Zhong Qifan; Xiao Jin,2; Huang Jindi; Yuan Jie; Yu Bailie

(1. School of Metallurgy and Environment, Central South University, Changsha 410083; 2. National Engineering Laboratory of Ef fi cient Utilization of Refractory Nonferrous Metal Resources, Central South University, Changsha 410083)

Desulfurization of Petroleum Coke by Calcination in Ammonia Atmosphere below 1 000 ℃

Zhong Qifan1; Xiao Jin1,2; Huang Jindi1; Yuan Jie1; Yu Bailie1

(1. School of Metallurgy and Environment, Central South University, Changsha 410083; 2. National Engineering Laboratory of Ef fi cient Utilization of Refractory Nonferrous Metal Resources, Central South University, Changsha 410083)

The desulfurization effciency and mechanism of the calcination of petroleum coke in ammonia atmosphere at lower than 1000 ℃ were investigated through a series of conditional experiments and comparison with other gases such as H2. The topics of effciency and reaction mechanism are usually discussed through investigation by means of the Fourier transform infrared spectroscopy (FT-IR), the Brunauer-Emmett-Teller (BET) technique, and the thermogravimetry coupled with the mass spectrometry (TG-MS). Results showed that in addition to H2, ammonia not only could retain a high desulfurization rate but could also reduce coke loss during the desulfurization process of petroleum coke. The best desulfurization conditions covered a petroleum coke particle size of less than 0.1 mm, a calcination temperature of 800 ℃ in ammonia atmosphere with a fow rate of 10 L/h, and a heating duration of more than 120 min. Ammonia decomposition, H2generation, decline in the activation energy of the carbon–sulfur bonds, and petroleum coke with a largest specifc surface area at 800 ℃are the key goals of desulfurization studied thereby. As proved by TG-MS analysis, given a large quantity of H2, ammonia can be decomposed at the same temperature to completely come into contact with the sulfur species in petroleum coke to generate H2S.

desulfurization, petroleum coke, ammonia, calcination, TG-MS

1 Introduction

As a cheap carbon source and high-quality raw material, petroleum coke has the advantages of low ash and high fxed carbon content[1-2]. In recent ten years, with the worsening of crude oil quality and the increased depth of crude oil processing at refnery, a quality problem related with the impurities in petroleum coke including the increased sulfur content has been arousing concern all over the world[3].

Most of the sulfur species in petroleum coke exist in the form of stable thiophenes[4-5], which can hardly be removed from the coke itself. Therefore, the desulfurization of petroleum coke is always a gut issue in the technology industry, and it tends to worsen. Various methods of desulfurization have been attempted in the past 50 years. Specifcally, solvent extraction has been proven ineffective. Phillips and Chao[6]found that only 20% of sulfur compounds can be removed byo-chlorophenol at 130 ℃. A good desulfurization rate is obtained when coke is mixed with NaOH and KOH and then calcined at 800—1 000 ℃[7-8]. However, the ash content of petroleum coke increases as a consequence of this reaction. Desulfurization through a gas–solid reaction at a certain temperature has been proven advisable as it yields effective results and ensures that the ash content does not increase. For example, Brandt and Kapner[9]found that the desulfurization rate could reach 90% by mixing coke and methane at 1 000 ℃ or higher. George, et al.[10]discovered that through coke fluidization, 90% or more sulfur compounds in coke could be removed through hydrodesulfurization (HDS) at more than 600 ℃. However, HDS presents a certain risk and causes coke loss. The most direct and effective means of petroleum coke desulfurization is calcination at a temperature of 1 400 ℃ or more[11]. A high temperature leads to a high desulfurization rate. However, high-temperature calcination results in low coke yield and diseconomy. Given thelow economic beneft of high-temperature calcination, the use of high-sulfur petroleum coke as a main burning fuel or as the starting material for gasification is of habitual occurrence in daily life[12].

