Infinite dilution activity coefficients of ethanol-n-alkanes mixtures

Infinite dilution activity coefficients of ethanol-n-alkanes mixtures

103 Fluid Phase Equilibria, 27 (1986) 103-118 Elsevier Science Publishers B.V.. Amsterdam - Printed in The Netherlands INFINITE DILUTION COEFFICIE...

794KB Sizes 0 Downloads 136 Views

103

Fluid Phase Equilibria, 27 (1986) 103-118 Elsevier Science Publishers B.V.. Amsterdam - Printed in The Netherlands

INFINITE

DILUTION

COEFFICIENTS

ACTIVITY

OF

ETHANOL-n-ALKfANES

MIXTURES L. .

CX3li1’and

Mont.edi

c Italv

pe

F:‘. DELUG!I’

Research

Centrc,

Via

San

F’iett-0

56,

Boll

ate,

Mi 1 an

).

Infinite ethanol

dilution

a.cti.vity

coefficients

o.f

the

system

formed

o .+ funct.j.on n--al kanec, have been measured as a and tempera.t.ure by a yas sir-ippj ng method. Li near dependence5 c!-f’ .the i n have been found ,for al 1 systems studied In y (D on temperature the 1 imj t a+ the e::pet-imental errors. Regular trends along thtn--*l.l.anes homologue series have been observed. Gibbs ex tees The paraxeters of the most popular models of en?rqy have been 0bt.a.i ned. They allowed the predi.ct.ion of vapo~w-3 ~.quid equil ibria of the systems studied. The comparison between cal.c~~lat~?d and experj.menta.1 data taken from the literature showed i..I/ ?i:t t.he four parameters Wilson model is at the moment tlie b e E-,1.. model for thc35x Eyst.ems, becac!E;e it is able to describe at %.l-lE:! z’iam~+ time the depende.nce of ym on temperature and composi 1:ion.

by

I

Nl’F?m.NJCT IC!N

5q’l t.e In thermodynamics A cnrrc?ct vnsal ved.

o+ of

descr

thi? qt-eat: number alcc!hols-.hydr-ocal-bons :I pti. on

cl+ sLlc:h

of

rzystwms

studi es solut.ions, must.

st i 11

devoted to probl.em the be

c3nai

.I:. h E’

o-f der t+c.i

i_t\*:> nature o-f alcohol. 5 severe1 y tries mol ecu1 ar i.nterractj on t he I j cp.1.i CA at.. their crapabi 1 i t.y tn .f or m mu 1 t i I:!1.~5 pal. +.ri ty and hydrugsn bonds make inadequate t.tle “physical ” thenries; t-he pur CI ar e a I.S!L) chwn1 cal thec.ir i ec t I7e i r more simple form j. n r e z1.1. :i ncuf.ficiw-+t.. INi.x ed thee-it5 sAuca.1 d better drzcri, be the behavi our , a~, they take into account both chemical. and phyelca.11. iI~i.:eracticws. tk!\ler-the1 ~!z-E not even t. her;c theorj e5 can cot-rec_t_l y r eprc~wnt, the who1.e set of experxmerltal obrervatj enc. ‘Thr ) j. 5 normal :t.y iL!dUEL.d qnodneac, of a theory ( or of a model When the par *met c’rci by 1 ts al;i lity tn fit t.he expel-j mental de1t.a. 1.2 -E d model da.t..a 1.15c?d t. c:> are dw-i vcd irnm t.he Sd”w! ~,tF”_‘l-irnerlt~aI what we real1.y test ic it.s dcrcti 1 i .ky mc.crc? tec,t the model itself, t h,ari zta agreement. t.o ttle physi.cu; of the phenomenon. A poly nutni $1. t he ~ec)l.rat i Clr, c3.f 5ui t.ahl e ab1.e tc1 fit degree ma \y be “H3l-e e 3 p e r i ,”C.! n t ill 1:t 15 i_.h E! dr1.a t I-Ian P f wdwment.al phyci cal model . a,-,i , c1se t. o “When consi.der inq the ~~a.1idit.y of ci t.e f:)l--au5r:l t.z : (4t_,&a? I_b,0.2 1 t II’ tIiE!ot-y o-f 501. ~4i.i onEi ii. ic; more impnrtclnt to incl!.!ire 5’c1 ;;,, + :I not based or, it cari be supported by evi dencec, I..1,&I;.,, L.!:’ pl~ysi cochemi c:.al. properties. ‘1he v b 1.I clj. t: y Cl:: iin j ;lQ-e<..‘l’,“, : I.. i.f it. can tcE> <.hc:,iw 1 o t!!:$ i r, 5 o 1 I..., t i on s iE: mluch enhanced (c;$2IA i. t i !:at:+ .!.i.~j. 1) nj 1-h obr;ervnd physi c:a:l. pr r:~pc=~ t..i r’~.: ~4..t ,t:!I 1.I 1;:tr) i..h ci5 +I’c. e::

The

i st i rly thew 5.t.at.c2; their

0378-3812/86/$03.50

chemj

j. es,

cal.

cm the

0 1986 Elsevier Science Publishers B.V.

