Development of Photocatalytic 3D-Printed Cementitious Mortars: Inﬂuence of the Curing, Spraying Time Gaps and TiO 2 Coating Rates

: This work evaluated the photocatalytic activity of 3D-printed cementitious mortar specimens functionalized with TiO 2 nanoparticles to obtain a multifunctional and smart concrete. This research aims to assess the inﬂuence of different parameters related to the functionalization process such as adsorption, coating time gaps, and coating rates on the degradation efﬁciency of the functionalized cementitious specimens. Each specimen was evaluated under the degradation of Rhodamine B (RhB) in an aqueous solution using a sun-light simulation. The obtained results showed a decrease in adsorption (under dark condition) with increasing the sample curing age. The highest photocatalytic efﬁciency was observed for coated samples aged 7 days. By increasing the coating rates, the photocatalytic efﬁciency is enhanced. Nonetheless, regardless of the coating rates, all the specimens showed an increase in photocatalytic efﬁciency for longer time periods of light exposition, i.e., after 8 h of irradiation.


Introduction
Air pollutants, in addition to soot fouling, are some agents for both construction's lifespan reduction and deterioration of the buildings' architectural aesthetics [1,2]. Therefore, the need for preventing or reducing environmental pollutants to mitigate their problems and consequences has become an important concern for several industries and research groups [3][4][5]. The Architecture, Engineering, and Construction (AEC) sector presents a crucial role in this field since structures/buildings/infrastructures, i.e., the built environment, covers a huge area of urban centers. Therefore, the photocatalytic functionalization of construction materials for the built environment has great potential by taking advantage of solar energy in a useful manner by the degradation of different types of pollutants, such

Research Significance
In spite of the photocatalytic properties' evaluation of construction materials having started in the early 1990s [16], the research in this field is still ongoing, based on the development of new materials and techniques. For instance, combining photocatalytic technology with new construction methodologies, such as 3D concrete printing (3DCP) may result in the development of functionalized materials through advanced manufacturing techniques. In this case, if a volume incorporation approach is adopted, only the outer layer(s), which are visible, might be functionalized, leading to a more cost-effective material. On the other hand, if the functionalization is carried out by a surface-coating methodology, Buildings 2021, 11, 381 3 of 13 this could be employed simultaneously in an automatized way during the printing process, potentiating the photocatalytic efficiency due to the application of TiO 2 nanoparticles at an early age of the cementitious composites, which might not be possible in conventionally cast concretes. To the best of the authors' knowledge, the research within the latter field is still scarce [16,23]. Furthermore, the smart and multifunctional cementitious mortar will present photodegradation abilities of adsorbed and air pollutants, and a self-shaping capability, which has practical and cost benefits with prompting construction speed and reducing labour force.
The main scope of this study was to evaluate the potential of adding photocatalytic functionalization with focus on the RhB dye photodegradation efficiency to cementitious materials that have been developed for 3DCP [28]. The 3DCP technology as a freeform construction methodology allows a multitude of advantages; in this context, the specimen surface can be coated right away after finishing the printing process. This possibility can improve the quality of the bonding of the nanoparticles onto the surface of the cementitious specimens when they are still fresh. Therefore, the time gap between the printing process and spraying of the TiO 2 nanoparticles could be studied using 3DCP technology, which cannot be applicable to traditional mould cast concrete elements. In the literature, this coating method has also been applied for mould cast concrete specimens. However, for the latter, the semiconductor particles could only be sprayed after removing the moulds, i.e., after an initial hardening of the concrete. In this case, the bond quality between the particles and the specimens' surface could be lower, resulting in the separation of the particles from the surface. This phenomenon was reported under real environmental conditions, in particular due to wind and rain [3,16,29].
In this study, the time gap periods for spraying of TiO 2 nanoparticles, the variation of adsorption by curing age, and the influence of coating rates on the photocatalytic efficiency of the 3D-printed cementitious materials were studied. This represents a novelty for the scientific community regarding the photocatalytic properties of coated cementitious concretes, since the spraying of TiO 2 nanoparticles can be performed before the hardening of the concrete, on opposition to traditional cast concrete. It is important to emphasize that, as far as the authors are aware, this is one of the first times that nanoparticles of semiconductor materials are applied to functionalize cementitious mortars obtained by a 3D printer (in order to add new functions and/or properties).

