such as a law of nature. In optical design,
it is also necessary to subtract the hidden assumptions to
reach the reality [14]. Concerning the growing demand of
various high NA applications and the recent trend of deep
learning-aided optical design, it becomes more important
to operate the Epoché to efficiently utilise the experience
in conventional microscope objective design. As a basic
step, we have collected most of the patented microscope
lenses for standardised applications. Systematic analysis
was conducted with respect to three concepts: aberration
correction, impact of application and considerations of
manufacture and technology. For each phenomenological
model in the objective design, each concept is discussed
with certain Epoché.
In this paper, a historical review of the system
development is given, and a systematic classification is
*Corresponding author: Yueqian Zhang, Institute of Applied
Physics, Friedrich Schiller University Jena, Albert-Einstein-Str. 15,
07745 Jena, Germany, e-mail: yueqian.zhang@uni-jena.de
Herbert Gross: Institute of Applied Physics, Friedrich Schiller
University Jena, Albert-Einstein-Str. 15, 07745 Jena, Germany
Open Access. © 2019 Yueqian Zhang et al., published by De Gruyter. This work is licensed under the Creative Commons Attribution-
NonCommercial-NoDerivatives 4.0 License.
www.degruyter.com/aot
© 2019 THOSS Media and De Gruyter
Unauthenticated
Download Date | 6/14/19 7:56 PM
2 Y. Zhang and H. Gross: Systematic design of microscope objectives. Part I
implemented to set the basis for systematic analysis. The
impacts of the most significant applications and general
manufacture and technology considerations are analysed
and summarised. Decoupling these effects, the detailed
design principles and lens modules would be introduced
in the connected paper Part II, and the system synthesis
and discussion of some special techniques are included
in Part III.
2 Patents collection and database
setup
We have collected 448 entries from patents of the United
States, Germany and Japan. Although some individual
entries could also be found from Russian, Chinese and
Korean patents, they are mostly designed for a specific
setup, not for standardised applications. Because there
is no special design principle applied, they are excluded
from the database. Designed for different setups, some
systems with identical structures were reported in mul-
tiple patents from different countries. These systems are
combined with the reference to the US patents; thus the
remaining German and Japanese patents indicate that
they are only patented in the corresponding countries.
Owing to their distinctive complexity, optical disk
objectives and recent in vivo endoscopic microscope
objectives are not considered in the patent collection.
The collected entries are patented for various research
and routine applications and mostly focus on the field of
biomedical research and semiconductor industry. When
it comes to the system structure, reflective and catadiop-
tric objectives are excluded, but objectives with diffractive
optical elements (DOE) are collected to analyse the func-
tionality of the DOE in system simplification.
From the first apochromatic objective patented by
Boegehold in 1926 [15] to the latest released high etendue
immersion objective in May 2018 [16], the trend of patents
is illustrated as Figure 1, which is sorted by the release
dates of the patents. Except the early developments in
Carl Zeiss before World War II, the major development of
microscope objectives started in the 1950s, and there are
four peak periods of publication as indicated in Figure 1.
From 1965 to 1975, most of the basic objective struc-
tures were invented for biomedical and metallurgical
applications. During this time, there were an increas-
ing number of systems designed for infinite-conjugate
instead of the conventional finite-conjugate system with
standardised tube length. Microscopes with infinity optics
then became the major type from 1980s. The second peak
period started from the mid-1980s, following the flourish-
ing demand of the semiconductor industry. The working
distance and corrected spectral range were extended
during this period, corresponding to the requirements for
semiconductor fabrication operation. From the 1980s, the
objective series for research and routine applications are
clearly separated. Although the advanced objectives for
semiconductor significantly change the routine applica-
tions, applying fluorescence microscopy to biology and
material science led to a revolution of objective design
for research application. To realise UV excitation for epi-
fluorescence, the UV transmittance of objectives was first
improved by utilising special glasses in the 1980s. Then
0
5
10
15
20
25
1926
1930
1934
1938
1942
1946
1950
1954
1958
1962
1966
1970
1974
1978
1982
1986
1990
1994
1998
2002
2006
2010
2014
2018
Fluorescence Semiconductor All patents
1st: 1965–1975
2nd: Late 1980s
3rd: Late 1990s
4th: Recent decade
Number of documents
Figure 1:Number of modern microscope objective patents from 1926 to 2018.
Unauthenticated
Download Date | 6/14/19 7:56 PM
Y. Zhang and H. Gross: Systematic design of microscope objectives. Part I3
in the 1990s, various advanced objectives with excellent
fluorescence contrast were further invented. Therefore,
the highest peak period of microscope objective publi-
cation was found around 2000, which results from the
development for both the research and routine applica-
tions. There are two major reasons resulting in the recent
peak period of development. For one thing, coming into
the 21st century, digital sensors were well developed and
widely utilised. Utilising digital imaging and postmagnifi-
cation, it is possible to obtain high-resolution image with
large visual field. Therefore, objectives with low/medium
magnification and high NA are preferred instead of the
high magnification objectives, to avoid the frequently
cumbersome changing of objectives. In addition, an
increasing number of advanced fluorescence microscopy
methods were utilised, such as nonlinear microscopy (e.g.
multiphoton microscopy), total internal reflection fluo-
rescent microscopy (TIRFM) and superresolution localisa-
tion microscopy. Owing to their essential requirement of
high contrast, high resolution and special system param-
eter (spectral range, working distance, etc.), the objective
structures were modified. Hence, combining these two
effects, a series of objectives with highest complexity was
reported in the recent decade.
To extract the lens modules or the building blocks for
microscope objective design, the collected systems should
be systematically analysed and compared. For system
analysis, we excluded the objectives only corrected for UV
or IR spectral range, which have different functionality
in chromatic aberration correction. It is also notable that
typically there are many embodiments included in one
patent. When we build up the system database, maximum
etendue objectives with each basic structure are selected.
Eventually, from the remaining 373 patents, 484systems
with different structures are built up within Zemax as a
database.
The throughput of an optical system is typically repre-
sented by an etendue G-value, which is defined by its NA
and field of view, e.g. object height yobj:
2
obj
G(
2y NA
).
4
π= (1)
The aperture size and field of view of the microscope
objectives are mostly specified by the object space NA
and intermediate image size (the intermediate image size
is denoted by SF with the unit of millimeter in the three
papers), respectively. However, because of the different
arrangement of the conjugate (infinite-conjugate vs. finite
conjugate) and the selection of tube lens focal lengths,
etendue of objectives from different vendors cannot be
compared directly. Therefore, we are using the object
height, which is calculated as the intermediate image
size divided by the system magnification, to evaluate the
system etendue.
Sorting the systems as a function of NA and object
height, the general throughput of conventional microscope
objectives could be demonstrated by two boundary con-
stant-etendue curves, G = 0.0243mm2 and G = 0.9503mm2,
which are shown in Figure 2A. More than 90% of the col-
lected objectives, with only 38 exceptions, locate within
the area formed by these two curves. Assuming identical
intermediate image size of 22mm, these two G-values rep-
resent a 50×/0.40 and a 20×/1.00 objective, respectively.
The maximum etendue value of the collected systems was
achieved by a 10×/0.90 objective with intermediate image
size of 25mm (SF25), which is used for virtual slide micro-
scopy. When it comes to the assignees, according to Figure
2B, the four major vendors of microscope objectives,
0.00.2 0.40.6 0.81.0 1.21.4 1.61.8
0
5
10
15
20
25
Object height (mm)
NA
G = 0.9503
G = 0.0243
DUV
lithographic
projector
10x/0.90 SF25 Fujimoto
USP 8350904
Max etendue G = 3.976
40%
33%
8%
4%
5%
5% 5%
Olympu
s
Nikon
Zeiss
Leica
AO
Mitutoyo
Others
AB
Figure 2:(A) The diagram of collected objective lenses as a function of numerical aperture and field size. The blue and pink curves indicate
the boundary G-value of conventional microscope objectives. Position of typical DUV lithographic projectors is plotted as a reference.
(B)Share of different assignees in the database.
Unauthenticated
Download Date | 6/14/19 7:56 PM
4 Y. Zhang and H. Gross: Systematic design of microscope objectives. Part I
Olympus, Nikon, Leica and Carl Zeiss, hold 85% of all
patents, but the Japanese companies patented much more
than the German companies. The other two main assign-
ees are American Optical Corporation (AO) and Mitutoyo.
AO was the first assignee that patented a series of clear-
three-group objectives with infinite-conjugate in 1970s,
whereas Mitutoyo is the major assignee of long working
distance objectives especially for semiconductor industry.
3 Lens evolution
To realise the systematic sorting of the systems, it is nec-
essary to review the historical development of the system
structures and their corresponding applications. In this
section, the first patented or the most characteristic system
with each milestone structure for each milestone applica-
tion is selected to illustrate the lens evolution. To demon-
strate the glass selection, in all the lens layout plots within
these three papers, the crown glasses are coloured with
light yellow, the flint glasses are coloured with orange and
the fluorites or fluorite glasses (Abbe number ν> 90) are
coloured with light green. With special consideration of
colour correction for wider spectral range, it is possible to
cement two crown glasses or two flint glasses together. In
the corresponding system layout, the cemented element
is plotted with two components having the same colour.
Figure 3 gives a representative microscope objec-
tive structure [3]. The objective lens has a relatively sharp
leading edge and a narrowed rear part with screw threads
behind the objective shoulder. During the historical devel-
opment of microscope objectives, two mechanical dimen-
sions were standardised: the parfocal length and the
thread diameter. The parfocal length is defined as the dis-
tance between the object plane and the objective shoulder,
which limits the axial dimension of the objective lens. The
thread diameter determines the maximum size of the exit
ray bundle. Initially, microscope objectives were designed
for a small etendue with only a few of elements. There-
fore, these two parameters were standardised early on to
be relatively small, e.g. 45mm parfocal length and 0.8
(20.32 mm) thread diameter. However, when the system
NA and etendue are extended, the short axial dimen-
sion is filled with lenses and the small thread size cannot
match the large exit pupil size. Because the conventional
standard significantly limits the degree of freedom for the
system design, vendors have selected various larger values
for their modern systems, e.g. 60mm parfocal length and
25mm thread diameter. Detailed summary of the parfocal
length will be given in Section 4.3.3.
3.1 From Lister to Abbe
Figure 4 illustrates the early stage of the microscope
objective evolution in the 19th century. Utilisation of an
aplanatic lens with achromatic correction, which was
first introduced by Lister in 1830with a pair of planocon-
vex cemented doublet, could be regarded as the start of
modern microscope objective development. Petzval modi-
fied the design in 1843 by optimising the planar surface to
compensate spherical aberration, coma and astigmatism.
Furthermore, because of the long separation of the two
doublets, according to Petzval’s law, the field curvature
could also be controlled under low aperture and small
field size. It is also notable that the aperture stop was
sometimes placed at the rear focal plane of the objective,
forming a telecentric object space. In fact, the modified
Petzval type is nowadays well known as the Lister type
two-group microscope objective.
However, these two simple two-group objective
types could only correct small NA (~0.25). To afford the
high NA, as the next evolutionary step, Amici introduced
an aplanatic-vertex lens (Amici lens) as the front lens in
1850. According to the functionality of an aplanatic lens,
the NA could be enlarged by a factor of approximately n2
without introducing spherical aberration, where n is the
refractive index of the lens material. Nevertheless, with
the increasing NA, the field curvature of a large field
becomes critical. To achieve a similar system throughput
as the Lister type, the field size of the Amici type objec-
tive must be reduced. Hence, the Amici type could only
be used for objectives with medium magnification and
medium NA, e.g. 40×/0.65 SF18. By the end of the 19th
century, the success of Carl Zeiss factory achieved lots of
breakthroughs in creating new lens structures. Utilising
Ob
ject
plane
Thread diameter
Leading
edge
Objective
shoulder
Telecentric pupil
Figure 3:Schematic drawing of microscope objective structure
including housing. For all the objectives within our database, if it is
feasible, the aperture stop is always fixed at the back focal plane to
realise telecentric object space.
Unauthenticated
Download Date | 6/14/19 7:56 PM
Y. Zhang and H. Gross: Systematic design of microscope objectives. Part I5
Lister
10x
/0.25
Petzval
10x/0.25
Amici
40x/0.65
Abbe
1886
70x/1.25
Figure 4:Early stage of two-group microscope objective development in the 19th century.
Schott glasses and anomalous dispersion material, Abbe
invented the first apochromatic oil immersion objective
lens (70×/1.25). Until the inventions of Abbe, the funda-
mental forms of two-group microscope objectives were
well developed without patenting.
Apart from the development of the objective struc-
ture, standard of microscope was also built up in the 19th
century. The mechanical tube length was fixed as 160mm,
and the thread size of the objective was standardised as
0.8 inches by Royal Microscopical Society (RMS) (special
tube lengths such as 180 and 210 mm are sometimes
used by some vendors). On the contrary, the intermedi-
ate image size (SF) was not standardised. Producers have
different choices between 10 and 30 mm for the small-
field or wide-field observation. Until now, each major
microscope objective manufacturer still selects several
different intermediate sizes between 18 and 26.5mm for
their various microscope setups.
3.2 Modern lens evolution before 1980s
Development of two-group objectives continued in the
first half of the 20th century. The conventional structures
were patented by various companies during this period.
An overview of the lens evolution before 1980s is given
in Figure 5. Achieving larger field of view with excellent
chromatic correction became a trend during this period,
and the field curvature must be well corrected. In 1938,
Boegehold first reported the well-known method for field
curvature correction by utilising a thick meniscus lens in
the rear group to compensate Petzval curvature [17]. After
Boegehold
1938(1940)
31x/0.65
Taira
1974
40x/0.55
Klein
1970
98x/1.30
Matsubara
1975
20x/0.40
Two to three group with field flatteing Double gauss type
Taira
1973
10x/0.25
Shoemaker
1972
100x/1.25
Embeded
front
lens
Gauss
type rear
group
“PNP” Type
Klein
1965
3x/0.06
Figure 5:Modern microscope objective development for field flattening and extended working distance. ‘PNP’ type very low magnification
telecentric parfocal objective was also invented during the first peak period.
Unauthenticated
Download Date | 6/14/19 7:56 PM
6 Y. Zhang and H. Gross: Systematic design of microscope objectives. Part I
the World War II, two-group objectives were further modi-
fied with a complicated front group that corrects spherical
and longitudinal chromatic aberration, and a remote rear
group with thick meniscus lenses for field curvature and
lateral chromatic aberration compensation, e.g. 98×/1.30
SF25 Klein 1970 [18]. The complicated front group could
be divided into a high NA collecting front group, which
typically consists of several aplanatic lenses, and an aber-
ration correcting middle group, which typically comprises
a set of cemented doublets or triplets. Development of
these highly sophisticated objectives further matured
between 1965 and 1975, which was the first peak period
of patenting. As a milestone, Shoemaker patented the
first ‘clear-three-group’ plan-apochromatic oil-immersion
objective [19], which consists of embedded front lens for
NA enlargement and a quasi-symmetric Gauss type rear
group for field correction. This system type became the
basic structure of high NA immersion microscope objec-
tive until now.
Another trend of microscope objective development
during this period was the extension of working distance
(W.D.), which was mostly driven by the demand for longer
operation distance and the inverted microscope. In the
conventional high NA objectives, utilisation of the Amici
lens results in a small working distance. To further extend
the working distance, lens modules from typical photo-
graphic objectives were adopted. Klein (4×/0.14, SF20)
[20] and Taira (10×/0.25, SF30) [21] first utilised a modi-
fied double-Gauss structure to design a low magnifica-
tion plan-achromatic microscope objective with a low NA.
When the NA increases, the classical double-Gauss type
cannot provide sufficient optical power with excellent
aberration correction. Thus aplanatic-concentric lenses
were inserted into the front group to enlarge the NA for
long working distance without introducing spherical
aberration, e.g. Taira 40 × 0.55 [22]. Furthermore, to better
control field curvature and coma, a thick meniscus lens
was also utilised as a rear group, which resembles the
structure in ‘clear-three-group’ system, e.g. Matsubara
20 × 0.4 [23].
Because the microscope setups are standardised for
a fixed tube length and interchangeable objectives with
different magnifications are mounted on the turret, the
objectives should be designed with identical parfocal
length. Concerning the very low magnification (1×–4×)
objectives, to fulfill the requirement of telecentric object
space and parfocality, a special ‘Positive-Negative-
Positive’ (‘PNP’) type was invented during the first peak
period, consisting of a positive front group as the field
lens, a negative middle group with cemented doublets
and a positive rear group with cemented doublets and
meniscus lenses [24]. Distributing the optical power into
these three groups, the field curvature could be con-
trolled and the telecentric object space could be achieved.
