Alkylpolyglycosides—Based Formulations for Sustainable Remediation of Contaminated Aquifers: Lab-Scale Process Study for NAPL Solubilization Assessment

Abstract

In the context of the surfactant-enhanced remediation of polluted sites, this work focuses on the development of non-ionic alkylpolyglucosidic (APG) surfactant formulations with different additives for the solubilization and mobilization of strongly adsorbed organic pollutants. The study involved three consecutive phases to evaluate the effect of the additives on surface behavior and the potential improvement in alkylpolyglycoside surfactant’s capability to solubilize toluene and perchloroethylene (PCE), selected as reference contaminants. After a chemical–physical characterization phase, the APG-based formulations were first used in a batch configuration test, in which the formulations’ solubilization ability was indirectly assessed by observing the effect on pollutants’ adsorption. Lastly, a continuous configuration column experiment was performed to simulate the flushing process of a synthetic matrix previously contaminated with strongly adsorbed toluene or PCE. The results showed that the presence of additives firstly reduced the ability of the surfactant to form micelles, increasing the CMC, but at the same time improved the ability to reduce surface tension. Moreover, the addition of the additives overall resulted in a significant improvement in adsorbed pollutant removal in a minimal volume of fed solution, reaching 96% and 99% efficiencies for toluene and PCE, respectively, compared with 76% and 92%, for toluene and PCE, respectively, in the presence of free-additive APG surfactant.

1. Introduction

The pollution of environmental matrices by hydrophobic organic compounds (HOCs), such as petroleum hydrocarbons or chlorinated solvents, is an environmental issue on a global scale due to adverse effects on natural ecosystems and human health [1,2]. They are highly hydrophobic and immiscible compounds in water and therefore, following a spill, they tend to behave as Non-Aqueous Phase Liquids (NAPLs). These compounds can be less dense than water (light NAPLs, LNAPLs), such as petroleum hydrocarbons, or denser than water (dense NAPLs, DNAPLs) such as chlorinated solvents [3,4]. The low solubility in water and high hydrophobicity of NAPLs, their complex architecture in the subsoil, and their presence in low-permeability layers or poorly accessible fractures [5,6] generate high-persistence slow-release secondary contamination sources [7,8]. Traditional pump-and-treat (P&T) techniques are ineffective, requiring long operational times and unsustainable costs [9,10]. In addition, the P&T technique does not work on the contamination secondary source [11,12]. Therefore, desorption and back diffusion phenomena from clay lenses or fractures are very common, which continue to release contaminants into groundwater also when aqueous concentrations become lower than legal limits, making the achievement of remediation goals extremely complex [13,14].

In the field of sustainable remediation, the development of innovative technologies to directly attack persistent contamination sources is crucial to promote their rapid depletion [15]. For example, in the operational context of Surfactant-Enhanced Aquifer Remediation (SEAR) technology, the use of green surfactant solutions in ex situ washing or in situ flushing processes is an excellent strategy for the removal of hydrophobic contaminants from very persistent source areas [16,17].

Surfactants are amphiphilic molecules, with a polar hydrophilic portion directly bonded to an apolar hydrophobic portion, giving simultaneously the affinity for both the aqueous and the organic phases [18,19]. Surfactants, therefore, can promote the detachment of trapped organic phase ganglia by interfacial tension reduction between immiscible phases. At the same time, it considerably increases the solubility of poorly soluble hydrophobic compounds through the formation of micelles, i.e., aggregates of surfactant monomers within which hydrophobic compounds can effectively allocate themselves [20,21].

In addition, scientific research also makes efforts to optimize formulations to improve surfactant characteristics. A good strategy is the addition of small amounts of additives to the surfactant solution [22]. It is known from the scientific literature that specific co-solvents or co-surfactants, depending on the type of additive and its concentration in the formulation, can modify the critical micelle concentration (CMC) of the surfactant and its ability to reduce surface/interface tension, determining an alteration of solubilization and mobilization power towards organic compounds [22,23,24,25]. Regarding the effect on micellation, the use of low-molecular-weight alcohols (C2–C4) tends to reduce the CMC value, especially for linear-chain alcohols. In the presence of higher molecular weight fatty alcohols (C > 5) or co-surfactants, on the other hand, there is an increase in the CMC value due to electrostatic repulsion effects and steric hindrance [26,27]. Significant effects also occur on surface or interfacial properties. In this context, the presence of co-solvents or co-surfactants in a surfactant formulation can offer a synergistic effect: specifically, mixed adsorption can occur at the liquid–air or liquid–liquid separation interface, which makes the formulation more efficient and effective in reducing surface/interface tension [26,27,28]. The properties of surfactant formulations can also be advantageously utilized in the context of SEAR technology for remediation, where they can enhance the solubilization of organic contaminants and their mobilization by reducing interfacial tension to much lower values than single surfactants [29,30,31].

Based on these insights, this work is a prosecution of two previous studies [32,33] in which two alklipoliglycosides (APG 1 and APG 2), a sophorlipide (SL), and two rhamnolipids (RL 1 and RL 2) were tested to promote solubilization and mass transfer in the aqueous phase of strongly adsorbed toluene or perchloroethylene (PCE). These previous studies showed that PCE was solubilized by synthetic APG 2 and biological RL 2 with comparable efficiencies (92% and 98%, respectively). On the other hand, it was also observed that the synthetic surfactant APG 2 was significantly less effective in mobilizing toluene than the biosurfactant RL 2, whose mobilization efficiencies were 75% and 86%, respectively. In this scenario, the feasibility of enhancing the performance of the synthetic surfactant APG 2 by adding suitable additives was evaluated in this work. Therefore, three different APG-based formulations were tested using a C6 branched alcohol (APG 2a) and two polyetoxylated co-surfactants with a different ethoxylation degree (APG 2b and APG 2c). The comparative study started with a chemical–physical characterization of formulations to evaluate the effect of additives on CMC and surface tension reduction ability. The second step involved in thermodynamic batch experiment, in which the adsorption of toluene and PCE on a pine-wood biochar (PWB) was evaluated in the presence of investigated formulations as an indirect method to assess their ability in NAPL-enhanced solubilization. Finally, a column experiment was performed to test directly APG 2-based formulations in a soil flushing simulation on a synthetic porous matrix with strongly adsorbed NAPLs. To ensure the proper feasibility of performance comparison between formulations and individual free-additive APG 2 surfactants, the entire study was conducted with the same experimental setups and the same reference pollutants as previous studies. In this way, it was possible to more accurately evaluate the positive (or negative) effect associated with the presence of the additives.

Overall, the present work aims to provide valuable information on the feasibility of modifying and optimizing the performance of a surfactant through the careful selection of specific additives that can be used in formulations, ensuring a more sustainable approach for soils and aquifer surfactant-enhanced remediation.

