In this study, the fine aggregate was sourced from Yogyakarta, Indonesia, and coarse aggregate with a nominal diameter of 12.7 mm was obtained from Central Java, Indonesia. The binder materials consisted of type F fly ash from Tanjung Jati B PLTU, Central Java, Indonesia, and slag obtained from PT. Krakatau Semen Indonesia, Banten, Indonesia. Both fly ash and slag must pass sieve no. 200 with a diameter of 75 μm, according to cement fineness specifications (SNI 15-2530-1991). X-ray fluorescence (XRF) testing was employed to determine the chemical compositions of fly ash and slag, and the results are shown in Table 1. The fly ash had a low calcium oxide (about 8.64%) but a high total content of SiO₂, Al₂O₃, and Fe₂O₃ (about 83.8%), so it includes type F fly ash according to ASTM C618 standards. On the other hand, the slag had a high calcium oxide content (about 62.1%).

Alkaline activators were prepared using a mixture of SH and SS, with varying ratios of SS/SH at 1.0, 1.5, and 2.0. The SH solution was created by dissolving 98% pure SH flakes in distilled water, maintaining a constant concentration of 2M across all mixtures. The SS solution used in this study was a commercially available gel with the production code BE-58. Figure 1 and Figure 2 show the raw materials and alkaline activators used in the manufacture the geopolymer concrete.

In order to achieve a 25 MPa design strength for normal concrete, the study employed a mixture design that followed the guidelines of SNI 7656:2012 (39), a modified version of ACI 211.1-91 (40) for Indonesian context. The GPC production involved replacing the cement with fly ash and slag, with the slag to total binder percentages ranging from 10% to 50%. To substitute the water-cement ratio, an alkaline activator was utilized with varying Al/Bi ratios of 0.40, 0.45, and 0.50. The aggregates and total binder were kept constant throughout the study. In Table 2, the results of the concrete compressive strength test according to the mix design are presented.

Ten mixtures were produced, consisting of nine GPCs and one OPC as a control (Table 2). The mixtures were designated as GPC1 to GPC9, each with different components. GPC1 to GPC5 were created to investigate the impact of slag percentage. Mixtures GPC3, GPC6, and GPC7 focused on the variations in the SS/SH ratio, while GPC3, GPC8, and GPC9 examined the effects of varying the Al/Bi ratio. The GPC mix designation represented the components of the mixture, for instance, S30-1.5-0.50, indicating a GPC mix with a 30% slag content, the ratio of SS/SH 1.5, and the ratio of Al/Bi 0.50. No additional water or superplasticizer was used in the mixtures.

The GPC mixing process began with the preparation of the necessary materials. A two-molar SH solution was prepared the day before its use, while the alkaline activator, comprising of SS and SH, was mixed and left to settle until it had reached room temperature. After that, a container was filled with a homogeneous mixture of precursor materials, which are fly ash and slag. Then, the GPC constituents were blended in a specific order: sand and coarse aggregate were mixed together for one minute, followed by the addition of a fly ash and slag binder and a two-minute mixing time. The alkaline activator solution was then added according to a precise ratio, and the mixture was given five minutes of blending time to ensure thorough homogeneity.

After the mixing process, the specimens underwent slump testing and were subsequently cast in cylindrical molds with dimensions of 100 mm in diameter and 200 mm in height. The casting was performed in two layers, ensuring compaction in each layer to avoid any cavities. The specimens were allowed to remain in the molds for 24 hours prior to being taken out and cured. For the OPC, curing is done by submerging the specimen in water, while GPC curing involves storing it in an airtight plastic container to maintain moisture at room temperature. Prior to testing, all specimens were shaped into cylindrical forms with a diameter of 100 mm and a height of 50 mm, a process that occurred after 28 days of curing.

The influence of various independent variables on porosity are presented in Figure 5. Based on Figure 5 (a), an increased in the slag content causes the porosity to decrease. The highest reduction value of 10.4% occurs at a slag percentage of 10% and the lowest (7.5%) occurs at 50% slag. The porosity value in OPC is 14.4%. Compared with the OPC, the GPC with slag recorded lower porosity.

The decrease in porosity with increased slag content is likely due to the binder reactions. The formation of C-A-S-H gel is due to the presence of slag, which is denser than geopolymer type gels, resulting in a less porous structure (14, 21). The slag used in this study contains a significantly higher CaO content than fly ash, at 62.1% and 8.6%, respectively (see Table 1). By increasing the amount of slag used in place of fly ash, more C-A-S-H gel will form, which will, in turn, decrease the number of pores in the concrete. In addition, at similar SS/SH ratio, GPC based on fly ash-slag has a larger amount (79%) of 10nm or smaller pores than GPC based on fly ash (16%). This can be attributed to the formation of C-S-H through slag activation in geopolymer concrete, which can fill most of the capillary pores (14). On the other hand, the increased porosity of the fly ash-based GPC is probably related to the gel’s binding, mainly in the form of N-A-S-H, which is less dense and more porous (13, 44).

