Eo

Ja

b

a

ARR2AA

KPCBGPR

1

uio1tCHtccra

h0

Journal of Analytical and Applied Pyrolysis 117 (2016) 64–71

Contents lists available at ScienceDirect

Journal of Analytical and Applied Pyrolysis

journa l h om epage: ww w.elsev ier .com/ locate / jaap

ffect of high-temperature pyrolysis on the structure and propertiesf coal and petroleum coke

in Xiaoa,b, Fachuang Lia,∗, Qifan Zhonga, Jindi Huanga, Bingjie Wanga, Yanbing Zhanga

School of Metallurgy and Environment, Central South University, Changsha, Hunan Province 410083, PR ChinaNational Engineering Laboratory of Efficient Utilization of Refractory Nonferrous Metal Resources, Changsha 410083, PR China

r t i c l e i n f o

rticle history:eceived 1 July 2015eceived in revised form2 November 2015ccepted 15 December 2015vailable online 22 December 2015

eywords:yrolysisrystallite structureET surface areaasification reactivityowder resistivityeal density

a b s t r a c t

In this study, for the purpose of substituting part of petroleum coke with low ash anthracite to preparecarbon anode used in aluminum electrolysis, the effects of high-temperature pyrolysis on the structureand properties of coal char and petroleum coke were studied at the temperature range of 1000 ◦C–1600 ◦C.Results showed that the carbon crystallite structure of coal char and petroleum coke became more orderedwith the increase of the temperature. However, the graphitization degree of coal char was lower thanthat of petroleum coke at the same pyrolysis temperature. The Brunauer-Emmett-Teller (BET) surfacearea of coal char decreased with the increase of temperature, whereas the BET surface area of petroleumcoke decreased first and then increased. The increase of pyrolysis temperature generally inhibited thegasification reactivity of coal char, whereas the gasification reactivity of petroleum coke was inhibitedfirst and then promoted. Moreover, the increase of pyrolysis temperature led to the rapid decrease ofpowder resistivity of coal char and petroleum coke. In particular, the powder resistivity of coal char wassignificantly higher than that of petroleum coke at the same pyrolysis temperature. Additionally, the

real density of coal char and petroleum coke showed a different trend with the increase of temperature.Specifically, the real density of the former decreased, whereas that of the latter increased. The causes ofthe above observations were also discussed. Overall, this study provides further understanding about thedifferences between coal char and petroleum coke, which will facilitate the preparation of carbon anodefrom the mixture of coal char and petroleum coke.

© 2015 Elsevier B.V. All rights reserved.

. Introduction

Petroleum coke is the main raw material of carbon anodesed in aluminum electrolysis. Hence, petroleum coke significantly

nfluences the energy consumption and environmental benefitsf aluminum electrolysis. The cost of carbon anode accounts for5–20% of the total cost of aluminum electrolysis production. Withhe rapid development of the electrolytic aluminum industry inhina, the demand for high-quality petroleum coke is also growing.owever, the quality of petroleum coke is deteriorating because of

he increasing proportion of imported heavy crude oil used in China,onsequently leading to a tight supply of high-quality petroleumoke used for aluminum electrolysis [1,2]. Thus, expanding backup

eserves of resources for carbon anode is beneficial to the sustain-ble development of the aluminum industry. Furthermore, finding

∗ Corresponding author. Fax: +86 15273113704.E-mail address: cwd818@163.com (F. Li).

ttp://dx.doi.org/10.1016/j.jaap.2015.12.015165-2370/© 2015 Elsevier B.V. All rights reserved.

low-cost raw materials to replace petroleum coke is necessary toreduce the cost of carbon anode.

As an important raw material for carbon products, high-metamorphic anthracite exhibits the characteristics of highcarbon content, compact structure, high strength, good thermalstability, and lower cost than petroleum coke [3–6]. Hence, high-metamorphic anthracite can be an ideal raw material to replacepetroleum coke for the preparation of carbon anode. To date,attempts have been made to produce carbon anode from coalextracts [7–10], which were obtained by solvent extraction tech-nology. However, this high-cost technology is complex and theyield is low. Additionally, good physical and chemical propertiesof prebaked anodes have been achieved using low-ash calcinedanthracite to partly substitute petroleum coke [1,11,12]. However,coal and petroleum coke presented obvious differences becauseof the different formation processes. Even after high-temperature

treatment, the properties of coal char and petroleum coke (such ascarbon crystalline structure, specific surface area, gasification reac-tivity, electrical resistivity and real density) remain different, andsuch difference will affect the properties and energy efficiency of

and Applied Pyrolysis 117 (2016) 64–71 65

chdca

pd(poccct

2

2

pwgpofl

hfcrcoTecteriarfi

2

hisiTapsipAPa

J. Xiao et al. / Journal of Analytical

arbon anode in aluminum electrolysis. Although, studies [1,11,12]ave demonstrated that high-quality carbon anodes can be pro-uced by adding an appropriate amount of coal char to petroleumoke, the effects of pyrolysis on the structure and properties of coalnd petroleum coke have not been thoroughly investigated.

