Upload
george-v
View
218
Download
4
Embed Size (px)
Citation preview
This article was downloaded by: [University of Sussex Library]On: 01 March 2013, At: 03:15Publisher: Taylor & FrancisInforma Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,37-41 Mortimer Street, London W1T 3JH, UK
Energy SourcesPublication details, including instructions for authors and subscription information:http://www.tandfonline.com/loi/ueso19
A Study of the Structure of Petroleum Asphaltenesand Related Substances by Proton Nuclear MagneticResonanceTeh Fu Yen a , Wen Hui Wu a & George V. Chilingar aa School of Engineering University of Southern California, Los Angeles, CaliforniaVersion of record first published: 12 Jul 2010.
To cite this article: Teh Fu Yen , Wen Hui Wu & George V. Chilingar (1984): A Study of the Structure of Petroleum Asphaltenesand Related Substances by Proton Nuclear Magnetic Resonance, Energy Sources, 7:3, 275-304
To link to this article: http://dx.doi.org/10.1080/00908318408908088
PLEASE SCROLL DOWN FOR ARTICLE
Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-conditions
This article may be used for research, teaching, and private study purposes. Any substantial or systematicreproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form toanyone is expressly forbidden.
The publisher does not give any warranty express or implied or make any representation that the contentswill be complete or accurate or up to date. The accuracy of any instructions, formulae, and drug doses shouldbe independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims,proceedings, demand, or costs or damages whatsoever or howsoever caused arising directly or indirectly inconnection with or arising out of the use of this material.
A Study of the Structure ofPetroleum Asphaltenes andRelated Substances by ProtonNuclear Magnetic Resonance
Teh Fu Yen, Wen Hui Wu*, and George V.ChilingarSchool of EngineeringUniversity of Southern CaliforniaLos Angeles, California
Abstract The distribution of hydrogen according to structural type in a series of representative asphaltic materials wasinvestigated by proton nuclear magnetic resonance (NMR). Determination of the fraction of total hydrogens, which are saturated, hs' by using the NMR technique was shown to be moreaccurate than by infrared spectrophotometry. Also, by utilizingdata on the diameter of the aromatic sheets, La' obtained by xray diffraction, the shape of the aromatic centers was shown tovary but, on the whole, to be highly condensed and in no casechainlike in nature. Approximately 48 to 70 percent of the positions on the edge of the aromatic centers are substituted byatoms other than hydrogen. While the presence of naphthenicgroups was confirmed, lack of reference data did not permit determination of size, shape, or degree of substitution. For thefractions prepared from crude oils, approximately 29 to 66 percent of the terminal carbons are in methyl or ethyl groups at-
Wen Hui Wu is 8 visiting scholar from the Research Institute of Petroleum Processing; Ministry ofPetroleum, P.O. Box 914, Beijing, People's Republic of China.
Energy Sources. Volume 7. Number 30090-8312/84/010275-00$02.0010Copyright © \984 Crane, Russak & Company, Inc.
275
Dow
nloa
ded
by [
Uni
vers
ity o
f Su
ssex
Lib
rary
] at
03:
15 0
1 M
arch
201
3
276 Teh Fu Yen, Wen Hui Wu, and George V. Chilingar
tached to the aromatic centers. NMR results confirmed theconclusion reached previously that mild thermalcracking resultsin the cleaving of chainsand dehydrogenation of the naphthenicstructures to aromatic structures.
IntroductionThe dark-colored, high-molecular-weight, nonhydrocarbon portion of crude oils and related bitumens is referred to as the asphaltic fraction. For a given bitumen the asphaltic fraction isbelieved to consist of structurally similar molecules varying mainlyin molecular weight. In crude oils the asphaltic fraction may represent less than the limit of detection, or as much as 50 percentof the total. Much larger amounts of asphaltic material, particularly of high molecular weight and consequently having poor solubility, are found in the source rocks in which the bitumen isbelieved to have originated. In studying the asphaltic fraction ofcrude oils, it is usually divided into two fractions based on solubility in n-pentane, the soluble portion being known as resin andthe insoluble portion as asphaltene.
The average structure of the molecules making up the asphalticfraction of crude oils and related bitumens has been under studyby the writers, using methods that have proved useful for the elucidation of the structure of polymers and other complex molecules. Densimetric, NMR, and infrared data have indicated acarbon-atom-to-total-ring ratio of about 6 [16]. The X-ray diffraction data have indicated aromaticities ranging from 0.2 to 0.6;diameters of the aromatic sheets, from 8.5 to 15 A; intersheet distances from 3.55 to 3.70 A; stacking of sheets to the extent of 4 or5; and a spacing between saturated structural units of 5.5 to 6.0A [17]. Studies of rates of oxidation with permanganate have indicated that theheteroatoms, 0, N, and S occur in chemicallystable configurations, probably rings [8]. Infrared data indicatedthat naphthenic groups are present as well as aromatic and paraffinic ones. Electron spin resonance (ESR) data has shown thatthe unpaired electrons (free radicals) are located in the aromaticsheets [18]. In conjunction with metalization studies, ESR investigation also showed that the sites for the unpaired electrons as
Dow
nloa
ded
by [
Uni
vers
ity o
f Su
ssex
Lib
rary
] at
03:
15 0
1 M
arch
201
3
Structure oj Petroleum Asphaltenes 277
well as for metal coordination are probably "imperfections" or"holes" in the aromatic sheets caused by the loss or replacementof internal carbons with heteroatoms such as nitrogen Pl.
These studies provided little or no quantitative information concerning the shape of the aromatic sheets, the extent of substitutionof the aromatic sheets and naphthenic groups, and the location ofmethyl groups relative to the condensed ring systems. In order toinvestigate these structural features, as well as to obtain independent values for certain other structural features previously determined, a study was carried out using proton NMR. This methodwas selected because of its capacity to distinguish between hydrogens in different structural environments within a molecule andto provide data for each in terms of the total number of hydrogens.
Williams in 1958 [14] and in 1959 [13] was one of the first touse NMR to study petroleum fractions. As early as 1960, Brownand Ladner [2, 3] studied the low-temperature vacuum carbonization products from coals. Despite difficulties stemming fromthe presence of large amounts of phenolic hydrogen, they wereable to make a separation into three hydrogen types equivalent tothose of Williams. From these they were able to approximate thenumber of substitutable positions on the edge of the aromaticsheets, the degree of substitution, and the aromaticity. Oth et al.[11] examined several solvent extracts of coal and calculated thevalues for the ratio of aromatic to saturated hydrogen. Sebor etal. [5] used NMR spectroscopy to characterize an asphalt-resinfraction. The aromaticity increased from maltenes to asphalt toasphaltenes with increasing fraction molecular weight. High-temperature proton NMR was reported by Yokeno et al. [6] for ethylene tar .pitch. Pyrolysis depolymerization and polycondensationcan be monitored by the high-temperature technique.
