Chemosphere 121 (2015) 62–67
Contents lists available at ScienceDirect
Chemosphere journal homepage: www.elsevier.com/locate/chemosphere
The mechanism of coal gas desulfurization by iron oxide sorbents Yi-Hsing Lin a, Yen-Chiao Chen a, Hsin Chu b,⇑ a b
Department of Environmental Engineering, National Cheng Kung University, 1 University Road, Tainan 701, Taiwan Department of Environmental Engineering and Research Center for Energy Technology and Strategy, National Cheng Kung University, 1 University Road, Tainan 701, Taiwan
h i g h l i g h t s
g r a p h i c a l a b s t r a c t
The existence of CO and H2 affects
equilibrium concentrations of H2S and COS. The presence of H2 has a negative effect on H2S removal. By-products including COS, CS2, CH4, and Fe3C have vital influence on the reaction. The pathways of Fe2O3 reacting with H2S were successfully established. The major route of the reaction mechanism is Fe2O3 reacting with H2S to form FeS.
a r t i c l e
i n f o
Article history: Received 9 June 2014 Received in revised form 30 October 2014 Accepted 6 November 2014 Available online 27 November 2014 Handling Editor: O. Hao Keywords: Sorbent Fe2O3 H2S IGCC Mechanism
Fe3C CO2+ CS2
CO
+
CO2 + H2
H2+COS
Fe+COS
CO
CO
H2O + CS2
COS
CH4+H2O COS+FeO
CO
H2S + Fe2O3 CO2
H2O+COS
CO 773K
FeS + CO2 + H2O
H2
Fe3O4 + H2S H2
FeS + H2O H2
Fe + H2S
a b s t r a c t This study aims to understand the roles of hydrogen and carbon monoxide during the desulfurization process in a coal gasification system that H2S of the syngas was removed by Fe2O3/SiO2 sorbents. The Fe2O3/SiO2 sorbents were prepared by incipient wetness impregnation. Through the breakthrough experiments and Fourier transform infrared spectroscopy analyses, the overall desulfurization mechanism of the Fe2O3/SiO2 sorbents was proposed in this study. The results show that the major reaction route is that Fe2O3 reacts with H2S to form FeS, and the existence of CO and H2 in the simulated gas significantly affects equilibrium concentrations of H2S and COS. The formation of COS occurs when the feeding gas is blended with CO and H2S, or CO2 and H2S. The pathways in the formation of products from the desulfurization process by the reaction of Fe2O3 with H2S have been successfully established. Ó 2014 Elsevier Ltd. All rights reserved.
1. Introduction Integrated Gasification Combined Cycle (IGCC) is one of the most ideal technologies with high power generation efficiency and environmental friendly performance (Song et al., 2013; Fan et al., 2013b). Coal currently is the most abundant fossil fuel with relatively low and constant price. However, reserves of fossil fuels have been limited and global warming crisis is increasing ⇑ Corresponding author. Tel.: +886 6 208 0108; fax: +886 6 275 2790. E-mail address: [email protected] (H. Chu). http://dx.doi.org/10.1016/j.chemosphere.2014.11.010 0045-6535/Ó 2014 Elsevier Ltd. All rights reserved.
H2
H2
H2O+C
(Monazam et al., 2012). Therefore, the technology improvement for energy issues has become a trend nowadays, especially coal poly-generation with IGCC. Sulfur contained in coal will be emitted to typically form 0.2–3 vol% H2S and lesser amounts of other sulfur compounds during the process of gasification (Monazam et al., 2012). Long-term exposure to these sulfur-containing species may cause the components of IGCC system severely corrosive damage. Therefore, multiple impurities must be removed to comply with environmental laws and regulations for protecting the public health and reducing maintenance costs. Nowadays, almost all commercial IGCC power plants use wet desulfurization processes to
Y.-H. Lin et al. / Chemosphere 121 (2015) 62–67
remove H2S and other deleterious gases from hot syngas by the use of large amount of water, resulting in the decreasing thermal efficiency of the system (Lee and Feng, 2012). Consequently, the development of dry desulfurization at high temperature becomes vitally crucial in this field, and can improve the thermal efficiency and reduce cost of capital without the procedure of coal gas cooling and heating (Wan et al., 2011). Several researchers have reported on the important role of metal sorbents on the performance of hot-gas desulfurization (Jung et al., 2013; Zhang et al., 2013; Fan et al., 2013a). The sorbents can be divided into noble metal sorbents (Ru and Rh) (Hulteberg et al., 2005) and base metal oxide sorbents (Mn, Cu, Fe, Zn) (Li and Flytzani-Stephanopoulos, 1997; Wan et al., 2011; Zhang et al., 2012; Zheng et al., 2012). Iron oxide is the most common material as a sorbent for hot gas desulfurization under a reducing atmosphere with hydrogen and carbon monoxide due to its high absorption capacity, abundant resources, lower price, excellent physical strength and easy regeneration (Wang et al., 2011; Fan et al., 2013a). In our previous study (Tseng et al., 2008), the Fe-based sorbents emerged as the leading candidates for hot coal gas desulfurization through thermodynamic screening. In this study, the influences of CO and H2 on the performance of 10 wt.% Fe2O3/SiO2 sorbents for H2S removal were evaluated in various simulated syngas. The syngas (H2S) before and after the reactor were recorded by a gas chromatograph (Shimadzu, GC14B) equipped with a flame photometry detector (FPD) and fitted with a GS-Q capillary column. The outlet gases were recorded through on-line Fourier transform infrared (FTIR) spectroscopy and GC-FPD. A further investigation into the desulfurization mechanism of the sorbents was also conducted. The results will provide useful information for the designs and applications of the sorbents in the IGCC system. 2. Experimental 2.1. Sorbent preparation The Fe2O3/SiO2 sorbents were prepared by incipient wetness impregnation. Commercial SiO2 spherical pellets (Alfa Aesar, stock #44740) were used as sorbent supports. On the basis of the predetermined contents of iron loading from our previous study Huang (2007) and Bukhtiyarova et al. (2007), 5.63 g (Fe(NO3)39H2O) was used as the precursors of Fe dissolved in 16 mL de-ionized water and then mixed with the 10 g SiO2 supports. The impregnated materials were kept at room temperature for 12 h to let solution fully penetrate into the pore of supports. Subsequently, the impregnated sorbents were dried in an oven at 393 K for 24 h. Finally, the impregnated sorbents were conducted in a calcination process at 973 K for 8 h in airflow conditions to become 10 wt.% Fe2O3/SiO2 sorbents. 2.2. Sorbent characterization In the present work, temperature-programmed reduction (TPR) with hydrogen was conducted in a TG/DTA (Pyris Diamond TG/ DTA, Perkin Elmer) to determine the phase transformation in the temperature range from room temperature to 1073 K with a heating rate of 50 K min1. Elemental analyses (EA) were provided with an Elementar Vario EL-III elemental analyzer. 2.3. Desulfurization experiment The H2S removal experiments were carried out in a bench-scale fixed-bed reactor (a quartz tube with 1.5 cm id, 1.8 cm od, and 85 cm long) with 10 000 ppm inlet H2S at atmospheric pressure
63
and 773 K. A thin layer of #200 mesh frit quartz disk was set in the reactor 45 cm below the top of the tube to support the catalyst and uniformly distribute the gas stream. The weight of the catalyst packing was 1 g (thickness 0.82 cm). The weight hourly space velocity (WHSV) was set at 2000 mL h1 g1. 773 K and 2000 mL h1 g1 were chosen because it provided the best performance for H2S removal according to previous studies (Ko et al., 2006a; Tseng et al., 2008). The experimental apparatus included a simulated syngas system, a high-temperature reaction system, and a gas analysis system. The simulated syngas had a typical composition of 1 vol% H2S, 25 vol% CO and/or 15 vol% H2, and N2 as a balance gas, supplied from gas cylinders, mixed in a mixer to ensure the gas mixture was well mixed, and then introduced to the reaction system. The high-temperature reaction system was consisted of a bench-scale fixed-bed reactor, an electrical furnace, a temperature controller (TTE, TM-4800), and a K-type thermocouple. The gas analysis system was composed of a GC-FPD and a FTIR. Diffuse reflectance FTIR spectra of outlet gases were measured by using a Perkin–Elmer Spectrum One FTIR spectrometer. The recorded IR spectra were operating in the spectral range 4000–700 cm1 at a resolution of 4 cm1 with the use of Time Base 2.0 software. The desulfurization capacity of Fe2O3/SiO2 sorbents was calculated by the equation of sorbent utilization (SU):
Rt SUð%Þ ¼
0
ðC in C out Þdt 100% Cin t
ð1Þ
where Cin and Cout are the inlet and outlet concentrations of H2S (ppm), respectively; t is the experimental breakthrough time (min), and t⁄ is the theoretical breakthrough time (min) of the process. The experimental breakthrough time was defined as the time from the beginning of the experiment to the point when the outlet H2S concentration reached 100 ppm, which is 1% of the inlet H2S concentration. The theoretical breakthrough time (min) was calculated by the equation:
WX t ¼
M
A
F
ð2Þ
where W is the weight of the sorbent placed in the reactor (g), X is the actual weight loading of iron oxide supported on SiO2 (g g1), M is the molecular weight of the iron oxide (g mol1), A is the moles of H2S that can be absorbed by one mole of iron oxide (mol mol1), and F is the molar flow rate of the inlet H2S (mol min1). 3. Results and discussion 3.1. Temperature-programmed reduction The TG/DTA curves as a function of temperature for the 10 wt.% Fe2O3/SiO2 sorbent with 15% H2 in N2 are displayed in Fig. 1. In the TG curve, the sorbent can be assigned into three stages. First, a small weight loss (0.5%) of the sorbent accompanies an exothermic process at about 323–473 K, which is attributed to the evaporation of adsorbed water (Tan et al., 2012). The weight loss (0.4%) in the second stage can be attained at 693–773 K that is due to the reduction of hematite to magnetite (Pineau et al., 2006).