High-temperature pyrolysis of coal and petroleum coke is arimary process for their carbon products. Moreover, pyrolysis con-itions significantly influence the performance of calcined cokechar), thereby affecting the performance indicators of the carbonroducts. The current study systemically investigated the effectsf high-temperature treatment on the structure and properties ofoal and petroleum coke. Accordingly, the differences in pyrolysisharacteristics and performances between two kinds of calcinedokes are determined. Overall, this study provides theoretical basiso prepare carbon anodes by substituting coal for petroleum coke.

. Experimental

.1. Raw materials

Low-ash anthracite (Shenhua Group Corp., China) andetroleum coke (Qilu Petrochemical Co., Ltd., China), whichere named A and PC, respectively, were used for the investi-

ation. Proximate analysis, ash compositions, and ash meltingroperties are listed in Tables 1 and 2. The ash compositionsf anthracite and petroleum coke were determined by X-rayuorescence (PANalytical Axios mAX equipment).

As shown in Table 1, the fixed carbon content of coal is slightlyigher than that of petroleum coke, whereas the volatile and sul-

ur contents of coal are significantly lower than those of petroleumoke. These features are the major advantages of coal as raw mate-ial to prepare carbon anode for aluminum electrolysis. However,ompared with petroleum coke, the disadvantage of coal is alsobvious, that is, a higher ash content. The impurity elements inable 2, such as Si and Fe, will significantly affect the quality oflectrolytic aluminum, because the impurities of Si and Fe in thearbon anode will turn into the aluminum liquid. Although the con-ent of aluminum is also high, aluminum is considered a valuablelement. The allowed impurity content in raw carbonaceous mate-ials for carbon anode is generally based on the standard of 2B graden SH/T0527-1992 [13]; this standard states that the contents of Sind Fe should be <0.08% each. The contents of Si and Fe in coal areelatively high, whereas those in petroleum coke are low. There-ore, when adding an appropriate amount of coal, the content ofmpurities in the carbon anode can also meet the requirements.

.2. Preparation of char samples

The pyrolysis of coal and petroleum coke was completed in aigh-temperature furnace. The inert atmosphere was maintained

n the furnace to prevent oxidation of coal and petroleum coke. Theamples were heated to a desired pyrolysis temperature at a heat-ng rate of 4 ◦C/min, and then held at this temperature for 120 min.he pyrolysis temperatures were 1000 ◦C, 1150 ◦C, 1300 ◦C, 1450 ◦C,nd 1600 ◦C. When the heat preservation time was over, the sam-les kept in the furnace were cooled to room temperature. Theamples were then removed from the furnace, broken, and gradedn different particle sizes. Coal chars prepared under different tem-

eratures were referred to as A1000, A1150, A1300, A1450, and1600, whereas the petroleum cokes were referred to as PC1000,C1150, PC1300, and PC1450. The proximate analysis and ultimatenalysis of coal chars and petroleum cokes are listed in Table 3.

Fig.1. The schematic diagram of the testing device for gasification reactivity.

2.3. Measurements of air gasification reactivity and CO2gasification reactivity

Air gasification reactivity and CO2 gasification reactivity areimportant indicators of coke chemical activity, which significantlyaffect the overconsumption of carbon materials in aluminum elec-trolysis [14,15]. In accordance with the principle of weight lossmethod, the air gasification reactivity and CO2 gasification reac-tivity of cokes were characterized in terms of residual rate afterreaction with air and CO2 following the test standard YS/T 587.7-2006 [16] in China.

The tests for air gasification reactivity and CO2 gasification reac-tivity were conducted in a quartz tube reactor testing equipment;a schematic of this testing device are shown in Fig. 1. The detailedmeasurement procedure is as follows: 5 g of coke with particlesize of 1 mm to 1.4 mm was placed in the reactor; air gasificationreactivity was performed in air atmosphere at 600 ◦C for 60 min,whereas CO2 gasification reactivity was measured in CO2 atmo-sphere at 1000 ◦C for 100 min; and the flow rate of both gases was50 L/h. At the end of the reaction, the cooled char was collected fromthe testing equipment and then weighed. The reactivity was indi-cated in terms of the ratio of the mass loss to the initial weight ofthe sample. The gasification percentage (X) is calculated as follows:

X =(

1 − mt

m0

)(1)

where m0 represents the initial weight of the sample, and mt is theresidual weight of the sample.

2.4. Analysis of carbon crystalline structure and BET surface areaof the samples

The crystallite structure of coke was characterized by X-raydiffraction (XRD) with Cu K� radiation in a Rigaku D/Max 2500diffractometer at a scan speed of 4◦ min−1.

