Korean Z Chem. EHg..

1911 I, 61-67 (20021

Adsorption of Chlorinated Volatile Organic Compounds in a FLxed Bed

of Activated Carbon
Seung Jai Kim*, Sung kbng Cho and Tae Young Kim

Department of Environmental Engineering, Chumlara National University, Gwangju 500-757~ Korea

(Recefi,ed 19 June 2001 9 a c c e p t e d 2 8 A u g u s t 2001 )

Abstract-Adsorption isotherms of dichloromethane and 1,1,2-trichloro-l,2,2-trifluoroethane on an activated carbon

pallet, Nofit B4, were studied. For these ch~nicals, the Sips equation gave the best fit for the single component adsorp-
tion isotherm. The adsorption affinity on activated carbon was greater for dichloromethane than that of 1,1,2-trichloro-
1,2,2-trifluoroethane. An experimental and theoretical study was made for the adsorption of dichloromethane and 1,1,2-
tricliloro-l,2,2-bifluoroethazle in a fixed bed. Experimental results w~:e used to examine the effect of operalioll var-
iables, such as feed concenb-ation, flow rate and bed height. Inb-aparficle diffusion was able to be a\plained by a surface
diffusion mechanism. An adsorption model based on the linear driving force approximation (LDFA) was found to be
applicable to fit the experimental data.
Key words: Adsoiplion, Didflororaethane~ 1A,2-tficlfloro-l,2,2-tfifluoroethane, Fixed Bed

INTRODUCTION ethane ICFC-1131 and dichloromethane IDCMI was studied theo-

In recent years, the adsorption process of removing and recover-
ing organic solvents in trace levels from air has attracted special
interest as a means of protecting the envirolmlent froln aft pollu-
tiort Volatile organic compounds (VOCs) are a family of carbon-
containing compounds which are emitted or evaporated into the
abnosphere where they participate in photochemical reactions [Jmg
et al., 19)8]. Some VOCs are toxic and carcinogenic. Most VOCs, to
varying degrees, coimibute to the fcmration of ground level ozone.
The VOCs which may cause stratospheric ozone depletion are most-
ly a chlorine-containing group of compounds imown as chloro-
flota-ocarbons or CFCs. They release chlorine and other halogens
into the stratosphere gradually [Wolf, 1992; Cicerone et al., 1974;
Cho and Choi, 1996]. These compounds are particularly effective
in destroying the ozone even when present in small quantities. VOCs
are used as solvents in manufacturing of countless products. Some
prorainent e,~amples of these products include crystallized organic
chemicals, printed materials, paints and coa~gs and dry-cleaning
agents in magnetic tape and semiconductor manufacttimN facili-
ties. Clile consequence of these raanufactumlg operatioils is tint large
amounts of organics are being emitted to the atmosphere. VOCs
have been implicated as a major conttibutic~lto photc~lemical sraog,
which can cause haze, damage plant and animal life, and lead to
eye irritation and respiratory problems [Ogura et al., 1992]. Several
technologies can be used for recovering VOCs flora gaseous wastes,
but one of the most important and effective methods for control-
ling the emission of VOCs is the adsorption process. Temperature
swing adsc~ption using an activated caduceibed is probably the most
common adsorption cyclic process for the separation of strongly
adsc~bed species. It has been widely used in industries for removal
and recovery of hydrocarbon and solvent vapors [Chihara et al.,
20(x)]. In this work~ the removal of 1,1,2-trichloro-1,2,2-Nfluoro-
tTo whom correspondence shotfld be addressed.
E-mail: sjk~in@cholmmn.cholmam,ac .la
retically and expedlnentally.
MATHEMATICAL MODEL
A mathematical model was developed that takes into account
the non-ideality of adsorbable species in the adsorbed phase under
equilibrimn. Applying the typical assumptions [Hwang, 1994] mid
taldng the mass balance of the gas phase in a packed be& the fol-
lowing governing equation for isothemlal adsoipfion is obtained:
ac, D o~'c' ac, 1-e aCL
-~ = ~a--~ -V~zz --7--OP-~- (1)
In Eq. ( ] ~, the gas-solid mass ~ansfer rate can be expressed as
the following linear driving force (LDF) model for surface diffu-
sion [Tien, 1994; Lee and Moon, 2001]:
? =k.( CL~-eL)= 3--kc(c,-c, ~) (2)
l) pp
where
15D:
k~ , (3)
r;
As denoted by Eqs. (21 and (3 I, the LDF model is an approzd-
marion to the solution of Fickian diffusion inside a spherical parti-
cle. The expressic~l assumes that the mass trmlsfer rate of adsorp-
tion is proportional to the difference between the equilibritrn con-
centration and the bulk concentration of the component. Since the
solution of the LDF model is much simpler than the solution of a
diffusion model, we employed this model in the present study. By
ffm-~tucing appropriate dknensionless vmiables, Eq. ( 1 ) cml be mit-
ten as follows:
a_s !a_~,
O~ 130S -~
and the dimensionless variables are defmed as follows:

61
62 S.J. Kim et al.

vt z vL c&:'~k'l'r'l-e'l=-=-=]
] "(-]-~) rr',
(5)

The associated initial conditions for 0<z<L is t ,.j

~(z, 0) 0 (6)
"-m
and the boundary conditions at z=0 and z=L for t>0 are:

= (r r (7)

~ ==0 (8)

As previous researchers [Moon and Tien. 1987; Tien. 1994; Wang

and Tiei~ 1982] pointed out in their work oi1 f~xecl-bed adsoiplioi~
if the inb-apartide diffmion is described by a homogeneous diffu-
sion model or a lumped lmmmeter model (such as the LDF modell.
one needs to calculate the concei~-alions in both pisses at the same
time. From Eq. (2), the following dynamic condition could be ob-
Fig. 1. Flow diagram of adsorption system.
tare& :1) Flow meter (i~ Heating band
c; +A~q,§ B, (9) :~) Pressure bottle ~ 3 Way valve
::~"Water bath 8, G.C.
where :4) Column !! RC.
:.~)',Thermo couple .l.-?,N2
3k:. and B, =c, +A,% (10)
(Jeio Teck ) was attached to the pressure bottle in order to keep the
EXPERIMENTAL temperature constant The concentrations of solvents were deter-
taRed froln their saturated vapor pressures by assuming vapar-liq-
The adsolption system in this study consisted of the activated uid equilibrium state. In order to increase the mi:dng eflSciency of
solving, 6 mm glass beads were packed in the pressure boNe. The
carbon (Norit B4, pellet type 1, and CFC-113 and DCM. The phys-
ical properties of the activated carbon and bed characteristics are gas mLxture of the saturated solvent vapor and the pure nitrogen
given in Tables 1 and Z respectively. A schematic diagram of the gas was delivered to the adsorption section. The gas flow to the col-
feted bed experimental setup is presented in Fig. 1. The apparatus umn was controlled by a flow meter that was precalibrated under
was constructed with staitfless steel tubes. The adsorption colunm the expmme~ltal pressure conditions. The gas mixture was heated
line is 0.42 cm ID and all other lines are 1/4" ID. Pure ifib-ogen gas to the experime~ltal temperature in the preheating coil section before
was sent to the solvent saturator. A constant temperature circulator b e r g charged to the frxed bed adsorber. The adsorber was made
of a stainless steel tube in 1 .(13 cm ID. A K-type thermocouple was
Rstalled at mlet section of the packed colunm, and the temperature
Table 1. Physical properties of activated carbon
of the system was monitored continuously on a display recorder.
Activated carbon The composition of the e:dt gas stream from the adsorption section
Supplier Norit (B4) was measured in the alalyzea sectioi1 For t t ~ purlmse, a gas ct~-o-
Pellet diameter (mm) 3.9 matogmpl~ GC-14B model (Shimadzul, equipped with a hy&ogen
Pellet length (mm) 6.7 flame ionization detector was used.
BET surface area (m2/g) * 826
Micropore fraction ( < 1 nm)** 45.16~ RESULTS A N D D I S C U S S I O N
Porosity 0.42
1. Adsorption Equilibrium
* from manufacture
The gravRletfic method is a well established tectmique for ob-
** from nitrogen adsoiplion at 77 K
taitmg adsorption equilibria of pure gases and vapors. I).l g of ac-
tivated carbon in the quartz basket is velNlated to remove impuri-
Table 2. Experimental conditions for a fLxed bed adsorption ties within the activated carbon at 503 K for approximately 4 in- by
Variables Unit Range using a vacuum pump. Adsolption amounts are measured at diffea-
ent temperatures of 303 K, 318 K, and 333 K. The adsorbent used
Bed length m 0.15-0.25
for this expelmlent is Norbit B4 and the adsorbates are CFC-113
Flow rate m/s 0.369. 10 2-0.890. 10 2
and DC3r The adsorption data were obtained by using a high-pres-
Bed porosity 0.12
sure microbalance. Isotherms of CFC-113 and DCM on the acti-
Packing density kg/m 3 693.5
vated carbon at tt~ee different temperatures are shown in Figs. 2
Bath temperature K 298.15-308.15
and 3, respectively. The isotherms of CFC-113 and DCM were fa-