The findings presented in Figure 5 (b) suggest that the Al/Bi ratio plays an integral role in porosity. When the ratio reaches 0.45, the minimum porosity of 8.82% is achieved. This can be attributed to the OH^- ions from the alkaline activator, which act as catalysts in the geopolymerization process and facilitate the release of Si⁴⁺ and Al³⁺ ions from the binder. If the amount of alkaline activator used is less, the leaching process of the aluminum silicate will also be slower, resulting in a poor geopolymeric structure with larger pores and lower strength (28). Meanwhile, at a higher Al/Bi ratio, the amount of alkaline activator used also increases so that the mixture is more dilute. The excess alkaline activator not utilized in the geopolymerization process will eventually become a pore in the concrete after it has hardened. The increasing number of pores causes the porosity to increase and, when compared with OPC, all GPC samples with various Al/Bi ratios have lower porosity.

Figure 5 (c) depicts an increased SS/SH ratio that reduces porosity. The highest porosity of 9.2% was observed for the SS/SH ratio of 1.0. On the other hand, porosity decreased to 8.8% and 7.6% for the SS/SH ratios of 1.5 and 2.0, respectively. The decrease in porosity is due to the higher SS/SH ratio and the higher proportion of sodium silicate results in a greater provision of free silicate. Increasing the amount of silicate in the mixture can increase the polymerization reaction and will form a N-A-S-H gel. Besides that, the presence of high calcium content in the slag will form a denser C-(A)-S-H gel. This is in line with the finding by other researcher (45), who stated that the smaller pores at high SS/SH ratios were due to the increased silica content resulted in a denser polymerized gel with excellent mechanical properties.

According to the description above, porosity is closely related to the number of pores in the concrete. GPC with low molarity SH (2M) and varying slag percentages, Al/Bi ratios, and SS/SH ratios show lower porosity than OPC. This indicates that GPC has fewer pores and a denser structure, which will increase the mechanical properties of concrete.

The sorptivity test results indicate that the water absorption rate on the concrete surface and the initial and secondary sorptivity have a strong relationship with the time^1/2. This can be seen from the data presented in Figure 6, which demonstrates the correlative relationship between the two variables.

The rate of water absorption can be seen in two different phases. Initially, a faster rate of absorption process take place and started to decrease after 1 day. This condition is most likely caused by the pores in the concrete, which are partially saturated (46). In Figure 6(a), the initial absorption of OPC is higher than the other GPC mixtures. However, after 1 day it showed that GPC with 10% slag content (GPC1) had higher absorption than OPC. Based on the data presented in Figure 6(b) and 6(c), it can be observed that all GPC, regardless of the Al/Bi ratios and SS/SH ratios, exhibit lower water absorption values compared to OPC. This trend is observed both at the initial stage of water absorption and throughout the extended water absorption period. This shows that the proportion of the GPC2-GPC9 mixture has resulted smaller pores than OPC.

In accordance with ASTM C1858, the slopes of the graphs in Figure 6 may be used to determine the initial and secondary sorptivity values. However, some requirements must be met, where the graph of the relationship of water absorption with time must show a linear relationship, with a coefficient of correlation of more than 0.98. The sorptivity values obtained through the analysis of these slopes are shown in Figure 7.

Figure 7(a) confirms that a higher percentage of slag used will result in lower initial and secondary sorptivity. Samples with 20%, 30%, 40%, and 50% have initial sorptivities of 10.4, 9.6, 8.2, and 6.0 x 10^-3 mm/s^1/2, respectively. The proportion of GPC1 mixture with 10% slag has a slightly different sorptivity value, which has an initial sorptivity of 9 x 10^-3 mm/s^1/2, lower than the initial sorptivity of GPC2 and GPC3. However, GPC1 has a greater secondary sorptivity compared to other GPC mixtures. The secondary sorptivity of GPC is much lower than OPC with a very significant difference, except for GPC1. Because GPC1 has the lowest slag content, relatively less C-A-S-H gel is formed, which causes the concrete to be less dense. As a result, GPC1 has the highest secondary sorptivity when compared to other geopolymer concretes. The secondary sorptivity of GPC2, GPC3 and GPC4 is 3.0, 2.0, and 0.8 x 10^-4 mm/s^1/2, respectively. Meanwhile, GPC1 has a secondary sorptivity of 12 x 10^-4 mm/s^1/2, almost the same as the secondary sorptivity of OPC of 13 x 10^-4 mm/s^1/2. Unfortunately, the coefficient of correlation of the GPC5 graph is less than 0.98, so it isn’t possible to figure out its secondary sorptivity.

