Understanding competitive adsorption of water and trichloroethylene in a high-silica Y zeolite

Understanding competitive adsorption of waterand trichloroethylene in a high-silica Y zeolite

James Farrell *, Chris Manspeaker, Jing Luo

Department of Chemical and Environmental Engineering, University of Arizona, Harsbarger Room 108, Tucson, AZ 85721, USA

Received 5 August 2002; received in revised form 24 February 2003; accepted 26 February 2003

Abstract

In this study, the adsorption of trichloroethylene (TCE) and water was investigated on a hydrophobic Y zeolite with

an Si to Al ratio of 80. Single adsorbate isotherms for water or TCE, and TCE isotherms under conditions of 100%

relative humidity (RH) were measured over the temperature range from 5 to 47 �C. Water adsorption isotherms were

well described by the Freundlich isotherm model with isotherm exponents of 1.5. Isosteric heats for water adsorption

were less exothermic than the enthalpy for water condensation, and ranged from )34 to )39 kJ/mol. Entropy changesassociated with water adsorption were less negative than those for condensation of bulk water, indicating that the

adsorbed phase had less structure than bulk water. Type V isotherms were observed for TCE adsorption under con-

ditions of 0% RH. Isosteric heats for TCE adsorption on the dry zeolite ranged from )40 to )56 kJ/mol, and showed

regions where the heats of adsorption both increased and decreased with increasing TCE loading. The Henry�s lawasymptote for TCE adsorption on the dry zeolite was not experimentally accessible at the lowest vapor concentrations

investigated. Type V isotherms with a linear region were observed for TCE adsorption under conditions of 100% RH.

The isosteric heats for TCE adsorption on the wet zeolite depended strongly on the adsorbed phase concentration, and

ranged from )35 to )64 kJ/mol. At adsorbed phase TCE concentrations below 0.01 g/g, the presence of water increased

TCE uptake by the zeolite. However, at all other TCE loadings the presence of water decreased TCE adsorption by up

to 83%.

� 2003 Elsevier Science Inc. All rights reserved.

Keywords: HY; DAY; Zeolite; TCE; Trichloroethylene; Water; Humidity; Adsorption; Hydrophobic zeolites

1. Introduction

There is currently a paucity of information

regarding adsorption of halocarbons by dealumi-

nated Y (DAY) zeolites from gas streams at high-

relative humidity (RH). Most studies have been

conducted at RHs below 50%, or have investigated

only the effect of water on the maximum uptake at

high-halocarbon concentrations [1,2]. Other stud-

ies measuring halocarbon vapor adsorption at

100% RH, or from aqueous solution, have been

conducted only at very low-halocarbon concen-

trations [3,4]. One study measured halocarbon

adsorption by silicalite-1 and DAY (Si/Al¼ 75–90)

from both aqueous solutions and from the vaporphase at 100% RH [3]. Halocarbon uptake by sil-

icalite-1 from the aqueous phase was only slightly

less than that from the vapor phase. However, the

*Corresponding author. Fax: +1-520-6216048.

E-mail address: [email protected] (J. Farrell).

www.elsevier.com/locate/micromeso

Microporous and Mesoporous Materials 59 (2003) 205–214

1387-1811/03/$ - see front matter � 2003 Elsevier Science Inc. All rights reserved.

doi:10.1016/S1387-1811(03)00315-9

mail to: [email protected]

presence of liquid water decreased the halocarbon

uptake by DAY by up to 95% compared to that

observed from the vapor phase at 100% RH. These

results agreed with those of Anderson [4] who

observed high uptakes from solution of chloro-

form and trichloroethylene (TCE) by ZSM-5 (Si/Al¼ 1000), but negligible uptake of these com-

pounds by DAY (Si/Al¼ 75).

The different behavior of pentasil and DAY

zeolites with regard to competitive adsorption of

water and halocarbons has been the topic of a

recent investigation [5]. The small inhibition of

halocarbon adsorption by liquid water on pentasil

zeolites was attributed to the small pore size in-hibiting the formation of a liquid-like water phase

inside the pores [5]. In contrast, the strong inhi-

bition of halocarbon adsorption by water on DAY

was attributed to the wider pores allowing for-

mation of a liquid-like water phase. A more

thorough understanding of competitive adsorption

of water and halocarbons on DAY could be

gained from a thermodynamic analysis of waterand halocarbon adsorption.

