Form of the Occurrence of Aluminium in Municipal Solid Waste Incineration Residue—Even Hydrogen Is Lost

Abstract

In the bottom ash (BA) of municipal solid waste incineration, the content of iron and aluminum is relatively high. The efficiency of eddy current extraction of non-ferrous metals (including aluminium) routinely used in incineration plants is limited. The determination of the form of occurrence of aluminium or aluminium-rich components in BA is important in terms of its recovery or utilisation. BA from a newly built incineration plant in Poland was analysed using chemical analysis, X-ray diffraction, optical microscopy, and scanning electron microscopy with chemical microanalysis. Samples of water-quenched BA were analysed. For comparison, a non-quenched sample (collected above a water tank) was analysed. The obtained results indicate that aluminium-rich components in BA are present in both the melt phase and quench phase. In the melt phase (glassy material), the content of aluminium is low (usually below 2 wt%). Aluminium-rich components present in glass, inherited after aluminium products are usually oxidised, and occur as platy or irregular forms. Aluminium components in the quench phase are significantly transformed with the common presence of Cl⁻ and SO₄²⁻ phases formed during reaction with the quench water. Secondary phases form simple or complex rims around metallic or slightly oxidised cores, of which the size is significantly reduced during transformations. The variety in the forms of aluminium occurrence in BA makes its recovery challenging and inefficient. The reduced content of metallic aluminium indicates that the potential for hydrogen generation of BA is low.

1. Introduction

Globally, aluminium is the second-most produced metal after iron (steel) and its demand is expected to continue growing [1,2]. The production of aluminium is energy-consuming (ca. 3.5% of global electricity) and is thus related to significant CO₂ emissions (ca. 1% of global greenhouse gas emissions are related to industrial and energy production processes) [2]. The recycling of aluminium is important not only because of the conservation of aluminium ore deposits. While the production of aluminium from scraps is related to much lower energy consumption and emission of CO₂, an additional advantage is strongly related to the purity of aluminium scraps [2,3,4].

The global metal content in solid waste is ca. 4% and varies by income level (from ca. 6% in high-income countries to 2% in middle- and low-income countries) and in terms of geography [4,5]. Aluminium as a component frequently found in everyday use products is commonly present in municipal solid waste and its content in waste is related to the effectiveness of curbside recycling. Aluminium in municipal solid waste can also be related to the production of various goods. When metallic aluminium is deposited in landfill sites, it can react with alkaline water to produce hydrogen. Heat production in this exothermic reaction can lower or stop microbial activity and ignite or pyrolyse municipal waste [6]. Hydrogen can then react with CO₂ to produce methane [7].

Aluminium for recycling can be collected at the pre-user stage or at the end-of-life stage of aluminium-containing products through curbside collection, the separation from waste, and from municipal waste incineration (MSWI) residues. Metals from MSWI bottom ash (BA) can be extracted using magnetic methods (ferrous metals) or eddy-current separation (non-ferrous metals including aluminium) [8,9]. The efficiency of using magnetic and eddy-current separation in MSWI BA is relatively low. The content of metals after processing using these methods remains significant. Only 30% of aluminium can be recovered from BA using these technologies [10].

The content of iron and aluminium in MSWI BA is usually high, but the ratio of their content differs significantly [11]. According to Biganzoli et al. [12], the content of ferrous and non-ferrous metals in the BA varies in the 7–15% and 1–2% range, respectively, with the domination of aluminium (ca. 60%) in non-ferrous metals’ fraction. MSWI BA with the dominance of aluminium over Fe is also noted [13,14]. The ratio of iron to aluminium in BA can decrease with the decreasing size fraction of the material [13].

Aluminium separation can be difficult because of the complex chemical form of occurrence and small size of aluminium-rich inclusions in slag. The effectiveness of aluminium extraction using eddy-current separation from MSWI BA is related to the grain size of the ash and is usually lower for more fine-grained ash [12]. According to Biganzoli and Grasso [12], the recovery of aluminium is related to the structure and mechanical properties of the BA material.

Difficulty with the separation of metallic aluminium from MSWI BA suggests other applications of its potential. Hydrogen generation from the metallic aluminium–water reaction in MSWI BA was postulated by several authors [15,16,17,18]. Hydrogen production in the reduction of water using aluminium or its alloys is often considered as a possibility for energy generation [19]. Published data related to the efficiency of hydrogen production during BA aluminium–water reaction differ significantly. Biganzoli et al. [20] report hydrogen production from 5.7 to 10.8 L/kg of dry BA according to the sample. Larsson [20] discusses the role of numerous factors in the process and his results are from 0 to ca. 5 L of hydrogen per 1 kg of ash. A higher value of 39.4 l per 1 kg of BA was obtained by Saffarzadeh et al. [17]. Hydrogen can also be generated from MSWI fly ash’s reaction with water [21] or as a result of photocatalytic reactions [22], as shown in Rumayor et al. 2022.

