Oxidative processes for the direct conversion of coal under mild conditions
Dissertation zur Erlangung des Grades eines Doktors der Naturwissenschaften
Vorgelegt der Fakultät für Chemie und Biochemie der Ruhr-Universität Bochum
Von Nadine Braun aus Duisburg
Bochum Februar 2016
Die vorliegende Arbeit wurde in der Zeit von September 2011 bis Januar 2016 am Max-Planck-Institut für Kohlenforschung in Mülheim an der Ruhr unter der Leitung von Herrn Dr. Roberto Rinaldi angefertigt.
Referent: Prof. Dr. Martin Muhler
Korreferent: Prof. Dr. Wolfgang Grünert
Acknowledgements
First of all, I would like to thank Dr. Roberto Rinaldi for giving me the opportunity to work on such a fascinating topic in his research group and the chance to develop myself on a professional and on a personal level.
I wish to express my gratitude to Prof. Dr. Martin Muhler for agreeing to be the first referee.
I appreciate his contribution of time, interest, valuable comments and his support during the submission of my thesis.
I would also like to thank Prof. Dr. Wolfgang Grünert for agreeing to be the second referee.
I wish to express my appreciation to Dr. Dirk Hollmann for his support with the EPR spectroscopy studies at the Leibniz-Institut für Katalyse e.V. in Rostock and also for the fruitful discussions during the project meetings.
I am deeply thankful to Dr. Christophe Farès for the technical training at the NMR laboratories and his unconditional support and all the helpful discussions over the last two years.
I would also like to thank:
Dr. Bodo Zibrowius and Wolfgang Endler at the NMR department for their help with the solid- state NMR spectrometer and measurements.
The SEM department, particularly Silvia Palm, for the nice aid to visualize the numerous coal samples.
Udo Richter, for all his support, good scientific advices and the lab organization.
André Pommerin for the introduction and advises on working with the TG analyzer.
Marc-Philipp Ruby for his help with the TG-MS measurements.
My present and past office mates and colleagues, Hebert, Jakob, JP, Gaetano, Michael,
Zhengwen, Alex, Marco, Sarah, Tobi G., Carolina N., Tobi Z., Caro G., Mario, Kameh, Kiki,
Niklas and everyone from AK Schüth and AK Rinaldi for the nice working atmosphere.
Mis queridos, Jorge, Ilton and Heitor, for the inspiring discussions about chemistry and life in general, all the nice and funny lunchbreaks, our shared addiction to GoT and the sincere friendship.
Annette Krappweis and Kirsten Kalischer for their support and engagement in administrative matters.
I gratefully acknowledge the founding by the SusChemSys cluster and the basic founding of the MPG.
Finally, I would like to thank my family and friends for all their love and encouragement. I deeply thank and dedicate this thesis to my mother for being a great source of motivation, guidance, for her patience and support during all these years.
II Content
1 Introduction ...... 1
2 State of the art ...... 5
2.1 Coal ...... 5
2.1.1 Coalification, coal rank and structure...... 6
2.2 Coal liquefaction- State of the art ...... 9
2.2.1 Direct coal liquefaction processes ...... 9
2.2.2 Indirect coal liquefaction processes ...... 12
2.3 Oxidation of coal ...... 22
2.4 Oxidations and industrial oxidants ...... 29
2.4.1 Industrial Oxidants ...... 29
2.4.2 Mechanism proposed by several authors ...... 36
2.4.3 The utilization of Group 3 metals and hydrogen peroxide as oxidant ...... 39
3+ 3 Oxidative conversion of lignite mediated by [Al(H2O)6] /H2O2 ...... 41
3.1 Introduction ...... 41
3.2 Lignite used for experiments ...... 41
3.3 Conversion of lignite at different temperatures ...... 48
3.3.1 Characterization of the product oil ...... 49
3.3.2 Characterization of the obtained coal residues ...... 63
3.3.3 Characterization of the water-soluble products ...... 69
3.4 Monitoring the evolution of the conversion of lignite ...... 72
3.4.1 Characterization of the product oil ...... 72
3.4.2 Analysis of lignite residues obtained from duration effect studies ...... 80
3.4.3 Characterization of the water-soluble products ...... 84
III 3.5 Conclusion ...... 87
4 Solvents and solvent effects on the oxidation of lignite mediated by 3+ [Al(H2O)6] /H2O2 ...... 88
4.1 Participation of reaction medium in the extent of coal oxidation ...... 90
4.1.1 Ultimate analysis of the coal residue samples ...... 90
4.2 Coal oxidation experiments in methanol-13C at 65 °C ...... 99
4.3 Hydrodeoxygenation (HDO) experiments of lignite residues obtained from reactions in methanol ...... 104
4.4 Conclusion ...... 109
5 Effect of coal rank...... 111
5.1 Analysis of obtained coal residue samples ...... 111
5.2 Conclusions ...... 120
3+ 6 Activation of hydrogen peroxide by the [Al(H2O)6] /H2O2 system - Part I: ...... 121
3+ 6.1 EPR investigations into the [Al(H2O)6] system ...... 125
3+ 6.2 EPR studies of the [Al(H2O)6] /H2O2 system in organic solvents ...... 129
6.2.1 2-Methyl-tetrahydrofuran ...... 129
6.2.2 Methanol ...... 132
6.2.3 Comparison of 2-Methyl-tetrahydrofuran and methanol as solvents used for 3+ reactions with the [Al(H2O)6] /H2O2 system ...... 134
3+ 6.3 EPR studies of the conversion of coal with Al(H2O)6] /H2O2 system in 2-Methyl- tetrahydrofuran ...... 135
6.3.1 Lignite ...... 135
6.3.2 Anthracite ...... 138
3+ 6.3.3 EPR studies of the reaction with coal and the [Al(H2O)6) /H2O2 system in methanol ...... 140
3+ 6.4 EPR studies of the [Al(H2O)6] /H2O2 system formed from aluminum perchlorate ...... 142
IV 6.4.1 EPR investigations into the activation of hydrogen peroxide by the 3+ [Al(H2O)6] /H2O2 system formed from aluminum perchlorate ...... 142
3+ 6.4.2 The [Al(H2O)6] /H2O2 system formed from aluminum perchlorate in the presence of tetrahydrofuran as the solvent ...... 145
3+ 6.4.3 The [Al(H2O)6] /H2O2 system formed from aluminum perchlorate in the presence of coal and tetrahydrofuran as the solvent ...... 147
6.5 Activation of hydrogen peroxide by other nitrates ...... 150
6.5.1 Decomposition of hydrogen peroxide mediated by different nitrates ...... 150
6.6 EPR characterization of hydrogen peroxide activation with other metals ...... 152
6.6.1 Comparison of In(NO3)3 and Ga(NO3)3 with Al(NO3)3 as an activator for
H2O2 …………………………………………………………………………………….152
6.6.2 Comparison of In(ClO4)3 and Ga(ClO4)3 with Al(ClO4)3 as an activator for H2O2 …………………………………………………………………………………….157
6.7 Conclusions EPR studies ...... 166
7 Activation of hydrogen peroxide- Part II: ...... 167
7.1 Thermodynamic parameters of the H2O2-H2O exchange process ...... 177
7.2 Activation of hydrogen peroxide- Part III: Model compound reaction ...... 186
7.3 Conclusions ...... 189
8 Summary ...... 190
9 Appendix ...... 193
9.1 Analysis of the obtained soluble products ...... 193
9.1.1 Soluble products obtained from experiments in methanol ...... 193
9.1.2 Analysis of obtained soluble products from experiments in ethanol ...... 195
9.1.3 Analysis of obtained soluble products from experiments in water ...... 196
9.2 EPR measurements ...... 197
3+ 9.3 Variable temperature NMR spectroscopy of the various [M(H2O)6] /H2O2 systems ...... 198
V 9.3.1 Complete data obtained from D-NMR experiments ...... 198
9.3.2 Variable temperature 27Al-NMR ...... 203
10 Experimental ...... 206
10.1 Chemicals ...... 206
10.1.1 Coals ...... 206
10.1.2 Solvents ...... 206
10.1.3 Inorganic chemicals ...... 207
10.1.4 Other chemicals ...... 207
10.2 Experiments ...... 208
10.2.1 Direct conversion of coal in 2-Me-THF ...... 208
10.2.2 Direct conversion of coal- Solvent effect...... 210
10.3 Hydrodeoxygenation of lignite and lignite residue samples...... 211
10.4 Activation of H2O2 ...... 211
10.4.1 Activation of H2O2 – Evaluation of different metals ...... 211
10.5 Analysis Methods ...... 211
10.5.1 Determination of water-content in product oil samples (Karl-Fischer Titration) ...... 211
10.5.2 Determination of moisture (ASTM D3173) ...... 212
10.5.3 Determination of Ash content (ASTM D3174) ...... 212
10.5.4 Determination of Volatile matter (ASTM D3175) ...... 212
10.5.5 Elemental analysis ...... 213
10.5.6 FTIR analysis ...... 213
10.5.7 NMR ...... 213
10.5.8 GPC analysis ...... 214
10.5.9 Thermal degradation analysis ...... 215
10.5.10 Determination of hydrogen peroxide (iodometric titration) ...... 215
VI 10.5.11 Determination of pH ...... 215
10.5.12 Scanning Electron Microscope (SEM) ...... 216
10.5.13 in situ EPR spectroscopy ...... 216
10.5.14 Gas chromatography ...... 217
11 Literature ...... 219
12 Publications & conference communications ...... 228
VII
Tables
Table 2.1: Coal recoverable reserves by region5 ...... 5
Table 2.2: Chemical composition, volatile and heat content of different coals9 ...... 8
Table 2.3: Comparison of different parameters and yields from direct liquefaction processes11 ...... 11
Table 2.4: Comparison of MTG Gasoline, FT products, H-coal products4, 38 ...... 21
Table 2.5: Survey of results on coal oxidation in aqueous media43 ...... 23
67-68 Table 2.6: Reactive species found in H2O2 solutions at different pH values ...... 34
Table 2.7: Reduction potentials of selected reaction species in CHP systems63 ...... 35
Table 3.1: Lignite from “Fortuna Garsdorf, 3. Sohle, tief” ...... 42
Table 3.2: Assignment of typical IR bands for coal 84a, 84f, 84g, 87 ...... 43
Table 3.3: Assignments of different chemical shift ranges in 13C NMR spectra88 ...... 45
Table 3.4: Mineral matter in lignite (unprocessed) “Fortuna Garsdorf, 3. Sohle, tief” ...... 47
Table 3.5: Yields, elemental composition of product oil obtained from reactions between 25 and 70 °C after 4 h...... 50
Table 3.6: Estimated composition ratio of the different samples of product oil obtained from reaction at different reaction temperatures for 4 h...... 59
Table 3.7: Weight loss in % observed during the thermal degradation of product oil samples and their solid residue compared to lignite and an oily product of 2-Me-THF degradation 61
Table 3.8: Yields, elemental compositions (daf) and ash content for lignite residues obtained from reactions at different reaction temperatures and for 4 h...... 66
Table 3.9: Weight of water-soluble products and elemental composition (daf) and their ash content ...... 71
Table 3.10: Weight yields of obtained product oil samples, their elemental composition (daf) from reaction with different time duration at 65 °C ...... 73
VIII Table 3.11: Possible composition ratios of obtained product oils from a reaction at 65 °C and different process duration ...... 77
Table 3.12: Weight loss in % observed during the thermal degradation of product oil samples and their solid residue compared to lignite and an oily product of 2-Me-THF degradation 78
Table 3.13: Yield and elemental composition of obtained reaction residues from experiments with different process duration at 65 °C ...... 83
Table 3.14: Weight of yield and elemental composition (daf) of obtained water-soluble products ...... 85
Table 4.1: Weight of lignite residues and their elemental composition (daf) obtained from experiments performed in varying solvent at 65 °C...... 91
Table 4.2. Comparison of weight loss in % observed for the thermal degradation of obtained lignite residues from experiment in different solvents compared to unprocessed lignite. .. 96
Table 4.3: Initial weight of substrate and lignite residues and their elemental composition obtained from processing in 13C-methanol at 65 °C for 5 min...... 100
Table 5.1: Yields and elemental composition (daf) of coal residue samples obtained from reactions with different coals, compared to the elemental composition (daf) of unprocessed coal ...... 115
Table 5.2: Weight loss [%] of different coal ranks (unprocessed and processed) obtained from TG analysis ...... 118
Table 6.1: Spin Hamilton parameters of radical species observed during the reaction. .. 128
Table 6.2: Spin Hamilton parameters of the radicals observed during the reactions ...... 138
Table 6.3: Spin Hamilton parameters of radical species observed during the reaction. .. 144
Table 6.4: Apparent first-order rate constants for the decomposition of H2O2 obtained at 55 °C ...... 151
Table 7.1: Summary of calculated exchange rates k [Hz] obtained from dynamic 1H-NMR
3+ studies of the various [M(H2O)6] /H2O2 systems at -55, -25, 0, 25, 45 and 65 °C and the observed coalescence temperature (Tc)...... 175
Table 7.2: Thermodynamic parameters of exchange process for the various
3+ [M(H2O)6] /H2O2 systems ...... 182
IX Table 7.3: Comparison of the various catalytic systems for the epoxidation of carvone at 80 °C and 3 h...... 187
Table 9.1. Weight and elemental composition (dry) of obtained soluble products at 65 °C...... 194
Table 9.2: Carbon mass balance calculation for the reactions in methanol at 65 °C for varying duration...... 194
Table 9.3: Weight and elemental composition of obtained ethanol soluble products and precipitated products at 65 °C...... 195
Table 9.4: Carbon mass balance calculation for the reactions in ethanol at different reaction duration and 65 °C...... 195
Table 9.5: Soluble products obtained from reactions in water at different reaction duration and 65 °C...... 196
Table 9.6: Carbon mass balance calculation for the reactions in water at different reaction duration and 65 °C...... 196
Table 9.7: Arrhenius plot parameters obtained from D-NMR experiments with the
Al(ClO4)3/H2O2 system ...... 198
Table 9.8: Arrhenius plot parameters obtained from D-NMR experiments with the
Ga(ClO4)3/H2O2 system ...... 199
Table 9.9: Arrhenius plot parameters obtained from D-NMR experiments with the
In(ClO4)3/H2O2 system ...... 200
Table 9.10: Arrhenius plot parameters obtained from D-NMR experiments with the
Al(NO3)3/H2O2 system ...... 201
Table 9.11: Arrhenius plot parameters obtained from D-NMR experiments with the
Al(NO3)3/H2O2 system ...... 202
X
Figures
Figure 2.1: Different ranks of coal.4 ...... 6
Figure 2.2: Main options for coal-to-liquids processes...... 9
Figure 2.3: Possible feedstock for syngas production and its following main options for the synthesis of liquid fuels...... 13
Figure 2.4: MTG reaction pathway43 ...... 19
Figure 2.5: EMRE MTG Process flow diagram adapted from the ExxonMobile MTG brochure41 ...... 20
Figure 2.6: Metal-oxygen species73 ...... 30
Figure 2.7: AMOCO commercial process for para-xylene to terephthalic acid74 ...... 31
Figure 2.8: Various ways for the activation of hydrogen peroxide.80 ...... 32
Figure 3.1: FTIR analysis of lignite from "Fortuna Garsdorf, 3. Sohle, tief" ...... 43
Figure 3.2: 13C CP MAS NMR of lignite (“Fortuna Garsdorf”)...... 44
Figure 3.3: a) SEM picture of lignite (unprocessed), 500x zoom; Figure 3.3b (right): SEM picture of lignite (unprocessed), 1.5k zoom ...... 46
3+ Figure 3.4: Scheme for direct liquefaction of coal mediated by [Al(H2O)6] ...... 48
Figure 3.5: FTIR analysis of product oil samples obtained from reactions at different temperatures after 4 h...... 51
Figure 3.6: 13C NMR of product oil samples obtained from reactions at different temperatures after 4 h...... 52
Figure 3.7: HSQC NMR spectrum of product obtained from an experiment with lignite in d- THF at 65 °C and 4 h...... 53
Figure 3.8: GPC analysis of product oil samples obtained from reaction at different temperatures after 4 h compared to the GPC analysis of the oily product from 2-Me-THF degradation (65 °C, 4 h). Detection of the analysis performed at a wavelength of 220 nm...... 55
XI Figure 3.9: a) TG and DTG curve of lignite; b) TG and DTG curve for the product oil obtained from lignite after a reaction at 65 °C and 4 h; c) TG and DTG curve for the oily product
3+ obtained from a reaction without lignite (2-Me-THF + Al /H2O2) at 65 °C and 4 h...... 57
Figure 3.10: Comparison of experimental and calculated TGA and DTG curves of oil obtained from a reaction with 2-Me-THF mixed with lignite (50:50 wt%), and a coal liquid obtained after a reaction at 65 °C for 4h...... 60
Figure 3.11: a) observed TGA curves for product oils and b) their corresponding DTG curves obtained at reaction temperatures between 70 and 40 °C for 4 h, compared the oily degradation product of 2-Me-THF...... 62
Figure 3.12: SEM pictures of unprocessed lignite (a) and reactions residuals obtained from reactions at b) 25 °C, c) 40 °C, d) 45 °C, e) 50 °C, f) 55 °C, g) 60 °C, h) 65 °C and i) 70 °C...... 64
Figure 3.13: 13C CP MAS NMR of lignite residues obtained from reactions at different reaction temperatures after 4 h...... 68
Figure 3.14: FTIR spectra of water-soluble products obtained from reactions at different temperatures...... 70
Figure 3.15: FTIR spectra of obtained product oil samples from reactions at 65 °C and different reaction times compared to the spectra of the oily product obtained from 2-MeTHF degradation reaction (65 °C, 4 h)...... 74
Figure 3.16: 13C SS-NMR spectra from obtained product oil samples at 65 °C and different reaction duration...... 75
Figure 3.17: GPC analysis of obtained product oil samples from reactions at 65 °C and different process duration compared to oily product obtained from 2-Me-THF degradation reaction (65 °C, 4 h)...... 76
Figure 3.18: a) TGA curves of product oil samples obtained from a reaction at 65 °C and different reaction times compared to the oily product from 2-Me-THF degradation (65 °C, 4 h); ...... 79
Figure 3.19: SEM pictures of lignite (a) and lignite residues obtained from experiments at different process duration: b) 30 min, c) 60 min, d) 90 min, e) 120 min, f) 150 min, g) 180 min, h) 240 min ...... 81
XII Figure 3.20: 13C CP MAS NMR of obtained lignite residues from reactions at 65 °C...... 84
Figure 3.21: FTIR spectra for the water-soluble products compared to the FTIR spectrum of aluminum nitrate...... 86
3+ Figure 4.1: FTIR spectra of lignite residue obtained after treatment with H2O2 or Al /H2O2 at different reaction duration at 65 °C in different solvents: a) methanol, b) ethanol, c) water, d) 2-Me-THF...... 93
Figure 4.2: 13C CP-MAS NMR of lignite residues obtained from experiments in methanol and different process duration at 65 °C compared to the residues from corresponding blank reaction (without catalyst), and unprocessed lignite...... 94
Figure 4.3: TG (a) and DTG (b) analysis of lignite residues obtained from reactions in
3+ methanol mediated by the Al /H2O2 system at 65 °C and various process durations (5-30 min)...... 97
Figure 4.4: TG (a) and DTG (b) analysis of lignite residues obtained from reactions in ethanol
3+ mediated by the Al /H2O2 system at 65 °C and various process durations (5-30 min). .... 97
Figure 4.5: TG (a) and DTG (b) analysis of lignite residues obtained from reactions in water
3+ mediated by the Al /H2O2 system at 65 °C and various process durations (5-30 min). .... 98
Figure 4.6: TG (a) and DTG (b) analysis of lignite residues obtained from reactions in 2-Me-
3+ THF mediated by the Al /H2O2 system at 65 °C and various process durations (5-30 min)...... 98
13 Figure 4.7: FTIR spectra of lignite residue obtained from experiments in CH3OH at 65 °C and 5 min reaction duration...... 101
13 13 Figure 4.8: C-NMR of lignite residue samples obtained from experiments in CH3OH at 65 °C and 5 min reaction duration. The peaks at about 0 and 100 ppm (indicated by asterisks) are spinning side-bands...... 102
Figure 4.9: TG-MS analysis of lignite compared to lignite residues obtained from reactions at 65 °C; data normalized by sample weight...... 103
Figure 4.10: Products obtained from blank experiment with unprocessed lignite and without catalyst. Reaction condition: 0.5 g lignite residue, 5 mL n-Octane, 300 °C, 100 bar H2, 20 h. Internal standard (ISTD): 8x10-5 mol n-hexadecane ...... 105
XIII Figure 4.11: Products obtained from blank experiment with processed lignite (Al/H2O2, MeOH, 65 °C, 5 min) and without catalyst. Reaction conditions: 0.5 g lignite residue, 5 mL
-5 n-octane, 300 °C, 100 bar H2, 20 h. Internal standard (ISTD): 8x10 mol n-hexadecane 106
3+ Figure 4.12: Products obtained from unprocessed lignite (a) and residue samples (Al /H2O2, MeOH, 65 °C): b) 5 min and c) 30 min process duration. Reaction conditions: 0.5 g lignite residue, 0.2 g catalyst, 5 mL n-octane, 300 °C, 100 bar H2, 20 h. Internal standard (ISTD): 8x10-5 mol n-hexadecane...... 108
Figure 5.1: SEM pictures of a) Gasflammkohle and b) its reaction residue ...... 112
Figure 5.2: SEM pictures of a) Fettkohle and b) its reaction residue ...... 112
Figure 5.3: SEM pictures of a) Magerkohle and b) its reaction residue ...... 113
Figure 5.4: SEM pictures of a) Anthracite and b) its reaction residue...... 113
Figure 5.5: 13C CP-MAS NMR spectra for coal residue samples (1), and unprocessed coals (2): ...... 116
Figure 5.6: TGA (a) and DTG (b) curves of coal residue samples of different coal ranks
3+ obtained from reaction with the [Al(H2O)6] /H2O2 at 65 °C and 4 h...... 119
Figure 5.7: TGA (a) and DTG (b) curves of unprocessed coal of different ranks...... 119
3+ Figure 6.1: Proposed mechanism of the [Al(H2O)6] catalyzed epoxidation of α,β- unsaturated ketones by Rinaldi et al.100 ...... 122
Figure 6.2: a) EPR spin trapping of superoxide and hydroxyl radical with DMPO and the decay from DMPO/·OOH product to DMPO/·OH; b) EPR spin trapping of superoxide with DEPMPO...... 123
Figure 6.3: EPR spectrum of formed DMPO-OH product from a control experiment performed on a H2O2 (70 wt%) solution in H2O...... 124
Figure 6.4: a) EPR-spectra of the DMPO-OH adduct obtained from the experiment with 1 mmol Al(NO3)3∙9H2O dissolved in an aqueous solution of H2O2 (7 wt%) at room temperature; b) Comparison of EPR data obtained by experiment and from computer simulation for DMPO-OH adduct...... 125
XIV Figure 6.5: EPR-spectra of the DEPMPO-OH adduct obtained from the experiment with 1 mmol Al(NO3)3∙9H2O dissolved in an aqueous solution of H2O2 (7 wt%, 1 mL) at room temperature...... 126
Figure 6.6: EPR spectra of DMPO-OH and DMPO-OOH adducts obtained from an
experiment with 70 wt% H2O2 (1 mL) and 1 mmol Al(NO3)3∙9H2O at room temperature. . 127
Figure 6.7: a) EPR spectrum of DMPO-OH and DMPO-OOH products at 1 min, and corresponding simulated EPR spectra of b) DMPO-OH product and c) DMPO-OOH product...... 128
Figure 6.8: a) Experimental EPR spectrum of DMPO-OOH150 and DMPO-2-Me-THF adducts
acquired from an experiment with 70 wt% H2O2 (0.1 mL) and 0.1mmol Al(NO3)3∙9H2O in 0.8 mL 2-MeTHF at room temperature; b) proposed mechanism for the formation of the 2-Me- THF radical and its addition to DMPO...... 130
Figure 6.9: EPR spectra obtained from an experiment with 0.1 mL H2O2 (7 wt%), 0.1 mmol
Al(NO3)3 9H2O in 0.8 mL 2-Me-THF at 55 °C...... 131
Figure 6.10: Simulated EPR spectra for different DMPO species observed during the reactions...... 132
Figure 6.11: EPR spectra obtained from an experiment with 0.1 mL H2O2 (7 wt%), 0.1 mmol
Al(NO3)3∙9H2O in 0.8 mL methanol at 55 °C...... 133
Figure 6.12: Comparison of EPR spectra obtained from reaction experiment with 0.1 mL
H2O2 (7 wt%), 0.1 mmol Al(NO3)3∙9H2O in 0.8 mL a) methanol and b) 2-Me-THF at 55 °C...... 134
Figure 6.13: Quartz cell with a) dispersion of lignite in 2-Me-THF before reaction and b) after
3+ addition of [Al(H2O)6] /H2O2 and reaction at 55 °C for 20 min...... 135
Figure 6.14: EPR spectra for the reaction with lignite in 2-Me-THF mediated by the
3+ [Al(H2O)6] /H2O2 system at 55 °C for 20 min...... 136
Figure 6.15: Simulation of EPR spectra obtained under reaction conditions with five different DMPO species...... 137
Figure 6.16: EPR spectra obtained from a reaction with anthracite in 2-Me-THF mediated by
3+ the [Al(H2O)6] /H2O2 system at 55 °C for 10 min; Radical species trapped by DMPO. ... 139
XV Figure 6.17: EPR spectra obtained from experiments performed with lignite in methanol with a) 1 mmol Al(NO3)3·9 H2O and with 70 wt% H2O2 and b) 0.1 mmol Al(NO3)3·9 H2O and 7 wt% H2O2 at 55 °C. Radical species trapped by DMPO...... 141
3+ Figure 6.18: EPR spectra obtained from the [Al(H2O)6] /H2O2 system (1 mmol Al(ClO4)3, 1 mL 7 wt% H2O2) at 55 °C for 20 min; Radical species trapped by DMPO. (Straight lines: measurement out of tune) ...... 143
3+ Figure 6.19: EPR spectra extracted from Figure 6.18 from [Al(H2O)6] /H2O2 system (1 mmol
Al(ClO4)3, 1 mL 7 wt% H2O2) at 55 °C for 20 min; Radical species trapped by DMPO. ... 144
3+ Figure 6.20: EPR spectra obtained from the [Al(H2O)6] /H2O2 system (0.1 mmol Al(ClO4)3,
0.1 mL 7 wt% H2O2) in 0.8 mL THF at 55 °C for 20 min; Radical species trapped by DMPO...... 146
3+ Figure 6.21: EPR spectra extracted from Figure 6.20 from the [Al(H2O)6] /H2O2 system
(0.1 mmol Al(ClO4)3, 0.1 mL 7 wt% H2O2) in 0.8 mL THF at 55 °C for 20 min; Radical species trapped by DMPO...... 146
3+ Figure 6.22: EPR spectra obtained from the [Al(H2O)6] /H2O2 system (0.1 mmol Al(ClO4)3,
0.1 mL 7 wt% H2O2) in 0.2 mL THF and 0.2 g lignite at 55 °C for 10 min; Radical species trapped by DMPO...... 148
3+ Figure 6.23: Extracted EPR spectra obtained from the [Al(H2O)6] /H2O2 system (0.1 mmol
Al(ClO4)3, 0.1 mL 7 wt% H2O2) in 0.2 mL THF and 0.2 g lignite at 55 °C for 10 min, b) EPR spectra extracted from Figure 6.22 ; Radical species trapped by DMPO...... 148
3+ Figure 6.24: a) EPR spectra obtained from the [Al(H2O)6] /H2O2 system (0.1 mmol Al(ClO4)3,
0.1 mL 7 wt% H2O2) in 0.8 mL THF and 0.2 g anthracite at 55 °C for 10 min, b) extracted spectrum from Figure 6.24a after the injection of DMPO spin trap. Radical species trapped by DMPO...... 149
Figure 6.25: a) Decomposition profiles of H2O2 catalyzed by different nitrates; b) first order fit for the decomposition of H2O2 catalyzed by different nitrates; Reaction conditions: H2O2 (70 wt%, 2 mL, 48 mmol), salt (2 mmol), 55 °C...... 151
Figure 6.26: a) EPR spectra from the reaction of H2O2 (70 wt %) and In(NO3)3 with DMPO spin trap; b) single spectrum obtained after 1 min reaction time from the kinetic study;
Reaction condition: 1 mmol salt dissolved in 1 mL H2O2, room temperature...... 153
XVI Figure 6.27: EPR spectra from the reaction of H2O2 and Ga(NO3)3 with DMPO spin trap; b) single spectrum obtained after 1 min reaction time from the kinetic study; data smoothed by
Savitzki-Golay method; Reaction condition: 1 mmol salt dissolved in 1 mL H2O2, room temperature...... 154
Figure 6.28: EPR spectra of DEPMPO radical adducts formed in experiments with a)
Ga(NO3)3, b) In(NO3)3, c) Al(NO3)3 (1 mmol) dissolved in H2O2 (7 wt%) and d) H2O2 (7 wt%) without addition of catalyst at room temperature...... 154
Figure 6.29: Comparison of EPR profiles of DMPO-trapping products obtained from reactions with a) Ga(NO3)3, b) In(NO3)3 and c) Al(NO3)3 (1 mmol)) dissolved in H2O2 (7 wt%, 1 mL) at 55 °C...... 156
Figure 6.30: Comparison of EPR spectra of DMPO radical adducts acquired from experiments with H2O2 (7 wt %, 1 mL) and a) Ga(NO3)3, b) In(NO3)3 and c) Al(NO3)3 (1 mmol) in methanol at 55 °C...... 157
3+ Figure 6.31: EPR spectra obtained from the kinetic study from a) [Al(H2O)6] /H2O2 system
3+ (1 mmol Al(ClO4)3, 1 mL 7 wt% H2O2); b) [Ga(H2O)6] /H2O2 system (1 mmol Ga(ClO4)3, 1
3+ mL 7 wt% H2O2), and c) [In(H2O)6] /H2O2 system (1 mmol In(ClO4)3, 1 mL 7 wt% H2O2) for 20 min at 20 °C; Radical species trapped by DMPO...... 159
Figure 6.32: Extracted EPR spectrum obtained from the kinetic study with NaClO4/H2O2 system (1 mmol NaClO4, 1 mL 7 wt% H2O2) at 55 °C for 20 min; Radical species trapped by DMPO...... 160
3+ Figure 6.33: EPR spectra obtained from the kinetic study from a) [Ga(H2O)6] /H2O2 system
3+ (1 mmol Ga(ClO4)3, 1 mL 7 wt% H2O2); b) [In(H2O)6] /H2O2 system (1 mmol In(ClO4)3, 1 mL
7 wt% H2O2) at 55 °C for 20 min; Radical species trapped by DMPO...... 162
Figure 6.34: Extracted EPR spectrum obtained from the kinetic study with NaClO4/H2O2 system (1 mmol NaClO4, 1 mL 7 wt% H2O2) in 0.8 mL THF at 55 °C for 20 min; Radical species trapped by DMPO...... 163
3+ Figure 6.35: EPR spectra obtained from the kinetic study from a) [Ga(H2O)6] /H2O2 system
3+ (1 mmol Ga(ClO4)3, 1 mL 7 wt% H2O2) in 0.8 mL THF; b) [In(H2O)6] /H2O2 system (1 mmol
In(ClO4)3, 1 mL 7 wt% H2O2) in 0.8 mL THF at 55 °C for 20 min; Radical species trapped by DMPO...... 165
XVII 1 Figure 7.1: H-NMR spectra of H2O2 (0.14 mol/L) and NaClO4/H2O2 (0.003/0.14 mol/L) in
THF-d8 at room temperature...... 168
1 3+ Figure 7.2: H-NMR of [Al(H2O)6] /H2O2 (0.003 mol/L Al(ClO4)3∙9H2O/ 0.14 mol/L H2O2) in THF-d8 at -55 °C...... 169
1 3+ Figure 7.3: Variable temperature H-NMR of [Al(H2O)6] /H2O2 (0.003 mol/L Al(ClO4)3∙9H2O/
0.14 mol/L H2O2) in THF-d8...... 170
1 3+ Figure 7.4: Variable temperature H-NMR of [Ga(H2O)6] /H2O2 (0.003 mol/L
Ga(ClO4)3∙9H2O/ 0.14 mol/L H2O2) in THF-d8...... 171
1 3+ Figure 7.5: Variable temperature H-NMR of [In(H2O)6] /H2O2 (0.003 mol/L In(ClO4)3∙9H2O/
0.14mol/L H2O2) in THF-d8...... 172
1 3+ Figure 7.6: Variable temperature H-NMR of [Al(H2O)6] /H2O2 (0.003 mol/L Al(NO3)3∙9H2O/
0.14 mol/L H2O2) in THF-d8...... 173
1 3+ Figure 7.7: Variable temperature H-NMR of [Al(H2O)6] /H2O2 (0.003 mol/L AlCl3*9H2O/
0.14 mol/L H2O2) in THF-d8...... 174
3+ Figure 7.8: 2D COESY spectra of the various [M(H2O)6] /H2O2 systems recorded at -55 °C...... 176
Figure 7.9: Arrhenius plots of ln(k) against 1/T obtained from d-NMR data of the various
3+ [M(H2O)6] /H2O2 systems; a) Al(ClO4)3, b) Ga(ClO4)3, c) In(ClO4)3, d) Al(NO3)3, e) AlCl3...... 178
Figure 7.10: “Hofmeister series” for anions170 ...... 184
Figure 9.1: EPR spectra obtained from the [Al(H2O)6/H2O2 system ( 1 mmol Al(ClO4)3, 1 mL
7 wt% H2O2) at room temperature. Radical species trapped with DMPO...... 197
Figure 9.2: Arrhenius plots for the Al(ClO4)3/H2O2 system...... 198
Figure 9.3: Arrhenius plots for the Ga(ClO4)3/H2O2 system ...... 199
Figure 9.4: Arrhenius plots for the In(ClO4)3/H2O2 system ...... 200
27 Figure 9.5: Variable temperature Al-NMR spectra obtained for the Al(ClO4)3/H2O2 (0.003/ 0.14 mol/L) system...... 203
27 Figure 9.6: Variable temperature Al-NMR spectra obtained for the Al(NO3)3/H2O2 (0.003/ 0.14 mol/L) system...... 204
XVIII 27 Figure 9.7: Variable temperature Al-NMR spectra obtained for the AlCl3/H2O2 (0.003/ 0.14 mol/L) system...... 205
Figure 10.1: Homemade setup for in situ EPR measurements ...... 216
XIX
List of Abbreviation
2-Me-THF 2-Methyltetrahydrofuran
2-PrOH 2-Propanol
DMPO 5,5-Dimethyl-1-Pyrroline-N-Oxide
DEPMPO 5-Diethoxyphosphoryl-5-methyl-1-pyrroline-N-oxide
EA Elemental analysis
EPR Electron paramagnetic resonance
EtOH Ethanol
FID Flame ionization detector
FTIR Fourier transform infrared spectroscopy
GC Gas chromatography
GCxGC-qMS Two-dimensional gas chromatography coupled with quadrupole mass spectrometry
GPC Gel-permeation chromatography
HDO Hydrodeoxygenation
ISTD Internal standard
MeOH Methanol
MNPA 4-Methyl-4-nitro-pentanoic acid
MS Mass spectrometry
NMR Nuclear magnetic resonance
Temp. Temperature
TGA Thermogravimetric analysis
THF Tetrahydrofuran
XX
Introduction
1 Introduction
Since 2003, the rise in world oil prices has renewed the interest in producing liquid fuels from unconventional resources, such as biomass, shale oil, and coal. Coal seems to offer one of the greatest promises because of its production potential and commercial readiness.1
It is the most abundant fossil fuel and well distributed worldwide.2 The overall recoverable reserves of coal are estimated at around one trillion tons which will last for approximately another 112 years.1, 2 In Europe, Germany is the biggest producer of lignite (brown coal) in
Europe with total deposits around 80 billion tons and a coal output with ca. 170 Mio tons per year.3
In general, there are two routes for coal-to-liquids (CTL) plants, (1) an indirect pathway, e.g.
Fischer-Tropsch process, or (2) direct liquefaction of coal, e.g. the Bergius process. In the indirect process, coal is gasified into hydrogen and carbon monoxide (syngas) and afterwards converted into liquids fuels. In the direct liquefaction process, coal is converted into synthetic fuels by hydrogenation. Normally, high severity conditions (> 400 °C, > 350
4 bar H2) are needed in direct liquefaction processes to achieve high oil yields, increasing process costs.
In this work, a novel oxidative approach for pretreatment of low-ranking coals under mild conditions is described. Lignite, a soft coal with low organic maturity was processed and liquefied into oily and solid products upon subjecting it to a treatment with hydrogen peroxide
(H2O2) and aluminum salt (catalyst) in a cyclic ether solvent (e.g. 2-Methyl-THF). Under
1 Introduction
atmospheric pressure and at temperatures as high as 50 °C, high yields were obtained.
Analysis of the obtained products showed that lignite structure is depolymerized and insertion of the solvent to the new formed coal fragments occurs in the reaction. In addition, further solvents (methanol, ethanol, water) and their effect on the conversion of lignite were examined. Due to the high solubility of the aluminum salt in those solvents, the focus of this study was to examine the conversion of lignite by analyzing the lignite residues obtained after reaction and not on the soluble products. Furthermore, the effect of the
3+ [Al(H2O)6] /H2O2 system was tested with coals of different maturity as well.
To examine the reaction mechanism, the activation and decomposition of hydrogen peroxide under different reaction conditions was studied. EPR studies of the radicals formed showed
3+ that the [Al(H2O)6] /H2O2 system strongly initiates the generation of hydroxyl and perhydroxyl radicals and show that the solvents act as trapping agents towards the formed radicals which influences the conversion of coal dramatically.
In Chapter 2, an overview of the fossil fuel coal, coal ranks, and its economic importance is provided. Furthermore, the state-of-the-art of different direct and indirect processes for coal- to-liquid plants is discussed and compared. Finally, a survey of published results on coal oxidation using different oxidants and catalysts is given.
In Chapter 3, the direct conversion of lignite mediated by an aluminum salt dissolved in hydrogen peroxide is introduced. In the first part of this chapter, the effect of temperatures on the conversion of lignite (a low-rank coal) is assessed in order to optimize the reaction to achieve high yields of product oil from coal. To lend insight into the complex processes
3+ occurring in the conversion of lignite in the presence of Al /H2O2, the product oil samples were characterized by elemental analysis (EA), nuclear magnetic resonance (NMR), Fourier
2 Introduction
transform infrared (FTIR) spectroscopy and gel permeation chromatography (GPC).
Furthermore, additional details on the coal degradation were obtained from coal residues and water-soluble products.
In the second part of Chapter 3, an analysis of the effect of process duration on the yield of product oil is given. Again, the product oil samples, coal residues and water-soluble products were characterized by the aforementioned techniques providing more information concerning the nature of the product oil.
3+ In Chapter 4, the effect of solvent on lignite conversion in the presence of the Al /H2O2 system is examined. Surprisingly, when replacing 2-Me-THF or THF by other solvents, lignite is converted into soluble products that are solid instead of oil. Due to the high amount of remaining aluminum salt in the obtained soluble products, the focus of this study was on the chemical and physical structure of the obtained coal residues which can be further upgraded into various compounds (e.g. phenols) by hydrodeoxygenation (HDO) process.
3+ In Chapter 5, the effect of the Al /H2O2 system was investigated on higher coal ranks and its ability to break down the more condensed structures. Since this study revealed that the obtained product oil samples consist of mostly 2-Me-THF degradation products the focus is again on the difference in chemical and physical structure of the obtained coal residues.
In Chapter 6, to gain further information on the reaction mechanism, the formation of radicals during the reaction and their effect on/within the solvent and coal was studied by electron paramagnetic resonance (EPR) spectroscopy. Further, the activation of H2O2 by different metals was studied and compared. This study reveals that the group 13 metals (aluminum, gallium and indium) are capable to activate and decompose H2O2. Their mechanism on the formation of radicals was examined by EPR spectroscopy, as well.
