&EPA
United States
Environmental Protection
Agency
Environmental Research
Laboratory
Duluth MN 55804
EPA-600 3 8(
Research and Development
Direct Photolysis of
Hexacyanoferrate
Complexes
Proposed
Applications to the
Aquatic Environment
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6, Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the ECOLOGICAL RESEARCH series. This series
describes research on the effects of pollution on humans, plant and animal spe-
cies, and materials. Problems are assessed for their long- and short-term influ-
ences. Investigations include formation, transport, and pathway studies to deter-
mine 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-600/3-80-003
January 1980
DIRECT PHOTOLYSIS OF HEXACYANOFERRATE COMPLEXES
Proposed Applications to the Aquatic Environment
by
Steven J. Broderius and Lloyd L. Smith, Jr.
Department of Entomology, Fisheries, and Wildlife
University of Minnesota
St. Paul, Minnesota 55108
Grant No. R805291
Project Officer
John E. Poldoski
Environmental Research Laboratory
Duluth, Minnesota 55804
ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
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
-------
FOREWORD
This work represents an effort to characterize the effect of sunlight and
various other environmental parameters on the decomposition of iron-cyanide
complexes to give hydrogen cyanide. Past work by the authors indicate that
hydrogen cyanide can be highly toxic to aquatic organisms. Therefore, this
work should provide additional insight for making hazard assessments of
iron-cyanide discharges into the environment.
J. David Yount
Acting Director
Environmental Research Laboratory-Duluth
111
-------
ABSTRACT
The theory and computations described by Zepp and Cline (1977) were
experimentally tested in predicting the direct photolysis rates of dilute
hexacyanoferrate (II) and (ill) solutions in the aquatic environment.
Essential information for these calculations includes the quantum yield for
the photoreaction, molar extinction coefficients of the complex ions for
wavelengths > 295 nm, solar irradiance data used to calculate specific
sunlight absorption rates, and the assumption that the photolysis reaction
obeys a first-order kinetic rate expression. Direct photolysis rates of the
irreversible photochemical reactions are calculated as a function of the time
of year, latitude, time of day, meteorological conditions, and depth in
natural water bodies. Light of wavelengths < 480 nm is active in the
photolysis reactions, and pH, temperature, and concentration all affect the
reaction to varying degrees. Assuming first-order kinetics, in which the rate
constant was approximately concentration independent within the range of
25-100 ug/1 total cyanide, the minimum quantum yields of HCN formation were
0.14 and 0.0023 for the iron (II) and (III) complexes, respectively. These
values correspond to minimum, nearsurface, midday half-lives at midsummer of
about 18 and 64 min at St. Paul, Minn. The photolysis rate at various fixed
depths in a natural water column, when compared with that at the surface,
decreases exponentially with depth. It is suggested that the photolysis
reactions are enhanced by suspended material in turbid waters because of the
forward scattering of light when compared with that theoretically calculated
from beam attenuation coefficients. Hexacyanoferrate (II) and (III) solutions
of equal initial total cyanide concentration respond photochemically quite
differently from one another in solutions prepared with deionized water, but
respond in a similar manner for solutions prepared with natural waters. The
potentially rapid photodecomposition of iron-cyanides with formation of HCN
suggests that this phenomenon may be of toxicological importance under certain
environmental conditions.
This report was submitted in fulfillment of Grant No. R805291 by the
Department of Entomology, Fisheries, and Wildlife, University of Minnesota,
under the sponsorship of the U.S. Environmental Protection Agency. This
report covers a period from July, 1977 to August, 1978, and work was completed
as of August, 1978.
-------
CONTENTS
Foreword ill
Abstract iv
Figures vi
Tables viii
Acknowledgments ix
1. Conclusions 1
2. Recommendations 3
3. Introduction 4
4. Materials and Methods 8
Analytical methods 8
Test procedures and apparatus 8
Computational approach 10
5. Results and Discussion 12
Volatilization of HCN 12
Free cyanide and HCN determinations 14
Kinetics of photodecomposition 14
Molar extinction coefficients 14
Specific sunlight absorption rate ..'........ 16
Properties of photochemical reaction 16
Quantum yield 23
Time of year and latitude 23
Diurnal change ...... 27
Attenuation by natural waters 27
Sky conditions and photolysis rate ......... 39
Photolysis in natural waters ....... 39
References ............. 45
Appendix
A. Beam Attenuation Coefficients 48
-------
FIGURES
Number Page
1 Electronic absorption spectra of hexacyanoferrate (II) and (III)
solutions 17
2 Specific sunlight absorption rates of hexacyanoferrate (II) and
(III) complexes as a function of wavelength at midday and mid-
summer, latitude 40° N 19
3 Relative photolysis rate constants normalized to the values
determined for solutions with an initial total cyanide concen-
tration of 25 ug/1 CN and as a function of the average iron-
cyanide concentration during the rate determination period ... 22
4 Midday half-lives for 100 Mg/1 CN hexacyanoferrate (II) solutions
at near surface depths for different times of the year under
full sunlight conditions at St. Paul, Minn 24
5 Midday half-lives for 100 ng/1 CN hexacyanoferrate (III) solutions
at near surface depths for different times of the year under
full sunlight conditions at St. Paul, Minn 25
6 Midday half-lives for direct photolysis of pure water hexacyano-
ferrate (ll) and (III) solutions (near surface) as a function of
the time of year for several northern latitudes. Values are
relative to July 1 rate at 45° N latitude 26
7 Diurnal variation of direct photolysis rates of pure water hexa-
cyanoferrate (II) solutions (near surface) relative to
photolysis rates at midday on July 1 at latitude 45° N,
longitude 93.2° W 28
8 Diurnal variation of direct photolysis rates of pure water hexa-
cyanoferrate (III) solutions (near surface) relative to
photolysis rates at midday on July 1 at latitude 45° N,
longitude 93.2° W 29
Time of day dependence of direct photolysis rates of pure water
hexacyanoferrate (II) solutions (near surface) relative to
photolysis rates for midday at St. Paul, Minn, on October 20,
1977. Theoretical relationship indicated by smooth line .... 30
VI
-------
Number Page
10 Time of day dependence of direct photolysis rates of pure water
hexacyanoferrate (III) solutions (near surface) relative to
photolysis rates for midday at St. Paul, Minn, on October 21,
1977. Theoretical relationship indicated by smooth line .... 31
11 Attenuation coefficients relative to deionized water for natural
water samples collected in north-central United States 32
12 Influence of turbidity on the experimentally determined to
theoretically calculated relative photolysis rate of hexa-
cyanoferrate (II) and (III) solutions 35
13 Calculated depth-dependence of the direct photolysis of hexa-
cyanoferrate (II) at midday and midsummer for latitude 40° N
when using beam attenuation coefficients and assuming complete
mixing of the water column 38
14 Relationship between photolysis rate and solar radiation, both
normalized to that predicted for a clear day 40
VII
-------
TABLES
Number page
1 Rate of HCN loss from sodium cyanide .solutions prepared with
deionized water and in open cylindrical jars under laboratory
conditions 13
2 The HCN concentration estimated by the indirect colorimetric
method and that determined by the direct vapor phase equili-
bration procedure for hexacyanoferrate (ll) and (ill) solutions
exposed to sunlight and initially containing 100 ug/1 CN . . . . 15
3 Molar extinction coefficients of hexacyanoferrate (II) and (III)
solutions 18
4 The linear relationship between exposure period (X) near midday
for different meteorological conditions and calculated log
iron-cyanide concentration (Y), for solutions with initial
complex concentrations up to 200 pg/1 as total cyanide, as
indicated by the regression correlation coefficient 20
5 Experimentally determined and theoretically computed relative
photolysis rates for hexacyanoferrate (II) and (ill) solutions
as normalized to deionized water controls and affected by
bentonite and wind-blown silt. Computed rates are based on
beam attenuation coefficient measurements 34
6 Physical properties of various natural waters and the theoretical
and determined rate of decrease in the depth-dependent direct
photolysis rate of hexacyanoferrate (II) and (III) solutions
prepared with deionized water and exposed to natural light at
specific fixed depths at latitude 45° N, longitude 93.2° W . . . 37
7 Determined photolysis rates for hexacyanoferrate (II) solutions
prepared with different waters relative to the rates for
similarly exposed solutions prepared with deionized water ... 41
8 Determined photolysis rates for hexacyanoferrate (ill) solutions
prepared with different waters relative to the rates for
similarly exposed solutions prepared with deionized water ... 43
9 Ratio of determined photolysis rates for hexacyanoferrate (III)
to (ll) solutions of equal initial total cyanide concentration
prepared with different water types and exposed to the same
natural light conditions 44
viii
-------
ACKNOWLEDGMENTS
The authors wish to express our appreciation to Drs. Richard Zepp, Donald
Bahnick, and William Swenson for reviewing this manuscript and to both Dr.
Zepp and David Cline of the Southeast Environmental Research Laboratory, U.S.
Environmental Protection Agency, Athens, Georgia for technical assistance and
use of their computer program for calculating the theoretical photodecompo-
sition of iron-cyanide complexes. We also wish to thank Mr. Walter Koenst for
his assistance in preparing figures for the final report.
IX
-------
SECTION 1
CONCLUSIONS
1. The evaporative loss of HCN from aqueous solutions is directly related to
the initial concentration and test temperature. This loss from natural
waters is relatively slow when compared with the potential photolysis
rate of hexacyanoferrate (II) and (ill) complexes for near surface
conditions.
2. The colorimetric method for the determination of free cyanide gave the
highest estimates of HCN in photodecomposed iron—cyanide solutions.
These values were essentially constant at about 1.18 and 1.56 times the
HCN determined by the volatilization method, regardless of the extent of
photodecomposition, for the hexacyanoferrate (II) and (III) solutions,
respectively.
3. Only light of wavelengths less than about 420 and 480 nm are active in
the photodecomposition of hexacyanoferrate (ll) and (ill) complexes,
respectively. The maximum absorption of solar radiation by these
complexes occurs at about 330 nm for hexacyanoferrate (II) solutions and
330 and 420 nm for hexacyanoferrate (ill) solutions.
4. The test pH, temperature, and concentration all have a varying affect on
the photolysis reaction of both iron-cyanide complexes.
5. The essentially irreversible photolysis reaction of hexacyanoferrate (II)
and (ill) complexes can be approximately described by first-order
kinetics for concentrations up to 100 pg/1 as total cyanide, with minimum
quantum yields of HCN formation determined to be 0.14 and 0.0023,
respectively.
6. The minimum, near surface, midday direct photolysis half-lives for hexa-
cyanoferrate (II) solutions containing 100 Mg/1 CN ranged from about 50
min in late fall to a minimum of about 18 min in midsummer at St. Paul,
Minn. In comparable hexacyanoferrate (ill) solutions the midday
half-lives ranged from about 160 min in late fall to a minimum of about
64 min in midsummer.
7. Midday haIf-lives for the direct photolysis of hexacyanoferrate (II) and
(III) complexes near the surface of an aqueous solution and the amplitude
of their time of year variation should increase with increasing northern
latitude.
8. The intensity of incident collimated sunlight that penetrates natural
waters is attenuated through absorption and scattering, thus diminishing
the photodecomposition rate of iron-cyanides by reducing the amount of
light available. 1
-------
9. The logarithm of photolysis rate at a specific depth in a water column,
as compared to the rate at the surface, was observed to linearly decrease
with depth in natural waters.
10. The photolysis reaction in turbid waters is enhanced by suspended
material due to forward scattering of light when compared with that
theoretically calculated from beam attenuation coefficients.
11. The relatively small concentration of HCN resulting from photolysis of
hexacyanoferrate (II) is of minimal toxicological importance below depths
of about 50 to 100 cm in most well-mixed natural waters likely to receive
this complex as a pollutant.
12. The photodecomposition of hexacyanoferrate (II) and (III) solutions is a
direct function of natural light intensity, under varying meteorological
conditions.
13. The hexacyanoferrate (ll) and (ill) complexes photochemically decompose
quite differently from one another in deionized water solutions but
produce similar amounts of HCN in solutions prepared with natural waters.
