Formulae to estimate the average percent reflectance (Rave) of a broadband antireflection (AR) coating as a function of
the bandwidth (B), the overall thickness (C), the index of refraction of the last layer (L), and the difference between the
indices of the high- and low-index layers (D) were reported in 1991. Various refinements of these formulae and other
insights into the underlying behavior of such coating designs have been reported up until the present time.
Dobrowolski, et al.6 and Tikhonravov, et al.7 have also added independent viewpoints to this subject over this period. In
the previous studies, the effects of the index of refraction of the substrate have mostly been ignored and have appeared to
be very minor. This study has investigated the influence of the substrate index on the Rave results. It has been found
that there seem to be two classes of designs with respect to the effect of substrate index. In the class of "step down"
AR designs, there is a significant effect, in the other class, there is no significant effect. Even in the step-down case,
there is no effect of substrate index if any and all indices of refraction for the coating materials are available from that of
the index of the substrate to the index of the media.
Narrow bandpass filters have historically been designs of quarter waves at the passband wavelength, and have been
monitored at the turning points using the passband wavelength. By direct monitoring at the passband wavelength, errors
have been shown to be primarily self compensated, and have allowed much better performance than could otherwise be
expected. The turning points are difficult to detect precisely and accurately because the change in transmittance with
thickness becomes zero at the desired termination point. By proper design with non-quarter-wave layers, essentially the
same spectral performance can be achieved by layer terminations that are far enough from turning points to be
significantly more sensitive termination points. The design approach is to maintain the optical thickness of the reflector
layer pairs at one half-wave of the passband wavelength, but change the ratio of the optical thicknesses of the high and
low index layers. These can be adjusted enough so that the thicker layers contain two turning points and the last turning
point in the layer can be more accurately and precisely determined. The error compensation benefit from the historic
method should be maintained. This leads to potentially improved control during deposition and monitoring of narrow
Mirrors at normal incidence have been successfully made for the 12.4 nm spectral region using molybdenum (Mo), silicon (Si), and other materials. At these wavelengths, a typical layer is of the order of 3 nm in physical thickness, and the atomic diameters of the materials are of the order of 0.3 nm. The implication is that the layers are of the order of 10 atomic layers thick. If the deposition of such films were done by atomic layer epitaxy (ALE) or other atomic layer deposition (ALD) techniques, the spectral results would fall in discrete patterns that could limit the potential design choices. The problem would be even more severe at shorter wavelengths. This work reports on the study of some of those possibilities and limitations.
A decade ago we introduced empirically derived formulas which allowed one to estimate the performance which can be achieved in the design of broad-band antireflection (AR) coatings based upon the indices of refraction of the materials to used, the bandwidth required, and the overall thickness of the coating. This has proved to be a useful tool to avoid attempting impossible designs and to guide the designer toward an optical result. It can also be helpful to non-designers who need to know what can and cannot be done before specifying a system AR coating requirement. In the new work reported here, forumlas with additional accuracy have been developed by further data generation and the application of modern statistical analysis tools. The overall thickness parameter used in the equations has also been better defined and understood, and the tendency of overall thicknesses to have quantization has been studied further. These findings are discussed in conventional thin film design terms and also from the Fourier synthesis/analysis viewpoint.
The final corrections which might be made in the last two antireflection (AR) layers in the deposition of narrow bandpass filter designs such as might be used for Dense Wavelength Division Multiplexing (DWDM) in the fiber optics communications field were discussed in a previous report. A broader range of techniques and simulations of those final layer adjustments are described here, how they can be done, and the benefits which might be obtained. A surprisingly simple new technique is given which should yield improved results.
A constrained optimization design procedure is described which gives good control of the spectral position of a filter edge and simultaneously optimizes pass band transmittance. The underlying principle of our previous study was to use constant level monitoring and position that constant layer termination photometric level for the least sensitivity to photometric and other layer termination errors. Additional layers are needed between the substrate and the periodic stack to bring the layer termination to the ideal level and also provide the antireflection coating of the pass band of interest. The requirement can be stated as: the preliminary layers must move the reflectance phase at the monitoring wavelength from that of the substrate to that of the start of the periodic stack needed for CLM. The constraints are that the magnitude and phase values at the end of the deposition of the preliminary layers satisfy these specific requirements. The constrained optimization can vary both the pre- and post-periodic matching layers while attempting to meet the transmittance targets and simultaneously satisfy the constraints. When this is done, the resulting design has optimized transmittance and satisfies the required constant level optical monitoring conditions for the most reproducible result in production.
An improved plasma source for use with oxygen and other gasses has shown great stability at ion power levels almost an order of magnitude above those of the other commercially available gridless sources. The original motivation of this work was to obtain a powerful a source of oxygen ions with neutralizing electrons (plasma source) for the reactive deposition of SiO2 from SiO. The results are also applicable to other gasses such as nitrogen, argon, etc., and other deposition materials. The characterization of the source in detail by the use of the design of experiments methodology is discussed.
It has been shown empirically that the lack of very low index of refraction materials is the major limitation to achieving a very low reflectance coating over a very broad spectral band. It has been shown in the same work that additional coating thickness can be employed to make up for this deficiency to a certain extent. Thicknesses which are an order of magnitude thicker than the minimum necessary for a reasonable very broad band antireflection (AR) coating can reduce the reflection to about one half that of the minimum thickness case. This result is empirically predictable to a satisfactory degree, but the underlying reasons for this have not been clear. This paper explores the principles which contribute to the understanding of this effect of additional thickness by empirical and comparative means. The Fourier viewpoint adds to the understanding of variations from the ideal design. We have previously shown that the ideal AR coating, when any and all indices are available, would be a smooth inhomogeneous `step-down' in index from the substrate to the medium. The form of the index profile is approximately a Gaussian decay from the substrate to the medium or similarly described by Southwell as a `quintic function'. When the medium is a vacuum or air and the lowest available index is represented by a real material such as MgF2 at index 1.38 rather than values very close to that of the medium, the discontinuity from the smooth step-down profile causes a reflection residual that cannot be overcome by adjustments in the rest of the smooth profile. Additional thickness and the appropriate index profile can be used to reduce, but no eliminate entirely, this residual reflection. We discuss our observations and findings on these effects in more detail in the body of this paper.