The spontaneous reaction of aluminium-containing MSWI ash with water can result in the uncontrolled accumulation of hydrogen (or hydrogen-containing gas mixture) and explosions in ash bunkers [20,21,23]. Aluminium in MSWI residue can reduce its use as an addition to concrete owing to the production of hydrogen, which lowers the mechanical properties of concrete [13,16,24,25].

The aim of this paper is to present the form of occurrence of aluminium in MSWI BA from incineration plants in Poland built in the second decade of the 21st century and to discuss the possibility of its utilisation. The form of occurrence and degree of transformation of metallic aluminium determine the potential for its extraction or hydrogen production. The study is based on a detailed phase analysis of BA.

2. Materials and Methods

Several samples of BA were collected from an MSWI plant in Poland. The sampling details are listed in

Table 1. List of the studied samples.

Sample	Sampling Site
RZ1	BA from bunker
RZ2	BA after crushing
RZ3	BA after ferrous metals’ separation
RZ4	BA after non-ferrous metals’ separation
RZ5	BA after sieving (0–16 mm)
RZ6	BA after sieving (16–63 mm)
RZ7	BA—fine-grained material mixed from two lines; three months of ageing
RZ8	BA—coarse-grained material mixed from two lines; three months of ageing
RZ9	BA—fresh material collected from conveyor belt
RZ10	BA after (0–16 mm); two months of ageing
RZ11	BA—slag collected without quenching
RZ12	BA from bunker
RZ13	BA from bunker
RZ1W–RZ11W	Samples RZ1–RZ11 after hand picking of glass, ceramics, and metallic fragments

. Samples were collected in two cycles and several samples from each cycle were studied.

The chemical composition of the BA samples was analysed using inductively coupled plasma mass spectrometry and inductively coupled plasma atomic emission spectroscopy, performed by Bureau Veritas Minerals in Vancouver, Canada. Loss of ignition (LOI) was determined by the mass difference after heating to 1000 °C. Sulphur and carbon content were determined using the Leco method.

X-ray diffraction was used for the mineralogical characterisation of the BA samples. The analyses were performed at the Institute of Geological Sciences, Jagiellonian University in Krakow, Poland, using a Philips X’Pert APD type diffractometer with a PW 3020 vertical goniometer equipped with a curved graphite crystal monochromator (CuKα radiation, analytical range 2–64° 2Θ, step 0.02°, counting time 2 s/step) (Almelo, The Netherlands). Phase identification was obtained using Philips X’Pert software (associated with the ICDD database).

Optical microscopy was employed for the observation of the samples’ mineral composition and structure using transmitting and reflected light.

A field emission scanning electron microscope (FE-SEM, HITACHI S-4700; Japan) fitted with an energy-dispersive spectrometer (EDS, NORAN NSS; Madison, USA) at the Institute of Geological Sciences, Jagiellonian University in Krakow, Poland, was used for detailed analysis of the BA samples’ phase and chemical composition. Both rough and polished preparations were coated with carbon or gold. An accelerating voltage of 20 kV and current of 10 μA were used for imaging and chemical analysis. Secondary electrons (SEs) and backscattered electrons (BSEs) were utilised for imaging. For the chemical analysis in spots, a 100 s acquisition time and the standardless method of quantification were used. Mapping of the selected chemical elements’ distribution was performed for the imaging of aluminium-rich compounds and their reaction products’ distribution.

3. Results and Discussion

3.1. Chemical Composition of BA

The results of the analyses of samples (Table 1 and

Table 2. Chemical composition of BA samples (content of all components in wt%; except for Cu, Zn, and Pb, in mgkg⁻¹; LOI—loss of ignition; samples’ symbols—see Table 1).

