Ind. Eng. Chem. Res.

2010, 49, 3597–3602 3597

Reaction Mechanism and Kinetic Analysis of the Decomposition of

Phosphogypsum via a Solid-State Reaction
Liping Ma,* Ping Ning, Shaocong Zheng, Xuekui Niu, Wei Zhang, and Yalei Du
Kunming UniVersity of Science and Technology, Kunming, Yunnan 650093, People’s Republic of China

Phosphogypsum decomposition is very complex because of its complicated compounds. In this study, using
high-sulfur-concentration coal as a reducer, thermogravimetric analysis, differential thermal analysis, scanning
electron microscopy, and X-ray diffraction were used to characterize the decomposition process of
phosphogypsum in a nitrogen atmosphere at different conditions. Mechanism analysis and experiment results
showed that the amount of the intermediate production of CaS depended on the heating rate and the size of
coal used, which was produced mainly through the reaction between CaSO4 and C, and production would
decrease with an increase of the coal size. 60 mesh was a suitable size of coal for phosphogypsum
decomposition to get high recovery of the main production of CaO. Using Kissinger, Flynn-Wall-Ozaw,
and Coats-Redfem methods, the mechanism model of phosphogypsum decomposition was confirmed, the
decomposition process belonged to the core-form and growing mechanism (n ) 2), and the kinetic model
was dR/dt ) 1.6 × 1016e-475.99×10 /RT × 2(1 - R)[-ln(1 - R)]1/2, where R is conversion.
3

1. Introduction microscopy (SEM) and X-ray diffraction (XRD) analysis to

deduce the decomposition reaction process and mechanism. A
Phosphogypsum is the gypsum formed as a byproduct of the
kinetic model was proposed, and experimental data were adopted
processing of phosphate ore into fertilizer with sulfuric acid
to validate the reaction mechanism and kinetic analysis.
using a one- (dihydrate) or two-step (hemihydrate-dihydrate)
process.
2. Mechanism Analysis and Kinetic Model
[Ca5(PO4)·3F(s) + 5H2SO4(aq) + 10H2O(l) f The main components in phosphogypsum are CaSO4 · 2H2O
5CaSO4·2H2O(s) + 3H3PO4(aq) + HF(aq) (calcium sulfate dihydrate) or CaSO4 · 1/2H2O (calcium sulfate
hemihydrate) and CaSO4 (calcium sulfate anhydrite). The
Over 30 million tons of phosphogypsum is produced per annum following reactions could take place in a reduction environment:
in China, and less than 10% has been reused.1,2 By the end of
2005, the stock pile of phosphogypsum was almost 100 Mtons,3 2CaSO4 + C f 2CaO + 2SO2+CO2 (I)
which caused various environmental and storage problems and
had become a urgent task for the government to reuse or recycle CaSO4 + 4C f CaS + 4CO (II)
this solid waste for the purpose of environment protection. The
main components in phosphogypsum are CaSO4 · 2H2O (calcium 3CaSO4 + CaS f 4CaO + 4SO2 (III)
sulfate dihydrate) or CaSO4 · 1/2H2O (calcium sulfate hemihy-
drate) and CaSO4 (calcium sulfate anhydrite). One of the CaSO4 + 3CaS f 4CaO + 4S (IV)
recycling methods for this solid waste is thermal decomposition
to recycle the sulfur source and produce cement with the residual CaS + 2SO2 f CaSO4 + 2S (V)
solid material.4-6 Phosphogypsum decomposition has been
S + O2 f SO2 (VI)
studied under various experimental conditions, such as using
reducing gases CO, H2S, H2, and CS2 and reacting with different
2C + O2 f 2CO (VII)
types of carbon such as coke, coal, charcoal, etc.7,8 However,
there is still a debate on the precise mechanism and on the
C + CO2 f 2CO (VIII)
elementary step of controlling the decomposition. The nature
of the solid and gaseous products varies depending on the
CO + CaSO4 f CaO + CO2 + SO2 (IX)
reaction conditions and reducing agents, making the process
difficult to control.9-13
1 1
In this study, the kinetic characteristics of phosphogypsum 2CO + CaSO4 f CaS + 2CO2 (X)
2 2
decomposition were analyzed via thermogravimetric analysis
(TGA) and a tube reactor by heating the mixture with high- Besides these, there are many components contained in phos-
sulfur-concentration coal at elevated temperatures in a nitrogen phogypsum (Table 1). Because of its complex components, the
atmosphere. The purpose of this study is to elucidate the reaction reaction process is uncertain14,15 and depends on the reaction
mechanism and reaction kinetics of the decomposition of conditions and atmosphere. It is very important to clearly state
phosphogypsum in a solid-state reaction. The solid raw and the decomposition of the reaction mechanism for its design in
production material were examined using scanning electron industry use.
* To whom correspondence should be addressed. E-mail: lip- According to the nonisothermal dynamic analysis of solid-
ingma22@ state reactions, the basic model for solid decomposition could
hotmail.com. Tel: 86-871-5170905. Fax: 86-871-5170906. be expressed as follows:16
10.1021/ie901950y  2010 American Chemical Society
Published on Web 02/22/2010
3598 Ind. Eng. Chem. Res., Vol. 49, No. 8, 2010

