Photocatalytic Lime Render for Indoor and Outdoor Air Quality Improvement

Abstract

This article reports a novel photocatalytic lime render for indoor and outdoor air quality improvement that is composed of a lime binder and doped TiO₂ (KRONOClean 7000^®) nanoparticles. These nanoparticles were distributed throughout the bulk of the finishing render, instead of as a thin coating, thus ensuring the durability of the photocatalytic properties upon superficial damage. The physical properties of these renders were not affected by the addition of nanoparticles except in the case of surface area, which increased significantly. In terms of their photocatalytic activity, these novel lime renders were shown to degrade up to 12% NO_x under UV light and up to 11% formaldehyde under visible light.

1. Introduction

Buildings account for 40% of the total energy consumption in Europe [1,2,3,4]. The current need for a reduction in total energy consumption and the resulting demand for more energy-efficient buildings through increased air-tightness pose a series of technological challenges including those related to ensuring indoor air quality [5,6]. In particular, it has been shown that volatile organic compound (VOC) pollutants can build up in the interior of these buildings due to the significantly reduced ventilation rates [3,7,8,9]. Poor quality of indoor air is linked to a phenomenon known as sick building syndrome (SBS), which is generally identified by irritation to different mucosae and skin together with general sickness symptoms that can be correlated to the amount of time spent indoors [10,11].

Nowadays, adsorption, electrochemical oxidation processes, and filtration are the most commonly employed methods for the reduction and removal of air pollutants [12,13,14]. However, these approaches involve physical removal, thus necessitating replacement, cleaning, and lastly disposal of any consumable items such as filters. The additional cost associated with these operations is often considered a drawback. By contrast, the mineralization of said VOC pollutants into non-hazardous or at least less dangerous compounds shows promise as a more viable solution for removal [15,16].

Therefore, researchers are now developing new construction materials (particularly for use in the interior of energy-efficient buildings) that degrade, by means of photocatalysis, the concentration of harmful VOCs, such as formaldehyde, in air [17,18,19,20].

Since Honda and Fujishima [21,22] discovered the photocatalytic (PC) properties of titanium dioxide (TiO₂) for water splitting, strands of research have developed concerned with the direct application of PC particles to improve the quality of indoor air [23,24,25,26,27].

The large-scale use of TiO₂ can be attributed to its favorable cost versus photocatalytic activity ratio, non-toxicity, ease of extraction and diversity in synthesis [28], and modification methods [29]. TiO₂ is photocatalytically active within the ultraviolet (UV) range of the electromagnetic spectrum due to its wide band gap of 3.2 eV. However, it is unable to utilize the full solar spectrum where the UV range counts for only 5% [30]. This is often considered to be the major drawback of TiO₂, especially for internal use, as the ability of pure TiO₂ to neutralize significant proportions of contaminants in the indoor environment is greatly reduced. Strategies to enhance the photocatalytic performance of TiO₂ involves modification by various methods including doping with noble metals, transition metals, and non-metallic elements [31] combining with graphene [32], and the production of reduced TiO_n (1 < n < 2) sub-oxide phases [33]. Amongst these methods, Hermann showed that doping with transition metals is ineffective because of an increase in electron–hole recombination [34]. In contrast, noble metal modification and doping of TiO₂ with anions were more effective methods for harvesting visible-light [35,36,37]. The former method is based on the surface plasmon resonance (SPR) phenomenon of noble metals [38], and the latter method has been attributed to the presence of a foreign element in the lattice, leading to the formation of an additional electronic state in the band gap [39]. This allows electronic transitions to be initiated by visible light, improving the PC performance [29]. In fact, the authors have recently reported a study on the photocatalytic activity of modified TiO₂ under indoor lighting conditions [40,41,42]. Other strategies to improve the absorption of visible light foresee the coupling of two or more semiconductor materials [43], which is outside the scope of the present work.

TiO₂ is fully compatible with traditional construction materials [44], so its long-time adoption in combination with building materials is no surprise at all. The literature reports the application of TiO₂ on the surface of a wide range of building materials [45] for both interior and exterior use, including heavy clay materials [46], architectural stones [47], wall papers [48], and mortars (sensu lato) [49,50].

