EPA/600/2-91/033
                                          July 1991
THE SWELLING PROPERTIES OF SOIL ORGANIC MATTER AND THEIR
  RELATION TO  SORPTION OF NON-IONIC ORGANIC COMPOUNDS
                           by

                     WILLIAM G. LYON

                           and

                     DAVID E. RHODES

         ManTech Environmental Technology, Inc.
                   Ada,  Oklahoma 74820
               EPA CONTRACT NO.  68-C8-0025
                     Project Officer

                       Roger Cosby

        PROCESSES  AND SYSTEMS RESEARCH DIVISION
    ROBERT S. KERR ENVIRONMENTAL RESEARCH LABORATORY
                   ADA,  OKLAHOMA 74820
    ROBERT S. KERR ENVIRONMENTAL RESEARCH  LABORATORY
           OFFICE OF RESEARCH AND DEVELOPMENT
          U.S. ENVIRONMENTAL PROTECTION AGENCY
                  ADA,  OKLAHOMA 74820

-------
                                    TECHNICAL REPORT DATA
                             (Please read Instructions on the reverse before complef
1. REPORT NO.
  EPA/600/2-91/033
                               2.
                                                                   PB91t217406
4. TITLE AND SUBTITLE
                                                             5. REPORT DATE
  THE SWELLING PROPERTIES OF SOIL ORGANIC MATTER AND THEIR
  RELATION TO SORPTION OF NON-IONIC' ORGANIC COMPOUNDS
             6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)

  William G. Lyon  and  David E.  Rhodes
                                                             |a. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
  ManTech Environmental  Technology, Inc.
  P.O. Box 1198
  Ada, OK 74820
                                                              10. PROGRAM ELEMENT NO.
                ABPC1A
             11. CONTRACT/GRANT NO.
                                                                  68-C8-0025
12. SPONSORING AGENCY NAME AND ADDRESS
  R.S. Kerr Environmental Research Laboratory
  U.S. ENvironmental  Protection Agency
  P.O. Box 1198
  Ada, OK 74820
             13. TYPE OF REPORT AND PERIOD COVERED
               Final Report  8/1/88-9/30/90 -
             14. SPONSORING AGENCY CODE
                EPA/600/15
 15. SUPPLEMENTARY NOTES
  Project Officer:   Roger L.  Cosby    FTS: 743-2320
16. ABSTRACT
       A method  has  been developed to measure the swelling properties  of  concentrated
  natural organic  materials in various organic liquids, and has been applied to various
  peat, pollen,  chitin and cellulose samples.  The swelling of these macromolecular
  materials  is the volumetric manifestation of bulk sorption, i.£., sorption by dissolution
  (or partitioning)  of the sorbed liquids into the macromolecular  solid phase.   Direct
  evidence for the existence of this category of sorbed materials  has  been obtained for
  soil organic materials by the present research; swelling in liquids  has long  been known
  in coals and polymers.
       Bulk  sorbed molecules are thought to be inaccessible to direct  biological attack,
  and may represent  a continuing source of low-level rebound contamination of groundwater
  at a polluted  site.   Equilibration of bulk-sorbed molecules with liquid phases surround-
  ing the particles  is kinetically slow (diffusion limited) relative to sorption and
  fluid movement,  and this sluggishness is probably responsible for some  nonequilibrium
  sorption phenomena seen in soil column flow experiments.
       Molecules with molar volumes greater than about 93 cnH mol~l appear to be strongly
  excluded from  sorption inside the soil organic materials studied in  this work.  In
  contrast,  cellulose excluded molecules with molar volumes greater than  about  88 cm^
  mo-1-1';
       Extensive bibliographies included.-
17.
                                 KEY WORDS AND DOCUMENT ANALYSIS
                   DESCRIPTORS
                                                 (.IDENTIFIERS/OPEN ENDED TERMS
                            L.  COSATi Field, Group
 SWELLING
 SOIL  ORGANIC MATTER
 SORPTION
 PARTITIONING
 HUMIC MATERIALS
 CELLULOSE
 CHITIN
                              POLLEN
 ORGANIC SOLVENTS
 NAPL
18. DISTRIBUTION STATEMEN1
  RELEASE TO PUBLIC
 EPA Form 2220-1 (R«v. 4-77)    PREVIOUS EDI T.ON is OBSOLE TE
19. SECURITY CLASS (This Report)
  UNCLASSIFIED
n NO. OF PAGE;
    138
20. SECURITY CLASS t Tins pate •

  UNCLASSIFIED
                                                                            22. PRICE

-------
                       DISCLAIMER

     The information in this document has been funded wholly
or  in  part  by the  United  States Environmental  Protection
Agency under Contract Number  68-C8-0025 to  NSI  Technology
Services Corporation.   It has been subject  to  the Agency's
peer and administrative review, and it has been approved for
publication as an EPA document.

Although mention of  trade  names,  commercial products  and
companies  is made throughout  this report  for  the  sake  of
completeness, such mention does not constitute endorsement or
recommendation for use.
                           11

-------
                             FOREWORD

     The  Environmental  Protection  Agency  was  established  to
coordinate administration of the major Federal programs designed to
protect the quality of our environment.

     An important part of the Agency's effort involves the search
for information about environmental problems,  management techniques
and new technologies through which optimum use of the Nation's land
and water resources can be assured, and the threat pollution poses
to the welfare of the American people can be minimized.

     EPA's Office of Research and Development conducts this search
through a network of research facilities.

     As one  of the facilities,  the Robert  S.  Kerr Environmental
Research  Laboratory  is  the Agency's  center   of  expertise  for
investigation of the soil  and subsurface environment.  Personnel at
the laboratory are responsible for management of research programs
to  (a)  determine  the  fate,  transport  and transformation rates of
pollutants in  the soil,  the unsaturated zone  and the  saturated
zones of the  subsurface environment; (b) define  the processes to be
used  in characterizing the soil and  subsurface environment  as a
receptor of pollutants; (c) develop techniques for predicting the
effect  of  pollutants  on  ground  water,  soil  and  indigenous
organisms; and (d)  define and demonstrate  the  applicability and
limitations of using natural processes,  indigenous to the soil and
subsurface environment, for the protection of this resource.

     This  report  contributes to  knowledge   of  the  environmental
compartments,  previously  lumped   together   as  "sorbed  phases,"
through which  many  pollutants pass during their transport in the
subsurface.  This information should  assist the  construction of
more realistic computer models involving  sorption of pollutants in
the subsurface by organic materials.
                                     Clinton W. Hall, Director
                                     Robert S. Kerr Environmental
                                       Research Laboratory
                               111

-------
                            ABSTRACT

     A method has been developed to measure the swelling properties
of  concentrated natural  organic materials in  various  organic
liquids, and has been applied to various peat,  pollen, chitin and
cellulose samples.   The swelling of these macromolecular materials
is the volumetric manifestation of bulk sorption, i.e., sorption by
dissolution  (or partitioning)  of the  sorbed  liquids into  the
macromolecular solid phase.   Direct  evidence for the existence of
this  category of  sorbed  materials has been  obtained for  soil
organic materials by the present research; swelling in liquids has
long been known in coals and polymers.

     Bulk sorbed molecules are thought to be  inaccessible to direct
biological attack, and may represent  a  continuing source  of low-
level,  "rebound" contamination of groundwater  at  a polluted site
following attempted pump-and-treat remediation.  Equilibration of
bulk-sorbed molecules with liquid phases surrounding the particles
is kinetically  slow  (diffusion limited)  relative  to sorption and
fluid movement, and this sluggishness  is probably responsible for
some  nonequilibrium  sorption  phenomena seen in soil  column flow
experiments.

     Molecules with molar volumes greater than about 93 cm3 mol"1
appear  to be  strongly  excluded  from sorption  inside the  soil
organic materials  studied in  this work.  In contrast,  cellulose
excluded molecules with molar volumes  greater than 88 cm3  mol"1.

     Besides the size exclusion factor, the degree of swelling of
soil organic materials in different liquids is controlled mainly by
site-specific,  generalized  acid-base interactions  between  the
sorbed molecules and the various acidic sites within soil  organic
materials.   The swelling spectra  observed for  soil materials are
complex, and completely unlike the simple Gaussian swelling spectra
observed for polymers like rubber (cross-linked polyisoprene) and
for  some  coals.   In these latter  materials  the intermolecular
forces  are  dominated by non-specific dispersion  forces  (van der
Waals  interactions),  and  can be  adequately  treated by  simple
equations  (Flory-Huggins-Rehner theory)  involving the solubility
parameters of the liquid and the swelling substrate.

     Swelling in morpholine  appears to be a characteristic of soil
organic materials containing  free cellulose.   Unfortunately, the
cellulose within natural ligno-cellulosic plant debris apparently
behaves differently  from  free cellulose, so that swelling alone
does  not  provide a  simple  measure of  humification  in soils or
peats.  We speculate that the intimate association of lignin with
the cellulose at the molecular level blocks  access to the specific
sites (alcoholic-OH groups)  on the cellulose with which morpholine
interacts most strongly.  Free cellulose does,  however, appear to
be present in pollen intine membranes.
                                IV

-------
                        TABLE OF CONTENTS


FOREWORD	  iii

ABSTRACT	   iv

ACKNOWLEDGEMENTS	 viii

INTRODUCTION	  1

     Tools available for studying insoluble organic matter...  2
     The swelling phenomenon	  5
     The connection between sorption and swelling	  5
     Operational definition of volumetric swelling	  8
     Theory of swelling for macromolecular materials	  9

EXPERIMENTAL	 12

     Standard operating procedures developed for this project 12
     Quality assurance considerations	 12
     Details and purpose of preparative steps	 13
     Provenance  and processing of organic materials used in
     the swelling experiments	 15
     Characterization data for organic materials used in the
     swelling experiments	 17
     Solvents used for swelling experiments	 19
     Swelling measurements	 23

INTERPRETATIONS AND CONCLUSIONS	 50

     General observations	 50
     Peat-Like Materials	 50
     Pine and Oak Pollen	 51
     Cellulose	 51
     Chitin	 52
     Molecular Size-Exclusion Effects	 52
     Sample Similarity Based on Swelling	 56
     Consequences for Environmental Studies	 61
     Suggestions for Future Work	 61

REFERENCES	 64

SUPPLEMENTARY DOCUMENTS

1.   Derivation of a Simplified Thermodynamic Equation of State for
     Swelling Materials

2.   Review:  Properties of Peats/ and Methods of Preparation and
     Characterization prior to Sorption and Swelling Studies

3.   Review:  Thermodynamics of Polymer Swelling as  an Analogy for

-------
     Soil Organic Matter

4.   Review:  The swelling of coal and Kerogen, and its Relation to
     the Swelling  and  Sorption  Characteristics  of Soil  organic
     Matter

5.   Review:   The  Swelling  of  Cellulose and its Relation  to  the
     Swelling and Sorption Characteristics of Soil  Organic Matter

6.   References on  Drago's E and C Formulation of Generalized Acid-
     Base Interactions  and its  Application to Sorption of Organic
     Molecules on Soils, Coals, Polymers and Minerals

7.   RSKSOP-104, Swelling Spectrum of Organic Matter

DIAGRAMS

     I.  Soil Constituents
     II.  Sorption Categories in Natural Soil Materials
TABLES
     I.   Basic Analytical  Data  for  Organic  Materials used  in
     Swelling Studies

     II.   Classification of Solvents for Swelling Experiments by
     Major Functional Groups

     III.  Properties of Solvents for  Swelling Experiments

     IV.   Correlation  Coefficients  (r-Values)  between  "Best"
     Spectra

     V.  Statistics of Differences between Duplicate Spectra

     VI.  Maximum Observed Swelling for each Sample and the Solvent
     for which this Occurred

     VII.  Swelling of Peat-Like Materials

          A.   Results Uncorrected for  Ash
          B.   Results Corrected for Ash

     VIII.  Swelling of Pine and Oak Pollen

          A.   Results Uncorrected for  Ash
          B.   Results Corrected for Ash

     IX.  Swelling of Chitin and Cellulose
                               VI

-------
FIGURES



     I.  Example of Duplicate Swelling Spectra for Pine Pollen



     II.  Absolute Deviation vs. Average Swelling for Pine Pollen



     IIIA.  Raw Swelling vs. <50/ Peat-Like Materials



     IIIB.  Ash-Corrected Swelling vs. £„,  Peat-Like Materials



     IV. Swelling vs. SQr  Acid and Ca-Exchanged Canadian Peat



     V.  Swelling vs. 6Q,  Pine and Oak Pollen



     VI.  Swelling vs. 
-------
ACKNOWLEDGEMENTS

     Many thanks are due  Stan  Shannon,  Joyce Bergin  and  all the
library staff  at  Kerr Lab  for  their considerable  assistance in
literature searching  and  in obtaining  copies of the references
cited and discussed herein.
                             Vlll

-------
                           INTRODUCTION

     Organic matter encountered in  soils and sedimentary rocks is
typically  a complex  mixture  of materials  of diverse  origins.
Characteristics of these mixtures  that  influence  the sorption of
various organic pollutants species from  mobile  fluid phases are of
great importance in computer  models of  pollutant  movement in the
subsurface,  because  the  organic  materials  in  rocks and  soils
usually far  surpass the mineral  constituents  in their ability to
sorb  non-ionic substances.    An understanding  of the  sorption
phenomena  associated  with  natural  organic  materials will  have
application  to environmental problems  in  a variety  of  settings
including  the  sorption  of pollutants  from contaminated  ground
water, and  the sorption of gasoline or  synfuel constituents near
leaking underground storage tanks.

     Much work has been devoted to  characterization of the soluble
fractions of natural organic materials.  Relatively little work has
emphasized  the sorption characteristics  of the  insoluble  humin
fraction of soils,  or the  kerogens  and coals of sedimentary rocks.
Most papers dealing with sorption onto soil organic matter propose
that sorption  for many water-insoluble,  non-ionizing compounds is
controlled by  an hydrophobic mechanism,  i.e.,  these compounds are
not so much  attracted  to  soil organic matter  as they are ejected
from  the  aqueous  phase  by the  strong  mutual  attractive  forces
between water molecules.  For such a mechanism,  the detailed nature
of  the  solid  organic  sorbents should  not matter very much,  and
this, perhaps, explains the  general lack  of  concern about  the
properties of  solid organic soil materials.

     A few  papers  have suggested that much of the sorption onto
soil  organic matter  should be treated  as  a  partitioning process
rather than true surface adsorption  (see especially the papers by
Chiou et  al. .  1990,   Mingelgrin and Gerstl, 1983,  Chiou et al. .
1983,  and  Freeman  and Cheung,  1981).    Many  aspects  of  these
treatments  remain  obscure  and somewhat  speculative,  however.   To
date  very  little  has been  done  to   investigate  directly  the
macromolecular properties  of solid soil organic  materials, although
analogous work on coals, kerogens, and other natural macromolecular
systems suggests that much information pertinent to environmental
problems  can be obtained  from  studies  of this  type,  especially
studies of swelling phenomena.1   We  believe the present work fills
this  gap   and  sheds  light  on  the nature  and  limits  of  this
partitioning form of sorption.

     Subsurface organic matter other than  soil organic matter is
also of potential  environmental  interest  concerning swelling and
sorption characteristics.  Although many aquifers consist  of rather
     1These  swelling  phenomena  of   soil   organic  matter  are
completely distinct  from the various  swelling-related phenomena
associated with clay minerals in aqueous systems (e.g.  shrinkage of
clay liners exposed to organic solvents).

-------
organic-poor sediments,  it should be noted that within the heavily
industrialized coastal plains of the  United States,  near-surface
lignite deposits are very commonly found in close association with,
or as a part of the aquifers.  For example, the lignite deposits of
the  Eocene Wilcox  formation in East  Texas and Louisiana  are
saturated  with  ground water  and are a  significant  part  of  the
aquifers in the region.  Indeed,  coal  exploration in this region is
frequently assisted  by  an examination  of  water well  drilling
records, because near surface lignite is so commonly encountered.
(See Kaiser, 1974, and Snider and Covay,  1987  for further details).

TOOLS AVAILABLE FOR STUDYING INSOLUBLE ORGANIC MATTER

     The  mixture of  insoluble  macromolecular  substances  that
compose  the  major  portion  of  soil  organic   matter  presents
significant difficulties in its  chemical analysis.   Furthermore,
these substances usually occur as a minor part of a  very complex
mixture containing mineral matter,  moisture and  partially soluble
organic materials that produce interferences and complications in
the  interpretation  of data taken  on the whole  mixture.   These
substances are difficult to remove in a sufficiently gentle manner
to study  the  macromolecular fraction by itself .(see  Diagram  I).
For this  reason,  the  present  work  is necessarily devoted  to  the
study of model materials, available  in concentrated form, which are
believed to provide reasonably faithful  analogs  of macromolecular
organic materials that  actually occur  in  soils  and  sedimentary
rocks.

-------
                            DIAGRAM I

                        SOIL CONSTITUENTS

     Soil on the microscopic  scale is a heterogeneous mixture of
solid and fluid materials:


                 Crystalline  (quartz, feldspar, etc.)

Mineral Matter
               \
                 Amorphous  (rusts, phytoliths, cutans, etc.)
                  Sorbed
      Moisture — Capillary
               \
                \
                  Water of Hydration  (e.g., gypsum, opal)
                   Water Soluble  (sugars, tannins,
                 /                some fulvic acid, etc.)
Organic Matter --- Soluble in Organic Solvents  (Bitumen8)
              \
               \               Humic and Fulvic Acids
                \            / (NaOH extractables)
                 \          /
                   Insoluble — Macromolecules:
                            \    Cellulose15
                             \   Chitinc
                              \
                               Cross-Linked Macromolecules
                                 Humin/Kerogend (lignin)
                                 Sporopollenin6 (carotene)

-------
Notes to Diagram I

     aThiscategory   includes   waxes,    resins,   lipids,   and
miscellaneous other  soluble products,  including some  humic and
fulvic acids in their acid forms.  The split obviously depends on
the solvents used.  Some of these materials (e.g., resin bodies and
waxy, leaf cuticles)  can be inherited as discrete,  separate phase
particles  and  coatings  from  the  parent,  structured  botanical
material.

     bCellulose  is  a  biopolymer composed of glucose repeat units,
which can be hydrolyzed by  strongly acidic solutions.   The three
slightly acidic  hydrogens  (alcoholic) per repeat unit are the major
sites of  interaction  with organic  bases.   See Kremer  and Tabb,
1990.

     cAlthough chitin is  usually thought of in  association with
arthropod exoskeletons, it  is  also found in the walls  of fungal
cells.  It is a polymer of amino-sugar units,  and like cellulose,
it can be hydrolyzed by strong acids.  See Muzzarelli,  1977.

     dHumin is commonly defined operationally as that soil organic
matter  fraction that  is  insoluble  in  dilute  NaOH  solutions.
Unfortunately, this definition is  not  a  very  discriminating one,
and  humin  so  defined may  represent both unaltered  and degraded
lignin-derived material,  and even cellulose.  (In our experience,
there is  no  clear-cut  endpoint in  extractions with NaOH;  more
extraction simply seems to produce more extract.)

     Kerogen  and coal  include humin-like materials  within their
broad definitions,  but also can  include  materials  of much higher
H/C  ratio  (e.g.,   resinite,  sporinite,  alginite)  than  found  in
typical soils.  Generally,  the terms kerogen and coal are reserved
for materials  of "higher rank" (i.e., more thermally mature organic
materials)  than soil  organic matter.   Also,  the  mineral matter
associated with coals and kerogens is dramatically altered by the
more extensive thermal  and geochemical  history of these materials;
pyrite,   formed  by  sulfate-reducing  bacteria  under  anaerobic
conditions, is very commonly associated with both kerogen and coal.

     eSporopollenin is found in the exine wall of spores and pollen
and is thought to be a condensation product  formed from carotenoid
pigments and  their esters.   In acidic  and anaerobic  soils and
sediments,  it  is quite resistant  to degradation; however, alkaline
conditions cause etching  of the exine.   Presumably washing with
NaOH or sodium pyrophosphate solutions to remove humic  acids can
destroy some of  the sporopollenin.   See Brooks and Shaw, 1968 and
1972.

-------
     Since the macromolecules are largely involatile and insoluble,
the major tools available for their study are various spectroscopic
techniques  (NMR,  FTIR,  etc.)/  various kinds  of pyrolysis  (PY-GC,
PY-MS,   PY-GC/MS,   etc.)/   flow  microcalorimetry  and  certain
techniques  developed to characterize polymers,  such  as solvent
swelling  spectroscopy  (chemo-elastic  effects)  and reversed-phase
gas and  liquid chromatography-  These all have  their  place in a
comprehensive  investigation of subsurface  organic matter.   The
technique that is the particular focus of this report  is the method
of  solvent  swelling spectroscopy which  probes the chemo-elastic
properties.

THE SWELLING PHENOMENON

     Macromolecular  substances swell  when placed in  contact with
various fluids1; the amount of  swelling depends both on the nature
of the fluid, and of the macromolecular material.   Strongly cross-
linked materials swell less than weakly cross-linked  materials.
Such swelling is a property of the  insoluble macromolecular network
itself,  and  it occurs  even when  all associated soluble materials
have been removed by exhaustive solvent extraction.

     Swelling  is the dissolution of  small molecules of more mobile
substances into the solid.  The macromolecules act as  a solvent for
these  smaller  molecules,  and the  swollen  phase represents  an
unusual  kind of  solution.   An alternate point  of view  is that
swelling  represents  the  solvation  of   internal  macromolecular
"surfaces" by  the smaller  fluid molecules.

THE CONNECTION BETWEEN SORPTION AND SWELLING

     Swelling  represents   a  volumetric  manifestation  of certain
kinds of sorption involving bulk sorption (partitioning) into soil
organic matter.  The various categories of sorption in soil organic
matter  are detailed in Diagram II.
     1Here we  restrict the  discussion to  fluids with  a single
molecular  constituent.   When  two constituents  are  present,  a
richer, more  complex  set  of phenomena can occur,  such  as gel-
collapse,   phase transitions,  where a  swollen gel  can  suddenly
decrease   in  volume   when   exposed  to  a  fluid  of  different
composition.

-------
                           DIAGRAM II

       8ORPTION CATEGORIES IN NATURAL SOIL ORGANIC MATTER

                                   Insoluble Solid Phases8
                                  / (Humin,  Cellulose,  Chitin)
                     Swelling &   /
                   Bulk Sorption<
                 / (partitioning)\
                /                 \ Extractable Phases
               /                    (Waxes,  Resins,  Lipids, etc)
Organic Matter<
             i  \                       Polarc & Hydrogen Bonding
                \                    /                 d
                 \ Surface Sorption <-  Cation Exchange0
                     (adsorption)     \
                                     \ Non-Polar (hydrophobic)


               	 Capillary Condensation6
                      (micropores)

Notes to Diagram II

     Dissolutionin  these  materials  is  accompanied  by  size
exclusion.   Also,  within the  solid macromolecular  structures,
different types of  organic molecules will preferentially solvate
various  sites:    polar,   non-polar,   hydrogen  bonding,  cation
exchange, etc.   See Tegelaar  et  al., 1989 for a concise summary of
known biomacromolecules and  their potential for  preservation in
soils and sediments.

     Dissolution   of  various   sorbed  molecules  occurs  without
significant size exclusion. Naturally, too many "sorbed" molecules
in the  form of  excess  solvent will extract and mobilize these
materials.  Here  we emphasize the role of native waxes, resins,
etc., rather than that of anthropogenic  residual separate-phase
liquid contaminants which can act similarly.

     cPolar is used here only  to denote a  class  of  molecules.
Various authors (for example, Fowkes, 1980) have shown that dipole-
dipole  interactions between polar  molecules  in  a   liquid   (in
contrast to the vapor phase)  represent only a very tiny portion of
the  intermolecular  interactions  compared   to  donor-acceptor
interactions.

     dCation exchange sorption phenomena would include replacement
of ionizable,  acidic hydrogens  by cationic  organic species,  but
also polarization  of some ligands by multiply charged, exchanged
cations.

     eSome size exclusion  effects  probably operate here also.

                                6

-------
     BULK SORPTION AND SWELLING

     Swelling  experiments  measure the maximum  capacity  for bulk
sorption of certain solvents that solid organic materials can hold
while in contact with the pure solvent. This limiting capacity can
be expressed in a variety of units,  such  as cm3 sorbed solvent per
cm3 sorbent.   Sorbed material in this form represents a potential
source  of residual  contamination  that  is  difficult to  remove
completely by any known process  (e.g., solvent extraction, vacuum
extraction, biodegradation, etc.).  This  residual contamination is
considered to be of greatest importance for the smaller, more polar
solvent  molecules  such  as  methanol   (See  later  section  for
experimental  evidence on  size exclusion supplied  by  the present
work,  especially Figure VII).    Some solid  macromolecular soil
organic materials  (e.g., cellulose  and chitin)  can themselves be
biodegraded, and would presumably release materials sorbed in bulk
as the macromolecular matrix is destroyed.

     The extractable  organic soil phases are also capable of bulk
sorption (dissolution) of various hydrophobic chemicals. Unlike the
insoluble  solid phases, however,  these materials are potentially
capable  of  mobilization   by  the  right  mixtures  of  nonaqueous
solvents.   The  sorbed species  are  probably also  somewhat more
accessible to remediation measures, especially if they are sorbed
into biodegradable lipid or wax fractions.

SURFACE SORPTION

     True  surface  sorption  can  be  defined  meaningfully  for
insoluble  soil  organic  matter  for  sufficiently  large  organic
molecules  that are effectively excluded  from the  macromolecular
framework of humin and other macromolecules.  This restriction to
surface   interaction  is   probably   very   important   for   many
agrochemicals  such as  insecticides,  fungicides and  herbicides,
which tend to be rather large molecules2.   Note, however, that the
lower molecular weight, microbial metabolites of these chemicals
may distribute themselves over the available sorption categories in
a completely different way than the parent chemical.

SORPTION IN MICROPORES

     The final category of  sorbed material is that which resides in
the  micropores of  the various  solid organic  materials without
causing any  volumetric increase.   No  doubt,  some  size exclusion
effects operate in this situation  as well, but probably not as
stringently  as  in  the bulk sorption  into  macromolecular organic
matter.  The difference between certain gravimetric determinations
of  sorption  and  volumetric  determinations  of  sorption   (via
     2Large in this context means somewhat larger than the aromatic
ring of benzene.

                                7

-------
swelling, see  below)  can indicate the magnitude  of the sorption
capacity in this form.  In many ways, this is the least understood
of  the  categories  in  Diagram  II.    Diffusion  in  and  out  of
micropores also could produce some of the non-equilibrium  effects
seen in column studies  (see Brusseau and Rao, 1989).

OPERATIONAL DEFINITION OF VOLUMETRIC SWELLING

     The volumetric swelling3,  Qv,  is defined as the ratio of the
swollen volume to the unswollen volume of the sorbent; a value  of
unity4 indicates no swelling.   The reciprocal  of the volumetric
swelling is equal to the volume fraction of macromolecular material
in the swollen, solution phase.  Thus,  the volumetric swelling  at
equilibrium is related  to the  saturated  solubility of the sorbed
fluid in the swollen  phase.   Knowledge of appropriate densities,
allows this volume-fraction solubility  to be  converted to the more
usual gravimetric concentration units.

     The   volumetric   swelling5    is   the   quantity  determined
experimentally  in  the  present  work.    A  swelling  spectrum  is
obtained when  the volumetric  swelling  for  various  solvents  is
plotted  versus some  pertinent  solvent  property,  such  as  the
conventionally chosen solubility parameter.
     3The volumetric swelling,  Qv, should not be confused with the
volume change, AVsweUing, associated with the swelling process:

             solvent + solid sorbent -»• swollen sorbent.

The values for Qy are positive values > 1.0  except  in the extremely
unlikely  situation  where  the  AVSHell.   is  negative and exceeds the
imbibed volume of solvent.  Many values of  AVSHell-  are potentially
negative, and they are usually not known because of the difficulty
of determination  (see, however,  RSKSOP-105,  a standard operating
procedure listed later in the  experimental  section, for one method
of  determination) .    Usually  AVSHell.   values  are  some   small
percentage of the volume of imbibed liquid.

     4Apparent swelling values  less  than unity could occur if some
material is extracted into solution by the applied solvent.

      We  note that swelling can also  be determined gravimetrically.
In  these cases,  the  amount   of  sorbed  material is  determined
gravimetrically, and converted to a volume basis using densities.
Usually the gravimetrically determined swelling is larger than the
volumetric quantity because it includes the filling of micropores
without any associated swelling.

-------
THE THEORY OF SWELLING OF MACROMOLECDLAR MATERIALS

SWELLING OF CROSS-LINKED, MACROMOLECULAR SOIL MATERIALS

     There are  two kinds of swelling  corresponding  to the cases
when the  macromolecular material  is cross-linked or  not cross-
linked.  Substances such  as humin and sporopollenin are  believed to
be  cross-linked  materials;  thus  no   "molecules"   of humin  or
sporopollenin exist  to  go  into solution in  any  fluid  applied to
these material.   In  such cases there is a  clear boundary between
phases containing macromolecules and the supernatant fluid, and the
maximum amount of swelling  is ultimately limited by the elasticity
of the macromolecular network.

THE EFFECTS OF IONIZABLE NETWORKS

     Additionally,  humins   have  cation-exchange  characteristics
because  of the  presence of carboxylic  acid  (-COOH)  groups  and
phenolic  (-OH)   groups.   The  protons  on  these  acid  groups  are
frequently  replaced  ("exchanged")  by  alkali  or  alkaline  earth
cations  in natural  materials.    Exchanged  or not,  these groups
represent ionizable groups whose degree of dissociation  will depend
on the  dielectric properties of  the swelling  fluid.  The varying
degree of dissociation will cause the apparent elastic properties
(e.g.,  bulk  modulus)   of  the  matrix  material   to  be  somewhat
different for each solvent applied6.  The hydrogen cations inside
the macromolecular network are  thought by various authors (Flory,
1953,  Tanaka  et al. , 1980)  to behave  somewhat as an  ideal gas,
contributing a linearly temperature-dependent term to the swelling
pressure.

