Investigation of Earth-Abundant Metal Salts for the Inhibition of Asphalt-Derived Volatile Organic Compounds

Asphalt is used globally in construction for roads, pavements, and buildings; however, as a fossil-derived material, it is known to generate volatile organic compounds (VOCs) upon exposure to heat and light that can be harmful to human health. Several heterogeneous strategies have been reported for the inhibition of these VOCs; however, the direct use of inexpensive, accessible Earth-Abundant metals has not been extensively explored. In this study, simple metal salts are examined for their coordination capability toward asphalt-derived VOCs. From UV–visible (UV–vis) spectroscopic studies, FeCl3 emerged relative to other metal salts (metal = Mn, Co, Ni, Cu, Zn) as a promising candidate for the adsorption and retention of Lewis basic compounds. Coordination of an example oxygen-containing VOC, benzofuran (Bf), to Fe yielded a paramagnetic semi-octahedral complex Fe(Bf)3Cl3. Evaluation by thermal gravimetric analysis (TGA) coupled to infrared spectroscopy (IR) demonstrated that the complex was stable up to 360 °C. Spectroscopic evaluation demonstrated the stability of the complex upon visible light irradiation and in the presence of a variety of organic pollutants. The potential application of Fe was demonstrated by subjecting biochar to FeCl3 followed by the addition of Bf. It was discovered that this Fe-rich biochar was successful at adsorbing Bf suggesting the possibility of introducing Fe to biochar late-stage in processing to deter asphalt degradation and VOC emissions. An understanding of the binding and stability of Fe salts to VOCs provides insight into how a sustainable infrastructure can be achieved.


COMPUTATIONAL METHOD
The interaction of FeCl3 with Bf and candidate VOCs was examined using DFT.3] Optimizations were performed with the convergence criteria of 2.0 × 10 -5 Eh, 4.0 × 10 -3 Eh/Å, and 5.0 × 10 -3 Å for energy, maximum force, and displacement, respectively.] Stabilization energies (ES) for formation of complexes were evaluated using Eq.S1.The electrostatic COSMO potential was integrated into the SCF procedure, so the optimized total energies of the product comprised solvent effects.ES = ΣEt (products) -ΣEt (substrates) (S1) In the above formulation, Et (products) is the total energy of the products, and Et (substrates) is the total energy of the substrates in a specific reaction.Negative values of ES indicate the thermodynamic stability of the products.

ELECTRON PARAMAGNETIC RESONANCE SPECTROSCOPY
Spin Hamiltonian.The EPR spectrum of the Fe compound was interpreted using a S = 5/2 spin Hamiltonian, H, containing the electron Zeeman interaction with the applied magnetic field B0, and the zero-field interaction. 6= βe S .g .B0 + h S .D .S (S2) Where, 'S' is the electron spin operator, 'D' is the zero-field interaction tensor in frequency units, 'g' is the electronic gtensor, 'βe' is the electron magneton, and 'h' is Planck's constant.
Fitting of EPR Spectra.To quantitatively compare experimental and simulated spectra, we divided the spectra into N intervals.We treated the spectrum as an N-dimensional vector R.Each component Rj has the amplitude of the EPR signal at a magnetic field Bj with j varying from 1 to N. The amplitudes of the experimental and simulated spectra were normalized so that the span between the maximum and minimum values of Rj was 1.We compared the calculated amplitudes Rj calc of the signal with the observed values Rj defining a root-mean-square deviation 'σ' by: where the sums are over the N values of j, and p's are the fitting parameters that produced the calculated spectrum.For our simulations, N was set equal to 2048.The EPR spectrum was simulated using EasySpin (v 5.2.35), a computational å j package developed by Stoll and Schweiger and based on Matlab (The MathWorks, Natick, MA, USA). 7EasySpin calculates EPR resonance fields using the energies of the states of the spin system obtained by direct diagonalization of the spin Hamiltonian (Eq.S2).The EPR fitting procedure used a Monte Carlo type iteration to minimize the root-meansquare deviation, σ (Eq.S3), between measured and simulated spectra.We searched for the optimum values of the following parameters: the isotropic g-value (giso), the zero-field splitting parameters (D and E), and the isotropic peak-topeak line width (ΔB).
The vial was fitted with a cap and the reaction was stirred at 300 rpm for 1 h at 60 °C.After heating, the reaction was filtered using filter paper placed in a glass funnel and the filtrate was concentrated in vacuo.Unless otherwise stated, this procedure was used for the synthesis of the complex used for the stability, stoichiometry, and selectivity studies (Section III-Section V).For the synthesis of the Fe-Bf complex and other experiments described below, no unexpected or unusually high safety hazards were encountered.

