PB84-112663
Development and Evaluation of Procedures far the
Analysis of Simple Cyanides, Total Cyanide, and
in Watep and Wastewater
SRI International
Menlo Park, CA
Prepaied for
Environmenfeal;MonitoriRg and Support
Lab.-Cincinnati, Of!
Ocb 83
U.S. Department
National
-------
EPA-600/4-83-054
October 1983
DEVELOPMENT AND EVALUATION OF PROCEDURES
FOR THE ANALYSIS OF SIMPLE CYANIDES,
TOTAL CYANIDE, AND THIOCYANATE IN
WATER AND WASTEWATER
by
D. Ingersoll, W. R. Harris, D. C. Bomberger, and D. M. Coulson
SRI International
Menlo Park, California 94G25
EPA Contract No. 68-03-2714
Project Officer
Gerald D. McKee
Environmental Monitoring and Support Laboratory
26 West St. Clair Avenue
Cincinnati, Ohio 45268
Environmental Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing}
1. REPORT NO.
EPA-600/4-83-054
3. RECIPIENT'S ACC§SSJOJSj C*O? 97
PHI fa £, i £($>'*& 'S>
4. T
TJevet'opmenY'ind Evaluation of Procedures for the
Analysis of Simple Cyanides, Total Cyanide, and
Thiocyanate in Water and Wastewater
5. REPORT DATE
October 1983
6. PERFORMING ORGANIZATION CODE
7. AUTHORIS)
D. Ingersoll, W. R. Harris, D. C. Bomberger, and
D. M. Coulson
B. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
SRI International
333 Ravenswood Avenue
Menlo Park, California 94025
10. PROGRAM ELEMENT NO.
CBCBIC
11. CONTRACT/GRANT NO.
EPA No. 68-03-2714
12. SPONSORING AGENCY NAME AND ADDRESS
Environmental Monitoring and Support Laboratory
U.S. Environmental Protection Agency
26 W. St. Clair Street
Cincinnati, OH 45268
13. TYPE OF REPORT AND PERIOD COVERED
Proj.Rep. .Sept 73 to Apr 80
14. SPONSORING AGENCY CODE
EPA-600/06
£-•• .-_ = ..:;NTARY NOTES
16.
Seven methods for the analysis of simple cyanides have been investigated.
Included are (1) an ion-exchange procedure, (2) a continuous-flow distilla-
tion, (3) and EDTA electrode method, (4) the American Iron and Steel Institute
aeration method, (5) an EDTA aeration method, (6) the modified Roberts-Jackson
method, and (7) the EPA procedure for Cyanides Amenable to Chiorination.
Of all of the procedures studied, the modified Roberts-Jackson method is the
best. It gives complete recovery from all but one of the simple cyanides
without decomposing the complex cyanides.
In addition to these methods for simple cyanides the EPA procedure has been
evaluated for the analysis of total cyanide. Procedures using ligand-exchange
and high temperature distillation have been developed and evaluated for analysis
of total cyanides.
Colorimetric high performance liquid chromatography and atomic absorption
spectrophotometric methods for the analysis of thiocyanates were investi-
gated. All these methods are based on the formation and extraction of a
thiocyanate-pyridine-copper (II) complex. ;
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
18. DISTRIBUTION STATEMENT
Release to public
19. SECURITY CLASS (Tliii Ktporl)
Unclassified
20. SECURITY CLASS (Tliispa.ne/
Unclassified
EPA Form 2220-1 (R»». 4 — 77) PREVIOUS EDITIOM is o BSOLETE
i
-------
NOTICE
This document has been reviewed in accordance with
U.S. Environmental Protection Agency policy and
approved for publication. Mention of trade names
or commercial products does not constitute endorse-
ment or recommendation for use.
11
-------
TABLES
Number Page
4.1 Solubilities of Metal-Cyanide'Compounds 6
4.2 Stability Constants of Metal-Cyanide Compounds 7
4.3 Compounds Studied as Potential Interferences 7
4.4 Logarithms of the Stability Constants of Cations with Various
Ligands 13
4.5 Cyanide Recoveries Obtained with the Ligand-Exchange Method. ... 14
4.6 Effects of Potential Interferences on Cyanide Recoveries Using
the Ligand-Exchange Procedure with 1m Ton~Selective" ETectfode Finish 15
4.7 Cyanide Recoveries Obtained with the EPA Total Cyanide Method
Using an Ion-Selective Electrode Finish 18
4.8 Effects of Potential Interferences on Cyanide Recoveries Obtained
with the EPA Total Cyanide Method With an Ion-Selective Electrode
Finish 20
4.9 Comparison of Cyanide Recoveries Obtained with the EPA and
Ligand-Exchange Total Cyanide Procedures 22
4.10 Comparison of Interferences on the EPA and Ligand-Exchange Total
Cyanide Procedures 23
5.1 _CqraiDounds Jncluded in Studv ,_^ . . . . ^ ... 24
5.2 Cyanide Recoveries Obtained with the Continuous-Flow Distillation
Procedure 30
5.3 Cyanide Recoveries Obtained with the EDTA-Electrode Procedure. . . 31
5.4 Summary of Results from the EDTA-Electrode Procedure 32
5.5 Cyanide Recoveries Obtained with the AISI Aeration Procedure ... 35
5.6 Cyanide Recoveries Obtained with the EDTA Aeration Method 41
5.7 Effects of Potential Interferences on Cyanide Recoveries Obtained
with EDTA Aeration Procedure 43
5.8 Cyanide Recoveries Obtained with EPA Method "Cyanides Amenable to
Chlorination" Using an Ion-Selective Electrode 45
Pages iii and iv are blank.
-------
Number Page
5.9 Standard Reduction Potentials of Chlorinating Species 46
5.10 Effects of Potential Interferences on Cyanide Recoveries Obtained
with the EPA Procedure "Cyanides Amenable to Chlorination" Using
an Ion-Selective Electrode Finish 49
5.11 Cyanide Recoveries Obtained with the Modified Roberts-Jackson
Procedure 51
5.12 Effects of Potential Interferences on Cyanide Recoveries with
the Modified Roberts-Jackson Procedure 52
5.13 Comparison of Cyanide Recoveries from the Simple Cyanide Methods. 55
5.14 Comparison of Interference Effects on the Simple Cyanide Methods. 56
5.15 Comparison of Various Operating Parameters for the Simple
Cyanide Methods 57
6.1 Extraction of Copper-Pyridine Solutions with Chloroform in the
Absence of Thiocyanate 70
6.2 Extraction Efficiency of Dithiocyanatodipyridylcopper(II) by
Chloroform 72
6.3 Extraction of Cupric Ion by Pyridyl Ligands in the Absence of
Thiocyanate 73
6.4 Copper Extracted by Aliphatic Amines at pH 7 74
VL
-------
CONTENTS
Abstract iii
Figures iv
Tables v
Contents vii
Acknowledgments x
1. Introduction 1
2. Conclusions 2
3. Recommendations -. . . . 5
4. Total Cyanide Methodology 6
4.1 Introduction 6
4.2 High Temperature Distillation 8
4.2.1 Introduction and Background 8
4.2.2 Procedure 8
4.2.3 Results and Discussion 8
4.2.3.1 Cyanide Recoveries 8
4.2.3.2 Interferences 10
4.2.4 Conclusions 10
4.2.5 Recommendations 10
4.3 Ligand-Exchange Total Cyanide Method 12
4.3.1 Introduction and Background 12
4.3.2 Results and Discussion 15
4.3.2.1 Cyanide Recovery 15
4.3.2.2 Interferences 15
4.3.3 Conclusions 16
4.3.4 Recommendations 16
4.4 EPA Total Cyanide Method 17
4.4.1 Introduction and Background 17
4.4.2 Results and Discussion 17
4.4.2.1 Cyanide Recoveries 17
4.4.2.2 Interferences 19
4.4.3 Conclusion 21
4.4.4 Recommendations 21
4.5 Comparison and Summary of Total Cyanide Methodology . 21
5. Simple Cyanides Methodology 24
5.1 Introduction 24
5.2 Ion Exchange Procedure 25
5.2.1 Introduction and Background 25
5.2.2 Procedure 25
5.2.3 Results and Discussion 27
5.2.4 Conclusions 27
5.2.5 Recommendations 28
5.3 Continuous-Flow Distillation 28
5.3.1 Introduction and Background 28
5.3.2 Procedure 28
vii
-------
5.3.3 Results and Discussion 28
5.3.4 Conclusions 30
5.3.5 Recommendations 30
5.4 EDTA Electrode Procedure 31
5.4.1 Introduction and Background 31
5.4.2 Procedure 31
5.4.3 Results and Discussion 31
5.4.4 Conclusions 33
5.4.5 Recommendations 33
5.5 AISI Aeration Procedure 33
5.5.1 Introduction and Background 33
5.5.2 Procedure 33
5.5.3 Results and Discussion 35
5.5.4 Conclusion 36
5.5.5 Recommendations 40
5.6 EDTA Aeration Procedure 40
5.6.1 Introduction and Background 40
5.6.2 Results and Discussion 40
5.6.2.1 Cyanide Recoveries 40
5.6.2.2 Interferences 42
5.6.3 Conclusions 42
5.6.4 Recommendations 44
5.7 EPA Procedure for Cyanides Amenable to Chlorination . 44
5.7.1 Introduction and Background 44
5.7.2 Results and Discussion 44
5.7.2.1 Cyanide Recoveries 44
5.7.2.2 Interferences 48
5.7.3 Conclusions 48
5.7.4 Recommendations 48
5.8 Modified Roberts-Jackson Procedure 48
5.8.1 Introduction and Background 48
5.8.2 Results and Discussion 50
5.8.2.1 Cyanide Recoveries 50
5.8.2.2 Interferences 50
5.8.3 Conclusions 53
5.8.4 Recommendations 53
5.9 Comparison of Simple Cyanide Methods 54
6. Thiocyanate Methodology 58
6.1 Introduction 58
6.2 Colorimetric Methods of Analysis 58
6.3 High Performance Liquid Chromatography (HPLC) .... 61
6.4 Atomic Absorption Spectroscopy 69
6.5 Conclusions and Recommendations 74
References 76
Appendices
A. EDTA Aeration Procedure for Simple Cyanide A-l
B. Modified Roberts-Jackson Method for Analysis of Simple
Cyanides B-l
C. EPA Procedure for Analysis of Cyanides Amenable to
Chlorination' C-l
viii
-------
D. Ligand-Exchange Method of Analysis for Total Cyanide. . . . D-l
E. EPA Procedure for Analysis of Total Cyanide E-l
F. Statistical Protocol F-l
LX
-------
ACKNOWLEDGMENTS
This work was supported by Contract No. 68-03-2714 with the U.J3 ._Envirpn-__
mental Protection Agency. We are especially grateful to _Gerald_D_.._McKee...__
Environmental Monitoring and Support Laboratory, for his valuable guidance and
support.
x
-------
SECTION 1
INTRODUCTION
The Federal Water Pollution Control Act as amended after 1972, Section
304(h), requires that the administrator promulgate guidelines establishing
test procedures for the analysis of pollutants. These test procedures must be
applicable to measure a specific pollutant in a wide variety of industrial ef-
fluents. The objectives of this work, were to review the pertinent literature
to determine the technical approaches to be taken for development and evalua-
tion of test procedures for analysis of cyanides and thiocyanates. Two new
methods for total cyanide were developed and evaluated during this study and
are compared with the EPA procedure for total cyanide. Three methods for thio-
cyanate were tested. These procedures depend on the chloroform extraction of
the neutral dithiocyanatodipyridylcopper(II) complex to separate the SCN~ from
the sample matrix. The thiocyanate was quantitated by colorimetry, high per-
formance liquid chromatography, or graphite furnace atomic absorption spectros-
copy. Statistical evaluation of the test results was obtained at several con-
centration levels for each of the methods studied. A description of the statis-
tical protocol used in this study is given in Appendix F.
Six existing procedures for the analysis of simple cyanides were evalu-
ated and compared to the Environmental Protection Agency (EPA) .method
"Cyanides Amenable to Chlorination, Manual of Methods for Chemical Analysis of
Water and Wastes, 1979." Statistical evaluation of the test results was ob-
tained at several concentration levels. The following chapters present and
discuss the development and evaluation of the analytical test procedures used
in this study.
The EPA method referred to is actually a modification of the approved
procedure. Among the modifications described on page 17, paragraph 4.4~.l, is
the use of an electrode finish instead of the titrimetric and colorimetric
finishes. The modified method is presented in Appendix E.
-------
SECTION 2
CONCLUSIONS
A number of analytical test procedures for the analysis of total cya-
nides, simple cyanides, and thiocyanates in water samples have been developed
and/or evaluated. The operating characteristics of each of these methods are
summarized below.
• Total Cyanides
High Temperature Distillation—The primary advantage of this pro-
cedure is its ability to recover cyanide from all of the compounds
studied, including the thermodynamically stable, kinetically inert
K3[Co(CN)6]. The method is subject to a number of interferences
including sulfide and thiocyanate. The lower limit of detection
using this method is estimated to be 6 ppb, higher than any of the
other total cyanide methods studied.
EPA Total Cyanides—A modification of the nrocednrR iispH _
by the EPA for the analysis of total cyanides was evaluated. Com-
plete recoveries of cyanide were possible from all of the compounds
studied except K3[Co(CN)g]. Included in this list are the stable
ferri- and ferrocyanide complexes. The lower limit of detection is
2 ppb ± 1 ppb. The relative standard deviation above 10 ppb is less
than ±10%. The main disadvantage of the procedure is interference
from a number of compounds. Of major concern, are sulfide and thio-
cyanate. Various procedures have been described to alleviate these
problems, but these procedures also introduce additional problems.
As such, this method should be used for the analysis of cyanide in
samples that do not contain either sulfide or thiocyanates.
Ligand-Exchange—A procedure was developed that gives not only
complete recovery of cyanide from all of the compounds studied
except K.3[Co(CN)g], but also is unaffected by the presence of either
sulfide or thiocyanate. This is a major development in the field
of cyanide analysis, since these compounds have historically been a
major problem. Another advantage is a 30% reduction in analysis
time. The lower limit of detection is 2 ppb ± 1 ppb. The relative
standard deviation above 10 ppb is less than ±10%.
• Simple Cyanides
Ion Exchange—An ion exchange procedure for the analysis of simple
cyanides was evaluated and found to be unacceptable. Among
its deficiencies are: (1) the lower limit of detection is well
-------
above 2 ppm, (2) analyses are time consuming, and (3) there is
incomplete recovery of cyanide from the column during the rinsing
operations. Only considerable developmental effort could improve
the performance of this procedure.
EDTA Electrode—Although this is a relatively simple procedure, its
use is limited only to samples that do not contain certain other
interfering compounds. The procedure is able to accurately measure
only easily dissociable cyanides. Only with additional developmen-
tal work could this procedure be made to operate efficiently.
Continuous-Flow Distillation—Although this procedure is fast and
easy to operate, the iron cyanides are partially decomposed, which
leads to spuriously high results. A vast amount of additional work
is required to improve the performance of this method. However,
this procedure could be adapted for use as an analytical method for
total cyanides.
AISI Aeration—The major disadvantage of this procedure is its in-
ability to obtain complete recoveries from the cyanides equivalent
to cyanides amenable to chlorination. Also of concern are its in-
creased lower limit of detection (5 ppb) and lower precision. These
problems are inherent in the technical design of the apparatus and
could probably be improved through modifications of the apparatus
and procedure.
EDIA Aeration—In an effort to take advantage of the beneficial
characteristics of the preceding two methods, an EDTA-Aeration
procedure was developed and evaluated. Incomplete recoveries of
cyanide were found from half of the simple cyanides studied and a lower
limit of detection of 5 ppb. Most significantly, neither thiocya-
nate nor sulfide (as PbS) interfere with an analysis.
EPA Procedure for Cyanides Amenable to Chlorination—The EPA proce-
dure for chlorination and distillation of the samples has been
evaluated in our laboratories. Incomplete recoveries of cyanide were
found from a number of simple cyanide species. These recoveries
are also concentration dependent. Recoveries could probably be
improved by altering the chlorination conditions. The lower limit
of detection is 2 ppb ± 1 ppb. Since the method relies on_the
"hPA~ mddi'f ied~~t6tal "cyrahi~de~lnethod"ologyj~ all of the deficiencies
associated with that method are also applicable here. Because of
its widespread use, additional effort should be expended to improve
the method.
Modified Roberts-Jackson--Incomplete recovery of cyanide is found only
-from—the—me-r-eu-r-y--cyanide-compounds T-he-addit-ion-of—eh-1-or-i-de—i-on-du-p-i-ng
analysis will probably overcome this deficiency. The procedure is
-unaffected, .by-.the-pr-esence-of—e-ithe-r—su-1-f-i-de-o-r—thioc-ya-na-t-e-.—Q-t-her
compounds that interfere are removed before analysis. A lower limit
_of_-2 ppb...+_. l-ppb.-is-pos.s-i-b-1-e-iv-i-th-a-prec-is-ion-o-f—jKlO%-a-bove—1-0-ppb—
-------
• Thiocyanates
Three different analytical techniques for the analysis of thio-
cyante have been investigated—colorimetry, high performance liquid
chroraatography, and atomic absorption spectroscopy. All three
methods rely on the formation of a mixed ligand complex between
copper(II), pyridine, and thiocyanate. The complex is normally
extracted into chloroform, at which point problems develop.
At concentrations below 200 ppb, the high background level
of copper interferes with detection of thiocyanate. Thus, even
though the analytical methods are quite sensitive, the effective
lower limit of detection of thiocyanate is over 100 ppb.
-------
SECTION 3
RECOMMENDATIONS
The following recommendations are made for the analysis of simple cya-
nides, total cyanides, and thiocyanates.
(1) Developmental work on the ligand-exchange method for analysis
of total cyanide should be continued. In particular, the con-
ditions of analysis should be improved and this optimized pro-
cedure should be ruggedized and evaluated on both laboratory
standards and field samples.
(2) If the analysis of those metal-cyanide complexes more stable
than either ferri- or ferrocyanide is needed, additional effort
should be expended on the development of the high temperature
distillation.
(3) Performance of the Roberts-Jackson method of analysis could be
improved with minor modification. A comprehensive evaluation of
the method should be conducted to determine the effect of chloride
ion on cyanide recoveries and interferences. On this basis, a pro-
cedure should be optimized and evaluated on laboratory-prepared
samples.
(4) Recoveries from the EPA procedure "Cyanides Amenable to
Chlorination" might be improved by modifying the chlorination
procedure. A study should be undertaken to determine the most
efficient chlorination conditions and a modified procedure should
be developed.
(5) We feel that the lower limits of detection for cyanide could be
lowered by one order of magnitude by adjusting the pH of the solu-
tion in the electrode finish. The optimum pH should be determined
and then used in all the electrode procedures.
(6) To accurately assess the performance of the more effective total
cyanide and simple cyanide methods, it is necessary to conduct a
limited sampling and analysis schedule of industrial effluents.
The sites selected for monitoring should represent widely diver-
gent sources and include sites where cyanide and common interfer-
ences are expected. Some sites that should be sampled include
refineries, coking operations, and coal gasification industries.
(7) The methods evaluated for thiocyanate analysis are unsatisfactory.
A number of recommendations, too numerous to mention here, are given
in Section 6.5.
-------
SECTION 4
TOTAL CYANIDE METHODOLOGY
4.1
INTRODUCTION
Total cyanide refers to all of the CX groups in a sample regardless
of the metal complex, if any, with which the CN~ is associated. There is wide
variation between the chemical characteristics of the different metal-cyanide
complexes included in this study. This variation is readily apparent in
Tables 4.1 and 4.2, which list thermodynamic constants of some of these metal-
cyanide complexes. An ideal method of analysis for total cyanide should assess
cyanide levels in a sample regardless of the metal-cyanide compounds making up
this sample, as well as minimize the effects of interferences.
