PE86-1C9485


                                                     EPA/60Q/1_85/013
                                                     June  1985
AN ISOTOPIC STUDY OF THE INHALATION TOXICOLOGY OF OXIDANTS
             John M0  Hayes  and  Jeffrey  Santrock
                     Indiana University
                   Department of Chemistry
                    Bloomington, IN 47405
      Cooperative Agreements CR807322 and CR811261
               HEALTH  EFFECTS RESEARCH LABORATORY
               OFFICE  r~ RESEARCH AND DEVELOPMENT
              U.S.  ENVlrx  IMENTAL PROTECTION AGENCY
                RESEARCH TRIANGLE PARK, NC 27711

-------
TECHNICAL REPORT DATA
/Please read /ntinicnons un the reverse before completing)
1. REPORT NO. 2.
EPA/600/1-85/013
4. TITLE ANDSUBTITLE
AN ISOTOPIC STUDY OF THE
^INHALATION TOXICOLOGY OF OXIC&NTS
7. AUTHOR(S)
Dr. John M. Hayes, Dr. Jeffrey Santrock
^,
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Department of Chemistry
Indiana University
Bloomington, IN 47405
12. SPONSORING AGENCY NAME AND ADDRESS
Office of Research & Development
Health Effects Research Laboratory
USEPA
Research Triangle Park, NC 27711
3. RECIPIENT'S ACCESSION NO.
rr, ': 1 -; '. G'57/iS
5-R5KReT?9&
6. PERFORMING ORGANIZATION CODE
8. PERFORMING ORGANIZATION REPORT NC
1O. PROGRAM ELEMENT NO.
C9GA1A
11. CONTRACT/GRANT NO.
CR807322 and CR311261
13. TYPE OF REPORT AND PERIOD COVERED
14. SPONSORING AGENCY CODE
EPA-600/11
is. SUPPLEMENTARY NOTES
16. ABSTRACT
The purpose of  these studies was to develop novel methods to investigate the biological
fate of  inhaled ozone and other oxygen-containing pollutants in animal and human tis-
sues using the  heavy isotope of oxygen, oxygen-18 (180).   Methods were developed which
facilitated the conversion of tissue oxygen to CO2 and the subsequent trapping of the
(X>2 so that it  could be subjected to isotope-ratio mass spectranetry.  The ratios of
the various masses of evolved COj were used to calculate the ±JO content of the orig-
inal .tissues, thus enabling the detection of isotopic enrichments as small as 0.4%.
The above procedures were performed by modification of a commercially available ele-
•».antal analyzer to include effluent columns and trapping devices, development of oxygen
isotopic standards, and by derivation of mathematical models for correction of blank
and memory effects originating during sample pyrolysis.    These techniques were applied
with success  to the determination of the biological fate of inhaled ozone, and to the  !
measurement of  tissue oxidation induced by  a model peroxidation initiator, carbon      I
tetrachloride.
17. KEY WORDS AND DOCUMENT ANALYSIS
». DESCRIPTORS

18. DISTRIBUTION STATEMENT
Release to Public
b. IDENTIFIERS/OPEN ENDED TERMS

19. SECURITY CLASS (This Report)
Unclassified
20. SECURITY CLASS (This page/
Unclassified
c. COSATI Field/Croup

21. NO.OR.PAGES
22.PHICE
CPA Ftrm 2220-1 (R»«. 4-77)   PREVIOUS EDITION is OBSOLETE

-------
                                                        PE86-1C9485


                                                     EPA/60Q/1_85/013
                                                     June  1985
AN ISOTOPIC STUDY OF THE INHALATION TOXICOLOGY OF OXIDANTS
             John M0  Hayes  and  Jeffrey  Santrock
                     Indiana University
                   Department of Chemistry
                    Bloomington, IN 47405
      Cooperative Agreements CR807322 and CR811261
               HEALTH  EFFECTS RESEARCH LABORATORY
               OFFICE  r~ RESEARCH AND DEVELOPMENT
              U.S.  ENVlrx  IMENTAL PROTECTION AGENCY
                RESEARCH TRIANGLE PARK, NC 27711

-------
                      NOTICE

This document has been reviewed in accordance with
U.S. Environmental Protection Agency policy and
approved for publication.   Mention of trade names
or commercial products does not constitute endorse-
ment or recommendation for use.
                       11

-------
     This is to certify  that the thesis submitted by Jeffrey Santrock



has been passed by the Ph.D. Advisory Committee as satisfactory in



partial fulfillment of the requirements for the Ph.D. degree.
                    Chairman
                            CResearch Advisor)
                             QlxruKxtxix
                                     iii

-------
                            TABLE OF CONTENTS

 TABLE OF CONTENTS	     v

.ILLUSTRATIONS AND FIGURES	vill

 ACKNOWLEDGMENTS	    xi



 CHAPTER 1   INHALATION TOXICOLOGY OF OXIDANTS	"...    1

             Inhalation Toxicology .,	    3

             Respirtory Transport and Absorption of Inhaled
             Oxidants	,    6

                Airway Structure 	    6
                Epithelial Cell Population 	    9
                Flow Dynamics  . . . .	   11
                                       4!
             Chemical Mechanisms	   12
                                         f
                Moleculat Targets  	   13
                Ozonolysis	   14
                Aatoxidation	'	   17
                Mechanism of Tissue Damage	   23

             Isotopic Studies 	   24

                Physiological Applications of Oxygen-18  ....   25

             References for Chapter 1	   28

 Chapter 2    DETERMINATION OF OXYGEN-18 IN ORGANIC SUBSTANCES .   33

              Oxygen Isotopic Standards  	   35

                 Pyrolysis of Benzole Acid	   36
                 Isotopic Standards	   46

              Mathematical Model	   46

                 Chemical Processes  	   46
                 Isotopic Mass Balance	   52
                 Correction Parameters 	   57
                 Correction of Analytical Results  	   65

              Oxygen Isotopic Analysis 	   66

                 Yield of Carbon Dioxide	   69
                 Correction Parameters 	   69
                 Evaluation of the Correction Parameters  ....   78
                 Standards	   85

-------
             Experimental Section 	 	   89

                 Isotopic Standards 	   89
                 Oxygen Isotopic Analysis 	   93

             References for Chapter 2	   98

Chapter 3    MASS SPECTROMETRIC DETERMINATION OF CARBON-13,
             OXYGEN-17, AND OXYGEN-18 IN CARBON DIOXIDE  .....  101

             Development of Isotope Ratio Equations 	  103

                Isotopic Distribution 	  103
                Relationships Between 17R and iaR	106

                   Theoretical Considerations 	  .  106
                   *'R and *'R in Natural Samples	110

             Calculation of Isotope Ratios   	  Ill

             Determination of Isotope Abundances  	  112

                Mixed Samples Enriched in ^0	113
                17R and »»R in Natural Samples	117

             Conclusions	123

             Experimental Section 	  124

                Preparation of Carbon Dioxide for Isotopic
                Analysis	124
                Mass Spectrometric Measurement	124

             References for Chapter 3	127

CHAPTER 4    DEPOSITION, DISPOSITION, AND CLEARANCE OF
             OZONE-DERIVED OXYGEN IN LUNG	12.9

             Detection and Quantitation of the Isotopic  Label  .  131

             Deposition in the Respiratory System 	  135

                Comparative Aspects of Ozone Uptake 	  135
                Distribution Within the Respiratory System   .  .  138

             Deposition and Clearance in the Lung	143

                Deposition in the Lung	143
                Clearance from the Lung	152

             Experimental Section 	  156

                Apparatus	156

-------
                Preparation and Purfication of ''O,	162
                Aminal Exposures  	  163
                Tissue Sampling and Preparation 	  163
                Elemental and Isotopic Analysis 	  164

             References for Chapter 4	165

CHAPTER 5    DETECTION AND QUANTITATION OF AUTOXIDATION
             IN VIVO;  A STUDY OF HEPATIC AUTOXIDATION INDUCED
             BY CARBON TETRACHLORIDE  	  167

             Lipid Peroxidation	168

                Peroxidation of Linolenic Acid  ........  168
                Measurement of Lipid Peroxidation In Vivo  .  .  .  171

             Carbon Tetrachi.oride Hepatotoxicity  	  172

                Microsomal Cytochrome P-450  	  173
                Carbon Tetrachloride-Induced Autoxidation  .  .  .  174

             Conclusions	183

             Experimental Section 	  184

                Protocol	184
                Tissue Sampling and Preparation 	  184
                Elemental and Isotopic Analysis 	  185

             References for Chapter 5	186

SUMMARY AND CONCLUSIONS 	  189
                              vii

-------
                        ILLUSTRATIONS. AND FIGURES

Figure

 1-1   Schematic representation of the  mammalian respiratory
       system  ....... ..................    8

 1-2   Reaction of ozone with unsaturated fatty acids  .....   16

 1-3   Reaction of nitrogen dioxide with unsaturrated fatty
       acids ....................... ...   19

 1-4   Autoxidation in the cell  ................   21

 2-1   Yield and isotopic composition of carbon dioxide
       produced by pyrolysis of benzoic acid ..........   39

 2-2-  Equilibrium constant and equilibrium isotopic
       fractionation factor for the reaction C02 + C *r± 2CO . .   45

 2-3   Production of C02 from organic material ....... . .   53
                                        i :
 2-4   Results Predicted for replicate  analyses of organic
       samples ............. .-...' ........   59

 2-5   Flow-chart for determination of  correction parameters . .   62

 2-6   Illustartion of the method for refinement of values for
            and n  .......................   68
 2-7   Effect of IjOj temperature on yield of carbon dioxide . .   71

 2-8   Determination of HQ and JiFQ ...............   75

 2-9   Determination of a and b  ............. ...   77

 2-10  Sample matrix effects ..................   84

 2-11  Calibration of working standards  ......... ...   91

 2-12  Diagram of modifications to elemental analyzer  .....   95

 3-1   Schematic representation of isotope ratio mass
       spectrometer   ......................  108

 3-2   Relative error in »»R as a function of oxygen-18
       abundance ........................  119
 3-3   Relative error in  "F as a function of oxygen-18
       abundance  ........................  121

 4-1   Distribution of ozone-derived oxygen in the respitatory
       system   .........................  142
                                 viii

-------
 4-2   Uptake of ozone-derived oxygen in the lungs	146

 4-3   Molecular distribution of ozone-derived oxygen in the
       lungs	148

 4-4   Clearance of ozone-derived oxygen from the lungs  ....   155

 4-5   Apparatus for preparation and purification of 1§03  . .  .   158

 4-6   Exposure system	161

 5-1   Products formed during peroxidation of linolenic
       acid	170

 5-2   Incorporation of Oa-derived oxygen into liver 	   181

Table

 1-1   Isotopes of oxygen	    26

 2-1   Effect of pyrolysis time and temperature on yield and
       isotopic composition of carbon dioxide  	    37

 2-2   Recovery and isotopic composition of carbon dioxide
       heated to 550°C in quatrz tubes 	    42

 2-3   Analysis of benzoic acid	    47

 2-4   Analysis of heteraromatic carboxylic acids   	    48

 2-5   Iterative determination of correction parameters  ....    73

 2-6   Comparison of isotopic analysis of benzoic acids by
       pyrolytic decarboxylation and by the Schutze-
       Unterzaucher procedure   	    81

 2-7   Comparison of isotopic analysis of heteroaromatic
       carboxylic acids by pyrolytic decarboxylation and by
       Schutze-Unterzaurher procedure  	   86

 2-8   Analysis of water standards and sucrose samples  	   88

 3-1   Isotopic molecular species of carbon dixide  	  104

 3-2   Analysis of CO, with oxygen derived from exchange with
       mixed water	115

 3-3   Analysis of CO, with oxygen derived from exchange with
       distilled water	122

 4-1   Contributions to the variance in »»F	134

 4-2   Deposition of ozone-derived oxygen in  the  respiratory
       system	136

                                  ix

-------
4-3   Oxygen pools in nasopharynx,  trahea,  lungs, and blood .  .  137

4-4   Uptake of ozone-derived oxygen in the respiratory
      system	140

4-5   Deposition of ozone-derived oxygen in the lungs	144

4-6   Mouse-lung oxygen poolr>	151

4-7   Clearance of ozone-derived oxygen in mouse lung . . . .  .  153

5-1   Summary of experimental protocol,  	  175

5-2   Abundance of oxygen-18 in liver	177

5-3   Relat've enrichment of oxygen-18 in liver 	  178

5-4   Oxygen pools in liver	179

-------
                            ACKNOWLEDGEMENTS

     My sincere appreciation goes to John Hayes,  his commitment ani

guidance made this thesis possible.

     Dr. Edward Baer provides much advice concerning the chemistry

and preparation of ozone.

     The technical support in our lab is outstanding.  Steve Studley

had a posative influence on aiy research in many,  innumerable ways.

A special thanks for the conscientious analysis of tissue samples, and

tissue samples, and tissue samples ... by David Andrews, David Archer,

and Martha J. Lars en is also well deserved.

     Many of the specialized pieces of equipment required in these
                                        4 :
studies were designed and constructed by the technical staff in the

department.  The glass blower, Don Fowler/ spent many hours building

rebuilding apparatus.  Each of the members of the machine shop, John

Dorsett, Kenny Bastin, Dick Martin, Bill Miller,  Larry Sexton, and Don

Swafford, contributed to this work.  Several devices had their origin

in the electronics: Bob Ensman, Steve Williamson, and Andy Quails.

     This work was supported by a cooperative agreement with the

Unites States Environmental Protection Agency.  The physiological

studies were conducted at the laboratories of the Inhalation

Toxicology Division of the EPA.  Their support and collaboration made

this work possible.  A special thanks must go to Dr. Gary E. Hatch for

his efforts.  He initiated this project, provided the necessary

administartive support from within the EPA, and had an equal and ever

present hand in these studies.

-------
            CHAPTER 1
INHALATION TOXICOLOGY OF OXIDANTS

-------
                                                                       2


     Oxidant gases, specifically ozone (Oa) and nitrogen dioxide (NOj),


are the most important toxic components of photochemical smog.  Toxicity


arises from oxidation of components of biological tissue.  Upon


inhalation, the chemice.1 lesion occurs in the respiratory system, at the


interface between inspired air and blood in the lung.  Damage to tissue


can occur by two processes:  direct reaction of the inhaled oxidant with


molecular components of the cell and free radical-induced oxidation


(autoxidation) of cellular components.  Although autoxidation is a


secondary process, initiated by free radicals produced in the initial

                                  4!
chemical insult, it is thought to account for the majority of damage to


tissue.  Moreover, a complex series 'of biochemical, physiological, and


anatomical events in the lung ensues, and it is difficult to distinguish


dysfunction resulting from oxidation from the response of the organism.


     A major emphasis of current research is the establishment of


relationships between effects in' laboratory animals and in humans.  A


realistic assessment of the risk to humans must involve, not only an


understanding of the chemical mechanisms of oxidant injury to tissue,


but an understanding of the chemical and physiological factors which


influence transport, absorption, and reactivity of oxidants in the


respiratory system.  A unified approach to the study of the inhalation


toxicology of oxidants utilizing cxygen-18 as a physiological tracer of


inhaled oxidants will be presented.  This approach relies on techniques,


developed in this work, for the quantitation of excess oxygen-18 in


complex organic material (physiological samples).  Two questions of


immediate concern are addressed directly:


     1. What is the retained dose of a specific oxidant, both in the


lung and at specific sites within the respiratory system?

-------
                                                                       3



     2. What are the target molecules for inhaled oxidants in the lung



and the chemical mechanisms of oxidant injury in tissue?



     This approach can be used to determine the influence of various



biochemical and physiological factors on the retained dose and oxidant



injury in the lung, and to provide a sound basis for extrapolation of



dose-response relationships frcj laboratory animals to humans.  Although



the primary focus of this research has been study of the inhalation



toxicology of photochemical oxidants, the experimental protocol



developed in the course of these investigations is generally applicable



to the study of any agent or pathological condition that causes aberrant



oxidation of the molecular components of tissue.





                          Inhalation Toxicology



     The physiological function of the lung is to provide a medium for



gas exchange between inspired air and the blood, while maintaining the



integrity of the organism.  Any process which impairs the ability of the



lung to mediate exchange of respiratory gases is detrimental.  In the



case of inhaled oxidants, pulmonary distress results mainly from tissue



damage in the respiratory regions of the lung.



     An important conclusion, derived from cytologic and biochemical



observations of lung tissue, is that the peak concentration of an



oxidant in the atmosphere, as opposed to exposure time or the time-



weighted-average concentration, correlates most closely with changes in



lung structure and biochemistry that are associated with toxicity (1).



Although it is difficult to establish a distinction between an acute and



a chronic dose of a particular oxidant or, more precisely, the



concentration which elicits an acute or a chronic response, the oxidant



levels typically found in highly polluted urban areas (ozone < 1 ppm and

-------
                                                                       4



nitrogen dioxide < 0.5 ppm) are well within the chronic range, except



for a small group of high-risk individuals (2).  Exposure to high



concentrations of oxidants for prolonged periods of time results in



extensive destruction of epithelial tissue in the lung and can be lethal



(3).  Symptoms of acute toxicity include chest pains,  stvere respiratory



distress, and accumulation of edemata in the lung (4).  If the exposure



is not lethal, prolonged structural and biochemical changes are observed



(5).  Since chronic toxicity is most prevalent, however less obvious, in



the general population, the effects of low levels of inhaled oxidants



are currently of greatest interest.



     Chronic effects encompass a complex sequence of events.  Exposure



to sub-lethal concentrations of ozone or nitrogen dioxide results in an



increase in breathing frequency with decreases in minute volume, total



ventilation, and oxygen consumption (6).  This rapid,  shallow breathing



is accompanied by an increase in pulmonary resistance and a decrease in



pulmonary compliance (6).  Respiratory function is impaired and



stressed.



     Susceptibility to infection is also increased.  Increased risk of



respiratory infection is a result of decreased phagocytyzing capacity of



the alveolar macrophage (7).  This effect has been observed in mice



exposed to concentrations of ozone as low as 0.08 ppm (7).  Furthermore,



the immune response of T and B lymphocytes (8) and the phagocytic rate



of polymorphonuclear leukocytes (9) are suppressed in humans exposed to



low concentrations of ozone.  These effects persist for at least two



weeks after the exposure (1-9).



     Decreases in the activities of enzymes which are particularly



susceptible to attack by oxidants, such as membrane proteins and

-------
                                                                       5

sulfhydryl-containing enzymes,  are observed initially (10-14).   As

regenerative processes begin,  biosynthesis increases, as do the

activities of typical microsomal and mitochondrial marker enzymes

(10,11).  A certain degree of  adaptation to oxidant injury occurs upon

repeated exposure (12,115), but prolonged changes in lung structure

(16,17), cell population (18),  enzyme activity (13,14),  and pulmonary

function (19) occur.  The degree to which the organism is affected is,

in part, a function of the dose of oxidant to tissue.  Tissue damage is

greatest in the centriacinar*  region of the lung (20-22) and, therefore,

exchange of respiratory gases  is greatly impaired.

     Attempts to assess the dose of inhaled oxidants to specific regions

within the respiratory system have relied on mathematical models which

account for factors controlling the transport and absorption of reactive

gases in the respiratory system (23).  Though predictions of these

models (24) are generally consistent with observations that the major

sites of damage are the respiratory bronchioles and the alveolar ducts

(20), direct experimental verification of the predicted dose is lacking.

Since, in principle, such models can be used to extrapolate doseresponse

relationships from laboratory animals to humans, it is imperative that

direct measurements of the retained dose in tissue be made and that the

mechanisms of respiratory transport and absorption of inhaled oxidants

be understood.
 •The centriacinar region collectively refers to the structural units in
the lung where exchange of gases between inspired air and blood occurs.
See pages 6-9.

-------
                                                                       6




                   Respiratory Transport and Absorption




     The transport and absorption of  gases such as  nitrogen and oxygen




in the respiratory system is well understood (25).   In contrast,




theories which describe the behavior  of  low concentrations of highly




reactive gases in the respiratory system are not well developed, even




though this issue is of central impor-ance to understanding the toxic




effects of inhaled pollutants.  The dose of a reactive gas to various




regions of the respiratory system and the resulting tissue damage depend



on airway structure, the cellular morphology of the lung,  and dynamics



of gas flow in the respiratory system.




     Airway Structure.  Conceptually, the respiratory system can be



divided into three functional zones:  a conducting zone, a transitory




zone, and a respiratory zone (26,27).  The conducting zone comprises the




nasal pharynx, the trachea, the bronchi, and the bronchioles, and



functions to distribute inspired gases into the lower lung.  The




transitory zone forms the link between the conducting zone and the



respiratory zone, and consists primarily of terminal and respiratory




bronchioles.  The respiratory zone, consisting of alveolar ducts and the




alveoli, is the site of exchange of gases between inspired air and the




blood.  A schematic representation of the lung is shown in Figure 1-1.



     The airways in the respiratory system form a series of bifurcated




tubes terminating in small sacs in the periphery of the lung.  Each




airway is approximately cylindrical with flattened ends in the dimension




perpendicular to the plane of the bifurcation (28).  Systematic




measurements of the size and branching angles of airways have shown that




the lung exhibits a regular, asymmetric branching pattern  (25-30).  If




the trachea is designated as the "rero-th generation" in the sequence of

-------
      Figure 1-1.  Schematic representation of the mammalian lungs •





     The conducting zone consists of the the trachea,  bronchi (br), and



bronchioles (bl).  The transitory zone consists of terminal bronchioles



(tbr) and respiratory bronchioles (rbl).  The respiratory zone consists



of the alveolar ducts (ad) and alveoli (ad).  Each bifurcation of the



airway is assigned a generation number N, where the trachea is



designated as the "zero-th generation" (26,27).  Human lungs typically



contain 23 generations between the trachea and the alveoli.

-------

-------
                                                                       9




bifurcations, the mean length and radius of each airway decreases




exponentially with generation number,  whereas total lung volume and




total surface area increase exponentially with generation number (24).




Although variations in the branching pattern, the size, and the geo-




metric shape of the conducting airways, thought to be a consequence of



evolutionary adaptation, are observed  in different species (29), lung




structure is similar in mammals.  Body mass is the single most important




factor which determines morphometric and physiological respiratory




parameters in mammals.  Linear relationships between body mass and total




lung volume, total lung weight, compliance, resistance, diffusing



capacity, oxygen consumption, airway diameter, and alveolar surface area



have been established (25).  The systematic morphometry of the mammalian




lung has been utilized in the development of a single convection-



diffusion model to describe the mass transport of inspired gases in the




lungs of mammalians (23).



     Cell Population.  Cell types with highly specialized functions are




found in epithelial tissue of the respiratory system.  The epithelium of




the conducting airways contains ciliated and nonciliated cells, and




receptor cells (25).  The nonciliated  cells synthesize and secrete onto



the surface of the epithelium an aqueous fluid which becomes the mucous




layer.  Mucus is cleared through the epiglotus into the gastrointestinal




tract by ciliary action.  This is a major mechanism by which particles




and aerosols are removed from the lung (30).  The function of the




receptor cells is not entirely clear.   It is suspected that they serve



as neuronal receptors for irritants.




     Transition to a different population occurs in the alveolar ducts.




The alveolar epithelium is composed of three cell types: Type I, Type

-------
                                                                      10



II, and Type III pneumocytes (25).   The majority of the alveolar



epithelium consists of type I cells.  The function of these cells is to



form the boundary between the alveolar lumen and the capillary endo-



thelium while contributing minimally to the barrier for diffusion of



gases between blood and inspired air.  Type I cells extend laterally, up



to 50 urn, from a single nucleus to form a thin cytoplasmic layer (0.1



vat) populated with relatively few organellas (25).  Type II cells are



cuboidal with no lateral cytoplasmic extensions.  They have a high



concentration of mitochondria, a well-developed endoplasmic reticulum, a



Golgi complex, and numerous multivesicular bodies (25).  These cells



account for most of the metabolic activity of the lung, their main



activity being the synthesis and secretion of a phospholipid surfactant



onto the surface of the alveolar epithelium.  Type III cells are very



rare and their importance to lung function is unknown.  It has bean



proposed that their characteristic brush microvilli extending into the



alveolar lumen serve as neuronal receptors, in a fashion similar to the



receptor cells found in the bronchi.



     A consequence of the low metabolic activity and large surface area



of the Type I alveolar epithelial cells is a high susceptibility to



damage by inhaled oxidants (11).  Cell damage progresses from severe



cytoplasmic swelling to a complete disintegration of Type I cells, thus



exposing the basement membrane of the intersitiuro (20).  Early stages of



tissue repair proceed by a proliferation of Type II cells, which



apparently give rise to the normal population of Type I cells (17).



Prolonged exposure to oxidants can result in a permanent increase in the



number of type II cells (16,17), together with fibrous intrusions into



the alveolar lumen, thickening of the respiratory bronchiolar and

-------
                                                                      11



proximal alveolar walls, and partial or complete occlusion of the



alveolar ducts (17,21).



     Flow Dynamics.  In cases of chronic exposure,  localization of



injury primarily to the centriacinar region is thought to be, in part, a



consequence of the flow characteristics of gases in the lung.  The



driving force for the transport of gases in the respiratory system is a



pressure gradient created by expansion (inhalation) and contraction



(exhalation) of the pleural cavity.  Convective flow predominates in the



conducting airways.  A transition from convection to diffusion occurs as



the flow front penetrates the lower portions of the lung.  Although the



average diameter of lung airways decrease exponentially as a function of



generation number, the total lung volume increases so that the velocity



of the flow front decreases as inspired, air expands into the lung.  Air



in the alveolar ducts and in the individual alveoli is stagnant.  Gases



diffuse along a concentration gradient between the airway lumen and the



capillary lumen (25).  Therefore, mass transport of gases in the lung



has both convective and diffusive components.



     Transport processes almost certainly play an important role in the



expression of toxicity.  The dose of trace reactive gases to the airway



epithelium is determined by the specific modes of transport which per-



sist in different regions of the respiratory system.  If fully developed



laminar flow existed in the conducting airways, then radial diffusion



would be the only mechanism for transport of gases to the airway wall.



Turbulent flow is common in the trachea and in the first several



bronchial generations  (31).  The asymmetric branching structure creates



nonlaminor flow profiles which are  initiated at the bifurcations and



propagated along the daughter branches  (32,33).  Increased transport  of

-------
                                                                      12



reactive gases to the airway wall would be  expected where turbulent flow



occurs and at the branching points,  where non-laminar  velocity fields



predominate.  As velocity decreases, the residence time of a pulse of



inspired air in a single generation  increases.   Radial diffusion becomes



more important.  The transitory and  respiratory regions in the lower



lung are also expected to receive a  high dose.



     Although some protection is provided by tha mucus and surfactant



layers, a large dose may penetrate the airway coating, with a maximum



local deposition to tissue at the bifurcations  and in  the respiratory



bronchi, alveolar ducts, and the proximal alveoli.  This view is



consistent with cytologic observations that cell damage accompanying



inhalation of oxidants is greatest at the branching points and in the



centriacinar region.  Not only is quantitation  of the  dose to the total



lung essential, it is important to assess non-uniform  distribution of



inhaled oxidants in the respiratory  system, as  the local dose to



specific epithelial sites may be disproportionately high.





                           Chemical  Mechanisms



     Studies of the reactions of ozone and  of nitrogen dioxide with



organic molecules in vitro have provided a  substantial framework from



which the present theories of damage to tissue  have been developed.



Although it has been shown that ozone and nitrogen dioxide can react



with a number of compounds found in  the cell, toxicity is thought to be



the result of destruction of the bronchiolar and alveolar epithelium,



associated with autoxidation of unsaturated fatty acids in the cell



membrane (34,35).  There are two essential  features to this scheme.



Initially, ozone or nitrogen dioxide reacts directly with carbon-carbon



double bonds in the fatty acid found in membrane lipids and possibly in

-------
                                                                      13




certain aromatic residues in membrane proteins.   These reactions,  in




turn, generate free radicals which initiate autoxidation in the cell




membrane.  Autoxidation is, however,  not limited to membrane lipids.




Uncontrolled free-radical reactions in the cell  may involve other



oxidizable compounds, causing changes in membrane fluidity and membrane-



bound protein structure, thus disrupting the normal function of the




membrane.  In fact, autoxidation may produce a chemical amplification of




initial oxidative damage caused by the inhaled toxicant, and, in that




way, account for most of the cell damage.



      Molecular Sinks for Ozone and Nitrogen Dioxide.  Studies of  the




reactions of ozone and of nitrogen dioxide with organic solutes and



macromolecules jn vitro have indicated molecular components of the cell




which are susceptible to direct oxidation by inhaled pollutants.  In



aqueous solution, ozone was found to add across the carbon-carbon double



bound in tryptophan and histidine, to oxidize the sulfhydryl group in




cytosine to sulfone, and to cause hydroxylation of the aromatic ring in




tyrosine and phenylalanine (36).  These amino acid residues were also



found to be susceptible to oxidation in proteins; however, the extent of




oxidation was a function of the accessibility of a particular residue to



the solution (36,37).  Cross-linking of sulfhydryl groups in proteins




and glutathione was also observed, and only a portion of the original




biological activity of the oxidized compounds could be restored by



enzymatic reduction  (36-38).  Aqueous solutions of NADH and NADPH were




oxidized by ozone; however, both oxidized species were biologically




active as coenzymes in the reduced form (39).