Because the photolysis reaction of the iron (II) complex was observed to
be unaffected by water type, effluents containing these iron-cyanides,
when discharged into receiving waters similar to those tested, are
expected to respond, with regards to HCN production, like that of the
hexacyanoferrate (II) complex.
14. The maximum amount of total cyanide that can be photochemically released
as HCN from dilute hexacyanoferrate (ll) and (III) solutions prepared
with deionized water was determined to be about 85 and 49%, respectively.
This indicates that for every mole of iron (II) and (III) complex, each
containing 6 moles CN, only 5 and 3 moles of CN, respectively, can be
released from the complex anions to form free cyanide through a
photolysis reaction in solutions prepared with deionized water.
15. This study indicates that sunlight photodecomposition of iron-cyanides
may provide an environmentally important pathway under certain conditions
for their conversion into toxic HCN.
-------
SECTION 2
RECOMMENDATIONS
1. Studies are needed to determine the penetration of sunlight into natural
waters as a function of light absorption and scattering characteristics
for wavelengths greater than 295 nm. These measurements could then be
used in mathematical models for predicting photolysis rates as a function
of depth.
2. Data collected during this study indicate that hexacyanoferrate complexes
may undergo photodecomposition in natural waters which can be of ecolog-
ical importance under certain circumstances. Since little is known about
the behavior of these complexes in natural aquatic environments, research
should be initiated to determine the fate of these compounds and HCN in
different receiving waters. This work could also be used to test the
proposed photolysis model in actual representative polluted waters.
3. A greater understanding of oxidation-reduction reactions involving hexa-
cyanoferrate complexes, as related to redox properties of natural waters,
is needed.
4. The toxicity of hexacyanoferrate (ll) and (ill) complexes to aquatic
organisms has been demonstrated to be essentially due to the presence of
HCN as derived from the photodecompositon reaction. Therefore, a
suitable analytical method for the determination of HCN in the ug/1 range
is needed in order to determine the actual adverse effects of cyanide on
aquatic organisms in waters receiving iron-cyanide wastes.
5. Nearly all of the lethal and sublethal effects of HCN on aquatic
organisms have been derived from continuous and constant toxicant
exposure tests. However, it is reasonable to expect HCN concentrations
in most natural waters to fluctuate because of intermittent waste
discharge, and due to the loss of cyanide and limited formation of HCN
from the photodecomposition of iron-cyanides at night. Therefore,
additional information on the toxicity of fluctuating HCN concentrations
is needed in order to more realistically establish the adverse effects of
cyanide on aquatic populations in natural waters.
-------
SECTION 3
INTRODUCTION
Appreciable amounts of ferro- and ferricyanides (i.e., hexacyanoferrate
(II) Fe(CN)£~, and (ill) Fe(CN)jT) may occur in effluents from
color film photographic processes, electrostatic conversion washes, the manu-
facture of iron and steel, the cracking of oil, and from plants which
manufacture these iron salts. It has also been established that the toxicity
to many aquatic organisms of aqueous solutions of various simple cyanides and
metallocyanide complexes is determined virtually alone by the concentration of
undissociated molecular HCN and not by the concentration of the cyanide ion
(CN~) or of most metallocyanide anions (Doudoroff, 1976; Broderius et al.,
1977). Hydrocyanic acid (HCN) is formed in hexacyanoferrate (II) and (III)
aqueous solutions mainly through photodecomposition of the iron-cyanide
anions, which are otherwise highly stable and relatively nontoxic, and through
hydrolytic reaction with water of the cyanide ions so liberated. Fish kills
in a New York stream (Burdick and Lipschuetz, 1950) and a Japanese river
(Kobayashi and Mori, 1973) were associated with the discharge of iron-cyanides
in industrial effluents at concentrations less than generally accepted as
nonlethal. The rapid development of a toxic situation was demonstrated to
;result from photodecomposition of the iron-cyanides by bright sunlight and
release of free cyanide (i.e., HCN + CN~). Similar findings were reported
by Myers (1950) for seepage from a ferromanganese blast furnace and for
chemicals from photographic processing by West (1970) and Terhaar et al.
(1972). The influence of sunlight and pH on degradation of iron-cyanides in a
polluted natural water was reported by Kongiel-Chablo (1966).
The photochemical behavior of potassium ferro- and ferricyanide in
aqueous solutions and the kinetics of their decomposition and reverse reaction
has been investigated extensively for various irradiation wavelengths in the
absorption spectral region. Solutions of both complexes when kept in the dark
or in diffused light are stable but when exposed to light of certain
wavelengths, decomposition occurs with the formation of aquopentacyanoferrates
and reduction of ferricyanide to ferrocyanide. According to Ohno and
Tsuchihashi (1965) and Ohno (1967), irradiation of aqueous hexacyanoferrate
(II) solutions with light of 366 nm causes a ligand substitution reaction
yielding aquopentacyanoferrate (II) (Fe(CN)5*H20^~) and cyanide ion.
Light of 253.7 nm resulted in a photo-oxidation reaction producing
hexacyanoferrate (III) and reducing species. Upon continued irradiation of
ferrocyanide solutions with wavelengths longer than 300 nm, Fe^+ and HCN
are also formed and the pH slowly increases during irradiation, as a result of
the hydrolysis of the cyanide ions produced, and after some time tends to
reach a constant value (Asperger, 1952). When the light is removed, the
reaction is apparently reversed under certain circumstances and the original
pH restored if the irradiation has not been too prolonged. The reverse
reaction apparently depends on the iron-cyanide concentration and is caused by
4
-------
a slower secondary reaction between the aquo salt formed and the parent
compound (MacDiarmid and Hall, 1953). Asperger ^t jQ. (1960) stated that the
reaction is apparently reversible only when no appreciable decomposition of
the aquopentacyanoferrate (II) ion occurred. Under prolonged irradiation on
aerated and relatively concentrated solutions, Fe(OH)3 (in alkaline
solutions) and Prussion blue (in acid solutions) are formed (Balzani and
Carassiti, 1970).
A mechanism for the photodecomposition of potassium ferrocyanides can be
represented according to Asperger (1952) and Mitra _et al. (1963) by the
primary reaction
Fe(CN)g~ + Fe(CN)|~ + CN~
CN- + H20 5 HCN + OH~
+ H20 t Fe(CN)5*
" + 2 H20 t Fe(CN)5'(H20)3~ + HCN + OH~ (1)
Asperger (1952) and Balzani and Carassiti (1970) have stated that in the
second phase the thermally and photochemically unstable FetClOg*(H20)3~
is decomposed by prolonged exposure to light with a progressive
photosubstitution reaction with increase in pH and release of Fe2+ plus
CN~ ions. In the presence of oxygen Fe(OH)-j can be formed.
There are a number of publications on the photochemistry of aqueous
ferricyanide solutions, but Balzani and Carassiti (1970) stated that the
complicated photochemical behavior of this complex is not yet completely
understood. Moggi £t a^. (1966) showed that the photochemical behavior of
Fe(CN)^ was qualitatively the same regardless of the wavelength of
irradiation (254, 313, or 405 run light). Spectral changes suggested that
Fe(CN)5*(H20)2~ was formed as the main product with the hydrolysis
step probably proceeding according to
2Fe(CN)|~ + 2H20 t 2Fe(CN)5'(H20)2~ + 2CN~ (2)
In the dark the complex slowly underwent an oxidation-reduction reaction which
was accelerated by light and the presence of CN~ according to
2Fe(CN)5*(H20)2- + CN~ + 20H~ t 2Fe(CN)5'(H20)3- + CNO~ + H20 (3)
with the overall hydrolysis reaction
2Fe(CN)|~ + H20 + 20H~ t 2Fe(CN)5'(H20)3~ + CN~ + CNO~ (4)
Upon irradiation by light the rates of the above reactions (equations 2 and 3)
increase, Fe2+ ions are formed, and the pH of the solution initially
increases.
-------
Balzani and Carassiti (1970) stated that in addition to a
photosubstitution reaction, a direct photoreduction of FeCCN)^" to
Fe(CN)5*(H20)^~" could occur. The oxidation-reduction reaction of hexacyano-
ferrate (III) ions was demonstrated by Adamson (1952) to be rapid whenever the
process involves a simple electron transfer and to be slow and of complex
mechanism if such a step cannot occur. The major product of the rapid reaction
was suggested to be hexacyanoferrate (II) ion (i.e., Fe(CN)&~ + e~ ->• Fe(CN)6~
0.36 volt) rather than an aquocyanide and no intermediate ions with a different
number of coordinate cyanide groups is involved. Since the nature and yields
of the products for photochemical reactions involving iron-cyanides are not
completely defined, the overall chemical changes and the reaction mechanisms
are not well established.
The first step in direct photolysis is the absorption of a light quantum
(photon) resulting in an electronically excited state of the molecule. There-
fore, only radiant energy which is absorbed by a molecule can be effective in
producing photochemical changes. Each light quantum absorbed activates one
molecule so that the efficiency of the photochemical process can be
represented by the quantum yield. This yield is defined by the ratio of the
number of molecules undergoing chemical reaction divided by the number of
photons absorbed by the reactants. In most cases this is experimentally
measured by the rate of the chemical reaction divided by the number of quanta
absorbed per second (Balzani and Carassiti, 1970).
According to a review by Balzani and Carassiti (1970), the reported
quantum yields for the photodecomposition of hexacyanoferrate (II) and (III)
complexes, as obtained by various authors using different experimental
conditions and techniques, are generally in disagreement. Upon irradiation of
aqueous hexacyanoferrate (II) solutions with wavelengths of about 313 or 365
nm, the ligand substitution quantum yield for aquopentacyanoferrate (ll)
formation was reported to be 0.1 using pH measurements in neutral unbuffered
solutions (Carassiti and Balzani, 1960) and 0.44 for a 0.5 M solution at pH
9.9 (Emschwiller and Legros, 1954). A value of 0.89 was determined from the
amount of Fe(CN)5*(H20) formed at pH 4.0 (Ohno and Tsuchihashi,
1965). The quantum yield was reported by Balzani and Carassiti (1970) to be
almost independent of Fe(CN)g~ concentration, temperature, light
intensity, and pH in the range 7-10. However, Ohno (1967) stated that the
quantum yield decreased with an increase in the hexacyanoferrate (II)
concentration.
The quantum yields for the formation of Fe(CN)5*(H20)2~ and Fe(CN)^~
from irradiation of hexacyanoferrate (ill) solutions with light of 365 nm were
reported by Balzani and Carassiti (1970) to be 0.009 and 0.014 at pH 4, and
0.0065 and approximately 0.018 at pH 10, respectively. However, these results
should be taken with reservation since questionable procedures were used in
their determination (Balzani and Carassiti, 1970).
The lack of correlation between photodecomposition tests conducted with
natural and artificial light can be due to the difference in the ultraviolet
spectral energy distribution and intensity of sunlight, and the various
artificial light sources. Because solar radiation at the earth's surface has
negligible intensity at wavelengths less than about 295 nm (Bener, 1972), the
-------
chemical of interest must have appreciable absorptivity at greater wavelengths
if significant photoreaction is to occur in sunlight. Most past studies of
the photodecomposition of hexacyanoferrate (II) and (ill) complexes were not
prompted by environmental considerations, and thus much of the information
derived from them cannot be extrapolated to natural water environments. The
primary purpose of this study was to determine the photochemical response of
iron-cyanides under natural light conditions in natural bodies of water, both
near the surface and as a function of depth. Such a study allowed us to
evaluate the relative importance of these compounds as sources of toxic
hydrocyanic acid (HCN) in natural waters. The approach used for predicting
the photochemical decomposition by solar radiation of such compounds in
aquatic environments was based on principles and equations described by Zepp
and Cline (1977).
