EPA-600/3-76-062b
July 1976                              Ecological Research Series
                EFFECT  OF  HYDROGEN SULFIDE ON
                          FISH AND  INVERTEBRATES
                        Part I - Hydrogen Sulfide
                    Determination and  Relationship
                  Between  pH and Sulfide Toxicity
                                     Environmental Research Laboratory
                                    Office of Research and Development
                                   U.S. Environmental Protection Agency
                                         Duluth, Minnesota 55804

-------
                 RESEARCH REPORTING SERIES

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

     1.    Environmental Health Effects Research
     2.    Environmental Protection Technology
     3.    Ecological Research
     4.    Environmental Monitoring
     5.    Socioeconomic Environmental Studies

This report has been assigned to the ECOLOGICAL RESEARCH series. This series
describes research  on the effects of pollution on  humans, plant  and animal
species, and materials. Problems are assessed for their long- and short-term
influences. Investigations include formation, transport, and pathway studies to
determine the fate  of pollutants and their effects. This work provides the technical
basis for setting standards to minimize undesirable changes in living organisms
in the aquatic, terrestrial, and atmospheric environments.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.

-------
                                          EPA-600/3-76-062b
                                          July 1976
EFFECT OF HYDROGEN SULFIDE ON FISH AND INVERTEBRATES

   Part II - Hydrogen Sulfide Determination and

   Relationship Between pH and Sulfide Toxicity
                        by

                Steven J. Broderius
                Lloyd L. Smith, Jr.
 Department of Entomology, Fisheries, and Wildlife
              University of Minnesota
            St. Paul, Minnesota  55108
                 Grant No. R800992
                   Project  Officer

               Kenneth E. F.  Hokanson
     Environmental  Research Laboratory - Duluth
            Monticello,  Minnesota  55362
        U.S. ENVIRONMENTAL PROTECTION AGENCY
         OFFICE OF RESEARCH AND DEVELOPMENT
          ENVIRONMENTAL RESEARCH LABORATORY
              DULUTH, MINNESOTA  55804

-------
                                 DISCLAIMER

This report has been reviewed by the Environmental Research Laboratory -
Duluth, U.S. Environmental Protection Agency, and approved for publication.
Approval does not signify that the contents necessarily reflect the views
and policies of the U.S. Environmental Protection Agency, nor does mention
of trade names or commercial products constitute endorsement or recommendation
for use.
                                      11

-------
                              ABSTRACT

An analytical method was developed for the direct determination of
yg/liter concentrations of molecular H_S.  The procedure involves
bubbling compressed nitrogen through an aqueous sulfide solution to
displace H«S which is collected in a glass bead concentration column
and measured colorimetrically.  The H-S concentration is calculated
from the determined sulfide displacement rate and by reference to a
log linear standard curve relating temperature with the H-S displace-
ment rate to the H~S concentration in standard solutions.  To permit
accurate determination of H_S from the determined dissolved sulfide
concentration and fraction of dissolved sulfide as H~S for specific
conditions of temperature and pH, the apparent linear relationship
between pK1 for H^S,  ,. and temperature was defined.  This procedure
          J.      £ (.aq;
of calculating H2S in various waters and effluents was confirmed by the
direct technique.
The described analytical technique was used to define the relationship
between test pH and sulfide toxicity to the fathead minnow.  Within the
pH range of 7.1 to 8.7, 96-hr LC50 values for molecular H-S decreased
linearly from 57.3 to 14.9 yg/liter with increasing pH.  However, the
log 96-hr LC50 values of dissolved sulfide increased linearly from 64.0
to 780.1 yg/liter with increasing test pH ranging from 6.5 to 8.7.

This report was submitted in fulfillment of Grant Number R800992 by
the Department of Entomology, Fisheries, and Wildlife, University of
Minnesota, under the sponsorship of the Environmental Protection Agency.
Work was completed as of March 1975.
                                   iii

-------
                                CONTENTS
                                                                   Page
Abstract                                                           iii
List of Figures                                                     vi
List of Tables                                                     vii
Acknowledgments                                                     ix
Sections
I     Conclusions                                                    1
II    Recommendations                                                2
III   Introduction                                                   4
        Theoretical Approach                                         4
        Literature Review                                            7
IV    Materials and Methods                                         26
        Determination of Sulfide in Aqueous Solution                26
        Direct Determination of Molecular H?S in                    36
          Aqueous Solution
        Acute Sulfide Bioassays                                     41
V     Results and Interpretations                                   46
        Determination of Sulfide in Aqueous Solution                46
        Direct Determination of Molecular H~S in                    57
          Aqueous Solution
        Equilibrium Constants for the First Dissociation            60
          of H9S,  .
              2 (aq)
        H2S Determination in Various Waters and Effluents           67
        Relationship Between Test pH and Sulfide Toxicity           75
          to the Fathead Minnow
VI    Discussion                                                    88
        Determination of Molecular H~S and K.,                       88
          lonization Constants of H0S,  N
                                   2 (aq)
        Modes of Toxic Action of Dissolved Sulfide                  89
          to Fish
VII   References                                                    95
VIII  Appendix                                                     103

-------
                               FIGURES
No.                                                                Page
 1  Relationship Between the First Dissociation Constant (K,)       12
      of H0S,  N and Temperature for Aqueous Solutions
          2 (aq)
      of Generally Low Ionic Strength

 2  Apparatus Used for Distribution Between Water and               37
      Nitrogen and Concentration of Molecular H~S
 3  Relationship Between Test pH and Dissolved Sulfide, HS ,        87
      and Molecular H~S Concentration at Levels Corresponding
      to the 96-hr LC50 for Fathead Minnows at 20 C
                                  VI

-------
                                 TABLES
No.                                                                Page
 1  First Dissociation Constants of H0S,  N in Aqueous              10
                                     2 (aq)
      Solution at Various Temperatures and Generally Low
      Ionic Strength
 2  Analysis of Laboratory Well Water                               42
 3  Linear Regression Analysis of Calibration Curves                51
      Relating Absorbance (Y) and Sulfide Concentration (X)
      in ug ELS per 25-ml for Solutions Prepared with Various
      Diamine Reagents and Under Different Acidity Conditions
 4  Recovery and Stabilization of H-S by Glass Bead Concen-         53
      tration Columns Coated with Various Metal Salts
 5  Stability of Metal Sulfides on Concentration Columns            55
      to Oxidation by Air
 6  H-S Displacement by Nitrogen Dispersed Through Test             58
      Solutions of Known Molecular H-S Concentration and
      Temperature
 7  Apparent K- Dissociation Constants and pK.. Values of            62
      H0S,  .  Determined for Test Solutions of Different Tem-
       2. (aq)
      peratures, pH Values, and Total Sulfide Concentrations
 8  Relationship Between Apparent K.. Dissociation Constants and     66
      pK.. Values of H^S..  . for Temperatures Ranging from  10
      to 25 C
 9  Fraction of Dissolved Sulfide as Molecular H~S in Aqueous       68
      Sulfide Solutions of Low Ionic Strength
10  Multiplication Factors for Converting H-S Calculated from       72
      Pomeroy's Factors for a "Typical Water Supply" to
      Corresponding Concentrations Based on This Study
11  Multiplication Factors for Converting H2S Calculated from       76
      Factors in the 1946 to 1965 Editions of Standard Methods
      to Corresponding Concentrations Based on This Study
                                   vii

-------
No.                                                                Page
12  Determination of Molecular H?S in Different Waters by           78
      Calculation from the Total and Dissolved Sulfide Con-
      centration and by a Direct Technique
13  Determination of Molecular H-S in Different Effluents by        80
      Calculation from the Total and Dissolved Sulfide Con-
      centration and by a Direct Technique
14  Summary of Test Conditions in Sulfide Bioassays at              82
      Different pH Values
15  Description of Fathead Minnows Used in Sulfide Bioassays        84
      at Different pH Values
16  Biological Assay by the BMD03S Probit Analysis Method of        85
      96-hour Fathead Minnow SuJ fide Bioassays Grouped
      According to Test pH
17  Summary of Lethal Concentration (LC) Analysis for 96-hour       86
      Fathead Minnow Sulfide Bioassays Grouped According
      to Test pH
                                  viii

-------
                           ACKNOWLEDGMENTS

The authors wish to thank David L. Lind for his assistance in performing
the acute sulfide bioassays and for aid in their analysis.  The assis-
tance of other supporting personnel is also acknowledged.
                                    ix

-------
                              SECTION I
                             CONCLUSIONS

1.  The direct determination of molecular H-S in aqueous solutions at
    levels as low as 4 yg/liter can be accomplished by using a vapor
    phase equilibration technique in conjunction with a glass bead
    concentration column coated with 0.1 M zinc acetate.

2.  The procedure of calculating molecular H«S concentrations in various
    waters and effluents from the determined dissolved sulfide concen-
    tration and the fraction of dissolved sulfide as H2S defined in
    this study, for specific conditions of temperature and pH, was
    confirmed by a direct method for H^S determination.

3.  The 96-hr LC50 values of molecular H_S for the fathead minnow de-
    creased linearly from 57.3 to 14.9 yg/liter with increasing pH
    within the range 7.1 to 8.7.

4.  A positive linear relationship was observed between log dissolved
    sulfide concentration and test pH ranging from 6.5 to 8.7, at
    sulfide concentrations corresponding to the 96-hr LC50 values for
    the fathead minnow at 20 C.

5.  The acute toxicity of sulfide solutions to fathead minnows does not
    depend entirely on the concentration of ambient molecular H?S.  The
    HS  ion appears to contribute to a much lesser extent to the toxicity
    of these solutions.

-------
                               SECTION  II
                           RECOMMENDATIONS

1.  Research should be conducted  to determine  if  the N,N-diethyl or
    N,ethyl-N-hydroxyethyl-p-phenylenediamine  reagents should replace
    the N,N-dimethyl-p-phenylenediamine reagent generally used in the
    colorimetric determination of sulfide.
2.  The factors corresponding to the decimal fraction of dissolved
                                                2
    sulfide as molecular H?S proposed by Pomeroy  and those presented
    in the 9th through 13th editions of Standard Methods for the Exami-
    nation of Water and Wastewater should be replaced by factors derived
    from tl
    study.
from the expression pK  = 7.252 - 0.01342 T (C) as defined in this
3.  The best method for the determination of molecular H-S, that has
    the widest application to practical situations, is based on calcula-
    tions from dissolved sulfide, pH, and temperature measurements.  It
    is recommended that this long-accepted procedure be continued and
    that use of the new proposed factors, sample preparation by filtra-
    tion rather than flocculation, and optimization of the colorimetric
    test be employed.

4.  Previous published H-S concentrations should be corrected to corre-
    spond to a common base derived from factors proposed in this study
    so that comparison of reported toxicity data will be consistent.

-------
5.  Reports of future sulfide toxicity tests in freshwaters should
    include molecular H S concentrations and in addition the pH, tem-
    perature, and dissolved sulfide values.

-------
                              SECTION  III
                              INTRODUCTION

Hydrogen sulfide is one  of  the  end  products which may  result  from the
bacteriological action on organic material containing  protein and from
various other chemical processes.   This toxic substance is found in
many ground and surface  waters  and  numerous  effluents are, either
directly or indirectly,  important sources of sulfides  in natural waters.
Because good toxicological  information on sulfides with respect to
aquatic species was lacking,  the United States Environmental  Protection
Agency awarded a research grant (No.  R800992) to the University of
Minnesota to investigate the  effect of hydrogen sulfide on various
freshwater species and life history stages.  The report on this project
has been divided into two parts: I, dealing with the the toxic effects
of hydrogen sulfide on aquatic organisms; II, dealing with the analy-
tical determination of molecular hydrogen sulfide (H~S,  ,) and defini-
                                                    ^  vaq/
tion of the relationship between test solution pH and  the toxicity of
dissolved sulfide to the fathead minnow (Pimephales promelas  Rafinesque)

THEORETICAL APPROACH
The accuracy of the method used to calculate molecular H_S throughout
Part I of this study was uncertain, suggesting the need for a method
for the direct determination of H_S.  Such a method could then be used
to define the relationship between pK, (i.e., -log K^) and temperature
so that the current procedure of calculating H?S concentrations from
dissolved sulfide,  pH,  and temperature measurements could be  corrected.

-------
An accurate means of determining molecular H?S could also be used to
investigate the relative importance of H.,3 and the HS  ion in con-
tributing to the acute toxicity to fish of sulfide solutions.

Determination of Molecular H^S and First lonization Constant (K..) of H^S
It has become common practice to define total sulfide as dissolved H-S
and HS, as well as acid-soluble metallic sulfides present in the sus-
pended matter.  The acid-insoluble sulfides such as copper sulfide and
 2-
S   ion, which is present in significant proportions only above about
pH 11, are not included in this definition.  The continued development
of new and refinement of existing analytical methods for the determina-
tion of small quantities of sulfide suggests that no one method is
entirely satisfactory or applicable to all types of samples.  Determina-
tions are also complicated by the fact that sulfides undergo oxidation
in the presence of air or oxygen.  The accepted procedure for deter-
mining molecular or un-ionized hydrogen sulfide in aqueous solutions is
by appropriate calculations with the known concentration of dissolved
sulfide, the pH of the sample, and the use of the first ionization
constant (K.) for H»S.  There are situations in which dissolved sulfide
cannot be accurately determined with presently accepted analytical
methods due to interferences or the presence of complex sulfides.  The
accuracy of the H_S determination depends, among other things, on the
accuracy of the value assumed for the first ionization constant of H~S
(K,).  A review of the scientific literature indicates that  the rela-
tionship between K.. and temperature is not well established and there-
fore the present method for H.,S determination may be inaccurate.

Only through chemical analysis of test water can the relationship be-
tween concentration of the toxicant and the observed harmful effects on
the test animals be definitely established.  A specific and  sensitive
independent analytical method for the determination of uhdissociated
molecular HjS which excludes other sulfide forms in water would be most
useful.  Such a method would be a tool  to aid research concerning the
toxicology of sulfides, for effective practical application  of research

-------
 results to waste disposal control in natural waters, and prediction of
 the toxicity of water polluted with sulfide compounds.   Therefore,  a
 major objective of this study was to develop a direct analytical method
 to determine molecular H-S which would be accurate for  concentrations
 as low as a few yg/liter.  Its utility for the prediction or explanation
 of adverse conditions for aquatic life in natural and waste waters  con-
 taining sulfide was also examined.   Since the present acceptable method
 for calculating molecular H_S is applicable in most situations,  a
 further objective of this study was to evaluate the relationship between
 pK...  and temperature to permit accurate determination of H-S by this
 well established procedure.

 The vapor phase equilibration method utilized during this study  for the
 determination of molecular H.S does not significantly disturb equilibria
 involving other sulfides since less than 1 per cent of  the dissolved
 sulfide is removed.   In this  procedure finely dispersed compressed
 nitrogen is bubbled continuously and at a regulated rate through a  test
 solution to displace H_S which is trapped out of  the nitrogen stream
 and quantified  by a conventional colorimetric method for sulfide.   The
 H.S  concentration in the tested solution is  obtained by reference to a
 log linear standard curve relating  yg of H-S displaced  per liter of
 nitrogen dispersed  at  various  temperatures between 10 and 25  C to the
 known H-S  concentration in standard Na2S solutions.

 Test pH  and Toxicity of Dissolved Sulfide
 The pK,  for H-S  is approximately  7,  thus at  a pH value  of  near 7  about
 50 per cent of  the dissolved sulfide will be as H-S  and 50  per cent  as
 the hydrosulfide  ion, HS.  In more  acidic solutions  the equilibrium will
 shift towards a  greater percentage  of  the sulfide  as  H-S and  in more
alkaline solutions towards that of  the HS  ion.  Therefore, in most
natural waters  the HS   ion can be expected to be the  principal dissolved
sulfide species.  It is  generally assumed that the  toxicity of sulfide
solutions  to fish is mainly due to  the penetration of the gills by  the

-------
undissociated H_S species and not of the HS  ion.  Thus a change in test
solution pH should have a drastic effect on the toxicity of a given con-
centration of dissolved sulfide.  For any bioassay study it is necessary
to know if the observed toxicity results from a specific toxicant known
or believed to be present.  Therefore, another objective of this research
was to define the relationship between test pH and the toxicity of dis-
solved sulfide with special reference to the molecular H?S concentration.
Continuous-flow 96-hr toxicity bioassays with fathead minnows were per-
formed at six pH values to define this relationship and to test the
validity of the assumption that the toxicity to fish of dilute sulfide
solutions depends on the concentration of ambient molecular H»S with
contribution of the HS  ion being negligible.

LITERATURE REVIEW
Equilibrium Constants for the First Dissociation of H^S/   N
_j	2-(aq)
When H-S gas is dissolved in water an ionization equilibrium  is estab-
lished that can be represented  by the equation:
                H2S(a  )  ^     H  + HS~ ^    s  2H  +  S                (1)

 The  proportion  of  sulfur existing in  aqueous solution  as undissociated
                                              _                     ?—
 molecular H-S  (H2S,  .), hydrosulfide ion (HS ),  and sulfide ion (S  )
 is determined by the chemical and physical conditions  in the solution
 and  the  equilibrium constants for the first and second dissociation of
 the  sulfide  species.   The  equilibrium expressions for  the above reactions
 are  given by:

                K,  =  [H*"] [HS~]/[H-S,   J  » 10"7 at 20  C             (2)
                 1                 2  (aq)
                     [H+]  [S2~]/[HS~] - 10"13'5 at 20 C              (3)
 Using the approximate values of K  and K« at 20 C, it can be demonstrated

-------
 that the second equilibrium constant is so small that the percentage of
                       2-
 dissolved sulfide as S   ion is less than 0.32 per cent when the hydrogen
 ion concentration is greater than 10    (i.e., pH less than 11).  There-
 fore,  for practical purposes it is assumed that no significant amount of
 sulfide ion is formed below a pH of 11 and the second dissociation step
 and the presence of sulfide ion will be neglected in calculations and
 discussions in this report.  The total concentration of dissolved sul-
 fide species in solution is thus given by the concentration of molecular
 H2S (H2S,  ,)  plus the hydrosulfide ion (HS~).  Theoretically then the
 concentration  of molecular  H-S in most freshwaters of low ionic strength
 (i.e.,  y less  than 0.01)  can be calculated when the dissolved sulfide
 concentration,  pH,  and temperature are known and the relationship be-
 tween  the equilibrium constant for the first dissociation of H S and
 temperature is defined.   This procedure for the indirect determination
 of  H-S  is the  proposed method of the American Public Health Association
     *•  i
 et  al.   and has appeared  in Standard Methods for the Examination of  Water
                                                              2
 and Wastewater since  1946.   This method is based on Pomeroy's  procedure
 for calculation of  the un-ionized hydrogen  sulfide concentration by mul-
 tiplying the determined dissolved sulfide  concentration by a factor
 representing the  proportion of  dissolved sulfide as molecular H2S at
 the pH,  temperature,  and  ionic  strength of  the  solution.   It should  be
 pointed  out that  in the APHA 1971 edition  no relationship  between K.
 and  temperature was employed  in  this  determination and  no  rationale  for
 the  use  of  the revised specified  K.,  (1.1 x 10~  )  at 25  C and ionic
 strength  0.02 was included.
Since about 1900 numerous investigators have employed a variety of tech-
niques to measure the ionization constants of H~S at various tempera-
tures.  These values in general are not in good agreement but the first
dissociation constant of H.S,  .  given in the literature approximates
1 x 10   at 25 C.  Equilibrium constants vary with temperature and
attempts to define this relationship over a range of temperatures occur
in the chemical literature.  Most values reported for K- by previous
investigators are shown in Table 1.  The single constant proposed by

-------
       2            -7
Pomeroy  of 1.7 x 10   at 25 C and y = 0, which has been used exten-
sively, is considerably higher than values of KI at 25 C reported by
other authors.  This difference would suggest that Pomeroy 's constant
and data derived from it may be in error.  A line defined by a linear
regression equation depicting the relationship between K- and tempera-
ture from 0 to 35 C for most of the dissociation constants in Table 1
is shown in Figure 1.  Thus the apparent equilibrium constant for the
first dissociation of H0S,  v in aqueous solution of low ionic strength
                       2 (aq)
in relation to temperature  (T) in degrees Celcius (C) is approximated
by the expression:

                        =  (0.31 + 0.029  T)  • 10~7.                   (4)
There have been a few attempts  to define K   over  a  range  of  temperatures.
Linear regression equations  summarizing these  findings  are:
   Temperature            Regression
     range, C _ equation _ Reference _
     5-30      K..  =  (0.28 + 0.032 T)-10~7        Wright & Maass6
                                        -77
    10 -  35      K..  =  (0.33 + 0.020 T)-10          Tumanova et al.
                                        -7                        4
     0-25      K.  =  (0.27 + 0.024 T) • 10          Loy &  Himmelblau
                  1                                             4
The equation representing  data  from  work by Loy and Himmelblau  defines
a relationship between  temperature and the  true or  absolute equilibrium
constants since  it  includes  corrections for experimental solution ionic
strength. The earlier  works of Wright and  Maass  and Tumanova, Mish-
chenko,  and  Flis   represent  relationships between the apparent equili-
 brium constant and temperature since corrections for ionic strength
 were not made.  Barnes,  Helgeson,  and Ellis  summarize the work of Rinj
 bom with the expression:

                   log OO = -7.05 + 0.0125 (T - 25)               (5)

-------
Table 1.  FIRST DISSOCIATION CONSTANTS OF H-S,  .  IN AQUEOUS SOLUTION
                                           2 (aq)

       AT VARIOUS TEMPERATURES AND GENERALLY LOW IONIC STRENGTH

Method
Various
Radioactivity
conductance
Thermodynamics
Conductance
Thermodynamics
Conductance
Potentiometric
Thermodynamics
Conductance
Thermodynamics
Conductance
Conductance
Conductance
Potentiometric
Thermodynamics
Conductance
Potentiometric
Thermodynamics
Tempera- Ionic
ture, strength,
C ,u
0
& 0 0

0
5 0
5
10 0
10 0.02-0.04
10
15 0
15
18
18
18
18
18
20 0
20
20
K,-107
0.1
0.271

0.434
0.471
0.501
0.574
0.534
0.579
0.747
0.668
1.2
0.57
0.91
3.31
0.729
0.896
0.873
0.772
Reference
3
Jellinek & Czerwinski
4
Loy & Himmelblau

Barnes et al.
Wright & Maass
Barnes et al.
Wright & Maass
Tumanova et al .
Barnes et al.
Wright & Maass
Barnes et al.
Paul (1899) In: Pomeroy2
Walker & Cormack (1900)
2
In : Pomeroy
Auerbach
9
Epprecht
Barnes et al.
Wright & Maass
Kubli10
Barnes et al.
                                  10

-------
Table 1 (continued).  FIRST DISSOCIATION CONSTANTS OF H0S,  ,
                                                       2 (aq)
       IN AQUEOUS SOLUTION AT VARIOUS TEMPERATURES AND
                 GENERALLY LOW IONIC STRENGTH

Temper- Ionic
ature, strength,
Method C ju
Thermodynamics
Conductance
Colorimetric
Colorimetric
Potentiometric
Potentiometric
Spectrophoto-
25
25
25
25
25
25
25
-
0
0
0.02
—
0.02-0.04
0
K, -107
j.
1.15
1.08
1.7
2.0
1.24
0.790
0.95
Reference
Lewis &
Wright &
2
Pomeroy
2
Pomeroy
Yui12
Randall11
Maass



Tumanova et al.
Ellis &
Golding13
  metric
Radioactivity &  25
  conductance
Thermodynamics   25
Potentiometric   25
Literature       25
  review
Conductance      30
Thermodynamics   30
Potentiometric   35
Thermodynamics   35
                        0.10
                         0.02
0.87

0.891
1.6
0.955

1.26
1.029
1.029
1.188
                                          Loy  & Himmelblau
Barnes et al.'
Hseu & Rechnitz
Chen & Morris15
                                          Wright  & Maass
                                          Barnes  et al.
Tumanova et al.
Barnes et al.
                                                          14
                               11

-------
                                                                        POMEROY
                                          TEMPERATURE   (C)

Figure 1.   Relationship between the first dissociation constant (K.)  of H-S,  . and temperature
           for aqueous solutions of generally low ionic strength.  Equation defining the regression
line is
                       (0.31 + 0.029 T) • 10~7.

