Optical coatings are critical for enhancing the transmission, reflection, polarization, and phase of light incident upon and passing through optical components. There are many different technologies used to fabricate optical coatings, but in the vapor-phase, they generally fall into two categories: physical or chemical. The most common physical deposition (PVD) methods are thermal evaporation and sputtering. No single coating technology is the best option for all situations, as different applications drive different requirements. This article explores the difference between several common PVD technologies relative to four unique use applications: highly reflective optics, high-laser-damage-threshold applications, cost-sensitive volume manufacturing, and ultrafast laser optics.

Coating Technology Overview

Table 1. Comparison of the key parameters of different coating technologies (E-Beam IAD: ion-assisted electron-beam evaporative deposition, IBS: ion beam sputtering, APS: advanced plasma deposition, and PARMS: plasma assisted reactive magnetron sputtering)

Within PVD, thermal deposition can be further split into standard electron beam (e-beam) and energy-assisted e-beam methods, such as ion assisted deposition (IAD) or advanced plasma source (APS) deposition. Sputtering can similarly be split into ion-beam sputtering (IBS) and magnetron methods, which can include energy assistance with techniques such as plasma assisted reactive magnetron sputtering (PARMS). The fastest rising chemical vapor technique is atomic layer deposition (ALD) and although it has many unique capabilities, it is not covered in this article. Table 1 provides a high-level overview of the differences between four PVD techniques: ion-assisted electron-beam deposition, ion beam sputtering, advanced plasma source assisted electron beam, and magnetron sputtering.

In ion-assisted electron-beam (IAD) evaporative deposition, bombardment from an electron gun vaporizes source materials in a vacuum chamber, which then condenses onto optical components in low-stress, and very low (UV) loss layers of a specified thickness with good density. This is thanks to the coincident ion-bombardment provided by an ion beam gun. E-beam coatings have the highest laser induced damage threshold (LIDT) in the near infrared (NIR) spectrum out of the listed coating technologies, as well as the widest range of useable materials, which allows for the highest flexibility in the coating design space. IAD e-beam evaporative deposition can also accommodate larger coating chamber sizes than the other coating methods and tend to have the capability of producing coatings at the lowest cost. On the down-side, depending on the type of ion source used, IAD can result in slightly lower film densities and coatings with limited smoothness and reflectivity. For the same reason, e-beam evaporative deposition may exhibit less repeatable optical properties, making precise control of layer thickness more challenging compared to ion beam or magnetron sputtering.

Figure 1. IBS is a highly-controllable process which utilizes a high-energy ion gun to sputter material off of a target onto rotating substrates, resulting in very accurate and repeatable optical coatings.

Ion beam sputtering (IBS) produces coatings with very high optical quality and stability. In the IBS process, a high-energy beam of ions is directed onto a target consisting of the desired coating material, causing the material atoms to “sputter” off the target and onto the optical surfaces (Figure 1). This deposition process results in deposition atoms with significant kinetic energy (~10's to 100's eV) packing the atoms into dense, hard, and smooth coatings with incredibly repeatable film properties, even over very long layers or run times. This, combined with high-speed substrate rotation, leads to exceptional layer thickness accuracy, and the capability of reproducing even the most challenging designs, including very high reflectivity and/or ultra-fast chirped mirrors, and very sharp transition filters. However, IBS has some drawbacks, such as higher stress and UV loss, which limit IBS to certain “well-defined” applications. Slower growth rates and uniformity challenges also lead to higher relative cost. Despite these limitations, IBS is uniquely suited to certain applications that the other techniques cannot match.

Advanced plasma sputtering (APS) is an enhanced version of e-beam evaporative deposition with a greater capability toward automated processing. Rather than using a directed ion-beam to assist in densification of the coating as in IAD, APS uses a hot cathode DC glow discharge plasma to generate ions inside the entire chamber of the process. It is ideal for situations requiring the same level of versatility as e-beam IAD, but where more stable optical properties allow for greater stability and film quality approaching that of sputtering. APS can produce dense, hard, and smooth coatings in volume at a similar price structure to e-beam evaporative deposition, but is limited by the fact that it still requires iterative process development (unlike sputtering) and suffers from the higher stress and higher UV losses due in part to the greater energy density of the process.

Plasma assisted reactive magnetron sputtering (PARMS) is another plasma generation-based, physical vapor deposition process in which a glow discharge plasma is generated, but instead of filling the entire chamber, the plasma is “confined” near a target by a magnetic field. Positive ions in the dense plasma near the target are accelerated and strike the surface with enough momentum to cause atoms at the surface to “sputter” off and deposit onto substrates. The confinement of the plasma near the target allows the sputtering process to run at a high efficiency while maintaining a relatively low chamber pressure. Low chamber pressures are required for coating a high volume of optics. In PARMS, reactive gasses are added to improve the stoichiometry of the films. The result is dense, hard films with excellent repeatability, but not as high a repeatability as IBS. It has a high throughput, however, so PARMS provides an often-optimal middle ground between price and performance. For example, PARMS can be an ideal technique to use in applications requiring relatively high optical performance and relatively high volume, such as fluorescence filters.

