Accepted Manuscript

Title: Alkaline hydrothermal stabilization of Cr(VI) in soil

using glass and aluminum from recycled municipal solid
wastes

Authors: Concetta Eliana Gattullo, Caterina DAlessandro,

Ignazio Allegretta, Carlo Porfido, Matteo Spagnuolo, Roberto
Terzano

PII: S0304-3894(17)30791-4
DOI: https://doi.org/10.1016/j.jhazmat.2017.10.035
Reference: HAZMAT 18940

To appear in: Journal of Hazardous Materials

Received date: 5-6-2017

Revised date: 16-10-2017
Accepted date: 17-10-2017

Please cite this article as: Concetta Eliana Gattullo, Caterina DAlessandro, Ignazio
Allegretta, Carlo Porfido, Matteo Spagnuolo, Roberto Terzano, Alkaline hydrothermal
stabilization of Cr(VI) in soil using glass and aluminum from recycled municipal solid
wastes, Journal of Hazardous Materials https://doi.org/10.1016/j.jhazmat.2017.10.035

This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.
Alkaline hydrothermal stabilization of Cr(VI) in soil using glass and aluminum from recycled
municipal solid wastes

Concetta Eliana Gattullo, Caterina D'Alessandro, Ignazio Allegretta, Carlo Porfido, Matteo
Spagnuolo, Roberto Terzano*

Department of Soil, Plant and Food Sciences, University of Bari, Via Amendola 165/A, 70126 Bari,
Italy

*corresponding author - e-mail: roberto.terzano@uniba.it; Tel. +39 080 5442852; Fax +39 080
5442850

Graphical abstract

Highlights

Glass and aluminium from municipal solid wastes can be used for soil remediation
Alkaline hydrothermal treatment reduced Cr(VI) concentration in soil by 98% in 7 days
Cr was stabilized in geopolymeric alluminosilicate aggregates

1
Abstract

Hexavalent chromium was stabilized in soil by using a mixture of glass and aluminum recovered
from municipal solid wastes under alkaline hydrothermal conditions. Cr(VI) concentration was
reduced by 94-98% already after 7 days of treatment. After the same period, more than 90% of total
Cr was stabilized in highly recalcitrant and scarcely mobile chemical forms, with 50% in the
residual fraction (when the samples were treated at 1/10 w/w mixture/soil ratio). Longer treatments
increased Cr stabilization. X-ray microanalyses revealed that Cr was stabilized in geopolymeric
structures within large aluminosilicate mineral aggregates (containing both amorphous and
crystalline phases). 3D microstructural analyses showed a limited compaction of the soil with still a
20% internal porosity in the neoformed aggregates. Increased pH and salinity after the treatment
can be restored by simple soil amendments and washing.

Key words: Remediation, Soil pollution, Chromium, Geopolymers, X-ray microanalyses

1. Introduction

Hexavalent chromium (Cr(VI)) is one of the most dangerous pollutants in the environment,

being very toxic for all living organisms and, in particular, carcinogenic and genotoxic for humans

and animals [1]. Soil contamination with Cr(VI) depends mainly on anthropogenic activities, such

as the uncontrolled discharge of wastes produced by metallurgical, chemical and refractory

industry, as well as chromite ore mining and refining [2]. Most of the Cr(VI) compounds present in

these wastes are soluble in water, with only few exceptions such as zinc chromate (insoluble),

strontium chromate (slightly soluble) and lead chromate (very low solubility, 58 g L-1) [3]. Once

released in soil, Cr(VI) may be naturally reduced to the trivalent form, Cr(III), and vice versa. Other

oxidation states are very unstable in soil and, consequently, are not involved in Cr biogeochemistry.

The speciation of Cr in hexavalent and trivalent forms and the stability of these two forms

determine the level of Cr hazardousness in soil. Indeed, opposite physico-chemical properties and

biological effects characterize the two oxidation states. While Cr(VI) is mobile, bioavailable and

very toxic to living organisms, Cr(III) is low soluble, scarcely bioavailable, and low toxic [4].

Chromium speciation is modulated by several soil parameters, especially pH, redox potential,
2
content of manganese (Mn) and iron (Fe) oxides, and organic matter [2,5]. Moreover, it is also

affected by the concentration and chemical form of Cr(III) and Cr(VI) species occurring in soil [5].

High Cr(VI) concentrations may overcome the natural capacity of soils to attenuate the

contamination through Cr(VI) reduction, thus causing serious environmental implications.

Therefore, decontamination of Cr(VI) polluted soils may be a priority in highly polluted areas, such

as some industrial sites.

Several physical, chemical and biological clean-up strategies have been tested in the last

decades for Cr(VI) remediation, mainly based on reduction, adsorption, precipitation, extraction and

ion exchange processes [2]. Biological treatments, especially using reducing bacteria, are largely

applied. However, their efficiency can be strongly limited in highly polluted sites where the high

concentrations of Cr(VI), often coupled with the lack of nutrients, may suppress the growth and

activity of microorganisms [6,7].

Stabilization/solidification (S/S) is a rapid and cheap technology widely applied for the

decontamination of heavy metal polluted soils and the stabilization of hazardous wastes. S/S

treatments are usually based on the chemical transformation of the contaminant into a less mobile

and toxic form (stabilization), followed by the physical encapsulation of the pollutant in a solid

matrix (solidification), so as to minimize the risks of diffusion in the surrounding environment [8].

For this purpose, soil is treated with a single or a combination of active compounds reducing the

pollutant mobility. The remediation of Cr(VI) polluted soils by S/S is typically achieved by treating

the soil with chemical reducers (sulfide, thiosulfate, zerovalent iron, ferrous iron and organic

compounds), before or concomitantly with the addition of binding materials, like fly ash and clay

minerals [9-11]. In some cases, the reducing step can be skipped and soil is treated directly with

aluminosilicate materials under hyperalkaline conditions, sometimes at moderately high

temperatures. Such conditions may promote the dissolution of aluminosilicates and the formation of

new binding phases of amorphous or crystalline nature, which adsorb and/or entrap the pollutant.

