Chemical composition of cool coatings and the influence on solar reflectance and thermal emittance

Composição química de revestimentos frios e a influência na refletância solar e emitância térmica

Abstract

Cool materials have higher values of solar reflectance and lower values of thermal emittance, although still high when compared to conventional coatings, thus, their surfaces are less heated. In this regard, the goal was to identify the chemical elements present on cool surfaces that influence solar reflectance and thermal emittance. To this end, the spectral reflectance and the thermal emittance were measured with methods and equipment standardized by the Cool Roof Rating Council (CRRC) of 23 elastomeric cool coatings commercialized in the Brazilian market, and the chemical composition of the surfaces with EDS (Energy-dispersive X-ray detector) coupled to the field emission scanning electron microscope (FE-SEM) was analyzed. The presence of the metallic chemical element (titanium) in greater quantities on the surfaces was not decisive for lower thermal emittance values. The surfaces of materials with lower carbon and higher oxygen content have higher near infrared and solar reflectance. Therefore, surfaces with a matte finish have higher solar reflectance values because of the higher ratio of pigments in relation to resin than those with a glossy finish, which are mistakenly associated with higher reflectance due to its high surface gloss.

Keywords:
Cool materials; Solar reflectance; Thermal emittance; Chemical composition.

Resumo

Os materiais frios possuem maiores valores de refletância solar e menores, apesar de elevados, de emitância térmica comparados aos revestimentos convencionais, desta forma, aquecem menos suas superfícies. Assim, o objetivo foi identificar os elementos químicos presentes nas superfícies frias que influenciam na refletância solar e emitância térmica. Para isso, foram medidas a refletância espectral e a emitância térmica com métodos e equipamentos normatizados pelo Cool Roof Rating Council (CRRC) de 23 revestimentos frios elastoméricos do mercado brasileiro e analisada a composição química das superfícies com EDS (detector de energia dispersiva de raios-x) acoplado ao microscópio eletrônico de varredura por emissão de campo (SEM-FEG). A presença do elemento químico metálico (titânio) em maiores quantidades nas superfícies não foi determinante para menores valores de emitância térmica. As superfícies dos materiais com menor quantidade de carbono e maior em oxigênio possuem maior refletância no infravermelho próximo e solar. Portanto, superfícies com acabamento superficial fosco têm maiores valores de refletância solar por causa da maior proporção de pigmentos em relação à resina do que as de acabamento brilhante, que erroneamente são associadas com maior refletância por causa do alto brilho superficial.

Palavras-chave:
Materiais frios; Refletância solar; Emitância térmica; Composição química

Introduction

The main mechanism responsible for cooling cool paints, according to Lim (2020), is light scattering. Thus, spherical particles, such as pigments, which have a minimum size of 0.3 μm and preferably between 0.5 μm and 2 μm more effectively spread the wavelengths of solar radiation according to the Mie scattering theory proposed by the German physicist Gustav Mie (Yamasoe; Corrêa, 2016).

Levinson, Berdahl, and Akbari (2005a) determined the spectral characteristics of light absorption and backscattering as a function of wavelength in the 300 nm to 2500 nm solar spectrum for various pigments. Therefore, using the Kubelka-Munk coefficients of backscattering (S) and absorption (K), in particular, the determination of these properties in the near-infrared characterizes a pigment as cool or warm. The authors found that solar absorption (or reflectance) has a greater impact on the thermal balance of surfaces than spreading due to the light scattering phenomenon.

The relative proportion between the pigments, their concentration, distribution, chemical composition, and the difference between their refractive indices (η) with the medium (resin) in which they are involved also influence the near-infrared reflectivity according to Dwivedi et al. (2020). In this case, pigments of inorganic chemical composition with nanometric dimensions are used for cool materials, which spread radiation better because of their higher refractive indices and also have greater opacity power than organic dyes (Ladchumananandasivam, 2007).

However, the smaller the unit size of the particles, the greater their surface area and high energy of attraction between them, which promotes greater agglomeration and overlapping of the particles, as observed by Werle (2015) and Preuss (2016). The use of high-volume concentrations of pigments also causes their agglomeration according to Levinson, Berdahl and Akbari (2005b). The agglomeration state of the overlapping particles alters the exposure to solar radiation, thereby reducing the efficiency of the nanometer pigments for light scattering. The homogeneous dispersion of inorganic pigments in polymers is considered to be an extremely significant property since it promotes better performance to solar reflection, according to Qi, Xiang and Zhang (2017).

Among the most used pigments, titanium dioxide (TiO₂) is a stable, inert, non-toxic, inexpensive, and extremely popular white metal oxide. It has spectral characteristics of strong dispersion (S) and weak absorption (K), either in the visible region because of its coloration, but especially in the near infrared (Levinson; Berdahl; Akbari, 2005b).

The cool surfaces, due to the addition of reflective pigments to near-infrared radiation, obtain increases in solar reflectance compared to conventional ones, with the greatest reflectance gains in cool materials of darker colors (Dornelles; Caram; Sichieri, 2014; Zinzi, 2016), heating less than conventional surfaces. The more reflective near-infrared components of cool materials retain more heat, which hinders the radiative heat exchange process in the form of medium- and long-wave infrared radiation, that is, thermal radiation. Thus, these materials have lower thermal emission values, although high, concerning conventional materials according to Dornelles, Caram and Sichieri (2014) and Zinzi (2016).

The contribution of cool coatings in the building envelope to reduce energy demand for cooling purposes was observed in roofs that are not thermally insulated and that have a greater surface area of the roof compared to other surfaces (Synnefa; Santamouris, 2012) and for climatic conditions with high seasonal thermal variation (Pisello; Rossi; Cotana, 2014).

Hence, due to the diversity of climatic contexts in Brazil and the use of traditional construction techniques without thermal insulation, buildings would benefit from energy efficiency and provide better thermal comfort conditions with the use of cool coatings. However, the solar reflectance and thermal emittance data provided by the manufacturers of cool materials marketed in Brazil are not conclusive, because they do not present the equipment and methods used or the information is incomplete. In this sense, the study aims to evaluate the optical properties and identify the chemical components that interfere with them and that are related to the increases in solar reflectance and reductions in thermal emittance inherent in cool materials.