Recently, Gillet et al. [9] used NMR spectroscopy to study somecrude oil and petroleum products and proposed some rules forobtaining reliable structural parameters. A method has been suggested to approximate the aliphatic and naphthenic saturated carbons.
Several petroleum-derived materials have been characterizedstructurally by Dickinson [10]. The structures of petroleum resi-
Dow
nloa
ded
by [
Uni
vers
ity o
f Su
ssex
Lib
rary
] at
03:
15 0
1 M
arch
201
3
278 Teh Fu Yen, Wen Hui Wu, and George V. Chilingar
dues were compared by constructing hypothetical average molecules based on these average parameters.
ExperimentalSamples
The preparation of the asphaltene fractions of the crude oils; gilsonite, and Kuwait visbreaker tar is described in a paper by Erdman and Ramsey [8], and of the resin fraction in a paper by Yenet al.[I7].
The reference polymers were prepared in the laboratory. Thepolyacenaphthylene, poly-o-vinylanaphthalene, poly-2,5-dimethylstyrene, poly-2-methyl-4-t-butylstyrene, and poly-2,6-dimethyl-4-t-butylstyrene were obtained from Drs. E. H. Gleasonand E. C. Hurdis, both associated with the Research Laboratoriesof the Koppers Co., Inc. The polybenzyl was. prepared by interfacial polymerization of benzene and formaldehyde. The poly-amethyl-styrene, polyindene, and poly-p-vinyltoluene were prepared by anionic polymerization of the monomers using n-butyllithium.
Instrumentation and Methods
Two Varian NMR spectrometers were used. The early portion ofthe study, involving some of the synthetic polymers and asphaltene substances, was carried out using an A-60 or T-60 NMR,whereas the latter portion was carried out using an HR-IOO NMR.For studies at elevated temperatures, a Varian variable temperature probe and closed loop temperature controller were used inconjunction with the A-60 NMR spectrometer.
Carbon disulfide and deuterated chloroform, and in some casesperdeuterated pyridine, were used as standard solvents. For asphaltenes and resin, the concentration in the solvent was 10 percent. Owing to their limited solubility, the polymers were used atsaturation.
Where resonance bands overlapped, the areas were separatedusing the method of Brown et al. [3]. The areas were measured
Dow
nloa
ded
by [
Uni
vers
ity o
f Su
ssex
Lib
rary
] at
03:
15 0
1 M
arch
201
3
Structure of Petroleum Asphaltenes 279
repeatedly with a planimeter until good reproducibility was attained.
DiscussionThe NMR spectra of petroleum asphaltenes are qualitatively similar. A typical curve is shown in Figure I. The region of strongerresonance between () = 8.0 to 6.5 ppm is assigned to hydrogensattached to aromatic ring systems. The absence of absorption between these regions rules out the presence of hydrogens attachedto carbons of isolated olefinic bonds. In the following discussion,the designation of the hydrogen types will be abbreviated to saturated, aromatic, and so on.
Qualitatively, a spectrum of asphaltics, for example, the Baxterville resin, is very similar to that of a polymer, such as poly (2methyl-4-t-butylstyrene). Figure 2 serves as an illustration, although there are no naphthenic protons in the polymer.
Inasmuch as in proton NMR the area under a peak is directlyproportional to the number of hydrogens contributing to the peak,the fraction of the total number of hydrogens that are saturated,i.e., Hsl H or b« [15], can be obtained from the areas under thecurve in the respective regions of resonance. In this respect, thecalculation is essentially the same as that used previously by theauthors to determine hs from infrared data [15]. The NMR valuesof hs for eight native petroleum asphaltenes, a petroleum resin, agilsonite asphaltene, and an asphaltene prepared from a refinerybottom producer are listed in the. third column of Table 1: .Thevalues, averaging a little above 0.9, are close to those obtained forthe same samples from infrared data. The h, values are shown inthe second column of Table 2.
In order to compare the accuracy of the two methods .of determining hs, nine polymers of known structure with hs values ranging from 0.25 to 0.90 were studied by both NMR and infraredmethods. The true values of hs as determined from the formulas,the values obtained from NMR, and the values indicated by infrared analysis are listed in Table 3 and plotted relative to eachother in Figure 3. The asphaltic materials have been plotted onthe basis of the two methods of determination. The data for the
Dow
nloa
ded
by [
Uni
vers
ity o
f Su
ssex
Lib
rary
] at
03:
15 0
1 M
arch
201
3
AC
ompo
site
Hig
hR
esol
utio
nP
roto
nN
MR
Spec
trum
ofA
spha
lten
eP
repa
red
from
Waf
raN
o:A
-IC
rude
Oil.
Typ
ical
Val
ues
for
the
Che
mic
alSh
ift
ofH
ydro
gen
Typ
esin
Pure
Com
poun
dsA
rePr
ovid
edat
the
Top
ofth
eC
hart
(Ref
eren
ces
I,12
).
~ ~ ::. ~ -. ? § l:l.. ~ ~ "' :-::: Q :::.: ~.~ ::r- ~ ~ .::.
-s,
:r "~ o OD
o ~ s i
H(O=U~5)
R H~M.
TICS
(6=
0.88
"1
1.0
~~~~r~~
o.~:'ji.!A
1~ii~~t
"~~~£~f..
~N~~!::!!N
"A,;
,E~~
G'£
£'~G
~g~
:!.
10
0
IHun
r-~
'i'ri
'~
~~e~
~y~;!'
!,.,
~
-s,~
~~
i~~
%~~i';
5.~
~r.
--
:r u .
3.0
20
0
£ ~.. e< .. ~. -. E~ :;
.--;
.e
:r!!
::!
,ij
Fig
ure
I
v,cp
s. 4.0
01pp
m.
30
0
< s; ~ \:. J' ] :;;
6.0
7.0H
A(0
=72
5)
~~~
i:..
"0
~
.i~;·t
..c:
--..
_:r
.:r.
E
~~
~x
~--
~
...!
'..
II:l:
0e
0
E~~~
..::0
":-
c_
:i:
X0
%0.