January, 2002
 Adsorption of Chlorinated VOCs in a Fixed Bed of Activated Carbon 63

Table 3. Adsorption equilibrium isotherm parameters of CI:FC-

CCI2F on activated carbon
Langsnuir Freundlich Sips
Temp. [K]
% b k n % b n
j..~llf. ~,,........ 0 ~,nl...... . . . . . . . . 303 3.48 5.76 2.64 5.18 3.60 8.79 1.13
./"o ~ ~ " Error (~ 3.4 17.9 3.8
318 3.12 2.59 2.04 5.42 3.33 1.75 1.09
I Error (%) 3.8 4.3 3.2
/ . ~ I
2'
333 2.69 1.89 1.51 4.38 2.6 2.17 1.01
Error (%) 3.3 4.55 2.5
9 303k
318K I
1 9 333K Table 4. Adsorption equilibrium isotherm parameters of CH2CI:
- - - - Predicted on activated carbon
Langxnuir FreuMlich Sips
! Temp. [K]
% b k n % b n
0 5 10 15 20
p (kpa) 303 5.91 2.15 3.25 3.83 6.44 1.51 1.03
Error (%) 5.9 11.7 3.5
Fig. 2. Adsorption equilibrium curves of CLFCCCIF2 on Norit B4
for different temperatures. 318 5.83 0.78 2.76 3.96 5.96 0.73 1.01
Error (%) 1.8 2.8 1.4
o i i 333 5.03 0.51 1.96 3.16 5.60 0.39 1.04
EiTor (%) 5.1 4.1 3.4

this pressure range, the affinity of the adsorbate toward the adsor-
6
bent is ct~-actelized by a Henry's constant The heats of adsorption
can be calculated from the Van't Hoff equation, which is defined as:

I/Pl~ AH.( 1 (11)

4

9 303k In the HenlT law region IC~=KrP), Eq. ( 11 ) can be replaced by

• 318K Eq. (12).
9 333K
i Predicted
0.7

C ~ i ...... i i

0 5 10 t5 20 25 30 0,6
p (Kpa}
Fig. 3. Adsorption equilibrium curves of CH2CL on Norit 134 for
different temperatures. 0.5

~la

vorable types. Single-species isotherm data were correlated by the 0.4

well-lmown Langmuir, Freundlich and Sips equations. The param-
etea-s of each isotheml were obtained by the least square tilting with
experimental data. These parameters and the average percent dif-
0.3
ferences between the measured and calculated values are given in
Tables 3 and 4, respectively. Among these isotherms, tile Sips equa-
ficxl with three iZal-ametels is more appropriate m prediclmg our data
0,2 i i i ............. i
compared to the Langmuir or Freundlich isotherms with two imra - 3.0 3.1 3.2 3,3
meters. Isothenns provide itffonnaticxl on tile affltfity of tile adsor-
I/T x 103(K-I)
bent toward the adsorbate as well as the heat of adsorptiort The ad-
sorption amounts increase linearly up to a pressure of _ 5 kPa In Fig. 4. Temperature dependence of Henry's constant.