-------
where  T = solution temperature in C.

       2
Pomeroy  reported that if the effect of ionic strength on the activities
of ions is considered when calculating the true or absolute ionization
constant, the relationship can be expressed by:
            log  (K^) = log  [HS~] - log  [H2S] - pH - 0.5^V~          (6)
where  p =  ionic  strength.
He also reported  that  an  increase  in  temperature has  the  same effect as
an increase in pH to the  extent  that  log K.  increases by  0.0146  unit per
degree centigrade.

Determination of  Sulfide  in  an Aqueous  Solution
The  accurate determination of small amounts  of sulfide is made  difficult
because sulfide  is oxidized  in the presence  of air or oxygen.   However,
there are various quantitative methods  to  determine microgram quantities
of aqueous  sulfide and H_S evolved from sulfur and sulfides.   These
methods fall into three broad categories - volumetric using iodometric
titration,  colorimetric with the formation of colored complex compounds,
and  a fluorimetric procedure.  Methods  based on  the isolation of evolved
H-S  usually employ alkaline  or metallic solutions  or suspensions as
trapping  agents.

Volumetric  Iodometric  Method—In most iodometric titration methods pre-
sented  in the  literature, dissolved  and acid-soluble sulfides are oxi-
dized  to  sulfur  in an  acid medium according  to the following reaction:
                                    13

-------
There  are many modifications  of  the  titration method but they generally
involve  either titration  of the  sulfide directly with iodine or iodide-
iodate mixture, or  addition of an  excess of  iodine or iodide-iodate
mixture  and back  titration with  standard sodium thiosulfate.

Pomeroy   and Bethge   have outlined  certain important sources of error
affecting the iodometric  titration method.   Summarizing these findings,
it can be stated  that since H^S will  escape  an acidified solution, a
moderately alkaline sodium sulfide solution  should be pipetted directly
into an  acidified iodine  solution, and the excess iodine back-titrated
at room  temperature with  standard  sodium thiosulfate to minimize the
loss of  H2S.  The iodine  solution  must be acidified to the proper extent
since  in alkaline solutions at room temperature iodine oxidizes a small
portion  of the sulfide to sulfate.  It is also important that sulfides
which  are stabilized with an absorbant form  acid-soluble precipitates
and in addition, the ionic strength of the absorbant should be kept as
low as possible.  Also critical to this method is that the absorbant
containing the sulfide be acidified before the addition of the iodine.
Bethge   observed that if sulfide was added  to a cadmium or zinc absor-
bant and the mixture in turn added to an iodine solution, the titri-
metric results were highly variable.  On the other hand if the absor-
bants without sulfide were added to the iodine solution, the amount of
thiosulfate titrant consumed was comparable with the titration of iodine
solutions alone.   Apparently the large sulfur-containing particles
formed in the presence of the metal absorbant include cadmium or zinc
sulfide besides some iodine.   Titrations should also be conducted
allowing minimum contact of sulfide solutions with air and dissolved
oxygen since sulfide is readily oxidized.

       18
Bethge   discussed methods for the volumetric determination of sulfides
in which the sulfide contained in a very strong alkaline or heated solu-
tion is oxidized by an excess of oxidant to sulfate.   After oxidation,
the excess oxidant is determined by adding potassium iodide and acidi-
fying,  and the iodine liberated is then back-titrated with sodium thio-
                                   14

-------
                18
sulfate.  Bethge   concluded that oxidation of sulfides to the sulfate
state by potassium iodate is quantitative within experimental error.
Therefore, his method forms a better basis for the estimation of sulfides
than do methods employing the oxidants sodium hypochlorite and potassium
permanganate, since these latter reagents are partially decomposed by
boiling with strong alkali while potassium iodate is not.  In methods in
which the sulfide is oxidized tt> sulfate, four times as much oxidant is
required, and so the sensitivity is increased fourfold over the methods
in which sulfide is oxidized to sulfur.

Colorimetric Methylene Blue Method-—The methylene blue reaction is
recognized as one of the most specific and most sensitive of the few
colorimetric procedures available for the determination of sulfides,
allowing for the determination of approximately 10 yg/liter sulfide-
sulfur.  This method is based on the specific reaction of an acidic
solution of N,N-dimethyl-p-phenylenediamine oxalate or sulfate  (p-amino-
N,N-dimethyl-analine) with sulfide in the presence of an excess amount
of iron  (III) oxidizing agent and chloride to cause complete color de-
velopment of methylene blue in about 1 min.  This is generally known as
Lauth's or Caro's reaction.  Following color development, diammonium
hydrogen phosphate is usually added to eliminate the ferric color.
Various combinations of reagents have been proposed for use in  the
color-forming reaction.  The standard method proposed by APHA et al.
is based on Pomeroy's   modifications.  Many applications of this method
are reported in the chemical literature but since new modifications
continue to appear, one should be aware that the method is reliable but
lacks perfection.

Stoichiometry of reaction—If a stoichiometric reaction between sulfide
and N,N-dimethyl-p-phenylenediamine is assumed, the reaction may be
represented by 2 moles diamine reacting with 1 mole sulfide in  the
presence of ferric chloride to form methylene blue.  According  to the
stoichiometric reaction, 1 mg sulfide should yield 9.97 mg dye.  Pome-
   18
roy   observed that 1 mg sulfide produced 6.57 mg methylene blue for a
                                    15

-------
yield of 66 per cent.  Therefore, it is evident that sulfide is not
quantitatively transformed to methylene blue but that some is lost in
                                        19
side reactions.  According to Gustafsson   the ferric ion together with
N,N-dimethyl-p-phenylenediamine gives a red oxidation product, Wurster's
red, which reacts in several ways with sulfide.  Among other products,
sulfide green, leuco methylene blue and methylene red are said to be
formed.  The first two are easily transformed to methylene blue, whereas
methylene red is not.  Methylene red and methylene blue are reported by
          19
Gustafsson   to be formed in the proportion 1:50.  He also calculated
that 66.7 + 0.5 per cent of the added sulfide is transformed to methylene
          ~ 20
blue.  Cline   stated that when a comparison is made of the apparent
molar absorptivity with that of pure methylene blue solutions under com-
parable conditions the reaction is approximately 62 per cent complete.
                              21
Recently, Zutshi and Mahadevan   observed the recovery of methylene blue
from a given sulfide sample to be 65 + 2.0 per cent of the theoretical
value when using methylene blue obtained from different sources.  When
the methylene blue samples were dehydrated before preparing the solu-
tions by keeping overnight in an oven at 80 C, the authors found that
the recovery of methylene blue from sulfide samples was reduced to 51
+2.0 per cent.  The loss in weight on drying is apparently due to loss
of absorbed moisture by the methylene blue powder or crystals.  This
loss in weight was reported to be dependent on environmental conditions
rather than the sources of manufacture.

                                                          19
Absorbance and molar absorptivity—According to Gustafsson,  Hofmann
        22      20                         21
and Hamm,  Cline,  and Zutshi and Mahadevan   the color formed during
the methylene blue reaction has an absorption spectrum with rather
steep peaks and wave lengths of maximum absorption at about 670 and
                         22
745 nm.  Hofmann and Hamm   stated that this maximum at 745 nm is not
caused by the formation of a second dye but probably by the interaction
between methylene blue and a chloroferrate (III) complex.  The absorp-
tion at the peak wave lengths is strongly dependent on the concentration
of acid in the mixture and as the pH of the color development solution
declines, the primary absorption band of the formed methylene blue shifts
                                    16

-------
from about 670 to 745 nm.

      2o                       23
Cline    and Grasshoff and Chan   concluded from their work and that
previously reported by other authors that the final absorbance after
dilution should be less than 1.0 in a 1-cm light path cuvette and pre-
ferably even less than 0.8 because aqueous methylene blue solutions do
not strictly conform to Beer's Law at higher sulfide concentrations and
that increasing the amount of reagents does not improve the linear range.
                                                    20
Departure from Beer's Law was also observed by Cline   when the molar
diamine concentration was less than seven times the total sulfide molar
concentration.  Dilution of more concentrated methylene blue solutions
to a solution of exactly the same acid concentration permitted Johnson
           24
and Nishita   to restore the linear relationship and thereby prove that
this lack of linearity is caused by deviations from Beer's Law and not
                                                         19
by a decrease in the yield of methylene blue.  Gustafsson   proposed an
explanation for this observed phenomenon which depends on the assumption
that the absorption spectrum of the formed methylene blue consists of
two overlapping absorption bands with maxima at 600 and 650 nm.  The
former is more pronounced in more concentrated solutions and is due to
the dimeric ion; the latter is due to the monomeric ion, predominating
                                                                     2+
in very dilute solutions.  The dissociation of the dimeric ion  ((MB)   )
is increased by diluting the solution or raising the temperature and
thus increasing the relative absorbance at 670 nm.
The molar absorptivity index (O of the methylene blue solution used by
                   24
Johnson and Nishita,  corresponding to the range 5 to 50 yg sulfur in
100 ml, was about 34,500 liter mole   cm,  very close to that for maximum
                                                        23
sensitivity of colorimetric methods.  Grasshoff and Chan   concluded that
the sensitivity of the methylene blue method is 0.2 yg at. H-S-S/liter
or 6.4 yg H2S-S/liter.

Factors affecting color development—The effect of various factors such
as reagent strength, acidity, temperature, light, and salt on the methy-
                                    17

-------
lene blue  reaction have been  investigated  by  numerous  authors.   It  is
well known that  the absorbance  of  a methylene blue  solution  formed  from
                                                       19
a given amount of sulfide  is  pH-dependent.  Gustafsson   observed that
the acid concentration may affect  both  the yield  of the reaction and the
extinction of methylene blue.   He  reported that the yield of methylene
blue increases with increasing  acidity, whereas its extinction at 667 nm
decreases.  This result is different  than  the assumption  made by Fogo and
        25
Popowsky    that variable acidity of the diatnine reagent influences  the
absorption spectrum of methylene blue rather  than the  reaction yield.
                            23
Tests by Grasshoff and Chan  showed  that  the absorbance  of  the  methylene
blue solution was essentially constant when the acidity was  in the  range
                         19
0.4 to 1.0 M.  Gustafsson   reported  that  the optimum  sulfuric acid con-
centration in the final solution is about  0.30 M  when  the reaction  and
measurement acidity are the same.  Examination of his  data also  reveals
that when  the reaction acidity  is  about 0.67  M and  the measurement  acid-
ity 0.1 M,  an even greater absorbance at 667  nm is  realized.  In general,
it can be  stated that most  studies have been  confined  to  acid concen-
trations of less than 0.7  M and sulfuric acid has been found to  be  a
better medium than hydrochloric acid for the  diamine reagent.

                    21
Zutshi and Mahadevan   found that  if the acid  concentration during
reaction and measurement is constant, only the 675  nm  peak is observed
at low acidity.  This peak increases with  increasing acidity, reaching
a maximum  at a sulfuric acid concentration of  about 0.4 M, and gradually
decreases  at higher acidity.  A peak also appears at 749  nm and  in-
creases in  intensity with  increasing acid concentration up to about 1.4
M, possibly indicating the  formation of another compound  produced by
the addition of a proton to the methylene blue cation.  At maximum  ex-
tinction,   the absorbance at this later peak is about 20 per cent greater
                                                          21
than the maximum at the 675 nm peak.   Zutshi  and  Mahadevan   proposed
that the sum of the absorbances at both 675 and 749 nm or that at 749 nm
alone be used to measure the sulfide concentration when the sulfuric
acid concentration at both reaction and measurement is approximately
1.5 M.   This suggestion is based on the fact  that the arithmetic sum of
                                    18

-------
the two absorbance values remains more steady than the separate values
when the acid concentration is varied.  It was also stated that Beer's
Law is very closely followed only at relatively high acid concentrations,
a fact which may account for the deviations reported by earlier inves-
tigators who employed lower acid concentrations.  Therefore, some care
is necessary to maintain reproducible acid conditions in both standard
solutions and actual determinations on samples so that a near optimal
color development is realized with the specific sample volumes and
dilutions employed.

The formation of methylene blue is not only a function of acidity but
may depend on temperature.  The reaction is known to be faster at higher
temperatures, but at the same time the possibility of H~S escaping before
                                  19
reacting is increased.  Gustafsson   stated that the precision of the
method is much improved by keeping the temperature constant for both the
                                                               21
reaction and the absorbance measurements.  Zutshi and Mahadevan   found
that at normal laboratory temperatures there is essentially no variation
in the absorbance with temperature.

         26
Patterson   stated that methylene blue solutions, although stable several
days in the dark, fade rapidly in sunlight.  However, Zutshi and Maha-
     21
devan   found that under normal laboratory lighting or  in laboratory
daylight there is no detectable photodecomposition effect on the methy-
lene blue during or after the reaction.
The presence of many different electrolytes other than those added as
                                  27
reagents was found by Sands et al.   to have no effect on color inten-
            20
sity.  Cline   stated that the method is without salt effect over the
salinity range of 0 to 50 parts per thousand.
Interfering substances—The effect of a variety of substances on color
development by the methylene blue procedure for sulfide has been inves-
tigated by various authors.  Pomeroy   stated that sulfites, expressed
as S0_, or thiosulfate and chlorine up to concentrations of 10 mg/liter
                                    19

-------
 have no  serious effect on color development.   By increasing the concen-
 tration  of  Fed., and extending the time for reaction,  accurate  results
                                                           2-         15
 may be obtained in the presence of 40 mg/liter SO- or  S^O-.  Pomeroy
 also stated that colloidal sulfur has no effect on the test.  According
                27
 to  Sands et al.    carbon disulfide,  the common types of organic sulfur
 compounds found in coke-oven gas, carburetted water gas,  natural gas,
 and thiophene  in concentrations about 40 times that of the sulfide  do
                                                   1 f\            97
 not appreciably affect color development.   Pomeroy,  Sands et aL^.,  and
                 28
 Marbach  and Doty   observed that many mercaptans produce  a pink color on
 addition of the diamine reagent in the methylene blue  method.   However,
 this effect is essentially eliminated by addition of ammonium phosphate
 and by proper  wave length selection,  since  the pink solutions do not
 significantly  absorb near 670 or 745  nm.  Because the  methylene blue
 colorimetric assay is conducted in strongly acidic conditions,  sulfides
                                                          29
 may be determined in the presence of  many metals.   Siegel   stated  that
  _2
 10    M zinc, cadmium,  magnesium,  and  manganese do not  interfere.  Pome-
 roy  concluded  that suspensions of iron, zinc,  and  lead  sulfides react
 just as  readily  as dissolved sulfide.   Copper,  however, the sulfide of
                                                 —36
 which possesses  a solubility product  of about 10,    binds the sulfide so
 tightly  that formation of methylene blue is almost completely inhibited.

                  2-
 Hyposulfite, S^O,,   and  nitrite are two materials  known to interfere
 with the methylene blue  procedure.  Hyposulfite  gives  a false positive
 sulfide  test since it  is  decomposed by  acid to H»S as  one of the pro-
 ducts.  When nitrite  is  present in an acidified  sample containing sul-
 fide, some of  the  sulfide  is  oxidized to sulfur.   Pomeroy  observed
 that  a sample  of  sewage  containing 1.0  mg/liter  nitrite and 0.5 mg/liter
 sulfide produced a color  corresponding  to 0.4 mg/liter sulfide.  In
 general,  however,  the methylene blue procedure is  not  seriously affected
by most compounds which might ordinarily be expected to be present  in
natural waters, sewage, and many  industrial effluents.

                          30  31
Diamine reagents—Zavodnov   '    stated that the reagents,  p-phenylene-
diamine and N,N-dimethyl-p-phenylenediamine, used  in the  colorimetric
                                   20

-------
determination of sulfide in formation of methylene blue,  can be replaced
by N,N-diethyl-p-phenylenediamine or N-ethyl-N-hydroxyethyl-p-phenylene-
diamine without any procedural changes in the colorimetric determination
                                            32
of sulfide.  Rees, Gyllenspetz, and Docherty   also investigated the use
of various diamine reagents and showed that N,N-diethyl-p-phenylene-
diamine produced a stable ethylene blue color which was about twice as
intense for a given amount of sulfide as that produced with the dimethyl
diamine reagent normally used.  The method employed was similar to the
methylene blue procedure except that no diammonium hydrogen phosphate
was added.  This omission was made because the acidity of the final
solution would not allow for the iron (III) phosphate (if present) to
dissolve.  The standard deviations at 100 and 1.0 yg sulfide per 100 ml
were reported to be about 3 and 0.08 ug, respectively.

                                   33
Isolation of H~S—Paez and Guagnini   described a method for the isola-
tion and ultramicro determination of H_S in gaseous mixtures or water
by sorption on an anion-exchange resin column of hydroxide  from Amber-
lite IRA 400, 20-50 mesh.  The collected sulfide is then eluted with
4 M sodium hydroxide and determined  colorimetrically by the methylene
blue method.  The technique was reported to permit the determination of
1 to 20 yg H?S present at dilutions  of 0.07 to 20 ppm in air and down
to 0.1 pg/liter in water.  The sulfide can be kept on the resin without
loss for as long  as 10 days before the analysis.  Many other investi-
gators have employed alkaline  or metallic solutions or suspensions  as
trapping agents in the isolation of  evolved H S.

Fluorimetric Method—The analytical  determination  of H_S,  ,.  in air  has
	           '                         2  (g)
received much attention in recent  years.  Jacobs,  Bravermann,  and Hoch-
       34
heiser   determined atmospheric H_S  in  the  ppb range  using  the methylene
blue method, but  required  large volumes  of  air,  since  either long  sam-
pling  periods or  high flow rates were necessary.   The high  flow rates
resulted  in  low trapping efficiencies.   Paper  impregnated with Pb  (II)
and Hg (II)  salts has been used  extensively to trap H~S.   Sensenbaugh
           35
and Hemeon  used Pb(OAc)~-impregnated  paper traps, but  this system could
                                    21

-------
be used  only  at  low flow rates  (<0.5 liter/min)  and the PbS  formed was
                                          O£
unstable (Smith,  Jenkins and Cunningworth  ).  Mercury (II)  chloride
forms  the more stable  HgS but,  when combined with the  methylene blue
method is not sensitive enough  to measure background levels  (Hoch-
                  37
heiser and  Elfers  ).
 In order  to measure  H-S,  -.  in the ppb  range,  an  indirect  fluorescence
 method  involving  fluorescein  mercuric  acetate (FMA)  has been  used.
                   oo
 Andrew  and Nichols   observed that fluorescein solutions  in dilute
 aqueous alkali  produced intense  yellow-green  fluorescence.  Addition of
 sulfide to FMA  solutions  linearly decreases the  fluorescence  intensity,
                                    2-
 producing a pink  coloration.   The S - FMA reaction  appears to be in-
                            39
 stantaneous.  Axelrod et  al.   observed that  the fluorescence intensity
 was not affected  by  variations in ionic strength, but was affected by
 changes in pH.  The  FMA intensity was  reported to fall off rapidly at
 NaOH concentrations  greater than 0.15  N.  However, NaOH concentrations
 between 0.05 and  0.1 N did not affect  the FMA intensity, and  1.0 N H.SO
 added to  neutralize  any excess NaOH did not alter the fluorescence.  The
 excitation and  emission wave  lengths generally utilized are about 499
 and 519 nm, respectively.  The useful  range of analysis of sulfide is
 0.5 to  10 x 10~8 M (0.16  to 3.2  vig-S/liter) in 0.1 N NaOH and 1 x 10~7 M
 FMA.  It  is important  to  note  that  the  FMA reagent must be standardized
 daily with dilute  sulfide solutions.
One of the most sensitive published methods for the determination of
                                       39
H0S, . in air is that of Axelrod et al.   who trapped the gas in an
 2 (g)                           	
alkaline aqueous solution and estimated the resulting sulfide by the
fluorimetric technique.  While the sensitivity of this method is ade-

                                                                      39
                                                              40
quate, the collected sulfide is unstable (Avrahami and Golding  ) and
necessitates that analysis follow soon after sampling.  Axelrod et al.'
also attempted to place FMA in a bubbler to capture the H_S directly from
the air.  However, the FMA was strongly affected by aeration and by ex-
posure to sunlight.  Therefore, the FMA reagent should be added to the
                                    22

-------
sample following the collection and stabilization of sulfide.   If pre-
cipitation is used to stabilize the sulfide,  the precipitate must be
redissolved prior to the addition of the FMA.

              41
Natusch et al.   used a modification of the above method and were able
to measure trace levels of atmospheric hydrogen sulfide as low as 5
                           -12
parts per trillion (ppt, 10   ).  According to their method H S, ,. is
extracted from air and stabilized as Ag~S by reacting with a AgNO_-
impregnated filter.  This efficient recovery technique was first used
               O£
by Smith et al.    The filters were prepared with Whatman No. 4 filter
paper soaked for 2 min in 0.01 M HNO  containing 2% AgNC-  and 2% ethanol
and allowed to dry.  The collected Ag?S is then dissolved with 0.1 M
NaCN - 0.1 M NaOH solution producing the non-interfering silver cyanide
complex Ag(CN)2 and free sulfide which is analyzed fluorimetrically
using very dilute fluorescein mercuric acetate.  The advantages of the
method include efficient collection of H_S, good stability of the col-
lected sulfide, and a sensitive and specific analysis method.  Mercap-
tans and high ozone levels were reported to interfere with mercaptans
reacting with AgNC-  to  form silver mercaptans which in  turn quench FMA.

Vapor Phase Equilibration Method
Various methods have been published for  the  determination  of  dissolved
gases in aqueous  solutions but  few  can provide  a precise measurement  of
existing concentrations since most  methods involve marked  disturbance
of  relevant  equilibria  through  removal of  all  or a large proportion of
the gas in question.   The vapor phase  equilibration procedure has proven
to  be a specific  and  sensitive  analytical  method for  the direct  deter-
mination of  molecular HCN  in  aqueous  solutions  without  material  dis-
turbance of  existing  ionic  equilibria  of the system.  With this  method
a distribution equilibrium is  established  between the concentration of
HCN in solution and in finely  dispersed  air  or  nitrogen bubbled  through
the sample.   Analysis of the  displaced HCN,  which is  collected on one of
several possible types of  concentration columns, can be accomplished by
                                     23

-------
                                         42
various procedures.  Schneider and Freund   used gas-liquid chroma-
tography incorporating a thermal conductivity detector.  Claeys and
      43
Freund   developed a more sensitive modification by utilizing a chro-
matographic procedure using a flame ionization detector while Nelson
         44
and Lysyj   measured the trapped HCN polarographically in a cell con-
taining a stationary platinum cathode and rotating gold anode.  The
procedure developed and utilized in this investigation for the direct
determination of molecular H~S most closely approximates in principle
                           45
that described by Broderius,  where the displaced and collected HCN was
determined by a colorimetric method.  Once an accurate method was de-
veloped for the direct determination of molecular H~S a means of defining
the relationship between pK.. for H-S,  , and temperature was available.
This ultimately would allow for the accurate calculation of H~S concen-
trations from dissolved sulfide, pH, and temperature measurements.
Toxicity of Sulfide Solutions in Relation to pH
Although detailed experimental work on the detrimental effect of H_S
in the ug/liter range on many aquatic species and certain life history
stages has been published in recent years, very limited work and only
at high sulfide concentrations has been done to define the relationship
                                                      46
between pH and toxicity of dissolved sulfide.  Jacques   showed that
the rate of sulfide penetration into cells of the alga Valonia macrophysa
was proportional to the concentration of molecular H~S in the external
solution.  It was also demonstrated that after extended exposure in
solutions of constant total dissolved sulfide, the equilibrium concen-
tration of total sulfide inside the cells varied with external solution
pH and by calculation can be shown to be proportional to the molecular
H^S content of the test solutions.  Therefore Jacques' work suggests
that the rate of entrance and equilibrium concentration of sulfide in
the cells is controlled by the diffusion of molecular H~S across the
cell membrane.

