The important role that hydrogen donor species play in quenching the thermally generated radical intermediates during thermal cleavage of chemical bonds in coal related structures has been studied (Meyer et al 1990a, b; Oviawe et al., 1993a,b).
Industrial processes for continous liquefaction of coal under hydrogen at high pressure and at elevated temperature (400-500°C) usually use a process-generated solvent (recycle oil) to disperse the coal in slurry. The role of the solvent is crucial to obtain steady-state usable coal-derived liquids over successive passes of the recycle oil. Solvent refining is one of the most attractive means for converting coal into liquid fuels. Various aspects of the liquefaction process have been studied extensively (Oviawe, 1989; Le-Roux et al., 1982; Ruixia et al., 2004a,b; Hong et al., 2003). The studies have varied widely, ranging from rate determination utilizing several types of coal related compounds such as benzyphenylether (BPE) and of donor solvents to attempt to identify the composition of specific fractions of the products.
In a previous study, it has been demonstrated that -CH2- and -CH2-CH2-
bridges connected aryl groups in coal (Benjamin et al., 1978). In a related
study, we had earlier explored the behaviour of model acceptors to rank solvents
for their effectiveness in stabilizing the radicals generated from the pyrolysis
of benzylphenlether (Oviawe et al., 1993b).
We had earlier (Oviawe et al., 1989) used the benzyl and phenyloxy radicals
generated thermally from benzylphenylether (BPE) to study hydrogen transfer
in model hydrogen donor solvents such as tetralin (T), indane (I), 9,10-dihydrophenanthrene
(DHP) and 1,2,3,4-tetrahydroquinoline (THQ).
In this study, a related aspect of hydrogen transfer reactions is explored
to determine if model acceptor experiments effectively mimic the behavior of
coal in liquefaction and if so, whether they can be utilized to rank solvents
for their coal liquefaction effectiveness. The direct measurement of coal conversion
for model donors in this work permits comparison with model acceptor results.
This study compares the results obtained using BPE with those obtained using
coal to define the most efficient solvents for use in industrial processes.
Materials and Methods
Materials and Solutions
Hydroaromatic substrates (1,2,3,4-tetrahydroqunoline, tetralin, indane,
9,10-dilydrophenanthrene (Aldrich 98% purity grade) and benzylphenylether (Fair
Field) were used without further purification. Freyming coal was generously
supplied to us by CERCHAR. Solutions were obtained as 2:1 molar mixtures of
BPE with each solvent (10 and 5 millimoles, respectively) and were introduce
into the microreactor.
Competition Between the Different Solvents in the Presence of Coal
Reaction of Freyming Coal with Hydroaromatic Compounds
A scheme of the micro reactor has been presented previously (Oviawe, 1989).
A glass tube (volume ~10 cm3) was fitted into the body of the reactor
to avoid coking of radical products on the steel walls and to enable quantitative
analysis of the products. The microautoclaves were filled with the reactants
and allowed to stand in a fluidized sand bath at 400°C for 30 min in a nitrogen
atmosphere while agitating at 700 cpm. The reactor and content were then cooled
rapidly and opened after reducing the residual pressure (~ 8 MPa) to one atmosphere.
Weight losses in the glass tube were less than 1%. The solution was divided
into aliquots for 1H nmr analysis. The microreactor allow both rapid
heating and cooling of the reaction mixture. For each reaction with coal as
substrate 0.5 g of Freyming coal ground to 200-270 mesh (53-74 μm) and
dried in vacuo overnight were introduced into the micro reactor and 5 millimoles
of hydroaromatic solvent were used. For each experiments with coal, the solvent
was composed of the hydroaromatic substrate of interest and 1-methyl naphthalene
as diluent. The % conversion of coal as a function of time is given in Table
3 (Fig. 1).
1H nmr Spectroscopy
1H nmr was carried out at 400 MHZ. A known quantity, Ws, of the
above solution was diluted with deuterated dimethylsulphoxide (99%, CEA) for
nmr analysis. The total number of moles, ni, of each compound, I
, in solution was calculated from the integrals Ii and Ir
of 1H signals of compound I and of reference r (hexamethylcyclotrisiloxane
Aldrich 98%) of molecular weight Mr added to the solution in known quantity
Where Hi and Hr are the number of equivalent protons in each signal. The expected accuracy is 1-2%.
|| Percentage conversion of coal as a function of time
Results and Discussion
A series of four hydroaromatic solvents were used to rank their efficacy for the dissolution of freyming coal. The ranking of the model donors according to their reactivity for coal conversion is given in Table 1. The lowest donor is indane while the highest apparent donor is THQ.
In comparing the results presented in Table 1, especially
taking note of the recycle index of each solvent it is clear that the best solvent
for the liquefaction of coal at 400°C is DPH (RI = 0.72).
It has been widely recognized that coal liquefaction proceeds by means of pyrolytic degradation of coal molecules and the stabilization of the degraded radicals by solvent to enhance liquefaction by the solvent (Kulik et al., 1980; Kuhlamann et al., 1985). The solvent can also stabilize the radicals through dispersion by dissolution and solvation and hydrogen transfer (Mochida et al., 1979; Curtis et al., 1984). Thus, an effective solvent for the liquefaction is required to have high dissolution and hydrogen donating abilities (Mochida et al., 1979). From our studies, THQ proved to be an excellent solvent. THQ and its dehydrogenated from quinoline, have high polarity which allow penetration of the coal structure and aid in dispersion of the dissolved coal. Its N-H bond is weaker than C-H bond in tetralin and in the other two solvents, thus its hydrogen donating activity is superior to those DHP, T and I .
