S IG U R D G R O E N N IN G S , Shell Development Company, Emeryville, Calif.
in determining the aromatic content of cracked gasolines by specific dispersion, correction must be made for the specific dispersion of the olefins. This correction has been calculated from available litera
ture data and found to vary considerably with the class of olefins, with the structure of olefins of the same class, and with the boiling point. A n estimate of the reliability of the recommended correc
tion factors is presented; also, as far as possible, the accuracy of the specific dispersion method has been determined experimentally.
T
H IS m ethod, developed by Grosse and W ackher (1), is based on th e fact th a t th e specific dispersion of the aro- m atics is appreciably higher th a n th a t of th e satu rate s (naph- thenes an d paraffins) which is nearly constant. Hence, the arom atic co n ten t of a gasoline m ay be determ ined by m easuring its specific dispersion increm ent over th a t of th e satu rate s. (This presupposes t h a t the specific dispersion of th e ty p e of arom atics present m ay be estim ated w ith satisfactory accuracy, which is particularly th e case w ith lower an d middle boiling fractions containing single arom atics—viz., benzene or toluene.)
T h e specific dispersion of th e olefins is also higher th a n th a t of th e saturates, and when present, as in cracked gasoline stock, olefins will cause a specific dispersion increm ent which, if not corrected for, will give too high a value for th e arom atic con
ten t.
T he arom atic content m ay be expressed by the following equation:
A rom atics, % w =
p S , - 98 - / , X B r No. - / , X M .A .V .l , 00 + c
L 8 a — 98 J
where
%w = per cent b y weight
S , = specific dispersion, n r , Uc X 104 of the sam ple a t 20° C.
S„ = specific dispersion of the arom atics present
98 = estim ated average specific dispersion of th e saturates.
[For greater accuracy, th e determ ined specific dis
persion of th e satu rate s as obtained by silica gel tre a tm e n t is used instead of 98. An a d ap ta tio n of the silica gel tre a tm e n t as described by M air and F orziati (2) is used.]
/ i = facto r correcting for th e specific dispersion of mono
olefins an d nonconjugated noncyclic diolefins present
fz = facto r correcting for th e specific dispersion of con
ju g ated diolefins present
C — correction for th e deviation from lin earity of th e re
lation of specific dispersion to arom atic content M .A .V . = maleic anhydride value
T h e brom ine num ber is expressed as gram s of brom ine con
sum ed by 100 gram s of sam ple, an d th e m aleic anhydride value as m illigrams of maleic anhydride consum ed b y 1 gram of sam ple.
T he necessity of a good evaluation of th e olefin correction m ay be illu strated by th e following example: If in calculating th e aro
m atic content of a toluene fraction containing 25% monoolefins, one uses a correction factor w hich is in error b y 15% of its tru e value, th e resulting error in th e arom atic content can be shown to am o u n t to 1 % , which is already th e expected accuracy of th e m ethod as applied to olefin-free m aterial. T his error increases directly w ith th e olefin content.
C alculation of th e correction for th e sim plest an d probably m ost comm only occurring olefins from available d a ta is outlined below.
C A L C U L A T I O N O F O L E F I N C O R R E C T I O N
T he specific dispersion increm ent due to th e presence of olefins is very nearly a linear function of th e olefin content; for m od
erately high olefin contents th e deviation m ay be considered negligible in view of th e accuracy of th e m ethod. Since the olefin content is m easured b y th e am o u n t of brom ine absorbed, th e increm ent is directly proportional to th e theoretical brom ine num ber. Hence, th e correction factor to be applied equals th e specific dispersion increm ent of olefins per u n it brom ine num ber an d is obtained1 by dividing th e increm ent by th e theoretical brom ine num ber.
Sp e c i f i c Di s p e r s i o n. Since th e in terest in specific dispersion of pure hydrocarbons is relatively recent, only a lim ited am ount of d a ta m ay be found in the literatu re. T he m ost comprehen
sive collection is probably contained in Grosse and W ackher’s publication (I ). Therefore, th e ir d a ta have been em ployed in th e present calculations, b u t augm ented an d in p a rt supple
m ented by d a ta accum ulated in these laboratories as selected best values from a critical lite ra tu re review.