TABLE 4.1 SOLUBILITIES OF METAL-CYANIDE COMPOUNDS
Compound
AgCN
Cd(CN)2
Co(CN)-
Cu(CN)
Fe(CN)2
Fe(CN)
Hg(CN)
Hs ( CN)
Mn(CN)2
Ni(CN)0
Zn(CN)2
K[Ag(CN)2]
K2[Cd(CN)4]
K3[Cr(CN)6]
K_[Fe(CN) ]
3 6
K [Fe(CN), ]
Name
Silver cyanide
Cadmium cyanide
Cobaltous cyanide
Cuprous cyanide
Ferrous cyanide
Ferric cyanide
Mercuric cyanide
Mercurous cyanide
ManganouH; cyanide
Nickel^, • cyanide
Zinc cyanide
Potassium dicyanoargentate(I)
Potassium tetracyanocadmate(II)
Potassium hexacyanochromate(III)
Potassium hexacyanof errate (III)
(Ferri cyanide)
Potassium hexacyanof errate (II)
(Ferro cyanide)
Solubility
(mole/L)
1.64 x 10~6 (20°C)a
1.51 x 10~5 (18°C)a
3.77 x 10~4 (18°C)a
2.90 x 10~5 (18°C)a
3.68 x 10'1 (20°C)a
1.79 x 10~6 (25°C)a
5.35 x 10~4 (18°C)a
4.90 x 10~5 (18°C)a
•u
1.3 (20°C)
>,
1 (20°C)
1 (4°C)b
0.7 (12°C)b
Reference 1.
Reference 2.
-------
TABLE 4.2 STABILITY CONSTANTS OF METAL-CYANIDE COMPOUNDS
Complex
K4[Mn(CN)6]
K2[Zn(CN)4]
K2[Cd(CN)4]
K[Ag(CN)2]
K3[Cu(CN)4]
K2[Ni(CN)4]
K2[Hg(CN)4]
K4[Fe(CN)&]
Log Stability constant
Name at 25 °C
Potassium hexacyanomanganate(ll)
Potassium tetracyanozincate(II)
Potassium tetracyanocadm.a.te.fjl")
Potassium dicyanoargentate (I)
Potassium tetracyanocuprate(I)
Potassium tetracyanonickelate (II)
Potassium tetracyanomercurate(II)
Potassium hexacyanof errate(II)
9.7a
16. 7a
16. 9a
20. 9b
30. 3a
31. 33
41. la
A7a
K... [Fe(CN) , ] Potassium hexacyanof errata [III)
"
(Ferricyanide)
Potassium hexacyanocobaltate' (III)
(Cobalticyanide)
52'
64C
, Reference 3.
Reference 4.
Comprehensive laboratory evaluations were planned to determine how
closely each method approached ideal behavior. The evaluations included study-
ing the response of the procedure on ten simple, Mi(CX)x, and six complex
MiM2(CN)x, cyanides. Solutions of these compounds were prepared in deionized
_water at four concentration levels; 2, 20, 200, and__2000_ ppb CN~. The ferrous,
'•_fe_r_ric and manganous cyanides, which are not stable as isolated salts, were _
_p_r.epared in solution from potassium cyanide and the appropriate metal chlorid_e
_o.r_s.u]Lfate. _. The effect_of interferences at two levels_f_qr each _inte_r£erence was
•. s.tudied. _The interferences are listed in Table 4.3. If, during preliminary
investigations, a procedure was found to exhibit numerous deficiencies, further
work on that procedure was curtailed.
TABLE 4.3 COMPOUNDS STUDIED AS POTENTIAL INTERFERENCES
Ca(OCl)2
NaN02
Mn02
CH3CH2CH2CHO
Na2S
CH3CH2CH2CH2SCN
KOCN
KSCN
Co(SCN)2
K2[Hg(SCN)J
K3[Co(SCN)6]
The methods developed and evaluated were manual methods useful for the
analysis of water and wastewater samples. The methods have sensitivities of
less than or equal to 2 ppb CN with a precision of ±1 ppb in the range of 2 to
10 ppb and ±10% above 10 ppb CN.
-------
This section describes the results of the evaluations conducted in our
laboratory.
4.2 HIGH TEMPERATURE DISTILLATION
4.2.1 Introduction and Background
The conventional reflux distillation procedures3"7 for the analysis of
total cyanides have gained wide acceptance, but they are not able to achieve
complete recovery of cyanide from the more stable cyanide complexes. For
instance, use of the analytical procedure currently recommended by the Environ-
mental Protection Agency5 gives only a 10% recovery of cyanide from hexacyano
cobaltate(III). To improve recoveries of cyanide from these highly stable and
kinetically inert cyanide compounds, a number of approaches have been devised.
These approaches include fusion with potassium in a nickel bomb,8 the Carius
sealed tube method,9 decomposition at elevated temperatures in a stream of
hydrogen,10 photodecomposition,1L and high temperature distillations.12'13
A procedure based on this last concept was developed for the determination of
total cyanide.
4.2.2 Procedure
After much development work, a procedure was standardized
that was subsequently used in all further laboratory evaluations. The appara-
tus developed is depicted in Figure 4.1. Briefly, the procedure is as follows.
The funnel is charged with 60 rnL of an aqueous mixture of 50% (by volume) of
85% orthophosphoric acid and 4% (by volume) of 50% hypophosphorous acid. To
flask 2 is added 2 mL of 85% orthophosphoric acid and enough water to cover the
tip of the inlet tube. To flask 1 is added 100 mL .of sample or an --
aliquot diluted to 100 "ml The gas scrubber is charged with 10 nil of 1.25 N
sodium hydroxide and enough water to give an adequate depth of liquid. The
apparatus is assembled as diagrammed and a stream of nitrogen is pulled through
the reaction flask at a flow rate of approximately 4 bubbles per second. After
dropwise addition of the acid mixture to the reaction flask, both flasks are
heated to boiling.
As the water from flask 1 is removed, the temperature in the flask
rises to 170°C, where it is maintained for 15 minutes. The heating of flask 1
is discontinued after this time, and the nitrogen flow is allowed to continue
for' an additional 15 minutes. The contents of the scrubber are then quantita-
tively transferred to a 50-mL volumetric flask, and this solution is analyzed
for cyanide.
4.2.3 Results and Discussion
4.2.3.1 Cyanide Recoveries—
Complete recovery of cyanide from KCN, K3Fe(CN)6, and K3Co(CN)g was pos-
sible using this procedure. Although the high temperature distillation procedure
is adequate for the analysis of cyanide from these more stable complexes, it
does have limitations, For example, when^this procedure was designed, the inten-
sion was to use the Ag"*"/S= ion-selective electrode for the analysis of the liber-
ated cyanide. It has been reported by others10 that cyanide could be detected
8
-------
'Funnel
250 ml
Flask 1
Stirring Bar
'Magnetic Stirrers
Gas Scrubber
Equipped with
a Medium
Porosity Frit
Port tor
Thermometer
.250 ml
Flask 2
Stirring Bar
SA-7854-6R
Figure 4.1 High temperature distillation apparatus.
-------
down to 1 ppb using this technique. Replication of their work was possible
with similar results (see Figure 4.2). However, during our investigations, it
was found that when analyzing a series of unknowns of random concentrations, the
electrode does not respond in the manner depicted by the calibration curve.
Because of the slow exchange rate of cyanide for the anions that make up the
pellet of the electrode, and the resulting carryover of cyanide from one solu-
tion to the next, it was not possible to analyze cyanide solutions with any degree
of confidence. Also, the equilibration time required to attain a stable volt-
age reading was often in excess of 15 minutes.
For these reasons use of this electrode was abandoned in favor of the
less sensitive cyanide ion-selective electrode. As a result, the lower limit
of detection of cyanide in a sample is 6 ppb, slightly above the 2 ppb method
development criteria set forth by the Environmental Protection Agency. By re-
designing the system, this limitation could be overcome. However, it was
anticipated that this procedure would exhibit other deficiencies as outlined
in the next section.
4.2.3.2 Interferences—
A major disadvantage of the high temperature distillation procedure is
the number of compcmnds that are expected to interfere, including thiocyanates,
sulfides, organicj^ocyanates,1fatty acids, "and aldehydes, as well as others. Tnis-
is largely due to the very harsh reaction_cpnditions of high temperature and high
acidity. Therefore, work on this procedure was discontinued in favor of tHe more
advantageous liganchexchange method^
4.2.4 Conclusions
The high temperature distillation procedure allows complete recoveries
of cyanide from all the stable complexes. Thus, it is the only method that
can quantitate CN~ bound to cobalt(III). However, since the dissociation of
the cobalt compounds is achieved only by using very harsh reaction conditions,
the method is subject to a number of interferences. It is also likely that
longer analysis times will be necessary to reach a lower limit of detection of
2 ppb. Thus, further development of this method is warranted only if there is
a definite need to detect cyanide from cobalt(III) or other refractory
compounds.
4.2.5 Recommendat ions
Modifications could be made to the procedure and design of the high tem-
perature distillation method that would substantially improve its performance.
For instance, if a moderately acidified solution of a lead, cadmium, or arsenic
salt were substituted for the solution currently used in the second flask, the
interferences from sulfide and thiocyanate could probably be eliminated. Addi-
tional developmental work in this area should yield a procedure for the analysis
of total cyanide that is capable of quantitating all types of cyanide compounds
from a wide variety of sample matrices.
10
-------
-550
-500
-450
UJ
o
>
-400
-350
-300
-250
0.001
CONCENTRATION OF CN~ (ppb)
10 100 1000
0.01 0.1 1.0
CONCENTRATION OF CN~ (ppm)
10,000
10
SA-7854-5R
Figure 4.2 Emf response of Orion's 94-16A Ag4/S ion selective electrode as a function of CM"
concentration.
11
-------
4.3 LIGAND-EXCHAXGE TOTAL CYANIDE METHOD
4.3.1 Introduction and Background
Historically, the procedures most often used for the analysis of total
cyanide have generally employed harsh reaction conditions to ensure complete
recovery of cyanide from the more stable cyanide complexes . s~9 ' 10~12 ' 1/t~18 The
procedure currently recommended by the U.S. Environmental Protection Agency5
consists of a reflux distillation procedure in a highly acidic solution. Using
this procedure, it is possible to attain complete recovery of cyanide from all
but the most stable cyanide complexes. Less than 15% of the cyanide from
K3[Co(CN)g] is recovered.
A major disadvantage of this procedure is that a number of compounds
commonly found in industrial effluents interfere with the analysis. Of major
concern, because of their positive interference and common occurrence in ef-
fluents, are sulfide, S=, and thiocyanate, SCN~. A method that will overcome
these problems is needed to accurately assess effluents for cyanide in the pre-
sence of these compounds. A procedure has been developed that not only is less
susceptible to interferences, but also reduces analysis time by as much as 30%.
The method consists of a one-half hour reflux distillation at moderate pH in
the presence of sequestering agents and lead acetate. The reason for each of
these conditions is discussed below.
_i i_
• Lead Acetate—This is used as a source of Pb . Any sulfide present
in the sample will precipitate as PbS, which has a solubility product of 7 x 10 ,
Since this precipitate is stable in hot water at moderate pH, any S= in the sam-
ple remains in the distillation flask instead of being transferred as H2S to the
scrubber solution and interfering with the various finishes.
• Moderate pH—The PbS precipitate would be appreciably dissociated in
strongly acidic solution, releasing H2S. Thus, the efficiency of this proce-
dure for trapping sulfide is dependent on obtaining satisfactory recoveries of
cyanide from the very stable metal cyanides at moderate pH. Operating at moder-
ate pH also prevents the decomposition of SCN~ to produce S=. This normally
occurs in highly acidic reflux distillation procedures and is responsible for
the large positive interference often found for SCN~. Moreover, the effective-
ness of most sequestering agents increases at higher pH. However, the CN~ must
be protonated for efficient aeration of HCN from the sample matrix. Thus, pH
4.5 is a compromise that allows aeration of HCN without unduly restricting the
sequestering abilities of the ligands.
• Distillation Time—The distillation time for the procedure is one-
half hour. Complete recovery of CN~ from some metal complexes is possible in
only 15 minutes. The half-hour distillation time reduced the extent of decom-
position of PbS compared to the normal one-hour distillation time. This distil-
lation time has the added advantage of substantially reducing the analysis time.
• Sequestering Agents—The ligands are used to aid decomposition of
the cyanide complexes by a ligand-exchange reaction. This displacement can gen-
erally be depicted by the equation:
M(CN)n + yL = M(L) + nCN~ (1)
12
-------
Because the pH is well below the pKa of CN , the cyanide so liberated is quick-
ly transferred as HCN to the sodium hydroxide scrubber solution. The use of a
ligand to displace cyanide from the inner coordination sphere of a metal is not
a new idea. Various procedures for the analysis of simple cyanides that make
use of EDTA and an ion-selective electrode have been reported.19"21 These
methods are entirely unsuited for total cyanide analysis. A procedure for the
analysis of total cyanide that makes use of EDTA has been described.18 However,
under the extremely acidic pH conditions employed in this procedure, EDTA is
largely ineffectual at displacing cyanide. The reported success with this pro-
cedure is no doubt due to the increased distillation temperatures.
It was assumed that complete cyanide recovery would be most difficult
to obtain from the iron, mercury, and cobalt (III) cyanides, since these com-
plexes have the largest stability constants. Thus a. successful ligand must
form very stable complexes with these metal ions. However, the ligand should
not strongly sequester lead, as this would have two detrimental effects. One,
the formation of a lead complex would reduce the effectiveness of the ligand
toward displacing cyanide from other metals. Two, if the complexation of lead
is too strong, the ligand will displace S~ and reduce the protection against
this interference.
The stability constants of several ligands are listed in Table 4.4
Initial results showed that EDTA, EGTA, and CDTA all gave complete recov-
ery of CN from the iron and mercury cyanides. However, they also increased
the interference from sulfide. Thus a more selective ligand was needed. Tiron
has an extremely high selectivity for Fe3+, and the addition of tiron to the
samples results in complete recovery of CN~ from both ferro- and f erricyanide,
but only 60% recovery from H
TABLE 4.4 LOGARITHMS OF THE STABILITY CONSTANTS OF CATIONS
WITH VARIOUS LIGANDS
Ligand
CN~
EDTA3
b
EGTA
CDTAC
TEPd
Q
Tiron
u 2+
Hg
41
21.5
22.9
24.8
27.7
19 .1(1:1)
Cd2+
—
16.4
16.5
19.8
14
= £
13.3(2:1)
Pb2+
—
18
14.5
20.4
10.5
f
15(2:l)r
Fe2+
47
14.3
11.8
18.9
9.9
—
Fe3+
52
25.1
20.5
30.0
__
f
45(3:l)r
Co3+
64
41.4
—
—
S (Ksp) 28 29
Ethylenediamine-N,N,N',N'-tetraacetic acid
2,2'-Oxybis[ethyliminodi(acetic acid)]
£_
trans-Cyclohexane-l,2-diamine-N,N,N',N'-tetra-acetic acid
Tetraethylenepentamine (3,6,9-triazaundecane,1,11-diamine)
1,2-' dihydroxy-3,5-benzenedisulfonic acid
Numbers in parentheses indicate ligand:metal ratio of tiron complex.
13
-------
TABLE 4.5 CYANIDE RECOVERIES USING THE LIGAND EXCHANGE METHOD
Compound
Studied
KCN
Cd(CN)2
Cu(CN)
Ni(CN)2
Hg(CN)2
K3[Cu(CN)lt]
K2lNi(CN)4]
K2[Hg(CN)lt]
K,4[Fe(CN).g]
K3[Fe(CN)6]
K3tCo(CN)G]
Concentration of CN
0.002 ppm
n
6
2
2
2
2
2
2
Mean
Recovery
120
-
-
-
115
124
119
147
147
132
-
Relative
Standard
Deviation
21
-
-
-
-
-
-
0.2 ppm
n
4
4
4
4
4
4
4
4
4
4
4
Mean
Recovery
98
101
99
99
99
99
105
99
104
104
0
Relative
Standard
Deviation
4
2
1
1
2
1
3
2
3
3
4
2 ppm
n
4
4
4
4
4
4
4
4
4
4
4
Mean
Recovery
100
100
100
99
98
100
102
99
101
101
0
Relative
Standard
Deviation
1
1
1
2
2
1
2
1
2
1
1
Note: n= Number of replicate analyses,
-------
The stability constants in Table 4.4 show that TEP has a very high
affinity for Hg2+. Therefore, a mixture of jDpth tiron and TEP was chosen for _•-
use. This procedure is described in Appendix D.
Cadmium ion was also evaluated as a sulfide precipitating reagent, but
was less effective than Pb2+. Since the Ksp values of CdS and PbS are com-
parable, the difference in effectiveness presumably is due to the much higher
stability constant of 10ll+ of the cadmium-TEP complex compared with only 10i0-5
for the Pb"+ complex. It appears that TEP can displace significant amounts of
S= from cadmium, whereas the lead sulfide species is stable in the presence of
TEP.
4.3.2 Results and Discussion
4.3.2.1 Cyanide Recovery—
An abbreviated study of the ligand-exchange method was conducted in
the laboratory, and the -results are shown in Table 4.5. This table includes
mean recoveries and their standard deviations. It can.be seen.from these re-
sults that complete recovery of cyanide is obtained from each of the compounds
investigated except KgfCoCCN^]. There is essentially no cyanide recovered
from this compound. Since recoveries from other procedures5"? are concentra-
tion dependent and do not normally exceed 15%, the lack of cyanide recovery from
this complex is not a major disadvantage.
The lower limit of detection of the ligand-exchange method is 2 ppb.
The method demonstrates precision within the guidelines outlined by the con-
tract; that is, ±10% above 10 ppb cyanide and zl ppb in the range between
0-10 ppb. Thus the performance of the method with respect to recovery of cya-
nide from the various compounds meets the EPA standards as outlined.
4.3.2.2 Interferences—
The performance of the ligand-exchange in the presence of suspected
interferences was also investigated. The results are depicted in Table 4.6.
TABLE 4.6 EFFECTS OF POTENTIAL INTERFERENCES ON
CYANIDE RECOVERIES USING THE LIGAND-EXCHANGE PROCEDURE
WITH AN ION SELECTIVE ELECTRODE FINISH
Compound level Apparent recovery from KCN
mole interference:mole CN percent
S (as Na S)
SCN~ (as KSCN)
NH, (as NH.C1)
Butanal
100
10
1
100
1000
1000
:1
:1
:1
:1
:1
:1
96
102
100
101
100
117
15
-------
From these results, it is apparent that the method overcomes the problems nor-
mally associated with the presence of sulfide and thiocyanate. Unfortunately^_the
comprehensive evaluation of suspected interferences was not completed. However,
on the basis of our experience using this and other analytical methodology,
the following statements can be made concerning likely interferences:
Chlorine - Chlorine can be expected to cause a negative inter-
ference in a manner similar to that found in the other reflux
distillation procedures investigated. This can be eliminated
by proper sample pretreatment.
Butylthiocyanate - This can be expected to interfere in a posi-
tive manner when using the electrode finish. The compound exhib-
its a large vapor pressure at room temperature and would be
expected to distill over using the procedure outlined and sub-
sequently interfere with the ion-selective electrode. It may
also interfere with the colorimetric and titrimetric finishes
by its production of a translucent scrubber solution. This can
probably be eliminated by using the extraction procedure outlined
for removing fatty acids.
Aldehydes - Aldehydes are expected to interfere because of their
reaction with cyanide to form cyanohydrins and further hydrolysis
to the corresponding acid and ammonia. In some instances, this
compound can be removed by the fatty acid extraction procedure.
In addition, the presence of the amine ligand TEP offers con-
siderable protection.
Fatty Acids - If the colorimetric finish is used, these compounds
are expected to interfere in a manner similar to that found for
the generally accepted procedures.5"7 However, the pH used in
this method and the short analysis time should minimize this prob-
lem and obviate the need for the extraction procedure that has
been described for removal of these compounds.
Other compounds commonly found in a sample that would be expected to interfere
are eliminated by the distillation procedure.
4.3.3 Conclusions
^_P^w_5ethpd for the analysis of total cyanide that overcomes a number
of problems normally associated with the generally accepted methods was
developed and evaluated.5-7 The procedure is simple and fast to operate, does
not require any exotic equipment or chemicals, can be readily adapted for use
on the equipment currently required by the approved method, and does not experi-
ence interferences from S= or SCN~. As such, it should be able to accurately
assess and monitor a wider variety of industrial effluents than was possible
using other procedures.
4.3.4 Recommendations
Further work should be conducted to determine the optimum operating
parameters of a method based on the ligand-exchange principle. Once this
16
-------
optimized procedure has been developed, a. comprehensive evaluation should be
conducted including ruggedness and round robin tests. This evaluation should
consist of determining the performance of the method on laboratory-prepared
samples of a number of compounds at various concentrations, the effect of sus-
pected interferences on the procedure, and the applicability of the method for
monitoring industrial effluents. It should also include a side-by-side com-
parison of this method with the procedure recommended for use by the EPA.
