Amorphous hybrid TiO2 thin films: The role of organic ligands and UV irradiation

Journal Pre-proofs Full Length Article Amorphous hybrid TiO2 thin films: the role of organic ligands and UV irradiation Maria Luisa Addonizio, Antonio Aronne, Claudio Imparato PII: DOI: Reference:

S0169-4332(19)32911-3 https://doi.org/10.1016/j.apsusc.2019.144095 APSUSC 144095

To appear in:

Applied Surface Science

Received Date: Revised Date: Accepted Date:

17 May 2019 16 September 2019 18 September 2019

Please cite this article as: M. Luisa Addonizio, A. Aronne, C. Imparato, Amorphous hybrid TiO2 thin films: the role of organic ligands and UV irradiation, Applied Surface Science (2019), doi: https://doi.org/10.1016/j.apsusc. 2019.144095

This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

© 2019 Published by Elsevier B.V.

Amorphous hybrid TiO2 thin films: the role of organic ligands and UV irradiation Maria Luisa Addonizioa, Antonio Aronneb, Claudio Imparatob* aCentro

Ricerche ENEA, Località Granatello, Portici (Naples), Italy. E-mail: [email protected]

bDipartimento

di Ingegneria Chimica, dei Materiali e della Produzione Industriale, Università di Napoli Federico II, P.le

Tecchio 80, 80125 Naples, Italy. E-mail: [email protected]; [email protected] *Corresponding author. E-mail: [email protected]. 'Declarations of interest: none'

Abstract The development of a simple, economic and low-temperature synthesis procedure for titanium oxide thin films is an important challenge, due to their diffusion and to the limitations imposed by high temperature annealing. The knowledge on the properties of amorphous TiO2 films, as well as on the effect of their functionalization by organic compounds, is still inadequate. Here we describe a hydrolytic sol-gel procedure to prepare TiO2-based hybrid (inorganic-organic) thin films containing four different complexing agents in their structure (acetylacetone, dibenzoylmethane, citric acid and diethanolamine). The process variables were explored in order to attain stable sols and to tune the features of the amorphous films resulting by drying at 80 °C. Their structural, morphological, optical and electrical properties were studied by SEM, AFM, XRD, ellipsometry, FT-IR and UV-visNIR spectroscopy and resistivity measurements. The effect of the organic component, precursors concentration, deposition parameters and annealing temperature was considered. Moreover, the modifications induced by the exposure of the hybrid films to UV light were deeply investigated. A short UV treatment has a strong impact on the microstructure and wettability of the films and promotes a striking increase in their electrical conductivity.

Keywords: titanium oxide; hybrid film; amorphous film; sol-gel; spin-coating; ellipsometry.

1

1. Introduction Titanium dioxide (TiO2) is a wide band gap semiconductor with a variety of applications. TiO2 films and coatings are employed as sensors, self-cleaning, antimicrobial, protective or anti-reflective layers, electrodes in electrochromic devices, electron transport layers in emerging photovoltaic cells (dye-sensitized, perovskite or polymeric cells), and so on [1–4]. Among the techniques for the synthesis of oxide thin films, wet chemical methods, in particular sol-gel coupled with spin- or dip-coating, offer several advantages: strict control on the composition, easy mixing with additives (e.g. dopants, templates or organic modifiers), inexpensiveness and mild process conditions [2,5,6]. The rapid gelation, which occurs during the rotation of the substrate or its withdrawal from the starting solution, provides uniform films with good substrate adhesion, which is not so easily achieved in the case of coatings of nanoparticles. A thermal annealing at temperatures higher than 400 °C is usually performed after the deposition from solution to obtain compact crystallized films and remove organic additives. Unfortunately, this thermal treatment step makes the manufacturing slower, costlier and energy consuming, and incompatible with some applications, such as flexible coatings on polymeric substrates or tandem solar cell architectures [4,7–9]. The growing interest toward these technologies has driven the research of lowtemperature annealing processes to obtain efficient electron transport materials. Amorphous TiO2 films are largely less studied than crystalline (anatase or rutile) ones, although they have shown interesting performances in applications where high crystallinity is traditionally considered necessary for a better electron/hole separation and mobility, for instance electron selective layers [8,10–12] or photocatalysts [13,14]. The functionalization of TiO2 coatings with organic compounds is also important in the mentioned fields, since it can enhance catalytic activity [15], extend light harvesting, a fundamental aspect in dye-sensitized solar cells [16] and visible light photocatalysis [17], or modify surface wettability and conductivity, improving the contact at the interface with another material. Interface engineering through organic functionalization is a promising way to increase the solar conversion efficiency [16], as was demonstrated for perovskite cells, in which a layer of aminoacid between TiO2 and the perovskite absorber induced a passivation effect and suppressed charge recombination [18,19]. There is therefore an urgency to develop low-temperature synthetic routes for chemically modified TiO2 films, keeping their functional properties, such as transparency, chemical stability, good electrical conductivity and charge carrier transport or photoresponsivity. The simplest way to obtain a TiO2 thin film by sol-gel is the spin coating of an alcoholic solution of a Ti alkoxide precursor, hydrolysed by atmospheric humidity. The process is therefore sensitive to the relative humidity and the condensation can be incomplete, leading to defective (TiOx) films [11,20]. In hydrolytic sol-gel processing the addition of complexing organic compounds, able to stabilize the sol, allows to enhance the homogeneity and tune the surface features of the resulting material [6,21–24], 2

whereas surfactants or polymers can act as templating or structure-directing agents, providing the desired porosity and morphology [25,26]. The organic additives are generally removed after the deposition, so hybrid (inorganic-organic) films are seldom investigated. However, it has been demonstrated that the nature of the complexant plays a role in determining the properties of the TiO2 films produced after thermal treatment [22,24]. They are also affected by the operative conditions, i.e. nature and concentration of the precursor, water amount, solution pH and deposition parameters. Among the treatments employed to modify the surface of TiO2 thin films, UV irradiation is often performed with the purpose to improve surface hydrophilicity and wettability, for example before the deposition of another material in multilayer devices [1,7,20,27]. High energy UV irradiation was recently proposed as an effective photochemical activation method of sol-gel amorphous oxide semiconductors, alternative to thermal annealing, allowing to achieve similar conductive performances and stability [28]. The effects of irradiation seem related to an increase in surface defectivity and hydroxylation, although a complete understanding of this phenomenon is still missing. In this work a series of amorphous hybrid TiO2-based thin films was prepared by a simple sol-gel procedure at low temperature. Acetylacetone, dibenzoylmethane, diethanolamine or citric acid were added to a titanium alkoxide as complexing ligands, obtaining stable sols deposited by spin coating. The influence of the organic component bound in the TiO2 structure, the spinning parameters, heating temperature and UV irradiation on the structural and surface features, optical behavior and electrical resistivity was analyzed.

