View Full Table | Close Full ViewTable 1.

Particle size, pH, and chemical composition or origin of the filter materials.

Material Diameter pH Composition or origin (dominant elements) Reference
Blast furnace slag 9.5–19 n.d.† 42%CaO, 6%MgO, 0.3%Fe2O3, 15%Al2O3 Mann and Bavor, 1993
Fly ash <0.5 n.d. 0.2%CaO, 0.3%MgO, 3.5%Fe2O3, 70.7%Al2O3
Granulated slag 0.08–4.75 n.d. n.d.
Gravel (R/G) 5–10/3–5 n.d. 0.08/0.01%MgO, 0.8/0.4%Fe2O3, 0.4/4.5%Al2O3
Black oxide <2 5.2 Taken from mineral sands processing industry Cheung et al., 1994
Fly ash (I/II) <2 5.9/7.6 From Kwinana (I) and Pillar Point (II) power station
Red mud gypsum <2 8.5 Red mud + 5% gypsum
FILTRALITE <2 10.4/10.3 24.8/31.0%Ca, 16.2/12.5%Mg, 3.0/3.4%Fe, 8.5/11.9%Al Zhu et al., 1997
LECA (I/II) <2 9.4/10.5 5.6/8.5%Ca, 3.3/3.6%Mg, 2.2/3.9%Fe, 6.0/6.5%Al
LehighCementVA <2 9.2 34.1%Ca, 7.3%Mg, 3.0%Fe, 11.6%Al
Sand (I/II) <2 5.8/5.1 4.2/3.9%Ca, 4.7/4.7%Mg, 6.1/3.0%Fe, 10.1/7.4%Al
UTELITE <2 10.1 45.6%Ca, 9.3%Mg, 3.3%Fe, 14.7%Al
Blast furnace slag n.d. n.d. 38–43%CaO, 5–8%MgO, < 1.3%FeO, 13–16% Al2O3, 32–37%SiO2 Sakadevan and Bavor, 1998
Steel furnace slag n.d. n.d. 35–45%CaO, 7–12%MgO, 20–30%FeO, 1–5%Al2O3, 10–15%SiO2
Zeolite (70% clinoptilolite) n.d. n.d. 2.1%CaO, < 1%MgO, < 1%FeO, 13%Al2O3, 66%SiO2
Bauxite 6–13 5.9 Al and Fe oxides Drizo et al., 1999
Fly ash 0–4 8.3 SiO2, Al2O3 and Fe oxides
LECA 6–14 8.2 Light expanded clay aggregates
Limestone 6–14 7.8 Calcium carbonate
Shale 2–13 4.5 Limestone derived
Zeolite 6–13 6.5 Aluminium-silicate
Amorphous slag (F/C)‡ 0–0.125/0.25–4 10.6/10.3 35%CaO, 13.4%MgO, 10.6%Al2O3, 36.2%SiO2 Johansson, 1999
Crystalline slag (F/C) 0–0.125/0.25–4 10.2/10.3 35%CaO, 13.4%MgO, 10.6%Al2O3, 36.2%SiO2
Fly ash n.d. 4.9 23.7%CaO, 23.0%Fe2O3, 12.5%Al2O3, 22.1%SiO2
Limestone 0.25–2 8.9 53%CaCO3, 4%MgO
Opoka 0.125–2 8.3 50%CaCO3, 4%MgO, 5%Fe2O3, 10.6%Al2O3, 36.2%SiO2
Spodosol n.d. 5.4 3.4%Fe2O3, 13%Al2O3
Calcareous soils n.d. 7.3–8.2 High CaCO3 content Zhou and Li, 2001
Limestone n.d. n.d. n.d.
Electric arc furnace steel slag 2.5–10 n.d. 21.7%Ca, 7.9%MgO, 24.3%Fe2O3, 2.5%Al2O3, 6.4%Si Drizo et al., 2002
Amorphous slag 0–0.1 n.d. 32.1–33.6%CaO, 15.1–15.6%MgO, 0.2–0.7%Fe2O3, 6.2–7.2%Al2O3, 40.2–40.4%SiO2 Kostura et al., 2005
Crystalline slag 0–0.1 n.d. 38.0%CaO, 13.7%MgO, 0.2%FeO, 6.6%Al2O3,38.6%SiO2
Fly ash <0.15 9.37 2.7%CaO, 1.5%MgO, 7.3%Fe2O3, 25.4%Al2O3,56.4%SiO2 Li et al., 2006
Red mud <0.15 11.70 46%CaO, 1.2%MgO, 12.8%Fe2O3, 6.9%Al2O3,19.1%SiO2
Bentonite <0.25 3.03 1.8%CaO, 3.1%MgO, 0.5%Fe2O3, 16.7%Al2O3, 67.5%SiO2 Xu et al., 2006
Fly ash <0.25 12.1 3.1%CaO, 0.9%MgO, 6.9%Fe-oxides, 27.8%Al2O3, 50.4%SiO2
Furnace slag 0–5 12.3 Furnace bottom ash with similar composition to fly ash
Sand (I-IV) 0–5 6.7–8.1 32.5–56.3%Ca, 0.9–1.6%Mg, 2.3–6.3%Fe, 1.3–2.0%Al
Filtralite P 0.5–4 10.7 3.1%Ca, 0.7%Mg, 0.6%Fe, 2.0%Al Ádám et al., 2007
Shell sand 3–7 8.82 32.8%Ca, 1.4%Mg, 0.05%Fe, 0.03%Al
Fe-coated sand 0.7–1 n.d. 0.3%Fe Boujelben et al., 2008
Fe-coated brick 0.8–2 n.d. 0.5%Fe
n.d., no data.
F, fine; C, course.