3Fe2 O3ðsÞ þ H2ðgÞ $ 2Fe3 O4ðsÞ þ H2 OðgÞ
ð3Þ
As the calcination temperature increases, the third weight loss about 0.6% occurs in the temperature range of 773–953 K, which is caused by the formation of iron oxide from magnetite (Munteanu et al., 1997).
Fe3 O4ðsÞ þ H2ðgÞ $ 3FeOðsÞ þ H2 OðgÞ
ð4Þ
Y.-H. Lin et al. / Chemosphere 121 (2015) 62–67
60
30
100.0 TG DTG DSC
40
20
0
TG (%)
DSC (mV)
99.5
25 20
Fe2O3
Fe3O4
Fe3O4
FeO 15
99.0 10
DTG (mg/min)
64
5
98.5
-20 0
-40
-5
98.0 400
600
800
1000
Temperature (K) Fig. 1. Temperature-programmed reduction profiles of a fresh 10 wt.% Fe2O3/SiO2 sorbent with 15% H2 in N2.
The summation of theoretical weight losses of the sorbent calculated from Eqs. (3) and (4) is 1.0% because the weight ratio of Fe2O3 to the sorbent is 10%. The summation of weight losses of two reduction steps obtained from the TGA curve is also 1.0%. Therefore, the weight loss observed is consistent with the weight loss calculated theoretically.
unfavorable influence on the removal of H2S (Ko et al., 2006a). The equilibrium constants for sulfidation reaction with various iron oxides were calculated based on thermodynamics, as shown in our previous study (Ko et al., 2006a). The sulfidation reactions are expressed as follows:
Fe2 O3ðgÞ þ 2H2 SðgÞ þ H2ðgÞ $ 2FeSðsÞ þ 3H2 OðgÞ
ð5Þ
3.2. Blank experiments
Fe3 O4ðsÞ þ 3H2 SðgÞ þ H2ðgÞ $ 3FeSðsÞ þ 4H2 OðgÞ
ð6Þ
A blank experiment was conducted to estimate the reaction of H2S with sorbents in absence of both CO and H2. Besides, a couple of blank experiments were also performed to study the possibility of homogeneous gas-phase reactions. The outlet gases of blank experiments with H2S sorption were measured by using real-time FTIR spectrometer. Fig. 2a shows the blank experiment results with H2S sorption by 10 wt.% Fe2O3/SiO2 sorbents at 773 K in absence of both CO and H2. In the beginning, there are no H2S peaks formed in the FTIR spectroscopy. As the reaction proceeds, significant peaks of H2S (peaks at 1320–1430 cm1) are observed after 0.7 min min1. The concentration of H2S increases as the experiment proceeds, which indicates that the breakthrough of H2S occurs within a time range of 0.9–1.1 min min1. From the XRD analysis (not shown), FeS has been identified in the reacted sorbents. Therefore, H2S can react with Fe2O3 to form FeS in absence of both CO and H2. A couple of FTIR spectroscopy experiments in the empty reactor without sorbents were also performed. The results are shown in Fig. 2b and c. Due to the high concentration of inlet CO, the scales of IR absorbance for H2S are magnified 5 times in the corresponding boxes of Fig. 2b and c. The existence of COS (at 2050–2070 cm1) right after the reaction begins and the slightly decreasing absorbance of H2S in Fig. 2b confirms the gas-phase reaction of CO with H2S. The influences of H2 on the gas-phase reaction of CO with H2S are displayed in Fig. 2c, and the results show the similar profiles and peak intensities as Fig. 2b.
As in our previous study, sulfidation equilibria of Fe3O4 and Fe2O3 are vastly superior to that of FeO (Ko et al., 2006a). Recent studies have indicated that the optimum desulfurization temperature is controlled between 523 and 823 K (Ko et al., 2006a; Dong et al., 2013). In this study, the optimal temperature for the removal of H2S in an IGCC system is around 773 K. Therefore, the sulfidation reactions combined Eqs. (3) and (6) to give Eq. (5), which indicates Fe2O3 may initially be reduced to Fe3O4, followed by a sulfidation reaction to form FeS in a reduction atmosphere. The reduction of Fe2O3 leads to form FeS accompanied with the production of H2O. At around 0.9 min min1 the spectra of H2O vanish, which could be concluded that the breakthrough of H2S takes place within a dimensionless time range of 0.9–1.1 min min1. In addition, it can be shown in Fig. 2d, CO2, COS and CS2 are not produced while CO is not in the system. The formation of Fe may be derived in Eq. (7).
3.3. Effect of hydrogen The effect of hydrogen on H2S removal without CO at 773 K is illustrated in Fig. 2d. In the beginning, some H2O molecules (peaks at 1400–1800 and 3550–3900 cm1) are produced in the system. The decreasing absorbance of H2O along the dimensionless time is assumed to be related to the exhaustion of iron oxides contained in the sorbent. Fe2O3 in the Fe2O3/SiO2 sorbents is able to be reduced to Fe3O4 or FeO in a reducing atmosphere at high temperatures (>773 K) (Ko et al., 2006b). FeO has been observed having an
FeSðsÞ þ H2ðgÞ $ FeðsÞ þ H2 SðgÞ ;
DG 773 K ¼ 58:8 kJ mol
The equilibrium constant of Eq. (7) is 1.07 10
1
ð7Þ
4
.
3.4. Effect of carbon monoxide Fig. 2e displays the FTIR spectroscopy of products and intermediates as a function of dimensionless time (t/t⁄) for H2S removal at 25% carbon monoxide without hydrogen. Due to the low absorbance of CS2 and H2O (and CH4 in Fig. 2f), the spectra of them are magnified in favour of the observation. The scales of IR absorbance are magnified 118 and 16 times in the corresponding boxes for Fig. 2e and f, respectively. In the beginning, a great amount of CO2 (peaks at 2340–2360 cm1) and some H2O (peaks at 3550– 3770 cm1) are produced. The decreasing absorbance of CO2 and H2O with dimensionless time is assumed to be related to the exhaustion of Fe2O3 contained in the sorbent. Therefore, reaction of Fe2O3 with H2S may be derived as Eq. (8), which indicates the reduction of Fe2O3 in the sorbent leading to form FeS accompanied with the production of CO2 and H2O. The FTIR spectra of the
65
Y.-H. Lin et al. / Chemosphere 121 (2015) 62–67
(a)
(d) H2O
0.1 min min-1
H2O 0.1 min min-1 0.2 min min-1
0.4 min min-1 0.7 min
H2S
min-1
0.4 min min-1
Absorbance
Absorbance
0.2 min min-1
0.7 min min-1 0.9 min min-1
-1 0.9 min min-1
1.1 min min-1 -1 1.1 min min-1
4000
(b)
3500
3000
2500
2000
1500
4000
COS
3500
3000
(e)
H2S
CO 0.1 min min-1
1000
2500
CO2
H2O
2000
CO
1000
0.1 min min-1
0.2 min min-1
0.2 min min-1
Absorbance
Absorbance
1500
0.4 min min-1 0.7 min min-1
0.4 min min-1 0.7 min min-1
COS
CS2
0.9 min min-1
0.9 min min-1
CS2
1.1 min min-1
1.1 min min-1
4000
3500
3000
2500
CO
(c) 0.1 min
2000
1500
1000 4000
H2S
COS
3000
2500
2000
CO
0.7 min min-1
0.1 min min-1
0.5 min min-1
COS
0.9 min min-1
CH4 3500
3000
0.9 min min-1 1.0 min min-1
1.1 min min-1
4000
1000
0.3 min min-1
Absorbance
0.4 min min-1
1500
CO2
H2 O
0.2 min min-1
Absorbance
3500
(c) (f)
min-1
2500
2000
1500
1000
4000
3500
1.1 min min-1
CS2
3000
CS2 2500
2000
1500
1000
Wavenumber (cm -1)
Wavenumber (cm -1)
Fig. 2. Identification of products and by-products as a function of dimensionless time during the sorption process by a FTIR spectroscopy: (a) inlet H2S = 1%, with 10 wt.% Fe2O3/SiO2 sorbents, (b) inlet H2S = 1%, CO = 25%, without sorbents, (c) inlet H2S = 1%, CO = 25%, H2 = 15%, without sorbents, (d) inlet H2S = 1%, H2 = 15%, with sorbents, (e) inlet H2S = 1%, CO = 25%, with sorbents, and (f) inlet H2S = 1%, H2 = 15%, CO = 25%, with sorbents.
