Ultrasonics, the sound you cannot hear, has emerged as a valuable tool in achieving the cleanliness required by todayís ever-advancing technology. Disc drives, silicon wafers and chips, medical implants, and all sorts of critical hardware require ultrasonic cleaning to function properlyóor at all. Traditional ultrasonic equipment operating in the 20 to 40kHz range served well for most of 50 years. Prior to the 1980s, ultrasonic technology was essentially ìmatureî with periods of excitement provided only by the occasional advancement in transducer and/or generator technology initiated by the availability of new fabrication materials for transducers and electronic devices; this improved the efficiency and reliability of generators. This period also saw the advancement of the increasingly affordable ultrasonic technology into more and more ìindustrialî applications including plating, surface finishing and metal fabrication.

Starting in the 1980s, however, a new ìwaveî of ultrasonic technology development was initiated as ultrasonic manufacturers began to expand the envelope of available ultrasonic parameters. One cannot be sure if the technology preceded the need or if the technology was prompted by need, but in any event, things started happening. Ultrasonic technology began to move forward and with this motion came a flurry of ìposition jockeyingî by the handful of primary ultrasonic suppliers.

Ultrasonic Generator Technology Sweep Frequency

The first major generator development was sweep frequency. Ultrasonic cleaning had long been troubled by the formation of standing waves which, under the right conditions, could produce parts with ìzebra stripes.î These stripes were created as the reflecting sound wave fell back on itself and reinforced its intensity in horizontal zones located approximately one half wavelength apart. At best, one saw these bands as areas of cleaning vs. non-cleaning on the surface of the part. At worst, parts, especially those fabricated from softer metals including aluminum, brass and copper, were actually etched by high ultrasonic intensity in the areas of reinforcement. A simple solution to this problem was to move parts vertically through the ultrasonic field to spread out the high intensity effect by scanning the part.

It was discovered that varying or ìsweepingî the ultrasonic generator frequency over time effectively broke up the standing waves and reduced the tendency for zebra striping to occur. It was assumed that since the spacing of the stripes was based on frequency, formation of damaging standing waves relied on a fixed ultrasonic frequency. Varying the ultrasonic frequency slightly up and down would cause movement of the bands of high intensity. The effectiveness of sweeping frequency in reducing zebra strips was demonstrated repeatedly when the rate of sweeping and the bandwidth of sweep frequencies were designed to prevent resonance of the liquid column.

Another effect of sweeping went almost unnoticed. Ultrasonic transducers of the time were all relatively ìHigh Qî devices. In simple terms, like a tuning fork, they operated well at their resonant frequency, but as the driving frequency was changed, performance deteriorated rapidly. Sweeping the ultrasonic frequency attempted to drive the transducer at a frequency somewhat off its resonant frequency during most of the sweep cycle. This effort effectively reduced the output power of the transducer without an apparent reduction in the power delivered from the ultrasonic generator. Interestingly, a reduction in ultrasonic output power was also shown to reduce the zebra striping effect.

In defense of sweep frequency, most ultrasonic transducers consist of an array of individual driving elements (also sometimes individually called transducers). Despite the considerable effort given in manufacturing to assure that these individual elements all operate at the same frequency, there is always a slight variance in frequency of the elements in any transducer array even if they are individually chosen and matched prior to bonding to the tank. When a transducer array is operated at a single, fixed frequency, the individual elements (which are wired in parallel), are forced to divide up the available power. Elements with resonant frequencies closest to the frequency provided by the ultrasonic generator will draw the largest portion of power and therefore provide the highest ultrasonic intensity. Meanwhile, elements operating off their resonant frequency will have reduced output. The effect is non-uniformity of the ultrasonic field.

With sweep frequency, each transducer element sees its preferred driving frequency twice during each sweep cycle (provided it is within range of the sweep). The result is that a much more uniform ultrasonic field is produced in the cleaning tank. In addition, the use of sweeping frequency provides more useful cavitation per watt of ultrasonic excitation as a larger bubble population is resonated as the frequency varies. These are undeniable benefits of sweep frequency ultrasonics.

