Membranes of Multiwall Carbon Nanotubes in Chitosan–Starch with Mechanical and Compositional Properties Useful in Li-Ion Batteries

: This work reports on membranes of a combination of chitosan–


Introduction
Lithium-ion batteries (LIBs) dominate the market for portable electronic devices and are becoming increasingly important in other markets, such as electric vehicles [1][2][3].However, many issues are being investigated to improve their security, efficiency and cost.One of the most promising approaches for improving the efficiency of this type of battery is the use of polymer electrolytes [4,5].The carrier transport properties of common ionic liquids are very comparable to those of a polymer electrolyte [6], with the particularity that the polymer can be prepared as a membrane, providing mechanical support and stability to the cells.A solid electrolyte in a battery cell separates the electrodes, avoiding short circuits but allowing for the transport of specific ions (Li + ) from one electrode to the other at acceptable operating speeds.Similar characteristics are necessary for electrolytes for other applications, like proton-conductive materials used in fuel cells [7].The electrolytes are required to exhibit high ionic conductivity and, at the same time, provide mechanical strength.This is the case of multivalent phosphate cross-linked chitosan biopolymer membranes in borohydride fuel cells [8], biodegradable polystyrene sulfonated-lignosulfonate polymers as an electrolyte in fuel cell membranes [9] and, recently, conductive polymers within a hydrogel matrix to obtain synthetic hydrogels with characteristics for applications, such as artificial biological tissues, flexible electronics and conductive membranes [10].Designing new engineered electrolytes like those is a challenge [11].Other classic examples of such membranes are those based on coordination compounds using high-molecular-weight polymers and Li salts, like LiClO 4 or LiN(CF 3 SO 2 ) 2 .Examples of these polymers in the literature are polyethylene oxide, polypropylene oxide, polybis, methoxy ethoxyethoxy-phosphazene, polydimethyl siloxane, polyacrylonitrile, polymethyl methacrylate, polyvinyl chloride and polyvinylidene fluoride [6,12].Most of the synthetic polymers mentioned above are difficult to degrade, causing environmental pollution problems.Because of this, today, there is an interest in natural polymers (carbohydrates), which degrade naturally in a normal environment in a short period of time.Such polymers are commonly referred to as biopolymers [13,14].Research on the use of biopolymers as an electrolyte is currently highly competitive but represents a fertile area for future work [13,15].This is the case with multivalent phosphate cross-linked chitosan biopolymer membranes in borohydride fuel cells [8], biodegradable polystyrene sulfonated-lignosulfonate polymers as an electrolyte in fuel cell membranes [9] and, recently, conductive polymers within a hydrogel matrix to obtain synthetic hydrogels with characteristics for applications, such as artificial biological tissues, flexible electronics and conductive membranes [10].
Carbon nanotubes as anodes in lithium batteries were used in a study from two decades ago.Two of the first works were one where the electrochemical discharge capacity of single-walled carbon nanotubes exceeded 1300 mAh/g after 30 charge/discharge cycles applying a current density of 20 µA/cm 2 [16] and a Japanese review where they concluded that single-walled carbon nanotubes have a high capacity of up to 3611 mAh/g [17].On the other hand, in the literature, one can also find reviews that talk about the use of multiwall carbon nanotubes as electrodes, introducing other materials, like Mn 3 O 4 , CoFe 2 O 4 , Fe x O y , Sn and CoSn, between carbon walls exhibiting capacities ranging from 100 to 1000 mAh/g [2].Additionally, in one of the latest publications, carbon nanotubes synthesized from used cooking oil at 900 • C showed excellent performance as anodes in a Li-Ion battery with a capacity of 213.75 mAh/g in cycle 30 [18].In recently published work, we found that some researchers designed nanoporous nanocomposites of silicon microparticles with carbon nanotubes (PSI-CNTs) that offer a high specific capacity of 3210.1 mAh/g at a rate of 1/15 C. Their initial coulomb efficiency (CE) is 87.3%, but after 100 cycles, the capacity stabilizes above 2000 mAh/g, and the CE is 99.5% [19].In recent studies, carbon nanotubes (CNTs) have been used as promising materials in different applications.Christwardana et al. [20] developed a photo-bioelectrochemical cell with a CNT-chlorophyll photo-biocatalyst, generating photo-currents 6-times higher than the current in darkness.On the other hand, Liu [21] presented a composite of polymer-CNT nanofibers as an anode for Li-ion batteries with a reversible capacity of 1105.2 mAhg −1 and almost 100% columbic efficiency.Meanwhile, Li et al. [22] reported defect-rich sulfurdoped short amorphous carbon nanotubes (SACTs-S) with a high reversible capacity of 1608.7 mAhg −1 at 50 mAg −1 and 538.0 mAhg −1 over 2500 cycles at 2000 mAg −1 .This remarkable performance demonstrates the potential of CNT-based materials for future energy applications.
With the aim to embed CNTs in a polymeric electrolyte, good candidates were searched among biopolymers, finding that cellulose, starch, chitosan, agar, pectin and gelatin are good options [23].For example, solid electrolytes based on polymeric pectin are used for the storage of anionic polysaccharide and ammonium iodide salt, exhibiting their maximum ionic conductivity (4.5 × 10 −3 S/cm) at room temperature [24].Similarly, a polymeric electrolyte based on corn starch with different percentages of sodium bisulfite (NaHSO 3 ) shows an ionic conductivity of 2.22 × 10 −4 S/cm at room temperature [25].On the other hand, biopolymers such as chitosan and starch have been studied by different authors in this context [26][27][28], explaining their properties in detail.Furthermore, some authors have proposed using them as a matrix for carbon nanotubes [29][30][31].The ionic conductivity of chitosan was found to be as high as 10 −4 S/cm after 1 h hydration when it was prepared as membranes [32] and 2.1 × 10 −3 S/cm when it was prepared as ionogel electrolyte with cellulose [33].On the other hand, starch is not far behind, being an excellent electrolyte when mixed with other materials.This is the case of potato starch with Mg(ClO 4 ) 2 , which exhibited high ionic conductivity (2.01 × 10 −2 S/cm), low relaxation time (55 µs) and a wide voltage window of electrochemical stability (about 3 V) [34].Another case is that of flexible electrolyte membranes of nanocomposites of polyethylene oxide/starch nanocrystals complexed with MgBr 2 salt, with an ionic conductivity of 1.16 × 10 −6 S/cm at room temperature [35].More recently, poly(ε-caprolactone) membranes have been synthesized with chitosan and bis(trifluoromethanesulfonyl)imide salt (LiTFSI to 50 wt%) to obtain an ionic conductivity of 7.7 × 10 −4 S/cm [36].
However, the use of combinations of chitosan and starch as solid electrolytes is new, taking advantage of the properties of these polymers.Chitosan provides strength, while starch provides good ionic transport.There is a report on an unplasticized solution of starch-chitosan doped with NH 4 I, which had a good conductivity at room temperature of up to (3.04 ± 0.32) × 10 −4 S/cm [37].In a different work, the synthesis of a biodegradable solid polymer electrolyte of a mixture of chitosan and potato starch, plasticized with glycerol and lithium perchlorate (LiClO 4 ), showed a maximum ionic conductivity of 6.5 × 10 −4 S/cm [38].In a more recent report, a chitosan-starch solid biopolymer with 9 wt% of oxidized graphene showed a maximum conductivity of around 10 −3 S/cm [39].
Combining these natural polymers with carbon nanotubes as electrodes provides a novel and eco-friendly approach to designing LIBs with improved performance, stability and reduced environmental impact.Moreover, the combination of chitosan and starch as solid electrolytes takes advantage of their respective strengths, resulting in a promising electrolyte matrix with enhanced ionic transport capabilities [39,40].This composite represents an innovative and sustainable solution to address some challenges of lithiumion batteries.For this reason, this article reports on the production of carbon nanotube membranes (MWNTs) embedded in a chitosan/starch matrix, with the intention of using them in a future development as an electrode for lithium-ion batteries.Furthermore, two unconventional routes for lithium-modified multi-walled carbon nanotubes with lithium hydroxide (LiOH) are presented: reflux (RLi) and sonication (SLi).The compositional, mechanical and electrochemical properties of the composites are thoroughly studied.

