APPROVAL SllEET 
Title of Dissertation: The Kinetic s a nd Mechani s m of 
Sedimentary Iron Sulfide Formation 
Name of Cand idat e : Alb e rt 
Doctor 
Dissertation a nd Abs t ract Approved. 
Dr. Sheldon E . Somm e r 
Associate Profe ss or 
Department of Ge olo gy 
Da tc Approved :~ /t}1 J1'7t.{ 
THE KINETI CS AND MECHANI SM 
OF 
SEDIMENTARY IRON SULFIDE FORMATION 
by 
Albe r t John Pyzik 
Di ssertation submitted to the Faculty of th e Graduate Sc hool 
of the Univer si ty of Maryland in partial fulfillment 
of th e r equir ement s for the degree of 
Docto r of Philo sophy 
1 976 
ABSTRACT 
Title of Diss e rtation: The Kinetics and Mechani s m of 
Sedimentary Iron Sulfid e Formation 
Albert John Pyzik, Doctor of Philo s ophy, 1976 
Di s sertation directed by: Dr. Sheldon E. Sommer 
As s ociate Professor 
Department of Geolo gy 
The reaction between goethit e , a-FeOOH, and aqueous 
bisulfide ion, HS-, was s tudied under conditions repre s en -
tative of estuarine sediment s . Th e conc e ntration-tim e 
curves of the following s pecie s were det e rmined by s pectra -
photometric method s : total s ulfide , di s solve d s ulfid e , 
precipitated sulfide, thiosulfate ion, s ulfit e ion, e lem ental 
sulfur, and di s solved ( <O.lw iron. Poly s ulfides (S~ and S~) 
were monitored by ultraviolet absorbanc e mea s urement s , whil e 
the hydrogen ion concentration wa s d e t e rmine d with a pH 
e lectrode . Elemental s ulfur, both a s fr e e and poly s ulfide 
sulfur was found to be the ma jor s ulfide ox idation product. 
Thiosulfate ion compri se d about 14? 8% ( e l ectron bal a nc e-
wi s e) of the ox idation product s . 
Concentration-time curves o f pr e cipitat e d s ulfide 
s ulfur were analy z d by the initial rat e 1ne thod to d e t e rmin e 
the rate expre ss ion. Th e rat e expr ess ion for th e r eac tion 
between a-FeOOH and HS 1s 
d [FeS]/dt 
where d[FeS]/dt is the r a te of precipitated iron s ulfide 
formation, [HS - J i is the initial total sulfide c once ntra-
tion, (H+)i i s the initial hydrogen ion activity, AF e OOH . 
2 l 
i s the initial goethite s urface area in m /1, and k i s th e 
rate constant with the value 31?10 M- 1 1 - l m- 2 min - 1 . 0. 9 7, 
0.82, and 1.1 are the reaction order s for the s p e cie s 
bi s ulfide ion, hydrogen ion, and goet h i te surface area 
respectively. 
A combination of hydrogen balance a nd electron 
transfer balance and stoichiome tric reaction s we re s tudi e d 
in view of the rate expres s ion to yield a mechani s m. 
The multistep mechanism con s isted of several parall e l 
and consecutive reaction s : (1) th e protona tion reaction of 
t h e goethite surface, (2) the parallel reduction r e action s 
of ferric iron to yield elemental sulfur and thio s ulfate as 
o x idation products, (3) the dis s olution of the f e rrou s 
hydroxide, and (4) the precipitation reaction of di ss olve d 
ferrous species and aqu e ous bi s ulfide ion . Th e rat e d e t e r -
m1n1 n g step in the reaction se quenc e wa s th e dis s olution s tep. 
Result s of thi s study indicat e tha t the ox idation of 
sulfide species by ferric iron may b e a s i gnifi c ant s ource 
of elemental s ulfur in th e s ediment. El e me ntal s ulfur i s 
neces s ary for th e form a tion of pyrit e (F eS 2), th e th e rmo -
dynamically s tabl e iron s ulfid e . 
In addition, the previous studie s of the interstitial 
waters of anoxic sedime nt s s how ed an excess of "di ssolved" 
iron which was grea ter th a n calculated from equilibrium 
so lubility product s . It i s s ug gested from particle size 
s tudies of the precipitat ed iron s ulfide that these high 
concentrations are a re s ult of the s ubmicron particles of 
ferrous s ulfide ( <O .1 ?). These particle s would obviously 
pass through the 0.45 ? filters which are traditionally u sed 
as the dividing line for di sso lv e d and particulate s p ec i es . 
11 
FO REWORD 
Any man's finest hour -- hi s grea t est 
fulfillment to a l l that h e hold s dear 
-- is tha t moment when h e has worked 
his h eart out in a good ca u se a nd lies 
exhausted on the f i e ld of battle --
victorious. 
V.L. 
lll 
DEDICATION 
To my wife and my parents: 
Without them, none of this would have been possibl e . 
lV 
ACKNOWLEDGMENTS 
Many p eopl e and institutions have aided me in my 
work. Their assistance and efforts are great ly appreciated. 
Yet, there are several persons who require a specia l note 
of thanks. 
The financial s upport of the Nat ural Resources 
In s titut e and the Office of Water Res earch and Technolo g y, 
through the Water Resources Research Center of the Univ e r s ity 
of Maryland (Project A026-Md) is g reatly appreciated . 
Th e Center for Materials Research (Electron Micro -
s cope Central Facility) is here by thanked for in s trument 
time on the tr a n s mi ssio n electron mi croscope . The assistance 
and instruction of Mr. Gen e Taylor is gratefully appr ec iat e d. 
Sr . M. Regina Alma and Dr. Harry O. Warren deserve 
my appreciation for their g uidance 1n my ch e mi ca l e ducation. 
The faculty of the Divi sion of Geochemi s try d ese rve 
my thank s for their friend s hip a nd intellectual s upport. 
Spec i a l thanks are ex tended to my re sea rch advisor, 
Dr. Sheldon So mmer, for the guidance, moral s upport, and 
financi a l s upport with which h e ha s provided me . I partjcu -
larly want to thank hin1 for the intellectual freedom that 
he allowed me 1n the pur s uit of my research intere s t s . 
A specia l not e of gratit ude i s exte nd ed to my parent s , 
Mr. and Mr s . Albert Py z ik, for the many y ea r s of love, g uid-
V 
ance , and s acrifi ce which we r e n e c ess ary to g ive me thi s 
opportunity. Thi s little not e cannot truly e x press my 
appreciation for their e ffort. 
To my wife, Iri s , I want to ex t e nd my s ince r e appr e ci -
ation for her love , under s tanding a nd s upport during my 
r es e a rch, but particularly for the final f e w month s of th e 
laborious ta s k. Al s o for h e r me ritoriou s s ervic e a bove 
and beyond the call of duty in th e typing of thi s ma nu s cript. 
Finally, as a fellow s cienti s t who se rve d a s a s ounding 
board for my many ideas, s h e de se rve s my thanks. 
vi 
TABL E OF CONTENTS 
FOREWORD ii 
DEDICATION iii 
ACKNOWLEDGMENTS l V 
TABLE OF CONTENTS vi 
LIST OF TABLES vii 
LIST OF FIGURES viii 
INTRODUCTION 1 
REVIEW OF THE LITERATURE 9 
EXPERIMENTAL 30 
RESULTS AND DISCUSSION 60 
CONCLUSIONS 107 
APPENDIX A: DATA TABLES FOR REACTION RUNS 115 
APPENDIX B: INITIAL RATE METHOD 1 3 5 
BIBLIOGRAPHY 138 
vii 
LIST OF TABLES 
1. Li st of d - spacings (in A) for Smythit e a nd Siderite 17 
2. Summary of Reported Sul f ide Products 29 
3. Palit zsc h Borax-Borate Buffer 36 
4. Variati on of% Thio s ulfate with Initial Reaction 
Conditio n s 63 
5. Rate of Reduction During Phase II (Rate of 
Di ss olution) 79 
6. Hydroge n Ba l a nce 92 
7. Rate of FeS Formation 99 
8. Par ti c l e Size St udy 103 
9 . "Di sso lved Iron" Concentration 104 
viii 
LIST OF FIGURES 
1. Stability Fields of Iron Minerals at 25?C and 1 Atm 
Total Pressure, pH 7 .5, Pco 10- 2 - 52 atm . 11 
2. Schematic Diagram of Reaction Vessel 34 
3. Flow Chart of Analytical Procedure s 38 
4. Sample Sulfide Calibration Curve 42 
5. Sample S? Calibration Curve 44 
6. Sample Thiosulfate Calibration Curve 46 
7. Sample so; Calibration Curve 47 
8. Sample Fe Calibration Curve 49 
9. Titration Curve for Borax-Borate Buffer 51 
10. Equilibrium Distribution of Polysulfide Ions n s= 54 
11. Real and Synthesized Polysulfide Spectra 56 
12. Change of pH Upon Dilution of Reaction Solution 
Due to Addition of Wa t er 58 
13. Concentration-Time Curves for Reactants a nd Product s 
(Run 31) 61 
14. Plot of Acid Extractable Sulfide Concentration vs. 
Reduced Iron Concentration 72 
15. Plot of Reduced Iron and Acid Extractable Sulfide, 
S~, vs. Time (Run 6) 74 
16. Plot of Reduced Iron and Acitl Extractable Sulfide 
S~, vs. Time (Run 24) ' 75 
17. Plot of Rat e of Fe Reduction vs. 
s= 
T- 76 
1 
18. Plot of Rate of Fe Reduction vs. Initial Hydro ge n 
Ion Concentration 77 
19. Plot of Rate of Fe Reduction vs. Initial Goethite 
Surface Area 78 
ix 
20. Plot of Reduced Iron and Acid Extractable Sulfide 
s~, vs. Time (Run 31) ' 81 
21. Idealized Reduction-Dissolution Curve 82 
22. Plot of Rate of Iron Sulfide Formation vs. 
Initial Bisulfide Ion Concentration 96 
23. Plot of Rate of Iron Sulfide Formation vs. 
Initial Hydrogen Ion Concentration 97 
24. Plot of Rate of Iron Sulfide Formation vs. 
In i tial Goethite Surface Area 98 
25. Detailed Reaction Mechanism 108 
--
1 
INTRODUCTION 
Th e formation of sedimentary iron s ulfid e o ccurs as 
a result of a s e ries of complex g eochemical reaction s and 
microbial processes in anoxic sediments. Mackinawit e (FeS), 
greigite (Fe s 4 ) and pyrite (F e S ) are the most comnon iron 3 2
sulfides found in anoxic sediments. These min e ral s form by 
the reaction of detrital iron minerals, particularly fine 
grained iron oxyhydroxide s, a nd bacteriogenic hydro ge n 
sulfide (Berner, 1970). Iron sulfides form in benthic en-
vironments which are characterized by low wav e and curr e nt 
energy, high biolo g ical productivity, restrict e d bottom cir-
culation and high depositional rates. The s e conditions re-
strict the horizontal and v e rtical transport of o x y ge n rich 
waters and permit the accumulation of or ganic matt e r 
(Richards, 1965). Aerobic ox idation of the or g ani c 1n att e r 
reduces the di s solved oxyg e n conc e ntration and r es ult s in 
the formation of anoxic conditions. Typical ano x ic ba s in s 
are the Black Sea (Casper s , 1957), Cariaco Tr e n c h (Richards 
and Vaccaro, 195ij), Santa Barbara Ba s in (Kaplan e t al., 196 3 ) 
and Saanich Inlet (Nissenbaum e t al., 1972). Hi g h biolo g ic a l 
productivity and high sedimentation rates in tida l flat s 
and salt water marshes al s o r es ult in dis s olve d ox y ge n 
depletion. 
In both s ituations , th e d e po s it e d organi c ma tt e r i s 
utilized by a seri es of 1nicroorgani s ms in a sequ e n ce of 
2 
respiratory a nd fermentation processes as follows: 
(1) Aerobic respiration 
CH 20 + o +CO2+ H 0 2 2
( 2) Ni trate reduction 
5 CH20 + 4 No;+ 4 H+ + 2 Nz + 5 CO2+ 7 HzO 
(3) Sulfate reduction 
2 cH 2o + so~+ Hs - + Hco; + H+ 
(4) Methane production 
CO + 4 H2 + CH2 4 
+ 2 HzO 
(In the above reactions, CH20 represents generalized organic 
matter in the form of carbohydrat e .) Reaction two is not 
quantitatively s i gnificant due to the low concentration of 
nitrate in the sediment (Richa rds, 1965). 
Several important effec t s of the sequence are (1) the 
removal of di ss olve d oxygen, (2) re duction o f the redox 
potential (Eh), a nd (3) th e formation of reduced su l f ur 
species. The first two effects are interr e l ated and result 
in the formation of aerobic and anaerobic zones in the sedi-
ment and water columns. Th e boundary betwe e n these biolog i -
cal regions usua lly occurs at or b e low th e sedime nt -water 
interface. several multioxida tion s tat e eleme nt s (Fe Mn 
' ' 
s, c, and N) are present in the sedime nt. The redox boundary 
for each element is a function of the redox s p ecies, con-
centration, and u s ually pH (Garr e l s and Christ, 19 65 ) . 
Sulfate reduction is a mi crobial pro cess that does 
not occur inorganica lly (Hem, 1960; Berner, 1970). The 
important sulfate reducing bacteria belong to the genera 
De s ulfovibrio, Des ul f otomac ulum a nd Clostridium (Trudinger 
';iLL-',r,14 . ,. 
3 
e t al., 1972; Zajic, 1969). These obligate a n aerobes are 
dissimulatory sulfate reducers, tha t is, they reduce s ulfate 
in excess of that needed for cellular growth. Sulfate i s 
the terminal e lectron acceptor in the respiratory process. 
In addition, it is a nece ssary s ourc e of oxygen a nd e n ergy 
for the microbes. Hydrogen sulfide is a re s piratory by-
product which reacts with available chalcophile e lement s to 
form metallic sulfides (Zajic, 1969). 
A few ppm of hydro ge n s ul fide impart s an unpl easant 
taste and odor to air and water . More important, it i s ex-
tremely toxic to most microor gani s ms and hi gh er organisms; 
the ceiling v a lue for human s is 20 ppm (Chri s t e n sen a nd 
Luginbyhl, 1974). It wa s r es pon s ibl e for massiv e fish kills 
in Norway a nd Canada when me teorologica l and hydro graphi c 
conditions resulted in the upw e lling of sulfide - rich bottom 
waters (Ozretich, 1975; Brongersma-Sanders, 1 957) . Di sso lved 
sulfide s hould diffuse out of the se dime nt as a r es ult of 
the concentration gradient that ex i s t s between th e a noxic 
sediment and the overlying wat er column. Fortunately, 
there are several processe s which can remove o r bind hydro ge n 
sulfide in the e nvironment: biological ox idation, chemical 
oxidation and me tal s ulfide precipitation. The l ast two 
me chanisms a re o f particular interest in this study. Iron 
sulfide precipit a tion would bind the hydro gen s ul fide or the 
bi s ulfide ion in the sedimen t since the meta l s ul fides are 
ex tremely insoluble. The process would onl y be effective if 
the rate o f r eac tive me tal additio n to the s e diment exceed s 
4 
the rate of sulfate reduction. In addition, the concentra-
tion of hydrogen sulfide can be controlled by the coupling 
of the oxidation of hydrogen sulfide with the reduction of 
ferric iron. This is extremely likely since hydrogen sulfide 
is a strong reducing agent. This redox process would convert 
toxic hydrogen sulfide into several sulfur oxyanions such 
as sulfite, thiosulfate, and sulfate. However, some of these 
oxidation products (e.g. thiosulfate) can be utilized by 
Desulfovibrio and h ence would be recycled by the bacteria. 
Both chemical oxidation and metal precipitation could reduce 
the concentration, the concentration gradient, and ultimately 
the diffusion of H7S out of the sediment. 
/.., 
Dissolved sulfate is the principal source of sulfur 
for iron su lfide formation . Marine organic matter contains 
approximat e ly 1% organic sulfur, and thus represents less 
than 1 -2% of the s ulfur budget for iron s ul fide formation 
(Kaplan, Emery, and Rittenberg, 1963). Dissolved sulfate 
is trapped with organic and inorganic detritus upon deposi-
tion and is reduced by bacterial processes. Additional sul -
fate diffuses into the sediment from the overlying water in 
response to the concentration gradient that develops because 
of the sulfate reduction. The diffusion of s ul fate is sub-
stantiated by sulfur isotope ratios which show a progre s sive 
enrichment of 32s in the reduced sulfur species . This en-
richment can only be explained by assuming acce s s by the 
sediment to the unlimited sulfate re s ervoir in the over -
lying waters (Thode and Kemp, 1968). 
-
5 
Two processes are important in t h e removal of sulfate 
fr om seawat er : calcium s ulfa te formation and sulfate re-
duction-pyrite formation (Ber n er, 1972). At present, there 
are no quantitatively signif icant evaporite basins for 
ca l c ium s ulfat e formation, h e n ce pyrite formation must be 
the major control of oc ea ni c sulfate concentration. Calc u -
lations by Berner (1972) for s ulfur upt ake due to pyrite 
format ion in pelag ic and sem i- euxin ic basins can only ac-
count for 5% of sulfate d e livered by river wa t er . Ber n er 
thu s proposed that either me t a l s ul f id e for mation is much 
high e r in shallow water e nvironment s or the seawater sulfate 
concentration i s incr eas ing. Therefore an increased und e r-
s tanding o f the formation of iron s ulfid es in nearshore a n d 
..,  
es tuarin e e nvironments may al low exp l anat ion of the dis- ,,, 
"~ 
crepanci es in the s ulfur budget. .  ?,  
Several po ss ibl e so u rces o f iron are present i n a ,'''    
.. ;'  
reducing sedim e nt: dissolved iron, organic iron, detr i tal .~ 
1 
iron minera l s , iron oxide a nd oxyhydroxide coatings on min - ''  j 
! 
e ral gra ins , and s tructural iron in c lay minera l s . Previous 
work ( Gibb s , 197 3 ; Carroll, 1 958 ; Drever, 1971) has shown 
that only th e last two s ources are quantitatively signifi-
cant. Iron oxide coatings are formed as a res ult of weath-
er1ng and s oil forming processes. These oxi d es are g oethite, 
lepidocrocite, ha ema tit e, a nd iron oxides of indefinite 
s tructur e (Carroll, 1 958 ). Altho u g h they are fo und o n all 
<l e trit a l particl es, th ey are mainly associated with the c lay 
s iz e d fraction (Berner, 19 6 4a). Von Straaten (1954) showed 
6 
that the principal source of iron su lfide formation is ad -
sorbed iron in the clay fraction. Precipitated and copre-
cipitated iron oxide coatings comprise 40-47 % of the river 
load of iron (Gibbs, 1973). 
These iron coating s are stable in an oxidizing envi -
ro nmen t but are removed by reducing conditions. Ferric 
hydroxide (Fe(OH) ) is the thermodynamically stable phase 
3
in oxidizing environments. However, the lowering of Eh be -
low +350 mv at pH 8 and O mv at pH 6 (Hem and Cropper, 1959) 
results in the reduction of ferric iron to ferrous iron. 
The latter is more soluble and more mobil e. The particular 
ferrous species which predominates 1s a function of the 
salinity and the pH . At pH 7 and in fresh water, Fe+ 2 is 
the most stable species, while at pH 8, FeOH+ and Fe+Z are 
equally important. In saline waters at pH 7, Feel+ +2 
' Fe , 
,,,I  
I 
:1 
and FeOH+ are important, while FeOH+ predominate s at pH 8 ;,1 ,,,,I  
(Kester et al .. , 1975). :,,,,:  / 
Structural iron (i.e. that within the crystal struc -
ture) comprises 45-48% of the total river load of iron 
(Gibbs, 1973). This iron species was assumed to be unavail-
able for reaction. However structural iron may b e important 
where other iron sources are not present . Drever (1971) 
determined from studies of nonexc hangeable ma g n es ium in clay 
minerals that s tructural iron in clay minera ls can b e an 
additional source of iron for FeS formation. Hi gher values 
of nonexchangeable magnesium were determined in reducing 
sediments than in mineralogically similar oxidizing 
7 
sediments. This was attributed to replacement of ferro u s 
iron in the octahedral l ayer by ma gnesium. This woul d be 
significant in sediments which contain clay minerals rich 
in iron, e.g. nontronite, glauconite, and chamosite. 
Iron sulfide formation affects the distribution of 
trace metals in reducing sediments. Reactant iron oxides 
and oxyhydroxides adsorb trace metals (Co, Ni, Cu and Zn) 
in oxidizing environments (Jenne, 1968). During iron s ul-
fide formation, the adsorbed trace and heavy metal s would 
be released upon dissolution of the iron oxyhydroxide s . 
Krauskopf (1956) indicated that the concentrations of Cu, 
' 
Zn, and Cd 1 n marine sediments could be explained by pre- ~ 
~ 
cipitation as the metal sulfides. On short time scales .??1  ' ~ 
~ 
J 
the concentration of Mo in anoxic sediments was shown to b e 
controlled by coprecipitation with the iron ~ulfides 
(Bertine, 1972). Thus in the ory, th e conc e ntrations of 
chalcophile elements are controlled by the formation of j 
! 
metal sulfides. Studies of trace metal conce11trations in 1 1, I  
interstitial waters of reducing sediments (Brooks et al., I 
1968; Preseley et al., 1972) have shown that the concentra-
tions of several chalcophile elements (Cd, Co, Cu, Fe, Ni, 
and Zn) exceeded the equilibrium concentrations by sever a l 
orders of magnitude. Polysulfide a nd bisulfiJe complexes 
were proposed to explain the g reat e r s olubility of the 
metals (Barnes and Czamanski, 1967; Krauskopf, 1956). 
A better understanding of the formation of sedini ent-
ary iron sulfides i s n eed ed b ecau se of their i mportan ce ia 
C::::-cT -? r: <0 . ~ ? ? - . 
8 
several environmenta l and geological processes. The distri-
bution of toxic H2S in the se d iment water column is affected 
by the precipitation of i ron sulfides . The rate of this 
process is important. The precipitation rate must equal or 
exceed the rate of sulfate reduction to preve nt the accumu -
lation of sulfide and eventual diffusion of H2S into the 
overlying waters where it could adversely affect aerobic 
organisms. 
The concentration, distribution, and mobility of 
chalcophile elements is theoretically controlled by the sol-
ubility of metal sulfides. Recent data (Brooks, Preseley, 
,, 
and Kaplan, 1968; Preseley, et al., 1972) indicate trace ,i;i  ,, 
I'~ 
metal concentrations i n interstitial waters greater than pre- :,~ 1"
,,,:~,
' 
 , 
,, 
dicted by solubility calculations. Possible intermediate 
1,.:,!  
species in the mechanism of iron s ulfide formatio~ such as :'!,  
?,,,l  
polysulfide~ could explain the high solubility. Unfortu- ,, 
,11 I 
nately, no data exist for intermediate species. 
:/:J,' I ,, 
Geologically, the oceanic sulfur budge t appears to ??:,,  
I 
be out of balance. There exists no quantitatively signifi - ! 
cant sink for dissolved sulfate; how ever, the rates for near-
shore formation of iron sulfides are not known. 
Thus additional data on the rates, kinetics, a nd 
mechanisms for the formation of the various iron s ulfid es 
are needed before their impact on these other processes can 
be adequately assessed. 
J 
9 
REVIEW OF THE LITERATURE 
"Iron sulfide" is u sed as a ge n eric term to de scr ib e 
the several known minerals of iron and s ulfur that form a t 
low termperature in nature. Mackinawite, pyrit e, g reigite, 
smythite, marcasite and pyrrhotite have b een identifi ed 111 
nature, although only mackinawite, greigite, a nd pyrit e have 
been found in recent sediments. Mackinawite (FeS) is the 
most common iron sulfide found in anoxic sediment s a nd pro-
duced in laboratory studies. In the past it ha s b een d e-
scribed by several name s : troilite, hydrotroilite , k a n s it e ' 
tetragonal FeS, precipitated FeS and amorphous FeS. Berner 
(1964a)and Rickard (1969aj agreed that all of these mat erials !:~ 
.,.,,  
were probably mackinawite. r 
,f 
Greigite (Fe s ) i s a cubic magnetic iron s ulfide ,:.l 3 4 , 
:.. ;'  
(Skinner et al., 1964). It does not form dir ec tly fro m so lu- ,, ,,, 
?'l 
tion, but by the reaction of mackinawit e with either ad - ::,J 
sorbe4 coprecipitated, or solution sulfide. Greigite is 
believed to be identical with reported cases of melnikovit e 
(Berner, 1964). Both mackinawi te a nd greig i te are thermo-
dynamically unstable relative to s toi c hiom e tr ic pyrrhotite 
and pyrite (Berner, 1967). Th e sedimentary record c ontains 
no occurences of e ither minera l in sedime nt s older tha n 
the Tertiary. 
Although pyrite is the thermodynamically s t ab l e p 11 as c, 
pyrrhotite may form if ther e i s i n s uf ficient elemental 
h 
10 
sulfur present. 
FeS + so + FeS ~G = -17.0 2 kcal 
Fe s + 2 so+ 3 FeS 2 ~G = -45.5 kcal 3 4 
FeS + FeS ~G = - 1.9 kcal mack pyrr 
However, there is no evidence for authigenic pyrrhotite for-
mation in recent sediments. Metastability of mackinawi t e 
and greigite is a problem (Berner, 1967). 
