Research in fields such as nanotechnology, biosensors, thin-film displays and high-speed optical communications has increased the complexity of thin-film structures. Although an antireflection coating consisting of a single layer of magnesium fluoride may have been sufficient 20 years ago, current designs may call for a 24-layer stack of films with alternating refractive indices. With high-speed optical communications, this stack increases tenfold, leading to filters comprising up to 256 layers. Monitoring the thickness of these coatings has become increasingly important and, although quartz crystals have been used for this purpose since the 1960s, many variables can degrade the process. New developments, however, have addressed this issue. From the introduction of quartz crystal thin-film sensors that operate at elevated temperatures to the refinement of existing quartz designs to account for the stresses and failure mechanisms caused by optical materials, this critical process-control element is catching up with the requirements for making the next generation of optics. Current practice Quartz crystal monitors may be the most misunderstood components of optical thin-film deposition systems. They provide coating rate and thickness data in real time, with angstrom resolution. When set up and calibrated correctly, they automatically control deposition sources, ensure repeatable and accurate thin-film coatings, and control optical film properties dependent on deposition rate. When used incorrectly, crystals drive process engineers to wits’ end with erroneous measurements and random failures. Quartz sensors measure film thickness by monitoring a change in the frequency of vibration of a test crystal coated simultaneously with process substrates (Figure 1). These substrates are positioned close to the crystal, ensuring that the amount of evaporant falling on both components is identical. If this is not the case, a geometrical correction, called the tooling factor, is applied. Figure 1. In a typical quartz crystal monitor setup, the crystal is contained in a water-cooled housing, mounted in a line-of-sight position relative to the coating source. Coupled to an electrical circuit, the crystal vibrates at its natural or resonant frequency, which for most commercial instruments is between 5 and 6 MHz. A microprocessor-based control unit monitors and displays this frequency, or derived quantities, continuously. As material coats the crystal during deposition, the resonant frequency decreases in a predictable fashion, proportional to the rate at which material arrives at the crystal and proportional to the material density. The frequency change is calculated several times per second, converted in the microprocessor to angstroms per second and displayed as deposition rate. The accumulated coating is displayed as total thickness. The sensitivity of the sensor is remarkable. A uniform coating of as little as 10 Å of aluminum will cause a frequency change of 20 Hz, easily measured by today’s electronics. As the density of the film increases, the frequency shift per angstrom increases. The useful life of quartz depends on the thickness and type of coating monitored. If a low-stress metal such as aluminum is deposited, layers as thick as 1 million Å have been measured. At the other extreme, high-stress dielectric films will reach less than 2000 Å before the crystal malfunctions. Shortcomings, limitations In the early days of crystal monitors, metallic films of copper, silver and gold were the most common materials deposited. These elements produced low-stress coatings and were condensed on substrates held at or near room temperature. Under these conditions, accurate determinations of film thickness and deposition rate were achievable. When the optics industry discovered crystal monitors, attention shifted from opaque metals to transparent materials such as magnesium fluoride and silicon dioxide, which, unfortunately, produced films with high intrinsic stresses and required high process or substrate temperatures. These were not welcome developments for crystal monitoring. Quartz is highly sensitive to stress and temperature changes. This sensitivity can be traced to a fundamental property of quartz called piezoelectricity. If a bar of quartz bends, it will develop a voltage on opposite faces. Conversely, if a voltage is applied, the bar will bend. When alternating voltage is applied, the bar will vibrate or oscillate in phase with the voltage. At a specific frequency of oscillation, quartz will vibrate with minimal resistance, much the way a tuning fork rings when struck. This natural resonance frequency is used as the basis for measuring film thickness. When coatings are added to the crystal surface, the resonant frequency decreases linearly. If the coatings are removed, the resonant frequency increases. To complicate matters, a quartz crystal also exhibits a frequency change when deformed by thin-film stresses or mechanical forces from the mounting holder. If process conditions heat or cool the crystal, a similar frequency shift occurs. Regardless of the origin, the shift is indistinguishable from that caused by the addition of a coating. Frequency shifts can be positive or negative, cumulative and random. The causes of resonant frequency changes include: • Vibrations introduced through the mounting hardware. • Variations in the voltage used to oscillate the crystal. • Changes in the film being monitored (acoustic impedance). • Adhesion failure of the monitored coating or quartz electrodes. • Radio-frequency interference in the monitoring circuit. What can go wrong? These effects introduce large errors in thickness and rate. Temperature swings in quartz can result in thickness variations of 50 Å or more. Adhesion failure causes 100-Å rate spikes, and extraneous vibrations can produce changes of about 1000 Å. For precision optical components, these errors result in major yield loss. The harsh conditions present during optical film coating have a deleterious effect on the operating life of the crystal. High-stress coatings can deform the crystal to the point that it ceases oscillation without warning. Splatters of material from the coating source lead to the same failure. High-energy plasmas used for substrate cleaning couple into the crystal electronics and cause severe electrical noise. High-temperature depositions overheat the crystal, driving it past its