XRD is an effective means to analyze carbon crystallite struc-ture. The interplanar spacing d002 and the stacking height of thecarbon crystal Lc are the basic parameters for analyzing the carboncrystallite structure. According to Eqs. (2) and (3), the parametersd002 and Lc were calculated from the (0 0 2) peak as follows:

d002 = �

2sin(

�002) (2)

Lc = 0.89�

�002cos(

�002) (3)

66 J. Xiao et al. / Journal of Analytical and Applied Pyrolysis 117 (2016) 64–71

Table 1Proximate analysis and ash character temperature of raw materials.

Sample Proximate analysis(wt%) Ultimate analysis(wt%) Ash character temperature/◦

FC V A St C H N St DT ST FT

A 89.99 7.12 2.89 0.14 88.39 2.40 0.73 0.19 1400 1460 >1500PC 88.8 11.0 0.40 3.7 92.15 1.29 1.15 3.43 1280 1300 1330

DF: initial temperature of melting; ST: softening temperature; FT: flow temperature.

Table 2The content of the impurity elements (ppm).

Sample A Al Si Fe Ti Ca Mg Na K P V Ni

A 4960 3440 1569 160 2933 457 727 80 95 – –PC – 66 68 12 160 – 67 – – 309 200

Table 3Proximate analysis and ultimate analysis of coal chars and petroleum cokes.

Sample Proximate analysis(wt%,d) Ultimate analysis(wt%,d)A V FC C H N St O*

A-1000 3.65 4.13 92.22 90.80 0.87 0.72 0.20 3.76A-1150 3.72 1.76 94.52 94.31 0.52 0.52 0.18 0.75A-1300 3.13 1.77 95.10 95.44 0.43 0.52 0.17 0.31A-1450 3.14 1.69 95.27 95.55 0.41 0.46 0.16 0.28A-1600 3.19 1.63 95.18 95.68 0.41 0.44 0.17 0.11PC-1000 0.42 1.85 97.73 92.97 0.44 1.69 3.20 1.28PC-1150 0.47 1.22 98.31 94.01 0.38 1.34 3.18 0.62PC-1300 0.43 1.04 98.53 94.88 0.37 0.76 3.19 0.37

97.09 0.37 0.36 1.39 0.34

A content.O

wom

u

2t

tdtptcO

�

wtms

th1

ti

dps

PC-1450 0.45 0.93 98.62

:ash content; V:volatile matter content; FC:fixed carbon content; St: total sulphur* = 100–C–H–N–S–A.

here � is the wavelength of the X-ray emission, �002 is the positionf the (0 0 2) diffraction peak, and �002 is the angular width at half-aximum intensity of peak (0 0 2).N2 adsorption–desorption measurements were performed

sing a Quantachrome instrument at −196 ◦C.

.5. Powder resistivity, real density, ultimate analysis andhermogravimetric analyses

The powder resistivity of the chars was tested using a GM-IIype resistivity meter. The testing method and basic principle areescribed as follows: 3.3 g of char with a particle size of 0.315 mmo 0.4 mm was placed into a cylindrical mold; and two conductinglates at the end of the mold were pressed with 784 N to connecthe char and then import direct current. The powder resistivity ofoke was automatically calculated according to Eq. (4) based onhm’s law:

= V S/Ih (4)

here � is the powder resistivity of coke, V is the voltage betweenhe two conducting plates, S is the inner cross-sectional area of

old, I is the current through the sample, and h is the height of theample bed.

The thermal properties of cokes were studied with the SDTQ600hermogravimetric (TG) analysis apparatus. The samples wereeated from room temperature to 1000 ◦C at a heating rate of0 ◦C min−1 under high-purity N2 at a flow rate of 100 ml/min.

Ultimate analysis was carried out by Vario MICRO cube elemen-ar (produced by Hanau, Germany). The analytical precision of thisnstrument is <0.1% abs for C, H, N, and S.

The real density of char with particle size of <74 �m wasetermined using the 3H-2000TD automatic real density analyzerroduced by Beishide Instrument-S&T Co., Ltd. This real den-ity analyzer has high testing accuracy (±0.04%) and repeatability

Fig. 2. TG–DTG plots of coal and petroleum coke during pyrolysis in N2.

(±0.02%). High-purity helium gas (99.999 v/v%) was used in thissystem.

3. Results and discussion

3.1. TG–DTG analysis

To understand the pyrolysis process of coal and petroleum coke,the variation in weight of the two samples from room temperatureto 1450 ◦C were studied by TG analysis. The TG–DTG curves of thecoal and petroleum coke are displayed in Fig. 2.

Fig. 2 shows that the weight losses of coal and petroleum coke

gradually increase with the increase of pyrolysis temperature. Inlow-temperature pyrolysis (below 350 ◦C), the weight losses of thetwo samples are small, which are mainly caused by dehydrationand removal of a small amount of volatile matter. At 350 ◦C–800 ◦C,

and Ap

amctArstsai

fowpilttda

3s

cfstps

c(ai(cauasd

toftcmhiiidTtwrgsr

J. Xiao et al. / Journal of Analytical

maximum amount of gas was discharged because of depoly-erization and decomposition reaction. The thermal instability

omponents, such as the side chains and functional groups aroundhe basic structure unit of samples, are continuously cracking.ccordingly, the low molecular compounds are formed and evapo-ated as volatile [17]. In high-temperature pyrolysis (above 800 ◦C),mall weight losses are still observed in the samples, which belongo the secondary devolatilization phase [18,19]. The polyconden-ation reactions, which mainly occur between aromatic structuresnd proceed to dehydrogenation simultaneously, play a major rolen this stage.