IO
.o
Nr:
:Z
Co
-
¥-~ .~
8.0
50
0 ,-
Figu
re1.
Dow
nloa
ded
by [
Uni
vers
ity o
f Su
ssex
Lib
rary
] at
03:
15 0
1 M
arch
201
3
Structure oj Petroleum Asphaltenes
BAXTERVILLE RESIN
281
TMS
.. Side-Bend of TMS •
2
a.o3
7.04
6.0
5
do6
4.0
7
3.0
B
2.0
9
1.0
10
0.0
Figure 2. A Comparison of NMR Spectrum of Poly (2-methyl-4-t-butylstyrene)vs. that of a Baxterville Resin.
Dow
nloa
ded
by [
Uni
vers
ity o
f Su
ssex
Lib
rary
] at
03:
15 0
1 M
arch
201
3
Tab
le1
""V
alu
esF
orS
ixH
ydro
gen
Ty
pes
Det
erm
ined
for
Pe
tro
lcu
mA
sph
nl
te
ne
s,
~A
Pet
role
um
Res
in,
AG
ilso
nit
eA
sph
alt
ene,
anrl
An
Asp
ha
lten
cP
rep
nre
dF
rom
aV
isb
rea
ke
rT
ar
asD
et
erm
ined
byN.
'tRr:
Nu
mb
erS
ampl
eh
sH
A/H
HIH
HN
/HH
R/H
HS>
le/H
Pet
role
um
Asp
ha
lten
es;;;l ::r-
1B
ax
terv
ille
0.9
30
.06
60
.22
O.1
60
.36
0.2
0~
2L
agu
niL
las
0.9
50
.05
50
.27
0.1
80
.31
0.1
93
Bur
gan
0.8
90
.11
40
.23
0.1
70
.32
0.17
~4
Waf
raN
o.A
-I0
.94
0.0
67
0.2
1O
.l6
0.3
8O
.1~
5M
ara
0.9
20
.08
30
.15
0.1
70
.38
0.2
2~
6If
afra
No.
170
.91
0.0
88
0.2
60
.18
0.3
20
.15
7R
aud
hat
ain
0.9
10
.08
60
.23
0.1
70
.36
0.1
5~
8R
agu
sta
0.9
60
.04
00
.18
0.1
90
.36
0.2
2::
Pet
role
um
Res
ins
~9
Ba
xte
rv
ille
0.9
50
.05
60
.27
0.1
70
.37
0.1
4-.
Gil
son
ite
Asp
ha
lten
e.~
10T
abor
vei
n0
.94
0.0
61
0.2
00
.19
0.3
"0
.21
I::l
::R
efin
ery
Asp
ha
lten
el:l.
. oII
Ku
wai
tv
isb
rea
ker
tar
0.8
60
.14
50
.31
0.1
60
.26
0.1
2ti
l0 ~ ti
l ;": Q :::.: S· ~
Dow
nloa
ded
by [
Uni
vers
ity o
f Su
ssex
Lib
rary
] at
03:
15 0
1 M
arch
201
3
enT
able
2~
Ref
eren
ceV
alu
esfO
tP
etro
leu
mA
sph
alt
enes
and
Rel
ate
dS
ub
sta
nce
sA
sD
eter
min
edfr
omE
lem
enta
ryA
na
lysi
s,In
fra
red
An
aly
sis,
and
X-R
ayD
ata
<') .... l:: ~
Sam
ple
h sH
/efa
HM
e/H
SL
a~
(In
frare
d)
(Ele
men
tary
(X-r
ay)
(In
frare
d)
(X-r
ay)
An
aly
sis)
~(1
5)(1
8)(1
7)(I
S)
(171
....P
etro
leu
mA
sPh
alt
enes
a ~B
ax
terv
ille
0.9
2L
OS
0.5
30
.42
13
.0l::
La
gu
nil
las
0.9
21
.13
0.4
10
.43
9.7
~B
urga
n0
.92
1.1
70
.38
0.4
59
.7A
Waf
caN
o.A
-I0
.91
1.1
80
.37
0.4
71
1.0
{;JM
ara
0.9
31
.19
0.3
50
.34
15
.0~
Haf
caN
04
170
.91
1.1
90
.35
0.4
89
.1I:l
Rau
dh
atai
n0
.92
1.1
70
.32
0.4
01
1.0
-....R
ag
us
ta0
.91
1.2
90
.26
0.3
71
2.0
~P
etro
leu
mR
esin
sa
Bax
terv
il1
e0
.97
1.3
20
.22
0.3
18
.5
Gil
san
ite
Asp
hal
ten
e
Tab
or
vein
0.9
81
.42
0.1
40
.41
9.2
Ref
iner
yA
sph
alt
ene
KU
wai
tv
isb
reak
er
tar
0.9
30
.84
0.5
90
.49
8.6
"" e
Dow
nloa
ded
by [
Uni
vers
ity o
f Su
ssex
Lib
rary
] at
03:
15 0
1 M
arch
201
3
Ta
ble
3E
va
lua
tio
nan
dC
omp
aris
ono
fNM
Ran
dIn
fra
red
Met
hod
sfo
rth
eD
eter
min
ati
on
of
the
Pro
po
rtio
no
fH
ydro
gen
sT
hat
are
Sa
tura
ted
~p
olym
erhS
ca
lcu
late
dfr
om
For
mu
la~
Infr
are
d(1
5)
1p
oly
acen
aph
thy
1en
e0
.25
0.2
00
.30
2p
oly
--v
iny
lna
ph
tha
len
e0
.30
0.2
50
.39
3p
o1
yb
enzy
10
.33
0.3
70
.50
4p
o1
y-
-meth
y1
sty
ren
e0
.50
0.5
60
.56
5p
o1
yin
den
e0
.50
0.5
10
.61
6p
oly
-p-v
iny
lto
luen
e0
.60
0.6
40
.69
7p
o1
y-2
,5-d
imeth
y1
sty
ren
e0
.75
0.7
60
.66
8p
o1
y-2
-meth
yl-
4-t
-bu
ty1
sty
ren
e0
.83
0.8
30
.92
9p
o1
y-2
,6-d
imeth
yl-
4-t
-bu
ty1
sty
ren
e0
.90
0.9
00
.83
Sta
nd
ard
dev
iati
on
0.0
30
.09
~ ~ ;:,. ~ ;;,< ,::l ~ ::l ~ -. ,~ § ~ ~ ~ <1
> ;::: Q -. :::- ~.