4.4 EPA TOTAL CYANIDE METHOD
4.4.1 Introduction and Background
Briefly stated, the EPA procedure is a one-hour catalytic reflux dis-
tillation procedure operated at a low pH. The cyanide compounds present in the
sample are dissociated and the cyanide, as hydrocyannic acid (HCN), is trans-
ferred to a sodium hydroxide scrubber solution. This solution is then quan-
titatively transferred to a volumetric flask and subsequently analyzed for
cyanide using either a titrimetric or colorimetric finish. A procedure
that is closely allied to that procedure recommended by the U.S. Environmental
Protection Agency for the analysis of total cyanides was evaluated. The modifi-
cations to the EPA method consist of the following: (1) addition of boiling
chips to the distillation flask, (2) transferring the scrubber solution to a
smaller volumetric flask, and (3) the use of an ion-selective electrode finish '
rather than a colorimetric or titrimetric finishes. (Refer to Appendix E for
the description of the procedure used.)
4.4.2 Results and Discussion
4.4.2.1 Cyanide Recoveries—
The performance of this method with respect to cyanide recoveries
on a number of compounds at each of four concentration levels (2, 20, 200,
and 2000 ppb) was evaluated. The results of this extensive investigation are
depicted in Table 4.7. This table includes several statistical parameters,
including the number of analyses performed, mean recoveries, and relative
standard deviations. It is apparent from these results that the procedure is
more than adequate on laboratory-prepared samples.
Only partial recovery of cyanide from K3[Co(CN)e] was found. The
reason for this is no doubt the slow kinetics of the decomposition of this
complex. At the pH used in the analysis, the decomposition of the complex is
thermodynamically favored. Since this complex may not be found in many indus-
trial effluents, incomplete recovery of cyanide from this compound may not be
a major problem.
From the results in Table 4.7, it is also apparent that there is no
recovery of cyanide from the nitriles. This is not an unexpected result.
Under the analysis conditions, these compounds can be expected to hydrolyze
to their corresponding acids as depicted below:
IT+ +
RCOOH + NH
17
-------
TABLE 4.7 CYANIDE RECOVERIES OBTAINED WITH THE EPA TOTAL CYANIDE METHOD
USING AN ION-SELECTIVE ELECTRODE FINISH
00
Compound
Studied
Simple Cyanides
tCCN
Cd(CN)2
Co(CN)2
CuCN
Fe(CN)2
Fe(CN)3
Hg(CN)2
Ma(CN)2
NI(CN)2
Zn(CN)2
Complex Cyanides
K,,[Fe(ci06]
K3[Fc(CN)6]
K^HgCCN),,]
K2[Ni(CN)M]
K3[Co(CN)6]
Organic Cyanides
CH3CN
CH3C1I2CH2CN
Concentration of CN
0.002 ppm
n
6
4
4
2
2
2
4
2
4
2
6
6
6
4
6
6
6
1
Mean
Recovery
87
48
127
92
96
96
73
100
101
76
130
105
110
25
116
0
0
0
Relative
Standard
Deviation
8
20
28
-
-
-
63
-
17
-
104
7.9
6
215
12
-
-
-
0.02 ppm
n
6
4
4
2
2
2
4
2
4
2
6
6
6
4
4
6
6
1
Mean
Recovery
99
101
65
99
100
100
99
100
97
99
96
101
95
99
104
7
0
0
Relative
Standard
Deviation
2
1
3
-
-
-
0.5
-
2.4
—
7.8
4.2
5
0.6
4
20
-
-
0 . 2 ppm
n
6
4
4
2
2
2
4
2
4
2
6
6
6
4
4
6
6
1
Mean
Recovery
100
102
35
94
100
100
100
100
98
100
97
104
111
99
102
4
0
0
Relative
Standard
Deviation
0.5
3
6
-
-
-
2.8
-
2.4
~
3.4
4.7
4.4
0.5
1.5
3
-
—
2 ppm
n
6
4
4
.2
2
2
4
2
4
2
6
6
6
4
4
6
6
]
Mean
Recovery
100
99
32
98
110
100
100
99
99
105
101
99
102
99
101
7
0
0
Relative
Standard
Deviation
0.5
0.5
5
-
-
-
0.5
-
1.1
"
4
3
3
1
2.5
7
1
"
Note: n = number of replicate analyses.
-------
Although one can expect to see nitriles in effluents, the inability of this
procedure to detect these compounds is not of major concern, since these
compounds are normally analyzed in effluents by other means.
4.4.2.2 Interferences—
The performance of this procedure was studied in the presence of
a number of suspected interferences. The results of our investigation, given
in Table 4.8, show that this procedure is sensitive to a number of interferences,
These are discussed below.
Thiocyanate—All of the thiocyanate compounds interfere with the
analysis._ This interference can be attributed to the decomposition of SCN~
to form S , which is transferred to the scrubber solution as H2S along with
the HCN. The presence of sulfide in the scrubber solution was verified by
visual observations of a yellow CdS precipitate upon treatment of the scrubber
with a cadmium sulfate solution. The decomposition of SCN~ does not proceed
rapidly, but occurs during the entire course of distillation. Hydrogen sulfide
could still be detected in the reaction flask after the analysis was complete.
This decomposition of thiocyanate to sulfide during distillation represents a
major interference. The presence of sulfide in the distillate affects all
three commonly used finishes: colorimetrie, titrimetric, and potentiometric.
.The sulfide can be removed from the distillate gas
stream by placing another gas scrubber in the vacuum train ahead of__the sodium
hydroxide gas scrubber. The solution in this additional scrubber is acidified,
• pH 4^j, and contains lead acetate^ ^p_.-PT^cipjltate_.the_jsjul.f±dg,.^^===f ,__...
A method that has been recommended for use in removing sulfide from
the caustic scrubber solution is precipitation as the cadmium or lead salt,
followed by filtration, which should produce a solution relatively free of
sulfide. However, this procedure has been shown to adversely affect cyanide
quantitation.15
Both of the above sulfide removal procedures add to the difficulty
in performing an analysis and exhibit still other problems.
Sulfide—Sulfide present in the sample will be distilled as H-S
during the analysis and will contaminate the sodium hydroxide scrubber solution.
As mentioned earlier, the sulfide can be removed from the sample solution by
precipitation as the insoluble cadmium or lead salt followed by filtration.
However, filtration also removes insoluble cyanides from the sample, and as
a result, the actual cyanide content of the sample may be underestimated.
Since the majority of cyanide compounds display rather small solubility prod-
ucts, this can be a serious problem.
Aldehyde—Aldehydes in the sample react with cyanide to form cyano-
hydrins, which are subsequently hydrolyzed to the corresponding acid and ammonia,
thereby, lowering the cyanide concentration. The procedure outlined for the
removal of fatty acids will also remove a number of aldehydes. However, the
problem cannot be completely alleviated by this procedure.
19
-------
TABLE 4.8 EFFECTS OF POTENTIAL INTERFERENCES ON CYANIDE RECOVERIES OBTAINED WITH
THE EPA TOTAL CYANIDE METHOD WITH AN ION-SELECTIVE ELECTRODE FINISH
Potential Interference
Compounds
KSCN
Co(SCN)2
K2[Hg(SCN)iJ
K3[Co(SCN)6]
n-butylthiocyanate
OCN~ (as KOCN)
NH,,+ (as NH^Cl)
Mn02
But anal
S= (as Na2S)
C12 (as Ca(OCl)2)
N02" (as NaN02)
Level of
Interference
Mole:Mole of CN~
100:1
1:1
100:1
1:1
100:1
1:1
100:1
1:1
100:1
1:1
100:1
1:1
100:1
1:1
100:1 "
1:1
100:1
1:1
10:1
1:1
10;1
1:1
100:1
1:1.
KCN
19000
160
32000
170
23000
330
100000
360
5800
170
100
99
100
60
92
2800
250
0
0
100
Apparen
from
CuCN
. 19000
120
19000
110
-
_
100
98
100
98
4700
240
0
0
97
: Recovery <
designated
Hg(CN)2
16000
210
28000
260
34000
380
94000
710
5500
180
100
100
100
100
3400
190
0
0
100
3f Cyanide (%)
Compound
K3[Cu(CNM
28000
120"
24400
125
_
-
-
105
110
110
88
105
3800
185
0
0
110
K^Fe^NJe]
36000
450
40000
600
_
-
_
105
100
105
100
3600
180
0
0
99
Note: Cyanide concentration level was 0.2 ppm In all test solutions.
20
-------
Chlorine—Chlorine present in a sample will oxidize
cyanide in the sample and produce abnormally low results. The procedure
recommended for removing excess chlorine is the incremental addition of
ascorbic acid. Although this procedure is effective in reducing the Cl£
content, an excess of ascorbic acid will produce a yellow scrubber solution,
thus interfering with the colorimetric finish. The ion-selective electrode's
response remains unaffected.
.Butylthi ocyanate—This compound, if present in the sample, will distill
during an analysis. It interferes with both the c.olor.ime.tri.c_and_ele.ctr.o.de
finishes ( the former by production of a cloudy scrubber solution). Because of
the limited solubility of this^ compound .in water.. _it .appears that_the_ extraction ^'
method recommended for removing fatty acids will also remove it.
Fatty Acids—Because of their finite vapor pressure under the
analysis conditions, fatty acids are distilled over into the caustic scrubber
solution, where they saponify and produce a cloudy solution. This interferes
with the colorimetric but not the potentiometric finish. An extraction pro-
cedure has been recommended for use to remove these substances.
4.4.3 Conclusions
The EPA total cyanide procedure is well suited for the analysis of sam-
ples not containing the above constituents. The procedure gives complete recovery
of cyanide from all of the compounds studied except K3[Co(CN)s]. The lower
limit of detection is 2 ppb. A major difficulty appears to be the large number
of compounds that interfere with the method. Thus a need exists for the develop-
ment of an analytical technique that can accurately assess cyanide levels in
other than clean samples.
This need is most accurately demonstrated by industrial effluents
from energy related industries. Refineries, coking operations, and coal
gasification industries produce effluents that can be expected to have high
levels of not only cyanide, but also thiocyanate and sulfide. (In view of the
recent expansion of the syn-fuel program, this need becomes most acute.) The
total cyanide procedure is inadequate for analyzing samples from such sources.
4.4.4 Recommendations
Barring any new improvements in this method, it is recommended that the
use of this procedure be restricted to samples not containing the above con-
stituents; most notably sulfide and thiocyanate. It is also recommended that
development and evaluation be continued on the alternative procedure described
in Section 4.2 for the analysis of samples for total cyanide, as it appears to
be most advantageous.
4.5 COMPARISON AND SUMMARY OF TOTAL CYANIDE METHODOLOGY
Of the two methods that have been evaluated in the most depth, the
ligand-exchange and catalytic reflux distillation procedures, both behave in a
comparable manner with respect to cyanide recoveries. Both procedures give
21
-------
complete recoveries of cyanide from all of the compounds studied except
K3[Co(CN)6] (See Table 4.9).
Both procedures also meet the criteria as set forth in the contract
for precision, accuracy, and sensitivity. The lower limit of detection for
both procedures is 2 ppb. In the range between 0-10 ppb, the precision is
±1 ppb. Above 10 ppb, the precision is within ±10%. (See Tables 4.5 and 4.7).
A comparison of their performance with respect to interferences reveals
that the ligand-exchange procedure outperforms the catalytic reflux distillation
method. This is most accurately depicted in Table 4.10. It can be seen from
this table that thiocyanate and sulfide interfere with the catalytic reflux
distillation procedure but not the ligand-exchange procedure. Since these
anions are common pollutants, this represents a significant improvement in
total cyanide analysis.
Another advantage of the ligand-exchange method is its ability to
reduce analysis time by approximately 30%. This will substantially reduce the
cost of an analysis since the greatest expense incurred is that associated
with labor.
TABLE 4.9 COMPARISON OF CYANIDE RECOVERIES OBTAINED WITH THE
EPA AND LIGAND-EXCHANGE TOTAL CYANIDE PROCEDURES
Cyanide Compound
KCN
Cd(CN)2
CuCN
Ni(CN)2
Hg(CN)2
K3[Cu(CN)4]
K2[Hg(CN)4]
K3[Fe(CN)6]
K3[Co(CN)6]
Concentration Level
PPB CN
2000 ppb CN
EPA
Complete
Complete
Complete
Complete
Complete
Complete
Complete
Complete
Complete
Complete
Partial
Ligand
exchange
Complete
Complete
Complete
Complete
Complete
Complete
Complete
Complete
Complete
Complete
None
200 ppb CN
EPA
Complete
Complete
Complete
Complete
Complete
Complete
Complete
Complete
Complete
Complete
Partial
Ligand
exchange
Complete
Complete
Complete
Complete
Complete
Complete
Complete
Complete
Complete
Complete
None
22
-------
TABLE 4.10 COMPARISON OF INTERFERENCES ON THE EPA AND THE
LIGAND-EXCHANGE TOTAL CYANIDE PROCEDURES
Compounds studied
for interference
SCN~ (as KSCN)
S= (as Na?S)
Butanal
NH4+ (as NH^Cl)
Level
Mole interference:
mole CN~
100:1
1:1
100:1
10:1
1:1
1000 : 1
100:1
10:1
1000:1
100:1
10:1
EPA
total cyanide
procedure
-1-
^
++
-
0
0
0
Ligand exchange
procedure
0
0
0
0
0
0
severe positive interference
slight to moderate positive
interference
slight to moderate negative
interference
0 no interference
23
-------
SECTION 5
SIMPLE CYANIDES METHODOLOGY
5.1 INTRODUCTION
Total cyanide has been used to indicate the sum of all the CN groups in
a sample, regardless of the nature of the metal complex with which the CN~ is
associated. A subset of total cyanide is a group of metal compounds commonly
referred to as "simple cyanides." Despite the widespread use of this term, no
single definition has gained complete acceptance. The term is intended to de-
note easily dissociable cyanide complexes, and is frequently used as a synonym for
"cyanides amenable to chlorination." The definition used in this report is that
from the contract statement of work—a definition based on the easily defined stoi-
chiometry of the cyanide complexes. Thus, the neutral stoichiometric cyanides
[Mn+(CN)n] are considered to be simple cyanides. The more highly associated
salts that dissolve in aqueous solution to produce [Mn+(CN),, ] ^Y~n) anions are
defined as complex cyanides. Table 5_._1 lists _the_compounds included in this
study.
TABLE 5.1 JJOMPOUNDS INCLUDED I'JsTSTUDY " '
Cyanide compound
KCN
Cd(CN)2
Co(CN)2
CuCN
Fe(CN)
Mn(CN)2
Ni(CN)2
Zn(CN)2
K2 [Ni (CN) ^ ]
R2[Hg(CN)4]
(continued)
24
-------
TABLE 5.1
Cyanide compound
K3[Fe(CN)6]
K4[Fe(CH)6]
K3[Co(CN)6]
Six analytical methods for simple cyanides have been studied; two of
these have been compared in detail with the EPA method, "Cyanides Amenable to
Chlorination." (Strictly speaking, comparisons were -made to a modified EPA
procedure. The modifications used were: addition of boiling chips during dis-
tillation, smaller volumetric flasks, and an electrode finish. These modifi-
cations are included in the method as it appears in Appendix C.) This labora-
tory work generally consisted of evaluating the performance of the method on
prepared standards of metal-cyanide compounds. The compounds studied included
ten simple and six complex cyanides. These compounds were prepared in deionized
water at four concentration levels. The effect of suspected interferences was
also evaluated. Table 4.3 lists the substances studied for interferences.
In some instances, severe problems were identified early ifi the
development phase. In these cases, only an abbreviated study was conducted.
since further work on an inherently deficient method was deemed inappropriate.
5.2 ION EXCHANGE PROCEDURE
5 . 2 .'1 Introduction and Background
A procedure developed by Gilath22 for the analysis of simple cyanides
appeared to be fast, simple, and efficient. Complete recoveries of cyanide
from K2[Zn(CN)J4], K2 [Cd(CN) 1+ ],: K'3 [CuCCN)^ ], and K[Ag(CN)2] were reported. On
the basis of these reported results and knowledge of the chemistry of the
various cyanide compounds, this procedure was considered to be ideally suited
for the analysis of simple cyanides.
The procedure is based on the absorption of cyanide ion on a strong
anion exchanger, followed by elution of the absorbed species by an acidic
solution. The cyanide so eluted is then quantitated by any of a variety of
techniques.
5.2.2 Procedure
The apparatus is depicted in Figure 5.1. The column was packed with 5 g
of the strong anion exchange resin, Amberlite IR-400. The column was then
backwashed to free the bed of entrapped air pockets, classify the resin
25
-------
( }—— Reservoir
•Ion Exchange Resin
-Sintered Glass Filter
Stopcock
Glass Tip
SA-7854-11
Figure 5.1 Ion-exchange apparatus.
26
-------
particles, and rid the bed of debris and resin fines. The resin was then con-
verted to the hydroxide form and rinsed.
After preparation of the column, a 20-mL_sample was allowed to flow
through the resin bed at the rate of 1 drop/second. The column was then rinsed
with 15 mL_ of distilled water at the same flow rates. The absorbed cyanide
was then eluted from the column by two consecutive acid elutions. The^first
elution was performed with 15niL_ of 2N HaSOt* and the second with 15 m~L of
4.5N l^SOit. The first acid elution was performed quickly to avoid the loss
of cyanide as HCN. The second acid wash was performed slowly in order to
ensure complete removal of cyanide from the resin. The resin was then rinsed
with 14 mL .of distilled water. The acid eluents were delivered to a magnetic-
ally stirred beaker containing 80 mL °f 2.05N sodium hydroxide. The delivery
tip of the column was placed below the sodium hydroxide surface during the
elutions to prevent loss of cyanide as HCN. The contents of this beaker were
then quantitatively transferred to a volumetric flask and analyzed. The
standards used for this analysis were prepared in solutions consisting of a
similar matrix as that found in the sample.
5.2.3 Results and Discussion
Recovery of cyanide from a KCN standard was found to be only 25 + 5%.
These results differ significantly from those reported by Gilath.22 Gilath
conducted her studies on solutions containing between 21 and 40 g/L cyanide'
The concentrations of cyanide in the solutions analyzed in our investigation
were approximately 1 mg/L. These levels are orders of magnitude lower
than those used by Gilath. Apparently, the lower limit of detection of the
method lies somewhere between 1 and 20,000 ppm. This lower limit is suffic-
iently high so that the method is not useful at the concentrations of major
concern to us, namely between 2 and 2000 ppb.
Other shortcomings of a technical nature became apparent during our
investigations. The procedure was found to be excessively time consuming.
Several hours were required to regenerate and rinse the column in order to
perform one analysis. There was some evidence of HCN evolution during the
acid elution, resulting in abnormally low cyanide values. Finally, because
of the low recoveries found, it would seem that some cyanide is still adsorbed
on the resin. If so, this would impair the ability to use the column for more
than one analysis.
Because of poor performance of the method during these initial inves-
tigations, further work using this method was discontinued.
5.2.4 Conclusions
The ion exchange method is deficient in several respects. Of major
concern is the incomplete recovery of cyanide from a 2-ppm potassium cyanide
sample. The method is also too time consuming for routine analyses because
of the slow regeneration of the column.
27
-------
5.2.5 Recommendations
In light of the performance of this procedure and that of the other
procedures investigated in this laboratory, further work using the ion
exchange method is inappropriate.
5.3 CONTINUOUS-FLOW DISTILLATION
5.3.1 Introduction and Background
An automated procedure designed by Goulden, Afghan, and Brooksbank11
for the analysis of simple cyanides has been suitably altered for use as a
manual method of analysis. The primary alterations to the system consist of
replacing the pump-driven delivery systems with a gravity feed system and
replacing the automated colorimetric finish with a manual potentiometric
finish.
An acidified sample is allowed to flow in a thin film down a heated
tube through which air is flowing. The cyanide so released is trapped in a
caustic scrubber solution and manually quantitated. Presumably, because of
the short residence time of the sample in the heated tube, only simple cyanides
are decomposed and subsequently analyzed.
5.3.2 Procedure
The apparatus used in our investigations is depicted in Figure 5.2.
