Professional Documents
Culture Documents
1 s2.0 S0956053X1830730X Main
1 s2.0 S0956053X1830730X Main
1 s2.0 S0956053X1830730X Main
Waste Management
journal homepage: www.elsevier.com/locate/wasman
a r t i c l e i n f o a b s t r a c t
Article history: The alkaline leaching of Phosphorous (P) from sewage sludge (SS) using either NaOH or KOH was inves-
Received 26 March 2018 tigated. The effect of type and concentration of alkali, solid liquid loading and leaching temperature were
Revised 13 November 2018 also investigated. The possibility to stabilise the leached SS residue via geopolymerisation was also
Accepted 25 November 2018
explored. The results showed that KOH leached more P than NaOH. The leaching conditions yielding
Available online 30 November 2018
the highest P recovery were 8 M KOH, 4% m/v solid to liquid loading, a leaching time of 2 h and a leaching
temperature of 95 °C resulting in a 85% P leached from SS. The leaching kinetics followed modified Jander
Keywords:
kinetics with the leaching process at 95 °C being controlled by diffusion of reactants through a porous
Phosphorous
Geopolymer
layer of reaction products, whilst at 80 °C and 70 °C was controlled by diffusion of reactants through a
Modified Jander kinetics dense layer of reaction products. A 10% replacement of gold mine tailings (GMT) with the leached SS resi-
Toxicity characteristic leaching procedure due yielded a geopolymer which met the minimum requirements for construction materials with regards
Sewage sludge to ASTM C67/14 and C129/14a. The geopolymer was cured at 100 °C for 5 days yielding a geopolymer
with strength of 14.2 MPa and an open porosity of 0.38. The geopolymer also met the minimum statutory
requirements for metal leachability. This study presents an opportunity for the increased alkaline recov-
ery of P from SS and the stabilisation of the SS residue thereby creating a no solid waste process.
Ó 2018 Elsevier Ltd. All rights reserved.
1. Introduction (Shiba and Ntuli, 2017) using either adsorption, sequential precip-
itation, nano filtration or ion exchange. In addition the Al and Fe in
An estimated increase of 1.8% in crop production up to the year the leachate have to be separated so that the P precipitation step
2050 (FAO, 2015), will inevitable result in an increase for phospho- can be efficient (Petzet et al., 2012). This ultimately makes the pro-
rous (P) based fertilisers. It is a well-known fact that the world cess expensive. Residual SS after P recovery can become a sec-
reserves of P are dwindling (Vaccari and Strigul, 2011) hence it is ondary waste whose effects on the environment have not
imperative to come up with alternative sources of P. It has been received much attention in research. Stabilisation of sewage sludge
shown that more than 80% of available phosphate rock now con- ash or sewage sludge to make construction materials is a well-
tains very low levels of P and high levels of radioactive elements researched topic (Yamaguchi and Ikeda, 2010; Vouk et al., 2017)
like uranium (Cieslik and Konieczka, 2017). This then calls for but the stabilisation of SS residue from P extraction is less studied.
research into alternative sources of P like sewage sludge (SS). The It is with this in mind that the stabilisation of the SS residue was
city of Johannesburg in South Africa produces over 100,000 tons investigated. This approach also satisfies the ‘no solid waste gener-
of SS annually (Shiba and Ntuli, 2017). It has been shown that sew- ation’ concept (Cieslik and Konieczka, 2017). Though the objective
age sludge contains the second highest amount of inorganic P apart of this paper was to deal with the solid waste, future studies of this
from P rock (Cieslik and Konieczka, 2017). P is a component of DNA project will need to investigate the effluent generated and its pos-
and hence essential to life (Oelkers and Valsami-Jones, 2008). The sible uses in geopolymerisation processes.
most common recovery of P from SS is the leaching of SS using Geopolymerisation is an alternative way to stabilise waste
acids and followed by precipitation of P as struvite, hydroxyap- materials and reduce heavy metal leachability in these materials
atites or calcium phosphates (Shiba and Ntuli, 2017). One of the (Zhang et al., 2009; Ahmari and Zhang, 2013). SS has been sta-
disadvantages of acid leaching is the co-leaching of heavy metals bilised with Portland cement (Zhan and Poon, 2015; Cry et al.,
resulting in the need for heavy metals removal from the leachate 2012) yet the use of SS in geopolymeric reactions offers a more sus-
tainable way in its stabilisation (Yamaguchi and Ikeda, 2010;
Chakraborty et al., 2017). Due to poor hydration qualities, SS
E-mail address: tfalayi@gmail.com has been mixed with several pozzolanic materials in order to
https://doi.org/10.1016/j.wasman.2018.11.041
0956-053X/Ó 2018 Elsevier Ltd. All rights reserved.