Photocatalytic Process
The dye degradation promoted by the use of TiO 2 nanoparticles in cementitious materials occurs due to two mechanisms related to the photocatalytic activity of TiO 2 nanoparticles under light irradiation and through a dye-sensitised approach [22]. This degradation process is based on breaking chemical bonds of the dye chromophore with different bond energies [30]. TiO 2 has three electronic band structures, which include the valence band, conduction band and band gap [5,31]. The energy range between the lowest unoccupied state of the conduction band and the highest occupied energy state of the valence band influences the electrical conductivity of a semiconductor material, which will affect the photocatalytic behaviour of such a semiconductor, as is the case of TiO 2 .
In the first mechanism, the photocatalytic process starts when light photons with a specific wavelength irradiate the semiconductor nanoparticles to activate the electrons of the valence band [14]. If the energy of the absorbed photon (hv) by the semiconductor is equal to or greater than the band gap energy, electrons from the valence band can be promoted to the conduction band, resulting in the separation of electron (e − ) and hole (h + ) charge carriers [14,32,33]: The h + carriers have a powerful oxidation ability and are capable of degrading organic and inorganic compounds into water and other small harmless molecules [5].
The number and lifetime of the generated free (e − /h + ) pairs depend on the size of the semiconductor particles [14]. For large particles, the volume recombination of (e − /h + ) pairs is the dominant process (should be avoided). On the contrary, for smallsized particles, the covered distance by the (e − /h + ) pairs during their movement from the crystal interface to the particle surface is short; under this condition, the migration rate of the (e − /h + ) pairs to the surface of small particles is increased, thus minimizing their electronic recombination, which enhances the occurrence of redox reactions and, therefore, improving the photocatalytic efficiency.
In the presence of moisture (water), the highly reactive hydroxyl radicals (HO • ) can be formed due to the reaction between h + and H 2 O [6], see Equation (2). On the other hand, a superoxide ion (O − 2 ) can be produced due to the reaction between electron carriers and oxygen [6]; see Equation (3).
Both hydroxyl radicals and superoxide ions are highly reactive species that will oxidize the organic/inorganic compounds adsorbed on the semiconductor surface [5]. Moreover, the hydroperoxyl radicals, also known as hydrogen superoxides (HO 2 ), are created from the reaction between H + and O − 2 : Organic compounds can be fully degraded to carbon dioxide (CO 2 ) and water (H 2 O), when they react with hydroxyl radicals [2]: The reaction on Equation (5) can be invoked for different organic materials, such as oil and grease [34] and bio-microorganisms [12]. Figure 1 shows the overall diagram of heterogeneous photocatalysis of organic and inorganic compounds promoted by a semiconductor material. sized particles, the covered distance by the ( ℎ ⁄ ) pairs during their movement from the crystal interface to the particle surface is short; under this condition, the migration rate of the ( ℎ ⁄ ) pairs to the surface of small particles is increased, thus minimizing their electronic recombination, which enhances the occurrence of redox reactions and, therefore, improving the photocatalytic efficiency.
In the presence of moisture (water), the highly reactive hydroxyl radicals (HO • ) can be formed due to the reaction between h + and H2O [6], see Equation (2). On the other hand, a superoxide ion (O ) can be produced due to the reaction between electron carriers and oxygen [6]; see Equation (3).

ℎ + H O → HO • + H
(2) Both hydroxyl radicals and superoxide ions are highly reactive species that will oxidize the organic/inorganic compounds adsorbed on the semiconductor surface [5]. Moreover, the hydroperoxyl radicals, also known as hydrogen superoxides (HO2), are created from the reaction between H and O : Organic compounds can be fully degraded to carbon dioxide (CO2) and water (H2O), when they react with hydroxyl radicals [2]: The reaction on Equation (5) can be invoked for different organic materials, such as oil and grease [34] and bio-microorganisms [12]. Figure 1 shows the overall diagram of heterogeneous photocatalysis of organic and inorganic compounds promoted by a semiconductor material.
During the second mechanism, the electrons in the highest occupied molecular orbital level (known as HOMO level) of the dye are promoted to the lowest unoccupied molecular orbital level (known as LUMO level) and afterwards inserted within the conductance band of TiO2 [22]. Then, oxygen uses these electrons to create oxidative species that reduce and subsequently degrade the dye molecules, which are already partially reacted [22,35]. Dye sensitization occurs at lower energies of the visible light that are not sufficiently enough to activate the TiO2 particles, but can sensitize and degrade the chromophores of the dye.  During the second mechanism, the electrons in the highest occupied molecular orbital level (known as HOMO level) of the dye are promoted to the lowest unoccupied molecular orbital level (known as LUMO level) and afterwards inserted within the conductance band of TiO 2 [22]. Then, oxygen uses these electrons to create oxidative species that reduce and subsequently degrade the dye molecules, which are already partially reacted [22,35]. Dye sensitization occurs at lower energies of the visible light that are not sufficiently enough to activate the TiO 2 particles, but can sensitize and degrade the chromophores of the dye.

