Mechanism and Thermochemistry of Coal Char Oxidation and

Dec 27, 2016 - The present study investigates the coal char combustion by a combination of thermochemical and X-ray photoemission spectroscopy (XPS) ...
0 downloads 0 Views 2MB Size
Subscriber access provided by UNIV OF CALIFORNIA SAN DIEGO LIBRARIES

Article

MECHANISM AND THERMOCHEMISTRY OF COAL CHAR OXIDATION AND DESORPTION OF SURFACE OXIDES Gianluca Levi, Mauro Causà, Paolo Lacovig, Piero Salatino, and Osvalda Senneca Energy Fuels, Just Accepted Manuscript • DOI: 10.1021/acs.energyfuels.6b02324 • Publication Date (Web): 27 Dec 2016 Downloaded from http://pubs.acs.org on January 1, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Energy & Fuels is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 30

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

MECHANISM AND THERMOCHEMISTRY OF COAL CHAR OXIDATION AND DESORPTION OF SURFACE OXIDES Gianluca Levi‡,†, Mauro Causà‡, Paolo Lacovig§, Piero Salatino‡, Osvalda Senneca∥,*



Dipartimento di Ingegneria Chimica, dei Materiali e della Produzione Industriale, Università degli Studi di Napoli Federico II, Napoli, Italy.

§

∥Istituto

Elettra - Sincrotrone Trieste S.C.p.A., AREA Science Park, Trieste, Italy

di Ricerche sulla Combustione, Consiglio Nazionale delle Ricerche, Napoli, Italy.

(*) corresponding author: [email protected] † present address: Technical University of Denmark (DTU), Department of Chemistry, DK-2800 Kgs. Lyngby, Denmark 1

ACS Paragon Plus Environment

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Keywords: coal, char, oxidation, surface oxide, combustion

ABSTRACT

The present study investigates the coal char combustion by a combination of thermochemical and XPS analysis. Thermoanalytical methods (DTG, DSC, TPD) are used to identify the key reactive steps that occur upon oxidation and heat up of coal char (chemisorption, structural rearrangement and switch-over of surface oxides, desorption) and their energetics. XPS is used to reveal the chemical nature of the surface oxides that populate the char surface and to monitor their evolution throughout thermochemical processing.

XPS spectra show the presence on the carbon surface of three main components. It is shown that the most abundant oxygen functionality in the raw char is epoxy. It decreases with preoxidation at 300°C and even more at 500°C where carboxyl and ether oxygen functionalities prevail. The rearrangement of epoxy during preoxidation goes together with activation of the more stable and less reactive carbon sites. Results are in good agreement with semi-lumped kinetic models of carbon oxidation which include 1. formation of “metastable” surface oxides, 2. complex switchover and 3. desorption into CO and CO2.

INTRODUCTION Early studies on oxidation of solid carbon were largely stimulated by the need to rationalize broad differences in combustion rates among various coals. Early kinetic models commonly 2

ACS Paragon Plus Environment

Page 2 of 30

Page 3 of 30

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

applied until late 50s were overly simplified and based on lumped kinetics. By the late 80s broad recognition of the relevance of intraparticle transport phenomena in coal combustion was put in order and several studies focused on the role of internal porosity and on the transport of oxygen from the outer surface of a coal particle to the internal reactive surface area. The further and definitive progress was made in late 90s, when full recognition of the role of active sites, a concept borrowed from heterogeneous catalysis, and the setup of semi-lumped models to express the kinetics of coal combustion came through1. The recognition of the role of surface oxides as reaction intermediates in the combustion of coals2–4 has been extensively addressed since then, but their chemical nature and mechanistic pathways are still open to debate. Active sites may arise from the presence of inorganic matter that exerts some kind of catalytic action, if the carbonaceous material is impure, otherwise activity can arise from defects in the carbonaceous organic matter, such as defects or edges of the aromatic domains. Coals are “turbostratic” materials, where small crystallites, each made up of aromatic domains, are mixed up and cross-linked in a rather disordered fashion. The porosity of coals arises from the voids between crystallites (macropores and larger micropores) or between the lamellae of the aromatic domains (smaller micropores). The extent of coalification over geological ages or upon heat treatments modifies the disordered structure of coals towards that of graphite. Active sites of different chemical nature may coexist on the same carbon. Even with the simplest carbon structure, graphite, constituted by parallel aromatic plains/graphene layers, the attack of oxygen may take place at armchair or zig-zag positions. When analysing coal oxidation Du et al.5 proposed the existence of two different types of active sites involved respectively in oxygen chemisorption, in the early stages of reaction, and combustion, in the later stages of 3

