Gas chromatographic separation, by carbon number and hydrocarbon type, of saturated hydrocarbon mixtures and different naphthas over molecular sieves 13X

CHROM.

6278

GAS CHROhIATOGR.WHIC HYDROCARBOX

TYPE,

.4XD DIlW3IEST

SEPAR.4TIOS, OF SATC‘RMED

XAPHTHAS

.A gas-solid ~llrr)mato~raylllic mixtures according to hydrocarbon sieves 13s.

O\‘ER

BY CARBOS

SUIBER

HYDROC.4RBOX

JIOLECI:L.L\R

ASD

MIXTURES

SIEYES

13X

method is described for separating hydrocarbon type and carbon number by the USC of molecular

Practical applicati~,,ls of this analytical metllod arc reported, including the anal>& of the charge and the end-products uf a platforming plant, and in addition, some quantitxtivc and qualitative determinatiow on virgin naphtha from different t:\pes of crude oil.

Tllis paper propows an cstcnsion of the mctlltrd dcscrilwd I~\- I_‘,lrr-ssocr< ASI) l*t-N for tic rapid separation of tlw II?-drocarbolis, according to carbon number, in saturat~~tl pctrolcwn distillates ul: to IS5 l‘llis prxcdurc permits tlw separation and determination of nal~l~tlwnes, isc-pnraitins, wpnraflins and results in improved accuracy* for the evalilz~tion oi str;ii,nllt-run wplitlias. liccausc tlw aromatic hydrv;carbons separately

are partially determined

and

irreversibly

according

adsorbed

by

molecular

sicvcs,

they

must

be

to the

ASTM D-2267 method. The object of this analytical investigation was to identify the various components of virgin naphtha and to rclntc them to tlw !kld of effluent products in tlh- refinery operations. The mctllods commonly used to determinate total iso- avd w-parafins (I’), okfills (0), ~q~l~tl~e~lcs (S) a11tl aromaiics (X) contcllt are usually* P.O.X..A. method (LOI Method 273-Q), or more sophisticated tcclmiques suc11 a; capillary column These methods gi1.c a series of data that cllronlatograplI\and mass spectrometry. are useful in rcscnrch, hut tlw long operating time required m&s routine application ditlicult . Tlw procedure dcscribcd in this paper for the dctcrmiuation of hydrocarbons by t\yes and by tlw number of carhn atr TM. is simple as ~~11 as a valid altcrnatlve to the methods mcntioncd ahovv.

The apparatus and experimental conditions \vere as follows. The gas chromatogap11 WT,Sa Perkin-Elmer Model 1: II with a flame ionization detector (single flame),

I

/

GC

SEPARhTIOS

OF HYDROCARBOS

MISi’IIRF_.;

connected to a Hitachi Perkin-Elmer chart speed of 5 mm/min.

Fig.

I. Schcnxttic

dinbpm

A?::? 0IFF13REST

Model 159 recorder

of chromatographic

SAI’HTHA;

5

with a I rn\T range and a

colunu~.

The chromatographic column was of length 90 cm and I.D. I/T~ in., made of annealed steel aud filled as shown in Fig. I. The components shown in Fig. I are the following. E is the carrier gas input, 0 is the carrier gas output, a is 1 in. of quartz wool, b is I 12 in. of Chromosorb P, 30-40 mesh, c is molecular sieves 13.X, 30-60 mesh (from Curtin, Lindc, Union Carbide, I_‘.S.X.), and d is I in. of quartz wooLThe molecular sieves (3040 mesh) were treated with XaOH (3 oh aqueous solution), rinsed with water to neutrality and dried in an oven at IOO for 24 11. Subscqucnt:! the sieves were kept in an oven at ZOO for 34 h and then sieved. The initial temperature was rSo -, and the tc-tnpcrature programming consisted

ASALYSIS OF

.\IITIFICI.\L

IIYDl
ZllS.I‘CKES

X0.

3

.\SD .+, E.\CH

\VlTII

.\Sl)

\VITIIOUT

SYLESES

(4 RVSS)

The xlsorption of s~lcncs is c\-idcnt (> yoO<, ) : the adsorption is less for tolucnc. The results the misturcs \vith and without s~lcncs confirm the mnrkcd adsorption of x~lcncs.

3

J

--2.77 _ -.