Although the quality of petroleum coke is deteriorating, its role in the carbon industry is irreplaceable. In nearly 10 years till now, researchers have been unable to find a material that can replace petroleum coke as the starting material for the production of carbon anodes[13]. Therefore, the economic use of high-sulfur petroleum coke in production presents a problem that currently needs to be solved.

The ammonia atmosphere-induced desulfurization of coke, which is an effective desulfurization method aside from desulfurization of coke in hydrogen atmosphere, is explored in this study on the basis of a series of experimental conditions for calcining coke at less than 1 000 ℃. Moreover, through analytical techniques such as FT-IR, BET, and TG-MS, the mechanism for desulfurization of petroleum coke by ammonia gas is investigated.

2 Experimental

2.1 Raw materials

The results of proximate and ultimate analysis, and analysis of trace elements are presented in Tables 1 and 2. Petroleum coke is classifed into four types according to its sulfur content ranging from high to low. Among these types, Tianjin (TJ) coke has the highest sulfur content, followed by Qingdao (QD) coke and Zhenhai (ZH) coke. Dongming (DM) coke shows the lowest sulfur content. With the exception of DM coke, all types of coke belong to the typical high-sulfur petroleum coke (with S content exceeding 4%). Judging from the analysis of the trace metal elements, the contents of all metal elements in petroleum coke are extremely low. Moreover, because of the following three reasons, the effects of metal elements on the desulfurization of cokes are not considered in this study. (1) Compared with the carbon and sulfur contents, the metal element contents are significantly lower. If some catalytic or inhibitory substance exists between metal elements and carbon or sulfur, then the effect on the desulfurization results will be limited. (2) The catalytic or inhibitory activity of metal elements in petroleum coke is mostly demonstrated in the presence of oxygen, such as the roles of metal elements identifed during study on the air/CO2reactivity of petroleum coke[14]. This experiment can be considered as a petroleum coke desulfurization reaction in an anaerobic environment, where no research in the literature has proved that trace metal elements in petroleum coke can impose signifcant catalytic or inhibitory effect on petroleum coke reactivity. (3) The effect of trace metal elements in petroleum coke on ammonia gas in the process of heating can also be ignored. In this study, only iron among other metal elements was confrmed to be capable of promoting the ammonia decomposition effciency[15].

Table 1 Proximate and ultimate analyses of four dried coke samples

Table 2 Analysis of trace metal elements in four dried coke samples

2.2 Calcination equipment and methods

The calcination equipment used in this study consists of a muffe furnace and a tube furnace. The structure of the muffe furnace and tube furnace is shown in Figure 1. The muffe furnace is made by the Yuandong Electric Furnace Company in Changsha, and the temperature difference of the heat preservation zone is less than ±2 ℃. The tube furnace is a Hefei Kejing OTF-1200X type tube furnace, and the temperature difference of the heat preservation zone is less than ±1 ℃. During the experiments, the muffe furnace can be regarded as the equipment operating under normal atmospheric pressure. When the gas stream passes through the furnace at a certain temperature, the internal andexternal pressure of the furnace is nearly the same. However, given the good sealing (through which the thermoblock and rubber ring interfaces can exhibit good sealing ability) during the expansion of gases in heating small, uniform gas channels, the pressure of the in-tube furnace is higher than the atmospheric pressure (usually > 0.12 MPa). In addition, the fne particle samples in the furnace are fuidized.

Figure 1 Calcination equipment: (a) muf fl e furnace and (b) tube furnace

Prior to calcination experiments in the presence of gas stream, a sample composed of 5 g of dried petroleum coke powder is evenly spread on a standard 4 cm×4 cm square fat porcelain boat. The thickness of the specimen in the porcelain boat is approximately 1.5 mm. The porcelain boat is then sent to the heat preservation zone of the furnace for calcination. Unless otherwise specifed in these experiments, the particle size of petroleum coke is less than 0.1 mm, and the fowrate of ammonia gas (with a NH3purity of ≥ 99.99%) is 10 L/h, while the calcination equipment consisting of the muffe furnace operates at a calcination temperature of 800 ℃, a furnace heating rate of 10 ℃/min, and a holding time of 120 min. In addition, a protection gas for the low-fow (5/h) calcination atmosphere is introduced into the system during the heating process to prevent the sample from being oxidized. When the temperature increases to the required calcination temperature, the protection fow is adjusted to the preset gas fow. At last, the gas fow is interrupted when calcination is completed, that is, until the in-furnace temperature is lower than 400 ℃.

