School of Chemical Engineering,Northwest University,Xi'an 710069,China International Scientific and Technological Cooperation Base of the Ministry of Science and Technology(MOST)for Clean Utilization of Hydrocarbon Resources,Xi'an 710069,China Chemical Engineering Research Center of the Ministry of Education(MOE)for Advanced Use Technology of Shanbei Energy,Xi'an 710069,China Shaanxi Research Center of Engineering Technology for Clean Coal Conversion,Xi'an 710069,China Collaborative Innovation Center for Development of Energy and Chemical Industry in Northern Shaanxi,Xi'an 710069,China

Keywords:Low-rank coal Coal direct liquefaction residue Co-pyrolysis Kinetics

ABSTRACT To reasonably utilize the coal direct liquefaction residue(DLR),contrasting research on the co-pyrolysis between different low-rank coals and DLR was investigated using a TGA coupled with an FT-IR spectrophotometer and a fixed-bed reactor.GC-MS,FTIR,and XRD were used to explore the reaction mechanisms of the various co-pyrolysis processes.Based on the TGA results,it was confirmed that the tetrahydrofuran insoluble fraction of DLR helped to catalyze the conversion reaction of lignite.Also,the addition of DLR improved the yield of tar in the fixed-bed,with altering the composition of the tar.Moreover,a kinetic analysis during the co-pyrolysis was conducted using a distributed activation energy model.The co-pyrolysis reactions showed an approximate double-Gaussian distribution.

1.Introduction

The low-rank coal reserves in China constitute approximately 52.3 billion tons,and account for 42%of the total coal proven reserves[1].Due to the high volatile content and reactivity of these reserves,pyrolysis of low-rank coals has attracted broad attention.Although there has been significant progress in the research on pyrolysis of low-rank coal,the low yield and poor quality of the tar product are persistent problems that need to be solved.It is widely believed that hydrogenation and condensation reactions of free radicals occur competitively during pyrolysis of low-rank coal[2].The former reactions form tar molecules,while the latter generate heavy tar and char.Consequently,stabilizing the free radicals and depressing crosslinking reactions could be a solution for enhancing tar yield which would upgrade the quality of the tar[3].

The DLR from the Chinese Shenhua's coal direct liquefaction plant contains an oil-rich fraction (HS,A and PA),carbon-rich substances and minerals (THFIS),which account for about 30% of raw coal feed[4].Efficient utilization of this material directly affects the economic and environmental sustainability of the coal direct liquefaction process.However,since DLR has a low softening point,poor feeding of the raw material adversely impacts its individual pyrolysis.On the other hand,compared to these coal,DLR has a lower oxygen content,but it has more alkali and alkaline earth metallic(AAEM)species,and solid liquefaction catalysts,which can improve coal pyrolysis[5-10].Therefore,co-pyrolysis of low-rank coals together with DLR has been proposed as a means for improving tar yields by exploiting the properties of DLR.The advantages of this approach are manifest:(1)making use of the oil in DLR,(2)overcoming the feed problem of DLR pyrolysis,and(3)the minerals and liquefaction catalysts in DLR can increase coal conversion.

To date,in most of the research conducted on this subject,the DLR that has been used originated from laboratory or pilot scale coal liquefaction plants.Studies employing industrial scale plant DLR have been very limited.Of the related research,Li et al.[4]suggested that the organic components of DLR were responsible for the interaction during the co-pyrolysis of HL and DLR.This organic matter played the role of a hydrogen donor during the co-pyrolysis.Xu et al.[5]claimed that the inorganic components of DLR exhibited a catalytic effect during pyrolysis in a N2atmosphere.Since DLR is generally separated into four fractions:HS(hexane soluble),A(asphaltene),PA(preasphaltene),and THFIS(tetrahydrofuran insoluble),the characteristics and interactions of the coal and DLR during co-pyrolysis are controlled by the properties and content of the four DLR fractions.However,the mechanisms of these interactions and the influence of the DLR fractions on the product distribution of the devolatilization are still ambiguous.Moreover,there have been no reported studies on the kinetics of co-pyrolysis to date.

