Environmental Protection Technology Series
ACTIVATED CARBON  ADSORPTION OF
         TRACE  ORGANIC COMPOUNDS
                  Municipal Environmental Research Laboratory
                       Office of Research and Development
                       U.S. Environmental Protection Agency
                              Cincinnati, Ohio 45268

-------
                RESEARCH REPORTING SERIES

Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:

      1.  Environmental Health  Effects Research
      2.  Environmental Protection Technology
      3.  Ecological Research
      4.  Environmental Monitoring
      5.  Socioeconomic  Environmental Studies
      6.  Scientific and Technical Assessment Reports (STAR)
      7.  Interagency Energy-Environment Research and Development
      8.  "Special" Reports -
      9.  Miscellaneous Reports

This report has been assigned  to the ENVIRONMENTAL PROTECTION TECH-
NOLOGY series. This series describes research performed to develop and dem-
onstrate instrumentation, equipment, and methodology to repair or prevent en-
vironmental degradation from point and non-point sources of pollution. This work
provides the new or improved technology required for the control and treatment
of pollution sources to meet environmental quality standards.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.

-------
                                            EPA-600/2-77-223
                                            December 1977
            ACTIVATED CARBON ADSORPTION OF
                TRACE ORGANIC COMPOUNDS
                          by

Vernon L. Snoeyink, John J. McCreary and Carol  J.  Murln
            Department of Civil  Engineering
                University of Illinois
                Urbana, Illinois  61801
                  Grant No.  R 803473
                    Project Officer

                    Alan A. Stevens
            Water Supply Research Division
      Municipal  Environmental  Research Laboratory
                Cincinnati, Ohio  45268
      MUNICIPAL ENVIRONMENTAL RESEARCH LABORATORY
          OFFICE OF RESEARCH AND DEVELOPMENT
         U.S.  ENVIRONMENTAL PROTECTION AGENCY
                CINCINNATI, OHIO  45268

-------
                                 DISCLAIMER
     This report has been reviewed by the Municipal Environmental Research
Laboratory, U.S. Environmental Protection Agency, and approved for publica-
tion.  Approval does not signify that the contents necessarily reflect the
views and policies of the U.S. Environmental Protection Agency, nor does
mention of trade names or commercial products constitute endorsement or
recommendation for use.

-------
                                  FOREWORD
     The Environmental Protection Agency was created because of increasing
public and government concern about the dangers of pollution to the health
and welfare of the American people.  Noxious air, foul  water, and spoiled
land are tragic testimony to the deterioration of our natural environment.
The complexity of that environment and interplay between its components
require a concentrated and integrated attack on the problem.

     Research and development is that necessary first step in problem
solution and it involves defining the problem, measuring its impact, and
searching for solutions.  The Municipal Environmental Research Laboratory
develops new and improved technology and systems for the prevention, treat-
ment, and management of wastewater and solid and hazardous waste pollutant
discharges from municipal and community sources, for the preservation and
treatment of public drinking water supplies, and to minimize the adverse
economic, social, health, and aesthetic effects of pollution.  This
publication is one of the products of that research; a  most vital communica-
tions link between the researcher and the user community.

     The ability of activated carbon to adsorb certain  types of organic
compounds has frequently been demonstrated.   In this publication the results
of research on the application of activated  carbon to reduce the levels of
organic contaminants in drinking water is examined.
                                      Francis T.  Mayo, Director
                                      Municipal  Environmental  Research
                                      Laboratory

-------
                                  ABSTRACT


     This research program was conducted to determine how effectively humic
substances and the trace contaminants 2-methylisoborneol, geosmin, the
chlorophenols and polynuclear aromatic hydrocarbons were adsorbed by acti-
vated carbon under the competitive adsorption conditions encountered in
natural waters.  Data were collected using isotherm tests and small-scale
laboratory columns.

     Humic substances were obtained from a commercial source, a well water,
leaf extract, and soil extract, and some of the materials were separated
into the humic acid and fulvic acid fractions.  The molecular weight distri-
butions of the fractions were determined by gel permeation chromatography
and ultrafiltration.  Significant differences in the adsorbability, haloform
formation potential, and fluorescence of the various fractions were observed.

     A procedure for easily synthesizing the earthy-musty odor-causing com-
pound 2-methylisoborneol (MIB) was developed, and a gas chromatographic
analytical technique for quantitative analysis of MIB and geosmin down to
0.1 yg/1 was successfully formulated and tested.  Humic substances compete
with MIB and geosmin for adsorption sites on activated carbon and signifi-
cantly reduce its capacity for these compounds.  These naturally occurring
odorous compounds were found to be much more strongly adsorbed than the
humic substances.

     Both the chlorophenols and the polynuclear aromatic hydrocarbons are
very strongly adsorbed.  Strong competition was observed between anionic
and neutral species of 2,4-dichlorophenol and 2,4,6-trichlorophenol, even
at the 1 yg/1-concentration level.  The Langmuir model for competitive
adsorption, or the Jain and Snoeyink modification of that model, conformed
well to the observed data, with one exception at pH 7.0.  The presence of
the various humic substances also caused a significant reduction in chloro-
phenol adsorption capacity.  Also, humic acid did not interfere with the
rate of adsorption of a model polynuclear aromatic hydrocarbon, anthracene.

     This report was submitted in fulfillment of EPA Grant No.  R 803473 by
the University of Illinois under the sponsorship of the U.S.  Environmental
Protection Agency.   This report covers the period January 6,  1975, to
July 5, 1977, and the work was completed July 5/1977.

-------
                                  CONTENTS

Foreword	iii
Abstract	   iv
Figures	   vi
Tables	   ix
Acknowledgment 	    x

     1.   Introduction 	    1
               Statement of Problem  	    1
               Objectives  	    1
               Adsorbate Characteristics and Background Information  .  .    2
               Approach to the Study	    8
     2.   Conclusions  	    9
     3.   Recommendations	   12
     4.   Materials and Methods	   13
               Adsorbents	   13
               Humic Substances	   13
               2-Methylisoborneol, Synthesis and Analysis  	   23
               Geosmin Source and Analysis 	   31
               Chlorophenols 	   32
               Polynuclear Aromatic Hydrocarbons 	   33
               Adsorption Test Procedures  	   36
     5.   Results and Discussion	   41
               Haloform Formation Potential of the Humic Substances  .  .   41
               Adsorption of Humic Substances  	   41
               MIB synthesis and Analysis	   53
               Adsorption of MIB and Geosmin	   65
               Adsorption of Chlorophenol  	   76
               Adsorption of Polynuclear Aromatic Hydrocarbons 	  107

References	110

-------
                                  FIGURES
Number                                                                  Page
  1   Reaction scheme for the chlorination  of phenol   	     6
  2   Upflow column for extracting humic material  	    14
  3   Organic carbon separation monitored by TOC analysis  	    16
  4   Typical  chloroform chromatogram 	    22
  5   Concentration apparatus 	    28
  6   Standard curve for 2,4-dichlorophenol  	  	    34
  7   Typical  chlorophenol  chromatogram 	    35
  8   Typical  anthracene chromatogram 	    37
  9   Chloroform formation from humic and fulvic acids   	    40
 10   Chloroform formation from soil  fulvic acid	    42
 11   Chloroform formation from soil  humic  acid	    43
 12   Adsorption isotherms for commercial  humic acid  measured  by
       fluorescence and UV	    44
 13   Adsorption of well water organic matter 	    46
 14   Effect of phosphate buffer concentration on  adsorption of
       soil  fulvic acid	    47
 15   Effect of pH  on the adsorption  of soil  fulvic acid	    48
 16   Adsorption of various types of  humic  substances  	    49
 17   Adsorption of molecular weight  fractions of  soil fulvic  acid   .  .    50
 18   Adsorption of molecular weight  fractions of  soil humic acid ...    51
 19   Adsorption of molecular weight  fractions  of  leaf fulvic acid   .  .    52
 20   Stereochemical  structures  of 2-methylisoborneol  and camphor
       enantiomers	54
                                     vi

-------
Number                                                                 Page

 21   Chromatogram from the product of the action of
       methyl!ithiurn on d-camphor  	  56

 22  Chromatogram from the product of the action of methyl!ithium
       on d-camphor after treatment with alkaline hydroxylamine  ...  57

 23  Chromatogram from the residual aqueous solution from the
       hydroxylamine reaction after hexane extraction  	  58

 24  Chromatogram from the hexane extract after the
       hydroxylamine reaction  	  59

 25  Chromatogram from the hexane extract in Figure 24 after
       exhaustive washing with 2 N sodium hydroxide  	  60

 26  Chromatogram obtained in a study of the recovery of MIB
       added to tap water at 0.1  pg/1  	66

 27  Adsorption of MIB	67

 28  Adsorption of geosmin	69

 29  Column breakthrough curves for MIB and geosmin  	  70

 30  Column breakthrough curve for geosmin 	  71

 31  Column breakthrough curve for MIB	73

 32  Adsorption of MIB--nonequilibrium 	  75

 33  Adsorption isotherms for 2,4-dichlorophenol 	  77

 34  Adsorption isotherms for 2,4,6-trichlorophenol  	  78

 35  Influence of pH on chlorophenol adsorption capacity  	  79

 36  Competitive adsorption capacities predicted by the Langmuir
       model for dichlorophenol at pH 5.2	96

 37  Competitive adsorption capacities predicted by the Langmuir
       model for trichlorophenol  at pH 5.2	97

 38  Competitive adsorption capacities predicted by the Langmuir
       model for dichlorophenol at pH 7.0	99

 39  Competitive adsorption capacities predicted by the Langmuir
       model for trichlorophenol  at pH 7.0	100

 40  Competitive adsorption capacities predicted by the Jain
       model for dichlorophenol at pH 9.1  	101

                                     vii

-------
Number                                                                 Page

 41   Competitive  adsorption  capacities  predicted by  the Jain
       model  for  trichlorophenol  at  pH  9.1	102

 42   Kinetics  of  anthracene  adsorption  with and without humic acid  .  .  108

-------
                                   TABLES

Number                                                                   Page

  1   Gel Molecular Weight Exclusion Limits and Ultrafiltration
       Membrane Molecular Weight Cut-Off Limits  	   17

  2  Finch Soil Humic Substance Fractionation  	   18

  3  Evaluation of G-50 Coarse Sephadex Fractions of Soil  Humic
       Acid by Ultrafiltration	19

  4  Fluorescence Intensity of 5 mg/1 TOC Fractions of Humic
       Substances	20

  5  Reagents, Solvents and Adsorbents for MIB Synthesis and Analysis.  .   23

  6  TLC and MIB and Some Other Camphor Derivatives	25

  7  Isolation of MIB by Column Chromatography of a Natural  Product  .  .   26

  8  Characteristics of the Chloro- and Bromophenols Studied 	   32

  9  Response Factor of MIB for Various Proportions with the
       Internal Standard and Amounts Gas Chromatographed 	   64

 10  Relative Recovery of MIB from Various Waters	65

 11   Values of Constants for Equation 3 for Chlorophenol Single
       Solute Data	81

 12  Chlorophenol Competition at pH 5.2	83

 13  Chlorophenol Competition at pH 7.0	87

 14  Chlorophenol Competition at pH 9.1  	91

 15  2,4,6-TCP Competition with Humic Substances - pH 5.2	104

 16  2,4,6-TCP Competition with Humic Substances - pH 9.1   	105

 17  Concentrations of Benzanthracene after Eight Hours  	  107

-------
                               ACKNOWLEDGMENT
     The assistance of Dennis  Beckmann,  Paul  Boening,  David Dunn,  Dennis
Herzing, Terry Temperly and Neville  Wood in  producing  material  for portions
of this report is acknowledged.   The adsorbents  used  in  this study were
supplied by the manufacturers.

-------
                                   SECTION 1

                                INTRODUCTION
STATEMENT OF PROBLEM

     Odor problems plague the majority of water treatment plants.   Granular
activated carbon (GAC) beds now are being used in the U.S.A.  predominantly
for the removal of odor-causing compounds and are reported to be effective
for certain types of odors of biological  origin for up to 3 to 5 years
(Love et al., 1973).  Although only limited data have been reported on  the
use of GAC for removal of odors of predominantly industrial origin  (Postal
et al., 1965), bed life may be much reduced.   Unfortunately very few data
are available to indicate the mechanism of removal  and to enable the applica-
tion of findings at one water treatment plant to odor problems at other
locations.  It also is not possible to say whether selected trace organics
will adsorb similar to the odor compounds or whether they will rapidly
saturate the GAC bed thus making thermal  regeneration necessary.

     Humic substances in water supplies are also of much concern.   These
materials react with chlorine to produce haloforms (Rook, 1976; Stevens
et al., 1976; Symons, 1976), they occupy adsorption sites on the carbon
surface thereby reducing the adsorption capacity for selected trace compounds
(Herzing et al., 1977), and they associate with metal ions, pesticides,
phthalates and possibly other organics (Schnitzer and Khan, 1972).   Humic
substances constitute the major fraction of organics in most natural waters.
There exists little information on the magnitude and nature of the  competi-
tive effects of humic substances, or information on how humic materials from
various sources can be removed by adsorption.  For these reasons humic
substances were an important part of this study.


OBJECTIVES

     The objectives of this study were to

     1.    Characterize humic substances from different sources and  to
          determine how these materials are adsorbed and the extent to  which
          they compete with selected trace organics for adsorption  sites on
          activated carbon.
        i

     2.    Determine how geosmin and 2-methylisoborneol (MIB), causative agents
          of earthy-musty odor in water supplies, adsorb on carbon  at their
          threshold odor level in the presence and absence of humic substances

-------
     3.   Determine how mixtures of chlorophenols, compounds which are odorous
          and which are representative of other undesirable  compounds, adsorb
          at the yg/1 level in the presence and absence of humic substances.
          Especially important is the degree to which the presence of one
          chlorophenolic species causes a reduction in the capacity of carbon
          for other chlorophenols.

     4.   Determine whether the adsorption of polynuclear aromatic hydro-
          carbons, some of which are proven carcinogens, is affected by the
          presence of humic substances.

     Achievement of these objectives will permit an assessment of whether
 past results on GAC adsorption of odor compounds are generally applicable,
 and whether selected, important trace compounds will adsorb similar to the
 odor compounds.


 ADSORBATE CHARACTERISTICS AND BACKGROUND INFORMATION

 Humic Substances

     According to Schnitzer and Khan (1972) humic substances are compounds
 which are amorphous, brown or black, hydrophilic, acidic, polydispersed
 substances of molecular weight of several hundred to tens of thousands.  In
 contrast, the nonhumic substances are those such as proteins, carbohydrates,
 carboxylic acids, etc. which exhibit recognizable chemical characteristics.
 The humic substances which are found in water can be classified into two
 broad categories.  The first is humic acid which is soluble in dilute
 alkaline solutions but precipitates in strongly acidic solutions.  The second
 is fulvic acid which is soluble in both acidic and basic solutions.

     These fractions are structurally similar but differ in molecular weight,
 ultimate analysis and functional  group content.  Both fractions generally are
 relatively resistant to microbial  degradation, form water soluble and insoluble
 salts and complexes,  and  interact with clays and organic matter in solution.
 It appears that the carboxyl,  hydroxyl, and carbonyl groups are the predom-
 inant functional  groups in humic materials and the relative proportions of
 these are an index of humic substance reactivity (Schnitzer and Khan, 1972).
The carbon content of humic acid ranges from 50 to 60 percent, the oxygen
 from 44 to 50 percent,  the nitrogen from 1 to 3 percent, and the sulfur from
near 0 to about 2 percent.

     Some adsorption studies have been conducted on humic substances, or on
fractions of organic matter which include the humic substances.  A pilot
plant with a 30 in.  GAC filter operated at 2 gpm/ft2 treating raw Ohio  River
water has been  evaluated for elimination of haloform precursors.   It was
shown that after  3 to 4 weeks  of operation, sufficient material was being
passed  through  the bed  to pr-oduce measurable amounts of chloroform when  the
effluent was chlorinated,  and  after 10 weeks the concentration of the GAC
bed effluent was  about  50% of  the concentration formed when  GAC was not  used
 (U.S.  Environmental  Protection Agency,  1975).

-------
     Sontheimer and Maier (1972) carried out an extensive evaluation of 10
different commercial carbons to determine their ability to remove organic
matter as measured by absorption of UV light at 240 nm.  River Rhine water
was used which had been filtered through the river bank, (this removes many
biodegradable organics and some dilution of river water with ground water
takes place), ozonated and filtered.  The conclusions which could be drawn
from their results were:

     1.   The relative positions and slopes of the isotherms were dependent
          upon the point on the lower Rhine where the sample was  taken and
          upon the time of the year and the rate of flow of the river.

     2.   The phenol number and the BET surface area do not provide a good
          indication of carbon effectiveness for organic removal  by carbons
          prepared from different raw materials or by different activation
          processes.  However, if the same raw material and the same activa-
          tion process are used, better adsorption properties are associated
          with the higher values of these parameters.  It is also likely that
          the more extensively activated carbons will cost more,  however.

     3.   The least effective of the 10 carbons in the past had proven to
          be very effective for odor removal.

Their results thus provide strong evidence that the adsorbability of this
organic matter is dependent on the type of material being adsorbed.

     Sontheimer and Maier (1972) further evaluated whether the relative
efficiencies of the carbons as indicated by the isotherms could be observed
in pilot scale tests using 1 m deep beds.  While the results were not
entirely conclusive, it was observed that the better carbons based on the
isotherm evaluation generally removed more material in the carbon beds.  The
total amount of adsorbed material as indicated by extraction with dimethyl-
formamide after a period of operation yielded results consistent  with those
obtained by analyzing the column effluent.  It was further observed that all
of the carbons tested in beds seemed equally effective for odor removal, but
that saturation of the carbon's capacity for total organics was reached much
before its capacity for odorous compounds was reached.  They also found
evidence indicating that some of the odor compounds were removed  in the
upper part of the bed by biological activity.

     In later work by Schweer and co-workers (1975), the removal  of sulfur
containing organics by GAC and other treatment processes was found to
parallel TOC removal.

     Sontheimer (1974) reports pilot plant data obtained at the waterworks
in Diisseldorf showing the efficiency of 3 different carbons for removal of
TOC and total chlorinated organic compounds (TOC1) as determined  by the
technique of Ku'hn and Sontheimer (1973a, 1973b), as well as specific
chlorinated organics.   TOC1  is a measure primarily of the lipophyllic organic
group of compounds.   This group of compounds practically never occurs in
nature, are difficult to decompose biologically, and are frequently hazardous
to health.   There are currently two methods used to determine these compounds.

-------
 The first  involves  adsorption of them onto carbon or a synthetic resin and
 then to extract  with  dioxane or dimethylformamide.  The organic chlorine in
 the extract  can  then  be  analyzed by combusting the sample in a quartz
 combustion tube  and then determining the chloride produced by microcoulometry
 (Klihn and  Sontheimer,  1973a).  The second involves direct combustion of the
 carbon sample  containing adsorbed organochlorine compounds and determination
 of the chloride  which  is produced by either microcoulometry or a chloride ion
 electrode.   Care must  be taken to eliminate interference from inorganic
 chloride adsorbed on  the carbon (Ku'hn and Sontheimer, 1973a, 1973b).  The
 phenol  number  and surface area were again found to have no relationship to
 the adsorption efficiency.  Also the carbon which adsorbed the most total
 organic matter from bank filtered and ozonated Rhine River water over a
 6  month period of time, as indicated by TOC measurements on the filter influ-
 ent and effluent as well as by extraction of the carbon with dioxane and
 dimethylformamide,  adsorbed the least TOC1.  The efficiencies of adsorption
 of specific  chlorinated organics, hexachlorocyclohexane, bis (2-chloropropyl)
 ether and  hexachlorobutadiene, paralleled the removal of TOC1.   These results
 indicate the importance of knowing the objective for which carbon is to be
 used prior to  selection of the carbon for an application.

 Geosmin and  2-Methylisoborneol

      The results of using 6AC for the removal of odors presumably of biolog-
 ical  origin  indicate that odor breakthrough occurs much later than break-
 through of organics measured by a more general parameter such as carbon
 chloroform extract  (CCE) or carbon alcohol extract (CAE) (Robeck, 1975).
 Because  most GAC adsorbers are now in use with the objective of removing
 odors of biological  origin, two compounds of biological  origin,
 2-methylisoborneol  (MIB) and geosmin, were chosen for in-depth study.   Close
 examination of the  adsorptive behavior of these species should  indicate
 why  the  bed-life for such compounds is so long and whether other trace
 compounds are  likely to be similarly adsorbed.

     Geosmin is produced by some actinomycetes and blue-green algae (Rosen
 et al.,  1970; Gerber and Lechevalier, 1965; Medsker et al., 1968) and MIB
 has  been identified  as a product of actinomycetes (Rosen et al., 1970;
 Medsker et al., 1969).  These compounds have been identified as causative
agents of the widespread problem of earthy-musty odors in water supplies
 (Rosen et al.,  1970; Medsker et al., 1969; Jenkins, 1973).   A survey in 1957
 showed that two-thirds of the water treatment plants surveyed described
their odor problems  as earthy-musty or, similarly, as that of decaying
vegetation.(Sigworth,  1957).   Dice (1976) also reports survey results
 indicating that odor from actinomycetes is a frequent problem.   MIB and
geosmin specifically have been found in surface waters with earthy-musty
odors in Europe by the Dutch (Piet et al., 1972),  and Japan (Kikuchi  et al
 1973a, 1973b).

     MIB was  identified by Medsker et al. (1969).   Geosmin  was  first  isolated
and named by  Gerber  and Lechevalier (1965),  while  Safferman et  al.  (1967)
 later isolated  an identical  compound.   Gerber (1967)  presented  a  partial
 identification  of this compound which was substantiated  by  Medsker  et  al.

-------
(1968).
0.1-0.2
below.
 The threshold odor concentrations of geosmin and MIB are about
yg/1 (Jenkins, 1973) and their structural representations are shown

                                                  CH,
                                              OH   -J
            2-Methylisoborneol
                                              CH3

                                            Geosmin
Chlorophenols

     Chlorophenols were selected as model compounds for our study for a
number of reasons.  First, they impart an objectionable taste and odor to
water when present and may have been partly responsible for the reported
chemical tastes and odors in the drinking water that triggered the Lower
Mississippi Study (U.S. Environmental Protection Agency, 1975).  In a study
performed to determine the halogenated compounds which result from
chlorinated secondary sewage effluents, trichlorophenol, as well  as a number
of other chlorinated aromatic compounds, was found (Glaze and Henderson,
1975).  In these two instances, Chlorophenols probably were formed by the
reaction of phenols with aqueous chlorine.  The primary sources of phenols
in natural waters include natural decay products, waste effluents of coking
plants, brown coal distillery plants, and the pulp and paper industry.
Phenols are used in the synthesis of a number of organic compounds resulting
in their presence in the effluents from many chemical plants.  It has been
estimated that the concentration of free phenols in unpolluted streams is
less than 50 yg/1 while that in rivers receiving industrial and municipal
wastewater is frequently greater than 100 yg/1 (Zogorski and Faust, 1974).

     According to Burttschell et al. (1959), the chlorination of phenol
proceeds by stepwise substitution of the 2, 4, and 6 positions of the
aromatic ring, in the manner shown in Figure 1.  Below each compound in
Figure 1 is listed its threshold odor concentration.  The compounds with the
strongest odor-producing potential are 2-chlorophenol, 2,4-dichlorophenol,
and 2,6-dichlorophenol which are detectable at concentrations from 2 to 3
yg/1.  These are the compounds primarily responsible for the taste and odor
in water.   The 1962 Public Health Service Drinking Water Standards set the
maximum level  for total phenols at 1 yg/1 to prevent odor problems.  Toxic
effects are thought to occur only at far higher concentrations partly
because phenols are largely detoxified in the mammalian body (McCaull and
Crossland, 1974).

-------
  OH
                               /\
                                                  OH
                                       Cl      Cl
                /\
                       Cl
cri
                   OH
                 y\
 V
[>iooo]
                                 OH
  C I
[250]
                V
                  OH
                /\
                                                 Cl
                                                               OH
                                                             v

                                                                     Oxidation
                Figure 1.  Reaction scheme for the chlorination of phenol.
                         [ ] indicates odor threshold concentration (yg/1)
                         (after Burttschell et al.,  1959)

-------
     The series of reactions leading to the formation and ultimate destruc-
tion of chlorophenolic odors involves complex kinetic interrelationships.
A short time after mixing chlorine and phenol, eight interdependent reactions
are occurring simultaneously.  In a study of the kinetics of these reactions
by Lee (1967), the expected concentrations of the individual chlorinated
phenols were determined as a function of time and other variables.  By
combining these data with information on the organoleptic properties of the
compounds, Lee determined the threshold odor values of chlorinated waters
which contained phenols as a function of time, pH, and relative concentrations
of chlorine and phenol.  Depending on the conditions which existed in the
water, Lee observed that different mixtures of chlorophenols may be present
at a given time.