SWELLING OF NON-CROSSLINKED MACROMOLECULES AND BIOPOLYMERS

     Other soil materials such  as humic acid,  cellulose, and chitin
are not covalently cross-linked,  but consist of mixtures of large
individual molecules with a spectrum of molecular weights.  Between
the large, covalently-linked molecular  blocks,  there may, however,
exist   weaker   cross-links   via   hydrogen  bonds,   molecular
entanglements, etc.

     True biopolymers such as cellulose and chitin are,  in addition
to being  composed of regular,   repeating units,  are  also usually
partially crystalline.  The swelling of these materials  is somewhat
like melting because of the  loss of  crystallinity;  the  supernatant
fluid phase  may contain some  smaller  polymer molecules  in true
solution.
           that this  kind of  coupling between elastic  and chemical
interactions   limits  the   accuracy   of   the   usual  separate
consideration given below.

-------
PHYSICAL CHEMISTRY OF SWELLING IN CROSS-LINKED. MACROMOLECULAR
MATERIALS

     The  equilibrium   between  a  liquid   and   a  cross-linked
macromolecular  soil  material  depends  on a  free energy  balance
involving roughly the following terms:

     1.   Strain  Energy  (isotropic  stretching   of  cross-linked
     matrix)

         Parameters  must  be  introduced  here  that  adequately
     describe the elastic properties of the macromolecular matrix
     (e.g.,  the bulk modulus).

     2.  Chemical Interaction Energy (between solvent and matrix)

         The  interaction terms  must   include  a.   non-specific,
     dispersive    (van  der Waals)  interactions  between solvent
     molecules and matrix, b.   site-specific interactions (acid-
     base, etc.) between solvent molecules and certain functional
     groups in the macromolecular structure,  and  c.  electrostatic
     interactions involving attraction between the ionized groups
     of the  matrix as moderated by the  dielectric  properties of the
     solvent molecules.  The dispersive interactions are usually
     described using the so-called  solubility  parameters for the
     fluid and matrix.   The other interactions have been variously
     described; no clear consensus  on  the best methods for their
     formulation exists8.

     3.   Entropy of Mixing  (entropy  increase  due to  mixing of
     solvent molecules with macromolecular matrix)

         This  term is  commonly  handled  with the  Flory-Huggins
     7When the process is conducted under  conditions of constant
volume and temperature, the Helmholtz free energy, A = E - TS, is
the  appropriate  function to  use.   For  the more  usual constant
temperature  and  pressure  conditions,  a volume  change  for the
process must be accounted for,  and the Gibbs free energy function,
G =  E  +  PV - TS ,  is  the appropriate function to  use.   We have
assumed separability, e.g.,  G « Gelastic +  Gchem1cal.

     °We  note that  the Drago C and E parameters used to describe
generalized acid-base interactions between molecules seem to have
some merit for introducing site-specific  interaction energies into
the  equations.      These  parameters   allow   semi-quantitative
calculations  of   enthalpies  of  interaction  which  agree  with
calorimetric  results.   Other  treatments such  as  various donor-
acceptor numbers, or extended solubility parameter schemes do not
seem to be  as well  grounded in real data,  and many times do not
predict even the  correct sign for the enthalpy of interaction.

                                10

-------
     approximation, which simply relates the entropy of mixing to
     the volume fractions of fluid and sorbent.  There is also an
     entropy contribution from the dissociation of ionic groups in
     the imbibed organic liquids.

     4.   Strain Entropy  (entropy  decrease due to  stretching of
     matrix)

         This is  a complex and important  term that depends very
     much  on  the  details  of macromolecular  structure.    In  the
     Flory-Rehner-Huggins theory  for  simple polymers,  additional
     parameters are introduced that describe the concentration of
     cross-links in the matrix and the coordination number around
     certain  bonds   that   rotate   during  elastic  deformation.
     Isotropic  expansion  of the matrix decreases the  degrees of
     freedom of the molecular  network in  the neighborhood of its
     cross-links.

          Semi-crystalline  biopolymers like cellulose  require an
     additional entropy (increase)  term to account for the loss of
     crystallinity in the swelling process.

THERMODYNAMIC MODELS OF SWELLING EQUILIBRIA;  CURRENT STATUS

     In modeling swelling equilibria,  the various terms described
above  are  combined into  a  free energy function.   This  is best
expressed as a  function of  swelling,  Q (or, equivalently,  of the
volume fraction of sorbent  matrix  in  the  swollen  material),  and
the value  of  Q that minimizes this function  is sought.   Various
models of  swelling have been developed for simple  polymers,  and
some of  these have been  applied rather naively to coals;  these
models have been reviewed  separately in supplementary documents on
polymers and coal, and will not  be discussed in  detail here.   At
present, there is no adequate model available to describe swelling
in materials as complex as  humin9.

     The reason for  the  current lack of a  model  is twofold:   1.
soil organic matter being a mixture, requires a different model for
each component of the mixture,  and 2.   even a single component of
the  mixture,   such as  humin,  requires  features  for  which  the
treatment is inadequate.   For any component it  is easy to see that
the model  will  a  priori contain many parameters  describing both
     'Recently,  Chin and Weber, 1989, and  Chin  et al. ,  1990 have
adapted an extended Flory-Huggins  model to  sorption of hydrophobic
organic compounds to dispersed aqueous humic acids with a view to
explaining  some  features  of  enhanced  transport  via  natural
colloidal materials. In this treatment, molecular interactions are
handled with  an extended version of  solubility parameters which
allows separate parameters to  account for   dispersive,  polar and
inductive interactions.

                                11

-------
solvent  and solid  sorbent  that  will  need to  be  specified to
calculate the equilibrium swelling.  Probably only a few of these
parameters can be obtained, even in principle, by curve-fitting the
model to experimental swelling data on real  samples; the remainder
must be obtained from other experiments.

     This  lack  of   an  adequate  model   makes  interpretation of
swelling spectra very difficult, and, moreover, makes prediction of
swelling in any given solvent extremely uncertain.  Therefore, in
the remainder of this report, when swelling data are plotted  as a
spectrum against some  solvent attribute, such as the solubility
parameter,  no simple functional relationship is implied.

                           EXPERIMENTAL

STANDARD OPERATING PROCEDURES DEVELOPED FOR THIS PROJECT10

     RSKSOP-101, OPERATION OF THE LECO WR-112 CARBON ANALYZER FOR
     THE DETERMINATION OF  CARBON  IN PEATS  AND OTHER HIGH CARBON
     MATERIALS, W.G. Lyon,  D.E. Rhodes,  R. Powell, L. Pennington

     RSKSOP-103, MOISTURE DETERMINATION OF PEATS AND SOIL ORGANIC
     MATTER CONCENTRATES, W.G. Lyon, D.E. Rhodes.

     RSKSOP-104, SWELLING SPECTRUM  OF ORGANIC MATTER,  W.G. Lyon,
     D.E. Rhodes  (See Supplementary Document 7)

     RSKSOP-105, PRECISE DENSITY  DETERMINATION  OF  SOIL ORGANIC
     MATTER AND OTHER ORGANIC-RICH MATERIALS,  W.G. Lyon

     RSKSOP-106, COMBINED MOISTURE AND ASH DETERMINATIONS  FOR PEATS
     AND SOIL ORGANIC MATTER  CONCENTRATES, W.G. Lyon, D.E. Rhodes


QUALITY ASSURANCE CONSIDERATIONS

     All research projects making  conclusions  or  recommendations
based on environmentally related  measurements and funded  by the
     10Standard operating procedures developed  for Robert S. Kerr
Environmental Research Laboratory are kept  in a file maintained by
the laboratory.  These supplementary documents can be obtained upon
request from the EPA Quality Assurance Officer at the Laboratory:

                    Jimmie L. Kingery
                    Quality Assurance Officer
                    Robert   S.   Kerr   Environmental   Research
                    Laboratory
                    P.O.  Box 1198
                    Ada,  OK 74821-1198
                                12

-------
Environmental Protection Agency are required to participate in the
Agency Quality Assurance Program.  This project was conducted under
an  approved  Quality Assurance  Project  Plan.    The  procedures
specified in this plan were followed with a  few exceptions.
The  list  of standard  operating  procedures  given  above indicate
those actually developed for the project.  Certain other procedures
which were  to have  been developed were dropped because  of time
constraints on the research program.  Two methods that could not be
developed were a  complementary  gravimetric method  for swelling
determination and application of certain statistical procedures for
un-mixing the spectral data into natural end-members.  RSKSOP-101
for total carbon,  RSKSOP-103 for moisture, and RSKSOP-106 were used
mainly  as  quality  control  checks  on  samples  sent  for  total
elemental  analysis  plus  ash  and  moisture  determinations to  an
independent commercial laboratory.   A standard operating procedure
for density  separation  of  soil organic matter was not specifically
developed for this research although the basic method was used in
processing  the Atoka sample of  pine  duff  as described earlier.
Further information on this method and its limitations is available
in  RSKSOP-107  (Density  Separation  of  Solid Organic Matter  from
Aquifer  Sediments)   that  was  developed  for a separate  research
project involving sandy  aquifer sediments. Many more solvents were
actually used  in the swelling determinations than were originally
thought necessary in the research plan.  Information on the quality
assurance activities and results  is available from the principal
investigator.

DETAILS AND PURPOSE  OF PREPARATIVE STEPS

Air-Drying, Diminution of  Particles and Sieving

     The swelling  experiment requires  samples of  finely powdered
organic  materials,  relatively  concentrated in  their  swelling
components.  Fibrous materials  such as  the Canadian peat cannot be
readily  "ground"  to a  fine  powder,  but  can  be  chopped to  a
sufficiently  small  size with the  blades  in a blender  so  that a
reasonable harvest of -100 mesh (i.e. less than 100 mesh) material
can be obtained by sieving.

Soxhlet Extractions

     Soxhlet  extractions  were  conducted  using  a  three-stage
sequence of solvents consisting of 1-propanol, 1-propanol-toluene
(28% propanol) mixture,  and toluene.   The mixed propanol-toluene
solvent was one of those tabulated  by Fuchsman, 1980  (citing older
work by Reilly)  giving  the greatest yield  of peat  bitumens (see
table on p.  29 of Fuchsman,  1980) .  However, it  should be noted
that this composition is apparently not the binary azeotrope at  1
atmosphere pressure; the azeotrope is closer to a mixture with ca.
50% propanol,  according to Horsley,  1973.    Composition shifts,
therefore,  must have occurred during the course  of this stage in
our soxhlet extraction procedure.

                                13

-------
     The purpose of this pre-swelling extraction procedure was two-
fold:

     1.   to remove as much of the extractable fraction as possible
     so that the various swelling solvents applied to these organic
     materials would not dissolve any further materials from the
     samples,  and

     2.   to remove hydrophobic,  low-melting waxes from the particle
     surfaces so that the samples could  be  dried at 105°C without
     blocking access to the particle interiors for some of the more
     hydrophilic swelling solvents.

     Unfortunately,  some highly swelling solvents such as DMSO and
the  various  amides  dissolved  considerable  amounts  of  humic
materials from some samples during the swelling determinations.  In
several instances these  solvents gave supernatant  solutions over
the  swollen  solid material  that were  as  dark as coffee.   The
extraction of humic substances  by non-aqueous solvents of various
types is discussed in detail by Hayes, 1985.  A pre-extraction step
including DMSO would eliminate  the humic acid dissolution problem;
however, it is very doubtful whether the relatively non-volatile
DMSO could be  entirely  removed from the remaining solids  by any
simple procedure.

     Additionally,  it must  be  noted that a  single vacuum drying
step at 105°C for 24 hours was  not sufficient to remove all traces
of extraction solvents:  the 1-propanol component was still clearly
visible in pre-extracted, vacuum-dried peat  samples subjected to
pyrolysis-GC/MS.

Acid Washing and Cation Exchange

     Acid washing with 0.1  M HC1  solutions was  performed when it
was  desired to have the acid  form  of  the organic materials for
study.  Washing with  0.5 M  CaCl2  solution was  performed when the
calcium-exchanged version of the organic materials was needed; this
was  followed  by water  washing to  a  chloride-free  condition as
determined by tests with aqueous AgNO3 solution.

Vacuum Oven Drying

     Peats which have not been pre-extracted cannot be oven dried
without risking major alteration  of their  wetting (and swelling)
characteristics.  Peat wax and resin components melt in the range
60-70°C  (see Fuchsman, 1980) and  spread on the particle surfaces
rendering  them hydrophobic.   De-waxed  peats are,  however, more
susceptible to oxidation; for the Michigan peat, heating in  air at
105°C caused  samples  of  extracted peat to ignite  and  burn.  The
upper limit of useful drying temperature is bounded by the onset of
pyrolysis during heating under inert gas or vacuum.
                                14

-------
     The above observations from the literature and from experience
led  us  to the  provisional  drying  treatment  for  all  organic
materials of vacuum oven drying at 105 °C for 24 hours  at a few Torr
pressure:  once  just before  loading  the powders in  the  swelling
tubes, and  a  second time after loading in the  tubes.   After the
second drying, the filled tubes were quickly capped with teflon
caps to prevent uptake of moisture from the air.

PROVENANCE  AND PROCESSING OF ORGANIC MATERIALS  USED IN SWELLING
EXPERIMENTS

Michigan Peat

     The original  sample of Michigan peat consisted of a 25 pound
bag of commercial horticultural peat from the Alpar Peat Company of
Ovid,  Michigan  (Clinton  County).    The  dark,  muck peat is  a
predominately reed-sedge peat with some contributions  from tamarack
trees.  The mined deposit is a  layer  approximately 8  to 10 feet in
thickness  in a region once used for farming.

     Further preparation of the as-received sample included coarse-
sieving to  remove  large  sticks, air-drying, sieving to -100 mesh,
soxhlet extraction with 1-propanol, 1-propanol-toluene mixture, and
toluene, vacuum  oven drying at 105°C for  24  hours,  acid washing
with 0.1 M  HC1 solution, and final drying at  105°C for 24 hours.

     A density value of  1.53 g cm'3 was determined for  this material
by RSKSOP-105 using methylcyclohexane.

Canadian Peat

     The original  sample of  Canadian peat  consisted  of  a bale of
commercial  horticultural  peat  with   a  light brown  color  and  a
distinctly  fibric  texture.   Information on  the origin  of this
material was unavailable.

     Further preparation of the as-received sample included coarse
sieving to  remove  sticks,  air-drying, dry  chopping  in a blender,
sieving  to -100  mesh,   soxhlet extraction  with 1-propanol,  1-
propanol-toluene mixture, and toluene, vacuum oven drying at 105°C
for  24  hours,  acid  washing with 0.1 M HC1  solution,  and final
drying at  105°C for 24 hours.

     Besides the acid-washed sample,  described above,  a calcium-
exchanged  version  of this sample was also prepared  by  washing a
portion of  solvent-extracted peat with 0.5 M  CaCl2  solution.

Atoka Pine  Duff

     The Atoka material  consisted of  composted pine needles  (Pinus
echinata)  that had collected in  small  pockets  on a rocky slope
located  in  Atoka  County,  Oklahoma   (S20,  T1S,    R13E).    The

                                15

-------
predominate pine  needle contribution to the  compost was readily
identified from the intact needles in the top layer.  The mineral
constituents were in the form of  fine, windblown, clay-sized dust.


     Preparative steps  for this  material  included coarse sieving
(10 mesh) to  remove large sticks and stones, water washing on  a
sieve  (200  mesh)  to remove  fine dust, density  separation using
carbon  tetrachloride-bromoform  mixtures   (density  2.0  g  cm"3),
chopping in a blender,  sieving to -100 mesh,  soxhlet extraction
with 1-propanol,  1-propanol-toluene  mixture,  and toluene, vacuum
oven drying at 105°C for 24  hours,  acid  washing  with  0.1 M HC1
solution, and final drying at 105°C for 24 hours.

Pine Pollen (Pinus echinata)  & Oak Pollen (Quercus stellata)

     The samples of pine pollen and oak pollen were obtained from
a commercial laboratory  that supplies various species of pollen and
mold spores to allergists11.

     Further preparation of the as-received pollen samples included
soxhlet extraction with 1-propanol,  1-propanol-toluene mixture, and
toluene, and vacuum oven drying at 105°C for 24 hours.

Cellulose

     The sample of  cellulose was obtained  from  Aldrich Chemical
Company, Inc.  (Cat.  31069-7, Lot #07809TV)  and consisted of 1 kg of
powdered cellulose,  nominally 20/i average diameter.   Data from the
supplier showed an average assay  of about 90% a-cellulose based on
acid hydrolysis and  an average residue on ignition ("ash") of about
0.05%.   Cellulose has the stoichiometry (C6H10O5)n (see Kremer and
Tabb, 1990)  leading  to theoretical values of 44.45% C, 6.22% H, and
49.34% 0.

     A density value of 1.57 g cm"3  was determined by RSKSOP-105
using methylcyclohexane.  The production  of static during drying
was a major problem for this density determination.


     The sample was  examined by optical microscopy at lOOOx and was
found to consist primarily of cylindrical  particles with lengths
usually several particle diameters.

     The sample was used for the swelling experiments after final
vacuum oven drying at 105"C for 24 hours.
     11Greer Laboratories Inc.
      P.O. Box 800
      Lenoir,  N.C. 28645
                                16

-------
Chitin fCrab Shell Chitin)

     The  sample of  crab shell  chitin was  obtained  from Sigma
Chemical Company (Catalog C3641,  Lot #107f-7115), and had been de-
mineralized and  purified by the method of Skujins  et  al. , 1965.
The sample was used as-is following vacuum oven drying at 105°C for
24 hours.   Our swelling  studies were  limited  to  25 solvents for
this material  because of the  small supply  of chitin  available.
Chitin has the stoichiometry (C8H13N05)n, with a monomer unit of 2-
acetamido-2-deoxy-B-D-glucose  (See Muzzarelli, 1977).

CHARACTERIZATION  DATA  FOR  ORGANIC  MATERIALS  USED IN  SWELLING
EXPERIMENTS

Elemental Analyses. Moisture and Ash

     Samples   of   all  organic  materials  were  submitted  to  a
commercial analytical  laboratory12  for determination of moisture,
ash,  and the  elements C,  H,  O,  N,  S,  and P.   These data  are
presented in Table I.  The  direct determination of oxygen permits
a    test    for   closure    by    comparison   of   the    sum
(%C+%H+%O+%N+%S+%P+%ASH)  with  100%.   These  values are given as
follows:

     Michigan  Peat, Acid               99.99 ± 0.28

     Canadian  Peat, Acid              100.24 ±0.18

     Canadian  Peat, Ca-Exchanged        99.66 ± 0.23

     Atoka Pine  Duff                    99.01 ± 1.28

     Pine Pollen                       100.40 ±0.32

     Oak Pollen                        100.06 ± 0.22

     Cellulose                         99.17 ± 0.07

     Chitin                             99.03 ± 0.50

The overall standard deviations were estimated as the  square-root
of the sum  of  the squares of the individually estimated standard
deviations of the analyses.  These values seem good considering the
assumed equivalence between "ash" and mineral matter.  The  direct
determination  of organic oxygen is  generally  less reliable, the
larger the ash content of the  sample becomes.
     12Huffman  Laboratories,  Inc.
      4630 Indiana Street
      Golden,  Colorado 80403
                                17

-------
                             TABLE I

           BASIC ANALYTICAL DATA FOR ORGANIC MATERIALS
                     USED IN SWELLING STUDIES
SAMPLE
MP
CP
CPCa
AT
PP
OP
C
CH
C
48.20
0.08
48.96
0.01
49.40
0.10
37.26
0.08
55.21
0.10
56.76
0.15
44.48
0.03
46.82
0.02
H
4.25
0.05
5.26
0.05
5.39
0.06
3.61
0.01
7.51
0.04
7.56
0.06
6.48
0.05
6.59
0.01
0
26.52
0.10
39.50
0.05
38.33
0.19
22.51
0.37
32.59
0.29
24.78
0.06
48.17
0.04
38.60
0.49
N
3.34
0.01
1.01
0.04
0.96
0.02
1.18
0.05
2.17
0.02
6.22
0.08
<0.01
0.01
6.87
0.01
s
0.74
0.05
0.14
0.01
0.12
0.01
0.09
0.01
0.24
0.01
0.40
0.01
0.01
0.01
0.02
0.01
p
0.17
0.07
0.08
0.05
0.07
0.02
0.32
0.01
0.27
0.07
0.68
0.07
0.01
0.01
0.03
0.03
ASH
16.77
0.23
5.29
0.15
5.39
0.03
34.04
1.22
2.40
0.05
3.66
0.09
<0.01
0.01
0.10
0.10
MOIST
13.45
0.22
12.50
0.01
13.33
0.10
5.80
0.09
7.60
0.01
4.00
0.03
5.06
0.03
12.30
0.04
Key: MP=Michigan Peat, Acid; CP=Canadian Peat, Acid; CPCa=Canadian
Peat, Calcium Exchanged, AT=Atoka Pine Duff;  PP=Pine Pollen, OP=Oak
Pollen, C=Cellulose; CH=Chitin.

     The tabulated elemental and ash compositions are presented on
a dry  sample basis  (24  hours at 105°C  in vacuum).   The moisture
contents are presented as a percentage of the original, air-dried
sample mass.  All  values given in bold  represent the average of
duplicate  elemental  determinations.     Uncertainties  are  given
immediately beneath the percentages as one half the difference in
duplicate determinations, or a minimum of 0.01.

     The values  reported were adjusted from  raw values by a factor
obtained from ratioing the standard theoretical values to the value
obtained.  These standards were interspersed with the sample runs
during analyses:  stearic acid, triphenylmethane, and anthracene (C
& H) ; acetanilide  (N); benzoic acid (O);  triphenylphosphine (P) ; 4
certified coal  and oil standards ranging from 0.18 to 3.93% S.
                               18

-------
Pvrolysis-GC/MS Analyses. FTIR Spectra and CEC Data

     All  organic  materials  from  these  swelling  studies  were
submitted for pyrolysis-GC/MS analysis  as part of other projects
involved with characterization of subsurface organic matter.  These
data will not be discussed further here, but are available as part
of the growing analytical information being gradually accumulated
on these known materials.  FTIR spectra in a variety of modes (KBr-
Pellet,  DRIFT,  ATR,  and  PAS)  will  also  be  obtained  on  these
materials.

     Accurate and  informative evaluation of the  cation exchange
capacity  (CEC)  of concentrated  organic materials  is  not easily
accomplished.  The best methods  for CEC  (discriminating carboxylic
acid and  phenolic  hydrogens)  apparently involve a combination of
chemical  and infrared analyses.   There are also serious questions
whether aqueous cations really have access to all  (or the greatest
number) of cation exchange sites in swelling macromolecular media.
Values  of  158  and  124 meq/100  g  (air-dried  basis)   have  been
obtained  for unextracted samples of Michigan  and Canadian peat,
respectively, by  a barium acetate  saturation  procedure (see  the
Peat Testing Manual. Day et al., 1979).

SOLVENTS  USED FOR  SWELLING EXPERIMENTS

Classification of  Solvents

     The  liquids selected for the swelling studies had to fulfill
two basic requirements:  a wide  range of solvent properties, and a
lack of chemical reaction with the organic matrix.   Selection even
within  these constraints  was not  random;  availability,  purity,
price, and  safety  in handling also had to be considered.  Within
these  limitations, we  believe  that we have  used  representative
liquids,  some of which  swell the organic materials to the maximum
limits  set  by  their elastic properties  and  compositions.   The
liquids used for the swelling experiments are classified in Table
II, and certain of their properties are given in Table  III.  They
were  obtained in  the  highest  available purity,  and  were  used
without further purification.
                                19

-------
                             TABLE II

        SOLVENTS  FOR SWELLING EXPERIMENTS:   CLASSIFICATION
                    BY MAJOR FUNCTIONAL GROUPS
Alkanes:
Pentane
Heptane
Cycloalkanes:
Cyclohexane
Methylcyclohexane
           Carbonyl Groups:  >C=O
           Acetone
           Acetylacetone
           Ethyl Acetate
           Propylene Carbonate
           Nitrile:  R-CSN
           Acetonitrile
Nitro-group:
N itromethane
Nitroethane
Nitrobenzene
R-N02
Chloro-group:  R-C1
Chlorobenzene
Dichloromethane

Aromatic Rings:
Benzene
Toluene
p-Xylene
Tetralin

Ether-group: R-o-R1
1,4-Dioxane
[Morpholine]
Amines:
Pyridine
Morpholine

Sulfoxide:  >s=O
Dimethylsulfoxide
sulfide:  S=C=S
Carbon Disulfide

Alcoho1s:  R-OH
(Water)
Methanol
Ethanol
1-Propanol
2-Propanol
1-Butanol
1-Pentanol
3-Methyl-l-Butanol
1-Octanol
1,2-Ethanediol
1,2-Propanediol
           Amides:  -C=O
                      \
                       NH2
           Formamide
           N-Methy1formamide
           N,N-Dimethy1formamide
           N,N-Dimethylacetamide
                                20

-------
                            TABLE III
          SOLVENTS  FOR SWELLING EXPERIMENTS:   PROPERTIES
    Solvent

n-Pentane
n-Heptane
Methylcyclohexane
Cyclohexane
p-Xylene
Toluene
Ethyl Acetate
Benzene
Tetralin
Acetylacetone c
Chlorobenzene
Dichloromethane
Acetone
Carbon Disulfide
1,4-Dioxane
Nitrobenzene
3-Methyl-l-butanol
1-Octanol
Pyridine
Morpholine
N,N-Dimethylacetamide
1-Pentanol
Nitroethane
1-Butanol
2-Propanol
Acetonitrile
1-Propanol
Dimethylsulfoxide
N,N-Dimethylformamide
Nitromethane
Ethanol  (99.9%)
Propylene Carbonated
Methanol
1,2-Ethanediol
1,2-Propanediol
N-Methy1formamide
Formamide
Water
6Q  (MPaV
14.3
15.1
16.0
16.8
18.0
18.2
18.6
18.8
19.4
19.5
19.6
19.8
20.2
20.4
20
20
20
21
21.9
22.1
22.1
22.3
22.7
23.3
23.5
24
   5
   5
  ,5
   1
   1
24.3
24.5
24.8
26.0
26.0
27.2
29.6
29.9
30.7
32.9
39.3
47.9
Molar Volume (cmymol)1

       116.3
       147.6
       127.6
       108.1
       123.3
       106.8
        98.5
        89.4
       137.1
       103.4
       102.6
        64.5
        74.0
        60.6
        85.7
       103.4
       108.9
       158.4
        80.9
        87.5
        93.0
       108.8
        71.9
        91.9
        77.0
        53.0
        75.2
        70.9
        77.0
        53.7
        58.7
        85.9
        40.7
        55.9
        73.7
        58.4
        39.7
        18.1
                                21

-------
                        NOTES TO TABLE III

"The delta values tabulated here are mostly the simple solubility
parameters of the liquids at 25°C tabulated by Barton,  1983;  in a
few  instances where  these  were missing,  the  total  solubility
parameter, «5t, from the same reference was used instead.

bThe values  for  the  molar volume  (cm3/mol)  were  in most  cases
computed from molecular weights (g/mol) and density (g/cm3)  values
for the liquids at 25°C tabulated by Barton,  1983.  A few missing
values  were   computed from  similar data  tabulated  in  standard
handbooks elsewhere (e.g., Weast, 1984)

C2,4-Pentanedione

dl,2-Propanediol  cyclic carbonate

     CH2-0
          \
           C=0
     CH2-0


     CH3
                               22

-------
SWELLING MEASUREMENTS

PROCEDURES

     The  detailed  procedures for  the swelling measurements  are
documented in RSKSOP-104 (see Supplementary Document 7) .  For each
sample of organic material at least a pair of spectra were obtained
at 30°C,  representing  successive measurements  taken  over several
weeks.  From these  data over time,  a "best" spectrum was taken, and
duplicate best  spectra were averaged for each sample.   Figure I
gives an example of paired  swelling spectra for the pine pollen in
which  volumetric   swelling   data   are   plotted   versus  solvent
solubility parameter.  The  multiple points vertically in each plot
represent measurements made at successive times.
                                23

-------
Legend for Figure I.  Pine Pollen Swelling Spectra

     These  paired   spectra  represent  the results  for  swelling
measurements at 30°C  on  duplicate samples of pine  pollen (pinus
echinata)   plotted   versus solvent  solubility  parameter.    The
sequence of  points labeled H2ONH1,  H3ONH1 etc.  denote  swelling
ratios based on heights  (H2/H1 etc.) taken at  successively later
times after  application  of the  solvent;  HI  denotes  the  initial
height of the column of unswollen powder.
                               24

-------
3.3
2.9
z
3 2.5
yj
_i
(0
u
ff 2.1
£
E
J
O 1 T
:> 1.7
3
t)1
SPECTRUM J2, PRE-EXTRACTED, 30 de9 C
-+- H3WH1
i
-
-





}
fl
	 7+ 1 *. /
i ' ' ' ' i ' ' • • i i
1
I
II
I!
'I
II
II f"
\ /
]\ V
• V


" " - ^f





^ "" -f

_l
-
-
-
:
!...-.!....!.. t
14
24                  34
        BELTA, SOUEKT SOLUBILITY IW«ETIR
                                                          44
54
                                   25

-------
     The reason for the choice of  30°C as the standard temperature
was the  avoidance of  kinetically sluggish swelling  observed in
cellulosic materials by Chitumbo et al..  1974, and interpreted by
them  as  phase  transitions.    (See  discussion  of cellulose in
Supplementary Document 5 for further details.)  Temperature control
was achieved  by  immersing  the teflon-capped  sample  tubes  in a
precision (± 0.1"C)  water  bath (Neslab Models RTE-220 and EX-210).
Absolute  temperatures  were   checked  periodically   against  a
calibrated  mercury   in   glass standard   thermometer  (Brooklyn
Thermometer Co., model 21256,  0.1'C div-)

ELECTROSTATIC EFFECTS

     Large electrostatic charges could be produced in the samples
by  vacuum  drying.    When  such samples  were  immersed in  non-
conductive liquids like pentane or  heptane the charges were very
slow to dissipate; the approach of a finger in  a rubber glove would
polarize the particles and move them around much like iron-filings
attracted by a magnet. We believe the presence of a static charge
on  a  sample at  times influenced the uptake  of  certain  organic
liquids.   However,  having no  way to  quantify the amount  of the
charge, we could not establish this effect with any confidence.  It
proved possible to lessen  the  build-up of static charge during the
drying   procedure   and  to    improve   greatly   the   handling
characteristics  of  the  dry  organic powders by  mounting ionizing
radioactive  strips   (Staticmaster  Ionizing Units,  Model  2U500)
inside the heated vacuum desiccator during the drying step.