Stability field diagrams of iron minerals in anoxic 
sediments are difficult to represent on the traditional Eh-
pH diagrams because the variable s PCOz and ps= are a l so in-
volved. Fortunately, the pH varies only from 6.9 to a. 3 
(Ben Yaakov, 1973) while the P ranges from 10- 3 ? 5 to 10 -1_ COz 
Thus Eh -pS= di ag rams at constant pH and PCOz yield the sta-
bility field diagram in Fi gur e 1. This figure represents 
the average situation of anoxic marine sed iments (pH= 7 _5 
,,, 
and pco = 10- 2 ? 5). The broken line in the pyrit e fie ld ~??, 
2 ',,
  
j 
represents th e Eh-pS = values r ecorde d for anoxic marine ! 
sediments by Berner (1964b). Th ese values indicat e that the 
Eh of reducing sulfide sedim ent s i s controlled by tli e s ul f ur-
s ulfide half cell 
HS + s0 + H+ + 2 e 
The Eh-pS= values indicate that th ermodynami c equil ibri um 
is be ing established and meta stab l e FeS is be ing a lt ered to 
s table pyrit e . 
0.00 
metastability / ~ 
boundary . 
-0.10 r / / ,' / . 
natural / . 
systems , / / 
/ 
.r'GREIGITE . / ~CKINAWITE 
/ 
-0. 20 \- / Fe 3s/ 4 / 
FeS 
/ 
/ / 
Eh I / MAGNETITE 
/ Fe 3o4 
-0.30 I- / PYRITE FeS
/ 2 ~?~  
/ ~ 
!- SIDERITE - 0 . 40 / I FeC0 ' 3 ~ ~ 
H - - ... ~ ........... ....... .............. .J....1 20 / FeS I I 
Hz - I I 
-0.50 
6 . 0 8. 0 1 0.0 1 2 .0 1 4 . 0 16.0 1 8. 0 
ps= ' 
Figure 1 . Stability f ields of _Iro~ Minerals at 25?C and 1 Atm Tot~l fr~ssure, pH 7. 5 , I-' 
Pc oz 1 0 - 2 ,;:iatm. This diagram represents the naverage situation" expected I-' 
for many anaerobic sediments (Berner, 196 4) . 
?--.:..~::::_~ ~ --~_:~-~ ':-. ~-~-~~ :...~ -~~.;.,~.;:.:...:. ..~. ,'--
~ 
12 
Previous studies of the sedimentary occurences of 
iron s ulfides have been limited by the physical and chemical 
characteristics of these minerals. Tetragonal and amorphous 
Fe s , and p yr i t e , ( 1 ) occur in 1 ow con c en tr at ions, ( 2 ) are p O O r 1 y 
crystalline, (3) are extremely fine grained and (4) are sus-
ceptible to air oxidation . The low concentration and poor 
crystallinity hinder the identification of these minerals 
by x - ray diffraction because they produce broad peaks or no 
peaks at all. Identification by microscopic analysis 1s limi-
ted by the fine grain size, while extreme care is needed in 
handling to prevent air oxidation. Thu s initial studies 
have been directed to the characterization of these mineral s 
and the physico-chemical conditions that are necessary for 
their formation. Several modern laboratory studies have 
been directed to this end. 
Berner (1964a)synthesized several iron s ulfides from 
aqueous solutions at atmospheric pressur~ t emperature s from ,, 
20 0 to 9 aoc and pH conditions from 3 to 9. The purpose was 
::J 
?,:,,  
to identify tl1 e l?ron sulfides that were preparecl  under simu -
1 
lated conditions by x - ray diffraction in li ght of the afore-
mentioned difficulties. The results showed that the nature 
of the iron sulfide products was affected by the pH, tem-
perature, presence of oxidizing agent, and type of iron source. 
Tetragonal FeS was the mo st commonly produced iron 
s ulfide. It was produced at pH 7-8 when reagent grade iron 
metal, ferrous sulfate solution and synthetic goe thit e were 
the reactants. Synthetic goethite of different crystac 1. 1?1 n1?  ti.e s 
> 
13 
was us ed to approximate natural limonite. Again g oe thit e 
produced a more crystalline FeS. 
Pyrite, marcasite and elemental sulfur were found 
only when an oxidizing agent such as air or ferric iron was 
present. This should not be unexpected, since an oxidizing 
agent would be needed to oxidize sulfide sulfur to e l emental 
sulfur. Elemental sulfur reacted rapidly with dissolved 
sulfide to form polysulfides (Teder, 1971). The sulfur in 
pyrite and marcasite is in a mixed oxidation state; both 
the zero and the plus two s tates are present. s2 is a poly-
sulfide and hence would require the presence of elemental 
sulfur to form. 
Results (Berner, 1964aj also indicate that the type 
of iron source material is important in controlling the 
nature of the product sulfide. The use of dissolved ferrous .,,
, 
,  
?l 
iron promoted rapid reaction and the formation of an amorph - .,. 
(, .,  
:1 
ous precipitate. The use of iron metal resulted in a slower ::J 
,,, 
:j 
reaction and a more crystalline product. :, 
?I 
Unfortunately, there are several experimental factors ! 
which make the results of questionable applicability to 
sedimentary environments. In many runs, the temperatur e ex-
ceeded 35?c and the pH wa s either 4 or 9. These concli tions 
lie outside the limits of temperature (<30?c) and pH values 
(6.9 to 8.3) of reducing marine and estuarine sediments. 
Also, most reactions were conducted in unbuffered solutions? 
' 
the recorded pH values were the final pH values. Ther e was 
no indication of the range of pH va lues over which the 
14 
solutions might have varied. In some experimental runs, a 
small air space was left i n the reaction vess e l which would 
have allowed volatilization of hydrogen s ulfide from the 
solution phase into the gas phase . This would have bee n 
particularly important in reactions below pH 7 where molec-
ular hydrogen sulfide is the stable species. More import-
antl? the presence of air resulted in the oxidation of 
sulfide to elemental sulfur and the subsequent formation of 
polysulfides. Air would not be present in reducing sedi-
ments since oxygen is a poison to s ulfate reducing bacteria. 
The goal of this study was to det ermine the important con -
ditions in the formation of iron sulfides and the identifi-
cation of those products. No attempts were made to d e termine 
the kinetics and the mechanisms of the reaction. ..~ 
?i 
Roberts e t al. (1969) also s tressed th e importance ?,I  
I
of elemental sulfur in their study of pyrite formation. ' , ( 
,,I ..  
I 
Sulfur was produced in situ by the oxidation of hydro ge n ,,j 
,,, 
sulfide by ferric iron. Pyrite was synthesized by the re- ?:'l .,,  
action of goethite and molecular hydro ge n sulfide at zsoc. I I 
The results indicated two mechani s ms for the formation of 
pyrite: (1) the reaction of ferrous iron with the disulfide 
ion cs=) and (2) the s ulfidi za tion of FeS with e l emental 
2 
s ulfur . The latter reaction wa s much s lower than the former. 
Several factors cast doubt upon their conclusions and the 
applicability to sedimentary environment s . The pr esence 
of oxygen could have changed the reaction mec hani s ms . This 
is exemplified by the difference in the reaction product s 
15 
that were observed in the reactions where all conditions 
except the addition or exclusion were the same. Little 
control was exercised over the pH of the reactions . Only 
a few pH values were listed. A great deal of emphasis was 
placed on the existence and importance of the disulfide ion. 
Studies of the equilibrium distribution of polysulfides 
(Schwarzenbach and Fischer, 1960; Teder, 1971; Giggenbach 
' 
19 7 2) showed that the disulfide ion would be stab 1 e only 
at extremely high pH levels. 
Rickard (1969a) used a "qualitative semi-kinetic 
approach'' to determine the mechanism of formation of the 
various iron sulfides from aqueous solutions at low tempera-
ture and pressure . This information was then used to define 
the physico-chemical conditions necessary for the formation 
of each iron sulfide. Results indicated that the mineral s .,  
1 
may be used as indicators of the conditions in the environ - : , I 
',,I  
I 
ment. However, this is subject to limitations since several ,,, 
,., 
conditions may permit the formation of a few iron s ulfides. :J '.,J  
The following iron phases were used as reactant s : l 
ferrous carbonate, ferrous sulfate and synthetic goethite. 
Reactant sulfide phases were sodium s ulfide, s odium poly-
sulfide, and sodium thiosulfate. Th e solutions were not 
bufferred and the pH values g iven were tho se measured at the 
end of the experiment. Previous work by Berner (1964a) 
showed that the pH of the s olution affects th e product 
iron sulfides. 
h 
16 
Reactio n s of ferrous carbonate with sodium sulfide 
s olutions at pH 6-10 produced smythite, with minor amounts 
of mackinawite. This was the only circumstance in which 
the rhombohedral Fe 3s4 wa s produced. At lower pH values, 
the ferrous carbonate dissolved and mackinawite was formed. 
Rickard thus concluded that the preexistence of siderit e 
(Feco ) was necessary for the formation of smythite. A 
3
comparison of the diffraction patterns of s mythite and 
siderite (Table 1) revealed a close similarity of the inter-
planar distances of the two minerals. This similarity 
could hav e caused the epitaxial growth of smythite on the 
surface of siderite particles. Natural occurences of 
smythite do not indicate a siderite precursor. 
Mackinawite was produced by the reaction of ferrous 
sulfate and sodium sulfide at pH 6.5 to 11.7 and also syn-
thetic goethite and sodium sulfide at pH 7.0 to 9.0. This 
mackinawite contained adsorbed or coprecipitated sulfide as 
shown by c h emical analysis. The Fe:S ratio was 1:1.1. 
When thi s materi al was dried a nd heated at 70 0 C, grei g ite 
was formed. At lower pH values, the reaction of ferrou s 
sulfate a nd sodi um sulfide produced greigite. This indicated 
to the author that there was a mackinawite to greigit e tran s -
formation which was pH dependent. pH value s of ano x ic sedi-
ments are such that greigite s hould not form. The u s e of 
lower pH values for the goethite reaction resulted in the 
formation of su l fur, rnarca s ite and pyrite. A check of the 
Eh-pl! diagram for the five s pe c ie s indicat e d tli at el e mental 
17 
Table 1 
List of cl-spacings (in A) for Smythite and Side rit e 
Siderite Smythite 
11. 6 
5.75 
3.59 3.82 
3.00, 2.96 
2.79 2.86, 2. 8 3, 2.75 
2. 56 2.56 
2.35 2.45, 2.29, 2. 26 
2.13 2.16 
1. 96 1.979 
1.897 
1.795 ;
?~t  
1. 734 1.732 ~~ 1?1 
1.687, 1. 6 7 2 j 
1.527 1.577, 1.546 
:?~,  
1.426 1.435, 1. 427 
'2 
1. 395 .: i 
1. 354 1.351 ., 
1.306 ..,   
1.281 1. 28 ,,,:  I 
1.258 1. 25 ' 
1.229 :l'  
1 . 1 5 , 1.10 ., 
1. 06 :,,:,J  
., 
'!  
Siderite: Data from ASTM Powd e r Diffraction File Card 
8-133. a 0 == 5.796 'A 
Smythite: Data from Erd, Evan s , a nd Richt e r (1957) 
a == 3.465 'A, c == 34.34 'A 
0 
;;>;::::, c ::se:c......::::-,;,.--..--..? . .._ _._ _________ _ 
18 
sulfur is a stable species at lower pH values. 
Pyrite, marcasite and sulfur were produced by the 
reaction of sodium polysulfide with ferrous sulfate, s yn-
thetic goethite and mackinawite a t sedimentary pH value s . 
This supported Rickard's major conclusion that the sulfur 
bearing phase was important in controlling the nature of 
the iron sulfide product. This was concluded because only 
ferrous iron was used as a reactant. This contention was 
made in spite of the fact that goethite was one of th e re-
actants. Ferric iron is present in goethite. 
The pyrite:marcasite product ratio was found to b e a 
function of pH. At pH 4.4, marcasite wa s a major product, 
but the ratio decreased as the pH increa s ed, until at pH 9.S 
no marcasite was observed . The variation in the product ..- ~~  
ratio was ascribed to the different mechanisms of f ormation ., ?1  
for the two minerals. Pyrite was believed to form by th e 1 
I 
"'I  
direct precipitation reaction betwe e n (1) dissolved ferrou s :! 
iron and polysulfide ions or (2) mackinawite and poly s ulfide. ?,;.;,  ,, 
j 
The formation of marcasite involved a s olid stat e ox idation I 
reaction between sulfur and a preex i s ting iron sulfide. 
This last reaction was slow at low t emperature and h e nce 
marcasite would not be found in recent ano x ic sediments. 
Pyrrhotite was not observed in any of the se 
experiments. 
These reactions were conducted with rigorou s ex-
clusion of air to prevent error s in int e rpretation du e to 
unknown side effects from air ox ida tion. How e v e r, no s u c h 
19 
precautions were taken with respect to the pH of t h e solu-
tions. In anoxic sediments, the pH of the interstitial 
water ranges from 6.9 to 8.3 (B e n Yaakov, 197 3 ) due to the 
buffer ing s ys tem s in the water. The se expe rim ent s by 
Rickard we re conducted in unbuffere d s olutions to pr e ve nt 
the possibility of reactions between th e precipitates and 
th e buffer components. Thus the pH of the solutions might 
have changed considerably, particularly in those ex per im ent s 
1n which goethite was used as a reactant. The ferr i c iro n 
1n goethite would have oxidized the hydro ge n s ulfide to 
either elemental sulfur or s ulfur oxyz nion s . Thi s would 
have produced consi<lerable hydro x ide ions as in th e following 
reaction 
.,,.,  
2 FeOOH + 3 HS-+ 2 FeS + s0 + 3 OH + HzO .. 
.,  
?, 
which would have changed th e pH of th e solution cons iderably. .I  
Thi s would have been import ant s inc e hi gh concentrations of ' ,,' 
iron (O. 9M) and s ulfur 'J (1. 851\1) would have produced .,'  
j,  
considerable amounts of hydroxyl ions. In hi s discussion, 
Rickard also not ed that the high concentrations of iron 
and sulfur re s ult ed in a hi gher rat e of precip itation for 
mackinawit e and pyrite than the rate of crysta lli zation. 
Th e re s ulting f ine gra ined s ul f id es could have affected the 
rates of tr a ns formation of macki nawi te to greigite and the 
mackinawit e to pyrit e transformation . It was s uggested 
that incons i s t en c i es in the res ult s co uld have been cau sed 
by the importance of grai n size on the reaction rates. 
'-.!!!!l!!l!i;::;::!.::::::===================- L.,..41"'"/~ _.~:., =j ~...,.._ - -~.-:::;---?-? - ??- ________. .:.:_ _ _:_____J 
20 
Thus concentrations that ar e more realistic should be used 
in order to make comparisons between laboratory e x periments 
and authigenic iron sulfide formation. 
The formation of iron sulfides is a biological ly 
mediated process insofar as the reactant hydrogen sulfide 
molecule or ion is produced by bacteria. Thus Rickard (196gb) 
conducted a series of experiments to determine if the mi-
crobial formation of iron sulfides differed from the abio-
genic reactions. Ferrous salts and synthetic goethite 
were biogenically sulfidized by the metabolic proce s ses of 
Desulfovibrio _desulfuricans at pH 6 - 8. Mackinawite was the 
initial iron sulfide produced from both iron reactants. 
After three months greigite was observed at pH values of 6 
and 7 in the experiments with ferrous salts. This wa s in 
agreement with the earlier observed mackinawite to greigite .,,,  
r 
transformation that was observed in the abiogenic study. 
Mackinawite was the predominant phase present at pH 8 aft e r :.l 
., 
:; 
nine months. ?.?,,  
l 
Marcasite and pyrite were observed at pH values 6 ( 
and 7 after three months, in the reactions where goethite 
was the reactant iron phase. Mackinawite persisted at pH 8 
for three months. After six months, pyrite and marcasit e 
were found at all pH values from 6 to 8; marcasite was the 
predominant phase at pH levels 6 and 7. This latter re s ult 
was in agreement with Rickard's earlier abiogenic result 
that showed that marcasite was formed at acid pH value s . 
No crystallographic difference s were detected?  b e t we en 
, - :""':7- r:s:-- --r~-.,..,,.....,..,.... _,..... ______ :_:__  _:__~ 
21 
the iron sulfides produced in the inorganic study and the 
microbial study. Pyrite framboids were not found in this 
study. This added credence to Berners (1969) statement 
that biological control was unnecessary for framboid formation. 
Rickard therefore concluded that there was no dif-
ference between the mechanisms of formation of biogenic and 
abiogenic iron sulfides. One important difference was the 
absence of rhombic sulfur as a product of the sulfidization 
of goethite in all but the initial products. The rapid 
disappearance of the sulfur was attributed to the formation 
of polysulfides initially, which then reacted with organic 
compounds to form organic s ulfur complexes. 
Although this study delin~ated the reactions and con -
ditions necessary for the formation of iron sulfides, rate ... ,~  
expressions and mechanisms for the reactions were not <level ope d . ...?   , 
Sweeney a nd Kaplan (1973) studi ed the formation of 
'I  
pyrite framboids, both in the laboratory and 1n nature. :.i 
., 
Freshly precipitated iron sulfides were reacted with excess ?: j ?:, , 
:; 
elemental sulfur in aqueous and anhydrous conditions at 60 0_ I 
so 0 c. X-ray diffraciton and scanning e l ectron microscopy 
were used to det ermine the mineralogy and the texture of the 
products. Framboids were produced only in the reactions 
with water. The mechanism for framboid formation suggested 
was: (l) precipitation of the iron mono s ulfide mackinawite, 
C2 ) reaction with elemental sulfur to form greigite, and 
(3) transformation of greigite spheres to pyrite frambo l? ds . 
The mackinawite to pyrite transformation was consistent with 
br 
22 
Rickard's (196 9aj work. However , no pH data wa s g ive n s o 
that direct comparison between the studies wa s not po s sible . 
The texture of sedimentary pyrite wa s both framboidal and 
non-framboidal, which indicated to the authors that two 
mechanisms for pyrite formation were involve d. 
Berner (1970) also studied the formation of pyrit e 
in the laboratory and in nature. Framboidal pyrite was syn -
thesized from O.l and 1 M solutions of sodium bisulfide , f e r-
rous sulfate, and elemental sulfur at 65?C, and pH conditions 
of 6.9 and 7.9. After two week s , pyrite wa s form e d in th e 
experiments where excess sulfur was us e d. No pyrit e wa s 
formed in those runs in which barely suffici e nt s ul f ur wa s 
added to form polysulfides. Pyrite framboid s we r e found on 
the surfac e of sulfur particles. Thus it wa s conc lud e d that 
sulfur served a s a nucleation surface for the formation ?,~,  ?, 
i 
of pyrite. .? ?,,  
Field studies by Berner (1970) of Connecticut co as tal 0 
?' 
sediments showed that pyrit e was formin g at the e x pen s e of ..: ;,  
??,I  
the iron monosulfides. The source of the e lemental s ul f ur I 
which was necessary for this r e action wa s not ide nti fi ed. 
In sediments overlain by a e robic bottom water s , s ulfur can 
b e formed in the aerobic zone by either the c hemi c al or bio-
logical ox idation of the hydr o ge n s ulfid e which di f fu ses 
into the aerobic zone . Howe v e r, in s edime nt s ov e rlain by 
anaerobic wat e rs such as the Black Se a, the proc ess f or th e 
formation of elemental s ulfur i s not known. Ye t, the f o r-
mation of pyrit e has b ee n ob se rv e d in these e nvironme nt s . 
23 
This transformation reaction of FeS to FeS 2 was predicted 
to be completed within several years at the concentration 
and temperature of the sediments (Berner, 1970). 
The first kinetic study of the ferric iron-sulfide 
system was conducted by Cheng (1972). The object of this 
study was to determine the affects of ferric ions on the 
oxygenation kinetics of reduced sulfur species in simulated 
natural water environments at pH values of 4, 7, and 10. 
Results showed that the reaction rate was directly propor-
tional to the iron concentration. In the presence and ab-
sence of air at pH 4, the ferric ion was reduced to the 
ferrous species. 
However, at higher pH values and in the presence of oxygen, 
the ferrous ion was not detected. This was probably clue to 
the rapid oxidation of ferrous to ferric ion at neutral and .. 
alkaline pH values (Stumm and Lee, 1961). One interesting ??:?,; ,,,  
result showed that the rate of the reaction increased direct- 'r  
ly with the ageing of the ferric solutions. This was attrib-
uted to the increased coordination of the ferric iron to 
hydroxyls during the longer ageing periods. 
No analysis of solids was reported in this study. 
The kinetics and mechanism of the sulfidation of 
goethite was studied by Rickard (l974) ? pH measurements 
were combined with mass balance and equilibrium relation-
ships to determine the rate of the reaction. A proposed 
24 
mechani s m consisted of a simple initial stage and a complex 
seconLlary stage. Th e rate express ion for the initial stage 
wa s s hown to be 
cl(FeS)/dt = k (1?-r +)
2 [S=] 3/ 2 (A) 
wh e re d(FeS)/dt was the rate of production of FeS in M- 1 sec-1, 
(H+) was the hydro ge n ion activity, [S= ] was the total s ulfi de 
-1 
ion concentration in M and (A) wa s the goethite surface 
area i? n cm 2 . T11e rate constant k was 1.5 x 10
7 11-1/2 1 -1/2 1? 
cm - 2 sec -1 
The black product was s hown to be identical wit h a 
mi xture of x- r ay amorphous ferrous sulfide a nd poo rly crys-
talized mackinawite. No pyrite or elemental s ulfur was ob-
served in the x -ray diffraction scans. The propo sed mec hanism 
included a dis s olution reaction in the first stage. The 
.;,, ,,  
::: 
seco nd order rate dependanc e in the rate expression on the 
,,,'   
hydro ge n ion ac tivity wa s interpreted as an indicator of a ' ? i 
,, 
dis s olution process. Thi s was concluded fr om a comp arison ::,?,;  
:; 
with a study of the opposing oxidation reaction of aqueo u s ( 
ferrou s iron (Stumm and Le e, 19 61) in which a second order 
rat e dependanc e was observed for hydroxy l ions . Lin e broad -
ening in the x-ray scans indicated to th e author a fi n e 
grained iron s ulfide product. Thi s was i nt e rpr eted as ad-
ditional evidence for a di ssolution reactio n. 
A dissolution step was proposed to pro duce ferric 
ions in s olution. 
brz 
25 
+3 
The stability field of ferric ion, Fe , (>750 mv, < pH 2 ; 
Hem, 1960) is incompatible with the reaction conditions u s ed 
by Rickard (1974; pH 7-9 and Eh <- 100 mv). Conditions in an-
oxic sediments, pH 6.9-8.3 and Eh <- 100 mv are also incom-
patible with the existence of dissolved ferric iron. Thus 
the dissolution step is more likely to be preceded by th e 
reduction of iron. Thermodynamically, ferrous ions are more 
stable then ferric ions at the pH and Eh conditions found 
1n anoxic sediments. 
However, Rickard proposed that th e s econd s t e p in 
the mechanism of iron sulfide formation involved the reduc -
tion of ferric iron by sulfide ion which was then followed 
by the precipitation of the iron monosulfide. 
Fe+ 3 
0 
+ 3/2 s= + FeS + 1/16 S8 
This work (Rickard, 1974) r e sulted in a rat e expre ss ion a nd 
rate constant. The proposed mechanism had insuffici e nt 
verification since the dissolved products, intermediates ' 
and elementary step were not identi f ied. The solid product s 
were studied and characteriz ed, but the s olution product s 
were not identified. Yet a study of the dissolved reaction 
products could give more detailed information about the 
mechanism and the rate of the reaction. Thi s i s true be-
cause the sulfide oxidation product s are ea s ier to identify 
than are the solid product s . Dissolved s pecies a re not af-
f ected by the problem of s mall grain s i ze or that o f poor 
crystallinity. In addition, th e c oncentrations of th e 
26 
d i ss olved species are easier to measure . 
Studies of the ox idation of s ulfide solutions by 
oxygen (Chen and Morris, 1972; Cline and Richards, 1969; 
O'Brien, 1974) have shown that severa l sulfur species were 
produced by the sulfide-oxygen reaction. Sulfate was the 
thermodynamically s table s ulfur ox idation produc t, but ele-
mental sulfur, polysulfides, thio sulfat e and s ul fite were 
also produced a nd persisted at sedim entary conditions for 
varying periods of time. Similar studies of the fe rric iron-
sulfide system are not available. 
The reaction order for each spec i es in the rate ex -
pr ess ion was de termined by maintaining the concentrations of 
all other species in the reaction constant except the one to 
be s tudi ed. The rate of the reaction wa s followed by the 
change in the pH. pH changes of 0.1 to 0 .4 pH unit s were .?.? 
observed. Yet hydro ge n ion wa s one of the species in the 
,,,,  
rate expression and as such s hould have been kept constant. j 
Thi s variation in the pH might have affected the reaction 
order of the other species. Mor e l ike ly, the rate con s t ant 
would have been altered. In addition, the experiment s were 
conducted at relative ly high total s ulfide concentrations 
(0.05-0.5 M). This i s 10 to 100 times th e max imum concentra-
tion of total s ulfide that could be expected in nat ure . 
Thi s could conceivably have altered the rate of the reactio n 
by changing the texture of the product. Ri ckard noted t hi s 
in an earlier study (1969a). Future s tudies s hould be con-
Juct e <l at mor e realistic or natura l s ul fi de values ( <0 . 005 M). 
27 
The complex secondary stage of the sulfidation of 
goethite, the formation of pyrite, was studied by Rickard 
(1975). Iron monosulfides were reacted with elemental sulfur 
in aqueous sulfide solutions to determine the kinetics of 
pyrite formation. The rate expression for this reaction was 
determined to be 
where d(FeS )/dt was the rate of pyrite formation in M-1 1 -1 
2
sec-l k was the rate constant, (FeS) wa s the surface area 
of FeS in cm 2 , (S) was the surface area of elemental sulfur 
in cm2, (ES=) was the sum of activities of the dissolved 
sulfide species, and (H+) was the hydrogen ion activity. 