2. Materials and Methods

2.1. Materials

2.1.1. Investigated Formulations

Three different formulations were taken under investigation, denoted by the acronyms APG 2a, APG 2b, and APG 2c. Each formulation was prepared by adding a small quantity of specific additives to the original alkylpoliglycosidic surfactant APG 2 [32,33]. In particular, the employed additives were a C6 branched fatty alcohol as a co-solvent, a commercial C10 polyethoxylated co-surfactant (POE-CS), and a commercial C10 polyethoxylated-propoxylated co-surfactant (POE-P-CS), respectively.

Table 1. Composition of investigated formulations and chemical structure of additives.

Formulation	Composition	Additive Chemical Structure
APG 2a	APG 2 + C6 branched alcohol (15:1)
APG 2b	APG 2 + C 10 POE-CS (15:1)
APG 2c	APG 2 + C 10 POE-P-CS (15:1)

shows the compositions of the investigated formulations.

2.1.2. Sorbent Material

Biochar from pine wood wastes was used as sorbent reference material for batch tests. PWB was produced by the gasification of wood at approximately 850 °C in V 3.90 Burkhardt and ECO 180 HG wood gas generator (Burkhardt GmbH, Plößberg bei Tirschenreuth, Germany). This microporous material is characterized by a total pore volume of 0.383 cm³ g ⁻¹, with a specific surface area of 343 ± 2 m² g ⁻¹ and a carbon content of 95.84 wt % [34]. Previous studies have shown that this material has a high adsorption capacity due to its high organic carbon content and remarkable surface development, making it a cost-effective alternative to activated carbon [34].

2.2. Methods

2.2.1. Critical Micelle Concentration (CMC) Determination

The study began with the characterization of the formulation by determining their critical micelle concentration (CMC). A series of surface tension measurements of the surfactants were performed, as a function of concentration, according to the ring detachment method, by an MWG LAUDA (Lauda-Königshofen, Germany) tensiometer using a platinum–iridium alloy ring (Du-Nouy ring) with a diameter of 0.95 cm [35,36]. During the experiment, each surfactant solution was diluted with high-purity milli-Q water, and the concentration gradually decreased, starting from 1 wt % with steps of 0.2 wt %. Since the surface tension measurement is sensitive to impurities [37], the ring was carefully heated with a flame before each measurement. The sample holder was cleaned first with distilled water, then with nitric acid, and finally with ultrapure milli-Q water.

Milli-Q water with electric resistivity equal to 18.2 MΩ cm was directly collected from a deionizer Smart-Mini Water Purification System (Shanghai Canrex Analytic Instruments, Shanghai, China).

2.2.2. Batch Configuration Study: NAPL Adsorption Isotherms

A thermodynamic study for adsorption isotherm was performed to assess the mobilization capacity of surfactants by evaluating the reduction in the adsorption capacity of contaminants on a reference sorbent material in the presence of surfactants. Two organic compounds were selected as target contaminants, toluene and PCE, representative of LNAPL and DNAPL constituents, respectively [38]. To achieve a maximum adsorption condition, far superior to any soil, a PWB was selected as the reference sorbent material.

A series of isothermal batch experiments with PWB/contaminant/surfactant were conducted at room temperature (23 ± 2 °C) and pressure in 20 mL batch reactors (VWR International glass vials, Milan, Italy) using different amounts of sorbent material (10, 20, 50, 80, 100, and 200 mg) and 50 mg L^–1 of contaminant. The contaminated solutions were prepared in distilled water and stored in gas-tight collapsible Tedral bags^® (Supelco, Bellefonte, PA, USA), which prevent the formation of headspace and ensure the constant maintenance of the concentration of the solution containing volatile compounds [9]. Different concentrations of each surfactant were investigated, just below and just over the critical micelle concentration, in particular half (0.5× CMC) and five times the CMC (5× CMC), respectively.

Reference tests were performed under the same operating conditions but without surfactant solutions. Samples of PWB and the contaminants in the presence and absence of surfactants were prepared in glass vials. The glass vials were filled without creating headspace to avoid the dispersion of contaminants in the gas phase of each glass reactor, hermetically sealed with a Teflon face gray butyl stopper (Wheaton, Millville, NJ, USA), crimped by an aluminum cap, and stirred on rotating plates for 24 h. Each test was performed in duplicate to evaluate reproducibility. The initial concentration of the contaminant was analyzed at time zero (C₀) and after 24 h (Ce). This is the time required to reach the thermodynamic equilibrium condition between the sorbent material and the dissolved contaminant in the liquid phase [34].

2.2.3. Continuous Configuration Column Experiments: Surfactant-Enhanced Desorption of NAPLs

A continuous configuration experiment was carried out to assess an evaluation of the ability of the selected surfactants to enhance the solubilization of hydrophobic organic pollutants strongly adsorbed on PWB. For this purpose, six plexiglass (polymethylmethacrylate, PMMA) columns (h = 14 cm, id = 2.8 cm, Vg = 68.68 cm³) packed with a granular fixed bed were set up. The experimental setup was made according to Rossi et al. [34]. Specifically, the first and the last 3 cm of the columns were packed with only silica sand in order to make the liquid phase front homogeneous. On the other hand, the central part (7 cm high) of each column was filled with a mixture of sand and 4% wt of PWB in which the adsorption/desorption process of organic pollutants occurs (reactive zone). At the bottom and the top of each column, two three-way sampling ports were installed (port IN and port OUT, respectively) to allow the entry and the exit of liquid phase and its sampling. A scheme of the experimental setup is shown in Figure 1.

Before starting the experiments, a tracer test for each column was carried out by feeding a 100 mg L⁻¹ KCl solution with a constant flow rate Q = 0.6 mL min⁻¹ (step signal) to calculate the fluid dynamic parameters of columns, such as the porosity (ε), the pore volume (VP), and the hydraulic retention time (HRT) [39].

Continuous column tests involved two consecutive operating steps: an initial contamination phase of the reactive zone followed by the flushing phase of adsorbed pollutants.

The contamination phase of the reactive zone was first performed in order to replicate and simulate the representative conditions of a polluted aquifer. During this phase, columns were continuously fed in an upward flow with toluene or PCE solutions prepared with distilled water into a self-collapsing Tedlar-bag^® with a concentration equal to 100 mg L⁻¹. The contamination phase was daily monitored by repeated sampling at regular times of both influent and effluent. This step was stopped at the equilibrium condition when the pollutant’s effluent concentration was equal to the influent one (C_OUT/C_IN = 1), indicating that the reactive bed adsorption capacity was at equilibrium with respect to the concentration of the fed solution, and no more toluene or PCE could be adsorbed at the same aqueous concentration.

