Surface treatment with nano-silica and magnesium potassium phosphate cement co-action for enhancing recycled aggregate concrete

: In this study, surface treatment, using blended materials of nano-silica (NS) and magnesium potassium phosphate cement (MKPC) slurries containing four dosages of NS, was applied to recycled aggregate (RA) with distinct RA maintenance schemes to enhance the RA concrete (RAC) performance and control NS dosage for economy. The results imply that the NS + MKPC slurry can e ﬃ ciently bolster the bonding with RA and the new interfacial transition zone (ITZ), contributing to a distinct enhancement of RAC ’ s macro-properties. Besides, the optimal NS dosage and RA maintenance scheme revealed by the simpli ﬁ ed linear weighted sum optimization method were 3% and no pre-curing procedure, respectively. On the basis of the aforementioned optimal conditions, the modi ﬁ ed RAC (C-3-N) achieved enhanced compressive strength by 40.75%, split tensile strength by 46.26%, and chloride ion penetration resistance by 65.93% in comparison with the untreated RAC0. Moreover, the advantages observed in C-3-N were attributed to the exceptional microstructural characteristics in both the NS + MKPC slurry and the new ITZ. This study establishes the potential to augment the e ﬃ cacy of nanomaterials in reinforcing RAC and enhance the economic viability and practicability of RAC applications. Simultaneously, these advancements contribute to fostering sustainable development within the construction industry and yielding environmental bene ﬁ ts.


