Content uploaded by Vladimir G. Kutcherov
Author content
All content in this area was uploaded by Vladimir G. Kutcherov on Sep 10, 2014
Content may be subject to copyright.
The evolution of multicomponent systems at high
pressures: VI. The thermodynamic stability of
the hydrogen–carbon system: The genesis of
hydrocarbons and the origin of petroleum
J. F. Kenney
†‡§
, Vladimir A. Kutcherov
¶
, Nikolai A. Bendeliani
储
, and Vladimir A. Alekseev
储
†Gas Resources Corporation, 11811 North Parkway, Floor 5, Houston, TX 77060; ‡Russian Academy of Sciences, Joint Institute of Earth Physics,
Bolshaya Gruzinskaya 10, 123810 Moscow, Russia; ¶Russian State University of Oil and Gas, Leninski Prospect 65, 117917 Moscow, Russia;
and 储Russian Academy of Sciences, Institute for High Pressure Physics, 142092 Troitsk, Moscow Region, Russia
Communicated by Howard Reiss, University of California, Los Angeles, CA, June 24, 2002 (received for review April 3, 2002)
The spontaneous genesis of hydrocarbons that comprise natural
petroleum have been analyzed by chemical thermodynamic-stability
theory. The constraints imposed on chemical evolution by the second
law of thermodynamics are briefly reviewed, and the effective pro-
hibition of transformation, in the regime of temperatures and pres-
sures characteristic of the near-surface crust of the Earth, of biological
molecules into hydrocarbon molecules heavier than methane is rec-
ognized. For the theoretical analysis of this phenomenon, a general,
first-principles equation of state has been developed by extending
scaled particle theory and by using the technique of the factored
partition function of the simplified perturbed hard-chain theory. The
chemical potentials and the respective thermodynamic Affinity have
been calculated for typical components of the H–C system over a
range of pressures between 1 and 100 kbar (1 kbar ⴝ100 MPa) and
at temperatures consistent with those of the depths of the Earth at
such pressures. The theoretical analyses establish that the normal
alkanes, the homologous hydrocarbon group of lowest chemical
potential, evolve only at pressures greater than ⬇30 kbar, excepting
only the lightest, methane. The pressure of 30 kbar corresponds to
depths of ⬇100 km. For experimental verification of the predictions
of the theoretical analysis, a special high-pressure apparatus has been
designed that permits investigations at pressures to 50 kbar and
temperatures to 1,500°C and also allows rapid cooling while main-
taining high pressures. The high-pressure genesis of petroleum hy-
drocarbons has been demonstrated using only the reagents solid iron
oxide, FeO, and marble, CaCO
3
, 99.9% pure and wet with triple-
distilled water.
Natural petroleum is a hydrogen–carbon (H–C) system, in
distinctly nonequilibrium states, composed of mixtures of
highly reduced hydrocarbon molecules, all of very high chemical
potential and most in the liquid phase. As such, the phenomenon
of the terrestrial existence of natural petroleum in the near-surface
crust of the Earth has presented several challenges, most of which
have remained unresolved until recently. The primary scientific
problem of petroleum has been the existence and genesis of the
individual hydrocarbon molecules themselves: how, and under what
thermodynamic conditions, can such highly reduced molecules of
high chemical potential evolve?
The scientific problem of the genesis of hydrocarbons of natural
petroleum, and consequentially of the origin of natural petroleum
deposits, regrettably has been one too much neglected by compe-
tent physicists and chemists; the subject has been obscured by
diverse, unscientific hypotheses, typically connected with the ro-
coco hypothesis (1) that highly reduced hydrocarbon molecules of
high chemical potentials might somehow evolve from highly oxi-
dized biotic molecules of low chemical potential. The scientific
problem of the spontaneous evolution of the hydrocarbon mole-
cules comprising natural petroleum is one of chemical thermody-
namic-stability theor y. This problem does not involve the properties
of rocks where petroleum might be found or of microorganisms
observed in crude oil.
This paper is organized into five parts. The first section reviews
briefly the formalism of modern thermodynamic-stability theory,
the theoretical framework for the analysis of the genesis of hydro-
carbons and the H–C system, as similarly for any system.
The second section examines, applying the constraints of ther-
modynamics, the notion that hydrocarbons might evolve sponta-
neously from biological molecules. Here are described the spectra
of chemical potentials of hydrocarbon molecules, particularly the
naturally occurring ones present in petroleum. Interpretation of the
significance of the relative differences between the chemical po-
tentials of the hydrocarbon system and those of biological mole-
cules, applying the dictates of thermodynamic-stability theor y,
disposes of any hypothesis of an origin for hydrocarbon molecules
from biological matter, excepting only the lightest, methane.
In the third section is described a first-principles, statistical
mechanical formalism, developed from an extended representation
of scaled particle theory (SPT) appropriate for mixtures of aspheri-
cal hard-body molecules combined with a mean-field representa-
tion of the long-range, attractive component of the intermolecular
potential.