                     47
Longwell and Pentelow   found that the acute toxicity of a standard
                                    24

-------
solution of 3,200 yg/liter total sulfide,  as demonstrated by overturning
time in minutes for 50 per cent of a brown trout sample, decreased as
the pH increased from 6.0 to 9.0.  This observed effect was attributed
by the authors to the greater percentage of dissolved sulfide as free
                                  48
H-S at the lower pH values.  Jones   also stated that the increased
toxicity to sticklebacks of sulfide solutions at low pH values is most
likely due to the greater proportion of the dissolved sulfide being
                                                      49
present in the form of molecular H2S.  Bonn and Follis   measured the
acute toxicity of sulfide from calculated molecular H»S concentrations
to various life history stages of the channel catfish (Ictalurus punc-
tatus) at pH values varying from 6.8 to 7.8.  They observed that the
toxic effects of sulfide appear to be independent of total sulfide in
solution but are related to the calculated quantity of molecular H~S as
controlled by the pH of the solution.  This relationship was then
                          49
applied by Bonn and Follis   in attempting to improve the productivity
of acidic lakes having a high H_S level by raising the  pH through
addition of limestone.  In summary,  it appears  that the  toxicity of
sulfide to fish may be largely attributed  to the toxic  action  of mole-
cular H-S, varying with pH and the dissolved sulfide concentration in
solution.
                                    25

-------
                               SECTION  IV
                        MATERIALS AND  METHODS

DETERMINATION  OF  SULFIDE  IN AQUEOUS  SOLUTION
The purpose of this  section is  to describe  the methods used  throughout
this research  for  the determination  of  sulfide.  The procedures used
included iodometric  titration and colorimetric methods.  As  noted in
the literature review, there are other  methods which would permit the
measurement of  H~S at much lower levels but these two procedures were
satisfactory for the purpose of this study  since they are widely used,
convenient, and allow for the determination of H_S at levels as low as
a few yg/liter.

Iodometric Titration Method
This method is based on the addition of excess standard iodine solution
to a sulfide solution under acidic conditions followed by titration with
standard thiosulfate to determine the unreacted iodine and the iodine
consumed by sulfide.  Because hydrogen  sulfide is volatile and sulfides
are oxidized by dissolved oxygen, exposure to air and manipulation of
the sulfide standard solutions was kept to a minimum.  When a stabilizing
absorbant was used the sulfide did not  form an acid-insoluble metal pre-
cipitate and the absorbant containing the sulfide was added to an acidi-
fied solution before the addition of the iodine.  The reagents in the
titrimetric procedure were prepared with reagent grade chemicals and
deionized water  which had been boiled and allowed to cool and stored
under a nitrogen atmosphere.   Most of the reagents were prepared and
standardized according to directions presented in the 13th edition of
                                    26

-------
Standard Methods (APHA ) under sections headed sulfide, titrimetric
(iodine) method (section 228 A-3) and oxygen (dissolved), azide modi-
fication (section 218 B-2).  The sodium thiosulfate titrant was stan-
dardized with both standard potassium biniodate and dichromate solutions
with the determined average thiosulfate normality used in appropriate
calculations.

Stock sulfide solutions were prepared by dissolving approximately 0.75 g
sodium sulfide (Na_S'9H»0) in oxygen-free boiled, cooled deionized water.
                  £,    £*
Crystals of the reagent grade sodium sulfide were rinsed free of oxi-
dation products such as sodium sulfite and thiosulfate with deionized
water, dried quickly on filter paper under an atmosphere of nitrogen,
and weighed.  The crystals are deliquescent but the rate of water absorp-
tion is slow.  The weighed sodium sulfide crystals were immediately
dissolved and diluted to 1.0 liter in a volumetric flask to form a solu-
tion containing approximately 0.1 mg S/1.0 ml.  If the weight of Na^S^H-
used was not 0.75 g, the sulfide concentration was calculated from

                        mg/liter S = (133.4) B                       (8)
where  S =  sulfide concentration  expressed as  sulfur
       B =  g Na2S-9H20/liter.

The  stock sulfide solutions were  standardized  by  adding  approximately
100  ml deoxygenated  deionized water  to  an Erlenmeyer  flask,  followed by
4  drops concentrated HC1.  Then 10.00 ml standard iodine solution,  fol-
lowed immediately by 20.00 ml stock  sulfide  solution, were delivered to
the  flask below the  solution surface.   After 1 to 3 minutes,  the  residual
iodine was  determined by titration with sodium thiosulfate using  starch
indicator.  The sulfide concentration,  which should be approximately
                                    27

-------
100 mg/liter as S or 1 ml = 100 yg, is calculated as follows:

                          = (A-B) 16 000 yg/millieq.
                6                  ml sample                          '
where  S = sulfide concentration expressed as sulfur
       A = (10 ml iodine) (milliequivalents iodine/ml)
                     2_
       B = (X ml S2
-------
cool and stored under an atmosphere of nitrogen.  Most of the reagents
were prepared according to directions presented in Standard Methods
(APHA ), under section headed sulfide, methylene blue visual color-
matching method (section 228 B-3).   Procedures for the preparation of
reagents not included in this section are as follows:

Diamine-Sulfuric Acid Stock Reagent—Dissolve 14.78 g N,N-dimethyl-p-
phenylenediamine oxalate, 17.06 g N,N-diethyl-p-phenylenediamine oxalate
or 18.67 g N-ethyl-N-hydroxyethyl-p-phenylenediamine sulfate in a cold
mixture of 50 ml concentrated H~SO, and 20 ml deionized water; cool,
then dilute to 100 ml with deionized water.  The stock reagents are
stored in dark glass bottles at 4 C.  The oxalate solutions have been
reported to be stable indefinitely while the sulfate solution is stable
for at least a few months when stored at 4 C.

Diamine-Sulfuric Acid Reagent—Dilute 5.0 ml diamine-sulfuric acid stock
solution with deionized water and 17.6, 57.7, or 77.7 ml concentrated
H?SO, to a final volume of 200 ml.  These solutions were used in sulfide
determinations when the desired reaction molar acidity was 0.25, 0.75,
or 1.0, respectively.  The solutions were stored in dark glass bottles
at room temperature and have been reported to be stable for at least
1 month at 25 C.

The preparation of the diamine reagents and  the amount used was done
such that the concentration of diamine at the time of color development
was the same as if the procedure outlined in Standard Methods  (APHA  )
had been followed.  About 10 times  as much diamine reagent was added as
is theoretically necessary to react with 15  ;ig H_S in 25 ml.

The procedure for  the colorimetric  test employed in  this study was
designed to minimize the production of interfering colors, while quickly
yielding near maximum blue color development.   This  end was accomplished
by keeping  the concentration of  diamine relatively low, using  a large
excess  of ferric  chloride, and by  adding diammonium  hydrogen  phosphate
                                    29

-------
 after  the  blue  color  had  developed.   This  latter  reagent  is  important
 since  it eliminates  the yellow color of  ferric  chloride,  the red color
 of  the ferric chloride-diamine combination,  and colors  formed by certain
 interfering  substances.   The  reagent also  prevents  further formation of
 the blue color  and neutralizes enough of the acid to  permit  the near
 maximum blue color development.   Color development  is complete within a
 few minutes  at  room  temperature  and  is stable for several hours if  the
 solutions  are kept in subdued light.

 Standard sulfide  solutions were  prepared and immediately  stabilized by
 forming an acid-soluble metal sulfide.   The  colorimetric  determination
 can subsequently  be completed either directly,  or after proper dilution.
 This technique  allows for color  development  in  dilute sulfide solutions
 with negligible loss  of the sulfide  by evaporation  or oxidation.  Many
 different  stabilizing solutions  have been  proposed  but  those which  con-
 tain either  zinc  or cadmium have received  the most  attention.  Several
 trapping solutions of these metals were  tested  to determine  their col-
 lecting efficiency and stability of  collected sulfide.  The  zinc stabi-
 lizing solutions  that were used  during this  investigation were prepared
 as  follows:

 Zinc Acetate Sulfide-Absorbing Solution  -  0.1 M Zn—Dissolve 21.95  g
 zinc acetate dihydrate (Zn(OAc)2'2H20) in  deionized water.   Acidify
 with about 3 drops of acetic  acid to  prevent hydrolysis and  dilute  to
 1 liter.

 Zinc Chloride Stabilizer  Sulfide-Absorbing Solution - 0.1 M  Zn-—Dissolve
 13.63  g zinc chloride (ZnCl2)  in deionized water, and 100 ml 1% (W/V)
 gelatin solution and  dilute to 1 liter.

 The cadmium stabilizing solutions that were  used  during this  investigation
were prepared as follows:

Alakline Cadmium Hydroxide Sulfide-Absorbing  Suspension - 0.017 M Cd—
                                    30

-------
Dissolve 4.3 g cadmium sulfate (3CdSO, -SH.,0)  in deionized water.   Add
0.3 g sodium hydroxide, dissolved in water,  and dilute to 1 liter.  The
final pH of the suspension was 7.7 and the suspension was mixed well
each time before using.
Cadmium Chloride Sulfide-Absorbing Solution - 0.1 M Cd-—Dissolve 22.84 g
cadmiui
liter.
cadmium chloride (CdCl2'2-1/2 H-O)  in deionized water and dilute to 1
Cadmium Sulfate Sulfide-Absorbing Solution - 0.1 M Cd—Dissolve 25.65 g
cadmium sulfate (3CdSO,.8H«0) in deionized water and dilute to 1 liter.

Calibration curves defining the relationship between absorbance and yg
sulfide expressed as H»S in a total volume of 25 ml were prepared for
later reference.  The standard sulfide solutions were prepared by di-
luting, with freshly boiled and cooled deionized water, a stock sodium
sulfide solution whose sulfide concentration was determined iodometri-
cally.  The curves were made immediately after the preparation of the
sulfide standard solution.  Aqueous standards were prepared by adding
to separate 25-ml volumetric flasks, each containing 0.5 ml of a sulfide-
stabilizing absorbant, the following volumes of a dilute aqueous sulfide
solution containing approximately 2.5 ug lUS/ml:  0  (reagent blank), 0.5,
1.0, 2.0, 3.0, 4.0 and 5.0 ml, in order to prepare a sulfide series  con-
taining approximately 0, 1.25, 2.5, 5.0, 7.5, 10.0, and 12.5 yg H2S,
respectively.  Portions of the standard solution were  transferred to the
volumetric flasks with a pipette, submerging the tip in the absorbant
before releasing the solution to avoid loss of sulfide.  The standard
sulfide solutions were then diluted in each flask with boiled and cooled
deionized water to a volume of 5.5 ml.  In all instances the standards
and samples were contained in the same volume before adding colorizing
reagents.

Color development is then accomplished by adding either 0.9 ml diamine-
sulfuric acid reagent, rapidly followed by the addition of  0.1 ml ferric
                                    31

-------
chloride reagent, or 1.0 ml fresh diamine and iron solution, prepared
by mixing nine parts of the diamine reagent with one part of the ferric
chloride solution.  After at least 1 min, and following mixing, 1.5 ml
diammonium hydrogen phosphate solution is added to eliminate the ferric
chloride color and the solution is mixed to dissolve the white ferric
phosphate precipitate.  Then the solutions are diluted to volume with
deionized water and the flasks are stoppered.  Reasonable care was taken
to use the specified amounts of reagents.  However, it was observed that
moderate variation with reagent concentrations would not markedly affect
color development.  The relative amounts of acid and ammonium phosphate
must be balanced in such a way that the ferric phosphate will remain in
solution at the end of the test.

The temperature of the solutions in the flask while adding reagents and
during absorbance measurements was between 19 and 21 C.  Following dilu-
tion to volume, the solutions were allowed to stand for about 30 min in
subdued light and then measurement of the absorbance in a 1-cm light
path cuvette was made using a Beckman DB-GT spectrophotometer against
the reagent blank, which is nearly colorless.  The wave lengths at which
absorbance measurements were made are 666,  668, and 672 nm for solutions
prepared with the N,N-dimethyl-p-phenylenediamine oxalate, N-ethyl-N-
hydroxyethyl-p-phenylenediamine sulfate, and N,N-diethyl-p-phenylene-
diamine oxalate color developing reagents,  respectively.

After completing one series of standards, the working sulfide standard
solution was discarded, a new standard solution was prepared and a
second set of standards were analyzed.  This procedure was repeated
several times and the relationship between micrograms of sulfide used
in preparing the solutions (expressed as H S) and average blank-corrected
absorbance values for the individual diamine reagents were defined by
linear regression equations.
                                    32

-------
Determination of Total and Dissolved Sulfide
The distinction between total and dissolved sulfide can be made by a
number of different procedures.   During this research total sulfide was
determined either by employing a direct colorimetric measurement on zinc
acetate-stabilized sulfide or by using a volatilization procedure fol-
lowed by adsorption.

Sulfide solutions to be used for direct colorimetric measurement were
stabilized by adding approximately 30 ml of the test solution,  followed
by a few drops of Na?CO~ solution when necessary to increase pH, to a
glass vial containing four drops of 2.0 N zinc acetate.  Color  develop-
ment on a well mixed 7.5-ml aliquot of the sample was then accomplished
according to directions presented in Standard Methods (APHA ),  under the
section headed methylene blue visual color-matching method (section 228
B-4a(2)).  Color intensity is determined with a spectrophotometer and
H-S calculated from an appropriate calibration curve.

The volatilization procedure involves stripping H2S from an acidified
solution by a stream of nitrogen.  The evolved sulfide is carried over
and quantitatively absorbed as zinc sulfide on a glass bead concentra-
tion column coated with 0.1 M zinc acetate.  The metal sulfide produced
is then determined colorimetrically.  The determination of total  sul-
fide is accomplished by addition of between 50 and  200 ml of sample  to
a 300-ml three-neck distilling flask containing 1.0 ml of 0.1 M  zinc
acetate solution.  The volume of sample in the flask is determined by
difference in weight before and after the addition  of sample and  from
the density of water at 20 C.  The flask is then connected to a  spray
trap by means of a standard taper  (i.e., S) ground  glass joint.   A
dropping funnel containing 1+1 l^SO^ is fitted in  the middle neck  of
the distilling flask by means of a S ground glass joint.  The glass  bead
concentration column, coated with approximately 0.5 ml of 0.1 M  zinc
acetate absorbant,  is then placed  in series with the spray trap  and  con-
nected with an 0-ring joint.  A  coarse porosity  glass  frit extending to
                                    33

-------
 the bottom of  the flask  is  inserted  in  the  third neck  and connected with
 a £ ground glass joint.  A  nitrogen  delivery  tube  is connected  to  the
 frit by means  of an 0-ring  joint.  Nitrogen is  then allowed to  pass
 through the system to displace all oxygen before releasing the  stabi-
 lized sulfide.  The rate of nitrogen flow through  the  apparatus is ad-
 justed to approximately  300 ml/rain.  This flow  is  sufficiently  rapid to
 sweep the system completely free of  H-S  in  less than 60 min.  Following
 this procedure  the stopcock in the dropping funnel is  opened and 20
 drops of 1 + 1  H.SO, is  allowed to run  slowly into the mixture, but the
 flow is stopped so that  some acid remains in  the bulb.  The acid con-
                             2-
 verts the dissolved HS   and S   ions as  well  as acid-soluble metallic
 sulfides to H»S because  of  the reduction in sample pH.  The steady stream
 of nitrogen is  maintained for 1 hr through  the  system  after acidification.
 During the stripping reaction no heat was applied  other than that pro-
 duced by addition of acid.  The H.S  displaced from the sample by the
 stream of nitrogen is then adsorbed  on  a glass  bead concentration column
 as a metal sulfide precipitate.  The displaced  sulfide is at room tem-
 perature and is protected from exposure  to  light during and following
 collection until analysis.  Samples  generally contained between 2 and 10
 Mg H2S and were analyzed as soon as  possible after collection to avoid
 loss of sulfide.  It should be emphasized that air was excluded at all
 stages to prevent oxidation and the  escape  of gaseous H_S.

 The precipitate in the concentration column is  treated with 5.0 ml of
 deionized water, 0.9 ml of the acidic N,N-dimethyl-p-phenylenediamine
 reagent,  and mixed by inversion.  Then 0.1 ml of the iron (III) reagent
 is added to the column, sealed with  parafilm, and mixed by inversion.
 The H_S evolved immediately reacts with  the resultant formation of methy-
 lene blue.  The H_S which escapes into  the  air space above the  liquid is
minimal since the liquid in the concentration column comprises essen-
 tially the entire volume.  After at  least 2 min the colorized solution
 in the concentration column is quantitatively removed into a 25-ral volu-
metric flask with three 5-ml washings of deionized water.  After approxi-
                                     34

-------
mately 30 min  the absorbance of the resulting methylene blue solution
is measured at 666 nm with a 1-cm light path cuvette and against a
reagent blank.  The corresponding quantity of sulfide expressed as pg
I^S is then determined from previously prepared calibration curves
developed from similar measurements on solutions prepared by adding a
known amount of sulfide directly to 25-ml volumetric flasks.  Since
absorbancy of methylene blue solutions is influenced by acid concentra-
tion, care was exercised in maintaining acid concentrations constant in
all determinations.

Dissolved sulfide was determined on samples following removal of sus-
pended solids by flocculation and settling, centrifugation, and fil-
tration.  The flocculation of suspended solids was accomplished by
filling a 125-ml reagent bottle to overflowing with test solution, fol-
lowed immediately by the addition of 0.2 ml of 6 N aluminum chloride
solution and 0.2 ml of 6 N NaOH.  The bottle was stoppered to exclude
air and then rotated back and forth about a transverse axis in order to
flocculate the contents.  The reagents used were prepared according to
directions presented in Standard Methods (APHA1) under the section headed
titrimetric (Iodine) method (section 228 A-3).  The flocculant was
allowed to settle for 15 min and then a portion of the clear supernatant
was removed and stabilized with zinc acetate.  The removal of suspended
solids was also accomplished by filling a 30-ml vial to overflowing with
test solution, capping, and centrifuging for 10 min.  A portion of the
supernatant was then removed and stabilized with zinc acetate.  Removal
of the suspended matter by filtration was accomplished by drawing a
20-ml sample into a 20-ml glass syringe, excluding as much air as possible,
and forcing the sample through a 0.45-micron millipore filter of 25 mm
diameter utilizing a swinny-type filter holder adapted by swedge lock
connection with the hypodermic syringe.  The filtrate was stabilized
immediately with zinc acetate.  Whenever the pH of the test solutions
was below about 7.5, two drops of 5% Na~CO  were added to the zinc sul-
fide solution to insure sulfide stabilization.  Color development and
sulfide determination on a well mixed 7.5-ml aliquot of the zinc-stabi-
                                    35

-------
lized samples was then accomplished according to directions previously
described.

DIRECT DETERMINATION OF MOLECULAR H2S IN AQUEOUS SOLUTION
The apparatus used for determination of molecular H.S is shown in Figure
2.  The flow rate of compressed nitrogen from a cylinder is maintained
by means of a two-stage gas regulator and a flow meter.  The nitrogen is
sparged through an approximate 16-in high column of test solution in a
bubbler immersed in a 20-liter Pyrex glass carboy.  The bubbler, des-
                              42
cribed by Schneider and Freund,  was designed so that the rising bubbles
cause circulation of the test solution in the container and prevent
significant local depletion of sulfide.  A medium porosity sintered
glass disc (30 mm diameter) produces the desired bubble size.

A spray trap is inserted between the bubbler and the concentration
column to ensure that no droplets of test solution are carried over in
the nitrogen and deposited  on the concentration column.

The concentration column is a 26-cm section of 10-mm diameter boro-
silicate glass tubing to which a three-way teflon buret stopcock has
been fused and which is packed with glass beads of 3-mm diameter.  The
acutal concentration section of the column containing the beads is 18
cm in length.  To facilitate installation and removal of the concen-
tration column, 12/5 0-ring joints were fused to each end.  The capil-
lary tip of the three-way teflon buret stopcock facilitates complete
delivery of the column washings into a 25-ml volumetric flask.

A 10-liter water displacement bottle is used to determine the volume
of nitrogen dispersed in the solution and passed through the concen-
tration column.  The bottle is inverted and mounted on a supporting
frame.  A graduated glass tube of 8-mm diameter is inserted in a 3/4-
inch hole drilled in the bottom of the bottle.  The glass tube extends
to within about 2 mm of the rubber stopper inserted in the neck of the
                                   36

-------
                          WATER FROM CONSTANT
                          TEMPERATURE HEAD BOX
                                 GRADUATED
                                 GLASS TUBE
                             y«i *
                         f      ^y__—10 LITER BOTTLE
                                                 NO 3
                                                                                            N2  CONTAINING H2S
                                                                                                      [—--• £o-RING JOINT
03
       2-STAGE GAS
       REGULATOR
2O LITER TEST x
SOLUTION
BOTTLE  -
          Figure 2.   Apparatus used for  distribution  between water  and nitrogen  and concentration of
                       molecular hydrogen  sulfide.   Insert shows top  view  of buret stopcock end  of the
                       concentration column.

-------
bottle.  An  inlet  tube  for  nitrogen  and  an  outlet  tube  for displaced
water are  inserted into the stopper  closing the  neck  of  the bottle.  A
two-way  teflon  stopcock controls  the flow of nitrogen into the bottle.
The  bottle need not be  removed  from  the  support  except  for cleaning
purposes,  because  it  can be filled through  the 3/4-inch  hole after each
run.

With proper  manipulation of the various  stopcocks,  the dispersed nitrogen
from the bubbler can  be passed  through the  concentration column and then
to the water displacement bottle.  By manipulation  of three stopcocks,
the  nitrogen flow  can be diverted from one  concentration column to
another, permitting continuous H2S determinations.

The  solution whose molecular H-S content is to be determined is placed
in a 20-liter carboy  and the circulating glass bubbler  is immersed in
the  solution.   The bubbler  is connected  to  the spray  trap, which is
connected  to the No.  1  three-way teflon  stopcock; all connections are
made by means of Buna-N 12/5 0-ring  joints.  The regulator on the nitro-
gen  cylinder is  opened  and  its pressure adjusted to 12 Ib/square inch.
The  needle valve on the  flow meter is then  adjusted so the compressed
nitrogen is  bubbled through the solution at a rate of usually between
25 and 50 ml/min.   The  No.  1 three-way teflon stopcock is positioned so
that the displaced  H2S will not pass  through the arm  of  the stopcock to
which the concentration  column is to  be connected, but instead will
escape through  the  third  arm, which  at that  time is open to the atmosphere.
Compressed nitrogen is  then bubbled  through  the test  solution for 30 min
to ensure equilibrium in  the system  before  collection of H-S is begun.