Abbreviations
untreated RA R-1-P RA treated by the S-1 under the pre-curing scheme R-1-N RA treated by the S-1 under the no pre-curing scheme R-2-P RA treated by the S-2 under the pre-curing scheme R-2-N RA treated by the S-2 under the no pre-curing scheme R-3-P RA treated by the S-3 under the pre-curing scheme R-3-N RA treated by the S-3 under the no pre-curing scheme R-4-P RA treated by the S-4 under the pre-curing scheme R-4-N RA treated by the S-4 under the no pre-curing scheme NAC0 concrete produced using N0 RAC0 concrete produced using R0 C-1-P concrete produced using R-1-P C-1-N concrete produced using R-1-N C-2-P concrete produced using R-2-P C-2-N concrete produced using R-2-N C-3-P concrete produced using R-3-P C-3-N concrete produced using R-3-N C-4-P concrete produced using R-4-P C-4-N concrete produced using R- Recycled aggregate (RA) processed from old demolished concrete structures has been adopted as a partial replacement of natural aggregate (NA) in the production of new concrete in numerous countries across the globe [1,2], which has been demonstrated to be an effective strategy for decreasing the quantity of construction and demolition waste [3,4], conserving natural resources [5,6], mitigating environmental impacts [7], as well as saving construction costs [8][9][10].Moreover, the residual mortar layer on RA surfaces exhibits high pore and micro-crack contents as a result of the crushing procedure [11].Consequently, it is generally acknowledged that RA concrete (RAC) made from RA has deteriorated mechanical properties [12][13][14][15][16][17][18] and durability [19][20][21][22][23][24][25] in comparison with conventional concrete.
Currently, the study of nanomaterials in enhancing RAC properties has garnered considerable attention within the construction sector [26][27][28][29] due to their potential function in concrete mixtures, i.e., the ability of nanoparticles to fill extremely small-sized pores as well as to react with calcium hydroxide (CH) facilitating the formation of an additional calcium silicate hydrate (C-S-H) gel within the system, as a result of their amorphous nature and ultrahigh surface area.Typically, nanomaterials have been employed in two approaches for RAC's mechanical properties and durability enhancement [30][31][32][33][34][35][36].Among them, one method is adding nanomaterials (e.g., nano-silica (NS) [30,31]) directly during the RAC mixing process to enhance the properties of the new cement mortar, leveraging the principle that nanomaterials can improve the pore structure of new cement paste.Nevertheless, incorporating nanomaterial as a supplementary cementitious material would deteriorate the workability of fresh concrete [31], which led to extra admixtures including efficient water reducers.Consequently, another modifying treatment has been proposed that can be attained through the pre-coating of RA with a nanomaterial layer [37][38][39] to enhance the new interfacial transition zone (ITZ) between RA and the new cement mortar, which has been confirmed [40][41][42][43][44] to exhibit a strong correlation with the durability and strength of RAC [45][46][47][48][49].
Notwithstanding, surface treatment of RA by nanomaterials' pre-coating requires a massive nanoparticle in which porous RA absorbs a certain amount of nanoparticle, resulting in the method not being economical.Thus, if nanomaterials tend to be extensively utilized as surface treatment materials for RAC modification, it is critical to control their dosages or identify a substance that can be utilized in conjunction with the nano-solutions to ensure the economic efficiency of RAC projects [49].Moreover, Zhang et al. [37] proposed the preparation of surface treatment slurries for RA utilizing NS mixed with ordinary cement.Based on a comparative study with nSi + nCa nano-slurry, it was determined that the cement + nSi slurry enhanced the new ITZ in RAC and had a secured modification effect on RAC's mechanical properties and durability.Consequently, this hybrid nano-slurry served as a superior substitute for the pure nano-slurry, which contained a greater quantity of nanomaterials, due to its cost-effectiveness and regulated nanomaterial dosage.Nevertheless, as a surface treatment slurry for RA, its bonding performance with RA has not been considered.In accordance with the damage mechanism of RAC, the hypothetical low bond strength of the surface treatment slurry to RA would create weak links in the RAC, enhancing the velocity of crack propagation and corrosive substance ingress when subjected to external loads or erosions.Furthermore, the cement usage in surface-coating materials likewise aids in the increased carbon emissions of RAC projects [50].Thus, the discovery of an environmentally friendly material with excellent bonding properties to RA that can be employed in conjunction with nanomaterials as a surface treatment slurry is essential in strengthening the performance of RAC composites as well as their practical applications.
Magnesium potassium phosphate cement (MKPC), recognized as an environmentally friendly cementitious substance [51], has found extensive application [52][53][54][55][56] in the strengthening and restoration of civil and road engineering.This is attributed to its superior bonding characteristics, surpassing those of ordinary Portland cement (OPC) [55,[57][58][59][60]. Typically, MKPC works on two levels for high-strength bonds with cementitious materials including concrete, i.e., excellent mechanical anchorage as a result of the micro-filling effect induced by substrate penetration [61][62][63], and chemical interactions with the alkali components of the substrate, which was explicitly demonstrated in this work.Accordingly, whether MKPC substitution for OPC to form NS + MKPC slurry can ensure the formation of a strong bond at the interface with RA and repair the RAC weak point-new ITZ, as well as what effects it may have on RAC macroperformance, needs to be further investigated.Moreover, in the case of surface treatment slurry consisting of two materials, it is necessary to examine the optimal proportion as well as reaction mechanism of the two materials, namely, NS and MKPC, for the purpose of obtaining the optimal enhancement of the slurry and the RAC properties.
Thus, the purpose of this study is to explore the innovative application of NS + MKPC in enhancing RAC properties through preparing NS + MKPC slurries with four NS dosing levels to pre-coating RA along with using two RA maintenance schemes to provide an efficient, practical, and cost-effective modification method for RAC.These are the aims of our research: 1) To assess the efficacy of surface treatment techniques in improving the macro-characteristics of RAC by conducting compressive and split tensile strength tests, as well as rapid chloride permeability tests (RCPTs) of NA concrete (NAC) and different types of RAC. 2) To select the optimal NS concentration and RA maintenance scheme without introducing any bias by using the linear weighted sum optimization method under comprehensive consideration of RAC's macro-performances [64,65].3) To comprehend the strengthening mechanism of surface treatment with NS + MKPC on RAC macro-properties and indicate why/how is the optimal reinforcement efficiency generated through micro-characteristic tests of the bonded interfaces, the new ITZ, and NS + MKPC hardened pastes.
Furthermore, regarding micro-testing, mention that interface models for simulating bonding interfaces and new ITZ were prepared to enhance testing accuracy and statistical efficiency, and scanning electron microscopy (SEM) coupled with an energy-dispersive spectrometer (EDS) was applied to characterize the composition, microstructure, as well as porosity evaluation of interfaces and NS + MKPC pastes.Furthermore, a double-mixing method (DM) [25,66] was adopted to further optimize the new ITZ and thus enhance RAC properties.
2 Materials and methods