Dow
nloa
ded
by [
Uni
vers
ity o
f Su
ssex
Lib
rary
] at
03:
15 0
1 M
arch
201
3
Structure of Petroleum Asphaltenes 285
a::
0.8
~-'----'------'a1.0
PolymersG NMR vs. true<:> Infrared vs. true
ASRhaltenes and RelatedSubstances 6
<:>• NMR vs. infrared 6 0
5<:>
404
'5
aI.ar-----r---=,...:----,-.::,--.---"--'r--,----,--..,------.
0.2
a:::2:z
(/)
.c 0.4
Figure3. Comparison of NMR and Infrared Methods for the Determination of. the Proportion of Hydrogens that are Saturated. (The Identifying
Numbers Correspond to those in the First Column of Table 3.)
Dow
nloa
ded
by [
Uni
vers
ity o
f Su
ssex
Lib
rary
] at
03:
15 0
1 M
arch
201
3
286 Teh Fu Yen, Wen Hui Wu, and George V. Chilingar
polymers, as shown in the table and the figure, indicate that NMRvalues of hs are approximately three times as accurate as thoseobtained by using infrared analysis.
In Figure I it is evident that the structure in the saturated andthe aromatic regions, particularly the former, is complex. For reference and comparison, the chemical shift values for the hydrogens of a number of compounds of known structure are listedacross the top of the figure. Williams [13] and Brown and Ladner[2] have investigated the structure of certain petroleum and coalfractions by grouping the hydrogen types according to chemicalshift and-correlating these data with the spectrum for the bitumenfraction. In each case, the saturated region of the spectrum wasdivided into two parts. For the petroleum asphaltenes and relatedbituminous substances under study, the complexity of the saturated hydrogen resonance seemed to justify a more comprehensiveanalysis. Accordingly, the saturated region was broken down intofour subgroups:
h =S
HexH
+ +
The assignments are given in Table 4, whereas the numerical values are·presented in the last four columns of Table 1.
In order to evaluate the above terms, it is assumed that the absorption corresponding to each subgroup is a symmetrical curve,or conversely, that the observed curve between 0 = 3.5 and 0.5may be considered as a composite of four symmetrical curves. Accordingly, the curves recorded at 60 Mc/s for each of the asphalticsubstances resolved into four symmetrical peaks as shown in Figure 1. This step was carried out following a procedure describedby Brown and Ladner, whereby the amplitude is deduced fromthe slope on the high field side of the observed resonance. Thegeometric procedure for resolving the experimental curve into itsfour symmetrical components is considered to be accurate to 5percent or better.
In the following discussion, the values of H N , H R , and HsM,areassumed to be due to the major contributors, i.e., to naphthenic,
Dow
nloa
ded
by [
Uni
vers
ity o
f Su
ssex
Lib
rary
] at
03:
15 0
1 M
arch
201
3
Han
geB
and
Cen
ter
Sy
mb
ol
In
ppm
10
pp
mM
ajo
r
Tab
le4
Str
uct
ura
lA
ssig
nm
ents
for
the
Su
bgr
oup
so
fth
eS
atu
rate
dH
ydro
gen
Res
onan
ce
Ha
HN HR liSM
e
2.9
0-i
.90
i.8
0-1
.40
i,3
5-i
.00
0.9
2-0
.80
2.4
0
1.
58
1.2
S
0.8
8
Ass
ign
men
tsB
a-m
eth
yl
a-m
eth
yle
ne
para
ffin
ica
-met
hin
e
met
hy
len
em
eth
ine
na
ph
then
iccris
perip
hera
lb
rid
ge
met
hin
e
met
hy
len
eo
ther
.th
ana
or
fj-p
ara
ffin
ic
met
hy
lo
ther
ths
na
Or~-paraffiniC
Min
or
~-methytene-paraffinic
Q-m
eth
yle
ne-
pa
raff
inic
tra
nsp
erip
her
albrid~e
met
hin
e-0
8p
bth
enic
tran
speri
ph
era
lb
rid
ge
met
hin
e-n
aph
then
ic
~ ~ ~ ~ ~ a ~ ;: s :t.. {3 f !Jl
BT
hele
tters
a~nd
Pin
dic
ate
po
siti
on
srela
tiv
eto
ana
rom
ati
cri
ng
.
~ 'I
Dow
nloa
ded
by [
Uni
vers
ity o
f Su
ssex
Lib
rary
] at
03:
15 0
1 M
arch
201
3
288 Teh Fu Yen, Wen Hui Wu, and George V. Chilingar
paraffinic methylene other than those 0 or fJ to an aromatic ring,and paraffinic methyl similarly other than 0 or fJ, respectively. Theprincipal departure-from this assumption comes from the fusedring naphthenic contributions to H R and H SMe' Whereas data onsuch compounds are as yet meager, the spectra for the decalins(II indicate that when bridge hydrogens are in the cis-configuration, the entire hydrogen resonance concentrates essentially in asingle peak at 0 = 1.41, i.e., within the range assigned to H N •
When the bridge hydrogens are in the trans-configuration, however, about one third of the resonance is spread out to lower 0values concentrating mainly in the range assigned to Hs Me ' No dataare known to exist for more highly condensed naphthenic ring systems, particularly where internal as well as peripheral bridge methine groups are present. Thus, it has been assumed that thenaphthenic contribution to H SMe and H R will be no more than thatof an equivalent 1:1 mixture of cis- and transdecalin, i.e., H N maybe low by as much as 1/6 H N (or 18 percent) and H SMe may behigh by as much as 1/6 H N that, for the asphaltic materials understudy, would range from 13 percent to 20 percent. The effect onH R also would be such as to make the values high, but only by afew percent.
In the experiment described below, a sample of known composition was examined and its spectrum correlated with theamounts of the several hydrogen types present, namely, H A , H a ,
HN , H R , and H SMe ' There is considerable difficulty in locating asample of fused-ring naphthenics; a natural hormone has servedthis purpose. Inasmuch as H; is required for this model, smallamounts of H A are also necessarily introduced. Thus, the modelmixture selected has to contain aromatic (HA ) , methyl and methylene substitution (Ha ) , naphthenic (HN ) , paraffinic methylene(HR ) , and paraffinic methyl substitution (HSMe) groups. The system selected is a mixture of cholestane (I), o-tocopherol (II), andpentamethylbenzene (III) (Figure 4). Experimentally, in order toobtain a mixture whose hydrogen types are close to those of thepetroleum asphaltics, a molar ratio of 5:10:3 for 1:11:111 was used.For this particular experiment, the weights of I, II, and III were0.166 g (0.3 mole), 0.386 g (0.6 mole), and 0.0399 g (0.18 mole),respectively. The detailed calculation was tabulated in Table 5,
Dow
nloa
ded
by [
Uni
vers
ity o
f Su
ssex
Lib
rary
] at
03:
15 0
1 M
arch
201
3
Structure of Petroleum Asphaltenes
Me
MeMe
289
¢Me Me
lUI) . s-: IMe::::-" Me
Me
Me
Figure 4. A Blend of Model Compound for the Simulation of Petroleum Asphaltene, I, Cholestane, II, a-Tocopherol, III, Pentamethylbenzene.
from which the actual hydrogen-type ratios of the blend can beobtained.
