In this interview, we had the pleasure of speaking with Nathan Rosenmann, a sixth-year PhD student in Materials Science and Engineering at Northwestern University in Professor Gianneschi’s group. Nathan’s research focuses on in situ liquid-cell TEM studies of nanomaterials, exploring how particles assemble and evolve in liquid environments. He shared his perspective on the unique opportunities and challenges of studying both metallic and soft nanomaterials, and how recent advances in liquid-cell design, imaging techniques, and complementary analytical methods are shaping the future of nanoscale research.
You are currently conducting your research at Northwestern University. Could you start by telling us a bit about your background and what initially motivated you to enter this field of nanomaterials and in situ TEM research?
Of course! I am a sixth-year PhD student in Materials Science and Engineering at Northwestern University in Professor Gianneschi’s group. I originally came from a physics background at the University of Illinois at Chicago, where I worked with Professor Robert Klie and was first introduced to electron microscopy. That experience motivated me to pursue in situ TEM as a way to directly observe nanoscale processes.
During my PhD, I became particularly interested in liquid-cell TEM and in understanding how nanomaterials behave in realistic liquid environments, such as during water purification processes. In Robert Klie’s group, I worked mostly with graphene liquid cells and high-Z contrast nanoparticles, but I was ready to move toward more “bulk-like” solutions using silicon nitride liquid cells. There was similar research happening around the country and around the world, but at the time, Professor Gianneschi’s group stood out to me for its focus on combining in situ studies with organic materials.
Graphene offers excellent resolution, but it also introduces strong confinement effects. With SiN liquid cells, it felt like I could study more organic and soft materials in environments that were less confined than the picoliter-scale pockets of graphene cells. That said, there are still many subtle physical effects and nuances we do not fully understand as a community, which means we need a wide range of experiments to really disentangle these effects.
In in situ liquid TEM, electron beam effects, such as radiolysis and beam-induced structural changes, are major challenges. How critical is electron dose control in your experiments, and what strategies do you use to manage or mitigate beam-induced artifacts?
Electron dose control is absolutely critical in our experiments. Many of the systems we study, such as proteins, polymers, peptides, and other materials, are extremely sensitive to the electron beam. In conventional TEM conditions but also in early liquid-phase work, dose rates would reach hundreds of electrons per square angstrom per second, but in our work we typically aim to stay well below that, often under about 0.5 e⁻/Ų/s. The trade-off is that image contrast becomes quite poor, which makes careful experimental design essential. Our group has done extensive work on understanding dose-rate effects in different liquid environments, and there is also excellent work from Hutzler, A. and Fritsch, B et al. that has really helped shape how the community thinks about dose control.
Over the course of my PhD, we have also seen that radiolytic effects depend strongly on imaging mode. TEM and STEM, for instance, can produce very different chemical environments in the liquid, even at similar nominal doses. Many earlier studies focused mainly on whether a particle remained visible or crystalline, but more detailed analyses show that significant chemical and structural changes can occur well before complete damage is obvious. For that reason, chemical information is just as important as physical or structural characterization. Overall, I feel it is important to understand beam effects and electron flux, but also make sure to report on it diligently so that others can understand and replicate work already performed.
What kinds of nanomaterials or chemistries are you currently focusing on, and what makes these systems particularly interesting or important for you to study?
Currently, a major focus of my work is in situ polymerization, where we directly observe the formation of latex nanoparticles within emulsions. These experiments allow us to study polymer nanoparticle nucleation and growth in real time under realistic liquid conditions, which I find especially exciting and something I would like to continue pursuing in the future.
At the previous group, I worked extensively on metallic nanoparticles like gold and silver, particularly in the context of catalysis and battery-related research. While those systems are incredibly important and very relevant to everyday technologies, I became interested in moving beyond high-Z materials toward more organic, biological, and soft-matter systems. Areas like drug loading, polymerization, and hydrogels are central to pharmaceutical and biomedical research, yet only a relatively small part of the in situ TEM community has focused on them so far.
These types of materials are most commonly studied using cryo-EM, which is excellent for high-resolution imaging and limiting beam damage, but it often misses key dynamic information. Cryo sample preparation is time-consuming, and in some cases our work has shown that preparation steps or temperature effects, can introduce artifacts or alter solvated structures. Liquid-cell TEM is also challenging, but it gives access to dynamics and chemical processes that are difficult to capture with cryo alone, making these systems particularly compelling for me to study.
You have previously carried out several nanomaterial studies using the Poseidon AX liquid-cell TEM system, what are the key factors you consider when selecting the appropriate E-chip configuration (e.g. liquid thickness, spacer size, electrode materials, or membrane type)?
In many of my experiments using the Poseidon AX liquid-cell TEM system, I focus on thermally driven processes, so I primarily work with the heating chips in combination with 150 nm spacer chips. Most of the systems I study involve particle growth, assembly, or polymerization, where I am interested in length scales on the order of tens of nanometers. Because of that, ultra-high spatial resolution is not always the primary requirement, as long as the features are roughly 30 nm, I can typically extract the information I need.
What is critical, however, is maintaining a very low electron flux. Many of these systems are highly beam-sensitive, so imaging at low dose is essential, even though it makes visualization more challenging. When higher resolution is needed, my main strategy is to reduce the liquid thickness rather than increase the dose. Practically, that means adjusting the cell configuration, working with thinner windows, and carefully managing flow conditions.
We have also spent time modifying and reproducing our own liquid-cell setups, particularly by changing window sizes. However, this is extremely challenging and time consuming as etching of the nanochip has to be done by eye. That kind of iterative optimization, balancing heating, liquid thickness, and electron flux, has been key to achieving our research goals. It is also an area where I expect continued development, both from our group and from companies like Protochips, as the next generation of liquid-cell designs pushes toward improved resolution while maintaining realistic liquid environments.
