LED Light Engines for Microscopy

Once LED’s became powerful enough for use in applications like traffic signals and bike lights, every microscopist has been wondering when they’d make their way into our corner of the world. Today there are several options for LED based illumination, of which I will explore in this post. The path towards the current market condition is also an interesting point.

Patents

Around the late 1990’s a few LED manufacturers began to produce diodes that, when powered and cooled properly, would emit enough light to be barely  useful for microscopy applications. This is about the time that a patent was filed, claiming the use of an led illumination source on a scope as a “novel idea”. In my mind this was a major disservice to the microscopy community. To begin with, the concept of placing any photon emitting device or source in a path so as to collect or transmit that source and be of use is inherent. This concept has been used with untold photon generating sources, and is not novel. Secondly, the concept of an LED replacing a conventional light source is not novel. In every other corner of our technological world we have seen, and are seeing, LED illuminators replace conventional filament and gas arc systems. Now of course all credit is due to whomever came up with this and actually did the work. It’s easy to talk about doing something, and a lot more difficult to be the person doing the exploratory work.  On the other hand, this patent, that as an aside was denied in European patent office, was blindly accepted by someone from the USPTO who must not have done enough homework to confirm that the claimed invention was in fact an invention, and not an application of common knowledge. In the end the patent is on the books, and luckily there are other solutions to the problem.

Heliophor from 89 North

A new approach

Enter stage left two manufacturers: 89 North and Lumencor. Both companies saw the ridiculous situation placed upon our community, both developed solutions around the limits by creating new ways of harnessing LED power. Lumencor has a product they call a “light engine”, which relies upon photon emission from crystal composite structures placed in front of an energizing LED. This circumvents the LED patent because no LED photon hits the scope. 89 North uses a phosphor to achieve a similar result. The phosphor is excited by an LED and then emits a longer wavelength light into the scope, however no LED-generated photon enters the microscope. Both systems are commendable for their ingenuity and in the high quality / performance result of the product.

Advantages

What are the advantages of an LED excited or LED driven system? Well, they are numerous. In both systems, the phosphor or composite emits the energy, thus the LED is the driver of the system. This translates into the fact that the LED can affect the photon output of the material used for emission. This then means that if the manufacturers turn down the LED a bit when it ships, they can turn the power up over time, as the emission system decays in it’s performance. What this means to customers is that the photon emission produced by these systems can be regulated. This is a major step beyond traditional sources, all of which lose 40-50% of their power during normal bulb lifetime. Think about this for a second: Why did your first run of wild type cells come out so much brighter than your second? Are you using a core facility scope? How new was the bulb? How long was it between your imaging sessions? Intensity deviations from the excitation source are possibly far more powerful an effect on image quality than most people realize.

• LED’s don’t contain the poisonous gas of gas-arc systems. Bulbs can and will blow. When they do it’s not something to be taken lightly! LED’s don’t have this problem.

• The lifetime of an LED, when properly used, far exceeds gas-arc based lamps. Most LED’s can last 10,000-40,000 hours, with the best arc systems achieving 5,000. This translates into lower operating cost and less hassle over time.

• Mechanical devices such as filter wheels and shutters are not required for use with LED systems. Some excitation systems do require the use of a filter, however this is typically included with the wavelength line package, so it’s not an externally-required item. Because of this the switch times of these light engines is infinitesimal compared to mechanical products. A typical wheel will state 10+ms switch, however this is usually a spec when the wheel isn’t loaded up with filters. So a typical filter wheel with 10 filters usually runs more like 50ms in switch times. Compare this to an LED-driven system with a 1ms switch time.

• LED’s can be manufactured for a much lower cost than a white light arc lamp. This of course translates to lower cost of a system. Lumencor has provided an interesting cost estimation tool to compare the Spectra light engine to other available light sources. I’m curious to see how the real world numbers compare to these estimates, but the point is a compelling one in a core facility or other high use environment.

Disadvantages

• While LED’s cost less, they are not white-light systems, so the user/buyer must specify which excitation lines will be used.

• LED’s still create heat. Of course they are more efficient than arc systems, but the heat they do create must be dealt with. This translates into weight, cooling systems, etc. It’s an incremental improvement on traditional systems, but when comparing the bulk of a traditional system to a multi-LED system that is cooled, they are about the same.

• There aren’t any suitable sources for 340/380 LED based excitation. Anyone working with FURA or who needs UV uncaging or ablation light will find that there aren’t any commercially available options. I’m not certain as to why this is. There are UV based LED’s being used for things like water purification. I’d imaging if these are available at the power required for purification, they’d be available at the power level we need, however it may be a wavelength-position problem, and not one of output energy.

Example Systems

89 North’s Heliophor system is a well thought-out package. Customers can purchase the base unit and three or more wavelength lines to start with, and each additional module for new wavelengths can be purchased at a later time, as needed. Modules for wavelengths can be swapped by the customer in the field, and they are warranted for life.

I had a chance to shoot a video of this system in action at the SPIE show last month.

httpv://www.youtube.com/watch?v=6TdvtZFVHRc

Lumencor offers a Spectra series of light engines that provide 4,7 or even a tunable selection of excitation lines. The Spectra X is as close to a white light system as any engine has yet reached, and is definitely an interesting option for users who need many excitation lines. Both systems use high speed device controllers to interface with cameras on the fly. (89 North’s Controller and Lumencor’s Controller) This is something that hasn’t been a problem with conventional systems, as the slowest component on the system was the filter wheel. Now the wheel can be removed, leaving the camera to be the slowest part. In order to achieve the fastest synchronization between the camera and the device, the computer can be eliminated from the equation. I frequently employ a similar system to this with ASI’s Stage Sequencer module.

Summary

LED illumination in our market totally fits the bill of a disruptive technology. I’m excited to see where it goes. I’m thankful to those who have found ways to innovate despite the barriers, and I believe that 5-10 years from now, using an arc illuminator will be a rare exception, rather than the rule.

-Austin