Table 1. Chemical Components of Phosphogypsum

composition
SO3 CaO SiO2 Al2O3 Fe2O3 MgO total P2O5 hydrotropic P2O5
content (%) 40.86 29.82 9.43 0.236 0.132 0.055 1.17 0.87

composition
total F hydrotropic F Na2O K2O MnO free water crystal water acid-insoluble material
content (%) 0.52 0.12 0.043 0.086 0.002 5.38 4.27 8.42

dR/dt ) Ae-E/RTf(R) (1) sample was about 0.074 mm. SEM analysis of the phospho-
gypsum sample is shown in Figure 1, from which it is clear to
where a is the conversion of phosphogypsum, t is time, A is a see the sliced crystal structure of phosphogypsum. The high-
constant, E is the reaction active energy, and R is the gas sulfur-concentration coal used in the experiment came from
constant. Yunnan Chuxiong; the compounds are listed in Table 2. The
The reaction active energy could be calculated by the coal sample also was broken and filtered before experiments.
Kissinger16 (eq 2) and Flynn-Wall-Ozawa (FWO)17 (eq 3) 3.2. Experiment Equipment. The experiment and analysis
multiheating rate methods, respectively. equipment used in this study are as follows: XXWRT-2C
thermogravimetric analyzer (Beijing Optician Plant); KM9106
ln(β/T2) ) ln(AR/E) - E/RT (2) complex fuel gas analyzer (KANE Company); D/max-3BPEX-
P96 powder X-ray diffractometer (Japan); SK2 tube resistance
log β ) log[AE/RG(R)] - 0.4567E/RT - 2.315 (3)
stove (Tianjing Zhonghuan Experiment Stove Company, Ltd.).
where β is the rate of temperature increase and G(R) is defined 3.3. TGA. TGA was performed on a XXWRT-2C thermo-
as follows: gravimetric analyzer. The phosphogypsum sample and coal were
mixed in definite proportions and used for kinetic analysis with
∫
R dR masses between 10 and 20 mg. Nitrogen was used as the
G(R) ) (4)
0 f(R) dynamic atmosphere, with a flow rate of 10 mL/min at different
heating rates β (5, 10, 15, 20, and 40 °C/min, respectively).
Therefore, f(R) is the differential form of the reaction mechanism 3.4. Experiments in a Tube Reactor. A tube reactor (SK2
function, and G(R) is the integral form. tube resistance stove) was used to analyze the decomposition
The advantage of the FWO method is that the active energy process of phosphogypsum mixed with definite sizes of high-
E could be calculated directly without using the reaction sulfur-concentration coal (200, 160, 100, and 60 mesh, respec-
mechanism function. It could be used to test the results tively), and data were collected to test the reaction mechanism.
calculated by the Kissinger method. The reactor was a quartz tube with 25 cm diameter. The sample
In order to confirm the mechanism function G(R), a constant
was placed in a little porcelain boat; when the temperature was
heating rate method was used:
increased about 800 °C, the boat was placed in the middle of
reactor. The output gas was analyzed using a KM9106 fuel gas
ln[G(R)/T2] ) ln[AR/βE(1 - 2RT/E)] - E/RT (5)
analyzer online.
If the mechanism function is fitted, ln[G(R)/T2] is linear with
1/T and G(R) will be the most probable mechanism function. 4. Results and Discussion
Combined with the active energy E calculated by the Kissinger
4.1. Effect of the Proportions of Phosphogypsum and
and FWO methods, the most probable mechanism function could
Coal. As mentioned in the prework, the proportions of phos-
be confirmed.
phogypsum and coal affect the decomposition reaction,18 and
the mole proportion of C:CaSO4 ) 0.5-1.0 is suitable for the
3. Experimental Section
decomposition of phosphogypsum. Figure 2 shows the TGA
3.1. Sample Prepared. The phosphogypsum sample used curve of phosphogypsum mixed with coal at different mass
in this study came from Yunnan Natural Gas and Chemical proportions. The mass loss observed between 120 and 130 °C
Engineering Company; the compounds are listed in Table 1. is due to the loss of the hydrated water of phosphogypsum, and
After drying and filtration, the size of the phosphogypsum the small mass loss between 750 and 800 °C occurs because of