Materials Synthesis
Multiwall carbon nanotubes (MWNTs) prepared via chemical vapor deposition were obtained from Sun Nanotech Co. (Nanchang City, Jiangxi province, China) The MWNTs had a diameter from 10 to 30 nm and a length of 1 to 10 µm, with a purity > 90% and a surface area of 90 to 350 m 2 /g [41].The other reactants were purchased from Sigma-Aldrich Co. LLC.(San Luis, MI, USA).These reagents did not undergo any further purification.
According to the literature cited [42][43][44][45][46][47], the purification/oxidation of the MWNTs was achieved in a 3:1 mixture of HNO 3 (75%) and H 2 SO 4 (98%) at 85 • C for 3 h in a reflux process.The resulting material was vacuum washed to a neutral pH with deionized water.The product is called MWOHs in this work.Afterwards, it was lithium-modified in two different ways, according to the study of [48].
Reflux-lithiated nanotubes (RLi) were synthesized using a carefully controlled procedure involving a specific weight ratio of 6:1 w/w between MWOHs and LiOH.LiOH was dissolved in an aqueous solution with a concentration of 0.01 M. The synthesis process was conducted in a round-bottom flask equipped with a reflux condenser.The flask was charged with the precursor mixture, and a stirring bar magnet was employed to ensure homogeneity.The reaction mixture was then subjected to reflux conditions, maintaining a temperature of 75 place, leading to the formation of lithiated nanotubes.Subsequent to the reaction, the resulting material underwent thorough purification.Vacuum washing with deionized water was employed, with multiple wash cycles being conducted until the pH of the solution reached a neutral state.This step was crucial to remove any residual reactants or byproducts.Finally, the obtained lithiated nanotubes were carefully collected and preserved in a hermetically sealed glass jar to prevent any unintended interactions with the surrounding environment.In contrast, the synthesis of sonicated-lithiated nanotubes (SLi) involved a different approach.The same precursor materials, with the same weight proportions, were used.However, instead of refluxing, the synthesis took place under sonication conditions.The precursor mixture was placed in a flask equipped with an ultrasonic homogenizer.The mixture was subjected to ultrasonic waves at room temperature for a duration of 2 h.The sonication process facilitated the incorporation of lithium ions into the nanotube structure.Similar to the reflux method, the resulting sonicated-lithiated nanotubes were subjected to thorough purification through vacuum washing with deionized water.The purification process was repeated until the pH of the solution reached neutrality, ensuring the removal of any residues.The purified nanotubes were then securely stored in an airtight glass jar to prevent any degradation or contamination.These distinct synthesis approaches, reflux and sonication, provided two sets of lithiated nanotubes (RLi and SLi) with controlled properties.These synthesized materials held promise for various applications due to their unique structural and electrochemical characteristics.
The polymer electrolyte was prepared with a mixture of chitosan-starch solutions.According to previous research [15,[49][50][51], the chitosan (Ch) solution (2% m/v) was prepared by dispersing chitosan in an acetic acid aqueous solution (1% v/v) and then stirring at 100 rpm.After the chitosan was completely dispersed, the solution was sonicated for 15 min to break any air bubbles present and then allowed stand for 1 h.The starch (St) solution (2% m/v) was prepared dispersing starch in glycerin and heating above its gelatinization temperature (90 ± 2 • C) [52], continuously stirring at 100 rpm for 20 min.The solution was then cooled down to room temperature by ceasing heat and stirring.Finally, the chitosan-starch films (ChSt) were obtained mixing equal volumes of chitosan and starch solutions and then sonicating to homogenize.
The complete membranes for each batch of MWNTs (pristine MWNTs, MWOHs, RLi and SLi) were prepared by dispersing them within the chitosan-starch polymer matrix at a concentration of 0.025% w/v.The dispersion, totaling 4.55 × 10 −3 L m −2 , was poured into a polyethylene container.Ultrasonic agitation was employed for 120 s, accompanied by mechanical vibration.The resulting dispersion was subsequently air-dried at room temperature in polystyrene molds within a fume hood for a duration of 24 h.Following this process, the samples underwent characterization.An overview of the complete fabrication procedure is provided in Figure 1.
byproducts.Finally, the obtained lithiated nanotubes were carefully collected and preserved in a hermetically sealed glass jar to prevent any unintended interactions with the surrounding environment.In contrast, the synthesis of sonicated-lithiated nanotubes (SLi) involved a different approach.The same precursor materials, with the same weight proportions, were used.However, instead of refluxing, the synthesis took place under sonication conditions.The precursor mixture was placed in a flask equipped with an ultrasonic homogenizer.The mixture was subjected to ultrasonic waves at room temperature for a duration of 2 h.The sonication process facilitated the incorporation of lithium ions into the nanotube structure.Similar to the reflux method, the resulting sonicated-lithiated nanotubes were subjected to thorough purification through vacuum washing with deionized water.The purification process was repeated until the pH of the solution reached neutrality, ensuring the removal of any residues.The purified nanotubes were then securely stored in an airtight glass jar to prevent any degradation or contamination.These distinct synthesis approaches, reflux and sonication, provided two sets of lithiated nanotubes (RLi and SLi) with controlled properties.These synthesized materials held promise for various applications due to their unique structural and electrochemical characteristics.
The polymer electrolyte was prepared with a mixture of chitosan-starch solutions.According to previous research [15,[49][50][51], the chitosan (Ch) solution (2% m/v) was prepared by dispersing chitosan in an acetic acid aqueous solution (1% v/v) and then stirring at 100 rpm.After the chitosan was completely dispersed, the solution was sonicated for 15 min to break any air bubbles present and then allowed stand for 1 h.The starch (St) solution (2% m/v) was prepared dispersing starch in glycerin and heating above its gelatinization temperature (90 ± 2 °C) [52], continuously stirring at 100 rpm for 20 min.The solution was then cooled down to room temperature by ceasing heat and stirring.Finally, the chitosan-starch films (ChSt) were obtained mixing equal volumes of chitosan and starch solutions and then sonicating to homogenize.
The complete membranes for each batch of MWNTs (pristine MWNTs, MWOHs, RLi and SLi) were prepared by dispersing them within the chitosan-starch polymer matrix at a concentration of 0.025% w/v.The dispersion, totaling 4.55 × 10 −3 L m −2 , was poured into a polyethylene container.Ultrasonic agitation was employed for 120 s, accompanied by mechanical vibration.The resulting dispersion was subsequently air-dried at room temperature in polystyrene molds within a fume hood for a duration of 24 h.Following this process, the samples underwent characterization.An overview of the complete fabrication procedure is provided in Figure 1.