In the fourth section, the thermodynamic Affinity developed
using this formalism establishes that the hydrocarbon molecules
peculiar to natural petroleum are high-pressure polymorphs of
the H–C system, similarly as diamond and lonsdaleite are to
graphite for the elemental carbon system, and evolve only in
thermodynamic regimes of pressures greater than 25–50 kbar (1
kbar ⫽100 MPa).
The fifth section reports the experimental results obtained using
equipment specially designed to test the predictions of the previous
sections. Application of pressures to 50 kbar and temperatures to
1,500°C upon solid (and obviously abiotic) CaCO
3
and FeO wet
with triple-distilled water, all in the absence of any initial hydro-
carbon or biotic molecules, evolves the suite of petroleum fluids:
methane, ethane, propane, butane, pentane, hexane, branched
isomers of those compounds, and the lightest of the n-alkene series.
1. Thermodynamic Stability and the Evolution of
Multicomponent Systems
Central to any analysis of chemical stability is the thermody-
namic Affinity, A({
i
}), which determines the direction of
evolution of a system in accordance with the second law of
thermodynamics as expressed by De Donder’s inequality, dQ⬘⫽
Ad
ⱖ0 (2). The Affinity of an n-component, multiphase system
of pphases involving rchemical reactions is given as
Abbreviations: STP, standard temperature and pressure; SPT, scaled particle theory; SPHCT,
simplified perturbed hard-chain theory.
§To whom reprint requests should be addressed. E-mail: JFK@alum.mit.edu.
10976–10981
兩
PNAS
兩
August 20, 2002
兩
vol. 99
兩
no. 17 www.pnas.org兾cgi兾doi兾10.1073兾pnas.172376899
A⫽
冘
⫽1
r
A
⫽⫺
冘
⫽1
r
冘
␣
⫽1
p
冘
i⫽1
n
i,
i
␣
共p,T,兵n
j
其兲,[1]
in which
i,
and
i,
are the chemical potential and stoichio-
metric coefficients of the ith component in the
th reaction,
respectively;
␣
designates the respective phase.
The second law states that the internal production of entropy
is always positive for every spontaneous transformation. There-
fore, the thermodynamic Affinity (Eq. 1) must always be posi-
tive, and the direction of evolution of any system must always
obey the inequalities:
dS
int
⫽
冦
1
T
冘
A
d
⫽⫺1
T
冘
⫽1
r
冘
␣
⫽1
p
冘
i⫽1
n
i,
i
␣
共p,T,兵n
j
其兲d
冘
k
F
k
dX
k
冧
ⱖ0.
[2]
The inequalities in Eq. 2express the irreversibility of spontaneous
transitions and state that for a spontaneous evolution of a system
from any state, A, to any other state, B, the free enthalpy of state
Bmust be less than that of state A, and at no point between the two
may the free enthalpy be greater than that of state Aor less than
that of state B.
The sum of the products on the second line of inequality in Eq.
2, of the thermodynamic Affinities and the differential of the
variables of extent, d
, is always positive, and the circumstance for
which the change of internal entropy is zero defines equilibrium,
from which there is no spontaneous evolution. This is De Donder’s
inequality.
The sum of products on the second line of inequality in Eq. 2
deserves particular note. In the second line of Eq. 2,F
k
and dX
k
are
general thermodynamic forces and f lows, respectively, and subsume
Newton’s rule, F
ᠬ⫽ma
ᠬ
, as a special case (3, 4). The expression in the
second line of Eq. 2states further that for any circumstance for
which the Affinity does not vanish, there exists a generalized
thermodynamic force that drives the system toward equilibrium.
The constraints of this expression assure that an apple, having
disconnected from its bough, does not fall, say, half way to the
ground and there stop (a phenomenon not prohibited by the first
law) but must continue to fall until the ground. These constraints
force a chemically reactive system to evolve always toward the state
of lowest thermodynamic Affinity.
Thus, the evolution of a chemically reactive, multicomponent
system may be determined at any temperature, pressure, or com-
position whenever the chemical potentials of its components are
known. To ascertain the thermodynamic regime of the sponta-
neous evolution of hydrocarbons, their chemical potentials must
be determined.
No consideration has been given in the foregoing discussion of
chemical thermodynamic stability to the rate of increase of the
variables of extent, d
. Such is the subject of chemical kinetics, not
stability theory. The rate at which a reaction might occur cannot
alter its direction as determined by the second law of thermody-
namics; otherwise the second law would not exist. The evolution of
a system can admit intermediate states, in which one (or more)
intermediate product might possess a chemical potential consider-
ably greater than that of any of the initial reagents. The presence
of a selected catalyst can enhance a fast reaction, and if the system
is removed rapidly from thermodynamic env ironment at which such
reactions proceeding to the final state occur, the intermediate
product(s) can be separated. The petrochemical industry routinely
operates such processes. However, such complex industrial pro-
cesses are not mimicked spontaneously in the natural world.
2. The Thermodynamic Energy Spectrum of the H–C System
and the Effective Prohibition of Low-Pressure Genesis
of Hydrocarbons
The thermodynamic energy spectrum of the chemical potentials
(molar Gibbs energies of formation, ⌬G
f
) of the H–C system at
standard temperature and pressure [(STP, 298.15 K; 1 atm
(1 atm ⫽101.3 kPa)] is available from tables of chemical data
(5). The chemical potentials of naturally occurring members of
the of the H–C system at STP are shown graphically in Fig. 1.
