ca

osra

g

ence

ll Si

emi

Received 13 February 2006; received in revised form 10 September 2006; accepted 12 September 2006Available online 13 November 2006

Keywords: MAP; Landll leachate; Ammonium removal; MAP reuse

potential in solving the above problems, they are still understudy and it is not yet possible to constantly control the nitri-cation process in the nitrite-formation dominating step(Jetten et al., 1997; Strous et al., 1997).

* Corresponding author. Tel.: +86 10 62923475; fax: +86 10 62923541.E-mail address: yangmin@mail.rcees.ac.cn (M. Yang).

Chemosphere 66 (2007)1. Introduction

Sanitary landlls have been widely used for municipalsolid waste disposal in China as well as in other countriesbecause of their low cost and eectiveness. Beijing, the cap-ital of China, has constructed seventeen landll sites in itssuburbs, and these sites produce a large amount of leachate(about 8.0 105 to 9.0 105 m3 every year). Landll leach-ate from these sites generally contains high concentrationsof NH4 N ranging from several hundred mg l

1 to severalthousand mg l1 with relatively low BOD to COD and

COD to NH4 N ratios (Lo, 1996), which poses a majorproblem in wastewater treatment.

Conventional nitricationdenitrication biological pro-cesses, which have been widely utilized for ammoniumremoval from municipal and many industrial wastewaters,are not appropriate for treating landll leachate because ofthe lack of sucient electron donors in leachate and the highenergy requirement for aeration (Grabinska-Loniewska,1991; Dogen et al., 2001; Li and Zhao, 2001).While new bio-logical techniques like short-cut nitrication/denitricationand anaerobic ammonium oxidization have demonstratedAbstract

The residues of magnesium ammonium phosphate (MAP) decomposed by heating under alkali conditions were repeatedly used as thesources of phosphate and magnesium for the removal of high ammonium concentration from landll leachate. Up to 96% of ammoniumin MAP powder could be released under the following conditions: NH4 :OH

molar ratio, 1:1; temperature, 90 C; heating time, 2 h.Fourier transform infrared spectra and X-ray diraction analysis of MAP before and after heating demonstrated that MAP was mainlytransformed to amorphous magnesium sodium phosphate (MgNaPO4), which makes it possible for the NH

4 to replace Na

+ inMgNaPO4 to form more stable struvite. Successful ammonium removal was achieved by using the MAP decomposition residues asthe sole phosphate and magnesium sources. The ammonium removal decreased gradually following the increase of MAP reuse cycles,and in the 6th cycle, ammonium removals of 84% and 62% were achieved for synthetic wastewater and landll leachate, respectively.Analysis of the surfaces of MAP powders acquired at dierent reuse cycles using scanning electron microscopy with energy dispersiveX-ray suggested that the existence of calcium, kalium and aluminum ions in landll leachate might have inhibited the formation ofMAP through competition with ammonium ions for phosphate ions. It is estimated that reuse of MAP for 3 cycles could save about44% chemical costs. 2006 Elsevier Ltd. All rights reserved.Techni

Repeated use of MAP decompof high ammonium concent

Shilong He a, Yu Zhang a, Min Yana SKLEAC, Research Center for Eco-Environmental Sci

b Liulitun Sanitary Landc Institute of Environmental Ch0045-6535/$ - see front matter 2006 Elsevier Ltd. All rights reserved.doi:10.1016/j.chemosphere.2006.09.016l Note

ition residues for the removaltion from landll leachate

a,*, Wenli Du b, Hiroyuki Harada c

s, Chinese Academy of Sciences, Beijing 100085, China

te, Beijing 100049, China

stry, SAGA University, Japan

www.elsevier.com/locate/chemosphere

22332238

Formation of magnesium ammonium phosphate (MAP),a crystal with a solubility as low as 0.0023 g per 100 ml H2O,has been considered to be an eective method for theremoval of ammonium because of its high reaction rateand low residual ammonium concentration (Li et al.,

2.2. Experiment

The decomposition of MAP was conducted in alkalisolution. Factors aecting MAP decomposition were eval-uated by varying the molar ratio of OH:NH within a

pH

2234 S. He et al. / Chemosphere 66 (2007) 223322381999). The method of MAP precipitation has been exten-sively studied to treat wastewater with high NH4 content(Altinbas et al., 2002; Ozturk et al., 2003), and the NH4 N concentration could be reduced from 5618 mg l1 to112 mg l1 within 15 min under a molar ratio of Mg2+:NH4 :PO

34 of 1:1:1 for treating landll leachate (Li and

Zhao, 2001). The main obstacle to widespread applicationof the MAP process is the high consumption of phosphateand magnesium salts, which leads to a high operating cost.So it is of great importance to reduce the consumption offresh phosphate and magnesium during MAP formation.