     The second reason for selecting chlorophenols for use in this study was
due to their similarity in structure to a number of pesticides which resist
biodegradation.  Roughly 25 percent of the pesticides on the world market
are compounds which possess a substituted phenol moiety which can be cleaved
from the molecule through hydrolysis in natural waters (Friestad et al.,
1969).  It has been reported that photodecomposition of the herbicide
2,4,5-T (2,4,5-trichlorophenoxyacetic acid) by sunlight in alkaline natural
waters may result in the formation of 2,4,5-trichlorophenol and 2,5-
dichlorophenol (Crosby and Wong, 1973b).  Studies by Crosby and Wong (1973a,
1973b) indicate that this pathway may be general for other commercial
phenoxy herbicides.  Therefore, the decomposition of pesticides in natural
waters may lead to trace amounts of chlorophenols.

     In spite of their importance little information is available to indicate
how chlorophenols adsorb from waters containing a mixture of chlorophenolic
species and humic substances.

Polynuclear Aromatic Hydrocarbons

     Polynuclear aromatic hydrocarbons (PAH) are ubiquitous and are found in
small but detectable concentrations in air, water and soil samples of all
types (McGinnes and Snoeyink, 1974).  They are natural products of organic
decomposition and are products of incomplete combustion, petrochemical,  coal
and chemical industrial processes.  The concentrations found in water typi-
cally range from 0.001 to 10 yg/1 and they are concentrated in the food
chain because of their favorable solubility in fatty material.  Some species
of PAH are demonstrated carcinogens at higher concentrations and because
of this they are a potential health hazard in water.

     Previous studies have shown naphthalene to be very strongly adsorbed
from Cincinnati tap water (Robeck, 1975), and that river bank filtration
followed by activated carbon treatment removed about 99 percent of PAH. from
River Rhine water (Andelman, 1973).

     In this study, our primary goal was to determine whether PAH associated
with humic substances.  If this does occur, it is possible that adsorption
onto carbon will be controlled by the adsorption characteristics of the humic
substances and not by those of the PAH.

-------
APPROACH TO THE STUDY

     The study was conducted using aqueous solutions prepared in the labora-
tory.  Sufficient quantities of the adsorbates were obtained from various
sources.  MIB was synthesized by us while the geosmin was obtained from the
U.S. EPA.  The chlorophenols and PAH were obtained from commercial sources.
Humic substances were obtained from a commercial source, a local well water,
and from leaf and soil extract.  Coal-base activated carbon was used as the
adsorbent.   In most cases it was necessary to develop  a quantitative anal-
ytical procedure for each of the species.

     Small-scale laboratory batch and column tests were used to obtain the
desired information on adsorption characteristics.  These tests are very
flexible and enabled us to look at a wide variety of conditions.

     The tests were conducted in such a way so that biological  activity did
not take place.   Short-term tests, and in some cases biocides were used to
insure the  absence of significant biological  growths.   It is recognized that
biological  growths are prevalent in GAC beds at water treatment plants,
and that such growths significantly affect the quality of the effluent from
these beds  (see  McCreary and Snoeyink, 1977 for a review) but the scope of
this study  did not permit us to examine it.

-------
                                  SECTION 2

                                CONCLUSIONS
     Activated carbon adsorbed humic substances in all  cases that were
studied but the adsorption properties of the substances from different
sources varied widely as did the extent to which they competed with selected
trace organics for adsorption sites on activated carbon.   Humic substances
from leaf and soil extract, a well water and a commercial  source were
examined in detail.  Extent of adsorption depended upon solubility, with the
less soluble humic acid (HA) fraction being more adsorbable than the fulvic
acid (FA) fraction from the same source.  The lower molecular weight species
from a given FA or HA fraction are more adsorbable than the high molecular
weight species presumably because more surface area is  accessible to them.
The adsorption characteristics of the humic substances  are also dependent
on the method of analysis used to quantify them.  The species which fluoresce
the most were found to be the lower molecular weight species and these adsorb
best.  UV absorbing species did not adsorb as well as those which fluoresce.
Solution pH and phosphate concentration also had a marked effect on adsorba-
bility of the humic materials with adsorption generally improving with
decreasing pH and increasing phosphate concentration.  The haloform formation
potential of the humic substances varied widely from source to source, with
only one exception, but no dependence on molecular weight was found for
fractions of FA or HA.  Based on the results of this work it is concluded
that when designing an adsorption system to remove haloform precursors or
trace organics, it is important that testing be done using the water to be
treated to determine design and operating parameters.  Because of the vari-
ability in adsorption characteristics of the humic substances it would appear
that adsorption results obtained at one location will likely not be those
obtained at another location.

     Concentrations of 2-methylisoborneol (MIB) and geosmin can be quanti-
tatively measured down to 0.1 yg/1 by a procedure consisting of extraction,
concentration and gas chromatographic analysis with a flame ionization
detector.  The MIB synthesized from d-camphor was found to be identical to
that produced by actinomycetes.

     MIB and geosmin are both strongly adsorbed by activated carbon.  When
present, the humic substances significantly reduce the capacity of carbon
for these compounds, more so before equilibrium is achieved than at equi-
librium.  Commercial HA and the humic substances from well water each had
differing competitive effects on MIB.  The capacity of carbon for geosmin
was reduced to a greater extent than was observed for MIB by commercial HA.
The performance of laboratory columns was consistent with the isotherm

-------
 results.  Application of distilled  water  to  a  partially  saturated  carbon  bed
 resulted in almost no elution  of MIB  indicating  that  it  was  strongly
 adsorbed.  Assuming complete  saturation of the carbon, no  leakage  and  no
 biological  activity, predicted bed  life for  reduction of MIB or geosmin  from
 10 yg/1 to its threshold odor  level of 0.1 yg/1  in  a  2-foot  deep bed is much
 greater (several  months  to  years) than for reduction  of  humic  substances
 from 5 to 1 mg/1  (1 to 2 months), for example.   When  both  MIB,  or  geosmin,
 and humic substances must be removed, humic  substance removal  will  control
 the life of the bed.  In field installations of  granular activated  carbon,
 the possibility that odor compounds are generated,  or in some  instances
 degraded, by biological  growths  within the bed cannot be overlooked.

      Chlorophenols are adsorbed  very strongly  by activated carbon  at the
 yg/1 level  which  is near the threshold odor  limit for these  compounds.  The
 extent of adsorption of  2,4-dichlorophenol (DCP) and  2,4,6-trichlorophenol
 (TCP)  is a  function of pH.  The  neutral species  of  these compounds  predomi-
 nate at pH  below  the pKa values  (7.85 and 6.00,  respectively,  at 25°C) and
 are adsorbed more strongly  than  the anionic  species.  As the number of
 chlorine atoms substituted  on  the phenol  increases, the  solubility  of  the
 neutral species decreases and  the adsorbability  increases; as  substitution
 increases,  the pKa of the species is lowered,  however.   When water  containing
 phenol  is chlorinated with  low levels of  chlorine,  a  mixture of chlorophenols
 will form and thus the extent  of adsorption  of one  chlorophenol  in  the
 presence of another chlorophenol is an important consideration.  Significant
 reductions  in adsorption capacity (up to  50  percent)  of  one  chlorophenol was
 caused  by the presence of a second chlorophenol.  The langmuir  model for
 adsorption  was found to  be  inadequate for fitting single solute  adsorption
 data over a  broad concentration  range.  To obtain Langmuir parameters  for
 use in  competitive adsorption  equations,  it was  necessary  to fit the single
 solute  data  with  a polynomial  equation and to  use the polynomial to calculate
 the Langmuir parameters  at  the desired concentrations.   Competitive adsorp-
 tion studies  between  DCP and TCP resulted in verification of the applica-
 bility  of the  Langmuir competitive adsorption  equation at  pH 5.2, and  the
 Jain and  Snoeyink  modification of this model at  pH  9.1.  At  pH  7.0, where
 neutral  DCP  competes  with anionic TCP, neither equation  was  satisfactory.
 Evaluation of  the  competitive  effects of  commercial HA,  soil  FA  and leaf FA
 showed  that  the presence of these materials  decreased the capacity  of  carbon
 for chlorophenol  and  that each of the materials  competed somewhat differently.
 However, even  in  the  presence of humic substances and another chlorophenol
 species,  it appears  that the adsorption capacity is even greater for chloro-
 phenol  th-an  it  is  for  MIB and that bed life for  chlorophenol will be greater
 than for MIB and much  greater than for humic substances.

     Limited experimentation with the polynuclear aromatic hydrocarbon (PAH)
anthracene led to  the  conclusion that there was  no  significant association
between it and humic  substances.  Thus it appears that the possibility of
PAH passage through  carbon beds because of its association with the more
poorly adsorbable  humic substances is not a cause for concern.

     The general conclusion made on the basis of this  study is that adsorp-
tion characteristics of organic substances which one may wish to remove
during water purification are highly variable.   Past experience  with full-


                                     10

-------
scale carbon beds at water treatment plants, used primarily to  remove  odor,
provides little indication of how humic substances and selected trace
organics will be removed, for example.   Adsorption tests  using  the water  to
be treated should be conducted prior to designing systems to accomplish a
specific objective.   On the basis of this study it is  expected, however,
that carbon bed life for removal of typical  levels of  chlorophenols, MIB  and
geosmin will be much longer than for humic substances.
                                    11

-------
                                  SECTION 3

                               RECOMMENDATIONS


     Laboratory and field studies should be initiated to develop simplified
 procedures for characterizing natural organic matter with respect to adsorb-
 ability, competitive effects and tendency to form haloforms when chlorinated.
 Fluorescence response, UV absorbance, TOC and density of selected functional
 groups should be related to both haloform formation potential and adsorba-
 bility and seasonal variability in type of organics should be taken into
 account.  Research should be conducted on procedures other than adsorption,
 such as coagulation, which may be effective in removing certain fractions
 of humic substances and thereby increasing carbon bed life.

     Field studies should be undertaken at water treatment plants where
 granular activated carbon is used and where geosmin or MIB is a problem.
 Carbon samples taken from several depths within the bed should be extracted
 and the extract should be analyzed for the odor compound to determine the
 concentration profile in the bed.  Determination of the profile at different
 times will  permit a determination of the rate of movement of the odor com-
 pound through the bed and verification of the prediction that bed life
 should be on the order of several months to years.   Biological activity in
the bed should also be examined to determine whether it is producing or
degrading the odor compound within the bed.

     Research should be undertaken to develop reliable small-scale adsorption
test procedures to be used on-site to determine the best design parameters
and operating conditions to treat a given water.  The small-scale test
results would necessarily have to be compared with large-scale results and
this could best be accomplished at locations where pilot- or full-scale
 studies are being initiated.  If small-scale tests can be used successfully,
 significant reductions in the time and funds required to obtain necessary
 design and operation information should be possible.
                                     12

-------
                                  SECTION 4

                            MATERIALS AND METHODS
ADSORBENTS
     Two bituminous coal base activated carbons, Filtrasorb 200 (F-200) and
Filtrasorb 400 (F-400) were used for this study.  Both carbons were prepared
by grinding, sieving to the desired size range (i.e., 40 x 50, 50 x 60, or
60 x 80 U.S. standard mesh), washing, and then drying to a constant weight
at 120-140°C.  See Calgon Activated Carbon Product Bulletin (1969) for
general carbon characteristics.


HUMIC SUBSTANCES

Isolation and Purification

     The humic substances used in this study were obtained from a commercial
source (Pfaltz and Bauer), and were extracted from leaves and soil.  Humic
substances in well water obtained from a deep well in the Civil Engineering
Building at the University of Illinois were also used.  The commercial humic
acid was purified using the procedure outlined by Narkis and Rebhun (1975),
with some modifications.  The material was dissolved in 1.5 percent NaOH,
filtered, acidified to pH 1 with HC1 to precipitate the humic acid, washed
until chloride free and dried at 60 to 65°C.  The procedure was later changed
to employ centrifugation rather than filtration and freeze-drying concentra-
tion instead of oven drying.  The purification step did not have a signifi-
cant effect on the humic acid adsorption isotherm as shown by comparison of
isotherms determined before and after the purification step.

     Well water from a deep aquifer at the University of Illinois was also
used as a source of humic material.  Prior to use it was aerated and settled
to remove the iron.  The well water had a yellowish-brown color and a
chemical oxygen demand (COD) (Standard Methods, 1975) of about 10 mg/1.  It
was also analyzed at the USEPA Municipal Environmental Research Laboratory
in Cincinnati and found to have a nonpurgeable total organic carbon (NPTOC)
concentration of 3.3-3.6 mg/1.

     A continuous upflow column (see Figure 2) was used to extract humic
material  from leaves and a Finch soil.  Eight liters of deionized water was
cycled continuously through a bed of leaves acquired from a hardwood forest.
Air was bubbled into the holding tank to prevent septicity.  The water
quickly became discolored and after about two weeks of cycling, the organics
                                      13

-------
 Soil
                                                        Pump
                                                       Reservoir
Figure 2.   Upflow column for extracting humic material
                             14

-------
which were extracted were concentrated by freeze-drying.  The pH of the
water was 6.8.  Very little material precipitated from the solution at pH 1
indicating that almost no leaf humic acid was present.

     Eight liters of a 0.1 M sodium pyrophosphate solution at a pH of 10.6
was cycled through the column with Finch soil.  The Finch series is a some-
what poorly drained sandy soil with a strongly cemented subsoil.  It was
obtained from an area near Traverse City, Michigan.  The solution became
immediately discolored and contained fine sediment that would not filter
readily.  After about three days of cycling, the solution was centrifuged in
150 ml polyethylene bottles at 8000 rpm for 30 minutes.  This served to
deposit the colloidal material.  The centrate was then acidified to pH 1
and stirred for approximately one hour.  This solution was centrifuged to
deposit the precipitated humic acid while the fulvic acid remained in
solution.

     In order to separate the fulvic acid from the remaining solution,
XAD-8 macroreticular resin from Rohm and Haas was used following the pro-
cedure of Leenheer and Huffman (1976).  The resin was initially washed with
methanol and then Soxhlet extracted with ether, acetonitrile, and methanol
for 8 hours each.  A column 3 cm in diameter and 30 cm long was used for the
resin.  The XAD-8 resin was followed by a similar bed of XAD-2 resin which
resulted in little additional removal (see Figure 3 for typical results).

     About one liter of fulvic acid solution at pH 1 was applied to the
XAD-8 bed at a flow rate of approximately 2.5 ml/min.  This was followed by
one liter of 0.1 M NaOH to regenerate the resin.  The desorbed fulvic
material moved as a band through the bed.  Backwashing followed by several
bed volumes of deionized water was sufficient to prepare the resin for
adsorption of more fulvic acid.  The fulvic acid solution obtained from the
resin was adjusted to pH 7 and freeze-dried.

     Dialysis tubing was used initially to prepare salt-free organics.
However, it was observed that the smaller molecular weight organics were
passing through the tubing and into the surrounding solution.  A similar
observation has been reported by Stevenson (1965).  Freeze-drying was
therefore conducted without prior dialysis.

Molecular Weight (Size) Fractionation

     Gel filtration was used for the separation of materials on the basis of
molecular weight or size.  Gjessing (1976) has shown that the non-excluded
fraction from a Sephadex gel column is of questionable value as far as a
molecular weight estimate is concerned due to irreversible adsorption inside
the gel  particle.  For the fractionation of humic material he used a series
of Sephadex gel columns and selected only the excluded fraction from each
column with subsequent concentration and reapplication of the non-excluded
fraction to the column with the next smaller molecular weight exclusion
limit.  We used a similar procedure accompanied by ultrafiltration to obtain
our molecular weight fractions.
                                     15

-------
           100% TOC initially extracted from Finch soil
                              I
                       pH adjusted to 1
                          centrifuge
                              \
                      61% TOC in solution
                                           humic acid precipitate
                                                39% of TOC
                   adsorption on XAD-8 resin
                                                fulvic acid
                                                41% of TOC
                    20% of TOC in solution
adsorption on XAD-2 resin
                       X
                    17% of TOC in solution
                                              adsorbed organics
                                                3% of TOC
Figure 3.   Organic carbon separation from Finch soil monitored  by
           TOC analysis
                               16

-------
     Table 1  gives the approximate molecular weight exclusion limits for the
gels and the molecular weight cut-offs for the ultrafiltration membranes
used.
      TABLE 1.  GEL MOLECULAR WEIGHT EXCLUSION LIMITS AND ULTRAFILTRATION
     	MEMBRANE MOLECULAR WEIGHT CUT-OFF LIMITS	

             Ami con Diaflo Filters             Sephadex Gels


             UM 10      10,000 MW          G-75      50,000 MW

             XM 50      50,000 MW          G-50      10,000 MW

                                           G-25       5,000 MW

                                           G-10         700 MW
     Our procedure was modified from Gjessing's in that a pH 10, 0.01  M
phosphate buffer was used as the eluant to decrease the adsorption of
charged groups on the Sephadex particles (Pharmacia, 1974).  Also, freeze-
drying was used to concentrate the eluted organics rather than roto-
evaporation at 40°C.

     Each of the gels was boiled in distilled water and allowed to swell
for at least the time specified by the manufacturer.  The columns were
poured through a Buchner funnel fitted on the top of the column with constant
stirring and a slow constant flow rate through the column.  Finally a glass
fiber filter was placed on top of the bed to prevent disturbance when adding
sample.  The column was thoroughly washed with the phosphate eluant prior to
use.  Several bed volumes of buffer were required to halt leaching as
monitored by TOC.  The absence of packing irregularities was confirmed by
the application of Blue Dextran dye and analysis of the resulting
chromatogram.

     One hundred milligrams of organic material dissolved in 2 ml of water
was applied to the top of the column.  When necessary the solution prepared
from the freeze dried organic was filtered to eliminate insoluble residue
before application.  Irreversible adsorption on the gels was noted when the
soil and commercial humic acid were applied and excessive head loss developed
with the fine-grade gels.  For this reason, only the coarse-grade Sephadex
was used for these materials.  The volume containing the excluded organics
was taken as equal to one bed volume after color appeared in the bed effluent.
The first bed volume, consisting of that liquid which passed after application
of the sample and before color breakthrough, was discarded.  The flow rate
was approximately 0.5-1.0 ml/min.  Sampling was done with an automatic SMI
fraction collector.
                                      17

-------
     To further purify the Sephadex fractions, ultrafiltration was used.
As shown in Table 1, the Sephadex G-75 exclusion limit is identical  to the
molecular weight cut-off for the Amicon  XM-50  membrane.   The  same  is  true Tor
Sephadex G-50 and the UM-10 filter.  The volume containing the excluded
organics from the G-75 gel was filtered through the XM-50 membrane.   The
solution that passed through the membrane was  discarded and the organics
retained by the filtration unit constituted the purified fraction of molecular
weight > 50,000.  The XM-50 filter was also used to purify the non-excluded
fraction of G-75 by collecting the organics that passed through the filter
and discarding those retained by the unit.  The filtrate was freeze-dried to
small volume and applied to the G-50 gel filtration column.  An identical
procedure was followed for the G-50 gel  and the UM-10 filter.  Molecular
weight fractions from the other gels were not  further purified by
ultrafiltration.

     A mass balance (see Table 2) was conducted on the Finch soil humic
material for the molecular weight fractionation using Sephadex gel filtration.
It was observed, as expected, that the humic acid consisted of generally
larger weight molecules than the more soluble  fulvic acid.


	TABLE 2.  FINCH SOIL HUMIC SUBSTANCE FRACTIONATION	

                Soil Humic Acid              Soil Fulvic Acid


           MW > 50,000    56.4%            MW  > 50,000    29%

   5,000 < MW < 50,000    32.3%    5,000 < MW  < 50,000    48%

           MW <  5,000    11.3%      700 < MW  <  5,000    20%

                                           MW  <    700     3%
     A mass balance was also made on the G-50 gel fractions of Finch soil
humic acid using the Amicon XM-50 and UM-10 ultrafiltration membranes.  From
Table 3 we observe that much of the humic substance excluded from this gel
passed through the UM-10 membrane which has the 10,000 MW cut-off.
Similarly, much of the volume of < 10,000 MW organics from this gel was
retained by the filtration unit.  Thus we see that the use of ultrafiltration
after Sephadex gel fractionation significantly improves the fractions.

     Gjessing (1976), observed the same results and attributed this as a
demonstration of the effectiveness of ultrafiltration over gels as a frac-
tionation tool.  However, experience in our laboratory demonstrated problems
with clogging of the pores of the membrane if prior gel filtration was not
used and little difference among the fractions in terms of adsorption
capacity when ultrafiltration was used alone as compared with the non-
fractionated material.  It appears from our results that gel  filtration
followed by ultrafiltration gives superior results.

-------
       TABLE 3.  EVALUATION OF 6-50 COARSE SEPHADEX FRACTIONS OF SOIL
                         HUMIC ACID BY ULTRAFILTRATION
   Procedure
Molecular Wt.
of Total
Molecular Wt.    % of Total
Gel Filtration    > 10,000
                  100
            < 10,000
                   100
Ultrafiltration


> 50,000
> 10,000
< 10,000
42.6
14.7
42.4
> 10,000
< 10,000

45.4
54.6


Fluorescence, UV Absorbance, Total Organic Carbon --

     Three primary analytical procedures, ultraviolet absorbance at 240  nm,
fluorescence, and TOC were used for determination of the concentrations  of
the humic substances.  The fluorescence measurements were made using a
Turner Model 110 fluorometer with an excitation wavelength of 365 nm, an
emission wavelength of 415 nm and a path length of 1 cm.   Fluorescence scans
were obtained with an Aminco-Bowman scanning spectrofluorometer-  Excitation
spectra and emission scans were obtained for the various humic materials at
identical NVTOC concentrations.  The solutions were buffered at pH 7. We
observed that the maximum excitation wavelength was near 360 nm, comparable
to the 365 nm wavelength used on the Turner fluorometer.   The maximum
emission peak was very broad and in the range 450-470 nm as compared to  the
415 nm wavelength used on the Turner fluorometer.  All  of the materials  had
essentially the same excitation and emission characteristics.

     The fluorescence intensity of the unfractionated and fractionated humic
materials at 5 mg/1 TOC concentration was found to correlate well with
molecular size (see Table 4).  The smaller molecular weight materials were
found to have a much higher fluorescence intensity than the larger molecular
weight substances.

     A Beckman ACTA III spectrophotometer was used for the UV measurements.

     For well water analysis, the average TOC of 3.45 mg/1 of an aerated,
settled sample was used.   A standard curve was then prepared by making
dilutions of this sample and determining the UV or fluorescence response of
the dilution.  Unknown concentrations were determined by measuring the
fluorescence of a sample and using this standard curve; mass concentrations
are thus based on equivalent TOC and on the assumption that UV absorbance/
unit TOC or fluorescence intensity/unit TOC was the same for the unknown as
for the standard curve samples.
                                     19

-------
           TABLE  4.   FLUORESCENCE  INTENSITY OF 5 MG/L TOC FRACTIONS
                             OF HUMIC SUBSTANCES        	
                    Sample                Fluorescence Intensity
Commercial
Ful vie Soil
Fulvic Soil
Ful vie Soil
Fulvic Soil
Soil Humic,
Soil Humic
Soil Humic
Soil Humic
Leaf Fulvic
Leaf Fulvic
Humic Unfractionated
, Unfractionated
>50,000
>5,000
<5,000
Unfractionated
>50,000
>10,000
<10,000
>5,000
<5,000
415
315
94
195
320
185
57
96
200
237
397
      Since  the soil and  leaf humic material was not entirely salt-free,
 50 mg/1 TOC solutions were prepared.  Standard curves were made by making
 dilutions of these solutions and determining the fluorescence response of
 the dilutions.  Concentrations of unknown solutions could be determined by
 comparing their fluorescence to this standard curve; mass concentrations are
 thus  based  on equivalent TOC and the assumption that the fluorescence
 intensity/unit TOC was the same for the unknown samples as for the standard
 curve samples.

 Haloform Formation Potential
     The haloform formation potential (Stevens and Symons, 1977) of the
various humic fractions was determined to further characterize the material.
The chloroform analysis procedure reported by Kaiser and Oliver (1976) was
modified to suit our needs.  All water used for chloroform analysis was
distilled in glass followed by exhaustive stripping with N2 heated over a
copper catalyst.  This water was then buffered at pH 7 with 0.001  M phosphate.
All glassware was thoroughly washed and baked overnight at 450°C before use.
Standard chloroform solutions were prepared by injecting chloroform
(Mallinckrodt Nanograde) into an appropriate amount of water followed by
stirring with minimal headspace overnight.  Solutions of the internal
standard, 1,1,1-trichloroethane, were prepared in a similar fashion.