., , 
,1..: 
The rate constant varied from 1 x 10- 12 (40?C) to 3 x 10 -14 
...~ ~  
6 -1 -1 -1 
?.,".  
(5?C) cm mole 1 sec ?: ,.. .  
A proposed mechanism included (1) di ss olution of both ,, 
? , 
sulfur and ferrous sulfide, (2) formation of polys ulfide ions. :: l 
,, 
and (3) precipitation reaction between aqueous ferrous ions :1'i.,
 
,  
and polysulfide ions. The pyrite produced wa s non-framboidal i 
although the reactant sulfur was classified as framboidal by 
the author. The reactant iron s ulfide particle s were spher i -
cal with a mean diameter of 0.02?. 
These studies have contributed grea tly to the under-
standing of the formation of iron sulfides. However, it is 
difficult to correlate the results because of the wide range 
of experimental conditions that wer e used by the various re-
serarchers. Different temperatures, pH values, reactant 
28 
concentrations, iron phases and sulfide reactants were used. 
Yet by limiting the range of some of the conditions (tempera-
ture to 20?-2s 0 c, and pH to 6-8) some agreement can be achiev-
ed. Results fr om similar experimental conditions are listed 
in Table 2. 
Several general statements can be made about these 
results: 
(1) mackinawite was the initial iron sulfide formed 
(2) greigite formed at acid pH values only 
(3) mackinawite transformed to greigite in acid pH 
conditions 
(4) marcasite and pyrite ne eded elemental s ulfur to 
form 
(5) mackinawite altered to pyrite when elemental 
sulfur was present. 
Other results not listed in the table showed that greigite 
transformed to pyrite. The texture of the pyrite was af-
fected by the presence of greigite as an intermediate. No 
adequate exp lanations have been advanced to consistently 
explain the nature of the products. 
The product distribution can only be explained by 
determining the elemental chemical steps of each reaction. 
In order to accomplish this, the dissolved products, inter-
mediates, and time-concentration curves of each species 
must be determined, in addition to characterizing the 
mineralogy of the solid iron sulfid es . 
-
29 
Table 2 
Summary of Reported Sulfide Product s 
Iron and 
sulfur pl-I 
reactant 
species <6 6 . 5 7 8 > 8 
Fe+ 2 s= 
' 
M 
initial G M M M 
3 month s G G G,M M 
M 
M 
6 months G G G M 
s= ,I Goethit e, .P 
,, : 
M .. , 
initial S,MS,P M M M :i ',.  
3 months MS,P MS,P,M MS,P,M M 
6 months MS,P MS,P MS,P P,MS .? 
' 
Fe+ 2 s= P,MS,S P,MS ,S P,MS ,S P,MS ,S 
' X 
s s s s s 
...,.  
Goethite, sx j ' 
M, s= P,MS X 
Fe concentration: 0.05-1 M 
S concentration range: 0.05 - 1.8 M 
G grei gite MS = marca s it e 
M = mackinawite s = s ulfur 
p = pyrite s: = poly s ulfides A 
30 
EXPERIMENTAL 
A mechanism of iron sulfide formation requires the 
identification of the solid and dissolved product s of the 
reaction between ferric oxyhydroxides and aqueous bi s u l fide 
ion. In order to accomplish this, several s eri es of experi-
ments were conduct ed. Goethite and poorly crystalline 
goethite were reacted with aqueo us bisulfide ion (0.002-0.09 M) 
to determine the kinetics of tl1e initial sulfide ox idatio n 
products under the fo llowing conditions : pH 7.0-8.0, 
Fetotal = 0.8-5.6 x 10 -3 , JJ = 0.295 (salinity 24-25 o / ), 
00
temperature= 25 .0 ?0.2?C. These values are r e pre sentative 
of typical estuarin e concli tions. 
A. Preparat ion of Reactant Iron Min e ral s 
''  
1. Coethi te ,..i  
Syn thetic goethite (a-F e OOH) wa s pr epar e d by th e l ' 
alkali n e hydrol ysis of ferric nitrate s olutions (Atkins on 
et a l., 1968). Ferric nitrate nonohydrate wa s di ss olved in 
distilled water fol low ed by the additio n of 2 . 5 M NaOH to 
g ive a hy dro x id e :iron ratio between zero a nd two. The s olu-
tion was n e x t hydro l yzed at room t emperatur e for 50 hour s . 
Conc entrated NaOH wa s added until the pH wa s g reater than 
10.6. The red brown s u spens ion was a ge d in polyethyl e n e 
bottles for 80 hours at 60?C. Aft e r a geing , th e sampl es 
were centrifuged at 2100 RPM for 52 minut es and wa s h e d with 
31 
distilled water until a n ega tive reaction with AgN0 3 was 
obtained. Samples were suspended in di sti lled water and 
store d 1n po l yet hyle n e bottles. 
The size of the particle s wa s determined by tran s mi s-
si on electron microscopy. Hi g h initial hydro x i d e :iron ratios 
produced s mall lath-like particle s (630xl60 A) while larger 
partic l es (5100x840 A) were produced by low hydroxide:iron 
ratios. The surface areas of these particles were found to 
be 66 and 12 m2/g, respectively. Both sample s gave electron 
and x-ray diffractio n patt ern s corresponding to goethit e, 
alt hou gh not all reflections were observed for both sampl es. 
J 
, r' 
The amount of FeOOH in s u s pens ion was determined by drying ;/ 
~' ,I 
:;: 
aliquots of the suspension at 90?-l00?C for 24 hour s. 
2. Amorphous Fe(OH) 3 
Poorly crystalline Fe(OH) 3 was prepared by dropwise ' ~ ,;' 
' 
addition of 2 M KOH to 2M FeC1 3 with constant stirr ing, un - ' ' = 
) 
til pH 7 was reached (Landa and Gast, 1973). The pH was 
monitor ed by a glass electrode r eferred agai nst a s ingle 
junction reference electrode. The brown-red pr ecipitate 
wa s centrifuged and washed with distilled wat er until a 
n egative chlor ide te s t was achieved. X-ray diffraction pat -
tern s (Cr Ka radiation) showed no peaks. Electron diffrac-
tion patterns s howed two broad, weak lines which were attribu-
ted to (021) and (002) reflections of goethite. Th e particle 
s ize wa s 150 A, and the surface area 13 3 m2/g. 
32 
B. Electrode Measur ements 
All pH measurements wer e made with a Ag -AgCl low 
sodium error glass electrode. This wa s referenced a gains t 
a double junction reference electrode with a 10 % KNo 3 outer 
filling solution, or a single junction reference el e ctrode 
with a saturated KCl-agar salt bridge. The glas s el e ctrode 
was calibrated and the slope was checked prior to each mea-
surement with pH 7 and pH 10 buffers. The experimental slope s 
ranged from 95% to 100% of the theoretical slope of 59.16 mv 
(at 25?c). The Eh was measured with a platinum bill e t e lec -
trode referenced against a doubl e junction reference el e ctrode . 
The Pt electrode was calibrated with fresh Zobell solution 
at +0.430 mv (Parks, 1968). 
The pS was measured with an Orion Ag -AgS el ec trod e . 
All measurements were made with an Orion 1 701 di gital pH I:~'' ?~  ,i? 
meter and a #605 el e ctrode switching box . ' " 
,. ,t 
C. Reagent s ::i 
1:1 
I I' 
: :1 
,:' 
All solutions were prepared from reagent grad e chemi-
cals. The solutions were prepared with distill e d water ex -
cept the sulfide stock solutions. The s e wer e pr e par e d with 
distilled water which was deoxygenat ed by bubbling with nitro-
gen gas for 8-12 hours. Stock sulfide s olutions we r e pr e-
pared by flushing a glass stopper ed fla s k with Nz f or several 
minutes. The crys talline sodium sul f ide monohydr a t e was 
rinsed with distilled deoxygena ted wat e r, to r emov e ox 1? o1. e 
coatings, and wip ed dry with Kimwip es . Th e necess ary we i ght 
33 
was dissolved in distilled, deoxygenated water , made to 
volume, and used within one hour of preparation. Tests for 
polysulfide s , thiosulfate, and sulfite of the s tock s ulfide 
solution were negative. 
D. Dissolved Reaction Products 
1. Reaction vessel 
The studies of the oxidation products were conducted 
1n a 0.5 liter plexiglas cylinder (Figure 2). (Th e int e rior 
2
surface area with the piston in, i s 426 cm .) The cylinder 
was sealed at top and bottom with plexiglas sheeting. Th e 
top had holes for a thermometer, gas bubbling tube, two 
sampling ports, and a glass pisto11. The piston permitted 
,,? 
the removal of a sample without the introduction of a gas 
phase to the chamber. 
The plexiglas bubbling and sampling tubes were sea l e d 
,:,1? ,.  
to the top lid with methylene chloride. The g lass thermo- .:.'   
meter and piston barrel were sealed to the lid with General ::; 1,:,1  
: :I 
Electric silicone cement. A gastight seal wa s mad e betwee n r 
the piston and the barrel with silicone stopcock grease . 
The reaction vessel was maintained a t 25.0?0.zoc by 
means of a thermostated water jacke t. All solutions wer e 
stirred with a glass stirring bar at constant rate for th e 
duration of the run. (Private communicati on with R. Berner 
indicated that sulfide could diffuse through Teflon and 
Tef lon coa t ed s t irr ing? b ) 
react with the ma gnet of a ar . 
The top of the reaction vesse l wa s co vered with pl ast i c 
34 
Electrode 
pH Meter ' ' 
Thermometer 
'V 
pH Electrode > 
I 
Plastic , 
Glove --) ? I I 
Bag :, Nz ; Water . --...,.. -- Out 
Water~ 0 
In 
, , 
,,, 
' " 
1:: 
I? 
:J 
? I..'  
,,,,., , ..  
: ~ 
:i 
,:1 
Stirring 
Plat e 
Figure 2 Schematic Diagram of Re a c tion Ve sse l 
35 
g lov e ba gs and flus hed with n itrogen . Thi s provided an 
iner t atmosp he r e for sampling . 
2 . Reaction medium 
The reaction was conducted in a buffer r ed sa line 
s olution with a n ionic strength of 0 . 295?0.005. Sodium 
c hloride was u se d to ad just the ionic strengt h. (The sal i-
nity vari ed fro m 23 . 8 ?/ 00 at pH 7. 45 to 24.9 ?/ 00 at pH 
8.55.) Palit zsc h bor ax -bo rate buffer (H a rvey, 1955) was 
used to maint a in essenti a lly constant pH fro m 7.45-8.55. 
Th e buffer capacity of th e medi um was determined experi ment -
al l y, and found to r a nge fro m 1.50 mM/0.1 pH unit s at pH .,.,?.  
,,.::,  
7 . 6 O to 4 . O8  mM / 0 . 1 p I-I uni t s at p I-! 8 . 5 0 . More ex t en s i v e ,, .~? ;,i 
:..1  
data on th e medium i s li sted 1n Table 3 . 
3. Proc edure s 
The re action vesse l was wa s he d with dilute nitri c 
ac i d , rinsed with disti ll ed water a nd dried. Appropr i ate 
volumes o f buffer , so dium c hlo ride s olution, di s till e d water, 
:1 
a nd ace tic acid were pipetted into t h e reaction ve ss e l a nd .,' 
bubbled with puri f ied nitro ge n f or 8 - 1 2 hour s prior to 
addition of s tock s ul fide so lution. Next, s ul fide s o luti on 
was a dded to gi ve the desired total s ul fide co ncentration 
(1-5 x 10 -3 M) . The s= , pol ysu lfide, t h ios ul fate and s ul fit e 
co ncentrations and pH were determined. (Th e s ulfit e te s t 
was l a t er di sc ontinu e d du e to ear l y negative re s ult s .) 
Af t er 30 - 6 0 minut es , an aq ueou s s u s pe ns ion of a r e actant 
iron s p ec i es wa s ad ded and mi xe d for 5 - 10 mi nut es . Th e 
36 
Table 3 
Palitzsch Borax - Borate Buffers 
Borax solution= 19.108 g Na2B407 ? 10 H2o / 1 
Boric acid solution= 12.404 g H3Bo 3 + 2.925 g NaCl/ 1 
ionic 
ml ml strength 
@ 
Borax solution Boric acid s olution pH 25?C jJ 
50 50 8.48 0.075 
45 55 8.38 0.072 
40 60 8.28 0.0 70 
35 65 8.17 0.067 
30 70 8.05 0.065 
25 75 7.92 0.063 
,. 
20 80 7.76 0.060 
.. 
1, 
:1 
I' 
15 85 7. 58 
,, 
0.05 7 
10 90 7.34 0.055 ,,, 
:i 
,:,1  
-
37 
first sample was taken after 15-30 minutes and at regular 
intervals for generally 24 hours. 
The sample was extruded into a syringe barrel. The 
I-I was measured and total sulfide was determineu1. ? Ne xt, 
P
0.2 ml was filtered through a 0.2? Nuclepore filter, and 
the solid was analyzed for precipitated sulfide. (Precipi-
tated sulfide refers to adsorbed or reactant sulfide.) Ap-
proximately 3 ml was removed, placed in UV cuvettes and 
centrifuged at 2100 RPM for 15 minutes. This procedure re-
moved particles larger than 1000 A. The UV spectra of the 
centrifuged solutions were recorded form 470-230 nm. In 
??' 
several runs, the UV spectrum was recorded after filterina .:,i .;.ib .  
1:..1'. 
?1 ?1 
through the 0.1? filter, but it was found that the filtering ;J. ?'.I 
? ' 
process resulted in loss of solution sulfide. Solution 
sulfide refers to dissolved sulfide and sulfide from?  par t?1 c 1 es "",,,   
,;J,  
smal l er than 1000 A. The supernatant liquid was sampled for ,, 
:? 
solution sulfide. Approximately 3-6 ml of suspension was ~ 
filtered through a 0.1? filter. 0.1 M Znc1 2 was added to 
the filtrate to precipitates= as ZnS (0.1 ml Znc1 2 ; ml 
sample). The filtered solid was refluxed for 30 minutes 
with 25 ml of acetone to extract elemental sulfur. This 
suspension was set aside, shaken after 30 minutes and fil-
tered throuao h a o.2? filter. The filtrate was analyzed f: or 
thiosulfate, sulfite, and dissolved iron. A flow chart of 
the analytical procedure is shown in Figure 3. 
38 
Figure 3 
Flow Chart of Analytical Procedures 
0.2 ml filtered ( pH measurement ) 0 .1 ml analyzed for s= through O. 2 ? T 
filter and 
ex tracted for s=p  I 1 \ 
3 ml centrifuged for 'v5 ml filtered 
15 min @ 2100 RPM through O. 1 ? 
t / filter 
UV spectra from 
470 - 260 nm for Sx filtrate treated filter extracted with 0.1 ml with 25 ml acetone 
t ZnC 1 2 / 1.0 ml for 30 min for s 0 
0.2 ml of filtrate 
s upernatant 
for Sc t resulting suspe nsio n 
filtered through 0.2? 
filter after 30 min 
J t 
., 
ST = total s ulfide= solution and acid extractable s olid " J 
s ulfide 
SC so lution sulfide 
S~ = acid extractable so lid s ul fide 
S 0 = ace tone extractab l e elemental s ulfur 
Sz03 = thio s ul fate ion 
= dissolved iron ( <O.l?) 
before addition of ZnClz 
FeBz 
FeAZ = dissolved iron 
( <O.l?) after addition of ZnClz 
s=X polysulfide, 
s= a nd s=5 
 = 4  
39 
4. Electron microscopy and electron spectro s copy 
In several runs the solids were studied by transmis-
s ion electron microscopy (200 KV Hitachi) to note change s 
in morphology and mineralogy. Here, the reaction suspension 
was centrifuge<l at 2100 RPM for 15 minutes. The s up e rnatant 
liquid was decanted and the solid was resuspended in di s til -
led deoxygenated water. This suspension was spotted on cop-
per electron microscope grids, with a collodion s ubstrate. 
After drying in a nitrogen atmosphere , the s ample wa s analyz-
ed by transmission electron micro s copy. Both micro graph s 
and diffraction patterns were recorded. The diffraction 
patterns were calibrated with tha llium chloride or a luminum. 
Several samples were also analyzed by ESCA (El ec tron 
' ,'i  
Spectroscopy for Chemical Analysis). Electron s pectro s copy I~:, 
...: ,~ ,l  
(Siegbahn, 1967) is a technique for measuring the bindin ba  ,..  ,..  
ener gy of ejected electrons after interaction o f the sampl e :i 
,, 
,, 
with high energy photons or x -rays . The values of the bind- ?:1 
,,,.,  
ing energy can be calculated from th e followin g equation: ' I 
.:,1,  
,:1,  
where Eb is the electron binding ene r gy, Ei i s tl1c e nergy of 
the exciting source, ?sp is the work function in the vie?1 n1. ty 
of the spectrometer, and Er i s the r ecoil ene rgy. Th e work 
function and the recoil ener gy are generally neg l ec t ed for 
li ght element s , and the binding ene r gy can be calculat e d 
from the knowledge of the exciting s ource a nd kine ti c en e r gy. 
The binding ener gy i s a function of th e atomi c numb e r of tl1 e 
b 
40 
e l ement, and the chemical environment. Thus information 
a bout the structur e of the molecule a nd the oxidatio n s tate 
of th e e l ement can be determined. ESCA measur es these prop -
er tie s of th e top 100 A only, a nd th erefore the s urface 
1nust be representative of the bul k of the sample. 
The s uspensio ns were filtered through a 0.2? filter 
under a nitroge n atmo spher e. After drying , the sample was 
transferred with doubl e sided scotch tap e to an a luminum 
sample hold er . The Fe 3p a nd S 2p spectra we r e meas ured on 
a Varian induced e l ec tron emi ssion e l ec tron spectromete r. 
The spectra were calibrated wit h a Au s t andard at 83 .4 eV 
., 
and s mooth ed by a standard computer pro g ran1 u s ing five point s. !:: 
: ': 
1'1 
I 
?'.I 
5. Analytical procedures 
Concentrations of di ss olved s pecies were de t ermin ed ',I 
,.. 
colorimetrically. All measurements were made on a Cary 15 
UV-visible spectrophotometer wit h a 1 cm quart z cell. Di s-
" 
,/ 
till ed water wa s use d in the reference beam. 
:1 
:1 
All me thods were checke d to ins ur e that there we r e ,, :1 
?' 
no interferenc es from reactant, buffe r, or other product . 
Procedures were modi fied when n ecessary , du e to s mall sample 
volume. Otherwi se , a ll methods were used as s tat e d in the 
referenced works in the sect i ons t hat f o I l ow. Necessary 
modi f ications , detailed pro cedures, a nd calibratio n c u rves 
for the different s pecies fo ll ow . 
41 
a. Sulfide sulfur 
Total, s olution and precipitate d s ulfide were de-
t e rmined by r eaction with N, N- dime thyl-p - ph enyl ened i ami ne 
s ulfate (Budd and Bewick, 1952). 0.1 ml of total or so lu-
tion s ample wa s fi xed in 5 ml of 2% zinc acet a t e . Pr ec ipi -
tated sulfide was measured after f iltration o f 0. 2 ml o f 
sample through a 0.2? Nuclepore filt er. The f ilt e r and 
pr ecipitat e were transferred to zinc a ceta te s olution. 
0.5 ml of dilute amine solution and 0.1 ml of s atura t e d 
f e rric chlo r ide were added, mi xe d and made up to 25 ml. 
Ab s orb ance was measured a t 670 nfil The mean r e l a tive s t a nd -
a rd deviation of the method wa s 4.5 % for tripli cat e a na ly ses , 
-5 
while the detection limit wa s 1 x 10 M. A repr esent a tive 
,,'  
calibration curve i s shown in Fi gur e . ,? 4. 
.,. ..  
0
b. El emental s ulfur (S p ? ) ,I 
I' 
A known volume of reaction s u s pe ns ion was f ilt ere d ,. 
?" 
,I 
through a o.1? Nuclepor e filt er. The precipit a t e wa s ex -
:1 
:1 
tracted with 25 ml of acetone by reflux ing for 30 minut es 
(Bartlett and Skoog, 1954). Te s t s s howed tha t thi s volume 
of acetone and reflux ing time was effective in di ss olving 
over 95% of the sulfur present. Py r it e was not ex t racte d 
under the s e conditions. 
After cooling to room temp eratur e , 2 ml of s ampl e 
were added to 5 ml of 0.1 % s odium cyanid e s olution i n a ce-
tone s olvent (5 % water) and se t asi de for two minu tes . Th is 
ac e tone s olvent wa s a dded to ma ke 10 ml a nd mi xe d. 5 ml 
of the r esulting s olution was a dd e d t o an e qua l vo lume of 
-
42 
0 . 5 
~ 0. 3 
~ 
cu 
..,D_.  
0 
l/l 
..D 
<(, 
0 . 2 
0 . 1 
0.0 .__ ____ ._ _____. .__ ____ ..__ ____. ,__ ___- J,._ _ _J. 
0 1. 0 0 2 .00 3 . 0 0 4 . 00 5 .00 
S = ( mM) 
Fi g u re 4 : Samp l e Sul fi d e Ca l i b ratio n Cu rve 
43 
of sa turat ed ferric chloride so lution (acetone s olvent) 1n 
opaque bottle s. The absorbance was measured at 465 nm . Th e 
detec tion limit wa s 1 ppm S0 or 1.5 x 10 - 4 M S0 The mean 
rel a tive s tandard deviation for dupli ca t e analyses was 2.7%. 
The calibration curves wer e pr epared from weighed amoun t s 
of reagent grade elemental sulfur and were linear over th e 
range 1 - 50 ppm (Figure 5). Thi s volume wa s relat ed to the 
original volume through the f ollowing equation: 
- 3 
ppm S0 x 25 ml x 1 x 10 
Ms  = 
ml sample x 32 g mol e- 1 
c. Thio s ulfate ion (S203) 
1.0 ml of sample was mi xe d with 1.5 ml of deoxyge nat ed 
distilled water and 0.5 ml of 1% sod ium cyanide s olution 1n 
as opaque bottl e (Urban, 1961). 0.3 ml of 0.1 M cupri c 
chloride solution wa s added with cons tant swirling to pre -
vent formation of an insolubl e pr ec ipitate. 0.5 ml of fer-
ric nitrate-nitric acid r eagent wa s a l s o added with constant 
swirling . 1.0 ml of deoxygenated di s tilled wat er wa s added 
and mixed. The absorbance of the so lution was meas ure d 
within one hou r at 460 nm. 
Te s ts s howed that s ul f ide gave a po si tive i11t erfe r e nce 
te s t and had to be r emove<l prior to tl1 e thio s ulfat c te s t. 
Two me thod s were tested to remove aq ueo us s ulfide f rom th e 
filtered s ampl es . Th e sampl e wa s acidified with 0.5 ml of 
1 M acetic acid per 5 ml of sampl e and the n bubbl ed with 
nitrogen gas f or thi rty minut es . /\ sec ond met hod rcqui red 
44 
1. 0 
,......., 
E:: 
~ 
0 . 8 
en 
\() 
tj-
'-" 
(1) 
u 
~ 
co 
,D 0 . 6 
H 
0 
(/) 
,D 
c:i:: 
0 . 4 
0 . 2 
0 . 0 L-___ _ _.__ ____ L._ _____. __ ____ ,__ ____ , 
50 
20 3 0 
40 
0 10 
S 0 (ppm) 
Samp l e S? Ca librat i on Curve 
Fi gur e 5 
45 
the addit io n of 0.5 ml of saturated zi nc chloride solution 
per 5 ml of sampl e, followed by f iltrat ion of the precipi-
tate through a 0.2? filter after 30 minut es . Both gave 
s imilar results, but th e l atter was c ho sen for simplic ity. 
The samples were ca librated against sodi um thiosulfate 
s olution (F i gure 6). The detection limit was l x 10-5 M 
S 20~, and th e mean relative standard deviation was 3.3%. 
cl. Sulfite ion (SO:D 
Sulfit e ion concentration wa s meas ur ed after treatment 
with zinc chloride (We s t and Gacke, 1956). One ml of sample 
wa s added to 5 ml of 0.1 M sodi um tetrachloromercura te. 
I :' :  
0 . 5 ml of 0.04 % hydrochlori c acid - bl eac he d p-ro s aniline was i I 
'I 
added and followed by 0.5 ml of 0.2% formaldehyde so lution. ! 
, I 
The absorbance was measured at 565 nm between 15 a nd 35 ?J 
minut es after addition of the formaldehyde sol u tion. The 
a bs orbanc e was fo und to increase up to 15 to 20 mi nu te s , 
a nd s lowly decrease after that time. 
?,''  
The concentration was determine d fro m a ca libratio n 
c urve (Figure 7) which was prepared from sodi um s ul fite 
s olutions. The detec tion limit was 1 x 10- 5 M so;. 
e. Iron 
Di ss olved iron ( <0.1 ? particle s ) wa s determined for 
tJ1e filtered sampl e (Charlot, 1964). 0.5 ml of s a mpl e wa s 
mi xe d with 0.2 ml of 0.1 M acetic acid, 0.5 ml of JO% hydro xyl -
annn e hydro ch lorid e s olution and O. 5 ml of O. 5% 1, 10 - ph e nan -
throline hydr oc hl oride. The s olution wa s mi x ed and se t as ide 
46 
0.5 
0.4 
,---, 
f::; 
~ 
C) 
't?j- 
'---' 
(l) 0.3 
u 
~ 
cu 
..D 
0'""  
Cf) 
..D 
<-r: 0.2 
0 .1 
5 
o.o 4 
0 l 
2 3 
s o 4(M x 10 ) 
2 3 
Samp l e Thio s ulfat e Calibratio n Curve 
Figure 6 
47 
0.6 ? 