In our previous study, recycling of phosphate and mag-nesium was proposed by heating MAP at 300 C to releaseammonia (Yang et al., 2004). Ammonia gas produced dur-ing MAP decomposition could be oxidized to nitrogen gasthrough selective oxidization, and the MAP residues mightbe able to be reused for ammonium removal as the phos-phate and magnesium sources. However, the conditionsfor MAP decomposition and the eciency of the decom-posed MAP for ammonium removal have not yet beenevaluated.

The objective of this study was to clarify the conditionsfor MAP decomposition, and to evaluate the possibility ofusing the MAP decomposition residues as the sources ofphosphate and magnesium. The eciency of ammoniumremoval from landll leachate was compared with thatfrom synthetic wastewater to investigate the eects of coex-isting compositions, and the reduction of chemical costs bythe method was estimated.

2. Materials and methods

2.1. Material

Landll leachate was taken from the Liulitun SanitaryLandll Site located in the northern suburb of Beijing.The composition of the landll leachate used in the exper-iments is shown in Table 1. Synthetic wastewater was pre-pared by dissolving 3.8207 g ammonium chloride into500 ml tap water to make a solution containing NH4 Nof 2000 mg l1.

Magnesium chloride (MgCl2 6H2O), ammoniumchloride (NH4Cl) and di-sodium hydrogen phosphate(Na2HPO4 12H2O) of analytical grade were purchasedfrom the Sigma Chemical Company (USA).

Table 1Average composition of leachate

Alk. (as CaCO3) COD BOD NH4 N TP SS11120 6800 2300 2400 8.16 2520 7.5

The unit for all items except pH is mg l1.4

range from 0.25 to 1.00, the heating temperature from60 C to 100 C, and the heating time from 0.5 h to 4.0 h.NH4 release eciency was used as the indicator for select-ing MAP decomposition conditions. Experiments for MAPpreparation, decomposition, and reuse for ammoniumremoval were performed as follows.

(1) MAP precipitate was generated by adding 18.8779 gMgCl2 6H2O, 3.8207 g NH4Cl and 28.1396 g Na2H-PO4 12H2O to 500 ml deionized water (molar ratioof Mg:N:P = 1.3:1:1.1; initial NH4 N, 2000 mg l

1),and agitated for 2 h at pH 9.0 (Stratful et al., 2001).

(2) The precipitate was collected using a 0.45 lm mem-brane lter and washed for three times with deionizedwater. The ltrate before washing was taken forquantication of NH4 , Mg

2+ and PO34 .(3) The collected solids (MAP) on the lter were heated

after NaOH was added at given temperatures.(4) All of the decomposed MAP residues were added to

500 ml synthetic wastewater containing 2000 mg l1

NH4 N or landll leachate, and agitated for 2 h atpH 9.0. Steps 2, 3 and 4 were repeated.

All of the experiments were performed in triplicate.

2.3. Analysis

The collected precipitates and the decomposed MAPresidues were washed with pure water for three times, driedin an oven at 40 C for 48 h, and then analyzed by X-raydiraction (XRD, D/max-RB, Rigaku, Japan), Fouriertransform infrared (FTIR) spectroscopy (PE2000, USA),and scanning electron microscopy with energy dispersiveX-ray analysis (SEM-EDS, S-3000, Hitachi, Japan). Theconcentrations of NH4 and PO

34 were measured according

Standard Methods (APHA, 1998), and the concentrationof Mg2+ was determined using an atomic absorption spec-trophotometer (AA-6300, Shimadzu, Japan).