Use Case 1: Highly Reflective Optics

Many laser systems require reflective optical components with reflectivity ranging from 99.8 - 99.999% for minimizing losses during beam steering in laser systems. Lower reflectivity values may lead to system performance degradation and scattering of high-intensity light, which causes both a loss of throughput and a safety hazard. In extreme cases, catastrophic system failure can occur. Due to its unparalleled capability to produce films of extreme quality and process control, IBS, above all the other coating methods, is uniquely suited for this application. The physics of the deposition process leading to an actual reduction of roughness as a function of layer-count, makes possible surface enhancements capable of specular reflectivity values above 99.99%. IBS coatings combined with super-polished substrates are an ideal match-up for best-in-class reflective optics.

Use Case 2: High Laser Damage Threshold Applications

The laser induced damage threshold (LIDT) of optical components is critical in many laser optics applications to prevent performance degradation and system failure. ISO 21254 defines LIDT as the “highest quantity of laser radiation incident upon the optical component for which the extrapolated probability of damage is zero”, but the exact definition of damage and the methods of detecting and analyzing it are not standardized. This, along with fluctuations in laser parameters and differences between test conditions and real-world conditions, makes LIDT a complicated performance metric.

Figure 2. As this example shows, a) beam with a 1/e2 diameter of 200μm and a peak fluence of 20J/cm2 has a 10J effective diameter of only 115μm, about half of the 1/e2 diameter.

LIDT testing and specification is a constantly evolving field. Currently, Gaussian laser beams are used predominantly for laser damage testing, but they greatly complicate the statistics of the test. Gaussian beams have intensity profiles that are devoid of regions of constant fluence anywhere except very near the peak. This uneven spatial distribution of intensity leads to a gross over-estimation of effective beam diameter. The intensity of a 200μm diameter beam, for example, near the edge is defined as 1/e2, or about 1/10th the peak intensity. It is well known that localized defects in real coatings are primarily responsible for initiating damage, especially at longer wavelengths. At sufficiently low defect densities, the statistics show that the probability of a defect occurring exactly at the peak of a Gaussian is vanishingly small. At fluences lower than the peak, a larger fraction of the beam's diameter may be involved in the initiation of laser induced damage, increasing the probability of damage at fluences lower than the peak (Figures 2 & 3). These issues can be overcome by increasing the number of test sites, increasing the diameter of the beam, and/or substituting the Gaussian for a “flat top” beam.

Figure 3. Plots of damage probability vs. laser fluence that demonstrate the Gaussian effect of lower defect density on LDT uncertainty

E-beam evaporative deposition is able to achieve consistently higher LIDT performance than the other coating methods, presumably because the coating stress tends to be lower, the coating purities can be better maintained, and the “fluence-strength” of the defects can approach that of the intrinsic coating, especially in the low loss wavelengths (i.e. 1064nm). Other methods, such as APS, can also be well suited for laser optical applications due to the density and relatively high optical quality of the films. IBS has also been seen to do well with significant LIDT values in the near UV-VIS range.

Use Case 3: Cost-Sensitive Volume Manufacturing

When manufacturing large quantities of optical components, coating technologies must be able to coat large lot sizes and be cost-competitive. E-beam evaporative deposition and magnetron sputtering benefit from larger coating chamber sizes than the other coating technologies and typically has the lowest relative price, making it ideal for cost-sensitive volume manufacturing. When large volumes of dense hard coatings with slightly more demanding specifications are required, APS can be preferable over e-beam. For the vast majority of volume manufacturing cases, flexibility and speed are the dominant cost drivers, and thus e-beam IAD is usually preferred.

Use Case 4: Ultrafast Laser Optics

Ultrafast lasers with short pulse durations on the order of picoseconds, femtoseconds, or attoseconds are highly advantageous for a variety of applications due to their ability to induce unique non-linear properties and high peak powers. Their short pulse durations are derived from broadband lasers and that allows them to hold superior dimensional tolerances for materials processing and medical laser applications compared to other laser types.

However, optical components used in ultrafast systems have demanding coating requirements to meet stringent tolerances on the phase change in the coatings over a broad band of wavelengths. The rate at which the phase changes in the coating leads to different amounts of group delay dispersion (GDD), which if not managed properly, tends to broaden the temporal profiles of the pulses and degrade the functionality of the system. As a result, some applications require very low phase change (low GDD) variant coatings over a broad wavelength range. For this case, metal dielectric coatings made via e-beam IAD are ideal. On the other hand, some applications require correcting broadened pulses by injecting large, oppositely sloped GDD coatings in the optics. These “chirped” mirror coatings require the realization of exceedingly accurate layer thickness. Here, IBS with its superior layer accuracy capability is the best-suited coating technique.

Key Takeaways

PVD methods such as e-beam IAD, APS, IBS, PARMS all have their own strengths that make them ideal for some specific and overlapping use cases, and no single coating type is the optimal choice for every application. An understanding of the advantages and disadvantages of each available coating technology allows optical designers to choose the best option to meet their individual performance and cost requirements.

This article was written by Chris Cook, Technical Fellow, and Cory Boone, Lead Technical Marketing Engineer, Edmund Optics (Barrington, NJ). For more information, contact Mr. Boone at cboone@edmundoptics or visit here .