Material and method

For the evaluation of the solar reflectance and thermal emittance properties and the identification of the chemical composition on the surface of these cool materials, the methodology was divided into three stages: spectral reflectance measurement with spectrophotometer; thermal emittance measurement with portable emissometer and microanalysis of the compositional spectrum with the EDS accessory (energy dispersive x-ray detector) coupled to the field emission scanning electron microscopy (FE-SEM).

Complementary analyzes of color parameter L* were carried out because greater part of cool materials acquired have same color tone (white).

Selection of cool materials

A survey was carried out in the national market of civil construction coatings, which indicated, in their technical specifications, that they present high values of solar reflectance and thermal emittance. These materials have the designations "thermal" or "reflective" in their nomenclature, with mention of the reduction of surface temperature or decrease in energy consumption with air conditioning.

The white powder additive (A-01), selected reflective material, was incorporated according to its manufacturer's instructions into water-based acrylic paints with a semi-gloss finish of four different colors for application in roofing and facade. In the first four samples, two coats of the additive paints were applied and in the third and last coat, the conventional paint (without additives) of a corresponding color was applied, because according to the manufacturer of the additive powder, a smooth finish is obtained as a surface coating following these steps. However, the solar reflectance property changes due to the finish, therefore, in the last four samples of additive paints, three coats were applied only with the additive ones. Thus, due to the difference perceptible by the human eye in the surface finish, they are called rough finish.

The cool materials selected consist of three waterproofing liquid membranes (acronym WLM), twelve thermal or reflective paints (acronym P), and eight additive paints (acronym AP), after the addition of the powder, with four colors and two types of finish, which sum up to 23 samples of cool materials (Table 1). For comparison, conventional paints (acronym P-REF) were also evaluated, and all materials were painted on a cement board for steel frame (CB-REF) with a flat and waterproofed surface, which was cut into squares in the dimension of 10 cm on the side.

Table 1
Materials of cool coatings and reference evaluated

Solar reflectance

The measurements were performed with the Perkin Elmer spectrophotometer, Lambda 1050 model, with an integrating sphere of 150mm in diameter from the Federal University of Santa Catarina (UFSC), which provides the spectral curves of solar reflectance following the recommendations of the E903-20 standard (ASTM, 2020a). The absolute values measured in the equipment were adjusted to the standard solar spectrum of G173-03 (ASTM, 2020b) according to the procedure adopted in Dornelles (2008).

According to equipment technical specifications, measurements in the ultraviolet and the visible wavelength have higher resolution (≤ 0.05 nm) and precision (± 0.080 nm) than near infrared region with lower resolution (≤ 0.20 nm) and precision (± 0.300 nm) (Marinoski et al., 2013) and the uncertainty associated with the method is up to ± 0.02 absolute according to E903-20 (ASTM, 2020a).

Color parameters

Hue or chromaticity (intersection a* x b* axes) and lightness (L*) geometric coordinates were measured by the reflectance colorimeter, Delta Color brand, model Colorium 2 and evaluated on Lab color scale of CIE (International Commission on Illumination) chromatic system to describe visual differences among materials of same tone (white). Readings were taken at three different points on the painted cement boards by calculating measurements‘ arithmetic mean. Pearson's correlation with the scripts available in the free software RBio (Bhering, 2017) was performed between lightness (L*), with reflectance values in the visible region (ρ_VIS) and solar (ρ_solar) to evaluate dependence between these parameters.

Thermal emittance

Thermal emittance was measured by the portable emissometer, model AE1, from the Devices & Services belonging to the Laboratory of Energy Efficiency in Buildings (LabEEE - Laboratório de Eficiência Energética em Edificações), from the Federal University of Santa Catarina (UFSC) as recommended by C1371-15 (ASTM, 2015). Total hemispherical emittance (0.00 < ε_T< 1.00) corresponds to a linear value to the multimeter display voltage, with resolution in two decimal places, related to output voltages differences of the reference standards used in calibration. Equipment precision is ± 0.01 emittance units, i.e., greater than indicated by C1371-15 standard of ± 0.02 (ASTM, 2015).

Chemical composition and surface micrograph

The surface of the samples was visualized in high resolution using Field Emission Scanning Electron Microscopy (FE-SEM), its brand was JEOL, and its model JSM 7200F; it belonged to the Chemistry Institute of São Carlos of the University of São Paulo (IQSC/USP). The black and white micrograph with a magnification level of 250 times was selected in a more representative area of the surface of each sample.

However, to perform the proper fitting on the equipment, new samples of the analyzed materials were produced with a dimension of 8 mm on the side, using a 3 mm thick laser cut MDF board as substrate. In terms of comparison, samples were positioned next to a ruler with the smallest measuring less than 1 cm placed nearby the largest sample (10 cm) used for color, solar reflectance and thermal emittance tests (Figure 1). The six additive paints, all colored, were excluded from the chemical analysis, because the observation, especially of the chemical element titanium, from the pigment present in the white materials was preliminarily prioritized.

Figure 2(a) shows the nineteen samples, counterclockwise, from the sample of the unpainted MDF board, called a reference (REF-00), before the carbon coating process and Figure 2(b) after the procedure. This procedure, before the observation by FE-SEM, is performed by vacuum evaporation of the chemical element carbon on the surface of the samples to provide images with good resolution.

The BRUKER energy dispersive X-ray detector (EDS) model XFLASH 6-60, coupled to the microscope, was used to emit electron beams with applied power at 20 kV, and soon after its interaction with the surface of the samples, the signals emitted by them in the form of secondary electrons were collected. This microanalysis system was performed in three different sites of the samples and identifies the chemical elements present across a compositional spectrum and quantifies them in percentage (%) with a margin of error in BRUKER's EDX Esprit 2.3 software.

Figure 1
Size and relative proportion of samples used in the tests

Figure 2
Samples analyzed in the SEM/EDS counterclockwise from the MDF board

Statistical and visual analysis

Pearson correlation analysis was performed to assess the degree of dependence between variables and was performed with the scripts available in the free software RBio (Bhering, 2017). This analysis allowed to evaluate the influence of the main chemical elements distributed and present on the surface of all samples, especially titanium, on the reflectance properties in the near-infrared (ρ_NIR) and solar (ρ_solar) as for thermal emittance (ε).