Examination of the energy spectrum of these chemical potentials
of the H–C system establishes at once that, at STP, the chemical
potentials of the entire hydrocarbon system are remarkable for
both their characteristic increase with degree of polymerization
as well as their linear, and almost constant, magnitude of such
increase with carbon number. With increasing polymerization,
the n-alkane molecules manifest increased chemical potential of
very approximately 2.2 kcal per added carbon atom or CH
2
unit.
(There exist also branched isomers, the chemical potentials of
which differ from such of the normal configuration by, typically,
2–4%.) Such increase in chemical potential with increased
degree of polymerization contrasts strongly with the thermody-
namic spectrum of the highly oxidized biotic carbon (‘‘organic’’)
compounds of the hydrogen–carbon–oxygen (H–C–O) system,
which manifest consistently decreasing chemical potentials with
increasing polymerization. This latter property allows the
high degree of polymerization and complex ity of the biotic
compounds.
Examination of the H–C–O system of oxidized carbon com-
pounds establishes that the chemical potentials of almost all biotic
compounds lie far below that of methane, the least energetic of the
reduced hydrocarbon compounds, typically by several hundred
kcal兾mol. Although there exist biotic molecules of unusually high
chemical potential such as

-carotene (C
40
H
56
), vitamin D
(C
38
H
44
O), and some of the pheromone hormones, such com-
pounds are relatively rare by abundance. They are produced by
biological systems only when the producing entity is alive (and at
formidable metabolic cost to the producing entity), and the pro-
duction ceases with the death of the entity. Such compounds are not
Fig. 1. Molar Gibbs energies of formation, ⌬Gf, of the naturally occurring
hydrocarbons at STP (5).
Kenney et al. PNAS
兩
August 20, 2002
兩
vol. 99
兩
no. 17
兩
10977
PHYSICSCHEMISTRY
decomposition products of other biotic compounds and are labile
and themselves decompose rapidly. For these foregoing reasons,
such compounds cannot be considered relevant to the subject of the
origin of natural petroleum.
The properties of the thermodynamic energy spectrum of the
H–C and H–C–O systems, together with the constraints of the
second law (Eq. 2) establish three crucial properties of natural
petroleum:
(i) The H–C system that constitutes natural petroleum is a
metastable one in a very nonequilibrium state. At low pres-
sures, all heavier hydrocarbon molecules are thermodynami-
cally unstable against decomposition into methane and carbon,
as similarly is diamond into graphite.
(ii) Methane does not polymerize into heavy hydrocarbon mole-
cules at low pressures at any temperature. Contrarily, increas-
ing temperature (at low pressures) must increase the rate of
decomposition of heavier hydrocarbons into methane and
carbon.
(iii) Any hydrocarbon compound generated at low pressures,
heavier than methane, would be unstable and driven to the
stable equilibrium state of methane and carbon.
These conclusions have been demonstrated amply by a century
of refinery engineering practice. The third conclusion has been
demonstrated by many unsuccessful laboratory attempts to convert
biotic molecules into hydrocarbons heavier than methane.
There are three generic chemical processes that deserve specific
consideration: the ‘‘charcoal burner’s,’’ ‘‘bean-eater’s,’’ and ‘‘oc-
tane-enhanced bean-eater’s’’ reactions. All describe limited reac-
tions by which a highly oxidized biotic molecule can react to
produce elemental carbon when ‘‘carried’’ by a more thorough
oxidation process. In both the following examples, the simple
carbohydrate, sugar C
6
H
12
O
6
, is used as a typical biotic reagent; the
same reasoning and results hold also for any of the highly oxidized
biotic compounds.
The charcoal burner’s reaction is:
C
6
H
12
O
6
36C ⫹6H
2
O. [3]
The chemical potential of water vapor at STP is ⫺54.636 kcal兾mol.
The thermodynamic Affinity for the charcoal burner’s reaction
(Reaction 3) to produce amorphous carbon, or graphite, is 109.10
kcal. Therefore, the genesis of coal from biological detritus in an
oxygen-poor environment is per mitted by the second law. However,
the thermodynamic Affinity for the charcoal burner’s reaction to
produce diamond is 105.02 kcal, the quantity of which is also
positive, and therefore not immediately prohibited by the second
law as expressed solely by de Donder’s inequality, the first of
equations (Eq. 2). Nonetheless, no charcoal burner ever scrabbles
through his ashes hoping to find diamonds. Such reasonable
behavior demonstrates an effective appreciation of the dictates of
the second law as expressed by Eq. 2. In this case, the generalized
force is the difference in thermodynamic Affinity between the
reactions for graphite and diamond, respectively,
␦
F⫽
␦
A兾T, which,
in the regime of temperatures and pressures of the near-surface
crust of the Earth, assures always the genesis of graphite, never
diamond. Similarly, for reactions involving hydrocarbons heavier
than methane, the same generalized force,
␦
A兾T, always drives the
system toward the state of lowest free enthalpy, i.e., methane plus
free carbon.