T he m agnitude of th e specific dispersion of the m ain hydro
carbon groups—viz., satu rates (naphthenes and paraffins),
aro-&
*O
aU.
i
4 0 6 0 8 0 100 120 140 160 180 2 0 0 2 2 0 BOILING POIN T, *C.
Figure 1. Specific Dispersion of Aromatics, Olefins, and Saturates vs. Boiling Point
361
362 I N D U S T R I A L A N D E N G I N E E R I N G C H E M I S T R Y Vol. 17, No. 6 m atics, and olefins—is shown graphically in Figure 1 where they
have been plotted against boiling point (d a ta for olefins from Tables I an d II ).
T he average values for naphthenes and for paraffins cannot be distinguished from one another. In general, th e specific disper
sion values for these tw o groups lie betw een 97 and 99, varying som ew hat w ith th e extent and mode of branching and w ith the boiling point. Any tre n d in this respect has n o t y e t been definitely ascertained, owing to the low accuracy w ith which m ost of these nffeasurements have been carried out in th e past.
T he arom atics in gasoline are alm ost all monocyclic (benzenes), the specific dispersion tren d of which is approxim ately as indi
cated in th e graph. (Since bicyclic arom atics have m uch higher specific dispersions, higher boiling fractions containing both mono- and bicyclic arom atics cannot be analyzed by th e present m ethod. Though th e lowest boiling bicyclic arom atic, n a p h th a lene, boils a t 218°, traces have been observed in m aterial boiling as low as 180° C., due to azeotropism. This introduces an error in
Table I. Specific Dispersion Increment, i, per Unit Theoretical Bromine Number of Monoolefins of the Gasoline Range error m ay be negligible in analysis of th e full range gasoline.)
T he specific dispersion of th e olefins is interm ediate between those of th e satu rate s and th e arom atics; th e conjugated diole
fins m ake an exception, as th eir specific dispersion is higher than th a t of th e arom atics. T here is a distinct difference betw een the line, which is impossible. (The exceptional direction of the curve for cyclic conjugated diolefins m ay be explained by th e combined exalting effect of cyclization an d olefinic double bonds.)
Th e o r e t i c a l Br o m i n e Nu m b e r. By the aid of theoretical th e specific dispersion increm ent due to monoolefins is expressed
June, 1945 A N A L Y T I C A L E D I T I O N 363 where num erous types of olefins m ay be present, th e portion of th e to tal olefin increm ent due to conjugated diolefins can be estab
lished b y a direct determ ination of the conjugated diolefin con
te n t in term s of the am ount of m aleic anhydride consumed.
(There are conjugated dienes which do n o t react w ith m aleic
an-Table II. Specific Dispersion Increment, r, per Unit Theoretical Bromine Number and Correction Factor, f 2for Diolefins of the Gasoline Range
& 1 CO 00
Table III. Olefin Correction Factors for Determination of Aromatic Content of Gasoline Fractions by Specific Dispersion
C o n ju g a te d D io le fin F a c to r fe ,
anhydride value is i = However, since the conjugated
1V1 .A . V .
diolefins also brom inate, this increm ent has already been p artly accounted for as mono-olefin increm ent, /i, an d th e la tte r, ex
pressed in term s of maleic anhydride value, is
f i =
Br No. (of conjugated diolefins) M .A .V .
BOILING POIN T, *C.
T he theoretical correction factor for conjugated diolefins is therefore
364 I N D U S T R I A L A N D E N G I N E E R I N G C H E M I S T R Y Vol. 17, No. 6
4 0 6 0 8 0 100 120 140 160 ISO 2 0 0 2 2 0 BOILIN G POINT, *C
Figure 3. Specific Dispersion Increment per Unit Theoretical Bromine Number of Olefins vs. Boiling Point
Actually, conjugated diolefins absorb only some 60% of th e theoretical am o u n t of brom ine by th e m ethods m ost reliable for determ ination of monodlefins (which employ essentially organic m edia—i.e., acetic acid, carbon tetrachloride). Hence, the factor
So - 98 - / i X 0.6 X B r No.
h ~ M .A .V .
will be more applicable.
Example. T h e noncyclic conjugated diolefin 2,4-heptadiene.
Boiling p o in t = 105° C., Si, = 214, theoretical B r N o. = 332.4, theoretical M .A .V . = 1019.
