**2. Materials and Methods**

Nanoparticles used in this experimental work were supplied by Sigma-Aldrich (Saint Louis, MO, USA). The mineralogical composition of nano-calcium oxide (NC) is calcium oxide (<180 nm), and that of nano-silica (NS) (0.007 μm) is amorphous silicon dioxide diffraction patterns that have been given in previous works [22].

For the X-ray Fluorescence elemental analysis of the cement I42.5N, an S8 Tiger, Bruker Instruments (Karlsruhe, Germany) was used, and the content of oxides is given in Table 1. Table 2 shows the cement granulometry recorded using a particle size analyzer, Mastersizer 2000, Malvern Instruments (Malvern, UK). It is observed that 50% of the cement powder is below 4.667 μm. The density of the cement was measured according to ASTM-C188-95.


**Table 1.** Elemental analysis of cement, expressed in oxides.


Compressive strength (loading rate: 0.5 KN/s) was tested at 28 and 90 days, using a Technik ToniNorm device (Toni Technik GmbH, Berlin, Germany), and the mean value of six measurements was calculated. Open porosity was tested at the same ages, according to RILEM CPC 11.3 method, in water under vacuum and the mean value of three specimens was calculated. For Scanning Electron Microscopy, a Jeol JSM-6390LV, Oxford Instruments (Abingdon-on-Thames, UK), was used to observe the microstructure at 28 days. The thermogravimetric (TG) method was used to determine Ca(OH)2%wt and carbonates %wt content of the samples at 28 days. A dynamic method was applied to determine the mass loss of the samples at a function of temperature. A Neutsch F5 Jupiter instrument (Juliet, TN, USA) was used in an N2 atmosphere (50 mL per minute) from 50 ◦C to 1000 ◦C and 20 ◦C per minute step. Differencial thermogravimetric analysis (DTG) curves were extracted from TG thermograms.

The composition of cement pastes and nano-modified cement pastes is given in Table 3. A capital letter "R" is used for reference cement pastes. The letters "NL" are used for cement pastes with NC addition, and letters "NSL" are used for cement pastes with a combination of NC and NS additions. The small letter "w" is used for the specimens cured in water immersion, and the small letter "c" is used for the specimens cured in wetting-drying cycles. The NS particles were added in a pre-weight quantity of water and were subjected to ultrasonication for 30 min. The suspensions were directly added to the cement powder and stirred up to homogenization. The NC particles were added in water and were directly used in the mixture. The ultrasonication step was not needed as NC is instantly hydrated in a pre-weighted amount of water. The pastes' consistency was determined using a Vicat apparatus (according to EN 196-3:2016). Superplasticizer (SP) was used in nano-modified mixtures to maintain the systems' consistency (Vicat 6 <sup>±</sup> 1 mm). Pastes were cast in (25 <sup>×</sup> 25 <sup>×</sup> 50) mm<sup>3</sup> molds to test the mechanical properties and the microstructure. Moreover, specimens of (40 <sup>×</sup> 40 <sup>×</sup> 160) mm3 were produced for pre-cracking and to observe the self-healing effect. All the samples, regardless of their size, were cured under two different conditions. One condition was the water immersion until the performing of the tests. The second condition was the conduction of wetting-drying cycles, where the samples daily were subjected to water for six hours and set on specific conditions for 18 h (temperature ~20 ◦C and RH 65%).


**Table 3.** Composition of cement pastes.

The "3-point bending method" [12,21,23] was used for the formation of the pre-cracks. The specimens were marked and notched in the middle of their length, and they were carefully cracked after 28 days of curing (Figure 1). A low displacement rate (0.3 mm/min) was used to form cracks of less than 0.5 mm in width. The observation and measurement of cracks were performed with the Dino-Lite2 microscope and the DinoCapture2.0 software. An optical microscope was used to evaluate the healing of the cracks under the two different curing conditions. The width of the cracks was measured at three different ages. Firstly, immediately after the cracking, and at 14 and 28 days after the cracking. The cracked specimens were subjected to their previous curing conditions (water immersion and cycles) until testing.

After the optical observation of the healed cracks, the newly formed material was cautiously removed and collected in small containers. These samples were measured using ATR (Attenuated Total Reflectance) Spectroscopy to identify the compounds. A Cary 630 FTIR instrument with an ATR probe of Agilent Technologies was used to determine reflectance patterns from 375 cm−<sup>1</sup> to 4000 cm−<sup>1</sup> wavenumber.

**Figure 1.** Specimen notched in the middle of its length and pre-cracked, at 28 days from their production.