CALCULATIONS

     For most samples, the averaged raw volumetric  swelling results
are reported as  they were  obtained.   For  cellulose, the  raw
swelling results contained so many values less than unity,  that a
renormalization  factor was applied  to the best averaged swelling
spectrum so  that the averaged,  background value for no-swelling was
then unity.   It is believed that the low apparent swelling values
represented the  effect of lubrication of  solvents in  producing a
slightly denser state of compaction than could be obtained in the
dried  powder cellulose by  standard  centrifugation  alone.   The
cylindrical  shape of the cellulose particles may have contributed
in some way to the apparently lower density of packing in the dry
state.

     For the samples of peat-like materials, a fair comparison of
swelling requires corrections  for ash.  From  the analytical  data
given in Table  I, approximate  ash corrections were calculated as
follows:

          Atoka Pine Duff          C-  r a  1.299*C-   - 0.299
                                    corn          raw

          Michigan Peat            Qcorr *  1.116*Qrau - 0.116

                                26

-------
          Canadian Peat, Acid      <)r « 1.033*QrBU  -  0.033
                                                 raw
                                      r
                                    corr
          Canadian Peat, Ca        Qcorr « 1.034*Qraw  -  0.034

These corrections were based upon the correction formula for inert,
non-swelling  diluents derived in Appendix  II  of RSKSOP-104  (See
Supplementary Document 7).   The conversion from mass fraction to
volume fraction was accomplished by the  formula:


                                            CO.
              Volume Fraction OM, q>* =	
                                      1 - «2  1  -
using approximate densities.  In this expression, the subscript 1
denotes  the  organic  matter, and  2  denotes  inert  diluent;  the
symbols  0,  u, and p  denote,  respectively,  volume fraction, mass
fraction, and density.  Corrected swellings are  given in Table VI,
and are  used  in Figure IIIB.

DUPLICATE SWELLING SPECTRA AND EVALUATION OF EXPERIMENTAL ERRORS

     The evaluation  of  experimental  errors  was  based  on  the
reproducibility  of  swelling values  between duplicate  swelling
spectra.  This  reproducibility was assessed in  two ways.    First,
the  correlation coefficient  was  computed  between  each  pair of
duplicate  spectra.    These  values  are  given  as  the  principal
diagonal elements in  Table IV, and range from 0.95 to 0.99.  Note
that the correlation coefficient should detect similarities between
spectra  independently of scaling  by a constant,  such as might occur
for a sample with more of an inert  diluent  than  some  other sample.
Since all points were  used for these calculations,  the correlations
are diluted by the inclusion of a great many  points with  a low
signal-to-noise  ratio  (low-swelling solvents).    Later  in  the
discussion  we give  a graphical  means of  comparing samples (star
plots) that emphasizes only selected high-swelling solvent values.

     Secondly,  the  absolute  values  of  the differences  between
duplicate spectra for each solvent  were plotted against the average
swelling given by the two spectra.  An example  is given in  Figure
II  for  the pine pollen sample.    This   type  of  plot  detected
departures  (probably  sample handling difficulties,  such as  static
problems)  from random  distribution of  errors  in  two  instances
(Michigan peat and cellulose); fortunately, the largest differences
involved  only  a  few low-swelling  solvents.     In  Table  V  the
statistics  associated with  these  differences  between  duplicate
spectra  are  summarized  for  the  eight  samples.    The  absolute
differences  between  spectra  were  in   all cases  not  normally
distributed; the histograms for  the differences are  highly  skewed
to the left  because  of the effect of the absolute value function  in

                                27

-------
folding the distribution over on itself about zero.13

     The  raw  swelling measurements  are computed  as a  ratio of
lengths measured  to the  nearest tenth of  a millimeter  using a
calibrated measuring reticle and magnifier.   These ratios have a
propagated measurement error  in the second  decimal  digit on the
order of one or two units.  The average deviations given in Table
V are typically larger than this by  a  factor of 3  or 4.  Several
factors probably  account  for  these larger  errors:   occasional
fuzziness  of   the  liquid/powder  boundary,   extraction  of  humic
materials by some solvents, electrostatic effects,  irregularities
in the calibration marks on the glass tubes, and  incomplete wetting
of  the powder  by  the  solvent.   Nevertheless, the large-scale
features of the swelling spectra are clearly visible even with this
higher than expected level of noise.
      Despite the  non-normality of this distribution,  it seemed
more appropriate to  treat the  differences  on an  absolute value
basis because treatment of a swelling value for a given solvent as
if  it  "belonged"  to  one spectrum  or the  other  is really  an
arbitrary decision.  Each swelling measurement with  a given solvent
is really an independent experiment.
                               28

-------
                           TABLE IV

    CORRELATION COEFFICIENTS  (r-VALUES)  BETWEEN  "BEST"  SPECTRA

          1234       56       78
1
2
3
4
5
6
7
8
0.95
0.92
0.97
0.92
0.89
0.84
0.67
0.90
***
0.97
0.93
0.91
0.83
0.85
0.57
0.92
***
***
0.97
0.89
0.88
0.86
0.71
0.95
***
***
***
0.98
0.86
0.74
0.47
0.75
***
***
***
***
0.99
0.77
0.71
0.81
***
***
***
***
***
0.97
0.78
0.91
***
***
***
***
***
***
0.95
0.77
***
***
***
***
***
***
***
0.98
1.  Michigan Peat, Acid Washed
2.  Canadian Peat, Acid Washed
3.  Canadian Peat, Ca-Exchanged
4.  Atoka Pine Duff, Acid Washed
5.  Pine Pollen
6.  Oak Pollen
7.  Cellulose
8   Chitin  (only 25 points)

     In the above table,  the correlations are computed between the
averaged spectra when the spectra are from different materials, but
are computed  between the individual, duplicate  spectra  when the
spectra are from the same  material.   This  is  the reason that the
correlations  along  the main diagonal are  not  identically unity.
Note that  these overall correlations  give equal weight  even to
noisy parts of the  spectra where  the  swelling is  barely above
background. Alternatives are presented later for comparing samples
based on swelling behavior with the most highly swelling solvents
only-
                                29

-------
                             TABLE V


       STATISTICS OF DIFFERENCES BETWEEN DUPLICATE SPECTRA
SAMPLE
Michigan Peat
Canadian Peat,
Acid
Canadian Peat,
Ca-Exchanged
Atoka Pine
Duff
Pine Pollen
Oak Pollen
Cellulose
Chitin
NUMBER OF
POINTS
38
38
38
38
38
38
38
25
MEAN ABSOLUTE
DEVIATION*
0.082
0.044
0.046
0.039
0.045
0.037
0.055
0.064
STANDARD
DEVIATION**
0.097
0.040
0.040
0.035
0.037
0.033
0.063
0.063
 The mean absolute deviation refers  to the mean  of  the absolute
values of the differences between duplicate swelling spectra.

**The standard deviation refers to the standard deviation computed
for the mean of the absolute differences.
                               30

-------
Legend for Figure II.  Absolute Deviation vs. Average Swelling for
Pine Pollen

     Here the absolute value  of  the  difference between duplicate
spectra  is  plotted versus  the  average value  of the  duplicate
swelling  determinations.    This  example  shows  little  sign  of
systematic trends in the distribution of errors.
                                31

-------
ABSOLUTE DEVIATION vs~ AVERAGE SWELLING
     PINE POLLEN  SPECTRA  1 AND 2

6.15
UJ
0
z
LLl
DC
UJ
U. al,
lL U2
N
._
Z
H
JO on
D.D7
J
UJ
Z
(0
U.
Q (U
Oi TO
UJ
D
J

-------
AVERAGED SWELLING DATA SUMMARIZED IN VARIOUS WAYS

     One simple property of the samples which can be derived from
the full data set is the maximum swelling that occurred for any of
the solvents  used.   These values  are  given in Table VI.   These
values   probably  reflect  the  intrinsic  elasticity   of   the
macromolecular matrix as well  as certain other chemical properties
of the samples.

     The complete  swelling  data for all solvents  are  split  into
Tables VII (A & B) ,  VIII  (A & B) , and IX, which correspond to peat-
like materials, pollens, and polysaccharides respectively; swelling
spectra are similarly subdivided in Figures III (A & B) , IV, and V.
In all  instances the best  averaged  swelling values at  30°C are
plotted against the solvent solubility parameter at 25°C.
                                33

-------
                             TABLE VI

  MAXIMUM OBSERVED SWELLING FOR EACH SAMPLE AND THE SOLVENT  FOR
                       WHICH THIS OCCURRED
SAMPLE
Pine Pollen
Michigan Peat
Chitin
Aldrich Cellulose
Oak Pollen
Canadian Peat, Ca
Atoka Pine Duff
Canadian Peat, Acid
RAW
MAXIMUM
SWELLING
2.86
2.46
2.35
1.93
1.89
1.85
1.79
1.78
SOLVENT
FOR
MAXIMUM
SWELLING
DMSO
DMSO
DMSO
DMSO
Formamide
DMSO
DMSO
DMF
ASH-
CORRECTED
SWELLING*
2.89
2.63
2.35
1.93
1.91
1.88
2.03
1.88
SOLVENT KEY:  DMSO = Dimethylsulfoxide, DMF =
N,N-Dimethylformamide.

*Ash corrections  were based on the analytical data given in Table
I,   using  the  formula given  in RSKSOP-104  (Appendix  II),  and an
"ash" density of  2.65 g cm"3.  Values  given in bold denote large
changes after ash correction.
                                34

-------
                            TABLE VIIA

                 SWELLING OF PEAT-LIKE MATERIALS
                    (No Ash Correction Applied)
                              HP
CP
CPCa
AT
n-Pentane
n-Heptane
Methylcyclohexane
Cyclohexane
p-Xylene
Toluene
Ethyl Acetate
Benzene
Tetralin
Acetylacetone
Chlorobenzene
Dichloromethane
Acetone
Carbon Disulfide
1, 4-Dioxane
Nitrobenzene
3-Methyl-l-butanol
1-Octanol
Pyridine
Morpholine
N , N-Dimethylacetamide
1-Pentanol
Nitroethane
1-Butanol
2-Propanol
Acetonitrile
1-Propanol
Dimethy Isul f oxide
N , N-Dimethyl f ormamide
Nitromethane
Ethanol (99.9%)
Propylene Carbonate
Methanol
1,2-Ethanediol
1, 2-Propanediol
N-Methy 1 f ormamide
Formamide
Water
0.98
0.95
1.09
1.06
1.03
1.01
1.11
1.08
1.01
1.18
1.03
1.13
1.24
1.04
1.13
1.02
1.04
1.04
1.65
1.17
1.82
1.09
1.04
1.19
1.25
1.38
1.33
2.46
2.18
1.11
1.47
1.03
1.49
1.76
1.49
2.38
2.29
1.53
0.96
1.07
1.14
1.08
1.11
1.07
1.08
1.08
1.06
1.18
1.11
1.17
1.17
1.11
1.16
1.17
1.13
1.22
1.61
1.19
1.71
1.16
1.18
1.22
1.38
1.21
1.33
1.76
1.78
1.25
1.51
1.14
1.51
1.71
1.63
1.71
1.74
1.57
1.05
1.09
1.10
1.10
1.03
1.05
1.10
1.07
1.09
1.23
1.05
1.05
1.29
1.04
1.08
1.08
1.13
1.12
1.45
1.20
1.51
1.06
1.07
1.07
1.16
1.20
1.17
1.85
1.75
1.11
1.40
1.01
1.43
1.65
1.34
1.78
1.75
1.52
1.08
1.09
1.12
1.16
1.09
1.12
1.23
1.15
1.09
1.32
1.15
1.28
1.33
1.18
1.27
1.19
1.12
1.13
1.69
1.35
1.79
1.18
1.23
1.27
1.34
1.32
1.38
1.80
1.80
1.26
1.38
1.20
1.43
1.52
1.47
1.72
1.65
1.24
Key:  MP=Michigan Peat, CP=Canadian Peat, CPCa=Canadian  Peat,  Ca-
Exchanged, AT=Atoka Pine Duff
                                35

-------
                           TABLE VIIB

                 SWELLING OF PEAT-LIKE MATERIALS
                      (Ash-Corrected Values)
                              HP
CP
CPCa
AT
n-Pentane
n-Heptane
Methyl eye 1 ohexane
Cyclohexane
p-Xylene
Toluene
Ethyl Acetate
Benzene
Tetralin
Acetylacetone
Chlorobenzene
Dichloromethane
Acetone
Carbon Disulfide
1,4-Dioxane
3-Methyl-l-butanol
Nitrobenzene
1-Octanol
Pyridine
Morpholine
N , N-Dimethylacetamide
1-Pentanol
Nitroethane
1-Butanol
2-Propanol
Acetonitrile
1-Propanol
Dime thy Isulf oxide
N , N-Dimethylf ormamide
Nitromethane
Ethanol(99.9%)
Propylene Carbonate
Methanol
1,2-Ethanediol
1 , 2-Propanediol
N-Methy 1 f ormamide
Formamide
Water
0.98
0.94
1.10
1.07
1.04
1.01
1.12
1.08
1.01
1.20
1.03
1.14
1.27
1.05
1.15
1.05
1.02
1.04
1.72
1.19
1.91
1.10
1.04
1.21
1.28
1.42
1.36
2.63
2.31
1.12
1.52
1.03
1.55
1.85
1.54
2.54
2.44
1.59
0.96
1.08
1.15
1.09
1.11
1.07
1.09
1.08
1.06
1.19
1.12
1.18
1.17
1.12
1.16
1.13
1.17
1.22
1.63
1.20
1.74
1.17
1.18
1.23
1.39
1.21
1.34
1.78
1.80
1.26
1.52
1.15
1.53
1.73
1.65
1.74
1.76
1.58
1.05
1.09
1.10
1.10
1.03
1.05
1.10
1.07
1.09
1.23
1.05
1.05
1.29
1.04
1.08
1.13
1.08
1.12
1.46
1.21
1.53
1.06
1.07
1.07
1.17
1.21
1.18
1.87
1.77
1.11
1.41
1.01
1.44
1.67
1.35
1.81
1.77
1.54
1.10
1.11
1.15
1.20
1.11
1.16
1.29
1.19
1.11
1.42
1.19
1.36
1.42
1.23
1.34
1.15
1.24
1.16
1.90
1.45
2.02
1.23
1.29
1.34
1.44
1.42
1.49
2.03
2.03
1.34
1.49
1.25
1.56
1.68
1.60
1.94
1.84
1.31
Key:  MP=Michigan Peat, CP=Canadian Peat, CPCa=Canadian Peat,  Ca-
Exchanged, AT=Atoka Pine Duff.   Ash corrections were computed as
follows:            QMP(ASH FREE) = 1.116*QMP - 0.116
                    QCP(ASH FREE) = 1.033*QCP - 0.033
                    QCPCa(ASH FREE) = 1.034*QCPCa - 0.034
                    QAT (ASH FREE) = 1.299*QAT - 0.299

                                36

-------
                           TABLE VIIIA

                 SWELLING OF PINE AND OAK POLLEN
                    (No Ash Correction Applied)
                                        PP
                         OP
n-Pentane
n-Heptane
Methylcyclohexane
Cyclohexane
p-Xylene
Toluene
Ethyl Acetate
Benzene
Tetralin
Acetylacetone
Chlorobenzene
Dichloromethane
Acetone
Carbon Bisulfide
1, 4-Dioxane
Nitrobenzene
3-Methyl-l-butanol
1-Octanol
Pyridine
Morpholine
N , N-Dimethylacetamide
1-Pentanol
Nitroethane
1-Butanol
2-Propanol
Acetonitrile
1-Propanol
Dimethylsulf oxide
N , N-Dimethyl f ormamide
Nitromethane
Ethanol (99.9%)
Propylene Carbonate
Methanol
1 , 2-Ethanediol
1 , 2-Propanediol
N-Methy 1 f ormamide
Formamide
Water
1.04
1.03
1.02
1.00
0.99
1.05
1.03
1.06
1.03
1.21
1.06
1.19
1.16
0.93
1.22
1.13
1.09
1.09
1.57
1.82
1.93
1.21
1.08
1.23
1.26
1.06
1.29
2.86
1.88
1.02
1.29
1.01
1.49
1.74
1.41
1.99
1.79
1.55
1.00
1.00
0.99
1.02
1.06
1.05
1.06
1.02
1.04
1.21
1.09
1.06
1.12
1.10
1.14
1.17
1.14
1.11
1.27
1.31
1.28
1.13
1.13
1.14
1.19
1.15
1.17
1.61
1.29
1.14
1.27
1.25
1.42
1.66
1.44
1.63
1.89
1.51
KEY:   PP=Pine  Pollen,
stellata
Pinus  echinata,  OP=Oak  Pollen,  Quercus
                                37

-------
                           TABLE VIIIB

                 SWELLING OF  PINE  AND OAK POLLEN
                     (Ash-Corrected Values)
                                        PP
OP
n-Pentane
n-Heptane
Methylcyclohexane
Cyclohexane
p-Xylene
Toluene
Ethyl Acetate
Benzene
Tetralin
Acetylacetone
Chlorobenzene
Dichloromethane
Acetone
Carbon Bisulfide
1,4-Dioxane
3-Methyl-l-butanol
Nitrobenzene
1-Octanol
Pyridine
Morpholine
N , N-Dimethylacetamide
1-Pentanol
Nitroethane
1-Butanol
2-Propanol
Acetonitrile
1-Propanol
Dimethylsulf oxide
N , N-Dimethy 1 f onnamide
Nitromethane
Ethanol(99.9%)
Propylene Carbonate
Methanol
1,2-Ethanediol
1 , 2-Propanediol
N-Methy 1 f ormamide
Formamide
Water
1.04
1.03
1.02
1.00
0.99
1.05
1.03
1.06
1.03
1.21
1.06
1.19
1.16
0.92
1.22
1.09
1.13
1.09
1.57
1.83
1.94
1.21
1.08
1.23
1.27
1.06
1.30
2.89
1.89
1.02
1.30
1.01
1.50
1.75
1.41
2.00
1.80
1.55
1.00
1.00
0.99
1.02
1.06
1.05
1.06
1.02
1.04
1.21
1.09
1.06
1.12
1.10
1.14
1.14
1.17
1.11
1.28
1.32
1.29
1.13
1.13
1.14
1.19
1.15
1.17
1.62
1.30
1.14
1.28
1.26
1.43
1.67
1.45
1.64
1.91
1.52
KEY:   PP=Pine  Pollen,  Pinus  echinata,  OP=Oak  Pollen,  Quercus
stellata.  Ash corrections were computed as follows:
               QPP(ASH FREE)  = 1.014*QPP - 0.014
               QOP(ASH FREE)  = 1.022*QOP - 0.022
                                38

-------
                            TABLE IX
                SWELLING OF CHITIN AND CELLULOSE
                                       CHITIN
CELLULOSE
n-Pentane
n-Heptane
Methylcyclohexane
Cyclohexane
p-Xylene
Toluene
Ethyl Acetate
Benzene
Tetralin
Acetylacetone
Chlorobenzene
Dichloromethane
Acetone
Carbon Bisulfide
1,4-Dioxane
Nitrobenzene
3-Methyl-l-butanol
1-Octanol
Pyridine
Morpholine
N,N-Dimethylacetamide
1-Pentanol
Nitroethane
1-Butanol
2-Propanol
Acetonitrile
1-Propanol
Dimethyl sulf oxide
N , N-Dimethy 1 f ormamide
Nitromethane
Ethanol (99.9%)
Propylene Carbonate
Methanol
1,2-Ethanediol
1 , 2-Propanediol
N-Methy 1 f ormamide
Formamide
Water
**
**
**
**
**
**
1.03
0.99
**
**
**
1.05
1.19
1.00
0.92
**
**
**
1.33
1.28
1.59
**
1.00
1.13
1.27
1.23
1.31
2.35
1.92
1.15
1.71
1.01
1.66
2.11
1.88
2.11
2.24
1.99
1.03
1.06
1.17
0.96
1.01
1.00
1.07
0.99
1.16
1.05
1.00
1.03
1.03
1.03
1.04
0.78
1.01
0.92
0.93
1.61
0.82
0.91
1.00
0.94
1.01
1.08
1.01
1.93
1.03
1.06
1.22
0.89
1.29
1.58
1.24
1.70
1.70
1.56
Value not measured.
                               39

-------
Legend for Figure IIIA.   Raw Swelling vs. £_,  Peat-Like  Materials
     The swelling spectra for Michigan  and  Canadian peat and for
the  Atoka  pine  duff  are compared  in  this  figure before  ash
corrections.   All  materials  were in  their  acid forms  for these
measurements.
                               40

-------
               SUELLIN6 vs  SOLVENT SQLUE1 TTV
                   iLi i i  i mnnuiu'i
    2.5 —
o
z
M
J
J
LU
Z
(0
DC
h
UJ
z
D
J
O
     1.3
                     PEAT-LIKE  MATERIALS WITH ASH
            i  I   I   i     !   I  I   I     I   1   I  I   I   r   i
    8.9
            I	I   I   I  I	I   I  I   I   I	I  1   |   |  I	L   J  I   !
        14
24
34
44
54
                  DELTA,  SOLVENT SOLUBILITY  PARAMETER

-------
Legend for Figure IIIB.  Ash-Corrected Swelling vs. SQ, Peat-Like
Materials

     The swelling spectra for Michigan and  Canadian peat and for
the  Atoka pine  duff  are  compared  in   this  figure  after  ash
corrections.
                               42

-------
     2.7  —
(D

I
M

J
J

LU
2

0)


O
h
U
E
D
J
O
fi
LLJ
h
O
UJ
K
ir
o
o
 i
i
to
2.4 —:
     2.1
1.8
1.5
1.2
                SWELLING vs SOLVENT SOLUBILITY PARAMETER

                   PEAT-LIKE MATERIALS, ASH-CORRECTED
               1	1	1	1	1	:	1	r
                                        MICHIGAN PEAT
     0.9
               i   i
                           '  i   i   i
                                            i   i
         14
                  24
34
44
                                                         J	1	L
                   DELTA, SOLVENT SOLUBILITY PflRflflETER
                                                                  54

-------
Legend for  Figure IV.   Swelling vs.  S0/  Acid  and  Ca-Exchanged
Canadian Peat

     The swelling spectra  of Canadian peat  in its acid  and Ca-
exchanged forms are compared in this plot.  No ash corrections were
applied to these  data.
                               44

-------
0
z
M
J
J
LU
3
0)

o
M
OC
h
UJ
E
D
J
O
      2 —
     1.9
     1.8 -
1.7
1.6
1.5
1.4
1.3
     1.2
     1.1
                SWELLING  vs  SOLVENT  SOLUBILITY  PARAfCTER
                  CANADIAN PEAT,  ACID  AND  CarEXCHANGED
        *$i	£	>;"
                ....!	1	
               4  n
     8.9
            I	I  I   I   I	I   I  I   I   I	I   I   I  I	I  1   I	I	I
         14
                 24
34
44
54
                  DELTA, SOLVENT SOLUBILITY PARAMETER

-------
Legend for Figure V-   Swelling vs.  6Q, Pine and Oak  Pollen

     The  swelling  spectra  of  Pine  Pollen  and  Oak Pollen  are
compared in this plot.  No ash  corrections were  applied to these
data.
                               46

-------
     2.9
     2.5 -
0
I
H
J
J
Ui
3
0)

0
H
IL
h
LU
E
D
J
0
2.1
1.7
     1.3
     8.9
         14
                SWELLING vs SOLVENT  SOLUBILITY PARAMETER
                  PINUS ECHINATA AND QUERCUS  STELLATA
                               _iiiii
            i   t   i
                          i   i  i   i   i _ 1111
                  24
34
44
                  DELTA,  SOLVENT SOLUBILITY PARAMETER
54

-------
Legend for Figure VI.   Swelling vs.  SQ, Cellulose and  Chitin

     This figure compares the swelling spectra  for cellulose and
chitin.   The raw spectrum for cellulose was  scaled by a constant
factor so that the  average  background value of the swelling was
shifted upwards to  1.0.   No corrections  for ash were applied to
either spectrum.
                               48

-------
     2.5  —
                SWELLING vs  SOLVENT  SOLUBILITY  PARAMETER
                           CELLULOSE  AND CHITIN
                           i   i   i   r
                                            CELLULOSE
(D
Z
M
J
J
LU
Z
(0

o
M
n
h
UJ
Z
D
J
O
     2.2  -
1.9
1.6
1.3
                                                  SCALED :StRLIN6
                                          i   i  i   i
         14
                   24
34
44
                   DELTA,  SOLVENT SOLUBILITY PARAMETER
54

-------
                 INTERPRETATIONS AND CONCLUSIONS

GENERAL OBSERVATIONS

     The choice of solvent  solubility parameter as the abscissa is
traditional in  plots  of swelling,  but solubility  parameter,  by
itself, has no special predictive power for the swelling in these
materials.  What is seen instead, are spectra with large jumps in
swelling for very small changes  in  solvent solubility parameter.
These jumps in swelling are largely attributable to site-specific
chemical interactions  of the donor-acceptor type that lower the net
free  energy  of the  swollen  state  for   some  swelling  agents.
Additionally,  there is a molecular size-exclusion effect (discussed
later) that excludes  some larger molecules from  swelling  these
materials even if the donor-acceptor characteristics were correct
for larger swelling.

     The  importance  of site-specific  interactions  in  no  way
diminishes the role  of dispersive interactions in the complete free
energy balance.   A significant  issue  is  the  calculation  of  a
properly corrected solvent  solubility parameter that might be used
to include them.  Fowkes,  1990  has discussed calculations of  5d,  a
modified solubility parameter  describing  dispersive interactions
only; note that this is completely unrelated to the  "<5d" notation
commonly used  for one of the empirical three dimensional solubility
parameters developed by Hansen (see Barton, 1983).

PEAT-LIKE MATERIALS

     The swelling spectra  (Figures  III  A, III B  and  IV)  for the
various  pre-extracted  peat-like  materials  studied  here  seem
interpretable in terms of a mixture of "humic materials," including
both humic  acids  and  humin,  and  of relatively  undegraded  ligno-
cellulosic materials.   The cellulose present in  these peats does
not  appear  to  behave  like  free cellulose  when  exposed to the
solvent morpholine.  The possible explanation for this difference
is that the intimate association of lignin with  cellulose  in the
undegraded plant debris may be sufficient to block the alcoholic-OH
sites on the cellulose chains.

     The  results  for  Canadian  Peat  in  its  acid  and  calcium-
exchanged form (Figure IV)  are  puzzling.   If, as supposed, organic
bases such as  DMSO  interact  strongly with phenolic hydrogens and
with  the  hydrogens associated with  carboxylic acid  groups, why
aren't  there  major  differences  in  swelling  when  these  acidic
hydrogens are  replaced  with  calcium?   We  can only guess at this
point  that  the  small differences  observed  might be  due  to  a
compensating strong coordination of DMSO with the calcium cations
in this  slightly humified peat.   Supporting evidence  for this
explanation could  possibly be obtained from a  functional  group
analysis for this peat, and  a  better  understanding of its cation
exchange capacity.  We note  that the very slight increase in ash

                               50

-------
for the Ca-exchanged sample compared to the acid washed peat does
not seem consistent with  complete  cation exchange to the calcium
form, assuming reasonable values for the cation exchange capacity
(i.e.,  CEC = ca. 124 meq/100 g).

PINE AND OAK POLLEN

     The swelling spectra for pine and oak pollen (Figure V) can be
interpreted in terms of the basic structure and chemistry of pollen
grains.     The  cell   wall  has  an   exine   layer  composed  of
sporopollenin, and an intine layer,  composed of cellulose.  In the
present experiments, solvent-soluble cell material was removed by
soxhlet extraction before swelling determination.

     According to Brooks  and Shaw, 1972 (see especially their Table
I) , the  proportion of the pollen  grain  due  to the  cell  wall is
unusually high in the pine pollens they examined, ranging from 26
to 31% of the  cell  mass.   Of  this  wall fraction of the cell, the
sporopollenin  represents the  larger  portion, and  cellulose the
smaller,  in a  ratio  of  about 3.8 to  1.   From the very  large
swelling observed in DMSO  (2.86), we infer that the sporopollenin
probably has  a very large swelling in this solvent because even
pure free cellulose (1.93) can't account for the magnitude of the
net swelling  observed  in this solvent.  There is,  however,  some
uncertainty concerning the swelling properties of other materials
still  present  in the cell interiors  of  solvent-extracted pollen
grains.

