Neutron Reflectivity for Testing Graphene Oxide Films Sorption of EuCl3 in Ethanol Solution

Neutron reflectivity (NR) was used to study the sorption of Eu(III) by graphene oxide (GO) films exposed to ethanol solution of EuCl3. Most of the earlier sorption studies have been performed using GO dispersed in solution. In contrast, layered structure of GO films imposes limitations for penetration of ions between individual sheets. The analysis of NR data recorded before and after sorption under vacuum demonstrates an increase of GO film thickness due to sorption by 35–40%. The characterization of chemical state of Eu(III) sorbed by GO films by X‐ray absorption near‐edge structure (XANES) in high‐energy resolution fluorescence detection (HERFD) method at the Eu L3 edge reveals that it remains the same as in anhydrous EuCl3. Analysis of all collected data including reference experiments with bulk GO samples allows to conclude that EuCl3 penetrates into GO interlayers with ethanol solution and remains trapped in interlayers after evaporation of ethanol. Sorption of EuCl3 results in nearly complete amorphization of film and likely formation of voids, thus making NR models based on specific volume of unit cell not valid for quantitative evaluation of Eu sorption. Limitations of NR method must be taken into account in future studies of sorption by thin films.


Introduction
Graphite oxide is a hydrophilic, layered material prepared by oxidation of graphite, most commonly using Hummers method. [1]Single-layered graphene oxide (GO) dispersions can be then prepared by mild sonication of graphite oxides in polar solvents.The surface and edges of GO flakes are functionalized with a variety of oxygen-containing groups, most importantly hydroxylic, epoxy, and carboxylic. [2]hese functional groups in combination with high specific surface area in polar solvents make GO an efficient adsorbent of various organic and inorganic pollutants.
GO is considered as a very promising material for treatment of radioactive waste and natural waters because of its high sorption capacity toward radionuclides. [3]GO was also extensively studied as a sorbent for a large variety of water contaminants, including, for example, dyes, heavy metals, and organics.In recent years, GO has been studied also for the adsorption of trivalent Eu. [3a,4] Eu(III) has been considered in some studies as a chemical analogue of other trivalent lanthanides and actinides in nuclear waste. [5]Therefore, understanding the sorption of Eu(III) is especially useful in order to develop more efficient adsorbents for nuclear waste treatment.
It should be noted that the research field related to sorption of radionuclides and heavy metals by graphene-related materials was affected in recent years by multiple retractions (see for example [6] ) and extensive corrections. [7]Therefore, confidence in some of the previous studies related to sorption of Eu(III) by GO was undermined.
Moreover, the sorption studies have been typically performed using only GO dispersions but not solid graphite oxides or multilayered GO laminates.GO dispersions can be deposited on suitable substrates (e.g., by spin coating [8] or drop casting [9] ) to make multilayered thin films.The dispersions can also be vacuum filtrated to make free-standing foils named according to expected applications as papers [10] or membranes. [11]The multilayered assemblies are formed by irregularly shaped and sized GO flakes packed approximately parallel to each other with random in-plane orientations.
Sorption properties of multilayered GO are expected to be affected by the size of interlayers in the c-lattice expanded due to swelling in water or other polar solvents used to dissolve DOI: 10.1002/pssb.202400069Neutron reflectivity (NR) was used to study the sorption of Eu(III) by graphene oxide (GO) films exposed to ethanol solution of EuCl 3 .Most of the earlier sorption studies have been performed using GO dispersed in solution.In contrast, layered structure of GO films imposes limitations for penetration of ions between individual sheets.The analysis of NR data recorded before and after sorption under vacuum demonstrates an increase of GO film thickness due to sorption by 35-40%.The characterization of chemical state of Eu(III) sorbed by GO films by X-ray absorption near-edge structure (XANES) in high-energy resolution fluorescence detection (HERFD) method at the Eu L 3 edge reveals that it remains the same as in anhydrous EuCl 3 .Analysis of all collected data including reference experiments with bulk GO samples allows to conclude that EuCl 3 penetrates into GO interlayers with ethanol solution and remains trapped in interlayers after evaporation of ethanol.Sorption of EuCl 3 results in nearly complete amorphization of film and likely formation of voids, thus making NR models based on specific volume of unit cell not valid for quantitative evaluation of Eu sorption.Limitations of NR method must be taken into account in future studies of sorption by thin films.
pollutants.The conditions of sub-nm confinement between GO sheets might also affect the sorption, making it different compared to single-layered flakes in dispersions.