However, because the length between the lens front to
the objective shoulder should be controlled within the
parfocal length, it is increasingly difficult to design with
decreasing magnification. Therefore, there are very few
objectives reaching 1× or lower magnification.
3.3 Advanced lens evolution after 1970s
Until 1970s, most of the basic structures of modern micro-
scope objectives were invented. In the past 40 years,
further developments were driven by the great variety
of applications. In this section, the milestone changes
are summarised, which is briefly illustrated in Figure 6,
where objectives for biomedical research applications
and semiconductor industry are utilised as characteristic
research application and routine application for illustra-
tion, respectively. The miscellaneous applications, such
as metallurgical microscope for research application, will
be introduced in Section 5, where a systematic discussion
of the impact of each application on the system design is
given. Moreover, objectives designed with special optical
elements, such as diffractive optical element (DOE) will be
discussed in Part III.
The first change was the creation of objectives with
correction function (CORR). The CORR objectives appeared
in the middle of 1970s, which were invented to adjust the
tolerance of the cover glass (CG) thickness, which deviates
from the standardised value 0.17mm [25]. Furthermore,
microscope objectives developed for multiimmersion were
first reported by Zeiss in 1975. Around 1980, because of
the increasing demand to observe samples in cell culture
dishes, inverted microscope became popular for biomedi-
cal applications. Consequently, the objectives should be
designed for the dish thickness, e.g. 0.8–1.2mm, and with
long working distance [26]. Because the thickness of the
dish was not standardised as the cover slip, it is neces-
sary to design the objective with a wider range of CG
adjustment.
From 1980s, all the manufacturers developed their
own standardised microscope systems, which utilise the
infinity optics instead of the objectives with finite tube
length: UIS (Universal Infinity System) for Olympus,
CFI (Chrome-Free Infinity-corrected system) for Nikon,
ICS (Infinity Color Corrected System) for Zeiss and HCS
(Harmonic Component/Compound System) for Leica
(also with DELTA system in the early 1990s). Figure 7
shows the difference between the finite-conjugate and
Unauthenticated
Download Date | 6/14/19 7:56 PM
Y. Zhang and H. Gross: Systematic design of microscope objectives. Part I7
infinite-conjugate system. Concerning both the incidence
angle on the tube lens and the mechanical track of the
whole system, the optimal focal length of the tube lens
would be selected between 150 and 250 mm. The four
major vendors are using 180, 200, 164.5, and 200 mm
in their standardised systems, respectively. As the light
path leaving the objective is collimated, the beam splitter
for epi-illumination and additional equipment for con-
trast methods could be inserted into the infinity space
without changing the image scale or the intermediate
image position. In most microscope systems, objective
lens with different magnification and NA are installed
on an objective turret. When the objectives are parfocal
and the focal length of the tube lens is fixed, independ-
ent of the additional elements inserted into the infin-
ity space, the sample-intermediate image conjugate is
always fixed. Therefore, observation with different con-
trast methods under different magnifications could be
compared easily. Taking this advantage, it is also simple
to add an epi-illumination setup into the microscope
system, particularly for fluorescence microscopy with
epifluorescence excitation.
1970s
1980s
Tojyo
1977
60x/0.95
Yonekubo
1979
20x/0.40
Kimura
1985
50x/0.60
Tojyo
1981
10x/0.40
Kenno
1987
50x/0.50
1990s
2000s
2010s
Kasahara
2013
100x/1.70
Arisawa
2000
200x/0.62
Abe SF16
2018
10.6x/1.35
Fujimoto SF25
2018
10x/0.90
Arisawa
1992
50x/0.50
Kashima
1993
100x/0.85
Chuang
2006
100x/0.80
Saito
1990
100x/0.80
Saito
1996
60x/1.15
Yamaguchi
1999
60x/0.85
Shi
2006
63x/1.20(1.30)
Wartmann
2014
40x/1.30
Shi
2007
100x/1.25
Objective with correction
(CG adjust)
Inverted microscope
Widely use of infinite-corrected system
Application segmentation
Research application
esp. biomedical research
Fluorescence microscopy (UV trans)
Improved chromatic correction
and CG adjust
Improved field flattening and index
matching for confocal microscopy
Reduced autofluorescence for better
fluoresence observation
Live cell molecule
observation
Multi-adjust
Multi-photon, IRDIC (UV-IR)
Advanced digital microscopy
e.g. vitual slides
IR objectives...
Low cost
TIRF
Cata systems...
Advanced aberration correction
Extended super-long working distance
DUV objectives...
Wide spectral range (UV-IR)
superapochromatic correction
Routine application
esp. semiconductor industry
Figure 6:Advanced modern microscope objective development corresponding to application segmentation.
Unauthenticated
Download Date | 6/14/19 7:56 PM
8 Y. Zhang and H. Gross: Systematic design of microscope objectives. Part I
After adopting infinity optics, the application segmen-
tation became clear, which could be classified as research
application, especially for biomedical use, and routine
applications, especially for semiconductor inspection.
Concerning the biomedical applications, fluorescence
microscopy is the first game changer, which appeared in
the 1980s. Because UV light is widely used for excitation,
the objective transmittance for specific UV spectral range
(e.g. 340, 365nm) should be improved for epifluorescence
setup. Besides, to efficiently collect the weak fluorescence
emitted from the specimen, the objective lens should be
designed with a larger NA, which enhances the resolution
simultaneously. From the beginning of 1980s, utilising
new type of glasses with better UV transmission, vendors
patented microscope objectives specifically designed for
fluorescence microscopy [27]. Through 1980s, invention of
objectives with improved chromatic correction benefited
from new type of glasses with anomalous dispersion.
Various correction methods for CORR objectives were
also developed, resulting in a set of highly sophisticated
clear-three-group objectives. In the 1990s, to fulfill the
requirement from the widely used confocal setup, special
design principles were exploited to create complicated
rear group, thus realising nearly flattened field. Consid-
ering the confocal depth imaging, many water immersion
objectives were designed to replace the oil immersion
objectives for index matching in biological sample obser-
vation [28]. Coming into the 21st century, the biomedical
research interest was changed from simple cell structure
imaging to both the observation of cell structure and
behavior of molecules, which should be observed with live
cells. However, the live cell observation requires different
environmental conditions from the conventional speci-
men observation, such as temperature and immersion
medium; thus multiadjustment should be incorporated
into the CORR objective design [29]. Another evolution
was driven by the requirement of high contrast in fluores-
cence microscopy. To avoid the strong autofluorescence,
glasses were carefully selected, and system structure was
also modified to reduce stray light.
When it comes to the semiconductor inspection objec-
tives, after adopting the infinity optics [30], the evolution
in the 1980s mostly resulted from extension of spectral
range. For one thing, because the YAG laser is used for
laser processing of the semiconductor device repair, the
microscope objective should be apochromatic corrected
for the visible spectrum (VIS) for observation, together
with the specific wavelength of YAG laser and its harmon-
ics for laser processing, which reaches IR at 1064nm and
UV at 355nm [31]. In objectives of this type, the glass pair
should be carefully selected to realise the superapochro-
matic correction. For another, because the resolution
increases with decreasing wavelength, the same as litho-
graphic projection systems, UV/DUV light sources were
widely utilised for semiconductor inspection and opera-
tion. However, optical materials with excellent transmit-
tance in DUV spectrum (below 280nm) are limited. Hence,
the following two types of objectives were invented:
1. DUV objectives corrected for a very narrow band-
width in DUV, e.g. 193, 248nm, and a conjugated sin-
gle visible or IR wavelength for autofocus. Because
only calcium fluoride and fused silica could be used
for this wavelength, the system complexity signifi-
cantly increases. Therefore, we excluded them from
discussion.
2. UV-capable objective achromatically corrected for
visible spectrum and a conjugated UV wavelength.
Because of the limit of material, the corrected UV
wavelength is typically around i-line [32] and could
reach 266 nm by using calcium fluoride and fused
silica with certain sacrifice of visible correction [33].
The next evolution focused on further extension of the
working distance. Applying the conventional methods
discussed in Section 3.2, the relative working distance
factor k, which is defined in Equation (2):
W.D.,k
f
= (2)
as the ratio between free working distance (W.D.) and
effective focal length of the objective (f), could reach 2.0.
However, with increasing magnification, the effec-
tive focal length reduces, resulting in insufficient working
distance for operation. Additional shell lenses and special
power distribution were then exploited to further enlarge
Objective
Eyepiece
RMS tube length
160 mm
Objective Eyepiece
Tube lens focal
length
Infinity space
Tube lens
Additional equipment
Pupil
Figure 7:Finite-conjugate microscope system with standardised
tube length and infinite-conjugate microscope system with
standardised tube lens. (Correct).
Unauthenticated
Download Date | 6/14/19 7:56 PM
Y. Zhang and H. Gross: Systematic design of microscope objectives. Part I9
the working distance. For some 200× objectives, the rela-
tive working distance could exceed 13.0 [34]. After 2000,
catadioptric layouts are used for semiconductor objectives
to realise extreme aperture and field size. Furthermore,
many designs focused on advanced field and chromatic
correction, including chromatic variation of coma, to gen-
erate uniform resolution over the full field [35]. It is worth
mentioning that modern microscopy technologies such
as fluorescence microscopy and confocal microscopy are
also applied to semiconductor microscopy through its
development from 1990s. They utilised the design prin-
ciples from corresponding microscope objectives for bio-
medical application.
In the recent decade, the latest development of micro-
scope objectives for biomedical use mostly focused on
the application of advanced fluorescence microscopy. It
is notable that many design principles utilised for semi-
conductor applications were also adopted. For instance,
in multiphoton microscopy, IR light is used for excita-
tion and the harmonic fluorescence is generated at UV
and visible range. Under the epi-illumination setup,
the microscope objective must be superapochromatic
corrected from UV to IR [36], where the glass selection
method is similar to what is used in superapochromatic
semiconductor objective. The next trend of the advanced
microscope objective design is to fulfill the requirements
from the superresolution methods. TIRF microscopy is
particularly considered because of its requirement of
system NA larger than 1.38. Utilising special immersion
oil, the 100× objective could reach extreme NA of 1.70
[37]. Other objectives designed for popular localization
microscopy methods were also proposed by the vendors.
Moreover, based on the development of digital sensor,
objective with high NA but low magnification is of interest
to achieve high resolution with wide field, particularly for
virtual slide microscopy. However, the abovementioned
objectives should be corrected for far larger etendue than
the conventional systems, which could be found as the
dots above the constant-etendue curve G = 0.9503mm2 in
Figure 2A. Their structures cannot be simply divided into
clear-three-groups. Many of the utilised special structures
significantly correct the induced higher order aberrations
but suffer from critical sensitivity. Another orientation in
the recent decade is to generate various new structures
that ensure better system tolerance for low cost micro-
scope objective lens.
In recent years, new microscope objectives with
highest etendue have been patented for virtual slide
microscopy. Some systems do not only realise high NA
under low magnification, but also eliminate vignetting
[38], which is similar to the advanced semiconductor
objective. However, as a consequence, the parfocal length
and thread diameter cannot be controlled within the con-
ventional value.
Generally speaking, lots of common design princi-
ples are applied to both the current microscope objec-
tives for biomedical use and semiconductor industry.
Because these high NA systems nearly reached the bound-
ary etendue, the application-based differences are over-
whelmed by the physical similarity.
4 System classification and
important parameters
To analyse the objectives and summarise the lens
modules, it is necessary to classify the collected objec-
tives into several classes and compare the structure of
individual objective lens concerning their functional-
ity resulted from aberration correction, application and
the considerations of manufacture and technology. The
most conventional classification method is based on the
system performance, which focused on the correction of
chromatic aberration and field curvature. Correction of
these two most important aberrations could partly indi-
cate the complexity of the objective; however, because
a systematic sorting of system NA and field, namely
etendue, is missing, a general systematic classification of
all the objectives cannot be achieved. It is only possible
to analyse the impact of aberration on the lens modules
by combining the etendue classification and conventional
performance classification.
4.1 System classification based on
performance
The most traditional approach to classify microscope
objectives is to define quality classes based on their lon-
gitudinal chromatic correction and field flatness, which is
briefly demonstrated in Table 1. The conventional classes
were defined according to their correction with respect
to the depth of focus (DoF), which is defined by nλ/NA2,
where λ is the central wavelength and n is the refractive
index of the immersion medium.
Achromate: Red–blue two colours’ longitudinal
chromatic aberration corrected within 2 × DoF. Typi-
cally, the F-line and C-line, or F-line and C-line are
corrected. The secondary spectrum, e.g. C-e and F-e,
are also limited within 1.5 × DoF. (F-C is considered
instead of F-C in following discussion.)
Unauthenticated
Download Date | 6/14/19 7:56 PM
10 Y. Zhang and H. Gross: Systematic design of microscope objectives. Part I
Fluorite (semiapochromate): Red–green–blue three
colours’ longitudinal chromatic aberration corrected
within 2.5 × DoF.
Apochromate: Red–green–blue at least three col-
ours’ longitudinal chromatic aberration corrected
within DoF. Typically, the apochromate is corrected
from g-line to C-line.
Plan: Best focus position at the field edge deviates
from the axial focus position within 2.5 × DoF [39].
The higher level of correction was achieved by combin-
ing the better chromatic and field correction. Thereby,
the Plan–Apochromate class typically represents the
best performance systems. However, according to the
lens evolution, from 1980s, there were an increasing
number of objectives corrected for extended spectral
range, even reaching IR and UV spectrum. For instance,
Carl Zeiss claimed that their APOCHROMATS are fully
colour-corrected for up to 7wavelengths from UV through
to IR [40]. There is not a standardised class for these
superb performance systems. Different vendors named
this advanced feature in distinctive way. Therefore, apart
from the traditional Ach-/Fluor-/Apo- classification, we
should carefully analyse the wavelength dependence of
the objectives. Furthermore, the classical ‘Plan’ definition
is also insufficient to evaluate the field correction. Accord-
ing to Section 3.3, some latest semiconductor inspection
lens does not only correct the field curvature but also fully
corrects coma and astigmatism, achieving consistent res-
olution through the full field. The extra complexity should
also be considered.
4.1.1 Wavelength dependence and colour correction
To classify the chromatic correction associated with wave-
length dependence of microscope objectives, the corrected
spectrum should be first classified, which is summarised
in Figure 8. Beside the conventional three classes: Achro-
mate, Fluorite and Apochromate, four new classes are
introduced concerning their extension of corrected spec-
trum in VIS, UV and IR. Furthermore, the chromatic cor-
rection strategies of semiconductor inspection lenses are
sometimes different from the general applications, which
are also demonstrated in Figure 8.
Improved VIS Apochromate, which extended
the corrected spectrum over the full visible range,
Table 1:Conventional classification of microscope objectives based
on performance.
Field
correction
Colour correction improved
Improved
None AchromateFluorite Apochromate
Plan-Ach  
Plan Plan-Fluor
Plan-Apo
..... .
General
applications
Se
miconductor
inspection
Achromate
Fluorite
Apochromate
IR supe rapo ch roma te
Improved VIS apochr omate
UV superapochromate
UV-IR superapo ch roma te
YAGYAG 2ndYAG 3rd
Obse rvation range
UV/DUV na rrow band
obse rvation
Conjuga ted
VIS/IR
autofocus
Obse rvation range e.g. Ther mal
detection
Figure 8:Spectrum of various chromatic correction classes.
Unauthenticated
Download Date | 6/14/19 7:56 PM
Y. Zhang and H. Gross: Systematic design of microscope objectives. Part I11
typically corrects longitudinal chromatic aberra-
tion from g-line to A-line. In some special examples,
the corrected spectrum could be further extended to
h-line approaching NUV and s-line reaching NIR.
IR Superapochromate extends the corrected spec-
trum to t-line, which is mostly required by mul-
tiphoton microscopy and IRDIC applications. It is
notable that in the recent Raman microscopy, IR
superapochromatic correction from the VIS to SWIR
around 2000nm is also required.
UV Superapochromate maintains full-spectrum
correction in NUV and visible range. It is sometimes
well corrected from i-line to C-line, which assures the
common focus of UV excitation and visible observa-
tion. This class is also widely used for semiconductor
inspection system.
UV-IR Superapochromate is the state-of-the-art chro-
matic correction class, which could be corrected from
i-line to t-line. Only a few objectives could reach this
class, and they are used for multiphoton microscopy.