3
Such approach has been reported for Cr stabilization in hazardous wastes [12-13], whereas only few

studies have evaluated its feasibility for Cr stabilization in soil [9].

Research for sustainable materials to employ in S/S treatments is a primary goal to reduce the

costs and minimize soil physicochemical alterations. Recently, Terzano et al. [14] suggested a

simple method to synthesize zeolites by using mixtures of glass and aluminum recovered from

municipal solid wastes, under alkaline hydrothermal conditions. Zeolites are widely used in S/S

treatments of heavy metal-polluted soils due to their ability to immobilize heavy metals through

cation exchange, adsorption, precipitation and encapsulation of the pollutant [15-17]. Some studies

have proved the efficiency of zeolites for the immobilization of Cr(VI) in soil [18] and, in general,

for the remediation of Cr(VI)-polluted soils [19-21].

Based on these premises, in this study a mixture of aluminum and glass from municipal solid

wastes was applied directly to a Cr(VI)-polluted soil under alkaline hydrothermal conditions and its

efficiency for the stabilization of Cr(VI) was evaluated. The potentialities of this technique as an

innovative S/S treatment were assessed by measuring Cr(VI) concentrations in soil during the

treatment as well as Cr mobility by means of sequential chemical extractions. In addition, chemical,

structural and mineralogical X-ray analyses were performed on soil samples to elucidate the

mechanisms of Cr stabilization. The combination of traditional analytical techniques with X-ray

investigations at microscopic level allowed a full insight into the dynamics of Cr stabilization in soil

and the immobilization mechanisms occurring during the S/S treatment.

2. Materials and methods

2.1. Soil characterization and contamination

An uncontaminated soil was sampled in a vineyard located in the South of Italy (Ginosa

Marina, Taranto), air dried, sieved at 2 mm and characterized for the main physico-chemical

parameters [22]. Major element content was determined by wavelength dispersive X-ray

fluorescence (XRF) spectrometry (Supermini 200, Rigaku Corporation) while trace element content
4
was measured by means of energy dispersive XRF (NITON XL3t GOLDD, Thermo Scientific).

Results of soil characterization and elemental composition are reported in Table 1.

Soil spiking was carried out by adding an aqueous solution of K2Cr2O7 ( 99.9%, Sigma

Aldrich) to the soil, in order to obtain a concentration of 1500 mg kg-1 Cr(VI). This value is

representative of Cr(VI) levels actually measured in highly polluted soils [23]. After the

contamination, the soil was aged for three months, under normal conditions of pressure and

temperature, until constant Cr(VI) concentration. During this period, soil was periodically

moistened with water (without exceeding the water holding capacity) and mixed up. Finally, it was

air-dried, sieved at 2 mm, and analyzed for Cr(VI) content in order to evaluate the level of

contamination achieved, which represented the starting concentration in remediation experiments.

2.2. S/S treatment

S/S treatment was performed in 500 mL high-density polyethylene (HDPE) vessels, treating the

contaminated soil (100 g) with a mixture (MIX) of glass and aluminum under alkaline hydrothermal

conditions. The MIX was prepared according to the procedure described in [14]. Briefly, white

glass, recovered from bottles and food jars, and aluminum from beverage cans were finely

pulverized with a planetary mill (Pulverisette 7, Fritsch) and mixed at a ratio of 16/1

(glass/aluminum, w/w). The MIX was added to the soil at two ratios, 1/10 and 1/20 (MIX/soil,

w/w), and mixed thoroughly along with pulverized KOH. The latter was used as mineralizing agent

at a final concentration of 5 M in the soil suspension. After the addition of 200 mL bidistilled water

to each soil sample, a highly exothermic reaction occurred with bubbles formation. Once the

effervescence stopped, soil suspensions were mixed again. Then the vessels were closed with screw

caps and divided in two sets, one stored at room temperature in the dark, and the other incubated in

oven at 90C. This last temperature was used since it was demonstrated to favor glass dissolution

and zeolite synthesis [14]. In order to assess the effect of sole KOH on Cr(VI) stabilization, a

treatment with KOH without MIX was carried out at both temperatures. In addition, the influence of

5
temperature alone on the fate of Cr(VI) in the polluted soil was evaluated by treating the soil with

bidistilled water at 90C (in the absence of MIX and KOH). All the treatments were replicated three

times. The relative standard deviation of analytical data (presented in Results and discussion

section) varied between 2% and 8%.

2.3. Soil analyses after S/S treatment

Ten grams of each treated soil were sampled after 1, 7, 30, 60 and 90 days, air-dried, gently

homogenized in a mortar and analyzed for chemical, mineralogical and structural parameters.

Chromium (VI) concentration was determined by alkaline digestion (USEPA Method 3060A

[24]) followed by colorimetric assay (USEPA Method 7196A [25]). Because of interferences

possibly connected to this assay [26], recovery tests from Cr(VI)-spiked soil were performed to

control the quality of results and correct Cr(VI) concentration values (Supplementary information).

Chromium chemical speciation in soil was assessed by a sequential extraction procedure

consisting of four steps. The first one extracted the exchangeable fraction of Cr by using 1M MgCl2

at pH=7 [27]. The remaining three steps were performed according to the BCR protocol [28] and

extracted the acid-soluble fraction (step 2) using 0.11 M acetic acid, the reducible fraction (step 3)

using 0.5 M hydroxylamine hydrochloride at pH=2, and the oxidisable fraction (step 4) using 1M

ammonium acetate at pH=2 (after two oxidation steps with H2O2). Total Cr concentration in soil

extracts was determined by inductively coupled plasmaatomic emission spectrometry (ICP-AES;

Thermo iCAP 6000 series, Thermo Fisher Scientific Inc.). Finally, the soil residue was air-dried and

analyzed by XRF to determine the Cr residual fraction.