Some studies reported possible selectivity in sorption of lanthanides by GO which could possibly be used for their separation from mixed solution. [12]This effect was reported for extremely thin 2-3 layer-thick GO films on water-air interface. [13]herefore, the GO films were studied at the conditions of saturated swelling in given solutions and dissolved species in hydrated state.Possible effects of sub-nm pore size provided by GO interlayers on sorption of lanthanides need to be verified using multilayered materials in order to be closer to practical applications related to possible separation of lanthanides using selective sorption.
Neutron reflectivity (NR) was demonstrated in our earlier studies as a powerful method to characterize sorption properties of relatively thin (few hundreds of Å) GO films.The interpretation of NR data was based in these studies on assumption that GO flakes remain parallel after sorption, thus allowing to estimate increase of interlayer distance using change of film thickness.The exact composition change of GO films was evaluated using NR for sorption of polar solvents from vapor in pure state or binary mixtures. [8,14]The GO films easily detach from Si substrate in water solutions but are stable in ethanol solutions.Therefore, the sorption studies have to be performed using ethanol as solvent, as in our recent studies of dye sorption by thin GO films. [15]n this work we applied NR method to study the sorption of Eu(III)Cl 3 onto GO films from ethanol solution.Some reference experiments were performed also using bulk GO powder samples.Additional characterization was performed using a variety of methods including X-ray diffraction (XRD), high-energy resolution fluorescence detection (HERFD)-X-ray absorption nearedge structure (XANES), thermogravimetric analysis (TGA), Fourier transform infrared spectroscopy (FTIR), and Raman spectroscopy.Demonstration of the NR method feasibility for quantitative evaluation of Eu(III) sorption by multilayered GO films would allow us to test also sorption of other lanthanides and to study possible selectivity effects related to penetration of ions into sub-nm interlayers of GO.Complex analysis of data demonstrates high sorption of EuCl 3 inside GO interlayers in chemically unmodified and desolvated state.Our experiments also demonstrated that using only NR is not sufficient for evaluation of relatively complex sorption processes which involve significant change in packing of GO sheets and amorphization of structure.

Results and Discussion
Sorption experiments were performed using GO films deposited on Si substrates.The films were immersed into ethanol solution of EuCl 3 for different periods of time and analyzed by NR after removing solution, drying the film, and brief rinsing with pure ethanol to clean the surface.NR scans were recorded from the dry thin films before and after exposure to EuCl 3 /EtOH solution.
The first set of experiments was performed with EuCl 3 solutions with concentrations (0.001, 0.01 and 0.05 M) by exposing GO thin films to excess of solution for 4 h followed by drying.Figure 1 shows an example of NR data collected from GO film before and after sorption experiment with 0.05 M solution.The NR data for pristine film (Figure 1a) show well-defined oscillations typical for good quality GO with homogeneous thickness.The NR curve recorded from the GO film after exposure to EuCl 3 /EtOH solution is distinctly different.[16] Expanding the GO lattice with ethanol also allows permeation of dissolved ions and molecules into the space between GO sheets.[16] Therefore, the change in NR curve shown in Figure 1 does not correspond to trivial swelling in ethanol.It occurs due to intercalation/sorption of Eu(III) which remains in the film.Analysis of the NR data allows to evaluate the change in film thickness due to sorption and change in the film's composition using a simple structural model.It is also possible to estimate the interlayer distance change due to the sorption using change in the film thickness.It is assumed that GO film is composed of GO sheets parallel to the substrate and that the number of layers remains the same after sorption.In this case, the change in the film thickness can be assigned to change of interlayer distance due sorption of ions or molecules.Fitting to the NR scans was performed using a three-layer model of GO film similar to our earlier studies of GO films.The Scattering Length Density (SLD) profiles (Figure 1b) were simulated using substrate Si layer, air layer and three layers representing GO films (including Si-GO interface layer, GO-air interface layer).The profiles simulated for the GO films after exposure to ethanol solution of EuCl 3 clearly demonstrate increase of the film thickness and change of chemical composition.Details of the calculations providing change in the film composition and thickness are provided in supporting information.The width of GO film interface layers also increased somewhat after sorption experiments, most likely reflecting stronger disorder induced by first strong expansion of GO lattice due to swelling in ethanol, followed by contraction of lattice due to evaporation of ethanol.We cannot also completely rule out somewhat inhomogeneous Eu sorption along the film cross section, for example, some excess at Si-GO interface layer.Similar observations have been reported previously for polymer films exposed to swelling in water/methanol mixtures. [17]e also performed reference experiment with GO film immersed in pure ethanol for 24 h.The NR scans recorded from this film on air before and after exposure to pure liquid ethanol appeared to be very similar (Figure S7, Supporting Information), demonstrating reversible removal of absorbed ethanol and overall stability of GO film in the process of accurate immersing into liquid ethanol and removal followed by air drying.