The degrees of weight losses of coal and petroleum coke are dif-erent. Petroleum coke and anthracite reached the maximum ratef weight loss at 608 ◦C and 661 ◦C, respectively; the maximumeight loss rate of coal is higher than that of petroleum coke. Beforeetroleum coke reaches the maximum weight loss rate (608 ◦C),

ts weight loss is higher than that of coal. However, the weightoss rate of coal is substantially higher than that of petroleum cokehereafter. This finding may be attributed to the fact that coal con-ains large amounts of stable aromatic rings, which are not easilyestroyed at low temperature, but can be decomposed at temper-tures above 600 ◦C [18].

.2. Effect of pyrolysis temperature on carbon crystallinetructure

Pyrolysis temperature significantly affects the structure ofarbonaceous materials, consequently determining their per-ormances. Therefore, understanding the changes in crystallinetructure of chars after high-temperature treatment is very impor-ant. The XRD patterns of coal char and petroleum coke at differentyrolysis temperatures are shown in Fig. 3; the correspondingtructural parameters of crystallite are displayed in Fig. 4.

As shown in Fig. 3, the intensities of the (0 0 2) peak of the twohars gradually increased with the increase of temperature, and the1 0 0) diffraction peak appeared. The stacking height Lc exhibitedn increasing trend, whereas the interplanar spacing d002 exhib-ted a declining trend with the increase of pyrolysis temperatureFig. 4). Thus, the crystallite structures of coal char and petroleumoke became more ordered with the increase of pyrolysis temper-ture. The interfacial defects between the adjacent basic structuralnits (BSUs) in the carbon structure disappeared gradually, and theromatic nucleus of chars increased in size because of the conden-ation reaction. However, the growth of BSUs is mainly in a verticalirection.

At the same pyrolysis temperature, the d002 of coal char is higherhan that of petroleum coke, but the Lc of coal char is less than thatf petroleum coke. In addition, the Lc of petroleum coke increasedaster than coal with the increase of temperature. The results showhat the graphitization degree of carbon structure in petroleumoke is higher than that in coal char after high-temperature treat-ent. The crystallite orientation in carbonaceous materials before

eat treatment or in the early stage of carbonization significantlynfluences the graphitization of carbonaceous materials. Accord-ng to the Franklin’s structure model [20], the crystallites arrangedn order belong to easily graphitized carbon, and the crystallitesistributed randomly are regarded as non-graphitizing carbon.herefore, crystallite orientation in petroleum coke is more orderedhan anthracite in the initial state. These results are also consistentith the TG analysis; the structural rearrangement of anthracite

equires higher temperatures. Several researchers have investi-ated the influence of pyrolysis temperature on the crystallitetructure of coal and petroleum coke, and they obtained similaresults [21–23].

plied Pyrolysis 117 (2016) 64–71 67

3.3. Effects of pyrolysis temperature on BET surface area

The effects of pyrolysis temperature on the BET surface area andaverage pore sizes of coal char and petroleum coke are presented inTable 4. The BET surface area of coal char decreased gradually from23.88 m2 g−1 to 4.53 m2 g−1, whereas the average pore sizes exhib-ited an increasing trend, with the increase of pyrolysis temperature(Table 4). The pore size distribution of coal char in Fig. 5 shows thatthe pore sizes of coal char concentrated in mesopore and micro-pore, which are less than 4 nm. Moreover, the peaks of the poresize distribution decreased. Therefore, the shrinkage degree of coalchar increases with the increase of temperature. This phenomenonresults in the reduction and disappearance of small pores (microp-ore and mesopore), as well as the reduction of BET surface area ofthe pores.

Contrary to the BET surface area of coal char, that of thepetroleum coke initially showed a decreasing trend, followed byan increasing trend. When the pyrolysis temperature was 1000 ◦C,the BET surface area of petroleum coke was 7.35 m2 g−1. Whenthe pyrolysis temperature increased to 1150 ◦C, the specific surfacearea decreased to 2.64 m2 g−1. However, when the pyrolysis tem-perature increased further, the BET surface area began to increase.This trend is opposite to that of the coal char in this temperaturerange. Wu and Gao [21] also found the similar trend of BET sur-face area of petroleum coke at temperatures above 1000 ◦C. Thus,the calcination temperature of petroleum coke should not exceed1350 ◦C [24]. Similar to coal char, the average pore size of petroleumcoke increases with the increase of temperature.

The pore size distribution of petroleum coke (Fig. 5) shows thatthe pores of petroleum coke are mainly mesopores and micropores(<10 nm); this pore size distribution is different from that of coalchar. The peak of the pore size distribution decreases first and thenincreases with the increase of pyrolysis temperature. This trendis consistent with the BET surface area. Furthermore, the pores ofpetroleum coke with >10 nm pore size increased significantly afterheat treatment at 1450 ◦C.