T h e specific dispersion increm ent per u n it maleic anhydride So — 98 214 — 98 n , , . j 11 . .I value is ^ A y - = — — = 0.114. I n th e presence of th e oth er olefins th is increm ent has been p a rtly corrected for by the noncyclic monoolefin facto r /i , which a t 105° C. is 0.170 (see Figure 4), or, expressed in term s of m aleic anhydride value, is , 0.6 X B r No. 0.6 X 332.4 _ n n o o , _ /■ M .A .V . 0 1 7 0 --- 1019---° 033- HenCe> h ~
0.114 - 0.033 = 0.081.
C alculated values of / 2 for conjugated diolefins are listed in T able I I a n d shown graphically in Figure 4. As expected, curves sim ilar to th e /i-c u rv e s were obtained an d th e factor of the
•cyclics is lower th a n th a t of th e noncyclics. (A lthough th e /*- curve of the cyclics is based on three points only, it m ay safely b e assum ed th a t its position is approxim ately as shown.)
Since factor f 2 is a differential correction, it is valid only when a brom ine num ber determ ination is carried out, which in analyses of cracked gasolines will alw ays be th e case.
Re c o m m e n d e d Ol e f i n Co r r e c t i o n Fa c t o r s. By aid of the curves in Figure 4, T able I I I has been m ade for more convenient use in analyses of an y fraction w ithin th e gasoline range. T able IV contains th e factors applicable to th e arom atic fractions, assum ing a 50/50 distribution between cyclic an d noncyclic ole
fins, as well as factors applicable to the analysis of full range
cracked gasolines (aviation and m otor stocks). Since gasolines contain more low-boiling th an high-boiling olefins, the mid-boiling po in t of th e olefins will be lower th a n th a t of th e gasoline; the estim ated figures are based on studies of olefin distributions made in these laboratories.
D I S C U S S I O N
I t is evident from these calculations th a t, for th e ty p es of ole
fins investigated, th e specific dispersion increm ent correction factor increases w ith th e boiling point. Thus, over th e distilling interval of gasoline it is nearly doubled. F urtherm ore, th e factor for noncyclic olefins is ab o u t one and a h alf tim es as high as for cyclic olefins of th e sam e class an d of the same boiling point.
T he olefins considered here undoubtedly constitute th e over
whelming m ajority of types present in cracked gasoline stock and therefore th e calculated correction factors m ay be sufficient for practical purposes. E ven if th e specific dispersion of other olefins were available for calculation of th e m any individual additional factors, th e y could n o t be applied because no m ethod exists for determ ination of such olefins in th e presence of other olefins.
(Straight-chain and cyclodiolefins, mono- an d diacetylenes, olefin-acetylenes, arom atics an d naphthenes w ith olefinic side chains, cyclomono an d diolefins w ith olefinic side chain, etc.)
These conclusions m ake it desirable to m odify the statem en t of Grosse an d W ackher (/, p. 616) regarding th e ratio betw een spe
cific dispersion increm ent an d brom ine num ber for monoolefins an d nonconjugated diolefins. T heir suggested value of 0.16 for th is ratio is derived from Figure 4 (1) where the increm ent of a few olefins was p lo tted against theoretical brom ine num ber, an d a
Table IV . Olefin Correction Factors for Determination of A r o matic Content of W hole Gasoline and of the Aromatic Fractions by
Specific Dispersion
( I t is a ssu m ed t h a t th e o le fin s are 50% cy c lic)
M id - M o n o C o n ju g a ted
B o ilin g B o ilin g o lefin D io le fin
A rom atic R a n g e, P o in t, F a c to r , F a c to r,
F r a ctio n ° C. 0 C. h h
B en ze n e 6 0 - 92 76 0 .1 3 5 0 .0 6 0
T o lu en e 9 2 -1 2 2 107 0 .1 5 0 0 .0 7 0
X y le n e s 1 2 2 -1 5 0 136 0 .1 6 0 0 .0 8 0
Cb a ro m a tics 1 5 0 -1 8 0 165 0 .1 7 0 0 .0 8 5
H igher a rom atics 1 8 0 -2 0 5 192 0 .1 8 0 0 .0 9 0
W h o le a v ia tio n A b o u t
g a so lin e 3 0 -1 8 0 7 5 ° 0 .1 3 5 0 .0 6 0
W h o le m otor A b o u t
g a so lin e 3 0 -2 0 5 9 0 ° 0 .1 4 0 0 .0 6 5
a E s tim a te d m id -b o ilin g p o in t of olefins p resen t.
