U.S. DEPARTMENT OF COMMERCE
National Technical Information Service
PB-255 852
Applicability of the Cyanide
Electrode for Measuring Free
and Total Cyanide
Central State Univ.
Prepared For
Environmental Monitoring & Support Lab.
June 1976
-------
TECHNICAL REPORT DATA
(Please nod Imtnictioru on the rtvtrte before completing}
REPORT NO.
EPA-600/4-76-020
3. RECIPIENT'S ACCESSION-NO.
TITLE AND SUBTITLE
The Applicability of the Cyanide Electrode for
Measuring Free and Total Cyanide
S. REPORT DATE
June 1976 (Issuing Date)
6. PERFORMING ORGANIZATION CODE
AUTHOR(S)
Albert Schlueter
8. PERFORMING ORGANIZATION REPORT NO.
. PERFORMING ORGANIZATION NAME AND ADDRESS
Central State University
Wilberforce, Ohio 45384
10. PROGRAM ELEMENT NO.
1HA323 and EHE62^
11. CONTRACT/GRANT NO.
R-802755-01
12. SPONSORING AGENCY NAME AND ADDRESS
Environmental Monitoring and Support Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
13. TYPE OF REPORT AND PERIOD COVERED
14. SPONSORING AGENCY CODE
EPA-ORD
15. SUPPLEMENTARY NOTES
16. ABSTRACT
The Orion model 94-06 cyanide electrode was evaluated to determine its
applicability to water and wastewaters. The calibration curve was Nernstian
over the concentration range of 0.26 to 26 ppm, and the slope of the curve
was 59 mv per decade change. This work consisted in studying the response
of the cyanide electrode to cyanide when this ion was present in solution in
both free and complex forms. The results show conclusively that the electrode
responds only to free cyanide in solution and not at all to that complexed to
metals.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
COSATi Field/Group
Cyanides
Water analysis
Monitors
Hydrogen cyanide
Iron cyanides
Cyanide electrodes
Free cyanide
13B
18. DISTRIBUTION STATEMENT
Release to public
10. SECURITY CLASS (This Report)
21. NO. OF PAGcS
29
20. «CT/RTTY'C.C&SS (Thlj page)
Unclassified
22. PRICE
EPA Form 2220-1 (t-73)
«USGPO: 1976 — 657.«95/5452 ««glon 5-11
PRICES SUBJECT TO CHANGE
-------
EPA-600/4-76-020
June 1976
APPLICABILITY OF THE
CYANIDE ELECTRODE FOR MEASURING FREE
AND TOT/i, CYANIDE
by
Albert Schlueter
Department of Chemistry
Central State University
Wilberforce, Ohio 45384
Grant No. R-802755-01
Project Officer
Morris Gales
Environmental Monitoring and Support Laboratory
Cincinnati, Ohio 45268
U.S. ENVIRONMENTAL PROTECTION AGENCY
OFFICE OF RESEARCH AND DEVELOPMENT
ENVIRONMENTAL MONITORING AND SUPPORT LABORATORY
CINCINNATI, OHIO 45268
-------
DISCLAIMER
This report has been reviewed by the Environmental Monitoring
and Support Laboratory, 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
Environmental measurements are required to determine the quality
of ambient waters and the character of waste effluents. The Environ-
mental Monitoring and Support Laboratory-Cincinnati conducts research
to:
o Develop and evaluate techniques to measure the presence and
concentration of physical, chemical, and radiological
pollutants in water, wastewater, bottom sediments, and
solid waste.
o Investigate methods for the concentration, recovery, and
identification of viruses, bacteria, and other microbio-
logical organisms in water. Conduct studies to determine
the responses of aquatic organisms to water quality.
o Conduct an Agency-wide quality assurance program to assure
standardization and quality control of systems for monitor-
ing water and wastewater.
There is an ever-increasing interest in the use of electrode
methods to analyze water and waste samples, whether the resulting
data are to be used for research, surveillance, compliance monitoring,
or enforcement purposes. Accordinaly, the Environmental Monitoring
and Support Laboratory has an ongoing methods research effort in the
development, evaluation, and modification of electrode procedures.
This particular report pertains to the evaluation of the cyanide elec-
trode. The method has potential routine application for the analysis
of free cyanide in surface waters and domestic and industrial wastes.
Dwight G. Ballinger
Director
Environmental Monitoring and Support Laboratory
111
-------
ABSTRACT
The Orion model 94-06 cyanide electrode was evaluated to deter-
mine its applicability to water and wastewaters. The calibration
curve was Nernstian over the concentration range of 0.26 to 26 ppm,
and the slope of the curve was 59 mv per decade change. This work
consisted in studying the response of the cyanide electrode to cya-
nide when this ion was present in solution in both free and complex
forms. The results show conclusively that the electrode responds
only to free cyanide in solution and not at all to that complexed to
metals.