Extensions of sweeping frequency technology include varying the bandwidth and frequency of the frequency sweep itself. In some cases the sweep is randomized to eliminate the potential for damaging resonance effects that may be created as a result of the sweep frequency itself. These enhancements have found value in applications where both delicate and seemingly robust components are prone to fatigue failure when excited at their resonant frequency.

Pulse

Within the same time frame as the development of sweep frequency, ìpulseî was identified as an important parameter in ultrasonic cleaning. In simple terms, pulse means turning the ultrasonic energy on and off. Variables are duty cycle (percentage of ON time), frequency of the repeating on-off cycle, and pulse amplitude. Although it would at first seem counter-intuitive to better cleaning to turn off the ultrasonic power for a portion of time, this action does result in better cleaning results. The reason is that there is a burst of high energy ultrasonic power generated each time the ultrasonic power is turned on, which occurs before the cavitation bubble field reaches saturation. During this time, sound passes quite freely through the liquid without being attenuated by the saturated cloud of bubbles which are released by the sound field after the ultrasonic energy is initiated. This effect is familiar to those who have witnessed ultrasonic energy in some very low surface tension solvents. Droplets of solvent are often driven several inches high during the initiation of cavitation in a solvent.

The impact of the recognition of this effect was somewhat softened when it was realized that most ultrasonic cleaning systems at that time already had some inherent degree of pulse at twice the power line frequency as a result of a one half wave rectified power supply. Many manufacturers, in fact, adjusted the duty cycle of this inherent pulse as a means of providing a form of ìpower control.î Pulse, by the way, should not be confused with ìdegas.î The degassing cycle uses considerably longer time periods to allow gas bubbles, formed as the result of cavitation, to float to the surface of a liquid prior to re-applying ultrasonic energy. Pulse usually means 50 to 150 on-off cycles per second while degas uses a cycle time up to several seconds between pulses. The effect of pulse is most pronounced in solvents but can also be seen in aqueous solutions.

Waveform Flexibility

In its simplest form, an ultrasonic generator is nothing more than a frequency converter. Just as a rectifier changes alternating current to direct current, an ultrasonic generator changes electrical energy at the power line frequency to electrical energy at the frequency required to drive the ultrasonic transducer.

Ever-advancing electronic technology has given designers of ultrasonic generators the tools to allow them to customize generator characteristics not limited to sweep and pulse. Today, nearly any imaginable wave characteristic can be customized (using techniques much like those used in musical synthesizers). Ultrasonic frequency and amplitude can be modulated instant by instant. Waveform patterns can be programmed or randomized depending on each individual application. Research continues to explore and define the proper use of all of these waveform parameters.

Transducer Technology

Many of the developments in ultrasonic generator technology were driven at least in part by developments in ultrasonic transducers. Transducer technology has advanced notably over the past several years.

Transducer Bandwidth

Ultrasonic transducers are typically designed to resonate at their operating frequency. Much like bells and tuning forks, however, they are less efficient when driven at a frequency even slightly different than the one at which they were intended to operate. The impedance vs. frequency characteristic of a resonant device expresses its sensitivity to frequency. The deeper and sharper the resonance, the more selective the device is to frequency. Figure 1 shows the impedance vs. frequency curves for a transducer with a narrow bandwidth and a transducer with a broad bandwidth.

Classical transducer technology would favor sharper and deeper resonance, as shown in the graph on the left in Figure 1, for a transducer designed to operate at a single, fixed frequency. Ultrasonic transducers, providing sharp resonance, provide high efficiency output and the feedback required to allow feedback loops that control and automatically adjust generator frequency in real time in response to the characteristics of changing loads.

