« Back To Vision & Imaging Industry Insights
AIA Logo

Member Since 1984


AIA - Advancing Vision + Imaging has transformed into the Association for Advancing Automation, the leading global automation trade association of the vision + imaging, robotics, motion control, and industrial AI industries.

Content Filed Under:

Packaging , Pharmaceutical , and Semiconductor Packaging , Pharmaceutical , and Semiconductor


UV Inspection Detects the Ultra-Small Details

POSTED 02/20/2017  | By: Winn Hardin, Contributing Editor

The usage of glue to join materials dates back to more than 6,000 years, when our ancestors used collagen-rich animal bones, hooves, horns, and hides to form the perfect adhesive. Early human inventors also discovered that fish scales and bones produced a clear adhesive undetectable by the naked eye, and likely the first of its kind to be used in photographic processes.

Flash forward to the 1960s, when manufacturers introduced ultraviolet (UV) light to instantly cure or harden adhesives, along with inks and coatings. UV light eventually expanded into the inspection realm, opening the door to a new segment within the machine vision industry that not only could detect invisible material properties but also inspect objects in ultra-small wavelengths.

It’s a niche market, but verticals such as pharmaceutical packaging, semiconductor inspection, and electronic leakage detection present opportunities for expanding UV lighting and imaging in machine vision.

Fluorescence Flourishes
When machine vision system designers decide to work with a narrow – or broad – band of the spectrum, they do it for a very specific reason. In the case of UV, “the nature of the UV wavelength gives a certain property of an object that’s important to see, whether it’s a marker, characteristic, or increased level in contrast,” says Greg Hollows, director of machine vision solutions at Edmund Optics (Barrington, New Jersey).

For most machine vision systems concerned with the UV spectrum, UV light is used to illuminate a target and excite a tracer or optically active molecule to fluoresce in the visible. This allows the system designer to capture the return signal by using more commonly available and cost-effective visible-light cameras. “For UV lights, customers gravitate toward 365 nm and 395 nm,” says Matt Pinter, Design Engineer and Cofounder of Smart Vision Lights (Muskegon, Michigan). “Most of the tracing materials call out 365 because it’s the peak of the bell curve of the fluorescent response.”

Today, the most common application for UV fluorescence continues to be inspecting adhesives. “Customers lay a bead of glue or are gluing together a couple of parts, and they want to find the presence [of the clear glue] and verify the right amount is there,” says Pinter. “We see something similar with grease, where they make sure the robot laid the bead of grease and that there were no gaps or the spot was big enough.”

By adding tracer chemicals that absorb UV light and fluoresce in the visible, system designers can use cut-out filters that only allow the fluorescent wavelength through, greatly enhancing the contrast between the glue/grease/adhesive and all surrounding materials.

Printing with “invisible” UV inks, most often in pharmaceutical packaging, is another application with high demand for UV machine vision solutions. “A bottle of eye drops, for example, looks like clear shrink wrap on top of it, but when you hit it with UV, it fluoresces the safety seal,” Pinter says.

UV inks are also used in similar applications where either regulatory control or anti-counterfeit/secure measures are necessary to track and protect products without the text or code compromising the package aesthetics.

Look Honey, UV Shrunk the Defect
In addition to fluorescence applications, another big reason to use UV light in machine vision inspection is the diffraction limit of light. “Diffraction limit is the overall limit of the smallest detail you can see or produced with an optical system,” explains Edmund Optics’ Hollows.

UV light, with its shorter wavelength than visible light, allows someone to see smaller objects than they could with visible light. The semiconductor industry, for example, leverages this physical relationship by employing UV lasers and imaging systems to create increasingly smaller circuits on microchips.

“As semiconductor feature sizes have continued to drop (now below 65 nm), the need for shorter UV wavelengths has risen,” says Rich Dickerson, Manager, Marketing Communications for JAI Inc. (San Jose, California). “A decade ago, UV sensitivity between 320-360 nm was usually sufficient. Now the latest systems are requiring UV sensitivity below 200 nm.” JAI manufactures several cameras capable of imaging in this range.

In semiconductor inspection, UV is used in both reticle inspection and inline inspection of wafers. “These inline systems are mostly used for unpatterned wafer surface inspections that look for polish marks, crystalline pits, terracing, voids or other defects that can affect the performance of IC devices created on those wafers,” Dickerson says.