The bean-eater’s reaction is:
C
6
H
12
O
6
33CH
4
⫹3CO
2
.[4]
The chemical potentials at STP for the simple carbohydrate
C
6
H
12
O
6
, methane, and carbon dioxide are ⫺218.720, ⫺12.130, and
⫺94.260 kcal兾mol, respectively. The thermodynamic Affinity for
the reaction accordingly is 100.42 kcal兾mol and therefore permitted
by the second law. Indeed, reactions of the type in Reaction 4are
typical of those by which methane is produced in swamps, sewers,
and the bowels of herbivores.
The octane-enhanced bean-eater’s reaction is:
C
6
H
12
O
6
32CH
4
⫹3CO
2
⫹1
8C
8
H
18
⫹7
8H
2
.[5]
Since the chemical potential of n-octane is 4.290 kcal兾mol at STP,
and that of molecular hydrogen is zero, the thermodynamic Affinity
for the octane-enhanced bean-eater’s reaction is A⫽(100.42 ⫺
12.130 ⫺4.290兾8) ⫽87.70 kcal兾mol, still positive and thereby not
prohibited outright by the constraints of De Donder’s inequality.
However, no biochemical investigation has ever observed a mole-
cule of any hydrocarbon heavier than methane resulting from the
decomposition of biological detritus. After a meal of, e.g., Boston
baked beans, one does experience biogenic methane, but not
biogenic octane. No such process produces heavier hydrocarbons,
for such a process would involve effectively a reaction of low-
pressure methane polymerization, similarly as the effective prohi-
bition of the evolution of diamonds by the charcoal burner’s
reaction. In the previous section, we described the industrial
technique by which useful intermediate products can be obtained
by controlling the reaction process. The Fischer–Tropsch process
uses reactions essentially identical to Reaction 5to generate liquid
petroleum fuels from the combustion of coal, wood, or other biotic
matter. However, the highly controlled industrial Fischer–Tropsch
process does not produce, spontaneously and uncontrolled, the
commonly observed large accumulations of natural petroleum.
The foregoing properties of natural petroleum and the effective
prohibition by the second law of thermodynamics of its spont aneous
genesis from highly oxidized biological molecules of low chemical
potentials were clearly understood in the second half of the 19th
century by chemists and thermodynamicists such as Berthelot and
later confirmed by others including Sokolov, Biasson, and Men-
deleev. However, the problem of how and in what regime of
temperature and pressure hydrogen and carbon combine to form
the particular H–C system manifested by natural petroleum re-
mained. The resolution of this problem had to wait a century for the
development of modern atomic and molecular theory, quantum
statistical mechanics, and many-body theor y. This problem now has
been resolved theoretically by determination of the chemical po-
tentials and the thermodynamic Affinity of the H–C system using
modern quantum statistical mechanics and has also now been
demonstrated experimentally with specially designed high-pressure
apparatus.
3. Calculation of the Thermodynamic Affinity Using SPT and
the Formalism of the Simplified Perturbed Hard-Chain
Theory (SPHCT)
To calculate the thermodynamic Affinity of a distribution of
compounds of the H–C system in general regimes of temperature
and pressure, one must use a rigorous mathematical formalism
developed from first-principles statistical mechanical argument.
No approximate, or interpolated, formalism developed for the
low-pressure regime can suffice. A sufficiently rigorous formal-
ism has been developed by extending the SPT equation of state
of Pavlı´cek, Nezbeda, and Boublı´k (6, 7), for mixtures of con-
vex hard-body systems, combined with the formalism of the
SPHCT (8).
Following the procedure enunciated originally by Bogolyubov (9)
and developed further by Feynmann (10) and Yukhnovski (11), a
factored partition function is used that employs a reference system:
Q⫽Q
ref
Q
vdW
. The reference system used is that of the hard-body
fluid as described rigorously by SPT (12–15). The description of the
hard-body fluid by SPT is one of the few exactly solvable problems
in statistical mechanics. This property is especially valuable because
the thermodynamic evolution of a system at high pressures is
10978
兩
www.pnas.org兾cgi兾doi兾10.1073兾pnas.172376899 Kenney et al.
determined almost entirely by the variable components that are
obtained from the reference system.
For mixtures of hard-body particles of different sizes and shapes,
SPT generates the following analytical expression for the contri-
bution to the pressure of the hard-core reference system:
p
ref
⫽k
B
T
冋
1⫹
冉
共1⫺
兲⫹rs
共1⫺
兲
2
⫹qs
2
共1⫺2
兲⫹5rs
2
3
共1⫺
兲
3
冊册
⫽共p
IG
⫹p
hc
兲,[6]
in which the geometric compositional variables, r,s, and
, are
defined by the Steiner–Kihara equations:
r⫽
冘
i
x
i
R
˜
i
,q⫽
冘
i
x
i
R
˜
i
2
,s⫽
冘
i
x
i
S
˜
i
,
⫽
冘
i
x
i
V
˜
i
,
⫽
冘
i
x
i
V
˜
i
⫽
冎
.
