*2.3. Methods*

2.3.1. Scanning Electron Microscopy (SEM)

Using a GEMINI (FESEM) CARL ZEISS scanning electron microscope (SEM) (LEICA, Madrid, Spain), with a Röntec M Series EDX Detector (LEICA, Madrid, Spain), belonging to the Scientific Instrumentation Centre of the University of Granada (CIC) the mineralogical, microstructural and textural characterization of the mixes was performed. The results of the EDX analysis were collected by matrix spotting of the indicated samples.

#### 2.3.2. Water Vapor Permeability

According to [67], sixty-four test pieces were analyzed for each paste type (two for each pigment) and two for each mix without added pigments (eight test pieces), after 120 days, in laboratory conditions (temperature: 20 ± 2 ◦C; RH: 65 ± 5%). Their dimensions were 40 × 40 × 20 mm. The edges of the samples were sealed using liquid paraffin; then they were placed in plastic recipients with covers, such that one part of the test piece was inside the recipient and the other was outside of it. The join between the test piece and the plastic recipient was sealed using liquid paraffin.

Granular calcium chloride was placed inside the plastic recipients as a drying agent, in sufficient volume in order to obtain a relative humidity of 0%. An air gap measuring approximately 10 mm was left between the drying agent and the base of each test piece.

After preparing the samples, the recipients were weighed in order to determine their initial mass.

The specimens were brought out from the climatic chamber to measure the weights. The same specimens were used to perform the water vapor permeability tests at both 28 and 120 days.

The test conditions were 50 ± 1% RH and 25 ± 0.5 ◦C, the samples being weighed every twenty-four hours until the weight difference every twenty-four hours did not exceed 5%.

#### 2.3.3. Mechanical Tests

The flexural strength and compressive strength tests were carried out on each type of test piece, following a hardening time of 120 days. The total of samples used were 216. In order to calculate the strength, regulation [68] was used, for prismatic test pieces with dimensions of 160 × 40 × 40 mm.

The press machine used to perform the break test was IBERTESTEUR TEST MD2 universal testing apparatus (Ibertest, Madrid, Spain). For a sampling interval of 64 mm, the test speed was 1 mm/min. The sample was broken by using a concentrated load on the central part, with the load cell set to 5 kN. In the compression testing, the speed used was 5 mm/min.

Using the average of the results for the three test pieces for each dosage and pigment used, the average mechanical strength value was obtained.

#### 2.3.4. Color Tests

After preparing the samples, their diffuse spectral reflectance curves were measured. Five colorimetric determinations were performed for each one. Using Bessel's correction, the standard deviation was obtained for the values acquired, without this ever exceeding 3% of the associated average value [69].

Finally, SpectraMagic NX Color Data Software (I.T.A. Aquateknica, S.A., Valencia, Spain) was used to present the simulations of the color variations for the samples studied.

#### **3. Results**

### *3.1. Scanning Electron Microscopy (SEM)*

Figure 3 shows both the morphological analysis and EDX analysis of the samples.

**Figure 3.** Scanning electron microscopy images corresponding to different pastes with the pigment, GE (Green Earth). The presence of gypsum lumps (**A**), several recrystallizations that are filling cracks (**B**,**D**) and other characteristic mineralization of the paste's matrix (**B**–**D**) are highlighted. The plot of the EDX analysis corresponds to the elemental composition of the matrix.

In the SEM study carried out 120 days after they were prepared, plaster crystallization was detected in specific areas of all of the samples, specifically in cavities with sizes of ≈20–35 μm, along with needle morphology in the case of the PPS + GE sample, where lumps of such crystallization were detected, and in PALS + GE and PHLS + GE samples. It may be speculated that this type of neoformed crystal generates a certain mechanical improvement in the pastes due to the filling and consolidation produced in cracks and micro-fractures.

Meanwhile, calcium carbonate crystallization was found in cavities in some samples, such as those of PALS + GE; belite crystallization (Ca2Si, calcium silicate hydrate) was found in the PHLS + GE samples; and neoformed sodium hydroxide (NaOH) in the plaster and sodium metasilicate matrix of the PWGS + GE samples.

The EDX analyses of the matrixes of all of the samples provided results which are consistent with the quality of the samples. Elements associated with the Green Earth (GE) pigment stood out in all of them. This artificial pigment is obtained by combining two types of hydrous phyllosilicate: celadonite (KMgFe3 + Si4O10(OH)2) and glauconite (Fe3 +, Al, Mg2Si, Al4O10(OH)2). Apart from this, the elements associated with the matrix are those which are characteristic of each mix. S-Ca associations stand out, as they are typical in plaster and air lime (PPS and PALS), along with increases in Si, which is characteristic in samples containing hydraulic lime and water glass.

#### *3.2. Water Vapor Permeability Test*

The results of the water vapor permeability tests for both test periods, after 28 and 120 days, are shown in Table 5 and Figure 4.

**Table 5.** Results of the water vapor permeability test and standard deviation for the different white pastes tested.


**Figure 4.** Water vapor permeability of the binders and their corresponding pigmented pastes (PPS, PALS, PHLS and PWGS). The green circles and the orange circles correspond to the permeability at 28 and 120 days respectively. The dotted line and the green letter B correspond to the specific binder at 28 days. The dotted line and the orange letter B correspond to the specific binder at 120 days.

After 28 days, the average permeability for the PPS + PIGMENT samples was 33kg/(m·Pa·s))·10<sup>−</sup>2. The maximum value was 33.3 ± 1.12kg/(m·Pa·s))·10−<sup>12</sup> for the sample containing the ZY pigment, and the minimum value was 32.3 ± 1.3kg/(m·Pa·s))·10−<sup>12</sup> for the sample containing the GE pigment.

In the case of the PALS + pigment samples, the average representative value for the group is 26.83kg/(m·Pa·s))·10<sup>−</sup>12. The maximum value was 27.4 ± 1.33kg/(m·Pa·s))·10−<sup>12</sup> for the sample containing the CG pigment, and the minimum value was 26.1 ± 1.20kg/(m·Pa·s))·10−<sup>12</sup> for the sample containing the GE pigment.

The average value of the PWGS + PIGMENT sample group was 24.8kg/(m·Pa·s)) ·10<sup>−</sup>12. There was a maximum value of 25.72 ± 1.8kg/(m·Pa·s)) ·10−<sup>12</sup> for the NS samples and a minimum value of 24.1 ± 1.11kg/(m·Pa·s)) ·10−<sup>12</sup> for the GE samples.

Finally, the lowest values were noticed in the plaster and hydraulic lime pastes (PHLS + PIGMENT), the average value being 20.8kg/(m·Pa·s)) ·10−12. The maximum value was 21.5 ± 0.94kg/(m·Pa·s)) ·10−<sup>12</sup> for the CG samples, and the minimum value was 20.0 ± 1.50kg/(m·Pa·s)) ·10−<sup>12</sup> for the GE samples.

It is deduced from the overall analysis that the substitution of the plaster (−15%) in the different pastes by air lime, sodium silicate and hydraulic lime resulted in lower permeability of the compound. In the different mixes, over time (120 days), it was observed that the values decreased in all cases, confirming compaction of the samples. We recorded reductions in the values of between 8.50% for the PPS + O samples and 0.60% for the PPS + NS samples, for the 28 and 120 days. For the remaining pastes, the values varied between 4% and 5% for each period of study.
