See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/235405724

Optimization of thickness uniformity of optical coatings on a conical

substrate in a planetary rotation system

Article  in  Applied Optics · February 2013

DOI: 10.1364/AO.52.000B26 · Source: PubMed

CITATIONS READS

12 1,165

4 authors:

Chun Guo Mingdong Kong

The institute of Optics and Electronics 22 PUBLICATIONS   82 CITATIONS   
21 PUBLICATIONS   132 CITATIONS   
SEE PROFILE
SEE PROFILE

Cunding Liu Bincheng Li

University of Science and Technology of China University of Electronic Science and Technology of China
18 PUBLICATIONS   151 CITATIONS    237 PUBLICATIONS   1,516 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Optical films for DUV lithography systems View project

Fabrication of ultralow stress optical coatings by plasma-ion assisted deposition View project

All content following this page was uploaded by Cunding Liu on 23 October 2015.

The user has requested enhancement of the downloaded file.

 Optimization of thickness uniformity of optical
coatings on a conical substrate
in a planetary rotation system

Chun Guo,1,2 Mingdong Kong,1 Cunding Liu,1 and Bincheng Li1,*

1
Institute of Optics and Electronics, Chinese Academy of Sciences, Chengdu 610209, China
2
University of Chinese Academy of Sciences, Beijing 100039, China
*Corresponding author: [email protected]

Received 5 September 2012; revised 21 November 2012; accepted 22 November 2012;

posted 26 November 2012 (Doc. ID 175226); published 2 January 2013

For a coating machine with a planetary rotation system and counterrotating shadowing mask configura-
tion, a shadowing mask was designed using a numerical optimization algorithm to control the thickness
uniformity of optical coatings formed on conical substrate. Single-layer magnesium fluoride (MgF2 ) and
antireflective (AR) coating at 193 nm were fabricated on a convex conical substrate holder (with diameter
225 mm, apex angle 140 deg, and height 41 mm) by thermal evaporation. Thickness distribution deter-
mined from the transmittance spectra of single-layer MgF2 thin films on BK7 slices showed that uni-
formities better than 99.3% were experimentally achieved with the designed counterrotating shadowing
mask. From the reflectance spectra, uniform optical performance was also obtained for the 193 nm AR
coating deposited on fused-silica substrates. © 2013 Optical Society of America
OCIS codes: 310.0310, 310.1860, 310.3840.

1. Introduction addition, the gradient of the coating thickness along

An axicon lens is a common optical element normally the radical direction of the conical substrate may
used to form a continuous focal segment along the have significant impact on image formation. It is
optical axis. Thanks to such performance, an axicon therefore crucial to control the thickness uniformity
lens plays a significant role in many optical systems of multilayer coatings for conical substrates in preci-
[1–5]. For instance, it is used to improve the resolving sion optical applications.
power and extend the depth of focus for off-axis illu- Conventionally, a rotating or fixed shadowing
mination in some optical systems. Normally, uniform mask is employed in a simple or planetary rotation
coatings are required to enhance the optical perfor- system in a coating machine to selectively obstruct
mance of the axicon lens. Nevertheless, if not prop- the deposition plume profile, and therefore optimize
erly corrected, for a large-diameter axicon lens, the thickness uniformity of optical coatings depos-
coating performance at the edge of the conical sub- ited upon optics [6–11]. Most recently, a gyro and
strate can be substantially different from that in modified planetary rotation motion without shadow-
the center due to nonuniform thickness profile, ing mask have been employed to achieve high degree
which possibly results in wavefront distortion and il- of thickness uniformity with thermal evaporation
lumination nonuniformity in the optical systems. In (TE) and ion beam sputtering coating machines,
respectively [12,13]. Thickness uniformity of 99.2%
was achieved by modifying the manner of substrate
1559-128X/13/040B26-07$15.00/0 rotation. For large-diameter optics with (strongly)
© 2013 Optical Society of America curved surfaces, the traditional planetary rotation