Sulfur is evenly distributed in coke particles[16]. Therefore, during analysis, the desulfurization rate of the coke sample is obtained by comparing the sulfur content in the samples before and after desulfurization. The desulfurization rate can be expressed by the following equation:

whereCbandCaare the sulfur contents in the samples before and after desulfurization, respectively.

2.3 Analytical devices

The analytical devices used in this study are as follows: (1) A Vario MICRO cube analyzer (Elementar Analysensysteme GmbH) with an analytical precision of less than 0.1% (abs) for C, H, N, and S elements and less than 0.2% (abs) for O element; (2) An inductively coupled plasma emission spectrometer (Thermo Electron Corporation) with an analytical precision of less than 0.001‰ (abs); (3) A Nicolet 6700 FT-IR spectrometer (Thermo Electron, USA) with an analytical precision of approximately 0.09 cm-1; (4) An automatic specifc surface and pore analyzer (Micromeritics Instrument Corp., USA) with a minimum range of analysis of 0.000 1 m2/g and no upper limit; and (5) an Evolution 16/18 type thermogravimetric analyzer (French SETARAM Instrumentation) coupled with an OMNI star type mass spectrometer (German Pierre Vacuum Instrument Co.).

3 Results and Discussion

3.1 Calcination desulfurization in different atmospheres

After ZH petroleum coke had been calcined, an inert gas (N2) and three kinds of reducing gases were introduced. The results are presented in Table 3. The corrected coke yield is the sum of the coke yield plus the sulfur content that has been removed in the desulfurization process. Given that the amount of sulfur removed by different desulfurization processes varies and that sulfur is present in certain proportions in petroleum coke (i.e., from 1% to 10%), the determination of coke yield performed by weighing only the petroleum coke samples after desulfurization is one sided. Therefore, the corrected coke yield can directly refect the loss of non-sulfur elements in various desulfurization processes. Table 3 shows that in N2atmosphere, the sulfur in coke cannot be removed at 800 ℃. However, sulfur can be partially removed in reducing gases (except N2). Among these gases, H2has the best desulfurization efficiency, followed by NH3, while the desulfurization effciency of CO is the weakest. However, the coke yields of H2- and NH3-calcined coke are lower than those of N2-calcined coke. When petroleumcoke is calcined, a series of additional reactions between petroleum coke and H atoms are speculated. Therefore, abundant aromatic carbon particles are pyrolyzed and gasified with the formation of low-molecular saturated and unsaturated hydrocarbons along with the remaining petroleum coke particles. By comprehensively considering all above-mentioned circumstances, we fnd that NH3is a type of gas that not only ensures the desulfurization rate of petroleum coke but also guarantees the good recovery effciency. This fnding is worthy of further study and discussion.

Table 3 Comparison of different gas desulfurization method

3.2 Experiment for desulfurization of petroleum coke in NH3atmosphere

The desulfurization rates of the four kinds of petroleum coke at different ammonia gas flow rates are illustrated in Figure 2(a). The desulfurization rate of TJ, QD, and ZH cokes reached their highest value at an ammonia fow rate of 10 L/h. However, the desulfurization rate of DM coke, which showed a lowest sulfur content among the four kinds of coke samples, exhibited a highest value at the gas flow, which ranged from 10 L/h to 15 L/h with minimal difference. We speculate that with a minimal ammonia fow, the desulfurization reaction cannot occur completely. However, if the ammonia fow is excessively large, then the ammonia gas will not be suffciently preheated, rendering the temperature of the coke?gas contact surface lower than the temperature point required by the normal coke?gas reaction. Therefore, in this study, a fow rate of 10 L/h is considered as the best ammonia gas fow. A duration of 0, 60, 120, 180, 240, and 480 min, respectively, is considered as the reaction time for ammoniainduced desulfurization calcination. The results are presented in Figure 2(b), which indicates that a longer heat preservation means a higher desulfurization rate. However, the desulfurization rate is stable in the range from 120 min to 240 min. A further increase in reaction time results inmore desulfurizing gas (ammonia) consumption and energy waste. Therefore, a duration of 120 min is selected as the heating time required in the ammonia desulfurization experiment.