[7]
(A thorough discussion of the Steiner–Kihara parameters may be
found in ref. 16.) The following geometric functions are introduced:
␣
⫽rs兾3
, and
␥
⫽具r
2
典兾具r典
2
. The geometric parameter
␣
is the
multicomponent analogue of the Boublı´k parameter of asphericity
for a single-component fluid,
␣
B
⫽R
˜S
˜兾(3V
˜), and may be inter-
preted as the system’s weighted degree of asphericity. The param-
eter
␥
has no analogue in a single-component fluid, for which it is
always equal to unity.
␥
might be interpreted as a parameter of
interference that measures the degree of difference in the mean
component dimensions of radii. When these definitions are used,
the Boublı´k equation (Eq. 6) can be written in a simple form as
p
ref
⫽k
B
T
冋
1⫹c
1
⫹c
2
2
⫹c
3
3
共1⫺
兲
3
册
,[8]
in which c
1
,c
2
, and c
3
are variables of composition that depend on
the combined geometries of the molecular components and their
fractional abundances: c
1
⫽3
␣
⫹1, c
2
⫽3
␣
(
␣␥
⫺1) ⫺2, and
c
3
⫽1⫺
␣
(6
␣␥
⫺5).
Similarly, the contribution of the reference system to the free
enthalpy may be written, as
G
ref
⫽Nk
B
T
冋
冘
i
x
i
ln
冉
共n
i
i
3
兲
V
冊
⫹
冉
I⫹J
⫹K
2
共1⫺
兲
3
冊
⫺c
3
ln共1⫺
兲
册
,
[9]
in which I⫽2c
1
⫺c
3
,J⫽⫺1兾2(3c
1
⫺3c
2
⫺5c
3
), and K⫽1兾6(3c
1
⫺
3c
2
⫺3c
3
). When these identities are used, the contribution of the
reference system to the pressure and the free enthalpy become
simplified functions of the packing-fraction,
, and the geometric
compositional variables,
␣
and
␥
.
The contributions to the pressure and the chemical potentials
from the long-range van der Waals component of the intermolec-
ular potential are described using the formalism of the SPHCT
(17–19). The SPHCT uses the mean-field technique (20) of the
Bethe–Peierls–Prigogine ‘‘lattice-gas’’ model, in which has been
applied the shape-independent scattering formalism (21). As dem-
onstrated previously (22) at elevated pressures, the pressure and
chemical potential are dominated by their respective hard-core
components, and the attractive component is several orders of
magnitude smaller and of little consequence. The representation of
the attractive components of the pressure and chemical potential
used has been that developed for the SPHCT by Sandler (23),
Donohue and Prausnitz (19), and Lee and Chao (24) using the
mixing rules of Kim et al. (25). The Prigogine shape cfactors used
by the SPHCT are related to the Boublı´k geometric parameters
such that c
i
⫽(1 ⫹3R
˜
i
S
˜
i
兾V
˜
i
) and V*
i
⫽V
˜
i
(1 ⫹
␣
i
). The values of V
i
and
␣
i
were taken from van Pelt et al. (26). The chemical potential
of the i-th specie of a multicomponent system is given by
i
⫽
i
⫺
⫹
i
ref
⫹
i
vdW
, in which
i
⫺
represents the reference value of the
chemical potential at STP.
4. The Evolution of the Normal Alkanes, Ethane, Hexane, and
Decane from Methane at High Pressures
At STP, methane possesses the lowest chemical potential and is
the only thermodynamically stable hydrocarbon. At low pres-
sures and all temperatures, all hydrocarbons are thermodynam-
ically unstable relative to methane or methane plus carbon
(either graphite or amorphous carbon). At normal temperatures
and pressures, the evolution of methane will dominate and
effectively exhaust the H–C system of its elemental components.
Because methane is the sole hydrocarbon specie that is thermo-
dynamically stable at low pressures, the chemical Affinities of
each of the heavier species have been calculated in comparison
with methane. Accordingly, the chemical Affinity calculated for
the thermodynamic stability of, for example, the methane 7
(n-octane ⫹hydrogen) system is that for the reaction CH
4
3
1兾8C
8
H
18
⫹7兾8H
2
.
The chemical potentials of the hydrocarbon and methane
molecules and the resulting thermodynamic Affinities of the
(methane 3hydrocarbon ⫹hydrogen) system have been eval-
uated for the n-alkanes from methane through C
20
H
42
. In Fig. 2
are shown the Gibbs energies for the set of hydrocarbons
methane (CH
4
) and the n-alkanes, ethane (n-C
2
H
6
), hexane
(n-C
6
H
14
), and decane (n-C
10
H
22
). These thermodynamic vari-
ables have been determined at pressures ranging from 1 to
100,000 bar and at the supercritical temperature 1,000 K, the
temperature of which corresponds conservatively to the geolog-
ical regime characterized by the respective pressures of transition.
The values of the SPHCT parameters, c,
, and Y, for the
individual compounds that have been used were taken from van
Pelt et al. (27), and the reference values of the chemical
potentials of the pure component were taken from standard
reference tables (3).