Materials and mix design
The coarse aggregates used in this research were RA procured from a supplier in Tianjin, China, and NA obtained from crushed limestone.The RA particle size distribution was essentially identical to that of the NA.Besides, the NA and untreated RA were marked as N0 and R0, respectively, and their particle size distributions, physical properties, and crush values are indicated in Figure 1 and Table 1, respectively.Ordinary river sand was used as the fine aggregate.
The raw materials for the NS + MKPC surface treatment slurries comprised dead burned magnesia (MgO, 1,600°C) powders, industrial-grade potassium dihydrogen Surface treatment with nano-silica and MKPC for enhancing RAC  3 phosphate (KH 2 PO 4 ) particles, as well as NS with purities of 92, 98.7, and 98.5%, respectively.Moreover, the apparent density of MgO was 590 m 2 /kg, and the mean particle size of NS was 102 nm.Additional 99.8% pure borax granules and the high-efficiency water reducer were incorporated to control the slurry's setting time and fluidity.In this study, 1, 2, 3, and 4% NS doping of the mass of MgO were investigated.Additionally, the proportions of the distinct raw materials in each series of slurries are illustrated in Table 2 for pre-treatment of 1,000 kg R0.On the condition that the NS ratio was 1, 2, 3, and 4%, and the corresponding NS + MKPC slurries were labeled S-1, S-2, S-3, and S-4, respectively.The subsequent methods were employed to produce four NS + MKPC slurries: first, the NS weighed in proportion, and some water along with surfactant added for the hard-to-disperse nature of the nanoparticles was blended thoroughly utilizing an ultrasonic disperser.Subsequently, in accordance with the ratios in Table 2, the corresponding NS solution was combined with the remaining water and the other raw materials for 120 s.As a result, four enhancing slurries exhibiting desirable dispersibility were produced.
The surface treatment process of the RA consisted of soaking R0 in the corresponding pre-made slurry for 5 min, accordingly, removing the RA from the tank, and adopting a screen to remove the excess slurry adhering to the RA.Besides, the treated RAs were divided into two groups: one group was not pre-cured after slurries initial setting prior to the casting of the corresponding concrete, and the other group was required to dwell in a curing room with a temperature of 20 ± 2°C and a relative humidity of 64 ± 5% for 28 days.Furthermore, the treated RAs were labeled as R-1-P, R-1-N, R-2-P, R-2-N, R-3-P, R-3-P, R-4-N, and R-4-N, respectively, where the front letters "R" imply the modified RA; the numbers "1", "2", "3", and "4" denote the NS proportions in percentage; and "P" or "N" symbolizes the RA maintenance schemes as pre-curing and no pre-curing, which were previously described therein.Except for that, the water absorption rates of R-1-P, R-2-P, R-3-P, and R-4-P were measured to be 5.2, 5.1, 5.3, and 5.4%, respectively.
As illustrated in Table 3, the treated RAs were used for producing eight groups of concrete mixtures, and N0 and R0 were employed to pour concrete NAC0 and RAC0, respectively.The P.O.42.5 cement with an apparent density of 3,090 kg/m 3 was adopted for the concrete mixture preparation in the experimental program, and the tests likewise employed a water reducer to adjust the fresh concrete slump.Each substance was combined in the following proportion of cement: water: aggregate: sand: water reducer = 431:216:1,020:731:1.2.To eliminate the influence of RAs' high water absorption on the water volume available for the chemical reactions, i.e., the effective "free" water (W ef ) [67,68], and, accordingly, the effective water-to-cement (W ef /C) ratio [69] and workability of the RAC, extra mixing water (W ex ) was added to RAC mixture [70,71].Meanwhile, for maintaining a consistent W ef /C, W ex in RAC0 was calculated as 29.88 kg/m 3 according to the following equation [72]: where w Ragg = 1.53% and w Nagg = 0.68%, indicating the rate of moisture content of R0 and N0, respectively; h Rabs = 5.40% and h Nabs = 1.62%, depicting the moisture absorption rate of R0 and N0, respectively; m agg represents the mass of aggregates.As the water absorption and moisture content of R-1-P, R-2-P, R-3-P, and R-4-P demonstrated no significant differences in comparison with those of R0, the same W ex value was used for all of them [28].For R-1-N, R-2-N, R-3-N, and R-4-N, it is challenging to assess the water absorption of RA in the absence of pre-curing, their W ex was determined to be identical as R-1-P, R-2-P, R-3-P, and R-4-P, respectively, to examine how modified RAC performance is affected by RA maintenance schemes.
As previously documented, all surface-treated RAs were poured into RAC using DM [66], as illustrated in Figure 2. The slumps of fresh concrete were tested in accordance with GB/T 50080-2016 [73], and the statistical results are given in Table 3.It was evident that the modified RAC had good workability, and yet the slump gradually decreased with the increase in NS content.

Macro-tests
For each concrete group, a series of cylindrical specimens (Φ100 × 50 mm) and cubes (100 × 100 × 100 mm 3 ) were cast and placed with vibration.Besides, all specimens were cured in a laboratory environment (20 ± 2°C, 95 ± 2% relative humidity) for 28 days.Moreover, cylindrical specimens were employed in RCPTs based on the American society for testing and materials C1202 standard [74], and vacuum saturation of the cylindrical specimens was required prior to carrying out RCPTs.Meanwhile, cubic specimens were used to measure the compressive strength and split tensile strength of concrete following GB/T 50081-2019 standard [75].As illustrated in Figure 3(a), the compressive and split tensile tests were conducted on the 2,000 kN servo-hydraulic compressional testing machine at a loading rate of 0.5 and 0.05 MPa/s, respectively, and the prefabricated mold employed for split tensile tests is demonstrated in Figure 3(b), which could ensure that cubic specimens are subjected to a pure split tensile loading condition [76].