ACS Paragon Plus Environment

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 30

reaction. The concept was expanded to account for larger populations of active sites characterized by continuous distributions of chemisorption activation energies6–8. Haynes and co-workers7,8 further expanded the consideration of the multiplicity of sites and proposed a semi-lumped reaction network embodying two types of sites:

−∗ +  → −  # 

−  # +  → −  +  ,  

−  # → −  +  ,  

−  → −∗ +  , 

(1)

According to this mechanism, chemisorption on an active carbon site results in nondissociative chemisorption of oxygen, with formation of a “metastable” surface oxide (reaction 1). The metastable oxide may further undergo complex switch-over in the presence of oxygen according to reaction 2. Alternatively, metastable oxides can be rearranged or isomerized into more energetically favourable forms along path 3. Finally, reaction 4 represents the abstraction of CO and CO2 from the oxidized carbon. Material scientists investigating oxidation of graphene by means of XPS esperiments and DFT modelling have further strengthened this kinetic mechanism and have also shown that chemisorption begins with the bonding of O atoms in bridge position over the C−C bonds, forming epoxy groups9–12. Due to the cumulative cleaving force exerted on the underlying C−C bonds, epoxy groups tend to unzip into ethers being ultimately incorporated into the C basal plane. Epoxy groups are the functionalities with the lowest thermal stability and are dominant group at the low oxidation stage, while ethers and semiquinones form prevail as oxidation proceeds. The ratio between ethers and epoxy groups determines the balance between

4

ACS Paragon Plus Environment

Page 5 of 30

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

epoxy−epoxy and epoxy−ether reactions, the latter promoting the removal of C atoms from the C backbone. As far as the thermicity of reaction is concerned, detailed thermochemical studies of the individual steps that contribute to carbon combustion are lacking. Some early work showed that oxygen chemisorption on carbon at moderate temperature is moderately exothermic (see e.g. Ismail and Walker13), while abstraction of CO and CO2 during the desorption step is considered endothermic14,15. Few researchers have performed experimental measures of the heat of desorption. Among these, Haynes and coworkers6–8 performed DTA experiments on Saran carbon after oxidation at T≥450°C. In the early stage of desorption, when mostly CO2 is released, an exotherm was recorded. However this was followed and balanced off by a pronounced endotherm in the later stages of desorption when CO was the prevailing product of desorption. Thermicity of desorption was overall endothermic but its value decreased with the temperature of the preliminary oxidation step. In a recent study Senneca et al.16 investigated the thermicity of reactions occurring during oxidation at 300°C and during temperature programmed desorption (TPD) of samples of bituminous coal. They found that oxygen chemisorption was moderately exothermic (with ∆H=-4 KJ/g), and TPD was overall exothermic (∆H=-14 kJ/g). Notably, oxygen chemisorption in this study was carried out at a temperature well below that used for oxidation tests by Haynes and coworkers. It was suggested that at such mild temperature only non-dissociative chemisorption occurs, with formation of a metastable complex. During the subsequent TPD, metastable oxides are rearranged or isomerized into more energetically favourable forms with a pronounced exotherm. This exothermicity could balance off and even overtake the endothermicity of CO and CO2 abstraction.

5

ACS Paragon Plus Environment

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The formation of metastable oxides at mild oxygen chemisorption temperature was further demonstrated by XPS analysis and discussed for different type of coal chars in a recent study by the same group17. The present study attempts to move one step further toward the characterization of the role of surface oxides in coal combustion. The range of oxidation temperatures is stretched from 300 to 500°C, and thermoanalytical methods, such as DSC and TPD techniques, are complemented by with XPS analysis. The chemical nature of the char surface oxides, their stability upon oxidation and desorption and the thermicity of the reaction steps are scrutinized.

