Solidification Effect and Mechanism of Marine Muck Treated with Ionic Soil Stabilizer and Cement

Abstract

In this study, an environmentally friendly ionic soil stabilizer (ISS) was adopted with combination of Portland cement to stabilize a marine muck. The macro and micro tests results demonstrated that the ISS was an effective stabilizer to improve the strength of marine muck when it was used combined with cement after adding the alkalizer NaOH. Except for the reduction in interlayer distance of clay minerals by ISS, Ca${}^{2+}$ and SO${}_4^{2-}$ dissolved from ISS promoted the production of ettringite (AFt), pozzolanic and carbonation reactions of Portland cement in the presence of NaOH. Meanwhile, the hydration products of curing reaction notably agglomerated soil particles, which caused an obvious decrease of pores and a high increase of strength for solidified soils. Furthermore, this combination of stabilizers can not only save the dosage of cement, but also accelerate the solidification speed, decrease the cement setting time within 7 days to meet the curing requirements, and enhance the strength of solidified soils.

1. Introduction

Marine soft soil deposits are widely distributed in the southeast of China, where the economy is highly developed and requires a large number of roads, railways, buildings, etc. However, because of the properties of high moisture content, high porosity and low strength in natural state of soft soil, it cannot be used directly as subgrade or foundation and is difficult to be recycled as building materials [1,2,3]. Adding suitable binders into soft soil is considered to be a very useful way to solve the above significant engineering problems and has been widely used [4,5,6,7].

Portland cement is one of the most widely used cementitious material, which can provide efficient soil solidification effect with high durability [8,9,10]. However, the Portland cement production process brings serious environmental problems [11,12]. Therefore, finding sustainable green cementitious materials to partially or completely replace Portland cement has become an appealing goal in the field of industry and environmental research. Currently, many industrial wastes and by-products, such as fly ash [13], lime [14], bagasse ash [15], recycled-glass [16], silica fume [17], and metakaolin [18], are added to Portland cement as mixtures to save resources and improve the strength of solidified soil.

Recently, ionic soil stabilizer (ISS), a by-product of citrus industry, has attracted much more attentions as a kind of soil curing agent [19,20,21]. It has been efficiently utilized alone for some kinds of soils, such as frozen soil [22], red clay [23], expansive soil [24,25], soft clay [26], etc. The theoretical study indicates that the ISS can change the double electric layer structure of clay and expulse various types of water from soil through an irreversible process [27], which has been supported by experimental studies [19,26,28,29]. Usually, cement mixture solidification method, which is considered in accord with construction standards, is still widely utilized in practice in the field of engineering [30,31,32,33]. However, few studies had been done on the efficiency and micro-mechanism of combined utilization of ISS and Portland cement for soil solidification [34,35]. Therefore, with the purpose of using ISS to replace partly Portland cement, meanwhile, aiming at the reduction of defects of soil solidification with only Portland cement (such as long cement setting time, low degree of hydration, high cement consumption for high-water-content soft soil), the study of combined utilization of ISS and Portland cement for soft soil stabilization is a challenging problem and worth to be studied further.

Taking the Ningde high-water-content marine muck as an example, our research was focus on the combined solidification effect of ISS-cement on marine muck and the ISS-cement-muck stabilization mechanism by using different macro and micro tests, including Atterberg limits, water content, pH, unconfined compression test (UCT), direct shear test (DST), scanning electron microscope (SEM), X-ray diffraction (XRD), and X-ray fluorescence (XRF).

2. Materials and Methods

2.1. Materials

Materials used in this study are assembled in Figure 1. The marine soft clay was dredged in Ningde city of Fujian province near the East China Sea as shown in Figure 1a,b. ISS, Portland cement (Pc), and sodium hydroxide (NaOH) used in this study as chemical additives for soil solidification.

ISS is a black liquid at normal temperature as shown in Figure 1c. It is an acyloxy compound with more than 1% content of sulfuric acid, 1 mg/cm${}^3$ content of sulphur and 1.05 pH-value. ISS is completely soluble in water as described in ISS specification issued by the manufacturer. The active chemical component of ISS is sulfonated oil. The ion concentration in ISS was tested by inductively coupled plasma optical emission spectrometer (ICP-OES, Thermo iCAP 6300, Thermo Fisher Scientific, Waltham, MA, USA) and ion chromatography system (ICS). The main cations and anions in ISS are presented in

Table 1. Concentration of main cations and anions in ISS.