The  concentration  column  is prepared by coating the glass beads with 0.1
M zinc acetate.   This coating is applied with the concentration column
held in a vertical  position, with the stopcock end down.  A separatory
funnel,  containing  0.1 M  zinc acetate, is connected to the buret stop-
cock at the lower end of  the column by means of tygon tubing and an
                                    38

-------
0-ring joint.  By raising and lowering the separator? funnel,  the column
can be filled and drained.  When this procedure has been repeated four
 times, the  0-ring joint  is disconnected and the excess  zinc acetate  is
 allowed  to  drip  out  of  the column. .  By weighing the  concentration column
 before and  after coating, the volume of 0.1 M zinc acetate adhering  to
 the  glass beads  and  column wall was  determined to be about 0.5  ml.
 Slight variations in the amount of  zinc acetate remaining in  the column
 were demonstrated to have no effect  on  intensity  of  color development
 when the methylene blue method is used  for the determination  of sulfide.
 The column  is then placed in position between the No. 1 and No. 2 three-
 way stopcocks and the 0-ring joint at each end of the column  is secured
 with a metal clamp.   The concentration column is  connected in series with
 a water  displacement bottle by rotating,  in the appropriate  manner,  the
 three-way teflon buret stopcock and the No.  2 and No. 3 teflon stopcocks,
 and by opening the two-way teflon stopcock in the nitrogen inlet tube of
 the water displacement bottle.  The No. 1 three-way  teflon stopcock then
 is rotated so the H»S equilibrated nitrogen coming from the bubbler
 passes through the concentration column and then continues through  the
 system to displace water from  the displacement bottle.  In the  concen-
 tration  column the H-S  in the  nitrogen reacts with the zinc acetate to
 from zinc sulfide.

 At  the end  of the concentration period,  the  column  is  removed  from  its
 collecting  position  and sulfide is  analyzed  according  to a procedure
 described  previously.   The  H»S concentration  in  the  tested solution is
 derived  by  reference to a standard  curve relating pg of  H_S  displaced
 per liter  of nitrogen  dispersed with the known H S  concentration in
 the standard solutions. The standard H-S solutions were prepared by
 diluting a known amount of  Na2S-9H20 with deoxygenated deionized water
 and lowering the pH of the  test  solution with HC1  to about 4 or 5,  where
 essentially all the sulfide present is as molecular H-S.  Since the
 quantity of H-S collected  on the column  is  directly related  to the  H2S
 concentration  in the test  solution, the  latter  concentration can be
                                      39

-------
determined by reference to a calibration curve defined during this study.

In making a molecular H2S determination, it is necessary to know pre-
cisely the amount of nitrogen that has passed through the test solution
and concentration column.  Since gas regulators do not always deliver
precisely the desired, sustained flow, it was decided to measure the
nitrogen volume by means of a water displacement bottle (Figure 2).  The
inverted bottle is filled through the 3/4-inch hole in its bottom with
deionized water at room temperature.  The graduated glass tube is then
inserted into this 3/4-inch hole and the rubber cap is removed from the
S-shaped outlet tube.  The water in the graduated glass tube drops to
the level of the discharge end of the outlet tube and the displaced
water is discarded.  The two-way stopcock on the inlet line now can be
opened and the nitrogen passing through the concentration column is
allowed to enter the displacement bottle.  The displaced water flows
out through the S-shaped outlet tube and is collected.  At the end of
the concentration period, the two-way inlet stopcock is closed and a
rubber cap is placed over the outlet tube opening.  The displaced water
is measured and corresponds to the total uncorrected volume of nitrogen
dispersed.  Water is then introduced into the graduated tube until the
water level in the tube rises to the water level in the 10-liter dis-
placement bottle.  The amount of water added by way of this tube is
referred to here as the "crude correction volume1.'  To obtain the "true
correction volume',1 it is necessary to determine how much water would be
required to fill the graduated tube to the level of water in the dis-
placement bottle if the ".orrection tube were sealed off at the bottom.
This value is called the "tube correction" and will not exceed 10 ml.
The true correction is equal to the crude correction minus the tube
correction.  When the level of water in the tube equals the level of
water in the displacement bottle, the gas above the water must be at
atmospheric pressure, because the correction tube is open to the atmos-
phere at the top.  Therefore the total volume of nitrogen dispersed is
equal to the total uncorrected volume of nitrogen dispersed minus the
true correction volume.
                                   40

-------
ACUTE SULFIDE BIOASSAYS
Experimental Water
Experimental water used in all sulfide bioassays came from a laboratory
well which draws water from the Jordon Sandstone stratum underlying the
Minneapolis-St. Paul metropolitan area.  Chemical analysis of the well
water by the Minnesota Department of Health, Division of Environmental
Health, and the National Water Quality Laboratory, Duluth, Minnesota is
given in Table 2.

Experimental Fish
Juvenile fathead minnows  (Pimephales promelas Rafinesque) were used as
test organisms to study the toxicity of solutions containing sulfide at
various pH values.  The fathead minnow was chosen as an experimental
organism because it can be cultured and maintained in a laboratory as
well as handled with ease, and because it has a wide distribution in
various chemically diverse natural waters from acid bog lakes  (Dymond
and Scott  ) to lakes  of  high pH  (Rawson and Moore,  McCarraher  and
      52
Thomas  ).

The juvenile fathead minnows used in  all the bioassays were  cultured
in the  fisheries  laboratory at the University of  Minnesota,  St.  Paul.
Our culture was originally started in January 1972 with  fathead  minnows
from the  U.S.  Environmental Protection Agency's National  Water Quality
Laboratory in  Duluth,  Minnesota.  It  was hoped  that  by  using an  inbred
laboratory cultured  strain of  fish a  consistent  sensitivity  to sulfide
would be  observed for  tests at different  times with  fish of  different
stocks  and  that  possible  effects  of disease stress  and/or treatment  of
wild fish stocks  would be eliminated.

Fathead minnows  used in all  tests were cultured  under a constant  photo-
period  in 30-liter glass  aquaria receiving a continuous supply of labora-
 tory well water  at 25 C and with pH of approximately 7.9.  Six lots of
 fish were tested at various  times during a 16-week period.  Fish in
                                     41

-------
                                                         a/
              Table  2.  ANALYSIS  OF LABORATORY  WELL  WATER-'
                           (milligrams/liter)
     Item
Concentration
Total alkalinity as CaCO,
Total hardness as CaCO_
Calcium as CaCO_
Magnesium as CaCCL
Iron
Chloride
Sulfate
Sulfide
Fluoride
Total phosphates
Sodium
Potassium
Copper
Manganese
Zinc
Cobalt, nickel
Cadmium, mercury
Ammonia nitrogen
Organic nitrogen
   230
   220
   140
    70
     0.02
     0.0
     0.22
     0.03
     6
     2
     0.0004
     0.0287
     0.0044
   < 0.0005
   < 0.0001
     0.20
     0.20
a/
— Water taken from well head after iron removal and before aeration
and heating; pH 7.5.
                                  42

-------
separate bioassays ranged in age from approxiamtely 12 to 15 weeks,  in
mean total length from 27.7 to 34.1 mm, and in mean weight of survivors
from 0.227 to 0.422 g.

One week prior to a bioassay fish were removed from rearing tanks and
transferred to 20-liter aquaria.  Well water at 20 C was introduced  to
each aquaria at the rate of 0.5 liter/min and throughout an acclimation
period.  The oxygen concentration during acclimation was maintained
above 6 mg/liter.  The fish were fed Oregon moist and Glencoe pelleted
food twice daily until 1 day prior to exposure to sulfide.

Experimental Apparatus and Conditions
Acute toxicity bioassays were performed in three identical diluter units
each including one control and four treatment chambers.  The test cham-
bers were constructed of double-strength window glass and General Electric
RTV adhesive, measure 50 x 25 x 20 cm deep, and contained 20 liters of
test solution.  The cyclic water-delivery and toxicant systems were modi-
fied from that described by Brungs and Mount   and Mount and Warner,
respectively.  Flow through each chamber was at the rate of approximately
500 ml/min, affording 90% replacement in about 90 minutes.

The pH of the test water was controlled by dispensing a sulfuric acid or
sodium hydroxide solution with a "dipping bird" into the head reservoirs.
The temperature of the test water was thermostatically controlled at
20 C by a hot water stainless steel heat-exchange coil in each head
reservoir.  The test water was aerated in the head reservoirs to main-
tain oxygen concentrations in the test chambers at near 7.5 mg/liter.
Test chambers were illuminated for 12 hr each day with a 40-watt incan-
descent bulb 10 inches above each chamber.

Experimental Design and Procedure
The 96-hr bioassays conducted in this study were designed to determine
the relationship between pH of  test solutions and apparent H~S toxicity.
                                    43

-------
Between two and three acute tests were performed with fathead minnows
at each of six pH values, ranging between 6.5 and 8.7.  Each series of
tests consisted of three acute bioassays conducted in three sets of
experimental chambers at the same time and on juveniles from the same
lot.  The desired concentrations of H?S within each set of treatment
chambers were arranged in an appropriate series so that various degrees
of percentage mortality would be observed in the four treatments.  The
toxicant concentrations were randomly assigned to treatment chambers
before each set of bioassays.  Stock solutions of sodium sulfide were
prepared with reagent grade Na^S^I^O crystals and deionized water.
One pellet of reagent grade sodium hydroxide was added to each liter of
stock solution to raise the pH, thus retarding evolution of H2S from
the "dipping bird" reservoirs.

After the concentration and amount of stock sulfide necessary to give
the desired concentration of dissolved sulfide in the test chambers for a
given pH series were determined from a trial run without fish, the test
chambers were flushed with well water.  Three days before initiation of
the bioassay 10 fish were randomly assorted into each of the 12 treat-
ment and three control chambers.  Sulfuric acid or sodium hydroxide was
then slowly added to the head reservoirs to attain the desired pH.  The
fish were acclimated to the specified pH for at least 2 days before
introduction of the sulfide.

At the beginning of the bioassay the sulfide concentrations were raised
to the desired levels within a period of less than 1 hr.  During each
bioassay water temperature, dissolved oxygen, and pH in each test chamber
were measured daily.  Alkalinity was determined by potentiometric titra-
tion with a standard H2SO,  solution to the successive bicarbonate and
carbonic acid equivalence points, identified by the inflection in the
titration curve.  Dissolved oxygen was measured with a Winkler stan-
dardized galvanic-type membrane electrode meter and pH with a Corning
model 12 immersion-type glass electrode meter.  Dissolved sulfide con-
centrations,  which were determined to be essentially the same as total
                                    44

-------
sulfide concentration, were measured in treatment chambers at least
twice daily.  Water samples taken from the center of each test chamber
were stabilized with zinc acetate and analyzed for sulfide by a colori-
metric procedure previously described in section 228 B-4a (2) of Standard
Methods (APHA ).  The concentration of molecular H«S in each treatment
chamber was calculated for each sulfide determination using the daily
pH and temperature measurements, and the K,  (H_S) equilibrium constants
derived during this study.

The number of mortalities in each test chamber was recorded at 24-hr
intervals.  Total lengths of dead and surviving fish were measured, and
survivors were weighed.  At no time during these bioassays was any mor-
tality in control chambers observed.  Estimates of the concentration of
H-S most likely to cause 50 per cent mortality  (LC50) after 96 hr of
exposure were made in this study from lines  fitted mathematically by
the BMD03S probit analysis computer program  to plots of percentage mor-
tality against log H2S concentration  (Dixon   ).  The 96-hr LC16 and LC84
values were also calculated from the probit  analysis regression equations
and 95% confidence intervals for LC50 values were computed according to
                                             C£       O
formulas proposed by Litchfield and Wilcoxon.    (Chi)  tests were applied
to each group  of data to determine variability  and acceptability.
                                     45

-------
                              SECTION V
                     RESULTS AND INTERPRETATIONS

DETERMINATION OF SULFIDE IN AQUEOUS SOLUTION
lodometric Sulfide Determination
The salt Na~S'9H?0 has been widely used as a source of sulfide in the
preparation of standard solutions.  Because there may be situations
where the sulfide concentration in stock solutions cannot be analyti-
cally determined, definition of the relationship between concentration
based on a weight measurement and that determined quantitatively by
titration may be useful.  A number of stock sulfide solutions containing
about 100 rag/liter as S (0.75 g Na2S-9H20/liter) were prepared and the
sulfide concentrations were determined by an iodate-iodide procedure
with thiosulfate titration using starch to detect the end point.  The
values obtained were then compared with the calculated concentrations
determined on a weight and volume basis.  In general, the weighed Na_S'
91^0 standards were quite accurate as determined by iodometric titration.
When 20 ml of the sulfide stock was added to approximately 100 ml of
boiled and cooled deionized water, followed by the addition of concen-
trated HC1 and 10 ml of standard iodine solution, the average percentage
of sulfide determined compared to that calculated for 29 samples was
97.18 + 0.94.  When the concentrated HC1 was added first to approximately
100 ml of boiled and cooled deionized water, followed by the addition of
the standard iodine and then stock sulfide solutions, the average per-
centage of sulfide determined compared to that calculated for 29 samples
was 98.17 + 0.59.  Therefore, when unstabilized stock sulfide solutions
                                    46

-------
are prepared on a weight and volume basis, the calculated sulfide con-
centrations should be about 2 to 3 per cent higher than the actual
values.  These data suggest that the best procedure for iodometric
standardization of unstabilized sulfide is acidification of deionized
water with concentrated HC1, followed by addition of the standard iodine
solution and then the stock sulfide, and titration of excess iodine with
thiosulfate.  The volume of concentrated HC1 added to the deionized
water before addition of the iodine and stock sulfide is not critical
and can vary from 2 to at least 10 drops with no change in titrant
volume.

The stability of stock sulfide solutions prepared with Na2S'9H20 was
examined over a 190-hr period for solutions prepared with boiled and
cooled deionized water (deoxygenated) and initially containing about
100 mg/liter as S.  The solutions were prepared and stored under an
atmosphere of nitrogen in 1-liter Pyrex glass volumetric flasks and
either exposed to laboratory light, kept in the dark, made alkaline to
0.1 N NaOH, or stabilized with zinc acetate to 0.025 M Zn.  Iodometric
analysis at various times after preparation on aliquots of the stock
solutions indicated that storage in light or dark had no effect on the
rate of degradation.  The decrease in sulfide concentration occurred
at a rate of about 0.025 mg/liter per hour as S following preparation.
Addition of base to a final concentration of 0.1 N NaOH essentially
doubled the rate of sulfide degradation to about 0.048 mg/liter per hour
as S following preparation.  The addition of zinc acetate had no effect
on the iodometric titration when a 20-ml aliquot of sulfide solution
in 100 ml of deionized water was acidified before addition of iodine.
The solution is turbid white before addition of acid but the end point
is the same as for similar sulfide solutions in which no zinc was present.
If acid is added following the iodine, the volume of thiosulfate required
to titrate the remaining iodine is somewhat less than when the acid is
added before the iodine.  The addition of zinc acetate to the stock
sulfide solution to a concentration of 0.025 M Zn stabilized the sulfide
                                     47

-------
so that over a 190-hr period essentially no decrease in sulfide concen-
tration could be detected.

Colorimetric Sulfide Determination
The stabilization of sulfide solutions with certain metal salts does
not appear to interfere with sulfide determination by the colorimetric
method utilized during this investigation and may even under some cir-
cumstances enhance the intensity of color by preventing sulfide degra-
dation before color development can be accomplished.  When 0.5 ml of a
zinc acetate solution (0.1 M Zn), zinc chloride plus gelatin (0.1 M Zn) ,
or cadmium hydroxide suspension (0.02 M Cd) were added to a known amount
of sulfide, the amount of sulfide determined was essentially the same as
that calculated to be present.  However, when 0.5 ml of a 0.1 M cadmium
sulfate or cadmium chloride solution was added to a known amount of
sulfide, the amount of sulfide determined was about 4 and 8 per cent
less, respectively, than the calculated concentration.  It appears that
under the conditions of color development employed during this study,
these latter cadmium salts bind sulfide to such an extent that not all
of the sulfide is released during the color development phase.

The colorimetric diamine reaction for sulfide determination is known to
be pH-dependent and different authors have proposed various sulfuric acid
molar concentrations at which maximum color production is realized.  The
molar acidity values discussed below merely represent calculated values
based on dilutions of reagents containing sulfuric acid.  The values
probably do not represent hydrogen ion concentration since other reagents
such as the diammonium hydrogen phosphate will buffer the sulfuric acid.

An attempt was made to define the molar acidity of reaction and measure-
ment that would maximize color development when the reaction volume is
6.5 ml and that of measurement is 25 ml.  No attempt was made to deter-
mine these precise conditions but it was felt that in order to have a
reproducible and sensitive sulfide method, color production should be
                                    48

-------
nearly optimal.  When utilizing the N,N-dimethyl diamine reagent the
optimum color development occurred with absorbance determined at 666
nm when reaction molar acidity was about 0.75 to 1.0 and the measure-
ment molar acidity was 0.25 to 0.26.  However, when the quantity of
N,N-dimethyl diamine reagent was varied from 0.8 to 1.1 ml with the
resulting reaction and measurement molar acidities varying from 0.89 to
1.22 and 0.23  to 0.32, respectively, the amount of color development
produced was essentially unchanged.  Since color development was also
not markedly affected when the volume of diammonium hydrogen phosphate
was kept between 1.25 and 1.75 ml,  slight variations in the amount  of
reagents added do not appear  to affect  color  development significantly.

Use of  the N,ethyl-N-hydroxyethyl-  or N,N-diethyl diamine reagents
produced near  optimum color development with  absorbance determined  at
668 or  672 nm  when the molar  acidity of reaction was about  0.75 or  0.25
and the molar  acidity of measurement was 0.20 to 0.25  or 0.20,  respec-
tively.  The minimum molar acidity  of measurement necessary to  produce
a  clear solution in a total volume  of 25 ml  following  the addition  of
1.5 ml  of diammonium hydrogen phosphate is 0.18.  According to  Rees et
   32
al.,  the optimum color development  when using the N,N-diethyl diamine
reagent is realized at a reaction and measurement molar acidity of  about
0.09 without  the addition of  the  phosphate reagent.   If the molar
acidity of reaction was 0.25  and  the molar acidity  of  measurement  0.20,
the optimum amount of diammonium  hydrogen phosphate necessary  to produce
maximum color  development using the N,N-diethyl diamine reagent was
determined in  this study  to be 1.5  ml.  However,  the color  development
is not  markedly affected when the volume of  phosphate  is kept between
1.25  and  1.75  ml.

Sulfide calibration  curve  solutions were prepared by dilution with
freshly boiled and cooled  deionized water of a stock solution  of known
concentration determined  iodometrically.  Each curve was defined  im-
mediately after  the  preparation of  a  sulfide standard  solution. Addition
of 0.5 ml zinc acetate (0.1 M  Zn)  to each flask stabilized  the  sulfide.
                                     49

-------
The regression equations  summarizing  the calibration curves prepared
with the various diamine  reagents and under different calculated acidity
conditions when the diamine and iron reagents are added separately or
in a 9 to 1 mixture are presented in Table 3.  In the regression analysis
X refers to sulfide concentration expressed as yg of H-S in a total
volume of 25 ml.  All absorbance values (Y), which never exceeded 1.0
unit, were corrected for  a reagent blank and were determined at 666 nm
for N,N-dimethyl, 668 nm  for N,ethy1-N-hydroxyethyl, and 672 nm for
N,N-diethyl diamine reagents.  If a known amount of sulfide was added
to a concentration column where color development occurred, the resul-
ting absorbance readings were essentially the same as those determined
for similar sulfide solutions in which color development was in a 25-ml
volumetric flask.  The slope of an N,N-diethyl diamine reagent calibra-
                                                       32
tion curve determined from data published by Rees et al.  was calculated
to be 0.07120.  This value applies when the reaction and measurement
molar acidity is 0.09, no diammonium hydrogen phosphate added, sulfide
expressed as yg H?S in a  total volume of 25 ml, and absorbance deter-
mined for a 10 mm light path.  When compared with the optimum slope of
0.06081 (Table 3) determined during the present study with N,N-diethyl
diamine reagent, Rees1 method is 1.17 times more sensitive to sulfide.
However, this increase in sensitivity can only be attained by omitting
the diammonium hydrogen phosphate, which may not be advisable in some
instances, and by color development at reduced acidity.

When the various diamine  reagents and iron chloride are added to the
standard solutions separately, the color development is essentially
unaffected by the age of  the reagents over a period of at least 1 or 2
months.  However, when the various diamine reagents and iron chloride
solutions are added as a  9 to 1 mixture, color development in a stan-
dard sulfide solution is markedly affected by the age of the mixed
reagent.  With the N,N-dimethyl diamine reagent,  used when the reaction
molar acidity was 1.0 and measurement molar acidity 0.26, a given mixture
will produce maximum color development for a period less than 30 min
after preparation.   With  the N,ethyl-N-hydroxyethyl and N,N-diethyl
                                    50

-------
 Table 3.  LINEAR REGRESSION ANALYSIS OF CALIBRATION CURVES RELATING
  ABSORBANCE (Y) AND SULFIDE CONCENTRATION (X) IN>UG H2S PER 25-ML
   FOR SOLUTIONS PREPARED WITH VARIOUS DIAMINE REAGENTS AND UNDER
                    DIFFERENT ACIDITY CONDITIONS
Diamine    Molar acidity
and iron           Meas-  Calibration
reagents   Reac-   ure-   curves in
added	tion    ment   regression
 Linear regression analysis—'
	(Y = A + BX)
   A       B        r
                           a/
               N,N-dimethyl-p-phenylene diamine oxalate
Separately  1.0    0.26       8         0.0094  0.04724  0.9992  0.0085
Mixture     1.0    0.26      12         0.0109  0.04117  0.9990  0.0130

          N,ethyl-N-hydroxyethyl-p-phenylene diamine sulfate
Separately  0.75   0.20       4
Separately  1.0    0.26       8
Mixture     1.0    0.26       2
0.0080  0.05908  0.9996  0.0070
0.0062  0.05880  0.9998  0.0053
0.0102  0.06756  0.9997  0.0094
               N,N-diethyl-p-phenylene diamine oxalate
Separately  0.25   0.20       2         0.0122  0.06081   0.9990  0.0125
Separately  0.75   0.20       8         0.0119  0.05816   0.9989  0.0120
Mixture     1.0    0.26       2         0.0207  0.05594   0.9989  0.0154
__                                                             ——
— Reagent blank corrected  absorbance  values  of less  than 1.0  and for  a
light path of  10 mm were determined at 666,  668,  and 672 nm when cali-
bration  curves were prepared with  the N,N-dimethyl,  N,ethyl-N-hydroxy-
ethyl, and N,N-diethyl-p-phenylene diamine reagents, respectively.
— Standard error of the estimate or deviation of  Y for  fixed  X.
                                   51

-------
diamine reagents, used when the reaction molar acidities are 0.75 and
measurement molar acidities are 0.20, the mixtures will produce maximum
color development for at least 4 hr after preparation.  With the N,N-
diethyl diamine reagent, used when the reaction molar acidity is 0.25
and measurement molar acidity is 0.20, the mixture will produce maximum
color development for only a few minutes after preparation.  Therefore,
it appears that for reproducible and maximum color development sulfide
solutions should be stabilized with zinc and the diamine and iron
reagents should be added separately.  An exception to this generaliza-
tion would be with use of the  N,ethyl-N-hydroxyethyl diamine reagent.
In this case the intensity of color development was greater at compa-
rable acidity levels when the reagents were added in a mixture than
when added separately.

