Waste Management 30 (2010) 885–892

Contents lists available at ScienceDirect

Waste Management

journal homepage: www.elsevier .com/locate /wasman

Impact of struvite crystallization on nitrogen losses during composting of pigmanure and cornstalk

Limei Ren a, Frank Schuchardt b, Yujun Shen a, Guoxue Li a,*, Chunping Li c

a College of Resource and Environmental Science, China Agricultural University, Beijing 100094, Chinab Heinrich von Thunen-Institute, Federal Research Institute for Rural Areas, Forestry and Fisheries, Institute of Agricultural Technology and Biosystems Engineering,Braunschweig 38116, Germanyc Department of Environmental Science and Engineering, Tsinghua University, Beijing 100084, China

a r t i c l e i n f o

Article history:Accepted 18 August 2009Available online 29 January 2010

0956-053X/$ - see front matter � 2009 Published bydoi:10.1016/j.wasman.2009.08.006

* Corresponding author. Tel.: +86 10 62733498.E-mail addresses: ligx@cau.edu.cn, aomei386@yah

a b s t r a c t

An absorbent mixture of magnesium hydroxide (Mg(OH)2) and phosphoric acid (H3PO4) was added tocompost mixtures of pig manure with cornstalk in different molar ratios (T1, 1:1; T2, 1:2; T3, 1:3) in orderto examine its effect on controlling ammonia losses during composting. Based on the principle of struviteprecipitation, and with an unamended trial as control (CK), an in-vessel composting experiment was con-ducted in fermenters (60 L with forced aeration) in which the absorbent mixture was added with propor-tions of 3.8%, 7.3% and 8.9% of dry weight for T1, T2 and T3, respectively. The results showed that the totalnitrogen loss was reduced from 35% to 12%, 5% and 1% of initial N mass, respectively. In the final compost,the total nitrogen content in T1, T2 and T3 was improved by 10, 14, 12 g kg�1, and NHþ4 —N in T1, T2 andT3 was improved by 8, 9, and 10 g kg�1, respectively, compared with the unamended trial. The results ofthe germination index test showed that the maturity of treatment T2 was best among the four treatmentsin the final compost, followed by T1, CK and T3. The results of X-ray diffraction (XRD) confirmed the for-mation of magnesium ammonium phosphate hexahydrate (MgNH4PO4�6H2O:MAP) in the T1, T2 and T3compost. Based on these results, the adsorbent mixture of Mg(OH)2 + H3PO4 could control nitrogen losseffectively during composting via struvite crystallization. However, an excess of phosphoric acid (1:3)had a negative influence on composting properties. The pH value decreased which led to reduced micro-organism activity, and which finally resulted in reduced biodegradation of the organic matter.

� 2009 Published by Elsevier Ltd.

1. Introduction

A key problem in composting is the nitrogen loss caused by therelease of ammonia gas. Initial nitrogen (16–74%) has been re-ported as lost during composting (Beck-Friis et al., 2001; Martinsand Dewes, 1992; Raviv et al., 2002; Tiquia and Tam, 2000), whichleads not only to the decline in value of the compost as a fertilizer,but also the malodor problem in full-scale composting facilities.Therefore, controlling the nitrogen loss in composting has beenthe focus of much environmental research in the last few years.

The factors that influence nitrogen loss include: temperature,moisture content, pH, aeration, particle size and nutrient content.A low C/N ratio is an important reason for nitrogen loss (Hongand Park, 2005; Tiquia and Tam, 2000; Hu et al., 2006). In particu-lar, composting of nitrogen-rich waste (e.g. manure of domesticanimals and poultry) can be associated with substantial gaseousnitrogen losses. Nitrogen loss can be controlled by adding car-bon-rich materials (straw, leaves, hardwood, softwood, paper,sphagnum and molasses) (Eklind and Kirchmann, 2000; Liang

Elsevier Ltd.

oo.com.cn (G.X. Li).

et al., 2006); however, a lower C/N ratio for composting is expectedin practical production to reduce the cost of carbon amendmentsand to increase the amount of nitrogen-rich waste to be treated(Zhu, 2007). As a result, pressure exists to reduce C/N ratios whilenot causing problems associated with nitrogen loss.