                                      20

-------
     Water with chloroform was spiked with a known amount of internal  standard
solution for analysis.  The sample was then placed in a 60 ml separatory
funnel with a 2 ml headspace and inverted in a 70° water bath.   After
equilibration for 45 minutes, about 100 yl of the gas was sampled by gas-
tight syringe and injected directly into a Hewlett-Packard gas  chromatograph
with a Ni63 ECD detector.  A six foot 0.4% Carbowax 1500 on Carbopack A column
was used with argon-methane carrier gas at a flow rate of 36 ml/min.  The
column was preconditioned at 210°C for one week after packing.   Temperature
settings were 220°C for the detector, 200°C for the injector port, and 110°C
for the column.  At these conditions, the retention times for chloroform and
1,1,1-trichloroethane were 2.2 minutes and 3.6 minutes respectively.  A
typical chromatogram  is shown in Figure 4.

     Chlorine solutions were prepared by bubbling Cl2> (Linde high purity
gas), into an alkaline solution.  The pH was then adjusted to 7 and the
solution was stored in a dark bottle at 4°C.

     Samples of humic material were prepared at 2.5 mg/1 TOC and chlorinated
with 10 mg/1 chlorine as Cl2 as measured by the DPD titrimetric method,
(Standard Methods, 1975).  All water was distilled, stripped, and buffered
at pH 7 with 0.001 M phosphate.  The reaction solution was mixed and immedi-
ately  transferred to 125 ml Wheaton Scientific glass vials.  Solution was
added to fill the bottle without headspace and the caps were crimped over a
a Teflon seal.  The reaction bottles were placed in an 18°C constant temper-
ature  water  bath until sampling.  At various times the bottles would be
reopened and transferred to a second vial containing an appropriate amount
of sodium thiosulfate to stop the reaction, crimped again without headspace,
and refrigerated until analyzed.

     To avoid problems with evolution of chloroform from the standard
solutions, the solution was transferred to a series of Wheaton  bottles and
crimped without headspace.  Before each series of separatory funnels was
set up, a new bottle of internal standard was removed from the  refrigerator
and used.  After opening and using, the remainder of the solution would be
discarded.

     It was found that the headspace technique for analysis of chloroform
offers distinct advantages over conventional extraction or purge and trap
techniques.  Several injections could be made from one funnel and the peak
height ratios of chloroform to internal standard are reproducible to within
2%.  Several funnels could be set up at staggered times to permit considerable
time savings over the purge and trap technique.  Like the purge and trap
technique, only peaks due to volatile species are observed and  no problem
exists with high retention solvent impurity peaks.  Due to the  non-linearity
of the ECD detector, concentrations above 50 yg/1 had to be diluted to a
suitable concentration range and the peak ratios had to be compared to a
standard curve of several known chloroform concentrations.
                                      21

-------
                         Concentration !
                            Chloroform = 40 ppb
                            1,1,1 Trichloroethane (Internal
                            Standard ) =  10.7 ppb

                         Retention  Time  !
                            Chloroform — 2.1 min
                            1,1,1 Trichloroethane - 3.25 min

                         Attenuation !
                            32 xlO
     Inject 100 y£ Gas

Figure 4.  Typical chloroform chromatogram
                            22

-------
2-METHYLISOBORNEOL, SYNTHESIS AND ANALYSIS^

     The 2-methylisoborneol (MIB) used in this investigation was synthesized
from d-camphor, and extraction, concentration, and GC analysis were used to
quantify concentrations down to 0.1 yg/1.

Reagents. Solvents and Adsorbents

     The reagents, solvents and adsorbents used in the synthesis and analysis
are given in Table 5.


            TABLE 5.  REAGENTS, SOLVENTS AND ADSORBENTS FOR MIB
                           SYNTHESIS AND ANALYSIS

              Substance                                Source


d-Camphor, mp 178-180°                  Eastman Kodak Company, Rochester, NY

Methyl!ithium, 1.7 M in ether           Ventron, 8 Congress Street, Beverly, MA

Ether, anhydrous analytical reagent     Mallinckrodt, St.  Louis, MO

Hydroxylamine hydrochloride, analytical
reagent                                 Mallinckrodt, St.  Louis, MO

Hexane, distilled in glass              Burdick and Jackson Laboratories,
                                        Muskegon, MI

Methylene chloride, distilled in glass  Burdick and Jackson Laboratories,
                                        Muskegon, MI

Silica Gel, 0.05/0.2 mm, non-activated  Brinkmann Instruments

Florisil, 100-200 mesh, non-activated   Fisher Scientific Supply Co.
Hater Samples

     The tap water used was that from the Civil Engineering Building, and well
water that from a well in the basement.  Surface water was that taken from a
polluted stream running through the center of Urbana and lake water that taken
from a small lake just outside the town.
1.  Much of the material in this subsection was taken from "2-Methylisoborneol
    Improved Synthesis and a Quantitative Gas Chromatographic Method for Trace
    Concentrations Producing Odor in Water," by N.F. Wood and V.L. Snoeyink,
    J. Chromatogr., 132, 405 (1977) with the permission of the copyright
    holder.

                                      23

-------
Polarimetry

     The specific rotation of both natural  and synthetic MIB was determined
on a Bendix-NPL Automatic Polarimeter which employs the Faraday electro-optic
effect to measure optical rotation.   To measure the very small  rotation
observed from the small sample of natural  MIB, the instrument had to be
operated at high sensitivity.  Full-scale  deflection of the meter read-out
needle to the right or the left then corresponded to a rotation of plus or
minus 0.1°, respectively, and rotations could be read to within 0.001°.
With 0.1% sucrose the instrument was found to have a calibration factor of
0.952.  The 1-cm cell used in this work would not fit the usual cell holder
for the instrument, so a cardboard holder  was constructed that fitted in the
recess in the base of the cell compartment.  In this way the cell could be
positioned inside the compartment in the same way every time, and a reading
from a particular solution could be  reproduced exactly.

     The specific rotation of the synthetic MIB was also determined using a
Carl Zeiss polarimeter.

Thin-Layer Chromatography

     Thin-layer chromatography (TLC) was performed on Merck precoated plates
of silica gel GOF-254 (0.25 mm).   (Stockists may supply the earlier version
of these plates from old stocks,  but results with these are distinctly
inferior in terms of sensitivity and resolution.)  Samples were spotted in
amounts up to 200 yg.  To develop the plates, suitable mixtures of ethyl
acetate and hexane were used.  Spots were  visualized by examination under
UV light, treatment with iodine vapor, and  by spraying with 1% vanillin in
sulfuric acid.   Sensitivity with  the spray  was 1  yg for 2-methylisoborneol,
5 yg for camphor oxime, and 10 yg for camphor.  Visualization of camphor
spots was quite sensitive to the  manner of spraying and sometimes required
several  hours of standing.  Best  results with camphor were obtained by a
very light spraying followed by a second spraying after 15 minutes.  Spray
reactions and some typical Rf values are shown in Table 6.

Gas Liquid Chromatography

     Gas liquid chromatography (GLC) was performed at 140° on a 183 x 0.2-cm
glass column containing Supelcoport  (60/80) coated with 3% SP2100 using a
F & M instrument (model 810) equipped with  a flame ionization detector.  The
glass column was treated initially with 10% dimethyldichlorosilane in
toluene for two hours, washed with methanol, dried, packed with the stationary
phase (Supelco, Inc., Bellefonte, PA), and  conditioned at 300°  for two hours.
Flow rates were:  nitrogen carrier,  35 ml/min; hydrogen, 30 ml/min; and air
350 ml/min.   Injections were made on-column at 270° with a Hamilton 10-yl
syringe set at  2-yl  with a Chaney adapter.   Typical retention times were:
camphor, 1.4 min; 2-methylisoborneol, 1.7  min; and camphor oxime, 3.4 min.
                                      24

-------
           TABLE 6.  TLC AND MIB AND SOME OTHER CAMPHOR DERIVATIVES
              Compound
                                         Spot formed with 1%
                                         vanillin in sulfuric acid
Developing solvent - ethyl acetate:hexane (1:4)

                                    0.30
Camphor Oxime


Borneol
        Unknown for HBr on
        2-methylisoborneol

        Isoborneol
        2-Methylisoborneol


        Camphor
0.34



0.38

0.41


0.49


0.59
Blue-gray, developing
slowly

Sharp translucent,
developing slowly

Bright crimson
developing immediately

Yellow-brown, turning
blue

Bright crimson,
developing immediately

Translucent, developing
very slowly
     GLC proved invaluable for analyzing and for monitoring the formation of
the camphor oxime, the course of liquid-liquid extractions, and the eluants
from the silica gel and florisil columns.  Before injection, small  aliquots
of both aqueous and organic solutions were diluted 1:250 with hexane.

Specific Rotation of Natural MIB

     Natural MIB investigated in this work originated from Streptomyces sp
CWW3 isolated from Lake Michigan.  It was contained in a by-product kindly
sent to us by the Environmental Protection Agency from the production of
geosmin by Dr. Nancy Gerber of Rutgers University for the Agency (Gerber,
1974).   The by-product, a dark tar, was found by GLC to contain traces of
what were probably olefins (0.6 min and 0.7 min), about 35 mg of geosmin
(4.6 min), and only 14.0 mg of 2-methylisoborneol (1.5 min).

     The MIB was isolated by liquid chromatography of the by-product on
silica  gel.  Though probably not essential to the operation, advantage was
taken of its convenient availability to use modern liquid chromatography
equipment.  A stainless steel column (60 cm x 0.71 cm id) was packed with
silica  gel and connected with stainless steel fittings to a Waters  pump
(model  M 6000 A), through a Waters loop injection (model U 6 K).  Methylene
chloride was pumped through the column at 2 ml/min and the by-product was
injected in about 1 ml of the same solvent.  Fractions of 10 ml were collected
and analyzed without dilution by GLC; results are shown in Table 7.  Total
                                      25

-------
  TABLE 7.   ISOLATION OF MIB BY COLUMN CHROMATOGRAPHY OF A NATURAL PRODUCT

                         Adsorbent:  silica gel
                         Eluant:  methylene chloride
                         Fractions:  10 ml
 Fraction
                Peak  Height  by  GLC	

     Unknown           Unknown            MIB            Geosmin
(r.t.T 0.6 min)    (r.t.  0.7  min)   (r.t.  1.5min)   (r.t.  4.6 min)
2
3
4
5
6
7
8
9
10
7.5 9.6
3.4 4.4
15.0
48.4
5.7
48.4
25.4
4.0
0.1
1.  r.t. = retention time
recovery of MIB was estimated as 14.5 mg.  Fractions 7-9, that were homogen-
eous in 2-methylisoborneol by GLC, were combined and found to be homogeneous
by TLC also.  Removal of methylene chloride was carried out on a rotary
evaporator with the bath temperature at 30° and the evaporation terminated as
soon as it appeared that all solvent had been removed.  The residue was a
clear gum weighing about 17 mg.

     The residue was transferred with ethanol to a quartz cell (1 cm x 1 cm
id) and made up to a calibration mark of 0.905 ml on the sidearm.  The cell
was sealed with a Teflon stopper and an air bubble in the sidearm manipulated
around the entire cell  to produce a homogeneous solution.  The observed
rotation was -0.023°, and allowing for the ethanol blank of -0.003° and the
calibration factor of 0.952, this corresponded  to a corrected rotation of
-0.019°.   The concentration of MIB in the cell solution was found to be
0.0136 g/ml after a 9.8 yl portion had been diluted to 1.0 ml  with ethanol
and analyzed by GLC.
                                      26

-------
                                 0.1
Mass Spectrometry
        The mass spectrum of MIB was recorded on a Finnigan 3300 quadrupole
instrument using the solid probe.

Infrared Spectroscopy

     Infrared spectra were recorded on a Beckman IR-20A from samples as
potassium bromide pellets (1 mg in 200 mg) or 10% solutions in chloroform.

Evaporations

     Analytical extracts of methylene chloride were concentrated using the
apparatus shown in Figure 5.  The concentration vessels were made from 25-ml
round-bottom flasks and are unavailable commercially.  Micro Snyder distil-
lation columns (Burke et^ al_. , 1966) like the one shown are supplied by
Kontes, Inc., Vineland, NJ.  The heating bath used consisted of a pan
completely filled with water held at 75°C by means of a small hot plate.
The cover was a round aluminum plate (0.5 cm x 25 cm diam.) with a small
hole in the center fitted with a thermometer and several holes (4 cm in diam.)
to accept concentration vessels.  The holes held the vessels securely around
the midpoint of the bulb and no additional support was necessary.

     Extracts were added to concentration vessels together with a small
boiling chip.  Glass joints were sealed with a thin film of water.  After
about 5 min. on the water bath, evaporation was complete and the vessel was
cooled in a stream of cold tap water or by allowing it to stand at room
temperature before removing the Snyder column.  The condensate was about
200 yl .  Where maximum sensitivity by gas chroma tog raphy was desired, this
solution was further concentrated to about 20 yl using a rotary evaporator
without a heating bath.  To avoid solutions of this small volume evaporating
to dryness on standing, about 20 yl of ethanol was sometimes added.

     Eluates from columns used in the purification of synthetic and natural
MIB were evaporated to dryness using a rotary evaporator with a bath
temperature at 30°.

Stock Solutions of MIB

     The following stock solutions of MIB were prepared:

                                             yg/ml

                            A
                            B
                            C
                            D
                                      27

-------
                                                              14/20
CO
                            14/20
                                                              12/18
                                                                                     >©<
                                                                                              14/20
                                                       B
              Figure 5.   Concentration  apparatus.   (A) and (B) are concentration vessels;
                         (C)  is  a  micro-Snyder distillation column.

-------
A solution was prepared by making 10.00 mg up to 100 ml with ethanol.   The
other solutions were prepared by successive 1 in 10 dilutions with ethanol.
A and B were mixed with similar camphor solutions to form reference mixtures,
C and D were used to make additions of MIB to water at 0.1-10 yg/1 during
recovery experiments.

Stock Solutions of d-Camphor, the Internal Standard

     The following stock solutions of d-camphor were prepared:
                            A1
                            B1
                            C1
                            D1

     Preparation was similar to that described for the MIB solutions.

Reference Mixtures of MIB and d-Camphor

     Usually solution A was mixed with A1 in the following ratios by volume:
2:1, 1:1, 1:2, 1:5, and 1:10, respectively.

Analytical Method for MIB in Water

     An exact volume of internal standard solution was added to 1 liter of
the water sample.  For an expected MIB concentration of 1-10 yg/1, 5.0 ml  of
C' was usually used; for 0.1-1 vg/1, 5.0 ml of D1 was usually used.   The
solution was extracted with methylene chloride (1 x 25 ml  and 1 x 10 ml) and
the extracts run directly into a concentration vessel.  Total volume of the
combined extracts was 20 ml since about 15 ml of methylene chloride  dissolved
in the water.  After concentration to about 20-200 yl, the extracts  were
analyzed by gas chromatography using a reference mixture with a similar
mixture of MIB to camphor.  Peak height ratios were determined by dividing
the peak height of the MIB by that of the camphor.

     Now assuming response of MIB relative to camphor is constant:


                   Wl
             hl =kw7


                    W
where:  h, and h2 are peak height ratios for sample and reference mixture,
        respectively.

             W-i is the weight of MIB in sample.



                                      29

-------
             w-i is weight of camphor added to sample.

             W2 is weight of MIB per ml  of A used  to make reference mixture.

             w2 is weight of camphor added to reference mixture per ml  of A.

             k is a constant

Eliminating k and rearranging:


                  hl   w
             wi = TT1
               I   ru


                -
                  h  x  cr x  2
where:  c is the concentration of camphor/ml  of A'

        d is the dilution factor for the particular stock solution of
        camphor used with the sample.

        v is the number of ml of camphor stock solution added to sample.

        r is the ratio of A1  to A in the reference  solution.
                                                                     °'01  x 5
Thus when 5 ml of C1  (d = 0.01)  was used:   2-methylisoborneol  = T~ x
        hl   5                                                   2 hl
x 100 = fp x 1 ug/1.   When 5 ml  of D1  (d = 0.001)  was used:   MIB = ^- x

         2                                                          2

Synthesis of MIB
     d-Camphor (60 g) in about 80 ml  of ether was added dropwise with
magnetic stirring to about 260 ml of 1.7 M methyl lithium in ether at such a
rate as to maintain the solution under gentle reflux.   Before use, the
reaction flask was flushed with dry nitrogen fed in through a three-way
stopcock in the top of the reflux condenser in one arm and out through the
dropping funnel in another arm.  Since d-camphor dissolves readily in ether,
the solution was prepared in the dropping funnel, so as to avoid exposure of
the ether to the atmosphere.  After completion of the  camphor addition, the
solution was refluxed for a further 2 hours and allowed to stand at room
                                     30

-------
temperature overnight.  The solution, containing a white precipitate, was
then poured onto about 400 g of crushed ice and adjusted to pH 6 with glacial
acetic acid.  The ether layer was collected by decantation, and the aqueous
layer was extracted with more ether (2 x 10 ml).  The combined ether solutions
were dried over sodium sulfate and evaporated to dryness using a rotary
evaporator.

     To the residue in 200 ml of ethanol was added 40 g of hydroxylamine
hydrochloride in 100 ml of water and 64 g of sodium hydroxide in 100 ml of
water to yield two clear layers of approximately equal volume.  The solution
was refluxed for 8 hours and allowed to stand at room temperature overnight.
Water (about 230 ml) was then added with shaking,resulting at first in the
formation of a homogeneous solution and finally in the appearance of a
slight permanent precipitate.  The solution was extracted with hexane
(2 x 100 ml) and the combined extracts washed first with 2N sodium hydroxide
(15 x 500 ml) and finally with water (2 x 50 ml).  The hexane was dried over
sodium sulfate and evaporated on a rotary evaporator to yield 24.87 g (37.5%)
of MIB as a white crystalline solid.

     The product was purified by chromatography on a column of silica gel
(6 x 34 cm, 500 g) using methylene chloride as eluant.  Eluation of MIB
(23.04 g) commenced after 1 liter of eluate had been collected and was
complete after 3 liters.  Only the first and last 3 g of eluted material
showed trace impurities by gas chromatography.  The remaining material
was homogeneous by both gas and thin-layer chromatography.

     The mass spectrum of the product was closely similar to that already
published (Medsker ejt al_., 1969) with a parent peak at m/e 168 and a very
strong base peak at m/e 95.  The infrared spectrum was also closely similar
to that published (Medsker e_t al_., 1969).  Specific rotation at 20° for the
D line using the Carl Zeiss polarimeter was -14.7° (c!5.5, ethanol) and
using the Bendix polarimeter was -14.9° (clO.76, ethanol) and -3.2° (clO.O,
hexane).


GEOSMIN SOURCE AND ANALYSIS

     The geosmin was supplied by the U.S. EPA Municipal Environmental  Research
Laboratory in Cincinnati.  Gerber (1974) used Streptomyces sp CWW3 to
produce this compound for the EPA.  The microorganism was grown at 28°C
in a broth medium and the resulting mixture was steam distilled and the
distillate extracted with methylene chloride and purified by column and gas
chromatography.  This species of actinomycetes yielded both MIB and geosmin.

     The analytical  procedure for determining geosmin concentrations was
almost identical to that used for MIB.   The only differences were the use
of 2-chloronaphthalene as the internal  standard and a gas chromatograph
column temperature of 140°C.
                                      31

-------
CHLOROPHENOLS

     2,4-Dichlorophenol  (2,4-DCP)  and 2,4,6-trichlorophenol  (2,4,6-TCP)
(Eastman Kodak) were used for the  chlorophenol  studies.

Chlorophenol Analysis

     Analysis of chlorophenols was performed on a 5750B  Hewlett Packard gas
chromatograph equipped with a pulsed Nib:i  electron capture detector.   A
one-foot coiled glass column packed with 10% DEGS on 80/100 mesh Supelcoport
(Supelco, Inc.) was used for the separation of 2,4-DCP and 2,4,6-TCP.   The
column was conditioned three hours at 200°C with a flow  of 20 ml/min  carrier
gas.  Operating conditions were:  Column temperature = 170°C, Detector
temperature = 230°C, Injection port temperature = 200°C, Carrier gas  = dried
5% CH4-95% Argon at 50 ml/min flow, Pulse  interval  = 150 ysec.
2,4-Dibromophenol  (2,4-DBP) obtained from  the Aldrich Chemical  Company was
selected as the internal standard  since it is eluted close to and completely
separate from 2,4-DCP and 2,4,6-TCP at these operating conditions; at  a given
concentration, the detector response is intermediate to  these two chloro-
phenols; and its physical properties are very similar to the chlorophenols
as shown in Table 8.


      TABLE 8.  CHARACTERISTICS OF THE CHLORO-  AND BROMOPHENOLS STUDIED
                    (Handbook of Chem.  & Physics, 1967)
Compound


M.W.


m.p.
°C

b.p.
°C

Solubility
g/100 ml
H20
PKa



2,4-DCP
2,4,6-TCP
2,4-DBP
163.01
197.46
251.92
45
68
40
210
244.5
238.9
0.46
0.08
0.19
7.85
6.00
7.30
     Nanograde toluene (Mallinckrodt)  was  chosen  as  the  solvent for sample
preparation, for GC analysis,  and for  extraction  of  the  chlorophenols from
aqueous solution for several  reasons:   the solvent peak  does not overlap
with any of the peaks of interest nor  are  there any  interfering impurities;
it is practically immiscible  with water; and  the  partition coefficients for
extraction of chlorophenols from water are high enough to indicate adequate
recovery (Korenman, 1974).

     Stock standard solutions  containing 1 yg/yl  2,4-DBP and another
containing 1 ug/yl  of both 2,4-DCP and 2,4,6-TCP  were prepared  as  outlined
by the U.S.  EPA (1971) for organic pesticides.  Working  standards  were
prepared from the stock solution using a micro-syringe and then stored at
                                      32

-------
5°C.  A typical working curve for 2,4-DCP is shown in Figure 6.  Ratios of
the peak heights of the chlorophenols to that of 75 yg/1 2,4-DBP were computed
and plotted versus chlorophenol concentration.  The chromatogram in Figure 7
showing 80 yg/1 of both 2,4-DCP and 2,4,6-TCP and 75 yg/1 2,4-DBP is typical
of the data used to obtain the curves.  Day-to-day variation in detector
response and column performance was significant enough that standard curves
had to be prepared prior to each set of chromatographic analyses.

     Extractions were performed using 100 ml of the adsorbate solutions
adjusted to pH 2.0 to ensure that the less soluble neutral species were
present.  When the chlorophenol concentration was high, dilutions were
prepared  to ensure peak heights fell on the standard curves.  Prior to
extraction, 7.5 yl of 50 mg/1 2,4-DBP in distilled water was added to the
solution.  Extractions were performed in 250 ml separatory funnels with 5 ml
toluene resulting in a concentrating factor of 20X.  The solutions were
shaken for five minutes as recommended by Korenman (1974), and 30 minutes
were allowed for complete separation of the phases.  Recoveries of the
chlorophenols relative to 2,4-DBP were found to be 100 percent for 2,4-DCP
and 104 percent for 2,4,6-TCP in the concentration range studies.  For this
procedure the estimated sensitivity limits for 2,4-DCP and 2,4,6-TCP were
found to be 1 yg/1 and 0.01 yg/1 in water, respectively.


POLYNUCLEAR AROMATIC HYDROCARBONS

     Various methods were evaluated to determine the concentration of the
polynuclear aromatic hydrocarbons (PAH), anthracene (Aldrich) and benz-
anthracene (Eastman).   A fluorometer (Turner Model  110) produced a linear
response with concentration.   This procedure was satisfactory for pure
solutions of PAH, but in the presence of humic substances which fluoresce,
the contribution to fluorescence by PAH alone would be impossible to deter-
mine.   Early attempts were made to extract the PAH from humic acid solution
with toluene and to measure the fluorescence of the extract.  Some of the
humic material  was also extracted, however, and good results using fluor-
escence were not obtained.   When the aqueous solution was made basic to a
pH of 11, less  material was extracted into the organic phase.  It was also
found that UV absorbance did not offer adequate sensitivity; a 5 cm path
length cell was satisfactory for the determination of 20 yg/1 of benzanthra-
cene as a lower limit, for example.

     Gas chromatography was chosen for the analytical determination of PAH
in solution with humic acid.   The solution was extracted with cyclohexane
(Mallinckrodt)  after being made alkaline with several drops of concentrated
NaOH to minimize the extraction of humic substances.   Cyclohexane was used
because it tended to produce less of an emulsion.   The SP-2100 column used
for the quantitation of MIB and geosmin was found to be suitable for the
PAH work.  A variety of internal standards was checked, among them carbazole,
phenanthrene, fluorene, fluoranthene, 2-methyl naphthalene, 9-methyl anthra-
cene and trans-stilbene.  Fluoranthene eluted between anthracene and benz-
anthracene and  thus showed promise as an internal  standard if competitive
experiments between these two PAH compounds were performed.  Fluorene was
masked by the solvent front and carbazole and phenanthrene eluted too close


                                      33

-------
    O.7
    0.6

CL-
OD
O

*  0.5
CM
O.
O
O
CM

O
    0.4
•55  0-2
JC.
O
0)
CL  0. I
Stondord  Curve For
2,4— Dichlorophenol
(2,4 — Dibromophenol =
                                        I
I )
              20       40      60       80      100     120      140      160

                       Concentration  2,4—DCP  in   Toluene,  /ig / I
      180
                    Figure 6.  Standard curve for  2,4-dichlorophenol.