Since the exposure to 0.05 M solution provided rather obvious changes in the film thickness and composition, the next set of experiments was performed using this concentration and varying the exposure time.The results from all relevant NR experiments are summarized in Table 1.
As evident from the change in neutron scattering length (NSL) and increase in thickness of the films, exposure to EuCl 3 results in intercalation of the GO structure, causing its expansion of the structure and change in chemical composition.However, NR does not provide direct information about the chemical nature of intercalated species.Therefore, we considered several possible chemical mechanisms of sorption and compared results anticipated in these scenarios with experimental data.Table 2 summarizes expected outcomes for several hypothetical sorption scenarios.The first scenario assumes sorption of pure Eu(III) by ion exchange (e.g., with hydrogen from carboxylic groups), leaving Cl outside of the film in the form of HCl.Next models assume that some ethanol also remains in the film due to incomplete evaporation (e.g., as a part of solvation shell of Eu(III)).Next variant considers possibility of hydration since Eu is hygroscopic and easily forms hydrates.Finally, the possibility of sorption in the form of EuCl 3 in pure or ethanol-solvated state was considered.
As follows from Table 1, rather large values of sorption in mg g À1 would be observed in all considered variants of sorption.
To distinguish between several possible variants of sorption listed in Table 1, we performed a simple Eu(III) sorption test with bulk GO powder (Figure 2).The sorption values for bulk materials and thin films are not expected to be exactly the same but considering very strong difference in the expected sorption, bulk test provides an important reference point.The change in concentration of Cl in solution over the GO powder was controlled in this experiment, allowing to determine sorption of Eu.The sorption versus concentration curve shows maximal value of Eu(III) Interlayer distance was taken to be 7.9 Å (in air) for all films in agreement with earlier reports. [8,14,15]ble 2. Possible intercalant compounds and corresponding sorption based on difference in NSL (presented in Table 1).All values are in mg g À1 GO. sorption about 202 mg g À1 .This value is significantly smaller than the sorption expected from NR data and proposed model of pure Eu(III) sorption (933 mg g À1 ).Moreover, increase of EuCl 3 concentration to 27 mM resulted in a significant drop of sorption down to 18 mg g À1 .Note that the method based on measurement of Cl concentration in remaining solution provides information about sorption by ion-exchange mechanism but is not sensitive to the sorption of EuCl 3 .Therefore, we suggested that the ion exchange mechanism of Eu (III) sorption is suppressed at high salt concentrations, resulting in a sorption of EuCl 3 , which remains in GO interlayers after evaporation of solution.This kind of sorption can be imagined as simple crystallization of desolvated chloride inside of the space between individual GO sheets.
The chemical nature of sorption was verified using synchrotron-based XANES spectroscopy in HERFD mode at the Eu L 3 edge (Figure 3).The HERFD detection mode significantly improves the total energy resolution of XANES and reduces experimental broadening in the spectra.As a result, the HERFD-XANES spectrum consists of the pre-edge structure (attributed to the 2p-4f corresponding transition) and main edge transitions (related to the 2p-5d excitations). [18]The pre-edge structure and position of the white line generally serve as indicators of the Eu oxidation state, while the post-edge features provide indications of the local structural changes near the absorbed Eu atom.
