Study of the Kinetics and Mechanisms of

Thermal Decomposition of Ellajjun Oil Shale

Dr. Omar S. Al-Ayed1 1
Mu’tah University
Department of Chemical Engineering
Karak – Mutah - 61710
Jordan

e-mail: omar.alayed@bau.edu.jo

Abstract
Oil Shale kinetic parameters have been determined for Ellajjun oil shale samples. Estimation of
activation energy and frequency factor was performed using different standard procedures. The
employed standard methods are used to estimate the kinetic parameters using TGA/DGA data.
Heating rates; 1, 3, 5, 10, 30, and 50 oCmin-1 are used with a 100 mlmin-1 nitrogen flow rate.
The kinetic parameters determined using Friedman, Kissinger-Akahira-Sunose (KAS) Method,
Flynn-Wall-Ozawa (FWO) and other methods. The activation energy and pre-exponential factor
estimated using Friedman, KAS, FWO did not produced the required matching results while
Coats & Redfern procedure was the best to result in good estimates of the kinetic parameters
that produced predicted curves in good match with experimental data.
Keywords: Pyrolysis, Oil Shale, TGA, Kinetics, Parameters.

Introduction
World reserves of conventional fossil fuels such as petroleum, natural gas, oil shale and coal are
finite. The supply of energy has become a worldwide problem in recent years. Many countries
have attempted to diversify their potential energy sources in order to alleviate the impact of oil
price increases and their effect on the economy. Oil shale represents an enormous potential as a
substitute but it needs technology development to improve the existing industry for
dissemination.
Thermogravimetric analysis (TGA) of oil shale samples has been extensively used as a means
of determining pyrolysis characteristics and also to determine kinetic parameters [1 -12].
Thermogravimetric analysis (TGA) is an analytical device used for accurately measuring weight
change of a material subjected to a temperature history. The weight loss information can be
used to construct kinetic models of the decomposition process.
Non-isothermal thermogravimetry (TG) has been applied in the study of the kinetics of thermal
decomposition of Green River oil shale kerogen [10]. It was shown that the decomposition

1
On leave from Al-Balqa Applied University

1
process involves two consecutive first order reactions. Kerogen decomposes to intermediate
bitumen and the pyrolytic bitumen further decomposes into oil and gas.
Torrente and Galan [4] studied the kinetics of thermal decomposition of oil shale from
Puertollano (Spain) by isothermal and non-isothermal thermogravimetry analyses. They
reported that decomposition of the shale could be described by overall first-order kinetics. Kök
[3] studied the thermal behavior of Seyitomer oil shale by thermogravimetry (TG/DTG) and
differential scanning calorimetry (DSC). It was reported that the decomposition took place in
more than one reaction region where the apparent activation energies of the first region are
higher. Higher heating rates resulted in higher reaction temperatures. It was shown that the
decomposition process of the oil shale could be described by a pseudo first order reaction. Al-
Ayed [13] investigated the kinetics of Ellajjun oil shale using a variable reaction order in which
the calculated order varied between 1.85 and 2. Tiwari, P. and Deo, M. [8] indicated in their
recent work that product evolution temperature is heating rate dependent, which negate iso-
conversional procedure and augment the assumption of activation energy variation with heating
rate.
Karabakan and Yurum [9] investigated the effect of mineral matrix of oil shales. They found out
that pyrolysis reactions were catalyzed by alkaline earth metal and inhibited by silicates. The
inhibition effect of the silicates seemed to be greater than the catalytic effect of the carbonates
in the pyrolysis reactions of the original Göynük and Green River oil shales. Jaber and Probert
[5] studied Jordanian oil shale samples using integral method in the analysis of TG data in order
to determine the pyrolysis kinetics. It was observed that the magnitude of the total mass loss
was mainly dependent on the final temperature, as well as, to a lesser extent, on the heating rate
employed. Williams and Ahmad [6] studied oil shale samples in a thermogravimetric analyzer
in relation to heating rate and temperature using non-isothermal and isothermal analysis
respectively. The main region of mass loss corresponding to hydrocarbon oil and gas release
was between 473–893 K.
The present research is an experimental study on Ellajjun oil shale samples using a TGA
apparatus under non-isothermal conditions. The objective was to investigate the effects of
heating rate and the nature of oil shale on the kinetic parameters. The obtained data are tested on
different available kinetic models to determine activation energy and frequency factors of the
decomposition kinetic equation. In order to achieve these objectives, kinetic models are
proposed which allow a good correlation of the results obtained.