Materials
The mixture composition was defined in previous experimental works regarding the development and characterization of the mechanical behaviour of ternary cementitious mortars for 3D concrete printing [28,36]. Table 1 includes the constituents used in the selected mortar composition. TiO 2 nanoparticles (nano-TiO 2 ) were chosen as semiconductor photocatalysts in order to functionalize the 3D-printed cementitious mortar specimens. The TiO 2 nanoparticles (80% anatase and 20% rutile) were purchased from Quimidroga Portugal Lda. (Lisboa, Portugal). Diffuse reflectance as a simple and powerful spectroscopic tool was used to calculate the optical properties of the TiO 2 nanoparticles (namely its band gap energy, Eg). The diffused reflection measurements were acquired from a spectrophotometer ScanSpecUV-vis, ScanSci equipped with an integrating sphere assembly where barium sulphate was used as a reference. The E g of the photocatalyst was calculated based on Kubelka-Munk theory [37]. Due to the presence of oxygen vacancies, titanium dioxide (TiO 2 ) is an intrinsically n-type semiconductor material [38] with an energy bandgap equal to 3.2 eV (anatase crystalline phase-indirect allowed transition) and 3.0 eV (rutile crystalline phase-direct allowed transition) [39]. Therefore, the production of an electron in its conduction band and hole in its valence band requires the absorption of a photon with energy in the UV region of the electromagnetic spectrum. In this work, the calculated value for the energy gap of nano-TiO 2 was 3.19 eV, which is an intermediate value for both crystalline phases (anatase and rutile) because it results from a mixture of these two crystalline phases, being slightly lower than the value of the anatase phase.
The nano-TiO 2 aqueous suspensions were prepared based on different concentrations of nanoparticles, namely, 4 g/L, 8 g/L and 16 g/L. A pH of 8 was adopted for all the suspensions. The choice for this pH value was based on its ability to prevent the aggregation of TiO 2 nanoparticles, which results in more stable colloidal suspensions. In fact, at a pH of 8 the zeta potential of the TiO 2 nanoparticles dispersion is negative [8] and, therefore, repulsive electrostatic forces (Coulomb forces) arise that help to promote a better dispersion of the TiO 2 nanoparticles.

Sample Preparation
The mortar specimens were manufactured through a small 3D concrete printer. The specification and characteristics of the 3D printer can be found elsewhere [28]. In each printing batch, several cementitious mortar strips were printed. Each strip was printed in two layers on top of each other with the dimensions of 250 × 25 × 15 mm 3 . The strips were cut right after printing in order to prepare the final specimens with adequate dimensions (i.e., 25 × 25 × 15 mm 3 ) for the photocatalytic tests ( Figure 2a). ous solutions onto the surface of the specimens. The distinct time gaps were selected because the matrices have distinct maturation times, from the time that they are still fresh up to the hardened stage. In this case, the first series of the specimens was coated after 1.5 h of the hydration process initiation. A preliminary test on the cementitious mortar mixture showed an initial setting time of about 5 h. Therefore, the second series was sprayed after 5 h at about the initial of setting time. The other four series were coated at 9, 24, 32 h, and 7 days after the start of hydration process.