Concentration Column Sulfide Absorbants
The efficiency of H~S collection was studied by using two zinc acetate
coated glass bead concentration columns placed in series with aliquots
of standard sodium sulfide contained in the total sulfide reaction
flask.  The H2S uptake by each column was measured after 60 min of
nitrogen stripping at 270 ml/min and the percentage efficiency of cap-
ture calculated.  Since no detectable amount of sulfide was collected
in the second column, it is assumed that all of the H-S reaching the
first column is absorbed.  There was no effect on amount of sulfide
displacement and collection efficiency when the time of displacement
varied between 60 and 180 min, however 30 min was inadequate.  The
recovery of a known amount of sulfide added to a total sulfide reaction
flask by the first concentration column when coated with different metal
salts was demonstrated to be reproducible but incomplete.  The results
summarizing the degree of H2S recovery by various metal salts, as indi-
cated by the percentage of sulfide collected compared to that added to
a reaction flask, are presented in Table 4.  The sulfide absorbants
prepared with zinc salts resulted in the highest recovery of sulfide
when compared with those containing cadmium.  The less than 100 per cent
                                   52

-------
      Table 4.  RECOVERY AND STABILIZATION OF H2S BY GLASS BEAD
        CONCENTRATION COLUMNS COATED WITH VARIOUS METAL SALTS
Sulfide
absorbant
Number
  of
tests
Range in sulfide
              a/
concentration,—
                  Percentage sulfide
                  recovered
                  Mean
                                  SD
Zinc acetate
  (0.1 M Zn)
Zinc chloride plus
  gelatin (0.1 M Zn)
Cadmium hydroxide
  suspension (0.02
  M Cd)
Cadmium chloride
  (0.1 M Cd)
Cadmium sulfate
  (0.1 M Cd)
  19
3.00-15.68
             5.18-10.80
            10.31-10.58
            10.61-11.83
             5.28-16.17
                     98.3
                  97.4
                  93.2
                  95.3
                  86.8
3.7
                                  3.7
                                  2.2
                                  2.1
                                  2.6
a/
— Known amounts of sulfide were added to a  total  sulfide reaction  flask.
                                   53

-------
sulfide recovery may have resulted from loss due to oxidation, adhesion
to glass walls between the reaction flask and concentrations column, and
transition to sulfides insoluble in acid due to impurities in reagents
or deionized water.  The stability of sulfide collected in a zinc acetate
coated concentration column was determined by storing the columns in
the dark or exposing them to laboratory fluorescent light for periods
of 2 to 6 hr.  No decline in sulfide recovery with storage time was
observed.

Previous experiments have indicated that some cadmium salts may combine
with sulfide and result in a non-quantitative release of sulfide during
the methylene blue colorimetric test and consequent reduction in sulfide
determined in comparison with quantities known to be present.  Therefore,
the reduction in percentage of sulfide collected from the amount added
to the total sulfide reaction flask may be due in part to decreased
efficiency in sulfide collection but is more than likely due to inter-
ference in the colorimetric test for determination of the cadmium
stabilized sulfide.

The stability of various metal sulfides was determined by passing air
over sulfide collected in a concentration column and comparing the per-
centage of sulfide collected following exposure to various amounts of
oxygen to that added and displaced by nitrogen from a total sulfide
reaction flask (Table 5).  When compared with the data in Table 4, it is
apparent that none of the metal stabilized sulfides are affected by
exposure to relatively large amounts of air.  Because nearly 100 per
cent recovery and complete stabilization of displaced H2S could be
attained by the zinc acetate sulfide absorbant, it was used as the metal
salt for coating of concentration columns in subsequent experiments, the
results of which are presented below.

Known amounts of sulfide were added to total sulfide reaction flasks,
displaced with nitrogen and collected on zinc acetate coated concentra-
tion columns.  These columns were then connected to carboys containing
                                    54

-------
   Table 5.  STABILITY OF METAL SULFIDES ON CONCENTRATION COLUMNS
                        TO OXIDATION BY AIR

H2S added
Air flow, to reaction
ml/min for Oxygen,— flask,
60 min moles ug

100
200
400

100
200
400

0.054
0.107
0.214
Zinc
0.054
0.107
0.214
Zinc acetate (0
10.43
10.39
10.43
H2S col- Sulfide col-
lected on lected after
column, exposure to air,
Jig %
.1 M Zn)
10.28
10.30
9.91

98.6
99.1
95.0
chloride plus gelatin (0.1 M Zn)
10.57
10.54
10.52
10.54
10.03
10.35
99.7
95.2
98.4
Cadmium hydroxide suspension (0.02 M Cd)
100
200
400

100
200
400

100
200
400
0.054
0.107
0.214

0.054
0.107
0.214

0.054
0.107
0.214
10.39
10.35
10.31
Cadmium chloride
10.65
10.61
10.63
Cadmium sulfate
10.77
10.77
10.73
9.40
9.84
9.77
(0.1 M Cd)
10.37
10.23
10.06
(0.1 M Cd)
9.62
9.35
9.01
90.5
95.1
94.8

97.4
96.4
94.6

89.3
86.8
84.0
a/
  Assume air is 20% 02 and 1 mole 02 occupies 22.4 liters.
                                   55

-------
20 liters of aerated well water and nitrogen was passed over the columns
after being dispersed through the water at approximately 140 ml/min for
4 to 5 hr.  The initial and final dissolved oxygen concentrations in the
carboys were about 10 and 0.3 mg/liter, respectively.  The oxygen
displaced from the well water and passed over the sulfide collected on
the concentration columns had no effect on the stability of the sulfide
collected since essentially 100 per cent of the sulfide added to the
total sulfide reaction flasks was subsequently determined on the concen-
tration columns.  The maximum amount of dissolved oxygen displaced from
20 liters of water containing 10 mg/liter 02 is 0.00625 mole.  This is
considerably less than the amount of oxygen demonstrated (Table 5) to
have no adverse effect on zinc stabilized sulfide.  Therefore, oxidation
of zinc stabilized sulfide on a concentration column by oxygen displaced
from test solutions would be negligible.

Because the recovery of sulfide by the indirect total sulfide method is
incomplete though reproducible, it was necessary to prepare a calibra-
tion curve from standards varying in volume from 50 to 200 ml and con-
taining HpS from 2.71 to 11.66 pg following the same procedure as for
the samples.  The standards were prepared one at a time and carried
through the sulfide evolution and collection procedure.  The relation-
ship between yg H2S (X) added to the reaction flask and that collected
(Y) on a zinc acetate coated concentration column was determined from
18 standard solutions and can be defined with r = 0.994 by the regression
equation:

                        Y = -0.156 + 0.984 X.                      (10)

The overall average percentage of H-S collected compared with that added
was 95.7 + 4.7 per cent.  From the above equation the yg of sulfide
expressed as H_S in a sample of known volume can be calculated from the
yg of l^S collected in a concentration column following recovery by the
indirect total sulfide method.  If a sample is stabilized with zinc
                                   56

-------
acetate, analysis for total sulfide by the indirect method can be made
without any decrease in percentage recovery with samples stored for at
least 5 hr.  However, sulfide stabilization should occur in a reaction
flask because when performed in a volumetric flask and then transferred
to the reaction flask, a lower percentage recovery was observed.

DIRECT DETERMINATION OF MOLECULAR H2S IN AQUEOUS SOLUTION
The precision with which the vapor phase equilibration method can be
used for the direct determination of molecular H«S depends in large
part on the effect which the height of the nitrogen bubbling column
and the rate of nitrogen dispersion have on the yg of H»S displaced per
liter of nitrogen dispersed.  Experiments conducted with six different
bubbling depths demonstrated that there is only a slight effect on H_S
displacement rate when the water column in a bubbler varies in height
from 13 to 95 cm.  During any single determination of molecular H~S the
maximum variation in column height ranged from about 35 to 50 cm.  There-
fore, at no time did the rate of H_S displacement have to be corrected
for bubbling column height.  It was also demonstrated that bubble size
was not critical since no measurable difference in displacement rate of
H^S was observed when the nitrogen was dispersed through a medium or a
coarse gas dispersion frit.  When the nitrogen dispersion rate  ranged
from 25 to 200 ml/min, the yg of H«S displaced per liter of nitrogen
dispersed was the same for solutions containing essentially equal H2S
concentrations.  Therefore, as long as the total volume of nitrogen dis-
persed through a test solution is known, no correction has to be made
for the rate of dispersion over the tested range.

The rate at which a  gas in solution can be displaced  is a function of
the solution's temperature.  The rate at which H^S is displaced per
liter of nitrogen dispersed was determined for a number of test solu-
tions containing different H~S concentrations ranging from about 5 to
200 yg/liter and at  temperatures of 10, 15, 20, or 25 C.  A summary of
these data is presented in Table 6 where partition coefficients relating
                                    57

-------
Table 6.  H2S DISPLACEMENT BY NITROGEN DISPERSED THROUGH TEST
SOLUTIONS OF KNOWN MOLECULAR H_S CONCENTRATION AND TEMPERATURE

Temper-
ature,
C
10.0
10.0
10.0
10.2
10.3
10.1
10.0


15.1
15.1
15.1
15.1
15.2
15.2
15.2


20.0
20.0
20.0
20.0
20.0
Determined H9S
a/
in solution, —
/ig/1
5.039
10.75
25.88
50.86
80.68
145.6
215.7


4.516
9.756
25.15
56.52
83.12
153.3
198.0


4.592
9.600
24.83
46.30
69.36
H-S displaced/
1 N2 dispersed,
/ig
1.243
2.852
6.959
13.61
20.34
37.66
56.33


1.297
2.964
7.304
16.38
25.51
45.74
62.69


1.613
3.208
8.315
15.13
23.10
Partition coefficient—
>ug H0S displaced/1
NO
,ug/l H2S in solution log,Q
0.2467
0.2653
0.2689
0.2676*
0.2521
0.2587
0.2611
Mean 0.2601
SD 0.0082
CV 3.15
0.2872
0.3041
0.2904
0.2898*
0.3069
0.2984*
0.3166
Mean 0.2991
SD 0.0108
CV 3.61
0.3513
0.3342
0.3349
0.3268
0.3330
-0.6078
-0.5763
-0.5704
-0.5725
-0.5984
-0.5872
-0.5832


-0.5418
-0.5170
-0.5370
-0.5379
-0.5130
-0.5252
-0.4995


-0.4543
-0.4760
-0.4751
-0.4857
-0.4776
                               58

-------
 Table 6 (continued).   H2S DISPLACEMENT BY NITROGEN DISPERSED THROUGH

 TEST SOLUTIONS OF KNOWN MOLECULAR H2S CONCENTRATION AND TEMPERATURE

Temper-
ature,
C
20.0
19.9


24.9
25.0
24.9
24.9
24.9
24.9
24.9
25.0


Determined H9S
a/
in solution, —
jug/1
127.0
218.1


5.143
8.861
25.57
42.45
47.51
82.15
155.5
191.7


H_S displaced/
1 N- dispersed,
MB
46.21
83.15


2.079
3.492
10.50
17.93
18.57
33.49
62.67
78.33


Partition coefficient—
/ig H0S displaced/1 N
/ag/1 H_S in solution
0.3639
0.3812
Mean 0 . 3465
SD 0.0199
CV 5 . 74
0.4042
0.3941
0.4106
0.4224*
0.3909
0.4077
0.4030
0.4086
Mean 0.4052
SD 0.0098
CV 2.42
-2
Iog10
-0.4390
-0.4188


-0.3934
-0.4044
-0.3866
-0.3743
-0.4079
-0.3896
-0.3947
-0.3887


a/
7-/PH values of sulfide test solutions were between 4 and 5.
—Partition coefficients marked with an asterisk correspond to solutions
prepared with deoxygenated well water; all others prepared with deoxy-
genated deionized water; SD - standard deviation; CV - coefficient of
variability (%).
                                    59

-------
the yg H9S displaced per liter nitrogen dispersed to the known ug/liter
H9S concentration in the test solution are calculated.  The partition
coefficients (Y) were log linear with respect to temperature (X) and a
regression equation defining this relationship with r = 0.9771 is:
                     log Y = -0.7188 + 0.01301 X.                  (11)

This relationship was subsequently used to calculate the concentration
of H,,S in a test solution when temperature and yg of H-S displaced per
liter nitrogen dispersed through the solution were known.  The rate at
which H?S is displaced from solution is independent of whether the solu-
tions are prepared with deionized or well water, so the above regression
equation is applicable to all test solutions prepared during this study.
EQUILIBRIUM CONSTANTS FOR THE FIRST DISSOCIATION OF H0S/  ,
                                                     2  (aq)
                                   2-
Assuming the concentration of the S   ion to be negligible in solutions
prepared during this study, the relationship between dissolved sulfide
species and pH for a test solution at particular temperatures can be
defined from the first dissociation constant of H0S,  N (equation no.
                                                 2 (aq)
2).  Therefore, the apparent first dissociation constants and pK... values
of H2S,  , were determined at 10, 15, 20, and 25 C.  The test solutions
used contained different total sulfide concentrations,  ranging from
about 25 to 2,800 yg/liter as H2S, prepared with either deoxygenated
deionized water or well water and having various pH values ranging from
about 6.1 to 8.7.  The pH of these solutions was controlled by the addi-
tion of 200 ml of the appropriate 1/15 M phosphate buffer and small
amounts of weak H~SO, or NaOH to a total volume of 20 liters.  Measure-
ments of temperature, pH, total sulfide, and H2S displacement rate were
made on each test solution.  From the relationship between the partition
coefficients and temperature, previously described in equation no. 11,
the molecular H_S concentration in each test solution could be calcu-
lated.  A summary of these results is presented in Table 7.
                                    60

-------
Linear regression analyses of the K  dissociation constants and pK..
values in Table 7 for temperatures ranging from 10 to 25 C are presented
in Table 8.  The slight difference in the linear regression equation for
solutions prepared either with deionized or well water may be due to
differences in test solution ionic strength or from the presence of
minute amounts of total sulfide occurring as metal sulfides in test
solutions prepared with well water.  The well water test solutions
represent natural waters of fairly high alkalinity and hardness.  There-
fore, the linear regression equation for the combined data from deionized
and well water test solutions is proposed for defining the relationship
between the apparent first dissociation constant of H^S,  , and tempera-
ture and would be applicable to most freshwaters of low ionic strength.
The relationship between pK.. and temperature  (T) in degrees Celcius  (C)
gives the best fit by linear regression analysis, therefore the equation

                     pK  = 7.252 - 0.01342 T  (C)                    (12)
was used  to define  the  first dissociation  constant  of  tLS,   x  at  various
                                                        2  (aq)
temperatures  in  subsequent  calculations.   Data  used in calculating  this
equation  were obtained  from solutions  prepared  over the temperature
range  of  10 to 25 C.  However, because of  the "good fit"  of these data
to the linear regression  equation,  it  was  felt  that extrapolation to
other   temperatures,  as justified  from the relationship presented in
Figure 1, would  permit  acceptable  predictions of pK..  at temperatures
ranging from  at  least 5 to  30  C.

By calculation it can be  demonstrated  that the relationship between
 [HS]/ [dissolved sulfide] is equal to  the  factor [H+]/K  + [H+] ,
   2.         2_                                          -1-
assuming  [S   ] to be  negligible.   When both the molecular H-S  and dis-
solved sulfide concentrations  are  expressed as H-S  and in the  same  units
 (i.e., yg/liter  H~S), dissolved  sulfide as yg/liter H?S times  the factor
will  equal molecular  H^S  in yg/liter.   Factors presented in Table 9
define the fraction of  dissolved sulfide as molecular H^S at various
temperatures  from  5 to  30 C in 1.0 degree  intervals and pH values from
                                    61

-------
OS
to
       Table  7.  APPARENT  ^  DISSOCIATION CONSTANTS AND pKj_ VALUES OF H2S(aq)  DETERMINED FOR TEST SOLUTIONS


                     OF  DIFFERENT TEMPERATURES,  pH VALUES,  AND TOTAL SULFIDE CONCENTRATIONS
Temper- Determined
ature, total sulfide
C pH jug/1 as H^S
H2S displaced/
liter N2 Determined
, dispersed, Partition H S,
/ ^
yUg coefficient— xie/1
Deionized water
10.0
10.0
10.0
10.1
10.2
10.2
10.0

10.1
10.1
10.0
9.9


6.235
6.270
7.066
7.021
7.678
7.674
8.497

7.509
7.674
8.581
8.516


25.52
85.55
47.96
156.9
168.0
598.3
1123

546.7
566.4
1370
2329


5.834
19.30
6.806
22.23
9.626
33.92
11.07

41.19
33.22
11.74
20.12


0.2578
0.2578
0.2578
0.2586
0.2594
0.2594
0.2578
Well water
0.2586
0.2586
0.2578
0.2571



22.63
74.85
26.40
85.98
37.11
130.8
42.95

159.3
128.4
45.53
78.28
Mean
SD
CV
Equilibrium constants—
K.-107 oK.
— 1 — 	 	
0.7422
0.7681
0.7014
0.7861
0.7406
0.7574
0.8009

0.7533
0.7222
0.7634
0.8764
0.7647
0.0462
6.04
	 «-_t 	
7.129
7.115
7.154
7.104
7.130
7.121
7.096

7.123
7.141
7.117
7.057
7.117
0.0255
0.359

-------
Table 7 (continued).  APPARENT    DISSOCIATION  CONSTANTS AND
                                                                               VALUES OF
                                                                                               .
CO
Temper-
ature,
c

15.0
14.9
15.0
15.0
14.9
15.0
14.9

15.2
15.1
15.0
14.9


Determined
total sulfide,
pH vug/1 as H^S

6.178
6.157
7.095
7.044
7.714
7.653
8.473

7.553
7.673
8.520
8.657


• — ™ 	 £ 	
31.14
83.09
57.80
141.9
155.7
570.9
1031

489.5
578.9
1198
2803


H2S displaced/
liter N2
dispersed, Partition
a/
>ug coefficient—

8.182
22.34
8.463
21.17
8.205
33.53
10.44

35.01
33.81
11.38
19.57


Deionized water
0.2995
0.2986
0.2995
0.2995
0.2986
0.2995
0.2986
Well water
0.3013
0.3004
0.2995
0.2986


Determined
H2S,
-ug/1

27.32
74.81
28.26
70.69
27.48
112.0
34.98

116.2
112.6
38.01
65.54
Mean
SD
CV
W
Equilibrium constants—
K -107

0.9289
0.7715
0.8400
0.9100
0.9012
0.9114
0.9582

0.8993
0.8797
0.9215
0.9200
0.8947
0.0505
5.65
P*l

7.032
7.113
7.076
7.041
7.045
7.040
7.019

7.046
7.056
7.036
7.036
7.049
0.0256
0.364

-------
             Table 7 (continued).   APPARENT Kn  DISSOCIATION CONSTANTS AND pKn VALUES OF H S
                                             1                               12 (aq)
05
Temper- Determined
ature, total sulfide,
C pH >ug/l as H0S

20.0
19.9
20.0
19.9
20.3
20.1
20.0

6.116
6.400
7.076
7.056
7.696
7.734
8.499

26.07
76.42
56.51
174.2
192.8
839.0
1163
H2S displaced/
liter N2
dispersed, Partition
a/
xig coefficient—

7
21
8
27
11
46
12

.985
.96
.633
.74
.12
.31
.89
Deionized
0
0
0
0
0
0
0
water
.3479
.3468
.3479
.3468
.3510
.3489
.3479
Determined
H2S,
jug/1

22
63
24
79
31
132
37

.95
.31
.81
.99
.68
.7
.06
Equilibrium constants—
K, -107
	 ± 	
1.041
1.037
1.072
1.035
1.024
0.9816
0.9628


6
6
6
6
6
7
7
PK,

.983
.984
.970
.985
.990
.008
.016
Well water
20.0
20.3
19.9
19.9


7.576
7.714
8.515
8.582


367.6
728.7
1143
2295


25
40
11
19


.81
.33
.55
.80


0
0
0
0


.3479
.3510
.3468
.3468


74
114
33
57


.18
.9
.30
.11
Mean
SD
CV
1.050
1.032
1.018
1.026
1.025
0.0303
2.96
6
6
6
6
6
0
0
.979
.986
.992
.989
.989
.0128
.184

-------
Table 7 (continued).  APPARENT    DISSOCIATION CONSTANTS AND
                                                                             VALUES OF
a>
en
Temper-
ature,
c
25.1
25.0
25.0
24.9
25.0
24.9
25.0

25.0
25.0
25.1
25.0


PH
6.307
6.311
7.059
7.063
7.688
7.674
8.472

7.534
7.557
8.424
8.518


Determined
total sulfide,
vug/1 as H^S
* ° 	 z 	
31.81
84.90
54.28
166.3
176.5
426.7
1115

474.7
497.1
1179
2256


H2S displaced/
liter N2 1
dispersed, Partition 1
a/
>ug coefficient—
10.40
27.13
8.944
28.87
10.07
26.13
11.86

39.71
38.58
13.22
21.47


Deionized water
0.4053
0.4041
0.4041
0.4029
0.4041
0.4029
0.4041
Well water
0.4041
0.4041
0.4053
0.4041


Determined
H?S, Equilibrium
>ug/l
25.66
67.14
22.13
71.66
24.93
64.86
29.35

98.26
95.46
32.62
53.13
Mean
SD
CV
K^IO7
1.183
1.293
1.268
1.142
1.247
1.182
1.247

1.120
1.167
1.324
1.258
1.221
0.0656
5.37
b/
constants—
PKX
6.927
6.889
6.897
6.942
6.904
6.927
6.904

6.951
6.933
6.878
6.900
6.914
0.0234
0.338
      —/Partition coefficients calculated from the equation log  Y -

      -'Calculated by assuming dissolved sulfide = [H2S] + [HS ] and KX = [H ] [HS ]/[H2S].

-------
 Table  8.   RELATIONSHIP  BETWEEN APPARENT KI  DISSOCIATION CONSTANTS AND
    p^ VALUES  OF  H-S,   ,  FOR TEMPERATURES  RANGING FROM 10 TO  25 C

Test
water

Delonized
Well
Combined

Delonized
Well
Combined
a/
Linear regression analysis—
(Y = A + BX)
A

0.4340-
0.4780-
0.4500-

7.261
7.238
7.252
B
Kl
10~7 0.03076-10"7
10~7 0.02882-10"7
10~7 0. 03005 -10~7
PK,
-0.01379
-0.01278
-0.01342
Correlation
coefficient,
r

0.9620
0.9495
0.9569

-0.9644
-0.9550
-0.9601
Standard error
of estimate,
Syx

0.04862
0.05324
0.05085

0.02107
0.02218
0.02180
a/
— Y refers to
                 or
values and X to temperature in degrees C.
6.0 to 9.0 in 0.1 unit intervals.  These factors were calculated from
equation 12.

A general acceptance by biologists of the apparently incorrect factors
                   2
proposed by Pomeroy  prompted preparation of a table to be used to con-
vert molecular H2S concentrations calculated from Pomeroy's factors to
normalized concentrations corresponding to values obtained if the factors
in Table 9 had been used originally.  Table 10 defines this relationship
between the fraction of dissolved sulfide as molecular H S derived in
this study (Table 9) and the fraction derived from Pomeroy's 1941 work
corresponiding to a "typical water supply" at different pH values ranging
from 6.0 to 9.0 and temperatures from 10 to 30 C.  Molecular H S concen-
trations have also in previous publications been calculated from constants
                                    66

-------
presented in Standard Methods from 1946 to 1965 (9th through 12th
editions).   These values can be converted to H^S concentrations based
on data collected during this study by multiplying the reported concen-
tration by the appropriate value in Table 11 at the corresponding pH.
The factors presented in the 9th through 12th editions of Standard
Methods of 0.29 at pH 7.1 and 0.23 at pH 7.3 do not correspond to the
original values as found in Table IV of Pomeroy's 1941 work of 0.28 and
0.20, respectively.  A similar correction can be made for H^S concen-
trations calculated from the factors presented in the 13th edition of
Standard Methods (APHA ).  Over the pH range of 6.0 to 8.8 and at 25 C
the ratio of fractions calculated in this investigation to the reported
values averaged 1.04 + 0.02.  Therefore, the fraction of dissolved sul-
fide as molecular H,jS as calculated from these most recent factors at
temperatures near 25 C would be very close to the values determined
from factors derived in this study.

H2S DETERMINATION IN VARIOUS WATERS AND EFFLUENTS
The feasibility of calculating molecular H-S concentrations in various
waters, spiked with known amounts of sulfide, from  the determined dis-
solved sulfide concentration and the fraction of dissolved sulfide as
H»S under the experimental conditions was evaluated by comparing cal-
culated values to direct determinations by the vapor phase equilibration
technique.  A summary of the tests performed in various  types  of waters
and effluents is presented in Tables 12 and 13, respectively.

