62 CHINESE OPTICS LETTERS / Vol. 8, Supplement / April 30, 2010
Adaptive manufacturing of high-precision optics based on
virtual deposition and hybrid process control techniques
H. Ehlers
∗
, S. Schlichting, C. Schmitz, and D. Ristau
Laser Zentrum Hannover, Hollerithallee 8, 30419 Hannover, Germany
∗
E-mail: h.ehlers@lzh.de
Received Octob er 30, 2009
The challenge in rapid pro duction of high-precision optical coatings is the need to realize a variety of com-
plex coating designs in one process environment. Two approaches to enhance a stable deposition process
are presented. First, a virtual deposition system is applied for a pre-selection of coating designs that result
in increased process stability using optical broadband monitoring strategies. Second, optical broadband
monitoring is combined with additional quartz crystal sensors to realize a hybrid process control for im-
proving layer thickness accuracy. Finally, a successful combination of both approaches is demonstrated by
comparative studies on virtual and real dep osition processes.
OCIS co des: 310.0310, 310.1860.
doi: 10.3788/COL201008S1.0062.
Under routine production conditions, iterative optimiza-
tion cycles are often needed if varying applications de-
mand very different highly complex coating designs.
This applies particularly if the process control requires
a control strategy development on an individual basis.
In contrast, the adaptive manufacturing concept should
enable a linear production chain without additional con-
sumption of resources by test runs. On one hand, pre-
cision and yield can be increased by choosing multilayer
designs with a higher chance of success. This decision
making can be supported by sophisticated simulation
software. Therefore, an effective virtual deposition sys-
tem will be presented. On the other hand, an enhanced
process control system for layer thickness determination
is essential to reduce the waste. This second approach
involves a hybrid combination of optical and non-optical
process control without the need for individual (design-
dependent) control strategies. In the present contribu-
tion, an optical broadband monitoring (BBM) system,
which evaluates in situ taken transmittance spectra, is
the key component in both approaches. The BBM sys-
tem allows for a fully automated process control based
on absolute transmittance values measured directly on
the moving substrates, as well as an online computation
of these data for a precise thickness determination.
This letter is organized as follows. Firstly, the con-
cepts of the virtual deposition system and the hybrid
process control will be outlined. The hybrid concept is
partly based on the alternating use of BBM and quartz
crystal monitoring for different layers in the stack, but
mainly on a new algorithm merging the two measure-
ments to stabilize the optical monitoring. Subsequently,
the results of real and virtual deposition processes will
be compared to prove the significance of the simulation
results. Finally, an experimental example will document
the positive effect of the combination of BBM and quartz
crystal monitoring.
The core idea of the virtual deposition system is the
use of the original process control software in combi-
nation with a simulation of layer growth and optical
measurement
[1,2]
. In the first step, the deposition simu-
lation is based on the given optical constants and rates for
the employed layer materials. For each simulated mea-
surement cycle, a transmittance spectrum corresponding
to the actual layer thickness is calculated and used as
input for the BBM software. In the second step, the
deviations caused by the main sources of error have to
be considered. Therefore, the optical constants (index
of refraction, extinction coefficient), as well as the depo-
sition rates, are varied by defined error parameter sets.
Furthermore, the simulation of the measured sp ectral
data reproduces the characteristics of the original spec-
trometer setup (noise, wavelength resolution). Besides
random errors, the error parameter sets include system-
atic effects, such as offsets or drifts.
Figure 1 shows the graphical user interface of the vir-
tual deposition system. It is divided into the BBM
interface (left window) and the simulation control win-
dow. On one hand, the BBM displays information in-
cluding transmittance spectra, target thickness, actual
thickness and rate, or status messages. On the other
hand, comprehensive adaptations of the process control
parameters are possible, if required. A more detailed
description of industrial environments in the well-proven
BBM system can be found in Ref. [3]. In addition, the
simulation control window allows for access to all pa-
rameters of the virtual deposition system. At present,
Fig. 1. Graphical user interface of the virtual deposition sys-
tem (left: BBM, right: simulation control).
1671-7694/2010/S10062-05
c
° 2010 Chinese Optics Letters