In this reported study,the characteristics of the co-pyrolysis of two kinds of low-ranked coals (Hulun Buir lignite and Shendong sub-bituminous coal)and DLR were explored using a TGA coupled with an FT-IR spectrophotometer and a fixed-bed reactor.The objective of this study was to investigate the mechanism during the co-pyrolysis of low-rank coal and DLR,and to reveal the source and mode of these kinetic interactions.The systematic research results of the pyrolysis reactivity,gas release,and the interactions during the co-feed of coal and DLR in a fixed-bed reactor,were recorded for a future simulation to optimize high-efficiency co-pyrolysis processes.

2.Experimental

2.1.Samples

The coal used in this experimental investigation was Hulun Buir lignite (HL)and Shendong bitumite (SD).The residue was Shenhua coal direct liquefaction residue(DLR)that was obtained from a Shenhua coal direct liquefaction industrial plant.To explore the relationship between the co-pyrolysis characteristics and DLR's structure and composition,the four organic fractions in the DLR was separated by the Soxhlet extraction which is an accurate and necessary research method to achieve the separation of these organic components with high boiling points.A DLR sample of 1 g was used in the successive extraction with hexane,toluene,and tetrahydrofuran.In each extraction run,the solvent of 150 ml was used to separate one fraction from the residual DLR for 48 h.When one extraction run was over,the solvent in the Soxhlet extractor was nearly colorless.The DLR sample was composed of heavy oil (HS,16.6%),asphaltene (A,28.5%),preasphaltene (PA,2.0%),and carbon-rich materials(THFIS,52.7%).The HL,SD and DLR,were pulverized and sized to 96-180 μm,and dried overnight in a vacuum oven at 110 °C to exclude moisture before the experiments.Table 1 shows the proximate analysis,ultimate analysis,and ash composition analysis of the experimental samples.

Table 1 Proximate and ultimate analyses along with ash composition(wt%)of samples

So as to exclude the effect of coal minerals on the co-pyrolysis reaction,HL and SD coal samples were washed in sequence with HCl and HF(Tianjin Fuyu Chemical Company,analytical grade)with stirring them in solution at normal temperature for 24 h[6].Following this step is washing the demineralized coal samples using distilled water till the filtrates were neutral.The ash content of the acid-washed coals was determined using GB/212-2008 method,and both were found to have less than 0.5 wt% mineral content.All the coal samples used in the tests were demineralized in this manner.

2.2.Pyrolysis experiments

The pyrolysis tests were proceeded using a quartz fixed-bed reactor(10 mm i.d.),which had an effective heated zone of 40 mm.In each experiment,about 3 g of sample was heated to a preset temperature at 5 °C·min?1and maintained at that temperature for 30 min with a concurrent,continuous N2(purity ≥99.999%)gas flow of 200 ml·min?1.The pyrolysis temperatures were within the scope of 400-800 °C.Liquid products were collected in a cold trap that was maintained at ?15°C,then washed with acetone.The gaseous products were analyzed online using a gas chromatograph (GC-9790 II,Fuli,China).Due to the block problem in the fixed-bed reactor during the DLR individual pyrolysis,the DLR sample was blended with quartz sand under the blending ratio of 1:9.The particle size of quartz sand was in range of 250-425 μm,so that the char was easy to separate from the quartz sand by screening.

A TGA(STA449 F3,NETZSCH)coupled with a FT-IR Spectrometer(Vertex 70,Bruker)was used to determine the devolatilization behavior of samples.For each run,a coal sample of 10 mg was pyrolyzed in the temperature range of 25 to 1000°C under 30°C·min?1in a N2of 60 ml·min?1.The kinetic analysis was investigated by four different heating rates(10,30,50,and 70°C·min?1).An FTIR spectrophotometer containing a 200 ml gas cell which was maintained at 200°C was used to analyze the gas products of samples during the pyrolysis.The spectral region of the FTIR was 400-4000 cm?1,and the resolution of the FTIR was 4 cm?1.During pyrolysis,to refrain from the condensation of tar,the Teflon pipe connecting the TGA and FTIR was kept at 200°C.The main peaks in the FTIR spectra were specified as follows:light aromatics at 3051 cm?1[11];C2-C4at 2930 cm?1;CH4at 3103 cm?1[12];CO at 2181/2110 cm?1[13];CO2at 2360/669 cm?1[14,15];and H2O at 1500/3700 cm?1[16].