Characterization
For the characterization of MWNTs, MWOHs, RLi and SLi, the analyses of X-ray photoelectron spectroscopy (XPS) were performed with a JEOL JPS-9200 Photoelectron Spectrometer (ESCA) (Tokyo, Japan).Infrared (FTIR) spectra were recorded using a Vector 33 Bruker spectrophotometer (Billerica, MA, USA) at 32 scans, with a resolution of 4 cm −1 .Solid samples were embedded in KBr disks.Raman spectra were recorded with a Dilor LabRAM Micro-Raman with a resolution of 0.5 cm −1 , with a 514.5 nm laser with 15 s of integration time.High-resolution transmission electron microscopy (HRTEM) of MWNTs and MWOHs was performed with a JEOL-JEM 2010FEG instrument (Tokyo, Japan); on the other hand, micrographs of RLi and SLi were recorded with a JEOL-JEM 2200FS instrument (Tokyo, Japan).All the samples for microscopy were prepared by depositing droplets of a suspension of the studied materials on 100-mesh microscope gold grids.
All the chitosan-starch examples were characterized via dynamic mechanical analysis (DMA) with the objective of analyzing the thermo-mechanical properties of the polymer electrolyte.Dynamic mechanical analysis (DMA) was recorded using an RSA III of TA instruments (New Castle, DE, USA).Measurements were run using tweezers to tension in a temperature range from 30 • C to 350 • C, with a frequency of 1.0 Hz.Scanning electronic microscopy (SEM) was recorded in a SEM JEOL 5200 (Tokyo, Japan) with 5.0 of resolution to 25 kV.Infrared (FTIR) spectra were recorded in a Vector 33 Bruker spectrophotometer (Billerica, MA, USA) at 32 scans, with a resolution of 4 cm −1 .
The electrochemical measurements were performed using an EG&G PAR VersaSTAT 3 Potentiostat/Galvanostat (Princeton, NJ, USA).Cyclic voltammetry studies were performed from −2.5 to 2.5 V, at a scan rate of 50 mV/s.Electrochemical impedance spectroscopy (EIS) experiments were carried out in potentiostatic mode in the 1 MHz to 1 Hz frequency range.The impedance spectra were registered with a logarithmic data collection scheme at 10 steps per decade at open-circuit potential with a small signal amplitude of 10 mV.

Results and Discussion
The results obtained are presented below, separating them into two sections.First, the results of the characterization of the different reinforcement used are shown, such as carbon raw and oxidized nanotubes (MWNTs and MWOHs), as well as lithium-modified ones (RLi and SLi).Subsequently, the characterization of the chitosan-starch (ChSt) membranes with the different kinds of nanotubes is shown.

Characterization of Carbon Nanotubes via FTIR
Figure 2 shows the normalized spectra of the MWNTs after the different treatments.It is possible to see the characteristic FTIR peaks of the carbon nanotubes [53][54][55][56][57], which describe the normal modes of vibration at ~1632 cm −1 for E 1u and ~800 cm −1 for A 2u in multiwall carbon nanotubes (MWNTs) in all cases.Additional bands can be identified in oxidized carbon nanotubes (MWOHs, RLi, SLi) between 1750 and 1550 cm −1 for v(C=O), 1466 cm −1 for δ(O-H) and between 1300 and 950 cm −1 for v(C-O) due to vibrations of the carboxyl groups and at 3443 cm −1 for isolated surfaces of v s (OH) [54,[57][58][59].The presence of lithium in the samples could be evidenced through the interaction of Li with oxygen and carbon.According to some authors [60][61][62][63][64][65], the molecular vibration bands suggest the presence of ROCO 2 Li, Li 2 O, Li 2 CO 3 , ROLi and LiOH.O-H bands are visible at 3700-3100 cm −1 , being broader in the lithium-modified samples than in the pristine MWOHs.st(C-H) at 3000-2800 cm −1 and st(C-O) at 1064 cm −1 correspond to ROCO 2 Li and ROLi, respectively.st(Li-O) of the Li 2 O is visible at 530 and 476 cm −1 .Additionally, the C-O band at 880 cm −1 may originate from Li 2 CO 3 .Finally, the bending vibrations CH 2 at 1466 cm −1 and C=O symmetric at 1628 cm −1 and asymmetric at 1367 cm −1 vibrations indicate the presence of ROCO 2 Li.These results indicate that lithium ions are interacting with carbon nanotubes at these normal modes of vibration, thus promoting their use as electrodes in lithium batteries [66].cm −1 may originate from Li2CO3.Finally, the bending vibrations CH2 at 1466 cm −1 and C=O symmetric at 1628 cm −1 and asymmetric at 1367 cm −1 vibrations indicate the presence of ROCO2Li.These results indicate that lithium ions are interacting with carbon nanotubes at these normal modes of vibration, thus promoting their use as electrodes in lithium batteries [66].