The reproducibility of the area measurements was within 5 percent. There is fair agreement between experimental and true valuesfor HA> H sM e, and H.; however, the HOi and H N agreement wasonly qualitative. The true value is also subject to certain experimental uncertainties, such as weighing, evaporation, and so on,but these are thought to be minor. Purity, however, is not knownand may introduce an appreciable error.
From the area under the component peaks, it is possible to estimate values for H,", HI> i.e., the average number of carbon atoms directly attached to an aromatic sheet divided by the numberof hydrogens that would be attached if the aromatic sheet weretotally unsubstituted. Expressed in another manner, C,", HI ratiois the degree of substitution of the aromatic sheets. The derivationis based on the assumption that for a unit of structure, C," is approximately equal to one half of the number of o-hydrogens, HOi,i.e.,
Dow
nloa
ded
by [
Uni
vers
ity o
f Su
ssex
Lib
rary
] at
03:
15 0
1 M
arch
201
3
Tab
le5
Foun
dan
dT
rue
Val
ues
of
Hyd
roge
nT
ypes
of
aP
hy
sical
Ble
nd
.I
IIII
IB
len
dFo
und
Mol
esin
ble
nd
0.3
0.6
0.1
81
.08
Mol
arR
atio
510
3H
IH--
-0
.13
0.1
70
.30
0.2
4H
N/H
0.1
60
.02
---
0.1
80
.14
HR
/H0
.05
0.2
9--
-0
.34
0.3
6H
SMe/
H0
.09
0.1
4--
-0
.23
0.2
5H
A/H
---
---0
.01
0.0
10
.01
~ ;;;l ::- ;p ~ ::s ~ ::s ~ -. ~ § ~ ~ oa "' ~ 9 -, :::- i'
Dow
nloa
ded
by [
Uni
vers
ity o
f Su
ssex
Lib
rary
] at
03:
15 0
1 M
arch
201
3
Structure of Petroleum Asphaltenes
C =(l/2)Ha,su
andHI = H +(1/2)HA a
291
Hence:Csu»: =
The C,u IH, values obtained for the native asphaltenes and relatedsubstances are provided in the second column of Table 6. For thenative asphaltenes the degree of substitution on the edge of thearomatic sheets range from a low of 0.48 to a high of 0.70. Theresin cut of the Baxterville crude oil shows a higher degree of substitution than does the corresponding asphaltene, i.e., 0.70 versus0.62. The degree of substitution for the asphaltene fraction of gilsonite centers in the range of the native petroleum asphaltenes,which provides another indication of the close structural similarityof gilsonite and the asphaltic fraction of crude oils. The value forthe asphaltene from Kuwait visbreaker tar is about the same asfor the Burgan (Kuwait) crude oil, indicating that light crackinghas not affected the degree of substitution on the edge of the aromatic sheets.
Data from NMR also permit the estimation of the Csi C,u ratio,the average number of carbon atoms attached to a position on theedge of an aromatic sheet. Alternately defined, it is the total number of saturated carbon atoms, Cs, divided by the number of carbon atoms attached to the edge of the aromatic sheet:
HR
+ EN HSMe
Cs = Csu + 2 + -3-
= 1 +(Hn -tHN) /2 + H"&..'Je/3
H /2a
Dow
nloa
ded
by [
Uni
vers
ity o
f Su
ssex
Lib
rary
] at
03:
15 0
1 M
arch
201
3
Tab
le6
Va
lues
for
Str
uctu
ra
lP
ara
met
ers
of
Pet
role
um
Asp
ha
lten
esan
dR
ela
ted
Su
bst
an
ces
Ca
lcu
late
dfr
omNH
RD
ata
!e
Sam
ple
Cs
u/1I
1C
S/C
su
III/
CA
IIN/C
NC
SMe/
CM
e~
Pet
role
um
Asp
ha
lten
es;:0
-
~B
ax
terv
ille
0.6
24
.00
.34
---
0.>
0L
ag
un
illa
s0
.71
3.2
OJ,3
l.25
0.4
7~
8u
rga
n0
.51
3.6
0.5
20
.97
0.4
22'-
Wah
.N
o.A
-l0
.61
4.1
0.4
21
.09
0.4
2M
ara
0.4
85
.60
.38
0.9
80
.71
~W
afr.
No.
170
.60
3.4
0.5
30
.98
0.3
4R
aud
hat
ain
0.5
73
.70
.48
0.8
20
.42
::sR
ag
us
ta0
.70
4.8
0.4
01
.01
0.6
3
~P
etro
leu
mR
esin
s-.
Ba
xte
rvil
le0
.70
3.4
0.6
10
.81
0.4
8?
Gil
son
ite
Asp
ha
lten
es::. ::s
Tab
orv
ein
0.6
24
.30
.60
0.8
10
.55
I:l..
Ref
iner
yA
sph
alte
ne
~ CK
uwai
tv
isb
rea
ker
tar
0.5
22
.60
.39
1.1
40
.28
.~ ~ Q :::
; -. ~ ~
Dow
nloa
ded
by [
Uni
vers
ity o
f Su
ssex
Lib
rary
] at
03:
15 0
1 M
arch
201
3
Structure of Petroleum Asphaltenes
and
293
~SCsu
=hS - HSWe/3H
H /H.- Ct
As shown in Table I, the values of hs are approximately four anda half times those of H SMe ' and in the above equation HSMe is divided by three. The effect on Csi Csu of naphthenic contributionto H S M " therefore, can be at most only a few percent. In the aboveequations, however, the assumption is made that there are twohydrogens (HR, H N, or H a ) attached to each methylene carbon,i.e., that the groups attached to the aromatic sheet are straightchains. Inasmuch as branching at the a carbons would decreasethe denominator for H; below 2, the value of Csi Csu as given abovewould be high. Branching, including naphthenic structure not involving a carbons, would decrease the denominators for H R andH Nbelow 2; hence, the above value for CsIC," would be low. Theterm "branching" is used here to include naphthenic structure.For the time being, it is assumed that the effects on the value ofCslCsu induced by a and non-a branching will tend to cancel.