The development of frequency sweep technology, therefore, presented a bit of a technical conundrum. Sharply resonant transducers suffer a significant reduction in power output when driven by an ultrasonic generator providing sweep frequency. In fact, maximum efficiency is only achieved at the instant the driving signal equals the resonant frequency of the transducer (twice every sweep cycle). Maximizing the effectiveness of sweeping frequency prompted the redesign of ultrasonic transducers to provide a wider frequency acceptance. Figure 2 shows one benefit of a wider bandwidth transducer.

Higher Frequencies

Prior to the early 1990s there was a notable gap in the utilized ultrasonic spectrum. This gap was bordered by the highest ultrasonic frequency (something just short of 100kHz) and the frequencies near 1mHz used for megasonic cleaning. It is generally agreed that megasonic cleaning is based on the phenomenon of acoustic streaming, which is a somewhat different mechanism than that of cavitation, which is associated with ultrasonic cleaning. Acoustic streaming does not necessarily involve the formation and violent collapse of cavitation bubbles. Megasonic technology is also a more ìline of sightî phenomenon than cleaning using cavitation.

This frequency gap started closing when the removal of micron and sub-micron sized particles became important. Increasing frequency has two effects that are beneficial to the removal of small particles. Higher frequencies produce smaller cavitation bubbles that are able to produce a force normal to the substrate of sufficient magnitude to dislodge and remove very small particles. In addition, the thickness of the ìboundary layerî present at the interface between a liquid and a substrate is reduced at higher frequency. Within the boundary layer, it is difficult or impossible to produce the relative shearing forces required to remove particles. Reducing the thickness of the boundary layer provides access to smaller and smaller particles for removal.

The operating frequency of a transducer is determined by its geometry. In general, a shorter transducer (just like a shorter tuning fork) will operate at a higher frequency. However, by utilizing harmonics, transducers of varying size can be designed to operate at a frequency other than that established by length alone. In any event, transducers became available that would operate at frequencies above 100kHz. These transducers quickly found favor in systems where the removal of very small particles was of primary importance. This area of cleaning was quickly dubbed ìprecision cleaning.î The frequency spectrum continues to grow, now having reached 300kHz. The benefit of higher frequency in the removal of small particles is well established. The down-side of higher frequency is that larger particles are often too well attached to be removed by the more gentle force provided by the implosion of smaller cavitation bubbles produced at higher frequencies.

Multiple Frequencies

The effectiveness of higher ultrasonic frequency for removing small particles has been demonstrated. Higher frequencies, however, may not provide cavitation implosions with sufficient energy to dislodge and remove larger particles. In fact, a given frequency will most effectively remove particles falling within a given size range.

The removal of a wider range of particle sizes can be achieved in one of two ways. Increased ultrasonic power at a single frequency may provide sufficient energy to remove particles in a broader size range. The risk of this approach is damage to the substrate as a result of extremely high power.

The other approach is to use multiple ultrasonic frequencies each of which removes particles in a targeted size range. A series of cleaning tanks operating at a discrete frequency can do this.

›ontinuing ultrasonic transducer development now offers ultrasonic transducers that resonate at more than one frequency. These multi-frequency transducers, used in conjunction with a digitally controlled multi-frequency generator provide the ability to remove particles with a wide range of sizes in a single process tank.

Application

The options for ultrasonic technology have advanced from a few, set frequencies to a veritable ìorchestraî of inaudible sound. Some of the advancements described above were developed in response to clearly defined needs such as improving the uniformity of cleaning or eliminating part damage due to induced resonance. Others are natural extensions of developing technology and, although available, do not yet have a clearly defined use.

Although cleaning has always been a primary target application for high power ultrasonic technology, new uses are being developed for growing ultrasonic technology as this article is written. De-agglomeration and particle size refinement of CMP slurries, micro-finishing of surfaces to change or enhance surface characteristics, applications in liquid particle counting, enhancement of plating and other deposition processes, and a multitude of other new developments are either at or nearing production stage.