Ion	Concentration (mg/L)
Fe${}^{3+}$	2388
Al${}^{3+}$	807
Ca${}^{2+}$	10,662
Mg${}^{2+}$	1433
K${}^{+}$	11,300
Na${}^{+}$	16,128
NO${}_3^{-}$	1920
SO${}_4^{2-}$	100,344

. Due to the high ion concentration, strong oxidation and dissolution ability as reported in ISS specification, ISS should be diluted with water before utilization. Another curing agent used in this study was traditional Portland cement of 32.5R grade (Pc32.5R) [36]. The main chemical components of the Portland cement were: 53.76% CaO, 22.72% SiO${}_2$, 6.70% Al${}_2$O${}_3$, 3.40% Fe${}_2$O${}_3$, 2.72% MgO, 2.63% SO${}_3$, 0.65% K${}_2$O, 0.14% Na${}_2$O and others. NaOH is a chemically pure (i.e., analytical reagent). It was a kind of alkalizer used in this study to adjust the pH-value of curing environment. Furthermore, in this study, deionized water (pH = 6.7, electrical conductivity <1 µS/m) was used to avoid the effects of other chemical components, which may affect the test results.

The basic physical and mechanical parameters of the natural Ningde marine muck are summarized in

Table 2. Basic physical and mechanical parameters of natural Ningde marine muck (UCS: unconfined compressive strength).

Parameter	Value
Natural water content (%)	65.02
Natural density (g/cm${}^3$)	1.38
Soil particle density (g/cm${}^3$)	2.73
Initial void ratio	2.26
Liquid limit, LL (%)	50.77
Plastic limit, PL (%)	31.67
Plasticity index, PI (%)	19.10
pH	5.6
UCS (kPa)	46

. Weakly acidic (pH = 5.6) Ningde marine muck had a high initial void ratio of 2.26 and a high natural water content of 65.02%, which exceeded its liquid limit (LL = 50.77%). Moreover, the unconfined compressive strength (UCS = 46 kPa) of this soil was very low, therefore, it was a kind of soil with extremely poor engineering properties as ground, subgrade or resource reuse materials. The grain size distribution curve of natural Ningde marine muck is presented in Figure 2. The soil showed relatively high clay fraction contents of 24% (particle size < 0.002 mm) and 37% (particle size < 0.005 mm) respectively. According to the Unified Soil Classification System (USCS) presented in ASTM D2487 [37], the Ningde marine muck can be classified as an elastic silt (MH).

2.2. Sample Preparation

The natural Ningde marine muck (abbreviated to M as described in

Table 3. Symbols and compositions of different types of samples (ISS:H${}_2$O ratio = 1:x, where x = 25, 50, 75, 100, 125, 150, 175, 200; Cement dosage (t%): the ratio of dry weight of Portland cement to dried soil powders; LL: liquid limit; A: the amount of NaOH added until pH-value of the sample reached neutral before adding cement).

Sample Symbol	Soil	ISS:H${}_2$O Ratio	Cement Dosage (t%)	NaOH	Water Content (%)
M	Natural	0	0	0	Natural content
MIx	Powder	1:x	0	0	LL
MI	Powder	Optimal ratio	0	0	LL
MPct	Powder	0	0, 3, 6, 9, 12, 15, 18	0	LL
MIPct	Powder	Optimal ratio	0, 3, 6, 9, 12, 15, 18	0	LL
MIAPct	Powder	Optimal ratio	0, 3, 6, 9, 12, 15, 18	A	LL

) was dried in an electro-thermostatic blast oven for 48 h with a temperature of 55 ${}^{°}$C, then it was milled and sieved through a 2-mm sieve to get dried soil powders in preparation for the subsequent different types of solidified samples.

Meanwhile, ISS should be diluted with water before utilization due to its high concentration [26]. Firstly, ISS concentrate was diluted by deionized water at the volume ratio (ISS:water = 1:x) as 0 (without ISS), 1:200, 1:175, 1:150, 1:125, 1:100, 1:75, 1:50 and 1:25. Then, the mass of dried soil powders and the volume of ISS diluents were measured in order to obtain samples with initial water content equal to liquid limit (LL = 50.77%), so as to simulate the high-water-content condition in natural state. Next, each ISS diluent with different dilution ratio was sprayed by a sprinkling can onto dried soil powders and mixed thoroughly in an electric blender for 10 min until a homogeneous ISS-solidified sample (abbreviated to MIx as described in Table 3) was obtained [38]. Afterwards, samples with different sizes and shapes were prepared by static pressure method on the basis of the experimental design. Finally, samples were wrapped in polyethylene plastic film and maintained in the standard curing box (temperature = 20 ± 2 ${}^{°}$C, relative humidity = 95%) for different periods according to the test requirements. The optimal volume ratio of ISS:water and the optimal ISS-solidified sample (abbreviated to MI as described in Table 3) were obtained through physical and mechanical tests.