The procedure is set in operation by charging the sample reservoir with water,
the acid reservoir with the acid mixture described in Figure 5.2, and the
absorber with 10 mL of 1.25 N NaOH diluted with enough water to give an ade-
quate depth of liquid. The water, acid, and air were allowed to flow through
the system at flow rates of 2, 0.7, and 20 mL/minute, respectively, for 5 minutes
to stabilize the system. The distillation tube was heated so that approximately
60% of the sample was vaporized. The water-cooled condenser ensures that the
water vapor is condensed and eliminated from the distillation tube via the
waste disposal lines rather than collecting in the scrubber. The sample was
transferred to the sample reservoir and allowed to pass completely through the
system. The sample lines were thoroughly flushed by passing 30:mL of water
through the system. The scrubber solution was transferred to a volumetric
flask and brought up to volume with water, and the contents of this flask were
then analyzed for cyanide using an ion-selective electrode.
5.3.3 Results and Discussion
Analyses of laboratory-prepared standards of KCN, K2[Ni(CN)4], and
Ko[Fe(CN)^] solutions were conducted with this system. The results of these
investigations are shown in Table 5.2.
28
-------
Sample
Funnel
Acid Funnel
(20% of 85% H3P04
and 4% of 50% H3P02)
Capillary-
F low-
Control
(2 ml/min)
-Capillary
Flow
Control
(0.7 ml/min)
5 mm ID
Check
Valve
Mixing
Coil
This Portion of the Tube
is Wrapped with Heating
Tape That Has a Resist-
ance of 25 n. Operating
Potential is 52 V.
20 cm
To Vacuum
Gas Scrubber
with Medium
Porosity Frit
(ft
3 mm \\y
Tubing
SA-7854-8
Figure 5.2 Continuous-flow distillation apparatus.
29
-------
TABLE 5.2 CYANIDE RECOVERIES OBTAINED WITH THE CONTINUOUS-FLOW
DISTILLATION PROCEDURE
Relative standard
Mean recovery deviation
Compound
KCX
K2[Ni(CN)4]
K3[Fe(CN)6]
97.9
74
13.5
4
6
5
Essentially, complete recovery'of cyanide from a KCN sample was
possible. However, only partial recoveries of cyanide from both the nickel
and iron cyanide complexes was found. To be a viable method for the analysis
of" simple cyanides, the procedure must be able to give complete recoveries
of cyanide from the less stable cyanide complexes and essentially no recovery
from the more stable cyanides. From the preceeding data, it can be seen that
this procedure does not meet this criterion.
It appears that the procedure would have other problems as well.
Among these are:
• Inability to analyze samples containing particulate matter
due to clogging.
• Interference from sulfide
• Long analysis times.
5.3.4 Conclusions
The continuous-flow distillation procedure evaluated in our laboratory
for the analysis of simple cyanides was found to be inadequate. Chief among
the deficiencies of the method were:
• Inability of the method to differentiate between simple and
complex cyanides.
• Inability of the method to achieve complete recovery of
cyanide from complexes of moderate stability.
• Long analysis time.
With sufficient developmental work, the procedure could be modified
to overcome these and other deficiencies.
5.3.5 Re commend ati ons
In view of the performance of other methods studied, this method should
be abandoned as a manual method for the analysis of simple cyanides. A
30
-------
significant amount of additional developmental work is necessary to produce
a viable method of analysis.
5.4
EDTA ELECTRODE PROCEDURE
5.4.1 Introduction and Background
Ethylenedinitrilotetraacetic acid, also known as ethylenediaminetetra-
acetic acid (EDTA), is an excellent ligand for most ions. By using this rea-
gent in a suitable environment (pH and temperature), it is possible to displace
cyanide from all of the cyanide compounds studied except potassium hexacyano-
cobaltate(III) (see Section 4.3). Appropriate selection of the environment in
which this displacement occurs should result in a system that is useful for
the analysis of simple cyanides. Use of this•chelate as a mask and sequestering
agent have been described.19"21 On the basis of these earlier reports, a sys-
tem was developed and evaluated in the laboratory.
5.4.2 Procedure
The procedure consists of adding EDTA to an aliquot of sample made basic
by the addition of 1.25 N sodium hydroxide. The sample was immersed in a water
bath regulated at 40°C and stirred by means of a magnetic stirrer. After a
sufficient period of time, the sample was removed from the water bath and
transferred to a volumetric flask. The contents of the flask were then
analyzed for cyanide using an ion-selective electrode.
5.4.3 Results and Discussion
and 5.4.
The results of these investigations are summarized in Tables 5.3
TABLE 5.3 CYANIDE RECOVERIES OBTAINED WITH THE
EDTA-ELECTRODE PROCEDURE
a
Cyanide compound
KCN
1
3
3
10
CuCN
1
3
3
10
, 20
, 20
, 45
, 60
-
, 20
, 20
.. 45
min
min
min
min
min
min
min
, 60 min
Concentration
range' studied Number
(ppm CN) of analyses
1
1
1
1
0.5-8.8
0.7-4.2
1.6-7.3
1.5-7.5
9
6
6
6
12
6
12
5
Mean recovery
99.
101.
100.
101.
71.
86.
65.
100.
9
1
0
3
5
5
5
1
Relative
standard
deviation
2.
11
3.
8.
29
8.
15
7
7
7
2
8
(continued)
31
-------
TABLE 5.3
Cyanide Compound'
Concentration
Range Studied
(ppm CN)
Number
of Analyses
Relative
Standard
Mean Recovery Deviation
K2[Ni(CN)4] • —
3 iiiL , 20 min
3~mL > 45 min
10 iiiL , 60 min
K [Fe(CN)6]
1 mL , 20 min
3 mL , 20 min
3 mL , 45 min
10'm"L , 60 min
K-[Co(CN)6]
1 mL , 20 min
3 mL_, 20 min
3 "ml, 45 min
0.6-2.8
1.4-2.8
1.1-5.4
1.2-1.9
0.6-3.6
2.5-3.2
1.0-4.2
0.7-2.2
1.2-4.3
0.5-5.7
98.3
88.6
110.1
29.3
1.1
2.2
0.0
0.0
0.0
0.0
11
6.5
5.5
14
49
40
0
0
0
0
Each solution was prepared by adding the amount of 0.257M EDTA solution
indicated and equilibrating for the time indicated.
Precautions were not taken to prevent photodecomposition. When uv
light is absent, the complex is stable.
TABLE 5.4 SUMMARY OF RESULTS FROM THE EDTA-ELECTRODE PROCEDURE
Cyanide compound
CuCN
K2[Ni(CS)4]
Mole ratio
(EDTA/metal)
< 100
> 100
< 600
>1000
Mean recovery
68
97
90
110
Number of
samples
24
8
8
6
Both ferricyanide and cobalticyanide remain unaffected by any of the
systems investigated. Recoveries of cyanide from KCN are essentially complete
and also unaffected by the different experimental parameters used. Difficul-
ties are only encountered when analyzing samples that contain CuCN or
K.2 [Ni(CN) 1+ ]. To obtain complete recoveries of cyanide from cuprous cyanide,
high EDTA:metal ratios are required. However, at these high levels_a positive
interference is encountered when analyzing solutions containing K2 [Ni(CN)[,.].
Thus, there is not a clear separation between simple and complex cyanides.
32
-------
To be a useful method of analysis, the method must accurately quanti-
tate cyanide with a relatively high degree of precision. From the data
included in Table 5.2, this is shown not to be the case. The relative standard
deviation was found to vary from 3.7 to 29%. Another major limitation of the
procedure is that it can analyze only clean samples. The cyanide is not
separated from the sample matrix, which often contains a number of compounds
that will interfere. In light of these disadvantages, further work using this
procedure was deemed inappropriate.
5.4.4 Conclusions
The EDTA electrode procedure for analysis of simple cyanide was found
to be deficient in a number of areas. Although it is a fast, simple method,
complete recovery of cyanide from complexes of moderate stability was not
possible and the precision was very poor. Also, the procedure is sub-
ject to a number of interferences and as a result, could be used on only the
cleanest samples. Further developmental work would be necessary to overcome
these difficulties.
5.4.5 Recommendations
In light of the amount of additional developmental work required by
the EDTA electrode procedure and the satisfactory results obtained using other
procedures, (see Sections 5.7 and 5.8), no further work along this line is
recommended.
5.5 AMERICAN IRON AND STEEL INSTITUTE 'AERATION PROCEDURE
5.5.1 Introduction and Background
An aeration procedure for the analysis of simple cyanides described
by Caruso23 has been evaluated in our laboratory. The procedure consists
of dissociating all but the most stable cyanide complexes by acidifying the
sample to pH 4. The cyanide released by these complexes is aerated as
hydrocyanic acid into a sodium hydroxide absorption trap. This scrubber
solution is subsequently analyzed for cyanide by one of any number of
analytical techniques.
5.5.2 Procedure
The apparatus used in this procedure is, but for a minor modification,
the same as that used in other reflux distillation procedures.5 This appara-
tus is depicted in Figure 5.3. To the absorber is added 50 mL of 1.25 N
sodium hydroxide diluted with enough water to obtain a satisfactory depth
of solution. The reaction flask is charged with 500 ml of sample or an
aliquot of sample diluted to 50QiiiL . The vacuum is adjusted so that the
air flow through the flask is approximately 3 liters/minute. A few drops of
methyl orange indicator are added to the sample through the air inlet tube.
33
-------
In
Cooling Water
Inlet Tube
Screw
Clamp
To Low
Vacuum
Source
250 mL Absorber
Medium
Porosity Frit
Distillation Flask
SA-7854-10
Figure 5.3 AISI aeration apparatus.
34
-------
The pH of the sample is then adjusted to and maintained at 4.5 by the addition
of sulfuric acid (1+9). Air is pulled through the sample for 2 hours, after
which time the scrubber solution is quantitatively transferred to a 250-mL volu-
metric flask. This solution is then analyzed using an ion-selective electrode.
5.5.3 Results and Discussion
An evaluation of the AISI aeration method was conducted on prepared
standards containing KCX, CuCN, K2[Ni(CN)^], K3[re(CN)6], and K3[Co(CN)6].
The recoveries, concentration range, and standard deviations found using this
procedure are shown in Table 5.5. Of the cyanide compounds studied, complete
recoveries of cyanide were obtained only with KCN. The reasons for the poor
performance of the method with respect to recovery efficiencies from the
simple cyanide species are discussed below.
TABLE 5.5 CYANIDE RECOVERIES OBTAINED WITH THE AISI AERATION PROCEDURE
Compound
KCN
CuCNa
K2[Ni(CNK]
K3[Fe(CN) 5]
K3[Co(CN)6]
Concentration
range studied
(ppb CN~)
1000
392-706
276-534
482-1122
558-1822
Mean recovery
(%)
95.8
1.5
60.2
0
0
Relative standard
Deviation
(%)
7.7
6
18.9
-
—
The CuCN could be visually observed on the inside walls of the reaction
flask above the liquid level.
The aeration of hydrocyanic acid from solution is dependent on pH,
temperature, air flow rate, ionic strength, and vapor pressure. This relation-
ship is described by the following equation.
a)
but PV = nRT
Rearrangement and substitution of this into equation (1) gives
(V) 4£ = -[l£-
or
-(?)(¥)
-
35
-------
where: V = volume of sample (L)
W = air flow rate (L minute )
H = Henry's law coefficient (torr mole L)
P = atmospheric pressure
Co = initial cyanide concentrations
C = cyanide concentration at tine t
R = gas law constant (62.358 to^r ^t
mole K
T = temperature
t = time.
Solution of this differential gives:
The ratio C/Co varies from 1 to 0. Substitution of these values into
equation (2) gives the curves shown in Figures 5.4, 5.5, and 5.6, which
represent aeration of cyanide under various conditions of analysis.
Based on these figures and equation (2), most of the hydrocyanic acid
will have been aerated from the sample at the end of 2 hours. However, most
of the compounds studied dissociate slowly at pH 4, even though the decompo-
sition of most cyanide complexes is thermodynamically favored.
As a result, the effective aeration time during an analysis is less than 2
hours and recoveries will not be complete. This is adequately depicted by the
recoveries of cyanide found for the analysis of the samples containing K.2 [Ni(CN) ^ ],
Low recoveries of cyanide from cuprous cyanide are also observed when
using this procedure. The compound is insoluble in water (refer to Table 4.1),
and during an analysis, CuCN was observed adhering to the inside walls of
the reaction flask well above the liquid level. As a result, this compound-was
not able to participate in the decomposition reactions.
From the above discussion, it becomes apparent that this procedure is
useful only for the analysis of soluble cyanides that are kinetically labile
with respect to the dissociation of cyanide. Only a relatively small number of
compounds fall into this category. Since .the procedure was inappropriate for
the analysis all simple cyanides, further evaluation of this procedure was
discontinued.
5.5.4 Conclusion
The AISI aeration method is inadequate for the analysis for s'imple
cyanides. Although decomposition of ferricyanide is avoided, recoveries of
cyanide from insoluble and/or kinetically inert complexes are incomplete. Con-
ditions could be altered to alleviate some, if not all, of these problems, but
substantial developmental work would be required (see Section 5.6).
36
-------
1.0
0
**y
** ?*
**>
15
20
75 90 105
TIME (min)
SA-7854-20
Figure 5.4 Effect of temperature on aeration of HCN from solution.
Sample volume = 0.5 L. Air flow rate = 3 L/min.
-------
00
1.0
0
0
15
30 45
60
75 90 105
TIME (min)
120
SA-7854-21
Figure 5.5 Effect of sample volume on aeration of HCN from solution.
Temperature = 298 K (25°C). Airflow rate = 3 L/min.
-------
u>
ID
1.0
0
15
30 45
60
75 90 105
TIME (min)
120 135 150 165 180 -O
SA-7854-22
Figure 5.6 Effect of air flow rate on aeration of HCN from solution.
Temperature = 298°k (25°C). Sample volume = 0.5 L.
-------
5.5.5 Recommendations
Further developmental work would improve the performance of this method.
In particular, the effect of varying key parameters such as temperature and air
flow rate should be evaluated. However, because other procedures^ give more
satisfactory performance, in particular the Roberts-Jackson distillation pro-
cedure, continued developmental work may not be desirable.
5.6 EDTA AERATION PROCEDURE
5.6.1 Introduction and Background
As stand-alone analytical methods of analysis for simple cyanides, the
EDTA electrode (Section 5.4) and AISI aeration (Section 5.5) procedures exhibit
serious shortcomings. To overcome the major flaws of these procedures and yet
take advantage of their beneficial operating characteristics, a procedure
based on a combination of these two methods was developed and evaluated. (A
similar procedure has been previously described.)2A The procedure evaluated
is described in Appendix A.
Basically, the method is a 2 hour room-temperature aeration procedure
carried out at a pH of 4.5 in the presence of EDTA. As previously stated,
EDTA is an excellent ligand for many metals. As a result it effectively com-
petes and displaces cyanide from the inner coordination sphere of the various
metals present in the sample. The liberated cyanide is aerated from solution
as hydrocyannic acid (HCN) and is subsequently trapped in a caustic scrubber
solution and quantitated either titrimetrically, colorimetrically, or
potentiometrically.
5.6.2 Results and Discussion
5.6.2.1 Cyanide Recoveries—
Complete recoveries of cyanide using the EDTA aeration method are
found for only a limited number of the compounds studied (see Table 5.6). The
remaining compounds give either partial or no recovery of cyanide when sub-
jected to this method of analysis.
Compounds K3[Fe(CN)6], K4[Fe(CN)6], and K3[Co(CN)6] give essentially
zero recovery of cyanide. These species are not considered simple cyanides
and as such should not respond. The partial recoveries of cyanide exhibited
by many of the remaining compounds is, however, a problem. These incomplete
recoveries arise because the rate of displacement of cyanide from these com-
plexes by EDTA appears to be relatively slow and because of the logarithmic
stripping rate of hydrocyannic acid from solution (see Section 5.5.3). Thus,
the aeration time is not long enough to give complete recoveries.
These incomplete recoveries could no doubt be improved by slight
modification of the analysis conditions. For example, a slight increase in tem-
perature not only would increase the stripping rate of the hydrocyannic acid
from solution, but also would increase the rate of displacement of cyanide from
the complexes. This temperature change would also increase the risk of decom-
40
-------
TABLE 5.6 CYANIDE RECOVERIES OBTAINED WITH THE EDTA AERATION METHOD
Compound
Studied
KCN
Cd(CN)2
Co(CN)2
CuCN
Fe(CN)2
Fe(CN)3
Hg(CN)2
Mn(CN)2
Ni(CN)2
Zn(CN)2
K [Cu(CN) ]
K2[NI(CN) ]
K2(Hg(CN) ]
K3[Fe(CN)6]
K [Fe(CN)6]
K3(Co(CN)6]
Concentration of CN
2 ppm
n
4
2
2
2
2
2
2
2
2
2
4
2
2
2
2
2
Mean
Recovery
(%)
98
100
18
81
98
•91
35
95
82
92
96
48
128
0
0
0
Relative
Standard
Deviation
CO
12
—
—
—
~
—
—
—
—
—
14
—
—
—
—
—
0.2 ppm
n
4
2
2
2
2
2
2
2
2
2
4
2
2
2
2
2
Mean
Recovery
(%)
112
94
18
102
98
93
43
90
36
102
93
57
123
.0
0
0
Relative
Standard
Deviation
CO
8
—
—
—
—
—
—
—
—
—
17
—
—
—
—
—
0.02 ppm
n
2
2
2
2
2
2
2
2
2
2
4
2
2
2
2
2
Mean
Recovery
CO
98
102
39
66
87
67
43
98
54
103
98
78
198
0
0
0
Relative
Standard
Deviation
CO
10
•
—
—
—
—
—
—
—
—
32
'
—
—
~
—
0.006 ppm
n
4
2
2
2
2
2
2
2
2
2
3
2
2
2
2
2
Mean
Recovery
a>
64
76
40
84
73
65
38
74
81
96
50
40
112
0
0
0
Relative
Standard
Deviation
CO
16
—
—
—
—
—
—
—
—
—
13
—
—
—
—
—
-------
position of the ferri- and ferrocyanide complexes. Both these factors must be
considered when devising and evaluating modifications.
Currently, the lower limit of- detection is 5 ppb, higher by a factor
of 2.5 than any of the other procedures evaluated. The relative standard devia-
tion, a measure of precision of the method, is also higher. This is apparently
due to the large number of experimental variables involved (see Sections 5.4
and 5.5). Suitable control of the analysis conditions should improve the pre-
cision of the method.
5.6.2.2 Interferences—
The EDTA aeration procedure is subject to a limited number of inter-
ferences, as shown by Table 5.7. The major interferences are briefly discussed
below.
Chlorine—Chlorine oxidizes cyanide and thus is a negative interfer-
ence. This interference can be eliminated by reduction of chlorine at the time
of sample collection with ascorbic acid.
Sulfide—Sulfide will cause a positive interference to an ion specific
electrode finish because of formation of hydrogen sulfide and its subsequent
trapping in the caustic scrubber solution. However, the sulfide is commonly
removed from the sample before storage as the lead, cadmium, or arsenic salt.
Sulfide, as PbS, will not interfere as seen from the results in Table 5.7.
Thus, filtration of the sample is not necessary.
.Butylthiocyanate •• Because of its high volatility, this organic compound-
_is found Jn~the~scr libber 'solution at the end of the analysis and_ interferes with the_
electrode finish. Because of its limited solubility in water, this substance may
__b.e_removeii from the sample by the extraction procedure outlined for fa
removal.
Co3+—When added as K3[Co(SCN)6] or Co(SCN)3, this substance produces
a negative interference. Apparently, cyanide exchanges with SCN~ to form the
more stable, inert Co(CN)63~ complex. This appears to be an irreversible pro-
cess, although the use of another sequestering agent offers a slim chance of
alleviating this problem. The SCN~ does not appear to interfere.
Hg +—When added as K2[Hg(SCN)^], this interferes in a manner similar
to that reported for Co3"1" by forming stable mercury cyanide complexes. However,
by suitable alterations in methodology this could probably be eliminated. The
changes that should be investigated include higher temperatures, higher air flow
rates, and different sequestering agents.