     Unsaturated fatty acids are particularly susceptible to oxidation




by ozone and nitrogen dioxide.  Oxidation of methyl esters of unsat-

-------
                                                                      14




urated fatty acids occurs in both thin films and aqueous dispersions




(40,41).  The major products of ozonolysis of lipid in the presence of




molecular oxygen indicate that lipid peroxidation had occurred.




Further, addition of phenolic: antioxidants (a-tocopherol, butylated



hydroxyanisole, and butylated hydroxytoluene) decrease the rate of




oxidation (40).  Products of lipid peroxidation have also been detected




in tissue from animals exposed to both ozone and nitrogen dioxide



(42-45).  The mode of initiation and the extent of autoxidation in vivo




is, at present, unknown.




      Direct Oxidation.  The mechanism by which ozone reacts with




carbon-carbon double bonds was first proposed by Criegee (46).  In this



scheme, Figure 1-2, ozone reacts directly with the carbon-carbon double




bond to form an initial adduct, a 1,2,3-trioxylane JT-complex (reaction



1).  At room temperature it is unstable.>and rapidly decomposes to an



aldehyde and a zwitterion (reaction 2).  In a non-participating solvent,




a non-polar solvent lacking active hydrogen, a resonance form of the



zwitterion (reaction 3) can react with the aldehyde to form a 1,2,4-




trioxylane ring system (reaction 4).  However, in a participating




solvent (the cell membrane) the zwitterion can react with any molecule



containing active hydrogen, yielding a peroxide (reaction 5).  Destruc-




tion of the lipid bilayer of the membrane by ozone alone could have




disastrous consequences for the cell; however, the peroxide products of




ozonolysis, including the 1,2,4-trioxylane compound, can initiate ion




autoxidation in the cell producing a chemical amplification of oxidative



damage.




     Nitrogen dioxide reacts with unsaturated lipids by addition to tJ-.e




carbon-carbon double bond or by abstraction an a-hydrogen (47).  The

-------
                                                                          15

IS
         Figure 1-2.  Mechanism for the reaction of ozone with
                             unsaturated fatty acids.
     (1)  reaction of ozone with carbon-carbon double bond to form a
          1,2,3-trioxylane

     (2)  decomposition of 1,2,3-trioxylane to an aldehyde and a Criegee
          zwitterion

     (3)  resonance form of Criegee zwitterion

     (4)  reaction of Criegee zwitterion and aldehyde to form
          1,2,4-trioxylane (molozonide)

     (5)  reaction of 1,2,4-trioxylane with molecular components of
          tissue containing active hydrogen to form a peroxide

-------
                                   16
    =+ 0
 (A)
-c-c-
     o-o-

     C+
     I
 A
 •?-?-
  9-0
-f
                        o
                                 (I)
                                 (2)
               (3)
                                (4)
— C-O-C-R
                                (5)
0-OH

C-S-R

-------
                                                                      17



important steps in both mechanisms are shown in Figure 1-3.   The addi-



tion reaction (Scheme I) predominates at extremely high concentrations



of nitrogen dioxide (>1%), whereas hydrogen abstraction (Scheme II)  is



the predominant mode of reaction at lower concentrations. Addition  of



nitrogen dioxide to the carbon-carbon double bond produces a carbon-



centered radical, reaction 1.   In the presence of molecular  oxygen,  a



carbon-centered radical reacts rapidly to form a peroxy radical



(reaction 2).  A peroxy radical can initiate autoxidation in the cell by



hydrogen abstraction from an unsaturated lipid (reaction 3).  Hydrogen



abstraction by nitrogen dioxide produces a delocalized carbon-centered



radical (reaction 4).  In a similar fashion, this radical reacts with



molecular oxygen to form a peroxy radical (reaction 5) thus  initiating



autoxidation.



      Autoxidation.  Lipid peroxidation in vivo is thought to be an



important component of oxidant injury accompanying exposure  to nitrogen



dioxide (42,43) and to ozone (44,45).  Presumably, carboneentered free



radicals and organic hydroperoxides produced by the reaction of lipids



with nitrogen dioxide and ozone, respectively, initiate autoxidation.  A



general scheme for autoxidation is shown in Figure 1-4.  Initiation



occurs by the production of carbon-centered free radicals in the cell,



where an organic hydroperoxide is the precursor for the initiation



process (reactions 1 and 2).  At low conversions, lipid peroxidation in



the membrane produces radical species which sustain the autocatalytic



cycle (48-50).  Propagation of the chain of free-radical induced oxida-



tions involves the reaction of the carbon-centered radical with



molecular oxygen to form a peroxy radical (reaction 3) followed by



hydrogen abstraction from an organic compound to regenerate a carbon-

-------
                                                                          18

18
      Figure 1-3.  Mechanism for the reaction of nitrogen dioxide with
                               unsaturated fatty acids
    Scheme I:   Addition mechanism

            (1)  addition of nitrogen dioxide to carbon-carbon double
                 bound to form a carbon-centered radical

            (2)  reaction of molecular oxygen with carbon-centered
                 radical to form a peroxy radical

   Scheme II:  Hydrogen abstraction mechanism

            (3)  abstraction o-hydrogen from unsaturated hydrocarbon
                 to form a delocalized carbon-centered radical

            (4)  reaction of molecular oxygen with delocalized
                 carbon-centered radical to form a peroxy radical

-------
                                             19
                  CsJ
                          ro
       o
        i
       o
        i
 o
 o
  I
•o-
  i
•o-
  I
                               y
                                o-
                                 I
+
X
                                     o
                                     o
                                     I
                                     o—
                                     I
                                     o
                                     II
 II
,0,

 +
•o—
 I
•o—
 I

 CVJ
o
                                O
                                 I
                                O
H



o
CO
                    o>
                    o
                   CO

-------
                                                                          20
20
                   Figure 1-4.  Autoxidation in the cell
     (1)  initiation:  ferric-ion catalyzed hemolysis of  hydroperoxide
          to yield an alkoxy radical and a hydroxide ion

     (2)  propagation:  abstraction of nydrogen from an alkane to yield
          a carbon-centered radical

     (3)  propagation:  reaction of carbon-centered radical with
          molecular oxygen to yield a peroxy-radical

     (4)  propagation:  abstraction of active hydrogen to yield an
          organic-hydroperoxide and a free radical

     (5)  production of a carbon-centered radical (QH « allylic
          hydrogen) will result in a chemical amplification of
          autoxidation

     (6)  reduction of ferrous ion by an organic hydroperoxide to yield
          a peroxy radical and a hydrogen ion

          RH « organic hydrogen; QH = active organic hydrogen; RCH *
          organic alcohol; ROOH = organic peroxide; R- and Q-  = carbon
          centered radical; ROD  * organic peroxy-radical; RO   » alkoxy
          radical; HO- = hydroxide ion; H* * hydrogen ion; FeJ* and
          Fe1* « iron ion redox couple.

-------
                               21
ROM
       ROOM
ROO + H+

-------
                                                                      22




centered radical and an organic hydroperoxide (reaction 4).  Propagation




can also occur by the production of alkoxy and peroxy radicals by the




metal ion-catalyzed oxidation of organic hydroperoxides (reactions 1 and




5).




     During the initiation phase, free radicals can be produced by



thermal hernolysis of bonds, one-electron redox reactions, -and highenergy




radiation (49).  The rate of hemolysis of organic hydroperoxides is




extremely slow at physiological temperatures.  However, soluble




transition-metal ions (51) and transition-metal complexes such as iron




protoporphyrins (52) catalyze the production of alkoxy radicals from




hydroperoxides (reaction 1).  AlkoxV radicals are, in general, extremely



reactive so that hydrogen abstraction (even'from saturated alkanes)




(reaction 2) proceeds at the diffusion-limited rate.  Peroxy radicals




are, however, relatively stable, and can only abstract hydrogen from




easily oxidizable compounds (53,54).  Autoxidation in the cell is thus




accelerated by reduced nucleotides, protein-thiol moieties, and poly-



unsaturated lipids.  Termination can occur by the combination of two




radical species to form nonradical products or by the formation of a




stable radical.  Nonradica.l products of termination reactions include



malonaldehyde (55), conjugated dienes (56), and volatile hydrocarbons



(57).




     Autoxidatio.. in tissue is limited by the concentrations of all




oxidizable compounds and the rates of all initiation and termination




reactions, but is independent of the partial pressure of molecular




oxygen (54).  As the concentrations of oxidizable compounds in the cell




are fixed, the various mechanisms that have evolved to protect the cell




against oxidation either limit initiation processes or increase termina-

-------
                                                                      23



tion reactions (58).  Initiation-inhibiting antioxidants act by



directing the decomposition of organic peroxides to nonradical products.



This type of  protective mechanism includes the glutathione peroxidase



enzymes (59,60), catalase, and superoxide dismutase (61), and require



NADPH, supplied by the pentose monophosphate shunt, for activity.  It is



therefore possible that this type of response can lead to a depletion of



reducing equivalents and inhibit mitochondrial production of adenosine



triphosphate.  Chain-breaking antioxidants act by forming a stable



radical.  For example, the o,0,7,6-tocopherols (vitamin E) are found in



the cell membrane and act by forming stable phenoxy radicals, thus



preventing chain propagation in the lipid bilayer of the membrane (62).



The reduced phenol can then be regenerated by reaction with ascorbate



(vitamin C) (63).  Recent evidence suggests that, at low concentrations



of oxygen, ^-carotene also acts as a chain-breaking antioxidant by the



formation of a biologically stable radical (64).



      Mechanism of Toxicity.  Autoxidation is an uncontrolled branching



chain reaction with potentially disastrous consequences to the cell.



Studies of the effects of ozone on cultured erythrocytes suggest that



toxicity is due to oxidation of the cellular membrane, and it is



unlikely that ozone penetrates the cell membrane (31,38).  The sequence



of events is:  i) free radical-induced autoxidation, initiated by



ozonolysis of unsaturated phospholipids in the cell membrane; ii) cross-



linking of membrane proteins, either by direct radical combination, or



by formation of Schiff-base adducts between carbonyl moieties (produced



by the peroxidation of unsaturated lipids) and primary amines; and iii)



the formation of disulfide linkages by the oxidation of protein-



sulfhydryl groups.  Disruption of membrane processes, especially in the

-------
                                                                      24



mitochondria, and utilization of reducing equivalents by the various



antioxidant mechanisms may seriously inhibit cellular production of ATP.



Therefore, cell death may ultimately result from depletion of the



available energy sources used to maintain normal anabolic and catabolic



processes.





                             Isotopic Studies



     At present, the key issues in the study of the toxicity of inhaled



oxidants fall into two categories: dosimetry and mechanisms of oxidant



injury in vivo.  The important aspects of dosimetry are quantitation of



the retained dose of an inhaled oxidant to epithelial tissue and



determination of depositional fine structure within the respiratory



system.  An understanding of the physicochemical factors controlling



respiratory absorption can be developed by correlating differences in



total and local dose with differences in airway structure in various



mammalian species.  In conjunction with dosimetry models, this informa-



tion can be used to obtain an accurate prediction of absorption in the



human lung.



     Activity in the area of oxidant mechanisms has involved, primarily,



in vitro studies of the oxidation of organic compounds and the effects



of a specific oxidant insult on cell cultures.  Although a substantial



framework for tissue damage by photochemical pollutants has been



proposed, the chemical mechanisms involved in oxidant injury in vivo are



still speculative.  While autoxidation in the cell membrane, initiated



by free radicals produced by reaction of nitrogen dioxide or ozone with



unsaturated lipids, has been suggested as the primary mode of oxidation



in vitro and in cell cultures, there is no direct evidence that this



mechanism is important in the toxicity of inhaled oxidants in vivo.  The

-------
                                                                      25



measurement of both direct oxidation and autoxidation in vivo is


therefore important because cell death may be the result of a general


depletion of reducing equivalents and inhibition of mitochondrial ATP


production.


     Inhalation of oxidants can result in tissue damage by one, or both,


of two distinct processes:  direct reaction with molecular components of


tissue and free radical-induced oxidation of tissue involving molecular


oxygen (02).  Direct reaction of nitrogen dioxide or ozone with tissue


components results in the fixation of toxicant-derived atoms in the


tissue, whereas free radical-initiated oxidation leads to the uptake of


Oj-derived oxygen in tissue.  This is: an important distinction which


allows these mechanisms to be resolved and studied in vivo by the use of
                                      »*             ~~*~~~~"~~

oxygen isotopic tracers.  The isotopic label provides a means to


determine the retained dose of a specific oxidant (03, N02, or 02) in


tissue, so that, by the appropriate use of labeled oxidant, questions of


dosimetry and mechanism can be addressed.


      Physiological Applications of Oxygen-18.  All radioactive isotopes


of oxygen are sufficiently short-lived (Table 1-1) that they are of


limited utility as physiological tracers.  Therefore, a stable isotope


of minor abundance, either oxygen-17 and oxygen-18, must be employed.


The natural abundance of both isotopes in biological material represents


an irreducible background, and becomes especially important if only


minute quantities of labeled compound enter the body (as is the case


with inhaled pollutants).  Although the use of oxygen-17 would offer


greater sensitivity because of its lower natural abundance, the methods


used to quantify oxygen isotopes in organic material are more  suited to



the determination of oxygen-18.  Moreover, compounds highly enriched in

-------
                                                                    26
                   Table 1-1.   Isotopes of  Oxygen
natural mass
nuclide abundance
% AU
»0 — 13b
I 4Q 	 14b
»0 — 15b
**0 99.759 15.99491502°
IT0 0.037 16.99913220C
liO 0.204 17.99915996°
b
i »0 	 19
a OQ 	 20b
half-life
s
0.0087
71.0
124
(stable)
(stable)
(stable)
29
14
decay
process6
0+, 17.8 MeV
/J+, 5.14 MeV
0+, 2.76 MeV
	
	
	
0-, 4.82 MeV
0-, 3.81 MeV
approximate abundance in mammalian tissue

nominal mass

Wapstra, A. H.; Grove, N. B.  1971 atomic mass evaluation.  I.
Atomic mass scale.  Nuclear Data, Sect. A. 9:267-301, 1971.

Handbook of Chemistry and Physics ,  53rd edition, Weast, R. C.  (ed.),
Chemical Rubber Company Press, 1973, p. B-249.

-------
                                                                      27




oxygen-17 are rare and expensive,  whereas compounds containing greater




than 99% oxygen-18 are readily available and comparatively inexpensive.




Oxygen-18 is, therefore, the best  choice for use as a physiological




tracer.




     In contrast to a radioactive  isotopic tracer where quantitation is




based on the detection of a specific decay product, difference in mass



is the only physical property which can be used to distinguish stable




isotopes of a particular element.   Mass spectrometry provides the most




accurate and precise determination of oxygen isotopic abundances.




However, analysis of complex organic mixtures, such as physiological




samples, requires that all oxygen  in a sample be converted to a simple,



volatile gas for mass spectrometric measurement.  Techniques for




preparation of carbon dioxide from organic material and for the mass



spectrometric analysis of carbon dioxide will be discussed in chapters 2




and 3, respectively.

-------
                                                                      28

                              •  REFERENCES

 1.   Melton,  C.  S.   Effects of long-term  exposure  to  low levels  of
     ozone:  A review.  Aviat. Space Environ.  Ked. 53:105-111,  1982.

 2.   Epidemiologic  Appraisal of  Photochemical  Oxidants,  Air Quality for
     Ozone and Other Photochemical Oxidants, Vol.  II,  Chapter 10, U.  S.
     Environmental  Protection Agency,  Office of Research and  Develop-
     ment, Washington, D.  C., EPA-600/S-78-004, 1978.

 3.   Stokinger,  H.  E.   Ozone toxicology.  Arch. Environ.  Health 10:
     719-731, 1965.

 4.   Griswold, S. S.;  Chambers L.  A.;  Motley,  H. L.   Report of case
     of exposure to high ozone concentrations  for  t*o hours.   Arch.
     Industr. Health Ib:108-118, 1957.

 5.  . DeLucia, A. J.; Mustafa, M. G.; Cross,  C. E.; Plopper, C. G.;
     Dungworth,  D.  L.; Tyler, W. S. Biochemical and  morphological
     alterations in the lung following ozone exposure.  Am. Inst. Chem.
     Eng. Symp.  Series, 71:93-100, 1975.
                                      4 '
 6.   Dungworth,  D.  L.; Cross, C. E.; Gillespie, J. R.; Plopper,  C.  G.;
     The effects of ozone on animals.   Ozone  Chemistry and Technology ,
     Chapter 2,  Murphy, J. S. and Orr, J'. R. (ed.s),  Franklin Institute
     Press,  Philadelphia,  1975.

 7.   Coffin,  D.  L.; Blonaner, E.  J.; Gardner, D. E.; Kolzman,  R.   Effects
     of air pollution on alteration of susceptibility to pulmonary
     infection.    Proceedings of the 3rd  Annual Conference on
     Atmospheric Contamination in Confined  Spaces, Aerospace  Medical
     Research Laboratories, Wright-Patterson Air Force Base,  Ohio,  p.
     71, 1968.

 8.   Savino,  A.; Peterson, M. L.;  House,  D.; Turner,  A.  G.; Jeffries, H.
     E.; Baker,  R.   The effect of  ozone on  human cellular and humoral
     imnunity: characterization  of T and  B  lymphocytes by rosette
     formation.   Environ.  Res. 15:65-69,  1978.

 9.   Peterson, M. L.; Harder, S.;  Rummo,  N.; and House,  D. Effect  of
     ozone on leukocyte function in exposed human  subjects.  Environ.
     Res. 15:485-493,  1978.

10.   Ospital, J. J.; Hacker, A.  D.; Mustafa, M. G. Biochemical changes
     in rat lungs after exposure to nitrogen dioxide.  J. Toxicol.
     Environ. Health 8:47-58, 1981.

11.   Witschi, H.; Cote, M. G. Primary pulmonary response to  toxic
     agents.  Crit. Rev. Toxicol.  5:23-65,  1977.

12.   Stokinger,  H.  E.; Coffin, D.  L.   Biological  effects of urban air
     pollution.   Air Pollution  ,  2nd  edition, Chapter 1, Stern,
     A. C. (ed.), Academic Press,  Hew  York, 1968.

-------
                                                                      29

13.  Buckley, R. D.;  Balchum, 0. J.   Enzyme  alterations  following
     nitrogen dioxide exposure.  Arch. Environ.  Health,  14:687-692,
     1967.

14.  Werthamer,  S.;  Penha,  P. D.; Amaral,  L.   Pulmonary  lesions induced
     by chronic  exposure to ozone.  Arch.  Environ.  Health,  29:164-166,
     1974.

15.  Frager, N.  B.;  Phalen,  R. F.; Kenoyer,  J.  L.   Adaptation to ozone
     in reference to mucociliary clearance.   Arch.  Environ. Health
     34:51-57, 1979.

16.  Stephens, R. J.;,  Freeman, G.; Wright,  P.  G.   Cytologic changes
     in dogs' lungs  induced by chronic exposure to  ozone.   Am. J. Path.
     73:711-726, 1973.

17.  Freeman, G.; Stephens,  R. J.; Coffin, D. L.; Stara, F. J.  Changes
    . in dogs' lungs  after long-term exposure to ozone.   Arch. Environ.
     Health. 26:209-216, 1973.

18.  Evans, M. J.;,  Cabral,  L. C.;, Stephens, R. J.;  Freeman, G.  Trans-
     formation of alveolar  type 2 cells  to type 1 cells  following
     exposure to N0a.    Exp. Mol. Pathol. 22:142-150,  1975.

19.  Bates, D. V.  Physiological effects on  man of  air pollutants.  Fed.
     Proc., 33:2133-2135, 1974.

20.  Stephens, R. J.; Sloan, M. F.; Evans, M. J.; Freeman,  G.  Early
     response of the lung to low levels  of ozone.   Am. J.  Pathol. 74:31-
     44, 1974.

21.  Hugod, C.  Ultrastructural changes  of the rabbit lung after a 5 ppm
     nitric oxide exposure. Arch. Environ.  Health  34:12-17, 1979.

22.  Plopper, C. G.j  Chow,  C. K.; Dungworth, D. L.; Tyler,  W. S.
     Pulmonary alterations  in rats exposed to 0.2 and 0.1 ppm ozone:  A
     correlated morphological and biochemical study.  Arch. Environ.
     Health 34:390-395, 1979.

23.  Miller, F.  J.   A mathematical model cf transport  and removal of
     ozone in mammalian lungs  .  Ph.D. thesis, North  Carolina State
     University, Raleigh, North Carolina,  1977.

24.  Miller, F.  J.;  Menzel, D. B.; Coffin, D. L. Similarity between
     man and laboratory animals in regional  pulmonary deposition of
     ozone.  Environ. Res.  17:84-101, 1978.

25.  Weibel, E.  R.  Morphological basis  of alveolar-capillary gas
     exchange.  Physiol. Rev.  53:419-495,  1973.

26.  Weibel, E.  R.;  Gomez,  D. M.  Architecture of the human lung.
     Science 137:577-585, 1967.

-------
                                                                      30

27.  Weibel, E. R.   Morphometry of the human lung .  Springer, Heidel-
     berg, 1963.

28.  Personal communication from Dr. R. Schreck,  Biomedical Science
     Department, General Motors Research Center,  Warren, Michigan.

29.  Horsfield, K.;  Gumming,  G.  Morphology of the bronchial tree in
     man.  J. Appl. I'hysiol. 24:373-383, 1968.

30.  Horsfield, K.; C'.uraming, G.  Functional consequences of airway
     morphology.  J. Appl. Physiol. 24:384-390, 1968.

31.  Chan, T. L.; Schreck, R.  M.; Lippman,  M.  Effect of the laryngeal
     jet on particle deposition in the human trachea and upper bronchial-
     airways.  J. Aerosol Sci. 11:447-459,  1980.

32.  Schroter, R. C.; Sudlow,  M. R.  Flow patterns in models of the
     hujian bronchial airways.   Respir. Physiol. 7:341-355, 1969.
                                4 '
33.  Olson, D. E.; Sudlow, K.; Horsfield, K.; Filley, G. F.  Convective
     patterns of flow during inspiration'.  Arch.  Intern. Med. 131:51-57,
     1973.

34.  Goldstein, B. D.; Balchum, 0. J.  Effects of ozone on lipid
     peroxidation in the red blood cell.  Proc. Soc. Exp. Med. 26:356-
     358, 1967.

35.  Menzel, D. B.; Roehm, J.  N.; Lee, S. D.  Vitamine E: the biological
     and environmental antioxidant.  Agr. Food Chea. 20:481-485, 1972.

36.  Mudd, J. B.; Leavitt, R.; Ongun, A.; McManus, T. T.  Reaction of
     ozone with amino acids and proteins.  Atm. Environ. 3:669-682,
     1969.

37.  Leh, F.; Warr, T. A.; Mudd, J. B.  Reaction of ozone with protease
     inhibitors from bovine pancrease, egg white, and human serum.
     Environ. Res. 16:179-190, 1978.

38.  Chan, P. C.; Kindya, R. J.; Kesner, L.  Studies on the mechanism of
     ozone inactivation of erythrocyte membrane (Na* + K*)-activated
     ATPase.  J. Biol. Chem. 252:8537-8541, 1977.

39.  Menzel, D. B.  Oxidation of biologically active reducing substances
     by ozone.  Arch. Environ. Health 23:149-153, 1971.

40.  Roehm, J. N.; Hadley, J.  G.; Menzel, D. B.  Oxidation of
     unsaturated fatty acids by ozone and nitrogen dioxide.  Arch.
     Environ. Health 23:142-148, 1971.

41.  Kesner, L.; Kindya, R. J.; Chan P. C.  Inhibition of erythrocyte
     membrane (Na* * K*)-activated ATPase by ozone-treated
     phospholipids.  J. Biol.  Chem. 254:2705-2709, 1979.

-------
                                                                      31

42.  Thomas, H. V.;  Mueller,  P.  K.;  Lyman, R.  L.   Lipoperoxidation of
     lung lipids in  rats exposed to  nitrogen dioxide.   Science 159:532-
     534, 1968.

43.  Csallany, A. S.;  Ayaz, K. L.  Long-term exposure  of  mice in the
     presence and absence of  vitamin E.   I.  Effects on body weights and
     lipofuscin pigments. Arch. Environ. Health  33:285-291,1978.

44.  Goldstein, B. D.; Lodi,  C.; Collinson,  C.; Balchum,  0.  J.  Ozone
     and lipid peroxidation.  Arch.  Environ. Health 18:631-635,  1969.

45.  Dumelin, E. E.; Dillard, C. J.; Tappel, A. L. Effect of vitamin E
     and ozone on pentane and ethane expired by rats.   Arch. Environ.
     Health 33:129-135, 1978.

46.  Criegee, R.  The  Course  of  ozonation of unsaturated  compounds.
     Rec. Prog. Chem.  18:111-120,  1957.

47.  Pryor, W. A.; Lightsey,  J.  W.;  Church,  D. F.  Reaction  of nitrogen
     dioxide with alkenes and polyunsaturated  fatty acids:  addition and
     hydrogen abstraction mechanisms. J. Am.  Chem. Soc.  104:6685-6692,
     1982.

48.  Demopoulos, H.  B.  The basis of free radical pathology.  Fed. Proc.
     32:1879-1861, 1973.

49.  Pryor, W. A. Free radical  reactions and  their importance in
     biochemical systems. Fed.  Proc. 32:1862-1869, 1973.

50.  Pryor, W. A. Free Radicals.  McGraw-Hill, New York, 1966.

51.  Kochi, J. K. Mechanism  of  organic  oxidation and  reduction by metal
     coa.olexes.  Science 155:415-424, 1967.

52.  Barber, A. A.;  Bernheim  F.   Lipid peroxidation: its  measurement,
     occurrence, and significance in animal  tissues.   Advan.
     Gerontological  Res. 2:355-403,  1967.

53.  Mayo, F. R.  Free-radical  autoxidation  of hydrocarbons.  Ace. Chem.
     Res. 1:193-201, 1968.

54.  Walling, C.  Limiting rates of  hydrocarbon autoxidations.  J. Am.
     Chem. Soc. 91:7590-7594, 1969.

55.  Dahle, L. K.; Hill, E. G.;  Holman,  R. T.  The thiobarbituric acid
     reaction and the  autoxidations  of polyunsaturated fatty acid methyl
     esters.  Arch.  Biochem.  Biophys. 98:265-261, 1962.

56.  Farmer, E. H.;  Koch, H.  P.; Sutton, D.  A. The course of
     autoxidation in polyisoprenes and allied  compounds.   Part V.
     Observations on fish-oil acids.  J. Chem. Soc. 1943, 122-125.

-------
                                                                      32

57.  Dumelin, E. E.;  Tappel,  A.  L.   Hydrocarbon gasses produced during
     in vivo peroxidation of  polyunsaturatod fatty acids and
     decomposition of preformed  hydroperoxides.  Lipids 8:194-198, 1973.

58.  Burton, G. W.; Ingold, K. U.   Autoxidation of biological molecules.
     1.  The antioxidant activity of vj.tamine E and related chain-
     breaking phenolic antioxidants in vitro.  J. Am. Chem. Soc.
     103:  6472-6477, 1981.

59.  Chow, C. &.; Tappel, A.  L.   An enzymatic protective mechanism
     against lipid peroxidation  damage to lungs of ozone-exposed dogs,
     Lipids 8:518-524, 1972.

60.  Chow, C. K.; Tappel, A.  L.   Activities of pentose shunt and
     glycolytic enzymes in lungs of ozone-exposed rats.  Arch. Environ.
     Health 26:205- 208, 1973.

61. • Fridovich, I.  Superoxide radicals and superoxide dismutase.  Ace.
     Chem. Res. 5:321-326, 1973.

62.  Beri, J. G.; Farrell, P. M.  Vitamine E. in  Vitamine and hormones,
     v. 34, pp. 31-74, Academic  Press, New York, 1976.

63.  Parker, J. E.; Slater, T. F.;  Willson, R. L.  Direct observation of
     a free radical interaction  between vitaraine E and Vitamine C.
     Nature (London)  278:737-738,  1979.

64.  Beri, J. G.; Farrell, P. M.  0-Carotene: an unusual type of lipid
     antioxidant.  Science 224:569-573, 1984.

-------
                                                          33
                   CHAPTER 2






DETERMINATION OF OXYGEN-18 IN ORGANIC SUBSTANCES

-------
                                                                      34



    Determination of the abundance of  l§0 in organic materials requires



conversion of oxygen to a volatile gas such as  C02  or Oj  for mass



spectrornetric analysis (1).   Up to now,  several techniques for the



preparation of carbon dioxide from oxygen in organic material have been



described (2-7).  Two problems commonly encountered have  been i) in-



accuracies linked to fractionation of  the oxygen isotopes between two or



more intermediates in the conversion pathway and ii) interferences



arising from oxygen contamination.  While these problems  have been



successfully overcome in procedures where the sample is pyrolyzed in a



sealed nickel tube (7), such methods require a great deal of skilled



manipulation and, at their present state of development,  are limited to



the analysis of compounds containing only carbon, hydrogen, and oxygen.