The values of Csi Csu ratio for the native petroleum asphaltenesand related substances are listed in the third column of Table 6.For the native asphaltenes and the resin, the values range from3.4 to 5.6, with the gilsonite asphaltene again falling well withinthis range. Comparison of the Kuwait visbreaker tar with the Burgan (Kuwait) crude oil shows a marked decrease in Csi Csu. Referring again to Table I, it will be seen that there is a decrease inHRIHand H SM' IH and little change in HNIH. In an earlier paper[171, Yen et al. have shown that the aromaticity,f.. i.e., the fraction of carbon atoms that are aromatic, is higher for the asphaltene from the visbreaker tar than for the corresponding Burgancrude oil. Taking into account the constancy of Csut H, ratio andthe fact that H decreases with increasing fa. it would appear thatlight cracking results mainly in the breaking of chains attached toaromatic rings. Further, inasmuch as HNIH does not increase, itis likely that dehydrogenation is taking place concurrently with thecleavage of paraffinic chains.
Dow
nloa
ded
by [
Uni
vers
ity o
f Su
ssex
Lib
rary
] at
03:
15 0
1 M
arch
201
3
294 Teh Fu Yen, Wen Hui Wu, and George V. Chilingar
The ratio of carbon to hydrogen of the saturated portion ofsample, fs/ hs, increases with increasing saturated chain length,C,I C.u , and with decreasing arornaticity, fa. As the carbon-to-hydrogen ratio of the saturated portion increases, the ratio of carbonto hydrogen of total sample decreases (Figure 5).
Thus far, very little has been done to determine the shape ofthe condensed aromatic sheets in petroleum asphaltenes, althoughsome research of this type has been done for coals by Brown andLadner (2). It is evident that the value of the ratio of H, to thenumber of aromatic carbon atoms, CM will, for a sheet with agiven long dimension, increase as the degree of condensation orsymmetry decreases. This ratio was determined from NMR dataand from the carbon/hydrogen ratio in the following manner:
_ _C_. CA Ra/H H.NIH. BRIE. HStvle lE= -8--- + 2 + ---2--- + 2 + 3
8.
H
and
2H1=HalH + 2HAIH
H1 H. 18a
Values for HIC are provided in the third column of Table 2.Again, it can be observed that naphthenic contribution to HS Me
would lead to slightly low values of H/CA • For the asphaltic materials under study, the error probably would not exceed 2 percent.
In order to derive the equation above, the assumption shouldbe made that there are two hydrogens attached to each carbon thatis a to aromatic ring or to each naphthenic carbon, i.e., HOI/ COl= 2, and HN/CN = 2. Figure 6 is the plot of two sets of H,/CA ,
which were calculated for different HOI/Ca and HN/CN values.The results signify that different values of HOI/COl and HN/CNdonot affect H,/CA value much. (See Figure 7.)
Dow
nloa
ded
by [
Uni
vers
ity o
f Su
ssex
Lib
rary
] at
03:
15 0
1 M
arch
201
3
f ~ ~ ~ a iii s:: s ~ ~ f2.
0I 1.8
I 1.6
I 1.4
I 1.2
I 1.0
HIe
I0
.8I
0.6
I0
.4
c:,10//
09
,/'"
~1~
.6)"
.
/°
24
Q//
0'
//
//
//
01 /
Io
0.2
1.0,
"I
0.8
1-
0.61
---
0.21
---
~ h S(N
MR
) 0.41
---
Figu
re5.
Var
iatio
no
fC
,IH
,O
ver
the
Ran
geof
Hie
ofA
spha
lten
esas
Stud
ied
byN
MR
(h.)
and
X·R
ay(f
J.'" ~
Dow
nloa
ded
by [
Uni
vers
ity o
f Su
ssex
Lib
rary
] at
03:
15 0
1 M
arch
201
3
296 Teh Fu Yen, Wen Hui Wu, and George V. Chilingar
~"u/ U::E ::Ii
II> II>IU
NII
a:1 noIU
II
z\ 2-IU 0.4If)
NII
"'I tlIU
-------HI ..IU
Figure 6. Effect of Variation of HOt/COt and HN/CNvalues on H,IC, ratio.
Dow
nloa
ded
by [
Uni
vers
ity o
f Su
ssex
Lib
rary
] at
03:
15 0
1 M
arch
201
3
1.0
1;1) -., ;: ...,
I-/
-;: ~
0.8
1-
/~
/~ -
II/
., <:)
'iI~
0.6
1-
4E"
;: s5.
6~32
;.
(Ha
HN
HR
HS
Me)
H~
r-7
'iI'iI
'iI1--+
-+
-+
---
fa(N
MR
)
.:jf!/
82H
2H2H
3HC
~
(H
aH
N'
HR
HSM
.)H
~
0.4
1-
10l::
l'iI
.:.
•::;:
-'iI
01
-2
.5H
+H
+2
H+
3HC
""0
°0
..
:::c
I-
0(H
aH
NH
RH
sMe)
Ha
8~
00
1-
2H+
H+
2H+
3'H
C0
.21
-/.
(h
s)
H•
1-/
•1-
1.65
c:-O
l/I
II
II
II
II
Ia
0.2
0.4
0.6
0.8
i.Ofa
(X-r
oy)
Figu
re7.
Eff
ect
ofV
aria
tions
ofD
iffe
rent
Ass
umed
Coe
ffic
ient
son
thef.
de
term
ined
byN
MR
Met
hod
and
X-r
ayA
naly
sis.
~ '1
Dow
nloa
ded
by [
Uni
vers
ity o
f Su
ssex
Lib
rary
] at
03:
15 0
1 M
arch
201
3
298 Teh Fu Yen, Wen Hui Wu, and George V. Chilingar
The calculated values of H,I CA are listed in the fourth columnof Table 6. In Figure 8, these values are plotted against the reciprocal of the layer diameter, L., calculated from X-ray data by themethod of Diamond and reported by Yen et al. [17J. The referencelines represent the values calculated for the two extremes of symmetry: a circular sheet and a linear katacondensed band. Theequations for these lines were derived as follows:
Circular
HI = (6CA)1/2
Linear BandCA
HI =~ + 3
'« =43r( CA - .z. )4 2
hence4
for an infinite hexagonal net (16)
t. 33/2. r, ..a.~cA--:1=-/=2- = -----:;--- = 1. 62
hence
HI=c;: 3.95 t: -1
·a
where r is the bond distance of the hexagons (r = 1.42..\). It canbe seen from Figure 8 that for the asphaltic materials under study,the average probable symmetry varies, but in no case approachesthat of a linear katacondensed band.
