Adlard E.R. (ed.) Chromatography in the Petroleum Industry
Подождите немного. Документ загружается.
122
Chapter
4
Other polymers that have been studied by packed-column HDC include dex-
tran [7], liposomes [41] and polystyrenes [42]. Mori
et
al.
[5] used a column that
is a hybrid between a packed column and a microcapillary column by sintering
together a bundle of small-diameter glass rods. The resulting column, consisting
of numerous parallel microcolumns with “hole” sizes of 0.9-1.4 pm, was used in
HDC
experiments on polystyrenes. Cheng [42] discussed the similarity and dif-
ference between size exclusion chromatography and packed-column
HDC,
using
separations
of
polystyrenes
in
columns packed with 16-20 and 1-4pm glass
beads.
The performance of column HDC can be expected to improve when very
small particles of a uniform size are used. Work in this field has been reported
by Kraak
et
al.
[8] and Stegeman
et
al.
[9] with separations of polystyrenes,
colloidal silica particles and proteins on impermeable silica spheres of 1.4-
2.67
pm
diameter. To compare their experiments with microcapillary HDC, they
used the hydraulic radius,
&,
i.e. the radius of a capillary having the same sur-
face-to-volume ratio as the packed bed column, to characterize the average value
of the flow channel radius. From this they calculated the aspect ratios of polysty-
renes in these separations.
&
can be obtained as
(4.16)
where
dp
is the particle size and
E
is the bed porosity. In their experiments with
the 1.4 pm silica beads column, with a porosity
of
E
=
0.380, the hydraulic radius
&
amounts to only 0.286 pm, a size that, up till now, has never been attained for
a true microcapillary
HDC
column where the smallest column radius reported is
0.6pm [18]. Good separation of six polystyrenes of molecular mass 2.75
X
lo6-
3.45
X
lo4 Da and toluene in
THF
were shown
in
less than
6
min analysis time
(Fig. 4.15). They fitted their experimental relative residence times to the calcu-
lated aspect ratios for polystyrenes and obtained
a
relation in the form of
Eq.
(4.7),
so
t
=
(1
+
U,
-
C$)-’,
but with a value of
C
=
3.7 instead of
C
=
2.698
that would be expected for random coil macromolecules in a good solvent in
a
cylindrical capillary. But as has already been pointed out in the theoretical sec-
tion, the exact value
of
C
is
not too important for
t
values in excess of say 0.95.
So
with their columns, it is possible to determine the size of eluting species from
relative residence times alone, for polymer sizes up to 16nm in their column
packed with 1.4 pm particles
(up
to 30 nm diameter in the column packed with
2.67
pm
material). This is important, as it indicates that a number of applications
that have been described
in
the section on microcapillary HDC can,
in
principle,
be performed on packed columns (although, at present, preparation of very uni-
form small particles and the efficient packing of these into a column is still a
problem). The main advantage
of
using a packed column is, of course, the fact
Hydrodynamic chromatography
of
polymers
123
I..,
0'
2
4
6
time
,
min
Fig.
4.15.
HDC
separation
of
polystyrenes and toluene in
a
column packed with 1.40pm diameter
non-porous silica spheres. Peaks
1-6,
polystyrenes
of
molecular mass
2750, 1260, 700,
310,
127
and
34.5
ma,
respectively; peak 7, toluene. Reprinted from
[9].
that detectors other than the ultraviolet detectors used in microcapillary
HDC
can be applied. Still, with the columns packed with these highly efficient, small
sized materials, not every standard liquid chromatographic detector can be used.
For instance, even a standard ultraviolet detector cannot be used as such, since a
detector cell volume of only a few microliters deteriorates the separation
of
the
very narrow eluting zones too much. Therefore, in
[8,9],
a small length of a
100pm internal diameter capillary was placed at the outlet of the
HDC
column
to act as the
UV
detector cell. Standard refractive index detectors cannot be used
either, but since this detection principle, using laser sources, has already been
described for use in capillary applications
[43-45],
this more or less universal
detection method could well be applied in
HDC
separations if the efficient, small
particle size packed columns were equipped with a capillary detection cell.