EXPERIMENTAL SECTION Thermoanalytical experiments Experiments have been carried out on chars from two medium rank coals: South African and Colombian coal. Chars were prepared by pyrolysing batches of coal in a lab-scale fluidized bed reactor at 850 °C in nitrogen flow. The proximate and ultimate analyses of coal samples and chars obtained from them are reported in Table 1. The heating value of the chars is also reported, as calculated by the isoperibolic calorimeter Parr 2000. Char samples were ground and sieved in the size range 100-150µm prior to further processing and stored in closed vials to prevent them from absorbing further oxygen or moisture. Char samples were then oxidized with air in an electrically heated tubular reactor under either set of oxidation conditions: A (300°C for 2 hr) or B (500°C for 30min). Raw and oxidized char samples were subjected to the following set of thermoanalytical experiments: •

Temperature programmed desorption (TPD) with analysis of evolved gas (EGA)

6

ACS Paragon Plus Environment

Page 6 of 30

Page 7 of 30

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels



Temperature programmed oxidation (TPO) coupled with differential scanning calorimetry (DSC)

Both types of experiments were performed using a NETSCH 409C TG/DSC apparatus connected with a NDIR CO and CO2 analyzer (ABB AO2020 Uras 14). About 20 mg of char have been used in each test. In TPD experiments the sample was heated up to 100 °C at the heating rate of 30 °C/min in a flow of high-purity argon (150 ml/min) and held at this temperature for 5 min to release moisture. Eventually the sample was heated up to 1300°C at the rate of 30°C/min. In TPO experiments the sample was heated up to 100 °C in a stream of 21% O2 in Ar metered at the flow rate of 150 ml/min and held at this temperature for 5 min. Eventually the sample was heated at the rate of 20°C/min up to 1000°C. TPD results were worked out to obtain the profiles of CO-CO2 released as a function of time/temperature, while TPO results were worked out to calculate the derivative of the weight loss (DTG) and the heat release curves (DSC). DTG and DSC curves measured during the experiments were also integrated to calculate the overall heat of combustion.

Structural characterization by XPS XPS measurements have been performed in the ultra-high vacuum chamber (UHV) (base pressure 8∙10-11 mbar) of the SuperESCA beamline at the Elettra synchrotron radiation facility (Trieste, Italy). SuperESCA exploits the full capabilities of high resolution core-level photoemission spectroscopy (HR-XPS) by combining high resolution feature with the high flux of linearly polarised photons in the 90 to 1500 eV energy range.

7

ACS Paragon Plus Environment

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The SuperESCA experimental station consists of two chambers separated by a gate valve: a sample preparation chamber, and the main chamber, where the sample is allocated during spectra acquisition. The main chamber is equipped with a 150 mm hemispherical electron energy analyser with variable entrance slit. In the main chamber, the sample can be heated by electron bombardment and, using a PID system, its temperature can be stabilized in the range between 40 and 1300 K. This special feature of SuperESCA allows to heat up the sample in situ. This feature was exploited to perform two different set of experiments: measurements on raw and oxidized samples, and measurements of TPD-XPS on oxidized samples, in which photoemission spectra are recorded before and after controlled thermal desorption up to 850 °C. Char particles of original size 1-2 cm were cut to obtain slabs with dimensions of approximately 6×6 mm2 and thickness of 1 mm. Char slabs were oxidized by air in an electric oven at 300°C for 2hr or at 500°C for 30 min. Eventually they were carefully fixed to a Ta frame with a 6×6 mm2 front side window to expose the sample surface. The mechanical and electrical contact between the Ta foil and the coal sample was secured by two Ta clips welded to the frame. An Au foil with dimensions of approximately 3×6 mm2 was placed in electrical contact with the sample. Au was used as a reference to calibrate the binding energy position. Finally, a type K chromel-alumel thermocouple was welded to the Ta frame. The sample holder was loaded in the main chamber and alignment of the sample surface with the X-ray beam and analyzer was accomplished by means of a manual manipulator. The spot of the X-ray beam on the sample surface had a diameter of was 100 x 20 µm2. Survey spectra in the binding energy range 550-0 eV were acquired to verify the alignment and investigate possible interferences arising from inorganic matter and contaminants. The measurements were performed with the beam impinging at an angle of 37° with respect to the 8

ACS Paragon Plus Environment

Page 8 of 30

Page 9 of 30

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

normal to the sample surface and the photoelectrons were collected at 20° emission angle. Then high-resolution C 1s and O 1s core level spectra were measured at room temperature. C 1s and O 1s core level spectra were measured at photon energies of 400 and 650 eV, with energy resolutions of 80 and 150 meV. For each spectrum the binding energy position was calibrated by measuring the Fermi level position of the Au reference sample. After acquiring the XPS scans at room temperature, the samples were heated inside the SuperESCA chamber up to 850 °C at a rate of 20 °C/min under vacuum. At the end of the heating, the sample was cooled down to room temperature and survey scans, high-resolution C 1s and O 1s core level spectra were again acquired at room-temperature.

