United States
Environmental Protection
Agency
Health Effects Research
Laboratory
Research Triangle Park NC 27711
Research and Development
EPA-600/S1-83-001 Mar 1983
4>EPA Project Summary
An Evaluation of the Adsorption
Properties of Silicalite for
Potential Application to Isolating
Polar Low-Molecular-Weight
Organics from Drinking Water
Colin D. Chriswell, Douglas T. Gjerde, Gerda Shultz-Sibbel, James S. Fritz, and
Ikue Ogawa
Isolation is the first step in the
determination of many organic species
in drinking water. An effective isolation
technique is therefore essential to
ascertaining whether or not potentially
harmful species are present in drinking
water. Conventional isolation
techniques yield only low recoveries
when applied to small, water-like
compounds. Because retention' of
components on a molecular sieve is
based primarily on molecular size,
studies were performed to determine if
low-molecular-weight organic
compounds could be isolated from
water matrices by adsorption on a
hydrophobic molecular sieve.
In this work the chemical and physical
properties of the adsorbent known as
Silicalite were explored, the utility of
this molecular sieve for accumulating
analytes from aqueous and gaseous
streams was elucidated, techniques
were developed for recovering
adsorbed components, and an
analytical protocol was developed for
determining low-molecular-weight
compounds such as dichloroacetoni-
trile in standard samples.
It was shown that Silicalite can be
used for the accumulation of a variety of
aldehydes, acids, esters, ethers, alco-
hols, ketones, nitriles, and halogenated
species from water. The accumulated
components can subsequently be
recovered from Silicalite by use of a
simple, convenient, and effective
elution procedure using a water-
methanol gradient as the eluent. Com-
bining accumulation and recovery
techniques into a protocol resulted in
recoveries exceeding 80% for
compounds as varied as phenol, acetic
acid, ethyl acetate, chloroform,
crotonaldehyde, propanal, acetalde-
hyde and butanal which were added to
standard solutions. Optimization of the
protocol for the determination of
dichloroacetonitrile resulted in
essentially quantitative recoveries from
standard solutions. Tests of the
procedure on drinking water samples
from Ames and Ottumwa, Iowa were
inconclusive. Compounds were recov-
ered having chroma tog ra ph ic
properties expected of dichloroaceto-
nitrile, but the levels were below those
allowing confirmation by gas chroma-
tography/mass spectrometry. Even
though the method should work on
drinking water samples when
detectable levels of the above
compounds are present, applicability to
"real samples" has not been
established.
This Project Summary was developed
by EPA's Health Effects Research Lab-
oratory, Research Triangle Park. NC, to
announce key findings of the research
project that is fully documented in a
separate report of the same title (see
Project Report ordering information at
back).
-------
Introduction
Hundreds of different organic
components have been identified in
various drinking water supplies in the
United States. Most of these compounds
are present only at ultra-trace concentra-
tions and are not believed to pose any
threat to human health or to the
environment. Some deleterious
compounds which have been detected
require appropriate control measures to
be taken Obviously before any controls
can be instituted, procedures must be
available for characterizing and
quantitatmg contaminants in drinking
water. Effective techniques are available
for most classes of organic compounds
found in water However, no effective
procedures are generally applicable to
the isolation and concentration of low-
molecular-weight, polar organic
compounds. Compounds such as
aldehydes, ketones, nitnles, alcohols and
esters are too water-like mtheirchemical
and physical properties for effective
accumulation by conventional
procedures, and are present at levels
below which they can be determined
without resorting to accumulation
procedures.
It has been suggested that a
hydrophobic molecular sieve introduced
by Union Carbide Corporation could be
used in the treatment of wastewater for
the removal of components such as
benzene, phenol, propanol and hexane It
was subsequently demonstrated that this
molecular sieve, known as Silicalite*, is
effective for removing chloroform from
drinking water and for recovering ethanol
from fermentation beer.
Based on the ability of Silicalite to
adsorb polar as well as non-polar organic
species of low-molecular weight from
aqueous solutions, studies were made to
determine if this adsorbent could be used
for isolating small, polar organic
compounds from drinking water prior to
their determination.