Ko]~an J. Chem. Engc(VoL 19, No. 1)

64 S.J. Kim et al.

( AH.,/1
1ntkpl)= 1
•Tt~, ~7 (12)
1.2

It can be seen from Fig. 4 t i n the H ~ y ' s constant for DCM is great- 1.0
er tilan tilat of CFC-113, and DCM is more strongly adsorbed on
the activated carbon than CFC-113. The heats of adsoiptic~l are al-
most comparable, although that for DCM ( 5.9 kcal/mol is slightly 0.8

greater than that for CFC-113 ( 5.0 kcal/mol 1.

2. Bed Adsorption o
0.6
A colunm adsorber has been used as commercial equlplnent for
separation by adsorption since it gives a sharp break'through due to
the diff~ence in affinity between gases and particles [Park et al., 0,4
2001 ; Hu et al., 21Y)I;Na et al., 2(x)l ]. The break'tllrough curves of A CL2FCCCIF z

all species, in gen~-al, depend on adsoipfion equilibrium, intrat:atticle 9 CF[~C[~

............... P r e d i c t e d
mass transf~-, arm hydi-odynalnic conditions in the colunm. There- 0.2
fore, it is reasonable to consider adsorption equilibrium and mass
transfer silnultaneously in simulating the adsorption behavior in z~
1 8 ........... ~ ~.-" : ' 8 .... ~
the fixed bed adsorbel: The operational factors such as input con- 0.0 0.5
. . . . .

1.0
. . . . . . . . . . . . . . .

1.5
. . . . . . . . . . . . .

2,0
centration, flow rate and L/D ratio are i m p o ~ n t in an adsorber de- Time (hr)
signing and op~nization, In tiffs work, breakthrough curves were
obtained under various ~iperimental conditions mentioned above. Fig. 6. Comparison of breakthrough curves of CI~FCCCIF~ and
The experimental <xmdidons are summarized in Table 2. Fig. 5 shows CH:CI: adsorption on Norit B4 (25 ~ L: 0.15 m, v: 0.369x
10-2 m/s).
the effect of film mass tr~lsfer coefficienk k~ on the breaktln-ough
curve of CFC-113. As the value of the mass h-ansfer coefficient in-
t.2
creases, the breakthrough curve becomes sharper and the break-
through behavior is very sensitive to k/value. This implies that the
adsolption equilibiium comols the breaktt~ough as the mass b-m~s-
1.0 ( ~ 1 ~ / ~ r moA
fer resistance dwindles, ks for CFC-113 is approxmlately 10 4 in/s.
The breakXhrough curves of CFC-113 and DCM at the same con-
ditions are shown in Fig. 6. The breakthrough ~ n e of CFC-113 is 0.8
short~- than DCM because the aff~fity of CFC-113 is smaller than
that for DCM. This figure also shows that the predicted br~&'through
curves by the LDFA inodel incorporated with the Sips equation are 0.6

fitted well with the fLxed bed data for single component systems.
9 L : 0.tSm
Fig. 7 shows the effect of bed height on the CFC-113 adsolption. O.4 O L : 0.20m
,L L : 0.25m
- - Predicted
1.2
0.2

1.0

03
// ~--/./-.--

/"
0,0
0,0 0.5

Time (hr)
1.0
...........
1.5 2.0

Fig. 7. Experimental and predicted breakthrough curves of CI:-

o

r,,3
0.6 ///
,V : k f ; 10~
FCCCIF: adsorption for different bed heights (25 ~ v:
0.369x10-2 m/s).

0.4
~= kf:lO 4 Tile break'through curves for diffelmlt bed heights are similar in stN;e.
--.. kf : 10 4 Since the breakXhrough curve retains a constant pattern, it may be
/
0.2 / interpreted that the adsorption zone has a constant length.
The effect of flow rate on adsc~ptic~l was saflied and the results
-~/'///, are shown in Fig. 8. This figure shows that the breakXhrough time
0,0' ...... i i , J decreased with increasing flow rate, and the breakXhrough curves
O.0 0,5 1,0 1,5 2.0
are slightly steeper for higher flow rates. These trends are consis-
Time (hr) tent with general behavior of a fixed bed adsorbeE Since the in-
Fig. 5. Effect of !k on single-specie breakthrough curves of CI2- traparticle ditSJsivity is usually inde~ndent of flow rate, fffis behav-
FCCCW:. ior is due to the external film mass transfer resistance. This resis-