-------
SECTION 4
MATERIALS AND METHODS
ANALYTICAL METHODS
Aliquots of hexacyanoferrate (II) and (III) solutions were tested
periodically for free cyanide (i.e., HCN + CN~) by the pyridine-pyrazolone
colorimetric method (APHA, 1971). Molecular HCN concentrations were
calculated from free cyanide and pH measurements, and dissociation constants
derived from pKgcN = 3.658 + 1662/T where T is temperature in degrees
Kelvin (Broderius and Smith, 1979). This procedure worked satisfactorily for
both the deionized and natural water solutions when appropriate corrections
were made. A method similar to that proposed by Broderius and Smith (1977)
for the determination of ^S and by Broderius (1973) for studies with
metallocyanide complexes was incorporated for the direct determination of
molecular HCN in the pg/1 range. This procedure utilizes a glass bead
concentration column coated with 0.1 N NaOH for collecting displaced HCN,
which is determined colorimetrically. The separation of HCN from a sample by
means of displacement with N2 bubbled through the solution does not upset
the chemical equilibrium involving the various cyanide forms.
Turbidity was measured on well-shaken samples with a Hach model 2100A
turbidimeter* and was reported in Nephelometric Turbidity Units (NTU). Total
and filterable residue, pH, and dissolved oxygen were also determined
according to standard procedures (APHA, 1971).
Light was measured with a LI-COR model LI-185 quantum meter**, a LI-192S
underwater quantum sensor with a spectral response based on photon absorption
between 400 and 700 nm, and a mv recorder. For experiments relating different
meteorological conditions with photolysis rate, these measurements were
integrated with exposure period to determine change in energy associated with
radiation incident to test solutions.
Molar extinction coefficients were determined with a Beckman DB-GT
spectrophotometer from electronic absorption spectra of the hexacyanoferrate
(II) and (III) complex anions in deionized water. The light from the
spectrophotometer was insufficient to cause any measurable photochemical
reaction.
TEST PROCEDURES AND APPARATUS
All chemicals were of reagent-grade quality and were used without
additional purification. Iron-cyanide solutions were usually prepared by
* Hach Chemical Company, Ames, Iowa.
**Lambda Instruments Corporation, Lincoln, Nebraska.
8
-------
weighing out potassium salts immediately before use and dissolving in
deionized water saturated with dissolved oxygen. Appropriate further dilution
under laboratory light gave the desired concentrations. Iron-cyanide
solutions with concentrations of 25 to 200 Mg/1 as total CN were generally
used. The pH in solutions prepared with potassium iron-cyanide and deionized
water was maintained with phosphate or borate buffer concentrations of about
10~^ M. Nearly all test solutions prepared with deionized water were
buffered at pH 7.8.
In photochemical experiments the reaction cells through which the light
travels to reach the dissolved reactant must be transparent to the exciting
radiation. This was accomplished by conducting the exposure experiments in
open vessels or in tightly sealed tubes which were irradiated by sunlight on
the roof of our laboratory in St. Paul, Minn, (latitude 45° N, longitude 93.2°
W). The test solutions were exposed to sunlight in 25 x 150 mm Teflon lined
screw-cap Pyrex test tubes (Corning No. 9826) unless stated otherwise. A
series of tubes for each concentration was filled to the brim and submerged in
40 mm of deionized water in a water bath thermostatically controlled, usually
at 20 _+ 0.1°C. The tubes were positioned horizontally above a black
background in a shallow tray to minimize reflected light, or for field
experiments at various fixed depths in a natural water column. It was
determined that the photolysis rates for solutions prepared with deionized
water and exposed in test tubes were not significantly different (P = 0.05)
from those determined for 8 1 of test solution in an open 2-gal Pyrex acid
battery jar (21 cm diameter x 25 cm high, Corning 6942) blackened on the
inside and tilted directly towards the sun's rays. Therefore, change in
photolysis rates due to internal reflection of sunlight, and slight absorption
of active light wavelengths in the Pyrex test tubes submerged in water
appeared to be insignificant factors. Losses of HCN to the atmosphere from
open vessels were also minimal during the relatively short exposure periods.
The concentration of free cyanide determined in an open vessel was the same
for solutions which were stirred occasionally or undisturbed.
The shadowing effect on the test solutions by the walls of the test jars
during low sun angles prevented using them in an upright position for the
determination of actual photolysis rates. However, the jars can be used for
comparative type experiments. To determine the effect of turbidity and
materials dissolved in natural waters on the photodecomposition reaction,
hexacyanoferrate (II) and (ill) solutions in test jars were exposed to natural
light filtered by about 40 mm of various solutions placed in open Pyrex
filters (27 cm square x 5 cm high) placed directly over the jars. The
photolysis rates calculated for these iron-cyanide solutions were divided by
the rates determined for comparable solutions subjected for the same exposure
period but in jars under light that had been filtered through 40 mm of
deionized water. In this way relative photolysis rates were calculated.
During the photolysis experiments, samples from test jars were frequently
taken or test tubes from a certain series were removed periodically and
analyzed spectrophotometrically for free cyanide. Control solutions kept in
the dark showed no measurable formation of free cyanide during the exposure
period of similar solutions. Based on the fraction of hexacyanoferrate (II)
and (III) that disappeared, as measured by the formation of free cyanide or
-------
HCN, photolysis rate constants and half-lives were calculated assuming first-
order kinetics.
COMPUTATIONAL APPROACH
The rate of HCN loss from open acid battery jars is proportional to the
remaining hydrogen cyanide concentration at any time. According to Palaty and
Horokova-Jakubu (1959), this loss can be expressed by a first-order rate
equation. Therefore, the half-life, or time required to reduce the original
concentration of hydrogen cyanide to one-half, is given by the equation
0.693 _ _ 0.693 t ...
\ = ~~k~ °r \ = ln(C0/C) C5)
where CQ is the initial and C the final HCN concentration after time t. If
In CQ/C calculated from the experimental data is plotted as a function of
HCN removal time, equation (5) is an expression of a straight line passing
through the zero coordinates. The slope of the line equals the rate constant
(k=ln(C0/C)/t) of HCN removal. The value of k is a function of the factors
affecting HCN loss and consequently it can be used to characterize and compare
these effects.
Quantitative calculations of direct sunlight photolysis rates and half-
lives were made from quantum yields and electronic absorption spectra of
hexacyanoferrate (II) and (III) aqueous solutions assuming first-order
kinetics. Equations defining the intensity and path lengths of direct and sky
radiation in air and water as described by Zepp and Cline (1977) were
employed. With this approach, photolysis rates under average intensities
derived from full sunlight and the whole sky on cloudless days were calculated
as a function of latitude and longitude, time of year, time of day, depth in a
water body, and other influencing parameters. The input data required to
calculate the theoretical photolysis rates are
(1) molar extinction coefficients of the hexacyanoferrate (II) and (III)
complex anions at wavelengths > 297.5 nm;
(2) the attenuation coefficient or absorbance per cm for the reaction
medium (a refractive index of 1.34 is assumed);
(3) the quantum yield for the direct photolysis reaction;
(4) the angular height of the sun as determined from the solar declin-
ation, solar right ascension, and sidereal time for the date of
interest as obtained from the American Ephemeris and Nautical
Almanac;
(5) the latitude and longitude; and
(6) the average atmospheric ozone layer thickness that pertains to the
season and location of interest (a value of 0.30 cm was assumed -
London, 1963).
10
-------
The average sunlight photolysis rate at a certain wavelength (x) is
proportional to the quantum yield (;\) for the reaction of the weakly
absorbing system. The kinetic expression for this formula as discussed by
Zepp and Cline (1977) is
where [P] is the hexacyanoferrate (II) or (ill) concentration in moles/liter
and ka, is the specific sunlight absorption rate as a function of wave-
length. The ka, term is equal to the wavelength specific average rate of
light absorption per unit volume (la>) times the molar extinction coeffi-
cient of the iron-cyanide (e^) divided by the attenuation coefficient of the
water body ("A) times a constant (i, 6.02 x 10 ) which converts the
intensity expressed in photons cm s into units compatible with [P]
(Zepp and Cline, 1977). Assuming that the quantum yield for the photolysis
reaction in solution is not wavelength dependent then the rate expression is
where kfl equals the sum of the ka» values integrated over all wavelengths
of sunlight absorbed by the reactants. This expression conforms to a first-
order rate equation in which the photolysis rate constant (<{>ka) is expressed
in units of reciprocal time. The value of ka changes during the day because
the surface spectral flux distribution of solar radiation is a function of the
solar zenith angle. For natural waters ka is also a function of competitive
light absorption by the water itself along with other absorbing materials and
light scattering. Because in most natural waters attenuation due to light
scattering is less important than that due to absorption, scattering was
initially ignored in the following computations. Surface reflection and
scattering from water bodies was also ignored because it is small (about 5 to
6 percent) for most solar zenith angles (Hutchinson, 1957).
For a first-order reaction the half-life (tu), or time required for one-
half of a compound to react, is independent of the initial concentration and
can be expressed by
ti = —T~, or t. =
where t is the time of exposure, and PQ is the initial iron-cyanide concen-
tration expressed as HCN with that after exposure (P) equal to PO minus the
determined HCN concentration. The half-lives represent the period of time
required to decompose the iron-cyanides with release of half the HCN that can
be potentially formed from the specific complexes. The photolysis rate
constant (ka) can be calculated from the expression ln(Po/P)/t.
A computer program written in Fortran IV, that uses the above mentioned
theory and equations to compute direct photolysis rates, is available on
request from the Environmental Research Laboratory-Athens, Georgia, U.S.
Environmental Protection Agency (Zepp and Cline, 1977).
11
-------
SECTION 5
RESULTS AND DISCUSSION
VOLATILIZATION OF HCN
The volatilization of HCN from 8 1 sodium cyanide solutions prepared with
deionized water and placed in 2 gal acid battery jars was determined in the
laboratory as a function of temperature and concentration. This rather slow
loss of cyanide, determined from 6 hr exposure periods, is apparently not
independent of the initial concentration as indicated by the variation in the
first-order rate constants in Table 1. This correlation between rate constant
for HCN loss and initial concentration implies that the evolution of HCN from
an open aqueous solution does not follow the first-order kinetics "as antici-
pated. However, from the limited number of tests conducted, it is concluded
that the loss is approximately first-order, especially when one considers the
possible analytical error introduced into the determinations for solutions
with the lowest initial cyanide concentrations.
At a given initial cyanide concentration a decrease in test temperature
resulted in higher residual cyanide concentrations. This is indicated by a
reduction with decreasing temperature in the rate constant for HCN loss to the
atmosphere from the surface of comparable solutions. Over the temperature
range 10 to 25° C, this decreased rate is approximately a linear function of
temperature such that for a 10° C decrease the rate constant is reduced by
about 55%. Experiments conducted with solutions placed in jars on the roof of
our laboratory demonstrated that the rate of HCN loss was about twice that
determined in the laboratory. This increase is most likely a result of
increased loss due to wind action, since it was demonstrated that HCN was not
subjected to photolysis by natural light. Also, solutions placed on the roof
and loosely covered with a sheet of glass showed approximately the same rate
of HCN loss when compared with similar solutions in the laboratory.
Since the loss of HCN from a solution into the atmosphere is through the
phase borderline, the rate of cyanide decrease is a function of the surface
area to solution volume ratio. For our containers with perpendicular walls
the HCN removal rate is inversely proportional to the solution depth. Thus
the time required for the removal of HCN to 1/2 its original concentration is
directly proportional to the solution depth, since the same type of container
and solution volume was used in all HCN rate loss experiments, temperature and
initial concentration are the only variables for which a comparison of HCN
removal can be made.
The total loss of cyanide from natural waters occurs not only by vola-
tilization into the atmosphere but also through chemical reactions and
biological oxidation to ammonia and C02« The above calculated rates
represent only minimal losses since the oxidative loss is negligible in
12
-------
TABLE 1. RATE OF HCN LOSS FROM SODIUM CYANIDE SOLUTIONS PREPARED WITH
DEIONIZED WATER AND IN OPEN CYLINDRICAL JARS UNDER LABORATORY CONDITIONS*
Initial free
cyanide cone.
(ug/1 CN)
25
50
100
200
25
50
100
200
10
Rate Constant
0.00624
0.0103
0.0136
0.0149
Half Life,
111.0
67.3
50.8
46.5
Temperature
15
, kChour'1)
0.0139
0.0117
0.0135
0.0173
tu (hour)
49.7
59.4
51.5
40.1
(°c)
20
0.0194
0.0225
0.0215
0.0254
35.8
30.8
32.2
27.3
25
0.0257
0.0263
0.0299
0.0315
27.0
26.3
23.2
22.0
*Test solution pH was about 7.9.