IV
-------
CONTENTS
Page
Foreword iii
Abstract iv
List of Tables vi
Acknowledgment vii
I Introduction 1
II Surciuary 2
III Conclusions 3
IV Recommendations 4
V Experimental 5
VI Data 7
-------
LIST OF TABLES
Number
1 Stability Equilibrium Constants (Ks) and Rates 8
of Ligand Exchange for Some Typical-Cyanide
Complexes
2 Response of the Cyanide Electrode to Metal-Cyanide 9
Complexes at Different pH Values
3 Relative Response in Millivolts of the Cyanide 10
Electrode to Successive 1 ml Additions of 1.0 x
10-3M Metal Ions to 100.0 ml of a 1.0 x 10'4M
(2.6 ppm) Cyanide Solution
4 Relative Millivolt Readings 12
5 Relative Response in Millivolts of the Cyanide 13
Electrode to the Addition of Mickel (11) and
Copper (11) to Solutions Containing Both EDTA
and Cyanic?- (pH13)
6 Relative Response in Millivolts of the Cyanide 14
Electrode to the Dissociation of the Nickel-
Cyanide Complex by EDTA at pH13
7 Response of the Cyanide Electrode to the Cyanide 16
Concentrations in Solutions Resulting From Acid
Hydrolysis-Distillation of Solutions of KCN and
Metal Complexes
8 Acid Hydrolysis-Distillation of Metal-Cyanide 16
Complexes
9 Effects of Metal Ions on the Recovery of Cyanide 17
With Acid Hydrolysis-Distillation
10 Response of the Cyanide Electrode to the Cyanide 19
Concentration in Several Natural Samples Using
Various Analysis Methods
11 Electrode Response to Solutions Spiked with Additional 19
Cyanide (Increased by 5 x 10"^M_)
12 Analysis of Fish from Fish Kill 19
VI
-------
ACKNOWLEDGMENT
The work of Michael Taylor, a senior Chemistry major at Central
State University, in collecting much of the data contained in this
report is gratefully acknowledged. The assistance of Gerald Pitts and
Ernest Lieis in the initial phase of this work is note.
Also acknowledged is the considerable help of Morris Gales, Jr.,
Project Officer of this grant, and others at the Environmental Monitor-
ing and Support Laboratory in Cincinnati, Ohio, for clearly defining the
scope of the work. The assistance of Mr. Gales specifically and the
U.S. Environmental Protection Agency generally has supplied Central
State University faculty and students a valuable opportinity to partici-
pate in important and relevant research work.
Vll
-------
SECTION I
INTRODUCTION
The classical cyanide determination uses acid-hydrolysis to convert
free and complexed cyanide into hydrogen cyanide, which is distilled
from solution for subsequent determination. The result of this analysis
thus gives the total amount of cyanide in solution (i.e. both free and
complexed cyanide). It can be argued that for evaluation of environ-
mental impact, in most cases, the free cyanide levels are of more signi-
ficance than the total cyanide content. Cyanide complexed with ferrous
or ferric iron, for example, is both kineticly and thermodynamicly in-
ert, and thus, the complexed-cyanide has no business being classified
with the free cyanide if the amount of chemically-active cyanide is being
evaluated. The important distinction between free and total cyanide
levels has not received sufficient attention by those doing cyanide
analyses or by those evaluating the results of cyanide determinations.
Though the classical hydrolysis-distillation cyanide determination was
only amenable to determinations of the total cyanide levels, the recent
development of the cyanide-selective electrode makes measurement of
free cyanide potentially feasible. It is essential that the meaning of
the cyanide concentrations measured with the cyanide electrode be care-
fully ascertained and their relationship to free and total cyanide de-
termined. Though the responses of the cyanide electrode to cyanide
levels in laboratory-prepared samples and its utility in the analysis of
many natural products have been amply reported, the meaning of the re-
sults as free or total cyanide remains ambiguous. The purpose of this
report is to clarify the meaning of the cyanide electrode responses.
This study investigates the effect of metal-cyanide complexes on the ion
selective electrode, the effects of metal-contaminats on the cyanide
electrode response, and the conditions under which the electrode can be
used to give evaluations of the free and total cyanide levels.
-------
SECTION II
SUKMARY
The work presented here consisted of studying the response of the
cyanide electrode to cyanide when this ion was present in solution in
both free and complexed forms. The results show conclusively that the
electrode responds only to free cyanide in solution and not at all to
that complexed to metals. Nickel and copper are shown to form the most
stable complexes, and ferrous and ferric iron, zinc, and lead form less
stable complexes.
Attempts were made to use the cyanide electrode to measure total
cyanide concentration (i.e. both free and complexed cyanide) by adding
a strong complexing agent to decompose the cyanide complexes. Both
hydroxide and ethylenediamine tetracetic acid (EDTA) were investigated,
but the thermodynamic and kinetic stability of the cyanide complexes
rendered determination of total cyanide with the cyanide electrode
unfeasible.
Studies with the cyanide electrode in the hydrolysis-distillation
procedure for determining total cyanide indicate that high metal ion
concentrations can give low results. In light of this result and the
demonstrated kinetic and thermodynamic stability of the metal-cyanide
complexes, the environmental significance of determination of total
cyanide levels is questioned. It is suggested that determination of
free cyanide concentrations with the cyanide elertrod0 T.lght be ,x>re
significant and meaningful.
Studies of the cyanide levels in natural water samples are found
to be at the limit of Nernstian response of the cyanide electrode.
Electrode responses s^em to correlate with free cyanide levels even
though the determinations contain larger errors.
The cyanide electrode is used to determine the cyanide levels in
two fish and one crayfish that died in a large fi^h kill. Though an
effluent containing cyanide was suspected, this study indicates normal
cyanide levels.