     The  swelling observed  in  DMSO  for  oak pollen  (1.61)  was
significantly  less than  that  for pine pollen  (2.86).   This is in
accordance with the substantially lower wall fraction (7.1%) found
in the single oak pollen  (Quercus robur) included in the tabulation
of Brooks and Shaw,  1972.  The measured proportion of sporopollenin
in the wall was larger  than that  for cellulose  by a ratio of 4.5 to
1.

     The swelling in morpholine,  characteristic of free cellulose,
is present  in both pollens  examined.   The  net swelling  of oak
pollen  (1.31)   and  pine  pollen (1.82)  in morpholine are  both
significant, and in the  order  expected;  however,  the swelling of
pine pollen in this solvent exceeds  the swelling of pure cellulose
(1.61)  in this  solvent,  so that  at least some pollen swelling in
morpholine must be  attributed  to sporopollenin itself  or to some
other non-extractable cell material.

CELLULOSE

     The  swelling  of  cellulose in  various   solvents has  been
investigated previously by various authors; much of this voluminous
literature  from  the paper industry is  reviewed in Supplementary
Document 5.   It is difficult to compare previously obtained results
with the present work since many of the solvents and methods used

                                51

-------
were quite different from those of the present study.  As mentioned
in an earlier section,  the  choice  of  30°C for the swelling studies
was prompted by the unusual behavior of cellulose-solvent systems
reported by  Chitumbo et al.,  1974.   These sensitive temperature
dependent effects and  the  effect  of  different samples (different
degrees of crystallinity,  etc.)  probably account for many seeming
discrepancies in the literature.

     The present results for cellulose (Figure VI) do not suggest
any global maximum in the swelling when it is plotted  as a function
of  solvent  solubility parameter.    Instead,  the  results  are
indicative of site-specific interactions apparently involving the
alcoholic-OH groups on the glucosidic rings.

     There are three hydroxyl groups per  glucose  subunit,  one of
which (C6) is a primary hydroxyl, and two  of which (C2 & C3) are
secondary hydroxyls.   According  to  Kremer  and Tabb,  1990,  the
acidity of the hydroxyls varies in the order C2 > C3 > C6.   Since
DMSO  (an  organic base)  readily  interacts with  slightly  acidic
groups like alcoholic hydroxyls, it probably interacts with these
hydroxyls.  Our swelling results indicate  an uptake of  0.93  cm3 of
DMSO per cm3 of cellulose.   Using the molar volume of DMSO  (70.94
cm3 mol"1)  and the  molar volume of cellulose per glucosidic subunit
(103.27 cm3 mol"1)  allows calculation of the swelling as ca. 1.35
mol of  DMSO per  mol  of glucosidic  subunit.   This  implies that
mainly the C2 hydroxyl is involved in bonding to DMSO, but perhaps
some  C3  is  also  involved.   Possibly,  the  peculiar temperature
dependent  swelling  effects seen  by  Chitumbo et  al. ,  1974  for
cellulose  in DMSO  are related to  the degree  of ionization of
hydroxyls other than C2.

CHITIN

     The swelling of chitin has not been previously investigated.
Austin,  1984  has  considered the best estimate  of the solubility
parameter characteristic of chitin itself; these estimates do not
have any apparent  relationship to the  swelling spectrum observed in
the present work, which does  not  show any clearly defined global
maximum swelling  with respect  to solvent  solubility  parameter.
Instead,  we observe  a pattern  for   chitin  somewhat  similar to
cellulose, but with larger swellings in most cases other than the
solvent morpholine.

MOLECULAR SIZE-EXCLUSION EFFECTS

     The  evidence  for molecular  size-exclusion  is  simple  and
direct.   Figure VII displays  a graph of  swelling values plotted
against solvent molar  volume  for  all materials except cellulose.
The largest solvent still  capable of significantly swelling these
materials is N,N-dimethylacetamide with a  molar volume of ca. 93
cm  mol" .    Many solvents  with  molar volumes less than  this can
swell these  materials.   For  cellulose,  the pattern  is similar

                                52

-------
except that  morpholine with  a molar volume  of ca.  87  cm3  mol"1
represents the largest swelling solvent observed.

     Since volume is only a partial constraint on molecular shape,
it is to  be  expected that the volume boundary  for  swelling  is a
somewhat blurry barrier;  some long rod-like molecules might retain
some swelling ability,  though more globular molecules with similar
donor-acceptor capabilities might not.  The magnitude of molecular
exclusion has  been  partially delimited by the  present  work,  but
these limits  really need to be  challenged with further swelling
studies  using carefully  chosen homologous  series  of  compounds
(e.g., substituted pyridines or sulfoxides).

     The volume exclusion effect introduces an additional term to
the  (strain)  energy portion  of the  free energy expression  for
swelling; one expects the energy to rise rapidly as  the molecules
exceed the size of entry-microporesu.  There may be  possibilities
of  pressure-induced "encapsulation"  phenomena  by  which  higher
pressures could force slightly size-mismatched molecules into the
macromolecular structure of the organic matter.
     uOne  uses this expression  only  for want of  a better term.
Probably,  the  micropores  providing entry into the macromolecular
structure  are  micropores on  a  smaller size  scale  from those
micropores  involved  in capillary  condensation phenomena such  as
discussed  in Diagram II.

                                53

-------
Legend for Figure VII.  Swelling vs. Solvent Molar Volume

     Swelling  for  Michigan  and  Canadian  peats,  Ca-exchanged
Canadian Peat,  Atoka pine duff, pine and oak pollen and chitin are
plotted versus solvent molar volume.   The  data exhibit an abrupt
drop-off   for   solvents   with   volumes   larger   than   N,N-
dimethylacetamide  (ca.  93 cm3 mol"1) .   Above  this  rather fuzzy
boundary,  swellings  tend  to  be  near 1.0  plus  some  background
imprecision.
                               54

-------
     2.9 -
(D
Z
M
J
J
UJ
3
0)

o
h
LU
E
D
J
O
2.1
1.7
     1.3
     8.9
                    AVERAGE SWELLING vs MOLAR  VOLUME
                     ALL MATERIALS EXCEPT CELLULOSE
1 ' '
•
- :
'

" i 0
:
- :


- : .
D
; x


•





i
i ' ' • i • ' ' i ' ' ' i
! • '. '. '.
: 0 : : :
• " • • •
: 00 : : 	 :
i D * *! * i u
x * ^ * ? i - i .
? 1 >*\ " ; I'
• 'A •* • •
0 X + |L : :
: X* jl o : : -
-| t »^
P | 515 ; • : -
x i i*- * « i •
y ^ J_ ' W W
To ^v

: ~ ^ ~X ~ 1 * ? 1 | : -t-:
i * ? x^i ***$ I3! ; * *i
' Di^i :JLY fl?^D:^ ^ D >:
' 0 O'^A TlX'nY X * -
: : -f : :
: - D : o : :
i 1 i i i 1 i i i 1 t i i 1
                                                  128
                                                            168
                       VBflR,  SOLVENT MOLAR  VOLUME

-------
SAMPLE SIMILARITY BASED ON SWELLING

     It was previously mentioned that comparisons of spectra might
be best made with a restriction of  the  comparison to the highest
swelling solvents.   One way of performing this comparison that has
a  good  visual appeal  is the method  of  "star  plots,"  which is
available in the STATGRAPHICS15 software package.   In this method
of plotting, several variables may  be  simultaneously plotted for
each sample.   In Figure VIII swelling values  for  eight highly
swelling solvents are plotted for all eight samples and for a blank
(ideal,  non-swelling  material).   Each  direction  corresponds to
swelling for a particular  solvent (note the key  given in Figure
VIII) that is scaled onto a plotting interval from 1 to 10 in the
following manner: the swelling for  the  blank is plotted as 1 unit
for all  solvents; the  maximum swelling (over all samples) for each
solvent is plotted as  10 units.   The polygonal  patterns for each
sample  allow a  simple visual comparison of  relatedness.    For
example, all the  peat-like samples have a characteristic "notch" in
the morpholine direction corresponding  to the  lower  swelling of
ligno-cellulose compared to free  cellulose.  The two pollens have
rather homogeneous swelling in all  eight  dimensions,  but the oak
pollen makes  a  much smaller polygon.   Cellulose  and  chitin are
similar in the bottom  five dimensions, but differ considerably for
the two amides and morpholine.

     A  full  statistical treatment  of   sample-based  similarities
would involve the methods  of  cluster analysis  and Q-mode factor
analysis.   These  complex  approaches  are  probably not justified for
a set of samples as small as the  present  one; two or three times
the present number of samples  would  be  desirable.
     15STSC, Inc.
      2115 East Jefferson Street
      Rockville, Maryland 20852
                               56

-------
Legend for Figure VIII.   "Star"  Plot of Swelling in Eight Solvents

     This plot exhibits the ash-corrected swelling of all samples
in eight  high-swelling solvents.   The  solvent  key is  given in
Figure IX.  Each direction corresponds to a different solvent; the
blank  indicates  unit  swelling  (i.e.,   no  swelling)   in  each
direction.   The  maximum  swelling  material for  any solvent is
assigned 10 units,  and intermediate values are scaled between 1 and
10.
                                57

-------
    OAK  POLLEN
    CELLULOSE
    CHITIN
CANADIAN PEAT, Ca
ATOKA PINE DUFF
PINE POLLEN
                                    nrAT  n
                                          11
                                        j  n
                           ^Al I ATI T Al I HPAT  I

                                    TLHIj

-------
Legend for Figure IX.  Solvent Key to Star Plot

     This  plot  represents  the  solvent  key  to  the  principal
directions used in the star plot of Figure VIII.
                               59

-------
                   N?N-Dimethylacetamide
      Morpholine
Pyridine
1,2-Ethanediol
N»N-Dimet hy Homamide
      DimethylsuHbxide
Water
                          Methanol

-------
CONSEQUENCES FOR ENVIRONMENTAL STUDIES

     It would be desirable in environmental studies of the  fate and
transport of organic  contaminants  in the subsurface to delineate
all possible categories of sorption onto soil and aquifer materials
in terms of capacity, equilibrium, energetics and kinetics.  This
is far  from being accomplished even  for  the simplest of typical
real systems.

     The  major  application  of   the  present  work   to  real
environmental questions centers on the direct demonstration of an
additional category of sorbed substances that can occur in  soil and
aquifer systems.  Some nonequilibrium partitioning effects seen in
certain  column experiments by  Brusseau  and Rao,  1989 may well
involve a diffusion-limited step  between  swollen organic particles
and the  external fluid phase.   We  wonder,  also about laboratory
biodegradation  studies using small molecules such as methanol as a
carbon  source.   Here we would expect  methanol  to  swell  any soil
organic  matter  that was present, and  thus,  add a  slow diffusion
step  to  the  overall  kinetics  of  the  degradation  process  for
methanol.    In  dilute systems,  this  would be  complicated  by
competitive sorption phenomena involving water also.

     The  bulk-sorbed   fraction   inside   macromolecular  organic
materials probably can serve as a source of hard-to-remove residual
contamination  in a pump and treat  remediation;  however,  not all
organic molecules  can participate in bulk sorption into the solid
organic  matter.   Size  exclusion  seems to  limit the  category to
fairly small,  polar molecules such as some alcohols  and amides.  If
these species were present in  a  contaminating mixture,  they also
might serve as  co-sorbents for  other small  molecules  with less
polar character,  such  as the smaller  chlorinated hydrocarbons.
Most of the alkanes and aromatic species present in fuels are not
expected to undergo significant bulk sorption into macromolecular
organic materials  in soils  and aquifers.

     The present work with its emphasis on macromolecular materials
does  not,  however,  rule  out  the  possible  bulk   sorption  of
hydrophobic organic molecules into the wax-resin-lipid fraction of
soils, nor  does it rule out significant  sorption into micropores
(i.e.,  capillary  condensation).    These   sorption  categories,
extractable  materials  (bitumens)   and microporosity,  represent
potentially important separate areas for  further study, and  may be
responsible for much supposed "partitioning"  of non-ionic organic
contaminants.

SUGGESTIONS FOR FUTURE WORK

     The  present  method  of obtaining swelling measurements is
fairly  labor  intensive,  and,  unfortunately,  rather imprecise.
Recently, various  instrumental methods have become available for

                                61

-------
studying the  particle-size distributions  of powdered  materials
dispersed in liquids.   It would be very worthwhile to explore the
use of these  instruments  for  obtaining swelling data on a given
material  from particle  size  distributions taken  in  different
solvents.   This  should be  relatively simple  to do  for powders
consisting of a  single  substance like cellulose or chitin.   The
swelling of a  heterogeneous mixture of insoluble organic materials
might still be successfully analyzed by such methods,  provided:

     1.    a   series  of  solvents  were used that  affected  the
     components differently,  and

     2.    the size  distribution  functions  for  the  different
     components were of relatively simple analytical types, such as
     lognormal or Gaussian.

     Many features of sorption on  organic  materials  have already
been  pointed  out  as not  well  understood; in  particular,  the
energetics of  site-specific sorption are not yet incorporated into
environmental  models.   The level  at  which  understanding  must
eventually occur has shifted recently from a concern with average
batch  sorption  properties  to  a  concern   with  molecular  level
interactions and  with the behavior of individual components of very
complex  mixtures.    This  level  of  understanding  is  probably
essential for  the modeling of  typical column sorption experiments
to  match the experimental data at  their  full precision.    A
compelling demonstration of adequate understanding of break-through
curves will require  a major increase in the characterization of
materials used in the preparation of columns.

     It became increasingly clear during the course  of  this  work
that the presence of ionizable groups containing oxygen is one of
the key characteristics of soil  and  sediment organic  matter that
will  affect   sorption characteristics.    Therefore,  the various
analytical techniques for determining the disposition of oxygen in
these materials  are of major  interest; among these are FTIR and
pyrolysis-GC/MS  methods.    These  spectroscopic  techniques,  in
combination with  appropriate multivariate statistical methods, are
probably  the   most  appropriate  techniques  available  for  the
development of  rapid  site-characterization tools  for  organic
matter.   Flow  microcalorimetry  or titration  calorimetry  with
various kinds  of probe molecules may provide methods of analyzing
for number  and  energetics of  different  sorption sites  based on
direct measurement of the enthalpies of sorption.

     Besides these newer instrumental methods, classical techniques
for the determination of cation  exchange capacity  (CEC)  of high-
organic materials need re-examination,  if only  to  clarify their
interpretation for  environmental  work.    Just  as  the  apparent
surface area  of  swelling materials is  dependent  on  the  nature of
the molecular probe  (see  Chiou  et al. ,  1990) ,  so should  the
apparent cation  exchange  capacities  of  macromolecular organic

                               62

-------
materials  depend on  the  nature of  the solvent  system;  cations
dissolved in the maximally swelling solvent DMSO should have much
greater access to the  cation exchange sites in these materials than
the  usual  aqueous  cations  (e.g.,  Ba+2, NH4+)  employed  for this
purpose. Unfortunately, highly swelling solvents such as DMSO also
have considerable dissolving power for humic  acids, and would tend
to mobilize these materials; therefore, the usual cation exchange
procedures, if they were adapted to a DMSO solvent system, probably
would have  to use  organic  matter  pre-extracted with DMSO.   The
competition between the cation exchange equilibrium, and the donor-
acceptor interaction  involving  DMSO and acidic hydrogens  is not
understood,  and investigations  here might  also  lead to  useful
analytical procedures.

     Development of CEC measurement  procedures  based on  a better
understanding of swelling materials may have  an improved chance of
measuring a true  total cation exchange capacity (at  least of the
humin  fraction)  that  bears  some  relation  to  bulk  chemical
composition, especially the 0/C ratio.  A statistical examination
of  CEC versus  O/C plots for  many different  kinds of  organic
materials would perhaps be of some use in constructing a sorption
typology for these materials in much the same way as H/C versus O/C
plots (Van Krevelen diagrams) have provided a  simple classification
of  coals  and  kerogens   corresponding  to   their  potential  for
generation  of hydrocarbon gases  and liquids (see,  for  example,
Tissot and Welte, 1984).
                                63

-------
                           REFERENCES

Austin, P.R.,  1984, "Chitin Solvents and Solubility Parameters", in
Chitin. Chitosan.  and Related Enzymes. Academic Press,  Inc., NY, p.
227-237-

Barton, A.F.M., 1983,  CRC Handbook of  Solubility Parameters and
other Cohesion Parameters. CRC Press, Inc., Boca Raton, Florida.

Brooks, J.,  and Shaw, G.,  1968, Chemical Structure of  the Exine of
Pollen Walls and a New Function for Carotenoids in Nature, Nature
V. 219, p. 532-533.

Brooks, J.,  and Shaw,  G.,  1972,  Geochemistry  of Sporopollenin,
Chem. Geol.  v. 10, p.  69-87.

Brusseau,  M.L., and Rao,  P.S.C,  1989, The  Influence  of Sorbate-
Organic Interactions on  Sorption Nonequilibrium,  Chemosphere v.
18(9/10),  p. 1691-1706.

Chin, Y.-P.,  Weber, W.J.,  Jr.,   1989, Estimating the Effects of
Dispersed Organic Polymers on  the  Sorption  of  Contaminants by
Natural Solids.   1.  A Predictive Thermodynamic Humic-Substance-
Organic Solute Interaction Model,  Environ. Sci. Technol. v. 23(8),
p. 978-984.

Chin, Y.-P., Weber, W.J.,  Jr., and  Eadie,  B.J.,  1990, Estimating
the  Effects of Dispersed Organic  Polymers on  the  Sorption of
Contaminants by Natural Solids.   2.   Sorption  in the  Presence of
Humic and other Natural Macromolecules, Environ. Sci.  Technol. v.
24(6), p.  837-842.

Chiou, C.T., Lee,  J.-F., and Boyd, S.A., 1990, The Surface Area of
Soil Organic Matter,  Environ.  Sci. Technol. v.  24, p.  1164-1166.

Chiou, C.T.,  Porter, P.E., and Schmedding,  D.W.,  1983, Partition
Equilibria of Nonionic  Organic Compounds between Soil Organic
Matter and Water,  Environ. Sci.  Technol. v. 17(4), p.  227-231.

Chitumbo,  K. ,  Brown,   W. ,  and  De  Ruvo,  A.,  1974,   Swelling of
Cellulosic Gels, J. Polymer Sci.,  Symposium No. 47, p. 261-268.

Day, J.H., Rennie, P.J., stanek, W.,  and Raymond,  G.P., 1979, Peat
Testing Manual. National  Research  Council of  Canada,  Associate
Committee on Geotechnical Research,  Technical Memorandum No. 125.

Flory,  P.J.,   1953,  Principles   of  Polymer  Chemistry.  Cornell
University Press,  Ithaca,  NY.

Fowkes, F.M., 1980, "Donor-Acceptor  Interactions at  Interfaces",
Polymer Science and Technology, v. 12A, p.  43-52,  Plenum Press, NY.
                                64

-------
Fowkes, F.M.,  1990,  Acid-Base Measurements of Solvents, Polymers
and Inorganic Surfaces, Preprint.

Freeman, D.H., and Cheung, L., 1981, A Gel Partition Model for
Organic Desorption from a Pond Sediment,  Science v. 214, p. 790-
792.

Fuchsman, C.H., 1980, Peat, Academic Press, NY.

Hayes, M.H.B., 1985,  "Extraction of Humic Substances from Soil",
chapter 13 in Humic Substances in Soil.  Sediment,  and Water. Aiken,
G.R., McKnight, D.M.,  Wershaw, R.L.,  and MacCarthy, P., editors,
John Wiley & Sons, NY.

Horsley, L.H., (ed.), 1973, Azeotropic  Data III.  American Chemical
Society, Washington, D.C., p.  202.

Kaiser, W.R.,  1974,  Texas Lignite:  Near Surface and Deep-Basin
Resources, Report of  Investigations No.  79,  Bureau  of Economic
Geology, The University of Texas at Austin, Austin, TX.

Kremer,  R.D.,  and  Tabb, D. ,  1990,  Paper:    The Beneficially
Interactive Support Medium for Diagnostic Test Development, Amer.
Lab. v. 22(3), p. 136,138-140,142-143.

Mingelgrin, U., and Gerstl, F.,  1983, Reevaluation  of
Partitioning as a Mechanism of Nonionic Chemicals Adsorption in
Soils, J. Environ. Qual.  v. 12(1), p. 1-11.

Muzzarelli, R.A., 1977, Chitin.  Pergamon  Press,  NY.

Skujins,  J.J.,   Patgieter,    H.J.,  and   Alexander,   M. ,   1965,
Dissolution of Fungal Cell Walls by a Streptomycete Chitinase and
6- (1-^3) Glucanase, Arch.  Biochem.  Biophys. v.  Ill,  p. 358-364.

Snider,  J.,  and  Covay,   K.J.,  1987,  Premining  Hydrology  of the
Logansport Lignite Area,  DeSoto Parish,  Louisiana, Water Resources
Technical Report No.  41, Louisiana Department of Transportation and
Development, Baton Rouge,  Louisiana.

Tanaka, T., Fillmore, D.,  Sun, S.-T., Nishio,  I., Swislow,  G., and
Shah, A., 1980, Phase Transitions  in Ionic Gels,  Phys. Rev. Lett.
V. 45(20), p.  1636-1639.

Tegelaar, E.W., de Leeuw,  J.W., Derenne, S., and Largeau,  C., 1989,
A Reappraisal of Kerogen Formation, Geochim. Cosmochim. Acta v. 53,
p. 3103-3106.

Tissot,  B.P.,  and  Welte,  D.H.,  1984,  Petroleum  Formation and
Occurrence, Springer-Verlag,  NY.

Weast,  R.C.,  1984,  CRC  Handbook  of  Chemistry  and Physics. CRC

                                65

-------
Press, Inc., Boca Raton, FL, p. F8,
                               66

-------
                                                      SUPPLEMENT  1
                                                      PAGE  1
        DERIVATION OF A SIMPLIFIED THERMODYNAMIC EQUATION
                      OF STATE FOR SWELLING
                          \\\\\\-
                            o       ft


     Consider  the  apparatus depicted in the  above diagram.  The
side  labeled a  denotes  a  sample  of a  solid,  swelling material
confined  by  a  frictionless  piston  producing pressure  Pa,  and
separated by a strong porous barrier from a chamber containing pure
solvent.  The pure solvent side of the apparatus,  denoted by jS, is
maintained at  a  constant  reference pressure,  P. (usually taken as
1 atmosphere or 1 bar) .  The porous barrier allows  solvent, but not
solid, to pass from  one side to the other.  It will be assumed in
all the  following derivations that  no  volume change accompanies
this solvent transfer from one side to another, so that dVa = -dV^.

STATE VARIABLES

     The state variables  to be considered  for this system are the
temperature, T, the applied pressure  P , and the volume occupied by
the swollen material, Vfl.   At  equilibrium,  these are sufficient to
specify the condition of  the system.

     When Pa = P»,  the system  is in a  state of "free swelling", in
which solvent will be sorbed by the material on the a side  of the
apparatus up to some limit set by the cross-linking and the elastic
properties of the material. This free swelling, Qv,  is essentially
what  is measured in the  usual swelling experiment,  in which the
fully  swollen volume,  Va*,  is compared  to the  original   volume
occupied by  the  dry material  (including its  micro-porosity) ,  V0;
_i.e.  Qv = Vtf*/vo'   Tne Pressure/ pa ~  Vft  needed to prevent solvent
from swelling the dry material is  known as  the "swelling pressure"
(for  the dry material)  and can be enormously large  for strongly
swelling systems;  for this reason,  the swelling  pressure  of dry
materials can seldom be measured directly -

WORK DUE TO A CHANGE IN Va

     The work due to a change  in Vc can be calculated as:
 (l)     dW =  -PadVa - PpdVp = -(Pa - Pp)dVa

The swelling pressure is denoted  as follows:
                                   if]

-------
                                                      SUPPLEMENT 1
                                                      PAGE 2
  (2)    n = pa -
ENERGY FUNCTION

     Consider the  internal energy E  = E(Va,T);  the total  energy
differential may be expressed as:
For reversible changes,


 (4)    dqiev = TdS


This allows the substitution:
Since dS is an exact differential,  we may also write the equation,


 ">    —
This allows identification of the terms as  follows


 (7)    ldS\



and

-------
                                                     SUPPLEMENT 1
                                                     PAGE 3
                           E
                           T
Taking cross-partial derivatives, we may equate:


  (9)
This yields the following final equation:
Equation  10  is  a  thermodynamic  equation  of state  completely
homologous to the usual thermodynamic equation of state for fluids.

CONNECTIONS TO CALORIMETRY

     The thermodynamic equation of state for the swelling situation
contains the  differential of  the  internal energy  with respect to
the volume, Va.   This  term can be related  to a set of calorimetric
experiments as  follows.   Assume that a series  of  samples  of the
swelling powder  are prepared with varying  amounts  of pre-sorbed
solvent.  These samples are then immersed in excess solvent at some
constant  temperature,  and  the  heat  evolved  in   each  case  is
measured.  The  free swelling  is known for this material,  so that
the internal energy changes  for each  sample  can  be normalized with
respect to the  final  volume of the swollen  paste.   A plot of the
volume-normalized internal energy  changes  versus the volume of the
material with pre-sorbed solvent gives a  curve  whose slope at any
given volume  is the volumetric energy differential  in question.

TOTAL ENTROPY CHANGE CONNECTED WITH SWELLING

     By substitution of equation 10 into  equation 8, we obtain an
expression for the  differential of the  total system entropy with
respect to volume changes:
     1The pre-sorbed solvent  is  assumed to be homogeneously spread
throughout the swollen sample.

-------
                                                     SUPPLEMENT 1
                                                     PAGE  4
          as
Isothermal integration of this expression from the original volume,
V0,  to  the final  swollen volume,  Va*,  yields the entropy
increment associated with swelling:
 (12)     AS-        dV
                            «
In  actual  modeling  of swelling  data,  this  entropy  is usually
calculated from separate statistical mechanical expressions for the
entropy of mixing,  strain entropy, etc.   Equation  12  allows an
alternative determination of the net effect,  at least  in principal,
from equation of state data along an isotherm.

-------
                                                     SUPPLEMENT 2
                                                     PAGE 1
                             REVIEW;
      PROPERTIES OF  PEATS, AND METHODS OF PREPARATION AND
     CHARACTERIZATION PRIOR TO SORPTION AND SWELLING STUDIES


                         William G. Lyon

               NSI Technology Services Corporation
        Robert  S.  Kerr  Environmental  Research Laboratory
                       Ada, Oklahoma 74820

INTRODUCTION

     Understanding the unique properties of peats and other organic
rich soils  is essential to  understanding  fate  and  transport of
pollutants, and ultimately to designing  remediation  measures for
pollution in wetlands of various  types.  Additionally, there is an
increasing appreciation that  the  sorption characteristics of peats
make them potentially useful materials for various kinds of waste
water treatment processes.

     The essential features of peats  which  affect their sorption
characteristics are as follows:

     1.   The  organic  constituents  of  Peat  are  a mixture of humic
     materials, cellulose,  spores, and various soluble fractions.

     2.   The  acidic depositional  environment and the antiseptic
     nature  of the  water-soluble  phenolic compounds  tends  to
     preserve  more and  different organic  materials than  other
     environments.

     3.   The wax and resin  fraction of peats can  be a fairly large
     portion of the organic mass (up to 10 or 15% sometimes).

     4.   Peat waxes  and resins melt  in the  60  - 70°C range,  and
     spread on the  peat surfaces, blocking  access  to sorption
     sites.  Such samples are difficult to rewet with water.

     5.    Peat  which has  been  extracted  of its wax  and  resin
     fraction is much more  subject to  air oxidation:  it can catch
     fire on drying in air at 105°C.

     6.   Peats tend to have large cation exchange capacities  from
     their  relatively large carboxylic  acid and phenolic  group
     contents.

     7.   Peats  are capable of swelling  in  organic solvents, and
     show size exclusion behavior  for molecules  larger than about
     93  cm3 mol"1.    The cellulosic  components   of  peats exclude

-------
                                                     SUPPLEMENT  2
                                                     PAGE 2
     molecules somewhat smaller than this.
     It is clear  from the  literature surveyed that peats are, in
fact, a very diverse group of materials with considerable variation
in sorption properties depending on their detailed chemical nature
as  shaped  by  source material,   depositional  environment,  and
subsequent diagenetic processes.  Classification schemes for peat
which were designed for fuel or agricultural purposes are clearly
inadequate for the more sophisticated uses of this resource which
are anticipated.

ANNOTATED BIBLIOGRAPHY

Aho, M.J., Tummavuori, J.L.,  Hamalainen,  and Saastamoinen, J.L.,
1989, Determination of Heats of Pyrolysis and Thermal Reactivity of
Peats, Fuel v- 68, p. 1107-1111.

Aho,  M.,  Kortelainen,  P.,  Rantanen,  J.,  and  Linna, V.,  1989,
Pyrolysis of Peat  Studied by Thermogravimetry and Fourier Transform
Infrared Spectroscopy, J. Anal.  Appl.  Pyrolysis  v. 15, p. 297-306.

Aho, M.J., Tummavuori, J.L.,  Hamalainen,  and Saastamoinen, J.L. ,
1989, Determination of Heats of Pyrolysis and Thermal Reactivity of
Peats, Fuel v. 68, p. 1107-1111.

Allen,  S.J.,  McKay,   G.,  and Khader,  K.Y.,  1989,  Intraparticle
Diffusion of  a Basic Dye  During Adsorption  onto  Sphagnum Peat,
Environ. Poll. v.  56, p. 39-50.

Allen, S.J.,  McKay,  G. ,  and Khader,  K.Y.,  1988,  Multi-component
Sorption Isotherms of Basic Dyes onto Peat,  Environ. Poll. v. 52,
p. 39-53.