B26 APPLIED OPTICS / Vol. 52, No. 4 / 1 February 2013

scheme plus the shadowing mask technique is still
the preferred solution to obtain required thickness
uniformity of optical coatings.
On the other hand, the successful applications of
the shadowing mask technology to achieve the re-
quired thickness uniformity of optical coatings rely
on the precise design of the shadowing mask. Cur-
rently, the model for the design of a fixed mask, no
matter in a simple or planetary rotation system, has
achieved great success [7,8,10]. However, the model
for the design of the moving mask is scare in litera-
ture. Up to now, Abzalova et al. have developed an
analytical model for the design of a moving mask
in a simple rotation system [14]. Kelkar et al. re-
ported the experimental results of applying a moving
mask to optimize the thickness uniformity of optical
coatings on convex or concave substrates in a plane-
tary rotation stage [9]. It was very tedious to obtain
thickness uniformity better than 95%, depending on
the geometric configurations of the coating machine
and the substrate to be coated. Fig. 1. General geometric configuration of a planetary rotation
In this paper, an accurate model is developed to de- system with a conical substrate.
sign the shape of the counterrotating shadowing
mask in a planetary rotation system, with the help With vector operation, Eq. (1) can be rewritten as
of a numerical optimization algorithm. Experimen-
tally, the theoretically designed shadowing mask is ur; r1 vj r; r1 
tr; r1 
A

: (2)
employed to control the thickness uniformities of
r − r1
j3
a single layer magnesium fluoride (MgF2 ) film as well




as an antireflective (AR) coating deposited onto a con- By definition, s
∇S∕
∇S
and p
∇P∕
∇P

vex conical substrate carrier with diameter 225 mm, are unitary vectors of the source normal and the
apex angle 140 deg, and height 41 mm, respectively. substrate normal; vr; r1 
s · r1 − r and ur; r1 

Thickness uniformity of 99.3% is achieved. p · r − r1  represent the source function and the sub-
strate function, respectively. In the coating process,
the physical thickness of the thin film deposited on
2. Theory the point of the substrate surface is a summation
According to Knudsen’s laws [6,7], the elementary of elementary thickness amounts as the motion of
thickness of a growing film at a given position of substrate through the vacuum chamber. Now our
the substrate surface can be expressed as problem includes determination of the functions
vr; r1 , ur; r1 , and the trajectory of the point in
cos φ cos j θ
t
A ; (1) the vacuum chamber, given a geometrical configura-
r22 tion. The common geometrical configurations of the
where φ is the angle between the substrate normal evaporation source were already well documented by
and the vector r2
r − r1 connecting a point on the Villa and Pompa [6]. Here, the source is approxi-
evaporation source with a point on the substrate sur- mated to a point source, which is placed at position
face, as shown in Fig. 1, namely the deposition angle x; y; z. For a point x1 ; y1 ; z1  on the substrate, the
of the evaporated material at the point on the sub- source function vr; r1  is described as
strate surface. θ is the angle between the source nor-
vr; r1 
z1 − z: (3)
mal and the vector r2. A is a constant. j is an emission
characteristic parameter of the evaporator, which On the other hand, for the geometrical configura-
can be determined by comparing the calculated tions of the planetary rotation system and a conical
and experimental thickness distributions when no substrate as shown in Fig. 1, the motion of the point
shadowing mask is applied. r2 is the distance be- x1 ; y1 ; z1  on the cone with planetary rotation and in-
tween the point on the evaporator and the point clination will form the path, which can be expressed
on the substrate surface. by the following equations:

( x
R cos α  l cos β sin γ cos α cosn  1α  l sin γ sin α sinn  1α  sin β cos αh − l cos γ
1
y1
R sin α  l cos β sin γ sin α cosn  1α − l sin γ cos α sinn  1α  sin β sin αh − l cos γ ; 4
z1
h1  l sin β sin γ cosn  1α − cos βh − l cos γ