The ISS-cement-solidified soil samples (abbreviated to MIPct as described in Table 3) were prepared by adding Portland cement to sample MI.The samples MIPct with different Portland cement dosage (the ratio of dry weight of Portland cement to dried soil powders in percentage) t% (t = 0, 3, 6, 9, 12, 15 and 18) were prepared. For uniformity, the Portland cement was also sieved to pass through a 2-mm sieve. Then, each mixture was stirred in an electric blender for another 10 min to achieve homogeneity [38]. The subsequent sample preparation method and curing condition were the same as sample MIx, but the curing period was 7 days for sample MIPct.

Due to that ISS and cement had the opposite acidity and alkalinity, which may affect the curing effect with each other in combined utilization for soil solidification [35,39], NaOH was selected in this study as the alkalizer to regulate the pH environment of cement hydration reaction. Thus, to prepare the ISS-NaOH-cement-solidified soil samples (abbreviated to MIAPct as described in Table 3), the alkalizer NaOH was added to the sample MI until it reached neutral. Then, the Portland cement dosage of t% was added to the neutral mixture. The following sample MIAPct preparation and curing methods were the same as sample MIPct.

For comparison, the control group of cement-solidified soil samples (abbreviated to MPct as described in Table 3) were also prepared under the same Portland cement dosage t%. The preparation and curing methods of sample MPct were the same as sample MIPct, except that ISS was replaced with deionized water in the sample preparation process.

2.3. Experimental Methods

For macroscopic analysis of solidification effect on different types of solidified samples, Atterberg limits test, water content test, pH test, UCT and DST were carried out.

Atterberg limits tests were carried out for sample MIx after 48 h of sample curing in the standard curing box [40]. Atterberg limits of the triplicate specimens of each sample MIx were measured according to ASTM D4318 [41]. Water content tests, pH tests and DST were adopted on samples MI, MPct, MIPct and MIAPct followed relevant standards [42]. For DST, in accordance with the above sample preparation methods for different types of solidified samples, each predetermined and prepared mixture was carefully poured into a standard oedometer ring (61.8 mm in diameter and 20 mm in height) with a palette knife in layered way. Then, the solidified samples were made by the static pressure method and maintained for 7 days before testing. After a 7-day cure, DST with the shearing rate at 0.8 mm/min was carried out on four parallel specimens of each solidified sample under the vertical pressure of 100, 200, 300, and 400 kPa, respectively. The equipment used in DST was ZJ strain-controlled direct shear instrument. Besides, UCT was performed on all types of solidified samples, including MIx, MI, MPct, MIPct and MIAPct. For UCT, standard cylindrical samples (39.1 mm in diameter and 80 mm in height) were prepared in the same curing condition followed the same sample preparation method as for DST. UCT (axial strain rate = 1–3%/min) was carried out by DW-1 electrodynamic strain-controlled unconfined pressure gauge [42].

For the purpose of evaluating the solidification effect from a microscopic point of view and understand the solidification mechanism, SEM, XRD, XRF were performed on all types of samples.

For SEM, specimens (about 1 cm${}^3$ in volume) were retrieved from the macroscopic test samples as described above. Then, the liquid nitrogen freeze-drying technique was employed to dehydrate the solidified samples [43,44]. After that, SEM samples were gold-plated. Finally, the surface morphology, pore structures and mineral components were observed by SEM with Hitachi Uhr FE-SEM SU 8010 machine (Hitachi Ltd., Tokyo, Japan). For XRD, freeze-dried samples were milled and sieved through a 0.075-mm sieve. Then, XRD (PANalytical X’ Pert Pro DY 2198 diffractometer, PANalytical B.V., Amsterdam, The Netherlands) was carried out for both samples prepared by powder pressed method (nondirectional sample) and on oriented sheet (directional sample) in a 2$\theta$ range of 3–65${}^{°}$ and 3–35${}^{°}$, respectively. In addition, XRF analysis was performed by the PANalytical Axios mAX X-ray fluorescence spectroscopy on the samples prepared by powder compressing machine (PANalytical B.V., Amsterdam, The Netherlands).

For clarity, experimental procedure for all types of samples is summarized in Figure 3.

3. Results and Analysis

3.1. Macroscopic Analysis for Solidification Effect

The variation curves of PI and UCS at different ISS:water volume ratio for sample MIx are shown in Figure 4. Both PI and UCS data had extrema, which means that there was an optimal value for soil solidification. To further verify this result, two dilution ratios of 1:40 and 1:60 for ISS:water were selected near the optimal dilution ratio of 1:50 to retest, and still the same result was gotten. This result was consistent with that reported in literature for various soils solidified by ISS [39,45,46,47]. The experimental results showed that the optimal ratio (ISS:water) for Ningde soil was 1:50, which corresponded to the minimum PI (13.24) and the maximum UCS (65 kPa) for optimal ISS-solidified soil sample MI. Compared to natural Ningde soil sample M, PI of sample MI reduced by 30.7% and its UCS increased by 41.3%. Our results indicated that the soil mechanical characteristics are closely related to the plasticity index, which has been confirmed by many previous studies [48], for example, plasticity index can be used as a reliable index to evaluate shear modulus decay characterisitics of different soils [49].