The qualitative analysis of the surface of the samples allowed the visualization of morphological aspects such as the roughness or smoothness of the black-and-white images. Through the color images provided by the EDX Esprit 2.3 software, the distribution and location of the chemical elements, which were previously mapped by the EDS accessory on the surface were observed. In this sense, Adobe Photoshop software version 23.0.0 was used to superimpose two color images (each chemical element is represented by a different color) with a level of transparency applied at 100% in the first layer and 60% in the second to facilitate the observation of two chemical elements superimposed on the same area.

Results and discussion

Solar reflectance

The reflectance values were measured in the spectrophotometer and subsequently adjusted to the standard solar spectrum in the three sub-regions: ultraviolet (300 to 380 nm), visible (380 to 780 nm), and near-infrared (780 to 2.000 nm) and solar (300 to 2.500 nm) for the cool materials, the reference ones, and the cement board are presented in Table 2. The results are presented with three decimal places as indicated in the manuals of the classification programs of cool roof products (CRRC, 2022a) and cool facades (CRRC, 2022b) of the Cool Roof Rating Council.

All spectral curves of the liquid materials of manufacturers A and B have similar behaviors according to Figure 3 and Figure 4 in that order. The curves of white reflective paints P-01 and P-04 are very close to each other and above the others, due to this, their reflectance values are very close and higher than those of the other materials (Figure 3). The thermal paint P-03 has the worst reflective performance from the ultraviolet spectrum, in the visible (the only one with a value below 0.80 reflectance) and near-infrared, followed by the waterproof liquid membrane WLM-02 (Table 2). According to manufacturer A, the P-03 paint with the designation "Extreme" is an improved version of the P-01 paint, but the results did not show an increase in reflectance (Table 2).

The waterproof liquid membrane WLM-03 has higher reflectance only in the ultraviolet spectrum and lower values in the visible and near-infrared region than the reflective paint P-06, which then has the highest solar reflectance (Table 2). As the distance between the curves of materials of manufacturer B is greater (Figure 4), which shows that they have greater differences in the reflectance of the near-infrared (Δρ_NIR = 4.8%) and solar (Δρ_solar = 7.4%) than those of manufacturer A (Figure 3), whose curves are closer and overlapping, with less variation in the reflectance of the near-infrared (Δρ_NIR = 3.7%) and solar (Δρ_solar = 3%).

Table 2
Solar and spectral reflectance of cool and reference materials

Figure 3
Spectral reflectance curves of manufacturer A's materials

Figure 4
Spectral reflectance curves of manufacturer B's materials

Among the materials of manufacturer C, the reflective white (P-02-W) and beige (P-02-B) paints for application in coverage are the ones with the highest reflectances, in this right order, and they have similar spectral behavior and descending aspect since 600 nm in the visible region (Figure 5). The reflective red paint (P-02-R), despite having lower visible reflectance due to its dark hue, shows an increase in the near-infrared wavelength from 1.650 nm, approaching white and beige paints while the curve of gray paint (P-02-G) remains almost constant and always below 0.15 in the spectral reflectance value (Figure 5).

Among the membranes and white paints of different manufacturers, no similar patterns were found between the curves, especially in the near infrared (Figure 6). The waterproof liquid membrane WLM-01 has good reflection in the visible spectrum (ρ_VIS = 0.884), however, its curve decays significantly, with the worst reflective performance in the near-infrared region (ρ_NIR = 0.787), in this group of materials, accompanied by the thermal paint P-07 (ρ_NIR = 0.798), which has the lowest reflectance values in the visible (ρ_VIS = 0.852) and solar (ρ_solar = 0.810).

Thermal paint P-08 is the most reflective (except in ultraviolet) in the visible (ρ_VIS = 0.949) and the only one, among all samples analyzed, that has values above 90% reflectance in the infrared (ρ_NIR = 0.927) and solar (ρ_solar= 0.921) spectrum followed by reflective paints P-09 and P-05, in that order. The near-infrared reflectance values for reflective paints P-09 (ρ_NIR = 0.867) and P-05 (ρ_NIR = 0.856) are close. However, the curve of the P-09 remains above in the visible region and in the near-infrared up to the wavelength at 1.670 nm, which contributes to the better solar reflective performance (ρ_solar = 0.880) in relation to the P-05 paint (ρ_solar = 0.851) (Figure 6).

The difference between the reflective curves of the additive paints and the conventional reference paints of the same color is very small (Figure 7 to Figure 10).

Figure 5
Spectral reflectance curves of manufacturer C's materials

Figure 6
Spectral reflectance curves of white materials from different manufacturers

The highest amplitudes are observed for beige (AP-01-B-2) and red (AP-01-R-2) smooth finish additives paints in relation to their conventional paints as shown at Figures 7 and 8, respectively. Beige paint has the largest increase in the near-infrared (∆ρ_NIR = +5.1%) and solar (∆ρ_solar = +3.1%) spectrum reflection than the red one (∆ρ_NIR = +3.8%) (∆ρ_solar = +2.9%). However, the largest variations in the visible spectrum are for red (∆ρ_VIS = +2.4%) and beige (∆ρ_VIS = +1.6%) additive paints, respectively (Table 2).

Among the whites, the curves of the reference paints (P-01-W-REF) and the additive with a smooth finish (AP-01-W-2) are superimposed on the visible, therefore, they have the same reflectance value in this spectrum, up to about 975 nm in the near-infrared, where the additive slightly exceeds it (Figure 9). So, additive smooth finish paint (AP-01-W-2) has an increase in the reflectance of the near-infrared spectrum (∆ρ_NIR = +1.1%) and solar (∆ρ_solar = +0.6%), although small. On the contrary, additive paint with a rough finish (AP-01-W) remains with the curve below the other paints up to the wavelength of 1.650 nm in the near-infrared, so it has the worst reflective performance among them (Table 2). Despite classification of reference paint white (P-01-W-REF) as conventional, it has a high concentration of titanium dioxide (TiO₂) in its chemical composition according to the manufacturer.

The distance of the curves in the visible and near-infrared spectrum between the gray paints is even smaller than that observed in the white ones (Figure 10). For this, the graph is represented on a smaller scale (value up to 0.55 in reflectance), to observe in an enlarged way the differences between the paints, with the red curve, which refers to the cement board (CB-REF) that was used as a substrate for the paintings, correlated to Figures 9 and 10. The increases in reflectance values in the additive paint with a smooth finish (AP-01-G-2) in visible region (∆ρ_VIS = +0.3%), near-infrared (∆ρ_NIR = +1.0%) and solar (∆ρ_solar = +0.6%) is very small in relation to the reference paint (P-01-G-REF). Thus, there is little contribution of the reflective-powder additive to increase solar reflection in paints, with worse performance in gray and white shades.