Extensive research [51,52] has explored the formulation of mortars containing pure TiO₂ in the form of nanoparticles (NPs) for improving the outdoor environment through the degradation of pollutants. However, this research mainly exploited PC activity using the UV portion of the solar irradiation and NO_x as model pollutants [52,53]. For instance, Park et al. investigated the NO_x removal of mortar mixed with TiO₂ using a UVA lamp [54]. Vieira et al. incorporated a commercial nano-TiO₂ (Evonik-previously Degussa, P25) into mortars and evaluated their PC activity against NO_x, using artificial solar light irradiation [55]. In addition, Lucas and co-workers tested PC activity against the NO_x degradation of mortars containing P25 TiO₂ using a lamp simulating solar radiation containing a UVA component [56]. Other research groups used organic dyes such as methylene blue or rhodamine B to monitor the PC activity of TiO₂-based mortars using UV lamps as the light source [56,57].

In this work, we report the use of nanotechnology in the development of a novel photocatalytic lime render with an aim to improve indoor air quality through the breakdown of formaldehyde [58,59] as well as outdoor air quality by NO_x abatement. Lime-based mortars and renders exhibit a number of important advantages over more commonly used cement-based materials such as vapor permeability and the ability to accommodate movement [60,61,62,63,64,65]. Their hardening over months or years by carbonation contributes to these properties along with the ability to self-heal (due to the limited solubility of calcium carbonate), which may enhance a long-term mechanical bonding with embedded nanoparticles [66,67,68,69,70]. Doped KRONOClean7000^® nanoparticles were selected in this study due to their suitability in terms of photocatalytic activity (under UV and visible light) as well as their commercial availability and potential for scale-up and real-scale use in buildings. Herein, we present a detailed account of the preparation of these novel photocatalytic lime renders, as well as their physical and chemical characterization and their photocatalytic activity in terms of indoor (VOC) and outdoor (NO_x) air purification, under visible and UV light, respectively. In conclusion, in this study, we show that a new generation of lime-based construction materials is now available for air quality improvement both in the interior and exterior of buildings.

2. Results and Discussion

2.1. Aesthetics: Color Variation

The color of the lime render is often of importance to the end user, and for this reason, the color variation of the test specimens was evaluated. Photographs of the macroscopic appearance of the photocatalytic samples ETDK2, ETDK3, and ETDK5 are compared to the reference sample ETD in Figure 1.

Color variation is represented using the CIELAB [71] color space described by three values a*, b*, and L* that describe all perceivable colors. The difference in color is achieved by comparison between sample values (L₂^*, a₂^*, and b₂^*) and a control (L₁^*, a₁^*, and b₁^*), which in this case was ETD. ΔE* is calculated using Equation (1) as reported by Sharma et al. [71]. When the value of ΔE* is over 2.3, it corresponds to a noticeable difference.

$${\mathsf{\Delta }\mathrm{E}}^{*}=\sqrt{{\left({\mathrm{L}}_2^{*}-{\mathrm{L}}_1^{*}\right)}^2+{\left({\mathrm{a}}_2^{*}-{\mathrm{a}}_1^{*}\right)}^2+{\left({\mathrm{b}}_2^{*}-{\mathrm{b}}_1^{*}\right)}^2},$$

(1)

The data from

Table 1. Color CIELAB parameters (ΔE* values over 2.3 correspond to a noticeable difference).

Sample	L*	a*	b*	ΔE*	Noticeable Difference
ETD	90.07	0.14	4.37	0	n/a
ETDK2	91.16	0.11	4.17	1.11	No
ETDK3	91.51	0.18	4.03	1.48	No
ETDK5	90.81	0.26	4.46	0.76	No

imply that the color change attributed to the presence of K7000 nanoparticles inside and on the surface of the photocatalytic mortar is not significant enough to be noticeable to the naked eye. Indeed, ΔE values 1 < ΔE < 2 mean that a difference in color can only be perceived by an experienced observer [72].