104

105

106 TABLE

I

Experimental activity system ethanci-n-altane.

coefficients

at infinite

dilution

for

_--___---__--___--_~~--~~~-~~~~--~~~~~~~~~~~_~~~~_--~~_-~~__~~~__ ALKANE

T/C

y”

Tl C

Ym

ETUANOL ALWANE _---__--___-____-_______________________________________________

n-pentane

36.5 37.0 50.5 b5.5 $6.0

47.0 49.5 33.0 21.0 23.0

26.5 31.0 37.2 39.0 40.0 49.0 51.5 62.0 69.0 79.2 El.0 81.2

8.9 ?.1 8.7 0.6 8.5 8.3 8.3 8.2 7.8 7.0 b.? 7.4

n-hexaoe

24.0 24.2 24.5 43.4 45.5 io. 8 bO.8 78.3 78.4

62.6 58.2 59.4 37.6 35.4 22.4 21.9 14.7 1488

47.0 81.0

ID.0 8.7

n-heptane

41.4 41.3 59.0 76.4 lb.4 93.6

34.6 36.3 23.2 14.6 15.1 9.8

49.b 64.5 81.b

1280 11.2 lb.P

49.0 b4.0 RD.5

14.5 13.1 13.0

27.9 47.7 47.7 82.0

20.8 IV. 8 18.9 15.5

n-octane

n-mane

22.9 22.9 29.2 39.2 39.1 41.8 42.3 so. 5 50.5 AD.! 6O.U 82.3 82.3

50. ? 51.7 42.4 32.3 32.6 30.4 30.4 24.2 23.8 18.6 lB,B 11.5 11.4

n-decane

33.2 48.2 65.5 84.3

37.5 25.5 15.9 ID.9

the

107 6

ETHANOL

0

10

20

-

JO

TEMPERATURE

c

nPENTANE

40

50

‘C

60 10 80 90 100

( Scale:

l/T

140 K J

ETHANOL

-

“OCTANE

ETHANOL

-

.OECANE

t

120

TEMPERATURE

‘C

I Scale:

l/l

K 1

108 to operate at temperature even higher than the That al 1 ow5 the Moreover t. h (2 normal bailing point of solvent. evaporation rate of the solvent can be controlled. the more the gas stripping technique only the y m of 2) By Luckjly, in can be easi 1 y measured. volatile component minimum azreotropic systems, 1iL:e ethanol-n al kanes, the more volatile component is alwa_ys that at high dilution. In t.1-1at ccw be obtained by the same technique. case bot.h the y o 5) The accuracy afforded by out- apparatus was Uy = 0.5 in the experimental range. The accuracies of the measured variables were: up = 1 mmHg , (I D - 4.0 ml. ./h, (IT =0,5 “C. Tt,e reprodcrcibi 1 i ty of the measures wac at worst f5i: of the mCari val ur.

RESCJLTS The obtained resultc, r7re shown in table I. Rlong the series o.i n-al i::anes C5-Cl0 only the activity coefficients of ethanol in Doctane were not measured at ~711, whi 1 e mcas~tr ement of y m of n -decane in ethanol were unsuccessful 1y because of the tried limits of our g;?c,c~,rornatoRraphic detector. Figure 2 shows how In y o varies with 1iT; it allowLl t he thr?& compar i 501, wi t h the few data found in the literature OrI systems (hlessi and k::ikic, 1985 ). The compari sun I ookrj quite good when the hydrocarbon is; the z.ia1ut.e; when ethanol i 5 the c D 1 1.1 te the Ii terature values are lower then those measured in this WOI-I.:. Howeverthe data from the li.terat.ure are qc!ite spread and the) don ‘t af ,Ford a reyul car t.rcnd al.onq the homnl,oque Eeries. Owing to t.he I inear dependence o-f: lrl y m on 1.11. ra-f t.he d3t.a eactf s,,stem have been linearly inter-palat.ed. Table 11 ShoGJL- 1.h~’ i,i and F . par amct er-ci of t.he eq1.rat.i en

I

C1 i

In ri* r Ai+Bi/T

obtained for- every c,yst.em. The sjtraiqht lines chown in fis. have been cal c~1.1ated with the parameter 5 of that table. The 51 ape D i is di.rectly related to the heat. of dilc!ti,orl

ethanol 17y d r UC C!r b on

_ _._. -- . ....___.--...-..-

hydror,srbnn

..__ - ..._.- ..._.