Figure 1. Phosphogypsum microstructures at different times.

 Ind. Eng. Chem. Res., Vol. 49, No. 8, 2010 3599

Figure 2. TGA and DSC curves of decomposition at different C:CaSO4 proportions.

Table 2. Elementary Components of the Coal Sample
Cr Hr Or Nr Sr
71.17 3.43 2.73 1.56 1.62

Table 3. TGA Results for the Reaction between Coal and Different
Calcium Sulfate Compounds

phosphogypsum:coal ofbeginning time phosphogypsum

decomposition decomposition mass loss
(mass) time (s)
(s) (mass %)
5:1 4600 1100 22.2
8:1 5900 1800 29.2
10:1 4300 800 40.1
12:1 4500 900 34.7
Figure 3. Gas production concentration change with the reaction temper-
ature: (a) SO2; (b) CO.
Table 4. Calculated Results of ∆rG and ∆rH of Reactions
reaction T (°C) ∆rG (kJ/mol) ∆rH (kJ/mol)
I 900 -26.540 581.000
II 500 -34.385 513.280
III 1200 -14.064 955.949
IV 1500 219.570 331.093
V 300 -155.658 -352.047
VI 300 -300.810 -298.089
IX 1000 -12.879 202.944
X 300 -175.506 -174.127

oxidation of some of the carbon in the mixture. From 950 to

1200 °C, the TGA curve is constantly decreasing because of
the decomposition of CaSO4. It is clear that the large mass loss
appeared at different times, and it increased with an increase Figure 4. Concentration change of SO2 with the reaction temperature.
of the C:CaSO4 proportion. The reaction time and mass loss
percentages are listed in Table 3. mass loss in the TGA curve. However, there is only a little
TGA shows that when the mass proportion of phosphogyp- mass loss between 500 and 850 °C (Figure 2). Except for
sum:coal is 10:1, the decomposition is almost complete; reaction IV, mass loss would take place from reaction I to
therefore, this proportion was chosen as the reactant in the next reaction X at high temperature (>900 °C). When this is
experiments. compared with the results by a tube reactor experiment in a N2
4.2. Reaction Mechanism Analysis. The decomposition atmosphere (Figure 3), it can be seen that the concentrations of
reaction of phosphogypsum is very complex because of its gas productions of SO2 and CO changed with temperature. The
complicated reaction system. Researchers have given different concentrations of both gases are slowly increased with an
conjectures,19-21 and until now, there was no integrated systemic increase of the temperature; when the temperature was higher
explanation for this process. According to thermodynamic than 750 °C, the concentration of SO2 increased abruptly, while
analysis,18 the beginning temperature of the possible reactions the concentration of CO changed mildly, which indicated that
taking place are listed in Table 4. From Figure 2, it is noted at lower temperature reactions II, V-VII, and X may take place
that a large mass loss appears at 1000-1150 °C, which means and reactions VI and VII could not take place because of the
that the main reactions take place in this temperature range. inert atmosphere. With an increase of the temperature, the
According to thermodynamic calculations (Table 4), the main decomposition of CaSO4 began (reaction I), and then the reaction
reaction I could take place at 900 °C and reaction II could take between CaSO4 and intermediate production of CaS (reaction
place at 500 °C. If these reactions took place, there would be a III) took place. When Figures 2 and 3 are compared, it is clear
3600 Ind. Eng. Chem. Res., Vol. 49, No. 8, 2010