Other methods besides carbon amendments have been used tocontrol nitrogen loss. The first is adjusting the properties of thecomposting material itself, such as lowering the pH by acid wash-ing the composting materials (Wang et al., 2006; Stevens et al.,1989; Wang et al., 2005). A second method is using materials thathave physical adsorption properties such as peat, zeolite, basalt(Kithome et al., 1999; Liao et al., 1997; Witter and Kirchmann,1989a; Eklind and Kirchmann, 2000), and active carbon (Linet al., 2005). In the third method of controlling nitrogen loss, achemical reagent is added whose principle is to react with orchange the properties of the materials. Reagents which have an im-pact on the fermentation process include CaCl2, CaSO4, MgCl2,MgSO4 and Al2(SO4)3 (Kithome et al., 1999; Witter and Kirchmann,1989), FeCl3 (Boucher et al., 1999), P and Mg salt (Jeong and Kim,2001), and potassium dihydrogen phosphate (Hu et al., 2006).The fourth is a biological method, adding an exogenous microor-ganism into the compost to change the metabolization of carbon

Table 1Properties of pig manure and cornstalk used in composting: values are means oftriplicates (dry weight basis), values in parentheses are standard error (n = 3).

Materials Pig manure Corn stalk

TOC (g kg�1) 370(3.18) 430(3.32)TN (g kg�1) 19.3(0.62) 11.1(0.04)NH4

+–N (g kg�1) 0.97(0.01) –EC (ms cm�1) 3.78(0.14) 2.15(0.09)C/N 19(0.49) 39(0.36)pH 8.5(0.22) 7.7(0.34)Moisture content (%) 72(0.63) 23(0.21)

886 L.M. Ren et al. / Waste Management 30 (2010) 885–892

and nitrogen, reduce the ammonia emission and preserve morenitrogen nutrients (Kuroda et al., 2004; Wang et al., 2005).

Struvite crystallization is usually applied to the recovery of Nand P from wastewaters (James and Simon, 2002; Suzuki et al.,2005; Sugiyama et al., 2005). According to the literature, in a givensolution, struvite will be formed if the concentrations of Mg2+,NH4

+ and PO3�4 exceed the solubility product for struvite (Ali and

Schneider, 2006; Nelson et al., 2003):

Mg2þ þ PO3�4 þ NHþ4 þ 6H2O!MgNH4PO4 � 6H2O ð1Þ

Struvite has many potential uses as a slow-release fertilizer thatcan be applied in a single high dose without damaging growingplants (de-Bashan and Bashan, 2004). Li and Zhao (2003) appliedstruvite crystallization to recover ammonium-nitrogen from land-fill leachate; the maximum NHþ4 –N recovered was 92%. Uludag-Demirer et al. (2005) removed ammonia from anaerobically di-gested dairy manure by struvite precipitation, with a maximumNHþ4 –N removal efficiency above 95%. Ganrot et al. (2007) alsoused struvite precipitation for recovery of N and P from human ur-ine. Korean researchers Jeong and Kim (2001) applied struvite pre-cipitation by adding Mg and P salts to control the N loss incomposting. Their research results showed that the loss of ammo-niacal nitrogen was clearly reduced and reached up to 1.4% of drymass. But the Mg and P salts may increase the total salinity of thecompost product, the microorganism activity is reduced, and thebiodegradation of organic matter decreased (Jeong and Hwang,2005). So, alternative Mg and P compounds should be considered.Schulze-Rettmer (1991) recommended the use of phosphoric acid(H3PO4) and magnesium oxide (MgO) for the formation of struviteto avoid the problem of high salinity in wastewater treatmentfacilities. Wu et al. (2001) applied Mg(OH)2 to an anaerobic sludgedigester and this resulted in a reduction of suspended solids andchemical oxygen demand (COD), a higher biogas production rate,and a lower level of phosphate and ammonia concentration.

Since H3PO4 and Mg (OH)2 can supply the Mg and P, and will notintroduce other ions, they should reduce salinity concerns. There-fore, this study utilized them as the additive; in addition, theycould adjust the pH value, which can be helpful both for the com-posting itself and struvite formation. In order to investigate andcompare nitrogen conservation, key compost quality parameterswere measured. Furthermore, X-ray diffraction analysis was usedto test the existence of struvite.

2. Material and methods

2.1. Composting materials and reactor

Pig manure and cornstalk were adopted in this experiment. Thecornstalk was cut into 1 cm fragments, the C/N ratio of the mixturewas adjusted to 15 by adding urea, and the initial water contentwas adjusted to 65% in all experiments. The properties of the com-posting materials are shown in Table 1. The adsorbents, Mg(OH)2

and H3PO4, were added as chemically pure reagents. In each treat-ment, Mg(OH)2 and H3PO4 were initially mixed before being addedto the composting materials, and the dry weight proportion of

GI % ¼ Seeds germination of compost sample� Root length of compoSeeds germination of control� Root length of control

composting materials and dosage of Mg(OH)2 and H3PO4 in eachtreatment are shown in Table 2.