B. GENERAL PROCEDURE FOR THE MAGNETIC MOMENT MEASUREMENT OF THE Fe-Bf COMPLEX:
The magnetic properties of the Fe-Bf complex were investigated to understand the paramagnetic susceptibility (χM) of the paramagnetic Fe-Bf complex using EPR spectroscopy and Guoy's method.In the EPR experiment the tube was filled with 500 µL DCM containing 5 mg of the Fe-Bf complex and EPR spectra was noted after sample was cooled to -125 °C.
To measure the gram susceptibility (χg) of the complex using Guoy's method, 70 mg of the complex was filled in the glass tube (Guoy's tube) and measurements were conducted using [Ni(en)3][S2O3] as a calibrant.First the instrument was set to zero by inserting a blank Guoy's tube and then 92.2 mg of the calibrant was filled in the tube and the reading of the instrument was noted.The value of 'χg' for [Ni(en)3][S2O3] was 0.00001103 emu/gm; using this value, 'C' was calculated employing the following equation: In equation S4, 'C' is the calibration constant calculated using the calibrant whose χg is known, 'L' is the length of the sample in the Guoy's tube in cm (≥1.5 cm), 'R' is the balance reading for the calibrant and sample (after calculating C using calibrant), 'Ro' is the balance reading in the absence of sample (the blank), and 'm' is the sample mass in g.
After calculating the value of 'C', the 'χg' for the complex was calculated and from the value of χg, the molar susceptibility (χm) of the complex was obtained using the following equation: where 'M' is the molecular weight of the complex in g/mol.The molecular weight of the complex was deduced from the possible structures of the Fe-Bf complex based on stoichiometric studies (Figure 7 in the manuscript).
The 'χm' can be used to calculate the effective magnetic moment (µeff) using the following relation in Bohr Magnetons: where 'χm' is measured in m 3 /mol and 'T' is measured in Kelvin.

A. GENERAL PROCEDURE FOR THE STUDY OF INTERACTION BETWEEN METAL SALTS AND Bf:
A 1-dram vial (VWR glass vials, 470151-622) equipped with a PTFE-coated stir bar (VWR spinbar micro, 3 x 10 mm, 58948-375) was charged with DCM (1.00 mL).To the reaction vial metal salt (0.050 mmol, 50.0 mM) and Bf (0.050 mmol, 50.0 mM) were added.The vial was fitted with a cap and the reaction was stirred at 300 rpm for 1 h at 60 °C.After stirring, 4 μL of the prepared solution was added to a quartz cuvette containing 1 mL of DCM.The UV-vis spectrum was collected from 250 nm through 800 nm for each of the prepared samples individually.Additionally, control samples for the iron salt and Bf were also recorded.

B. GENERAL PROCEDURE FOR THE STUDY OF THE STABILITY OF THE Fe-Bf COMPLEX UPON HEATING
Five 20 mL scintillation vials equipped with PTFE-coated stir bars were each charged with 10 mL of a 50 mM FeCl3 solution in DCM.To the vials varying amount of Bf (50 mM, 100 mM, 150 mM, 250 mM and 500 mM) were added to obtain a mixture of FeCl3 and Bf in ratios of 1:1, 1:2, 1:3, 1:5 and 1:10, respectively (Table S1).The vials were fitted with caps and the reactions were stirred at 300 rpm for 1 h at 60 °C.After stirring, 1.00 mL of the complex solution from each of the above solutions was transferred to three 1-dram vials (VWR glass vials, 470151-622) equipped with a PTFE-coated stir bar (VWR spinbar micro, 3 x 10 mm, 58948-375).Identical solutions were then stirred at 300 rpm for 1 h at different reaction temperatures (60 °C, 80 °C, and 100 °C) for 1 h.In order to achieve high temperature stirring of the complex in DCM at 80 and 100 °C, the vials were sealed using electrical tape and closed caps, rather than the septa-fitted caps.
After stirring was completed, the vials were cooled down to room temperature.10 μL of the prepared solution was added to a quartz cuvette containing 1.00 mL of DCM and the UV-vis absorption spectrum was collected from 250 nm through 800 nm for each of the prepared samples individually.
Table S1.Various proportions of FeCl3 and Bf were employed to form the complex.

