UCLA researchers unveil the first-ever structure of supramolecular light-harvesting nanotubes

Posted on

Researchers led by Professor Justin Caram and alumna Dr. Arundhati Deshmukh (Ph.D. ’21) have settled an almost 30 year old question about the nanoscale structure of self-assembled light-harvesting nanotubes. The nanotubes exhibit a brick-like arrangement of molecules that enables highly unusual photophysical properties, giving them extraordinary light harvesting capabilities on par with natural photosynthetic pigments. The findings offer vital insights for designing and stabilizing chromophore self-assemblies for next-generation technologies.

The research is the result of a groundbreaking collaboration between UCLA researchers and Professor Edward H. Egelman’s group at the University of Virginia, Charlottesville.

The team’s research was recently published in the prestigious journal Nature Chemistry. Dr. Arundhati Deshmukh (Ph.D. ’21 Caram group), now a TomKat Center Postdoctoral Fellow in Sustainable Energy at Stanford University, was first author.  She wrote about the collaboration in a blog post on “Behind the Paper” (see below).

Co-authors include Professor Ellen Sletten (UCLA Chemistry & Biochemistry), Caram and his group members Austin Bailey and Jillian Williams, Egelman and postdoc Dr. Weili Zheng (UVA), and Professor Chern Chuang from the University of Nevada, Las Vegas.

A physical chemist, Caram joined the UCLA faculty in 2017. His research program focuses on developing new materials and spectroscopic methods for studying near and shortwave infrared photophysics.

From “Behind the Paper” blog post by Dr. Arundhati Deshmukh:

(Not) Just another brick in the wall: Cryo-EM solves the mystery of excitonic light-harvesting nanotubes

Organic chromophores self-assemble into polymeric architectures like nanotubes, sheets or bundles via weak non-covalent interactions. Within the self-assemblies, excitations on each molecule Coulombically interact over long distances, forming collective excited states covering hundreds of molecules, called delocalized Frenkel excitons. Due to the highly directional nature of Coulombic interactions, the relative geometric arrangement becomes a crucial deterministic factor for their absorption and emission properties. One widely studied model system is the light-harvesting nanotubes (LHNs), self-assembled from an amphiphilic cyanine dye (C8S3-Cl).1,2 Similar excitonic couplings allow the photosynthetic pigment complexes in nature to efficiently absorb sunlight and channel the energy to the reaction center.3,4

Unlike their natural counterparts that are well protected within a protein, the LHNs are much more sensitive to their environments. Small changes in solvent, concentration, humidity or deposition on a substrate can destroy the underlying packings and thereby, the excitonic couplings. As a result, despite the vast research efforts on the LHNs, there was no solved structure. All the structure-property investigations relied on indirect structural models that were benchmarked against the optical spectra. Due to the lack of a better solution, researchers had to resort to extrapolating structural models from indirect evidence such as the crystal structures of monomeric dyes, molecular dynamics simulations or fitting linear/circular dichroism spectra. These methods led to several ambiguities in the structural assignment, puzzling the field for decades, as both brick layer and herringbone-type molecular geometries seemed plausible from the models.

Our previous work showed temperature dependent spectroscopy as a comprehensive tool to gauge the underlying molecular packing and the excitonic band structure.5 However, we would repeatedly get one question – why not just solve a crystal structure of the aggregates? However, these are inherently solution state self-assemblies held together by weak dispersive interactions. Despite the overall long-range order, there is smaller-scale local disorder. Such solution self-assemblies often do not conform to crystallographic space groups that only allow integral symmetry parameters, something not that appreciated in the chemistry community.6 Biological polymers (also inherently solution state) are well-known to have non-integral screw symmetry parameters.7 Indeed, we later found the helical symmetry parameters for the LHNs to be 33.6° rotation and 9.9 Å translation along the tube axis. So, even if we did somehow manage to get a crystallographic solution for the LHNs, it would not represent the native solution state geometry that determines the excitonic coupling. What we really needed was a nanoscale picture in the dynamic solution state, which sounds quite contradictory. Coupled with the technical limitations posed by their sensitivity, any kind of structural solution was almost a pipe dream for us.