Additional mineralogical, chemical and structural X-ray analyses were performed on selected

soil samples. X-ray diffraction (XRD) analysis was performed with a Miniflex II (Rigaku

Corporation) X-ray diffractometer. Soil thin sections were analyzed by micro X-ray fluorescence

(XRF; M4 Tornado, Bruker Nano GmbH) and by scanning electron microscopy coupled with

microanalysis (FEGSEM-EDX, Zeiss igma SUPRA 300 VP, Carl Zeiss NTS GmbH). The internal
6
structure and macro/micro porosity of soil aggregates formed during the S/S treatment were

investigated by high resolution micro X-ray computed tomography (CT, SkyScan 1272, Bruker

Gmbh). Additional information about the analytical conditions adopted for X-ray analyses are

reported as Supplementary information.

3. Results and discussion

3.1. Cr(VI) concentration

Total Cr concentration in uncontaminated soil was below XRF detection limit (<25 mg kg-1;

Table 1). After soil spiking with 1500 mg kg-1 Cr(VI) and aging for three months, the concentration

of Cr(VI) decreased to 579 mg kg-1 because of natural reduction of Cr(VI) to Cr(III). The

attenuation of Cr(VI) concentration was not ascribable to Cr(VI) leaching, since soil vessels were

closed at the bottom and total Cr concentration was confirmed by XRF analysis. Reduced organic

matter, Fe(II), Mn(II) and sulfides may act in soil as electron donors and, consequently, drive

Cr(VI) reduction [29]. Soil microorganisms may also enhance the intrinsic decontamination

capacity of the soil by reducing Cr(VI) both directly and indirectly [7].

All soil treatments at 90C caused a considerable reduction of Cr(VI) concentration already

after 1 day (Fig. 1). In particular, the hydrothermal treatment with only water at 90C was able to

reduce Cr(VI) concentration by 73%, the alkaline hydrothermal treatment (KOH at 90C) by 48%,

and the alkaline hydrothermal treatments with MIX by 87% (MIX 1/20) and 82% (MIX 1/10). After

7 days, Cr(VI) concentration was further reduced in all samples treated at 90C, especially in those

added with MIX 1/20 (98% reduction) and MIX 1/10 (94% reduction). The extension of the

treatment up to 90 days produced a further slow decrease of Cr(VI) concentration in almost all

samples, with 98% of Cr(VI) reduced in all treatments, except for the water treatment, where the

reduction accounted to 91%.

Unlike the treatments at 90C, the treatments at room temperature caused only a moderate

reduction of Cr(VI), which did not exceed 38% in samples treated with MIX at both concentrations,
7
and 15% in samples treated with sole KOH (Fig. 1). At this temperature, a slight tendency to Cr re-

oxidation also occurred in some samples during the process, as revealed by the fluctuating trends of

Cr(VI) concentration in the 90 days of soil treatment.

These results revealed that curing soil at high temperature was crucial to reduce Cr(VI) in soil,

however only the association of high temperature and alkaline conditions caused an almost

complete Cr(VI) decontamination, which was further accelerated in the presence of MIX.

Similarly, Qafoku et al. [30] observed a significant reduction of Cr(VI) concentration in a

sediment-aqueous phase system under hyperalkaline and saline conditions at a temperature of 50C,

and attributed it to the reducing activity of Fe(II) released by the base-induced dissolution of

sediment minerals. Also soil organic matter has been reported to promote Cr(VI) reduction in soil at

80-105C [31]. In our study, hydrogen gas was produced during aluminum dissolution [14]

(evidenced by a strong effervescence at the beginning of the treatment) thus strongly increasing the

reducing potential of the solution when MIX was applied. In addition, other elements released by

KOH-dissolution of MIX may have concurred to the reduction of Cr, thus contributing to their

higher decontamination efficiency compared to S/S treatments without MIX. In particular,

aluminum used for the preparation of MIX contained 1.0% Fe and 2.1% Mn [14], which were

possibly involved in redox reactions. Further details about Cr(VI) reduction process are described in

Section 3.4.

3.2. Cr mobility

Results of sequential chemical extractions for total Cr from samples treated at 90C are

presented in Fig. 2. Before the treatment (Time 0), most of the Cr was extracted from the soil after

the first (59%) and second step (24%), while only 11% was extracted in the other two steps and 6%

remained in the residue. This result suggested that, although about two thirds of Cr(VI) in the

untreated spiked soil were "naturally" reduced to the less toxic Cr(III) form during soil aging, the

mobility of Cr was still high. Indeed, only a minor part of Cr was stabilized in the most recalcitrant

8
soil fractions, possibly precipitated as oxides and hydroxides or strongly bound to organic matter.

After one day of treatment at 90C, the percentage of Cr extracted in the first two steps was still

quite high for the treatments without MIX and accounted to 64% (water at 90C) and 75% (KOH at

90C). Differently, in the samples added with MIX, the amount of Cr extracted in the first two steps

was only 9% for the MIX 1/20 and 22% for the MIX 1/10, and decreased almost to zero after seven

days. In the two alkaline hydrothermal treatments with MIX, the amount of Cr in the most

recalcitrant and unavailable fractions (fourth step and residue) reached a maximum of about 98%

after 90 days, being 68% in the residual fraction. For comparison, after 90 days, the amount of Cr in

the same fractions was 69% for the hydrothermal treatment without KOH and 80% for the alkaline

hydrothermal treatment with sole KOH, but being respectively just 44% and 31% in the residual.

With regard to soil treated at room temperature, sequential extractions revealed that most of the

Cr remained quite available in all samples throughout the experiment (Supplementary information,

Fig. S.2).

Some criticisms were evidenced by a number of authors about the inaccuracy of BCR

sequential extractions for the speciation of Cr and other redox-sensitive elements [32-33]. In

particular, a biased distribution of Cr between the reducible and oxidisable fractions may occur.

