326
The Chemistry and Technology of Petroleum
yield (and, therefore, on an estimation of crude oil composition). However, the resin occluded within
(or adsorbed onto) the asphaltene during the separation procedure has a considerable effect on the
observed molecular weight and, presumably, on the degree of association of the asphaltenes. Indeed,
the speculative concept that asphaltenes release the final vestiges of this resin only upon swelling by
a solvent such as pyridine is also worthy of consideration.
Attention should be focused upon the delineation of various structural types in the asphaltenes.
Since petroleum is actually a continuum of structural types, there is little to be inferred from this
study other than the influence of degree of association or on the degree of polymerization (for those
concepts based on the concept of a polymer structure).
If, however, the structural types vary from the resins to the asphaltenes, it may be necessary to
reassess the concepts of molecular types that have been proposed heretofore. In addition, the concept that the resins may be entrapped within the asphaltene matrix (and may even be analogous to a
clathrate compound) warrants further investigation.
It is worthy of mention at this point, since the theory of the physical structure of petroleum is
based upon the results of the method, that the specific viscosity of asphaltene solutions has also been
employed to determine asphaltene molecular weights.
The method essentially requires a determination of the relative viscosity, nr, of known solute–
solvent systems, when the constant k is derived by the relationship
log n r = ckM
where
c is the concentration (g/100 mL)
M is the molecular weight
Thus, subsequent application of the relationship to solute–solvent systems involving asphaltenes and
maltenes allowed Mack, in 1932, to derive molecular weights of the order of 5000 at 0°C (32°F) to
1800 at 120°C (250°F). Thus, it is not surprising that the postulate of the existence of asphaltene
clusters within the maltenes evolved and has remained to this time.
Application of this procedure to systems involving Athabasca asphaltenes allows derivation of
values for the constant k, but transposition of k from one solute–solvent system to another gives
inconsistent results. For example, the constants derived from naphthalene–benzene or biphenyl–
benzene systems give unexpectedly low molecular weights when applied to asphaltene–benzene
systems. Indeed, application of these constants to petrolene–solute systems gives molecular weights
of naphthalene and biphenyl in error by some several hundred percent.
Application of the constant k for the biphenyl–benzene system to the asphaltene–petrolene system
gives molecule weights within an acceptable range. However, the validity is suspect because of the
unsatisfactory results obtained when the constants from one solute–solvent system are transposed
to a different solute–solvent system in which the molecular weight of the solute is known. This is
emphasized to an even greater extent by application of the constant derived from the asphaltene–
benzene system to the asphaltene–petrolene system, which yields molecular weights of the order of
17,000–60,000.
In view of these inconsistencies, the viscosity method of determining asphaltene molecular
weights must be regarded as suspect when the constants k, as derived for one solute–solvent system,
is applied to a completely different solute–solvent system. Indeed, Mack’s original work involved
transposition of data from earlier work (Kendall and Monroe, 1917) and must be regarded as questionable, since these authors consider their own results to be indecisive because data from different
solvent systems containing the same solute are in conflict. Even though preliminary application
of some of the viscosity data yields asphaltene molecular weights within an acceptable range, the
remaining data indicate several inconsistencies in the method, and recognition of this causes some
revision of previous theories of the physical structure of crude oil.
The Chemistry and Technology of Petroleum
yield (and, therefore, on an estimation of crude oil composition). However, the resin occluded within
(or adsorbed onto) the asphaltene during the separation procedure has a considerable effect on the
observed molecular weight and, presumably, on the degree of association of the asphaltenes. Indeed,
the speculative concept that asphaltenes release the final vestiges of this resin only upon swelling by
a solvent such as pyridine is also worthy of consideration.
Attention should be focused upon the delineation of various structural types in the asphaltenes.
Since petroleum is actually a continuum of structural types, there is little to be inferred from this
study other than the influence of degree of association or on the degree of polymerization (for those
concepts based on the concept of a polymer structure).
If, however, the structural types vary from the resins to the asphaltenes, it may be necessary to
reassess the concepts of molecular types that have been proposed heretofore. In addition, the concept that the resins may be entrapped within the asphaltene matrix (and may even be analogous to a
clathrate compound) warrants further investigation.
It is worthy of mention at this point, since the theory of the physical structure of petroleum is
based upon the results of the method, that the specific viscosity of asphaltene solutions has also been
employed to determine asphaltene molecular weights.
The method essentially requires a determination of the relative viscosity, nr, of known solute–
solvent systems, when the constant k is derived by the relationship
log n r = ckM
where
c is the concentration (g/100 mL)
M is the molecular weight
Thus, subsequent application of the relationship to solute–solvent systems involving asphaltenes and
maltenes allowed Mack, in 1932, to derive molecular weights of the order of 5000 at 0°C (32°F) to
1800 at 120°C (250°F). Thus, it is not surprising that the postulate of the existence of asphaltene
clusters within the maltenes evolved and has remained to this time.
Application of this procedure to systems involving Athabasca asphaltenes allows derivation of
values for the constant k, but transposition of k from one solute–solvent system to another gives
inconsistent results. For example, the constants derived from naphthalene–benzene or biphenyl–
benzene systems give unexpectedly low molecular weights when applied to asphaltene–benzene
systems. Indeed, application of these constants to petrolene–solute systems gives molecular weights
of naphthalene and biphenyl in error by some several hundred percent.
Application of the constant k for the biphenyl–benzene system to the asphaltene–petrolene system
gives molecule weights within an acceptable range. However, the validity is suspect because of the
unsatisfactory results obtained when the constants from one solute–solvent system are transposed
to a different solute–solvent system in which the molecular weight of the solute is known. This is
emphasized to an even greater extent by application of the constant derived from the asphaltene–
benzene system to the asphaltene–petrolene system, which yields molecular weights of the order of
17,000–60,000.
In view of these inconsistencies, the viscosity method of determining asphaltene molecular
weights must be regarded as suspect when the constants k, as derived for one solute–solvent system,
is applied to a completely different solute–solvent system. Indeed, Mack’s original work involved
transposition of data from earlier work (Kendall and Monroe, 1917) and must be regarded as questionable, since these authors consider their own results to be indecisive because data from different
solvent systems containing the same solute are in conflict. Even though preliminary application
of some of the viscosity data yields asphaltene molecular weights within an acceptable range, the
remaining data indicate several inconsistencies in the method, and recognition of this causes some
revision of previous theories of the physical structure of crude oil.