5.6.3 Conclusions
As a method of analysis for simple cyanides, the EDTA aeration method
exhibits certain deficiencies. Prime among these are its partial recoveries of
cyanide from some simple cyanides and its low precision. However, through suit-
able modification of analysis conditions, these problems could probably be eli-
42
-------
TABLE 5.7 EFFECTS OF POTENTIAL INTERFERENCES ON CYANIDE RECOVERIES
OBTAINED WITH EDTA AERATION PROCEDURE
Potential interference
compounds
SCN~ (as KSCN)
Co(SCN)2
K2[Hg(SCN)4]
K3[Co(SCN)6]
OCN~ (as KOCN)
Nll,,+ (as NH^Cl)
Mn02
n-butylthiocyanate
Butanal
N0a~ (as NaN02)
S~ (as NazS)
S= (as PbS)
Cl2 (as Ca(OCl)2)
Level of
interference
mole :mole CN
100:1
100:1
100:1
100:1
100:1
100:1
100:1
100:1
100:1
100:1
10:1
1:1
100:1
10:1
1:1
Apparent Recovery of Cyanide (%)
from Designated Compound
KCN
90
45
0
53
95
90
103
1225
38
90
5000
850
100
0
0
CuCN
93
50
0
40
93
93
85
3250
60
93
4750
500
-
0
0
Hg(CN)2
95
90
0
105
55
45
53
2350
43
48
4500
450
-
0
0
K3[Cu(CN)lt]
90
75
0
75
75
75
100
4000
75
88
5000
650
-
0
0
K4[Fc(CN)6]
0
0
0
0
0
0
0
2750
0
0
4000
600
-
0
0
Note: Concentration of cyanide was 0.2 ppm in all test solutions.
-------
minated. The procedure is unaffected by the presence of thiocyanate and sulfide,
as lead sulfide. The other interferences can be eliminated from the sample be-
fore analysis.
5.6.A Recommendations
Through suitable modification, the performance of this procedure
with respect to cyanide recoveries could be substantially improved. Studies
of these modifications should be conducted, especially in light of its satis-
factory performance in the presence of thiocyanate and lead sulfide. These new
lines of investigation should include a study of the effect of temperature, air
flow rate, analysis time, and different sequestering agents. After optimiza-
tion of conditions, the procedure should be evaluated on a number of compounds
in the presence and absence of suspected interferences.
5.7 EPA PROCEDURE FOR CYANIDES AMENABLE TO CHLORINATION
5.7.1 Introduction and Background
A procedure for the analysis of simple cyanides, or more appropriately
"cyanides amenable to chlorination" (CATC), that has attained widespread use
and is currently recommended by the Environmental Protection Agency was evalu-
ated. This procedure is based on the difference between two analyses for total
cyanide. An unadulterated aliquot of the sample and an aliquot of the sam-
ple that has been chlorinated (thereby destroying the CATC) are subjected to
the total cyanide distillation procedure. The difference in the cyanide con-
tent of these two samples is defined to be the CATC. The cyanide in the absorp-
tion solution is determined by any one of a number of different analytical
techniques.
This procedure relies on the total cyanide methodology presented in
Section 4.4. All the information contained in that section is equally applic-
able here. The method is described in detail in Appendix E.
5.7.2 Results and Discussion
5.7.2.1 Cyanide Recoveries—
The results of•the evaluation of the CATC procedure are shown in
Table 5.8. Each cyanide compound falls into one of three categories: those
that are not chlorinated at all; those that are partially chlorinated; and those
that are completely chlorinated. This latter category is restricted to those
species that are easily dissociable in water and of only moderate stability.
Inlcuded in this group are: KCN, Cd(CN)2, Cu(CN), Fe(CN)2, Fe(CN)3, Mn(CN)2,
Zn(CN)2, and K3[Cu(CN)4]. The compounds that are not at all chlorinated are
K3[Fe(CN7)6L Kit[Fe(CN)6], K3[Co(CN)6], and nitriles. These complexes are very
stable and/or kinetically inert, whereas the nitriles are easily hydrolyzed to
the corresponding acid and ammonia under the conditions of analysis.
The remaining compounds—Ni(CN)2, Hg(CN)2, K2[Ni(CN) ij, and
K.2 [Hg(CN) L,.]—are only partially chlorinated and so fall into the second category.
The degree of chlorination of these compounds depends on the initial concentra-
tion of the compound. Presumably only dissociated CN~ groups are chlorinated,
44
-------
TABLE 5.8 CYANIDE KECOVERIES OBTAINED WITH EPA METHOD "CYANIDES
AMENABLE TO CHLORINATION" USING AN ION-SELECTIVE ELECTRODE
-C-
Ln
Compound
Studied
Simple Cyanides
KCN
Cd(CN)2
Co(CN)2
CuCN
Fe(CN)2
Fe(CH)3
Hg(CN)2
Mn(CN)2
Ni(CN)2
Zn(CN)2
Complex Cyanides
K3[Cu(CN),J
K,,[Fe(CM)6]
K3[Fe(CN)6]
K2lHg(CN),,]
K2[Ni(CN),,]
K3(Co(CN)6]
Organic Cyanides
CH3CN
CH3CH2CH2CN
Concc-.iitration of CN
0.002 ppm
n
6
4
4
1
1
1
4
1
4
1
6
6
6
4
6
6
6
1
Mean
Recovery
87
48
127
92
96
46
48
50
101
76
130
-41
-30
30
70
0
0
0
Relative
Standard
Deviation
8
20
24
-
-
-
93
-
19
-
104
9
6
157
11
-
-
-
0.02 ppm
n
6
4
4
1
1
1
4
1
4
1
6
6
6
4
4
6
6
1
Mean
Recovery
99
101
45
99
100
83
93
93
76
99
96
-5
-2
99
59
2.5
0
0
Relative
Standard
Deviation
2
14
3
_
-
-
42
-
2
-
7. a
5
6
0.6
5
14
-
-
n
6
4
4
1
1
1
4
1
4
1
6
6
6
4
4
6
6
1
0 . 2 ppm
Mean
Recovery
100
102
3.5
94
100
90
89
97
64
99
97
0
0.3
92
48
-0.1
0
0
Relative
Standard
Deviation
0.5
2
8
-
-
-
6
-
4
-
3.4
4
3
2
3
12
~
—
2 ppm
n
6
4
4
1
1
1
4
1
4
1
6
6
6
4
4
6
6
1
Mean
Recovery
100
99
-1
98
108
99
59
96
33
104
101
-0.4
1
81
30
0.1
o
0
Relative
Standard
Deviation
0.5
4
3
-
-
-
b
-
2
—
4
3
3
3
. 2
5
~
—
-------
so that metal-bound groups are protected. The degree of dissociation is concen-
tration dependent. At lower concentrations, a higher percentage of cyanide will
be dissociated, which results in chlorination of a larger fraction of the total
sample. Figure 5.7 shows the recoveries of cyanide from several complexes as a
function of their initial concentration.
The mechanism of chlorination generally proceeds along the following
lines. The cyanide present in the sample is oxidized to cyanogen chloride
(CNC1), which is then hydrolyzed to cyanate (CNO~). The hydrolysis of cyanogen
chloride is both pH and time dependent. At pH 9, with no excess chlorine pre-
sent, cyanogen chloride may persist for 24 hours. The cyanate produced by CNC1
hydrolysis can be further oxidized with chlorine at near neutral pH to carbon
dioxide and nitrogen. Upon acidification, cyanate will be converted to ammonia.
Although the method exhibits a number of deficiencies, there are two
areas that should be explored further in an effort to improve the method. These
are briefly discussed below.
The oxidizing ability of the hypohalous acid used is directly depen-
dent on pH. The dissociation constant of HOC1, 3.4 x 10~8, and the standard
oxidation-reduction potentials for reactions of the halogens (tabulated below)
indicate that chlorination of cyanide would proceed more readily at a lower pH
than that currently being used.
TABLE 5.9 STANDARD REDUCTION POTENTIALS OF CHLORINATING SPECIES
H+ + HOC1 + e~ - — >• 1/2 C12 + H20
1/2 C12 + e~ " Cl~
C10~ + H20 + 2e~ -< -PI"" + 20H~
1.63
1.36
0.89
If the rate-limiting step of the chlorination reaction is the dissoci-
ation of the cyanide complexes, the use of stronger oxidizing agents may have no
effect on the chlorination. If, however, direct oxidation of the metal cyanide
compounds becomes a major contributing pathway for the chlorination reaction as
a result of the increased oxidation potentials, then the use of different pH
values should prove beneficial. A laboratory investigation to determine the pH
dependence and/or mechanism of this reaction should be conducted to develop a
more efficient chlorination procedure.
Another avenue of investigation is based on a short-cut method for
the analysis of simple cyanides.33 This procedure involves the chlorination
of free cyanide or cyanide that h£.s been displaced from various metal-cyanide
compounds by EDTA. It may be possible to adapt this type of chlorination pro-
cedure for use here.
Either one or both of these approaches should be investigated in an
attempt to alleviate the deficiencies of this procedure with respect to
recoveries.
46
-------
100
80
60
LU
O
u
LU
tr
LU
o
QC
LU 40
0_
20
O Hg(CN)2
A Ni (CN}2
D K2 [Hg (CN)4]
0 K2[Ni(CN)4]
I
I
20 200
CONCENTRATION OF CN~ (ppb)
2000
SA-7854-9R
Figure 5.7 Recovery of cyanide from mercury and nickel cyanide compounds as a function of initial
cyanide concentration, using EPA CATC procedure.
47
-------
5.7.2.2 Interferences—
This method, as previously stated, uses the EPA method for total
cyanides (see Section 4.4). Therefore, it is subject to the same interferences
as the method for total cyanide, as shown by the results listed in Table 5.10.
A discussion of these interferences in given in Section 4.4.
5.7.3 Conclusions
The method of analysis "Cyanides Amenable to Chlorination," was origi-
nally designed to indicate the treatability of cyanides by the alkaline chlori-
nation process. It has become apparent during this study that this method
exhibits a number of deficiencies. There are primarily two major areas of con-
cern; these are (A) the method is subject to a number of interferences, and
(b) the method is unable to definitely classify some of the cyanide compounds
studied as either treatable or not treatable by the alkaline chlorination pro-
cess. This latter problem is most apparent when one attempts to classify the
compounds Hg(CN)2, Ni(CN)2, K2[Hg(CN)4], and K2[Ni(CN)4 ] ; in these cases, the
percentage of the compound chlorinated varies over a wide range and is directly
dependent on the initial concentration of the compound in the sample (see
Figure 5.7). It should be possible to alleviate some of these problems through
a more judicious choice of chlorination conditions and/or digestion-distillation
procedures.
5.7.4 Recommendations
Because of the widespread use of the EPA CATC method, its deficiencies
should be corrected. A different total cyanide method should be used to over-
come the interference problems (see Section 4.3). An evaluation to determine
the most efficient chlorination conditions is also recommended.
5.8 MODIFIED ROBERTS-JACKSON PROCEDURE
5.8.1 Introduction and Background
The procedure investigated in this study was a slight variation of the
Wood-River modification of the procedure developed by Roberts and Jackson,25
which has been reported to measure cyanides equivalent to "cyanides amenable to
chlorination." The method evaluated in this study is 'based on converting to hydro-
cyanic acid all but the most refractory metal-cyanide complexes from a slightly
acidified sample during a one-hour reflux distillation. The liberated gas (HCN)
is absorbed in a sodium hydroxide solution, which is subsequently analyzed for
cyanide either volumetrically, colorimetrically, or potentiometrically. The
procedure avoids the dissociation of iron cyanides, compounds not considered to
be amenable to chlorination, by the addition of zinc and lead acetates. The
Zn^+ and Pb^"*" presumably form insoluble double salts with the iron cyanides,
thereby preventing their decomposition.
Inclusion of such "fixing agents" in distillation procedures to prevent
the decomposition of the iron cyanides is not a new idea. The use of lead ni-
trates,26 lead acetate,27 and zinc acetate28 have been reported in earlier publi-
cations. For various reasons, the procedure developed by Roberts and Jackson
and later modified at Wood River has gained the widest acceptance, and it was
48
-------
TABLE 5.10 EFFECTS OF POTENTIAL INTERFERENCES ON CYANIDE RECOVERIES OBTAINED WITH
THE EPA PROCEDURE "CYANIDES AMENABLE TO CHLORINATION" USING AN ION-
SELECTIVE ELECTRODE FINISH
Apparent recovery of cyanide (%)
from designated compound
Level of
Potential interference interference
compounds mole:mole of CN
KSCN
Co(SCN)2
K2[Hg(SCN)4]
K3[Co(SCN)6]
n-butylthiocyanate
Oof (as KOCN)
NHi,+ (as NH^Cl)
N02~ (as NaN02)
Mn02
Butanal
S= (as Na2S)
C12 (as Ca.(OCl)/,)
^
100:1
1:1
100:1
1:1
100:1
KCN
19000
159
31400
128
17800
1:1 301
100:1 99000
1:1
100:1
1:1
100:1
1:1
100:1
1:1
100:1
1:1
100:1
1:1
100:1
1:1
10:1
1:1
10:1
1:1
320
5
50
100
-
57
100
66
100
99
-
55
92
2800
250
0
0
CuCN
19000
119
18800
66
-
-
-
-
_
-
99
-
96
94
81
99
99
-
97
-
4680
240
-0.5
0
Hg(CN)2
16000
200
27700
246
22500
342
93200
695
700
70
83
- 84
45
80
58
84
81
-
.,.««„
28000
120
24300
118
-
-
-
KJt[Fe(CN)6]
35000
120
39500
300
-
-
-
| -
_
-
103
-
106
-
103
-
109
i
93
-
3370
190
-11
-17
88
105
3770
185
-0.5
-0.4
_
-
-5
-
-5
-
1
-
-5
-
5
-
3420
75
-100
-98.6
Note: Cyanide concentration level was 0.2 ppm in all test solutions.
-------
a slight variation of this Wood River procedure that was investigated in our
laboratory. The modification, which was made at the request of the Project
Officer, consisted of substituting lead acetate for one-half of the zinc ace-
tate. The procedure used is described in Appendix B.
5.8.2 Results and Discussion
5.8.2.1 Cyanide Recoveries—
The cyanide recoveries from this method are shown in Table 5.11. A
useful method for the analysis of "simple" cyanides must meet two criteria.
First, the method must show complete recovery from all the metal cyanides equi-
valent to "cyanides amenable to chlorination." In addition, little or no re-
covery should be observed from the most refractory metal cyanides, which in-
includes K3[Fe(CN)6], K^[Fe(CN)6], and K3[Co(CX)6]. This method comes close to
fulfilling these requirements. The main deficiency is the incomplete recovery
of cyanide from the mercuric compounds. This is not a surprising result, since
the mercury compounds are quite stable (3it for Hg(CN)J7~ is Id1**). The recoveries
are inversely dependent on the concentration of the metal complex. The higher
recoveries observed at lower concentration are presumably due to the greater de-
grees of dissociation of the metal complex, which is expected in more dilute
solutions.
It should be possible, by lowering the pH, to improve recoveries of
CN~ from the mercury compounds without adversely affecting the response to the
iron cyanides. However, in a brief study, the same recoveries were obtained
at pH values of 3.8 and 2.5.
On the basis of the performance of the ligand-exchange procedure for
total cyanide, it was felt that improved recoveries of cyanide from the mercury
compounds would be possible if a selective sequestering agent was added. An
abbreviated study was conducted to evaluate this and it was found that in the
presence of chloride ion, as sodium chloride, the recoveries of cyanide from
Hg(CN)2 at 2 ppm CN increased from 22% to approximately 80%. These improved
recoveries are due to the displacement of cyanide by chloride to form the
insoluble HgCl2.
5.8.2.2 Interferences--
A number of experiments were conducted to determine the effect of
various compounds on the performance of the modified Roberts-Jackson method.
As shown in Table 5.12, the procedure is subject to only a relatively small
number of interferences, which are breifly discussed below.
MnQ?—At high levels, Mn02 appears to cause a slight positive inter-
ference, as observed during our analysis of samples of ferrocyanide. However,
the magnitude of this interference may be statistically insignificant.
Chlorine—Under the conditions of analysis, chlorine oxidizes all
cyanide species and thus produces very low recoveries. Since recovery of cya-
nide from ferrocyanide is already low, the presence of chlorine does not have
any adverse effect on this sample. Procedures for removing chlorine from the
sample before analysis are described in Appendix B.
50
-------
TABLE 5.11 CYANIDE RECOVERIES OBTAINED WITH THE MODIFIED ROBERTS-JACKSON PROCEDURE
Compound
Studied
KCN
Cd(CN)2
Co(CN)2
CuCN
Fe(CN)2
Fe(CN)
3
Hg(CN)2
Mn(CN)
f.
Ni(CN)
2
Zn(CN)
z.
K3[Cu(CN)A]
K2[N1(CN)/.]
K [Hg(CN) ]
K3[Fe(CN)6l
K4[Fe(CN)6]
K3[CO(CN)61
Concentration of CN~
0.002 ppin
n
4
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
Mean
Recovery
U)
115
120
120
115
120
125
140
135
120
125
155
135
153
0
0
0
Relative
Standard
Deviation
(%)
13
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
0.02 ppm
n
4
2
2
2
2
2
2
2
2
2
4
2
2
2
2
2
Mean
Recovery
(%)
104
110
51
102
100
100
100
100
100
100
101
100
100
7
6
0
Relative
Standard
Deviation
(%)
9
—
—
—
—
--
—
—
—
—
2
—
—
—
—
—
0.2 ppm
n
4
2
2
2
2
2
2
2
2
2
4
2
2
2
2
2
Mean
Recovery
(%)
100
100
7
99
100
100
82
100
100
100
99
100
74
2
1
0
Relative
Standard
Deviation
(%)
0
—
—
—
—
—
—
—
—
—
2
—
—
—
—
"
2 ppm
n
4
2
2
2
2
2
2
2
2
2
4
2
2
2
2
2
Mean
Recovery
(2)
100
100
9,34
96
96
100
22
100
100
100
100
100
75
1
1
0
1 1
Relative
Standard
Deviation
(%)
0
—
—
—
—
—
—
—
—
—
0
—
—
—
—
"
-------
TABLE 5.12 EFFECTS OF POTENTIAL INTERFERENCES ON CYANIDE RECOVERIES WITH
THE MODIFIED ROBERTS-JACKSON PROCEDURE
t-n
ho
Apparent recovery of cyanide (%)
from designated compound
Potential interference
SCN~ (as KSCN)
N02~ (as NaNO )
+ ^
NHi, (as NH^Cl)
OCN~ (as KOCN)
Mn02
Co(SCN)2
K2[Hg(SCN),1]
K3[Co(SCN)6]
Butanal
C12 (as Ca(OCl)2)
S~ (as Na2S)
n-butylthiocyanate
Level
mole: mole CN
100:1
100:1
100:1
100:1
100:1
100:1
10:1
100:1
100:1
100:1
10:1
1:1
10:1
1:1
100:1
10:1
KCS
100
100
100
100
96
72
100
0
62
100
0
0
100
100
4500
1500
CuCN
98
100
100
100
82
60
70
0
58
94
0
0
100
90
7400
1700
Hg(CN)2
84
82
68
56
68
68
0
100
75
0
0
100
100
9000
1900
K3[Cu(CNK]
100
108
100
100
100
88
60
0
95
100
0
0
92
100
10000
1400
Kit[Fe(CN)6]
0
0
0
0
0
0
0
0
0
0
0
0
0
10000
1100
Note: Cyanide concentration was 0.2 ppm in all test solutions.
-------
Suit'ide—Sulfide does not interfere with any of the analyses except
for samples containing mercury complexes. The sulfide is precipitated out of
the sample as the insoluble PbS. This precipitate is stable under the condi-
tions of analysis and hence does not interfere. The positive interference ob-
served during analysis of the Hg(CN7)2 sample is somewhat misleading. Recoveries
of only 82% have been found on samples containing only Hg(CN)2. In the presence
of sulfide, there is 100% recovery of CM. Apparently, the cyanide is displaced
from mercury by the formation of HgS. Any additional sulfide remaining after
this reaction is then removed from solution by precipitation as lead sulfide.
It should be stressed that no sulfide is found in the scrubber.
The absence of sulfide interference is significant. Sulfide should
still be removed from solution when the sample is collected by precipitation as
lead sulfide to prevent oxidation and autocatalytic reaction Xs'ith cyanide to
form thiocyanate. However, filtration to remove this precipitate is not neces-
sary. This avoids the loss of insoluble cyanides during the filtration step.
K2 [Hg(SCN) i> ] — This produces a substantial negative interference on
the analysis. Apparently, the tetracyanomercurate, or a mixed ligand system con-
sisting of cyanide and thiocyanate is formed. The reason for these incomplete
recoveries from mercury compounds has previously been discussed.