The results of water content tests, pH tests and UCT are summarized in Figure 5. The variation curves of water content and pH-value at different Portland cement dosage of different samples are shown in Figure 5a. The water content obviously decreased with the increasing of Portland cement dosage for all types of cement solidified samples (MPct, MIPct and MIAPct). Under the same Portland cement dosage, the water content of different samples ranged from high to low as: MPct > MIPct > MIAPct, which indicated that ISS, when used in conjunction with Portland cement, can reduce the water content of cement solidified soils, regardless of whether the NaOH was added to the sample. Nevertheless, the addition of NaOH further significantly reduced the water content of the solidified soil as shown in Figure 5a with solid square. ISS can expulse various types of water (such as interlayer, bound and free water) from soils through an irreversible process as demonstrated in the previous study for silt solidified with single ISS [26]. In contrast, the pH-value firstly significantly increased with the increasing of Portland cement dosage from 0 to 6%, and then increased slowly with the increasing of Portland cement dosage from 6% to 18% for all types of samples (MPct, MIPct and MIAPct). Under the same Portland cement dosage, the pH-value of different samples ranged from high to low as: MIAPct > MPct > MIPct, which indicated that ISS, when used in conjunction with Portland cement without adding NaOH, would decrease the pH-value of solidified soils compared to the natural Ningde soil sample and sample MPct as shown in Figure 5a with hollow triangle.

Figure 5b shows the same influence of ISS used in conjunction with Portland cement from the point of view of the strength changes of cement solidified samples. The UCS of different types of cement solidified samples under the same Portland cement dosage ranged from more to less as: MIAPct > MPct > MIPct. In addition, with the increasing of Portland cement dosage, the UCS of samples MPct and MIAPct increased followed the power function as Equation (1) for sample MPct and Equation (2) for sample MIAPct:

$$y=0.072x^{1.765}$$

(1)

$$y=0.206x^{1.424}$$

(2)

where y is UCS expressed in MPa, and x is Portland cement dosage expressed in %. The changes of strength development were consistent with previous study [50] for cement stabilized dredged sediment. However, for samples MIPct as shown in Figure 5b with hollow triangle, it did not follow the power function, but showed the sudden change in UCS occurred when the Portland cement dosage of MIPct was 9% as indicated in Figure 5b with a black arrow. This phenomenon may be due to the neutralization reaction between alkaline cement and acidic ISS, that is, with the increase of Portland cement dosage, there will be excess cement to solidify the soil and increase the UCS of samples MIPct. Consequently, combined using of ISS and Portland cement without adding NaOH will result in reducing the effectiveness of cement-soil solidification.

Figure 5c shows an unified power function relationship between UCS and pH-value for all types of cement solidified samples. The generalized strength development of cement solidified soils with the increasing of pH-value can be expressed in the following form:

$$y=5.027\times {10}^{-13}·x^{12.509}$$

(3)

where y is UCS expressed in MPa, and x is pH-value. The result indicated that, the pH-value had a great influence on the cementitious effect rather than the water content. This conclusion and power function relationship agreed with the previous findings for cement and fly ash solidified silt [51] and cement stabilized kaolin [52].

Figure 5d shows the typical stress-strain curves. A decreasing trend of failure strain (${\epsilon }_{\mathrm{f}}$) with the increase of UCS can be obviously found in Figure 5f for all types of cement solidified soils. Such a relationship can be simulated by the unified natural logarithmic function as Equation (4), and this kind of correlation between ${\epsilon }_{\mathrm{f}}$ and UCS had also been reported in previous literature for solidified soils with MgO [53]. The UCS values of samples MIAPct suggested that the efficiency of strength improvement was significantly enhanced by the combined use of ISS and NaOH for the single cement solidified soils.

$$y=-1.723ln\left(x\right)+5.483$$

(4)

where y is failure strain (${\epsilon }_{\mathrm{f}}$) expressed in %, and x is UCS expressed in MPa.