January, 2002
 Adsorption of Chlorinated VOCs in a FLxed Bed of Activated Carbon 65

1.2 R~ ale Scbmidt and Reynolds ntrnbers, respectively. Molecular dif-

fusion coefficients, D,~, for CFC-113 and D C M are obtained by Fuller
et al. [1969]. Under the experimental conditions of this study, the
1,0
estimated values of ~rdal dispersion coefficient and extmaal film

o
0.8

0.6
tf b raass laansfei coefficient, and raolecular diffusion in the f~xed bed
are listed in Table 5.
Fig. 9 shows file effect of tile flow rate on tile adsolption of DCM
in the fixed bed. As shown in this figure, the effluent concenlration
increases with the flow rate. This is due to the fact that at a higher
flow rate, file contact time between tile adsolbent arid tile gas phase
in the bed is shorter. Fig. 10 illustrates the effect of inlet concel-m'ation
0,4 A Y : 0 r 3 6 9 X 1 0 "2 rll/S
O V : 0,616xl0"~'m/z
'1.2
9 v : 0.890xl0;m/s
0,2 - - Predicted

"LO -~,~ ~ ~, ill i fill _ "2_

0.0 ~ i t
0.0 0.5 1.0 1.5 2.0

Time lhr) 0.8

Fig. 8. E.,cperimental and predicted breakthrough curves of CL-

FCCCIF~ adsorption for different flow rates (25 ~ L: 02 Q
0.6
m). r...)

tance decreased with the flow rate, so that the length of the mass 9 V : O,369xlO'Zi~IS
0,4
transfer zone is reduced, and a sharper breakthrough curve is gen- 0 V : 0.616xlO'Zm/s

el-ate& hi modeling for a packed bed adsorbel, the main paralne- 9 V : O.890xlOar~:s

ters concerning the transport of adsorbates are the axial dispersion 0.2 - - predicted
coefficient and die exteailal fllln mass transfer coefficient [Row and i

Choi, 1990; Markovska et al., 2001]. Usually, the axial dispersion

0.0 ~ ,.0 • #.; ~ , ,
coefficient comes from the contribution of molecular ditfiasion and 0.0 0.5 1.0 1.5 2.0 2.5 3.0
turbulent raixing arising fi-oln the spitting and recombination of flows Time (hr)
around the adsorbent particle. In this study, the ~cdal dispersion co-
efficient, Dz, was calculated from the following correlation given Fig. 9. Experimental and predicted breakthrough curves of CH2CI:
adsorption for different flow rates (25 ~ L: 0.2 m).
by Edwards and Richardson [1968].

1 rzez _+ 1 (13) 1.2

Pe Re Sc Pc' [1+ [ 3 r ~
~'l_ Re ScJ
1.0
The constant values for this calculation are
r,=0.73, [3=13.0, Pe~,=2.0
0.8
The fllln raass transfer coefficient, ks, in a f~xed bed adsober is ob-
tained by Wakao and Funazkri [1978]:

Sh 2k~.Rp 2.0+l.lScZ3ReD0 (14) ~9 1

D~
9 Co : [ 1.51 mol/ms
where, D,, is raolecular diffi_tsivity, S~ is Shelwood numbel; Sc and 04
,2, Co : 14.36:iml/rn~
9 Co : [7.93 mol/m ~
Table 5. Mass transfer parameters for a fkxed bed model simu- -- predicted
lation
0.2
.!
Paralneters Symbols CLFCCC1F~ CH,CL
~ I
(unit) 00= s c=~L ~ - = =-~ , , ,
0.0 0,5 4.0 1.5 2.0 2.5 3.0
A ~ a l dispersion coefficient D L (m2/s) 4.20 9 10-5 4.43 9 10-~ Time (hr)
Filraraetsstrarlsfercoefficient k:,(m/s) 1.56. 10 ~ 2.10. 10 3
Surface diffusivity D: (lll'-/s) 1.10 9 10-~ 1.50 9 10-9 Fig. 10. Experimental and predicted breakthrough curves of CH.~-
CL. adsorption for different inlet concentrations (L: 0.15
Moleculardiffusivity D,, (m~/s) 1.15 9 10 ~ 1.56. 10
m, v: 0.369x10 -2 m/s).