3 Introduction
In Chapter 7, further investigation into the reaction mechanism were performed by temperature-variable 1H proton NMR to study the interaction between hydrogen peroxide
3+ 3+ and various [M(H2O)6] systems (M = Al, Ga, In) and their influence on the proton- exchange which allows the formation of radical species. In addition, the D-NMR studies will also reveal that the salt’s anion has a tremendous effect on the proton-exchange as well.
4 State of the art
2 State of the art
2.1 Coal
Coal is the most abundant fossil fuel available on earth. Its global recoverable reserves are estimated at one trillion tons according to the World Energy Council.5 This figure corresponds to three times the current world’s recoverable reserves of petroleum.5
Compared to gas or oil occurrence, coal is better distributed across the globe. Table 2.1 lists the eight regions with the largest reported proven recoverable reserves of coal. On top of the list is Europe with the total coal reserves of about 30.8%. Within Europe, Russia (74.5 gigatonnes of oil equivalent, Gtoe) and Germany (19.2 Gtoe) are the countries with the largest recoverable coal reserves5.
Table 2.1: Coal recoverable reserves by region5
Region Percentage of global reserves Europe 30.8 North America 27.5 East Asia 13.2 South & Central Asia 11.2 Southest Asia & Pacific 12 Africa 3.5 Latin America & The Caribbean 1.6 Middle East & North Africa 0.1
5 State of the art
Globally, most of the produced coal is locally used for electric power generation.6
Nonetheless, the international coal trade represents about 15% of the global production, and is ruled by the energy demand from Taiwan, Japan and South Korea.1
2.1.1 Coalification, coal rank and structure
Coal is not just coal, but the altered remains of prehistoric vegetation that originally accumulated in swamps and peat bogs.7 As coal matures, different ranks of coal occur in
Earth. This coalification has an important meaning for the physical and chemical properties of coal. Figure 2.1 shows the ranking of coal according to its carbon (C-) content.
Figure 2.1: Different ranks of coal.4
6 State of the art
Originally, peat is transformed into lignite, a coal which is of low organic maturity.
Compared with other coals, lignite is quite soft and its color can range from dark black to various shades of brown. When peat and lignite are subjected to the effects of temperature and pressure underneath the earth for geological times, these organic materials undergo further chemical and physical changes. Upon maturation of the organic matter, the content of oxygen decreases and the structure progressively becomes more aromatic. High ranks of coal show a high carbon content, and, therefore, possess a high heating value. High rank coals show a high carbon content, and, therefore, possess a high heating value.
Coal is a network heteropolymer that consists mainly of organic material. It is a large, complex molecule with mostly cross-linked aromatic ring structure. In coal a varying amount of inorganic and organically bonded sulfur (1 to 6%), nitrogen (1 to 2%) and oxygen can be found. Sulfur and nitrogen are mainly found in heterocyclic moieties of coal.4, 8 Raw coal also contains moisture and solid particles of mineral matter. Most of the high ranking coals contain ca. 10 wt% mineral matter. Coal is almost non-volatile, insoluble and a non- crystalline solid.4
Table 2.2 shows the chemical compositions, volatile and heat content of different coal ranks and the classifications by ASTM (USA) and DIN (Germany). It summarizes the changes in elemental composition of coal upon maturation. The carbon content of the dry-and-ash-free
(daf) coal increases from lignite (65 wt%) up to anthracite (91 wt%), whereas the hydrogen content decreases (from 8 to 3 wt %), respectively.
7 State of the art
Table 2.2: Chemical composition, volatile and heat content of different coals9, 10
USA German Carbon Hydrogen Oxygen Volatiles Heat Classification content Classification [%] [%] [%] [%] (ASTM) MJ/kg (DIN)
Lignite Braunkohle 65 – 75 8 – 5.5 30 – 12 60 – 43 7 - 13
Subbituminous Flammkohle 75 – 81 6.6 – 5.8 > 9.8 45 – 40 < 32 coal
High volatile Gasflammkohle 81 – 85 5,8 – 5,6 9.8 – 7.3 40 – 35 33 – 34.2 Bitum. coal Gaskohle 85 – 87.5 5.6 – 5.0 7.3 – 4.5 35 – 28 33.9 – 34.8
Medium Fettkohle 87.5 – 5.0 – 4,5 4.5 – 3.2 28 - 19 34.5 – volatile bitum. 89.5 35.6 coal
Low volatile Esskohle 89.5 – 4.5 – 4.0 3.2 – 2.8 19 -14 35.2 – Bitum. coal 90.5 35.6
Semi- Magerkohle 90.5 – 4.0 – 3.75 2.8 – 2.5 14 – 12 35.2 – Anthracite 91.5 35.5
Anthracite Anthrazit > 91.5 < 3.75 < 2.5 < 12 32.8
8 State of the art
2.2 Coal liquefaction- State of the art
In this section, the various coal-to-liquid (CTL) process will be presented. The focus is on the differences between direct and indirect liquefaction processes and their development and improvement over the last century.1, 11 The different advantanges and disadvantages will be discussed as well. Figure 2.2 shows the main option for CTL processes based on coal as feedstock, which will be discussed in the following sections.
Figure 2.2: Main options for coal-to-liquids processes.
2.2.1 Direct coal liquefaction processes
The history of coal liquefaction and coal chemistry was summarized by Speight.12
The technology for coal liquefaction exists since the early 20th century. In 1913, Friedrich
Bergius developed a process to directly liquefy coal into synthetic fuels by hydrogenation of high-volatile bituminous coal. This process requires high temperatures and pressures for high yields of oils and fuels to be achieved.1, 13 Finely ground and dried coal is mixed with
9 State of the art
heavy oil recycled from the process and usually subjected to iron oxide (catalyst) at 400 to
500 °C under 20 to 70 MPa hydrogen pressure.13, 14 The obtained products are heavy oils, middle oils, gasoline and gases.15 Next, the immediate product from the reactor is catalytically hydrotreated. Naphthenes and aromatics, in addition to low yields of paraffins and olefins are obtained.
From the original Bergius process, several variants of direct coal liquefaction process have been developed. These processes can be classified into two main categories: single- stage and two-stage processes. A comparison of process conditions and yields (oil, gas, residuals) for different processes are shown in Table 3. In the single-stage processes, liquid fuels are obtained by using one reactor or more which are operating under the same conditions. Unfortunately, these processes were unable to make coal liquefaction economically attractive, the single-stage processes were only used from the middle 1960s to the beginning of the 1980s. The typical catalysts are iron-based ones, cobalt-molybdenum
(Co-Mo) and nickel-molybdenum (Ni-Mo) supported on aluminum oxide. Notably, liquefaction in absence of a catalyst is also possible, such as the Exxon Donor Solvent
Process.16 By this technology, dried and crushed coal is fed into a reactor along with a hydrogen donor solvent under hydrogen pressure (17.5 MPa) operating at temperatures between 425-470 °C. The products are separated by a series of distillation steps. High yields of low-sulfur liquids are obtained.
Another example for a one-step process is the H-Coal process.17 A preheated slurry of crushed coal and recycle oil is introduced into the bottom of an ebullated-bed reactor.
Depending on the desired products, e.g. crude oil or low-sulfur residual oils, the operating
17, 18 conditions can be altered (435-465 °C, 20 MPa H2). Because of the ebullated-bed, it is
10 State of the art
possible to use a conventional solid hydrotreating catalyst (e.g. Co-Mo/Al2O3) with a slurry feed.
Table 2.3: Comparison of different parameters and yields from direct liquefaction processes12 Process Kohleoel H-Coal EDS BCL CTSL LSE Ruhrkohle, HRI, Exxon, NEDO, DOE, HTI, British Coal Germany USA USA Japan USA Corp., UK Process type One-stage Two-stage
Temperature 1. stage: 1st stage: 1. stage: 427 – 410 – 430 °C 435 – 425 – 448 °C, 400 °C, 470 °C 2. stage: 465 °C 470 °C 2. stage: 2nd stage: 435 – 360 – 400 – 400 °C 400°C 440 °C
Pressure 150 – 200 10 – 20 300 atm 200 atm 175 atm 170 atm atm atm; 200 atm
1. stage: Fe, Reactor/ 1. stage: Co- 2. stage: jet- catalyst No catalyst Fe Mo/Al2O3, No Fe-slurry, stirred 2. stage: (red slurry) Ni- catalyst Ca-Ni-Mo reactor, 3-phase- Mo/Al2O3 Ni- fluidised- Mo/Al2O3 bed-reactor Yields (%)
Gases (C1- C4) 19 10,5 5 13.6 --- 15.4
Light oils 25.3 13.3 14 36.7 60 – 65 49.9 (C5 – 200 °C) Heavy oils 32.6 19.7 10 15.1 --- 12.4 (200 - 325 °C) Residuals 22.1 37.1 48 8.1 7 - 22 24.7
Characteristics Added Lignite First catalyst process commercial during (moisture plant reaction ca. 60 %) EDS = Exxon Donor Solvent, BCL = Brown Coal Liquefaction, CTSL= Catalytic Two-stage Liquefaction, LSE= Liquid Solvent Extraction
11 State of the art
The principles involved in the H-Coal process were further developed and effectively incorporated into the Catalytic Two-Stage Liquefaction (CTSL) by the US DOE and HTI.4
The reason why two-stage processes constitute the process option most explored nowadays is that optimum conditions for the dissolution of coal (1st stage) and upgrading of the coal liquids (2nd stage) are different in severity. Compared with a single-stage process, a two- stage process leads to increased selectivity to distillates with low heteroatom content (Figure
2.2).18 In general, the composition of the coal-derived liquids depends strongly on the character of the used coal, the process conditions and the degree of hydrogen addition.4, 12
2.2.2 Indirect coal liquefaction processes
Indirect coal liquefaction (ICL) has enormous potential applications.19 Over the last two decades, a large amount of papers have been published on ICL, including e.g. new research on Fischer-Tropsch (FT) synthesis, syngas to methanol, and methanol to olefins processes. ICL processes mostly consist of two main steps. The first is the gasification of the feedstock into hydrogen and carbon monoxide (syngas). In the second step, the syngas is converted into liquid fuels. Syngas can be produced from many carbonaceous sources, such as coal, biomass, natural gas, or almost any hydrocarbon feedstock by processing it with steam or oxygen. Figure 2.3 shows the general steps from feedstock (fossil or renewable) to syngas to fuels. Syngas is a crucial intermediate product for the production of hydrogen, ammonia, methanol and fuels in general. This section will discuss some of the processes mentioned in Figure 2.3.
12 State of the art
Figure 2.3: Possible feedstock for syngas production and its following main options for the synthesis of liquid fuels.
2.2.2.1 Lurgi (-Sasol) gasifiction process
In Germany before World War II, the Lurgi process was developed for large-scale commercial production of synthetic natural gas. It was later improved by Sasol which is has been used at the Sasol commercial plant in Africa. In the Lurgi-Sasol process, crushed (6 –
50 mm) noncaking coal is given in a pressurized vertical furnace. Inside the furnace, a moving bed of coal, steam and either air or oxygen react. The coal is gasified at temperatures of 425-600 °C with a residence time of 1 h.12, 20
The hydrogen is supplied by injected steam and the heating is produced by the combustion of product char. At the bottom of the gasifier, a revolving grate supports the bed of coal, removes the ash and allows steam and oxygen to be introduced. Afterwards, the obtained syngas is cleaned to remove contaminants, e.g. particles, sulfur, ammonia, and further processed to adjust the H2/CO ratio by using the water-gas shift reaction. The purified syngas can then be further converted into methanol, multiple products and synthetic fuels by using e.g. Fischer-Tropsch synthesis.20
13 State of the art
2.2.2.2 Fischer-Tropsch Synthesis
Indirect liquefaction process was invented by Franz Fischer and Hans Tropsch in 1925.21
The coal feed is gasified into synthesis gas which is converted to liquid hydrocarbons by cobalt or iron catalysts. Depending on the desired products, two different approaches are possible, a low temperature process (200-240 °C) or a high temperature process (300-
350 °C). With the low temperature process, linear waxes of high-quality are the main products. They can be further processed to high-quality Diesel fuel. In general, the products from FT synthesis have many outstanding and advantageous properties. FT products are sulfur and nitrogen free and comprise only a few aromatic components. Overall these properties make FT fuels environmentally friendly, since they lead to low emissions of SOx and NOx in addition to suppression of soot formation.
Only the metals iron (Fe), Nickel (Ni), Cobalt (Co) and Ruthenium (Ru) own the required activity for commercial FT applications, with iron and nickel being the cheapest metals.11 However, since Ni produces too much methane and Ru is too expensive for large- scale applications, only Fe and Co constitute the practical choice as catalysts.21, 22 Over the years, different types of reactors have been developed for FT process, such as fixed-bed
(FBR), slurry bubble column and fluidized-bed reactor.23 The best option for producing high yields of gasoline is a high capacity fixed fluidizized bed (FFB) reactor, which operates at temperatures of about 340 °C in presence of iron catalysts. The products are straight run gasoline (40%) and propene and butene as FT product (20%). Propene and butene can be further oligomerized to gasoline.24 Due to the highly branched oligomers, products with high octane numbers are obtained, while the products from the straight run gasoline have low octane numbers because of their mainly linear structure and low aromatic content. Typically,
14 State of the art
the C5/C6 cuts of straight run gasoline need to be further hydrogenated and isomerized to increase the octane value.
For the large-scale production of linear olefins, high temperature fluidized bed FT reactors with iron catalysts proved to be the best methodology. The typical olefin content of
25 the C3, C5-C12 and C13-C18 cuts are about 85%, 70%, and 60%, respectively. The obtained cuts can be further converted into longer chain olefins, and therefore, sold as petrochemicals at higher prices than fuels.
The selectivity of FT products depends on the process parameters, such as temperature, pressure, catalyst type and promoters.21, 22 In general, an increase in process temperature shifts the selectivity toward lower carbon number products and more methane, while simultaneously the number of branched products increases.21 These shifts agree with thermodynamic expectations and the relative stability of the products. Since Co catalysts are more active for hydrogenation, the methane selectivity increases faster with higher temperatures than it does with Fe catalysts.
Additionally, promoters also play a key role in changing the selectivity.26 For instance, alkali metals have an electronic effect and may change the electronic properties of Fe-based catalysts and modifying the adsorption pattern of reactants (H2 and CO), on the active sites, which often increases the 1-alkene selectivity, reaction rate, growth of hydrocarbon chains.
Moreover, the addition of external water could enhance the CO conversion and cause an increase in the average carbon number in the chains as well as the 1-alkene selectivity.27, 28
Furthermore, the impact of Group I alkali metal modifiers on iron catalysts through different process conditions with respect to CO conversion has been studied by many research groups. Zhang et al. showed that potassium promoter severely suppresses the hydrogen
15 State of the art
adsorption, but apparently simplifies the CO adsorption. As the carburization is improved, the formation of methane is suppressed while the selectivity to hydrocarbons increased.29
Group I alkali metal promoters have less effect on Co catalysts, compared to Fe- counterparts for Co-catalysts. Noble metals, transition metals and rare earth oxides are mostly used as promoters. Studies showed that the use of proper promoters enhance the adsorption of CO as well as the chain growth.30, 31 De la Osa et al. reported that small additions of calcium oxide as promoter to cobalt oxide catalyst improved its reducibility, and led to increased CO conversion, to C5+ products. Furthermore, the addition of calcium
32 shifted the distribution to mainly C16-C18 hydrocarbon fractions.
However, Fe-based catalysts seem to be more suitable due to their low costs and the ability to increase the production of olefins, paraffins and oxygenates on a large-scale over a wide range of temperature. Additionally, high temperature FT synthesis with Fe- based catalysts allows using hydrogen-deficient syngas (low H2/CO ratio), produced from coal gasification.33 However, the addition of alkali metals also has some disadvantages. Due to their lower melting points and high mobility, alkali metals could easily be removed from the catalyst surface, which results in poor stability for long-term use FT synthesis.34 Studies showed that alkaline-earth metals have similar properties to alkali metals while their oxides show more similarities with silica and alumina, e.g. higher melting points. It was found that alkaline-earth metals could strengthen Fe-O bonds of α-Fe2O3 in fresh catalysts, suppressing H2 adsorption while CO adsorption was simultaneously enhanced. Although the alkaline-earth metals had no apparent influence on the Co conversion of used catalysts, the formation of methane could be suppressed. Moreover, the selectivity to olefin and higher molecular weight products was increased.35
16 State of the art
Investigations on copper (Cu) as a promoter for Fe-based catalysts revealed an improved rate of catalysts activation and shorter induction period. Furthermore, Cu as a promoter shifts the product distribution to heavy hydrocarbons and the olefin/paraffin ratio is enhanced due to indirect enhancement of basicity of the surface.36
Aforementioned, cobalt-based catalysts are only used in low-temperature FT (LTFT) process because at high temperatures a large amount of methane is produced. Due to the higher price of Co, it is preferred to reduce its amount used in the process, but maximize the available surface area on the metal. Therefore, Co-supported Al2O3, SiO2 or TiO2 catalysts
37 are used. The selectivity to C5+-products was found to be dependent on the particle size.
A decrease in Co-particle (from 6 to 2.6 nm) reduced the selectivity to C5+-products (from 85 to 51 wt%).
To avoid catalysts deactivation, especially by oxidation, two-stage operations, combined with gas recycling are needed in order to achieve high conversion (>90%). Also, catalyst poisoning (by sulfur), sintering, coking, fouling, and attrition are often causes of catalyst deactivation. The first three causes are due to chemical failure, while fouling and attrition are due to mechanical failure.38 Sulfur is known to be the main poison for FT- reactions, although it does not affect the product selectivity obtained by Fe-based catalysts.
Co-based catalysts show more resistance to oxidation, but are more sensitive to sulfur.
17 State of the art
2.2.2.3 Methanol Synthesis
Methanol is obtained by synthesis of carbon monoxide/hydrogen mixtures in presence of a catalytic system. At present, methanol from syngas is carried out on CuO-ZnO-Al2O3 catalysts. The syngas-to-methanol synthesis is described by the chemical equations:
CO + 2 H2 CH3OH
CO2 + 3 H2 CH3OH + H2O
The conversion is thermodynamically favored at relatively low temperatures and under high pressures. To produce methanol from coal, temperatures about 30 –375 °C and pressures about 27.5 – 36 MPa are used.39 Although this process is generally highly selective, the obtained raw methanol needs to be purified by distillation to removes side products (e.g. dimethyl ether and higher molecular weight alcohols).
Methanol itself can be used as a motor fuel (e.g. racing fuel), but its direct use as fuel finds some problems.40 Methanol is hygroscopic and contains only half the energy content compared to gasoline. Therefore, the further conversion of methanol, for instance into gasoline, is a more desirable approach.
2.2.2.4 Methanol-to-gasoline (MTG)41
Methanol can be further converted into gasoline by catalytic processes.40-42 For instance, Methanol-to-Gasoline (MTG) process, developed by ExxonMobil Research and
Engineering (EMRE) Company, produces clean gasoline of high quality from coal by coupling coal gasification and methanol synthesis technologies. MTG process was
18 State of the art
developed in the 1970’s and commercialized in the mid-1980’s in New Zealand. According to Exxon Mobile, the obtained gasoline is fully compatible with conventional refinery gasoline. While the FT synthesis produces a wide-range of linear hydrocarbons that requires upgrading to produce a commercial gasoline, MTG process selectively converts methanol into a high quality liquid fuel.41, 43 Figure 2.4 shows reaction path way for the MTG process in detail.
Figure 2.4: MTG reaction pathway43
The commercialized MTG plant is a fixed-bed process in which the reaction is engineered by splitting the conversion into two parts. Figure 2.5 shows a simple flow diagram of the EMRE MTG process. First, methanol is converted into an equilibrium mixture of methanol, dimethyl ether (DME) and water. Within this step, 15-20% of the overall heat of reaction is released. In the second step, the mixture is mixed with recycle gas and passed over specially designed ZSM-5 catalyst to produce hydrocarbons (mostly gasoline range) and water.
The inlet temperatures of the conversion reactor are individually controlled by adjusting the flow of reactor effluent passing through the recycle gas/reactor effluent heat exchangers, and by regulating the temperature difference across the exchangers. The reactor effluent is
19 State of the art
used as well in order to preheat, vaporize and overheat the methanol feed to the DME reactor. Afterwards, the reactor effluent is further cooled to 25 – 35 °C and brought to the product separator in which gas, liquid hydrocarbon, and water are separated. The gas phase
(light hydrocarbons) is returned to the recycle gas compressor. The obtained aqueous phase can be treated afterwards or recycled within the CTL complex. The obtained liquid product
(raw gasoline) contains mostly gasoline boiling range material and dissolved H2, CO2 and light hydrocarbons (C1-C4). Through distillation, the non-hydrocarbons, C1, C2, C3 and partly
41 C4 hydrocarbons are removed to obtain gasoline within the required specifications.
Additionally, produced methane, ethane and some propane are removed in the de- ethanizer. The obtained liquid product is given to a stabilizer in which propane and partly butane are overhead to fuel gas. At the end, the stabilized gasoline is splitted into light and heavy gasoline fractions.
Figure 2.5: EMRE MTG Process flow diagram adapted from the ExxonMobile MTG brochure41
20 State of the art
Table 2.4 compares the obtained product distributions from low-temperature (LT) and high- temperature (HT) Fischer-Tropsch synthesis, with the products obtained from MTG process and from direct liquefaction (H-coal TM). As aforementioned, a broad spectrum of products are obtained for the FT process, where for the LTFT the main products are heavy oil and waxes, and for the HTFT C5+ selectivity about 36% is obtained. The direct liquefaction process (H coal) produces no C1-C4 products but almost equal amounts of C5+ and
Distillates. The MTG process produces primarily C5+ products (82.3%) and only minor amounts of light hydrocarbons. It is noteworthy, that the liquid products obtained from FT and H Coal processes require upgrading steps to obtain transportation fuels.
Table 2.4: Comparison of MTG Gasoline, FT products, H-coal products4, 41 LT-FT HT-FT H Coal TM MTG Co catalysts Fe catalyst (DCL process) Methane 5 8 0.7 Ethylene 0 4 - Ethane 1 3 0.4 No C1-C4 yields Propylene 2 11 0.2 reported Propane 1 2 4.3 Butylenes 2 9 1.1 Butane 1 1 10.9 C5 – 160C 19 36 36.5 82.3 Distillates 22 16 43.2 - Heavy oil/wax 46 5 0.3 0.1 Water Sol. Oxyg. 1 5 0.3 0.1 Total 100 100 100 100
21 State of the art
2.3 Oxidation of coal
The transformation of coal in soluble products has been studied by many scientists over the last 240 years. Oxidation has been one of the more widely used techniques for the degradation of coal into smaller fragments. The principal types of oxidation are those performed in acidic and basic media. Commonly used oxidants are HNO3, HNO3-K2Cr2O7,
KMNO4, O2, H2O2 and O3, which converted most of the aromatic compounds of coal into benzene-carboxylic acids, like the strong oxidizing agent nitric acid (HNO3). With HNO3 as an oxidant, side-reactions occur, leading to excessive amounts of nitrogen in the isolated products. Practically, this kind of oxidation is not very valuable regarding coal valorization to chemicals, but provides information about the coal structure, especially its aromaticity and degree of substitution as the obtained number of carboxylic acid groups per ring correlates with the degree of substitution or condensation.8 Carboxylic acid groups are also obtained by reaction with alkaline potassium permanganate (KMnO4). As oxidation proceeds, the obtained products show an increasing in acid content and become soluble in aqueous potassium hydroxide. This solution can be flocculated with acids, which leads to a rusty colored precipitation. The precipitate shows an oxygen-content of 35-40%. Then extensive oxidation is performed, soluble acids are formed. During the oxidation process, the color of the solution changes from44, 45 brown to yellowish-green and completely disappears altogether when about 90% of the original carbon content has been oxidized.44, 45, 46, 47 Table
2.5 summarizes most important investigations on coal oxidation in aqueous media. Some of them will be explained in more detail.
22 State of the art
Table 2.5: Survey of results on coal oxidation in aqueous media48-55 Medium Oxidant Reaction Products Most important conditions investigators
Alkaline O2 270 °C (water-)soluble Fischer and 60 bar products (carboxylic Schrader (1920) acids), mostly Howard and co- benzene carboxylic workers (1939-55) acids, CO2 Montgomery and Holly (1956-58)
KMnO4 100 °C “humic” acid as Bone and co- 1 bar intermdiate workers (1926-37) crystalline Ward and Lawson benzenoid acids, (1947-74) oxalic and acetic acids
NaOCl 50 – 60 °C CO2, benzene Chakrabartty et al. 1 bar carboxylic acids (1972-74)
Acid HNO3 100 °C Aliphatic acids (e.g. Fuchs et al. (1928- acetic, oxalic, 40) succinic), benzene Jüttner et(1935-65) carboxylic acids, Hayatsu et al. (1978) picric acid Deno et al. (1981)
K2Cr2O7 100 °C acetic acids, Kinney (1947) methane Honda and Hirose (1958) Peroxy-acids Soluble “sub-humic Schnitzer and Acetic acids + H2O acids”; Skinner (1975) 80 % of methanol- Hayatsu et al. (1978) soluble acids (benzene-, toluene-, furan-, methylfuran-, carboxylic acids) Performic acids 50 °C Benzene derivates Raj (1976) Bimer et al. (1977- 78) Trifluoro acetic acid + acetic, succinic and Deno et al. (1977) H2O2 glutaric acids and methanol 2+ H2O2 + Fe 25 °C Untreated: Heard, I., (Fenton reaction) 308 h Small amount of Senftle, F.E. (1983) pH= 0-11.5 oxidation (pH< 1.5); Untreated & Demineralized: demineralized Oxidation even in anthracite basic solutions
H2O2 + acids* 23-100 °C Reduced pyritic Boron, D., (coal desulfurization) 1-3 h sulfur content Taylor, S.R. (1984) (with all acids); *acids= H2SO4, Reduced organic H3PO4, HOAc sulfur content (only with H3PO4)
H2O2 60 °C Carboxylic acids Miura et al. (1997) 2 h
23 State of the art
Table 2.5 Continued Medium Oxidant Reaction Products Most important conditions investigators 2+ Acid H2O2 + Fe (from 50 °C Maleic and oxalic Sugano, M. et al. mineral matter) 6 h acids, 9-phenan- (2002) 0.1 Mpa N2 threnol, aromatic (Coal liquefaction –COOH, small fatty residue) acids, small carboxylic acids (>50%), CO2
H2O2 60 °C Aliphatics, aryl Yujiao et al. (2012) (unprocessed coal 4 h ketones, quinones; and coal, CS2/N- extracted coal is methyl-2-pyrolidinone richer in: phenols, extracted) alcohols, ethers, esters, carboxylic acids, anhydrides
H2O2 20-40 °C CS2 extracts: Thamasebi et al. 4 h aliphatic (2015) Stepwise hydocarbons; oxidation and Ethanol & acetone solvent extracts: extraction Mostly oxygen- containing compounds, aliphatic hydrocarbons Neutral Electrochemical Ambient “ulmic acids” Lynch and Chollet (1932) Belcher (1948)
O3 Ambient Water-soluble acids Kinney (1950-52)
O3 ambient Increased content Patrakov et al. in: (2006) Ether, quinone (low/medium-rank), Carboxylic, phenolic groups (high-rank coal) Photochem. Ambient Benzene carboxylic Hayatsu et al. (1975- O2 + UV acids (lignite); 78) Malonic, succinic, glutaminic acids (bit. Coal)
O2 30-180 °C CO, CO2, H2O Clemens et al. (main compounds); (1990) Peroxides, aldehydes, carboxyls, esters
Na2Cr2O7 250 °C Mostly aromatic Hayatsu et al. (1975- (autoclave) acids 78) Stephens et al. (1985)
24 State of the art
In 1920, Fischer and Schrader showed that coal suspended in an alkali solution can be oxidized with oxygen at 200–250 °C. This reaction produces soluble products (e.g. benzene carboxylic acids).56 Almost 20 years later, Howard et al. took the investigation a bit further.57 They subject coal to alkaline aqueous solutions under pressure of oxygen (6 MPa) at 270 °C. They found that nearly 50% of coal was converted into carbon dioxide. The residual compounds were determined to be mostly water-soluble aromatic acids of relatively low molecular weight (less than 450 Da), with one-third of the compounds being identified as benzene carboxylic acids.45, 58, 59 From 1927-1937, Bone and co-workers probably did the most important quantitative investigation of permanganate oxidation in alkaline solution of coal.60 Dried and pulverized coal was suspended in 1 % potassium hydroxide (KOH) solution to which KMnO4 was added and heated at 70 °C until boiling point, when the typical permanganate color disappeared. After the filtration and acidification, the obtained products were analyzed. Due to step-wise oxidation, it was found that “humic” acids are formed as an intermediate product which is oxidized further to crystalline benzenoid acids, oxalic acid and acetic acid, as well.60-62
In the period of 1925 to 1956, coal oxidation with nitric acid provided important information on the coal structure. As already mentioned, side reactions and nitration is a negative side-effect of this type of coal oxidation. Nevertheless, it was shown that simple aliphatic acids (e.g. acetic, oxalic, succinic, etc.) are produced besides benzene carboxylic acids from coal oxidation.44, 45, 47
The studies from Kinney et al. on coal oxidation with a mixture of nitric acid and potassium dichromate (K2Cr2O7) lead to the interesting conclusion that a correlation between the yield of acetic acid and the methane amount produced by low-temperature carbonization. It is obvious that structures in coal yielding acetic acids due to oxidation are
25 State of the art
mostly decomposed by treating coal at higher temperatures (e.g. 500 °C). It was also found that high rank coals render low yields of acetic acid.63, 64
Ray and Bimer et al. investigated the oxidation of coal with performic acid. They showed that this method results in several benzene derivate products. Unfortunately, the gained information are not very consistent, and the mechanism still needed to be elucidated at that time.65, 66
Deno et al. studied the oxidation effect of a peroxy-trifluoro-acetic acid mixture
(CF3COOH + H2O2 (30 wt%)). It was shown that this combination of oxidants is specific for aromatic compounds. Almost 70% of aliphatic compounds have been preserved in the form of acetic, succinic and glutaric acids and methanol (from –OCH3). Moreover, aromatic molecules (e.g. anthracene, phenanthrene, pyrene, chrysene) do not produce benzene carboxylic acids under those reaction conditions.47
Over the last 30 years, many studies using hydrogen peroxide as oxidizing agent were performed. In 1983, Heard and Senftle studied the effect of H2O2 (30 wt%) solutions in varying pH ranges (from 0 to 11.5) using the comprised mineral matter, primarily pyrite
52 (FeS2), to form hydroxyl radicals (Fenton reaction). It was found that the oxidation of untreated anthracite was minimal and only occurred at pH values below 1.5, whereas demineralized anthracite was more effectively oxidized, even in alkaline solutions. However, since all iron materials were removed from anthracite’s surface, the oxidation mechanism must have been different. In this case, the decomposition of H2O2 is explained by activated carbon sites on coal’s surface. One year later, Boron and Taylor showed that desulfurization of coal is possible by mild oxidation with H2O2 at temperatures up to 100 °C and reaction
51 duration of 1-3 h and in addition of an acid (H2SO4, H3PO4, HOAc). They found that in
26 State of the art
presence of all acids the pyritic sulfur content was significantly reduced, whereas the organic sulfur content was only diminished in the presence of phosphoric acid.
In 2012, Yujiao and co-workers processed untreated coal and samples obtained from
50 coal extraction (CS2/NMP) with a solution of H2O2 at 60 °C and for 4 h. The results showed that the extracted samples are more reactive with H2O2, leading to higher yields of oxygen- containing compounds, e.g. phenols, alcohols, esters, ethers, carboxylic acids and anhydrides compared to the untreated coal.
Electrochemical studies revealed that the use of a copper anode leads only to formation of “ulmic acid” (alkaline extractable, alcohol soluble fraction of humic acids), which cannot be further oxidized electrochemically. In contrast, platinum and lead anodes lead to a fast oxidation and degradation of coal to carbon dioxide.67 Several years later, the formation of “ulmic acid” was confirmed by Belcher. He also found that a small amount of water-soluble compounds are produce. The oxidation products formed on lead and platinum anodes are completely different than those formed on a copper anode.68
In 1916, Fischer converted about 92% of coal into dark-brownish water-soluble acids by ozonization of an aqueous suspension of coal.56 Almost four decades later, Kinney and co-workers took further investigations using ozone as oxidant into coal oxidation and further oxidation of humic acids derived from coal. For the ozonization of humic acids, they reported that 65% of the carbon content was converted into carbon dioxide and oxalic acid or dicarboxylic acid. The residual carbon was found to be ozone-resistent acids, most likely benzene carboxylic acids. Additionally, they confirmed Fischer’s results for ozonization of coal. Again, dark-brownish water-soluble acids have been obtained, which contained small amounts of acetic acid and some traces of oxalic acids.63, 69
27 State of the art
Further studies on the oxidative degradation of coal were carried out by Hayatsu et al. in the 1970s.70 They investigated the effect of ultraviolet light at ambient temperature on different coal ranks. No degradation of the anthracites’ structure was observed, whereas lignite and bituminous coals were oxidized about 25–30%, respectively. Products derived from photochemical degradation consisted mostly of benzene carboxylic acids. High volatile bituminous coals yielded not only in a wider range of aromatic acids, but also malonic, succinic, glutaminic acids and small amount of saturated fatty acids.46, 71
The oxidation with sodium dichromate (Na2Cr2O7) seems to be the most specific oxidation process for coals. Na2Cr2O7 almost mineralize the aliphatic components of coal structure and leaves the aromatics more or less intact. For 100 g of an Illinois bituminous
46 coal, Hayatsu found ca. 53 g aromatic acids after coal oxidation by Na2Cr2O7.
28 State of the art
2.4 Oxidations and industrial oxidants
2.4.1 Industrial Oxidants
Oxidations continue to play an important role in organic chemistry, especially for the production of basic chemicals derived by the conversion of oil- or gas-based feedstocks. In general, oxidizing agents, such as permanganate or dichromate are used, but their use is associated with formation of large amounts of waste. Nowadays, the environmental regulation demands cleaner technologies, which involves, for instance, the use of molecular oxygen as the primary oxidant. This section provides some insight into the various oxidants used in industrial processes, regarding their transferable oxygen content and their environmental aspects.
2.4.1.1 Nitric acid
Nitric acid has been industrially used as an oxidizing agent for (aromatic) hydrocarbons and alcohols for the production of carboxylic acids (e.g. adipic acid). The effectiveness of nitric acid is possibly caused by two factors. The first is the homolytic scission of the nitric acid, producing free radicals which act as a reaction initiator (Eq.2.1).72
HONO2 HO• + •NO2 (Eq. 2.1) RH + initiator R• (Eq. 2.2)
29 State of the art
2.4.1.2 Oxygen
The ground state of dioxygen (O2) is a triplet state with two unpaired electrons with parallel spins. Such an electronic structure forbids the direct reaction of O2 singlet and organic
3 molecules. For oxidation purposes, the triplet-oxygen ( O2) needs to be converted into a
1 singlet-oxygen ( O2). One way to bypass the activation energy barrier involves a free radical pathway, for instance by using paramagnetic (transition) metal ions which afford superoxometal complexes. Subsequently, inter- or intramolecular electron-transfer processes can lead to the formation of various metal-oxygen species which can contribute to the oxidation of organic substrates.73
Figure 2.6: Metal-oxygen species73
An example for metal-induced oxidation is the reaction of para-xylene to terephthalic acid using a homogeneous catalytic oxidation system of cobalt, manganese and bromine.
Herein, Bromine from hydrobromic acid (HBr) is used as a promoter for the generation of radicals. Acetic acid and oxygen/air are used as solvent and oxidant, respectively. This process is commercially known as the AMOCO process. The oxidation operates at 175-
225 °C and 1.5-3.0 Mpa oxygen pressure. Yields about 95% are obtained in reaction duration of 8-24 h (Figure 2.7).
30 State of the art
Figure 2.7: AMOCO commercial process for para-xylene to terephthalic acid74
2.4.1.3 Hydrogen peroxide
Hydrogen peroxide is a bulk chemical annually produced on a megaton scale. It is available in aqueous solutions (3–85 wt% H2O2) for a variety of technical and industrial applications
(e.g. pulp bleaching, water treatment). H2O2 is often referred to as a “green oxidant” because
75 it generates water as a reaction byproduct. H2O2 may generate different types of free radicals and other reactive species that are able to oxidize or decompose organic substrates. The reactive species formed by the reaction of H2O2 and a catalyst or activation can act either as oxidants or as reductants. Over the years, catalytic oxidations using H2O2 systems have proven particular useful in environmental applications due to their ability to degrade organic contaminants and thus detoxify wastewater. Despite the many industrial applications of H2O2, detailed understanding about the several reactive species formed by its decomposition is still missing. In the early 1990’s, the use of H2O2 for environmental applications, e.g. soil, remediation drove the research of new catalysts for H2O2 activation.
Several research groups started to explore the role of soil mineralogy and different chemical
76-79 amendments on H2O2 effectiveness. Furthermore, a wide range of metal-based catalysts have been designed and investigated for the activation of H2O2, as well as some organic, non-metal compounds. Figure 2.8 provides a general overview of different pathways to
80 obtain active oxidants derived from H2O2.
31 State of the art
Figure 2.8: Various ways for the activation of hydrogen peroxide.80
In 1894, Fenton discovered that soluble iron (II) salts could act as a catalyst with
H2O2 solutions under mild acidic conditions (pH 3-5), and mediate a strong oxidation process which produces hydroxyl radicals according to the following reaction (“Fenton’s reagent”):81
2+ 3+ - • Fe + H2O2 Fe + OH + OH
Fenton’s discovery has been evolved over time, and several other catalysts and radicals that are not present in the original “Fenton process” have been found to play key roles in degradation of organic contaminants in wastewater and soils.
When H2O2 is used as a direct oxidant it refers to a direct electron-transfer reaction between a reactant and H2O2. While it is presumed that H2O2 has a little direct reactivity towards organic compounds, it does have a high degree of direct reactivity towards inorganic compounds. Inorganic compounds showing high degrees of direct reactivity towards H2O2 are for instance transition metals, inorganic anions, and mineral surfaces.82
32 State of the art
Indeed, transition metals are typically responsible for the activation of H2O2 and initiation of free radical reaction mechanisms. As aforementioned, hydrogen peroxide is able to act either as an oxidizing or a reducing agent, as described by the half-equations:
+ - 0 H2O2 O2 + 2 H + 2e (E = -0.682 V)
+ - 0 H2O2 + 2H + 2e 2 H2O (E = 1.776 V)
For instance, under acidic conditions, H2O2 readily oxidizes iron(II) to ferric iron
[Fe(III)]; however it also reductively dissolves manganese dioxide [Mn(IV)] to dissolved
Mn(II).83
Furthermore, the rate of radical formation and reactivity often depends on the pH of the reaction medium. Earlier studies have shown that the formation of HO· radicals mediated
2+ by Fenton’s reagent (Fe + H2O2) occurs preferably at acidic conditions, whereas in neutral medium no generation of HO· was observed.84, 85 Instead, the formation of the less reactive iron (IV) has been postulated.86, 87 Furthermore, a survey of literature shows that a variety of
- - reactive species (e.g. HOO , HOO·, O2· ; s. Table 2.6), and not only HO· radicals, can be
88-90 obtained from H2O2 decomposition. Their appearance and concentrations also depend strongly on the pH of the reaction medium.
Table 2.6 lists the pH areas in which the formation of the different reactive species is favored. It was found that the HO· can be best promoted in pH values below 11.9, while the
HOO· rather prefers pH value above 4.8. In order to obtain a high concentration of HOO- anions, pH values above 11.6 seem to be required.