-------
SECTION III
CONCLUSIONS
This work has demonstrated four things:
1. that the cyanide electrode responds to free cyanide levels as
opposed to the total cyanide concentration in aqueous solution;
2. th-ic a simple procedure for determining total cyanide using
the cyanide electrode is not feasible because of the great
stability of metal-cyanide complexes;
3. that the classical acid hydrolysis-distillation procedure
for determining total cyanide may not accurately yield true
total cyanide levels in the presence of large metal ion con-
centrations;
4. that the concentrations of cyanide in natural surface water
samples are at the limit of Nernstian response of the cyanide
electrode and beyond.
The unusually great kinetic stability of metal-cyanide complexes,
even the very labile nickel-cyanide complex, coupled with their great
thermodynamic stability renders determination of total cyanide levels
with the cyanide electrode extremely unlikely. Concomitantly, this
exceedingly great stability brings to question the significance of de-
terminations of total cyanide. This work indicates that evaluation of
free cyanide, or perhaps cyanide freed by EDTA using the cyanide
electrode, might be the more significant measurement of the cyanide
content in water samples. Such measurements would more accurately re-
flect the level of chemically active cyanide present in sar.ples.
-------
SECTION IV
RECOWENDATIONS
Several additional avenues of research on measurement of cyanide
levels in water samples need to be explored:
1. the indication from this work that the acid hydrolysis-
distillation procedure for total cyanide gives low cyanide
results when metal ions are present in solution need; to be
confirmed;
2. a procedure for concentrating cyanide in natural water
samples must be developed so accurate determinations with
the cyanide electrode can be made;
3. determination of the mechanism of decomposition of metal-
cyanide complexes (if indeed any decomposition occurs) in
natural and wastewaters should be investigated. If it can
be shown that these complexes are nearly totally stable, or
if their decomposition mechanism results in destruction of
the cyanide (e.g. to methane and ammonia), then it can be
convincingly argued that free cyanide and NOT total cyanide
is the only environmentally meaningful measurement.
-------
SECTION V
EXPERIMENTAL
Measurements
All readings in this report were collected using an Orion Model 801A
DigitalpH/raVolt meter, an Orion Model 90-02 double junction reference
electrode and an Orion Model 94-06 cyanide electrode.
Solutions
All solutions were prepared from reagent grade potassium cyanid**
Hydroxide concentrations were established using standard solutic . p-.e-
pared from Acculutes prepared by Anachemia Chemicals Ltd. and obtained
from the Sargeant-Welch Company. Sodium nitrate was used to adjust the
ionic strength of solutions to 0.10M_ although most readings were carried
out in 0.10M NaOH (pH = 13) solutions. The outer chamber of the double
junction reference electrode was filled with a 10% solution of sodium
nitrate.
All other chemicals used were of reagent grade except the potassium
tetracyanonickelate (II) which was prepared by the procedure given in
Inorganic Synthesis, volume 1.
Distilled water was used to prepare all solutions. Checks of the
cyanide level in the distilled water with the cyanide-electrode always
indicated considerably less than 10~^M_ concentration.
Stock and Standard Solutions
A 0.10M_ stock solution of potassium cyanide at a pH of 13 was pre-
pared at least once a month for most of the succeeding readings. From
this solution were prepared a series of standard solutions ranging in con-
centration from 1 x 10"3M (26 ppm) to 1 x 10'7M (2.6 ppb) cyanide. These
standard solutions were prepared fresh at least once every two weeks.
All solutions were stored in polyethylene bottles. No sign of decom-
position was observed for any solution. A 1 x 10"5M (.26 ppm) cyanide
solution at a pH of 13 gave the same electrode response even after a four
month storage period.
Operating Procedure
The instrumental techniques and general operating procedures used
with the cyanide electrode and digital pH meter were those suggested by
-------
the manufacturers of the equipment. In general, readings were taken
while stirring and after a wait of one or two minutes for stabilization.
Under certain conditions (e.g. dilute solutions, high pH studies, etc.)
drifting and fluctuating results were observed requiring longer waiting
periods.
Calibration curves of electrode potentials versus pCN (the negative
logarithm of the cyanide molarity) were prepared on the average of once a
week and were always made immediately prior to taking readings deemed im-
portant. The curves were always linear and Nernstian over the concen-
tration range of 10"-* (26 ppm) to 10'^M (.26 ppm) cyanide. The slopes of
these plots ranged from 56 to 65 although usually they were close to 59,
the theoretical value for 24°C. The range of values for the slope are
partially a result of poor temperature control in the room this work was
carried out in during the summer months. The temperature of the samples
in the course of this study ranged from 23°C to 30°C although for most,
the temperature was 24°C. No attempt was made to take thermostatted
readings.
-------
SECTION VI
DATA
The results of the applicability of the cyanide electrode for
measuring free and total cyanide will be reported in three sections:
1. Does the cyanide electrode respond to free cyanide alone?
2. Can the cyanide electrode be used to directly determine
total cyanide content?
3. Can the cyanide electrode be used with the hydrolysis-
distillation procedure to determine total cyanide?
Does the Cyanide Electrode Respond to Free Cyanide Alone?