Chemical and Physical
Properties of Silicalite
Molecular sieves are porous, solid
adsorbents having pores of consistent
diameters in the range of the solution
diameters of molecules. Only molecules
small enough to enter the pore structure
can be retained by molecular sieves.
Interactions between the pore surfaces
and adsorbed compounds determine the
•Mention of trade names or commercial products
does not constitute endorsement or recommendation
for use by the U S Environmental Protection Agency
degree of retention. Conventional
molecular sieves contain metal ions and
hydroxyl groups which interact with polar
materials such as water Thus, molecular
sieves are commonly used for removing
water from organic solvents In contrast
with conventional molecular sieves,
Silicalite contains only silicon and oxygen
and no polar functionalities. It is hydro-
phobic and can be used to accumulate
organic components from aqueous solu-
tion.
The pores in Silicalite are six
Angstroms in diameter. Molecules
approximately the size of benzene or
smaller can enter the pores and be
retained. Linear molecules much longer
than the six Angstrom pore diameter can
enter the pore structure so long as they
can assume a conformation such that
their diameter in one direction is smaller
than six angstroms.
Silicalite is a polymorph of silica and
has properties similar to those of quartz. It
is stable in the presence of most corrosive
agents except for strong bases and
hydrofluoric acid. It is unaffected by
solvents. Silicalite is stable at tempera-
tures in excess of 1000°C. At about
1300°C it reverts to amorphous silica.
Silicalite is produced as a fine powder
with particle sizes of about 20 microns
diameter A binder is used to agglomerate
these particles into granules of about 20
to 80 mesh. Various silicate and alumina-
silicate clays have been used by Union
Carbide as binders. A material designated
as LZ-115 which contains 10% of an
alumina-silicate clay binder was used m
this work.
Adsorption of Analytes
from Water
The primary requirement of an
adsorbent is having an affinity for
components of interest. The distribution
coefficient is a measure of the relative
affinity of compounds for an adsorbent.
The distribution coefficient Dg, is the
ratio of the concentration of a species of
an adsorbent to the concentration of the
same species in water at equilibrium.
Distribution coefficients for typical low-
molecular-weight organic compounds
(Table 1), show a general trend that the
most polar compounds such as acetic
acid have the lowest distribution coeffi-
cients This is to be expected because
such compounds have a high affinity for
water. Within a homologous series such
as the aldehydes there appears to be an
optimum chain length leading to the
highest distribution coefficient. Thus,
pentanal has a much higher distribution
coefficient than does acetaldehyde which •
is more polar, and pentanal also has a
much higher distribution coefficient than
does decanal which must assume a linear
conformation to enter the pore structure
The breakthrough capacity of an
adsorbent is the amount of material that
is adsorbed before the bed effluent
reaches a certain percentage of the
influent concentration. Table 2 contains
data on the 1%, 10% and 50%
breakthrough capacity of Silicalite for
selected compounds The 1% and 10%
breakthrough capacities provide an
indication of the amount of material that
can be accumulated on a bed while
retaining 99% and 90% respectively of
the analyte. The value for 50% break-
through is of more interest in-water
treatment applications than for analysis.
In actual use excess capacity must be
provided because a minimum bed depth
is required for contact, and this minimum
is dependent on factors such as flow rate,
analyte concentration, and particle size.
Desorption of Materials from
Silicalite
If an adsorption technique is to be
useful as part of an analytical protocol,
the adsorbed components must be
recovered from the adsorbent in a form
amenable to their subsequent determin-
ation Solvent elution, Soxhlet extraction,
high pressure Soxhlet extraction,
adsorbent dissolution, microwave
desorption and thermal desorption were
investigated as potential techniques for
recovering adsorbed species from
Silicalite Of these techniques solvent
elution was found to be the most conven-
ient and applicable to the widest range of
components.