Despite of this, sequential chemical extractions are widely considered a valid tool to understand the

partitioning of elements, including Cr, during a stabilization treatment [33]. In the samples treated

with MIX under alkaline hydrothermal conditions, the rapid and remarkable reduction of the Cr

extractable in the first two steps and the progressive and substantial increase of Cr in the residual

fraction were indicative of Cr stabilization. This immobilization began concomitantly with the

drastic decrease of Cr(VI) concentration (Fig. 1), and was even enhanced after the almost complete

reduction of Cr(VI) in soil. Therefore, the stabilization of Cr was reasonably ascribable, in the first

days, to the reduction of Cr(VI) to the less mobile trivalent form and, then, to the immobilization of

Cr(III) species in stable minerals. Conversely, the mobility of Cr remained always high in all soil

samples treated at room temperature, in agreement with the scarce reduction of Cr(VI) measured in
9
these samples. The maximum Cr immobilization was registered in soil treated with MIX 1/10 under

alkaline hydrothermal conditions, with 50% Cr measured in the residual fraction already after 7

days of treatment. However, after 90 days, also the treatment with MIX 1/20 gave results

comparable to those with MIX 1/10. Possibly, silicon and aluminium released by the MIX

components enhanced the formation of new binding phases able to entrap Cr and, consequently,

reduce its availability (see Section 3.4).

3.3. X-ray analyses and microanalyses

Mineralogical analyses were carried out on selected soil samples characterized by low

concentrations of Cr(VI) and reduced Cr mobility after the alkaline hydrothermal treatment and, for

comparison, on the untreated soil. The latter showed a very high content of amorphous phases

(45%) and quartz (31%), and a moderate content of clay minerals, specifically illite (14%) (Fig. 3).

The alkaline hydrothermal treatments caused significant alterations of soil mineralogy. In particular,

quartz was dissolved during the treatments and, after 90 days, it was reduced to 6.5% in the

treatment with MIX 1/20 and to 2% in the other two treatments (Fig. 3). An analogous trend was

observed for illite, which was almost completely dissolved already after 30 days, independently of

MIX addition. In a similar way, amorphous components were initially dissolved, although their

variation after 30 days was masked in the treatments with MIX (also containing amorphous

material), especially at the higher dose. However, after 90 days of treatment, the level of amorphous

phases was similar to the initial value in all samples. Simultaneously with dissolution processes,

new crystalline phases formed. A K-zeolite, edingtonite, was detected in relevant amounts in

samples treated with KOH alone and, only after 90 days, in soil treated with KOH and MIX 1/20.

The formation of edingtonite in these experimental conditions was in agreement with Terzano et al.

[14]. No zeolites formed by treating the soil with MIX 1/10. Alkaline hydrothermal treatments

promoted also the formation of kalicinite (KHCO3) in almost all samples. This mineral possibly

formed from the reaction of CO2 (from the atmosphere and soil organic matter mineralization) in
10
the highly alkaline environment, as well as from carbonates already present, at undetectable

concentrations, in the amorphous fraction. Plagioclase and K-feldspar, already present in the

untreated soil, were subjected to moderate variations during the treatments. In conclusion, XRD

analyses revealed that the high temperature and alkaline conditions adopted promoted the

dissolution of some mineral phases and the formation of new secondary minerals, in accordance

with findings reported in previous studies on the weathering of soil and sediment minerals

[30,34,35].

Micro XRF analyses were performed on the same samples subjected to XRD analysis, and the

corresponding maps are reported in Fig. 4. In the untreated soil, Cr was randomly distributed with

isolated small hotspots (Fig. 4a). The hydrothermal alkaline treatment without MIX caused the

alteration of soil structure and the redistribution of Cr, especially after 90 days (Fig. 4d). As can be

observed in Fig. 4d, the intense hotspots of Cr were no longer visible but larger areas of lower Cr

concentration appeared. These areas, large few millimeters, contained also Si and Al (Fig. 4e and f).

Alkaline hydrothermal treatments with MIX showed similar results as those obtained with KOH

(Fig. 4g-i), but the process was faster and large Cr-Si-Al aggregates appeared already after 30 days

(data not shown). After 90 days, the number and size of aggregates in soil treated with MIX

appeared larger compared to the soil treated for 90 days without MIX, in the presence of KOH (Fig.

4g-i). Conversely, in the untreated soil, no Cr-Si-Al aggregates were visible, but Si and Al were

randomly dispersed in the map and not associated to Cr (Fig. 4b and c). In order to elucidate the

possible relations between Cr, Si and Al in the treated samples, correlation diagrams of Cr-Si and

Cr-Al were calculated on the entire map area. No evident correlation was found among these

elements, even in the samples where the aggregates were numerous and large (i.e., treatment with

KOH alone or KOH plus MIX 1/10 for 90 days) (Supplementary information, Fig. S.3). Based on

XRF results, it can be assumed that Cr stabilization occurred along with the formation of

aluminosilicate aggregates which likely adsorbed or entrapped the pollutant. These aggregates,

11
according to -XRF and XRD analyses, were not constituted by a single mineral, but they were

possibly a combination of different crystalline and amorphous phases.

To better elucidate the nature of these aggregates, soil thin sections of samples treated for 90

days under alkaline hydrothermal conditions with only KOH or with KOH and MIX 1/10 were

analyzed by FEGSEM-EDX. In the sample treated with KOH and MIX 1/10, Cr was homogenously

dispersed in the large aggregates and its signal was very low, often below the detection limit. The

analysis at different resolutions of one of the large aggregates (evidenced with a circle in Fig. 4h

and whose image at the optical microscope is shown in the Supporting information Fig. S5),

allowed to clarify the nature of this type of formations and the distribution of Cr inside these

structures (Fig. 5). From a morphological point of view, the aggregate was composed of large

compact mineral structures, along with small fragmented minerals and a binding phase (Fig. 5a and

f). The chemical analyses revealed that the large and compact mineral formations were silicates,

mostly aluminosilicates (Fig. 5b-e). Potassium was distributed over the entire aggregate (Fig. 5d),

and appeared the main component of the binding phase. The distribution of Cr in the aggregate

appeared homogenous on the particles surfaces (Fig. 5e), however, being the Cr signal very low,

single spectra with longer acquisition time were collected in different points of the aggregate (Fig.