Figure 6. Concentration change of CO with the reaction time at different

coal sizes: (a) 200 mesh; (b) 100 mesh; (c) 60 mesh.

Figure 7. Microstructures of solid production at different coal sizes.

impurity compounds contained in phosphogypsum. From reac-

tions IV and V, there will be a production of S; however,
reaction IV will not take place at this temperature range
according to thermodynamic analysis, and from XRD analysis
of the decomposition production (Figure 5), the elementary S
could not be found in the production; therefore, the reaction
Figure 5. XRD characterization of the solid production. rate of reaction V is very slow at this condition.
In order to further improve the reaction mechanism, CO was
that the decomposition temperature is lower than that of used as the reduction reactant instead of coal in the experiment
theoretical calculations (Table 4); this may be due to the in a tube reactor; the results are shown in Figure 4. It is clear
 Ind. Eng. Chem. Res., Vol. 49, No. 8, 2010 3601

Figure 8. TGA (a) and DTG (b) curves of phosphogypsum decomposition with coal at different heating rates β.

Figure 9. ln β/Tp2 - 1/Tp of the Kissinger method (a) and log β - 1/Tp of the FWO method.

Table 5. Active Energies and Linearities Calculated with Different Mechanism Functions
β ) 2.5 °C/min β ) 5 °C/min β ) 10 °C/min β ) 15 °C/min
G(R) E (kJ/mol) r E (kJ/mol) r E (kJ/mol) r E (kJ/mol) r
R 2
1410.55 0.9668 1403.24 0.9734 1396.33 0.9694 1387.11 0.956
(1 - R) ln(1 - R) + R 1561.37 0.9787 1551.72 0.9826 1545.49 0.9804 1538.34 0.9708
[1 - (1 - R)1/3)]2 1849.34 0.9893 1829.08 0.9853 1823.76 0.9844 1825.09 0.9852
-ln(1 - R) 971.91 0.9912 963.68 0.9907 962.01 0.9933 961.43 0.9917
[-ln(1 - R)]1/2 479.81 0.9909 475.56 0.9913 474.56 0.9939 474.01 0.9915
[-ln(1 - R)]1/3 312.19 0.9923 308.67 0.9886 308.11 0.9924 308.11 0.9928
1 - (1 - R)1/2 820.76 0.9823 813.52 0.9827 811.80 0.9847 809.15 0.9778
1 - (1 - R)1/3 913.13 0.9892 902.73 0.9838 899.9 0.9837 900.82 0.9854
(1 - R)-1 - 1 1358.42 0.9736 1346.35 0.9693 1346.70 0.9786 1351.94 0.9865
(1 - R)-1/2 - 1 1168.78 0.9888 1155.65 0.9835 1156.56 0.9911 1159.22 0.9953