_..... _......_.__. _._.._. ___-_....__

‘2 at

109 Ini-inite dilution mear,w-emcnts n+

Lf the

usually enthalpy

Xt is

.

excecs Li”

=lim

experimental. val ue!s The system UC ihc c: c:,mp c 5 i t..i ~7n 5 sseqvat I on 1 i I..:e the I Rc!qal 51: i. , I.977 1 eqc!a?: i nn

=

o-f

from the

calorimet.ric equation (. 7; , .

dXi

measured nf: hE a,- e i nterpol atcd Symmetrical ! sc!m ci f

mWLxlg2 :I

obtained means

SF_

If0

$

hy

5 + ( wL

at by

1

di+feren-t e,icI,t.abl e Functi ens>

, .a 1 -:a

VL’2

2 )

?

O.F.

=

NEXPl r i-1

GXP

-YPjl&

2

NEXPL 4.

2 i-1

"&XP

OD - ‘2iCAL

I2

i. :‘, )

113

In

I,==

cl2

In

yzBe

c21

(6)

results That dependence

in of

a

1 inear

parameters

Cl2

C1, vs temperature, in and agreement with the theory of equation I F’rausni t.z , t hc Ii i I.969 ). The ~ i and shown in table II qive directly t.he dependence of C:ii The on temperature. prediction of the dependence activity Of the coefficient5 nn composi. t i on is very good. Fnr t.he not system ethanol --n-heptane the Ei bbs energy of mixing ha2 calculated following been the rel ati ens:

FIG. 4. Ethanol-n-hcptane system at ‘T=40°C. Gibbs energy 0.f mixinq computed by Van Laar model.

(7)

,-Al3 RT

id

ciE obtained

which

doesn

cw-ve,

‘t

actual

k!.i.I.z!E _es..~!tAx!,

x, In '1 +

(9)

12x1 4 C2? “2

shown in 1 y ex i St.

fiq.

At. infinite In

"2 'nx?

Cl2 c21 Xl ‘2

RTCC The

L

4

4,

predict

djlution

81 -In Aii--Aii

it

is

a miscibility

( Wilson,

gap

1964

)

i : 1.2

j : 2.1

where

*xp(_$)

n&i “i

! 11)

1 /T impossible t.o deecribe the dependence It is of lnym vsi with this equation by I_eeping A i .i ‘5 independent of ten!perat.ure,. Neverthel err, 5 I. i near dependence on T t..h e a reproduces experimental data wj thin the experimental error, 5 shc!ws as f1.q. for the system ethanol-n-pental-le. Table IV q i vec5 i.tlc= kJi 1 son o_i I‘ constants der i ved from experimental data for all ttle systems studied. s);ct.amci ‘The VLE for ethanol s-n-l>exane, n-heptane and n-nnnane ha ;e been calculated at the same temperature where e:.:perl merlt.al cwd5 ) C3at.a Table IV rtlew5 the root rne~n 5quare are available.

tht? ac:t.i vity relation5

cnwff

icients


0

71

=

K+l

=

K+l K

(D

y2

derivat.i

\‘es

on

di 1ut.i on

temper-atc.lre bin 70D K 1 =I-bu / T ) K+l ilInyT W/T

1

1

Ah R Ah

=-K+l

R

are

ylvcn

bit

c.he

113

gE=LIE +gEthem phyr

1)

1:t 5~iOUlCl ,,,clr-f? accI..lr-at.e1> describe t.he dependence 04 ~2E over any t il E’ -F1.11 1 r a 1-1 gc 0.f composj t i an. It 5hfml dn : C prrdi ct. VL f.: d d t. ir ‘I m.i.5 c i t2i 1 i t..f it z.h cx! I.d i rnpr exe t. he! f i t. 04: g t\p j avoiding !systemat.ic: trnndr, 0.1 error5 with composit.ion.