operating limit. Early crystal failure can be a great inconvenience or an unmitigated disaster. In the case of thin-film stacks of 100 layers or more, venting the chamber to replace crystals is not an option, because of the undesirable effects of atmospheric gases on film chemistry. For the very thick films used in laser-power or infrared optics, short crystal life may prevent completion of the coating. For high-speed roll coating systems, abrupt crystal failure can cause great numbers of ruined substrates. Over the past 30 years, some weaknesses of crystal monitors have been addressed with design changes to crystals, mounting holders and control units. The use of water-cooled holders, or sensor heads, combined with temperature-insensitive quartz (AT-cut), reduced thermally induced frequency shifts for low-temperature processes. Planoconvex circular crystals mounted in spring contact heads minimized extraneous vibrations. Improved electronics and shielding eliminated radio-frequency interference and voltage variations. Corrections to the thickness calculation algorithm have accounted for the acoustic impedance. Multiple crystal sensor heads are used as a solution to premature crystal failure. Nevertheless, issues such as film stress, adhesion failure and extreme temperature effects have not been adequately addressed. These are difficult problems, and there has not been sufficient economic incentive to pursue solutions. In 2003, however, the demands of nanotechnology, thin-film displays and high-speed optical communications are forcing a re-engineering of the quartz crystal monitor. A promising future For optical device fabrication, dealing with thin-film stress and adhesion effects on quartz is an area with the greatest short-term benefit. Until recently, the solutions for erratic crystal behavior have been more art than science. Users have employed special crystal inspection, cleaning and mounting procedures to achieve reproducible runs. More sophisticated users have ascribed benefits to crystal electrode coloration and surface topography, with mixed results. However, the science is catching up. Studies on new quartz types, electrodes and sensor heads have led to promising designs for reliable sensors. One improved design is the development of the aluminum or aluminum alloy electrode crystal. For SiO2 coatings, this electrode can extend the useful life of the sensor by 100 percent or more compared with the industry-standard gold crystal. Furthermore, frequency shifts due to electrode adhesion failure are decreased up to 90 percent under standard laboratory conditions. The benefit of this electrode appears to be material- and deposition-specific, because it is not the same for all coatings. Nonetheless, it may prove to be a test-bed design for thin-film stress studies. A second design change involves alternative orientation angles of quartz for monitor crystals. AT-cuts exhibit little frequency change with increasing temperature, or what is called frequency-temperature dependence, and are used to minimize the resonant frequency shift caused by process heating (Figure 2). The angle of cut can be varied to allow stable operation at higher temperatures as well. Figure 2. The standard AT-cut crystal is derived from cutting wafers from a quartz bar at an angle of 35¼ and 15 minutes from the Z-axis. A more promising crystal is based on the stress-compensated, or doubly rotated, cut of quartz. This orientation exhibits similar frequency-temperature behavior to that of the AT-cut, with the added advantage that it shows essentially no change of frequency when the crystal is stressed. A monitor crystal fabricated from stress-compensated-cut material, in initial coating trials, exhibits none of the frequency changes induced by high-stress dielectrics on AT-cut crystals. Historically, it has been a more expensive version of quartz, but the benefits for the optical process engineer may outweigh the cost penalty. A third area of improvement is a redesign of the crystal sensor head. Although quartz crystals are capable of subnanometer measurement and are sensitive to mass accumulations of as little as a picogram, they are traditionally finger-handled when installed in the deposition system. This is surprising because the crystal is 0.010 in. thick and can be easily cracked prior to mounting. In addition, potential contamination of the crystal surface has disastrous effects on the adhesion of the coating being monitored. Adhesion failure results in frequency shifts opposite those of film coating, introduces thickness errors and degrades the crystal’s lifetime. Recently introduced crystals are premounted in a holder, with the measurement surface protected from human contact. Less obvious but equally rewarding improvements deal with the thermal environment of the crystal. Much as stress effects lead to erratic crystal performance, temperature effects lead to equally significant thickness errors. An examination of the frequency-temperature curve for AT-cut quartz illustrates how variable the effect of temperature is (Figure 3). If one increases the crystal temperature 20° from room temperature, the frequency will shift approximately 20 Hz. For aluminum, this is equivalent to 10 Å of coating. Larger temperature shifts generate accelerating changes because of the steep slope of the frequency-temperature curve. Figure 3. This AT-cut quartz frequency temperature curve shows how temperature variation affects frequency. Recent studies of standard sensor heads show that, even with water-cooling, the crystal temperature can rise 20° to 30° within a 10-minute process. For extended runs with high chamber temperatures, this becomes considerably larger. A new dedicated quartz crystal sensor head temperature monitoring system retrofits existing thin-film quartz crystal controllers. With data in hand, the true thermal environment of the crystal sensor can be measured, eliminating the guesswork and adding more science to the art of optical coatings. A paradigm shift is under way in quartz crystal process monitoring. In many applications, crystals are the keys to success. No matter how big a breakthrough may be — whether in materials, geometry, process design or application — if a thin-film coating of any sophistication is needed, the weak link is how accurately that film can be measured. Meet the author Scott Grimshaw is founder and president of Cold Springs R&D Inc. of Syracuse, N.Y.