Figure 7
Spectral reflectance curves of additive beige paints and their conventional reference paint

Figure 8
Spectral reflectance curves of additive red paints and their conventional reference paint

Figure 9
Spectral reflectance curves of additive white paints and their conventional reference paint

Figure 10
Spectral reflectance curves of additive gray paints and their conventional reference paint

Color parameter L*

To analyze differences in the reflectance of same tone materials, i.e., with the same visual sensation perceived by users, an analysis of color parameter L* was carried out. Concerning reflectance, all white materials in the visible region have a sharp increase from 380 nm to 420 nm, when reflective curves maintain linear up to wavelength 780 nm. Reflectance values in this region vary from values close to 0.80 to around 0.95,as represented in Figure 11 with a difference of up to 0.158 (∆ρ_VIS = 15.8%).

Lightness parameter (L*) in the CIE Lab system is represented on the vertical axis and indicates how much white there is in the sample color composition; therefore, its gradient varies from absolute black (value zero) to absolute white (value one hundred), whose midpoint (value fifty) is the color gray in accordance with Figure 12(a). Therefore, materials with light tones have a lot of white in their composition; therefore, all white materials evaluated have high luminosity values, that is, greater than ninety (from 90.44 to 96.80) on the CIE Lab color scale as shown in Figure 12(b).

Thermal paints P-08 and P-09 have the highest lightness values according to Figure 12(b) as well as visible and solar reflectance. The P-03 paint and WLM-03 liquid membrane are the white materials with the lowest visible and solar reflectance values (less than 0.80) and have some of the lowest lightness values coloras seen in Figure 12(b); therefore, the Pearson correlation was performed.

Figure 11
White materials represented only in the visible electromagnetic spectrum

Figure 12
Graphical representation of white materials lightness parameter L*

According to the graphs, the correlation (r) and determination coefficients (R²) present p values much lower than 0.05 (p < 0.001), which gives them high statistical significance. Furthermore, there was a positive correlation with correlation coefficients (r) equal to 0.9385 (p < 0.001) and 0.9087 (p < 0.001), respectively, which indicates a high adjustment between visible reflectance values (ρ_VIS) and solar (ρ_solar) with lightness in Figure 13(a) and 13(b). That is, as materials lightness value L* increases, there is a corresponding increase in the visible and solar reflectance values. For lightness values (L*) greater than 90, an increase of 1 absolute point on lightness scale material increases visible reflectance by 0.024 (2.4%) and solar reflectance by 0.0272 (2.72%), in this order, according to the equations (Figure 13).

Therefore, colors choice of same tone, such as white, for example, should not be carried out only visually by human eye because of chromatic differences that must be evaluated by quantitative parameters. Therefore, it is recommended to choose a color that has lightness parameter (L*) with high value because of the greater visible reflectance, and consequently higher solar reflectance.

However, for colors aesthetic option with darker tones, materials that have higher reflectance values in the near infrared spectrum (ρ_NIR) should be chosen with the addition of some rare organic and inorganic pigments such as metal sulfides and oxides and special fillers such as aluminum and mica flakes as indicated by Levinson et al. (2007). In this way, chemical components investigation, particularly pigments, of the evaluated materials was carried out to establish the relationship with near infrared reflectance (ρ_NIR), as presented in topic Statistical and visual analysis.

Thermal emittance

The values of the arithmetic mean of the thermal emittance measured at three points on the surface by the portable emissometer for the cool materials, the reference materials, and the cement board are shown in Table 3. The numerical value is in a dimensionless unit and represented with two decimal places as determined by the manuals of the Cool Roof Rating Council (CRRC, 2022a; 2022b) and the standard deviation (σ) that represents three measurements dispersion in relation to the average are represented with four decimal places.

All liquid materials were applied on the reference cement board (CB-REF) and, therefore, they present the same type of surface composition for the substrate. Among all materials, only liquid membranes WLM-01 and WLM-02 have thermal emittance values greater than 0.90 (ε = 0.91) and WLM-03 has a slightly lower value (ε = 0.88) and, therefore, present a difference of ± 0.03. The variation in thermal emittance is ±0.04 absolute (from 0.86 to 0.90) among the group of cool paints, the same variation found for the cool white coatings studied by Synnefa, Santamouris and Livada (2006), but these have slightly higher values of thermal emittance (between 0.89 and 0.93).

Figure 13
Correlation between lightness L* of white materials

The final coat was not a relevant aspect considering that all additive paints, despite the difference in finish, presented an emittance value greater than +0.01 than their respective reference paints. The greatest variation (Δ = ±0.02) occurred only among gray paints because of the gradual increase in emittance from 0.86 of the reference paint (P-01-G-REF) to 0.87 and 0.88 in relation to the smooth (AP-01-G) and rough (AP-01-G-2) finish samples, respectively.

Thus, the variation of the thermal emittance property between cool paints in relation to the corresponding conventional ones in colors is in line with that found in Synnefa, Santamouris, Apostolakis (2007), Revel et al. (2014), Zinzi (2016) and Silva (2017). However, changes in these values are considered derisory due to the use of reflective pigments in the infrared as well as within the uncertainty in the measurement related to the equipment (margin of error ± 0.01).

The greatest variation was found in Dornelles, Caram and Sichieri (2014) between the thermal emittance values of conventional paints (from 0.90 to 0.91) and cool paints (from 0.81 to 0.89), whose lowest values refer to paints imported from a Greek manufacturer, but despite this they are classified as good heat emitters in the form of thermal radiation. Therefore, in the literature, despite the high values of thermal emittance in cool paints, they tend to be lower than in conventional paints, which differs from the results found with the highest values, although subtle, observed in paints incorporated with the reflective additive powder.

Chemical composition and surface micrograph

Next, because of the condition inherent to the precepts of correlation analysis, only the five chemical elements detected by EDS are identified, which are present in all materials evaluated, in order of increasing atomic number, and their margin of error (Table 4). Therefore, the elements carbon (C), oxygen (O), silicon (Si), calcium (Ca) and titanium (Ti) together total approximately 90% of the chemical constitution of these samples. The values and margin of error correspond to the three points arithmetic mean collected by the EDS beam in each sample.