33 State of the art
88-91 Table 2.6: Reactive species found in H2O2 solutions at different pH values
Species Species formula pH where present
Hydrogen peroxide H2O2 pH < 11.6 Hydroxyl radical HO· pH < 11.9
- Superoxide anion O2· pH > 4.8 Perhydroxyl radical HOO· pH > 4.8 Hydroperoxide anion HOO- pH > 11.6
According to literature, formed HOO- anions obtained from the hydrogen exchange reaction
85, between H2O2 and water, initiate the formation of HO· and HOO· radicals (Scheme 2.1). 86
- + HO2H + HOH HOO + H2OH HOOH + HOH
- - H2O2 + HOO HO· + HOO· + HO
Scheme 2.1: Formation of hydroxyl and hydroperoxide radicals
In addition, the pH of the reaction system also determines the rate of radical production and reactivity. Some radical species, such as superoxide and perhydroxyl radical, are weak acids and bases, which will deprotonate in alkaline solutions, and therefore, affect their reaction pathways. Moreover, deprotonation will affect their standard reduction potential, which is the driving force for oxidation and reduction reactions.82 Selected half-reactions and their standard reduction potentials are given in Table 2.7.
34 State of the art
Table 2.7: Reduction potentials of selected reaction species in CHP systems82
Reaction Standard reduction pH range potential E0
1.776 Acidic
-0.682 Alkaline
0.878 Alkaline
2.59 Acidic
1.64 Alkaline
1.495 Acidic
-0.33 Alkaline
35 State of the art
2.4.2 Mechanism proposed by several authors
The activation of hydrogen peroxide has been an interesting field of study due to its high potential, e.g. for application in wastewater treatment or the synthesis of (fine) chemicals, particulary the functionalization of alkanes to their corresponding alcohols or ketones.92-94
The oxidation requires the use of catalysts (e.g. transition metal complexes, metal oxides) in presence of peroxo compounds as oxidizing agent, reacting either through a ionic or through radical mechanism. Many research groups explored the Fenton system under different conditions. However, recently, also group 3 metals (e.g. Al, Ga, In) became the subject of deeper investigation.
The first proposal for the generation of hydroxyl radicals from H2O2 in the presence of Fe(II) aqua complexes was postulated by Haber and Weiss.95 A one-electron oxidation of
Fe(II) to Fe(III) is involved and accompanied by the homolytic cleavage of the O-O-bond in the H2O2 molecule. The initial Fe(II) complex may afterwards be restored by reduction of the
96 Fe(III) species with H2O2. This proposal has been modified later by Kozlov et al. , suggesting that two H2O2 molecules are involved in the process, with the first one being responsible for the generation of HOO- anion, which is ligated to the iron center. The second
H2O2 molecule is coordinated to the Fe-cation and undergoes the O-O bond cleavage
(Scheme 2.2).96
Scheme 2.2: Mechanism of the iron-catalyzed HO• radical generation involving two H2O2 molecules96
36 State of the art
In 1980, Oshima et al. showed that stereoselective epoxidation of allylic alcohols and conversion of secondary alcohols into ketones are obtained in presence of organo- aluminumperoxides (e.g. (t-BuO)3Al/t-BuOOH). The yields for epoxides were over 70% and those for ketones over 90%.97 They proposed that a three-coordinated aluminum intermediate is formed during the reaction resulting in the obtained products.
The same oxidizing system was further studied by Dodonov et al., e.g. for the reaction of tri- and tetra-substituted alkenes under mild conditions (20 °C).98 It was found, that the reaction proceeds via the formation of tertiary allylic hydroperoxides which transform subsequently into unsaturated alcohols, epoxy alcohols and carbonyl compounds. Several pathways for the formation of hydroperoxides are proposed involving a radical oxidation process.
In 2008, Lei and co-workers proposed a Baeyer-Villiger oxidation of ketones by using
99 AlCl3 as catalyst dissolved in ether an aqueous H2O2 (30 wt%). It was observed that acyclic and cyclic ketones were transformed into the corresponding lactones or esters with high conversion and selectivity.
Mechanism for the generation of HO· radicals mediated by the simple aqua-complex
3+ 100 101 [Al(H2O)6] was almost simultaneously reported by Rinaldi et al. and Mandelli et al. ,
3+ but for different reaction mixtures. Rinaldi found that the [Al(H2O)6] system efficiently catalyzes the epoxidation of olefins and α,β-unsaturated ketones in THF, while Mandelli and co-workers showed that oxidation of alkanes in acetonitrile (CH3CN) is also possible in
3+ presence of Al /H2O2. The reaction mechanism for the latter has been recently proposed by density functional theory (DFT) calculations of Novikov et al.93 They proposed the
3+ substitution of one of the water ligands in the [Al(H2O)6] system for the H2O2 molecule,
37 State of the art
followed by protolysis of the coordinated H2O2. Then, another H2O molecule is substituted for H2O2. The following homolytic HO-OH bond cleavage forms the HO radicals. According to their calculations, this mechanism is also valid for the group 3 metals gallium (Ga), indium
(In), and scandium (Sc).
3+ Scheme 2.3: Proposed mechanism of the [Al(H2O)6] catalyzed epoxidation of α,β-unsaturated ketones by Rinaldi et al.100
Over the past years, a large number of results about water exchange reactions of metal ions in aqueous solutions have been published and revealed a wide spectrum of kinetic behavior.102, 103 Likewise a very informative review about structure and dynamics of hydrated ions was published in 1993 by Ohtaki and Radnal or by Helm et al. in 2005.104, 105
The water exchange reaction between the first and second coordination sphere is generally described as followed106, 107:
z+ z+ [M(H2O)n] + nH2O* [M(H2O*)n] + nH2O
38 State of the art
Amongst others, it was reported that the Al3+ ion consists of high charge, small ionic radius and is quite inert, and therefore, water exchange reactions, or mechanism where water is
104 replaced by H2O2, are kinetically hindered. Furthermore, if water exchange is slow, the protons exchange faster than oxygen and water, and the observed exchange rate constant
+ and reaction mechanism depend strongly on the acidity and pKa of the aqua ion and the [H ] concentration.105
2.4.3 The utilization of Group 3 metals and hydrogen peroxide as oxidant
Experimental data on the activation of hydrogen peroxide by further group 3 metals, e.g. Ga,
In, is very limited. Mostly, Al(III) and Ga(III) salts or complexes are reported to efficiently catalyze epoxidation reactions.108-110
In 2006, Uguina and co-workers showed that alumina is an efficient catalyst for
111 regioselective epoxidation of terpenic diolefins in presence of H2O2. Herein, they propose that Al-OOH is formed as active species which is reacting with the substrates to Al-OH and the corresponding epoxides. One year later, DiPasquale and Mayer studied the ability of
H2O2 to build metal-H2O2 complexes by trying to generate gallium-porhydrin peroxide
109 complexes in dichloromethane (CH2Cl) and CD2Cl, respectively. The NMR studies revealed that H2O2/CD2Cl does not react with the gallium complex.
Moreover, Pescarmona and Jacobs published a comparative study about epoxidation of alkenes with transition-metal-free heterogeneous catalysts (e.g. Ga2O3, Al-MCM-41, Ga-
110 MCM-41) and aqueous H2O2. It was shown that gallium oxide gives the highest epoxide yield for each of the tested substrates. Further, Goldsmith et al. report the first homogeneous
39 State of the art
Ga(III) catalyst for olefin epoxidation.108 It was observed that the activity of the
+ [Ga(phen)2Cl2] complex is fast and, with a 1% catalyst loading, highly selective for the epoxide product.
As mentioned before, experimental studies on In(III) as catalyst in presence of H2O2 is scarcely reported. Whereas Al(III) and Ga(III) are repeatedly reported to be good catalysts for epoxidation reactions, In(III) is often described to be an efficient catalyst for the rearrangement, deoxygenation and ring-opening of epoxides.112-115
However in 2001, Dimitrova et al. reported that indium-modified boron- and aluminum- beta zeolites are able to activate H2O2 for liquid-phase epoxidation of cinnamyl alcohol
(Scheme 2.4).116
Scheme 2.4: Probable mechanism of InO+/H[B]-beta induced epoxidation of cinnamyl alcohol adapted from Dimitrova et al.116
40 3+ Oxidative conversion of lignite mediated by [Al(H2O)6] /H2O2
3 Oxidative conversion of lignite mediated by 3+ [Al(H2O)6] /H2O2
3.1 Introduction
Due to continued and increasing reliance on oil, research into alternative resources to obtain fuels is highly necessary. Besides conversion of biomass, the interest in coal liquefaction has intensified, especially for countries which are short of oil resources, but have large coal reserves, (e.g. China or the United States). Most research into the direct liquefaction of coal has been done using heterogeneous catalysts, high pressures and temperatures. In the first part of this chapter, the direct conversion of lignite at temperatures between 25 to 70 °C is introduced. We found that a viscous product oil is obtained from subjecting lignite suspended in THF or 2-Me-THF in the presence of aluminum nitrate (Al(NO3)3) and hydrogen peroxide.
In the second part, the effect of process duration on lignite suspended in 2-Me-THF in
3+/ presence of the Al H2O2 system is introduced. In both experiment series, the obtained product oils as well as the corresponding lignite residues will be analyzed and discussed in detail.
3.2 Lignite used for experiments
The raw lignite obtained from the German pit mine “Fortuna Garsdorf” was dried and extracted in a Soxhlet extractor with an azeotrope mixture of toluene-ethanol (water-free),
68:32, until the solvent was colorless. Afterwards, the extracted coal was dried at a high
41 3+ Oxidative conversion of lignite mediated by [Al(H2O)6] /H2O2
vacuum pump (10-3 torr) at room temperature until constant weight was obtained. Table 3.1 summarizes the ultimate and approximate analysis of the raw and the processed lignite.
Table 3.1: Lignite from “Fortuna Garsdorf, 3. Sohle, tief”
Ash Volatile Water C H S N Odiff. [%] [%] [%] [%] [%] [%] [%] [%]
Unprocessed 1.4 19.5 62.2 65.40 4.72 1.18 28.70
Processed 4.3 54.0 4.78 66.99 5.21 1.43 0.76 25.6
Elemental Analysis = dry and ash free values
Figure 3.1 shows the FTIR spectrum of lignite. In assessment of the FTIR-bands was performed by comparison with the data published in literature.117-123 The broad absorption band observed between 3600 to 3200 cm-1 appears to be mainly O-H groups. The characteristic peak for aromatic hydrogen band at 3040 cm-1 is missing. This fact gives a hint that lignite shows a highly substituted and condensed structure.122 Furthermore, the
-1 peaks at 2920 and 2850 cm are assigned to aliphatic and alicyclic CH3, CH2 and CH groups. The peak around 1700 cm-1 is typical for carbonyl (C=O) functional groups as well as the strong peak at 1600 cm-1 is characteristic for aromatic C=C and vinylic C=C groups.122-
124 -1 The smaller peaks around 1440 and 1420 cm are probably due to CH3 asymmetric deformation and CH2 group in bridges. However, they may be also attributed to C=C and O-
-1 H groups. The slightly more intense peak about 1370 cm indicates CH3 symmetric deformation. Most of the peaks in FTIR spectra of coals between 1100 and 400 cm-1 are assigned to clay minerals.125 The assignments of IR bands for coal are listed in Table 3.2.12,
125
42 3+ Oxidative conversion of lignite mediated by [Al(H2O)6] /H2O2
O-H (N-H) CH3, CH2, CH
C=O
C=C
CH3,CH2 deform.
Figure 3.1: FTIR analysis of lignite from "Fortuna Garsdorf, 3. Sohle, tief"
Table 3.2: Assignment of typical IR bands for coal 117, 122, 123, 126 [cm-1] Assignment
3030 aromatic CH 2978 CH3 2940 Aliphatic CH 2925; 2860 CH3, CH2, CH 1600 Aromatic C=C 1575 Condens. arom. ring C=C 1460 Aliphatic CH2, CH3 1370 CH3 group, cyclic CH2 group 870; 814, 760 Hydrogen atoms on subst. benzene rings
Another common way to gain further information concerning the chemical structure of coal is by means of solid state 13C CP-MAS NMR. The obtained spectrum reveals the general distribution of aromatic (150–100 ppm) and aliphatic (90–10 ppm) region, which is roughly about 50:50% in lignite. As already observed in the FTIR spectrum, a high number of carbon- oxygen functions can be found. The peaks around 150 ppm and 70 ppm indicate the
43 3+ Oxidative conversion of lignite mediated by [Al(H2O)6] /H2O2
presence of phenols and aliphatic alcohols, respectively; while the peaks at 160 ppm is typical for carboxyl and ester functions, and the peaks at 210 and 190 ppm for carbonyl groups in ketones and aldehydes, respectively. Additional features are the peaks at about
30 ppm and 17 ppm which are most probably originating from methylene carbons and terminal methyl groups.
The resonance assignments are indicated in the spectra and listed in detail in Table 3.3.127-
129
Figure 3.2: 13C CP MAS NMR of lignite (“Fortuna Garsdorf”).
44 3+ Oxidative conversion of lignite mediated by [Al(H2O)6] /H2O2
Table 3.3: Assignments of different chemical shift ranges in 13C NMR spectra127-129
Entry Chemical Shift Functional groups [ppm]
1 0 – 25 Methyl CH3
2 25 – 50 Methylene CH2
3 50 – 60 Methoxy, methyne, quaterny CH3O-, CH-NH, CH, C
4 60 – 90 Alcohol, ether CHOH, CH2OH, CH2-O- 5 90 – 120 Aromatic CH 6 120 – 135 Aromatic CH, C 7 135 – 160 O-substituted aromatic C-O, C-OH 8 160 – 185 Ester, carboxyl COO, COOH 9 190 – 200 Aldehydes HC=O 10 200 – 205 Ketones >C=O
Microstructural and chemical composition investigations of lignite and its ash were carried out using SEM and EDS. Figure 3.3a and 3.3b show SEM pictures of the unprocessed lignite. The coal particles are round and oval in form. Also some larger particles, either round or somewhat angular can be seen, especially in the 1.5k magnification (Figure 3.3b). This picture reveals that the particles own a more or less smooth surface with small particles distributed on the top. In turn, the coal ash particles, obtained after combustion at 850 °C for
2 h, exhibit pits and craters within the particles which give a sponge-like appearance (Figure
3.3c).
45 3+ Oxidative conversion of lignite mediated by [Al(H2O)6] /H2O2
Figure 3.3: a) SEM picture of lignite (unprocessed), 500x zoom; Figure 3.3b (right): SEM picture of lignite (unprocessed), 1.5k zoom
Figure 3.3c: SEM picture of coal ash obtained from lignite, 3.0k zoom
The inorganic constituents of lignite from “Fortuna Garsdorf”, given by the supplier and confirmed by EDX analysis, are shown in Table 3.4. Some of it may be removed during coal preparation and/or conversion process. However, a certain amount will invariably always stay embedded in the organic matrix. This mineral matter can be removed by different ways when coal is used or processed. The inorganic constituents of coal are often given as ash and sulfur content.12
46 3+ Oxidative conversion of lignite mediated by [Al(H2O)6] /H2O2
Table 3.4: Mineral matter in lignite (unprocessed) “Fortuna Garsdorf, 3. Sohle, tief”
Mineral matter in lignite [%]
(Determined from ash content as oxides)
sodium (Na2O) 2.8
potassium (K2O) 0.7
calcium (CaO) 22.6
magnesium (MgO) 4.4
aluminum (Al2O3) 14.2
silicon (SiO2) 1.8
iron (Fe2O3) 14.9
sulfur (SO3) 38.3
47 3+ Oxidative conversion of lignite mediated by [Al(H2O)6] /H2O2
3.3 Conversion of lignite at different temperatures
3+ First, the lignite, dispersed in 2-Me-THF, and in presence of the Al /H2O2 system was stirred at the preferred reaction temperature (40-70 °C). After the reaction (30-240 min), the reaction mixture was cooled down in an ice bath and the liquid and solid phases were separated by centrifugation or filtration. The obtained fractions were separated and the solid residues (lignite residues) were washed and centrifuged again two times with 2 mL of solvent. The organic solution was washed twice with 5 mL distilled water. The aqueous fractions were collected separately. Here, the residual H2O2 was removed by adding carefully manganese (II) oxide. The solution was filtered and the water was slowly removed under reduced pressure. The organic solution was heated up to 85 °C with Deperox®
Molecular Sieve (Sigma-Aldrich) in order to remove residual H2O2. In sequence, the mixture was also filtered and the solvent was removed under reduced pressure (rotary evaporation).
After that, the obtained oil was dissolved in acetone to precipitate residual mineral matter, which was then removed by centrifugation. The acetone was removed by rotary evaporation.
The obtained product oil was additionally dried by a high vacuum pump at room temperature for several hours until constant weights were obtained. Figure 3.4 presents a scheme of steps which were taken after the reaction.
3+ Figure 3.4: Scheme for direct liquefaction of coal mediated by [Al(H2O)6] .
48 3+ Oxidative conversion of lignite mediated by [Al(H2O)6] /H2O2
3.3.1 Characterization of the product oil
To examine the effect of temperature on the conversion of lignite, experiments were performed at temperatures of 25-70 °C. In addition, a blank experiment without catalyst at
65 °C and for 4 h was performed. No product oil was obtained.
Table 3.5 summarizes the obtained yields of product oil and the elemental composition. Very low amounts of product oil are obtained at 25 and 40 °C, while nearly complete conversion of lignite into product oil is reached at temperatures between 55 and 70 °C after 4 h. The product oil comprises 98.5–100% of volatile products. The product oil shows a significant increase in the H-content in their elemental composition, while the O-content is only slightly higher than that found for lignite. The results from Table 3.5 show that the method also reduces the sulfur content by about 10 times, which is a significant benefit for an eventual further downstream processing of the liquid product by heterogeneous catalysis.
Furthermore, the results show that the N-content in the product oil was reduced from about
0.8 for the unprocessed lignite to 0.3%. The C/H ratio of lignite (≈ 1) indicates that this substrate is highly aromatic. However, in the obtained oils, the amount of hydrogen increases by around 70%, implying that the liquid products possess a more saturated structure than lignite. The obtained product oils were highly viscous and dark brownish in color. In addition, a blank experiment without catalyst at 65 °C and for 4 h was performed.
No product oil was obtained.
49 3+ Oxidative conversion of lignite mediated by [Al(H2O)6] /H2O2
Table 3.5: Yields, elemental composition of product oil obtained from reactions between 25 and 70 °C after 4 h.
Reaction Yield C H N S Odiff. C/H C/O ratio temperature oil ratio [%] [%] [%] [%] [%] [g]
Lignite / 67.0 5.2 0.8 1.4 25.6 C100H92 C100O36
± 0.5 ± 0.1 ± 0.1 ± 0.1
25 °C 0.02 60.1 7.5 0.5 0.3 31.6 C100H149 C100O39 ± 0.5 ± 0.1 ± 0.1 ± 0.1
40 °C 0.15 58.6 7.7 0.32 0.2 33.2 C100H158 C100O43 ± 0.7 ± 0.1 ± 0.1 ± 0.1
45 °C 0.3 59.0 7.8 0.27 0.2 32.7 C100H160 C100O42 ± 0.7 ± 0.1 ± 0.1 ± 0.01
50 °C 0.4 58.8 7.5 0.25 0.2 33.2 C100H153 C100O42 ± 0.4 ± 0.3 ± 0.1 ± 0.1
55 °C 0.4 60.1 7.8 0.2 0.1 31.8 C100H155 C100O40 ± 1.3 ± 0.1 ± 0.1 ± 0.1
60 °C 0.5 57.1 7.4 0.3 0.3 34.9 C100H155 C100O46 ± 0.6 ± 0.1 ± 0.1 ± 0.1
65 °C 0.6 57.8 7.8 0.2 0.2 34.1 C100H162 C100O44 ± 0.2 ± 0.1 ± 0.1 ± 0.1
70 °C 0.7 57.3 7.6 0.3 0.2 34.3 C100H160 C100O45 ± 0.3 ± 0.2 ± 0.1 ± 0.1
FTIR analysis (Figure 3.5) verifies the hypothesis that the oil is less aromatic than lignite. The characteristic band at 1600 cm-1 in the spectrum for lignite (Figure 3.1), typical for stretch of C=C bond for aromatics, almost completely disappears in the spectra of the product oil samples. Additionally, the peaks at 2920 and 2850 cm-1, assigned to aliphatic
50 3+ Oxidative conversion of lignite mediated by [Al(H2O)6] /H2O2
and alicyclic CH3, CH2 and CH groups, became more distinct in the spectra for the oil samples compared to the one collected from lignite. Another feature in the spectrum is the presence of an intense band at 1775 cm-1, related to stretch of C=O bond in 5-membered lactone rings (e.g., butyrolactone). This observation suggested that there should be some incorporation of solvent, into the structure of the low-molecular weight coal fragments, followed by oxidation on the 2nd position in the (2-Me)THF ring. However, as it will be presented in the next sections, not only the oxidative coupling of THF or 2-MeTHF with lignite fragments, but also degradation of THF forming an oily matrix of poly-(THF) or poly
(2-Me-THF) occurs through this methodology. The latter chemical event poses great problems to the identification of the nature of the chemical structures present in the product oil.
O-H CH3, CH2, C=C CH C=O
Figure 3.5: FTIR analysis of product oil samples obtained from reactions at different temperatures after 4 h.
51 3+ Oxidative conversion of lignite mediated by [Al(H2O)6] /H2O2
To verify the loss of aromaticity, 13C CP-MAS NMR was also performed. The spectrum (Figure 3.2) of lignite shows a presence of ca. 50 % aromatic structures (signals between 150 – 110 ppm), while the spectra of the product oil samples (Figure 3.6) indicate almost no aromatic structure. A distribution of carbon-oxygen functionalities is seen in the spectra. Based on those spectra, the product oil samples comprise many alcohol and ether functions as well as aldehydes, carboxyl and carbonyl groups. Also, some unsaturated non- aromatic compounds seem to be present (110-100 ppm).
40 °C
45 °C
50 °C
55 °C
60 °C
65 °C
70 °C
Lignite
Figure 3.6: 13C NMR of product oil samples obtained from reactions at different temperatures after 4 h.
52 3+ Oxidative conversion of lignite mediated by [Al(H2O)6] /H2O2
In addition, a small coal liquefaction experiments was performed in d-THF and analyzed by HSQC-NMR (Figure 3.7). The spectrum mostly mirrors the 13C CP-MAS NMR spectra (Figure 3.6) for the product oils and shows that a small amount of -COOH, -COOR aromatic compounds (7.97/162.4 ppm) are still present. Furthermore, the product comprises unsaturated aliphatic C=C (4.3/101.5 – 5.3/106.3 ppm), many alcohol and ether functions
(3.9/61.1 – 4.4/76.9 ppm), though the ketone compounds are missing in Figure 3.7. The 1H-
NMR spectra for the oily product obtained from the blank reaction show similarities in the presence of new formed compounds with aldehyde- and ketone functional groups (11 –
9.5 ppm).
Figure 3.7: HSQC NMR spectrum of product obtained from an experiment with lignite in d-THF at 65 °C and 4 h.
53 3+ Oxidative conversion of lignite mediated by [Al(H2O)6] /H2O2
To estimate the number average molecular weight (Mn) and the weight average molecular weight (Mw) of the product oil samples, the samples were analyzed by Gel
Permeation Chromatography (GPC, Figure 3.8). Surprisingly, the distributions of apparent molecular weight distributions (relative to polystyrene standards) are similar for the samples processed at the studied temperatures.
Although the absolute values have a limited physical meaning due to the differences in chemical structure of coal fragments and polystyrene standards, the similarity in the chromatogram profiles is thought-provoking, and suggests that (1) the temperature would not affect the product composition, which is difficult to believe, or (2) the solvent degradation products are majority in identity, while the very distinguished fragments of coal show low individual levels in the mixture.
Since all obtained spectra of obtained product oil samples show strong similarities, a blank experiment with 2-Me-THF under the same reaction conditions used for the reaction with lignite (65°C, 4 h) was performed. A yellowish oily product was obtained. GPC analysis of the oily product obtained from 2-Me-THF degradation is also presented in Figure 3.8.
54 3+ Oxidative conversion of lignite mediated by [Al(H2O)6] /H2O2
Figure 3.8: GPC analysis of product oil samples obtained from reaction at different temperatures after 4 h compared to the GPC analysis of the oily product from 2-Me-THF degradation (65 °C, 4 h). Detection of the analysis performed at a wavelength of 220 nm.
To gain further information on the thermal stability of lignite and product oil samples, thermogravimetric analysis (TGA) was performed under argon atmosphere. TGA and their first derivative (DTG) curves of lignite, and product oil obtained at 65 °C, and the oily product
3+ obtained from the degradation of 2-Me-THF in the presence of Al /H2O2 are shown in Figure
3.9 a, b and c, respectively. The pyrolysis of lignite is characterized by a three-stage thermal degradation, as detected in the TGA and DTG curves (Figure 3.9a). At the first stage (30-
150 °C), the TG curve shows a loss of physically absorbed water on lignite. At the second stage (150-525 °C), alkyl aromatics compounds are volatilized,130 and primary carbonization takes place, releasing gaseous products and tars.131 At the third stage (>525 °C), the catalytic process develops at a slower rate (secondary carbonization). Typical products
132, 133 released are CO, CO2 and C1-C3 hydrocarbons.
55 3+ Oxidative conversion of lignite mediated by [Al(H2O)6] /H2O2
Figure 3.9b shows TGA and DTG curves of the product oil obtained from a reaction at 65 °C for 4 h. A three-stage degradation is observed. A weight loss of 50 % occurs in the first stage (30-254 °C). After 254 °C, it seems there is an overlapping or blending of volatilization of different kind of compounds since the DTG curve is rising again at ca. 300 °C in stage 2 (254-360 °C) and, after a short decrease, further increasing at the beginning of stage 3 (~400 °C), indicating still the presence of coal-like fragments or structures. Finally,
Figure 3.9c displays the TGA and DTG curves of an oily product, obtained from a blank reaction with 2-Me-THF. Interestingly, comparing Figure 3.10b and Figure 3.10c, it shows a substantial weight loss at about 143 °C. Karl Fischer analysis of the product oil and 2-Me-
THF degradation products indicate an H2O-content below 0.01% for both products.
Therefore, this weight loss cannot be assigned to water vaporization.
The results from TG analysis of lignite, product oil, and 2-Me-THF degradation product show that only about 50% of the weight of the product oil should be composed of coal fragments. To verify this hypothesis the elemental analysis data was revisited in more detail.
56 3+ Oxidative conversion of lignite mediated by [Al(H2O)6] /H2O2
a)
b)
c)
Figure 3.9: a) TG and DTG curve of lignite; b) TG and DTG curve for the product oil obtained from lignite after a reaction at 65 °C and 4 h; c) TG and DTG curve for the oily product obtained from a 3+ reaction without lignite (2-Me-THF + Al /H2O2) at 65 °C and 4 h.
57 3+ Oxidative conversion of lignite mediated by [Al(H2O)6] /H2O2
Using the elemental composition of 2-Me-THF, coal and product oil, and considering that the product oil comprises both 2-Me-THF and lignite fragments, an estimate of the product oil composition can be obtained through the following 3x3 equation system:
a C2-Me-THF + b CCoal + 0 = CCoal oil Eq. (3.1)
a H2-Me-THF + b HCoal + c HOH = HCoal oil Eq. (3.2)
a O2-Me-THF + b OCoal + c OOH = OCoal oil Eq. (3.3)
Table 3.6 lists the product oil composition with respect to the formal content of 2-Me-THF, coal and “OH groups”. The latter contribution to the product oil composition is needed for accounting for the products formed by oxidation. The data from Table 3.6 shows that the content of 2-Me-THF degradation products in the product oil samples is estimated at 35 ±
2% irrespective of the reaction temperature. The content of the coal fragments in the product oils is estimated at 51 ± 2%. In turn, the content of “OH groups”, created by oxidation, slightly increases (from 12 to 16%) with reaction temperature. The average value for all samples is estimated at 15 ± 2%.
58 3+ Oxidative conversion of lignite mediated by [Al(H2O)6] /H2O2
Table 3.6: Estimated composition ratio of the different samples of product oil obtained from reaction at different reaction temperatures for 4 h.
Reaction 2-Me-THF Coal OH temperature [%] [%] [%]
25 °C 34 54 12
40 °C 36 50 15
45 °C 38 48 14
50 °C 34 53 15
55 °C 37 52 13
60 °C 31 53 17
65 °C 37 48 16
70 °C 35 49 16
In the light of the GPC results, the resemblances of the first stage of the TGA curves (Figure
10b and 10c) strongly suggests that the coal fragments should not be chemically coupled with solvent molecules. In fact, the fragments should be dissolved in the oily products derived from oxidative degradation of 2-Me-THF.
To check this hypothesis, lignite and the oily product (2-Me-THF degradation) were mixed
(50:50 wt%) and analyzed by TG. A predicted curve of this mixture was calculated using Eq.
3.4, which corresponds to the average value of the obtained data for the TG curve of Lignite and the TG curve of 2-MeTHF-oil.
TGALignite + TGAOily Product = TGAestimated Eq. (3.4) 2
59 3+ Oxidative conversion of lignite mediated by [Al(H2O)6] /H2O2
Figure 3.10 shows that the TG curves for the product oil and the mixture of the oily product from 2-Me-THF degradation with coal are very similar at the beginning until about 150 °C.
Likewise, the temperature peak of the product oil derived from lignite degradation reaction
(=140 °C) in the DTG curve is very close to that of the oily product and coal mixture (138 °C).
The simulated DTG curve for the mixture has its peak temperature around 143 °C and is slightly lower (-0.67) than the measured ones (-0.74). Overall, the experimental TG curve agrees well with the simulated TGA curve, confirming the hypothesis that coal fragments
3+ are dissolved in the oily products of 2-Me-THF. Therefore, the coal conversion by Al /H2O2 should not be referred to as direct coal liquefaction, but as oxidative coal conversion.
Figure 3.10: Comparison of experimental and calculated TGA and DTG curves of oil obtained from a reaction with 2-Me-THF mixed with lignite (50:50 wt%), and a coal liquid obtained after a reaction at 65 °C for 4h.
60 3+ Oxidative conversion of lignite mediated by [Al(H2O)6] /H2O2
Figure 3.11a presents the TG curves of the product oil samples obtained from experiments performed at different temperatures. By the analysis of the DTG curves, three stages of weight loss were found. The first stage (53-275 °C) is related to the volatilization of 2-Me-
THF degradation products. The weight loss found for the first stage slightly increases (from about 47 to 60%, Table 3.7) with the reaction temperature. The values found by TGA agree well with those estimated from the elemental composition of the samples (Table 3.5). The sum of weight losses occurring within the second and third stage is maintained at 83-89% in this experiment series. Regarding the solid residue found at the end of the temperature program of the TG analyses (1000 °C), a decrease in the amount of residue with the reaction temperature up to 55 °C was observed; however, at higher temperatures (60 - 75 °C), the amount of residue is slightly high and accounts for 13 ± 1% of the sample.
Table 3.7: Weight loss in % observed during the thermal degradation of product oil samples and their solid residue compared to lignite and an oily product of 2-Me-THF degradation Weight loss [%] Stage 1 Stage 2 Stage 3 Sample Residue [%] (53 – 275 °C) (275 –500 °C) (500 – 1000 °C)
Lignite 13.4 27 29.5 30.1 2-Me-THF oil 87.1 8.7 1.4 2.8 Product oil
obtained at 40 °C 47.2 26.2 9.5 17.1 45 °C 53.4 24.5 8 14.1 50 °C 51.9 24.6 11 12.5 55 °C 60.0 21.3 8.1 10.6 60 °C 53.4 25.7 6.8 14.1 65 °C 58.9 24.2 4.9 12 70 °C 59.0 24 4.3 12.7
61 3+ Oxidative conversion of lignite mediated by [Al(H2O)6] /H2O2
a)
b)
Figure 3.11: a) observed TGA curves for product oils and b) their corresponding DTG curves obtained at reaction temperatures between 70 and 40 °C for 4 h, compared the oily degradation product of 2-Me-THF.
62 3+ Oxidative conversion of lignite mediated by [Al(H2O)6] /H2O2
3.3.2 Characterization of the obtained coal residues
3+ As already mentioned in section 3.3.1, subjecting lignite to Al /H2O2 system leads to the partial solubilization of lignite in THF or 2-Me-THF. In this section, the focus is on the
3+ characterization of lignite residue remaining in the experiments performed with Al /H2O2 at various temperatures for a duration of 4 h. In order to assess the particle morphology of the lignite residues, the samples were analyzed through scanning electron microscopy (SEM).
Figure 3.12 displays the pictures taken from unprocessed lignite, and the lignite residues
3+ obtained by treatment of lignite with Al /H2O2 in 2-Me-THF at various temperatures for a duration of 4 h. Lignite particles show a rough surface. This feature changes after subjecting
3+ lignite to Al /H2O2. The lignite residue obtained from the experiment at 25 °C still displays particles with a rough surface; however, it is also possible to visualize particles with flat, smooth surfaces. This observation is more apparent for the particles of the residues from the experiments performed at 40, 45 and 50 °C (Figure 3.12 c, d, f). In contrast, the residues obtained from the experiments at 55, 60, 65 and 70 °C show a progressive increase in roughness of the particles because of the formation of holes and crates within the particles.
The visualization of the features of the particles of lignite residues strongly suggests that the
3+ organic matter is leached by treating lignite with H2O2/Al in 2-Me-THF. To verify this hypothesis, the contents of C, H, N, S and O (by weight difference) as well as the ash content were determined.
63
a) b) c)
d) e) f)
g) h) i)
Figure 3.12: SEM pictures of unprocessed lignite (a) and reactions residuals obtained from reactions at b) 25 °C, c) 40 °C, d) 45 °C, e) 50 °C, f) 55 °C, g) 60 °C, h) 65 °C and i) 70 °C.
64 3+ Oxidative conversion of lignite mediated by [Al(H2O)6] /H2O2
Table 3.8 lists the amount of lignite residues (relative to the initial weight of lignite) as well as their elemental composition and ash content. The C-content dramatically decreases from
67% for lignite to values as low as 34.2% for the residues obtained by processing lignite with
3+ Al /H2O2 system at 70 °C. It is clear that the treatment at higher temperatures (55-70 °C) causes a more significant solubilization of the organic matter of lignite. However, it is also evident that the remaining organic matter is of a more oxidized structure as revealed by the high O-content (determined by difference) found for the samples treated at high temperatures (55-70 °C, 50-55% O) in addition to the decrease in H-content found for these samples. The ash content of the residues are ca. 3 to 3.5 times higher (i.e. 11-15%) than that determined for the lignite (4.3%). There is no correlation between ash-content of the
3+ lignite residues and the degree of solubilization of lignite through processing with Al /H2O2
3+ system at different temperatures. Interestingly, the Al /H2O2 system is a homogeneous system. Therefore, the washing procedure applied to the lignite residue (i.e. washing of the solids with 2-Me-THF) should already suffice for the removal of Al(NO3)3 residues. However,
Table 3.8 also shows that the lignite residues are richer in N-content (3.9–4.7%) compared with unprocessed lignite (0.76 %). Assuming the N-content (% N) to stem from nitrate ion
3+ associated stoichmetrically with Al , one could estimate the contribution of Al(NO3)3 to the ash content (as AlsO3) by following equation:
% 푁 푤퐴푙 푂 = 푥 101.9582 Eq. (3.5) 2 3 14.007 푥 6
65 3+ Oxidative conversion of lignite mediated by [Al(H2O)6] /H2O2
Comparing the determined ash content and the estimated ash content as Al2O3, Table 3.8 shows that the determined ash content cannot be fully attributed to the presence of Al(NO3)3 adsorbed on the lignite residues. In fact, the content seems to derive mostly from the original ash content of lignite.
Table 3.8: Yields, elemental compositions (daf) and ash content for lignite residues obtained from reactions at different reaction temperatures and for 4 h.
Reaction Reaction C H N S Odiff. Ash Ash as
temp. residues [%] [%] [%] [%] [%] [%] Al2O3 [g] [%]
Lignite --- 67.0 5.2 0.8 1.4 26 4.3 --- ± 0.5 ± 0.1 ± 0.1 ± 0.1
25 °C 0.63 47.1 4.9 3.9 1.7 42 12 4 ± 1.3 ± 0.1 ± 0.1 ± 0.1
40 °C 0.57 43.5 5.0 4.6 1.7 45 14 5 ± 0.6 ± 0.1 ± 0.3 ± 0.1
45 °C 0.46 47.8 5.3 3,83 1.5 42 11 4 ± 0.8 ± 0.1 ± 0.1 ± 0.2
50 °C 0.21 45.0 5.0 3.5 1.2 45 12 3 ± 1.2 ± 0.3 ± 0.4 ± 0.2
55 °C 0.17 39.2 5.0 5.0 1.4 50 15 5 ± 0.4 ± 0.1 ± 0.0 ± 0.3
60 °C 0.10 37.0 4.9 5.7 1.6 50 17 6 ± 0.2 ± 0.1 ± 0.1 ± 0.3
65 °C 0.06 40.6 5.2 5.9 1.9 46 15 6 ± 1.7 ± 0.4 ± 0.2 ± 0.2
70 °C 0.03 34.2 4.9 4.7 1.6 55 - 5 ± 0.2 ± 0.1 ± 0.2 ± 0.2
66 3+ Oxidative conversion of lignite mediated by [Al(H2O)6] /H2O2
As discussed in the previous section, the S-content present in the product oils are in the range of 0.1–0.3% (daf), which is much lower than that found for the unprocessed lignite
(1.43%, daf). In contrast, the elemental composition of the lignite residues shows slightly higher values for S-content (1.5–1.9%), compared to lignite. This observation indicates that part of S-content of lignite stays in the lignite residue, but the majority of the S-content is most likely leached as sulfates upon working up of the mixture with water (s. Chapter 3.3.3).
Further information on the nature of the organic matter present in the reaction residue was obtained by 13C CP MAS NMR of those residues.
Figure 3.13 reveals the loss of aromatic compounds (140–110 ppm) with increasing reaction temperatures. Due to the small amount of reaction residue obtained at 70 °C, NMR measurement was not possible. A peak at ca. 168 ppm progressively becomes more prominent for the lignite residues obtained from processing at 55, 60 and 65 °C. This peak indicates the presence of carboxylic acids in the residual organic matter. In the light of these results, we can draw the conclusion that the obtained lignite residue comprises not only mineral matter, but also oxidized organic compounds, which are insoluble in 2-Me-THF at temperatures as high as 65 °C.
67 3+ Oxidative conversion of lignite mediated by [Al(H2O)6] /H2O2
3
R
-
OH
H
C
OH
3
2
-
-
-
-
O
ar
-
ar
ar
al
COOR, COOH
C
R OCH C
CH
C
C
CH
>C=O
- -
-
- -
-
-
-
- -
65 °C
60 °C
55 °C
50 °C
45 °C
40 °C
25 °C
Lignite
Figure 3.13: 13C CP MAS NMR of lignite residues obtained from reactions at different reaction temperatures after 4 h.
68 3+ Oxidative conversion of lignite mediated by [Al(H2O)6] /H2O2
3.3.3 Characterization of the water-soluble products
Using water as an extracting phase, the water-soluble products from lignite in addition to
Al(NO3)3 and residual H2O2 were isolated from the reaction mixture containing lignite fragments dissolved in 2-Me-THF. The aqueous phase was treated with manganese dioxide
(MnO2) in order to decompose residual peroxides. The suspension was filtered. The aqueous solution was dried under reduced pressure at 60 °C. The isolated fraction was a dark brownish to black, sticky solid. Table 2.1 summarizes the weight of water-soluble products, their elemental composition and ash content. The weight of water-soluble products shows no direct correlation with the reaction temperature. Since Al(NO3)3 is highly soluble in water, it is expected that a large fraction of the residue is constituted of Al(OH)x(NO3)3-x species. Again, assuming the N-content or Al2O2(NO3)2 to stem from nitrate, one can estimate the content of Al(NO3)3 to the ash content (as Al2O3) using Eq. 3. The estimate ash contents (as Al2O3) to correspond roughly to half the value of ash determined by calcination of the samples. The most important source of error in this estimation is the fact that even at temperatures as high as 60 – 70 °C, Al(NO3)3 decomposes, forming undefined
134 Al(OH)x(NO3)3-x or Al2O2(NO3)2 species. Logically, these species possess a different Al:N stoichiometric ratios from Al(NO3)3. The presence of Al(NO3)3 as a major component of the water-soluble products is confirmed by the similarity of the FTIR spectra of these products with that of Al(NO3)3 •9H2O (Figure 3.14).
69 3+ Oxidative conversion of lignite mediated by [Al(H2O)6] /H2O2
Figure 3.14: FTIR spectra of water-soluble products obtained from reactions at different temperatures.