Allen, S.J., McKay, G.,  and Khader, K.Y., 1988, The Adsorption of
Acid Dye onto Peat from Aqueous Solution - Solid Diffusion Model..
J. Colloid Interface Sci.  v. 126(2),  p. 517-524.

      [Concerned with  the ability of peat to treat textile effluents
     containing  dyes,  and  focuses  on  theoretical  models  of
     sorption.  Not a sufficiently detailed characterization of the
     sphagnum peat used.  Would removal of  the peat wax fraction
     have improved the sorption  characteristics? Did drying at 60°
     spread the bitumens over the particle surfaces?]

Allen, S.J.,  1987, Equilibrium Adsorption Isotherms for Peat, Fuel
v- 66, p. 1169-75.

      [Sorption  isotherms  for  acid  dyes,  basic  dyes,   and  zinc
     cations.]

Andrejko, M.J., Fiene,  F.,  and Cohen,  A.D.,  1982,  Comparison of

-------
                                                     SUPPLEMENT 2
                                                     PAGE 3

Ashing Techniques  for Determination of the  Inorganic  Content of
Peats, in  Testing of  Peats  and Organic Soils.  Proceedings  of a
Symposium  sponsored by  ASTM  Committee  D-18  on Soil  and  Rock,
Toronto,  Canada, 23 June 1982.

      [It is  the mineral matter content  of peats that  is really
     desired, not the "ash" content.  Examines  the difficulties in
     methods borrowed from coal analysis, and the assumptions used
     in converting from ash content to mineral matter content.]

Anonymous,  1990,  Simple Solutions  May Soon Solve  Contamination
Problems, Geotimes v.  35(7), p. 10.

      [A  naive  publicity piece  concerning  the  work of  Cohen and
     Durig at the University  of South  Carolina which claims that
     peats can be used to clean up various kinds of waste water.]

Antworth, C.P., Yates, R.R., and Cooper,  W.T.,  1989, Applications
of   Inverse  Chromatography   in  Organic   Geochemistry  -   I.
Characterization of Polar Solute-Soil Organic Matter Interactions
by High Performance Liquid Chromatography, Org. Geochem. v. 14(2),
p. 157-164.

      [Valuable information on a new technique, but rather skimpy on
     experimental details.   The characterization  of the commercial
     peat used for these experiments is inadequate.  No mention was
     made of the possibility that the cellulose component of peat
     might be sorbing polar materials.]

Arpiainen,  V.,  and  Lappi,  M. , 1989,  Products  from   the  Flash
Pyrolysis of Peat and Pine Bark, J.  Anal. Appl.  Pyrolysis v. 16, p.
355-376.

Bailey, A., and Rosters,  E.G.,  1983, Silicate Minerals in Organic-
Rich Holocene  Deposits in Southern Louisiana,  Proc.  Workshop on
Mineral Matter in Peat:  Its  Occurrence. Form and Distribution. R.
Raymond  and  M.J.  Andrejko,  eds,  Sept. 26-30,  1983,  Los Alamos
National Laboratory, Los Alamos, NM.

Bohlin, E.,  Hamalainen,  M. ,  and Sunden, T. , 1989,  Botanical and
Chemical Characterization of Peat Using Multivariate Methods, Soil
Sci.  V. 147(4), p. 252-263.

     [Use of principal components analysis to separate the 41 peat
     samples  into  8  groups.    Ultimately  one  would  hope  for a
     comprehensive typology of peats and histosols  to arise from
     this sort of chemometric analysis of a  sufficiently large data
     base.]

-------
                                                     SUPPLEMENT 2
                                                     PAGE 4

Bracewell,  J.M.,   Robertson,   G.W.,   and  Williams,  B.L.,  1980,
Pyrolysis-Mass Spectrometry Studies of Humification in a Peat and
a Peaty Podzol, J. Anal. Appl. Pyrolysis v. 2, p. 53-62.

Brown, A.,  Mathur,  S.P.,  Kauri,  T. , and  Kushner, D.J.,  1988,
Measurement and Significance of  Cellulose  in Peat Soils, Can. J.
Soil Sci. v. 68,  p. 681-685

Cloutier,  J.-N.,   Leduy,  A.,  and  Ramalho,  R.W.,  1985,  Peat
Adsorption of Herbicide 2,4-D  from Wastewaters, Can. J. Chem. Eng.
V. 63(2), p. 250-257.

Cohen, A.D., and Spackman,  W., 1977,  Phytogenic Organic Sediments
and Sedimentary Environments  in  the  Everglades-Mangrove Complex
Part II.   The Origin, Description, and Classification of the Peats
of Southern Florida, Palaeontographica, Part B, v.  162(4-6), p. 71-
114.

Cohen, A.D., and Spackman,  W.,  1972, Methods  in Peat Petrology and
their Application  to Reconstruction of Paleoenvironments, GSA Bull.
v. 83, p. 129-142.

Day, J.H., Rennie, P.J., Stanek,  W., and Raymond, G.P., 1979, Peat
Testing Manual. NRC  Canada,  Associate Committee  on Geotechnical
Research, Technical Memorandum No. 125,  193 pp

      [An  extremely useful compendium  of  standard methods  for
     testing organic rich sediments and soils, especially peat.]

Durig, D.T., Esterle, J.S., Dickson,  T.J., and Durig, J.R., 1988,
An Investigation of the  Chemical Variability of Woody Peat by FT-IR
Spectroscopy, Appl. Spec. v.  42(7), p. 1239-1244.

Durig, J.R., Calvert, G.D., and Esterle, J.S., 1989,  Development of
a   Pyrolysis-Gas   Chromatographic-Fourier    Transform   Infrared
Spectroscopic  Technique for  the  Study of Woody  Peats,  J.  Anal.
Appl. Pyrolysis v. 14,  p. 295-308.

Forsberg, S., and  Alden, L., 1989,  Dewatering of Peat, Fuel v. 68,
p. 446-455.

Fuchsman,  C.H., 1980,  PEAT Industrial Chemistry  and Technology.-
Academic Press, NY.

      [A very valuable survey of analytical  results  accumulated over
     the years.  The section (Chapter 3)  on solvent extraction of
     peat  bitumens gives comparative yields  for  various solvent
     systems.  Table  XIII gives softening points and various other
     properties of peat bitumens.]

-------
                                                     SUPPLEMENT 2
                                                     PAGE 5

Gafni, A.,  and Brooks, K.N.,  1990,  Hydraulic Characteristics of
Four Peatlands in Minnesota, Can. J. Soil Sci. V- 70, p. 239-253.

     [Contains an  excellent set  of references  on  the hydraulic
     properties of peat deposits.]

Hatcher, P.G., Breger,  I.A., Maciel,  G.E.,  and Szeverenyi,  N.M.,
1985, Geochemistry  of Humin,  chapter  11 in  Humic  Substances in
Soil.   Sediment   and   Water:	Geochemistry,   Isolation.   and
Characterization. John Wiley & Sons, NY, pp.  275-302.

Hatcher, P.G., Lerch, H.E., III, Kotra, R.K., and Verheyen,  T.V.,
1988,  Pyrolysis  G.C.-M.S.  of  a Series  of  Degraded  Woods  and
Coalified Logs that  Increase in Rank  from  Peat to Subbituminous
Coal, Fuel v. 67, p. 1069-1075.

Holmgren, A., and Norden, 1988, Characterization of Peat  Samples by
Diffuse Reflectance FT-IR Spectroscopy, Appl. Spect. v. 42(2), p.
255-262.

Joshi,  H.C.,  and   Misra,  S.G.,   1983,   Adsorption   of  Three
Organophosphorus Insecticides on Different Soils,  Environ. & Ecol.
v- 1, p. 75-81.

     [One peat soil sample with a TOM of 20.7% was included.  The
     three pesticides investigated were malathion, parathion, and
     phosphamidon.]

Katase, T., and Kondo, R., 1989, Vertical Profiles of trans-
and  cis-4-Hydroxycinnamic  Acids  and   Other   Phenolic  Acids  in
Horonobe Peat Soils, Japan, Soil Sci. v- 148(4), p. 258-264.

Kornder, S.C., and Carpenter, J.R., 1984, Application of a Linear
Unmixing Algorithm to the Normal Alkane Patterns  from Recent Salt
Marsh Sediments,  Org. Geochem. v. 7(1), p.  61-71.

Rosters,  B.C.,  and  Bailey, A.,  1953,  Characteristics  of  Peat
Deposits in the Mississippi  River Delta Plain, Trans.  Gulf Coast
Assoc. Geol. Soc. v. 33, p. 311-325.

Kumari,  D.,   1987,  Analysis  of  Wax and  Resin  Components  from
Minnesota Peat Bog, Int. J. Coal Geology v. 8, p. 99-109.

Leger, S., Chornet, E., and Overend, R.P.,  1987,  Characterization
and  Quantification  of  Changes  Occurring   in  the  Low-Severity
Dewatering of Peat, Int. J. Coal Geology v- 8, p. 135-146.

Loxham, M. , 1981, Pollution  in Peats,  Proc.  10th Int.  Conf. Soil
Mech. Found.  Eng.  Stockholm,  Sweden, June  15-19,  1981, v.  2, p.
345-348, A.A. Balkema, Publisher, Rotterdam,  Neth. and  Salem, NH.

-------
                                                     SUPPLEMENT  2
                                                     PAGE 6

MacCarthy,  P.,  and  Djebbar,  K.E.,  1986,  Removal  of Paraquat,
Diquat, and Amitrole from Aqueous Solution by Chemically Modified
Peat, J. Eriviron.  Qual. v- 15(2), p. 103-107.

McCarthy, T.S., Mclver,  J.R.,  Caimcross, B.,  Ellery, W.N.,  and
Ellery,  K.,  1989,  The  Inorganic  Chemistry  of  Peat from   the
Maunachira Channel-Swamp System, Okavango Delta,  Botswana, Geochim.
Cosmochim. Acta v. 53, p. 1077-1089.

Martin,  A.M.,  and  Manu-Tawiah,  W.,  1989,   Study  on the  Acid
Hydrolysis  of Peat:    Composition  of the Extracts from Sphagnum
Peat, J. Chem. Tech. Biotechnol. v. 45, p. 171-179.

Moers,  M.E.C.,  Boon,   J.J.,  de Leeuw, J.W.,  Baas,  M.,  and P.A.
Schenck, 1989,  Carbohydrate  Speciation and Py-MS Mapping of Peat
Samples  from  a  Subtropical  Open  Marsh  Environment,  Geochim.
Cosmochim. Acta v. 53, p. 2011-2021.

Moore, P.O., 1989, The  Ecology of Peat-Forming Processes: A Review,
Int. J. Coal Geol. v.   12, p. 89-103.

Orem, W.H.,  and Hatcher, P.G.,  1987,  Early Diagenesis of Organic
Matter  in a Sawgrass  Peat from the Everglades,  Florida,  Int. J.
Coal Geology v. 8, p.   33-54.

Preston, C.M.,  Axelson, D.E.,  Levesque,  M.,  Mathur, S.P., Dinel,
H. ,   and   Dudley,   R.L.,   1989,   Carbon-13   NMR   and  Chemical
Characterization of Particle-Size Separates of Peats Differing in
Degree of Decomposition,  Org. Geochem. v. 14(4), p.  393-403.

Roy, C., Chornet, E.,  and Fuchsman,  C.H.,  1983, The Pyrolysis of
Peat, A Comprehensive Review  of  the Literature,  J.  Anal.  Appl.
Pyrolysis v. 5, p. 261-332.

Ryan, N.J.,  Given,  P.H.,  Boon, J.J.,  and de Leeuw,  J.W.,  1987.
Study  of the  Fate of Plant  Polymers  in Peats  by Curie-Point
Pyrolysis, Int. J. Coal Geol. v. 8, p. 85-98.

M. Schnitzer, M.,  and Levesque, M., 1979, Electron Spin Resonance
as a  Guide  to the Degree of Humification of  Peats,  Soil  Sci. v.
127(3), p. 140-145.

      [A partially successful attempt to apply ESR spectroscopy of
     free radicals to the estimation of degree of humification of
     peats.]

Sheppard, S.C., Gibb, C.L.,  and  Hawkins,  J.L.,  1989,  Fate of
Contaminants  during  Utilization of  Peat Materials,  J.  Environ.
Qual. v. 18, p. 503-506.

-------
                                                     SUPPLEMENT 2
                                                     PAGE 7

Stevenson, F.J.,  and Butler, J.H.A., 1969, Chemistry of Humic Acids
and Related Pigments, Chapter 22 in Organic Geochemistry
edited by G. Eglinton and M.T.J. Murphy, Springer-Verlag, NY.

Ting,   F.T.C.   1977,   Microscopical   Investigation   of   the
Transformation (Diagenesis) from Peat to Lignite, J. Microscopy v.
109(1),  p. 75-83.

      [Estimates approximately a 4:1 compaction of peat passing to
     the lignite stage of coalification.]

van  Smeerdijk,  D.G., and  Boon,  J.J.,  1987,  Characterization of
Subfossil Sphagnum Leaves,  Rootlets of Ericaceae and their Peat by
Pyrolysis-High Resolution Gas Chromatography-Mass Spectrometry, J.
Anal. Appl. Pyrolysis v. 11, p. 377-402.

Viraraghavan, T. ,  and Mathavan, G.N. ,  1988,  Use of Peat  in the
Treatment  of Oil-In-Water Emulsions,  Proceedings of  the  42nd
Industrial Waste Conference, Purdue University, Lafayette, Indiana,
May  12-14, 1987.
                                    77

-------
                                                     SUPPLEMENT  3
                                                     PAGE 1

                             REVIEW;

                THERMODYNAMICS  OF POLYMER SWELLING
              AS AN ANALOGY FOR SOIL ORGANIC MATTER

                         William G. Lyon
               NSI Technology Services Corporation
         Robert S.  Kerr Environmental  Research Laboratory
                       Ada, Oklahoma 74820

INTRODUCTION

     The thermodynamics of polymers in contact with orgarfic liquids
might  seem  remote from the  sorption  properties  of  soil organic
matter (SOM);  nevertheless, the analogy has been successfully drawn
between the sorption behavior of coal and of cross-linked polymers.
With some  significant modifications,  several  of  the theoretical
constructs for polymers are expected to  apply  also to soil organic
matter.

SWELLING

     For historical reasons, polymeric systems were studied at an
early stage in  the development  of physical chemistry,  because it
was known that solutions  of large molecules  departed  from ideal
behavior at extremely small molar  concentrations.   Cross-linked
polymeric systems, which swell rather than dissolve, were studied
during this period, and it is  these systems  especially which are
analogous to certain components of soil organic matter.  Swelling
is  simply  defined  as the ratio  of  the   swollen  volume  (at
equilibrium) to the unswollen volume of the polymer.

THE LIMITS OF THE POLYMER-SOM ANALOGY

     The limitations of the Polymer-SOM  analogy are useful to keep
in mind throughout the following discussion.   Soil organic matter
is  a  mixture  of  components of  different origin,  some  of  them
soluble, and  some  insoluble.    The  cross-linked macromolecular
components are of primary  concern here,  so we  will assume that all
soluble  components have been  removed when  swelling in  a  given
liquid is being considered.  In addition to being a mixture, soil
organic matter is polymeric,  in the  strict sense of the term, only
for certain  constituents   such  as cellulose  and chitin.   Other
components such as humin  or humic acid  are macromolecular rather
than polymeric:   there is  no  regularly repeating unit  in these
materials.   Another manner in which macromolecular soil components
differ from synthetic polymers is in their cross-linkages:  these
are not  regular in soil organic matter either, but  consist of  a
hodge-podge of  covalent cross-links,  hydrogen-bond cross-links,
molecular entanglements, and perhaps even  pi-bonded cross-links.

-------
                                                     SUPPLEMENT 3
                                                     PAGE 2
SOME RELATIONS BETWEEN SOM AND COAL
     The intrinsic  irregularity  of  soil  organic matter is shared
with coal;  however,   soil organic matter,  typically,  is  a much more
complex mixture of terrestrially derived macromolecular materials
than is coal.   The  diagenetic processes  occurring in the initial
coalification  of plant  debris  removes  many  sorptive materials
(e.g..   cellulose)    which   can   be  readily   metabolized   by
microorganisms, leaving a  fairly inert, and simpler collection of
bio-resistant molecules as lignite.  As coal passes from lignite to
sub-bituminous  rank,  surface  functionalities,  like  -COOH groups,
responsible  for cation exchange  properties are  lost, sometimes
abruptly,  through  abiotic diagenesis coming into play under the
increasing severity  of surrounding conditions.

PHYSICAL CHEMISTRY OF CROSS-LINKED, POLYMER SWELLING

     The equilibrium between  a liquid and a cross-linked polymer
depends on a free  energy  balance involving roughly  the following
terms:

     1.    Strain Energy   (isotropic  stretching  of  cross-linked
     matrix)

     2.  Chemical Interaction  Energy  (between solvent  and matrix)

     3.   Entropy of Mixing  (entropy  increase due  to mixing of
     solvent with macromolecular matrix)

     4.  Strain Entropy (entropy decrease due to  stretching)

     The above subdivision is described as a rough one, because it
is not clear experimentally whether the free energy can always be
legitimately subdivided into elastic and  chemical  components.  The
assumption of separability of  the free energy into  these components
was introduced by Flory and Rehner, 1943.

     It may  be  significant that the known apparent disagreements
with the  Flory-Rehner assumption,  appear to show  up  in systems
involving a polymer  in equilibrium with a vapor.  As discussed by
Vink, 1983, vapor phase equilibria with homogeneous gels seem to be
inadmissable  thermodynamically,  because  of  the  formation  of
heterogeneities  in  the  gels  (i.e.  the  assumption of  a  single
homogeneous  gel  phase used in certain thermodynamic  derivations
breaks down).   At  present,  the Flory-Rehner  assumption remains
largely untested for Polymer-Liquid systems (binary or ternary).
For the moment we will not worry about this unresolved dilemma, but
merely note  that it  is  currently  an  item of  active  debate and
research even  for  such  well-studied  polymers  as  rubber.    (See
especially the papers by Neuburger and Eichinger,  1988, and  by Gee
et al., 1965).

-------
                                                     SUPPLEMENT 3
                                                     PAGE 3

CAVEATS CONCERNING APPLICATIONS TO COMPLEX SUBSTANCES

     The polymer literature (especially on rubber) may be somewhat
misleading on what  to  expect for the  relative  importance of the
four  free  energy  terms,  in  applications  involving  naturally
occurring macromolecular substances.   In general,  the energetic
components  of  the  free  energy balance  are thought to  be less
important for the swelling of  rubbery  polymers  than the entropic
contributions;  this is not necessarily the case  for swelling in
other types of substances such as cellulose etc.  Therefore, models
appropriate  to   these  more  complex  substances  should not,  in
general, either  ignore  (elastic strain  energy) or oversimplify the
energetic  effects  (chemical  interaction energy).   This  caveat
applies  especially  to  the uncritical  application  of  unmodified
Flory-Huggins-Rehner theory to humic materials as has been reported
by  Chin  and  Weber, 1989.   The  parameters derived   from such
treatments may be of dubious validity.

ELASTIC STRAIN ENERGY

     The strain energy  stored in  a stretched  spring  (i.e.  the  iskx2
term  for the potential energy)  has a similar, but generalized,
formulation (see Kittel,  1966)  for isotropically expanded swollen
polymers:  the  strain energy  density, U,  is equated  to a quadratic
expression involving the bulk modulus,  B, and the volume dilation,
d,  i.e.  U = ^Bd2.   The volume  dilation,  d,  is  defined  as  the
increase in volume,  divided by the original volume,  V0, and can be
easily related to the  swelling,  Q.  Departures from Hooke's law
behavior are expected to become serious when swelling greater than
about 1.3 is observed.  In  this region  of large strain, the theory
might be adjusted either  by allowing  B = B(d),  or by including
higher order terms in the expansion of the strain energy.

     It should be noted that calorimetric determinations of heats
of  immersion of  a cross-linked,  macromolecular material in  a
swelling solvent determine a net heat effect, which is the sum of
strain  energy  and  chemical  interaction energy terms.   Careful
calorimetric  work  on  model   systems  with  simple  chemical
interactions would allow a direct determination of the importance
of a strain energy term in the swelling equilibrium.

CHEMICAL INTERACTION ENERGY AND SOLVENT PARAMETER (6)

     When  two components  are  admixed  on a molecular  scale,  the
energy change involved  represents a difference in the final  state,
component 1 and  2 molecules  interacting with each other, and the
initial  state,  component  1  and  two  molecules interacting with
themselves.   This  can be  expressed  in  a  number  of  ways,  the
simplest being in terms of differences in solvent parameters for
the  liquid  and   the  polymeric  matrix.  Such   a  description is
sufficient only in instances for which the molecular interactions

-------
                                                     SUPPLEMENT 3
                                                     PAGE 4

are very non-specific and dominated by dispersive contributions.

SOLVENT PARAMETERS FOR LIQUIDS

     The solvent parameter  for  a  liquid can be simply calculated
from its energy of vaporization and its molar volume; it represents
the square-root of a cohesive energy density, and is found useful
in  approximating  non-specific, van der  Waals-type  interactions
between  molecules.    Empirical  generalizations  of  the  solvent
parameter concept break the overall parameter into  components which
describe, for example,  the dispersive, polar, and hydrogen-bonding
interactions  between  molecules.     These  parameters  have  been
conveniently tabulated for many substances by Barton, 1983.

SOLVENT PARAMETERS FOR MACROMOLECULAR MATERIALS

     Solvent parameters for macromolecular or polymeric substances
are defined and determined quite differently than  those for simple
liquids; Huglin and Pass, 1968 have applied many of these diverse
methods of determination to the  polymer, polytetrahydrofuran.  One
of the simplest methods involves identifying the solvent parameter
of the solvent which gives the largest swelling, with the solvent
parameter of the macromolecular matrix.  This method seems to work
fairly  well  for  simple  polymers,   but  is  misleading,  in  many
instances, when applied  naively to heterogeneous  mixed materials
like coal  or soil  organic  matter which  also  have site-specific
interactions with many polar molecules.  It is almost certainly
true that no single measure of cohesive energy density is entirely
adequate for describing one of these complex materials, or even an
isolated pure component of one of these materials.

INCLUSION OF SITE-SPECIFIC INTERACTIONS

     A modification of the simple Flory-Huggins-Rehner picture is
needed to handle the large chemical interaction energies associated
with  site-specific,  generalized   acid-base interactions.    The
quantitative empirical,  "E-C" formulation  suggested by  Drago et
al. . 1971,  and applied by  Fowkes  et al. .  1984 to polymers, has
great promise as a means of correcting the equations for swelling
of cross-linked polymers.  The additional  parameters added by this
treatment would be  the appropriate  E and C parameters describing
the liquid molecules (many of these are already known  from the work
of  Drago  et  al. ,  1971) ,  another  set  of  E  and C  parameters
describing  the polymer  interaction  sites,  and  a concentration
parameter,  defining how many  moles  of  interaction  sites are
available per unit mass of unswollen polymer.

VOLUME OF MIXING

     It should  be mentioned  that systems  for  which  there  is a
significant volume of mixing,  require a "P delta V" term  to adjust

-------
                                                     SUPPLEMENT 3
                                                     PAGE 5

for the difference between energy and enthalpy.  Volume of mixing
in swelling systems  can be determined for  powdered materials by
careful pycnometric  determinations  in swelling  and non-swelling
liquids.

ENTROPY OF MIXING

     Admixture  of substances  also  produces  an  entropy  effect.
Fortunately, this can largely  be  calculated as a configurational
entropy of  mixing term  of the sort recommended  independently by
Flory and Huggins for molecules of greatly dissimilar size.  Their
amazingly simple  reformulation of the mixing  entropy in terms of
volume fractions  rather than mole fractions has  proved extremely
useful whenever molecules of dissimilar sizes are involved.

     Modifications of the simple Flory-Huggins entropy are needed
when some of the  mixing is  not entirely  random,  as for instance,
when strong hydrogen-bonded interactions  exist between some liquid
molecules and some sites on the macromolecular matrix.  The usual
calculation of mixing entropy is expected to overestimate the true
mixing entropy in such cases.

ELASTIC ENTROPY

     When  a  polymer is   stretched,  the  molecular  chains  are
generally   forced to   occupy   positions  with   fewer  possible
configurations; hence, the entropy decreases in such a process, and
opposes the entropy  increase due to mixing.   The calculation of
such strain entropy requires a  model of the polymer  structure.

TERNARY AND MULTICOMPONENT SYSTEMS

     Thus far, we have  considered systems  consisting of a single
liquid and  an  insoluble,  cross-linked polymeric  component.   The
next level  of  complexity  is a  ternary system  with two, miscible
liquids  and a  single  cross-linked polymer.    In  general,  such
systems are more complex,  because of the possibility of the mixed
liquids changing their relative proportions as they pass from the
supernatant liquid phase  to the  swollen  polymer  phase.  In other
words,  various  additional  phenomena  involving  selective  and
competitive sorption are possible in such systems.

     The phenomenology of simple  ternary systems involving mixed
solvents and  polymers  has  been  studied sporadically,  since the
exemplary early experimental work of Bronsted and Volqvartz, 1940.
Until quite recently, such work  has not proven  popular,  perhaps
because of the lack of a firm theoretical foundation for analyzing
the results.  In principle, the phase behavior cap be mapped out by
examining both swelling and changes in composition in a series of
mixtures applied to a polymer specimen, as was done  in the classic
work of Bronsted and Volqvartz, 1940.

-------
                                                     SUPPLEMENT 3
                                                     PAGE 6
ANNOTATED BIBLIOGRAPHY
Barton, A.P.M., 1983, Handbook of Solubility Parameters and Other
Cohesion Parameters. CRC Press, Inc., Boca Raton, FL.

Bastide, J., Picot, C.,and Candau, S., 1981, Some Comments on the
Swelling of Polymeric Networks in Relation to their Structure, J-
Macromol. Sci.-Phys., v. B19(l), p.  13-34.

Brandup,J., and Immergut,  E.H.,eds,  1975,  Polymer Handbook,  John
Wiley & Sons, NY.

Bronsted, J.N., and Volqvartz, K. , 1940, Solubility and Swelling of
High Polymers  in  Ternary Mixtures,  Trans.  Faraday Soc.  v. 36, p.
619-624.

     [One  of  the  first, and simplest  papers  on ternary mixtures
     with  a  swelling polymer as  one constituent.   The  value of
     triangular diagrams for concisely and completely summarizing
     the  isothermal  experimental  phase  information about  such
     systems is shown here.]

Brown, H.R., 1978, Flory-Huggins-Rehner Theory and the Swelling of
Semicrystalline  Polymers  by Organic  Fluids,  J. Polymer  Sci.,
Polymer Physics ed., v. 16, p. 1887-1889.

Campos, A., Gavara, R.,  Tejero, R., Gomez, C.,  and Celda, B., 1989,
A  Flory-Huggins  Thermodynamic  Approach for  Predicting  Sorption
Equilibrium in  Ternary  Polymer  Systems,  J.  Polymer Sci:   Part B:
Polymer Physics,  v. 27, p. 1569-1597.

Campos, A., Gavara, R.,  Tejero, R., Gomez, C.,  and Celda, B., 1989,
A Procedure for Predicting Sorption Equilibrium  in Ternary Polymer
Systems from  Flory-Huggins Binary  Interaction Parameters and the
Inversion Point of Preferential Solvation, J. Polymer Sci:  Part B:
Polymer Physics,  v. 27, p. 1599-1610.

Candau, S., Bastide, J., Delsanti, M., 1982, Structural, Elastic,
and Dynamic Properties  of  Swollen  Polymer Networks,  Adv. Polymer
Sci. v. 44, p. 27-71.

Chin, Y.-P-, and Weber,  W.J.,  Jr.,  1989,  Estimating the Effects of
Dispersed  Organic  Polymers  on the  Sorption  of  Contaminants by
Natural Solids.   1.  A  Predictive Thermodynamic Humic Substance  -
Organic Solute Interaction Model,  Environ.  Sci.  Technol. v. 23(8),
p. 978-984.

Cowie,  J.M. G., Dey, R., and McCrindle, J.T., 1971, A Comparative
Study of Preferential  Adsorption in Bromoform (1) , Benzene (2), and
Polymer  (3)  Systems  by  Light Scattering  and  Density  Gradient
Ultracentrifugation, Polymer J. v. 2(1), p. 88-93.

-------
                                                     SUPPLEMENT 3
                                                     PAGE 7

Cowie, J.M.G., and Bywater,  S.,  1966,  Preferential Adsorption in
the Ternary Systems Solvent(1), Nonsolvent(2), Polystyrene(3), J.
Macromolecular Chem.  v. 1(3), p. 581-594.

DiPaola-Baranyi,  G. ,  and Guillet, J.E.,  1978,  Estimation of Polymer
Solubility Parameters  by  Gas  Chromatography, Macromolecules v.
11(1), p. 228-235.

Drago, R.S.,  O'Bryan, N., and Vogel, G.C., 1970, A Frequency Shift-
Enthalpy  Correlation for  a  Given  Donor with Various Hydrogen-
Bonding Acids, J. Am. Chem. Soc. v- 92, p. 3924-3929.

Drago, R.S.,  Vogel, G.C., and Needham, T.E., 1971, A Four-Parameter
Equation  for Predicting  Enthalpies of Adduct Formation,  J.  Am.
Chem. Soc. v. 93, p.  6014-6026.

Drago, R.S.,  Parr,  L.B.,  and  Chamberlain,  C.S.,  1977,  Solvent
Effects  and  Their Relationship to  the E and C Equation,  J.  Am.
Chem. Soc. v. 99, p.  3203-3209.

Ewart, R.H.,  Roe, C.P., Debye,  P., and McCartney,  J.R., 1946, The
Determination of  Polymeric Molecular Weights by Light Scattering in
Solvent-Precipitant Systems,  J. Chem. Phys. v. 14(11),  p. 687-695.