Co (SCN) 2 and K.3[Co(SCN)5] — These two compounds both exhibit a negative
interference for the same reason as K.2 [Hg(SCN) 4 ] . A cobalt-cyanide complex
forms that, because of its high stability, will not release cyanide during the
distillation.
Butylthiocyanate—During the distillation, butylthiocyanate is distil-
led over into the scrubber solution and interferes with the electrode finish.
Because of its limited solubility in water, this compound can probably be removed
by the extraction procedure described for the removal of fattv ar.-i.Hs,
5.8.3 Conclusions
The modified Roberts-Jackson procedure for the analysis of simple cya-
nides is far superior to any other method investigated. The procedure pro-
vides a clear distinction between the different types of cyanides in solution
and is relatively free of interferences. For those compounds that do interfere,
other pretreatment methods have been described that will alleviate the problem.
Most significantly, the interferences from sulfide and thiocyanate have been
completely eliminated. A deficiency of the procedure is the incomplete recovery
of cyanide from Hg(CN)2 and K2Hg(CN)it from samples with greater than 20 ppb CN~.
5.8.4 Recommendations
On the basis of its performance in this study, the modified Roberts-
Jackson method comes nearest to being a universal method for cyanides equivalent
to "cyanides amenable to chlorination." Its only serious deficiency is the in-
complete recovery of cyanide from the mercury complexes. This could probably
be overcome by the addition of a highly selective sequestering asent. such as
53
-------
TEP. Because of the low affinity of this ligand for ferric and ferrous ions, its
addition would probably not affect cyanide recoveries from the iron cyanide
complexes. Further work should be supported to evaluate such a modification.
5.9 COMPARISON ^F SIMPLE CYANIDE METHODS
Qualitative comparisons of the performance of those methods studied in
depth with respect to recoveries, interferences, and statistical parameters are
depicted in Tables 5.13, 5.14, and 5.15, respectively.
The best method of analysis for simple cyanides appears to be the
modified Roberts-Jackson method. The procedure not only gives complete
recovery of cyanide from most of the simple cyanides, but also is unaffected
by the presence of sulfide and thiocyanate. The other interferences can be
removed before distillation. Some additional development work is warranted
on the basis of these results, but not to the degree required by the other
methods.
As previously discussed, the procedure for cyanides amenable to
chlorination and the EDTA-aeration procedure could be improved by further
developmental work. The -major concern is improvement of cyanide recoveries
from the simple cyanides. Significant benefits could be obtained from some
developmental work in this area. However, the major limitationn of the CATC
procedure is its inability to accurately assess cyanide levels in the presence
of either sulfide or thiocyanate. It appears that these interferences can be
eliminated only with extensive developmental work.
-------
TABLE 5.13 COMPARISON OF CYANIDE RECOVERIES FROM THE SIMPLE CYANIDE METHODS
Compound Studied
KCN
Cd(CN)2
Co(CN)2
CuCN
Fe (CN) 2
Fe(CN)3
Hg(CN)2
Mn(CN)2
Ni(CN)2
Zn(CN)2
K3[Cu(CN)lt]
K2[Ni(CN)4]
K2[Hg(CN)1+]
K3[Fe(CN)5]
Klt[Fe(CN)6]
K3[Co(CN)6]
Concentration
2000
A
C
C
N
C
C
C
M
C
s
C
C
s
M
N
N
N
ppb
B
C
C
S
M
C
C
S
C
M
C
C
s
C
N
N
N
CN~
C
C
C
S
C
C
C
S
C
C
C
C
C
200
A
C
C
C
C
C
C
M
C
M
C
C
M
M j M
N
N
N
N
N
N
ppb
B
C
C
s
C
C
C
s
C
s
C
C
M
C
N
N
N
Level
CN~
C
C
C
s
C
C
C
M
C
C
C '
C
C
M
N
N
N
20
A
C
C
S
C
C
M
C
C
M
C
C
M
C
N
N
N
ppb
B
C
C
S
M
M
M
S
C
M
C
C
M
C
N
N
N
CN
C
C
C
M
C
C
C
C
C
C
C
C
C
C
N
N
N
A = EPA—Cyanides Amenable to Chlorination
B = EDTA-Aeration
C = Modified Roberts-Jacks on
N = None (<10%)
S = Slight (m-50%)
M = Moderate (51%-90%)
C = Complete (>90%)
55
-------
TABLE 5.14 COMPARISON OF INTERFERENCE EFFECTS
ON THE SIMPLE CYANIDE METHODS
Compound studied Level Modified
for interference mole:mole CN EPA EDTA-aeration Roberts-Jackson
SCN~
N02~
NH4+
OCN~
MnOa
Co(SCN) 2
K2[Hg(SCN)4]
K3[Co(SCN)5]
Butanal
C12 (as Ca(OCl)2)
S= (as Na2S)
Butylthiocyanate
100 : 1 ++ 0
10 : 1 ++ 0
100 : 1 0 0
100 : 1 0 0
100 : 1 0 0
100 : 1 0 0
100:1 ++
10:1 ++
10 : 1 -H-
100 : 1 -H-
10:1 ++
100:1
10:1
10 :.l
1:1
10 : 1 -H- ++a
1:1 -H- -H-a
100 : 1 -H- ++
10 : 1 ++ -H-
0
0
0
0
0
0
_
_
-
::
0
0
-H-
If present as PbS, there will be no interference.
++ = Severe positive interference.
+ = Slight to moderate positive interference.
0 = No interference.
- = Slight to moderate negative interference
— = Severe negative interference.
56
-------
TABLE 5.15 COMPARISON OF VARIOUS OPERATING PARAMETERS
FOR THE SIMPLE CYANIDE METHODS
Lower limit of detection
EPA
2
EDTA-aeration
5
Modified
Roberts- Jacks on
2
(ppb)
Relative standard devia-
tion above 20 ppb CN~ -8% -10%
Length of analysis
(minimum) 1.5 hr 2.5 hr 1.5 hr
57
-------
SECTION 6
THIOCYANATE METHODOLOGY
6.1 INTRODUCTION
In the early 1960s, Ayres and Baird29 reported an analytical method
for cupric ion _that_was based_ on__the_extraction of_a_ brightly colored.,, neutral
dithipcvanatodipyridylcopper(II) complex. * L_ater_ Danchik. and Bo.l.tz
adapted this procedure to use for thiocyanate analysis, and included quantita-
tion of low levels of flame atomic absorption spectroscopy of the copper.
The colorimetric procedure for thiocyanate was reinvestigated in
addition, evaluations were made of two new variations of this method that in-
volve either high performance liquid chromatography or graphite furnace atomic
absorption spectroscopy.
6.2 COLORIMETRIC METHODS OF ANALYSIS
In the presence of excess pyridine, the Cu(Py)2(SCN~)2 complex is readily
extracted into chloroform, with a distribution ratio reported by Ayres and Baird
of 10^. Figure 6.1 shows the visible absorption spectra of the chloroform ex-
tract at various thiocyanate concentrations. The spectrum is characterized by
a A at 407 nm with an extinction coefficient of 810 L mole"1cm"1. Conform-
ity to Beer's law was observed between 2-40 ppm SCN~. The chloroform extract
is stable for more than 24 hours if properly stored.
Quantitation based on this visible absorption is limited by the small
extinction coefficient of the complex. Assuming an original aqueous sample
volume of 20 mL , the lower limit of detection is 0.2 ppm SCN~ in the aqueous
sample.
An attempt was made to lower the detection limit by modifying the colori-
imetric procedure. The pyridine ligands do not directly contribute to the visi-
ble absorption spectrum of the complex, but they do exhibit a complex set of
aromatic H^-* II* UV bands between 230-270 nm, as shown in Figure 6.2. To make
use of the absorption bands fqr the analysis of thiocyanate, it is necessary to
use a solvent that has a cutoff wavelength sufficiently below the X of these
absorption bands. (The cutoff wavelength of chloroform is 250 nm.) In the
spectrum of the copper complex in methanol, the II —* II* UV bands of pyridine are
slightly broadened and are superimposed on the tail of an intense band at higher
30
"ihe copper is not extracted into CHCL, in the absence of excess pyridine.
Therefore, it may be that the species' that is actually extracted is
Cu(SCN~)2Pyit species.
58
-------
0.6
0.5
0.4
LLJ
U
£ 0.3
o
in
CD
0.2
0.1
3700
40 ppm
Blank
I
3900 4100 4300
WAVELENGTH (A)
4500
SA-7854-12
Figure 6.1 Absorption spectra of dithiocyanatodipyridylcopper(II) in chloroform.
59
-------
2500
2000
1500
u
1000
500
210
230
250
X (nm)
270
290
SA-7854-13
Figure 6.2 UV spectrum of pyridine in methanol.
60
-------
energy, giving the spectrum shown in Figure 6.3 with an e of 5,700 L M^cnT1 at
257 nm. Thus, a spectrophotometric analysis based on this UV band should pro-
vide approximately a three-fold increase in sensitivity compared with the
standard analysis at 407 nm.
This rather modest increase in sensitivity would not by itself extend
the range of the method down to the desired limit of 2 ppb SCN~. Other short-
comings^^ the method include: (1) increased sample handling necessary to
remove the chloroform and excess pyridine extracted into the chloroform, (2)
instability of this complex in the absence of excess pyridine.
An obvious way to increase the lower limit of detection of this method
is to use an aromatic amine whose extinction coefficient is much greater than
pyridine. Unfortunately, the coextraction of excess nonvolatile ligand into
the chloroform along with the copper complex was a recurring problem. The UV
bands are not shifted enough by complexation to permit one to distinguish be-
tween free and coordinated ligand. However, high performance liquid chroma-
tography appeared to offer a convenient means of separating the complex from
excess ligand and taking full advantage of the more intense absorbance in the
UV region of the spectrum.
6.3 HIGH PERFORMANCE LIQUID CHROMATOGRAPHY (HPLC)
A high performance liquid chromatographic (HPLC) analysis should sepa-
rate the copper complex from free ligand, thus allowing quantitative analysis
of the complex with a UV detector. In addition, the expanded scale of an HPLC
detector system, compared with the 1 absorbance until full-scale recorder on the
Gary 14, should increase the apparent sensitivity of the method. Finally, it
was considered desirable to investigate the possibility of achieving on-column
concentration of the sample by a judicious choice of solvent systems.
Chromatograms were recorded using a Du Pont 848 liquid chromatograph
equipped with a variable wavelength model 837 UV-Vis detector with an 8-mm
pathlength. Maximum sensitivity was 0.01 AU full scale, although excessive
noise limited the useful attenuation to 0.04 AU full scale. Taking into consid-
eration the difference in pathlengths, this represents a twentyfold scale ex-
pansion compared to. the Gary 14. A series of Waters and Du Pont columns were
used with a variety of solvent systems, as described below.
In the initial experiments, a reverse phase system was used, consisting
of a nonpolar permaphase ODS column and a polar 80% MeOH/20% H20 mobile phase.
The choice of solvents was based on the relatively high solubility of the
Cu(Py)2(SCN~)2 complex in this solvent system. Typical chromatograms are shown
in Figure 6.4. Injection of a methanolic solution of the copper complex resulted
in a solvent peak at 1.1 min, followed closely by a pyridine peak at 1.3 min.
No peak was observed that could be assigned to the copper complex. Since the
original sample contained no excess pyridine, the detection of the free ligand
in the analysis indicated that the copper complex was at least partially disso-
ciated in this solvent system.
The dissociation of the copper complex was confirmed by recording the
visible spectrum of Cu(Py)2(SCN~)2 in pure methanol and in an 80% MeOH/20% H20
61
-------
9000
8000
7000
6000 -
•7 5000 -
E
u
„ 4000 -
3000 -
2000 -
1000 -
210
290
SA-7854-14
Figure 6.3 UV_ spectr_um_of dithiocyanatodipyridylcopper(II) complex
In methanol.
62
-------
SA-7854-15R
Figure 6.4 KPLC chroma to grams of the dithiocyanatodipyridylcopper (II)
complex"." ~~ ... . . .
Column — permaphase ODS
Flow Rate - 1.25
Solvent - 80% methanol, 20% water
Sample Size - 50
Detector wavelength -260 nm
Samples dissolved in methanol
63
-------
mixture, as shown in Figure 6.5. Curve 1 represents the intact complex in
methanol, while Curve 2 shows the spectrum resulting from a 25% dilution with
H20. Instead of a straightforward reduction in absorbance due to dilution,
there is a 70% decrease in absorbance at 390 nm and a shift in Am_v to shorter
llld A
wavelengths.
Further study revealed that even in pure methanol there is a slow de-
crease in absorbance, with approximately a 20% loss in intensity over the first
1-1/2 hours. Presumably, the solvent is displacing pyridine from the inner
coordination sphere of the copper. In the absence of any coordinated amines,
cupric ion in the presence of thiocyanate is unstable with respect to reduction
to cuprous ion. Since the cuprous complex lacks any charge transfer spectrum,
there is a decrease in absorbance at 390 nm.
The second HPLC analysis was performed with a bondapak CIS column.
Chloroform was chosen as the solvent, since previous workers had established
that the copper complex is stable in this solvent in the presence of excess
pyridine.30 Pyridine itself elutes from this column with a retention time of
16.5 min. A solution of^_u_(Py)2(J5CN )2 dissolved in a pyridine/CHCl3 mixed
solvent gives only a single peak at 16.5 min. The addition of more solid
Cu(Py)2(SCN )2 to the sample solution caused a steady increase and slight shifting
of this band to "^15.5 min, as shown in Figure 6.6, but no additional peaks were
observed. The coelution of the complex and pyridine was ruled out on the basis
of the chromatogram of the most concentrated copper solution, with the detector
set to the 407 nm ^ax of the complex. No peaks were observed, indicating that
the increase in the 16.5 min peak observed with the detector at 260 nm is due
to free pyridine.
Our experience tends to corroborate previous reports that the copper
complex is stable in a pyridine/CHCla solvent. The apparent decomposition of
this complex during HPLC analysis could be due to strong interactions of cupric
ion with the column packing. Alternatively, excess pyridine may be a necessary
factor for long-term stability. Thus, the on-column separation of the complex
from free pyridine may lead to dissociation of coordinated pyridine.
These results do not prove that the complex is completely dissociated
during the analysis. Because of the rather low solubility of the copper complex
in pure chloroform, it is possible that the separation of excess pyridine could
also lead to on-column precipitation of the intact complex. No quantitative
data are available from which to estimate the percent of dissociation versus
precipitation.
The direct analysis of Cu(Py)2(SCN~)2 by liquid-solid chromatography has
also been investigated with a Waters' porasil column and 1:1 THF/CHC13 as the
mobile phase. Injection of pyridine dissolved in the mobile phase resulted in a
peak at 7 min that tailed somewhat, presumably because of interaction of the
basic pyridine nitrogen with the acidic column material. Injection of
Cu(Py)2(5CN-)2 dissolved in 1:1 THF/CHC13 produced no peaks at all. Switching
to a 50:50:1 dioxane/CHCl3/isopropanol mobile phase produced the same results.
Injection of the copper complex dissolved directly into the mobile phase with
no excess pyridine resulted in a blank chromatogram, with no peak for either
the complex or free pyridine. The retention time for pyridine was determined to
64
-------
LLI
CQ
DC
O
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
I f
360 380
400
X
420 440
SA-7854-16
Figure 6.5 "Absorption spectra of the dithiocyanatodipyridylccpper (II)
..complex in _.(!)_methanol... and- (2) 20 mL methanol and 5mL~water.
65
-------
LU
CD
CC.
o
CO
00
Cu(Py)(SCN-
-INJECT
10 12
RT
14
16
18
20
22
SA-7854-17
Figure 6.6 HPLC chrQmatograms_p.f__the,.._dithiocv_anatodipyridylcopper (II)
Column - Bondapak C18
Sample Size — 20
Flow Rate — 1 ~~
Solvent System — Chloroform
Detector Wavelength - 260 nm
66
-------
be 10 min by the separate injection of pyridine alone. Finally, pure chloroform
was used as the mobile phase, but it, too, failed to elute the copper complex
from the porasil column.
Because of the continuing difficulties in the analysis of Cu(Py)2(SCN~)2,
a new ligand, 1,10-phenanthroline, was substituted for pyridine. The choice of
this ligand was based on two factors. First, because it is bidentate, it forms
much more stable copper complexes than does pyridine, and thus is less likely
to dissociate from the metal ion under very dilute conditions. In addition,
phenanthroline has a very intense absorption band at 265 nm, as shown in
Figure 6.7, with an z of 28,000 LM " cm"1, which should increase the sensitivity
with a UV detector.
The addition of KSCN to a 1:1 solution of copper and phenanthroline re-
sulted in the immediate precipitation of the green mixed-ligand complex. How-
ever, unlike the bis(pyridine) analogue, the phenanthroline complex is virtually
insoluble in all common organic solvents. Thus, because it is not possible to
use the routine extraction into chloroform, the initial extraction of SCN~
from water was performed in the normal manner using pyridine. The chloroform
layer was then re-extracted with an aqueous solution of phenanthroline, re-
sulting in the formation of the blue Cu(phen)^ complex in the aqueous layer.
The absorption spectrum of this complex is also shown in Figure 6.7. Based on
an e per copper of 58,000 M~^cm~•*•, there is a potential thirtyfold increase in
sensitivity compared to the 407 nm band of Cu(Py)2(SCN )2«
Unlike the neutral Cu(Py)2(SCN )2 complex, Cu(phen)2+ is a divalent
cation so that different strategies apply to its analysis by HPLC. The primary
goal was the separation of the Cu(phen)2 complex from excess phenanthroline.
The most obvious approach was to use a reverse phase system that would retard
the hydrophobic ligand but would not interact strongly with the charged copper
complex. However, the injection of an aqueous solution of Cu(phen)2(1^3)2
onto a bondapak ODS column with a methanol mobile phase resulted in no peaks.
Further investigation revealed that phenanthroline itself was not eluted from
this column by methanol, even though pyridine was observed in 5 min. The effect
of increased molecular weight on retention time was estimated by injecting
anthracene, which was easily detected by 5.5~ min. Thus, the larger molecular
weight of phenanthroline does not account for its strong retention by this
column. Further attempts to elute the Cu(phen)2 complex using 20% MeOH/80%
0.01 phosphate buffer at pH 7 were also unsuccessful. Ion pair chromatography
was also investigated as a technique for analysis of the bis(phenanthroline)
copper(II) complex. The commercial reagent PIC B5 (n-pentanesulfonic acid)
was added to the methanol-phosphate mobile phase, and the pH was readjusted to
7.0. The anionic sulfonate groups should form ion pairs with the cationic copper
complex. It was hoped that this neutral species could then be eluted from a
reverse phase column—in this case, permaphase ODS. Unfortunately, this was not
the case, and neither the copper complex nor free phenanthroline could be
eluted from the column with this system.
The final attempt to analyze for the copper-phenanthroline complex in-
volved ion exchange chromatography using a Du Pont Zipax-SCX strong cation ex-
change column, with a pH 9.5 borate buffer as the mobile phase. However, as in
the previous methods, neither the free ligand nor the copper complex eluted from
the column. This column also retained a sample of bis(ethylenediamine)copper(II).
67
-------
60,000
50,000
7 40,000
u
s. 30,000
20,000
10,000
I
I
I
210 220 230 240
250 260
X (nm)
I
270 280 290 300
SA-7854-18
Figure 6.7 Spectra of Cu(phen). and phenanthroline in water.
68
-------
Thus, there appeared to be two problems! (1) some unexpected interaction of
phenanthroline itself with the packing material, which may be decomposing the
copper complex, and (2) very strong retention of divalent cations.
An obvious, recurring problem in the analysis of phenanthroline complexes
is the strong interaction of the free ligand with the packing materials. Thus,
phenanthroline does not appear to be a suitable ligand,' despite its strong
sequestering abilities and attractive uv spectrum. There are still other ligands
that might merit future consideration, such as the 0-diketonates and 3-ketoamines ,
which have been used previously to obtain analyses of -various metal ions, as well
as the ethylene-bis-(salicylaldimines) (salen) ligandsj which have not been used
in thj.3 capacity. However, any future programs must take into consideration
some serious problems recently discovered in the chloroform extraction of the
pyridine complex. These are discussed in detail in the following section on
atomic absorption spectroscopy.