The classes correct chromatic aberration through a certain
spectral range. However, when it comes to the semicon-
ductor inspection systems, despite the visible range for
observation, correction of specific wavelengths in UV and
IR range, instead of a full spectrum correction, is consid-
ered according to applications.
YAG laser (1064nm) and its harmonics are widely
used for semiconductor repairing. To assure the
repairing, laser beam is focused onto the observation
plane, achromatism of the visible range (including
YAG second harmonic 532 nm) and 1064 nm and/or
third harmonic 355nm should be realised.
Longer wavelengths in NIR and SWIR, such as
1300, 1550, and 1970nm, are often used to test the
thermal behavior of high frequency circuit.
UV/DUV observation is mostly used in the modern
semiconductor industry instead of the traditional vis-
ible observation, because of the higher resolution.
Because of the limit choice of materials, which has
good transmittance in DUV, this type objective could
only be achromatically corrected for a narrowband of
spectrum. However, the objective should also be cor-
rected for a conjugated wavelength in visible or IR
range, which is used for autofocusing. Because the
correction of this type objective is different from the
conventional broadband system, as we mentioned in
Section 3.3, they are excluded from discussion.
Concerning the applications utilising YAG laser, because
1064 and 355 nm are not far away from the visible
spectrum, the design principle is similar to that of IR/UV
superapochromate systems. Concerning the second case,
it is impossible to perfectly correct the longitudinal chro-
matic aberration for full spectrum from the visible range
to SWIR. Nevertheless, because the depth of focus of SWIR
light is far larger than that of visible light, it is possible to
realise the focal plane within the depth of focus of both
spectra. An example 100×/0.50 objective is illustrated in
Figure 9, where the best focus plane could be found within
the depth of focus of F-line through 1800nm, which means
Apo’. The boundary wavelength 1970nm is also corrected
within 2 × DoF, which means ‘Semiapo’. The complexity of
this system type is not critical, but the design only works
for medium and low NA applications.
Conventionally, only the longitudinal chromatic aber-
ration correction was considered to evaluate the chro-
matic correction of the microscope objective. However,
when we carefully compare the performance of the
modern systems, the difference between each class is not
only indicated by the longitudinal chromatic aberration
correction, but it could also be seen from the spherochro-
matism correction. Moreover, the strategy of colour cor-
rection is also different, resulting in the different shape
of the chromatic focal shift curve. The detailed differences
could be illustrated by the plots of longitudinal aberration
and chromatic focal shift shown in Table 2. The numerical
1970 nm
C-line
F-line
–8 –4 04812
Chromatic focal shift (µm)
1550 nm
1300 nm
100x/0.50 SF24 LD Nakamura JP H11-174338
Best focus
Figure 9:Chromatic focal shift of the 100×/0.50 objective, which
is measured in the object space. The blue curve shows the paraxial
focal shift, and the red area indicates the depth of focus of each
individual wavelength.
Unauthenticated
Download Date | 6/14/19 7:56 PM
12 Y. Zhang and H. Gross: Systematic design of microscope objectives. Part I
Table 2:Representative objectives from each chromatic correction class with their chromatic focal shift and longitudinal aberration plots.
Longitudinal aberrationChromatic focal sh
i
Object
i
v
e system
(a) Achromate
60x
/0.85 Yamaguchi USP 5861996
(b) Fluorite
160x
/0.95 Klein USP 4009945
(c) Apochromate
20
x/0.95 Fujimoto USP 8350904
(d) Improv
ed VIS Apochromate (g-s)
55.9
x/1.40 O Suzuki JP 2003-015046
(e) UV
Superapochromate (351nm-C)
100x
/0.85 Kashima JP H05-196874
(f) IR Superapochromate (h-t)
40
x/0.95 Yamaguchi JP 2010-134218
(g) UV
-IR Superapochromate (i-t)
40x
/1.30 O Wartmann USP 9645380
In the chromatic focal shift plot, the red dashed lines give the corrected spectrum, whereas the green lines indicate the depth of focus (DoF).
Unauthenticated
Download Date | 6/14/19 7:56 PM
Y. Zhang and H. Gross: Systematic design of microscope objectives. Part I13
aperture of the selected systems is narrowed down to 0.8,
and the longitudinal aberration is calculated in the object
space. Thereby, identical depth of focus (DoF) could be
used as a reference for comparison. For instance, at the
central wavelength d-line, it is DoF = 0.918 μm for dry
objective and DoF = 1.392 μm for oil immersion objective.
Concerning the conventional Achromate, Fluorite
and Apochromate from class (a) to class (c), based on
the basic achromatism principle, typically the chromatic
focal shift curve only has one inflection point within
the corrected spectrum. The slope around the inflection
point, as well as saddle point, is relatively small; thus
the chromatic focal shift from g-line to C-line or F-line to
C-line could be controlled within 1 ×~ 2.5 × DoF. For some
advanced Apochromate objectives, such as the example
class (c) system, utilising special glass combinations, the
area around the saddle point could be corrected rather
flat. Consequently, the maximum chromatic focal shift is
only half of the DoF. By comparing the longitudinal aber-
ration curves, typically there is a large residual sphero-
chromatism in the objectives of Achromate and Fluorite
class. But when it comes to the Apochromate, both the
paraxial chromatic error (longitudinal chromatic aberra-
tion) and the chromatic aberration at the full aperture are
well corrected, which means that the spherochromatism
is removed.
When the corrected spectrum is extended to the UV
and/or IR range, the longitudinal aberration should
be corrected with at least two inflection points on the
chromatic focal shift curve. The class (d) improved VIS
Apochromate 55.9×/1.40 objective has three inflection
points. The chromatic focal shift is therefore very flat and
controlled within the DoF through the full spectrum. The
class (e) UV superapochromate only considers the short
wavelength side; three saddle points could be found from
i-line to C-line, which utilises glasses with specific blue
side partial dispersion. However, because the red side is
not controlled, the focus of the wavelength above C-line
is shifted significantly. When it comes to the class (f)
system, only the chromatic focal shift in the visible range
is apochromatic corrected within the DoF. The correc-
tion of C-line to t-line, although takes the advantage of
DoF enlargement of long wavelength, only realises ‘Fluo-
rite’ level correction. As the state-of-the-art, the class (g)
system selects glasses with rather equivalent partial dis-
persion in blue side and red side. Hence, the two saddle
points are found in NIR and NUV, and the general curve
is flat. Consequently, the chromatic focal shift from i-line
to t-line is controlled within the DoF. Comparing the
examples from class (d) to class (g), they could be always
classified as ‘Apochromate’ according to conventional
definition. However, their exact correction of the bound-
ary wavelength might be distinctive. The difference mostly
results from glass selection. After this careful classifica-
tion, the glass selection strategies should be further ana-
lysed and discussed in Part II.
The high-performance objectives above the Apochro-
mate class (c) mostly correct spherochromatism at the full
aperture. Some extreme cases even correct the sphero-
chromatism for all the aperture zones, such as the system
(d). This advanced feature is useful in the objective with
iris and the application utilising laser with apodization.
Under these circumstances, although the effective NA is
smaller than the designed value, nearly identical longitu-
dinal correction could be maintained.
4.1.2 Field correction
In the conventional classification based on performance,
field correction level was basically classified according to
their field curvature correction. Apart from field curvature
correction, in high-performance systems, correction of
other field aberrations, especially coma, should also be
considered. Therefore, the field correction could be clas-
sified into the following seven classes, shown in Table3.
Table 3:Seven classes of field correction level of modern microscope objectives.
Field curvature Field aberration NA Etendue
Non-plan (class ) No correction No correction Arbitrary Arbitrary
Plan
Class  Corrected Corrected Low Medium/low
Class  Corrected Not well corrected High Medium/low
Class  Corrected Corrected with vignetting High Medium/low
Class  Corrected Corrected without vignetting High Medium/low
Class  Corrected Corrected with vignetting High High
Class  Corrected Corrected without vignetting High High
Unauthenticated
Download Date | 6/14/19 7:56 PM
14 Y. Zhang and H. Gross: Systematic design of microscope objectives. Part I
Owing to the small field, coma and astigmatism in
low NA systems are usually not critical; thus they are well
corrected in the Plan objective (class 2). However, because
the primary transverse coma increases linearly with field
size and quadratically with NA, it would be tremendous
at the boundary field of high NA system. Some high NA
systems only utilise a small field with corrected field cur-
vature and sagittal coma to fulfill isoplanatism. Although
the tangential coma is not well corrected, because of the
small field, it is still acceptable (class 3). When it comes to
the high NA system with relatively large field (high mag-
nification system with medium etendue), vignetting is a
useful tool to control the field deviation by shrinking the
effective NA of the boundary field (class 4). Coma could
also be well corrected without vignetting, but special
lens modules must be used especially in the rear group,
resulting in enormous complexity (class 5). According to
Section 3.3, microscope objectives with extremely high
etendue were invented recently for virtual slides applica-
tion. Utilising similar intermediate image size (SF20-30),
the magnification is very low (10×—20×) and the NA is
extremely high (dry 0.9–0.95, oil immersion 1.40–1.45).
Consequently, with standardised infinity optics, the exit
pupil size is 5–10 times larger than that of high magnifi-
cation high NA systems. Hence, in class 6systems, lens
modules in class 5 must be utilised, and the vignetting
is inevitable. As the state-of-the-art (class 7), in the high
etendue system, field aberration could also be fully cor-
rected without vignetting. However, the dimension of the
objective is enlarged, and/or the standardised infinity
optics is abandoned. The representative systems of these
seven classes and their field performance are shown in
Table 4. All the systems are apochromatic corrected from
g-line to C-line. All the transverse aberrations are calcu-
lated for the whole system, which is a combination of the
objective and its tube lens.
According to the transverse aberration fans of the
axial field, all the systems are nearly perfectly corrected
for spherical aberration and longitudinal chromatic aber-
ration. In the class 1system (a), only the rear cemented
meniscus triplet compensates the field curvature; thus
the positive power is still too strong to generate negative
Petzval curvature. As a result, the focal shift of the edge
field is approximately 5 × DoF, two times larger than the
‘Plan’ limit. The system also suffers from large amount
of astigmatism. However, because the cemented menis-
cus triplet contributes a lot to coma compensation, with
a little vignetting, although the field curvature is large,
the resolution of the off-axial field is acceptable. The class
2system (b) utilised the double-Gauss structure. The field
curvature could be compensated by the meniscus lenses,
and coma is also well controlled by the quasi-symmetric
layout. Adopted from photographic objective design, the
double-Gauss type is useful in working distance exten-
sion and field correction, but it only works for the low
NA applications. The field curvature of class 3system (c)
is perfectly corrected for very small object diameter of
0.066mm. Coma of F-line and d-line are also controlled
well for this small field. However, the objective suffers
from chromatic variation of coma, resulting in the tremen-
dous coma of g-line and C-line. The class 4system (d) is
a typical high magnification high NA objective, with low
etendue G = 0.044 mm2. The field curvature is corrected
perfectly, and 20% vignetting was introduced to cut off
the exploding coma at the edge field. On the contrary, the
class 5system (e) is designed for identical magnification
and NA as (d) but with larger field. Comparing the rear
part of these two systems, the system (e) did not use the
popular Gauss type rear group. The thick positive lens
forms a special air lens together with the following strong
negative lens, which compensates field curvature and cor-
rects coma simultaneously. System (f) and system (g) were
invented for virtual slides application, with extremely high
etendue Gf = 3.976 and Gg = 3.715 mm2, respectively. The
system (f) is designed for standardised infinity optics and
with parfocal length of 90mm. The chief ray height at the
rear group is large, which induced coma, and the vignet-
ting of the off-axial fields is introduced by both the first
and last cemented meniscus doublet as field diaphragms.
However, according to the transverse aberration plot, even
if 20% vignetting is utilised, coma rises dramatically at
the boundary aperture. Field curvature correction is also
hampered. Although the best image shell still lies within
2.5 × DoF, because of the large astigmatism, the tangential
shift exceeds the limit. In the state-of-the-art system (g),
all the field aberrations are well controlled. Nevertheless,
the parfocal length is sacrificed. The distance from the
object to the objective shoulder is approximately 300mm,
far larger than the standardised parfocal length from 45
to 105 mm. Furthermore, the diameter of the objective
exceeds 100mm, which is also five times larger than the
typical diameter of NA = 1.45 objectives.
Transverse chromatic aberration (lateral colour)
should also be considered when we evaluate the perfor-
mance of field correction. Taking the advantage of the
infinity optics, the lateral colour is usually well corrected
in modern Plan-Apochromates. However, the vendors
have different strategies in correcting lateral colour.
Nikon and Olympus fully correct lateral colour in both
objective and tube lens. Carl Zeiss leaves 1.5% lateral
colour in the objective, which should be compensated
by the tube lens. Leica follows a similar strategy but
Unauthenticated
Download Date | 6/14/19 7:56 PM
Y. Zhang and H. Gross: Systematic design of microscope objectives. Part I15
Table 4:Representative objectives from each field correction class with their field curvature and transverse aberration fan plots.
Transverse aberrationField curvature
Objective system
(a) Class 1
100x
/1.35 O SF20 Goto USP 3912378
(b) Class 2
20x
/0.40 SF29 Tojyo USP 4232941
(c) Class
3
100x
/1.65 O SF6.6 Suzuki USP 5659425
(d) Class 4
100x
/0.95 SF20 Saito USP 6188514
*Rear
cemented doublet with two dierent
flint glasses
(e) Class 5
100x
/0.95 SF25 Matthae DE 10316415
*Cemented
quartet with two dierent c rown
glasses cemented together
(f) Class
6
10x
/0.90 SF25 Fujimoto USP 8350904
(g) Class 7
20
x/1.45 SF30 Abe USP 9746658
Field curvature is plotted for central wavelength d-line. In the plot, the S stands for sagittal focal shift and the T represents the tangential
focal shift. According to different definition, the best image shell could be found between the sagittal shell and tangential shell. Except
Class 2, where the DoF is very large under low NA, 2.5×DoF is shown as the reference to evaluate the level of field curvature correction.
Unauthenticated
Download Date | 6/14/19 7:56 PM
16 Y. Zhang and H. Gross: Systematic design of microscope objectives. Part I
sometimes with additional compensation in eyepiece,
which idea is also utilised in field curvature compensa-
tion [41]. The different strategies were mostly originated
from their correction strategies before the introduction of
infinity optics. Without the help of tube lens, it was dif-
ficult to correct lateral colour, whereas the correction of
other aberrations is realised simultaneously. It was dis-
covered very early that lateral colour and astigmatism
are best corrected by delegating correction work to the
objectives and eyepieces [42]. Before 1980s, only Nikon
corrects lateral colour in the objective. The other vendors
compensate it by the eyepiece. In the modern infinity
systems, Leica and Carl Zeiss transferred the compensa-
tion from eyepieces to tube lens, but Olympus turned to
Nikon’s philosophy.
4.2 System classification based on etendue
In addition to the above classification based on perfor-
mance, to analyse the lens modules, it is also necessary
to further classify the systems based on NA and field size,
which mostly determine the system complexity. According
to the NA vs. field plot in Figure 2A, the general behavior of
microscope objectives could be seen; however, the different
complexity of structures corresponding to different correc-
tion level cannot be distinguished. The systems could be
further classified into six zones, demonstrated in Figure 10,
which is modified from the former 5-zone classification [12].
Zone 1: typical ach/apochromatic two-group systems.
Zone 2: typical Plan-ach/apochromatic clear-three-
group systems.
Zone 3: novel three-group systems with special correc-
tion lens modules.
Zone 4: systems with extremely high NA or etendue,
which sacrifice other system parameters, such as par-
focal length and immersion liquid type.
Zone 5: very low magnification parfocal telecentric
systems.
Zone 6: very high magnification systems.
There are three solid boundaries in Figure 10: (a) magni-
fication of 4, (b) magnification of 100 and (c) NA of 1.5,
which defines Zone 5, Zone 6 and partly defined Zone 4,
respectively. Below the boundary (a), nearly all the Zone
5 systems are designed with ‘PNP’ structure, which was
introduced in Figure 5, to fulfill the requirement of parfo-
cality and telecentric object space. Concerning the Zone 6,
only with few total internal reflection microscopy (TIRF)
objectives as exceptions, most of the very high magnifi-
cation systems are designed with long working distance
for semiconductor-related or metallurgical applications.