The results of the analysis are shown graphically for the temper-
ature 1,000 K in Fig. 2. These results demonstrate clearly that all
Fig. 2. Gibbs energies of methane and of the H–C system [(1兾n)CnH2n⫹2⫹
(n⫺1)兾nH2].
Kenney et al. PNAS
兩
August 20, 2002
兩
vol. 99
兩
no. 17
兩
10979
PHYSICSCHEMISTRY
hydrocarbon molecules are unstable chemically and thermodynam-
ically relative to methane at pressures less than ⬇25 kbar for the
lightest, ethane, and 40 kbar for the heaviest n-alkane shown,
decane.
The results of this analysis, shown graphically in Fig. 2, establish
clearly the following:
(i) With the exception of methane, heavier hydrocarbon mole-
cules of higher chemical potentials are not generated sponta-
neously in the low-pressure regime of methane synthesis.
(ii) All hydrocarbon molecules other than methane are high-
pressure polymorphs of the H–C system and evolve sponta-
neously only at high pressures, greater than at least 25 kbar
even under the most favorable circumstances.
(iii) Contrar y to experience of refiner y operations conducted at
low pressures, heavier alkanes are not unstable and do not
necessarily decompose at elevated temperatures. Contrarily, at
high pressures, methane transforms into the heavier alkanes,
and the transformation processes are enhanced by elevated
temperatures.
The theoretical analyses reported here describe the high-pressure
evolution of hydrocarbons under the most favorable chemical
conditions. Therefore, although this analysis describes the thermo-
dynamic stability of the H–C system, it does not explicitly do the
same for the genesis of natural petroleum in the conditions of the
depths of the Earth. The chemical conditions of the Earth, partic-
ularly near its surface, are oxidizing, not reducing; of the gases in the
Earth’s atmosphere and crust, hydrogen is significantly absent, and
methane is a very minor constituent.
Although both methane and heavier hydrocarbons were present
in the carbonaceous meteorites that participated in the accretion
process of the formation of the Earth, such molecules were unlikely
to have surv ived in their initial composition. The heat of impact that
accompanied accretion most likely would have caused decompo-
sition of heavier hydrocarbons and the release of methane. For both
the theoretical analyses described in this section and the experi-
mental investigations described in section 5, the conservative per-
spective has been taken that hydrocarbons evolve from the solid,
abiotic carbon compounds and vestigial water present in the upper
mantle of the Earth.
5. Experimental Demonstration of Hydrocarbon Genesis Under
Thermodynamic Conditions Typical of the Depths of the Earth
Because the H–C system typical of petroleum is generated at
high pressures and exists only as a metastable me´lange at
laboratory pressures, special high-pressure apparatus has been
designed that permits investigations at pressures to 50 kbar and
temperatures to 1,500°C, and which also allows rapid cooling
while maintaining high pressures (28). The importance of this
latter ability cannot be overstated; for to examine the sponta-
neous reaction products, the system must be quenched rapidly to
‘‘freeze in’’ their high-pressure, high-temperature distribution.
Such a mechanism is analogous to that which occurs during
eruptive transport processes responsible for kimberlite ejecta
and for the stability and occurrence of diamonds in the crust of
the Earth.
Experiments to demonstrate the high-pressure genesis of petro-
leum hydrocarbons have been carried out using only 99.9% pure
solid iron oxide, FeO, and marble, CaCO
3
, wet with triple-distilled
water. There were no biotic compounds or hydrocarbons admitted
to the reaction chamber. The use of marble instead of elemental
carbon was intentionally conservative. The initial carbon com-
pound, CaCO
3
, is more oxidized and of lower chemical potential,
all of which rendered the system more resistant to the reduction of
carbon to form heavy alkanes than it would be under conditions of
the mantle of the Earth. Although there has been observed igneous
CaCO
3
(carbonatite) of mantle origin, carbon should be more
reasonably expected to exist in the mantle of the Earth as an
element in its dense phases: cubic (diamond), hexagonal (lonsda-
leite), or random-close pack (chaoite).
Pressure in the reaction cell, as described in ref. 25, of volume 0.6
cm
3
was measured by a pressure gauge calibrated using data of the
phase transitions of Bi, Tl, and PbTe. The cell was heated by a
cylindrical graphite heater; its temperature was measured using a
chromel-alumel thermocouple and was regulated within the range
⫾5°C. Both stainless steel and platinum reaction cells were used; all
were constructed to prevent contamination by air and provide
impermeability during and after each experimental run.
The reaction cell was brought from 1 bar to 50 kbar gradually at
a rate of 2 kbar兾min and from room temperature to the elevated
temperatures of investigation at the rate of 100 K兾min. The cell and
reaction chamber were held for at least1hateach temperature for
which measurements were taken to allow the H–C system to come
to thermodynamic equilibrium. The samples thereafter were
quenched rapidly at the rate of 700°C兾sec to 50°C and from 50°C
to room temperature over several minutes while maintaining the
high pressure of investigation. The pressure was then reduced
gradually to 1 bar at the rate of 1 kbar兾min. The reaction cell was
then heated gently to desorb the hydrocarbons for mass spectrom-
eter analysis using an HI-120 1B mass spectrometer equipped with
an automatic system of computerized spectrum registration. A
specially designed high-temperature gas probe allowed sampling
the cell while maintaining its internal pressure.