Innovation of modeled interfaces
In order to ascertain the efficacy and mechanism by which the surface treatment method strengthens RAC, it is   Surface treatment with nano-silica and MKPC for enhancing RAC  5 necessary to characterize the microstructures and element distributions at the interface of adhesion between the surface treatment slurry and RA, as well as the new ITZ, across both RA maintenance schemes.Nevertheless, the complex characteristics of RA pose great difficulties for quantitative testing in actual concrete.For example, as revealed in Figure 4, the irregular shape of the aggregate in the direction of penetration will affect the accuracy of micro-tests including EDS.Additionally, with regard to statistical results, attributed to factors including internal leakage around the aggregate, similarly, the microscopic properties of substances surrounding the same aggregate are heterogeneous [77], complicating the acquisition of dependable mean properties while minimizing the testing workload [77,78].For these reasons, the interface model was employed to simulate the actual interfaces in the RAC referencing previous studies [64], attempting to avoid the variability of attached mortar of RAs [79] and consequently acquiring more reliable interface microstructure testing results within a low statistical effort range.Moreover, the ITZs were predominantly attributed to the wall effect [77], and accordingly, the simulated interfaces were not noticeably distinct from the ones in actual concrete.As depicted in Figure 5, the interface models were produced as follows: 1) A number of Φ100 × 10 mm slices were cut out from old concrete specimens; 2) Four MKPC + NS slurries with 1 mm thickness were painted on the surface of the old concrete slices, respectively; 3) A 9 mm thick cement mortar with a 0.2 water-to-cement ratio was poured on the MKPC + NS slurry at setting or 28 day pre-curing.Accordingly, the interface conditions could be created, which were comparable to those of the real RAC using the two RA maintenance schemes and DM.
As a point of comparison, an additional set of MD-0 models was cast as a reference group, excluding the surface treatment slurry, in which the water-to-cement ratio of the new mortar was adopted as 0.57 in consistent with RAC0.Besides, the designations of all models and their respective specific meanings are depicted in Table 4. Upon casting, all specimens were cured at 20 ± 2°C and 95 ± 2% relative humidity.

Micro-tests
SEM (TESCAN MIRA LMS, Czech Republic) was employed to apply backscattered electron (BSE) mode to interface models to  characterize the microstructures of the bonding interfaces between old concrete slices and NS + MKPC pastes, along with the new ITZ.Besides, elemental distributions in the aforementioned testing regions were investigated by using EDS (Oxford Xplore) equipped in SEM to qualitatively determine the components of the matrices and indicate the chemical interactions between the pastes.The SEM was operated at a 15 mm working distance and 15 kV accelerating voltage.For the purpose of quantitatively analyzing the reinforcement efficiency of NS + MKPC slurries on the new ITZ microstructure, 40 high-magnification BSE images (×500) covering the new ITZ captured by each type of model were used to conduct a statistical analysis of the span and porosity [80] of the new ITZ [81,82] according to the distinct gray values in the images.Sequential calculation regions were partitioned within each BSE image for this objective, and each region was a 10 μm wide strip; moreover, the total span of the calculation regions was 240 μm to cover the new ITZ, as depicted in Figure 6(a), for the specific distribution of the strips.Furthermore, the porosity inside of each strip was derived from the pore area percentage (which is approximately equal to the pore volume percentage [83]) in the gray binary image, and the "overflow method" was used to ascertain the pore gray threshold [84].On the condition that the boundary of the new ITZ has been determined, the final results were plotted on the basis of the relative distance away from the interface.Except for that, a graphical representation of the BSE image processing and analysis is demonstrated in Figure 6.Detailed descriptions of the method could be found elsewhere [85].
The BSE samples were fabricated by taking 10 mm × 10 mm × 20 mm slices from Φ100 × 20 mm interface model specimens, as demonstrated in Figure 5.Then, these slice samples were immersed in anhydrous ethanol to impede cement hydration and dried at 45°C.Successively, following the vacuum impregnation of low-viscosity epoxy resin onto the dried samples, the samples were polished after the epoxy-dried to produce smooth test surfaces.
The micro-morphology, as well as the components of the four hardened NS + MKPC pastes, was characterized and analyzed by using SEM and EDS to expand upon the impact of NS percentage on the operational characteristics of the NS + MKPC composite slurry and the modified RAC.Moreover, the test samples were developed from broken nuggets acquired after RAC strength testing, dehydrated, dried, as well as stored in a vacuum chamber.
The research framework, including methodology, materials used, preparation processes of test specimens, and corresponding measurement equipment in this study, is demonstrated in Figure 7.  Additionally, it was observed that the mechanical strengths of the RAC exhibited a consistent increase as the dosage of NS rose from 1 to 3%, but subsequently decreased as the dosage increased to 4%.Furthermore, the maximum strength improvement was observed at 3% NS dosing in the case of RA no pre-curing, which raised 40.75 and 46.26% in compressive and split tensile strengths, respectively, in comparison with RAC0.