2. Materials and methods 2.1

Sols preparation and deposition

Titanium (IV) n-butoxide (97+%), acetylacetone (2,4-pentanedione, 99+%), dibenzoylmethane (1,3-diphenyl-1,3propanedione, 98%), citric acid monohydrate (99.0%), diethanolamine (98+%), 1-propanol (99.8+%), cyclohexane (99.5%) and hydrochloric acid (37 wt%), were purchased by Sigma-Aldrich and used as received. Acetylacetone (Hacac), dibenzoylmethane (dbm), citric acid (cit) and diethanolamine (dea), illustrated in Scheme 1, are the complexing compounds chosen to prepare hybrid TiO2 sols. In a typical procedure, a solution of Ti(IV) n-butoxide, one of the complexing agents and 1-propanol with the desired molar ratios was prepared and stirred for 15 min. Then a solution of deionised water (containing 0.1 mol/L HCl when Hacac or dbm was used as complexant) and 1-propanol was slowly added to the first one. The resulting mixture was stirred for 2 h at room temperature. In the case of dbm, a 3:1 cyclohexane/1-propanol mixed solvent was used in order to avoid precipitation. A Ti(IV) n-butoxide solution in 1-propanol was used as precursor for the preparation of bare TiO2 3

films (named T), taken as reference. The conditions of sols preparation are reported in Table 1. All the stock solutions were left ageing at least one day prior to deposition, performed by spin coating usually within one week from preparation. Boron-aluminium silicate glass substrates were carefully cleaned with diluted detergent, 0.75 M NaOH, acetone, ethanol and a cleaning solution (H2O/H2O2/NH4OH = 5:1:1 v/v), with the aid of sonication. A Laurell WS-650 Series spin coater was used for the deposition, at 2000 or 4000 rpm spin rate for 30 s. The films were dried in an oven at 80 °C for 10 min. The hybrid TiO2 films are named A, B, C or D according to the complexant added, as described in Table 1. In some cases, a double layer was deposited to increase the thickness, repeating the spinning and drying procedure. A bilayer film composed of a bottom layer of bare TiO2 and a top layer of TiO2-acac (T+A) was also prepared, with the idea of reducing the organic content while keeping the functionalization on the superficial part of the film. In order to follow the evolution of the films properties with the heating temperature, some samples were also annealed in air at 150 °C for 1 h (named A-150, B-150, C-150 and D-150) or at 400 °C for 1 h.

Table 1. Principal conditions for the preparation of the hybrid TiO2 films by sol-gel.

2.2

Sample

Ligand

Ligand/Ti (mol/mol)

H2O/Ti (mol/mol)

[Ti] (mol/L)

A

acac

0.5

4

0.30

B

dbm

0.2

4

0.15

C

cit

1

2

0.30

D

dea

0.5

2

0.30

Thin films characterization

The surface morphology of the films was examined by optical microscopy (Zeiss, mod. Axiopot), Scanning Electron Microscopy (SEM, Leica-Cambridge mod. S360), and Atomic Force Microscopy (AFM, Bruker mod. NSIV). X-ray Diffraction in grazing incidence mode (GIXRD) was performed by means of a Philips X’Pert PRO MRD diffractometer, using CuKα radiation (λ = 0.154056 nm) and an incident angle of 0.5°. Fourier-Transform Infrared (FT-IR) spectroscopy was performed by means of a Nicolet system, Nexus model spectrometer, equipped with a DTGS KBr detector. The transmittance spectra were recorded at room temperature, in the 4000−400 cm−1 range, with a spectral resolution of 2 cm−1, on films deposited on Si <100> substrate. The spectrum of each sample represents an average of 64 scans, corrected for the spectrum of the blank (a clean Si wafer). 4

The optical properties of the films were analysed by UV-visible-NIR spectroscopy, using a double beam Perkin Elmer (Lambda 900) spectrophotometer equipped with an integrating sphere. Transmittance and reflectance spectra were recorded in the 200-1500 nm range and absorbance was calculated by difference. The thickness and optical constants of the TiO2 films were measured using spectroscopic ellipsometry (SE, Woollam mod. WVASE32). The SE spectra (, ) were recorded in the wavelength range 300-1000 nm. The electrical resistivity of the films was evaluated at room temperature by two-point probe measurements in longitudinal geometry. Two parallel silver contacts were placed on the film’s surface. Increasing voltages were applied, in the range 1-80 V, measuring the current intensity (in the dark). The electrical resistivity, ρ (Ω·cm), was then calculated by the following equation: 𝜌=

𝑉𝐿 𝑆 𝐼 𝑑

where: V is the potential (V), I is the current intensity (A), L is length of the Ag contacts (cm), d is the distance between the Ag contacts (cm) and S is the film thickness (cm). In order to investigate the effect of UV treatment on the films, some of them were irradiated by using a Mercury lamp (100 mW/cm2) as UV source with an exposure time of about 15 min, and then characterized.

3. Results and discussion 3.1

Sol-gel preparation of hybrid films

The first step in the sol-gel synthesis of thin films is the preparation of a stable stock solution. Ti alkoxides in alcohol solution, the most used precursors, are highly reactive toward hydrolysis and the contact with water vapour may start their hydrolysis and condensation, leading to the formation of aggregates and precipitates. The reference T films were prepared starting from Ti(OBu)4 diluted in 1-propanol (0.15 M). As expected, the stability of these stock solutions when shortly exposed to air was limited. During spin coating and subsequent ageing and drying atmospheric humidity can cause a sufficient degree of hydrolysis to yield a titanium oxide film, although non-stoichiometric (TiOx) [11,20]. On the other hand, in this approach hydrolysis and polycondensation reactions are strictly dependent on the relative humidity, which is not controllable outside a glove box, and they may be incomplete, causing a variability in the resulting structure. The use of organic compounds able to coordinate Ti ions is known as an effective way to moderate the hydrolysis rate of Ti alkoxides and then stabilize sols allowing the addition of a proper water amount, which ensures a high degree of hydrolysis, so that upon spinning or dipping condensation reactions are favoured, providing a better 5

cross-linking and uniformity of the gel film. The complexant is usually removed by annealing at high temperature to obtain the pure oxide. With the purpose to study the hybrid inorganic-organic films, in this work the samples were only dried at 80 °C, which is sufficient to volatilize most of residual alcohol and water trapped in the thin coatings. Some of the samples were annealed at 150 °C, since heating is believed to enhance the compactness and electrical conductivity of oxide films, and this temperature is still compatible with the adoption of some polymeric substrates, to produce light and flexible coatings, and with multijunction solar cells [8,27]. The chemical features of the complexing molecule are expected to affect the properties of the hybrid films, owing to their different size, polarity, functional groups, binding mode to Ti and charge transfer character of the obtained complex. Therefore we explored the effect of two characteristics of the ligand, hydrophilicity and coordination kind to metallic centre, on the properties of the hybrid film using two β-diketones of different hydrophilicity (acetylacetone and dibenzoylmethane), to parity of their bidentate coordination to Ti, and two polydentate compounds (citric acid and diethanolamine) with different acid-base characteristics. Each of the four organic complexants required different conditions for the formation of stable homogeneous sols, as indicated in Table 1 and discussed in the following. Acetylacetone (Hacac, Scheme 1a) is an efficient chelator for most transition metals, and probably the most common stabilizing ligand for titanium in sol-gel chemistry [6,22,23,26,29,30]. The acetylacetonate (acac) anion coordinates Ti ions in a bidentate chelating mode, forming 6-membered rings stabilized by partial charge delocalization. The complexing ratio (acac/Ti) can be varied to direct the evolution of the sol-gel process towards gelation or sol stabilization. We have previously studied TiO2-acac bulk gels, finding peculiar properties, among them the stable surface adsorption of superoxide radicals and the consequent activity in the degradation of organic pollutants [31,32]. We observed that sols with high Ti concentration (about 1.0 mol/L) and acac/Ti molar ratio = 1 remained unaltered for weeks [33]. However, coatings of sols with [Ti] > 0.5 mol/L formed cracks during drying or annealing. When the mixture was diluted to [Ti] = 0.3 mol/L, thinner, more uniform and resistant layers were produced, and a lower complexation ratio (c = 0.5) could be used, whereas at higher Ti concentration a 0.5 ratio would not prevent gelation. With a small amount of HCl (0.1 mol/L in the hydrolytic solution), resulting in a pH 5-6, the sols are stable for more than one month even with excess water, H2O/Ti molar ratio (h) = 10. Varying h between 10 and 4 did not produce a significant effect on the main properties of the TiO2-acac films. Besides acetylacetone, a small and relatively hydrophilic diketone, we tested another β-diketone with larger substituents (two phenyl groups) and a marked hydrophobic character: dibenzoylmethane (dbm), 1,3-diphenyl1,3-propanedione (Scheme 1b). Although metal complexes of dbm are studied for different applications, the only literature reports about materials containing dbm bound to TiO2 are some works by the groups of Tohge and Segawa. They prepared photosensitive hybrid TiO2 or ZrO2 films by sol-gel adding Hacac, benzoylacetone or dbm 6