various dimensionless times were operated at room temperature. When the hot deleterious gases passed through the pipe to reach FTIR detection devices, water vapors would condense in the pipe. Therefore, the detectable of CO2 is more than H2O. With the coexistence of CO and H2O, CO2 and H2 may be produced by the water–gas shift reaction with a decrease of H2O as shown in Eq. (9).
within a dimensionless time range of 0.7–1.1 min min1. The increasing absorbance of COS and CS2 along the dimensionless time indicates that the breakthrough of H2S is gradually achieved as the experiment proceeds. The formation of COS may be derived in Eqs. (10)–(13). Gas reaction:
Fe2 O3ðsÞ þ 2H2 SðgÞ þ COðgÞ $ 2FeSðsÞ þ 2H2 OðgÞ þ CO2ðgÞ
H2 SðgÞ þ COðgÞ $ COSðgÞ þ H2ðgÞ ;
ð8Þ
1
DG 773 K ¼ 21:9 kJ mol
ð10Þ COðgÞ þ H2 OðgÞ $ H2ðgÞ þ CO2ðgÞ 1
ð9Þ
At around 0.7 min min the spectra of COS (at 2050– 2070 cm1) and CS2 (at 1490–1550 cm1) can be observed, which could be concluded that the breakthrough of H2S takes place
H2 SðgÞ þ CO2ðgÞ $ COSðgÞ þ H2 OðgÞ ;
1
DG 773 K ¼ 32:3 kJ mol
ð11Þ Gas–solid reaction:
66
Y.-H. Lin et al. / Chemosphere 121 (2015) 62–67
FeSðsÞ þ COðgÞ $ FeðsÞ þ COSðgÞ ;
1
DG 773 K ¼ 80:6 kJ mol
FeSðsÞ þ CO2ðgÞ $ COSðgÞ þ FeOðsÞ ;
ð12Þ 1
DG 773 K ¼ 75:1 kJ mol
ð13Þ Our previous study has demonstrated that the formation possibility of COS in the gas reaction is higher than that in the gas–solid reaction (Ko et al., 2006a). The result can be explained by the calculation of the thermodynamic equilibrium constants (Barin et al., 1993). The equilibrium constants of reactions (10)–(13) are 3.33 102, 6.61 103, 3.55 106 and 8.42 106, respectively (Chen, 2010). These values imply that the formations of COS of the gas reactions are more favorable than those of the gas–solid reactions, and mostly from the reaction of H2S and CO. Clark et al. (2001) indicated that the formation of COS by Eq. (10) dominated at temperature below 1273 K. The equilibrium constant of Eq. (10) at 773 K is much greater than that of Eq. (11) confirms their conclusion. According to Eq. (10), the formation of COS is preferred when the system has plenty of CO but is absent from H2. However, since the COS peaks appear only after 0.7 min min1, metal Fe may need to be presented as a catalyst for the gas reaction after the time needed for the reduction of Fe2O3 through Fe3O4 and FeO to Fe (Chang et al., 2009) or the reduction of FeS to Fe according to Eqs. (7) and (12). Clark et al. (2001) also proposed two pathways for the formation of CS2 as following:
H2 SðgÞ þ COSðgÞ $ CS2ðgÞ þ H2 OðgÞ
ð14Þ
2COSðgÞ $ CS2ðgÞ þ CO2ðgÞ
ð15Þ
The increasing absorbance of CS2 after the formation of COS as shown in Fig. 2e is consistent with Eqs. (14) and (15). 3.5. Combined-effects of CO and H2 In order to further investigate the combination effect of CO and H2 on H2S removal at 773 K, a series of FTIR observations have been performed and the results are shown in Fig. 2f. It can be found that CO2 only appears in the beginning. At the moment, CO2 may be produced by the desulfurization reaction of Eq. (8). After a dimensionless time around 0.3 min min1, CO2 disappears. This may be due to that the desulfurization reaction of Eq. (5) prevails during this stage because a part of Fe2O3 may be reduced to Fe3O4, which may not react with CO quick enough compared with H2. At around
0.9 min min1 the spectrum of COS appears. The increasing absorbance of COS along the dimensionless time is assumed to be related to the exhaustion of Fe2O3. It could be concluded that the breakthrough of H2S takes place within a dimensionless time range of 0.9–1.1 min min1. Due to the low absorbance of CS2 and CH4 (at 3010–3020 cm1), the spectra are magnified in favour of the observation. Consisting with the results of Section 3.4, CS2 appears after COS. However, a new species, CH4, is also produced in this combined CO and H2 system which is different from that of the CO only system. This may happen when Fe2O3 is reduced further to Fe and a part of CO is reduced by H2 to C after the breakthrough of H2S (Zhao et al., 2002; Chang et al., 2009). The reaction can be shown as Eq. (16). From the elemental analysis, it is found that the C contents of reacted sorbents are 0.060, 0.086, and 0.143 wt.% for 0%, 25%, and 40% CO conditions, respectively.
COðgÞ þ H2ðgÞ $ CðsÞ þ H2 OðgÞ
Therefore, the reaction of Fe with C in this condition may be derived as Eq. (17) (Zhao et al., 2002; Chang et al., 2009).
3FeðsÞ þ CðsÞ $ Fe3 CðsÞ
COðgÞ þ 3H2ðgÞ $ CH4ðgÞ þ H2 OðgÞ
In order to further confirm the reactions derived above, a series of breakthrough determination of H2S and COS were conducted during the desulfurization processes performed in Sections 3.2– 3.5. The results are shown in Fig. 3. For the case of CO only condition, COS appears after 0.7 min min1, which is consistent with the results shown in Fig. 2e. However, breakthrough of H2S is not observed until 1.0 min min1. Using the equilibrium constants of Eqs. (10) and (11), the equilibrium concentration of COS for Eq. (10) can be calculated as 6752 ppm for the CO only case, which is much greater than that for Eq. (11). From Fig. 3, it can be found that the equilibrium concentration of COS after the breakthrough of H2S for the CO only case is around 7150 ppm. Therefore, it could be concluded that Eq. (10) is the dominating pathway to produce COS, which is in concordance with the findings of Clark et al. (2001). For the case of H2 only condition, COS is not obtained and the breakthrough of H2S can be observed at around 0.9 min min1. This is in conformity with the finding in Fig. 2d. In the case of CO coexisting with H2 condition, the breakthrough
Concentration of H2S & COS (ppm)
6000 4000 2000 0 0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
-1
Dimensionless time, t/t* (min min ) Fig. 3. Breakthrough curves for H2S removal using 10 wt.% Fe2O3/SiO2 sorbents at various CO and H2 concentrations.
ð18Þ
3.6. Breakthrough characteristics of the sorbents
H2S (CO = 0%, H2= 0%) H2S (H2 = 15%) H2S (CO = 25%) COS (CO = 25%) H2S (CO = 25%, H2 = 15%) COS (CO = 25%, H2 = 15%)
8000
ð17Þ
The production of CH4 may take place by the reaction of CO and H2 as shown in Eq. (18) (Seo et al., 2011).
12000 10000
ð16Þ
Fig. 4. The Pathways for the reaction of H2S with Fe2O3.