The reduction of BET surface area of coal char is partly causedby ash melting when the temperature increased to 1450 ◦C, whichis higher than the ash melting temperature. The ash melting blockspart of the char pores. Hence, the BET surface area decreased. Theeffect of ash melting on the surface area of petroleum coke is neg-ligible because of its low ash content.

3.4. Effects of pyrolysis temperature on air gasification reactivityand CO2 gasification reactivity

The effects of pyrolysis temperature on air gasification reactivityand CO2 gasification reactivity are illustrated in Fig. 6 . The gasifi-cation rate of coal char generally shows a decreasing trend withthe increase of pyrolysis temperature. Specifically, the variationtrend of air gasification reactivity is more significant, indicating thatthe increase of pyrolysis temperature can reduce the gasificationreactivity of coal char, which is very beneficial to reduce over-consumption of carbon products. Researchers generally believethat the carbon crystallite structure and BET surface area of car-bonaceous materials are the two main factors affecting gasificationreactivity. Smaller BET surface area or more ordered carbon crys-tallite structure decreases gasification reactivity [21,25–27]. XRDand BET analyses showed that the carbon crystallite structure ofchars tends to be ordered and the surface area decreases with theincrease of temperature. They work together to inhibit the gasifi-cation reactivity of coal char. Therefore, the gasification rate of coal

char generally decreases.

Both the air and CO2 gasification rates of petroleum cokedecreased with the increase of temperature, and the gasifica-tion rate reached the minimum value when the temperature

68 J. Xiao et al. / Journal of Analytical and Applied Pyrolysis 117 (2016) 64–71

Fig. 3. XRD patterns of coal chars and petroleum coke at different pyrolysis temperature (a)A-chars (b) PC-chars.

Fig. 4. Effects of the pyrolysis temperature on the carbon crystalline structure parameters.

Table 4The effect of pyrolysis temperature on BET surface area and average pore size.

Samples A1000 A1150 A1300 A1450 A1600 PC1000 PC1150 PC1300 PC1450

BET/(m2/g) 23.88 10.82 9.28 4.66 4.53 7.35 2.64 4.48 9.41Average pore size /nm 2.62 3.73 4.60 8.37 9.45 6.13 9.83 8.89 14.37

pore s

ibttt

Fig. 5. Effect of the pyrolysis temperature on the

s up to 1300 ◦C. The carbon structure of the petroleum coke

ecomes ordered with the increase of pyrolysis temperature,hereby decreasing the gasification rate. However, the gasifica-ion reactivity of petroleum coke is enhanced when the pyrolysisemperature is higher than 1300 ◦C because of the increase of

ize distribution of coal char and petroleum coke.

BET surface area. Accordingly, the carbon crystallite structure of

petroleum coke plays a major role in gasification reactivity attemperatures below 1300 ◦C. The effect of BET surface area on gasi-fication reactivity is stronger than that of the carbon structure whenthe temperature exceeds 1300 ◦C.

J. Xiao et al. / Journal of Analytical and Applied Pyrolysis 117 (2016) 64–71 69

F

etabiea[cm

hcpicah[lnthfs

3

tdaforcrnft

hrce

ig. 6. Effects of the pyrolysis temperature on the air and CO2 gasification reactivity.

In addition, the inherent minerals exert remarkable catalyticffects on gasification reactivity, consequently increasing the reac-ion rate. Thus, the difference between air gasification reactivitynd CO2 gasification reactivity of petroleum coke can be explainedy the catalytic effect of impurity elements. Impurity elements

n petroleum coke, such as V, Ni, and Na, have strong catalyticffects on carbon–air reaction. Moreover, impurity elements suchs Ca, Ni, and Na significantly influence the carbon–CO2 reaction14,15,28–30]. Table 2 shows that the V content in the petroleumoke is obviously higher; hence, the air gasification reactivity isore obvious than the CO2 gasification reactivity.Furthermore, the CO2 gasification rate of coal char is slightly

igher than that of petroleum coke, indicating that the petroleumoke is less reactive than coal char for CO2 gasification at this tem-erature range. However, the air gasification rate of petroleum coke

s much higher than that of coal tar. On the one hand, coal mainlyontains impurity elements, such as Al, Fe, Si, and Ca, they inter-ct with each other to form compounds with no catalytic activity;ence, they play an inhibiting role on air gasification reactivity24,31,32], and V and Ni with strong catalytic abilities are relativelyess or do not exist. On the other hand, ash melting of coal char sig-ificantly influences the gasification reaction. The ash melts coverhe surface and block the pores of the coal char particles, therebyindering the gasification reaction [33,34]. Considering all these

actors, the capability of coal char to resist air gasification is moreignificant than that of petroleum coke.