3. Results and discussion

3.1. Determination of MAP decomposition conditions

The eects of OH:NH4 ratios on MAP decompositionwere investigated under the following conditions: heating

Mg Ca K Al Ni Fe Cr8.1 268 78.5 1750 13.5 0.74 1.82 3.21

magnesium sodium phosphate (MgNaPO4). Due to theextensive face sharing of PO34 tetrahedra and Mg(H2O)

2+

octahedra in MgNaPO4, Na+ might be less stable than

0.25 0.5 0.75 1.251

5

080604020

2 ( ) Fig. 2. FTIR and XRD analyses of MAP before and after heatingdecomposition. (a) FTIR of MAP (upper curve) and MAP decompositionresidues (lower curve). (b) XRD of MAP (upper curve) and MAPdecomposition residues (lower curve).

heretime, 4 h; heating temperature, 100 C, and the results areshown in Fig. 1. The ammonium release eciency increasedwith the increase of OH:NH4 ratio, and reached a level of>97% at the OH:NH4 ratio of 1:1, which was used for thefollowing experiments.

MAP could also be decomposed by direct heating at300 C (Abdelrazig and Sharp, 1998; Yang et al., 2004).However, we found that the MAP residues obtained atsuch a high temperature resulted in a low ammoniumremoval. Zhang et al. (2004) found that MAP powdercan release ammonium to form MgHPO4 at pH 6 5.0and temperature P40 C, and the resulting decompositionresidues could be eectively used for ammonium removal.However, this method allowed about 4% of phosphateand magnesium to be lost each time of MAP decom-position.

The eects of heating temperature and heating time on

Molar Ratio (OH-:NH4+)Fig. 1. Eect of OH:NH4 ratio on ammonium release from MAPpowder.0

20

40

60

80

100 N

H 4+-N

Rel

ease

Effi

cienc

y (%

)

S. He et al. / Chemospammonium release were further investigated at the xedOH:NH4 ratio of 1:1. The conditions permitting anammonium release eciency of >96% were determined asfollows: OH:NH4 ratio, 1:1; temperature, 90 C; heatingtime, 2 h.

3.2. Surface analysis on MAP decomposition residues

FTIR and XRD analysis before and after heating wereperformed to identify the transformation of MAP duringdecomposition, and results are shown in Fig. 2a and b,respectively. As shown in Fig. 2a, the infrared spectrumof the MAP decomposition residues shows a similar shapewith that of MAP. However, the characteristic ammoniumbands at 1441 cm1 and 1480 cm1 and the characteristicNH band at 3300 cm1 almost completely disappearedafter heating (Stefov et al., 2004, 2005), further conrmingthe eective removal of NH4 from MAP by heating underalkali conditions. XRD analysis indicated that the charac-teristic MAP crystal disappeared after heating, formingamorphous solids that could be assigned to amorphous4000 3500 3000 2500 2000 1500 1000Wavenumber (cm-1)

Abso

rban

ce

14801441

99610741069

3300

25

20

15

10

103

66 (2007) 22332238 2235those of larger univalent cations like NH4 (Mathewet al., 1982). Therefore NH4 can replace Na

+ in MgNaPO4to form more stable struvite, which makes it possible torepeatedly use the MAP residues for removing ammoniumfrom wastewater.

3.3. Reuse of MAP decomposition residues for ammonium

removal

The MAP residues were repeatedly used for ammoniumremoval from synthetic wastewater and landll leachatewithout further supplementation of Mg2+ and PO34 , andthe results are shown in Fig. 3. For the treatment of syn-thetic wastewater, the ammonium removal eciencydecreased gradually with the increase of reuse cycles, whichwas 84% in the 6th cycle. The residual concentrations ofMg2+ and PO34 varied in a range between 1020 mg l

1

and 210 mg l1, respectively, in the reuse cycles. Thelosses of Mg2+ and PO34 during former cycles might bemainly responsible for the decrease of ammonium removal(accounting for 6%).