Although isosteric heats of adsorption have

been measured for halocarbons on dry DAY zeo-

lites, there has yet to be a study measuring isosteric

heats for water adsorption, or for halocarbon ad-

sorption at high RH. The shortage of good ex-

perimental data on water adsorption on DAY

zeolites may in part be attributed to the interferingeffects of binders. The one study presenting water

adsorption isotherms on DAY attributed much

of the uptake to the binder [2]. Clay and silica

binders are often used to aggregate zeolite crys-

tals into pellets [6]. These materials are much more

hydrophilic than high-silica zeolites, and may

therefore interfere with water adsorption mea-

surements.This research investigated single-adsorbate ad-

sorption of water and TCE on a DAY zeolite, and

also investigated the adsorption of water and TCE

mixtures at 100% RH. Single adsorbate experi-

ments were performed in order to characterize

the thermodynamics of water and TCE interac-

tion with the DAY surfaces. The primary goal of

this study was to understand competitive adsorp-tion between water and TCE on DAY over a wide

range in TCE concentrations.

2. Experimental

All experiments were performed on the Y zeolite

CBV780 [7] obtained from Zeolyst (Valley Forge,

PA). The zeolite was used as received and had asilicon to aluminum ratio of 80 and an exchange

cation of Hþ. The zeolite powder contained no

binder and was composed of 1–2 lm diameter

crystals. The experiments were performed in 8 cm

long by 0.9 cm inner diameter stainless steel col-

umns fitted with 0.5 lm stainless steel frits at each

end. The columns were loosely packed with 0.25–

0.5 g of powdered adsorbent, and were thenpurged for one day with dry ultra-high-purity ni-

trogen gas at 110 �C in order to remove any

compounds adsorbed during the packing process.

Water adsorption isotherms were determined

gravimetrically by measuring the increase in weight

of the columns as a function of the water vapor

concentration. Columns filled with the oven-dried

adsorbents were weighed to the nearest 10�5 g usingan Ohaus analytical microbalance, and were then

purged with humidified nitrogen gas at RH ranging

from 2% to 100%. The RH values in the purge

stream were controlled by combining gas streams

at 0% and 100% RH. The 100% RH gas stream

was generated by purging dry nitrogen at flow

rates between 1 and 15 mL/min through a series

of three 1.0 L gas washing bottles each contain-ing 500 mL of deionized water. The total flow

rate through the columns ranged between 5 and

15 mL/min, and was controlled by mass flow

controllers. Each column was purged with the

humidified gas stream for three to ten days until a

constant weight was achieved. Each experiment

was performed in duplicate in a thermostated water

bath or oven at temperatures of 5, 15, 20, 35 and47 �C.TCE breakthrough fronts were measured in

columns packed with the oven dried adsorbent at

0% RH, and on zeolites that were pre-equilibrated

with a 100% RH nitrogen stream. For the experi-

ments conducted at 0% RH, a drying column was

installed on the nitrogen carrier gas, and the liquid

TCE (Aldrich) of 99+% purity was filtered usingsilica gel (Davisil) in order to remove any trace

amounts of water. Purge streams containing TCE

were generated by passing dry or humidified ni-

206 J. Farrell et al. / Microporous and Mesoporous Materials 59 (2003) 205–214

trogen gas through a series of three gas washing

bottles containing dry TCE or TCE plus water.

The gas saturators were in the same water bath or

thermostated oven as the adsorbent columns.

Breakthrough flow rates between 5 and 20 mL/min

were used and were controlled with mass flowcontrollers. TCE concentrations in the column

effluent were measured using a flame ionization

detector (FID) on an SRI Instruments gas chro-

matograph. The mass of TCE adsorbed on the

zeolite was determined from integration of the

breakthrough fronts.

Isotherms for TCE desorption from the zeolites

were measured using a continuous purge technique[8]. After measuring a TCE breakthrough front,

the purge gas was switched to clean nitrogen at 0%

or 100% RH. Effluent TCE concentrations from

the column were then monitored over time using

the FID. The effluent vapor concentrations (Cv,) as

a function of time (t) were integrated to obtain the

adsorbed phase TCE concentrations (Cs) accord-

ing to:

Cs ¼ ðf =mÞZ t

t0

CvðtÞdt ð1Þ

where f is the volumetric flow rate of the purge

gas, m is the mass of adsorbent in the column, andt0 is the time required for the effluent vapor con-

centration to return to zero. A sampling frequency

of 1 Hz resulted in high-resolution isotherms com-

posed of thousands of data points. This method is

capable of determining a complete isotherm from a

diffuse front under the conditions that the shape

of the front is dominated by the shape of the

isotherm and not by diffusion or mass transferlimitations [8].

