338
The Chemistry and Technology of Petroleum
sulfur) casts serious doubts upon the authenticity of the hypothetical large polynuclear aromatic
systems that have been postulated as a major part of the asphaltene structure.
There is now adequate evidence to show that asphaltenes contain smaller polynuclear aromatic systems than has previously been, or even continues to be acknowledged. In addition, spectroscopic studies (before and after derivative formation) have indicated the presence of carbazole nitrogen and amide
nitrogen. Nonaqueous potentiometric titration has confirmed the presence of these nitrogen types and
has also provided evidence for the presence of pyridine nitrogen and indole nitrogen in asphaltenes.
Oxygen fractions vary from the phenol oxygen and polyhydroxyphenols to quinone and amide oxygen.
These observations require some readjustments to the previous postulates of the asphaltene structure. Thus, previously conceived structures in which the sole aromatic system is a large polynuclear
alkylated system are considered unlikely in view of these data. In fact, the current evidence is
indicative of a much more open (rather than highly condensed) structure. From these investigations,
a key structural feature of petroleum asphaltenes is the occurrence of small polynuclear aromatic
systems, benzothiophenes, dibenzothiophenes, and n-paraffins within the asphaltene molecule.
The manner in which these moieties appear within the asphaltene molecule must for the present remain largely speculative, but above all it must be remembered that asphaltenes are a solubility class and, as such, may be an accumulation of (literally) thousands of structural entities.
Hence, caution is advised against combining a range of identified products into one (albeit hypothetical) structure.
Highly condensed polynuclear aromatic structures, which have been used to explain the high
yields of thermal coke from asphaltenes, are not consistent with the natural product origins of petroleum asphaltenes. Such entities require conditions of temperature and pressure for their formation
that are more extreme than those currently recognized as responsible for petroleum maturation.
There is evidence that can be used in support of the formation of a nonvolatile residue (coke) being
formed from small condensed (≤6) ring systems rather than requiring a large condensed aromatic
ring system. Heat-resistant polymers containing aromatic and heterocyclic units, such as polyquinolines and polyquinoxalines, have a strong tendency to form large condensed systems during pyrolysis and finally carbonize. Formulae that include the occurrence of smaller polynuclear aromatic
hydrocarbon systems in asphaltenes are consistent with such behavior when it is recognized that the
majority of the nitrogen, oxygen, and sulfur species accumulate in the nonvolatile residue.
Thermal studies using model compounds confirm that volatility of the fragments is a major
influence in carbon residue formation and a pendant-core model for the high molecular weight constituents of petroleum has been proposed (Wiehe, 1994). In such a model, the scission of alkyl side
chains occurs thereby leaving a polar core of reduced volatility that commences to produce a carbon
residue (Speight, 1970b, 1994; Wiehe, 1994).
The danger in all of these studies is the in attempts to link together fragmented molecules produced
by thermal processes into a structure that is believed to be real. Functionality and polynuclear aromatic
ring systems are real, but combination of these parameters into one or even several structures can be
misleading. Molecular models can be an excellent aid to understanding process chemistry and physics.
As long as the model allows predictability of process behavior, use of the model is warranted. Caution
is advised, however. Because of the complexity of the asphaltene fraction, there is no one method that
will guarantee the development of a suitable molecular model. The effort must be multidimensional,
taking into account all the properties and characteristics of the material. Extreme caution is advised
when average structures are used since averages can be meaningless and, worse case, misleading
(Speight, 1994). Indeed, even when a multidimensional approach is employed for the derivation of a
model, there is no guarantee that the model is a true reflection of the molecular types in the asphaltene
fraction. It must also be recognized that any model developed may not be the only, or final, answer to
the problem. As techniques improve and knowledge evolves so must the model.