RESULTS Thermoanalytical characterization Figure 1 shows the DSC and the DTG profiles recorded during Temperature Programmed Oxidation of raw chars. The profiles of CO2, the main gaseous product of the reaction, are not reported since they closely mirror DTG curves. Two interesting results can be obtained from the analysis of the figure: 1. Two DTG (at ≅700 °C and 900 °C) and three DSC (at ≅400 °C, 800 °C and 1000 °C) peaks are observed for both coal chars. Notably, DTG peaks correspond to minima in the DSC curves. 2. For both chars a clearly detectable exotherm is recorded already at 200 °C, when the mass loss profile is flat and well before the sample starts to release CO2.

9

ACS Paragon Plus Environment

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 30

The exothermic reaction taking place below 300 °C, when no mass loss is observed, might reasonably be related to rearrangement/transformation of the oxidized complexes originally present on the sample. When, at higher temperature, desorption of CO-CO2 takes place, the correspondence with minima in the DSC curves suggests that abstraction of these species from the carbon surface is endothermic. The endothermic heat associated with abstraction of CO and CO2, however, is relatively modest when compared with the overall heat release associated with combustion. The effect of preoxidation is shown in Figure 2 for both chars. Preoxidation of the South African char at 300 °C for 2hr makes the exotherm below 300°C less evident, whereas the heat release between 350 and 900 °C is increased. Moreover, the third DSC peak at 1000°C vanishes in the preoxidized sample. Upon oxidation at 500 °C for 30 min the exotherm below 300°C is further suppressed, but the most noticeable change concerns the loss of both the second and third DSC peaks. Preoxidation of Colombian char at 300 °C for 2 hr decreases the third DSC peak which, however, does not disappear completely as in the case of South African char. Further oxidation at 500°C again results in the loss of the second and third DSC peaks at odds with the expectation that partial combustion at 500°C would reduce the first peak. Integration of the DSC curves over the entire TPO experiment allows to estimate the heat of combustion. For the SA and Colombian chars raw and pre oxidized at 300°C the heat of reaction differs from the heat of combustion measured by isoperibolic calorimetry and reported in tab.1 by less than ± 10%. On the contrary pre oxidation at 500°C results in a significant (over 60%) loss of heat of reaction compared to the raw char, both for SA and Colombian.

10

ACS Paragon Plus Environment

Page 11 of 30

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

Figure 3 shows the profiles of CO and CO2 concentrations recorded during Temperature Programmed Desorption of raw char, char preoxidized in air at 300 °C for 2 hr, char preoxidized at 500 °C for 30 min. For South African char, CO2 release starts already above 200 °C. Apparently a modest shoulder in CO2 release profiles is observed at 400 °C and a marked peak at 800 °C. A late release of CO2 occurs around 1300 °C. Release of CO starts at higher temperature with three peaks at T = 850-1100 and 1300 °C. Preoxidation of South African char at 300 °C results in an increase of the CO and CO2 concentrations throughout TPD, which keep however the same shape as for the raw char. Oxidation at 500 °C results in much more pronounced changes in both the CO and CO2 profiles, which increase consistently and peak up around 700 and 800 °C, respectively. Also for raw Colombian char the CO2 release starts above 200 °C, with two peaks at 400 and 750 °C. CO release starts at 400 °C with two major peaks at 600-800 °C and 1300 °C. Again preoxidation at 300°C affects the amount of both CO and CO2 release, but the shape of the curves is only moderately affected, while remarkable changes are observed after oxidation at 500 °C with noticeable peaks of CO and CO2 between 700 and 800 °C.