0. 5 
,.--, 
8 ? 
i=:! 0.4 
t.n 
\() 
t.n 
'--' 
. I 
? I 
Q) 
u I 
i=:! 
ell 0. 3 
.a : I ??I 
H 
0 ,1.1 ? 1 
U) ,, . 
.a 
<t; 
0.2 ,/ 
? 
0.1 ? 
0.0 2 3 4 5 
0 1 
so= (M 10 4X ) 3 
Curve 
Figure 7 Sampl e 
S03 Calibration 
48 
for one hour, then made up to 10 ml. The a b sorban ce was 
meas u red at 510 nm. The calibration c urv e was linear from 
1-50 ppm a nd had a detection limit of 1 ppm or 1.7 9 x 10 - 5 M. 
Tests with dissolved iron stan dards were not affected 
by the zi n c chloride treatment to r emov e disso l ved s ul fide . 
11owever, several time s during the reaction r un s, the iron 
test was conduct e d with and without th e zi n c c hlorid e treat-
me nt. The untr ea ted sampl es alway s showed hi g her c onc e ntra -
tions of iron. This would indicat e that th e pr ec ipi tat ion 
of zi n c s ul fide caused a coprecipit a tion or ad s orption of 
iron from the so lu t ion. (Calibration c urv e in Fi g ure 8.) 
f. Hy drogen ion (I-I+) 
The reaction of iron ox ide s , hydro x ide s a nd oxyhydrox-
ide s result s in th e formation of hydro x ide ions, according to 
the followin g reactions: 
0 
Fe 2o3 + 3 I-IS + 2 FeS + S + 3 OH 
2 0 Fe(OH) + 3 HS- + 2 Fe S + S + 3 OH + 3 II zO 
3 
0 
2 FeOOH + 3 HS + 2 FeS + S + 3 OH + I-1 20 
A pH c hange will b e ob se rved as a result during the reaction. 
The amount of OH produced can be d etermi n e d fro m t h e pl! 
c hange and the buffer capacity of the sys t en1. The int eg r a l 
of the buff e r c a pacity S, from pHini tial to p ll final will 
yie ld th e titration c urve for th e buffer . Th e titration 
curve for the borax-bo r ate buff e r (Perrin et a l. , 1974) wa s 
d e termine d experime ntally by titrating a so lu tion of t h e 
49 
0.5 
0.4 
"~i=  ; 
~ 
0 
.-1 
en 
'--' 
Q) 0.3 
u 
~ 
ro 
...0 
1--< 
0 
(/) 
..D 
<t: 0.2 
0 .1 
0.0 L------..L------&...------'------1-------'--1 
0 10 20 30 40 50 
Fe (pp m) 
Figure 8 Sampl e Fe Ca libration Curve 
so 
buffer, salt solution and distilled water (in the propo r t?i ons 
used in the reaction runs) with a strong base. The ti tra -
t i on curve for the region 7.2 to 8.0 is shown in Figure 9 . 
The equation for the line is 
2 
Y = 376 - 106. 7 X + 7.76 x
where y is the number of millimoles per liter of added 
In addition to the borate 
hydroxide, and xis the pH. 
buffer, the reaction solution is buffered by the H 2S-Hs-
this weak acid 15 6.75 at the 
system, since the pK1 for 
reaction system (Goldhaber and Kaplan, 
ionic strength of the 
capacity however, is small due to the 
1975). The buffering 
low total sulfide concentrations used in this series of 
experiments (2-9 x 10-3 M). The buffering by this couple 
I 
"'I 
is based on the fact that the reacting species in the equa- ,I,! ' ' 
tions HS - , the predominant species at pH greater than 
15 
6.75. As a result, a change in the total sulfide concentra- . :I 
tion or a pH change will result in the release of the excess 
hydrogen ion which can then react with excess hydroxide io ns. 
The amount of excess hydrogen ion that is released can be 
determined by the following expression 
51 
5 . 0 
4 .0 
,---, ,L. s 
'-' 3. 0 
I 
:I: 
0 
2. 0 
1. 0 
o.o 7.4 7 . 6 7.8 
7 . 0 7 . 2 
p l! 
Fi gure 9 : Titratio n Curve for Borax- Borat e Buffer 
2 
[OH- ] = 367 - 106.7 pH+ 7.76 pH
(mM) 
52 
where K and K are the first and second di ss ociation con-
1 2 
stants for hydro ge n s ul fide. At the pH conditions use d in 
this study, the denomina tor simpli f ie s because the third 
t e rm is small and the equation reduces to 
The buffer capacity of the system i s thus the s um of the 
two buffers (Perrin et al., 1974). 
g . Polysulfides 
The concentrations of the polysulfide ions, s4 and 
' ' 
' ' 
s , were determined from hydro ge n ion, tot a l sulfide a nd :1 
5 , I , I 
e l emental sulfur concentrations, ultraviol et ab so rpt ion i 
' ' 
,I 
measurement s at 300 nm and 370 nm, and equilibrium ca l c ul a- ,,1 
tions. At any wavelength, the total absorbance is equa l to 
the s um of the absorbances for all of the a bsorbing species . 
At 300 nm and 370 nm, polysulfides are the only known ab-
s orbing species that are expected to be present in the re-
action solutions. UV s pec tra of thiosulfate, s ul fide , bu f-
fe r s olutions , sulfite, ferrous and ferric chloride so lutions 
s howed no absorbances over the range 290-400 nm. Thus, the 
polysulfide concentrations can be determined fro m t he eq uatio n 
wher e 1 is the cell path leng th ' a nd E4 E 5 are the mol ar 
absorptivities for the s; ands; species respectively , a nd 
5 3 
[S;] and [s;J are the concentrations of the re s p ective 
species. The molar absorptivit ie s of the poly s ulfides are 
3420 and 8000 cm-l M-l at 300 nm, and 960 and 2560 cm - 1 M-l 
at 370 nm for S~ a nd S~ respectively (Gig genback, 1972). 
At the conditions of the reaction for pH, s ulfide concentra-
tion and sulfur concentration, the tetra- and p e nta s ulfid e 
s pecies are the only polysulfide s pec ies present. Th e 
amount of each can be determined with the aid of the ex-
pression: 
The B range for the experimental runs was -4.5 to -6 .6, 
with the tetrasulfide predominating below -6.0 . Th e B 
value can then be used with Figure 10 from Giggenbach (1972). 
The ratio between the two peak s at 300 nm a nd 370 nn1 
can be calculated to rang e from 3.35 to 3.12. How ever, early 
expe riments s howed that the ratio between th e two peaks 
wa s a lways l ess than 3 .1. Either a n unknown absorbing 
species wa s present or s mall particl es were pr esent which 
sca ttered th e li ght and rais ed the backgro und equally over 
the entire spectral range . Tests were conducted to eva luat e 
the two possibilities. Poly sulfide so lutions which had the 
proper ratio for the two peak s were u sed . Produ ct materia l 
fro m the reaction of ferrou s iron and s ulfide so lution were 
added. Thi s s olution was used to remove th e possibility 
that s ubmicron goet hit e parti c l es were responsible for th e 
observed ph e nomenon. Th e additio n of s n1 a ll aliquots of 
-- -~----===---
54 
II 
U) 
U) 
II 
U) 
~ 
U) 
~ 
0 
H 
b l) 
,0-,  
C'sl :t 
II ?II 
U) 
(/) 
/ / .... , 
/ ' 
/ ' ..... 
/ 0 
/ ' ' ,-, 
/ ' 
Q) 
H 
;::l 
b.() 
0 0 ?ri 
0 N 
0 0 P ... 
0 00 '? 
rl 
55 
thi s s u s p e n s ion resulted in the ra1? s1. ng o f th e e n ti. re UV 
s pectrum over th e desired range. Th e difference b etween 
the p eaks a nd the s hap e of the spec trum did not c ha n?b. re ? Th e 
ratio betwe e n the two peaks decrease d w1? th t 1 e add.i tion of 
each aliquot. Thus the correct po ssibility was the sca t-
t e ring by s ubrnicron particles of iron s ulfide product. Thi s 
would be con s ist ent with th e fact tl1 at as the reactio11 pro-
ceeded the backgr ound incr eased. 
The polysulfide concentrations were t~1 en determine d 
by s ubtracting th e same amount from bo t h p eaks until th e 
Proper ratio wa s obtained. The pro per ra t 1? 0 was determin ed 
from the following expression 
X 8000 + y 3420 
y = X 2560 + Y 960 
wher e xis th e percentage of s; a nd y is th e percentage of 
s . The percentages x and Y can be determined from th e 
4
value of S and Fi gure 10 from Gi gge nbach ( 19 7 2) . 
A check of thi s entire procedure wa s mad e by taking 
the concentrations det ermined f or the two poly s ulfide 
s peci es , the peak po s itions, th e molar absor p tivit i es , and 
the peak widths to synthe s i ze a s pectrum. Thi s sy nth esis 
was p e rformed on a Dupont curve r eso lve r, which was pro -
vided by the Center f or Mater i a l s Re search. Fi gur e 11 
s how s the validity of the met hod by th e good fit that wa s 
achieved by the synthetic cu r ve a nd th e "r ea l s pectrum", 
over the significant r eg ion 30 0 - 38 0 nm. 
t 
(1) 
u 
h 
Cil 
...0 
;....; 
0 
Ul Synthesized 
.D 
< Spectrum 
M Experimental .~ Spectrum 
250 300 350 400 450 (nm) 
Figure 11 Real and Synthesized Polysulfide Spectra 
c.n 
(J\ 
j 
57 
The peak characteri s tic s for the poly s ul f id es are: 
Species vn'  En'  v1' ;2 vn" E "  v" n 1/2 
3030 3420 595 3676 960 595 
2990 8000 520 37 40 2560 520 
where the v value s are the waveleng th s of th e two pe aks, 
the E values are the molar ab s orptivitie s , and th e v112 
value s are the halfwidths of th e peaks. Pea ks a nd half-
widths are in angstroms (Giggenbach, 1972). 
6. Dilution effects 
The pH of the medium wa s ob served to increa se upon 
the addition of the ferric reactants. A significant por-
I 
:J 
tion of this change wa s probably due to th e dilution of th e 
buffer. In order to de termine th e mag nitud e of thi s effec t, 
a se ries of measurement s wa s ma de. Res ults s how ed that the 
dilution effec t was +0.01 pH unit s / 10 ml of so lution 
added to 500 ml f inal volume . There was no signific ant di f-
ference between the addition of deoxyge na t ed di st ill ed 
water or goethite s us pen s ion. (Fi gur e 12.) 
E. Parti c l e Si ze Study 
Determination of th e s i ze of s ubmi cro n partic l es of 
FeS was attempted by s uc cess ive filtration of product s us -
pens ions . Several (10-15) ml of reaction product s us pe n -
s ion wa s withdrawn from th e reaction vessel and s uccessive l y 
filtered through 0.8, 0. 2 , 0 .1 , a nd 0.01? filters. The 
58 
? 
0.10 
0. 08 
::g_, 0.06 ? 
<J 
+ 
0.04 
??  
0.02 
0.00 1 0 0 
0 20 40 
60 8 0 
ml H20 added/ 500 ml s olution 
Fi gure Change of pH Up on Dilution of Reac tion 12: Solution Du e to Addition of Wat e r 
59 
fir s t thr ee filters (Nuclepore) were polyc a rbona t e wh i l e 
the 0 . 01? filter was a membrane filt e r (Schl e iche r and 
Schuell). 
After each filtration s tep, the polys ulfide s p ect r a , 
total "dissolved" sulfide concentration (i.e. di ss olved 
sulfide and particulate iron s ulfide smaller than the filt e r 
por e size) and iron concentrations were determined by the 
s pe c trophotometric and colorimetric methods a s previous ly 
di s cus sed. 
A decrease in both iron concentr a tion and s ulfide 
concentration would be attributed to filtr a tion of iron 
s ulfide particle s . Measurement s of s ulfide conc entrations 
were complicated by the re sults which s howed a dec r e a se of 
di ss olved s ulfide without a concomitant decr e a se in iron 
concentration in the 0.8 and 0.2? filtration s t eps. Thi s 
effect was attributed to the out ga ss ing of H2S f rom s olu -
tion during the filtration proce s s. As a re s ult it wa s 
neces s ary to provide a blank to corr ec t for th e out ga ss ing . 
The 0.1? filtration wa s repeated two times in pa r a ll e l with 
the 0.01? filtration to corr ect for thi s e f fec t. The amount 
of HzS lost was a s sumed to be independent of th e f ilt e r 
s ize and only dependent on the filtration pro cess i tse l f . 
Inves ti gation of thi s r es ult indica t es that thi s wa s not 
a good a ss umption. consequently , the s ul f ide va lues a r e no t 
very accurate and c annot a i d in the inte rpr e t a tion of the 
Only the iron curve and a bs orb ance va lues wet~ 
r es ult. 
use d i n the particl e s i ze ana l ys i s . 
60 
RESULTS AND DI SC USS ION 
The r es ults of 19 reaction run s were s tudi ed by the 
initi a l rate me thod to determine the rate expressio n a nd 
r a t e constant for the formation of se dimentary iron su l f ides . 
Concentration-time curves, electron transfer balance, hydro-
ge n balance, and s toichiometric relation s hip s were u se d to 
d e t e rmin e a consistent mech a ni s m. 
Data for the reaction run s u sed can b e found in 
Ap p e ndix A. All but two of th e reaction runs u sed goethit e, 
Amorphous Fe(OI-1)
a - FeOOH, as the reactant iron material . 3 
wa s u se d in runs 11 and 12. Th e g o e thit e s u spensio n u se d 
in all the other runs was prepared by the same metho d , but 
thr ee reaction runs (33, 35, and 36) u sed mate ri a l wh ich 
was prepared at a differ ent time . 
Studies by J effe r s on e t a l. (1975) s h owed that coat-
ings on the s urface of iron s t a ined kaolinite minera l s co n -
sis t e d of crystalline goethit e and a morph o u s iron oxide 
c oatings . The equivalent sp h erica l d i a meter of th e goethite 
particles wa s det ermined to b e 300 A . 
Figur e 13 s how s a sampl e co n ce n tration-time curve 
f or th e variou s reactant s a nd products. This semi - logarit h -
mic plot i s u se d to s how the b e h av iou r of a ll the s pecies 
meas ur e d during th e reaction. The co ncentration of reactant 
bisulfide ion ( i. e . so luti on s ulfide) decays rapidly at first 
a nd th e n s low s . This type of b e haviour is typical of a 
61 
s~ 
0 400 800 1200 
Time (min) 
Fi gure 13 : Concentration -Tim e Curves for 
Reactant s and Produ cts (Run 31) 
62 
multiorder rate expression. Similar polynomial g rowth 
curves are observed for the product species. 
Oxidation products 
The reaction between goethite and aqueous bisulfide 
ions resulted in the formation of several sulfur ox idation 
products: elemental sulfur, polysulfides, and thio s ulfate. 
Thiosulfate was found to comprise 14?8 % (mean of 18 runs) 
of the oxidation products on the ba s is of total e lectrons 
transferred. The reaction which can produce thiosulfat e i s 
3 
3 H+ + 8 FeOOH + 2 HS-+ S 2 0-3 + 13 OH + 8 Fe+
Variation of percent thiosulfate as a function of initi a l 
sulfide, hydrogen ion concentration, and total goethite 
A linear c or-
initial surface area is shown in Table 4 
relation coefficient of r = .97 wa s found for the initial 
sulfide concentration vs. % thiosulfate. A line ar corr e la-
tion coefficient of r = .74 wa s found for initial goe thite 
surface area vs. % thiosulfate . The linear corr e lation 
coefficient between initial hydro gen ion and % thio s ulfat e 
was r = .091. These correlations will be expl a ined lat e r 
in terms of the initial reduction reaction. 
Unfortunately, such a comparison o f th e data of ox-
idation product s (conc entration and %) with data from se di -
ment s is limited since th e re 1s only one s tudy (Ro zonov e t 
al., 1971) in which pos s ibl e s ulfide ox idation produ c t s ar e 
reported. The lack of information in thi s a r ea may be a 
63 
Table 4 
Variation of% Thiosulfate with Initial Reaction Conditions 
[Si ] i and% Thiosulfate 
[Sy] i (mM) 1. 8 2.6 4.7 8.7 
!le 
0 thiosulfate 4 3 7 18 
3 
% thiosulfate = -1. 35 + 2.14 X 10- [ST ] i r .97 
Initial Surface Area (Ai) and% Thio s ulfate 
2.51 5.45 6.52 15.2 18.5 32.6 
% thiosulfate 22 22 36 7 5 6 
% thiosulfate 27.3 .824 A? r = . 74 
l 
[f?I+ ] i and % Thio s ulfate 
[H+ ] i (M X 10 8) 1. 06 1. 86 2.79 3.81 4.65 6. 7 5 9. 3 0 
!le 
0 thiosulfate 10 20 36 25 17 19 22 
% thiosulfate 20.1 + 2.52 X 10 7 [1-1+] i r = .091 
I 
~ 
64 
r es ult of s everal experimental dif f icultie s or the ore tical 
pr e conceptions. Experimentally, int e r s titi a l wat e r s o f re-
ducing s ediments are difficult to s tudy due to probl ems o f 
rapid air ox idation (Troup e t a l ., 1974; Br ay e t a l., 19 73). 
Difficultie s in me a surement of small quantiti e s ( <l mM) of 
s ulfate, sulfite or thiosulfate may al s o be a fa c to r in the 
small number of studies of sulfide ox idation produc t s . An 
additional reason for the lack of da ta for thiosulfa t e c on -
centrations in s ediment s i s that Des ulfov i brio can utili ze 
thio s ulfate ion as well as s ulfat e ion for it s r es pir a tory 
process es (Murakami, 1952). Thu s , thiosul fa t e c ould be 
r ecycl e d back to s ulfide and would not be ob se rve d as a n 
ox idation product. 
Ro zanov et al. ( 1 97 1), however, meas ur e d th e con cen -
tration of H sand thiosulfa t e in int erstiti a l wa t e r s of 
2 
anox ic marine s ediments. Conc entra tions of thio s ul fa t e and 
s ulfide ranged from 4 x 10 - 6 to 3 .5 x 10 ~4 M, and fr om 1 x 
10-5 to 4.5 x 10-3 M r es pective ly. Calculations we r e ma de 
us ing thi s data to det e rmine the import a nce of thio s ul fa t e 
a s an ox idation product in the se dim ent. El ement a l s ul f u r 
and thio s ulfat e wer e assume d to be th e only s ulfide ox ida -
tion products. El emental s ul f ur and thio s ul fa te were a l s o 
ass ume d to be produced only by the r eduction of Fe OOH by 
H2S . These calcul a tions s how e d th at o. 9 - 32% of th e ox ida -
tion product s a r e thio sulfa t e (on t he basis of e l ec t ro n 
trans f e r ba l ance ) with an ave r age of 6% f or 1 5 va lues . Thi s 
1s low c ompar e d to the 14 % fo und i n t he pr ese nt s t udy , bu t 
... 
65 
conside ring the fact that some elemental s ulfur ma y be de -
rived from the microbial oxidation of HS- by Desulfovibrio 
and Thiobacillus (Zajic, 1969) the agreement is rather g ood. 
In addition, a linear regression analysis wa s con-
ducted on the concentration of sulfide and % thiosulfate 
(electron balance-wise) reported by Rozonov et al. (1971) 
to see if any correlation existed. Calculations using 15 
points yielded a straight line, with the equation 
= . 7 8 3 + 4 . 3 2 X 1 0 3 [ S ;) 
l 
This line had a linear correlation coefficient of r = .77, 
and these results agree with those of this study which s howed 
a linear relationship between% thiosulfate and total initial 
sulfide concentration. 
Polysulfides, s; and S~, and e lemental s ulfur were 
grouped together s ince polysulfide s are formed by th e follow-
ing rapid reaction of elemental sulfur with di ss olv e d s ulfid e 
ions (Teder, 1971): 
where x = 2, 3, 4, or 5. At th e pH, s ulfid e , a nd elemental 
sulfur concentrations u s ed in thi s s tudy, th e pr e dominant 
polysulfide species ares; and S ~. Other po ss ibl e poly-
sulfides are present only at hi g h e r pH value s (Gi gg enbach, 
1972). The total concentration of e l e me ntal s ulfur can b e 
determined from th e following e qu a tion: 
66 
[S 0 ]t == [S 0 ] + 3 [s;i) + 4 [S5] 
wh e r e (S 0 ] i s the fr e e sulfur a s de t e rmin ed by the a c e ton e 
e x traction. The last two terms repr e sent the compl exe d 
elemental sulfur. 
The overall reaction r es pons ible for the form a tion 
of the elemental sulfur as th e ox idation product 1s : 
2 FeOOH +HS-+ Fe+Z + S 0 + 3 OH + HzO 
Polysulfide s are formed by the rapid subs e qu ent r e action of 
the elemental sulfur as shown above . 
Elemental sulfur is the major ox idation produc t in 
the reaction and as such can s erve a s a qu antitative ly 
significant s ource of S 0 in the s ediment. How ev er, th e 
above reaction for the chemical produ c tion of e l eme nt a l 
s ulfur could provide only half the el ement a l s ulfur n ee d e d 
for the transformation of iron mono s ulfide , mackin awit e 
or amorphou s FeS, to pyrite. Thi s a ss um es that a ll of th e 
ferrous iron i s initially reduc e d by s ulfide s p ec i es . Mi -
crobial proce s ses are probably r es pon s ibl e for the bulk o f 
elemental sulfur in ano x ic sedim ent s . Thiobacillu s de nitri -
ficans and other sulfide ox idi zing ana e rob es c an ox idi ze 
s ulfide to el emental sulfur, although these ba ct eri a prefe r 
to ox idize elemental sulfur to thio s ulf a t e a n<l s ulf a t e 
(Zajic, 1969). 
Data by Berne r (1964 and 1 97 0) a nd Ka pl a n e t a l. 
(1963) s how s. that pyrit e was th e ma j o r iron s ulfi<l e found 
67 
1n r ecent marine sediment s . Sulfur derived from th e chemi-
cal oxidation of sulfide could have accounted for only 50% 
of the e lemental s ul fur found as free and pyrite s ul f ur . 
This calculation is based on th e amount of total iron s ul-
fides , both mono- and di s ul fides, and th e concentration of 
total e lemental s ulfur. 
Reduced iron 
What , then, is the predominant reaction respo ns ibl e 
for the reduction of iron 1n the sediment? Several bac t eria 
of the genu s Pse udomonas can r educ e l arge quantiti es of fer-
ric to ferrous iron in soils and bo gs . Howe ver, littl e is 
known about the iron reduci ng capabil iti es of mi c rob es in 
anoxic sedim ent s . 
Thermodynamically, th e reduction of iron s houl d oc-
cur prior to the r e duction of s ulfat e (and formation of 
s ul fide) . The redox r eac tions for th ese two species are: 
F e +3 + e - F +2 -+ ?e 
(Stumm a nd Morgan, 19 7 0) 
Thus the reduction of iron would not pro ceed by th e co n -
comitant ox idation of s ul fide spec i es . Th e thermodynami c 
seq ue nce of reactions a bove would occur in the sed im e nt i n 
which the oxic-anox ic boundary occ ur s well be low the sedi-
ment-water interface. Thi s t ype of sed im ent co lumn would 
have a n Uh profile as s hown in Fi gure A be low. 
68 
Dh 
Eh 
0 + 
- - ,:-:;- ------- -
X 
y 
+ 
B 
A 
At point x , the reduction of ferric iron would occu~ while 
s ulfate reduction would occur a t pointy. 
In many a noxic sediments, the boundary be tw ee n the 
ox idizing and reducing zo nes occurs at or s li ghtly a bov e 
the sediment-water interface, and th e entire se d i ment c olumn 
ha s a nega tive Eh. An Eh profile for s uch a sediment c olumn 
In thi s environm ent, both s ul -
is shown in Figure B above. 
fate ion and ferric iron s peci es could be r educ ed s imultane-
In thif type of anox ic se diment, the rcduc tion of 
ously. In ex-
Fe+ 3 by HS- could occur, produc ing e l emental s ul f ur. 
trem e cases, the overlying bottom wat e r s a r e a noxic , and HzS 
i s fo und in the water co lumn. (The Bl ack Sea i s the c l assic 
ex ampl e of this type of environment.) Here th e r educ tion 
proce ss can occur in th e wat er co lumn. Brewer a nd Spe nce r 
, 
(1974), studied the part_i'f: ..i on f 
0 !:race 
dissolved Bllc/ part.iculate phase ? 
? s ..1.n t-he 
served a rap.id r.ise .ill the concen t:ra t: .ion 
jn the anox. ic part of the water co..I um.1
1 ~ wh 
!;/ 8 rap.id decrease, as they descended .in t:o 
, joJ1 of the water colunm. 
0 Th.is beha v.iour h 
7 
?t1idence for the rapid reduction and d..z?sso.Z 
s-aturat.ioll w.ith respect to .iron s u..If..z?des (. 
;tation of .iron sulf.ides occurred when t:he 
1j'/1r,na' ' tion was exceeded. 
qJJJOUJlt of ferrous iron can be det:erm.ined 
9 C. 
.o JJ of the sulfide oxidation product:s and t: 
er Changes that occur in the spec.i f.i c redo_ 
ox.ida tion number changes for '!:he fo..l ..I ow..z?n,g 
6, 8, 8 , 6, and 8 . 
