Good day, Dr. Jyotsana Kala. I'm Arjun. I've had a look through your background and your professional record. I'm interested in hearing more about your work and how you see yourself fitting in here with our faculty. Ready to start?

Please start by introducing your professional journey in academia.

Hi, thank you for the opportunity. So my name is Doctor Jyothnakala and I have completed my PhD from Indian Institute of Technology, Delhi and from the Department of Physics. So my basic research background was to study functional materials for energy storage and convergent devices using first principles and molecular dynamics simulations. So I have studied simple perovskite, double perovskite, high entropy perovskite.

Also explored oxide materials for battery applications using Ebenezer molecular dynamics simulations. In addition, I've also expanded my research experience by having a Commonwealth website fellowship at Imperial College London. At Imperial College London, I have explored conjugated polymers using molecular dynamics simulation for battery applications.

I have expertise research expertise in computational tools like DfT, tools like VAS, CP2K and Gaussian and MD simulation tools like Lamps Grow Max and AMD and VMD Pymol.

In addition, currently I'm a postdoctoral research associate at Imperial College London. In addition.

Thank you for the opportunity to to introduce myself.

Thank you, Professor Kala, for sharing your background. Let's delve into the details of your profile and expertise. Starting with computational modeling, could you outline the specific methodologies or frameworks you frequently employ in your research, particularly when investigating oxide materials for battery applications?

So I have used DfT that is density functional theory and MD that is molecular dynamics simulations to explore materials for battery applications. In my specific project which was on oxide materials for battery applications, I have studied spinel cobalt oxide and effect of doping on its performance next day. I have explored its oxygen evolution, reaction energetics and also find out the found out the rate limiting factors.

And how the epidoping will affect this properties. So I have used the methods like climbing image.

Nudge elastic methods to explore the reaction pathways and also the adsorption kinetics on the surface slabs to explore the adsorption of the reaction intermediates. So for this I have used the tools like VASP and CP2K to explore oxide materials for battery applications.

That’s quite methodical, Professor Kala. You mentioned using VASP and CP2K for studying oxide materials. Could you share an instance where these tools provided critical insights into reaction pathways or adsorption kinetics that might have otherwise been unavailable with alternative approaches?

So a wasp is a DF wasp NCP to our DfT computational tools and softwares we can say they can help us to run simulation on a system. Let's say we had oxide system. So for the first thing in DfT is to optimize the system. So this these tools help us to calculate its.

Its ground state configuration, it can provide its electronic information and also let's say the adsorption kinetics. If we are interested in in that in which we can model the system in initial state and final state. Let's say when a system is alone itself as a surface lab and the system when the reaction intermediates are adsorbed on the surface. So we can model these two systems, calculate the energies and get insights how? The energetics will change on.

Adsorbing different different reaction intermediates. In addition, there are various methods, let's say nudge elastic method that can give us the intermediate images also in the transition image also for the reaction, which can give us the barrier for reaction intermediate to adsorb and desorb. So that can provide us very detailed information about the system at electronic and atomic scale.

Understood, Professor Kala. Your approach demonstrates a strong grasp of computational tools for probing material properties at a fundamental level. Since method selection is integral in computational research, what are the primary factors you consider when choosing between techniques like Density Functional Theory and Molecular Dynamics simulations for your research objectives?

Uh, so like, uh, choosing the computational technique depends on our interest. Let's say, if, uh, we are interested in getting information at electronic level, in that case, I will be choosing the density functional theory. Density density functional theory is a many body, uh, many body, uh, quantum mechanical theory in which we can solve the Schrodinger equation based on multiple approximations made on, made on that.

And we can solve the systems energy and we can get its output properties at electronic level and also so that is a approximated Schrodinger equation approach that is the density functional theory. So there are limitation on the size of the system also. Let's say we can model a system size up to 100 or 200 or few 100 atoms only.