13
-------
deionized water solutions as well as the loss through agitation since the
solutions were unstirred.
Lure and Panova (1964) determined that for river water test solutions
initially containing 10 to 25 mg/1 CN and exposed in open vessels, the
concentration of cyanide during standing declines after 7 days to about 10% of
the original with the complete disappearance of cyanide occurring in 10-12
days. These authors concluded that in the natural water they investigated,
the loss of cyanide by volatilization is the most significant of the means of
HCN decline.
The rate of HCN loss has been demonstrated to depend on temperature but
Palaty and Horokova-Jakubu (1959) observed that the intensity of agitation and
ratio of the solution volume to surface area are also very important factors.
In fact, they determined the removal rate in vigorously air-agitated solutions
to be one order of magnitude greater (12 to 14 times) than the nonagitated
solutions. However, even though the decomposition of simple cyanides
dissolved in natural and deionized water occurs differently, it is likely that
the rapid loss of HCN under the most favorable circumstances is still consid-
erably slower than the formation of HCN from photolysis of hexacyanoferrate
(II) and (III) complexes for midday and near surface conditions.
FREE CYANIDE AND HCN DETERMINATIONS
It has been assumed by previous workers (Burdick and Lipschuetz, 1950)
that the pyridine-pyrazolone method for the measurement of free cyanide does
not upset the chemical equilibria by liberating cyanide from the hexacyano-
ferrate complex ions or intermediate photolysis forms. Comparison of the
direct vapor phase method and the indirect colorimetric methods for the
determination of HCN (Broderius, 1973) demonstrated that the indirect pyridine-
pyrazolone method gives a relatively high estimate of free cyanide in
iron-cyanide solutions. This may be due to some cyanide which is liberated
from the hexacyanoferrate complex ions or intermediate photolysis forms by the
pyridine-pyrazolone method. In five tests with each complex, however, the
ratios between HCN concentrations determined by the direct vapor phase
equilibration procedure and those estimated by the indirect colorimetric method
were relatively constant, regardless of the degree of decomposition, at about
84.6% and 64.2% for the hexacyanoferrate (II) and (ill) solutions, respectively
(Table 2). Therefore, the indirect colorimetric method for the estimation of
HCN concentrations can be used when the relative (normalized) photolysis rates
of hexacyanoferrate (II) and (ill) solutions exposed to natural light are
determined. When the proper correction is made the results from the indirect
colorimetric method can also be used to determine the actual photolysis rate of
these complexes as indicated by HCN formation.
KINETICS OF PHOTODECOMPOSITION
Molar Extinction Coefficients
The electronic absorption spectra of freshly prepared and dilute potassium
Fe(CN)g and Fe(CN)g aqueous solutions were measured in the wavelength
14
-------
TABLE 2. THE HCN CONCENTRATION ESTIMATED BY THE INDIRECT COLORIMETRIC
METHOD AND THAT DETERMINED BY THE DIRECT VAPOR PHASE EQUILIBRATION
PROCEDURE FOR HEXACYANOFERRATE (II) AND (III) SOLUTIONS EXPOSED TO
SUNLIGHT AND INITIALLY CONTAINING 100 ng/1 CN*
Test
1
2
3
4
5
1
2
3
4
5
Determined HCN concentration, yg/1
A B Percentage
Indirect Direct vapor phase B of
colorimetric equilibration initial total Ratio
method method cyanide B/A
Ferrocyanide solutions
33.9 29.9
38.8 31.4
60.4 51.2
83.1 68.9
98.2 84.8
Ferricyanide solutions
25.7 17.2
45.2 27.7
47.2 31.7
60.9 37.5
64.8 41.4
29.9 88.2
31.4 80.9
51.2 84.8
68.9 82.9
84.8 86.4
mean 84 . 6
^2.9
17.2 66.9
27.7 61.3
31.7 67.2
37.5 61.6
41.4 63.9
mean 64.2
*Test temperature was 25° C and pH 7.9.
15
-------
region close to the strong ultraviolet absorption bands. As seen in Figure 1,
these complexes are photosensitive at wavelengths <420 nm for the ferro and
<480 nm for the ferri ions. Therefore, photolysis rates of these complexes
would be expected to reflect fluctuations in intensity of the shorter
wavelength component of sunlight. The molar extinction coefficients (e)
(Table 3) derived from the absorption spectra of freshly prepared aqueous
solutions of potassium Fe(CN), and Fe(CN)^ (approximately
10 M) agreed closely with those reported by Ibers and Davidson (1951),
Adamson (1952), and Asperger (1952). A possible reaction product,
Fe(CN)5*(H20)3~, has also been reported to absorb photoactive
spectral light (Asperger, 1952). However, this product is a transient
intermediate and since absorption spectra of reaction solutions were quite
similar in shape and reductions in these spectra were proportional to the
degree of decomposition, it was assumed that the predominant cyanide
components contributing to the total absorption were the hexacyanoferrate (II)
and (III) ions.
Specific Sunlight Absorption Rate
The rate of a photochemical reaction in an aqueous solution'is dependent
upon the solar spectral irradiance at the solution surface, radiative transfer
from air into the solution, and the transmission of sunlight in the solution.
Through absorption by components in the atmosphere, the intensity of sunlight
is decreased such that the ultraviolet and visible region decreases with
decreasing wavelength so that essentially no light is transmitted to the
earth's surface at wavelengths <295 nm (Bener, 1972). Molar extinction
coefficients of hexacyanoferrate (II) and (III) ions can be coupled with
actinic irradiance data of Leighton (1961) and Bener (1972) to calculate the
specific rate of sunlight absorption (kfl) (Zepp and Cline, 1977). The
magnitude of these rates depends upon the degree of spectral overlap between
the electronic absorption spectra and the spectrum of sunlight at the earth's
surface. The specific sunlight absorption rates (ka*) of hexacyanoferrate
(II) and (III) complexes were computed as a function of wavelength for shallow
depths and apply to midsummer and midday at latitude 40° N. Only wavelengths
of less than about 480 nm are photometrically important (Figure 2). The
maximum interaction with solar radiation occurs at about 330 nm for ferro-
cyanide solutions and at 330 nm and 420 nm for ferricyanide solutions. The
ratio of integrated ka, values (ka) for the wavelength region of 297.5 -
490 nm indicates that the sunlight absorption rate constant for ferricyanide
is about 17 times larger than that for ferrocyanide (Figure 2).
Properties of Photochemical Reaction
The photolysis of hexacyanoferrate (II) and (III) complexes in solutions
of certain initial total cyanide concentrations follows first-order kinetics
as indicated by typical linear regression correlation coefficients in Table 4.
These values are for the relationships between midday exposure period and
calculated log iron-cyanrde concentration for solutions with initial complex
concentrations up to 200 ug/1 as total cyanide. It has been proposed that the
reaction is reversible in darkness (Asperger et^ _a_l., 1960). However, deter-
minations of the amount of released cyanide that is recombined or lost showed
that the reaction is irreversible in the dark for ferricyanide solutions, but
16
-------
o
OC
UJ
Q.
UJ
U
CO
a:
o
}
00
4-
Fe(CN)
I.9XIO~3M
280 3OO 320 34O 36O 380 4OO 420 440 460 480 500
WAVELENGTH (nm)
Figure 1. Electronic absorption spectra of hexacyanoferrate (II) and (III)
solutions.
17
-------
TABLE 3. MOLAR EXTINCTION COEFFICIENTS OF HEXACYANOFERRATE (II) AND
(III) SOLUTIONS
Wavelength
(nm)
297.5
300.0
302.5
305.0
307.5
310.0
312.5
315.0
317.5
320.0
323.1
330.0
340.0
350.0
360.0
370.0
380.0
390.0
400.0
410.0
420.0
430.0
440.0
450.0
460.0
470.0
480.0
490.0
Extinction coefficient,
Fe(CN){T
483.7
428.5
387.9
363.2
346.9
335.3
330.6
329.0
328.5
328.5
326.4
309.5
252.7
174.8
105.3
55.8
25.8
11.0
5.3
2.6
1.6
1.6
1.0
1.0
0.5
0.5
0.0
0.0
liter/mole- cm
Fe(CN)^~
1589
1673
1710
1677
1575
1410
1279
1215
1193
1195
1175
938
571
334
336
434
577
770
945
1007
1053
932
597
261
80
18
4
0
18
-------
10
-s
10'
-5
10
-6
I//
_J I 1_
375 300 335 350 375 40O 425 450 475 500
WAVELENGTH (nm)
Figure 2. Specific sunlight absorption rates of hexacyanoferrate (II) and
(III) complexes as a function of wavelength at midday and mid-
summer, latitude 40° N.
19
-------
TABLE 4. THE LINEAR RELATIONSHIP BETWEEN EXPOSURE PERIOD (X) NEAR MIDDAY FOR
DIFFERENT METEOROLOGICAL CONDITIONS AND CALCULATED LOG IRON-CYANIDE
CONCENTRATION (Y), FOR SOLUTIONS WITH INITIAL COMPLEX CONCENTRATIONS UP
TO 200 ug/1 AS TOTAL CYANIDE, AS INDICATED BY THE REGRESSION CORRELATION
COEFFICIENT
Total Exposure
cyanide, period,
iig/1 CN min
25 0
15
45
75
50 0
15
45
75
100 0
15
45
75
200 0
15
45
75
25 0
15
45
75
135
50 0
15
45
75
135
100 0
15
45
75
135
200 0
15
45
75
135
Determined Calculated
HCN iron-cyanide
concentration, concentration as
ug/1 CN ug/1 CN
0
4.95
13.0
15.1
0
11.2
27.6
36.2
0
36.8
69.4
82.7
0
68.4
128.8
156.3
Fe(CN)|~
1.87
5.87
7.96
10.8
0
4.46
11.5
16.2
22.3
0
8.22
21.6
31.5
43.9
0
28.9
63.7
82.0
94.9
25.00
20.05
12.0
9.9
50.0
38.8
22.4
13.8
100.0
63.2
30.6
17.3
200.0
131.6
71.2
43.7
25.00
23.13
19.13
17.04
14.2
50.00
45.54
38.5
33.8
27.7
100.00
91.78
78.4
68.5
56.1
200.00
171.1
136.3
118.0
105.1
Linear
correlation
coefficient,
r
0.962
0.999
0.992
0.990
0.972
0.981
0.986
0.897
20
-------
is slightly reversible though incomplete for ferrocyanide solutions that have
undergone only minimal decomposition. After prolonged exposure the reaction
becomes irreversible, even for ferrocyanide solutions.
The photolysis reactions for both iron-cyanide complex solutions
initially containing 100 ug/1 CN were virtually unaffected by dissolved oxygen
in the range 2-8 mg/1. There was a moderate effect of pH on the photolysis
reactions with the rates increasing with a decrease of test pH over the range
9.0 - 6.6. The relative rates normalized to that determined at pH 6.6
decreased to about 0.84 and 0.70 at pH 9.0 for the ferri- and ferrocyanide
solutions, respectively. The decrease in relative rate with increase in pH is
linear over the pH range tested for ferricyanide solutions but decreases in a
linear manner only from pH 6.6 to about 8.0 and then remains fairly constant
to pH 9.0 for ferrocyanide solutions.
The photochemical reaction rate for iron-cyanide solutions initially
containing 100 pg/1 CN and at pH 7.8 was measured at 5.6, 12.1, and 23.5° C.
The slight negative temperature relationship between the relative photolysis
rate (Y) normalized to the rate determined at 23.5° C, as a function of test
temperature in ° C (X), can be represented by
Log Y = -0.104 + 0.00464 X (r = 0.962)
and Log Y = -0.136 + 0.00574 X (r = 0.999)
for hexacyanoferrate (ll) and (ill) solutions, respectively. Balzani and
Carassiti (1970) stated that this observed temperature dependence may be due
to secondary thermal reactions which contribute to the overall quantum yield.
Varying the initial concentration of iron-cyanide salt affected the
photodecomposition rate constant. The dependence of this constant on concen-
tration demonstrates, according to Asperger (1952), that the decomposition is
due not only to the light energy, but also to collision of molecules.