114 c’)

L

predict the correct dependence 0i: I II y a nn rhoul d 1.t e qu i 1.i. h I_1.um throuqh that of t-he associat.ion temperatcrre of: temperature in place the ar-hitrar-,constar7t.s on param~l.ere on ‘17. I. inear dependence of physical. interaction has (3 paramet..wE3 ct:: “, ) A 11* < l-he model of: Nath and Drnder

1::: *“, , AhAm , AI.,:~ , duo,, pas-arnet:.et-s can be estimated

, Au;,,

, AL,;*

1: a ma:x~m.~mCT+: ;I

i 61 par wwters model , the binary assnci c&t-ion. Physi. cal y).

parameters par amciterc

‘i *i-e al I. physxcitl. r7re est.imat.ed .f’rarri

.From

the val ues and i r-om t.hk?1 I-. so , it 1s necessary tc3 f:j.:: 3 pr1.cwa dependence on t.emperature. d I-, 4 ‘S_r we,-p Ii; ‘=._ and t: h E! vi\1 c!es of at. l.east 4 paramet.et-s. arid IWI~W, 1’7t?l i, i I obtained .i:rom the proper-Ci,es of pure A i Nath F’r~r t.he other b par amrt er r, WE:have t I- i ed t tie f:ol I ow3 ng clloi cfzfc :.

nn crossed

TkBLE IV Yilson node1parametersfor ethanol Ill-n-alkane (21 aysters

n-dkane

Parametersfron this work i calicole

42

cal /sole

n-pentane n-hexane n-heptane

2757.2 2b99.2

n-nonane

A0 32 21 cal/nolei°C caliaale

Cwparisonnith data frost the literattire

i Gr cal/mole/°C

343.24 363.96

-0.8421 -G.X%

273b.b

-9.986 -9.512 -10.025

457.56

-0.3723

2681.8

-9.221

579.lb

-0.0256

Refwences

T/Y

RI6 on y RHC; on P lLPai

Janaszewskl1982 Janaszewsii1982 Rerro 1982 Berm 1982

40 40 70 76

O.OiiS4 0.0055 0,0109 o.co25

3.4E: c.23 z.01 7.39

115

> c c.2 <

-

ETHANOi

-

nHEPTANE

8 0

IO

20

30

40

50

TEMPERATURE

F1.G.

7.

Effect of temperature coeff icier-its for the between c::perj,mental model

nlodel

5. val

uez.:

(-)

C

60 70 80 90 100

( Scale:

on the infinite ethanol ---n--hcptane a~ld romputed model a;

by (-a-*-)

140

K 1

dilution system. Mal:h and modal

activity Cornpat-i E.O~ Ecender b; C....)

c.

0.m

X

FIG ” 8.

120

l/T

\‘LE predi.ction Nsth and b ; C----J

Render model.

.For ethanol models: C.

1

ETHANOL

--n-heptane !-_)

model

system a;

at

(----_)

4Cf’C

by

model.

116 CUICLUS

1 ONS

allow a. strict te~.t 05 the gE data shown in this paper No clf’,e systems ethanol --,r~-..,alkarres. ity to represent the a t: t hE’ 5<3”,E tj me used Cli-l predict normal 1 y I C The assnci .t.“~~~ and compocl ti on. of y ’ or3 temperatLu-e 1:h (L’ bctC aC tt1e moment iurt.her, model 5 s,cem to be wotrt,h worki.ng parameter:: model q wJ.t h .fit5 lrii 1 cion t!eat: are obtained by the The set_ o.(’ data obt.a.ined ic: dependent, on temperature. 1 inenrly anti it allows the eval uirtinn of the behax/iotr! cp1 te homogcncou~, I ti c OLII. d be t a 1..en P s> iz, ~3I.on !:I tli v home! 1 ng~lk3 se,. i e6 of n--it 1 kan ~3. !ztart.inq point For. <:ur-ther improvements on the exiE.t.1 nq tnodc?.Is. The new mndels abll o -f t: hnsc dependence

117 non arsscjci ati c:~~~c.~ssc-cl a5snci

. E

M i ci 0

ng

cumpanent

at i on

at infinite dilution e:.:c.c5s f:unction mi ): i nq j. d a a 1 Cemperature t.emperakwe

independent dependent

term term

118 ki., Leroj q J.C., Masson, J.C., Fabriee, J.F., Muacurcment 0.f Activity Coefficjents 1.977. fkcurate dilution by Inert Eae Strippinq and Gas Chramat.ography. lncl. Eny. Chem. Process Des. Dev., I&: 13Y--144.

Renon,

Rngal ski 1: 137.

Ynung,

,

C.L.,

W.

,

Ma1annwski

1868.

Chromatag.

,

5. ,

Rev.,

1977.

Flcti cl Phase

Sanni

er,

at.

Infini

I#.

,,

tt2

Equi 1.i. bri 3,

1f.l:129.

Wi 15on , G.M., 1964. Vapaur,-Li quid expt-esfion Tut- the exce55 .ft-ee energy of 80: 1L‘27- 1 :;o _ J . Am. Chem. Sot.

Eqctil ibrictm mir:incj.

XI.

F,

r,F O&J