So we can get electronic level information from that. And the next thing molecular dynamics, that provides us information at the atomic level in which we solve the equation, the Newton's equation of motion. And we can track the atomic motions atomic throughout the template at a finite temperature at like with time also so we can have a time scale. Analysis. Let's say we can have nanoscale analysis on a system.

By solving the Newton's equation of motion. So that will not provide as the electronic level information, but that will provide us the atomic level information. And for molecular dynamics simulation if we have forces already available or we can also optimize the forceful power assimilation. So that will be a plus point. So DfT and MD will depend on the type of problems we are interested in and also the size of problem we can. Simulate. Let's say DfT we can simulate few 100 atoms in MD we can simulate.

In which how we can track the atomic motions?

2000 to few lakh atoms also.

Thank you, Professor Kala. Your explanation distinguishes the strengths and limitations of both methods clearly. You’ve covered both the scale and level of detail these computational techniques provide effectively. Let’s shift focus a bit. How have you applied machine learning or AI in conjunction with computational tools like DFT or MD to enhance or accelerate your research outcomes?

So, uh, currently my project at Imperial College London as a postdoctoral research research associate is to apply the machine learning methods for MD simulations. In that case, the first priority method in the MD is to get a force built. Let's say the force build a mathematical framework which will tell us how the atomic interaction will happen, how the atom motion will happen, and we can get the potential energy of the system. So these posts will can be.

Obtained by empirically. Let's say we have some experimental output and we can.

Like parameterize are forceful based on that experimental output and that will give us empirical forceful parameters. Next is the thing. We can parameterize these force fields using Aminatio simulation that is density functional theory in which we can parameterize the different interaction let's say dihedral or angle interactions using the FT simulation and can parameterize it within a force field. The next point is that in addition.

Limitations let's say in the case of empirical it is not always accurate and also it is not transferable from one system to another system. So the forceful that is developed for a particular system cannot be used for other system. In case of DfT this is also the same case and the problem was also like we are dealing with the static charges and we are dealing with the static properties but actually these are not static. The charge also vary with time.

In this case, these cases had some.

And also, uh.

Parameterizing the, uh, the forceful parameters based on our data from Atomic Energy and atomic configurations and then using it in a force field so that that that is how we are creating the machine learned interatomic potentials for a system which could be transferable, which could be scalable, which could be fast and accurate.

The uh, structure also vary with time, which can also affect the properties. So the my my next project is to use machine learning methods to parameterize the force fields. So in which we are like, uh.

That's impressive, Professor Kala. Your emphasis on machine-learning-derived interatomic potentials reflects a cutting-edge approach. Beyond parameterization, how do you validate the accuracy and transferability of these machine-learned force fields, particularly when applied to systems distinct from the training dataset?

So the first of all the accuracy in the transferability can be done like we are doing it at like by running multiple test simulations on our systems. So what we are doing here that we are, we are having a data on first of all how it is able to make the system stable. So we are tracking its thermodynamic stability and structural stability first of all using the. Obtained machine learnt interatomic potentials. So that is the first point of checking whether it is.

It's useful for this material or not. The second point is that that we also already have some fader distribution radial.

The ideal distribution pair functions data from experiments and we are now using that data to obtain it from the like MD simulation with the parameterized machine learning interatomic potentials. The secondly, we are also expanding our work to the organic crystals in which we already have the obtained X-ray diffraction patterns on the organic crystals and we are also checking our forceful parameters whether this can produce the XRD reflection.

Of that organic pistols or not accurately so that means these forceful parameters are retaining the structural integrity of that particular material and.

Transferability. We are just now building our new tools and new connection methods in which we can transfer it from one building block to another building block.

Thank you.

Thank you, Professor Kala. Your approach to validating thermodynamic stability, structural integrity, and experimental consistency exemplifies a rigorous methodology. I'll now proceed with additional questions to assess the breadth and depth of your expertise further. Shifting focus, could you walk me through your experience in teaching computational theory and laboratory courses? Specifically, how do you ensure that students engage deeply with both the theoretical and practical aspects of computational modeling?