2.2. Experimental performance

The experiment was carried out in four insulated fermenters(inner diameter 0.36 m, height 0.6 m, net volume 60 L, fillingheight of the compost about 0.5 m). The air flow was0.57 L min�1 kg�1 with continuous aeration, temperature wasmeasured and recorded by a temperature sensor connected to acomputer. The installation of the fermenter is shown in Fig. 1.Ammonia gas released during composting was captured in boricacid and was determined by 0.02 mol/L H2SO4 titration. Sampleswere collected seven times over the duration of the experiment(0, 3, 6, 10, 15, 20, 26 days) at three sampling ports and mixed thor-oughly. The samples were divided into two parts, one part wasanalyzed immediately, and the other was air-dried, and thenpassed through a 1-mm sieve. The materials were turned on day7 and day 14, and water was added to keep the moisture contentaround 65%. On day 26, the compost masses were removed fromthe fermenters.

2.3. Chemical analysis

The pH, inorganic N and germination index (GI) were deter-mined in fresh samples. Total nitrogen and total organic carbon(TOC) were analyzed in air-dried samples. Inorganic N in the sam-ples was extracted with 2 mol L�1 KCl and the NHþ4 —N deter-mined by distillation in alkaline medium (MgO). NOx –N wasdetermined by a colorimetric method. The pH, EC and GI weredetermined in water extracts (20 g of dry weight compost wereextracted with 200 ml of distilled water, stirred for 1 h and thencentrifuged at 4000 rpm). pH was measured with a pH meter,electrical conductivity (EC) was measured by a DDS-12A conduc-tivity meter.

Seed germination index was used to assess phytotoxicity (Sel-lami et al., 2008): 20 Lepidium sativum seeds were evenly distrib-uted on filter paper in Petri dishes (10 cm diameter) andmoistened with 8 ml of the compost extract. Three replicatedishes for each sample were incubated at 25 �C for three days.The number of germinating seeds and root length were measured.As a control, 8 ml of distilled water replaced the extract at everytreatment. The GI was calculated according to the following for-mula:

st sample� 100 ð2Þ

Table 2Dry weight proportion of composting materials and the dosage of adsorbents.

Item Treatment no.

CK T1 T2 T3

Pig manure (%, DM) 55 53 51 50.1Corn stalk (%,DM) 42.7 41.1 39.6 38.9Urea (%,DM) 2.3 2.2 2.1 2.1Adsorbents (%,DM) 0 3.8 7.3 8.9Moisture content (%) 65 65 65 65C:N ratio 15 15 15 15pH 7.6 7.4 6.6 6.4Dosage of Mg(OH)2 (% of initial nitrogen) 0 15.4 18.7 17.1Molar ratio of Mg (OH)2:H3PO4 0 1:1 1:2 1:3

0

10

20

30

40

50

60

70

80

0 5 10 15 20 25 30

Composting time (d)

Tem

pera

ture

()

Ambient CK T1

T2 T3

Fig. 2. Changes of temperatures of ambient and four trials during composingprocess.

L.M. Ren et al. / Waste Management 30 (2010) 885–892 887

Dry matter content was analyzed after drying at 105 �C to con-stant weight. Kjeldahl N was measured by the Kjeldahl digestionmethod (Kjeltec, Tecator, Sweden) (Lu, 2000; Nanjing AgriculturalUniversity, 1992). The theoretical total N concentration of the com-posting materials was calculated by adding the Kjeldahl N with theNOx–N, whereas the organic N concentration was derived by sub-tracting the NHþ4 —H from the Kjeldahl N. Total organic C in themixtures was measured by potassium dichromate (K2Cr2O7) andsulfuric acid (H2SO4) (Yeomans and Bremner, 1988). The struvitecrystals in the final compost product were analyzed using X-raydiffraction (D/MAX-2400) after the compost had been dried inair. Masses of N and C were determined (N or C concentra-tion �weight of dry matter) (Tiquia et al., 2002), and the losseswere computed as follows:

N loss ð%Þ ¼ ðN1 �M1 � N2 �M2Þ=ðN1 �M1Þ ð3ÞC loss ð%Þ ¼ ðC1 �M1 � C2 �M2Þ=ðC1 �M1Þ ð4Þ

Here N1 and C1 are the initial N-Kjeldahl and C concentrations, N2

and C2 are the final N-Kjeldahl and C concentrations, and M1 andM2 are the initial and final dry mass weight.

2.4. Statistical analysis

The mean and standard deviation of three replicates are re-ported and One Way Analysis of Variance (ANOVA) statistical test-ing was performed and shown for chemical parameters to comparethe variations of different treatments in composting, and multiple

Fig. 1. Compost reactor (1) air pump; (2) gas flowmeter; (3) sieve plate; (4) heat insulcomputer monitor system; (9) sampling ports and (10) filtrate outlet.

comparisons between every two treatments were compared usingLeast Significant Difference test (LSD-t). SPSS for Windows, release15.0 (SPSS, 2007) was used to perform all statistical analyses.