The vapor phase equilibration technique for the direct determination of
molecular H»S is not strictly applicable  to certain test  solutions under
static conditions since the degradation of sulfide  occurs at  too rapid
a rate to estimate accurately initial H_S levels.   However, by making
repeated direct H2S determinations and interpolating, the molecular  H2S
concentration could be calculated at the  time of sampling for  total  or
dissolved sulfide.  The oxygenated well water  test  solution did not  show
a rapid decline in total sulfide during the vapor phase  equilibration
                                    67

-------
               Table  9.  FRACTION OF DISSOLVED SULFIDE AS MOLECULAR H2S IN AQUEOUS  SULFIDE  SOLUTIONS
                                               OF LOW IONIC STRENGTH^
oo
Temperature,
pH
6.0
6.1
6.2
6.3
6.4
6.5
6.6
6.7
6.8
6.9
7.0
7.1
7.2
7.3
7.4
7.5
5
.9387
.9241
.9063
.8848
.8591
.8289
.7938
.7535
.7083
.6586
.6051
.5489
.4915
.4344
.3789
.3264
6
.9369
.9219
.9036
.8816
.8554
.8245
.7886
.7477
.7019
.6516
.5977
.5413
.4838
.4268
.3716
.3196
7
.9351
.9196
.9009
.8783
.8515
.8200
.7835
.7419
.6954
.6445
.5902
.5336
.4761
.4192
.3644
.3129
8
.9332
.9173
.8981
.8750
.8476
.8154
.7782
.7359
.6888
.6374
.5827
.5259
.4684
.4117
.3573
.3063
9
.9312
.9149
.8952
.8716
.8435
.8107
.7728
.7298
.6821
.6303
.5752
.5182
.4607
.4043
.3502
.2998
10
.9292
.9125
.8923
.8681
.8394
.8059
.7673
.7237
.6754
.6230
.5676
.5105
.4531
.3969
.3432
.2934
11
.9272
.9100
.8893
.8645
.8352
.8010
.7618
.7175
.6686
.6158
.5600
.5028
.4454
.3895
.3363
.2870
C
12
.9250
.9074
.8862
.8608
.8309
.7960
.7561
.7112
.6617
.6084
.5524
.4950
.4378
.3822
.3295
.2807

13
.9229
.9048
.8830
.8571
.8265
.7910
.7504
.7048
.6548
.6010
.5448
.4873
.4302
.3749
.3227
.2745

14
.9206
.9021
.8798
.8533
.8220
.7858
.7445
.6983
.6477
.5936
.5371
.4796
.4226
.3677
.3160
.2684

15
.9184
.8994
.8765
.8494
.8175
.7806
.7386
.6918
.6407
.5861
.5294
.4719
.4151
.3605
.3093
.2624

16
.9160
.8965
.8731
.8454
.8128
.7752
.7326
.6852
.6335
.5786
.5217
.4642
.4076
.3534
.3028
.2565

17
.9136
.8936
.8697
.8413
.8081
.7698
.7265
.6785
.6263
.5711
.5140
.4565
.4002
.3464
.2963
.2506

-------
Table 9  (continued).   FRACTION OF DISSOLVED SULFIDE AS MOLECULAR H2S IN AQUEOUS SULFIDE
                             SOLUTIONS OF LOW IONIC STRENGTR-
Temperature ,
PH
7.6
7.7
7.8
7.9
8.0
8.1
8.2
8.3
8.4
8.5
8.6
8.7
8.8
8.9
9.0
5
.2779
.2341
.1954
.1617
.1329
.1085
.08815
.07131
.05749
.04621
.03706
.02966
.02371
.01892
.01509
6
.2717
.2286
.1906
.1576
.1293
.1055
.08570
.06929
.05584
.04487
.03597
.02879
.02300
.01836
.01464
7
.2657
.2232
.1859
.1535
.1259
.1027
.08331
.06733
.05423
.04356
.03492
.02794
.02232
.01781
.01420
8
.2597
.2179
.1812
.1495
.1225
.09985
.08098
.06541
.05267
.04229
.03389
.02711
.02165
.01728
.01377
9
.2538
.2127
.1767
.1456
.1193
.09711
.07871
.06355
.05115
.04106
.03289
.02631
.02101
.01676
.01336
10
.2480
.2076
.1722
.1418
.1160
.09443
.07650
.06174
.04967
.03986
.03192
.02553
.02038
.01626
.01296
11
.2423
.2025
.1679
.1381
.1129
.09183
.07434
.05997
.04823
.03869
.03098
.02477
.01978
.01577
.01257
C
12
.2366
.1976
.1636
.1345
.1099
.08928
.07224
.05825
.04683
.03756
.03007
.02403
.01918
.01530
.01219

13
.2311
.1927
.1594
.1309
.1069
.08680
.07020
.05658
.04547
.03646
.02918
.02332
.01861
.01484
.01182

14
.2257
.1880
.1553
.1274
.1040
.08438
.06821
.05495
.04415
.03539
.02832
.02263
.01806
.01440
.01147

15
.2203
.1833
.1513
.1241
.1011
.08203
.06627
.05337
.04286
.03435
.02748
.02195
.01752
.01396
.01112

16
.2151
.1787
.1474
.1207
.09834
.07973
.06439
.05183
.04161
.03334
.02667
.02130
.01699
.01354
.01079

17
.2099
.1742
.1435
.1175
.09564
.07749
.06255
.05033
.04040
.03236
.02588
.02066
.01648
.01314
.01046

-------
Table 9  (continued).  FRACTION OF DISSOLVED SULFIDE AS MOLECULAR H2S IN AQUEOUS SULFIDE
                             SOLUTIONS OF LOW IONIC STRENGTR-

Temperature
PH
6.0
6.1
6.2
6.3
6.4
6.5
6.6
6.7
6.8
6.9
7.0
7.1
7.2
7.3
7.4
7.5
18
.9111
.8906
.8661
.8371
.8032
.7643
.7203
.6717
.6191
.5635
.5063
.4489
.3928
.3394
.2899
.2448
19
.9086
.8876
.8625
.8328
.7983
.7587
.7141
.6648
.6117
.5559
.4985
.4412
.3855
.3326
.2836
.2392
20
.9060
.8845
.8588
.8285
.7933
.7530
.7077
.6579
.6044
.5482
.4908
.4336
.3782
.3257
.2773
.2336
21
.9033
.8813
.8550
.8241
.7882
.7472
.7013
.6509
.5970
.5406
.4831
.4261
.3709
.3190
.2712
.2281
22
.9006
.8780
.8511
.8195
.7830
.7413
.6948
.6439
.5895
.5329
.4754
.4185
.3638
.3123
.2651
.2227
23
.8978
.8747
.8472
.8149
.7777
.7353
.6882
.6368
.5820
.5252
.4677
.4110
.3566
.3057
.2591
.2174
24
.8949
.8712
.8431
.8102
.7723
.7293
.6815
.6296
.5745
.5175
.4600
.4036
.3496
.2992
.2532
.2122
, c
25
.8920
.8677
.8390
.8054
.7668
.7231
.6748
.6223
.5669
.5098
.4523
.3962
.3426
.2928
.2474
.2071

26
.8890
.8641
.8348
.8005
.7612
.7169
.6679
.6151
.5593
.5020
.4447
.3888
.3357
.2864
.2417
.2021

27
.8859
.8605
.8305
.7956
.7556
.7106
.6611
.6077
.5517
.4943
.4371
.3815
.3288
.2801
.2361
.1971

28
.8827
.8567
.8261
.7905
.7498
.7042
.6541
.6003
.5440
.4866
.4295
.3742
.3220
.2739
.2306
.1923

29
.8795
.8529
.8216
.7853
.7440
.6977
.6471
.5929
.5364
.4789
.4219
.3670
.3153
.2678
.2252
.1875

30
.8762
.8490
.8170
.7801
.7380
.6912
.6400
.5854
.5287
.4712
.4144
.3599
.3087
.2618
.2198
.1829

-------
   Table  9   (continued).  FRACTION OF DISSOLVED SULFIDE AS MOLECULAR H_S IN AQUEOUS SULFIDE
                                SOLUTIONS OF LOW IONIC STRENGTH-/
Temperature ,
pH
7.6
7.7
7.8
7.9
8.0
8.1
8.2
8.3
8.4
8.5
8.6
8.7
8.8
8.9
9.0
18
.2048
.1698
.1398
.1143
.09300
.07531
.06076
.04888
.03922
.03141
.02511
.02005
.01599
.01274
.01015
19
.1998
.1655
.1361
.1112
.09042
.07319
.05902
.04746
.03807
.03048
.02436
.01945
.01551
.01236
.00984
20
.1949
.1613
.1325
.1082
.08792
.07112
.05733
.04608
.03696
.02958
.02364
.01887
.01505
.01199
.00955
21
.1901
.1572
.1290
.1053
.08547
.06911
.05568
.04474
.03587
.02871
.02294
.01831
.01460
.01163
.00926
22
.1854
.1531
.1256
.1024
.08309
.06714
.05408
.04344
.03482
.02786
.02225
.01776
.01416
.01128
.00898
23
.1808
.1492
.1222
.09959
.08076
.06523
.05252
.04218
.03379
.02703
.02159
.01723
.01373
.01094
.00871
24
.1763
.1453
.1189
.09685
.07850
.06338
.05101
.04095
.03280
.02623
.02095
.01671
.01332
.01061
.00845
C
25
.1718
.1415
.1157
.09418
.07629
.06157
.04953
.03975
.03183
.02545
.02032
.01621
.01292
.01029
.00819

26
.1675
.1378
.1126
.09158
.07414
.05981
.04810
.03859
.03090
.02470
.01972
.01573
.01253
.00998
.00794

27
.1632
.1341
.1096
.08904
.07205
.05809
.04670
.03746
.02998
.02396
.01913
.01526
.01216
.00968
.00770

28
.1590
.1306
.1066
.08657
.07001
.05642
.04535
.03636
.02910
.02325
.01856
.01480
.01179
.00939
.00747

29
.1549
.1271
.1037
.08416
.06803
.05480
.04403
.03529
.02824
.02256
.01800
.01435
.01144
.00911
.00725

30
.1509
.1237
.1009
.08181
.06609
.05322
.04274
.03425
.02740
.02189
.01747
.01392
.01109
.00883
.00703
— Sulfide solutions with ionic strength y less than 0.01.

-------
 Table 10.  MULTIPLICATION FACTORS FOR CONVERTING H2S CALCULATED FROM
   POMEROY'S FACTORS FOR A "TYPICAL WATER SUPPLY" TO CORRESPONDING
                  CONCENTRATIONS BASED ON THIS STUDY
      	Temperature, C	
pH     10    11    12    13    14    15    16    17    18    19    20
6.0   1.03  1.03  1.04  1.04  1.04  1.04  1.05  1.05  1.05  1.05  1.06
6.1   1.05  1.05  1.05  1.05  1.06  1.06  1.06  1.06  1.07  1.07  1.07
6.2   1.06  1.06  1.06  1.06  1.07  1.07  1.07  1.07  1.07  1.07  1.07
6.3   1.07  1.07  1.07  1.07  1.08  1.08  1.08  1.08  1.09  1.09  1.10
6.4   1.08  1.08  1.09  1.09  1.09  1.09  1.10  1.10  1.10  1.10  1.11
6.5   1.09  1.09  1.10  1.10  1.11  1.11  1.12  1.12  1.12  1.12  1.13
6.6   1.11  1.11  1.12  1.12  1.13  1.13  1.14  1.15  1.16  1.16  1.17
6.7   1.14  1.14  1.15  1.16  1.16  1.16  1.17  1.17  1.18  1.18  1.19
6.8   1.16  1.16  1.17  1.17  1.18  1.18  1.20  1.20  1.21  1.21  1.22
6.9   1.19  1.19  1.20  1.20  1.22  1.22  1.23  1.23  1.25  1.25  1.26
7.0   1.22  1.22  1.24  1.24  1.25  1.25  1.26  1.25  1.27  1.26  1.28
7.1   1.24  1.24  1.25  1.25  1.27  1.27  1.29  1.29  1.31  1.31  1.33
7.2   1.28  1.28  1.30  1.30  1.32  1.32  1.34  1.33  1.35  1.35  1.37
7.3   1.32  1.32  1.34  1.34  1.35  1.35  1.36  1.35  1.37  1.36  1.38
7.4   1.34  1.33  1.35  1.34  1.36  1.36  1.38  1.37  1.39  1.39  1.41
7.5   1.36  1.35  1.38  1.37  1.38  1.37  1.39  1.38  1.39  1.38  1.40
7.6   1.36  1.35  1.37  1.36  1.38  1.37  1.39  1.38  1.40  1.40  1.42
7.7   1.37  1.36  1.38  1.38  1.40  1.40  1.43  1.43  1.46  1.46  1.49
7.8   1.41  1.41  1.45  1.45  1.47  1.45  1.47  1.46  1.47  1.47  1.49
7.9   1.44  1.43  1.45  1.44  1.46  1.45  1.47  1.46  1.49  1.49  1.51
8.0   1.45  1.44  1.47  1.46  1.48  1.47  1.49  1.48  1.50  1.50  1.52
8.1   1.46  1.45  1.48  1.47  1.49  1.47  1.49  1.48  1.50  1.49  1.51
8.2   1.46  1.45  1.47  1.46  1.47  1.46  1.47  1.46  1.47  1.46  1.49
8.3   1.44  1.43  1.44  1.43  1.45  1.44  1.47  1.46  1.49  1.49  1.52
8.4   1.44  1.44  1.47  1.47  1.48  1.46  1.48  1.46  1.47  1.46  1.48
                                  72

-------
   Table 10  (continued).  MULTIPLICATION FACTORS FOR CONVERTING H?S
    CALCULATED FROM POMEROY'S FACTORS FOR A "TYPICAL WATER SUPPLY"
         TO  CORRESPONDING CONCENTRATIONS BASED ON THIS STUDY
        	Temperature. C	
2l	21    22    23    24    25    26    27    28    29    30
6.0     1.06  1.06  1.06  1.07  1.07  1.07  1.07  1.07  1.08  1.08
6.1     1.07  1.07  1.07  1.08  1.08  1.08  1.08  1.08  1.09  1.09
6.2     1.08  1.09  1.09  1.09  1.09  1.09  1.10  1.10  1.10  1.10
6.3     1.10  1.10  1.10  1.11  1.11  1.11  1.11  1.11  1.12  1.12
6.4     1.11  1.12  1.12  1.13  1.13  1.13  1.14  1.14  1.15  1.15
6.5     1.14  1.15  1.15  1.16  1.16  1.16  1.17  1.17  1.18  1.18
6.6     1.17  1.18  1.18  1.19  1.18  1.18  1.19  1.19  1.21  1.21
6.7     1.19  1.20  1.20  1.22  1.22  1.22  1.23  1.23  1.25  1.25
6.8     1.22  1.24  1.24  1.25  1.25  1.25  1.27  1.27  1.28  1.27
6.9     1.26  1.27  1.27  1.28  1.27  1.27  1.29  1.29  1.31  1.31
7.0     1.28  1.30  1.30  1.32  1.32  1.32  1.34  1.34  1.36  1.36
7.1     1.33  1.35  1.35  1.37  1.37  1.36  1.38  1.38  1.39  1.38
7.2     1.36  1.38  1.37  1.39  1.38  1.38  1.39  1.39  1.41  1.40
7.3     1.37  1.39  1.39  1.41  1.41  1.40  1.42  1.41  1.42  1.42
7.4     1.40  1.41  1.40  1.41  1.41  1.40  1.41  1.41  1.42  1.42
7.5     1.39  1.41  1.40  1.42  1.42  1.41  1.44  1.43  1.46  1.46
7.6     1.42  1.45  1.45  1.48  1.48  1.48  1.51  1.50  1.52  1.51
7.7     1.48  1.50  1.49  1.50  1.49  1.48  1.50  1.49  1.52  1.51
7.8     1.48  1.50  1.49  1.52  1.51  1.51  1.53  1.52  1.54  1.53
7.9     1.49  1.52  1.51  1.53  1.52  1.52  1.54  1.52  1.54  1.53
8.0     1.50  1.52  1.51  1.53  1.52  1.51  1.53  1.51  1.53  1.51
8.1     1.49  1.51  1.49  1.51  1.49  1.48  1.50  1.49  1.52  1.51
8.2     1.47  1.50  1.49  1.52  1.51  1.51  1.54  1.52  1.53  1.52
8.3     1.50  1.51  1.50  1.51  1.49  1.48  1.50  1.49  1.51  1.50
8.4     1.47  1.49  1.48  1.52  1.51  1.50  1.53  1.52  1.53  1.52
                                   73

-------
    Table  10  (continued).  MULTIPLICATION FACTORS FOR CONVERTING H S
     CALCULATED FROM POMEROY'S FACTORS FOR A  "TYPICAL WATER SUPPLY"
          TO  CORRESPONDING CONCENTRATIONS BASED ON THIS STUDY

Temperature,
PH
8.5
8.6
8.7
8.8
8.9
9.0
10
1.44
1.44
1.45
1.50
1.51
1.49
11
1.42
1.43
1.44
1.50
1.49
1.48
12
1.44
1.46
1.47
1.55
1.50
1.50
13
1.43
1.46
1.46
1.55
1.49
1.50
14
1.45
1.47
1.49
1.56
1.51
1.51
15
1.44
1.46
1.48
1.54
1.50
1.49
C
16
1.46
1.48
1.52
1.54
1.52
1.50

17
1.46
1.47
1.52
1.53
1.51
1.49

18
1.49
1.49
1.57
1.54
1.53
1.51

19
1.48
1.49
1.57
1.52
1.53
1.49

20
1.51
1.52
1.60
1.54
1.54
1.52
determination.  Therefore, it appears that the sulfide demand exhibited
by some of the other test solutions is due to biological and other chemi-
cal processes and is not due appreciably to oxidation from dissolved
oxygen.

The procedure used in preparing the sample for dissolved sulfide deter-
mination may have an effect on the calculated H.S concentration.  The
flocculation and centrifugation techniques for isolation of dissolved
sulfide were not entirely satisfactory since in certain instances it
appeared that not all of the suspended sulfides were removed and that
during processing some decline in sulfide levels may also occur.  For the
types of samples prepared the millipore filtration technique is the best
means of isolating and stabilizing dissolved sulfide.  When this latter
technique is used in conjunction with the factors defining the fraction of
dissolved sulfide as H2S determined in this study, the calculated H_S
concentration is very close to the H-S concentration determined by the
direct method for all types of samples tested.
                                   74

-------
   Table 10 (continued).   MULTIPLICATION FACTORS  FOR CONVERTING H2
    CALCULATED FROM POMEROY'S FACTORS FOR A "TYPICAL WATER SUPPLY"
         TO CORRESPONDING CONCENTRATIONS BASED ON THIS  STUDY
Temperature, C
PH
8.5
8.6
8.7
8.8
8.9
9.0
21
1.50
1.51
1.58
1.53
1.53
1.50
22
1.51
1.55
1.59
1.55
1.54
1.53
23
1.50
1.54
1.57
1.54
1.52
1.52
24
1.52
1.59
1.58
1.56
1.54
1.55
25
1.52
1.59
1.56
1.55
1.53
1.55
26
1.51
1.59
1.54
1.55
1.51
1.55
27
1.54
1.62
1.57
1.57
1.54
1.57
28
1.53
1.60
1.55
1.55
1.52
1.55
29
1.57
1.61
1.57
1.56
1.55
1.56
30
1.
1.
1.
1.
1.
1.
56
59
56
54
54
54
RELATIONSHIP BETWEEN TEST pH AND SULFIDE TOXICITY TO THE FATHEAD MINNOW
A series of acute tests with fathead minnows to determine the relation-
ship between test pH and the toxicity of sulfides was conducted at pH
levels from about 6.5 to 8.7 and 20 C.  A summary of the data and analysis
for 16 bioassays of 96-hr duration is presented in Tables 14 through 17.
Inspection of the data in Table 17 for composite tests grouped according
to pH reveals that, with the exception of tests done at a pH of about
6.5, the ambient molecular H2S concentration required to give a 96-hr
median tolerance limit (LC50) response decreased as the pH of the test
solution increased.  The 96-hr LCSO's in yg/liter H2S ranged from 57.3 at
pH 7.101 to 14.9 at pH 8.693.  The anomalous results for the experiments
performed at a pH of about 6.5 may be due to the interaction between sul-
fide and the relatively high C0_ concentration in these test solutions.
It is also quite feasible that the 96-hr LC50 values for H-S are rela-
tively constant in  test solutions over a pH range of about 6.5 to 7.1
and with the accompanying dissolved  sulfide and C02 concentrations.
Within the pH range of about 7.1 to  8.7, a 0.1 unit increase in pH was
                                     75

-------
 Table 11.  MULTIPLICATION FACTORS AT  25 C FOR CONVERTING H2S CALCULATED
    FROM FACTORS  IN THE 1946 TO  1965 EDITIONS OF  STANDARD METHODS
         TO CORRESPONDING CONCENTRATIONS BASED ON THIS  STUDY

pH
6.0
6.2
6.4
6.5
6.6
6.7
6.8
6.9
7.0
7.1
7.2
Factor
1.07
1.10
1.14
1.19
1.20
1.24
1.29
1.31
1.37
1.37 (1.41)-/
1.43
PH
7.3
7.4
7.5
7.6
7.7
7.8
7.9
8.0
8.2
8.4
8.8
Factor
1.27
1.46
1.48
1.56
1.55
1.59
1.60
1.59
1.60
1.59
1.64

(1.46)^










a/
— Factors in parenthesis apply to values originally published in Table
IV of Pomeroy's 1941 work?
calculated by linear regression to result in a 2.7 yg/liter decrease in
the molecular H2S 96-hr LC50 value.  The linear regression correlation
coefficient (r) was equal to -0.9945.  From a practical standpoint the
importance of this change in apparent toxicity of H_S is reduced since
as pH is increased, the concentration of dissolved sulfide required to
produce an acutely toxic solution is logarithmically increased.  This
change in concentration of dissolved sulfide, HS~ ion, and molecular H.S
concentration required to give the observed 96-hr LC50 toxic response at
the pH of the test solution is shown in Figure 3.  It should be empha-
sized that under relatively alkaline pH conditions the major portion of
the dissolved sulfide consists of the HS  ion while only a small propor-
tion exists as molecular H?S.
                                   76

-------
When test pH within the range 6.5 to 8.7 for the composite fathead
minnow sulfide bioassays at 20 C was compared with the log dissolved
sulfide concentration at the 96-hr LC50 level, a positive linear
relationship with a regression correlation coefficient of 0.9991 was
calculated.  The regression applying to dissolved sulfide levels ranging
from 64-780 yg/liter as H~S is described by the equation:

                       log Y = -1.278 + 0.477 X                     (13)
where  Y = 96-hr LC50 of dissolved sulfide as ug/liter
       X = pH of test solution.
                                     77

-------
         Table 12.  DETERMINATION OF MOLECULAR H2S IN DIFFERENT WATERS-7 BY  CALCULATION FROM THE  TOTAL AND
                          DISSOLVED SULFIDE CONCENTRATION AND BY A DIRECT TECHNIQUE
                                 (concentrations expressed as jug/liter H0S)
-q
00
Item
Temperature, C
PH
Fraction dissolved
sulfide as H2S
Sulfide added
Total sulfide^7
Direct
Indirect
Dissolved sulfide after:
Flocculation
Centrifugation
Filtration
H_S calculated from:
Total sulfide - direct
- indirect
Deionized—
20.0
7.654
0.176

221.4

227.8
220.5

195.2
212.7
211.5

40.2
38.8
Well^7
19.8
7.615
0.204

184.7

190.9
168.1

177.6
178.2
178.2

38.9
34.2
Well^7
19.8
7.820
0.128

181.2

175.8
174.7

171.6
162.5
166.1

22.6
22.4
Fish
. c/
aquarium—
20.0
7.680
0.168

174.5

134.7
114.9

150.4
123.8
122.0

22.7
19.3
50%
pond
19.8
7.564
0.209

338.6

134.7
139.7

129.8
135.9
126.2

28.2
29.2
Pond
19.8
7.700
0.162

519.3

157.6
154.5

127.4
137.1
135.9

25.6
25.0
Mississippi
River^7
20.1
8.040
0.0806

317.3

91.9
93.7

82.0
72.1
88.6

7.40
7.55

-------
          Table 12 (continued).   DETERMINATION OF MOLECULAR H-S IN DIFFERENT WATERS-  BY CALCULATION FROM
                      THE TOTAL  AND DISSOLVED SULFIDE CONCENTRATION AND BY A DIRECT TECHNIQUE

                                     (concentrations expressed as /ag/liter H^S)
-q
CO

Item
H_S calculated from
dissolved sulfide after:
Flocculation
Centrifugation
Filtration
H~S determined directly
Fish 50% Mississippi
Deionized^ Well- Well— aquarium^ pond Pond River-

34.4 36.2 22.0 25.3 27.2 20.6 6.60
37.5 36.3 20.8 20.8 28.4 22.2 5.80
37.3 36.3 21.3 20.5 26.4 22.0 7.14
37.6 35.7 21.1 22.3 27.2 21.0 7.17
        a/
        —Results represent one sample per test water.  Test solutions prepared with less than 100% sample
        were diluted with deoxygenated well water.
        —^Deoxygenated.
        -r^Oxygenated.
        —ySample was taken below the Interstate 494 bridge, St. Paul, Minnesota.
        — Direct colorimetric sulfide determination on sample or indirect  determination following dis-
        placement and collection of sulfide on concentration column.