Chloride ion penetration resistance
Figure 9 illustrates the average cumulative electrical flux passed through each set of six concrete samples during 6 h in RCPTs, implying the chloride ion penetration resistance of the concrete.On the condition that using R0 as a substitute for N0, attenuation remained observed regarding resistance to chloride ion permeability.Moreover, the tendency of chloride ion permeation resistance was consistent with the mechanical strengths.The cumulative flux of RAC decreased by 31.57,38.50, 44.75, and 37.35%, respectively, as NS concentration increased, in comparison with RAC0 under the RA pre-curing regimen, and demonstrated further reductions under the RA no pre-curing regimen.Furthermore, the permeability of each concrete group was qualitatively ranked as specified in the American society for testing and materials C1202 standard.Untreated RAC0 is classified as "moderate," whereas all modified RAC are classified as "very low" or "low." The elevated resistance to chloride ion penetration suggests that NS + MKPC modification for RAC enhanced its durability.
Through micro-characterization analyses of the modeled interfaces and the NS + MKPC pastes, the strengthening mechanism of various NS + MKPC slurries for RAC macroproperties under either of the RA maintenance schemes will be interpreted in detail in the following sections.

Determination of the optimal NS dosage and RA maintenance scheme
The optimal NS dosage and RA maintenance scheme concerning the distinct mechanical properties as well as chloride ion penetration resistance efficiency (η) were determined by employing the linear weighted sum optimization method [65].η of the various modified RAC specimens were described as the percentage of the electrical flux decline in comparison with RAC0.Additionally, the derived weights displayed in Table 5 were used to multiply the average mechanical properties and η listed in Table 6 using equation ( 2).The linear weighted sum optimization method used the statistical P-value of each modified RAC compared to the control (RAC0).Moreover, the advantages of this approach are the elimination of bias introduction into the computation, simpler implementation, comprehensive evaluation of mechanical properties and durability, and the ranking of final results to ascertain the optimization result.
where W i is the weighted value for each type of modified RAC properties, and P i for each type of modified RAC = Pvalue obtained on the condition that each property of modified RAC was referenced to the control value for each respective property.All the experimental data sets were in comparison with the control data set in pairs to determine the respective P-value one following the other at a 5%  threshold (α) value.The final weighted outcome and ranking for all modified RAC tested are indicated in Table 7.
From the optimization result and ranking demonstrated in Table 7, C-3-N and C-3-P possess the maximum performance, ranked number one and five, respectively, under the two RA maintenance schemes.Moreover, C-1-N, C-2-N, C-3-N, and C-4-N without pre-curing recorded higher ranking in comparison with C-1-P, C-2-P, C-3-P, and C-4-P.Based on the mechanical properties and durability of the different modified RAC, this result indicates that the optimal NS dosage and RA maintenance scheme are 3% and no pre-curing scheme, respectively.

Micro-properties of bonding interfaces
To validate the outstanding bonding performance of NS + MKPC slurry to RA, the typical BSE-EDS test results at a high magnification of a total of four bonding interfaces in models are illustrated in Figure 10.Besides, the bonding interfaces can be identified based on the unique micro-  morphologies of the two matrices: the interfaces' left sides consisted of old concrete matrices, whereas the right sides consisted of various NS + MKPC matrices.From the BSE photographs, certain micro-cracks could be observed in part of the NS + MKPC matrices resulting from polishing throughout the process of sample fabrication, as reported in several works of literature [61,86,87].
From the figures, it can be determined that all the bonding interface regions were dense, with no apparent micro-cracks or large pores, i.e., the NS + MKPC slurry was firmly bonded to the old concrete.Moreover, BSE-EDS analysis reveals the following two factors as the chief reasons for this: (1) the presence of surface defects on the old concrete, which symbolized RA, provided the NS + MKPC slurries with a strong mechanical anchorage with it, and (2) there was a favorable chemical bonding between the two matrices predicated on the overlap of P and Ca elements manifesting in the vicinity of the interface, i.e., the soluble phosphate in the NS + MKPC slurry penetrated through the micro-defects of the old concrete and reacted with the alkali components, including CH, to produce cementitious substances, which is consistent with findings reported in the prior literature [62].Furthermore, the NS particles with small particle size and high activity in the slurry would likewise infiltrate into the cracked or loose layers of the old mortar and bonding interface; by functioning as a filler and undergoing a reaction with CH to produce C-S-H gels, it enhanced the chemical bonding at the interface.On the basis of the elemental graphs, a partial overlap of Mg, O, and Si elements are likewise observed, associated with forming the M-S-H gel, further enhancing the nano-slurry's cementitious properties.Additional findings associated with this will be covered in detail in Section 3.4.