Measurement of a single isotherm using the

purge technique required 10–20 h. Mass balance

errors between adsorption and desorption were

generally between 5% and 10%. Runs with greater

mass balance errors were repeated. Heating the

columns after termination of the purge experi-

ments showed that all the TCE had been removedby the time the detector signal had returned to its

baseline value. This indicates that there was no

irreversible adsorption. Experiments were also re-

peated at different desorption flow rates between 5

and 10 mL/min in order to determine if the effluent

TCE concentrations were at equilibrium with the

adsorbent. Differences in isotherms measured at 5

and 10 mL/min were similar to differences in rep-

licate runs at the same flow rate [9], as shown in

Fig. 1. Additionally, several isotherms determinedfrom integration of the breakthrough fronts at dif-

ferent feed concentrations agreed with those de-

termined using the continuous desorption method

[9]. This indicates that there was no significant

hysteresis between adsorption and desorption, and

that mass transfer limitations had a negligible ef-

fect on the isotherm measurements. The absence of

hysteresis for TCE sorption on DAY zeolites hasbeen previously reported [12].

Y-type zeolites consist of roughly spherical 12.3�AA diameter supercages that are connected through

windows that are 7.3 �AA in diameter [10]. The su-

percages are constructed from 8 b-cages (sodaliteunits) that contain �16% of the total void volume

[6]. Water is able to penetrate both the super- and

b-cages, and is therefore capable of occupying thefull void volume of the zeolite, which is 0.38 mL/g

for high-silica HY [6]. Larger adsorbates such as

TCE cannot penetrate the b-cages, and therefore

can occupy only a fraction of the void volume.

TCE can be considered a planar molecule with van

der Waals radii based dimensions of 6.8 by 6.5 �AA[11]. The window sites in DAY are therefore wide

enough to accommodate only a single TCE mol-ecule, while each supercage can accommodate

slightly more than four TCE molecules [12]. Given

that the molar volume of hydrocarbons in zeolites

0

0.1

0.2

0.3

0.0 0.2 0.4 0.6 0.8 1.0

Relative Vapor Saturation (P/P˚)

Cs

(g/g

)

Test C

Test B Test A

Fig. 1. Flow rate and repeatability tests conducted at 35 �C.Tests A and B were conducted at a flow rate of 10 mL/min and

test C was conducted at a flow rate of 5 mL/min.

J. Farrell et al. / Microporous and Mesoporous Materials 59 (2003) 205–214 207

are often similar to their liquid molar volumes

[3,13,14], the theoretical maximum TCE adsorp-

tion capacity for DAY is 0.47 g/g.

3. Results and discussion

3.1. Water isotherms

Isotherms for water adsorption on the HY zeolite

are shown in Fig. 2. Differences in water uptake

between duplicate columns were within �10%. Theincreasing slopes with increasing vapor phase con-

centrations classify them as type III isotherms in theIUPAC classification system [15]. Type III iso-

therms for water adsorption have also been ob-

served on other DAY zeolites [2]. Although

saturation of the zeolites was not observed in

the isotherm experiments conducted at RH up to

90–95%, saturation was observed in experiments

measuring the rate of water uptake at 100% RH.

Columns purged with water vapor at 100% RH at30 �C showed 0.46� 0.02 g/g of water uptake after

40 h elapsed, and showed no further uptake after an

additional 180 h. This water loading was close to

the 0.47 g/g observed by Giaya and Thompson [5]

on this same zeolite, but is greater than the maxi-

mum theoretical uptake of 0.38 g/g in the intra-

crystalline pores. This suggests that there was

significant adsorption on the external surfaces ofthe zeolite, most likely at silanol sites on the crystal

surfaces [5]. The kinetic experiments also showed

that the rates of water adsorption slowed down

dramatically after the zeolites had adsorbed 0.39

mL/g of water. This volume is close to the intra-

crystalline pore volume, and suggests that filling of

intracrystalline pores occurred prior to significant

extracrystalline adsorption.

As indicated by the good fit of the data to the

solid lines in Fig. 2, the water adsorption data was

well described by the Freundlich isotherm model,as given by:

Cs ¼ KFCnv ð2Þ

where Cs is the adsorbed phase concentration in g-

adsorbate/g-zeolite, Cv is the vapor phase concen-

tration in g/mL, and KF and n are the Freundlich

capacity coefficient and exponent, respectively.Table 1 summarizes the Freundlich isotherm pa-

rameters. The n values greater than 1 indicate that

water–water attractions were stronger than water-

adsorbent attractions.