-------
        Table 13.   DETERMINATION OF MOLECULAR H2S IN DIFFERENT EFFLUENTS-7 BY CALCULATION FROM THE TOTAL AND
                             DISSOLVED SULFIDE CONCENTRATION AND BY A DIRECT TECHNIQUE
                                (concentrations expressed asyug/liter H_S)
oo
o
Item
Temperature, C
PH
Fraction dissolved
sulfide as H0S
2.
Sulfide added
Total sulfide^
Direct
Indirect
Dissolved sulfide after:
Flocculation
Centrifugation
Filtration
H_S calculated from:
Total sulfide - direct
- indirect
50%
sewage—
19.9
7.732
0.152


207.3

271.4
277.7

219.4
259.9
225.4

41.4
42.2
Sewage—
20.0
7.478
0.243


201.4

325.2
303.5

285.9
304.0
243.6

78.9
73.7
10% Kraft paper
processing
19.7
7.974
0.0936


1056

327.0
336.5

321.0
282.3
314.9

30.6
31.5
10% hard board
wood processing
20.0
7.738
0.150


990.8

48.9
-

13.6
24.6
57.7

7.33
-
Lagooned oil
refinery 	
20.1
7.856
0.118


538.3

189.1
190.0

171.0
158.9
177.0

22.3
22.4

-------
           Table 13 (continued).   DETERMINATION OF MOLECULAR H2S  IN DIFFERENT  EFFLUENTS^  BY  CALCULATION

                    FROM THE TOTAL AND  DISSOLVED SULFIDE  CONCENTRATION AND  BY  A DIRECT TECHNIQUE

                                     (concentrations  expressed  asyug/liter  H-S)
OD


H2




H2
Item
S calculated from
dissolved sulfide after:
Flocculation
Centrifugation
Filtration
S determined directly
50%
b/
sewage—


33.
39.
34.
34.


4
5
3
6
Sewage—


69.
73.
59.
61.


4
8
1
1
10% Kraft paper
processing


30.
26.
29.
25.


0
4
5
3
10% hard board
wood processing


2
3
8
6


.04
.69
.64
.81
Lagooned oil
refinery


20.2
18.8
20.9
19.0






       —Results represent  one  sample  per  test  effluent.   Test  solutions  prepared with  less  than  100% sample
       were diluted with deoxygenated  well water.
       — Sewage samples from Minneapolis-St.  Paul  Metropolitan  Sewage  Treatment  Plant were taken  following
       secondary treatment.
       — Direct colorimetric sulfide determination on  sample  or indirect  determination  following  displace-
       ment and collection  of sulfide  on concentration column.

-------
                 Table 14.   SUMMARY OF TEST CONDITIONS IN SULFIDE BIOASSAYS AT DIFFERENT pH VALUES
oo
to
PH
Test
4C
5C
6C
Mean
2B
3B
6B
Mean
1A
4A
5A
Mean
IB
4B
Mean
Mean
6.471
6.474
6.442
6.462
7.101
7.105
7.097
7.101
7.701
7.673
7.720
7.698
8.148
8.154
8.151
SD
.088
.030
.092
.070
.009
.020
.060
.030
.052
.018
.038
.036
.047
.012
.030
Temp
Mean
20.0
20.0
19.9
20.0
20.1
20.0
20.0
20.0
19.9
20.0
20.2
20.0
20.1
20.0
20.0
., c
SD
.02
.01
.13
.05
.05
.06
.00
.04
.39
.01
.19
.20
.17
.05
.11

DO,
Mean
7
7
7
7
7
7
7
7
7
7
.68
.61
.56
.62
.58
.68
.73
.66
.50
.78
7.68
7.65
7.26
7.72
7.49
mg/1
SD
.07
.07
.11
.08
.10
.09
.09
.09
.12
.06
.03
.07
.39
.08
.24
Alkalinity,,
Total Phenol.
65
63
60
62
124
126
133
127
198,
198,
197.
.0
.1
.2
.8
.0
.0
.8
.9
.1
.8
.6
198.2
229 . 2
229.
229.
0
1
Free C02 in
mg/1 CaCO test solution^
Bicarbonate me/1 mm HO
65
.0
63.1
60,
62,
124,
126.
133.
127.
198.
198.
197.
198.
229.
229.
229.
.2
•8 44.0 19.5
.0
,0
,8
9 20.5 9.1
1
8
6
2 7.9 3.5
2
0
1 3.2 1.4

-------
          Table  14  (continued).   SUMMARY OF TEST CONDITIONS IN SULFIDE BIOASSAYS AT DIFFERENT pH VALUES
oo
CO
PH
Test
2A
3A
Mean
2C
3C
6A
Mean
^Free
20 C,
Mean
8.445
8.416
8.430
8.693
8.707
8.679
8.693
SD
.025
.011
.018
.010
.011
.020
.014
CO evaluated
then 1 mg/liter
Temp . , C
Mean
20.1
19.9
20.0
20.1
20.1
20.1
20.1
SD
.05
.05
.05
.21
.03
.07
.10
DO, mg/1
Mean SD
7.53
7.57
7.55
7.39
7.47
7.60
7.49
by nomographic method
CO- = 0.444 mm Hg CO-
.13
.09
.11
.12
.14
.06
.11
(APHA1) .
pressure
Free
Alkalinity, mg/1 CaC00 test
W _^. . 	
Total Phenol. Bicarbonate mg/1
238.2
238.2
238.2
242.9
243.4
241.6
242.6
Assuming K
(Stumm and
4.0
3.5
3.8
9.3
9.5
8.3
9.0
= H2C03/
Morgan,
234
234
234
233
233
233
233
'PCO;
P
.2
.7
.4 1.7
.6
.9
.3
.6 1.0
2 and -log K =
. 1485/).
CO in
a/
solution—
mm Hg


0.75



0.44
1.41 at

-------
Table 15.  DESCRIPTION OF FATHEAD MINNOWS USED IN SULFIDE BIOASSAYS AT
                         DIFFERENT pH VALUES

Test
4C
5C
6C
Mean
2B
3B
6B
Mean
1A
4A
5A
Mean
IB
4B
Mean
2A
3A
Mean
2C
3C
6A
Mean

All fish
28.3
29.3
33.8
30.5
28.9
28.5
32.8
30.1
31.6
27.7
29.5
29.6
31.6
29.6
30.6
29.7
29.0
29.4
28.5
28.3
34.1
30.3
Mean length, mm
Survivors
28.7
28.7
33.2
30.2
27.7
28.8
32.2
29.6
31.8
27.6
29.5
29.6
31.6
29.3
30.4
29.6
29.0
29.3
28.4
28.7
33.8
30.3

Mortalities
28.0
30.3
34.5
30.9
30.2
27.6
33.6
30.5
30.8
28.0
29.5
29.4
31.5
30.1
30.8
29.9
29.0
29.4
28.8
27.2
36.8
30.9
Mean weight
survivors, g
0.277
0.252
0.418
0.316
0.240
0.259
0.351
0.283
0.326
0.227
0.263
0.272
0.314
0.281
0.298
0.268
0.276
0.272
0.243
0.271
0.422
0.312
                                  84

-------
        Table 16.   BIOLOGICAL ASSAY BY THE BMD03S PROSIT ANALYSIS METHOD^ OF 96-HOUR FATHEAD MINNOW SULFIDE

                                       BIOASSAYS GROUPED ACCORDING TO TEST pH
oo
01
Parameters in
Average of
b/
grouped tests—
pH Temp., C
6.462 20.0
7.101 20.0
7.698 20.0
8.151 20.0
8.430 20.0
8.693 20.1
probit
Y =
A
-9.847
-15.094
-19.891
-32.082
-31.486
-12.139
c/
equation—
A + BX
B
8.775
11.431
15.716
26.254
28.826
14.598

Chi-square
statistic
9.14
4.85
25.57*'
2.33
9.53
26.67*'
Degrees
of
freedom
9
9
10
6
6
10
Log 96-hr
LC50,
Ug/1 IUS
1.6920
1.7579
1.5839
1.4125
1.2657
1.1741
Standard
error of
log LC50
0.1140
0.0875
0.0636
0.0381
0.0347
0.0685
7- /See Dixon.
       — ,See table 14 for individual test conditions.
       — A probit of 4.0, 5.0, and 6.0 corresponds  to 16, 50, and  84% mortality,  respectively, when Y is  the
       maximum likelihood probit value and X is log H~S concentration in ug/liter.
       — The (chi)  of the probit curve exceeds the value of  (chi)  for P = 0.05,  thus  the data are signifi-
       cantly heterogeneous.

-------
        Table  17.   SUMMARY  OF  LETHAL CONCENTRATION (LC)  ANALYSIS FOR 96-HOUR FATHEAD MINNOW SULFIDE BIOASSAYS

                                            GROUPED ACCORDING TO TEST pH

                                             (expressed  as yg/liter H_S)
CD

Average of
grouped tests
PH
6.462
7.101
7.698
8.151
8.430
8.693
Temp . , C
20.0
20.0
20.0
20.0
20.0
20.1
a/
Fraction-
dissolved H_
sulfide LC
as H0S
0.769
0.433
0.162
0.0637
0.0346
0.0191
16%
37.8
46.8
33.1
23.7
17.0
12.8
S 96-hr
values
50%
49.2
57.3
38.4
25.8
18.4
14.9
b/ 95%
Slope— confidence
function, limits for
84%
64.0
70.0
44.4
28.2
20.0
17.5
S
1.30
1.22
1.16
1.09
1.08
1.17
H0S LC50
45.
52.
34.
24.
17.
13.
1-53.7
5-62.6
4-42.9^
4-27.2
5-19.3
5-16.4^
Dissolved
sulfide
at LC50
level
64.0
132.3
237.0
405.0
531.8
780.1
HS at
LC50
level
14.
75.
198.
379.
513.
765.
8
0
6
2
4
2
        a/
        7-,Obtained from Table 9  for average pH and temperature.
        ^yS  = (LC84£LC50 + LC50/LC16)/2.                    2
        — The (chi)   of probit curve exceeds value of (chi)   for p = 0.05, thus the data are significantly
        heterogeneous and a special means of calculating the confidence limits was employed (Litchfield and
        Wilcoxon5b).

-------
            800
       CO
        CM
       UJ
            600
            400
oo
       CO
       id
       £    200
       CO
                DISSOLVED   SULFIDE
                HS
                                H2S
      Figure  3.
                                  TEST   PH

Relationship between test pH and dissolved sulfide, HS,  and molecular
at levels corresponding to the  96-hr LC50 for fathead minnows at 20 C.
                                                                                  concentration

-------
                              SECTION VI

                              DISCUSSION
DETERMINATION OF MOLECULAR H0S AND K, IONIZATION CONSTANTS OF H_S,  N
                            21                          2  (aq)
The close agreement between the expression for K  of (0.31 + 0.029 T  (C))
  _7                                            L
10   derived from the literature (equation 4) and that of (0.45 + 0.030
T (C))-10~  defined during this study gives support to the validity of
the methods and the accuracy of the equations proposed in this report.
It can be demonstrated that what might be considered a slight change  in
KI can have a dramatic effect on the calculated percentage of dissolved
sulfide as molecular H-S.  Therefore, it is critical that a correct
expression for the relationship between K- and temperature be employed
when calculating molecular H_S concentrations from dissolved sulfide.
Equation (12) ,

                     p   = 7.252 - 0.01342 T (C) ,                    (12)
derived from data obtained in this study is believed to be an accurate
expression for most freshwaters.  The data employed in its calculation
were obtained over the temperature range of 10 to 25 C.  It is felt  that
the expression can be extended for use with temperatures ranging from at
least 5 to 30 C with an acceptable loss in accuracy of calculated pK...
values in the extrapolated temperature regions.  Use of the expression
at temperatures very far removed from those used in defining the equation
is not recommended since the relationship between pK.. and temperature
                                                    •*•           g
may not be linear over extremes in temperature (Wright and Maass ) .
                                    88

-------
The KI values determined in this study are "apparent" values and are not
"true" ionization constants extrapolated to zero ionic strength.  Since
the values represent the average relationship determined from sulfide
solutions prepared with deionized water and with a well water of rela-
tively high alkalinity and total hardness, it is felt that the combined
expression is applicable to most freshwaters of normal ionic strength
(y = 0.001 to 0.010).  If no correction in the pK. temperature expression
is made for ionic strength, it can be demonstrated that less than a 5%
error in the calculated fraction of dissolved sulfide as molecular H»S
would be realized at most combinations of temperature, pH, and ionic
strength encountered in normal freshwaters.  However, it is proposed
that if the ionic strength of the solution is greater than about 0.01
but less than 0.10, the fraction of dissolved sulfide as molecular H^S
should be calculated from the expression derived in the Appendix.

The accuracy of calculating molecular H_S concentrations from determined
dissolved sulfide with factors derived in this study  (Table 9), which
correspond to the fraction of dissolved sulfide as molecular H«S, was
confirmed for many different types of freshwaters and industrial ef-
fluents.  A technique utilized during this study  also allows for the
direct determination of molecular H9S at  levels as low  as  a few yg/liter.
However, this method may be of limited value since in many static  test
solutions sulfide could undergo  oxidation during  the  tUS  displacement
and collection phase.  The procedure  could conceivably  be used  to  monitor
H»S in various waters when the gas stripping procedure  is combined with
a  continuous-flow liquid phase sampler  and a suitable sulfide detector
                                                                      58
system.  A similar  approach has  been  used by Garber,  Nagano,  and Wada
to measure H^S in sewer atmosphere and  liquid.

MODES OF  TOXIC ACTION OF DISSOLVED SULFIDE TO FISH
It is generally  recognized  that  the  gill  is the primary site  designed
for gas  exchange between blood  and water  and that undissociated molecules
will  penetrate living  tissues more readily than charged ions.   The
                                     89

-------
pronounced toxicity of sulfide solutions in studies with fish has been
assumed by numerous researchers, even though there is no conclusive
direct experimental evidence, to be attributable  to the action of undis-
sociated molecular H_S, varying with the pH and its concentration in
                             -     2-
solution, and not with the HS  or S   ions.  Recent recommendations of
safe levels of sulfide have even been expressed in terms of concentra-
tions of molecular H_S rather than dissolved sulfide.

                 59
Lloyd and Herbert   discuss the possible effect which CO  excreted via
the gills of fish may have in shifting the CO^-bicarbonate-carbonate
chemical equilibrium and lowering of the pH of water in contact with
the gills.  These authors also proposed that the  toxicity of ammonium
salts is not strictly dependent on the pH value of the bulk solution but
on the pH of the water at the gill surface.  If Lloyd and Herbert's
explanation is correct, the toxicity of sulfide solutions may be in-
creased by respiratory depression of the pH in gills of fish due to
excretion of CO- since the concentration of molecular H~S due to a
conversion of HS  ions may be higher in the solution in contact with the
gills than that surrounding the fish.  According  to their proposal, the
pH at the gill surface can be calculated from the bicarbonate alkalinity,
temperature,  and free C0~ concentration in the water, and the free C02
excreted by the gills of fish by use of the standard nomographic method
and assuming that equilibrium of C0~ hydration is rapid.  The increase
in the concentration of excreted (X>2 in the respiratory water (as mg CO^/
liter) is given by the following relation:
                                        mol. wt. C02    P
            Increase in CO,, = DO x RQ x mol, wt      x ^           (14)
                                   90

-------
where  DO = the dissolved oxygen concentration of the water in mg/liter
       RQ = the respiratory quotient of the fish
        P = the percentage of oxygen removed from the respiratory water
            by the fish.

Kutty   has determined that the respiratory quotient is essentially
unity when the fish are spontaneously active and in near air-saturated
water.  Since the C0» is excreted along the surface of the lamellae, it
is possible that there is a pH gradient formed in the gills.  Lloyd and
Herbert proposed that the average pH shift occurred for the condition
when half the C0~ was excreted and came to equilibrium with the carbonate
system.  When their proposed explanation was applied to the sulfide
bioassay data in this study, it was calculated that the anomalous change
in H-S toxicity with test solution pH could best be explained by their
theory when assuming a respiratory quotient of unity and that the fat-
head minnow absorbs about 60% of the dissolved oxygen available through
the gills.  If a respiratory quotient of 0.8 is used, as was proposed by
Lloyd and Herbert, then about 75% of the dissolved oxygen  available
through the gills would need to be absorbed to satisfy  their theory.

The above proposed explanation  for the  toxicity  to fish of weak acids
and bases is indeed unique but  for a number of reasons  it  may not be
                                                               59
appropriate.  First of  all,  it  was assumed by Lloyd  and Herbert   in
their calculations that the  utilization of oxygen  in the water passing
over  the gills of rainbow trout is about 80%.  Recent  studies have  shown
that  the percentage utilization of oxygen  is variable  between  fish  and
for a given fish under  different conditions.  According to information
presented in a review by Shelton,  almost  all of  the reported  utilization
values are less than 80% under  near  ideal  environmental conditions.
During the stress occurring  in  an acute toxicity bioassay, it  is  not
unreasonable to assume  that  the utilization might  even be  considerably
less  than the  60% value necessary to justify  the pH  drop at  the  gill
theory for the sulfide  bioassays.
                                    91

-------
A second and most important criticism of this theory arises when one
examines the manner in which CCL is excreted at the gills and the amount
                                        62
which is excreted.  According to Randall,  the rate at which CO  is
exchanged across the gills depends on the dimensions of the epithelium,
the concentration gradient, and the diffusion coefficient of CO-.  These
factors are such that the expected changes in C0~ tension in the inspired
and expired water passing over the gills are small and at most a few mm
Hg.  To satisfy Lloyd and Herbert's explanation of the sulfide bioassay
data, the increase in CO- at the gill surface when assuming a 60% utili-
                                                                 62
zation would need to be about 6.2 mg/liter or 2.8 mm Hg.  Randall   also
stated that the mechanism for the excretion of C0_ from the cells of
freshwater teleosts includes the conversion to bicarbonate of some of
the CO- in the blood by carbonic anhydrase located in the gill epithelium.
Therefore,  along with the free C0~ entering the water, bicarbonate passes
across the gill epithelium by an exchange diffusion mechanism which
involves chloride.  Stumm and Morgan   indicate that the hydration/dehy-
dration reaction
                      C°2(aq) + H2° ;==T H2C°3

proceeds very slowly and the establishment of the hydration equilibrium
at pH values near 7 requires a finite time on the order of many seconds.
It should also be noted that the formation of CO- from the bicarbonate
actively diffusing out of the gill epithelium is slow.  Therefore, since
a volume of water is generally considered to be in contact with the gill
epithelium for less than 2 seconds, and the hydration of CO- and forma-
tion of CO- from bicarbonate in water is on the order of many seconds,
the major portion of the rise in PCQ  and ultimate pH shift at equili-
brium should occur after the water has left the respiratory surface.
This process tends to maintain the necessary PCO  gradient between blood
                                                2
and water.
Molecular species are known to penetrate membranes more readily than
charged ions.  If it is assumed that molecular H»S is the major internal
                                    92

-------
toxic sulfide species,  then an explanation for the observed relationship
between test solution pH and sulfide toxicity may include the penetration
of the gill epithelium mainly by molecular H~S accompanied by a change
in blood and intracellular pH in accord with ambient C0? tensions.
      f O                                               ^
Albers   has stated that the relationship between fish blood pH and log
Prn  is linear and may differ in slope, depending on buffer capacity and
absolute values with different species.  In his review it can be seen in
Figure 11 (p. 197) that for Cyprinus carpio, as the blood P    is increased
from about 2 to 12 mm Hg, the pH decreases from about 7.9 to 7.4.  The
blood pH of the tested fathead minnows probably decreased with increasing
ambient C09 tensions (decreasing test pH).  The intracellular pH is
generally lower than that of the blood.  Therefore, the observed increase
in molecular H_S toxicity with increasing test solution pH may partially
be explained by the difference in the degree of ionization of molecular
H»S following penetration of the gill epithelium for fish exposed to
different test pH values.  A similar expanation for the effect of ambient
C09 tensions on the toxicity of ammonia was proposed by Warren and
        64
Schenker.   A change in the permeability of the gill to molecular H»S
may also contribute to the observed apparent change in H^S toxicity.
However, the nearly fourfold change in penetration rate over the pH
range of 7.1 to 8.7 necessary to account entirely for the change in H^S
LC50 values from 57.3 to 14.9 yg/liter is not very likely.

Among  its various modes of toxic action, poisoning by sulfide  species
includes the formation of metal complexes.  It has been documented  (White,
Handler, and Smith  ) that sulfide, including  its anionic species,  can
inhibit certain enzymes by formation of  complexes with  essential metal
ions contained in  the enzyme.   Sulfide species are known  to  inhibit
iron-containing enzymes such as peroxidase  and catalase.   They also bind
to  the ferric ion  of cytochrome oxidase  and thereby  inhibit  0- metabolism.
The sulfmethemoglobin formed from  the  combination of  sulfide species with
the ferric  ion of  methemoglobin cannot be  further metabolized  and  remains
until  the  cell is  phagocytized.
                                     93

-------
If dissolved sulfide species other than molecular H~S can penetrate the
gills and since they can inhibit certain enzymes, the toxicity of sulfide
solutions to fish should not be entirely related to the ambient molecular
H7S level but should be more closely correlated with the internal total
dissolved sulfide concentration.  The negative relationship between pH
and molecular H-S LC50 levels can possibly be explained by assuming that
the HS  ion penetrates the gill epithelium, though presumably to a much
lesser extent than molecular H^S, and contributes to the toxicity of
sulfide solutions to a greater degree as the pH increases.  Therefore,
the results obtained in this study demonstrated that the acute toxicity
to fathead minnows of sulfide solutions does not depend entirely on the
concentration of ambient molecular H«S, but that the HS  ion may con-
tribute to a much lesser extent to the toxicity of these solutions.
However, definition of the relative toxicity of these two dissolved
sulfide species awaits further physiological and toxicological evalua-
tion.
                                   94

-------
                             SECTION VII
                              REFERENCES

1.  American Public Health Association, American Water Works Associa-
    tion, Water Pollution Control Federation.  Standard Methods for
    the Examination of Water and Wastewater.  13th ed.  New York,
    American Public Health Association, Inc., 1971.  874 p.