.5. Effect of pyrolysis temperature on powder resistivity

The effect of pyrolysis temperature on powder resistivity is plot-ed in Fig. 7 . Both coal char and petroleum coke show a rapidecrease in their powder resistivity with the increase of temper-ture. The final carbonization temperature is the most importantactor affecting the powder resistivity of coke. The rapid decreasef powder resistivity with the increase of temperature is mainlyelated to several factors [35,36]. First, the volatile substances areontinuously discharged from the material. Second, hydrogen iseleased with the breaking of C H bond, consequently formingew free electrons. Third, the conductivity of coke depends on the

ormation of conjugated � bond, which increases with the aroma-ization of coke.

As shown in Fig. 7, the powder resistivity of coal char is far

igher than that of petroleum coke. According to China’s nonfer-ous industry standard YS/T 625–2007, the powder resistivity ofalcined coke for carbon anode should be ≤610 � � m. However,ven after calcination at 1600 ◦C, the powder resistivity of coal char

Fig. 7. The effect of pyrolysis temperature on electrical resistivity.

remains very high; hence, improving the heat treatment tempera-ture becomes necessary. The powder resistivity of petroleum cokemeets the requirement for carbon anode even when the pyrolysistemperature is only higher than 1150 ◦C.

The results showed that pyrolysis temperature has significanteffect on both powder resistivity and structural parameters ofthe carbon crystallites of chars. Thus, a relationship should existbetween the powder resistivity and structural parameters of crys-tallites. The correlations between powder resistivity and structuralparameters of carbon crystallite (Lc, d002) are shown in Fig. 8. A closeexponent correlation is observed between Lc and powder resis-tivity, and d002 demonstrates a good linear dependence relationwith powder resistivity, especially both the correlation coefficientsare >0.93. Therefore, the carbon crystalline structure plays a vitalrole in determining powder resistivity. Larger Lc or smaller d002values indicate higher ordered crystalline structure, resulting inlower electrical resistivity [37]. Hence, powder resistivity can beconsidered a structure-sensitive parameter reflecting the internalstructure of chars [22].

3.6. Effect of pyrolysis temperature on real density

Real density is related to the types of materials and heat treat-ment conditions, as well as the porosity of materials. In the processof aluminum electrolysis, using carbon anode with higher real den-sity can reduce the replacement frequency; thus operational costsare also reduced.

The effect of pyrolysis temperature on the real density of thechars are shown in Fig. 9 . The real density of petroleum cokeincreases with the increase of pyrolysis temperature, which isattributed to the molecular structure rearrangement of petroleumcoke, while hydrogen, oxygen, nitrogen, and other impurity atomsare continuously discharged.

Notably, the real density of coal chars declined with the increaseof temperature, and this result appears to conflict with the XRDand BET surface area analyses, which showed that carbon crys-tallite structure of coal char tends to be ordered, and the BETsurface area decreases. To explain the decrease of real density ofcoal chars, proximate analysis and ultimate analysis of coal charsand petroleum cokes are added in Table 3. With the increase oftemperature, the content of C increases, and the contents of N,H, O, and S gradually decrease (Table 3). The spatial arrangement

and abundance of the elements: C, H, N, O, and S often correlateor directly influence the coal properties, including helium density[38]. With the increase of graphitization degree, the side chainsand functional groups on the molecules are reduced, oxygen and

70 J. Xiao et al. / Journal of Analytical and Applied Pyrolysis 117 (2016) 64–71

Fig. 8. Correlation between the carbon crysta

noimbdtpwtccgc

oditsrr

4

pc

Fig. 9. The effect of pyrolysis temperature on real density.

itrogen elements in molecules are also rapidly reduced simultane-usly. Although, the reduction of side chains and functional groupss beneficial to the enhancement of density, the relative atomic

ass of oxygen and nitrogen is relatively larger than that of car-on. As a result, the decrease of oxygen and nitrogen contents has aominant effect on the real density of coal chars at 1000 ◦C–1600 ◦C,his effect results in the decrease of the real density. For the calcinedetroleum coke, a very compact structure of aromatic compoundsas formed with the rapid increase of the order degree of the crys-

allite structure. Accordingly, the real density of calcined petroleumoke increased rapidly. The different trends of real density betweenoal char and petroleum coke may be attributed to the fact that theraphitization degree of coal char is lagging behind the petroleumoke.

As a result, the real density of petroleum coke is larger than thatf coal char. For instance, after heat treatment at 1300 ◦C, the realensity of petroleum coke is 2.085 g cm−3, whereas that of coal char

s only 1.749 g cm−3. According to Chinese nonferrous metal indus-ry standard YS/T 625–2007, the real density of a carbon materialhould be >2.01 g cm−3. Therefore, the real density of coal char iselatively small, which is a disadvantage for a carbonaceous mate-ial in preparing carbon anode.