So as a PhD student at IIT Delhi, I had a vast experience as a teaching assistant. So I have been a teaching assistant, a sole teaching assistant for courses like solid-state physics, for applied optics. And even I was a teaching assistant for a large cohort like let's say the classroom of more than 500 students for quantum mechanics, I was also a teaching assistant for. Laboratory course like electrical physics laboratory course.

Now I can use this expertise when I'll join this as a teaching position at bit where I will be like making like let's say we can design or we can have a course which will have mostly like.

We have to teach them startingly from the basic Schrodinger equation, Newton equations of motion and then gradually shift from basic to advanced level. In the basic to advanced level, we can then let's say in DfT, we can tell them about how this Schrodinger equation is not solvable when we are dealing with many electron system. Then we can make some certain approximations. Let's say we can decouple the electronic and nuclear motion and then we can. Deal with only electronic motion. Then cone and champagne and Hohenberg and cone cane and they.

Like focus on the molecular modelling. So engaging student in a theoretical and practical aspect will start with teaching them firstly the basic concepts of that particular method, let's say starting the DfT and MD simulation.

Proposed are equations and theorems and then how they have came up with the approach that can now solve many electron system and how then further the functionals are introduced and so we can introduce everything to us to a student in a particular course. But starting with the basics to advance and I think engaging them with the time sequence, let's say engaging and teaching them in a history wise pattern like firstly. Few scientists came and they.

Came up with this discovery, but that didn't work. And then gradually shifting towards the modern computational technique and how the computational technique and computational simulation level is at present. And then bringing to the practical, practical part like along with that particular course, we can design some assignments and some.

The massage maintain system tutorials that will teach them to to like to to a simulation, simple simulation. Let's say we can teach them how to run an energy minimization in MD or how to run an energy minimization minimizations simulation and DfT. And then further we can go to advanced level like how they can now extract the properties, let's say in DfT, how they can extract the density of states plots. Of a system or of a crystal which is well known. We can teach them and we can teach.

Than their particular properties details also and also how they can do that an addition in MD we can teach them how to.

Let's say plot the radial distribution functions in MD for a particular system for which we will already have a data and people and student can compare their data and output that with that already available data. And lastly, we can also go to the advanced levels. In that pattern we can teach them from basics to advanced and we can give them tutorials, assignments in a group project also where they can come up with their own project which will be like. Research based projects by which they can learn some.

Research problems and if they have interest in pursuing research further, they can do their research with particular faculty on the particular or maybe on a computational methods and also with they can go for the graduate studies.

Thank you, Professor Kala. Your structured teaching approach, combining theoretical grounding with progressive, hands-on assignments, demonstrates a solid commitment to student learning and engagement. Encouraging students to progress from fundamentals like the Schrödinger equation to advanced, research-driven projects is indeed an excellent strategy for fostering deep understanding and interest. Could you also elaborate on your experience guiding student projects or research, specifically emphasizing how you ensure they develop both technical skills and independent thinking?

So as a senior PhD student at IIT Delhi, I have mentored few students. Let's say I have mentored a student within my group also to whom I have taught the basic, uh, adsorption energy calculations on an oxide system for battery uh, applications. So I insured him like, let's say I started with how to start with scripting for that particular problem, how to start with a script and what properties we're interested in and.

And thirdly, uh, we can go for the literature review, like what people have already done and what we can do and what is that, what will be approach. So I taught him in a, taught him that how in a strategic manner we can complete a research project. So.

That was one of my experience and that successfully completed this junior PhD student has successfully completed that project and which had LED us to a collaborative project published in Advanced Functional Materials. And I have also collaborated with my other experimental collaborators. Let's say I have taught that PhD students, let's say one of them was in two of them were in IIT Hyderabad to whom I have taught like let's say the first student I have taught the.

How to perform the simple calculations for high entropy materials and how to come up with the materials firstly for the starting dissimulation. And I have also taught another girl about the diffuse and extracting the diffusion properties using MD using tools like lamps and BMD app also at my visiting PD.