2.2. Mineralogy

X-Ray Diffraction (XRD) analyses determined that the cured reference Tradical Décor lime render (ETD) is composed mainly of quartz (SiO₂) accompanied by calcite (CaCO₃). Minor amounts of Portlandite (Ca(OH)₂), vaterite (CaCO₃), aragonite (CaCO₃), and rutile (TiO₂) were also detected as reported in Figure 2. The presence of Portlandite implies that full carbonation of the lime was not reached when analyses were performed after 28 days. This is quite common in limes where carbonation takes place over a period of time ranging from a few weeks to several years [73]. The rutile TiO₂ detected in the XRD patterns comes from the original Tradical Décor reference sample (ETD) and is present in the material, as it is needed for aesthetic purposes as a whitener (pigment). The addition of K7000 nanoparticles to the renders (ETD2, ETDK3, and ETDK5) did not lead to measurable changes in the XRD pattern. Hence, the reflection at a 2θ value of 25.5° which can be attributed to the anatase structure within the K7000 was not detected, this is likely because even at 5 wt % addition this was under the detection limit of that technique.

2.3. SEM Inspection of Lab-Scale Formulated Lime Renders

The cured Tradical Décor reference sample (ETD) has a smooth surface (Figure 3a) but when examined by Scanning Electron Microscopy (SEM) at high magnification (Figure 3b), the surface is perceived to be nodular and rough with abundant irregular pores. Energy Dispersive X-ray (EDX) analysis (Figure 3c–e (maps) and f (spectrum)) shows that this material contains mainly Ca, Si, O, C, and Ti. Other minor elements (K, S, Al, and Na) are also found. This composition agrees with the mineralogical analysis. Figure 3c shows that Ti is already present in the commercial Tradical Décor product and is uniformly distributed throughout the lime render sample. EDX maps in Figure 3c–e and Figure 4c–e correspond to the respective areas in Figure 3a and Figure 4a.

SEM inspection of the three Tradical Décor lime renders with added K7000 nanoparticles (ETDK2, ETDK3, and ETDK5) shows that the original microstructure of the reference lime render (ETD) is preserved. Images from ETDK3 are shown in Figure 4. Accordingly, the surface of these samples is smooth (Figure 3a and Figure 4a), with numerous nodules and irregular pores that add micro-roughness. These materials also contain mainly Ca, Si, O, and C (plus Ti, K, S, Al, and Na as minor elements). This composition agrees with the mineralogical results. EDX analysis shows that the titanium present in all these samples has a uniform distribution. It is not possible to distinguish the distribution of the added TiO₂ (K7000) because its EDX footprint is identical to that of the titanium in the Tradical Décor sample (Figure 3c and Figure 4c). However, from EDX analysis, it can be seen that the intensity of the Ti peak grows as the amount of K7000 is increased (Figure 3f and Figure 4f).

2.4. Mercury Intrusion Porosimetry of the Lime Renders

The porosity, density, average pore size, and pore distribution of the four Tradical Décor-based lime renders was measured by mercury intrusion porosimetry, the results of which are presented in

Table 2. Formulated lime renders: Mercury porosimetry analysis results.

Sample	Porosity (%)	Skeletal (g/cm3)	Bulk (g/cm3)	Average Pore Ø (µm)
ETD	29	2.5	1.8	1.0
ETDK2	30	2.5	1.7	0.9
ETDK3	30	2.5	1.7	0.9
ETDK5	33	2.5	1.7	0.8

.

The reference sample (ETD) has a porosity of 29% and an average pore size of 1.0 µm. The K7000 nanoparticle-containing renders have different porosity values. In the case of ETDK2 and ETDK3 (2 and 3 wt % K7000), the porosity is barely changed with respect to the blank sample (ETD), but in the case of ETDK5 (5 wt% K7000), an increase in porosity from 29% to 33% is observed. This porosity increase is coupled with a reduction of the average pore size (from 1.0 µm for ETD to 0.8 µm for ETDK5), as can be seen in Table 2 and Figure 5 below.