However, the eciency for the treatment of landllleachate was much lower than that for synthetic wastewa-

ter, especially from the 2nd cycle, although the raw waste-water even contains a relatively high concentration ofMg2+ and some PO34 , as shown in Table 1. It is possiblethat accumulation of some leachate components of MAPmight inhibit the formation of MAP in later cycles.Another possible reason is the reduced free Mg2+ concen-tration due to the formation of Mg2+ complex. The surfacecomposition of the MAP residues obtained in dierentreuse cycles was analyzed by SEM-EDS, and the resultsare shown in Fig. 4. With the increase of reuse cycle, thepeaks of Mg, P, O, and Na (the major elements of MAPdecomposition residues) decreased, whereas the peaks forcalcium, kalium and aluminum appeared and increased,suggesting that one or more newly-appeared metal ionsmight be responsible for the decreasing ammonium removaleciency. Although the Ca2+ concentration was quite lowin comparison with other ions, as shown in Table 1, itseects on MAP formation could not be neglected (Kristellet al., 2005). The K ion, on the other hand, might competewith NH4 to form struvite analogs (MgKPO4) (Mathew

0

20

40

60

80

100

0 1 2 3 4 5 6Recycling Times

NH4+

-N

Rem

oval

(%

)

0

20

40

60

80

Resi

dual

Co

n (m

g l-1

)Fig. 3. Reuse of MAP decomposition residues for removing ammoniumfrom synthetic wastewater and landll leachate (d, j, mNH4 Nremoval, residual Mg2+ and residual PO34 in synthetic wastewater; s, h,nNH4 N removal, residual Mg

2+ and residual PO34 in landllleachate).

2236 S. He et al. / Chemosphere 66 (2007) 22332238Fig. 4. SEM-EDS analysis of MAP residues in dierent reuse cycles.

Technol. 46, 271278.APHA, 1998. Standard Methods for the Examination of Water

hereet al., 1982), since its concentration in leachate was veryhigh in this case. Since Al3+ is able to form Al2(PO4)3 pre-cipitates, its eects on the formation of MAP should not beneglected, either. However, further research is required toclarify the main inhibiting factors.

3.4. Estimation of chemical costs

The chemical costs for treating one ton of landll leach-ate (ammonia concentration 2400 mg l1) are estimated,and shown in Fig. 5. The prices of MgCl2 6H2O, Na2HPO4and NaOH are assumed to be $56 ton1, $200 ton1 and$288 ton1 (2006 market price), respectively.

Fig. 5 indicates that the total chemical costs would be$4.5 ton1 if the MAP residues are not reused. The reuseof MAP residues could save the costs of MgCl2 6H2Oand Na2HPO4 addition signicantly. However, the needfor adding NaOH counteracts the reduction of costs. Thetotal chemical costs decreases with the increase of reuse

00

1

1

2

2

3

3

4

4

5

5 6

Chem

ical C

ost(

T-1 )

Recycle Times

Costs of Mg2+ and PO43-

Cost of NaOHAverage cost

Fig. 5. Chemical costs for treating landll leachate in dierent reusecycles.

S. He et al. / Chemospcycles, and about 44% chemical costs could be saved byreuse the MAP residues for 3 cycles. However, perceptiblereduction of chemical costs could not be expected when theMAP residues are further reused. So a reuse cycle of 3should be reasonable for the treatment of landll leachate.

4. Conclusions

The possibility of using MAP decomposition residuesfor the removal of ammonium from landll leachate asthe phosphate and magnesium sources were investigated,and the following conclusions could be obtained.

(1) The conditions permitting >96% ammonium releaseeciency were determined as: OH:NH4 ratio, 1:1;temperature, 90 C; heating time, 2 h.

(2) The MAP decomposition residues could be reused forthe removal of ammonium from synthetic wastewateras the sole phosphate and magnesium sources, andthe ammonium removal was 84% in the 6th cycle.and Wastewater, 20th ed. American Public Health Association,Washington, DC.

Dogen, U., Jetten, M.S.M., Losdrecht, M.C.M., 2001. The SHARON-ANAMMOX process for treatment of ammonium rich wastewater.Water Sci. Technol. 44, 153160.

Grabinska-Loniewska, A., 1991. Denitrication unit biocenosis. WaterRes. 25, 15651573.

Jetten, M.S.M., Hom, S.J., Loosdreecht, M.C.M., 1997. Towards a moresustainable municipal wastewater treatment system. Water Sci. Tech-nol. 35, 171180.

Kristell, S., Le, C., Eugenia, V.J., Phil, H., Parson, S.A., 2005. Impact ofcalcium on struvite crystal size, shape and purity. J. Cryst. Growth283, 514522.