Fowkes, F.M., McCarthy, D.C., and Tischler, D.O., 1983, Predicting
Enthalpies of Interfacial  Bonding  of  Polymers  to  Reinforcing
Pigments, in  Molecular Characterization of  Composite  Interfaces.
edited by H.  Ishida and G.  Kumar, Plenum Press,  NY.

Fowkes, F.M., Tischler, D.O., Wolfe, J.A., Lannigan, L.A., Ademu-
John,  C.M.,  and  Halliwell,  M.J.,  1984,  Acid-Base  Complexes of
Polymers, J.   Polymer Sci.,  Polymer Chemistry Edition, v.  22, p.
547-566.

Flory, P,J., 1970,  Thermodynamics of Polymer Solutions,  Disc.
Faraday Soc.  v. 49, p. 7-29.

Flory,  P.J.,  1953,   Principles  of Polymer Chemistry.  Cornell
University Press, Ithaca, NY.

Flory,  P.J.,  1961,   Thermodynamic Relations for   High  Elastic
Materials, Trans. Faraday Soc.  v. 57,  p. 829-838.

Flory, P.J.,  1950,  Statistical Mechanics of Swelling of Network
Structures, J. Chem.  Phys.  v.  18, p. 108-111.

Flory, P.J.,   and  Rehner, J., Jr., 1944, Effect  of Deformation on
the Swelling Capacity of  Rubber, J.  Chem.  Phys. v.  12(10), p.  412-
414.

Flory, P.J.,   and  Rehner, J., Jr., 1943, Statistical Mechanics of

-------
                                                     SUPPLEMENT 3
                                                     PAGE 8

Cross-Linked Polymer Networks  II.  Swelling, J. Chem. Physics v.
11(11), p. 521-526.

Flory, P.J., and  Rehner,  J.,  Jr.,  1943,  Statistical Mechanics of
Cross-Linked Polymer Networks I.  Rubberlike Elasticity, J. Chem.
Phys. v. 11(11), p. 512-520.

Flory, P.J., 1942, Constitution of Three-Dimensional Polymers and
the Theory of Gelation, J. Phys. Chem. v. 46, p. 132-140.

Flory, P.J.,  1942,  Thermodynamics of High  Polymer Solutions, J.
Chem. Phys. v. 10, p. 51-61.

Flory, P.J.,  1941,  Thermodynamics of High  Polymer Solutions, J.
Chem. Phys. v. 9, p. 660-661.

Gee,  G. ,  1946,  The  Thermodynamic  Study of  Rubber Solutions and
Gels, Adv. Colloid Sci. v.  II, p. 145-195.

Gee,  G.,  1946,  The Interaction between  Rubber  and Liquids.   IX.
The  Elastic  Behavior of Dry  and Swollen Rubbers,  Trans. Faraday
Soc.  v. 42, p. 585-598.

Gee,  G. ,  1944,  VI.   Swelling  and  Solubility  in  Mixed Liquids,
Trans. Faraday Soc. v. 40, p. 468-480.

Gee,  G., 1944, The Interaction between Rubber and  Liquids V.  The
Osmotic Pressures  of  Polymer  Solutions  in Mixed Solvents, Trans.
Faraday Soc. v. 40, p. 463-468.

Gee, G. ,  1943,  Interaction between Rubber and Liquids  IV.  Factors
Governing the Absorption of Oil by Rubber, Trans. I.R.I.,  v.  18, p.
266-281.

Gee, G. ,  1942,  IV.  The Micellar Theory of the Structure of Rubber,
Trans. Faraday Soc. v. 38, 109-115.

Gee,  G., 1942, The  Interaction  between  Rubber and Liquids.   III.
The  Swelling of  Vulcanised  Rubber  in  Various  Liquids,  Trans.
Faraday Soc. v. 38, p. 418-422.

Gee,  G. , 1942,  The Interaction between  Rubber  and Liquids.  II.
The Thermodynamical Basis  of the Swelling and Solution of Rubber,
Trans. Faraday Soc. v. 38, p. 276-284.

Gee, G.,  and Treloar, L.R.G.,  1942,  The  Interaction between Rubber
and Liquids.   1.   A Thermodynamical  Study  of the System Rubber-
Benzene,  Trans.  Faraday Soc. v. 38, p. 147-165.

Gee, G.,  1940,  II.  Osmotic Pressure and  Viscosity of  Solutions  of
Raw Rubber, Trans. Faraday Soc. v. 36, p. 1171-1178.

-------
                                                     SUPPLEMENT  3
                                                     PAGE 9

Gee,  G.,   1940,  The  Molecular  Weights  of  Rubber and  Related
Materials.  I.  Experimental Methods, Trans. Faraday Soc. v.  36, p.
1162-1171.

Gee,  G. ,  Herbert,  J.B.M.,  and  Roberts,  R.C.,  1965, The  Vapor
Pressure of a Swollen Crosslinked Elastomer,  Polymer v. 6, p. 541-
548.

Harland, R.S., Klier,  J.,  and Peppas,  N.A., 1990, Thermodynamic
Models for  Swelling of Heterogeneous Networks,  J. Appl.  Polymer
Sci. V. 41, p. 249-265.

Hearst, J.E., and Vinograd,  J., 1961, A Three Component Theory of
Sedimentation Equilibrium in a Density Gradient, Proc. Nat. Acad.
Sci. V. 47, p. 999-1004.

Heil, J.F., and Prausnitz, J.M.,  1966, Phase  Equilibria in Polymer
Solutions, A.I. Ch. E. Jour. v. 12(4), p. 678-685.

Higgs,  P.G.,  and  Ball,  R.C.,  1989, Some Ideas  Concerning  the
Elasticity of Biopolymer Networks, Macromolecules v. 22,  p. 2432-
2437.

Horkay, F., and Zrinyi, M., 1983, Comparative Study of Polymer Gels
Based   on   Theories   of   Critical  Phenomena   and   Classical
Thermodynamics, Acta Chim.  Hungarica v.  114(3-4), p. 261-281.

Huggins, M.L.,  1942,  Some  Properties of  Solutions of Long-Chain
Compounds, J. Phys. Chem. v- 46, p. 151-158.

Huggins, M.L., 1941,  Solutions  of Long  Chain Compounds,  J. Chem.
Phys. v. 9, p. 440.

Huglin, M.B.,  and Pass, D.J.,  1968, Cohesive  Energy  Density of
Polytetrahydrofuran, J. Appl. Polymer Sci. v. 12, p. 473-485.

Kirkwood, J.G., and Goldberg,  R.J.,  1950,  Light  Scattering Arising
from Composition Fluctuations in Multi-Component Systems, J. Chem.
Phys. v. 18(1), p. 54-57.

Kittel, C.,  1966,  Introduction to Solid  State  Physics.  3rd Ed.,
John Wiley & Sons, NY, p. 118.

Koningsveld, R. ,  1989,  Thermodynamics of Macromolecular Systems,
Pure & Appl. Chem. v. 61(6), p. 1051-1064.

Kovac, J.,  1978,  Modified  Gaussian  Model for Rubber Elasticity,
Macromolecules v. 11(2), p.  362-365.

Krigbaum, W.R.,  and Carpenter,  O.K.,  1954,  Phase Equilibria  in
Polymer-Liquid 1-Liquid 2 Systems,  J. Polymer Sci. v- 4, p. 241-

-------
                                                     SUPPLEMENT 3
                                                     PAGE 10
259.
McKenna, G.B., Flynn, K.M., and Chen, Y.-H., 1988, Mechanical and
Swelling Behavior of Crosslinked Natural Rubber:  Consequences of
the Flory-Rehner Hypothesis, Polymer Commun. v. 29, p. 272-275.

McKenna, G.B.,  and  Zapas, L.J., 1983,  Experiments  on the Small-
Strain  Behavior of  Crosslinked Natural  Rubber:    1.   Torsion,
Polymer v. 24(11), p. 1495-1501.

McKenna, G.B.,  and  Zapas, L.J., 1983,  Experiments  on the Small-
Strain Behavior of Crosslinked Natural Rubber:  2.  Extension and
Compression, Polymer v. 24(11), p. 1502-1506.

Masegosa, R.M., Prolongo,  M.G., and Horta, A.,  1986, g Interaction
Parameter of Polymer-Solvent Systems, Macromolecules v. 19, 1478-
1486.

Merk,  W. ,   Lichtenthaler,   R.N.,  and  Prausnitz,   J.M.,  1980,
Solubilities  of  Fifteen   Solvents  in  Copolymers  of  Poly(vinyl
acetate) and Poly(vinyl chloride) from Gas-Liquid Chromatography.
Estimation of Polymer Solubility Parameters,  J.  Phys. Chem. v. 84,
p. 1694-1698.

Nagy, M. ,  Some  New Aspects of Research  on  Polymer Gels,  Colloid
Polymer Sci. v. 263(3), p. 245-265.

Nagy,  M. ,   Wolfram,   E.  ,  and  Gyorfi-Szemerei,   A.,   1972,  A
Refractometric  and  Gel Sorption  Study of the  System Poly(Vinyl
Alcohol)-Water-n-Propanol, J. Polymer Sci.:  Part C,  v.  39, p.  169-
177.

Neuburger, N.A.,  and Eichinger,  B.E.,  1987,  The Status of Flory-
Rehner Theory, Amer.  Chem. Soc.,  Div. Polymeric  Materials Preprint
v- 56, p. 6-10.

Neuburger, N.A., and Eichinger, B.E., 1988, Critical Experimental
Test of the Flory-Rehner Theory of  Swelling, Macromolecules v. 21,
p. 3060-3070.

Nozari, M.S.,  and Drago,  R.S.,  1970,   Spectral and Calorimetric
Studies of Hydrogen Bonding with Pyrrole,  J.  Am. Chem.  Soc. v. 92,
p. 7086-7090.

Okita, K., Teramoto,  A., Kawahara,  K.,  and Fujita, H.,  1968, Light
Scattering and  Refractometry of a Mondisperse Polymer in Binary
Mixed Solvents, J. Phys.  Chem. v. 72, p. 278-285.

Pouchly,  J.,  1989,   Sorption  Equilibria  in  Ternary   Systems:
Polymer/Mixed Solvent,  Pure and Appl. Chem. v. 61(6), p. 1085-1095.

-------
                                                     SUPPLEMENT 3
                                                     PAGE 11

Pouchly,  J.,  and  Zivny,  A.,  1982,  Correlation  of  Data  on
Preferential Sorption  Using the Modified Flory-Huggins Equation,
Makromol. Chem. v. 183, p. 3019-3040.

Pouchly, J., 1969, Specific Interactions  in Solutions of Polymers.
I.  General  Formulation  of the Flory-Huggins Lattice Statistics,
Collect. Czech. Chem. Commun. v. 34, p. 1236-1253.

Scott,  R.L.,  and Magat,  M.,  1945,  The  Thermodynamics  of High-
Polymer Solutions:  I.  The Free Energy of Mixing of Solvents and
Polymers of Heterogeneous Distribution, J. Chem. Phys. v. 13(3), p.
172-177.

Scott, R.L., 1943, The Thermodynamics of High-Polymer Solutions:
II.  The Solubility and Fractionation of a Polymer of Heterogeneous
Distribution, J. Chem. Phys. v. 13(3), 178-187.

Scott, R.L., 1949, The Thermodynamics of High Polymer Solutions.
IV-   Phase  Equilibria in the  Ternary  System:   Polymer-Liquid 1-
Liquid 2, J. Chem. Phys.  v. 17(3), p. 268-284.

Shultz, A.R.,  and Flory.-  P.J. ,  1955,  Polymer Chain Dimensions in
Mixed-Solvent Media,  J. Polymer Sci. v. 15, p. 231-242.

Starkweather, H.W., 1977, The Sorption of Chemicals by
Perfluorocarbon Polymers, Macromolecules v. 10(5), p. 1161-1163.
Stockmayer,  W.H.,  1950,  Light  Scattering  om  Multi-Component
Systems, J. Chem. Phys. v. 18(1), p. 58-61.

Strazielle, C., and Benoit, H., 1961, Experimental Study of Light
Scattering in the System  Polystyrene-Benzene-Cyclohexane, J. Chim.
Phys. v- 58, p. 678-681.

Tanaka, T., 1981, Gels, Sci. Amer. v. 244(1), p. 124-138.

Treloar, L.R.G., 1975, The Physics of Rubber Elasticity. Clarendon
Press, Oxford.

Tuzar,  Z.,  and Bohdanecky,  M. ,  1969, The Properties  of Diluted
Solutions  of  Poly(Ethylene  Glycol  Monomethacrylate),  Collect.
Czech. Chem. Commun.  v- 34, p. 289-294.

Tuzar, Z., and Kratochvil, P.  1967,  Light Scattering. XVIII.  The
Behavior  of Polymers  in Mixed  Solvents,  Collect.  Czech.  Chem.
Commun. 32, p.  3358-3370.

Vetere, A.,  1987,  A Simple Modification of the Flory-Huggins Theory
for Polymers in Non-Polar or Slightly Polar Solvents, Fluid Phase
Equilibria v.  34(1),  21-35.

Vetere, A., 1988, Erratum, Fluid Phase Equilibria v. 39, p. 225.

-------
                                                     SUPPLEMENT 3
                                                     PAGE 12

Vink, H. ,  1983, Thermodynamics of Swelling and Partition Equilibria
in Gels, Acta Chem. Scand. v. A37, p. 187-191.

von  Tapavicza,  S.,  and Prausnitz, J.M.,  1976,  Thermodynamics of
Polymer Solutions:  An Introduction, Int. Chem. Eng. v- 16(2), p.
329-340.

Wall, F.T.,  1943,  Statistical  Lengths of Rubber-Like Hydrocarbon
Molecules, J. Chem. Phys. v. 11, p. 67-71.

Wall, F.T.,  1944,  Osmotic Pressures  for Mixed Solvents, J. Amer.
Chem. Soc. v. 66, p. 446-449.

Zivny, A., Pouchly, J.,  and Sole,  K.,  1967, The Preferential and
the   Overall   Sorption   in   a   Macromolecular   Coil   System:
Polymethylmethacrylate-Benzene-Methanol,  Collect.  Czech.  Chem.
Commun. v. 32, p. 2753-2765.

-------
                                                      SUPPLEMENT 4
                                                      PAGE  1
                             REVIEW;
     THE SWELLING OF COAL AND OF KEROGEN.  AND ITS RELATION TO
 THE SWELLING AND SORPTION CHARACTERISTICS OF SOIL ORGANIC MATTER

                         William G. Lyon
               NSI Technology Services Corporation
         Robert S. Kerr Environmental  Research Laboratory
                       Ada, Oklahoma 74820
INTRODUCTION:

     The  swelling properties of  coal  provide  a  useful analogy,
intermediate  in  complexity  between  soil  organic  matter  and
synthetic polymers.  Many experimental techniques and theoretical
constructs  applicable to  coal,  might have  direct  and obvious
applications to soil organic matter.  It is often overlooked that
coal itself also has some intrinsic  interest from an environmental
point  of  view:    many  shallow lignite deposits occur within the
highly industrialized, coastal plains of the Southern and Eastern
United States.   In many instances these lignite  deposits are in
contact with, or are part of aquifer systems; hence, their sorptive
characteristics are of interest in modeling ground water  systems in
these regions (see Kaiser, 1974, and Snider and Covay,  1987).

     Soil organic matter, typically, is a much more complex mixture
of  terrestrially  derived macromolecular materials than is coal.
The diagenetic processes occurring in the initial coalification of
plant debris remove many sorptive  materials (e.g.. cellulose) that
can  be readily metabolized  by  microorganisms, leaving a fairly
inert,  and  simpler  collection  of  bio-resistant  molecules  as
lignite.   As coal  passes from  lignite to sub-bituminous rank,
surface functionalities,  like -COOH groups, responsible  for cation
exchange properties are lost, sometimes abruptly,  through abiotic
diagenesis operating under the increasing severity of surrounding
conditions.

     Coal is most assuredly  not a simple material;  rather, it is
itself a complex mixture  of cross-linked, macromolecular materials
and sorbed molecules of varying origin.  Coal, like soil organic
matter, is microporous  in its structure, a fact  that makes sorption
and swelling phenomena much  more  difficult to  interpret from the
experimental  data than  those obtained  for  synthetic  polymeric
materials.   Also,  coal  lacks  a  repeating  polymeric  unit  like
synthetic polymers; therefore, many formulas  describing extensive
polymer properties in units "...  per monomer formula unit" are
not so simply applicable to coals.
     Kerogen, the  insoluble  organic matter found  in sedimentary
rocks,  differs  from typical coals  in  several  aspects,  the most
important being the occurrence  of more hydrogen-rich components,

-------
                                                     SUPPLEMENT 4
                                                     PAGE 2

usually  of marine  origin.   The  full range  of  kerogens  does,
however,  span the  full  range  in  organic  matter  content,  from
exclusively terrestrial to exclusively marine in origin.  The bound
lipid material in kerogen,  especially from marine algal sources, is
known to be the optimal source material for generation of petroleum
deposits (see Durand, 1980).   Relatively little has been published
on  kerogens  concerning  their  swelling or  other macromolecular
properties; the sole example to date is the work  of Shadle et al. .
1989, who  studied the swelling  of kerogens  isolated from Western
and Eastern oil shales.

COAL STRUCTURE AND SWELLING IN SOLVENTS

     The known facts of coal structure that are of greatest concern
in  the fundamental  description of  processes such as  swelling,
sorption,  and extraction are as follows:

     1.    Coal contains  a  cross-linked,  macromolecular  matrix
     material.

     2.   The  cross-links  are of  at least  two,  distinct  kinds,
     covalent and hydrogen-bonds.

     3.  Some hydrogen-bonding solvents are capable of disrupting
     the hydrogen-bond cross-links of coal,  causing a very large
     swelling of the matrix structure.

     4.  Coal also contains somewhat mobile, extractable molecules
     of  a  wide  range  of  composition,  that   are  sorbed    or
     "clathrated" by the macromolecular matrix.

     5.  The coal  matrix material is microporous,  and is capable of
     sorbing chemical vapors by capillary condensation as well as
     by swelling-sorption.

     6.  The "surfaces" of the coal matrix that  are accessible to
     solvents contain functional groups capable  of specific, non-
     bonded  interactions   (i.e.. not  covalent bonds)  of  various
     types with appropriate solvents.

Later  sections  will elaborate  on  these various  aspects  of coal
structure  and their  impact on swelling, sorption and extraction.
Many excellent reviews of coal structure  in general  have been
published;  several  better  ones  are  listed  in the  Annotated
Bibliography-

VOLUMETRIC VERSUS GRAVIMETRIC SWELLING

     Experimentally, coal swelling has been determined by two major
methods:   the volumetric  method,  and the  gravimetric method.   In
the former, coal  is  put  in direct contact with  a liquid,  and the

-------
                                                     SUPPLEMENT 4
                                                     PAGE 3

change  in the  volume of  the coal  is measured  directly.   The
swelling, Qv, determined  by this method  is simply the  ratio of
swollen to unswollen volume.

     The usual  gravimetric  method  involves equilibration of coal
with  saturated  vapors  of  the  chosen   liquid,  and  successive
weighings to determine when equilibrium is achieved.  The limited
rate of mass transport via  the vapor phase makes  this a slow and
tedious procedure,  although it can be automated  with continuous
weighing  on a  specially  modified  electrobalance.   The  masses
involved can be converted to an equivalent volumetric swelling, QH,
provided the densities of the matrix and sorbed liquid are known.

     A much faster gravimetric method (apparently not yet used for
coal) would involve putting coal powder  in direct contact with a
liquid  for  some pre-determined period, and  centrifuging  off the
excess liquid through a special centrifuge tube (or an extraction
thimble with a  fritted bottom) under standard conditions.  Here the
liquid  retention  mass is an  indication  of swelling, and  can be
converted to an equivalent volumetric swelling.

     The  fact  that,   generally,  the  gravimetric  method  gives
swelling considerably in  excess of the volumetric measurements for
materials like coal has been attributed to the microporous nature
of coals by Nelson et al.. 1980.   That is, a major portion of the
weight  gain observed  in  the gravimetric  method,  is  capillary
condensation of solvent vapors into the micropores of coal.  This
seems to subdivide sorbed solvent into two categories:  capillary
condensed solvent,  and solvent  imbibed  into  the  macromolecular
structure.  Presumably, the  factors governing the equilibria in the
two  cases  are  different,  and  the  macroscopic  thermodynamic
description of the  system would have  to  take both processes into
account  to  calculate the  overall  vapor  sorption  equilibrium
isotherms.

     An alternate  view is that the micropore  system constitutes
merely a "dead" volume, VQ,  which must be filled before the matrix
starts to stretch.  In the  case  of  equilibrium by direct contact
with a liquid,  there would be no real distinction between the two
categories of sorbed liquid, and the thermodynamic treatment would
merely have to  recognize that the dead volume can be filled with
liquid  without producing  any strain  energy  or  strain  entropy
effects (see discussion under Physical Chemistry of Coal Swelling)
provided the molecules of the liquid can fit into the dead volume.
Some dependence of VQ on  the molecular dimensions  of the swelling
liquid relative to the micropore size distribution is thus expected.

COAL, SOLVENTS, AND SWELLING:  AN HISTORICAL PERSPECTIVE

     During the classical period of coal chemistry,  much work was
expended searching for "coal solvents" that would dissolve all the

-------
                                                     SUPPLEMENT 4
                                                     PAGE 4

organic materials contained in coal  (leaving the mineral matter),
and  thus  produce a  liquid  feedstock that could be  more readily
handled in the refineries and chemical plants designed for liquid
materials like petroleum.   This  work culminated in the extensive
investigations of Dryden, and  of Van Krevelen in the fifties and
early sixties which pretty much ruled out the  existence of such
magic solvents,  at least among the commonly known organic liquids.
There thus appeared to be definite limits to the  solubility of coal
without using rather aggressive reagents capable of  breaking up its
bonded structure.  The phenomenon of swelling in organic solvents
was well known from the work of this period, and  some rough notions
of the physical  chemistry of this process were inferred by analogy
with properties  of synthetic and  natural,  cross-linked, rubbery
polymers.

     The  energy  crisis of the mid-seventies produced  a  surge in
coal  research  aimed  especially at  solving  some of  the  many
engineering problems  associated  with producing  liquid fuels from
coal.   Studies  of  swelling were  undertaken,  because it  seemed
plausible that coal in its fully expanded, swollen state would be
more  susceptible to  reagents capable  of breaking  up  the  coal
structure.   Certainly it appeared to be true that solvents that
produced the greatest swelling of coal,  also seemed  to yield the
greatest amount of extract.

     For a variety  of reasons, such work on  liquefaction has_ not
yet met with economic success, and has  been  largely  halted since
the oil price collapse of the eighties.  Despite  the heavy emphasis
on engineering  during this period, many   interesting aspects of
coal  chemistry   and  structure were restudied,  and  more  firmly
established by some  of the  newer methods for characterization of
complex solids,  such  as  FTIR, proton and  13C-NMR,  and the various
pyrolysis based  techniques  (Py-GC, Py-GC/MS, etc.).  The conceptual
model of  coal  as a  macromolecular matrix plus a  sorbed,  mobile
phase of extactable constituents was evolved and  popularized during
this post-classical period (see Given, 1986).

PHYSICAL CHEMISTRY OF COAL SWELLING

     For the purposes of the following discussion,  it is presumed
that the macromolecular matrix of coal,  exhaustively extracted of
any  small  sorbed species is  the subject of intrinsic interest.
When swelling studies are attempted on coal without  pre-extraction,
soluble material dissolves from the coal into the test  solvents for
long periods of time  (months) , and shifts the final equilibrium.
In the following discussion,  "matrix" or  "coal matrix" will denote
an ideal, totally extracted sample.

     Equilibrium volumetric swelling of coal matrix  in  contact with
an organic  liquid  represents a free  energy  balance   involving
several contributions, more or less separable:

-------
                                                     SUPPLEMENT 4
                                                     PAGE 5

     1.   Strain Energy  (isotropic stretching  of macromolecular
     matrix)

     2.  Chemical Interaction Energy (between solvent and matrix)

     3.   Entropy of  Mixing (entropy  increase  due to  mixing of
     solvent with macromolecular matrix)

     4.  Strain Entropy (entropy decrease due to stretching)

     Depending  on  the details  of  the  statistical  mechanical
treatment of these  energetic and entropic contributions, various
levels of realism can be achieved.

     The  simplest  statistical  mechanical  model  that   has  been
applied to coal is that of Flory and Rehner,  1943 that ignores term
1,  treats only  van der  Waals   interactions  in  term  2   (i.e.,
specific, site-dependent interactions are  completely ignored), and
treats terms 3  and  4  with expressions  involving a mixing entropy
computed  from  volume  fractions,  and a  strain  entropy calculated
from a Gaussian model of cross-linked chain structure.

     For  the  most  part,   the  Flory-Rehner  model  fails in  its
application  to coal,  and  the most obvious  flaw  is the lack of
inclusion of site-specific  interactions.  The  inclusion of these
terms  via the  Drago, E  & C  formulation  of  general  acid-base
interactions has been  suggested by Fowkes,  1980, and are definitely
needed to account for  the  exothermic heats of wetting observed for
coals in many solvents.

     It  is  not  so  obvious that  the  strain energy term can be
ignored in coal-solvent systems.  The observed heats of wetting in
coal immersion  experiments  should  be a net  heat effect  including
both chemical interaction  terms,  and a  strain energy term; careful
calorimetric studies are needed here to determine the importance of
the strain energy term directly.

ANNOTATED BIBLIOGRAPHY

Allen,  S.J.,  McKay,  G. ,  and Khader,  K.Y.H.,   1989,  Equilibrium
Adsorption Isotherms  for  Basic Dyes onto  Lignite, J.  Chem.  Tech.
Biotech, v. 45, p. 291-302.

     [The authors  conclude that  the  sorption  isotherms exhibit
     deviations  from  the  models,  because the  sorbed substances
     expand the sorbing particles by swelling, and  thus create "new
     surface."]

Brenner,  D.,  1983,  In  Situ Microscopic  Studies   of  the Solvent
Swelling of Polished Surfaces of Coal,  Fuel v.  62, p.  1347-1350.

-------
                                                     SUPPLEMENT  4
                                                     PAGE  6

Brenner,  D. ,   1984,  Microscopic  in-situ  Studies of  the  Solvent
Induced  Swelling of Thin Sections of Coal,  Fuel v.  63, p.  1324-
1328.

Brenner,  D. ,  1985,  The Macromolecular Nature of Bituminous  Coal,
Fuel v.  64, p. 167-173.

Callanan, J.E.,  Filla,  B.J., McDennott, K.M. , and Sullivan,  S.A.,
1987, Enthalpies of  Desorption  of Water from Coal Surfaces,  Amer.
Chem. Soc., Div.  Fuel  Chem. Preprints v- 32(1),  p. 185-192.

Cody, G.D., Jr., Larsen,  J.W. ,  and Siskin, M. ,  1988, Anisotropic
Solvent  Swelling of  Coals, Energy & Fuels v. 2,  p. 340-344.

Collins,  C.J., Hagaman. E.W., Jones. R.M. , and Raaen, V.F.,  1981,
Retention of Pyridine-^C  and other f4C-Labelled Amines by Illinois
No. 6 Coal, Fuel v.  60, p. 359-360.

Cooke,  N.E.,   and Gaikwad,  R.P.,  1984,  Removal of  Pyridine and
Quinoline from Coal  and Coal Extracts, Fuel v. 63, p. 1468-1470.

Dryden,   I.G.C.,  1951,  Action   of  Solvents  on  Coals   at  Lower
Temperatures I- A Qualitative Survey of the Effects of Liquids upon
Bright Coals of  Low  Rank, Fuel  v. 30, p. 39-44.
Dryden,  I.G.C.,  1951,  Action of  Solvents  on  Coals  at
Temperatures  II- Mechanism  of Extraction  of Coals  by Specific
Solvents and the Significance of Quantitative Measurements, Fuel v.
30, p. 145-158.

Dryden,  I.G.C.,  1951,  Action of  Solvents  on  Coals  at  Lower
Temperatures  III- Behavior  of  a  Typical Range  of  British Coals
Towards Specific  Solvents, Fuel v. 30, p. 217-233.

Dryden,  I.G.C.,  1951,  Action of  Solvents  on  Coals  at  Lower
Temperatures IV- Characteristics of Extracts and Residues from the
Treatment of Coal with Amine Solvents, Fuel v. 31, p.  176-199.

Durand,  B.  (ed.), 1980,  KEROGEN   Insoluble  Organic Matter from
Sedimentary Rocks. Editions  Technip, Paris.

Fowkes, F.M. , 1980, Donor-Acceptor Interactions  at Interfaces,  in
Polymer Science and Technology vol. 12A, Plenum  Press,  NY, p. 43-
52.

Given, P.H. ,  1984, An Essay  on the Organic Geochemistry of Coal,
Coal Science, vol. 3,  Academic Press, Inc., p. 63-252.

Given, P.H.,  1986, The  Concept of  a  Mobile  or Molecular Phase
within the Macromolecular Network of Coals:   A Debate,  Fuel v. 65,
p. 155-163.

-------
                                                     SUPPLEMENT  4
                                                     PAGE 7

Green,  T.K.,  Kovac,  J. ,   and  Larsen,  J.W.,  1984,  A  Rapid and
Convenient Method for Measuring the Swelling of  Coals by Solvents,
Fuel v- 63, p. 935-938.

Green,  T.K.,  and  Larsen,  J.W.,  1984,  Coal  Swelling  in Binary
Solvent Mixtures:   Pyridine-Chlorobenzene and N,N-Dimethylaniline-
Alcohol, Fuel V- 63, p. 1538-1543.