Methodology for Evaluating the Photocatalytic Efficiency
The photocatalytic behaviour of the functionalized cementitious mortar specimens was evaluated based on their capability to promote the photodegradation of a particular organic compound. In this work, RhB dye (a cationic dye, acquired from Merck S.A., Algés, Portugal) has been selected as the pollutant model (due to their wide use in photocatalytic tests) to evaluate the photocatalytic ability of the TiO2 coated-cement mortar specimens. The RhB, which has quite low sensitivity to the alkaline nature of the cementitious materials [4,26], was used in the form of an aqueous solution with concentration of 5 mg/L (5 ppm). The RhB degradation rate in this aqueous solution was measured based on the maximum absorbance of light at a wavelength of 554 nm. All samples were immersed in 30 mL of the RhB solution. After 3 h of immersion under dark conditions, the test specimens were irradiated by an OSRAM UltraVitalux lamp with 300 W, which simulates the solar irradiation. By keeping a vertical distance of 25 cm between the top surface of the samples and the light source, the specimens were irradiated by light power density The spraying coating technique was used as the functionalization method of the samples. The coating was applied by spraying a TiO 2 aqueous suspension onto the specimens' surfaces based on our previous work carried out over asphalt mixtures developed by [8,40,41]. This method of functionalizing material surfaces consists of spraying the TiO 2 aqueous suspension onto the specimens' surfaces using an atmospheric air compressor (Figure 2b). The distance between the tip of the air compressor and the top surface of the specimens was kept constant at 20 cm, and the spraying was continued for 30 s. Using these parameters, the speed of the aqueous suspension jet was equal to 100 mL/min, which could create coating rates of about 5 mg/cm 2 to 80 mg/cm 2 onto the surface of specimens, based on the known volume of TiO 2 aqueous suspensions and the sprayed area. Moreover, in order to evaluate the influence of spraying time gaps on the photocatalytic behaviour of cementitious materials, six different time gaps were chosen to spray the TiO 2 aqueous solutions onto the surface of the specimens. The distinct time gaps were selected because the matrices have distinct maturation times, from the time that they are still fresh up to the hardened stage. In this case, the first series of the specimens was coated after 1.5 h of the hydration process initiation. A preliminary test on the cementitious mortar mixture showed an initial setting time of about 5 h. Therefore, the second series was sprayed after 5 h at about the initial of setting time. The other four series were coated at 9, 24, 32 h, and 7 days after the start of hydration process.

Methodology for Evaluating the Photocatalytic Efficiency
The photocatalytic behaviour of the functionalized cementitious mortar specimens was evaluated based on their capability to promote the photodegradation of a particular organic compound. In this work, RhB dye (a cationic dye, acquired from Merck S.A., Algés, Portugal) has been selected as the pollutant model (due to their wide use in photocatalytic tests) to evaluate the photocatalytic ability of the TiO 2 coated-cement mortar specimens. The RhB, which has quite low sensitivity to the alkaline nature of the cementitious materials [4,26], was used in the form of an aqueous solution with concentration of 5 mg/L (5 ppm). The RhB degradation rate in this aqueous solution was measured based on the maximum absorbance of light at a wavelength of 554 nm. All samples were immersed in 30 mL of the RhB solution. After 3 h of immersion under dark conditions, the test specimens were irradiated by an OSRAM UltraVitalux lamp with 300 W, which simulates the solar irradiation. By keeping a vertical distance of 25 cm between the top surface of the samples and the light source, the specimens were irradiated by light power density equal to 1 mW/cm 2 (measured by a Quantum Photo Radiometer HD9021 Delta Padova). During the irradiation phase, a transparent plastic film covered all the cups containing the RhB solution and the immersed specimens. The plastic film was used to prevent the evaporation of RhB aqueous solution due to the generated heat flux by the lamp. The type of used plastic film could transmit at least 90% of the incident light in the wavelength range between 292 and 900 nm, so it presents a negligible effect on the light wavelength and density received by the test specimens to initiate the photocatalysis.
The test specimens were irradiated for 8 h. After each time period of 1, 2, 4, and 8 h, an aliquot of 3 mL of RhB aqueous solution was extracted from the cup, and its light absorbance was measured in the wavelength ranges of 300 to 800 nm using a Shimadzu UV-310PC scanning spectrophotometer. Afterwards, the photocatalytic efficiency was calculated according to Equation (6) based on the maximum absorbance value at the wavelength of 554 nm using the Beer-Lambert law [8] where Φ is the photocatalytic efficiency [%]; A 0 and A are the maximum absorbance of the light by RhB aqueous solution at the time zero and time t, respectively. The time zero was considered as the time of starting the light irradiation at the end of adsorption/desorption equilibrium process. Figure 3 shows the variation in the specimens' adsorption (reference and functionalized) based on their curing ages, as well the average value (avg) of adsorption for the different series. The results show a reduction of about 74% in the adsorption of the specimens by increasing the curing ages from 2 days to 90 days. However, the reduction rate of the adsorption decreases with the increase in the curing age. The adsorption decreases about 14% from a curing age of 28 days to 90 days. The reduction in the adsorption can be ascribed to the change of the microstructure in cementitious materials during time due to the ongoing hydration process decreasing even their hygroscopic behaviour [7]. During the hydration process, the water inside the pores is substituted by the accumulation of hydration products [11]. Therefore, by increasing the curing time, the decrease in the porosity results in a reduction in the internal absorption and surface adsorption of the specimens. A reduction in the amount of both capillary pores and total volume of pores is reported by [11,42] due to the increase in the curing age in cementitious materials. In a previous study [28] which concerns the mechanical/physical properties of the same cementitious mixture composition of this research, a reduction of about 46% was observed in the porosity of 3D-printed specimens from the curing age of 3 days to 28 days.

