**1. Introduction**

Concrete, as one of the most commonly used construction materials, is exposed to plenty of aggressive environmental factors [1]. Transport of external media is conducted by a connected pore structure [2]. This is the reason why the description of the pore network deserves special attention in the analysis of concrete's durability [3,4].

Gas permeability is one of the parameters that describes the pore structure of materials, therefore, it is considered one of the most popular parameters to determine their quality and durability [3,5]. Among the variety of methods used to determine the gas permeability, the most common are the CEMBUREAU method [6] and the oxygen permeability index. The principle of these tests is to measure the amount of gas passing through the sample with a known cross-sectional area per unit of time. The most commonly used gases are oxygen, nitrogen, or dry air. These methods of measurement refer to laboratory tests. In situ testing by Figg's method [7] or others methods described in the literature [8,9] is acceptable but the main disadvantage of them is the lack of possibility to control the impact of temperature and of moisture on the material. These parameters strongly affect the result of permeability measurements [10,11].

**Citation:** Duzy, P.; Choinska ˙ Colombel, M.; Hager, I.; Amiri, O. The Effect of Preconditioning Temperature on Gas Permeability of Alkali-Activated Concretes. *Mater. Proc.* **2023**, *13*, 38. https://doi.org/ 10.3390/materproc2023013038

Academic Editors: Katarzyna Mróz, Tomasz Tracz and Tomasz Zdeb

Published: 20 February 2023

**Copyright:** © 2023 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

Gas permeability measurements are basic to the determination of the durability of constructions and modelling the service life [12]. The CEMBUREAU method is considered a reference method for air permeability for other methods. However, laboratory methods require temperature preconditioning of the specimens [6,13]. The temperature of drying ranges from 50 ◦C to 105 ◦C. It is clear that the results of measurements will not be the same for variable conditions. The definition of the temperature impact on the results obtained may allow researchers to adjust the condition to the tested materials and compare it with others.

The temperature impact on geopolymers' microstructure is the subject of many investigations [14,15]. Many of them focus on behaviour in high temperatures up to 1000 ◦C. [16,17]. However, it is a well-known fact that the temperature of curing strongly affects geopolymer properties, especially alkali-activated fly ashes [18,19]. In addition, for alkali-activated slag, temperature of curing influences its microstructure and porosity [20]. Taking into consideration the influence of preconditioning of specimens on gas permeability, the question about its applicability to geopolymer concretes is an unsolved problem.

The temperature that activates fly ashes has been investigated by many researchers in recent years [21,22]. Palomo et al. [23] presented the evolution of properties of alkaliactivated fly ash cured at temperatures between 65 ◦C and 85 ◦C. The ingress of compressive strength with high temperature reached 60 MPa. With these changes, it is obvious to assume that all the properties connected with porosity were also affected by curing temperature. These studies only confirmed the theories made much earlier [24]. Similar research was conducted to define the influence of the activator used [25] for mixes and their behaviours in temperatures between 75 ◦C and 95 ◦C. Unfortunately, there is no possibility to specify one general rule to describe the influence of curing temperature on porosity and microstructure of alkali-activated materials. It depends on the type of precursor and activator and on the temperature applied. In the case of gas permeability measurement, it is more necessary to compare the results obtained for materials with different compositions than to obtain accurate values of permeability. In this paper, the influence of preconditioning temperature on gas permeability measurements and comparison of this influence on different compositions of alkali-activated concretes are presented.

#### **2. Materials**

The studies were conducted on six compositions of alkali-activated concretes based on fly ash (Połaniec powerplant, Poland) with ground granulated blast furnace slag (Ekocem D ˛abrowa Górnicza, Poland) replacement by mass. Three levels of GGBFS dosage were examined: 5%, 20%, and 35%. Two types of coarse aggregate were used to detect its influence on gas permeability. The tested blends were denoted AAC5B, AAC20B, and AAC35B for basalt aggregate and AAC5D, AAC20D, and AAC35D for dolomite aggregate with 5%, 20%, and 35% of slag, respectively. According to ASTM C618 Standard Specification for Coal Fly Ash and Raw or Calcined Natural Pozzolan for Use in Concrete [26], the FA was categorised as class F, and is mainly composed of silicon dioxide and aluminium oxide. In the case of the slag used, it mainly consists of calcium oxide and silicon oxide. The chemical compositions of precursors are presented in Table 1. For activation, the diluted sodium silicate solution Geosil® 34417, supplied by Woellner, was used.