The elution technique developec
consists of using a gradient going frorr
100% water to 100% methanol in aboul
15 minutes. During elution, analytes are
generally completely retained until the
methanol concentration reaches 100%
Despite the fact that analytes are no
eluted until methanol concentrations
reach 100%, the use of a gradient i;
critical. This is because the pores ir
Silicalite are initially filled with air
During adsorption of components frorr
water, some of this air is replaced by th<
accumulated components, but at the en<
of a typical adsorption cycle the pores stil
contain about 5mL of air per gram o
Silicalite. This air is displaced durini
elution by methanol When gradien
elution is used, the air will be displace^
slowly and will dissolve in the wate
which comprises most of the initie
-------
Table 1. Distribution Coefficients Between Silicalite and Water
Compound Dg Compound
Acetic acid
Pyruvic acid
Tnchloroacet/c acid
Ethanol
Propanol
Phenol
Bis-(2-chloroethyl) ether
Acetone
Isophorone
Methyl isobutyl ketone
Acetonitnle
Dichloroacetonitnle
Chloroform
Acrylonitnle
72
29
47
65
250
170
270
270
1
2770
750
>600
1230
220
Methyl formate
Ethyl acetate
Acetaldehyde
Acrolem
Crotonaldehyde
Furfural
Propanal
Butanal
Pentanal
Hexanal
Heptanal
Octanal
Nonanal
Decanal;
2 - Chloroacetaldeh yde
2090
4970
WO
580
1340
1100
1350
88 O2000)*
2800
940
440
570
130
240
23
*2000 based on column equilibrium.
Table 2. Breakthrough Capacities for Compounds in Water
Compound
Capacity in mg/g at given % breakthrough*
1% 10% 50%
Acetic acid
Phenol
Ethanol
Acrolein
Crotonaldehyde
Propanal
Butanal
Pentanal
Furfural
Ethyl acetate
Acetone
Acetomtrile
Chloroform
02
18
120
22
7
63
42
36
24
90
20
150
1
3
130
high
29
12
83
55
high
42
WO
29
high
3
9
757
high
high
27
high
81
high
58
no
52
high
10
«240 )
«220J
{ <240 1
«240 )
(<240 )
*10% or 50% breakthrough did not occur with some compounds before the run was terminated
10% and 50% breakthrough capacities for these compounds are higher than the 1% capacity
eluent. If, however, the bed is eluted
directly with methanol, the air would be
displaced rapidly and create air pockets m
the bed. The solvent would channel
around those pockets and portions of the
Silicalite would not be eluted.
Acetonitrile has been found to work
better than methanol for recovering
decanal from Silicalite.
Determination of Low-
Molecular-Weight Organic
Species in Water
A liquid chromatograph equipped with
a three solvent gradient elution capa-
bility was used for accumulation and
desorption of organic compounds from
standard water samples. A diagram of the
system is depicted in Figure 1. Initially,
organic-free, deaerated water is passed
through the solvent-selection valve and
the pump to flush the system. A sample is
then pumped through a column contain-
ing Silicalite where organic materials are
accumulated. When a sufficient volume
of sample has passed through the Silic-
alite bed, elution begins with a gradient of
water and methanol. The eluate from the
Silicalite column passes through a RP-8
chromatographic column, which serves
to partially separate the eluted
components, and then through a UV
detector. Eluate fractions are collected
from the detector outlet Components in
these fractions are determined by gas
chromatography.
The first test of the applicability of this
protocol to the determination of dichloro-
acetomtnle involved adding dichloro-
acetonitrile directly to the column and
then eluting with methanol. It was found
that recoveries were quantitative and it
was confirmed by GC/MS that no
artifacts or degradation products
interfered with the determination of
dichloroacetonitnle in standard samples.
The ability to elute small volumes of
dichloroacetonitrile was confirmed in
studies in which 100, 10, and 1 fjg
amounts were loaded on a bed and eluted
using a water-methanol gradient A?
shown in Table 3, essentially quantita-
tive recoveries were obtained. The total
protocol was tested by loading various
volumes of water containing various
concentrations of dichloroacetonitnle on
Silicalite, eluting with a methanol-water
gradient, and determining recoveries by
gas chromatography. Table 4 shows near
quantitative recoveries obtained at
concentrations ranging from 10 to 100
jug/L using sample volumes ranging from
100 to 780 mL
The procedure was applied to drinking
water from Ames, Iowa and indicated a
dichloroacetonitnle concentration of 0 1
fjg/L. The amount found is below the
detection limit required for confirmation
by GC/MS and, thus, the identity of the
recovered material as dichloroacetoni-
trile could not be confirmed
The protocol was also applied to
samples of raw water, finished water at
the treatment plant, and finished water m
the distribution system in Ottumwa,
Iowa. No dichloroacetonitnle was
detected in the raw water. Peaks having
retention times corresponding to those of
dichloroacetonitrile were present inchro-
matograms of the finished and distribu-
tion system water. The identities of those
peaks could not be determined by GC/MS.