Therefore, photolysis of the hexacyanoferrate (II) and (ill) complexes is not
truly first-order. For practical purposes, however, we can assume that the
reactions follow first-order kinetics for complex concentrations that are
likely to be found in natural waters. The relative photolysis rate constants
normalized to those determined for solutions at 20° C with an initial total
cyanide concentration of 25 Mg/1 CN and as a function of the average iron-
cyanide concentration during the rate determination period are represented in
Figure 3. The pH of all solutions was about 7.8. The rate constants of the
decomposition reaction decrease with increasing initial cyanide complex
concentration up to some apparent limiting concentration. The relative
photolysis rate constant normalized to that determined for 25 pg/1 total
cyanide (Y) was a function of average hexacyanoferrate (II) concentration (X)
up to 1000 Ug/1 CN and can be represented by the empirical expression, Y =
2.099 x~0-222" (r = -0.954). For hexacyanoferrate (ill) solutions the
above relationship can be represented by Y = 1.390 x~°-1538 (r = -0.904).
Photolysis rates were determined for comparable hexacyanoferrate (II) and
(ill) solutions prepared with deionized water and exposed to the same light
conditions. The ratio of these rates for ferro- to ferricyanide solutions
over the initial total cyanide concentration range of 25 - 200 yg/1 CN
averaged 2.05 _+ 0.32 for 32 such comparisons. However, the rate of change in
photolysis rate constants (Figure 3) is such that at initial total cyanide
concentrations of approximately 2.0 mg/1 CN and greater the decomposition rate
21
-------
o
o
UJ
I
CO
CO
1.0
0.9
0.8
O.7
0.6
0.5
O.4
0.3
0.2
O.I
0
Y-2.099 X
r-0.954
Fe(CN)*"
-0.2220
100
200
300
400
500
600
700
800
9OO
Q,
UJ
I
_l
UJ
1.0
O.9
O.S
O.7
O.6
0.5
0.4
0.3
O.2
O.I
0
Y-1.390 X
r« 0.904
-0.1538
Fe(CN)
3-
IOO 200 300 400 500 600 700 800
AVERAGE IRON-CYANIDE CONCENTRATION (fJG/L CN)
9OO
Figure 3. Relative photolysis rate constants normalized to the values
determined for solutions with an initial total cyanide concentration
of 25 ng/1 CN and as a function of the average iron-cyanide
concentration during the rate determination period.
22
-------
for ferricyanide solutions is faster than for comparable ferrocyanide
solutions.
The maximum amount of total cyanide that could be photochemically
released as HCN from prolonged exposure of dilute hexacyanoferrate (II) and
(III) solutions was determined to be about 85% and 49%, respectively. This
indicates that for every mole of iron (II) and (ill) complex, each containing
6 moles CN, only 5 and 3 moles of CN, respectively, can be released as free
cyanide from the complex anions through a photolysis reaction. These results
are not consistent with the reaction pathways for the photodecomposition of
hexacyanoferrate complexes as proposed in the introduction (equations 1 and
4). This supports the contention of Balzani and Carassiti (1970) that the
overall chemical changes and the reaction mechanisms for the photolysis of
hexacyanoferrate complexes is not well defined.
Quantum Yield
Minimum direct photolysis rates of hexacyanoferrate (II) and (ill)
complex solutions prepared with deionized water were determined empirically at
near-surface depths and midday for different times of the year under full
sunlight at St. Paul, Minn. The experimental midday half-lives for
hexacyanoferrate (II) solutions initially containing 100 yg/1 CN at pH 7.8 and
20° C ranged from about 50 min in late fall to a minimum of about 18 min in
midsummer (Figure 4). For comparable hexacyanoferrate (III) solutions the
midday half-lives ranged from about 160 min in late fall to a minimum of about
64 min in midsummer (Figure 5).
The quantum yields (4>) for the photodecomposition of the iron-cyanide
complexes were estimated by a visual best fit analysis of observed date-
dependent midday half-lives, determined from the change in HCN concentration
(eq. 8, p. 11), to the theoretical curves calculated for certain $ values and
specific sunlight absorption rates (kfl) of the reactants as determined from
the Zepp and Cline computer program. From our investigation it was experi-
mentally determined that the iron-cyanide disappearance quantum yield as
indicated by HCN formation for the hexacyanoferrate (II) and (III) complexes
are approximately 0.14 and 0.0023, respectively (Figures 4 and 5). This
calculation assumes that the quantum yields are wavelength independent in the
region of sunlight absorption. Our experimentally determined quantum yields
are in reasonable agreement with those reported in the review by Balzani and
Carassiti (1970) (see p. 6).
Time of Year and Latitude
The intensity and spectral distribution of sunlight on a horizontal
surface generally decreases with decreasing angular height of the sun.
Therefore, intensity decreases from midday to^sunset, from summer to winter,
and from the tropics to higher latitudes. Midday half-lives for direct
photolysis of hexacyanoferrate (ll) and (III) complexes near the surface of an
aqueous solution and as a function of time of the year and latitude are shown
in Figure 6. The results were computed as relative values with the half-life
of each complex on July 1 at latitude 45° N assigned a value of unity. For
both complexes the photolysis half-lives at the midlatitudes are predicted to
23
-------
CO
UJ
UJ
u,
_l
I
u.
_J
o
Q
JFMAMJJASONDJ
TIME OF YEAR
-40
30
-20
- 10
Figure 4. Midday half-lives for 100 yg/1 CN hexacyanoferrate (II) solutions
at near surface depths for different times of the year under full
sunlight conditions at St. Paul, Minn.
24
-------
tu
=}
UJ
<
X
o
Q
zoo
180
I6O -
I4O-
200
IOO •
JFMAMJJASONDJ
TIME OF YEAR
Figure 5. Midday half-lives for 100 ug/1 CN hexacyanoferrate (ill) solutions
at near surface depths for different times of the year under full
sunlight conditions at St. Paul, Minn.
25
-------
o
Q
UJ
<
X
UJ
CE
-0.8
JFMAMJ JA
0 N
TIME OF YEAR
Figure 6. Midday half-lives for direct photolysis of pure water hexacyano-
ferrate (II) and (III) solutions (near surface) as a function of
the time of year for several northern latitudes. Values are
relative to July 1 rate at 45° N latitude.
26
-------
be minimal during the summer and maximum during the winter months. Both the
half-lives and the amplitude of the time of year variation increase with
increasing northern latitude. In the tropical zone photolysis rates should be
relatively constant throughout the year. Variations in rates during the
summer are also expected to be minimal with less than a 1.3-fold increase from
the equator to latitude 60° N. The relative half-lives for both iron-cyanide
complexes are quite similar at all latitudes and times of the year. During
the ice-free months, the variation in relative midday half-lives is expected
to be less than a value of 3 on sunny days and for latitudes to 60° N.
Diurnal Change
The direct photolysis rate of iron-cyanide complexes changes diurnally as
the intensity of sunlight increases and then decreases throughout the day,
with maximum rates occurring at midday. This variation in computed rates for
shallow depths and relative to photolysis rates at midday on July 1 at
latitude 45° N, longitude 93.2° W is illustrated for the first of various
months in Figures 7 and 8. Experimentally determined photolysis rates
relative to midday values and as determined at different times of the day for
pure water iron-cyanide solutions were close to the theoretically computed
values as indicated in Figures 9 and 10.
Attenuation by Natural Waters
Light falling directly on the surface of a water body is both reflected
at an angle equal to the angle of incidence and penetrates with a change in
direction due to refraction. The fraction of direct sunlight and sky
radiation that is reflected is small when compared with that which penetrates
(Wetzel, 1975). The reflected light is essentially the same in spectral
composition as incident light but the spectrum of light penetrating the
surface of the water is altered. The intensity of incident collimated sun-
light that penetrates natural waters, when compared with that of pure water,
is attenuated through absorption and scattering. The reduction in trans-
mission of light and shift in absorption selectivity depends on the
wavelengths of the incident light and the depth to which the light has
penetrated. In inland surface waters absorption is due mainly to dissolved
metallic io<.s and natural organics. Attenuation coefficients (e^) were
measured for different water bodies with a Beckman DB-GT spectrophotometer and
are presented in Appendix A. In all cases the attentuation of light was wave-
length dependent and varied considerably from one water body to another
(Figure 11). This is especially true for the ultraviolet region where
attenuation of light intensity increases with decreasing wavelength.
Both suspended and dissolved materials within a water body induce
variations in the depth of light penetration. Increasing turbidity decreases
transparency and shifts maximum wavelength transmission to longer wavelengths.
Substances in a solution that absorb or scatter light should diminish the
photodecomposition rate of iron-cyanides by reducing the amount of light
available. This effect was demonstrated by using turbid suspensions prepared
with deionized water and bentonite colloidal clay or wind-blown silt from
melted snow. The experiments were conducted with an apparatus consisting of a
Pyrex glass filter containing 40 mm of the suspension and horizontally
27
-------
I.Or
8 10 12 2 4
TIME OF DAY (C.S.T.)
Figure 7. Diurnal variation of direct photolysis rates of pure water hexa-
cyanoferrate (II) solutions (near surface) relative to photolysis
rates at midday on July 1 at latitude 45° N, longitude 93.2° W.
28
-------
8 10 12 2 4
TIME OF DAY (C.S.T.)
Figure 8. Diurnal variation of direct photolysis rates of pure water hexa-
cyanoferrate (III) solutions (near surface) relative to photolysis
rates at midday on July 1 at latitude 45° N, longitude 93.2° W.
29
-------
LU
Sc
K
>
5
111
en
I.Or
0.9-
0.8
0.7 •
0.6
0.5
0.4
O.2 -
8 10 12 2
TIME OF DAY (C.S.T.)
Figure 9. Time of day-dependence of direct photolysis rates of pure water
hexacyanoferrate (II) solutions (near surface) relative to
photolysis rates for midday at St. Paul, Minn, on October 20, 1977,
Theoretical relationship indicated by smooth line.
30
-------
i.o
0.9
0.8
0.7
0.6
Ul 0.5
UJ
tr
0.4
0.3
0.2H
O.I
FC (CN)
10
12
TIME OF DAY (C.S.T.)
Figure 10. Time of day dependence of direct photolysis rates of pure water
hexacyanoferrate (III) solutions (near surface) relative to
photolysis rates for midday at St. Paul, Minn, on October 21, 1977,
Theoretical relationship indicated by smooth line.
31
-------
CO
NJ
E
u
o
UJ
o
u
.eoor
.900
.400-
.300
.200
.100
MISSISSIPPI
V \ RIVER
EL£PHA
CAK
LAKE PHALEN
SQUARE LAKE
300 320 340
360 360 400 420 440
460
480
WAVELENGTH (nm)
Figure 11. Attenuation coefficients relative to deipnized water for natural water samples
collected in north-central United States.
-------
positioned over an open cylindrical jar which contained the iron-cyanide
solutions at pH 7.8 and 20° C. In this manner, the suspensions were not in
direct contact with the iron-cyanide solutions and all of the light reaching
an iron-cyanide solution had to first pass through a filter. During the short
exposure periods a negligible amount of suspended material settled out on the
bottom of the filters.
Relative photolysis rates were defined as those normalized to rates
determined for iron-cyanide solutions for which the overlying Pyrex filters
contained 40 mm of deionized water. The ratios of the experimentally
determined and theoretically computed relative photolysis rates of hexacyano-
ferrate (II) and (ill) solutions as affected by suspensions of benttmite clay
and wind-blown silt are presented in Table 5. The relationships between the
ratios expressed as logarithms and turbidity were determined, as represented
in Figure 12, to be linear at least to 40 NTU. These lines can be charac-
terized for the bentonite solutions by a slope function (g) of 0.0618 and
0.0502 for the hexacyanoferrate (II) and (ill) solutions, respectively. The
slope for the relationship determined when light had to pass through wind-
blown silt solutions before reaching hexacyanoferrate (ll) solutions is
0.0456.
The total attenuation of radiance passing through a medium is due to
absorption and to a redirection or scattering of some of the beams radiance.
