Ion solvation as a predictor of lanthanide adsorption structures and energetics in alumina nanopores

Adsorption reactions at solid-water interfaces define elemental fate and transport and enable contaminant clean-up, water purification, and chemical separations. For nanoparticles and nanopores, nanoconfinement may lead to unexpected and hard-to-predict products and energetics of adsorption, compared to analogous unconfined surfaces. Here we use X-ray absorption fine structure spectroscopy and operando flow microcalorimetry to determine nanoconfinement effects on the energetics and local coordination environment of trivalent lanthanides adsorbed on Al2O3 surfaces. We show that the nanoconfinement effects on adsorption become more pronounced as the hydration free energy, ΔGhydr, of a lanthanide decreases. Neodymium (Nd3+) has the least exothermic ΔGhydr (−3336 kJ·mol−1) and forms mostly outer-sphere complexes on unconfined Al2O3 surfaces but shifts to inner-sphere complexes within the 4 nm Al2O3 pores. Lutetium (Lu3+) has the most exothermic ΔGhydr (−3589 kJ·mol−1) and forms inner-sphere adsorption complexes regardless of whether Al2O3 surfaces are nanoconfined. Importantly, the energetics of adsorption is exothermic in nanopores only, and becomes endothermic with increasing surface coverage. Changes to the energetics and products of adsorption in nanopores are ion-specific, even within chemically similar trivalent lanthanide series, and can be predicted by considering the hydration energies of adsorbing ions.

The manuscript deals with an interesting topic. The text is well written and fairly easy to read, but nevertheless there are a number of comments that should be eliminated before publication. Experiment: Al2O3 materials and temperature-controlled batch adsorption experiments Why was the pH 6.5choosen? Since lanthanides are quite active in complexation, did the authors check the effect of the buffer on sorption? Has the sorption on the lanthanide filters been tested? It could have a strong effect on the sorption value (I mean Kd vs T) Where can we see the data obtained to calculate the thermodynamic parameters X-ray absorption fine structure analysis Why pH 6 was chosen? The pH value should be written somewhere in the text, not only in the experimental part The obtained EXAFS spectra should be present somewhere (may be in SI) in the k-space. Adsorption complexes on corundum and nanoconfined Al2O3 surfaces Table 1. Could the authors explained why they use so different parameters for the fitting^ in particular K range and R range. Fir comparison the EXAFS spectra usually use the same condition for the whole dataset. Why the ΔE0 differs so much from sample to sample? And where are the Tb spectra? What about the influence of strong and weak sorption site? How this fits in with the authors' hypotheses? It would also be interesting to see the switching of sorption mechanisms for Nd not only on exafs data but also on experimental sorption data? Do the authors not have such data? That would be an excellent confirmation Reviewer #2 (Remarks to the Author): In this paper, the authors use X-ray absorption fine structure spectroscopy and operando flow microcalorimetry to determine nanoconfinement effects on the energetics and local coordination environment of trivalent lanthanides adsorbed on Al2O3 surfaces. The presented work illustrates that the energetics and products of adsorption could be predictably controlled by changing the size of a reactive nanopore. So, I suggest that this manuscript can be considered after minor revision.
1. The XAFS results on Nd-O, Tb-O, Nd-Al, Tb-Al and so on are reported in this work. What are the reference and criterion when these XAFS data are analyzed? Them should be metioned.
The manuscript by Ilgen et al describes experimental studies of lanthanide adsorption to free and confined mineral surfaces. Using XAFS and calorimetry techniques, the adsorption of Lu, Nd, Tb to nanoporous alumina and corundum were investigated. Important and very interesting differences between nanopore adsorption and adsorption to the free mineral surface are observed. The adsorption corundum is endothermic for all cations, while adsorption in nanopores was always exothermic. Subtle changes in the structure of surface complexes are also observed, which seem to indicate tighter surface bonding in nanopores. The properties of ions in nanopores are poorly understood and detailed, atomic level research is needed to understand these complex systems better. The experiments and analysis are carefully carried out and the conclusions are sound and novel.

Point-by-point response to the reviewers' comments,
Manuscript submitted to Communications Chemistry, title: "Ion solvation as a predictor of lanthanide adsorption structures and energetics in alumina nanopores", Anastasia G. Ilgen, Nadine Kabengi, Kadie Sanchez, and Jacob Smith Reviewers' comments are shown in regular black font, and authors' responses are in blue italics.

May 28, 2023
Reviewer #1 (Remarks to the Author): The manuscript deals with an interesting topic. The text is well written and fairly easy to read, but nevertheless there are a number of comments that should be eliminated before publication.
Response: We are grateful to the reviewer for their constructive feedback. To address the comments below, we re-fitted the data using the same k-value in the beginning of the range for all samples. The kvalue for the end of the range was chosen based on the noise level in the data. The R-range was selected to include 1 st and 2 nd shells, therefore R-range varies slightly between the three elements since their Ln-O and Ln-Al distances vary slightly. For the same element we use the same R-range in the dataset. We hope that these revisions make our work acceptable for publication in Communications Chemistry.