Component	RZ1	RZ2	RZ3	RZ4	RZ5	RZ6	RZ7	RZ8	RZ9	RZ10	RZ11	RZ12	RZ13	RZ1W	RZ2W	RZ3W	RZ4W	RZ5W	RZ6W	RZ7W	RZ09W	RZ10W	RZ11W
SiO2	63.08	57.97	60.38	60.14	57.87	62.67	60.44	57.41	60.00	61.88	59.47	56.02	57.80	50.69	53.51	54.93	56.39	52.45	59.47	56.00	50.22	52.98	46.24
TiO2	0.32	0.63	0.54	0.53	0.51	0.49	0.50	0.55	0.59	0.52	1.03	0.67	0.60	0.69	0.82	0.78	0.73	0.71	0.68	0.65	1.03	0.85	2.39
Al2O3	4.87	6.67	6.69	5.97	5.74	8.38	5.75	8.27	6.42	5.38	9.03	9.00	5.30	11.15	9.12	7.90	7.31	8.02	7.00	7.62	10.37	9.03	14.48
Fe2O3	2.34	4.14	2.85	3.27	4.43	4.26	3.48	5.02	4.92	3.73	2.39	3.38	3.96	5.85	4.23	4.45	3.73	4.73	5.64	4.74	3.65	4.40	3.96
MgO	1.44	1.74	1.80	1.75	1.68	1.47	1.78	1.73	1.59	1.70	2.25	1.85	1.94	1.83	1.94	1.94	1.79	1.86	1.88	1.71	1.70	1.95	2.39
CaO	13.82	14.77	13.78	14.28	14.69	11.56	13.89	14.54	13.60	13.25	15.36	15.24	15.30	16.27	16.56	16.10	16.24	16.54	13.99	14.90	19.01	15.91	20.41
MnO	0.14	0.08	0.06	0.10	0.08	0.06	0.06	0.10	0.08	0.07	0.06	0.07	0.08	0.25	0.08	0.08	0.08	0.09	0.10	0.09	0.06	0.08	0.09
Na2O	9.39	6.66	7.63	7.35	7.32	6.95	7.33	3.87	7.68	8.10	7.59	7.34	7.63	5.10	4.81	5.36	5.83	5.10	7.52	5.57	4.36	4.74	3.25
K2O	0.85	1.03	1.04	0.97	0.98	1.19	0.92	1.18	1.00	0.98	1.13	0.92	0.85	1.06	1.09	1.13	1.01	1.07	0.94	1.02	1.11	1.17	1.26
P2O5	0.62	0.81	0.70	0.67	0.73	0.34	0.69	0.32	0.67	0.58	0.64	0.83	0.66	1.27	1.08	1.16	0.91	1.17	0.60	0.96	0.69	0.43	0.62
LOI	2.50	4.60	3.80	3.90	5.10	1.80	4.50	6.50	2.70	3.20	0.10	3.70	5.20	5.10	5.90	5.20	5.10	7.20	1.50	5.90	6.10	7.00	2.60
Total	99.52	99.36	99.48	99.17	99.34	99.41	99.50	99.63	99.37	99.52	99.34	99.27	99.49	99.38	99.36	99.28	99.36	99.08	99.52	99.30	99.34	99.30	98.75
C	0.64	1.11	0.93	1.00	1.35	0.49	1.15	2.20	0.63	0.94	0.72	0.99	0.84	1.58	1.53	1.21	1.21	1.64	0.50	1.38	1.82	2.40	0.64
S	0.51	0.59	0.51	0.52	0.55	0.29	0.47	0.25	0.30	0.31	0.33	0.47	0.49	1.05	0.79	0.70	0.69	0.79	0.33	0.59	0.65	1.03	0.59
Al	2.58	3.53	3.54	3.16	3.04	4.43	3.04	4.38	3.40	2.85	4.78	4.76	2.80	5.90	4.83	4.18	3.87	4.24	3.70	4.03	5.49	4.78	7.66
Fe	1.64	2.90	1.99	2.29	3.10	2.98	2.43	3.51	3.44	2.61	1.67	2.36	2.77	4.09	2.96	3.11	2.61	3.31	3.94	3.32	2.55	3.08	2.77
Cu	1017	1072	1016	1968	1156	358	451	939	1500	799	1135	1101	604	983	827	1121	711	1331	734	904	1209	1149	1948
Pb	132	686	253	893	436	111	365	181	434	240	311	624	322	1288	421	698	648	448	148	293	211	455	1626
Zn	470	929	970	2373	986	838	1232	282	1054	1003	1285	1269	1239	1707	1367	1370	1537	2201	990	1833	1574	1511	3168

) indicate variation in the chemical composition of the studied samples. In the chemical composition of the samples presented as the content of oxides, SiO₂ dominates in all samples (56–63 wt%), followed by CaO as the second-most-important component (11.56–15.30 wt%). The variation can be related to changes in composition in the waste stream, as well as the BA treatment (Table 1). The ratio of Al/Fe for all samples is greater than 1 (1.01–2.86), except for samples RZ5 and RZ9 at 0.98 and 0.99, respectively. The BA sample collected without quenching exhibits the lowest LOI value. Samples after three months of ageing are characterised by significantly higher LOI values. The removal of ferrous and non-ferrous metals results in a lowering of the content of iron and aluminium (Table 2). For Fe, the process seems to be more effective (see sample RZ3, Table 2).