Y.-H. Lin et al. / Chemosphere 121 (2015) 62–67
of both H2S and COS is observed at around 0.9 min min1, which is consistent with Fig. 2f. This is different from the finding for the case of CO only condition, which COS appears before H2S. Therefore, it could be concluded that the formation of COS through Eqs. (10) and (11) is faster in the system while H2 is not present. The equilibrium concentration of COS for Eq. (10) can be calculated as 568 ppm for the CO and H2 coexistence case, which is also much greater than that for Eq. (11). From Fig. 3, the equilibrium concentration of COS after the breakthrough of H2S is found at around 660 ppm. Again, it is confirmed that Eq. (10) dominates the pathways to form COS. 3.7. Pathways for reactions of H2S with Fe2O3/SiO2 sorbents With the results of Sections 3.2–3.6, an overall reaction mechanism of the Fe2O3/SiO2 sorbent in a sulfur-containing coal gas system can be summarized in Fig. 4. The solid arrows represent the reactions of H2S, Fe2O3, FeS, COS, H2O, Fe, C, or CO individually with other compounds. The dashed arrows represent the pathways of the reactions of both H2S and Fe2O3, both Fe3O4 and H2S, and both FeS and CO2, respectively, with other compounds. It can be found that both CO and H2 play important roles, which affect the breakthrough time of H2S and COS. The water–gas shift reaction and the formation of Fe3C by the carbon deposit may enhance or deactivate the sorbents, respectively. Besides, CS2 and CH4 may also be produced after the breakthrough of H2S. 4. Conclusion This study evaluated the influence of CO and H2 concentrations on the removal of H2S in a coal gas by the Fe2O3/SiO2 sorbents. The experiments were conducted in a fixed-bed reactor with various CO and H2 concentrations. Through the FTIR and GC-FPD analyses, an overall reaction mechanism of this system has been developed. Both CO and H2 are found playing important roles, which affect the breakthrough time of H2S and a by-product COS. By-products including COS, CS2, CH4, and Fe3C are also found having vital influence on the reaction. The major pathways in the formation of FeS from the desulfurization process by the reaction of Fe2O3 with H2S have been successfully established. Acknowledgement This study was funded in part by the National Science Council, Republic of China, under Grant NSC 102-3113-P-006-002. References Barin, I., Sauert, F., Schultze-Rhonhof, E., Sheng, W.S., 1993. Thermochemical Data of Pure Substances. VCH Weinheim. Bukhtiyarova, G., Delii, I., Sakaeva, N., Kaichev, V., Plyasova, L., Bukhtiyarov, V., 2007. Effect of the calcination temperature on the properties of Fe2O3/SiO2 catalysts for oxidation of hydrogen sulfide. React. Kinet. Catal. Lett. 92, 89–97.
67
Chang, L.P., Zhang, Z.Y., Ren, X.R., Li, F., Xie, K.C., 2009. Study on the stability of sorbents removing H2S from hot coal gas. Energy Fuels 23, 762–765. Chen, Y.C., 2010. High-Temperature Sorption of Hydrogen Sulfide and Hydrogen Chloride by Fe2O3/SiO2. Master Thesis, National Cheng Kung University, Taiwan. Clark, P.D., Dowling, N.I., Huang, M., Svrcek, W.Y., Monnery, W.D., 2001. Mechanisms of CO and COS formation in the Claus furnace. Ind. Eng. Chem. Res. 40, 497–508. Dong, Y., Ren, X., Wang, M., He, Q., Chang, L., Bao, W., 2013. Effect of impregnation methods on sorbents made from lignite for desulfurization at middle temperature. J. Energy Chem. 22, 783–789. Fan, H.L., Shangguan, J., Liang, L.T., Li, C.H., Lin, J.Y., 2013a. A comparative study of the effect of clay binders on iron oxide sorbent in the high-temperature removal of hydrogen sulfide. Process Saf. Environ. Prot. 91, 235–243. Fan, H.L., Sun, T., Zhao, Y.P., Shangguan, J., Lin, J.Y., 2013b. Three-dimensionally ordered macroporous iron oxide for removal of H2S at medium temperatures. Environ. Sci. Technol. 47, 4859–4865. Huang, J.Y., 2007. High-Temperature Sorption of Hydrogen Sulfide and Carbonyl Sulfide by Fe2O3/SiO2. Master Thesis, National Cheng Kung University, Taiwan. Hulteberg, P.C., Brandin, J.G.M., Silversand, F.A., Lundberg, M., 2005. Preferential oxidation of carbon monoxide on mounted and unmounted noble-metal catalysts in hydrogen-rich streams. Int. J. Hydrogen Energy 30, 1235–1242. Jung, S., Lee, S., Jun, H., Kim, J., 2013. A study of the Zn-based desulfurization sorbents for H2S removal in the IGCC. Catal. Surv. Asia 17, 85–102. Ko, T.H., Chu, H., Lin, H.P., Peng, C.Y., 2006a. Red soil as a regenerable sorbent for high temperature removal of hydrogen sulfide from coal gas. J. Hazard. Mater. 136, 776–783. Ko, T.H., Chu, H., Tseng, J.J., 2006b. Feasibility study on high-temperature sorption of hydrogen sulfide by natural soils. Chemosphere 64, 881–891. Lee, J., Feng, B., 2012. A thermodynamic study of the removal of HCl and H2S from syngas. Front. Chem. Sci. Eng. 6, 67–83. Li, Z., Flytzani-Stephanopoulos, M., 1997. Cu–Cr–O and Cu–Ce–O regenerable oxide sorbents for hot gas desulfurization. Ind. Eng. Chem. Res. 36, 187–196. Monazam, E.R., Siriwardane, R., Breault, R.W., Tian, H., Shadle, L.J., Richards, G., Carpenter, S., 2012. Kinetics of the reduction of CuO/bentonite by methane (CH4) during chemical looping combustion. Energy Fuels 26, 2779–2785. Munteanu, G., Ilieva, L., Andreeva, D., 1997. Kinetic parameters obtained from TPR data for a-Fe2O3 and Au/a-Fe2O3 systems. Thermochim. Acta 291, 171–177. Pineau, A., Kanari, N., Gaballah, I., 2006. Kinetics of reduction of iron oxides by H2: Part I: Low temperature reduction of hematite. Thermochim. Acta 447, 89–100. Seo, H.K., Park, S., Lee, J., Kim, M., Chung, S.W., Chung, J.H., Kim, K., 2011. Effects of operating factors in the coal gasification reaction. Korean J. Chem. Eng. 28, 1851–1858. Song, H., Shah, K., Doroodchi, E., Wall, T., Moghtaderi, B., 2013. Analysis on chemical reaction kinetics of CuO/SiO2 oxygen carriers for chemical looping air separation. Energy Fuels 28, 173–182. Tan, Z., Su, S., Qiu, J., Kong, F., Wang, Z., Hao, F., Xiang, J., 2012. Preparation and characterization of Fe2O3–SiO2 composite and its effect on elemental mercury removal. Chem. Eng. J. 195–196, 218–225. Tseng, T.K., Chang, H.C., Chu, H., Chen, H.T., 2008. Hydrogen sulfide removal from coal gas by the metal-ferrite sorbents made from the heavy metal wastewater sludge. J. Hazard. Mater. 160, 482–488. Wan, Z.Y., Liu, B.S., Zhang, F.M., Zhao, X.H., 2011. Characterization and performance of LaxFeyOz/MCM-41 sorbents during hot coal gas desulfurization. Chem. Eng. J. 171, 594–602. Wang, D., Yu, J., Chang, L., Wang, D., 2011. Effects of addition of Mo on the sulfidation properties of Fe-based sorbents supported on fly ash during hot coal gas desulfurization. Chem. Eng. J. 166, 362–367. Zhang, F.M., Liu, B.S., Zhang, Y., Guo, Y.H., Wan, Z.Y., Subhan, F., 2012. Highly stable and regenerable Mn-based/SBA-15 sorbents for desulfurization of hot coal gas. J. Hazard. Mater. 233–234, 219–227. Zhang, Y., Liu, B.S., Zhang, F.M., Zhang, Z.F., 2013. Formation of (FexMn2x)O3 solid solution and high sulfur capacity properties of Mn-based/M41 sorbents for hot coal gas desulfurization. J. Hazard. Mater. 248–249, 81–88. Zhao, J., Huang, J., Zhang, J., Wang, Y., 2002. Influence of fly ash on high temperature desulfurization using iron oxide sorbent. Energy Fuels 16, 1585–1590. Zheng, X., Bao, W., Chang, L., Jin, Q., He, R., Xie, K., 2012. Interaction among metal components of Zn–Mn–Cu-based sorbents prepared by high-pressure impregnation method and its effect on the removal of H2S from hot coal gas. Energy Fuels 26, 3393–3398.