2.3.Error and analysis

All the tests were repeated more than 3 times to confirm the modifiability of the results,as well as to evaluate the test errors.In the process,all the results were figured up to produce the arithmetical averages,and the standard error of the yield was found to be 0.2.

The tar composition was determined using a GC-MS instrument(GCMS-QP 2010 plus,Shimadzu)with an Rtx-5ms capillary column(Restek,30 m×0.25 mm;i.d.,0.25 μm).XRD analysis of the DLR and the THFIS fraction was executed by an XRD-3000 powder diffractometer(Shimadzu)between 10 and 85°C.

3.Results and Discussion

3.1.Effect of DLR on product distribution during the co-pyrolysis

The effect of blending ratio on co-pyrolysis was investigated using thermal analysis.TG and DTG curves from the blended samples employing various blending ratios under a N2atmosphere are shown in Fig.1.For the co-pyrolysis of HL and DLR,all the blended samples exhibited a higher mass loss than what had been calculated.Moreover,the temperature of the maximum mass loss rate during co-pyrolysis was about 30 °C lower than the calculated mass loss,indicating that the activation energy during the co-pyrolysis might have become lower with the addition of DLR to the mixture.A further discussion of this will follow.

For both HL and SD,the effect of blending ratio was negligible.This was probably the result of the small differences between the blending ratios that were used.Taking into account the annual production of DLR in China,only blending ratios under 15 wt%were investigated in this work.A DLR blending ratio of 10 wt%was used in the remainder of this research.

Fig.1.TG and DTG curves of blending samples of HL,SD and DLR with different blending ratios.

Fig.2.Effect of temperature on co-pyrolysis yields and H/C ratio of char in fixed bed.

As exhibited in Fig.2,the tar,char,gas yields and H/C ratio of char in co-pyrolysis (10 wt% DLR+90 wt% HL)varied with temperature.According to Fig.2,the char yield decreased from 80 wt%to 67 wt%as the temperature of the process was increased.By contrast,the tar yield increased with temperature,and attained a peak value at 550°C,after that reduced.When the DLR was melted,a few volatiles were stuck in the DLR plastic mass,which produced a reduction of the tar yield.As the temperature was raised,the fluidity of the plastic mass was lowered,which produced a higher mass transfer resistance.Therefore,the decrease of co-pyrolysis tar yield was a result from the secondary reaction and the inferior flow of the DLR,which increased the mass transfer resistance and the retention period of the volatile compounds.That is to say,two factors affected the co-pyrolysis:heating and the mass transfer processes.The gas yield increased with temperature,perhaps due to the escape of small molecules from the tar during secondary cracking [20].Moreover,the H/C ratio of the char decreased with the temperature increased.The greatest tar yield occurred at 550°C,so further analysis was performed with the tar at 550°C.

As illuminated in Fig.3,during co-pyrolysis of DLR and HL,the tar was produced more than the calculated yield (+2.4 wt%),while the char was produced less (?1.2 wt%).Both the organic and inorganic components in the DLR influenced the co-pyrolysis.On the one hand,the organic compounds produced a plastic mass,which increased the mass transfer resistance,resulting in a decrease of the tar yield.For another,the minerals such as calcium and iron were enriched in the DLR,which promoted hydrogen transfer to the coal.The pyrite increased the efficiency of hydrogenation of radicals during co-pyrolysis.In an overall view,there was a positive interactive effect during co-pyrolysis of DLR and HL.

3.2.Effect of DLR on the components in liquid products

Fig.3.Comparison between experimental and calculated yields of tar,gas,and char in the co-pyrolysis(DLR and HL)in fixed bed.

Fig.4.GC-MS total ion chromatogram of tar from the co-pyrolysis at 550°C.