1 February 2013 / Vol. 52, No. 4 / APPLIED OPTICS B27

 
with limit 1; φ ≤ 90°
Bl; α
: (10)
0; φ > 90°
l ∈ 0; h∕ cos γ; (5)
Finally, the model for the design of the shadowing
where R is the radius of the planetary orbit, α is the mask can be described as
angular position of the planet in its orbit, β is the an-
gle of inclination of the planetary holder, γ is the Z
2πC
semivertical angle of the conical substrate, h1 is tl; αBl; αMl; αdα
t0 ; (11)
the height of the center of planetary holder along
0
z-axis direction, h is the height of the cone, and l
is the distance between the coated position and
where t0 is a constant thickness. Ml; α is a logical
the conical vertex along the radical direction of the
function, which describes the relation between the
cone surface. n is the gear ratio between the plane- position of the cone and that of the shadowing mask.
tary and solar rotations, which determines the de- Theoretically, it is difficult to give an explicit expres-
gree of random motion of the substrate and is sion for the logical function. However, the design of
crucial to achieve a high degree of thickness unifor- the shadowing mask can be optimized by a numerical
mity [8]. optimization algorithm with the following iteration
Then, the substrate function ur; r1  and the dis- steps: (1) an initial shape of the shadowing mask
tance jr − r1 j are described as, respectively, is introduced and then the thickness distribution
of the optical coating on the substrate is calculated.
ur; r1 
fcos β cos γ cos α cosn  1α In the thickness calculation the effect of the shadow-
 cos γ sin α sinn  1α ing mask on the thickness distribution is taken into
consideration as follows: when the projection of the
 sin β sin γ cos αgx − x1  straight line connecting the point of the substrate
and the point source at the mask plane is outside
 fcos β cos γ sin α cosn  1α
the shadowing mask, the evaporated molecules accu-
− cos γ cos α sinn  1α mulate at the point of the substrate and the corre-
sponding thickness increases. On the other hand,
 sin β sin γ sin αgy − y1  when the projection at the mask plane is blocked
 fsin β cos γ cosn  1α − cos β sin γg by the shadowing mask, no evaporated molecules ac-
cumulate at the point of the substrate. (2) The shape
× z − z1 ; (6) of the shadowing mask is optimized with the calcu-
lated thickness distribution in step (1) by a numeri-




r − r1

x − x1 2  y − y1 2  z − z1 2 1∕2 : (7) cal optimization algorithm. Step (2) is repeated until
the best possible thickness uniformity is achieved.
Afterward, the deposition thickness at a specific 3. Experiment
radical distance l in Eq. (2) is obtained by integration
To design the shadowing mask, the emission charac-
of the angular position of the planet α. That is,
teristic parameter of the evaporation source has to be
determined first. As previously mentioned, the emis-
Z
2πC
sion characteristic parameter can be determined by
tl
tl; αdα; (8) fitting the calculated thickness distribution to the
0
experimental result without the shadowing mask.
A conical substrate holder is prepared to measure
with C the circle number of planetary revolution. the thickness distribution of the optical coating along
On the other hand, the microstructural and optical the radical direction of the cone surface. The struc-
properties of the deposited thin film depend on the tural parameters of the conical carrier are as follows:
deposition angle of arrival molecules with respect diameter 225 mm, apex angle 140 deg, and height
to the substrate normal [15–17], as shown in Fig. 1. 41 mm. Two BK7 slides (size: 27 mm × 55 mm ×
Practically, when the deposition angle is larger than 0.5 mm) and three fused-silica substrates [size:
90 deg, there is no arrival molecule accumulated at Φ25.4 mm × 4 mm, with a root-mean-square surface
the point of the substrate. Taking into consideration roughness of 0.5 nm by an atomic force microscope
this point, Eq. (8) can be modified as (AFM)] are fixed onto the conical carrier and used
to measure the thickness uniformity and the spec-
trum performance of the deposited thin films, respec-
Z
2πC
tively. Figure 2 shows the conical substrate holder
tl
tl; αBl; αdα: (9) with slides and substrates assembled.
0 Experimentally, single-layer MgF2 thin films and
AR coatings at 193 nm are fabricated by TE in a
Here, Bl; α is a correcting function of deposition commercial coating plant (SYRUS pro 1110 DUV,
angle, defined as Leybold Optics) equipped with a planetary rotation