After the equilibrium, a flushing phase was carried out by switching the column feeding from pollutant solution to APG 2 formulations (APG 2a, 2b, or 2c) with a concentration equal to five times their CMC (5× CMC). Each flushing solution was prepared into polyethylene terephthalate (PET) flasks by an opportune dilution of the initial stock solution with distilled water. The daily sampling of the effluent at regular time intervals allowed the control of the flushing process of sorbed pollutants by surfactants. The flushing phase was stopped when the measured pollutant concentration in the effluent stabilized at a constant value of 1–2 mg L⁻¹.

In both the contamination and flushing phases, a peristaltic pump (Gilson Miniplus evolution, Middleton, WI, USA) was used to allow the entry of the solution into the column. The pump’s speed was regulated in order to have a constant flow rate (Q) equal to 0.6 mL min⁻¹.

Table 2. Characteristics of fixed-bed reactors.

	g of PWB (Reactive Zone)	Contaminant	Flushing Agent
PWB-TOL-a	3.00	Toluene (LNAPL)	APG 2a
PWB-TOL-b	2.53	Toluene (LNAPL)	APG 2b
PWB-TOL-c	2.83	Toluene (LNAPL)	APG 2c
PWB-PCE-a	2.62	Perchloroethylene (DNAPL)	APG 2a
PWB-PCE-b	2.88	Perchloroethylene (DNAPL)	APG 2b
PWB-PCE-c	2.89	Perchloroethylene (DNAPL)	APG 2c

summarizes the characteristics of all the set-up fixed-bed reactors in terms of PWB amount in the reactive zone, pollutants employed in the contamination phase, and mobilizing agents used in the flushing phase.

2.2.4. Analytical Methods

Toluene and PCE were determined with gas chromatography (GC) DANI MASTER, equipped with a DANI 86.50 headspace auto-sampler (DANI Instruments, Contone, Switzerland). A capillary column (30 m × 0.53 mm ID × 3 µm, TRB624) and a Flame Ionization Detector (FID) were used. The headspace analysis program was performed as follows: oven temperature 80 °C, manifold T 120 °C, transfer line temperature 180 °C, and soft shaking for 1 min. The GC conditions were as follows: helium (He)carrier gas (flow 10 mL min⁻¹); 180 °C injector temperature; split injection 1:2; 300 °C detector temperature with air; and nitrogen (N₂) and hydrogen (H₂) for the FID (flows 240, 25, and 60 mL min⁻¹).

The oven temperature was programmed as follows: setting at 70 °C for 0.5 min, raising at 30 °C min⁻¹ speed up to 90 °C and then at the same speed up to 180 °C. For the quantitative determination of toluene and PCE, a calibration curve was obtained by the dilution of a stock solution in standards within a concentration range of 0.1–5 mg L⁻¹.

Chloride concentration for the tracer test was continuously monitored by measuring the conductivity of the reactor outlet using a Handylab^® 330 (SI-analytics, Weilheim, Germany) conductometer.

2.3. Data Processing and Calculation

To calculate the CMC value, the surface tension vs. the logarithm of the surfactant concentration (logC) was reported. The data were fitted with two linear regressions, one below the CMC where surface tension decreases with increasing concentration and the other above the CMC where surface tension remains almost constant, and the CMC values were calculated from the breakpoints of the two regions [32].

To compare the performance of the different surfactants, the efficiency and effectiveness parameters were evaluated for both the adsorption to the liquid–air surface and the reduction in surface tension. In this regard, the following parameters were calculated:

• The efficiency of surface adsorption was evaluated by the maximum (monostrate) surface excess concentration (Γm) [40,41] using Equation (1).

$${\Gamma }_1\cong {\Gamma }_m=-{\displaystyle \frac{1}{2.3 nRT}}{\left({\displaystyle \frac{\partial \gamma }{\partial \log C_1}}\right)}_T$$

(1)

where n is the number of the solute species (for non-ionic surfactant, n = 1); R = 8.314 is the gas constant [J mol⁻¹ K⁻¹]; T is the temperature [K]; and (∂γ/(∂ logC₁) is the slope of the linear fit of the data below the CMC.

• The effectiveness in surface tension reduction was evaluated by the surface pressure at CMC (π_CMC) and by the difference between the liquid–air surface tension of pure solvent (γ₀) and the surfactant solution at surfactant’s CMC (γ_CMC) [42,43], as reported in Equation (2).

$${\pi }_{CMC}={\gamma }_0-{\gamma }_{CMC}$$

(2)

• Efficiency in surface tension reduction calculating the value of the negative logarithm of the surfactant concentration needed for the reduction in the surface tension of water by 20 mN m⁻¹ (pC₂₀ = −logC₂₀) [40,41], as described in Equation (3).

$${-\log C_{20}=p C}_{20}={\displaystyle \frac{{\pi }_{CMC}-20}{2.3RT{\Gamma }_m}}-\log CMC$$

(3)

For the thermodynamic batch study, the solid phase equilibrium concentration (q_e) was calculated considering the correspondent aqueous equilibrium concentration (C_e) using Equation (4).

$$q_e={\displaystyle \frac{\left(C_0-C_e\right)V}{m}}$$

(4)

where C₀ is the initial NAPL concentration in the aqueous phase (synthetic polluted water), [mg L^–1]; V is the volume of liquid phase used to fill each batch reactor [L]; and m is the mass of PWB used for each test [g].

Then, the Langmuir isotherm model was used for equilibrium data fitting purposes. The model is reported in Equation (5).

$$q_e=q_{max}{\displaystyle \frac{K_L{C}_e}{1+K_L{C}_e}}$$

(5)

where q_max is the maximum quantity of adsorbed species [mg g^–1], K_L is the Langmuir thermodynamic constant [L mg^–1], and C_e and q_e are the equilibrium concentrations in aqueous [mg L^–1] and solid [mg g⁻¹] phases, respectively. Parameter calculation was performed by Sigma Plot 12. Then, these parameters were finally used to assess formulates’ ability in NAPL solubilization through an investigation of the formulates’ effect on NAPL adsorption.