Characterization of Carbon Nanotubes via Raman Spectroscopy
Raman spectra of MWNTs, MWOHs and lithium-modified multiwall carbon nanotubes (RLi and SLi) are shown in Figure 3.For some authors, Raman spectroscopy is considered the fingerprint of carbon nanotubes [67][68][69][70][71][72][73], allowing one to produce a deep analysis of the oxidation, purification and modification of the structure of their outer walls.The most prominent peaks of the spectra at ~1570 cm −1 (G band), ~1340 cm −1 (D band) and ~2684 cm −1 (G' band) and their overtones found as small peaks at 1233 and 2898 cm −1 are all attributed to carbon [59,71,73-82] and give a good indication of the state of CNTs.In some publications [83,84], the G band (the name G comes from "Graphite") is used as an indicator of purity since the CNTs should ideally present a graphitic order with carbon in sp 2 hybridization.When broken bonds are present, carbon can hybridize sp 3 , giving rise to the D band (the name D comes from "Defect").Thus, this band can be used to identify if there are dangling bonds or impurities on the surface of the CNTs [85,86].The ratio between the intensities of these two bands gives a hint of the purity of carbon nanotubes, but it says nothing of the kind of material (graphite, fullerene, CNTs); for this last thing, the FWHM of the peaks should also be considered (e.g., the FWHM of graphite is usually larger than that of CNTs) [86].In this work, to consider the width of the peaks, the area of the peaks was used instead of the intensity.The areas of the different peaks are tabulated

Characterization of Carbon Nanotubes via Raman Spectroscopy
Raman spectra of MWNTs, MWOHs and lithium-modified multiwall carbon nanotubes (RLi and SLi) are shown in Figure 3.For some authors, Raman spectroscopy is considered the fingerprint of carbon nanotubes [67][68][69][70][71][72][73], allowing one to produce a deep analysis of the oxidation, purification and modification of the structure of their outer walls.The most prominent peaks of the spectra at ~1570 cm −1 (G band), ~1340 cm −1 (D band) and ~2684 cm −1 (G' band) and their overtones found as small peaks at 1233 and 2898 cm −1 are all attributed to carbon [59,71,73-82] and give a good indication of the state of CNTs.In some publications [83,84], the G band (the name G comes from "Graphite") is used as an indicator of purity since the CNTs should ideally present a graphitic order with carbon in sp 2 hybridization.When broken bonds are present, carbon can hybridize sp 3 , giving rise to the D band (the name D comes from "Defect").Thus, this band can be used to identify if there are dangling bonds or impurities on the surface of the CNTs [85,86].The ratio between the intensities of these two bands gives a hint of the purity of carbon nanotubes, but it says nothing of the kind of material (graphite, fullerene, CNTs); for this last thing, the FWHM of the peaks should also be considered (e.g., the FWHM of graphite is usually larger than that of CNTs) [86].In this work, to consider the width of the peaks, the area of the peaks was used instead of the intensity.The areas of the different peaks are tabulated in Table 1, for the different samples.The G' band is related to the G band, whose intensity is proportional to the purity of the multiwall carbon nanotubes due to the absence of nanocarbons (disordered phase) in the samples [87].The ratio between areas of the different peaks (G'/G, G'/D and D/G) was calculated, as shown in Table 1.The values suggest that the oxidation of MWNTs cured defects, since the ratio D/G decreases and ratio G'/D increases.The small difference in G'/G values could be due to the removal of carbonaceous material during oxidation [88].
in Table 1, for the different samples.The G' band is related to the G band, whose intensity is proportional to the purity of the multiwall carbon nanotubes due to the absence of nanocarbons (disordered phase) in the samples [87].The ratio between areas of the different peaks (G'/G, G'/D and D/G) was calculated, as shown in Table 1.The values suggest that the oxidation of MWNTs cured defects, since the ratio D/G decreases and ratio G'/D increases.The small difference in G'/G values could be due to the removal of carbonaceous material during oxidation [88].The samples intercalated with lithium show a drop in G'/D compared to MWOHs due to the decrement in the area of G' caused by the exfoliating action of LiOH, which degrades carbonaceous materials [89,90].The exfoliated surface of MWNTs is more reactive [91], which is positive for capturing Li-ions.Moreover, Li-ions can be intercalated in MWCNT's inner walls because of coulombic repulsion effects [92].A significant decrement in the D/G ratio in comparison with the MWNTs is observed in the lithium-modified samples, in accordance with various publications [92][93][94][95][96][97][98][99], indicating a possible exfoliation of carbon nanotubes.It is also important to mention that the small peaks that appear in the spectra refer to normal vibration modes that are affected in position due to the oxidation of multiwall carbon nanotubes [100].Furthermore, the G'/G and G'/D are larger in RLi than in SLi, indicating that there is an increment in the G' band with respect to the  The samples intercalated with lithium show a drop in G'/D compared to MWOHs due to the decrement in the area of G' caused by the exfoliating action of LiOH, which degrades carbonaceous materials [89,90].The exfoliated surface of MWNTs is more reactive [91], which is positive for capturing Li-ions.Moreover, Li-ions can be intercalated in MWCNT's inner walls because of coulombic repulsion effects [92].A significant decrement in the D/G ratio in comparison with the MWNTs is observed in the lithium-modified samples, in accordance with various publications [92][93][94][95][96][97][98][99], indicating a possible exfoliation of carbon nanotubes.It is also important to mention that the small peaks that appear in the spectra C 2023, 9, 87 8 of 22 refer to normal vibration modes that are affected in position due to the oxidation of multiwall carbon nanotubes [100].Furthermore, the G'/G and G'/D are larger in RLi than in SLi, indicating that there is an increment in the G' band with respect to the other bands in this sample [101].The greater contribution of this band to the RLi spectrum could suggest a higher graphitic character of this sample, as this band is observed as an overtone of graphitic samples, and it is not affected by defects [102].The graphitic nature of the sample, in addition to allowing for Li ion intercalation, enhances electron transfer, promoting its use as an electrode in lithium batteries [103].SLi may have undergone a higher degree of oxidation during Li incorporation, as indicated with the higher proportion of the D band in this sample in comparison with RLi.This effect is common when oxidizing graphene [102].