HS- -+ S 0 +2 + 2 Fe + 
eOOlf + 4 0/-/ 
0fi + 4 HS - -+ S/ + 6 Fe +
2 + .12 0/-/ -
7
.rf+SlfS--+S5 + 8 Fe +
2 + .16 0.1-/-
+ 2 f/S -+ S O = + 8 Fe + 
2 + .1 3 0/-/ -
2 3 
f/S - -+ SOJ + 6 Fe + 2 + 9 0/-1 
+2 
f/S -+ so; + 8 Fe + .1 2 0/-/ -
75 that are t.ralls ferred fro/JJ th e 
ag e11t, which in 't:hes e react:.ions 
?TJJlined from '!:he fo.1.1 ow.ing 
69 
(1974), studied the partition of trace elements b e tween the 
dissolved and particulate phases in the Black Sea. They ob-
serve d a rapid rise in the concentration of dissolved iron 
1n the anoxic part of the water column, which was followed 
by a rapid decrease, as they descended into a more anoxic 
region of the water column. This behaviour was interpreted 
3
as evidence for the rapid reduction and dissolution of Fe+ , 
until saturation with respect to iron s u lfide s occurred. 
Precipitation of iron sulfides occurred when the Ksp for 
their formation was exceeded. 
The amount of ferrous iron can be determined from the 
concentration of the sulfide ox idation products and th e ox-
idation number changes that occur in the specific redox re-
changes for the following 
actions . The oxidation number 
reactions are 8' 8 ' 6' and 
8. 
2' 6' 
so +2 + Fe + 4 OH 
1-t + 2 FeOOH + HS- -+ 
2 
HS -+ s= + 6 Fe+Z + 12 OH 2 1-t + 6 FeOOH + 4 4 
r + z 
1-r+ + 8 FeOOH + 5 HS -+ s= + 5 8 ? e 
+ 16 OH 
3 
Fe+ 2 + 13 OH 
3 H+ + 8 FeOOH + 2 HS 
-+ s o; + 8 2
+ FeOOH + HS -+ so=3  + 6 
r? e + 2 + 9 OH 
3 1-i+ 6 
-+ so= + 2 + 8 Fe + 12 0 1-1-
3 1-t + 8 FeOOH + HS 4 
Thus the number of electrons that are transfer red from th e 
sulfide ion to the oxidizing agent, which in the s e reactions 
is th e ferric 1011, can be determined f ram the fol lowing 
e quation: 
7 0 
2 [S 0 ] + 6 [S4] + 8 [S5 ] + 8 [S o == ] 
2 3 
+ 6 [ S0 3] + 8 [ SO~JJ 
Thi s i s base d on the a ss umption th a t ir o n i s the onl y ox-
id iz ing agent pre sent. On e me thod of t est ing t his ass ump -
ti on i s to see i f the e l ect ron ba l a n ce exceed s th e initi a l 
c onc e ntr a tion of ox idi z ing a ge nt - th e co ncen t r at ion of 
g o e thit e . A c h e ck of 90 e l ec tron b a lance meas ur e me nt s 
s howe d tha t 74 % of the me a s ur e me nt s indic a t e d a n e l ec tron 
b a l a n ce l ess th a n the initi a l iron c on centr a t io n. Th e re-
ma ining 26% o f the readings , whi c h exceed ed th e t h e o r eti ca l 
limit, we r e t a k e n during r eaction r un s in whi c h th e ra t e o f 
s ulfide f orma tion was hi gh. Thi s i s s hown by th e hi g h r e -
duce d iron values in r eaction runs 11, 1 2 , 33 , 34, a n d 35, 
which a r e s hown in App e ndi x A . Thi s ex c ess i s t o o l a r ge to 
b e a ttribut e d to experim e nt a l e rror. 
Compl e t e c onve r s ion o f g o e thi te to i ron s u lf i de was 
o b se rve d within 30 minut es f or r un s 11 a nd 1 2 . In t hese 
two runs , th e e lectron ba l a n ce exceed e d th e to t a l reducibl e 
iron by 1.8 m~ Th e el e ctron bal a n ce in these two r un s , h o w-
e v e r, d ec r ease d over 24 hour s until th e va lu es were 1. 8 a nd 
0. 7 mM f or runs 11 a nd 1 2 r espec t ive l y . A simil ar decrease 
was a l s o ob se rve d in th e ac i d ext r ac t a bl e so l id s ul f i de ( Sp) 
f rom a max imum o f 1.46 mM to 1.0 3 mM ( in reaction r un 11). 
Thi s ad s orb e d or copr e cipitat e d s ul f id e c ould ha v e b ee n 
o x idi ze d during the aceton e ex t ract ion to for m e l e me n ta l 
s ulfur. Thi s would h a v e i n c r eased t h e e l eme n ta l s u lfu r 
71 
c onc e ntration and total reduc e d iron values . "Excess " 
r e duc e d iron wa s al s o ob se rved in runs 33, 34, and 35, in 
whi c h 1.3, 0.4, a nd 0.5 mM of e x c es s r e duced iron wer e meas -
ured. The s e thre e run s had the hi g hest r a t e con s tant f or 
f ormation of iron s ulfide. During thi s r a pi d form a tion, 
bi s ulfide ion could have been copr e cipitated a s in runs 1 1 
a nd 12. This ad s orbed s ulfide c ould e xplain in a simi l a r 
f as hion the high level s of reduc e d iron calculat e d for 
runs 33, 34, and 35. Such a phenome non was ob s erv e d by 
Be rner (1969) in a study o f California co as t a l se d i me nt. 
Rickard ' s (1974) sugg ested me chani s m for th e formation 
o f s edime ntary iron sulfide s includ e d di ss oluti o n, r e du c -
tion, and precipitation step s . Idea lly, the way to d e t e rmin e 
the orde r of the s equence would b e to d e t e rmin e th e ra t e s 
of the individual r e action s . Qu a lit a tiv e a n a ly s i s o f th e 
c oncentration-time plot s for th e r eac tion runs ma y g ive s om e 
in s i ght into the r e actions involve d. Plot s of r e duc e d iron 
v s . acid e x tractable sulfide (S~) for se v era l r e a c ti o n s ar e 
g iven in Figur e 14. The 1 ine incl i ca t es e qu a lity b e tw ee n th e 
a mount of reduced iron a nd the a mount of FeS produced. I f 
th e point s fall above the line, thi s indi cates th e se qu e n ce 
of precipitation (adsorption) follow e d by re du c ti o n. Con -
v e r s ely, if th e point s fall b e low th e line , r e du c tion pr e -
cede s precipitation. T}1 e time int er v a l b e tw ee n the two r e -
ac tion s i s propo r tional to the d i s tan ce of th e poin t f r om 
th e lin e . Thi s plot s how s tha t r e du c tion prece d es t h e pr e-
c ipitation r e action by a s i gnifi c ant time int e rva l. 
72 
? 
? 
? 
2.0 
11 0... 
U) 
'--' 1. 6 ? 
? 
(l) 
'tJ 
?r-1 ? ? 
4-, ? 
rl 
;J 
U) 
(l) 1. 2 
rl 
..D 
ro 
+J 
u 
ro 
H 
+J 
>< 
?.:i 0.8 
? ? 
? ? ? 
'tJ 
?r-1 
u 
<i:: . ? 
. ? ? ? 
0.4 ? ? . ? ? ? ? 
? ? ? 
?.. ?  ? ? ? ? ? . ? ? ? . ? 
? ? ? ? 
? 
0.0 
0.0 0.4 0 . 8 1. 2 1. 6 
Reduced Iron (mM) 
Figure 14 Plot of Acid Ex tractabl e Sul f id e 
Co ncentration v s . Reduced Iron 
Co ncentration 
73 
Several concentration-time plots of FeS and electron 
balance are s hown in Figures 15, and 16. These plots also 
s how that reduction preceeds precipitation. The reduction 
process occurs rapidly at first but then slows down as the 
available reducible iron (i.e. s urfac e iron) is depleted. 
Addition a l reduction of iron can only occur after the dis-
sol ution of the surface ferrous iron. 
Quantitatively, the reduction reaction was studied 
by analyzing the e l ec tron balance-time plots by the initial 
rate method. A detailed description of the initial rate 
method can be found in Appendix B. Slop es of the p lot s were 
first determined by simple linear regression. All but four 
reaction runs exceeded the 95% confidence level for the cor-
relation coefficient; those rejected were not used in the 
determination of the rate constant. The slope s were then 
plotted against the initial concentrations to determine the 
reaction order with respect to the particular sp ecie s . Thus , 
lo g -log plots are shown in Figuresl7, 18, and 19 for HS-, 
+ The reaction orders were determined to b e 
H ' and AFeOOH" 
0.60, 0.49, and 0.89, respectively. The rat e co nstant wa s 
calculated on the basis of 10 runs (Table 5) to be 0.0179 
Thus the rate expres s ion was 
determined to be: 
cl[Red Fe]/dt 
0.8 
0 . 6 
,---. 
~ 
(..:: 
'--' 
0 . 4 
Sp 
0 . 2 
0 . 0 
0 200 400 600 8 00 1 00 0 1 20 0 1 400 
T i me (m i n ) 
F i gur e 15 : P l o t o f Reduce d Iron and Acid Ex tr a ctab l e Sulf i de, si , vs. Time (Run 6) "-l +>-
~ 
- - .. 
3.2 
2 . 4 
,--, 
~ 
E 
'--' 
1.6 L Red Fe 
0 . 8 
0 . 0 -=-_........._ _ _,__--1._ _. .___  _J___  _.__-1.._ _L __..J..__  __l..__--1.._  __JL__..L__  _j__  _j__  __J 
0 200 400 600 800 1000 1200 1 400 
Time (m in) 
Figure 16: Plot of Reduced Iron and Ac id Extractable Sulfide, S~, vs. Time (Run 24) 
'-J 
u, 
76 
.? 
'l,j 
'-.. 
,---, 
(1) 
eL, 10-6 
'l,j 
(1) 
~ 
'D 
10-7 i----~-_.._....__.__._.._._.....L.Ji-__ _ _L._  _L_.L-...L-..L-L....L.....1-...1 
10-3 10 - 2 10 - 1 
Sr. (M) 
1 
Fi gur e 17: Plot of Rat e 0 ? Fe Re du c tion vs . ST. 
1 
77 
,....., 
~ 
?r-1 
E:: 
......... 
:2 
'--' 
.? 
'U 10- 6 ......... 
,-, 
(l) 
?.. 
'U t 
(l) 
.0__::_: , 
'U 
Figure 18 : Plot of Rat e of Iron Redu c tion vs . 
Initial Hydro ge n Ion Conce ntr a tion 
78 
r-, 
~ 
?rl 
E; 
'---
~ 
'--' 
.? 
'D 
'--- f 
(1) 
r..r... 
10 - 6 
'D 
(1) 
0::: 
'--' 
'U 
10 - 7 L- --'-- -l-- -'-_._-'-L--L-1-l_ ___ i,__  _L_J__.J_j_LJ_..J_J 
1 10 100 
A-  
2
Fe OOJ-I. (m /l ) 
1 
Fi gure 1 9 : Pl ot of Ra t e of Re du c ti on of Fe v s . 
I nitia l Co e t l1 i t e Sur f ace Ar ea 
79 
Table 5 
Rate of Reduction During Pha se II (Ra te of Dissolution) 
49 
d[Red Fe ]/dt = kred [HS- ] i 60 (H+)f Aft~gOH . =mt+ b 
1 
Correlation 
Coef fici ent 
Run rn b kr ecl 
r 
6 1.10 10-
7 6.50 X 10- 4 2.68 X 10-
3 . 7 2 
X 
10- 7 10- 4 2 .04 10-
2 
7 5. 30 1.40 X X 
.96 
X 
7 4 3 .75* 
9 6.25 X 10- 6.50 X 
10- 4.02 X 10 -
6 10-4 4.37 2 17 1. 40 10- 1. 40 X X 
10 - * * .99 
X 
7 5.90 10- 4 6 2 
2 
10- 1. X 10 - .99 21 3.07 X X 
7 4 2 .97 
22 6.50 X 10 ~ 3 . 90 X 
10 - 1. 84 X 10 -
6 10- 4 10 - 2 10- 6.00 X 1. 69 X . 98 24 1. 50 X 
6 4 2 
25 1. 09 X 10- 2.86 X 10 - 1. 98 X 
10 - .96 
6 4 
28 1. 03 10- 5.25 X 10 - 1. 66 
2 
X 10- . 96 
X 
6 10- 4 10- 6.90 X 1. 64 10-
2 
X .98 29 2.18 X 
7 4 2 . 97 
30 8 .49 X 10- 9.30 X 
10 - 1. 68 X 10 -
6 10 - 4 10 - 2 10 - 5.50 1. 73 X . 96 31 1. 88 X X 
6 4 10- 4. 70 V 10 - 2.05 X 10 -
2 .4 2* 
3 2 1. 78 X " 
6 4 2 
33 1. 48 X 10 - 3 . 28 10 - 1. 99 X 
10 - . 96 X 
4 2 
34 7.50 X 10-
7 6. 70 10 - 1.1 2 X 10 - . 79* X 
7 10- 4 1. 24 10 - 2 35 6.69 X 10 - 6 . 70 
. 89 
X X 
36 Not De t ermined 
*not s tatistically s i gni f i cant **discarded 
80 
Visual examination of th e plots of e l ectron transfer 
balance against time, revealed that the data point s were 
more likely to be fitted by a quadratic curve. (See Figures 
15, 16, 20.) The curves through the se point s were determined 
by a lea s t squares polynomial fit. However, when th e initial 
s lopes were used to determine the c oefficient s for the rate 
expression, no significant linear fit was obtained. As a 
result, the obviously polynomial curves wer e analyzed with 
the reactions responsible for the generation of this dat a . 
It was noted that the initial electron balance me as ur eme nt s 
( taken at times less than 40 minut es ) were in some reaction 
runs, as much as 30 % of the final e l ect ron balance meas ur e-
ment s . Ther e fore, the polynomial curve was int erp r eted as 
resulting from a two s tep reduction r eac tion 1 which is graphi-
ca lly represented in Figure 21 . In the first step of the 
reaction, Phas e I, the rapid r e duction of surface iron occur s . 
This is indicated by the lar ge initial s lop e o f th e c urve at 
In the second s tep, Phase II, th e und e rlyin g 
time zero. 
layers of the ferric iron are reduc ed as the s urface layer 
of ferrous ions is di sso lved. Thus the r ate of r eduction 
in Phas e II is actually controlled by the rate of dissolution 
of the reduced s urface layer s , and the rate of reduction 
during Phase II gives the rat e of dissolution. 
The rate of the initial s urface reduction of iron in 
Phase I i s difficult to assess because of the rapidity of 
the reaction, and the in s uffici ent number of data point s 
during the first 30 minute s. Additional studi es, how ever, 
3 . 2 
2.4 Red Fe 
,--... 
~ 
E; 
'--' 
1. 6 
Sp 
0 . 8 
o. o---_._---'.___  _,_ _____ '---.......L...----1--...,1_--1_ _. .J__  ___,J_ _. ...1__----1._ _. J__  __._ _. 1__  _J 
0 200 400 600 800 1 000 1200 1 400 
Time (m in ) 
00 
Figur e 20 : P l o t of Reduced Iron and Acid Ex tractab le Sulfide, S~, vs. Time (Run 31 ) f--1 
82 
------ - - - -----------------------, 
Initi a l 
Re du c t io n Re duc tion and 
Rat e Di ss olution 
Pha se II 
Re du- c-ib-l e- T-r-o-n- L-j-rn-jt = - ? 
--~> 11 ?--- ---- P 11 J s (' 1 r 
Phase I 
Fi gure 21 : Ide a li ze d RcJu ct i o n - ili ssolution Curve 
83 
may be able to determine the rate expre ss ion and rat e c on -
s tant for the r eduction process. 
Any postulated mechanism for the reduction reaction 
must explain the rate expression, s ulfide oxidation product s ' 
and variation 1n the products with respect to initial con-
ditions. Two basic approaches can be used to develop a 
mechanism: the rate determining step approach, or the st e ady 
s tate hypothesis (Edwards et al., 1968). In thi s s tudy the 
former was adopted, since data by Rickard (1974) indicated 
that a rate determining step might have been involved in 
the reaction. 
Any complex chemical reaction proceeds by a mechani sm 
of several elementary steps. The overall rate of the reac-
tion is determined by the rate of the s lowest s tep, th e rate 
determining step. 
At this point it is necessary to clarify thre e con -
Order 1s an 
order molecularity and stoichiometry. 
cepts: ' 
experimentally determined value in the rate expre s sion, whil e 
molecularity is the number of reacting specie s or mol ecul es 
involved in the formation of the activated c ompl ex (Pilling , 
1975). Stoichiometry is the value of the coeffici ent s in 
the reaction. These three quantitie s mayormay not have the 
s ame numerical value, depending on the r eaction. 
The redox reaction between FeOOH and aqueou s sulfide 
s pecies required the approach of reduc ed s ulfide s pec i e s to 
the s urface of the goethite, wh er e the e lectron tr a ns f e r 
can occur. The rate of approach and d i s t a nce from the 
84 
s urface are affect ed by the surface char ge of the s o l id and 
po ss ibly by physical and chemical adsorption (Reynold s a nd 
Lurn r y, 1966). The surface of an ox ide or oxyhydrox ide has 
a s urface charge as a result of the interaction of the s ur-
fa c e oxygen atoms with water in the bulk s olution. Ba s i cally, 
the mechanism for this reaction involves the protona tion 
and deprotonation of surface ad s orption s ites (P a rks a nd De 
Bruyn, 1962). The surface charge is a function of pH , a nd 
the pH at which the surface charge i s zero i s c a ll e d the 
ze ro point of cha rge. However, when H+ and OH - a r e the 
potential de termining ions, this pH i s call ed the i s oe l ec tri c 
point (IEP) (Berner, 1970). 
The double layer consi s t s essentially of a c ha rge d 
s urface layer of pot ential determining ions , a nd a layer of 
oppo s itely charged counterions in s olution pa ra lle l to the 
s urface of the s olid. This layer of mobil e count er i ons i s 
ca lled the Gouy layer . The thickness of the Gouy l ayer i s 
a ffected by the ionic strength and s tirring of the so lution. 
An increase in either or both of these f ac tor s will r educe 
the thickness of the Gouy layer. The c ompo s ition of both 
the Helmholtz laye r of potential determining ions a nd the 
Gouy layer of counterions c an change in res ponse t o cha nges 
1n the composition of the bulk s olution (B erner, 1 97 1). 
Physica l and chemical ad s or ption may a l s o affec t t he 
rate of electron tr ans fe r acro ss the s ol i d -so lut i on i nt erface 
85 
a-Fe o ha s been shown to chemi s orb HzS from the gas phase 
2 3 
resulting in the formation of HS- and H+ (Blyholde n a nd 
Richardson, 1962). No similar studies have been conduct ed 
on the H S- goethite sys tem, but r esult s by Gast et al. c1974 ) 
2
s howed that goethit e hydrogen bonds the first layer of ad-
sorbed water to the surface . This hydrogen bonding wa s 
s hown to be s tronger than observed in a-Fe 2o3 . The adsorbed 
water on goethite was also shown to be readily exc hanged 
With DzO. 
The physica l adsorption of ions is dependent on the 
charges of the ions and the surface of the particl es . As 
s hown above, the surface of goe thit e is ne ga tively charged 
Therefore, the phy s ical ads orption of water 
above p rJ1  6 ? 7 ? 
is hindered above pH 6.7 by coulombic interaction. Below 
the IEP however, the surface is positively charged and physi -
conse quently reduction 
cal adsorption s hould be rapid 
would be rapid if the adsorption of HS- were the rate deter-
mining step in the reaction. 
The rapid adsorption of HS - be low the I EP was demon -
s tr at e d in react ion runs 11 a nd 1 2 ' where Fe ( 0 H) 3 was us e cl 
5 
a s the reactant iron phase. The IEP for Fe(OH) 3 is s.
(Parks, ). Results for the se two runs s how ed that the 
196 8
formation of iron s ulfide wa s essentially c ompl ete within 
8 5
the first 30 minutes even at pH ? ? 
A comparison of reaction run 11 with one of comp arabl e 
conditions with goethi te as a reactant (run 6) shows a drama -
t1?c d" te of iron s ul fide format1? 011 . At 
1fference in t 1 e ra 
86 
30 minutes, 1.48 mM of FeS wa s produced in reac tion run ll 
(pH= 7.58) while in run 6 (pH= 7 . 554), only 0.02 mM of 
FeS wa s produced (extrapolated from Fi gure 15) . Al t hough 
there is a difference between the surfac e a r eas of th e r e-
2
actant iron phases (6 . 5 m2/l in run 6, and 14 m /l in run 
11), this difference in initial surfac e a rea could only ac -
count for 1% of the difference in the amount of Fe S produc e d . 
This would then indicate that the rapid adsorption of HS -
in the cas e of run 11, with a pH below . the IEP 
do es occur 
of Fe(OH) . Adsorption of HS- does occur above th e IEP of 
3
goe thite, but to a less er ex tent than it would be low th e I EP . 
The r eduction of haematit e would al s o be exp ec t ed t o 
oc cur rapidly, since the IEP for haematit e i s 8.5 (P a rks, 
1968). Thus HS- can exchange with OH - in the pot enti a l de -
termining laye r but the exchange would be favor e d a t pH con -
ditions e i ther ne a r or below the I EP . Thi s would expl ai n 
the increase in the rate of the r eduction as th e in i ti a l pH 
- i? s pre s ent in th e pot ential <l e t e rnn? ning 
i s r educ ed. Once I-Is
layer, the e lectron transfe r can oc cur. How ev er, the ox i da-
0 
t i? on change?-  for Fe +3 goi? ng to Fe +
2 and s = to S (pri nc ipa l 
ox idation product) differs by 1 . Ther e for e , th e r educ tion 
of one ferric iron by one bisulfide ion will r es ul t in a 
hi ghly unstabl e ox idation s t a t e of -1 f or s ul f ur (L aure nce 
and Elli s , ). The se author s s tudi e d the oxi dat i on of 
1972
aqueou s iodide ions by aqueous f er r ic ion . The mechan ism 
they po s tul a ted requir ed the forma ti on of I 2 whic h r eacted 
Wi th f er ri c ion by th e r eaction: 
87 
Fe+ 3 + C -+ I?'? e 1+ 2 
F 1+ 2 +2 ?e + I -+ r e + J 2 
Fe+ 3 + r; +2 ->- Fe + I 2 
HS- could react as doe s I .1n the ;1hove reactions , to yield 
HS 0 , equivalent of I~. However, in the reaction of goethi te 
with aqueous sulfide species, t he nt craction of 1IS 0 i with 
anoth e r ferric iron coordination s ph ere is not necessar ily 
required, as it ' s a nalo g in t he Jbovc me chani s m. Th e n 
HS- can be thought of as be .in g ;1ssoc L~1ted with a s urfac e 
adsorption s ite, s inc e the purpose o r the fixed laye r i s to 
balance the charg e of the p r o to n ~1 t c d s ite. The structur e 
of goethite cons i s t s of he xagonal l y cl ose- pa cke d ox yge n 
atoms with iron in the oc t ahedral inte rstices (Deer, Howi e , 
and Zussman, 1966). Thu s , ecJc h su r [ ;1l e oxyge n i s s hared by 
two iron octahedra, a nd th e pot e nti ;1 J det ermining HS i s 
a s soc i ate d with two iron o ct ah c d 1? a . both octahedra co ntain 
ferric ion which can be r educed. !I S- would then transf e r 
one electron to on e of th e ferr i c i 0 11 :-; , resulting in th e 
formation of J-I S 0 I-I S 0 ? r e a cts i 111rn cd i ; 1 tely by tran s ferrin g 
an electron to th e other fcrrL c 10~ with which it i s a ss oci -
ated. This would produ ce c ] cmcnt;1! s ulfur, S 0 , hydroge n i o n , 
and two ferrou s ion s . Th e st ru ct ur e of the goethite favor s 
a two e lectron exc han ge . Th i s 111 ccll;111 '. Sm would exp l a in why 
s ulfur is the prin cipa l ox i clat ion pr odu c t. 
How, the n, i s thio su] ra t c produce d? Th e for matio n 
o f th i o s u 1 fat e wo  u 1 cl re q u i r e c o 11 e: i t i o ns  wh  i c 11 :fa v o r 
88 
interaction between two or mor e unstable ox idation species 
HS 0 ? An increase in HS - concentration would increase the 
0 
number of HS -occupied s urface adsorption sites, which 
would also increase the number of HS 0 species. An increas e 
in the total dissolvetl s ulfide, at constant iron reactant 
s ur face area, should result in an increase in% thiosulfat e. 
This was observed in the results of the present st udy and 
in the results of Rozanov et al. (1971). 
Conversely, an increase in the total number of 
s urface adsorptio n sites would decrea se the% thio s ulfate of 
the oxidation products, by decreasing the interaction b e-
tween HS 0 species. A low correlation (r = 0.09) wa s ob-
served between the% thiosulfate and the initial pH. Th i s 
would indicate that sulfide speciation was not an important 
factor in the formation of oxidation products. 
It appears that both HzS and HS- are equa ll y re-
active. Thi s interpretation was made in view of the fact 
that the relative proportions of H2S a nd HS- c h a n ge s i gnifi -
cantly over the pH range of this s tudy. If the spec i ation 
of reduced s ulfur was important, the re lative proportion 
of oxidation products should change in re s ponse to a varia-
tion in pH. 