The water adsorption data were also fit to the

Dubinin–Ashtakov (DA) model [16] for micropore

filling, as given by:

W ¼ W0 exp

�� A

E

� �n�ð3Þ

where W is the volume adsorbed per unit mass of

adsorbent, W0 is the micropore volume per unit

mass of adsorbent, E is a characteristic energy of

adsorption, n is a heterogeneity parameter, and

A ¼ RT lnðP 0=P Þ is the molar work of adsorption

[17]. The molar work of adsorption is the differ-

ence in Gibbs energy between the adsorbate as an

ideal gas at a vapor pressure of P , and the ideal gasadsorbate at its saturation vapor pressure, P�. Asshown in Fig. 3, the experimental data were rea-

sonably well-fit by the DA model, with a v2 valueof 0.33. Chi-squared values less than unity are

generally indicative of a good fit [18]. Fitting the

model to the experimental data yielded an E pa-

rameter of 1503 J/mol and an exponent of 0.87.

0

0.1

0.2

0.3

0.4

0.5

0 10 20 30 40 50Cv ( g/mL)

Cs

(g/g

)

5˚C

10˚C

20˚C

35˚C

47˚C

µ

Fig. 2. Water adsorption isotherms on the HY zeolite. Solid

lines were generated using the Fruendlich isotherm parameters

given in Table 1.

Table 1

Freundlich isotherm parameters for water adsorption

Temperature (�C) LogðKFÞ n

5 7.51 1.53

10 7.34 1.53

20 7.10 1.56

35 7.12 1.58

47 6.04 1.50


Isosteric heats (Qiso) for water adsorption can be

determined from the data in Fig. 1 according to:

Qiso ¼ Rd lnCv

dT�1

� �Cs

ð4Þ

where R is the gas constant, and T is the temper-

ature. The Qiso values for water adsorption are

shown in Fig. 4. For all water loadings the heats of

adsorption were less than the enthalpy for water

condensation, which is )44.2 kJ/mol at 20 �C [19].

The hydrophobicity index proposed by Olson et al.[20] indicates that this HY zeolite is more hydro-

phobic than high-silica pentasil zeolites. Olson

et al. use the fraction of the micropore capacity not

occupied by adsorbed water at 25% RH as an in-

dicator of zeolite hydrophobicity. For this HY at

20 �C, 13% of the micropore capacity was occu-

pied by adsorbed water at 25% RH, yielding a

hydrophobicity of 87%. This compares to a hy-drophobicity of 75% for a HZSM-5 with an Si/Al

ratio of 75 [20]. Less water vapor adsorption for

DAY compared to pentasil zeolites has been at-

tributed to greater fluid-pore attractions in the

smaller pore pentasil zeolites [5].

Isosteric heats of adsorption can also be calcu-lated from the DA model according to [17]:

Qiso ¼ �DH vap � E lnW0

W

� �1=nð5Þ

where DH vap is the enthalpy of vaporization for the

adsorbate. In Fig. 4 the Qiso values calculated from

Eq. (5) are compared to those calculated using Eq.(4). The Qiso values from the DA model consis-

tently overestimate the magnitude of the true

isosteric heats of adsorption. This disparity in Qiso

values indicates that the assumptions in the DA

model [17] are not appropriate for water adsorp-

tion on this zeolite.

Analysis of the Gibbs energy change accompa-

nying the adsorption process indicates that theadsorbed water phase had a greater entropy than a

bulk liquid phase of water. The difference in Gibbs

energies for a water molecule in the vapor and

adsorbed phases can be determined from the

concentrations of water in both phases from [21]:

DG ¼ �RT lnCv

Cs=W0

� �ð6Þ

and the entropy change associated with adsorption

can then be determined from [22]:

DS ¼ Qiso � DGT

ð7Þ

The entropy differences between a water molecule

in the adsorbed and vapor phases are shown in

Fig. 5 for the 20 �C data. For all adsorbed phaseloadings the entropy loss upon adsorption was less

than the entropy loss for water condensation,

which is 59.2 J/molK at 20 �C. This contrasts withthe behavior in hydrophilic micropores where the

entropy loss associated with water adsorption may

be several times greater in magnitude than that for

water condensation [23]. The smaller entropy los-

ses observed for adsorption on HY are consistentwith the weak interactions of water with the hy-

drophobic zeolite. The greater entropy of adsorbed

-55

-50

-45

-40

-35

-30

0 0.1 0.2 0.3

Isotherms

DA model

0.4 0.5

Cs (g/g)

Qis

o(k

J/m

ol)

Fig. 4. Isosteric heats (Qiso) for water adsorption determined

from the adsorption isotherms using Eq. (4), and from the DA

parameters using Eq. (5).