     A general technique for oxygen isotopic analysis of  physiological



samples, which utilizes the Schutze-Unterzaucher procedure (8,9) for the



preparation of carbon dioxide, is described in this chapter.  In this



procedure, organic oxygen is converted to carbon dioxide  in a sequence



of reactions: i) pyrolysis of the organic sample, ii) conversion (by



equilibration with solid carbon at high temperature) of the oxygen-



containing pyrolysis products to carbon monoxide, and iii) oxidation of



the carbon -monoxide by iodine pentoxide to produce carbon dioxide.  An



unavoidable blank and memory (2) associated with the pyrolytic



production of CO, and the addition of  oxygen from iodine  pentoxide



during the production of CO,, complicate the isotopic analysis.



Ultimately, the abundance of oxygen-18 in the initial sample must be



reconstructed from that measured in carbon dioxide contaminated with



oxygen from other sources.  Interferences in the isotopic analysis



arising from contamination of sample-oxygen have been addressed by the

-------
                                                                      35




development of a mathematical model based on the physical processes




occurring in the conversion of organically bound oxygen to carbon




dioxide.  The model facilitates accurate and precise determinations of




oxygen-18 in organic material.




     Determination of the effects of non-sample-derived oxygen requires



the use of samples with known isot.^ic compositions.  Because there is




no standard method for analysis of oxygen isotopes in organic materials,




no organic compounds which could be classified as "oxygen isotopic




standards" have been available.  To solve this problem, a procedure for




the quantitative pyrolytic decarboxylation of aromatic carboxylic acids



has been developed.  The isotopic composition of the pyroly^ically



produced carbon dioxide can be measured directly, and the isotopic




composition of the organic compound thus determined independently and



accurately.  These new standards have allowed accurate quantitation of




contaminants and sensitive tests of the performance of the method.






                        Oxygen Isotopic Standards




     Benzole acid has been used to define optimum conditions for




pyrolysis and to test the fidelity of the isotopic analysis.  The major



products of the thermal decomposition of benzoic acid are benzene and




carbon dioxide.  At approximately 500*C, decarboxylation of benzoic acid




has been shown (10) to occur predominantly by a unimolecular process






                        C,H,COOH  .      C.H. + CO,                (2-1)






with small quantities of CO, K,, and biphenyl being produced by minor




side reactions involving radical intermediates.  The occurrence of




heterogeneous catalysis cannot be excluded (10), especially when group-




VIII transition metals are present (11).  Further, carbon dioxide

-------
                                                                      36

produced by reaction 2-1 can also yield carbon monoxide if excess carbon

is present (12)


                        COj + C  -, '.•»• 2CP                        (2-2)


Under the conditions that prevail during pyrolysis, reaction 2-1 is

irreversible, whereas reaction 2-2 is reversible and may approach

equilibrium.

     The yield of carbon dioxide is affected by both processes.  Kinetic

isotope effects (KIE) associated with reaction 2-1, with both the

forward and reverse processes in reaction 2-2, and an equilibrium
                                  4!
isotope effect (EIE) associated wi.th reaction 2-2 can shift the isotopic
                                   r>
composition of the carbon dioxide from that of the carboxyl group if

products other than COj are formed''from it during pyrolysis.

      Pyrolysis _of .Benzoic,Acid.  With pyrolysis of benzoic acid, the

yield and isotopic composition of the C02 were affected by pretreatment

of the quartz tubes.  If the quartz tubes were used without prior

treatment, yields of carbon dioxide were low and variable (90 ± 5%), and

the isotopic composition of the C02 was irreproducible (± 3%e).  Use of

the cleaning procedure described  in the experimental section improved

both the yield of carbon dioxide  and the consistency of the isotopic •

analysis.  Only quartz tubes treated by this procedure were used to

obtain the results summarized here.

     Samples of benzoic acid were pyrolyzed at 500, 550, and 600»C for

5, 10, 15, and 20 minutes.  The yields and isotopic compositions of

carbon dioxide are given in Table 2-1 a.id their variation with reaction

conditions is graphically  summarized in Figure 2-1.  A maximum yield of

approximately 97% was attained in 5 rnin at 600eC and at  longer times

-------
                                                                      37
   Table 2-1:  Effect of Pyrolysis Time and Temperature on Yield and
                   Isotopic Composition of Carbon Dioxide


       time       yielda          613C_..  vs PDBa      $l'0rri  vs SMOWa
        min         %                 C°2                  C°2
   500'C

         5     76.1 ± 4.0        -32.71 ± 0.92         +25.72 ± 0.09

        10     94.0 ± 0.4        -26.33 +. 0.14         +25.72 ± 0.03

        15     95.2 ± 0.4        -25.93 ± 0.05         +25.68 + 0.05

        20     94.9 ± 0.1        -25.74 ± 0.02         +25.62 ± 0.04




   550'C

         5     94.7 ± 0.1        -26.48 ± 0.13         +25.67 t 0.02

        10     98.4 ± 1.0        -25.56 ± 0.01         +25.61 t 0.02

        15     96.1 ± 0.4        -25.67 ± 0.04         +25.57 ± 0.03

        20     94.0 + 0.5        -25.51 ± 0.04         +25.66 ± 0.05




   600«C

         5     98.8 t 0.2        -25.56 ± 0.02         +25.59 ± 0.02

        10     97.7 ± 0.1        -25.62 ± 0.08         +25.61 ± 0.13

        15     93.5 i 0.3        -25.64 ± 0.04         +25.30 ± 0.02

        20     92.0 t 0.2        -25.79 ± 0.04         +25.17 + 0.05




aMeans and  standard errors of means of 3 replicate analyses.

-------
                                                                          33




38
     Figure 2-1.  Yield and isotopic composition of carbon dioxide




produced by pyrolysis of benzoic acid.  Each point indicates the mean



and standard error of the mean of 3 replicate analyses:  (@) 500°C, dD




550»C, and (V) 600eC.

-------
                                             39
100
                 10            15
                      time, min
20

-------
                                                                      40


for lower temperatures.  Initial low yields of  carbon dioxide were the


result of incomplete decarboxylation; residual  benzoic acid was


observed.  The systematic decrease in yield of  C02  under more rigorous


conditions was apparently associated with the partitioning of carbon and


of oxygen between CO ard CO, (reaction 2-2).  In all cases, benzene


(m/z 78} condensed on the inside of the quartz  tube as it cooled.


Carbon monoxide (m/z 27.9951) was also present  in the liquid-nitrogen-


noncondensible fraction of all samples.


     A kinetic isotope effect is associated with decarboxylation.  The


decarboxylation of aromatic carboxylic acids in aqueous solution has


been studied extensively (13-16).  In general,  the mechanism involves


displacement of the carboxyl group by a proton.  At low acidity, the


rate-limiting step is protonation of the ring carbon to form a


delocalized carbanion, and carbon isotopic fractionations are small.  At


high acidities, protonation is fast and the rate-limiting step is

   *•   *•
breaking of the C-C bond.  Kinetic isotope effects range between 2.5 and


3.5% (13-16).  For pyrolysis of benzoic acid at 500 and 550°C, initial


low yields at short pyrolysis times (5 and 10 min,  and 5 min,


respectively) were accompanied by depletion of  carbon-13 in the CO,


(Fig. 2-1, Table 2-1).  A carbon-12/carbon-13 kinetic isotope effect of


approximately 1.5% was calculated from these results.  The smaller value


for the KIE is consistent with (i) the high temperature of the reaction


and/or (ii) a concerted mechanism where the rate-limiting step involves


intramolecular displacement of the carboxyl moiety by hydrogen.  No


secondary KIE for oxygen was observed.


     The magnitude of isotopic fractionation between the carboxyl group


and carbon dioxide depends on the magnitude of the isotope effect and

-------
                                                                      41



the yield of CO,.  If decarboxylation were quantitative, no isotopic




fractionation would occur, and the isotopic composition of the COj would




be identical to that of the carboxyl group.  Isotopic fractionations




diminished as decarboxylation approached completion.  However,




quantitative yields of CO, were never observed.  A maximum yield was




attained in 5 min at 600°C (98.8%), in 10 min at 550°C (98.4%), and in




15 min at 500»C (95.2%).  At maximum yield, the carbon isotopic




composition of the CO, was constant (6ls^co2 vs roB * -25.65%»).  The



isotopic composition of the oxygen in the CO, was independent of yield




at 500 and 550"C (<Oco  vs SHOW « +25.64%0).  At 600»C, the oxygen-18



content decreased with yield.  These variations in yield and oxygen-



isotopic composition can be attributed to pyrolytic decomposition of CO,



and an accompanying equilibrium isotope effect.




     To examine isotopic fractionaticns associated with decomposition of



COj, 30-umole aliquots of CO, were transferred to quartz tubes r^nd




heated to 550°C for intervals of 5 to 30 minutes.  The recoveries and



isotopic compositions of the residual carbon dioxide are summarized in




Table 2-2.  Although small amounts of CO, were lost, no statistically




significant isotopic shifts were observed (p < 0.05).  However, the data



in Table 2-2 suggest that decomposition of CO, under conditions of




pyrolysis resulted in an enrichment of carbon-13 and a depletion of




oxygen-18 in the residual carbon dioxide.  These trends differ from




those observed with pyrolysis of benzoic acid (Table 2-1, Fig. 2-1), and




were probably due to a KIE associated with decomposition of C02 to CO




(the forward process of reaction 2-2).  Acceleration of reaction 2 in




both directions in the prasence of elemental carbon produced by




decomposition of benzene during pyrolysis of benzoic acid (10) allows

-------
                                                                  42
Table 2-2:  Recovery and Isotopic Composition of Carbon Dioxide
                     Heated to 550°C in Quartz Tubes
time    n     recovery3      6lscco2 vs fDn&      51*°C02 vs SM°tjl?a
                  %                 %«                   %o
                 100         -10.409 ± 0.09         -*-31.33 ± 0.04
10      4    99.8 ± 0.1       -9.96 ± 0.02         +31,28 ± 0.03



20      4    99.5 ± 0.2       -9.94 ± 0.03         +31,28 ± 0.02



30      4    99.4 4 0.1       -9.92 ± 0.06         +31.27 ± 0.04




   ± standard errors of means of n replicate analyses.

-------
                                                                      43



this reaction to approach equilibrium.   In this case,  the isotopic



compositions of the products are controlled by the equilibrium isotope



effect associated with reaction 2-2.



     Estimates of isotopic fractionations between carbon dioxide and



carbon monoxide were obtained from calculations of the thennodynamic



equilibrium constant and of the equilibrium oxygen isotopic fractiona-



tion factor for reaction 2-2 (Figure 2-2).  The isotopic composition of



the product gas mixture at 550°C was calculated using values of 14A»a



= 1.000735, K = 0.12.  The temperature and pressure of the gas inside



the quartz tube during pyrolysis were estimated assuming decarboxylation



was quantitative.  This calculation indicated that, at equilibrium,  the



composition of gases derived from the carboxyl group would be 94.1% C02



and 5.S% CO, resulting in a shift in the oxygen isotopic composition of



the COj from that of the carboxylic acid moiety of AS^CLg  vs



carboxylic acid = -0.30%0.  Although the decrease in yield and the shift



in isotopic composition observed during thermal decomposition of benzoic



acid at 600eC (Table 2-1, Fig. 2-1) were significantly greater (p >



0.05) than the predicted values, the model calculations serve only to



estimate isotopic fractionations arising from an EIE.  The differences



between the experimental and theoretical results ar? not serious.  Since



the rates of the forward and reverse steps in reaction 2-2 are



temperature dependent (17), only at 600eC does this reaction approach



equilibrium and do the isotopic fractionations become significant.



Isotope effects associated with decarboxylation are unimportant when



near-quantitative (maximum) yields of carbon dioxide are obtained.  Only



with pyrolysis at 600°C does the oxygen isotopic composition of carbon



dioxide change significantly urom that at maximum yield.

-------
                                                                          44




44
     Figure 2-2.  Calculated values for the thermodynaaiic equilibrium



constant (K) and the equilibrium oxygen isotopic fractionation factor



(a) for reaction 2 over the temperature range of 300 to 1500°K.

-------
                              45
log(Kp)

-------
                                                                      46




     Suitable conditions for isotopic analysis are those which result in




no isotopic fractionation from either incomplete decarboxylation or




production of carbon monoxide..  At 550eC there is a considerable range,




between 10 and 20 minutes, where there is no apparent isotopic




fractionation, which is ideal for isotopic analysis.  Any combination of




time and temperature of pyrolysis which gives a maximum, riear-




quantitative yield of carbon dioxide also satisfies the condition of




minimal isotopic fractionation.  The isotopic composition of the carbon




dioxide is, therefore, identical to that of the carboxyl group.



      Isotopic Standards.  Results of isotopic analyses of CO, derived




by replicate pyrolytic decarboxylations (550°C, 10 min) of five benzoic



acids are summarized in Table 2-3.  Isotopic analysis of CO, obtained by




pyrolysis (550°C, 10 min) of 2,6-pyridinecarboxylic acid, pyrrole-




2-carboxylic acid, and 2-thiophenecarboxylic acid are shown in Table




2-4.



                            MATHEMATICAL MODEL




     The following description of the chemical processes involved in the




production of carbon dioxide from organic material by the Schutze-



Unterzaucher procedure explains the physical basis for the mathematical




treatment.






Chemical Processes



      Pyrolysis of organic material.  Depending on the composition and




the  structure of the initial sample, pyrolysis products  generated  in the




first step of the process may include CO, C03, H,0, NO,  COS,  H,, N,,




CH,, and traces of H,S and C,Ht  (18,19).  Some generalizations  can be




made about the pyrolytic  behavior of specific oxygen functionalities.




Carboxyl groups yield their oxygen predominantly  in the form  of carbon

-------
                                                                     47
           Table 2-3:  Analyses of Benzole acid Standards
    sample                   %-yielda                 1§FCO  x 10'
benzole acid-0             97.0 ± 0.2              2.03548 ± 0.00015
                               t!

benzole acid-1             96.9 ±,0.2  •            2.04756 ± 0.00009



benzole acid-2             96.5'i 0.2              2.11813 i 0.00015



benzole acid-3             96.3 ± 0.3              2.18844 ± 0.00015



benzole acid-4             96.2 ± 0.3              2.24562 ± 0.00040




      ± standard errors of means of 5 replicate analyses.

-------
                                                                     48
Table 2-4:  Analysis of  Heteroaromatic Carboxylic Acids


                                                ?C02
       sample                  %-yielda              1§FCO  x 10*
   2f6-pyridine-
   dicarboxylic acid          99.1 ± 0.4          2.03740 ± 0.01017
   pyrrole-2-
   carboxylic acid            95.6 ± 0.2          1.97134 ± C 00006
   2-thiophene-
   carboxylic acid            99.0 ± 0.2          1.97804 ± 0.00005
Means ± standard errors of 5 replicate analyses.

-------
                                                                      49




dioxide, whereas aldehydes, ketones,  and alcohols yield carbon monoxide.




Carbohydrates are the only compounds  to produce water in appreciable




amounts.  Traces of carbonyl sulfide  are present when the sample




contains organic sulfur.  The interactions of the various oxygen-



containing pyrolysis products with carbon in the second step are of




central importance in formation of carbon monoxide and in the appearance




of the blank and the memory.



      Production of CO from pyrolysis products.  A helium carrier gas is




employed to sweep the pyrolysis products through a quartz tube packed



with carbon and maintained at 1060°C.  The oxygen-containing compounds



react with the carbon surface to yield carbon monoxide as the




thermodynamically favored product (12).  This reaction proceeds through




a surface-oxide intermediate.  The reactions of interest are (20):






                    CO, + C  «    *  CO + Cx(0)        fast        (2-3)






                   2ND + 2C  *    *  N, •«• 2CX(0)       fast        (2-4)






                    H,0 + C  •    V H, •*• Cx(0)        fast        (2-5)






                  2COS ••• 2C  «    *  CS, + 2CX(0)      fast        (2-6)






where Cx(0) represents oxygen chemisorbed CO on the surface of the



carbon and x is the number of atoms in the carbon matrix which




participate in the surface-oxygen complex.  The surface oxide then




undergoes a unimolecular decomposition






                       Cx(0) 	^ CO + C            slow        (2-7)






However, the carbon surface is microscopically heterogeneous, containing




two physically distinct reaction sites (21-27): a short-lived inter-

-------
                                                                      50




mediate that rapidiy decomposes to yield carbon monoxide (reaction 2-3),




and a more stable surface oxide (reaction 2-7) in which oxygen is




strongly adsorbed to the surface (25).  The activation energy for




desorption at both sites is inversely proportional to the local surface



coverage (23,26,27).  Thus, oxygen accumulates on the carbon surface as




chemisorbed carbon monoxide.



     Reactions 2-3 through 2-6 mediate the exchange of oxygen between




the carbon surface and the gas phase.  Desorption of CO from the surface




(reaction 2-7) is rate limiting in the formation of carbon monoxide, and




is essentially irreversible (25), whereas reaction 2-3 is readily



reversible (22-24) during the time when the pyrolysis products are in



contact with carbon.  Thus, any oxygen entering the system will be




distributed between the surface oxide and gaseous carbon monoxide.  As



multiple samples are sent through the reactor, slow desorption of carbon




monoxide effectively generates a sorbed-oxygen pool charged with oxygen




from previous inputs.  As subsequent samples are produced, exchange of



oxygen with this pool can buffer the isotopic composition of the carbon




monoxide produced.



      The analytical blank.  The blank arises from a reaction between




carbon and silica (the quartz tube used to contain the carbon bed).  A




direct interaction






                   SiO, + C 	•- CO •*• SiO f C                  (2-8)






is apparently of minimal importance  (28).  Significant amounts of carbon




monoxide are produced by reaction 2-8 at temperatures in excess of




1200°C, but a large blank is observed at lower temperatures.  A process




involving a silicon-hydroxyl moiety as the active oxygen species

-------
                                                                      51




             2SiO(OH) + 2C 	*• 2CO + 2SiO + H,  •*• 2C         (2-9)






has a lower overall energy of activation and could explain the blank




observed at lower temperatures.  Surface hydroxyl groups are likely to




be produced by reaction of acidic gases and HJf produced in the




pyrolysis reactor, with the walls of the quartz tube  (20):"






                 SiO, + HX 	*• SiO(OH) + X                 (2-10)






Due to its lower energy requirements, a mechanism involving reactions



2-9 and 2-10 is a more likely explanation of the blank (20).




     Two factors influence the magnitude of the blank:  contact between



carbon and the quartz tube in the pyrolysis reactor,  and the temperature




of the the reactor in the region of the contact.  Each has been utilized



by other investigators in schemes to reduce the blank.  The quartz tube




has been lined with platinum foil (2) to separate the carbon from the



quartz, but carbon from the sample is inevitably deposited on the




internal surface of the tube in the region where it undergoes volatil-




ization and initial pyrolysis.  Smaller blanks are observed at temper-



atures below 1200*C, but use of a metal catalyst in direct contact with




the carbon is required to ensure a quantitative yield of carbon monoxide




in a dynamic system (20).  The metal on the carbon surface acts as a



catalyst for reaction 2-7, increasing the rate of evolution of carbon




monoxide at the lower temperature (26,29).




     The presence of low levels of a volatile chlorinated hydrocarbon in




the carrier gas has been found to improve the yield and reduce the




tailing of carbon monoxide (30).  The mechanism leading to this improve-




ment is not known.  However, acidic gases are adsorbed by carbon (31),




and it seems likely that HC1 produced by pyrolysis of the halocarbons

-------
                                                                      52

(19) competes successfully for sorption sites on the carbon surface.  As

a result, interferences arising from the sorbed-oxygen pool are reduced

and stabilized.

      Oxidation of CO to yield CO;.   Carbon monoxide produced in the

pyrolysis reactor is oxidized to carbon dioxide as it is swept through a

tube packed with iodine pentoxide (9) and maintained at 120°C.  Under

these conditions, hydrocarbons, methane in particular, are not oxidized

by IjOs, nor is oxygen exchanged between either carbon monoxide or

carbon dioxide and the reagent (2).   Therefore, no interferences to the

isotopic analysis other than the addition of oxygen from iodine

pentoxide accompany this reaction.

                                                   I
Isotopic Mass Balance                             ;

     An isotopic mass balance is required to understand the relationship

between the isotopic composition of  carbon dioxide produced by the

Schutze-Unterzaucher procedure and those of the components of the carbon

dioxide-oxygen pool.  The mass-balance equations must reflect the

specific reactions that result in the production of carbon dioxide from

an organic material.  As noted in Figure 2-3, oxygen in the carbon

dioxide derives from several sources: the sample being analyzed,

previous samples (memory effect), silica in the quartz tube which

contains the carbon packing (blank), and iodine pentoxide.

      Mass Balance for CO.  The abundance of oxygen-18 in the total

oxygen pool is the weighted sum of the abundances in each component.

The mass-balance equations are


               nil'FT = HS^FS + ngi'FQ * Injl'Fj                 (2-11)


and

-------
                                                                          53



53
     Figure 2-3.  Schematic representation of the production of C02 from

                                       i!


an organic sample.  The sorbed-oxygen pool is represented by {c (0)}.
                                                r               X
                                         f

Oxygen from the present sample, from previous samples (memory), and from




the quartz tube (blank) are incorporated in the carbon monoxide through




the surface-oxide.  Carbon monoxide is oxidized to carbon dioxide by




1,0, with the addition of a single atom of oxygen to each molecule of




CO.

-------
                                                    54
 o
•g
*x
O
                                     O
                                      CVJ
 CD

(T
 in
J*?
 Q.

 e
 o
 O)

 13

 O
o


CD

6
 CO

 'en
  O
  v.
  >

 CL
                  o  ^O
                  X  o   c\j
                  <->  O  X
                                     C.
                                              CJ

                                             O
                            CO
                            o
                            o

-------
                                                                      55



                       nT = ng + nQ + In.                         (2-12)





where l»F is the fractional abundance of oxygen-18 and n is the amount



of oxygen (MIDO! 0).  The subscript T designates the total oxygen pool; S



designates oxygen derived from the sample  itself;  Q is used to designate



oxygen from the blank, thought to be derived largely from the quartz



reactor tube; and j is used to designate oxygen derived ultimately from



a previous sample carrying the index j,  that is, a "memory effect" (32).



Only a finite number of samples is considered to be significant in the



sorbed oxygen pool.



     An expression for the fractional abundance of oxygen-18 in the



carbon monoxide can be obtained if it is assumed that no isotopic frac-



tionation accompanies the partitioning of  oxygen between the gas phase



and the sorbed phase in the reduction reactor.  The carbon monoxide is



then a representative sample of the total  oxygen pool, lgF   = **F , and
                                                          \s\J      1


equations 2-11 and 2-12 yield the following relationship
                                       + Zn »•?.

               »Tro- - - - 2 - - - 3 - J - ±-         (2-13)

                              1 + (nQ + Znj)-l/ns





where  lt?rn is the fractional abundance of oxygen-18 in the carbon



monoxide oxygen pool.  Equation 2-13 can be rearranged to yield the



fractional abundance of oxygen-18 in the sample:
     1§FS =  I§FCO +   + Inj
-------
                                                                      56

      Mass Balanca for CO, .   The oxidation  of  carbon monoxide by iodine

pentoxide results in the combination of  two oxygen pools  of  equal size
                       nC02l'FC02 " nC01>FCO * n!l'FI            <


                              nC02  * nCO * nl                   (2-16)


                                   nco - nj.                      (2-17)


where the subscripts (CO, and I) designate the total carbon-dioxide

oxygen pool and that portion of it derived from iodine pentoxide,

respectively.  It follows that, since the size of the oxygen pool  is

doubled, isotopic fluctuations in the carbon monoxide are reduced  by a

factor of two in the carbon dioxideu'


                        »'FCO = 2-»Fc'02 - "Fj-                 (2-18)


Note that ltF]. is the only quantity required for definition of the

relationship between I*FCO and 1§FCO .

     Combining equations 2-13 and 2-18 yields a relationship which

relates the isotopic composition of the carbon dioxide to those of its

component cxygen pools:
             i »FCO  - - : - 3 - = - = - —            (2-19)
                           1 + (n  + Zn
Equation 2-19 summarizes the dependence of the analytical result on

experimental parameters, only one of which is the isotopic composition

of the sample.  The relationship between *'FS and 1§FCO  is complex.  In

particular, it will be different for each sample because the isotopic

composition of the sorbed oxygen pool depends on the history of the

-------
                                                                      57



reactor.  Nevertheless, if ail other terms in equation 2-19 are known,



1§FS can be calculated from observed values of IIFCO •  In practice,  it



is most convenient to use equations 14 and 16 to calculate 1§Fg .





Determination of Correction Parameters



     In devising a procedure to determine J*PI,  lt?Qt  nq* .^^ an



expression for n^ it is helpful to consider the relationship between



1»FCO  and l'Fs for each member in a series of analyses of a single



material where each sample differs only in size.  Equation 2-19



indicates the effects of each parameter on the observed results.  A plot



of 2'1*FCO  as a function of l/ns for a hypothetical series of replicate



analyses of a single organic material with a true isotopic composition,



1>FS, is shown in Figure 2-4.  Several different curves appear as  a



result of changes in the isotopic composition of the sorbedoxygen  pool.



If a sample is infinitely large d/ng approaches zero), the effects of



the blank and the memory will be insignificant.  Accordingly, all  Curves



converge to a common y-intercept at 14FS + 1$Fi'  For samples of finite



size, curve A represents the locus of all possible results of the  first



analysis in the series, curve B the second, curve C the third, etc..



Each curve is displaced from its predecessor because each sample changes



the isotopic composition of the sorbed oxygen pool, making it more like



that of the sample.  A small input affects the sorbed oxygen pool  less



strongly than a large one; therefore, as shown in Figure 2-4, the



magnitude of the shift between curves depends on the size of the



previous sample (recall that a small value of 1/n corresponds to a large



sample).  Curve Z defines the limit reached when, after repeated



analyses of a single material, the isotopic composition of the sorbed-



oxygen pool differs from that of the input sample only due to

-------
                                                                         58




58
     Figure 2-4.  Results predicted" for a plot of 2*18FCQ2 vs.  1/n for



replicate analyses (A, B, C, D, ... s Z) of"an organic compound with a



true isotopic composition of 1*FS. .Equation 2-17 defines the locus of



all possible results for a single analysis.  In this example, both the



blank and the initial sorbed-oxygen pool have been assumed to be



depleted in cxygen-18 relative to-the sample, so all analyses fall below



up  + itp   tne y-intercept.  Curve Z defines the limit where memory
   D      JL


effects have been eliminated by the continuous, repetitive processing of



samples of the same material.

-------
  59
.
 o

-------
                                                                      60
contributions from the blank.   The condition where the effects of memory
have been suppressed by saturation of the sorbed-oxygen pool with oxygen
from the sample will be referred to as "memory equilibration".  In the
construction of Figure 2-4,  it was assumed that both the isotopic
composition of the blank and the sorbed-oxygen pool were depleted in
oxygen-18 with respect to the sample.
      Values of the correction parameters (i. e.,  all terms in equation
2-19 except 1§Fj» l'Fco ' anci °S^ can be determined from results of
analyses of materials for which I*FS is known.  However, since the terms
used to correct for the blank and the memory are not separable an
iterative procedure is required.  In practice, it is useful to
manipulate experimental conditions so that the effects of the blank or
the memory are maximized in a particular experiment, thus improving the
precision with which the correction parameters can be determined.
     An outline of the procedure used to determine l*F-ri 1§Fof "O'
an expression for n  is shown in Figure 2-5.  Initial estimates of
»»FQ, and nQ are obtained from a single experiment in which memory
equilibrium has been established (step I).  Subsequently, the effects of
the memory can be quantitated in a separate experiment (step II).  These
estimates are then refined by a procedure of successive approximations
until the values for all parameters converge within predetermined limits
(step III).  A detailed description of this procedure follows.
      Blank and I20,-added Oxygen.  Under conditions of memory
equilibration, the difference between »«FS and ''FCO? is controlled by
the isotopic composition and relative size of the blank and by the
isotopic composition of the oxygen derived from the 1,0,.  The latter
quantity O'Fj) can be determined from the y-intercept of a plot of

-------
                                                                          61





61
     Figure 2-5.  Flow-chart of the procedure for determining, the



correction parameters.

-------
                                                                     62
    2'18FC0.,  vs l/ns for two
     standards where memory
equilibrium has been established
           determine:
          I'
             18
FQ,  and nQ
    1 fl
      F^Q  vs sample number
     for a series including
 isotopically enriched samples
1
n^ and
r
determine :
an expression
for n •
                                                next approximation
                                               using current values
                                             of correction parameters
                  yes
       final values for:
       F!»   FQ>  ng,  and  n:

-------
                                                                      63




2-l'FCo2 vs l/ns if i'Fg is knovm:






                     {y-intercept} «= lt?z + ''Fj                (2-20)






     The size and the isotopic composition of the blank are determined




from memory-equilibrated curves for two samples of different but knovm



isotopic compositions.  Comparison of difference between the observed
and true isotopic compositions (1 , a parameter whose



values are, furthermore, unknown at this point.  An iterative procedure



can be adopted to derive a consistent set of correction parameters once



the characteristics of the memory have been derived.



      Memory.  The amount of a previous sample which is present in the



sorbed oxygen pool at any time is a function of the size of that  sample



and of the rate of its displacement by subsequent throughputs.  As an



approximation

-------
                             nj
                                                                      64



                                                                 (2-23)
where n* 0 is ^e initial size of the jth previous sample and f(j) is an



exponential function of the index, j.  The quantitative characteristics



of f(j) cannot easily be derived from a specific model of the sorbed



oxygen pool since the surface oxide is heterogeneous.  In practice, the



exponential function which best describes the wash-out of a particular



sample (see below) is determined empirically (32).