Influence of Curing Age on Adsorption
The results also reveal that the reduction in the adsorption has occurred for both the reference and the TiO 2 -coated specimens. Nonetheless, the adsorption of TiO 2 -coated specimens was higher than the reference ones for each specific curing age. The only exception was observed at the curing age of 14 days, in which the reference specimens showed higher adsorption than the functionalized specimens. The higher adsorption of functionalized specimens may be attributed to the presence of the TiO 2 nanoparticles on the specimen surfaces. The high specific surface area of the TiO 2 nanoparticles on the coated specimens could enhance the adsorption through the negatively charged surface of TiO 2 particles [43] when compared to the reference specimens without any kind of coating. For longer curing times (i.e., 28 and 90 days) the reduction in the dye concentration is mainly due to the decomposition of dye adsorbed on the surface of the specimen through the photocatalytic activity of semiconductors. However, at early ages, a higher amount of dye is absorbed by the porous structure of the material due to the existence of larger pores and a higher amount of pore volume. Therefore, it is more difficult to distinguish between the real photocatalytic behaviour of the material due to semiconductor inherent photocatalytic activity and the dye degradation due to internal absorption. Moreover, larger pores present mainly during the early curing ages may absorb the dye molecules into the matrix in deeper zones that neither the semiconductor particles (in the coating approach) nor the light can reach, which was reported as an adverse effect of the large pores against the photocatalytic efficiency [9]. The results in tabular format can be found in the supplementary material as Table S1. photocatalytic activity and the dye degradation due to internal absorption. Moreover, larger pores present mainly during the early curing ages may absorb the dye molecules into the matrix in deeper zones that neither the semiconductor particles (in the coating approach) nor the light can reach, which was reported as an adverse effect of the large pores against the photocatalytic efficiency [9]. The results in tabular format can be found in the supplementary material as Table S1.

Influence of the Spraying Time Gap on the Photocatalytic Efficiency
The influence of the spraying time gap on the photocatalytic behaviour was assessed for the same coating rate of 5 mg/cm 2 . The coating time gap was defined based on the initiation of the hydration process, namely, since the initial instant when water was added to the mixture composition. The distinct spraying time gaps were defined in order to have different surface conditions of the 3D-printed specimens, from the early ages, in which the mixture is completely fresh, up to when it is hardened (See Section 4.2). The specimens coated at distinct time gaps of 1.5, 5, 9, 24, 32 hours and 7 days were named as SP_1.5 h, SP_5 h, SP_9 h, SP_24 h, SP_32 h, and SP_7 d, respectively. The photocatalytic test was carried out for all the specimens after 28 days of curing process. Figure 4 shows the photocatalytic efficiency of the specimens based on the distinct spraying time gaps. All the specimens showed an enhancement of the photocatalytic efficiency from the beginning of light irradiation. During the entire time of the testing procedure, the specimens coated at the age of 7 days showed the highest photocatalytic efficiency when compared to the specimens coated with fewer days of curing. At the end of the testing procedure, after 8 h of light irradiation, the series SP_9 h showed the lowest photocatalytic efficiency of about 19%, which translates to a difference of about 47% between the maximum (i.e., SP_7 d) and the minimum (i.e., SP_9 h) photocatalytic efficiencies.
Regardless of the specimen SP_7 d, the results obtained by the other specimens revealed similar photocatalytic behaviour. In this case, the maximum difference between the efficiencies of specimens coated at the ages of 32 (SP_32 h) and 9 h (SP_9 h) was about 15%. On the other hand, the photocatalytic efficiency of specimens SP_32 h was just 4% higher than the efficiency of specimens coated at the age of 1.5 h (SP_1.5 h), which could be considered negligible.
The similarity of the results reveals that the difference in coating time does not extensively influence the photocatalytic efficiency. However, the better behaviour observed by specimen SP_7 d shows the need for further research for larger time gaps. Nonetheless, the coating at the beginning of the hydration process, when the mixture is still fresh, can improve the immobilization of TiO2 nanoparticles to the surface of mortar specimens. Moreover, the alkaline nature of the cementitious matrices can properly create a suitable