The hand picking of fragmented glass, ceramics, and metallic fragments results in a decrease in the SiO₂ content and an increase in the Al₂O₃, Fe₂O₃, and CaO content (Table 1 and Table 2). These changes are related to the high frequency of glass (from 20 to over 52 wt%), ceramics (from 2.8 to 28 wt%), and metal fragments (from 0.6 to 3.9 wt%).

3.2. Mineral Composition of BA

Besides amorphous glassy material, the analysed BA contains quartz, wollastonite, diopside, calcite, anhydrite, mullite, and feldspar. Quartz and other minerals (e.g., zircon) are inherited from waste material, while others are formed in high-temperature processes.

MSWI BA usually contains two phases—the phase formed during the solidification of melt and the phase formed during the quenching of the slag in water tanks [17,23]. In the studied BA, both phases are present (Figure 1A). The quench phase occurs around the glassy material. The quench phase was also noticed in the RZ11 sample (collected without quenching), but its content is significantly lower (Figure 1B), which is also confirmed by the lower LOI value (Table 2).

Metallic fragments in the quench phase are surrounded by a reaction rim (Figure 1C,D). However, the reaction rim is absent where the metallic fragment is in contact with the glass phase (Figure 1D).

3.3. Form of Occurrence of Aluminium in BA

Aluminium is present in the studied samples in different forms. Below, we briefly present typical examples.

The content of aluminium in the glass phase in BA is commonly below 2 wt% (Figure 2A, Table 3), although higher values are also noted (ca. 4 wt%; Figure 2B, Table 3). Glass is dominated by silicon and calcium. Glass often contains detrital quartz grains, which are cracked during the rapid changes in temperature (Figure 2A), and rarely other detrital components (e.g., zircon, Figure 2A). Differences in the composition of the melt phase formed during waste incineration (and different aluminium content) are related to local variation in the waste composition.

Aluminium can be noted in glass in the form of remains inherited after different products (e.g., aluminium foil or other aluminium packing materials). Aluminium-rich phases of this type commonly occur as platy forms of different size, often below 50 μm (Figure 2C,D; Table 3). The glass phase between aluminium-rich elongated fragments is also relatively rich in aluminium (Figure 2C,D; Table 3). This could be related to local enrichment of the melt phase in aluminium or partly to the analytical ambiguity caused by a mixed X-ray signal from the area situated in close proximity to aluminium-rich fragments. The aluminium-rich fragments (Figure 2C) differ in chemical composition from almost calcium-free (spot 1) to Ca-enriched (spots 2, 6, and 8), but with a low content of silicon to aluminium-, calcium-, and silicon-containing phases of variable composition (spots 7 and 5). Phases containing aluminium with calcium (and oxygen) represent Al-rich calcium aluminate (with non-stoichiometric composition) or a mixture of different calcium aluminates. Phases containing aluminium, silicon, and calcium locally exhibit the presence of iron, titanium, and magnesium (Figure 2C, Table 3). Similar aluminium-containing phases, but occurring in plates larger than 50 mm, are presented in Figure 2D.

Aluminium-rich components also occur as irregularly shaped forms present in glassy material (Figure 3 and Figure 4).

Table 3. EDS chemical analyses in spots and areas indicated in Figure 2 and Figure 5 (content in wt%; “-”— not determined).