3. Results and discussion

3.1. Temperature and pH

From the temperature profile in Fig. 2, the mixtures in the fourtrials reached a peak temperature (>55 �C) on the second day. Afterturning and moisture addition (7th and 14th), the temperaturesrose again. Overall, the thermophilic phase (more than 50 �C)lasted for 12, 13, 13 and 6 days respectively in CK, T1, T2 and T3.

In the initial phase of composting (0–3 days), pH increased withtemperature (Fig. 3). Subsequently, pH decreased when the tem-perature declined, but on the 10th day, pH reached its peak valuewith temperature increasing again, and then decreased with thematuration of the compost. Comparing the different fermenters,the pH of T1 was highest among the treatments (7.4–8.1), whichshowed that the adsorbent at Mg(OH)2/H3PO4 ratio of 1:1 had noeffect on lowering pH, and in fact, it increased the pH value slightlyabove the pH of the unamended trial (CK). However, the pH of T3

ation layer; (5) composting material; (6) temperature sensor; (7) gas washing; (8)

280

300

320

340

360

380

400

420

0 5 10 15 20 25 30

Composting time (d)

Tot

al o

rgan

ic c

arbo

n (g

·kg

-1)

CK

T1

T2

T3

Fig. 5. Changes of organic carbon during the composting. Mean and standarddeviation are shown. Differences between treatments are not significant (P > 0.05).

5.0

5.5

6.0

6.5

7.0

7.5

8.0

8.5

0 5 10 15 20 25 30

Composting time (d)

pH

CK T1 T2 T3

Fig. 3. Changes of pH during the composting. Mean and standard deviation areshown. Differences between treatments are highly significant (P < 0.001).

888 L.M. Ren et al. / Waste Management 30 (2010) 885–892

(6.3–6.7) was the lowest among the four fermenters, less than 7during all composting days, and the variation was small. One-way analysis of variance (ANOVA) showed a highly significant dif-ference among the four trials (P = 0.000).

3.2. Ammonia (NH3) emission

In all treatments, ammonia emission showed a similar trend asshown in Fig. 4. Ammonia was emitted sharply with the rapid in-crease in temperature on days 1–4, caused by rapid hydrolysis ofurea and ammonification of organic nitrogen compound, and onthe 4th day, the NH3 loss rate reached its first peak, and then de-creased with the fall in temperature (5–7 days). However, withthe rise in temperature again (7–9 days), the rate of ammonia lossreached a maximum value on the 9th day. The ammonia loss inamended treatments was much lower than that of CK and therewere remarkable differences among treatments. The total amountsof ammonia lost in CK, T1, T2 and T3 were 54, 18, 7 and 1 g(Fig. 4(b)), equivalent to 28%, 9%, 4% and 0.5% of initial nitrogen,respectively.

3.3. Organic carbon

As Fig. 5 shows, the total organic carbon (TOC) concentrationdecreased with composting time. Over the whole composting per-iod, the degradation rate of T3 was lowest among the four treat-ments. In the final compost, TOC values in CK, T1, T2 and T3were 306, 298, 295, 319 g kg�1, respectively. The low degradation

0

50

100

150

200

250

300

350

400

450

Composting time (d)

Am

mon

ia lo

ss r

ate

(g·d

-1)

CK

T1

T2

T3

0 5 10 15 20 25 30

(a)

Fig. 4. Changes in (a) ammonia loss rate and (b

rate in T3 could be attributed to the additives which affected theproperties in the composting materials such as the pH value; alow pH value may restrain the activity of microorganisms. After26 days, 50%, 53%, 55% and 43% of initial C mass was degraded.

3.4. Nitrogen changes during composting

As shown in Fig. 6(a), in the first 3 days, the total nitrogen (TN)concentration decreased slightly in CK and T1 as a result of NH3

emission (Fig. 4), from day 4th to 15th, TN of CK showed continu-ous decrease as a result of large NH3 emission (Fig. 4). Changes ofTN in T1 were not great, and losses were lower than CK, whichcould be attributed to the adsorption of NH3 by adsorbents. After15 days, the TN showed increase both in CK and T1 because of con-centration effect caused by organic matter degradation (Fig. 5). TheTN showed a continuous rise in T2 and T3 due to only a smallquantity of NH3 loss in T2 and T3. TN concentration of T2 was high-er than T3, which could be attributed to more complete organicmatter degradation in T2 than T3 (Fig. 5). In the final compostproduct, the TN of T1, T2 and T3 was improved by 10, 14 and12 g kg�1 compared with CK. One-way analysis of variance (ANO-VA) showed a significant difference among the four trials(P < 0.05). LSD-t showed the significant difference except betweenCK and T1.