Green, T.K.,  and West,  T.A.,  1986,  Coal Swelling in Straight-Chain
Amines:  Evidence  for Specific Site Binding, Fuel v.65, p. 298-299.

Green, T.K.,  1987,  The  Macromolecular Structure of Coal, J. Coal
Quality v- 6(3), 90-93.

Grillet, Y.,  and  Starzewski, P.,  1989,  Thermocheroical Studies of
Wetting Phenomena of Coals  by Organic  Solvents such as Methanol or
Tetralin, Fuel v.  68, p. 55-57.

Flory, P.J.,  and  Rehner, J., Jr.,  1943,  Statistical Mechanics of
Cross-Linked  Polymer Networks   II.   Swelling,  J.  Chem. Phys. v-
11(11), p. 521-526.

Hall, P.J.,  Marsh, H., and  Thomas,  K.M.,   1988,  Solvent Induced
Swelling of Coals  to Study Macromolecular Structure,  Fuel v. 67, p.
863-866.

Hombach, H.-P-,  1980, General Aspects of Coal Solubility, Fuel v.
59, p. 465-470.

Howell, J.M., and Peppas,  N.A., 1987,  Macromolecular Structure of
Coals   8.   Viscoelastic Behavior  of  Coal  Networks  Determined by
Thermomechanical Analysis,  Fuel v. 66, p.  810-814.

Kaiser, W.R., 1974,  Texas  Lignite:   Near  Surface and  Deep-Basin
Resources, Report of Investigations  No.  79,  Bureau  of Economic
Geology, The University of Texas at Austin, Austin, TX.

Kirov,  N.Y.,  O'Shea,  J.M.,   and Sergeant,   G.D.,   1967,  The
Determination of  Solubility  Parameters  for Coal, Fuel  v.  46, p.
415-424.

Kybett, B. ,  Potter, J., Etter, M., and Krahe, M.,  1987,  The Effect
of  Solvent Extraction  on  the  Reflectance  of   Coal and Coal-Oil
Mixtures, Amer.  Chem.  Soc., Div.  Fuel  Chem.  Preprints v.  32(1), p.
9-11.

Larsen, J.W. , and  Kuemmerle, E.W.,  1978,  Heat of Wetting of Coal by
Tetralin:  Evidence for Structural Disruption at 25°C, Fuel v. 57,
p. 59.

Larsen,   J.W.,    Kennard,   L.,   and   Kuemmerle,   E.W.,    1978,

-------
                                                     SUPPLEMENT 4
                                                     PAGE 8

Thermodynamics of Adsorption of Organic Compounds on the Surface of
Bruceton Coal Measured by Gas  Chromatography, Fuel v.  57, p. 309-
313.

Larsen, J.W.,  and Lee,  D. ,  1983,  Effect of  Solvent Swelling on
Diffusion Rates in Bituminous  Coal, Fuel v. 62, p. 1351-1354.

Larsen, J.W.,  and  Lee,  D. ,  1985,  Steric Requirements  for Coal
Swelling by Amine Bases, Fuel  v. 64, p. 981-984.

Larsen, J.W., Green, T.K., and Kovac, J., 1985, The Nature of the
Macromolecular Network Structure of Bituminous Coals,  J. Org. Chem.
V. 50, p. 4729-4735.

Larsen, J.W., and Wernett,  P.,  1988, Pore Structure of Illinois No.
6 Coal, Energy & Fuels v. 2, p.  719-720.

Lucht, L.M., and Peppas, N.A., 1984, The Molecular Weight between
Crosslinks of Selected American Coals,  Preprints Amer.  Chem. Soc. ,
Div. Fuel Chem. v- 29(1), p. 213-219.

Mastral,  A.M.,   Izquierdo,  M.T.,  and Rubio,  B. ,  1990,  Network
Swelling of Coals, Fuel  v. 69, p. 892-895.

Nelson,  J.R.,  Mahajan,  O.P.,  and  Walker,  P.L.,   Jr.,  1980,
Measurement  of  Swelling of  Coals  in  Organic Liquids:    A  N<^w
Approach, Fuel v. 59, p. 831-837-

Painter, P.C., Nowak, J., Sobkowiak, M., and Youtcheff, J., 1987,
Hydrogen Bonding  and Coal Structure, Amer. Chem.  Soc, Fuel Div.
Preprints 32(1), p.  576-582.

Painter, P.C.,  Sobkowiak, M.,  and Youtcheff, J.,  1987,  FT-I.R.
Study of Hydrogen Bonding in Coal, Fuel v. 66, p. 973-978.

Reucroft,   P.J.,   and   Patel,   K.B.,   1983,   Surface  Area  and
Swellability of Coal, Fuel v.  62, p. 279-284.

Sanada, Y. , and  Honda,  H.,  1966, Swelling Equilibrium of Coal by
Pyridine at 25°C,  Fuel v. 45,  p.  295-300.

Sanada, Y., and  Honda,  H.,  1967, Equilibrium Swelling of Coal by
Various Solvents, Fuel v. 45,  p. 451-456.

Schafer, H.N.S.,  1984, Determination of Carboxyl Groups in  Low-Rank
Coals, Fuel v. 63, p. 723-726.

Shadle, L.J.,  Khan,  M.R.,  Zhang,  G.Q., and  Bajura,  R.A.,  1989,
Investigation  of Oil Shale  and Coal  Structures  by Swelling  in
Various Solvents,  Amer.  Chem.  Soc.,  Div-  Petroleum  Chem,  Inc.
Preprints v. 34(1),  p. 55-61.

-------
                                                     SUPPLEMENT  4
                                                     PAGE  9

Shibaoka, M.,  Stephens,  J.F.,  and Russell, N.J., 1979, Microscopic
Observations of the Swelling of a High-Volatile Bituminous  Coal in
Response to Organic Solvents,  Fuel v. 58, p. 515-522.

Snider,  J.,  and Covay,  K.J.,  1987,  Premining Hydrology  of the
Logansport Lignite Area,  DeSoto Parish, Louisiana, Water Resources
Technical Report No. 41, Louisiana Department of Transportation and
Development, Baton Rouge, Louisiana.

Szeliga, J., and Marzec,  A., 1983, Swelling  of  Coal in Relation to
Solvent Electron-Donor Numbers, Fuel v. 62,  p. 1229-1231.

Van Krevelen,  D.W., 1965, Chemical Structure  and Properties  of  Coal
XXVIII - Coal Constitution and Solvent Extraction, Fuel v. 44, p.
229-241.

Van  Krevelen,   D.W.,   1981,   COAL     Typology-Chemistry-Physics
Constitution.  Elsevier Scientific Publishing Co.,  N.Y.

Weinberg, V.L.,  and  Yen, T.F.,  1980,  'Solubility Parameters' in
Coal and Coal Liquifaction Products, Fuel v. 59,  p. 287-289.

Wightman, J.P.,  Glanville, J.O., Hollenhead,  J.B.,  Phillips,  and
Tisa,  K.N.,   1987,  Heats  of  Immersion  of  Bituminous Coals in
Liquids, Amer. Chem.  Soc.,  Div. Fuel Chem.  Preprints v. 32(1), p.
205-208.

-------
                                                     SUPPLEMENT 5
                                                     PAGE 1
                             REVIEW;
          THE SWELLING OF CELLULOSE AND ITS RELATION TO
 THE SWELLING AND SORPTION CHARACTERISTICS OF SOIL ORGANIC MATTER

                         William G. Lyon
               NSI Technology Services Corporation
         Robert S.  Kerr Environmental  Research Laboratory
                       Ada, Oklahoma 74820

INTRODUCTION

     Cellulose  is  a  biopolymer  of  glucose,  and an  important
constituent  of  soil organic matter  or peat which  has  been only
slightly humified.   The  polymeric chains of cellulose are cross-
linked  by hydrogen-bonds,  which  in  most instances  permit only
solvent-swelling to occur when common  organic  liquids are applied,
rather than formation of true polymer solutions.

     The  sorption/swelling  characteristics   of  cellulose  are
dominated by site-specific, hydrogen-bonded interactions with the
glucose monomer units.  In  mixtures of organic liquids capable of
hydrogen-bonding,  complex  preferential   sorption  phenomena  are
observed.  These are expected also in soils containing cellulose as
a component.

CRITICAL POINT PHENOMENA

     The occurrence  of  critical temperature phenomena in swollen
cellulosic gels seems to have been tentatively established for time
scales  on  the order  of a  day,  but it  has not  yet  been widely
recognized.   These phenomena  can give  rise  to  very  puzzling
sorption  behavior,   such  as  the  sudden  doubling  of  saturation
sorption when the temperature is increased by a small amount.  The
fortuitous location  of these  transition  points near ambient room
temperature  for  many  solvent-cellulose  systems  makes  careful
control  of  temperature in  determination of  swelling  spectra or
sorption isotherms an experimental necessity.

     Critical composition phenomena are known to occur in many gel
systems; these are  also to  be  expected in cellulosic gels.  This
kind of behavior should be  sought in ternary systems:  H-bonding
solvent (1)  - Miscible Solvent  (2) - Cellulose.  For example, H20  -
 Acetone - Cellulose,  or  DMSO - Acetone - Cellulose  (above 24°C)
would be reasonable candidates for such  a  transition.   In these
examples, it might be expected that as the acetone concentration of
the external solution is gradually increased, the  initially large
swelling of the cellulose will abruptly decrease over a very small
range of composition.

-------
                                                     SUPPLEMENT 5
                                                     PAGE 2

     On  time  scales  much  longer  than  a  day,  the  critical
temperature phenomena, noted earlier, apparently disappear.  This
means that these phase transitions are probably kinetic artifacts
of a very slow solvation step in which a hydrogen-bonding solvent
disrupts the interchain, hydrogen-bonded cross-links.


BIBLIOGRAPHY

Calahorra, M.E.,  Cortazar,  M., Eguiazabal, J.I., and Guzman, G.M.,
1989, Thermogravimetric Analysis  of Cellulose:   Effect  of the
Molecular Weight on Thermal Decomposition, J. Appl.  Polymer Sci. v.
37, p. 3305-3314.

     [These results are the best available guide to maximum drying
     temperature for cellulosic  soil  organic matter; the oft-cited
     limit of 105°C appears reasonable (but only in vacuum or inert
     gas!); however,  some celluloses  may not have a truly constant
     dry mass without pyrolysis.]

Chitumbo,  K. ,  Brown,  W. ,  and  De Ruvo,  A.,  1974,  Swelling  of
Cellulosic Gels, J. Polymer Sci., Symposium No. 47, p.  261-268.

     [A key set of results which firmly established the existence
     of phase transitions in swelling versus temperature data for
     a variety of strongly hydrogen-bonding solvents,  such as DMSO.
     DMF etc. interacting with cellulose.  Two aspects which seem
     to have  been  left uninvestigated were  the  possibilities of
     kinetic sluggishness and of hysteresis being associated with
     these transitions; such accompanying phenomena do seem to be
     present in Coal-Pyridine systems.]

Golub,  N.V.,  Kaputskii,  F.N.,  and Yurkshtovich,  T.L.,  1986,
Mechanisms  of  Swelling of  Cellulose  and  Diethylaminohydroxy-
cellulose in Water-Organic Solvent Systems, Kolloid.  Zhur. v. 48,
p. 1009-1014.  [English Transl.  Plenum Publ. Corp., 1987]

     [Equilibrium isotherms are  given for Methanol-Water-Cellulose
     and for 2-Propanol-Water-Cellulose Systems.  The data for the
     preferential sorption  of Methanol-Water solutions on Cellulose
     agree qualitatively with our swelling data taken at Kerr Lab:
     there   exists   an   intermediate   composition   (crossover
     composition)  of aqueous methanol which  is  sorbed  without
     change  in  composition.   The  published data do not include
     total sorption data, and are thus,  not sufficient to construct
     the complete ternary phase  diagrams for the systems studied.]

Hudson, S.M., and Cuculo, J.A.,  1980, The Solubility of Unmodified
Cellulose:  A Critique of the Literature, J. Macromol. Sci. - Rev.
Macromol. Chem. v. C18(l),  p. 1-82.

-------
                                                     SUPPLEMENT  5
                                                     PAGE 3

      [A very useful review of the literature.   Emphasizes the lack
      of true solubility in single, non-reactive organic solvents.]

Koura, A., Schleicher, H., and  Philipp, B., 1972,  Untersuchungen
zur   Quellung   und   Losung   von   Cellulose   in   aminhaltigen
Flussigkeitsgemischen, 2.  Mitt.:  Untersuchungen zur Loslichkeit
von  Cellulose  in  Dimethylsulfoxid/Amin-Gemischen,  Faserforsch.
Textiltech. v. 23(3),  p.  128-133.

      [Investigation of Swelling and Solution of Cellulose in Amine-
      Containing Liquid Mixtures  2.  Investigation of Solubility of
      Cellulose in Dimethlsulfoxide/Amine Mixtures.]

Koura, A., Schleicher, H.,  and Philipp, B.,  1973, Untersuchungen
zur   Quellung   und   Losung   von   Cellulose   in   aminhaltigen
Flussigkeitsgemischen,  6.   Mitt.:   Zur  Strukturveranderung  von
Cellulose durch Amine und  Aminlosungen, Faserforsch. Testiltech. v.
24(5), p. 187-194.

      f...6.  On Structural Modification of Cellulose by Amines and
      Amine Solutions.]

Kremer,  R.D.,  and  Tabb, D. ,   1990,  Paper:     The  Beneficially
Interactive Support Medium for  Diagnostic Test Development, Amer.
Lab.  v- 22(3), p. 136-143.

      [An excellent introduction to the hydrogen-bonded structure of
      cellulose, and the various hydroxyl-group binding sites.]

Larsson, A.,  and Johns, W.E., 1988, Acid-Base Interactions between
Cellulose/Lignocellulose  and Organic Molecules, J. Adhesion v. 25,
p. 121-131.

      [A   chemometric  approach,   which  attempts  a   canonical
      correlation  between  swelling and  tensile energy  absorption
     with  various  organic  fluid descriptors.    These  authors
      recognize the importance of the Drago C&E model of acid-base
      interactions  for describing  the  site-specific  interactions
      that  dominate   the   enthalpic  portion  of  the  chemical
      interactions; unfortunately, they  are forced to use the less
     adequate formulations of donor-acceptor parameters for their
     solvent descriptors.]

Lokhande,  H.T.,   1978, Swelling Behavior  of  Cotton  Fibers  in
Morpholine and Piperidine, J.  Appl.  Polymer  Sci. v.  22,  p. 1243-
1253.

Moers, M.E.C.,  Boon,  J.J.,  de  Leeuw,  J.W.,  Baas,  M. ,  and P.A.
Schenck, 1989, Carbohydrate  Speciation and Py-MS Mapping of Peat
Samples  from  a  Subtropical  Open  Marsh Environment,  Geochim.
Cosmochim. Acta v. 53, p. 2011-2021.

-------
                                                     SUPPLEMENT 5
                                                     PAGE 4

Nayer, A.N.,  and  Hossfeld,  R.L., 1949, Hydrogen  Bonding and the
Swelling of Wood in Various Organic Liquids, J.  Amer.  Chem. Soc. v.
71, p. 2852-2855.

Philipp,  B.,   Schleicher,  H. ,  and  Wagenknecht,   W. ,  1973,  The
Influence of  Cellulose Structure  on  the Swelling  of Cellulose in
Organic Liquids, J.  Polymer Sci., Symposium No. 42, p. 1531-1543.

      [One of the most striking results of this work, is the kinetic
     study of DMSO sorption as a function of temperature depicted
     in Fig.  2.   The discontinuity  between the 20°  and the  40°C
     result  is apparently   connected  to   the  phase  transition
     observed  by   Chitumbo  et  al.   1974.   The discontinuity in
     swelling  is  observed on time  scales up  to 20 hours,  but
     apparently disappears on time  scales of 200  hours  or more.
     The  implication  is that the apparent  equilibrium  swelling
     achieved after one day.  slowly shifts  to a markedly different
     value, given sufficient time.]

Philipp, B., Schleicher,  H.,  and Wagenknecht, W., 1973, Quell- und
Loseprozesse   bei   Zellulose  mit   unterschiedlicher  Struktur,
Zellstoff und Papier v- 23(11),  p. 324-330.

      [Swelling and Solution Processes in Celluloses with Different
     Structures.]

Robertson,  A.A.,  1970,  Interactions  of Liquids  with Cellulose,
Tappi v. 53(7), p. 1331-1339.

Schleicher, H., 1982,  Zur Abhangigkeit der Cellulosequellung von
den Donor- und Acceptorzahlen der Quellmittel,  Acta Polymerica v.
34, p. 63-64.

      [The  Dependence   of Cellulose-Swelling  on  the Donor-  and
     Acceptor-Numbers of the Swelling Agent.

     Has plot of mols swelling medium per mole glucose-unit versus
     the sum of donor and acceptor numbers,  squared.  The data were
     taken at 20°C;  therefore, the swellings observed for many of
     the polar molecules  (including DMSO) were not  at their maximum
     values.]


Thode, E.F., and Guide, R.G.,  1959, A Thermodynamic Interpretation
of the Swelling of  Cellulose in Organic Liquids:   The Relations
Among Solubility Parameter,  Swelling, and Internal Surface, Tappi
v- 42(1), p. 35-39.

      [This  is  one  of  the   more   lucid  treatments   of  the
     thermodynamics of swelling.]
                                      lot

-------
                                                     SUPPLEMENT 6
                                                     PAGE 1

    REFERENCES ON DRAGO'S E AND C FORMULATION OF GENERALIZED
   ACID-BASE  INTERACTIONS  AND  ITS APPLICATIONS  TO SORPTION OF
    ORGANIC MOLECULES  ON SOILS.  COALS. POLYMERS AND MINERALS
                         William G. Lyon
               NSI Technology Services Corporation
         Robert  S.  Kerr Environmental  Research Laboratory
                       Ada, Oklahoma 74820

     The empirical theory developed by Drago and his collaborators
attempts to reproduce  quantitatively  the  enthalpies  of mixing of
liquids, including a broad range of site-specific associations in
dilute solution.  The Drago equations, although not as economical
of parameters as  some  other  donor-acceptor formulations,  seem to
reproduce  the  calorimetric  data  better  than  other  contending
models,  and  to  have  an  intimate connection  with  vibrational
frequency shifts observed for active functional groups by infrared
spectroscopy.

     The  heart  of  the  Drago  approximation  is  the  empirical
equation:


                         -AH = EaEb + CaCb
where  the enthalpy  of mixing  acid  A and  base B  per mole  of
interaction is represented  as a  dot product of parameter vectors
describing the generalized acid and generalized base properties of
the substances being mixed.  The  parameters Ea and Eb are said to
describe the electrostatic contributions, and the parameters Ca and
Cb are  said to represent the covalent contributions from  acid A and
base B, respectively.  As recently pointed out by  Fowkes, 1990, the
acid  A  also  has  a   set  of  parameters   describing   its  basic
characteristics,  and  likewise,  the  base   B  also has  a  set  of
parameters  describing  its acid  characteristics.   While  these
usually  represent  much  smaller  contributions  to  the  total,
recognition  of  this  "amphipathic"  character  for  most  organic
solvents other than hydrocarbons is important.

     The  parameters  involved in  the Drago  approximation  can be
obtained  for  sites on  polymers  and  solid surfaces by means of
"probe" molecules whose properties have already been established.
These test substances must have significantly different values of
E and C so that the simultaneous equations  lead to precise values
for the unknown parameters.  Several different approaches to this
measurement problem have  been reviewed recently by Fowkes,  1990.
Among  the more useful  approaches are measurements  of frequency
shifts of interacting functional  groups with respect to  a standard
                                      (03

-------
                                                      SUPPLEMENT 6
                                                      PAGE  2

dilute solution by fourier transform infrared spectroscopy.

     The   frequency   shift  method  depends   on  the   empirical
observation that the shift "in frequency, Av, is  a  linear function
of the enthalpy of interaction described above as a function of the
E and C parameters.   The frequency-shift relationship for a  test
substance must  be calibrated against calorimetric data, but  once
this has been accomplished, the frequency-shift method can  be  used
to  derive E  and  C parameters  for unknown  substances  including
polymers and  solid surfaces.   This method seems to have the  most
potential for defining the nature  and  number of binding sites  on
very complex  sorbents such as soils.
Chamberlain,   C.S.,   and  Drago,   R.S.,  1979,   Comparison  of
Thermodynamic  Parameters  Obtained  by  Gas  Chromatographic and
Spectrophotometric Methods  for the  Interaction of  a  Lewis Acid
Transition Metal Complex with a Lewis Base, J. Amer.  Chem. Soc. v.
101(18), p. 5240-5245.

Doan, P.E., and Drago, R.S.,  1982, An E and C Modification of the
0 -  TT*  Solvation Approach, J.  Amer. Chem. Soc.  v.  104,  p.  4524-
4529.

Doan, P.E., and Drago,  R.S.,  1984, Requirements  and Interpretation
of Linear  Free Energy Relations,  J. Amer. Chem.  Soc.  v. 106, p.
2772-2774.

Drago, R.S., O'Bryan, N., and Vogel, G.C., 1970,  A Frequency Shift-
Enthalpy  Correlation  for  a  Given  Donor with  Various  Hydrogen-
Bonding Acids, J. Am.  Chem. Soc. v. 92, p. 3924-3929.

Drago, R.S., Vogel, G.C., and Needham, T.E., 1971, A Four-Parameter
Equation  for Predicting Enthalpies  of Adduct  Formation,  J. Am.
Chem. Soc. v. 93, p. 6014-6026.

Drago,  R.S.,  Nusz,  J.A.,  and  Courtright, R.C.,  1974,  Solvation
Contributions  to  Enthalpies  Measured  in Methylene  Chloride, J.
Amer. Chem. Soc. v.  96(7), p. 2082-2086.

Drago, R.S., 1977, Modern Acid-Base Chemistry/ Proc.  Summer School
on Stability Constants, Florence,  Italy, June 1974,  p. 117-123.

Drago,  R.S.,  Parr,  L.B.,  and  Chamberlain,  C.S.,  1977, Solvent
Effects and  Their  Relationship to  the E and C Equation,  J. Am.
Chem. Soc. v. 99, p. 3203-3209.

Drago, R.S., 1980, The Interpretation of Reactivity in Chemical and
Biological Systems with the E and C Model, Coord. Chem.  Rev. v- 33,
p. 251-277.

-------
                                                     SUPPLEMENT 6
                                                     PAGE 3

Fowkes, F.M.,  1968,  Comments on "The Calculation of Cohesive and
Adhesive Energies,"  by J.F- Padday  and  N.D.  Uffindell,  J.  Phys.
Chem. v. 72(10), p.  3700.

Fowkes, P.M.,  1973,  Donor-Acceptor Interactions at Interfaces,
Recent Adv. Adhes., Proc. Amer.  Chem. Soc. Symp.,  p. 39-44 (meeting
1971).


Fowkes, P.M.,  and  Mostafa,  M.A.,  1978,  Acid-Base Interactions in
Polymer Adsorption, Ind. Eng. Chem. Prod. Res. Dev-  v- 17(1), p. 3-
7.

Fowkes,  F.M.,  1979,  Donor-Acceptor Interactions  at  Interfaces,
Amer. Chem. Soc.,  Div.  Org.  Coatings end Plastics Chem. Preor. v.
40, p. 13-18.

Fowkes, F.M.,  1980,  Donor-Acceptor Interactions at Interfaces, in
Polymer Science  and  Technology  vol.  12A, Plenum Press,  NY,  p. 43-
52.

Fowkes, F.M.,  Tischler, D.O.,  Wolfe, J.A.,  and  Halliwell,  M.J. ,
1981, Acid-Base  Complexes  of Polymers  with Solvents,  Amer.  Chem.
Soc., Div. Org.  Coatings and Plastics Chem. Prepr.  v.  46, p. 1-6.

Fowkes, F.M.,  1982,  "Characterization  of Solid Surfaces by Wet
Chemical  Techniques",  chapter  5  in Industrial Applications  of
Surface Analysis.  L.A. Casper,  C.J. Powell,  eds.,  ACS Symposium
Series 199, Amer.  Chem. Soc., Wash.  D.C., pp 69-88.

Fowkes, F.M.,  McCarthy, D.C.,  and Wolfe,  J.A.,  1983,  Predicting
Enthalpies  of  Interfacial  Bonding  of  Polymers  to  Reinforcing
Pigments, Amer.  Chem.  Soc. Div. Polymer Chem. Prepr. v. 24(1), p.
228-9.

Fowkes, F.M.,  McCarthy, D.C., and Tischler, D.O., 1983, Predicting
Enthalpies  of  Interfacial  Bonding  of  Polymers  to  Reinforcing
Pigments,  in Molecular Characterization of Composite Interfaces.
edited by H. Ishida  and G. Kumar, Plenum Press, NY, pp 401-411.

Fowkes, F.M., Tischler, D.O., Wolfe, J.A., Lannigan, L.A., Ademu-
John,  C.M.,  and Halliwell,  M.J.,  1984,  Acid-Base Complexes of
Polymers, J.  Polymer Sci.,  Polymer  Chemistry Edition,  v.  22, p.
547-566.

Fowkes,  F.M.,   1984,  Acid-Base Contributions to  Polymer-Filler
Interactions, Rubber Chem. Technol. v. 57(2), p. 328-343.

Fowkes, F.M.,  1984, Spectral and Calorimetric Determination  of the
Intermolecular  Interactions of  Solvents,  Amer.  Chem.  Soc. Div.
Polymeric Mater. Prepr. v- 51,  p. 522-7.

-------
                                                     SUPPLEMENT  6
                                                     PAGE 4

Fowkes, P.M., Lloyd, T.B.,  Li,  G. ,  Jones,  K.L., and Wolfe, J.A. ,
1985,  Final  Report,  untitled,  [Techniques  for the  study of the
surface sites of coal  powders],  DOE/PC/50809-T11 under DOE Grant
No. DE-FG22-82PC50809,  27 pp.

Fowkes, F.M., Jones, K.L., Li, G., and Lloyd, T.B., 1989, Surface
Chemistry of Coal by Flow Microcalorimetry, Energy & Fuels v- 3, p.
97-105.

Fowkes, F.M., Riddle,  F.L., Jr., and Cole,  D.A.,  1989, Chemical
Characterization of the Surface Sites of Coal, Report DOE/PC/79925-
4    under Contract No. DE-FG22-87PC79925.

Fowkes, P.M., 1990, Acid-Base Measurements of Solvents, Polymers,
and Inorganic Surfaces, Preprint, 36 pp.

Guidry, R.M., and  Drago, R.S.,  1973, An Extension  of the E and C
Equation to Evaluate Constant Contributions to a Series  of Observed
"Enthalpies of Adduct Formation", J.  Amer. Chem. Soc. v. 95(3), p.
759-763.

Joslin, S.T., 1984, Characterization of the Acidic Surface Sites on
Alpha-Ferric  Oxide Using  Flow  Microcalorimetry,  Ph.D.  Thesis,
Lehigh University,  210 pp.

Karger, B.L., Snyder,  L.R., and Eon, C. , An Expanded Solubility
Parameter Treatment for Classification and Use of Chromatographic
Solvents and Adsorbents, J. Chromatog.  v. 125, p. 71-88.

Rolling,  O.W.,  1978,  Comparisons between  Hydrogen Bond Donor-
Acceptor Parameters and Solvatochromic Red Shifts,  Anal. Chem. v.
50(2), p. 212-215.

Kroeger, M.K., and Drago, R.S., 1981, Quantitative Prediction and
Analysis of  Enthalpies  for the  Interaction  of Gas-Phase Ion-Ion,
Gas-Phase  Ion-Molecule,  and Molecule-Molecule Lewis Acid-Base
Systems, J. Amer. Chem. Soc. v. 103, p. 3250-3262.

Lim, Y.Y.,  Drago,  R.S.,  Babich, M.W.,  Wong, N. , and Doan, P.E. ,
1987,  Thermodynamic  Studies  of Donor  Binding  to  Heterogeneous
Catalysts, J. Amer. Chem. Soc. v. 109,  p. 169-174.

Marmo, M.J., Mostafa, M.A., Jinnal, J., Fowkes, F.M.,  and Manson,
J.A.,  1976,  Acid-Base  Interaction  in Filler-Matrix Systems, Ind.
Eng. Chem., Prod. Res.  Dev. v. 15(3), p. 206-210.

Nozari, M.S.,  and  Drago,  R.S.,  1970,   Spectral  and Calorimetric
Studies of Hydrogen Bonding with Pyrrole, J.  Amer.  Chem. Soc. v.
92(24), p. 7086-7090.

Nozari, M.W.,  and  Drago,  R.S.,  1972,  Elimination  of Solvation

-------
                                                     SUPPLEMENT 6
                                                     PAGE 5

Contributions to the Enthalpies of Adduct Formation in Weakly Polar
Solvents.     II.    Adducts   of  Bis(hexafluoroacetylacetonato)
copper(I), Inorg. Chem. v. 11(2), p. 280-283.

Nozari,  M.S.,  Jensen,  C.D.,  and Drago, R.S.,  1973,  Eliminating
Solvation  Contributions  to the  Enthalpy of Adduct  Formation in
Weakly Polar, Acidic Solvents, J.  Amer.  Chem.  Soc. v.  95(10), p.
3162-3165.

Pugh,  R.J.,   and  Fowkes,  F.M.,   1984,  The  Dispersibility  and
Stability of Carbon Black in Media of Low Dielectric Constant.  2.
Sedimentation Volume of  Concentrated  Dispersions, Adsorption and
Surface Calorimetry  Studies, Coll. Surf, v- 9, p. 33-46.

Riddle,  F.L., and Fowkes, F.M.,  1988, Spectral Shifts in the Acid-
Base Chemistry  of Polymers.   I.,  Amer. Chem.  Soc.  Div.  Polymer
Chem. Prepr.  v.  29(1), p. 188-189.