to study the patterning and methacrylic acid polymerization induced by dissociation of the metal-chelate complex under UV irradiation [34–36]. A few patents mention the possibility of using dbm among other chelating ligands for the preparation of metal oxide films and nanostructures. Due to the hydrophobic nature of dibenzoylmethane, and the formation of complexes and clusters with Ti(OBu)4, precipitation occurred in 1-propanol after hydrolysis. The polarity of the solvent was therefore reduced switching to a cyclohexane/propanol solution, which favoured solvent-ligand interactions, allowing sol stabilization. The sols had a bright yellow colour as a result of the ligand-to-metal charge transfer in the Ti-dbm complex. It is worth noting that a relatively low complexation ratio (dbm/Ti = 0.2 molar ratio) provided a sol unaltered for at least 3 months, attesting a superior stabilizing ability of dbm over the other ligands. Higher ratios were not explored, because the aim was to adopt the lower possible content of organic ligands in the films. Among the chosen polydentate ligands, citric acid (Scheme 1c) is a triprotic acid, safe, naturally abundant and cheap, useful as chelating and capping agent in the synthesis of a wide range of nanomaterials, to control their size and morphology or to give surface functionality [5,37]. Citric acid is commonly employed in the sol-gel route and its variants, such as combustion synthesis and Pechini method, but has been less studied for film processing. We observed that, in spite of its polydentate nature, citrate ligand was not able to effectively stabilize the sol with c = 0.5 and [Ti] = 0.3 mol/L, as gelation started in a few days, thus c = 1 was used. Lowering [Ti] to 0.15 mol/L, a good stability was obtained with c = 0.5. The pH of these sols was about 2. Diethanolamine (dea, Scheme 1d) is a weak base, which can act as a tridentate ligand [38,39], and is also frequently used for the stabilization of TiO2 sols for coatings [7,10,21,22,25,40]. For the complexation with dea, c ratios 0.5 and 1 were tested, with [Ti] = 0.3 mol/L and h = 2. With c = 0.5 a transparent sol was formed, but it did not show a long-lasting stability: in about two weeks it became opalescent and its viscosity increased until partial gelation. The sol with c = 1, a ratio often reported in literature [21,25] was stable for a longer time. Its pH was about 10. Films were prepared starting from both stock solutions, showing comparable characteristics. The influence of most spin coating parameters was evaluated on the TiO2-acac system. A key variable is the spinning rate, directly related to the film thickness. The values used were 2000 or 4000 rpm; a higher thickness was obtained, as expected, by either a decrease in the spinning rate or an increase in the precursor concentration. However, at 1000 rpm some films resulted non-uniform because the rotation was too slow to spread the solution evenly on the whole substrate surface. The spinning acceleration was found to have a smaller influence on the film thickness and morphology.

7

3.2 3.2.1

Effect of the organic complexant Structural and morphological properties

A first inspection of the films was made through optical microscopy. When a deposited layer was too thick (for example with [Ti] > 0.5 mol/L in the sol), cracks were observed on the surface, that increased after drying or thermal treatment. The mechanism of crack formation for sol–gel-derived films is attributed to the generation of tensile stresses, induced by capillary force during the drying and densification process, which can also be responsible for some inhomogeneities and other signs of stress on the surface [41]. The synthetic conditions were optimized in order to improve the quality of the obtained films, so that the films appeared crack-free and clean and a compact structure of the film with uniform thickness was observable in the SEM image of Fig. 1. AFM image and related profile of Fig. 2a show for the TiO2-acac (A) sample a flat and uniform surface on the nanometric scale. An exception respect to the smoothness and regularity of the others coatings was observed with the TiO2-dbm (B) sample, which exhibits a deeply rough surface (see Fig. 2b), with a root mean square (RMS) roughness estimated to be as high as 52 nm (with a thickness of 142 nm). After annealing at 400 °C most of the organic has been removed and crystal nucleation has started, but the B film retains its morphology, with a RMS roughness about 28 nm, decreased proportionally to the thickness (73 nm), attesting the structure-directing function of dbm. The TiO2-dbm sol is the only one in a mixed cyclohexane/1-propanol solvent. It is plausible that the assembly of dbm ligands on the surface of molecular Ti clusters favoured the formation of relatively large primary particles prior to the spin coating. The higher volatility of cyclohexane with respect to 1-propanol could also contribute to the disordered film structure formed during the evaporation of the solvent, as it occurred with an increase in the polarity of the mixture (because of the faster evaporation of cyclohexane), possibly inducing further aggregation of the rather hydrophobic dbm-coated particles. Although many applications request a flat and even surface, others can benefit from such a rough morphology, for example when a large surface area or an enhanced light scattering is needed. The marked hydrophobicity of the TiO2-dbm film compared to the other studied systems can be appreciated by the smaller radius of a water droplet deposited on the surface, indicating a larger contact angle (Fig. S1a). The hybrid films dried at 80 °C, as well as those treated up to 150 °C, have an amorphous structure, as shown by GIXRD (Fig. 3). In the diffractograms the wide band between 15 and 30° is principally due to the amorphous halo of the glass substrate, although the amorphous phase of TiO2 can give a contribution. Annealing in air at 400 °C for 1 h the A and T films induces at least partial crystallization, with the formation of anatase nanocrystals, as

8

attested by the position of the main diffraction peak of the anatase polymorph (25.33°). Their grain size is about 8-10 nm, depending on the preparation conditions of the film. Thickness has a crucial role in determining several properties of a thin film [42], even some intensive (bulk) ones, such as electrical conductivity. The thickness of selected films, measured by ellipsometry, is reported in Table 2.

Table 2. Thickness (t), average transmittance (measured in 380-1000 nm range), refractive index at 550 nm (n), porosity (estimated by ellipsometric data) and electrical resistivity of representative samples, before and after 15 min UV irradiation. as prepared Sample

after UV irradiation

t (nm)

aver. T (%)

n

porosity (%)

ρ (106 Ω cm)

t (nm)

aver. T (%)

n

porosity (%)

ρ (106 Ω cm)

A

65

86

1.84

55

10

60

86

1.82

57

0.5

B

145

84

1.70

65

>10000*

106

84

1.79

59

5

C

320

91

1.66

67

20

57

81

1.99

45

1

D

140

87

1.78

60

100

55

81

2.01

43

0.2

T

45

87

1.84

55

5

45

87

1.80

58

0.5

* B sample did not show ohmic conductive behaviour.

The expected trend of a thickness increase with the increase of Ti(OBu)4 concentration and with a reduction of the spinning rate is confirmed. A strong dependence on the type and concentration of ligand is also observed. In similar deposition conditions, at the same ligand and Ti concentration the thickness follows the trend: acac < dea < cit. This is likely due not only to the size of the molecule, but also to its ability to stabilize smaller or larger molecular aggregates and to influence the viscosity of the stock solution. Citrate proved to be the least efficient stabilizing agent, so the TiO2-citrate sol has probably higher degree of polymerization and viscosity, producing thicker layers. Heating at 150 °C induces a limited thickness decrease (about 10 %), while the annealing at 400 °C causes a large decrease, proportional to the amount of organics evacuated from the structure. Also the T films showed a thickness reduction of at least 30 % upon annealing at 400 °C, suggesting that a significant amount of alkoxide was not hydrolysed by humidity and remained in the structure of the dried films. The high temperature treatment is anyway expected to induce a densification and sintering of the xerogel films. The hybrid composition of the films is verified by FT-IR spectra, which present the typical features of the respective organic compounds. The A film (Fig. 4a) displays the same bands seen in TiO2-acac [31,33] and ZrO29