Characterization of surface oxides by XPS Figure 4 reports the O 1s and C 1s high-resolution photoemission spectra recorded at roomtemperature of: a) South African char; b) South African char oxidized at 300°C; c) South African char oxidized at 500°C; d) South African char oxidized at 300°C and then subjected to thermal desorption up to 850°C; e) South African char oxidized at 500°C and then subjected to TPD up to 850°C. Decomposition of the spectra was achieved by fitting the experimental data with Doniach−Šunjić functions convoluted with Gaussians according to the procedure outlined in 11

ACS Paragon Plus Environment

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 30

reference17. Peak assignment was made on the base of calculated values of binding energy found in the literature and also detailed in reference17. The O 1s spectrum of the raw char shows three components: the main component at 532.2 eV is associated with electrons photoemitted from epoxy oxygen10,16. The component on the high binding energy side (533.4 eV) is assigned to ether and carboxylic groups and the component on the low energy side (at 531.2 eV) to carbonyls. The C 1s spectra show a main component located at about 284.3 eV representative of sp2 carbon in C=C bonds in aromatic rings and in aliphatic chains. Other components due to sp3 and oxidized carbon are identified on the high binding energy side: at 288.2 eV related to HO-C=O, at 286.1 eV related to epoxides and C-OH, at 285 eV related to C sp3 and C-C(O) and finally a 283.7 eV to carbon vacancies. Upon oxidation at 300°C the O 1s spectrum still reveals the existence of three main components, but the epoxy component is shifted by 0.4-0.5 eV compared to the raw char. In parallel the C 1s spectrum shows a reduction of the main C sp2 component (at 284.3 eV) and of carbon vacancies (283.8 eV), balanced by the increase of carboxylic and lactone groups (288.2 eV). Results suggest that oxidation at 300°C mainly induces a structural rearrangement of epoxides towards carboxyl/lactone groups. When the South African char oxidized at 300°C is subjected to thermal desorption, we observe a sharp attenuation of the intensity of the O 1s spectra, in particular the peak of the epoxy group decays significantly, while the decrement of the ether-hydroxyl and carbonyl components is less substantial. Thermal desorption induces also a recovery of carbon vacancies and sp2 carbon . Upon oxidation at 500°C the analysis of the O 1s and C 1s spectra reveals a significant increase of the overall oxygen uptake on the carbon surface, with formation of ethers, carboxyls, carbonyls and lactones. The carbon vacancies also increase as a consequence of carbon 12

ACS Paragon Plus Environment

Page 13 of 30

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

consumption and formation of new active sites. Altogether results indicate that oxidation at 500°C promotes combustion, oxygen build up and transformation of epoxy into ether. After thermal treatment, only part of the oxygen is lost, mainly associated with epoxy functionalities. Desorption leaves the surface populated by ethers, hydroxyls, carbonyls and a smaller fraction of epoxides. Figure 5 reports the O 1s and C 1s high-resolution photoemission spectra recorded at roomtemperature of: a) Colombian char; b) Colombian char oxidized at 300°C; c) Colombian char char oxidized at 500°C; d) Colombian char oxidized at 300°C and then subjected to thermal desorption up to 850°C; e) Colombian char oxidized at 500°C and then subjected to TPD up to 850°C. Fig. 6 compares the O1s spectra of South African and Colombian chars after different treatments. Compared to the South African char, Colombian char has a distinctively larger population of epoxy already in the raw sample, and consistently lower vacancies and sp2 carbon. Further oxidation of Colombian char at 300°C and 500°C progressively reduces epoxy, so that after oxidation at 500°C ether becomes the main component of the O 1s spectrum. The effect of thermal desorption of preoxidized Colombian char is to reduce mainly the epoxy component. Char oxidized at 300°C after desorption at 850°C loses the epoxy component almost completely. This loss goes together with important changes in the C 1s spectrum which displays loss of carbon-oxygen components and sharpens up. The char oxidized at 500°C retains a larger amount of oxygen after desorption, as compared to the char oxidized at 300°C. The persistence of large amounts of oxygen after thermal desorption of char oxidized at 500°C is even more remarkable in the case of South African char. In this case also some epoxy is retained, at odds with the idea that epoxy is a metastable and labile complex and as such should be unstable above 300°C. 13

ACS Paragon Plus Environment

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 30

In order to check the stability of oxygen complexes as a function of temperature, the O 1s spectrum was acquired in the continuous mode during heat up in the SuperEsca chamber. The spectrum corresponding to the desorption temperature of about 400°C is reported in Fig. 6 for both the South African and Colombian chars. It can be observed that at 400°C a small portion of carbon oxides is lost in the case of Colombian char, and an even smaller fraction is lost in the case of South African.