T. Falayi / Waste Management 84 (2019) 166–172 167
Intensity
P (Petzet et al., 2012), yet this research looked into the effect of 500
mined. The SS residue was then stabilised with gold mine tailings CP
(GMT) via alkaline activation. GMT and SS residue were the only 300
0 10 20 30 40 50 60 70 80 90
2. Materials and methods 2 theta
2.1. Materials Fig. 1. XRD analysis of SS (CP = Calcium Phosphate oxide, Q = Quartz, A = Albite)
Operating conditions (Voltage 40 kV, Current 30 mA and scan speed 1.0000°/min).
Constituent Sewage Sludge Sewage Sludge 2.3.5. Geopolymerisation of SS residue and GMT
(% m/m) Residue (% m/m) The SS residue was stabilised via geopolymerisation with gold
P2O5 14.6 2.83 mine tailings (GMT) using KOH as a solvent and potassium alumi-
SiO2 40.4 30.45 nate as an activator. The SS residue was added to GMT at 10%, 20%
CaO 6.51 8.46 and 30% m/m to make the precursor mixture. The SS residue and
Fe2O3 10.2 15.66
the GMT were thoroughly mixed with potassium aluminate (KA).
SO3 5.42 7.87
Al2O3 15.4 8.44 A workable paste (that could easily be poured into the mould)
MgO 1.25 2.44 was then made by adding 10 M KOH to the dry solids. The liquid
K2O 1.96 17.44 to solid ratio of the paste was 0.26. The paste was cast into a
MnO 0.84 2.15 100 100 100 mm mould. The cast moulds were cured at room
TiO2 0.92 1.96
temperature for 90 days and at 100 °C for five days respectively.
168 T. Falayi / Waste Management 84 (2019) 166–172
The above procedure was repeated with pure GMT (GMT-0). This AlPO4 þ 4MOH ! 4M þ þ PO3
4 þ AlðOHÞ4 ð2Þ
was based on the previous research work done by the authors
(Falayi et al., 2018). The Unconfined compressive strength (UCS) There was an increase in P recovery with an increase in the con-
test was conducted with the aid of a stress controlled compression centration of alkali as shown in Fig. 3. This showed that P leaching
machine with a stress rate of 0.25 MPa/s. The reported UCS was an was alkali concentration dependent as the single factor ANOVA for
average of 3 samples. Toxicity characteristic leaching procedure KOH gave a p value of 1.06 1010. The highest recovery was at
(TCLP) was done in accordance with the method used by 10 M KOH though the result at 8 M KOH was within the error limits
Anastasiadou et al. (2012). The modified ASTM C373-14a method of the analysis and hence the difference between 8 and 10 M KOH
was used to determine open porosity. The cured geopolymer was was statistically insignificant, therefore 8 M KOH was chosen to
soaked in water for 24 h. After 24 h the geopolymers were removed determine the effect of temperature on P leaching
from water and visible water on the composite was wiped using (p = 0.05961 > 0.05). There was a slight decrease in P yield as NaOH
soft cloth. The wet geopolymers were weighed within 5 min of concentration was increased from 4 to 10 M. The decrease was
being removed from water. Open porosity, f, was then calculated as within error limits of the analysis although the NaOH
(p = 0.0789 > 0.05) slurries became more viscous with an increase
f ¼ W s W d =V a ð1Þ in concentration leading to reduced leaching efficiencies. 4 M
where Ws is the mass of the soaked geopolymer, Wd was the mass of NaOH was chosen for the determination of the effect of tempera-
the dry geopolymer, V was the volume of the geopolymer and a was ture on P leaching.
the density of water. Water absorption rate was determined using The increase in P yield with temperature was because at higher
ASTM C373-14a. temperatures, the rate of reaction increases as more particles col-
lide (Li et al., 2017) (Fig. 4). A single factor ANOVA for temperature
2.4. Statistical analysis dependent leaching of P showed that temperature was a significant
factor in P leaching (p = 1.41 1024 < 0.05). Leaching by refluxing
All reported results are an average of three test specimens. The yielded the best conditions as it resulted in 85 and 67% P leached
error bars in all graphs at 95% confidence interval. Analysis of vari- using KOH and NaOH respectively. This is in agreement with
ance (ANOVA) using Microsoft Excel was used to determine signif- Petzet et al., 2012, that direct alkaline leaching at low tempera-
icance of difference in results. tures yield less P. The refluxing temperature was chosen as the
one which gave the highest P yield.