**Table 1.** Oxide composition of fly ash and ground granulated blast furnace slag.


The mixing procedure was performed based on prior experience of the research group [27]. The first step was to prepare pastes. Fly ash was mixed with diluted alkaline solution until a homogenous consistency was obtained. Slag was added with constant stirring. At that time, a mixture of coarse aggregate (basalt or dolomite) and quartz sand was prepared in a larger mixer. When the paste was ready, it was added to the mixer with the aggregate and distributed evenly throughout the mass. The samples were cast and compacted on a shaking table in cylindrical moulds with diameter 11 cm and height 22 cm. They were stored in ambient conditions at 20 ± 2 ◦C and protected from water evaporation by plastic film. For the prepared samples, no thermal curing was applied.

#### **3. Methods**

The CEMBUREAU method is still one of the most popular methods of gas permeability measurements [28]. The values of inlet and outlet pressures are steady during the measurement. Identification of gas permeability is based on stabilised permanent gas flow. The atmospheric pressure and temperature, which may affect the results, are also taken into account for permeability estimation. Lateral pressure of 8 bars is applied to the rubber gum surrounding the specimen to exclude gas leaking around the sample.

Measurements in accordance with this standard were initially carried out within samples which had never been dried before. The value of gas permeability started to stabilise after 240 days and this continued for up to 360 days, as given in Figure 1.

**Figure 1.** Gas permeability development in time for samples stored in room conditions: (**a**) basalt aggregate; (**b**) dolomite aggregate.

In order to exclude test disturbance caused by the age of the materials, the impact of preconditioning temperature started to be investigated once the values of permeability stabilised (after 1 year). AAMs are highly sensitive to external temperature application. Referring to this fact, three temperatures of preconditioning were taken into consideration: 40 ◦C, 80 ◦C, and 105 ◦C.

After the first step, corresponding to the measurement performed on the natural state of the material (denoted as 20 ◦C, corresponding to data at 360 days, see Figure 2), further stages of research were run. For each step of temperature, the drying process was carried out in order to reach a stable mass of samples (relative mass loss Δm less than 0.1% during 24 h). In order to cool down the specimens, they were placed in a desiccator fulfilled with hydrophobic pellets to reduce moisture. After 24 h of cooling down, the mass of each material was verified. Then, the permeability test was performed. For each material, two specimens were tested. Just after measurements, the specimens were placed in a dryer at the next level of temperature and the procedure was repeated. All of the tests were conducted on the same specimens to avoid impact of heterogeneity of concretes. The total drying process and changes in specimens' mass are shown in Figure 2.

**Figure 2.** Scheme of drying process.

The values of gas permeability coefficients were calculated according to modified Darcy's law. The analysis of the obtained results consists of influence of temperature on the values of apparent and intrinsic gas permeability and the occurrence of boundary slippage effect related to the Klinkenberg effect [29].

#### **4. Results and Discussion**

Apparent permeabilities as a function of inversed mean pressure show permeability increased with temperature and followed Klinkenberg's law. The pore fineness increased, probably because of the creation of newly accessible fine pores during drying and cooling. In parallel, a competitive action is normally due to expansion of already present and accessible pores, caused by water evaporation from internal surfaces [30,31].

Table 2 presents intrinsic permeability values for initial, not oven-dried, specimens (20◦C) of all the tested concretes. These intrinsic permeabilities were calculated following Klinkenberg's law, using apparent permeabilities. The values for low slag content materials (AAC5) are the lowest, independently of aggregate type, in comparison to other materials. Globally, there is no tendency in permeability evolution as a function of aggregate type.

**Table 2.** The intrinsic permeability values for specimens preconditioned at 20 ◦C.


The analysis presented below is based on intrinsic gas permeability values and the relation between them for appropriate drying temperatures. The ratio proposed for the further analysis is ki/ky, where i, y mean drying temperatures. Values of ki/k20 are shown in Figure 3.

