We had the pleasure of interviewing Dr. Patrick Wellborn, our Product Development Engineer, based at our headquarters in North Carolina, USA. Over the past few years, Patrick has been deeply involved in the technical design, refinement, and development of the Sol for Atmosphere AX system, working closely with the broader product development and manufacturing teams. In this interview, we aimed to gain insights into the development process and explore what the future holds for this exciting new field of research.
What inspired the development of Sol for Atmosphere AX?
Like all of our projects, the concept for Sol was driven by market research. The team works with researchers to identify the topics they are interested in and compiles those needs into a Marketing Requirements Document (MRD). As an engineer, that is the point where I get involved.
That first step is really about asking: “Is this something we can actually build?”. For Sol, that included factors such as the type of light, wavelength ranges and power levels customers needed. And of course, a part of my role is to understand if these requirements are even feasible!
If the answer is yes, the project moves into the feasibility stage, where we gather the data needed to demonstrate that it can become an actual product. It then progresses to the design stage, where I use CAD tools to develop and test the design, before moving into implementation. At that point, we hand off the hardware designs, manufacturing documentation, and design notes to the manufacturing team so they can build the product to be tested internally and externally.
How do you know whether an idea is technically feasible, especially when it’s outside your area of expertise?
Of course, we already have the engineering and physics background to understand the underlying concepts, but photoillumination, lasers, and optics were a completely new territory for us! There were a lot of aspects to consider, from laser sources and fibers to lenses and how to guide and focus light so it reaches the sample in the right way.
Rather than trying to learn everything through trial and error, or spending months figuring things out ourselves, we brought in laser and optics experts. This allowed us to move the project forward which much more confidence. They helped us understand the differences in hardware we needed, including laser sources, fibers, and the optics (there are so many different lenses). We understood the concepts, but having experts in the field helped us turn those concepts into something practical and manufacturable.
What was the biggest engineering challenge in bringing controlled illumination into in situ TEM?
One of the biggest challenges was understanding that introducing light also introduces heat. It wasn’t enough to simply deliver controlled illumination, we also had to know the exact temperature changes on the sample. Without this information, it would be impossible to determine whether a reaction was being driven by the light or by the heat.
We set strict targets for temperature accuracy that we wanted to achieve. Achieving those targets required extensive calibration of the E-Chips under laser illumination and several design iterations. That process ultimately allowed us to decouple the effects of light and heat and provide researchers with confidence in the accuracy of their experiments. There was definitely an “aha moment” when we were able to solve that problem and prove that we could separate the effects of light and heat.
In this example, the software is placed in the photothermal mode, reflecting temperature changes due to changing light intensity. The illumination provides enough localized heating to initiate a reaction that normally requires temperatures of 350–400 °C. As the light intensity increases, the E-chip records a corresponding temperature rise, while the reaction rate accelerates. Once the light is removed, the reaction stops.
How does the AXON software support temperature control?
AXON gives users the option to enable or disable temperature compensation. As the laser introduces heat to the sample, the system’s closed-loop temperature control can automatically compensate by removing that additional heat and maintaining the target sample temperature. Users can also turn that compensation off, allowing the sample temperature to rise naturally under illumination. This makes it possible to distinguish whether a reaction is driven by the light itself or by the additional heat, giving researchers much greater control over their experiments. That ability to decouple those two effects is really powerful for users in the field.
Sol isn’t simply the integration of a fiber optic illumination, it contains an entire workflow. Could you walk us through the user’s workflow?
We developed a separate ‘Ex Situ Power Characterization Station’ that allows users to accurately determine how much light reaches the sample. The workflow starts by aligning the optics to maximize the laser power delivered through the system, as there can be losses between the laser source and the sample. Users first verify the output with a laser sensor before connecting the fiber to the holder.
From there, the light passes through a dedicated aperture E-chip, which replicates the optical path of the experimental E-chip. Combined with an optical power meter, this allows us to accurately calculate the power density at the sample rather than just the laser output. Without that calibration, users would not know how much light is actually reaching the specimen. You would essentially be shining light at the holder without knowing exactly what is reaching the sample.
The workflow is supported by two software applications. The calibration application guides users through the characterization process and generates a calibration file. That file is then imported into the Atmosphere AX application, where users simply enter the desired power density. Using the calibration data, the software automatically calculates the laser settings needed to deliver the requested power at the sample.
By enabling proper measurements through the holder, researchers can apply accurately quantified light intensities across experiments, improving reproducibility, enabling direct comparison between datasets, and increasing confidence in photochemical and photocatalytic results.
Could you describe the ex situ power density characterization station a bit more: What are the components and how does it work?