Table 3.9 also reveals that the residue contains a substantial amount of carbon (20 – 45%).
This C-content is mostly like associated with the presence of aluminum carboxylates. This hypothesis is verified by the band at around 1450 cm-1, which is characteristic of Al3+- carboxylate species.134 Finally, the elemental compositions of the water-soluble products also show that at temperatures as high as 60 °C most of the S-content of lignite is converted into water-soluble products.
70 3+ Oxidative conversion of lignite mediated by [Al(H2O)6] /H2O2
Table 3.9: Weight of water-soluble products and elemental composition (daf) and their ash content
Temp. Water- C H N S Odiff. Ash Al Ash soluble [%] [%] [%] [%] [%] [%] [%] as
products (estimated Al2O3 [g] from N)
25 °C 0.40 20.8 4.9 14.0 0.2 60.2 31 9.0 17.0 ± 0.1 ± 0.1 ± 0.2 ± 0.1
40 °C 0.37 21.6 5.7 11.3 0.3 61.2 31 7.2 13.7 ± 0.3 ± 0.4 ± 0.1 ± 0.2
45 °C 0.47 25.3 5.1 11.2 0.5 57.9 24 7.2 13.7 ± 0.6 ± 0.1 ± 0.1 ± 0.1
50 °C 0.62 29.6 5.8 11.6 0.3 52.7 27.8 7.5 14.2 ± 1.0 ± 0.2 ± 0.1 ± 0.1
55 °C 0.70 23.7 4.5 9.6 0.3 61.9 13 6.2 11.7 ± 0.2 ± 0.1 ± 0.1 ± 0.1
60 °C 0.70 28.3 5.4 11.7 0.9 53.7 28 7.5 14.3 ± 0.2 ± 0.2 ± 0.1 ± 0.1
65 °C 0.41 45.6 6.6 6.35 0.2 41.3 18 4.1 7.8 ± 0.3 ± 0.2 ± 0.2 ± 0.1
70 °C 0.57 39.8 5.6 8.2 1.5 45.0 20 5.2 10.0 ± 0.2 ± 0.1 ± 0.1 ± 0.1
71 3+ Oxidative conversion of lignite mediated by [Al(H2O)6] /H2O2
3.4 Monitoring the evolution of the conversion of lignite
3+ Studies of the effect of temperature on coal conversion using the Al /H2O2 system showed that the highest yields of product oil were obtained at reaction temperatures between 65 and
70 °C. However, degradation of 2-Me-THF or THF concomitantly occurs with the coal conversion. To lend further insight into the conversion of lignite, the temporal evolution of the lignite solubilization was studied. Ultimately, this study aims finding the optimal duration of the process in order to mitigate the degradation of the reaction solvent while maximizing the yield of coal products.
3.4.1 Characterization of the product oil
Table 3.10 summarizes the weight yield of product oil in addition to their composition.
Analyzing the weight yield of product oil, one can see a slightly increase in the weights throughout the process duration. Nonetheless, it appears that the lignite conversion already
3+ took place within the first 30 min of processing with the Al /H2O2 system at 65 °C. The C- content of the product oil decreases (by about 2%) throughout the process duration. In turn, the H-content found for the product oils is sustained at about 7.9%. Likewise, the N-contents found for the product oil samples remain constant at about 0.2% throughout the process duration. Conversely, the S-content of the product oil decreases (from 0.24–0.09 %) throughout the process duration. Lastly, the O-content of the product oil increases (from 32–
35%). Overall, C/H ratio increased from C100H157 to C100H164, while the C/O ratio progressively increases from C100O41 to C100O46 throughout the process duration. All obtained product oil samples comprise 98–100 % volatile products.
72 3+ Oxidative conversion of lignite mediated by [Al(H2O)6] /H2O2
Table 3.10: Weight yields of obtained product oil samples, their elemental composition (daf) from reaction with different time duration at 65 °C
Reaction Weight C H N S Odiff. C/H C/O ratio Time yields [%] [%] [%] [%] [%] ratio of oil [g]
30 min 0.34 59.5 7.8 0.2 0.2 32.2 C100H157 C100O41 ± 0.9 ± 0.1 ± 0.1 ± 0.1
60 min 0.47 58.3 8.0 0.2 0.2 33.4 C100H165 C100O43 ± 1.2 ± 0.1 ± 0.1 ± 0.1
90 min 0.40 59.5 8.1 0.2 0.2 32.0 C100H163 C100O40 ± 0.9 ± 0.1 ± 0.1 ± 0.1
120 min 0.44 57.0 7.9 0.3 0.3 34.6 C100H166 C100O45 ± 0.26 ± 0.1 ± 0.1 ± 0.1
150 min 0.48 56.8 7.8 0.2 0.2 35.0 C100H165 C100O46 ± 1.3 ± 0.1 ± 0.1 ± 0.1
180 min 0.33 57.0 7.8 0.2 0.1 34.8 C100H164 C100O46 ± 0.8 ± 0.1 ± 0.11 ± 0.1
240 min 0.48 57.0 7.8 0.2 0.1 34.9 C100H164 C100O46 ± 0.1 ± 0.1 ± 0.1 ± 0.1
The observations summarized in Table 3.10 clearly show the process duration to effect in the further oxidation of lignite fragments. The oxidative process decrease the S- content of the sample. However, this positive side-effect of the coal processing with the
3+ Al /H2O2 system results also in the oxidation of the product oil. This last feature is less desirable because it implies in a higher demand for hydrogen in the further downstream processing for the upgrade of the product oil.
FTIR analysis provides an overview of chemical structures present in the product oil samples. Surprisingly, Figure 3.15 shows that the product oil spectra are identical to the spectrum from the oily product obtained from 2-Me-THF degradation reaction. As discussed
73 3+ Oxidative conversion of lignite mediated by [Al(H2O)6] /H2O2
in Chapter 3.3.1, the strong peak for C=O at 1715 cm-1 together with a shoulder peak at
1735 cm-1 suggest the presence of 5-membered lactone ring structures. This observation indicates that the conversion of lignite takes place already within the initial 30 min of process duration. Moreover, the decomposition of the complex structure of lignite generate a complex mixture of products where the individual yields are low. Accordingly, FTIR can detect only the degradation products of 2-Me-THF, since they are more abundant and more uniform in chemical structure.
CH3, CH2,CH C=C C=O
Figure 3.15: FTIR spectra of obtained product oil samples from reactions at 65 °C and different reaction times compared to the spectra of the oily product obtained from 2-MeTHF degradation reaction (65 °C, 4 h).
In an attempt to lend insight into the chemical and molecular structure, 13C SS-NMR analysis of the product oil was carried out. Compared with the lignite 13C NMR spectrum, all product oil samples show the complete loss of aromatic structure (150–110 ppm)
74 3+ Oxidative conversion of lignite mediated by [Al(H2O)6] /H2O2
irrespectively of process duration. Carboxylic acids and esters (180–170 ppm), ketones and aldehydes (210–205 ppm) and aliphatic alcohols and ethers (85–60 ppm) are the main functional groups present in the product oil samples.
3
R
-
3
2
O
O
-
-
al
CH
OCH
CH
R
COOR, COOH
C=C
C
-
-
-
-
>C=O
- -
- -
240 min
180 min
150 min
120 min
90 min
60 min
30 min
Lignite
Figure 3.16: 13C SS-NMR spectra from obtained product oil samples at 65 °C and different reaction duration.
75 3+ Oxidative conversion of lignite mediated by [Al(H2O)6] /H2O2
To obtain information on the dependence of the depolymerization extent of lignite
3+ through its processing with the Al /H2O2 system at varying process duration, the apparent molecular weight (MW) distribution of the product oils was estimated by GPC analysis.
Figure 3.17 shows the GPC curves for product oils and the oily degradation product from 2-
Me-THF. Apart from the contribution of the 2-Me-THF degradation products to the apparent
MW distribution curves (tree peaks at 43, 47 and 53 min), the distribution of the apparent
MW values (main peak) slightly shifts towards lower values with an increase in process duration. This observation suggests that the breakdown of coal fragments is slightly enhanced throughout the process course.
Figure 3.17: GPC analysis of obtained product oil samples from reactions at 65 °C and different process duration compared to oily product obtained from 2-Me-THF degradation reaction (65 °C, 4 h).
76 3+ Oxidative conversion of lignite mediated by [Al(H2O)6] /H2O2
As discussed in Chapter 3, the obtained product oil samples comprise fragments from lignite and 2-Me-THF degradation. In order to determine the contribution of 2-Me-THF fragments the same 3x3 equation system (Eq. 3.4) was used.
Table 3.11 lists the product oil composition with respect to the formal content of lignite, 2-Me-THF and “OH groups”. The data shows that the content of 2-Me-THF degradation products in the product oils is estimated between 38-42%, for coal between 44-
49% and for the “OH groups” between 13-17% for process duration between 30-120 min.
Surprisingly, the composition ratios for 2-Me-THF, coal and “OH group” remains the same for the process duration of 150 min to 240 min.
Table 3.11: Possible composition ratios of obtained product oils from a reaction at 65 °C and different process duration
Process 2-Me-THF Coal OH duration [%] [%] [%]
30 min 38 49 14
60 min 40 45 15
90 min 42 45 13
120 min 39 44 17
150 min 37 46 17
180 min 37 46 17
240 min 37 46 17
77 3+ Oxidative conversion of lignite mediated by [Al(H2O)6] /H2O2
In addition, TG analysis was applied. Figure 3.18 presents the TGA curves of the products oil samples obtained from experiments at various process durations. The DTG profiles of the product oils show a three-stage degradation. Again, the weight loss for the first stage (53–275 °C) is related to the volatilization of 2-Me-THF degradation products
(water content below 0.01%, Karl Fischer). Herein, the weight loss found slightly increases from about 60-66% (Table 3.12). The sum of the weight loss found within the second and third stage is about 83-88%. In matters of the solid residue obtained at the end of the TG analysis, the amount is about 13 ± 1% of the sample for all product oils.
Table 3.12: Weight loss in % observed during the thermal degradation of product oil samples and their solid residue compared to lignite and an oily product of 2-Me-THF degradation
Weight loss [%]
Stage 1 Stage 2 Stage 3 Sample Residue [%] (53 – 275 °C) (275 –500 °C) (500 – 1000 °C)
Lignite 13 40 70 30
2-Me-THF oil 87 9 1 3
Product oil
obtained after
30 min 60 23 5 12
60 min 60 23 5 13
90 min 58 24 4 14
120 min 61 24 3 12
150 min 64 18 4 14
180 min 64 21 3 12
240 min 66 18 4 12
78 3+ Oxidative conversion of lignite mediated by [Al(H2O)6] /H2O2
a)
b)
Figure 3.18: a) TGA curves of product oil samples obtained from a reaction at 65 °C and different reaction times compared to the oily product from 2-Me-THF degradation (65 °C, 4 h); b) DTG curves of product oil samples obtained from a reaction at 65 °C and different reaction times compared to the oily product from 2-Me-THF degradation (65 °C, 4 h).
79 3+ Oxidative conversion of lignite mediated by [Al(H2O)6] /H2O2
3.4.2 Analysis of lignite residues obtained from duration effect studies
As already discussed in Chapter 3.3.2, the partial solubilization of lignite in THF or 2-Me-
3+ THF by the Al /H2O2 system at different temperatures showed that the organic matter is most likely leached out of the coal particles, leaving a rough surface with holes and crates behind. Since the aim of this study was the temporal evolution of lignite solubilization, SEM pictures of the lignite residues were recorded to study the effect of process duration on the
3+ particle morphology by treatment lignite with Al /H2O2.
The SEM pictures taken from lignite residues are similar to those obtained from reactions at temperatures between 65 and 70 °C. The particles show a rough surface with crates and holes. As previously stated in section 3.3.1, the main solubilization of lignite takes place within the first 30 min at 65 °C. That is why, only slight changes in the particles appearance (e.g. more crates, holes) can be observed for the particles obtained from the experiments with 30 to 150 min process duration. The visualization of the particles obtained from the experiments with 180 and 240 min process duration show no significant changes in their morphology. This fact supports the hypothesis that the solubilization of lignite in 2-
3+ Me-THF mediated by the Al /H2O2 system at 65 °C is completed after ca. 150 min process duration.
80
a) b) c)
d) e) f)
g) h) Figure 3.19: SEM pictures of lignite (a) and lignite residues obtained from experiments at different process duration: b) 30 min, c) 60 min, d) 90 min, e) 120 min, f) 150 min, g) 180 min, h) 240 min
81 3+ Oxidative conversion of lignite mediated by [Al(H2O)6] /H2O2
Table 3.13 lists the obtained weight of solid reaction residues, their elemental composition and ash content. As assumed, the weight of lignite residues decreases with process duration, likewise their carbon content. After time duration of 30 min, the residue shows a C-content of ca. 50%, which decreases further to 37 ± 2% for the reaction duration
150 to 240 min. However, the H-content (5.1–5.5%) and S-content (1.6-2.1%) are more or less the same and do not differ much from those from lignite. As concluded in section 3.3.2, most part of the S-content stays in the lignite residue, but is also leached as sulfates upon working up processes.
Again, it is obvious that the lignite residues comprise more oxidized compounds as shown by the increased values for the O-content (from 40 to 50%). Especially for the samples obtained after 150 min to 240 min process duration, the O-content was redoubled (50± 2%) compared to lignite (26%). The amount of ash content of the lignite residues are 2 to 3 times higher (i.e. 10-15%) than that determined for coal. As already mentioned, there is no correlation between ash-content of the lignite residues and the degree of solubilization of
3+ lignite through processing with the Al /H2O2 system. However, Table 3.13 also shows that the obtained lignite residues are richer in N-content (3.2-5.3%) compared with the origin coal
(0.8%). The contribution of possible remaining Al(NO3)3 was estimated using Eq. 3.5. The ash content cannot be fully attributed to the presence of Al(NO3)3 adsorbed on the lignite residues.
82 3+ Oxidative conversion of lignite mediated by [Al(H2O)6] /H2O2
Table 3.13: Yield and elemental composition of obtained reaction residues from experiments with different process duration at 65 °C
Reaction Reaction C H N S Odiff. Ash Ash as duration residues [%] [%] [%] [%] [%] [%] Al2O3 [g] [%]
Lignite --- 67.0 5.2 0.8 1.4 26 4.3 ---
30 min 0.30 49.3 5.5 3.2 1.6 40.5 10.1 3.9 ± 1.1 ± 0.1 ± 0.1 ± 0.1
60 min 0.21 44.1 5.3 4.2 1.6 44.9 11.4 5.1 ± 0.9 ± 0.3 ± 0.1 ± 0.1
90 min 0.15 40.6 5.1 4.9 1.7 47.7 12.9 6.0 ± 0.3 ± 0.1 ± 0.1 ± 0.1
120 min 0.12 40.4 5.5 5.3 1.6 47.2 15.4 6.4 ± 0.8 ± 0.1 ± 0.1 ± 0.1
150 min 0.09 35.2 5.2 5.2 2.1 52.3 15.0 6.3 ± 0.3 ± 0.1 ± 0.2 ± 0.1
180 min 0.08 39.0 5.1 5.8 2.0 48.0 14.7 7.1 ± 0.2 ± 0.2 ± 0.2 ± 0.1
240 min 0.06 37.1 5.2 5.9 1.2 49.9 15.3 7.2 ± 1.7 ± 0.4 ± 0.2 ± 0.2
More information about the lignite residues was gained by 13C CP-MAS NMR analysis. Figure 3.20 presents the 13C NMR spectra on the lignite residues compared to the spectrum of lignite. The spectra reveal the dramatic loss of aromatic compounds (140–110 ppm) within a process duration of 30 min. The peak showing at ca. 168 ppm, typical for carboxylic acids, becomes more intense for the lignite residues obtained from experiments with 150, 180 and 240 min process duration. This observation agrees with the increase in
O-content, as determined from elemental analysis (Table 3.13). Therefore, the lignite residues not only comprise mineral matter, but also 2-Me-THF insoluble, oxidized organic materials.
83 3+ Oxidative conversion of lignite mediated by [Al(H2O)6] /H2O2
Figure 3.20: 13C CP MAS NMR of obtained lignite residues from reactions at 65 °C.
3.4.3 Characterization of the water-soluble products
Table 3.14 lists the weight of water-soluble products, their elemental composition and ash content. From the increasing weight of yield obtained after the several reactions, it can be observed that the amount of water-soluble products is increasing with process duration. This is supported by the C-content of the water-soluble products, which is growing as well (from
26 to 34%). The H-content is about 5-6% for all samples, whereas the S-content is increasing and varies between 0.3-1% with process duration. Furthermore, the N-content and O-content are decreasing (from 12 to 9%, and 56 to 50%), and correspondingly, the ash content is diminished as well (from ca. 29 to 26%). As previously hypothesized
84 3+ Oxidative conversion of lignite mediated by [Al(H2O)6] /H2O2
(Chapter 3.3.3), the highly soluble Al(NO3)3 is likely to form Al(OH)x(NO3)3-x species.
Therefore, the content of Al(NO3)3 or Al2O3, respectively, was estimated as well using Eq.
3.5. The estimated ash values relate to almost half of the ash content determined by calcination of the water-soluble samples.
Table 3.14: Weight of yield and elemental composition (daf) of obtained water-soluble products
Weight C H N S O Ash Ash Al [%] Reaction diff. of yield [%] [%] [%] [%] [%] [%] as (estimated duration [g] Al2O3 from N) [%]
30 min 0.54 26.2 5.2 12.1 0.5 56.1 29.3 14.6 7.7 ± 0.6 ± 0.1 ± 0.1 ± 0.1
60 min 0.53 28.2 5.4 11.6 0.3 54.6 29.2 14.1 7.4 ± 0.1 ± 0.1 ± 0.1 ± 0.1
90 min 0.67 31.0 5.82 10.5 0.4 52.4 27.6 12.7 6.7 ± 0.3 ± 0.2 ± 0.1 ± 0.1
120 min 0.77 32.5 5.7 9.9 0.4 51.6 26.4 12.0 6.4 ± 0.6 ± 0.1 ± 0.1 ± 0.1
150 min 0.82 33.8 5.9 10.1 0.8 49.4 28.7 12.2 6.5 ± 0.2 ± 0.1 ± 0.1 ± 0.1
180 min 0.78 34.4 5.9 9.0 0.8 49.9 26.8 10.9 5.8 ± 0.7 ± 0.1 ± 0.1 ± 0.1
240 min 0.83 32.4 6.0 9.0 1.0 51.7 26.2 10.9 5.7 ± 0.3 ± 0.2 ± 0.2 ± 0.1
Again, the FTIR spectra for the aqueous phase residues show strong similarities among themselves and the spectrum for Al(NO3)3 (Figure 3.21). The similarities seem to stem mostly from aluminum carboxylates, which is confirmed by the band at ca. 1450 cm-1, typical for Al3+-carboxylated species.
85 3+ Oxidative conversion of lignite mediated by [Al(H2O)6] /H2O2
1450
Figure 3.21: FTIR spectra for the water-soluble products compared to the FTIR spectrum of aluminum nitrate.
86 3+ Oxidative conversion of lignite mediated by [Al(H2O)6] /H2O2
3.5 Conclusion
In light of these results, we propose that the described procedure is successful in breaking down lignite into smaller, volatile fragments. The resulting product oil samples are mainly composed of saturated compounds. We tentatively suggest that hydroxyl radicals are involved in catalysing the oxidative coupling of the cyclic ether solvent with lignite and its fragments, leading to liquid products. It was shown that the conversion depends strongly on the reaction temperature. Moreover it was shown that the apparent molecular weight distribution decreases for experiments with higher reaction temperatures. CHNS/O values determined by elemental analysis showed only minor differences within the samples, equally the 13C NMR spectra. This can also be due to the products from 2-Me-THF degradation which are additionally obtained and comprised within the product oil samples. Since the amount of product oil yields is also rising with higher reaction temperature, this indicates a correlation between the activation of hydrogen peroxide and the liquefaction of lignite.
Furthermore, the results presented and discussed within this chapter showed that the main conversion of lignite is happening within the first 30 min of process duration. Again, a dramatic loss of aromatic compounds is found for the obtained product oil samples and can be observed for the lignite residues with increasing process duration. Moreover, this series of experiments revealed that a higher amount of 2-Me-THF degradation products instead of coal fragments are added to the new formed compounds after process duration longer than 60 min, increasing the weight yield of product oil.
In summary, high yields of product oil samples are only obtained due to the high amount of oily product from 2-Me-THF degradation which dissolves the obtained coal fragments.
87 Solvent effects on the oxidation of lignite
4 Solvents and solvent effects on the oxidation of 3+ lignite mediated by [Al(H2O)6] /H2O2
The varying effects of solvents, e.g. the hydrogen donor ability135, 136, on direct coal conversion and liquefaction have been studied and known for a long time. The quality of products obtained may be influenced by the characteristics of a solvent, e.g. their solubility parameters. In the context of this work, as radical oxidations are solvent-dependent, the
3+ conversion of lignite mediated by [Al(H2O)6] /H2O2 was also performed and studied in different solvents. Most importantly, the activation of hydrogen peroxide constitutes a
3+ 100 process involving the second coordination sphere of [Al(H2O)6] /H2O2. Accordingly, it is expected that solvent will exert effects upon the dynamics of hydrogen peroxide activation.
Moreover, regarding the high amount of obtained THF or 2-Me-THF degradation products comprised within the products oil samples (see Chapter 3), the focus on this study is on the difference in chemical structure of the obtained coal residues from the experiments in the different solvents.
To examine the effect of solvent on the conversion and oxidation of lignite by the
3+ [Al(H2O)6] /H2O2 system, experiments were performed in methanol, ethanol, water and 2-
Me-THF at 65 °C. Since the duration experiments showed that the main conversion of lignite is occurring within the first thirty minutes, the experiments in this study were performed for a duration of 5, 10 and 30 minutes. Furthermore, blank experiments in absence of aluminum
88 Solvent effects on the oxidation of lignite
nitrate were performed to investigate the effect of hydrogen peroxide, by its own, on coal oxidation.
Scheme 4.1 shows the experimental steps of the coal conversion experiments.
Scheme 4.1: Scheme for coal conversion/benefication process.
Lignite was dispersed in a solvent. The suspension was heated up to 65 °C under
3+ magnetic stirring. Before the solution of [Al(H2O)6] /H2O2 was added, the whole system was flushed for several minutes with Argon to remove air and then connected to a washing flask with Ba(OH)2 solution for the sequestration of CO2 eventually liberated from the oxidation processes. After the reaction, the mixture was cooled down and the solid and liquid phases were separated via filtration. The obtained lignite residue was first washed with the solvent to collect the soluble-products. Next, the coal residue was washed once with 10 mL of water to remove remaining aluminum salt. The residues were dried overnight in a vacuum oven at
60 °C.
The work-up of the obtained organic solutions included first the elimination of residual hydrogen peroxide with manganese dioxide. After filtration, a stoichiometric quantity of ammonia solution (25%) was added to the solution in order to remove aluminum ions as aluminum hydroxide Al(OH)3. Again, the mixture was filtered. To collect all organic matter, the obtained brownish solid was washed with solvent until it remained colorless. The light- yellowish solid was dried overnight in a vacuum oven at 65 °C. The organic phases were
89 Solvent effects on the oxidation of lignite
collected and the solvent was removed using a rotary evaporator (60 °C, 200 rpm, 20 min).
A brownish solid was obtained.
All reaction products were analyzed by elemental analysis (CHNS/O). The lignite residues were additionally analyzed using FTIR and 13C SS-NMR spectroscopy.
4.1 Participation of reaction medium in the extent of coal oxidation
4.1.1 Ultimate analysis of the coal residue samples
4.1.1.1 Lignite residues
Table 4.1 lists the weight of the lignite residues, their elemental composition and ash content for the blank experiments (performed in absence of catalyst, but in presence of
H2O2), and for coal residues obtained from the experiments carried out with the
3+ [Al(H2O)6] /H2O2 system. Regardless of the solvent, the dataset reveals that the experiments performed without the aluminum salt results in a lower increase in the O- content (from ca. 26 to 31-35%), compared with those carried out with aluminum nitrate
(increase in O-content from ca. 26 to 37-43%). This result indicates that aluminum nitrate is, indeed, activating H2O2 for the oxidation of lignite. As a result of the oxidative decomposition of lignite structure, the C-content decreased from ca. 67 to about 55% (vs. 62% - blank experiment). Surprisingly, however, regarding the H-content, the values are centered around the initial value found for lignite (5.2%). Clearly, in the set of experiments in the presence of
Al(NO3)3, a considerable increase in the N-content was observed, compared with respective blank experiments. This result set shows that part of the aluminum salt remained (adsorbed)
90 Solvent effects on the oxidation of lignite
on the lignite surface. Finally, the values of S-content are very scattered (0.9-1.4%), suggesting that, in some cases, the oxidative decomposition of lignite was effective at removing sulfurous species occurring in lignite.
Table 4.1: Weight of lignite residues and their elemental composition (daf) obtained from experiments performed in varying solvent at 65 °C
Entry t Solvent Catalyst Lignite H C N S O[a] Ash (min) residue (g) (%)
1 Lignite none none --- 5.2±0.1 67.0±0.5 0.8±0.1 1.4±0.1 25.6 4.3
2 5 MeOH none 0.48 4.0±0.1 64.2± 0.5 0.7±0.1 0.8±0.1 29.4 3
3 5 MeOH Al(NO3)3 0.34 4.9±0.1 55.9±1.5 0.7±0.1 0.7±0.1 37.9 1.5
4 10 MeOH none 0.46 4.8±0.1 62.2± 0.4 0.7±0.1 0.7±0.1 31.6 3
5 10 MeOH Al(NO3)3 0.25 4.9±0.2 55.4±1.2 0.6±0.1 1.2±0.6 37.5 1
6 30 MeOH none 0.43 4.7±0.1 61.2±1.9 0.7±0.1 1.3±0.4 32.1 3
7 30 MeOH Al(NO3)3 0.16 5.3±0.1 55.6±1.5 0.5±0.1 1.5±0.1 37.1 <0.1
8 5 EtOH none 0.46 4.9±0.1 63.0±0.4 0.8±0.1 0.9±0.1 30.5 4.4
9 5 EtOH Al(NO3)3 0.34 4.9±0.3 52.4±3.8 0.7±0.1 1.7±0.1 40.4 3
10 10 EtOH none 0.47 4.7±0.1 60.6±1.1 0.7±0.1 1.4±0.7 32.3 4
11 10 EtOH Al(NO3)3 0.21 5.2±0.1 56.3±1.9 0.6±0.1 1.5±0.1 36.3 3
12 30 EtOH none 0.46 4.9±0.1 59.6±0.8 0.7±0.1 1.3±0.7 33.4 4
13 30 EtOH Al(NO3)3 0.07 5.9±0.1 55.4±0.8 0.4±0.1 1.4±0.1 36.9 1
14 5 H2O none 0.47 4.8±0.1 58.4±0.8 0.7±0.1 1.5±0.7 34.7 2.5
15 5 H2O Al(NO3)3 0.43 5.1±0.1 57.9±1.0 1.0±0.1 1.8±0.6 37.3 1
16 10 H2O none 0.46 4.8±0.1 58.1±0.8 0.7±0.1 1.1±0.6 35.4 2
17 10 H2O Al(NO3)3 0.36 5.0±0.1 55.8±1.1 1.7±0.1 1.1±0.7 36.6 1
18 30 H2O none 0.37 4.7±0.1 64.3±0.6 0.8±0.1 1.0±0.6 23.3 2
19 30 H2O Al(NO3)3 0.19 4.9±0.1 55.7±1.2 1.6±0.1 0.7±0.1 37.0 1
20 5 MTHF none 0.50 4.9±0.1 60.4±1.0 0.6±0.1 0.1±0.1 33.0 4
21 5 MTHF Al(NO3)3 0.68 4.9±0.1 48.0±1.9 2.9±0.1 1.5±0.5 42.7 11
22 10 MTHF none 0.48 5.3±0.1 61.6±0.6 0.6±0.1 0.9±0.1 31.7 4
23 10 MTHF Al(NO3)3 0.43 5.1±0.1 49.1±0.4 3.0±0.1 1.1±0.4 41.7 12
24 30 MTHF none 0.45 5.1±0.1 61.0±0.5 0.7±0.1 1.9±0.1 31.4 4
25 30 MTHF Al(NO3)3 0.30 5.5±0.0 49.3±1.1 3.2±0.1 1.6±0.1 40.5 10
91 Solvent effects on the oxidation of lignite
This hypothesis of a different reaction mechanism is also verified by FTIR (Figure
4.1) and 13C CP MAS analysis (Figure 4.2). In contrast to the product oil samples discussed in Chapter 3, both analyses still indicate the presence of aromatic compounds, e.g. the peak
-1 around 1600 cm for the FTIR spectra, and the signals between 140 and110 ppm for the
NMR spectra. In case of the FTIR spectra (Figure 4.1a-d) the peak around 1700 cm-1, typical for C=O bonds, becomes more intense, while the one for the aromatic C=C bonds (1600 cm-1) suffers a decrease in intensity.
Furthermore, the 13C-NMR spectra reveal that the lignite residue samples from all blank reactions remained almost unchanged compared to the spectrum of the unprocessed
3+ lignite, but in presence of the [Al(H2O)6] /H2O2 system the signals typical for aromatic structure (140-110 ppm) start to decrease. The most obvious changes in chemical structure can be observed for the reactions performed in ethanol and 2-Me-THF as solvent (Figure
4.2b and d). Herein, a great degradation of aromatic compounds can be already found for the lignite residue samples obtained after 5 min reaction duration. In addition, the intensity of this NMR band decreased with longer reaction duration. For the lignite residue samples obtained from experiments in methanol and water, the oxidation of coal seems to be milder; their NMR spectra shows less structural changes with only slightly decreased aromatic signals (Figure 4.2a and c). However, in all spectra an increase of the signal about 200-170 ppm, typical for RCOOR and RCOOH, was observed as well.
The results obtained from FTIR and 13C-NMR analyses agree with the information obtained from elemental analysis and are clearly showing that the increased oxygen content can be assigned to new formed carbonyl and carboxyl functional groups.
92 Solvent effects on the oxidation of lignite
a) Methanol b) Ethanol
c) Water d) 2-Me-THF
3+ Figure 4.1: FTIR spectra of lignite residue obtained after treatment with H2O2 or Al /H2O2 at different reaction duration at 65 °C in different solvents: a) methanol, b) ethanol, c) water, d) 2-Me- THF.
93 Solvent effects on the oxidation of lignite
a) Methanol b) Ethanol
30 min
30 min, Blank
10 min
10 min, Blank
5 min
5 min, Blank
Lignite
ppm c) Water d) 2-Me-THF 30 min
30 min, Blank
10 min
10 min, Blank
5 min
5 min, Blank
Lignite
Figure 4.2: 13C CP-MAS NMR of lignite residues obtained from experiments in different solvents and different process duration at 65 °C compared to the residues from corresponding blank reaction (without catalyst), and unprocessed lignite.
94 Solvent effects on the oxidation of lignite
To study the change in thermal degradation, TG analyses of the lignite residue samples were performed and compared with the TG curve of unprocessed lignite. Figure
4.3a to Figure 4.6a show the TG curves and corresponding DTG curves. The analysis of the
DTG curves indicates a five-stage thermal degradation process. At the first stage (30-
105 °C), only a slight decrease in weight loss can be observed which shows that not much physically absorbed water is present in the lignite (ca. 2%) and the lignite residue samples
(3-4%). Within the second stage (105-200 °C), the TG curve of lignite remains almost straight until about 200 °C whereas the curves of the residues already start to decay at
105 °C. This change in thermal degradation behavior can also be observed in the corresponding DTG curves (Figure 4.3b to Figure 4.6b). With beginning of the second stage
(200-300 °C), the weight loss of the processed lignite samples is more intense than that of the unprocessed lignite and the maximum weight loss is observed during the third stage
(300-500°C). This observation suggests that the structure of the processed lignite underwent substantial degradation, creating (weak) C-O bonds that are more thermolabile than the C-
C bonds occurring in lignite.137 At the end of the fifth stage (500-800 °C), the overall weight loss for the unprocessed lignite is about 30%, whereas the weight loss for the lignite residues is in general over 50%.Table 4.2 summarizes the weight loss [%] found for the five stages observed during the thermal degradation of the lignite residue samples compared to the unprocessed lignite.
95 Solvent effects on the oxidation of lignite
Table 4.2. Comparison of weight loss in % observed for the thermal degradation of obtained lignite residues from experiment in different solvents compared to unprocessed lignite. Weight loss [%]
Stage 1 Stage 2 Stage 3 Stage 4 Stage 5 Residue Solvent Sample (30- (105- (200- (300- (500- [%] 105 °C) 200 °C) 300 °C) 500 °C) 800 °C)
Unproc. none 1.5 3.9 7.9 23.7 31.4 68.6 Lignite
MeOH 5 min 3.5 7.6 19.3 43.2 53.2 46.8
MeOH 10 min 3.5 8.6 21.0 47.5 58.3 41.7
MeOH 30 min 3.0 6.2 22.2 48.3 57.2 42.8
EtOH 5 min 3.9 9.6 21.8 44.8 55.4 44.6
EtOH 10 min 4.3 10.5 24.2 49.5 58.6 41.4
EtOH 30 min 3.2 9 24.2 54 60.7 39.3
Water 5 min 4.4 9.7 19.4 41.2 52.6 47.4
Water 10 min 4.4 9.8 19.9 43.3 54.7 45.3
Water 30 min 4.3 10 21.3 48.2 59.1 40.9
MeTHF 5 min 6.1 15.1 26.2 47.5 57.2 42.8
MeTHF 10 min 2.5 10.7 24.4 53.0 58.2 41.8
MeTHF 30 min 5.1 11.7 24.6 51.9 57.7 42.3
96 Solvent effects on the oxidation of lignite
a) b)
Weight [%] Weight
Temperature [°C] Temperature [°C] Figure 4.3: TG (a) and DTG (b) analysis of lignite residues obtained from reactions in methanol 3+ mediated by the Al /H2O2 system at 65 °C and various process durations (5-30 min).
a) b)
Weight [%] Weight
Temperature [°C] Temperature [°C]
Figure 4.4: TG (a) and DTG (b) analysis of lignite residues obtained from reactions in ethanol 3+ mediated by the Al /H2O2 system at 65 °C and various process durations (5-30 min).
97 Solvent effects on the oxidation of lignite
a) b)
Weight [%] Weight
Temperature [°C] Temperature [°C]
Figure 4.5: TG (a) and DTG (b) analysis of lignite residues obtained from reactions in water 3+ mediated by the Al /H2O2 system at 65 °C and various process durations (5-30 min).
a) b)
Temperature [°C]
Figure 4.6: TG (a) and DTG (b) analysis of lignite residues obtained from reactions in 2-Me-THF 3+ mediated by the Al /H2O2 system at 65 °C and various process durations (5-30 min).
98 Solvent effects on the oxidation of lignite
As aforementioned at the introduction of this chapter, the obtained soluble and precipitated products and were also briefly analyzed by elemental analysis (see Appendix,
Table 9.1). It was found that, the samples comprise low C- and H-content yet high N- (ca.
27%) and O-content (ca. 60%) of the precipitated products supports the assumption that most of the aluminum (III) was effectually precipitated by an ammonia solution. For the soluble products, the determined values for the C-content differed from about 7 to over 30% and very high O-content over 50-80%. Herein, the higher values for C-content were obtained for soluble products from reactions in ethanol and 2-Me-THF. These facts agree with the previously made observation that the conversion of lignite seems to be more efficient in ethanol and 2-Me-THF, forming higher amounts of soluble products.
4.2 Coal oxidation experiments in methanol-13C at 65 °C
To investigate whether methanol is inserted into lignite or lignite fragments, a set of experiments in 13C-labelled methanol were performed. Table 4.3 compares the results
3+ obtained from the experiments in the presence of the Al /H2O2 system with those performed in the absence of catalyst or H2O2. The elemental composition of lignite samples obtained from varying treatments are similar to those from large scale experiments carried out in unlabeled methanol (Table 4.1). This observation indicates that the chemical processes on lignite is not considerable affected by the experiment scale.
99 Solvent effects on the oxidation of lignite
Table 4.3: Initial weight of substrate and lignite residues and their elemental composition obtained from processing in 13C-methanol at 65 °C for 5 min. Experiment Initial Weight H C N S O weight residue [%] (by diff.) [g] [g] 1 Lignite ------5.2 67.0 0.8 1.4 25.6 ± 0.1 ± 0.5 ± 0.1 ± 0.1
2 With H2O2 0.2124 0.1952 4.5 57.3 0.6 1.2 36.3 ± 0.1 ± 1.1 ± 0.1 ± 0.6
3 With Al3+ 0.2030 0.2000 4.5 53.7 1.9 1.5 38.4 ± 0.1 ± 2.3 ± 0.1 ± 0.6
3+ 4 With Al /H2O2 0.2100 0.0675 5.2 55.1 0.6 1.1 38.0 ± 0.1 ± 0.7 ± 0.1 ± 0.1
To gain further information about the processes occurring on the chemical structure of the obtained lignite residues, FTIR and 13C CP-MAS NMR were also performed. The FTIR spectra of the residues compared to unprocessed lignite are presented in Figure 4.7. Here,
3+ 13 the spectrum for the experiments with just Al / CH3OH show no significant changes, e.g. the typical peak for aromatic C=C bonds at 1600 cm-1 is still very prominent. Whereas the
13 spectrum for the experiment with H2O2/ CH3OH reveals a decrease in aromatic compounds and an increase in C=O bonds (1700 cm-1). This peak becomes more intense in the sample
3+ of the residue obtained from reaction with the Al /H2O2 system.
100 Solvent effects on the oxidation of lignite
13 Figure 4.7: FTIR spectra of lignite residue obtained from experiments in CH3OH at 65 °C and 5 min reaction duration.
In addition, Figure 4.8 compares the 13C CP-MAS NMR spectra of the lignite residues and unprocessed lignite. Slight changes in the peak intensities, especially in the region for the aromatic compounds (140-110 ppm) can be observed for the spectrum 3
13 (H2O2/ CH3OH), where also a strong peak at about 50 ppm, typical for methoxy group, is clearly visible. This strong peak indicates a minor addition of 13C-methoxy groups to the lignite and lignite fragments during the reaction with H2O2.
Clearly, the insertion of 13C-methoxy groups is dramatically enhanced when aluminum nitrate is added to the reaction mixture. The 13C-NMR spectrum No. 4 shows that the peak at about 50 ppm markedly increased, compared to the experiment with only H2O2 added to the dispersion of lignite in methanol-13C. Moreover, the band at 140-110 ppm is almost completely vanished as indicated by its low intensity.
101 Solvent effects on the oxidation of lignite
4) Lignite,
13 Al + H2O2/ CH3OH
3) Lignite, 13 H2O2/ CH3OH
2) Lignite, 13 Al/ CH3OH
1) Lignite
13 13 Figure 4.8: C-NMR of lignite residue samples obtained from experiments in CH3OH at 65 °C and 5 min reaction duration. The peaks at about 0 and 100 ppm (indicated by asterisks) are spinning side-bands.
Another method to study the addition of methoxyl groups mediated by the
3+ [Al(H2O)6] /H2O2 system is thermogravimetric analysis coupled with mass spectrometry
13 (TGA-MS). Herein, the m/z 45 for CO2 was monitored in the analysis. The residue samples were slowly heated from 35 to 800 °C (10°C/min) with a flow of 60 mL of O2 per minute.
Figure 4.9 presents the obtained curves for the m/z 45 for the lignite residues and unprocessed lignite. For the unprocessed lignite, a low intensity curve was obtained, where
102 Solvent effects on the oxidation of lignite
a signal at about 300–450 °C appeared. This signal dramatically increases for the
3+ 13 experiment performed for 5 min in presence of the [Al(H2O)6] /H2O2 system in CH3OH.
13 The curve shown for the residue sample from the experiment in CH3OH and only H2O2 shows a slightly increased signal in this region, compared to the one of the unprocessed lignite.
Figure 4.9: TG-MS analysis of lignite compared to lignite residues obtained from reactions at 65 °C; data normalized by sample weight.