B28 APPLIED OPTICS / Vol. 52, No. 4 / 1 February 2013

 corresponding best fit. From the best fit the physical
thickness of the MgF2 film is determined to be
281.74 nm. To determine the thickness uniformity,
the thicknesses of the MgF2 films at 11 different po-
sitions evenly distributed (with 10 mm spacing)
along the radical direction of the cone are evaluated
and compared.
Fig. 2. (Color online) Assembly of the conical substrate holder To analyze the spectrum uniformity of the depos-
and substrates for thickness uniformity experiment. ited AR coating, the reflectance spectra of the two-
side AR-coated fused silica substrates are measured
by a high-precision deep/vacuum ultraviolet (DUV/
system, two e-beam guns, four TEs, three substrate VUV) spectrophotometer (VUVaS 2000, McPherson)
heaters, and an advanced plasma source. in the spectral range of 170–240 nm. The spectrum
The geometric parameters of the coating machine measurements are performed under vacuum envir-
are R
300 mm, β
0°, h1
750 mm, and onment with a pressure p < 1.0 × 10−2 Pa and at
n
131∕19, and the evaporation source is placed an angle of incidence of 10 deg.
at x
305 mm, y
0 mm, and z
0 mm. In the
coating process, the deposition vacuum pressure is
higher than 3.0 × 10−4 Pa by a cryopump set. The 4. Results and Discussion
conical substrate is placed at the center of the plane- To determine the emission characteristic parameter
tary holder. The substrate temperature is 250 °C, of the evaporation source, the influence of the emis-
heated with ceramic heaters. The deposition rate sion characteristic parameter on the thickness distri-
and the thickness of single-layer MgF2 film con- bution of the film deposited on the conical substrate
trolled by a quartz monitor are 0.2 nm∕s and 300 nm, is calculated with the model developed in Section 2,
respectively. The AR coating is designed with a and the results are presented in Fig. 4(a). As shown
seven-layer structure having non-quarter-wave in the figure, the thickness uniformity of the film on
optical thickness, and optimized by a commercial the conical substrate improves with the increasing
FilmWizard software to maximize the transmittance emission characteristic parameter. On the other
at 193 nm for the angle of incidence ranging from 0 to hand, the film thickness is approximately a linear
45 deg with a target transmittance of 100% [18]. The function of the deposited position along the radical
resulting design of the AR coating is sub/33.1Lp direction of the conical substrate. The slope of the
13.03Hp 20.31Lp 19.43Hp 41.22Lp 29.91Hp linear dependence is a monotonic function of the
36.27Lp/air. The deposition rates are set to 0.2 nm∕s emission characteristic parameter. The calculated
and 0.04 nm∕s for MgF2 and LaF3 layers in AR coat- dependence of the slope on the emission characteris-
ing preparation. As the starting material, MgF2 and tic parameter and the corresponding polynomial fit
LaF3 (Merck) are used and deposited by electron- are shown in Fig. 4(b). The emission characteristic
beam evaporation and resistive TE, respectively. parameter of the evaporation source can be deter-
BK7 and fused-silica substrates are manually mined from the experimentally measured slope. In
cleaned with alcohol and acetone, and then irra- this experiment, the emission characteristic param-
diated by a commercial UV photo cleaner for 40 min eter of the evaporation source is determined to be
before deposition to remove hydro-carbon contami- 1.78  0.02.
nations at the substrate surfaces [19,20].
For the prepared coating samples, a spectrophot-
ometer (Lambda 1050, Perkin Elmer) is used to de-
termine the physical thickness of the single-layer
MgF2 films deposited on the BK7 slides. The refrac-
tive index of BK7 substrate versus wavelength is
evaluated from the transmission spectrum of un-
coated slides. Then the transmission spectra of the
single-layer MgF2 films in 400–800 nm wavelength
range are used to determine the thicknesses of the
MgF2 films deposited by electron-beam evaporation.
Here, the MgF2 films are taken as nonabsorbing
material, as the extinction coefficient of the MgF2
film is lower than 1.0 × 10−4 in the measured wave-
length range and has negligible impact on the trans-
mission. The refractive index versus wavelength is
fitted using the Cauchy law. Afterward, the physical
thicknesses of the MgF2 films are determined from Fig. 3. (Color online) Experimental (dashed curve) and fitted
the measured spectra. As an example, Fig. 3 shows (solid curve) transmittance spectra of a single-layer MgF2 film
one of the measured transmission spectra and the deposited on a BK7 substrate.