Another application of packed columns is the
HDC
separation that can take
place in the interstitial volume in a size exclusion chromatographic column
[46].
Thus, there is a possibility
of
having
a
chromatographic column that can handle
Refirences pp.
125-126
124
Chapter
4
r
I
I
I
1
I
0
10
20
30
40
50
time,
min
Fig. 4.16.
HDC-SEC
separation
of
polystyrenes and toluene on a column packed with
3
pm
porous
silica particles (pore size
6
nm). Peaks
1-8,
polystyrenes
of
molecular
mass
4000,
2200,
775,
336,
127,
43.9,
12.5
and
2.2
kDa,
respectively; peak
9,
toluene. Reprinted
from
[47].
a very wide range
of
molecular masses, the smaller components being separated
in
the pores of the packing material by a
GPC
mechanism, while the very large
molecules are separated by
HDC.
This has been studied by Stegeman
et
al.
[47]
with
GPC
columns containing
3
,urn
porous silica particles
(6
nm pore size) and
3
pm
cross-linked
poly(styrene-divinylbenzene)
particles
(10
and
30
nm pore
sizes). The separation of eight polystyrenes
and
toluene
on
the silica column by
this mixed mechanism is shown
in
Fig.
4.16.
It is claimed that the separation
range of this column
is
for molecular masses ranging fiom a few hundred up to
2
X
lo7
Da, a dynamic range that cannot be attained, without sacrificing selec-
tivity and efficiency, with columns working with a pure size exclusion mecha-
nism.
4.5
CONCLUSIONS
From this overview we have seen that
HDC
is used mainly for the separation
of
discrete particles rather than for polymeric species. This technique, which
uses only simple instrumentation, is potentially very useful for the separation
of
Hydrodynamic chromatography ofpolymers
125
polymers, especially those with very high molecular masses. It has been shown
that both with columns packed with very small impermeable spheres
of
uniform
size, and with microcapillary columns, information on the size of separated spe-
cies can be obtained without recourse to calibration standards. The working
range is for molecular radii ranging from say
1
nm up to about
500
nm. Cur-
rently, the detection in microcapillary HDC is limited to
UV
absorption, but
promising new detection principles based on laser technology are emerging.
4.6
REFERENCES
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
H. Small, J. Colloid Interface Sci., 48 (1974) 147-161.
H.
Small, Adv. Chromatogr., 15 (1977) 113-129.
H.
Small, Anal. Chem., 54 (1982) 892A-898A.
K.O. Pedersen, Biochem. Biophys. Suppl.,
1
(1962) 157-168.
S.
Mori, R.S. Porter and J.F. Johnson, Anal. Chem., 46 (1974) 1599-1602.
R.K. Prud’homme, G. Froiman and D.A. Hoagland, Carbohydrate Res., 106 (1982) 225-233.
D.A. Hoagland, K.A. Larson and R.K. Prud’homme, in Modern Methods of Particle Size
Analysis (H.G. Barth, Ed.), Wiley, New York, 1984, pp. 277-301.
J.C. Kraak,
R.
Oostervink,
H.
Poppe and
K.K.
Unger, Chromatographia, 27 (1989) 585-590.
G. Stegeman, R. Oostervink, J.C. Kraak and
H.
Poppe, J. Chromatogr., 506 (1990) 547-561.
M.E. Mullins and C.
On;
Int. J. Multiphase Flow, 5 (1979) 79-85.
R.J. Noel, K.M. Gooding, F.E. Regnier, D.M. Ball, C.
Orr
and M.E. Mullins, J. Chromatogr.,
A.W.J. Brough, D.E. Hillman and R.W. Perry, J. Chromatogr., 208 (1981) 175-182.