The performed analysis show that an insertion of methoxy groups is happening during the
3+ 138 reaction with the [Al(H2O)6] /H2O2 system. Finkelstein and Rosen already reported in
1980 that methanol reacts as a strong radical scavenger, trapping OH• and forming the more stable CH3O•. This observation will be further investigated and confirmed in chapter 6 by
EPR analysis.
103 Solvent effects on the oxidation of lignite
4.3 Hydrodeoxygenation (HDO) experiments of lignite residues obtained from reactions in methanol
The previous Chapter 4.1 showed that the properties of lignite can be highly improved by
3+ the [Al(H2O)6] /H2O2 system in presence of methanol as a reaction medium. These results motivated us to perform upgrade reactions, e.g. a hydrodeoxygenation process on the coal products. The aim was to cleave the still complex coal fragments into smaller molecules like phenols, benzenes, naphthenes, etc.
During the last decade, silica supported nickel phosphide catalysts (Ni2P/SiO2) have attracted attention in hydrotreating reactions for bio-oil products due to their high stability and active behavior.139 Especially studies about HDO processes with lignin model compounds, e.g. guaiacol, have attracted much interest.140, 141 Also, several studies about naphthalene hydrogenation using this kind of catalyst have been reported.139, 142 Since lignite is a low-maturated coal, it still comprises much oxygen44 which is further increased after
3+ using our introduced process with the [Al(H2O)6] /H2O2 system. Therefore, the Ni2P/SiO2 catalysts seemed to be the right choice for performing HDO experiments with lignite residues obtained from coal conversion experiments in methanol, to reduce the increased oxygen content and further cleave the complex coal fragments into single molecules.
To test the effectiveness of the Ni2P/SiO2 catalyst, blank experiments with unprocessed lignite and lignite residue obtained from reaction in methanol with the
3+ [Al(H2O)6] /H2O2 system after five minutes process duration were performed as well.
Figure 4.10 shows the 2D GCxGC image of the products obtained from a blank experiment with unprocessed lignite in octane and without catalyst (300 °C, 100 bar H2, 20 h). A small consumption about 14% of lignite (initial weight: 0.5 g) was observed. The
104 Solvent effects on the oxidation of lignite
products found, as indicated in Figure 4.10, where phenol, substituted phenols, acyclic
C8-ketones and C13-C18 hydrocarbons are detected as trace products.
Figure 4.10: Products obtained from blank experiment with unprocessed lignite and without catalyst. Reaction condition: 0.5 g lignite residue, 5 mL n-Octane, 300 °C, 100 bar H2, 20 h. Internal standard (ISTD): 8x10-5 mol n-hexadecane
3+ The second blank experiment was performed with processed lignite (Al /H2O2,
MeOH, 5 min), to show the impact of the prior benefication of coal. It was found that the comsumption for the processed lignite was improved more than two times to about 28%.
Although the processing of coal clearly enhanced the ability for further upgrading processes, the GCxGC image and products obtained are still very similar to the ones obtained from unprocessed coal: only Phenols, C8-ketones and C14-C18 hydrocarbons were found in the reaction mixture.
105 Solvent effects on the oxidation of lignite
Figure 4.11: Products obtained from blank experiment with processed lignite (Al/H2O2, MeOH, 65 °C, 5 min) and without catalyst. Reaction conditions: 0.5 g lignite residue, 5 mL n-octane, 300 °C, -5 100 bar H2, 20 h. Internal standard (ISTD): 8x10 mol n-hexadecane
The HDO reactions with Ni2P/SiO2 with unprocessed lignite and processed lignite
3+ obtained from previous reaction (Al /H2O2, MeOH, 65 °C, 5 and 30 min) resulted in more complex product mixtures (Figure 4.12 a,b,c). The variety in products increases with processing (reaction duration) of lignite. For the HDO reaction with unprocessed lignite
(Figure 4.12a), a consumption about 28% was found. Besides the recurring compounds, e.g. phenol and C8-ketones, the peaks identified by EI-MS show the presence of phenols with C2-3 chains or two methyl groups as substituents. In addition, the amount of unsaturated hydrocarbons was increased from C9-C25. Almost 50% conversion of processed lignite
3+ (Al /H2O2, MeOH, 65 °C, 5 min) was observed after the HDO reaction.
106 Solvent effects on the oxidation of lignite
Figure 4.12b shows a larger variety of aromatic compounds with different kind and levels of substituents as well as more saturated compounds. Among others, the main compounds identified were benzenes, phenols, naphthalenes and C9-C20+ saturated
3+ hydrocarbons. The HDO experiment with the second process lignite (Al /H2O2, MeOH,
65 °C, 30 min) showed slightly lower conversion (about 43%). This might be due to the higher ratio of mineral matter per gram lignite residue, compared to the residue obtained from shorter processing reaction in which less coal was converted to soluble products.
However, in Figure 4.12c it can be observed that HDO with processed lignite from longer reaction (30 min) results in more saturated compounds, e.g. linear and branched hydrocarbons as well as tetralin and decalin compounds.
107 Solvent effects on the oxidation of lignite
a)
b)
c)
3+ Figure 4.12: Products obtained from unprocessed lignite (a) and residue samples (Al /H2O2, MeOH, 65 °C): b) 5 min and c) 30 min process duration. Reaction conditions: 0.5 g lignite residue, -5 0.2 g catalyst, 5 mL n-octane, 300 °C, 100 bar H2, 20 h. Internal standard (ISTD): 8x10 mol n- hexadecane
108 Solvent effects on the oxidation of lignite
4.4 Conclusion
The results obtained from this study show clearly how the choice of the solvent can control the outcome of the product or in this case the degree of conversion of coal. It was presented that methanol and water used as a reaction medium, leads to lignite residue samples with most of the aromatic compounds still intact after 30 minutes reaction duration, whereas ethanol and 2-Me-THF lead to coal residue samples with almost none aromatic structure and low yields of solids left. The reason for this is the formation of different radical species during the reaction in the various solvents, which will be explored in more detail in chapter
6. Furthermore it was revealed that in general only five minutes of reaction duration in
3+ presence of the [Al(H2O)6] /H2O2 system is efficient enough to slightly change the chemical composition to positively affect the thermal degradation behavior of the residue samples.
Thus, the amount of volatile products was enhanced from about 31% (unprocessed lignite) to 50-60% for the processed samples. However, the most promising results were obtained from the experiment series performed in methanol since it is easy to remove, cheap and, as aforementioned, results in lignite residue samples with minor changes in chemical composition and nature and relatively high yields of solids.
Furthermore, the potential of silica supported nickel phosphide catalysts for hydrodeoxygenation (HDO) of unprocessed and processed lignite samples was explored.
The Ni2P/SiO2 catalyst demonstrated high effectiveness in HDO of processed lignite samples (ratio cat:coal = 1:2.5) with a conversion about 50% at 300 °C, 100 bar H2 and 20 h. GCxGC analyses of the soluble products revealed a wide range of benzenes, phenols, naphthalenes and C9-C20+ saturated hydrocarbons (unbranched and branched). Further, the results indicated that the amount of saturated compounds was increased for processed coal
109 Solvent effects on the oxidation of lignite
samples of 30 min processing. Noteworthy, full conversion can never be fully achieved due to comprised insoluble mineral matter in the coal samples.
However, these first results demonstrate that it is possible to cleave the complex structure of lignite into small molecules by just applying a 5-min-processing as an entry point process.
Certainly, HDO processes for the conversion of lignite still need to be optimized and the recyclability of the catalyst must be investigated. Nevertheless, this very first study appears to be very promising for the utilization of lignite, showing another option then its combustion in electric power plants.
110 Effect of coal rank
5 Effect of coal rank
In past years, various differences have been noted between high rank coals, particularly in dissolution rates and product yields with differences in weight yield and chemistry.In this
3+ chapter, we studied the effectiveness of the [Al(H2O)6] /H2O2 system in 2-Me-THF on more matured coals than lignite, e.g. “Gasflammkohle” (high volatile bituminous coal), “Fettkohle”
(medium volatile bituminous coal), “Magerkohle” (semi-anthracite) and “Anthrazit”
(anthracite). As a result, the comprehensive analysis of obtained product oil samples (yield
<0.1 g) showed that the oil samples are only a formation of poly-Me-THF from solvent degradation.
3+ Therefore, the focus of this chapter is a brief discussion if the [Al(H2O)6] /H2O2 system affects the coals’ structure at all and may have a positive effect on the properties, e.g. thermal degradation properties.
5.1 Analysis of obtained coal residue samples
Figure 5.1 to Figure 5.4 illustrate SEM pictures of the various coal types and their reaction residues obtained from reactions at 65 °C and 4 h in 2-MeTHF. No real changes in their appearance, physical structure, can be observed. All coal samples have smooth surfaces with smallish particles distributed on the surface, like their reaction residues. Only the SEM
111 Effect of coal rank
picture for the residue sample from the reaction with “Gasflammkohle” (Figure 5.1b) shows very small holes, indicating that some organic matter might be leached out during the process.
a) b)
Figure 5.1: SEM pictures of a) Gasflammkohle and b) its reaction residue
a) b)
Figure 5.2: SEM pictures of a) Fettkohle and b) its reaction residue
112 Effect of coal rank
a) b)
Figure 5.3: SEM pictures of a) Magerkohle and b) its reaction residue
a) b)
Figure 5.4: SEM pictures of a) Anthracite and b) its reaction residue
Table 5.1 summarizes the obtained weight of coal residues, their elemental composition and ash content. The coal residue samples were washed three times with 10 mL of water to remove remaining aluminum nitrate and dried over 48 h at 60 °C in a vacuum oven and kept under argon atmosphere. The list shows, that the obtained weight for the
“Gasflammkohle”-residue (0.9 g) after reaction much higher than the initial weight of 0.5 g.
The high weight, N-content (3.8%), the vastly increased amount of oxygen (30.3%) and high ash content strongly indicate the presence of remaining aluminum nitrate attached to the coal surface. Moreover, the C-content was decreased from 81 to 60.5% for the residue
113 Effect of coal rank
sample. Its sulfur content was also significantly reduced to more than the half (from 2.1 to
1%), also the H-content was slightly changed to about 4.4%. For the other coal residual samples, the weight of residue were between 0.5–0.6 g. For them, only minor changes in the elemental composition were found. The C-content found was decreased (84–86%) compared to the unprocessed coals (87–94%). The values determined for the H-content also did not change (5– 3%). Again, the residue samples show increased amounts of oxygen and also slightly increases nitrogen content, indicating the presence of more oxidized compounds and remaining catalyst. The N-content was increased at about 0.5-0.6% for all residuals, and the O-content was enhanced about 2-3% for entry 2 and 3, but over 9% for entry 4. To find out the contribution of remaining aluminum salt to the ash content, the Al2O3 content was determined (Eq. 3.5) as well. The ash content cannot fully be attributed to the presence of adsorbed Al(NO3)3 on coal residues, except for entry 3. It seems that the coals
3+ structures were indeed slightly oxidized after the process with the [Al(H2O)6] /H2O2 system for 4 h at 65 °C.
114 Effect of coal rank
Table 5.1: Yields and elemental composition (daf) of coal residue samples obtained from reactions with different coals, compared to the elemental composition (daf) of unprocessed coal
Entry Reaction Mass C H N S Odiff. Ash Ash as
No. residue residue [%] [%] Al2O3 obtained from [g] [%]
1 Gasflammkohle 0.9 60.5 4.4 3.8 1.0 30.3 12.2 4.6 ± 0.6 ± 0.1 ± 0.1 ± 0.1
2 Fettkohle 0.6 84.2 4.4 2.1 1.4 7.9 0.4 2.5 ± 0.7 ± 0.1 ± 0.1 ± 0.1
3 Magerkohle 0.5 86.2 3.7 2.2 1.3 6.6 2.9 2.7 ± 0.1 ± 0.1 ± 0.1 ± 0.1
4 Anthracite 0.6 84.1 3.1 2.0 1.4 9.4 5.0 2.4 ± 1.3 ± 0.1 ± 0.1 ± 0.1
Unprocessed coals
5 Gasflammkohle -- 80.7 5.1 6.0 2.1 6.2 4.3 -- ± 0.4 ± 0.1 ± 0.1 ± 0.1
6 Fettkohle -- 86.6 5.0 1.4 1.6 5.5 4.1 -- ± 0.6 ± 0.2 ± 0.1 ± 0.1
7 Magerkohle -- 89.5 4.1 2.1 0.6 3.8 7.7 -- ± 0.4 ± 0.1 ± 0.2 ± 0.2
8 Anthracite -- 93.9 3.4 1.6 0.7 0.3 6.8 -- ± 0.8 ± 0.1 ± 0.1 ± 0.1
Furthermore, the 13C NMR spectra (Figure 5.5-1) for the coal residues only reveal
significant changes in the spectrum for the residue sample of “Gasflammkohle”. The new
peaks in the region of aliphatic C-O-bonds (90–55 ppm) and carboxylic functions (170-
160 ppm) are revealed, while the spectra for “Fett-, Magerkohle” and Anthracite (1b-d) show
115 Effect of coal rank
the same three main peaks, around 60-30 ppm (carbohydrates), 128 ppm (aromatic C) and
205 ppm (ketone and aldehyde C), which have been also observed for the spectra of their
3+ corresponding coals (Figure 5.5- 2b-d). On first sight, this indicates that the [Al(H2O)6] /H2O2 system seems to be less effective for the conversion of high rank coals under used reaction conditions, compared to the conversion experiments with lignite.
2 1 a)
b)
c)
d)
Figure 5.5: 13C CP-MAS NMR spectra for coal residue samples (1), and unprocessed coals (2): a) Gasflammkohle, b) Fettkohle, c) Magerkohle, d,) Anthracite
To investigate, if the reaction process affects the thermal degradation behavior of the different coal samples, TG analyses were performed as well. Figure 5.6a and b show the obtained TG and their corresponding DTG curves. Table 5.2 summarizes and compares the
116 Effect of coal rank
weight loss [%] of the coal residue samples obtained at different stages and the remaining solid residues [%] after the measurements for the processed, compared to their corresponding unprocessed coals (Figure 5.7).
The TG and DTG curves of “Gasflammkohle” residue sample show the biggest weight loss of all coal residue samples. Its DTG curve indicates a four-stage thermal degradation. For better comparison, those four stages were also employed for the other coal residue samples, although they consist of only two to three stages devolatilization.
During the the first stage (30-125 °C), low volatile compounds are realeased which is mainly physically absorbed water within the sample. The residue samples for “Gasflammkohle” and
Anthracite show a weight loss [%] about 4.3 and 1.6% within this temperature range. The other residue samples show almost none weight loss. At the end of the second stage
(330 °C), the residue sample obtained from “Magerkohle” still reveals no real weight loss
(~0.1%) whereas the samples of “Fettkohle” and “Anthracite” lose 3-5% weight, and the residue from “Gasflammkohle” about 23%. During the third stage (330-640 °C), the weight loss for the “Gasflammkohle” residue samples is almost doubled to about 44%. Also the residue sample from “Magerkohle” starts to degrade slowly about 450 °C and reaches a loss about 5% at 640 °C. Furthermore, the TG and DTG curve for “Fettkohle” residue sample clearly shows a bigger devolatilization about 420 °C to 640 °C. Herein, its total weight loss is about 16% at 640 °C. The loss for Anthracite residue sample within this stage was found to be about 5%.
At the end of the measurement (800 °C), it was found that the thermal degradation of the residue samples from “Gasflammkohle” and “Anthracite” could be enhanced about 6-15%, whereas the residues obtained from “Fett- and Magerkohle” showed no real changes compared to the corresponding unprocessed coals.
117 Effect of coal rank
Table 5.2: Weight loss [%] of different coal ranks (unprocessed and processed) obtained from TG analysis
Weight loss [%]
Stage 1 Stage 2 Stage 3 Stage 4 Residue [%] Sample (30 – (125 – (330 – (640 – 125 °C) 330 °C) 640 °C) 800 °C)
Gasflammkohle 4.3 22.5 43.5 46.8 53.2
Fettkohle 0.5 3.1 16.1 19.9 80.1
Magerkohle 0 0.1 5.2 8.2 91.8
Anthracite 1.6 4.9 9.2 12.6 87.4
Unprocessed coals
Gasflammkohle 2.1 3.3 26.8 32.4 67.6
Fettkohle 0 0 15 20 80
Magerkohle 0 0 5.3 9.0 91.1
Anthracite 0 0 2.5 6.1 93.9
118 Effect of coal rank
a) b)
Figure 5.6: TGA (a) and DTG (b) curves of coal residue samples of different coal ranks obtained 3+ from reaction with the [Al(H2O)6] /H2O2 at 65 °C and 4 h.
a) b)
Figure 5.7: TGA (a) and DTG (b) curves of unprocessed coal of different ranks.
119 Effect of coal rank
5.2 Conclusions
3+ In this chapter, the effect of the [Al(H2O)6] /H2O2 system at 65 °C and 4 h process duration was investigated in presence of high matured coal samples. The experiments showed that the low yields of oily product samples obtained from the reaction consists of 2-Me-THF degradation products. Moreover, only slight changes in chemical composition were determined for the obtained coal residue samples, with the most significant changes found for the residue sample of “Gasflammkohle”, a high volatile bituminous coal. Furthermore, the analyses of the residue samples from “Fett- and Magerkohle” revealed no considerable changes either in chemical composition or thermal degradation behavior. Surprisingly, it was found that the oxygen content of anthracite was increased from 0.3 to over 9%. Its pyrolysis showed an improved behavior about 6% compared to the unprocessed sample. However,
3+ this study showed that the [Al(H2O)6] /H2O2 system shows less impact on the degradation of the condensed aromatic ring structure, yet was able to slightly improve the thermal degradation properties by forming more low volatile compounds.
120 Activation of H2O2 – EPR spectroscopy investigations
6 Activation of hydrogen peroxide by the 3+ [Al(H2O)6] /H2O2 system - Part I:
Evaluation of the reaction mechanism by Electron Paramagnetic Resonance (EPR) spectroscopy
Interaction between hydrogen peroxide (H2O2) and aluminum salts displays an important role in hydroperoxidation reactions. However, the mechanism of radical formation remains uncertain. According to Rinaldi et al.100, controllable formation of HO· and HOO· radicals is possible in cyclic ether solvents (e.g. THF) through the decomposition of –OOH species
3+ formed by a mechanism involving the second coordination sphere of [Al(H2O)6] (Figure
6.1). The aluminum-coordinated water exchanges slowly with the water molecules in the second coordination sphere. In this chemical environment, an interaction between the
3+ [Al(H2O)6] and H2O2 through hydrogen bonds takes place, resulting in a proton exchange.
In contrast, Shul’pin et al.143 proposed that the HOO· radical bonds directly to the Al3+ species and reacts further into two hydroxyl radicals in acetonitrile.
121 Activation of H2O2 – EPR spectroscopy investigations
3+ Figure 6.1: Proposed mechanism of the [Al(H2O)6] catalyzed epoxidation of α,β-unsaturated ketones by Rinaldi et al.100
The current results obtained from coal conversion strongly suggest that the lignite structure is converted by radical oxidation. Additionally, an incorporation of the cyclic ether solvent seems to take place. In this Chapter, the mechanism involving HO· and HOO· radicals in coal conversion was examined by Electron Paramagnetic Resonance (EPR) measurements. EPR spectroscopy is selective and sensitive for the characterization of paramagnetic species. The following results and discussion show how the aluminum salt activates hydrogen peroxide, including the reaction of radical and coal.
The free radicals HO· and HOO· have a short life-time. This feature poses a problem for the direct detection of these radicals at room temperature. To overcome this problem, an analytical technique called spin trapping is employed for the detection and identification of the short-lived radicals through the use of EPR at room temperature.144 The most common spin trap reagent is DMPO (5,5-dimethyl-1-pyrroline N-oxide). It forms long-lived radical
122 Activation of H2O2 – EPR spectroscopy investigations
adducts with O-, C-, N-, S-centered radicals. Figure 6.2 shows the characteristic product radical formed by the reaction of oxygen radicals and spin trap reagents. It involves the addition of the reactive free radical, here HO· and HOO·, across the double bond of a diamagnetic spin trap to form the more stable free radical adduct of DMPO. Due to the short lifetime, the detection of DMPO/·OOH is not without problems. In this manner, DEPMPO (5-
(diethoxyphosphoryl)-5-methyl-1-pyrroline-N-oxide) is often the best choice for HOO· radicals and HO· to be detected and identified by EPR.
a) b) DEPMPO
DMPO/·OOH
DMPO
DMPO/ ·OH DEPMPO/ ·OOH
Figure 6.2: a) EPR spin trapping of superoxide and hydroxyl radical with DMPO and the decay from DMPO/·OOH product to DMPO/·OH; b) EPR spin trapping of superoxide with DEPMPO.
The activation of H2O2 by aluminum salt was investigated in absence of solvent and coal, so that the effect of solvent and the overall interaction with coal can be compared and elucidated. Initial EPR studies indicated a very fast reaction of DMPO and 70 wt% H2O2 at
65 °C. Additionally, oxygen bubbles interfered the EPR measurements. Thus, the concentrated solution of H2O2, which is often used for the coal conversion, was diluted ten times for the experiments performed at 55 °C.
123 Activation of H2O2 – EPR spectroscopy investigations
Since minor decomposition of H2O2 about 1-2% per year can be still observed despite added
145 stabilizers in commercially available H2O2 solutions , a blank EPR experiment was performed to check whether radicals can already be observed in absence of any activators.
Therefore, only the H2O2 solution (7 wt%) was given into the EPR flat cell. After injecting the
DMPO spin trap, the formation of the radical adducts and their time evolution was monitored.
Figure 6.3 shows the obtained EPR spectrum. The EPR spectrum of DMPO-OH consists of a characteristic 1:2:2:1 quartet138 and can be easily compared and identified by data provided from literature or from simulation experiments using EPRSim3 (Figure 6.4b). A low but stable signal intensity was observed which indicates low concentration and formation of hydroxyl radicals in the H2O2 solution, which are not consuming the DMPO spin trap.
Nevertheless, the presence of hydroxyl radicals in H2O2 solution in absence of an activator was confirmed.
Figure 6.3: EPR spectrum of formed DMPO-OH product from a control experiment performed on a H2O2 (70 wt%) solution in H2O.
124 Activation of H2O2 – EPR spectroscopy investigations
6.1 EPR investigations into the [Al(H2O)6]3+ system
Next, the effect of aluminum nitrate as an activator for H2O2 was studied. Figure 6.4a illustrates the EPR spectra from the experiment performed on a solution of 1 mmol aluminum nitrate dissolved in 7 wt% H2O2 at room temperature. In contrast to the blank experiment, a fast increase in the peak intensity is noticeable after adding the DMPO. Interestingly, the intensity of the EPR signal decreases with time due to the decomposition of the radical adducts. This indicates the ongoing formation of new radicals which are fully consuming the
DMPO reagent. The same observations were made for an experiment using the same reaction mixture at 55 °C. Only herein, the decomposition of the DMPO-OH adducts was even faster due to the increased formation of the radical species.
a) b)
Figure 6.4: a) EPR-spectra of the DMPO-OH adduct obtained from the experiment with 1 mmol Al(NO3)3∙9H2O dissolved in an aqueous solution of H2O2 (7 wt%) at room temperature; b) Comparison of EPR data obtained by experiment and from computer simulation for DMPO-OH adduct.
125 Activation of H2O2 – EPR spectroscopy investigations
Moreover, an experiment with spin trap DEPMPO at room temperature was carried out due to its higher stability as well to investigate the formation of HO· and HOO· radicals or the DEPMPO adducts, respectively. Figure 6.5 shows the obtained EPR spectra in the same radical generating system as used for the experiment with DMPO. The spectra shows a higher signal/noise ratio of DEPMPO adducts compared to the spectra with DMPO adducts under identical conditions, due to the higher stability of DEPMPO spin adducts.
Once again, only the formation of DEPMPO-OH spin trap products is revealed. This is most probably due to the diluted solution of hydrogen peroxide which minimizes the quantities of other reactive species present, as the perhydroxyl radical.146 The perhydroxyl radical is the dominant form of superoxide present under acidic pH conditions, e.g. in the typical Fenton’s system conducted at pH 3. However, the superoxide anion seems to be only available in appreciated concentrations when the pH is above 4.8.147
Figure 6.5: EPR-spectra of the DEPMPO-OH adduct obtained from the experiment with 1 mmol Al(NO3)3∙9H2O dissolved in an aqueous solution of H2O2 (7 wt%, 1 mL) at room temperature.
126 Activation of H2O2 – EPR spectroscopy investigations
In addition, an EPR experiment was performed with the commercially available hydrogen peroxide (70 wt%). Figure 6.6 reveals the fast formation of different DMPO adducts after adding the spin trap to the reaction mixture and their fast decomposition. The spectra show an overlay of signals for DMPO-OH and DMPO-OOH products, which was confirmed by spectra simulation (Figure 6.7). For better visualization and inspection, one spectrum is presented in Figure 6.7. In addition, Table 6.1 lists the spin Hamilton parameters for the spin trap adducts DMPO-OH and DMPO-OOH. The g values are in the range of 1.99 and 2.01 which are typical for organic radicals.148 Furthermore, the hyperfine coupling constants are listed as well as the acquired values for the resonance line width. All those data are in good agreement with data reported in literature and confirmed the formation of HO· and HOO· radicals in the reaction medium.146, 149, 150
DMPO-OH
DMPO-OOH
Figure 6.6: EPR spectra of DMPO-OH and DMPO-OOH adducts obtained from an experiment with
70 wt% H2O2 (1 mL) and 1 mmol Al(NO3)3∙9H2O at room temperature.
127 Activation of H2O2 – EPR spectroscopy investigations
b)
a) c)
Figure 6.7: a) EPR spectrum of DMPO-OH and DMPO-OOH products at 1 min, and corresponding simulated EPR spectra of b) DMPO-OH product and c) DMPO-OOH product.
Table 6.1: Spin Hamilton parameters of radical species observed during the reaction. g value coupling line width constants B A /G
DMPO-OH 2.0056 AN=14.7G B=1.4G
AHβ=13.9G
AN=13.9G DMPO-OOH 2.0057 B=1.1G AHβ=11.2G
AH=1.2G
128 Activation of H2O2 – EPR spectroscopy investigations
6.2 EPR studies of the [Al(H2O)6]3+/H2O2 system in organic solvents
6.2.1 2-Methyl-tetrahydrofuran
Rinaldi et al.100 showed that an abstraction of hydrogen from the C-2 position (for 2-Me-THF,
C-4) of THF by HO· radical leads to a radical 2-tetrahydrofuranyl which can be trapped by diethyl maleinate (Scheme 6.6.1).
Scheme 6.6.1: Addition of 2-tetrahydrofuranyl to diethyl maleinate.100
EPR experiments were performed to confirm the formation of this radical, likewise investigate the possible presence of other radical species in other solvents. Figure 6.8a presents the fast formations of a DMPO adduct which is stable over a time period of 10 min.
3+ Unlike the results from [Al(H2O)6] /H2O2 in water, which show that HO· and HOO· radicals are formed in the reaction mechanism, only a conformer for the DMPO-OOH adduct and no
DMPO-OH adduct can be observed, indicating that HOO· is the most energetic radical present in the 2-Me-THF medium. Nonetheless, there is another present in this spectrum which corresponds to the adduct of DMPO with 2-Me-THF. This information confirm that 2-
Me-THF (or THF) act as a scavenger radical for only HO· radicals via an abstraction of
129 Activation of H2O2 – EPR spectroscopy investigations
hydrogen.100 Figure 6.8b shows the proposed mechanism for the formation of the 2- tetrahydrofuranyl and its addition to DMPO spin trap.
b) a)
Figure 6.8: a) Experimental EPR spectrum of DMPO-OOH150 and DMPO-2-Me-THF adducts acquired from an experiment with 70 wt% H2O2 (0.1 mL) and 0.1mmol Al(NO3)3∙9H2O in 0.8 mL 2- MeTHF at room temperature; b) proposed mechanism for the formation of the 2-Me-THF radical and its addition to DMPO.
Since the reaction with concentrated hydrogen peroxide solution (70 wt%) was to fast at
55 °C, the measurements were performed with diluted hydrogen peroxide (7 wt%). Herein, a fast formation of DMPO-OH and DMPO-OOH products could be observed as well as the radical adduct of DMPO with 2-Me-THF (Figure 6.9). After 5 min of monitoring, a degradation product of DMPO was also detected. According to literature and computer simulations
(Figure 6.10), those signals refer to the two conformers of a ring-opened structure of DMPO.
The decomposition product is shortly named “MNPA” (4-methyl-4-nitroso-pentanoic acid)
130 Activation of H2O2 – EPR spectroscopy investigations
and its formation is presented in Scheme 6.6.2. It is reported that the lifetime of the DMPOX is dependent on the presence of dioxygen species. If those are absent, the DMPOX will decay slowly into several decomposition products.151
DMPO DMPOX MNPA
Scheme 6.6.2: Proposed reaction mechanism for the oxidative ring opening of spin trap agent DMPO.151
Figure 6.9: EPR spectra obtained from an experiment with 0.1 mL H2O2 (7 wt%), 0.1 mmol Al(NO3)3 9H2O in 0.8 mL 2-Me-THF at 55 °C.
131 Activation of H2O2 – EPR spectroscopy investigations
Figure 6.10: Simulated EPR spectra for different DMPO species observed during the reactions.
6.2.2 Methanol
Coal conversion experiments performed in different solvents (Chapter 4) revealed a milder oxidative degradation of lignite, especially for the experiments performed in methanol, where a soluble product, containing aromatic compounds, was obtained, was obtained.
Finkelstein and Rosen138 showed that solvents, e.g. methanol and ethanol, have an inhibitor effect in radical reactions, as they trap the HO· radical, forming low energy radicals, e.g. α- hydroxymethyl or –ethyl radicals.
Figure 6.11 shows the spectra acquired from the experiment in methanol at 55 °C. At the very left and right side of the spectra the signal for the DMPO-hydroxymethyl (∙CH2OH) radical adduct is shown, whereas in the middle the DMPO-OH and –OOH signals and their
132 Activation of H2O2 – EPR spectroscopy investigations
evolution over time are displayed. Comparing the evolution of the DMPO spin products, it seems that the formation of the DMPO-hydroxymethyl radical is more pronounced. In addition, blank experiments without catalyst were performed with 7 and 70 wt% H2O2 in methanol at 55 °C to confirm that the aluminum nitrate initiates the formation of the various radical species. No formation of DMPO adducts were observed for both blank experiments.
Figure 6.11: EPR spectra obtained from an experiment with 0.1 mL H2O2 (7 wt%), 0.1 mmol Al(NO3)3∙9H2O in 0.8 mL methanol at 55 °C.
133 Activation of H2O2 – EPR spectroscopy investigations
6.2.3 Comparison of 2-Methyl-tetrahydrofuran and methanol as
solvents used for reactions with the [Al(H2O)6]3+/H2O2 system
3+ From the investigation and comparison (Figure 6.12) of the [Al(H2O)6] /H2O2 system in 2-
Me-THF and methanol, the following conclusion are drawn:
3+ 1) The presence of HO· and HOO· radicals in the [Al(H2O)6] /H2O2 and 2-Me-THF
indicates that the reaction of these species and 2-Me-THF (solvent) is a slow
reaction. Moreover, this reaction takes place to a short extent, generating 2-Me-THF
radicals.
2) In methanol, the most abundant radical species is the ·CH2OH radical. However, the
presence of HO· and HOO· radicals, in lower yields than ·CH2OH, is also detected.
a)
b)
Figure 6.12: Comparison of EPR spectra obtained from reaction experiment with 0.1 mL H2O2 (7 wt%), 0.1 mmol Al(NO3)3∙9H2O in 0.8 mL a) methanol and b) 2-Me-THF at 55 °C.
134 Activation of H2O2 – EPR spectroscopy investigations
6.3 EPR studies of the conversion of coal with Al(H2O)6]3+/H2O2 system in 2-Methyl-tetrahydrofuran
6.3.1 Lignite
In order to investigate the evolution and behavior of the radicals in the presence of solvent and lignite, three series of experiments were carried out at room temperature, 55 °C and
65 °C. At room temperature, a very slow conversion of DMPO adducts were observed, whereas at 55 °C the reaction was very fast. In this case, no data could be recorded due to too fast reactions which severely disturbed the EPR tuning.
Figure 6.13 shows the quartz cell used for the measurements before (a) and after the reaction (b) at ca. 55 °C. First, lignite was dispersed in 2-Me-THF (ca. 0.3 mL) into the cell.
3+ 3+ The solution of [Al(H2O)6] /H2O2 (50µL of 1 mmol Al in 1 mL 70 wt% H2O2) was added via a capillary when the temperature of 55 °C was reached. Almost full conversion of coal into soluble products was achieved.
Figure 6.13: Quartz cell with a) dispersion of lignite in 2-Me-THF before reaction and b) after 3+ addition of [Al(H2O)6] /H2O2 and reaction at 55 °C for 20 min.
135 Activation of H2O2 – EPR spectroscopy investigations
The EPR spectra for the experiment at 55 °C are displayed in Figure 6.14. At the beginning, only the signal for a carbon radical of lignite can be observed. This signal rapidly decreases
3+ upon the addition of the [Al(H2O)6] /H2O2 solution, indicating an immediate reaction of carbon radicals of lignite and HO· and/or HOO· radicals. Furthermore, the following addition of DMPO reveals the formation of several DMPO adducts.
MNPA
2nd addition of DMPO
1st addition of DMPO
C-radical (lignite)
Figure 6.14: EPR spectra for the reaction with lignite in 2-Me-THF mediated by the 3+ [Al(H2O)6] /H2O2 system at 55 °C for 20 min.
Since it is possible to obtain different conformers of each DMPO adduct, additional spectra simulation was performed to identify the obtained peaks. Herein, the various spectra for
DMPO-OH, DMPO-OOH, DMPO-CRx (from coal and/or solvent) and the DMPO decomposition product MNPA, were simulated. Figure 6.15 shows the simulated EPR
136 Activation of H2O2 – EPR spectroscopy investigations
spectra which agree with the experimental data. According to the simulation, the EPR signals obtained from experiments’ complex mixture of the 1st conformer of DMPO-OH and
DMPO-OOH, both conformers of the DMPO decomposition product MNPA and a carbon radical (DMPO-CRx). To confirm this assumption, the data of the various single spectra were simulated together, resulting in an EPR signal identical to the EPR signal obtained from the reaction with coal.
Table 6.2 lists the spin Hamilton parameters for all the identified radicals and DMPO- adducts, respectively. All the presented values are in good agreement with data shown in literature.146, 149-152
Figure 6.15: Simulation of EPR spectra obtained under reaction conditions with five different DMPO species.
137 Activation of H2O2 – EPR spectroscopy investigations
Table 6.2: Spin Hamilton parameters of the radicals observed during the reactions g value coupling constants line width A /G B C-radical (lignite) gǁ=2.004 - Bǁ =3.5G g=2.004 - B =10.3G
DMPO-OH 2.0056 AN=14.7 G, AHβ=13.9G B=1.4G
DMPO-OOH 2.0057 AN=13.9 G, AHβ=11.2G B=1.1G
AH=1.2 G
DMPO-CRx 2.0055 AN=15.5 G, AHβ=21.1G B=1.6G
AH=0.2G
MNPA 2.0055 AN=14.4 G B=1.4G Conformer 1
MNPA AN=14.9 G B=2.0G Conformer 2
6.3.2 Anthracite
The coal conversion experiments performed on different coal ranks showed that conversion decreases with the coal ranking (i.e. increasing maturity time). This observation holds particularly true for the conversion of anthracite, a highly aromatic and condensed coal. In this case, the product oil mostly comprised oily products from the 2-Me-THF degradation and also the aromatic compounds of the obtained coal residue remained mostly untouched.
To shed light on the reaction of anthracite and the radicals generated by the
3+ [Al(H2O)6] /H2O2 system, EPR studies on the anthracite conversion was performed.
Figure 6.16 shows the EPR spectra obtained from the reaction with anthracite in 2-Me-THF
3+ and [Al(H2O)6] /H2O2 at 55 °C. The very strong signal of the C-radical signal dominates the
138 Activation of H2O2 – EPR spectroscopy investigations
3+ spectra. This signal does not decrease upon addition of the [Al(H2O)6] /H2O2 solution, confirming the very stable chemical structure of anthracite. Furthermore, a fast formation of
DMPO adducts with the HO· and HOO· radicals are observed. Under the experiment conditions, the DMPO-adducts are quickly decomposing into MNPA products.
Figure 6.16: EPR spectra obtained from a reaction with anthracite in 2-Me-THF mediated by the 3+ [Al(H2O)6] /H2O2 system at 55 °C for 10 min; Radical species trapped by DMPO.
139 Activation of H2O2 – EPR spectroscopy investigations
6.3.3 EPR studies of the reaction with coal and the [Al(H2O)6)3+/H2O2 system in methanol
Due to the observations made in previous Chapters 4 and 6.2, the oxidation and conversion
3+ of coal mediated by the [Al(H2O)6] /H2O2 system proceeds milder in methanol as a solvent than in THF or 2-Me-THF. It was revealed that the obtained coal residues still comprise a high amount of aromatic structure, which is more accessible for further upgrading processes.
3+ As aforementioned, EPR studies of the [Al(H2O)6] /H2O2 system in presence of methanol showed that the formed hydroxymethyl radical is the most abundant radical species and only a low concentration of the hydroxyl and perhydroxyl radicals were detected. The purpose of this study was to observe the interaction of the formed radical species with coal.
Therefore, the experiments were performed first under coal conversion conditions with 70 wt% hydrogen peroxide and afterwards with lower concentration of H2O2 (7 wt%) and catalyst to slow down the reaction for better investigations. Figure 6.17 compares the obtained spectra from those EPR experiments, and reveals that lignite undergoes immediate radical oxidation. The strong carbon signal stemming from coal which can be observed at
3+ the start is directly fading away due to the addition of the [Al(H2O)6] /H2O2 solution. In addition, both experiments show the fast formation of the different radical species with hydroxymethyl radical being in the majority. For the reaction with lower concentration of
3+ [Al(H2O)6] /H2O2 (Figure 6.17b), the built DMPO-adducts are stable over a period of several minutes, indicating a slower formation of new radicals, whereas in Figure 6.17a, the MNPA degradation product is rapidly formed. For the reaction under normal coal conversion conditions (Figure 6.17a), a fast and abundant formation of hydroxymethyl radicals was detected.
140 Activation of H2O2 – EPR spectroscopy investigations
MNPA a)
b)
DMPO-CRx
DMPO-OH DMPO-OOH
Figure 6.17: EPR spectra obtained from experiments performed with lignite in methanol with a) 1 mmol Al(NO3)3·9 H2O and with 70 wt% H2O2 and b) 0.1 mmol Al(NO3)3·9 H2O and 7 wt% H2O2 at 55 °C. Radical species trapped by DMPO.
141 Activation of H2O2 – EPR spectroscopy investigations
6.4 EPR studies of the [Al(H2O)6]3+/H2O2 system formed from aluminum perchlorate
The activation of hydrogen peroxide by different aluminum (III) salts for the various kinds of reactions have been the ongoing subject of studies for several years now. Among others, it was reported that for instance aluminum perchlorate nonahydrate (Al(ClO4)3·9H2O) is an efficient catalyst for the epoxidation of α,β-unsaturated ketones.100 Assuming the hypothesis that the activation of H2O2 occurs via a mechanism involving the second-coordination sphere, it is also important for the understanding of the process, to investigate the effect of
3+ the anion on the reactivity of the [Al(H2O)6] /H2O2 system. Therefore, our EPR studies were further extended. First, the activation of hydrogen peroxide on the formation of radical species in absence of solvents was investigated, followed by measurements with solvent,
3+ and lastly, the effect of this [Al(H2O)6] /H2O2 system on the conversion of coal.
6.4.1 EPR investigations into the activation of hydrogen peroxide by
the [Al(H2O)6]3+/H2O2 system formed from aluminum perchlorate
The experiments were performed under the same conditions used for the studies described in Chapter 6.1. Since it was already known that EPR measurements at reaction temperatures about 55 °C with high concentrated hydrogen peroxide solution (70 wt%) were difficult to measure. Therefore, the EPR experiments of the current study were also performed with diluted hydrogen peroxide solutions (7 wt%).