Experiment:
Al2O3 materials and temperature-controlled batch adsorption experiments Why was the pH 6.5choosen?
Response: We chose pH 6.5 because it is environmentally relevant: due to the atmospheric CO 2 dissolution into natural waters, this pH is typical in natural systems. Second important reason for choosing pH <7 is because it is below the pKa values for lanthanides (>7), therefore they exist as aquacomplexes with a +3 charge in solution. We made a note about these two reasons in the manuscript text.
Since lanthanides are quite active in complexation, did the authors check the effect of the buffer on sorption?
Response: In the reported results, we only used HEPES buffer in the temperature-controlled experiments to avoid opening the vials and causing temperature fluctuations during the experiments. For microcalorimetry and XAFS samples the pH was adjusted with inorganic compounds as noted in the Methods section. Based on the reported studies for lanthanides, the complexation constant with HEPES buffer is measurable, but it is not significant, and the vast majority of a lanthanide in solution is predicted to exist as aqua-ions. For example, Eu(III) complexation constant with HEPES is 10 -4.1 (Mandal et al, 2022).

Has the sorption on the lanthanide filters been tested? It could have a strong effect on the sorption value (I mean Kd vs T)
Response: We did not measure lanthanide adsorption onto nylon membrane filters, as such filtration is considered a "gold standard" for natural and synthetic samples intended for ICP-MS analysis. Based on the previous studies, filtration through nylon filter with 0.2 um pores removes 0.2% of lanthanide (lanthanum 3+) during sample filtration (Weltje et al., 2003)

. Our overall errors (including pipetting, weighting, dilution and ICP-MS analytical errors) are ~3%, and the 0.2% error is included in this propagated error value since stock solutions containing lanthanides are also filtered in the same manner as the samples.
Where can we see the data obtained to calculate the thermodynamic parameters Response: We included the dataset in the SI file.
X-ray absorption fine structure analysis Why pH 6 was chosen? The pH value should be written somewhere in the text, not only in the experimental part

Response: We chose pH 6 because it is environmentally relevant: due to the atmospheric CO 2 dissolution into natural waters, this pH is typical in natural systems. Second important reason for choosing pH <7 is because it is below the pKa values for lanthanides (>7), therefore they exist as aqua-complexes with +3 charge in solution. We made a note about these two reasons in the manuscript text.
The obtained EXAFS spectra should be present somewhere (may be in SI) in the k-space.

Response: We included all XAFS spectra plotted in k-space in the SI file.
Adsorption complexes on corundum and nanoconfined Al2O3 surfaces Table 1. Could the authors explained why they use so different parameters for the fitting^ in particular K range and R range. Fir comparison the EXAFS spectra usually use the same condition for the whole dataset.
Response: The noted differences in the selected ranges do not impact our fitting results. To address this comment, we adjusted our XAFS data fits to use the same k-range (2.6-10) for all samples. The R-range was selected to include the 1 st and 2 nd shells, therefore R-range varies slightly between the three elements since their Ln-O and Ln-Al distances vary slightly. For "sample sets" -e.g. the two Nd samples and the two Lu samples we use the same R-ranges for consistency.
Why the ΔE0 differs so much from sample to sample?
Response: To address this comment, we re-visited Tb XAFS data analysis and adjusted ΔE 0 value to be more consistent with the other two lanthanides. This resulted is slightly longer bond lengths for all shells, which were increased by 0.03-0.04 Å. In our final XAFS data fits the ΔE 0 values are similar across the whole series: 6.4 eV and 5.9 eV for Nd; 6.3 eV for Tb; and 7.5 eV and 6.5 eV for Lu.
And where are the Tb spectra?

Response: We added Tb spectra and corresponding fits to the SI file.
What about the influence of strong and weak sorption site? How this fits in with the authors' hypotheses?
The potential presence of sites with different affinities and therefore energetics does not alter the conclusions of the manuscript. Specifically, in flow microcalorimetry experiments, a return of the calorimetric signal to the initial baseline indicates a completion of the reaction, which is then assumed to have reached equilibrium. Therefore, the ΔH measured represents a summation of adsorption on the totality of the sites occupied at that particular concentration. The effect of confinement stands, particularly as it is backed up by changes in the surface complexes. However, we cannot rule out the possibility that should the experiments on confined surfaces be run at higher concentrations (similar to those for corundum) a higher surface coverage and in fact a different equilibrium state may be obtained potentially sampling more sites, both with higher and lower energetics, and resulting in different magnitude of the ΔH.
It would also be interesting to see the switching of sorption mechanisms for Nd not only on exafs data but also on experimental sorption data? Do the authors not have such data? That would be an excellent confirmation Response: We agree with the reviewer that an extensive investigation of lanthanide adsorption onto Al 2 O 3 nanopores would be very interesting. We envision studies on pH-and ionic strength-dependent adsorption, as well as coverage-dependent adsorption thermodynamics to further clarify the phenomenon reported in our current manuscript; however, our research goal was to connect adsorption thermodynamics with molecular structures observed on the surfaces. Therefore, the extensive pH-and ionic-strength-dependency studies are outside of the current scope of work and constitute future research.
Reviewer #2 (Remarks to the Author): In this paper, the authors use X-ray absorption fine structure spectroscopy and operando flow microcalorimetry to determine nanoconfinement effects on the energetics and local coordination environment of trivalent lanthanides adsorbed on Al2O3 surfaces. The presented work illustrates that the energetics and products of adsorption could be predictably controlled by changing the size of a reactive nanopore. So, I suggest that this manuscript can be considered after minor revision.
Response: Thank you for the careful review of our work and providing useful comment to further improve our manuscript.