C.D. Shuster, J.R. Schroeder and D. Mclntyre, Rubber Chem. Techno]., 54 (1981) 882-891.
P.
Elie and M. Renaud, Entropie, 136 (1987) 33-40.
C.A. Silebi and J.G. DosRamos, J. Colloid Interface Sci., 130 (1989) 14-24.
J.G. DosRamos and
C.A.
Silebi, J. Colloid Interface Sci.,
133
(1989) 302-320.
R.
Tijssen, J.P.A. Bleumer and M.E. van Kreveld, J. Chromatogr., 260 (1983) 297-304.
R. Tijssen, 5. Bos and M.E. van Kreveld, Anal. Chem., 58 (1986) 3036-3044.
J.
Bos, R. Tijssen and M.E. van Kreveld, Anal. Chem., 61 (1989) 1318-1321.
J. Dodds, Analusis, 10 (1982) 109-119.
A.J. McHugh, CRC Crit. Rev. Anal. Chem., 15 (1984) 63-117.
E.A. DiMarzio and C.M. Guttman, J. Polyrn. Sci., Part B, Polyrn.
Lett.,
7 (1969) 267-272.
E.A. DiMarzio and C.M. Guttman, Macromolecules,
3
(1970) 131-146.
E.A. DiMarzio and C.M. Guttman, J. Chromatogr., 55 (1971) 83-97.
H.
Brenner and L.J. Gaydos, J. Colloid Interface Sci., 58 (1977) 312-356.
G.M. Mavrovouniotis and
H.
Brenner, J. Colloid Interface Sci, 124 (1988) 269-283.
M.V. Volkenstein, Configurational Statistics of Polymeric Chains, High Polymers Series, Vol.
XVII, Wiley Interscience, New York, 1963.
M.E. van Kreveld and N. van den Hoed, J. Chromatogr., 83 (1973)
11
1-124.
R.
Tijssen and J. Bos, in: Theoretical Advancement in Chromatography and Related Separa-
tion Techniques
(F.
Dondi and G. Guiochon, Eds.), NATO AS1 Proceedings, Ferrara 1991,
Kluwer, Dordrecht.
R. Tijssen, Sep. Sci. Techno].,
13
(1978) 681-720.
166 (1978) 373-382.
126
Chapter
4
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
G.V.
Schulz and H. Baumann, Makromol. Chem.,
I14 (1968) 122-138.
W. Mandema and
H.
Zeldenrust, Polymer,
18
(1977) 835-839.
F.
Zarrin and N.J. Dovichi, Anal. Chem.,
57 (1985) 1826-1829.
A.
Revillon and P. Boucher, J. Appl. Polym. Sci.: Appl. Polym. Symp.,
43 (1989) 115-128.
M. Tazaki and T.
Homma,
Rep. Prog. Polym. Phys. Jpn,
29 (1986) 29-3
1,
R.K.
Prud’homme and D.A. Hoagland, Sep. Sci. Technol.,
18 (1983) 121-134.
D.J.
Nagy, J. Colloid Interface Sci.,
93 (1983) 590-592.
G.R. McGowan and M.A.
Langhorst,
J. Colloid Interface Sci.,
89 (1982) 94-106.
M.A. Langhorst and F.W. Stanley, Polym. Mater. Sci. Eng.,
53 (1985) 446-448.
J.
Lecourtier
and
G.
Chauveteau, Macromolecules,
17 (1984) 1340-1343.
F.J.
Molinas Lucas, A.O. Vila
Buso,
P. Dieguez Bosch and J.E. Figueruelo Alejano, J. High
Resolut. Chromatogr.,
12 (1989) 560-561.
W. Cheng,
J.
Chromatogr.,
362 (1986) 309-324.
D.J. Bomhop
andN.J.
Dovichi, Anal. Chem.,
59 (1987) 1632-1636.
R.E.
Synovec, Anal. Chem.,
59 (1987) 2877-2884.
A.E.