Tenure at Imperial. I have also taught a junior undergraduate student on how to calculate the free energy free energies for a conjugated polymer system using Gromacs. So these.

These experiences taught me how to start teaching. First of all you have to start with the basics and also as it is a research project you have to tell them about how to do this literature review and come up with some relevant literature and then read them thoroughly. And from that literature review we can extract the information and also implement into our research project by by appropriately citing them also.

So these experiences I've had already have in mentoring and I've also experiences in teaching. Let's say I've already told you about my teaching experience as teaching is a variety of courses.

Thank you, Professor Kala. Your mentoring experience highlights a thoughtful capacity to guide students through structured learning paths that incorporate both technical acumen and research maturity. Transitioning them from foundational topics to collaborative publications and advanced analysis demonstrates a well-rounded approach. Now, let’s address your publication record. Could you discuss the impact of one of your key publications in reputed journals and how it contributes to advancing knowledge in computational modeling or materials science?

Sorry to interrupt, we're down to the last 5 minutes.

So like my best publication till now is is published in Journal of Materials Chemistry. That paper is about high entropy materials in which let's say in the first few project of my PhD I I studied the perovskite materials for energy application. But I identified A degrading segregation phenomena. But we come up with a different compositions and different type of periscite oxides double simple, but we couldn't omit it completely.

So the next option I have done, I have explored novel high entropy concept. So to start with we have to come up with a strategy also. So in that paper I came up with a novel strategy to narrow down the selection of the atoms to be doped a particular at a particular site for a high entropy materials that could lead to a synthesizable single phase high entropy materials. So we came up with the strategy and we found out that the material have three to four orders higher diffusion coefficient and.

As compared to its parent material, so the Electro catalytic property was enhanced tremendously and to the plus point was that the segregation phenomena which we were observing for the parent material and also for the other double prescribed materials is completely gone here. It was negligible in that particular material. So we have came up with a material, we came up with a strategy and we also came up with the composition which is giving us very high Electro catalytic properties. So this was one of the.

Let's say ACS Applied Energy Materials and Applied Surface Science and few of my project, let's say one of project of my project is still in JMCA in peer review, peer revision and few are in preparation.

Best publications I'm also a publication in physical chemistry, chemical physics as the first author and also two advanced functional materials papers and also some collaborative papers on.

Thank you.

Thank you, Professor Kala. Your work on high-entropy materials and the innovative strategies you developed for improving electro-catalytic properties are impressive and clearly impactful in the field. It is commendable how your research not only advances theoretical methodologies but also addresses practical challenges like segregation phenomena. Given the time constraints, I believe we've covered some vital aspects of your academic and research experience. Would you like to continue discussing any remaining topics, or shall we proceed to conclude the interview?

From you like how the student in game engagement is uh is like tackled at EBIT. Like let's say if a student is interested in computational research, then but support the department or institute provide to him and what support the department and institute provide to the faculty for interdisciplinary collaboration or in uh mentoring and guiding students on research projects.

So I have a question.

Thank you for the thoughtful question, Professor Kala. At our institution, we prioritize both student engagement in specialized research areas and faculty support for interdisciplinary collaborations. For students, the department offers tailored support to those interested in computational research. This includes access to well-resourced computational laboratories, high-performance computing facilities, and mentorship opportunities from faculty actively working in relevant fields. We also encourage students to participate in multidisciplinary research projects, often in collaboration with industry partners, to gain practical exposure. As for faculty, the institution actively promotes interdisciplinary research by fostering collaborations across departments and with external organizations. Funding support is available for innovative projects, and there’s a strong focus on facilitating industry-academia partnerships. Faculty members are encouraged to engage in consultancy services and contribute to high-impact research, for which administrative and logistical support is provided. If you have specific ideas for integrating computational modeling with other emerging research areas, our institution would welcome such initiatives enthusiastically. Does this address your question, or is there any particular aspect you'd like me to elaborate on?

Yes, that has answered me correctly.

Thank you.

Yes, sure.