Referring to Table 2, it can be seen that the asphaltenes prepared from Lagunillas and Burgan crude oils both have the sameL. value but, according to Figure 9, vary considerably in symmetry. In order to get a better picture of the average. aromaticsheet in each of these asphaltenes, the H,I CA values were calculated for all possible ring arrangements that would give the closestlong axis, or L., value. The results are shown in Figure 9, whereit is evident that the average aromatic sheet for the Lagunillas asphaltene corresponds closely to ovalene (m), while the average
Dow
nloa
ded
by [
Uni
vers
ity o
f Su
ssex
Lib
rary
] at
03:
15 0
1 M
arch
201
3
Structure of Petroleum Asphaltenes 299
1.0r--------------------~
o -Native Petraleum AsphalteneYJ - Refinary Asphalteneci-Gilsanite Asphaltene[J - Native Petroleum Resin• - Typicol Aromotic HydrocarbonI I
0.10 0.15 0.2CL -I
o
I0.05
0.2-
------~ ...---..-- -- TerpMnyl _____Links, __ ., ...-
07 t- -- >- Quint:uiot.enyl ~·5 Anthracene /'
• -- -- ~.'II Coeranthen! /~ r Tetracene . ;
PIUr.e .y 9 • u Tnangulene
061- kala -Band ........--""- 10~h [J em ./,... Pyrene
';' I HEnIP~r"flfnE Rubicene Benlfluoranlhene. Perytene..........- • Ttr)'lene z 6-... "
,.........- '.J'Clanthfenf e... 9 .0· 30 \:) JAntr,anlhrene
A 5~ QU2rt.:~yleni' .Anlhro\tlanClUlene. c /' .0• 0 70 k Coronene
Tt'rpher.yllfrylene n 4 ...d ~nthrodianthrene. e:::> v ovatene
ClftUfficintl1rene e~8 2......50' .~ IIYJ
Pentccinexaene b 1 tlisletranrenelOy:" Dinapt",tho avalene
0.3 - . C' I /' Oianthrocircumanlhreneperi- IrCU or
Disk"", .:
»:/
0.11- /
V/°O~-----;::-:~-----;:::-~-----;:::-~-----;::;-;'
0.8f-
0.4 -
0.91-
Figure 8. Estimation of Degree of Ring Condensation of the Aromatic Sheets.The Lines Represent the Theoretical Extremes for Fused Ring Systems.(The open points represent the values for the asphaltic fractions. Thesolid circles represent typical PNA.)
Dow
nloa
ded
by [
Uni
vers
ity o
f Su
ssex
Lib
rary
] at
03:
15 0
1 M
arch
201
3
300 Teh Fu Yen, Wen Hui Wu, and George V. Chilingar
I~
OAJ ------DXir-LOgunillaS 1m!
04at-_ @........ '<,
I
(nl
Figure 9. Comparison of the H,/CA Values of Lagunillas and Burgan Asphaltenes with Those for Condensed Aromatic Ring Systems with an Approximately Equal Long Axis or L. Value.
Dow
nloa
ded
by [
Uni
vers
ity o
f Su
ssex
Lib
rary
] at
03:
15 0
1 M
arch
201
3
Structure of Petroleum Asphaltenes 301
value for the Burgan asphaltene falls close to anthro- 3' , 5/, I,10 -tetracenyl (k) or 2,3-benznaphtho- 2/, 7/, 1,8 anthryl [I].Whereas it is recognized that these correlations represent the average structure, it should be pointed out that, for the Lagunillasasphaltene at least, it would take nine naphtho- 2/ , 7', 1, 13 ovalene (n) nuclei to permit the existence of one tetracene band(a). These findings seem to confirm earlier conclusions [15,17] thatthe aromatic sheets of petroleum asphaltenes approach the shapeof discs in contrast to coals, where the aromatic structure is believed to exist largely in the bent or branched katacondensedbands.
Turning now to the naphthenic portion of the molecule, datafrom NMR and X-ray permit an estimation .of HN/CN, i.e., theratio of the number of naphthenic hydrogens to naphthenic carbons. One assumption, however, is necessary, i.e., that on the average each paraffin carbon is linked to two hydrogens, i.e., Cpl H;= 112. Accordingly,
R'N H Rp .JLHS H-- -.-+-..- - H.H C H C C
Hp .Cp HN HCN (s
-- -+ f{ =r;C Hp C fIN
C tc = 1 fa ' and-s .
HN!tNIH
= t;I!:{ 1 - H- HS/2H + NIHwe 2 - hs
Values for fa as determined by X-ray analysis are provided in thefourth column of Table 2.
The values calculated for the HNI CN ratio using the above equation are listed in the fifth column of Table 6. Structural iriterpretation of H N I CN is made difficult at this time because, as pointedout previously, there is uncertainty concerning the effect of bothperipheral and internal bridge hydrogens on the NMR spectrum
Dow
nloa
ded
by [
Uni
vers
ity o
f Su
ssex
Lib
rary
] at
03:
15 0
1 M
arch
201
3
302 Teh Fu Yen, Wen Hui Wu, and George V. Chilingar
of condensed polynuclear naphthenes. If it is assumed that theaverage naphthenic group contains the same number and arrangement of rings as exist in the aromatic sheets and that H N representsonly naphthenic methylene hydrogens, then the extent of substitution could be estimated by comparing the observed HNI CN
values with those for methylene of the unsubstituted naphthenicgroups, which would be given by 2H,I CA. Comparison of the values given in Table 6 shows that in a number of instances HNI CN
> 2H,I CA. Consequently, either the assumption that the naphthenic and aromatic groups are comparable in number and arrangement of rings or that H N can be assumed to represent onlythe methylenic hydrogens of a condensed naphthenic group is incorrect. Studies involving the progressive dehydrogenation of asphaltenes, as well as a better understanding of NMR spectra ofcondensed naphthenic molecules, should permit verification of oneor rejection of both of the above assumptions and permit a clearcut structural interpretation.
