Characterization of Oxygen-Containing Species in Methanolysis

Aug 21, 2014 - Key Laboratory of Coal Processing and Efficient Utilization, Ministry of Education, China University of Mining & Technology, Xuzhou 221...
0 downloads 0 Views 2MB Size
Article pubs.acs.org/EF

Characterization of Oxygen-Containing Species in Methanolysis Products of the Extraction Residue from Xianfeng Lignite with Negative-Ion Electrospray Ionization Fourier Transform Ion Cyclotron Resonance Mass Spectrometry Fang-Jing Liu, Xian-Yong Wei,* Rui-Lun Xie, Yu-Gao Wang, Wei-Tu Li, Zhan-Ku Li, Peng Li, and Zhi-Min Zong Key Laboratory of Coal Processing and Efficient Utilization, Ministry of Education, China University of Mining & Technology, Xuzhou 221116, Jiangsu, People’s Republic of China ABSTRACT: Methanolysis of an extraction residue (ER) from Xianfeng lignite was carried out to obtain extracts 1−4 (E1−E4). The molecular compositions (MCs) of oxygen-containing species (OCSs) in E1−E4 were characterized using negative-ion electrospray ionization Fourier transform ion cyclotron resonance mass spectrometry (ESI FT-ICR MS). In addition, solid-state 13 C nuclear magnetic resonance and X-ray photoelectron spectrometry were used to analyze the carbon types and oxygen functional groups (OFGs) in the ER. The results show that the carbon skeleton structures in the ER are dominated by aliphatic (50.2%) and aromatic (44.9%) carbons. Methylene and methoxy carbons are the most abundant among the aliphatic carbons. Each aromatic cluster contains three rings on average with two substituent groups on each ring. The OFGs in the ER include hydroxy, ether, carbonyl, and carboxyl groups, among which the hydroxy group is the most abundant. The ESI FT-ICR MS analysis shows that the molecular mass distributions of E1−E4 range from 150 to 500 u. The On (n = 1−6) class species are the predominant OCSs in E1−E4, with 0−14 double bond equivalent (DBE) values and 9−34 carbon numbers (CNs). The most abundant On class species in E1−E4 are O2, O2, O2−O3, and O3, respectively. The OCSs in E4 contain low abundances of O1 and O2 class species but relatively high abundances of O4−O6 class species. In addition, the On class species in E4 have narrower ranges of DBE values and CNs than those in E1−E3. A series of acidic species with different DBE values and CNs are assigned to alkanoic acids, alkanedioic acids, alkanetricarboxylic acids, alkylarenols, alkylarenediols, alkylarenetriols, and alkylarenoic acids. With high resolving power and mass accuracy, ESI FT-ICR MS is an effective technique for characterizing MC of the soluble portion from lignites, which will facilitate producing important chemicals from lignites.

1. INTRODUCTION Dramatically dwindling petroleum resources along with severe environmental pollution caused by coal combustion make it imperative to develop non-fuel uses for producing chemicals and carbon materials from coals. Because lignites are rich in oxygencontaining species (OCSs) and most of them are value-added chemicals or specialty chemicals, they show great potential as feedstock for producing such chemicals.1 OCSs can be separated from coal-derived liquids (CDLs) generated from pyrolysis,2 coking,3 liquefaction,4 or oxidation5 of lignites. Therefore, knowledge of the molecular composition (MC) of OCSs in CDLs is critical for the selection of refining processing and separation conditions for OCSs. However, detailed compositional characterization of OCSs in CDLs at the molecular level still faces many difficulties. To a great extent, the difficulties are attributed to the extreme complexity in MC of OCSs and the lack of appropriate separation and analytical techniques. Therefore, developing an efficient conversion process combined with advanced analytical techniques will facilitate the effective use of lignites. In comparison to conventional lignite conversion processes under severe conditions, such as high temperature and pressure, methanolysis could be more feasible to produce OCSs from lignites.6,7 Makabe et al.8−10 investigated coal methanolysis in the presence of alkali to evaluate coal structures. They found that © 2014 American Chemical Society