Figure (Sample)	Spot	Si	Al	Fe	Ti	Mg	Ca	Mn	Ba	K	Na	O	S	Cl	P	Zr	Cu	Cr	Zn
Figure 2A (RZ11)	1	66.69	-	0.20	-	-	0.22	-	-	-	-	32.89	-	-	-	-	-	-	-
	2	66.08	-	-	-	-	-	-	-	-	-	33.92	-	-	-	-	-	-	-
	3	45.81	1.66	0.87	0.13	1.67	12.48	-	1.37	1.67	7.56	26.6	0.18	-	-	-	-	-	-
	4	16.88	-	-	-	-	-	-	-	-	0.22	15.4	-	-	-	67.50	-	-	-
Figure 2B (RZ2)	1	30.47	-	-	-	-	48.83	-	-	-	-	20.70	-	-	-	-	-	-	-
	2	39.83	3.65	1.19	0.91	1.08	12.21			4.08	10.85	25.61	-	-	0.59	-	-	-	-
Figure 2C (RZ9)	1	0.11	71.05	1.77	-	-	0.29	0.36	-	-	-	25.78	-	-	0.64	-	-	-	-
	2	0.56	61.65	-	-	-	12.44	-	-	-	-	25.35	-	-	-	-	-	-	-
	3	21.57	15.80	5.33	9.18	0.31	12.17	1.86	-	0.78	0.61	13.1	0.13	-	9.19	-	9.97	-	-
	4	28.09	40.05	25.72	-	-	0.20	5.94	-	-	-	-	-	-	-	-	-	-	-
	5	15.08	26.26	2.30	1.01	1.15	27.37	-	-	0.74	1.39	24.09	-	-	0.61	-	-	-	-
	6	1.06	60.62	-	-	-	11.40	-	-	-	-	26.92	-	-	-	-	-	-	-
	7	7.80	46.10	-	-	-	25.75	-	-	-	-	19.07	0.19	-	1.09	-	-	-	-
	8	1.23	61.54	-	-	-	10.01	-	-	-	-	27.22	-	-	-	-	-	-	-
Figure 2D (RZ9)	1	-	62.46	-	-	-	10.75	-	-	-	-	26.79	-	-	-	-	-	-	-
	2	0.07	60.02	-	-	-	8.71	-	-	-	-	31.20	-	-	-	-	-	-	-
	3	26.91	12.55	2.43	4.18	1.77	23.38	-	-	1.01	2.67	24.65	-	-	0.45	-	-	-	-
	4	25.56	12.55	2.86	4.08	2.01	24.53	-	-	0.82	2.57	24.43	-	-	-	0.59	-	-	-
Figure 5A (RZ10)	1	-	98.93	-	-	-	0.24	-	-	-	-	0.83	-	-	-	-	-	-	-
	2	1.14	52.66	-	0.91	-	0.65	-	-	0.60	1.37	40.53	0.71	1.29	-	0.14	-	-	-
	3	0.88	51.60	1.36	-	-	6.81	-	-	0.26	0.52	36.49	1.24	0.84	-	-	-	-	-
	4	1.25	8.29	-	-	-	38.75	-	-	-	-	35.34	15.93	0.44	-	-	-	-	-
	5	0.13	12.89	-	-	-	50.61	-	-	-	-	20.52	0.16	15.69	-	-	-	-	-
Figure 5B (RZ11)	1	1.44	92.97	-	-	-	-	-	-	-	-	5.59	-	-	-	-	-	-	-
	2	6.54	61.73	10.79	-	-	-	19.88	-	-	-	0.45	-	-	-	-	-	0.61	-
	3	2.22	36.07	-	-	-	3.69	0.44	-	-	-	54.73	0.42	-	-	-	-	-	2.43
	4	-	40.85	-	-	-	0.53	-	-	-	-	58.62	-	-	-	-	-	-	-
Figure 5C (RZ3)	1	0.60	42.31	1.98	0.36	-	11.97	-	-	0.30	0.19	36.92	3.14	2.23	-	-	-	-	-
	2	-	56.81	0.81	-	-	3.20	-	-	0.23	-	36.84	0.37	1.74	-	-	-	-	-
	3	1.06	13.54	-	-	-	48.21	-	-	-	-	21.30	0.29	15.6	-	-	-	-	-
	4	0.43	35.74	-	-	-	17.73	-	-	-	-	41.48	0.28	4.34	-	-	-	-	-
	5	7.29	5.71	43.79	1.62	0.80	22.56	0.93	-	-	-	16.84	-	0.46	-	-	-	-	-
	6	18.03	11.49	0.63	-	1.23	48.07	-	-	-	-	20.55	-	-	-	-	-	-	-
Figure 5D (RZ3)	1	0.41	13.36	-	-	-	47.39	-	-	-	-	26.72	0.38	11.74	-	-	-	-	-
	2	2.17	8.30	-	-	0.58	42.17	-	-	1.10	1.10	29.35	12.37	2.57	0.29	-	-	-	-

Table 3. EDS chemical analyses in spots and areas indicated in Figure 2 and Figure 5 (content in wt%; “-”— not determined).