The constituents of the asphaltene fraction are macromolecules insofar as the molecular weights
are generally accepted to be at least several hundred and more often approaching 2000. In addition,
The Chemistry and Technology of Petroleum
sulfur) casts serious doubts upon the authenticity of the hypothetical large polynuclear aromatic
systems that have been postulated as a major part of the asphaltene structure.
There is now adequate evidence to show that asphaltenes contain smaller polynuclear aromatic systems than has previously been, or even continues to be acknowledged. In addition, spectroscopic studies (before and after derivative formation) have indicated the presence of carbazole nitrogen and amide
nitrogen. Nonaqueous potentiometric titration has confirmed the presence of these nitrogen types and
has also provided evidence for the presence of pyridine nitrogen and indole nitrogen in asphaltenes.
Oxygen fractions vary from the phenol oxygen and polyhydroxyphenols to quinone and amide oxygen.
These observations require some readjustments to the previous postulates of the asphaltene structure. Thus, previously conceived structures in which the sole aromatic system is a large polynuclear
alkylated system are considered unlikely in view of these data. In fact, the current evidence is
indicative of a much more open (rather than highly condensed) structure. From these investigations,
a key structural feature of petroleum asphaltenes is the occurrence of small polynuclear aromatic
systems, benzothiophenes, dibenzothiophenes, and n-paraffins within the asphaltene molecule.
The manner in which these moieties appear within the asphaltene molecule must for the present remain largely speculative, but above all it must be remembered that asphaltenes are a solubility class and, as such, may be an accumulation of (literally) thousands of structural entities.
Hence, caution is advised against combining a range of identified products into one (albeit hypothetical) structure.
Highly condensed polynuclear aromatic structures, which have been used to explain the high
yields of thermal coke from asphaltenes, are not consistent with the natural product origins of petroleum asphaltenes. Such entities require conditions of temperature and pressure for their formation
that are more extreme than those currently recognized as responsible for petroleum maturation.
There is evidence that can be used in support of the formation of a nonvolatile residue (coke) being
formed from small condensed (≤6) ring systems rather than requiring a large condensed aromatic
ring system. Heat-resistant polymers containing aromatic and heterocyclic units, such as polyquinolines and polyquinoxalines, have a strong tendency to form large condensed systems during pyrolysis and finally carbonize. Formulae that include the occurrence of smaller polynuclear aromatic
hydrocarbon systems in asphaltenes are consistent with such behavior when it is recognized that the
majority of the nitrogen, oxygen, and sulfur species accumulate in the nonvolatile residue.
Thermal studies using model compounds confirm that volatility of the fragments is a major
influence in carbon residue formation and a pendant-core model for the high molecular weight constituents of petroleum has been proposed (Wiehe, 1994). In such a model, the scission of alkyl side
chains occurs thereby leaving a polar core of reduced volatility that commences to produce a carbon
residue (Speight, 1970b, 1994; Wiehe, 1994).
The danger in all of these studies is the in attempts to link together fragmented molecules produced
by thermal processes into a structure that is believed to be real. Functionality and polynuclear aromatic
ring systems are real, but combination of these parameters into one or even several structures can be
misleading. Molecular models can be an excellent aid to understanding process chemistry and physics.
As long as the model allows predictability of process behavior, use of the model is warranted. Caution
is advised, however. Because of the complexity of the asphaltene fraction, there is no one method that
will guarantee the development of a suitable molecular model. The effort must be multidimensional,
taking into account all the properties and characteristics of the material. Extreme caution is advised
when average structures are used since averages can be meaningless and, worse case, misleading
(Speight, 1994). Indeed, even when a multidimensional approach is employed for the derivation of a
model, there is no guarantee that the model is a true reflection of the molecular types in the asphaltene
fraction. It must also be recognized that any model developed may not be the only, or final, answer to
the problem. As techniques improve and knowledge evolves so must the model.
The constituents of the asphaltene fraction are macromolecules insofar as the molecular weights
are generally accepted to be at least several hundred and more often approaching 2000. In addition,