There was an increase in P yield with time due to the increased
3. Results and discussion
interaction between the SS and the alkali solution resulting in fur-
ther leaching of P (Fig. 5). P leaching was time dependent from
3.1. Leaching experiments
30 min to 120 min (p = 2.01 1014 < 0.05 using single factor
ANOVA) whilst from 120 min to 180 min the increase in P leached
There was an increase in P yield with an increase in the solid to
was statistically insignificant (p = 0.23 > 0.05). After 180 min the P
liquid loading up to 4% m/v as shown in Fig. 2. This was due to an
yield was 85.9 and 68% for KOH and NaOH based leaching respec-
increase in P available from the SS. There was a decrease in P yield
tively as compared to 75% yield for the Japanese waste water treat-
from 6 to 8%, this might have been caused by reduction in mass
ment plants (Nakagaw and Gotou, 2010) and less than 50% for raw
transfer efficiency due to more solid being available for a fixed
sewage sludge (Schaum et al., 2007) with both using NaOH.
quantity of alkali (reagent starvation) (Nosrati et al., 2013; Li
et al., 2012; Hosseini et al., 2017). The ANOVA two factor with
replication analysis showed that the difference in P leaching 3.2. Leaching kinetics and thermodynamics
between NaOH and KOH was statistically significant
(p = 0.004 < 0.05). P leaching was also dependent on solid loading The KOH based leaching was studied further to establish the
as the ANOVA single factor analysis yielded a p value of 0.001. kinetics and thermodynamics of the reactions involved. The leach-
On average KOH yielded more P than NaOH due to that KOH has ing outcome fitted the modified Jander kinetics. The linear equa-
higher alkalinity as compared to NaOH. Alkalinity is important in tion for the modified Jander equation is shown in Eq. (3)
the leaching of P as shown in Eq. (2) (Takahashi et al., 2001). The lnð1 ð1 X ÞÞ1=3 ¼ 1=NlnK N þ 1=Nlnt ð3Þ
4% m/v solid to liquid loading was then chosen for the rest of the
experiments. where t was the leaching time, X was the fraction of P leached, K has
the dimensions of a diffusion coefficient. The value of N is explained
35 100
90
30
80
25 70 KOH
% P leached
% P leached
20 60
NaOH
50
15 KOH
40
NaOH
10 30
20
5
10
0 0
2 4 6 8 2 4 6 8 10
Solid to liquid loading (%m/v) Alkali concentraon (M)
Fig. 2. Variation in P yield with alkali type and solid loading (Time 120 min, alkali Fig. 3. Variation in the yield of P with alkali concentration (Time 120 min, S/L 4% m/
concentration 2 M, leaching temperature 95 °C). v and leaching temperature 95 °C).
T. Falayi / Waste Management 84 (2019) 166–172 169
100 ln t
0
90
0 1 2 3 4 5 6
80 -0.2
95°C
70 -0.4
80°C
y = 0.5479x - 2.8862
% P leached
ln (1-(1-X))1/3
50
KOH -0.8
40
NaOH -1
30
y = 0.4796x - 2.7911
20 -1.2
R² = 0.9794 (80ºC)
10 -1.4
0 -1.6 y = 0.4382x - 3.1223
50 60 70 80 95 R² = 0.9395 (70ºC)
-1.8
Temperature (ºC)
Fig. 6. Modified Jander plots (Alkali concentration 8 M KOH, S/L 4% m/v and
Fig. 4. Effect of leaching temperature on P yield (Time 120 min, alkali concentration
leaching time 180 min).
8 M KOH and 4 M NaOH and S/L 4% m/v).
100
1/T (K-1)
90
0
80
0.0027 0.00275 0.0028 0.00285 0.0029 0.00295
70 -0.1
60 -0.2
%P
50 -0.3
KOH
40
NaOH -0.4
30
ln k
y = -1129.6x + 2.4672
20 -0.5 R² = 0.9998
10 -0.6
0 -0.7
30 60 90 120 150 180
-0.8
Time (mins)
-0.9
Fig. 5. Effect of time on P yield (Alkali concentration 8 M KOH and 4 M NaOH, S/L
4% m/v and leaching temperature 95 °C). Fig. 7. P leaching Arrhenius plot. (Alkali concentration 8 M KOH, S/L 4% m/v and
leaching time 180 min).
Table 2
XRF of Ca-P precipitates.
2500 20
TCP 18
16
2000 14
UCs (MPa)
12
10
1500 8
6
TCP
Intensity
4
TCP
1000 2
TCP
0
GMT-0 GMT-10 GMT-20 GMT-30
KP Geopolymer
500 Q Q
KP KP
Q A
Q Q
Q TCA TCA Fig. 9. Variation of UCS with SS residue content (Curing at 100 °C for 5 days, using
10 M KOH, L/S ratio of 0.26 and KA:KOH = 2.8, KA is potassium aluminate).
0
0 10 20 30 40 50 60 70 80 90 10
2 theta 9
8
Fig. 8. XRD of Ca-P precipitate TCP-Tricalcium Phosphate, Q = Quartz, KP = Potas-
7
sium Phosphate, TCA tricalcium aluminate, A = Albite.