DISCUSSION A first remarkable result obtained in this campaign by thermal analytical experiments is that a pronounced exotherm is recorded already at 200°C, when the DTG profile associated with mass loss of the samples is flat and well before the sample starts to release CO2. When, at higher temperature, desorption of CO and CO2 takes place, minima in the DSC curves are recorded. These two circumstances suggest that chemisorption and rearrangement of surface oxides towards more stable species might be exothermic while, as expected, abstraction of CO and CO2 is inherently endothermic. The comparison of results obtained with samples preoxidized at 300 °C and at 500 °C is helpful to provide additional mechanistic insight into the formation and role of surface oxides. The simplified schemes depicted in Fig. 6 give a conceptual frame of the thermochemistry associated with the interplay of different reaction paths included in the kinetic scheme equation (1): Case A. oxidation at low temperature followed by desorption; Case B. oxidation at moderate temperature followed by desorption.

14

ACS Paragon Plus Environment

Page 15 of 30

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

When oxidation is performed at low temperature (Case A), chemisorption is likely to occur extensively, but the energy barrier for the rearrangement of the early oxides towards more stable forms is not overcome. This is likely to be the case for chemisorption and mild oxidation up to 300°C. Upon further heating up and in the inert conditions typical of TPD experiments, reaction step 3 occurs in parallel with release of CO/CO2 (reaction step 4). The overall heat of reaction is the result of the balance between the heat associated with reaction steps 3 (exothermic) and the heat associated with reaction step 4 (endothermic). The balance results in ∆H 450°C. The framework provided by Fig. 6 should only be regarded as a broad reference. It assumes the existence of only one type of native chemisorption sites, and only one path toward site rearrangement/stabilization, at odds with the well known existence of a wide population of a sites on the surface of coal char. The XPS measures performed in the present work confirm the existence of different types of surface oxides. The O 1s spectra show in particular the existence of three main groups: epoxy oxygen, ether and carboxyl groups and finally carbonyls. It is interesting to observe that the 15

ACS Paragon Plus Environment

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 30

existence of three groups of surface oxides may be related to the occurrence of three peaks (at T≅400°C, 800°C and >1000°C) in DSC profiles of TPO experiments and in CO/CO2 profiles of TPD tests. A detailed mechanistic study would be necessary to further elucidate the relationship between the stability of different functional groups and the occurrence of distinct TPD peaks. The most abundant surface oxides in the raw char are of the epoxy type, which could in the first instance be associated to the “metastable” oxygen complexes, indicated as −  # in eq. (1) and Figure 6. It is interesting to note, though, that epoxy decreases with preoxidation at 300°C and even more at 500°C where carboxyl and ether prevail. It is noteworthy that DSC and TPD profiles of the chars after oxidation at 300°C and 500°C do not lose the first peak, which would be in the first instance associated to the more labile epoxides complexes, but instead loose the third and second peak. This finding can be explained by assuming that the rearrangement of epoxy during preoxidation goes together with activation of less reactive carbon sites which, in the absence of peroxidation, would require overcoming higher energy barriers, hence higher temperatures, to decompose into CO and CO2. Indeed, the beneficial role of preoxidation has already been highlighted in previous studies, where it was indicated as a measure to reactivate and enhance the combustion rate of carbon with intrinsic low reactivity, such as residual carbon in ash from PF boilers18,19. The proposed interpretation is in agreement with results of DFT models on epoxy configurations and stability on the hexagonal grid. Theoretical calculations in fact showed that on small graphene patches, epoxy groups form aligned rows and, due to the cumulative strain exercised on the underlying C−C bonds, tend to unzip being ultimately incorporated into the C basal plane11. This process can be roughly associated to the rearrangement reaction 3. On larger graphene patches, however, epoxy oxygens are preferentially adsorbed as second neighbours and 16