Korean J. Chem. Eng~(Vol. 19, No. 1)

66 S.J. Kim et al.

on experimental breakthiough curves for DC2VI. The breakthrough S : dimensionless axial distance
time decreases with the increase of inlet concenlratiorL The result t :time [hr]
con be explained by the concept of the mass transfer zone IMTZI T : temperature [K]
velocity [Ruthven, 1984]. The velocity of MTZ is a function of in- v : superficial velocity [m/s]
terstitial velocity, particle density, bed porosity, and the slope of the z : axial distance coordinate [m]
equilibiium isotheml. For a linear isotherm adsolption system, the
velocity of MTZ is constant Therefore the break~arough time is Greek Letters
not affected by inlet conceIm'ations at constant MTZ velocity. How- c(, : dimensionless variable [-]
eve1, the adsolption isotheml of CFC-113 on activated c~-bon is [3 :dimensionless [-]
favorable as shown in Fig. 2. As the inlet concentration increases, : bed void fraction [-]
the value of slope of the equilibiiunl isotheml decreases and the : dimensionless concentration [-]
MTZ velocity increases. Therefore, the breaktt~ough time becomes 9p 9particle density [kg/m ~]
shorter under this circumstance. "c : dimensionless time [-]

CONCLUSIONS REFERENCES

Adsorption capacities of solvent vapors such as 1,1,2-11ichloro- Chitin, K., Mellot, C. F., Cheetham, A. K., Harms, S., Mangyo, H.,
1,2,2-trifluoroethane and dichloromethane on an activated carbon C~note, M. and I~lfiyama, R., "Molecular SinMation for Adsolp-
pellet (Ncdt B4 ) were measured exFelmlentally at various telnper- tion of ClflolJalatedHych-oc~rbonin Zolites~"I<_'oteanJ Chem. Eng.,
atures. Single-species equilibrium data show typical type ~ isothemas. I7, 649 (2001).
The Sips equation gave the best fit for the single-component ad- Cho, S. Y and Choi, D. K, "'L~lgmuir Parameters for Adsolpfion of
sciption of 1,1,2-tiichlcio-l,22-tiifluoroe~ioneand ctidfloiometheaie Two Halogellated Chemicals oi1an Activated Carbon Pellet~'Kotean
on activated carbon. The adsorption affinity on activated carbon J. Chem. Eng,. 13, 409 (1996).
was greater for dichloionlethone than that of 1,1,2-tlichloro-l,Z2- Cicerone, R. J., Stolars!d, R. S. and Waiters, S., "'Stratospheric Ozone
trifluoroethane. Adsciption behavior in the fixed-bed was saldied Destruction by Man Made Chlorofluoromethanes;" Science, 185,
in terms of flow rates, heights of the bed, and concentrations of the 1165 (1974).
adsci-bates. The adsorption processes of 1,1,2-tiichloro-l,2,2-tiiflu- EdwNds, M. E and Richa-dsoll, J. E, "'Gas Dispersion in Parked Beds7
oroethane and dichloromethane vapors on activated carbon were Chem. Eng. Sci., 23, 109 (1968).
found to be cciilrolled by intl-aparticle mass ll-ansfea-,and surface HE H. B., Yao, S. J., Zhu, Z. Q. and Hur, B. IC, "'Comparison of the Ad-
diffusion was a dominant intvalzo_rticlemass ti-,msfer mechanism. soiption Ch~-actelistics of Expanded Bed Adsorbent with Conven-
Adsorption dynamics of 1,1,2-trichloro-l,2,2-trifluoroethane and tional Chromatographic Adsorbent7 L'o~eanZ C'hem.Eng., 18, 357