Silicon dioxide has many desirable properties as a coating material for the visible and near IR spectrum. These include low index, high transmittance, good adhesion, and environmental durability. The deposition of the material from solid discs or granular forms using an electron beam gun tends to have a fluctuating angular distribution. This causes reproducibility problems in some critical optical coating processes. Silicon monoxide can be evaporated from a resistance source with a highly reproducible deposition pattern, but it does not have the low index and high transmittance required. We report on attempts to develop processes which convert the SiO to SiO2 during deposition by the use of ion assisted deposition. The additional oxygen must be supplied in a sufficiently energetic process to provide the material conversion during or after the SiO deposition on the surface to be coated. The deposition rate (10 angstrom per second), the uniformity, and the repeatability of the processes must also be adequate for the production of the product at an economical rate. These experiments have been done with both an End-Hall and a Cold Cathode ion source. A comparison of the results and behavior observed with each type of ion source are reported along with recommended process parameters to be used with each.
Prior works by the author and others point to the observation that the ideal antireflection coatings would be a family of inhomogeneous index of refraction versus thickness profiles. Our previous and recent studies have been from the design perspective and of an empirical nature. The pursuit of the understanding of the nature and properties of these profiles has employed optimization tools, various modes of display of the properties, and Fourier techniques. The multiple views of the nature serve to better define and clarify the properties of these profiles. The basic nature and fundamentals of antireflection coatings are described from these several viewpoints. The possibilities and limitations of the inhomogeneous profiles are discussed and compared with their approximation by homogeneous layer systems.
Previously published works on the use of Fourier transforms to synthesize thin film designs have mentioned two problems. One has been a lack of a relational function of reflectance to the index profile, or `Q-function,' which gives satisfactory results for both high and low reflectance cases. The other is that the reflectance versus frequency profiles of well known results, from the matrix approach, showed a significant distortion for high reflectors. Our investigations have concluded that the transform of the simple reflectance amplitude versus optical thickness gives correct results in reflectance versus frequency for all cases when multiple reflections are properly taken into account. The limitation of using only non- dispersive and non-absorbing media still applies to this work. The challenging antireflection coating problem posed for this meeting (a 400 to 900 nm bandwidth and less than 1% reflectance from 0 to 30 degrees) is used as an example to see what insight may be gained by the use of Fourier and related viewpoints.
Front surface metal mirrors need protection of the inherently fragile metal film deposited on a glass substrate. Conventional evaporated dielectric thin-film overcoats provide limited protection because of their less than dense packing. These films usually have a columnar structure with voids between the columns. The voids give access to the underlying metal film for humidity and corrosive gases or liquids. Some progress in developing better coatings was made in the early 1980s with ion-assisted deposition. Front surface aluminum mirrors with dielectric thin films deposited by reactive low-voltage ion plating have a still higher survival time, by a factor of up to 3 under comparable test conditions. The transmission of our best samples increased to only 10% when immersed in 0.2 M NaOH for 20 h. By comparison, an unprotected aluminum film dissolves in less than 5 mm. Electron beam evaporated dielectric coatings provide protection for about 1 .5 to 2 h in the same test solution. The reason for the significant improvement brought about by reactive low-voltage ion plating deposition, and its advantage for large-scale production is discussed.
The focus of this paper is concerned with the practical aspects of designing optical instruments which are intended for production in large or even small quantities. Certain aspects of performance are required of an instrument which is its reason to exist. It is usually expedient, plus fiscally and ecologically sound, to strive for an overall efficient use of resources in the life cycle of the instrument (i.e., keep the net cost down). This includes the processes from concept through design, prototypes, production, use, and disposal. The design and tolerancing aspects of the process have a major effect on the life cycle cost and efficiency of the system and that is the principal subject of this paper.
We discuss what makes up the cost of a lens and the effects of tolerances and other factors on that cost. This results in a new lens cost estimation formula. We describe the interactions of lenses and lens cells from the tolerance viewpoint. We then explain the principles whereby the system tolerances can be determined which will give the minimum cost system which meets the performance requirements. We conclude with an example from real life of the preliminary application of the principles.
We show that when designers and fabricators understand each other''s art there are ways to combine their techniques to achieve the best results with the minimum difficulty. We share some problems that we have encountered and sometimes caused ourselves in hopes of helping the reader avoid the same pitfalls. 1 .
The NRCC Fourier transform synthesis method has been applied for the first
time to an optical thin film design problem in which the external media are different.
Graded index AR coatings for germanium substrates have been synthesized and
compared to multilayer solutions. No limits have been imposed on the refractive
indices. Although such systems cannot be put into practice, they provide basic
insight into the general nature of broadband AR coatings. Admittance diagrams have
been calculated for some of the systems to gain further understanding of their
SC209: Practical Design and Production of Optical Thin Films
Advanced optical thin films are being used in optical systems, communications, and light control applications. The industry is advancing rapidly to meet demands for performance and production capacity. This short course describes optical thin films, components and processes. New viewpoints, equipment, and processes are available to support advanced capability and efficiency.