5f) to assess if Cr was associated to the more compact formations or to the fragmented ones. Results

of single spectra revealed that Cr was slightly more concentrated in the large compact structures,

compared to fragmented ones. In addition, the correlations between Cr and other elements were

calculated using 14 points of analysis collected from three different aggregates. Significant

correlations were found between Cr and Si (r=0.77, P 0.01), and Cr and Al (r=0.56, P 0.05)

(Supplementary information, Fig. S.4), whereas no correlation was observed between Cr and K, Ca,

Fe and Mn. These results further suggest that Cr was stabilized in newly-formed aluminosilicate

structures after the dissolution of glass and aluminum (along with other soil minerals) under the

alkaline hydrothermal conditions.

12
 A different Cr distribution was observed in the sample treated with only KOH. Here, along

with Cr dispersions in the large aggregates (similarly to what observed in the sample treated with

MIX), small isolated hotspots of Cr at very high concentration (about 6.5%) were observed. The

associations of Cr with other elements (Si, Al, K, Fe, Mg, Ca) changed depending on the spot

considered, without any clear correlation (data not shown). The presence of high concentrations of

Cr not associated to the aluminosilicate neoformations in these samples could explain the higher

availability of Cr, as observed with sequential extractions.

The 3D structure of one of these soil aggregates after 30 days of treatment with KOH and MIX

1/10 was visualized by CT (Supplementary information, Fig. S.6 and video). As can be seen from

the micrograph of a sagittal section (Fig. 6a), the aggregate possessed an intrinsic porosity, which

was estimated as 20% of the total volume. The thresholding of a volume of interest (VOI, Fig. 6b)

representative of the entire aggregate evidenced the presence of portions characterized by different

X-ray absorption coefficients (Fig. 6c). The fraction characterized by the lowest absorption values

was attributed to the pore spaces (shown in blue in Fig. 6c), while the fraction corresponding to the

highest absorption values was attributed to aluminosilicates and other mineral phases (shown in red

in Fig. 6c). The portion corresponding to the intermediate range of absorption values, which was

estimated as 2% of VOI (shown in yellow in Fig. 6c), was possibly attributed to the binding phase,

being present mainly at the interface between mineral particles and pores.

Microanalyses also revealed that the synthesis of zeolites was not directly related to Cr

stabilization, differently from what observed for other similar S/S processes for Cu and Cd [17,36].

Indeed, the formation of zeolite edingtonite occurred mainly in the treatment with KOH alone,

which was the least efficient among the alkaline hydrothermal treatments tested, whereas zeolite

was not detected in soil treated with KOH and MIX 1/10, which produced the highest Cr

stabilization. Also kalicinite was not involved in Cr stabilization, as proved by its decrease in

samples treated with KOH alone and KOH with MIX 1/20, between 30 and 90 days of treatment

when, conversely, sequential extractions revealed an increase of Cr stabilization. In addition,

13
chemical maps did not show a correlation between K and Cr in the newly formed aggregates. In

general, no correlation can be found between Cr stabilization and any new crystalline phase formed

during the S/S process. Therefore, it is more likely that the stabilizing phases are amorphous

aluminosilicates with a geopolymeric structure [12,37].

Beside stabilizing Cr, the alkaline hydrothermal treatment significantly modified some soil

properties, such as pH, electrical conductivity (EC), cation exchange capacity (CEC) and texture.

On average, after the treatment with MIX, soil pH increased to about 10.0, EC to 30 dS m-1, CEC to

30 cmol(+) kg-1 and the amount of silt fraction increased by almost 100%, with a consequent

reduction of sand, while the clay fraction remained substantially unaltered. Total organic carbon

reduced by about 20%, while C/N ratio did not change significantly; available P increased to about

70 g kg-1. Therefore, the most relevant problems for a reuse of the treated soil may derive from the

high pH and salinity. Anyway, from our preliminary experiments, we observed that pH can be

reduced to about 8.0 by treating the soil with 100 ml kg-1 H2SO4 2M and 1.1 g kg-1 CaSO4, while

the salinity can be reduced by repeatedly washing the soil with water after pH correction.

3.4. Mechanism of Cr(VI) stabilization

In order to describe the mechanism of Cr(VI) stabilization, the following steps are

hypothesized:

1) after addition of water to the soil/MIX mixture, aluminum from beverage cans is dissolved in the

alkaline KOH solution by the following fast reaction [38]:

3
() + + 32 ()
4 + 2 2() (a)

2) At the same time, glass starts dissolving with the following slow reaction [39]:

2() + + 2 3 4 (b)

this reaction is slow at ambient temperature, but is faster at higher temperatures (i.e. 90C).

14
3) Al and Si in solution react to form a gel from which geopolymers are successively synthesized

[39]:

() +
4 + 3 4 + + [ (2 ) (2 ) 2 ]() (c)

where , , , , , are generic stoichiometric coefficients, depending on the final

composition of the geopolymeric gel.

4) Hydrogen gas generated by reaction (a) can directly reduce Cr(VI) to Cr(III) through the

following reaction:

242 + 32() + 22 2()3() + 4 (d)

but H2 can also react with other compounds in soil, acting as a reducing agent. For example it can

reduce soil organic matter (which is strongly solubilized and hydrolized by the high alkaline

conditions) as well as iron and manganese compounds (e.g., Fe2O3, MnO2). On their turn, all these

reduced compounds can react with Cr(VI) and reduce it to Cr(III) [2].

5) At the pH of the treatment, once Cr(VI) is reduced to Cr(III), it quickly precipitates as hydroxide.

Chromium hydroxide can then react with the geopolymeric gel (synthesized with reaction (c)),

forming microscopic and/or nanoscopic clusters included in the geopolymeric matrix, as already

described for other heavy metals by Terzano and Spagnuolo [40].