C. GENERAL PROCEDURE FOR THE STUDY OF THE STABILITY OF THE Fe-Bf COMPLEX UPON IRRADIATION:
Five 20 mL scintillation vials equipped with PTFE-coated stir bars were each charged with 10 mL of a 50 mM FeCl3 solution in DCM.To the vials varying amount of Bf (50 mM, 100 mM, 150 mM, 250 mM and 500 mM) were added to obtain a mixture of FeCl3 and Bf in ratios of 1:1, 1:2, 1:3, 1:5 and 1:10, respectively (Table S1).The vials were fitted with caps and the reactions were stirred at 300 rpm for 1 h at 60 °C.After stirring, 1.00 mL of the complex solution from each of the above solutions was transferred to three 1-dram vials (VWR glass vials, 470151-622) equipped with a PTFE-coated stir bar (VWR spinbar micro, 3 x 10 mm, 58948-375).Identical solutions were then stirred at 300 rpm for 1 h while being irradiated at different wavelengths (390 nm, 456 nm, and 525 nm).After stirring was completed, the vials were cooled to room temperature.20 μL of the prepared solution was added to a quartz cuvette containing 1.00 mL of DCM and the UVvis absorption data was collected from 250 nm through 800 nm for each of the prepared samples individually.

D. GENERAL PROCEDURE FOR THE STUDY OF THE STABILITY OF THE Fe-Bf COMPLEX IN DIFFERENT SOLVENTS:
In the experimental set-up 7.5 mM solution of FeCl3 and a 7.5 mM solution of Bf were prepared in two different DCM: methanol solvent system (9:1 and 1:1; v/v ratio).Then, six 1-dram vials (VWR glass vials, 470151-622) equipped with a PTFE-coated stir bar (VWR spinbar micro, 3 x 10 mm, 58948-375) were charged with 2.00 mL of the 7.5 mM FeCl3 solution.To the reaction vials some of the Bf stock solution was added in volumes to obtain different ratios of Fe to Bf as shown in Table S2.The vials were fitted with a cap and were stirred at 300 rpm for 1 h at 60 o C.After stirring was completed, the vials were cooled down to room temperature.10 μL of the prepared solution was added to a quartz cuvette containing 1.00 mL of DCM and the UV-vis spectrum was collected from 250 nm through 800 nm for each of the prepared samples individually.In addition, UV-vis spectrum was collected for FeCl3 and Bf solutions as controls.Similarly, the experiment was repeated at higher concentration (50 mM) as shown in Table S2.
Table S2.Calculations for the preparations of the Fe-Bf complex for the stability studies.
In a different experiment, the stability of the Fe-Bf complex was evaluated in the presence of various organic solvents.
The complex was prepared using 50 mM FeCl3 and 50 mM Bf in DCM after stirring at 300 rpm for 1 h at 60 °C (Section III-A).After stirring was completed, the vials were cooled down to room temperature.For UV-vis analysis of the different samples, 20 µL of the complex was added to a quartz cuvette charged with DCM: solvent (1:1; v/v ratio).The sample was analyzed in the presence of different organic solvents including cyclohexane, methanol, toluene, DCM, ethyl acetate, dimethylformamide, acetonitrile, dimethyl sulfoxide, tetrahydrofuran, N, N-dimethylacetamide and diethyl ether.The UVvis spectrum was collected from 250 nm through 800 nm for each of the prepared samples.

E. GENERAL PROCEDURE FOR THE STUDY OF THE FORMATION OF THE Fe-Bf COMPLEX AT DIFFERENT STOICHIOMETRIC RATIOS OF Fe AND Bf
To a 20 mL scintillation vial, 50 mM stock solution of FeCl3 and 50 mM stock solution of Bf were prepared in DCM.The stock solutions were then mixed in the correct proportions to obtain Fe:Bf ratios of 1:1, 1:2, 1:3, 1:4, 1:5 and 1:10.The vials were then stirred at 300 rpm for 1 h at 60 °C.After stirring was completed, the vials were cooled down to room temperature.10μL of the prepared solution was added to a quartz cuvette containing 1.00 mL of DCM and the UV-vis absorption spectrum was collected from 250 nm through 800 nm for each of the prepared samples individually.