Although the % conversion of coal using THQ is very high but the recycle index is also very low (RI = 0.38). We had earlier showed that THQ starts to decompose at a temperature of 350°C leading to the formation of degradation products such as orthotoludine (Oviawe et al., 1989).
The thermal cleavage of activated bonds in BPE results in the formation of
phenoxyl and benzyl radicals.
||Reactant and products issued from the liquefaction of coal
in the presence of hydroaromatic solvents at 400°C (30 min)
||Reactant and products from the thermal decomposition of BPE
and hydroarmatic solvent (THQ)
These radicals are then stabilized by the hydrogen donated by the hydroaromatic
solvents to form mainly phenol and toluene as the principal products. The reactants
and products from the thermal decomposition of BPE in the presence of hydroaromatic
solvents are given in Fig. 2. The 1H nmr spectra
of the reactants and products are given in Fig. 3 and 4.
THQ is dehydrogenated by the radicals to form quinoline.
|| Decomposition of BPE at 300° (without catalyst using
tetralin (NMR Spectra)
|| Decomposition of BPE at 300°C (in the presence of a catalyst
, SaCl2) using tetralin
Similarly, Tetralin and 9, 10-dihydrophenanthrene are dehydrogenated to naphthalene
and phenanthrene respectively. Comparing Fig. 3 and 4,
one finds that the presence of a catalyst, SnCl2, helps to increase
the liquefaction of BPE, as no trace of BPE was found at 300°C whereas there
was still undecomposed BPE at the same temperature in the absence of a catalyst.
In the study using THQ to stabilize the radicals generated by the thermal decomposition
of BPE, the radicals abstract the hydrogen atom attached to the nitrogen atom
of THQ. Some of the benzlyl radicals then attack the nitrogen atom radical,
resulting in the formation of N-benzyltetrahydroquinoline especially at the
beginning of the reaction.
The radicals form at the beginning of the reaction using THQ are thus stabilized
either by abstraction of hydrogen. This process is termed abstraction (AB),
or are stabilized by addition to the double bonds in the aromatic ring of the
solvent. This process is termed addition (AD).
These side reactions delay the dehydrogenation process of 1,2,3,4-tetrahydroquinoline to quinoline thereby resulting in poor Recycle Index (RI).
As expected, the % conversion of coal using tetralin is better than that of indane (54.80 and 36.62%, respectively). In addition, no trace of indene was detected (RI = 0). An important factor in coal liquefaction is the possibility of reusing the same solvent in successive dehydrogenation-rehydrogenation cycle (Whitehurst et al., 1980). This means that in each dehydrogenation step studied, the hydroaromatic solvent SH2 is quantitatively converted to its aromatic analog S. This ideal situation is far from being reached for all the solvents considered. Thus, a performance index, the Recycle Index (RI) is defined as the ratio.
Further more, the % conversion using a non-donor solvent (1- methylnaphthalene) was the least (25.56%). If one compares these results with reference to the hydrogen donating ability of each of the solvents, it becomes more evident that DHP is the most efficient of the solvents used. The recycle index and the % conversion found in this study are practically comparable with those found in our earlier work using BPE in place of coal (Oviawe et al., 1993b; Oviawe, 1989).
In the experiments carried out at 400°C using DHP and tetraline, DHP donates its hydrogen faster than tetralin, Δ[DHP]/Δ[T] = 2.61 against 1.33 in the presence of BPE (Table 2). The experiment with tetraline and indan at 400°C show that tetralin donates its transferable hydrogen faster than indane, Δ[T]/Δ[ I ] = 1.98 against 1.13 in the presence of BPE. THQ was obviously excluded from this second set of experiments because of its instability above 350°C.
Comparing the ratio, Δ[DPH]/Δ[T] and Δ[T]/Δ[I] in the presence of coal and in the presence of BPE and taking critical note of the recycle index in the two cases, it is clear that the results obtained using coal is in direct agreement with those obtained using BPE. The slight difference observed may be explained by the fact that hydrogen donor content of a solvent alone may not be sufficient to completely define its effectiveness as a coal liquefaction solvent (Curtis et al., 1984).
The % conversion of coal is determined by the solubility of its product with
hydroaromatic solvent in tetrahydrofurane (THF) according to:
||Competition between the hydroaromatic solvents in the presence
of coal: concentrations of reactants and products at 400O C (15
|[Dihydrophenanthrene] = 5x10-3 moles, [Tetralin]
= 5x10-3 moles, [Indane] = 5x10 -3 moles, Weight of
coal = 0.5 g
|| Percentage conversion of coal, in the presence of tetrsalin,
as a function of time at 400°C
The weight of the organic matter is calculated by subtracting the weight of ash from coal knowing that the percentage of ash in freyming coal is 5.5% and that majority of the ash is present in the residue.
For each experiment, the solvent was composed of the hydroaromatic substrate of interest and 1-methylnaphthalene as diluent. Coal conversion was measured by ash balance using THF as the wash solvent. The error obtained from duplicate analysis was generally between 1-4%.
This study was supported by a grant from CROUS, France.