Determination of the forward scattering function requires artificial light
which yields a collimated beam (Tyler and Preisendorfer, 1962). Therefore,
this parameter could not be measured from our data. However, from Figure 12
it is apparent that the photolysis reaction in iron-cyanide solutions is
enhanced by suspended materials when compared with that theoretically
calculated from beam attenuation coefficients. In fact, one might expect up
to a threefold to fourfold increase in photolysis rates when compared with
those calculated for iron-cyanide solutions in natural waters with a turbidity
of about 10 NTU. The photolysis reaction for ferrocyanide solutions is
affected to a greater degree by the presence of bentonite than for
ferricyanide solutions. The presence of wind—blown silt had less of an
enhancement effect than that of bentonite on the photodecomposition of
hexacyanoferrate (II) solutions (Figure 12). This may be due to a reduction
in the forward scattering of light by silt solutions when compared with those
prepared with bentonite.
The decrease in photolysis rate with increasing depth is dependent upon
the relative magnitude of the attenuation coefficients of the water body, the
molar extinction coefficients of the iron-cyanide complex, and the intensity
of sunlight, all as a function of spectral wavelength. In general, the
shorter wavelengths are more readily removed by the surface layer of most
natural waters. Thus, the photochemically active light which decomposes iron-
cyanides is expected to penetrate only a short distance in natural waters.
Field experiments on the photolysis of iron-cyanide solutions in natural
waters were conducted by suspending test tubes containing cyanide solutions
prepared with pH 7.8 phosphate buffered deionized water for a specific time
period at known depths in a water column. Using beam attenuation coefficients
and assuming photolysis at a specific depth, the depth-dependence of the
33
-------
TABLE 5. . EXPERIMENTALLY DETERMINED AND THEORETICALLY COMPUTED RELATIVE
PHOTOLYSIS RATES FOR HEXACYANOFERRATE (II) AND (ill) SOLUTIONS AS
NORMALIZED TO DEIONIZED WATER CONTROLS AND AFFECTED BY BENTONITE AND
WIND-BLOWN SILT. COMPUTED RATES ARE BASED ON BEAM ATTENUATION
COEFFICIENT MEASUREMENTS
Test
solution and
concentrat ion ,
(mg/1)
Bentonite
Bentonite
Bentonite
Bentonite
Silt
Silt
Silt
Bentonite
Bentonite
Bentonite
Bentonite
Bentonite
Bentonite
Bentonite
75
250
600
750
—
-
-
75
100
200
250
400
500
600
Relative photolysis rate
normalized to controls
Turbidity
NTU
6.7
24.0
57.0
68.0
23.0
43.0
63.0
5.6
9.9
18.5
23.0
36.4
46.0
54.0
A
Determined
FERROCYANIDE
0.916
0.632
0.326
0.260
0.403
0.186
0.0824
FERRICYANIDE
0.958
0.960
0.805
0.872
0.528
0.603
0.421
B
Computed
0.365
0.0155
0.000147
0.0000264
0.0431
0.00197
0.000107
0.471
0.240
0.0906
0.0405
0.00936
0.00244
0.00111
Ratio of
A to B
2.51
40.8
2218
9848
9.35
94.4
770.1
2.03
4.00
8.88
21.5
56.4
247.1
379.3
34
-------
«5
QC
UJ
UJ
QC
O
UJ
ID
O
<
O
O
UJ
a:
UJ
h-
UJ
O
= 0.0456
4-
• Fe(CN)K BENTONITE
Fe(CN)g" SILT
Fe(CN)g BENTONITE
10 20 30 40 50
TURBIDITY (NTU)
60
70
Figure 12. Influence of turbidity on the experimentally determined to
theoretically calculated relative photolysis rate of hexacyano-
ferrate (II) and (III) solutions.
35
-------
direct photolysis of hexacyanoferrate (II) and (ill) solutions relative to
near-surface rates were computed with the Zepp and Cline (1977) computer
program for latitude 45° N, longitude 93.2° W. The photolysis rates
normalized to that at the surface decreased exponentially with depth. The
linear regression lines defining these relationships were calculated by
forcing the line through the log relative rate of 1.0 at depth 0 cm. The
theoretical and determined rates of decrease in the depth-dependent photolysis
of hexacyanoferrate (II) and (ill) solutions are presented in Table 6. The
regression lines apply to relative photolysis rates as low as one-hundredth
of that determined at the surface. The photolysis rate for hexacyanoferrate
(II) solutions decreases more rapidly with increasing depth than the rate for
hexacyanoferrate (III) solutions. The rates differ because the predominant
wavelengths affecting photolysis of the iron (II) complex are shorter and thus
penetrate less into natural waters than those activating the iron (III)
complex.
The ratio of the theoretical to determined rate of decrease in photolysis
rates varied from near 1.0 for Square Lake and bog waters of low turbidity to
3.52 for the turbid Minnesota River. The ratios are apparently a function of
turbidity or suspended solids,-since for Square Lake and bog water the
observed and theoretical rates of decrease were almost identical. The
theoretical computations assume that the natural materials in a water body act
only as photochemically inert sun screens. Reactions by "sensitizers" and the
bouncing of light or light scattering in turbid solutions which may increase
photolysis over that predicted is ignored. However, it was observed that in
the more turbid waters the photolysis reaction was relatively enhanced with
depth because of light scattering and thus the rate of decrease in photolysis
rate with depth is less than theoretically predicted by the Zepp and Cline
model. These results confirm the relationship between turbidity and
photolysis rate as presented in Figure 12.
By using the attenuation coefficients- in Appendix A and assuming complete
mixing of the water column of interest, the depth-dependence of the average
direct photolysis rates for hexacyanoferrate (ll) solutions at midday and
midsummer relative to near-surface rates were calculated for latitude 40° N by
the computer program of Zepp and Cline (1977). These relationships are shown
in Figure 13. From the results in Table 6 for photolysis rates at a point for
various depths in a water column, it is proposed that the theoretical
relationships actually overestimate the rate at which the relative photolysis
rate decreases when mixing is assumed. For waters like the Minnesota River,
the rate of decrease may in fact be overestimated by a factor of about 3.5
(Table 6). For the natural waters tested the average observed decrease is
about one-half as rapid as that theoretically predicted in Figure 13. If this
factor is applied to the computed depth-dependence curves where mixing of the
water column was assumed (Figure 13), it is proposed that below a depth of
about 50 - 100 cm photolysis of hexacyanoferrate (II) is insignificant from a
toxicological standpoint in all tested natural waters. A relationship for the
hexacyanoferrate (ill) complex was not included since the concentration of HCN
formed by photolysis of this complex in solutions prepared with natural waters
was approximately the same as that produced in comparable hexacyanoferrate
(ll) solutions.
36
-------
TABLE 6. PHYSICAL PROPERTIES OF VARIOUS NATURAL WATERS AND THE THEORETICAL AND DETERMINED RATE OF
DECREASE IN THE DEPTH-DEPENDENT DIRECT PHOTOLYSIS RATE OF HEXACYANOFERRATE (II) AND (ill)
SOLUTIONS PREPARED WITH DEIONIZED WATER AND EXPOSED TO NATURAL LIGHT AT SPECIFIC FIXED
DEPTHS AT LATITUDE 45° N, LONGITUDE 93.2° W
Secchi
disk
Water measurement,
body cm
Square Lake
Lake Phalen 202
Lake Como 79
Elephant ISO
Lake
Mississippi -
River
Minnesota 45
River
Bog water
Square Lake
Lake Phalen 202
Lake Como 79
Elephant 180
Lake
Mississippi
River
Minnesota 45
River
Bog water
Residue, mg/1 Linear regression slope values*
Turbidity,
NTU
0.43
2.7
5.4
2.9
3.6
14.0
0.9
0.43
2.7
5.4
2.9
3.6
14.0
0.9
Total Total non A
filterable filterable Total theoretical
FERROCYANIDE
-0.0118
223 3.0 226 -0.0364
190.5 11.0 201.5 -0.0884
74 <1 74 -0.0886
-0.130
495 24.5 519.5 -0.206
-0.290
FERRIC* ANIDE
-0,00516
223 "3.0 226 -0.0201
190.5 11.0 201.5 -0.0590
74 <1 74 -0.0338
-0.0939
495 24.5 519.5 -0.121
-0.112
B
determined
-0.0107
-0.0190
-0.0392
-0.0453
-0.0850
-0.0586
-0.260
-0.00482
-0.00956
-0.0314
-0.0226
-0.0461
-0.0444
-0.120
R2
0.984
0.990
0.998
0.927
0.947
0.939
0.978
0.956
0.925
0.998
0.972
0.976
0.996
0.982
Ratio of
A to B
1.10
1.92
2.26
1.96
1.53
3.52
1.12
1.07
2.10
1.88
1.50
2.04
2.73
0.93
*Linear regression analysis for log relative photolysis rates (Y) normalized to those determined for comparable solutions at the surface
and depth (X) in era. Regression lines are forced through the log relative photolysis rate of 1.0 at depth of 0 cm. Log Y = gX with g =
I (Steel and Torrie, p. 179, 1960).
-------
A - PURE WATER
B - SQUARE LAKE
C - LAKE PHALEN
0 - LAKE COMO
E-ELEPHANT LAKE
F - MISSISSIPPI RIVER
G - MINNESOTA RIVER
H - BOG WATER
DEPTH (CM)
Figure 13. Calculated depth-dependence of the direct photolysis of hexa-
cyanoferrate (II) at midday and midsummer for latitude 40° N
when using beam attenuation coefficients and assuming complete
mixing of the water column.
38
-------
In making the depth-dependent computations, it was assumed that the iron-
cyanide is isotropically distributed and all that is in a water layer is
exposed to the same amount of light during a given time period. This
assumption has been demonstrated according to Zepp and Cline (1977) to be
valid at depths in which a small fraction of the incident light is absorbed.
However, if at a given depth almost all of the light is absorbed in the upper
part of the water column, the assumptions are valid only if mixing is more
rapid than entry of iron-cyanide into or loss from the upper layer of the
water body. If entry to the upper layer is more rapid than mixing, then the
concentration of iron-cyanide and thus HCN will be higher than predicted near
the surface. If a situation exists in which the iron-cyanide is initially
uniformly distributed in the water column but mixing is incomplete, then it
would be expected that the photolysis rate would be slower with increasing
depth than is indicated in Figure 13. The above calculations apply to near
midday and summer situations. As the sun moves lower in the sky as a function
of the time of day or latitude, the underwater path length of direct sunlight
is longer and depth dependence increases.
Sky Conditions and Photolysis Rate
The daytime radiation received at any point on the earth's surface
consists of direct solar radiation or sunlight, and indirect solar radiation
or scattered light of the sky. The spectral distribution of irradiance
reaching the earth's surface depends on the sun's altitude and the
meteorological conditions of the intervening atmosphere. The relative
spectral composition of the energy distribution of combining solar and sky
radiation between 315 and 800 nm is essentially constant during the day
(Robinson, 1966). However, a reduction in atmospheric clarity contributes to
a decrease in intensity of light falling on the surface of a water body.
Measurements of the relative spectral distribution of the radiation which
penetrates a cloud or an overcast indicate that it will be essentially the
same as that which enters (Hull, 1954, and Leighton, 1961). Therefore, the
relative energy/wavelength curve of sun and sky radiation together is nearly
the same on a clear and overcast day. This phenomenon was tested by
determining the photolysis rate as a function of solar radiation, both
normalized to that predicted by interpolation from results for clear days.
This relationship is depicted in Figure 14. Since the experimentally
determined observations are close to the 45° line in Figure 14, it can be
concluded that the photodecomposition of hexacyanoferrate (ll) and (ill)
solutions is a direct function of natural light intensity for various
meteorological conditions.
Photolysis in Natural Waters
The rate of photodecomposition at 15° C for hexacyanoferrate (II)
solutions initially containing 100 ug/1 CN and in Pyrex test tube cells
exposed to natural light was essentially the same when different waters were
used in preparation of the test solutions. This is demonstrated in Table 7 by
the similarity in photolysis rates for solutions prepared with different
unfilterd natural waters relative to the rates determined for comparable
solutions prepared with deionized water. The slight reduction (Table 7) in
relative photolysis rate for solutions prepared with Minnesota or Mississippi
39
-------
Ill
5
tr
LU
1
UJ
(T
10 20 30 40 50 60 70 80 90 100
PERCENTAGE SOLAR RADIATION
Figure 14. Relationship between photolysis rate and solar radiation, both
normalized to that predicted for a clear day.