Rider, P.E.,  1980,  A Two-Parameter Model for Estimating Hydrogen
Bond Enthalpies of Reaction, J.  Appl. Polymer Sci. v. 25, p. 2975-
2984.
                                       107

-------
                                            SUPPLEMENT 7
                                            RSKSOP-104
                                            Revision No. 1
                                            Date:   March 25, 1991
                                            Page l of 22
                                            William G. Lyon
                                            David E. Rhodes

                  STANDARD OPERATING PROCEDURE

               SWELLING SPECTRUM OF ORGANIC MATTER

   Disclaimer;

     This Standard Operating Procedure has  been prepared for the
     use of the Robert S. Kerr Environmental  Research Laboratory of
     The U.S.  Environmental Protection  Agency and  may not  be
     specifically  applicable   to   the   activities   of   other
     organizations.

1. Purpose;   (Scope and Application)

     To determine the swelling spectrum of an organic-rich material
(concentrated soil organic matter,  peat,  pollen, spores, etc.)  in
a wide range of organic solvents and water.

2. Background;

     The swelling spectrum of an insoluble organic material gives
quantitative information on the  interaction between the material
and organic  solvents.   Swelling is the volumetric manifestation of
bulk  sorption;  that   is,  swelling   is   concerned  with  those
sufficiently small molecules  that can partition  into  the  solid
organic material. The  swelling spectrum can be used qualitatively
for "finger  printing" a complex mixture of solid organic materials
for comparison with  standards,  and  also  to determine  the  size-
exclusion limits of a macromolecular material.
3. Summary of Method;

     Powdered organic  matter is  placed in  small  glass  tubes and
exposed to liquid organic solvents.  Length measurements are made
before and after wetting with each solvent.   The basic method has
been described by Green et al.,  1984 for  coal samples.

4. Equipment Needed;

     A. A 7 or 8 cm3 sample of organic matter  sieved to pass 100
     mesh standard  sieve.    (Sample  may  be dry or  partly  moist
     depending on the purpose of the determination).  The procedure
     works best on materials  from which all  soluble materials have
     been removed by extraction.

     B.  Disposable  Wintrobe Tubes,  3 mm  ID,  mm  etched scale.
     (Scientific Products,  Cat. No. B4449).  These represent about

-------
                                            SUPPLEMENT 7
                                            RSKSOP-104
                                            Revision No. 1
                                            Date:  March 25, 1991
                                            Page 2 of 22
                                            William G. Lyon
                                            David E. Rhodes

     the smallest tubes for which this technique is practical.

     C.  Centrifuge  Adaptor so  Wintrobe tubes  can  be spun  in a
     standard centrifuge.

     D. Centrifuge  capable  of  2000  rpm.   (Most bench top models)
     swivel head.

     E. Magnifier with measuring reticle capable of reading to 0.1
     mm.  (Bausch & Lomb,  Cat.  No. 81-34-35, 81-34-38) .  Note that
     we use the measuring reticle to interpolate precisely between
     the graduations etched on the Wintrobe tubes.

     F. 2 ml luer-lok syringes and 8" pipetting  needles (18 gauge)
     for dispensing solvents.

     G.  Wide  selection  of  solvents  (see  Appendix  I)  spanning
     solvent parameter values from 14 to 48 MPa*.

     H. Teflon  caps,  drilled from  short lengths  of 5/16" Teflon
     rods, are  very much  better than Teflon tape for sealing the
     Wintrobe tubes.  The holes drilled in the  caps need to taper
     so that the caps will fit despite the slightly varying OD of
     the Wintrobe tubes.

     I. Fume Hood and protective equipment for  handling solvents.
     (See  MSD   sheets   for  information  on  proper  protective
     measures.)

     J. Coded sample rack with holes for Wintrobe tubes.

     K. Constant temperature bath or incubator for equilibration of
     loaded tubes.

     L. Radioactive strips to discharge static  build-up  (e.g.,
     Staticmaster Ionizing Units, Model 2U500).

     M. Heated  vacuum desiccator,   capable  of  reaching  105°C  and
     pressures of a few Torr.
5. Procedure;
     A. Load  the necessary Wintrobe  tubes  (20-50)  with powdered
     organic matter to about the  20% level (a 2 cm column of powder
     is equivalent to ca. 0.14 cm3).

-------
                                       SUPPLEMENT 7
                                       RSKSOP-104
                                       Revision No. 1
                                       Date:  March 25, 1991
                                       Page 3 of 22
                                       William 6. Lyon
                                       David E. Rhodes

     Note:   If it  is  desired to  run the  powdered organic
     material in the dried condition,  it is usually necessary
     to redry the material  in  the  tubes by heating at  105°C
     for  24  hours in  a  vacuum desiccator.   The  tubes are
     capped as soon as possible after vacuum is broken through
     a drierite  tube.   It is  recommended  that radioactive,
     anti-static strips  be  placed  in the  desiccator  during
     heating  to discharge  the build-up  of  static on the
     powdered samples.

B. Spin the loaded,  capped tubes at 1800 rpm for five minutes
on the centrifuge to get a standard packing.  (This speed is
a provisional  value;  further  experimentation is  needed  to
determine if higher values give more reproducible values for
some materials.  See Appendix V for discussion of this issue.)


C. Measure initial height, HI, for  each tube  with magnifier
and measuring reticle.   Record results.

D. Apply solvent to each sample. Sample  should be tapped onto
the side of the tube so that pipetting needle can be used to
introduce solvent starting at  the bottom of the  tube.   With
many samples, the solvent can be stirred into the powder with
the syringe needle.   Fill tubes with solvent to  80% mark.  Cap
immediately with Teflon caps.

E. Rotate and tap tube  to release small bubbles from powder.

F. Allow tube to sit undisturbed in sample rack overnight in
a nearly horizontal position  with  the  sample  spread out.
Check again for bubbles.  Tubes should be placed in a constant
temperature  bath or an  incubator  to  equilibrate  between
readings.

G. Before each height reading, spin tubes at 1800  rpm for five
minutes on Centrifuge.

H. Measure swollen  heights  H2, H3  etc. for each tube as in
step C.  Record heights,  date and time.

I. Final height measurement is taken when measurements  agree
for successive measurements  within the typical uncertainty of
the measurement.  Otherwise, repeat steps G and H.  Note any
changes in color of solvent above sample.
                               \\0

-------
                                            SUPPLEMENT 7
                                            RSKSOP-104
                                            Revision No. 1
                                            Date:  March 25, 1991
                                            Page 4 of 22
                                            William 6. Lyon
                                            David E. Rhodes
6. Calculations;
   A sequence of raw volumetric swelling values  is calculated from
the height  ratios H2/H1, H3/H1, etc.   These swelling values are
plotted versus  the solvent parameter,  6,  for each  solvent. (See
Appendix  I) .   Corrections can be made  for inert diluents by the
method outlined in Appendix II.  If  other soil minerals such as
Kaolinite or  Montmorillonite  are present,  the corrections become
much more complex.  Various strategies for  this  type of correction
are discussed in Appendix III.

     Besides  the  above elementary  manipulations  of the  data,
several additional transformations are useful.  The quantity C> lar,
which represents the number of moles of imbibed solvent per  cirr of
sorbent can be  calculated from the volumetric swelling, Q and the
solvent molar volume (cm3 mol"1) for each solvent (see Appendix I) :
If additionally,  the density of the  solid  sorbent is known  (see
RSKSOP-105) and some composition data are available, further useful
transformations are possible.  For example,  sorption on Cellulose,
a biopolymer  of glucose, can be  recalculated  as  moles of sorbed
solvent  per  mol  of  monomer  unit,   X  lar;  this   transformation
emphasizes the  formation of  specific molecular complexes  (if  such
are present).   Other transformations of this type might be  based on
moles of carboxylic  acid or phenolic  groups per unit mass should
these seem  to be  involved  in sorption.   Finally,  if the carbon
content of the  sorbent (RSKSOP-101)  has been measured, values of
the amount sorbed  can be put on a per gram organic carbon basis.
Examples of various  calculations on a swelling spectrum are shown
as figures in Appendix IV-

7. Quality Control:

     The most  important  consideration in working  with this  many
samples in small tubes is the potential for mixing them up during
handling.  Coding  of loading rack and of Centrifuge tube adaptor
will assist  in this  bookkeeping.   Labeling on the tubes is the
safest measure.

     A second  factor that  is important in obtaining  reproducible
                                        If

-------
                                            SUPPLEMENT 7
                                            RSKSOP-104
                                            Revision No. 1
                                            Date:  March 25, 1991
                                            Page 5 of 22
                                            William G. Lyon
                                            David E. Rhodes

values is the  total elimination of  gas bubbles  from the wetted
samples before measurement  of length.   Various  tricks  that can
assist this process include lowering  the  pressure to enlarge the
bubbles,  and  using ultrasonic  vibrations to  help  fluidize the
powder and allow bubbles to float up.

     A third factor that can be a source of error  in the procedure
as currently described is the  variability  of the etch marks on the
Wintrobe tubes.  Since the measuring reticle is capable of spanning
only 1 cm,  our  measurements  are an  interpolation procedure that
relies on the accuracy  of the etched  graduations for part of the
total height.   Alternatives  to  the  present  procedure  would  be
calibration of the etch marks against a standard probe, or use of
a  precision  cathetometer   (or  similar  instrument)   for  the
measurement of total heights.   Either of these measures would add
significantly to the time and expense of obtaining swelling data,
and would be recommended  only when the  need  for higher precision
justifies the added difficulties.

     Finally, the  proper conditions of compaction for comparison of
swollen powder heights with dry powder heights is an issue raised
by one of the reviewers of the project report.   It has been found
that constant starting amounts of powder helps achieve consistent
results.   Our use  of a constant 1800 rpm value for both wetted and
dry powders merely  followed  the published procedure  of Green et
al., 1984.  Appendix V  contains a discussion  of a correction for
the variable buoyancy component to the  net compacting force from
the various organic liquids.   Our opinion is that this additional
correction would  multiply the time  and expense  of  the method at
least 10 fold, and would greatly undermine the usefulness of what
is, at best, a semi-quantitative survey technique.

     Purified cellulose (Aldrich 31,069-7)  is  recommended  as a
standard for checking  out the  procedure.   It is fairly hygroscopic
when dried,  and must  be dried carefully  under  vacuum to prevent
oxidation.   Prevention  of the build-up of static charges during
drying by means of a radioactive source mounted inside the vacuum
drying apparatus is essential.

8. Safety Considerations;

     All operations involving manipulation of  solvents should be
conducted under a  fume hood.  Many of the suggested solvents  on the
attached list (Appendix I) are flammable,  toxic or corrosive; the
experimenter should be  familiar with the  hazards associated with
all solvents used,  and  wear appropriate eye  and hand protection.

-------
                                            SUPPLEMENT 7
                                            RSKSOP-104
                                            Revision No. 1
                                            Date:  March 25, 1991
                                            Page 6 of 22
                                            William 6. Lyon
                                            David E. Rhodes
See MSD sheets for details.
     Nitromethane  and  nitroethane apparently have some potential
for detonation under adiabatic compression; thus, it is recommended
that  these substances  not  be  exposed  to  ultrasonic vibrations,
especially  while  in  contact  with  organic matter.    We  have
cautiously tested  these substances ultrasonically while contacted
with several types of peat,  and have found  no problem (so far!) in
an ordinary laboratory ultrasonic bath.

     Chlorinated solvents must be disposed of in a separate waste
container from other solvents.  Carbon Disulfide wastes must now be
kept separate  from all other flammable  solvent wastes.

9. References;

Barton, A.F.M., 1983, Handbook of Solubility Parameters and Other
Cohesion Parameters, CRC Press, Boca Raton, FL.

Green,  T.K.,   Kovac,  J. ,  and  Larsen,  J.W., 1984,  A  Rapid  and
Convenient Method  for Measuring the Swelling  of  Coals by Solvents,
Fuel v. 63, p. 935-938.

Greene-Kelly,  R.,  1955, Sorption of Aromatic Organic Compounds by
Montmorillonite, Trans. Faraday Soc. v. 51, p. 412-424.

Lyon, W.G., 1989,  Precision Density Determination of Soil Organic
Matter and Other Organic - Rich Materials, RSKSOP-105.

Lyon, W.G., Rhodes, D.E., Powell, R.M., and Pennington, L., 1990,
Operation of the Leco WR-112 Carbon Analyzer  for the  Determination
of Carbon in Peats and Other High Carbon Materials,  RSKSOP-101.

Olejnik, S., Aylmore, L.A.G.,  Posner,  A.M., and  Quirk, J.P., 1968,
Infrared Spectra of Kaolin Mineral - Dimethyl Sulfoxide  Complexes,
J. Phys. Chem. v.  71(1), p. 241-249.

Olejnik, S.,  Posner,  A.M.,  and Quirk,  J.P-,  1968,  Swelling of
Montmorillonite in Polar Organic Liquids, Clays and  Clay Minerals
V. 22, p. 361-365.

Starkey, H.C., Blackmon, P.O., and Hauff, P.L., 1984, The Routine
Mineralogical  Analysis  of Clay-Bearing Samples,  U.S.  Geological
Survey Bulletin 1563, U.S. Govt. Printing Office, Washington  D.C.

Thompson, J.G., and Cuff,  C.,  1985,  Crystal Structure of Kaolinite

-------
                                            SUPPLEMENT 7
                                            RSKSOP-104
                                            Revision No. 1
                                            Date:  March 25, 1991
                                            Page 7 of 22
                                            William G. Lyon
                                            David E. Rhodes

- Dimethylsulfoxide  Intercalate,  Clays and Clay  Minerals  v. 33,
490-500.

Weast,  R.C.,  1984,   CRC  Handbook of  Chemistry and  Physics. CRC
Press, Inc., Boca Raton,  FL, p. F8.

-------
                                            SUPPLEMENT 7
                                            RSKSOP-104
                                            Revision No. 1
                                            Date:  March 25, 1991
                                            Page 8 of 22
                                            William 6. Lyon
                                            David E. Rhodes
                            APPENDIX I
             Solvents for Obtaining Swelling Spectra
    Solvent
n-Pentane
n-Heptane
Methy1eye1ohexane
Cyclohexane
p-Xylene
Toluene
Ethyl Acetate
Benzene
Tetralin
Acetylacetone c
Chlorobenzene
Dichloromethane
Acetone
Carbon Bisulfide
1,4-Dioxane
Nitrobenzene
3-Methyl-l-butanol
1-Octanol
Pyridine
Morpholine
N,N-Dimethylacetamide
1-Pentanol
Nitroethane
1-Butanol
2-Propanol
Acetonitrile
1-Propanol
Dimethylsulfoxide
N,N-Dimethy1formamide
Nitromethane
Ethanol (99.9%)
Propylene Carbonated
Methanol
1,2-Ethanediol
1,2-Propanediol
N-Methy1formamide
Formamide
Water

-------
                                            SUPPLEMENT  7
                                            RSKSOP-104
                                            Revision No.  1
                                            Date:  March  25,  1991
                                            Page 9 of 22
                                            William 6.  Lyon
                                            David E. Rhodes
                       NOTES TO APPENDIX I
8The delta  values tabulated here are mostly the simple solubility
parameters of the liquids at 25°C tabulated by Barton, 1983;  in  a
few  instances  where  these  were missing,  the  total solubility
parameter,  St,  from the same reference was used instead.
     values  for  the molar volume  (cm/mol)  were in  most  cases
computed from molecular weights (g/mol)  and density (g/cm3) values
for the liquids at 25°C tabulated by Barton, 1983.  A  few missing
values  were computed  from  similar data  tabulated  in standard
handbooks elsewhere (e.g. ,  Weast, 1984)

C2 , 4-Pentanedione

dl,2-Propanediol cyclic carbonate

     CH2-0

           c=o
     !     /
     CH?-0
     I  '

     I
     CH,

-------
                                             SUPPLEMENT  7
                                             RSKSOP-104
                                             Revision No.  1
                                             Date:  March  25,  1991
                                             Page  10 of  22
                                             William G.  Lyon
                                             David E. Rhodes
                          APPENDIX  II.

       Correction for the Effect  of Non-Swelling  Diluents

     If the sample contains  a  small amount  of  a non-swelling
diluent such as sand, a  simple correction can  be  made  to obtain
the swelling for pure organic  matter,  Q°.   This simple correction
will not be sufficient if clays such as  montmorillonite or
kaolinite are  present;  clays such as these can form crystalline,
interlayer complexes with certain polar  organic solvents.

Notation:   Qnet measured swelling,  impure organic matter

            Q°  swelling of pure organic matter

            V^ volume of organic  matter  in  sample

            V^ volume of mineral  matter  in  sample (assumed
                  inert)

            z^ volume-fraction of organic matter  in sample

            z  volume-fraction of mineral matter  in sample
             rorn

                 on    mri

            Qnet is defined as:


                      (Volume Swollen Sample)
                     (Volume Unswollen  Sample)
            Therefore,


                              (O °*V  + V )
                              \<^ * V - -r t'm-y
                                    +  V )
                                  oin    mm'
          noting that:

-------
                                            SUPPLEMENT 7
                                            RSKSOP-104
                                            Revision No. 1
                                            Date:  March 25, 1991
                                            Page 11 of 22
                                            William 6. Lyon
                                            David E. Rhodes
                                  *'on
                               (V  + V  )
                               v vom  vmn'
                                om
          and
                            _
                                 om   ram'
          allows the simplification:
In  other words,  the  observed  swelling  is  a  volume-fraction,
weighted-average of the swelling of the minerals (Q^ assumed =
1.000) and of the organic matter.  Hence, the correction for non-
swelling diluents is given by the formula,
                                  zom
     For a truly inert ingredient such as  a small amount of quartz
or feldspar sand estimates of z^ can be based on determinations of
"ash content" of the  sample  ignited in air plus estimates of the
mineral grain density and organic matter density.   More complex
calculations are required to  correct ash content to mineral matter
content when thermally unstable minerals such as calcite, opal or
gypsum  are  present;  in  these  cases,  the  ASTM  methods  and
approximations used for coal calculations seem appropriate. Powder
x-ray  diffractograms  of  the organic  matter  and ash  are useful
guides to the correction procedure.

-------
                                            SUPPLEMENT 7
                                            RSKSOP-104
                                            Revision No. 1
                                            Date:  March 25, 1991
                                            Page 12 of 22
                                            William 6. Lyon
                                            David E. Rhodes

                          APPENDIX III.

  Corrections for the Effect of Swelling Clay Mineral Diluents

     If the sample  is an organic-rich soil rather than a
concentrate, the corrections for swelling of mineral species
become a major problem.  Both kaolinite and smectite minerals are
capable of swelling in polar organic solvents; see papers by
Greene-Kelly, 1955, Olejnik et al.. 1968, 1974, and Thompson and
Cuff, 1985.  Two approaches seem feasible and need further
experimental investigation.

     The first correction method requires having volume-fraction
estimates for each  swelling mineral and for all inert minerals
(lumped together);  these could be obtained from careful x-ray
diffraction analysis.  Additionally, swelling spectra would be
needed for pure representatives of each swelling mineral.
Expanding the formula of Appendix II yields the equation,
that can be rearranged to give a formula for Q°.

     A second alternative would be to use chemical treatment with
dilute acid to remove any carbonate minerals (non-swelling)
followed by hydrogen peroxide to remove the organic matter.  A
separate swelling spectrum could then be obtained directly from
the residual material that might be representative of the
lumped-together swelling clays and  inert  minerals present in the
sample.

     Several difficulties should be mentioned concerning both these
methods.    Quantitative  x-ray  diffraction analysis  is  highly
dependent on the  standards  used  for the pure clay minerals; clay
minerals found in soils  are usually not as well-ordered in their
crystals as the standards and may differ substantially from the
standards in their composition as well.  Moreover, organic matter
frequently must be removed from soil samples to get a clean x-ray
diffractogram.  (See Starkey et al., 1984).

     The chemical removal of organic matter by oxidation presents
three problems:  formation of oxalate minerals from any soluble
calcium minerals present (such as calcite or gypsum), chemical
alteration of the swelling character of the clays and removal of

-------
                                            SUPPLEMENT 7
                                            RSKSOP-104
                                            Revision No. 1
                                            Date:  March 25, 1991
                                            Page 13 of 22
                                            William 6. Lyon
                                            David E. Rhodes

organic coatings that can limit the  swelling of clay minerals in
their native, mixed state.   The first of these problems can be
greatly lessened by an acid treatment to remove carbonates etc. ;
however, this acid wash procedure  inevitably alters the exchange
cation composition of the clays, and  that, in turn, changes their
swelling behavior in polar solvents.   This effect makes some sort
of standardization of the exchangeable cations on both the
original sample  (containing organic  matter)  and the acid-washed,
oxidized sample a practical necessity;  an exchange to the sodium
form following acid washing is provisionally recommended.

     Chemical oxidation of clay minerals also can occur during
peroxide destruction of organic matter.  This can occur when
ferrous iron is present in a clay,  for example, substitution of
Fe+3 onto some Al+3 sites of a dioctahedral smectite.  Oxidation of
this ferrous iron to ferric causes a decrease in cation
exchange capacity, and thus an alteration of solvent swelling
characteristics.  Fortunately, smectite minerals in the oxic zone
of surface sediments are typically completely oxidized already;
the main difficulty is expected from anoxic  sediments or soils.

     The removal of organic coatings from the  clay minerals in a
sample activates them and produces  a  material more likely to swell
in organic  solvents  than the  minerals in the  original mixture.
Naturally, corrections based  on the swelling  of  such cleaned-up
mineral matter  will  be too large;  however,  this  is  probably no
worse than corrections based on swelling properties of pure mineral
standards.

-------
                                            SUPPLEMENT  7
                                            RSKSOP-104
                                            Revision No.  1
                                            Date:  March  25,  1991
                                            Page  14 of  22
                                            William 6.  Lyon
                                            David E. Rhodes

                          APPENDIX  IV.

  Examples of Calculations  on  a  Swelling spectrum for Cellulose


     A.  Qraw versus Delta

     This  figure  displays  the  raw  volumetric  swelling versus
     solvent solubility parameter.

     B.  Q     versus Delta
     The  raw swelling data were adjusted by a factor so that the
     background  swelling  for  non-swelling solvents  averaged to
     unity  (i.e., no  swelling).

     C.   0,0^ versus  Delta

     The  scaled  swelling data  minus one1 were  divided  by the
     solvent molar volume so  that  the units  are now  moles of
     solvent per  cm3 of solid sorbent.

     D.   X^ar versus  Delta

     The values for Q^^,. were multiplied by the molar volume of the
     cellulose  repeat unit  (ca. 103  cm3 mol"1)  so that the  final
     value has units  of moles of solvent per mole of  repeat  unit.

     E.  (g sorbed liquid/g sorbent) versus Delta

     Densities  of liquids and of the  solid sorbent (1.57 g cm"3)
     were used to convert the scaled swelling  values (Q-l)  to more
     typical  gravimetric  sorption units:  g  sorbed liquid  per  g
     solid sorbent.
     1Note that the quantity Q-l represents volume of sorbed liquid
per volume solid sorbent.

-------
     1.9
                        RffiH VOLUMETRIC SHELL!H6 vs=  DELTA
                                CELLULOSE? 30 dege  C
             T	1   I    1
                                                                  1	1	T
(D
Z
M
J
J
LU
Z
0)

o
M
c
h
u
E
D
z

-------
     2.2
CD
I
M
J
J
UJ
z
(0

0
M
o:
E
D
J
0
0
111
J

-------
h
I
UJ
ft
tL
0
(0
\
h
z
111
:>
j
o
0)

w
UJ
ft
M
ft
E
(0
UJ
J
o
E
       e.w
       e.e3
6.02
e.ei
                     MOLES  IMBIBEB SOLVENT PER  GRAM SORBEHT

                                     vs. DELTA
               T	1	T
                                                              1	1	T
o
E
O
       -e.ei
            14
                   24
44
                      DELTft, SOLVENT SOLUBILITY PARAMETER
                                                                       54

-------
                                Xmohr vs. DELTA

                              CELLULOSE, 38 deg.  C
       3.5
                       1   I
                             I   I   i  r
                                              i   i   i
                                                         i   i   i   i
Z
D

LI
0)
o
o
D
J
CD
 0
 E
 \
h
Z
HI
:>
J
o
(0
0
E
2.5
1.5
       6.5
o
E
X
      -6.5 — •:••
                    J	I
                             I   I  I   I
                                           1   I   I   I
                                                         I   I   I   I
                                        34
                                               44
                     DELTA,  SOLVENT SOLUBILITY PARAMETER
54

-------
4*
c
i.
0
(0
\
u
•H
•H

J
II
L
0
(0
      -e.2 -
           14
                    (g  Sorbed Liquid/g  Sorbent) vs. DELTA

                            CELLULOSE,  38 deg. C
                            I   I  .   ,   I
24
34
44
                                                                   54
                    DELTA,  SOLVENT SOLUBILITY PARAMETER

-------
                                            SUPPLEMENT 7
                                            RSKSOP-104
                                            Revision No. 1
                                            Date:  March 25, 1991
                                            Page 20 of 22
                                            William G. Lyon
                                            David E. Rhodes
                           APPENDIX V.
Consideration  of  the Different  Standard Compaction  Conditions
needed for Dry and Wetted Powder Columns:  Buoyancy Effects

     In this section we consider the conditions of compaction of a
column of  powder spun in  a centrifuge in both  dry  and solvent-
wetted conditions.  We specifically  ignore:

     1.  any effects  of particle-to-particle adhesion,

     2.   any variable effects of particle  "lubrication"  by the
     different solvents, and

     3.  compaction re-bound following removal of  the powder column
     from the centrifuge.

These possibly important additional effects do not lend themselves
to simple analysis at present.

     The reality of adhesive forces  seems proved by the fact that
dry cellulose powder  always is under-compacted compared to wetted
cellulose powder in non-swelling liquids,  yielding apparent values
of swelling less than unity. Without some framework for including
(or  even  estimating) these  effects,   we cannot  wholeheartedly
recommend  the corrections  for  buoyancy, derived  below,  as  an
improvement upon the  raw volumetric  swelling values.
     The compaction of a powder column  achieved at a given angular
velocity on  a  centrifuge when the  free  pore space  is air-filled
should be different from  the state of compaction when the free pore
space  is  liquid-filled.   The  generally  smaller  effective stress
causing  compaction in the  liquid-saturated case  is  due  to the
substantial buoyant force on the particles exerted by the liquid.

Calculation of the Buoyant Force Contribution

     For a particle of powdered material at the top of a column of
solvent-wetted powder, the net compacting force on  the particle
will depend  on  the  densities  of  solid  and liquid,  and  on the
effective  "gravitational"  force  constant,  g,   which  will  be  a
function of the angular velocity of  the tube, and the distance of
the particle from the axis of rotation of the centrifuge.

     Noting that g = g(o>,R)  = Ru2, where  w =  angular velocity, and
R = distance from axis of rotation, we may write for the net force:
                                     12-7

-------
                                            SUPPLEMENT 7
                                            RSKSOP-104
                                            Revision No. 1
                                            Date:   March 25, 1991
                                            Page 21 of 22
                                            William G. Lyon
                                            David E. Rhodes

Rearranging this expression slightly,  yields:
                                      ro
where F   = net force on a particle of mass nu and density  p0 when
immersed in  a  liquid of density pL.   Strictly  speaking,  samples
should be spun at different angular velocities to achieve the same
packing when the density of the liguid varies.  Similarly, when the
pore space is  filled with  air, the angular velocity should be at
its lowest to account for the very low buoyant force from air.

Derivation when Swelling Occurs

     The above derivation assumed that no swelling of the particles
in the liquid  occurs.  Even  with the assumption of swelling, the
same formula for net force  holds approximately due to cancellation
of terms  involving the  swelling,  Q,  as  long  as  the AV  of the
swelling process can be taken as approximately zero.  That is, the
mass of the imbibed liquid in a swollen particle of original mass
m0 can  be  taken exactly as:

                                                     p
   Mass of Imbibed Liguid inside Swollen Particle = mQ —- (Q - 1)
                                                     Po

Correction to Conditions of Equivalent Applied Force

     Correction  of  swelling  measurements to a  standard state of
equivalent applied force should be possible with  additional  data on
powder compaction at different angular velocities.  We have  taken
some preliminary data indicating that the effective swelling of a
column of  solvent wetted powder is  approximately  an exponential
function of  the angular velocity,  <•>,  over the 1600  to 4000 rpm
range on our centrifuge.  That is,

-------
                                             SUPPLEMENT  7
                                             RSKSOP-104
                                             Revision  No.  1
                                             Date:  March  25,  1991
                                             Page 22 of  22
                                             William G.  Lyon
                                             David  E.  Rhodes
                          ln(Qaff) = A + BCD
Here,  the heights of a wetted powder column were each divided by
the height of the corresponding dry powder column compacted at the
standard  angular velocity to  yield  effective swelling values at
different angular velocities.

     Operationally, one would take column height data on the wetted
powders at increasingly higher angular velocities  (starting with
the standard velocity of 1800 rpm)  up  to the velocity needed to
overcome  the buoyant force, and then apply a correction to a final
standard  state for each solvent.   If the standard angular  velocity
is given  by o)0,  then the adjusted angular velocity,
solvent on a given sorbent would  be  given by:
0)
    for each
              Adjusted Angular Velocity, GO" =
                                               1--
                                                  Po
From these  data,  an adjusted value of the swelling, Qadj. could be
computed by the expression:
                                           O)n)
where   B   represents    a   parameter   describing   compaction.
Unfortunately,  it appears  from our preliminary  tests  that  each
solvent which  swells the powder,  in  effect,  produces a  distinct
material with a different B value; hence,  powders  wetted with each
solvent would require a whole  series of measurements  at different
rpm values to estimate B  from  least-squares regression.

-------