The data obtained from the DST after calculation and analysis are presented in Figure 6. Figure 6a shows the typical shear stress-displacement curves, under the normal stress of 100, 200, 300, and 400 kPa, from which the parameter shear strength (${\tau }_{\mathrm{f}}$) can be evacuated. The ZJ strain-controlled direct shear instrument and the failure modes of some samples after DST are presented in Figure 6b. As shear displacement increased, the shear stress increased first and then decreased, and the peak strength was called as the shear strength (${\tau }_{\mathrm{f}}$). A good linear relationship between ${\tau }_{\mathrm{f}}$ and normal stress can be found as shown in Figure 6c. Furthermore, the cohesion (c) and the internal friction angle ($\phi$) were calculated from fitting curves and presented in Figure 6d. Similar phenomenon can be also discovered as described above, that was, under the same Portland cement dosage, c and $\phi$ of different types of cement solidified samples ranged from more to less as: MIAPct > MPct > MIPct.

Compared to the UCS results, an unified linear function relationship between c and UCS for all types of cement solidified samples can be obtained as expressed in Equation (5) shown in Figure 6e, which can be used for strength prediction of other similar cured samples with different Portland cement dosage.

$$y=17.226x+2.347$$

(5)

where y is cohesion (c) expressed in kPa, and x is UCS expressed in MPa.

The above macroscopic observations suggested that, adding ISS alone can reduce the water content and improve the strength of natural soils. On the contrary, adding ISS and cement in combination for soil solidification will reduce the strength of solidified soils, due to the opposite pH-value between the two curing agents, which was not conducive to the hydration reaction of cement. However, the strength of solidified soils can be greatly increased by adding the alkalizer NaOH to adjust the pH-value of the reaction environment prior to the combined use of ISS and cement. For example, the UCS of sample MIAPc9 (4.574 MPa) was about 7.0 times MIPc9 (0.65 MPa) and 1.6 times MPc9 (2.845 MPa), and the cohesion of sample MIAPc9 (84.726 kPa) was about 4.1 times MIPc9 (20.531 kPa) and 1.5 times MPc9 (57.272 kPa), respectively. Comparing the strength of sample MIAPct with MPct and M, which indicated that, the combined use of ISS, NaOH and cement can not only save the amount of cement but also efficiently enhance the strength of solidified soils, for instance, UCS for sample MIAPc9 was 99.4 times higher than sample M.

3.2. Microscopic Analysis for Solidification Effect

To better understand the solidification effect from a microscopic point of view, SEM tests were conducted on different types of samples, meanwhile, SEM micrographs were applied to evaluate the micro-structure change qualitatively.

Figure 7 shows the surface morphology and pore structures of different samples at SEM micrographs 2000×. Compared to the solidified samples MPc9, MIPc9 and MIAPc9, the clay minerals in sample M were loosely packed together with large pores between them. For sample MIAPc9 as shown in Figure 7, pores significantly decreased, and larger aggregates were formed resulting in more stable structures, which may be attributed to the combined utilization of ISS, NaOH and cement for soil solidification.

Quantitative analyses in pore size and structure changes were performed by using Particles (Pores) and Cracks Analysis System (PCAS) [54,55]. Figure 8 shows the pore vector images from PCAS corresponding to the SEM micrographs (2000×) of different samples. The physical true values converted from image processing data [56] are listed in

Table 4. Quantitative pore characteristics of different samples (Smax: maximum pore area; S: total pore area).

Sample Symbol	Actual Porosity (%)	Image Analytical (2000×)
		Porosity (%)	Smax (µm${}^2$)	S (µm${}^2$)
M	69.33	65.48	258.88	1860
MPc9	50.50	50.01	160.87	1422
MIPc9	52.38	51.06	165.75	1451
MIAPc9	42.86	40.91	63.67	1164

. The small error between actual and image analytical porosity (about −0.97% to −5.55%) demonstrated the credibility of PCAS.

Quantitative data in Table 4 also suggests that sample MIAPc9 had the best curing effect among the three types of cement solidified samples. Compared to the natural soil of sample M, the maximum pore area decreased by 75.41%, the total pore area decreased by 37.42%, and the actual porosity reduced by 38.18% for sample MIAPc9. On the other hand, the curing effect of sample MIPc9 was lower than that of sample MPc9. This may be due to the acid-base neutralization reaction between ISS and cement. All the microscopic results were consistent with the aforementioned physical and mechanical test results of corresponding samples.