6.4 ATOMIC ABSORPTION SPECTROSCOPY
In the proposed ASTM Method B,32 cupric nitrate and pyridine are added to
a thiocyanate sample, and this aqueous solution is extracted with CHCls- The_
SCN is quantitatively transferred to the chloroform layer as the Cu(Py)2(SCN )2
complex. Next, the CHC13 is removed and the copper complex is redissolved in
ethyl acetate. This solution is then analyzed by flame atomic absorption spec-
troscopy. The method is used when the original SCN sample is in the 0.05-2.0
ppm concentration range.
A revised mettiod that uses graphite furnace atomic absorption instead
of flame atomic absorption was studied. It was anticipated that this modi-
fication would have two beneficial effects. One, the sample could be analyzed
directly in the CHC13 matrix, eliminating the time-consuming change in solvent.
More important, however, is the greater sensitivity of graphite furnace AA, which
can detect copper in the low ppb range.
Data were collected using a Varian AA6 spectrometer equipped with a
graphite furnace attachment and 20-mL carbon rods. The program consisted of
drying at 100°C for 40 sec, ashing at 700°C for 20 sec, and atomization at 2200°C
for 2 sec. The ramp rate was 600°C/sec. Samples of 4-8 mL were slowly injected
after the carbon rod had reached the 100°C drying temperature to prevent the
chloroform from creeping out the ends of the tube. To keep the spectrometer
reading on-scale over the wide range of concentrations studied, it was necessary
to use different sample volumes. The reported intensities are normalized accord-
ing to the equation
intensity = Absorbance
Volume Injected
which gives intensity values in the general range of 1-200.
A bulk sample of Cu(Py)2 (SCN~) 2 was prepared by adding 0.483 g KSCN
(5 mmol) dissolved in a minimum volume of water to a 50-nL aqueous solution of
0.483 g CuCl2 (2.8 mmol) and 3 mL of pyridine. This resulted in the immediate
precipitation of a dark-green powder, which was filtered from solution, washed
with water, and air-dried.
69
-------
The copper content of this material was determined by weighing 20-30 mg
samples into 50-mL beakers, then covering and gently heating them for 2 hours
in vLO mL concentrated HN03. These solutions were then quantitatively transferred
to 50-mL volumetric flasks and diluted to volume with distilled water, giving
solutions with copper concentrations in the 4-8 ppm range, which were then
analyzed by flame AA using commercial copper standards. An average value of
21.5 ± 0.4% copper by weight was obtained for four samples (calculated for
CuC12HioN2S2 = 20.5%.)
A carefully weighed sample of Cu(Py)2(SCN~)2 was dissolved in 100 Ti'L of
95% CHCl3/5% pyridine solvent to give a stock solution that contained 21.6 pg
Cu/ml. Thereafter, daily calibration curves were prepared by first diluting this
stock solution lilOO with pyridine/CHCl3 to give a solution containing 0.216 yg
Cu/ml. This dilute solution was then used to prepare a series of standards.
needed to construct a calibration curve such as the one in Figure 6.8. The
slopes of these curves were fairly constant for 1 day. However, the intercepts
tended to drift, so that it was necessary to frequently recalibrate the instrument
during the analysis of several unknowns.
Based on the slope of the calibration curve and the observed reproduci-
bility of the spectrometer readings, one should be able to detect ^15 ppb copper.
This corresponds to 8 ppb SCN in the original aqueous solution. Initial efforts
to extract SCN from aqueous solutions and analyze the CHC13 layer by this method
gave very erratic results. The background level of copper being extracted into
the CHC13 was determined by using 10 mL of CHC13 to extract samples of 0.00262 M
Cu(N03)2 containing various amounts of pyridine but no thiocyanate. The results
are shown in Table 6.1.
TABLE 6.1 EXTRACTION OF COPPER-PYRIDINE SOLUTIONS WITH CHLOROFORM
IN THE ABSENCE OF THIOCYANATE
Sample No.
1
2
3
4
5
6
ml Pyridine
1.0
1.0
1.0
1.0
0.5
0.5
AA Intensity
137 + 8
150 + 5
125 + 6
182 + 6
63+7
69+2
ppb Cu
188 +
200 +
162 +
249 +
66 +
76 +
(est.
12
8
9
9
11
4
Intensity =0.65 (ppb Cu) +19.9
Clearly, it is not possible to obtain accurate low ppb analyses when the
background fluctuates from 150-250 ppb, with an average of 200 ± 40 ppb Cu. It
70
-------
200
Slope = 0.65 Au/ppb
int = 19.9
40
80 120 160 200
CONCENTRATION OF Cu(ppb)
240 280
SA-7854-19
Figure 6.8 Graphite furnace calibration curve of dithiocyanatodipyridylcopperfll).
71
-------
would be necessary to have at least 80 ppb copper just to be significantly above
the noise level. This corresponds to 73 ppb SCN~ as the Cu(Py)2(SCN~)2 complex.
The results for samples 5 and 6 "clearly establish a correlation between
the background copper level and the concentration of pyridine in the aqueous layer.
An effort was made to reduce the background copper to an acceptable level by
extracting 25-mL samples of 101 ppb SCN~ following the addition of only 0.5 mL__
pyridine and various amounts of copper. These results are shown in Table 6."2.
TABLE 6.2 EXTRACTION EFFICIENCY OF DITHIOCYANATODIPYRIDYLCOPPER(II)
BY CHLOROFORM
Aqueous Copper Concentration
First 10-mL extract
blank
net
n
% recovery of SCN"
Second extraction (ppb Cu)
ppb
0.00262 M
131
71
60
48%
20
Cu in 10 mL Q
0.00116 M
82
38
44
35%
19
l\j -L f)
0.00604 M
57
32
25
20%
20
Calculated assuming all copper in first 10-mL extract was Cu(Py)2(SCN )*.
The results indicate that adding 0.5 mL of pyridine does not result in complete
extraction of the SCN , and that any attempts to reduce background levels by
reducing the initial copper concentration also have an adverse effect on the
extraction efficiency. Multiple extractions are of limited use in obtaining
complete extraction of the copper complex due to the almost complete removal of
free pyridine in the initial 10-mL extraction.
Despite the low recovery of SCN , an attempt was made to obtain a cali-
bration curve by_extracting a series of standard KSCN solutions using 0.00116 M
copper and 0.5 mL pyridine. However, the points were widely scattered and
indicated that although the level of 38 ppb Cu reported in Table 6.2 is close to
the average, the individual background levels range from 20 to 70 ppb. Any
further reduction in copper or pyridine concentrations does not appear to be
reasonable in light of the already low extraction efficiency for SCN .
The results discussed above all indicate that standard extraction procedures
simply are not acceptable for use with thiocyanate at the low ppb level. Although
it is still not known exactly what species are being extracted into chloroform,
there is clearly at least one copper-pyridine complex that is slightly soluble
in CHC13. Pyridine itself is very soluble in chloroform. It was felt that by
adding substituents to pyridine that reduce its extraction coefficient, one might
also impede the extraction of its copper complexes as well. Thus, 50-mL
72
-------
aliquots of 0.00143 M copper solution were extracted with 10 TiL of CHC13 after
the addition of 12.4 mmoles of the appropriate pyridyl ligand (L/Cu = 170) but
no thiocyanate. The results are shown in Table 6.3.
TABLE 6.3a EXTRACTION OF CUPRIC IO\T BY PYRIDYL LIGANDS
IN THE ABSENCE OF THIOCYANATE
Ligand ppb Copper
Pyridine 350
2-Aminopyridlne 580
3,5-Dichloropyridine 160
2-Hydroxy-5-nitropyridine
2,6-Diaminopyridine 660
S50mL"of 1.43 MM Cu(N03)2 extracted into lo'_mL~ CHC13
in the presence of 12.4 mmoles of ligand.
This ligand is virtually insoluble in water.
Both the 2,6-diaminopyridine and 3,5-dichloropyridine are difficult to
work with because of their limited water solubility. A large portion of each
of these ligands failed to dissolve in the original 50-mL .aqueous layer, but the
solid 3,5-dichloropyridine did dissolve in the chloroform layer. The 2-hydroxy-
5-nitropyridine was so insoluble in water that no analysis of the chloroform layei
was conducted. It is obvious that none of the substituted pyridines are capable
of reducing the copper background to an acceptable level.
Since the atomic absorption method does not depend on the intense color
associated with the pyridine complex, aliphatic amines were also considered as
extracting agents. The usual extraction procedure was followed using the amines
listed in Table 6.4.
73
-------
TABLE 6.4 COPPER EXTRACTED BY ALIPHATIC AMINES AT pH 7
ppb Cu Extracted5
Lieand ppb SCN Cu + SCN Cu only Net
NH*
n-Butylamine
Diethylamine
Trie thy lamine
327
503
503
503
8
42
10
14
4
32
7
18
4
10
3
-3
50-~mL "samples; 0.00135 M Cu(NC>3)2
bAdded 0.9mL concentration NH^OH.
Added 2.0mL neat amine.
Although these ligands were very successful in reducing the copper background
to an acceptable level, they do not facilitate the extraction of thiocyanate
into the organic layer. The color of these solutions at pH 7 is quite pale
compared to their intense blue color at higher pH. Thus, it is not clear whether
their failure to extract SCN is due to a low distribution coefficient of the
Cu(SCN )2(amine)2 complex or to the low degree of formation of such z. complex.
6.5 CONCLUSIONS AND RECOMMENDATIONS
^Thiocyanate can be readily removed from the sample matrix by extraction
of the Cu(P_y)2^_SCN _) 2 complex. The copper can be quantitated by atomic
absorption, and the concentration of SCN can be determined using the 2:1
stoichiometric ratio of SCN :Cu. Carbon rod AA analysis of copper is suffi-
ciently sensitive to permit quantitation of SCN at 2 ppb if one can obtain a
two- to threefold concentration in the extraction step. However, the excess
pyridine extracts additional copper in addition to that bound to thiocyanate..
This simultaneous extraction of excess copper must be drastically reduced, and
this appears to be a formidable task. Any ligand that forms an extractable
mixed-ligand complex with copper thiocyanate is also likely to form soluble
binary copper complexes as well. Thus, it is necessary to find a ligand that
will extract the thiocyanate complex but still give a low background of copper.
The results presented above indicate that pyridine derivatives are not
suitable for SCN extractions at low levels. Although the results for the
aliphatic amines are not particularly encouraging, to date there has been no
systematic investigation of the effects of pH on this process. One can contin-
uously vary the effective binding constant of these amines by varying the pH.
A multitude of other compounds also deserve consideration, such as substituted
bidentate diamines, aniline derivatives, or imidazole derivatives. Other sol-
vents may also be considered. It is probable, however, that any program intended
74
-------
to optimize an extraction procedure adaptable to a low ppb SCN"
quire a considerable level of effort.
analysis will re-
An improvement of the HPLC method of analysis also requires a suitable
extraction procedure to separate the SCN~ from the sample matrix. Assuming such
a procedure is developed, HPLC still has high potential as a method of quanti-
tating SCN . The best approach is to react the extracted mixed-ligand complex
with a suitable multidentate ligand to form a stable, easily detected copper
complex. This was the approach taken here using phenanthroline, which appears
to be an unfortunate choice because of its unexpected interaction with the
packing materials. However, a number of other ligands could be useful, such as
those shown below.
II
III
IV
Ligands I-IV all form neutral complexes with cupric ion. In addition,
it is fairly straightforward to add substituents to the aromatic rings so as to
alter the solubility of the complex as needed and also to increase the absorp-
tivity in the uv region; the latter would serve as a convenient basis for
detection.
Another method that offers the possibility of direct analysis of SCN
without an extraction step is ion chromatography. Because of its size, SCN is
strongly retained on most columns and thus easily separated from common anions such
as Cl , Br , and No3. By use of a precolumn, significant concentration of the
sample is possible. This technique has allowed analysis of Cl at the 5-ppb level.
To alleviate the interferences from metal ions, a suitable chelating agent could
be added to sequester the metals and free coordinated SCN . The most likely
choices would be polyamines, since they are cations at neutral pH and thus would
not be retained by the column.
75
-------
REFERENCES
1. W. F. Linde and A. Seidell, Solubilities of Inorganic Compounds, 4th edi-
tion (American Chemical Society , "Washington, _ DC, 1965).
2. R. E. West, Editor, Handbook of Chemistry and Physics, 54th edition (CRC
Press, Cleveland, Ohio, 1974).
3. R. G. Kunz, J. P. Casey, and J. E. Huff, Hydrocarbon Processing (October
1978), pp. 98-106.
4. L. S. Sillen and A. E. Kartell, Stability Constants of Metal-Ion Complexes
Supplement No. 1, The Chemical Society, London 1971.
5 . EPA Methods for Chemical Analysis of Water and Wastes, Environmental Moni-
toring and Support Laboratory, Cincinnati, Ohio (1979).
6. Standard Methods for the Examination of Water and Wastewater, 14th edition
(American Public Health Assoc. , Washington, DC, 1976), pp. 361.
7. 1978 Annual Book of ASTM Standards (American Society for Testing and Mater-
ials, East on, Maryland, 1978), p. 583.
8. C.J.L. Lock and G. Wilkenson, J. Chem. Soc. , 2281 (1962).
9. B. Jaselskis and J. G. Lanese, Anal. Chim. Acta, 23, 6, (I960).
10. B. L. Gilbert, B. L. Olson, and W. Revter, Anal. Chem., 46, 170 (1974).
11. P. D. Goulden, B. K. Afgan, and P. Brooksbank, Anal. Chem., 44, 1845 (1972).
12. G._W. Latimer, Jr., L. R. Payne, and M. Smith, Anal. Chem., 46_, 311 (1974)_._
13. j_._N._P.. .Kelada.,.. C... Lue^Hins, ami D. _T._Lprdi, "Cyanid_e_Sp_acie3_.and Thiocyanate---
Methodology in Water and Wastev/ater . " The Metropolitan Sanitary District of
____ Gr.eater_Chi.cago_, _ Re.p_o.r.t_No_._7_7_^2Jl,_CAug^_19_7J.) _______________
14. C. T. Elly, J. Water Pollution Contr. Fed., 40_, 848 (1968).
15. "P. J. Barton, C. A. Hammer, and D. C. Kennedy, J. Water Pollution Contr. Fed.,
_50, 234 (1978).
16. H. E. Williams, Soc. Chem. Lnd. Jour., 31, 468 (1912).
17. E. J. Ludzack, Anal. Chem., 26_, 1978 (1954).
18. J. M.Kruse and M. G. Mellon, Anal. Chem., _25_, :-A4~6_(1953) .
76
-------
19. M. S. Frant, J. W. Ross, and J. H. Riseman, Anal. Chem., 44, 2227 (1972).
20. I. Sekerka and J. F. Lechner, Water Research, 10, 479 (1976).
21. A Schleuter, U.S. NTIS, PB Report, 1976 PB 255852, available from Gov.
Rep. Announce Index (U.S.) 76(21), 187 (1976).
22. I. Gilath, Anal. Chem., 49, 516 (1977).
23. Private communication from C. Caruso, Carneigie Mellon Institute.
24. Private communication from L. E. Lancy, Lancy Co., Zelienople, Pennylvania.
25. R. F. Roberts and B. Jackson, Analyst, 96, 209 (1971).
26. F. L. Ludzack, W. A. Moore, and C. C. Ruchhoft, Anal. Chem., 26_, 1784 (1954).
27. Official Standardized and Recommended Methods of Analysis, S. C. Jolly,
Editor (W. Heffer and Sons, Ltd., 1963), p. 252.
28. F. R. Russel and N. T. Wilkinson, Analyst, 84, 751 (1959).
29. G. H. Ayers and S. S. Baird, Talanta, 7_, 237 (1961).
30. F. J. Welc'ner, Organic Analytical Reagents, Volume III, (D. van Nostrand
Co., Inc., New York, 1947), p. 18.
31. R. S. Danchik and D. F. Boltz, Anal. Chem., _40, 2215 (1968).
32. New Method of Test for Thiocyanate in Water, reported by ASTM Committee D-19,
Subcommittee D-19.06, L. E. Lancy, Chairman, April 1979.
33. Colorimetric Method for Cyanides Amenable to Chlorination and Thiocyanate,
Without Distillation, reported by ASTM Committee D-19, Subcommittee D-19.06,
L. E. Lancy, Chairman, 1979.
77
-------
-------
Appendix A
EDTA AERATION PROCEDURE FOR SIMPLE CYANIDES
A-l
-------
1. Scope and Application
1.1 This method is applicable to the determination of cyanides in drink-
ing, surface, and saline waters, and domestic and industrial wastes.
1.2 The titration procedure using silver nitrate with p-dimethylamino-
benzal-rhodanine indicator is used for measuring concentrations of
cyanide exceeding 1 mg/L (0.1 mg/100mL of absorbing liquid).
1.3 The colorimetric procedure is used for concentrations below 1 mg/niL
and is sensitive to about 0.02 mg/L.
1.4 The potentiometric procedure is used for concentrations between 26
and 0.02 mg/L. The lower limit can be extended through the use of
a calibration curve to 0.002 mg/L. Higher concentrations can be
measured but since these cause increased erosion of the membrane,
measurements above 26 mg/L cyanide should be done only occasionally.
2. Summary of method
2.1 The cyanide as hydrocyanic acid (HCN) is released from cyanide com-
pounds by means of a 2-hour aeration in the presence of EDTA at pH 4.5.
The cyanide so released is absorbed in a scrubber containing sodium
hydroxide solution. The cyanide ion in the absorbing solution is
quantitated volumetrically, colorimetrically, or electrometrically.
2.2 The titrimetric measurement uses a standard solution of silver
nitrate to titrate cyanide in the presence of a silver sensitive
indicator.
2.3 In the colorimetric measurement the cyanide is converted to cyanogen
chloride (CNC1) by reaction with chloramine-T at a pH less than 8
without hydrolyzing to the cyanate. After the reaction is complete,
color is formed on the addition of a pyridine-pyrazalone or a pyri-
dine-barbituric acid reagent. The absorbance of the resulting
colored solution is read at 620 nm when using pyridine-pyrazalone
and 578 nm when using the pyridine-barbituric acid reagent. It is
essential to have comparable ionic strenghs in both the sample and
the standards.
2.4 The potentiometric measurement uses a cyanide ion selective electrode
and double junction reference electrode to quantitate the cyanide'
ion. It is essential that both the sample and the standard have
comparable ionic strengths.
3. Definitions
" Simple^cya"nides~in thTs'metno'd" are defined as cvanide inn
" - - --
dissocia^ted_ complex cyanide .converted, to hydrp.cvanic. acid _hv .reaction_ . ._
eration system_of an__acetate buffer in the presence of EDTA.
A-2
-------
4. Sample handling and preservation
4.1 The sample should be collected in plastic or glass bottles of 1 liter
or larger size. All bottles must be thoroughly cleansed and rinsed
to remove soluble material.
4.2 Oxidizing agents such as chlorine decompose most of the cyanides.
Test a drop of sample with potassium iodide-starch test paper (KI-
starch paper); a blue color indicates the need for treatment. Add
ascorbic acid, a few crystals at a time, until a drop of sample pro-
duces no color on the indicator paper. Add an additional 0.6 g
ascorbic acid for each liter of sample volume.
4.3 Sulfides slowly convert the cyanide in the sample to thiocyanate.
The reaction rate is greatly increased at high pH. Sulfide there-
fore interferes and should be removed as soon as the sample is col-
lected and before adjustment of the pH. When sulfides are present
in the sample, it may be assumed that oxidizing agents are absent.
Test for the presence of sulfide by placing a drop of the sample on
a strip of lead acetate test paper that has been previously moist-
ened with the acetic acid solution. Darkening of the test paper
indicates the presence of sulfide.
4.3.1 Sulfide is removed by treating the sample with small incre-
ments of powdered lead carbonate (PbCC^), cadmium carbonate
(CdCOs), or with the dropwise addition of lead nitrate
[Pb(N0_)9] solution. (When significant quantities of sulfide
• •""must 6~e "removed, the addition of PbC03, or CdC03 is preferred.
Pb (103)2 maY unduly depress the pH, and with Pb(OAc)2 addi-
tions, the acetic acid that distills over may neutralize
too much NaOH in the absorber). Black PbS precipitates in
samples containing sulfide. Repeat the operation until no
more lead sulfide forms, as indicated by testing the super-
natant liquid with Pb(OAc)2 test paper. It is not necessary
to filter the sample because S=, as PbS, does not interfere.