The data which follow indicate a definite "yes" as an answer to
this question. The specifications supplied by the Orion Company with
the cyanide electrode require work at pH values greater than 9.5 where
all cyanide is in an ionic, unprotenated form. The electrode does not
respond to cyanide when it is in the form of hydrogen cyanide in solu-
tion. The manufacturer makes no statement about the electrode response
to solutions containing metal-cyanide complexes, however.
In considering potential metal-cyanide complexes for study, it was
felt important to choose both stable, kineticly labile complexes, and
stable, kineticly inert complexes. The tetracyanonickelate (II) complex,
Ni(CN)4^, was chosen as an example of a kinetic labile, thermodyanamically
stable complex for this and subsequent studies. The ferrocyanide Fe(CN)£4>
and ferricyanide, Fe(CN)^, ions were chosen for their kinetic inertness
a factor that needed careful consideration if the conclusions of this
study were to be applied to complexes generally. The thermodynamic and
kinetic characteristics of these and other cyanide complexes are given in
Table 2. In this table the thermodynamic stability of the metal-cyanide
complexes can be seen from their very large stability constants. The
kinetic inertness of the iron-cyanide complexes and the kinetic lability
of others is apparent from the rates of ligand exchange given.
-------
TABLE 1. Stability equilibrium constants (Ks) and rates of ligand
exchange for some typical metal-cyanide complexes
Complex
Ni(CN)42
Mn(CN)63
-4
Fe(CN)6
Fe(CN)63
Hg(CN)42
Ks
io30
io27
37
10
io44
io42
Exchange rate
Very fast
Measurable
Very slow
Very slow
Very fast
-4
TV.» -,AU >roi.3), both uncorrected for sodium error, were chosen for sub-
sequent measurements. The response of the cyanide electrode to 1.00 x
10"^M_ (2.6 ppm) cyanide at these pH values were -157.5 millivolt at pH
= 12.7 and -158.4 millivolt at pH = 9.8. The 9.8 value represents a
near lower limit for the pH before protonation begins to appreciably re-
duce the syanide concentration. The good agreement of the millivolt
readings experimentally verifies that essentially all cyanide is in the
non-protenated form at these pH values.
The effect of the response of the cyanide electrode to the addition
of various cyanide-metal complexes to a 1.00 x 10"4M (2.6 ppm) cyanide
solution was investigated. Specifically, reagent grade ferrocyanide,
reagent grade ferricyanide and the tetracyanonickelate (II) ion (prepared
according to the procedure in Inorganic Synthesis) were used. In Table 2
are some typical results.
-------
TABLE 2. Response of the cyanidfi electrode to metal-cyanide complexes
at different pH values
p_H _ Relative millivolts*
pH 9.8
1.00 x 10~4M KCN -62.2
100 ml 1 x 10~4M KCN + .lg KFe(CN) -57.2
100 ml 1 x 10"4M KCN + .lg K4Ke(CN)6 -61.8
100 ml 1 x 10"4M KCN + .lg K2Ni(CN)4 -62.0
pH 12.70
1.00 x 10"4M KCN -62.0
100 ml 1 x 10"4M KCN + .lg K3Fe(CN)6 -61.1
100 ml 1 x 10~4M KCN -«• .lg K4?e(CN)6 -61.3
100 ml 1 x 10"4M KCN * .lg K2Ni(CN)4 -6?.. 8
*In this table and most succeeding tables relative millivolt units are
used since it is relative changes that are of interest.
-4
In all cases addition of cyanide complexes to 10 M (2.6 ppm) CN"
solutions result in an increase in cyanide potential implying a decrease
in the cyanide concentration as sensed by the cyanide electrode. While
all decreases are very small and are nearly within experimental error,
these decreases could be resulting from cyanide in solution complexing
with uncomplexed metal impurities in the metal-cyanide complexes used.
If the cyanide added in the above complexes were free, a cyanide molarity
of about 2.5 x 10~2M (650 ppm) would be realized. This is a 250-fold
increase over the 1 x 10"*M_ (2.6 ppm) cyanide present in the original
solution. It thus appears that the cyanide electrode does not respond to
complexed cyanide, but only the free and uncomplexed c>anide.
The lack of an increase in cyanide in these solutions might be due
to the kinetic inertness of these complexes. For example, while the
tetracyanonickelate is very labile with an exchange half-life of only 30
seconds, the ferrocyanide, with considerable ligand field stabilization
energy, has an exchange half-life of about 25 days. To better evaluate
the response of the electrode to cyanide complexes without the compli-
cation of kinetic effects, additions of aqueous solutions of metal ions
-------
into cyanide solutions and the response of the cyanide electrode to
indicate formation of cyanide complexes was observed. The responses of
the cyanide electrode to additions of 1.0 x 10~3M (about 60 ppm for
first transition series metals) aqueous metal ions to 100 ml of l.uO
x 10"4W[ (2.6 ppm) cyanide were made at two pH values. The results are
given in Table 3.