F. GENERAL PROCEDURE FOR THE MOLAR RATIO METHOD AND JOB PLOT
The stoichiometry of the Fe-Bf complex (50 mM) was investigated using two different UV-vis absorption spectroscopy methods.Two different plots were drawn including a Molar Ratio Plot and a Job Plot.
In the molar ratio method, the absorption of the complex is plotted against the molar ratio of the two interacting species while keeping the concentration of Fe constant.A 50 mM stock solution of FeCl3 was prepared in DCM.1-dram vials (VWR glass vials, 470151-622) equipped with a PTFE-coated stir bar (VWR spinbar micro, 3 x 10 mm, 58948-375) were charged with 1.0 mL of the FeCl3 solution.To prepare the Fe-Bf complex with varying Fe to Bf ratios (1:1, 1:2, 1:3, 1:4, 1:5, 1:6, 1:7, 1:8, 1:9, and 1:10), the vials were subjected to increasing concentrations of Bf ranging from 50 mM, 100 mM, 150 mM, 200 mM, 250 mM, 300 mM, 350 mM, 400 mM, 450 mM and 500 mM respectively.The vials were then stirred at 300 rpm for 1 h at 60 °C.After stirring was completed, the vials were cooled down to room temperature.10 μL of the prepared solution was added to a quartz cuvette containing 1.0 mL of DCM and the UV-vis data was collected from 250 nm through 800 nm for each of the prepared samples individually.The variation in the absorbance intensity of the absorption peak at 554 nm were observed and plotted with respect to the molar ratio of the complex and the stoichiometry of the complex formed was deduced from the position of breaks in the absorption curve.
In the Job plot method, the Fe-Bf complex was prepared using different mole fractions of the FeCl3 and Bf while keeping the molar concentration of the complex constant.A 100 mM stock solution of FeCl3 and a 100 mM stock solution of Bf were prepared in DCM.Next, 1-dram vials (VWR glass vials, 470151-622) equipped with a PTFE-coated stir bar (VWR spinbar micro, 3 x 10 mm, 58948-375) were charged with FeCl3 and Bf in proportions given in Table S3.The vials were then stirred at 300 rpm for 1 h at 60 °C.After stirring was completed, the vials were cooled down to room temperature.10 μL of the prepared solution was added to a quartz cuvette containing 1.0 mL of DCM and the UV-vis data was collected from 250 nm through 800 nm for each of the prepared samples individually.Variations in the absorbance intensity of the absorption peak at 554 nm were observed.The Job plot was deduced from the absorption studies by assigning the xaxis of the plot to the mole fraction of Fe(III), and the y-axis to the absorbance intensity of the absorption maxima at 554 nm.The stoichiometry of the complex was deduced from the maxima of the Job plot curve.
Table S3.Various proportions of FeCl3 and Bf used to form the samples for the Job plot.

G. GENERAL PROCEDURE FOR THE EVALUATION OF THE BINDING CONSTANT FOR THE COMPLEX USING THE BENESI-HILDEBRAND RELATION
Furthermore, the extent of the binding of Bf toward Fe(III) ion is calculated from an experimental plot of the absorption data using the Benesi-Hildebrand relation.A vial was charged with 2.0 mL of 150 mM Bf solubilized in DCM and 10 µL of this was added to a quartz cuvette charged with 1.0 mL DCM and the UV-vis spectrum was noted.To the vial, aliquots of FeCl3 (0-50 mM) were added and 10 µL of the formed solution was drawn after each addition into the cuvette charged with 1.0 mL DCM and the variations in absorption spectrum were noted.The variation in the absorption band at 554 nm was observed and a linear fitted curve was then obtained with logarithm of concentration of FeCl3 on the x-axis and logarithm of the variation in the absorption intensity plotted on the y-axis.The binding constant (Kb) was determined from the intercept value of the linear fitting absorption curve using the following Benesi-Hildebrand equation: in which A0, A and Af are the absorption values, in the absence of, at the intermediate, and at the saturation of the interaction of ferric ion, respectively.For the experiment, UV-vis absorption spectrum of Bf was collected in the presence of increasing amounts of FeCl3 until saturation was observed (Figure S6).The higher the value of the binding constant, the stronger the interaction between the metal ion and the ligand.Also, the experiment was conducted three times to validate the binding affinity of Bf towards ferric ions.