Bruno, B. Krattiger,
F.
Maystre and H.M. Widmer, Anal. Chem.
63 (1991) 2689-2697.
P.G.
Squire,
A.
Magnus and M.E. Himmel, J. Chromatogr.
242 (1982) 255-266.
G.
Stegeman, J.C.
Kraak
and
H.
Poppe, 5. Chromatogr.
550
(1991)
721-739.
E.R.
Adlard
@id.),
Chromatography
in
the Petroleum Industry
Journal
of
Chromatography
Library
Series,
Vol.
56
Q
1995
Elsevier Science
B.V.
All
rights reserved
127
CHAPTER
5
Chromatography in petroleum
geochemistry
S.J.
Rowlanda
and
A.T.
Revillb
apeboleurn and Environmental Geochemistry Group, Department
of
Environmental Sciences, University
of
Plymouth, Drake Circus, Plymouth,
PL4
8AA,
UK;
bCSIRO
Division
of
Oceanography, Castray Esplanade,
Hobart9 Tasmania, Australia
5.1
INTRODUCTION
Petroleum is formed by the natural degradation of a heterogeneous macro-
molecular substance
known
as kerogen. Whilst a proportion of kerogen can be
made amenable to examination by chromatographic techniques if it is first de-
graded to more volatile compounds, it is for the analysis
of
petroleum itself that
chromatography has found most widespread use. Indeed, the complex nature of
petroleum makes the use of chromatographic methods more or less essential for
any detailed geochemical study. Pioneering studies, prior to the advent of mod-
ern day chromatography, employed distillation as the principal separation
method (see e.g.
[
13)
but whilst distillation remains valuable, the large scale of
apparatus and samples required, the expertise necessary and the time-consuming
nature of the work have made chromatography-based methods much more
popular for geochemistry
in
recent times. This chapter reviews some of the most
recent developments.
5.1.1
Recent reviews
The use of chromatographic and other methods
to
characterize kerogen has
been reviewed by Rullkotter and Michaelis
[2],
whilst the applications of high
temperature gas chromatography
(GC)
to geochemistry have also been discussed
recently
[3].
Horsfield
[4]
has reviewed the applications of pyrolysis techniques,
References
pp.
139-111
128
Chapter
5
such as pyrolysis-GC, to petroleum geochemistry and reviews published regu-
larly in
Analytical
Chemistry
include useful summaries of some of the applica-
tions of chromatography in this field up to about 1991 (see e.g. [5,6]). Several
up-to-date texts have appeared which discuss the analysis
of
so-called “bio-
logical marker” compounds in petroleum
[7-91.
These compounds are valuable
for
oil-source rock and oil-oil correlations and numerous applications of chroma-
tography in this area have been reported. Indeed, some techniques, such as GC
and GC-MS are used
so
frequently in this area that in the present review we have
selected only examples which are intended to illustrate particular advances.
Other chromatographic methods are used more rarely but represent areas of great
potential for future geochemical research and these are also discussed. Thus, we
have attempted
a
somewhat selective review, confining ourselves mainly to re-
cent (i.e. post-
1990)
research.
5.2
KEROGEN
AND
OTHER PETROLEUM MACROMOLECULES
By definition, kerogen is insoluble in organic solvents
[2]
and
is
not amenable
to conventional chromatographic fractionation. Many studies have used bulk
characterization techniques but these have provided relatively little molecular
information. However, chromatography has played an increasingly important
role, usually in conjunction with degradation techniques such as thermal or se-
lective and non-selective chemical degradation. The latter reduce the macro-
molecular kerogen to smaller sub-units which are more amenable to chroma-
tographic analysis. Oxidative degradation usually produces functionalized com-
pounds such as acids which require derivatization prior to GC or GC-MS
[
10,111. More selective chemical degradations have included sequential use of
desulphurizing agents such as nickel(0)cene and Li/EtNH2 for the characteriza-
tion of sulphur-rich kerogens and other macromolecular fractions of
oils
fol-
lowed by
LC,
TLC and GC
[
12,131. The more severe thermal techniques, such as
flash pyrolysis (py) are often interfaced directly to
a
chromatography system, for
which GC and GC-MS are most commonly used. Such techniques require small
samples, have a fast sample throughput and reduce the need for sample work-up.