Additional complexity has been introduced to control the
more severe chromatic aberration resulted from the reduc-
ing focal length and increasing working distance. The
design principles utilised in most of the Zone 6 systems
are similar. The boundary condition (c) gives the limit of
normal oil immersion, under which circumstance the oil
index is typically around 1.515 at d-line. To realise the
extremely high NA (state-of-the-art 100×/1.70), special oil
must be used. For instance, the oil with d-line refractive
index 1.78035 is used by Olympus (100×/1.651.70) [37, 43],
and 1.80914 is used by Nikon (100×/1.65–1.67) [44, 45].
Most of the conventional microscope objectives,
which hold 86% share in our database, are classified
into Zone 1, Zone 2, and Zone 3. On the semilog coordi-
nate in Figure 10, these zones are defined by four lines,
which could relatively represent the etendue. With respect
110100
NA
Magnification
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
1.8
0.05 24 20 40 60 250
Limit of normal oil immersion
Zone
1
Zone
2
Zone
3
Zone
4
Zone
5
Zone
6
Figure 10:6-zone classification of microscope objectives based on etendue.
Unauthenticated
Download Date | 6/14/19 7:56 PM
Y. Zhang and H. Gross: Systematic design of microscope objectives. Part I17
to 22 mm intermediate image diameter, the four lines
approximately represent G = 0.025 mm2, G = 0.051 mm2,
G = 0.086mm2 and G = 0.950mm2. In each zone, most of
the systems are designed with similar structural complex-
ity from basic two groups to sophisticated three groups. In
Zone 4, except the extremely high NA region defined by
the NA = 1.5 boundary, the other systems have extremely
large etendue, which are also the exceptional systems in
Figure 2A.
4.3 Other important parameters
Based on the classification introduced in Sections 4.1 and
4.2, during our modular analysis, we first focus on an
etendue zone and then pick up the systems with identical
or comparable colour correction class and field correction
class to investigate the structural complexity. However,
beside these three factors, which mostly determine
the objective structure, there are three other important
parameters that influence the complexity of objective:
the free working distance, CORR function and parfocal
length. Furthermore, the pupil fixation is also necessary
to be considered, which determines the telecentricity and
is important for microscope systems applying contrast
methods.
4.3.1 Working distance
The working distance (W.D.) is the most influential system
parameter. Because the correction of spherical aberration
and longitudinal chromatic aberration are more chal-
lenging with enlarged NA, typically the working distance
extension results in dropped NA or additional complexity.
For instance, it could be seen from Figure 10 that immer-
sion objectives with magnification 50× are not found and
the number of 50× low NA objectives is far larger than
the 40× or 60× (63×) low NA objectives. Some of these
Zone 1 low NA 50× objectives have similar complexity
as the Zone 2systems. The reason is that compared with
the 40× and 60×, which are standardised magnification
for both biomedical applications and industrial applica-
tions, the 50× objectives are usually only designed for
industrial inspection use. Thus, the systems are required
to have long working distance and only work in air. There-
fore, when we compare the systems within one etendue
zone, the working distance should also be considered
simultaneously.
W.D. of an optical system mostly depends on its focal
length and NA. Because the microscope objectives are
designed for standardised optical tube length, the magni-
fication could be considered instead of focal length. W.D.
reduces with increasing NA and decreasing focal length,
which means increasing magnification. The dependence
of W.D. on magnification and NA is illustrated in Figure11.
Assuming the objectives are corrected with infinity optics,
the high magnification objectives, which have short
focal length, have small exit pupil diameter. To restrain
the spherical aberration, the beam diameter through the
whole system should be small. The conventional micro-
scope objectives are designed with quasi-aplanatic front
lens with great power. Consequently, spherical aberra-
tion is well controlled, but the W.D. is very small, which
is shown by red path in Figure 11. On the contrary, in low
magnification objectives, because the exit pupil size is
comparably large, under the same NA, the free working
distance d2 is larger than d1 of high magnification. When
it comes to the low NA, according to Figure 11, the free
working distance d3 and d4 are far larger than d1 and d2
in high magnification and low magnification systems,
respectively. In the modern objectives, particularly for
semiconductor inspection application, extended W.D.
is required with high magnification. Therefore, quasi-
aplanatic shell lenses are used to generate the front group
as the low magnification case, and the system is designed
with a retrofocus layout to realise the small exit pupil size.
However, since the beam diameter in the middle group
increases, the optical power of middle and rear group
must be enlarged, typically generating larger aberration.
The residue aberration from the front group is therefore
difficult to be corrected. Consequently, trade-off between
extended W.D. and reduced NA for correction should be
made. Utilising the retrofocus structure, high magnifica-
tion objectives could extend the W.D. to two to eight times
High magn ification
Low magnification
Middle and
rear group
d3
d4
d1
d2
Figure 11:Ray paths of microscope objectives with different values
of NA and magnification. The red path shows the high magnification
case, and the blue path demonstrates the low magnification case.
The dashed cone angles indicate the low NA, whereas the solid cone
angles show the high NA. From the high NA setup to low NA setup,
fixing clear aperture, the curvature of front surface could be adapted
to match the aplanatic condition. The green element and ray path
show the retrofocus layout of objectives with extended W.D.
Unauthenticated
Download Date | 6/14/19 7:56 PM
18 Y. Zhang and H. Gross: Systematic design of microscope objectives. Part I
the focal length. The state-of-the-art system even realised
relative working distance of 13f under 200× magnifica-
tion. Retrofocus is not the unique basic structure designed
for working distance extension. As it was introduced in
Section 3.2, before 1980s, double-Gauss structure was first
used to enlarge working distance. However, it only works
for low and medium NA case and can only extend the rela-
tive working distance from 0.5f to 2f, which is less effective
than the retrofocus type.
The illustration above only considered the dry objec-
tives; however, when it comes to the high NA immersion
objectives, there is a different philosophy to control the
working distance. Concerning the requests from applica-
tions, the front lens of immersion objectives is designed
with an embedded structure. The smaller component is
made of index matching material and has a planar front
surface. According to its layout demonstrated in Figure12,
to reduce the generated spherical aberration, the cemented
surface is designed quasi-concentric to the object plane.
Since the thickness of cover glass is standardised, the W.D.
does not only depend on the NA and focal length, but it is
also influenced by the thickness of the small embedded
component, which typically suffers from critical manufac-
turability problem. Furthermore, the thicker the immer-
sion layer, the more the temperature-specific change in
the refractive index impairs the image [41]. Therefore, for
fixed NA, the W.D. of the high NA immersion objectives is
usually set around a nominal value and thus nearly inde-
pendent of magnification. Figure 13 gives a comparison
of the W.D. of immersion objectives with different NA,
with reference to the typical NA = 0.90 dry objectives. The
selected systems are all plan-apochromatic corrected from
g-line to C-line with vignetting. The immersion objectives
show a departure from the normal behavior of dry lens.
Concerning the systems with high NA = 1.40, on one
hand, examples with medium magnification have smaller
working distance than the normal case, because of the
thermal consideration. On the other hand, the high mag-
nification examples slightly extend the free working dis-
tance for better operability.
It is notable that in recent development, for live
cell observation, water dipping objectives are widely
used. These relatively high NA immersion objectives are
designed with long working distance but without the front
planar surface. The design principle of these systems is
similar to that utilised in the dry lenses, but specific
modification of the front element is involved. In all, the
W.D. could be classified into six classes shown in Table5,
which are based on the different design considerations
discussed above. The relative working distance factor k
defined in Equation (2) is used as a measure for quantita-
tive classification.
During the system complexity and modular analysis,
NA, field size (magnification), spectrum with chromatic
correction level, field correction level and working dis-
tance are the five most significant objective parameters,
which should be first noticed and classified.
4.3.2 Objective with correction function (CORR)
When designing the microscope objectives, the system
correction is based on certain assumptions on the envi-
ronmental conditions. However, in practical use, these
assumptions are often violated because of environmen-
tal change or bias use of supplies (e.g. cover glass and
immersion liquid). Therefore, a correction collar is often
Embeded rear flint lens
Observat ion sample
Em
beded front crown lens
Immersion medium
Cover glass
Figure 12:Embedded front lens in high NA immersion microscope
objectives. The front small lens is planoconvex, and the rear large
lens is meniscus with rear surface quasi-aplanatic. The cementing
surface is nearly concentric to the object plane. For oil, silicon oil
and glycerin immersion, refractive index of cover glass, immersion
medium and front small lens could be matched within deviation of
0.01. However, for water immersion, as refractive index of optical
material is typically at least 0.1 larger, curvature of the cementing
surface is adapted.
020 40 60 80 100
0.2
0.4
0.6
0.8
1.0
Magnification
NA = 1.
40
NA = 1.
30
NA = 1.20
NA = 0.
90
W.D. (mm)
Figure 13:The W.D. dependence on magnification of dry and oil
immersion objectives. The nominal W.D. of NA=1.20, NA=1.30 and
NA=1.40 objectives are 0.3, 0.24, and 0.15mm, respectively.
Unauthenticated
Download Date | 6/14/19 7:56 PM
Y. Zhang and H. Gross: Systematic design of microscope objectives. Part I19
introduced into the high-performance microscope objec-
tives to adjust the changes with correction (CORR) func-
tion. According to the introduction of CORR objectives in
Section 3.3, as application-oriented, correction and adjust-
ment should be made for four most crucial parameters:
1. Thickness of the cover glass (CG)
2. Immersion liquid type
3. Operation temperature
4. Imaging depth for Z-stack scanning
The CORR objectives were first invented in the 1970s to
correct the CG thickness. For biomedical applications, as
early as in the 1960s, the CG is standardised as 0.17mm
in most of the countries (JIS R 3702 in Japan, DIN 548884
in West Germany and ASTM Designation E211-65T in
USA) [25], there is a tolerable thickness range prescribed,
for instance, ±0.02 mm. This small deviation has little
impact on optical systems with low NA but significantly
hampers the imaging under high NA by introducing a
great amount of spherical aberration. Furthermore, in the
inverted microscope (IM) or other cytodiagnosis setups,
the cell culture vessels are used instead of the slide and
cover slip. However, the bottom thickness of the vessel
is not standardized. Therefore, concerning the observa-
tion with different vessels, the objective should work for a
large range of CG thickness, typically 0.6–1.2 and 0–2mm.
In the state-of-the-art system for cytodiagnosis, even CG
adjustment for 0–10mm is realised [46]. When it comes
to the industrial applications, particularly for the observa-
tion of liquid crystal devices, the objective should also be
adjustable for a large range of material thickness, such as
2–5mm [47]. We also consider this large thickness range
adjustment as CG correction.
The focus of microscope objectives in object space is
schematically demonstrated in Figure 14.
The longitudinal spherical aberration of the imaging
could be expressed as [48]:
22 22
00
22 22
NA NA
,
NA NA
II
II
SG
SG
nn
nn
ss
ds
d
nn
nn

−−
∆=
+−+

−−

(3)
where NA =nS sin uS=nG sin uG=nI sin uI. According to
Equation (3), the well-known index matching princi-
ple in confocal depth imaging could be obtained. When
immersion medium has identical refractive index as the
specimen nI=nS, the longitudinal spherical aberration is
independent of the terms with s0, thus, focusing in differ-
ent depth of the specimen does not induce spherical aber-
ration. Assuming the index matching is fulfilled, taking
the lower order of Taylor expansion, the longitudinal
spherical aberration of cover glass could be expressed as
22
2
3NA
.
2
GI
GI
nn
sd
nn
∆=
(4)
To assure the CG thickness deviation does not influ-
ence the image quality, the induced longitudinal spherical
aberration should be controlled within the depth of focus
nSλ/NA2, where nS is the refractive index of the specimen.
Therefore, the acceptable maximum thickness deviation,
Δd could be defined as Equation (5):
3
22 4
4.
NA
GI S
GI
nn n
dnn
λ
∆=
(5)
Concerning typical cover glass made of K5 glass,
under the central wavelength at d-line, the corresponding
acceptable thickness deviation range for four cases: dry
Table 5:Six classes of working distance of modern microscope objectives.
W.D. classes System type Relative working distance factor k
Class  Normal W.D. microscope objectives .–.
Class  Immersion high NA objectives
Class  Medium NA, long working distance double-Gauss dry objectives .–.
Class  Medium NA, long working distance retrofocus dry objectives .–. (state-of-the-art .)
Class  High NA, long working distance retrofocus dry objectives .–.
Class  Long working distance immersion objectives, especially water dipping lens .–.
d
Front lens
Cover glass n
G
Specimen nS
Immersion
medium n
I
uI
uG
uS
s0
Figure 14:Sketch of the object space focus through the cover glass
with immersion.
Unauthenticated
Download Date | 6/14/19 7:56 PM
20 Y. Zhang and H. Gross: Systematic design of microscope objectives. Part I
objectives with specimen in air, dry objectives with speci-
men in water, water immersion objective with specimen in
water and typical oil immersion objectives with specimen
in water could be calculated as
∆∆
∆∆
≤≤
≤≤
44
44
0.00630.0084
mm
,m
m,
NA NA
0.0267 0.7342
mm
,m
m,
NA NA
DA DW
WO
dd
dd
(6)
respectively. Figure 15 illustrates the tolerable CG thick-
ness of these four cases with different NA.
It is self-evident that for dry objective with NA >0.75,
0.02mm deviation of the CG thickness would hamper the
image quality. Thus, to realise high performance, they
must be designed with CORR function, typically utilis-
ing movable groups in the objective, to compensate the
effect. Concerning the applications requiring a large toler-
able CG thickness range, such as the inverted microscope,
according to Figure 15, the objective lens cannot have
NA above 0.4 (critical NA). Therefore, most of the pat-
ented IM objectives have a correction collar incorporated.
When it comes to the water immersion objective used for
biological specimen, the critical value of NA is around
1.10; therefore, the off-the-shelf water immersion objec-
tives with NA > 1.10 are designed with CG CORR function.
Because of the small index gap between the immersion
medium and cover glass, the oil immersion objectives are
sparsely affected by the CG thickness deviation. However,
regarding the total internal reflection fluorescence (TIRF)
microscopy, because the illuminating light beam should
be accurately focused onto the front surface of cover glass
to generate evanescent wave for fluorescence excitation,
the extremely high NA objectives should be capable to
correct CG thickness.
The immersion correction objective was invented
in 1975 [40]. Before 1990s, microscope objectives, which
could be utilised for water, glycerin, silicone oil, and oil
immersion, were well developed. Under high NA, the
change of the refractive index of different immersion
liquid varies the working distance and induces spheri-
cal aberration. Similar to the idea of CG correction, the
induced spherical aberration should be compensated by
movable groups. Thereby, this class of high-performance
objectives could work for multitasks, especially to match
the refractive index of specimen to realise depth imaging
with confocal microscopy.
The invention of temperature-correctable systems
is mostly oriented to the application of live cell observa-
tion, which would be further discussed in Section 5.4. The
temperature of live cells is 37°C, which is 14°C higher than
the room temperature 23°C. To compensate the induced
spherical aberration, these water immersion objectives
should be designed with correction collar. Moreover,
because the extremely high NA oil immersion objectives
are very sensitive to the thermal expansion and refractive
index deviation of the oil, they are not only designed with
a relatively larger working distance but also capable to
compensate the thermal-induced aberration. Therefore,
temperature correction collar could be found in most of
the oil objectives with NA larger than 1.40.
It is well known that for depth imaging, refractive
index of the immersion liquid and the specimen should
be matched to avoid tremendous induced spherical
0.00.1 0.20.3 0.40.5 0.60.7 0.80.9 1.01.1 1.21.3 1.41.5
1E–3
0.01
0.1
1
10
100
Tolerable thickness d (mm)
Numerical aperture
Dry in water
Water immersio
n
Oil immersion
0.17 ± 0.02 mm
IM 0–2 mm
IM 0.8–1.6 mm
0–10 mm
Dry in air
TIRF
Figure 15:The tolerable CG thickness of dry, water immersion and oil immersion objectives. Dry objective could reach maximum NA of 0.95.
Water immersion objectives are typically used for NA from 0.8 to 1.2. The oil immersion objectives usually work for NA from 1.1 to 1.5. Typical
values of the CG tolerance of conventional biomedical microscopes and cytodiagnosis setups are plotted.
Unauthenticated
Download Date | 6/14/19 7:56 PM
Y. Zhang and H. Gross: Systematic design of microscope objectives. Part I21
aberration [3]. Therefore, to observe biological cells, water
immersion objectives should be used. However, biologi-
cal specimen sometimes has refractive indices different
from immersion water, resulting in induced aberration.
Furthermore, concerning the method to operate Z-stack
scanning, the conventional method by moving tube optics
or the whole objective, utilising adaptive optics, Alvarez
plate [49, 50], and DOE suffer from problems about opera-
tion speed, limit of application and system robustness
[51]. Therefore, utilising a simple moving group in the
objective, mostly a single element, for Z-stack scanning
became an effective approach to compensate the aberra-
tion and realise automatic confocal 3D imaging.