6) If not stabilized within the geopolymeric matrix, Cr(III) can be reoxidized to Cr(VI) by the

following reaction:

4()3() + 32() + 8 442 + 102 (e)

At room temperature, it is likely that once reduced, Cr(VI) is not properly stabilized by inclusion in

geopolymers and, therefore, it can be then partly reoxidized. This might explain why only about 30-

40% of Cr(VI) was reduced at this temperature, as well as the fluctuating trend observed over time

(Fig. 1). At 90C, a higher amount of geopolymers formed because of the higher dissolution rate of

glass and other soil aluminosilicates (limiting step) at this temperature. In addition, also all the

reactions leading to Cr(VI) reduction are accelerated by higher temperatures, thus quickly lowering

15
Cr(VI) concentration. At last, the processes of geopolymers formation and consolidation are faster

at higher temperatures, resulting in a more efficient solidification/stabilization.

EXAFS studies conducted on Cu stabilized in an aluminosilicate matrix originating from

silicates and aluminates pure chemical reagents in alkaline solution at 60C, revealed the presence

of different Cu phases even at the short-range molecular level and suggested the presence of nano-

or micro-sized Cu hydroxides precipitates trapped in the aluminosilicatic matrix, with some of these

occlusions even bound to the aluminosilicate framework [41]. The same behavior has been

suggested for other heavy metals [40] and can therefore be hypothesized also for Cr(III).

Additionally, this general behavior can be useful to stabilize in the geopolymeric matrix other

potentially toxic metals likely to be present in the soil or in the reaction mixture.

As described by Terzano et al. [14], in the view of possible applications of the treatment, the

issue of maintaining high temperatures in the reaction system can be partly overcome by exploiting

the high exothermic reactions at the beginning of the treatment (KOH and Al dissolutions), which

increase the soil temperature to about 65C, and by using the hydrogen gas generated during Al

dissolution for energy production [42]. An adiabatic reactor as well as a device to collect and store

excess hydrogen gas (e.g. for use in fuel cells) might be two solutions to keep the costs of the

treatment low.

In this study, the treatment was applied to a soil polluted with a soluble Cr(VI) salt but it is

likely to be applicable also to soils polluted by less soluble Cr(VI) compounds since the alkaline

hydrothermal conditions can solubilize almost all Cr(VI) compounds [24], thus making them

available for chemical reduction. In fact, it has been demonstrated [43] that heated alkaline

solutions can extract from soil all Cr(VI) forms, including PbCrO4, with the only exception of

BaCrO4, which is usually solubilized by 60-80%, depending on soil type. Therefore, the method

proposed in this paper should be able to solubilize almost all Cr(VI) from soil (probably only with a

not complete extraction of BaCrO4), thus making it available for subsequent reduction to Cr(III),

precipitation and stabilization within the geopolymeric matrix.

16
4. Conclusions

Soil treatments with a mixture of recycled glass and aluminum (MIX) caused a quick

reduction of Cr(VI) concentration already after 1 day, reaching 94% and 98% after 7 days for the

treatments with MIX 1/10 and MIX 1/20, respectively. Both treatments increased the amount of Cr

in the more stable soil phases already after 7 days, but MIX 1/10 was more efficient in sequestrating

Cr in the more recalcitrant residual fraction. Hexavalent chromium stabilization occurred by a

combination of chemical reduction (fast) and Cr immobilization (slower) within large

aluminosilicate aggregates. A treatment period of 7-30 days is sufficient to obtain very good results.

Longer incubation periods improve Cr stabilization. Some soil properties are impacted by this S/S

method, especially pH and salinity, but the soil still maintains a good structure (with large pores

visible in the formed aggregates). Simple and cheap soil amendments can restore the altered soil

features for a safe soil reuse.

Acknowledgments

The work was supported by the Research Programme FutureInResearch (Regione Puglia, Italy).

X-ray analyses were performed at the Micro X-ray Lab of the University of Bari.

17
References

[1] IARC (International Agency for Research on Cancer), A review of human carcinogens. Part C:

Arsenic, metals, fibres and dusts / IARC Monographs on the Evaluation of carcinogenic risks to

humans, Vol. 100C, IARC Press, France, 2012.

[2] B. Dhal, H.N. Thatoi, N.N. Das, B.D. Pandey, Chemical and microbial remediation of

hexavalent chromium from contaminated soil and mining/metallurgical solid waste: A review, J.

Hazard. Mater. 250251 (2013) 272 291.

[3] D.G. Barceloux, Chromium, Clin. Toxicol. 37 (1999) 173-194.

[4] K.R. Cook, In situ treatment of soil and groundwater contaminated with chromium, EPA/625/R-

00/005, Ohio, 2000.

[5] R.J. Bartlett, Chromium redox mechanisms in soils: should we worry about Cr(VI)?, in: S.

Canali, F. Tittarelli, P. Sequi (Eds.), Chromium Environmental Issues, FrancoAngeli s.r.l., Italy,

1997, pp. 1-20.

[6] M. Megharaj, S. Avudainayagam, R. Naidu, Toxicity of hexavalent chromium and its reduction

by bacteria isolated from soil contaminated with tannery waste, Curr. Microbiol. 47 (2003) 51

54.

[7] C. Viti, L. Giovannetti, Bioremediation of soils polluted with hexavalent chromium using

bacteria: a challenge, in: S.N. Singh, R.D. Tripathi (Eds.), Environmental Bioremediation

Technologies, Springer-Velrag, Germany, 2007, pp. 57-76.

[8] C.N. Mulligan, R.N. Yong, B.F. Gibbs, Remediation technologies for metal-contaminated soils

and groundwater: an evaluation, Eng. Geol. 60 (2001) 193207.

[9] D. Dermatas, X. Meng, Utilization of fly ash for stabilization/solidification of heavy metal

contaminated soils. Eng. Geol. 70 (2003) 377394.

18
[10] L. He, M. Wang, G. Zhang, G. Qiu, D. Cai, Z. Wu, X. Zhang, Remediation of Cr(VI)

contaminated soil using long-duration sodium thiosulfate supported by micronano networks, J.

Hazard. Mater. 294 (2015) 6469.

[11] S.B.K. Kanchinadham, L.M. Narasimman, V. Pedaballe, C. Kalyanaraman, Diffusion and

leachability index studies on stabilization of chromium contaminated soil using fly ash, J.