Li, X.Z., Zhao, Q.L., 2001. Eciency of biological treatment aected byhigh strength of ammonium-nitrogen in leachate and chemicalprecipitation of ammonium-nitrogen as pretreatment. Chemosphere44, 3743.

Li, X.Z., Zhao, Q.L., Hao, X.D., 1999. Ammonium removal from landllleachate by chemical precipitation. Waste Manage. 19, 409415.

Lo, I.M.C., 1996. Characteristics and treatment of leachates fromdomestic landlls. Environ. Int. 22, 433442.

Mathew, B.Y.M., Kingsbury, P., Takagi, S., Brown, W.E., 1982. A newstruvite-type compound, sodium phosphate heptahydrate. Acta Cryst.38, 4044.

Ozturk, I., Altinbas, M., Koyuncu, I., Arikan, O., Gomec, C.Y., 2003.Advanced physico-chemical treatment experiences on young municipallandll leachate. Waste Manage. 23, 441446.

Stefov, V., Soptrajanov, B., Spirovski, F., Kuzmanovski, I., Lutz, H.D.,Repeated losses of Mg2+ and PO34 during ammo-nium removal might be mainly responsible for thedecrease of ammonium removal.

(3) The MAP decomposition residues could also bereused for the removal of ammonium from landllleachate as the sole phosphate and magnesiumsources, and the ammonium removal was 62% in the6th cycle. The accumulation of some leachate compo-nents, such as Ca, K, and Al, on MAP might beresponsible for the lower ammonium removal e-ciency.

(4) Through the estimation of chemical costs, a reusecycle of 3 should be reasonable for the treatment oflandll leachate and about 44% chemical costs couldbe saved.

Acknowledgements

This work was supported by the State High TechResearch and Development Project of China for youngerresearchers (2004AA649280) and the National Nature Sci-ences Foundation of China (50525824).

References

Abdelrazig, B.E.I., Sharp, J.H., 1998. Phase changes on heating ammo-nium magnesium phosphate hydrate. Therm. Acta 129, 197215.

Altinbas, M., Yangin, C., Ozturk, I., 2002. Struvite precipitation fromanaerobically treated municipal and landll wastewaters. Water Sci.

66 (2007) 22332238 2237Engelen, B., 2004. Infrared and Raman spectra of magnesiumammonium phosphate hexahydrate (struvite) and its isomorphous

analogues. I. Spectra of protiated and partially deuterated magnesiumpotassium phosphate hexahydrate. J. Mol. Struct., 110.

Stefov, V., Soptrajanov, B., Spirovski, F., Kuzmanovski, I., Lutz, H.D.,Engelen, B., 2005. Infrared and Raman spectra of magnesiumammonium phosphate hexahydrate (struvite) and its isomorphousanalogues. III. Spectra of protiated and partially deuterated magne-sium potassium phosphate hexahydrate. J. Mol. Struct., 6067.

Stratful, I., Scrimshaw, M.D., Lester, J.N., 2001. Conditions inuencingthe precipitation of magnesium ammonium phosphate. Water Res. 35,41914199.

Strous, M., Gerven, E., Zheng, P., Kuenen, J.G., Jetten, M.S.M., 1997.Ammonium removal from concentration waste streams with theanaerobic ammonium oxidation (AMMOX) process in dierentreactor congurations. Water Res. 31, 19551962.

Yang, M., Wu, C.Q., Zhang, C.B., He, H., 2004. Selective oxidation ofammonium over coppersilver-base catalysts. Catal. Today 90, 263267.

Zhang, S.J., Yao, C.H., Feng, X.X., Yang, M., 2004. Repeated use ofMgNH4PO4 6H2O residues for ammonium removal by acid dipping.Desalination 170, 2732.

2238 S. He et al. / Chemosphere 66 (2007) 22332238

Repeated use of MAP decomposition residues for the removal of high ammonium concentration from landfill leachateIntroductionMaterials and methodsMaterialExperimentAnalysis

Results and discussionDetermination of MAP decomposition conditionsSurface analysis on MAP decomposition residuesReuse of MAP decomposition residues for ammonium removalEstimation of chemical costs

ConclusionsAcknowledgementsReferences