The substrate used for the application of the liquid materials is MDF (REF-00), a board of wood fibers with medium density; therefore, composed only of organic material, carbon and oxygen (Table 4). From the microscopic images, the material presents a web of fibers in several non-oriented directions composed of oxygen and some well-defined points of carbon (Figure 14).

Therefore, the penetration depth of the X-ray beam by the EDS fitting is about 2 μm, i.e., less than the total thickness of the materials painted on the MDF board, after drying. Then, the quantification of the organic elements carbon (C) and oxygen (O) in the paints refer only to their chemical composition, mostly to resin.

In addition, oxygen is also related to metal oxides, since they are binary compounds consisting of two chemical elements, oxygen (O) bound to another element, usually inorganic. As an example, titanium (Ti) and silicon (Si) form the dioxides of titanium (TiO₂) and silicon (SiO₂), or just silica, which are pigments added to paints. The detection of calcium (Ca) is related to calcium carbonate (CaCO₃) or calcite, a fine white powder used as a pigment or its by-product, calcium oxide (CaO), known as quicklime, which is also used in painting processes.

Table 3
Thermal emittance of cool and reference materials

Table 4
Percentage (%) of the five chemical elements - and their margin of error - present on the surface of all samples analyzed by SEM/EDS

Chemical element (atomic number) and margin of error in percentage (%)
Acronym	C (06)	Error (±)	O (08)	Error (±)	Si (14)	Error (±)	Ca (20)	Error (±)	Ti (22)	Error (±)
REF-00	60.46	8.67	39.54	6.15	-	-	-	-	-	-
WLM-01	37.02	5.95	36.22	6.15	1.00	0.15	16.69	1.52	6.75	0.70
WLM-02	60.66	10.08	25.62	4.53	2.33	0.31	3.24	0.33	4.82	0.51
WLM-03	62.67	10.56	23.69	4.71	0.13	0.04	5.31	0.55	5.59	0.61
P-01	53.70	9.43	24.91	4.45	6.88	0.96	0.08	0.04	5.52	0.62
P-02-W	59.29	9.88	22.25	4.12	0.47	0.08	0.15	0.04	17.05	1.71
P-02-B	57.05	9.33	22.98	4.11	1.02	0.15	0.16	0.04	16.69	1.61
P-02-G	73.17	10.45	17.54	2.87	0.26	0.04	0.05	0.02	8.47	0.80
P-02-R	81.82	11.47	15.00	2.47	0.15	0.03	0.12	0.03	1.18	0.14
P-03	47.31	8.25	31.94	5.60	6.42	0.90	1.01	0.13	4.70	0.52
P-04	59.20	10.58	22.26	4.18	4.06	0.60	3.35	0.38	5.29	0.59
P-05	42.81	7.20	36.14	5.95	15.93	2.04	0.35	0.07	3.80	0.41
P-06	46.29	8.18	28.49	5.23	3.82	0.55	3.11	0.34	13.63	1.41
P-07	44.04	7.62	29.94	5.22	5.11	0.71	0.22	0.06	10.48	1.05
P-08	26.09	4.11	38.86	6.32	2.27	0.30	21.00	1.86	8.70	0.82
P-09	39.60	6.17	31.97	5.37	0.45	0.08	10.46	0.95	16.04	1.46
AP-01-W	42.94	7.18	29.20	5.14	2.78	0.38	2.17	0.23	19.89	1.91
P-01-W-REF	44.53	7.04	28.39	4.74	2.29	0.30	2.12	0.21	19.72	1.76
AP-01-W-2	40.99	6.78	29.34	5.07	3.79	0.50	2.20	0.23	20.05	1.89

• Caption:
• WLM (waterproofing liquid membrane)
• P-REF (conventional paint)
• P (thermal or reflective paint)
• AP (additive paint w/ smooth finish)
• AP-2 (additive paint w/ rough finish)
• P-REF (conventional paint)
• CB-REF (cement board non painted)
• Color acronym W (white); B (beige); G (gray); R (red)

Figure 14
Microscopic images (250x magnification) of the MDF base used as a substrate for the application of the liquid materials provided

Statistical and visual analysis

The relationship between the properties of solar reflectance and thermal emittance with the concentration of the chemical elements identified on the surface of the samples was performed for all materials evaluated in the electron microscopy technique, except for the MDF reference (REF-00). Thus, the values of near-infrared (ρ_NIR) and solar (ρ_solar) reflectance and thermal emittance (ε) were correlated with the concentration, in percentage (%), of the five chemical elements present in all samples and which predominantly constitute the chemical composition of these materials (Table 5).

There was no statistically significant relationship between the concentration of the chemical elements titanium (Ti) and silicon (Si), which are present in the metal oxides that constitute the cool pigments, and reflectance, although Levinson, Berdahl and Akbari (2005a) found for titanium dioxide (TiO₂) spectral characteristics of weak absorption (K) and strong backscatter (S) of near-infrared radiation.

Instead of this, near-infrared (ρ_NIR) and solar (ρ_solar) reflectance showed strong negative and positive correlation, respectively, for the chemical elements carbon (C) and oxygen (O), and significant at 0.1% probability. Therefore, there is an increase in reflectance in the near-infrared and solar spectrum as the proportion of carbon (C) decreases and that of oxygen (O) increases. In view of this, in the white paint P-08, which has the highest reflectance in the near-infrared (ρ_NIR = 0.927) and solar (ρ_solar = 0.921), the lowest amount of carbon (26.09%) and the highest amount of oxygen (38.86%) was detected by the EDS accessory.

Inversely related, the colored paints for application in tiles in gray (P-02-G) and red (P-02-R) are those with the highest amount of carbon (C) and lowest in oxygen (O). Therefore, the gray and red paints have the lowest reflectances in the near-infrared spectrum (ρ_NIR = 0.113) (ρ_NIR = 0.476) and solar (ρ_solar = 0.110) (ρ_solar = 0.360), respectively, but are also influenced by the low values in the visible spectrum (ρ_VIS = 0.108) (ρ_VIS = 0.273), in this order, considering that they are samples with low lightness in the color parameter. To visually exemplify the differences in the ratio between carbon (C) and oxygen (O), the micrographs of white paint (P-08) with predominantly rough surfaces and gray (P-02-G) and red (P-02-R) paints are presented, which are the smoothest observed by the field scanning electron microscope among all samples (Figure 15).