-------
                 2,4,6-TCP
                        2,4-DBP
Retention Time 1
  2,4-DCP =0.75min.
  2,4,6-TCP - 1.35 min.
  2,4-DBP =1.85 min.
Concentrations  !
  2,4-DCP = 80 fig//
  2,4,6-TCP  = 80 fig//
  2,4-DBP = 75 fig//
    t          t
 Inject     Atten. = 10x16
 Atten.  10x4
Figure 7.  Typical chlorophenol chromatogram
                           35

-------
to the anthracene peak.  Trans-stilbene eluted after the solvent peak and
before anthracene and was a logical choice.   A portion of it was lost during
evaporation, however, possibly due to partial isomerization to the cis-
isomer.  9-Methyl anthracene (Aldrich) which eluted after the anthracene
was used instead.  Tests were performed with standard solutions and it was
found that extraction of a 75 ml  sample successively with 10 ml, 5 ml and
5 ml volumes of cyclohexane removed 99 percent of anthracene from solution.
It was also shown that removal of solvent with the micro Kuderna-Danish
(K-D) evaporator resulted in no absolute losses of PAH.

     A 6 foot, 3 percent SP-2100 (Supelco) column was used with a nitrogen
carrier gas at a flow rate of 40 ml/min.  The temperature settings were
210° for the column, 260° for the injector,  and 280° for the detector.  At
these conditions, retention times for the anthracene and methyl-anthracene
were 1.9 minutes and 2.5 minutes, respectively, as shown in Figure 8.

     Solutions of anthracene to be analyzed  were made alkaline to pH of 11
and extracted with three volumes of cyclohexane.  The cyclohexane phases
were combined and spiked with a known amount of internal standard.  This
was then concentrated to about 200 yl in a micro K-D evaporator.  If further
concentration was necessary it was done on a rotary evaporator without a
heating bath similar to the MIB analytical procedure.


ADSORPTION TEST PROCEDURES

Isotherm Tests

Humic Substances --

     The tests were conducted by adding an accurately weighed dose of carbon
to a series of 250 ml bottles containing solution of known concentration.
The solution volume was 150 ml.  The bottles were placed on a shaker for 4-7
days at room temperature, after which time the equilibrium solution concen-
tration was determined.  Several  identical samples were  analyzed daily to
verify that equilibrium was achieved.

     A buffer was not used for nonfractionated commercial humic acid or well
water,  but it was necessary to use it for those humic substances which were
fractionated using a phosphate eluant.  In order to prepare a phosphate
buffered solution at a fixed concentration of phosphate  when the organics
also contain phosphate, the Vanadomolybdophosphoric Acid test was employed
to determine the proper amount of phosphate  to add (Standard Methods, 1975).
The phosphate concentration of the solution  containing only the organics was
measured and the additional required phosphate was then  added.  Equilibrium
concentrations were determined by fluorescence or, in some cases, UV
absorbance.

MIB and Geosmin --

     Stock MIB and geosmin solutions were made by dissolving the compound in
ethanol  and then diluting with deionized distilled water.  The desired

                                     36

-------
                       Retention Time !
                          Anthracene ! 1.87 min.
                          Methyl Anthracene ! 2.5 min.

                       Concentrations !
                          Anthracene =640mg/j?
                          Methyl Anthracene = 965 mg/Jl
   t
Inject  2fj.f  Atten.  8x10^

Figure  8.   Typical anthracene  chromatogram,
                          37

-------
quantity of solution was prepared in this manner immediately before starting
the tests.  MIB is very stable in ethanol and the ethanol thus introduced
into the adsorption test  systems, on the order of 1  to 5 mg/1, had no notice-
able effect on the isotherms as shown by comparing isotherms prepared using
different ethanol  concentrations, consistent with its poor adsorption charac-
teristics.
     Most of the batch data for MIB and geosmin were  obtained using 1 liter
samples in 2-1/2 liter cylindrical bottles.   The tests were generally con-
ducted by setting up a series of bottles each with the same concentration of
adsorbate but with different doses of carbon.  Blanks were also used which
contained no carbon.  It was determined that 8-9 days were necessary for
equilibration at the yg/1 concentration level on the  gyratory shaker while
only 4-5 days were required when there was more vigorous shaking on a
reciprocating shaker or if the bottles were tilted about 30° from the
horizontal on the gyratory shaker.  Other tests were  conducted by putting a
large dose of carbon into a single bottle with a high concentration of
adsorbate.  After equilibration, the solution was decanted and another
solution containing adsorbate was contacted with the  carbon and the sample
was then re-equilibrated.  This process was  repeated  several times to
establish an adsorption isotherm.

     A considerable amount of data were gathered for  MIB using a procedure
which did not result in equilibrium.   Some of these data are presented in
this paper because they are important when carbon is  used under nonequi-
librium conditions.

Chlorophenols --

     Chlorophenol  stock solutions were prepared just  prior to use by dis-
solving a known amount of compound in distilled water adjusted to about pH 9
with NaOH to ensure that the compound was dissolved.   2,4-DCP and 2,4,6-TCP
are weak acids and were studied in both their dissociated and undissociated
forms.   To ensure the predominance of either the neutral  or aniom'c form,
the pH of the solution was adjusted to at least one unit below or above the
compound's pKa value (Ward and Getzen, 1970).  The solutions were buffered
with 10-2 M phosphate salts.  A series of 2-1/2 liter bottles were filled
with one liter of the adsorbate solution.  Amounts of Filtrasorb 400 carbon
were added to each bottle to yield the desired equilibrium concentration
with at least two  bottles containing  no carbon to serve as blanks.   The
bottles were sealed with Teflon-lined caps and placed on a model  G10 gyra-
tory shaker (New Brunswick Scientific) operated at 200 rpm.   The shaker
compartments were  angled to ensure adequate  mixing.   At least 12 days were
allowed for equilibration.   All studies were performed at room temperature
which varied ±2°C.   Prior to extraction and  analysis, the solutions were
filtered through fiberglass filters to eliminate carbon fines.   Studies were
performed which indicated that chlorophenols did not  adsorb on the filter.

Column  Tests

     MIB and geosmin breakthrough curves were developed using continuous
flow column systems and deionized water.   Two grams of 40 x 50 mesh carbon
(unless stated)  were placed in 1.27 cm diameter columns.   In some cases, the


                                      38

-------
carbon was sandwiched between several  cm of sand,  also of 40 x  50  mesh  size,
although this had little effect on column performance.  Bed  depth  was 2.7  cm
and empty bed contact time was on the order of 0.1  minute.   The flow  rate  of
2 liter/hr (6.5 gpnvft2) was kept constant by use  of a metering pump.   The
tests were conducted at room temperature of approximately 23°C.  Samples were
collected on a regular basis and then analyzed to  obtain a record  of  effluent
concentration vs. time.
                                     39

-------
                                  SECTION  5

                           RESULTS AND  DISCUSSION


HALOFORM FORMATION POTENTIAL OF THE HUMIC  SUBSTANCES

     The haloform formation potential of the  humic substances  used in this
study was determined to further characterize  these materials.   In  Figure 9
chloroform formation with  time by the various  humic substances is  shown.  It
is interesting to note that not all  humic  materials will  produce the same
amount of chloroform per unit weight of TOC.   The  yield  of chloroform from
the soil humic acid is lower than from  the soil  fulvic  acid while  soil
fulvic acid and leaf fulvic acid give nearly  the same  result.   The large
difference between commercial  humic acid and  soil  humic  acid illustrates
that the source of the material  is important.

     In Figure 10 we observe that various  molecular weight fractions of soil
fulvic acid yield about the same amount of chloroform  as  the unfractionated
soil fulvic acid.  This would seem to indicate that molecular  weight frac-
tions have the same density of those functional  groups  which produce chloro-
form as the unfractionated group of compounds.   In Figure 11,  similar
results are shown for soil  humic acid fractions.  The  differences  between
the curves are not considered significant.

     Another experiment was performed to evaluate  the  removal  of haloform
precursor by activated carbon.  One liter  of  5 mg/1 TOC  commercial humic
acid was equilibrated with an amount of activated  carbon  to give about 50
percent removal as indicated by fluorescence.   After one  week  the  blank
solution was diluted to give an identical  fluorescence  reading as  the
equilibrated solution and  both were chlorinated using  the above procedure.
After 8 hours, the chloroform concentration of the diluted blank was
115 yg/1.  The chloroform  yield of the  carbon equilibrated solution was
198 ug/1, a 72 percent increase over the blank.  Because  it has been shown
that the various molecular weight fractions have the same haloform formation
potential as the unfractionated material,  this finding  is consistent with
our adsorption results which show that  carbon preferentially removes the low
molecular weight compounds which are also  the most highly fluorescent.   Thus
the decrease in fluorescence is not matched by an  equal  decrease in chloro-
form formation and is not  matched by an equal  decrease in TOC.

ADSORPTION OF HUMIC SUBSTANCES

     The results of batch  studies using commercial humic  acid  are  shown in
Figure 12.  The data are based on analysis by fluorescence and UV  absorbance,
                                     40

-------
    300
 -  200
c
o
c
V
o
c
o
O
e
t_
o

o

jO
£L
O
100
                10  mq/J(  C\2

                2.5 mg/Jj TOC

                pH =7.0
                                             Commercial Humic Acid

                                                       	

                                                       Soil Fulvic Acid
0
1 1 1 1 1 1
) 10 20 30 40 50 60 70 80
1
90 10
                                            Time ,  hours



                 Figure  9.   Chloroform formation from humic and fulvic  acids.

-------
 o»
 c
 o
c
0)
o
c
o
o
E

o
H—
O

_o

JC.
u
              10 mg/   CI2

              2.5 mg/!l TOO

              pH = 7
Unfroctionated

MW > 50,000

MW > 5000

MW < 5000
                                                      60
           7O
80
               Figure 10.
                   50



               Hours



Chloroform formation  from  soil  fulvic acid.
90
100

-------
-pi
GO
                  240r

                  220

                  200

                  ISO

                  160

             ppb "40
0  Unfroctionoted
•  MW > 50,000
•  MW < 10,000
                                                                        10 mg/5?  CX2
                                                                        pH = 7
                                                                        TOC =2.5 mq/j
                                                                                                   IOO
                                Figure 11.  Chloroform formation from soil  humic acid.

-------
       1000
    o>
    o»
    E
100
    o
    o
    00
         10
                A  F- 400, pH 5.5 (FLUOR.)
                D  F- 200, pH 8.5 (FLUOR.)
                O  F-200, pH 5.5 (FLUOR.)
                V  F-200, pH 5.5 (UV)
             0.01
                        O.I                  I                 10
                        Equilibrium  Concentration,  mg/X
100
Figure 12.  Adsorption  isotherms for commercial humic acid  measured by fluorescence and UV.

-------
as indicated.  F-400, which has a larger average pore size and about 200
m2/g more surface area than F-200, showed a much higher capacity for humic
acid.  The data show no difference in the adsorption isotherm using F-200 as
the pH is changed from 5.5 to 8.5, however.  The analytical  technique used
to determine humic acid concentrations is shown to have a very significant
effect on the results.  The material which absorbs UV light adsorbs less
strongly than the material which fluoresces.

     Figure 13 shows data for adsorption of organic matter on F-200 from
well water compared to the isotherm for commercial humic acid reproduced
from Figure 12.  There appears to be no significant effect of pH on adsorp-
tion of humic material from the well water.  There is also more fluorescent
organic matter removed per gram of carbon from well water than from the
commercial humic acid solution.  Based upon our results presented earlier
which showed lower molecular weight material  to fluoresce more and be
adsorbed to a greater degree, the presence of a greater number of low molec-
ular weight compounds is indicated.   However, this remains to be demonstrated

     Figure 14 demonstrates the effect of phosphate buffer concentrations on
the soil fulvic acid adsorption at pH 7.  The higher buffer concentrations
yield higher capacities, the specific reason  for which is not apparent.

     Figure 15 demonstrates the effect of pH  on soil fulvic acid adsorption
in the presence of 0.001 M phosphate buffer.   Due to the acidic character
of the natural organics, a decrease in pH renders them less soluble and
thus more readily adsorbed on activated carbon.  This rather significant
effect is in contrast to the absence of a pH  effect in the pH 5 to 8.5 range
reported previously for commercial and well water humic substances.  A pH of
7 was chosen for all subsequent isotherms.

     Figure 16 compares the adsorption of the various humic materials on
F-400.  It is interesting to note the different capacities.   The soil humic
acid adsorbs better than the soil fulvic acid due to its lower solubility
in water.  This cannot be generalized to all  fulvic and humic material,
however, as the leaf fulvic acid was adsorbed best of all.  It should be
able to penetrate smaller pores that are inaccessible to the humic acid
molecules.

     Figures 17, 18 and 19 show the extent of adsorption of various molecu-
lar weight fractions on adsorption capacity.   In all cases we observe a
steady decrease in extent of adsorption as the molecular weight of the
material increases.  This probably is due to  the inability of the large
molecules to enter the smaller pores of the carbon.
                                      45

-------
 O>   10
X.
 o»
 E
 c
 o
O
C
o
O
4)
O
O
«4—
W
3
CO
Well Water Organic Matter
pH 5.5 and  10
                                     Humic Acid  (From Fig. 12}
                                     pH 5.5 and 8.5
                      F - 200  Analysis By Fluorescense
                    l                               10

                  Equilibrium   Concentration, mg/fl


      Figure  13.   Adsorption of well water organic matter.
                                 46

-------
   100
o>
E
c
o

-------
                 IOO
-pi
oo
                  to
                                                                  p H = 3.0
                             F—400

                             0.001 M  PO.
                 O.I
                  .01
                                       O.I
10
100
                          Figure 15.  Effect  of pH on the adsorption of soil fulvic  acid.

-------
    100
                                             Leaf Fulvic Acid-
o>
X
o>
E
c
V
u
c
o
o
0)
u
o
     10
                                       -Soil Humic Acid
                                       Soil Fulvic  Acid
                              -Commercial  Humic Acid
                                               p H =7.0

                                               P04 = O.OOIM

                                                F -400
      0.01
O.I                      I                       10


 Equilibrium   Concentration,  mg/J  As  TOC
                                                                                                    100
                    Figure 16.   Adsorption of  various types of humic  substances,

-------
           IOO
en
O
        o>
        E
o


i  ">
4>
O
C
o
O
        o
        o
                    p H = 7.0

                    P04=O.OOI M

                     F -400
             O.Ol
                                           Unfractionaled




                                           MW  < 5000


                                           MW  >5000





                                           MW  > 50,000
O.I                      I                      "0


 Equilibrium   Concentration,  mg/*  As  TOC
                                                                                                  too
                      Figure  17.  Adsorption  of  molecular  weight  fractions  of soil  fulvic acid.

-------
    100
 o>
 E
o


I


«
u
c
o
o
Q>

U

O
•*-
w

=J

in
10
     pH  =7.0

     P04 =0.001  M

      F  -400
                                     MW < 10,000-
                                                               Unfractionated



                                                             MW  10,000 —50,000
                                                             MW > 50,000
      O.Ol
                        O.I                       I                      10


                         Equilibrium   Concentration ,  mg/>  As   TOC
                                                                                                 100
                Figure 18.   Adsorption of molecular weight fractions of soil humic  acid.

-------
    100
 O»
^s

 o>

 E
 c
 o
     I0
u

o
O
o
o
o
                                     Unfroctionated



                                  MW < 5000



                                     MW > 5000
                                       p H =7.0

                                       P04 =0.001

                                         F -400
                                                    I
      0.01
O.I                     I                      10


Equilibrium   Concentration ,   mg/J  As   TOC
100
                Figure 19.   Adsorption of molecular weight fractions of leaf fulvic acid.

-------
MIB SYNTHESIS AND ANALYSIS1

Enantlomeric Form of Natural 2-Methylisoborneol

     MIB can exist in two optically-active forms with the current Chemical
Abstracts names of (1-R-exo_)- and (1-S-expJ- 1,2,7,7-tetramethylbicyclo
[2.2.1] heptan-2-ol as shown in Figure 20 (Personal communication from D.
Weisgerber, Chemical  Abstracts Service, 1976).  These are derived from d-
and 1-camphor, respectively (Figure 20).  In the literature natural MIB has
always been depicted as the S^ form, although no evidence for this has ever
been reported.  Since we were interested in studying not only the adsorption
of the natural form on carbon beds but also in examining for any concurrent
biological activity,  we needed to determine which of the two forms occurred
naturally so as to synthesize the appropriate one for our experiments.

     Accordingly, we have measured the specific rotation of a natural sample
of MIB and so determined its enantiomeric form.  We isolated the sample by
liquid chromatography on silica gel of a product obtained by Dr. Nancy Gerber
of Rutgers University from the culture of Streptomyces sp.  CWW3 (Gerber,
1974), for the U.S. EPA and supplied to us by the U.S. EPA Municipal  Environ-
mental Research Laboratory-  The product was mostly geosmin and only  14 mg
of pure MIB was obtained.  By measuring the optical rotation in ethanol to
within 0.001°, a specific rotation of -14° was obtained, in good agreement
with the -14.8° reported in the literature (Malkonen, 1964) for MIB obtained
from d-camphor.  Thus natural MIB exists in the R form.

Synthesis of 2-Methylisoborneol from d-Camphor

     Since the first preparation was reported in 1901 (Zelinsky, 1901) MIB
has been prepared many times by the action of methylmagnesium halides or
methyllithium on camphor (Medsker et al., 1969; Rosen et al., 1970; Malkonen,
1964; Capmau et al. ,  1968; Fieser and Ourisson, 1953; Zeiss and Pease, 1956).
The difficulty with this seemingly easy preparation is that much of the
camphor reacts to form an enolate and during the work-up this reverts to
camphor, which until  now has been difficult to eliminate from the product.

     Direct separation of the MIB from the camphor by chromatography  has
been reported.  Liquid chromatography on alumina (Malkonen, 1964; Capmau
et al., 1968) has been used, though it is apparently difficult to eliminate
the camphor entirely (R. G. Webb and N. N. Gerber, personal communications,
1976).  Preparative gas liquid chromatography  (Medsker et al., 1969;  R. G.
Webb and N. N. Gerber, personal communications, 1976) apparently yields a
pure product but is limited to the preparation of milligram quantities.

     Another approach (Malkonen, 1964) to the problem has been to repeatedly
1.  Much of the material in this subsection was taken from "2-Methyliso-
borneol, Improved Synthesis and a Quantitative Gas Chromatographic Method
for Trace Concentrations Producing Odor in Water," by N. F. Wood and V. L,
Snoeyink, Jour. Chromatogr. 132, 405 (1977), with the permission of the
copyright owner.

                                     53

-------
       d — Camphor
I— Camphor
                                  HO




                 2— Methylisoborneol


        R                                    S
Figure  20.  Stereochemical  structures of 2-methylisoborneol and
           camphor enantiomers.
                              54

-------
treat the crude product with more reagent until the camphor is reduced to
such a small proportion that it can be eliminated by recrystallization.

     Finally, MIB has been separated from the camphor through its chromate
ester (Fieser and Ourisson, 1953; Zeiss and Pease, 1956).  Although reduc-
tion of the ester with hypophosphorous acid or saponification yielded impure
MIB (Fieser and Ourisson, 1953), reduction with lithium aluminum hydride on
a small scale appeared to yield pure material (Zeiss and Pease, 1956).

     In our synthesis of MIB we chose methyl lithium as reagent rather than
a methylmagnesium halide, partly because this was conveniently available
commercially, and partly because we hoped it would give the higher yield.
Figure 21 shows the gas chromatogram from the crude product obtained from
60 g of camphor and 1.3 equivalents of methyllithium after reaction at
reflux for 2 hours and overnight at room temperature.  Since the response of
MIB relative to camphor was 0.90, the observed peak height ratio of 0.50
corresponds to 67 percent of unchanged camphor.  When the amount of reagent
was increased to 2.0 equivalents, the observed peak height ratio increased
to 0.56 corresponding to only a slightly decreased amount of 64 percent of
unchanged camphor.  Decreasing the reaction time to 2 hours without reflux
had no effect on the amount of unchanged camphor.  Malkonen (1964) has
reported only 37 percent of unchanged camphor using 3.0 equivalents of methyl'
magnesium iodide.  Thus, although the Grignard reagents may not be as con-
venient to use as methyl!ithiurn, they appear to give better yields of MIB.

     To eliminate the large amount of camphor in our crude product we
decided to selectively react the camphor quantitatively to form a derivative
that would be more easily separated than the camphor itself.  Many reagents
were tested but most were unsuitable because of the particularly unreactive
nature of camphor.  The exception was alkaline hydroxylamine at reflux.
Using the reaction conditions of Lenz (1911) and monitoring the reaction
mixture by gas chromatography, it was found that 19 percent camphor remained
after 2 hours but reaction to the oxime was complete after 7 hours.  A chro-
matogram of the reaction product is shown in Figure 22.

     With the camphor in the form of the oxime it was rather easily elimi-
nated by taking advantage of its acidic properties.  First the alkaline
reaction solution was diluted with water and extracted twice with a small
proportion of hexane.  Gas chromatography on the residual solution (Figure
23) and the combined extracts (Figure 24) showed that less than 50 percent
of the oxime and more than 98 percent of MIB was in the extracts.  Finally
repeated washing of the extracts with aqueous sodium hydroxide gradually
removed all of the oxime without loss of MIB.  The product was essentially
pure by gas chromatography (Figure 25).

     In another experiment the reaction solution containing the oxime was
not diluted with water as above but extracted directly with hexane.  Par-
titioning into the hexane layer was reduced but selectivity improved.  Four
extractions removed a total of 92 percent of the MIB and only 20 percent of
the oxime from the reaction solution, so that complete elimination of the
oxime was achieved with far fewer washings of aqueous sodium hydroxide.
                                      55

-------
                         Minutes
Figure 21.   Chromatogram  from  the product of the action of
            methyl lithium on d-camphor.
                            56

-------
                                    o
                                    o>
                                    c
                                    k_
                                    o
                                    .0
                                    o
                                    in
                                   I CO
              0
                                    Minutes
Figure 22.  Chromatogram from the product of the action of methy11ithium
            on d-camphor after treatment with alkaline hydroxylamine.
                                   57

-------
                                  4

                                Minutes
8
Figure 23.   Chromatogram from the residual  aqueous  solution from the
            hydroxylamine reaction after hexane extraction.
                                 58

-------
                              4         6

                            Minutes
Figure 24.   Chromatogram from the hexane extract after the
            hydroxylamine reaction.
                            59

-------
                  0
8
                                     Minutes
Figure 25.   Chromatogram from the hexane extract in Figure 24 after
            exhaustive washing with 2 N sodium hydroxide.
                                60

-------
     Thin-layer chromatography in addition to gas chromatography was used to
monitor the above work.  Table 6 gives details for MIB and related camphor
derivatives.  Note that the bright crimson color reaction of MIB is quite
characteristic and develops immediately with a sensitivity of less than 1 yg.
Unfortunately camphor gives only an uncharacteristic translucent spot that
develops erratically over a long period with a sensitivity of 10 yg at best.
(Other sprays for camphor give even poorer sensitivities [Stahl, 1969].)
Camphor oxime does give a characteristic color reaction but this develops
only slowly and again with a sensitivity of 10 yg.   Despite the  limitations
with camphor and its oxime, thin-layer chromatography was useful in that it
detected impurities in the MIB not revealed by gas  chromatography, and thus
indicated the need for final purification of the product by liquid chroma-
tography.

     Liquid chromatography on alumina has usually been used for  the purifi-
cation of MIB, but our attempt to use basic aluminum oxide (Woelm) led to
complete decomposition.  This also occurred when an old preparative column
of silica gel was used.  Even on Florisil  there appeared to be  slight
decomposition as elution was delayed for 10 hours.   This  was  evidenced by the
appearance of several very slow developing translucent spots on  thin-layer
chromatograms of the product.  However, MIB homogeneous by thin-layer
chromatography was obtained by chromatographing on  fresh silica  gel  and
eluting with methylene chloride within 2 hours.

Attempted Preparation of Electron-Capturing Derivatives from 2-Methyliso-
borneol

     Since we were interested in analyzing aqueous  solutions of  MIB down to
0.1 yg/1 by gas chromatography, the use of the sensitive electron-capture
detector seemed appropriate.  Accordingly, attempts were made to prepare a
suitably electron-capturing ester.