Figure 2 Desulfurization rates of petroleum coke under different treatment conditions

The desulfurization rate of the four kinds of petroleum cokes at different calcination temperatures is illustrated in Figure 2(c). All types of cokes attained their highest desulfurization rate at approximately 800 ℃. The desulfurization reaction began at above 500 ℃. However, a temperature which was higher than 800 ℃ would result in a poorer desulfurization effect.

Under the same calcination condition, the desulfurization rate of four kinds of petroleum coke with different particle sizes is shown in Figure 2(d). When the particle size of coke sample was greater than 1 mm, the increase in desulfurization rate was extremely limited despite the decrease in particle size. When the particle size of coke sample was less than 1 mm, the desulfurization rate increased rapidly with a decreasing coke particle size. When the particle size was less than 0.1 mm, approximately 70% of the sulfur species in high-sulfur petroleum coke could be removed.

3.3 FT-IR analysis of the four kinds of petroleum cokes

The FT-IR spectra of the four kinds of petroleum coke samples before and after desulfurization are shown in Figure 3. The spectra of all four kinds of green cokes showed absorption peaks at approximately 2 900 cm-1, 1 080 cm-1, and 755 cm-1. This result means that the thiophene-like groups existed in petroleum coke. Absorption peaks were also found at approximately 1 640 cm-1, 1 380 cm-1, and 800 cm-1. This result means that the benzene-like aromatic groups existed in petroleum coke. Further analysis of the spectra of coke samples showed that most of the sulfur species existed in the thiophene-like groups and virtually nowhere else. Therefore, organic sulfur species mostly existed in the thiophene-like groups of the coke. As illustrated in Figure 3, the spectra of the four types of desulfurized coke samples presented several identical phenomena. For example, the benzene-like aromatic groups exhibited minimal change, and the thiophene-like groups still existed in the coke. However, the characteristic absorption peaks of thiophene-like groups decreased signifcantly at around 755 cm-1. This result indicates that slight changes occurred in the thiophene-like groups because of the ammonia-induced desulfurization of coke sample.

Figure 3 FT-IR spectra of (a) TJ coke, (b) QD coke, (c) ZH coke, and (d) DM coke

The characteristic absorption peaks of thiophene-like groups still existed at approximately 650 cm-1, indicating that some sulfur species still existed in the coke. Further analysis of the spectra indicates to the lack of absorption peaks of the new organic groups[17]. This finding means that ammonia-induced desulfurization of coke does not add impurities to the coke. Overall, the ammonia-induced desulfurization of coke only slightly affects the quality of petroleum coke.

3.4 Desulfurization equipment (fluidization experiment)

As discussed in Section 2.3, the coke sample in the tube furnace is fluidized in order to enhance the desulfurization reaction. We use another equipment (tube furnace) aside from the muffe furnace to study the relationship between the ammonia-induced desulfurization of coke and the desulfurization equipment. The results of desulfurization rate using the muffe or tube furnace are shown in Table 4. Upon using the tube equipment, more than 90% of the sulfur in the four kinds of green cokes is removed. This result can prove that under optimum reaction conditions (e.g., proper calcination temperature, gas flow, reaction time, etc.), the final factor that limits the improvement of the desulfurization rate depends on the degree of petroleum coke particles capable of coming in contact with the ammonia atmosphere. Fluidized desulfurization is one of the most effective approaches that can increase the contact degree between coke and ammonia atmosphere. Therefore, it is also one of the most effective approaches to coke desulfurization.