It is evident from the data shown in Figure 3 that the chemical state of Eu(III) in the GO films is the same as in precursor powder.The post-edge features exhibit remarkable similarity throughout, leading to the conclusion that no structural modifications occurred during the varying exposure times.
In order to exclude possibility of simple EuCl 3 precipitation, we performed a reference test by exposing graphite plate to the solution with the same concentration for the same time.Analysis of graphite sample after exposure to EuCl 3 solution showed absence of precipitation as evidenced by XRD scan and atomic force microscopy (AFM) imaging of sample surface (see Supporting Information file).We also have not revealed surface contamination with EuCl(III) using XRD and AFM scans of GO films surface.We were also unable to detect clear signatures of Eu species in FTIR and Raman spectra recorded from samples studied after sorption (Figure S2 and S4, Supporting Information).However, analysis of FTIR recorded from powder sample showed slight upshift for C-OH peak, indicating possible interactions of EuCl 3 related to hydroxyl groups.Raman spectra of GO do not allow to monitor the types of functional groups of GO and did not provide information about sorbed species.
Therefore, we assumed that the high abundance of EuCl(III) is not related to the surface of films but occurs through the whole bulk of the multilayer sample.The increase of film thickness after sorption by 20-30% allows to suggest that EuCl 3 is intercalated in GO interlayers.In this case, XRD method can be used to verify interlayer distance using d(001) value.
However, diffraction patterns recorded from the GO films after sorption experiments did not reveal (001) peak.Therefore, we performed additional experiments with bulk GO powder once again.Exposure of GO powder to EuCl 3 solution followed by vacuum filtration removal of ethanol and drying resulted in increase of d(001) value from 7.8 to 9.2 Å (Figure 4) but also in very strong drop in (001) peak intensity.Longer exposure of GO powder to EuCl 3 solution resulted in nearly complete  sample amorphization.Therefore, the change in the GO film thickness cannot be anymore interpreted assuming parallel GO layers.The values of increase in interlayer spacing listed in Table 1 can be considered only as a hint to "expected" change for idealized model with parallel GO layers.Note also nearly complete absence of (010) reflection in the sample exposed to EuCl 3 solution for 4 h.The data presented in Figure 3 can be interpreted by assuming that GO layers are not anymore parallel to each other after sorption and strongly deviate from planar shape.
After analysis of all data obtained in our experiments, we suggest that sorption at relatively high concentrations studied in our experiments results in inhomogeneous intercalation of EuCl 3 between GO layers and nearly complete amorphization of structure due to corrugation of flake shape.This suggestion is in agreement with some literature reports on decreased ordering of GO films after sorption of certain other ions. [19]Strong deviation of flake shapes from planar in combination with non parallel orientation of flakes relative to each other explains observed increase in thickness of GO films.Therefore, the models used to evaluate sorption in Table 1 are not valid.The models include certain unit cells with the size in c-direction (normal to GO planes) provided by the interlayer distance between individual GO sheets.The irregular shape of GO flakes does not allow anymore to use this model for exact calculations of chemical composition of samples.
Amorphization of GO films in the process of EuCl 3 sorption explains also why the sorption values listed in Table 1 are exceptionally high compared to literature data.The most reliable study reports maximal Eu sorption by GO (from water solution) of 78 mg g À1 .This would correspond to sorption of EuCl 3 of 133 mg g À1 .This value is in good agreement with bulk tests performed using TGA and bulk GO powder material exposed to 0.05 M EuCl 3 for 16 h (Figure S5, Supporting Information).Precursor GO are composed of carbon and oxygen, leaving only trace amount of noncombustible impurities after TGA scans performed on air at temperatures above 600 °C.The "ash" content of GO material after sorption was about 147 mg g À1 GO.This value is in good agreement with another study [4a] considering different solvents and mechanisms of sorption.