2.  Pomeroy, R.  Hydrogen Sulphide in Sewage.  Sewage Works J. 13(3);
    498-505, May 1941.
                                      ii
3.  Jellinek, K., and J. Czerwinski.  Uber  die Dissoziation von H^S,
    Na2S und NaHS in WSsseriger Lbsung.   (Dissociation  of  H2S, Na2S
    and NaHS in Aqueous Solutions).  Z.Phys. Chem.   3.02:438-479,  1922.

4.  Loy, H. L., and D. M. Himmelblau.   The  First  lonization Constant of
    Hydrogen Sulphide in Water.  J.  Phys. Chem.   (>5_:264-267,  February
    1961.

5.  Barnes, H.L., H. C. Helgeson, and A.  J. Ellis.   lonization Constants
    in Aqueous Solutions.   In:  Handbook  of Physical Constants,  Clark,
    S. P.,  Jr.  (ed.).  New  York, Geological Society of  America,  Inc.,
    1966.   p.  401-412.

6.  Wright, R. H.,  and 0. Maass.  The  Conductivity of Aqueous Hydrogen
    Sulphide.  Can. J. Res.  6^:588-595, 1932.
                                    95

-------
 7.  Tumanova, T. A., K. P. Mishchenko, and I. E. Flis.  The Dissociation
    of Hydrogen  Sulphide  in Aqueous  Solutions at Various Temperatures.
    Zh. Neorg. Khim.   2^1990-1997, 1957.

 8.  Auerbach, F.   Der  Zustand  des  Schwefelwasserstoffs  in Mineral-
    quellen.   (The State  of Hydrogen Sulfide  in Mineral Springs).
    Z. Phys. Chem.   49:217-223,  1904.

 9.  Epprecht, A. G.  Die  Dissoziationskonstante der Ersten  Stufe des
    Schwefelwasserstoffs.  (The  Dissociation  Constant  of  the  First  Step
    of Hydrogen  Sulfide lonization).  Helv. Chim.  Acta.  2^:205-211,
    1938.

10.  Kubli,  H.  Die Dissoziation  von  Schwefelwasserstoff.   (The Dissoci-
    ation of Hydrogen  Sulfide).   Helv. Chim.  Acta. ^9:1962-1973, 1946.

11.  Lewis,  G.  N.,  and  M.  Randall.   Thermodynamics and the Free Energy
    of Chemical  Substances.   New York, McGraw-Hill, 1923.  653 p.

12.   Yui,  N.  The lonization Constant of Hydrogen Sulfide.  Sci. Rep.
     Tohoku Univ.,  First  Ser.  ^5:53-61, 1951.

13.   Ellis, A. J.,  and R. M. Golding.  Spectrophotometric Determination
     of the Acid Dissociation Constants of Hydrogen Sulphide.   J. Chem.
     Soc.  (London).  1959:127-130, January 1959.

14.   Hseu, T., and G. A.  Rechnitz.  Analytical Study of a Sulfide Ion-
     Selective Membrane Electrode in Alkaline Solution.  Anal. Chem.
     40(7):1054-1060, June 1968.

15.  Chen, K. Y., and J.  C. Morris.  Oxidation of  Sulphide by  02: Cataly-
     sis and Inhibition.  Am.  Soc. Civil Eng. Sanitary  Eng. Div.  J.  98:
     215-227, February  1972.
                                    96

-------
   16-   Pomeroy, R.  The Determination of Sulphides  in  Sewage.   Sewage Works
        J.   1(4) .-572-591, July 1936.

   17-   Bethge,  P.  0.   On the Volumetric Determination of Hydrogen Sulfide
        and  Soluble Sulfides.  Anal. Chim.  Acta.  j):129-139, 1953.

  18.   Bethge,  P.  0.   On the Volumetric  Determination of Hydrogen Sulfide
        and Soluble Sulfides.   Anal.  Chim. Acta.   K):310-311,  1954.

  19.  Gustafsson, L.  Determination of  Ultramicro Amounts  of Sulphate as
       Methylene Blue.  I.  The Colour Reaction.  Talanta.   4^-227-235, 1960.

  20.   Cline,  J. D.  Spectrophotometric Determination of Hydrogen Sulfide
       in Natural  Waters.  Limnol. Oceanogr.  J.4/3) :454-458, May 1969.

 21.   Zutshi, P. K., and T.  N. Mahadevan.   Spectrophotometric Method for
       the Standardization  of Sulphide  Radicals in Solutions.  Ind. J.
       Chem.  9^:719-720,  July 1971.

 22.  Hofmann, K., and R. Hamm.   Zur Bestimmung von  Schwefelwasserstoff
      mit  N,N-Dimethyl-p-Phenylenediamin und Eisen(III)-Chlorid.  (De-
      termination of Hydrogen Sulfide with N,N-Dimethyl-p-Phenylenediamine
     and  Iron (III) Chloride).  Z. Anal. Chem.  _2_32:167-172, June 1967.

23.  Grasshoff, K. M., and K. M.  Chan.   An  Automatic Method for the  Deter-
     mination of Hydrogen  Sulphide in  Natural  Waters.   Anal.  Chim. Acta.
     .53:442-445, 1971.

24.  Johnson, C. M., and H.  Nishita.  Microestimation of Sulfur—in Plant
     Materials,  Soils, and Irrigation Waters.  Anal.  Chem.  24:736-742,
     April 1952.
                                  97

-------
25.  Fogo,  J.  K.,  and M. Popowsky.  Spectrophotometric Determination of
     Hydrogen Sulfide, Methylene Blue Method.  Ind. Eng. Chem. Anal. Ed.
     ^1:732-734, June 1949.

26.  Patterson, G. D., Jr.  Sulfur.  In:  Colorimetric Determination of
     Nonmetals, Vol. 8, Boltz, D. F. (ed.).  New York, Interscience,
     1958.   p. 261-308.

27.  Sands, A. E., M. A. Grafius, H. W. Wainwright, and M. W. Wilson.
     The Determination of Low Concentrations of Hydrogen Sulfide in
     Gas by the Methylene Blue Method.  U.S. Bur. Mines Rep.  Invest.
     4547,  p. 1-18, 1949.

28.  Marbach, E. P., and D. M. Doty,  Sulfides Released from  Gamma-
     Irradiated Meat as Estimated by Condensation with N,N-Dimethyl-p-
     Phenylenediamine.  J. Agr. Food Chem.  4_: 881-884, October 1956.

29.  Siegel, L. M.  A Direct Microdetermination for Sulfide.  Anal.
     Biochem.  11(1):126-132, April 1965.

30.  Zavodnov, S. S.  Colorimetric Determination of Small Amounts  of
     Hydrogen Sulfide in Mineral  waters.   Sovrem. Metody Analiza Pri-
     rodn. Vod., Akad.  Nauk SSSR.  1962.   p. 63-66.

31.  Zavodnov,  S. S.  New  Indicators  for  Colorimetric Determination of
     Small Amounts  of Hydrogen  Sulfide  in Mineral  Waters.   Gidrokhim.
     Mater ialy.   3.5:203-206,  1963.

32.  Rees, T. D., A.  B.  Gyllenspetz,  and  A.  C. Docherty.   The Determina-
     tion  of  Trace  Amounts of Sulphide  in Condensed  Steam with N,N-
     Diethyl-p-Phenylenediamine.   Analyst. JJ6:201-208, March 1971.

33.  Paez, D. M., and 0.  A. Guagnini.   Isolation and  Ultramicro  Deter-
     mination of  Hydrogen Sulphide in Air or Water by Use of Ion-Exchange
                                    98

-------
         Resin.  Mikrochim. Acta.   .2:220-224,  1971.

    34.   Jacobs, M. B., M. M. Bravermann, and  S. Hochhelser.  Ultramicro-
         determination of Sulfides  in Air.  Anal. Chem.  ^9:1349-1351,
         September 1957.

   35.   Sensenbaugh,  J.  D.,  and W.  C.  L. Hemeon.  A Low-cost Sampler for
        Measurement of Low Concentration of Hydrogen Sulfide.   Air  Repair.
        1:1-4,  1954.

   36.   Smith, A. F.,  D. G. Jenkins, and  D. E. Cunningworth.  Measurement
        of Trace Quantities of Hydrogen Sulphide in  Industrial Atmospheres.
        J.  Appl. Chem. (London).  LL:317-329> 1961.

  37.   Hochheiser, S., and L. A. Elfers.  Automatic Sequential Sampling of
        Atmospheric H2S by Chemisorption on Mercuric Chloride-treated Paper
        Tape.  Environ. Sci. Technol.  4^:672-676,  1970.

  38.   Andrew, T. R.,  and P.  N.   R.  Nichols.  The Determination of Hydrogen
       Sulphide in the Atmosphere.  Analyst.  ^0:367-370, june 1965.

  39.  Axelrod,  H.  D., J.  H.  Gary,  J.  E. Bonelli, and J. P.  Lodge,  Jr.
      Fluorescence Determination of Sub-Parts  per Billion Hydrogen  Sul-
      fide  in the Atmosphere.   Anal. Chem.  41. .-1856-1858, November  1969.

 40.   Avrahami, M., and R. M. Golding.  The Oxidation of the Sulphide Ion
      at Very Low Concentrations in Aqueous Solutions.  J. Chem. Soc. (A),
      1968:647-651,  1968.

41.  Natusch, D.  F. S.,  H. B. Klonis, H.  D. Axelrod, R. J.  reck, and
     J.  P.  Lodge, Jr. Sensitive Method for Measurement of  Atmospheric
     Hydrogen Sulfide.  Anal. Chem.  44(12):2067-2070,  October 1972.

42.  Schneider, C. R., and H. Freund.  Determination of Low Level Hydro-
                                    99

-------
     cyanic Acid in Solution Using Gas-liquid Chromatography.  Anal.
     Chem.  _34:69-74, January 1962.

43.  Claeys, R. R., and H. Freund.  Gas Chromatographic Separation of
     HCN on Porapak Q Analysis of Trace Aqueous Solutions.  Environ.
     Sci. Tech.  ,2:458-460, June 1968.

44.  Nelson, K. H., and I. Lysyj,  Analysis of Water for Molecular
     Hydrogen Cyanide.  J. Water Pollut. Contr. Fed.  43_: 799-805, May
     1971.

45.  Broderius, S. J.  Determination of Molecular Hydrocyanic Acid in
     Water and Studies of the Chemistry and Toxicity to Fish of Metal-
     cyanide Complexes.  Ph.D. Thesis, Oregon State Univ., Corvallis.
     1973.  287 p.

46.  Jacques,  A. G.  The Kinetics of Penetration.  XII Hydrogen Sulfide.
     J. Gen. Physiol.  19:397-418, 1936.

47.  Longwell, J., and F. T. K. Pentelow.  The Effect of Sewage on Brown
     Trout (Salmo trutta L.) .  J. Exp. Biol.  3^2:1-12, January 1935.

48.  Jones, J. R. E.  A Further Study of the Reactions of Fish to Toxic
     Solutions.  J. Exp. Biol.  25(1):22-34, March 1948.

49.  Bonn, E. W., and B. J. Follis.  Effects of Hydrogen Sulfide on
     Channel Catfish  (Ictalurus punctatus).  Trans. Amer. Fish. Soc.
     96:31-37, January 1967.

50.  Dymond, J. R., and W. B. Scott.  Fishes of Patricia Portion of  the
     Kenora District, Ontario.  Copeia.  1941(4):243-245, November  1941.

51.  Rawson, D. S., and J. E. Moore.  The  Saline  Lakes of Saskatchewan.
     Can. J. Res.  22(6):141-201, December 1944.
                                   100

-------
52.  McCarraher,  D.  B.,  and R.  Thomas.   Some Ecological Observations  on
     the Fathead Minnow, Pimephales promelas, in the Alkaline Waters  of
     Nebraska.  Trans. Amer. Fish. Soc.  j^:52-55, January 1968.

53.  Brungs, W. A., and D. I. Mount.  A Water Delivery System for Small
     Fish-holding Tanks.  Trans. Amer. Fish. Soc.  99/4):799-802,
     October 1970.

54.  Mount, D. I.,  and  R. E. Warner.   A  Serial-dilution Apparatus  for
     Continuous Delivery  of Various  Concentrations of Materials in Water.
     U.S.  Public Health Serv.,  Cincinnati..  Publ.  No.  999-WP-23.  1965.
      16 p.

 55.   Dixon, W.  J.   BMD  Biomedical Computer  Programs.   3rd ed.  Berkeley,
      U. of California Press,  1973.  773  p.

 56.   Litchfield,  J. T., Jr.,  and F.  Wilcoxon.   A Simplified Method of
      Evaluating Dose-Effect Experiments.  J. Pharmacol. Exp. Ther.
      ^6(2):99-113,  June 1949.

 57.   Stumm, W., and J.  J. Morgan.  Aquatic Chemistry.  An Introduction
      Emphasizing Chemical Equilibria in Natural Waters.  New York, John
      Wiley and Sons, Inc., 1970.  583 p.

 58.  Garber, W. F., J. Nagano, and F. F. Wada.   Instrumentation for
      Hydrogen Sulfide Measurement.  J. Water Pollut.  Contr.  Fed.
      j4_2(Pt. 2):R209-220, May 1970.

 59.  Lloyd, R., and D. W. M. Herbert.   The Influence of Carbon Dioxide
      on the Toxicity of  Un-ionized  Ammonia to  Rainbow Trout (Salmo
      gairdnerii Richardson).   Ann.  Appl. Biol.   4^(2):399-404, June
      1960.
                                    101

-------
60.  Kutty, M. N.  Respiratory Quotients in Goldfish and Rainbow Trout.
     J. Fish. Res. Board Can.  ^5:1689-1728, August 1968.

61.  Shelton, G.  The Regulation of Breathing.  In:  Fish Physiology,
     Vol. IV.  The Nervous System, Circulation, and Respiration,
     Hoar, W. S., and D. J. Randall (ed.).  New York, Academic Press,
     1970.  p. 293-352.

62.  Randall, D. J.  Gas Exchange in Fish.  In:  Fish Physiology, Vol.
     IV.  The Nervous System, Circulation, and Respiration, Hoar, W. S.,
     and D. J. Randall (ed.).  New York, Academic Press, 1970.  p.
     253-286.

63.  Albers, C.  Acid-Base Balance.  In:  Fish Physiology, Vol. IV.  The
     Nervous System, Circulation, and Respiration, Hoar, W. S., and D.
     J. Randall (ed.).  New York, Academic Press, 1970.  p. 173-205.

64.  Warren, K. S., and S. Schenker.  Differential Effect of Fixed Acid
     and C02 on Ammonia Toxicity.  Amer. J. Physiol.  203:903-906, 1962.

65.  White, A., P. Handler, and E. L. Smith.  Principles of Biochemistry.
     4th ed.  New York, McGraw-Hill, 1968.  1187 p.
                                   102

-------
                             SECTION VIII



                               APPENDIX





DERIVATION OF AN EQUATION TO CALCULATE THE FRACTION OF DISSOLVED  SULFIDE


AS MOLECULAR H2S WHEN THE IONIC STRENGTH OF THE SOLUTION  IS LESS  THAN  0.10



The dissociation of H-S in aqueous solution can be represented  by:
                        Kl     +     -  K2      +    2

               H2S(a ) ^      H  + HS  ^       2H  + s
The equilibrium constants K.. and K_ are given by:
                        V  '
                   __     n     no     t *r      i      J                  /1 /• \
                   K, * 	 and K0 = 	               (16)
                    la              2

                          H2S(aq)
where        K-  =  10~   at  20  C

                    -ITS
             K2  *  10 1J   at  20  C


       au        =  activity of undissociated molecular H_S dissolved in
        H0S,   x                                         *•
         2  (aq)
            v 4'    water


           a   -  =  activity of hydrosulfide ion
            no

           a 2-  •  activity of sulfide ion


            a..+  =  activity of hydrogen ion.
                                     103

-------
The concentration of sulfide species is determined by the first and


second dissociation constants and the equilibrium solution conditions.


By calculation it can be demonstrated that when the pH is less than


about 11 the second equilibrium constant K~ is so small that it and the

                           2-
presence of sulfide ions (S  ) can be neglected in equilibrium calcula-


tions in the following discussion.
The total concentration of dissolved sulfide species in solution may be


expressed by:




            Dissolved sulfide =   [DS] = [H,S,  v] + [HS~].           (17)
Since
where   a  = activity of the species A
         J\

       [A] = molar concentration


        f  = dimensionless number called the activity coefficient


then
                            _ aR+.[HS


                         Kl =
                               1 r«">J  H2S(aq)





From the dissolved sulfide expression  (equation 17)
                      [HS ] =  [DS] - [H2S(aq)]                        (19)
                                   104

-------
                         1"   [H2S(aa)]'fH S
                                 2 (aq)   H2S(aq)
                        ir  = \ **	"•*•*	x  N^ +*•	fto	^  \^Q/  _s

                         1 ~         \K"     '  "
     Kl
        [H2S(aq)] fHS) + (V fHS"  tH2S(aq)])  - aR+ £RS- [DS]
                [H2SCaq)]  Kl fHS     + V  fHS-   = V
                                      foe'  [DS]
                                     ,   ,
                                     (aq)
                                     V fHS~
                                   2  (aq)
In freshwaters of relatively  low  ionic strength, it can be assumed that


for the molecular species H0S,  x
                            2  (aq)
Thus
                                    105

-------
This expression may also be written in concentration units where C
corresponds to yg/liter as H_S:
                                 V"
By definition a + = antilog   (-pH) when standard buffers prepared
according to the National Bureau of Standards recommendations are used
to standardize the pH meter.  Therefore:
                            antilog (-PH)-fuc-
                                                  . r                (24)
                H2S(aq)   Kl + antil°8 (-
From the above expression, it is apparent that the concentration of
molecular H2S in an aqueous solution of known dissolved sulfide concen-
tration can be calculated for various solution pH and temperature values
by knowing the following equilibrium solution parameters:
     K., - the relationship between the equilibrium constant for the
          first dissociation of H0S,  , in aqueous solution with tempera-
                                 2 (aq)
          ture.  From research performed in this study this relationship
          can for all practical purposes be defined by equation (12)
                     pKx = 7.252 - 0.01342 T  (C).

   fHg_ - the activity coefficient of the HS" ion which varies with
          solution ionic strength and temperature.

In dilute solution of electrolytes, the individual ion activity coef
ficient is given by the extended Debye-Huckel expression:
                        antilog
                                     A'ZHS-
(25)
                                    106

-------
  where     A - a constant
            B = a constant
           a° = a constant
         ZRS_ = charge of the HS~ ion equal to 1
            I = ionic strength of the solution.

 Garrels  and Christ1 give the value of ajs_ as 3.5 x 10*  and  give  tables
 of data  for A and  B which for the temperature range of  0 to 30 C can be
 expressed by:

                      A  =  0.4880 + 0.0082  T

                      B  =  (0.3241 + 0.00016 T)-108
where   T = temperature in C.
Thus
                            [~  (0.4880 + 0.00082  T)  y/T  ~|
            fHS- ~ antll°8  [_  1 +  ,/T (1.134 + 0.00056  T)J
Therefore, the fraction of dissolved sulfide  (DS) as molecular H_S is
given by:

                           V
                            CDS
                                   107

-------
                -,  /  „>   ,--,   F   (0.4880 + Q.00082 T) y/Tl
where   f = antilog(-PH).antilog|- x +X/T(1.134 + 0.00056 T)J

        k = antilog - [7.252 - 0.01342 T]                             (29)

       pH = final pH of the solution
        T = temperature of the solution in C
        I = ionic strength of the solution.

Much of the information used in  this derivation was previously  included
                                  2
in an information report by Clarke.

LITERATURE CITED
1.  Carrels, R. M., and C. L. Christ.  Solutions, Minerals and  Equilibria.
    New York, Harper and Row, 1965.  450 p.

2.  Clarke, T. R.  Physical, Chemical, and Biological Effects of H-S
    Releases to Lake Huron.  The Hydro-Electric Power Commission of
    Ontario, Thermal Generation Division,  Central Nuclear Services.
    Report No. CNS-IR-191.  March 1974.  62 p.
                                    108

-------
                                   TECHNICAL REPORT DATA
                            (Please read Instructions on the reverse before completing)
 1. REPORT NO.
 EPA-600/3-76-062b
                             2.
                                                           3. RECIPIENT'S ACCESSION-NO.
4. TITLE AND SUBTITLE
 Effect  of Hydrogen Sulfide on Fish  and Invertebrates
 Part II - Hydrogen Sulfide Determination and Relation-
 ship between pH and Sulfide Toxicity
                             5. REPORT DATE
                                July 1976  (Issuing Date)
                             6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
 Steven J.  Broderius
 Lloyd L.  Smith, Jr.
                                                           8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
  Department of Entomology, Fisheries,  & Wildlife
  University of Minnesota
  St.  Paul, Minnesota 55108
                             10. PROGRAM ELEMENT NO.

                                1BA608
                             11. CONTRACT/GRANT NO.
                                                              R800992
 12. SPONSORING AGENCY NAME AND ADDRESS
  Environmental Research  Laboratory
  U.S.  Environmental Protection  Agency
  Office of Research and  Development
  Duluth, Minnesota 5580*1
                             13. TYPE OF REPORT AND. PERIOD COVERED
                             Final  (Aug. 1972-Mar.  1975)
                             14. SPONSORING AGENCY CODE
                             iPA  -  ORD  (OHEE)
 15. SUPPLEMENTARY NOTES
  See  Part I,  EPA-600/3-76-062a
 16. ABSTRACT
       An analytical method was  developed for the direct determination  of ug/liter
  concentrations of molecular  H2S.   The procedure involves bubbling  compressed nigrogen
  through an aqueous sulfide solution to displace H2S which is collected in a glass
  bead concentration column and  measured colorimetrically.  The R^S  concentration is
  calculated from the determined sulfide displacement rate and by reference to a log
  linear standard curve relating temperature with the E^S displacement  rate to the
  H S concentration in standard  solutions.   To permit accurate determination of HpS
  from the determined dissolved  sulfide concentration and fraction  of dissolved
  sulfide as H2S for specific  conditions of temperature and pH, the  apparent linear
  relationship between pK]_ for H2S(aq)  and temperature was defined.   This procedure of
  calculating H2S in various waters  and effluents was confirmed by  the direct technique

       The described analytical  technique 'was used to define  the  relationship between
  test pH and sulfide toxicity to the fathead minnow.  Within the pH range of 7.1
  to 8.7, 96-hr LC50 values for  molecular H2S decreased linearly  from 57;.3 to lU.9
  ug/liter with increasing pH.  However, the log 96-hr LC50 values  of dissolved
  sulfide increased linearly from 6U.O to 780.1 ug/liter with increasing test pH
  ranging from 6.5 to 8.7-
17.
                               KEY WORDS AND DOCUMENT ANALYSIS
                  DESCRIPTORS
  *Analytical Chemistry
   Colorimetric  analysis
   Quantitative  analysis
  ^Equilibrium constants
   Microanalysis
   Thermochemistry
pH
*Toxicity
minnows
*Hydrogen  Sulf-
ide.
                                              b.lDENTIFIERS/OPEN ENDED TERMS
Partition coefficient
Fathead minnow
96-hour LC50
Displacement rate
Vapor phase equilibratior
                                          c.  COSATI Field/Group
07/B
06/T
13. DISTRIBUTION STATEMENT

  Release to  Public
                19. SECURITY CLASS (This Report)
                   Unclassified	
                          21. NO. OF PAGES
                                119
                                              20. SECURITY CLASS (Thispage)
                                                 Unclassified
                                                                        22. PRICE
EPA Form 2220-1 (9-73)
                                            109
                                                              OUSGPO: 1976-657-695/5455 Region 5-1

-------