Characterization of Carbon Nanotubes via XPS
XPS was employed to quantify the relative lithium content of RLi and SLi compared to carbon and oxygen, referencing MWNT and MWOH spectra for analysis.Carbon (C1s) exhibited binding energies within a range of 295 to 281 eV, oxygen (O1s) within 540 to 528 eV and lithium (Li1s) within 61 to 57 eV.The deconvolution of C1s, O1s and Li1s peaks was based on established data from previous XPS studies on lithium-modified multiwall carbon nanotubes [63,[104][105][106][107] In terms of the C1s peak, the hydroxyl (-OH at 286.6 eV) and carboxyl (-COOH at 288.6 eV) bands of lithium-modified multiwall carbon nanotubes were found to be more prominent compared to pure nanotubes.This phenomenon may be attributed to the presence of lithium carbonate.An increase in π-π* electronic transitions (at 290.9 eV) in MWOHs indicated potential CNT wall exfoliation due to acids [57].Conversely, lithiummodified nanotubes demonstrated reduced transitions, implying non-covalent interactions.
Regarding the O1s band, the intensity of carboxyl, carbonyl and carbonate bands were considered (-COOH at 531.2 eV, O=C at 532.1 eV, and C-O*-C=O at 534.3 eV).Comparing the spectra of RLi and SLi, one could observe that the contribution of C-O and C=O bands in SLi is much higher, indicating that the sample was oxidized due to the L incorporation method, as suggested by the Raman spectra.Nevertheless, both samples are well lithiated, as indicated by the presence of the Li 2 CO 3 band at 56 eV.On the other hand, analyzing the Li 1s line, it is possible to observe the presence of Li 0 and Li 2 O on lithited surfaces.Furthermore, strong electrostatic interactions between lithium ions and CNTs led to a decrease in the intensity of π-π* transitions (at 290.9 eV) [104] in lithiated samples with respect to the non-lithiated samples.The presence of lithium indicated the potential of these samples for facilitating electrode-to-electrode transfer in lithium batteries.Figure 4 (subscripts A and B) provides an illustrative deconvolution of spectra encompassing the aforementioned bands according to different reports [108,109].

Characterization of Carbon Nanotubes via HRTEM
The HRTEM micrograph in Figure 5, of the as-received MWNTs, shows the walls of a multiwall carbon nanotube and other allotropic forms of carbon as impurities.An FFT pattern (Fast Fourier Transform) was obtained from two zones: "1" is part of a multiwall carbon nanotube and "2" belongs to graphite.In both areas, the FFT shows the presence of fine dots (defined peaks in the line scan), confirming the presence of ordered carbon.On the other hand, the diffuse rings in the FFT patterns and the corresponding wide peaks in the line scans indicate sections of low crystallinity due to impurities, likely corresponding to low-order graphite originating from MWNTs [110,111].
The HRTEM micrograph of MWOHs in Figure 6 clearly shows defects on the walls of the nanotubes.By analyzing two areas of these MWOHs micrographs with FFT, both display more diffuse spots, and the line scan shows a rather poor crystallinity, revealed by the wide peaks and low intensity of the central spot.This is a characteristic of defective CNTs exhibiting distorted layers [59,112], confirming the oxidation of the multiwall carbon nanotubes.

Characterization of Carbon Nanotubes via HRTEM
The HRTEM micrograph in Figure 5, of the as-received MWNTs, shows the walls of a multiwall carbon nanotube and other allotropic forms of carbon as impurities.An FFT pattern (Fast Fourier Transform) was obtained from two zones: "1" is part of a multiwall carbon nanotube and "2" belongs to graphite.In both areas, the FFT shows the presence of fine dots (defined peaks in the line scan), confirming the presence of ordered carbon.On the other hand, the diffuse rings in the FFT patterns and the corresponding wide peaks in the line scans indicate sections of low crystallinity due to impurities, likely corresponding to low-order graphite originating from MWNTs [110,111].HRTEM micrographs and ED patterns of RLi and SLi are shown in Figure 7. From the micrographs, it is possible to see the presence of a higher amount of material covering the nanotubes in RLi than in SLi.This material must be Li 0 and Li 2 O, as confirmed via XPS.In SLi, Li becomes more internal, forming Li 2 CO 3 .Thus, ultrasound induces the insertion of Li.The crystal planes of the nanotubes are evident in both cases, indicating that the crystallinity of the material remains, even after the oxidation and lithiation processes.In a recent publication [113], it was confirmed that the crystallinity of MWNTs increases after oxidation.The diffraction patterns indicate that the material is polycrystalline, in accordance with the HRTEM micrographs, showing bending crystalline planes.The two inner rings in the ED patterns originate in MWNTs [101,114].The smallest one denotes the d-spacing of graphite oxide, which, in the present case, is 10.44 Å.The second ring corresponds to the d-spacing between (002) graphite planes, being, in the present case, 3.92 Å.The d values of outer diffuse rings (2.33, 1.91 and 1.38 Å) are in accordance with d values found for VAST-MWNTs [115], which, in XRD with Cu Ka, produce reflexes at 67.97 • , 47.65 • and 38.5 • 2θ, respectively.These last values are characteristic of platinum, being from its ( 111), ( 200) and (220) planes [116].Thus, the as-purchased VAST-MWNT contains Pt.Therefore, with this, we confirm that by analyzing the diffraction patterns of the HRTEM micrographs in the lithiated carbon nanotubes, we corroborate that they can be used as electrodes in lithium batteries, as has been previously reported [117].The HRTEM micrograph of MWOHs in Figure 6 clearly shows defects on the walls of the nanotubes.By analyzing two areas of these MWOHs micrographs with FFT, both display more diffuse spots, and the line scan shows a rather poor crystallinity, revealed by the wide peaks and low intensity of the central spot.This is a characteristic of defective CNTs exhibiting distorted layers [59,112], confirming the oxidation of the multiwall carbon nanotubes.HRTEM micrographs and ED patterns of RLi and SLi are shown in Figure 7. From the micrographs, it is possible to see the presence of a higher amount of material covering the nanotubes in RLi than in SLi.This material must be Li 0 and Li2O, as confirmed via XPS.In SLi, Li becomes more internal, forming Li2CO3.Thus, ultrasound induces the insertion of Li.The crystal planes of the nanotubes are evident in both cases, indicating that the crystallinity of the material remains, even after the oxidation and lithiation processes.In a recent publication [113], it was confirmed that the crystallinity of MWNTs increases after oxidation.The diffraction patterns indicate that the material is polycrystalline, in accord-   The HRTEM micrograph of MWOHs in Figure 6 clearly shows defects on the walls of the nanotubes.By analyzing two areas of these MWOHs micrographs with FFT, both display more diffuse spots, and the line scan shows a rather poor crystallinity, revealed by the wide peaks and low intensity of the central spot.This is a characteristic of defective CNTs exhibiting distorted layers [59,112], confirming the oxidation of the multiwall carbon nanotubes.HRTEM micrographs and ED patterns of RLi and SLi are shown in Figure 7. From the micrographs, it is possible to see the presence of a higher amount of material covering the nanotubes in RLi than in SLi.This material must be Li 0 and Li2O, as confirmed via XPS.In SLi, Li becomes more internal, forming Li2CO3.Thus, ultrasound induces the insertion of Li.The crystal planes of the nanotubes are evident in both cases, indicating that the crystallinity of the material remains, even after the oxidation and lithiation processes.In a recent publication [113], it was confirmed that the crystallinity of MWNTs increases after oxidation.The diffraction patterns indicate that the material is polycrystalline, in accordance with the HRTEM micrographs, showing bending crystalline planes.The two inner rings in the ED patterns originate in MWNTs [101,114].The smallest one denotes the dspacing of graphite oxide, which, in the present case, is 10.44 Å .The second ring corresponds to the d-spacing between (002) graphite planes, being, in the present case, 3.92 Å .