A
significant advance in this technique was the quantification
of
pyrolysis prod-
ucts via the addition of internal standards
[
141.
The latter approach is illustrated
in
Fig.
5.1
which shows a py-GC-MS total ion current chromatogram of
a
sample
of
asphaltenes from an Iranian bunker oil admixed with poly-t-butylstyrene in-
ternal standard. Pyrolysis produces the t-butylstyrene monomer which can
be
used to semi-quantify the pyrolysis products of the asphaltenes.
Py-GC methods have found increasing
use
in the studies of kerogen
or
other
macromolecules
[
15,161. For example in the determination of sulphur-rich kero-
Chromatography
in
petroleum geochemistry
129
Internal standard
100
90
80
70
60
50
40
30
20
10
0
Scan
R.T.
TIC
Yane
I
I
I
/
n-lcosane
J
n-Triacon tane
500
26:
21
1000
52:
46
1500
79:
11
Fig.
5.1.
Pyrolysis-gas chromatography-mass spectrometry total ion current chromatogram
of
as-
phaltenes
(ca.
200
pg) isolated from Iranian bunker oil mixed with t-butylpolystyrene internal
standard
(10
pg)
and pyrolysed at
600°C
for
20
s.
Components from methane to triacontane can be
measured if the p+olysis behaviour
of
the internal standard is reproducible and linear. Kratos
MS
25
GC-MS operated with CDS
120
pyroprobe (modified). GC oven temperature
40
to
+300"C
at
5°C
/min. GC column:
30
m
X
0.32
mm
DB-1
(J
&
W).
gens, unique distributions
of
pyrolysis compounds have provided
prima facie
evidence that unsaturated kerogen moieties derived from algae take up sulphur
and
form
di- and polysulphide linkages
[
17,181.
In
other studies, py-GC-MS has
aided identification of the origins of polycyclic oligoterpenoids in crude oils via
resinous polymers
[19]
and has established that some kerogens are formed by
selective preservation
of algaenan-containing thin outer walls of Chlorophyceae
algae
[20].
The availability
of
a range of detectors is also important for the progress of
py-GC characterization of macromolecular geochemical materials. The most
common detector for
GC
is the flame ionization detector
(FID),
which is non-
selective and responds to almost any compound containing
C-H
bonds. The
References pp.
139-1
41
130
Chapter
5
main use
of
py-GC-FID is the “typing” of kerogens
[
141
on the basis
of
the dis-
tribution of pyrolysate compounds and their relative quantities. However, ele-
ment specific detectors also have important applications in py-GC of kerogens.
Perhaps the most common of these
is
the flame photometric detector (FPD)
which is selective for sulphur-containing compounds. The sensitivity
of
both the
FID and FPD allows the utilization
of
column splitting technology to obtain con-
current chromatograms from both detectors
[
181.
A
less widely used selective detector is the nitrogen-phosphorus detector
(NPD), generally used for nitrogen containing compounds. This has enabled
pyridines, quinolines and benzoquinolines to be observed in the pyrolysates
of
asphaltenes, coals and source rocks
[21].
The atomic emission detector (AED), which is a multi-element detector, has
also found increasing use and has typical limits
of
detection in the pg/s range for
several elements
[22].
For example, alkylpyrroles have been detected recently in
kerogen pyrolysates by py-GC with FID-AED and
MS
detection
[24].
AED al-
lows monitoring
of several elements at once and importantly these include car-
bon, sulphur, nitrogen and oxygen, making it potentially a valuable tool in kero-
gen analysis
[23].