When planning a nanomaterial synthesis experiment in liquid-cell TEM, how do you think about imaging configurations such as TEM versus STEM or acceleration voltage?
Yes, imaging configuration is something we think about carefully from the start. We have tested lower accelerating voltages (like 80 kV), and while they do increase contrast, we have found that radiolysis and beam damage also become significantly worse. For our systems, higher accelerating voltages tend to work better, largely because of improved penetration depth in thicker liquid layers and more reliable usable resolution at low dose.
Equally important is having complete confidence in the microscope setup itself. Good TEM alignment, stable imaging conditions, and well-understood beam effects are essential. If you do not fully understand the reaction mechanism, the radiolytic chemistry, and the microscope alignments, there are simply too many variables to interpret the experiment reliably.
That is why we also rely on complementary techniques outside the liquid cell, such as mass spectrometry, IR spectroscopy, X-ray scattering, DLS, cryo-EM, and conventional TEM. With this additional techniques it becomes easier to understand the chemistry, structure, and kinetics in advance. The IR spectroscopy is quite interesting because we can use it to simulate beam damage in the TEM, but then ex situ. In general, the more we know beforehand, the more confident we can be that what we observe in the liquid cell reflects the real process rather than an imaging artifact.
From your in situ TEM synthesis studies, what types of mechanistic insights have you been able to uncover that would have been difficult or even impossible to obtain using only ex situ techniques?
One clear example comes from a study we are just finishing on lower critical solution temperature polymers, which are widely used polysaccharide-based materials. When heated in water, these polymers aggregate into fibrils and then into larger hierarchical structures. They’ve been used for decades in applications like eye drops, yet the assembly process wasn’t well understood.
About ten years ago, cryo-TEM showed the presence of small fibrils, but that was essentially the limit of what could be resolved. Using in situ liquid-cell TEM, we were able to directly observe how those fibrils further assemble in real time, first into bundles and then into larger fiber-like structures. These higher-order assemblies were largely missed in cryo studies.
We also found important differences in flexibility. In cryo, the fibrils appear highly curved or “spaghetti-like,” whereas in liquid they are much straighter, suggesting that sample preparation and temperature effects can significantly influence the observed structure. These insights helped clarify the hierarchical assembly pathway and better explain the structure, property relationships of this class of materials, information that would have been very difficult to obtain from ex situ techniques alone.
With the introduction of new systems offering extended temperature capabilities in liquid EM, (going from -50°C to 300°C), what new research opportunities do you foresee?
A lot of the work we have done so far has focused on thermoresponsive materials, which we heat up and then expect to return to their original state at room temperature. In practice, we often see only partial reversal, so having access to lower temperatures would allow us to fully observe processes like solution–gel transitions or other morphology changes in real time.
There are also materials that respond to cooling rather than heating. For example, some polymers or gels that only form or polymerize at low temperatures. These systems have not been fully explored in liquid-cell TEM, so extended temperature capabilities open up new opportunities to study those dynamics.
Another exciting aspect is that radiolysis slows down at lower temperatures, which could let us observe longer reactions without as much beam-induced damage. Combining this with different solvents, like IPA, could further tune imaging conditions, improve resolution, and give us more control over flux. Overall, being able to explore a wider temperature range really expands the types of chemical and physical processes we can study directly.
If you were to advise a new PhD student entering the field of in situ liquid TEM, what lessons learned or common pitfalls would you highlight for designing successful synthesis experiments?
If I were advising a new PhD student entering in situ liquid TEM, the first thing I would emphasize is the importance of really knowing your microscope and trusting your alignments. Many issues in liquid-cell experiments come back to basic setup and imaging reliability. Alongside that, you need to deeply understand your experiment, have a clear scientific question, a solid hypothesis, and be familiar with the relevant literature. Every observation should be intentional, contributing to answering that question rather than just collecting data blindly.
Hands-on practice is also critical, especially with organic or soft materials, which often have low contrast and are sensitive to radiolysis. Learning to recognize beam-induced effects, understanding your sample’s response under different conditions, and knowing when postmortem techniques like EDS or STEM imaging are useful can make a huge difference. Success in liquid-cell TEM comes from a combination of careful planning, repeated practice, and a strong understanding of both your sample and the microscope, so that what you observe reflects real processes, not artifacts.
Finally, looking ahead, what developments or breakthroughs would you most like to see in nanoscale research, e.g. in instrumentation, software, methodology, or scientific understanding?
Looking ahead, one of the developments I would most like to see is continued improvement in liquid-cell design, particularly reducing both membrane and liquid thickness. Thinner cells would allow higher resolution imaging while still maintaining realistic liquid environments, which would open up a lot of possibilities for studying nanoscale processes in greater detail. Alongside that, a good workflow to integrate spectroscopy, like EELS and EDS, would let us probe not just the structure, but also the chemistry of materials in real time.
Scientifically, I think there is still a lot we do not fully understand about nanomaterial synthesis. Observing morphology alone is often not enough; we also need to understand changes in crystallinity, chemical structure, and how these factors evolve during reactions. I am also excited by the potential for better control over beam-induced effects, mitigating dose and damage will be crucial for studying more sensitive or organic systems. Ultimately, combining tailored cell designs, improved analytical tools, and deeper mechanistic understanding could transform what we can learn about nanoscale materials and their dynamic behavior.
Thank you so much, Nathan, for sharing your insights and giving us a fascinating look into nanomaterial synthesis with liquid-phase TEM!