Then, carbon (C), an organic element, predominantly composes the resin with oxygen (O), also organic, which is incorporated into the chemical formula of inorganic metal oxides such as calcium oxide (CaO) and titanium (TiO₂) and silicon (SiO₂) dioxides. Therefore, it forms chemical bonds with the elements calcium (Ca), titanium (Ti), and silicon (Si) which were also detected. Thus, although there was no statistically significant relationship between the inorganic chemical elements, titanium (Ti) and silicon (Si), with solar reflectance, there is an association between the greater amount of oxygen (O), which is also present in the pigments, and the increase in reflectance.

Therefore, the highest reflectance values are related to the lower proportion of resin in relation to the amount of pigments, referring to the surface finish of the paints, that is, the matte finish paint in Figure 15(a), whose micrograph is the roughest, reflects more than the paints with high surface gloss in Figure 15(b) and Figure 15(c) of smoother images.

Table 5
Correlation matrix between reflectance and emittance properties with the chemical elements present on the surface of the materials

Figure 15
Microscopic images at 250x magnification for paint samples

This quantitative relationship between the elements carbon (C) and oxygen (O) present on the surface of the paints can clarify the observations made in Dornelles (2008). In this study, acrylic paints with a semi-gloss finish had higher solar absorptance values (i.e., lower solar reflectances) than paints (of the same color and from the same manufacturer) with a matte finish, with the highest reflectance differences observed in the near-infrared spectrum.

Due to this, the paints of the same color (white) from the same manufacturer (manufacturer C), the high gloss finish (P-02-W) for coverage application and the reference (P-01-W-REF) with semi-gloss finish, were also compared. Semi-gloss white (P-01-W-REF) has higher reflectance in the near-infrared (ρ_NIR = 0.794) and solar (ρ_solar = 0.835) and has a lower amount of carbon and a higher amount of oxygen in relation to high gloss finish paint (P-02-W) (ρ_NIR = 0.758) (ρ_solar = 0.805), which is consistent with the statistical results. Thus, as verified by Dornelles (2008), the paints analyzed with high gloss finish do not reflect more solar radiation than semi-gloss and probably do not reflect more than matte finish.

Thermal emittance (ε) showed a statistically significant correlation (p < 0.001), with a negative and moderate coefficient (-0.605) with the chemical element titanium (Ti), a transition metal. This result indicates that the lower the titanium content on the surface of the material, the thermal emittance increases and vice versa. The thermal emittance values, in descending order, are presented associated with the percentage of titanium (%) detected in the evaluated materials (Table 6).

The P-05 paint has the lowest percentage of titanium (3.80%) on its surface, among white materials, and has one of the highest thermal emittance values (ε = 0.90) and the liquid membranes WLM-01 and WLM-02 are the samples with the highest thermal emittance values (ε = 0.91) obtained among all analyzed. Although it does not verify on its surfaces the lowest percentages of titanium, however, its quantities are still low for the membranes WLM-01 (6.75%) and WLM-02 (4.82%). In this sense, all samples with a percentage greater than 17% in titanium (Ti) on the surface have the lowest thermal emittance values (ε = 0.85 or ε = 0.86).

Table 6
Thermal emittance values and the percentage amount of the chemical element titanium detected for the materials

However, this observation cannot be extended to paints for coverage application from manufacturer D (acronym P-02). Reflective white paint (P-02-W) has the highest amount of titanium (17.05%) and a slightly lower thermal emittance value (ε = 0.86) compared to other colored paints of the same line and manufacturer. Despite the gradual and considerable decrease in the percentage of titanium (Ti) in values from 16.09% to 8.47% to 1.18% present on the surface of the paints in beige (P-02-B), gray (P-02-G) and red (P-02-R) colors, respectively, the thermal emittance value is the same (ε = 0.87), that is, it does not change for all these samples.

Therefore, the relationship between thermal emittance (ε) and titanium (Ti) concentration, although statistically significant by the t-test, does not corroborate the same value of thermal property (ε) with the significant reduction in the amount of the metallic chemical element, since the degree of correlation is only moderate.

Other examples that do not validate the correlation are P-09 paint that has a high amount of titanium (16.04%) and one of the highest thermal emittance values (ε = 0.89), on the other hand, P-08 paint has a moderate amount of the chemical element (8.70%) and one of the lowest values measured for the property (ε = 0.87).

Then, the thermal emittance (ε) is not influenced solely by the concentration of the metallic chemical element present in the pigment titanium dioxide (TiO₂). Therefore, other characteristics such as roughness, oxidation, temperature, among others, as stated by Perin (2009) should be analyzed in future studies to understand which factor has the greatest influence on the radiative behavior of materials.

In addition to the quantitative evaluation, the visual qualitative analysis was performed from observations of the elementary maps provided by the EDX Esprit 2.3. software from the BRUKER manufacturer of the EDS accessory. Thus, the location and distribution of the chemical elements on the surface of the samples were analyzed.

The inorganic chemical element titanium (Ti) is present in the chemical composition of the materials in the form of titanium dioxide (TiO₂). Sodium polyacrylate [(C₃H₄O₂)x.xNa] is specified as the dispersant in manufacturer's liquid materials datasheets A and reflective paints P-05 and P-07. The chemical element sodium (Na) is present in all samples, except in the liquid membrane WLM-03 and in the paint P-08, in the proportion between 0.10% and 0.55%.

Its function, as well as solvents, is to help the segregation of agglomerated particles, but its excessive use can change the morphological characteristics of the pigments. There are presented the black and white micrographs and color images of the individual chemical element titanium (Ti), in yellow, and superimposed on sodium (Na) in the manufacturer's materials A (Figure 16) and in the paints P-05 and P-07 (Figure 17) in which this chemical element is presented as the dispersant.

In the micrographs of the element titanium (Ti) (3^rd column), it is homogeneously distributed over the entire surface of the samples, with the exception of the paint P-05 in Figure 17(a), where it is concentrated in some well-defined points by the micrograph, nor is the element sodium (Na) evenly distributed in this material.