The characterization station serves two purposes. First, it calibrates the laser power density at the sample. Second, it acts as a safety enclosure for the class 4 laser.
From a safety perspective, the enclosure is designed to prevent stray laser light from escaping. All seams are sealed, with additional foam around the hinges and brush seals around cable feed-throughs to block any light leakage. It also incorporates a laser interlock, which immediately cuts power if the enclosure is opened. The idea is that we want the system to be safe by design, not just safe through user behavior. Of course, sometimes the enclosure needs to be opened while the laser is on, such as during optical alignment. In that case, the system limits the laser output to a safe level, providing enough power to align the optics without exposing users to the full laser intensity.
The calibration side of the station starts with a temperature-controlled laser source. An integrated thermoelectric cooler (TEC) keeps the laser at a constant temperature, ensuring a stable power output. The beam then passes through collimating and focusing lenses before entering the optical fiber. A fiber junction allows users to switch between direct power measurements and holder calibration without disturbing the optical alignment.
The station also includes a holder stand and sensor mount that simplify alignment and ensure repeatable measurements. Together, these components allow users to accurately characterize the laser power reaching the sample while keeping the workflow straightforward. Because the calibration measures the complete optical path, it also accounts for any transmission losses introduced by components such as fibers or changes in cable length, ensuring the power density at the sample remains accurate. Additional tools, such as alignment keys and dust caps, are included to maintain the optical system and support routine operation.
How critical is it to perform calibrations like these for in situ TEM measurements?
Calibration is essential because researchers need to know exactly how much light reaches the sample. If we do not know that value, we do not know what is actually driving the reaction. Many photo-induced processes depend on both wavelength and power density, so inaccurate measurements can lead to uncertainty about what is actually driving a reaction at the nanoscale.
The challenge is that there are losses throughout the entire optical path. The laser output alone does not tell us how much power is reaching the sample, since losses can occur from lens alignment, fiber length, optical junctions, and components inside the holder.
To account for these losses, the calibration workflow uses a dedicated aperture E-chip that closely replicates the experimental E-chip. The silicon nitride window introduces its own losses through absorption and reflection, so having the same material in the calibration setup ensures those effects are included. The aperture itself matches the sample area of an experimental E-chip, and the backside of the chip is coated with aluminium. This ensures that any light passing through the aperture is reflected toward the sensor rather than spreading out and being lost, allowing us to accurately measure the power reaching the sample area.
By calibrating through the complete optical path, users can determine the true power density at the sample before an experiment. This provides confidence that the light conditions are accurately controlled, even when components are changed, the system is moved, or the optical alignment is adjusted.
Were there any moments during development when the team realized they had solved a particularly difficult problem?
I would say there were two major moments. The first was solving the heating challenge. Understanding how to separate the effects of light and heat and achieving accurate temperature control was a major milestone for the team.
The second was realizing how critical the fiber alignment within the holder was. We developed what we call the “snorkel,” which positions the fiber at the correct angle while fitting within the limited space of the TEM environment. The challenge was not only designing it to work, but also making the alignment repeatable with extremely tight tolerances. To achieve this, we used imaging techniques to precisely determine the correct position of the fiber and lens within the snorkel. Small changes in alignment, even on the micron scale, can significantly affect the power profile at the sample. Getting that alignment process right became one of the highlights of the project. It was one of those problems where once you solve it, you really appreciate how much work went into making something that appears simple to the user.
What part of Sol are you personally most proud of?
I would say the characterization station is what I am most proud of. I had a major role in designing it, and it was a really interesting challenge. It was fun because I had never designed something where the goal was not only to control something, but also to keep something out, keeping the light exactly where we wanted it. It took three or four different concepts, many discussions, and design iterations to determine the right form factor, how it should open, and what components needed to be included. Seeing all of those ideas come together into a functional product was really rewarding.
I am also proud of how we transformed the calibration workflow. Early on, we were literally typing values into spreadsheets by hand, so seeing that become an automated workflow has been really satisfying. The user can run the calibration, generate the file, and transfer it directly into the Atmosphere AX software to run an experiment. Seeing that evolution from a manual process to a simple workflow has been one of the most exciting parts of the project.
What do you hope to hear users saying about Sol that would make you feel like its doing its job?
I would love to hear users say that it is easy to use and that it just works. For me, making the calibration process simple and intuitive is a huge part of the product’s success. After working through all the complexity behind the scenes, having a user say, “this is easy to use,” or “it just works,” would be music to my ears.
Of course, negative feedback is also valuable because that is how we continue to improve the product. Understanding where users run into challenges helps us make Sol better and continue developing a workflow that truly supports researchers.
Thanks to Patrick to share his knowledge and enthusiasm on this new system!