Hazard. Mater. 297 (2015) 52-58.

[12] J. Zhang, J.L. Provis, D. Feng, J.S.J. van Deventer, Geopolymers for immobilization of Cr6+,

Cd2+, and Pb2+, J. Hazard. Mater. 157 (2008) 587598.

[13] J. Chen, Y. Wang, S. Zhou, X. Lei, Reduction/immobilization processes of hexavalent

chromium using metakaolin-based geopolymers, J. Environ. Chem. Eng. 5 (2017) 373380.

[14] R. Terzano, C. DAlessandro, M. Spagnuolo, M. Romagnoli, L. Medici, Facile zeolite

synthesis from municipal glass and aluminum solid wastes, CLEAN 43 (2015) 133140.

[15] W. Shi, H. Shao, H. Li, M. Shao, S. Du, Progress in the remediation of hazardous heavy metal-

polluted soils by natural zeolite, J. Hazard. Mater. 170 (2009) 16.

[16] R. Terzano, M. Spagnuolo, L. Medici, F. Tateo, P. Ruggiero, Zeolite synthesis from pre-treated

coal fly ash in presence of soil as a tool for soil remediation. Appl. Clay Sci. 29 (2005) 99110.

[17] R. Terzano, M. Spagnuolo, L. Medici, B. Vekemans, L. Vincze, K. Janssens, P. Ruggiero,

Copper stabilization by zeolite synthesis in polluted soils treated with coal fly ash. Environ. Sci.

Technol. 39 (2005) 62806287.

[18] Z. Ioannou, A. Molla, A. Dimirkou, Remediation of polluted with chromium waters and soils

cultivated with wheat (Triticum durum) using zeolites modified with iron oxide. Water Air Soil

Pollut. 227 (2016) 116.

[19] V. Antoniadis, T. Polyzois, E.E. Golia, S.A. Petropoulos, Hexavalent chromium availability

and phytoremediation potential of Cichorium spinosum as affect by manure, zeolite and soil

ageing, Chemosphere 171 (2017) 729-734.

19
[20] R.B. Fu, F. Liu, C.B. Zhang, J. Ma, Effects of permeable reactive composite electrodes on

hexavalent chromium in the electrokinetic remediation of contaminated soil, Environ. Eng. Sci.

30 (2013) 17-22.

[21] B. Fonseca, M. Pazos, T. Tavares, M.A. Sanromn, Removal of hexavalent chromium of

contaminated soil by coupling electrokinetic remediation and permeable reactive biobarriers,

Environ. Sci. Pollut. Res. 19 (2012) 1800-1808.

[22] D.L. Sparks, Methods of soil analysis - Part 3 Chemical Methods, SSSA Book Series No 5.

Soil Science Society of America and American Society of Agronomy, Wisconsin, 1996.

[23] M. Chrysochoou, D.R. Ferreira, C.P. Johnston, Calcium polysulfide treatment of Cr(VI)-

contaminated soil, J. Hazard. Mater. 179 (2010) 650657.

[24] USEPA (United States Environmental Protection Agency), Method 3060A, Alkaline digestion

for hexavalent chromium, Washington, D.C. 1996.

[25] USEPA (United States Environmental Protection Agency), Method 7196A, Chromium,

hexavalent (colorimetric), Washington, D.C. 1992.

[26] M. Pettine, S. Capri, Digestion treatments and risks of Cr(III)-Cr(VI) interconversions during

Cr(VI) determination in soils and sediments a review, Anal. Chim. Acta 540 (2005) 231-238.

[27] A. Tessier, P.G.C. Campbell, M. Bisson, Sequential extraction procedure for the speciation of

particulate trace metals, Anal. Chem. 51 (1979) 848-851.

[28] A. Sahuquillo, J.F. Lpez-Snchez, R. Rubio, G. Rauret, R.P. Thomas, C.M. Davidson, A.M.

Ure, Use of certified reference material for extractable trace metals to assess sources of

uncertainty in the BCR three-stage sequential extraction procedure, Anal. Chim. Acta 382

(1999) 317-327.

[29] C.D. Palmer, P.R. Wittbrodt, Processes affecting the remediation of Cr contaminated sites,

Environ. Health Perspect. 92 (1991) 2540.

[30] N.P. Qafoku, C.C. Ainsworth, S.M. Heald, Cr(VI) fate in mineralogically altered sediments by

hyperalkaline waste fluids, Soil Sci. 172 (2007) 598-613.

20
[31] Y.L. Wei, H.F. Hsieh, Y.S. Peng, K.W. Chen, C.Y. Lin, H.P. Wang, Chromium speciation in

mildly heated Cr(VI)-doped latosol soil, J. Synchrotron Radiat. 17 (2010) 173-178.

[32] E. Fernndez, R. Jimnez, A.M. Lallena, J. Aguilar, Evaluation of the BCR sequential

extraction procedure applied for two unpolluted Spanish soils, Environ. Pollut. 131 (2004) 355-

364.

[33] N. Papassiopi, A. Kontoyianni, K. Vaxevanidou, A. Xenidis, Assessment of chromium

biostabilization in contaminated soils using standard leaching and sequential extraction

techniques, Sci. Total Environ. 407 (2009) 925-936.

[34] N.P. Qafoku, C.C. Ainsworth, J.E. Szecsody, D.L. Bish, J.S. Young, D.E. McCready, O.S.

Qafoku, Aluminium effect on dissolution and precipitation under hyperalcaline conditions: II.

Solid phase transformations, J. Environ. Qual. 32 (2003) 2364-2372.

[35] Y. Deng, J.B. Harsh, M. Flury, J.S. Young, J.S. Boyle, Mineral formation during simulated

leaks of Hanford waste tanks, Appl. Geochem. 21 (2006) 13921409.

[36] R. Terzano, M. Spagnuolo, L. Medici, W. Dorrin, K. Janssens, P. Ruggiero, Microscopic

single particle characterization of zeolites synthesized in a soil polluted by copper or cadmium

and treated with coal fly ash, Appl. Clay Sci. 35 (2007) 128138.