40
-------
TABLE 7. DETERMINED PHOTOLYSIS RATES FOR HEXACYANOFERRATE
(II) SOLUTIONS PREPARED WITH DIFFERENT WATERS RELATIVE
TO THE RATES FOR SIMILARLY EXPOSED SOLUTIONS
PREPARED WITH DEIONIZED WATER
Type of
water
Deionized
Well
Lake Phalen
Lake Como
Elephant Lake
Mississippi River
Minnesota River
Test
PH
7.9
8.1
7.6
8.5
7.4
8.5
7.5
Relative
photolysis rate
1.000
1.044
0.988
1.096
0.991
0.819
0.960
41
-------
River water is probably due to the decreased light penetration into the 2 cm
path length of the cells containing natural water solutions relative to those
prepared with deionized water. Therefore, the relationships presented in
previous sections, as determined for deionized water solutions, are appropriate
for hexacyanoferrate (II) solutions prepared with several different natural
waters.
Various natural waters and chemicals were used to determine empirically
the effect of materials in different waters on the photolysis rate of
Fe(CN)g. These results are summarized in Table 8 where the photolysis
rate at 15° C of hexacyanoferrate (III) solutions initially containing 100 Mg/1
CN and prepared with various waters are compared to the rates determined for
similarly exposed solutions prepared with deionized water. The Fe++ ion
from the addition of ferrous chloride under near-zero dissolved oxygen levels
enhanced the photolysis rate, whereas Fe ion added as ferric chloride
or that derived from the oxidation of Fe depressed the reaction. The
presence of ammonia (ammonium chloride) and nitrate (sodium nitrate) had no
effect on the photolysis rate, but dissolved sulfide (sodium sulfide)
dramatically enhanced the photodecomposition reaction. The photolysis rates
for ferricyanide solutions prepared with several different natural surface
waters and well water were about twice as great in the natural water as in
deionized water (Table 8). This acceleration may be attributed to the rapid
reduction of hexacyanoferrate (ill) to the more photochemically active
hexacyanoferrate (II). This possibility is supported by the fact that the
ratios of photolysis rates at 15° C for comparable ferri- and ferrocyanide
solutions initially containing 100 yg/1 CN and prepared with the same natural
surface water from various sources averaged 0.92 (Table 9). If our results for
experiments conducted with natural waters are typical of other waters, then it
is believed that iron-cyanides in effluents being discharged into various
natural waters will photochemically respond like hexacyanoferrate (II) ions,
since essentially no Fe(CN)j?~ ions will be present in the receiving
waters.
This study has demonstrated a means by which the cyanide in relatively
nontoxic iron-cyanide complexes may, under certain conditions, be largely
liberated as toxic free cyanide in natural waters. The rate of this process
and the concentration of free cyanide produced can be approximately predicted
from information derived in our study and by utilizing the computer program of
Zepp and Cline (1977). These calculations can be made as a function of
longitude, latitude, time of day and year, and depth in a natural water body.
The extent of the photolysis reaction will be a function of the iron-cyanide
concentration and the degree and duration of illumination. Photolysis may be
negligible in deep, turbid, or shaded waters, and the slowly liberated
nonpersistent free cyanide may decrease or escape as rapidly as it is released.
However, in relatively clear shallow waters, or in situations where effluents
are stratified in the light penetration zone, the photodecomposition of
iron-cyanides may produce free cyanide concentrations in the ug/1 range that
could adversely affect the distribution and abundance of aquatic organisms
(Smith et al., 1979).
42
-------
TABLE 8. DETERMINED PHOTOLYSIS RATES FOR HEXACYANOFERRATE
(III) SOLUTIONS PREPARED WITH DIFFERENT WATERS RELATIVE
TO THE RATES FOR SIMILARLY EXPOSED SOLUTIONS PREPARED
WITH DEIONIZED WATER (DW)
Type of
water3
Deionized
DW + 0.5 mg/1 Fe+++ (DO)
DW + 0.5 mg/1 Fe++ (DO)
DW + 0.5 mg/1 Fe++ (near zero DO)
DW + 5.0 mg/1 Fe++ (DO)
DW + 0.5 mg/1 NH4+ (DO)
DW + 0.5 mg/1 N03~ (near zero DO)
DW + 0.5 mg/1 sulfide (near zero DO)
WW - after Fe removal0 (DO)
WW - after Fe removal0 (near zero DO)
WW - before Fe removal0 (DO)
Elephant Lake
Lake .Como
Minnesota River
Mississippi River
Bog
Test
PH
7.8
6.2
6.6
7.3
6.0
7.5
7.2
7.6
7.6
8.7
8.4
7.3
8.5
7.7
8.3
7.2
Relative
photolysis rate
1.000
0.790
1.486 •* 0.450b
2.995 •* 1.690b
approx. 0
1.095
0.970
2.309
1.370
1.692
, 2.228
2.299
2.157
1.988
1.758
1.528d
aDO refers to solutions that were aerated prior to exposure and it is
assumed that dissolved oxygen concentrations were near saturation
Near zero DO refers to solutions that were stripped with N2 prior to
exposure and it is assumed that dissolved oxygen concentrations were
near zero.
"Rate decreased during exposure period.
cRemoval of iron from the well water (WW) was accomplished by a
Culligan catalytic system which converts Fe(ll) to Fe(lII). Iron
concentrations were approximately 0.3 mg/1 before and <0.05 mg/1 after
removal.
°Low because of strong light absorption (Figure 11) by bog water
relative to deionized water solutions in test tube exposure cells.
43
-------
TABLE 9. RATIO OF DETERMINED PHOTOLYSIS RATES FOR
HEXACYANOFERRATE (ill) TO (II) SOLUTIONS OF EQUAL
INITIAL TOTAL CYANIDE CONCENTRATION PREPARED
WITH DIFFERENT WATER TYPES AND EXPOSED TO
THE SAME NATURAL LIGHT CONDITIONS
Type of Ratio of
water photolysis rates
Deionized 0.417
WW - after Fe removal 0.694
WW - before Fe removal 1.090
Lake Como 0.869
Elephant Lake 0.921
Mississippi River 0.938
Minnesota River 0.868
Bog 0.987
44
-------
REFERENCES
Adamson, W. 1952. Electron Transfer Processes and the Oxidation-Reduction
Reactions of Hexacyanoferrate (ill) Ion in Aqueous Solution. J. Phys.
Chem. 56: 858-862.
Alich, Sister M. A., D. T. Haworth, and Sister M. F. Johnson. 1967. Spectro-
photometric Studies of Hexacyanoferrate (III) ion and its Reaction with
Iron (III) in Water and Ethanol. J. Inorg. Nucl. Chem. 29(7):
1637-1642.
American Public Health Association, American Water Works Association, and
Water Pollution Control Federation. 1971. Standard Methods for the
Examination of Water and Wastewater. 13th ed. Washington, D.C. 874 p.
Asperger, S. 1952. Kinetics of the Decomposition of Potassium Ferrocyanide
in Ultra-Violet Light. Trans. Faraday Soc. 48: 617-624.
Asperger, S., I. Murati, and D. Pavlovic. 1960. Kinetics and Mechanism of
the Decomposition of Complex Cyanides of Iron (II) and Molybdenum (IV).
J. Chem. Soc. 730-736.
Balzani, V., and V. Carassiti. 1970. Photochemistry of Coordination
Compounds. Academic Press, New York, N.Y. 432 p.
Bener, P. 1972. Approximate Values of Intensity of Natural Ultraviolet
Radiation for Different Amounts of Atmospheric Ozone. U.S. Army Report
DAJA 37-68-C-1017, Davos Platz, Switzerland.
Broderius, S. J. 1973. 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 University, Coryallis, Oregon.
287 p.
Broderius, S. J., and L. L. Smith, Jr. 1977. Direct Determination and
Calculation of Aqueous Hydrogen Sulfide. Anal. Chem. 49: 424-428.
Broderius, S. J., and L. L. Smith, Jr. 1979. Lethal and Sublethal Effects of
Binary Mixtures of Cyanide and Hexavalent Chromium, Zinc, or Ammonia to
the Fathead Minnow (Pimephales promelas) and Rainbow Trout (Salmo
gairdneri). J. Fish. Res. Board Can. 36: 164-172.
Broderius, S. J., L. L. Smith, Jr., and D. T. Lind. 1977. Relative Toxicity
of Free Cyanide and Dissolved Sulfide Forms to the Fathead Minnow
(Pimephales promelas). J. Fish. Res. Board Can. 34: 2323-2332.
45
-------
Burdick, G. E., and M. Lipschuetz. 1950. Toxicity of Ferro- and Ferricyanide
Solutions to Fish, and Determination of the Cause of Mortality. Trans.
Am. Fish. Soc. 78: 192-202.
Carassiti, V., and V. Balzani. 1960. Photochemistry of Complex Salts in
Solution: IV. Kinetics of the Photochemical and Thermal Decomposition
of Potassium Ferrocyanide. Annali Di Chimica (Rome). 50: 782-789.
Doudoroff, P. 1976. Toxicity to Fish of Cyanides and Related Compounds. A
Review. EPA-600/3-76-033, U.S. Environmental Protection Agency, Duluth,
Minn. 155 p.
Emschwiller, G., and J. Legros. 1954. The Photolysis of Potassium Ferro-
cyanide in Aqueous Solution. Comptes Rendus. 239: 1491-1493.
Hull, J. N. 1954. Spectral Distribution of Radiation from Sun and Sky.
Trans. Ilium. Eng. Soc. (London). 19: 21-28.
Hutchinson, G. E. 1957. A Treatise on Limnology Vol. I - Geography, Physics,
and Chemistry. Wiley and Sons, New York, N.Y. 1015 p.
Ibers, J. A., and N. Davidson. 1951. On the Interaction Between Hexacyanato-
ferrate (III) Ions and (a) Hexacyanatoferrate (II) or (b) Iron (III)
Ions. J. A. Chem. Soc. 73: 476-478.
Kobayashi, T., and K. Mori. 1973. Formation of Compounds Toxic to Fish from
Ferro- and Ferricyanides by Photodecomposition (Occurrence of Toxicity to
Fish by Photodecomposition of Complex Cyanides). Nihon Mizu Shori
Gijutsu Kenkyuhai, Mizu. Shori. Gijutsu. 14(6): 575-579.
Kongiel-Chablo, I. 1966. Behavior of Complex Cyanides in Natural Water with
High Rate of Contamination. Roczniki Panstwowego Zakladu Hig. 17(1):
95-102.
Leighton, P. A. 1961. Photochemistry of Air Pollution. Academic Press, New
York, N.Y. 300 p.
London, J. 1963. The Distribution of Total Ozone in the Northern Hemisphere.
Beitr. Phys.*Atmos. 36: 254-263.
Lure, Y. Y., and V. A. Panova. 1964. Behavior of Cyanide Compounds in Water
Ponds. Gidrokhim Materialy. 37: 133-143.
MacDiarmid, A. G., and N. F. Hall. 1953. Illumination - pH Effects in
Solutions of Complex Cyanides. J. Am. Chem. Soc. 75: 5204-5207.
Mitra, R. P., D. V. S. Jain, A. K. Bannerjee, and K. V. Raghavachari. 1963.
Role of Free Radicals in the Photo-Oxidation of Fe"1"^ £n Acidic
Solutions of Ferrocyanide. Nature. 200: 163-164.
46
-------
Moggi, L., F. Bolletta, V. Balzani, and F. Scandola. 1966. Photochemistry of
Coordination Compounds-XV Cyanide Complexes. J. Inorg. Nuclear Chem.
28: 2589-2597.
Myers, C. S., and T-. lezzi. 1950. A Report of Experiments on the Effects of
Certain Toxicants on Fish Life with Especial Reference to the Effects of
Photodecomposition. Ind. Wastes Division, Bureau of San. Engr., Penn.