acac gels [43]. Those at about 1595 and 1435 cm-1 are attributed to the antisymmetric and symmetric stretching of the C=O bond, largely shifted to lower wavenumbers than in free acetylacetone (where they are found about 1700 cm-1). This shift is compatible with a chelating coordination of the enolate to Ti4+ ions, forming a 6membered ring stabilized by resonance [29,31,44]. The band at 1533 cm-1, also absent in free Hacac, can be assigned to the enol C=C bonds [44], the one at 1372 cm-1 should be related to another C-O stretching [43,44]. Most bands in the FT-IR spectrum of a B film (Fig. 4b) are not noticeably shifted compared to pure dbm. It was expected, because the tautomeric keno-enol equilibrium in dbm is largely shifted towards the enol form, thanks to the formation of a strong intramolecular hydrogen bond and the consequent charge delocalization [45]. The major difference is the intense band appearing in the hybrid film at 1360 cm-1, that has to be associated to the Ti-dbm complex. The bands in the range 1300–1600 cm-1 are similar to those observed in TiO2-acac samples. They are mainly due to the stretching of C=O and C=C of the enol ring, and some of them result from combinations involving vibration modes of the phenyl rings [45]. A bidentate chelation of dbm, analogous to acac, can be argued, considering the correspondence of the observed bands with the enol structure and the charge transfer complex formation, evidenced by UV-vis spectra (see below), in agreement with previous reports [36]. The spectra of the hybrid films heated at 150 °C (Fig. S2) show that the organic compounds are retained in the structure after the thermal annealing, without significant changes compared with the as-prepared films. For the A-150 sample the bands are slightly shifted with respect to the untreated film (e.g. C=O stretching at 1582 and 1428 cm-1), suggesting that the coordination may be somewhat strengthened after annealing. The spectrum of D-150 shows characteristic features of alkanolamines: the N-H deformations above 1600 cm-1, the C-H bending at 1455 cm-1, the sum of C-O, C-N vibrations and C-H bending around 1080 cm-1 and the N-H stretching centred at 3250 cm-1 [22,39]. Diethanolamine is generally described as a tridentate ligand, able to coordinate metal ions through its two hydroxyl groups with a contribution from the nitrogen atom [38,39]. The spectrum of C-150 contains both bands assigned to pure citric acid (especially the intense band at 1740 cm-1, due to C=O stretching) and several overlapping bands that can be assigned to coordinated carboxylate groups. In particular, the band at 1565 cm-1 may be due to the asymmetric stretching of coordinated COO-, and those at 1435 and 1400 cm-1 to the symmetric stretching [37,46]. The splitting between these bands (around 130–160 cm-1) indicates a probably bidentate bridging coordination [46]. The existence of other contributions in the surrounding range suggests the presence of various binding modes and some carboxylic groups which remained free (as can be expected considering the citrate content, citrate/Ti = 1) or more weakly bound, for example through hydrogen bonds. The coordination by the central α-hydroxyacid moiety of citrate, forming a 5-membered chelate, is another possible binding mode [47]. Deep modifications of the organic components occur during UV irradiation in some of the studied films, as discussed below. 10

3.2.2

Optical properties

The interaction with solar light is fundamental in photovoltaic and correlated applications. In regular architecture solar cells, the anode is the transparent front electrode, so the TiO2 electron transport layer, intercalated between the transparent conductive oxide (typically FTO or ITO) and the absorbing layer, is required to show high transmission, allowing the largest possible fraction of solar radiation to reach the absorber. The transmittance and reflectance spectra in the UV/visible/NIR range for samples with different compositions are shown in Fig. 5, with the calculated absorption spectra. All the samples present comparable characteristics: a high transmittance, between 80 and 90 %, almost constant in the whole visible and NIR range (see the average transmittance values in Table 2), and an absorption edge around 350 nm. The corresponding band gap energies, evaluated by alpha plot linearization for indirect transitions, are between 3.3 and 3.4 eV, in line with the typical values for amorphous TiO2 films, which have generally slightly larger band gaps than bulk and crystalline TiO2 materials [3,30]. The TiO2-dbm films present an observable yellow coloration and an evident additional absorption band, centred at 390 nm and extending up to 480 nm (Fig. 5b). It is related to a ligand-to-metal charge transfer from the ground state (highest occupied molecular orbital, HOMO) of dbm to TiO2 conduction band. Such direct transition was demonstrated to occur when Ti ions are complexed by some bidentate ligands, like acetylacetone and catechol [48], and in the case of TiO2-acac gels it was associated with unusual properties, in particular the generation of stable superoxide radicals on the surface [31]. The films with acac and citrate as complexants would be expected to have a similar visible light absorption observed in xerogels with the same composition (dark yellow with acac and light yellow with citrate), although less pronounced. Conversely, they appear transparent and their absorption edge is only marginally shifted to higher wavelengths with respect to bare TiO2 films (see Figs. 5b, 5d and 5f). At the nanoscale some properties can differ from those encountered in bulk materials. An example is the absorption edge at lower wavelengths generally found in thin films, indicating a wider valence-conduction band gap. This could imply also a larger energy gap between the HOMO of the ligands and the conduction band edge, compared to the corresponding xerogels. Heating at 150 °C induces small variations in the transmittance fringes, but the average value is practically unaltered. The annealing at 400 °C, the subsequent densification and partial crystallization are accompanied by a broadening of the absorption band in pure TiO2 films, compatible with anatase formation, while almost no changes are seen in TiO2-acac (data not shown). A bilayer T-A film exhibits similar optical spectra as TiO2-acac alone (Fig. 5e and 5f). The consequences of UV irradiation on the optical behaviour are described in Section 3.3. 11

In general, the amorphous hybrid films present a high transparency in the visible and NIR range, and show limited variations in the optical characteristics with their composition and treatment temperature. Spectroscopic ellipsometry was used to gain a deeper insight into various optical and structural properties of TiO2 thin films, like film thickness, refractive index (n), extinction coefficients (k) and porosity, which can be derived from these data [30,49,50]. After the ellipsometric data have been acquired, a fitting procedure, utilizing the Cauchy model, was performed to calculate optical parameters and thickness by minimizing the mean square error (MSE). Sol-gel prepared films have been considered homogeneous structures in the computation. At 550 nm, our films have a refractive index ranging from 1.70 up to 1.84, which is much lower than the value reported for anatase (2.52) [49], but closer to amorphous TiO2 (2.03). This is consistent with the amorphous structure attested by XRD results. Lower refractive indexes may also indicate TiO2 thin films with low density and with a significant content of organic complexant, as attested also by FT-IR analysis. In Fig. 6a the refractive index and the extinction coefficient are shown. It can be observed that n strongly decreases with wavelength in the range 300 – 500 nm whereas, for longer wavelengths, it slightly decreases for all the samples. Spectral extinction coefficient k is near zero for TiO2 and TiO2-acac films. It decreases with increasing wavelength and reaches values less than 0.01 at wavelengths higher than 450 nm for films C and D. Sample B shows a different behaviour, with k decreasing at the increase of the wavelength, varying from 0.1 down to 0.02. The porosity values of the TiO2 films are determined from the following equation [51]:

(

Porosity (%) = 1 ―

𝑛2 ― 1

)

𝑛2𝑑 ― 1

∙ 100

where n denotes the refractive index of the porous thin film and nd is the refractive index of dense (pore free) anatase phase (nd = 2.52). The porosity values are lower for T and A films (55 %) and higher (≥ 60 %) for the other hybrid films fabricated in this work. These differences in porosity can be related to the structure, size and amount of the organic complexant retained in the structure of the films, as well as to the size of the clusters formed during condensation reactions. The most porous samples show a stronger dependence of the refractive index on UV irradiation, as commented in Section 3.3.

