Nowadays, microbatteries are in demand as power source to drive small devices such as smartcards, medical implants, sensors, etc. To date, the electrochemical performances of these all-solid-state batteries are limited because planar thin films are employed as electrode and electrolyte materials. In general, the total thickness of the stacking films is below 15 μm, and the resulting battery reveals relatively low power and energy densities. In order to ensure significant advances for extended applications, it is crucial to improve the electrochemical performances by investigating new materials and manufacturing processes. In this context, the large specific area offered by nano-architectured electrodes represents a promising alternative to improve the general performances of these micro-power sources 1 .

Particularly, better rate capability, capacity, and cycling behavior have been observed for self-organized nanostructures such as titania nanotubes (TiO₂nt) 2 3 4 5 6 7 8 9 . However, when targeting 3D microbatteries, the conventional top-down approach to deposit solid electrolyte (e.g., lithium phosphorous oxynitride) 10 11 12 is not really suitable due to the accumulation of the electrolyte at the top of the nanotubes 13 . Certainly, with this accumulation of electrolyte, the 3D paradigm of microbatteries cannot be realized. Thus, investigating the deposition of polymer electrolytes into nanostructures by electrochemical techniques is a convenient way to ensure the desired filling of the nanostructures 14 . Indeed, electropolymerization is particularly powerful to control the deposition of different polymers into various porous materials 15 16 17 18 19 . Very recently, the use of electrodeposition to fill TiO₂nt with a layer of poly(methyl methacrylate)-polyethylene oxide, i.e., PMMA-(PEO)₄₇₅ has been reported 13 20 . We have demonstrated that this simple bottom-up approach is adequate to deposit a homogeneous copolymer layer into titania nanotubes while improving the electrochemical performance. However, conformal coating of the nanotubes by the polymer electrolyte is required to design a 3D microbattery. In this work, it is reported that the conformal deposition of a PMMA-PEO electrolyte into self-organized TiO₂nt can be obtained by controlling the electrodeposition parameters. The morphology and the chemical composition of the resulting copolymer-embedded TiO₂nt materials are characterized by scanning electron microscopy (SEM) and transmission electron microscopy (TEM). The incorporation of lithium bis(trifluoromethanesulfone)imide, so-called LiTFSI salt, into the electrodeposited polymer is studied by X-ray photoelectron spectroscopy (XPS).

Synthesis of self-organized TiO₂nt has been widely reported for a wide range of applications 21 22 23 24 . In the present work, TiO₂nt layers were produced by the electrochemical anodization of Ti foils using the Modulab potentiostat from Solartron Analytical (Hampshire, UK). Before the anodization, Ti foils with the 99.6+ % purity and 0.125-mm thickness were cut into pieces with the desired dimensions and sonicated sequentially during 10 min in acetone, propanol, and methanol sequentially. After that, the foils were rinsed with deionized water and dried with compressed air. The anodization process was carried out in an electrochemical cell containing a solution of 1 M H₃PO₄, 1 M NaOH, and 0.4 wt.% of HF. The setup consisted of Ti foil as the working electrode, a Pt grid as the counter electrode, and a Hg/Hg₂SO₄, K₂SO₄ (saturated) (E=0.64V vs NHE) reference electrode. A constant voltage of 20 V was applied during 2 h. The material was rinsed with deionized water and dried with compressed air immediately after the anodization process.

Then, an aqueous electrolyte containing 0.035 M LiTFSI was introduced into the cell and purged with N₂ gas for 10 min before adding 4 g of the MMA-(PEO)₄₇₅ monomer provided by Sigma Aldrich (St. Louis, MO, USA). It can be noticed that no initiator was added into the solution. The copolymer-embedded TiO₂nt was obtained by cyclic voltammetry (CV) using the as-prepared TiO₂nt layers as the working electrode and a Pt grid as the counter electrode. The CV curves were carried out in the potential window ranging from −0.4 to −2.5 V vs Hg/Hg₂SO₄, K₂SO₄ (saturated) with the scan rate of 25 mV/s. The number of cycles was varied from 1 to 10 in order to observe the influence of cycle number on the polymer electrolyte layers. After electropolymerization, the samples were dried at room temperature to evaporate part of the residual water. The morphology of the copolymer-embedded TiO₂nt was studied by SEM and TEM analyses using a JEOL 6320 F SEM and a JEOL 2010 F TEM (JEOL Ltd., Akishima, Tokyo, Japan). XPS measurements were carried out with a Kratos Axis Ultra spectrometer (Kratos Analytical Ltd., Manchester, UK), using focused monochromated Al Kα radiation (hν = 1,486.6 eV). The XPS spectrometer was directly connected to an argon dry box through a transfer chamber to avoid moisture/air exposure of the samples. The analyzed area of each sample was 300 μm × 700 μm. Peaks were recorded with constant pass energy of 20 eV. The pressure in the analysis chamber was around 5 × 10⁻⁸ Pa. Short acquisition time spectra were recorded before and after each normal experiment to check that the samples did not suffer from degradation under the X-ray beam during measurements. Peak assignments were made with respect to experimental reference compounds, namely bulk anatase and/or rutile TiO₂. The binding energy scale was calibrated from hydrocarbon contamination using the C 1 s peak at 285.0 eV. Core peaks were analyzed using a non-linear Shirley-type background. The peak positions and areas were optimized by a weighted least-square fitting method using 70% Gaussian and 30% Lorentzian line shapes. Quantification was performed on the basis of Scofield's relative sensitivity factors.