View Full Table | Close Full ViewTable 2.

Batch experiment parameters used in the selected studies.

Materials Mass Volume Material-to- solution ratio Electrolyte P concentration Time Rotation Temperature Estimation of P sorption capacity Reference
g mL mg L−1 h rpm °C
Blast furnace slag, Fly ash, Gravel 20 40 1:2 KCl (0.01 mol L−1) 5–100 24 (30) 1500 25 Langmuir, Freundlich Mann and Bavor, 1993
Black oxide, Fly ash, Red mud gypsum 5 25 1:5 KCl (0.01 mol L−1) 0–800 25 n.d.† 25 Langmuir, Freundlich Cheung et al., 1994
FILTRALITE, LECA, Lehigh Cement VA, Sand, UTELITE 8 200 1:25 Water 0–320 24 n.d. 22 P-sorbed at highest initial P concentration Zhu et al., 1997
Blast furnace slag, Steel furnace slag, Soil, Zeolite 3 30 1:10 KCl (0.01 mol L−1) 10–1000 (10,000) 48 100 25 Langmuir Sakadevan and Bavor, 1998
Bauxite, Fly ash, LECA, Limestone, Shale, Zeolite 20 ?(500) 1:25 CaCl2 (0.01 mol L−1) 2.5–40 24 60 21 Langmuir; Column test (35–45 mg P L−1) at 3 L d−1 Drizo et al., 1999
Fly ash, Furnace slag, Limestone, Opoka, Spodosol 1 50 1:50 NaNO3 5–25 24 70 Room (?) P-sorbed at highest initial P concentration Johansson, 1999
Calcareous soils, Limestone 3 30 1:10 KCl (0.05 mol L−1) 0.1–1250 24 n.d. Room (?) Langmuir, Freundlich;One-point isotherm Zhou and Li, 2001
Electric arc furnace steel slag 35 700 1:20 Water 1–320 24 175 24 Langmuir; Column test (350–400 mg P L−1) at 1.73 L d−1 Drizo et al., 2002
Amorphous slag, Crystalline slag 0.5 100 1:200 Water 50–500 150(1min) 0(200) 20 Langmuir, Freundlich Kostura et al., 2005
Fly ash, Red mud 0.1 20 1:200 KCl (0.01 mol L−1) 0.31–3100 4 180 25 Langmuir, Freundlich Li et al., 2006
Bentonite, Fly ash, Furnace slag, Sand, Soil 2 40 1:20 NaCl (0.01 mol L−1) 10–10010–1000 24 200 25 Langmuir Xu et al., 2006
Filtralite P, Shell sand 3 90 1:30 Water 0–480 24 n.d. n.d. P-sorbed at highest P concentration; Column test (5–10 mg P L−1) at 4.5–5.5 L d−1 Ádám et al., 2007
Fe-coated sand, Fe-coated brick 5 250 1:50 NaOH HNO3 5–30 2 n.d. 20 Langmuir, Freundlich Boujelben et al., 2008
n.d., no data.

View Full Table | Close Full ViewTable 3.

Phosphorus sorption capacity of the materials investigated.