In the regions where they are marked with white circles, the largest and brightest yellow spots in the distribution of the element titanium (Ti) were identified (Figure 16). In the micrographs of the elements titanium and sodium (2^nd column), on these circles, there is no identification of the element sodium (Na). Therefore, one explanation is that there was no segregation of the titanium (Ti) particles due to the absence of the dispersant in these regions. In this way, these spots are likely identified as agglomerated or aggregated particles of the titanium dioxide pigment (TiO₂).

The titanium (Ti) points, where the presence of sodium is not detected, in paints P-05 and P-07 were also delimited (Figure 17). Therefore, the highlighted points of titanium (Ti) as the brightest may represent the agglomerated or aggregated particles of the pigment, where segregation of the particles has not occurred. However, there were also many large and bright titanium spots on P-05 paint that are superimposed on sodium (Na), which differs from the images of the other materials analyzed.

In the study by Pereira (2010), the influence of different dispersants on suspensions of titanium dioxide nanoparticles (TiO₂) with the same concentration (100 μg/ml) was analyzed. The separation of the solid (pigment) suspended in the liquid (dispersant) caused by the sedimentation process was verified. The best result obtained was with oxalic acid [(COOH)₂.2H₂O] which kept the particles in suspension, not agglomerated, for longer. However, scanning electron microscopy (SEM) images visually show that even in the presence of the best surfactant, the individual pigment particles do not present themselves, but in the form of small agglomerates and no significant change in the individual dimension of the nanoparticles was observed.

In addition, the materials were evaluated as final products intended for commercialization. Therefore, there is no information regarding the production process, or the characteristics of the raw materials used, such as the size of the pigment and its crystalline structure for a more detailed basis. Thus, qualitative visual analysis is a finding that refers to the repeated observation of the distribution of chemical elements on the surface of the evaluated samples.

In Preuss (2016), the pigments were compacted at large concentrations and optical depths for the diffuse reflectance test so that multiple scattering of radiation between the various nanoparticles occurs. Therefore, the reflectance results do not refer to just an individual particle, but a set of them. Thus, titanium dioxide (TiO₂) pigments sold commercially as those treated in the laboratory (with a higher degree of purity) have reflectance values above 70% in the visible region and the near infrared (up to the wavelength at 1.700 nm) in almost 100% reflectance due to the multiple scattering between the particles.

Thus, despite the high values of solar reflectance related to the pigment titanium dioxide (TiO₂) in its powder form, no correlation was found in this study between the concentration of this chemical element (Ti) detected by the EDS accessory on the surface of cool materials with reflectance. Despite this, the association of the highest amount of oxygen (O), which is also present in the chemical formula of metal oxides (TiO₂), with the highest reflectance values demonstrates that the reflectance gains are related to the presence of the pigment, although indirectly. Therefore, a hypothesis that evidences the indirect relationship with reflectance is the apparent overlap of titanium (Ti) particles, because there was no adequate distribution of the pigment particles due to the absence of dispersant on the surface, where the yellow titanium spots were circulated.

Conclusions

Cool materials are mostly white, and because of this, they have high solar reflectance as a result of higher reflectances in the visible spectrum. Despite the significant difference in the visible (15.8%), the greatest variation in reflectance is verified in the near-infrared region (24.1%). The paints for application in tile in gray and red colors, darker, reflect more in the near-infrared than in the visible region, however it does not help so much in improving solar reflectance. After the addition of the reflective white powder, the greatest solar reflectance gains were, in this order, in beige and red paints, the latter being the darkest of all, but it does not extend to gray additive paint, followed by white, with the lowest reflectance gains.

All materials have a high thermal emittance value, so they are good heat emitters in the form of thermal radiation despite having been detected by EDS concentration of metallic chemical elements such as titanium (Ti) and silicon (Si).

Figure 16
Microscopic images with 250x magnification of manufacturer materials A

Figure 17
Microscopic images with 250x magnification of paints P-05 and P-07

By statistical analysis, paints with a lower amount of carbon (C), that is, lower content of organic resin and higher amount of oxygen (O), an element also incorporated into metal oxides, are more reflective. Thus, the detection of the chemical element titanium (Ti), present in the pigments, is related to higher reflectance values indirectly, considering that no statistically significant relationship was found. According to the observation of the images, bright spots of the element titanium (Ti) were identified on the surfaces of the materials, after mapping the elements by EDS, in regions where the chemical element sodium (Na) of the dispersant was not located, which evidences the overlap of pigment particles. Therefore, it causes a decrease in the exposure of pigments to solar radiation and consequently the performance of light scattering and reflection.

Despite the identification of a metallic chemical element on the surfaces, the greater amount of titanium (Ti) is not necessarily related to the lower thermal emittance values since the degree of correlation is moderate for this statistically significant relationship. The uncertainty associated with thermopile differential method and portable emissometer may be related to the correlation admitted as moderate, considering that all assessed materials have high thermal emittance and near values, with a maximum absolute variation of ± 0.06, with the equipment accuracy of ± 0.01. Therefore, in this case, the concentration of metallic chemical elements was not a determining aspect for thermal emittance property of the evaluated elastomeric materials by the portable emissometer.

Acknowledgements

The present experimental research was funded by the São Paulo State Research Support Foundation (FAPESP) that granted a master's degree scholarship (Process n° 2019/20050-9). The opinions expressed in this material are those of the authors and do not necessarily reflect FAPESP views. Special thanks to Professor Deivis Marinoski, from the Federal University of Santa Catarina (UFSC) for his contribution to solar reflectance and thermal emittance measurements.