142 Activation of H2O2 – EPR spectroscopy investigations
Figure 6.18 shows EPR spectra obtained from the kinetic study of aluminum perchlorate (1 mmol) dissolved in hydrogen peroxide (7 wt%), and the evolution of different radical species trapped by DMPO over 20 minutes at 55 °C. After adding the spin trap, the typical four signal spectrum of DMPO-OH adduct can be observed. These signals are accompanied by weak, but steady growing signals caused from the DMPO degradation product “DMPOX” (Scheme
6.6.2). For better investigation, Figure 6.19 shows the extracted spectra obtained after 40,
300, 600 and 1200 s from Figure 6.18. At the beginning of the measurement (40 s) the
DMPO-OH signal are clearly dominating over the weak signals stemming from DMPOX.
However, after 300 s the intensities of the signals changes into the opposite situation with
DMPOX signal ruling the spectrum. After 600 s, the DMPO-OH signal has almost completely vanished, whereas the DMPOX signal remains stable until the end of the measurement
(1200 s).
3+ Figure 6.18: EPR spectra obtained from the [Al(H2O)6] /H2O2 system (1 mmol Al(ClO4)3, 1 mL 7 wt% H2O2) at 55 °C for 20 min; Radical species trapped by DMPO. (Straight lines: measurement out of tune)
143 Activation of H2O2 – EPR spectroscopy investigations
1 DMPO X
DMPOX2
1200 s
600 s
300 s
DMPO-OH 40 s
3+ Figure 6.19: EPR spectra extracted from Figure 6.18 from [Al(H2O)6] /H2O2 system (1 mmol Al(ClO4)3, 1 mL 7 wt% H2O2) at 55 °C for 20 min; Radical species trapped by DMPO.
EPR spectra simulation using EPRSim3, confirmed the presence of DMPO-OH and
DMPOX radical adducts. The spin Hamilton parameters for the radical species are given in
Table 6.3.
Table 6.3: Spin Hamilton parameters of radical species observed during the reaction. g value coupling line width constants B A /G
DMPO-OH 2.0063 AH=AN=14.9 G B=1.1 G
DMPOX 2.0074 AN=7.4 G B=0.6 G
2x AHβ= 4.0 G
144 Activation of H2O2 – EPR spectroscopy investigations
An additional EPR experiment performed with aluminum perchlorate and hydrogen peroxide at room temperature showed low concentration of DMPO-OH adducts. A slow formation of DMPOX degradation product was observed after several minutes as well (s.
Appendix, Figure 9.1). These results indicate that only HO· radicals are formed at room temperature and also at temperatures as high as 55 °C.
6.4.2 The [Al(H2O)6]3+/H2O2 system formed from aluminum perchlorate in the presence of tetrahydrofuran as the solvent
The EPR experiment with aluminum perchlorate dissolved in the aqueous hydrogen peroxide solution showed that different radical species are present and a totally different mechanism seems to take place compared to the activation of H2O2 with aluminum nitrate
(Al(ClO4)3: HO·, Al(NO3)3: HO· and HOO· radicals). Therefore, to gain more information on the system reactivity, the system was also studied in presence of a solvent. (e.g. THF). In
Figure 6.20, the EPR spectra for the reaction over 20 min is presented. It can be observed that after the addition of the spin trap, strong EPR signals for DMPO-OH adduct are present
(four most intense signals). These are again accompanied by the DMPOX degradation product, but in lower concentration compared to the experiment without solvent. This is due to the formation of solvent-radicals (first signal from the left and right) which are also present in every single spectrum. At the beginning, the signals for the DMPOX appear relatively strong; however, they decreased after several minutes, as shown in Figure 6.21. After 600 s the formation of the several radical species seems to be equilibrated since the signal intensities do not change until the end of the measurement.
145 Activation of H2O2 – EPR spectroscopy investigations
3+ Figure 6.20: EPR spectra obtained from the [Al(H2O)6] /H2O2 system (0.1 mmol Al(ClO4)3, 0.1 mL 7 wt% H2O2) in 0.8 mL THF at 55 °C for 20 min; Radical species trapped by DMPO.
DMPO-OH DMPO-CRx DMPO-CRx (DMPO-MeTHF)
3+ Figure 6.21: EPR spectra extracted from Figure 6.20 from the [Al(H2O)6] /H2O2 system (0.1 mmol Al(ClO4)3, 0.1 mL 7 wt% H2O2) in 0.8 mL THF at 55 °C for 20 min; Radical species trapped by DMPO.
146 Activation of H2O2 – EPR spectroscopy investigations
6.4.3 The [Al(H2O)6]3+/H2O2 system formed from aluminum perchlorate in the presence of coal and tetrahydrofuran as the solvent
Since the aforementioned EPR analyses revealed different radical species for the
3+ [Al(H2O)6] /H2O2 system formed from aluminum perchlorate, further experiments in presence of coal were also performed. Noteworthy, an experiment under original reaction conditions (70 wt% H2O2, 1 mmol Al(ClO4)3) was not feasible due to strong exothermic reactions after addition of the of Al(ClO4)3/H2O2 solution to the dispersion of coal in 2-Me-
THF. For this, 0.2 g lignite were dispersed in only 0.2 mL 2-Me-THF to obtain a thick slurry for detecting a strong carbon signal. Next the solution of Al(ClO4)3/H2O2 was added, quickly followed by the addition of the DMPO trap agent. Figure 6.23a shows the EPR spectra recorded for a duration of 20 min. For better investigation, single spectra were extracted after 40, 300 and 600 s (Figure 6.23b). At the beginning, the strong carbon signal stemming from coal can be observed. Surprisingly, unlike the Al(NO3)3-based system which leads to the instantaneous disappearance of the coal radical signal, the intensity of the coal radical is dramatically reduced after the addition of Al(ClO4)3/H2O2 (Figure 6.23b, 40 s); however, the signal stays present in the spectra background even after 600 s. Furthermore, the spectrum reveals again the prompt formation of DMPO-OH and DMPO-CRx radical species.
The first peak shown, originates most likely from the DMPO-2-Me-THF species, as already witnessed several times. The second peak stems from DMPO-OH adduct, followed by further DMPO-CRx radical species, either from coal or the degradation product DMPOX.
After 300 s, the signals and the intensities of the formed radical species are dramatically reduced so that the carbon signal from coal is thus clearly visible as the dominating species in the spectrum. This observation is also different compared to the EPR studies with
147 Activation of H2O2 – EPR spectroscopy investigations
aluminum nitrate. In that case, the signals for the DMPO degradation product MNPA remained stable until the end of the measurement (10-20 min).
3+ Figure 6.22: EPR spectra obtained from the [Al(H2O)6] /H2O2 system (0.1 mmol Al(ClO4)3, 0.1 mL 7 wt% H2O2) in 0.2 mL THF and 0.2 g lignite at 55 °C for 10 min; Radical species trapped by DMPO.
DMPO-CRx DMPO-OH
DMPO-CRx (DMPO-MeTHF)
Carbon signal from coal
3+ Figure 6.23: Extracted EPR spectra obtained from the [Al(H2O)6] /H2O2 system (0.1 mmol Al(ClO4)3, 0.1 mL 7 wt% H2O2) in 0.2 mL THF and 0.2 g lignite at 55 °C for 10 min, b) EPR spectra extracted from Figure 6.22 ; Radical species trapped by DMPO.
148 Activation of H2O2 – EPR spectroscopy investigations
To study whether the solution radical species formed from Al(ClO4)3/H2O2 have a similar or different effect on the high matured coal “anthracite”, another EPR experiment was performed, by using the same reaction conditions as for the experiments with lignite. Figure
6.24a presents the EPR spectra obtained over a time period of 10 min. Firstly, the very intense carbon signal stemming from coal can be observed. This signal remains stable throughout the whole measurement. After addition of the Al(ClO4)3/H2O2 solution, followed by the DMPO spin trap, small signals for DMPO-OH and DMPO-CRx are detected but are completely vanished after ca.1 min, confirming again the very stable chemical structure of anthracite. These results differ from the observations from experiments with aluminum nitrate as an activator for hydrogen peroxide. In those experiments, the lifespan of the radical species is long and therefore the occurrence of radicals in the medium seems to be sustained throughout the reaction duration.
10 G
3+ Figure 6.24: a) EPR spectra obtained from the [Al(H2O)6] /H2O2 system (0.1 mmol Al(ClO4)3, 0.1 mL 7 wt% H2O2) in 0.8 mL THF and 0.2 g anthracite at 55 °C for 10 min, b) extracted spectrum from Figure 6.24a after the injection of DMPO spin trap. Radical species trapped by DMPO.
149 Activation of H2O2 – EPR spectroscopy investigations
6.5 Activation of hydrogen peroxide by other nitrates
We extended our studies to other compounds with representive elements in order to assess whether the formation of the HO· and HOO· radicals also occur in the presence of these ions. The decomposition and activation of hydrogen peroxide was tested with the following different types of nitrates:
NaNO3
LiNO3
In(NO3)3·H2O
Mg(NO3)2
Zn(NO3)2·6 H2O
Ga(NO3)3·H2O
6.5.1 Decomposition of hydrogen peroxide mediated by different nitrates
In Figure 6.25a, the decomposition of H2O2 [% of initial concentration] was measured during and after the reaction at 55 °C in absence of an organic compound (solvent). On the one hand, the salts of sodium, lithium, magnesium did not display catalytic activity for the decomposition of H2O2, whereas the presence of zinc nitrate lead to 4% decomposition of this oxidant after 7 h. Surprisingly, the decomposition profiles for the reactions with indium and gallium nitrate show almost the same trend. Herein, the decomposition of H2O2 was about 67 % for both. For all the experiments, the initial rate of decomposition of H2O2 could
150 Activation of H2O2 – EPR spectroscopy investigations
be described by an apparent first order kinetic law with respect to H2O2 concentration. Table
6.4 lists the initial and final concentration of H2O2 for the reactions with aluminum, indium and gallium nitrate and their apparent first order rate constants k (s-1) calculated (Figure
6.25b) at 55 °C. However, the studies pointed out that the activation of H2O2 is best mediated by the aluminum nitrate with ca. 6% residual H2O2 after 7 h at 55 °C.
a) b)
Figure 6.25: a) Decomposition profiles of H2O2 catalyzed by different nitrates; b) first order fit for the decomposition of H2O2 catalyzed by different nitrates; Reaction conditions: H2O2 (70 wt%, 2 mL, 48 mmol), salt (2 mmol), 55 °C.
Table 6.4: Apparent first-order rate constants for the decomposition of H2O2 obtained at 55 °C
-1 Salt [Salt] [H2O2]initial [H2O2]420min k (s )
2 mmol 48.6 mmol 3.2 mmol -3.39E-3 Al(NO3)3 · 9H2O
2 mmol 48.2 mmol 32.3 mmol -9.12621E-4 In(NO3)2 · H2O
2 mmol 47.8 mmol 31.1 mmol -9.35581E-4 Ga(NO3)3 · 9H2O
151 Activation of H2O2 – EPR spectroscopy investigations
6.6 EPR characterization of hydrogen peroxide activation with other metals
6.6.1 Comparison of In(NO3)3 and Ga(NO3)3 with Al(NO3)3 as an
activator for H2O2
As already discussed, a strong formation of HO· and HOO· radicals was confirmed by EPR spectroscopy when aluminum nitrate is added to the H2O2 solution. Therefore, the revealed differences in the decomposition profiles of H2O2 mediated by different metals, lead us to the question how this might affect the formation of radicals as well. In order to get more insight into H2O2 activation further EPR studies were performed with indium and gallium nitrate as catalysts.
Figure 6.26 shows the kinetic study (a) and an extracted single EPR spectrum (b) of a reaction mixture of In(NO3)3 (1 mmol) dissolved in H2O2 (70 wt %, 1 mL). The EPR spectra of the kinetic study show a fast formation of DMPO radical adducts, without any signs of decomposition for a duration of 20 min. The signals seem not to be as intense as the ones obtained for the reactions with aluminum salt as the catalyst. This is a result of the less reactive behavior of indium nitrate. The DMPO radical adducts are longer stable due to the slow formation of new HO· radicals. For better visualization, Figure 6.26b shows the single spectrum obtained after 60s reaction time. Herein, the overlapping signals of DMPO-OH and
-OOH adducts are denoted.
With gallium nitrate on the other hand, the spectra and kinetic study of the reaction show surprisingly weak signals, indicating almost none radicals or rather DMPO adducts. At the first sight, this was not expected due to the fact that the degradation of H2O2 with indium and
152 Activation of H2O2 – EPR spectroscopy investigations
gallium nitrate showed almost identical decomposition profiles (Figure 6.25). A single signal was extracted and its data smoothed using Savitzki-Golay method. The obtained signal indicates the presence of low amount of DMPO-OOH adduct. In order to support that observation, further measurements with the more stable and HOO· sensitive DEPMPO trapping agent were performed (Figure 6.28a). As known the DEPMPO trapping agent generates signals in the kinetic profile that are more prominent.149, 153
EPR measurements with DEPMPO as the spin trap confirmed the generation of HO· and
HOO· radicals for the reaction with indium nitrate as well as gallium nitrate (Figure 6.28b).
Likewise, the experiments with the DMPO spin trap shows that the DEPMPO adducts are stable and do not decompose over a certain time. In addition, an experiment without catalyst was performed. Small signals for the DEPMPO-OH adduct were determined, which indicate a slight formation of HO· radicals for the H2O2 (7 wt %) solution. As already discussed, a slow decomposition of H2O2 can be always observed at room temperature.
a) b)
DMPO-OOH
DMPO-OH
Figure 6.26: a) EPR spectra from the reaction of H2O2 (70 wt %) and In(NO3)3 with DMPO spin trap; b) single spectrum obtained after 1 min reaction time from the kinetic study; Reaction condition: 1 mmol salt dissolved in 1 mL H2O2, room temperature.
153 Activation of H2O2 – EPR spectroscopy investigations
a) b)
Figure 6.27: EPR spectra from the reaction of H2O2 and Ga(NO3)3 with DMPO spin trap; b) single spectrum obtained after 1 min reaction time from the kinetic study; data smoothed by Savitzki-Golay method; Reaction condition: 1 mmol salt dissolved in 1 mL H2O2, room temperature.
a)
DEPMPO-OH DEPMPO-OOH
b)
c)
d)
DEPMPO-OH only
Figure 6.28: EPR spectra of DEPMPO radical adducts formed in experiments with a) Ga(NO3)3, b) In(NO3)3, c) Al(NO3)3 (1 mmol) dissolved in H2O2 (7 wt%) and d) H2O2 (7 wt%) without addition of catalyst at room temperature.
154 Activation of H2O2 – EPR spectroscopy investigations
6.6.1.1 Effect of temperature on EPR spectra for experiments with different nitrates
Since former experiments have already shown that the increase in the reaction temperature
3+ has an effect on the formation of radicals mediated by the [Al(H2O)6] system, the effect of temperature on the EPR spectra for experiments with gallium and indium nitrate were also investigated at 55 °C. Figure 6.29 shows a comparison of single EPR spectra from reaction of H2O2 with gallium, indium and aluminum nitrate obtained after 40, 600 and 1200 s, following the addition of the DMPO spin agent at 55 °C. As aforementioned, the formation of DMPO-OH adducts and their fast decomposition can be observed in the presence of aluminum nitrate. For the reactions with gallium and indium nitrate as an H2O2 activator, a slow conversion and low yields of DMPO-OH species are observed. These results agree with the corresponding decomposition profiles (Figure 6.25) obtained for the reaction at
55 °C. In addition, a control experiment with sodium nitrate (NaNO3) was performed. Weak but stable signals for DMPO-OH adduct, and no DMPO decomposition, was observed over a time period of 20 min, like for the blank experiment with only hydrogen peroxide solution.
The stable DMPO-OH signal indicates that a low concentration of HO· radicals seems to be present which is not further increasing since the DMPO spin trap is not consumed and no
- DMPO degradation products were detected. Therefore, the NO3 anion is demonstrated to not have a role in the reaction by its own right.
155 Activation of H2O2 – EPR spectroscopy investigations
Figure 6.29: Comparison of EPR profiles of DMPO-trapping products obtained from reactions with a) Ga(NO3)3, b) In(NO3)3 and c) Al(NO3)3 (1 mmol)) dissolved in H2O2 (7 wt%, 1 mL) at 55 °C.
6.6.1.2 Effect of solvent on EPR spectra for experiments with different nitrates
In Chapter 6.1.1, the effect of solvent on the EPR spectra was studied. Due to the fact, that gallium and indium nitrate show none to very low solubility in THF and 2-Me-THF, only a comparison of the different salts in presence of methanol as a solvent was possible. Figure
6.30 indicates that the formation of DMPO radical adducts is more pronounced in presence of methanol. The spectra also reveal the formation of methanol-1-yl radicals for the experiments with gallium and indium nitrate, although the overall generation of radicals in the reaction mixture with gallium still seems to be very little. However, compared to the reaction with the aluminum nitrate, the EPR spectra for the experiment with indium nitrate show again a slower and lower amount of radical species.
156 Activation of H2O2 – EPR spectroscopy investigations
Figure 6.30: Comparison of EPR spectra of DMPO radical adducts acquired from experiments with H2O2 (7 wt %, 1 mL) and a) Ga(NO3)3, b) In(NO3)3 and c) Al(NO3)3 (1 mmol) in methanol at 55 °C.
6.6.2 Comparison of In(ClO4)3 and Ga(ClO4)3 with Al(ClO4)3 as an
activator for H2O2
Aluminum perchlorate was tested as a hydrogen peroxide activator. It was found that only hydroxyl radicals are formed within this reaction system. Since the EPR studies of gallium and indium nitrate showed that the same radical species are formed as with aluminum nitrate, only in lower concentrations, it was interesting to examine if corresponding results can be found for gallium, indium and aluminum perchlorate. First, EPR experiments were performed at room temperature to assess whether a catalytic activity of the salts can already be detected. For all reaction systems, very low signal intensities were detected, and single spectra had to be extracted from the kinetic study for deeper investigations.
157 Activation of H2O2 – EPR spectroscopy investigations
Figure 6.31a shows the EPR spectra obtained from the catalytic system Al(ClO4)3/ H2O2.
The fast formation of DMPO-OH radical species can be observed (40 s). After 600 s of reaction duration, signals for the degradation product DMPOX start to appear beside the signals for DMPO-OH which on the other hand are slowly decreasing. At the end of the measurement (1200 s), the EPR spectra is a wild mixture of DMPO-OH and DMPOX signals, with the DMPOX slightly in majority.
Totally different spectra are obtained for the experiment containing the
Ga(ClO4)3/H2O2 system (b). Firstly, the generation of hydroxyl radicals seems to be much slower compared to Al(ClO4)3/H2O2, and secondly, no DMPO degradation products were detected. The DMPO-OH signals remain stable after 600 s until the end of the experiment
(1200 s), indicating that the DMPO spin trap is not fully consumed by hydroxyl radicals, and therefore, no DMPO adducts are attacked and decomposed.
Surprisingly, very weak signals were detected for the EPR experiment with
In(ClO4)3/H2O2. This was not expected, since the experiments with the nitrates showed that indium seems to be a better activator for H2O2 than gallium nitrate. However, as Figure 6.31c demonstrates, the formation of DMPO-OH adducts is low and the weak signals start to disappear already after 600 s. In addition, no DMPO decomposition products were detected.
158 Activation of H2O2 – EPR spectroscopy investigations
Al(ClO4)3/ H2O2
Ga(ClO4)3/ H2O2
In(ClO4)3/ H2O2
3+ Figure 6.31: EPR spectra obtained from the kinetic study from a) [Al(H2O)6] /H2O2 system (1 mmol 3+ Al(ClO4)3, 1 mL 7 wt% H2O2); b) [Ga(H2O)6] /H2O2 system (1 mmol Ga(ClO4)3, 1 mL 7 wt% H2O2), 3+ and c) [In(H2O)6] /H2O2 system (1 mmol In(ClO4)3, 1 mL 7 wt% H2O2) for 20 min at 20 °C; Radical species trapped by DMPO.
159 Activation of H2O2 – EPR spectroscopy investigations
6.6.2.1 Effect of temperature on EPR spectra for experiments with different perchlorates
In addition, the effect of temperature on the generation of radical species in the
Ga(ClO4)3/H2O2, In(ClO4)3/H2O2 was also investigated. For this, the same solutions as prepared for EPR experiments at room temperature were used and heated within the EPR spectrometer to 55 °C. As usual, DMPO spin trap was injected shortly after the data recording was started. Furthermore, a control experiment with sodium perchlorate was performed which showed no catalytic activity in previous tests. The EPR experiment showed that a weak but stable signal for DMPO-OH adduct was detected for the measurement over
20 min, like e.g. for the blank experiment (only H2O2) or with sodium nitrate. Figure 6.32 presents a single EPR spectrum obtained from the kinetic study with sodium perchlorate, showing the typical four signal spectrum for the DMPO-OH adduct.
Figure 6.32: Extracted EPR spectrum obtained from the kinetic study with NaClO4/H2O2 system (1 mmol NaClO4, 1 mL 7 wt% H2O2) at 55 °C for 20 min; Radical species trapped by DMPO.
160 Activation of H2O2 – EPR spectroscopy investigations
Figure 6.33 presents extracted EPR spectra (40, 600, 1200 s) obtained from the reactions with Ga(ClO4)3/H2O2 and In(ClO4)3/H2O2 at 55 °C. The EPR experiment at 20 °C, the spectra for the Ga(ClO4)3/H2O2 system shows slow generation of DMPO-OH adducts.
Contrary to the former experiment at 20 °C, the concentration of DMPO-OH seems to slowly decrease after 600 s reaction duration. Again, no DMPO degradation products, e.g.
DMPOX, were detected. On the other hand, the spectra for the In(ClO4)3/H2O2 system reveal now strong similarities with the spectra obtained for the Al(ClO4)3/H2O2 (Figure 6.19), showing also prompt formation of DMPO-OH as well as DMPOX adducts. After 600 s, the signals for the DMPOX are dominating the spectrum, and are the only radical species detected after 1200 s. These results clearly show that the In(ClO4)3/H2O2 system needs higher temperatures to activate the hydrogen peroxide whereas the generation of hydroxyl radicals within the Ga(ClO4)3/H2O2 seems to be only increased at 55 °C. Compared to
Al(ClO4)3/H2O2, the In(ClO4)3/H2O2 system seems to be equally effective for the activation of
H2O2.
161 Activation of H2O2 – EPR spectroscopy investigations
Ga(ClO4)3/ H2O2
Only DMPO-OH
DMPOX
In(ClO4)3/ H2O2
DMPO-OH
3+ Figure 6.33: EPR spectra obtained from the kinetic study from a) [Ga(H2O)6] /H2O2 system (1 mmol 3+ Ga(ClO4)3, 1 mL 7 wt% H2O2); b) [In(H2O)6] /H2O2 system (1 mmol In(ClO4)3, 1 mL 7 wt% H2O2) at 55 °C for 20 min; Radical species trapped by DMPO.
162 Activation of H2O2 – EPR spectroscopy investigations
6.6.2.2 Effect of solvent on EPR spectra for experiments with different perchlorates
To complete this EPR study, further experiments in presence of a solvent needed to be performed as well. For this, the solvent tetrahydrofuran was chosen since it was also the reaction medium for the model compound and dynamic NMR studies (s. Chapter 7).
Once again, a control experiment with sodium perchlorate was performed to avoid false data interpretation (Figure 6.34). The obtained EPR spectra show the prompt formation of DMPO-
OH and various DMPO-CRx radical species (solvent and DMPOX) which are slowly equally decreasing over time. The formation of DMPO-CRx species was expected since the solvent is a good and stable radical scavenger, as already discussed at the beginning of this chapter, resulting in the generation of DMPOX degradation products.
DMPO-OH
DMPO-CRx
Figure 6.34: Extracted EPR spectrum obtained from the kinetic study with NaClO4/H2O2 system (1 mmol NaClO4, 1 mL 7 wt% H2O2) in 0.8 mL THF at 55 °C for 20 min; Radical species trapped by DMPO.
163 Activation of H2O2 – EPR spectroscopy investigations
Figure 6.35 compares the spectra obtained from EPR experiments with Ga(ClO4)3 and In(ClO4)3 as catalyst in the presence of tetrahydrofuran as reaction medium. For the
Ga(ClO4)3/H2O2 system a slowly growing formation of DMPO-OH, DMPO-CRx and DMPOX radical species can be observed. The concentration of the DMPO-OH adduct seems to be slightly higher than that of the DMPO-THF adduct. Only weak signals for DMPO degradation products can be observed throughout the measurement. The results for the EPR measurement with the In(ClO4)3/H2O2 system are once again quite similar with the spectra recorded for the Al(ClO4)3/H2O2 system (Figure 6.21). A fast and high formation of DMPO-
OH, DMPO-CRx and minimal DMPO degradation products can be seen. Moreover, the signals are stable and only slightly changing over time. The results indicate that also in presence of a solvent the In(ClO4)3/H2O2 is efficient in building several radical species which can be useful for the conversion of coal.
164 Activation of H2O2 – EPR spectroscopy investigations
1200 s
3+ [Ga(H2O)6] /H2O2 in THF
600 s
DMPO-THF 40 s
DMPO- OH DMPOX
DMPO-THF
3+ [In(H2O)6] /H2O2 in THF
3+ Figure 6.35: EPR spectra obtained from the kinetic study from a) [Ga(H2O)6] /H2O2 system (1 mmol 3+ Ga(ClO4)3, 1 mL 7 wt% H2O2) in 0.8 mL THF; b) [In(H2O)6] /H2O2 system (1 mmol In(ClO4)3, 1 mL 7 wt% H2O2) in 0.8 mL THF at 55 °C for 20 min; Radical species trapped by DMPO.
165 Activation of H2O2 – EPR spectroscopy investigations
6.7 Conclusions EPR studies
3+ The radical species formed in the [Al(H2O)6] /H2O2 system by aluminum nitrate dissolved in hydrogen peroxide and their utilization in the lignite conversion was examined by EPR spectroscopy. The results demonstrate the formation of HO· and HOO· radicals in the reaction medium. Moreover, methanol used as a solvent seems to be a better radical scavenger than the solvent 2-Me-THF, which leads to a higher population of hydroxymethyl radicals than hydroxyl or perhydroxyl radicals. Since hydroxymethyl radicals are less reactive than hydroxyl radicals, the milder degradation of lignite into soluble aromatic products might be explained on the basis of this observation. In 2-Me-THF, the interaction of H2O2 with Al(NO3)3 generates a coal structure, leading to mostly aliphatic lignite residue samples, whereas by choosing a different kind of solvent, e.g. methanol, the prevalence of
·CH2OH instead of HO· and HOO· radicals, result in a coal conversion or benefication in which part of the aromatic structure is preserved. Altogether, the final product nature, oil or solid, aliphatic or aromatic, can be tuned by the choice of solvent.
3+ Furthermore, it was found that H2O2 can also be activated by the [M(H2O)6] systems of gallium and indium nitrates and perchlorates, respectively. Interesting results were obtained
3+ from further EPR experiments of their [M(H2O)6] systems. The EPR spectra of the three hexaaqua-complexes (Al, Ga, In) in H2O2 revealed completely different behaviors regarding
3+ the formation of radical species. First, it was revealed that the [M(H2O)6] systems formed
3+ by nitrates generate HO· and HOO· radicals, whereas for the [M(H2O)6] systems formed by perchlorates, only HO· radicals were detected.
In general, the EPR studies showed that the gallium salts seem to be less effective for the activation of H2O2 and the generation of radicals than indium or aluminum nitrate.
166 Activation of H2O2 – Dynamic NMR spectroscopy investigations
7 Activation of hydrogen peroxide- Part II:
Variable temperature proton NMR (1H-NMR) studies of different [M(H2O)6]3+/H2O2 systems
3+ Former NMR studies of the [Al(H2O)6] /H2O2 system by Rinaldi et al. already showed that the proton exchange between H2O2 and H2O induces the nucleophilicity of H2O2 through a
3+ proton-exchange mechanism involving the second coordination sphere of [Al(H2O)6] .
Compelling evidence supporting this mechanism is the epoxidation of α,β-unsaturated ketones, which is a reaction that occurs by using HOO- anion as an oxidant. Furthermore,
1 3+ interesting H-NMR spectra were obtained of the [Al(H2O)6] /H2O2 system in THF-d8 at variable temperatures. It was shown, how the proton exchange enhances with increasing reaction. Peak broadening of H2O2 and H2O signals was observed, resulting in one broad and coalescent resonance. This chemical shift is the weight average of the chemical shifts of the individual states, indicating a very fast proton exchange.154
3+ 3+ 3+ 3+ As demonstrated by the EPR studies, [M(H2O)6] /H2O2 systems (M= Al , Ga , In ) can generate radicals. Since the radicals HO∙ and HOO∙ are formed by the decomposition of
- H2O2 and HOO , possible differences regarding the proton-exchange between H2O2 and H2O
1 3+ was also investigated. For this, H-NMR variable temperature spectra of [Al(H2O)6] /H2O2
167 Activation of H2O2 – Dynamic NMR spectroscopy investigations
obtained from Al(NO3)3, Al(ClO4)3 and AlCl3 in THF-d8 were collected. In addition,
3+ [M(H2O)6] /H2O2 variable temperature spectra were recorded from Ga(ClO4)3 and In(ClO4)3 as an activator for hydrogen peroxide. Next, the exchange rates for each various
3+ [M(H2O)6] /H2O2 system were calculated using TopSpin 3.2 “D-NMR” software. In addition, a blank and a control experiment with only H2O2 and NaClO4/H2O2 were performed at room temperature. Figure 7.1 compares the spectra obtained for the NMR experiments and shows that both spectra look the same. No proton exchange can be observed which is crucial for the activation of hydrogen peroxide. These results agree with data obtained from the EPR studies in which also no activation of H2O2 was observed.
H2O H2O2
THF
1 Figure 7.1: H-NMR spectra of H2O2 (0.14 mol/L) and NaClO4/H2O2 (0.003/0.14 mol/L) in THF-d8 at room temperature.
168 Activation of H2O2 – Dynamic NMR spectroscopy investigations
Two proton-exchanges have to be considered for the catalytic proton-exchange of
H2O2 (A) and H2O (B) in presence of a group 3 metal salt. A first order chemical exchange process is assumed (Scheme 7.1) and gives rise to a two-site exchange in a dynamic NMR
1 3+ experiment. Figure 7.2 shows the H-NMR spectrum of the [Al(H2O)6] /H2O2 system obtained from Al(ClO4)3. Sharp peaks can be observed for H2O2 (9.75 and H2O (3.29, indicating a rather slow proton exchange which was confirmed by D-NMR calculations. The exchange rate at this temperature was determined to be about 46 Hz.
Yet, when the temperature of the reaction system increases (Figure 7.3), a line broadening for both resonances are found (intermediate exchange) until they seem to disappear at about 45 °C (coalescence, k= 5070), and start to reappear at 65 °C as one broad resonance, due to the very fast proton exchange (13800 Hz). Moreover, the resonance for the hexaaquaaluminum system (9.54) only shows minor changes in its line shape and indicates no proton exchange process between H2O2 or bulk water.
THF
Scheme 7.1: Two-site exchange model for proton exchange of H2O H2O2 and H2O
H2O2 3+ [Al(H2O)6] Probe
1 3+ Figure 7.2: H-NMR of [Al(H2O)6] /H2O2 (0.003 mol/L Al(ClO4)3∙9H2O/ 0.14 mol/L H2O2) in THF-d8 at -55 °C.
169 Activation of H2O2 – Dynamic NMR spectroscopy investigations
65 °C
Coalescence 35 °C
15 °C
5 °C 0 °C -5 °C -15 °C
Zoom of 1H-NMR spectra obtained at 35 to -35 °C 65 °C (5 °C steps).
-55 °C
1 3+ Figure 7.3: Variable temperature H-NMR of [Al(H2O)6] /H2O2 (0.003 mol/L Al(ClO4)3∙9H2O/ 0.14 mol/L H2O2) in THF-d8.
NMR spectra obtained for the system with gallium perchlorate suggest a faster
exchange process (Figure 7.4). Two signals of H2O2 (9.8) and H2O ( 3.29) already appear
wider than the ones observed for the system with aluminum perchlorate at -55 ° C and with
a calculated exchange rate about 410 Hz, which almost 10 times higher as for the
3+ [Al(H2O)6] /H2O2 system. Again, it is shown that the signals move together, coalescense
(25 °C, k= 5300 Hz) and finally join to build a single resonance at 35 °C. The final exchange
rate was determined to be 44600 Hz at 65 °C.
170 Activation of H2O2 – Dynamic NMR spectroscopy investigations
65 °C
45 °C
25 °C, Coalescence
5 °C
0 °C
-5 °C Zoom of 1H-NMR spectra -15 °C obtained at 15 to 35 °C.
-35 °C
-55 °C
1 3+ Figure 7.4: Variable temperature H-NMR of [Ga(H2O)6] /H2O2 (0.003 mol/L Ga(ClO4)3∙9H2O/ 0.14 mol/L H2O2) in THF-d8.
For the system with indium perchlorate (Figure 7.5), different observations within the NMR spectra can be made. First of all, the resonance of the bulk water shifted from about 3.3 to
3.7 ppm, overlapping with one of the solvent peak (3.5 ppm). Further, the peaks appear
3+ sharper compared to that of the [Ga(H2O)6] /H2O2 system, showing a slower proton exchange (144 Hz). In addition, a peak close to the one of H2O2 cannot be observed such as for the Ga- and Al-systems. Although the proton-exchange between H2O2 and H2O seems to be slower compared to the gallium system, coalescence can be already observed at 15 °C
(k= 3700 Hz). Again, a joint peak appears afterwards at ca.6-4 ppm. At the end of the measurement (65 °C), the calculated exchange rate was about 26000 Hz.
171 Activation of H2O2 – Dynamic NMR spectroscopy investigations
65 °C
35 °C
15 °C, Coalescence
0 °C
-1 5 °C Zoom of 1H-NMR spectra obtained at 0 to 65°C. -35 °C
-55 °C
1 3+ Figure 7.5: Variable temperature H-NMR of [In(H2O)6] /H2O2 (0.003 mol/L In(ClO4)3∙9H2O/ 0.14mol/L H2O2) in THF-d8.
Surprisingly, totally different dynamic behavior is detected for the reaction systems with aluminum nitrate and chloride. The NMR spectra and data obtained for the exchange rates
(k) show a faster proton-exchange at low temperatures than at higher temperatures.
In case of aluminum nitrate (Figure 7.6), the peak shape of H2O2 (9.66 ppm) and
H2O (.68 ppm) appears already at -55 °C very broad, having a determined k about 1360
Hz, which is slowly increasing as indicated by the line broadening of the peaks. Then about
0 °C, the line shape starts to become narrower again. At 65 °C, the peaks for H2O2 (9.08 ppm) and H2O (2.56 ppm) are strongly emerged, indicating slower proton exchange (ca.
600 Hz) within the reaction mechanism.
172 Activation of H2O2 – Dynamic NMR spectroscopy investigations
Analyzing the data for the system with aluminum chloride (Figure 7.7), one can see that the peaks for H2O2 (9.68 ppm) and H2O (3.25 ppm) are also slightly broadening from -55 to -35 °C, but then also start to become narrower with increasing temperature. At this point, the determined exchange rate is about 1480 Hz, increasing to ca. 1900 Hz. At the end of the measurements (65 °C), two sharp peaks for H2O2 (9.08 ppm) and H2O (2.29 ppm) are clearly visible and the calculated value of k was about 190 Hz.
For both reaction systems, no coalescence was observed, indicating that the overall proton exchange rate seems to be only slightly affected by the change of temperature.
65 °C
15 °C
0 °C
-15 °C
-35 °C
-55 °C
1 3+ Figure 7.6: Variable temperature H-NMR of [Al(H2O)6] /H2O2 (0.003 mol/L Al(NO3)3∙9H2O/ 0.14 mol/L H2O2) in THF-d8.
173 Activation of H2O2 – Dynamic NMR spectroscopy investigations
65 °C
35 °C
15 °C
0 °C
-15 °C
-35 °C
-55 °C
1 3+ Figure 7.7: Variable temperature H-NMR of [Al(H2O)6] /H2O2 (0.003 mol/L AlCl3*9H2O/ 0.14 mol/L H2O2) in THF-d8.
Table 7.1 lists some of the determined exchange rates k [Hz] and the observed coalescence temperature for all reaction systems (s. Appendix for full data lists, Table 9.7-
11). The fastest proton-exchange was found for the reaction system with gallium perchlorate as catalyst. However, the largest increase in exchange rate from -55 to 0 °C is observed for the reaction with indium. This explains why the peaks coalescense at lower temperatures
(15 °C) than the ones for the gallium system with the highest exchange rates.
For the reactions with aluminum nitrate and chloride the exchange rate descends with increasing temperature and no coalescence was observed during the NMR measurements.
For the aluminum nitrate, the exchange remains quite stable for temperatures between -55
174 Activation of H2O2 – Dynamic NMR spectroscopy investigations
to 0 °C, yet starts to decrease to ca. 590 Hz. The determined k values for the aluminum chloride are continuously decreasing to about 190 Hz, and therefore, showing the slowest proton exchange at 65 °C.
Table 7.1: Summary of calculated exchange rates k [Hz] obtained from dynamic 1H-NMR studies of
3+ the various [M(H2O)6] /H2O2 systems at -55, -25, 0, 25, 45 and 65 °C and the observed coalescence temperature (Tc).
3+ [M(H2O)6] /H2O2 k [Hz] k [Hz] k [Hz] k [Hz] k [Hz] k [Hz] Tc
-55 °C -25 °C 0 °C 25 °C 45 °C 65 °C
Al(ClO4)3 46 141 296 1468 5075 13800 45 °C
Ga(ClO4)3 412 861 1288 5300 18200 44600 25 °C
In(ClO4)3 144 521 1731 5215 13485 26000 15 °C
Al(NO3)3 1363 1516 1353 1152 873 590 -
AlCl3 1481 1788 1197 684 480 190 -
For better visualization of the proton-exchange between hydrogen peroxide and
3+ water, 2D COESY spectra of the various [M(H2O)6] /H2O2 systems were recorded as well.
The proton exchange is indicated by crossing the involved peaks/dots in the spectra (Figure
7.8).
Furthermore, variable temperature 27Al-NMR spectra were recorded for the various
3+ [Al(H2O)6] /H2O2 systems. In all cases, the obtained spectra show no differences and changes during the measurements and no additional information were gained (s. Appendix,
Figure 9.5-7).
175 Activation of H2O2 – Dynamic NMR spectroscopy investigations
Al(ClO4)3/H2O2 Ga(ClO4)3/H2O2
In(ClO4)3/H2O2
Al(NO3)3/H2O2 AlCl3/H2O2
3+ Figure 7.8: 2D COESY spectra of the various [M(H2O)6] /H2O2 systems recorded at -55 °C.
176 Activation of H2O2 – Dynamic NMR spectroscopy investigations
7.1 Thermodynamic parameters of the H2O2-H2O exchange process
From the NMR data and TopSpin3.2 software exchange rates for all measurements and temperatures were calculated and the relevant kinetic parameters, e.g. activation energy Ea, were determined by an Arrhenius plot (ln(k) vs. 1/T) using following equation:
−(푠푙표푝푒) ∗ 푅 푘퐽 퐸 = = [ ] 푎 1000 푚표푙
Surprisingly, no linear fits are obtained for all reaction systems (Figure 7.9), indicating that the temperature dependence did not follow simple Arrhenius behavior. In case of the perchlorates, one can see concave upward Arrhenius plots which bend at about 0 °C and seem to connect two straight slopes of different activation energies (as indicates in the figures), although the bend within the plot for In(ClO4)3/H2O2 is more smoothed compared to the ones found for the Al(ClO4)3/H2O2 and Ga(ClO4)3/H2O2 systems. For the systems with
Al(NO3)3/H2O2 and AlCl3/H2O2, convex upward plots are obtained. Normally, concave/convex Arrhenius plots are rare and the most common interpretation is that at least two different rate-limiting steps are involved.155 Regarding literature, this phenomenon can be observed at times for enzymes-catalyzed reactions where two competing enzymatic forms are present.156 These enzymatic dynamics are well described by a proposed model known as Michaelis-Menten kinetics which explains how enzyme can cause kinetic rate enhancement of a reaction and how the reaction rates depend on the concentration of enzyme and substrate.157 Furthermore, concave/convex Arrhenius plots were also reported for some hydrogenation, hydroformylation reactions under water gas-shift reactions conditions using for instance rhodium-based catalysts.158
177 Activation of H2O2 – Dynamic NMR spectroscopy investigations
Al(ClO4)3/H2O2 Ga(ClO4)3/H2O2
In(ClO4)3/H2O2
Al(NO3)3/H2O2 AlCl3/H2O2
Figure 7.9: Arrhenius plots of ln(k) against 1/T obtained from d-NMR data of the various 3+ [M(H2O)6] /H2O2 systems; a) Al(ClO4)3, b) Ga(ClO4)3, c) In(ClO4)3, d) Al(NO3)3, e) AlCl3.