As it was introduced in Section 3.3, the modern
high-performance microscope objectives could have
multicorrection function for the above parameters. The
state-of-the-art system could adjust CG thickness, immer-
sion liquid, and operation temperature simultaneously
by moving five individual lens groups with complicated
mechanical structure [52]. Considering all the objectives
with CORR function from the simple CG adjustment to the
advanced multiadjustment, there are three solution types
to realise the correction function:
1. Applying removable element
2. Moving components in the middle group, with vari-
ous power distributions
3. Utilising air lens effect
Except the type 1, the other CORR objectives utilise correc-
tion collar and movable lens groups in the objective. The
type 1systems inserted a removable compensator at the
front part of the objective, which does not hamper other
properties of the objective. However, it could only work
for medium NA case with small range discontinuous CG
adjustment; therefore, its application is limited. Most of
the CORR objectives utilised the type 2solution. The idea
was adopted from photographic objectives. However, the
original approach only works for medium NA system and
may vary the focal length of the objective. From 1980s,
various setups with different power distributions were
developed, and the drawbacks were overcome. However,
although it is possible to correct the parameters for a large
range, typically this type system could only adjust one
or two parameters. Furthermore, because a compound
moving group is utilised, automatic adjustment is not
easy to realise. Therefore, the advanced CORR objectives
patented in the recent decades mostly utilised the air lens
effect, which is a structure with great sensitivity to the
spherical aberration. The manufacturability of this sensi-
tive system is assured by sophisticated mechanical design.
The details of these three solutions will be discussed in
Part III. Generally speaking, except the type 1, which
used additional elements, the type 2 and type 3systems
focused on the arrangement of optical power and special
structures, which does not significantly change the com-
plexity of the advanced microscope objective systems.
4.3.3 Parfocal length and tube lens arrangement
The last most significant system parameter that influ-
ences the objective structure is the parfocal length, which
is defined as the distance between the object and the
objective shoulder. By designing microscope objectives
with identical parfocal length, the focus position could be
fixed when changing the objectives with different magnifi-
cations on a turret by turning the nosepiece. The parfocal
system was invented in 1911, originated from a suggestion
of August Köhler, but was not standardised in the first
half of the 20th century. Later, 45mm parfocal length was
widely used by most of the vendors as an internationally
recognised convention until 1990s. However, after each
vendor developed their standardised infinity optics, they
have chosen their own parfocal length for their major
series production. Although the parfocal length basi-
cally fixed the mechanical dimension of the objective,
arrangement of the tube lens with infinity optics deter-
mined the focal length of the objective. According to the
discussion in Section 3.3, each manufacturer also chose
their own tube lens focal length when standardise their
infinity optics. Because of these differences and based on
different technology roadmaps, objectives from different
manufacturers have distinctive structure under the same
magnification and NA. The parfocal length and the focal
length of tube lens of the main production of each vendor
are summarised in Table 6.
The 45mm parfocal length is still used by the German
vendors and Olympus, but Nikon holds a longer parfocal
length of 60mm. Mitutoyo, who mostly produce industrial
microscope objectives, utilises a longer parfocal length of
Table 6:Parfocal length standard and tube lens focal length of
major manufactures.
Vendor Parfocal length Tube lens focal length
Carl Zeiss mm .mm
Leica mm mm
Nikon mm mm
Olympus mm mm
Mitutoyo mm mm
AO /mm mm
Unauthenticated
Download Date | 6/14/19 7:56 PM
22 Y. Zhang and H. Gross: Systematic design of microscope objectives. Part I
95mm. Based on the relaxed mechanical dimension, it is
more realistic to design the retrofocus type long working
distance objectives. The objectives from the American
Optical Corporation (AO) used their own 34 mm parfo-
cal length before 1985 and changed to 45mm afterwards.
45 mm parfocal length was used for finite microscope
with 160 mm mechanical tube length (RMS standard),
which equivalent tube lens focal length is approximately
150mm. When the vendors applied infinity optics, to keep
the flexibility of system design, the parfocal length should
be scaled with the new focal length of tube lens. However,
because of the consideration of convention and ergonom-
ics, only Nikon made the change.
The parfocal length basically determines the amount
of space for integrating lens elements. With a longer
choice, it is possible to design the objective with relaxed
arrangement of the components. However, as a drawback,
increasing the parfocal length would lead to an adverse
effect on ergonomics that the viewing port of upright
microscopes and the specimen stage position of inverted
microscopes are raised in proportion to the increase of
parfocal distance [41].
The arrangement of tube lens of different vendors
depends on their strategy for lateral colour correction,
which was discussed in Section 4.1.2, and its focal length
is selected within the optimal range from 150 to 250mm.
With longer focal length, the chief ray angle of the off-axial
field could be reduced, but the mechanical dimension of
the whole microscope is enlarged. The selection of tube
lens focal length determines the shortest objective focal
length of Carl Zeiss and longest objective focal length of
Nikon under the same magnification.
Taking the advantage of the relaxed mechanical
dimension and longer focal length, Nikon objectives
are usually designed with a less sensitive structure.
Furthermore, it is also feasible to design 1× objective and
even 0.5× macroobjectives using a special two-part struc-
ture [53] for the standardised microscope system. Compar-
ison of two 100×/1.45 oil immersion objectives is shown in
Figure 16. The 45 mm parfocal system is from Carl Zeiss,
consisting of 14 elements and semiapochromatic cor-
rected from g-line to C-line. The 60mm parfocal objective
is from Nikon, consisting of 17 elements. It is slightly better
corrected, with apochromatic performance from g-line to
C-line, mainly because of the utilisation of more anoma-
lous dispersive materials. Both these two systems are well
corrected for large field, but the 45 mm parfocal system
is at class 5without using vignetting, whereas the 60mm
objective, although corrected for a larger field, applied
vignetting to control coma. Comparing the front and
rear group of these two systems, elements in the 45 mm
parfocal system have stronger curvatures. This is due to
the consideration of aberration correction, particularly
for more effective NA enlargement with reduced spheri-
cal and chromatic aberration and balance of coma. Fur-
thermore, the axial track could be shortened to match the
parfocal length. However, these strong curved elements
always result in greater sensitivity. Therefore, although
there are more elements used in the 60 mm parfocal
system, because of the better tolerance, the manufactur-
ability is better, and the cost might be reduced.
It is worth mentioning to realise extended etendue,
particularly up to Zone 4, and extended W.D., all the
vendors also extended the parfocal length. 75mm par-
focal length is often used by Nikon and Olympus. Some
105mm parfocal systems are patented by Carl Zeiss. Con-
cerning the extremely high etendue objectives for virtual
slides application, the parfocal length is far beyond the
conventional limit, reaching 300mm.
As a conclusion, systems with smaller parfocal length
typically have a smaller number of elements but more
critical sensitivity. Utilising longer parfocal length, within
the larger space, although more elements are used, better
tolerance, and reduced cost could be achieved. Some
special system properties could be also realised, such as
the extreme etendue and very low magnification.
4.3.4 Pupil fixation
In most of the collected patents, information about aper-
ture stop is not given. However, in the practical microscope
45 mm
60 mm
100x/1.45 O SF22 mandai USP 7046451
100x/1.45 O SF20 matthae USP 6504653
Figure 16:Layouts of the 45mm parfocal and 60mm parfocal
100×/1.45 oil immersion microscope objectives.
Unauthenticated
Download Date | 6/14/19 7:56 PM
Y. Zhang and H. Gross: Systematic design of microscope objectives. Part I23
objectives, selecting appropriate stop position is crucially
significant. It determines the telecentricity and fixes the
pupil position, which should be manipulated for con-
trast methods. A schematic drawing of practical system
is shown in Figure 17. A rear physical stop is often placed
at the exit of the objective to filter the stray light, but it
is deviated from the aperture stop position and exit pupil
position, which usually locate inside the objective. There-
fore, the exit pupil is not accessible for accurate manipu-
lation. It is feasible to shift the pupil position outside the
system behind the glass, but two to three lenses should
be added. In our objective database, most of the collected
systems have exit pupil inside the system.
We only used the criteria of telecentricity to fix the
pupil position for our analysis. When the telecentric
object space is realised, because the chief ray is parallel to
the optical axis, the finite field is not impacted at the cover
glass. Furthermore, it is advantageous to realise volume
imaging because the chief ray height is identical at differ-
ent depth, thus maintaining invariant magnification.
4.4 Examples for system complexity
comparison
After systematically summarising the classification based
on the most significant parameters, etendue, correction
performance, working distance, CORR function and parfo-
cal length, practical examples are given in this section to
illustrate the comparison of system complexity by fixing
some of these parameters.
Example I
We selected the 54 oil immersion microscope objectives
with class 4 and class 5 field correction from etendue
Zone 2 and Zone 3 and compared the number of elements
regarding colour correction and system etendue.
According to Figure 18, we could draw the following
conclusions:
1. Plan-Achromate, typically with low cost, requires
three to four less elements for correction compared
with Plan-Apochromate.
2. For medium etendue systems, difference of element
number between Plan-Semiapochromate class and
Plan-Apochromate class cannot be clearly distin-
guished, because the major difference is the material
selection but not additional element utilisation. The
principle is also applied to the high etendue systems
that objectives apochromatic corrected for extended
spectrum select special anomalous dispersive mate-
rial and use similar number of elements as g-line to
C-line corrected Plan-Apochromate.
3. The parfocal length significantly influences the num-
ber of elements. All the 45mm parfocal systems utilise
Exit
pupilRear
stop
Aperture
stop
Chie
f
ray
Figure 17:Schematic drawing of microscope objective with pupil
fixation.
0.01563 0.03125 0.0625 0.125 0.25 0.
51
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
Etendue (mm
2
)
Plan
Plan
Plan
Plan
Plan-ach
Plan-apo
Plan-semiapo
Plan-superapo
60 mm
parfocal
45 mm
parfocal
40x/1.40 SF30
150 mm parfocal
40x/1.40 SF20
45 mm parfocal
Number of elements
Figure 18:Number of elements vs. etendue of selected oil immersion objectives.
Unauthenticated
Download Date | 6/14/19 7:56 PM
24 Y. Zhang and H. Gross: Systematic design of microscope objectives. Part I
13–15 elements, whereas the Nikon 60 mm parfocal
systems utilise 15–18 elements. The 40×/1.40 SF20sys-
tem [54] realised the maximum etendue of standard-
ised parfocal length system of G = 0.385 mm2. The
40×/1.40 SF30system [55] uses conventional structure
but extended the parfocal distance to 150mm by using
25 elements, and thus it nearly reaches the boundary
etendue to conventional system as G = 0.866mm2.
Example II
We selected all the systems of Plan-Apochromate class
corrected from g-line to c-line and analysed the number of
elements regarding the different levels of W.D. extension.
We also focused on the Zone 1, 2, 3 for specific comparison.
Utilising the relative working distance, as shown in
Figure19, the systems could be clearly classified into the
five levels. The first level 0–0.5f systems correspond to the
class 1 and class 2 of normal working distance objectives
and immersion high NA objectives. The left side of the
second level 0.5–1f systems and third level 1–2f systems rep-
resent the class 3medium NA systems with double-Gauss
structure [56]. It could be seen that with one more lens,
the W.D. extension could upgrade one level. However, to
further upgrade to 2–4f level, the system structure has to
be changed to retrofocus [57], and more elements should
be used. When it comes to the class 4systems, all the three
100× objectives are designed with retrofocus structure,
which could extend the W.D. to 2–8f. One more element is
required to upgrade one level as well.
With the systematic classification of microscope objec-
tives, the design complexity could be better understood. By
selecting appropriate systems for comparing, the detailed
functionality of each lens modules in aberration correction
could be described, which is demonstrated in Part II.
5 Impact of applications
According to the lens evolution introduced in Section 3,
before 1980s, design of microscope objectives mostly
focused on the development of basic structure for NA
enlargement with chromatic correction, field correction,
and working distance extension. The strong application-
oriented phenomenon appeared after the application seg-
mentation and the introduction of standardised infinity
optics. The historical evolution of the advanced applica-
tions has been demonstrated in Figure 6. Each of the new
application brought with specific requirement of etendue
(NA and magnification), spectrum with colour correction,
field correction, working distance and CORR function,
which correspond to our systematic classification intro-
duced in Section 4. With the help of our systematic under-
standing of the system complexity, in this section, we could
systematically summarise the impact of applications.
5.1 Conventional inverted setup vs.
conventional upright setup
An inverted microscope (IM) is a microscope with its light
source and condenser on the top, above the stage, while
050 100 150 200 250
Magnification
0
5
10
15
20
25
30
35
40 0–0.5f W. D.
0. 5f–1f W.D.
1f–2f W.D.
2f–4f W.D.
4f–8f W.D.
>8f W.D.
9 Elements
20 x/0.4 SF25 W.D.=1 5.5mm 10 Elements
20 x/0.4 SF25 W.D. =25.7mm 11 Elem ents
100x/0.8 SF25 W.D. = 3.64 mm 13 Elements
100x/0.8 SF25 W.D. = 5.5 mm 14 Elements
100x/0.7 SF30 W.D. = 8.1 mm 15 Elements
20x/0.4 SF25 W.D. = 4.46 mm
9 Elements
20x/0.4 SF25 W.D. = 15.5 mm 10 Elements
20x/0.4 SF25 W.D. = 25.7 mm 11 Elements
Zone 1, 2, 3
Working distance (mm)
Figure 19:Number of elements vs. magnification of Plan-Apochromate with different working distance (W.D.) extension level.
Unauthenticated
Download Date | 6/14/19 7:56 PM
Y. Zhang and H. Gross: Systematic design of microscope objectives. Part I25
the objectives are below the stage pointing up. The IM has
been invented for a long time, first introduced in 1850 by
Smith [58]. It has been applied to metallurgical applica-
tion in the middle of the 20th century, but concerning
biomedical applications, the popular application with
IM started from 1980s because of the increasing demand
of cell culture observation. A comparison of the conven-
tional inverted setup and the upright setup is shown in
Figure 20. The conventional upright microscope objec-
tive images the specimen between the cover slip and slide
glass. Thickness of the cover slip was standardised, which
is typically 0.17 mm for biomedical applications. But the
thickness of slide glass was not standardised and usually
varies between 1.2 and 1.7mm for different applications.
The cover glass thickness typically has a tolerance of
±0.02mm. Therefore, as it was introduced in Section 4.3.2,
for high NA systems, the high-performance objectives must
be able to compensate the induced spherical aberration.
When it comes to the conventional inverted microscope
objective, it images the cell floating in the culture medium
through the bottom of the cell culture dish. The cell does
not perfectly locate at the bottom surface, leaving a vari-
able distance, and the bottom thickness of the dish was
also not standardised, varying between 0.5 and 1.5mm.
Furthermore, the inverted objectives often work with slide
glass or without substrate. Consequently, the conven-
tional inverted microscope objective with NA > 0.4must be
designed with correction ring for 0–2mm CG correction.
As the correction range is large, the objective must be
flexible for a large scale of working distance. Thus, the
conventional inverted objectives were designed with rela-
tively longer working distance, typically 0.5–1 f. The longer
working distance always bring with more challenging cor-
rection of chromatic aberration, spherical aberration and
coma. Because of this effect and the large-range CORR,
the correctable maximum etendue is limited around
G = 0.1 mm2. Figure 21 gives the off-the-shelf objectives
with long working distance (LD) and long CORR range
(>0.6mm) from the catalogs of the major vendors Nikon,
Olympus and Carl Zeiss. 20×/0.40, 40×/0.60, 60×/0.70
are the typical magnification and NA combinations. The
state-of-the-art 40×/0.78 objective utilised a special sensi-
tive element in the middle group and realised 0–1.6mm
CG adjusted under the high NA [59].
It is worth mentioning, in the current microscopy
applications, the LD and large CORR range objectives
are not only used for IM. Biological routine applications,
particularly for cytodiagnosis, have similar requirements
and this kind of LD objectives are also used in upright
setups.
Recently, for confocal fluorescence microscopy,
culture dishes with thickness of 0.14–0.20 mm and low
autofluorescence have been widely used. Therefore,
typical high NA systems with narrow range CORR could
also be used for current advanced inverted microscopes.
5.2 Confocal setup
The confocal microscopy originated from the idea of M.