NHþ4 —N (Fig. 6b) is a key parameter of nitrogen change andammonia gas emission. The KCl extraction method used to mea-sure NHþ4 —N means that the NHþ4 —N results combine the watersoluble NHþ4 —N and the NHþ4 —N in the struvite. In the four

0

10

20

30

40

50

60

0 5 10 15 20 25 30

Composting time (d)

Cum

ulat

ed N

H3

loss

(g)

CK

T1

T2

T3

(b)

) cumulative ammonia during composting.

Table 3Mass changes of nitrogen (N) and N Losses after 26 days of composting.

Composting time N Mass

CK T1 T2 T3

Initial (g) 195 195 195 195Final (g) 127 172 185 193Balance (g) �68 �23 �9 �2N loss (%) 35 12 5 1N fixation ratio (%) 0 65.7 85.7 97.1

15

20

25

30

35

40

45

Composting time (d)

Tot

al N

(g·

kg-1

)

CK

T1

T2

T3

2

4

6

8

10

12

14

16

18

Composting time (d)

NH

4+-N

(g·

kg-1

)

CK

T1

T2

T3

(a)

0

0.1

0.2

0.3

0.4

0.5

Composting time (d)

NO

x- -N (

g·kg

-1)

CK

T1

T2

T3

10

15

20

25

30

35

0 5 10 15 20 25 30 0 5 10 15 20 25 30

0 5 10 15 20 25 30 0 5 10 15 20 25 30

Composting time (d)

Org

anic

N (

g·kg

-1)

CK

T1

T2

T3

(c)

(b)

(d)

Fig. 6. Dynamic changes of different forms of N during composting. Mean and standard deviation are shown. Data are expressed on a 105 �C dry weight basis. (a) Total N,differences between treatments are significant (P < 0.05); (b) NHþ4 –N, differences between treatments are highly significant (P < 0.001); (c) (NO�3 þ NO�2 )–N, differencesbetween treatments are not significant (P > 0.05); (d) organic N, differences between treatments are not significant (P > 0.05).

1.5

2.0

2.5

3.0

3.5

4.0

4.5

0 5 10 15 20 25 30

Composting time (d)

Ele

ctri

cal c

ondu

ctiv

ity

(ms·

cm-1

)

CK

T1

T2

T3

Fig. 7. Changes of electrical conductivity of four trials. Mean and standard deviationare shown. Differences between treatments are highly significant (P < 0.01).

L.M. Ren et al. / Waste Management 30 (2010) 885–892 889

treatments, NHþ4 —N increased at first (0–3 days) possibly due toconversion of organic N to NHþ4 —N via the ammonification process.Then it decreased when the degradation speed slowed down(Fig. 5) caused by the decline in temperature and pH (3–6 days).However, NHþ4 —N continually decreased in CK as a result of ammo-nia gas emission. The NHþ4 —N content in T3 was highest among thethree amended treatments. At the end of composting, the NHþ4 —Ncontents of CK, T1, T2 and T3 were 3, 11, 12, and 13 g kg�1, respec-tively; this increase was 3.7–4.3 times greater in the amendedtreatments than in CK. One-way analysis of variance revealed ahighly significant difference among the four trials (F = 17 and

35

45

55

65

75

85

95

105

0 5 10 15 20 25 30

Composting time (d)

Ger

min

atio

n in

dex

(%)

CK

T1

T2

T3

Fig. 8. Changes of germination index of four trials. Mean and standard deviation areshown. Differences between treatments are not significant (P > 0.05).

890 L.M. Ren et al. / Waste Management 30 (2010) 885–892

P < 0.01), and LSD-t showed a highly significant difference betweenCK and the other three amended trials.

From Fig. 6c, ((NO�3 þNO�2 )–N was very low (<0.5 g kg�1)throughout the composting. However, the organic N (Fig. 6d) hadan obvious high proportion in the total N; this indicated that or-ganic N was the major nitrogenous constituent (Tiquia and Tam,2000). At day 3, the organic N decreased in all treatments possiblydue to ammonification of organic N to NH3. Then the organic N

Fig. 9. XRD patterns of the compost after 26 days with the mixtu

increased in all treatments as a consequence of strong degradationof organic C compounds (Fig. 5).

3.5. N loss

As regards the N loss (as shown in Table 3), it is evident that theCK showed a higher N loss (about 35%) than the amended treat-ments. The high pH throughout the process was responsible for

re of the SiO2 and struvite crystals in T1(a), T2(b) and T3(c).