[37] P. Duxson, A. Fernndez-Jimnez, J.L. Provis, G.C. Lukey, A. Palomo, J.S.J. van Deventer,

Geopolymer technology: the current state of the art, J. Mater. Sci. 42 (2007) 29172933.

[38] T. Hiraki, A. Nosaka, N. Okinaka, T. Akiyama, Synthesis of zeolite-X from waste metals, ISIJ

Int. 49 (2009) 16441648.

[39] H. Xu, J.S.J. Van Deventer, The geopolymerisation of alumino-silicate minerals, Int. J. Miner.

Process. 59 (2000) 247-266.

[40] R. Terzano, M. Spagnuolo, Heavy metal stabilization by promoting zeolite synthesis in soil, in:

I. Sherameti, A. Varma (Eds.), Detoxification of Heavy Metals, Springer-Verlag, Germany,

2011, pp. 423-438.

21
[41] R. Terzano, M. Spagnuolo, L. Medici, F. Tateo, B. Vekemans, K. Janssens, P. Ruggiero,

Spectroscopic investigation on the chemical forms of Cu during the synthesis of zeolite X at

low temperature, Appl. Geochem. 21 (2006) 993-1005.

[42] T. Hiraki, S. Yamauchi, M. Iida, H. Uesugi, T. Akiyama, Process for recycling waste

aluminum with generation of high-pressure hydrogen, Environ. Sci. Technol. 41 (2007) 4454

4457.

[43] B.R. James, J.C. Petura, R.J. Vitale, G.R. Mussoline, Hexavalent chromium extraction from

soil: a comparison of five methods, Environ. Sci. Technol. 29 (1995) 2377-2381.

22
Figure captions

Fig. 1. Mean residual Cr(VI) concentration in soil during the stabilization treatment at room
temperature (dotted lines) or 90C (continuous lines) with only KOH, or KOH and a glass-
aluminum mixture (MIX) at two ratios (1/10 and 1/20, w/w). Soil hydrothermal treatment with only
water is also shown for comparison.

600
Cr(VI) concentration (mg kg-1)

500

400

300

200

100

0
0 1 7 30 60 90
Days
Water KOH KOH+MIX 1/20 KOH+MIX 1/10

Fig. 2. Percentages of Cr extracted from soil after each step of sequential extraction at different
incubation times of hydrothermal treatment (90C) with i) water (control), ii) only KOH, or iii-iv)
KOH and the glass-aluminum mixture (MIX) at two ratios. Step 1: exchangeable; Step 2: acid-
soluble; Step 3: reducible; Step 4: oxidizable.

23
 Water KOH
100% 100%
Extracted fraction (%)

Extracted fraction (%)

80% 80%

60% 60%

40% 40%

20% 20%

0% 0%
0 1 7 30 60 90 0 1 7 30 60 90
Days Days

KOH + MIX 1/20 KOH + MIX 1/10

100% 100%
Extracted fraction (%)

Extracted fraction (%)

80% 80%

60% 60%

40% 40%

20% 20%

0% 0%
0 1 7 30 60 90 0 1 7 30 60 90
Days Days

Step
Step 11 Step 2 Step 3 Step 4 Residue

Fig. 3. Soil mineralogical composition after 30 and 90 days of alkaline hydrothermal treatment at
90C with i) KOH, ii) KOH and MIX 1/20, iii) KOH and MIX 1/10. Mineralogy of untreated soil is
reported for comparison. Am: amorphous; Qtz: quartz; Pl: plagioclase; Ill: illite; Kal: kalicinite;
Kfs: K-feldspar; Edi: edingtonite; MIX: glass-aluminum mixture.

24
 Untreated KOH KOH+MIX 1/20 KOH+MIX 1/10

100%

80%
Wt (%)

60%

40%

20%

0%
0 30 90 30 90 30 90
Days

Am Qtz Pl Ill Kal Kfs Edi

Edi

Fig. 4. Micro-XRF maps of i) untreated soil (a-c), ii) soil treated for 90 days under alkaline
hydrothermal conditions with KOH (d-f), iii) soil treated for 90 days under alkaline hydrothermal
conditions with KOH and the glass-aluminum mixture (MIX) at 1/10 ratio (g-i). Chromium
distribution and its associations with Si or Al are shown. The soil aggregate within the white circle
in (h) was analyzed by FEGSEM-EDX.

25
 Untreated KOH KOH+MIX 1/10
a d g

b e h

c f i

26
Fig. 5 (a) FEGSEM-EDX image of a portion of a soil aggregate (white circle in Fig. 4h) formed
after 90 days of alkaline hydrothermal treatment at 90C with KOH and glass-aluminum mixture
(MIX) at 1/10 ratio. (b-d) Distribution maps of Si, Al, K and Cr. (f) Expanded view of the central
area of (a) (red rectangle) and points (red spots) where EDX spectra were collected.

27
Fig. 6. (a) Micro-CT image of a sagittal section of a soil aggregate formed after 30 days of alkaline
hydrothermal treatment with KOH and glass-aluminum mixture (MIX) at 1/10 ratio. A subvolume
(0.42 mm3) of the aggregate and the relative thresholding are reported in (b) and (c), respectively.
Different colors in (c) represent different phases: blue-pores, yellow-binding phase, red-soil
minerals.

a) b) c)

500 m

28
Table 1
Physico-chemical characteristics of the soil used in the experiments.

Texture pH (KCl) EC Corg Total N CaCO3 P 2O 5 CEC

(dS m-1) (g kg-1) (cmol(+) kg-1)

Loamy sand 7.7 1.7 4.4 0.7 11.0 61.6 11.9

Major and trace elements composition (mean values)

LOI Na2O MgO Al2O3 SiO2 K2O CaO TiO2 Fe2O3 Mn Cu Zn Cr Ni Cd

% mg kg-1

3.2 1.2 1.2 10.4 73.1 2.2 1.2 0.3 3.2 560 27 33 <LOD 77 23

LOI: loss on ignition

LOD: limit of detection (Cr: 25 mg kg-1)

29