     The turnover of oxygen in the sorbed oxygen pool can be determined



by examining the response of the system to isotopic perturbations (32).



In this approach, a series of samples of known isotopic composition



(1»F1) is interrupted by an input of material highly enriched in
                                  ,*


oxygen-18 (1'FJ).  If the sizes of the samples are constant, then the



amount of oxygen from the enriched input appearing in the k-th sample



after that input is given by
              (ns
                         lip   _
                           rco
- "F
                  (2-24)
where  l*FCo is calculated using equation 2-13.  Since f(j) is initially



unknown, values for n. cannot be determined and, in the first



approximation, are set equal to zero in equations 2-13 and 2-24.  The



values of n^ are then fit to an exponential function of the index (j).



Further approximations are obtained by using equation 2-23 to calculate



values for n* in equations 2-13 and 2-24 aM recalculating values for n^-



     Tt is also possible that more than one enriched sample appears in



the  series.  In this case, equation 2-24 must be modified to account for



the  introduction of multiple isotopic perturbations at different times.



The  most recent enriched sample causes the greatest shift in  ''FCQ  , so

-------
                                                                      65
its effects alone are considered explicitly.   Components from all



previous enriched samples are grouped in a single summed term which



treats these effects as a small contribution  to the observed shift
   nk *
                   '•FCO -
                                        ?.  - 1§F,
(2-25)
where H. « 1 when j corresponds to an enriched sample prior to the most



recent and Hj * 0 for all other samples,  independent of their isotopic



composition.  As in the previous example, initial estimates of n  are



obtained by setting n* • 0 and only isotopic perturbations caused by the



most recent enriched sample are considered in the calculation of n^.



This procedure is only an expedient adopted during study of the sorbed-



oxygen pool.  Once the functional form of f(j) has been specified, the



effects of all previous enriched samples  enter into the calculation.



     To complete the procedure, the first approximations of ''FT* ru/



**FQ, and n. are used to calculate values for all unspecified terms.



The procedure is repeated until convergent values for all correction



parameters are obtained.





Adjustment of Analytical Results



     Application of the correction procedure can be visualized most



simply in terms of equations 2-14, 2-18,  and 2-23.  Since 1'FI is known,



**FCO can be determined from equation 2-18.  The characteristics of the



blank (HQ and I§FQ) and of the memory (nj) have been determined directly



by the procedure outlined above.  Solution of equation 2-14 for i*Fs



requires, in addition, values for ns and lIFj.  The size of the sample



(n,) is determined from the elemental analysis, and values for **F* are
  •J                                                               J


given by the results of previous analyses in the series.  Equation 2-23

-------
                                                                      66




is used to determine values for n.,  and liFj  is,  in each case, set equal




to 1§Fg for the j-th previous sample.




     In the correction of a series of analytical  results, equations 2-14




and 2-16 are used to calculate values of *•*?$ based on the final set of




correction parameters and the observed values of  1*FC02'  Tne initial



isotopic composition of the rorbed oxygen pool is assumed to be that of




atmospheric 02 (1'FAIR » 2.04435 x. 10-') because  the introduction of air




into the system is unavoidable when the sampling  head is loaded.  Values




for i§Fs are calculated sequentially.  In this way, values for n$ and




J>FS for all previous samples are available and provide the infurination




necessary to correct for the effects of the memory.  A limit of 30 is



placed on the index j.  Software allowing rapid completion of these




calculations has been developed.



      Refinement of correction parameters.   The  inclusion of three




samples of a working standard in each sequence of analyses allows




refinement of parameters initially determined using the more elaborate



procedure outlined above.  It is of particular interest to check the




values assigned to 1§F]. and HQ.  The procedure is graphically summarized




in Figure 2-6.  The value assigned to nq is adjusted until the value



calculated for lfFs is independent of n
-------
                                                                          67

67
     Figure 2-6.  A schematic illustration of the method for refinement
                                       t!
of values for I§FI and nq.  The latter parameter is first adjusted to

minimize the sum of the residuals noted here as &F.  The value of ltFIis

then adjusted until the calculated value'of J*FS matches the known

value.

-------
                                                  68
                                   I
                H
               U_
              CO

               +
                CO

              CO
-»-•

 /
/
     *-
     
-------
                                                                      69



tests, the benzole acid samples listed in Table 2-3,  the three hetero-



aromatic carboxylic acids (pyrrole-2-carboxylic acid, 2-thiophene



carboxylic acid, and 2,6-pyridine dicarboxylic acid)/ and three primary



standard water samples, SHOW, NBS--1, and KBS-1A, were used to determine



the accuracy and precision of this technique.





Yield of Carbon Dioxide



     Kanonetric techniques were employed to determine the yield of



carbon dioxide from the Schutze-Unterzaucher procedure for samples of



from 3 to 60 wnoles oxygen.  Quantitative yields of carbon dioxide were



obtained from samples containing less than 50 ymoles oxygen at I20S-



reactor temperatures above 125°C (see Figure 2-7).  Incomplete yields of



carbon dioxide were obtained at lower temperatures even though the



thermal conductivity detector just downstream from the pyrolysis reactor



indicated that quantitative yields of CO were being produced.  It. is



evident, therefore, that oxidation of CO to C0a by iodine pentoxide can



be slow and can lead to non-quantitative yields of carbon dioxide at the



flow rate required by the analyzer (25 mL/min.).  The isotopic



composition of the C0a, however, did not vary even when the oxidation of



CO was incomplete.  This result demonstrates that isotope effects in the



oxidation of carbon monoxide by iodine pentoxide are minimal.  During



routine analyses, the temperature of the iodine pentoxide was maintained



between 120 and 140°C to insure quantitative yields.





Correction Parameters



      Blank, 1,0^.  Values of l'Fn, nn, and 1§FT were determined by five
      	      —               y   W          •>•


replicate analyses each of samples of benzoic acid designated 0 and 1 in



Table 2-3 (step I).  The sample sizes in each set of analyses varied

-------
                                                                          70


70
                                       4


     Figure 2-7.  A plot of yield of carbon diojude (/anole CO,) vs
                                         ^

oxygen input (ymole 0) for sucrose at various temper&tures of the 1,0,


reactor: 25"C (©), 50'C (0), 75«C (A), 100»C (O), and 125«C  (D)«

-------
                                        71
3|OLU/7'

-------
                                                                      72

from 30 to 2 wool 0.  Memory equilibration was obtained by processing

five samples containing 40 nmol 0 of the same material prior to each

set.  A plot of 2-1*FCO  vs l/ns for these data is shown in Figure 2-8.

The curves were obtained by fitting each set of data to an equation of

the form y * {O'Fg * Ax)/(l + Bx)} + C (that is,  the form of equation

2-19).

     The isotopic composition of the iodine pentoxide-derivsd oxygen was

determined from the y-intercept of each curve (equation 2-20).  The size

of the blank was determined from the difference between the two curves
    '
at l/ns = 0.1 (tfDol O)-1 (equation 2-21), and the isotopic composition

of the blank was determined from the slope of each curve at the same

point (equation 2-22).  These results are shown in the "initial

estimates" column of Table 2-5.

      Memory.  The washout of a single contribution to the sorbed-oxygen

pool was quantitated by analysis of a series of 23 samples of glycine

which was interrupted at positions 9 and 16 with glycine highly enriched

in oxygen-18 (step II).  Results are summarized Figure 2-9.  The

contribution (n.) of the enriched sample to each total oxygen pool was

initially determined using equation 2-25. The values of n, were fit to

an equation of the form f(j) = a(j)b  (32).  Once initial estimates of a

and b were established, equation 2-23 was used to calculate values for n,

in equation 2-24, and values for a and b were refined.

      Values for l*Fj» 1§F0» and nn were then recalculated using these

values of a and b to specify n  in equations 2-21 and 2-22 (step III).

The results appear in the "first iteration" column in Table 2-5.  Three

iterations were necessary to obtain convergent values for the correction

parameters (Table 2-5).  For every recalculation 1§FI, X*FQ» and "o'

-------
                                                                     73
    Table 2-5:  Iterative determination of correction parameters
initial
estimates
10' x »»Fi 1.9895
10' x *»FQ 2.0413
Hq 0.71
1 0.072
a 2 0.080
3 0.082
1 -1.80
b 2 -1.72
3 -1.75
iterative determinations
1 2
1.9895 1.9895 1.
1.9766 1.9745 1.
1.00 1.01 1.
0.080 0.081
0.081 0.082 0.
0.081 0.081
-1.56 -1.60
-1.61 -1.60 -1.
-1.61 -1.60

3
9895
9750
02
081
60
Three iterations of the memory-correction parameters (a and b) were
performed for each refinecent of the blank parameters (nn and 1*Ff)).

-------
                                                                           74
  74
     Figure 2-8.  A plot of 2-F^ vs. 1/ns for five replicate

analy.es of two benzole acids, -FBA.O = 2.03547 x 10- and "F

2.04757 K 10-; uncorrected:  BA-0 <•),  BA-! («. corrected for'th^

blank and the 1,0,-added oxygen:   BA-0 O,  BA-1 (O>.
BA-1

-------
    2.05

    2.04
                                                                                    75
    4.02
    4.01
2
 cvi
    4.00
    3.99
                                                              -i	L
                       O.I
0.2
0.3
                                                                     0.4
                               05

-------
                                                                          76



76
     Figure 2-9.  A plot of *»FS vs sample number for a series of



glycine (ltFl, = 2.06012 x 10-3) analyses interrupted at positions 9 and



16 with highly enriched material (1§Fj «* 4.489 x 10-3); corrected for



the blank and the IjO,-added oxygen (©); corrected for the blank, the



IjO,-added oxygen, and for the memory (O)» The sample size was constant



at 26 wrool 0.  The filled symbols are the raw results corrected only for



the blank and the iodine pentoxide-added oxygen.

-------
77
1 I 1 I
—
-
—
—
• 	 • 	
^
—

"
—
-
mm^ ^h
\ - ^^^^^S^S.
—
•—
V
—
_J — 1 J 1
IIIIIIII[|IIIP
* o -
• c
• (

• 0
_«...._ ^ '
• (
0


5 -
D—
"

) -
O—

• o -
* c
• o
o
0>

-
—
^ ®
(
«
V
<
1
1
t

'
I ••••
» -
t—
) -
^_
• -
1 1 1 1 1 1 1 1 1 1 1 1 1 I
3 8 !Q 2 ?
U o

CJ

0
r\i
CD
•
^


C\J

O
•
CO
*
CM
O
5
J cJ evi cvi






0
XI
1
©
"o.
E







-------
                                                                      78

three successive refinements were necessary to adjust the parameters a

and b.  The calculated value of the isotopic composition of the iodine

pentoxide-added oxygen remained constant during this calculation,

whereas the other values changed substantially between the first and the

second iterations (Table 2-5).  Convergence was rapid once values for

all terms were specified.

     The limits for convergence were chosen so that the total error due

to uncertainty in the correction parameters was smaller than the

observed precision of repeated measurements, s* = 2 x 10-' (see be low).

Convergence limits accepted for the various correction parameters were:

nQ , 0.05 wool 0; I>^QI 10-*; I§FI» 10-7; a, 0.002; and b, 0.02.  The
                              <;
standard deviation of calculated values of 1$FS will be less than 2 x

10-' under these circumstances, provided ns is greater than 5 wool 0 and

**F is within a factor of 2 of-all other components.  In general* the

mathematical correction is most sensitive to 1'FI, and this value must

be highly accurate under all conditions.  Uncertainties in the terns

specifying the contribution of the memory are much less important when

memory effects are small.  Thus, uncertainties in the values of a and b

can be as large as 0.5 and 0.1, respectively, when samples within the

natural abundance range are analyzed.


Evaluation of the Correction Procedure

     The correction procedure outlined here incorporates some assump-

tions that are essentially untestable.  We cannot be certain, for

example, that sample-derived oxygen retained in the sorbed-oxygen pool

does not differ in isotopic composition from that in the initial sample.

Although it would be extremely difficult to show that all such assump-

tions are correct, it is possible to show at least that the correction

-------
                                                                      79
procedure performs adequately.  Three questions arise and will be
discussed in sequence: i) is the functional form of the correction
procedure essentially correct, ii) what levels of accuracy and precision
can be obtained, and iii) are correction parameters derived with one set
of test samples generally applicable?
      Validation of the Model.  Results summarized in Figures 2-8 and
2-9 show that the correction procedure has accurately adjusted results
obtained during experimental runs designed for evaluation of correction
parameters.  The open symbols in both figures are the results of these
calculations.  The correction procedures accurately describe the
contribution of the blank over a range of sample size from 2 to 30 umol
0.  The means of the calculated values of 1§FS in both sets of data
(open symbols in Figure 2-8), 1'FBA_0 * 2.03512 ± 0.00030 x 10*J and
ltpBA-l = 2-04768 * 0.00028 x 10-' (x ± S.E.), do not differ
significantly (p < 0.05) from the true isotopic compositions of both
materials, as determined by the pyrolytic decarboxylation procedure.
The effect of the blank on the isotopic composition of the carbon
dioxide (the filled symbols) becomes more pronounced, shifting 2«1>FCO
further from the y-intercept, as the size of the sample decreases.  A
comparison of the slopes for each curve also reveals that the effect of
the blank is, as expected, greater for benzoic acid-1.  However, the
correction procedure effectively eliminates these variations.
     The effects of the memory are also adequately described and taken
into account by the mathematical correction.  The exponential relation-
ship, a(j) , closely described the wash-out of samples 9 and 16 in the
total oxygen pool, with a correlation coefficient for a least-squares
fit of r» - 0.995 (Figure 2-9).  Thus, equation 2-23 can be used to

-------
                                                                      80



estimate n<, and to provide an accurate indication of the amount of a



particular component in the sorbed-oxygen pool for the memory correc-



tion.  The open symbols in Figure 2-9 are the calculated values of lfFs



for samples 10-15 and 17-23.  Only the values for samples 10 and 11



differ significantly (p > 0.05) from the mean of samples 1-8, which is



unaffected by the memory.  This example is an extreme demonstration of



the memory effect, the great difference in the isotopic compositions of



the two materials resulting in a large shift in the abundance of



oxygen-18 in carbon dioxide produced from samples following each



enriched sample.  The effect of the memory will be small and more easily



guantitated when samples of similar isotopic composition are analyzed.



      Accuracy and Precision.  Accurate results were obtained in



analyses of the benzoic-acid standards (Table 2-6).  The correction



parameters were refined using the sample designated 0; the only value



found to differ from those in the original parameter set was that for



1(F0.  The mean isotopic abundances of the remaining samples (1 through



4) were compared to the results obtained by the pyrolytic



decarboxylation procedure.  A paired t-test indicates that the only



significant difference between calculated and true isotopic compositions



occurs for sample 1 (p > 0.05).  This discrepancy, albeit small, is



possibly the result of impurities in this sample, which is unique in



that it was used "as supplied," and was not equilibrated with water and,



consequently, was not recrystallized.  If the impurities happened to



yield CO, on pyrolysis, the pyrolytic result could have been affected.



     The precision of the analyses in Table 2-6 is typical of this tech-



nique.  In general, precision decreases as the sample becomes more



enriched or depleted in oxygen-18 with respect to the blank and  1,0,-

-------
                                                                      81
    Table 2-6:  Comparison of Isotopic Ana]yses of Benzole Acids
                    by Pyrolytic Decarboxylation and by the
                        Schutze-Untarzaucher Procedure
                          Pyrolytic   a
                       Decarboxylation

                         1IFBAX10>    .
                              Schutze-  b>c>d
                            Unterzaucher
                             1SFBA x 101
Benzoic Acid-0
2.03548 ± 0.00015
2.03544 ± 0.00022
      (-0.02)
Benzoic Acid-1
2.04756 ± 0.00009
2.04957 ± 0.00018
      (+0.98)
Benzoic Acid-2
2.11705 ± O'.OOOIS
2.11782 ± 0.00087
      (+0.36)
Benzoic Acid-3
2.18844 ± 0.00015
2.18976 ± 0.00095
      (+0.60)
Benzoic Acid-4
2.24562 ± 0.00040
 2.24429 ± 0.00066
      (-0.59)
 oean ± S.E. of 5 replicate analyses
b
 mean ± S.E. of 4 replicate analyses

 The correction parameters were optimized for Benzoic Acid-0:
 nQ  « 1.0 wnol 0, IIFQ = 1.977 x 10-J, »»FT « 1.9740 x 10-»,
 a » 0.081, and b « -1.60.  All other isotopic abundances were
 calculated with these values.
                             /l«F          -1> X
          Schutze-Unterzaucher    pyrolysis   '
                           1000

-------
                                                                      82



added oxygen.  This occurs because small variations in the contaminant



oxygen pools have, under these circumstances,  a more pronounced effect



on the calculated result.  For samples in the  range of natural abundance



(e. g., benzoic acid samples 1 and 2), the standard deviation of *•? is



typically 2 x 10-' (equivalent to 0.1 %„).  The trend toward increasing



standard deviations is evident in the results  obtained from the more



highly enriched samples.



      General Applicability of Correction Parameters.  Sanple matrix



effects have been identified.  That is, the chemical form in which the



oxygen is introduced can affect the analytical result.  Differences in



the reactivity of the initial oxygen-containing pyrolysis products with



the carbon, which persist even at high temperatures (32), and



interaction of non-oxygen containing gases such as Na, K,, and H,S (18,



20} with the surface oxide are probable causes.  Doping the carrier gas



(i. e., with hydrocarbons and halocarbons) minimizes these effects, but



does not eliminate them completely.



     An example of these effects is shown in Figure 2-10, which plots



2*"FCO  vs. l/ns for a series of analyses of acetanilide, benzoic acid,



and glycine (filled symbols, lower panel).  Three different curves,



indicative of different values for Tin, are obtained.  These differences



are particularly clear in the upper panel of Figure 2-10, which plots



the calculated (symbols) and the true  (lines) isotopic compositions of



each sample.  The causes of these differences are not known in detail,



but it is notable that, to obtain the corrected values in the upper



panel, the correction parameters have been optimized for benzoic acid,



which yields COj as its principal pyrolysis product, whereas acetanilide



yields CO and glycine yields a mixture of CO and CO,.  The deviations

-------
                                                                          33




83
     Figure 2-10.  A plot of 2-1»Fco  vs l/ng for benzole acid,  glycine,



and acetanilide; corrected for the IaO,-added oxygen:  benzole acid (©),



glycine 3D, and acetanilide (A); corrected for the blank, the memory,



and the I20,-added oxygen: benzole acid (O), glycine O, and acet-



anilide (A).  The correction parameters were optimized for the benzole



acid.

-------
                                                                                       84
      2.07  -
IO
 O
&    2.06 h
      2.05

      406
r 	 1 	 1 	 1 	 1 	 1 	 1 	 ' 	 1
A
A
- A
a
: a


O
- , 9 ...
1
•
-
p"

a ;
B:

A I
„ '
.
      4.05
10
 O
 CVJ
      4.04
      4.03
          0.0
O.I
                                        J_
 0.2

l/ns
                                                       0.3
                                              0.4
0.5

-------
                                                                      85



for acetanilide and glycine at the smaller sample sizes *ie plausibly



the result of differences in the size of the blank for CO and C02.   The



presence of N, apparently also influences the blank.



     The effects of nitrogen and sulfur on analytical results are



evident in the analyses of heteroarorcatic carboxylic  acids (Table  2-7).



The correction parameters were established using benzole acid, which, of



course, yields no sulfur- or nitrogen-containing pyrolysis products.



The results calculated for the Schutze-Unterzaucher procedure for  the H-



and S-containing materials differ significantly from  the accurate  values



determined by pyrolytic decarboxylation.  Sensitivity to heteroelement



content is also observed for the blank in oxygen elemental analysis



using the Schutze-Unterzaucher procedure; the blank is found to be



smaller for C,H,0 compounds than for C,H,N,0 or C,H,0,S compounds



(18-20, 33).  The differences seen in Table 2-7 arise from an increase



in the size of the blank due to the presence of nitrogen or sulfur in



the sample.  Paired t-tests indicate that all differences are



significant (p > 0.05) if n  « 1.0 wool 0, but that the results of the



two procedures do not differ significantly (p < 0.05) if n » 3.0 insol 0.



The mathematical model accurately reflects dependence of the size of the



blank on heteroelement content of the sample.  The importance of



choosing a working standard which closely matches the composition and



structure of the sample in order to correctly establish the size of the



blank and obtain the greatest accuracy is thus demonstrated.





Standards



      Analysis of Standard Waters and Carbohydrates.   The primary



standard for oxygen isotopic abundances is "Standard Mean Ocean Water"



(34).  Samples of sucrose intended as supplementary standards for the

-------
                                                                      66
Table 2-7:  Comparison of Isotopic Analyses of Heteroaroaatic Carboxylic
                  Acids by Pyrolytic Decarboxylation and by the
                        Schutze-Unterzaucher Procedure
                      Pyrolytic   a               Schutze-  b>c
                   Decarboxylation              Unterzaucher
                      up x 10'                   *«F x 10»
                                                     A

                                                 d                  c
                                         n  » 1.0           n  « 3.0
   pyrole-2-
carboxylic acid  1.97134 ± 0.00006  1.97756 ± 0.00018  1.97229" ± 0.00018
                                         (+3.15)            (+0.51)
  2-thiophene                 s
carboxylic acid  1.97804 ± O.C0005  1.98241 ± 0.00018  1.97771 ± 0.00020
                                         (+2.21)            (-0.17)
benzoic acid-0   2.03548 ± 0.00015  2.03530 i 0.00039. 2.03557 ± 0.00041
                                          (-0.09)             (+0.04)
 2,6-pyridine
carboxylic acid  2.03740 ± 0.00017  2.03827 1 0.00010  2.03875 ± 0.00012
                                          (+0.43)             (+0.66)


 mean ± 5.E. of 5 replicate analyses
b
 mean ± S.E. of 4 replicate analyses

CA' ^''Schutze-ttiterzauche/^pyrolysis-" « 100°
d
 n0» 1.0 Mmol 0, all other correction parameters were optimized for
 behzoic acid: 1§F0« 1.975 x 10-», »»FX » 1.98500 x 10-J, a -
 0.077, and b » -1.60.  All other isotopic abundances were calculated
 using these values.

 nQ« 3.0 MiBOl 0, all other correction parameters were optimized for
 bfchzoic acid-0: l'FQ- 1.975 x 10-', »»Fj» 1.98975 x 10-', a «
 0.077 and b « -1.60.  All other isotopic abundances were calculated
 using these values.

-------
                                                                      87



analysis of oxygen isotopes  in organic materials have been distributed



by I. Friedman (4).  Because sucrose should yield water as its



predominant oxygen-containing pyrolysis product (19), direct comparison



of results obtained for water and sucrose standards should be possible.



     Results of the analysis of cane and beet sugar (4) and standard



waters (35) are shown in Table 2-8.   The correction parameters were



optimized for V-SMOW (34).   A paired t-test indicates that the results



calculated for KBS-1 and NBS-1A do not differ significantly (p < 0.05)



from their accepted "0 abundances (34).  The calculated values for cane



and beet sugar are both slightly greater than previous measurements (4-



6,36).  At least four different laboratories have analyzed these



samples, each reporting a slightly higher oxygen-18 content than the



previous measurement, and each accounting for the discrepancy by virtue



of reduction of blank and memory effects in their respective procedures.



Our measurements extend the trend.  The accuracy of our results for NBS-



1 and NBS-1A in the same series of analyses suggests that the effects of



the blank and of the memory have been eliminated in the mathematical



correction and that the isotopic abundances for cane and beet sugar



reported in Table 2-6 are,  in fact,  correct.



      Calibration of Working Standards.  Although small variations in



the size of the blank are unimportant if large samples are used, the



effects of these variations can be minimized for samples of all sizes if



working standards yielding pyrolysis products similar to those of the



sample are available.  The provision of such standards is always



possible because the accurate isotopic composition of any sample can b«



determined by obtaining multiple analyses and fitting a curve similar to



that shown in Figure 2-4.  Such multiple analyses are tedious, but need

-------
                                                                      88
      Table 2-8:  Analysis of Water Standards and Sucrose Samples
                                    a,b
                      Accepted value                   Schutze-  c
                                                     Unterzaucher
                        »»P x 10J                       »»P x 10*
                      (S^'O vs SHOW)                 (&**0 vs SHOW)
                           °/6o                            %c
cane sugar            2.0627 - 2.0685            2.06978 ± 0.00073
                      (+31.3 - +34.1)                 (+34.7)

beet sugar            2.0489 - 2.0547            2.05373 ± 0.00055
                      (+24.3 - +27.2)                  (+26.7)

V-SMOW                    2.00044                2.00042 ± 0.00004
                          (0.00)                      (-0.01)

KBS-1                     1.98499                1.98496 ± 0.00035
                          (-7.74)                     (-7.75)

NBS-1A    '                1.9519                  1.9526 ± 0.0010
                          (-24.3)                     (-24.0)

a
 Samples o2 cans and beet sugar were provided by Dr. I. Friedman (4).
 The range indicates results obtained by other investigators (4-6,32).
b
 The absolute abundance of oxygen-18 in V-SMOW was measured by
 Baertschi (30), and the abundances of oxygen-18 in NBS-1 and NBS-1A
 relative to V-SMOW were measured by Gonfiantini (31).

 Results of the triplicate analysis of samples containing approximately
 50 junol 0.  The correction parameters were optimized for V-SMOW:
 n « 1.2 wnol 0,l'FQ » 1.977 x 10-», 1§FT « 1.9890 x 10-',
 a = 0.077, and b B -1.50.  All other isotopic abundances are
 calculated with these values.

-------
                                                                      89




not be repeated once a standard allowing adequate determination of




correction parameters has been established.  As an example,  a plot of




2'1'FCQ   vs. l/ns for benzoic acid,  glycine, mouse lung hcmogenate,  and




mouse blood is shown in Figure 2-11.   Samples of benzoic acid were used




to determine 1§Fj and the values of l*Fg for each of these materials




were determined from the y-intercept.  These materials can now be used



as working standards in the analysis  of similar tissue samples, where




the values of nQ and 1*FI for a particular set of analyses are optimized




by the procedure previously discussed in the section titled "Adjustment




of Analytical Results".






                           EXPERIMENTAL SECTION



Isotopic Standards                '




      Pyrolysis of Benzoic Acid. .-Benzoic acid (elemental analysis




standard, Erba, Inc.) was pyrolyzed in sealed, evacuated quartz tubes.



Prior to use, the quarlr tubes (o.d.  » 6 mm, length « 15 cm) were heated




in air at 550»C for 12 hours, soaked  in concentrated hydrochloric acid



for 12 hours, and again heated in air at 550*C for 12 hours.




Immediately before use, each tube was evacuated, heated to 800*C with a



flaae, and allowed to cool under vacuum.  A sample of benzoic acid,



approximately 3.6 mg (30 wnoles), was weighed to the nearest ug with  a




Cahn electrebalance.  The tube was vented, the sample was added, and  the




tube was then evacuated for 60 to 120 sec prior to sealing.  The sealed,




evacuated tube was placed in a preheated muffle furnace for the




specified time, removed from the furnace and allowed to cool in air.




     In order to investigate reaction 2 under the conditions of this




work, approximately 30-ttnole aliquots of carbon dioxide (99.99%,

-------
                                                                           90
90
     Figure 2-11.  A plot of 2-1§Fco vs l/ns for benzole acid (©),  acet-
anilide (^), mouse blood 03>, and mouse-lung homcgenate (A).  All results
were corrected for the I,0t-added oxygen.

-------
91

-------
                                                                      92



Matheson) were sealed in quartz tubes cleaned as described above and



placed in a muffle furnace, preheated to 550'C, for times from 0 to 60



minutes.  The tubes were removed from the furnace and allowed to cool in



air.



      Preparation of Carbon Dioxide.  Carbon dioxide from sealed-tube



experiments was purified by cryogenic distillation.  The yield of C03



was measured using a gas buret and mercury manometer.



      Mass Spectrometric Analysis.  Isotopic analysis of carbon dioxide



was performed with a Nuclide 6-60 RMS isotope-ratio mass spectrometer



and isotopic abundances were calculated as described in Chapter 3.



Pyrolysis products other than carbon dioxide were identified using a



Kratos MS-80 high-resolution mass spectrometer.



      Theoretical Calculations.  The equilibrium constant (K) for •



reaction 2 for temperatures between 300 ar.d 1500'K was calculated from



values of the enthalpies and entropies of formation of CO and CO, (37)



and values of the heat capacities of CO, CO,, and graphite (36).  The



equilibrium oxygen isotopic fractionation factor for this reaction was



calculated by means of the Bigeleisen-Mayer formulation (39), with



literature values of the force constants for CO and CO, (40).  These



results were used to predict '"he yield of carbon dioxide and the shift



in isotopic composition of the carbon dioxide from that of the carboxyl



group.



      Isotopic Standards.  Five samples of benzoic acid with different



isotopic compositions were prepared by exchange with water (41).  For



each of five isotepically different waters, approximately 4 g of benzoic



acid were added to 100 mL of water.  Gaseous HC1 was bubbled through the



mixture until the concentration of HC1 was 5 N.  The mixture was sealed

-------
                                                                      93



under an atmosphere of N, and kept at 80»C fcr 120 hours.   The benzole



acid was extracted in ether and the ether was removed under vacuum.