0.01

0.1

1

0 1 2 3 4 5RTln(Po/P))/E

W/W

o

DA Parametersn = 0.87E = 1.5 kJ/mol

Fig. 3. Water adsorption isotherms plotted in DA format.


water as compared to bulk water may be attrib-

uted to confinement resulting in a less ordered

structure for adsorbed water than for bulk water.This may arise from molecular-scale packing ef-

fects interfering with hydrogen bonding between

adjacent water molecules.

3.2. Dry TCE isotherms

TCE desorption isotherms measured at 0% RH

are shown in Fig. 6a and b. Each isotherm con-tained two inflection points, most dramatically il-

lustrated in the 10 �C isotherm. The inflection

point that occurs in the 10 �C isotherm near 0.41 g/

g is close to the maximum theoretical capacity of

HY for TCE of 0.47 g/g, and close to the 0.45 g/g

maximum TCE uptake that was reported for HY

in a previous investigation [24]. Therefore, this

inflection point likely represents the transitionfrom primarily intracrystalline adsorption to ad-

sorption on external surfaces and in intercrystal-

line pores. Between TCE loadings of 0.33 to 0.41 g/

g, the 10 �C isotherm is concave upward. The in-

flection point occurring at 0.33 g/g adsorbed likely

arises from increasingly favorable TCE–TCE in-

teractions as the micropores fill. Two repeat ex-

periments at 10 �C also showed similar behavior.However, in the repeat experiments the sharp in-

crease in slope due to pore filling was more gradual

and occurred over TCE loadings from 0.25 to 0.41

g/g. The 5 �C isotherm shows inflection points at

TCE loadings of 0.37 and 0.45 g/g. The upward

concavity over a loading range of 0.8 g/g corre-

sponds to approximately one TCE molecule per

supercage [12]. The 20 �C isotherm was concave

upward over a similar loading range from 0.34 to

0.42 g/g.

The experimental technique used to measure

TCE adsorption on the dry zeolite was not able to

obtain any points in the linear region of the iso-

therms. When viewed on a logarithmic scale, the

TCE isotherms in Fig. 6b show two distinct re-gions. The isotherms are nearly vertical for TCE

loadings below 0.01 g/g. This behavior is consis-

tent with the volume filling of micropores theory

proposed by Polanyi [25] and Dubinin et al. [26] in

that a minimum threshold vapor concentration

must be reached before there is any significant

amount of TCE adsorption. This same behavior

has been seen in grand canonical Monte Carlosimulations of TCE and methanol adsorption in

slit micropores [23,27].

Although the DA model is based on Polanyi�stheory, it was not capable of describing the dry

TCE isotherms. This can likely be attributed to the

narrow pore size distribution in the HY zeolite.

The DA model assumes that there is a continuous

Weibull distribution of adsorption energies [15].

-55

-45

-35

-25

0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4

Cs (g/g)

∆S

(J/m

ol-K

)

Fig. 5. Entropy changes associated with water adsorption on

zeolite HY at 20 �C.

(a)

(b)

Fig. 6. (a) Linear plot of the TCE isotherms on dry zeolite HY

and (b) logarithmic plot of the TCE isotherms in part (a).


However, the adsorption energies in the HY zeo-

lite are likely to be bimodal and discontinuous due

to the bimodal and discrete pore size distribution.

To get around the distribution continuity problem

associated with adsorption in zeolites, some re-

searchers have divided the zeolite pores into twoindependent regions and applied separate DA

models to each region [16]. The two-region DA

model takes the form:

W ¼ W0;1 exp

�� A

E1

� �n1�

þ W0;2 exp

�� A

E2

� �n2�ð8Þ

where the subscripts 1 and 2 denote, respectively,

parameters for micropore regions 1 and 2. Fig. 7a

shows the fit of the DA model to adsorbed phase

loadings of 0.13 g/g and below. The best-fit het-erogeneity parameter of n1 ¼ 29 is indicative of

very homogeneous adsorption sites. Fig. 7b shows

the fit of the DA model to adsorbed phase loadings

above 0.13 g/g. The best-fit n2 ¼ 2:9 is close to the

range of 3–6 that is commonly observed for zeo-

lites [16]. The very high n1 value of 29 is consistentwith very homogeneous adsorption energies, and

suggests that TCE adsorption at low concentra-tions occurs only in the window sites, since a

window site can accommodate only a single TCE

molecule. The small value of n2 indicates that ad-sorption energies at higher concentrations are

more heterogeneous. This is consistent with ad-

sorption in the supercages, which can accommo-

date approximately 4 TCE molecules [12], and

thus a range of adsorption energies is expected assites in the supercages fill. The E1 ¼ 12 kJ/mol for

the low-concentration data is 74% greater than the

characteristic energy for the high-concentration

data which had E2 ¼ 6:9 kJ/mol. This is consistent

with higher adsorption energies in the smaller di-

ameter window sites as compared to the supercage

sites.