TABLE 3. Relative response in millivolts of the cyanide electrode to
successive 1 ml additions of 1.0 x lO-^M metal ions to 100.0 ml
of a 1.0 x 10"^ (2.6 ppra) cyanide solution
1 x 10"4M CN~
(100 ml)
+ 1 ml
metal ion
+2 ml
metal
+ 3 ml
metal
+4 ml
metal
at pH = 9.8
Fe*3
Fe*2
Ni*2
Cu+2
Ba*2
s/2
Zn+2
Pb*2
Mn*2
-63.0
-61.0
-63.0
-61.2
-61.6
-61.8
-62.3
-61.5
-62.6
-60.5
-56.2
-52.3
-53.4
-60.8
-61.0
-58.9
-59.8
-61.2
-54.8
-51.8
-36.5
-45.3
-59.6
-60.4
-55.0
-57.6*
-60.5
-44.4
-46.0
- 5.6
-34.1
-58.8
-59.8
-51.9
-55.0*
-28.7
-37.8
—
-18.6
-58.1
-59.4
—
at pH = 12.70
Fe*3
Fe*2
Ni*2
Cu*2
-61.2
-60.7
-60.5
-61.6
-59.8
-57.2
-56.1
-58.5
-58.3
-56.2
-47.8
-55.7
-57.6
-55.3
-35.5
—
-56.9
—
-19.2
-46.7
*Precipitate forms
10
-------
These data at pH 9.8 show that Ni and Cu quite strongly complex
with cyanide removing it from solution as far as the cyanide electrode
response is concerned. For a 3 x 10~5M Ni*2 concentration (1.76 ppm),
the free cyanide concentration has been reduced from 1.00 x 10"4M (2.6 ppm)
to 1.1 x 10~SM (.29 ppm). This implies that each Ni*2 has complexed with
3 cyanide ions, or that three out of every four nickel ions exists as a
tetracyano complex. Again, it is very clear that the cyanide electrode is
NOT responding to complexed cyanide. Moreover, at a pH of 9.8 it appears
that 1 ppm iron (II), and 2 ppm iron (III) complex with cyanide to a
significant level, while zinc (II) and lead (II) show yet smaller effects.
Barium and strontium ions which could not conceivably complex with cyanide
show the slight positive shift in potential observed earlier for addition
of metal complexes. This small positive shift is possibly due to ionic
effects or the presence of metal ion impurities. Dilution cannot account
for the magnitude of decrease observed.
At a pH of 12.70 the addition of nickel shows a significant reduc-
tion in the free cyanide level. Copper (II) shows a smaller, but definite
effect, while Fe (II) and Fe (III) may exhibit a very small effect. Only
by working at a pH value of 14 did the nickel-cyanide complex prove un-
staole:
1 x 10"4M CN-
(100 ml)
-143 millivolts
+1 ml 10"3M
• (Ni*2)
-146
+ 2 ml
-144
+ 3 ml
-142
+4 ml
-140
At this pH, however, the data had larger errors because of larger fluctu-
ations in the digital readings.
These data demonstrate that the cyanide electrode does not respond to
complexed cyanide, but only free cyanide ions. It must be kept in mind
that the readings must be taken at pH values of about 10 and above and that
caution must be taken in relating a free cyanide reading at a pH of 10 to
that in the sample at pH of 7. As will be shown later in this report, how-
ever, the free cyanide concentration in natural water samples with its usu-
al range of metal contaminents is probably nearly the same at pH = 10 as
pH = 7.
Can the Cyanide Electrode Be Used to Directly Determine the Total Cyanide
Content?
In the previous section the cyanide electrode was shown to respond to
free cyanide only. Since the classical hydrolysis-distillation procedure
for determination of total cyanide is a time consuming (at least one hour
of distillation) one, it was considered profitable to investigate the
possible use of the electrode for determining the total cyanide content of
solutions directly without prior distillation of hydrogen cyanide from
solution. To measure total cyanide a complexing agent capable of decom-
posing the metal-cyanide complexes would be necessary. The two agents
11
-------
investigated in this study were hydroxide and ethylenediaminetetraacetic
acid (EDTA). The pH dependence of the formation of the metal-cyanide
complexes is clearly evident in the previous discussion. The question
here is not one of the formation of the complexes, but rather, one of
their decomposition. Their demonstrated thermodynamic instability at
pH = 13 (except nickel) does not necessarily require their kinetic in-
stability. The stability of the ferrocyanide and ferricyanide as a
function of pH is shown in Table 4.
TABLE 4. Relative millivolt readings
After heating
p_H Arter mixing 1 day later 1 hour
2 x 10"5M Fe(CN)"3 at:
— o
pH 7 250 205 210
pH 10 156 198 181
pH 11 131 139 120
pH 12 136 133 106
pH 13 134 131 109
2 x 1C"5M Fe(CN)"4 at:
— o
pH 10 177 --- 118
pH 12 206 --- 61
pH 13 227 132
If all of the cyanide had been released to the solution as free
cyanide a relative millivolt reading of -10 millivolts would have been
realized. These data demonstrate the tremendous kinetic stability of
the iron-cyanide complexes. After two days at room temperature and one
hour under reflux, less than 1% of the ferrocyanide-cyanide was re-
leased and less than 10% of the ferrocyanide-cyanide. These results
imply that distillation times of 24 hours or lorger would be required
to obtain near quantitative release of cyanide from iron-cyanide com-
plexes. A procedure requiring distillation times of this magnitude
obviously represents no improvement over the classical acid-hydrolysis-
distillation procedure. Further investigation of hydroxide-decom-
position of metal-cyanide was dropped at this point although additional
12
-------
work ought to be carried out to find conditions or catalyst to facilitate
this decomposition.