Minsky in 1957 [60] and confocal imaging was developed
Cell culture dish
Culture medium
Cell
Slid
e
Cover slip
Inverted
microscopic
objective
Upright
microscopic
objective
0.5–1.5 mm
Variable
0.17 ± 0.02 mm
1.2–1.7 mm
AB
Figure 20:Comparison of (A) conventional inverted microscope observation and (B) upright microscope observation.
20 40 60 80 100
0.4
0.5
0.6
0.7
0.8
0.9
Nikon
Olympus
Carl zeiss
Collected
NA
Magnification
Figure 21:Off-the-shelf objectives with long working distance and
large range CORR from the major manufactures and our collected
database.
Unauthenticated
Download Date | 6/14/19 7:56 PM
26 Y. Zhang and H. Gross: Systematic design of microscope objectives. Part I
during the 1960s. But after the introduction of laser illu-
mination in the 1970s [61] and the invention of confocal
laser-scanning microscope (CLSM) in the middle of 1980s
[62], the confocal setup was eventually incorporated into
the standardised microscope systems of major vendors
and widely used for biomedical applications. The princi-
ple of confocal microscopy has been well discussed in lit-
eratures [63, 64]. Utilising the pinhole filtering, the major
advantages of confocal microscopy are the improved axial
and radial resolution and the feasibility of 3D volume
imaging with optical sectioning.
To realise the optical section, according to the discus-
sion in Section 4.3.2, there are three requests for CORR
function. First, to avoid the spherical aberration induced
by different Z-stacks, the refractive index of the immer-
sion liquid and the specimen should be matched. For
biomedical applications, typically, water immersion is
utilised with its d-line refractive index varying between
1.33 and 1.38 with certain solution. However, the con-
ventional material cover slips, such as K5 glass, typically
have a refractive index gap at least 0.1 compared with the
water. Therefore, according to Figure 15, water immersion
objectives with NA larger than 1.10must be designed with
CG correction ring. One special solution to overcome this
problem is to utilise special cover slip made of fluorocar-
bon resin, e.g. cytop [65]. Because the difference between
the refractive index of this material and water is only 0.02
or less, a larger deviation of CG thickness is acceptable,
which is similar to the conventional oil immersion case.
Moreover, the material has far lower water absorption
degree than other materials with similar index, such as
PMMA, and good UV transmittance with low autofluores-
cence. Therefore, it is a good choice for precise confocal
fluorescence microscopy. Second, the high NA objectives
used in confocal setup sometimes do not only work with
water immersion; immersion liquids such as silicone oil,
glycerin and oil are also applied to match the index of
other specimens or realise better resolution. Therefore, an
advanced class of objectives was invented with immersion
correction. Third, to realise fast and robust scanning, the
recently developed objectives are designed with a light
weight simple moving group for depth correction.
When it comes to the impact of confocal setup on the
basic properties of microscope objective, to realise more
efficient 3D reconstruction of the object with high reso-
lution, particularly with the help of postmagnification,
objectives with high NA and large observation field are
preferred. Furthermore, because laser scanning is used
to obtain the image, the requirement of field flatness is
higher than conventional systems. The field aberrations
should also be controlled, and the relative illumination
of the image edge should be improved with less vignet-
ting. Therefore, the advanced microscope objectives used
in confocal setup are designed with medium etendue and
excellent field correction of class 5.
5.3 General consideration of fluorescence
microscopy
The idea of fluorescence microscopy originated from
August Köhler in 1904, but it was not widely used until
1970s. From the beginning of 1980s, some microscope
objectives are specifically designed for fluorescence
microscopy, particularly for i-line UV excitation. The
importance of fluorescence microscopy significantly
increases from 1990s, especially for biomedical applica-
tions. The demand of live cell observation has grown, in
which the observation target turned from the structure
of the cells towards the behavior of molecules. The avail-
ability of green fluorescent protein (GFP) and its deriva-
tives assured the possibility of observation and redefined
fluorescence microscopy. Nowadays, the development
of advanced fluorescence microscopy techniques is still
flourishing.
Most of the current fluorescence microscopes are
implemented with reflected light fluorescence setup (epi-
fluorescence), which utilises the epi-illumination setup
consisting of a light source, an excitation mirror to filter
the operating wavelength, a dichroic mirror letting the
longer wavelength fluorescence pass and an emission
filter for further filtering. The epifluorescence setup has
great advantage in analysing opaque material with higher
fluorescence intensity, and it is possible to be combined
with other transmitted light methods.
Generally, the epifluorescence microscopy has the fol-
lowing six major requirements for the objective:
1. High NA in the object space
2. Relatively large NA in the image space
3. High transmittance over wide spectrum
4. Favorable chromatic correction over wide spectrum
5. Favorable field correction
6. Low autofluorescence
To intensify the fluorescence, the intensity of the excita-
tion illumination light should be enlarged. Therefore,
under the epi-illumination setup, the corresponding
object space NA is required to be high. The resolution is
improved simultaneously for more detailed observation.
Furthermore, because the fluorescence emitted from the
specimen is typically weak, to increase the brightness of
image for more efficient observation, a larger image space
Unauthenticated
Download Date | 6/14/19 7:56 PM
Y. Zhang and H. Gross: Systematic design of microscope objectives. Part I27
NA is also required. Concerning the first two requirements
simultaneously, the optimal magnification should be
reduced.
When it comes to the system performance, first the
specific excitation and emission wavelengths for fluo-
rescence significantly change the requirement of system
correction regarding spectrum. For one thing, the working
wavelengths of popular fluorescent dyes and fluorescent
proteins cover a wide range. For instance, Fura-2 for Ca2 +
detection is exited at 340 and 380nm in UV, Kaede and
PAGFP are exited around 400nm, and CFP, GFP, YFP have
their peak excitation wavelengths at 430, 488 and 514nm,
respectively. Because of the high fluorescent efficiency,
the UV excitation is of great interest for all the applica-
tions. Therefore, to assure the high intensity of illumi-
nation light, which passed through the objective, high
transmittance over the wide spectrum from UV to VIS is
required. Because of the limit of material, the UV trans-
mission of conventional microscope objectives is usually
low. In the 1980s, the specific objectives designed for
fluorescence microscopy could realise over 60% transmit-
tance at i-line (365nm). The advanced fluorescent micro-
scope objectives could realise high transmittance of 80%
from i-line in UV to s-line in NIR and more than 60% trans-
mittance at 340nm. Some special systems even realised
good transmission for 300nm excitation. Beside the illu-
mination light path, when we consider the imaging path,
for another, because the peak emission wavelength of the
fluorescent dyes and proteins also covers a wide range
from UV to VIS, to realise fluorescence observation for
various applications, particularly for multicolour imaging
with multicoloured fluorescent tag, the chromatic aberra-
tion should be superapochromatic corrected for this wide
spectrum, corresponding to the class (d) and (g) intro-
duced in Section 4.1.1.
To improve the operability of the microscope and
observe large field at the same time, improved field cor-
rection is also required, typically as class 4 introduced in
Section 4.1.2. The low autofluorescence is the last require-
ment in the current advanced fluorescence microscopy to
further improve contrast. In the epi-illumination setup,
because the illumination light coming from the rear
part of the objective is intensified, if the intrinsic auto-
fluorescence of the optical elements is high, stray light is
induced and thus degrades image quality and hampers
the image contrast. Therefore, in the advanced objectives,
with respect to both the requirement 3 and requirement
6, optical materials with high UV transmission and low
autofluorescence are selected. However, because of the
limit of material choice, excellent chromatic correction
becomes difficult. To overcome the contradiction, special
glass pairs have been discovered and utilised in special
lens modules, which will be discussed in Part II.
5.4 Live cell and specimen observation
In the old microscopic observation for biological applica-
tions before 1950s, the cells must be stained. The invention
of phase contrast microscopy makes it feasible to observe
the structure of live cells. However, when the behavior
of molecules in the live cells became the major interest,
despite the general considerations of fluorescence micro-
scopy, special requirements are involved in the design of
live cell and specimen observation apparatus. Moreover, to
obtain the 3D image of the live tissues, it is difficult to use
conventional confocal setup for Z-stack scanning. Taking
the advantage of the operating spectrum as well, therefore,
utilisation of multiphoton excitation is becoming popular.
This nonlinear microscopy approach also imposed several
new requirements on the objective design.
5.4.1 General consideration
When the fluorescence observation is performed on
live cells, the fluorescent substance is used to stain the
living specimen. Corresponding to the general require-
ment of fluorescence microscopy, the objectives should
be designed with large object space NA, which is also
required in live cell observation to detect as much infor-
mation from the cell at a time. Furthermore, it is also
desired to keep a state of live cell behavior in the sight for
a longer time observation; thus a wider object field size is
required. Consequently, the etendue of the live cell obser-
vation objectives should be large.
A specific requirement of live cell and specimen fluo-
rescence observation is to reduce the cell toxicity. When
some stimulus, such as the excitation light, is given to the
live specimen, there is the possibility that the stimulus
itself adversely affects an active state of the cell. Although
UV light has high fluorescent efficiency, it results in critical
phototoxicity. Therefore, light in IR range is usually used
in live cell observation with new approaches for various
applications, including multiphoton, CARS, SHG, IRDIC
and optical tweezers. Conventional fluorescence obser-
vation methods could also be operated within IR range,
using fluorescence substance with long wavelength exci-
tation. There are three major advantages of utilising IR
light in live cell observations:
1. Beneficial from the longer wavelength, the scatter-
ing and absorption of IR light in the live specimen is
Unauthenticated
Download Date | 6/14/19 7:56 PM
28 Y. Zhang and H. Gross: Systematic design of microscope objectives. Part I
reduced. Consequently, a longer penetration depth
could be realised for depth imaging.
2. Less phototoxicity.
3. It is comparably easier to achieve good transmittance
in IR range compared with UV range.
A series of microscope objectives have been specifically
corrected only for IR, typically from 700 to 1300 nm.
Because of the different glass selection considerations,
they are excluded from our systematic analysis. Some
other objectives, particularly for IRDIC and multiphoton
microscopy, are corrected for VIS-IR or UV-IR, correspond-
ing to the class (f) and class (g) chromatic correction level
introduced in Section 4.1.1. Further considerations of
multiphoton microscope objectives will be introduced in
Section 5.4.2.
As discussed in Section 4.3.2, 5.1, and 5.2, as the
refractive index of the cell culture medium or live tissue
is close to water, to avoid the tremendously induced
spherical aberration, a water immersion objective is pre-
ferred for live cell and specimen observation. Different
from conventional structure, a kind of water dipping
objective was specifically designed for live cell obser-
vation to assure the manipulation of live cells for elec-
trophysiology. Figure22 gives the structure of the water
dipping objective in comparison with a conventional
water immersion objective.
The water dipping objective is designed with the
leading end with a large access angle and long working
distance, which left sufficient free space to use the patch
clamp. The access angle should be at least 35° and typi-
cally 45° in most of the off-the-shelf objectives, and
the working distance is usually extended to 0.5–1f, cor-
responding to the class 6, introduced in Section 4.3.1.
However, to achieve the large access angle and long
working distance, the correctable NA must be sacrificed.
The conventional objective in Figure 22 could realise
NA = 1.15with magnification of 40×, but the maximum NA
of the water dipping lens could only reach 0.80. Another
major structural change could be found at the front
surface. In the conventional water immersion objectives,
the front surface is usually designed as a plane to avoid
generating bubbles and considering easy cleaning. The
front embedded lens is made of material with its refrac-
tive index matched to the immersion medium to reduce
spherical aberration and field curvature and avoid total
internal reflection under epi-illumination. When it comes
to the water dipping lens, as the objective is dipped into
the culture medium and the size of front surface is typi-
cally enlarged, if the front surface is not strongly curved,
the bubble and cleaning problem are no longer critical.
Therefore, the front surface could be designed with small
curvature to enlarge working distance.
Regarding the environmental condition of the live cell
observation, particularly for the water dipping lens, it is
necessary to design the objective to be capable to work
under both the room temperature 23°C and body tempera-
ture 37°C. Typically, the high-performance high etendue
systems are designed with correction ring for temperature
adjust, such as 20×/1.00 [66].
It is notable that the special requirement of the water
dipping lens results in the special demand of mechanical
design. With the slender mount, it should be designed
with inert and scratch-resistant surface of low surface
conductivity and thermal conductivity [41].
5.4.2 Multiphoton microscopy
Multiphoton microscopy is a nonlinear approach. During
multiphoton excitation, a fluorescent object is illuminated
with a light beam with wavelengths of integral multiples of
an inherent absorption wavelength. Thereby, the resulting
excitation is nearly equivalent to that caused by the light
with wavelength of the inherent absorption wavelength.
Based on this nonlinear phenomenon, the IR light could
be used for excitation concerning deeper penetration and
less phototoxicity, and fluorescence emission is generated
as typical UV or VIS excitation with higher efficiency. For
instance, serotonin is excited at the wavelength of 260nm
and produces emission at 300–380 nm. It is complicated
to design such a UV-capable system. However, utilising
three-photon excitation, it could be excited at 750 nm;
thus the fluorescence observation could be realised in
normal system [67].
The excitation light collected by the objective has
a light intensity, which is inversely proportional to the
Immersion
medium
Specimen
Cell culture dish
Culture
medium
Access
Angle
45°
20°
Water dipping Conventional
AB
Figure 22:Comparison of (A) water dipping objective and (B)
conventional water immersion objective.
Unauthenticated
Download Date | 6/14/19 7:56 PM
Y. Zhang and H. Gross: Systematic design of microscope objectives. Part I29
square of the distance from the focal plane in geometri-
cal approximation. Therefore, the multiphoton excita-
tion only occurs in the vicinity of the focal point, and
thus, fluorescence is only emitted from the focal plane.
Advantageously, fluorescence in samples is subject to
less discolouration, and the confocal setup with confocal
pinholes is no longer necessary to realise depth imaging.
Consequently, multiphoton microscopy became the most
popular tool for 3D imaging of live cell and specimen.
However, also owing to these effects, design of mul-
tiphoton microscope objectives became more challeng-
ing. First, to reduce the area of excitation and emission,
larger NA is preferred to achieve better resolution. Second,
because the excitation (IR) and emission (UV/VIS) with
different wavelengths occurred at the same position, the
chromatic aberration of wide spectrum from UV to IR must
be well corrected, as the class (g). The objective should
also have excellent transmittance through the full spec-
tral range. Third, although the IR excitation light has low
scattering, the emission light, which lies within UV and
VIS, may become diffused in the specimen. Therefore, to
collect the scattered fluorescence without loss, the objec-
tive should be designed with large field. Combining with
the first requirement, the etendue is enlarged. Lastly, for
a large depth imaging, the working distance should also
be comparably enlarged, but typically still within 0–0.5f.
Concerning the different depth, to reduce the spheri-
cal aberration induced by the index change of specimen
and realise fast autoscanning, the objective should be
designed with a moving group for depth correction.
5.5 Total internal reflection fluorescence
microscopy (TIRFM)
The TIRFM utilises total internal reflection (TIR) in prism
or cover glass and thus could be classified as trans-TIRFM
(prism- and lightguide-based TIRFM) and cis-TIRFM
(through-objective TIRFM), respectively. In this section,
we only focus on the cis-TIRFM utilising high NA objective.
The excitation of TIRFM is produced by the TIR with the
evanescent wave at the surface between the cover glass
and specimen. Therefore, the penetration depth is very
small, approximately as the wavelength of the illumina-
tion light. By only exciting such a thin section, extremely
high SNR could be achieved. To realise the TIR at the cover
glass, the NA of TIRF objective should be larger than 1.38,
which is the typical refractive index of cells. The practi-
cal TIRF objectives are typically designed with NA > 1.42,
and according to the discussion in Section 4.2, Nikon and
Olympus utilised special oil to make the NA exceeding
1.65, even reaching 1.70. Nevertheless, the special oils
have strong autofluorescence, hampering the image con-
trast. TIRF objectives utilising typical Type A oil with NA
between 1.45 and 1.49 are favorable.
However, belonging to the high NA objectives, the
requirement of TIRF objectives is different from the typical
high NA objectives for confocal fluorescence microscopy.
According to Sections 5.2 through 5.4, the typical high NA
systems require large object field for more efficient obser-
vation. On the contrary, to further improve the SNR, the
TIRFM only illuminates a small area; thus the TIRF objec-
tives are typically designed with large magnification, typi-
cally from 60× to 160×.
Another special requirement of TIRF objective is the
CORR function of CG thickness and temperature. Owing
to the application principle of TIRF, the excitation laser
beam should be perfectly focused onto the front surface
of cover slip. Therefore, when the CG thickness is under
bias use, the induced spherical aberration should be cor-
rected. Furthermore, immersion oil always suffers from
critical index thermal change and thermal expansion.