ACS Paragon Plus Environment

Page 17 of 30

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

preserve the unzipped geometry12. Ether groups residing in the graphite basal plane likely arise from the inclusion of O atoms via unzipping processes, whereas at the vacancy and edge sites result from the oxidation of undercoordinated C atoms. With increasing oxidation level other functionalities such as semiquinones and lactones form, which require the presence of unsaturated C−C bonds at the edges of the basal plane or at the periphery of vacancies. Carbon gasification during the reduction of oxidized graphite is determined by the nature and concentration of the oxidizing groups and the manifold of thermally activated surface reactions that compete with each other. If the surface is oxidized only by a low density of epoxy groups, as it happens on epitaxial graphene, the thermal reduction proceeds by releasing solely molecular oxygen around 370 K via a cycloaddition reaction from epoxy-epoxy pairs9. The evolution of the ether groups depends on whether the epoxy−ether interactions are predominant over the epoxy−epoxy recombination: they remain locked in the C grid even above 800 K unless reactive epoxides catalyze gasification channels that remove most of them from the C backbone. The formation of ether-epoxy pairs at high O coverage promotes CO/CO2 desorption, hence C gasification9,10. The release of CO and CO2, which above 300-400 °C arises from the desorption of O atoms bonded to graphitic edges or vacancy sites, can be emphasized by the presence of nearby epoxy groups. This mechanistic framework presents clear similarities with earlier studies20 where it was inferred that small pores act as oxygen “reservoirs” which promote combustion via complex spill over and oxygen surface migration.

CONCLUSIONS The formation and role of carbon-oxygen complexes during oxidation of coal chars have been characterized by a combination of thermoanalytical techniques (DTG, DSC) and XPS. The 17

ACS Paragon Plus Environment

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 30

reference frame is represented by a semilumped kinetic model based on four steps: formation of metastable surface oxides by oxygen chemisorption, rearrangement of metastable oxides into stable complexes, oxygen-driven complex switch-over, desorption of surface oxides with release of CO and CO2. The influence of oxidation temperature on complex formation is scrutinized by Temperature Programmed Oxidation and Desorption on either raw or preoxidized (at 300°C and 500°C) char samples. The chemical nature of carbon-oxygen functionalities has been clarified by analysis of XPS spectra on raw and pretreated char samples. Results of thermoanalytical experiments are consistent with the mechanistic framework assumed as a reference. Oxidation at low temperature results into the formation of “metastable” surface oxides that are stable at temperatures in the order of 300°C. Early chemisorption is moderately exothermic. At higher temperatures, “metastable” surface oxides evolve into more stable carbon-oxygen complexes either by thermally activated rearrangement/isomerization or by complex-switch-over driven by molecular oxygen. Both reactive pathways are characterized by a pronounced exotherm. As temperature increases in inert atmosphere, desorption of surface oxides and release of CO and CO2 take place, desorption being inherently endothermic. Correlating thermoanalytical and structural (XPS) characterization, suggests that the “metastable” oxides may prevailingly consist of epoxy functionalities, whereas the more stable oxides would be composed by ether/carboxyl and carbonyl moieties. Altogether, the study further substantiates the semilumped kinetic model that was assumed as the reference and provide additional insight into the role and energetics of the individual reaction steps.

ACKNOWLEDGMENTS

18

ACS Paragon Plus Environment

Page 19 of 30

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

The Authors acknowledge the support of Stefano Lizzit (Elettra) and Luciano Cortese (IRC) in the experimental campaign.

ABBREVIATIONS DTG = differential thermogravimetry DSC = differential scanning calorimetry XPS = X-ray photoemission spectroscopy DFT = density functional theory DTA = differential thermal analysis TPD = temperature programmed desorption EGA = evolved gas analysis TPO = temperature programmed oxidation UHV = ultra-high vacuum PID = proportional–integral–derivative SA = South African

REFERENCES (1)

Laurendeau, N. M. Prog. Energy Combust. Sci. 1978, 4 (4), 221–270.

19

ACS Paragon Plus Environment

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

(2)

Page 20 of 30

Wang, H.; Dlugogorski, B. Z.; Kennedy, E. M. Prog. Energy Combust. Sci. 2003, 29 (6), 487–513.

(3)

Zhuang, Q.; Kyotani, T.; Tomita, A. Energy & Fuels 1995, 9 (4), 630–634.

(4)

Pan, Z.; Yang, R. T. Ind. Eng. Chem. Res. 1992, 31 (12), 2675–2680.

(5)

Du, Z.; Sarofim, A. F.; Longwell, J. P.; Mims, C. A. Energy & Fuels 1991, 5 (1), 214– 221.

(6)

Lear, A. E.; Brown, T. C.; Haynes, B. S. Symp. Combust. 1991, 23 (1), 1191–1197.

(7)

Haynes, B. S. Combust. Flame 2001, 126 (1-2), 1421–1432.

(8)

Brown, T. C.; Lear, A. E.; Haynes, B. S. Symp. Combust. 1992, 24 (1), 1199–1206.