Contents lists available at ScienceDirect
Chemosphere journal homepage: www.elsevier.com/locate/chemosphere
The mechanism of coal gas desulfurization by iron oxide sorbents Yi-Hsing Lin a, Yen-Chiao Chen a, Hsin Chu b,⇑ a b
Department of Environmental Engineering, National Cheng Kung University, 1 University Road, Tainan 701, Taiwan Department of Environmental Engineering and Research Center for Energy Technology and Strategy, National Cheng Kung University, 1 University Road, Tainan 701, Taiwan
h i g h l i g h t s
g r a p h i c a l a b s t r a c t
The existence of CO and H2 affects
equilibrium concentrations of H2S and COS. The presence of H2 has a negative effect on H2S removal. By-products including COS, CS2, CH4, and Fe3C have vital influence on the reaction. The pathways of Fe2O3 reacting with H2S were successfully established. The major route of the reaction mechanism is Fe2O3 reacting with H2S to form FeS.
a r t i c l e
i n f o
Article history: Received 9 June 2014 Received in revised form 30 October 2014 Accepted 6 November 2014 Available online 27 November 2014 Handling Editor: O. Hao Keywords: Sorbent Fe2O3 H2S IGCC Mechanism
Fe3C CO2+ CS2
CO
+
CO2 + H2
H2+COS
Fe+COS
CO
CO
H2O + CS2
COS
CH4+H2O COS+FeO
CO
H2S + Fe2O3 CO2
H2O+COS
CO 773K
FeS + CO2 + H2O
H2
Fe3O4 + H2S H2
FeS + H2O H2
Fe + H2S
a b s t r a c t This study aims to understand the roles of hydrogen and carbon monoxide during the desulfurization process in a coal gasification system that H2S of the syngas was removed by Fe2O3/SiO2 sorbents. The Fe2O3/SiO2 sorbents were prepared by incipient wetness impregnation. Through the breakthrough experiments and Fourier transform infrared spectroscopy analyses, the overall desulfurization mechanism of the Fe2O3/SiO2 sorbents was proposed in this study. The results show that the major reaction route is that Fe2O3 reacts with H2S to form FeS, and the existence of CO and H2 in the simulated gas significantly affects equilibrium concentrations of H2S and COS. The formation of COS occurs when the feeding gas is blended with CO and H2S, or CO2 and H2S. The pathways in the formation of products from the desulfurization process by the reaction of Fe2O3 with H2S have been successfully established. Ó 2014 Elsevier Ltd. All rights reserved.
1. Introduction Integrated Gasification Combined Cycle (IGCC) is one of the most ideal technologies with high power generation efficiency and environmental friendly performance (Song et al., 2013; Fan et al., 2013b). Coal currently is the most abundant fossil fuel with relatively low and constant price. However, reserves of fossil fuels have been limited and global warming crisis is increasing ⇑ Corresponding author. Tel.: +886 6 208 0108; fax: +886 6 275 2790. E-mail address: [email protected] (H. Chu). http://dx.doi.org/10.1016/j.chemosphere.2014.11.010 0045-6535/Ó 2014 Elsevier Ltd. All rights reserved.
H2
H2
H2O+C
(Monazam et al., 2012). Therefore, the technology improvement for energy issues has become a trend nowadays, especially coal poly-generation with IGCC. Sulfur contained in coal will be emitted to typically form 0.2–3 vol% H2S and lesser amounts of other sulfur compounds during the process of gasification (Monazam et al., 2012). Long-term exposure to these sulfur-containing species may cause the components of IGCC system severely corrosive damage. Therefore, multiple impurities must be removed to comply with environmental laws and regulations for protecting the public health and reducing maintenance costs. Nowadays, almost all commercial IGCC power plants use wet desulfurization processes to
Y.-H. Lin et al. / Chemosphere 121 (2015) 62–67
remove H2S and other deleterious gases from hot syngas by the use of large amount of water, resulting in the decreasing thermal efficiency of the system (Lee and Feng, 2012). Consequently, the development of dry desulfurization at high temperature becomes vitally crucial in this field, and can improve the thermal efficiency and reduce cost of capital without the procedure of coal gas cooling and heating (Wan et al., 2011). Several researchers have reported on the important role of metal sorbents on the performance of hot-gas desulfurization (Jung et al., 2013; Zhang et al., 2013; Fan et al., 2013a). The sorbents can be divided into noble metal sorbents (Ru and Rh) (Hulteberg et al., 2005) and base metal oxide sorbents (Mn, Cu, Fe, Zn) (Li and Flytzani-Stephanopoulos, 1997; Wan et al., 2011; Zhang et al., 2012; Zheng et al., 2012). Iron oxide is the most common material as a sorbent for hot gas desulfurization under a reducing atmosphere with hydrogen and carbon monoxide due to its high absorption capacity, abundant resources, lower price, excellent physical strength and easy regeneration (Wang et al., 2011; Fan et al., 2013a). In our previous study (Tseng et al., 2008), the Fe-based sorbents emerged as the leading candidates for hot coal gas desulfurization through thermodynamic screening. In this study, the influences of CO and H2 on the performance of 10 wt.% Fe2O3/SiO2 sorbents for H2S removal were evaluated in various simulated syngas. The syngas (H2S) before and after the reactor were recorded by a gas chromatograph (Shimadzu, GC14B) equipped with a flame photometry detector (FPD) and fitted with a GS-Q capillary column. The outlet gases were recorded through on-line Fourier transform infrared (FTIR) spectroscopy and GC-FPD. A further investigation into the desulfurization mechanism of the sorbents was also conducted. The results will provide useful information for the designs and applications of the sorbents in the IGCC system. 2. Experimental 2.1. Sorbent preparation The Fe2O3/SiO2 sorbents were prepared by incipient wetness impregnation. Commercial SiO2 spherical pellets (Alfa Aesar, stock #44740) were used as sorbent supports. On the basis of the predetermined contents of iron loading from our previous study Huang (2007) and Bukhtiyarova et al. (2007), 5.63 g (Fe(NO3)39H2O) was used as the precursors of Fe dissolved in 16 mL de-ionized water and then mixed with the 10 g SiO2 supports. The impregnated materials were kept at room temperature for 12 h to let solution fully penetrate into the pore of supports. Subsequently, the impregnated sorbents were dried in an oven at 393 K for 24 h. Finally, the impregnated sorbents were conducted in a calcination process at 973 K for 8 h in airflow conditions to become 10 wt.% Fe2O3/SiO2 sorbents. 2.2. Sorbent characterization In the present work, temperature-programmed reduction (TPR) with hydrogen was conducted in a TG/DTA (Pyris Diamond TG/ DTA, Perkin Elmer) to determine the phase transformation in the temperature range from room temperature to 1073 K with a heating rate of 50 K min1. Elemental analyses (EA) were provided with an Elementar Vario EL-III elemental analyzer. 2.3. Desulfurization experiment The H2S removal experiments were carried out in a bench-scale fixed-bed reactor (a quartz tube with 1.5 cm id, 1.8 cm od, and 85 cm long) with 10 000 ppm inlet H2S at atmospheric pressure
63
and 773 K. A thin layer of #200 mesh frit quartz disk was set in the reactor 45 cm below the top of the tube to support the catalyst and uniformly distribute the gas stream. The weight of the catalyst packing was 1 g (thickness 0.82 cm). The weight hourly space velocity (WHSV) was set at 2000 mL h1 g1. 773 K and 2000 mL h1 g1 were chosen because it provided the best performance for H2S removal according to previous studies (Ko et al., 2006a; Tseng et al., 2008). The experimental apparatus included a simulated syngas system, a high-temperature reaction system, and a gas analysis system. The simulated syngas had a typical composition of 1 vol% H2S, 25 vol% CO and/or 15 vol% H2, and N2 as a balance gas, supplied from gas cylinders, mixed in a mixer to ensure the gas mixture was well mixed, and then introduced to the reaction system. The high-temperature reaction system was consisted of a bench-scale fixed-bed reactor, an electrical furnace, a temperature controller (TTE, TM-4800), and a K-type thermocouple. The gas analysis system was composed of a GC-FPD and a FTIR. Diffuse reflectance FTIR spectra of outlet gases were measured by using a Perkin–Elmer Spectrum One FTIR spectrometer. The recorded IR spectra were operating in the spectral range 4000–700 cm1 at a resolution of 4 cm1 with the use of Time Base 2.0 software. The desulfurization capacity of Fe2O3/SiO2 sorbents was calculated by the equation of sorbent utilization (SU):
Rt SUð%Þ ¼
0
ðC in C out Þdt 100% Cin t
ð1Þ
where Cin and Cout are the inlet and outlet concentrations of H2S (ppm), respectively; t is the experimental breakthrough time (min), and t⁄ is the theoretical breakthrough time (min) of the process. The experimental breakthrough time was defined as the time from the beginning of the experiment to the point when the outlet H2S concentration reached 100 ppm, which is 1% of the inlet H2S concentration. The theoretical breakthrough time (min) was calculated by the equation:
WX t ¼
M
A
F
ð2Þ
where W is the weight of the sorbent placed in the reactor (g), X is the actual weight loading of iron oxide supported on SiO2 (g g1), M is the molecular weight of the iron oxide (g mol1), A is the moles of H2S that can be absorbed by one mole of iron oxide (mol mol1), and F is the molar flow rate of the inlet H2S (mol min1). 3. Results and discussion 3.1. Temperature-programmed reduction The TG/DTA curves as a function of temperature for the 10 wt.% Fe2O3/SiO2 sorbent with 15% H2 in N2 are displayed in Fig. 1. In the TG curve, the sorbent can be assigned into three stages. First, a small weight loss (0.5%) of the sorbent accompanies an exothermic process at about 323–473 K, which is attributed to the evaporation of adsorbed water (Tan et al., 2012). The weight loss (0.4%) in the second stage can be attained at 693–773 K that is due to the reduction of hematite to magnetite (Pineau et al., 2006).