**Figure 3.** Values of ki/k20 ratios for temperatures of preconditioning.

For almost all of the temperatures (40, 80, 105 ◦C), the values of permeability for dolomite-based concrete increased more than for basalt-based concrete. This behaviour is probably due to physical changes in the microstructure of the zone between paste and aggregate, as the latter is not intrinsically impacted at these temperatures. The impact of the drying process at 40 ◦C is much higher for AAC35B and AAC35D than for AAC5B and AAC5D, which is clearly visible in Figure 4. For AAC35B, the value of permeability increased from 6.07 × <sup>10</sup>−<sup>17</sup> <sup>m</sup><sup>2</sup> to 34.8 × <sup>10</sup>−<sup>17</sup> <sup>m</sup><sup>2</sup> for 20 ◦C and 40 ◦C, respectively (the k40/k20 ratio was equal to 5.74). For AAC35D, permeability at these temperatures reached 4.65 × <sup>10</sup>−<sup>17</sup> m2 and 38.2 × <sup>10</sup>−<sup>17</sup> m2, making the k40/k20 ratio equal to 8.22. However, low slag content materials show lower sensitivity when heated to a temperature of 40 ◦C. The values of the k40/k20 ratio for AAC5B and AAC5D were 2.71 and 2.64, respectively. The impact of the drying process at over 40 ◦C is related to complex physical and chemical changes and will be discussed below.

**Figure 4.** Impact of drying temperature on gas permeability of AAC.

Materials with the highest amount of fly ashes reveal a significant response to temperatures between 40 ◦C and 105 ◦C. As was explained in the Introduction, these materials are sensitive to temperature treatment. On the one hand, this may be due to the activation of the FA and therefore the creation of new phases, but on the other hand, it may be due to a rapid release of water from the pores of the material and generation of microcracks in the material, amplifying gas flow. All the absolute values of gas permeability are summed up in Figure 4.

As the results presented before are mainly related to aggregates' nature, the results presented in Figure 4 are strongly focused on slag content impact. Three families of results, depending on slag amount, may be highlighted. In the case of 5% and 20% slag content, the influence of the drying temperature is slight in comparison to 35% slag. Thermal expansion strain differences between high-content slag paste and aggregate should be more significant than for other pastes. It provides geometrical incompatibilities and therefore probably microcracking, causing an increase in permeability of up to two orders of magnitude.

The analysis carried out has brought many valuable observations. The difference in drying temperature impact shows distinct sensitivity of materials to drying. Furthermore, a meaningful impact of slag addition is highlighted.

#### **5. Conclusions**

The tests carried out were aimed at comparing the influence of the sample preconditioning temperature on the result of the gas permeability test. The variety of standards used for measurements may significantly affect the obtained result, which has been proven in this paper. Based on the obtained experimental results, the following conclusions can be drawn:

• The drying temperature of the samples significantly affects the obtained permeability values.


Experimental research has clarified the possibility of comparing the results obtained for materials with different binders and aggregates. Conclusions drawn from these studies may provide guidance for further analysis and interpretation of permeability test results performed on alkali-activated concrete.

**Author Contributions:** Conceptualisation, I.H., M.C.C. and O.A.; methodology, M.C.C. and P.D.; validation, I.H. and O.A.; formal analysis, M.C.C. and I.H.; investigation, P.D.; data curation, P.D. and M.C.C.; writing—original draft preparation, P.D. and M.C.C.; writing—review and editing, M.C.C., I.H., O.A. and P.D.; visualisation, P.D. and M.C.C.; supervision, M.C.C.; project administration, I.H.; funding acquisition, I.H. and M.C.C. All authors have read and agreed to the published version of the manuscript.

**Funding:** The authors of this research were financially supported by SMART-G Smart Geopolymers project financed by National Centre for Research and Development (NCBR)–ERA-MIN2-3/SMART-G/1/2022.

**Data Availability Statement:** The data presented in this study are available on request from the corresponding author.

**Acknowledgments:** The publication cost of this paper was covered with funds of the Polish National Agency for Academic Exchange (NAWA): "MATBUD'2023–Developing international scientific cooperation in the field of building materials engineering" BPI/WTP/2021/1/00002, MATBUD'2023.

**Conflicts of Interest:** The authors declare no conflict of interest.
