**2. Materials and Methods**

The RTPF were obtained from a tire recycling company located at Rionegro, Colombia. Steel, rubber, and nylon wastes were mechanically separated. The nylon fiber has an average diameter of 22.4 µm. The residual rubber has an average size of 46 µm and a concentration of 60 wt% with respect to the rubber weight content. Molasses were obtained from a sugar cane company, located at Colombia. Table 1 summarizes the chemical composition of molasses. Kaolin soil in natural condition was supplied by Sumicol S.A.

Figure 1a shows polymer fibers on macro-scale image, where fibers look like cotton pads with fine residual rubber particles from the recycling process. Figure 1b shows a higher magnification of these fibers, revealing they are connected to one another due to the recycling process, which destroys the polymer textile and forms polymer fibers. Figure 1c shows the consistency of molasses. Figure 1d shows the supplied raw soil.


**Table 1.** SEM-EDS chemical composition for molasses.

**Figure 1.** Materials used in this research, (**a**) polymer fibers, (**b**) magnification on polymer fibers, (**c**) molasses, (**d**) soil used in this research.

Figure 2 shows the optimum moisture (water) content showing the optimum moisture content, where soil contains 32% of water and compacted according to ASTM D1557-12e1 achieving 1.42 MPa as its highest dry density.

**Figure 2.** Optimum moisture content.

The dry kaolin clay was crushed using a hand roller in order to decrease agglomeration and have better particle size distribution. Molasses, RTPF, and water were measured by dry soil weight. Molasses and the fiber were hand mixed for three minutes until a homogenous consistency was reached; then the mixture and the water were mechanically mixed using a Hobart N50 apparatus at a speed of 126 rpm for five minutes. For the compaction of the soil, the ASTM D1557-12e1 Method A was used [39]. Three samples were selected from each compacted formulation, obtained by hammering a 2" diameter aluminum tube, and a total of 21 soil samples. Using a hydraulic jack, soil samples were removed from the tubes. These samples were remolded and cut to 4" height, and then, covered with plastic membrane in order to maintain the room temperature of 20 ± 3 ◦C during an air-cured process for 28 days.

The sample fabrication and all tests were conducted at 20 ◦C, typical room temperature in Medellin. Viscosity of molasses decreases with temperature as shown below [40], from about 8 to 12 Pa.s (8000 to 12,000 cPs). Since molasses change with temperature, the current solution is for areas where the temperature variation is low, such as areas where people do not need AC or heater, like Caracas, Bogota, San Diego, Medellin, etc., [41]. In Medellin, for instance, the temperature variation is between 18 and 28 ◦C, which in the soil could be just 30% variation (20 to 23 ◦C), a more acceptable temperature change. Future developments may consider thermal conductivity tests in order to increase insulation via materials science and thus decreasing even more the thermal changes.

For the characterization of the soil, specific gravity, liquid limit, optimum moisture content, plastic limit, and particle-size analysis, tests, have been performed according to respective ASTM test. The soil classification was done according to the ASTM D2487-17 by the Unified Soil Classification System (USCS) [42]. The density of soil samples was estimated based on weight and cylinder volume. Weight was measured using a Mettler Toledo balance, while the cylinder volume was measured using 0.05 mm precision caliper. Soil improvement was tested following the ASTM D2166-16 for the unconfined compression strength (UCS) [43]. For these tests, a universal testing apparatus Shimadzu AG250KN was used at a head speed of 0.68 mm/s.

X-ray diffraction (XRD) characterization was performed in a diffractometer X0Pert PRO, with λ = 1.5406 Å, Cu Kα radiation, voltage of 45 kV, and angles between 5◦ and 70◦ . Scanning electron microscopy with energy dispersive spectroscopy (SEM-EDS) was used to understand the microstructure of the samples in a JEOL JSM 6700 R equipment (Medellín, Colombia). For SEM, soil samples were dried at 30 ◦C for 24 h in a furnace, and coated with gold in a sputtering system at 15 mA AC for 30 s. Using the SEM images, the fiber diameter, rubber grain-size, and the mixed soil size distributions were estimated via image analysis with the Image-J software (version Java 1.8.0\_172).