Dept. of Health, Harrisburg, Penn. 10 p.
Ohno, S. 1967. The Photochemistry of Aqueous Hexacyanoferrate (II)
Solutions. I. Photo-Aquation Reaction at 3660 A. Bull. Chem.
Soc. Japan. 40: 1765-1769.
Ohno, S., and G. Tsuchihashi. 1965. The Photochemistry of Hexacyanoferrate
(II) Ions in Aqueous Solutions. Bull. Chem. Soc. Japan. 38(6):
1052-1053.
Palaty, J., and M. Horokova-Jakubu. 1959. The Course and Rate of Removal of
Cyanides from Water Under Natural Conditions. Sbornik Vscht, Prague,
Oddil Fak. Tech. Paliv A Vody. 3(1): 83-110.
Robinson, N. 1966. Solar Radiation. Elsevier Publishing Company, New York,
N.Y. 347 p.
Smith, L. L., Jr., S. J. Broderius, D. M. Oseid, G. L. Kimball, W. M. Koenst,
and D. T. Lind. 1979. Acute and Chronic Toxicity of HCN to Fish and
Invertebrates. EPA-600/3-79-009, U.S. Environ. Prot. Agency, Duluth,
Minn. 115 p.
Steel, R. G. D., and J. H. Torrie. 1960. Principles and Procedures of
Statistics with Special Reference to the Biological Sciences.
McGraw-Hill Book Co., Inc. New York, N.Y. 481 p.
Stumm, W., and J. J. Morgan. 1970. Aquatic Chemistry. An Introduction
Emphasizing Chemical Equilibria in Natural Waters. John Wiley and Sons,
Inc., New York, N.Y. 583 p.
Terhaar, C. J., W. S. Ewell, S. P. Dziuba, and D. W. Fassett. 1972. Toxicity
of Photographic Processing Chemicals to Fish. Photogr. Sci. Eng. 16:
370-377.
Tyler, J. E., and R. W. Preisendorfer. 1962. Transmission of Energy Within
the Sea. In: The Sea, Vol. I, M. H. Hill, ed. Interscience, New York,
N.Y., pp. 397-451.
West, L. E. 1970. Disposal of Waste Effluents from Motion-Picture Film
Processing. J. Soc. Motion Picture Tel. Eng. 79: 765-771.
Wetzel, R. G. 1975. Limnology. W. B. Sanders Co., Philadelphia, Penn. 743
P«
Zepp, R. G., and D. M. Cline. 1977. Rates of Direct Photolysis in Aquatic
Environment. Environ. Sci. Technol. 11(4): 359-366.
47
-------
APPENDIX A
BEAM ATTENUATION COEFFICIENTS
oo
Water body
Turbidity (NTU)
Wave length (nm)
297.5
300.0
302.5
305.0
307.5
310.0
312.5
315
.0
317.5
320
323
330
340
350
360
370
380
390
400
410
420
430
440
450
460
470
480
.0
.1
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
Deionized
0.0028
0.0028
0.0026
0.0025
0.0024
0.0023
0.0022
0.0021
0.0020
0.0019
0.0018
0.00152
0.00122
0.00100
0.00082
0.00069
0.00056
0.00043
0.00035
0.00030
0.00026
0.00023
0.00020
0.00017
0.00016
0.00016
0.00016
Bentonite Solutions
5.6
0.118
0.114
0.113
0.110
0.108
0.107
0.105
0.103
0.102
0.101
0.099
0.096
0.093
0.090
0.088
0.085
0.082
0.079
0.076
0.072
0.070
0.067
0.065
0.061
0.060
0.057
0.056
6.7
0.122
0.121
0.119
0.116
0.114
0.112
0.110
0.109
0.108
0.107
0.106
0.102
0.098
0.096
0.095
0.092
0.089
0.086
0.082
0.080
0.076
0.073
0.070
0.068
0.066
0.064
0.061
9.9
0.221
0.216
0.213
0.208
0.203
0.201
0.199
0.195
0.192
0.189
0.187
o.iai
0.174
0.169
0.166
0.160
0.154
0.149
0.143
0.138
0.134
0.130
0.125
0.121
0.118
0.114
0.112
23.0
0.509
0.496
0.487
0.478
0.469
0.462
0.455
0.450
0.444
0.438
0.429
0.419
0.399
0.387
0.377
0.365
0.351
0.339
0.328
0.317
0.308
0.299
0.291
0.281
0.274
0.268
0.260
24.0
0.527
0.516
0.509
0.498
0.491
0.481
0.475
0.469
0.463
0.457
0.450
0.438
0.420
0.405
0.391
0.377
0.364
0.349
0.337
0.327
0.316
0.305
0.294
0.286
0.277
0.269
0.263
46.0
0.963
0.951
0.924
0.910
0.893
0.879
0.866
0.857
0.848
0.833
0.824
0.799
0.770
0.745
0.721
0.697
0.674
0.652
0.631
0.609
0.590
0.575
0.559
0.545
0.535
0.521
0.509
54.0
1.081
1.060
1.041
1.018
1.004
0.991
0.971
0.963
0.951
0.939
0.924
0.900
0.866
0.839
0.815
0.788
0.759
0.735
0.710
0.690
0.672
0.654
0.635
0.618
0.604
0.590
0.577
57.0
1.131
1.102
1.086
1.066
1.046
1.032
1.009
1.000
0.991
0.983
0.963
0.943
0.907
0.876
0.848
0.818
0.790
0.764
0.742
0.719
0.699
0.678
0.660
0.642
0.629
0.618
0.602
68.0
1.387
1.367
1.328
1.310
1.292
1.276
1.252
1.237
1.222
1.208
1.187
1.155
1.114
1.081
1.046
1.013
0.979
0.947
0.914
0.886
0.857
0.836
0.815
0.790
0.772
0.750
0.735
-------
APPENDIX A
BEAM ATTENUATION COEFFICIENTS
Water body
Turbidity (NTU)
Wave length (nm)
297.5
300.0
302.5
305.0
307.5
310.0
312.5
315.0
317.5
320.0
323.1
330.0
340.0
350.0
360.0
370.0
380.0
390.0
400.0
410.0
420.0
430.0
440.0
450.0
460.0
470.0
480.0
Wind-blown silt
23.0
0.346
0.343
0.338
0.334
0.330
0.327
0.322
0.319
0.316
0.313
0.309
0.301
0.291
0.281
0.272
0.261
0.253
0.246
0.238
0.231
0.226
0.220
0.214
0.209
0.203
0.200
0.194
43.0
0.688
0.680
0.674
0.664
0.658
0.650
0.642
0.635
0.631
0.622
0.613
0.599
0.580
0.558
0.539
0.523
0.506
0.492
0.479
0.465
0.452
0.441
0.428
0.419
0.408
0.398
0.388
63.0
1.022
1.000
0.996
0.991
0.971
0.959
0.951
0.947
0.928
0.921
0.910
0.886
0.857
0.830
0.801
0.780
0.754
0.738
0.719
0.697
0.678
0.660
0.644
0.627
0.614
0.599
0.585
Square
Lake
0.43
0.023
0.022
0.021
0.020
0.018
0.018
0.017
0.016
0.016
0.015
0.014
0.013
0.012
0.009
0.008
0.007
0.005
0.005
0.004
0.004
0.004
0.003
0.003
0.003
0.003
0.002
0.002
Lake
Phalen
2.7
0.066
0.061
0.059
0.056
0.052
0.051
0.048
0.046
0.045
0.043
0.041
0.039
0.034
0.030
0.026
0.023
0.021
0.020
0.018
0.017
0.016
0.015
0.014
0.013
-0.013
0.013
0.013
Lake
Como
5.4
0.123
0.119
0.114
0.109
0.107
0.103
0.101
0.097
0.096
0.092
0.089
0.084
0.077
0.071
0.066
0.061
0.057
0.053
0.051
0.048
0.047
0.046
0.044
0.041
0.040
0.039
0.039
Elephant
Lake
2.9
0.171
0.164
0.156
0.149
0.143
0.137
0.131
0.126
0.121
0.117
0.111
0.102
0.087
0.074
0.064
0.056
0.047
0.041
0.036
0.032
0.028
0.025
0.022
0.019
0.018
0.016
0.015
Mississippi
River
3.6
0.206
0.197
0.190
0.182
0.176
0.170
0.166
0.159
0.154
0.149
0.143
0.131
0.115
0.101
0.088
0.078
0.070
0.063
0.057
0.052
0.048
0.045
0.041
0.039
0.036
0.035
0.032
Minnesota
River
14.0
0.283
0.275
0.268
0.260
0.251
0.243
0.237
0.230
0.224
0.218
0.210
0.198
0.178
0.161
0.147
0.134
0.125
0.114
0.108
0.101
0.096
0.091
0.086
0.083
0.081
0.077
0.075
Bog Water
0.9
0.548
0.527
0.509
0.491
0.470
0.455
0.438
0.420
0.407
0.395
0.372
0.337
0.292
0.248
0.208
0.177
0.148
0.125
0.107
0.090
0.076
0.066
0.057
0.049
0.043
0.038
0.034
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-600/3-80-005
2.
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
Direct Photolysis of Hexacyanoferrate Complexes:
Proposed Applications to the Aquatic Environment
5. REPORT DATE
January; 1980 issuing date
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Steven J. Broderius and Lloyd L. Smith, Jr.
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Department of Entomology, Fisheries, and Wildlife
University of Minnesota
St. Paul, Minnesota 55108
1O. PROGRAM ELEMENT NO.
1BA608
11. CONTRACT/GRANT NO.
Grant No. R805291
12. SPONSORING AGENCY NAME AND ADDRESS
Environmental Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Duluth, Minnesota 55804
13. TYPE OF REPORT AND PERIOD COVERED
14. SPONSORING AGENCY CODE
EPA/600/03
15. SUPPLEMENTARY NOTES
16. ABSTRACT
The theory and computations described by Zepp and Cline (1977) were experimentally tested in predicting the
direct photolysis rates of dilute hexacyanoferrate (II) and (III) solutions in the aquatic environment. Essential
information for these calculations includes the quantum yield for the photoreaction, molar extinction coefficients
of the complex ions for wavelengths > 295 nm, solar irradiance data used to calculate specific sunlight absorption
rates, and the assumption that the photolysis reaction obeys a first-order kinetic rate expression. Direct
photolysis rates of the irreversible photochemical reactions are calculated as a function of the time of year,
latitude, time of day, meteorological conditions, and depth in natural water bodies. Light of wavelengths < 480
nm is active in the photolysis reactions, and pH, temperature, and concentration all affect the reaction to
varying degrees. Assuming first-order kinetics, in which the rate constant was approximately concentration
independent within the range of 25-100 Mg/1 total cyanide, the minimum quantum yields of HCN formation were 0.14
and 0.0023 for the iron (II) and (III) complexes, respectively. These values correspond to minimum, nearsurface,
midday half-lives at midsummer of about 18 and 64 min at St. Paul, Minn. The photolysis rate at various fixed
depths in a natural water column, when compared with that at the surface, decreases exponentially with depth. It
is suggested that the photolysis reactions are enhanced by suspended material in turbid waters because of the
forward scattering of light when compared with that theoretically calculated from beam attenuation coefficients.
Hexacyanoferrate (II) and (III) solutions of equal initial total cyanide concentration respond photochemically
quite differently from one another in solutions prepared with deionized water, but respond in a similar manner for
solutions prepared with natural waters. The potentially rapid photodecomposition of iron-cyanides with formation
of HCN suggests that this phenomenon may be of toxicological importance under certain environmental conditions.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
D.IDENTIFIERS/OPEN ENDED TERMS C. COSATI Field/Group
Cyanides
Photolysis
Aquatic
Environment
06/F
18. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
19. SECURITY CLASS (This Report)
UNCLASSIFIED
21. NO. OF PAGES
60
2O. SECURITY CLASS (This page)
UNCLASSIFIED
22. PRICE
EPA Form 2220-1 (Rev. 4-77) PREVIOUS EDITION is OBSOLETE
50
a US GOVERNMENT PRINTING OFFICE: 1980 -657-146/5539
------- |