Figure (Sample)	Spot	Si	Al	Fe	Ti	Mg	Ca	Mn	Ba	K	Na	O	S	Cl	P	Zr	Cu	Cr	Zn
Figure 2A (RZ11)	1	66.69	-	0.20	-	-	0.22	-	-	-	-	32.89	-	-	-	-	-	-	-
	2	66.08	-	-	-	-	-	-	-	-	-	33.92	-	-	-	-	-	-	-
	3	45.81	1.66	0.87	0.13	1.67	12.48	-	1.37	1.67	7.56	26.6	0.18	-	-	-	-	-	-
	4	16.88	-	-	-	-	-	-	-	-	0.22	15.4	-	-	-	67.50	-	-	-
Figure 2B (RZ2)	1	30.47	-	-	-	-	48.83	-	-	-	-	20.70	-	-	-	-	-	-	-
	2	39.83	3.65	1.19	0.91	1.08	12.21			4.08	10.85	25.61	-	-	0.59	-	-	-	-
Figure 2C (RZ9)	1	0.11	71.05	1.77	-	-	0.29	0.36	-	-	-	25.78	-	-	0.64	-	-	-	-
	2	0.56	61.65	-	-	-	12.44	-	-	-	-	25.35	-	-	-	-	-	-	-
	3	21.57	15.80	5.33	9.18	0.31	12.17	1.86	-	0.78	0.61	13.1	0.13	-	9.19	-	9.97	-	-
	4	28.09	40.05	25.72	-	-	0.20	5.94	-	-	-	-	-	-	-	-	-	-	-
	5	15.08	26.26	2.30	1.01	1.15	27.37	-	-	0.74	1.39	24.09	-	-	0.61	-	-	-	-
	6	1.06	60.62	-	-	-	11.40	-	-	-	-	26.92	-	-	-	-	-	-	-
	7	7.80	46.10	-	-	-	25.75	-	-	-	-	19.07	0.19	-	1.09	-	-	-	-
	8	1.23	61.54	-	-	-	10.01	-	-	-	-	27.22	-	-	-	-	-	-	-
Figure 2D (RZ9)	1	-	62.46	-	-	-	10.75	-	-	-	-	26.79	-	-	-	-	-	-	-
	2	0.07	60.02	-	-	-	8.71	-	-	-	-	31.20	-	-	-	-	-	-	-
	3	26.91	12.55	2.43	4.18	1.77	23.38	-	-	1.01	2.67	24.65	-	-	0.45	-	-	-	-
	4	25.56	12.55	2.86	4.08	2.01	24.53	-	-	0.82	2.57	24.43	-	-	-	0.59	-	-	-
Figure 5A (RZ10)	1	-	98.93	-	-	-	0.24	-	-	-	-	0.83	-	-	-	-	-	-	-
	2	1.14	52.66	-	0.91	-	0.65	-	-	0.60	1.37	40.53	0.71	1.29	-	0.14	-	-	-
	3	0.88	51.60	1.36	-	-	6.81	-	-	0.26	0.52	36.49	1.24	0.84	-	-	-	-	-
	4	1.25	8.29	-	-	-	38.75	-	-	-	-	35.34	15.93	0.44	-	-	-	-	-
	5	0.13	12.89	-	-	-	50.61	-	-	-	-	20.52	0.16	15.69	-	-	-	-	-
Figure 5B (RZ11)	1	1.44	92.97	-	-	-	-	-	-	-	-	5.59	-	-	-	-	-	-	-
	2	6.54	61.73	10.79	-	-	-	19.88	-	-	-	0.45	-	-	-	-	-	0.61	-
	3	2.22	36.07	-	-	-	3.69	0.44	-	-	-	54.73	0.42	-	-	-	-	-	2.43
	4	-	40.85	-	-	-	0.53	-	-	-	-	58.62	-	-	-	-	-	-	-
Figure 5C (RZ3)	1	0.60	42.31	1.98	0.36	-	11.97	-	-	0.30	0.19	36.92	3.14	2.23	-	-	-	-	-
	2	-	56.81	0.81	-	-	3.20	-	-	0.23	-	36.84	0.37	1.74	-	-	-	-	-
	3	1.06	13.54	-	-	-	48.21	-	-	-	-	21.30	0.29	15.6	-	-	-	-	-
	4	0.43	35.74	-	-	-	17.73	-	-	-	-	41.48	0.28	4.34	-	-	-	-	-
	5	7.29	5.71	43.79	1.62	0.80	22.56	0.93	-	-	-	16.84	-	0.46	-	-	-	-	-
	6	18.03	11.49	0.63	-	1.23	48.07	-	-	-	-	20.55	-	-	-	-	-	-	-
Figure 5D (RZ3)	1	0.41	13.36	-	-	-	47.39	-	-	-	-	26.72	0.38	11.74	-	-	-	-	-
	2	2.17	8.30	-	-	0.58	42.17	-	-	1.10	1.10	29.35	12.37	2.57	0.29	-	-	-	-