Samples of 2,6-pyridinedicarboxylic acid, pyrrole-2-carboxylic acid, and



2-thiophenecarbcxylic acid (Aldrich) were used as purchased.  In order



to determine their oxygen-isotopic compositions, the products were



pyrolytically decarboxylated in sealed,  evacuated quartz tubes at 550eC



for 10 minutes.  7he carbon dioxide was purified by cryogenic



distillation and tfe abundance of urygen-18 in the CO, was measured mass



spectrometrically.





Oxygen Isotopic Analysis



      Apparatus.  A Carlo Erba model 1106 elemental analyzer (Erba



Instruments, Inc.) was modified for the preparation of organic samples



for isotopic analysis (Figure 2-12).  These modifications do not



interfere with the normal operation of the elemental analyzer, a



description of which has appeared elsewhere  (42,43), and allow the



preparation of CO, for the determination of oxygen-18 or for the



preparation of CO,, N, and H,0 for determination of carbon-13,



nitrogen-15, and deuterium, respectively.  Here, only procedures



pertinent to oxygen isotopic analysis will be discussed.



     The effluent of the chromatographic column was redirected so that



it passed through an oxidation reactor before entering an auxiliary



vacuum line for collection, purification, and packaging of  carbon



dioxide.  The oxidation reactor was a pyrex  tube, 10 mn o.d. x 40 cm  in



length, 30 cm of which were packed with  an iodine pentoxide reagent



(44).  The remaining 10 cm were filled with  dry  silica gel.  The  I,0S



was heated to 135'C.  A  series of traps  and  a vacuum  system were



constructed, so that carbon dioxide could be  trapped from  the effluent of

-------
                                                                            94
94
     Figure 2-12.  Schematic diagram of modifications to the Carlo Srba

elemental analyzer used for the preparation of carbon dioxide from

organic material for isotopic analysis.  Details of the operation of the

elemental analyzer appear elsewhere (42).  Valves V2, V3, and V4 were

connected to the high-vacuum system at positions A, B, and C.  Carbon

dioxide was trapped' from the affluent of the 1,0, reactor in T3.

-------
                                                                                     95
-g
v
cr

-------
                                                                      96



the oxidation reactor, purified,  and transferred to a sample tube for



mass spectrometric analysis.  All valves were stainless steel and



utilized l/8"-low dead volume fittings (Valco Instruments Co.; VI, V2,



and V3 were type HP-6a; V5 was type HP-4a; and V4 was type HP-lOa).



Traps Tl, T2, T3, and T5 were multicoiled, containing 6 loops each, and



were constructed from type 321 stainless steel tubing, 0.125" o.d. x



0.010" wall (Tube Sales).  Trap T4 was constructed from 6-mm pyrex



tubing and contained 0.1 g of molecular sieve 5A, 60/80 mesh ^Analabs).



The connections between the valves were made with teflon-lined aluminum



tubing, 1/8" o.d. x 1/16" i.d. (Analabs).  Connections to the highvacuum



system were made using one-piece, graphite-impregnated Vespel ferrules



(Supeltex M-2A, Sup«lco).  Standard two-piece Swagelok Teflon ferrules



(Crawford Fitting Co.) were used for other connections.



      Reagents.  High-purity helium, 99.995% (Air Products) was used as



the carrier gas for the elemental analyzer.  The carrier gas was doped



with approximately 550 ppm of CC14 and 1500 ppm of C7H,,.  These



materials ware mixed with the helium by volatilization from a toroidal



pyrex vessel with a 1-mm orifice, containing a 5% (v/v) solution of



carbon tetrachloride  (Fisher, Spectranalyzed) in n-heptane (Fisher,



Spectranalyzed).  The vessel was placed in the sample dispenser, just



upstream from the reaction chamber.



     Iodine pentoxide was prepared by mixing a solution containing 50 g



of 1,0,  (Mallinckrodt Analytical Reagent) in 100 mL of distilled water



with 100 g of 8-12 mesh silica gel  (Preiser Scientific, Inc.).  This



mixture was allowed to dry at 80°C  for 96 hours.  Concentrated  sulfuric



acid (25 mL) was added to the dried material and the mixture was again



allowed  to dry.  Finally, the mixture was heated to a temperature  of

-------
                                                                      97



220»C for 4 h in a dry helium stream (44).  The active reagent, a ccarse



yellow material, was kept dry and not exposed to light.



      Isotopic Analyses.  SampJes were weighed into silver-foil cups



(Erba, Inc.) on a Cahn 25 Electrobalance (Cahn Instruments, Inc.)  and



placed in the s.impling-head of the instrument.  The analyzer was



cperated normally, so that oxygen elemental abundances in .the samples



were also obtained.  A dry ice-acetone bath was placed around Tl to



remove residual amounts of I,, a byproduct of the oxidation of CO by



1,0,, which were present in the carrier gas.  To prevent aerosol forma-



tion and ensure quantitative recovery, T3 was ixnaersed to half its



length in liquid nitrogen when trapping carbon dioxide from ths analyzer



effluent.  Valve V3 was switched to-direct the effluent of the oxidation



reactor through T3 60 sec after a carbon dioxide peak was detected by



thermal conductivity detector Dl (the delay was determined using D2).



Helium was removed and the carbon dioxide was transferred to a sample



tube for mass spectrometric analysis.  Isotopic analysis of carbon



dioxide is described in Chapter 3.

-------
                                                                      98

                               REFERENCES

 1.  Mook,  W.  G.;  Grootes, P. M.  The measuring procedure and
     corrections for  the high-precision mass-spectroaetric analysis  of
     isotopic  abundance ratios, especially referring  to  carbon,  oxygen,
     and nitrogen. Int. J. Mass Spectrom. Ion Phys.  12:273-298,  1973.

 2.  Taylor, W.; Chen, I-J.  Variables in oxygen-18 isotcpic  analysis  by
     mass spectronetry.  Anal. Chem. 42:224-228,  1970.

 3.  Fong,  B.  J.;  Smith, S. R.; Tanaka, J.  Reevaluation of the  prepara-
     tion of organic  compounds for mass spectrometric analysis for
     oxygen-18.  Anal. Chem. 44:655-659, 1972.

 4.  Hardcastle, K. G.; Friedman, I.  A method for oxygen isotope
     analysis  of organic material.  Geophys. Res. Lett.  1:165-167,
     1974.

 5.  Per hi, A. M.; Letolle, R. R.; Lerman, J. C.  Oxygen isotope ratios
     of organic  matter: Analysis of natural compositions.  Proceedings
     of the Second International Conference on Stable Isotopes,  716-724,
     1976.

 6.  Thompson, P.; Gray, J.  Determination of  II0/»»0 ratios  in
     compounds containing C, H, and O.  Int. J. Appl. Hadiat. Isot.  23:
     411-415,  1976.

 7.  Brenninkmeijer,  C. A. M.; Hook, W. G.  A batch process for  direct
     conversion  of organic oxygen and water to CO, for  l»0/140 analysis.
     Int. J. Appl. Radiat. Isot. 32:137-141 1981.

 8.  Unterxaucher, J. The direct micro-determination of oxygen  in
     organic  substances.  Analyst 77:584-595,  1952.

 9.  Schutze,  M.  Die direh.e bestimmung des Sauerstoffs in Zinkoxyd.
     2. Anal.  Chem. 118:241-245, 1939.

10.  Winter, K.; Barton, D.  The thermal decomposition  of benzoic acid.
     Can. J. Chem. 48:3797-3801, 1970.

11.  Maier, W. P.; Roth, R. V.; Thies,  I.; Schleyer,  P.  v. R. Gass
     phase decarboxylation of carboxylic acids.   Chem.  Ber. 115:808-115,
     1982.

12.  Rhead, T. F.  E.; Wheeler, R. V.  The effect  of temperature  and  of
     pressure  on the  equilibrium 2CO ~   — CO, * CO.  J. Chan.  Soc. 99:
     1140-1153,  1911.

13.  Stevens,  K. H.;  Pepper, J. M.; Lounsbury, M.  The  decarboxylation
     of anthranilic acid.  Can. J. Chem. 30:529-540,  1952.

14.  Dunn,  G.  E.;  Dayal, S. K.  Mechanism od decarboxylation  of
     substituted anthranilic acids at high acidity.   Can. J.  Chem.
     48:3349-?353, 1970.

-------
                                                                      99

15.  Dunn, G. E.; Lee,  G.  K.  J.   Kinetica and mechanism of the
     decarboxylation of pyrrole-2-carboxylic acid in aqueous solution.
     Can. J. Chen. 49:1032-1035,  1971.

16.  Dunn, G. E.; Gordon,  K.  J.  L.;  Thimm, L.  Kinetics and mechanism of
     decarboxylation of some  pyridinecarboxylic acids in aqueous
     solution.  Can. J. Chen.  50:3017-3027, 1972.

17.  Rhead, W. F.; Wheeler, R.  V. The rate of reduction of carbon
     dioxide by carbon. J. Cheia. Soc.  101:831-845,  1912.

18.  Haraldson, L.  An  investigation of the products of pyrolysis in the
     determination of oxygen  in sulphur-containing organic substances.
     MikrochiB. Acta 1962, 651-670.

19.  Belcher, R.; Ingram,  G.; Majer, J. R.  Direct determination of
     organic compounds  in  the determination of oxygen.  Mikrochim. Acta
    . 1968, 418-426.

20.  Belcher, R.; Ingram,  G.; Majer, J. R.  Direct determination of
     oxygen in organic  materials-I.   Talanta 16:881-892, 1969.

21.  Bonner, F.; Turkevich, J.  J. Study of the carbon dioxide-carbon
     reaction using C14 as a  tracer.  J. Am. Chem. Soc. 73:561-564,
     1951.  '

22.  Oring, A. A.; Sterling,  E.   Oxygen transfer between carbon dioxide
     and carbon monoxide in the presence of carbon.   J. Phys. Chem. 58:
     1044-1047, 1954.

23.  Ergun, S.  Kinetics of the reaction of carbon dioxide with carbon.
     J. Phys. Chen. 60:480-485,  1956.

24.  Vastola, F. J.; Hart, P. J.; Walker, P. L.  A study of carbon-
     oxygen surface complexes using  Oi§ as a tracer.   Carbon 2:65-71,
     1964.

25.  Manser, M; Ergun,  S.   Kinetics  of oxygen exchange between CO, and
     CO on carbon.  Carbon 5:331-337, 1967.

26.  Grabke, H. J.  Oxygen transfer  and carbon gasification in trie reac-
     tion of different  carbons  with  CO,.  Carbon 10:587-599, 1972.

27.  Treoblay, G.; Vastola, F.  J.; Walker, P. L.  Thermal desorption
     analysis of oxygen surface complex on carbon.  Carbon 16:35-39,
     1978.

28.  Gouverneur, P.; Schreuders, M.; Degens, P. N.  The direct
     determination of oxygen  in organic compounds.  Anal. Cl^im. Acta 5:
     293-312, 1951.

29.  Tamai, Y.; Watanabe,  H.; Totaita, A.  Catalytic gasification of
     carbon with steam, carbon  dioxide, and hydrogen.  Carbon 15:103-
     106, 1977.

-------
                                                                     100

30.  Xirsten, W. J.  Micro and trace determination of oxygen in organic
     compounds.  Anal. Chim.  Acta 100:279-288,  1978.

31.  Freeman, G. B.; Reucroft, P. J.  Adsorption of HCK and H,0 vapor
     mixtures by activated and impregnated carbons.  Carbon 17:313-316,
     1979.

32.  Hartley, P. E.  Corrections for memory effects in analytical
     instruments.  Anal. Chem. 54:148-150, 1962.

33.  Belcher, R.; Ingrain, G.   Some observations on the microdetermina-
     tion of oxygen in organic materials.   Microchem. J. 11:350-357,
     1966.

34.  Baertschi, P.  Absolute  "0 content of Standard Mean Ocean Water.
     Earth Planet. Sci. Lett. 31:341-344,  1976.

35.  Gonfiantini, R.  Standards for stable isotope measurements in
     natural compounds.  Nature 271:534-536, 1978.

36.  Wedeking, K. W.; Hayes,^J. M.  unpublished results.

37.  Rossini, E. D.; Wagman,  D. W.; Evans, W. H.; Lavine, S.; Jaffe,  I.
     Selected values of thermddynamic properties, National Euro of
     Standards Circular 500,  1952.

38.  Spencer, H. M.  Emperical heat capacity equations of gasses and
     graphite.  Ind. Eng. Chem. 40:2152-2154, 1949.

39.  Bigeleisen, J.; Mayer, M. G.  Calculation of equilibrium constants
     for isotope exchange reactions.  J. Chem.  Phys. 15:261-267, 1947.

40.  Stern, M. J.; Spindel, W.; Monse, E.  V.  Temperature dependence of
     isotope effects.  J. Chem. Phys. 48:2908-2919, 1968.

41.  Roberts, I.; Urey, H. C.  Kinetics of the exchange between benzole
     acid and water.  J. Am.  Chem. Soc. 61:2580-2584, 1939.

42.  Pella, E.; Colombo, B.  Improved instrumental determination of
     oxygen in organic compounds by pyrolysis-gas chromatography.  Anal.
     Chem. 44:1563-1671, 1972.

43.  Pella, E.; Colombo, B.  Study of carbon, hydrogen, and nitrogen
     determination by ccmbustion-gas chromatography. Mikrochim. Acta
      Wien  1973, 697-719.

44.  Smiley, W. G.  Note on reagent for oxidation of carbon monoxide.
     Nucl. Sci. Abstr. 3:391-392, 1949.

-------
                                                        101
                  CHAPTER 3

MASS SPEC7ROMETRIC DETERMINATION OF CARBON-13,
  OXYGEN-17, AND OXYGEN-18 IN CARBON DIOXIDE

-------
                                                                     102



     Measurement of the mass spectrum of a volatile molecule can



quantify only the relative abundances of  molecular species  of



differing mass.  Commonly, elemental isotope ratios are not measurable



directly, but must be determined by calculations based on the observed



molecular spectrum (1).  In the case of carbon dioxide, for example, ion



currents appear at m/z 44 through m/z 49.  With the exception of m/z 44



(11C1»0,), all measurable ion currents represent sums of multiple



isotopic contributions (e. g.,  »»C1T01«0 + »»C1<01'0 + 13C1T0, at m/z



46).  A problem common to all mathematical formulations (2-4) is that a



measurement of two ion-current ratios (45/44 and 46/44) does not provide



sufficient information to obtain unique solutions for three isotope



ratios (i3C/12C, 1?0/140, and "O/^'O).



     The oxygen-isotope ratios have, therefore, commonly been treated as



functionally related; that is,  an expression for 170/1'0 as a function



of '•0/ltO-has been written and used as the third equation required for



solution of the problem.  This approach is based on the fact that proc-



esses (other than admixture of separated isotopes) which have affected



the abundance of 1»0 must also have affected the abundance of 1-I0, and



vice versa (5).  Recently, however, it has been shown (i) that no



single, fixed relationship correctly describes the fractionation of J'0



relative to 1»0 in all processes, and (ii) that the relationship



commonly employed by isotopic mass spectrometrists does not correctly



describe even the "average relationship" prevailing in natural systems



(6).  Here, we explore the significance of these findings for isotopic



analysis and show that a simple system of equations for accurate



calculation of isotopic abundances can be developed.

-------
                                                                     103


                 DEVELOPMENT OF ISOTOPE-RATIO EQUATIONS


      Notation.  The letter R is consistently used to designate the


ratio of a particular isotopic species to the related isotopic species


of lowest mass
                           isotopic species of mass n
                      R *	             (3-1)
                            base species related to n
Left superscripts will be consistently used to designate isotopic

species by mass.  For n * 13, the base species is 11C; for n * 17 or 18,

*«0; and, for r. » 45 through 49, the base species is 11C1»0,.  The

letter F is consistently used^in the sane way to designate fractional

abundances
                          isotopic species of mass n
                   F - 	-r—          (3-2)
                        sum of all species related to n
for n » 13, the denominator is the sum (1SC •*• 11C); ior n • 17 or 18,


(i*0 •+• i'o * 1§0); and, for n« 45 through 49, the denominator is the sum


of all species of carbon dioxide.  The symbol a is used to designate the


fractionation factor, in this case, the ratio of isotope ratios


characteristic of the isotope effect associated with a physical or


chemical process

                                n
                                 "products
                          o « 	                      (3-3)
                               nR
                                "reactants

for n = 13, 17, or 18.


      Isotopic Distribution.  The isotopic species present in carbon


dioxide are shown in Table 3-1.  Since the vibrational modes in CO, are


nearly harmonic and coupling between modes is weak (7), the distribution

-------
                                                               104
  Table 3-1: Isotcpic Molecular  Species  of  Carbon Dioxide
44          45          46          47           48          49

-------
                                                                     105




a particular isotopic species is accurately described by a simple




probability function <1).  The abundances of the various molecular




species are, therefore, related to the isotopic abundances by the




following equations






                         44p » I2f.ltf.ltf                      (3-4)







                 4»p s up.itp.itF 4. 2-1 IF-*- •F- * 'F              (3-5)







         44p B itf.iif.iif + 2*l2F-**F-l'F + 2«13F>1*F*J'F      (3-6)







         tif m iif.tif.nf 4- 2- * JF' * *F* * *F * 2- 13F*17F' * *F      (3-7)






                 4*p B iap.np.i»p ^ 2>"F*I'F*1*F              (3-8)






                         4»F . up.i.p.i.p                      (3_9)







In practice, it is nost convenient to express these relationships in




terms of ratios.  Dividing equations 3-5  through 3-9 by equation 3-4




yields






                         4.R » 1JR * 2-»'R                      (3-10)







                   44R « 2-l«R * 2-»JR-»'R •«• »'R»               (3-11)






              47R « 1JR.1TR1 4. 2-»>R-»«R + 2-17R«>»R            (3-12)







                     ««R - >«R» + 2-1>R'»'R'*»R                 (3-13)






                           4»R . iJR.liRJ                       (3-14)







For samples with isotopic abundances near natural levels, ion currents




large enough to allow precise, rapid measurement occur only at masses




44, 45, and 46.  Modern instruments employ a triple-collector system to

-------
                                                                     106


facilitate simultaneous measurement of  *'R and 44R (Figure 3-1).

Relationships Between 1J>R and **R

      Theoretical Considerations.   Fractionations of multiple isotopes

of the same element are proportional to differences in mass (5).   A

quantitative relationship between the values of »'R and 1§R for two

oxygen pools (designated by subscripts  1 and 2) linked by a single,

mass-dependent process has been presented by Clayton and coworkers (6)
                                                                  (3-15)
where a « ln(* 'a)/ln(» §a) .  Note, however, that (i) a is not constant
                               4 !
and  (ii) equation 3-15 refers to instantaneous precursor -product

relationships, not to the integrated compositions cf separated or mixed

isotopic pools.  Both of these points are significant in the isotope-

analysis problem.

     The value of o depends on the process causing the isotopic

fractionation.  Further, variations of »*a and »«o are not linked in a

way  that leads a to assume a constant value.  As a result, relative

fractionations of the oxygen isotopes can depend on the molecular

species involved or on the mechanism of a particular reaction.  In

general, 0.50 < 2 < 0.53 (6).  For a series of reactions, the value of a

for  the overall process is the weighted average of the values for each

step

                              I{a InO'cO}
                         a  . -                       (3-16)
 where a^ is  the  exponent and »»ai is the cxygen-18 fractionation

-------
                                                                           107
107
     Figure 3-1.  Schematic representation of an isotope ratio aass

spectrometer.  Reference and sample gases are alternatively introduced

into the mass spectroneter, and ion-currents at m/z 44, tn/z 45, and m/r

46 are measured for each gas.

-------
               108
^^   r^   Oy

~~o   o ~o
ro   cvj   cvj
  o   o
ro   eu
          en



     8   >
     D   %-
          o>
         cr
            ^^^^_Zk


            ?
            ">•





            ?

-------
                                                                     109



factor for each step in the series of n reactions.   In cases where a



reaction sequence imparts a large net change in the relative abundance



of one heavy isotope and a small net change in the  abundance of the



other heavy isotope, the value of a2 is not restricted to the range



defined by the az, and may be very large or very small.  Moreover, if a



net enrichment in one isotope accompanies a net depletion in the other,



then as will be negative.



     Natural samples are never the products of a discrete process, or



even a single sequence of processes.  For polyisotopic elements,



isotopic variations associated with mixing processes are complex.  On



the one hand, relationships governing isotopic fractionations associated



with mass-dependent isotope effecSts are correctly cast in terms of



ratios.  On the other hand, accurate calculation of the composition of



mixed pools must be based on fractional abundances.  Even if a is known



accurately, equation 3-15 will only approximately describe isotopic



relationships associated with the mixing of oxygen  pools with similar



isotopic compositions (i.e., natural samples of terrestrial origin).



In general, a  will be some average of the ai weighted by the relative



contribution of each component to the total oxygen  pool.  If one or more



of the oxygen pools involved in a mixture has been  selectively enriched



or depleted in a single isotope, serious inaccuracies can result from



application of equation 3-15 (6).



     Given these considerations, it is evident that there is nc basis on



which an equation  exactly  relating 11R and 1(R in all pools of oxygen



might be developed.  In particular, the relationship underlying most



oxygen-isotopic analytical calculations:

-------
                                                                     110

                                                                 (3-17)
where "unknown" and "standard" refer to a sample of unknown isotopic
composition and to some primary isotopic standard, is inexact and,
because 0.50 represents a minimal value for a, may not be the most
satisfactory approximation.
      1TR and **R in natural Samples.  Independent measurements
 of 61''0 and 6ltO in lunar and terrestrial silicate minerals and samples
of natural waters quite closely define a single relationship (6,8,9)

                     S^OSMOW" 0.52«6»»OSJ!ow                    (3-18)

As indicated by use of the d notation, this relationship is based on
measurements of differences between isotopic compositions of unxnown
samples and isotopic standards.  As such, it carries no direct
information regarding absolute values of isotope ratios.  Calculations
of 17R and 1§R for unknown samples — and, for that matter, accurat e
calculations based on equations 3-10 and 3-11 — require knowledge of
»'R and 1»R for the isotopic standards.  Although there have been recent
measurements of the absolute value of I*RSHOW ^and ****¥ are in 9CXX*
agreement, see refs 2,10,11), there have been no direct determinations
°f ItRSMOW
     It is, therefore, of great interest that the isotopic compositions
of all natural terrestrial and lunar oxygen pools appear to be related
by equation 3-18, which is closely related to equation 3-15.  The 6
terms are identical to the first terms in the series expansions of the
In terms in:

-------
                                                                     Ill



                                                                 (3-19)

                                        \  »-R       I
                       SHOW  /


Higher-order tern.s are negligible when l'Runknown anr^ l 'Unknown are


close to l1RsMOW zx^ llpSMOW-  Jt is evident that equation 3-19 is a


more general form of equation 3-18,  which can be recast as



                          *'R «  »«Ra-K                           (3-20)



where K is a constant characteristic of the relationship between 17R and


l»R in the terrestrial oxygen pool.   Because the isotopic compositions


of all samples appear to be^related by this equation, evaluation of K


requires only that both 1-IR a^d ^R-be known accurately in, a single,


representative sample.


     There has been only one set of measurements of the absolute


abundances of oxygen-16, oxygen-17,  and oxygen-18 in samples of


terrestrial origin.  Nier. determined values of the ratios ^C^'O/^'O,


(J)R) and (1101*0 * 11!0,)/l«0, (J«R) for samples of air and of 0, from a


commercial gas cylinder (12).  Discordant results were obtained for the


atmospheric sample (perhaps because total air, not purified atmospheric


0,, was admitted to the ion source), but, for the latter analysis, Nier


reports (corrected mean and probable error) *JR « 7.55 t 0.01 x 10-' and


"R » 4.103 ± 0.005 x 10-J, which yield isotope ratios (mean and


standard error of mean) of  l'Ro2 •  (3.775 ± 0.074) x 10-« and  l»Ro2 *


(2.0514 * 0.0037) x 10-».   It follows that, for a « 0.52, K «  (9.43 ±


0.18) x 10-».



Calculation of  Isotope  Ratios


     Combining  equations 3-10 and 3-11 and  substituting equation  3-20

-------
                                                                     112




for i'R yields an expression for »'R in terms of ion-current ratios,  K,




and a.  Apart from uncertainties in K and a,  the equation is exact.
                            2K'4iR-»«Ra •«• 2-1§R - «'R - 0        (3-21)





Rearrangement of this equation to yield an explicit solution of the form



l'R = f(a, K, 45R, «'R)  is not possible except where a is exactly 0.5.



An accurate value for 1(R can, however, be easily and rapidly obtained



using numerical techniques.  A value for 11R can be determined from



equation 3-20 or, if a direct solution in terms of ion-current ratios is



required, a derivation similar to that of equation 3-21 yields





               -3-i'R» + 2-«'R-"R * (l'R/K)1/Q  - «'R » 0        (3-22)





The calculation of J1R follows directly from «'R using equation 3-10.



     For clarity in subsequent discussions, we will refer to the



following sequence of calculations as "routine:"



     1.  »»R is calculated from «»R and ««R (eq. 3-21).



     2.  "R is calculated from »'R (eq. 3-20).



     3.  »'R is calculated from * 'R and «*R (eq. 3-10).





                   DETERMINATION OF ISOTOPE ABUNDANCES



     The accuracy of calculated isotope ratios will depend on how well



equation 3-20 actually describes the linked variations of 17R and 1§R



and on the development of representative values for a and K.  Here, we



will explore, by means of the application of equations 3-20 - 3-22 to



calculation of results from analytical data, two particularly



interesting cases:



          (i) oxygen pools derived by mixing natural-abundance material

-------
                                                                     113



     with material highly enriched in 1§0



          (ii) natural-abundance material with a range of oxygen-



     isotopic compositions wide enough that correct assignment of the



     values of a and K will be important



     Mixed Samples Enriched in **0.  If a sample has bean selectively



enriched in l»0, inaccuracies will result from use of the routine



sequence of calculations.  Because ««R is primarily dependent on 1»R,



solution of equation 3-21 will yield a roughly correct value for »'R.



The value of *'R will, however, be overestimated by equation 3-20.



Insertion of that overestimate in equation 3-10 will lead to an



incorrectly low value for IJR.  Further, use of equation 3-20 in the



derivation of equation 3-21 causes the initially calculated value of 1§R



(termed "roughly correct" above) to be low.  Because the relative



contribution of the term {2«l'S} to «*R is greater than that of



{2-i»R.iiR + »'Ra} to «'R, the error in »»R will be greater than the



error in 1§R.



     Clearly, this cascade of errors can be most appropriately prevented



by "uncoupling" "R and  »»R (that is, by not using equation 3-20 to



express the relationship between 17R and »»R).  Two approaches are



possible.  If 11R is known independently, the sequence of calculations



that we will term "»»C-known" can be employed:



     1.  1TR is calculated froa «»R and »»R (eq. 3-10).



     2.  *»R is calculated from «*R, »»R, and »7R (eq. 3-11).  If 1->R is



known independently, or if it is expedient to adopt an estimated value



in order to reduce systematic errors in the analysis of samples for



which neither 1JR nor "R is known, but in which »'0 has been enriched



well above natural abundance, a sequence of calculations that might be

-------
                                                                     114


termed "1'0-known" can be employed:


     1.  1JR is calculated from «'R and "R (eq. 3-10).


     2.  *'R is calculated from ««R, »>R,  and 17R (eq. 3-11).


     To explore these problems and possibilities, we have prepared ar.d


analyzed mass spectrometrically a series of carbon dioxide samples


enriched in the heavy isotopes of oxygen.   Oxygen in these samples was


derived by exchange with waters prepared by mixing 99% 1*0-water with


distilled tap water.  Because only oxygen-isotopic compositions could be


affected by exchange, the true carbon-12/carbon-13 ratio in all samples


was identical.
                            4'

     Results of the mass spectrometric analyses are summarized in Table

                              r»
3-2, which includes normalized ion-current ratios as we.ll as two


different sets of calculated'isotopic compositions.  The columns headed


"R" tabulate  apparent  isotopic compositions derived from the routine


sequence of calculations.  Samples bearing higher numbers have higher


abundances of "0, and it can be seen that, as the oxygen-18 content of


the samples increases, the apparent value of 6tlCpDg does not remain


constant, but, instead, decreases systematically.  For samples enriched


in 1§0 by approximately 300%., the error amounts to almost -10%, (see


sample 21: the routine calculation yields -19.11%, even though the


correct result must still be -10.17%,,).


     Use of the "l»C-known" sequence of calculations is appropriate in


this case.  Columns headed "CK  in Table 3-2 tabulate results.  Because


routine analysis of the unexchanged material indicated a carbon-isotopic


composition of -10.17%. vs PDB, this value was assigned to all samples


ir. which "0 had been enriched by exchange.  Comparison of cxygen-


isotopic compositions reported for each sample in the  "R" and "CK"

-------
                                                                                                             115
           1
           t/2
           o
M
Li
01
•g
 K
tb

I
 O
a


 01
 e>
 pi
 X
o

 0)
•o
-*-l
 K
 O
3
 01
i-H
J3
            1
                 u
                  a
 ca
 a
 o.