     Heptafluorobutyric anhydride alone did not react significantly with
MIB at room temperature, and even at 60°, 12 hours was required before most of
the MIB had reacted.  The product was a mixture of  two substances whose
retention times by gas chromatography relative to that of MIB (0.56 and
0.60) indicated that they were olefins.  In benzene in the presence of tri-
ethylamine at room temperature, reaction was virtually complete  in 1 hour but
a similar product resulted.  Thin-layer chromatography showed only one major
spot (Rf 0.94), which had a bright crimson appearance.  After 24 hours the
initially predominant substance (relative retention time 0.60) appeared to
have partially converted to the other substance (relative retention time
0.56).

     Hydriodic acid (49 percent) reacted immediately with a solution of MIB
in acetone to give a product that showed a single crimson spot (Rf 0.96) on
thin-layer chromatography.  The product in chloroform showed infrared
absorption at 3080 (medium), 1740 (weak), 1655 (medium), and 880 cnT1
(strong) corresponding to a methylene with standard absorption at 3080
(medium), 1800 ^ 1750 (medium), 1655 (medium), and  890 cm'1 (strong)
(Nakanishi, 1962).  Again an olefin was produced rather than the desired
ester.

                                     61

-------
     In contrast, hydrobromic acid (22 percent) reacted almost completely
with MIB in ethanol in 30 min at room temperature to produce little olefin.
The major product had a gas chromatographic retention time relative to that
of methylisoborneol of 1:10 and gave a crimson spot on thin-layer chromatog-
raphy with a slightly smaller Rf (0.42) than that of MIB (0.49).  The major
change in the infrared absorption in potassium bromide was a shift of the
C-0 stretching band from 1095 cnH in MIB to 1030 crrf'  in the product, and
this indicated transformation of the tertiary alcohol to a secondary
(Nakanishi, 1962).  Both the chromatographic and the infrared evidence
indicate rearrangement of the MIB to a secondary alcohol, possible 4-methyl-
isoborneol (Toivonen, 1968).

     These failures to develop a suitable derivative to use with the
electron-capture detector meant that we were forced to use the flame ioniza-
tion detector.  To get the desired sensitivity for trace concentrations of
MIB in water, the techniques explained below had to be developed to make
extreme concentrations of water extracts.

Extraction

     MIB resembles its precursor camphor in volatility, in gas chromato-
graphic retention times, and in having extremely high distribution ratios
for partitioning into organic solvents from water.   Camphor then quickly
became our choice as internal standard in the analysis  of MIB.

     The favorable partitioning into organic solvents meant that usually
only small proportions of extractant (1-2 percent)  were necessary to extract
MIB from water, and this helped both to reduce the amount of coextractives
and also the degree of concentration necessary before analysis.   However,
considerable concentration was still  necessary and this severely limited
the choice of solvent for extraction.  MIB co-distills  with some solvents
and with others the grade readily available has impurities that  concentrate
sufficiently to cause interference with the analysis.  Chloroform was quite
satisfactory for analysis down to about 1 yg/1  but below this solvent
impurities tended to interfere.   Methylene chloride was entirely satisfactory
and in addition was readily removed because of its  outstanding volatility.

Concentration of Extracts

     To achieve the desired sensitivity of 0.1  yg/1  for MIB in water, it was
evident that the extract from 1  liter of water would have to be  concentrated
to as little as 20 yl  before gas chromatography.   Because both MIB and
camphor readily sublime, special attention had to be given to the method of
concentration.

     Removal  of solvent through  a micro Snyder distillation column (Figure 5)
was first tried.   This method has the advantage that several  samples can be
concentrated simultaneously with little attention and at the end of evapora-
tion the appartus is self-rinsing (Burke et al.,  1966).   Concentration
vessels of the type A and B shown in Figure 5 were  used.   These  were fabri-
cated from 25 ml  round-bottomed  flasks and comfortably  held the  20 ml of
                                     62

-------
solvent from the extraction of 1 liter of water for MIB.  With type A,
access to the concentrate with a syringe was a little awkward but it was
generally preferred because of its simplicity.  Type B had to be assembled
with water as a seal on the lower joint to avoid small losses during the
concentration, but was advantageous in that the lower tube could be removed
later for easy access to the concentrate.  The smallest volume attainable
with this apparatus was about 200 yl, which was sufficiently low to allow
analyses down to about 1 yg/1.

     Both relative and absolute recoveries for the evaporation were deter-
mined using 1-10 yg of MIB in 20 ml of various solvents.  Absolute recover-
ies were determined as near as possible from the volume of concentrate as
calculated from its weight.  Disappointingly, pentane and hexane gave only
about 50 percent absolute recoveries and relative recoveries were very
erratic.  However, methylene chloride, chloroform, and carbon tetrachloride
gave quantitative absolute and relative recoveries.  Of these latter sol-
vents, methylene chloride was the obvious choice for the analysis because
of its greater volatility and availability in pure grades.

     To attain the highest sensitivity desired, it was necessary to determine
how to reduce the initial concentrate of 200 yl to about 20 ul  without loss
of MIB.  This was attempted initially in the lower tube of vessel B by
equipping this tube with a small Snyder column.  Although relative recoveries
were excellent absolute losses occurred which largely nullified the effect
of solvent removal.

     However, concentration of the initial concentrate in vessel  A or the
lower tube of B on a rotary evaporator without a heating bath was successful
without loss of MIB or camphor.  The same result was possible by simply
connecting the vessels to a water aspirator, but then there was risk of
losing the sample by bumping.

Evaporation of Column Eluates

     Before carrying out the purification of the few milligrams of natural
MIB available by liquid chromatography, losses were anticipated during the
evaporation of the column eluate to dryness.  However, preliminary tests
with synthetic MIB showed that, although 2.5 mg in 50 ml  of hexane was
evaporated to dryness on a rotary evaporator at 45° with a recovery of only
30 percent, similar evaporation at 35° using methylene chloride gave nearly
quantitative recovery.  Accordingly, the liquid chromatography was worked
out on the basis of methylene chloride as eluant.

Gas Chromatography

     The initial  parts of this work were carried out by gas chromatography
using a commercially prepared glass column of 3 percent OV-1.  First tests
on the use of camphor as internal  standard gave satisfactory results in
the determination of MIB in water at 100 yg/1.  However at 10 yg/1  unfavor-
able adsorption effects were observed with the smaller amount of internal
standard used.   Peak height response relative to that of MIB decreased
markedly and peak width and retention time increased to the point that

                                    63

-------
resolution from MIB was mostly destroyed.

     The column was therefore replaced with one of DMCS-treated glass packed
with 3 percent SP-2100, a similar stationary phase to that of 3 percent OV-1
but with improved characteristics.   No untoward adsorption effects were
noted with the column.   Providing it was  not subjected to harsh heat treat-
ment, this type of column proved to be very stable in performance over a
period of many months.

     In the quantitation of MIB in  this work, peak height relative to camphor
for the sample was compared with that of  a standard mixture showing a similar
ratio.  Standard mixtures in ethanol  were  prepared by mixing a stock solution
of MIB (10 mg/100 ml)  with that of  camphor (8 mg/100 ml)  in various propor-
tions.  When stored in stoppered volumetric flasks at room temperature, they
proved to be very stable over a period of  many months.   It was assumed that
there was a linear relationship between the amount of MIB in the sample and
the peak height ratio,  and that this ratio was unaffected by dilution of the
solution.

     To substantiate these assumptions the response factor was determined
for MIB from some standard reference mixtures and their corresponding 1:10
dilutions as shown in  Table 9.   Evidently  the response factor is not appre-
ciably affected by dilution.  Also, although there is a slight fall in the
response factor with the undiluted  mixtures as the proportion of MIB is
reduced greatly, this  is not significant enough to affect accuracy if widely
differing peak height  ratios are not compared.
       TABLE 9.   RESPONSE FACTOR OF MIB  FOR VARIOUS  PROPORTIONS  WITH
           THE INTERNAL STANDARD AND AMOUNTS GAS  CHROMATOGRAPHED

Ratio of (A/A1)
volumes MIB
stock solution to
that of internal
standard in
reference mixture^
Amount of ?
chromatographed, pg
Internal
MIB Standard
(W)
(Wi)
Peak height
of MIB
(H) relative to
that of internal
standard (Hi)
Response
Factor
H Wi
Hi W


2

1

0.2

133.3
13.33
100.0
10.00
33.33
3.333
53.33
5.333
80.0
8.000
133.3
13.33
2.15
2.14
1.04
1.05
0.196
0.211
0.860
0.856
0.832
0.840
0.784
0.844

 Stock solution of MIB (A)  = 10.0 mg/100 ml  in ethanol.
2Internal  standard solution (A1)  = 8.0 mg/100 ml
 Two microliters of each reference mixture was chromatographed both undi-
 luted and as a 1:10 dilution with ethanol.
                                    64

-------
Recoveries from Various Water Types

     Recoveries of MIB added to various waters in a small amount of ethanol
are given in Table 10.  Blanks equivalent to 0.1 yg/1 were carried out for
these waters and also for a sample of lake water, but no interference was
noted.  A typical chromatogram from a recovery at 0.1 yg/1 is shown in
Figure 26.
         TABLE 10.  RELATIVE RECOVERY OF MIB FROM VARIOUS WATERS

Added,
MIB
yg/1
Camphor
(internal
standard)
Recovery of MIB, %, Relative to Camphor
Type of Water Used
Distilled Tap Well


Polluted
Stream

10
5
1
1
0.5
0.1
4
4
4
0.4
0.4
0.8
100.0 97.8 98.2; 97.5
97.7; 94.9 97.4; 97.3
100.4 100.6 100.7
96.6
100.1
101.8
99.0
103.7




Mean relative recovery,
% ± standard deviation   100.3±0.2    98.5±2.4
98.2+1.4
101.4+3.3
ADSORPTION OF MIB AND GEOSMIN

Results
     Isotherm data for MIB are shown in Figure 27.  Included are the iso-
therms for MIB on F-200 carbon in the presence of both 10 and 100 mg/1  of
commercial humic acid.  Because of the small dosages of carbon used in  the
tests, negligible reductions in humic acid concentration were observed  and
final concentrations were approximately the same as the initial
1.   Much of the material in this subsection has been taken from "Activated
Carbon Adsorption of the Odorous Compounds 2-Methylisoborneol  and Geosmin"
by D. R. Herzing, V. L. Snoeyink and N. F- Wood, Jour. Amer.  Water Works
Assoc., 69, 223 (1977), with the permission of the copyright owner.
                                     65

-------
                                Minutes
Figure 26.   Chromatogram obtained in  a  study  of the recovery of MIB
            added to  tap water at 0.1  yg/1.
                                66

-------
                             100
CTi
                          0>
                          ^

                          o>

                          E
                          c
                          o
                          o
                          c
                          o
                          o
                           o
                           o
                              10
O D I — Dist. ,  pH  5.5

D D I - Dist. ,  pH  8.5

V 100  mg/^  Humic Acid

A 10  mg /^  Humic Acid

• Well  Water
                                                VA.
                                                               * D I = Oeionized

                                                             ** Dist = Distilled
                              0.1
                                O.I
            1.0
10
                                                                                            IOO
                                           Equilibrium  Concentration,
                                           Figure 27.   Adsorption of MIB.

-------
concentrations.  One hundred mg/1 is unrealistic for most natural  waters
(typical values are on the order of 1  to 10 mg/1) but the data are useful
for emphasizing the magnitude of the competitive effect.  As the equilibrium
concentration of MIB increases, the isotherms for MIB in the presence of
humic acid converge to the isotherm for which there is no competition.
However, the isotherm for MIB in well  water does not converge in the con-
centration range studied, probably owing to the nature of the organic matter
in the well water vs. the commercial humic acid used.

     Batch adsorption tests using F-200 carbon were conducted with geosmin
in a manner similar to that used for MIB and the results are shown in
Figure 28.  The data show that at an equilibrium geosmin concentration of
0.1 yg/1 the capacity of the carbon is 0.54 mg/g in deionized-distilled
water, twice the capacity shown in Figure 27 for MIB at the same equilibrium
concentration.   The data for adsorption of geosmin from 10 and 40 mg/1 humic
acid solution are also shown and indicate a greater degree of competition
than there was  with MIB.  In fact, at 1 yg/1 equilibrium concentration in
the presence of 10 mg/1  humic acid, the surface concentration of geosmin is
essentially the same as  for MIB under the same conditions.

     Adsorption data were also obtained for MIB and geosmin in deionized-
distilled water using a  single test bottle and sequential  addition of
adsorbate, with equilibration after each addition, as described in Section
4.  The data obtained in both cases varied only slightly from the data which
are shown in Figures 28  and 29.

     Figure 29  shows the results of several  continuous flow column experi-
ments in which  MIB was the adsorbate and F-200 was the carbon.   An obvious
characteristic  of these  curves is the manner in which they level  off rather
than converge to C/C0 =  1.  This suggests that a portion of the capacity is
being utilized  very slowly and that a long operation time would be required
for complete saturation.  This effect is probably attributable to a combi-
nation of the high flow  rate, short contact time and a slower rate of mass
transport in the small pores of the carbon.   As predicted from the isotherms,
the presence of 10 mg/1  of humic acid greatly reduces the capacity of the
carbon for MIB.  The data show very little effect due to variation in pH
from 5.5 to 8.5, also consistent with the isotherms.  The divergence of the
curve for pH 5.5 with no humic acid from the other curves  after 40 hours  is
attributed to a decreased flow rate because of column plugging.  Copper ion
which was introduced to  eliminate possible biological activity in the column
appears to have no significant effect on the breakthrough curve.

     Figure 29  also shows data for a column test in which geosmin was the
adsorbate.  The greater  capacity observed for geosmin compared to MIB is
consistent with the batch equilibrium data shown in Figures 27  and 28 which
show geosmin to be more  strongly adsorbed.

     The data used to determine the C/Co plot for geosmin in Figure 29 are
shown in Figure 30.  As  is apparent, geosmin loss from the open reservoir
was significant and amounted to more than 40 percent over the 190 hr of the
test.  By comparison, MIB is much less volatile from aqueous solution with
typical losses  amounting to less than 10 percent over 100 hours.


                                     68

-------
    100
o>
\
o>
E
o
V
o

o
o
o
s-
V-
    1.0
    0. I
      O.I
                              Dl -Dist-
                                            10 mg/Jt  Humic Acid
                                       40 rr\g/!( Humic Acid
l.O
                                               10
100
                   Equilibrium   Concentration ,
                     Figure 28.  Adsorption of geosmin.
                                  69

-------
                                              MIB C0 = 50 ug /
                                             pH 8-8.5, 2mM, HCO,
                                          O  pH 5.0, 50mg A  Cu
                                             pH 5.5
                                             pH 5-6, IOmg/2 Humic Acid
                                             pH5-6, lOmg/JP Humic Acid
Geosmin  C0  = 50 u.g / J(
           A  pH 5.5
                       80       100       120

                          Time ,  Hours
                                                                 180
Figure 29.   Column breakthrough  curves  for MIB  and  geosmin.

-------
u
c
o
o
                                               IOO


                                          Time  , Hours
150
                    200
                       Figure 30.  Column breakthrough curve for geostnin,

-------
Discussion

     Even though earthy-musty odors  are a common problem at water treatment
plants and activated carbon is commonly used to remove these odors, it is
difficult to say whether adsorption  or biological  activity, or some other
mechanism, is effecting the removal.   Our results  show that two causative
agents of earthy-musty odor, geosmin and MIB,  are  strongly adsorbed even in
the presence of interfering natural  organic matter.

     Silvey et al.  (1976) report that biological degradation of geosmin by
8a.cx££a6 ceAetM readily takes place;  because of the  biological  growth which
develops in beds of GAC it is possible that some removal of earthy-musty
odor can be attributed to it.  Nevertheless, in our  study biological  activity
did not contribute  to geosmin or MIB removal.   Most  of the MIB (70-80 per-
cent) and the geosmin (80-90 percent) was recovered  unchanged through simple
dioxane extraction  of the carbon.  No difference in  percent recovery  was
noted when low pH systems were used  with 50 mg/1 of  Cu^"1" serving as a biocide
as compared with systems near neutral pH without a biocide.  Also,  break-
through curves were essentially the  same when  Cu2+ was used as  when it was
not.

     It is not possible to conclude  that the capacity of carbon in  the
column is the same  as the carbon used in the batch tests because the  columns
were not run until  saturation was  achieved, although agreement  at saturation
is expected.  When  the column tests  were terminated, approximately  50 per-
cent of the expected saturation capacity for the column had been achieved.
Lower flow rates would have resulted in less severe  tailing for the break-
through curve but longer experimental operation times would then be required
to achieve a given  degree of saturation and this would severely limit the
number and variety  of experiments  which could  be run.

     The commercial humic acid used  in this study  significantly reduced the
amount of the two odor compounds which would adsorb, with the effect  being
somewhat greater for the geosmin.  Since humic acids are larger than  fulvic
acids and much larger than the odor  compounds, it  is possible that  a  greater
reduction in adsorption capacity will take place in  waters which have a
larger amount of fulvic acids.   The  fulvic acids can penetrate  smaller pores
and thus may interfere with adsorption to a greater  extent.  However, as
shown in the following subsections dealing with chlorophenols,  the  source of
the fulvic acid is  also likely to  be important.  Adsorption of MIB  from well
water showed a competitive effect  somewhat different than that of commercial
humic acid, and there was a strong  indication that  the organics  in well water
were more adsorbable.

     An important question to be answered when GAC beds are used to remove
trace organics concerns "unloading"  of the carbon, or appearance of concen-
trations of substances in the effluents of GAC beds  greater than are  in the
influent.  This phenomenon is commonly observed for  organic matter parameters
such as COD and CCE while the GAC  beds continue to remove odor (Love  et al.,
1973).  To determine whether MIB would appear in the effluent if the  humic
acid concentration  in the influent were suddenly increased, a column  test
was conducted, the  results of which  are shown  in Figure 31.  The bed

                                      72

-------
CO
                      1.0
                 e?   0.9
                 O
                  c
                  o
                      0.8
O   0.7

"c
0>
£   0.6
o
u


—   °-5
c


;=:   0.4
                  C
                  0)
                 ^   0.2
                       0. I
                             D I  Water With

                             MIB ,  pH  7-8
/~'0 mg/^  Humic  Acid

  Added  To MIB  Solution
                                                                             Begin  Dl Water  Only
                                   100
200        300        400       500


            Time ,  Hours
                600
                                                                                   700
                                       Figure 31.  Column breakthrough curve  for MIB.

-------
contained 5 grams of F-200 carbon in a 1.27 cm diameter column.  Bed depth
was 7.6 cm and the flow rate was 2 liters/hr.  As C/C0 approached 0.4,
10 mg/1 of humic acid was added to the influent MIB solution.  The humic acid
rapidly saturated the bed, with effluent concentration essentially equal to
influent concentration after 70 hours.  The effluent concentration of MIB
increased more rapidly, as shown, but did not exceed the influent concen-
tration.  Unloading is expected only when no adsorption capacity remains to
re-adsorb the MIB displaced by the humic acid.

     After 430 hours all  humic acid and MIB were eliminated from the influent
solution and only deionized water was applied to the column.   As shown in
Figure 31, the effluent concentration dropped very rapidly.  After passage
of more than 200 bed volumes of water only a very small fraction of the
total adsorbed MIB was released.  The column run was terminated because of
mechanical difficulties but the results indicate that MIB is  not easily
eluted from the carbon after adsorption.

     A series of tests using MIB was conducted in a manner similar to the
procedure normally used for determining isotherms except that they were not
carried to equilibrium.  All test solutions were agitated on  a gyratory
shaker for approximately 4 days.  The results are presented in Figure 32,
with the equilibrium curve for MIB adsorption from deionized-distilled water
from Figure 27 reproduced for comparison.  The data for these tests had much
more scatter than did the equilibrium data and thus only the  line of "best
fit" is shown.  A comparison of the curves for pH 5.5 in deionized-distilled
water shows nearly 10 times as much MIB is adsorbed at equilibrium as at
non-equilibrium.  It is interesting to note that changing the pH from 5.5
to 8.5 again had little effect on the amount adsorbed, and that humic acid
had a more significant effect than was observed at equilibrium conditions
(see Figure 27).  A major difference with the interference effect was that
the curves with and without humic acid did not converge at high MIB concen-
trations as in the equilibrium case.  Inclusion of 50 mg/1  Cu*+ in the test
solution as a biocide apparently increased the rate of adsorption, although
there appears to be no good explanation for this phenomenon.   As was pre-
viously noted, this effect was not observed in the column tests.

     The importance of the non-equilibrium data is that when  carbon is used
under conditions where equilibrium is not achieved, such as is likely the
case in many instances when PAC is used,  the humic substances still have a
major effect on adsorption capacity for MIB.

     The results of our study show MIB and geosmin to be much more strongly
adsorbed than natural  organic matter.  To illustrate this,  if it is assumed
that 10 yg/1 of MIB and 10 mg/1 of humic acid are present in  water entering
a 2 foot deep carbon bed at a flow rate of 2 gpm/ft^, it would take on the
order of 200 months to completely saturate the bed with MIB while saturation
with the humic acid would take place in about 1 month.  This  calculation
assumes that MIB and humic acid are present on a continuous basis and that
all of the substance entering the bed is  adsorbed until the bed is saturated.
Of course, in an actual situation some leakage would take place before the
bed was completely saturated; also, some  trace compounds may  be present
                                     74

-------
                            Dl -  Dist. pH  5.5-

                            At  Equilibrium
o>
 -  i.o

o
c
0>
o
c
o
o

-------
which also compete with MIB for adsorption sites.   Factors which would tend
to make the bed life for MIB even longer are that earthy-musty odors gener-
ally occur periodically through the year and that biological  growths may
develop in the bed which would destroy a fraction of the MIB.   Larger scale
studies are needed to verify these predictions.   All  indications are that if
removal of MIB or geosmin is the sole objective  for which GAC  beds are used,
the bed life will  be very long.  Currently,  data on MIB or geosmin distri-
bution in full scale carbon beds are needed  to verify that penetration of
these compounds through the bed is extremely slow.


ADSORPTION OF CHLOROPHENOL

     As phenol becomes substituted with chlorine,  its solubility decreases
and it is adsorbed more strongly onto carbon as  was found by  Gauntlett and
Packham (1973).  Their results also indicated that the dissociated molecules
found at pH values above a weak acid's pKa value are  less strongly adsorbed
than the undissociated form.  Ward and Getzen (1970)  observed  the same
effect of pH in a study performed on the adsorption of aromatic acids in
aqueous solution.   They found that maximum adsorption occurred near the
point where pH = pKa for each compound.  Their data suggested  that greater
quantities of undesirable solutes may be removed from waters  by the addition
of carbon at a pH level slightly below this  optimum point.

     In the research reported herein, adsorption of chlorophenols is examined
down to the threshold odor level in systems  containing single  solute,
bisolute chlorophenol, and single solute chlorophenol in the  presence of
various humic substances.   Previous research has not  examined  these aspects
of chlorophenol adsorption.

Single Solute Systems

     Equilibrium data for single solute isotherms  of  2,4-DCP  and 2,4,6-TCP
in distilled water at pH 5.2, 7.0 and 9.1  are shown in Figures 33 and 34.
At pH 5.2 and 7.0, 2,4-DCP is primarily undissociated (pKa =  7.85) while
at pH 9.1 it is primarily anionic.  At pH 5.2, 2,4,6-TCP is primarily
undissociated (pKa = 6.0)  while at pH 7.0 and 9.1  it  is primarily anionic.
The effect of pH on adsorption capacity for  an equilibrium concentration of
2 x 10~8 M chlorophenol species is shown on  Figure 35.  The results confirm
findings of Zogorski and Faust (1974) for 2,4-DCP in  the millimolar concen-
tration range.  Adsorption of 2,4-DCP and 2,4,6-TCP rises to  a maximum at
pH values near the compound's pKa and falls  off  rapidly at pH  values above
the pKa.  The peak in adsorption near pH - pKa where  the species is 50
percent ionized results because of synergistic adsorption of  the anionic
and neutral species (Ward and Getzen, 1970).  Jain and Snoeyink (1973) pro-
pose the alternative explanation that the anionic and neutral  species adsorb
at two different kinds of sites thereby increasing the total  amount of
adsorption when both species are present above that when one  species pre-
dominates .