(9)

Larciprete, R.; Fabris, S.; Sun, T.; Lacovig, P.; Baraldi, A.; Lizzit, S. J. Am. Chem. Soc. 2011, 133 (43), 17315–17321.

(10)

Larciprete, R.; Lacovig, P.; Gardonio, S.; Baraldi, A.; Lizzit, S. J. Phys. Chem. C 2012, 116 (18), 9900–9908.

(11)

Li, J. L.; Kudin, K. N.; McAllister, M. J.; Prud’homme, R. K.; Aksay, I. A.; Car, R. Phys. Rev. Lett. 2006, 96 (17), 176101.

(12)

Yan, J.-A.; Chou, M. . Phys. Rev. B 2010, 82 (12), 21–24.

(13)

Ismail, I. M. K.; Walker, P. L. Fuel 1989, 68 (11), 1456–1460.

(14)

Frankcombe, T. J.; Bhatia, S. K.; Smith, S. C. Carbon N. Y. 2002, 40 (13), 2341–2349.

20

ACS Paragon Plus Environment

Page 21 of 30

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

(15)

Sendt, K.; Haynes, B. S. Proc. Combust. Inst. 2011, 33 (2), 1851–1858.

(16)

Senneca, O.; Salatino, P.; Cortese, L. Proc. Combust. Inst. 2013, 34 (2), 2787–2793.

(17)

Levi, G.; Senneca, O.; Causà, M.; Salatino, P.; Lacovig, P.; Lizzit, S. Carbon N. Y. 2015, 90, 181–196.

(18)

Senneca, O.; Salatino, P.; Sorrenti, R. Exp. Therm. Fluid Sci. 2012, 43, 76–81.

(19)

Gao, Y.; Kulaots, I.; Chen, X.; Aggarwal, R.; Mehta, A.; Suuberg, E. M.; Hurt, R. H. Fuel 2001, 80 (5), 765–768.

(20)

Salatino, P.; Senneca, O.; Masi, S. Carbon N. Y. 1998, 36 (4), 443–452.

Tab. 1 Proximate and ultimate analysis South African char

Colombian char

4.6

7.5

Fixed carbon

75.0

84.9

Ash

20.4

7.6

Carbon

75.4

83.9

Hydrogen

1.2

1.1

Nitrogen

1.8

1.8

LHV (kJ/g)

28

30.5

Residual matter

volatile

21

ACS Paragon Plus Environment

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

22

ACS Paragon Plus Environment

Page 22 of 30

Page 23 of 30

South African Char raw

Colombian Char raw

DSC DTG T

DSC DTG

-0.5

500

-1.0

0

20

40

60

(103°C)

(338°C)

(736°C) t (min)

80

100

(1000°C)

(1000°C)

0.0

1000

-0.5

500

-1.0

0

20

40

60

80

100

(103°C)

(338°C)

(736°C)

(1000°C)

(1000°C)

T(°C)

1000

T(°C)

0.0

Normalized DTG/DSC (exo-)

T

Normalized DTG/DSC (exo-)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

t (min)

A

B

Figure 1. Comparison of DTG and DSC during TPO experiments for South African char (A) and Colombian char (B)

23

ACS Paragon Plus Environment

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

A

Page 24 of 30

B

Figure 2. Results of TPO experiments of raw char and char oxidized at 300°C and at 500°C. A. South African; B. Colombian

24

ACS Paragon Plus Environment

Page 25 of 30

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

A

B

Figure 3. Concentration profiles of CO and CO2 (arbitrary units) released during TPD experiments of raw char and char oxidized at 300°C and at 500°C. A. South Africna , B. Colombian

25

ACS Paragon Plus Environment

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 4. O 1s and C1s XPS spectra of South African chars after different treatments 26

ACS Paragon Plus Environment

Page 26 of 30

Page 27 of 30

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

Figure 5. O1s and C1s XPS spectra of Colombian chars after different treatments 27

ACS Paragon Plus Environment

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 6. O 1s XPS spectrum of South African and Colombian chars after different treatments

28

ACS Paragon Plus Environment

Page 28 of 30

Page 29 of 30

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

A

B

Figure 7. Outline of the energetics of oxidation/desorption. A. Oxidation at low temperature followed by TPD. B. Oxidation at medium temperature followed by TPD

29

ACS Paragon Plus Environment

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

30

ACS Paragon Plus Environment

Page 30 of 30