3.2.3

Electrical properties

A suitable electron conduction is one of the pivotal properties of a semiconductor, especially in photovoltaic devices. The electrical resistivity was measured in longitudinal geometry, by means of two parallel silver contacts drawn on the film’s surface. The samples normally showed ohmic behaviour, with direct proportionality between the applied voltage and the measured current. A selection of electrical resistivity data on different kinds of films 12

are listed in Table 2 and the corresponding conductivity values are reported in Fig. 7. The electrical resistivities measured vary widely and result comparable with those reported in literature for amorphous and also crystalline TiO2 films (106 – 108 Ω·cm) [3,9,11]. With respect to the reference TiO2 (T), most amorphous films present a relatively higher resistivity, although some A samples show values in the same range. A strong correlation was observed between thickness and resistivity: when films with similar composition but different thickness were compared, an inverse proportionality between the resistivity and thickness was always recorded. This is a common characteristic of very thin films, where the surface electrical conduction is more relevant than in thicker layers with regard to the bulk conduction. Beyond a defined thickness threshold, the resistivity should return to behave as a bulk property. In this work this dependence was found to be kept at least up to about 150 nm. It affected also crystallized films, which are supposed to exhibit better conductivity than the amorphous counterpart. The thinnest films (below 20 – 30 nm) annealed at 400 °C, despite the partial crystallization did not show ohmic behaviour, so a reliable resistivity value was impossible to determine. On the other hand, some TiO2-acac films thicker than 100 nm gave excellent resistivity in the order of 105 Ω·cm, with some further increase after annealing at 400 °C. It is also interesting to compare the as prepared films (dried at 80 °C) with those subsequently heated at 150 °C for 1 h. According to the common conception, even a mild thermal treatment at that temperature should increase the density and compactness of the oxide framework, and, in the case of solution processed coatings, remove residual solvent molecules, with a positive influence on electrical conductivity. Nevertheless, in most studied samples the inverse trend was observed, as can be noticed in Fig. 7. This agrees with the report of Deng et al., who measured resistivities in the range of 106 – 107 Ω·cm, growing with the heating temperature, in amorphous TiO2 films prepared by a simple sol-gel procedure, similar to the T films studied here [11]. They affirmed that the substoichiometric composition, i.e. oxygen defectivity, favours the electrical conduction in the low-temperature processed films, and that the relatively high resistivity did not hamper the efficiency of the perovskite cells fabricated on those TiOx films [11]. It may be hypothesized that heating in air at 150 °C removes part of the structural defects, such as oxygen vacancies and surface –OH groups, likely abundant in the low temperature dried films, thus decreasing the conductivity. A double-layer film, with bare TiO2 on the bottom and a hybrid TiO2-acac layer on the top (T-A) was also tested, with the purpose to enhance the performances. In this way, only the side closer to the surface is functionalized, while the foundations of the coating could gain better stability and conductivity. The first results on such bilayer samples (about 70 nm thick) showed reduced resistivity compared to the single components (Fig. 7). As usual, the thickness of both layers can be tuned varying the precursor concentration and the spinning rate. This may 13

prove as a simple but effective way to extend the applicability of the hybrid sol-gel route proposed in this work to the functionalization of metal oxide surfaces with organic complexes.

3.3

Effect of UV irradiation

Particular attention has been devoted to the effect of the irradiation with UV light on the hybrid films. Such treatment is often performed on TiO2 surface before the deposition of another material in multilayer devices to improve the wettability of the surface, because high energy irradiation increases its defectivity and hydroxylation. High energy UV irradiation was proposed as an effective photochemical activation method of solgel amorphous oxide semiconductors, in alternative to thermal annealing, allowing to achieve similar conductive performances and stability [28]. Hence this method was found useful to obtain efficient compact and crystallized TiO2 layers for solar cells [7,52]. From a structural viewpoint, a UV treatment up to 30 min does not induce crystallization of the films (Fig. 3), which was conversely found by some authors after a longer irradiation time of 90 min [7]. The surface morphology does not seem to undergo substantial modifications. It is interesting to note the effect on the film thickness, which varies with the composition. T and A films result practically unchanged, while the others undergo a marked thinning, with a variation trend: cit > dea > dbm. These data point at a modification and at least a partial removal of the organic component under intense UV irradiation, along with a density increase. The most obvious thickness reduction occurs on the samples with citrate and dea, consistent with the loss of large part of the organic phase, in agreement with FT-IR and ellipsometric data (see below). FT-IR spectra recorded on hybrid coatings containing acac and dbm after 15 min UV irradiation are reported in Fig. 4, compared with the as prepared samples. Both show a decreased absorption intensity, but while in A the band positions are almost unmodified, in B a change is evident: the band at 1360 cm-1, which can be ascribed to the stretching of a chelating carbonyl, disappeared together with the band at 1480 cm-1, and a new one appeared at 1414 cm-1. Concurrently, the visible light absorption band around 400 nm disappeared (Fig. 5b). The same transformation was reported by Segawa et al. upon UV irradiation up to 60 min of a TiO2-dbm film [36]. They deduced the breaking of the Ti-dbm coordination bonds and argued that released dbm can undergo a C-C cleavage, forming radicals that initiate methacrylic acid polymerization. In our case in the UV/vis/NIR spectrum an extended absorption shoulder remains, associated to the residual dbm and derived organic products, which disappears after annealing at 400 °C. The FT-IR spectra of hybrid films prepared with citrate and dea recorded after UV exposure (not reported) did not clearly show the features related to these ligands anymore. In agreement with the large thinning, it has to 14

be concluded that these compounds undergo photodecomposition, supported by the well-documented photocatalytic activity of TiO2 in the UV range. In these films (C and D) a similar marked reduction of transmittance and an increase in reflectance occur in the visible range (Fig. 5c). This change can be due to decomposition products of citrate and dea ligands, or to the altered structure of the films. On the contrary, the optical spectra of A are practically unaffected, confirming the TiO2-acac coatings as the most stable under irradiation. Although the concentration of acac and dbm in the samples seems to decrease and the Ti-dbm complex is mainly dissociated, diketones result to be the least susceptible to photoinduced degradation, possibly thanks to the higher stability of their structure and of the chelate bond formed with Ti. Fig. 6b shows the spectral refractive index (n) and extinction coefficient (k) behaviours for films exposed to UV irradiation for 15 min. Except for films T and A, where n results almost unchanged, for the other films n increases and consequently the porosity decreases after UV treatment (Table 2). These changes in the refractive index seem to be driven by film densification, as confirmed by the reduction of porosity. For samples with citrate, dea, and to a lesser extent dbm, a strong densification can be assumed, consistent with the thickness reduction and the total or partial removal of the organic components. The UV treatment revealed a dramatic influence on the electrical properties, causing in all cases an outstanding decrease of resistivity, from one to three orders of magnitude (Fig. 7). A similar effect was observed in sol-gel TiO2 films by some authors and correlated again with an increase of n-type conductivity due to the formation of lattice or electronic defects, like oxygen vacancies or interstitial Ti4+ ions [7]. In the hybrid films the removal of excess organics, achieving a denser and more ordered structure, probably contributes to reduce the resistance to electron conduction, together with the induced defectivity. Different irradiation times (5, 15 or 30 min) did not strongly influence the variation of conductivity on A samples. The effect of irradiation on dark conductivity suggest that the amorphous films might also generate relevant photocurrent intensities under illumination. Finally, the effect of UV irradiation on the wettability of the surfaces was observed. Photoinduced superhydrophilicity is a well-known phenomenon consisting in the enhancement of surface hydrophilicity of some metal oxides upon illumination (with UV light for TiO2). Its origin is still debated and may be linked to several factors, mainly the increase of surface hydroxyl groups and the photodecomposition of adsorbed hydrocarbon contaminant, but also the formation of oxygen vacancies and the variation of surface roughness [1,53]. The extent and duration of the superhydrophilic behaviour after the irradiation of TiO2 film has been the object of some studies [20,24,54]. After UV illumination we noticed the neatly higher spreading of a water droplet on the surface of all the hybrid films (Fig. S2), pointing at an increased wettability. The extent of this increment was comparable on all samples, including B, which was initially the most hydrophobic. An enhanced 15

hydrophilicity is important for antifogging coatings and to improve the contact between TiO2 and other layers in multilayer devices.