     The results also confirm findings of Gauntlett and Packham (1973) in
that the more highly substituted the molecule is with chlorine, the better


                                     76

-------
 in
 at

 o
 E
X


 c
 o

"o

-------
CO
                   .d2
              o

              ^   ,6'
              X
              o
c
0>
o
o
o

o>  id4
o
              3
              CO
                   165
                    .
                           ,68
                                                      T = 28
                                                      0.01 M  P04
                                                      F-400
                                                                                    pH 5.2
                                                            10
,66
                                        Equilibrium  Concentration,  C   (moles
                                                                                          pH 9.1
,65
                               Figure 34.  Adsorption isotherms  for  2,4,6-trichlorophenol

-------


'b
X
5
£.
o
J
X
c
o
o
h_
c
«
o
c
o
p
o
o
•*-
J)

\lL
II
10
9

8
7


6

5


4

3
2
1
A
•. 	 'v '
pj\
~ VTCP
\
\
\
\


	 t
_ /" ^ \ i N^"DCP

a \ ° N
\ V
— _ A
T =28 °C \
— 0.01 M P04 \
F-400 8 ^
Ceq,DCP = 2 xlO' M \
— Ceq , TCP = 2 x 10 M N
III 1
45678 9 1C
                                 PH
Figure 35.   Influence of pH on chlorophenol adsorption capacity.
                                 79

-------
is the adsorption of the undissociated molecule.   At pH  5.2,  2,4,6-TCP
adsorbs three times as strongly as  2,4-DCP at an  equilibrium  concentration
of 2 x 10"8 M.   It should be noted  that at each pH value studied,  10'2 M
phosphate buffer was employed to ensure that the  solution pH  would remain
constant throughout the time required to reach equilibrium.   Zogorski  and
Faust (1976) reported that the adsorption of undissociated 2,4-DCP is
unaffected by the presence of 5 x 10~2 M phosphates while the equilibrium
capacities of dissociated 2,4-DCP were enhanced by about 10 to 20  percent.
Due to the sensitivity of adsorption to pH fluctuations  in the region  near
the pKa of the chlorophenols, it was felt that use of the buffer was more
important than its potential for causing a shifting adsorption capacity.

     An attempt was made to fit the experimental  single  solute data with  the
Langmuir adsorption equation shown  below.  The model  assumes  that  a mono-
layer of solute molecules is adsorbed on the surface, the energy of adsorp-
tion is constant, and that no interaction occurs  between adsorbed  molecules
(Langmuir, 1918).
                      X =
                          XmbC
eg.
                          1  + b C
                                                                        (1)
                                 eq
where:
          X   = amount of solute adsorbed per  unit weight  of adsorbent

          X   = surface coverage corresponding to  a monolayer of  adsorbate
                molecules on the adsorbent surface

          b   = constant related to energy of  adsorption,  where  1/b  is  the
                adsorbate concentration  at which adsorption  attains  one-
                half of the monolayer coverage

          C   = equilibrium solute concentration
Linearization of Equation  1  results  in  the  following  form:

                                       1

                               "m    b Xm  Ceq
                           1
                           X
1
X.
                                        (2)
Fit of the data to the Langmuir model  is  indicated  if the  plot of 1/X  versus
l/Ceq is linear.   It was found that the  data  presented in  Figures 33 and  34
did not adhere to the model  over the concentration  range studied, as was
observed in several  other studies (Zogorski  and Faust, 1974;  Snoeyink  et  al.,
1969).  Instead,  a computer  program was  used  to fit the data  to a polynomial
by Gaussian elimination and  least-squares analysis.  The data fit a  second
order polynomial  of the form:
                     In (X)  = J  + K In (Ceq)  + L (In  Ceq)'
                                        (3)
                                    80

-------
The constants obtained for each isotherm are listed in Table 11
             TABLE 11.  VALUES OF CONSTANTS FOR EQUATION 3 FOR
                      CHLOROPHENOL SINGLE SOLUTE DATA
Compound    Solution pH

2,4-DCP


2,4,6-TCP


5.2
7.0
9.1
5.2
7.0
9.1

-8.4904
-6.9126
-14.056
-11.646
-13.201
-0.19919

-0.48105
-0.22529
-1.2068
-0.96124
-1.1995
+0.66134

-0.02470
-0.01375
-0.048154
-0.038926
-0.049262
+0.0086726

Bisolute Systems

     Many studies of carbon adsorption from single solute systems  have been
conducted but natural waters contain a mixture of organic substances.   As
the study of the kinetics of phenol chlorination by Lee (1967)  showed, there
is commonly a number of chlorophenols present at a given time,  for example.
The presence of other compounds results in the occurrence of competition  for
adsorption sites which can markedly affect the adsorption of a  particular
substance.  Mutual inhibition of adsorptive capacity occurs if  the adsorp-
tion affinities of the solutes do not differ by more than a few orders of
magnitude, and if there is no specific interaction between solutes which
enhances adsorption (Weber and Morris, 1964).  Some information can be
obtained from simplified systems, such as bisolute systems, and other  infor-
mation can be obtained from studies using natural waters and simulated
natural waters.

     The most common model used to predict equilibrium concentrations  in  a
multi-solute system is the Langmuir model for competitive adsorption,  first
developed by Butler and Ockrent (1930).  The model permits calculation of
the amount of species "1" adsorbed per unit weight of adsorbent at equi-
librium concentration Ceqsi in the presence of other species.
                     Xl =
Xm.l b!  Ceq,1


  l  
(4)
                                     81

-------
The constants Xm and b in Equation 4 are those obtained from single solute
systems.  The Langmuir competitive model is not expected to apply to systems
where adsorption of either component of a bisolute system occurs on sites
that are either inaccessible or unavailable to one of the species, e.g.,
where a portion of the adsorption occurs without competition.   If the
Langmuir model for competitive adsorption does predict adsorption capacities
from a bisolute system when Xmj  f %m,2> i* 1S probably because the species
compete for all available sites.   The'difference in X^ values  in this
instance would most likely be caused by a difference in surface area covered
by one adsorbate as compared with the competing adsorbate (Jain and Snoeyink,
1973)

     Jain and Snoeyink (1973) studied competitive adsorption of bisolute
mixtures of organic anions arid neutral  species.   Typical results showed that
the presence of 5 x 10"3 M p-bromophenol reduced the capacity  for p-nitro-
phenol by nearly 2 orders of magnitude  at p-nitrophenol equilibrium concen-
tration of 5 x 10~5 M.  It was found that the Langmuir competitive model,
which assumes competition among species for various sites could accurately
describe the adsorption from a solution of these compounds in  the neutral
form.  The data for other mixtures were varied,  however.  Some adsorbed
with very little competition and in other cases, electrostatic repulsive
forces were of importance.  These latter cases are more difficult to des-
cribe with a model, although a modification of the Langmuir equation to
account for some adsorption without competition was moderately successful.
Their model is based on the assumption  that adsorption occurs  without com-
petition when Xm,i > Xmj2 and that the  number of sites for which there is
no competition is equal to the quantity (Xmj  - Xp^)-  The equation for
the amount of species "1" adsorbed takes the form b'elow while  the equation
for the amount of species "2" adsorbed  is the same as that in  the Langmuir
model.
     .  - (Xm,1  ' Xm,2)  b1  Ceq,1  .      "m.2 b1  Ceq.l                     ,,>
      1  "      ! * bl  <=e,.l         1  + b,  Ceq>1  + b2  Ceq_2               I5'


     Tables 12,  13 and 14 show equilibrium data  obtained  for competitive
adsorption studies using 2,4-DCP  and  2,4,6-TCP  at pH  5.2,  7.0,  and 9.1.
Three different isotherms were set  up at each pH to attempt to  show the
magnitude of competition between  each species by increasing the equilibrium
concentration of one species relative to the other.  To indicate the degree
of competition,  the amount which  would be adsorbed per gram if  the species
were present in solution alone, Xss>  is tabulated. Also  tabulated is the
extent of adsorption as  predicted by  the Langmuir and the Jain  and Snoeyink
competitive models, Equations 4 and 5, respectively.   A computer program
was developed to evaluate the constants for the models using the polynomials
which were fitted to the single solute data. The program calculates Xm  and
b for each individual  concentration and then uses these values  in Equations
4 and 5  to calculate the amount of  species adsorbed in the presence of the
other chlorophenol.
                                    82

-------
              TABLE 12.   CHLOROPHENOL COMPETITION AT pH 5.2
    Ceq,DCP = °'35 Ceq,TCP



     Co,DCP= 2.70 mg/1




     Co,TCP = 20'80 mg/1
                            2,4-Dichlorophenol

Bottle
1
2
3
4
5
6


Bottle
1
2
3
4
5
6
Ceq,DCP
mole/1
5.52xlO"9
Z.OlxlO"8
3.99xlO"8
l.SlxlO"7
6.44xlO"7
3.37xlO"6

Ceq,TCP
mole/1
2.28xlO"8
7.09xlO"8
1.47xlO"7
3.92xlO"7
1.54xlO"6
7.34xlO"6
Ceq,TCP
mole/1
2.28xlO"8
7.09xlO"8
1.47xlO"7
3.92xlO"7
1.54xlO"6
7.34xlO"6
2,4
Ceq,DCP
mole/1
5.52xlO"9
2.01xlO"8
3.99xlO"8
1.61xlO"7
6.44xlO"7
3.37xlO"6
Xobs
mole/g
1.62xlO"4
2.13xlO"4
2.71xlO"4
3.31xlO"4
3.77xlO"4
3.77xlO"4
Xss
mole/g
2.47xlO~4
4.61xlO"4
6. 09x1 O"4
9.20xlO"4
1.25xlO"3
1.66xlO~3
V
Langmuir
mole/g
l.SlxlO"4
2.38xlO"4
2.82xlO"4
3.94xlO~4
3.60xlO"4
3.99xlO"4
XJain
mole/g
l.SlxlO"4
2.38xlO"4
2.82xlO"4
3.94xlO"4
3.60xlO"4
3.99xlO"4
,6-Trichlorophenol
Xobs
mole/5
l.OSxlO"3
1.36xlO"3
1.72xlO"3
2.12xlO"3
2.56xlO"3
2.80xlO"3
Xss
mole/g
1.22xlO"3
1.87xlO"3
2.20xlO"3
2.50xlO"3
2.91xlO"3
3.42xlO"3
V
Langmuir
mole/g
7.64xlO"4
1.15xlO"3
1.45xlO"3
1.82xlO"3
2.48xlO"3
2.63xlO"3
Jain
mole/g
1.04xlO"3
1.55xlO"3
1.91xlO"3
2.32xlO"3
2.84xlO"3
2.89xlO"3
X .   = observed X
XLangmuir = X as
                                                     from Equation 4



Xcc = single solute X    Xla_-   = X as calculated from Equation 5
 ss                       jain
                                                               (continued)
                                     83

-------
                           TABLE  12  (continued)

co
Co
q,DCP 1-1 Ceq,TCP
,DCP = 3'68 mg/1
jTCp =8.11 mg/1
2,4-Dichlorophenol
Bottle
1
2
3
4

Bottle
1
2
3
4
Ceq,DCP
mole/1
4. 60x1 O"9
1.47xlO"8
3. 80x1 O"8
1.87xlO"7

Ceq,TCP
mole/1
8.78xlO"9
1.57xlO"8
3.14xlO"8
1.16xlO"7
Ceq,TCP
mole/1
8.78xlO"9
1.57xlO"8
3.14xlO"8
1.16xlO"7
2,4
Ceq,DCP
mole/1
4.60xlO"9
1.47xlO"8
3. 80x1 O"8
1.87xlO"7
Xobs
mole/g
2.14xlO"4
3.39xlO"4
4.46xlO"6
6.31xlO"4
Xss
mole/g
2.28xlO"4
3.95xlO"4
e.ooxio"4
9.61xlO"4
V
Langmuir
mole/g
1.54xlO"4
2.52xlO"4
3.51xlO"4
5.41xlO"4
XJain
mole/g
1.54xlO"4
2.52xlO"4
3.51xlO"4
5.41xlO"4
,6-Trichlorophenol
Xobs
mole/g
3.89xlO"4
6.17xlO"4
8.13xlO"4
1.16xlO"3
Xss
mole/g
7.40xlO"4
l.OlxlO"3
1.41xlO"3
2.10xlO"3
V
Langmuir
mole/g
4.77xlO"4
5.94xlO"4
7.69xlO"4
1.1 6x1 0~3
XJain
mole/g
6. 58x1 O"4
8.24xlO"4
1.07xlO"3
1.60xlO"3

Note:   See Table 12A  for  definition of terms.
                                                              (continued)
                                    84

-------
                         TABLE 12 (continued)

c. c
c
c
f\j
eq,DCP ~
o,DCP = 8<
o,TCP = 6<
'9 Ceq,TCP
,47 mg/1
17 mg/1




2,4-Dichlorophenol

Bottle
1
2
3
4
5
6



Bottle
1
2
3
4
5
6
Ceq,DCP
mole/1
2.61xlO"8
9. 20x1 O"8
1.75xlO"7
4.75xlO"7
1.47xlO"6
5.24xlO"6

p
ueq,TCP
mole/1
3.80xlO"9
1.24xlO"8
2.03xlO"8
5.06xlO"8
1.27xlO"7
5.76xlO"7
Ceq,TCP
mole/1
3.80xlO"9
1.24xlO"8
2.03xlO"8
5.06xlO"8
1.27xlO"7
5.76xlO"7
2,4
p
Ueq,TCP
mole/1
2.61xlO"8
9.20xlO"8
1.75xlO"7
4.75xlO"7
1.47xlO"6
5.34xlO"6
Xobs
mole/g
4.98xlO~4
6.82xlO"4
8.55xlO"4
l.OOxlO"3
1.24xlO"3
1.50xlO"3
Xss
mole/g
5.21xlO"4
8.12xlO"4
9.49xlO"4
1.16xlO"3
1.49xlO"3
1.96xlO"3
Langmuir
mole/g
3. 68x1 O"4
5.45xlO"4
6.54xlO"4
8.23xlO"4
1.02xlO"3
1.12xlO"3
XJain
mole/g
3.68xlO"4
5.45xlO~4
6.54xlO"4
8.23xlO"4
1.02xlO"3
1.12xlO"3
,6-Trichlorophenol
V
Xobs
mole/q
2.99xlO"4
4.11xlO"4
5.51xlO"4
6.07xlO~4
7.65xlO"4
9.89xlO"4
y
Ass
mole/g
4.65xlO"4
8.90xlO"4
1.14xlO"3
1.66xlO"3
2.15xlO"3
2.61xlO"3
Y
Langmuir
mole/g
2.55xlO"4
4.42xlO"4
5.31xlO"4
7.27xlO"4
9.15xlO"4
l.SlxlO"3
Y
Jain
mole/g
3.14xlO"4
5.73xlO"4
6. 98x1 O"4
9.91xlO"4
1.29xlO"3
1.87xlO"3

Note:   See Table 12A for definition of terms.
                                                                (continued)
                                     85

-------
                            TABLE 12 (continued)

D. Data Summary
Dichlorophenol
y y Y
C_ obs obs Aobs
Bottle
<=H
mole/1
Isotherm A:
1 5
2 2
3 3

4 1

5 6

6 3
.52x10
.01x10
.99x10

.61x10

.44x10

.37x10
Isotherm B:

1 4

2 1
3 3
4 1

.60x10

.47x10
.80x10
.87x10

Ceq
-9
-8
-8
_7
/
_7
/
-6
\J
Ceq
-9
-/
-8
u
-8
-7
Isotherm C: C

1 2
2 9

3 1
4 4
5 1
6 5

.61x10
.20x10

.75x10
.75x10
.47x10
.34x10
-8
(J
-8
_7
/
-7
-6
-6
Xss
,DCP =
66%
46
44

36

30

23
,DCP =

94

86
74
66
a,
,DCP =

96
83

90
86
83
77
V
Langmui
r XJain

Ceq
mole/1
Trichlorophenol
Xobs Xobs Xobs
Xss
V
Langmuir
XJain
°'35 Ceq,TCP
107%
89
96

84

105

94
1J Ceq,

139

135
127
117
8.9 C
eq,

135
125

131
122
122
134
107%
89
96

84

105

94
TCP

139

135
127
117
TCP

135
125

131
122
122
134
2
7
1

3

1

7


8

1
3
1


3
1

2
5
1
5
.28xlO"8
.09x10
-8
.47xlO"7

.92x10

.54x10

.34x10


.78x10

.57x10
.14x10
.16x10


.80x10
.24x10

.03x10
.06x10
.27x10
.76x10
_7
/
-6
\J
-fi
o

_Q
J
-fi
O
-8
-7

-Q
3
-8
0
o
-8
-7
-7
84%
73
78

85

88

82


53

61
58
55


64
46

45
37
36
38
135%
118
119

116

103

106


82

104
106
100


117
93

97
83
84
75
99%
88
90

91

90

97


59

75
79
73


95
72

74
61
59
53

Note:   See Table 12A for definition of terms.
                                     86

-------
              TABLE  13.   CHLOROPHENOL  COMPETITION AT pH  7.0

A" Ceq,DCP = °'14 Ceq,TCP
Co,DCP = 6'58 m9/]
Co,TCP = 10'76 m9/]


Bottle
1
2
3
4
5
6



Bottle
1
2
3
4
5
6
r
ueq,DCP
mole/1
5.37xlO"9
1.17xlO~8
2. 58x1 O"8
6. 87x1 O"8
2. 58x1 O"7
1.27xlO"6

r
Leq,TCP
mole/1
4.25xlO"8
9.62xlO"8
2.15xlO"7
5.19xlO"7
1.65xlO"6
6.58xlO"6
r
eq,TCP
mole/1
4.25xlO"8
9.62xlO"8
2.15xlO"7
5.19xlO"7
1.65xlO"6
6.58xlO"6
2
r
Leq,DCP
mole/1
5.37xlO"9
1.1 7x1 0~8
2. 58x1 O"8
6.87xlO"8
2.58xlO"7
1.27xlO"6
2 ,4-Di chl orophenol
y
xobs
mole/g
4.61xlO"4
5.81xlO"4
7.19xlO"4
8. 50x1 O"4
1.02xlO"3
1.18xlO"3
y
SS
mole/g
5.40xlO"4
6.90x.O"4
8.80xlO"4
l.llxlO"3
1.37xlO"3
1.71xlO"3
Y
Langmuir
mole/g
3.99xlO"4
4.64xlO"4
5.22xlO"4
5.77xlO"4
.5.39xlO"4
2.54xlO"4
y
AJain
mole/g
3.99xlO"4
4.64xlO"4
5.23xlO"4
5.77xlO"4
5.39xlO"4
2.54xlO"4
, 4, 6-Tri chl orophenol
V
xobs
mole/g
6.22xlO"4
7.83xlO"4
9.67xlO"4
1.14xlO"3
1.34xlO"3
1.45xlO"3
y
xss
mole/g
9.60xlO"4
1.25xlO"3
1.37xlO"3
2.15xlO"3
1.39xlO"3
1.70xlO"3
V
Langmuir
mole/q
4.10xlO"4
G.lOxlO"4
8.57xlO"4
l.lSxlO"3
1.73xlO"3
2.39xlO"3
x
AJain
mole/g
6.71xlO"4
9. 74x1 0"4
1.31xlO"3
1.67xlO"3
2.09xlO"3
2.49xlO"3
Note:   See Table 12A for definition of terms.
                                                              (continued)
                                    87

-------
                            TABLE  13   (continued)

B" Ceq,DCP = °'45 Ceq,TCP
Co,DCP = 8'16 mg/1
Co,TCP = 6-^mg/l


p
Leq,DCP
Bottle mole/1
1
2
3
4
5



Bottle
1
2
3
4
5
7.05xlO-9
1.23xlO~8
7.36xlO"8
3.25xlO"7
1.44xlO"6

r
ueq,TCP
mole/1
1.06xlO"8
4.35xlO"8
1.49xlO"7
5.98xl0^7
2.53xlO"6
r
Leq,TCP
mole/1
1.06xlO"8
4.35xlO"8
1.49xlO"7
5.98xlO"7
2.53xlO"6
2
r
Leq,DCP
mole/1
7.05xlO"9
1.23xlO"8
7.36xlO"8
3.25xlO"7
1.44xlO"6
2,4-Dichlorophenol
V
Xobs
mole/g
4.75xlO"4
7.15xlO"4
9. 94x1 O"4
1.21xlO"3
1.46xlO"3
,4,6-Trichl
Y
obs
mole/g
2.95xlO"4
4. 44x1 O"4
6.15xlO"4
7.40xlO"4
8.55xlO"4
Y
xss
mole/g
5.84xlO"4
7.00xlO"4
1.12xlO"3
1.41xlO"3
1.75xlO"3
orophenol
X
Ass
mole/g
3.65xlO~4
8.70xlO"4
1.50xlO"3
2.20xlO"3
3.09xlO"3
x
Langmuir
mole/g
4.84xlO~4
5.12xlO"4
7.33xlO"4
8.33xlO"4
6.19xlO"4

x
Langmuir
mole/q
1.64xlO"4
3.91xlO"4
6.61xlO"4
l.llxlO"3
1.87xlO"3
x
AJain
mole/g
4.84xlO"4
5.12xlO"4
7. 33x1 O"4
8.33xlO"4
6.19xlO~4

Y
Jain
mole/g
2.27xlO"4
6.20xlO"4
l.OOxlO"3
1.52xlO"3
2.10xlO"3

Note:   See Table  12A  for  definition of terms.
                                                              (continued)
                                    88

-------
                            Table  13  (continued)

C> Ceq,DCP 1>4 Ceq,TCP
Co,DCP= 12.58mg/l
Co,TCP = 4'98 mg/1
2,4-Dichlorophenol

Bottle
1
2
3
4
5
6



Bottle
1
2
3
4
5
6
Ceq,DCP
mole/1
2.70xlO~8
7.67xlO"8
1.53xlO"7
6.29xlO"7
2.21xlO"6
5.40xlO"6

p
Leq,TCP
mole/1
2.03xlO"8
6.08xlO"8
1.42xlO"7
3.92xlO"7
1.37xlO"6
3.14xlO~6
Ceq,TCP
mole/1
2.03xlO"8
6.08xlO"8
1.42xlO"7
3.92xlO"7
1.37xlO"6
3.14xlO"6
2
r
Sjq.DCP
mole/1
2.70xlO"8
7.67xlO"8
1.53xlO"7
6.29xlO"7
2.21xlO"6
5.40xlO~6
Xobs
mole/g
8.45xlO"4
l.OBxlO"3
1.28xlO~3
1.49xlO"3
1.74xlO"3
1.92xlO"3
Xss
mole/g
8.90xlO"4
1.15xlO~3
1.28xlO"3
1.54xlO"3
1.74xlO"3
2.20xlO"3
Langmuir
mole/g
6.78xlO"4
8.14xlO"4
8.83xlO"4
1.06xlO"3
l.OlxlO"3
6.87xlO"4
XJain
mole/g
6.78xlO"4
8.14xlO"4
8.83xlO"4
1.06xlO"3
l.OlxlO"3
6.87xlO"4
,4,6-Trichlorophenol
X
obs
mole/g
2.76xlO"4
3.43xlO"4
4.16xlO"4
4.82xlO"4
5.55xlO"4
5.90xlO"4
x
ss
mole/g
5.70xlO"4
l.OSxlO"3
1.50xlO"3
2.02xlO"3
2.39xlO"3
2.52xlO"3
x
Langmuir
mole/g
2.28xlO"4
4.05xlO"4
6.02xlO"4
8.55xlO"4
1.33xlO"3
1.91xlO"3
x
Jain
mole/g
2.97xlO"4
5.80xlO"4
8.81xlO"4
1.19xlO"3
1.64xlO"3
2.06xlO"3

Note:   See Table 12A for definition  of  terms.
                                                               (continued)
                                    89

-------
                            TABLE 13 (continued)

D. Data Summary
Dichlorophenol
C. Xobs Xobs Xobs
Bottle mole/1
Isotherm A: C
eq
1 5.37xlO"9
2 1.17xlO"8
-8
3 2.58x10 °
-8
4 6.87x10 °
5 2.58xlO"7
6 1.27xlO~6
Isotherm B: C
eq
-9
1 7.05x10
2 1.23xlO"8
3 7.36xlO"8
4 3.25xlO"7
5 1.44xlO"6
Isotherm C: C
eq
1 2.70xlO"8
2 7.67xlO"8
3 1.53xlO"7
4 6.29xlO"7
5 2.21xlO"6
6 5.40xlO"6
Xss
a/
,DCP ~
85%
84

82

77
74
69
,DCP =

81
102
89
86
83
,DCP =
95
91
100
97
100
87
V
Langmui
0-14 Ceq
116%
125

137

147
189
465
0.45 Ce

98
140
136
145
236
1-4 Ceqj
125
129
145
141
172
279
r XJain
,TCP
116%
125

137

147
189
465
,TCP

98
140
136
145
236
TCP
125
129
145
141
172
279
Ceq
mole/1

4.25xlO"8
9.62xlO"8
_7
2.15x10 '
_7
5.19x10 '
1.65xlO"6
6.58xlO"6

-8
1.06x10 °
4.35xlO"8
1.49xlO"7
5.98xlO"7
2.53xlO"6

2.03xlO"8
6.08xlO"8
1.42xlO"7
3.92xlO"7
1.37xlO"6
3.14xlO"6
Trichlorophenol
Xobs Xobs Xobs
Xss

65%
63

71

53
96
85


81
51
41
34
28

48
33
28
24
23
23
V
Langmuir

152%
128

113

97
78
61


180
114
93
67
46

121
85
69
56
42
31
XJain

93%
80

74

68
64
58


130
72
62
49
41

93
59
47
41
34
29
Note:   See Table 12A for definition  of terms.
                                    90

-------
               TABLE 14.  CHLOROPHENOL COMPETITION AT pH 9.