. Conclusion

The effects of high-temperature pyrolysis on the structure andhysicochemical properties of low-ash anthracite and petroleumoke were investigated in this study. The differences in physical and

lline structure and electrical resistivity.

chemical properties were also analyzed. The results are presentedas follows:

(1) The crystallite structures of coal char and petroleum cokebecame more ordered with the increase of temperature. Com-pared with petroleum coke, structural rearrangement of coalchar requires higher temperatures. Therefore, the graphitiza-tion degree of coal char is lower than that of petroleum coke atthe same pyrolysis temperature.

(2) With the increase of pyrolysis temperature, the BET surface areaof coal char decreased gradually, whereas the BET surface areaof the petroleum coke exhibited a decreasing trend after ini-tial increasing trend. The average pore sizes of coal char andpetroleum coke showed an increasing trend.

(3) The carbon crystallite structure and BET surface area generallyinhibited the gasification reactivity of coal char. The gasificationreaction was first inhibited and then promoted for petroleumcoke. The air gasification reactivity of coal char is remarkablylower than that of petroleum coke, but the CO2 gasificationreactivity of coal char is slightly stronger.

(4) The powder resistivity of coal char and petroleum coke showeda rapid decrease with the increase of temperature, and the pow-der resistivity of coal char is considerably higher than that ofpetroleum coke. Moreover, the structural parameters of carboncrystals (Lc and d002) are well correlated with powder resistiv-ity.

(5) The real density of petroleum coke increased with the increaseof pyrolysis temperature. By contrast, the real density of coalchar declined.

Compared with petroleum coke, coal char presents disadvan-tages in terms of ash content, powder resistivity and real density.For this reason, we attempt to prepare the carbon anode by blend-ing part of coal char with petroleum coke. In this way, the propertiesof the mixed coke will be controlled within allowable ranges, andthe requirements of carbon anode will be satisfied.

Acknowledgments

This work was funded by the National Natural Science Founda-tion of China (51374253). The authors are also grateful to Miss Wufor the advice of this paper.

References

[1] Y. Luo, Z. Su, Z. Luo, L. Zhao, Research on producing anode with materialreplacing petroleum coke, Light Met. 8 (2011) 48–50 [Chinese].

and Ap

[

[

[

[[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[37] P.A. Eidem, M. Tangstad, J.A. Bakken, Determination of electrical resistivity of

J. Xiao et al. / Journal of Analytical

[2] R. Andrews, D. Jacques, T. Rantell, Mild coal extraction for the production forthe production of anode coke, Light Met. (2009) 963–968.

[3] Y. Zhao, Y. Zhang, H. Zhang, Q. Wang, Y. Guo, Structural characterization ofcarbonized briquette obtained from anthracite powder, J. Anal. Appl. Pyrol.112 (2015) 290–297.

[4] Y.J. Kim, H.J. Yang, S.H. Yoon, Y. Korai, I. Mochida, C.H. Ku, Anthracite as acandidate for lithium ion battery anode, J. Power Sources 113 (2003) 157–165.

[5] C. Nieto-Delgado, F.S. Cannon, P.D. Paulsen, J.C. Furness, R.C. Voigt, J.R.Pagnotti, Bindered anthracite briquettes as fuel alternative to metallurgicalcoke: full scale performance in cupola furnaces, Fuel 121 (2014) 39–47.

[6] C.E. Burgess-Clifford, D.L. Narayanan, D.T. Van Essendelft, P. Jain, A. Sakti, A.D.Lueking, The effect of calcination on reactive milling of anthracite as potentialprecursor for graphite production, Fuel Process. Technol. 90 (2009)1515–1523.

[7] R.J. Andrews, T. Rantell, D. Jacques, J.C. Hower, J. Steven Gardner, M. Amick,Mild coal extraction for the production of anode coke from blue gem coal,Fuel 89 (2010) 2640–2647.

[8] M. Hamaguchi, N. Okuyama, N. Komatsu, J. Koide, K. Kano, Prebaked anodefrom coal—utilization of coal extract as a coke feedstock, Light Met. (2010)923–928.

[9] M. Hamaguchi, N. Okuyama, T. Shishido, K. Sakai, N. Komatsu, T. Inoue, J.Koide, K. Kano, Prebaked anode from coal extract(2)–effects of the propertiesof hypercoal-coke on the performance of prebaked anodes, Light Met. (2011)913–916.

10] M. Hamaguchi, N. Okuyama, T. Shishido, K. Sakai, N. Komatsu, N. Kikuchi, T.Inoue, Prebaked anode from coal extract(3)–carbonization properties ofhypercoal its blends with binder pitch, Light Met. (2012) 1219–1221.

11] Y. Ren, Z. Sun, J. Tian, W. Lai, Study on electrolytic aluminium carbon anodepreparation with calcined anthracite, Proceedings of the InternationalConference on Mining Science & Technology 1 (2009) 694–700.

12] J. Xue, M. Han, J. Zhu, L. Feng, H. Ma, Investigation on air reactivity andelectrolysis consumption of anode carbons with anthracite additions, LightMet. (2012) 1229–1233.