For the tracer test, F-curves were built by plotting outlet Cl⁻ concentration normalized by inlet Cl⁻ concentration (C_OUT/C_IN) as a function of time. The first derivate method was used to determine the flex point of the F-curve, representing the experimental hydraulic retention time (HRT). The pore space’s volume (V_P) was calculated with Equation (6) by multiplying the HRT and the flow rate (Q). Then, the ratio between V_P and the geometrical volume (V_G) gives the porosity (ε) of the fixed bed, as reported in Equation (7).

$$V_P \left(mL\right)=HRT \left(min\right) \ast Q \left({\displaystyle \frac{mL}{min}}\right)$$

(6)

$$\epsilon \left(\%\right)={\displaystyle \frac{V_P\left(mL\right)}{V_G\left(mL\right)}} \ast 100$$

(7)

A three-parameter sigmoidal equation (Equation (8)) uploaded in Sigma Plot 12.0 was used to fit concentration data for obtaining the breakthrough curves.

$$y=a/(1+\exp \left(x-x_0\right)/b)$$

(8)

Contamination phase data were processed in order to calculate the total amount of adsorbed pollutants when the equilibrium was reached (C_OUT/C_IN = 1), which is equal to the value of the area between the breakthrough curve and the Y axis. For this purpose, it was used a graphical method based on the subtraction of the breakthrough curve’s integral (mass of contaminant leaving the column) to the total area of the graph (total mass of fed contaminant) [34]. Flushing-phase concentration data instead were processed to calculate the total amount of solubilized pollutants, corresponding to the flushing curve integral.

From the ratio between the solubilized amount and the adsorbed amount, it was possible to calculate the efficiency removal parameter, useful for comparing the ability of surfactants in NAPL removal.

3. Results and Discussions

3.1. Critical Micelle Concentration (CMC) Determination

The plots of the surface tension vs. log of concentration are reported in Figure 2a–c for the aqueous solutions of the formulations APG 2a, APG 2b, and APG 2c, respectively, showing the typical double linear trend of surface tension in the function of surfactant’s concentration, whose intersection provides the CMC values. Using the calculation methods reported in Section 2.3, surface/interfacial efficiency and effectiveness parameters (i.e., γ_CMC, π_CMC, Γm, pC20, and CMC/C₂₀) were also evaluated and compared with the same parameters obtained for the APG 2 synthetic surfactants without additives. These values are listed in

Table 3. Micellar and surface properties of surfactant aqueous solutions at 298.15 K and atmospheric pressure compared with the APG 2 of the previous study [32].

Material	CMC (% wt)	γCMC (mN m−1)	πCMC (mN m−1)	Γm (mol cm−2)	pC20	CMC/C20
APG 2a	6.13 × 10−2 ± 7.8 × 10−3	29.15 ± 0.05	43.65 ± 0.05	2.69 × 1010 ± 7.9 × 108	2.06 ± 0.02	7.08 ± 0.65
APG 2b	3.33 × 10−2 ± 1.8 × 10−3	30.97 ± 0.32	41.83 ± 0.32	2.42 × 1010 ± 6.4 × 108	2.27 ± 0.02	6.24 ± 0.38
APG 2c	3.74 × 10−2 ± 3.6 × 10−3	32.84 ± 0.25	39.86 ± 0.25	2.10 × 1010 ± 2.4 × 108	2.37 ± 0.05	8.74 ± 0.41
APG 2 [32]	7.10 × 10−3 ± 5.6 × 10−4	32.95 ± 0.13	39.95 ± 0.13	4.21 × 1010 ± 3.3 × 108	2.98 ± 0.03	6.70 ± 0.32

.

Observing Table 3, it is immediately visible that the CMC of the three formulations increased by an order of magnitude compared to the CMC of the single surfactant. Specifically, the measured CMC values were 6.13 × 10⁻², 3.33 × 10⁻², and 3.74 × 10⁻² % wt for the APG 2a, APG 2b, and APG 2c formulations, respectively, against the 7.1 × 10⁻³ % wt of the APG 2 single surfactant.

From these early data, it is evident that the presence of the additive results in an impediment to micellation rather than an enhancement of this mechanism. The APG 2a formulation was characterized by the absolute highest CMC (6.13 × 10⁻² % wt). This value is probably because the additive used is a branched fatty alcohol (C6), which by adsorbing at the micelle-water interface exhibits a steric hindrance effect that depresses micellation, greatly increasing the CMC value [26,27]. In the APG 2b and APG 2c formulations, the CMC values were about half the CMC value of the APG 2a formulation. In this case, there is a synergistic effect between the main surfactant APG 2 and the co-surfactant used in the formulation, unlike in the presence of fatty alcohol C6. However, the CMC was higher than that of the single surfactant APG 2, probably due to electrostatic repulsions between the hydrophilic groups of the two non-ionic components of the formulation (APG 2 + co-surfactant) and to the branched structure of the co-surfactants used [26,40,44]. In addition, looking at the CMCs of formulations 2b and 2c (3.33 × 10⁻² and 3.74 × 10⁻² % wt, respectively) the influence of the chemical structure of the co-surfactant is evident, which is crucial in explaining the different CMC values [41,45,46]. As shown in Table 1, the C9 POE-P-CS additive is characterized by a propylene oxide group that acts as a linker between a C9 alkyl chain and a polar hydrophilic group composed of a chain of 5–14 (average) ethylene oxide units, which undergoes a greater steric cluttering effect during micellation than the C9 POE-CS additive that lacks the linker and has a hydrophilic group composed of a chain of 3–9 (average) ethylene oxide units.

Although the CMCs of the formulations were greater than the CMC of the single APG 2 surfactant, and they adsorbed to the liquid–air surface less effectively (Γm = 2.69 × 10¹⁰, 2.42 × 10¹⁰, and 2.1 × 10¹⁰ mol cm⁻² for the APG 2a, 2b, and 2c formulations, respectively, and Γm = 4.21 × 10¹⁰ mol cm⁻² for APG 2), they were significantly more effective in reducing surface/interface tension. In fact, the surface tension values measured when the system is at the CMC (γ_CMC) were lower for the formulations than for the basic surfactant APG 2, and specifically 29.15 mN m⁻¹ for APG 2a, 30.97 mN m⁻¹ for APG 2b, and 32.84 mN m⁻¹ for APG 2c, against 32.95 mN m⁻¹ for the non-additive APG 2 surfactant. Among the three additives used in the formulations, the greatest surface tension reduction efficacy (lowest value of γ_CMC and greatest value of π_CMC) was observed for the APG 2a formulation characterized by the presence of branched C6 alcohol, which, since it is structurally the smallest, is the one that adsorbs most effectively to the liquid–air surface. In contrast, the APG 2c formulation resulted in the least reduction in surface tension due to the bigger structure of the additive used (C9 POE-P-CS), which adsorbs less effectively on the surface due to the very large size of the hydrophilic group and the oxypropylene linker, which exerts additional steric hindrance preventing the adsorption process on the surface [26,40,44].

3.2. Thermodynamic Batch Study: NAPL Adsorption Isotherms

The adsorption isotherm tests were further conducted to evaluate the mobilization capacity of the three investigated APG 2- based formulations (APG 2a, APG 2b, and APG 2c) in the presence of toluene from the BTEX (Benzene–Toluene–Ethylbenzene–Xylene) group, and PCE for the chlorinated solvent family. In this context, the ability of formulations to reduce the adsorption capacity on reference sorbent material (PWB) was studied to assess a preliminary evaluation of their ability to mobilize organic hydrophobic pollutants.