3Fe2 O3ðsÞ þ H2ðgÞ $ 2Fe3 O4ðsÞ þ H2 OðgÞ
ð3Þ
As the calcination temperature increases, the third weight loss about 0.6% occurs in the temperature range of 773–953 K, which is caused by the formation of iron oxide from magnetite (Munteanu et al., 1997).
Fe3 O4ðsÞ þ H2ðgÞ $ 3FeOðsÞ þ H2 OðgÞ
ð4Þ
Y.-H. Lin et al. / Chemosphere 121 (2015) 62–67
60
30
100.0 TG DTG DSC
40
20
0
TG (%)
DSC (mV)
99.5
25 20
Fe2O3
Fe3O4
Fe3O4
FeO 15
99.0 10
DTG (mg/min)
64
5
98.5
-20 0
-40
-5
98.0 400
600
800
1000
Temperature (K) Fig. 1. Temperature-programmed reduction profiles of a fresh 10 wt.% Fe2O3/SiO2 sorbent with 15% H2 in N2.
The summation of theoretical weight losses of the sorbent calculated from Eqs. (3) and (4) is 1.0% because the weight ratio of Fe2O3 to the sorbent is 10%. The summation of weight losses of two reduction steps obtained from the TGA curve is also 1.0%. Therefore, the weight loss observed is consistent with the weight loss calculated theoretically.
unfavorable influence on the removal of H2S (Ko et al., 2006a). The equilibrium constants for sulfidation reaction with various iron oxides were calculated based on thermodynamics, as shown in our previous study (Ko et al., 2006a). The sulfidation reactions are expressed as follows:
Fe2 O3ðgÞ þ 2H2 SðgÞ þ H2ðgÞ $ 2FeSðsÞ þ 3H2 OðgÞ
ð5Þ
3.2. Blank experiments
Fe3 O4ðsÞ þ 3H2 SðgÞ þ H2ðgÞ $ 3FeSðsÞ þ 4H2 OðgÞ
ð6Þ
A blank experiment was conducted to estimate the reaction of H2S with sorbents in absence of both CO and H2. Besides, a couple of blank experiments were also performed to study the possibility of homogeneous gas-phase reactions. The outlet gases of blank experiments with H2S sorption were measured by using real-time FTIR spectrometer. Fig. 2a shows the blank experiment results with H2S sorption by 10 wt.% Fe2O3/SiO2 sorbents at 773 K in absence of both CO and H2. In the beginning, there are no H2S peaks formed in the FTIR spectroscopy. As the reaction proceeds, significant peaks of H2S (peaks at 1320–1430 cm1) are observed after 0.7 min min1. The concentration of H2S increases as the experiment proceeds, which indicates that the breakthrough of H2S occurs within a time range of 0.9–1.1 min min1. From the XRD analysis (not shown), FeS has been identified in the reacted sorbents. Therefore, H2S can react with Fe2O3 to form FeS in absence of both CO and H2. A couple of FTIR spectroscopy experiments in the empty reactor without sorbents were also performed. The results are shown in Fig. 2b and c. Due to the high concentration of inlet CO, the scales of IR absorbance for H2S are magnified 5 times in the corresponding boxes of Fig. 2b and c. The existence of COS (at 2050–2070 cm1) right after the reaction begins and the slightly decreasing absorbance of H2S in Fig. 2b confirms the gas-phase reaction of CO with H2S. The influences of H2 on the gas-phase reaction of CO with H2S are displayed in Fig. 2c, and the results show the similar profiles and peak intensities as Fig. 2b.
As in our previous study, sulfidation equilibria of Fe3O4 and Fe2O3 are vastly superior to that of FeO (Ko et al., 2006a). Recent studies have indicated that the optimum desulfurization temperature is controlled between 523 and 823 K (Ko et al., 2006a; Dong et al., 2013). In this study, the optimal temperature for the removal of H2S in an IGCC system is around 773 K. Therefore, the sulfidation reactions combined Eqs. (3) and (6) to give Eq. (5), which indicates Fe2O3 may initially be reduced to Fe3O4, followed by a sulfidation reaction to form FeS in a reduction atmosphere. The reduction of Fe2O3 leads to form FeS accompanied with the production of H2O. At around 0.9 min min1 the spectra of H2O vanish, which could be concluded that the breakthrough of H2S takes place within a dimensionless time range of 0.9–1.1 min min1. In addition, it can be shown in Fig. 2d, CO2, COS and CS2 are not produced while CO is not in the system. The formation of Fe may be derived in Eq. (7).
3.3. Effect of hydrogen The effect of hydrogen on H2S removal without CO at 773 K is illustrated in Fig. 2d. In the beginning, some H2O molecules (peaks at 1400–1800 and 3550–3900 cm1) are produced in the system. The decreasing absorbance of H2O along the dimensionless time is assumed to be related to the exhaustion of iron oxides contained in the sorbent. Fe2O3 in the Fe2O3/SiO2 sorbents is able to be reduced to Fe3O4 or FeO in a reducing atmosphere at high temperatures (>773 K) (Ko et al., 2006b). FeO has been observed having an
FeSðsÞ þ H2ðgÞ $ FeðsÞ þ H2 SðgÞ ;
DG 773 K ¼ 58:8 kJ mol
The equilibrium constant of Eq. (7) is 1.07 10
1
ð7Þ
4
.
3.4. Effect of carbon monoxide Fig. 2e displays the FTIR spectroscopy of products and intermediates as a function of dimensionless time (t/t⁄) for H2S removal at 25% carbon monoxide without hydrogen. Due to the low absorbance of CS2 and H2O (and CH4 in Fig. 2f), the spectra of them are magnified in favour of the observation. The scales of IR absorbance are magnified 118 and 16 times in the corresponding boxes for Fig. 2e and f, respectively. In the beginning, a great amount of CO2 (peaks at 2340–2360 cm1) and some H2O (peaks at 3550– 3770 cm1) are produced. The decreasing absorbance of CO2 and H2O with dimensionless time is assumed to be related to the exhaustion of Fe2O3 contained in the sorbent. Therefore, reaction of Fe2O3 with H2S may be derived as Eq. (8), which indicates the reduction of Fe2O3 in the sorbent leading to form FeS accompanied with the production of CO2 and H2O. The FTIR spectra of the
65
Y.-H. Lin et al. / Chemosphere 121 (2015) 62–67
(a)
(d) H2O
0.1 min min-1
H2O 0.1 min min-1 0.2 min min-1
0.4 min min-1 0.7 min
H2S
min-1
0.4 min min-1
Absorbance
Absorbance
0.2 min min-1
0.7 min min-1 0.9 min min-1
-1 0.9 min min-1
1.1 min min-1 -1 1.1 min min-1
4000
(b)
3500
3000
2500
2000
1500
4000
COS
3500
3000
(e)
H2S
CO 0.1 min min-1
1000
2500
CO2
H2O
2000
CO
1000
0.1 min min-1
0.2 min min-1
0.2 min min-1
Absorbance
Absorbance
1500
0.4 min min-1 0.7 min min-1
0.4 min min-1 0.7 min min-1
COS
CS2
0.9 min min-1
0.9 min min-1
CS2
1.1 min min-1
1.1 min min-1
4000
3500
3000
2500
CO
(c) 0.1 min
2000
1500
1000 4000
H2S
COS
3000
2500
2000
CO
0.7 min min-1
0.1 min min-1
0.5 min min-1
COS
0.9 min min-1
CH4 3500
3000
0.9 min min-1 1.0 min min-1
1.1 min min-1
4000
1000
0.3 min min-1
Absorbance
0.4 min min-1
1500
CO2
H2 O
0.2 min min-1
Absorbance
3500
(c) (f)
min-1
2500
2000
1500
1000
4000
3500
1.1 min min-1
CS2
3000
CS2 2500
2000
1500
1000
Wavenumber (cm -1)
Wavenumber (cm -1)
Fig. 2. Identification of products and by-products as a function of dimensionless time during the sorption process by a FTIR spectroscopy: (a) inlet H2S = 1%, with 10 wt.% Fe2O3/SiO2 sorbents, (b) inlet H2S = 1%, CO = 25%, without sorbents, (c) inlet H2S = 1%, CO = 25%, H2 = 15%, without sorbents, (d) inlet H2S = 1%, H2 = 15%, with sorbents, (e) inlet H2S = 1%, CO = 25%, with sorbents, and (f) inlet H2S = 1%, H2 = 15%, CO = 25%, with sorbents.