3.4

Conclusions

The proposed synthetic procedure represents a single-step route to realize amorphous films based on titanium dioxide with an extended organic functionalization which is not limited to the surface. We prepared hybrid films with different organic ligands coordinating Ti4+ ions (acetylacetone, dibenzoylmethane, citric acid and diethanolamine), optimizing the conditions to achieve long-term stability of the precursor sol for each system. Choosing the appropriate complexant and conditions makes it possible to tune the thickness, porosity, roughness and hydrophilicity of the hybrid films produced at low temperature. The presence of the organic component does not strongly affect the film transparency, except for the TiO2-dbm samples, which feature a visible light absorption band. Conversely, the electrical conductivity of the amorphous films is influenced by the employed complexant. Moreover, it increases with film thickness and decreases after a low temperature annealing (150 °C). A short UV irradiation (15 min) induces considerable changes in the film properties, likely related to the photoinduced decomposition of the organic ligands and concurrent densification; diketones seem to be the most resistant to the irradiation, allowing to retain a hybrid composition, while the other complexants are mainly removed from the TiO2 matrix. A remarkable increase in electrical conductivity of at least one order of magnitude is measured for all the films after UV illumination. These findings may turn useful for various possible applications of TiO2 thin coatings with organic functionalization and tailored properties.

Acknowledgements The present work was supported by the Italian Ministry of Economic Development in the framework of the Operating Agreement with ENEA for the Research on the Electric System.

16

References [1]

S. Banerjee, D.D. Dionysiou, S.C. Pillai, Self-cleaning applications of TiO2 by photo-induced hydrophilicity and photocatalysis, Appl. Catal. B Environ. 176–177 (2015) 396–428. doi:10.1016/j.apcatb.2015.03.058.

[2]

G. Varshney, S.R. Kanel, D.M. Kempisty, V. Varshney, A. Agrawal, E. Sahle-Demessie, R.S. Varma, M.N. Nadagouda, Nanoscale TiO2 films and their application in remediation of organic pollutants, Coord. Chem. Rev. 306 (2016) 43–64. doi:10.1016/j.ccr.2015.06.011.

[3]

N.R. Mathews, E.R. Morales, M.A. Cortés-Jacome, J.A. Toledo Antonio, TiO2 thin films - Influence of annealing temperature on structural, optical and photocatalytic properties, Sol. Energy. 83 (2009) 1499– 1508. doi:10.1016/j.solener.2009.04.008.

[4]

G. Yang, H. Tao, P. Qin, W. Ke, G. Fang, Recent progress in electron transport layers for efficient perovskite solar cells, J. Mater. Chem. A. 4 (2016) 3970–3990. doi:10.1039/c5ta09011c.

[5]

A.E. Danks, S.R. Hall, Z. Schnepp, The evolution of ‘sol–gel’ chemistry as a technique for materials synthesis, Mater. Horiz. 3 (2016) 91–112. doi:10.1039/C5MH00260E.

[6]

M. Epifani, C. Giannini, L. Tapfer, L. Vasanelli, Sol–Gel Synthesis and Characterization of Ag and Au Nanoparticles in SiO2, TiO2, and ZrO2 Thin Films, J. Am. Ceram. Soc. 83 (2000) 2385–2393. doi:10.1111/j.1151-2916.2000.tb01566.x.

[7]

L. Huang, J. Xu, X. Sun, C. Li, R. Xu, Y. Du, J. Ni, H. Cai, J. Li, Z. Hu, J. Zhang, Low-temperature photochemical activation of sol-gel titanium dioxide films for efficient planar heterojunction perovskite solar cells, J. Alloys Compd. 735 (2018) 224–233. doi:10.1016/j.jallcom.2017.11.027.

[8]

D. Yang, R. Yang, J. Zhang, Z. Yang, S. Liu, C. Li, High efficiency flexible perovskite solar cells using superior low temperature TiO2, Energy Environ. Sci. 8 (2015) 3208–3214. doi:10.1039/C5EE02155C.

[9]

K. Wojciechowski, M. Saliba, T. Leijtens, A. Abate, H.J. Snaith, Sub-150 °C processed meso-superstructured perovskite solar cells with enhanced efficiency, Energy Environ. Sci. 7 (2014) 1142–1147. doi:10.1039/C3EE43707H.

[10]

T. Kuwabara, T. Nakayama, K. Uozumi, T. Yamaguchi, K. Takahashi, Highly durable inverted-type organic solar cell using amorphous titanium oxide as electron collection electrode inserted between ITO and organic layer, Sol. Energy Mater. Sol. Cells. 92 (2008) 1476–1482. doi:10.1016/j.solmat.2008.06.012.

[11]

X. Deng, G.C. Wilkes, A.Z. Chen, N.S. Prasad, M.C. Gupta, J.J. Choi, Room-Temperature Processing of TiOx Electron Transporting Layer for Perovskite Solar Cells, J. Phys. Chem. Lett. 8 (2017) 3206–3210. 17

doi:10.1021/acs.jpclett.7b01466. [12]

A. Huang, J. Zhu, J. Zheng, Y. Yu, Y. Liu, S. Yang, S. Bao, L. Lei, P. Jin, Room-temperature processible TiO2 electron selective layers with controllable crystallinity for high efficiency perovskite photovoltaics, Sol. Energy Mater. Sol. Cells. 163 (2017) 15–22. doi:10.1016/j.solmat.2017.01.006.

[13]

K. Eufinger, D. Poelman, H. Poelman, R. De Gryse, G.B. Marin, Photocatalytic activity of dc magnetron sputter

deposited

amorphous

TiO2

thin

films,

Appl.

Surf.

Sci.

254

(2007)

148–152.

doi:10.1016/j.apsusc.2007.07.009. [14]

J. Seo, H. Lee, H.J. Lee, M.S. Kim, S.W. Hong, J. Lee, K. Cho, W. Choi, C. Lee, Visible light-photosensitized oxidation of organic pollutants using amorphous peroxo-titania, Appl. Catal. B Environ. 225 (2018) 487– 495. doi:10.1016/j.apcatb.2017.12.009.

[15]

S. Biran Ay, N. Kosku Perkgoz, Nanotechnological Advances in Catalytic Thin Films for Green Large-Area Surfaces, J. Nanomater. 2015 (2015). doi:10.1155/2015/257547.

[16]

M. Grätzel, R.A.J. Janssen, D.B. Mitzi, E.H. Sargent, Materials interface engineering for solution-processed photovoltaics, Nature. 488 (2012) 304–312. doi:10.1038/nature11476.

[17]

P. Łabuz, R. Sadowski, G. Stochel, W. Macyk, Visible light photoactive titanium dioxide aqueous colloids and coatings, Chem. Eng. J. 230 (2013) 188–194. doi:10.1016/j.cej.2013.06.079.

[18]

Y. Ogomi, A. Morita, S. Tsukamoto, T. Saitho, Q. Shen, T. Toyoda, K. Yoshino, S.S. Pandey, T. Ma, S. Hayase, All-solid perovskite solar cells with HOCO-R-NH3+I- anchor-group inserted between porous titania and perovskite, J. Phys. Chem. C. 118 (2014) 16651–16659. doi:10.1021/jp412627n.

[19]

Y.C. Shih, L.Y. Wang, H.C. Hsieh, K.F. Lin, Enhancing the photocurrent of perovskite solar cells via modification of the TiO2/CH3NH3PbI3 heterojunction interface with amino acid, J. Mater. Chem. A. 3 (2015) 9133–9136. doi:10.1039/C5TA01526J.

[20]

S.-H. Nam, S.-J. Cho, C.-K. Jung, J.-H. Boo, J. Šícha, D. Heřman, J. Musil, J. Vlček, Comparison of hydrophilic properties of TiO2 thin films prepared by sol–gel method and reactive magnetron sputtering system, Thin Solid Films. 519 (2011) 6944–6950. doi:10.1016/j.tsf.2011.04.144.

[21]

Y. Djaoued, S. Badilescu, P.. Ashrit, D. Bersani, P.P. Lottici, J. Robichaud, Study of Anatase to Rutile Phase Transition in Nanocrystalline Titania Films, J. Sol-Gel Sci. Technol. 24 (2002) 255–264.