Thus the mechani s m fdr the reduction reaction i s 
postulated to b e : 
89 
(1) protonation of surface adsorption s ites 
(2) exchange of SH with OH in the fi xe d l ayer 
of the iron phase 
(3) consecutive transfer of two electrons from 
ads orbed HS- to surface ferric iron 
(4) formation of a protona ted layer of Fe(OH)z 
(5) dissolution of a Fe(OH) 2 layer. 
After reduction of the surface lay e r of ferric iron ' 
the reduction of ferric ions in the underlying bulk of s olid 
can only occur after di ss olution of the s urface l aye r of 
ferrous ions. Ferrous hydroxide , Fe(OH)z is metastabl e at 
the pH values of this s tudy, and in natural a noxic sediment s. 
The s olubility product for Fe(OH)z is 
Fe(OH)z ! Fe +2 + 2 OH-, lo gKs p = -15 .15 
(B aes and Mesmer 19 76 ) ' 
The concentration of iron in equilibrium with so lid 
Fe(OH)z at pH 7-8 is 10?1.15 to 10 ?3.15 Fe? 2 Thu s the 
s urface layer should dissolve and expose th e remainder of 
the s olid (Walton, 1967). Ki ne tic s of diffusion co nt ro ll e d 
dissolution ar e first order with respect to s ur face ar ea. 
However , ther e is ample prec edent for r eac tions in which 
di ss olution is controlled by the ch emi ca l reaction (Moelwyn -
Hughes, . Th e rate expression for the reduction and 
19331
a?i sso .l ution phase was determine d to b e 
= d Di ss olution/cit = 
d [Red Fe]/dt .89 _ . 60 (H+) . 49 ApeOOH 
kD [HS \ i l 
90 
He r e , a first order (0.89) dependanc e on s urface 
area was seen, but a half -order dependance on both HS- (0.60) 
and H+ (0.49) was also observed. This hi gher order kinetic s 
indicate s a chemical control for the dissolution reaction. 
From rate determining step theory, the compo s ition 
of the activated complex is FeOOH ? 1/2 HS- ? 1/2 H+. Si nce 
molecules do not react by halves, the compo s ition i s there-
fore 2 FeOOl?I ? HS- ? H+. This i s the activated comple x for 
the reduction reaction. 
I/ 
/le""' 
0 01-1 
, 1 / 
Fe 
/ oI  ""' 0H+s1r 
I / 
Fe 
/ 
""' 0
I  ""' 
FIe / 
OH 
/I " 
After reduction, the product e l emental s ul fur would 
diffuse out of the s urfac e of the s olid (Mo e lwyn-Hu ghes , 19 33) . 
The hydro gen ion from the HS- would remain to protona t e th e 
s urface since the IEP for Fe (OI-1)2 i s 12 (Park s , 1965). 
Hydro gen balance 
Large quantities of hydro xy l ion a r e produced by th e 
r e duction reaction of goethit e by HS- . In addition, 
hydroxyl ions are produced by the prec ipitation reaction of 
ferro u s iron with aq ueous bi s ulfide a nd th e formation of 
91 
polysulfide from elemental sulfur. 
S 0 + 2/7 HS-+ 1/7 s~ + 1/7 s; + 2/7 H+ 
Fe+Z +HS-+ FeS + H+ 
Th e overall reactio~therefore would be 
14 FeOOH + 23 HS + 14 FeS + 19 OH- + 9 HzO 
+ S4 + S5 
8 OH- + 5 HO+ s 0 =
8 FeOOH + 10 HS-+ 8 FeS + 2 2 3 
 
Thus the hydroxyl:iron sulfide product r atio s hould 
be 1.36 and 1 respectively. Since sulfur was the principal 
oxidation product, the ratio s hould be closer to 1. 36 . Table 
6 g ives the ratio of rrrillimO:tes of hydro xy l 1011 and the milli -
moles of FeS as Sp? For all but two runs, the ratios of 
these value s is less than one. Ei ther there exists an 
additional sink for OH- or the re ac tion does not proceed 
as s imply as is stated above. 
Chemical analysis by Berner (19 64a) of hi s product ' 
iron monosulfide, showed that the ratio of Fe:S wa s not 
1:1, but rather o.9:1 to 1.1:1. The iron deficient mono -
sulfide was hypothes i zed to form by the adsorption or co-
uct 
precipitation of NazS with the FeS. Analyse s of the prod 
nd 
iron s ulfide wer e conducted for Fe, Na a S . The excess 
s ulfide was then determined by s ubt racti ng 0.5 Na+. Th e 
However 
ratio was then determined as 0.9 at a minimum. ' 
the excess sulfide was assumed to be present as NazS, but 
Goldhaber and Kaplan (1975) have s hown thats= does not 
92 
Table 6 
Hydrogen Balance 
'\.,350 '\.,400 '\.,600 
'\.,750 '\J l440 
Time '\., 30 '\.,2 00 
(min) -
Run 
0.30 0.30 
0.62 
6 0 
o.74 0.59 
0.84 
9 0.40 
2.95 1.19 
1. 2 
22 6.80 1.16 
0.40 0.54 0.32 
24 0 0 0.51 
1.48 0 . 7 4 
25 0.37 0.54 
0.10 0.10 
0.19 
28 0 0 0.50 0.57 
29 o.71 o.52 
o.36 
0.10 0.19 0.19 
30 0.33 0.14 
0.41 0.57 0.46 
31 1. 29 o.48 
0.80 0.71 o.ss 
33 o.64 1.07 0.94 
1.47 1.49 
0.84 
34 1. 36 
0.93 1. 34 
35 6.60 4.00 0 . 43 1. 2 7 
1.70 1. 26 
0.93 
36 2 .40 
Ratio of mM +tOH-/mM S~ 
Exact times may be found in Appendix A. 
93 
ex i s t at neutral pH value s . Rath er , N: iHS would be pre sent, 
a nd the Na:HS ratio would be 1:1. Thu s the minimum ratio of 
Fe : S would b e o. 8. Thi s factor or O. 8 'N" ould reduce the 
01-!- :FeS in a lar ge number of dat;i sets ) Ut sti ll some of 
the data would not h ave th e requjr cll r; itio of 1 to 1.36. 
Previous mec hani s ms a nd reaction s 1 0 1? the formation of iron 
s ulfide (Rickard, 1974, Berner, 19(1 2 ) d id not acco unt for 
the product 01-i- ion s , as 1n Ri.c lc i rd ' s (1974) mechani s m 
Fe+ 3 + e -+ Fe+Z 
FeOOI-1 + e -+ Fe+Z 
An additional s ink fo r hydrn xy l mi ght be adsorp tio n 
of hydroxyl within th e structure or th e iron mono s ul f i des . 
This proce ss would r emov e add i t i on;1 l hydroxide ion fro m 
so lution s o that it would not h e me;1s ure d by th e h ydroge n 
e l ec trode. Quantitativ e l y thi s co u I L: only acco unt for a 
fe w percent o f th e excess h ydro xy I I o ,is . Thi s c oprecipi t a t e cl 
hydroxy l might account for t h e C;1 c t th.i t the initi al i ron 
s ulfide is amorphou s (Berner , l '..JC14;i) h ) hind e ring c r ysta lli -
za tion. Adsorbed h ydroxy l cou l d po s ::-; i1 l y b e determi n e d by 
infrared s p ec tro sc opy, but t h e re ;1r e n r, a vailabl e dat e for 
the pres e nce of hydroxy l group s in i r o il s ul fides . In 
addition, two ex perime nt a l prohl c rn s 111n ul d interfere with 
the determination of odsor b ed 0 11 - . 1 n ~tia ll y formed ir o n 
s ulfide s are extreme l y suscept jh l c t u air oxidatio n (Berner, 
1964a), thu s great ca r e mu s t h e t;1kc11 : o preve nt ox i. clatio n. 
Al s o, hydro xyls are pre s ent in th e l"l' :1c tant iro n pha ses, 
94 
FeOOH and Fe(OH) . The product sulfide material must b e 
3
free from contamination with reactant material which would 
give a positive test for hydroxyl. 
Formation of iron s ulfide 
After reduction of goethite, the ferrou s hydrox ide 
dissolves to produce aqueou s ferrous ion s and hydroxide io1n5 s. 15
So lubility product calculations for Fe(OH)z (Ksp - 10? ? ; 
Baes and Mesmer, 1976) show that the equilibrium concentra-== 
3 15 The concentration of S 
tion of Fe+ 2 at pH 8 is 10- ? M. 3 
s pecies at these conditions (pH 8 and s;? 4 x 10 ? M) can 
be calculat ed from the expression 
s=T=  
where 1 ? 10 ? 7 ? 1 and Kz ? 10? 14 are th? first and second 
1 
di ss ociation constants for H2S respectively (Stum1n a nd Morgan, 
1970). These calcula tions showed that the concentration of 
s == i s 10- 7 ? 5 or 3.16 x 10- 8 M. The s olubility product of 
FeS has a value of 10- 10 ? 65 , or s i x orders of mag nitud e 
greater than the mo st so 1 uble s ulfide, amorphou s iron 
16 9 
s ulfide, with an ion activity product of 10 - ? (Doyl e, 
1968). As a re s ult of the di ssolution of Fe (OE) 2 , the so lu ? 
tion is s up ersaturated with respect to iron s ulfide and pre-
cipitation s hould occur rapidly. 
The rate of formation of i ro n s ulfide was studi e d 
by the initial rate method. QuaJratic eq uations were de? 
95 
t e rmined by a leas t squares re g re ssi on analysis, whi c h 
described th e concentration-time curves for acid ex tr act -
able s ul fide sulfur. Th e initial rate for ea ch reaction 
run was de termine d by the me thod descri be d 1n Appendix B. 
Log - lo g plots of the initial rates for th e formation 
of FeS vs. th e initial bi s ulfide ion co ncentration, the 
initi a l hydrogen ion concentration, and th e initial s urf ace 
area of goethit e, can be fo und in Figur es 22, 23, a nd 24. 
Th e slopes, and h e nce the reaction orders f or th ese spec i es, 
were found to be 0.97, 0.82 and 1.1 respective l y. Thi s 
res ult ed in a rate expression : 
d[FeS] /dt 
The rate constant, k, wa s calculated on th e ba s i s of 16 
2 
experime ntal runs to have a mean value of 31?10 M- 1 1 i - m-
. - 1 
min The rate s and rate c ons tant s for the 16 individu a l 
runs may be found in Tabl e 7 . 
A comp arison of the result s of thi s s tudy with tho se 
obtained in a previous kinetic s tudy by Rickard (1974), 
reveal s s ome di sag reement. Ri ckar d determined the overall 
reaction to b e 9/2 order: first order in goet hit e s urf ace 
area , 3/ 2 order in total s ulfid e co nc e nt ra tion, a nd s eco nd 
order in hydrog e n io n activity. Results of t h e pre se nt 
s tudy indicate a third orde r r eact ion, first order in eac h 
of the three re ac tant s : goe thit e s ur face area, tot a l i niti al 
s ulfide conc ent ration, a nd hyd roge n ion a c tiv ity. Co mpari s on 
96 
10- 5 .....- ---------------------, 
U) 
(1) 
w... 
I 
HS~ (M) 
1 
Fi gure 2 2 : Plot of Rate of Iron Sulfide Formation 
vs . Initial Bisulfide Ion Co nce ntra t ion 
91 
r--. 
~ 
?rl 
...s....._  t 
~ 
'-' 
.? 
.'.".."
6 
..d.._  10-
U) 
(1) 
1-L, 
'""d t l 
H+.  (M) 
l 
Figure 23 Plot of Rat e of Iron Sulfid e Pormation 
vs. Initi a l Hydro gen Ion Concentration 
98 
t t 
t t 
10 
A (m2/l) 
FeOOHi 
Fi gure 24 : Pl ot of Rat e of Iron Sul f id e form a tion 
v s . Initial Go e thit e Sur face Ar ea 
99 
Table 7 
Rate of FeS Formation 
82 
d[FeS]/dt = k [HS-]: 97 1 (H+):1 AFle.Ol OHi 
k = 31?10 M- l 1-l m- 2 min - l 
2 
d[FeS ]/ <lt =a+ bt + ct
mean 
% thio -
s ulfate 
b C k 
Run a 
2 . 70 10- 5 6.07 X 10 -
7 -2 . 26 X 10 - 10 23.0 36 
6 X 
10 - 5 
7 10-ll 
3.11 X 10 - -6.4 9 X 25.0 
10 
7 - 1.55 X 
6 -5 10- X 10 9 34.6 6 
9 - 3.14 
-1. 98 
X 10 - 4.5 8 X 
5 10- 7 
-2.85 10- 5.52 X - 1.10 
X 10 - ll 29 . 3 22 
17 X 
4 10- 7 -4 .4 6 10 - 11 X 38.5 22 
21 - 1. 01 X 10- 3.97 X 
5 10- 7 -1. 7 7 X 10 - 11 26 . 8 20 
22 2 .4 3 X 10- 5 . 55 X 
4 10- 6 -9. 34 10 -11 10- 2.83 X 40.4 
7 
24 -1. 15 X X 
5 7 1 7 
25 3.97 X 10- 9.19 X 
10- r = 0.99 ** 21. 3 
5 10- 6 -5. 64 X 10 - 10 48.l 4 
28 -8 . 37 X 10- 1. 86 X 
-
10- 4 
6 9 
5.54 X 10 - -1.5 3 X 10 42 . 6 
18 
29 -1.65 X 
5 10 - 6 -3 .7 8 X 10 - 10 40.9 4 
30 -9 . 76 X 10- 1.13 X 
4 6 10 - 10 36 .0 5 
31 - 1. 53 X 10- 3.24 X 
10- - 8 . 5 6 X 
5 10 - 7 1. 53 X 10-10 3 .4 3 
32 * 8 . 44 X 10- 1. 61 X 
10- 4 1. 81 X 10-
6 - 7 .6 7 X 10 - 10 26 . 1 22 
33 - 1.65 X 
4 10- 6 - 7 . 21 X 10-10 28.2 19 
34 - 1. 01 X 10- 1. 76 X 
3 10- 7 r = 0.91 ** 1 3 . 6 25 
35 - 1. 6 7 X 10 - 5.68 X 
5 10 - 7 6.99 r = 0. 99 
,~ * 17.l 
36 -5.07 X 10- X 
[i_tt ed 
"'run c onduct e d without stirr in g 
**lin e ar l y 
100 
of rate constant s from the two studie s is not meaningful, 
since the rate expressions are of different orders. 
The difference in the two rate expre ss ions for th e 
formation of iron sulfide can be explained by the s peci e 
that was measured to determine the rate of formation. 
Rickard (1974) measured the amount of OH- produced. These 
experimental values were then used in ma s s balance and 
equilibrium expressions to calculate the amount of product 
FeS. Hydroxyl is produced (or H+ cons umed) by s everal 
reactions in Rickard ' s mechanism: 
FeOOH + H+ + Fe+3 + 2 oH-
Fe+3 + 3/2 HS- + FeS + 1/2 S 0 + 3/2 H+ 
Thus, the rate of formation of OH is a mea s ur e of the 
entire mechanism - dissolution, reduction, and pr e cipita -
tion. In the present study, th e rate express ion for the 
dissolution and precipitation reactions we r e d e t e rmin e d 
separately; that is, Rickard' s rate e x pre ss ion wa s s epa -
rated into two parts in the present s tudy. 
As stated previously, th e number of s p ec i es in th e 
activated complex is give n by th e order o f th e rat e ex -
pression. Rickard ' s (1974) rate expre ss ion impli es an ac t i-
vated complex of 9/2 mol e cul es which could form by pr eequi -
libria prior to the rate determining s tep, th e di ss olution 
of goethite. The rate expre ss ion determine d in th e pr esent 
s tudy indicate s a three mol ec ul e ac tivat e d c ompl ex , of th e 
101 
form Fe OOH - HS - ? H+. 
Termolecular activated complexes us ually form by an 
eq uilibrium step prior to the rate determining step (Wilkins, 
1974). It is unu s ual tha t the initial s urfac e area of 
goe thit e is involved in the precipitation of the two dissol-
ved species Fe+ 2 ands=. Yet, if the protonation of the 
Fe(OI-I) 2 is considered to be the pre-equilibrium step, the 
effect of the surface area may be explained. The protona -
tion of Fe(OH)z at pH 7-8 does occur, since the IEP for 
ferrous hydroxide is 12 (Parks, 1965). After protonation 
the ferrous hydroxide dissolves to form either Fe+2 or 
FeOH+ (Baes and Mesmer, 1976; Kester e t a l., 1 975 ) depend-
ing upon which s tability constants are used . Fe+ 2 is 
indicated in this mechanism since the protonation of the 
Fe(OH)z s urface layer would yield Fe0zH! 2 or Fe+ 2 ? 2 HzO. 
The second, and rate determining step, in th e reaction i s : 
+2 
Fe + HS + FeS(s) 
Thi s reaction is in agreement with Pohl (1954). Hi s s tudy 
of dissolution and precipitat ion of meta l s ul fides indicat e d 
that HS - was the s ulfid e specie involved in the for mation 
of the activated complex, by the reaction s : 
(M ? xHzO)+Z + HS- so1+w  [MSH 
[MSH (x- l)HzO] fist MSJ, + r-r+ (x - l)r-r 2o 
102 
reactions and mechanisms (Berner, 196 4a . 
Pr evi? ously proposed ' 
either H S or S == as the reactant sulfur-
Rickard, 1974) used 2
containing species. 
The reaction for the formation of iron sulfide is 
initially an equilibrium reaction, since it involves the 
formation of small (S-20 A) ionic clusters of FeS (Walton ' 
19 67). Th ese clusters grow and decay until a critical 
cluster size is achieved. The critical cluster size i s 
defined as the 1aroest cluster which can exist before 
b 
spontaneous crystallization occurs (Walton, 1967) . 
Particle size study 
Sequential filtration studies wer e conducted at the 
ends of four reaction runs to determine the s iz e of th e 
submicron particles of iron sulfide in s olution. The re-
s ults of these measurements (Ta ble 8) showed that the pre -
dominant particles size of FeS was <0.01~ but a s i gnificant 
number of particles were betwe en 0.01 and 0.1? in diameter. 
Th i? s was demonstrated by the o.Ol? f1? ltrat1?o n step which 
removed %, IOO% and 30% of the particulate iron in runs 
30
29, 30, and 31, respectively. Run 32 was the only reaction 
run that was not stirred, and it was the only run in which 
the particles of FeS were all <0.01~ This would indicate 
that s tirring increases th? size of the particles of iron 
sulfide formed. th
These results were also s upport ed by e UV meas ur e-
ment s which showed meaningful changes in a bsorba nce only 
103 
Table 8 
Particle Size Study 
Filter [Sx_] T Absorbance Ab s orbance [Fe] [ Sy] 
Run Size ]J (M) 300 nm 370 nm ppm (rnM) 
29 0.8 1. 2 4 X 10- 1.264 0.812 6 6.36 
0.2 1. 2 X 10- 4 1.266 0.796 6 5.11 
0.1 1. 2 X 10-4 1.264 0.789 6 3.16 
0.1* Not Determined 
0.01? 1.3 X 10 - 4 1.125 0.633 4 3.60 
30 0.8 3 10- 5 X 0.537 0.417 5 0. 57 
0.2 2 X 10- 5 0.507 0.419 4 0.4 3 
0.1 4 X 10- 5 0.426 0.407 5 0.1 9 
0.1* 4 X 10- 5 0.426 0.354 4 0.19 
0.01?0 0.046 0.016 0 0.00 
4 
31 0.8 1. 3 X 10- 1.490 0.968 10 1. 86 
0.2 1. 3 X 10- 4 1.502 0.968 10 1. 66 
0.1 1. 3 X 10- 4 1.509 0.965 10 1. 32 
0.1* 1. 3 4 X 10 - 1 . 482 0.945 10 1 .0 3 4 
0.01? 1.1 X 10 - 1.222 0.748 7 0.80 
32 0.8 6 5 X 10- 0.411 0.175 2 3. 1 9 
0.2 6 X 10 - 5 0.401 0.169 2 2.72 
0.1 6 X 10- 5 0.401 0.169 2 2.08 
10- 5 0.1* 6 X 0.389 0.150 2 2.01 5 
0.01? 6 X 10- 0.311 0.080 2 1. 8 2 
0.1* filtered t hroug h 0.1? filter twi ce 
0.01? Sch l eicher and Schuell nit roce llul ose membrane fi l ter 
All other filter s polycarbonate 
104 
Table 9 
"Di sso lved Iron" Concentration 
(mM) 
Time 0 'v 30 'v 2 0 0 'v 3 0 0 
'v 4 0 0 'v 7 5 0 'v l450 
(min) - - --
Run 
0.03 0.03 
6 
7 0.01 0.06 0.03 
0.03 0.01 0.03 
9 0.03 
0.02 0.03 
0.01 
11 0.01 0.01 
0.03 0 .0 3 0.06 
17 
0.04 
21 
0.08 
22 
0.02 0.02 0.03 0.03 24 
0.03 0.08 
25 0.01 
0.03 
28 0.01 
0.02 0.06 0.12 
29 0.01 0.02 
0.04 0.10 0.11 
0.01 0.01 0.01 0.01 30 0.01 0.01 
32 
l Fe A z] = dis so 1 v e d Fe + part i cu 1 at e Fe ( < 0 . 1 ?) 
Exact times ?or measureme nts may be fo und in 
Appendix A. 
105 
af t er fi ltration with 0.01? fi lters . The decrea se i n ab-
sorbanc e af t er filtration was attributed to the removal of 
submi cro n part i c l es wh ich scattered the UV radiation, and 
therefore increased th e absorbance. This ver i fied, at lea s t 
qualitative ly, that th e unus ua l polysulfide s pectra were 
ca use d by scattering from s ubmicron particles . 
The iron concentrations in a ll four fi ltration r un s 
Th er mo -
were 2- lOppm after f iltration t hrou gh a 0.2? filter. 
dynamic ca lculations by Brooks et a l . (1968) and Pre se J y et 
a l. (1972) indicated th a t the conce ntrati on of di ss olved 
iron at IO- 3M tot a l s ulfide and pH 8 s hould b e 0.03 pph. 
Th e fi ltration study r esult s s how that the solutio n i n the 
present study is not in thermodynamic equilibrium, o r t hat 
Tradition -
th e definition of di sso lved spec i es is i ncorrect . 
ally, the definition of dissolved species was a ny materia l 
that passe d through a O. 45 ? fi lt er . Yet the result 0? t h i s 
s tudy and others (Kennedy et al., 1974) s howed that t he r e 
are signi f ic ant l eve l s of iron-containing particl es at <0.45 ?. 
'fhe interpretation of co nc entrat io n s of di sso l ved s pe cie s 
( by th e tr ad itional definiti on ) fro m inter s titial waters 0? 
a no x ic or oxic sediment s could be se riou s l y in error if t h i s 
i s no t c onsi dere d . 
Di s s o 1 v e d i r o n ( < 0 . 4 5 y) con c en tr at ion s i n i n t e r -
stitial waters of a no xic sediment s , hav e bee n meas ur ed as 
Th e concentration s 
high as O.  2 ppm (Pr es e 1 y e t a 1 . , l 9 7 2) . 
in the reaction runs (Tab l e 9; 0.02 m~! = 1 ppm Fe ) and in 
106 
the particle size studi es indicate a co ncentration of iron 
as much as an order of ma gnitude greater than that mea s ur e d 
1n natural sediments. If one co n siders that the reaction 
conditions 1n this study are ideal for th e for mation of 
iron s ulfide, then the agreement in concentrations betwee n 
natura l interstitial water s and this study is quite good. 
107 
CONCLUSIONS 
The rates and rate expressions for reaction of FeOOH 
and HS- leave little room for interpretation, since th e y 
are based solely on experimentally measured quantiti es , 
However, the mechanism for the reaction i s open to seve ral 
interpretations . This is true because the development of 
any mechanism is based on both experimental data and chemi? 
cal insight. "Indeed it is impossible to prove any s ing le 
mechanism . However , much favorable data may be ama ssed 
for a mechanism so that one can be fairly certain of vali? 
dity" (Wilkins, 1974) . 
The mechanism for the formation of FeS involve s 
dissolution, reduction, and precipitation ste ps . A comp a ri-
son of morphology and crystallinity of reactant s olids and 
product solids (Rickard, 1974) indicates that a me chani s m 
includes a dissolution step, while chemical princ iple s in ? 
dicate that a precipitation reaction i s nec es s a ry. In 
addition, there is no question of the oxidation s tat e o f 
iron in iron sulfides; hence, a reduction re action i s a l s o 
required. The sequence of these processes i s defined in 
the mechanism. The reduced iron - time a nd pr ecipitat ed 
s ulfide-time data indicated that the r e du c tion r eac tion 
preceded the precipitation reaction. However, th e c ompo ? 
sition of the activated compl ex for th e r eduction r e a c tion 
was based on the a ss umption that th e mec hani s m for th e 
108 
Fi gure 25 
Detailed Reaction Mechanism 
1) Protonation of s urfac e adsorption site 
'-...,. 1 / H 
Fe 
/ 1 "" 
0 0 
I / 
/Fe + H20 
6' \_OH 
"/"'y' ,/ 
2 n hancre o f lL . w It OIi l. f ixed layer 
OI-1 OIi 
"' "'I/ F'e /  Fe 
/ I 
II + 1-r /I "" ""' + Ji 0 Oil+ Sil + Ol l -"'- ' / i / Fe Fo 
/ 1 '\_ / I ""
0 OH QI  /. 0
 11 
"'-1~ / Fe 
/I "- / I ~ 
C ont. J 
109 
Figur e 25 (cont.) 