2.5. BET Surface Area Analysis Results

In Figure 6a, the Brunauer-Emmett-Teller (BET) adsorption/desorption isotherms of the reference and the three photocatalytic renders are presented. An isotherm represents the relationship between the amount of N₂ adsorbed by a unit mass of material and the equilibrium pressure at known temperature following the BET equation [74]. A change in the N₂ adsorption/desorption isotherms is observed for the different amounts of TiO₂ added to the lime renders, which led to different surface areas. The specific surface area values and respective standard deviation bars are presented in Figure 6b. The reference sample ETD showed the lowest surface area of 1.6 m²/g. In summary, the BET results (Figure 6) showed that surface area increased upon addition of the nanoparticles, i.e., an increase of 65% in the surface area was observed for the ETDK5 lime render with respect to the reference sample (ETD).

2.6. Adhesion Test Results

Given that photocatalytic reactions take place on the surface of the materials, the most efficient and economic use of a photocatalytic mortar is as a rendering material. Hence, a good adhesion of the photocatalytic render to the substrate is a key feature and will ensure that the rendering layer is firmly supported.

The results of the adhesion tests are shown in

Table 3. Adhesion test data for the formulated photocatalytic lime renders and the reference sample.

Sample	Test Point	Test Surface			Load (N)	Resistance (MPa)	Fracture Zone
		(mm)	(mm)	(mm2)
ETD	1	48	46	2208	231	0.10	100% PF80 substrate
	2	51	46	2346	258	0.11
	3	50	46	2300	270	0.12
	4	49	53	2597	307	0.12
Average					207	0.11
ETDK2	1	47	50	2350	307	0.13	100% PF80 substrate.
	2	49	46	2254	248	0.11	70% PF80 substrate–30% coating limit.
	3	48	47	2256	294	0.13	90% PF80 substrate–10% coating limit.
	4	50	45	2250	269	0.12	80% PF80 substrate–20% coating limit.
Average					280	0.12
ETDK3	1	50	50	2500	307	0.12	100% PF80 substrate
	2	49	44	2156	301	0.14
	3	50	50	2500	434	0.17
	4	48	48	2304	454	0.20
Average					374	0.16
ETDK5	1	48	47	2256	348	0.15	100% PF80 substrate
	2	51	45	2295	343	0.15
	3	50	49	2450	366	0.15
	4	50	46	2300	308	0.13
Average					341	0.15

. The (ETDK) photocatalytic lime renders in addition to the reference (ETD) all demonstrated good adhesion to the underlying PF80 substrate with a resistance ranging from 0.12 to 0.15 MPa. In addition, the photocatalytic renders have a superior adhesion to the substrate compared to the reference sample ETD. In fact, the adhesion tends to increase slightly (from 0.11 to 0.15 MPa) when the amount of K7000 increases.

The fracture surface of these samples usually takes place within the substrate, which implies that the formulated photocatalytic renders adhere strongly to the PF80 substrate.

2.7. Photocatalytic NO_x Degradation under UV Light

The results of the photocatalytic NO_x degradation tests for ETDK2, ETDK3, and ETDK5 samples are shown in Figure 7 and

Table 4. Results of the photocatalytic abatement of NO_x under UV light.

Sample	NO Reduction (%)	NOx Reduction (%)
ETD	7.1	5.4
ETDK2	7.7	5.4
ETDK3	11.2	10.6
ETDK5	13.1	12.3

. The average NO and NO_x conversion has been calculated for every formulation according to Equations (10) and (11) in Section 3.3.1.

The graphics in Figure 7 show the evolution of the concentration of the NO, NO₂, and NOx (NO + NO₂) during the test. The NO pollutant is fed into the NO_x reactor and when initially the light is off, it is adsorbed by the mortar, decreasing its concentration until it stabilizes once the pores are saturated. As soon as the UV light is turned on, the photocatalysis begins and the NO concentration decreases again, because it is converted to NO₂, whose concentration increases continuously while light and NO supply is kept until the test is interrupted after six hours. The NO_x degradation test results show that a 2 wt % addition of K7000 did not contribute any photocatalytic performance to the lime render, the NO and NO₂ reduction being the same as that measured in the blank ETD specimen—the measured reduction has to be ascribed to adsorption. However, increasing photocatalytic nanoparticle content from 2 to 5 wt % yields approximately twice the amount of NO_x degradation, i.e., sample ETDK5 was shown to degrade up to 12% NO_x under UV light.