The experimental data were processed using the Langmuir adsorption isotherm model, from which the characteristic thermodynamic parameters, such as q_max (mg g⁻¹) and K_L (L mg⁻¹), were optimized for each setup system. Figure 3 shows the graphical trends of the adsorption isotherms related to the systems contaminated with both toluene and PCE in the presence and absence (reference) of the three alkylpolyglycoside formulations at different concentrations with respect to their CMC. It was also compared with the adsorption isotherm obtained in the previous study with the same contaminants in the presence of the surfactant APG 2 (Cs = 5× CMC) without the addition of the additives. The equilibrium data are well fitted to the Langmuir model. This peculiarity is also confirmed by the high regression coefficient R².

3.2.1. Adsorption Data for Toluene-Contaminated Systems

Table 4. Optimized parameters of the Langmuir model of toluene adsorption on PWB in the presence and absence of APG formulations compared with the APG 2 and H₂O test of the previous study [32].

		Formulates’ Concentration
Formulate	Langmuir Parameters	5× CMC	0.5× CMC
APG 2a	qmax (mg g−1)	17.08 ± 1.1	45.78 ± 2.3
	KL (L mg−1)	11.4 × 10−2 ± 36 × 10−3	24.6 × 10−2 ± 25 × 10−3
	R2	0.95	0.94
APG 2b	qmax (mg g−1)	20.56 ± 1.5	65.97 ± 1.5
	KL (L mg−1)	12.5 × 10−2 ± 8 × 10−3	41.1 × 10−2 ± 28 × 10−3
	R2	0.90	0.95
APG 2c	qmax (mg g−1)	17.86 ± 1.1	31.46 ± 1.8
	KL (L mg−1)	13.9 × 10−2 ± 16 × 10−3	37.2 × 10−2 ± 25 × 10−3
	R2	0.92	0.98
APG 2 (5× CMC) [32]	qmax (mg g−1)	20.14 ± 1.4
	KL (L mg−1)	15.3 × 10−2 ± 36 × 10−3
	R2	0.91
Reference PWB-TOL [32]	qmax (mg g−1)	77.71 ± 2.5
	KL (L mg−1)	15.4 × 10−2 ± 55 × 10−3
	R2	0.97

lists the thermodynamic parameters of the Langmuir model optimized for toluene-contaminated batch systems, whose adsorption isotherms are shown in Figure 3a–c for the APG formulations 2a, 2b, and 2c, respectively. From a very first view of Table 4 and the graphical trends of the adsorption isotherms, it can be seen that the presence of each formulation, at a concentration of 5× CMC or 0.5× CMC, leads to a substantial decrease in the thermodynamic parameter q_max compared with the reference system (without neither surfactant or formulation). This can be explained by the effect of the surfactant, which, by increasing the apparent solubility of hydrophobic compounds, physically prevents their adsorption, resulting in a decrease in q_max. At the same time, it was also observed that in the presence of the formulations, the K_L value decreases with increasing surfactant concentration, indicating a decrease in the affinity of toluene for PWB.

Specifically, the formulation that caused the greatest decrease in toluene adsorption on the PWB in this study was the formulation APG 2a at a concentration of 5× CMC for which q_max was equal to 17.08 ± 1.1 mg g⁻¹ compared with 77.71 ± 2.5 mg g⁻¹ in the reference system in the absence of surfactant (78% decrease). At the same time, the APG 2a formulation at a 5× CMC also showed the greatest effect on the thermodynamic parameter K_L. In this case, there was a 26% decrease in K_L from 15.44 × 10⁻² L mg⁻¹ in the reference test to 11.4 × 10⁻² L mg⁻¹ in the presence of APG 2a at a 5× CMC.

Even when working at a concentration of the APG 2a formulation equal to 0.5× CMC, the reduction in toluene adsorption on the PWB was observed; however, it was less intense than in the system with formulation concentration above its CMC. In this case, q_max was reduced by only 41% compared to 78% obtained by working above CMC. Differently, the K_L value was found to be higher than the value calculated for the reference test (K_L = 24.6 × 10⁻² L mg⁻¹ and K_L = 15.44 × 10⁻² L mg⁻¹, respectively).

Comparable behavior to the APG 2a formulation was observed for the system containing the APG 2c formulation at a 5× CMC. In this case, compared with the reference system, toluene q_max on PWB decreased by 77%, from 77.71 ± 2.5 mg g⁻¹ to 17.86 ± 1.1 mg g⁻¹. Instead, for the same system, K_L decreased by 9% in comparison with the reference system, from 15.4 × 10⁻² L mg⁻¹ to 13.9 × 10⁻² L mg⁻¹. When the 2c formulated was used at a concentration of 0.5× CMC, the q_max of toluene was equal to 31.46 ± 1.8 mg g⁻¹, so decreased if compared with the reference system without surfactant, but less intensely if compared with the system above CMC. Similarly to the observations for the APG 2a system, even in the presence of APG 2c at a concentration below its CMC, the Langmuir constant was 150% greater than that measured in the reference system.

Among the three formulations tested, APG 2b produced the least promising effects on reducing the adsorption and affinity of toluene on PWB. In fact, the presence of APG 2b at 5× CMC caused the reduction of q_max from 77.71 ± 2.5 mg g⁻¹ (reference test) to 20.56 ± 1.5 mg g⁻¹ (73% reduction). Also on the value of K_L, the formulated APG 2b at 5× CMC showed the least intense effect, reducing its value from 15.4 × 10⁻² L mg⁻¹ to 12.5 × 10⁻² L mg⁻¹ (19% reduction). In comparison with the other formulations, APG 2b was less promising also at a concentration below its CMC. In this case, q_max was reduced by 15% from 77.7 ± 2.5 mg g⁻¹ of the reference system to 66 ± 1.5 mg g⁻¹. Langmuir’s constant increased about 3-fold (an increase of about 300%) to a value of 41.1 × 10⁻² L mg⁻¹.

Focusing on Lamguir constant (K_L) behavior, for each investigated system, it can be seen that the presence of the formulations at 0.5× CMC always caused an increase in this parameter compared to the system without surfactant (reference), indicating an increase in the affinity between toluene and adsorbent. This behavior could be due to the presence of hemimicelles, which are monolayer, or “head-on” oriented formations of surfactants on the surface of a solid, typically at concentration values just below the CMC. Since they are aggregates of surfactant monomers, hemimicelles could retain toluene, which would then not be directly adsorbed by the active sites (decrease of q_max) but at the same time lead to an “apparent” increase in the affinity between adsorbate and adsorbent [47,48,49].