Isosteric heats for TCE adsorption on the dryzeolite are shown in Fig. 8. For all TCE loadings,

the Qiso values were more exothermic than the

enthalpy for TCE condensation, which is )35 kJ/

mol at 25 �C [19]. For loadings between 0 and 0.27

g/g, the Qiso values became increasingly exothermic

with increasing amounts adsorbed. This may be

attributed to increasing TCE–TCE attractions in

the adsorbed phase. At higher loadings above 0.27g/g adsorbed, there was an increasing trend of Qiso

values with TCE loading. This may be attribu-

table to molecular-scale packing restrictions that

Fig. 7. (a) Fit of the two region DA model to the experimental

data for adsorbed TCE concentrations less than 0.13 g/g and (b)

fit of the two region DA model to the experimental data for

adsorbed TCE concentrations greater than 0.13 g/g.

-58

-54

-50

-46

-42

-38

0.001 0.01 0.1 1Cs (g/g)

Qis

o(k

J/m

ol)

Fig. 8. Isosteric heats for TCE adsorption on zeolite HY under

dry conditions.


preclude the most energetically favorable arrange-

ment of the adsorbed molecules. The Qiso values in

Fig. 8 show the same trends and are within �5 kJ/mol of those reported by other investigators for

TCE adsorption on high-silica DAY [12,28].

3.3. Wet TCE isotherms

Isotherms for TCE desorption from the HY ze-

olite at 100% RH are shown in Fig. 9. Neither the

DA model, nor the bimodal DA model, were able

to adequately describe the TCE adsorption behav-

ior on the wet zeolites. The data in Fig. 10a for the

35 �C isotherm illustrates that there were three ad-sorption regimes on the wet zeolites. In the low-

concentration region the isotherms approached

linearity for TCE vapor pressures below 0:001P 0, as

shown in Fig. 10b. For TCE vapor pressures be-

tween 0.005 and 0:01P 0, the slope on the logarith-

mic plot was much greater than 1. This corresponds

to increasing isotherm slopes with increasing con-

centrations, and therefore increasingly favorableTCE adsorption with increasing concentrations.

However, as shown in Fig. 11, the Qiso values for

adsorption in this range (0.002–0.07 g/g adsorbed)

actually became less exothermic with increasing

TCE loadings. This indicates that entropic effects

were responsible for the increasing isotherm slopes

with increasing TCE loadings. This may be due to

entropy of mixing effects that contributed toincreasing configurational entropies with the in-

creasing concentration of a second component in

the micropore phase.

The third adsorption regime occurred at adsorbed

phase concentrations above 0.1 g/g adsorbed. In this

part of the isotherm the slope decreased with in-

creasing TCE concentration. This is normally at-

tributed to occupation of the most energetically

favorable adsorption sites at the lowest concentra-

tions. However, the trend in Qiso values between 0.1

and 0.35 g/g adsorbed show that adsorption becameFig. 9. Adsorption isotherms for TCE on zeolite HY under

conditions of 100% RH.

0.00001

0.0001

0.001

0.01

0.1

1

0.00001 0.0001 0.001 0.01 0.1 1

Cs

(g/g

)

0

0.00005

0.0001

0.00015

0.0002

0.00025

0 0.0002 0.0004 0.0006 0.0008 0.001


Cs

(g/g

)Low Concentration Data

(a)

(b)

Fig. 10. (a) Logarithmic plot of the TCE isotherm at 35 �C at

100% RH and (b) linear plot of the low-concentration region of

the TCE isotherm in part (a).

-70

-60

-50

-40

-30

0.001 0.01 0.1 1Cs (g/g)

Qis

o(k

J/m

ol)

Fig. 11. Isosteric heats for TCE adsorption on zeolite HY at

100% RH.


more enthalpically favorable with increasing TCE

concentrations. Therefore, the decreasing isotherm

slopes with increasing TCE concentrations must be

due to decreasingly favorable entropy changes with

increasing TCE loading. This trend in entropy

changes may also arise from entropy of mixing ef-fects, which for an ideal binary solution reach a

maximum at equimolar concentrations.