At pressures below 10 kbar, no hydrocarbons heavier than
methane were present. Hydrocarbon molecules began to evolve
above 30 kbar. At 50 kbar and at the temperature of 1,500°C, the
system spontaneously evolved methane, ethane, n-propane,
2-methylpropane, 2,2-dimethylpropane, n-butane, 2-methylbutane,
n-pentane, 2-methylpentane, n-hexane, and n-alkanes through
C
10
H
22
, ethene, n-propene, n-butene, and n-pentene in distribu-
tions characteristic of natural petroleum. The cumulative abun-
dances of the subset of evolved hydrocarbons consisting of methane
and n-alkanes through n-C
6
H
14
are shown in Fig. 3 as functions of
temperature. Methane (on the right scale) is present and of
abundance ⬇1 order of magnitude greater than any single com-
Fig. 3. Cumulative abundances of n-alkanes through n-C6H14 on left ordinate,
methane abundance on right, as functions of temperature at the pressure of
40 kbar. (Scales are in ppm.)
10980
兩
www.pnas.org兾cgi兾doi兾10.1073兾pnas.172376899 Kenney et al.
ponent of the heavier n-alkanes, although as a minor component of
the total H–C system. That the extent of hydrocarbon evolution
becomes relatively stable as a function of temperature above
⬇900°C, both for the absolute abundance of the individual hydro-
carbon species as well as for their relative abundances, argues that
the distributions observed represent thermodynamic equilibrium
for the H–C system. That the evolved hydrocarbons remain stable
over a range of temperatures increasing by more than 300 K
demonstrates the third prediction of the theoretical analysis: Hy-
drocarbon molecules heavier than methane do not decompose with
increasing temperature in the high-pressure regime of their genesis.
6. Discussion and Conclusions
The pressure of 30 kbar, at which the theoretical analyses of
section 4 predicts that the H–C system must evolve ethane and
heavier hydrocarbon compounds, corresponds to a depth of
more than 100 km. The results of the theoretical analysis shown
in Fig. 2 clearly establish that the evolution of the molecular
components of natural petroleum occur at depth at least as great
as those of the mantle of the Earth, as shown graphically in Fig.
4, in which are represented the thermal and pressure lapse rates
in the depths of the Earth.
As noted, the theoretical analyses reported in section 4 describe
the high-pressure evolution of hydrocarbons under the most favor-
able chemical conditions. The theoretical calculations for the evo-
lution of hydrocarbons posited the presence of methane, the genesis
of which must itself be demonstrated in the depths of the Earth
consistent with the pressures required for the evolution of heavier
hydrocarbons. Furthermore, the multicomponent system analyzed
theoretically included no oxidizing reagents that would compete
with hydrogen for both the carbon and any free hydrogen. The
theoretical analysis assumed also the possibility of at least a
metastable presence of hydrogen. Therefore, the theoretical results
must be considered as the determination of minimum boundary
conditions for the genesis of hydrocarbons. In short, the genesis of
natural petroleum must occur at depths not less than ⬇100 km, well
into the mantle of the Earth. The experimental observations
reported in section 5 confirm theoretical predictions of section 4,
and demonstrate how, under high pressures, hydrogen combines
with available carbon to produce heavy hydrocarbon compounds in
the geochemical environment of the depths of the Earth.
Notwithstanding the generality and first-principles rigor with
which the present theoretical analysis has used, the results of the
theoretical analyses here reported are robustly independent of the
details of any reasonable mathematical model. The results of this
theoretical analysis are strongly consistent with those developed
previously by Chekaliuk and Kenney (29 –32) using less accurate
formal tools. The analysis of the H–C system at high pressures and
temperatures has been impeded previously by the absence of
reliable equations of state that could describe a chemically
reactive, multicomponent system at densities higher than such of
its normal liquid state in ordinary laboratory conditions and at
high temperatures. The first analyses used the (plainly inade-
quate) Tait equation (33); later was used the quantum mechan-
ical Law of Corresponding States (34); more recently has been
applied the single-fluid model of the SPHCT (31, 32). None-
theless, all analyses of the chemical stability of the H–C system
have shown results that are qualitatively identical and quantita-
tively very similar: all show that hydrocarbons heav ier than methane
cannot evolve spontaneously at pressures below 20–30 kbar.
The H–C system does not spontaneously evolve heavy hydrocar-
bons at pressures less than ⬇30 kbar, even in the most favorable
thermodynamic environment. The H–C system evolves hydrocar-
bons under pressures found in the mantle of the Earth and at
temperatures consistent with that environment.
This article is dedicated to the memory of Nikolai Alexandrovich
Kudryavtsev, who enunciated what has become the modern Russian–
Ukrainian theory of abyssal, abiotic petroleum origins (35), and to the
late Academician E. B. Chekaliuk (J.F.K.).