2.8. Photocatalytic Formaldehyde Degradation under Visible Light

In the course of this test for photocatalytic formaldehyde degradation under visible light, four different phases were studied. In Phase 1, the chamber was flushed with neutral air when the intrinsic emissions from the specimen without exposure to light could be detected. In Phase 2, the chamber was flushed with neutral air, and the lamps (visible light) were turned on, irradiating the specimen surface in order to examine the light-induced emission of formaldehyde from the samples. In Phase 3, additional gas with an airflow of 1.65 L/min and a concentration of formaldehyde of approximately 90 ppb equivalent to a total of 12.2 µg/h of formaldehyde was fed over the surface of each sample with the lamps still turned on. Thus, at this stage, the photocatalytic degradation of formaldehyde by the sample could be detected. In Phase 4, the lights were switched off, and the addition of the formaldehyde was continued. Thus, the adsorption of formaldehyde by the sample could be determined at this stage. All phases were maintained until a stable formaldehyde concentration was reached for all samples. A total of 839 h was required to carry out all phases of the test. The formaldehyde degradation test profile is shown in Figure 8.

From the equilibration formaldehyde concentrations in the different testing phases, the desired variables are defined as shown below. For every phase, the resulting formaldehyde concentration can be calculated by Equation (2):

C_Res = C_in + C_ProdO + C_ProdL − C_RedL − C_RedA,

(2)

where

C_Res: resulting formaldehyde concentration

C_in: formaldehyde concentration measured in the empty reference chamber

C_ProdO: formaldehyde concentration from the self emission of the specimen

C_ProdL: formaldehyde concentration from sample emission caused by radiation

C_RedL: reduction of formaldehyde concentration caused by radiation

C_RedA: reduction of formaldehyde concentration caused by adsorption.

For the different phases of the experiment (see Figure 7), the resulting formaldehyde concentration can be calculated using the following equations:

C_Res(phase1) = C_ProdO,

(3)

C_Res(phase2) = C_ProdO + C_ProdL,

(4)

C_Res(phase3) = C_in + C_ProdO + C_ProdL − C_RedL − C_RedA,

(5)

C_Res(phase4) = C_in + C_ProdO − C_RedA.

(6)

The specimen performance parameters are obtained from the following equations:

C_RedL = C_Res(phase4) − C_Res(phase3) + C_Res(phase2) − C_Res(phase1),

(7)

C_ProdL = C_Res(phase2) − C_Res(phase1),

(8)

C_RedA = -C_Res(phase4) + C_Res(phase1) + C_in.

(9)

The equilibrium formaldehyde concentrations at the end of each experimental phase for all samples are listed in

Table 5. Formaldehyde equilibrium concentrations.

Sample	CRes(Phase 1) Own Emission without Light (ppb)	CRes(Phase 2) Own Emission with Light (ppb)	CRes(Phase 3) Own Emission with Light + Addition of Formaldehyde (ppb)	CRes(Phase 4) Own Emission without Light + Addition of Formaldehyde (ppb)
ETDK2	1	10	75	73
ETDK3	1	11	75	73
ETDK5	1	11	75	72

. Control experiments of the blank specimen (ETD) were also assessed; however, ETD data are not shown in Table 5, as this showed a negligible reduction in the formaldehyde initial concentration (<5%) due to adsorption.

Based on the equilibrium concentrations, the reduction rates of formaldehyde (either caused by photocatalytic reactions or by adsorption) can be calculated and are listed in

Table 6. Reduction of formaldehyde by photocatalysis and by adsorption of all samples.