3) Consecutive reduction of iron 
~ I OH OI-1 "" I / 
Fe/ Fe +2 
/10 "" OH+ sr-r- / 0I " "O H+ H+ + so 
ie/ ~ I/ Fe +2 
/6 ""OI-l / 6 ""OH 
te/ !e/ 
/ I "" /I ~ 
4) Protonation of Fe(OH) 2 s urfac e layer 
I / OH I ,("OH z 
+2 Fe +2 
Ae,~""  + /1 ~ + ""'0 OH 2 ~ o oH 2 I 
Fe + 2 + 2 H+ Fe +
1  
/1 "" /\ I0 ~ + 0 OH OH 2 
l _(z I / 
Fe Fe +2 
/ I ""- /! 
S)D i s s olution of Fe (OI-1); and pr ecipitation of FcS 
) l"- e + 2 s ( aq) + 1-L --4) FellS+~FcS ( ) + 1-l+ 5
s l ow .F a s t 
110 
r eduction reaction involved a r a te de t e rmining s t ep. Th is 
ha d to be inferre d since little exp e rimental data was av a i l-
abl e for the reduction reaction. Thus severa l othe r me cha -
nisms may exist. One such mechani sm would involve th e r e-
plac ement of a hydroxyl in the coordination s pher e of Fe in 
goethite by HS-. However, ther e i s no da ta to support 
thi s mechani s m. 
The inclusion of the dis s olution proc ess i n the 
mechani s m wa s inferred on the bas i s of the f ollowing argu-
If di ssolution had pr ec eded r eduction, c on s ide r abl e 
ment. 
quantiti es of aqueous ferric speci es woul d have bee n pr o -
duced. The s e ferric s peci es would then have be en r educ ed 
rapidly to an aqueous f e rrous sp eci es , as ind ica t ed by t he 
hi gh conc entration of sulfide ox idation produc t s, in t he 
f i. rs t s everal minut es of the reac ti.o n. Then these ferro u s 
ions could have r eacted r apidly with di ss olve d s ul f ide t o 
However, the r e i s a c ons i de r abl e time 
for m i? ron s ulfide. 
interval between the r eduction reaction and th e a ppearanc e 
of the precipitat ed iron s ulfide . Thu s , the dis s olu t i on 
reaction would have had to f ollow the r educ tion react i on. 
In addition, thermodynamic c a lculations i ndica t e that ferric 
s peci es would not be expected a t these pH, Eh, and pS 
conditions . 
The r eduction r eaction was infe r re d t o occu r a s a 
the r educ ti on of s ur face iron, and t he 
two s tep proc ess : 
r eduction of the r ema inder of fer ri c iron aft e r di ssol utio n 
of the s urfac e f errous 1ayer , The rate of the r e ductio n 
111 
reaction of the subsurface iron wa s interpr e t e d a s b e ing 
a measure of the rate of dis s olution of the s urfac e layer. 
A comparison of the rate of reduction during thi s se c ond 
reduction process with the initial rate of iron s ulfide for-
mation showe d that the pr e cipitation reaction wa s appro x i-
mately as fa s t as the dissolution reaction. Thu s it wa s 
inferred that the dissolution s tep was the rat e d e t e rmining 
step for the overall reaction, sinc e the initial rat e of 
reduction was shown to be hi gh. 
The overall mechani s m for th e form a tion o f iron 
s ulfide was determined to be a s outlined in Fi gur e 2 5 . 
The result s of this study are believed to b e s i g ni -
ficant in the understanding of two geological probl e ms . Th e 
initial rate of iron sulfide formation was found to range 
from 1.6 x 10- 7 (Run 32) to 5.54 x 10- 6 (Run 29) M/min. 
Conversion of this data to unit s comparabl e to Be rner ' s 
(1972) would yield rates of 2.7 - 58.5 mg sulfur/ c m2 - y r. 
Berner (1972) reported a rang e o f s ulfur uptak e by ma rine 
sediments of .05-11.3 mg/cm 2-yr. Th e high es t v a lue was 
found in a polluted Maine fjord. Thu s the lowes t ra t es in 
the present study are well with i n the r a nge o f Be rner ' s 
data. From his data, Berner ca l c ulat e d tha t th e Bl ac k Sea 
could remove one me gaton of s ul f ur p e r y e ar, which c ombin e d 
with the s i x megatons of sulfur uptake for th e h e mip e la g i c 
s ediments could not account fo r th e input of s ul f ur b y 
s tream load. Th e u se of the l a r ges t valu e from thi s s tud y 
(58.5 ms /cm2-yr) a nd a s ur face area of a no x i c s e d i me nt 
112 
c ompar a bl e to th e Black Se a ( 42 3 ,000 krn 2) could remove 248 
meg atons of s ul fur per year. Thi s could account for the 
129 megatons of sulfur per y e ar added by s tream input. How -
e ver, thi s higher rate of sulfur uptake would not b e likely . 
Rather, the value of 11.3 mg/cm 2-yr would be more f eas ibl e. 
Ca lculatio n s using this latter valu e show that anoxic sedi-
me nts with a sulfur upt ake of 11.3 mg /cm 2 - yr could remove 
45 megatons of sulfur per year, or 35% of th e oceani c s ulfur 
budget. Iron s ul fide formation could be i mportant i n th e 
s ulfur budge t assuming the rat e of iron s ulfide for mation in 
es tuaries and deltaic sediments ha s thi s v a lu e, a nd s ur face 
areas of anox ic se diments are comp arab l e to th e Black Sea . 
Another significant r es ult o f thi s s tudy wa s the 
for mation of eleme ntal sulfur in anoxic sedim e nt s by a c h e mi -
cal reaction. The postulated me chani s m for the fo rm at ion 
of iron rnonosulfide would yield elemental s ul f ur in a ratio 
of one part s ulfur to two part s iron s ulfid e . Thi s s ul f ur 
co uld be used to convert the iron mono s ul f ide to pyrite, 
the thermodynamic a lly stab l e iron s ulfide . 
Previous s tudie s of trace element co n centrations in 
interstitial water s of a no x i c se dim e nt s h ave b ee n u se <l to 
infer a solid phase controlling the s olubility of th e 1neta ] 
and h e nce the pre sence of that so lid in the sed i me nt. Tl1 i s 
s tudy s how e d that the di s tinct ion betwee n the dis s olved a nd 
particulat e phases is not as s impl e as pr ev i o u sly thought. 
Particle size mea s u rements s how e d t h a t there we re measur a bl e 
c onc e ntr a tion s of p a rti c ulat e FeS present, in the fi l tered 
1 1 3 
s olution, e ven after filt e ring through a 0.01 ? f il ter . Th e s e 
conc e ntrations were in the ppm r a n ge , whil e equilibrium 
c oncentr a tions from the Ks p indi ca t e d tha t the d is s o l v e d 
i r on conc e ntrations s hould b e in th e ppb r a n ge . Ir o n s ul f i de 
mu s t b e c ontrolling the s olubility o f Fe in th e p re sent 
s tudy, s inc e it i s the only iron ph as e present a t t h e e n d of 
th e exp e rim e nt. Thus futur e s tud i es o f the int er s t i t i a l 
wa t er che mi s try mu s t cons ide r th e p r e se n ce of non -f il terabl e 
p a rticl es in the s olution. 
All r esea r c h s hould b e o pe n e nde d, t h a t i s , i t s h o ul d 
s u gge s t sev e ral n e w a r e a s to be i nv e s ti g a t e d . On e p o ss ibl e 
a r e a f or f uture inve s ti ga tion i s th e stu dy of the effec t of 
ioni c s tr e n g th on the re ac tion ra t e s . Th e t hi c k n ess of the 
Gouy laye r, a nd h e nc e the c l o s e s t dis t a n ce of a pproa c h of 
b is ul f ide s p e ci es to th e s ur face of t h e iron r eactan t phase , 
i s r e duce d by a n inc r ease in i on ic s t r e n g th. Thu s it wo uld 
b e ex p ec ted tha t cha n ges in i oni c s tre n g th wo ul d aff ec t t h e 
r a t e of iron s ul f ide f orm a ti o n. Th i s i s impor t a n t in es t u -
a rine environme nt s s inc e the re i s co n s i d e r ab l e dif f e r e n ce 
in ionic s tr e n g th as one pro c e e d s d own t h e es tuary to th e 
op e n oc ean. 
Anoth e r po ss ibl e s tudy c ould i nvo lve t h e d e t e r mina-
tion of the r a t e o f the re du c t io n r e a c tio n. Th e i n iti a l 
r edu c tion r eac t i on o ccur r e d rapid l y , a n d a s a r es ul t, f e w 
meas u reme nt s were t a k e n dur i n g t h is time p e riotl , s o t h a t 
only qua litative s t a t eme n t s coul d b e mad e wit h r es p e ct to 
th e ra t e a n d mec h a ni s m of t h e r e d uction r e a c tion. Prior 
114 
to further studies, new dev e lopments in the ana l yti c al t ec h-
niques for the measurement of sulfide oxidation product s 
must be made in order to decrease the time requir e d to make 
these mea s urements. 
115 
APPENDIX A 
Data Table s for Reac tion Run s 
Thes e tables li s t the time eac h series of meas ur e ment s 
was taken (in minutes), the pH measurement at that time, 
the concentrations of the variou s s pecies meas ur ed (in rr~), 
and the calculated values for released OH- and reduced iron 
(electron balance) (also in rnM). Initial condition s are 
g iven for pH, total s ulfide (in mM), total iron (in mM), and 
the surface area of the reactant iron spec i es (in m2/l). 
Run 6 pHi = 7.554 
s= = 4.50 
T. 
l 
Fer.= 1.12, 6.52 m2/ l 
l 
Reduced 
Time pH +~OH - s= s= S203 so s= 4 T s= s= 5 C p Iron 
- -
I 
0 7.561 4.27 0.22 
27 7 .SS4 0. OS 0.13 4.51 4.38 0.68 '\ 
19 8 7.S24 0.01 0.01 o.os 0.24 4.54 4.07 0.15 1. 02 
386 7 .S40 0.06 0.01 0.01 0.06 0.18 4. 31 3.80 0.21 0.99 
73S 7 .SSO 0.11 0.02 0.01 0.07 0.07 4.07 3.84 0.36 0.94 
1449 7 .S63 0.27 0.03 0.02 0.07 0.09 3.64 3.19 0.43 1. 09 
2907 7.S64 0.28 0.04 0.02 0.07 0.09 3.87 3. 21 0.57 1.16 
I-' 
I-' 
?' 
Run 7 pHi = 7 . 9 74 
s= 
T. = 4 . 65 
1 2 
F e 1 _ = 1 . 1 2, 6. 5 2 m / 1 
1 
Reduced 
T i me pH +t:.OH- s= s= so s= 4 5 Sz03 ST s=C  p I r on 
-
0 7 . 953 0 . 00 4 . 66 0 . 04 
19 7 . 9 74 0 . 01 4 . 60 4 . 4 0 0. 1 0 
1 98 7 .9 7 3 0 . 01 0 . 01 0.00 0 . 00 0 . 08 4.2 5 4.30 0 . 2 7 
3 76 7 . 95 9 0 . 01 0.00 0 . 01 0 .11 4 . 3 7 4 . 31 0 . 09 0. 4 6 
742 7 . 9 59 0.01 0.0 2 0 . 01 0 . 0 2 0. 1 7 4 . 23 4.09 0 . 18 0 . 7 0 
1 4 51 7 . 960 0 . 0 1 0 .0 2 0 . 01 0 . 01 0 . 27 4 . 46 4. 1 2 0.33 0.88 
I-' 
I-' 
--..J 
Run 9 pHi = 7 . 583 
s = = 4.00 
Ti 
Fe~' = 5 . 62, 32.6 m2  /1 
1 
Reduced 
Time pH +L'IOH s = s = S203 so s = s= s= 4 5 T C p Iron 
-
0 7.506 0 . 02 4 . 93 0.18 
21 7.583 0 . 01 0.00 0 . 02 0.25 4 . 02 4.16 0 . 10 0. 7 2 
214 7 . 598 0 . 32 0.04 0 . 02 0.13 4.05 3.06 0. 79 0.61 
370 7.6 47 1.00 0.04 0 . 03 0 . 02 0.35 3.97 2.36 1. 36 1. 39 
72 5 7 .6 6 3 1. 32 0 . 07 0.05 0 . 04 0.2 7 3.23 1. 29 2.24 1. 61 
1455 7 . 708 1. 98 0.03 0.03 0.03 0.44 3.6 7 0 . 55 2.36 1. 54 
f-' 
f-' 
00 
Run 11 pHi = 7.58 
s= = 4.30 
T1?  
FeT? = 1.19 , 14 . 0 m2/l 
1 
Reduced 
Time pH +fl OH - S4 s= Sz03 so s= s= s= 5 T C p Iron ii 
I 
0 7.580 4.56 
24 7.689 0.03 0.02 0.03 0.94 3.81 3.11 1.17 2.45 
193 7.683 0.03 0 .0 3 0.03 1.05 4.10 3.03 1. 40 2 . 73 
383 7 .668 0.03 0.03 0.04 1. 06 4.2 1 3.00 1.46 2.86 
7 20 7 .656 0.03 0.03 0.02 1.00 3.58 3.02 1.13 2. 56 
1366 7 . 70 2 0.04 0.02 0.03 1. 90 3.42 2.88 1. 03 2.44 
\ 
I 
I-' 
I-' 
\0 
Run 12 pHi = 8.Sl 
s = = 4 . 7 3 
?_ 
Fet = 1 . 19, 14.0 m2;1 
1 
Reduced. 
Time pH +~ OH s = s =s  S203 so s4 T=  s = s =p  C Iron 
--
0 8.51 S . 03 
29 8.SS 0.04 0.02 0.04 4.3 7 3 . 38 0 . 96 0.70 
19 7 8 . SS 0 . 04 0.02 0 . 04 1. OS 4 . 10 3 . 1 7 1.13 2 . 81 
379 8 . 57 0.04 0.02 0.05 0.68 4 . 23 3 . 09 1.00 2 . 40 
727 8.53 0 . 04 0.02 o.os 0.38 3.93 3 . 08 1.00 1. 54 
1 461 8 .53 0.04 0 .0 2 o.os 0.44 4. 7 9 3.06 l . Sl 1.69 
I-' 
N 
0 
Run 1 7 pHi = 7 .605 
s= = 4 .1 6 
T-1 
FeT. = S .1 0, 5.45 m2/l 
1 
Reduced 
Time pH + ~OH - s= s= so s= s= s= 
4 5 s o; T C p Iron 2
-
0 7 . 467 5.54 
30 7 . 605 0 . 30 0.01 0 .00 4 . 20 0. 1 0 
190 7 .5 85 0 . 02 0 . 01 0 . 09 4.68 0.08 0.40 
363 7 . 618 0.44 0.04 0 . 02 0.1 8 4 . 36 0 .16 0 . 71 
71 9 7 . 581 0 . 06 0 . 03 0.04 0.18 4 . 23 0 . 37 1. 28 
14 44 7 . 620 0 .49 0 .0 7 0 . 03 0 . 05 0 . 49 4 .1 7 0. 74 2 .11 
f-J 
N 
f-J 
l 
Run 21 pH-1 = 7 . 5 29 
Si. = 4 . 80 
1 ,., 
Fe _ = 2.36, 2 .51 mL / 1 1
1 
Reduced 
Time pH + ?DH - s= 4 s=5  S20 3 so s= s=T C  sp=  Iron 
--
0 7.486 5.56 
33 7 .5 29 0.01 0 . 00 0.02 0.15 4.82 4 .6 4 0.54 
39 0 7 .5 27 0.0 2 0.01 4 .55 4 . 45 0.05 
72 5 7 .515 0.02 0 .01 4.80 4.25 0. 1 6 
1 3 51 7 .503 0 . 03 0.0 2 0.03 0. 28 4.94 4 . 44 0.35 1. 09 
21 68 7 . 4 97 0 . 03 0 . 0 2 4.73 4.02 0.56 
2827 7 . 51 7 0 . 04 0.0 2 0.02 4.60 4 . 09 0 . 66 
36 1 2 7 . 49 3 
42'1-5 7 .55 2 0 . 04 0 . 0 3 0 . 04 0 .5 7 4 .60 3 . 60 0. 78 1. 86 
t-' 
N 
N 
Run 22 pHi = 7 . 736 
s= = 4 . 90 
T. 
Fe1T  -- 1. 04, 6.34 m 2/ 1 
1 
Reduced 
Time pH +6 OH - s= s= S203 so 4 s= s= s5 T C p=  I ron 
--
0 7 . 71 6 4.99 
26 7 . 736 0 . 27 0 . 01 0 . 00 0 . 01 0 . 0 7 4.89 4.76 0.04 0 . 30 
184 7 . 7 26 0 . 1 4 0 . 01 0.00 0 . 01 0.12 5 . 34 4 . 78 0.1 2 0 . 44 
34 8 7 . 7 59 0.68 0.02 0.01 0 . 02 0.1 2 4.1 7 3.82 0.23 0.6 1 
81 4 7 . 7 53 0 . 55 0.03 0.02 0.02 0. 2 5 4.86 4.30 0.46 0.9 7 
1503 7 . 78 1 0.99 0 . 03 0 . 02 0 . 03 0.54 4.24 3.92 0. 8 2 1. 70 
30 1 0 7 . 79 5 1. 20 0 . 0 5 0 . 03 0 . 03 0. 72 4.79 3.91 1. 40 2 .20 
I-' 
N 
vs 
Run 24 pH.= 7 . 564 
1 
s~ = 4.7o 
Fe\ -- 2.62, 15. 2 m2  / 1 
Reduced 
Time pH +6 OH- s = 4 S5 S203 so s=T  s = C Sp Iron 
-
0 7 .554 0.01 5.16 0.08 
45 7 .564 0.01 0.00 0 . 01 0.12 4.72 4.52 0.06 0.42 
90 7 .574 
221 7 .5 45 0.04 0.02 0.01 0 . 46 4 . 70 4.44 0 . 46 1.14 
375 7 .5 76 0 . 23 0 .03 0.02 0.03 0.46 4.59 3.91 0. 72 1. 44 
74 0 7 .604 0 . 6 1 0.04 0.03 0.02 0 .56 4 .69 3.54 1. 53 1. 76 
1 756 7 . 634 1. 06 0 . 04 0.04 0.04 1. 1 9 4.66 2 .84 1. 9 7 3 . 28 
I-' 
N 
+:s 
Run 25 pH. == 7 . 3 3 2 
1 
s == 
T. == 4 . 80 
1 
FeT == 1.05, 6.52 m2/ l 
1 
Reduced 
Time pH +6 OH- s == s == S203 so 5 sT == s == s=4  C p  Iron 
- --
0 7.329 5.26 
19 7 .33 2 0 .0 0 0.00 0 . 00 4.78 4.53 0.0 2 0.08 
105 7 .340 0 .10 0 .01 0.01 4 . 36 0.27 0 .1 0 
192 7 . 31 7 0.02 0.01 0.01 0 .1 2 4.63 4 .32 0 .5 4 
242 7 .3 44 
280 7 . 328 0 . 02 0 . 01 4 . 05 0 . 24 
396 7 . 368 0 . 49 0 . 02 0.02 0 . 02 0 .1 8 4 .1 0 3 . 69 0 . 33 0.82 
76 2 7 . 366 0.55 0.03 0 . 03 0 . 02 0.39 3.44 3 . 26 0 . 75 1. 36 
1452 7 . 388 0 . 70 0 . 05 0.04 0 . 04 0 . 43 4 . 26 3.54 1. 38 1. 72 
f-J 
N 
u, 
Run 28 pHi = 7 .5 82 
ST? = 2 . 60 
1 
Fe1_ = 2.62, 15. 2 m
2/l 
1 
o; Reduced Time pH +ti OH - s= S5 s so s= s= s= Iron 4 2 T C p 
-
0 7 .52 4 3.10 
35 7.582 0.19 2.56 2.31 0.03 0 . 38 
140 7 . 563 0 . 00 0 . 00 0.26 2 . 95 2 .5 8 0.15 0 . 58 
255 7 . 5 7 0 0.03 0 . 02 0 . 01 0.00 0 . 23 2.82 2 . 38 0.32 0 . 69 
425 7.57 1 0.08 0.02 0 .0 2 0 . 38 2 . 75 2 . 02 0.54 1. 0 2 
790 7 . 588 0.21 0.03 0.03 0 . 01 0.56 3.1 1 1. 56 1.1 0 1. 6 2 
152 7 7 . 632 0. 77 0.04 0 . 0 4 0.01 0 . 68 2 . 7 5 1. 21 1. 43 1. 98 
1--' 
N 
?' 
~ 
Run 29 pHi = 7 .5 40 
ST. = 8 . 70 
l 
Fe 21. = 2.60, 15. 2 m / l l 
Reduced 
T i me pH +6. 0H S4 s= S203 so s= s= s= 5 T C p Iron 
-
0 7 .5 28 9.4 5 
30 7 .5 40 0.0 1 0 . 00 0.30 8 . 71 8 . 7 4 0 . 05 0.71 
1 60 7 . 566 0 . 40 0 . 03 0 . 02 0 . 01 0 . 36 9 .1 6 8.28 0.55 1.11 
323 7 .5 95 0 . 82 0 . 05 0.02 0 . 03 0.35 9 .5 8 7 .86 1. 58 1.40 
4 76 7 . 580 0 . 7 5 0 . 06 0.03 0.08 0 . 39 8 . 75 6 . 89 2 .08 1 . 98 
755 7 .5 85 0.99 0 . 0 7 0 . 04 0 . 06 0 . 36 5.53 1. 99 1.9 4 
13 7 5 7 . 69 5 2.61 0 . 06 0 . 0 5 0 . 06 1. 28 8 . 85 3 . 71 4 . 55 3.82 
;-., 
N 
---J 
Run 30 pH . == 7 . 5 5 5 
1 
ST ? == 1. 8 0 
1 
FeT == 2 .6 0, 15. 2 m2; 1 
1 
Re duced 
Time pH +6 OH- s~ s::: S203 so s::: s::: s::: 5 T C p Iron 
-
0 7 .516 0 . 00 2 .11 0.04 
39 7 . 555 0 . 00 0 . 00 0 . 00 0 . 4 5 1. 77 1. 86 0 . 99 
1 75 7 .5 68 0 . 02 0 . 01 0.0 1 0 . 00 0.55 1. 8 5 1. 7 5 0 .0 6 1. 22 
336 7 .51 8 0 . 03 0 . 01 0 . 01 0 .0 1 0 . 5 2 1. 94 1. 51 0 . 23 1. 29 
403 7 . 55 1 
465 7 . 55 1 0 . 04 0.0 1 0.0 1 0 . 02 0. 63 1. 80 1. 37 0.44 1. 59 
47 0 7 . 545 
60 7 . 529 0 . 09 0 . 02 0 . 02 0 . 02 0 . 56 1. 7 2 1.11 0 . 49 1. 53 
15 00 7 . 536 0 . 14 0 . 02 0 . 03 0.0 3 0 . 85 1. 70 0 . 78 0 . 75 2 . 32 
f--' 
N 
co 
Run 31 pHi = 7 .5 42 
s = = 4. 76 
T1?  
Fe~ = 3.17, 1 8 .5 m2/l 
1 
Reduced 
Time pH +~OH- S4 s = S203 so s = s = s = 5 T p Iron C 
-
0 7 .501 0.02 5. 40 0 . 16 
27 7 .542 0 .00 0.00 0.01 0 .1 7 4 . 77 4 . 74 0.50 
1 78 7 .5 64 0.37 0.02 0.01 0.01 0 . 3 7 4.49 3.89 0.29 1. 0 2 
333 7 .5 77 0.51 0.02 0.02 0.02 0.54 4 .8 2 4.01 1. OS 1. so 
475 7 .5 68 0.48 0 . 03 0 . 02 0.04 0.70 4 . 78 3.46 1.1 S 2.12 
605 7 .5 93 0 . 82 0 . 04 0 . 03 0 . 03 0.46 4 .62 3 .15 1. 4 3 1. 64 
1 476 7 . 620 1. 2 7 0.05 0.06 0.03 1. 20 4.44 2.05 2 . 77 3.42 
!-' 
N 
<.D 
~ 
Run 3 2 pHi = 7 . 5 72 
s;. = 4. 7 s 
1 - - 2 
Fe~ - 1. 94, 11. 5 m / 1 
1 
Reduc ed 
T i me pH +t.OH- s= s=4 s  S203 so s= s= s= T C p Iron 
---
0 7 .5 26 0 . 00 5. 20 0 . 0 1 
so 7 .5 72 0 . 3 7 0 . 02 0 . 01 0 . 01 4.26 0 . 26 
1 70 7 . 603 0 . 76 0.0 2 0 . 01 0 . 01 0 .1 2 4 . 45 4 .1 4 0 . 52 
31 6 7 . 564 0 . 0 2 0 . 0 1 0 . 0 2 0 .1 8 4 . 38 4 .11 0 .1 5 0 . 7 3 
46 5 7 . 6 20 0 . 9 7 0 . 0 2 0 . 01 0 . 01 0 . 25 4 . 82 4 . 0 7 0 . 20 0 . 82 
6 15 7 . 6 09 0 . 87 0 . 0 3 0 . 02 0 . 02 0 .1 9 4. 1 8 3 . 82 0 . 24 0 . 80 
1 46 5 7 . 61 8 1. 0 2 0 . 0 5 0 . 02 0 . 00 0 .1 0 4 .1 2 3 . 39 0 . 65 0 . 66 
15 56* 7 . 6 50 1. 4 7 0 . 04 0 . 0 3 0 . 02 0 . 50 4 . 22 3 .1 0 0 . 8 2 1. 6 2 
*S t irred a t 11. S hrs ( 690 mi n ) 
I-' 
lN 
0 
Run 33 pHi = 7 .050 
Sr. = 4 .50 
1 
Fe~= 1.06, 6 . 52 m2/1 
1 
Reduced 
Time pH +6 OI--r- s = s = S203 so s = s= s = 4 5 T C p Iron 
-
0 7 .0 70 0 . 00 0.00 0.02 4.90 0.16 
40 7 .050 0.00 0 . 00 0.02 4.54 4.48 0.2 1 
1 70 7 .0 74 0.10 0.01 0.01 0.03 0.17 4.31 4.51 0. 1 5 0.72 
310 7 .0 95 0 . 39 0.02 0.01 0.04 0 . 32 4.21 3.89 0.30 1. 1 7 
473 7 . 097 0 .4 7 0 . 02 0.02 0.06 0.30 4.50 3 .89 0.53 1. 4 0 
600 7 .11 4 0 .5 6 0 .03 0.02 0.06 0.32 4 . 25 3. 82 0.69 1. 49 
1450 7. 136 0 . 69 0.04 0 . 04 0.0 7 0 . 71 4.41 3.34 0.91 2 . 51 
>-' 
(..,.J 
>-' 
Run 34 pH . = 7 .1 73 
l 
s= 
T . = S . 1 7 
l 
FeT_ = 1. 06 , 6 .5 2 m2  / 1 
l 
Reduc ed. 