Micro-properties of the new ITZ and NS + MKPC pastes
Typical BSE photographs and corresponding elemental distribution maps of the interface regions between the NS + MKPC matrices and new mortars in models MD-2-P and MD-2-N are illustrated in Figure 12 to analyze the reinforcing efficiency and mechanism of NS + MKPC slurries on the new ITZ under the two RA maintenance schemes.For comparison purposes, Figure 11 presents a typical BSE image of the new ITZ in an MD-0 sample.It is evident from the image that numerous visible pores and extensive cracks (with a maximum width of approximately 78 μm) are presented within the new ITZ.These features result from the elevated water content and a significant abundance of CH crystals, thereby adversely impacting the macro-properties of RAC0.Nevertheless, no large-sized pores were found within the new ITZ obtained by both RA maintenance schemes following modification by S-2 slurry, and microcracks were all substantially reduced, while only a few tiny micro-cracks (with a width of approximately 3-11 μm) were observed under the RA pre-curing scheme.As a result, the formation of a denser microstructure in the new ITZ explains the significant enhancement observed in the mechanical strengths and durability of RAC following slurry strengthening.
Examining the elemental distributions in Figure 12 reveals significant overlaps of Ca and P elements in the NS + MKPC slurries near the interface.Additionally, there are limited overlaps of Ca, Mg, and P elements, as well as Ca and Si elements.These observations suggest that Ca ions in the new mortar permeated the NS + MKPC slurries, engaging in reactions with k-struvite, unhydrated phosphate, or NS particles, resulting in the generation of novel gelling substances.It can likewise be noted that the NS + MKPC slurries near the interface increased in compactness due to the penetration of Ca ions and the associated chemical interactions.Upon examination, it was evident that the penetration depth of Ca ions was more extensive in the absence of pre-curing.The same trend was observed under the no pre-curing condition for NS dosages of 1, 3, or 4%, as illustrated in Figure 13, by typical BSE and EDS images of the new ITZ.It can be recognized that the new ITZ modified by NS + MKPC slurries had superior mechanical bonding as well as chemical anchoring, leading to a substantial reduction of defects including de-bonding cracks, while under the no pre-curing method, enhanced chemical bonding resulted in a further reduction of micro-cracks in line with the BSE observations.Additionally, NS particles deposited on the strengthening slurry surface may disintegrate and permeate  the new ITZ surrounding it, thus inducing a micro-filling effect and exerting a volcanic ash effect at the same location, which reduced the quantity and size of CH crystals in the new ITZ and generated extra C-S-H gels, contributing to a denser new ITZ.The width and porosity of the new ITZ acquired from the surface treatment with the four slurries under two maintenance methods were rigorously quantified to substantiate this viewpoint.
Based on ITZ porosity distribution characteristics, the porosity distribution graphs of the test regions were accomplished by identifying the location where relatively high porosity occurred as the initiating boundary of the new ITZ and setting the origin of the coordinate there.Besides, Figure 14 illustrates the porosity distribution profiles for all modeled new ITZs and their vicinities: positive and negative regions along the x-axis, respectively, represent the NS + MKPC matrix and the new mortar.The figure allows for the determination of the width of the new ITZ, defined as the horizontal distance from the origin of the coordinate to the point of porosity stabilization.This enables the statistical averaging of porosity within the new ITZ, as illustrated in Figure 15.From Figures 14 and 15, it is obvious that the new ITZ in all enhanced treated samples had narrower widths and lower porosities in comparison with MD-0, with the width and porosity progressively diminishing as the NS dosage increased.Additional reductions of widths and porosities achieved through the use of the RA no precuring method occurred at approximately 9.52-17.86and 26.72-42.25%for the four NS doping levels, respectively, as in comparison with the RA pre-curing approach.The chief reason for this is that the RA no pre-curing method facilitated the penetration of NS particles into the new ITZ due to the hydration reaction in the NS + MKPC slurry, which was relatively incomplete.It is evident that with the rise in the total amount of NS infiltration caused by the increased NS doping or easier infiltration, the microstructure of the new ITZ was further enhanced, which demonstrates evidence for the penetration of NS particles in the new ITZ.Additionally, it can likewise be seen in the figure that the NS + MKPC slurry in the vicinity of the new ITZ under the RA no pre-curing scheme had a lower porosity within a certain range with little variation in width at different NS doping levels.Besides, the region had a compact Surface treatment with nano-silica and MKPC for enhancing RAC  13 microstructure owing to the cementitious phase produced by Ca ions and NS + MKPC, thus presenting additional evidence of a robust chemical anchorage between the new mortar and the NS + MKPC slurry.Synthesizing, outstanding chemical bonding, and NS penetration contributed significantly to the formation of a minute proportion of pores and fissures in the new ITZ that contributed to yielding strengthened mechanical and durability properties for modified RAC.