The GC-MS total ion chromatogram of tar from 550°C during copyrolysis is shown in Fig.4.As can be seen,the first three abundant components were phenols(22.3 wt%),indenopyrene(13.4 wt%)and naphthalenes (7.7 wt%).Table 2 shows the detailed content of the major components of the tar from the co-pyrolysis.The tar from the co-pyrolysis contained 2-6 ring polycyclic aromatic hydrocarbons(43.1 wt%),phenolic compounds (22.3 wt%),long-chain alkanes(13.6 wt%,C20,C21,C32),and a few light aromatics(2.3 wt%).

Table 2 Content of major components of tar during co-pyrolysis of DLR and HL at 550°C

Fig.5.Comparison of the major component content of the tar from the co-pyrolysis of DLR and HL and from the individual pyrolysis of HL at 550°C.

The comparison of the tar from the co-pyrolysis of DLR and HL and from the individual pyrolysis of HL at 550°C is shown in Fig.5.The percentage of the polycyclic aromatics of tar during the co-pyrolysis was higher than that in individual HL pyrolysis tar.This occurred for the following reasons:(1)DLR was rich in macromolecular compounds,such as asphaltene and preasphaltene.The oxygen-containing functional groups and chains in DLR were nearly destroyed during coal liquefaction;consequently,the aromaticity of the DLR was relatively higher.Generally speaking,the aromatic hydrocarbons tended to condense,rather than decompose,to form polycyclic aromatic hydrocarbon macromolecules.(2)The plastic mass from the DLR impeded the diffusion of free radicals,resulting in more recombination,which generated the heavier volatiles.(3)The AAEMs in the DLR produced an increase in the heavy components in the tar[22].However,in comparison to the DLR tar,the content of phenolic and aliphatic compounds increased in the co-pyrolysis product,indicating that co-pyrolysis was conductive to upgrading the DLR tar components.

3.3.Effect of DLR on evolution of gaseous products

There is a popular belief that pyrolysis processes include decomposition reaction and secondary gaseous reaction [17].Fig.6 illuminates the evolution of the gaseous products during co-pyrolysis of DLR and HL.The contents demonstrated in Fig.6 were the instantaneous volume percentages at the exit of the fixed bed.The carboxyls of HL and DLR (Fig.6(d))largely decomposed to CO2at low temperature.Then,the aliphatic chains in the coal and DLR molecules were cracked to produce CH4(Fig.6(c))and C2-C4.In the meantime,the evolution of some CO was caused by the break of the ether structures (Fig.6(b)).After devolatilization and repolymerization of the aromatics at about 600 °C,char was formed.Secondary reactions between the volatiles,such as CO2,and the char might have occurred at higher temperature (700 °C)producing much more CO [21].Furthermore,AAEMs enriched in DLR could more or less promote secondary reactions [22].

Above 450°C,the rate of production of H2during co-pyrolysis was much slower than that during the HL pyrolysis alone (Fig.6(a)).However,the tar yield increased during co-pyrolysis.Therefore,this was an indication that DLR delayed the release of H2,rather than simply depressing it.In this way,more hydrogen free radicals from the decomposition reactions combined with the tar fragments.This may have occurred,because the Ca,Fe,and pyrite in the DLR catalyzed the transfer of hydrogen to the coal[6].

Fig.6.Evolution of gas products in the co-pyrolysis of HL and DLR in fixed-bed reactor.

3.4.Effect of DLR on functional groups of solid products

Fig.7.FTIR spectra of HL and co-pyrolysis char.

Fig.7 demonstrates the FTIR spectrogram of the char formed in the co-pyrolysis(HL and DLR).As shown,there were some common features between the FTIR spectrogram of the char generated from various temperatures:the vibration at 1451 cm?1(methylene or methyl group),as well as the stretching vibration of hydroxy at 3429 cm?1.In addition,a few obvious variations existed in the spectrogram of the char and the HL.The stretching vibration of aliphatic(2920 cm?1)and aromatic C--H(3041 cm?1)nearly disappeared in the char,and the substituted functional groups of the aromatic rings in the range from 900 to 700 cm?1[31]moved toward smaller wavenumbers.The vibration of Si--O at 1151 cm?1gradually faded away at 800°C.The vibration of the carbonate stretching (494 cm?1)and the vibration of methyl group(1310 cm?1)also gradually became smoother.As the pyrolysis temperature was increased,some weak peaks in the FTIR spectra of the char disappeared.Observations showed that the degree of graphitization of the char increased with temperature,particularly above 600°C,which agreed with the observed change of H/C ratio of the char.