V    o
       K
•O
                  I-
                  re
                  U)
       ro o  P- O  «»     CM rM o  in r»     mrMcneoco
       en *»  i*» en  fH     in o m  in »-*     en vo  ro CM i»"*

       inocncoin     coinr-i-ir-     vo 10  ro in en
       m if\  vo p»  en     o CM *"o  in 10     r* en  »"4 CM ro
                             _| ,_, ,-t  ,_!,_(     rliHfMrMrM



vo     CM fr^  CM *v  co     vo f^ f*s  ^y ^j>     ^4 r*1  ro ^n co
in     enc*>r^coo     «*cn*a|'«j|O     oo^»oocn

tM     inocncoin     eo«»f«.i-ir-     vovorJinco
f^     ro u^  10 r**  cn     ^o CM '*1^  in vo     p* CTV  ^H CM F^
                             rHf<>-li-l>-<     rHiHCMCMrM







       mo**rMO     cotnin^eo     foin*^orM
       vor-vocoo     rofMioco-*     nrMinoro

       ^* oo  oo en  vO     c? en en  en c?     en t*4  ^^ ^D ^?
       i-l«H>-fH     rovocor-^
en     vo O  m «*  in     CM rn *f  en en     vo o  ** vo CM
  •       •••••      ••'!•      *••••
vo     COG«*OCO     inmC'ivoro     cccom>-ico
•H     i-i CM  m «•  «•     invovor-co     in c?\  o •-( —i
                                                          i-4 rH >-*







       p^r*t**r*p»     r*r*r*p*r*     r-r^r-p*"^
       fHi-i>-<>-4>-l     i-l>-lt~(^tH     i-4iHf-4r~4>-l

       C3OOC3O     O O O  O O     OOC3OC3

        I   I    I   I   I       I   I   I   I   I      I   I   I   I   I



r-     •vvocnvovo     vorMCOiMco     i~r-<-4inr-
i-(     CMVD«HVOO     mO«^O»rO     p-fMOOCMVO
  •       •••••      •••••      •••••
O     O o  »H •—t  CM     cJror*»fo«*     ^ininvovo

 I      I   I    I   I   I       I   I   I   I   I      I   I   I   I   I
                            CO

                            O
                            O
                            o
                             in
                             09
                             Ov
                             en
                             cr>
                         O  r-  •» m  co
                         p-  10  rs PJ  in
                         <*)  m  r« o\  \o
                         ^  CO  ID in  r-t
                         O  -^  m «.  10
                         O  O  O O  O
                             vo  r~ in co  —i
                             o  10 «» m  in
                             u>  in vo IN  n
                             •w  o CM vo  •-»
                             r-  o\ o   r» m  eo  t-«
O  v o  \o  r>i
eo  co o  in  o
O  c\ •»  t~  r->
*f  m r»  oo  o
                         in  o fM n  n
                         <»)  o m m  co
                         CO  C7^ O^ O\  (7)
                         O\  O\ CTv OA  C\
                             in  10 o  i*> rM
                             m  o 10  p- >-t
                             CA  o CK  o% o
                             en  o ox  c\ o
                             en  o en  o\ o
                             en  o en  co o
o  o en  CM  r»>
«»  vo O  co  O
en  o o  en  o
en  o o  en  o
 »
                                               vo r- oo  en o     r-< IM  ro ^- in

-------
116
•o
!*:
1
to
? "a:
>o
*? u
1

CM
CM
r-4
VO
«•>
O
o
o
0
»4
CO
•^
CO
(M
O
VO
CO
CM
O
CO
O
co
rn


o
1
CV
r*
i
o
OV
(M
CM
W
CM
000050 1
CO
fH
VO
CM
CO
CM
O
CO
CM
en
in
CM
in
vo
m


0
1
••
CD
1
VO
CO
«~<
to
m
CM
000078 1
^4
o
CM
•H
ro
••
O'
o
l-(
n
e>v
CM
CO
in
f4
in


O
r»
f-
co
I
3
r-
O
r—
CM
000106 1
*-»
CM
iM
in
CM
p~
OV
CM
CM
CM
f>
CM
CO
C7>
in

«-<
0
i
*-t
t-<
01
1
o
CM
CO
CM
000184 1
CM





• r-
eo p>
in in
r-t C\
co «
CO —I
-4 O
O O
« 1
R Rl
C C
ea D
tn in
K K
m «
• •
T3 "O
•a -a

u
u
i
c
0
Jt
\
u
d CK indicates "'










8
1
u
01
s
a
n
VI
O
rt
in
u

-------
                                                                     117


columns shows (i) that differences in the abundance of l§0 determined by
                                     •


the routine sequence and the "13C-known" sequence of calculations are,


indeed, small, and (ii) that much larger changes in the calculated


abundance of '- '0 result from uncoupling its abundance from that of 1'0.


     To generalize these observations,  the difference between the


solution to equation 3-21 and the true value of l*R was determined for


all possible values of ltF ("0 and 1JC were assumed to be at nat ural


abundance).  Results are summarized in Figure 3-2.  More important in


tracer studies, the relative difference between the calculated end true


values of 1§F was also determined and is shown in Figure 3-3.


      "R and **R in Natural paroles.  To examine relationships between


*'R and 1>R typical of those found in natural systems, a series of water
                               f        :

samples was produced by fractional distillation.  Subsequent exchange of


oxygen between these samples and carbon dioxide yielded a series of


samples of CO, with widely varying "0 content.  Because the same stock


of CO, was used to prepare all samples, the carbon-isotcoic compositions


of all samples were identical.


     Results of mass spectrcraetric analyses of these samples are


summarized in the first two columns of Table 3-3.  The third column,


based on routine calculations with a - 0.52 and K * 0.0094330 (although


the number of significant figures exceeds the precision with which K can


presently be estimated, the value specified exactly reflects the mean


isotopic abundances reported in ref. 14, and was employed in these


calculations), indicates the range of oxygen-isotopic compositions.  The


fourth column tabulates carbon-isotopic abundances as determined in the


same set of calculations.  Inspection of the latter results shows that

-------
                                                                          IK
118
     Figure 3-2.  Relative error in ^R as a function oC  oxygen-18
enrichment:  I*KZQ  " i-12372 x 10~3  **& 17RC02 * 3-73 x  10""•
solutions to equations 14 and 15 are  obtained only at natural abundance
(»'RCO  * 2.00520 x 10-').

-------
                                                                     CO
                                                                     O
                                                                     h-;

                                                                     O
                                                                             119
                                                                     (O
                                                                     O
                                                                     IT)

                                                                     O
                                       0>
                                       O

                                       O
                                       "O
                                       c
                                                                            go
                                                                             i
                                                                             c
                                                                     ro
                                                                     O
                                                                     cvj
                                                                     O
                                                                             X
                                                                             O
CVJ
•f
CVJ
 I
                                     JOJJ3
ro
 I
in
 i

-------
                                                                           120

120
     Figure 3-3.  Relative error in »'F as a function of oxygen-18
                                     4 !
abundance:  1JRco  e 1-12372 x 10-' and ''^co? = 3*73 x 10~4-  Values of

l*FCO  were calculated from equation 18 using solutions to equations 14

and 15.  Exact solutions to equations''14 and 15 are obtained only at

natural abundance (I*FCO  * 2.000443 x 10-J).

-------
                         121
         I    '
    OJ
     I


U! JOJJ8
ro
 I
                   cp
                   d
                    rq
                    d
                    OJ
                    d
                          0)
                          o

                          o
                         CO

                          I
                          c

-------
                                                                     122
    Table 3-3:   Carbon Dioxide-Oxygen Derived From Distilled Waters
  Scunole
                         ««nb
6»«0,
                                     SHOW'
                                                0.52     0.50     0.52
                                              .009433  .008335   .010090
oc
1
2
3
4
5
6
7
8
9
10
11
12
13
14
0
0
0
0
0
0
0
0
.0
0
0
0
0
1
1
.999857
.996008
.996375
.996499
.996168
.996495
.996547
.996855
.997078 .
.997128
.997114
.997289
.997242
.000050
.000953
1.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
1.
001108
888515
897943
900305
900698
901266
906725
912028
915067
916327
916420
919115
919410
999867
026080

-82
-72
-70
-69
-69
-63
-57
-53
-53
-53
-50
-50
+32
f59

.30
.56
.11
.71
.12
.48
.99
.92
.55
.45
.67
.36
.83
.93

-10.
-10.
-10.
-10.
-10.
-10.
-10.
-10.
-10.
-1C.
-10.
-10.
-9.
-9.

18
14
09
46
13
17
15
06
02
04
12
01
93
89

-10
-10
-10
-10
-10
-10
-10
-10
-10
-10
-10
-10
-9
-9

.28
.23
.19
.55
.22
.26
.23
.13
.09
.11
.19
.08
.88
.81

-10.03
-10.02
-9.98
-10.34
-10.02
-10.07
-10.06
-9.98
-9.94
-9.96
-10.05
-9.94
-10.04
-10.06
      1  "Runknovn/ ""standard'  "Standard = 0-01188158.

b"Q "  ""unknown/'""standard-  ""standard » 0.004149577.

C5antple 0 was initial, unexchanged C02.

-------
                                                                     123

the calculated carbon-isotopic composition is not indeptndent of the

oxygen-18 abundance, as it should be.   In fact,  regression of 61JC on

BltO indicates a slope of 0.0023%0C/%eO,  a result which differs

significantly (p < 0.008) from zero.

     Small variations in assigned values  of a and/or K have almost no

effect on calculated values of 6»»0, but  can substantially affect the

apparent variations in 61JC. As an example, the fifth column in Table

3-3 summarizes apparent carbon-isotopic compositions determined by use

of the routine sequence of calculations with a » 0.50 and K « 0.008335.

This corresponds to the present conventional approach (e. g., ref.  4).

Regression of 61JC on 6l»0 in this case yields a slope of 0.0036 %eC/%<>,
                               4:
a systematic error even greater than that encountered in the preceding

calculation.  The sixth column in Table 3-3 reports apparent carbon-

isotopic compositions determined' by use of the routine sequence of

calculations with a » 0.52 and K = 0.010090.  In this case, K was

adjusted to eliminate covariance between 61JC and 61§0.



                               CONCLUSIONS

     We have presented a simple approach to the calculation of isotope

ratios from the mass spectra of carbon dioxide.  A rigorous examination

of the effect of assuming a quantitative relationship between »*R and

»'R has shown (i) that determination of isotope ratios by the routine

sequence of calculations for carbon dioxide highly enriched in oxygen-18

will result in errors in the isotopic analysis, and (ii) when a « 0.52

is used in the routine sequence of calculations, a value of K » 0.010090

more accurately describes the mass-dependent fractionations of oxygen

isotopes by distillation and exchange.  In the first case, errors in the

isotopic analysis can be eliminated by decoupling »'R and  1§R in the

-------
                                                                     124



calculation, either with the "*• JC-known", or the "11 0-knovm" sequence of



calculations.  In the second case, the value of K which eliminates



covariance between 61JC and 6^0 when a «= 0.52 requires that 1'RSMov«/'"



3.99 x 10-«.  It is, however,  necessary to reevaluate the oxygen-17



content of SHOW by more direct methods.  A systematic study of oxygen-17



and oxygen-18 abundances in many different natural samples and



determination of values of a characteristic of representative processes



in nature are also suggested.





                           EXPERIMENTAL SECTION



      Preparation of Carbon Dioxide for Isotopic Anal/sis.  A series of



samples of water selectively enriched in oxygen-18 was prepared by



mixing varying amounts of 99%-H21*0 (Stohler Isotopic Chemicals) with



distilled, deionized water.  A second series of water samples with mass-



dependent enrichment of both 1'0 and i$0 was produced by fractional



distillation of tap water.  For the exchange of oxygen between water and



carbon dioxide, 200 vL of thoroughly degassed water and 20 lonoles of



carbon dioxide were placed in a sealed, evacuated Pyrex tube (6 mm



o.d.).  The tubes were allowed to stand at 25*C for 48 hours to ensure



that exchange of oxygen between water and carbon dioxide was complete



(13).  The carbon dioxide was purified by cryogenic distillation.



      Mass Spectrometric Measurement.  Values of 41R and 4*R used in



equations 3-16 and 3-17 must be numerically equivalent to the ratios of



molecular species in the sample.  There are two sets of reasons why



recorded ion-current ratios may differ from absolute molecular ratios.



Specific corrections are required in both cases.



     First, there is a set of measurable phenomena such as background



ion currents, amplifier-zero offsets, scattering of major-beam ions into

-------
                                                                     125


minor-beam collectors, leakage of gas-switching valves, etc.  Optimal


procedures for measurement and correction of the effects of these


artifacts have be>en extensively discussed (2-4).  In the present work.


only corrections for background ion currents and amplifier-zero levels


have been required, and these have been applied prior to the calculation


of ion-current ratios.


     Second, the efficiency of the production of ions varies for the


different isotopic species of carbon dioxide.  If, however, this "mass-


discriEination" is constant over time, then this problem can be


eliminated using a reference measurement using carbon dioxide with a

                           4!
known isotopic composition at a working standard.  Reference and sample


gases are alternatively introduced into the mass spectrometer, and ion-


currents at m/z « 44, m/z » 45, and m/r * 46 are measured for each gas


(Figure 3-1).  The true molecular-isotope ratios of the sanple are


determined as follows



                 _             ** sample       _
                 ^sample * ~	 *  ^standard          <3-23)
                              "standard



where °fc refers to the ratio of recorded ion currents, "R refers to the


true ratio of molecular species as defined in equations 3-7 through


3-11, and m and n « 45 or 46.  Calibration of the working standard


requires only determination of 4§Rstandard and 4'Rstandard-  A primary

reference provides a  sample of carbon dioxide with a known isotopic


composition.  Values  of 4IR and «*R for the primary reference are


calculated from the known »'R, »'R, and i'R using equations 3-7 and 3-8.


     The fractional abundance cf a particular isotope  (j) is calculated


directly from the isotope ratios

-------
                                                                     126
                                                                 (3-24)
where the index i is an integral value indicating all stable isotopes of


a particular element; in the case of carbon i * {12, 13} and in the case


of oxygen i « {16, 17, 18}.   For example,  »»P = »»R/(1 + -»'R +»»R).


Delta values are calculated from their definition (14)
                           ^sample
                          —	1   x   1000             (3-25)
                          Rstandard


Values of delta determined in this way do not incorporate mathematical


approximations.


     Mass spectrometric measurements were made using a Finnigan MAT


Delta E isotope ratio mass spectrometer (Finnigan Corp.).  Ion currents


at m/z » 44, o/z = 45, and m/z * 46 were measured for the sample and a


working standard.  The contribution of instrumental background was


subtracted from the ion current and the molecular ratios 48R and "R for


both gases were determined.  The working standard was calibrated using


carbon dioxide produced by phosphorolysis of K-2 and TKL-1 (15).

-------
                                                                     127

                               REFERENCES

 1.   Margrave,  J.  L.;  Polansky, R.  B.   Relative abundance calculations
     for isotcpic  molecular  species.   J.  Chem.  Ed.  39:789-793,  1962.

 2.   Craig,  H.   Isotopic  standards  for carbon and oxygen and correction
     factors for tnass-spectrotnetric analysis of carbon dioxide.
     Geochim. Cosmochim-  Acta  12:133-149, 1957.

 3.   Deinis, P. Mass  spectronetric correction  factors fot the
     determination of  snail  isotopic  composition variations of  carbon
     and oxygen.  Int. J. Mass Spectrom.  Ion Phys.  4:283-295, 1970.

 4.   Mook,  W. G.;  Grootes, P.  M.  The measuring procedure and
     corrections for the  high-precision mass-spectrometric analysis  of
     isotopic abundance ratios, especially  referring to carbon, oxygen,
     and nitrogen.  Int.  J.  Mass  Spectrum.  Ion  Phys. 12:273-298, 1973.

 5.   Bigeleisen, J. The  effects  of isotopic substitution on the rates
     of chemical reactions.  J. Phys.  Chem. 56:823-828, 1952.

 6.   Matsuhisa, Y.; Goldsmith, J. R.;  Clayton,  R. N.  Mechanisms of
     hydrotheriaal  crystallization of  quartz at  250"C and 15 kbar.
     Giochiin. Cosmochim.  Acta  42:173-182, 1978.

 7.   Bigeleisen, J.; Ishida, T.;  Lee,  M.  W.  Correlation of the isctopic
     chemistry  of  hydrogen,  carbon, and oxygen  with molecular forces by
     the WIMPER (2) method.  J. Chen.  Phys. 74:1799-1816, 1981.

 8.   Clayton, R. N.; Grossman, L.;  Mayeda,  T. K. A component of
     primitive  nuclear composition  in carbonaceous  meteorites.  Science
     182:485-487,  1973.

 9.   Clayton, R. N.; Onuma,  H.; Mayeda, T.  K.  A classification of
     meteorites based  on  oxygen isotopes.  Earth Planet. Sci. Lett.
     30:10-18,  1976.

10.   Craig, H.  Standard for  reporting concentrations of deuterium and
     oxygen-18  in  natural waters.  Science  133:1833-1835, 1961.

11,   Baertschi, P.  Absolute "0  content  of standard mean ocean water.
     Earth Planet. Sci. Let. 31:341-344,  1976.

12.   Nier,  A. 0.  A redetermination of the  relative abundances of the
     isotopes of carbon,  nitrogen,  oxygen,  argon, and potassium.  Phys.
     Rev. 77:789-793,  1950.

13.   O'Neil, J. R.j Epstein, S.   A  method of oxygen isotope analysis of
     milligram  quantities of water  and some of  its applications.  J.
     Geophys. Res. 71:4955-4961,  1966.

-------
                                                                     128

14.  McKinney, C. R.; McCrea,  J.  K.;  Epstein,  5.; Allen, H. A.; Urey,  H.
     C.  Improvements in mass  spectrometers for the measurement of small
     differences in isotope abundance ratios.   Rev. Sci. Inst. 21:724-
     730, 1950.

 15. Blattner, P»; Hulston, J. R.  Proportional variations of
     geochemical 6l*0 scales.   An interlaboratory comparison.
     Geochim. Cosnochim. Acta  42:59-62,  1978.

-------
                                                      129
                CHAPTER 4

  DEPOSITION, DISTRIBUTION, AND CLEARANCE OF
OZONE-DERIVED OXYGEN IN THE RESPIRATORY SYSTEM

-------
                                                                     130



     In order to provide experimental  measurements of the dose of ozone



to the respiratory systeri,  the retention and clearance of ozone-derived



oxygen in the respiratory system has been examined using oxygen-18 as a



tracer for inhaled ozone.   Extrapolation of health effects from labor-



atory animals to humans begins with a  prediction of the uptake of ozone



in the human respiratory system based  on mathematical dosimetry models



which describe the transport and absorption of reactive gases in the



lung (1).  This approach requires an understanding of the origin of



species differences in the dose of ozone to tissue and an understanding



of the factors controlling the distribution of ozone-derived oxygen in



the respiratory system.  Correlation of systematic differences in airway



structure and ventilation patterns (2) with quantitative differences in



local tissue dose in different mammalian species will provide a more



complete understanding of the factors  controlling transport and removal



of ozone" in the respiratory system. The first two studies discussed in



this chapter were designed to assess i) the feasibility of utilizing



1(0] to determine species differences  in deposition and ii) the ability



to resolve depositional patterns within the respiratory system.



     The primary events of injury to the lung must involve interaction



of oxidants with biological molecules.  Following the initial insult,



molecules altered by ozonolysis may fail to fulfill their biological



function (3).  Products of ozonolysis  may initiate autoxidation directly



or may mediate cell damage as chemotactic agents for alveolar



macrophages and polymorphic neutrophile (4-8), both events resulting in



changns in cellular function, and possibly cell death.  Mechanisms of



protection against oxidant injury (9,10) and of extrapulmonary effects



(11) have only recently been considered.  The dependence of dose on

-------
                                                                     131

exposure tiae at a single concentration has been examined.  The

distribution of czone-derived oxygen in proteins, lipids, and organic

solutes and the clearance of ozone-derived oxygen from the lung also

have been determined.


           Detection and Quantitation of the Oxygen-18 Label

     Stable isotopes are ubiquitous in nature.  Approximately 0.2% of

oxygen in biological material is oxygen-18.  Detection and quantitation

of tracer-derived oxygen in tissue therefore requires precise knowledge

of the isotopic composition of i) the initial tissue-oxygen pool and ii)

the tissue-oxygen pool after incorporation of the tracer.  An isotopic
                              4'
mass balance of oxygen in tissue from an animal exposed to labeled ozone

yields an expression which can be used to determine the dose of ozone-

derived oxygen to tissue.  For cfzone labeled with oxygen-18, the mole

fraction of ozone-derived oxygen in the tissue oxygen pool is



                                    ,E - I§FTI,NA
                       "TI      "FTR ~ 1$FTI,NA


where »»F is the fractional abundance of oxygen-18 and n is the quantity

of oxygen ( e . g . , micromoles 0) in the i-th oxygen pool; the subscripts

designate the oxygen pools {tissue, natural abundance (TI,NA); tissue,

enriched (TI,E); and tracer (TR)}.  The numerator in the right-hand side

of equation 1 is the enrichment of oxygen-18 in the tissue-oxygen pool,

the amount in excess of the natural background.  The denominator in the

right-hand side of equation 1 quantifies the potential of the tracer to

produce an enrichment in the tissue.  This mathematical formalism can be

used in studies employing ozone which is less than 100% oxygen-18.

     In this work, "dose" will refer to the total amount of ozonederived

-------
                                                                     132



oxygen in a tissue sample,  whereas "relative enrichment"  refers to the



mole fraction of ozone-derived oxygen in the tissue-oxygen pool.  Dose



is determined from relative enrichment and the size of the tissueoxygen



pool





                dose * (relative enrichment) x (oxygen pool)     (4-2)





where oxygen pool refers to the amount of atomic oxygen (moles) in the



tissue.  Utilization of oxygen-18 as a physiological tracer (in fact,



utilization of any stable isotope which is normally present in



biological material as a physiological tracer) requires both elemental



and isotopic analyses to quantify dose.



      A fundamental limitation on detection and quantitation of an



isotopic tracer is imposed by variations in the natural abundance of



oxygen-18, which include both variations in individual animals with time



and variations among animals in the same population.  These variations



appear as unavoidable "noise" on the "stable isotopic signal" in any



biological system.  The precision of the measurement of oxygen-18



derived from labeled material in a physiological sample is a function of



i) the precision of the mass spectrometric measurement, ii) isotopic



variations introduced by sample preparation (including conversion of



oxygen to carbon dioxide),  and iii) naturally occurring variations in



the abundance of cxygen-18 in the sample material.  Mathematically, this



can be expressed as a sum of variances






                         V  " V  * V  * V                  <«-3>





where Sj2 is the total analytical variance, SMI is the variance



assoc_ated with the mass spectrometric measurement, SpJ is the variance

-------
                                                                     133



arising from sample preparation,  and s^1  expresses natural variations in



oxygen isotopic abundance as a variance.   The level of isotopic



enrichment also affects the precision (enrichments just above natural



abundance can be measured more precisely  than large enrichments),  but



the exact relationship between enrichment and analytical precision has



not been determined.



     Ideally, sample preparation and mass spectrcmetric measurement



procedures should contribute insignificantly to uncertainty in the



isotopic analysis.  In practice,  this condition is rarely achieved.



Contributions to the variance in *»F from these sources are summarized



in Table 4-1.  Each entry represents the  variance obtained from five



replicate analyses.  The contribution of  the mass spectroinetric



measurement (s*) was determined from analysis of identical samples of



carbon dioxide.  The variance associated  with sample preparation (Sp1)



has two sources:  preparation of carbon dioxide from organically bound



oxygen and sampling of an inhomogeneous powder (tissue homogenate).  The



first contribution was determined by analysis of carbon dioxide prepared



from crystalline glycine and the second contribution was determined from



analysis of carbon dioxide prepared from  samples of a single lung homog-



enate.  Variations in the abundance of oxygen-18 between animals (SNI)



were examined by analysis of lung homogenates from different animals.



Temporal variations of oxygen-18 were not considered here.  The prepara-



tion of carbon dioxide from organic material is the greatest single



source of uncertainty in the isotopic analysis, accounting for 36% of



the analytical variance.  Tissue sampling and variations among animals



account for 34% and 29% of the analytical variance, respecively.  The

-------
                                                                     134
            Table 4-1.  Contributions to the Variance in l'F
source
a
mass spectrometer
sample preparation
tissue sampling0
d
natural abundance
sj Is1
x 10l» x 101§
6 6
157 e 163
14£e 309
126 e 435
detection limit
ppm
0.10
0.50
0.69
0.82
^replicate analyses of carbon dioxide.

^replicate analyses of glycine.

creplicate analyses of a single lung horaogenate.

 replicate analyses of lung homogenates from different animals.

Individual contributions to the variance were calculated sequentially
 froa the total variance (Is3).

 95% confidence limit for relative enrichment, where sj B 4.35 x
 10-» for »»FTT „. and »»F_, r, 1§F-~ «= 0.99, and N - 5.
             TI,HA        il,t     IK

-------
                                                                     135



mass spectrometric measurement contributes only 1% of the observed



variance.  Detection limits corresponding to the cumulative variance at



each step were calculated.   A detection limit of < 1 ppm (that is, one



atom of excess oxygen-18 per 10' atoms of oxygen in the tissue-oxygen



pool) is expected if no sources of uncertainty other than those



considered are present.  However, differences in physiological states of



individual animals can cause differences in the retention of label in



tissue.  Variations in the abundance of oxygen-18 observed in an



apparently homogeneous test population may be significantly greater than



predicted.  The detection limit for excess oxygen-18 in a given test



population will be effected by experimental conditions and must be



evaluated in each case.





                   Deposition in the Respiratory System



      Comparative Aspects of Ozone Uptake.  This study comprises a



comparison of the uptake of ozone-derived oxygen in the respiratory



system of the rabbit (6.5 kg), the rat (380 g), and the mouse (21.5 g).



Tissue samples included head (nasopharynx), trachea, lungs, and blood.



In two separate exposures, three animals (one of each species) were



placed in an exposure chamber containing 1 ppm of 1>0a for 60 minutes.



Unexposed animals (two of each species) were used to establish the



natural abundance of oxygen-18 in each sample.



     Results of isotopic analysis are shown in Table 4-2.  Relative



enrichments were calculated using equation 1, where ^F-pj  j^ was  the



abundance of oxygen-18 in tissue from unexpoSed animals, *'FTI  £  was the



abundance of oxygen-18 in tissue from exposed animals, and  "FTR* 0.99.



Dose was determined from the relative enrichment and an estimate  of the



size of the oxygen pool in each tissue, shown in Table 4-3.

-------
                                                                     136




Table 4-2.  Deposition of Ozone-Derived Oxygen in the Respiratory System
tissue
species
head




mouse
rat
rabbit
xlO>

2
2
2

.0286
.0302
.0330
'•FTI.E
xlO1

2
2
2

.0341
.0339
.0334
enrichment
xlO«

5
3
0

.5
.7
.3
dose
nanomoles

8.0 ± 0.
36 ± 6
0

8


trachea



lung



mouse
rat
rabbit

mouse
rat
rabbit
2
2
2

2
2
2
.0302
.0322
.0296

.0326
.0323
.0335
2
2
2

2
2
2
.0463
.0489
.0373

.0520
.0454
.0438
16
19
7

19
13
10
.4
.9
.7

.6
.2
.4
6.2 ± 1.
Ill ± 32
63 ± 34

8.6 ± 0.
51.7 ± 2.
286 i 22
7



8
9

blood



mouse
rat
rabbit
2
2
2
.0243
.0235
.0220
2
2
2
.0245
.0237
.0225
0
0
0
.2
.2
.5
0
0
0



ax, SE = 0.0003 x 10- J, and N » 2.




 enrichments were calculated using the following equation 1, where
0.99,
             j ,JA
                  was the fractional abundance of oxygen-18 in the
 unexposed animals, and the l'FTj E was the fractional abundance of



coxygen-18 in the exposed animals, SE • 0.7 x 10-*.



 dose is total tissue uptake, uncertainty is propagated SE.



 isotopic enrichment was not statistically significant (p < 0.05).

-------
                                                                     137
   Table 4-3.  Oxygen Pools in nasopharynx, Trachea, Lungs, and Blood
tissue
     species
                     weight
                        g
                       % oxygen
                         wt/wt
                  oxygen pool
                   millimoies
head

     mouse

    ' rat

     rabbit


trachea

     mouse

     rat

     rabbit


lung

     mouse

     rat

     rabbit
blood
     mouse

     rat
0.126 ± 0.012

 0.90 ± 0.15

 5.89 ± 0.10





0.003 ± 0.001

0.045 ± 0.013'

0.673 ± 0.029"
0.032 ± 0.003

0.289 ± 0.018

 2.03 ± 0.14
16.1 ± 0.5

17.3 ± 0.9

19.5 ± 1.3





20.5 ± 0.9

19.8 ± 0.4

19.6 - 0.6





21.9 1 0.7

21.7 i 0.5

21.7 ± 0.7





19.3 * 0.1



19.4 _ 0.1
                                                          1.27 i 0.12

                                                           9.7 ± 1.7

                                                            72 ± 5





                                                          0.38 ± 0.13

                                                           5.6 ± 1.6

                                                           8.2 ± 4.4





                                                           0.44 ± 0.04

                                                           3.92 ± 0.24

                                                           27.5 1 2.1

-------
                                                                     138



     Species differences in the uptake of ozone have their origin in



differences in respiratory rate, lung size,  total surface area, and



ventilation.  Since these parameters are proportional to body mass (13),



isotopic enrichment and dose also should be a regular function of body



mass.  When relative enrichment is considered, the mole fraction of



ozone-derived oxygen in the tissue-oxygen pool decreases with body mass.