[22]

A. Verma, S.B. Samanta, A.K. Bakhshi, S.A. Agnihotry, Effect of stabilizer on structural, optical and electrochemical properties of sol-gel derived spin coated TiO2 films, Sol. Energy Mater. Sol. Cells. 88 18

(2005) 47–64. doi:10.1016/j.solmat.2004.10.006. [23]

B. Babiarczuk, A. Szczurek, A. Donesz-Sikorska, I. Rutkowska, J. Krzak, The influence of an acid catalyst on the morphology, wettabillity, adhesion and chemical structure properties of TiO2 and ZrO2 sol-gel thin films, Surf. Coatings Technol. 285 (2016) 134–145. doi:10.1016/j.surfcoat.2015.11.030.

[24]

R. Sun, Z. Chen, J. Peng, T. Zheng, The effect mechanisms of pH, complexant and calcination temperature on the hydrophilicity of TiO2 films prepared by the sol-gel method, Appl. Surf. Sci. 462 (2018) 480–488. doi:10.1016/j.apsusc.2018.08.163.

[25]

J. Yu, X. Zhao, Q. Zhao, Effect of surface structure on photocatalytic activity of TiO2 thin films prepared by sol-gel method, Thin Solid Films. 379 (2000) 7–14. doi:10.1016/S0040-6090(00)01542-X.

[26]

A.L. Anderson, R. Binions, The effect of Brij® surfactants in sol-gel processing for the production of TiO2 thin films, Polyhedron. 85 (2015) 83–92. doi:10.1016/j.poly.2014.08.050.

[27]

F. Di Giacomo, V. Zardetto, A. D’Epifanio, S. Pescetelli, F. Matteocci, S. Razza, A. Di Carlo, S. Licoccia, W.M.M. Kessels, M. Creatore, T.M. Brown, Flexible perovskite photovoltaic modules and solar cells based on atomic layer deposited compact layers and UV-irradiated TiO2 scaffolds on plastic substrates, Adv. Energy Mater. 5 (2015). doi:10.1002/aenm.201401808.

[28]

Y.H. Kim, J.S. Heo, T.H. Kim, S. Park, M.H. Yoon, J. Kim, M.S. Oh, G.R. Yi, Y.Y. Noh, S.K. Park, Flexible metaloxide devices made by room-temperature photochemical activation of sol-gel films, Nature. 489 (2012) 128–132. doi:10.1038/nature11434.

[29]

L.F. Cueto, E. Sánchez, L.M. Torres-Martínez, G.A. Hirata, On the optical, structural, and morphological properties of ZrO2 and TiO2 dip-coated thin films supported on glass substrates, Mater. Charact. 55 (2005) 263–271. doi:10.1016/j.matchar.2005.05.004.

[30]

E. Blanco, J.M. González-Leal, M. Ramírez-del Solar, Photocatalytic TiO2 sol-gel thin films: Optical and morphological characterization, Sol. Energy. 122 (2015) 11–23. doi:10.1016/j.solener.2015.07.048.

[31]

F. Sannino, P. Pernice, C. Imparato, A. Aronne, G. D’Errico, L. Minieri, M. Perfetti, D. Pirozzi, Hybrid TiO2acetylacetonate amorphous gel-derived material with stably adsorbed superoxide radical active in oxidative degradation of organic pollutants, RSC Adv. 5 (2015) 93831–93839. doi:10.1039/c5ra21176j.

[32]

A. Aronne, M. Fantauzzi, C. Imparato, D. Atzei, L. De Stefano, G. D’Errico, F. Sannino, I. Rea, D. Pirozzi, B. Elsener, P. Pernice, A. Rossi, Electronic properties of TiO2-based materials characterized by high Ti3+ selfdoping and low recombination rate of electron–hole pairs, RSC Adv. 7 (2017) 2373–2381. doi:10.1039/C6RA27111A. 19

[33]

C. Imparato, A. Aronne, L. Minieri, G. D’Errico, D. Pirozzi, F. Sannino, P. Pernice, Sol–Gel Synthesis and Characterization of Hybrid TiO2-Acetylacetonate Materials, Adv. Sci. Lett. 23 (2017) 5912–5915. doi:10.1166/asl.2017.9067.

[34]

N. Tohge, R. Ueno, F. Chiba, K. Kintaka, J. Nishii, Characteristics of diffraction gratings fabricated by the two-beam interference method using photosensitive hybrid gel films, J. Sol-Gel Sci. Technol. 19 (2000) 119–123. doi:10.1023/A:1008786827194.

[35]

H. Segawa, S. Adachi, Y. Arai, K. Yoshida, Fine patterning of hybrid titania films by ultraviolet irradiation, J. Am. Ceram. Soc. 86 (2003) 761–764. doi:10.1111/j.1151-2916.2003.tb03371.x.

[36]

H. Segawa, S. Inque, K. Watanabe, R. Ohashi, H. Nitani, M. Nomura, A study of photoreactions in photosensitive TiO2 hybrid gel films induced by UV irradiation, J. Ceram. Soc. Japan. 123 (2015) 793–799. doi:10.2109/jcersj2.123.793.

[37]

I.A. Mudunkotuwa, V.H. Grassian, Citric Acid Adsorption on TiO2 Nanoparticles in Aqueous Suspensions at Acidic and Circumneutral pH: Surface Coverage, Surface Speciation, and its Impact on Nanoparticle Nanoparticle Interactions, J. Am. Chem. Soc. 132 (2010) 14986–14994.

[38]

G.I. Spijksma, D.H.A. Blank, H.J.M. Bouwmeester, V.G. Kessler, Modification of different zirconium propoxide precursors by diethanolamine. Is there a shelf stability issue for sol-gel applications?, Int. J. Mol. Sci. 10 (2009) 4977–4989. doi:10.3390/ijms10114977.

[39]

R. Hayami, N. Endo, T. Abe, Y. Miyase, T. Sagawa, K. Yamamoto, S. Tsukada, T. Gunji, Zinc–diethanolamine complex: synthesis, characterization, and formation mechanism of zinc oxide via thermal decomposition, J. Sol-Gel Sci. Technol. 87 (2018) 743–748. doi:10.1007/s10971-018-4768-x.

[40]

A. Verma, M. Kar, D.P. Singh, Aging effect of diethanolamine derived precursor sol on TiO2 films deposited at different annealing temperatures, J. Sol-Gel Sci. Technol. 54 (2010) 129–138. doi:10.1007/s10971-0102167-z.

[41]

A. Mehner, W. Datchary, N. Bleil, H.W. Zoch, M.J. Klopfstein, D.A. Lucca, The influence of processing on crack formation, microstructure, density and hardness of sol-gel derived zirconia films, J. Sol-Gel Sci. Technol. 36 (2005) 25–32. doi:10.1007/s10971-005-4792-5.

[42]

C.Y. Wu, Y.L. Lee, Y.S. Lo, C.J. Lin, C.H. Wu, Thickness-dependent photocatalytic performance of nanocrystalline TiO2 thin films prepared by sol-gel spin coating, Appl. Surf. Sci. 280 (2013) 737–744. doi:10.1016/j.apsusc.2013.05.053.

[43]

A. Aronne, F. Sannino, S.R. Bonavolontà, E. Fanelli, A. Mingione, P. Pernice, R. Spaccini, D. Pirozzi, Use of 20

a new hybrid sol-gel zirconia matrix in the removal of the herbicide MCPA: A sorption/degradation process, Environ. Sci. Technol. 46 (2012) 1755–1763. doi:10.1021/es203223s. [44]

I. Diaz-Acosta, J. Baker, J.F. Hinton, P. Pulay, Calculated and experimental geometries and infrared spectra of metal tris-acetylacetonates: Vibrational spectroscopy as a probe of molecular structure for ionic complexes. Part II, Spectrochim. Acta - Part A Mol. Biomol. Spectrosc. 59 (2003) 363–377. doi:10.1016/S1386-1425(02)00166-X.