Characterization of the Chitosan-Starch (ChSt) Membranes via DMA
To test the mechanical properties of the polymeric matrix and how this property is improved by introducing multiwall carbon nanotubes, the membranes were characterized via dynamic mechanical analysis (DMA).Only the membranes with the lithiated nanotubes were omitted, due to the predominant amount of oxide groups linked to lithium, as confirmed via XPS characterization.
Dynamic mechanical analysis (DMA) was used to characterize the electrolyte ChSt and the nanocomposites of this electrolyte with MWNTs and MWOHs, with the purpose of giving an idea about the mechanical stability of the composites, away from the lithiation of the carbon nanotubes.DMA is a thermo-mechanical test where viscoelastic data of the materials under test are collected at different temperatures [118,119].Figure 8 shows the values of the modulus E' at different temperatures.It is possible to observe that for MWOHs, E' is around 445% the value of the other samples at 35 The d values of outer diffuse rings (2.33, 1.91 and 1.38 Å ) are in accordance with d values found for VAST-MWNTs [115], which, in XRD with Cu Ka, produce reflexes at 67.97°, 47.65° and 38.5° 2θ, respectively.These last values are characteristic of platinum, being from its ( 111), ( 200) and ( 220) planes [116].Thus, the as-purchased VAST-MWNT contains Pt.Therefore, with this, we confirm that by analyzing the diffraction patterns of the HRTEM micrographs in the lithiated carbon nanotubes, we corroborate that they can be used as electrodes in lithium batteries, as has been previously reported [117].

Characterization of the Chitosan-Starch (ChSt) Membranes via DMA
To test the mechanical properties of the polymeric matrix and how this property is improved by introducing multiwall carbon nanotubes, the membranes were characterized via dynamic mechanical analysis (DMA).Only the membranes with the lithiated nanotubes were omitted, due to the predominant amount of oxide groups linked to lithium, as confirmed via XPS characterization.
Dynamic mechanical analysis (DMA) was used to characterize the electrolyte ChSt and the nanocomposites of this electrolyte with MWNTs and MWOHs, with the purpose of giving an idea about the mechanical stability of the composites, away from the lithiation of the carbon nanotubes.DMA is a thermo-mechanical test where viscoelastic data of the materials under test are collected at different temperatures [118,119].Figure 8 shows the values of the modulus E' at different temperatures.It is possible to observe that for MWOHs, E' is around 445% the value of the other samples at 35 °C.Even at 200 °C, the E' of this sample practically doubles that of chitosan-starch and is 1.6-times that of MWNTs.The fact that the storage modulus of the membrane increases as the temperature rises, up to the critical point of 250 °C, for ChSt and the membrane with MWNTs, is because there is an increment in crystallinity of the films produced by the arrangement of the starch molecules and because the interactions between the OH-groups of the glucose The fact that the storage modulus of the membrane increases as the temperature rises, up to the critical point of 250 • C, for ChSt and the membrane with MWNTs, is because there is an increment in crystallinity of the films produced by the arrangement of the starch molecules and because the interactions between the OH-groups of the glucose units and the H+ groups of the water molecules minimize [120].The storage modulus increases linearly from 35 to 150 • C, but the slope of the curve is higher for MWNTs than for ChSt.Then, from 150 to 250 • C, there is a steady state for both samples, indicating that a maximum ordering is reached.On the other hand, when oxidized carbon nanotubes (MWOHs) are incorporated into the chitosan-starch film, a phenomenon similar to adding a plasticizer to the membrane is observed.As the temperature increases, the thermal energy begins to overcome these intermolecular forces.The polymer chains start to experience more vibrational motion and increased molecular mobility.This phenomenon is often referred to as "softening" or "rubbery behavior".As a consequence, the material becomes less rigid and more compliant, leading to a decrease in the storage modulus [121].
Through the relationship between the storage and loss modulus, the internal friction of materials be determined, known as Tan (δ).High internal friction is reflected as high Tan (δ) values.Depending on the degree of adhesion between the phases, the internal energy is dissipated at the interface.Enhanced adhesion results in increased friction, as evidenced by a Tan (δ) value of 0.8 at 35 • C for the sample with MWOHs, while the other two samples exhibit a value of 0.55.Thus, oxidizing the MWNTs greatly enhances the adhesion properties between phases, probably due to -OH groups that promote hydrogen bonds.Stability in their mechanical properties gives us the confidence to use these materials in lithium batteries, as used in other reported works [122].Sample MWOH was considered a model for RLi and SLi, taking into account that the oxygen content and the surface characteristics of these samples are similar, as evidenced through FTIR and Raman spectroscopies and XPS.