The Qiso value of )48.3 kJ/mol at low loadings in

Fig. 11 is close to the Qiso value measured in the

linear region of the isotherm in a previous inves-

tigation of liquid phase TCE adsorption on this

same zeolite [29]. Gas and liquid phase Qiso values

for TCE adsorption should differ by the enthalpychange associated with TCE vapor dissolution into

water, which is )37.3 kJ/mol as determined from

the temperature dependence of the Henry�s law

constant [19]. The Qiso value for aqueous phase

TCE adsorption of )11.5 kJ/mol [29] plus the

)37.3 kJ/mol for the phase transfer yields a Qiso

value for gas phase TCE adsorption of )48.8 kJ/

mol. This indicates that the limiting Qiso value of)48.3 kJ/mol measured in this study is close to thatfor TCE adsorption in a water-filled HY zeolite at

infinite dilution.

Fig. 12a and b compare TCE vapor isotherms at

20 �C measured at 0% and 100% RH. Also shown

in Fig. 12a is a 20 �C isotherm measured for TCE

adsorption from solution on this same zeolite [29].

At high concentrations the presence of watereliminated the concave upward regions of the iso-

therms that occurred under dry conditions at

loadings near 0.35 g/g. At adsorbed phase con-

centrations between 0.01 and 0.25 g/g, the presence

of water greatly diminished TCE adsorption, as

shown in Fig. 12b. However, the presence of water

increased TCE adsorption for concentrations be-

low 0.01 g/g adsorbed. This can be attributed toattractions between TCE and water that promote

TCE adsorption at very low-TCE concentrations.

Water enhancement of TCE adsorption at low

concentrations accompanied by water inhibition of

TCE adsorption at high concentrations has also

been observed in grand canonical Monte Carlo

simulations of TCE adsorption in hydrophobic

silica micropores [23].Comparison of the 100% RH and aqueous

phase isotherms shows that the presence of liquid

water decreased TCE adsorption much more than

vapor phase water at 100% RH. A similar, but

more pronounced, effect has been previously re-

ported for TCE adsorption on DAY at concen-

trations below 3 mg/L [3]. Although both liquid

water and water vapor at 100% RH have the same

chemical potential, surface energy effects make the

chemical potential for water adsorbed in the mi-cropore opening depend on whether there is liquid

water or water vapor outside the micropore. This

effect has recently been explained by Giaya and

Thompson [5] who showed that the pore size and

the strength of the fluid-wall interactions deter-

mine the density of water in adsorbent micropores.

DAY pores were shown to be sufficiently large to

accommodate water with a liquid-like densitywhen liquid water was present outside the pores.

However, when 100% RH water vapor was present

outside the DAY pores, water could not form

a liquid-like phase within the adsorbent. This

0

0.1

0.2

0.3

0.4

0.5

0.6

0 0.2 0.4 0.6 0.8 1

Cs

(g/g

)

100% RH

Dry

Aqueous

0

0.05

0.1

0.15

0.2

0.25

0 0.02 0.04 0.06 0.08 0.1 0.12


Cs

(g/g

)100% RH

Dry (b)

(a)

Fig. 12. (a) TCE vapor adsorption isotherms at 20 �Cmeasured

under conditions of 0% and 100% RH. Also shown is a 20 �Cisotherm for TCE adsorption from aqueous solution. (b) Ex-

panded scale of the low-concentration region of the TCE iso-

therms in part (a).


indicates that the phase external to the micropores

can promote water structuring within the mi-

cropores through interactions at the pore opening.

4. Conclusion

Data from this study showed that water vapor at

100% RH can both increase and decrease TCE

adsorption compared to dry conditions. For ad-

sorbed concentrations below 0.01 g/g adsorbed, the

presence of water increased TCE adsorption due to

attractions between adsorbed water and TCE.

However, at higher TCE loadings water vapordecreased TCE adsorption by up to 83% compared

to dry conditions. The decrease in TCE adsorption

due to water vapor at 100% RH was less than the

decrease due to bulk water. The Qiso values for

water adsorption were less exothermic than the

enthalpies for water condensation, and the entropy

of the adsorbed water phase was greater than that

for bulk water. The Qiso values for TCE adsorptionunder both wet and dry conditions were more

exothermic than those for TCE condensation. On

the wet zeolite, entropy of mixing effects appeared

to enhance the favorability for TCE adsorption for

loadings less than 0.07 g/g, and hinder TCE ad-

sorption for loadings between 0.1 and 0.35 g/g.