The size of the tissue-oxygen pool in each sample increases with body



mass (Table 4-3), so that, with the exception of tissue from the naso-



pharynx, the total amount oi ozone-derived oxygen in each sample



increases with body mass (Table 4-2).



     Measurements of the tracheal concentration of ozone in rats, guinea



pigs, rabbits (14), and dogs (15,16) indicate that approximately 50% of



inhaled dose is removed in the head.  Nasopharyngeal deposition is



therefore related to minute volume and should be directly proportional



to body mass.  In this study, however, an inverse trend was observed.



Measurement of isotopic enrichment in the nasopharynx was complicated by



the sampling procedure.  Head samples included portions of extraneous



muscle and skeletal material which was more difficult to remove



reproducibly from the larger samples.  Incorporation of small amounts of



label into an increasingly large oxygen pool composed of oxygen from



unexposed tissue (extraneous material) rssulted in tin apparent decrease



in the amount of ozone-derived oxygen retained in the nasopharynx.



Problems of dilution in larger animals have taen circumvented with the



use of different sampling techniques.



      Distribution Within _the Respiratory System. The distribution of



ozone-derived oxygen in the respiratory system was determined in



rabbits.  One animal was exposed to 1 ppiu 1'0, for 60 minutes and one

-------
                                                                     139


animal was retained to establish natural abundance.  Results of isotopic


analysis of airway epithelium from both animals are shown in Table 4-4.


The relative enrichment in each sample .was calculated using equation 1;


1§FTI NA was tne oxY?511"18 abundance in tissue from the unexposed


animal, l'FTI E *'as the oxygen-18 abundance in tissue from the exposed


animal, and 1'FT, = 0.99 (Table 4-4).  The dose to various epithelial


regions was determined using equation 2, where the size of the oxygen


pool was determined from the elemental analysis and the weight of the


sample.  Dose is expressed as picomoles of ozone-derived oxygen per


milligram of dry epithelial tissue (Table 4-4).


     As expected, the dose is highest in areas of greatest turbulence,


the nasal turbinates and the trachea immediately below the laryngeal jet
                             <>

(17), and in the conducting airways of the lung (refer to Figure 4-1 for


proper anatomical orientation).  A small, but statistically significant


(p > 0.05) isotopic enrichment is observed in the peripheral parenchyma


of the lung.  Based on the preceding discussion of variance, the detec-


tion limit is 20 picomoles of ozone-derived oxygen per milligram of


epithelium.  Although the dose is greatest in the. nasopharynx and the


conducting airways of the lung, a small amount of ozone does reach the


centriacinar region.  Since oxidant injury to respiratory tissue is


responsible for pulmonary toxicity, the ability to measure the dose to


peripheral regions of the lung is necessary in establishing a valid


dose-response relationship for inhaled pollutants.  However, the finding


that the dose to the centriacinar region of the lung is extremely small,


even with a high atmospheric concentration of ozone, brings into ques-


tion the ability of ozonolysis alone to account for injury to tissue.


It is therefore likely that other mechanisms of tissue damage occur.

-------
                                                                     140
  Table 4-4.  Uptake of  Ozone-Derived  Oxygen in the Respiratory System
    tissue
                         .
                       10s
enrichment3      dose
  x 10*       piccmole/mg
nasal turbinates
nasal septum
nasal pharynx
olfactory mucosa
trachea 1
2
3
4
5
lung 7mm airway
lung parenchyma
2.0402
2.0377
2.0311
2.0344
2.0301
2.0301
2.0301
2.0301
2.0301
2.0262
2.0312
2.1359
2.1179
2.1124
2.0651
2.3518
2.2251
2.1762
2.1554
2.1369
2.1267
2.0365
96.9
81.2
82.3
31.1
325.6
197.4
149.9
126.9
158.7
101.8
5.3
1150
930
930
440
3660
2220
1690
1430
1790
1590
70
aenrichments were calculated using equation 1,  where **FTR » 0.99,
 14FTj JJA was the fractional abundance of oxygen-18 in the unexposed
 animals, and the l*FTI E was the fractional abundance of oxygen-18 in
 the exposed animals.

 dose is the amount of ozone-derivei oxygen/milligram of dry epithelium.

-------
                                                                           141





141
     Figure 4-1.  Distribution of ozone-derived oxygen in the



respiratory system.  The dose in a rabbit exposed to 1 ppta 1'0J for 60



minutes is expressed as picoraoles/milligram dry epithelial tissue (refer



to Table 4-2).

-------
142

-------
                                                                     143



Finally, the demonstrated ability to resolve spatial distributions of



ozone-derived cicygen within the respiratory system by this technique



offers great promise in the validation and definition of mathematical



dosimetry models.





                   Deposition and Clearance in the Lung



      Deposition in the Lung.  To determine the effect of exposure time



on retained dose in the lungs,  groups of five mice were exposed to 1 ppm



'•0, for times up to 60 minutes.  A portion of the lungs from each



animal was retained for analysis.  The rema.-'.ning tissue from the five



animals in an exposure group was pooled and chemically fractionated into



an acid precipitable fraction (macromolecules), chloroform extractable



fraction (lipids), and a water:methanol soluble fraction (solutes).  The



natural abundance of oxygen-18 in each of these samples was established



using five unexposed animals.



     Results of isotopic analysis are shown in Table 4-5.  All variables



were analyzed within a linear regression framework.  The dependent



variable was »»F and the independent variables were exposure time, the



actual concentration of ozone,  and their interaction.  There was no



indication that body weight (21.2 ± 0.7 g), lung wet weight (144 ± B



mg), or lung dry weight (30 ± 2 mg) varied significantly (p < 0.05) by



exposure group.  Although there was some evidence of lack of fit (p =



0.01), this appears to be due primarily to the anomalously high average



isotopic composition of the group at the 30-minute exposure.  Polynomial



fits did not suggest the existence of anything but a linear trend.  The



least-squares fit for ozone-derived oxygen in whole lung vs exposure



time is shown in Figure 4-2 and those for each chemical fraction are



shown in Figure 4-3.  The uncertainty indicated in Figure 4-3 is the 95%

-------
                                                                     144
      Table 4-5.  Deposition of Ozone-Derived Oxygen in the Lungs
exposure
time
min
0

6

17

21

30

32

35 -

44

60

intercept
slope d
r'd

lung
homogenate3
2.033?
± 0.0007
2.0350
± 0.0001
2.0398
± 0.0015
2.0395
± 0.0014
2.0486
± 0.0014
2.0412
t 0.0019
- — --

2.0453
± 0.0017
2.0521
1 0.0024
2.0337
0.0003
0.663
i«F
acid
precipitate"
2.0323
± 0.0003
2.0336
± 0.0012
2.0352
± 0.0013
2.0352
± 0.0011
2.0444
* 0.0017
2.0346
± 0.0019
2.0363
± 0.0014
2.0449
± 0.0020
2.0466
± 0.0022
2.0315
0.0002
0.726
x 10'
chloroform
extract b
2.0353
± 0.0004
2.0351
± 0.0003
2.0388
± 0.0006
2.0448
1 0.0005
2.0569 c
± 0.0007
2.0469
± 0.0007
2.0473
± 0.0003
2.0493
± 0.0004
2.0579
± 0.0006
2.0342
0.0004
0.767

water :methanol
extract53
2.0442
± 0.0004
2.0435
± 0.0006
2.0546
± 0.0007
2.0476
i O.OOOS
2.0527
± 0.0006
2.0505
± 0.0006
2.0503
± 0.0008
2.0570
± 0.0010
2.0554
± 0.0009
2.0451
0.0002
0.669
 x ± SE for five animals in a single exposure group.



 x ± SE for triplicate analyses of pooled samples.



cexcluded form the linear regression.



 values of linear regression

-------
                                                                          145


145
     Figure 4-2.  Uptake of ozone-derived oxygen in the lung.  A plot of


oxygen-18 abundance in lung tissue vs exposure time for mice exposed to
                                     ,*

1 ppn> '"Oj is shown.  Each point is the mean and standard error of the


mean for analysis of tissue from five mice in an exposure group.  The


regression line with 95% confidence limit for the regression estimate is


also indicated.

-------
                                                                       146
                            i  i  i  i  i—i  i  i  i  i  'i—m—r~r
                          i  i  i  I  I  i   iiliii\i\    i\i
                                             o
                                             t0
                                                                  O
                                                                  ID
                                                                  O
                                                                  O
                                                                  CM
                                                                  O
CD
q
CM
in
                                6un|
                     o
                     cvj
ro
O
CM

-------
                                                                          147
147
     Figure 4-3.  Molecular distribution of ozone-derived oxygen in the
lung.  Plots of oxygen-18 abundance in the water:methanol extract,  the
acid precipitate, and the chloroform extract vs exposure tine are shown.
Each point is the mean and standard error of the mean for three
replicate analyses of fraction obtained froa peeled lung tissue.  The
regression lines and 95% confidence limits for the regression estimates
are indicated.

-------
                                                                      1435
    2.06
    2.05
 u.
CD
    2.04
     2.03
     2.05
10
 Q

 *   2.04
 u.
s
     2.03
                    I     ^    I     '     I
              water -methanol extract
               acid precipitate
                    10        20        30        40        50        60

                                 exposure time, min

-------
                                                                     149



confidence limit for the entire population,  indicating real variation



among individuals.  Since pooled tissue samples were used to obtain the



chemical fractions, uncertainties indicated  in Figure 4-3 represent



analytical variation.



     The regression explained approximately  73% of the total variation



in the fractional abundance of oxygen-18 in  whole lung,  73% in the acid



precipitate, 67% in the chloroform extract,  and 77% in the methanolwater



extract.  There was evidence of an interaction (p « 0.002) between dose



and exposure time in whole lung, but it was  not significant when only



exposed animals were considered.  In all cases, regression of 1§F on the



concentration-time product explained less of the total variability than



did exposure time alone.  Within a narrow concentration range, exposure



time is the most important factor controlling uptake of ozone-derived



oxygen in the lung.



     In the statistical analysis, it was assumed that the variance of



1
-------
                                                                     150



burying their nose in fur,  thus filtering inspired air,  while others



regained active throughout the entire exposure.



     The dose of ozone-derived oxygen to the lungs and to each fraction



was determined using regression estimates of "F as a function of



exposure time.  Equation 1 was used to determine the relative enrichment



in each sample, where the abundance of oxygen-18 in samples from



unexposed animals was used to establish 1'FTj jj^, the regression



estimate of isotope abundance was used to determine 1§FTI E, and 1§FTR *



0.99.  Dose was determined from relative enrichment and the size of the



oxygen pool.  Estimates of the sizes of the oxygen pools are shown in



Table 4-6.  Whole lung retained approximately 100 picomoles per minute



of ozone-derived oxygen.  Of this amount, 63% was in the acid precip-



itate, 10% was in the chloroform extract, the 12% was in the methanol-



water extract, and 15% was lost.



     These results are somewhat surprising in light of the suggestion



that membrar.e lipids are the primary targets for ozone (18,19).  Most of



the retained label appears in the acid precipitate rather than in the



chloroform extract, suggesting that proteins rather than lipids are the



primary targets for attack by ozone.  Appearance of the isotopic label



in proteins can result from termination reactions between radicals



produced by ozonolysis of unsaturated lipids and proteins; however,



direct attack on proteins is also a possible explanation.  The un-



accounted portion could have been lost by volatilization of labile



molecules or acidification (small organic acids and bicarbonate) and by



exchange of oxygen with water  (20).  Loss of the isotopic label could



also have occurred during thi  exposure and, depending on the products  of



ozonolysis, resulted in selective depletion of label from one of the

-------
                                                                     151
                     Table 4-6.  Lung Oxygen Pool
        sample     dry weight   % of total     % oxygen     oxygen pool
                       mg                                    microraoles
       whole lung      30          100           21.9           411



  chloroform extract    3           10           12.9            24



   acid precipitate    20           65           31.1           288



water:methanol extract  7           25        .26.8           117





abody weight * 20.97 ± 0.13 g, lung wet weight = 154 ± 2 mg.

 x ± SE, N » 45.

-------
                                                                     152



chemical fractions.  A complete mass  balance of  inhaled ozone-derived



oxygen would help answer questions  of tissue dose and molecular targets.



Further work in this area is therefor.2 needed.



     Pulmonary toxicity is more closely related  to the concentration of



ozone, as opposed to exposure time  or the time-weighted-average concen-



tration (12).  Future studies employing different concentrrtions of



ozone will indicate whether tissue-dcse is related to atmospv ^ric



concentration or whether saturation of the ability of the lung to



protect against injury caused by oxidants occurs at higher doses.



      Clearance from the Lung.  To  examine the characteristics of the



clearance of ozone-derived oxygen from the respiratory system, 30 mire



were exposed to 1 ppm itO, for 45 minutes.  Groups of five randomly



selected animals were sacrificed at 0, 3, 6, 24, and 48 hours post



exposure.  The natural abundance of oxygen-18 was established using six



unexposed animals.



     Results of isotopic analysis are shown in Table 4-7.  A plot of the



natural logarithm of relative enrichment versus  time post exposure is



shown in Figure 4-4.  A biexponential decay of ozone-derived oxygen from



the lung is observed.  The clearance  equation,  indicated by the curve in



Figure 4-4, was determined using a  curve-stripping procedure (21)





                R(t) - 0.83-exp(-0.31t) * 0.17-exp(-0.01t)        (4-4)





where R(t) is the fractional retention of ozone-derived oxygen in the



lung and t is time in hours.  The first phase accounts for the clearance



of 83% of the isotopic label present  in the lung at the end of the



exposure (zero hours) and has a half-life of 2.2 hours.  The second



phase accounts for the clearance of 17% of the isotopic label initially

-------
    Table 4-7.,  Clearance of Ozone-Derived Oxygen in Mouse Lungs
                                                                    153
       time
  post exposure
       h
body weight3    lung wet weight3     ^Flung3
     g                 mg            x  103
natural abundance    24.3 ± 0.4
                   151 ± 7
2.0327 ± 0.0003
                     23.9':! 0.6
                   143 1 6
2.0459 ± 0.0030
3
6
24
48
25
25
25
24
.6 ±
.3 1
.9 ±
.7 ±
0
0
1
1
.7
.8
.2
.0
139 ±
139 ±
155 ±
150 ±
5
2 '
15
9
2
2
2
2
.0426
.0369
.0346
.0341
* 0
± 0
1 0
.0036
.0019
.0011
± 0.0005
x i SE for five animals in a single exposure group.

-------
                                                                          154
154
     Figure 4-4.  Clearance of ozone-derived oxygen fron the lung.  A

plot of the natural logarithm of relative enrichment vs time post

exposure for mice exposed to.l ppra ''O, for 45 minutes is shown.  Each

point is the mean and standard error of the mean for analysis of tissue

from five mice.  The clearance equation is also indicated in the figure.

-------
155

-------
                                                                     156


present and has a half live of 70 hours.


     The clearance of aerosols from the lung (22) has received much


attention because of the continued health risk posed by particulates


lodged in the centriacinar region.  Lung clearance of diesel exhaust


particles is described by a three compartment model, where the three


phases have half-fives of 1 day, 6 days, and 80 days (23).  Particles


deposited in the trachobronchial tree and the proximal bronchioles are


transported to the gastrointestinal tract by the mucocillary escalator


(phases 1 and 2, respectively), whereas particles reaching the alveolar


region are removed by endocytosis, adsorption, dissolution, and metab-


olism (phase 3) (24).  In contrast, clearance of ozone-derived oxygen


from the lung involves catabolism of ozonolysis products and adsorption
                            r»

directly into the systemic and lymphatic circulation, with mucocillary


clearance and phagocytosis playing minor roles.  Exchange of oxygen


between organic molecules and water (20) also results in removal of the


isotopic label from lung tissue and could be a significant factor in


lung clearance.  These data indicate that ozone-derived oxygen is


rapidly cleared from the lung; however, evidence for a two-compartment


model is minimal.  Therefore, at this time no assignment of clearance


mechanisms is possible.  Unambiguous definition of the phase(s) of


clearance and the mechanism(s) for removal of ozone-derived oxygen from


the lung require a closer examination.




                           EXPERIMENTAL SECTION


      Apparatus.  Isotopically enriched ozone was prepared from 99*


oxygen-18 0, in the apparatus shown in Figure 4-5.  The ozonizer was an


annular pyrex vessel with a sealed bottom and a cold finger for


condensing liquid oxygen and ozone.  Cylindrical copper electrodes were

-------
                                                                          157




157
     Figure 4-5.  Apparatus for preparation and purification of Ia0,.



Below the manifold, the apparatus consisted of an annular ozonizer and



two distillation traps for the-synthesis and purification of labeled



ozona.  Above the manifold were a 22-L ballast flask for storage of N,



and a 2-L flask for storage of an "0,/H, mixture.  See the text for



details of construction and procedure.

-------
                               158
Q
                             o
                             to
                             S
                             o
                             to
                             5
 o 5
 -8

-------
                                                                     159

placed on the inner- and outer-external surfaces of the ozonizer (25).

The distillation traps were pyrex U-tubes.  All valves were high-vacuum

valves with Teflon plugs and Kel-F 0-rings,  with a maximum clearance of

4 mm (Kontes).  The manifold was fitted with a pressure gauge which

measured absolute pressure from 0 to 760 torr (Matheson) and with a

thermal conductivity vacuum gauge (Varian) which operated in the range

from 10-3 to 2 torr (not shown in Figure 4-5).  A guard trap filled with

copper turnings heated to 100*C was placed between the vacuum manifold

and the pumping system to prevent ozone from reacting with oil in the

mechanical pump.

     Two exposure chambers and a syringe pump for metering isctopically
                           t!
labeled ozone into the chambers were constructed specifically for

exposure of mice to l*0,.  A schematic diagram of the exposure system is

shown in Figure 4-6.  A small exposure chamber, which held five mice,

was constructed to minimize internal volume (3 I.), thus minimizing

isotope consumption and expense.  A larger exposure chamber (30 L) was

constructed to contain up'to 40 mice.  Both chambers consisted of a

series of plenums separated by perforated stainless steel sheets, which

ensured laminar flow of gas over the animals (26).  A dispersion element

in the upper plenum ensured a uniform distribution of ozone in the

laminar flow region.  The syringe had a capacity of 500 mL (Precision

Sampling).  It was constructed of borosilicate glass with a Teflon cap

and plunger.  A stepping-motor-drive system displaced the plunger of the

syringe.  A variable compression rate from 0.8 to 800 mL/min was

available.  The transfer line from the syringe to the exposure chamber

was Teflon tubing.

-------
                                                                          160




160
     Figure 4-6.  Exposure system.  This system consisted of a 500-mL



stepping-inotor-driven syringe, ah exposure chamber, and an ozone



monitor.  See the text for details of construction and procedure.

-------
                                                             161
o
/N

0
s!=*
  UJ
  IT
                                                            UJ
                                                            §
                                                            s

                                                            cc

                                                            o

-------
                                                                     162

      Preparation and Purification of l'Q,.  To synthesize 140,, the


ozonizer was izsaersed in liquid nitrogen and labeled oxygen was admitted

to the system.  A potential of up to 25 XV was applied to the

electrodes.  The vapor pressure of oxygen at liquid nitrogen temperature

(165 torr) was sufficient to maintain a diffuse discharge at approxi-

mately 15 kv.  Ozone that formed in the discharge region immediately

condensed on the walls of the ozonizer and filled the cold finger.  At

completion of the reaction, the pressure inside the ozonizer was that of

ozone at liquid nitrogen temperature (5 X 10~J torr) and was

insufficient to maintain a diffuse discharge at full voltage.  The

procedure was stopped when this occurred, and a nearly quantitative

yield of ozone was obtained. , Residual oxygen and other possible

contaminants were removed by two cryogenic distillations.  The ozone wris
                              f
transferred to a clean 2-L pyrex flask, diluted with nitrogen from th-^

22-L flask at a concentration of 1% to a pressure of. 740 torr, and used

within 4 hours of preparation.

     As ozone is both explosive and toxic, precautions to prevent injury

in case of explosion and to minimize the risk of inhalation were taken.

The danger of explosion is greatest when ozone is a liquid or a solid,

where rapid warming or cooling of the condensed phases could cause a

detonation.  Rapid temperature changes should therefore be avoided.  The

ozonizer and distillation traps were enclosed by an explosion shield, a


Plexiglas housing with a screen back.  The explosion shield was designed

to direct any glass fragments resulting froa detonation of the ozone


away from the experimenter.  Two controlled explosions were performed to


ensure that the explosion shield functioned adequately.  A ventilation


system was added to remote ozone that might escape during preparation

-------
                                                                     163



aid transfer.



     Animal Exposures.  Female CD mice,  male Fisher 344 rats, and male



Hew Zealand white rabbits (Charles River) were used in these studies.



Animals were placed in an exposure chamber and allowed to adjust to



their surroundings for 10 minutes prior to beginning the exposure.



Filtered, dry air was supplied to the chamber at & flow rate of one



chamber volume per minute.  The syringe-pump was started, and the



measurement of exposure time began when the desired ozone concentration



was established in the laminar flow region of the chamber.  The



concentration of ozone inside the chamber was measured using a Bendix



Model 8002 ozone monitor (Bendix).  The stepping-rate of the syringe



pump was adjusted to maintain an ozone concentration of 1 i 0.1 ppm "0,



in the laminar flow region.  Animals were removed promptly at the end of



an exposure.



      Tissue Sampling and Preparation.  After removal from the exposure



chamber, animals were individually anesthetized with Halothane and



approximately 3 mL of blood were removed from a heart puncture.  The



trachea and lungs were immediately excised.  The lungs were homogenized



in distilled water using a Pclytron homogenizer (Brinkmann Instruments).



The head was skinned, muscle and skeleton not part of the nasopharyngeal



region were removed, and the remaining tissue was homogenized in



distilled water.  Weights of all samples were obtained prior to homogen-



ization.  The tissue homogenates were immediately frozen in liquid



nitrogen and lyophilized to dryness.  The samples remained frozen during



lyophilization.  Residual water was removed by placing the samples under



high vacuum for 24 hours, or until a pressure of < 10-3 torr was main-



tained when the samples were isolated from vacuum.

-------
                                                                     164



     Using the following procedures,  lung tissue was separated into



fractions roughly corresponding to lipids,  proteins, and organic solutes



(27).  Lung homogenates from animals  in each group were pooled and



resuspended in distilled water so that the final mixture was 80% (wt/wt)



water and 20% (wt/wt) tissue.  Volumes of solvent in the following



procedure are given for one gran of tissue suspension.   Ten drops of



concentrated hydrochloric acid and 10 mL of methanol were added to the



tissue suspension.  The mixture was homogenized, 20 mL  of chloroform was



then added, and the mixture was again homogenized.  The precipitate was



removed by suction filtration through a fritted glass filtering funnel



(ASTM 40-60 C), resuspended in 30 mL of chloroform-methanol (2:1 v/v),



homogenized, and separated by suction filtration.  This procedure was



repeated and the precipitated solids were lyophilized to dryness.  The



three filtrates were combined, an amount of water equal to 25% of the



total volume of filtrate was added, and the solution was mixed



thoroughly and was centrifuged at 10,000 x g for 5 minutes to separate



the twc layers.  The upper layer (methanol-water) was removed by



aspiration and the lower layer (chloroform) was extracted twice with 1/4



volume of oethanol-water (1:1 v/v).  The three methanol-water fractions



were pooled.  Solvents were removed by evaporation in a stream of dry



nitrogen.  The acid precipitate contained proteins and other



mscromolecules, the chloroform extract contained primarily lipids, and



the methanol-water extract contained polar organic  solutes.



      Elemental and Isotopic Analysis.  Both oxygen elemental and



isotopic analyses of each sample were obtained.  Procedures for isotopic



analysis were described in Chapters 2 and 3.  Working  standards (tissue



identical to the samples) were calibrated as described in Chapter 2.

-------
                                                                     165

                                REFERENCES

 1.   Miller,  F.  J.   A mathematical  model  of  transport  and removal  of
     ozone in mammalian lungs.   Ph.D.  thesis,  North  Carolina  State
     University, Raleigh,  North Carolina, 1977.

 2.   Supplement: Comparative Biology  of  the Lung.   Am.  Rev.  Respir.
     Dis. 128, 1983.

 3.   Janoff,  A.   Biochemical link between cigarette  smoking and
     pulmonary emphysema.  J. Appl. Physiol. 55:  285-293, 1983.

 4.   Tate, R. M.; Repine,  J. E.  Neutrophiles and the  adult respiratory
     distress syndrome. Aoer.  Rev. Respir.  Dis.  128:552-559,  1983.

 5.   Pantone, J. C.; Ward, P. A. Role of oxygen-derived free radicals
     and metabolism in leukocycte-dependent  inflammatory reaction.
    'Amer. J. Pathol. 107:397-418,  1982.

 6.   Weiss, S. J.;  Lo Buglio, A. F. Phagocyte-generated oxygen
     metabolites and cellular injury.   Lab.  Invest.  47:5-18,  1982.
                               t'
 7.   Diegelmann, R. F.; Cohen,  I. K.;  Kaplan,  A.  M.  The role of
     macrophages in wound  repair:  A review.  Plastic  and Reconstruc.
     Surgery 68:107-113, 1931.    '

 8.   Transwell,  A.  K.; Smith, B;' T. Human fetal  lung  type II
     pneumocytes in monolayer cell  culture:   The  influence of oxidant
     stress,  cortisol environment,  and soluble fibroblast factor.
     Pediatric Res. 13:1097-1100, 1979.

 9.   Chow, C. K.; Tappel,  A. L.  An enzymatic protective mechanism
     against lipid  peroxidation damage to lungs of ozone-exposed dogs.
     Lipids 8:518-524, 1972.

10.   Chow, C. K.; Tappel,  A. L.  Activities  of pentose shunt  and
     gli'colytic enzymes in lungs of ozone-exposed rats.   Arch. Environ.
     Health 26:206-208, 1973.

11.   Graham,  J.  A.; Menzel, D.  B.;  Miller, F. J.; Illing, J.  W.;
     Gardener, D. E.  Influence of  ozone on  pentobarbitol-induced
     sleeping time  in mice, rats, and  hamsters.  Toxicol. Appl.
     Pharmacol.  61:64-73,  1981.

12.   Melton,  C.  E.   Effects of  long-term exposure to low levels of
     ozone: A review.  Aviat. Space Environ. Med. 53:105-111, 1982.

13.   Weibel,  E.  R.   Morphological basis of alveolar-capillary gas
     exchange.  Physiol. Rev. 53:419-495, 1973.

14.   Miller,  F.  J.; McNeal, C.  A.;  Kirtz, J. M.;  Gardener, D. E.;
     Coffin,  D.  L.; Menzel, D.  B.  Nasopharyngeal removal of  ozone in
     rabbits and guinea pigs.   Toxicology 14:273-281,  1979.

-------
                                                                     166

15.  Yokoyama, E.; Frank,  R.   Respiratory uptake of  ozone  in dogs.
     Arch. Environ. Health 25:132-138,  1972.

16.  Morrman, w. J.; Chmiel,  J.  J.;  Stara,  J.  F.; Lewis, T.  R.
     Comparative decomposition of ozone in the nasopharynx of beagles.
     Arch. Environ. Health 26:153-155,  1973.

17.  Chan, T. L.; Schreck, R. M.; Lippmann, M.  Effect  of  the
     laryngeal jet on particle deposition in  the human  trachea  and  upper
     bronchial airways.  J. Aerosol. Sci. 11:447-459, 1980.

18.  Goldstein, B. D.; Balchum,  0. J.  Effects of ozone on lipid
     peroxidation in the red  blood cell.  Proc. Soc. Exp.  Med.  26:356-
     358, 1967.

19.  Menzel, D. B.; Roehm, J.; Lee,  S.  D.  Vitamine  E:  The  biological
    .and environmental antioxidant.   Agr. Food. Chen. 20:431-485, 1972.

20.  Murphy, R. C<; Clay,  K.  L.   Synthesis and back  exchange of *'0
     labeled amino acids for  use as  internal  standards  with  mass
     spectrometry.  Biomed. Mass Spectiom.  6:309-314, 1979.

21.  Jauquez, J. A.   Compartmental  Analysis  in Biology and  Medicine .
     Elsevier Publishing Co., New York, pp. 103-104, 1972.

22.  Morrow, P. E.  Alveolar  clearance of aerosols.  Arch. Intern.  Med.
     131:101-108, 1973.

23.  Lee, P. S.; Chan, T.  C.; Hering, W. E.  Long-term  clearance  of
     inhaled diesel exhaust particles in rodents. J. Tox. Environ.
     Health 12:801-813, 1983.

24.  Green, G. M.  Alveolobrcnchiolar transport mechanisms.   Arch.
     Intern. Med. 131:109-114, 1973.

25.  Myers, B. F.; Bartle, E. R.; Erickson, P. R.; Meckstroth,  E. A.
     Laboratory generator for batch  synthesis of pure ozone. Anal.
     Chem. 39:415-416, 1967.