Theoretical approach
The application of dynamic TG methods holds great promise as a tool for unraveling the
mechanisms of physical and chemical processes that occur during solids decomposition. In this
investigation, known developed methods have been used to analyze the non-isothermal kinetics
of Ellajjun oil shale. The pyrolysis process may be represented by the following reaction
scheme:

𝐎𝐢𝐥 𝐬𝐡𝐚𝐥𝐞 (solid) → 𝐒𝐡𝐚𝐥𝐞 𝐨𝐢𝐥 (liquid) + 𝐇𝐲𝐝𝐫𝐨𝐜𝐫𝐚𝐛𝐨𝐧𝐬 (gas) (1)

The rate of transformation of material during pyrolysis is defined in-terms of conversion 𝑥; and
is given as:

2
𝑑𝑥
= 𝑘(𝑇)𝑓(𝑥) (2)
𝑑𝑡
𝑤0− 𝑤𝑡
Where; 𝑥 = � � (3)
𝑤0

where 𝑤0 is total weight loss of sample recorded at 550oC for each TG experiment at constant
heating rate. The heating rate β is defined as:
𝑑𝑇
𝛽= (4)
𝑑𝑡

Equation (2) is expressed interms of heating rate as:

𝑑𝑥 𝑑𝑥
=𝛽 = 𝑘(𝑇)𝑓(𝑥) (5)
𝑑𝑡 𝑑𝑇

Where f (x) and k(T ) are functions of conversion and temperature, respectively. The function k
(T), is the temperature dependency of the rate of mass loss, is often modeled successfully by the
Arrhenius standard equation:

𝑘(𝑇) = 𝑘𝑜 𝐸𝑥𝑝(− 𝐸 ⁄𝑅𝑇) (6)

where E is the apparent activation energy, 𝑘𝑜 the pre-exponential factor and R the universal gas
constant.

Kissinger-Akahira-Sunose Method (KAS method)

Mathematical Development

This method (14, 15) is based on equation (2) in which upon re-writing to obtain the following:
𝑘𝑜
𝑑𝑥⁄𝑓(𝑥) = 𝑒𝑥𝑝(−𝐸 ⁄𝑅𝑇)𝑑𝑇 (7)
𝛽
If equation (7) is integrated with the initial conditions of 𝑥 = 0 at 𝑇 = 𝑇0 to obtain the
following expression:

𝑥 𝑑𝑥 𝑘0 𝑇 −𝐸 𝑘𝑜 𝐸 𝐸
𝑔(𝑥) = ∫𝑜 = ∫𝑇 𝑒𝑥𝑝 � 𝑅𝑇 � 𝑑𝑇 = 𝑝� � (8)
𝑓(𝑥) 𝛽 𝑜 𝛽𝑅 𝑅𝑇

Most of technique assumes that the pre-exponential factor 𝑘𝑜 , the conversion function 𝑓(𝑥)
and activation energy 𝐸 are independent of temperature while frequency factor and activation
energy are independent of conversion, 𝑥. The KAS method is based on the Coats and Redfern
approximation (16)

𝑙𝑛(𝛽⁄𝑇 2 ) = 𝑙𝑛(𝑘𝑜 𝑅⁄𝐸𝑔(𝑥)) − 𝐸 ⁄𝑅𝑇 (9)