Table 4 Desulfurization rates using muffle or tube furnace

3.5 Analysis of the mechanism of ammonia-induced desulfurization

The mechanism of ammonia in removing sulfur species from petroleum coke at 800 ℃ is considered to be based on the following points of view:

(1) Decomposition reaction of ammonia. The decomposition reaction (Reaction 1) occurs when ammonia is heated to a certain temperature. In the literature[18-19], the temperature at which ammonia begins to decompose is at approximately 100 ℃ in an open environment, and the decomposition rate reaches 90% at 300 ℃ through calculation. Through Reaction 1, strongly reducible H2is formed, and it has been proved to be capable of desulfurizing the coke. In order to verify if the decomposition reaction of ammonia occurs in the process of heating and explore the mechanism of the desulfurization reaction between ammonia and petroleum coke, QD coke is designed to undergo heating experiments by TG-MS analysis from 25 ℃to 1 000 ℃ in the gas atmosphere consisting of 1% NH3+ 99% Ar and 100% Ar, respectively. Using 1% NH3+ 99% Ar instead of pure NH3is conducted to prevent the corrosion of the analytical equipment. TG is used for heating the sample at a temperature increase rate 10 ℃/min. MS is used to detect the content of H2, H2S, CS2, COS, S(g), and SO2in the off-gas discharged from the desulfurization reactions between C and S, NH3and S, or some other elements in the coke sample. Among them, petroleum coke should be considered frst because of the strong reducing property of carbon. Finally, the results show that CS2, COS, and S(g)are not detected, thus verifying that the coke itself does not play a role during the desulfurization reactions of coke in the presence of ammonia gas.

The results for analysis of H2, H2S, and SO2in the offgas are presented in Figure 4. Figure 4 (a, c, and e) shows that a large amount of H2gas and a small amount of H2S in the off-gas are formed during heating of QD petroleum coke in ammonia atmosphere, but SO2is not found in the effuent gas. This fnding proves the cause showing that desulfurization of coke in ammonia gas can convert sulfur species into H2S gas. Figure 4 (b, d, and f) also indicates that a small amount of H2gas and a trace amount of SO2in the off-gases are formed during heating QD petroleum coke in pure Ar atmosphere (viz.: the petroleum coke pyrolysis process), but H2S is not found in the off-gas. The mechanism for formation of SO2detected during the petroleum coke pyrolysis process might be explained by the decomposition of sulfate species in the coke during heating or the reaction of H2S gas with trace O2, which mightpossibly remain in the mass spectrometer.

The comparison between Figure 4(a) and Figure 4(b) shows that during the petroleum coke pyrolysis process, H2is released at around 500 ℃, and the reaction reaches its maximum rate at 760—800 ℃, and then declines gradually. However, when the coke is heated in the presence of NH3gas, H2is detected synchronously and its concentration increases sharply at the beginning of heating. Thereafter, it reaches a stable value gradually after 300 ℃, with its growth rate leveling off. Although the TG-MS test results cannot serve as the quantitative analysis data used to calculate the ammonia decomposition rate, the Y (abundance) data of the test results can be used as a reference on the relative content. By contrast, the abundance of H2(Figure 4(a)) is different from the small one (Figure 4(b)), and this difference is signifcant. This amount of H2cannot be released by coke itself. Therefore, during the process of coke heating in the presence of ammonia, a large amount of H2is released. Moreover, it is inferred that H2could be formed during the ammonia decomposition reaction.

Figure 4 TG-MS analysis of gases from calcination of QD petroleum coke in the atmosphere of 1% NH3+ 99% Ar and 100% Ar: (a) H2; (c) H2S and (e) SO2contained in off-gases from calcination of coke in 1% NH3+ 99% Ar atmosphere, and (b) H2; (d) H2S and (f) SO2contained in off-gases from calcination of coke in 100% Ar atmosphere at temperatures ranging from 25 ℃ to 1 000 ℃

The curve in Figure 4(a) is similar to the ammonia decomposition rate data provided by the literature. Therefore, during the desulfurization process, the ammonia decomposition rate may reach more than 90% when the temperature is higher than 300 ℃. Figure 4(c) shows that during the coke heating in ammonia atmosphere, H2S is released at around 500 ℃, and its concentration peaks at around 800 ℃, declines gradually, and then begins to fatten after 900 ℃. This curve is similar to the desulfurization rate curve versus temperature presented in Figure 2(c) of Section 3.2, and it also proves that most of the sulfur in coke is removed after being transformed into H2S. Obviously, the only source of hydrogen must be originated from the H2gas formed during ammonia decomposition.