The maximal sorption of EuCl 3 calculated using NR data is about four times higher.Of course, sorption by GO films and bulk powder does not have to be the same, but so significant difference seems to be unlikely.Therefore, we believe that interpretation of NR data can be trusted only in the part related to increase of film thickness but not in the part related to overall sorption.The increase of film thickness by ≈35-40% is likely to occur not simply due to addition of EuCl 3 but due to the strong corrugation of GO flake shape and formation of voids.Therefore, the increase in film thickness cannot be correlated to the change of interlayer distance in GO.The assumption about existence of voids can be supported by the simple calculation which assumes that 40% increase of volume relates to EuCl 3 with density 4.89 g cm À3 while the rest is GO with density of ≈2.0 g cm À3 .The EuCl 3 -to-GO weight ratio would be then close to 1 (≈1000 mg g À1 ).According to TGA data, the true weight proportion between EuCl 3 and GO is about 10 times smaller, suggesting that the true increase of film thickness is largely due to formation of voids between randomly oriented and irregularly shaped GO sheets.

Conclusion
In summary, our experiments demonstrated limitations of the NR method for evaluation of Eu(III) sorption from ethanol solutions.Analysis of NR data allows to evaluate increase of film thickness due to sorption up to ≈40% after prolonged exposure to EuCl 3 solution followed by vacuum drying to remove ethanol.This would correspond to increase of interlayer distance by ≈2.2 Å if the GO sheets comprising the films remained aligned parallel to Si substrate.However, extensive amorphization of GO was revealed by XRD with only rather weak (001) reflection corresponding to significantly smaller increment of interlayer distance.Analysis of HERFD-XANES spectra demonstrated that EuCl 3 is sorbed by GO film without change in chemical state.However, analysis of all collected data suggests that EuCl 3 penetrates into GO interlayers with ethanol solution, but remains inside of interlayers when ethanol evaporates.Our experiments demonstrated that NR is not well suited to analysis of chemical composition of GO films due to numerous possibilities for explaining experimentally observed changes in SLD.The models based on certain compositions of unit cells with parallel GO sheets provide unrealistically high values of maximal sorption, which are not in agreement with bulk weight change tests of sorption capacity.In conclusion, we believe that Eu(III) is sorbed by GO films from ethanol solution in the form of EuCl 3 confined between individual GO sheets and noncrystalline according to XRD.

Experimental Section
Materials: Graphite oxide was synthesized according to the slightly modified Hummers procedure described in details elsewhere. [20]The as-prepared GO showed C/O = 2.46 and some remaining sulfur (0.34 at%), as determined by X-ray photoelectron spectroscopy (XPS) (Figure S1, Supporting Information).Hummers GO by Abalonyx, Oslo, Norway, was used for bulk sorption experiments requiring large amounts of material (1 g) and XRD measurements.C/O for the precursor GO was 2.2 (excluding oxygen from sulfate impurities) as found by XPS.Anhydrous EuCl 3 (99.99%pure) was used for all experiments and the anhydrous powder was handled in inert atmosphere.
Dispersion was prepared from GO powder by sonication for 12 h in ethanol/water mixture (90% ethanol by volume) and centrifuged at 4.4 krpm for 60 min.Initial concentration of GO powder in solution was 3 mg mL À1 , and the final dispersion had a typical concentration of 1 mg mL À1 .Typically 25-100 mL dispersions were prepared.
The GO solution was deposited onto clean 5 Â 5 cm 2 (or 7 Â 7 cm 2 ) Si substrates using spin coating followed by drying at ambient conditions for several days.Spin coating was performed using 1000 rpm for 1 min followed by 1900 rpm for 10 s.The acceleration rate of 500 rpm s À1 was used for the first step and 2000 rpm s À1 for the second step.The procedure was repeated 4-7 times using ≈0.5 mL of solution manually dropped over the surface of the substrate.The resulting GO films were almost transparent with a slight brownish color.The typical thickness in ambient condition ranged from 300 to 600 Å, depending on the number of dropping cycles.The hydrophilic nature of GO along with its flexibility prevented the formation of voids as a significant fraction of the total film volume. [21]herefore, precursor GO films can be considered as a dense material with GO flakes parallel to the surface of the substrate.GO films prepared using a very similar procedure were extensively characterized in our previous studies. [8,14]The films selected for NR characterization exhibited a uniform thickness (except for corners of the rectangular substrate) and a smooth surface.
Characterization Methods: NR experiments were performed at the reflectometer SuperADAM at the Institute LaueÀLangevin (ILL), Grenoble, France, using a monochromatic neutron beam with wavelength 5.19 Å.The fitting of neutron data was made using BoToFit and Igor software.