The electropolymerization of MMA-(PEO)₄₇₅ monomer was carried out by cyclic voltammetry (CV) given in (Figure 1a). For comparison, the CV curves recorded in monomer-free electrolyte are given in Figure 1b. It is important to note that the monomer is not electrochemically active, and the polymerization process is expected to be initiated by free hydrogen radicals produced at applied cathodic potentials higher than −1 V vs Hg/Hg₂SO₄, K₂SO₄ (saturated) 13 20 25 . The similar shape of the CVs recorded in the two different electrolytes can be explained by the reactions involving the Ti⁴⁺/Ti³⁺ and H⁺/H₂ redox couples. However, variations of the current densities obtained in the monomer-free electrolyte are almost independent of cycle number, suggesting that all electrochemical reactions are reversible without modification of the surface. In contrast, the cathodic current density decreases slightly with cycle number when the monomer is present. This effect on current density can be attributed to the thin copolymer layers successively deposited onto the titania nanotube walls and, consequently, acting as electrical insulators.

The SEM images of the as-formed TiO₂nt and the copolymer-embedded TiO₂nt with different numbers of CV cycles are displayed in Figure 2. The evolution in thickness of the copolymer layer can clearly be observed. The SEM image of the as-formed TiO₂nt shown in Figure 2a reveals very thin tube walls (around 9 nm) and presents some degree of perforation at the top. As the cycles of CV increase from 0 to 1, 5, and up to 10 according to Figure 2b,c,d, the average thickness of the nanotube walls increases from 9 to 13, 17, and 18 nm (see Table 1), respectively, while the inner diameters of the tubes and spaces between the tubes decrease from 90 to 84, 72 and 70 nm (Table 1), respectively. It can also be noticed that the perforation in the tube walls has clearly disappeared. These results confirm that the growth of a copolymer layer occurred on the inside and outside of each nanotube wall.

The deposition of the copolymer was further confirmed by examination of the cross-sectional SEM images given in Figure 3. Although the thin layer of copolymer is not obvious in cross-section after the first cycle (Figure 3a), it can easily be discerned along the nanotubes after 10 cycles of CV (Figure 3b). The inter-tube spaces are filled with the copolymer, and the nanotube walls become significantly thicker as compared to the sample obtained after 1 cycle of CV. It can also be noticed in these cross-sectional images (Figure 3a,b) that the perforations at the top of the tube are no longer visible owing to the filling and covering with the copolymer.

In order to verify the conformal deposition of the copolymer, TEM imaging was performed on a single nanotube. The TEM images of the TiO₂nt after 5 cycles of CV are displayed in Figure 4. From Figure 4a, it can clearly be observed that a homogeneous copolymer layer of 6 nm in thickness has been deposited onto the inner and outer walls of the nanotube. This conformal coating along the nanotube wall can be confirmed by examining the bottom part of the tube (Figure 4b). A thicker deposit is observed at the bottom of the tube, suggesting that the copolymer initially grows from the bottom to the top of the tube.

In order to confirm the presence of the copolymer and lithium salt (LiTFSI) in the self-organized TiO₂nt structure, XPS measurement was performed. The XPS spectra of C 1 s and O 1 s binding energies of TiO₂nt and copolymer-embedded TiO₂nt are displayed in Figure 5. Concerning the C 1 s core peak (Figure 5a), the binding energy of C-H and C-O is clearly different around 285 and 286.5 eV. For copolymer-coated TiO₂nt, the main contribution of this peak consists of a component located at 286.7 eV. This component is associated to the presence of oligomeric species of PEO (−CH₂-CH₂-O-) _n for which all carbon atoms are in a one-oxygen environment 26 . A sole peak at 292.8 eV can be assigned to CF₃-like carbon atoms in LiTFSI (LiN(SO₂CF₃)₂) as previously observed by Leroy et al. 27 in agreement with Ensling et al. 28 . The C 1 s peak for TiO₂nt could be decomposed into three components respectively assigned to C-C (285.0 eV), CO (286.5 eV), and O-C = O (288.8 eV). For the O 1 s spectra (Figure 5b), two main peaks of as-formed TiO₂ at 530.2 (Ti-O) and 531.5 eV (O-H hydroxyl) are absent in the copolymer-embedded TiO₂nt sample. The absence of these peaks results from the coverage of TiO₂nt with the copolymer. The broad peak of the copolymer-embedded TiO₂nt sample may be assigned to O in PEO and LiTFSI.

The presence of LiTFSI is confirmed by the XPS spectra of the copolymer-embedded sample at S 2p, F 1 s and N 1 s binding energies as shown in Figure 6. According to Leroy et al. 27 , the XPS core peak characterization of LiTFSI was achieved. In their work, the binding energy of N 1 s spectrum is located at 399.6 eV (399.4 eV in this work). The F 1 s, Li 1 s, and S 2p_3/2 core peaks appear respectively at 688.6, 56.6, and 169.0 eV (688.7, 56.7, 168.8 eV in the present study). Concerning C 1 s spectrum, a sole peak at 293.0 eV can be observed for LiTFSI (LiN(SO₂CF₃)₂), assigned to CF₃-like carbon atoms. Here, the C 1 s core peak corresponding to C-F₃-like carbon atoms is located at 292.8 eV. In conclusion, our results are in perfect agreement with those previously published. We can, thus, conclude to the evident presence of LiTFSI in the copolymer-embedded sample.

The highly conformal electrodeposition of copolymer electrolyte has been successfully achieved on titania nanotubes. It is demonstrated that control of the electropolymerization parameters allows to homogeneously cover the nanostructures without closing the tubes. By this technique, the copolymer-embedded titania nanotubes retain the 3D structure which is advantageous for the further fabrication on high-performance 3D microbatteries.

The authors declare that they have no competing interests.

NP conducted the experiments and prepared the manuscript. NAK participated in the experiments. ME, FD, LS, and PK participated in the discussion. HM carried out the XPS studies. TD supervised the work. All authors read and approved the final manuscript.