When the TIRF is used under body temperature, the cor-
responding error should also be corrected. As a typical
range, TIRF objectives have CORR function of CG thick-
ness 0.13–0.19mm at 23°C and 0.140.20mm at 37°C.
5.6 Virtual slide microscopy
The digital image sensor developed rapidly in the recent
decade. The number of pixels of the sensor used for digital
microscopy has increased remarkably. There is a growing
demand for the microscope apparatus to achieve both
a wide field of view and high resolving power in obser-
vation. Fulfilling these requirements, the virtual slide
microscope became the state-of-the-art research tool for
brightfield and fluorescence microscopy.
The basic idea of a virtual slide microscope is the
combination of wide field fluorescence observation and
confocal Z-stack scanning. However, compared with
conventional microscope objective for confocal setup,
the magnification is further reduced to 10×–20 × achiev-
ing object diameter of 1.5–2.5mm, and the NA is further
enlarged reaching 0.9–0.95 for dry lens and 1.35–1.45 for
oil immersion objectives. Thereby, the extreme etendue of
microscope objectives is required.
When it comes to the system performance, to realise
efficient field scanning, the same as conventional confo-
cal setup, all the field aberrations should be corrected. The
objective should also be at least apochromatic corrected
for VIS and with excellent transmittance from UV to VIS,
Unauthenticated
Download Date | 6/14/19 7:56 PM
30 Y. Zhang and H. Gross: Systematic design of microscope objectives. Part I
which correspond to the requirement of general fluores-
cence microscopy. However, because of the extremely high
etendue, it became more challenging to correct both the
axial aberrations (spherical aberration and longitudinal
chromatic aberration) and the field aberrations. For best
case, the longitudinal aberrations should be corrected as
class (c) or class (d), and field aberrations should be cor-
rected without vignetting to ensure uniform illumination
of the full field as class 7. Nevertheless, as introduced in
Section 4.3.3, parfocal length must be sacrificed to achieve
this level correction. Furthermore, with the small mag-
nification, as the exit pupil size increases significantly,
the coma cannot be controlled. Therefore, some state-of-
the-art systems [55] utilised retrofocus structure, which
increased the ray height in the middle group but reduced
the relative ray height in the rear group to control coma,
while some other state-of-the-art systems [38] abandoned
the infinity optics to achieve perfect axial and field cor-
rection. Generally speaking, to realise the extremely high
etendue correction, trade-off between vignetting, parfocal
length and entire microscope size must be made to deter-
mine the basic objective structure.
The virtual slide microscope was invented to achieve
high scanning speed. However, to operate Z-stack scan-
ning, the conventional method that moves the entire heavy
objective would induce vibrations and the speed is low.
Introducing special optical elements with zooming effect,
such as adaptive elements and Alvarez plate, would bring
with necessity of pupil arrangement, thus increasing
system complexity. Utilising DOE is also feasible, but typi-
cally results in critical stray light. Consequently, a light-
weight moving group should be used in the virtual slide
microscope objectives for efficient depth adjustment.
5.7 Semiconductor industrial applications
From 1980s, following the flourish of semiconductor
industry, a series of microscope objectives was specifi-
cally designed for industrial inspection, repairing and
fabrication. The observation sample was not fixed as
semiconductor chips, and the application was not fixed as
inspection, but in this paper, we mostly named this series
of objectives as semiconductor inspection objective.
The semiconductor inspection lens is used to resolve
the fine structure of the sample. Therefore, except some
very low magnification objectives for positioning, the
conventional semiconductor inspection lens requires
large magnification (50×–25 0×) and high NA without
immersion. Moreover, a long working distance is also
required for repairing and fabrication. However, it is
well known that with increasing magnification (reduc-
ing focal length), the chromatic aberration gets critical
and with increasing working distance the spherical aber-
ration, field curvature, coma and chromatic aberration
gets severe [68]. Combining these two effects, therefore,
as the retrofocus structure is applied, the enlargement of
NA has contradiction with the extension of W.D. Compar-
ing the class 4 and class 5 introduced in Section 4.3.1,
trade-off between NA and W.D. should be made for the
high magnification industrial objective. Typical class
4systems are designed as 100×/0.50with k= 6.5, and the
class 5systems are designed as 100×/0.90with k= 1.12.
Recently, utilising advanced digital image sensor, high-
resolution wide-field image could be obtained with
postmagnification. Therefore, similar to virtual slide
microscope, the low magnification high NA objective
with long working distance has become a new interest.
However, owing to the complexity, the number of inven-
tions is still limited.
According to the introduction in Section 4.1.1, regard-
ing operation spectrum, except the DUV objectives, the
semiconductor inspection objectives should work for
VIS and possible extension to UV or IR, which are mostly
based on the wavelength of YAG laser and its harmonics
for laser repairing. The semiconductor industrial objec-
tives extended the corrected spectrum and improved the
transmittance before the biomedical fluorescence micro-
scope objectives. Concerning the correction level, most
of the biomedical fluorescence objectives at least real-
ised Apochromate correction in VIS, but some industrial
objectives only, achieved Achromate or Fluorite correction
through the full spectrum.
The semiconductor inspection objectives typically
require better field correction than the biomedical objec-
tives with similar etendue and colour correction. For one
thing, the illumination uniformity should be superb;
thus, vignetting cannot be accepted. In addition, higher-
order field aberrations including the chromatic variation
of coma, which are not considered in other applications,
could also influence the precise inspection. The class
5 field correction is usually required for the industrial
objectives.
The conventional semiconductor industrial objectives
were mostly designed without CORR function. However,
in the recent 20years, there is an increasing demand of
inspection under glass plate, particularly for the inspec-
tion of liquid crystal substrate. The variable thickness of
the substrate is typically around 2–5 mm. Therefore, the
corresponding inspection lens should be designed with
large-range CG correction, the same as the objective used
for biological IM setup.
Unauthenticated
Download Date | 6/14/19 7:56 PM
Y. Zhang and H. Gross: Systematic design of microscope objectives. Part I31
Table 7:Impact of applications on the systems parameters of microscope objectives.
Applications
Etendue
Spectral range Field correction Working
distance
CORR function Others
NA Magnification TransmittanceColour correction
Conventional
objective
VIS Basic Ach-/Fluor-/Apo-
Class (a)–(c)
Only field curvature
Class –
CG
Conventional IM VIS Basic Ach-/Fluor-/Apo-
Class (a)–(c)
Only field curvature
Class –
LDLarge range CG
Confocal setup VIS Apo-
Class (c)–(d)
Petzval+field aberration
Class –
Immersion/
depth
Fluorescence
microscopy
 
General UV-VIS Apo-/UV-VIS (i-C) Superapo-
Class (c)–(g)
Petzval+field aberration
Class –
CG/immersion Autofluorescence
General live cell UV-IR VIS-IR (g-t) Superapo-
Class (d), (f), (g)
Petzval+field aberration
Class –
LD Temperature Possible CG/immersion/
temperature CORR
Multiphoton UV-IR UV-IR (i-t) Superapo-
Class (f), (g)
Petzval+field aberration
Class –
LD Depth
TIRF ↑↑↑UV-VIS Apo-
Class (c)–(d)
Only field curvature
Class –
Fixed CG/
temperature
Virtual slide
microscopy
↑↑ ↓↓↓ UV-VIS Apo-
Class (c)–(d)
Petzval + field aberration↑↑
Class –
Depth
Semiconductor
inspection
VIS+UV/IR Basic Ach-/Fluor-/Apo-
UV Superapo-/IR Superapo-
Class (a)–(f)
Petzval+field aberration↑↑↑
Class 
LD↑↑↑ (Large range
CG)
-
, Enlarged/improved; ↑↑, significantly enlarged/improved; ↑↑↑, greatly enlarged/improved; , reduced; ↓↓↓, greatly reduced; IM, inverted microscope; TIRF, total internal reflection
fluorescence microscopy; CORR, environmental correction; CG, cover glass thickness; VIS, visible spectrum; UV, ultraviolet; IR, infrared radiation; Ach-, achromate; Fluor-, fluorite; Apo-,
apochromate; Superapo-, superapochromate; LD, long working distance.
Unauthenticated
Download Date | 6/14/19 7:56 PM
32 Y. Zhang and H. Gross: Systematic design of microscope objectives. Part I
5.8 Miscellaneous
Most of the remarkable biomedical and industrial appli-
cations have been discussed above. Some additional
important applications also have slight impact on the
basic parameter and system complexity of the microscope
objectives.
Metallurgical microscope utilised epi-illumination to
observe the opaque substances with brightfield or dark-
field observation. The metallurgical microscope has been
developed over centuries, and the typical magnification,
NA and system structures are similar to the conventional
biomedical objectives. Concerning the observation of
metal, under the epi-illumination, the veiling glare gen-
erated by the air-to-lens interfaces, which are not well
coated, would significantly influence the metallurgical
analysis. Therefore, typically more cemented elements are
used in the metallurgical microscope objectives instead of
single lens to control the stray light.
Polarization microscope utilised polarised light for
observation. Therefore, to achieve the best image, it is impor-
tant to reduce the intrinsic influence of the objective on the
polarization. For one thing, utilisation of crystal should be
avoided, including calcium fluoride, which is commonly
used to correct secondary spectrum for Apochromate. For
another, the impact of strain and birefringence from the
optical elements, optical cements and coatings should
be reduced. The objective should be designed with fewer
elements. Because of the limit of material and number of
elements, typically the polarization microscope objectives
could only reach a medium NA and small etendue with low-
level chromatic correction at class (a) or class (b) and low-
level field correction at class 1 and class 2.
Contrast methods, such as phase contrast and DIC,
are widely used in modern microscopy. The system struc-
ture should ensure the manipulation on the pupil of the
objective. However, for the high-performance systems, the
space in the objective is usually fully filled with lenses,
and the exact pupil position is inside the objective. One
solution is to put an artificial aperture at the rear part of
the objective, as shown in Figure 17, but the manipula-
tion is not accurate. Another solution is introducing two
th three more special lenses to design the objective with
accessible pupil behind the glass, but the system com-
plexity is slightly increased. It is also feasible to utilise
pupil relay system, but the mechanical track of the whole
system is enlarged. Except this effect, the contrast method
does not have other impact on the complexity of micro-
scope objective.
Recently, following the development of the super-
resolution localization microscopy methods, many
manufactures specifically produced specific high-per-
formance objectives. The basic design consideration is
the same as basic high NA high magnification confocal
and fluorescence microscope objective. Both the chro-
matic and field correction should reach high level class
(d) and class 5 with excellent CORR function of CG and
temperature correction. A new feature of the localization
microscopy is the strong illumination intensity. Under the
high NA, the strong power gathered at the front group may
damage the surface and cementing (the glue is usually
sensitive to the UV light). Therefore, the front group is
sometimes specifically rearranged.
5.9 Summary
Impacts of the discussed applications on the design
ofmicroscope objectives are summarised in Table 7.
6 Impact of manufacture and
technology consideration
Compared with the impact of applications, the impact of
manufacture and technology consideration is far more dif-
ficult to be systematic summarised. Most of the considera-
tions originated from the technique roadmaps and process
details of the manufacturer, which are typically kept as a
secret. Because our work is to analyse the general lens
modules influencing system complexity, we would focus
on the most important four considerations as follow:
1. Manufacture accuracy of the optical elements
2. Element mounting and air gaps alignment
3. Dimensional restriction
4. Stray light control
The detailed discussion of each lens modules with these
considerations will be given in Part II. In this section, two
examples are used to illustrate the consideration of 1&2
and 3&4, respectively.
Example I
Figure 23 shows three different front group structures of
high magnification oil immersion objectives from prac-
tical patented systems. All the objectives are apochro-
matic corrected for visible range from g-line to C-line.
The front group (A) utilised a simple hemisphere made
of crown glass with matched index to the immersion oil.
The front group (B) is the most popular layout, utilis-
ing an embedded small crown lens with matched index,
Unauthenticated
Download Date | 6/14/19 7:56 PM
Y. Zhang and H. Gross: Systematic design of microscope objectives. Part I33
which is cemented with a meniscus flint lens with strong
power. The front group (C) is based on the new technique
‘optical contact bonding’. The strong positive crown lens
is designed as a hyper-hemisphere and bonded with a
plane plate.
The details of structural functionality in aberra-
tion correction will be discussed in Part II. Concerning
the manufacture accuracy of optical elements, in the
popular front group (B), although the embedded lens has
great advantage in field curvature correction and spheri-
cal aberration restraint, owing to its small clear aperture
(~1mm), it is difficult to produce. The ‘ball technology’ is
usually used for the production, but the cost is relatively
high, particularly to realise high accuracy of the cement-
ing. Therefore, for the cost-driven systems, the front lens
is designed with a single lens. However, if the refractive
index of the lens is not large enough, to achieve high NA,
the shape of the lens tends to be hyper-hemisphere, which
cannot be mounted to the leading edge of the objective.
The front group (A) nearly reaches the maximum NA
(1.25), which could be achieved by single hemispherical
lens with d-line refractive index around 1.52.
According to Section 5.8, for superresolution localiza-
tion microscopy, owing to the strong illumination inten-
sity at the front group, the cementing surface in the (B)
type front group is vulnerable. However, if a single lens
is used, the maximum NA cannot fulfill the required
strong excitation. Utilising front group (C), the NA could
be further enlarged with the hyper-hemisphere, and the
optical contact bonding is not sensitive to the strong UV
illumination. Furthermore, the mounting of the front
group could be fixed at the edge of the thick plane plate;
thus the system assembly is feasible. The only risk is the
robustness of the ‘optical contact bonding’ technique.
According to this comparison, it could be seen that to
determine the shape of the front group, all three design
concepts are involved. Functionality in aberration correc-
tion determines the basic shape. The application fixed the
boundary condition. The final trade-off is associated with
the consideration of manufacture and technology.
Example II
Comparison shown in Figure 24 is similar to that in
Figure 16, but the difference between these two 60×
TIRF objectives is the choice of doublets or triplets in
ABC
Cover slip Oil immersion
Figure 23:Front groups of high magnification oil immersion microscope objectives. (A) 100×/1.25 Shi US 7907348, (B) 100×/1.25
Shoemaker US 3902793, (C) 160×/1.43 Bauer US 9488818.
45 mm
60 mm
60×/1.48 O SF22 Fujimoto US 7199938
60×/1.45 O SF22 Mandai US 7046451
Figure 24:Layouts of the 45mm parfocal and 60mm parfocal 60×
oil immersion TIRF objectives from Olympus and Nikon.
Unauthenticated
Download Date | 6/14/19 7:56 PM
34 Y. Zhang and H. Gross: Systematic design of microscope objectives. Part I
correcting spherical and chromatic aberrations. Both
these two systems are apochromatic corrected for visible
range from g-line to C-line and utilised vignetting for field
correction.
The different parfocal length determines the mechan-
ical dimension to fill lenses. The 60×/1.48 45mm parfo-
cal objective from Olympus used two cemented triplets in
the middle group, whereas the 60×/1.45 60 mm parfocal
objective utilised four doublets. These two setups have
similar functionality in spherical and chromatic aberra-
tion correction. But the triplet setup could relatively save
space (also beneficial from the stronger front lens).
The selection of cemented doublets or triplets also
highly depends on the coating technique of the manufac-
turers. Because of the strong veiling glare generated by
the air–glass interface, if excellent antireflection cannot
be realised by coating, the air–glass interface with strong
curvature should be avoided. Consequently, except the
functionality of aberration correction, the cemented tri-
plets utilised in the high-performance objectives are typi-
cally designed with stronger inner cemented surface and
flatter outer surfaces. On the contrary, if the veiling glare
could be controlled by coating, utilising cemented dou-
blets instead of triplets could introduce additional degree
of freedom for correction, which is advantageous in relax-
ing the system sensitivity.
7 Conclusion
As the basic step of modular analysis and microscope
objectives synthesis, the systematic review of modern
systems and a new systematic classification have been
implemented. Six-zones-classification based on etendue,
seven classes of chromatic correction, seven classes of
field correction, five classes of working distance exten-
sion, four types of correction and strategies of vendors in
parfocal length and tube lens selection have been sorted
systematically, which is helpful to understand the system
complexity.
Based on the classification, following the historical
review, impacts of applications and manufacture and
technology considerations are summarised. The phe-
nomenological Epoché could be operated by decoupling
these two factors, and thus, we could concentrate on
the functionality of lens modules in aberration correc-
tion, which will be discussed in Part II. For the users of
microscope, understanding the impact of applications is
also helpful to evaluate the complexity and cost of their
setup.