XRD patterns were recorded using a Panalytical X'pert X-ray diffractometer with Cu Kα radiation (λ = 1.5418Å).The XRD data were recorded at ambient air humidity conditions, which were observed to stay within 22-51% from day to day.
FTIR spectra were recorded by Bruker Vertex 80v FT-IR spectrometer with DTGS detector at vacuum in attenuated total reflectance (ATR) mode.
XPS spectra were recorded using a Kratos Axis Ultra electron spectrometer equipped with a delay line detector.A monochromatic Al Kα source operating at 150 W, a hybrid lens system with a magnetic lens, providing an analysis area of 0.3 Â 0.7 mm, and a charge neutralizer were used for the measurements.The binding energy scale was adjusted with respect to the C1s line of aliphatic carbon, set at 285.0 eV.All spectra were processed with the Kratos software.HERFD-XANES measurements were performed at beamline BM20 [22] of the European Synchrotron Radiation Facility in Grenoble.The incident energy was selected using the <111> reflection from a double-Si crystal monochromator.XANES spectra were measured in HERFD mode using an X-ray emission spectrometer. [23]The sample, analyzer crystal, and photon detector (silicon drift detector, Ketek) were arranged in a vertical Rowland geometry.The Eu HERFD-XANES spectra at the L 3 edge (≈6977 eV) were obtained by recording the maximum intensity of the Eu Lα 1 emission line (≈5849.5 eV) as a function of the incident energy.The emission energy was selected using the <333> reflection of five spherically bent striped Ge crystal analyzers (with 1 m bending radius) aligned at 77°Bragg angle.The intensity was normalized to the incident flux.A combined (incident convoluted with emitted) energy resolution of 1.0 eV was obtained as determined by measuring the full width at half maximum of the elastic peak.The paths of the incident and emitted X-rays through the air were minimized by use of a helium bag installed at the spectrometer, in order to avoid losses in intensity due to absorption.
Bulk Sorption Tests: A 50 mM EuCl 3 stock solution in 96% EtOH was prepared.This solution was split into aliquots and further diluted with EtOH to get Eu solutions with concentrations varying from 50-0.1 mM and a total volume of 13 mL each.From each solution, an aliquot of 3 mL was removed and measured by inductively coupled plasma mass spectrometry (ICP-MS) to get C 0 (Eu).To each solution, a weighted amount (m(GO)) of GO (not Abalonyx) was added and then the samples were shaken for 24 h.After this, 24 h of equilibration time, the samples were centrifuged and 3 mL of the supernatant were analyzed again by ICP-MS to obtain C eq (Eu).A eq (Eu/GO) was calculated by dividing the difference of C 0 (Eu) and C eq (Eu) by C (GO) to obtain the sorption loading of the GO.

Figure 1 .
Figure 1.Example of NR data recorded from GO film in vacuum before (black) and after (red) exposure to EuCl 3 /ethanol solution (0.05 M). a) NR data (including error bars).The solid lines in the figure represent the result of fitting overlaid on experimental scan.b) SLD profile obtained by fitting.L 0 and ρ 0 correspond to thickness of film and SLD of the film before sorption, L 1 and ρ 1 the same parameters after sorption.

Figure 2 .
Figure 2. Sorption of Eu(IIII) by GO powder exposed to EuCl 3 /EtOH solutions in different concentrations for 24 h.

Figure 3 .
Figure 3. Eu L 3 HERFD-XANES spectra recorded from reference powder sample of anhydrous EuCl 3 and GO films exposed to 0.05 M solution of EuCl 3 for different periods of time.Spectra recorded on Eu(II) and Eu(III) references are included for clarity.

Figure 4 .
Figure 4. XRD patterns recorded from samples of bulk GO powder before and after sorption experiments with 0.05 M solution of EuCl 3 .a) Precursor GO powder before sorption and after exposure to 0.01 M EuCl 3 -ethanol solution followed by vacuum drying.b) The same after prolonged exposure for 5 days.

Table 1 .
Summary of data collected using NR for GO thin films before and after exposure of 0.05 M EuCl3/EtOH solution for different periods of time.Difference in NSL corresponds to a change in chemical composition of the film as a result of sorption.