Characterization of the Chitosan-Starch (ChSt) Membranes via FTIR
For the nanocomposites of carbon nanotubes in a chitosan-starch matrix, when we compare the different FTIR spectra, they are very similar between the ChSt with the different modifications of multiwall carbon nanotubes: raw (MWNTs), oxidized (MWOHs) and lithium-modified (CRLi and CSLi).For this reason, we only show the FTIR spectra of starch, chitosan and their mixture in Figure 9, as shown in other reports, where the reinforcement of nanostructures in polymeric matrices is practically imperceptible [123].We can remember that they have a reinforcement in the polymer matrix of chitosan-starch at 0.025% w/v.For starch, a band at 3274 cm −1 of st(O-H) of glycosidic chains is present.Also characteristic of starch, peaks at 2921 and 2926 cm −1 are present.These are st(CH) vibrations associated with the methine ring hydrogen atoms [124][125][126][127].For chitosan, the peak at 3496 cm −1 is due to the OH group (ν OH ), the one at 3345 cm −1 is due to NH group-stretching vibration (ν NH ) and the ones at 2926, 2873, 1421, 1322 and 1249 cm −1 are due to symmetric or asymmetric CH 2 stretching vibrations attributed to pyranose rings (ν CH ).The peak at 1646 cm −1 is due to C=O in amide groups (amide I band).The one at 1593 cm −1 is due to NH 2 bending vibration in an amino group (δ NH2 ).The peak at 1421 and 1322 cm −1 is due to vibrations of OH in a CH ring, the one at 1381 cm −1 is due to CH 3 in amide group, the one at 1249 cm −1 is due to C-O group, the one at 1156 cm −1 is due to -C-O-C-in glycosidic linkage, the ones at 1096 and 1030 cm −1 are due to C-O groups (ν C -O ) in amides and the one at 897 cm −1 is due to the CH 3 COH group [128][129][130].A linear combination of the spectra of starch, chitosan and MWNTs is observed in the composites, depending on their composition.Although these reinforcements could be imperceptible in this FTIR characterization, their use as electrolytes [123] or electrodes [131] for their proposal in lithium batteries has been reported.the one at 1249 cm −1 is due to C-O group, the one at 1156 cm −1 is due to -C-O-C-in glycosidic linkage, the ones at 1096 and 1030 cm −1 are due to C-O groups (νC-O) in amides and the one at 897 cm −1 is due to the CH3COH group [128][129][130].A linear combination of the spectra of starch, chitosan and MWNTs is observed in the composites, depending on their composition.Although these reinforcements could be imperceptible in this FTIR characterization, their use as electrolytes [123] or electrodes [131] for their proposal in lithium batteries has been reported.Regarding the absence of SEM images for the lithiated carbon nanotubes (RLi and SLi), there is a specific rationale for this omission.Lithiated nanotubes were not included in the SEM characterization due to their relatively low percentage of reinforcement in the composite.Given the minute amount of lithium-modified nanotubes, any observable morphological differences would likely be negligible and difficult to distinguish from the background matrix.Therefore, the decision was made to focus the SEM analysis on the most significant reinforcements, MWNTs and MWOHs, to provide clearer insights into the composite structure.

Characterization of the Chitosan-Starch (ChSt) Membranes via SEM
Additionally, it is important to note that the organic nature of the samples required a thin gold coating prior to SEM imaging.This coating was applied to enhance the conductivity of the samples and minimize the risk of charging effects, ensuring accurate and reliable imaging under the electron beam.
SLi), there is a specific rationale for this omission.Lithiated nanotubes were not included in the SEM characterization due to their relatively low percentage of reinforcement in the composite.Given the minute amount of lithium-modified nanotubes, any observable morphological differences would likely be negligible and difficult to distinguish from the background matrix.Therefore, the decision was made to focus the SEM analysis on the most significant reinforcements, MWNTs and MWOHs, to provide clearer insights into the composite structure.Additionally, it is important to note that the organic nature of the samples required a thin gold coating prior to SEM imaging.This coating was applied to enhance the conductivity of the samples and minimize the risk of charging effects, ensuring accurate and reliable imaging under the electron beam.Figure 11 depicts the assembly of the device for electrochemical characterization, which consists of two aluminum electrodes sandwiching the chitosan-starch (ChSt) membrane of 1 × 1 cm.The cell was hot sealed with polyethylene terephthalate.The tests were performed with a membrane of reflux-lithiated nanotubes (RLi), considering that these nanotubes are less oxidized than the SLi and may enhance electron transfer when used electrochemically.A membrane of oxidized carbon nanotubes (MWOHs) was also measured as a reference.Figure 12 shows cyclic voltammetry curves of using the assembly from Figure 11 and membranes of MWOHs and RLi.The third cycle is shown for both samples.As can be seen, the voltammograms are relatively symmetrical, as expected when using a symmetric electrochemical cell.Both the membranes with MOWHs and RLi present pairs of oxidation and reduction peaks close to 0 V.The difference in potential either between oxidation or oxidation peaks is of 0.84 V for MWOHs and 0.75 V for RLi.The peaks of the RLi sample are just shifted, having a lower potential difference between oxidation and reduction peaks (the processes are facilitated).This could be related to a lower ohmic drop.In fact, the parallel resistance, most probably related to the charge transport in the electrolyte, is Figure 12 shows cyclic voltammetry curves of using the assembly from Figure 11 and membranes of MWOHs and RLi.The third cycle is shown for both samples.As can be seen, the voltammograms are relatively symmetrical, as expected when using a symmetric electrochemical cell.Both the membranes with MOWHs and RLi present pairs of oxidation and reduction peaks close to 0 V.The difference in potential either between oxidation or oxidation peaks is of 0.84 V for MWOHs and 0.75 V for RLi.The peaks of the RLi sample are just shifted, having a lower potential difference between oxidation and reduction peaks (the processes are facilitated).This could be related to a lower ohmic drop.In fact, the parallel resistance, most probably related to the charge transport in the electrolyte, is lower in the sample with lithium.This can be denoted in the impedance spectra of cells with different membranes (See Figure 13).As can be observed, the diameter of the semicircle is smaller for the RLi sample.This is associated with parallel resistance.Then, the conductivity of the membrane is enhanced due to the presence of Li in the carbon nanotubes.Figure 12 shows cyclic voltammetry curves of using the assembly from Figure 11 and membranes of MWOHs and RLi.The third cycle is shown for both samples.As can be seen, the voltammograms are relatively symmetrical, as expected when using a symmetric electrochemical cell.Both the membranes with MOWHs and RLi present pairs of oxidation and reduction peaks close to 0 V.The difference in potential either between oxidation or oxidation peaks is of 0.84 V for MWOHs and 0.75 V for RLi.The peaks of the RLi sample are just shifted, having a lower potential difference between oxidation and reduction peaks (the processes are facilitated).This could be related to a lower ohmic drop.In fact, the parallel resistance, most probably related to the charge transport in the electrolyte, is lower in the sample with lithium.This can be denoted in the impedance spectra of cells with different membranes (See Figure 13).As can be observed, the diameter of the semicircle is smaller for the RLi sample.This is associated with parallel resistance.Then, the conductivity of the membrane is enhanced due to the presence of Li in the carbon nanotubes.The pairs of peaks close to 0 V may be related to the oxidation and dissolution of aluminum.It has been reported that the Al deposition and dissolution occur at about 0.35 V when using a three-Al electrode electrochemical cell and an Al ion-conducting electrolyte [102].On the other hand, double oxidation and reduction peaks with a potential dif- The pairs of peaks close to 0 V may be related to the oxidation and dissolution of aluminum.It has been reported that the Al deposition and dissolution occur at about 0.35 V when using a three-Al electrode electrochemical cell and an Al ion-conducting electrolyte [102].On the other hand, double oxidation and reduction peaks with a potential difference of about 0.4 V have been observed using an Al 2 O 3 working electrode in aqueous solution [132].The voltammetry peaks shifted depending on the pH of the solution.The peaks most probably appeared due to the oxidation and dissolution of Al in alkaline and acidic solutions.The findings of the past two reports could be supported by the theory of Al electrochemical etching.Depending on the conditions of the electrolyte (conductivity, pH, viscosity, etc.), it is possible to change the potential of the electrochemical cell, producing either oxidation or dissolution of Al and aluminum oxides, or a combination of both processes [133].