various dimensionless times were operated at room temperature. When the hot deleterious gases passed through the pipe to reach FTIR detection devices, water vapors would condense in the pipe. Therefore, the detectable of CO2 is more than H2O. With the coexistence of CO and H2O, CO2 and H2 may be produced by the water–gas shift reaction with a decrease of H2O as shown in Eq. (9).
within a dimensionless time range of 0.7–1.1 min min1. The increasing absorbance of COS and CS2 along the dimensionless time indicates that the breakthrough of H2S is gradually achieved as the experiment proceeds. The formation of COS may be derived in Eqs. (10)–(13). Gas reaction:
Fe2 O3ðsÞ þ 2H2 SðgÞ þ COðgÞ $ 2FeSðsÞ þ 2H2 OðgÞ þ CO2ðgÞ
H2 SðgÞ þ COðgÞ $ COSðgÞ þ H2ðgÞ ;
ð8Þ
1
DG 773 K ¼ 21:9 kJ mol
ð10Þ COðgÞ þ H2 OðgÞ $ H2ðgÞ þ CO2ðgÞ 1
ð9Þ
At around 0.7 min min the spectra of COS (at 2050– 2070 cm1) and CS2 (at 1490–1550 cm1) can be observed, which could be concluded that the breakthrough of H2S takes place
H2 SðgÞ þ CO2ðgÞ $ COSðgÞ þ H2 OðgÞ ;
1
DG 773 K ¼ 32:3 kJ mol
ð11Þ Gas–solid reaction:
66
Y.-H. Lin et al. / Chemosphere 121 (2015) 62–67
FeSðsÞ þ COðgÞ $ FeðsÞ þ COSðgÞ ;
1
DG 773 K ¼ 80:6 kJ mol
FeSðsÞ þ CO2ðgÞ $ COSðgÞ þ FeOðsÞ ;
ð12Þ 1
DG 773 K ¼ 75:1 kJ mol
ð13Þ Our previous study has demonstrated that the formation possibility of COS in the gas reaction is higher than that in the gas–solid reaction (Ko et al., 2006a). The result can be explained by the calculation of the thermodynamic equilibrium constants (Barin et al., 1993). The equilibrium constants of reactions (10)–(13) are 3.33 102, 6.61 103, 3.55 106 and 8.42 106, respectively (Chen, 2010). These values imply that the formations of COS of the gas reactions are more favorable than those of the gas–solid reactions, and mostly from the reaction of H2S and CO. Clark et al. (2001) indicated that the formation of COS by Eq. (10) dominated at temperature below 1273 K. The equilibrium constant of Eq. (10) at 773 K is much greater than that of Eq. (11) confirms their conclusion. According to Eq. (10), the formation of COS is preferred when the system has plenty of CO but is absent from H2. However, since the COS peaks appear only after 0.7 min min1, metal Fe may need to be presented as a catalyst for the gas reaction after the time needed for the reduction of Fe2O3 through Fe3O4 and FeO to Fe (Chang et al., 2009) or the reduction of FeS to Fe according to Eqs. (7) and (12). Clark et al. (2001) also proposed two pathways for the formation of CS2 as following:
H2 SðgÞ þ COSðgÞ $ CS2ðgÞ þ H2 OðgÞ
ð14Þ
2COSðgÞ $ CS2ðgÞ þ CO2ðgÞ
ð15Þ
The increasing absorbance of CS2 after the formation of COS as shown in Fig. 2e is consistent with Eqs. (14) and (15). 3.5. Combined-effects of CO and H2 In order to further investigate the combination effect of CO and H2 on H2S removal at 773 K, a series of FTIR observations have been performed and the results are shown in Fig. 2f. It can be found that CO2 only appears in the beginning. At the moment, CO2 may be produced by the desulfurization reaction of Eq. (8). After a dimensionless time around 0.3 min min1, CO2 disappears. This may be due to that the desulfurization reaction of Eq. (5) prevails during this stage because a part of Fe2O3 may be reduced to Fe3O4, which may not react with CO quick enough compared with H2. At around
0.9 min min1 the spectrum of COS appears. The increasing absorbance of COS along the dimensionless time is assumed to be related to the exhaustion of Fe2O3. It could be concluded that the breakthrough of H2S takes place within a dimensionless time range of 0.9–1.1 min min1. Due to the low absorbance of CS2 and CH4 (at 3010–3020 cm1), the spectra are magnified in favour of the observation. Consisting with the results of Section 3.4, CS2 appears after COS. However, a new species, CH4, is also produced in this combined CO and H2 system which is different from that of the CO only system. This may happen when Fe2O3 is reduced further to Fe and a part of CO is reduced by H2 to C after the breakthrough of H2S (Zhao et al., 2002; Chang et al., 2009). The reaction can be shown as Eq. (16). From the elemental analysis, it is found that the C contents of reacted sorbents are 0.060, 0.086, and 0.143 wt.% for 0%, 25%, and 40% CO conditions, respectively.
COðgÞ þ H2ðgÞ $ CðsÞ þ H2 OðgÞ
Therefore, the reaction of Fe with C in this condition may be derived as Eq. (17) (Zhao et al., 2002; Chang et al., 2009).
3FeðsÞ þ CðsÞ $ Fe3 CðsÞ
COðgÞ þ 3H2ðgÞ $ CH4ðgÞ þ H2 OðgÞ
In order to further confirm the reactions derived above, a series of breakthrough determination of H2S and COS were conducted during the desulfurization processes performed in Sections 3.2– 3.5. The results are shown in Fig. 3. For the case of CO only condition, COS appears after 0.7 min min1, which is consistent with the results shown in Fig. 2e. However, breakthrough of H2S is not observed until 1.0 min min1. Using the equilibrium constants of Eqs. (10) and (11), the equilibrium concentration of COS for Eq. (10) can be calculated as 6752 ppm for the CO only case, which is much greater than that for Eq. (11). From Fig. 3, it can be found that the equilibrium concentration of COS after the breakthrough of H2S for the CO only case is around 7150 ppm. Therefore, it could be concluded that Eq. (10) is the dominating pathway to produce COS, which is in concordance with the findings of Clark et al. (2001). For the case of H2 only condition, COS is not obtained and the breakthrough of H2S can be observed at around 0.9 min min1. This is in conformity with the finding in Fig. 2d. In the case of CO coexisting with H2 condition, the breakthrough
Concentration of H2S & COS (ppm)
6000 4000 2000 0 0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
-1
Dimensionless time, t/t* (min min ) Fig. 3. Breakthrough curves for H2S removal using 10 wt.% Fe2O3/SiO2 sorbents at various CO and H2 concentrations.
ð18Þ
3.6. Breakthrough characteristics of the sorbents
H2S (CO = 0%, H2= 0%) H2S (H2 = 15%) H2S (CO = 25%) COS (CO = 25%) H2S (CO = 25%, H2 = 15%) COS (CO = 25%, H2 = 15%)
8000
ð17Þ
The production of CH4 may take place by the reaction of CO and H2 as shown in Eq. (18) (Seo et al., 2011).
12000 10000
ð16Þ
Fig. 4. The Pathways for the reaction of H2S with Fe2O3.