UV systems also are used in darkfield wafer inspection, which typically looks at patterned wafers using the scatter from a low-angle UV laser to identify surface-pattern imperfections. “Instead of the old photo-multiplier tubes used in previous systems, these systems use UV cameras with high magnification optics to provide better resolution and higher throughput – greater than 1 gigapixel per second,” Dickerson explains.

In the automotive space, UV light can detect scratches or digs on drivetrain components, continues Dickerson. “These can increase friction on these moving parts and therefore reduce performance, increase wear, and decrease operating lifespans.”

UV cameras also are used to detect coronas, UV electromagnetic waves created by electrical power leakage from power transmission towers, which could indicate that something is electrically wearing and needs to be repaired. Detecting electronic leakage in transmission towers used to require a solar blind camera in the 125-150 nm range.

“The camera relied on intensified tubes to get enough amplification of that signal, but now, it’s commonly done using a solar-blind filter on a standard UV camera, as long as the camera has a reasonably low-end range,” according to Dickerson. “The filter blocks out visible light and the higher UV range as well. All that is left is really low UV, which is known as the corona effect. Even the UV that’s produced by the sun doesn’t get into that low range.

“It’s similar to looking at circuitry in flat-panel inspection, where some sort of emission in the UV spectrum shows leakage between pixels or between control circuits laying behind the visible layer of the flat-panel display,” Dickerson says.

Overcoming UV Limitations
Because standard glass covers on sensors will block UV light, UV cameras require either glassless or crystalline covers. While crystalline covers protect the sensor against physical damage, Dickerson notes that they make the camera more expensive.

What’s more, a Lumogen coating placed over the pixels can act as a UV amplifier. “Lumogen is a phosphor coating that reacts to being struck by UV light by emitting light in the 500-650 nm range,” Dickerson explains. “Since UV cameras are equipped with sensors that are sensitive to visible as well as UV wavelengths, imagers that have been Lumogen-coated have their UV quantum efficiency (QE) boosted by 30-50% from this phosphorescence.”

Glass isn’t just used to protect the sensor, but also to increase the light-gathering power of each pixel. When it comes to laser-profiling applications, for example, “users are willing to sacrifice the increase in sensitivity provided by polymer microlenses that increase the pixel’s fill factor to avoid distortion of the light source they are imaging,” Dickerson says. “Microlenses do not inhibit UV light the way a standard cover glass does. They do introduce some normal ‘fringe’ effects, which is why laser profiling typically requires imagers with no microlenses.”

Another critical consideration when working with UV imaging is that the shorter wavelengths mean UV light carries more energy than visible or IR light. According to JAI’s Dickerson, the lifespan of UV sensors is much shorter than those dealing with normal visible lighting.

“The actual lifespan depends on the intensity of the UV light, but typically sensor degradation will occur at around five times faster than for visible applications,” Dickerson estimates.

UV lights contend with their own set of drawbacks as well. Key among them is poor efficiency. According to Smart Vision Lights’ Pinter, a typical UV light will have a 50,000-hour life span, compared to 100,000 for visible lights.

“The energy we put in goes right to heat,” Pinter says. “UV lights burn up quickly. Whether they’re curing glue or fluorescing, people can never get enough UV light. So they will run the lights is hard as they can.”

Much like the relationship between the machine vision and microprocessor industries, machine vision has benefited from the growth in adhesive curing, even though it doesn’t sell lights for that purpose. “The growth in UV curing has dropped the price of UV LED lights to under $5 a piece, whereas four or five years ago we were paying $30 a piece,” Pinter says.

While the world has yet to develop the “killer app” that puts a UV camera in every phone and drives the volume of UV cameras up and prices down, there is enough demand for UV applications to keep industry interested. And like the advent of the microbolometer, which opened IR imaging to the masses, a little interest from the machine vision industry can go a long way to incubating next-generation technologies for as-yet unknown applications.

So while UV lighting may represent a niche segment in machine vision, supported by the semiconductor, packaging, and pharmaceutical industries, machine vision designers are still asking the same question of their component vendors: “How low can you go?”

Vision in Life Sciences This content is part of the Vision in Life Sciences curated collection. To learn more about Vision in Life Sciences, click here.