dichloromethone in the frxed-bed can be silnulated by the LDFA (2001).
(Lineor Driving Force Ap1~-oximation 1. Hwang, K. S., "'Separation of HJCO,, H,/CO Gas Mkx~res by Pres-
sure Swing Adsorption and Fixed-Bed Dynamics;" Ph.D disserta-
ACKNOWLEDGEMENT tiol~, KAIST (1994).
Jang, B. H., Lee, S. S., Yeon, q2 H. and Tie, % E., "'Effects of Base Metal
Tiffs research was financially supported by the Korea Science Promoters in VOC Catalysts for Chlorocarbons and n-Hexane Oxi-
and Engineering Foundation (No. 98-1111-2066-2 ). datioi~"K'orean Z Chem. Eng., 15, 516 (2001).
Lee, D. H. and Mool~ H., "'Adsorption Equilibrium of Heavy Metals
NOMENCLATURE on Natural Zeolites~"Korean Z Chem. Eng., 18, 625 (2001).
Markovska, L., Meshko, V and Noveski, x~, "'Adsorption of Basic Dyes
A, : dynamic condition of component i in a Fixed Bed Columl~" k'orean Z Chem. Eng., 18, 190 (2001).
B, : dynamic condition of component i Moon, H. and Tern, C., -Further Work on Multicompone~lt Adsolption
c, : bulk phase concentration of component i [mol/m s] Equilibria Calculations Based on the Ideal Adsorbed Solution The-
c,' 9 equilibrium concentration of component i [mol/m 3] Oly:'Ind Eng. Chem. Res., 26, 2042 (1987).
c, 9 adsorbed concentration on solid phase [moFm 3] Na, B. K., Koo, K. tC, Eurr~ H. M., Lee, H. and Song, H. K., "'CO~ Re-
DL : axial dispersion coefficient [m/s] covery from Flue Gas by PSA Process using Activated CarborC
D, : surface diffusion coefficient [m-'/s] Kot~anZ Chem. t~ng., 18, 220 (2001).
kt : film mass transfer coefficient [m/s] 9 I~, Kobayashi, V~, Migita, C. % and Kakum, K., "'Complete Pho-
k~, 9Henry's constant [mol/kg kpa] todecomposition ofCFC-113, CHCI~ and CC14 and Scavenging of
k, : LDF mass transfer coefficient [1/s] Generated Reactive SpeciesTt~m'imn. Technol., 13, 81 (1992).
L : bed length [m] Park~ I. S., Kwak~ C. and Hwang, Y G., "'Frequency Response of Ad-
p : pressure [kpa] soiption of a Gas onto Bidisperse Pore-Stuctured Soild with Modu-
p~ : Peclet number lation of Inlet Molar Flow-Rate~" L'o~ean Z Chem. Eng., 18, 330
q, : adsorbed amount of component i at equilibrium [mol/kg] (2001).
: adsorbed amount of component i at pure state [inol/kg] Row, K. H., Choi, D. K. and Lee, Y u -Sinmlation of the Conthined
r. :particle radius [m] Continuous and Preparative Separation of Ttwee Close-Bdfling Corn-

January, 2002
 Adsorption of Chlorinated VOCs in a FLxedBed of Activated Carbon 67

ponents in Gas-Liquid Chromaography:"Ko~ean ~ Chem. Eng., 7,
287 (1990).
Ruthven, D. M., "'prindples of Adsorption and Adsorption Processes:"
Wiley, New York, U.S.A. ( 1995 ).
Tiei~ C., "'Adsorption Calculations and Modeling:" Butterwolths-Hei-
nemann (1994).
Wang, S. C. and Tiell, C., "'Furthtx-Work Oll MulEcomponellt Liquid
Phase Adsorption in Fixed Beds:'AIChg ,t, 28, 565 (1982).
V~okao, N. and Funazl~i, T., "'Effect of Fluid Dispersion Coeflid~lt on
Particle to Fluid Mass Transfer Coefficients in Packed Bed~"o( of
Chem. Sci., 33, 1375 (1978).
Wolf, K., "'Ozone Depletion and the use of CFCs:" ~ Em'in)n. Sci., 32,
41 (1992).

Korean J. Chem. Eng.(VoL 19, No. 1)