[45]

S.F. Tayyari, H. Rahemi, A.R. Nekoei, M. Zahedi-Tabrizi, Y.A. Wang, Vibrational assignment and structure of dibenzoylmethane. A density functional theoretical study, Spectrochim. Acta - Part A. 66 (2007) 394– 404. doi:http://dx.doi.org/10.1016/j.saa.2006.03.010.

[46]

Z. Guo, C. Li, S. Lu, Y. Pan, H. Gu, Citrate/F− assisted phase control synthesis of TiO2 nanostructures and their photocatalytic properties, RSC Adv. 5 (2015) 74230–74237. doi:10.1039/C5RA08816J.

[47]

J.M. Collins, R. Uppal, C.D. Incarvito, A.M. Valentine, Titanium(IV) citrate speciation and structure under environmentally and biologically relevant conditions, Inorg. Chem. 44 (2005) 3431–3440. doi:10.1021/ic048158y.

[48]

S. Varaganti, G. Ramakrishna, Dynamics of Interfacial Charge Transfer Emission in Small-molecule Sensitized TiO2 Nanoparticles: Is it Localized or Delocalized?, J. Phys. Chem. C. 114 (2010) 13917–13925.

[49]

D. Mergel, Modeling thin TiO2 films of various densities as an effective optical medium, Thin Solid Films. 397 (2001) 216–222. doi:https://doi.org/10.1016/S0040-6090(01)01403-1.

[50]

E. Blanco, M. Domínguez, J.M. González-Leal, E. Márquez, J. Outón, M. Ramírez-del-Solar, Insights into the annealing process of sol-gel TiO2 films leading to anatase development: The interrelationship between microstructure

and

optical

properties,

Appl.

Surf.

Sci.

439

(2018)

736–748.

doi:10.1016/j.apsusc.2018.01.058. [51]

B.E. Yoldas, D.P. Partlow, Formation of broad band antireflective coatings on fused silica for high power laser applications, Thin Solid Films. 129 (1985) 1–14. doi:https://doi.org/10.1016/0040-6090(85)90089-6.

[52]

I. Jeong, H. Jung, M. Park, J.S. Park, H.J. Son, J. Joo, J. Lee, M.J. Ko, A tailored TiO2 electron selective layer for high-performance flexible perovskite solar cells via low temperature UV process, Nano Energy. 28 (2016) 380–389. doi:10.1016/j.nanoen.2016.09.004.

[53]

J. Schneider, M. Matsuoka, M. Takeuchi, J. Zhang, Y. Horiuchi, M. Anpo, D.W. Bahnemann, Understanding TiO2

Photocatalysis :

Mechanisms

and

Materials,

doi:10.1021/cr5001892. 21

Chem.

Rev.

114

(2014)

9919–9986.

[54]

C. Garlisi, G. Palmisano, Radiation-free superhydrophilic and antifogging properties of e-beam evaporated TiO2 films on glass, Appl. Surf. Sci. 420 (2017) 83–93. doi:10.1016/j.apsusc.2017.05.077.

22

Captions to the scheme and figures.

Scheme 1. Molecular structures of the organic compounds used in the synthesis of the hybrid films. a) acetylacetone; b) dibenzoylmethane; c) citric acid; d) diethanolamine. Figure 1. SEM micrograph of the edge of a representative TiO2-acac film. Figure 2. AFM images and related profile of textures obtained of representative TiO2-acac (A) and TiO2-dbm (B) films. Figure 3. GIXRD patterns of TiO2-acac films annealed at 150 °C for 1 h (A-150), the same after 30 min UV irradiation (A-150-uv) and annealed at 400 °C for 1 h (A-400). Figure 4. FT-IR spectra of TiO2-acac and TiO2-dbm films as prepared (A and B) and after UV irradiation for 15 min (A-uv and B-uv). Figure 5. UV/visible/NIR reflectance, transmittance (a, c, e) and absorbance (b, d, f) spectra of TiO2-acac (A), TiO2dbm (B), TiO2-cit (C), TiO2-dea (D), TiO2 (T) and bilayer (T-A) films, as prepared and after UV irradiation. Figure 6. Refractive index (n) and extinction coefficient (k) vs. wavelength of TiO2 and hybrid TiO2 films: a) as prepared and b) after UV irradiation for 15 min. Figure 7. Electrical conductivity of selected films, as prepared and after UV irradiation.

23

In te n s ity ( a . u .)

(1 0 1 )

A -4 0 0 A -1 5 0 -u v A -1 5 0 1 0

2 0

3 0

4 0

2 θ(d e g re e s )

5 0

6 0

7 0

A -u v

B -u v

T r a n s m itta n c e ( a . u .)

T r a n s m itta n c e ( a . u .)

A

1 2 8 5 1 4 3 5

1 3 7 2

1 5 9 5

2 0 0 0

1 8 0 0

1 6 0 0

B 1 4 1 4 1 1 0 2 1 4 8 0

1 3 6 0

1 5 9 5

1 5 4 2 1 5 2 6

1 5 3 3

1 4 0 0

W a v e n u m b e r (c m

1 2 0 0 -1

)

1 0 0 0

8 0 0

2 0 0 0

1 8 0 0

1 6 0 0

1 4 0 0

W a v e n u m b e r (c m

1 2 0 0 -1

)

1 0 0 0

8 0 0

1 0 0

1 0 0

b

8 0

8 0

A 6 0

A -u v B

4 0

B -u v

A b s o rb a n c e (% )

R e fle c ta n c e , T r a n s m itta n c e ( % )

a

2 0

A A -u v

6 0

B B -u v

4 0

2 0

0 0 2 0 0

4 0 0

6 0 0

8 0 0

W

1 0 0 0

1 2 0 0

1 4 0 0

2 0 0

3 0 0

4 0 0

a v e le n g th ( n m ) W

5 0 0

6 0 0

a v e le n g th ( n m )

1 0 0

1 0 0

d 8 0

8 0

C C -u v

6 0

D 4 0

D -u v

A b s o rb a n c e (% )

R e fle c ta n c e , T r a n s m itta n c e ( % )

c

C C -u v

6 0

D D -u v

4 0

2 0

2 0

0 0 2 0 0

4 0 0

6 0 0

8 0 0

W

1 0 0 0

1 2 0 0

2 0 0

1 4 0 0

3 0 0

4 0 0

W

a v e le n g th ( n m )

5 0 0

6 0 0

a v e le n g th ( n m )

1 0 0

1 0 0

f 8 0

8 0

6 0

A b s o rb a n c e (% )

R e fle c ta n c e , T r a n s m itta n c e ( % )

e

T T -A

4 0

T

6 0

T -A 4 0

2 0

2 0

0 0 2 0 0

4 0 0

6 0 0

W

8 0 0

1 0 0 0

a v e le n g th ( n m )

1 2 0 0

1 4 0 0

2 0 0

3 0 0

4 0 0

W

a v e le n g th ( n m )

5 0 0

6 0 0

2.4

0.1

a)

T A B C D

as grown

2.3

--------- n ----- k

2.2

0.06

k

n

2.1

0.08

2

0.04

1.9 1.8

0.02

1.7 0

1.6 400

480

560

640

720

Wavelength (nm)

800

880

960

2.4

0.1

b)

T A B C D

after UV

2.3

--------- n ----- k

2.2 2.1

0.08

k

n

0.06

2

0.04

1.9 1.8

0.02

1.7 0

1.6 400

480

560

640

720

Wavelength (nm)

800

880

960

Graphical Abstract

Highlights

- TiO2-organic hybrid thin films with four different ligands were prepared by sol-gel - The addition of organic complexants gives long-term stability of the precursor sol - Drying at 80 °C gives hybrid amorphous films with enhanced functional properties - The nature of complexing ligand influences the microstructural and surface features - UV irradiation of the films modifies their structure and increases their conductivity