1. Lomonosov, M. V. (1757) Slovo o Reshdinii Metallov ot Tryaseniya Zemli (Akadimii Nauk,
St. Petersburg).
2. De Donder, T. (1936) L’Affinite´(Gautier-Villars, Paris).
3. Kondepudi, D. & Prigogine, I. (1998) Modern Thermodynamics: From Heat Engines to
Dissipative Structures (Wiley, New York).
4. Prigogine, I. & Defay, R. (1954) Chemical Thermodynamics (Longmans, London).
5. U.S. Bureau of Standards (1946 –1952) Selected Properties of Hydrocarbons, A.P.I. Project 41
(U.S. Bureau of Standards, Washington, DC).
6. Pavlı´cek, J., Nezbeda, I. & Boublı´k, T. (1979) Czech. J. Phys. B29, 1061–1070.
7. Boublı´k, T. (1981) Mol. Phys. 42, 209 –216.
8. Prigogine, I. (1957) Molecular Theory of Solutions (North–Holland, Amsterdam).
9. Bogolyubov, N. N. (1946) Problems of a D ynamical Theory in Statistical Mechanics (GITTL, Moscow).
10. Feynman, R. P. (1972) Statistical Mechanics: A Set of Lectures (Addison–Wesley, New York).
11. Yukhnovskii, I. R. (1987) Phase Transitions of the Second Order: The Method of Collective
Variables (World Scientific, Singapore).
12. Reiss, H., Frisch, H. L. & Lebowitz, J. L. (1959) J. Chem. Phys. 31, 369 –380.
13. Reiss, H., Frisch, H. L. & Lebowitz, J. L. (1964) in The Equilibrium Theor y of Classical Fluids,
eds. Frisch, H. L. & Lebowitz, J. L. (W. A. Benjamin, New York), pp. II-299 –II-302.
14. Reiss, H. (1965) in Advances in Chemical Physics, ed. Prigogine, I. (Interscience, New York), Vol.
9, pp. 1–84.
15. Reiss, H. (1977) in Statistical Mechanics and Statistical Methods in Theory and Application: A
Tribute to Elliott W. Montroll, ed. Landman, U. (Plenum, London), pp. 99 –140.
16. Kenney, J. F. & Deiters, U. K. (2000) Phys. Chem. Chem. Phys. 2, 3163–3174.
17. Vera, J. H. & Prausnitz, J. M. (1972) Chem. Eng. J. 3, 113.
18. Beret, S. & Prausnitz, J. M. (1975) AIChE J. 21, 1123–1132.
19. Donohue, M. D. & Prausnitz, J. M. (1978) AIChE J. 24, 849 –860.
20. Bethe, H. A. & Kirkwood, J. G. (1939) J. Chem. Phys. 7, 578 –582.
21. Huang, K. & Yang, C. N. (1957) Phys. Rev. 105, 767–775.
22. Kenney, J. F. (1998) Fluid Phase Equilibria 148, 21–47.
23. Sandler, S. I. (1985) Fluid Phase Equilibria 19, 233–257.
24. Lee, R. J. & Chao, K. C. (1987) Mol. Phys. 61, 1431–1442.
25. Kim, C.-H., Vimalchand, P., Donohue, P. & Sandler, S. I. (1986) AIChE J. 32, 1726 –1734.
26. van Pelt, A., Peters, C. J. & de Swaan Arons, J. (1991) J. Chem. Phys. 95, 7569–7575.
27. van Pelt, A., Peters, C. J. & de Swaan Arons, J. (1992) Fluid Phase Equilibria 74, 67–83.
28. Nikolaev, N. A. & Shalimov, M. D. (1999) Patent RF 1332598, MKE 601; (1999) Bull. Izobretenii
(Russia) 7, 257, 23兾20.
29. Chekaliuk, E. B. (1975) in Regularities of Formation of Commercial Oil and Gas Fields (Nauk ova
Dumka, Kiev, Ukraine), pp. 66 –76.
30. Kenney, J. F. (1993) Bull. A. P. S. 38, 1508.
31. Kenney, J. F. (1995) in Joint XV AIR兾APT International Conference on High-Pressure Physics and
Technology (Warsaw), pp. 843–845.
32. Kenney, J. F. (1997) in The Prospects for Oil and Gas Potential of the Crystalline Basement in the
Territory of Tatar stan and Volga-Kama Region, ed. Romanov, G. V. (Russian Academy of
Sciences, Kazan, Russia), pp. 43–52.
33. Chekaliuk, E. B. (1971) The Ther modynamic Basis for the Theory of the Abiotic Genesis of
Petroleum (Naukova Dumka, Kiev, Ukraine).
34. Chekaliuk, E. B. & Kenney, J. F. (1991) Proc. Am. Phys. Soc. 36, 347.
35. Kudryavtsev, N. A. (1951) Petroleum Econ. (Neftianoye Khoz yaistvo) 9, 17–29.
Fig. 4. Pressure and temperature in the depths of the Earth.
Kenney et al. PNAS
兩
August 20, 2002
兩
vol. 99
兩
no. 17
兩
10981
PHYSICSCHEMISTRY