Sample	CRedL (Equation (7)) (ppb) Cin = 90	Reduction by Photocatalysis (%)	Standard Deviation (%)	CRedA (Equation (9)) (ppb) Cin = 90	Reduction by Adsorption (%)	Standard Deviation (%)
ETD	6	6.7	2.7	16	17.8	2.4
ETDK2	9	10.0	2.5	18	20.0	2.5
ETDK3	10	11.1	3.5	18	20.0	2.9
ETDK5	9	10.0	2.5	19	21.1	2.4

for the different samples. Equations (7) and (9) were used, and a normalization in relation with the incoming air (90 ppb) was carried out to compare the results.

The reduction of formaldehyde by photocatalysis was found to be ca. 10–11% (Table 6) for all lime renders. This indicates that increasing the K7000 concentration from 3 to 5 wt % does not lead to an increase in photocatalytic performance. So, ETDK3 was selected as the optimum sample in terms of cost efficiency versus photocatalytic activity.

3. Experimental

3.1. Materials and Sample Preparation

The Tradical Décor lime render (supplied by BCB—Lhoist) was used to prepare the photocatalytic coating. Four specimens of these materials were prepared with different contents of carbon-modified TiO₂ nanoparticles KRONOClean 7000^® (K7000) with respect to the dry mixture (0, 2, 3 and 5 wt %). These titanium dioxide nanoparticles have a BET specific surface area >225 m²/g and anatase as the only crystalline phase. This consisted of a small fraction of anatase domains in the size range 11.9–15.5 nm, co-existing with a much larger quantity of domains of size 3.8 to 4.4 nm, with an amorphous phase content between 6.5 and 8.1 wt %. An aromatic carbon compound (likely aryl carboxylate species) is at the origin of the visible-light absorption as previously characterized by the authors of this manuscript [40].

For the photocatalytic render preparation, all the dry components were first mixed for approximately 3 min (using an automatic mixer), without any previous treatment; then, water was added in two steps, as described in the Tradical Décor technical data sheet [75]. Then, the resulting mixture was further mixed in an automatic mixer, which meets the European Norm: EN 196-1 [76]. Then, the fresh render was poured into molds (210 mm × 297 mm in size). The render thickness could range from 2 to 20 mm. For the photocatalytic characterization, the renders were prepared as 20 mm thick samples. Whereas for the adhesion measurements, the renders were prepared as 3 mm thick samples by manual application (i.e., spatula) onto a PF80 lime substrate. This substrate was made cutting fiber boards to normalized A4-size that were inserted in wood frames where later Tradical PF80 mortar (provided by BCB -Lhoist) was added as a 30 mm layer on the framed fiber boards and later cured for 28 days.

All the samples prepared were cured for at least 28 days in an environmental chamber at 22 °C (±2 °C), humidity of 53% (±3%) before any testing and characterization were carried out.

The samples obtained are referred to as ETD, ETDK2, ETDK3, and EDTK5 corresponding to the percentage of K7000 incorporated into the lime render (0, 2, 3 and 5 wt % respectively). The water/lime render ratios employed in their preparation were as follows: ETD and ETDK2: 0.22, ETDK3: 0.24, ETDK5: 0.26.

All these samples were characterized in terms of their physical properties and of their photocatalytic activity in the degradation of NO_x under UV light and of formaldehyde under visible light.

3.2. Characterization

Color variation of the photocatalytic samples was measured using a Color Eye XTH from Gretag Macbeth spectrophotometer with a D65-10 light source and compared to the blank sample. Three points at different locations on the surface of each sample were analyzed. XRD patterns and mineralogical information were obtained using a Philips X’Pert Pro MPD PW3040/60 X-ray diffractometer equipped with a copper anode operating with a voltage of 40 kV and 40 mA (1.6 kW). The scans were performed in a 2θ range from 2° to 75° with a step size of 0.02° 2θ. Representative chips of the surface of each sample were taken with a mass of approximately 2 g, ground with an automatic micro mill, and analyzed by X-ray powder diffraction. Scanning electron microscopy (SEM) images with correlated energy-dispersive X-ray (EDX) spectra were acquired using a FEI Quanta 200 ESEM with a Genesis 4000 system X-ray energy dispersive fluorescence spectrophotometer. Both the SEM images and the EDX spectra were acquired using uncoated samples at a sample chamber pressure of 60 Pa, accelerating voltage of 25 keV, and a working distance of 11 mm. For this study, representative flat fragments (size 20 × 20 mm) were taken from the surface of each sample.