3.5.Effect of four fractions of DLR on the devolatilization and on the evolution of functional groups

The DLR was comprised of four fractions:HS,A,PA,and THFIS.The effect of DLR on co-pyrolysis behavior of the coal was controlled primarily by the property and proportion of the four fractions in the DLR.To determine the interaction between the DLR and the coal,the HS,A,PA,and THFIS fractions were individually blended with the coal in a blending ratio of 10 wt%.The theoretically calculated mass loss was obtained using the following equation:

where Wcis the calculated mass loss,Wirepresents the individual pyrolysis mass loss of each fraction in DLR,and Wjrepresents the individual pyrolysis mass loss of coals.

The deviation (ΔWt)between experimental (WE)and calculated(WC)mass loss was calculated as follows:

It can be concluded from this expression that when ΔWtwas large,a more intense interaction occurred between the various DLR fractions and the coal.For example,if ΔWtwas greater than zero,the interaction would be conductive to the co-pyrolysis;while in case that ΔWtwas negative,the interactive effect would become adverse to pyrolysis.

Fig.8 shows the different ΔWtof DLR fractions and coal for this system.As can be seen,the THFIS fraction favored co-pyrolysis devolatilization,particularly in the case of the HL.Overall,organic components,such as HS and PA,had a negative effect on the devolatilization of both coal specimens.The DLR organic components exhibited two different effects on co-pyrolysis:first,they played the role of a hydrogen donor during devolatilization [4]and second,the poor fluidity of the system hindered the mass transfer process inside the particles.To effectively achieve hydrogenation of the free radicals,the donors and receptors needed to be in contact on the molecular level[17].In the case of the solid-solid reactions between the DLR and the coal,it was difficult for this contact to materialize.Moreover,organic substances in the DLR contained limited active components,which offered few hydrogen free radicals following the extreme direct liquefaction[18].Considering that the plastic mass from the organic fractions might plug the coal pores,the positive effect from the organic components was partly counteracted.On the other hand,enriched minerals and the residual liquefaction catalysts(Fe1?xS)in the THFIS were probably conductive to devolatilization[19].

Fig.8.Comparison of various ΔWtduring co-pyrolysis of DLR fractions and coals.

The type of coal in the co-pyrolysis process was an important factor that affected the co-pyrolysis behavior[2].In the case of SD,the positive effect from the THFIS was not sufficient to offset the negative effect of the HS and PA.The Fe-based catalyst offered no remarkable effect on SD during pyrolysis,which agreed with the results reported by Jiang[30].In addition,the SD coal was used in the Shenhua direct liquefaction plant to produce liquid products[19].The active components in the DLR were‘homogenous’with the SD.However,numerous decomposition and condensation of SD coal macromolecules had already been produced during the direct liquefaction process.The active components experienced an ultra-liquefaction condition with the catalyst and high hydrogen pressure.So,the residual substances in the DLR,were more stable during pyrolysis for the SD.Therefore,it was found that co-pyrolysis might not improve the SD devolatilization.By contrast the higher H/C ratio of HL in comparison to SD,might make the HL more sensitive to the active components in the DLR,especially the liquefaction catalyst.As shown in Fig.8,the minerals and liquefaction catalysts that were enriched in the DLR actually intensified the conversion of the HL.Li et al.suggested that this interactive effect may be caused by the THFS(the tetrahydrofuran soluble)in the DLR,because the addition of THFIS did not impact the tar yield.In their work,a comparison was made between the yield of the co-pyrolysis tar and the lignite pyrolysis tar.Since the tar yield from the THFIS was so much lower than that from the lignite,the massed average value must be used in this comparison.In fact,in their work,the tar yields from the experiments were all higher than those calculated,indicating that THFIS was conductive to the co-pyrolysis process.