Comparing the optimized thermodynamic parameters for the formulations with the thermodynamic parameters obtained in the presence of the single APG 2 surfactant, both at a concentration of 5× CMC, we saw that the presence of the additives had an overall positive effect. In fact, both the formulated APG 2a and APG 2c resulted in a greater reduction in toluene adsorption than the single surfactant APG 2 (q_max = 17.08 ± 1.1 mg g⁻¹ for APG2a, q_max = 17.86 ± 1.1 mg g⁻¹ for APG 2c and q_max = 20.14 ± 1.4 mg g⁻¹ for APG 2). In contrast, in the presence of the APG 2b formulation, toluene adsorption on the PWB remained essentially unchanged compared with the system with non-additive APG 2, with optimized values of q_max = 20.56 ± 1.5 mg g⁻¹ and q_max = 20.14 ± 1.4 mg g⁻¹, respectively.

Regarding the Langmuir constant, the application of additives at a greater concentration than CMC (5× CMC) resulted in a reduction in this parameter in all the systems compared with the single APG 2 surfactant, demonstrating the advantage of the synergy between surfactants and co-solvents on interfacial and adsorption processes.

3.2.2. Adsorption Data for PCE-Contaminated Systems

The thermodynamic parameters of the Langmuir model optimized for systems containing PCE as the reference contaminant are summarized in

Table 5. Optimized parameters of the Langmuir model of PCE adsorption on PWB in the presence and absence of APG formulations compared with the APG 2 and H₂O test of previous study [32].

		Formulates’ Concentration
Formulate	Langmuir Parameters	5× CMC	0.5× CMC
APG 2a	qmax (mg g−1)	37.05 ± 3.5	43.68 ± 4.9
	KL (L mg−1)	49.9 × 10−3 ± 12 × 10−3	31.83 ×10−2 ± 17 × 10−3
	R2	0.93	0.96
APG 2b	qmax (mg g−1)	40.31 ± 5.1	58.65 ± 6.7
	KL (L mg−1)	60.6 ×10−3 ± 24 × 10−3	79.28 ×10−2 ± 28 × 10−3
	R2	0.95	0.97
APG 2c	qmax (mg g−1)	38.64 ± 4.1	47.39 ± 5.2
	KL (L mg−1)	53.3 ×10−3 ± 21 × 10−3	55.4 ×10−2 ± 18 × 10−3
	R2	0.96	0.98
APG 2 (5× CMC) [32]	qmax (mg g−1)	39.42 ± 4.7
	KL (L mg−1)	85 × 10−3 ± 25 × 10−3
	R2	0.91
Reference PWB-PCE [32]	qmax (mg g−1)	114.12 ± 7.8
	KL (L mg−1)	29 × 10−2 ± 48 × 10−3
	R2	0.98

. Graphical trends of adsorption on the PWB are shown in Figure 3d–f concerning the presence of the formulations APG 2a, APG 2b, and APG 2c, respectively.

Looking at the data shown in Table 5 and the corresponding adsorption isotherms, it is evident that all the formulations tested effectively reduced the adsorption of PCE on the PWB. In particular, the APG 2a formulation at a 5× CMC caused the most intense effect on PCE adsorption compared to the reference system in the absence of surfactant, reducing q_max from 114.12 ± 7.8 mg g⁻¹ to 37.05 ± 3.5 mg g⁻¹ (68% reduction). Following in order of efficiency there are the APG 2c and APG 2b formulations, both at 5× CMC, with 66% and 64% q_max reduction, respectively (q_max = 38.64 ± 4.1 mg g⁻¹ for APG 2c and q_max = 40.31 ± 5.1 mg g⁻¹ for APG 2b). Even at 0.5× CMC formulation concentration, the adsorption of PCE was significantly reduced compared with the system without active component, and the effect on the adsorption mechanism follows the same order as the system setup over CMC: APG 2a > APG 2c > APG 2b with q_max reductions of 61%, 58%, and 48%, respectively.

Considering the effect on the Langmuir constant, the decrease in this parameter was significantly more intense than that observed for toluene. Working at a concentration of formulation equal to 5× CMC, the values of K_L were lowered by an order of magnitude compared to the reference system, demonstrating how the presence of surfactant significantly reduced the PCE-PWB affinity. Also, in terms of K_L reduction, the efficiency of the formulations follows the order APG 2a > APG 2b > APG 2c. Similarly to the observations with toluene, the presence of the formulations below their CMC caused the opposite effect, i.e., an increase in K_L value probably due to the formation of hemimicelles on the surface of the solid that retain PCE, which, not effectively being adsorbed on the PWB, shows a greater “apparent” affinity for the PWB [47,48,49].

From the comparison between formulations and single APG 2 surfactant, both with a concentration five times greater than CMC (5× CMC), it is clear that the presence of additives produced a positive effect on the adsorption reduction capacity on PWB and PCE-PWB affinity, although not very intense. In particular, the calculated q_max for APG 2a was 6% lower than the calculated q_max for APG 2; with formulated APG 2c, q_max reduction was just 2%; and with formulated APG 2b PCE, adsorption on PWB was substantially the same. Instead, the reduction in K_L is definitely more pronounced, and, in particular, compared to the system with single APG 2, reductions in this parameter of 42%, 28%, and 37% were measured in the presence of the APG 2a, 2b, and 2c formulations, respectively.

3.3. Continuous Configuration Column Experiments: Surfactant-Enhanced NAPL Desorption

3.3.1. Fluid-Dynamic Characterization of Fixed Bed Reactors

The fluid dynamic characterization of the continuous systems was carried out by tracer testing using the step signal method. Fluid dynamic parameters, i.e., mean hydraulic retention time (HRT), pore volume (V_P), and porosity (ε), were calculated from the F-curve using the first derivative approach.

For toluene-contaminated columns, the HRT values were 71 min, 46.3 min, and 70.2 min for PWB-TOL-a, PWB-TOL-b, and PWB-TOL-c, respectively. Following the same order, the estimated porosities were 62.1%, 42.5%, and 59.0%. In parallel, HRTs and porosities of 46.0 min, 45.8 min, 50.4 min, 42.2%, 41.6%, and 44.1% were calculated for PWB-PCE-a, PWB-PCE-b, and PWB-PCE-c, respectively (where a, b, and c indicate the formulation used for the flushing step).

The fluid dynamic parameters calculated for each system are summarized in

Table 6. Fluid dynamic parameters for all six fixed-bed systems. Geometrical volume, V_G = 68.68 cm³. Flow rate, Q = 0.6 mL min⁻¹; ε = porosity; V_P = pore volume HRT = hydraulic retention time.