3
Equation (9) when plotted will lead to activation energy and frequency factor estimation.
Plotting LHS on y-axis for constant selected conversion values (iso-conversional points) against
inverse of pyrolysis temperature on x-axis will culminate in kinetic parameters determination.
Equation (9) is plotted for kinetic parameters estimation and the obtained graphs are depicted in
figure 1 below. In figure 1, plotting 𝑙𝑛(𝛽⁄𝑇 2 ) on y-axis for selected conversion points indicated
as horizontal lines and different heating rates represented by inclined vertical lines. The
activation energy values for each selected conversion point are indicated by the slope of the
trend-line obtained. On the other hand, the frequency factor is calculated from the intercept
value after substituting the appropriate quantities in the term, (𝑘𝑜 𝑅⁄𝐸𝑔(𝑥)) . The estimated
values of activation energy and frequency factor are given in table I below:

1/T, K-1
0.001 0.0011 0.0012 0.0013 0.0014 0.0015 0.0016 0.0017
-8
ln(β/T2) 0.25= -24235/T + 27.524
R² = 0.9745
-9
ln(β/T2) 0.3= -23768/T + 26.024
R² = 0.984
-10 ln(β/T2)0.4 = -22953/T + 23.787
R² = 0.9864
ln(β/T2)

-11 ln(β/T2)0.5 = -22733/T + 22.759

R² = 0.9853
ln(β/T2)0.6 = -23035/T + 22.612
-12 R² = 0.985
ln(β/T2) 0.7= -23363/T + 22.495
R² = 0.9839
-13 0.3 0.25
ln(β/T2) 0.8= -24177/T + 22.983
0.9 0.80.70.60.5 0.4
R² = 0.9823
ln(β/T2) 0.9= -25138/T + 23.196
-14
R² = 0.9476

Figure 1. Estimated activation energy values with different iso-conversional points using
KAS method.

The estimated values of activation energy and frequency factor values as tabulated in Table
I are plotted as indicated in figure 2 The average values of activation energies at different
iso-conversional points are calculated to be 197 kJ.mol-1 , the estimated values are plotted
and compared with those calculated using Friedman method. From figure 2 it is quite clear
that activation energy passes through a minimum value at 50%. Activation energy obtained
through KAS method is varied within ± 8 kJ.mol-1 of the average value, while, the
fluctuations in activation energy is much higher in those values obtained according to
Friedman procedure.

4
Table I. Activation energy and frequency factor values obtained from figure 1

Activation Energy, Frequency

Conversions Intercepts R-value
Symbol kJ/mol factor
E 0.25 201.5 0.25 7.3E+15 36.521 0.9745
E 0.3 197.6 0.3 2.0E+15 0.9841
E 0.4 190.8 0.4 3.3E+14 33.423 0.9864
E 0.5 189.0 0.5 1.7E+14 32.791 0.9853
E 0.6 191.5 0.6 2.3E+14 33.062 0.985
E 0.7 194.2 0.7 3.2E+14 33.401 0.9839
E 0.8 201.0 0.8 9.3E+14 34.462 0.9823
E 0.9 209.0 0.9 2.7E+15 35.525 0.9476

The activation energies calculated values are for the hydrocarbon evolution region. This fact is
clear from the selected iso-conversional points at the beginning and at the end of each run.
The initial weight loss of sample is attributed to the loss of moisture and structural water
associated with clay. As a result, the initial conversion point selected for the determination of
the kinetic parameters is 25% while the final point is taken at 90% which corresponds to
550 oC final pyrolysis temperature. The conversion is tailing with temperature at values higher
than 90%.

230.0
220.0
Activation Energy, kJ/mol

210.0
200.0
190.0
Friedman
180.0 KAS
170.0
160.0
150.0
0 0.2 0.4 0.6 0.8 1
Conversion

Figure 2 Estimated activation energy values with iso-conversion points using KAS method.