6
UCS (MPa)
GMT-0
3.3. Precipitation of calcium phosphate (CaP) 5
GMT-10
4
GMT-20
The equation detailing the precipitation of P with Ca(OH)2 can 3
be written as (Petzet et al., 2012) 2 GMT-30
1
3Ca2þ þ 2PO3
4 ! Ca3 ðPO4 Þ2 ð5Þ 0
14 21 28 56 90
None of the precipitates did show heavy metals (Table 2) anal-
Curing Time (days)
ysis which is one of the advantages of direct alkaline leaching as
compared to acidic leaching (Shiba and Ntuli, 2017). The Ca-P pre- Fig. 10. Variation of UCS with curing time (Curing at room temperature, using 10 M
cipitate synthesised with 1 g of Ca(OH)2 had the highest P2O5 con- KOH, L/S ratio of 0.26 and KA:KOH = 2.8).
tent. The objective was to obtain a precipitate with the highest
quantity of P2O5. At levels greater than 1% Ca(OH)2 there was a nate (1.2388%), Albite (0.7324%) and Quartz (6.1829%). The Ca-P
reduction in the overall content of P2O5 because the solids were precipitate was bioavailable to plants as it had a 92.7% dissolution
dominated by Ca(OH)2 hence an increase in the CaO content of in citric acid.
the precipitates. The silica and alumina were present in the precip-
itate because they both dissolve in KOH.
The mineralogical structure of the Ca-P precipitate was domi- 3.4. Stabilisation of SS residue
nated by tri-calcium phosphate (TCP) and potassium phosphate
(KP) (Fig. 8). The PDxL quantification showed that the precipitate GMT-0 was a geopolymer synthesised with pure acidic gold
constituted of TCP (78.5125%), KP (12.6835%), Tricalcium alumi- mine tailings, whilst GMT-10, GMT-20 and GMT-30 represented
Table 3
Physical properties of various geopolymers (Curing at 100 °C for 5 days, using 10 M KOH, S/L ratio of 0.26 and KA:KOH = 2.8, KA is potassium aluminate).
Table 4
Geopolymer TCLP (Curing at 100 °C for 5 days, using 10 M KOH, S/L ratio of 0.26 and KA:KOH = 2.8, KA is potassium aluminate).
Parameter GMT (ppm) SS Residue GMT-0 GMT-10 GMT-20 GMT-30 Statutory limits (ppm)
(South African water
Guideline, 1996)
Fe 12.75 6.45 1.8 2.52 4.87 5.67 5
Ni 4.51 2.46 0.23 0.34 0.41 0.42 0.4
Cu 2.69 0.45 0.13 0.18 0.32 0.34 2
Mn 3.51 1.32 0.5 0.23 0.26 0.45 10
Zn 2.62 ND ND ND ND ND 4
Sulphates 10,387 1288 2340 2376 2386 2365 6000
pH 3.84 11.45 10.8 11.1 11.2 11.3 –
T. Falayi / Waste Management 84 (2019) 166–172 171
1000 the specific gravity of GMT was 2.99 as compared to 1.03 for SS
900 residue. The addition of SS residue to GMT significantly changed
800 the UCS of the geopolymers (p = 2.8 1011 < 0.05 using single fac-
700 tor ANOVA). Room temperature curing (Fig. 10) yielded geopoly-
600 mers with lower UCS due less thermal energy available for the
500 dissolution of silica and alumina. The geopolymerisation was slow
400 GMT with setting/hardening of the geopolymer paste happening after
300
5 days. Maximum strength was achieved after 56 days. GMT
200
CP replacement with 30% SS residue resulted in 84.6% reduction in
100 SSR
UCS as compared to the GMT-0 geopolymer.
0
0 10 20 30 40 50 60 70 80 90
3.4.1. Wet UCS durability
2 theta
An increase in SS residue content resulted in increase in water
Fig. 11. XRD of GMT-10 geopolymer (SSR-sewage sludge residue, CP calcium absorption accompanied by a reduction in durability (Table 3).
Phosphate, S = quartz, P = Pyrite, M = Haematite) Curing at 100 °C for 5 days, using GMT-30 geopolymer recorded over 70% reduction in UCS indicat-
10 M KOH, L/S ratio of 0.26 and KA:KOH = 2.8, KA is potassium aluminate. ing that it was not durable. The minimum water absorption
required for a construction material is 25% (ASTM C67/14,
Fig. 12. SEM micrograph of GMT (A), Sewage sludge residue (B) and GMT-10 (C). Curing at 100 °C for 5 days, using 10 M KOH, L/S ratio of 0.26 and KA:KOH = 2.8, KA is
potassium aluminate.
172 T. Falayi / Waste Management 84 (2019) 166–172