Previous work has shown that dichloro-
acetonitrile decomposes rapidly at basic
pH levels and Ottumwa water has a pH of
about 9. Thus, it is unlikely that these
peaks are due to dichloroacetonitrile.
Currently the status of application of
the protocol to real water samples is
simply that the procedures should work if
a water supply is located containing
detectable amounts of dichloroacetoni-
trile, although applicability has not been
established
In addition to dichloroacetonitrile, the
protocol was applied to standard samples
containing chloroform, phenol, ethyl
acetate, crotonaldehyde, propanal,
acetaldehyde and butanal. In all cases
recoveries in excess of 80% were
obtained. Decanal could also be
recovered, but required the use of an
acetomtrile-water gradient m place of the
methanol-water gradient for elution from
Silicalite
-------
Sample
Methanol
Water
Solvent-Selection/Gradient-Formation Valve
High Pressure Metering Pump
Sample to Waste During Flush Cycle
Silicalite Adsorption Column
Effluent to Waste During Accumulation Cycle
RP-8 Column
Detector
Eluate Fractions Collected During
Elution Cycle
Figure 1. Experimental apparatus
Table 3. Recovery of Dichloroacetonitrile
from Silicalite
Amount
Loaded, /jg % Recovered
RSD
100
JO
1
96
99
98
4
1
2
Table 4. Recovery of Dichloroacetonitrile from Standards
Concentration
added, fjg/L
WO
WO
WO
10
10
Sample Volume
mL
WO
500
780
WO
780
Amount, ug, of
Dichloroacetonitrile
W.
50.
78.
1.
7.8
% Recovery
102
98
W5
105
98
Conclusions and
Recommendations
The present work has been successful
in elucidating the basic properties of the
hydrophobic molecular sieve known as
Silicalite and has led to the development
of what appear to be viable analytical
protocols for determining low-molecular-
weight organic components in drinking
water. However, as is often the case with
research, this study has provided more
questions than answers. The protocol
was applied to standard samples but
applicability was not established for
drinking water.
Continued development of the use of
this adsorbent as an analytical agent
should continue and would most likely
lead to an entire family of methods for
determining low-molecular-weight
compounds in real water samples.
Silicalite has been shown to be an
excellent adsorbent for a diverse
assortment of organic and inorganic
gases. The potential utility of this
adsorbent for sampling ambient air and
gaseous effluents should be explored
further.
There is a significant probability that
Silicalite could be used in inert gas
purging or closed-loop-strippi ng
procedures as a replacement for
currently used adsorbents or in
combination with them. In this
application it is expected that Silicalite
would trap materials that are not retained
by conventional adsorbents. An
evaluation of Silicalite for this application
is strongly recommended.
-------
Colin D. Chriswell, Douglas T. Gjerde, Gerda Shultz-Sibbel, James S. Fritz, and
Ikue Ogawa are with Ames Laboratory USDOE, Iowa State University, Ames, IA
50011.
W. Em He Coleman is the EPA Project Officer (see below).
The complete report, entitled "An Evaluation of the Adsorption Properties of
Silicalite for Potential Application to Isolating Polar Low-Molecular-Weight
Organics from Drinking Water," (Order No. PB 83-148 502; Cost: $8.50,
subject to change) will be available only from:
National Technical Information Service
5285 Port Royal Road
Springfield, VA22161
Telephone: 703-487-4650
The EPA Project Officer can be contacted at:
Health Effects Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
S. GOVERNMENT PRINTING OFFICE-. 1983/659-095/1913
-------
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
Postage and
Fees Paid
Environmental
Protection
Agency
EPA 335
Official Business
Penalty for Private Use $300
------- |