References
[1] A. H. Bennett, J. Opt. Soc. Am. 33, 123–128 (1943).
[2] H. Beyer, in ‘Handbuch der Mikroskopie’, (VEB Verlag Technik,
Berlin, 1988) pp. 13–23.
[3] H. Gross, F. Blechinger and B. Achtner, in ‘Handbook of Optical
Systems’, (Survey of Optical Instruments, Wiley, Weinheim, 2008).
[4] H. Boegehold, in ‘Das optische System des Mikroskops’, (VEB
Verlag Technik, Berlin, 1958).
[5] W. Klein, Jahrbuch für Optik und Feinmechanik. 47–86 (1976).
[6] H. Riesenberg, in ‘Handbuch der Mikroskopie’, (VEB Verlag
Technik, Berlin, 1988) pp. 24–107.
[7] B. G. Broome, Proc. SPIE 10263, 102630H (1992).
[8] H. Takenaka, Proc. SPIE 0554 (1986).
[9] A. Miks and J. Novak, Appl. Opt. 49, 3403–3410 (2010).
[10] Frolov, J. Opt. Tech. 69, 614 (2002).
[11] H. C. Claussen, Appl. Opt. 3, 993–1003 (1964).
[12] Y. Zhang and H. Gross, Proc. SPIE 10590, 105901G (2017).
[13] E. Husserl, in ‘The Paris Lectures’, (Springer, Netherlands,
DenHaag, 1964).
[14] D. Shafer, in ‘My 50Years of Optical Design Using Husserl’s
Phenomenology’, (Rochester, 2016).
[15] H. Boegehold, US 1578259 (1926).
[16] C. Schulz and C. Schumann, DE102017108595 (2018).
[17] H. Boegehold, US 2206155 (1940).
[18] W. Klein, US 3524694 (1970).
[19] A. H. Shoemaker, US 3700311 (1972).
[20] W. Klein, US 3537773 (1970).
[21] A. Taira, US 3744881 (1973).
[22] A. Taira, US 3806231 (1974).
[23] M. Matsubara, US 3925910 (1975).
[24] W. Klein, US 3176583 (1965).
[25] T. Tojyo, US 4059342 (1977).
[26] K. Yonekubo, US 4174151 (1977).
[27] T. Tojyo, US 4279477 (1981).
[28] Y. Saito, US 5530590 (1996).
[29] R. Shi, US 7349162 (2008).
[30] T. Kimura, JP S60-63512 (1985).
[31] T. Togino, JP S62-49313 (1987).
[32] S. Kashima, JP H05-196875 (1993).
[33] I. Ito, US 6181480 (2001).
[34] K. Arisawa, US 6016226 (2000).
[35] Y.-H. Chuang, D. Shafer and J. J. Armstrong, US 7245438 (2007).
[36] R. Wartmann, US 9645380 (2014).
[37] T. Kasahara, US 9323036 (2013).
[38] K. Abe, US 9939622 (2016).
[39] R. Shi, US 7672057 (2010).
[40] Carl Zeiss Microscopy GmbH, Objectives from Carl Zeiss
Exeeding your Expectations (2007).
[41] K. H. Schade, P. Euteneuer and A. Müller-Rentz, Mitteilungenfür
Wissenschaft und Technik, (Wetzlar, 1998).
[42] C. Metz, Zeitschr. für wiss. Mikroskopie. 37, 49–52 (1920).
[43] T. Suzuki, US 5659425 (1997).
[44] K. Watanabe, JP 2002-098903 (2002).
[45] K. Yamaguchi, US 6519092 (2003).
[46] N. Murayama and Y. Shimizu, US 5757552 (1995).
[47] K. Arisawa, JP H04-220615 (1992).
[48] S. F. Gibson and F. Lanni, J. Opt. Soc. Am. A. 9, 154–166 (1992).
[49] L. Alvarez, US 3305294 (1967).
Unauthenticated
Download Date | 6/14/19 7:56 PM
Y. Zhang and H. Gross: Systematic design of microscope objectives. Part I35
[50] J. Babington, Proc. SPIE 9626, 962615 (2015).
[51] T. Nobis, US 9696536 (2016).
[52] I. Fahlbusch and G. Herbst, US 8988797 (2011).
[53] Y. Ouchi, US 6366398 (1996).
[54] R. Wartmann, DE 102009037743 (2009).
[55] K. Abe, US 9746658 (2016).
[56] T. Tojyo, US 4232941 (1977).
[57] K. Arisawa, JP H05-196873 (1993).
[58] J. L. Smith, The Inverted Microscope (a New Form of Microscope)
American Journal of Science and Art, 2nd series, Vol. XIV, (1852).
[59] M. Akahane, US 20170184830 (2015).
[60] M. Minsky, US 3013467 (1961).
[61] P. Davidovits and M. D. Egger, US 3643015 (1972).
[62] K. Carlsson, etal., Opt. Lett. 10, 53–55 (1985).
[63] J. B. Pawley, in ‘Handbook of Biological Confocal Microscopy’,
(Springer, Boston, MA, 2006).
[64] T. Wilson, ‘Confocal Microscopy’, (Academic Press, New York,
1990).
[65] S. Kashima, US 5406421 (1992).
[66] M. Matthae and G. Herbst, US 20130100537 (2013).
[67] S. Kashima, US 7042648 (2003).
[68] T. Suzuki, US 5132845 (1990).
Yueqian Zhang
Institute of Applied Physics, Friedrich
Schiller University Jena, Albert-Einstein-Str.
15, 07745 Jena, Germany
yueqian.zhang@uni-jena.de
Yueqian Zhang did his undergraduate study in Optical Engineering
at Zhejiang University, Hangzhou, China. He received his Master
degree in Photonics from Friedrich-Schiller-Universität Jena,
Germany, in 2015. Since 2016, he has been working in the Optical
Design Group at the Institute of Applied Physics in Friedrich-Schiller-
Universität Jena. His research interests are classical system design,
microscopic application and system development.
Herbert Gross
Institute of Applied Physics, Friedrich
Schiller University Jena, Albert-Einstein-Str.
15, 07745 Jena, Germany
Herbert Gross studied Physics at the University of Stuttgart. He
received his PhD on Laser Simulation in 1995. He joined Carl Zeiss
in 1982where he worked as a scientist in optical design, modelling,
and simulation. From 1995 to 2010, he headed the central depart-
ment of optical design and simulation. Since 2012, he has been a
professor at the University of Jena in the Institute of Applied Physics
and holds a chair of Optical System Design. His main working areas
are physical optical simulations, beam propagation, partial coher-
ence, classical optical design, aberration theory, system develop-
ment, and metrology. He was editor and main author of the book
series ‘Handbook of Optical Systems’.
Unauthenticated
Download Date | 6/14/19 7:56 PM
... Compensating field curvature requires complicated, large and heavy objective lenses 39 . A fully uncompensated field curvature led to about 10 µm axial distance between the axial focal point and the outer part of the FOV. ...
Article
Full-text available
Advances in head-mounted microscopes have enabled imaging of neuronal activity using genetic tools in freely moving mice but these microscopes are restricted to recording in minimally lit arenas and imaging upper cortical layers. Here we built a 2-g, three-photon excitation-based microscope, containing a z-drive that enabled access to all cortical layers while mice freely behaved in a fully lit environment. The microscope had on-board photon detectors, robust to environmental light, and the arena lighting was timed to the end of each line-scan, enabling functional imaging of activity from cortical layer 4 and layer 6 neurons expressing jGCaMP7f in mice roaming a fully lit or dark arena. By comparing the neuronal activity measured from populations in these layers we show that activity in cortical layer 4 and layer 6 is differentially modulated by lit and dark conditions during free exploration.
... While the NODO imaging path can provide sub-micrometer resolution, the FOV of the multi-immersion primary objective (O1) is restricted to 1 mm, which is insu cient for fast mesoscopic imaging. This tradeoff between NA and FOV is standard across all currently available microscope objectives (e.g., no current clearing-compatible objectives can simultaneously offer sub-micrometer resolution over a mesoscopic FOV) (Supplementary Fig. 8) [49]. Therefore, in our hybrid system, we achieve low-resolution imaging with a second independent ODO imaging path. ...
Preprint
Full-text available
Light-sheet microscopy has emerged as the preferred means for high-throughput volumetric imaging of cleared tissues. However, there is a need for a user-friendly system that can address imaging applications with varied requirements in terms of resolution (mesoscopic to sub-micrometer), sample geometry (size, shape, and number), and compatibility with tissue-clearing protocols and sample holders of various refractive indices. We present a ‘hybrid’ system that combines a novel non-orthogonal dual-objective and conventional (orthogonal) open-top light-sheet architecture for versatile multi-scale volumetric imaging.
... For instance, a shorter wavelength can be used to achieve the same resolution with a lower NA than a longer wavelength. A low illumination NA can increase the light sheet length, and a low detection NA can make the detection objective lens design easier [24]. On the other hand, a longer wavelength is scattered less and therefore can provide a larger imaging depth [25,26]. ...
Article
Scattering-based light sheet microscopy (sLSM) is a microscopy technique that can visualize cellular morphologic details based on the scattering signal. While sLSM was previously shown to image animal tissues ex vivo at a cellular resolution, the wavelength used was chosen based on other in vivo microscopy technologies rather than through a comparison of the sLSM imaging performance between different wavelengths. In this paper, we report the development of a multi-wavelength sLSM setup that facilitates the investigation of different wavelengths for sLSM imaging. Preliminary results of imaging human anal tissues ex vivo showed that the sLSM setup allowed for comparisons of the cellular imaging performance at the same tissue location between different wavelengths. Both the quantitative analysis of the image contrast and the visual assessment by a pathologist showed that the imaging depth increased with wavelength, and the imaging depth increase was most notable around 600 nm. The preliminary results showed that the multi-wavelength sLSM setup could be useful in identifying the optimal wavelength for the specific tissue type.
Chapter
This chapter discusses how infinity-corrected microscopes work, as well as the principles of optics that are applied to their development. Proper designing and good alignment of an optical microscope are essential for accurate studies of cells, observation of cellular growth, identification, and counting of cells. To design an optical microscope the two important aspects are, namely, a better understanding of the function of each component and how their control influences the resulting images. The design of the infinity-corrected optics is routinely incorporated into multiple lenses, filters, polarizers, beam-splitters, sensors, and illumination sources. This chapter discusses the development of microscope with infinity optics and the design of infinity-corrected optics with optical ray diagrams. The microscope design parameters and aberrations are discussed to understand the necessity of multiple lens objective system. To get the best resolution and contrast, the condition for Köhler illumination should be maintained within a microscope. The unstained sample is unable to image in bright field microscopy. The chapter also discusses label-free techniques with infinity-corrected optics, such as dark field microscopy, Zernike phase contrast microscopy, differential interference contrast (DIC), and digital holographic microscopy and their applications to study the various type of specimens without dye or label.
Article
Recent advancements in additive manufacturing have enabled new methods of fabricating gradient-index (GRIN) optics by blending multiple materials in the deposition process. A design study highlighting the advantages of multi-material GRIN optics is presented. It is shown that additional materials in the GRIN allow for higher orders of color correction. A new multi-material refractive index representation, which constrains the GRIN to real materials, is also presented.
Article
Full-text available
The correction of modern microscope objectives is not usually discussed in literature. We have reported a system review and summarized the design principles in a series of papers in 2019 [1-3]. Here we are introducing the systematic view of microscope objective design with an extension of the database till 2021. Furthermore, a systematic synthesis approach aided by AI will also be discussed.
Article
Fast volumetric imaging of large fluorescent samples with high-resolution is required for many biological applications. Oblique plane microscopy (OPM) provides high spatiotemporal resolution, but the field of view is typically limited by its optical train and the pixel number of the camera. Mechanically scanning the sample or decreasing the overall magnification of the imaging system can partially address this challenge, albeit by reducing the volumetric imaging speed or spatial resolution, respectively. Here, we introduce a novel dual-axis scan unit for OPM that facilitates rapid and high-resolution volumetric imaging throughout a volume of 800 × 500 × 200 microns. This enables us to perform volumetric imaging of cell monolayers, spheroids and zebrafish embryos with subcellular resolution. Furthermore, we combined this microscope with a multi-perspective projection imaging technique that increases the volumetric interrogation rate to more than 10 Hz. This allows us to rapidly probe a large field of view in a dimensionality reduced format, identify features of interest, and volumetrically image these regions with high spatiotemporal resolution.
Chapter
SEM analysis attains a special role and wide acceptance in biomedical researches as it is helping in the better imaging and analysis of different biological specimens. Preparation of the biological samples is one of the crucial steps which in turn affects the result. This chapter briefs the important techniques and procedures involved in the processing and preparation of the biological samples for the SEM analysis. The sampling process starts with fixation where it allows the sample to adhere properly to the following processes. Various types of chemicals, traditional chemical, physical fixatives are employed for the same along with vapour fixation techniques. The sample preparation will be then further proceeded with dehydration and drying methods to ensure the complete removal of water from the biological samples. Air drying, critical point drying, freeze-drying, chemical drying etc. are a few such processes before the mounting where it provides a firm binding of the sample to the holder. Then the biological samples can be analyzed for their features based on SEM images generated.
Chapter
This chapter provides an overview of light and fluorescence microscopy, beginning from the basic principles of optics such as transmission, absorption, diffraction, refraction, and behavior electromagnetic theory. To understand the concepts of magnification and resolution, this chapter explores the theory of formation of real and virtual images and some properties of the lenses that compound the optical system of a microscope, including the aberration problems and corrections applied in the objectives. On the other hand, this chapter explores the concepts of Abbe resolution and the difference with the magnification of the optical system in the microscope. The description of the mechanical components and light sources of optical and fluorescence microscopes allow understanding of the difference between microtechniques applied to each of these microscopes and utilization in biology and materials science. These basic principles of microscopy will allow nonexpert users to contextualize the first results of observations in the microscope. The main limitations in the microscopist formation are the high cost of the equipment and the unavailability of the expert users to teach, it is for these reasons we consider that the training should begin with a rapid guide of basic microscopy concepts.KeywordsMicroscopyLightFluorescenceWavelengthSample holder
Article
Full-text available
We analyze and synthesize one type of planachromatic microscope objective lens (planachromats). Formulas are described for the calculation of basic parameters of such optical systems for various configurations of the optical system. The application of the described technique is shown on an example of the optical design of planachromat.
Book
The state-of-the-art full-colored handbook gives a comprehensive introduction to the principles and the practice of calculation, layout, and understanding of optical systems and lens design. Written by reputed industrial experts in the field, this text introduces the user to the basic properties of optical systems, aberration theory, classification and characterization of systems, advanced simulation models, measuring of system quality and manufacturing issues. In this Volume: Volume 4 presents a survey of optical systems, based on the principles of image formation, optical system setup and quality control which are covered by the first three volumes. Starting with the human eye, the chapters discuss all systems, from telescopes and binoculars to projection, spectroscopic and illumination systems. All these systems are characterized and described using coherent schemes and criteria to provide readers with a thorough background for their own developments. Other Volumes: Volume 1: Fundamentals of Technical Optics; Volume 2: Physical Image Formation; Volume 3: Aberration Theory and Correction of Optical Systems; Volume 5: Advanced Physical Optics.
Conference Paper
We consider using the Alvarez lens concept to perform focal length change in conventional optical systems. The Alvarez pair are a good example of freeform surfaces that are used to imprint a deformation into the propagating wavefront. In addition, we try to better understand the paraxial theory of each freeform component in building up to a composite lens system. An example dual field of view system in the medium wave infrared is presented. An inherent feature of the Alvarez pair is the axial symmetry breaking due to both the finite air gap between the cubic surfaces and the transverse movement of the pair. This has implications for the wavefront at the image plane. Having developed the first order theory one can better understand misalignment tolerances and how these produce certain wavefront aberrations. Most notably, misalignments lead to simple expressions in terms of the Zernike polynomials.
Book
Foundations of Confocal Scanned Imaging in Light Microscopy.- Fundamental Limits in Confocal Microscopy.- Special Optical Elements.- Points, Pixels, and Gray Levels: Digitizing Image Data.- Laser Sources for Confocal Microscopy.- Non-Laser Light Sources for Three-Dimensional Microscopy.- Objective Lenses for Confocal Microscopy.- The Contrast Formation in Optical Microscopy.- The Intermediate Optical System of Laser-Scanning Confocal Microscopes.- Disk-Scanning Confocal Microscopy.- Measuring the Real Point Spread Function of