The last feature to be investigated by NMR was CSMJCM" theratio of the total number of methyl groups that are neither Ci nor{3 to an aromatic ring to the total number of methyl groups. Thisratio can be evaluated quite easily by means of the following equation:
=
where HM,lHs is determined from infrared absorption data [15].Values for HM,IH s are provided in the fifth column of Table 2,whereas those for CSM'I CM, are presented in the sixth column ofTable 6. It is evident that the full effect of transnaphthenic contribution to H S M ' will be reflected in the values of CSM'I CM, ratio;hence, the values of the latter may be high by some 15 or 20 percent. The values listed in Table 6 for the native petroleum asphaltenes, the petroleum resin, and the gilsonite asphaltene suggestthat from 1/3 to more than 2/3 of the methyl groups are moreremote than {3 to the aromatic sheets. The low value of 0.28 forthe asphaltene from the Kuwait visbreaker tar compared to 0.42for the asphaltene fraction of the equivalent Burgan crude oil again
Dow
nloa
ded
by [
Uni
vers
ity o
f Su
ssex
Lib
rary
] at
03:
15 0
1 M
arch
201
3
Structure of Petroleum Asphaltenes 303
suggests that mild cracking results in breaking off from the aromatic sheets of relatively long chains and/or the dehydrogenationof naphthenic groups to which methyls or ethyls are attached.
As further insight into the assignment to the HN band, the NMRspectra of a large number of perhydropolynuclear aromatic hydrocarbons have been obtained. For most pericondensed naphthenic systems (even for pure compounds) the signal wandersthrough a wide range in the space of NMR spectra. This is due tospin-spin interaction of protons in a mixture of conformers. Naphthenics whose signal concentrates in the {) = 1.8 - 1.4 ppmrange are those of kata type, especially the stearane type. Inasmuch as a number of terpenoids and related compounds have beenfound in petroleum, this may imply that the naphthenic structurein asphaltic fraction is perhaps similar to the structure of stearaneduring the course of diagenesis.
Finally, it is always dangerous to propose the application ofNMR analysis for structural elucidation of such complex classesof substance as petroleum asphaltics. For example, if a proposedhypothetic structure is adopted, at least most of the structure parameters, such as cun; cue: HI/CA , HN/CN, CSMe/CMe, f ..and so on, have to be very close. In many cases, more than onetool should be used to derive such a structure.
Acknowledgment
A portion of this paper originally appeared in the preprint of theAmerican Chemical Society, 7(3): 99, 1962, issued by the Divisionof Petroleum Chemistry. The original version has completelymodified into the present form and some errors have been corrected.
References
I. American Petroleum Institute Research Project 44. 1978. Selected NuclearMagnetic Resonance Spectral Data. Serial No. 14 and 15.
2. Brown, J. K., and Ladner, W. R. 1960. Hydrogen distribution in coal-likematerial by high-resolution nuclear magnetic resonance spectroscopy, II. Acomparison with infrared measurement and the conversion to carbon structure. Fuel, 39: 87-%.
Dow
nloa
ded
by [
Uni
vers
ity o
f Su
ssex
Lib
rary
] at
03:
15 0
1 M
arch
201
3
304 Teh Fu Yen, Wen Hui Wu, and George V. Chilingar
3. Brown, J. K., Ladner, W. R., and Sheppard, N. 1960. Hydrogen distribution in coal-like material by high-resolution nuclear magnetic resonancespectroscopy, I. The measurement and interpretation of the spectra. Fuel,39: 79-86.
. 4. Bhacca, N. S., Johnson, L. F., and Shoolery, J. N. 1962. NMR SpectraCatalog. Varian Associates.
5. Seber, G., Lang, I., Hajek, M., Weiser, 0., and Mosteeky, J. 1977. ProtonNMR structure analysis of asphaltene-resin petroleum fractions. Chem.Prum., 27(9): 455-459.
6. Yokono, J., Miyazawa, K., Sanada, Y., Yamada, E., and Shimokawa, S.1979. High temperature 'H NMR of ethylene-tar pitch. Fuel, 58(3): 237-238.
7. Erdman, J. G., and Harju, P. H. 1963. Capacity of petroleum asphaltenesto complex heavy metals. .[. Chem. Eng. Data, 8: 252-258.
8. Erdman, J. G., and Ramsey, V. G. 1961. Rates of oxidation of petroleumasphaltenes and other bitumens by alkaline perrnanganate. Geochim. et Cosmochim. Acta, 25(3): 175-188.
9. Gillet, S., Rubini, P., Delpuech, Jean-Jacques, Escalier, Jean-Claude, andValentin; P. 1981. Quantitative carbon-13 and proton nuclear magnetic resonance spectroscopy of crude oil and petroleum products, I. Some rules forobtaining a set 'of reliable structure parameters. Fuel, 60(3): 221-225.
10. Dickinson, E. M. 1980. Structure comparison of petroleum fraction usingproton and »c NMR spectroscopy. Fuel, 59(5): 290-294.
II. Oth, J. F. M., De Ruiter, E., and Tschamler, H. 1961. Hydrogen in coaland coal extracts, I. The ratio of aromatic to aliphatic hydrogen from proton-spin resonance and infrared measurements. Brennstaff-Chem., 42:378-380.
12. Sadder Research Laboratories, Inc. 1978. The Sadtler Handbook of ProtonNMR Spectra.
13. Williams, R. B. 1959. Nuclear Magnetic Resonance. Spectrochem. Acta, 14:24-44.
14. Williams, R. B. 1958. Characterization of hydrocarbon in petroleum by nuclear magnetic resonance spectrometry. A STM Spectral Technical Publication No. 224: 168-194.
15. Yen, T. F., and Erdman, J. G. 1962. Investigation of the structure of petroleum asphaltenes and related substance by infrared analysis. Arn. Chem.Soc., Div. Pet. Chem., Preprints 7(1): 5-18.
16. Yen, T. F., Erdman, J. G., and Hanson, W. E. 1961. Reinvestigation ofdensimeter methods of ring analysis. J. Chern. Eng. Data, 6: 443-448.
17. Yen, T. F., Erdman, J. G., and Pollack, S. S. 1961. Investigation of thestructure of petroleum asphaltenes by X-ray diffraction. Anal. Chem., 33:1587-1594.
18. Yen, T. F., Erdman, J. G., and Saraceno, A. J. 1962. Investigation of thenature of free radicals in petroleum asphaltenes and related substances byelectron spin resonance. Anal. Chem., 34: 694-700.
Dow
nloa
ded
by [
Uni
vers
ity o
f Su
ssex
Lib
rary
] at
03:
15 0
1 M
arch
201
3