T i me pH +~OH s= s 5 S 203 so s= s= s= 4 T C p Iron 
--
0 7 . 1 86 5 . 41 0 . 00 
33 7 .1 73 0 . 00 0 . 00 0. 00 0.1 7 4 . 89 5.1 2 0 . 39 
16 1 7 . 1 80 0 . 2S 0 . 01 0.0 1 0 . 02 0 . 2 5 S . 07 4 . 41 0.1 8 0 . 7 S 
31 2 7 . 22S O. S6 0 . 02 0 . 02 0. 03 0 . 28 4 . 84 4 . 43 0 . 38 1.0 6 
4SS 7 . 237 0.74 0 . 02 0 . 02 0 . 04 0 . 30 4.8 1 4 . 06 0. 49 1. 20 
61 0 7 . 21 7 0.63 0 . 03 0 . 02 0 . 04 0 . 4S 4 . 69 3 . 94 0 . 7 S 1. S3 
1331 7 . 244 0 . 91 0 . 04 0 . 03 o.os 0 . 29 4 . 42 3 . 56 0 . 97 1 . 47 
\-I 
v,l 
N 
Run 35 pHi = 7 . 4 1 9 
s= 
T . = 5.50 
1 2 
FeT = 1.06 , 6 .5 2 m / 1 
1 
Reduced 
Time pH +L'i OH s= s= so s= s= s= 4 5 s2o; T C p I ron 
-
0 7 . 4 10 0.0 2 5. 72 0 .16 
30 7 .4 1 9 0.03 0. 22 5.52 5.40 0. 69 
1 66 7 . 433 0 . 20 0. 00 0 . 00 0 . 04 0 . 20 5. 4 5 5.05 0 .03 0.83 
316 7 . 4 5 2 0 . 40 0 . 01 0 . 00 0 . 0 5 5. 29 4. 98 0 .1 0 
4 6 0 7 . 434 0 . 01 0 . 01 0 . 0 5 0 . 3 5 5. 32 5. 01 0 . 23 1. 34 
48 1 7 . 4 7 2 0 . 02 0. 01 
60 5 7 . 447 0 . 42 0 . 02 0.0 1 0 . 06 0 . 36 5. 28 4 . 94 0 . 4 5 1. 4 5 
1 4 50 7 . 498 1. 0 2 0 . 0 5 0 . 0 3 0 . 09 0 . 23 4 . 77 4 . 24 0 . 76 1. 6 7 
1--' 
vs 
vs 
Run 36 pl-Ii = 7 .400 
= 
ST? = 5.40 
1 
FeT = 1.06, 6.52 m2; 1 
1 
Reduced 
Time pH +6 0H S4 s =s  S203 so s = s = s = Iron T C p 
- -
0 7 . 400 0.01 5.75 
35 7 . 400 0 . 01 0 . 00 0.0 2 5.3 8 4.96 
1 71 7 . 41 1 0 .1 2 0.0 1 0 . 01 0 . 03 5 .1 0 5.31 0 . 05 
315 7 . 41 8 0.22 0 . 02 0 . 01 0.0 4 5.23 5.14 0 .1 3 
46 5 7 . 42 5 0 . 38 0 . 02 0 . 0 1 0 . 04 4 .9 0 4 .6 7 0.30 
602 7 . 465 0.81 0 . 03 0.02 0 . 06 5. 16 4 .5 6 0 . 43 
1 22 5 7 . 4 79 1. 00 0 . 05 0 . 03 0 . 06 4 .66 4 . 3 3 0 . 78 
I-' 
l..N,.  
I 
1 3 5 
APPE NDI X B 
Initi a l Rat e Me tho d 
The kine tic s for th e r e a c tion of aqueou s bi s ulfid e 
ion a nd goethit e wa s s tudied by th e initi a l rat e me tho d . 
The rat e l a w for thi s r eac tion ca n be cons ide r ed to be : 
wh e r e Ri i s the rate o f f or ma tion of iron s ul f id e , (H+) i s 
the hydro gen ion a ctivity, [HS- ] is th e bi s ul f ide i on co n -
centration, AFe OOH i s the s ur face a r ea of th e r eactan t 
go e thit e , and k i s the r a t e con s t ant. Th e c oef fi c i ent s x , 
y, a nd z a re th e r eaction orders f or th e re s pecti v e s peci es . 
The initi a l r at e me thod or differe n t i a l me t hod i s 
base d on th e fa c t tha t: 
= ln k + x ln (H+) + y ln [HS- ] + z ln A 
FeOOII 
The e qu a tion i s r a th e r compl ex f or a r eact i on d epe ndent on 
sev e ral r eacta nt s , but it i s po ssi bl e to simp l ify t he ex -
pre ss ion by ma intaining all but one of th e r eac tan t s co n -
s tant. Thi s r es ult s in 
ln R? = ln k* + x ln (H+) 
l 
wh e r e 
ln k* = ln k + y ln [HS- ] + z ln A1~ OO ?e H 
1 36 
A plot of the lo g of the initial rate v s. the log of 
the particular variable, (H+) in thi s case, will y i e ld a 
s traight line with a slope x . Thi s proc ss i s repeated 
until th e r eaction order s for the reactants are determined. 
The initial rat e in thi s s tudy wa s det e rmined from 
plot s of the acid ex tractabl e iron s ulfide (>.2~ co nc e ntra -
tions v s . time. Then a s mooth curve or s traight lin e was 
fitted to the c urve by a lea s t s quare s regre ss ion method. 
The initial rate was determined from thi s equ a tion by takin g 
the first derivative of th e curve's equation a nd se tting the 
value of x equal to zero. For example: 
y =a+ bx+ cx 2 
dy/dt b + 2 ex 
dy/dt = b at X = 0 
For the several experimental runs i n whi ch one r e-
actant wa s varied, a lo g- lo g plot of b v s . th e c once nt rat i on 
of the species varied wa s made . A s trai ght lin e wa s drawn 
through the points by a least s quar es r egressio n. All but 
thre e of th e curves wer e fit by a par aboli c function. Th e 
remaining thre e were fit by a linear regr ession. Th e 
"goodne ss of fit" of the se curv es wa s determi ne d by on e of 
two me thod s . The linear e qu at ion s were t ested by a n equal 
tails test of th e correlation coeff i c i e nt a t th e 95% confi -
dence lev e l. Th e quadrati c curves we r e t es t ed by t he F- test 
me thod (Kreysz i g , 1970) a t th e 95 % co nf id e nce l eve l. Al l o f 
the curves fi t tl1 e cla ta a t th e a bove co n? ide nce limits . 
-
137 
The rate constant , k, for th e r eac tion could th e n 
be determined from the knowledge of the r eac tion order a nd 
concen tratio n s of the variou s reactant s in the rat e exp res-
s ion and th e initial r ate, R - . 
l 
k = 
All calculations, curve fitting, s ignificance te s ting 
a nd regre ss ion a nalyses were performed on a Hew l ett-Packard 
HP - 65 calculator. All programs were found in their S tat Pa c l. 
S ignificance of linear correlation coefficient s were det e r-
mined from tabl es in Crow et a l. (1960). F-t es ting of para-
bolic curves was performed fro m tabl es in Kreysz i g (1970). 
-
138 
BIBLIOGRAPHY 
Atkins on R.J., Posner A. M. and Quirk J.P. (1968) Crystal 
nucleation in Fe(III) so lutions and hydrox ide ge l s . 
J. Inor g . Nuc l Chem. 30, 2371-2381, 
Baas Becking L.G.M. (1956) Biological processes in th e 
estuarine environment . VI- Th? state of iron in th e 
estuarine mud iron sulfides . Konikl- Nederl. Akad. 
Wetensc happen ~' 1 81 - 189. 
Baes C.F. and Mesmer R. E. (1976) The Hydrolysi s of Cations. 
Wiley. 
Barnes H. L. and czamanski G. K. (1967) Solubilities and 
transport of ore mineral s . Geochemi s try o{ Hydroth e rma l 
Or? Deposit s (editor H. L. Barnes), Holt, Rinehar t, a nd 
Wilson, 334- 381 . 
Bartlett J.K. and Skoo? D.A. (1954) Colorimetric deterrni na ? 
t1. on of elemental sub lfur in hydrocarbons . Anal. Chern. 
~. 1008-1011. pH buffering of pore water of rec ent 
Ben-Yaakov S . (1973) Limn. and Qceanog. li, 86-94 . 
a noxic se diment s . 
Berne r R.A. (1964a) Iron sulfides formed from aqu eo us 
solution at low t emp eratures and atmospheric press ur e . 
J. Geol. !_l_, 293 - 306 . 
Be rner R.A. (1964b) Stability fields of iron min er a l s i n 
anaerobic marine sediment s . J. Gaol- 7]. , 826?83 4. 
Berner R.A. (1964c) Distribution a nd diagenesis of sulfur 
in some sediment s from the Gulf of California. Mar. eco l. 
~. 117-140. 
Berner R.A. (1967) Thermodynamic stability oE sedimentary 
iron s u 1 f id e s . Am . J . Sc i . ill? 7 7 3 - 7 8 5 . 
Berner R.A. (1969) The synth es i s of fra mboidal pyrite. 
Econ. Geol. ~' 383-384 , 
Berner R.A. (1970) sedimentary pyr ite formation. Am. J. Sc i. 
~. 1-23. 
Berner R.A. (1971) Principles of Chemica l Sedim entology. 
McGraw - Hill. 
1 39 
Be rner R. A. (1 972 ) Sul fa t e redu ction, pr r ite fo~matio n a nd 
th e o ceani c s ulfur budge t. Th e Chang ing Ch e~i s t r y of t h e 
Oc e an s ( e ditor s D. Drysse n a nd D. Jag n e r) Wi l e y , 3 4 7 - 36 1 . 
Be rtine K.K. (1972) Th e d e po s ition of molyb de num in a nox i c 
wat e r s . Mar . Chem.!, 43-54. 
Bl y ho ld e n G. a nd Ri cha rd s on E . A . (1 9 6 2) I n fra red a nd 
v olume t r i c d a ta on the a d s o r pt i on o f a mm o n i a , water , a nd 
oth e r gase s on a ctiva t e d i r on ( Ill) ox id e . J . Phys. Ch e m. 
66, 2 5 97 -2 6 02 . 
Bray J.T., Bri c k e r O.P. a n d Troup B . N. ( 1 973 ) Ph o s ph a t e in 
int e r s titi a l wa ters of a no x i c sed im e nt s : oxi d a t i on e ff ec t s 
during samp ling proc e dur es . Sc i e n ce 1 8 0 , 1 3 6 2 - 1 363. 
Br e we r P.G. a nd Sp e nc e r D.W. ( 19 74) Di s tribution of some 
trac e e l e me nt s in Black Sea a n d their f lux b e t we e n 
di s s olve d and p a rticulat e phase s . Am. Ass o c . Pe t ro l. Ge ol . 
Me m. 7 0, 13 7-14 3 . 
Brook s R . R., Pr es l e y B . J. a nd Ka pl a n I.R. ( 1 968) Trac e 
e l e me nt s in the inter s t i ti a l wa ters of ma rine se d i me n t s . 
Ge o c him. Co s mochim. Act a ll , 3 9 7- 414. 
Br on ge r s ma - Sand e r s M. ( 1 957 ) Ma ss mort a lity i n th e se a . 
Tre a ti se on Marine Ecolo y a nd Pa leo ecol ogy ( e d i t o r 
J. W. He d g p e th Geol. Soc . Am. ! , 9 41 
Bu d d M. S . a nd Be wi c k H. A. (1 9 5 2) Photometric det e r mi n a tion 
o f s ul fi d es a nd reduc ib le s ul f ur in al kalie s . Ana l. Ch e m. 
~' 15 3 6-1540 . 
Ca rrol D. (1 9 58 ) Rol e of c l ay minera l s in t h e tran s po rtation 
of iron. Ge o c him. Co s mo c h i m. Ac t a !_!, 1-27 . 
Cas per s H. (1957)In Tr ea ti se on Ma rine Ec ol ogy a nd Pa l e o -
ecolo gy ( e ditor J. W. He d g p e th) Ge ol. Soc . Am. ! , 8 01 
Charlot G. ( 1 9 64) Colorim e t ric De t ermina t io n of El eme n t s . 
E l se v e r e . 
Ch e n K. Y. a n d Morris J.C. (1 9 7 2) Kine ti cs of ox id a tion of 
aqu e ou s s ul fi de by Oz . Env i ro n. Sci . Te c h. ~' 5 29 - 537 . 
Ch e n g M. . (1 9 72) Eff ec t o f f er r o u s iron on t h e o x yge natio n 
of r e duce d s ul f ur s p e ci es . M. S . thes i s , U. Md . 
Ch r i s t e n se n 1-l. E . a nd Lug inb y hl T . T. ( 1 974 ) Th e Tox i c 
S ub s t a n ce s Li s t 19 7 4 Editi o n . USDHEW . 
Chiu J.~. a ~d Ri c hard s F . A . (1969) Oxyge natio n of h ydr oge n 
s ulfi de in s eawa t e r a t co n s t a n t s a l i n ity, t emp e ratur e 
a nd pH. Envir o n Sci. Te c h. 1, 838 - 84 3 . ' 
-
140 
Crow E .L ., Davis F. A. and Maxf ield M.W. (1960) 
Statistics Manual. Dover. 
De e r W. A., Howie R.A. and zussman J. (1966) An Introduction 
!__O the Rock Forming Mineral_2. . Wil ey. 
Doyle R. W. (1968) Identification and solubility of iron 
sulfide in a naerobic Jake sediment. Am. J. Sci. 266, 
980-994. -
Drever J. I. (1971) Magnesium-iron replacement in c Jay 
minerals in anoxic marine sediments. Sc i ence 172, 
1334-1336. 
Edwards J.O., Greene E.F . and Ro ss J. (1 968) From stOichiom -
etry and rate law to mechani sm. J. Chem. Ed . 42, 381-385 -
Erd R.C., Evans H.T. and Richter D.H. (1957) Smythite, a 
new iron sul fide and associated pyrrhotit e from Ind iana . 
Am . Mineral. ~' 309-333. 
Garrell? R.M. and Christ c.L. (1965) Solutions, Minera l s 
and Equilibria . Harper and Row. 
Gast R.G., Landa E.R . and Meyer G.W. (]974) The interaction 
of water with goethite (o-FeOOH) and amorphous hydrated 
ferric ox ide s urfaces. Clays and Clay Min. ~ . 31-39-
(1973) Mechani sm of trac e met a l trans port in 
Gi bbs R.J. 
rivers. Science 180, 71- 73 . 
Giggenbach W. ( 1972) Optical spectra and equilibrium 
distribution of polys ulfide ions in aqu eous s olution 
at 20? . Inorg. Chem. _!l_, 1201-1207. 
Goldhaber M.B. and Kaplan I.R. (19 75) Apparent dissociation 
cons tants of hydrogen sulfide in chlorid e so lutions . 
Mar. Chem.}, 83 - 104. 
Harvey H.W. (1955) The Chemi s try and ~ tilitY of Sea Wa~? 
Cambridge Univ. Press. 
Hem J. D. (1960) Some chemical relationships among s ulfur 
species and dissolved ferrous iron- USGS Water Suppl y 
Paper 1459-C, 57-73. 
Hem J.D. and Cropper W.H. (1954) s urvey of ferro us- fe r ri c 
c hemical equilibria and redox pot enti a l s . 
USGS Water Supply Paper 145 9-A, 1 -31 . 
Jefferson D.A., Tricke r M.J. and Winterbottom A. P. (19 75) 
Electron-microscopic and mH ss bauer s pectros copic 
s tudies of iron s t a ined kaolinite mi neral s . Cl ay s and 
Clay Min . Il, 355-360 . 
141 
Jenne E .A. (1968) Controls on Mn, Fe, Co, Ni, Cu a nd Zn 
concentrations in soils and water: the significant rol e 
of hydrous Mn and Fe oxides. Trace Jnorganics in Water 
(editor R. F. Gould) Am. Chem~ Soc. Adv. in Chem. Ser. 
]_l, 337-387 . 
Kaplan I .R., Emery K.0. and Rittenberg s.c. (1963) The 
distribution and isotopic abundance of sulfur in recent 
marine sediment off southern California. Geochim-
Cosmochim. Acta'!:}___, 297 - 332. 
Kennedy V.C . , Zellweger G.W. and Jones B.F. (1974) Filter 
pore-size effects on the analysis of Al, Fe, Mn, and Ti 
in water . Water Resources Research !Q., 785-790. 
Kester D.R., Byrne R.H.Jr. and Liang Yu-Jean. (1975) Redox 
reactions and solution complexes of iron in marine 
systems. Marine Chemistr in the Coastal Environment 
(editor T. M. Church). Am . Chem. Soc., 56-79. 
Krauskopf K.B. (1956) Factors controlling the concentrations 
of thirteen rare metals in sea water - Geochim. co smochim. 
Acta ~' 1-32. 
Krauskopf K. B. (196 7) !71troduction !Q.Jceochemistry. 
McGraw-Hill. 
Kreyszig E. (1970) Introductory Math~tical Statisti0_ ? 
Wiley. 
Landa E.R. and Gast R.G . (1973) Evaluation of crys tallinitY 
in hydrated ferric oxides . Clays and Clay Min. Q, 1 21 ? 1 3 0. 
Laurence G. S . and Ellis K. S. (19 7 2) Oxidation of iodide ion 
by iron (Ill) in aqueous solution- J . Chem. Soc. London 
Dalton, 2229-2233. 
Moelwyn-Hughes E.A. (1933) The Kinetics of Reaction s in 
Solutions . Oxford at the Clarendon Pre ss . 
Murakami E. (1952) Su]fate?producing bacteria. Int- symp. 
Enzyme Chem . (Tokyo) J__, 53-54. 
Nissenbaum A. , Presley B.J. and Kaplan J.IL (19 72 ) DarlY 
diagenesis in a reducing fjord, Saanich Inl et, Briti sh 
Col=bia. J. Chemical and isotopic changes in major 
component s of interstitial water . Geochim. Co smo chim-
Acta ~. 1007-1027. 
O'Brien D.J. (1974) Kinetics of the oxida tion of r educ ed 
s ulfur species in aqueous solution. Ph.D- thes i s , U. Md. 
14 2 
Oz retich R.J. (1975) Mechanism for deep water r e n e wa l in 
Lake Nitinat, a permanently anoxic fjord . Est. Coast. 
Mar. Sci. l, 189-200. 
Pa rk s G.A. and De Bruyn P.L. (1962) The zero point of 
char g e of oxides . J. Phys. Chem. ~' 967-973 . 
Park s G.A. (1965) The isoelectric points of solid ox ide s , 
s olid hydro x id e s, and a queous hydroxo compl ex sy s tem s . 
Chem. Reviews 165 177-198. 
' 
Park s G.A. (1968) Aq u eo u s s ur face chemist~y of ox ides an~ 
complex oxide minerals isoelectric point and zero point 
of charge. Equilibrium 'concept s in Natural Water Sy s t e ms 
(editor W. Stumm). Am. Ch e m. Soc . Adv. Chem . Ser . 67, 
121-161. 
Perrin D.D. and Dempsey B. (1974) Buffers for p H and Met a l 
Ion Control. Wiley. 
Pilling M.J. (1975) Re action Kinetics. Clar e ndon Pr es s . 
Po hl H.A. (1954) The formation a nd dissolution of me tal 
s ulfides . J. Am. Chem . Soc. J__j_, 2182-2184 . 
Presley B.J., Kolodny Y., Nissenbaum A. and Kaplan I.R. 
(1972) Early diag enesis in a reducing fjord, Sa ani c h 
Inlet, British Columbi a. II. Trace element di s tribution 
in interstitial water and sediment . Geochim. Cosmochim. 
Acta~' 1073-1090. 
Re y1:olds W.L. and Lumry R.W. (1966) Mechani s m o f El ec t r on 
1ransfer. Ronald Pres s Co. 
Richards F.A . and Vaccaro R . F . (1956) The Ca riac o Tr e n c h, 
an anaerobic basin in the Caribbean Sea. Dee p Se a Re s . 
1, 214-228. 
Richards F . A. (1965) Anoxic b a sins and fjord s . Ch e mi ca l 
Oc eanography (editors J. P . Riley and G. Skirrow). 
Academic Press!, 611-645 . 
Ricka rd D. T. (1969a) The chemi s try of iron s ulfid e f o r ma tion 
at low temperatures. Stockholm Contr. Ge ol. ~ ' 6 7- 9 5 . 
Rickard D.T. (1969b) Th e microbial forma tion o f iron 
s ulfides. Stockho lm Contr. Geol. ~' 49 - 66. 
Rickard D.T. ( 1 974) Kinetic s and me cha ni s m of the s ul f ida-
t i o n o f go e th i t e . Am . J . Sc i . 2 7 4 , 9 41 - 9 5 2 . 
Rickard D.T. ( 1 975) Kinetics and me chanism of pyrit e for ma -
t ion a t 1 ow tern per at u r e s . Am . J . Sc i . 2 7 5 , 6 3 6 - 6 5 2 . 
-
143 
Roberts W.M.' Walker A.L. and Buchanan A.S. (1969~ The 
~hernistry of pyrite formation in aqu~ous soluti?n and. 
its relation to the d e positional environment. Mineral1um 
Deposits?, 18 - 29. 
Rozanov A.G.' Volkov I.I.' Zhabina N.N. and Yagodinskiy:T.A. 
(1971) Hydrogen sulfide in the sedime~t~ of the co5n5tinental 
slope, northwest Pacific ocean. Geokhim1ya ~' 543- o. 
S chwarzenbach G. and Fischer A. (1960) Die Aciditat de~ 
S~lfane u nd die Zusammensetzung wasseriger Polysulfid-
losungen. Helv. Chim. Acta~' 1365-1390 . 
S iegbahn K. (1967) ESCA: atomic, molecular and solid s tat e 
st ructure studied by means of electron spectroscopy. 
Nova. Acta R. Soc. Sci. Upsal. Ser. IV, 20. 
S kinner B.J., Erd R.C. and Grimaldi F . S . (1964) Greigite, 
the thiospinel of iron: a new mineral. Am. Min . ~' 
543-555. 
s t umm W. and Lee G.F. (1961) Oxygenation of ferrous iron. 
Ind. Eng. Chem.~' 143-146. 
Stumr:1 W. and Morgan J. J. (1970) Aqua tic Chemistry. 
Wiley-Interscience. 
Sweei:iey R.E. and Kaplan I .R. (1973) Pyrite framboid forma -
tion : laboratory synthesis and marine sediment s . Econ. 
Geol . 68, 618-634. 
T e der A. (1971) The equilibrium between elementary s ulfur 
and aqueous polysulfide solutions. Acta Chem. Scand. 
25, 1722 - 1728. 
Thode H . G. and Kemp A.L.W. (1968) The mechanism of bacterial 
reduction of sulfate and of sulfite from isotopic fra c-
tionation studies . Geochim. Cosmochim. Acta g, 71-91. 
Trud~nger P.A., Lanbert I.B . and Skyring G.W. (19 72) 
Biogenic sulfide ores: a feasibility study. Econ. 
Geol . ?_]___, 11 1 4-1127. 
Troup B.N . , Bricker O.P. and Bray J.T. (1974) Ox idation 
effects on the analysis of iron in the interstitial 
waters of recent sediments. Nature 249, 237. 
Urban P. J. (1961) Colorimetry of sulfur anions. Z. Anal. 
Chem. 179, 415-422. 
Van Straaten L . M.J.V . (1954) Composition and s tructur e of 
recent sediments of the Netherlands. Leid s e Ge ol. 
Me dedel _!2., 1-110. 
-
,. 
144 
Walton AG ? f 
Pr . ?. ? (1967) The Formation and Propertie s o 
~- lnterscience. 
Wesdt i P ?.W ?  and Gaeke G.C. (1956) Fi? xati?o n o f?  su If ur 
oxi~e as disulfitomercurate (II) and subsequent 
Colorimetric estimation. Anal. Chem. ~' 1816-1819. 
Wilkins R.G . (1974) The Study of Kinetics and Mechanism of 
~ction of Transition Metal Complexes . Allyn and Bacon. 
Zajic J.E. ( cl ? P 
1969) Microbial Biogeochemistry. Aca emic re s s.