3.3. Microscopic Analysis for Solidification Mechanism

3.3.1. SEM Results

As shown in Figure 9, the calcium hydroxide (CH), calcium aluminate hydrate (CAH) gel and calcium silicate hydrate (CSH) gel can be observed visually in samples MPc9 and MIAPc9, but hard to be found in sample MIPc9. These results indicated that the new substances of cement hydration were difficult to form in the acidic environment of ISS. In addition, large amounts of ettringite (AFt) presented in sample MIAPc9 rather than in sample MPc9 demonstrated that the SO${}_4^{2-}$ in ISS promoted the production of AFt. The consistency between the above macroscopic and microscopic results suggested that ISS can promote the formation of cement hydration products, accelerate the solidification of soils, and increase the strength of cement solidified soils when the pH-value of solidification reaction was adjusted to neutral or alkaline by adding NaOH. On the contrary, the combined use of ISS and cement without NaOH for soil solidification may hinder the hydration reaction of cement, and it was difficult to generate gels to enhance soil strength, the solidification effect of which was even lower than that of single cement solidified soil.

3.3.2. XRD Results

XRD pattern for natural Ningde marine muck as shown in Figure 10a pointed out that quartz, albite and orthoclase were the main crystal phases for sample M. Besides, clay minerals (e.g., montmorillonite, illite, and chlorite) with high content were detected in sample M. Figure 10b shows that the main components of cement were alite (C${}_3$S), larnite (C${}_2$S), tricalcium aluminate (C${}_3$A), celite (C${}_4$AF), calcite, and gypsum, which will react with soil, water, ISS, and NaOH.

Samples prepared on oriented sheet and XRD patterns of directional samples are shown in Figure 10c. Almost the same XRD patterns for samples M and MI indicated that there was no new product generated after single ISS-soil stabilization. However, compared to 001 reflection in sample M, the d${}_{001}$-value of montmorillonite in sample MI reduced from 1.417 nm to 1.408 nm after single ISS solidification, which testified that the interlayer water was exhausted from montmorillonite through the irreversible action of ISS, the principle has been already proved by the previous literature on the research of single ISS solidified soil [26].

The XRD patterns of nondirectional samples (MPc9, MIPc9, and MIAPc9) in Figure 10d further confirmed the above SEM observation and suggested that, CH, CAH, and CSH were more easily formed in samples MPc9 and MIAPc9. Although AFt was not observed in samples MPc9 and MIPc9 in SEM images, however, through XRD analysis, AFt was found in all samples as MPc9, MIPc9 and MIAPc9. We can qualitatively explain these phenomena by SO${}_4^{2-}$ in ISS, the interaction between acidic ISS and alkaline Portland cement, and the cement hydration reaction. Nevertheless, in order to establish correlation with macro mechanical test data, the quantitative analysis of new products needs to be further studied in detail.

3.3.3. XRF Results

To further explain the change in composition during solidification reactions, XRF tests were also carried out for different samples and the results are summarized in

Table 5. XRF results of different samples (LOI: loss on ignition).

Sample Symbol	Oxide Composition and Content (%)
	SiO${}_2$	Al${}_2$O${}_3$	Fe${}_2$O${}_3$	MgO	CaO	Na${}_2$O	K${}_2$O	MnO	TiO${}_2$	P${}_2$O${}_5$	SO${}_3$	LOI
Pc	22.72	6.70	3.40	2.72	53.76	0.14	0.65	0.13	0.29	0.12	2.63	6.64
M	61.15	17.58	5.46	1.82	0.56	1.45	3.49	0.05	0.79	0.09	0.64	6.84
MI	61.85	17.78	5.26	1.79	0.29	1.41	3.61	0.04	0.79	0.09	0.90	6.11
MPc9	55.82	15.82	5.01	1.78	4.60	1.26	3.29	0.06	0.76	0.09	0.73	10.71
MPc12	54.04	15.38	4.89	1.77	5.63	1.19	3.12	0.06	0.70	0.09	0.78	12.29
MIPc9	56.12	16.20	4.98	1.83	4.88	1.30	3.17	0.06	0.73	0.09	0.97	9.59
MIPc12	55.12	16.13	4.98	1.84	5.58	1.29	3.05	0.06	0.73	0.09	1.01	10.04
MIAPc9	52.43	16.17	4.91	1.91	4.32	2.31	3.21	0.05	0.79	0.08	1.09	12.21
MIAPc12	52.43	15.49	4.78	1.83	5.24	2.11	3.18	0.06	0.74	0.08	1.16	12.84

. Compared to sample M, there was no significant change of oxide compositions and contents for sample MI except for SO${}_3$ content, which was attributed to the high SO${}_4^{2-}$ concentration in ISS. This result tied well with previous studies wherein single ISS solidified soil samples [26,29,39,47]. However, compared samples M and MI with cement solidified samples (MPc9, MPc12, MIPc9, MIPc12, MIAPc9, and MIAPc12), the SiO${}_2$, Al${}_2$O${}_3$, CaO, Na${}_2$O, SO${}_3$, and LOI (loss on ignition) had more obvious changes than other oxides. Considering the XRF results of Portland cement in Table 5, the high content of CaO in sample Pc (53.76%) caused the increase of CaO content in each cement solidified sample, resulting in the formation of new hydration products containing Ca${}^{2+}$. Moreover, the increase of Na${}_2$O in samples MIAPc9 and MIAPc12 was attributed to the addition of NaOH containing Na${}^{+}$. And the increase of SO${}_3$ in solidified samples was caused by the element S both in ISS and in cement. The content of LOI increased from 6.64% (sample M) to 12.84% (sample MIAPc12), partly due to the element C in hydrocarbon radical of ISS [26], and partly due to the carbonation of hydration products.