5
The activation energy (E) and frequency factor (k o ) estimated through KAS procedure are
related to each other according to the following equations:

𝑙𝑛(𝑘𝑜 ) = 0.292𝐸 − 22.44 , 0.25 ≤ 𝑥 ≤ 0.5, R2 = 0.998 (10, a)

𝑙𝑛(𝑘𝑜 ) = 0.142𝐸 − 5.714 , 0.5 ≤ 𝑥 ≤ 0.9, R2 = 0.998 (10, b)

Equations (10) are obtained with R2 greater than 0.99. The two equations are probably due to
the nature of the curve concavity where 50% conversion point is fitting both lines extremely
well

Friedman method (17)

Mathematical Manipulations:
This method is a differential iso-conversional method and is directly related to equation (2).
Friedman procedure is based on selected constant conversion points in the whole range of
experiment irrespective of heating rate. In the present work, conversion points 0.25, 0.3, 0.4,
0.5, 0.6, 0.7, 0.8 and 0.9 have chosen.
In this method the conversion function 𝑓(𝑥) has been assumed to be of the standard form:

𝑓(𝑥) = (1 − 𝑥)2 (11)

Equation (2) and introducing heating rate into the equation in its usual way of introduction, and
if the logarithm is taken, the following form of equation is obtained:

𝑑𝑥 𝐸
ln(𝛽 )=− + ln(𝑘𝑜 𝑓(𝑥)) (12)
𝑑𝑇 𝑅𝑇

Activation energies are estimated from equation (12), when plotting LHS at a selected constant
conversion points that have been inferred from experimental data at different heating rates,
(iso-conversional) against inverse of pyrolysis temperatures measured in Kelvin.
Figure 3 shows the determination of activation energies from the slopes of different conversion
points. On the other hand, the pre-exponential factor is calculated from the trend-line equation
of the plotted data. The obtained trend-line linear equation for the whole data is of the form:
𝑦 = 𝑚𝑥 + 𝑐 (13)

Where: 𝒚 represents the LHS of equation (12)

𝒎 slope that gives activation energy.
𝒙 inverse of pyrolysis temperature, K-1.

6
 𝒄 intercept; the quantity, ln (𝑘𝑜 𝑓(𝑥 )) , used to determined frequency factor at a
given conversion function form and at selected iso-conversional experimental points.
Figure 3 is plotted to estimate activation energies at the studied heating rates and selected
conversion points. It is clear from the graph and subsequently the calculated activation energy
values as tabulated in Table I, that activation energies varies between 177 and 222 kJmol-1.

1/T , K -1
0.0012 0.0013 0.0014 0.0015 0.0016 0.0017
0
ln(β(dX/dT))0.25 = -26108/T + 37.81
R² = 0.985
-1 ln(β(dX/dT)) 0.3= -24288/T + 34.48
R² = 0.988
ln(β(dX/dT))0.4 = -21268/T + 29.19
-2
R² = 0.976

ln(β(dX/dT))0.5 = -23589/T + 32.30

ln( h(dx/dT))

-3 R² = 0.984

-4
0.7 0.6 0.5 0.4
-5 ln(β(dX/dT))0.7= -23774/T + 31.38 0.3
R² = 0.979
0.8 0.25
ln(β(dX/dT)) 0.8= -26727/T + 34.71
-6 R² = 0.985
0.9
ln(β(dX/dT))0.9 = -25361/T + 30.71
R² = 0.914 ln(β(dX/dT))0.6= -23544/T + 31.696
-7 R² = 0.9843

Figure 3 Friedman Method for activation energy and frequency factor estimation.