178 Activation of H2O2 – Dynamic NMR spectroscopy investigations
Complete analysis of the activation energy parameters requires also the use of the Erying equation. Since we know k at any single temperature, it is possible to calculate ∆G‡ for each temperature and determine afterwards the average value for ∆G‡. The thermodynamic functions were calculated according to the following equations (enthalpy ∆H‡, entropy ∆S‡, the Gibbs energy ∆G‡ of activation):
푅푇 푘푏푇 ∆퐺‡ = ( ) [ln ( ) − ln (푘 )] 1000 ℎ 푟푎푡푒
‡ ∆퐻 = 퐸푎 − 푅푇푐
(∆퐻‡ − ∆퐺‡) ∆푆‡ = 푇푐
with kb = Boltzmann constant, h = Planck constant, R= gas constant, Tc = coalescence temperature
For reaction systems with Al(NO3)3 and AlCl3 following equations was used:
푘 ln ( 1) 푘 퐽 퐸 = − [ 2 ] = [ ] 푎 1 1 − 푚표푙 푇1 푇2
Since all the reaction systems show a bend about the melting point of water (0 °C) in their
Arrhenius plots, it is hard to dismiss the idea that water and its properties/ anomaly may play a role in proton-exchange during the reaction. Between 0 °C and 4 °C (temperature of maximum density) the hydrogen bond collapse dominates over the normal thermal expansion159 which may cause the sudden change in proton-exchange. Furthermore with
179 Activation of H2O2 – Dynamic NMR spectroscopy investigations
increasing temperature, the resulting partial collapse of the hydrogen bonded network within the water clusters allows non-bonded molecules to approach more closely, increasing the
160 number of nearest neighbors (e.g. hexagonal ice = 4 neighbors). However, H2O2-H2O is a binary system and the melting point of an aqueous solution of H2O2 (60 wt%) is about -55 °C161, so it seems rather a coincidence that the non-linear Arrhenius plots bend at about 0°C.
Considering these observations and facts, thermodynamic parameters have been calculated for each of the two indicated slopes for the different reaction systems from -55 to
0 °C and 0 to 65 °C.
The determined parameters for the several reaction systems are listed in Table 7.2. Since no coalescence was observed for the reactions with Al(NO3)3 and AlCl3, no values for enthalpy and entropy of activation can be determined from the NMR-data.
All reaction systems, except the ones with Al(NO3)3 and AlCl3, have positive values for Ea,
∆H‡ and ∆G‡ and show that the reactions require an input of energy. Further, all systems
‡ show negative values for ∆S . Regarding the perchlorate systems, the lowest values for Ea,
‡ ‡ ∆H and ∆G are found for the Ga(ClO4)3/H2O2 system. Herein, the Ea for the first slope (-55 to 0 °C) is about 10 kJ/mol and for the second slope (0 to 65 °C) about four times higher
(46 kJ/mol). The estimated values for the activation energy barrier ∆G‡ are about 46 and 51 kJ/mol, and for the activation enthalpy ∆H‡ 7 and 38 kJ/mol, showing that the system is a non-spontaneous and endergonic reaction. The determined values for the activation entropy
‡ ∆S increases from -132 to -46 J/mol for the two slopes. In case of the Al(ClO4)3/H2O2 system, the energy barriers are slightly higher than that for the gallium system. The Ea for the first slope is about 16 kJ/mol and increases about 3 times (46 kJ/mol) for the second one, whereas ∆G‡ was determined to be about 50 and 55 kJ/mol for slope 1 and 2. The calculated values for ∆H‡ 1 and 2 are about 14 and 44 kJ/mol, and for ∆S‡ -117 and -37
180 Activation of H2O2 – Dynamic NMR spectroscopy investigations
J/mol. Since the plot for the In(ClO4)3/H2O2 system is almost a straight line, the general increase in values from slope 1 to slope 2 is lower. The obtained Ea values are about 22 and
32 kJ/mol, and ∆G‡ is about 47 and 52 kJ/mol. Furthermore, the values for ∆H‡ are 20 and
30 kJ/mol. In addition, the numbers for ∆S‡ slightly increase from -95 for slope 1 and -77
J/mol for slope 2.
Combining and comparing the obtained data, one can conclude following assumption. The more ordered the system, the lower the activation energy barrier. In case of the perchlorates, gallium needs less energy to overcome the barrier for a temperature range from -55 to 0 °C than aluminum and indium, whereas from 0 to 65 °C, the indium system requires less energy.
‡ For the system with Al(NO3)3 and AlCl3, the calculated data for Ea and ∆G are very
‡ different. The Ea shows light negative values, whereas the ∆G values are more or less in the same range like the perchlorate systems. However, the negative Ea values agree with the obtained spectra and corresponding exchange rates which seem to decrease with increasing temperature. These data are also consistent with the results from EPR studies in which also totally different mechanism on the formation of radical species was revealed.
181 Activation of H2O2 – Dynamic NMR spectroscopy investigations
3+ Table 7.2: Thermodynamic parameters of exchange process for the various [M(H2O)6] /H2O2 systems
3+ ‡ ‡ ‡ ‡ ‡ ‡ M / H2O2 Ea 1 Ea 2 ∆H 1 ∆H 2 ∆S 1 ∆S 2 ∆G 1 ∆G 2 TC (-55– (0–65°C) (average) (average) 0°C)
[kJ/mol] [kJ/mol] [kJ/mol] [kJ/mol] [J/molT] [J/molT] [kJ/mol] [kJ/mol] [°C]
Al(ClO4)3 16 46 14 44 -117 -37 50 55 35
Ga(ClO4)3 10 41 7 38 -132 -46 46 51 25
In(ClO4)3 22 32 20 30 -95 -77 47 52 15
Al(NO3)3 -0.06 -10 n.a. n.a. n.a. n.a. 43 55 n.a.
AlCl3 -2 -22 n.a. n.a. n.a. n.a. 45 58 n.a.
*Tc = coalescence temperature
The differences for the various cations can be roughly explained due to increasing ionic radii (Al 3+ In , the experimental values of pKa for their hydrolysis are about 5.0 (Al), 2.6 (Ga) and 4.0 (In), showing that the Ga3+ cation is more acidic than the In3+ and Al3+ cations.162 So far, the structures of aqueous solutions of gallium and indium salts have been poorly studied and information on this topic is rather limited. Still in 2011, Smirnov and Trostin gave a good overview about the structural parameters of the close environment of group III metal ions in aqueous solutions.163 They reported that in general the Al3+, Ga3+ and 182 Activation of H2O2 – Dynamic NMR spectroscopy investigations 3+ In ions form a first coordination sphere of six water molecules, surrounded by a second coordination sphere including 12 water molecules. Nevertheless, this condition seems to 3+ depend on the present anion. In case of an aqueous solution of Ga(ClO4)3, the Ga ion forms a second coordination sphere of 18 water molecules instead of 12, assuming that each water molecule from the first sphere forms three hydrogen bonds with molecules of the 164 second sphere. For aqueous solutions of Al(ClO4)3 and In(ClO4)3 no such behavior was observed so far. Analyzing different aluminum salts, it was also found that the Al3+ ion of an aqueous solution of AlCl3 interacts with eight water molecules in the first sphere, independent of the salt’s 165 concentration. This might be a first hint about the differences found for the AlCl3/H2O2 system. However, the effect of different anions in aqueous solutions has been studied for many different applications, e.g. micellar properties and growth for surfactants166, 167 or ions for controlling protein dynamics.168, 169 Often the properties of some ions are mentioned together with the “Hofmeister series” which refers to the effect of several ions on the precipitation of proteins. It is generally accepted that ions affect the dynamic and structure of water. Some of them act as “structure maker” or “structure breakers”, respectively. Figure 7.10 shows the “Hofmeister series” for anions, pointing out their kosmotropic (structure maker) and chaotropic (structure breaker) effects on the structure of water.170 In the same paper by Yizhak Marcus, it was shown that Al3+ and Ga3+ ions belong to the group of kosmotropes. Kosmotropes are also reported as strongly hydrated ions whereas chaotropes demonstrate the opposite behavior. 183 Activation of H2O2 – Dynamic NMR spectroscopy investigations Strongly hydrated Weakly hydrated Figure 7.10: “Hofmeister series” for anions170 Moreover, Metrick and MacDonald found that decreased 1H/2H exchange can be observed 2- in the presence of kosmotropic, strongly hydrated anions (e.g. SO4 ), suggesting less free water molecules, whereas the 1H/2H exchange increases in presence of chaotropic, weakly - 171 hydrated anions (e.g. ClO4 ). Recently, Kitadai and co-workers published how various ions effect the O-H stretching band of water in an ATR-IR spectrum.172 By calculating the differences in area between the lower (2600-3420 cm-1) and the higher (3420-3800 cm-1) 2- infrared regions (∆MAL-H), compared with SO4 anion (difference= 0), they found that ions increasing the lower O-H stretch absorbances, and reducing the higher O-H stretch absorbances, are expected to strengthen the H2O-H2O hydrogen bonds in their hydration layers, while ions having negative values weaken those bonds. The obtained results are showing following trend from zero to increasing negative values: 2------SO4 > Cl > NO3 >Br >I >ClO4 Regarding these observations, the perchlorate anion (∆MA= -2584) seems to have the strongest effect on the H2O-H2O hydrogen bonds. The nitrate (-1770) and chloride (-1224) anions are more in the lower region and seem to be less effective in weakening the hydrogen bonds. These facts agree with our observation made from the D-NMR studies where the 184 Activation of H2O2 – Dynamic NMR spectroscopy investigations perchlorate systems proved to be good activators for proton-exchange reactions between hydrogen peroxide and water. In case of the AlCl3/H2O2 system, the observed downward-trend for the proton- exchange could be due to a possible ligand exchange reaction of the water molecules from the first coordination sphere with chloride which might act as a bridging ligand, connecting to aluminum ions, and consequently, disturbing the proton-exchange. However, some of these theories are just providing first ideas and explanations about the different behavior in proton-exchange and formation of radical species for the various 3+ [M(H2O)6] /H2O2 systems, and further studies are needed to fully elucidate the different reaction mechanisms. Especially in case of aluminum nitrate and chloride, the peculiar data obtained from the D-NMR studies need to be further investigated and confirmed using other techniques, e.g. (ATR-) FTIR spectroscopy. 185 Activation of H2O2 – Model compound reactions 7.2 Activation of hydrogen peroxide- Part III: Model compound reaction Carvone epoxidation experiments with the various [M(H2O]6]3+/H2O2 systems The interesting results from the several investigations by EPR and NMR spectroscopy, we 3+ already could conclude that the various [M(H2O]6] /H2O2 systems undergo different reaction pathways. Former work by Rinaldi et al., showed that Al(ClO4)3 efficiently catalyzes the epoxidation of carvone to carvone epoxide. Since this is a simple molecule which does not 3+ produce many side-products, the efficiency of the other [M(H2O]6] /H2O2 systems was tested as well for this reaction. Scheme 7.2 shows the epoxidation of carvone, resulting in carvone epoxide (1b) and a hydroxylated byproduct (1b). Scheme 7.2: Oxidation of (R)-carvone Table 7.3 lists the obtained results for the carvone epoxidation reaction, the yield and selectivity of carvone epoxide product and the overall conversion of (R)-carvone after 3 h at 80 °C. For the perchorate salts, the obtained results are very similar; the general conversion 186 Activation of H2O2 – Model compound reactions is between 74-77%, yields of epoxide are about 42-45% and the selectivity is about 55-61% whereas the reactions with aluminum nitrate and chloride only showed small conversion of the carvone. Herein, 20% conversion was found for the reaction with Al(NO3)3 which yielded in 5% epoxide product with a selectivity of 0.2%. For the reaction with AlCl3, a higher conversion of ca. 37% was determined, but yielding in less epoxide (1.6%) and selectivity below 0.1% after 3h at 80 °C. The main conversion for the latter two reactions can be contributed to the inversion of the (R)-configuration of carvone to (S)-configuration as determined by GC-MS analysis. The yield of the hydroxylated product (1b) was found to be about 1% for all reactions with M(ClO4)3 as catalyst and between 0.2-0.5% for the reactions with Al(NO3)3 and AlCl3. This product is formed by a Markownikoff addition of HO• or HOO• radicals to the α-,β-unsaturated structure of the ketone.100 Table 7.3: Comparison of the various catalytic systems for the epoxidation of carvone at 80 °C and 3 h. Reaction conditions: (R)-carvone (25 mmol), ISTD (Di-n-butyl ether, 12.5 mmol), 60 wt% H2O2 (56 mmol), Cat. (0.5 mmol), THF, 80 °C 187 Activation of H2O2 – Model compound reactions The obtained results support the ones of the previously discussed dynamic NMR studies. Since the epoxidation proceeds at the electron-deficient C=C-bond of carvone, a nucleophile such as HOO- seems to be involved and needed for this reaction. The low conversion of carvone and yields in epoxide product obtained from reaction with Al(NO3)3 and AlCl3 suggest that the amount of produced HOO- anions is much less than for the reaction system using M(ClO4)3 as catalyst due to the much lower proton exchange rates. 188 Activation of H2O2 – Model compound reactions 7.3 Conclusions This chapter described the application of dynamic NMR as an effective tool for investigating 3+ the proton-exchange between hydrogen peroxide and water for several [M(H2O)6] /H2O2 systems (M= Al3+, Ga3+, In3+). The exchange rates and thermodynamic parameters of the reaction systems were determined by variable temperature NMR experiments (-55 to 65 °C). It was found that the anions enormously affect the proton-exchange, while the differences between the cations of the group III metals are considerably smaller. The different M(ClO4)3 salts (M=Al, Ga, In) showed great potential for activating proton-exchange. The correspondingly nitrate and chloride demonstrated rather minor proton-exchange which seemed to decrease with increasing reaction temperature. Furthermore, the obtained data revealed concave/convex Arrhenius plots for all reaction systems, surprisingly exhibiting a bend at 0 °C, indicating that the anomalous property of water may have an influence in the proton exchange and its activation, respectively. However, further studies are needed to support obtained results and to get more insights into the different reaction mechanism, especially in case of the aluminum nitrate and chloride systems. In addition, reactions with (R)-carvone, an electron-deficient compound, were performed to further study the efficiency and differences on its epoxidation by the various 3+ 3+ 3+ 3+ [M(H2O)6] /H2O2 systems (M= Al , Ga , In ). The obtained results supported the observation made by D-NMR experiments, revealing good yields and conversions for the reactions using the perchlorates as catalyst and very low conversion/yields for the reaction with aluminum nitrate and chloride. 189 Summary 8 Summary In this thesis, a novel oxidative approach for the conversion of coal, particularly lignite, with activated hydrogen peroxide as oxidative species under mild conditions was explored. At the beginning, the optimum reaction temperature for the liquefaction of lignite mediated 3+ by the [Al(H2O)6] /H2O2 system was explored. It was discovered that the conversion of coal depends strongly on the reaction temperature. The optimum temperature for this methodology is about 65-70 °C where mostly the mineral matter and low amounts of insoluble acidic compounds are remaining. Surprisingly, the obtained product oil samples consist of mostly saturated compounds, comprising not only fragments from coal but also about 35 % products from 2-Me-THF degradation which seems to dissolve the coal fragments. The next chapter discussed the optimum process duration for the conversion of lignite 3+ mediated by the [Al(H2O)6] /H2O2 system at 65 °C. Herein, it was shown that the main conversion of lignite takes place within the first 30 min of process duration. Furthermore, the calculation of comprising 2-Me-THF/coal/OH groups within the new formed product oil samples revealed that the conversion of coal is completed after 120 – 150 min process duration and the ratios of the participating compounds within the product oil do not change afterwards. More interesting results were obtained from studying the effect of solvent for this reaction system. Pretests showed that no product oil samples were obtained from reactions in 190 Summary methanol, ethanol and water. Furthermore, the soluble products comprised high amounts of remaining catalysts. Therefore, the effect of the solvent was studied for coal benefication processes. Herein, the reaction conditions remained, only the process duration was changed to 5, 10 and 30 minutes. The results showed that processing lignite in methanol and water leaves most of the aromatic structure intact, whereas reaction in 2-Me-THF and ethanol revealed a dramatic loss of aromatic compounds. All obtained lignite residue samples showed improved thermal degradation behavior. Since the structure of lignite residue samples obtained from process in methanol were mostly untouched, hydrodeoxygenation experiments were performed to investigate whether further upgrading of the lignite samples is possible. The HDO experiments in presence of Ni2P/SiO2 catalysts presented promising first results. The complex coal structure could be further cleft into smaller fragments, such as phenols, benzenes, naphthenes, hydrocarbons. 3+ In addition, the capability of the [Al(H2O)6] /H2O2 system on higher coal ranks was explored. The experiments showed that only high volatile bituminous coal (here: “Gasflammkohle”) was converted in low amounts. Coals of higher maturity were mostly unaffected and the very low yields of derived product oil samples consist of 2-Me-THF degradation products. Only “Gasflammkohle” showed improved properties regarding thermal degradation. The last chapters of this thesis are dedicated to the mechanistic investigation into the reaction system. Herein, the formed radical species were determined by EPR spectroscopy and variable temperature dynamic NMR studies were performed as well to investigate the proton exchange between activated hydrogen peroxide and water within the 3+ [Al(H2O)6] /H2O2 system. 191 Summary By means of EPR spectroscopy, it was shown that aluminum, gallium and indium nitrate activate hydrogen peroxide and generate hydroxyl and perhydroxyl radicals, with gallium being less effective than aluminum and indium. Surprisingly, only hydroxyl radical species were detected for the reaction systems using aluminum, gallium and indium perchlorate as activator. Moreover, it was confirmed that 2-Me-THF and methanol act as a radical scavengers, whereas the fast formation of the hydroxymethyl radical results in its majority of all radical species found, which explains the milder oxidation of coal in methanol as solvent. Additional D-NMR studies support the obtained data from EPR spectroscopy by showing great differences for the proton exchange between hydrogen and water for the perchlorate, nitrate and chloride anion. Herein, the perchlorate anion shows great potential for activating proton exchange whereas nitrate and chloride anion are much less effective. Minor difference between the various cations (Al3+, Ga3+, In3+) have been also detected. The fastest exchange was observed for the Ga3+ system, followed by indium and aluminum. The size of the cations and their differences in the 1st and/or 2nd coordination sphere only slightly affects the proton exchange whereas the anions reveal a greater influence due to their different abilities in stabilizing or weakening the hydrogen bonds in H2O-H2O molecules. Overall, the conversion of coal under low-severity conditions mediated by the 3+ [Al(H2O)6] /H2O2 system was explored within this thesis. It was shown that lignite can be effectively converted, either into an oil or soluble product, dependent on the choice of solvent. However, this thesis suggests that the conversion of coal mediated by the 3+ [Al(H2O)6] /H2O2 system seems to be more suitable for coal benefication processes, either to improve chemical and physical properties, or combined with an additional further upgrading, such as HDO process for the production of fine chemicals. 192 Appendix 9 Appendix 9.1 Analysis of the obtained soluble products 9.1.1 Soluble products obtained from experiments in methanol As previously described Chapter 4.1 the organic phase obtained from the reaction mixture was treated with an solution of ammonia (24 wt%) to remove Al(III) as a precipitate of Al(OH)3. After removing the solid phase, the solvent (methanol) was removed by rotary evaporation. The obtained products were solid and brownish in appearance. For calculating the carbon mass balance the elemental composition of these fractions (i.e. methanol soluble products and precipitated soluble products) were also analyzed. To estimate the carbon balance, the produced carbon dioxide was sequestrated by a Ba(OH)2 and precipitated as BaCO3. The masses of obtained products (m =[g]) multiplied by the corresponding carbon content (C= [%]) were computed by using following Equation: 풎풄풐풂풍 × 푪풄풐풂풍 = (풎풓풆풔풊풅풖풆 × 푪풓풆풔풊풅풖풆) + (풎풔풐풍풖풃풍풆 풑풓풐풅풖풄풕 × 푪풔풐풍풖풃풍풆 풑풓풐풅풖풄풕) + (풎푨풍(푶푯)ퟑ ∗ 푪푨풍(푶푯)ퟑ) + (풎푩풂푪푶ퟑ ∗ 푪푩풂푪푶ퟑ) Table 9.2 lists C-balance estimated for each reaction. It can be observed that the C- balance is about 0.6 to 0.7 for the conversion of 0.5 g coal. The lack of carbon content is most likely associated with losses in the work-up procedure of the several fractions. 193 Appendix Table 9.1. Weight and elemental composition (dry) of obtained soluble products at 65 °C. Entry Duration Weight H C N S O yield (by [g] difference) 1 Lignite --- 5.2 67.0 0.8 1.4 25.6 ± 0.1 ± 0.5 ± 0.1 ± 0.1 Precipitated solid products 2 5 min 0.41 4.7 5.6 27.6 0 62.1 ± 0.1 ± 0.2 ± 0.5 3 10 min 0.36 4.8 5.1 28.1 1.6 60.5 ± 0.1 ± 0.4 ± 0.8 ± 0.1 4 30 min 0.38 4.7 3.6 27.5 0 64.3 ± 0.1 ± 0.2 ± 0.2 Methanol-soluble products 5 5 min 0.27 2.4 6.4 2.8 0.4 88.0 ± 0.2 ± 0.6 ± 0.3 ± 0.1 6 10 min 0.31 2.6 7.9 4.2 0.4 84.8 ± 0.2 ± 0.7 ± 0.4 ± 0.1 7 30 min 0.43 4.4 26.2 4.9 0 64.3 ± 1.3 ± 7.4 ± 0.6 Table 9.2: Carbon mass balance calculation for the reactions in methanol at 65 °C for varying duration. Experiment C-Balance [%] 5 min 70 10 min 60 30 min 70 194 Appendix 9.1.2 Analysis of obtained soluble products from experiments in ethanol Table 9.3: Weight and elemental composition of obtained ethanol soluble products and precipitated products at 65 °C. Entry Duration Weight H C N S O soluble (by product [%] difference) [g] 1 Lignite --- 5.21 66.99 0.76 1.43 25.6 Precipitated solid products 2 5 min 0.36 5.1 8.3 28.3 0 58.3 ± 0.1 ± 0.3 ± 0.5 3 10 min 0.33 5.5 13.0 26.3 0 55.2 ± 0.1 ± 0.6 ± 0.2 4 30 min 0.37 5.2 8.1 26.8 0 59.9 ± 0.1 ± 0.4 ± 0.8 Ethanol-soluble products 5 5 min 0.52 3.5 21.2 4.0 0.6 70.7 ± 0.2 ± 1.1 ± 0.4 ± 0.1 6 10 min 0.44 3.7 24.3 5.7 0.7 65.6 ± 1.2 ± 0.2 ± 1.1 ± 0.1 7 30 min 0.79 4.5 34.6 6.5 0.8 53.6 ± 0.8 ± 2.7 ± 1.1 ± 0.1 Table 9.4: Carbon mass balance calculation for the reactions in ethanol at different reaction duration and 65 °C. Experiment C-Balance [%] 5 min 90 10 min 80 30 min 98 195 Appendix 9.1.3 Analysis of obtained soluble products from experiments in water Table 9.5: Soluble products obtained from reactions in water at different reaction duration and 65 °C. Duration Weight H C N S O soluble (by product [%] difference) [g] Lignite --- 5.21 66.99 0.76 1.43 25.6 Precipitated solid products 5 min 0.35 4.2 1.0 26.3 0 68.4 ± 0.06 ± 0.1 ± 0.6 10 min 0.29 4.5 1.6 27.6 0.2 66.3 ± 0.1 ± 0.1 ± 0.2 ± 0.1 30 min 0.35 4.4 5.2 24.7 0.5 65.2 ± 0.1 ± 0.9 ± 1.3 ± 0.1 Water-soluble products 5 min 0.27 2.2 2.1 3.4 0.1 92.20 ± 0.5 ± 0.2 ± 0.7 ± 0.1 10 min 0.22 2.2 2.8 3.8 0.2 91.06 ± 0.3 ± 0.05 ± 0.6 ± 0.1 30 min 0.26 2.2 7.1 4.7 0.8 85.2 ± 0.1 ± 0.3 ± 0.4 ± 0.1 Table 9.6: Carbon mass balance calculation for the reactions in water at different reaction duration and 65 °C. Experiment C-Balance [%] 5 min 70 10 min 60 30 min 50 196 Appendix 9.2 EPR measurements Figure 9.1 illustrates EPR spectra obtained from the kinetic study of the Al(ClO4)3/H2O2 system at room temperature. The EPR spectra show the formation of only DMPO-OH adduct and later the degradation product DMPOX while at the same time the signals for DMPO-OH is decreasing. Figure 9.1: EPR spectra obtained from the [Al(H2O)6/H2O2 system ( 1 mmol Al(ClO4)3, 1 mL 7 wt% H2O2) at room temperature. Radical species trapped with DMPO. 197 Appendix 9.3 Variable temperature NMR spectroscopy of the various [M(H2O)6]3+/H2O2 systems 9.3.1 Complete data obtained from D-NMR experiments 9.3.1.1 Al(ClO4)3/H2O2 system Table 9.7: Arrhenius plot parameters obtained from D-NMR experiments with the Al(ClO4)3/H2O2 system T(K) 1/T k(Hz) ln(k) k/T ln(k/T) G‡ (kJ/mol) G‡ (kcal/mol) 218 0,004587 46,1 3,83 0,21 -1,55 45,88 10,97 228 0,004386 70,4 4,25 0,31 -1,18 47,27 11,30 238 0,004202 106,7 4,67 0,45 -0,80 48,60 11,62 248 0,004032 141,1 4,95 0,57 -0,56 50,16 11,99 258 0,003876 188,8 5,24 0,73 -0,31 51,64 12,34 268 0,003731 246,8 5,51 0,92 -0,08 53,13 12,70 273 0,003663 295,8 5,69 1,08 0,08 53,75 12,85 278 0,003597 378,8 5,94 1,36 0,31 54,20 12,96 288 0,003472 787,0 6,67 2,73 1,01 54,49 13,02 298 0,003356 1467,9 7,29 4,93 1,59 54,92 13,13 308 0,003247 2679,1 7,89 8,70 2,16 55,31 13,22 318 0,003145 5074,8 8,53 15,96 2,77 55,50 13,26 328 0,003049 7969,7 8,98 24,30 3,19 56,10 13,41 338 0,002959 13800,0 9,53 40,83 3,71 56,35 13,47 R= 8.3144 J, Tc= 318 °C 218-273 K 273-338 K 6,00 15,00 4,00 10,00 y = -1959,4x + 12,847 R² = 0,9985 2,00 5,00 y = -5564,8x + 25,982 R² = 0,9989 0,00 0,00 0,003 0,0035 0,004 0,0045 0,005 0,0025 0,003 0,0035 0,004 Figure 9.2: Arrhenius plots for the Al(ClO4)3/H2O2 system 198 Appendix 9.3.1.2 Ga(ClO4)3/H2O2 system Table 9.8: Arrhenius plot parameters obtained from D-NMR experiments with the Ga(ClO4)3/H2O2 system T(K) 1/T k(Hz) ln(k) k/T ln(k/T) G‡ (kJ/mol) G‡ (kcal/mol) 218 0,004587 411,6 6,02 1,89 0,64 41,91 10,02 228 0,004386 572,0 6,35 2,51 0,92 43,30 10,35 238 0,004202 728,5 6,59 3,06 1,12 44,80 10,71 248 0,004032 860,9 6,76 3,47 1,24 46,43 11,10 258 0,003876 1066,3 6,97 4,13 1,42 47,92 11,45 268 0,003731 1029,8 6,94 3,84 1,35 49,94 11,94 273 0,003663 1288,2 7,16 4,72 1,55 50,41 12,05 278 0,003597 1716,7 7,45 6,18 1,82 50,71 12,12 288 0,003472 3820,5 8,25 13,27 2,59 50,70 12,12 298 0,003356 5300,0 8,58 17,79 2,88 51,74 12,37 308 0,003247 10500,0 9,26 34,09 3,53 51,81 12,38 318 0,003145 18200,0 9,81 57,23 4,05 52,12 12,46 328 0,003049 21550,0 9,98 65,70 4,19 53,38 12,76 338 0,002959 44600,0 10,71 131,95 4,88 53,05 12,68 R= 8.3144 J, Tc= 298 °C 218-273 K 273-338 K 10,00 15,00 9,00 y = -1136,3x + 11,307 R² = 0,9653 10,00 8,00 7,00 5,00 y = -4886,3x + 25,076 6,00 R² = 0,9916 5,00 0,00 0,003 0,0035 0,004 0,0045 0,005 0,003 0,0032 0,0034 0,0036 0,0038 Figure 9.3: Arrhenius plots for the Ga(ClO4)3/H2O2 system 199 Appendix 9.3.1.3 In(ClO4)3/H2O2 system Table 9.9: Arrhenius plot parameters obtained from D-NMR experiments with the In(ClO4)3/H2O2 system T(K) 1/T k(Hz) ln(k) k/T ln(k/T) G‡ (kJ/mol) G‡ (kcal/mol) 218 0,004587 144,3 4,97 0,66 -0,41 43,81 10,47 228 0,004386 234,8 5,46 1,03 0,03 44,99 10,75 238 0,004202 356,9 5,88 1,50 0,41 46,22 11,05 248 0,004032 520,9 6,26 2,10 0,74 47,46 11,34 258 0,003876 887,7 6,79 3,44 1,24 48,32 11,55 268 0,003731 1327,1 7,19 4,95 1,60 49,38 11,80 273 0,003663 1731,2 7,46 6,34 1,85 49,74 11,89 278 0,003597 2343,4 7,76 8,43 2,13 49,99 11,95 288 0,003472 3675,2 8,21 12,76 2,55 50,80 12,14 298 0,003356 5215,3 8,56 17,50 2,86 51,78 12,38 308 0,003247 9039,5 9,11 29,35 3,38 52,19 12,47 318 0,003145 13485,3 9,51 42,41 3,75 52,91 12,65 328 0,003049 18203,3 9,81 55,50 4,02 53,84 12,87 338 0,002959 26000,0 10,17 76,92 4,34 54,57 13,04 R= 8.3144 J, Tc= 288 °C 218-273 K 273-338 K 10,00 15,00 y = -2661,9x + 17,115 8,00 R² = 0,9936 10,00 6,00 4,00 5,00 2,00 y = -3839,5x + 21,534 R² = 0,9979 0,00 0,00 0,0035 0,004 0,0045 0,005 0,0025 0,003 0,0035 0,004 Figure 9.4: Arrhenius plots for the In(ClO4)3/H2O2 system 200 Appendix 9.3.1.4 Al(NO3)3/H2O2 system Table 9.10: Arrhenius plot parameters obtained from D-NMR experiments with the Al(NO3)3/H2O2 system T(K) 1/T k(Hz) ln(k) k/T ln(k/T) G‡ (kJ/mol) G‡ (kcal/mol) 218 0,004587 1362,6 7,22 6,25 1,83 39,74 9,50 223 0,004484 1291,1 7,16 5,79 1,76 40,80 9,75 228 0,004386 1327,5 7,19 5,82 1,76 41,70 9,97 233 0,004292 1342,3 7,20 5,76 1,75 42,64 10,19 238 0,004202 1501,4 7,31 6,31 1,84 43,37 10,37 248 0,004032 1516,2 7,32 6,11 1,81 45,26 10,82 258 0,003876 1745,3 7,46 6,76 1,91 46,87 11,20 268 0,003731 1500,8 7,31 5,60 1,72 49,10 11,74 273 0,003663 1353,6 7,21 4,96 1,60 50,30 12,02 278 0,003597 1303,3 7,17 4,69 1,55 51,35 12,27 288 0,003472 1282,8 7,16 4,45 1,49 53,32 12,74 298 0,003356 1151,7 7,05 3,86 1,35 55,52 13,27 308 0,003247 1009,7 6,92 3,28 1,19 57,80 13,82 318 0,003145 872,5 6,77 2,74 1,01 60,15 14,38 328 0,003049 720,4 6,58 2,20 0,79 62,65 14,97 338 0,002959 591,3 6,38 1,75 0,56 65,20 15,58 201 Appendix 9.3.1.5 AlCl3/H2O2 system Table 9.11: Arrhenius plot parameters obtained from D-NMR experiments with the Al(NO3)3/H2O2 system T(K) 1/T k(Hz) ln(k) k/T ln(k/T) G‡ (kJ/mol) G‡ (kcal/mol) 218 0,004587 1481,2 7,30 6,79 1,92 39,59 9,46 228 0,004386 1803,8 7,50 7,91 2,07 41,12 9,83 238 0,004202 1905,0 7,55 8,00 2,08 42,90 10,25 248 0,004032 1788,7 7,49 7,21 1,98 44,92 10,74 258 0,003876 1548,2 7,34 6,00 1,79 47,12 11,26 268 0,003731 1327,5 7,19 4,95 1,60 49,38 11,80 273 0,003663 1196,7 7,09 4,38 1,48 50,58 12,09 278 0,003597 1038,2 6,95 3,73 1,32 51,87 12,40 288 0,003472 821,9 6,71 2,85 1,05 54,38 13,00 298 0,003356 684,4 6,53 2,30 0,83 56,81 13,58 308 0,003247 607,5 6,41 1,97 0,68 59,11 14,13 318 0,003145 480,5 6,17 1,51 0,41 61,73 14,75 328 0,003049 336,4 5,82 1,03 0,03 64,73 15,47 338 0,002959 190,2 5,25 0,56 -0,57 68,39 16,35 202 Appendix 9.3.2 Variable temperature 27Al-NMR 9.3.2.1 Al(ClO4)3/H2O2 system 27 Figure 9.5: Variable temperature Al-NMR spectra obtained for the Al(ClO4)3/H2O2 (0.003/ 0.14 mol/L) system. 203 Appendix 9.3.2.2 Al(NO3)3/H2O2 system 27 Figure 9.6: Variable temperature Al-NMR spectra obtained for the Al(NO3)3/H2O2 (0.003/ 0.14 mol/L) system. 204 Appendix 9.3.2.3 AlCl3/H2O2 system 27 Figure 9.7: Variable temperature Al-NMR spectra obtained for the AlCl3/H2O2 (0.003/ 0.14 mol/L) system. 205 Experimental 10 Experimental 10.1 Chemicals All chemicals were purchased from commercial sources and used as received without further purification if not indicated differently. 10.1.1 Coals Lignite (Braunkohle), Fortuna Garsdorf, 3. Sohle, tief Gasflammkohle, Proper/Haniel, Chriemhilt Magerkohle, Niederberg Fettkohle, Westerholt Anthrazit, Sophia Jacoba Coal Reference Material (Leco®, No 502-682) 10.1.2 Solvents Tetrahydrofuran (Aldrich, 99.9 %), 2-Methyltetrahydrofuran (Aldrich, 99.9 %), Methanol (Aldrich, 99.5 %), Ethanol (Aldrich, 99.8 %), 206 Experimental Acetone (Aldrich, 99.9 %) Acetone-d6 (Aldrich, 99.9 %), Dimethyl sulfoxide-d6 (Aldrich, 99.9 %), Tetrahydrofuran-d8 (Aldrich, 99.9 %), Methanol-13C, (Aldrich, 99.9 %) 10.1.3 Inorganic chemicals Manganese dioxide, activated (Aldrich, ~85 %) Aluminum nitrate ∙9H2O(Fluka, 99.9995 %) Gallium nitrate ∙xH2O (Aldrich, >99.999 %) Indium nitrate ∙ xH2O (Aldrich, >99.99 %) Sodium nitrate (Aldrich, >99.99 %) Aluminum perchlorate ∙9H2O (Aldrich, >99,99%) Gallium perchlorate ∙ xH2O (Aldrich, >99,99%) Indium perchlorate ∙ xH2O (Aldrich, >99,99%) Sodium perchlorate ∙H2O (Aldrich, >99,99%) Potassium iodide (Aldrich, >99.5 %) Ni2P/SiO2 (Prepared at the Max-Planck-Institut für Kohlenforschung by Z. Cao) 10.1.4 Other chemicals Tungsten trioxide, (elementar, >99.99 %) Molecular-Sieve Deperox® (Fluka) 207 Experimental 10.2 Experiments 10.2.1 Direct conversion of coal in 2-Me-THF 10.2.1.1 DCL- Effect of temperature In a 5 mL vial 0.69 g of Al(NO3)3 · 9 H2O (1,8 mmol, Fluka) were dissolved in an aqueous solution of hydrogen peroxide (2 mL, 70 wt%, Evonik). This solution was added to a two- neck round-bottom flask containing 15 mL of 2-methyltetrahydrofuran and 0.5 g of Lignite (German mine pit ‘Fortuna Garsdorf, 3.Sohle, tief’). The reaction mixture was heated to temperatures from 40 to 70 °C under magnetic stirring for four hours. Next, the mixture was transferred into a centrifugation tube and the round-bottom flask was well washed with 2- methyl-THF in order to quantitatively transfer the product. Next, it was centrifuged with 6000 rpm for 20 min. The obtained fractions were separated and the solid residues were washed and centrifuged again two times with 2 mL of solvent. The organic solution was washed twice with 5 mL distilled water. The water fractions were collected separately. Here, the residual H2O2 was removed by adding carefully manganese (II) oxide. The solution was filtered and the water slowly removed under reduced pressure. The organic solution was heated up to 85 °C with Deperox® Molecular Sieve (Sigma-Aldrich) in order to remove residual hydrogen peroxide. In sequence, the mixture was also filtered and the solvent was removed by using rotary evaporation (60 °C, 200 rpm, <10 mbar, 15 min). After that, the obtained oil was dissolved in acetone in order to precipitate residual mineral matter, which was then removed by centrifugation. The acetone was removed by rotary evaporation (60 °C, 200 rpm, <10 mbar, 20 min). The samples were additionally dried by a high vacuum pump at room temperature for several hours until constant weights were obtained. 