A. C
eq,DCP = °'°44 Ceq,TCP
Co,DCP = 3'23 mg/1
Co,TCP = 4'64 m9/]
2,4-Dichlorophenol

Bottl
1
2
3
4
5



Bottl
1
2
3
4
5
Ceq,DCP
e mole/1
1.38xlO"8
1.84xlO"8
2.94xlO"8
3.68xlO"8
7.51xlO"8

r
Leq,TCP
e mole/1
1.62xlO"7
3.73xlO"7
6.84xlO"7
1.52xlO"6
3.85xlO"6
Ceq,TCP
mole/1
1.62xlO~7
3.73xlO"7
6.84xlO"7
1.52X10"6
3.85xlO"6
2
r
eq,DCP
mole/1
1.38xlO"8
1.84xlO"8
2.94xlO"8
3.68xlO~8
7.51xlO~8
Xobs
mole/g
1.90xlO"4
2.57xlO"4
3.30xlO"4
4.12xlO"4
5.45xlO"4
,4,6-Trichl
y
xobs
mole/g
2.24xlO"4
3-OOxlO"4
3.80xlO"4
4.58xlO~4
5.43xlO"4
Xss
mole/g
3.35xlO"4
3.95xlO"4
S.lOxlO"4
5.70xlO"4
7.53xlO"4
orophenol
y
Ass
mole/g
2.37xlO"4
3.30xlO"4
4.25xlO"4
5.80xlO"4
8.70xlO"4
V
Langmuir
mole/g
1.85xlO"4
2.26xlO"4
3.00xlO"4
3.46xlO"4
S.OlxlO"4

Y
Langmuir
mole/g
1.65xlO"4
2.21xlO"4
2.66xlO"4
3.57xlO"4
4.74xlO"4
XJain
mole/g
2.63xlO"4
2.86xlO"4
3.53xlO"4
3.49xlO"4
S.OlxlO"4

x
Jain
mole/g
1.65xlO"4
2.21xlO"4
2.66xlO"4
3.57xlO"4
5.50xlO"4

Note:   See Table 12A for definition of terms.
                                                               (continued)
                                     91

-------
                            TABLE  14   (continued)

R r - n ?i r
dl Leq,DCP u-lil ueq,TCP
Co,DCP = 7'67 mg/1
Co,TCP = 2'89 mg/1
2,4-Dichlorophenol

Bottle
1
2
3
4
5



Bottle
1
2
3
4
5
Ceq,DCP
mole/1
2.61xlO"8
5.67xlO"8
1.29xlO"7
3.60xlO"7
1.23xlO"6

r
eq,TCP
mole/1
7.50xlO"8
2.58xlO"7
l.OOxlO"6
2.63xlO"6
5.19xlO"6
Ceq,TCP
mole/1
7. 50x1 O"8
2.58xlO"7
l.OOxlO"6
2.63xlO"6
5.19xlO"6
2
r
beq,DCP
mole/1
2.61xlO"8
5.67xlO"8
1.29xlO"7
3.60xlO"7
J.23xlO"6
Xobs
mole/g
3.60xlO"4
5.35xlO"4
7.81xlO"4
9.79xlO"4
1.24xlO"3
,4,6-Trichl
Y
obs
mole/g
l.llxlO"4
1.64xlO"4
2.27xlO"4
2.52xlO"4
2.56xlO"4
Xss
mole/g
4.75xlO"4
6.90xlO"4
8.80xlO"4
l.lOxlO"3
1.32xlO"3
orophenol
Y
Ass
mole/g
1.75xlO~4
2.87xlO"4
4.90xlO"4
7.30xlO"4
9.60xlO"4
V
Langmuir
mole/g
2.62xlO"4
4.10xlO"4
6.15xlO"4
9.11xlO"4
1.27xlO"3

V
Langmuir
mole/g
1.16xlO"4
1.66xlO"4
2. 44x1 O"4
2.76xlO"4
1.93xlO"4
XJain
mole/g
4.08xlO"4
5.58xlO"4
7.19xlO"4
9.55xlO"4
1.27xlO"3

V
Jain
mole/g
1.16xlO"4
1.66xlO"4
2.44xlO"4
2.76xlO"4
2.49xlO"4

Note:   See Table  12A  for  definition of terms.
                                                               (continued)
                                    92

-------
                           TABLE  14  (continued)

C' Ceq,DCP=°-45Ceq,TCP
Co,DCP = 10'68 ^
Co,TCP= K 56m9/1


Bottle
1
2
3
4
5
6


r
Leq,DCP
mole/1
3.76xlO"8
7.21xlO"8
1.44xlO"7
3.22xlO"7
6.60xlO"7
1.35xlO"6

Ceq,TCP
Bottle mole/1
1
2
3
4
5
6
8.36xlO"8
1.61xlO"7
3.70xlO"7
8.55xlO"7
1.60xlO"6
2.23xlO"6
r
beq,TCP
mole/1
8.36xlO"8
1.61xlO"7
3.70xlO"7
8.55xlO"7
1.60xlO"6
2.23xlO"6
2
Ceq,DCP
mole/1
3. 76x1 O"8
7.21xlO"8
1.44xlO"7
3.22xlO~7
6.60xlO"7
1.35xlO"6
2,4 Dichlorophenol
y
xobs
mole/g
5.89xlO"4
7.18xlO"4
8.85xlO"4
l.OSxlO"3
1.19xlO"3
1.33xlO"3
,4,6-Trichl
Xobs
mole/g
7.04xlO"5
8.49xlO"5
1.02xlO"4
1.17xlO"4
1.16xlO"4
l.lSxlO"4
x
xss
mole/g
5.90xlO"4
7. 50x1 O"4
9.00xlO"4
l.OSxlO"3
1.22xlO"3
1.33xlO"3
orophenol
Xss
mole/g
1.82xlO"4
2.37xlO~4
3.31xlO"4
4.65xlO"4
G.OOxlO"4
6.87xlO"4
V
Langmuir
mole/g
3.22xlO"4
4.55xlO"4
6.29xlO"4
8.61xlO"4
l.OSxlO"3
1.28xlO"3

V
Langmuir
mole/g
1.15xlO"4
1.34xlO"4
1.61xlO"4
1.82xlO"4
1.76xlO"4
1.26xlO"4
y
Jain
mole/g
4.89xlO"4
6.39xlO"4
8.05xlO"4
l.OOxlO"3
1.17xlO"3
1.33xlO"3

XJain
mole/g
l.lSxlO"4
1.34xlO"4
1.61xlO"4
1.82xlO"4
1.76xlO"4
1.26xlO"4
Note:  See Table 12A for definition of terms.
                                                              (continued)
                                    93

-------
                            TABLE 14  (continued)

D.
Data Summary
Ceq
Bottle mole/1
Isotherm A: C
eq
1

2

3
4
5
1.38xlO"8
-8
1.84x10 °
-8
2.94x10 °
3.68xlO"8
7.51xlO"8
Isotherm B: C
eq
1

2

3

4

5
2.61xlO"8
-8
5.67x10 °
_7
1.29x10 '
_7
3.60x10 '
-fi
1.23x10 D
Isotherm C: C
1

2
3
4
5
6
3.76xlO"8
-8
7.21x10 S
1.44xlO"7
3.22xlO"7
6.60xlO"7
1.35xlO"6



Dichlorophenol
Xobs Xobs Xobs
Xss
,DCP =
57%

65

65
72
72
,DCP =
76

78

89

89

94
,DCP =
100

96
98
100
98
100
X X
Langmuir Jain
0.044 Ceq
103%

114

no
119
109
,0.21 C „
eq,
137

131

127

108

98
°'45 Ceq
183

158
141
125
110
104
,TCP
72%

90

93
118
109
TCP
88

96

109

103

98
TCP
120

112
110
108
102
100

Ceq
mole/1

1.62xlO"7
_7
3.73x10 '
_7
6.84x10 '
1.52xlO"6
3.85xlO"6

7.50xlO"8
_7
2.58x10 '
_c
1.00x10 D
C
2.63x10"°
-fi
5.19x10 D

8.36xlO"8
_7
1.61x10 '
3.70xlO"7
8.55xlO"7
1.60xlO"6
2. 23x1 O"6

Tri
Xobs
Xss

95%

91

89
79
62

63

57

46

35

27

39

36
31
25
19
17


chlorophenol
Xobs Xobs
V
Langmuir

136%

136

143
128
115

96

99

93

91

132

61

63
63
64
66
94
XJain

136%

136

143
128
99

96

99

93

91

103

61

63
63
64
66
94
Note:   See Table 12A for  definition  of  terms.
                                   94

-------
Competition at pH 5.2 --

     At pH 5.2 both 2,4-DCP and 2,4,6-TCP are primarily undissociated.
Table 12 shows data obtained in competitive studies.  As the equilibrium
concentration of 2,4,6-TCP rises relative to the equilibrium concentration
of 2,4-DCP the extent of adsorption of 2,4-DCP decreases relative to the
single solute value, as summarized in Table 12D.  The same phenomenon was
observed for 2,4,6-TCP in relation to its single solute surface concentra-
tion values which indicates that competition for the same sites on the
carbon is occurring.  For both species, as the equilibrium concentrations
increase with the ratio of their concentrations remaining constant, the
extent of adsorption becomes increasingly lower than the single solute value
suggesting that at the higher surface coverage the competition is more
intense.

     Table 12D shows the fit of the Langmuir and Jain and Snoeyink competi-
tive models.  The Langmuir model does a fairly good job of fitting both data
sets.  Since Xmjycp is greater than XmjDCP over the entire concentration
range, the Jain model was applicable to the 2,4,6-TCP data.   The Jain and
Snoeyink model provided a better fit of the data where Ceq,TCP > ceq,DCP
but did not improve the fit where Ceg}TCP — ^eq,DCP-  As the amount of
2,4,6-TCP increased the amount of anionic species increases  since the pH is
fairly close to the pKa.   The better fit of the Jain model where there is a
significant portion of anionic TCP may indicate that there is some adsorption
occurring without competition because of differences in adsorption sites for
anionic versus neutral 2,4,6-TCP.  Figures 36 and 37 show Langmuir model
predictions for 2,4-DCP and 2,4,6-TCP as simulated using the computer and
the single solute data previously presented in Figures 33 and 34.

Competition at pH 7.0 --

     At pH 7.0, 2,4-DCP is primarily neutral  while 2,4,6-TCP is primarily
anionic.  Table 13 shows  data obtained in three studies at this pH value.
As at pH 5.2, competition occurs between the species as evidenced by the
increasing reduction of X0bs relative to XSs as the equilibrium concentra-
tion of the competing species increases.  As at pH 5.2, as the individual
equilibrium concentrations increase with their ratio remaining approximately
equal, the extent of adsorption is lowered significantly. The Langmuir model
fits the 2,4-DCP data fairly well at low Ceq but at high Ceq values the fit
is poor.  The same is true for 2,4,6-TCP.  At all of the concentrations
studied, Xm,TCP is greater than Xm,DCP so the Jain and Snoeyink model
applied.  The Jain and Snoeyink model fit the data well for  low 2,4,6-TCP
concentrations but did not provide an improved fit at high concentrations.
The Langmuir model predicted much less competition than was  observed for
2,4-DCP; in some cases where Ceq,DCP > 10"6 M the observed extent of adsorp-
tion was as much as 200 percent of that predicted by the model as shown
in Table 13D.  At high equilibrium concentrations of 2,4,6-TCP the observed
extent of adsorption was  less than predicted while at low equilibrium con-
centrations the opposite  was true.  For isotherm A where Cen}DCP = °-^4
ceq,TCP> at low equilibrium concentrations, the Jain model aid a good job
of predicting extent of adsorption.  Apparently the anionic  2,4,6-TCP and
neutral 2,4-DCP which predominated here did not compete entirely for the


                                     95

-------
cn
                   ,62
                0>

                en
                «>  10-
                o
                E
                c
                o
                O)
                O   -4
                §  I0
               o
                ft)
                o
                o
,69
                                  Single Solute
                                         ,68
                                                            2,4,6-TCP = 10°  M-
                                                                                          -7
                                                                            2,4,6-TCP=IO  M
                                                                        -6
2,4,6- TCP = 10"  M
              pH 5.2
              F-400
              0.01 M PO,
   ,a7
,66
                                    Equilibrium   Concentration, Ce  (moles
,d5
                        Figure  36.   Competitive adsorption capacities  predicted by the Langmuir
                                     model  for dichlorophenol at pH  5.2.

-------
     ,62
0>
O
E
c
o
c
Q>
O
C
o
O
.o4
                Single  Solute
                                            r2,4- DCP = I0"8 M
                                                                          -7
                                                         2,4- DCP = 10  M
                                        2,4- DCP = I06 M
u
o
                                              pH 5.2

                                              F —400

                                              0.01 M  PO,,
     ,b5
,o9
                           ,68
                                               -7
                                         10
 -6
10
                   Equilibrium    Concentration ,  Ceq  (moles/J[)
,o5
          Figure 37.  Competitive adsorption  capacities predicted by the Langmuir
                      model for trichlorophenol  at pH 5.2.

-------
same sites.  However at higher equilibrium concentrations the model  failed
since 2,4-DCP extent of adsorption was much greater than predicted while
2,4,6-TCP adsorbed lower than predicted.  These same trends are evident in
isotherms B and C which represent increasing concentrations of 2,4,6-TCP
relative to 2,4-DCP.  Since the pH is near the pKa of both compounds the
system contains four types of species, anionic 2,4-DCP and 2,4,6-TCP and
neutral 2,4-DCP and 2,4,6-TCP which complicates the search for a reason for
this behavior.  In studies performed by Jain and Snoeyink (1973) on
neutral p-nitrophenol and anionic benzenesulfonate, at pH 3.8, their model fit
well for data collected in the concentration range 10~5 to 10~2 M which
suggested that some adsorption occurred without competition.  Neither species
competed with the other to a great extent.  However with neutral 2,4-DCP and
anionic 2,4,6-TCP, 2,4-DCP competes with 2,4,6-TCP while 2,4,6-TCP affects
2,4-DCP only slightly.  A possible explanation is that the anionic 2,4-DCP
which is present at pH 7.0 effectively competes with the anionic 2,4,6-TCP
while the anionic 2,4,6-TCP present does not compete with the neutral  2,4-DCP
for the same sites.  The small degree of competition of 2,4,6-TCP with
2,4-DCP occurs between the neutral species of which 2,4-DCP is present to a
greater extent.  The effective competition of anionic 2,4-DCP with 2,4,6-TCP
occurs, as explained by Ward and Getzen (1969), since the anionic 2,4-DCP
adsorption is favored at pH values just below a compound's pKa-  An  alterna-
tive explanation may be that the chlorophenol species are changing on  the
carbon surface.  Computer predictions for pH 7.0 are shown in Figures  38 and
39.

Competition at pH 9.1 --

     At pH 9.1 both species are primarily anionic.  Table 14 indicates that
as the concentration of 2,4-DCP approaches that of 2,4,6-TCP the extent of
adsorption of 2,4,6-TCP decreases.  When Ceq,DCP - 0-45 Ceq5jCP tne  observed
value of the extent of adsorption of 2,4-DCP is equal to the values  which
would be observed if 2,4,6-TCP were not present while those of 2,4,6-TCP are
approximately 30 percent of those if no 2,4-DCP were present.  Since the
adsorption capacities for 2,4-DCP at this pH are nearly four times greater
than those for 2,4,6-TCP, an explanation for which is not apparent,  the Xm
value also was greater indicating that some adsorption of 2,4-DCP could occur
without competition.  This was indicated by the fit of the Jain and  Snoeyink
model to the 2,4-DCP data for all three isotherms as illustrated in  Table 14D.
The model fit the 2,4,6-TCP fairly well except when Ceq,DCP became close to
that of Ceq,TCP-  ^n this instance the adsorption of Ceq,TCP was somewhat
less than predicted indicating that the stronger adsorbing species predomi-
nated, and that electrostatic repulsion on the carbon surface may have come
into play causing less 2,4,6-TCP to be adsorbed.  When Ceq,TCP - 23  Ceq,DCP
the predictions for 2,4,6-TCP were slightly lower than observed which  may
also indicate that less repulsion on the surface was occurring since less of
the stronger adsorbing species was present.  Figures 40 and 41 show  adsorp-
tion capacities for pH 9.1 as predicted by the Jain and Snoeyink model which
was the model which provided the best fit for the system.  Apparently  for
this system the species adsorbed on the same sites but due to the widely
differing capacities of the carbon for the two species, some adsorption
occurred without competition.
                                      98

-------
0>
^
v>
_0>


1
c
o
    .0'
o
o
0)
o
o
    io4
    10"
	   Single Solute
                          .o8
                                              2,4,6- TCP = IO8 M
                                                                          -7
                                                           2,4,6- TCP = 10  M
                                     2,4,6- TCP = 10" M
                                            pH  7.0

                                            F-400

                                            0.01 M  PO,
                                             10
                                                         ,66
,65
                       Equilibrium
                                     Concentration ,   Ce   (moles/X)
           Figure  38.   Competitive adsorption capacities predicted by the Langmuir
                       model  for dichlorophenol  at pH 7.0.

-------
                   .62
o
o
               o>
              x.
               in
               9>


               I
c
0)
o
c
o
o
              0>
              o
     ,63
                              Single Solute
                                                                                    -7
                                                          2,4- DCP =10  M
                                                                2,4-DCP = 10  M
                                                       2,4- DCP =I06 M
                                                               pH 7.0

                                                               F-400

                                                               0.01  M PO.
                   .65
                     -9
                    10
                           -8
                           10
                                     Equilibrium
 -7
10
 -6
IO
                                     Concentration,  C     (moles/5()
 -5
10
                        Figure  39.
                      Competitive adsorption capacities predicted by the Langmuir

                      model  for trichlorophenol  at pH 7.0.

-------
o
                   io2
0>

M



1  ,6=
o

5
               0)
               u
               o
 -9
10
                                                                              Single Solute
                              2,4,6- TCP = 10
                                                                        -7
                                                          2,4,6-TCP = IO  M
                                            -2,4,6-TCP = 10  M
                                                                          pH 9.1

                                                                          F-400

                                                                          0.01 M PO,
                                          -8
                                        10
                                              -7
                                             10
 -6
10
                                                                                                  10
                                    Equilibrium    Concentration,   Ceq  (moles/50
                         Figure 40.  Competitive adsorption capacities predicted  by  the  Jain
                                     model for dichlorophenol at pH 9.1.

-------
    .62
                pH 9. I

                F-400

                0.01 M PO,
 in


~o
J


X


 c
 o

"o

 c
 0)
 u
 c
 o
u
u
o
    io3
    io4
                              2,4-DCP = IO8 M
                                                         2,4-DCP = IO'6 M
                                   -2,4- DCP = 10  M
    l65
      10"
.68
                                             .67
                    Equilibrium    Concentration,  Ceq  (moles/JO
,65
           Figure 41.  Competitive adsorption capacities predicted by the Jain

                       model for trichlorophenol at pH 9.1.

-------
     According to Weber and Morris (1964), when the concentration of solute
is small the term (E  b^ Cn-) in the denominator of the Langmuir equation,

Equation 4, is much less than unity so direct proportionality between solute
equilibrium concentration and the amount adsorbed, X-j, is anticipated.   No
competitive effects would then be observed.  However, this was not the  case
at pH 5.2 or at 7.0 or 9.1.  In all instances for both compounds the term
(bi C-| + b2 C2) was approximately equal to unity; as the value of Ci
increased, the corresponding value of bi decreased thus causing the entire
denominator to remain greater than unity.   In the Langmuir equation, the
parameter b is proportional to exp(-AH/RT) where AH is the adsorption energy.
Adsorption energy varies with surface coverage; high-energy sites are
occupied first with subsequent adsorption occurring at increasingly lower
energy sites as the surface coverage increases.  The net result is signifi-
cant competition between adsorbates even at low concentration.

Chlorophenols in the Presence of Humic Substances

     Data showing the adsorption of 2,4,6-TCP from solutions containing
humic substances at pH 5.2 and 9.1 are shown in Tables 15 and 16, respec-
tively.  Three types of humic substances were used:  commercial humic acid,
leaf fulvic acid, and soil fulvic acid, the characteristics of which are
described in Section 4 and previously in Section 5.  Tests at each pH were
performed with two initial concentrations of humic material, 10 and 50  mg/1
as TOC, and also with two F-400 carbon doses to obtain different 2,4,6-TCP
equilibrium concentrations.  Since initial tests showed that 2,4-DCP behaved
in a manner quite similar to 2,4,6-TCP in the presence of humic material,
data were collected using 2,4,6-TCP as the sole chlorophenol adsorbate.

     At pH 5.2, the presence of humic materials resulted in significant
reductions in the capacity of carbon for 2,4,6-TCP as compared with that
achieved in distilled water systems.   Leaf fulvic acid  proved to be the
most effective competitor of the three humic substances tested while soil
fulvic acid and commercial humic acids were nearly equally effective as
competitors.  As shown in Table 15, 10 mg/1 of leaf fulvic acid reduced
the carbon's capacity at Ceq = 1-99 x 10-6 M 2,4,6-TCP by 52 percent.
Increasing the fulvic acid concentration to 50 mg/1 resulted in a slightly
greater reduction in 2,4,6-TCP capacity.  Data for commercial humic and
soil fulvic acids show the same trends.  At a 2,4,6-TCP equilibrium con-
centration of nearly an order of magnitude higher resulted in capacity
reductions which were not quite as large.

     It was expected that leaf fulvic acid and commercial humic acid would
compete better with the chlorophenols than would soil fulvic acid due to
their greater adsorbability as illustrated in Figure 16.  The data in
Figure 24 were collected at pH 7.0, however, so direct comparisons to the
data collected at pH 5.2 cannot be drawn.   Humic substance removal as
measured by fluorescence showed the greatest removals for commercial humic
acid and the least for soil fulvic acid.  However, leaf fulvic acid was the
best competitor with 2,4,6-TCP for adsorption sites while the other two
materials resulted in roughly equivalent competition.  Hence, percent
                                     103

-------
       TABLE 15.   2,4,6-TCP COMPETITION WITH HUMIC SUBSTANCES  -  pH 5.2
                              CoTCR = 23.80 mg/1
                              TO"2 M phosphate buffer

Humic Substances


Type
Commercial
Humic
A/- T H
MC 1 U

Leaf
Fulvic
Arirl


Soil
Fulvic
AriH


Initial
Cone. % ,
mg/1 TOC Removed
10 83
70
50 66
48
10 56
37
50 28
16
10 48
23
50 27
0

Ceq
mole/1
4.41xlO"7
5.01xlO"6
l.OSxlO"6
9.18xlO"6
1.99xlO"6
1.04xlO"5
3.80xlO"6
1.73xlO"5
8.80xlO"7
5.72xlO"6
1.47xlO"6
9.94xlO"6
2,4

Xobs
mole/g
1.50xlO"3
2.32xlO"3
1.29xlO"3
1.96xlO"3
1.48xlO"3
2.21xlO"3
1.26xlO"3
l.SOxlO"3
1.50xlO"3
2.30xlO"3
1.27xlO"3
1.94xlO"3
,6-TCP

Xss
mole/g
2.63xlO~3
3.40xlO"3
2.68xlO"3
3.55xlO"3
S.llxlO"3
3.60xlO"3
3.31xlO~3
3.72xlO"3
2.81xlO"3
3.43xlO~3
3.02xlO~3
3.58xlO"3


Y Y
xobs/xss
57%
68
45
55
48
61
38
48
53
67
42
54

 Percent removal  as  measured  by  fluorescence  at  pH  5.2.
Note:   See Table  12A for definition  of terms.
                                    104

-------
      TABLE 16.   2,4,6-TCP COMPETITION WITH HUMIC SUBSTANCES  -  pH 9.1
                             CoTCp = 23.80 mg/1
                             10~2 M phosphate buffer

Humic


Type
Commercial
Humic
An' rl


Leaf
Fulvic
An' H


Soil
Fulvic
Acid

Substances
Initial
Cone. % 1
mg/1 TOC Removed
10 75
59
50 57
39
10 61
42
50 44
22
10 30
14
50 21
11
2,4,6-TCP

C
I ec!
1 mole/1
1.05xlO"7
1.32xlO~6
2.35xlO"7
2.55xlO~6
l.OlxlO"7
1.53xlO"6
1.72xlO"7
2.25xlO"6
5.70xlO"8
8.61xlO"7
6.58xlO"8
1.04xlO"6

Xobs
mole/g
1.32xlO"4
3.25xlO"4
l.OSxlO"4
2.45xlO"4
1.30xlO"4
3.21xlO"4
1.09xlO"4
2.50xlO"4
1.30xlO"4
2.35xlO"4
l.lOxlO"4
2.69xlO"4

Xss
mole/g
2.06xlO"4
5.30xlO"4
l.SOxlO"4
6.90xlO"4
2.04xlO"4
5.55xlO"4
2.47xlO"4
6.40xlO"4
1.65xlO"4
4.50xlO"4
1.76xlO"4
4.81xlO"4


obs/Xss
64
61
60
36
64
58
44
39
79
74
62
56

 Percent removal  as measured by fluorescence at pH 9.1.
Note:   See Table  12A for definition of terms.
                                     105

-------
removal as determined by fluorescence apparently is not a good indicator of
the ability of the humic substances tested to compete with chlorophenols.
Further research is required to determine a parameter which better indicates
the ability of the humic materials to compete with the chlorophenols.