Meanwhile, the calculated values of frequency factor were 1.3*1013 and 2.9*1016 min-1
correspondingly. Since activation energy is reflect of the mechanism of chemical reactions,
there is no direct relationship between conversion and the values of activation energy obtained.
The pre-exponential factor and activation energy are related to each other in the following form:

𝑙𝑛(𝑘𝑜 ) = 0.178𝐸 − 1.267, R2 = 0.951

P P (14)

7
Table II. Activation energy and frequency factor values obtained from figure 3

Activation Energy, Frequency factor

Conversions Intercepts R-value
Symbol, kJ/mol
E 0 . 25 217.1 0.25 38.4 4.7E+16 0.985
E 0.3 201.9 0.3 35.2 1.9E+15 0.988
E 0.4 176.8 0.4 30.2 1.3E+13 0.976
E 0.5 196.1 0.5 33.7 4.3E+14 0.984
E 0.6 195.7 0.6 33.5 3.6E+14 0.984
E 0.7 197.7 0.7 33.8 4.7E+14 0.979
E 0.8 222.2 0.8 37.9 2.97E+16 0.985
E 0.9 210.8 0.9 35.3 2.2E+15 0.914

Figure 3 Shows that no definite relationship between activation energy calculated

according to Friedman procedure and conversion. It is clear from the figure that
activation energy decreases from: 217 to 177 kJ/mol from initial values of used
conversions to about 40%. Low conversion is characterize by Pyrolysis reactions with
low activation energies, the pyrolysis of oil shale resulted mainly from the rupture of
weak chemical bond, probably the rupture of weak cross linked bonds, such as, C-O, C-S
bonds (19,20) etc. Also, rupture of branched functional groups in kerogen long molecular
structure. These weak bonds have low rupture energy which commensurate with
pyrolysis gaseous products such as H 2 O, CO 2 , H 2 S, H 2 and light hydrocarbons.
Campbell (21) studied the rate of evolution of CH 4 , H 2 , CO, CO 2 , and C 2 , C 3
hydrocarbons during pyrolysis of Colorado Oil Shale at linear heating rates varying from
0.5 to 4.0oCmin-1. More of hydrogen release was reported at lower heating rates.
Methane formation and evolution initiated at temperature slightly lower than 350 oC and
reaching a maximum value at 445 oC. Rate of methane release increased with a decrease
in heating rate. Ethane and ethene (C 2 ) production increased to a maximum value at
450 oC and stopped at slightly higher than 550 oC. C 3 (propane and propene) production
was found to occur at 450 oC and its formation was stopped at 525 oC. Oil release
profile was corresponding closely to that observed for C 1 and C 2 profiles.

8
Flynn-Wall-Ozawa Method (FWO): (22, 23)
Mathematical Development

All the four methods examined in this research work assumed equation (2) is the main equation
for oil shale pyrolysis. The Flynn-Wall-Ozawa method is derived from the integral iso-
conversional method. Using Doyle’s approximation for the integral that allows the quantity

𝑙𝑛�𝑝(𝐸 ⁄𝑅𝑇)� ≅ −5.331 − 1.052 𝐸 ⁄𝑅𝑇 (15)

𝑘𝑜
𝑑𝑥⁄𝑓(𝑥) = 𝑒𝑥𝑝(−𝐸 ⁄𝑅𝑇)𝑑𝑇 (16)
ℎ

If equation (16) is integrated with the initial conditions of 𝑥 = 0 at 𝑇 = 𝑇0 to obtain the

following expression

𝑥 𝑑𝑥 𝑘0 𝑇 −𝐸 𝑘𝑜 𝐸 𝐸
𝑔(𝑥) = ∫𝑜 = ∫𝑇 𝑒𝑥𝑝 � 𝑅𝑇 � 𝑑𝑇 = 𝑝� � (17)
𝑓(𝑥) 𝛽 𝑜 𝛽𝑅 𝑅𝑇

Combining equations (15) and equation (17) to obtain the final form used by FWO procedure as
follows:

𝑙𝑛(𝛽) = 𝑙𝑛(𝑘𝑜 𝐸 ⁄𝑅𝑔(𝑥)) − 5.331 − 1.052 𝐸 ⁄𝑅𝑇 (18)

Hence, for iso-conversional points, the plot of LHS of equation (18) verses inverse of pyrolysis
temperature obtained from experimental data recorded at different heating rates, would be a
straight line whose slope can be used to evaluate the activation energy and the corresponding
frequency factor. Figure 4 exhibit such a plot and the calculated kinetic parameters, namely,
activation energy and pre-exponential factor have been tabulated in Table III.
Inspecting the values of activation energy as reported in table III, similar trend of those values
obtained by KAS is observed. The activation energy passes through a minimum value at 0.5
conversions. On the other hand, the calculated frequency factors are much similar to those
obtained by KAS procedure. The following equations are obtained for frequency factor and
activation energy for FWO method of analysis.