13] SH/T0527-1992, the standard of delayed petroleum coke. [Chinese].14] F. Chevarin, L. Lemieux, D. Ziegler, M. Fafard, H. Alamdari, Air and CO2

reactivity of carbon anode and its constituents: an attempt to understanddusting phenomenon, Light Met. (2015) 1147–1152.

15] K.N. Tran, A.J. Berkovich, A. Tomsett, S.K. Bhatia, Influence of sulfur and metalmicroconstituents on the reactivity of carbon anodes, Energy Fuels 23 (2009)1909–1924.

16] YS/T587. 7-2006, Detection method of calcined petroleum coke for carbonanode. [Chinese].

17] Z.P. Zhu, J.S. Gao, Coal Chemistry, Shanghai Scientific and TechnicalPublishers, Shanghai, 1984 [Chinese].

18] J. Wang, J. Du, L. Chang, K. Xie, Study on the structure and pyrolysis

characteristics of Chinese western coals, Fuel Process. Technol. 91 (2010)430–433.

19] H. Hui, X. Fu, X. Wang, Y. Wei, J. Jia, X. Shu, Effect of washing process on thestructure andpyrolysis characteristics of coal, Proceedings of the CSEE 33(2013) 68–74 [Chinese].

[

plied Pyrolysis 117 (2016) 64–71 71

20] R. Franklin, Crystallite growth in graphitizing and non-graphitizing carbons,Proc. Roy. Soc. 209 (1951) 196–218.

21] Y. Wu, S. Wu, J. Gu, J. Gao, Differences in physical properties and CO2

gasification reactivity between coal char and petroleum coke, Process Saf.Environ. 87 (2009) 323–330.

22] B. Feng, S.K. Bhatia, J.C. Barry, Structural ordering of coal char during heattreatment and its impact on reactivity, Carbon 40 (2002) 481–496.

23] Y. Wu, S. Wu, S. Huang, J. Gao, Physicochemical properties and structuralevolutions of gas-phase carbonization chars at high temperatures, FuelProcess. Technol. 91 (2010) 1662–1669.

24] P. Wang, Z. Gong, L. Jia, Q. Li, Y. Luo, Z. Tan, The Aluminum Electrolysis CarbonAnode Production and Application, Press of Metallurgy Industry, Beijing, 2010[Chinese].

25] A. Zolin, A.D. Jensen, P.A. Jensen, K. Dam-Johansen, Experimental study of charthermal deactivation, Fuel 81 (2002) 1065–1075.

26] M.J.G. Alonso, A.G. Borrego, D. Alvarez, A. Menendez, A reactivity study ofchars obtained at different temperatures in relation to their petrographiccharacteristics, Fuel Process. Technol. 69 (2001) 257–272.

27] L.M. Lu, C.H. Kong, V. Sahajwalla, D. Harris, Char structural ordering duringpyrolysis and combustion and its influence on char reactivity, Fuel 81 (2002)1215–1225.

28] S. Hume, W. Fischer, R. Perruchoud, B. Welch, A model for petroleum cokereactivity, Light Met. (1993) 525–531.

29] G. Houston, H. Øye, Consumption of anode carbon during aluminiumelectrolysis, Aluminium 61 (1985) 251–254.

30] S.M. Hume, Influence of Raw Material Properties on the Reactivity of CarbonAnodes Used in the Electrolytic Production of Aluminium, Aluminium-Verlag,Auckland, New Zealand, 1993.

31] X. Jing, Z. Wang, Q. Zhang, Z. Yu, C. Li, J. Huang, Y. Fang, Evaluation of CO2

gasification reactivity of different coal rank chars by physicochemicalproperties, Energy Fuels 27 (2013) 7287–7293.

32] K. Miura, K. Hashimoto, P.L. Silveston, Factors affecting the reactivity of coalchars during gasification, and indices representing reactivity, Fuel 68 (1989)1461–1475.

33] J. Bai, W. Li, C. Li, Z. Bai, B. Li, Influences of mineral matter on high temperaturegasification of coal char, J. Fuel Chem. Technol. 37 (2009) 134–138 [Chinese].

34] S.Y. Wu, J. Gu, X. Zhang, Y.Q. Wu, J.S. Gao, Variation of carbon crystallinestructures and CO2 gasification reactivity of Shenfu coal chars at elevatedtemperatures, Energy Fuels 22 (2008) 199–206.

35] M. Kumar, R. Gupta, Electrical resistivity of Acacia and Eucalyptus wood chars,J Mater. Sci. 28 (1993) 440–444.

36] K. Mochidzuki, F. Soutric, K. Tadokoro, M.J. Antal, M. Toth, B. Zelei, G.Varhegyi, Electrical and physical properties of carbonized charcoals, Ind. Eng.Chem. Res. 42 (2003) 5140–5151.

dry coke beds, Metall. Mater. Trans. B 39 (2007) 7–15.38] J.P. Mathews, V. Krishnamoorthy, E. Louw, A.N. Tchapda, F. Castro-Marcano, A

review of the correlations of coal properties with elemental composition, FuelProcess. Technol. 121 (2014) 104–113.