L.M. Ren et al. / Waste Management 30 (2010) 885–892 891

the higher N loss caused by ammonia emission; about 85% of Nlosses can be attributed to NH3–N volatilization. This was similarto the results of some studies mentioned earlier: a lower initialC:N ratio contributed to losses of N via NH3 volatilization. How-ever, the addition of adsorbents in this study reduced the loss ofNH3. Nitrogen fixed by adsorbents was estimated by the followingformula:

N fixation ratioð%Þ ¼ ðN loss of CK� N loss of treatmentÞ=N loss of CK ð5Þ

It was assumed that the theoretical NH3 output of every trial wasequal to the emission amount of CK. As a result, about 66%, 86%and 97% of N was fixed in T1, T2 and T3 respectively (Table 3).

3.6. Phytotoxicity test

In this study, electrical conductivity (EC) and germination index(GI) were used to test the salinity and phytotoxicity of compost. ECvalue can reflect the degree of salinity in the compost, indicatingits possible phytotoxic/phyto-inhibitory effects on the growth ofplant if applied to soil (Huang et al., 2004). Results (Fig. 7) showedthat EC values of T1, T2 and T3 reduced to 2.38, 1.6 and2.07 mS cm�1, which were lower than CK 2.56 mS cm�1. EC valuesof all trials were lower than 3.00 mS cm�1, which has been identi-fied as the limit for safe growth of plants (Garcia, 1991). The resultssuggest that addition of magnesium hydroxide and phosphorusacid will not increase the total salinity of compost. The decreasein EC appears to be due to the precipitation of Mg2+, NHþ4 and PO3�

4 .

The Germination Index (GI) is an important maturity indicator.When GI exceeds 80–85%, it is generally considered that phytotox-icity is minimal (Tiquia and Tam, 1998a). In this study, at thebeginning of composting, the GI (Fig. 8) of the four treatmentswas less than 50%, and then increased gradually. In the final com-post, the GI of CK, T1, and T2 were more than 80%. The low GI of T3could be caused by low pH (Fig. 4) which restrained the activity ofmicroorganism (Sundberg et al., 2004), as a result, degradation oforganic matter was slower than other trials (Fig. 5).

3.7. Magnesium ammonium phosphate hexahydrate test by XRDanalyses

Powder X-ray diffraction (XRD) analyses confirmed the forma-tion of magnesium ammonium phosphate hexahydrate (MAP), asstruvite crystals in the three amended treatments, as shown in Fig. 9.

XRD analyses indicated that MAP existed in the final compostproducts of T1, T2 and T3, which suggested that the mechanismfor N conservation was ammonia combining with adsorbent toform struvite crystals during composting. A small quantity ofSiO2 has also been found which could originate from the cornstalk.According to James and Simon (2002), struvite precipitation is con-trolled by pH, degree of supersaturation, temperature and the pres-ence of other ions in solution such as calcium and can occur whenthe concentrations of magnesium, ammonium and phosphate ionsexceed the solubility product for struvite. Struvite formation canbe observed over a wide pH range from 7 to 11, and many studieshave shown the optimal pH for struvite precipitation to be be-tween 8.5 and 9.2 in wastewater (Uludag-Demirer et al., 2005).

4. Conclusion

Throughout this study, the effect of N conservation was im-proved when an excess of molar phosphoric acid was used overmagnesium hydroxide. However, a ratio of Mg(OH)2/H3PO4 = 1:3may lead to incomplete maturity caused by low pH (6.3–6.7).

Under the conditions of this experiment, it is recommended thatthe molar ratio of Mg (OH)2/H3PO4 is 1:2; this dosage of Mg(OH)2

on a molar basis is equivalent to 18.7% of total nitrogen in the com-post mixture.

de-Bashan and Bashan (2004) concluded that mixing struvitewith phosphoric acid might even yield a superior fertilizer; partslow-release MgHPO4, and part fast release, highly soluble ammo-nium phosphate (NH4)2HPO4. This might have considerably morenutritional value than commonly used diammonium phosphatefertilizers. Therefore, a higher ratio of phosphoric acid to Mg(OH)2, up to a certain point, would be expected to have a positiveimpact on compost.

The optimal conditions for struvite precipitation should bedetermined for different systems. Under composting conditions,temperature, pH and ions show more dynamic changes than inwastewater. The lack of replications in this study limits the abilityto generalize. In particular, further work is needed to see if nitro-gen addition in the form of ammonium nitrate rather than ureachanges the conclusions.

In a practical production process, the application method of theadditives should be determined based on practical requirementssuch as cost and composting periods. Therefore, further study isneeded on the effect of different materials, the effect of a widerange of C:N ratios and a full cost-benefit analysis are also needed.The market potential of compost via the struvite crystallizationreaction is much greater because rich nutrients such as N, P andMg are all present.