Acknowledgements

This work was supported by the National Sci-

ence Foundation Chemical and Transport Systems

Program through Grant Number CTS-962472.

References

[1] R. Schumacher, S. Ernst, J. Weitkamp, in: R. von

Ballmoos, J.B. Higgins, M.M.J. Treacy (Eds.), Separation

of Gaseous Tetrachloroethene/Water Mixtures by Adsorp-

tion on Zeolites, Proceedings from the Ninth International

Zeolite Conference, vol. 2, Butterworth-Heinemann, Mon-

treal, 1992, p. 79.

[2] G. Weber, O. Bertrand, E. Fremont, S. Bourg, F. Bouvier,

D. Bissinger, M.H. Simonot-Grange, J. Chim. Phys. 93

(1996) 1412.

[3] A. Giaya, R.W. Thompson, R. Denkewicz Jr., Micropor.

Mesopor. Mater. 40 (2000) 205.

[4] M.A. Anderson, Environ. Sci. Technol. 34 (2000) 725.

[5] A. Giaya, R.W. Thompson, J. Chem. Phys. 117 (2002)

3464.

[6] D.M. Ruthven, Principles of Adsorption and Adsorption

Processes, John Wiley-Interscience, New York, 1984, pp.

9–20.

[7] http://www.Zeolyst.com.

[8] C. Burgisser, M. Cernik, M. Borkovec, H. Sticher, Envi-

ron. Sci. Technol. 27 (1993) 943.

[9] C. Manspeaker, Master�s Thesis, University of Arizona,

2003.

[10] J.F. Denayer, A. Bouyarmaouen, G.V. Baron, Ind. Eng.

Chem. Res. 37 (1998) 3691.

[11] J. Farrell, J. Luo, P. Blowers, J. Curry, Environ. Sci.

Technol. 36 (2002) 1524.

[12] B. Clausse, B. Garrot, C. Cornier, C. Paulin, M.H.

Simonot-Grange, F. Boutros, Micropor. Mesopor. Mater.

25 (1998) 169.

[13] D.W. Breck, Zeolite Molecular Sieves: Structure, Chemis-

try, and Use, John Wiley-Interscience, New York, 1974,

pp. 628–699.

[14] X.S. Zhao, Q. Ma, G.Q. Lu, Energy Fuels 12 (1998) 1051.

[15] K.S.W. Sing, D.H. Everett, R.A.W. Haul, L. Moscou,

R.A. Pierotti, J. Rouquerol, T. Siemieniewska, IUPAC

Pure Appl. Chem. 57 (1985) 603.

[16] M.M. Dubinin, V.A. Ashtakov, Bull. Acad. Sci. USSR 20

(1971) 3.

[17] D.D. Do, Adsorption Analysis: Equilibria and Kinetics,

Imperial College Press, London, 1998, pp. 159–170.

[18] I. Miller, J.E. Freund, Probability and Statistics for

Engineers, third ed., Prentice Hall, Englewood Cliffs, NJ,

1985.

[19] http://webbook.nist.gov/chemistry/.

[20] D.H. Olson, W.O. Haag, W.S. Borghard, Micropor.


[21] R.P. Schwarzenback, P.M. Gschewend, D.M. Imboden,

Environmental Organic Chemistry, John Wiley and Sons,

New York, 1993, pp. 524–525.

[22] G.V. Tsitsishvili, T.G. Andronikashvili, S.H.D. Sabelash-

vili, T.A. Chumburidze, Natural Zeolites, E. Horwood

Publishing, New York, 1992, pp. 71–84.

[23] J. Luo, J. Farrell, Environ. Sci. Technol. 37 (2003) in press.

[24] J. Pires, A. Carvalho, M.B. de Carvalho, Micropor.


[25] M. Polanyi, Trans. Faraday Soc. 28 (1932) 316.

[26] M.M. Dubinin, E.D. Zaverina, L.V. Radushkevich, Zh.

Fiz. Khim. 21 (1947) 1351.

[27] A.V. Shevade, S. Jiang, K.E. Gubbins, Mol. Phys. 97

(1999) 1139.

[28] C.F. Mellot, A.K. Cheetham, S. Harms, S. Savitz,

R.J. Gorte, A.L. Myers, J. Am. Chem. Soc. 120 (1998)

5788.

[29] B. Hauck, Masters Thesis, University of Arizona, 1999.


Understanding competitive adsorption of water and trichloroethylene in a high-silica Y zeolite

Documents

Transcript of Understanding competitive adsorption of water and trichloroethylene in a high-silica Y zeolite