Column	HRT (min)	VP (cm3)	ε (%)
PWB-TOL-a	71.0	42.6	62.1
PWB-TOL-b	46.3	29.3	42.5
PWB-TOL-c	70.2	40.7	59.0
PWB-PCE-a	46.0	29.1	42.2
PWB-PCE-b	45.8	28.7	41.6
PWB-PCE-c	50.4	30.2	44.1

below.

3.3.2. Contamination Phase of the Fixed Bed: Contaminant Adsorption

Data processing of the contamination phase consisted of plotting the C_OUT/C_IN ratio of the target pollutants against the pore volume (PV) (volume of the solution passing through the column normalized by the total pore volume (V_P) of the fixed bed) to obtain the breakthrough curves for both the toluene and PCE columns, reported in Figure 4.

For the PWB-TOL-a, PWB-TOL-b, and PWB-TOL-c columns, the breakthrough points were reached after 48.2, 74.6, and 47.6 PV of the fed solution, and the equilibrium after 191, 151, and 183 PV, respectively. For the “PWB-PCE” columns, the breakthrough points were reached after 102.7, 167.7, and 171.6 PV, and the equilibrium after 209, 348, and 359 PV for PWB-PCE-1, PWB-PCE-2, and PWB-PCE-3, respectively.

The mass of the adsorbed contaminant at equilibrium with the inlet concentration can be calculated graphically by the difference between the total mass of the contaminant fed to the column until any PVs after the equilibrium condition and the integral of the breakthrough curve (representing the mass of contaminant leaving the column).

The calculated adsorbed masses of toluene were 349.18 mg, 241.63 mg, and 306.70 mg for the PWB-TOL-a, PWB-TOL-b, and PWB-TOL-c columns, respectively. On the other hand, the adsorbed masses of PCE were 382.96 mg, 450.12 mg, and 627.70 mg for the PWB-PCE-a, PWB-PCE-b, and PWB-PCE-c columns, respectively.

3.3.3. Flushing Phase: Contaminants Solubilization

The last experimental stage of the continuous test is the flushing phase, which began when each column had reached equilibrium condition and was carried out by feeding formulations into each column to solubilize adsorbed pollutants. All the formulations were employed with a concentration equal to fivefold their CMC (5x CMC).

Toluene flushing curves are reported in Figure 5 describing the behaviors of the formulations APG 2a, 2b, and 2c towards the adsorbed toluene (Figure 5a, Figure 5b, and Figure 5c, respectively). The flushing curve in the presence of the single surfactant APG 2 is also reported in Figure 5d for comparison purposes. Figure 6 shows the flushing curves of PCE in the presence of formulations (Figure 6a for APG 2a, Figure 6b for APG 2b, and Figure 6c for APG 2c) compared with the PCE flushing curve in the presence of the single APG 2 surfactant (Figure 6d).

For the PWB-TOL-a column, flushing with the formulated APG 2a (Cs = 5× CMC), resulted in a very rapid increase in the toluene output concentration, reaching a very high concentration peak with a maximum value C_OUT,MAX = 973.8 mg L⁻¹ in just 2.4 pore volumes of the fed solution, corresponding to about 4.5 h of treatment. The value reached by the concentration peak was found to be about 10 times greater than the equilibrium concentration value and about 4 times greater than the value reached by the concentration peak after treatment with non-additive APG 2, in which C_OUT,MAX = 237.7 mg L⁻¹ (Figure 5d).

On the other hand, in the presence of the same formulation (APG 2a), the PCE concentration in the effluent reached a peak with a maximum value of C_OUT,MAX = 583.6 mg L⁻¹ in five pore volumes of feeding, which corresponds to about 8 h of treatment. In this case, the peak concentration reached a value approximately two times higher than the peak obtained with APG 2 without additives, where C_OUT,MAX = 324.6 mg L⁻¹ (Figure 6d).

From the maximum point, the outlet concentration of both toluene and PCE decreased slowly until reaching a plateau around 1–2 mg L⁻¹. The flushing phase for the PWB-TOL-a and PWB-PCE-a columns was stopped after 321.7 PV and 309.3 PV, respectively. At the end of APG 2a flushing, by integrating the under-curve area, 334.3 mg of toluene and 380.4 mg of PCE were solubilized with a flushing efficiency of 96.0% and 99.8%, respectively. From the removal efficiency data, the positive contribution of the used additive which was able to increase significantly the performances of APG 2, characterized by removal efficiencies equal to 74.5% and 91.8% for toluene and PCE, respectively, is evident.

Similar behavior was also observed as a consequence of flushing with the APG 2b formulation for both toluene (in Figure 5b) and PCE (Figure 6b). For the PWB-TOL-b column, feeding the formulation APG 2b led to a rapid increase in the output concentration of toluene to a maximum C_OUT,MAX = 392.6 mg L⁻¹ reached after 12 PV of the fed solution, resulting in approximately twice the maximum value measured at the output following flushing with APG 2 for the same contaminant. Also, for the PWB-PCE-b column, the passage of the formulated APG 2b produced a significant increase in the output concentration until a maximum peak was reached on C_OUT,MAX = 537.2 mg L⁻¹. In this case, the concentration peak was reached slowly, i.e., after 45 PV of feeding, corresponding to 26.5 h of treatment. Again, the passage of the formulation resulted in a significantly greater increase in the output concentration of PCE than the single surfactant APG 2, specifically about 1.8 times higher. In terms of performance, flushing with APG 2b removed 209.5 mg of toluene in 246.4 PV and 427.7 mg of PCE in 222.2 PV, corresponding to the removal efficiencies of 86.7% and 95.1%, respectively.

Finally, for the PWB-TOL-c and PWB-PCE-c columns, both fed with the APG 2c formulated solution, the worst performance was found among the different system setups. Looking at the graphs shown in Figure 5c and Figure 6c, it can be seen that the output concentrations of toluene and PCE reached their respective peaks at C_OUT,MAX = 215.8 mg L⁻¹ and 284.7 mg L⁻¹ in 7 PV and 15 PV of feeding, respectively. Although the peaks were reached quite rapidly, they were the lowest of all for both toluene and PCE. Moreover, in the presence of APG 2c, the maximum output concentration reached is also lower than the peak concentration measured during flushing with APG 2. As a consequence, this behavior resulted in significantly lower removal efficiencies compared with the other formulations and compared with single APG 2. In fact, by integrating the areas under the flushing curves, the removal of 179 mg toluene and 264.1 mg PCE were calculated, corresponding to the removal efficiencies of 58.4% and 68.6%, respectively.

Below,

Table 7. Summarized data of continuous tests for both contamination and flushing phases. All data from formulation flushings are compared with APG 2 and H₂O flushings of previous study [33].