References

• AMERICAN SOCIETY FOR TESTING AND MATERIALS. C1371-15:standard test method for determination of emittance of materials near room temperature using portable emissometers. West Conshohocken, 2015.
• AMERICAN SOCIETY FOR TESTING AND MATERIALS. E903-20:standard test method for solar absorptance, reflectance and transmittance of materials using integrating spheres. West Conshohocken, 2020a.
• AMERICAN SOCIETY FOR TESTING AND MATERIALS. G173-03:standard tables for reference solar spectral irradiances: direct normal and hemispherical on 37o tilted surface. West Conshohocken, 2020b.
• BHERING, L.L. RBio: a tool for biometric and statistical analysis using The R platform. Crop Breeding and Applied Biotechnology, Viçosa, v. 17, p. 187-190, jun. 2017.
• COOL ROOF RATING COUNCIL. CRRC-1 Roof product rating program manual. Portland: CRRC, 2022a.
• COOL ROOF RATING COUNCIL. CRRC-2 Wall product rating program manual. Portland: CRRC , 2022b.
• DORNELLES, K. A. Absortância solar de superfícies opacas: métodos de determinação e base de dados para tintas látex acrílica e PVA. Campinas, 2008. 160 f. Tese (Doutorado em Engenharia Civil) - Faculdade de Engenharia Civil, Arquitetura e Urbanismo, Universidade Estadual de Campinas, Campinas, 2008.
• DORNELLES, K. A.; CARAM, R. M.; SICHIERI, E. P. Absortância solar e desempenho térmico de tintas frias para uso no envelope construtivo. Paranoá: Cadernos de Arquitetura e Urbanismo, Brasília, v. 12, p. 55-64, 2014.
• DWIVEDI, C. et al Infrared radiation and materials interaction: active, passive, transparent and opaque coatings. In: DALAPATI, G. K.; SHARMA, M. (ed.). Energy Saving Coating Materials. Amsterdan: Elsevier, 2020.
• LADCHUMANANANDASIVAM, R. A natureza da luz e a sua interação com a matéria. In: LADCHUMANANANDASIVAM, R. Processos químicos têxteis: Ciência da Cor. 2. ed. Natal: Universidade Federal do Rio Grande do Norte, 2007.
• LEVINSON, R. et alMethods of creating solar-reflective nonwhite surfaces and their application to residential roofing materials. Solar Energy Materials and Solar Cells, Amsterdan, v. 91, n. 4, p. 304-314, 2007.
• LEVINSON, R.; BERDAHL, P.; AKBARI, H. Solar spectral optical properties of pigments: part I: model for deriving scattering and absorption coefficients from transmittance and reflectance measurements. Solar Energy Materials and Solar Cells , Amsterdan, v. 89, n. 4, p. 319-349, 2005a.
• LEVINSON, R.; BERDAHL, P.; AKBARI, H. Solar spectral optical properties of pigments - part II: survey of common colorants. Solar Energy Materials and Solar Cells , Amsterdan, v. 89, n. 4, p. 351-389, 2005b.
• LIM, Y.-F. Novel materials and concepts for regulating infra-red radiation: radiative cooling and cool paint. In: DALAPATI, G. K.; SHARMA, M. (ed.). Energy Saving Coating Materials . Amsterdan: Elsevier , 2020.
• MARINOSKI, D. L.et alAnálise comparativa de valores de refletância solar de superfícies opacas utilizando diferentes equipamentos de medição em laboratório. In: ENCONTRO NACIONAL DE CONFORTO E AMBIENTE CONSTRUÍDO, 12.; ENCONTRO LATINO-AMERICANO DE CONFORTO E AMBIENTE CONSTRUÍDO, 8., Brasília, 2013. Anais [...] Brasília: ANTAC, 2013.
• PEREIRA, J. C. Estudo do comportamento de nanopartículas de dióxido de titânio em diferentes suspensões. Lisboa, 2010. 112 f. Dissertação (Mestrado em Ciência dos Materiais) - Faculdade de Ciências e Tecnologia, Universidade Nova de Lisboa, Lisboa, 2010.
• PERIN, A. L. Desenvolvimento de um equipamento para medição de emissividade. Porto Alegre, 2009. 115 f. Dissertação (Mestrado em Engenharia) - Escola de Engenharia, Universidade Federal do Rio Grande do Sul, Porto Alegre, 2009.
• PISELLO, A. L.; ROSSI, F.; COTANA, F. Summer and winter effect of innovative cool roof tiles on the dynamic thermal behavior of buildings. Energies, Basiléia, v. 7, n. 4, p. 2343-2361, 2014.
• PREUSS, N. L. Efeito dos aspectos morfológicos do pigmento TiO₂ nas propriedades ópticas de tintas base água. Porto Alegre, 2016. 100 f. Dissertação (Mestrado em Engenharia) - Escola de Engenharia, Universidade Federal do Rio Grande do Sul, Porto Alegre, 2016.
• QI, Y.; XIANG, B.; ZHANG, J. Effect of titanium dioxide (TiO₂) with different crystal forms and surface modifications on cooling property and surface wettability of cool roofing materials. Solar Energy Materials and Solar Cells , Amsterdan, v. 172, n. 2016, p. 34-43, 2017.
• REVEL, G. M. et alCool products for building envelope - part I: development and lab scale testing. Solar Energy, Amsterdan, v. 105, p. 770-779, 2014.
• SILVA, I. L. M. Estudo de durabilidade de pinturas “frias” e convencionais expostas ao envelhecimento natural. São Paulo, 2017. 170 f. Dissertação (Mestrado em Ciências) - Escola Politécnica, Universidade de São Paulo, São Paulo, 2017.
• SYNNEFA, A.; SANTAMOURIS, M. White or light colored cool roofing materials. In: KOLOKOTSA, D.; SANTAMOURIS, M.; AKBARI, H. (org.). Advances in the development of cool materials for the built environment. Sharjah: Bentham Science Publishers, 2012.
• SYNNEFA, A.; SANTAMOURIS, M.; APOSTOLAKIS, K. On the development, optical properties and thermal performance of cool colored coatings for the urban environment. Solar Energy , Amsterdan, v. 81, n. 4, p. 488 497, 2007.
• SYNNEFA, A.; SANTAMOURIS, M.; LIVADA, I. A study of the thermal performance of reflective coatings for the urban environment. Solar Energy , Amsterdan, v. 80, n. 8, p. 968-981, 2006.
• WERLE, A. P. Vida útil de revestimento frio e autolimpante. São Paulo, 2015. 272 f. Tese (Doutorado em Ciências) - Escola Politécnica, Universidade de São Paulo, São Paulo, 2015.
• YAMASOE, M. A.; CORRÊA, M. P. Processos radiativos na atmosfera: fundamentos. São Paulo: Oficina de Textos, 2016.
• ZINZI, M. Characterisation and assessment of near infrared reflective paintings for building facade applications. Energy and Buildings, Amsterdan, v. 114, p. 206-213, 2016.

Edited by

Editor:

Enedir Ghisi

Publication Dates

• Publication in this collection
  07 Oct 2024
• Date of issue
  2024

History

• Received
  31 Jan 2024
• Accepted
  23 Mar 2024