Influence of the Spraying Time Gap on the Photocatalytic Efficiency
The influence of the spraying time gap on the photocatalytic behaviour was assessed for the same coating rate of 5 mg/cm 2 . The coating time gap was defined based on the initiation of the hydration process, namely, since the initial instant when water was added to the mixture composition. The distinct spraying time gaps were defined in order to have different surface conditions of the 3D-printed specimens, from the early ages, in which the mixture is completely fresh, up to when it is hardened (See Section 4.2). The specimens coated at distinct time gaps of 1.5, 5, 9, 24, 32 hours and 7 days were named as SP_1.5 h, SP_5 h, SP_9 h, SP_24 h, SP_32 h, and SP_7 d, respectively. The photocatalytic test was carried out for all the specimens after 28 days of curing process. Figure 4 shows the photocatalytic efficiency of the specimens based on the distinct spraying time gaps. All the specimens showed an enhancement of the photocatalytic efficiency from the beginning of light irradiation. During the entire time of the testing procedure, the specimens coated at the age of 7 days showed the highest photocatalytic efficiency when compared to the specimens coated with fewer days of curing. At the end of the testing procedure, after 8 h of light irradiation, the series SP_9 h showed the lowest photocatalytic efficiency of about 19%, which translates to a difference of about 47% between the maximum (i.e., SP_7 d) and the minimum (i.e., SP_9 h) photocatalytic efficiencies.
Regardless of the specimen SP_7 d, the results obtained by the other specimens revealed similar photocatalytic behaviour. In this case, the maximum difference between the efficiencies of specimens coated at the ages of 32 (SP_32 h) and 9 h (SP_9 h) was about 15%. On the other hand, the photocatalytic efficiency of specimens SP_32 h was just 4% higher than the efficiency of specimens coated at the age of 1.5 h (SP_1.5 h), which could be considered negligible. tained for the specimens coated at the age of 7 days could also demonstrate the efficiency of this type of functionalizing method for hardened cementitious specimens. The spraying of TiO2 aqueous suspension after drying the cementitious material was also tried [3]. However, a poor dye degradation behaviour was reported, which results from the partial separation of the semiconductors' surface layer due to exposure to environmental conditions such as wind and rain [3]. The results in tabular format can be found in the supplementary material as Table S2.  Figure 5 shows the photocatalytic efficiency of functionalized specimens with different TiO2 coating rates of 10, 20, 40 and 80 mg/cm 2 (designated as CR10, CR20, CR40 and CR80, respectively). An increase in the photocatalytic efficiency was observed from the beginning of the light irradiation procedure for all the TiO2-coated specimens regardless of their coating rates. Moreover, the reference specimen without surface coating also showed an increase in degradation efficiency up to the end of 8 h of light irradiation.