that, with an increase of the reaction temperature, the production At 200 mesh size of coal, the crystal structure is floral and
through SO2 decreased, which means that the reaction rate of coarse on the surface and many intermediate productions such
reaction X is faster than that of reaction IX and an amount of as CaS exist, while a small satin particle appears when the
CaS is produced. coal size is 60 mesh. Therefore, 60 mesh is suitable for
4.3. XRD-SEM Analysis for the Decomposition phosphogypsum decomposition to obtain the main production
Production. The decomposition production at different coal of CaO.
sizes was analyzed with XRD and is shown in Figure 5. With 4.4. Kinetic Model. In order to confirm the mechanism
an increase of the coal size, the compound intensity of CaSO4 function, the TGA data (Figure 8) obtained from different
and CaS decreased distinctly and the production of CaO heating rates were used for calculations in eq 5 with different
increased, which further explained the reaction process analyzed mechanism functions; when the linearities of ln[G(R)/T2]-1/T
above. A high decomposition recovery at a defined size of coal are compared and the best linearity mechanism function is
can be obtained; 60 mesh is suitable for decreasing the chosen as the most probable model for this process, the active
intermediate production of CaS and increasing the production energy of the reaction could be calculated. The results are listed
of CaO, while a smaller coal size may helpful for gasification Table 6. Active Energies Calculated by a Mechanism Function
of C to CO and then produced CaS (reaction X). The change
β (°C/min)
of the concentration of CO at different coal sizes further attested
to this phenomenon (Figure 6). 2.5 5 10 15 average
Figure 7 shows the SEM analysis for the decomposition E (kJ/mol) 479.81 475.56 474.56 474.01 475.99
production from above experiment with different coal sizes. ln(A/s-1) 38.70 37.57 36.75 36.22 37.31
3602 Ind. Eng. Chem. Res., Vol. 49, No. 8, 2010