During the coal liquefaction process,an Fe-based catalyst was added and the main component of which was a complicated Fe-S compound(FeS2)[7].Fig.9 shows the XRD spectra of the THFIS and DLR.Once the catalyst was added the following reaction occurred during coal liquefaction[18]:

Fig.9.XRD spectra of THFIS and DLR.

Fe7S8was the major compound of remanent liquefaction catalyst in the DLR[7].

The release of gas products in the co-pyrolysis is reflected by the FTIR absorption intensity of CO2,C2-C4,CH4,light aromatics,and CO shown in Fig.10(f).With regard to co-pyrolysis,as illustrated in Fig.10(f),at low temperature(300-500 °C),the decomposition of carboxy[2,20]contributed to the first peak of CO2.The initial temperature of CO2evolution,caused by decarboxylation reactions,was as low as 200°C[22].In Fig.10(f),the second peak of CO2evolution was attributed to the decomposition of carbonate in THFIS,as illustrated in Fig.10(a).

Although the yield of C2-C4was relatively lower than the yield of CO2,the most amount of C2-C4was released by HS,as shown in Fig.10(b).During the co-pyrolysis process it is widely believed that C2-C4originates from the cracking of longer chain alkyls at about 500°C,which is slightly lower than the cracking temperature of methyl groups at 550°C.On the contrary,the higher inert and lower alkyl chain hydrocarbon content in the THFIS resulted in lower evolution of light aliphatic compounds.The light aromatic compounds formed by the decomposition of coal macromolecules,and the carbonyl group were primarily generated by the aldehydes,esters,and ketones at 500-600°C as well.HS fraction,which contained more reactive components,generated more light aromatic compounds(Fig.10(d)).It was found that CH4was released intensively at 580°C as a result of the breakage of alkyl chains[23].As illustrated in Fig.10(c),HS had more aliphatic chains so it produced more CH4,while,on the other hand,THFIS had more heavy components and released minimal amounts of CH4.

Fig.10.Evolution of various gaseous products during individual pyrolysis of DLR fractions(a-e)and during the co-pyrolysis of HL and 10 wt%DLR(f).

As shown in Fig.10(e),the evolution of CO originally rose and then reduced when the temperature was raised.The sources of CO were as follows:at around 400 °C,CO was produced by the cracking of carbonyls;above 500 °C,CO resulted from the cleavage of oxygen from the heterocyclic ring compounds;above 650°C,the source of CO was the Boudouard reaction between CO2and C;and above 800 °C,the water-gas shift and the Boudouard reaction occurred rapidly,leading to an increase in the CO yield[17].The results in Fig.10(e)show that HS,A,and PA produced more CO.On the contrary,the higher mineral content of THFIS resulted in a weaker intensity of absorption,which illustrated that too high a mineral content in the coal matrix would mitigate the gas-solid reaction.THFIS contributed to the generation of CO in the secondary reaction stage(700-900°C).Even so,a low yield of CO was obtained at 500-600°C,since the temperature was too low to launch carbon dioxide reduction reactions.

3.6.Kinetic parameters

Fig.11.Relationship of 1/T and ln(β/T2)under the specified conversion rate(V/V*)ratios during(a)co-pyrolysis of HL and DLR and(b)HL pyrolysis alone.

To conduct the kinetic analysis of the system,the distribution activation energy model(DAEM)was employed using TG data at four different heating rates.The DAEM was appropriate for some intricate systems,covering fossil fuel pyrolysis[24-27].It was assumed that a number of first-order,irreversible parallel reactions with various continuously distributed activation energies occurred synchronously.This concept could be expressed as f(E),a distribution function.

The release of volatiles was given by the following equation:

where V and V* were the content of instantaneous volatiles and the content of effective volatiles,respectively,E was the activation energy,β was the heating rate,A was the frequency factor,and R was the perfect gas constant.