The decline in sorptivity value as the slag content increases signifies that the inclusion of slag in the fly ash-based GPC mixture enhances the microstructural performance, which, in turn, makes the concrete last longer, in terms of water access so that the sorptivity value becomes low (28). The increase of slag will cause the formation of denser C-A-S-H gel with finer pore sizes, thereby preventing water from entering deeper into the concrete (18). In general, the presence of C-A-S-H and N-A-S-H gels in fly ash/slag concrete describes the beneficial behavior that occurs in this material. The formation of N-A-S-H gel during the geopolymerization process is instrumental in reducing the porosity of concrete. This gel fills the voids in the paste and interfaces between the paste and the aggregate, leading to a denser material. The C-A-S-H gel is associated with greater densification and compactness, due to the fact that fly ash/slag concrete has fewer porous capillaries or pores (9). In addition, the less porous matrix is also associated with a smaller slag particle size than the fly ash particles which develops a micro filler effect (22).

As shown in Figure 7(b), the sorptivity coefficient exhibits an increase in conjunction with the rise in the Al/Bi ratio. This can be attributed to the higher volume of liquid present, which results in a more diluted mixture. After the concrete hardens, this dilution leads to the formation of additional capillary pores. The lowest initial sorptivity value of 8.1 x 10^-3 mm/s^1/2 was achieved at an Al/Bi ratio of 0.40, followed by Al/Bi ratios of 0.45 and 0.50, which yielded initial sorptivity values of 9.6 x 10^-3 mm/s^1/2 and 10.6 x 10^-3 mm/s^1/2, respectively. When compared with the initial sorptivity of OPC, GPC has a lower value, but the difference is not significant. This indicates that GPC and OPC have almost the same ability to absorb water at the beginning of the period. In GPC, variations in the Al/Bi ratio have no effect on secondary sorptivity, whereas the three variations have the same secondary sorptivity. Meanwhile, the value of secondary sorptivity on GPC and OPC differs very significantly. The GPC with all variations of the Al/Bi ratio has a much lower secondary sorptivity of 2 x 10^-4 mm/s^1/2, while the OPC is 13 x 10^-4 mm/s^1/2. This is because GPC has fewer pores and a denser structure which is indicated by a lower rate of water absorption value than OPC, as seen in Figure 6(b).

Figure 7(c) illustrates the effect of the SS/SH ratio on the sorptivity coefficient. As the SS/SH ratio increases, the initial sorptivity decreases. This decrease was caused by the fact that the alkaline activator was mostly composed of sodium silicate which has a high viscosity. In the end, it leaves fewer pores after the concrete hardens. In addition, the smaller pores at high SS/SH ratios are because the increased silica amount produces denser polymerized gels with superior mechanical characteristics (45). In contrast, the secondary sorptivity shows a direct relationship with the SS/SH ratio. As the SS/SH ratio increases, the secondary sorptivity tends to increase as well, although the observed increase is not substantial. It is worth noting that all GPCs with varying SS/SH ratios exhibited lower initial and secondary sorptivity values compared to OPC. When the SS/SH ratio was 1.0, 1.5, and 2.0, the initial sorptivity value was 9.8, 9.6, and 7.2 x 10^-3 mm/s^1/2 and the secondary sorptivity was 2.0, 2.0, and 3.0 x 10^-4 mm/s^1/2, respectively. The OPC has an initial and secondary sorptivity value of 10.6 x 10^-3 mm/s^1/2 and 13 x 10^-4 mm/s^1/2, respectively.

Based on the previous description, it can be seen that the sorptivity coefficient is closely related to the level of water absorption in concrete. The higher the water absorption level, the more the sorptivity coefficient increases. In general, GPC with low molarity SH (2M) and with variations in slag percentage, Al/Bi ratio, and SS/SH ratio has lower sorptivity than OPC. This shows that GPC has a lower rate of water absorption, which is related to the smaller number of pores. So it can be concluded that the sorptivity coefficient is directly proportional to the porosity of the concrete.

Through the RCPT test, the chloride permeability of GPC was evaluated, as illustrated by the total charge passed value in Figure 8. To further ascertain the depth of chloride penetration, silver nitrate solution (AgNO₃) was sprayed on the GPC surface, and the results of this test are displayed in Figure 9.