Figure S1:
The binding activity of the potentially interfering VOCs with the Fe-Bf complex and chemical structures of all the VOCs investigated for their potential to displace Bf from Fe-Bf complex.
In this experiment FeCl3 (0.1 mmol, 1.0 equiv.),Bf (0.3 mmol, 3.0 equiv.), and other VOC (0.3 mmol, 3.0 equiv.)were mixed in 2.0 mL DCM and were stirred at 300 rpm for 1 h at 60 °C.After stirring was completed, the vials were cooled down to room temperature.10 μL of the prepared solution was added to a quartz cuvette containing 1.0 mL of DCM and UV-vis spectrum was collected from 250 nm through 800 nm for each of the prepared samples individually (Figure S7-S8).In addition, the controls were run along with the complex to gain insight into the possible interactions of the VOCs with the complex or FeCl3.UV-vis spectra of the VOCs and FeCl3 in the presence of other VOCs were recorded to analyze possible interactions of the VOC with FeCl3.It was observed that some VOCs interacted with FeCl3 evidenced from new peaks in the UV-vis absorption spectrum.In order to further investigate the interaction between the complex and those VOCs (quinoline and triethylene glycol), titration studies were conducted.Small aliquots (5 mM each) of the VOCs were added into the pre-formed complex solution (50 mM) stepwise and the absorption spectra were noted after each addition.

I. GENERAL PROCEDURE FOR THE DIRECT APPLICATION OF Fe-INCORPORATED BIOCHAR FOR Bf INHIBITION:
In this experiment, 100 mg of the biochar was solubilized in 10 mM FeCl3 solution prepared in DCM.The solution was then stirred vigorously at 1700 rpm for 24 h for adsorption to take place.Meanwhile, known concentrations of FeCl3 were spectroscopically analyzed to obtain a calibration plot as shown in Figure S15.The solution was then filtered using a Buchner funnel with a fritted disc (VWR, 10545-880) and the concentration of the Fe(III) ions in the filtrate was calculated using the calibration plot.The adsorption capacity of the biochar was then calculated by subtracting the amount in the filtrate from the initially fed concentration.The Fe-treated biochar was analyzed using TGA and FTIR after drying the residue in oven at 80 °C for 15 minutes.The finely grounded dry material (10-30 mg) was weighed into 70 μl Al2O3 pan with an Al lid.The samples were then heated under a flow of 20 mL/min N2 from 25 °C to 1000 °C at a ramp rate of 10 °C/min.The residue was also analyzed using FTIR spectroscopy from 400-4000 cm -1 .Biochar and FeCl3 were also analyzed individually by TGA and FTIR as controls.The residue was then solubilized in DCM followed by addition of Bf to form Fe-Bf complex.From the calibration plot, the adsorption capacity was obtained to be 13.02 mg/g of biochar which means that out of the 10 mM FeCl3 which was subjected to 100 mg of biochar, 4 mM of the FeCl3 was adsorbed.[Fe] (µM)

Figure S8 .
Figure S8.UV-visible absorption spectra of Fe-Bf complex in the presence of 3 equiv.of various VOCs.

Figure S9 .Figure S10 .
Figure S9.UV-vis absorption spectra of Fe-Bf complex in the presence of (A) anthracene, (B) 9-hydroxyfluorene, (C) catechol, and (D) salicylic acid along with controls of VOCs, VOCs + FeCl3 and FeCl3 solubilized in DCM displaying distinct absorption bands along with disappearance of complex absorption peak at 554 nm except for salicylic acid.

Figure S11 .
Figure S11.UV-vis absorption spectra of Fe-Bf complex in the presence of (A) 1-methyl naphthalene, and (B) fluoranthene along with controls of VOCs, VOCs + FeCl3 and FeCl3 solubilized in DCM.

Figure S12 .
Figure S12.UV-vis absorption spectra of the Fe-Bf complex in the presence of increasing concentrations of 2,3dihydrobenzofuran.

Figure
Figure S14: (A) FTIR spectra of FeCl3, Algae biochar and Algae biochar + FeCl3 displaying distinct stretching and vibration frequencies.(B) TGA-DSC and DTG comparison graphs for Algae biochar in the absence and presence of FeCl3 displaying unique thermal degradation pathway in the presence of FeCl3 suggesting physical adsorption of Fe species on the surface of biochar.

Figure S15 .
Figure S15.Calibration plot of FeCl3 displaying linear progression coefficient of 0.99557 with increasing concentration of Fe(III) ions.

Figure S16 :
Figure S16: TGA, DSC and DTG curves of (A) Algae biochar, (B) FeCl3, and (C) Algae biochar in the presence of FeCl3 displaying distinct thermal degradation pathways as well as characteristic heat flow signatures when heated up to 1000 °C at a ramp rate of 10 °C/min.

Table S4 .
Summary of the Selectivity Data for the Complex in Presence of Various VOCs Obtained Using UV-vis Absorption Spectroscopy.