Influence of the TiO2 Coating Rates on the Photocataytic Efficiency
When comparing all the series, a similar photocatalytic behaviour by all the TiO2coated and reference specimens was observed up to 4 h of irradiation. Increasing the time period of light irradiation up to 8 h produced more significant differences in the photocatalytic efficiency of the different series. After 8 h of light irradiation, the highest efficiency of about 21% is ascribed to CR80. The lowest photocatalytic efficiency was observed for CR10 and the reference series, both around 17%. Therefore, for a larger irradiation period of 8 h, a difference of about 26% was observed between the maximum and the minimum photocatalytic efficiencies of the specimens. Moreover, regarding the coating rates, an increase in the photocatalytic efficiency by increasing the coating rates of TiO2 nanoparticles was observed. The increase in photocatalytic efficiency by increasing the coating rates was also reported by [8] for the TiO2 surface coated asphalt samples.
The lack of enough TiO2 nanoparticles coverage on the surface of the specimens of these series can be pointed out as a hypothesis for the similarity between the photocatalytic behaviour of the reference and the CR10 series. Nonetheless, further assessment is necessary to find the main reason for this behaviour.
The similar photocatalytic efficiency for all the series up to 4 h of light irradiation may be ascribed to the accumulation of adsorbed pollutants and the produced dye decomposition products onto the specimens' surface, which could partially inactivate the reaction sites [44]. Nonetheless, by increasing the light exposure period up to 8 h, the The similarity of the results reveals that the difference in coating time does not extensively influence the photocatalytic efficiency. However, the better behaviour observed by specimen SP_7 d shows the need for further research for larger time gaps. Nonetheless, the coating at the beginning of the hydration process, when the mixture is still fresh, can improve the immobilization of TiO 2 nanoparticles to the surface of mortar specimens. Moreover, the alkaline nature of the cementitious matrices can properly create a suitable substrate for the immobilization of the TiO 2 nanoparticles [6]. Nevertheless, these aspects need to be evaluated. As the 3D-printing technology is a free-form construction process [28], it is feasible to spray the TiO 2 aqueous suspension immediately after printing the elements in order to improve the bonding properties. On the other hand, the results obtained for the specimens coated at the age of 7 days could also demonstrate the efficiency of this type of functionalizing method for hardened cementitious specimens. The spraying of TiO 2 aqueous suspension after drying the cementitious material was also tried [3]. However, a poor dye degradation behaviour was reported, which results from the partial separation of the semiconductors' surface layer due to exposure to environmental conditions such as wind and rain [3]. The results in tabular format can be found in the supplementary material as Table S2. Figure 5 shows the photocatalytic efficiency of functionalized specimens with different TiO 2 coating rates of 10, 20, 40 and 80 mg/cm 2 (designated as CR10, CR20, CR40 and CR80, respectively). An increase in the photocatalytic efficiency was observed from the beginning of the light irradiation procedure for all the TiO 2 -coated specimens regardless of their coating rates. Moreover, the reference specimen without surface coating also showed an increase in degradation efficiency up to the end of 8 h of light irradiation.

Influence of the TiO 2 Coating Rates on the Photocataytic Efficiency
When comparing all the series, a similar photocatalytic behaviour by all the TiO 2coated and reference specimens was observed up to 4 h of irradiation. Increasing the time period of light irradiation up to 8 h produced more significant differences in the photocatalytic efficiency of the different series. After 8 h of light irradiation, the highest efficiency of about 21% is ascribed to CR80. The lowest photocatalytic efficiency was observed for CR10 and the reference series, both around 17%. Therefore, for a larger irradiation period of 8 h, a difference of about 26% was observed between the maximum and the minimum photocatalytic efficiencies of the specimens. Moreover, regarding the coating rates, an increase in the photocatalytic efficiency by increasing the coating rates of TiO 2 nanoparticles was observed. The increase in photocatalytic efficiency by increasing the coating rates was also reported by [8] for the TiO 2 surface coated asphalt samples. reaction products generated by the degradation mechanism had enough time to be removed from the surface of the specimen, and the TiO2 nanoparticles could be reactivated in order to degrade a higher quantity of pollutants. Clearer differences between the photocatalytic efficiencies after 8 h of irradiation can be a confirmation for the observed behaviour. The results in tabular format can be found in the supplementary material as Table  S3.

Conclusions
In this study, the photocatalytic behaviour of 3D-printed cementitious mortar specimens coated with TiO2 nanoparticles was evaluated under their photocatalytic efficiency. Different parameters, such as adsorption against curing age, coating time gap (based on the initiation of the matrix hydration), and coating rates were evaluated regarding their influence on photocatalytic behaviour. The obtained results can be summarized as follows: • Adsorption in cementitious materials decreases with increasing curing age. This can be ascribed to the change in the matrix microstructure while the hydration process is ongoing. However, the adsorption of the functionalized specimens coated with TiO2 nanoparticles was higher when compared to the reference specimens, regardless of the curing ages that have been analysed. The high specific surface area of TiO2 nanoparticles could be the main reason for higher adsorption on the functionalized specimens.