The doping ratio of NS, in addition to influencing the microstructural properties of the new ITZ, would likewise significantly influence the micro-properties of NS + MKPC composite slurries [88] and, thus, the modification efficiency of the slurries for RAC. Figure 16 displays SEM-EDS images of NS + MKPC hardened pastes with varying dosages of NS at a magnification of 5,000.As the NS concentration increased from 1 to 3%, the microstructure of the paste became more compact, as illustrated in the figure.Besides the pore-filling effect of small spherical NS particles in the NS + MKPC paste, the chemical interaction between NS and MKPC generated new gelling substances, which progressively increased in proportion to the NS dosage.Upon analyzing Mg and Si elemental distribution maps and EDS results at the marked point of S-3, it can be identified that the gel was primarily composed of Mg, O, and Si elements and consequently regarded as a newly formed M-S-H phase.Furthermore, in the NS + MKPC system, the M-S-H phase could tightly connect the paste matrix, as well as encapsulate and cement the unreacted MgO better [89], contributing to the NS + MKPC matrix microstructure densification and a further decline in porosity, which contributed benefits to the enhancements of strengths and chloride ion penetration resistance of modified RAC.Nevertheless, on the condition that the NS content increased to 4%, the microstructure of the paste became looser and more porous due to the decreased fluidity of the slurry for the excessive incorporation of nanoparticles.Another reason for the deterioration of microstructure is that the amount of k-struvite playing an important bonding role in the matrix was reduced, as a result of decreased MgO available for reacting with KH 2 PO 4 due to the excessive chemical reaction between NS and MgO.In general, the microstructure of the new ITZ with 4% NS doping is denser, yet not markedly different from that with 3% NS.Conversely, the microstructure of S-4 is notably inferior to that of S-3.Therefore, the optimal level of NS incorporation was determined to be 3%, aligning with the test results of the macro-properties of RAC in the preceding section.
In light of the microstructural characterization of modified RAC and interface models, it can be hypothesized that the reason for the enhancement of modified RAC's macroproperties could be the superior bonding properties of NS + MKPC to RA, the new ITZ, and NS + MKPC slurry strengthening.
Overall, surface treatment with NS + MKPC enhanced the compressive strength, split tensile strength, and chloride ion penetration resistance of RAC.Nevertheless, the influence of the modification method on the long-term properties of RAC is a primary concern for RAC's application in structural use [90][91][92][93][94][95][96] and possesses a lack of clarity.On this basis, it is recommended to develop studies to monitor the modified RAC's long-term properties, as well as resistance to the freezing-thawing cycles, carbonation, and sulfate attacks in the future, to furnish crucial guidance for the application of the modified RAC in structures subjected to certain environmental impacts.Likewise, the effect of the surface treatment on particular aspects of RAC's microstructure has also been examined in this study.Nevertheless, the microhardness variation, which is an essential characterization of the microstructure [64], has not been surveyed.Accordingly, additional research must be undertaken to investigate the microhardness of modified RAC and compare the findings with those of previously published studies in the field.Notably, this study likewise introduced an interface model for microanalysis of interface regions, contributing to enhancing testing accuracy within a low statistical effort range and avoiding the variability of attached mortar of RAs.Yet, this interface model was developed by using old concrete slices obtained by cutting to simulate RA acquired by crushing and was tested under the assumption of no difference between them.Apparently, there are certain differences between concrete cutting and crushing, which can lead to variations in the micro-properties of the bonding interface between NS + MKPC paste and RA, thus affecting RAC's macro-properties.Likewise, note that straight plain areas not bonded/surrounded from all sides were used on the condition that simulating interfaces, which is distinct from those of the RAC.In future research endeavors, it is recommended to conduct experimental investigations into the variations in porosity, roughness, and other properties of RA surfaces obtained through concrete cutting and crushing.Additionally, the impact of these differences, coupled with the interfacial shape, on the interfacial microstructure should be explored.The aim is to develop a model that facilitates convenient statistical analysis and provides a more accurate representation of the actual scenario.Furthermore, to further prove the financial viability of the modified RAC, enhance its acceptability in the market, and provide a reference for the formulation of relevant policies, it is recommended to quantify the costs and environmental impacts of the aggregates' treatment and subsequently use that information to conduct a simplified analysis of life cycle costs and life cycle assessment.