in Tables 5 and 6. From the results, the differential form of the China, 2007AA06Z321) and Nature Science Foundation of
suitable mechanism function could be Yunnan (2007E182M), which are gratefully acknowledged.
Literature Cited
f(R) ) 2(1 - R)[-ln(1 - R)]1/2 (6)
(1) Tao, J. F. The development direction for sulfuricic acid and phosphate
fertilizer industry of Yunnan Province. Phosphate Compd. Fertilizer 2007,
and the integral form is 22 (2), 8–10.
(2) Wang, S. J. The synthesis applicuation methods of phosphogypsum
and attached important problems. Sulfuric Phosphorus Bulk Mater. Handling
G(R) ) [-ln(1 - R)]1/2 (7) Relat. Eng. 2003, 12 (1), 29–31.
(3) Li, Y. The synthesis applicuation of phosphogypsum. Yunnan Chem.
The active energies calculated with the Kissinger and FWO Eng. (Chinese) 2007, 34 (6), 74–80.
methods are shown in Figure 9. It is clear that the results are (4) Zhang, X. L.; Han, L. H. Studies on technology of cement integrated
with sulfuric acid produced from phosphogypsum. Shandong Chem. Ind.
consistently calculated with these two methods, which means (Chinese) 2007, 36, 26–27.
that this mechanism function is suitable for explaining the (5) Xi, M. Y. Comprehensive utilization of phosphogypsum. EnViron.
process of phosphogypsum decomposition. The kinetic equation Sci. Technol. 2001, 11 (3), 11–13.
for phosphogypsum decomposition is as follows: (6) Rei, J. L. Cleaner phosphogypsum coal combusition and waste
incineration ashes for application in building materials: A review. Building
EnViron. 2006, 42 (2), 1036–1042.
dR
) 1.6 × 1016e-475.99×10 /RT × 2(1 - R)[-ln(1 - R)]1/2
3
(7) Strydoma, C. A.; Potgieter, J. H. Dehydration behaviour of a natural
dt gypsum and a phosphogypsum during milling. Thermochim. Acta 1999,
(8) 332, 89–96.
(8) Taher, M. A. Influence of thermally treated phosphogypsum on the
properties of Portland slag cement. Resour., ConserV. Recycl. 2007, 52,
5. Conclusion 28–38.
Phosphogypsum decomposition is very complex because of (9) Smadi, M. M.; Haddad, R. H.; Akour, A. M. Potential use of
phosphogypsum in concrete. Cem. Concr. Res. 1999, 29, 1419–1425.
its complicated compounds. On the basis of TGA experiments (10) Saloua, S.; Mohamed, L. O.; Fatima, C. S.; Jilali, A.; Larbi, B.
and XRD analysis, the reaction process and mechanism of Thermal behaviour of Moroccan phosphogypsum. Thermochim. Acta 1997,
phosphogypsum decomposition have been studied. The inter- 302, 69–75.
mediate production of CaS formed depends on the heating rate (11) Canut, M. M. C.; Jacomino, V. M. F. Microstructural analyses of
phosphogypsum generated by Brazilian fertilizer industries. Mater. Charact.
and the size of coal used, which was produced mainly through 2008, (59), 365–373.
a reaction between CaSO4 and C, and it will decrease with an (12) van der Merwe, E. M.; Strydom, C. A. Thermogravimetric analysis
increase of the coal size. 60 mesh is a suitable size of coal for of the reaction between carbon and CaSO4 · 2H2O, gypsum and phospho-
phosphogypsum decomposition to get high recovery of the main gypsum in an inert atmosphere. Thermochim. Acta 1999, 431–437.
production of CaO. (13) William, C. B.; Alan, W. E. Nuclide migration and the environ-
mental radiochemistry of Florida phosphogypsum. J. EnViron. Radioact.
The reaction kinetic model of phosphogypsum decomposition 2001, 54, 27–51.
in a solid-state reaction was researched based on a chosen (14) Ozturk, A.; et al. Decomposition of phosphogypsum with oil shale
mechanism function. Using the Kissinger, FWO, and Coats- in a fluidized-bed reactor and its industrial suitability. Cimento Beton
Redfem methods, the mechanism model of phosphogypsum Dunyasi 1997, (6), 30–37.
(15) Yang, M.; Qian, J. S. Effect of impurities on the working
decomposition was confirmed; the decomposition process performance of phosphogypsum. Mater. ReV. 2007, 21 (6), 104–106.
belongs to the core-form and growing mechanism (n ) 2) and (16) Yuan, J.; Deng, M. J. Un-isotherm thermal decomposition kinetic
the kinetic model is dR/dt ) 1.6 × 1016e-475.99×10 /RT × 2(1 -
3
for poly-hydroxy butyric-valeric acid. J. Phys. Chem. (Chinese) 2005, 21
R)[-ln(1 - R)]1/2. (9), 988.
(17) Rafiq, R. A.; Anthony, P. S. Thermal analysis of reactions and
transformations in the solid state. Experimental evaluation of published
Appendix kinetic methods using thermogravimetry. Thermochim. Acta 1978, 26, 672–
688.
Nomenclature
(18) Ning, P.; Ma, L. Z. Decomposition of high-sulfur coal to restore
A ) constant the technical basis of phosphogypsum; Metallurgical Industry Press: Beijing,
E ) reaction active energy 2007.
f(R) ) differential form of the reaction mechanism function (19) Chen, J. M; Yang, R. T. Kinetics and mechanism of the regeneration
G(R) ) integral form of the reaction mechanism function of lime sorbent for fluidized bed combustion. Ind. Eng. Chem. Fundam.
1979, 118, 134–138.
R ) gas constant (20) Kikuchi, E. I. T.; Hat, S. T.; Morita, Y. Fundamental studies on
T ) temperature of the reaction, °C the regeneration of fluidizing material for desulurization in furnance. Kinetics
t ) time, s for regeneration of calcium sulfate by carbon monoxide. Nenryo Kyokaishi
Greek Letters 1983, 62, 204–210.
(21) Zhang, J. X.; Peng, J. H.; et al. Determination of organic matters
R ) conversion of the reactant
in phosphogypsum and their influence on the properties of cement. J.
β ) rate of temperature increase, °C/min Sichuan UniV. (Eng. Sci. Ed.) (Chinese) 2006, 38 (3), 110–113.
ReceiVed for reView December 08, 2009
Acknowledgment ReVised manuscript receiVed January 28, 2010
Accepted February 03, 2010
Financial support for this project was provided by the National
High Technology Research and Development Plan (863 of IE901950Y