4. Discussion

According to the cement hydration reaction mechanism [3,57] and the above-mentioned SEM and XRD results as shown in Figure 9 and Figure 10, the main hydration products of Portland cement were CH, CAH, and CSH for samples MPc9 and MIAPc9 after 7 days curing period.The main chemical equations for cement hydration are as follows:

$$2(3\mathrm{CaO}·{\mathrm{SiO}}_2)+6{\mathrm{H}}_2\mathrm{O}⟶\mathrm{C}-\mathrm{S}-\mathrm{H}+3\mathrm{Ca}{\left(\mathrm{OH}\right)}_2$$

(6)

$$2(2\mathrm{CaO}·{\mathrm{SiO}}_2)+4{\mathrm{H}}_2\mathrm{O}⟶\mathrm{C}-\mathrm{S}-\mathrm{H}+\mathrm{Ca}{\left(\mathrm{OH}\right)}_2$$

(7)

$$3\mathrm{CaO}·{\mathrm{Al}}_2{\mathrm{O}}_3+6{\mathrm{H}}_2\mathrm{O}⟶3\mathrm{CaO}·{\mathrm{Al}}_2{\mathrm{O}}_3·6{\mathrm{H}}_2\mathrm{O}$$

(8)

$$4\mathrm{CaO}·{\mathrm{Al}}_2{\mathrm{O}}_3·{\mathrm{Fe}}_2{\mathrm{O}}_3+2\mathrm{Ca}{\left(\mathrm{OH}\right)}_2+10{\mathrm{H}}_2\mathrm{O}⟶3\mathrm{CaO}·{\mathrm{Al}}_2{\mathrm{O}}_3·6{\mathrm{H}}_2\mathrm{O}+3\mathrm{CaO}·{\mathrm{Fe}}_2{\mathrm{O}}_3·6{\mathrm{H}}_2\mathrm{O}$$

(9)

CAH and CSH hydration products identified in samples MPc9 and MIAPc9 were interlaced with soils to form a more stable structure. Moreover, some early CSH gels also generated and enveloped soils, displaying an obvious transition interface zone [58,59]. Besides, more hydration products were observed in sample MIAPc9 than in sample MPc9, which may be attribute to the role of ISS and NaOH. However, the absence of these hydration gel products in sample MIPc9 was due to the acid-base neutralization reaction between H${}^{+}$ in ISS and Ca(OH)${}_2$ in cement. ISS can dissolve cations (such as H${}^{+}$, K${}^{+}$, Na${}^{+}$, Ca${}^{2+}$, Mg${}^{2+}$, Fe${}^{3+}$ and Al${}^{3+}$) and anions (such as SO${}_4^{2-}$) as verified in our previous research [26]. Hence, the acid-base neutralization reaction is:

$$2{\mathrm{H}}^{+}+\mathrm{Ca}{\left(\mathrm{OH}\right)}_2⟶{\mathrm{Ca}}^{2+}+2{\mathrm{H}}_2\mathrm{O}$$

(10)

As a result of the neutralization reaction, the hydration of cement and pozzolanic reaction were prevented, which led to a lower strength of sample MIPct than that of samples MPct and MIAPct as confirmed in Figure 5 and Figure 6.

On the other hand, the presence of a large number of Ca${}^{2+}$ and SO${}_4^{2-}$ in ISS promoted the production of AFt presenting needlelike and pillared morphological characteristics as shown in Figure 9 [60]. AFt was generated followed the chemical equation as:

$$3({\mathrm{CaSO}}_4·2{\mathrm{H}}_2\mathrm{O})+3\mathrm{CaO}·{\mathrm{Al}}_2{\mathrm{O}}_3+26{\mathrm{H}}_2\mathrm{O}⟶3\mathrm{CaO}·{\mathrm{Al}}_2{\mathrm{O}}_3·3{\mathrm{CaSO}}_4·32{\mathrm{H}}_2\mathrm{O}$$

(11)

Due to the use of ISS in the solidified soils, at the curing age of 7 days, AFt was more easily formed during solidification process. The pillared or needlelike AFt efficiently stabilized soil by constructing a skeleton structure. In the meantime, cementitious hydration products aggregated soils, resulting in a sharp pore reduction. The results were in line with previous studies [61]. Therefore, sample MIPc9 had a certain strength increase compared to sample M, but still had the less strength compared to samples MPc9 and MIAPc9 as shown in Figure 5b. Although the strength of sample MIPc9 did not increase much, the SEM and PCAS results as shown in Figure 7 and Figure 8 and Table 4 illustrated that the pores of sample MIPc9 still had an obvious decrease which was close to the decrease in sample MPc9. This phenomenon was due to the water removal and agglomeration of ISS, which was the effect of single ISS on soil solidification before adding cement [26].