Aluminium-rich components in the quench component or in fine-grained ash material are characterised by circular reaction rims (Figure 5, Table 3). Different stages of the transformation process can be observed from slightly advanced with a metallic aluminium core inside to complete substitution by secondary products (Figure 5). Figure 5A represents a relatively rare case of reaction rims around almost non-oxidised metallic aluminium (spot 1, Table 3). The metallic aluminium is surrounded by the oxidised phase (spot 2, Figure 5A), probably aluminium hydroxide (gibbsite). The rim is characterised by an irregular net of cracks (Figure 5A) and veins of the iron-rich phase exsolved in aluminium passing from the metallic aluminium core, similarly to the structure described by [17]. The rim also contains chlorine, sulphur, and a low content of aluminium. The next rim is very thin and more rich in calcium (spot 3). The following rim is relatively broad (up to 100 μm) and characterised by a high content of sulphur and calcium, but lower in comparison with the internal rims’ content of aluminium (spot 4, Figure 5A). A high content of S and Ca can suggest the occurrence of ettringite or calcium sulphate in the mixture with the calcium and aluminium phase, while the presence of other components in the mixture cannot be excluded (e.g., kuzelite Ca₄Al₂(OH)₁₂[SO₄]·6H₂O or Kuzel’s salt 3Ca·Al₂O₃·½ CaSO₄·½ CaCl₂·11H₂O). The outermost rim contains less sulphur, but significantly more chlorine (spot 5, Figure 5A). The composition of this rim corresponds to the Friedel’s salt. The formation of the Friedel’s salt in calcium-aluminate-rich material in the reaction with chloride solution is a typical process occurring in cements [26,27]. The transition from a sulphur-rich rim to a chlorine-dominated rim can be related to an increase in the chlorine concentration in the solution or to a depletion of S during rim formation.

A relatively simple reaction rim composed of oxidised aluminium (possibly hydroxide) is presented in Figure 5B (spot 2 and 3; Table 3). The rim is formed around slightly oxidised aluminium (spot 1, Figure 5B; Table 3). Compared with the previous example (Figure 5A), a lack of calcium-, sulphur-, or chlorine-containing rims is visible. The RZ11 sample was collected without quenching in a water tank, thus it solidified without contact with calcium-, sulphur-, and chlorine-containing solution.

At a more advanced stage of exchange reactions, the core of the Al-rich form is composed of material containing aluminium, calcium, oxygen, sulphur, and chlorine (spot 1, Figure 5C; Table 3). Only small fragments of the material poor in calcium are noted (spot 1, Figure 5C; Table 3). The following zones are characterised by a variable proportion between aluminium and calcium, with a relatively broad zone of the composition similar to the Friedel’s salt (spot 3, Figure 5C; Table 3). The irregular accumulation of material rich in Fe or Si can be noted (spot 5 and 6, Figure 5C; Table 3).

In the final stage of the exchange reaction, patches of the material of a composition similar to the Friedel’s salt (spot 1, Figure 5D; Table 3) are present, including S-rich material (phase of a chemical composition similar to ettringite) (area 2, Figure 5D; Table 3).

The abovementioned results indicate that of most importance to the transformation of aluminium (or aluminium-rich) metallic components in BA is the composition of the quench water. Newly formed phases around the metallic core are rich in calcium, chlorine, or sulphur, which indicates a high concentration of these components in the solution. The BA is a Ca-rich material with CaO content varying between 11.56 and 15.36 wt% (Table 2), the second-most-dominant element after SiO₂. It is possible to assume that Ca²⁺ can be present in the solution in a relatively high concentration after the reaction of the molten slag with the quench water. The sulphur content is between 0.26 and 0.55 wt% (Table 2), which is related to the primary content of sulphur in the BA and the exchange efficiency during the slag–water reaction. According to [27,28], the presence of the Friedel’s salt or chlorides can be related to the significant concentration of Cl⁻ and SO₄²⁻ in quench water.

Newly formed phases are usually present as more or less circular rims (Figure 6, Figure 7 and Figure 8). A reaction rim composed mainly of material containing calcium and sulphur (probably sulphates) is presented in Figure 6. The content of aluminium in this material is low. The rim containing more chlorine is incomplete and very thin. A more complex reaction zone contains two chlorine- and two sulphur-rich rims (Figure 7). One chlorine-rich rim occurs directly on an oxidised aluminium core, while the second is separated from the first by a rim rich in sulphur. The rims rich in chlorine contain more aluminium in comparison with the sulphur-rich rims. The most complex noted reaction zone contains three chlorine-rich rims separated by sulphur-rich material (Figure 8). Similarly to the previous example, the rims rich in chlorine are enriched in aluminium in contrast to the sulphur-containing material.