Porosity, skeletal density, bulk density, and average pore size were measured by a Mercury Intrusion Porosimetry (MIP) using a Micromeritics Autopore IV 9500 model with a mercury-filling pressure range between 0.0036 and 228 MPa with an equilibration time of 10 s. For this work, representative fragments, each with a mass of approximately 2 g, were taken from the surface of each sample. This technique uses a glass bulb where the sample resides, and that is filled with mercury at increasing pressures while the volume of the mercury is monitored, yielding a mercury intrusion curve that correlates intruded mercury at different pressures and serves for calculating the different measurements by the software of the equipment (Autopore IV v1.09).

To measure the surface area, the amount of adsorbed N₂ required to cover the material surface is determined. The amount of gas adsorbed by the mass of material being tested is dependent on the relative pressure and temperature isotherm. The nitrogen BET surface area of the renders was measured using a Micrometrics 3Flex surface characterization analyzer following BS ISO 9277:2010 [77]. In total, 0.7 g was used for each sample. Prior to the surface area measurements, all samples were degassed to remove any water or other volatile molecules physically adsorbed on the surface. The degas phase was carried out in situ at 10 °C/min until 105 °C was reached after 120 min. Following the degassing phase, the adsorption and desorption cycle of nitrogen began. Then, the nitrogen BET surface area values were calculated using the 3Flex software version 3.02. Three samples of each specimen were analyzed for statistical purposes. The adhesion of the four lime renders to the Tradical PF 80 mortar substrate was measured. The layer of the lime render had a thickness of 3 mm. The adhesion tests were performed according to the norm EN 1015-12:2000 [78].

3.3. Evaluation of Photocatalytic Activity

3.3.1. NO_x Degradation Test

The test set-up used for the photocatalytic oxidation experiment for NO_x was according to the ISO Standard 22197-1: 2007 [79], which describes a test method for determining the air-purification performance of semiconducting photocatalytic materials. The experimental set-up is composed of the reactor, light source, target gas pollutant (NO) supply, transport gas (free pollutant air) supply, NO_x analyzer (CM2041 from Casella Monitors), flow meters, relative humidity control, and parameter measurement apparatus such as temperature and relative humidity.

The experimental procedure followed to carry out the NO_x degradation tests is also based on the test procedure described in ISO 22197-1, which concerns the development, comparison, quality assurance, characterization, reliability, and design data generation of photocatalytic materials. The method described is intended to assess the air-purification performance of photocatalytic materials by exposing a test piece to model polluted air under illumination by ultraviolet (UV) light. The supply gas flow rate (free pollutant air + NO) for these tests was 3 L/min, and the NO gas concentration was 1000 ppb. Nitric oxide (NO) is chosen as a target pollutant. The test piece, placed in a flow-type photoreactor, is activated by UV light, and it adsorbs and oxidizes gas-phase NO to form nitric acid (or nitrate) on its surface. A part of the NO is converted to nitrogen dioxide (NO₂) on the test piece.

Since the degradation of nitrogen oxides by photocatalysis leads to the oxidation of NO to NO₂ in the first reaction stage, the photocatalytic performance of these lime renders was assessed by two degradation rates calculated as follows (Equations (10) and (11)):

$$NO, conv \left(\%\right)= \frac{N{O}_{in}-N{O}_{out}}{N{O}_{in}}\ast 100,$$

(10)

$$N{O}_x, conv \left(\%\right)= \frac{N{O}_x{}_{in}-N{O}_x{}_{out}}{N{O}_x{}_{in}}\ast 100,$$

(11)

where NO_x = NO + NO₂.

The samples tested were ETDK2, ETDK3, and ETDK5.