•
The coating time measured from the beginning of hydration process did not have a significant influence on the photocatalytic efficiency. The only exception occurred for the specimens SP_7d, i.e., coated for 7 days, which showed 47% higher photocatalytic efficiency when compared to the other specimens. Nonetheless, the coating at the beginning of the hydration process when the mixture is still fresh can improve the bonding of TiO2 particles to the surface of the cementitious mortar specimens. Therefore, using the advantages of the 3D printing technology as a free-form construction methodology, the TiO2 aqueous suspension can be sprayed immediately after printing the elements. On the other hand, adequate results obtained for the specimens coated and aged for 7 days could also demonstrate the efficiency of this type of functionalizing method for hardened cementitious matrices. However, TiO2 particles sprayed onto the dried cementitious surfaces may be more prone to be separated from the specimens, which may lead to a decrease in photocatalytic efficiency of samples exposed to real environmental conditions. • The photocatalytic efficiency could be enhanced by increasing the coating rates of the TiO2 nanoparticles onto the surface of the specimens. After 8 h of light irradiation, photocatalytic efficiencies of 17% and 21% were obtained for the lowest and the The lack of enough TiO 2 nanoparticles coverage on the surface of the specimens of these series can be pointed out as a hypothesis for the similarity between the photocatalytic behaviour of the reference and the CR10 series. Nonetheless, further assessment is necessary to find the main reason for this behaviour.
The similar photocatalytic efficiency for all the series up to 4 h of light irradiation may be ascribed to the accumulation of adsorbed pollutants and the produced dye decomposition products onto the specimens' surface, which could partially inactivate the reaction sites [44]. Nonetheless, by increasing the light exposure period up to 8 h, the reaction products generated by the degradation mechanism had enough time to be removed from the surface of the specimen, and the TiO 2 nanoparticles could be reactivated in order to degrade a higher quantity of pollutants. Clearer differences between the photocatalytic efficiencies after 8 h of irradiation can be a confirmation for the observed behaviour. The results in tabular format can be found in the supplementary material as Table S3.

Conclusions
In this study, the photocatalytic behaviour of 3D-printed cementitious mortar specimens coated with TiO 2 nanoparticles was evaluated under their photocatalytic efficiency. Different parameters, such as adsorption against curing age, coating time gap (based on the initiation of the matrix hydration), and coating rates were evaluated regarding their influence on photocatalytic behaviour. The obtained results can be summarized as follows: • Adsorption in cementitious materials decreases with increasing curing age. This can be ascribed to the change in the matrix microstructure while the hydration process is ongoing. However, the adsorption of the functionalized specimens coated with TiO 2 nanoparticles was higher when compared to the reference specimens, regardless of the curing ages that have been analysed. The high specific surface area of TiO 2 nanoparticles could be the main reason for higher adsorption on the functionalized specimens.

•
The coating time measured from the beginning of hydration process did not have a significant influence on the photocatalytic efficiency. The only exception occurred for the specimens SP_7d, i.e., coated for 7 days, which showed 47% higher photocatalytic efficiency when compared to the other specimens. Nonetheless, the coating at the beginning of the hydration process when the mixture is still fresh can improve the bonding of TiO 2 particles to the surface of the cementitious mortar specimens. Therefore, using the advantages of the 3D printing technology as a free-form construction methodology, the TiO 2 aqueous suspension can be sprayed immediately after printing the elements. On the other hand, adequate results obtained for the specimens coated and aged for 7 days could also demonstrate the efficiency of this type of functionalizing method for hardened cementitious matrices. However, TiO 2 particles sprayed onto the dried cementitious surfaces may be more prone to be separated from the spec-imens, which may lead to a decrease in photocatalytic efficiency of samples exposed to real environmental conditions. • The photocatalytic efficiency could be enhanced by increasing the coating rates of the TiO 2 nanoparticles onto the surface of the specimens. After 8 h of light irradiation, photocatalytic efficiencies of 17% and 21% were obtained for the lowest and the highest coating rates of 10 and 80 mg/cm 2 , respectively. Similar photocatalytic behaviour for all the series up to 4 h of irradiation was observed. This behaviour may be ascribed to the accumulation of adsorbed pollutants and the produced dye decomposition products onto the specimens' surface, which could partially inactivate the reaction sites. Nonetheless, by increasing the light exposure period up to 8 h, the reaction products created from the degradation mechanism have enough time to be removed from the surface of the specimen and the TiO 2 nanoparticles could be reactivated again in order to degrade a higher quantity of pollutants.
The combination of two techniques, namely the application of TiO 2 and 3D printing, is an innovative topic of smart materials by the development of two new capabilities to mortars, which are photocatalytic and also 3D printable. These functionalized surfaces could photodegrade gases, such as SO x and NO x , and adsorb organic compounds, cleaning the air and maintaining the aesthetic aspects of buildings and infrastructures.