Conclusions
This study evaluated the feasibility of using NS + MKPC in RAC modification and indicated its function mechanism, controlled dose of NS, and a more effective surface treatment method, and the main conclusions are as follows: 1) RAC's compressive strength, splitting tensile strength, and chloride ion penetration resistance were enhanced by RA surface treatment with any NS + MKPC slurries and RA maintenance schemes used in this study.Besides, the optimum dosage of NS in NS + MKPC slurries for improving RAC macro-performances was determined to be 3.0% by the linear weighted sum optimization method.
In relation to the pre-curing scheme for surface-treated RA, the no pre-curing approach demonstrated a more pronounced enhancement.2) Based on the micro-test results of the interface models, it was determined that the superior bonding interfaces between NS + MKPC slurries and old concrete slices (representing RAs) were responsible for the enhanced macroproperties of the modified RAC.Moreover, another crucial reason is that the new ITZ in interface models containing NS + MKPC slurries was substantially enhanced, with noticeable reductions in width and porosity.Furthermore, the enhanced bonding of the NS + MKPC slurry to new mortar in combination with the micro-filling and volcanic Surface treatment with nano-silica and MKPC for enhancing RAC  15 ash effects of NS particles on the new ITZ may be the dominant reason for that result.3) For the optimal NS dosage and RA maintenance scheme, the superiority toward enhancing RAC macro-properties was predominantly attributed to the denser new ITZ, as evidenced by the distinctive decrease in porosity.
Additionally, the advantages of the NS + MKPC slurry doped with 3% NS were ascribed to the superior microstructure of the slurry.
In particular, the optimal NS dosage of 0.614 wt% of the aggregate in this study was less than the recommended dosages of 1 and 2 wt% of the cement [97,98] for NS and NC [76], and the 3% optimal dosage for NS suspension with a SiO 2 content of 34.3%, whereas the optimum compressive strength improvement efficiency of 40.75% is higher than that of 13.9% [28].Hence, the surface treatment proposed in this study decreased the cost of modification and enhanced the reinforcement efficiency of RAC, thus raising the economic feasibility of RAC application in practice.The low-cost magnesium phosphate cement fabricated from industrial by-product boron muds [99] may be contemplated in the future as a means to further reduce modification expenses.Besides, the proposed RA no pre-curing scenario was more convenient for on-site construction, which simplified the RA surface treatment process and shortened the duration of construction, thus enhancing the operability of RA application in practice.Notably, the additional cement in RAC as one method to enhance its compression strength and durability in considerable studies [100] was avoided precisely owing to the high modification efficiency of this surface treatment method.This resulted in a reduction of cement used in RAC, the production of which emits large amounts of carbon dioxide [101] and is responsible for 5-7% of total global warming impacts [102].Consequently, the modification of RAC using NS + MKPC is conforming to the policy goals of peaking carbon dioxide emissions and achieving carbon neutrality.In general, the surface treatment with NS and MKPC co-action demonstrates a promising potential for the extensive utilization of RAC in practical engineering applications.This approach also presents prospects for fostering sustainable development within the construction industry, offering environmental advantages.
-N MD-0 interface model simulating the interfaces in the RAC0 MD-1-P interface model simulating the interfaces in the C-1-P MD-1-N interface model simulating the interfaces in the C-1-N MD-2-P interface model simulating the interfaces in the C-2-P MD-2-N interface model simulating the interfaces in the C-2-N MD-3-P interface model simulating the interfaces in the C-3-P MD-3-N interface model simulating the interfaces in the C-3-N MD-4-P interface model simulating the interfaces in the C-4-P MD-4-N interface model simulating the interfaces in the C-4-N 1 Introduction

Figure 3 :
Figure 3: (a) Compressive and split tensile test device and (b) the mold for split tensile strength tests.

Figure 5 :
Figure 5: Procedure for the preparation of interface model samples.

Figure 6 : 7 3Figure 8
Figure 6: BSE image processing and analysis for porosity statistic: (a) specific distribution of the strips and (b) BSE image binarization process.

Figure 9 :
Figure 9: Average cumulative electrical fluxes of various concrete types.

Figure 14 :Figure 15 :
Figure 14: Porosity distributions in the test regions based on the distance to the start boundaries of the new ITZ.

Table 1 :
Physical properties and crush values of N0 and R0

Table 2 :
Blending ratios for pre-treatment of 1.000 kg R0

Table 3 :
Types and slumps of concrete

Table 4 :
Interface model types

Table 5 :
Weighted values for mechanical properties and η of each type of modified RAC

Table 6 :
Average values of mechanical properties and η

Table 7 :
Optimization result and ranking