𝑙𝑛(𝑘𝑜 ) = 0.303𝐸 − 24.54, 0.25 ≤ 𝑥 ≤ 0.5, R2 = 0.997

𝑙𝑛(𝑘𝑜 ) = 0.138𝐸 − 6.637, 0.5 ≤ 𝑥 ≤ 0.9, R2 = 0.998 (19)

Whilst KAS correlation obtained earlier for the kinetic parameters are as follows:

𝑙𝑛(𝑘𝑜 ) = 0.292𝐸 − 22.44 , 0.25 ≤ 𝑥 ≤ 0.5, R2 = 0.997

𝑙𝑛(𝑘𝑜 ) = 0.142𝐸 − 5.714 , 0.5 ≤ 𝑥 ≤ 0.9, R2 =0.998 (20)

It is apparent from values as indicated by equations (19) and (20) that both methods are in
agreement and predict same values for the kinetic parameters.

9
Table III Activation energy and frequency factor values obtained from FWO method.

Activation Energy,
Conversions Intercepts Frequency factor R-value
Symbol, kJ/mol
E 0.25 201.6 0.25 36.7 7.59E+15 0.977
E 0.3 198.0 0.3 35.4 2.3E+15 0.986
E 0.4 191.9 0.4 33.7 4.16E+14 0.988
E 0.5 190.4 0.5 33.1 2.34E+14 0.987
E 0.6 192.9 0.6 33.4 3.09E+14 0.987
E 0.7 195.7 0.7 33.7 4.36E+14 0.986
E 0.8 202.3 0.8 34.7 1.22E+15 0.984
E 0.9 210.3 0.9 35.8 3.51E+15 0.953

230.0
220.0
Activation Energy, kJ/mol

210.0
200.0
190.0 Friedman

180.0 KAS
FWO
170.0
160.0
150.0
0 0.2 0.4 0.6 0.8 1
Conversion

Figure 4 Activation energy as function of conversion calculated by different methods.

10
 Conclusions
In the present research paper various developed methods for estimating kinetic
parameters have been tested and checked their suitability to model Ellajjun Oil Shale
pyrolysis kinetics.

Three models have been tested after brief mathematical manipulations to extract the final
form of equation employed to estimate the activation energy and pre-exponential factor.

In summary, the examined procedures were found that some partial fitting to data was
obtained. In conclusion, it cannot be said that any particular procedure fitted the
experimental data completely.