Electrochemical Characterization of the Chitosan-Starch (ChSt) Membranes
Additionally, the RLi sample presents an oxidation-reduction couple at higher potentials (indicated with dashed arrows in Figure 12b).These peaks appear at ±1.34 V.These peaks may related to the alloying/de-alloying process of Li with Al.Using nanostructured Al as an electrode in a Li ion battery, it has been possible to have Li storage through the formation of Li-Al alloys.The difference in potential between Al and Li-Al alloys has been reported as 0.5 V in water-free Li ion-conducting electrolytes [134].In the present case, the difference in potential is larger; however, it is well known that the nature of the electrolyte could have an effect on the potential of the battery.For example, when using a water-containing polymer electrolyte of polymethylmetacrylate in a Li ion battery, it was possible to change the voltage of the battery from 3 to 3.86 V when the electrolyte was UV-cured [135].The fact that Al-Li alloying occurs is an indication that the lithiated carbon nanotubes could work as a source of Li ions and that these ions could be transported through the chitosan-starch matrix.

Conclusions
This comprehensive analysis involving XPS, FTIR and Raman spectroscopies and HRTEM effectively validated the successful oxidation and lithium modification of multiwalled carbon nanotubes.Notably, the presence of hydroxyl and carbonyl groups within oxidized multi-walled carbon nanotubes facilitates the binding of lithium ions originating from LiOH.Particularly, the carbon nanotubes lithiated with the reflux methodology exhibit good Li incorporation while oxidation is minimized, maintaining the graphitic character.Such nanotubes are optimal for Li ion exchange and electron transfer.
Furthermore, the discernible enhancement in mechanical strength, as evidenced by the storage modulus measured through dynamic mechanical analysis (DMA), is particularly pronounced in the polymeric membranes with MWOHs compared to the ones with MWNTs.Thus, some oxidation of the carbon nanotubes is positive.This enhancement is reflected in the dispersion of carbon nanostructures within the biopolymer matrix, as observed in SEM micrographs.
Electrochemical characterization in symmetrical cells with Al current collectors indicates an enhancement in ionic transport through the membranes when the carbon nanotubes are lithiated.On the other hand, Al-Li alloying was inferred, indicating that the lithiated carbon nanotubes could work as a source of Li ions and that these ions could be transported through the chitosan-starch matrix.

Figure 1 .
Figure 1.Schematic of the fabrication process of the different modifications to the carbon nanotubes and the synthesis of the chitosan-starch matrix and the different nanocomposites.

C
2023, 9, x FOR PEER REVIEW 10 of 22

C
2023, 9, x FOR PEER REVIEW 10 of 22

Figure 8 .
Figure 8. Storage modulus (E') at different temperatures via dynamic mechanical analysis (DMA).The dashed lines and arrows indicate reference temperatures used for the discussion.

3. 7 .
Characterization of the Chitosan-Starch (ChSt) Membranes via SEMThe SEM micrographs in Figure10display the surface (S) and cross-sectional (C) morphologies of the ChSt, MWNTs and MWOHs nanocomposite samples.The ChSt sample exhibits a smooth and continuous surface in both morphologies, indicating a homogeneous structure without observable fissures.In contrast, the samples reinforced with raw carbon nanotubes (MWNTs) and oxidized carbon nanotubes (MWOHs) display distinct slit formations in the cross-sectional views, indicative of the incorporation of nanotubes into the membrane matrix.

3. 8 .
Figure 11  depicts the assembly of the device for electrochemical characterization, which consists of two aluminum electrodes sandwiching the chitosan-starch (ChSt) membrane of 1 × 1 cm.The cell was hot sealed with polyethylene terephthalate.The tests were performed with a membrane of reflux-lithiated nanotubes (RLi), considering that these nanotubes are less oxidized than the SLi and may enhance electron transfer when used electrochemically.A membrane of oxidized carbon nanotubes (MWOHs) was also measured as a reference.

Figure 12 .
Figure 12.Cyclic voltammograms of membranes of (a) MWOHs and (b) RLi.The assembly of Figure 11 was used for the experiments.The red continuous arrows indicate processes of aluminum oxidation and dissolution, while the dashed arrows indicate processes of Li-Al alloying/de-alloying.

Figure 12 . 22 Figure 13 .
Figure 12.Cyclic voltammograms of membranes of (a) MWOHs and (b) RLi.The assembly of Figure 11 was used for the experiments.The red continuous arrows indicate processes of aluminum oxidation and dissolution, while the dashed arrows indicate processes of Li-Al alloying/de-alloying. C 2023, 9, x FOR PEER REVIEW 16 of 22

Figure 13 .
Figure 13.Impedance spectroscopy spectra of the assembly of Figure 11 using membranes of MWOHs and RLi.

Funding:
Y.E.-M., thanks the economic support of Consejo de Ciencia y Tecnología del Estado de Puebla (CONCyTEP) and Tecnológico Nacional de México through the 2021 Call for Projects of Technological Development and Innovation (Project: Caracterización y Ensamble de Baterías Li-S con Nanotubos de Carbono de Multipared Funcionalizados).On the other hand, E.Q.-G., thanks the economic support of Vicerrectoría de Investigación y Estudios de Posgrado of BUAP (VIEP-BUAP) through the project 00185-VIEP2023 and the National Council of Science and Technology (CONACyT) through the project of Frontiers of Science 21077 and the project PRONACES 316537.
• C for a duration of 2 h.During this period, chemical reactions took

Table 1 .
Areas of the Raman bands and ratios between them for the different modifications of MWNTs.
• C.Even at 200 • C, the E' of this sample practically doubles that of chitosan-starch and is 1.6-times that of MWNTs.