2.7’) 1 .o.# C,..j’,

--c).J

AZ.32 -+Z.?1 -. <>,iT -0.rKj

/ .(‘O

I

to.32

+

f-o.62

-+-0.41

+0.14

LO.32

-to.10

12,jc)

--

-1.70 -

+o.47 -1.33 -

1 .j3

20.00

+

IS.99 q.os 1 2 2_1

2O.O.j

+1.sq

22.42

29.12 26.25 .-

1.02

"O.jc)

lS.S2

.“.21

+3.-14 -3.?3 -1.41 -0.4s -4.96

30.70 3CLS.j r3.S7

-

15.27

for

?j

27

t-2.39 A- I.37 - IO.94

23.2s 11.31

Z2.O.j 22.01 2s..jg 26.25 -

2.W)

-1-0.02 i-L’.44 +0.40 - 1.76

-

I-O.64 +-a36 -“.14 - 1.94 -

-k-3.48 i-2.90 i-3.01 Z:;,”

j-I.29 i-o.55 j-o.90 -1.70 -

1:. GAIULLI,

6 TABLE KESI’LTS

L. 1-‘.\1:1.\s1. c.

1;11.1.\, i-. f.c!,‘Sl

I\ FOK

OCCIULSf.\L

VIKtiIS

S.\I’HTH.\

111Tables: IV to VIII: aromatics were dctcrminetl by the :\ST.\I D-2rb7 mcthotl ad the results in 0 o : y rolume arc trausformctl iuto ‘IO I)\- weight; the ~alucs obtaiucd from the analysis ou molecular sicrcs (MS.) 13X and ou a cnpky column (C.C.) arc csprcssccl in I:,, by weight; C, includes iso- awl u-butane. Lkxsity. I j,_1 = o.;,jt,(.\l’i W>.ll.

c;c

SEPAIL~TIOS

RESULTS

OF HYDKOCAR~OS

FOR SASS.\S

KAS

ThSURh

JIISTUKES

VIKGIX

AKD

ljIFFEliEST

SAPHTHAS

NAPHTHA

For explanation, see Table IV. Density, x5/4O= 0.7’94 (API 65.1). Idol;”.O

B.P. (‘C)

recovered

Initial

hSTI\I D-86 :

j3

7s

j

10

84

98

30 j0

110

70 90 95 lTinal Analysis (\Ol.-q,O) : 1' = __ ._--_-_ Carbon

nwnbrv

I’3 140 146 ljb T_I_(>L ; 0 .-

0.3 j ; S

=

Ib.+G;

.\I..%.x3-Y

X

___-_-..-

i.c.

_-

.1I.S.

r3.Y

G.02

3.78

.+.3s

,j,ZO

(1.I3

Il.?-)

SCI Il.jj

1 LA)6

12.07

0,j.l .j.t

I

3.‘,‘) 0.O.j

j

4.42 _$uJG

_j,?;O

5.33 4.72

-zzzS.jj.

_~----

iso-l’wafjiizs

Saphtlwrltis

C.C.

= _--

“-34

7.71 3.4s

0.21

0.1s

c6.10

LY.22

u-Pamifim .-~ X.S. rg;Y C.C.

(;c

SEPARATION

OF HYDROCARHOS

JIISTl:RlSS

a-\SD l)II~l:I:RI~ST

SAPHTHAS

9

01 sampling to 330” at 5”/min, the:1 from 330” to 450” at 2”/min. The evaporator temperature was 2o0°. The carrier gas was helium at an inlet pressure of 1.2 kg/cm’, splitter 300 ml/mm. A series of artificial hydrocarbon mixtures was prepared and analyzed by the present method to check for eventual irreversible adsorption of heavy compounds and to lay the basis for quantitative analysis. The results are shown in Tables I, II and III. It can be seen that, with the exception *>fthe aromatics, the other hydrocarbons have a response coefficient to each other which can be considered to be unity in all instances. For the. aromatics, the followin, ff correction factors, F, were found 1.000; benzene, oS929; tolue,lt, 0.9345; and sylenes, experimentally : Iz-heptane, 0.97so. As shown in Tables I, II and III, by applying the correction factors relative to these components in the flame ionization detection, a good correlation is seen between the expected values and those found experimentally, although some adsorption of toluene .md ?z-decane is observed. The results for sylcncs and heavier aromatics obtained by this method are unreliable. Therefore, this method is applicable only to saturated hydrocarbons up to n-decane. Other methods (e.g., ASThf D-2267) must be used for the determination of the aromatics. actual samples of virgin naphtha !roin vari::uu crude oils Subsequently, (Occidental, hIiddlc East, Sassan and Sarir) were analyzed. The separation by carbon number and hydrocarbon type was accomplished by using molecular sieves 13X and the c.‘lnrlnatogram is shown in I;&. 2. The samples went: also analyzed by P.O.S.A. according to the (&.(’ \Iethod 273-64) and capillary column (C.C.) chromatography met 110ddescribed by LEVEQUI?. TIIC results are given in Tables IV to VII. It should be noted that values obtained 1)~molcwlar sieves and capillary rolumn~ are expressed in o. 1~7 weight, while values obtained by P.O.S.X. method are in ‘,‘, b!v vulumc. The molecular sieve method was also applied to characterize the feed and the effluent products, at various stages of reaction (liquid phase only), of a p!atforming plant (Table VIII %nd Fig. 3).

Fig.

P. Gas chromatogram

of 3ccidental

virgin

naphtha.

F. GARILLI,

t.

F_UilASI,

u

I-J

I

I

I i

!

1

i

V. FILI.4.

V. CU.51