An advantage of the technique is that the carbon chroma-
togram is very similar to that from an FID, making the
two
readily comparable.
In addition, the emission spectrum of
a
compound can be obtained to ensure it
contains the element
in
question. A drawback is that response is directly related
to
the strength
of
the emission line which for some elements, e.g. oxygen, is very
weak.
One impressive multidetection study involved py-GC coupled with FTIR and
FID
[25].
Fifty-nine pyrolysate compounds were determined in a peat deposit,
indicating a variety of decomposition reactions in the woody tissues. The use of
integrated chromatography detectors in similar studies may prove valuable in
future kerogen research.
5.3
GEOCHEMISTRY
OF
PETROLEUM
The degradation
of
kerogen which,
as
described above, is utilized by geo-
chemists to produce compounds which are easier to identify, is also accom-
plished by nature over much greater time spans and generally speaking, less se-
vere conditions. The product of this natural degradation or catagenesis is liquid
petroleum which is somewhat simpler to characterize by existing techniques than
the precursor macromolecules and a wide variety of chromatographic methods
have assisted in this characterization. Since fresh petroleum comprises compo-
nents which range from simple hydrocarbons to bulky micelles, a significant
proportion can be lost by evaporation if care is not taken in sampling and analy-
Chromatography
in
petroleum geochemistry
131
sis (e.g. use of special cylinders, such as Welker bombs, where the sample is
kept at its original high pressure) or pressurized oil syringe sampling methods
[26,27].
For this reason, most geochemical analyses are conducted on a distillate
fraction
(so
called CIS+ fraction)
of
petroleum where the volatile components
have been removed in a controlled manner. However, this may imply a signifi-
cant loss of information and “whole oil analysis” is usually to be preferred if at
all possible
[28].
5.3.1
LC, TLC
and
TLC-FID
The instrumentation and equipment available for column LC and planar
chromatography has been reviewed
[6].
Column LC with silica and/or alumina
stationary phases is a standard technique for petroleum fractionation, with or
without assisted gas pressure, and typically results in the isolation of so-called
“saturate”, “aromatic” and “polar” compound classes
[7].
If care
is
taken to
avoid losses during subsequent evaporation of solvents from these fractions (e.g.
by use of Kuderna-Danish apparatus
[29]),
the resulting data can be reasonably
quantitative, particularly if internal standards are used to calibrate the measure-
ments
[7,30].
However, some losses are unavoidable and recovery from the col-
umn is seldom
100%.
Of
the classes eluted, most detailed studies have continued
to concentrate on the examination of the “saturate” and “aromatic” hydrocarbon
fractions. However, a notable expansion of research into the “heavy ends” of
crudes (e.g. asphalt, tar and pitch fractions) has seen quite widespread applica-
tion of LC, for instance for the preparative isolation of acid-base fiactions of
asphaltenes and maltenes
[3
11
and for characterization of asphaltenes before and
after visbreaking
[32].
LC has also proved extremely useful for the isolation of compounds from
complex geochemical extracts of oils, coals and shales for structure elucidation.
The use of LC with silica gel and alumina columns and a variety of solvent
mixtures, followed by HPLC, allowed the isolation of seven alkylporphyrins and
thirteen porphyrin carboxylic acids from Messel shale
[33],
whilst pure vanadyl
porphyrins were isolated from other lacustrine sediments by LC on alumina fol-
lowed by propanesulphonic acid-bonded silica and HPLC
[34].
Others have iso-
lated aromatic compounds such as basic nitrogen compounds (azaarenes) on sil-
ica
[35]
and
17
synthetic trimethylbiphenyls on silica and alumina
[36].
Novel
C-ring cleaved triterpenoid aromatics were identified in Tertiary brown coal af-
ter LC fractionation of extracts on deactivated alumina and silica
[37].
TLC is
a
cheap and rapid technique (especially if automated) which has found
widespread application for oil fractionation. However, losses of volatiles are
usually significant.
References
pp.
139-1
41