	Contamination Phase			Flushing Phase
Column	Adsorbed Contaminant (mg)	Adsorbed Contaminant (mg g−1)	PV *	Flushing Agent	PV **	Solubilized Contaminant (mg)	Removal (%)
PWB-TOL-a	349.2	116.3	148.6	APG 2a	51	334.3	95.7
PWB-TOL-b	241.6	95.6	194.1	APG 2b	102	209.5	86.7
PWB-TOL-c	306.7	108.4	141.7	APG 2c	169	179.1	58.4
PWB-TOL-APG 2 [33]	296.3	98.6	145	APG 2	131	223.4	75.4
PWB-TOL-H2O [33]	292.2	97	202	H2O	250	145.4	49.7
PWB-PCE-a	382.9	146.3	234.5	APG 2a	94	380.4	99.3
PWB-PCE-b	450.1	156.1	356.6	APG 2b	118	427.7	95.0
PWB-PCE-c	527.7	132.2	414.3	APG 2c	152	314.1	59.5
PWB-PCE-APG 2 [33]	399.8	132.9	339	APG 2	123	367	91.8
PWB-PCE-H2O [33]	427.3	144	349	H2O	380	217.8	50.9

* refers to the first equilibrium point (C_OUT/C_IN = 1) detected during the contamination phase. ** refers to the pore volume required for the solubilization of 95% of the total adsorbed amount.

summarizes the performance data of the six columns set up for both the contamination and flushing phases. The different data are compared with the systems treated with simple APG 2 and with water investigated in the previous study [32]. In this way, comparative evaluations can be made between the formulations and the single APG 2 surfactant based on the co-solvent or co-surfactant used in the preparation of the formulation.

The data show that the use of branched C6 fatty alcohol and C10 POE-CS as additives in the APG 2a and APG 2b formulations, respectively, gave a concrete advantage to the mobilization and solubilization of the target pollutants.

First of all, the removal efficiencies of toluene and PCE were definitely increased compared to what was observed in the presence of the single surfactant APG 2. In particular, compared to flushing with APG 2, toluene was solubilized with 20% and 10% higher efficiency when treated with APG 2a and APG 2b, respectively. On PCE, the effect was less intense, but still positive, with efficiency increases of 7% with APG 2a and 3% with APG 2b compared with the surfactant APG 2.

A very important aspect to note regards that these higher efficiencies were achieved by feeding a much smaller pore volume of solution than was required with APG 2. Most importantly, with APG 2a, the mobilization of toluene occurred in just 51 PV of feed (versus 131 PVs required in the presence of APG 2). For PCE, 94 PVs were required to achieve 99% removal compared with 123 PVs fed in the presence of APG 2 for 91.8% removal.

In opposition was the behavior of the APG 2c formulation, prepared with C10 POE-P-CS, which caused a reduction in the efficiency of the APG 2 surfactant. The calculated removal efficiencies were much lower than those calculated in the presence of the single APG 2 surfactant. A kinetic limitation of the process was also observed in this case, as the lower overall pollutant mobilization occurred in a larger pore volume of the fed solution than the APG 2 flushing.

This different behavior of the APG 2c formulation can be attributed to the nature and structure of the additive used. Because it is a co-surfactant with a relatively larger structure than the other additives tested, it is definitely characterized by lower performance in terms of micellization and surface/interface properties [40,50] as already highlighted in the characterization study and also finding experimental confirmation following the batch study.

4. Conclusions

This study investigated the technical possibility of increasing the performance of a synthetic alckylpoliglosicidc surfactant in solubilizing and mobilizing organic pollutants. The surfactant, named APG 2, was selected based on the results of a previous study in which it resulted in the less efficient mobilization of toluene and perchloroethylene adsorbed onto a porous solid matrix. For this reason, three APG 2-based formulations were prepared by adding three different additives aimed to increase APG 2 efficiency in terms of surface/interfacial behavior and in terms of solubilization power toward organic contaminants. A C6 branched alcohol, a C10 polyethoxylated co-surfactant (C10 POE-CS), and a C10 polyethoxylated-propoxylathed co-surfactant (C10 POE-P-CS) were used as additives for the formulations APG 2a, 2b, and 2c, respectively. Thus, for comparison purposes, the same experimental protocol of the previous work was adopted.

Physicochemical characterization demonstrated that the additives produced an increase in the CMCs that were one order of magnitude greater than the CMC of the original APG 2. At the same time, they were able to enhance the surface and interfacial phenomena more than the single surfactant APG 2. In this context, APG 2a and APG 2b were the most effective in surface tension reduction with γ_CMC = 29.15 mN m⁻¹ and 30.9 mN m⁻¹, respectively, against the value of γ_CMC = 32.9 mN m⁻¹ estimated for APG 2.

These findings were confirmed with a preliminary batch configuration test which studied NAPL adsorption on reference material in the presence or the absence of surfactant formulations. In this test, our findings demonstrated that all the formulations induced a great adsorption reduction in NAPLs if compared to the reference test. In particular, the q_max of toluene was reduced by 78% in the presence of APG 2a, by 73% in the presence of APG 2b, and by 77% with APG 2c. Following the same order, the q_max of PCE was reduced by 68%, 64%, and 66%. On the other hand, the formulation performances were slightly greater with respect to APG 2 for both toluene and PCE adsorption.

Lastly, a continuous column test was performed to evaluate the formulations’ ability in adsorbed NAPL solubilization, allowing for a direct assessment of the additives’ impact on surfactant performances. The C6 branched alcohol in APG 2a was the best enhancer of APG 2 performances, leading to a very significant improvement in NAPL solubilization, especially for toluene, raising its solubilization efficiency from 76% to 96% in a much shorter pore volume of the fed solution than APG 2 (51 and 131 pore volume, respectively).

Significant effects were also detected for PCE solubilization. In this case, APG 2a produced a solubilization efficiency equal to 99% in 93 pore volume of the fed solution, resulting in more effective than APG 2 which solubilized 91% of PCE in 124 pore volume. On the contrary, the C10-POE-P-CS additive in APG 2c led to a performance worsening in terms of NAPL removal and process kinetics with respect to APG 2: the toluene solubilization efficiency was just 58% in 169 pore volume against 76% in 131 pore volume in the presence of APG 2. For the PCE solubilization, the efficiency was about 60% in 152 pore volume compared to 92% in 123 pore volume of APG 2.

In conclusion, this study highlighted the feasibility of increasing surfactant properties by adding specific additives in formulations, exploiting a synergistic effect between surfactants and additives in interfacial and micellar processes. This synergy, depending on the additive, can improve surfactant properties that can also be advantageously applied for environmental remediation from both technical and kinetic points of view, since formulations can give better removal efficiencies in a shorter volume of feeding.