10.2.1.2 DCL- Effect of process duration In a 5 mL vial 0.69 g of Al(NO3)3 · 9 H2O (1,8 mmol, Fluka) were dissolved in an aqueous solution of hydrogen peroxide (2 mL, 70 wt%, Evonik). This solution was added to a two- neck round-bottom flask containing 15 mL of 2-methyltetrahydrofuran and 0.5 g of Lignite (German mine pit ‘Fortuna Garsdorf, 3.Sohle, tief’). The reaction mixture was heated to 65 °C under magnetic stirring for 30, 60, 90, 120, 150, 180 and 240 min. Next, the mixture 208 Experimental was transferred into a centrifugation tube and the round-bottom flask was well washed with 2-methyl-THF in order to quantitatively transfer the product. Next, it was centrifuged with 6000 rpm for 20 min. The obtained fractions were separated and the solid residues were washed and centrifuged again two times with 2 mL of solvent. The organic solution was washed twice with 5 mL distilled water. The water fractions were collected separately. Here, the residual H2O2 was removed by adding carefully manganese (II) oxide. The solution was filtered and the water slowly removed under reduced pressure. The organic solution was heated up to 85 °C with Deperox® Molecular Sieve (Sigma-Aldrich) in order to remove residual hydrogen peroxide. In sequence, the mixture was also filtered and the solvent was removed by using rotary evaporation (60 °C, 200 rpm, <10 mbar, 15 min). After that, the obtained oil was dissolved in acetone in order to precipitate residual mineral matter, which was then removed by centrifugation. The acetone was removed by rotary evaporation (60 °C, 200 rpm, <10 mbar, 20 min). The samples were additionally dried by a high vacuum pump at room temperature for several hours until constant weights were obtained. 10.2.1.3 DCL- Effect of coal rank In a 5 mL vial 0.69 g of Al(NO3)3 · 9 H2O (1,8 mmol, Fluka) were dissolved in an aqueous solution of hydrogen peroxide (2 mL, 70 wt%, Evonik). This solution was added to a two- neck round-bottom flask containing 15 mL of 2-methyltetrahydrofuran and 0.5 g of coal. The reaction mixture was heated to 65 °C under magnetic stirring for 240 min. Next, the mixture was transferred into a centrifugation tube and the round-bottom flask was well washed with 2-methyl-THF in order to quantitatively transfer the product. Next, it was centrifuged with 6000 rpm for 20 min. The obtained fractions were separated and the solid residues were washed and centrifuged again two times with 2 mL of solvent. The organic solution was washed twice with 5 mL distilled water. The water fractions were collected separately. Here, the residual H2O2 was removed by adding carefully manganese (II) oxide. The solution was filtered and the water slowly removed under reduced pressure. The organic solution was heated up to 85 °C with Deperox® Molecular Sieve (Sigma-Aldrich) in order to remove residual hydrogen peroxide. In sequence, the mixture was also filtered and the solvent was removed by using rotary evaporation (60 °C, 200 rpm, <10 mbar, 15 min). After that, the obtained oil was dissolved in acetone in order to precipitate residual mineral matter, which 209 Experimental was then removed by centrifugation. The acetone was removed by rotary evaporation (60 °C, 200 rpm, <10 mbar, 20 min). The samples were additionally dried by a high vacuum pump at room temperature for several hours until constant weights were obtained. 10.2.2 Direct conversion of coal- Solvent effect Lignite was dispersed in a solvent (MeOH, EtOH, H2O, 2-Me-THF). The suspension was 3+ heated up to 65 °C under magnetic stirring. Before the solution of [Al(H2O)6] /H2O2 was added, the whole system was flushed for several minutes with Argon to remove air and then connected to a washing flask with Ba(OH)2 solution for the sequestration of CO2 eventually liberated from the oxidation processes. After the reaction, the mixture was cooled down and the solid and liquid phases were separated via filtration. The obtained lignite residue was first washed with the solvent to collect the soluble-products. Next, the coal residue was washed once with 10 mL of water to remove remaining aluminum salt. The residues were dried overnight in a vacuum oven at 60 °C. The work-up of the obtained organic solutions included first the elimination of residual hydrogen peroxide with manganese dioxide. After filtration, a stoichiometric quantity of ammonia solution (25%) was added to the solution in order to remove aluminum ions as aluminum hydroxide [Al(OH)3]. Again, the mixture was filtered. To collect all organic matter, the obtained brownish solid was washed with solvent until it remained colorless. The light- yellowish solid was dried overnight in a vacuum oven at 65 °C. The organic phases were collected and the solvent was removed using a rotary evaporator (60 °C, 200 rpm, 20 min). A brownish solid was obtained. All reaction products were analyzed by elemental analysis (CHNS/O). The lignite residues were additionally analyzed using FTIR and 13C SS-NMR spectroscopy. 210 Experimental 10.3 Hydrodeoxygenation of lignite and lignite residue samples -5 0.5 g lignite or lignite residue sample, 0.2 g Ni2P/SiO2, 5 mL n-Octane and 8x10 mol n- hexadecane (ISTD) were placed in an autoclave insert (10 mL). The autoclave was flushed and loaded with H2 (100 bar, r.t.). The experiment was performed at 300 °C for 20 h. 5 mL ethyl acetate/octane (50:50, v,v) was added to the reaction mixture after reaction. A sample for GC analysis was taken, dried with anhydrous Na2SO4 and filtered before the analysis. 10.4 Activation of H2O2 10.4.1 Activation of H2O2 – Evaluation of different metals 1,8 mmol of the metal nitrate was dissolved in H2O2 (2 mL, 70 wt%) and given into mL THF (15 mL) in a 2-neck round bottom flask which was already heated up to 55 °C. The reaction mixture was stirred for ca. 1-2s before an aliquot was taken for H2O2 determination by iodometric titration. Further aliquots for titration were taken after: 15, 30, 45, 60, 90, 120, 150, 180, 210, 240, 300, 360 and 420 min process duration. 10.5 Analysis Methods 10.5.1 Determination of water-content in product oil samples (Karl-Fischer Titration) The residual water-content of obtained product oil samples was determined by Karl-Fischer Titration using a Methrom Coloumetric Titrator 756. 211 Experimental 10.5.2 Determination of moisture (ASTM D3173) 50-100 mg coal or coal residue sample is weighed and placed in the preweighed quartz glass crucible. The sample is then placed in a furnace operating at 107 °C for 1 h. 10.5.3 Determination of Ash content (ASTM D3174) 50-100 mg coal or product oil or coal residue sample is weighed and placed in the preweighed quartz glass crucible. The sample is then placed in a cold muffle furnace, and the temperature is raised at such a rate that it reaches 450-500 °C at the end of 1 h. Heating is continued so that a temperature of 700-750 °C is reached at the end of 2 h and this temperature is held for an additional 2 h. During the ashing procedure, a sufficient supply of air must be supplied to the furnace. Calculation of ash content: 푎 × 100 100 퐴푠ℎ (푤푓) = × 푏 100 − 퐻2푂 wf= water-free, a= Ash content of wet sample [g], b= initial sample weight (wet), H2O in wt% of sample 10.5.4 Determination of Volatile matter (ASTM D3175) 50-100 mg coal or product oil or coal residue sample is weighed and placed in the preweighed quartz glass crucible with a close-fitting cover. The crucible is then placed in a muffle furnace operating at 950 ± 20 °C. After heating for exactly 7 min, the crucible is removed and cooled down. Calculation of volatile matter: 푎 − 푏 푉푀 = × 100 − 퐻 푂 푎 2 VM= volatile matter, a= wet sample [g], b= residue sample [g], H2O in wt% of sample 212 Experimental 10.5.5 Elemental analysis For the determination of CHNS a “Vario Micro Cube” from elementar was used. 2-3 mg of a sample was weighed into a tin ship. To enhance the combustion, a small spatula tip of WO3 (ca. 1mg, powder) was added. During the measurement, the combustion tube was supplied with oxygen for 80 s. Calibration substances for the device have been: Sulfanilamide (Merck), acetanilide (Merck), coal standard (Leco®). Equation for dry and ash free (daf) values: 퐸퐴𝑖 ∗ 100 퐸퐴(푑푎푓) = 100 − (퐴푠ℎ + 퐻2푂) EA= value for C, H, N or S, EAi = determined value (C,H,N or S), Ash= determined Ash- content, H2O= determined water-content 10.5.6 FTIR analysis FTIR analyses were performed unsing “Hyperion 3000” with ATR-Tensor 27 (Bruker). Spectra were recorded with Opus Software. Wavelength from 4000 – 600 cm-1, 128 scans. 10.5.7 NMR 10.5.7.1 Liquid-NMR The 1H, 13C were performed on 400 MHz NMR, the two dimensional experiments on 600 MHz NMR Spectrometer (Bruker). The product oil sample was dissolved in acetone-d6 (ca. 40 mg, 0.7 mL). 213 Experimental 10.5.7.2 Solid state-NMR The 13C SS-NMR spectra for the product oil samples were performed on 300 MHz NMR spectrometer (Bruker), at a rotation rate of 2 kHz (7 mm rotor, d=60 s, 2000 scans). The 13C SS-NMR spectra for lignite and lignite residue samples were performed on 300 MHz NMR spectrometer (Bruker) at a rotation rate of 4 kHz (7 mm rotor, d=2 s, 64 scans). 10.5.7.3 Dynamic 1H-NMR studies of the [M(H2O)6]3+/H2O2 systems 1 3+ The H-NMR spectra of the solutions of the various [M(H2O)6] /H2O2 in THF-d8 were recorded at temperatures from -55 to 65 °C using a Bruker AV VIII 400 microbay spectrometer operating at 400 MHz. Low temperature calibration was performed using a NMR reference standard of 4% methanol in methanol-d4 (Sigma Aldrich). The test solutions 3+ were prepared in 0.7 mL THF-d8 (Aldrich). The concentrations of [M(H2O)6] and H2O2 were -1 3+ always 0.003 and 0.14 mol L , respectively. The [Al(H2O)6] /H2O2 system was tested using Al(NO3)3•9H2O (99,999%, Sigma Aldrich), Al(ClO4)3•9H2O (98%, Sigma Aldrich) and AlCl3•6H2O (≥ 99%, Sigma Aldrich) and 60 wt% H2O2 (Evonik). 10.5.8 GPC analysis To analyze the apparent molecular weight distribution of obtained product oils from coal conversion, the samples (10-20 mg) were dissolved in THF (2 mL) and filtered prior to injection. The GPC analysis were performed at 25 °C on a Perkin-Elmer HPLC 200 equipped with 4 coloumns (2xTSKgel Super HZ1000, TSKgel Super HZ2000, TSKgel Super HZ3000, 4.6 mm x ID 15.0 cm, Tosoh Bioscience), and using inhibitor-free THF as eluent (0.25 mL min-1, Aldrich). For detection, a diode-array detector was used. The reported results show the chromatogram at 225 nm. The DAD response was normalized by the sample weight. The apparent molecular weight distribution is given to polysterene standards (200 – 66,000 Da, Aldrich), and thus is only for a relative assessment of the overall changes in the apparent molecular weight distributions. 214 Experimental 10.5.9 Thermal degradation analysis The weight loss profile of the indicated samples was measured on a Mettler Toledo TGA/DSC 1 Star System operating from 30 to 1000 °C at 5 °C min-1 under argon (60 mL min-1). DTG data were smoothed using Origin 9.1, “Savitzky-Golay”-Method, 50 points, 2nd polynomial order. 10.5.10 Determination of hydrogen peroxide (iodometric titration) To an Erlenmeyer flask, 50 mL of aqueous 20 wt% acetic acid and 20 g of dry ice for deaeration of the solution were added. After 2 min, ca. 2.0 g of potassium iodide (Aldrich; p.a.) and 3 drops of a 1-wt% ammonium molybdate solution were added. To this mixture, a 100-150 mg aliquot of the reaction mixture, collected before and during the reaction, was added. The iodine formed was titrated with a 0.1 mol L−1 solution of sodium thiosulfate (Merck; p.a.). The endpoint was detected using a manual titrator “Titrino 848 plus” from Metrohm. The amount of H2O2, n(H2O2) [mmol] in the reaction mixture was calculated by 퐶(푆 푂2−) × 푉(푆 푂2−) 푚(푟푒푎푐푡𝑖표푛 푚𝑖푥) 푛(퐻 푂 ) = 2 3 2 3 × × 103 2 2 2 푚(푎푙𝑖푞푢표푡) C=concentration of sodium thiosulfate solution [mol/L], V=Volume of thiosulfate [L]; m = mass of reaction mixture and aliquot taken during the reaction 10.5.11 Determination of pH For the metering the pH values, a pH (glass-) electrode and the Titrino 848 plus were used. 215 Experimental 10.5.12 Scanning Electron Microscope (SEM) SEM pictures of coal residue samples were obtained using a Hitachi S-3500N Scanning Electron Microscope with EDS Detector. Sample preparation and measurements were performed by the Electron Microscopy department of the Max-Planck-Institut für Kohlenforschung. 10.5.13 in situ EPR spectroscopy In situ EPR measurements in X-band (microwave frequency 9.8 GHz) were performed at 300 K with a Bruker EMX CW-micro spectrometer equipped with an ER 4119HS-WI high- sensitivity optical resonator with a grid in the front side. The samples were measured in a home-made setup which compromise an aqua-cell equipped with an GC capillary (see Fig.S1). g values have been calculated from the resonance field B0 and the resonance frequency using the resonance condition h = gB0. The calibration of the g values was performed using DPPH (2,2-diphenyl-1-picrylhydrazyl) (g = 2.0036±0.00004). EPR simulation were performed using the program EPRsim32.173, 174 Figure 10.1: Homemade setup for in situ EPR measurements 216 Experimental 10.5.14 Gas chromatography 10.5.14.1 GCxGC-qMS of HDO reaction products The reaction mixtures were analyzed using 2D GC×GC−MS (1st column: ZB-1HT Inferno 30 m, 0.25 mm ID, df 0.25 µm; 2nd column: BPX50, 1 m, 0.15 mm ID, df 0.15 µm) in a GC- MS 2010 Plus (Shimadzu) equipped with a ZX1 thermal modulation system (Zoex). The injector temperature was 300 °C. The temperature program began at 40 °C for 5 min, and subsequently was increased at a rate of 5.2 °C/min until reaching a temperature of 300 °C; the program culminated with an isothermal step at 300 °C for 5 min. The modulation applied for the comprehensive GC×GC analysis was a hot jet pulse (400 ms) every 9000 ms. The 2D chromatograms were processed with GC Image software (Zoex). The products were identified by a search of the MS spectrum with the MS libraries NIST 08, NIST 08s, and Wiley 9. The semi-quantification of the products was performed using integration of GC peaks. 10.5.14.2 GC-FID/MS analysis of carvone epoxidation products The reaction mixtures were analyzed using GC-FID/qMS (1st column: ZB-1HT Inferno 30 m, 0.25 mm ID, df 0.25 µm; 2nd column: BPX50, 1 m, 0.15 mm ID, df 0.15 µm) in a GC-MS 2010 Plus (Shimadzu) equipped with a ZX1 thermal modulation system (Zoex). The injector temperature was 230 °C. The temperature program began at 40 °C, and subsequently was increased at a rate of 6 °C/min until reaching a temperature of 180 °C. The heating rate was increased to 18 °C/min until reaching a temperature of 280 °C. Lastly, the heating rate was decreased to 5 °C/min until 350 °C was reached. The final temperature of 350 °C was hold for 5 min. The products were identified by a search of the MS spectrum with the MS libraries NIST 08, NIST 08s, and Wiley 9. The semi-quantification of the products was performed using integration of GC peaks. 217 Experimental Preparation of carvone epoxide as GC-standard substance A solution of R-carvone (50 g, 0.332 M) in methanol (120 mL) was added to a three-necked round bottomed flask (1 L) fitted with a pressure equalizing addition funnel. An aqueous solution of 30% H2O2 (100 mL) was added, and the reaction mixture was cooled to 5°C. A solution of 0.289 N K2CO3 (100 mL) was then added in a dropwise manner over a 45-min period with the temperature maintained below 10 °C. When addition was complete, the reaction was allowed to warm to 25 °C and stirred until GC analysis (OV-17) showed the reaction to be complete (ca. 5 h). The reaction mixture was then poured into ice water (400 mL), salted (NaCl), and extracted with ether (5 x 150 mL). The extracts were combined, dried (Na2SO4), and concentrated in vacuo to afford a colorless oil (55.0 g) which was distilled (70 °C/0.6 mmHg) to yield a distillate (47.5 g, 86%) whose physical and spectral properties were identical with those reported by Wycpalek' for 2. GC analysis (6 ft 3% OV- 17,100 °C 4 min to 180°C at 7.5° per min) of the product showed two peaks in ca. a 19 to 1 ratio.18 GC/MS analysis (6 ft, 5% FFAP, 145 "C) showed both peaks had M+ = 166 and similar fragmentation patterns. 218 Literature 11 Literature 1. J. T. Bartis, Camm, F., Ortiz, D. S., Journal, 2008. 2. www.worldcoal.org, Where is coal found?). 3. Bundesverband Braunkohle (DEBRIV), Braunkohle in Deutschland, 2011. 4. S. Vasireddy, B. Morreale, A. Cugini, C. Song and J. J. Spivey, Energ Environ Sci, 2011, 4, 311-345. 5. Word Energy Council, Coal Recoverable Reserves By Region, 2011. 6. W. C. Association, Uses of coal, http://www.worldcoal.org/coal/uses-of-coal/, (accessed 12/03/2013, 2013). 7. D. W. Van Krevelen, in Coal, Elsevier, 1993, pp. 1-4. 8. R. C. Neavel, M. L. Gorbaty, J. W. Larsen and I. Wender, Coal Science Vol. 1, 1982. 9. BGR, Statistik der Kohlenwirtschaft e.V. 10. E. Lindner, Chemie für Ingenieure, Lindner Verlag Karlsruhe. 11. E. Jin, Y. Zhang, L. He, H. G. Harris, B. Teng and M. Fan, Applied Catalysis A: General, 2014, 476, 158-174. 12. J. G. Speight, The Chemistry and Technology of Coal, Third Edition, CRC Press, Taylor and Francis Group, 2013. 13. F. Bergius, Chemical reactions under high pressure, http://www.nobelprize.org/nobel_prizes/chemistry/laureates/1931/bergius- lecture.html, (accessed 10/17/2011, 2011). 14. H. Shui, Z. Cai and C. Xu, Energies, 2010, 3, 155-170. 15. J. G. Speight, in The Chemistry and Technology of Coal, Third Edition, CRC Press, Taylor and Francis Group, LLC, 2013, ch. 19, p. 595. 16. W. N. Mitchell, K. L. Trachte and S. Zaczepinski, Ind Eng Chem Prod Rd, 1979, 18, 311-314. 17. J. E. S and W. R. H, Journal, 1970. 18. D. o. t. a. industry, Coal Liquefaction Technology Status Report, 1999. 19. M. Höök and K. Aleklett, International Journal of Energy Research, 2010, 34, 848- 864. 219 Literature 20. N. A. o. Science, Journal, 2009. 21. M. E. Dry, Catal Today, 2002, 71, 227-241. 22. in Studies in Surface Science and Catalysis, eds. A. I. E. W. Q. I. P. S. Toshiaki Kabe and K. Yaeko, Elsevier, 2004, vol. Volume 150, pp. 181-267. 23. Y.-N. Wang, Y.-Y. Xu, Y.-W. Li, Y.-L. Zhao and B.-J. Zhang, Chemical Engineering Science, 2003, 58, 867-875. 24. Y. Bi and A. K. Dalai, Canadian Journal of Chemical Engineering, 2003, 81, 230- 242. 25. R. H. Crabtree, in Studies in Surface Science and Catalysis, eds. H. E. Curry-Hyde and R. F. Howe, Elsevier, 1994, vol. Volume 81, pp. 85-92. 26. J. G. Speight, in The chemistry and Technology of Coal, Third Edition, CRC Press, Taylor and Francis Group, 2013, ch. 19, p. 599. 27. R. B. Anderson, B. Seligman, J. F. Shultz, R. Kelly and M. A. Elliott, Industrial & Engineering Chemistry, 1952, 44, 391-397. 28. M. E. Dry, T. Shingles, L. J. Boshoff and G. J. Oosthuizen, Journal of Catalysis, 1969, 15, 190-199. 29. H. Zhang, H. Ma, H. Zhang, W. Ying and D. Fang, Catal Lett, 2012, 142, 131-137. 30. J.-Y. Park, Y.-J. Lee, P. R. Karandikar, K.-W. Jun, J. W. Bae and K.-S. Ha, Journal of Molecular Catalysis A: Chemical, 2011, 344, 153-160. 31. J. Xiong, Y. Ding, T. Wang, Y. Lü, H. Zhu, H. Luo and L. Lin, Chinese Journal of Catalysis, 2005, 26, 874-878. 32. A. R. de la Osa, A. De Lucas, A. Romero, J. L. Valverde and P. Sánchez, Fuel, 2011, 90, 1935-1945. 33. J. Yang, Y. Sun, Y. Tang, Y. Liu, H. Wang, L. Tian, H. Wang, Z. Zhang, H. Xiang and Y. Li, Journal of Molecular Catalysis A: Chemical, 2006, 245, 26-36. 34. G. Connell and J. A. Dumesic, Journal of Catalysis, 1985, 92, 17-24. 35. J. Li, C. Zhang, X. Cheng, M. Qing, J. Xu, B. Wu, Y. Yang and Y. Li, Applied Catalysis A: General, 2013, 464–465, 10-19. 36. C.-H. Zhang, Y. Yang, B.-T. Teng, T.-Z. Li, H.-Y. Zheng, H.-W. Xiang and Y.-W. Li, Journal of Catalysis, 2006, 237, 405-415. 37. V. R. S. E. E. M. T. J. K. A. K. Dalai, International Journal of Chemical Reactor Engineering, 2012, 10. 38. C. H. Bartholomew, Applied Catalysis A: General, 2001, 212, 17-60. 39. J. G. Speight, The Biofuels Handbook, RSC Publishing, 2011. 40. R. H. Williams and E. D. Larson, Energy for Sustainable Development, 2003, 7, 103- 129. 220 Literature 41. ExxonMobil, Methanol to Gasoline (MTG)-Production of Clean Gasoline from Coal. 42. T. N. A. Press, Journal, 2009. 43. M. Hindman, presented in part at the World CTL Conference, 2010. 44. W. Fuchs, Die Chemie der Kohle, Springer Verlag, Berlin, 1931. 45. B. Juettner, R. C. Smith and H. C. Howard, Journal of the American Chemical Society, 1935, 57, 2322-2326. 46. R. Hayatsu, R. G. Scott, L. P. Moore and M. H. Studier, Nature, 1975, 257, 378-380. 47. N. C. Deno, B. A. Greigger and S. G. Stroud, Fuel, 1978, 57, 455-459. 48. D. W. Van Krevelen, in Coal Science, Elsevier, 3 edn., 1993, ch. 20, pp. 605-626. 49. A. Tahmasebi, Y. Jiang, J. Yu, X. Li and J. Lucas, Fuel Processing Technology, 2015, 129, 213-221. 50. Y. Tian, Z. Qin and B. Li, International Journal of Mining Science and Technology, 2012, 22, 731-734. 51. D. J. Boron and S. R. Taylor, Fuel, 1985, 64, 209-211. 52. I. Heard and F. E. Senftle, Fuel, 1984, 63, 221-226. 53. M. Sugano, R. Ikemizu and K. Mashimo, Fuel Processing Technology, 2002, 77–78, 67-73. 54. A. H. Clemens, T. W. Matheson and D. E. Rogers, Fuel, 1991, 70, 215-221. 55. Y. F. Patrakov, O. N. Fedyaeva, S. A. Semenova, N. I. Fedorova and L. V. Gorbunova, Fuel, 2006, 85, 1264-1272. 56. F. S. Fischer, H., Gesammelte Abhandlung zur Kenntnis der Kohle, 1920. 57. N. W. Franke, M. W. Kiebler, C. H. Ruof, T. R. Savich and H. C. Howard, Industrial & Engineering Chemistry, 1952, 44, 2784-2792. 58. A. N. Roy and H. C. Howard, Journal of the American Chemical Society, 1952, 74, 3239-3242. 59. B. Juettner, R. C. Smith and H. C. Howard, Journal of the American Chemical Society, 1937, 59, 236-241. 60. W. A. Bone, L. Horton and S. G. Ward, Proceedings of the Royal Society of London. Series A, Containing Papers of a Mathematical and Physical Character, 1930, 127, 480-510. 61. W. A. Bone and R. Quarendon, Proceedings of the Royal Society of London. Series A, Containing Papers of a Mathematical and Physical Character, 1926, 110, 537- 542. 221 Literature 62. W. A. Bone, L. G. B. Parsons, R. H. Sapiro and C. M. Groocock, Proceedings of the Royal Society of London. Series A, Mathematical and Physical Sciences, 1935, 148, 492-522. 63. C. R. Kinney, Journal of the American Chemical Society, 1947, 69, 284-289. 64. C. R. Kinney and K. F. Ockert, Industrial & Engineering Chemistry, 1956, 48, 327- 332. 65. H. H. Honda, Y., Fuel, 1958, 37. 66. J. G. Bimer, P.H.; Raj, S., Am. Chem. Soc. Symp. Ser., 1987, 71. 67. C. S. C. Lynch, A.R., Fuel, 1932, 11. 68. R. Belcher, J. Soc. Chem. Ind. (London), 1948, 67. 69. C. R. Kinney and L. D. Friedman, Journal of the American Chemical Society, 1952, 74, 57-61. 70. R. Hayatsu, R. E. Winans, R. L. McBeth, R. G. Scott, L. P. Moore and M. H. Studier, Nature, 1979, 278, 41-43. 71. R. Hayatsu, R. G. Scott, L. P. Moore and M. H. Studier, Nature, 1975, 257, 378-378. 72. O. K. L and G. L. V. E, in Oxidation of Organic Compounds, AMERICAN CHEMICAL SOCIETY, 1968, vol. 76, ch. 54, pp. 382-388. 73. R. A. Sheldon, I.; Hanefeld, U., in Green Chemistry and Catalysis, WILEY-VCH Verlag GmbH & Co. KGaA, 2007, ch. 4, pp. 133-212. 74. N. A. M. Fadzil, M. H. A. Rahim and G. P. Maniam, Chinese Journal of Catalysis, 2014, 35, 1641-1652. 75. J. Wahlen, D. E. De Vos and P. A. Jacobs, Organic Letters, 2003, 5, 1777-1780. 76. R. J. Watts, M. D. Udell, P. A. Rauch and S. W. Leung, Hazardous Waste and Hazardous Materials, 1990, 7, 335-345. 77. R. J. Watts, M. D. Udell, S. Kong and S. W. Leung, Environmental Engineering Science, 1999, 16, 93-103. 78. J. J. Pignatello, Environmental Science & Technology, 1992, 26, 944-951. 79. J. J. Pignatello and K. Baehr, J. Environ. Qual., 1994, 23, 365-370. 80. in Applications of Hydrogen Peroxide and Derivatives, eds. C. W. Jones and J. H. Clark, The Royal Society of Chemistry, 1999, DOI: 10.1039/9781847550132-00037, pp. 37-78. 81. H. J. H. Fenton, Journal of the Chemical Society, Transactions, 1894, 65, 899-910. 82. B. Petri, R. Watts, A. Teel, S. Huling and R. Brown, in In Situ Chemical Oxidation for Groundwater Remediation, eds. R. L. Siegrist, M. Crimi and T. J. Simpkin, Springer New York, 2011, vol. 3, ch. 2, pp. 33-88. 222 Literature 83. A. Neaman, F. Mouélé, F. Trolard and G. Bourrié, Applied Geochemistry, 2004, 19, 973-979. 84. C. Walling, Accounts of Chemical Research, 1975, 8, 125-131. 85. R. G. Zepp, B. C. Faust and J. Hoigne, Environmental Science & Technology, 1992, 26, 313-319. 86. J. D. Rush, Z. Maskos and W. H. Koppenol, in Methods in Enzymology, ed. A. N. G. Lester Packer, Academic Press, 1990, vol. Volume 186, pp. 148-156. 87. W. H. Koppenol and J. F. Liebman, The Journal of Physical Chemistry, 1984, 88, 99- 101. 88. D. R. Lide in Handbook of Chemistry and Physics, CRC Press, Taylor and Francis Group, 2006, ch. 5. 89. S. H. Bossmann, E. Oliveros, S. Göb, S. Siegwart, E. P. Dahlen, L. Payawan, M. Straub, M. Wörner and A. M. Braun, The Journal of Physical Chemistry A, 1998, 102, 5542-5550. 90. I. B. Afanas'ev, Superoxide Ion Chemistry and Biological Implications, CRC Press, Boca Raton, FL, 1989. 91. G. V. Buxton, C. L. Greenstock, W. P. Helman and A. B. Ross, Journal of Physical and Chemical Reference Data, 1988, 17, 513-886. 92. R. H. Crabtree, Chem. Rev., 1995, 95, 987. 93. A. S. Novikov, M. L. Kuznetsov, A. J. L. Pombeiro, N. A. Bokach and G. B. Shul’pin, ACS Catalysis, 2013, 3, 1195-1208. 94. G. B. Shul’pin, Mini-Rev. Org. Chem., 2009, 6, 95. 95. F. Haber and J. Weiss, Naturwissemschaften, 1932, 20, 948. 96. Y. N. Kozlov, A. D. Nadezhdin and A. P. Purmal’, Kinet. Katal., 1973, 14, 141. 97. K. Takai, K. Oshima and H. Nozaki, Tetrahedron Lett, 1980, 21, 1657-1660. 98. I. M. Martynova, L. P. Stepovik and V. A. Dodonov, Russian Journal of General Chemistry, 2001, 71, 736-741. 99. Z. Lei, G. Ma, L. Wei, Q. Yang and B. Su, Catal Lett, 2008, 124, 330-333. 100. R. Rinaldi, H. F. N. de Oliveira, H. Schumann and U. Schuchardt, J Mol Catal a- Chem, 2009, 307, 1-8. 101. D. Mandelli, K. C. Chiacchio, Y. N. Kozlov and G. B. Shul’pin, Tetrahedron Lett., 2008, 49, 6693. 102. J. Burgess, Journal, 1984, 24. 103. D. T. Richens, The Chemistry of Aqua Ions, Wiley, Chichester, 1997. 104. H. Ohtaki and T. Radnai, Chemical Reviews, 1993, 93, 1157-1204. 223 Literature 105. L. Helm, G. M. Nicolle and R. E. Merbach, in Advances in Inorganic Chemistry, Academic Press, 2005, vol. Volume 57, pp. 327-379. 106. L. Helm and A. E. Merbach, Coordination Chemistry Reviews, 1999, 187, 151-181. 107. T. Kowall, P. Caravan, H. Bourgeois, L. Helm, F. P. Rotzinger and A. E. Merbach, Journal of the American Chemical Society, 1998, 120, 6569-6577. 108. W. Jiang, J. D. Gorden and C. R. Goldsmith, Inorganic Chemistry, 2012, 51, 2725- 2727. 109. A. G. DiPasquale and J. M. Mayer, Journal of the American Chemical Society, 2008, 130, 1812-1813. 110. P. P. Pescarmona and P. A. Jacobs, Catal Today, 2008, 137, 52-60. 111. M. A. Uguina, J. A. Delgado, A. Rodríguez, J. Carretero and D. Gómez-Díaz, Journal of Molecular Catalysis A: Chemical, 2006, 256, 208-215. 112. B. C. Ranu and U. Jana, The Journal of Organic Chemistry, 1998, 63, 8212-8216. 113. M. Mahesh, J. A. Murphy and H. P. Wessel, The Journal of Organic Chemistry, 2005, 70, 4118-4123. 114. J. R. Rodrı́guez and A. Navarro, Tetrahedron Lett, 2004, 45, 7495-7498. 115. B. Gao, Y. Wen, Z. Yang, X. Huang, X. Liu and X. Feng, Advanced Synthesis & Catalysis, 2008, 350, 385-390. 116. R. Dimitrova, Y. Neinska, M. Mihályi, T. Tsoncheva and M. Spassova, Reaction Kinetics and Catalysis Letters, 2001, 74, 353-362. 117. J. G. Speight, Applied Spectroscopy Reviews, 1994, 29, 117-169. 118. H. L. C. Meuzelaar, Advances in Coal Spectroscopy, Plenum Press, 1992. 119. M. M. Hesse, H.; Zeeh, B., Spectroscopic Methods in Organic Chemistry, Georg Thieme Verlag, Stuttgart, 2nd Edition edn., 2008. 120. W. Geng, T. Nakajima, H. Takanashi and A. Ohki, Fuel, 2009, 88, 139-144. 121. J. Ibarra, E. Muñoz and R. Moliner, Organic Geochemistry, 1996, 24, 725-735. 122. P. R. Solomon and R. M. Carangelo, Fuel, 1988, 67, 949-959. 123. P. R. Solomon and R. M. Carangelo, Fuel, 1982, 61, 663-669. 124. B. Saikia, R. K. Boruah and P. K. Gogoi, J Earth Syst Sci, 2007, 116, 575-579. 125. B. K. B. Saikia, R. K.; Gogoi, P. K., Earth System Science, 2007, 116, 575-579. 126. J. G. Speight, Applied Spectroscopy Reviews, 1972, 5, 211-263. 127. S. Kalaitzidis, A. Georgakopoulos, K. Christanis and A. Iordanidis, Geochimica et Cosmochimica Acta, 2006, 70, 947-959. 224 Literature 128. B.-O. Erdenetsogt, I. Lee, S. K. Lee, Y.-J. Ko and D. Bat-Erdene, International Journal of Coal Geology, 2010, 82, 37-44. 129. T. H. Fletcher, A. R. Kerstein, R. J. Pugmire, M. S. Solum and D. M. Grant, Energy & Fuels, 1992, 6, 414-431. 130. P. R. Solomon, M. A. Serio and E. M. Suuberg, Progress in Energy and Combustion Science, 1992, 18, 133-220. 131. S. Ma, J. Hill and S. Heng, Journal of Thermal Analysis and Calorimetry, 1989, 35, 977-988. 132. W.-C. Xu and A. Tomita, Fuel, 1987, 66, 627-631. 133. R. J. Tyler, Fuel, 1979, 58, 680-686. 134. R. Rinaldi, F. Y. Fujiwara and U. Schuchardt, Journal of Catalysis, 2007, 245, 456- 465. 135. Y. Kamiya, S. Futamura, T. Mizuki, M. Kajioka and K. Koshi, Fuel Processing Technology, 1986, 14, 79-90. 136. F. J. Derbyshire, P. Varghese and D. D. Whitehurst, Fuel, 1982, 61, 859-864. 137. E. Dorrestijn, L. J. J. Laarhoven, I. W. C. E. Arends and P. Mulder, Journal of Analytical and Applied Pyrolysis, 2000, 54, 153-192. 138. E. Finkelstein, G. M. Rosen and E. J. Rauckman, Archives of Biochemistry and Biophysics, 1980, 200, 1-16. 139. X. Zhang, Q. Zhang, A. Zhao, J. Guan, D. He, H. Hu and C. Liang, Energy & Fuels, 2010, 24, 3796-3803. 140. Z. He and X. Wang, Journal, 2012, 1, 28. 141. J.-S. Moon, E.-G. Kim and Y.-K. Lee, Journal of Catalysis, 2014, 311, 144-152. 142. X. Zhang, Q. Zhang, J. Guan, D. He, H. Hu and C. Liang, Asia-Pacific Journal of Chemical Engineering, 2009, 4, 574-580. 143. D. Mandelli, K. C. Chiacchio, Y. N. Kozlov and G. B. Shul’pin, Tetrahedron Lett, 2008, 49, 6693-6697. 144. E. G. R. Eaton S.S., Berliner L.J., Biomedical EPR, http://www.bruker.com/fileadmin/user_upload/8-PDF- Docs/MagneticResonance/EPR_brochures/superoxide.pdf, (accessed 08/26/2013, 2013). 145. E. Industries, About Hydrogen Peroxide - Storage, http://h2o2.evonik.com/product/h2o2/en/about-hydrogen-peroxide/basic- information/storage/pages/default.aspx, (accessed 10/06/2014, 2014). 146. G. R. Buettner, Free Radical Bio Med, 1987, 3, 259-303. 225 Literature 147. I. Afanas'ev, Superoxide Ion: Chemistry and Biological Implications, CRC Press, 1989. 148. W. R. Hagen, Biomolecular EPR spectrocopy, CRC Press, 2009. 149. S. Dikalov, J. Jiang and R. P. Mason, Free Radical Research, 2005, 39, 825-836. 150. S. I. Dikalov, I. A. Kirilyuk, M. Voinov and I. A. Grigor'ev, Free Radical Research, 2010, 45, 417-430. 151. H. A. O. Hill and P. J. Thornalley, Inorganica Chimica Acta, 1982, 67, L35-L36. 152. P. R. Marriott, M. J. Perkins and D. Griller, Canadian Journal of Chemistry, 1980, 58, 803-807. 153. G. Gosset, J.-L. Clément, M. Culcasi, A. Rockenbauer and S. Pietri, Bioorganic & Medicinal Chemistry, 2011, 19, 2218-2230. 154. O. Zerbe, Jurt, S., in Applied NMR Spectroscopy for Chemists and Life Scientists, WILEY-VCH Verlag GmbH & Co. KGaA, 2014, ch. 16, pp. 335-349. 155. P. J. Crickmore, The Canadian Journal of Chemical Engineering, 1989, 67, 392-396. 156. D. G. Truhlar and A. Kohen, Proceedings of the National Academy of Sciences of the United States of America, 2001, 98, 848-851. 157. A. Liese, L. Hilterhaus, M. Howaldt and H. Chmiel, in Bioprozesstechnik, ed. H. Chmiel, Spektrum Akademischer Verlag, Heidelberg, 2011, DOI: 10.1007/978-3- 8274-2477-8_3, pp. 67-97. 158. A. J. Pardey, M. Fernández, J. Alvarez, M. C. Ortega, M. Canestrari, C. Longo, P. Aguirre, S. A. Moya, E. Lujano and P. J. Baricelli, Boletín de la Sociedad Chilena de Química, 2000, 45, 347-354. 159. J. Barrett, in Inorganic Chemistry in Aqueous Solution, ed. J. Barrett, The Royal Society of Chemistry, 2003, vol. 21, pp. 1-12. 160. M. Chaplin, Water structure and science- Anomalous properties of water, http://www1.lsbu.ac.uk/water/water_anomalies.html, (accessed 2015/09/07, 2015). 161. U. Technologies, Hydrogen Peroxide - Boiling points and freezing points, http://www.h2o2.com/technical-library/physical-chemical-properties/physical- properties/default.aspx?pid=22&name=Boiling-Points-and-Freezing-Points, (accessed 01/08, 2016). 162. B. W. Pfennig, in Principles of Inorganic Chemistry, John Wiley & Sons, Inc., 2015, ch. 14.2, p. 461. 163. P. R. Smirnov and V. N. Trostin, Russian Journal of General Chemistry, 2013, 83, 15-25. 164. P. R. Smirnov and T. Yamaguchi, Russian Journal of Physical Chemistry, 2006, 80, 84-89. 226 Literature 165. T. Sakuma, S. Hoshino and Y. Fujii, Solid State Communications, 1978, 25, 145-147. 166. L. Abezgauz, K. Kuperkar, P. A. Hassan, O. Ramon, P. Bahadur and D. Danino, Journal of Colloid and Interface Science, 2010, 342, 83-92. 167. E. Leontidis, Current Opinion in Colloid & Interface Science, 2002, 7, 81-91. 168. B. Szalontai, G. Nagy, S. Krumova, E. Fodor, T. Páli, S. G. Taneva, G. Garab, J. Peters and A. Dér, Biochimica et Biophysica Acta (BBA) - General Subjects, 2013, 1830, 4564-4572. 169. A. Kumar and P. Venkatesu, International Journal of Biological Macromolecules, 2014, 63, 244-253. 170. Y. Marcus, Chemical Reviews, 2009, 109, 1346-1370. 171. M. A. Metrick Ii and G. MacDonald, Colloids and Surfaces A: Physicochemical and Engineering Aspects, 2015, 469, 242-251. 172. N. Kitadai, T. Sawai, R. Tonoue, S. Nakashima, M. Katsura and K. Fukushi, J Solution Chem, 2014, 43, 1055-1077. 173. F. E. Mabbs and D. Collison, Mol. Phys. Rep., 1999, 26, 39-59. 174. T. Spalek, P. Pietrzyk and Z. Sojka, J. Chem. Inf. Model., 2005, 45, 18-29. 227 Publications 12 Publications & conference communications Posters (presenting author is underlined) [1] Braun, N., Rinaldi, R. (2013): Direct coal liquefaction under mild conditions. 46. Jahrestreffen der Deutschen Katalytiker, Weimar, Deutschland, März 2013 [2] Braun, N., Rinaldi, R. (2013): An alternative method for direct coal liquefaction under low‐severity conditions. EuropaCat, XIth European Congress on Catalysis, Lyon, Frankreich, September 2013 [3] Braun, N., Rinaldi, R. (2013): Direct Liquefaction of Low-rank Coals under Mild Conditions. DGMK-Konferenz “Shale Gas, Heavy Oils and Coal”, Dresden, Deutschland, Oktober 2013 Conference paper [1] Braun N, Rinaldi R, 2013, Direct liquefaction of low-rank coals under mild conditions, Pages: 195-197, ISSN: 1433-9013 Patent [1] Rinaldi, R., Braun, N. (2013): Verfahren zur direkten Kohleverflüssigung, DE 10 2013 107 865 A1 (Anmeldetag: 23.07.2013, Offenlegungstag: 30.01.2014) 228