Acknowledgements

This work was supported by the National Natural ScienceFoundation of China (No.30571084) and the National Science &Technology Pillar Program (Nos. 2006BAD10B05, 2007BAD89B00and 2007BAD89BP7), and a joint Sino-German transfer project:‘‘Recycling of organic residues from agricultural and municipalorigin in China”

References

Beck-Friis, B., Smars, S., Jonsson, H., Kirchmann, H., 2001. Gaseous emissions ofcarbon dioxide, ammonia and nitrous oxide from organic household waste in acomposting reactor under different temperature regimes. Agr. Eng. Res. 78,423–430.

Boucher, V.D., Revel, J.C., Guiresse, M., Kaemmerer, M., Bailly, J.R., 1999. Reducingammonia losses by adding FeCl3 during composting of sewage sludge. Water AirSoil Pollut. 112, 229–239.

de-Bashan, L.E., Bashan, Y., 2004. Recent advances in removing phosphorus fromwastewater and its future use as fertilizer (1997–2003). Water Res. 38 (19),4222–4246.

Eklind, Y., Kirchmann, H., 2000. Composting and storage of organic household wastewith different litter amendments. II: nitrogen turnover and losses. Bioresour.Technol. 74, 125–133.

Ganrot, Z., Dave, G., Nilsson, E., 2007. Recovery of N and P from human urine byfreezing, struvite precipitation and adsorption to zeolite and active carbon.Bioresour. Technol. 98, 3112–3121.

Garcia, C., 1991. Study on water extract of sewage sludge composts. Soil Sci. PlantNutr. 37 (3), 399–408.

Hong, J.H, Park, K.J., 2005. Compost biofiltration of ammonia gas from bincomposting. Bioresour. Technol. 96, 741–745.

Hu, T.J., Zeng, G.M., Huang, D.L., Yu, H.Y., Jiang, X.Y., Dai, F., Huang, G.H., 2006. Use ofpotassium dihydrogen phosphate and sawdust as adsorbents of ammoniacalnitrogen in aerobic composting process. J. Hazard Mater. 141, 736–744.

Huang, G.F., Wong, J.W.C., Wu, Q.T., Nagar, B.B., 2004. Effect of C/N on composting ofpig manure with sawdust. Waste Manage. 24, 805–813.

James, D.D., Simon, A.P., 2002. Struvite formation control and recovery. Water Res.36, 3925–3940.

Jeong, Y.K., Hwang, S.J., 2005. Optimum doses of Mg and P salts for precipitatingammonia into struvite crystals in aerobic composting. Bioresour. Technol. 96,1–6.

Jeong, Y.K., Kim, J.S., 2001. A new method for conservation of nitrogen in aerobiccomposting processes. Bioresour. Technol. 79, 129–133.

Kithome, M., Paul, J.W., Bomke, A.A., 1999. Reducing nitrogen losses duringsimulated composting of poultry manure using adsorbents or chemicalamendments. Environ. Qual. 28, 194–201.

892 L.M. Ren et al. / Waste Management 30 (2010) 885–892

Kuroda, K., Hanajima, D., Fukumoto, Y., Suzuki, K., Kawamoto, S., Shima, J., Haga, K.,2004. Isolation of thermophilic ammonium tolerant bacterium and itsapplication to reduce ammonia emission during composting of animalwastes. Biosci. Biotechnol. Biochem. 68, 286–290.

Li, X.Z., Zhao, Q.L., 2003. Recovery of ammonium–nitrogen from landfill leachate asa multi-nutrient fertilizer. Ecol. Eng. 20, 171–181 (in Chinese).

Liang, Y., Leonard, J.J., Feddes, J.J.R., McGill, W.B., 2006. Influence of carbon andbuffer amendment on ammonia volatilization in composting. Bioresour.Technol. 97, 748–761.

Liao, P.H., Jones, L., Lau, A.K., Walkemeyer, S., Eagan, B., Holbek, N., 1997. Compostingof fish wastes in a full-scale in-vessel system. Bioresour. Technol. 59, 163–169.

Lin, X.F., Li, G.X., He, Q., Zou, J.X., 2005. Materials of controlling nitrogen loss duringcomposting. J. Agro-Environ. Sci. 24, 975–978 (in Chinese).

Lu, R., 2000. Soil Agricultural Chemical Analysis Methods. China Agricultural Scienceand Technology Press, Beijing (in Chinese).

Martins, O., Dewes, T., 1992. Loss of nitrogenous compounds during composting ofanimal wastes. Bioresour. Technol. 42, 103–111.