Y.-H. Lin et al. / Chemosphere 121 (2015) 62–67
of both H2S and COS is observed at around 0.9 min min1, which is consistent with Fig. 2f. This is different from the finding for the case of CO only condition, which COS appears before H2S. Therefore, it could be concluded that the formation of COS through Eqs. (10) and (11) is faster in the system while H2 is not present. The equilibrium concentration of COS for Eq. (10) can be calculated as 568 ppm for the CO and H2 coexistence case, which is also much greater than that for Eq. (11). From Fig. 3, the equilibrium concentration of COS after the breakthrough of H2S is found at around 660 ppm. Again, it is confirmed that Eq. (10) dominates the pathways to form COS. 3.7. Pathways for reactions of H2S with Fe2O3/SiO2 sorbents With the results of Sections 3.2–3.6, an overall reaction mechanism of the Fe2O3/SiO2 sorbent in a sulfur-containing coal gas system can be summarized in Fig. 4. The solid arrows represent the reactions of H2S, Fe2O3, FeS, COS, H2O, Fe, C, or CO individually with other compounds. The dashed arrows represent the pathways of the reactions of both H2S and Fe2O3, both Fe3O4 and H2S, and both FeS and CO2, respectively, with other compounds. It can be found that both CO and H2 play important roles, which affect the breakthrough time of H2S and COS. The water–gas shift reaction and the formation of Fe3C by the carbon deposit may enhance or deactivate the sorbents, respectively. Besides, CS2 and CH4 may also be produced after the breakthrough of H2S. 4. Conclusion This study evaluated the influence of CO and H2 concentrations on the removal of H2S in a coal gas by the Fe2O3/SiO2 sorbents. The experiments were conducted in a fixed-bed reactor with various CO and H2 concentrations. Through the FTIR and GC-FPD analyses, an overall reaction mechanism of this system has been developed. Both CO and H2 are found playing important roles, which affect the breakthrough time of H2S and a by-product COS. By-products including COS, CS2, CH4, and Fe3C are also found having vital influence on the reaction. The major pathways in the formation of FeS from the desulfurization process by the reaction of Fe2O3 with H2S have been successfully established. Acknowledgement This study was funded in part by the National Science Council, Republic of China, under Grant NSC 102-3113-P-006-002. References Barin, I., Sauert, F., Schultze-Rhonhof, E., Sheng, W.S., 1993. Thermochemical Data of Pure Substances. VCH Weinheim. Bukhtiyarova, G., Delii, I., Sakaeva, N., Kaichev, V., Plyasova, L., Bukhtiyarov, V., 2007. Effect of the calcination temperature on the properties of Fe2O3/SiO2 catalysts for oxidation of hydrogen sulfide. React. Kinet. Catal. Lett. 92, 89–97.
67
Chang, L.P., Zhang, Z.Y., Ren, X.R., Li, F., Xie, K.C., 2009. Study on the stability of sorbents removing H2S from hot coal gas. Energy Fuels 23, 762–765. Chen, Y.C., 2010. High-Temperature Sorption of Hydrogen Sulfide and Hydrogen Chloride by Fe2O3/SiO2. Master Thesis, National Cheng Kung University, Taiwan. Clark, P.D., Dowling, N.I., Huang, M., Svrcek, W.Y., Monnery, W.D., 2001. Mechanisms of CO and COS formation in the Claus furnace. Ind. Eng. Chem. Res. 40, 497–508. Dong, Y., Ren, X., Wang, M., He, Q., Chang, L., Bao, W., 2013. Effect of impregnation methods on sorbents made from lignite for desulfurization at middle temperature. J. Energy Chem. 22, 783–789. Fan, H.L., Shangguan, J., Liang, L.T., Li, C.H., Lin, J.Y., 2013a. A comparative study of the effect of clay binders on iron oxide sorbent in the high-temperature removal of hydrogen sulfide. Process Saf. Environ. Prot. 91, 235–243. Fan, H.L., Sun, T., Zhao, Y.P., Shangguan, J., Lin, J.Y., 2013b. Three-dimensionally ordered macroporous iron oxide for removal of H2S at medium temperatures. Environ. Sci. Technol. 47, 4859–4865. Huang, J.Y., 2007. High-Temperature Sorption of Hydrogen Sulfide and Carbonyl Sulfide by Fe2O3/SiO2. Master Thesis, National Cheng Kung University, Taiwan. Hulteberg, P.C., Brandin, J.G.M., Silversand, F.A., Lundberg, M., 2005. Preferential oxidation of carbon monoxide on mounted and unmounted noble-metal catalysts in hydrogen-rich streams. Int. J. Hydrogen Energy 30, 1235–1242. Jung, S., Lee, S., Jun, H., Kim, J., 2013. A study of the Zn-based desulfurization sorbents for H2S removal in the IGCC. Catal. Surv. Asia 17, 85–102. Ko, T.H., Chu, H., Lin, H.P., Peng, C.Y., 2006a. Red soil as a regenerable sorbent for high temperature removal of hydrogen sulfide from coal gas. J. Hazard. Mater. 136, 776–783. Ko, T.H., Chu, H., Tseng, J.J., 2006b. Feasibility study on high-temperature sorption of hydrogen sulfide by natural soils. Chemosphere 64, 881–891. Lee, J., Feng, B., 2012. A thermodynamic study of the removal of HCl and H2S from syngas. Front. Chem. Sci. Eng. 6, 67–83. Li, Z., Flytzani-Stephanopoulos, M., 1997. Cu–Cr–O and Cu–Ce–O regenerable oxide sorbents for hot gas desulfurization. Ind. Eng. Chem. Res. 36, 187–196. Monazam, E.R., Siriwardane, R., Breault, R.W., Tian, H., Shadle, L.J., Richards, G., Carpenter, S., 2012. Kinetics of the reduction of CuO/bentonite by methane (CH4) during chemical looping combustion. Energy Fuels 26, 2779–2785. Munteanu, G., Ilieva, L., Andreeva, D., 1997. Kinetic parameters obtained from TPR data for a-Fe2O3 and Au/a-Fe2O3 systems. Thermochim. Acta 291, 171–177. Pineau, A., Kanari, N., Gaballah, I., 2006. Kinetics of reduction of iron oxides by H2: Part I: Low temperature reduction of hematite. Thermochim. Acta 447, 89–100. Seo, H.K., Park, S., Lee, J., Kim, M., Chung, S.W., Chung, J.H., Kim, K., 2011. Effects of operating factors in the coal gasification reaction. Korean J. Chem. Eng. 28, 1851–1858. Song, H., Shah, K., Doroodchi, E., Wall, T., Moghtaderi, B., 2013. Analysis on chemical reaction kinetics of CuO/SiO2 oxygen carriers for chemical looping air separation. Energy Fuels 28, 173–182. Tan, Z., Su, S., Qiu, J., Kong, F., Wang, Z., Hao, F., Xiang, J., 2012. Preparation and characterization of Fe2O3–SiO2 composite and its effect on elemental mercury removal. Chem. Eng. J. 195–196, 218–225. Tseng, T.K., Chang, H.C., Chu, H., Chen, H.T., 2008. Hydrogen sulfide removal from coal gas by the metal-ferrite sorbents made from the heavy metal wastewater sludge. J. Hazard. Mater. 160, 482–488. Wan, Z.Y., Liu, B.S., Zhang, F.M., Zhao, X.H., 2011. Characterization and performance of LaxFeyOz/MCM-41 sorbents during hot coal gas desulfurization. Chem. Eng. J. 171, 594–602. Wang, D., Yu, J., Chang, L., Wang, D., 2011. Effects of addition of Mo on the sulfidation properties of Fe-based sorbents supported on fly ash during hot coal gas desulfurization. Chem. Eng. J. 166, 362–367. Zhang, F.M., Liu, B.S., Zhang, Y., Guo, Y.H., Wan, Z.Y., Subhan, F., 2012. Highly stable and regenerable Mn-based/SBA-15 sorbents for desulfurization of hot coal gas. J. Hazard. Mater. 233–234, 219–227. Zhang, Y., Liu, B.S., Zhang, F.M., Zhang, Z.F., 2013. Formation of (FexMn2x)O3 solid solution and high sulfur capacity properties of Mn-based/M41 sorbents for hot coal gas desulfurization. J. Hazard. Mater. 248–249, 81–88. Zhao, J., Huang, J., Zhang, J., Wang, Y., 2002. Influence of fly ash on high temperature desulfurization using iron oxide sorbent. Energy Fuels 16, 1585–1590. Zheng, X., Bao, W., Chang, L., Jin, Q., He, R., Xie, K., 2012. Interaction among metal components of Zn–Mn–Cu-based sorbents prepared by high-pressure impregnation method and its effect on the removal of H2S from hot coal gas. Energy Fuels 26, 3393–3398.