For sample MIAPc9, due to the addition of NaOH before adding cement, neutralization reaction occurred between H${}^{+}$ in ISS and NaOH as:

$${\mathrm{H}}^{+}+\mathrm{NaOH}⟶{\mathrm{Na}}^{+}+{\mathrm{H}}_2\mathrm{O}$$

(12)

After that, the addition of cement in a neutral environment for soil solidification will not influence the hydration reaction of cement [12,62]. In addition, as proposed in references [26,47], ISS can cause a cation exchange reaction on the soil surface, leading to the reduction of d-value (as shown in Figure 10c), decrease of double electric layer and expulsion of water. Furthermore, the large amount of Ca${}^{2+}$ in ISS will promote the pozzolanic reaction and carbonation reaction during the combined solidification process of ISS and cement followed the chemical equations as:

$${\mathrm{SiO}}_2+\mathrm{Ca}{\left(\mathrm{OH}\right)}_2+{\mathrm{nH}}_2\mathrm{O}⟶\mathrm{CaO}·{\mathrm{SiO}}_2·(\mathrm{n}+1){\mathrm{H}}_2\mathrm{O}$$

(13)

$${\mathrm{Al}}_2{\mathrm{O}}_3+\mathrm{Ca}{\left(\mathrm{OH}\right)}_2+{\mathrm{nH}}_2\mathrm{O}⟶\mathrm{CaO}·{\mathrm{Al}}_2{\mathrm{O}}_3·(\mathrm{n}+1){\mathrm{H}}_2\mathrm{O}$$

(14)

$$\mathrm{Ca}{\left(\mathrm{OH}\right)}_2+{\mathrm{CO}}_2+{\mathrm{H}}_2\mathrm{O}⟶{\mathrm{CaCO}}_3\downarrow +{\mathrm{H}}_2\mathrm{O}$$

(15)

As a result, clay minerals closely aggregated, pores decreased, and the strength of ISS-NaOH-cement-solidified soil samples significantly enhanced by formation and cementation of hydration products.

5. Conclusions

Atterberg limits, water content, pH, UCT, DST, SEM, PCAS, XRD, and XRF tests were carried out to study the solidification effect and mechanism of marine muck treated with the combination of ISS and Portland cement, under the different conditions of ISS:water ratio, Portland cement dosage, and whether the alkalizer NaOH has been added before cement. The conclusions are as follows:

• ISS is an eco-friendly and effective stabilizer to enhance the strength of high-water-content soils (such as marine muck) when it was used combined with cement after adding the additives such as NaOH. Under the same Portland cement dosage, the strength of solidified samples ranged from more to less as: MIAPct > MPct > MIPct.
• There was an unified function relationship for all types of solidified samples (such as MPct, MIPct, and MIAPct) between UCS and pH-value by power function, UCS and failure strain by natural logarithmic function, and UCS and cohesion by linear function, respectively, which can be used to predict the strength of other solidified samples with different proportion of the additives of ISS, cement and NaOH. However, the relationship between UCS and Portland cement dosage cannot be described by an uniform functional equation, which was attributed to the different amount of cement involved in acid-base neutralization reaction or hydration reaction varying from sample to sample.
• With the exception of d-value decreasing for clay minerals, Ca${}^{2+}$ and SO${}_4^{2-}$ dissolved from ISS promoted the production of AFt, pozzolanic reaction and carbonation reaction of cement in the presence of NaOH. AFt efficiently stabilized soil by constructing a skeleton structure. In the meantime, cementitious hydration products aggregated soils, resulting in a sharp pore reduction and a high strength improvement for solidified soils.
• As a partial substitute, the best reinforcement effect of the ISS-NaOH-cement-solidified soil samples was determined. Furthermore, this combination of stabilizers can not only save the dosage of cement, but also accelerate the solidification speed, decrease the cement setting time within 7 days to meet the curing requirements, and improve the strength of solidified soils. Thus, this solidification method can be applied in engineering practice for the advantages of high strength, resource reuse and environmental protection.