As seen in Figure 8(a), when the proportion of slag increases as a replacement for fly ash, the chloride permeability of the GPC decreases. This is indicated by the smaller total charged passes. Mixes with slag content of 10%, 20%, 30%, 40%, and 50%, the total charged successively passed was 845.33 C, 739.80 C, 527.18 C, 438.08 C, and 379.13 C, respectively. The presence of calcium-based hydrated materials is responsible for the decreased chloride permeability as the percentage of slag increases, which form together with the geopolymer gel and produce a strong and compact microstructure, thus preventing the entry of chloride ions through pore capillaries. In addition, in the presence of slag, alumina and silica ions (which are very soluble and widely available) will be more easily soluble, thereby increasing the polycondensation mechanism (22). The matrix formed with a decrease in pore size diameter is also associated with a smaller slag particle size than fly ash particles, which develops a micro filler effect (23). The correlation between chloride ion penetration and other pore characteristics, such as tortuosity, is evident according to (18). When fly ash is replaced with slag, the tortuosity of the pore structure increases, which could be a major factor in the decrease of chloride penetration. According to the category in ASTM C 1202-12, all GPC mixtures with varying percentages of slag and OPC have chloride permeabilities in the very low category because they have total charged passes of less than 1000C.

Figure 8(b) shows that the Al/Bi ratio 0.45 produces the least chloride permeability, as indicated by the charged passed value of 527.18 C while the charged passed values of 0.40 and 0.50 Al/Bi ratios were 565.43 C and 628.43 C, respectively. The general trend is that higher Al/Bi ratios will lead to higher porosity and a higher rate of chloride penetration (47). However, as stated in the sub-discussion on porosity, at an Al/Bi ratio of 0.40, the amount of alkaline activator is insufficient for the geopolymerisation process to produce a geopolymeric structure with larger pores (28). Likewise, if the Al/Bi ratio increases from 0.45 to 0.50, the excess alkaline activator that is not used in the geopolymerisation process will leave pores with a larger size and volume. This condition causes higher chloride permeability.

When the SS/SH ratio is increased, the chloride permeability is lessened, as illustrated in Figure 8(c). The highest total charged passed of 560.25 C was noted at SS/SH 1.0, followed by SS/SH 1.5 and SS/SH 2.0, with the total charged passed of 527.18 C and 353.03 C, respectively. At the same SS/SH ratio, the reduction in chloride permeability is in line with the decrease in porosity and sorptivity in concrete. This is because a high SS/SH ratio will increase the silica content and cause the gel to polymerize into a denser matrix, making the pores smaller. In GPC, with all variations of the SS/SH ratio and OPC, it has a very low chloride permeability category because it has a charged passed value of less than 1000 C. When compared to OPC, which has a total charged passed of 472.50 C, the mixture of GPC with slag percentages of 40%, 50%, and SS/SH ratio of 2.0 have lower chloride permeability.

Based on Figure 9, the depth of penetration of chloride at various percentages of slag, variations in Al/Bi ratios, and variations in SS/SH ratios is directly proportional to the value of porosity, rate of water absorption, and chloride permeability.

As illustrated in Figure 9(a), an increase in slag content as a replacement for fly ash leads to a decrease in chloride penetration depth. The depth was found to be the greatest (31.5 mm) when 10% slag was used and the lowest (5.7 mm) when 50% slag was used. This phenomenon can be attributed to the dense matrix that slags with a high CaO content creates, resulting in a reduction in the size and number of the concrete pores. Subsequently, the slag in GPC promotes a lower amount of chloride attack since the chlorides penetrate concrete through capillary pores. When compared to OPC, GPC based on fly ash-slag has better durability in a chloride environment, as evidenced by a much lower depth of chloride penetration (OPC has a chloride penetration depth of 39.6 mm).

The Al/Bi ratio that produced the lowest chloride penetration depth is 0.45, with a depth of 16.1 mm. At 0.40 and 0.50 ratios of Al/Bi, the GPC has chloride penetration depths of 24.6 mm and 18.0 mm, respectively, as shown in Figure 9(b). Meanwhile, Figure 9(c) shows that geopolymer concrete with a minimum chloride penetration depth of 16.1 mm is noted at an SS/SH ratio of 1.5. The SS/SH ratios of 1.0 and 2.0 result in deeper penetration depths of 24.3 mm and 18.2 mm, respectively. Compared with OPC, which has a chloride penetration depth of 39.6 mm, GPC exhibits a much lower penetration depth for all variations of the Al/Bi ratio and SS/SH ratio. These levels of chloride penetration depth can be seen in Figure 10.

The depth of chloride ion penetration in concrete relates to the amount and size of pores, water absorption, and the presence of external pressure. Based on the previous description, GPC with SH 2M molarity has fewer and smaller pores, as well as lower water absorption. As a result, GPC has a lower chloride ion penetration depth and greater resistance to chloride attack than OPC.