REFERENCES

1. Al-Ayed, O.S. “Variable Reaction Order for Kinetic Modeling of Oil Shale Pyrolysis”.
Oil Shale, 2011, Vol. 28, No. 2, pp. 296–308.
2. Aboulkas, A. and E. El Harfi. “Study of the Kinetics and Mechanism of Thermal
Decomposition of Moroccan Tarfaya Oil Shale and its Kerogen”. Oil Shale, Vol. 25, No.
4, 2008. pp. 426-443.
3. Kok, M. V. “Thermal Investigation of Seyitomer Oil Shale”. Thermochim. Acta. Vol.
369, 2001. No. 1–2. pp. 149–155.
4. Torrente, M.C., and Galan. “Kinetics of the Thermal Decomposing of Oil Shale from
Puertollano (Spain)”. Fuel, 2001. Vol. 80, No. 3. Pp. 327-334.
5. Jaber, J. O., Probert, S. D.” Non-isothermal Thermogravimetry and Decomposition
6. Kinetics of Two Jordanian Oil Shales under Different Processing Conditions” Fuel
Process. Technol. 2000. Vol. 63, No. 1. pp. 57–70.
7. Williams, P. T., Ahmad, N. “Investigation of Oil-Shale Pyrolysis Processing Conditions
Using Thermogravimetric Analysis”. Appl. Energy. 2000. Vol. 66, No. 2. pp. 113–133.
8. Haixiang, C., L. Naian, and Z. Weitao. “Critical Study on the Identification of Reaction
Mechanism by the Shape of TG/DTG Curves”, Solid State Sciences xxx (2010) 1–6
9. Pankaj Tiwari, Milind Deo, Compositional and Kinetic Analysis of Oil Shale Pyrolysis
Using TGA–MS”. Fuel (2011). doi:10.1016/ j.fuel.2011.09.018
10. Karabakan, A., Yurum, Y. “Effect of the Mineral Matrix in the Reactions of Oil Shales:
1. Pyrolysis Reactions of Turkish Göynük and US Green River Oil Shales” Fuel. 1998.
Vol. 77, No. 12. pp. 1303–1309.
11. Rajeshwar, K. “The kinetics of The Thermal Decomposition of Green River Oil Shale
Kerogen by Non-isothermal Thermogravimetry”. Thermochim. Acta 1981.Vol. 45, No.
3. pp. 253–263.
12. Campbell JH, Kokinas GJ, Gallegos G, Gregg M. “Gas Evolution during Oil Shale
Pyrolysis 1: Non-isothermal Rate Measurements”. Fuel 1980, Vol. 59 pp.718–726.
13. Campbell JH, Gallegos G, Gregg M. “Gas Evolution during Oil Shale Pyrolysis. 2:
Kinetic and Stoichiometric Analysis”. Fuel 1980. Vol. 59. pp. 727–732.

11
14. Al-Ayed, O.S. “Variable Reaction Order for Kinetic Modeling of Oil Shale Pyrolysis”,
Oil Shale J. 2011, Vol. 28(2), P. 296-308.
15. Kissinger, H. E. Reaction kinetics in differential thermal analysis // Anal. Chem. 1957.
Vol. 29, No. 11. P. 1702–1706.
16. Akahira, T., Sunose, T. Trans. Joint Convention of Four Electrical Institutes, Paper No.
246, 1969 Research Report // Chiba Institute of Technology Sci. Technol. 1971. Vol. 16.
P. 22–31.
17. A.W. Coats and J.P. Redfern, “Kinetic Parameter from Thermogravimetric Data",
Nature, 1964. Vol. 201, 68-69.
18. Friedman, H. Kinetics of thermal degradation of char-forming plastics from
thermogravimetry. Application to a phenolic plastic // J. Polym. Sci., Part C. 1964. Vol.
6. P. 183–195.
19. Djurcic M., Murphy, R.C., Vitorovic D., Biemann K.," Organic acids obtained by
alkaline permanganate oxidation of kerogen from the Green River (Colorado) shale",
Geochim Cosmochim Acta, vol 35 (12), 1971, 1201-1207.
20. Young D.K., Yen T.F.," The nature of straight-chain aliphatic structures in green river
kerogen ", Geochim Cosmochim Acta, Volume 41, Issue 10, October 1977, 1411-1417.
21. Campbell J.H., Gallegos, G. and Gregg M. “Gas Evolution during Oil Shale Pyrolysis.
Kinetic and Stoichiometric Analysis”. 1980, Fuel, 59, 727-732.
22. Flynn, J. H., Wall, L. A. A quick, direct method for the determination of activation
energy from thermogravimetric data // Polym. Lett. 1966. Vol. 4. P. 323–328.
23. Ozawa, T. A new method of analyzing thermogravimetric data // Bull. Chem. Soc. Jpn.
1965. Vol. 38, No. 11. P. 1881–1886.

12