1 February 2013 / Vol. 52, No. 4 / APPLIED OPTICS B29

 thickness distribution without the shadowing mask
correction has been used to determine the emission
characteristic parameter of the evaporation source
as before. The experimental thickness uniformity
of single-layer MgF2 film with the shadowing mask
correction is better than 99.3%, in agreement with
the theoretical value, 99.5%. The small difference
between the experimental thickness uniformity
and the theoretical result might be caused by several
factors, such as the measurement errors of the geo-
metric parameters of the coating machine, the posi-
tioning and rotation ratio errors of the shadowing
mask, and, last but not least, the stability of the coat-
ing process. For instance, during the coating process
some craters are formed in the evaporated coating
material by the electron-beam evaporator, and

Fig. 4. (Color online) Dependence of the uncorrected thickness

distribution on the emission characteristic parameter j. (a) Calcu-
lated (lines) and experimental (circles) thickness distributions
along the radical direction l. (b) Calculated and experimental
slopes (squares and circle), and corresponding polynomial fit (solid
curve).

After the emission characteristic parameter of the

evaporation source is determined, the shape and size
of the shadowing mask is optimized by a simulated
annealing algorithm [21] for the specific geometric
configuration of the coating machine used in the ex-
periment to achieve the optimum uniform thickness
distributions on the conical substrate. In the ex-
periment, the shadowing mask counterrotates with
respect to the solar rotation of the planet, and the
rotation velocity of the mask equals that of the solar
rotation. The calculated thickness distributions on
the whole conical substrate without and with the
shadowing mask are shown in Fig. 5. With the sha-
dowing mask correction, the calculated thickness
uniformity is improved from 88.3% to 99.5%.
Obviously, excellent thickness uniformity is achieved
with the help of the shadowing mask.
Figure 6 shows the calculated and experimental
thickness distributions along the radial direction Fig. 5. (Color online) Calculated thickness distributions on
of the conical substrate holder without and with the conical substrate without (a) and with (b) shadowing mask
the shadowing mask correction. The experimental correction.