The literature[16]indicates that between the C–S and C–C bond structures, the H atom prefers to react frst with the former, thereby generating H2S (Reaction 2). The sulfur which constitutes the organic compounds in coke is indicated as organic-S.

As for the inorganic sulfur (Fe2S3, sulfate, and sulfte with their concentration being less than 0.1 wt.% in coke), Fe2S3, and CaSO4, and the remaining sulfate and sulfite (with the metal ions being Fe2+, Fe3+, Ni2+, or Ti4+) can be removed by transforming into SO2through reactions (3) and (4), which are feasible based on the thermodynamic calculation. However, in a hydrogen-rich environment at high temperature, SO2can be transformed to H2S through reaction (4).

(2) Characteristics of thiophene-like structures in coke. The other organic-S (sulfone, mercaptan, and thioether, with a concentration of usually less than 10% of the total sulfur content) that exists in coke can be easily removed. However, the thiophene that exists in coke not only accounts for most of the total sulfur content but is also a recalcitrant compound, which can hardly be removed. The essence of thiophene desulfurization is to break the two C-S bonds in thiophene. However, because of the π bond formed by the carbon and sulfur elements in thiophene, the binding energy of thiophene is high and can hardly be broken down. In addition, thiophene does not exist alone, and it combines carbon chains together, thus making itself harder. According to Guo Wenping[20], a C—S bond in thiophene can be easily broken in a hydrogenrich environment at low temperature because of the low activation energy of the bond cleavage. However, another C—S bond in thiophene has significantly higher activation energy. Only a high temperature can break the C—S bond coupled with the low activation energy of the bond cleavage. This fnding can explain why the effective desulphurization of petroleum coke, in which the thiophene is the main existing form of sulfur, needs to be performed at a certain high temperature.

(3) Porous structure of petroleum coke. The high temperature reduction and desulfurization of petroleum coke and the thermodynamics had been investigated by Xiao, et al.[21]According to thermodynamic calculation and the law of the higher calcination temperature, a higher desulfurization rate was found. Both thermodynamic results and ammonia decomposition rate have pointed out that the desulfurization rate should be increased with a rising temperature. However, it can be seen in Figure 2(c) that the desulfurization rates achieved at above 800 ℃ could not match up to the expected value at that temperature. Thus, there must be some reasons that hindered the desulfurization reaction dominating at above 800 ℃.