26.  Laskin, S.; Drew, R.  T.   An inexpensive  portable inhalation
     chamber.  Am. Ind. Ass.  J.  31:6^45-646, 1970.

27.  Folch, J., Lee, M., and  Sloane-Stanley,  G. H.  A simple method for
     the isolation and purification  of total  lipids-fron animal tissue.
     J. Biol. Chen. 226:497-509, 1957.

-------
                                                                 167
                           CHAPTER 5

     DETECTION AND QUANTITATION OF AUTOXIDATION IN TISSUE:
A STUDY OF HEPATIC AUTOXIDATION INDUCED BY CARBON TE7RACHLORIDE

-------
                                                                     168



     Evaluation of autoxidation in biological systems has been based on



the measurement of oxidation products of olefinic fatty acids (1-4).



Damage is not limited to lipids, however.  Generally not considered but



equally important in the assessment of cellular dysfunction is oxidation



of other cellular components, particularly oxidation of proteins.  In



this study, ''Oj was used to determine incorporation of oxygen into



residual organic material during autoxidation in the hepatocytes in



vivo.  Determination of isotopic enrichment provides a sensitive



measurement of autoxidation of all cellular components.  Conditions



producing hepatic autoxidation were created by interperitoneal injection



of carbon tetrachloride.  Metabolism of carbon tetrachloride by the



Bicrososnal cytochrome P-450-mediated monooxygenase system in the liver



produces CC1, and Cl radicals (5,6) which initiate autoxidation (5-7).



Correlation between autoxidation and cytochronie P-450 activity is



demonstrated by modulation of isotopic enrichment in liver tissue.





                            Lipid Peroxidation



     Peroxidation of unsaturated fatty acids results in the formation of



aldehydes, kotones, ethers, esters, epoxides, carboxylic acids, and



hydrocarbons (8).  The peroxidation of linolenic acid (18:3A*> iaf**),



for example, can result in the production of propane, 1,3-propion-



dialdehyde (cialonaldehyde), and 9,11-dodecadienic acid during autoxida-



tion of olefinic fatty acids (Figure 5-1).



      Peroxidation of Linolenic Acid. . Initiation of lipid peroxidation



(reaction 1) occurs by abstraction of hydrogen from an allylic position



by a free radical (8).  Intramolecular rearrangement of the organic



radical (reaction 2) yields a more s'.able conjugated system (9,10),



which reacts with molecular oxygen (reaction 3) to form a peroxy radical

-------
                                                                          169





169
     Figure 5-1.  Peroxidation of linolenic acid.  Peroxidation of



linolenic acid leads to the formation of hydrocarbons,  conjugated



dienes, malonaldehyde, and fluorescent products during autoxidation.

-------
        170
(I)
(2)
(3)
(4)
(5)
(6)
(7)

-------
                                                                     171




(8,11,12).  Only one of four possible iscmeric peroxy radicals is shown



in this example (1,8).  Intramolecular cyclizaticn of the peroxy radical



(reaction 4) leads to the formation of a five-membered ring peroxide



(1).  Propagation of autoxidation proceeds by abstraction of hydrogen



from other cellular components.  Formation of an a,7~diperoxy radical



(reaction 5) is followed by formation and subsequent decomposition ot a



diperoxy radical (reaction 6),  yielding propane, 1,3-propiondialdehyde



(malonaldehyde), and 9,11-dcdecadienic acid (1).  Hydrocarbon production



also .results from decomposition of mono-hydropercxides where ethane and



pentane are the major products  (4).  Malonaldehyde readily reacts



(reaction 7) with primary amines to form a di-Schiff base adduct (3).



Less than 2% of the malonaldehyde produced during autoxidation of



cellular material remains unbound (1).



      Measurement^ of Lipid Peroxidaticn in vivo.  The presence of



malonaldehyde has been considered indicative of lipid peroxidation in



biological matrices.  Addition  of 2-thiobarbituric acid to a solution



containing malonaldehyde produces a water-soluble colored complex, which



is a colorimetric indicator of  the concentration of malonaldehyde.



Analysis of lipid peroxidation  in tissue requires liberation of bound



malonldehyde, usually by heating in an acidic medium (1).  Alterna-



tively, production of conjugated dienes, inferred from an increased



ultra violet absorption, in tissue has been used as evidence of lipid



peroxidation.  Absorption occurs between 220 and 250 na, with 
-------
                                                                     172



to fluoresce with an excitation maximum at 360 nm and an emission



maximum at 430-440 nm.  Measurement of the fluorescent products in the



chloroform-methanol extract of tissue also has been used as an indica-



tion of lipid peroxidation of biological material (16-18).  Most



recently, measurement of volatile hydrocarbons evolved in the breath has



been used as an assay for lipid peroxidation (4).



     Determination of the products of peroxidation of unsaturated lipids



as a measure of autcxidation neglects oxidation of all other classes of



compounds.  The most general assay for autoxidation would be measurement



of the incorporation of oxygen into organic material.  Although the



formation of Schiff-base adducts results in the loss of 0,-derived



oxygen to water, other pathways, not shown on Figure 5-1, lead to the



formation of stable products which retain 0,-derived oxygen (8,19).



Retention of the 0,-derived oxygen in the organic residue is also



expected with oxidation of other cellular constituents.





                   Carbon Tetrachloride Hepatotoxicity



     Hepatotoxicity associated with carbon tctrachloride poisoning is



characterized by the rapid intracellular accumulation of triglycerides.



This dysfunction arises from inhibition of incorporation of lipids into



very low density lipoprotein in plasma (5).  Injury to hepatocytes



proceeds from swelling of the rough endoplasmic reticulum, with



decreases in microsossal enzyme activity and protein synthesis, to



mitochondrial swelling and eventual necrosis (6).



     Controversy arises, however, as to the nature of the histochemical



lesion.  Two distinct mechanisms have *-een proposed.  One hypothesis



holds that CC1, and Cl radicals, produced during metabolism of CC1, by



the microsomal cytochrome P-450-mediated monooxygenase system, initiate

-------
                                                                     173


autoxidative processes that lead to loss of structure and function of


macromolecules, primarily membrane proteins (6).   Initial damage is


localized to the rough endoplasmic reticulum,  the intracellular site of


microsomal monooxygenases, but spreads throughout the cell as


autoxidation progresses (6).  An alternative hypothesis suggests that


irreversible binding of carbon trichloride and chlorine radicals to


cellular components, particularly microsomal and mitochondrial membrane


proteins, is the cause of histological and biochemical dysfunction (20).


In both views, hepatotox:city is related to metabolism of carbon


tetrachloride.  The occurrence of lipid peroxidation, as indicated by


the production of conjugated dienes (6,20) and ethane (21,22) during

                             ^!
exposure to carbon tetrachlcride, is undisputed.  Autoxidation of other


cellular components has only been inferred.


      Microsomal Cytocnrorne P->450.  The hepatic microsomal cytcchrome


P-450-mediated monooxygenase system contains at least six distinct


isoenzymes with different, but somewhat overlapping, substrate


specificties (23).  Levels of specific isoenzymes can be increased by


repeated administration in vivo of a variety of drugs and toxins which


are normally substrates for these enzymes (24).  Induction by


phenobarbital affects four isoenzymatic forms and increases the rate of


carbon tetra- chloride metabolism (25).  Phenobarbital induction


increases polysomal P-450 monocxygenase mKNA by 30-fold (26); however,


it is not associated with the Ah (aromatic hydrocarbon) locus (27).


Thus, activities of NADPH-dependent cytochrome c reductase, HADPH-


dependent cytochrome P-450 reductase, epoxide hydroxylase, and


glutathione S-transferase are generally unaffected  (28).


     A wide variety of compounds has been found to inhibit cytochrome

-------
                                                                     174



P-450 activity (29).  Complete inhibition of carbon tetrachloride



metabolism in vivo can be obtained by administration of piperonal



butoxide.  Inhibition occurs by metabolic activation of piperonal



butoxide to generate an intermediate which has a high affinity for Fe»*



contained in the heme group of cytochrome P-450 (29-33).  Metabolic



activation requires HADPH and 0,, and a carbanion-cytochrome P-450



complex, which exists in a pH-dependsnt equilibrium, is formed (29,32).



The carbanion-cytochrome P-450 complex is stable in vivo (30), but, as



indicated by the Type I difference spectrum, it does not involve



covalent binding or destruction of cytochrome P-450 (32,33).



      Carbon-Tetrachloride-Induced Autoxidaticn.  In this study,



oxygen-18 labeled 0, has been used as a tracer for molecular oxygen



incorporated into biomolecules in the liver during autoxidation.



Autoxidation was induced by metabolism of carbon tetrachloride by



microsomal cytochrome P-450-mediated monooxygenases.



     Kale Fisher 344 rats were used in this study.  Animals were divided



into five groups.  A summary of treatment of each group appears in Table



5-1.  Animals in Group I were untreated and used to establish the



natural abundance of oxygen-18.  Animals in Groups  II through V were



ventilated by a mechanical respirator containing a mixture of H, and



"0, (80:20) Cor 60 minutes.  Animals in Group II represented an experi-



mental control, and received l*0, as well as I.P. injections of corn



oil, but did not receive CC1,.  Animals in Groups III through V received



I.P. injections of a solution of CC1« in corn oil.  Hepatic nicroscmal



mixed-function oxidase activity was normal in the animals in Group III,



induced in th* animals in Group IV, and inhibited in the animals in



Group V.  A portion of the liver from each animal was retained.  Another

-------
                                                                     175
              Table 5-1.   Sunrnary  of Treatments  of  Groups
                                                 pheno-      piperonal
   Group     »«0,        com oila    CC14        barbitalc     butoxide
     II
    III
     IV
adose » 5 mL/kg body weight.

 dose • 1 mL/kg body weight of carbon tetrachloride.

cdose • 20 ng/day of phenobarbital  for 3 days prior to the exposure.

 dose * 25 lag/kg body weight  of piperonal butoxide 60 oinutes prior to
 receiving carbon tetrachloride.

-------
                                                                     176



portion of the liver was fractionated into an acid precipitate, a



chloroform extract, and a nethanol-water extract.  Oxygen elemental and



isotopic analyses were performed on ill samples.



     Results of isotopic analysis are shown in Table 5-2, and relative



enrichments are s.hown in Table 5-3.  Relative enrichments were



calculated using equation 4-1, where I*FXI NA was tne avera9e isotopic



composition of Group I, 1§F   _ was the average isotopic composition of
                           X .L f LJ


the experimental group, and **F__ • 0.99.  Estimates of the sizes of the
                               IR


oxygen pools were derived from the weights of the sample and oxygen



elemental analyses (Table 5-4).  The amount of 0,-derived oxygen in



whole tissue and in each fraction was calculated using equation 4-2 from

                           v!

the relative enrichment and the size of the oxygen pool in the sample.



Results of these calculations'are shown in Figure 5-2.  Oxygen derived



from 0, is expressed as nanomoles per gram of liver.



     Incorporation of 0,-derived oxygen in animals not receiving carbon



tetrachloride (Group II) arises, in part, from basal activity of hepatic



nonooxygenases and mixed-function oxidases normally involved in metab-



olism, detoxification (34), and phagocytic processes (35).  Isotopic



enrichment in control animals may also reflect indirect incorporation of



0,-derived oxygen into organic molecules by exchange of oxygen (36) with



H,"0 produced by oxidative respiration. The background enrichment in



the livers from animals in Group II corresponds to an average incorpora-



tion of 650 nancooles of 0,-derived oxygen per gram of tissue.  The acid



precipitate retained 18% of the label, the chloroform extract retained



11% of the label, the methanol-water extract retained 32% of the label,



and 39% of the label was lost in the chemical work-up.

-------
                                                                     177
              Table 5-2.  Abundance of Oxygen-18 in Liver
Group     N     	*«F x 10'	
                   liver          acid      chloroform   methanol-water
                 homogenate3  precipitate3   extract3       extract
   I      5        2.0497        2.0416        2.0456        2.0432
                  ± 0.0010      ± 0.0005      ± 0.0009      ± 0.0015
  II      5        2.235         2.092         2.291         2.81
                  1 0.031       ± 0.002       ± 0.027       ± 0.16
 III      6        2.404         2.176         3.07          3.34
                  ± 0.049       ± 0.040       ± 0.15        ± 0.12
  IV      3        2.618         2.261         4.10          3.60
                  * 0.098       ± 0.035       ± 0.29        * 0.12
   V      2        2.24          2.093         2.40          2.780
                  1 0.04        ± 0.017       t 0.13        ± 0.075
ax ± SE for N animals.

-------
                                                                     178
                                                       a
               Table 5-3.  Relative Enrichment in Liver
Group                 	  Relative Enrichment x 10*

I
II
III
IV
V
liver
homogenate
0
190 ± 30
360 * 50
580 ± 100
200 ± 40
acid
precipitate
0
50 ± 2
140 ± 40
220 t 40
50 ± 20
chioroiorm
extract
0
250 ± 30
1040 ± 150
2080 ± 300
360 4 130
water imethanol
extract
0
780 ± 160
1310 ± 120
1780 ± 370
750 ± 80
aReiative enrichments were calculated from average isotopic
 compositions of the samples from each group, where the isotopic
 composition of samples from Group I was used as natural abundance.
 Tha uncertainties are propagated standard errors of the means.

-------
                   Table 5-4.   Oxygen Pools in Liver
                                                                     179
    sample
  weight
    rag
% oxygen
oxygen pool
micromoles
    liver
238 ± 18
 23 * 1
3400 ± 300
    acid
 precipitate
168 ±23^
 23 i 1
2400 ± 350
CHC1, extract3      2.2 ± 0.2
                     14 ± 1
                    19 ± 2
  CHjOH-H,0
   extract"
2.2 ± 0.3
 28 ± 1
  38 ± 5
 analysis of 0.07 aliquot of total chloroform extract.

3analysis of 0.15 aliquot of total methanol-water extract.

-------
                                                                          180
180
     Figure 5-2.  Incorporation of Oj-derived oxygen into liver.  The



incorporation of 0,-derived oxygan during free-radical-initiated



autoxidation is shown.  The amount of 0,-derived oxygen in whole liver,



and the distribution in the acid.precipitate, the chloroform extract,



and the methanol-water extract of liver are shown.  These values are



expressed as nanomoles 0,-derived oxygen/gram liver.

-------
                                                  131
UJDJ6-S8|OUJOUDU '9I-U9BAXQ SS90X3

-------
                                                                     182



     A direct correlation between P-450 activity and incorporation of



02-derived oxygen in the liver by autoxidation was observed in Groups



III through V.  Autoxidation resulted in an average incorporation of 580



nanomoles Oj-derived oxygen/gram liver in Group III in excess of



background levels (Group II).  The molecular distribution of label



within liver was different from that in the control group, in that the



acid precipitate and chloroform extract retained 38% and 36% of the



label, whereas the methanol-water extract retained only 26% of the



label.  A two-fold increase in the amount of 0,-derived oxygen was



observed in the livers of animals in which cytochroms P-450 had been



induced with phenobarbital.  Autoxidation resulted in an average incor-



poration of 1330 nanomoles 02-derived oxygen/gram liver in Group IV in



excess of background levels (Group II).  The distribution of the label



among these fractions was identical to that found in Group III.  A



decrease in isotopic enrichment to levels found in Group II occurred in



the livers of animals in which cytochrome P-450 activity had been



inhibited with piperonal butoxide (Group V).  All 0,-derived oxygen



incorporated into liver by autoxidation (identified by enrichment above



control levels) was recovered in the chemical fractions.  Insignificant



amounts of label were lost during the chemical work-up.



     Isotopic enrichments in the chloroform and methanol-water extracts



represent peroxidation products of unsaturated lipids.  Upon decomposi-



tion of lipid peroxides, Oj-derived oxygen can be retained in the major



portion of the lipid, thus appearing in the chloroform extract, or can



be retained in a small organic fragment (8) which, because of the polar



oxygen-containing functional group, would appear in the methanol-water



extract.

-------
                                                                     183



     The extent of incorporation of label in the acid precipitate is



surprising.  This suggests that macromolecules, proteins in particular,



are also a major target for autoxidation, an observation missed in



previous studies in which only peroxidation products of olefinic lipids



were considered.  Incorporation of the isotopic label in the acid



precipitate could result from binding of carbon trichloro-alkcxy



radicals directly to macromolecules (20).  In fact, species differences



in hepatotoxicity correlate most closely with the incorporation in



macromolecules of radioactivity from 14CC1« as opposed to lipid



peroxidation measured by absorbance at 243 run (20).  Distinction between



these two mechanisms is of minor importance, in that incorporation of



02-derived oxygen in macromolecules by e-ther process most likely



results in loss of structure and function.  Isotopic enrichment of the



tissue-oxygen pool is then a good indication of damage resulting from



autoxidation.





                               CONCLUSIONS



     This is the first study in which the incorporation of oxygen into



tissue during autoxidation has been measured directly.  This development



is significant for the following reasons:  i) this method provides a



sensitive, quantitative assay for autoxidation, ii) these techniques are



applicable to studies of autoxidation both in vitro and in vivo, and



iii) determination of autoxidation is not limited to measurement of the



products of lipid peroxidation, for any class of compounds which retains



oxygen from l»0, can be identified by determination of isotopic



enrichment.  Thus, the procedures discussed in this chapter outline a



general approach to the determination of autoxidation and provide a



unifying approach to understanding the relationship between the

-------
                                                                     184


environmental concentration of a compound and the physicochemical

processes controlling transport, adsorption, and reactivity of the toxic


agent in vivo.



                           EXPERIMENTAL SECTION

      Protocol.  Fivu groups of male Fisher 344 rats (Charles River)

were used in this stady; see Table 5-1 for a summary of the experimental

protocol.  Group I was untreated and used to determine natural

abundance.  Group II was injected I.P. with corn oil at a dose of 5

mL/kg body weight.  Groups III, IV, and V were ir.jected with 5 mL/kg

body weight of a corn oil solution of carbon tetrachloride (20%).  Group

IV received I.P. injections of phenobarbitol at a dose of 20 mg/day for

3 days prior to the exposure.  Group V received an I.P.  injection of
                                 f
piperonal butoxide at a dose of 25 mg/kg body weight 60 nir.utes prior to

receiving carbon tetrachloride.

     Groups II through V were anesthetized with Nembutal at a dose of 50

mg/kg body weight.  The trachea was surgically exposed and cannularized,

and the animal was connected to an artificial respirator (Harvard

Apparatus) containing a mixture of N, and 1§0» (80:20).  The animal was

then injected I.P. with Anectine at a dose of 0.06 mi..  Once the

respiratory rate and heart rate had stabilized, the animal received an

I.P. injection of corn oil or the corn oil solution of carbon tetra-


chloride.  While on the respirator, some animals received additional

Anectine  (0.03 mL) as needed to maintain even and unstressed breathing.


All animals were removed from the respirator 60 minutes after receiving


carbon tetrachloride and allowed to expire.


      Tissue Sampling and Preparation.  Two one-gram liver samples were


taken.  One sample of liver was homogenized in 3 mL of distilled water

-------
                                                                     185



using a Polytron homogenizer (Brinkmann Instruments).  The homogenate



was immediately frozen in liquid nitrogen and lyophilized to dryness.



The sample remained frozen during lyophilization.  Residual water was



removed by placing the tissue homogenates under high vacuum overnight.



     The other sample of liver was used for chemical separation.  One



gram of liver t?as homogenized in 7 mL of chloroform-methanol-water



(2:4:1), with the addition of 2 drops of concentrated HC1.  The



homogenate was centrifuged at 20,000 x g and 5"C for. 20 minutes.  Two



clear layers separated by a pellicle were visible.  The upper layer was



methanol-water and the lower layar chloroform.  The pellicle was removed



with forceps and lyophilized to dryness, and the layers were separated.



      Elemental and Isotopic Analysis.  Samples of the chlorofors.



extract were obtained by evaporation of the solvent in a stream of dry



nitrogen.  The residual material was redissolved in 1.00 mL chloroform,



70 ML of this solution (0.070 aliquot) was added to a silver cup, and



the solvent was removed by evaporation in a stream of dry nitrogen.  A



similar procedure was used to obtain samples of the methanol-water



extract.  Solvent was removed by evaporation in a stream of dry



nitrogen.  The residual material was redissolved in 200 uL of water, 30



uL of this solution (0.150 aliquot) was added to a silver cup, and the



solvent was removed by evaporation in a stream of dry nitrogen.  The



pellicle was weighed, powdered using a mortar and pestle, and



approximately 2 mg of the powder was placed in a silver cup.



     Both oxygen-elemental and isotopic analyses of each sample were



were obtained.  Procedures for isotopic analysis were described in



Chapters 2 and 3.  Working standards (tissue identical to the samples)



were calibrated as described in Chapter 2.

-------
                                                                     186

                                REFERENCES

 1.   Dahle,  L.  K.;  Hill,  E.  G.;  Holman,  R.  T.   The thiobarbituric acid
     reaction and  the autoxidation  of  polyunsaturated fatty acid methyl
     esters.  Arch. Biochem. Biophys.  98:2:53-251,  1962.

 2.   Farmer, E. H.; Koch,  H. P.; Sutton, D. A.   The course of
     autoxidation  in polyisoprenes  and allied compounds.   Part V.
     Observations  on fish-oil  acids.   J. Chem.  Soc. 1943,  122-125.

 3.   Malshet, V. G.; Tappe.1, A.  L.  Fluorescent products  of lipid
     peroxidation:  I.  Structural requirement for  fluorescence in
     conjugated Schiff bases.   Lipids  8:194-198, 1973.

 4.   Dumelin,. E. E.; Tappel, A.  L.  Hydrocarbon gases produced during iji
     vitro peroxidation of polyunsaturated  fatty acids and decomposition
    • of preformed  hydroperoxides.  Lipids 12:894-900, 3°77.

 5.   Recknagel, R.  0.  Carbon  tetrachloride hepatotoxicity. Pharmacol.
     Rev. 19: 145-1  , 1967.

 6.   Recknagel, R.  0.; Glende, E. A.   Carbon tetrachloride hepato-
    • toxicity:  An  example of lethal cleavage.   Crit. Rev.  Toxicol.
     2:263-297, 1973.

 7.   Di Luzzio, N.  R.  Antioxidants,  lipid  peroxidation,  and chemical-
     induced liver injury.  Fed. Proc. 32:1875-1881, 1973.

 8.   Frankel, E. M.  Hydroperoxides.    Lipids and Their Oxidation.
     Schultz, H. W.; Day,  E. A.; Sinnhuber, R.  0.  (eds),  The Avi
     Publishing Company, Inc., Westport, Connecticut, 1962, pp. 51-78.

 9.   Farmer, E. H.; Koch,  H. P.; Sutton, D. A.   The course of
     autoxidation  in polyisoprenes  and allied compounds.   Part VII.
     Rearrangement of double bonds  during autoxidation. J. Chem. Soc.
     1943, 541-547.

10.   Farmer, E. H.; Sutton,  D. A.  The course of autoxidation  in
     polyisoprenes and allied  compounds. Part  XI.  Double bond movement
     during autoxidation of  a  monoolefin.  J. Chem. Soc.  1946, 10-13.

11.   Farmer, E. H.; Sutton,  D. A.  The course of autoxidation  in
     polyisoprenes and allied  compounds. Part  IV.  The isolation and
     constitution  of photochsmically  formed methyl oleate peroxide.  J.
     Chem. Soc. 1943, 119-122.

12.   Bolland, J. L.; Koch, H.  P. The course of autoxidation in
     polyisoprenes and allied  compounds.  Part  IX.  The primary thermal
     oxidation products of ethyl linoleate.  J. Chen. Soc. 1945, 445-
     447.

13.   Bidlack, W. R.; Tappel, A. L.  Fluorescent products of
     phospholipids during peroxidation.   Lipids 8:203-207, 1973.

-------
                                                                     187

14.  Dillard, ('.. J.; Tappel,  A.  L.   Fluorescent products from reaction
     of peroxiciizing polyunsaturated fatty acids with phosphatidyl
     ethanolamine and phenylalanine.  Lipids 8:183-189/ 1973.

15.  Reiss, U.; Tappel, A. L.  Fluorescent product formation and changes
     in structure of DKA reacted with peroxidizing aracnidonic acid.
     Lipids 8:199-202, 1973.

16.  Fletcher, B. L.; :.'appel, A. L.  Fluorescent modification of serum
     albumin by lipid peroxidation.  Lipids 6:172-175, 1971.

17.  Dillard, C. J.; Tappel,  A.  L.   Fluorescent products of lipid
     peroxidation of mitochondria and microsomes.  Lipids 6:715-721,
     1971.

18.  Trembly, R.; Tappel, A.  L.? Coniglio, J. G.; Grogan,Jr., W. M.;
    . Rhamy, R. K.  Fluorescent products and polyunsaturated fatty acids
     of human testes.  Lipids 10:591-596, 1975.

19.  Keenen, M.  Secondary degeneration products.  Lipids and Their
     Oxidation.  Schultz, H.  W.; Day, E. A.; Sinnhuber, R. 0. (eds), The
     Avi Publishing Company,  Inc.,  Westport, Connecticut, 1962, pp.
     51-78.                      *!

20.  Diaz Gomez, M. I.; de Castro,  C. R.; D'Acosta, de Fenos, de
     Ferreyra, E. C.; Castro, J. A.  Species differences in carbon
     tetrachloride-induced hepatotqxicity: The role of CC14 activation
     and of lipid peroxidation.   Tbxicol. Appl. Pharm,. 34:102-114, 1975.

21.  Roster, V.; Albrecht, D.; Kappos, H.  Evidence for carbon
     tetrachloride- and ethanol-induced lipid peroxidation in vivo
     demonstrated by ethane production in mice and rats.  Toxicol. Appl.
     Pharmacol. 41:639-648, 1977..

22.  Burk, R. F.; Lane, J. M.  Ethane production and liver necrosis in
     rats after administration of drugs and other chemicals.  Toxicol.
     Appl. Pharmacol. 50:467-478, 1979.

23.  Lu, A. Y. H.; West, S. B.  Multiplicity of mammalian microsomal
     cytochrcmes P-450.  Pharmicol. Rev. 31:277-295, 1980.

24.  Herbert, D. N.  Multiple forms of inducible drug-metabolizing
     enzymes:  A reasonable mechanism by which any organism car cope
     with adversity.  Mol. Cell. Biochem. 27:27-46, 1979.

25.  Walz, F. G.; Vlas-jk, G.  P.; Omiecinski, C. J.; Bresnick, E.;
     Thomas, P. E.; Ryan, D.  E.; Levin, W.  Multiple, immunoidentical
     forms of phenobarbitol-induced rat liver cytochrome P-450 are
     encoded by different mRNAs.  J. Biol. Chem. 257:4023-4026, 1982.

26.  Adesnik, M.; Bar-Nun, S.; Maschio, F.; Zunich, M.; Lippman, A.;
     Bard, E.  Mechanism of induction of cytochrome P-450 by
     phenobarbitol.  J. Biol. Chem. 256:10340-10345, 1981.

-------
                                                                     188

27.  Pelkonen, 0.; Herbert,  D.  W.   Metabolism of polycyclic aromatic
     hydrocarbons: Etiologic role  in carcinogenesis,  Pharmicol. Rev.
     34:189-222, 1982.

28.  Herbert, D.-W.; Jensen, N. M.  The Ah locus:  Genetic regulation of
     the metabolism of carcinogens, drugs, and other environmental
     chemicals by cytochrome P-450-mediated monooxygenases.  Crit.
     Rev. Biochem. 6:401-437, 1979.

29.  Testa, B.; Jenner, P.  Inhibitors of cytochrome P-450s and their
     mechanisms of action.  Drug Metab. Rev. 12:1-117,1981.

30.  Ullrich, V.; Schnabel,  K.  H.   Formation and binding of carbanions
     by cytochrome P-450 of  liver  microsomes.  Drug Metab. Dispos.
     1:176-183, 1973.

31. . Hodgson, E.; Philpot, R. M.  Interaction of methylenedioxy-phenyl
     (1,3-benzodioxole) compounds  with enzymes and their effects on
     mammals.  Erug Metab. Rev. 3:231-301, 1974.

32.  Franklin, M. R.  Inhibitors of mixed function oxidases by
     substrates forming reduced cytochrome P-450 metabolic-intermediate
     complexes.  Phanaacol.  Ther.  A 2:227-245, 1977.

33.  Ullrich, V.  The mechanism of cytochrome P-450 action.   Microsomes
     and Drug Oxidation . Ullrich, V.; Roots, I.; Hildebrandt, A.;
     Estabrook, R. W.; Coony, A. H., (eds.), Pergaman Press, Oxford,
     1977, pp. 192-201.

34.  Coon, M. J.  Oxygen activation in the metabolism of lipids, drugs,
     and carcinogens. Nutrit. Rev. 36:319-328, 1978.

35.  Johnson, R. B.  Oxygen  metabolism and the microbicidal activity of
     oacrophages.  Fed. Proc. 37:27-59-27-64, 1978.

36.  Murphy, R. C.; Clay, K. L.  Synthesis and back exchange of 100
     labeled amino acids, for use as internal standards with mass
     spectrometry.  Biomed.  Mass Spectroia. 6:309-314, 1979.

-------