B30 APPLIED OPTICS / Vol. 52, No. 4 / 1 February 2013

 algorithm to design the shadowing mask. A shadow-
ing mask has been designed with the developed
thickness model and applied to correct the thickness
uniformities of single-layer MgF2 film and seven-
layer AR coating at 193 nm prepared on a conical
substrate holder with diameter 225 mm, apex angle
140 deg, and height 41 mm. For the single-layer
MgF2 film, thickness uniformity of better than
99.3% has been experimentally achieved, in agree-
ment with the theoretical result, 99.5%. While for
the AR coating deposited on fused silica substrates,
excellent reflectance spectrum uniformity has
been obtained. The model and method presented
in this paper could find applications in the design
of a shadowing mask for correcting the thickness
uniformity of optical coatings deposited on various
Fig. 6. (Color online) Calculated and experimental thickness
distributions of single-layer MgF2 film along the radial direc-
substrates.
tion of the conical substrate without and with shadowing mask The authors thank Yundong Zhang and Dawei Lin
correction. for their fruitful discussions.
References and Notes
1. J. A. Monsoriu, W. D. Furlan, P. Andrés, and J. Lancis, “Fractal
conical lenses,” Opt. Express 14, 9077–9082 (2006).
2. Z. Ding, H. Ren, Y. Zhao, J. Nelson, and Z. Chen, “High-
resolution optical coherence tomography over a large
depth range with an axicon lens,” Opt. Lett. 27, 243–245
(2002).
3. L. Sara, I. Umberto, and C. Robert, “Analysis of the combined
impact of the laser spectrum, illuminator miscalibrations, and
lens aberrations on the 90 nm technology node imaging with
off axis illumination,” Proc. SPIE 6154, 615434 (2006).
4. I. Huang, L. Lin, and C. L. Lin, “Novel illumination apertures
for resolution-enhanced technology and through-pitch critical
dimension control,” Proc. SPIE 5754, 1395–1404 (2005).
5. A. Bouwers, “Improvement of resolving power of optical
systems by a new optical element,” Appl. Sci. Res. 3, 147–148
(1954).
6. F. Villa and O. Pompa, “Emission pattern of a real vapor
sources in high vacuum: an overview,” Appl. Opt. 38, 695–703
(1999).
Fig. 7. (Color online) Measured reflectance spectra of the 193 nm 7. F. Villa, A. Martínez, and L. E. Regalado, “Correction masks
for thickness uniformity in large-area thin films,” Appl. Opt.
AR coating on fused silica substrates deposited at different
39, 1602–1610 (2000).
locations. 8. J. B. Oliver and D. Talbot, “Optimization of deposition unifor-
mity for large-aperture national ignition facility substrates in
a planetary rotation system,” Appl. Opt. 45, 3097–3105
these craters will therefore influence the emission (2006).
characteristic parameter [14]. 9. P. Kelkar, B. Tirri, R. Wilklow, and D. Peterson, “Deposition
Figure 7 exhibits the measured reflectance spectra and characterization of challenging DUV coatings,” Proc.
SPIE 7067, 706708 (2008).
of the AR coatings formed at three different locations 10. B. Sassolas, R. Flaminio, J. Franc, C. Michel, J. L. Montorio, N.
along the radical direction of the conical substrate Morgado, and L. Pinard, “Masking technique for coating
carrier comprehensively described as above. With thickness control on large and strongly curved aspherical
the counterrotating shadowing mask correction, the optics,” Appl. Opt. 48, 3760–3765 (2009).
reflectance spectra at different locations match each 11. J. B. Oliver, P. Kupinski, A. L. Rigatti, A. W. Schmid, J. C.
Lambropoulos, S. Papernov, and A. Kozlov, “Large-aperture
other very closely. The obtained uniformity of the re- plasma-assisted deposition of inertial confinement fusion
flectance spectrum further suggests that the shadow- laser coatings,” Appl. Opt. 50, C19–C26 (2011).
ing mask designed with the developed thickness 12. M. Gross, S. Dligatch, and A. Chtanov, “Optimization of
model and numerical optimization algorithm works coating uniformity in an ion beam sputtering system using
a modified planetary rotation method,” Appl. Opt. 50,
well for the thickness uniformity correction. C316–C320 (2011).
13. F. L. Wang, R. Crocker, and R. Faber, “Large-area uniformity
5. Conclusions in evaporation coating through a new form of substrate mo-
We have developed a thickness model for a coating tion,” in Optical Interference Coatings, OSA Technical Digest
system with a planetary rotation stage and counter- (Optical Society of America, 2010), pp. 1–3.
14. G. I. Abzalova, R. S. Sabirov, and A. V. MikhaÏlov, “Depositing
rotating shadowing mask to control the thickness uniform-thickness coatings on large surfaces by means of
distribution of optical coatings deposited on conical electron-beam evaporation in vacuum,” J. Opt. Technol. 72,
substrates, and employed a numerical optimization 799–801 (2005).

1 February 2013 / Vol. 52, No. 4 / APPLIED OPTICS B31

15. C. Zaczek, S. Müllender, H. Enkisch, and F. Bijkerk, “Coatings 19. J. Wang, R. Maier, P. G. Dewa, H. Schreiber, R. A. Bellman,
for next generation lithography,” Proc. SPIE 7101, 71010X and D. D. Elli, “Nanoporous structure of a GdF3 thin film
(2008). evaluated by variable angle spectroscopic ellipsometry,” Appl.
16. Y. J. Jen and C. F. Lin, “Anisotropic optical thin films finely Opt. 46, 3221–3226 (2007).
sculptured by substrate sweep technology,” Opt. Express 20. B. C. Li, D. W. Lin, Y. L. Han, C. Guo, Y. D. Zhang, and H. X.
16, 5372–5377 (2008). Liu, “Anti-reflective fluoride coatings for widely tunable deep-
17. C. C. Jaing, M. C. Liu, C. C. Lee, W. H. Cho, W. T. Shen, ultraviolet diode-pumped solid-state laser applications,”
C. J. Tang, and B. H. Liao, “Residual stress in obliquely Chin. Phys. Lett. 27, 044201 (2010).
deposited MgF2 thin films,” Appl. Opt. 47, C266–C270 21. B. Suman and P. Kumar, “A survey of simulated annealing as
(2008). a tool for single and multiobjective optimization,” J. Oper. Res.
18. FilmWizard, optical coating software. Soc. Am. 57, 1143–1160 (2006).

B32 APPLIED OPTICS / Vol. 52, No. 4 / 1 February 2013

View publication stats