Previous experiments (Section 3.2, Figure 2(d)) have confrmed that more than 90% of the sulfur content in petroleum coke is removed when petroleum coke is ground up to a certain size (＜ 0.1 mm). This fnding proves that a specifc surface area is also a constraint for increasing desulfurization rate. However, petroleum coke is a typical porous carbon material[22]. Abundant holes can induce contact between the outer atmosphere and the internal structure of coke and the complete desulfurization reaction, which can be refected by the specifc surface area. Thus, in analyzing specific surface area changes of petroleum coke during ammonia-induced desulfurization calcination, nine QD coke samples were calcined at 700 ℃, 725 ℃, 750 ℃, 775 ℃, 800 ℃, 825 ℃, 850 ℃,875 ℃, and 900 ℃, respectively. Thereafter, data on the desulfurization rate and specific surface area of each sample with similar particle size (0.1 mm) were analyzed. The results are shown in Figure 5, indicating that after calcination for 2 h, the specifc surface area at 700 ℃ is still low and is equal to only 15.09 m2/g. However, petroleum coke obtains a largest specifc surface of around 273.745 m2/g at 800 ℃. Therefore, with the calcination temperature further rising sequentially, the specifc surface area of coke declines slowly. The specifc surface area curve and the desulfurization rate curve have similar trend of fuctuations at the same temperature. Therefore, some relationships exist between the desulfurization rate and the specifc surface area of coke. However, the desulfurization reaction can occur when the temperature is higher than 500 ℃, but only when the temperature reaches 800 ℃ can the cokes achieve their maximum specifc surface area. During calcination of coke at temperatures below 800 ℃, not only the temperature itself but also the specifc surface area of coke can limit the desulfurization reaction. Similarly, the higher the calcined temperature (> 800 ℃) is, the lower the activation energy of the bond cleavage would be. Therefore, the sulfur species contained in coke are easier to be removed, and it would result in a lower specifc surface area of coke. The contraction of petroleum coke particles and the decline in specifc surface area could restrict the contact between the sulfur in coke and hydrogen in the gas stream. This restriction reduces the petroleum coke desulfurization rate when the calcination temperature is higher than 800 ℃. This phenomenon can explain why the petroleum coke can achieve a highest desulfurization rate at 800 ℃ in the ammonia-rich atmosphere.

Figure 5 Desulfurization rates and speci fi c surface areas of QD cokes at different ammonia-induced calcination temperatures

Therefore, the ammonia-induced desulfurization can be considered as the reaction of H2with the sulfur species contained in the coke. During desulfurization, three main gases, viz. NH3, H2, and N2, are present in the furnace, and the mass ratio of gaseous H2and N2is approximately 3:1. As the reaction proceeds, the effective gas (H2) is diluted by other gases in the furnace. Given the limited quantity of gas, H2prefers to react first with C-S bonds than with C-C bonds. Therefore, compared with H2, NH3not only can maintain a high desulfurization rate but also can reduce coke loss during the desulfurization process. In addition, during transportation and stockpiling, liquid ammonia is safer and more economical than H2. Overall, aside from H2, NH3is another gas that can be used for petroleum coke desulfurization.

4 Conclusions

This study explores the feasibility of ammonia gas to be used for petroleum coke calcining desulfurization. The results are as follows:

(1) Ammonia is a gas that not only can achieve a high desulfurization rate but also can reduce coke loss during the desulfurization process of petroleum coke as compared with H2and CO.

(2) After performing a series of experiments by using petroleum coke with a particle size of less than 0.1 mm, calcination at 800 ℃ in ammonia atmosphere with a fow rate of 10 L/h and under heating for 120 min or more, we have determined that most of the sulfur (60%–80%) in petroleum coke has been removed. In addition, through other means of improving the gas–solid contact, such as fuidized calcination, the desulfurization rate of petroleum coke can be further improved to exceed 90%.

(3) Through TG-MS analysis, the curve data of H2and H2S generated during ammonia desulfurization show an increasing trend from 25 ℃ to 1 000 ℃ clearly and accurately. This fnding proves that one of the key components of desulfurization is H2, which is formed from the decomposition of ammonia and can react with sulfur species in coke to generate H2S.

(4) Although the activation energy of the C-S bonds declines with the increase in temperature, the specific surface area of petroleum coke decreases (the particlesshrink) when the temperature is higher than 800 ℃. This phenomenon limits the contact area between sulfur and gas. Therefore, the desulfurization rate of petroleum coke reaches its maximum value at 800 ℃.

Acknowledgements: This work was financially supported by the National Natural Science Foundation of China (Projects No. 51374253 and No. 51574289).