Figure 5. (A) Reaction rims around metallic Al (also see Figure 8)—numbers of spots correspond to the analyses presented in Table 3 (SEM BSE image, sample RZ10); (B) relatively simple reaction rim around a slightly oxidised Al core—numbers of spots correspond to the analyses presented in Table 3 (SEM BSE image, sample RZ11); (C,D) advanced stages of Al components’ transformation—numbers of spots and areas correspond to the analyses presented in Table 3 (SEM BSE images, sample RZ3).

Figure 5. (A) Reaction rims around metallic Al (also see Figure 8)—numbers of spots correspond to the analyses presented in Table 3 (SEM BSE image, sample RZ10); (B) relatively simple reaction rim around a slightly oxidised Al core—numbers of spots correspond to the analyses presented in Table 3 (SEM BSE image, sample RZ11); (C,D) advanced stages of Al components’ transformation—numbers of spots and areas correspond to the analyses presented in Table 3 (SEM BSE images, sample RZ3).

Figure 6. SEM BSE image of the reaction zone and distribution of selected elements—sample RZ3.

Figure 6. SEM BSE image of the reaction zone and distribution of selected elements—sample RZ3.

Figure 7. SEM BSE image of the reaction zone and distribution of selected elements—sample RZ4.

Figure 7. SEM BSE image of the reaction zone and distribution of selected elements—sample RZ4.

Figure 8. SEM BSE image of the reaction zone and distribution of selected elements—sample RZ10.

Figure 8. SEM BSE image of the reaction zone and distribution of selected elements—sample RZ10.

Variation in the composition of the reaction zones forming regular circular structures is probably caused by fluctuation in the quench water composition during the addition of a new portion of molten slag. The range of fluctuations can differ during the progression of the waste incineration process, which results in differences between different samples.

The reaction zones formed around aluminium-rich cores are generally similar to those described in the literature, but differ owing to the much higher content of Ca, S, and Cl [17,29]. The described reaction zones correspond to the corrosion textures analysed by [15], formed in the presence of salt in pore water. Yang et al. [30] compared air-cooled and water-quenched MSWI BA. The difference between the reaction zones in the RZ11 sample (without water quenching) and other samples is much more significant. This can be related to a higher concentration of elements in the quench water or a longer residence time of the slag in the quench water tank.

The described forms of occurrence of aluminium in the BA (in glass representing solidified melt phase) and in the quench phase exhibit a very strong variation in composition and size, which indicates that the attempts of aluminium recovery appear to be inefficient.

Other utilisation methods of aluminium metallic components in BA can be considered. Several authors describe hydrogen production during the reaction of metallic aluminium and water [16,17,18]. The usage of the described BA for hydrogen production was not tested, but a high degree of transformation of metallic aluminium to various secondary phases significantly reduced the possible effectiveness of the process. It is worth mentioning that metallic aluminium or slightly oxidised aluminium is shielded against contact with water by secondary products (see the process of passivation described by [17]). It is possible to conclude that hydrogen was spontaneously produced in the reaction of the slag with quench water and dispersed in the atmosphere.

MSWI BA can be used as an additive for concrete or in road construction [31,32]. However, the presence of metallic aluminium is a factor limiting this application because of the potential for hydrogen gas emission after contact with water in the alkaline environment and the lowering of the mechanical properties of the material [33,34,35]. The evolution of hydrogen gas before the adding of MSWI BA to the concrete mixture allows hydrogen bubbles in concrete to be avoided [36,37]. It is possible to conclude that the studied material had previously passed the hydrogen-generation stage.

4. Conclusions

• Aluminium-rich components occurring in the studied MSWI BA contain both the melt phase and quench phase.
• Aluminium-rich components occurring in the solid phase formed from melt (i.e., in glass) are often oxidised and occur in platy or irregular forms. The Al content in the glass phase is low (usually below 2 wt%).
• Aluminium components in the quench phase are significantly transformed with an abundance of Cl and SO₄²⁻ phases, which is a result of the reaction with water in a tank during cooling. The volume of metallic or slightly oxidised cores is strongly reduced.
• The broad variation in the composition of aluminium components makes its recovery challenging and inefficient.
• The high degree of transformation of metallic aluminium fragments indicates that the potential for hydrogen generation is limited.