most of the treated coals were soluble in pyridine. Alkali, such as NaOH and KOH, could promote the methanolysis and hence increased the yield of the soluble portion (SP) from lignites.8,9,11−13 Elemental analysis, Fourier transform infrared (FTIR) spectrometry, 1H nuclear magnetic resonance (NMR), and gas chromatography/mass spectrometry (GC/MS) were usually used to characterize the SP from methanolysis.6−9,13 However, elemental analysis, FTIR, and 1H NMR only provide information on the element composition and functional groups. OCSs, such as phenols and esters, were identified with GC/MS in the SP from lignite methanolysis.6,7 However, large amounts of strongly polar and/or less volatile OCSs in the SP cannot be detected with GC/MS. In addition, it is difficult to resolve isomers of organic compounds or compounds with similar physicochemical properties by GC/MS analysis. Fourier transform ion cyclotron resonance mass spectrometry (FT-ICR MS), which has witnessed its rapid development in recent years, proved to be a powerful analytical technique for compositional characterization of petroleum-derived liquids.14−26 It possesses ultrahigh resolving power (m/Δm50% > 300 000, in which Δm50% denotes the mass spectral peak full width at Received: March 17, 2014 Revised: August 20, 2014 Published: August 21, 2014 5596

dx.doi.org/10.1021/ef501414g | Energy Fuels 2014, 28, 5596−5605

Energy & Fuels

Article

Table 1. Proximate and Ultimate Analyses (wt %) of XL and the ERa proximate analysis

ultimate analysis (daf)

sample

Mad

Ad

Vdaf

C

H

N

Odiff

St,d

H/C

XL ER

25.67

18.45

36.52

63.07 62.77

6.01 6.01

1.79 1.61

>28.73 >29.15

0.40 0.46

1.1356 1.1414

a

diff, by difference; daf, dry and ash-free basis; Mad, moisture (air-dried basis); Ad, ash (dry basis, i.e., moisture-free basis); Vdaf, volatile matter (dry and ash-free basis); and St,d, total sulfur (dry basis). The spectral width, recycle delay time, and contact time were set to 10 kHz, 0.5 s, and 1 ms, respectively. With an acquisition time of 10 ms, a total accumulation of over 12 000 transients was made and 2000 data points were collected for the spectrum. Curve fitting of the SS 13C NMR spectrum was performed using PeakFit software to obtain the distributions of different carbon types (CTs) in the ER. The OFG distributions in the ER were measured with Thermo Fisher ESCALAB 250Xi XPS, which was equipped with a monochromatized Al Kα X-ray source and operated at 150 W. XPS analysis was conducted in a fixed analyzer transmission mode, and the calibration was carried out to the main C 1 s peak at 284.8 eV. Peak fitting and semi-quantitation of XPS spectra were performed using XPS PeakFit software. 2.4. ESI FT-ICR MS Analysis. E1−E4 were analyzed with a Bruker APEX-ULTRA FT-ICR MS equipped with a 9.4 T superconducting magnet and an ESI source, which was operated in the negativeion mode. The sample was dissolved in acetone and then diluted with toluene/methanol (1:3, v/v) mixed solvent to 0.02 mg/L. The sample solution was injected through an Apollo II ESI source with a syringe pump. The voltages at the emitter, capillary column front end, and capillary column end in ESI were set to −4.0 kV, −4.5 kV, and −320 V, respectively. Ions produced in the ESI were accumulated for 0.01 s in a radio-frequency hexapole, allowed to transfer through a quadrupole mass analyzer, and then introduced into an argonfilled collision cell, in which ions were accumulated for 0.2 s before injection into the ICR cell. The data with 4 000 000 words were obtained, and a total of 64 scans were accumulated to enhance the signal-to-noise ratio. 2.5. Mass Calibration and Data Processing for ESI FT-ICR MS. Mass spectra were calibrated using sodium formate and then recalibrated with phenols and/or fatty acids, which show relatively high abundances of negative-ion ESI mass spectra peaks. The peaks for m/z 120−800, with relative abundances greater than 6 times the standard deviation of the baseline noise, were exported to an Excel spreadsheet. Data analysis was carried out using custom software, as reported elsewhere.27,31,45 In brief, the measured masses were converted from International Union of Pure and Applied Chemistry mass scale to Kendrick mass scale, and the Kendrick mass defect (KMD) was calculated to promote the identification of the homologous series.45,46 The homologous series, which share the same class (number of heteroatoms N, O, and S) and same type (DBE value) but differ by multiples of CH2 groups, have identical KMD. Then, two mass scaleexpanded segments (MSESs) within 1 u in the middle of the spectra were selected, followed by molecular formula assignments using a mass calculator program in the in-house ESI FT-ICR MS analysis software. The assigned peaks in MSESs contain at least one of the compounds for each heteroatom class species and were arbitrarily selected as references for other compounds in each class species. Each class species and corresponding isotopes with successive DBE values (with a mass interval of 2 u) and/or CNs (differing by 14 u in mass, i.e., CH2 group) could be searched within a set ±0.001 KMD tolerance.45

half-maximum peak height) and high mass accuracy (