4 0 6 0 8 0 100 120 140 160 ' 180 2 0 0 BOILING F*0INT, *C.
Figure 4. Specific Dispersion Correction Factors for Olefins vs. Boiling Point
June, 1945 A N A L Y T I C A L E D I T I O N 365
ence in calculated arom atic content of a sam ple, as illu strated by the following example: and maleic anhydride value determ inations. However, the factors are subject to the following errors:
Figure 5. Specific Dispersion Increment vs. Theoretical Bromine Number of Monoölefins and Noncyclic Nonconjugated Diolefins
As m ay be seen from th is tabulation, the greatest source of
mine num ber of the pure olefins were allowed to accum ulate.
Table V . respectively—so th a t a calculated value of specific dispersion m ay be used. F o r higher boiling fractions, however, th e error will increase rapidly owing to th e ever-increasing num ber of monocyclic arom atics of undeterm inable representation an d dis
tribution. (As previously m entioned, in th e upperm ost fraction th e situation m ay be fu rth er com plicated by the presence of the bicyclic arom atic naphthalene.)
366 I N D U S T R I A L A N D E N G I N E E R I N G C H E M I S T R Y Vol. 17, No. 6 arom atics applicable to the arom atic cuts, together w ith the prob
able error in these values an d the errors reflected in the calculated arom atic content. Since the m agnitude of the la tte r error de
pends on th e arom atic content, a sam ple containing 35% by w eight of arom atics is again used as an example in order to m ake a com parison of all conceivable errors possible. T he specific
Table V II. Error in Calculated Aromatic Content Due to Uncer
tainty in Estimation of Specific Dispersion of the Aromatics
A ro m a tic B o ilin g
Table V III. Determination of Aromatics by Specific Dispersion of Unsaturated Gasoline Fractions
P r o p e rties ascertain because the arom atic distribution varies. T he figures given in T able V II should therefore be considered as rough esti
m ates only. I t is conceivable th a t a fairly good estim ate of the arom atic distribution can be obtained by m eans of qu an titativ e silica gel adsorption an d desorption. I t m ay th en develop th a t for straig h t-ru n —viz., olefin-free—gasoline of a given source and for cracked gasoline processed in a given m anner, th e specific dis
persion of the arom atic aggregate m ay be considered fairly con
sta n t. Analyses by fractions should yield more reliable results th a n full range analyses because all factors applicable to frac
tions are more nearly correct.
Ex p e r i m e n t a l Ev i d e n c eo f Ac c u r a c y. F o r this dem onstra
tion have been chosen gasoline fractions of considerable olefin con
ten t, the arom atic content of which can be accurately determined by a n independent referee m ethod. Employing th e ultraviolet absorption spectrophotom etric m ethod, the arom atic content can be determ ined w ith a n accuracy of =*= 1% of th e arom atic con
ten t, provided the sample contains only one ty p e of arom atics.
Therefore, th e te s t has been restricted to the analyses of the ben
zene an d the toluene fractions. R esults are shown in Table V III.
I t m ay be noticed from the last line, entitled A arom atics, th a t this change alone will, in the case of high olefin content, alter the arom atic value by 3% . This discrepancy emphasizes th e im
portance of reliable corrections for th e specific dispersion incre
m ent due to th e presence of olefins.
(A lthough n o t required for com putation of arom atic content, th e olefin content has been included in T able VI I I as a m a tte r of orientation. T he value of the olefin content likewise depends on th e distribution of the tw o types of olefins, cyclic an d noncyclic, because for a given boiling range their difference in molecular weight m ay am ount to as m uch as te n units. Since the molecular w eight of th e olefins present in a petroleum fraction cannot be determ ined, it m u st be estim ated.)
A similar te s t of th e accuracy of the specific dispersion m ethod as applied to the higher boiling polyarom atic fractions and to full range gasolines is not y e t feasible because (possibly w ith th e ex
ception of the xylenes fraction) the spectrophotom etric m ethod is no longer sufficiently accurate in these ranges to serve as a ref-