Material P sorption capacity Conclusions/Practical application Reference
g P kg−1
Blast furnace slag 0.42 The industrial waste substrate (Blast furnace slag, Fly ash) may have potential as substrate suited to phosphorus removal in a constructed wetland (CW) system. Mann and Bavor, 1993
Fly ash 0.26
Granulated slag 0.16
Gravel (R/G) 0.03/0.05
Black oxide 0.89 The results suggest that alkaline fly ash may be a promising amendment for coarse sand bed to enhance P removal. Batch experiments should only be considered as an initial estimate of the PSC. Cheung et al., 1994
Fly ash (I/II) 1.19/3.08
Red mud gypsum 5.07
FILTRALITE 1.39/2.21 The PSC of different LWA products and sands varies by two orders of magnitude. The results are not sufficient for predicting the substrate life-length as a P-sink for a CW system. Zhu et al., 1997
LECA (I/II) 0.16/0.57
Lehigh Cement VA 2.91
Sand (I/II) 0.43/0.44
Blast furnace slag 44.25 CW systems with 60 cm substrate depth, surface area of 6.25 m2 and hydraulic load rate of 1250 L d−1 with 8 mg P L−1 would last for 11 yr for zeolite and 58 yr for blast furnace slag. Sakadevan and Bavor, 1998
Steel furnace slag 1.43
Zeolite 2.15
Bauxite 0.61 The results are not sufficient to forecast accurately the lifetime of a CW system. Batch experiments can be used to obtain an initial estimate. Based on column results, a lifetime of 7 yr is suggested for shale in a 5 m2 × 0.6 m CW with 440 g P yr−1 load. Drizo et al., 1999
Fly ash 0.86
LECA 0.42
Limestone 0.68
Shale 0.65
Zeolite 0.46
Amorphous slag (F/C)† 0.42/0.15 Crystalline slag could be considered a suitable filter material for P sorption in ecologically-engineered treatments plants. However, further investigation is needed to study the PSC over a long period of time (column experiments). Johansson, 1999
Crystalline slag (F/C) 1.00/0.65
Limestone 0.25
Opoka 0.10
Spodosol 1.00
Calcareous soils 0.59–5.55 P sorption values estimated from the one-point isotherm are very similar to those calculated from the Langmuir isotherm. Zhou and Li, 2001
Limestone 0.01 (0.42)
Electric arc furnace steel slag 3.93 (batch) P adsorption capacity derived from batch experiments can vary by several orders of magnitude. Life expectancy of a full-scale system (Drizo et al., 1999) would be 13–37 yr (based on column results). Drizo et al., 2002
1.35–2.35 (column)
Amorphous slag (I/II) 6.47/8.50 P sorption by slags can be described by the Langmuir equation. Acidification, neutralizing and sorption capacities are well correlated. Kostura et al., 2005
Crystalline slag 18.94
Fly ash 63.22 (78.44) Both substrates are inexpensive and can be economically used for the removal of P from wastewater. The PSC of a sorbent can be improved by ‘activation’ either thermally or with acids (results in brackets). Li et al., 2006
Red mud 113.87 (345.50)
Bentonite 0.93 Lifetime of CW can be estimated using the Langmuir P sorption maximum. Furnace slag shows promise as a substrate for CW and may have a potential lifetime of 22 yr. The Langmuir isotherm may provide a quick screening tool but full-scale supplementary research is needed to accurately forecast PSC. Xu et al., 2006
Fly ash 8.81
Furnace slag 8.89
Sand (I-IV) 0.13–0.29
Filtralite P 2.50 Batch and column studies suggest that shell sand has a more durable PSC than Filtralite P due to a persistently higher concentration of Ca. Ádám et al., 2007
Shell sand 9.60
Fe-coated sand 1.50 Fe-coated sorbents can efficiently remove P from contaminated water. Langmuir and Freundlich equations fitted experimental data satisfactorily. Boujelben et al., 2008
Fe-coated brick 1.75
F, fine; C, course.

View Full Table | Close Full ViewTable 4.

Classification of the filter materials investigated in the selected studies according to the P sorption capacity and particle size.

P sorption capacity Filter materials
P sorption Fine ( <1 mm) Coarse ( >1 mm)
g P kg−1
Very low <0.1 Soils Gravels
Low 0.1–0.5 Sand, Soils LECA‡, Limestone, Opoka
Moderate 0.5–1 Bentonite, Calcareous soils, Fly ash†, Spodosol Bauxite, BFS†, Zeolite†
High 1–10 BFS† ‡, Fly ash†, Fe-coated sand and brick BFS†, EAF‡, Filtra P§, Filtralite P, Polonite§, Shell sand, UTELITE
Very high >10 BFS†, Fly ash†, Polonite§, Red mud n.d.‡
Depending on chemical composition.
BFS, blast furnace slag; EAF, electric arc steel furnace slag; LECA, light expanded clay aggregates; n.d., no data.
§ Based on other studies (Brogowski and Renman, 2004; Cucarella et al., 2007; Gustafsson et al., 2008).