     Results for 2,4,6-TCP competition with humic substances at pH 9.1  are
shown in Table 16.  Commercial  humic acid and leaf fulvic acid caused  the
greatest reduction in capacity  for 2,4,6-TCP while soil fulvic acid was not
as effective.   Competition between the humic substances and 2,4,6-TCP  was
significant.  For example, 10 mg/1 leaf fulvic acid reduced the carbon's
capacity at Ceq = 1-01  x 10~7 M by 36 percent while 50 mg/1 leaf fulvic acid
reduced the capacity at Ceq = 1.72 x 10-7 M by 56 percent.  At a 2,4,6-TCP
equilibrium concentration which is higher by an order of magnitude slightly
greater reductions in capacity  resulted which were unexpected.  At pH  5.2,
the opposite occurred,  i.e., less competition was observed as the 2,4,6-TCP
equilibrium capacity was increased.  The reasons for this behavior are not
clear at present.  Percent removal of the humic substances as measured by
fluorescence at pH 9.1  were lower than those at pH 5.2 which is consistent
with the data shown in  Figure 15 for the adsorption of soil fulvic acid as
a function of pH.

     The effect of pH on the ability of the humic materials to compete with
2,4,6-TCP is noteworthy.  At pH 5.2 leaf fulvic acid was the strongest
competitor while commercial humic and soil fulvic acids resulted in nearly
equal competition.  At  pH 9.1,  commercial humic and leaf fulvic acids  com-
peted more strongly with 2,4,6-TCP than did soil fulvic acid.  Although more
complete studies are needed, it is expected that the different molecular
size distributions and  functional group contents of the humic materials used
in these studies lead to differences in adsorbabilities as a function  of pH.
This would explain the  differences in competitive ability of the humic
materials observed at pH 5.2 and 9.1.

     Gauntlett and Packham (1973) investigated the adsorption of monochloro-
phenol at the 0.1 to 1  mg/1 level in the presence of humic acid, fulvic acid
and in Thames River water.  They show a significant reduction in capacity
(approximately 40 percent for the river water as compared with that achieved
in distilled water systems) owing to the presence of these materials which
is consistent with the  results  presented in this section.  Unfortunately,
they did not state the  pH of their studies.  In our study, for example,
10 mg/1 soil fulvic acid at pH  5.2 reduced the  capacity at Ceq = 8.80 x
10-7 M by 47 percent, while at  pH 9.1 the capacity at Ceq = 8.61 x 10"7 M
was reduced by only 26  percent.  This can be accounted for by considering
differences in adsorbabilities  of soil fulvic acid at the two pH values;
at pH 5.2, 48 percent of the soil fulvic acid was adsorbed while at pH 9.1
only 14 percent was removed. Therefore at pH 9.1 less material was com-
peting with 2,4,6-TCP for adsorption sites than at pH 5.2

     When the findings  on competitive adsorption between chlorophenols and
humic substances together with  the information reported previously on  com-
petition between individual chlorophenols are considered, it is apparent
that any testing to determine the best carbon and design criteria for  a
                                    106

-------
given application should be done using the natural water to be treated.
Such factors as the nature and concentration of competing organic materials,
pH and salt content of the water, among others, play a major role in deter-
mining the removal one can expect of a certain component.


ADSORPTION OF POLYNUCLEAR AROMATIC HYDROCARBONS

     An initial test of the adsorbability of benzanthracene from water was
made to first evaluate the effectiveness of the carbon.  Solutions of various
concentrations of benzanthracene were shaken with 10 mg of F-400 activated
carbon for eight hours.  After this time the concentrations were measured by
fluorescence (see Table 17).  Thus these extremely insoluble, non-polar com-
pounds are adsorbed very well on active carbon.
        TABLE 17.  CONCENTRATIONS OF BENZANTHRACENE AFTER EIGHT HOURS

Volume
(ml)
150
150
150
c
o Concentration Remaining
(yg/D (yq/D
242
50
10
12
<1
<1

     Because the PAH was removed to below detection limits at such low
dosages of carbon, the competitive effect of natural organic matter was
examined.  A kinetic study was performed to establish whether the presence
of high concentrations of commercial humic acid would interfere with the rate
of adsorption.  Bottles with and without humic acid were set up with 1 mg/1
of anthracene and taken off each day.  A plot of the residual concentration
vs. time both with and without humic acid is shown in Figure 28.  No effect
of the humic acid was observed.  A series of bottles with 10 mg/1 of carbon
was  set up with half of the bottles containing humic acid and anthracene
and the other half with anthracene alone.  Two bottles, one with and one
without humic acid, were taken off the shaker each day and the anthracene
concentrations were compared.  Plots of residual concentration vs. time are
shown in Figure 42.  No effect of the humic acid was observed.

     In our experiments we were using abnormally high concentrations of PAH.
The anthracene was present as a particulate because concentrations exceeded
the aqueous solubility of these extremely hydrophobic compounds.  Even with
large aqueous concentrations and extremely low carbon dosages we experienced
removal of the PAH to below the detection limit, approximately 1 yg/1, after
shaking for several days.  It was believed that the PAH might associate with
natural organic material  in water and experience a decreased rate of adsorp-
tion or a decreased capacity on carbon.  This was not found to be the case.


                                     107

-------
        IOOO
         100
     0)
     u
     c
     o
    o
          10
                                     A  Without Humic Acid

                                     O  With Humic Acid
                                     Days
Figure Q2.   Kinetics of  anthracene adsorption with and without  humic acid
                                     108

-------
However, it should be stressed that our solutions  were not natural  and  that
the particles of anthracene in solution will  not behave as individual mole-
cules with respect to association with humic  acid.

     On this basis it does not appear likely  that  PAH will associate with
relatively poorly adsorbable humic substances and  result in the  "leakage"
of PAH from carbon beds.   Based on a review of the literature  as  presented
in Section 1, it is possible that other organics such as pesticides may
penetrate carbon beds in  this fashion.
                                    109

-------
                                 REFERENCES

Andelman, J. B.   1973.  World Health Organization, European Standard for
     Organic Matter in Drinking Water.  In:  Proc. 15th Water Quality
     Conference, Organic MatteA -in (UateA Sapptiu'-  OccuAAence, S+.gnifii.c.anc.e.
     and ContAol (V.  L. Snoeyink, ed.), University of Illinois, Urbana, IL.
     70:55.

Burke, J. A., P. A. Mills, and D. C. Bostwick.   1966.  JOUA. A^oc.
         .  Chum.  49:999.
Burtschell, R. H., A. A.  Rosen, F. M. Middleton, and M. B. Ettinger.  1959.
     Chlorine Derivatives of Phenol Causing Taste and Odor.  JOUA. AmeA.
           VJoAkA A64oc.   51:205.
Butler, J. A. V. and C. Ockrent.   1930.  Studies in Electrocapillarity.  III
     JOUA. Phy&. Cham.   34:2841.

Capmau, M. L., W. Chodkiewicz, and P. Cadiot.  1968.  Stereochimie de 1 'Addi
     tion d'Organometalliques a et 3 Insatures sur le (t) Camphre.
     Soc. ChMn.  Trfianc-d.  p.  3233.
Calgon Corp.  1969.  Calgon Activated Carbon Product.  Bull. No. 20-2a,
     Calgon Corp., Pittsburgh, PA.

Crosby, D. G.  and A.  S.  Wong.   1973a.  Photodecomposition of p-Chlorophenoxy-
     acetic Acid.  JOUA.  AgA.  Pood Ckzm.   21:1049.
Crosby, D. G. and A. S.  Wong.  1973b.  Photodecomposition of 2,4,5-Trichloro-
     phenoxyacetic Acid  (2,4,5-T) in Water.  JOUA.  Ag/i. food Ck&n.  21:1052.

Dice, J. C.  1976.  The  Challenge to the AWWA Taste and Odor Control Com-
     mittee.   Proc. AWWA Conference on Water Quality, Atlanta, Dec. 1975,
     AWWA, Denver, CO.

Dostal , K. A., R. C. Pierson, D. G. Hager and G. G. Robeck.  1965.  Carbon
     Bed Design Criteria Study at Nitro, West Virginia.  JOUA. AmeA.
           AMCC.  57:663.
Fieser, L. F. and G.  Ourisson.  1953.  Chromates d'alcools Tertiaires.
     Butt. Sec.  Ckim.  f fiance.,   p.  1152.

Friestad, H.  0., D.  E.  Ott and F.  A. Gunther.  1969.  Automated Colorimetric
     Microdetermination of Phenols by Oxidative Coupling with 3-Methyl-2-
     Benzothiazoline Hydrazone.  AnaJL. Ckzm.   41:1750.
                                    110

-------
Gauntlett, R. B. and R. F. Packham.  1973.  The Use of Activated Carbon in
     Water Treatment.  In:  Proc. Conference on Activated Carbon in Water
     Treatment, University of Reading, Water Research Association,
     Medmenham, England.

Gerber, N. N.  1974.  Microbiological Production of Geosmin.  EPA-670/2-74
     094, U. S. Environmental Protection Agency, Cincinnati, OH.
Gerber, N. N.  1967.
     Actinomycetes.
                      Geosmin, an Earthy-Smelling Substance Isolated from
                     &Lote.ch. and Rioting.  9:321.
Gerber, N. N. and H. A. Lechevalier.  1965.
     Substance Isolated from Actinomycetes.
                                             Geosmin, an Earthy-Smelling
                                             Appt. blLcAob-iot.  13:935.
Gjessing, E. T.
     Humus.   Ann
                 1976.  Physical and Chemical Characteristics of Aquatic
                 Arbor Science Publishers, Ann Arbor, MI.
Glaze, W. H. and J. E. Henderson.  1975.  Formation of Organochlorine
     Compounds from the Chlorination of a Municipal Secondary Effluent.
     Joufc. WateJi Polt. Con&iot f&d.  47:2511.
Handbook o£ Cke.mJJ>&Ly and
     Cleveland, OH.
                                    1967.  48th Ed., Chemical Rubber Co.,
Herzing, D. R., V. L. Snoeyink and N. F- Wood.  1977.  Activated Carbon
     Adsorption of the Odorous Compounds 2-Methylisoborneol and Geosmin.
     JOU/L. AmcA. WateA. Wonkt, /Woe.  69:223.

Jain, J. S. and V. L. Snoeyink.  1973.  Adsorption from Bisolute Systems on
     Active Carbon.  JOUA. Watax. Pott. Con&tot Fed.  45:2463.

Jenkins, D.  1973.  Effects of Organic Compounds - Taste, Odor, Color, and
     Chelation.  In:  Proc. 15th Water Quality Conference, O^igan^a Matt&i -in
     Wat&L SupptieA'-  QdusJiA.ino.iL, S-ignifi-icanzz and Con&iot (V. L. Snoeyink,
     ed.), University of  Illinois, Urbana,  IL.  70:15.
Kaiser, K. L. E. and B. G. Oliver.  1976.
     genated Hydrocarbons in Water by Gas
     48:2207.
                                           Determination of Volatile Halo-
                                          Chromatography.  Anat. Ck&m.
Kikuchi, T., T. Mimura, Y. Itoh, Y. Moriwaki , K. Negoro, Y. Masada and
     T. Inoue.  1973a.  Odorous Metabolites of Actinomycetes, Biwake-C and
     -D Strain Isolated from the Bottom Deposits of Lake Biwa.  Identifica-
     tion of Geosmin, 2-Methylisoborneol, and Furfural.  Ck&tn. and Pkatm.
     KvJUL.   21:2339.

Kikuchi, T., T. Mimura, K. Haranaya, H. Yano, T. Arinoto, Y. flasada and
     T. Inoue.  1973b.  Odorous Metabolite of Blue-Green Algae-'  Sck
-------
Korenman, Y.  1974.  Distribution of Chlorophenols between Organic Solvents
     and Water.  JOUA. Appi. Ckm. - USSR.  47:9 (part 2).

Ku'hn, W. and H. Sontheimer.  1973a.  Several Investigations on Activated
     Charcoal for the Determination of Organic Chloro Compounds,  l/om Wo6-6eA.
     41:65.

Ku'hn, W. and H. Sontheimer.  1973b.  Analytic Determination of Chlorinated
     Organic Compounds with Temperature Programmed Pyrohydro lysis,  l/om
              41:1.
Langmuir, I.  1918.  The Adsorption of Gases on Plane Surfaces of Glass, Mica
     and Platinum.  JOUA. AmeA. Ckm. Son..  40:1361.

Lee, C. F-  1967.  Kinetics of Reactions Between Chlorine and Phenolic
     Compounds.  In:   Psbinc/iptu and Application* o£ UateA Ckem'iA&ty (S.
     Faust and J. Hunter, eds.), Wiley and Sons, Inc., New York.

Leenheer, J. A. and E. D. W. Huffman.  1976.  Classification of Organic
     Solutes in Water Using Macroreticular Resins.  Submitted to JOUA. ojj
     ReieoAc^ for publication.

Lenz, V. W.  1911.  Zur Prufung des Kampfers.  AAC/U.V. dun. P/ioAmaz-te..
     249:286.

Love, 0. T., Jr., G.  G. Robeck , J. M. Symons and R. W. Buelow.  1973.
     Experience with Activated Carbon in the U.S.A.  Proc. Conference on
     Activated Carbon in Water Treatment, University of Reading, Water
     Research Association, Medmenham, England,  p. 279.
Malkonen, P.  1964.  Ann. Ac.ad. Sex. FemUcae SeA. All.  No. 128.

McCaull, J. and J. Crossland. 1974.  Wat&i Pollution (B. Commoner, ed.),
     Harcourt Brace Jovanovich, Inc., New York.

McCreary, J. J. and V. L. Snoeyink.  1977-  Granular Activated Carbon in
     Water Treatment.  JOUA. AmeA. Wo£e.A WoAfci AMOC., in press.

McGinnes, P. R. and V. L. Snoeyink.  1974.  Determination of the Fate of
     Polynuclear Aromatic Hydrocarbons in Natural Water Systems.  Res.
     Rept. No. 80, Water Resources Center, University of Illinois at
     Urbana-Champaign, Urbana, IL.

Medsker, L. L., D. Jenkins and J. F. Thomas.  1968.  Odorous Compounds  in
     Natural Waters:  An Earthy-Smelling Compound Associated with Blue-
     Green Algae and Actinomycetes.  Env-ision. Sex., and Tzcknol.  2:461.

Medsker, L. L., D. Jenkins, J. F. Thomas and C. Koch.  1969.  Odorous Com-
     pounds in Natural Waters:  2-exo-hydroxy-2-methylbornane, the Major
     Odorous Compound Produced by Several Actinomycetes.  Env-iion. Se-c.
     and Tachnol.   3:476.
                                    112

-------
Nakanishi, K.  1962.  In^fcuizd AbAostption Spuc&ioAcopy, Holden-Day, Inc.,
     San Francisco, p. 24.

Narkis, N. and M. Rebhun.  1975.  The Mechanism of Flocculation Processes in
     the Presence of Humic Substances.  JOUA. Ame/i. Uatoji WoAfe4 Awoc.
     67:101.

Pharmacia.  1974.  Sephadex:  Gel Filtration in Theory and Practice.
     Upplands Graf i ska AB, Sweden.

Piet, G. J.,  C.  J. Zoetman and J. A. Kraayeveld.  1972.  Earthy-Smelling
     Substances in Surface Waters of the Netherlands.  PAOC.  Soc.  Wo^eA
     T/izat. and Exam.   21 :281.

Robeck, G. G.  1975.  Evaluation of Activated Carbon.  Water Supply Research
     Laboratory, National Environmental Research Center, Cincinnati, OH.

Rook, J. J.  1976.  Haloforms in Drinking Water.  JouA. AmeA. (riat&i
             68:168.
Rosen, A. A., C. I. Mashni and R. S. Safferman.  1970.  Recent Developments
     in the Chemistry of Odor in Water:  The Cause of Earthy-Musty Odor.
     PAOC. Wat&i Tsizat. and Exam.  10:106.

Safferman, R. S., A. A. Rosen, C. I. Mashni and M. E. Morris.  1967.  Earthy-
     Smelling Substance from a Blue-Green Algae.  Envision. Sc.i. and Tzcknot.
     1:429.

Schnitzer, M. and S. U. Khan.  1972.  Hamic. SubAtanceA -in tkn. Environment,
     Marcel Dekker, Inc., New York.

Schweer, K. H., F.  Fuchs and H. Sontheimer.  1975.  Untersuchungen zur
     Summarischen Bestimmung von Organisch Gebundenem Schwefel in Wassern
     and auf Aktivkohlen.  l/om Wo^eA.  45:29.
Sigworth, E. A.  1957.  Control of Taste and Odor in Water Supplies.
                               49:1507-
Silvey, J. K. G., D. E. Henley, R. Hoehn and W. C. Nuney.  1976.  Musty-
     Earthy Odors and Their Biological Control.  Proc. Conference on Water
     Quality, Atlanta, Dec. 1975, AWWA, Denver, CO.

Snoeyink, V. L., W. J. Weber and H. B. Mark.  1969.  Sorption of Phenol and
     Nitrophenol by Active Carbon.  Envision. Sex., and Te.c/mo£.  3:918.

Sontheimer, H.  1974.  Use of Activated Carbon in Water Treatment Practice
     and its Regeneration.  Special Subject No. 3, Lehrstuhl fur Wasser-
     chemie der Universitat Karlsruhe, Karlsruhe, West Germany.
                                    113

-------
Sontheimer, H. and D. Maier-  1972.  Untersuchungen zur Verbesserung  der
     Trinkwasseraufbereitungstechnologie am Niederrhein (1.  Bericht).   Go*
     inn A ll\n L f.n H Unnln \\\n A A o h /AhlilnAA Oh .  1 1 "\ ' 1 ff7 .
     and WaAAHfi^ack Woi-ieA/AbwiaMe/i.  113:187.
Stahl , E. (ed.).  1969.  Tkin-Layeji Ckwmatogsiaphy.   Springer-Verlag,  New
     York.  p. 224.
Standard ttnthodb ^on tke. EKcmi.Yiat.J-on ofi Wat&i and Ua&tewateJi,   1975.   14th
     Edition, American Public Health Association, Washington,  D.C.

Stevens, A. A. and J. M. Symons.  1977.  Measurement of Trihalomethane and
     Precursor Concentration Changes Occurring during Water Treatment and
     Distribution.   Accepted by JouA. AmeA. UateA Wonkh A^^oc.  for  publi-
     cation, October 1977.

Stevens, A. A., C. J. Slocum, D. R. Seeger and G. G. Robeck.   1976.   Chlori-
     nation of Organics in Drinking Water.  JOUA. Ame/t. Watasi  Woife/i  Ai^oc.
     68:615.

Stevenson, F. J.  1965.  Gross Chemical Fractionation of Organic Matter.
     In:  M&tkodA o& Soit knalyA-u , Part 2, Agronomy Monograph #9 (C.  A.
     Black it aJL. , eds . ) , American Society of Agronomy, Ma'dison, WI,  p.  1409

Symons, J. M.  1976.  Interim Treatment Guide for the Control  of Chloroform
     and Other Trihalomethanes.  Water Supply Research Division, Municipal
     Environmental Research Laboratory, Environmental Protection Agency,
     Cincinnati , OH.

Toivonen, H.  1968.  Uber Die Reaktionen Des 2-Methyl isoborneols und  der
     2-Methylfenchols in Konzentrierter Salpetersaure.  T&&iake.dAon  L&ttzu.
     p. 3041.

U.S. Environmental Protection Agency.  1971.  Methods for  Organic Pesticides
     in Water and Wastewater.  National Environmental Research Center,
     Cincinnati , OH.

U.S. Environmental Protection Agency.  1975.  Preliminary  Assessment  of
     Suspected Carcinogens in Drinking Water - Interior Report to Congress.
     Office of Toxic Substances, Washington, D.C.

Ward, T. M. and F. W. Getzen.  1969.  A Model for the Adsorption of  Weak
     Electrolytes on Solids as a Function of pH--l .  Joun.. Cottoi.d and
               SCA-.  31:441.
Ward, T. M. and F. W. Getzen.  1970.   Influence of  pH  on  the  Adsorption of
     Aromatic Acids on Activated Carbon.   EniMAon.  Sc-t. and Jichyiot.   4:64.

Weber, W. J., Jr. and J. C. Morris.  1964.  Adsorption in Heterogeneous
     Aqueous Systems.  Jou/i. Ame/z.. W
-------
Zeiss, M. M. and D. A. Pease.  1956.  Chromate Esters.   II.   Reactions of
     Di-2-Methylfenchyl Chromate.  JOUA.  AmeA. Ckw.  Sec.   78:3182.

Zelinsky, N.  1901.  Uber eine Synthese der Cyclischentertiaren Alkohole mit
     Hulfe von Magnesiumhalogenalkylen.  BeA.  V&>ck.  Cham.  Ge/s.  34:2877.
Zogorski, J. S. and S. D. Faust.  1974.  Removal of Phenols from Polluted
     Waters.  New Jersey Water Resources Research Institute, Rutgers Univ.

Zogorski, J. S. and S. D. Faust.  1976.  The Effect of Phosphate Buffer on
     the Adsorption of 2,4-Dichlorophenol and 2,4-Dinitrophenol .   JOUA.
     Environ. Set. H&altk.  All (8&9):501.
                                    115

-------
                                   TECHNICAL REPORT DATA
                            (Please read Instructions on the reverse before completing)
1. REPORT NO.
    EPA-600/2-77-223
                              2.
                                                           3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
    ACTIVATED  CARBON ADSORPTION OF  TRACE
    ORGANIC  COMPOUNDS
              5. REPORT DATE
               December 1977  (Issuing Date)
              6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
    Vernon L.  Snoeyink,  John J. McCreary  and
    Carol J. Murin
                                                           8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
    Dept. of  Civil  Engineering
    University  of Illinois
    Urbana, Illinois  61801
              10. PROGRAM ELEMENT NO.

                  1CC614
              11. CONTRACT/GRANT NO.

                 R-S03473
 12. SPONSORING AGENCY NAME AND ADDRESS
    Municipal  Environmental Research Laboratory- Cin., Of
    Office of  Research and Development
    U.S. Environmental Protection Agency
    Cincinnati,  Ohio  45268
              13. TYPE OF REPORT AND PERIOD COVERED
               Final, 1/6/75  to 7/5/77
              14. SPONSORING AGENCY CODE
               EPA/600/14
 15. SUPPLEMENTARY NOTES
    Project  Officer:   Alan A. Stevens,  513/684-7228
 16. ABSTRACT
         Research  was conducted to determine  how effectively humic  substances
    and the  trace  contaminants 2-methylisoborneol (MIB), geosmin, the
    chlorophenols  and polynuclear aromatic  hydrocarbons were adsorbed  by
    activated  carbon under the competitive  adsorption conditions  encountered in
    natural  waters.   Humic substances  compete with MIB and geosmin  for
    adsorption sites on activated carbon  and  significantly reduce its  capacity
    for these  compounds.   These naturally occurring odorous compounds  were
    found to be much more strongly adsorbed than the humic substances.
         Both  the  chlorophenols and the polynuclear aromatic hydrocarbons are
    very strongly  adsorbed.  Strong competition was observed between anionic
    and neutral species of 2,4-dichlorophenol and 2,4,6-trichlorophenol.  The
    presence of the  various humic substances  caused a significant reduction in
    chlorophenol adsorption capacity.  Humic  acid did not interfere with the
    rate of  adsorption of a model polynuclear aromatic hydrocarbon, anthracene.
17.
                                KEY WORDS AND DOCUMENT ANALYSIS
                  DESCRIPTORS
                                              b.IDENTIFIERS/OPEN ENDED TERMS
                           c.  COSATI Field/Group
    Activated  Carbon Treatment,
    Potable Water,  Odor Control,
    Organic Compounds,  Actinomyces
   Competitive Adsorption,
   Humic  Acids, Chloro-
   phenols,  Geosmin,
   Methylisoborneo.1,
   Taste  and Odor
       13 B
 3 DISTRIBUTION STATEMENT

  Release  to  Public
 19. SECURITY CLASS (This Report)
  Unclassified
21. NO. OF PAGES
    126
                                              20. SECURITY CLASS (This page)
                                                Unclassified
                                                                         22. PRICE
EPA Form 2220-1 (9-73)
16
                                                   •ft U. S. GOVERNMENT PRINTING OFFICE: 1978-757- 1W6660 Region No. 5-1

-------