We had already tried many standard techniques such as nuclear magnetic resonance, solution small angle X-ray scattering, and even applied for beamtime on serial femtosecond X-ray diffraction, but none gave remote hints of success. Then, in January 2020, Prof. Ed Egelman visited UCLA for a seminar. My advisor, Prof. Justin Caram set up a meeting with him and asked me to join in. I could tell there was an initial disconnect between the communication styles and the jargon. I instantly knew that Ed, as a structural biologist, had a very different point of view from us physical chemists which turned out to be pivotal. Over time, I would go on to ask him such peculiar fundamental questions, to which, he would always have the most refreshingly clarifying answers. We managed to get him interested in the LHNs during that short meeting, and he asked us to send a sample. This turned out to be the beginning of a groundbreaking collaboration. I was able to send a sample to Ed’s postdoc, Dr. Weili Zheng just a few weeks before the COVID-19 lockdown. We had to co-ordinate the timing of the sample preparation and shipping as the nanotubes slowly convert into bundles over longer times. Weili was able to freeze them immediately upon receiving using the liquid ethane plunge freezing method, known to preserve the solvation shell and conserve the LHNs in their native solvated state.

We were in completely unchartered territory for both the teams involved. I still remember our excitement at seeing the reconstructed maps for the first time. We had never seen anything like it before. Justin would talk about it to our colleagues with such enthusiasm! We also got Dr. Chern Chuang on board for modeling the excitonic spectra, and he too was incredibly excited to work on this project despite being in the middle of a faculty job search at the time (he is now an Assistant Professor at UNLV). Once we had the maps with the correctly assigned symmetry, fitting a molecular model to the map turned out to be another challenge, mainly due to lack of any prior knowledge for a starting point. This was an unusual challenge for Ed’s team as well. Nevertheless, we knew that the molecules must be planar to maintain their conjugation. Otherwise, their optical properties would be completely different. We knew that the sulfonate chains must point outside the bilayers. Using these, we were able to get a good fit to the map of the inner wall.

 Most strikingly, we could now clearly ‘see’ the molecules in brick layer arrangement, finally putting the question of brick layer vs herringbone arrangement to rest. The sulfonate interlocking is something the field had never thought of earlier. But as it turns out, this interlocking may be crucial for the slip-stacked arrangement of the molecules. Slip-stacking gives the LHNs their well-known J-aggregate character and all the interesting excitonic properties. We are now using these findings to design self-assemblies with similarly desirable properties, albeit for the shortwave infrared spectral region (1000-2000 nm). We hope this work inspires the ongoing research on other soft materials as well. Just like the interlocking sulfonates, there are many more surprises waiting to be discovered and utilized in chemical design.


  1. Pawlik, A., Ouart, A., Kirstein, S., Abraham, H. W. & Daehne, S. Synthesis and UV/Vis spectra of J-aggregating 5,5′,6,6′-tetrachlorobenzimidacarbocyanine dyes for artificial light-harvesting systems and for asymmetrical generation of supramolecular helices. European J. Org. Chem. 2003, 3065–3080 (2003).
  2. Didraga, C. et al. Structure, Spectroscopy, and Microscopic Model of Tubular Carbocyanine Dye Aggregates. J. Phys. Chem. B 108, 14976–14985 (2004).
  3. Brixner, T., Hildner, R., Köhler, J., Lambert, C. & Würthner, F. Exciton Transport in Molecular Aggregates – From Natural Antennas to Synthetic Chromophore Systems. Adv. Energy Mater. 7, 1700236 (2017).
  4. Fenna, R. E. & Matthews, B. W. Chlorophyll arrangement in a bacteriochlorophyll protein from Chlorobium limicola. Nat. 1975 2585536 258, 573–577 (1975).
  5. Deshmukh, A. P. et al. Bridging the gap between H- and J-aggregates: Classification and supramolecular tunability for excitonic band structures in two-dimensional molecular aggregates. Chem. Phys. Rev. 3, 021401 (2022).
  6. Egelman, E. H. The iterative helical real space reconstruction method: Surmounting the problems posed by real polymers. J. Struct. Biol. 157, 83–94 (2007).
  7. Wang, F., Gnewou, O., Solemanifar, A., Conticello, V. P. & Egelman, E. H. Cryo-EM of Helical Polymers. Chem. Rev. 122, 14055–14065 (2021).

Article written by Dr. Arundhati Deshmukh.