EDTA, an excellent chelating agent for most metals, was next investi-
gated as a potential general agent for decomposing metal-cyanide com-
plexes and freeing all cyanide. It was anticipated that the EDTA would
complex with metal ions in solution, preventing metal-cyanide complex
formation and thus giving cyanide readings corresponding to the total
cyanide concentration in the solution. In order to evaluate whether
EDTA formed metal complexes more stable than the metal-cyanide complexes,
addition of 1.0 ml portions of 1.0 x 10-3M (58 ppm) Ni+2 to 100 ml of
1.00 x 10"4M (2.6 ppm) cyanide containing 1.00 ml of 1.10M disodium EDTA
were made at pH = 13. The results are given in Table 5.
TABLE 5. Relative response in millivolts of the cyanide electrode to
the addition of nickel (II) and copper (II) to solutions con-
taining both EDTA and cyanide (pH 13 j
ml 10'3-
Ni*2 added
0.0 ml
1.0
2.0
3.0
4.0
10" MCN-
without
EDTA
-60.5 mVolt
-56.1
-47.8
-35.5
-19.2
10"4MCN-
with
EDTA
-58.0 mVolt
-55.0
-54.0
-52.0
-51.0
ml 10"3M
Cu+2 ~
added
0.0
1.0
2.0
3.0
4.0
10"4M CN-
without
EDTA
-61.6
-58.5
-55.7
-46.7
10 M CN~
with
EDTA
-59.0
-56.0
-54.0
-52.0
-51.0
These results show that EDTA appears to have some effect, although
by no means complete, in blocking the interferences of both copper and
nickel, the two metals found to form the cyanide complexes most stable
in strongly basic solutions. This implies that the metal-EDTA complexes
are slightly more stable than the metal-cyanide complexes and that EDTA,
thus, might dissociate the metal-cyanide complexes to allow the cyanide
electrode to measure total cyanide in solutions.
Again the reverse experiment is the relevent one. While the above
data demonstrate the EDTA-metal complexes to be more stable, the im-
portant question is whether EDTA will kinetically decompose the metal-
cyanide complexes to release complexed cyanide. To answer this question
the following experiments were run with ferrocyanide and tetracyanonic-
kelate (II): to 100 ml of 1.00 x 10"^ (2.6 ppm) cyanide were added
successive 1.00 ml increments of 1.0 x lO"3^ (about 60 ppm) metal ions.
Reduction of the free cyanide concentration was observed as metal-
cyanide complexes formed as reported earlier in the first section. After
waiting until metal-cyanide complex formation was complete, 1.00 ml of
13
-------
0.10M disodium EDTA was added and the response of the cyanide electrode
to cyanide as the metal-complexes decomposed was observed. Data for
nickel are given in Table 6.
TABLE 6. Relative response in millivolts of the cyanide electrode to
the dissociation of the nickel-cyanide complex by EDTA at
pH 13
Solution
100.0 ml 1.00 x 10"4M CN~
+1.00 ml 1.0 x 10"3M Ni*2
+2.00 ml Ni*2
+3.00 ml Ni*2
+4.00 ml Ni*2
+5.00 ml Ni*2
Above solution after 20 minutes
Ab~'ve solution after 1 hour
Above solution after 1 1/2 hours
Solution + 1.00 ml 0.10M EDTA
Above solution after 1/2 hour
Above solution after 21 hours
Above solution after 26 hours
After 47 hours and 1 hour reflux
Relative millivolt reading
0.0
3.6
6.7
10.8
17..,
24.0
29.5
45.2
50.0
47.5
45.0
33.0
33.0
3.3
Since the electrode gives nearly the same response at the end of
the experiment as at the beginning, these data show that EDTA will give
virtually quantitative decomposition of the nickel-cyanide. Refluxing,
however, is necessary to effect this decomposition. The unexpected
kinetic stability of the very labile nickel-cyanide complex renders
EDTA decomposition of inert iron-complexes an unlikely prospect. A
similar study to the above, using iron (II) is place of nickel (II)
seemed to confirm the very slow iron (Il)-cyanide complex decomposition,
14
-------
but since the iron (II)-cyanide stability is less than that for nickel
the magnitute of the changes were much smaller and thus more difficult
from which to draw conclusions.
At this point, neither hydroxide nor EDTA have proven to be
effective in quickly decomposing metal-cyanides to enable direct
electrode measurement of total cyanide in solution. While these ex-
periments demonstrate the electrodes inability to simply measure total
cyanide, they also dramatically show that the free-cyanide response of
the cyanide electrode is to a large degree, pH independent. If the
free-cyanide level in a natural water sample at a pH of 7 is needed,
adjusting the pH to 10 will have little change in the free-cyanide
level. The cyanide electrode, thus, represents a means of accurately
assessing the free cyanide concentration in natural water samples.
The high thermodynamic stability of nickel and copper cyanide com-
plexes renders them stable at this pH whereas the kinetic stability of
the iron-cyanide complexes prevents their dissociation in the time re-
quired to make a reading.
Can the Cyanide Electrode Be Used with the Hydrolysis-Distillation
Procedure to Determine Total Cyanide?