[1] Leion H, Mattisson T, Lyngfelt A. The use of petroleum coke as fuel in chemical-looping combustion[J].Fuel, 2007, 86(12): 1947-1958

[2] Xing P, Wang J, Lyu T, et al. Ultrasound-assisted impurity removal from petroleum coke[J].Separationt and Purifcation Technology, 2015, 151: 251-255

[3] Xiao J, Zhong Q, Deng S, et al. Mechanism and effect of sulfur on coke air reactivity and mutual restraint between calcium and sulfur[J].Journal of Central South University (Science and Technology), 2015, 46(10): 3603-3610

[4] Xiao J, Zhong Q, Li F, et al. Mutual inhibition between catalytic impurities of sulfur and those of calcium in coke during carbon-air and carbon-CO2reactions[J].Energy and Fuels, 2015, 29(3): 1961-1971

[5] Xiao J, Deng S Y, Zhong Q F, et al. Effect of sulfur impurity on coke reactivity and its mechanism[J].Transactions of Nonferrous Metals Society of China, 2014, 24(11): 3702-3709

[6] Phillips C R, Chao K S. Desulphurization of Athabasca petroleum coke by (a) chemical oxidation and (b) solvent extraction[J].Fuel, 1977, 56(1): 70-72

[7] Kawano T, Kubota M, Onyango M S, et al. Preparation of activated carbon from petroleum coke by KOH chemical activation for adsorption heat pump[J].Applied thermal engineering, 2008, 28(8): 865-871

[8] Zhan X, Zhou Z, Wang F. Catalytic effect of black liquor on the gasifcation reactivity of petroleum coke[J].Applied Energy, 2010, 87(5): 1710-1715

[9] Brandt H, Kapner R. Desulfurization of petroleum coke[J].Preprints-American Chemical Society. Division of Petroleum Chemistry, 1984, 29(2): 453-455

[10] George Z M, Schneider L G, Tollefson E L. Desulphurization of a fuid coke similar to the Athabasca oil sands coke[J].Fuel, 1978, 57(8): 497-501

[11] Martins M A, Oliveira L S, Franca A S. Modeling and simulation of petroleum coke calcination in rotary kilns[J].Fuel, 2001, 80(11): 1611-1622

[12] Zhang J, Zhou J, Yan H. Kinetic model on coke oven gas with steam reforming[J].Journal of Central South University of Technology, 2008, 15(1): 127-31

[13] Megahed S, Ebner W. Lithium-ion battery for electronic applications[J].Journal of Power Sources, 1995, 54(1): 155-162

[14] Engvoll M A. Reactivity of anode raw materials and anodes for production of aluminium[D].Trondheim: Institute for Kjemi, 2002: 87-97

[15] Lu A H, Nitz J J, Comotti M, et al. Spatially and size selective synthesis of Fe-based nanoparticles on ordered mesoporous supports as highly active and stable catalysts for ammonia decomposition[J].Journal of the American Chemical Society, 2010, 132(40): 14152-14162

[16] Xiao J, Zhong Q, Li F, et al. Modeling the change of green coke to calcined coke using Qingdao high-sulfur petroleum coke[J].Energy and Fuels, 2015, 29(5): 3345-3352

[17] Zubot W, Mackinnon M D, Chelme-Ayala P, et al. Petroleum coke adsorption as a water management option for oil sands process-affected water[J].Science of the Total Environment, 2012, 427: 364-372

[18] Yin S, Xu B, Zhou X, et al. A mini-review on ammonia decomposition catalysts for on-site generation of hydrogen for fuel cell applications[J].Applied Catalysis A: General, 2004, 277(1):1-9

[19] Hansgen D A, Vlachos D G, Chen J G. Using first principles to predict bimetallic catalysts for the ammonia decomposition reaction[J].Nature Chemistry, 2010, 2(6): 484-489

[20] Guo W. Density functional theory study on the thiophene cracking mechanism over Bronsted acidic zeolites[D].Changsha: Central South University, 2006: 13-48

[21] Xiao J, Zhang Y, Zhong Q, et al. Reduction and desulfurization of petroleum coke in ammonia and their thermodynamics[J].Energy and Fuels, 2016, 30(4): 3385-3391

[22] Heintz E A. The characterization of petroleum coke[J].Carbon, 1996, 34(6): 699-709

Received date: 2016-06-24; Accepted date: 2016-09-27.

Professor Xiao Jin, Telphone: +86-13-607445108; E-mail: changshaxiaojin@126.com.