3.3.2. Formaldehyde Degradation Test

A custom-made experimental set-up based on the ISO/18560-1 [80] was developed for the evaluation of photocatalytic formaldehyde degradation performance. Samples were placed in 28.2 L test chambers sealed on the top by a special type of borosilicate glass with a 100% transmission down to 350 nm. Adjustable slot floors in the test chamber allowed the surface of each individual sample to be located at a distance of 5 mm below the glass panel regardless of the sample thickness. The fluorescent tubes were placed on top of the glass windows so that the entire specimen surface was irradiated homogeneously. The air flow was fed through the gap between the glass panel and the surface of the sample in the test chamber, following which the exhaust was divided into two streams, allowing both on-line and off-line analysis. The air supplied into the test chamber was humidified and enriched with formaldehyde as required by blending humidified neutral air with the test gas. Gases were homogeneously mixed before entering the test chambers, as shown in Figure 9. The whole system included a total of six test chambers for parallel operation. A flow meter was connected in line with the gas supply to each chamber and adjusted in order to ensure that an equal flux of test gas was delivered to all six test chambers, as seen in Figure 10.

Neutral air that did not contain any formaldehyde or VOCs was provided by the institute’s compressed air system and guided through several cleaning steps to guarantee purity and ultimately through a humidification unit in which a dew point of 12.5 °C was set. This corresponds to a relative humidity of 50% at 23 °C. The test gas was enriched through the addition of calibration gases that contained approximately 15 ppm of formaldehyde in nitrogen. That calibration gas was diluted by cleaned neutral air to produce a test gas with a formaldehyde concentration of 90 ppb. The flow of both neutral air and the enriched test gas was regulated by mass flow controllers with a resulting flow through each chamber of 1.6 L/min. Tests were conducted under irradiation from Osram Lumilux warm white 36 W tubular fluorescent lamps of 26 mm diameter mounted to G13 bases, which ensured the entire irradiated area was illuminated evenly. This type of lamp is one commonly used for indoor lighting in offices and elsewhere.

The concentration of formaldehyde was measured using an AL4021 analyzer manufactured by Aero-Laser GmbH (Gesellschaft für Gasanalytik, Unterfeldstraße 12, D-82467 Garmisch-Partenkirchen, Germany). Fluorescent emission after the reaction of formaldehyde with acetylacetone and ammonia was measured in a linear detection range between 0.1 and 3000 ppb with a detection limit of 0.1 ppb. The analyzer was calibrated beforehand for a measurement range between 0.1 and 400 ppb.

4. Conclusions

Novel photocatalytic lime renders containing doped TiO₂ KRONOClean7000^® nanoparticles have been successfully prepared and fully characterized. Their photocatalytic activity, in the degradation of NO_x under UV light and of formaldehyde under visible light, has been studied; therefore, their suitability for air quality improvement both in outdoor and indoor environments has been examined.

No detrimental changes in physical properties (e.g., color) were observed upon the addition of 2, 3 and 5 wt % of these nanoparticles to the lime render. The results of mineralogical and EDX analysis showed few significant changes between the traditional and photocatalytic lime renders due to the presence of the anatase TiO₂ nanoparticles. In addition, the porosimetry studies only showed a small increase in porosity in the case of sample ETDK5 (5 wt %) with respect to the reference sample (ETD), i.e., porosity increased from 29% for ETD to 33% for ETDK5. The BET results showed that the surface area increases upon the addition of nanoparticles to the lime render, e.g., an increase of 65% in the surface area was observed for the ETDK5 lime render with respect to the reference sample (ETD). More importantly, it was also demonstrated that the presence of the K7000 nanoparticles within the chosen lime render does not negatively affect its adhesion to the lime mortar (PF80) substrate. Finally, these novel materials were shown to degrade up to 12% NO_x under UV light and 11% formaldehyde (VOC) under visible light.

In conclusion, this study has demonstrated that these novel photocatalytic lime renders are a technically viable solution for air quality improvement, both in the outdoor and indoor environments, under UV light and under conventional indoor lighting conditions, respectively.