The inability to find a chemical agent for decomposition of metal-
cyanide complexes led to studies of the classical hydrolysis-distilj.a-
tion procedure for total cyanide determination. It was felt that the
cyanide electrode would at least supply a faster and simpler method for
analysis of cyanide after distillation than the spectrophotometric or
titration procedures usually employed. Moreover, the remarkable
stability of the metal-cyanide complexes observed to this point led to
questions of their stability in acid solutions. Specifically of
interest was the question of whether these complexes were completely
decomposed in the acid-hydrolysis-distillation procedure used in cyanide
determinations (Methods for Chemical Analysis of Water and Wastes, 1971,
Environmental Protection Agency, pages 41-52).
The reliability of the method was first checked using potassium
cyanide solutions. In the procedure used, 50.0 ml of a cyanide solution
was nydrolyzed and pulled by aspirator vacuum through a sintered glass
bubbler into 50.0 mi of a 0.1 M NaOH solution. The cyanide concentrations
of the solutions before and after hydrolysis thus should be the same if
the apparatus is working correctly. Verification of this is seen in
Table 7.
15
-------
TABLE 7. Response of the cyanide electrode to the cyanide concentrations
in solutions resulting from acid hydrolysis-distillation of
solutions of KCN and metal-cyanide complexes
Determination number
Acid hydrolysis-distillation of KCN solutions
1 2 3
Cyanide millivoltage
before distillation
Cyanide millivoltage
after distillation
CN~ cone.
-253.0
-250.0
lxlO"4M
(2.6 ppm)
-214.0
-217.0
3xlO"5M
-281.0
-270.0
2.5xlO~4M
(.78 ppm) (6.5 ppm)
When solutions of accurately weighed cyanide complexes were treated
by acid hydrolysis to investigate their stability under acidic conditions,
the results were not consistent, however. Typical results can be seen in
Table 8.
TABLE 8. Acid hydrolysis-distillation of metal-cyanide complexes
K-Fe(CN), K.Fe(CN),3H 0 K.Ni(CN) .2H .0
OD 462 242
CN- exp.
found
2.51xlO"4M
1.26xlO"4M
0.95xlO"4M
1.55xlO"4M
1.01xlO~4M
CN' exp. CN' exp.
CN~ calc. found CN' calc. found CN" calc.
.99xlO'4M .89X10'4 1.36xl(T4 0.8xlO'4M l-21xlO'4M
1.16xlO"4M --- --- 0.75xlO"4M 1.44xlO"4M
1.46xlO"4M --- --- 0.31xlO~4M 1.44xlO~4M
1.15xlO"4M
1.15xlO"4M
Generally the tetracyanonickelate complex gave results consistently
lower in cyanide than it should have. The ferricyanide results were
spurious once giving 2.5 times as much as experimentally calculated, and
once only 75%. The difficulty in making accurate measurements of this
type lies in the fact that a shift in pCN from 4.0 to 4.1 means a change
from 1.0 x 10'4 to 0.8 x 10'4. Thus, a 2.5% error in the potential
16
-------
reading in this case is reflected as a 20% error in cyanide concen-
tration. The daily shift in the working millivolt versus pCN curve is
frequently greater than 2.5% introducing significant errors. The lack
of consistent air conditioning during the summer months has also in-
troduced errors by greater than desired temperature fluctuations.
Nonetheless, there is some evidence that the presence of nickel will
cause low cyanide results in acid hydrolysis determinations.
To reduce experimental error and observe the interference of metal
ions in the acid hydrolysis determination of cyanide the following
determinations were made. The cyanide electrode response to a known
solution containing cyanide was measured, additions of metal ions to
the solution were made, and the solution immediately acid hydrolysed
and distilled to release the cyanide into an identical volume of
0.10M NaOH. The response of the cyanide electrode to this final
solution was then made. Since the initial and final solution volumes
are kept the same, and since the readings are made close in time to
each other, the interference of the metal ions should be reflected with
minimum experimental error. Results of these determinations with 0.1 g
samples of nickel (II), copper (II) and iron (II) nitrates added are
shown in Table 9.
TABLE 9. Effect of metal ion on the recovery of cyanide with acid
hydrolysis-distillation
Item Run #1 Run »2 Run *3 Run 04 Run *5
Millivolt before
adding metals -165 -165 -100 -253 -271
Millivolts after
adding metals § -155 -153 - 83 -235 -260
distilling
CN- Molarity IxlO"4 IxlO"4 IxlO*5 IxlO"4 2.5xlO~4
(2.6 ppm) (2.6 ppm) (.26 ppm) (2.6 ppm) (6.5 ppm)
These data clearly show that metal ions will cause low results in
the acid hydrolysis-distillation determination of cyanide. The results
nearly always are 40% to 50% low. Extended distillation times on several
samples failed to release significantly larger cyanide concentrations.
At this point it appears that metal ions cause low cyanide results
even when the acid-hyJlrolysis-distillation procedure is used. While it
is true that the 250 ppm metal ion concentration used in the last sequence
of determinations is very high, the data demonstrate an effect that is
very important and needs further study. At this point it appears that
even the classical acid-hydrolysis, distillation procedure does not always
17
-------
yield the total cyanide levels in samples with large metal-ion concen-
tration.
One of the principle purposes of this work was to check the response
of the cyanide electrode with natural water samples as well as laboratory-
prepared samples. The original intention was to use samples collected by
the Miami Conservancy District and arrangements for this had been made.