The following Eq.(5)was obtained from simplification of Eq.(4)[28]:

Most investigators suppose that the distribution function follows a Gaussian distribution with an E0(average activation energy)and σ(standard deviation).Then,with assumption of the dependency between A and temperature,it can be expressed as follows:

where the constant a and the constant b were related to pyrolysis system.

Here,to approximate the distribution function and the frequency factor,the Miura-Maki method was employed.The following equation was obtained[29]:

With the four different sets of thermogravimetric mass-loss data,the distribution function and the frequency factor were confirmed,without assuming any functions.The detailed procedure for this has been reported elsewhere[29].

Fig.11 shows the relationship between 1/T and ln(β/T2)under the chosen conversion rate.From the gradient of a fitted curve,E(activation energy)was obtained using the least square method.The parameter A was obtained from intercept according to Eq.(7).

Fig.12 demonstrates the relationship of conversion rate and activation energy as well as the corresponding distribution function.The E increased as V/V*increased.With regard to co-pyrolysis(Fig.12(a)),when V/V*was in the range of 1%-55%,the E was within the scope of 147.2-396.9 kJ·mol?1.For the individual pyrolysis of HL (Fig.12(b)),when V/V* was in the range of 1%-70%,the activation energy was in range of 289.4-445.8 kJ·mol?1.As V/V* increased,more stable macromolecules were generated by condensation of the residual volatiles in the char,which resulted in an increased E.Moreover,the distribution function approximated a double-Gaussian distribution [26]during co-pyrolysis (Fig.12(a)),and the E0and σ were 287.9 kJ·mol?1and 19.9 kJ·mol?1.For the HL individual pyrolysis(Fig.12(b)),the E0and σ were 316.7 kJ·mol?1and 1.4 kJ·mol?1.The data in Fig.13(a)confirmed that there was a linear relationship between lnA and activation energy during the co-pyrolysis.The fitting parameters for Eq.(6)were as follows:for co-pyrolysis of HL and DLR they were e2.2for a,0.13 for b;for HL individual pyrolysis they were e5.2for a,0.12 for b.

Fig.12.Relationship of V/V*vs.E and relationship of f(E)vs.E during(a)co-pyrolysis of HL and DLR and(b)HL pyrolysis alone.

Fig.13.The linear relationship between lnA and E for pyrolysis.

4.Conclusions

It was experimentally confirmed that the interactive effects between HL coal and DLR were existed during their co-pyrolysis.It was found that THFIS,contained enriched minerals and liquefaction catalysts that were conductive to devolatilization of the coal components.The organic components of DLR negatively affected copyrolysis.

The type of coal in the co-pyrolysis process was an important factor that affected the co-pyrolysis behavior.In the case of HL,the THFIS fraction favored co-pyrolysis mass loss.The higher H/C ratio of HL in comparison to SD,might make the HL more sensitive to the active components in the DLR,especially the liquefaction catalyst.During the co-pyrolysis of HL and DLR,more hydrogen free radicals from the decomposition reactions combined with the tar fragments.The minerals such as Ca and Fe were enriched in the DLR,which promoted hydrogen transfer to the coal.The pyrite increased the efficiency of hydrogenation of radicals during copyrolysis.In the case of SD,the lower H/C ratio makes the SD less sensitive to the active components in the DLR.In addition,the positive effect from the THFIS was not sufficient to offset the negative effect of the HS and PA.The Fe-based catalyst offered no remarkable effect on SD during pyrolysis.Moreover,the active components in the DLR were ‘homogenous’ with the SD,so DLR was more stable during pyrolysis for the SD.

Although the yield of tar during the co-pyrolysis of HL and DLR increased,the content of the polycyclic aromatic hydrocarbons in the co-pyrolysis tar was higher than that in individual HL pyrolysis tar.A kinetic estimation during the co-pyrolysis conducted by DAEM showed that the activation energy during co-pyrolysis was within the scope of 147.2-396.9 kJ·mol?1,when the conversion rate was in the range of 1%-55%.

Nomenclature

DLR coal direct liquefaction residue

THFS tetrahydrofuran soluble