Ali, Md.I., Schneider, P.A., 2006. A fed-batch design approach of struvite system incontrolled supersaturation. Chem. Eng. Sci. 61, 3951–3961.

Nanjing Agricultural University, 1992. Soil Agricultural Chemical Analysis, seconded. Agriculture Press, Beijing. in Chinese.

Nelson, N.O., Mikkelsen, R.L., Hesterberg, D.L., 2003. Struvite precipitation inanaerobic swine lagoon liquid: effect of pH and Mg: P ratio and determinationof rate constant. Bioresour. Technol. 89, 229–236.

Raviv, M., Medina, S., Karasnovsky, A., Ziadna, H., 2002. Conserving nitrogen duringcomposting. Biocycle 43, 48.

Schulze-Rettmer, R., 1991. The simultaneous chemical precipitation of ammoniumand phosphate in the form of magnesium–ammonium–phosphate. Water Sci.Technol. 23, 659–667.

Sellami, F., Hachicha, S., Chtourou, M., Medhioub, K., Ammar, E., 2008. Maturityassessment of composted olive mill wastes using UV spectra and humificationparameters. Bioresour. Technol. 99, 6900–6907.

Stevens, R.J., Laughlin, L.J., Frost, R.P., 1989. Effect of acidification with sulfuric acid onthe volatilization of ammonia from cow and pig slurries. J. Agr. Sci. 113, 389–395.

Sugiyama, S., Yokoyama, M., Ishizuka, H., Sotowaa, K.S., Tomidaa, T., Shigemotob, N.,2005. Removal of aqueous ammonium with magnesium phosphates obtainedfrom the ammonium-elimination of magnesium ammonium phosphate. J.Colloid Interf. Sci. 292, 133–138.

Sundberg, C., Smars, S., Jönsson, H., 2004. Low pH as an inhibiting factor in thetransition from mesophilic to thermophilic phase in composting. Bioresour.Technol. 95, 145–150.

Suzuki, K., Tanaka, Y., Kuroda, K., 2005. Recovery of phosphorous from swinewastewater through crystallization. Bioresour. Technol. 96, 1544–1550.

Tiquia, S.M., Richard, T.L., Honeyman, M.S., 2002. Carbon, nutrient, and mass lossduring composting. Nutr. Cycl. Agroecosys. 62, 15–24.

Tiquia, S.M., Tam, N.F.Y., 2000. Fate of nitrogen during composting of chicken litter.J. Environ. Pollut. 110, 535–541.

Tiquia, S.M., Tam, N.F.Y., 1998. Elimination of phytotoxicity during co-compostingof spent pig-manure sawdust litter and pig sludge. Bioresour. Technol. 65, 43–49.

Uludag-Demirer, S., Demirer, G.N., Chen, S., 2005. Ammonia removal fromanaerobically digested dairy manure by struvite precipitation. ProcessBiochem. 40, 3667–3674.

Wang, D.Q., Zeng, Q.F., Zuo, H., Pan, S., Zhao, M., 2006. Nitrogen conservation incomposting swine waste by bamboo vinegar. J. Guilin Univers. Technol. 26, 37–40 (in Chinese).

Wang, W.P., Ma, J.W., Fu, J.R., Jiang, L.N., Ye, J., Wang, J.M., 2005a. Influences ofacidification of chicken manure on its bio-fermentation characteristics. ActaAgriculturae Zhejiangensis 17, 300–302 (in Chinese).

Wang, W.P., Zhu, F.X., Qian, H., Zhu, F.X., 2005b. Effect of inoculating microorganismagent and adding different attendant agents on changes of nitrogen duringcomposting of pig manure. Acta Agriculturae Zhejiangensis 17, 276–279 (inChinese).

Witter, E., Kirchmann, H., 1989a. Peat zeolite and basalt as adsorbent of ammonicalnitrogen during manure decomposition. Plant Soil 115, 53–58.

Witter, E., Kirchmann, H., 1989b. Effects of addition of calcium and magnesiumsalts on ammonia voltailzation during manure decomposition. Plant Soil 115,53.

Wu, Q., Bishop, P.L., Keener, T.C., Stallzrd, J., Stile, L., 2001. Sludge digestionenhancement and nutrient removal from anaerobic supernatant by Mg(OH)2

application. Water Sci. Technol. 44, 161–166.Yeomans, J.C., Bremner, J.M., 1988. A rapid and precise method for routine

determination of organic carbon in soil. Commun. Soil Sci. Plant Analyst. 19(3), 1467–1476.

Zhu, N.W., 2007. Effect of low initial C/N ratio on aerobic composting of swinemanure with rice straw. Bioresour. Technol. 98, 9–13.