Unfortunately, due to cut-backs in financial support, the Miami Conser-
vancy District reduced their sample collection from daily to monthly.
While they would have supplied us with samples, their sampling schedule
did not agree with our needs. Samples were obtained instead from the
Springfield Water Pollution Control. They supplied us with three water
samples: one from the Mad River above Springfield, one from the Mad
River below Springfield, and one from Buck Creek in Springfield. They
also supplied us with three frozen fish which died in a large fish kill
in the Mad River in Springfield. An industry using pickling solutions
with high cyanide as well as nickel and copper ion concentrations was the
principle suspect in the kill. We were asked to determine cyanide levels
in the fish using our cyanide electrode.
While the meaning and significance of the following results of
natural samples are not always clear, they are presented to initiate a
realistic evaluation of some of the limitations of the cyanide electrode
in measuring cyanide concentrations. Readings were made on water samples
with the cyanide electrode in two ways. With the first method, the
solution was adjusted to a pH of 13 and a reading taken immediately with
the cyanide electrode. According to the preceding arguments, this read-
ing should give some measure of the free-cyanide content of the water.
Some evaluation of the total cyanide was obtained in the second approach
using the hydrolysis-distillation method and subsequent measurement of
the cyanide concentration in the solution obtained. In addition, direct
measurements with EDTA addition were made even though previous investi-
gations indicate the problems with this approach. Results for the Buck
Creek water sample and downstream Mad River samples are given in Tables
10, 11, and 12.
18
-------
TABLE 10. Response of the cyanide electrode to the cyanide concen-
tration in several natural samples using various analysis
methods
Sample
Buck Creek
Downstream Mad River
Direct elect.
determination
4.2xlO"7M
(11 ppb)
5.6xlO"?
(14 ppb)
Cyanide Molarity
Direct det. Hydro 1 . dist.
EDTA addition determination
1.5xlO"6M 1.58xlO"?-
(39 ppb) (4.1 ppb)
l.SxlO"6 6.3xlO~6
(39 ppb) (163 ppb)
TABLE 11. Electrode response to solutions spiked with additional
cyanide (increased by SxlO'^M)
Cyanide Molarity
Sample
Buck Creek
Mad River
Direct
electrode
response
2.2xlO~5M
(.57 ppmj
3.2xlO"5
(.83 ppm)
Direct
electrode
response
with EDTA
2.2xlO"5M
(.57 ppmj
3.2xlO"5
(.83 ppm)
Electrode response to
cyanide solution obtained
after hydrolysis-dis-
tillation
3.2xlO"5M
(.83 ppmT
3.5xlO"5
(.91 ppm)
TABLE 12. Analysis of fish from fish kill
Types of fish
CN'M obtained
ppm CN" in fish
Small fish
Large fish
Crayfish
l.lxlO~4M
1.2xlO"5M
7.94xlO"6M
2.4
0.30
0.15
19
-------
The upstream Mad River sample gave cyanide concentrations even less
than the figures in Table 10. Because these readings were well beyond
the region of Nernstian behavior of the electrode and because considerable
drifting occurred before a reading could be taken their validity is very
questionable. Even the results in the table are at the extreme limit of
application of the cyanide electrode and contain greater errors than
previous readings. The data do indicate, however, that addition of EDTA
and heating does release additional cyanide. The result of the hydrol-
ysis-distillation determination of cyanide in Buck Creek is suspicious
since it is less than that found by direct electrode measurement of the
original sample. Only large errors in these readings can explain this
inconsistency.
To measure the recovery of cyanide from these natural water samples
the Buck Creek and downstream Mad River samples were spiked with cyanide
to increase the cyanide concentration by SxlO'^M (1.3 ppm). The spiked
solutions were then analyzed for cyanide by direct electrode measurement
and by the hydrolysis-distillation procedure, followed by electrode
measurement. The results are given in Table 11. In no case was all the
added cyanide recovered, although the hydrolysis-distillation procedure
gave 64% and 70% recovery. Again, it seems as if the classical cyanide
procedure gives low results when contaminents are present. The reason
for the low results by direct electrode response can be rationalized as
a result of aqueous metal-ions in the samples complexing with the
cyanide added. The low results for the hydrolysis-distillation while
not easily explained are consistent with previous determinations made in
this study. It is very clear from the data for the spiked solutions that
the known addition method of analysis is entirely unapplicable to
cyanide determinations.
The fact that the direct electrode measurements give lower cyanide
concentrations is consistent with the earlier argument that these read-
ings are measuring free and not total cyanide.
In the final study using the cyanide-electrode the cyanide content
of two fish and one crayfish which died in a large fish kill in Spring-
field was determined. Each fish was liquified using a Waring blender.
No attempt was made to measure the cyanide level in the resulting
solution since it contained enormous concentrations of organic con-
taminants. Rather, the solution was subjected to the acid hydrolysis-
distillation procedure and the cyanide concentration of the solution
obtained determined with the cyanide electrode. The results are given
in Table 12. The small size of the one fish precluded confirmation of
the high cyanide concentration. The cyanide concentrations in the cray-
fish and large fish and reasonable copper and nickel ion concentrations
in the fish as determined by atomic absorption led the Springfield Water
Pollution Control to doubt a pickling solution to be the cause of the fish
kill.
20
------- |