4.4 Samples must be preserved with 2 mL of ION sodium hydroxide per liter
of sample (pH 12) at the time of sample collection. (Additional
base may be necessary to ensure that the pH of the sample is ^13.)
4.5 Samples should be analyzed as soon as possible. If storage is re-
quired, the samples should be stored in a refrigerator or in an ice
chest filled with water and ice to maintain the temperature at 4°C.
4.6 Minimize exposure of the samples to ultraviolet radiation. Photo-
decomposition of the iron cyanides may significantly increase the
cyanide content of the sample.- (Remove interferences in the hood
under incandescent light conditions, etc.)
5. Interferences
5.1 Interferences are eliminated or reduced by using the aeration pro-
cedure described in Sections 8.1 and 8.7.
A-3
-------
5.2 Organic thiocyanates will be transferred to the scrubber solution
by the air stream. These compounds adversely affect quantitation
by electrochemical means. (It may be possible to remove these com-
pounds from the sample prior to aeration by the extraction procedure
described for fatty acid removal.)
5.3 Aldehydes react with cyanide to produce nitriles under the analysis
conditions. These nitriles are further hydrolyzed to their corres-
ponding acid and ammonia.
5.4 The presence of heavy metals (e.g., Kg"1"*) inhibit the transfer of
cyanide to the scrubber solution by forming stable mercury-cyanide
compounds.
5.5 Other possible interferences include substances that might otherwise
contribute color or turbidity. In most cases, the aeration pro-
cedure will remove these.
6. Apparatus
*
6.1 The aeration apparatus is shown in Figure 1. The boiling flask
should be of 1-liter size with an inlet tube and provision for a
condenser.
6.2 Microburet, 5.0mL (for titration).
6.3 Spectrophotometer suitable for measurement at 578 nm or 620 nm with
a 1.0 cm cell or larger.
6.4 A cyanide ion selective electrode, a double junction reference elec-
trode, an expanded scale mV meter or specific ion meter, and a mag-
netic mixer with TFE fluorocarbon-coated stirring bar.
7. Reagents
7.1 Sodium hydroxide solution, 1.25 N: Dissolve 50 g NaOH in distilled
water and dilute to 1 liter with distilled water.
7.2 Cadmium carbonate: Powdered.
7.3 Ascorbic acid: Crystals.
7.4 Dilute sodium hydroxide solution, 0.25 N: Dilute 200 mL of sodium
hydroxide solution (7.1) to 1 liter with distilled water.
7.5 Methyl red indicator: Prepared as 0.02 g in 60 ml H20 and 40 ml
acetic acid.
Figure 1 is given at the end of the Appendix.
/
A-4
-------
7.6 Sodium dihydrogenphosphate, 1 M: Dissolve 138 g Na^PO^-l^O in
1 liter of distilled water. Refrigerate this solution.
7.7 Stock cyanide solution, 1 mg/:mL': Dissolve 2.51 g KCN and 2 g KOH
in 1 liter distilled water. Standardize with 0.0192N AgN03. Dilute
to appropriate concentration so that 1 mL"~= 1 mg CN~.
7.8 Standard cyanide solution, intermediate, 50 mg/L: Dilute 50.0mL of
stock solution (7.7) to 1 liter with distilled water.
7.9 Standard cyanide solution, 5 mg/L: Prepare fresh daily by diluting
100.0mL~of intermediate solution (7.8) to 1 liter with distilled
water and store in a glass-stoppered bottle.
7.10 Standard silver nitrate solution, 0.0192 N: Prepare by crushing
approximately 5 g AgNOa crystals and drying to constant weight at
40°C. Weigh out 3.2647 g_dried AgNOj, dissolve in distilled water,
and dilute to 1 liter (1 ~m'L_ = 1 mg CN~) .
7.11 Rhodanine indicator: Dissolve 20 mg p-dimethylaminobenzalrhodanine
in 100 mL of acetone.
7.12 Chloramine-T solution: Dissolve 1 g of white, water-soluble chlora-
mine-T in 100 mL of distilled water and refrigerate until ready to
use. Prepare fresh weekly.
7.13 Color reagent - One of the following may be used:
7.13.1 Pyridine-barbituric acid: Place 15 g of barbituric acid in
a 250-mL volumetric flask and add just enough distilled
water to wash the sides of the flask and wet the barbituric
acid. Add 75 mL of pyridine and mix. Add 15 mL of HC1
(sp gr 1.19); mix and cool to room temperature. Dilute to
250 mL with distilled water and mix. This reagent is stable
for approximately six months if stored in a cool, dark place,
7.13.2 Pyridine-pyrazolone solution:
7.13.2.1 3-methyl-l-phenyl-2-pyrazolin-5-one reagent
saturated solution: Add 0.25 g of this compound
to a 50 mL distilled water; heat to 60°C with
stirring. Cool to room temperature.
7.13.2.2 3,3'-dimethyl-l,l'-dipenyl-(4,4I-bi-2-Pyrazoline)-
5,5'-dione (bispyrazolone): Dissolve 0.01 g of
bispyrazolone in 10 mL pyridine.
7.13.2.3 Pour solution (7.13.2.1) through non-acid-washed
filter paper. Collect the filtrate. Through
the same filter pour solution (7.13.2.2), col-
lecting the filtrate in the same container as
the filtrate from-(7.13.2.1). Mix until the
A-5
-------
filtrates are homogeneous. The mixed reagent
develops a pink color but does not affect the
color production with cyanide if used within
24 hours of preparation.
7.14 Acetate buffer solution: Dissolve 410 g sodium acetate trihydrate
(NaC2H302'3H20) in 500'mL of water. Add glacial acetic acid to
pH 4.5, approximately 500 mL.
7.15 EDTA solution: Dissolve 66 g disodium(ethylenedinitrilo)tetra-
acetate dihydrate (C1oH1i+N2Na208'2H20) in 1 liter water.
8. Procedure
8.1 Place 500 mL...of sample, or an aliquot diluted to 500mL , in the
1-liter boiling flask. Add 50 mL of sodium hydroxide (7.1) to the
absorbing tube and dilute if necessary with distilled water to
obtain an adequate depth of liquid in the absorber. Connect the
boiling flask, condenser, absorber, and trap in the vacuum train.
8.2 Turn on the vacuum and adjust the air flow rate through the flask
to approximately 3 liters per minute or greater.
8.3 Add 2-3 drops of the methyl red indicator to the reaction flask.
8.4 Add 10 mL each of the acetate buffer (7.14) and EDTA solutions (7.15)
through the air inlet tube.
8.5 Rinse the air inlet tube with a few ml of water and allow the air
flow to mix with the contents of the flask. (If the solution is
not pink, add acetic acid dropwise through the air inlet tube until
there is a permanent color change.)
8.6 Allow the air to flow through the sample for 2 hours at the rate of
at least 3 liters per minute. (Do not heat the sample.)
8.7 After 2 hours, stop the air flow and quantitatively transfer the
absorption liquid to a 250JJiL volumetric flask.
8.8 This solution, or an aliquot of this solution, is then analyzed for
cyanide using the colorimetric, titrimetric, or potentiometric method
of analysis.
8.9 Titrimetric method of analysis
8.9.1 If the sample contains more than 1 mg of CN , transfer the
distillate, or a suitable aliquot diluted to 250 mL to a
500-mL Erlenmeyer flask. Add 10-12 drops • of the rhodanine
indicator (7.11).
8.9.2 Titrate with standard silver nitrate (7.10) to the first
change in color from yellow to brownish-pink. Titrate a
A-6
-------
distilled water blank using the same amount of sodium
hydroxide and indicator as in the sample.
8.9.3 The analyst should familiarize himself (herself) with the
end point of the titration and the amount of indicator to
be_used before actually titrating the samples. A 5 or
10 mL microburet may be conveniently used to obtain a more
precise titration.
8.10 Colorimetric method of analysis
8.10.1 Withdraw 50_mLor less of the solution from the flask an_d_
transfer to a 100-niL volumetric flask. If less than 50 mL
is taken, dilute to 50 mL with 0.25 N sodium hydroxide
(7.4). Add 15 mL of sodium phosphate solution (7.6) and mix.
8.10.1.1 Pyridine-barbituric acid method: Add 2 mL of
chloramine-T (7.12) and mix. After 1 to 2 minutes,
add 5 mL of pyridine-barbituric acid solution
(7.13.1) and mix. Dilute to mark with distilled
water and mix again. Allow 8 minutes for color
development and then read absorbance at 578 nm
in a 1-cm cell within 15 minutes.
8.10.1.2 Pyridine-pyrazolone method: Add 0.5inL chlora-
mine-T (7.12) and mix. After 1 to 2 minutes add
5mL of pyridine-pyrazolone solution (7.13.2)
and mix. Dilute to mark with distilled water
and mix again. After 40 minutes read the absor-
bance at 620 mn in a 1-cm cell. NOTE: More
than 0.5 mL chloramine-T will prevent the color
from developing with pyridine-pyrazolone.
8.10.2 Prepare a series of standards by pipetting suitable volumes
of standard solution into 250-mL volumetric flasks. To each
.___standard add 50mL of 1.25 N sodium hydroxide and dilute to
250 mL with distilled water. Prepare as follows:
mL of Standard Solution Concentration_CN
(5 ug/mL CN") (mg CN~/250'm'L~)
0 blank
1 0.005
2.0 0.010
5.0 0.025
10.0 0.050
15.0 0.075
20.0 " 0.100
8.10.2.1 Prepare a standard curve by plotting absorbance
of standard versus cyanide concentration.
A-7
-------
8.10.2.2 It is not necessary that all standards be dis-
tilled in the same manner as the samples. It is
recommended that at least two standards (a high
and low) be distilled and compared to similar
values on the curve to ensure that the distil-
lation technique is reliable. If distilled
standards do not agree within ±10% of the undis-
tilled standards, the operator should find the
cause of the apparent error before proceeding.
8.10.3 To check the efficiency of the sample distillation, add an
increment of cyanide from either the intermediate standard
(7.8) or the working standard (7.9) to ensure a level of
20 yg/L or a significant increase in absorbance value. Pro-
ceed with the analysis as in Section 8.1, using the same
flask and system from which the previous sample was just
distilled.
8.11 Potentiometric method of analysis
8.11.1 Prepare a series of standards by pipeting suitable volumes
of a standard solution_into 250-mL volumetric flasks. To
each standard add 50 mL of 1.25 X NaOH and dilute to 250_mL
with distilled water. Prepare as follows:
Concentration. CM
mL of Standard Solution (yg CN~/250.mL )
(1 pg/mL~CN~)
2 2
3 3
5 5
10 10
(10 iig/m_L_CN~)
2 20
5 50
10 100
(100 yg/mL~ CN~)
2 200
5 500
10 1000
8.11.1.1 Transfer the standard solutions into 150-_mL
beakers prerinsed with a small portion of the
standard being tested. Immerse the cyanide and
double junction reference electrodes in the solu-
tion and mix well on a magnetic stirrer. Main-
tain as closely as possible the same stirring
rate and temperature for all solutions.
A-8
-------
8.11.1.2 After equilibrium is reached (at least 5 minutes
and not more than 10 minutes) record the milli-
volt reading and plot the CN~ concentrations
versus millivolt reading on semi logarithmic graph
paper. A straight line with a slope of 59 mV
indicates that the instrument is operating
properly.
8.11.1.3 It is not imperative that all standards be dis-
tilled in the same manner as the samples. It
is recommended that at least two standards (a
high and low) be distilled and compared to simi-
lar values on the curve to ensure that the dis-
tillation technique is reliable. If the distil-
led standards do not agree within ±10% of the
undistilled standards, the operator should find
the cause of the apparent error before proceeding.
8.11.2 Place the absorption liquid into a 150-mL beaker and proceed
with the analysis as in Section 8.9.1. Determine the CN~
concentration by observing the millivolt reading and refer-
ring to the calibration curve established in Section 8.9.1.
The method of known addition can be used for measuring
occasional samples, since the preparation of a calibration
curve is not required.
8.11.3 To check the efficiency of the sample distillation, add an
increment of cyanide from either the intermediate standard
(7.8) or the working standard (7.9) to ensure a level of
20 ug/L or a significant increase in values . Proceed with
the analysis as in Section 8.1, using the same flask and
system from which the previous sample was just distilled.
9. Calculations
9.1 Using the titrimetric procedure, calculate the concentration of CN
as follows :
CN" (me/L) = _ (A-B) (1000) (250) _
g CnL of original sample) (ml of aliquot titrated)
where:
A = volume of AgNOs for titration of sample
B = volume of AgN03 for titration of blank
9.2 Using the colorimetric procedure, calculate the concentration of CN
as follows :
rw
CN
( 0/^ - (A) (1000) (250)
(yg/L) - - - -
A-9
-------
where:
A = yg CN read from standard curve
B = mL of original sample taken for distillation
C = mL of scrubber solution taken for colorimetric analysis.
9.3 Using the potentiometric procedure, calculate the concentration of
CN7~ as follows:
or (ug/D - (A) dooo) (250)
where:
A = vg CN read from standard curve
B - mL of original sample taken for analysis
A-10
-------
In
Cooling Water
Inlet Tube
Screw
Clamp
To Low
Vacuum
Source
250 mL ..Absorber
Medium
Porosity Frit
Distillation Flask
Figure 1 AISI AERATION APPARATUS.
SA-7854-10
A-11
-------
-------
Appendix B
MODIFIED ROBERTS-JACKSON METHOD FOR ANALYSIS OF SIMPLE CYANIDES
B-l
-------
1. Scope and application
1.1 This method is applicable to the determination of easily dissociated
cyanides in drinking, surface, and saline waters, and domestic and
industrial wastes.
1.2 The titration procedure using silver nitrate with p-dimethylamino-
benzal-rhodanine indicator is used for measuring concentrations of
cyanide exceeding 1 mg/L (0.1 mg/100.mL of absorbing liquid).
1.3 The colorimetric procedure is used for concentrations below 1 mg/mL
and is sensitive to about 0.02 mg/L.
1.4 The potentiometric procedure is used for concentrations between 26
and 0.02 mg/L. The lower limit can be extended through the use of
a calibration curve to 0.002 mg/L. Higher concentrations can be
measured, but since these increase erosion of the membrane, measure-
ments above 26 mg/L cyanide should be done only occasionally.
2. Summary of method
2.1 The method is based on converting to hydrocyanic acid all but the
most refactory metal-cyanide complexes from a slightly acidified
sample in a 1-hour reflux distillation, and absorbing in a NaOH
solution the HCN that has been purged from the sample by an air
stream. The procedure avoids the decomposition of iron cyanides
that are not amenable to chlorination by omitting the catalyst
(MgCl2) and by the addition of zinc and lead acetates which form
insoluble double salts with the iron cyanides. The cyanide ion in
the absorption solution..is quantitated either volumetrically, color-
imetrically, or electrometrically.
2.2 The titrimetric measurement uses a standard solution of silver
nitrate to titrate cyanide in the presence of a silver-sensitive
indicator.
2.3 In the colorimetric measurement the cyanide is converted to cyanogen
chloride (CNC1) by reaction with chloramine-T at a pH less than 8
without hydrolyzing to the cyanate. After the reaction is complete,
color is formed on the addition of a pyridine-pyrazalone or a pyridine-
barbituric acid reagent. The absorbance of the resulting colored
solution is read at 620 nm when using pyridine-pyrazalone and 578
nm when using the pyridine-barbituric acid reagent. It is essential
that both the sample and the standard have comparable ionic strengths.
2.4 The potentiometric measurement uses a cyanide ion selective electrode
and double junction reference electrode to quantitate the cyanide ion.
It is essential that both the sample and the standard to have com-
parable ionic strengths.
B-2
-------
3. Definitions
3.1 .Simple cyanides in this method are defined as cyanide Ton and "easily
dissociated cpmplex_cyanide_ cgnvert^d_tg_J1ydrocy_anic acid by reaction
in a reflux system of an acetate buffer in the presence "of lead and
zinc ion.
4. Sample handling and preservation
4.1 The sample should be collected in plastic or glass bottles of 1
liter or larger size. All bottles must be thoroughly cleansed
and rinsed to remove soluble material.
4.2 Oxidizing agents such as chlorine decompose most of the cyanides.
Test a drop of the sample with potassium iodide-starch test paper
(Kl-starch paper); a blue color indicates the need for treatment.
Add ascorbic acid, a few crystals at a time, until a drop of sample
produces no color on the indicator paper. Add an additional 0.6 g
ascorbic acid for each liter of sample volume.
4.3 Sulfides slowly convert the cyanide in the sample to thiocyanate.
The reaction rate is greatly increased at high pH. Sulfide there-
fore interferes and should be removed as soon as the sample is
collected and before adjustment of the pH. When sulfides are
present in the sample, it may be assumed that oxidizing agents are
absent. Test for the presence of sulfide by placing a drop of the
sample on a strip of..lead acetate test paper that has been pre-
viously moistened with the acetic acid solution. Darkening of the
test paper indicates the presence of sulfide.
4.3.1 Sulfide is removed by treating the sample with small
increments of powdered lead carbonate (PbCO 3), cadmium
carbonate (CdCO 3), or with the dropwise addition of lead
nitrate [Pb(N03)2] solution.. When significant quantities of
sulfide must be removed, the addition of PbC03 or CdC03 is
preferred. Pb(N03)2 may unduly depress the pH and with
Pb(OAc)2 additions, the acetic acid that will distill over
may neutralize too much NaOH in the absorber. Black PbS
precipitates in samples containing sulfide. Repeat the
operation until no more lead sulfide forms, as indicated
by testing the supernatant liquid with Pb(AOc)2 test paper.
It is not necessary to filter the liquid, since sulfide
will not distill over into the scrubber during distillation.
This avoids the removal of insoluble cyanides.
4.4 Samples must be preserved with 2 mL of ION sodium hydroxide par
liter of sample (pH 12) at the time of sample collection. (Ad-
ditional base may be necessary to ensure that the sample pH >_ 12 ).
4.5 Samples should be analyzed as soon as possible. If storage is re-
quired, the samples should be stored in a refrigerator or in an
ice chest filled with water and ice to maintain the temperature at 4°C.
B-3
-------
4.6 Minimize exposure of the samples to ultraviolet radiation. Photo-
decomposition of the iron cyanides may significantly increase the
cyanide content of the sample. (Remove interferences in the hood
under incandescent light conditions.)
5. Interferences
5.1 Interferences are eliminated or reduced by using the distillation
procedure described in Sections 8.1 through 8.5.
5.2 Fatty acids, which distill and form soaps in the alkaline scrubber
solution, make quantitation by titrimetric or colorimetric means
difficult. (This is not a problem if the ion-selective electrode
is used.) The fatty acids should be removed by extraction before
distillation. (Caution: This operation should be performed in a
fume hood and the sample left there until it can be made basic
again after extraction.) Acidify the sample with acetic acid
(1 + 9) to pH 6-7. Extract with iso-octane, hexane, or chloroform
(preference in order named) with a solvent volume equal to 20%
of the sample volume. One extraction usually is adequate to
eliminate the interference. Avoid multiple extractions or a long
contact time at low pH to minimize the loss of HCN. When the
extraction is completed, immediately raise the pH to Si 2 with NaOH
solution.
5.3 Organic thiocyanates will distill over and form turbid solutions
that adversely affect quantitation by titrimetric or colori-
metric and electrochemical means. These compounds may be mini-
mized by the extraction procedure outlined for removal of fatty
acids.
5.4 Aldehydes react with cyanide to form nitriles, which are further
hydrolyzed to their corresponding acids and ammonia under the
distillation conditions. Some of the aldehydes may be removed
by the extraction procedure outlined in Section 5.2.
5.5 Other possible interferences include substances that might con-
tribute color or turbidity. In most cases, the distillation
procedure will remove these.
6. Apparatus
^
6.1 The reflux distillation apparatus is shown in Figure 1. The
boiling flask should be of 1-liter size with an inlet tube and
provision for a condenser.
6.2 Microburet, 5.0mL (for titratior.).
6.3 Spectrophotometer suitable for measurement at 578 nm or 620 nm with
a 1.0 cm cell or larger.
k
Figure 1 is given at the end of the Appendix.
B-4
-------
6.4 A cyanide ion selective electrode, a double junction reference
electrode, an expanded scale mV meter or specific ion meter, and a
magnetic stirrer with TFE fluorocarbon-coated stirring bar.
7. Reagents
7.1 Sodium hydroxide solution, 1.25 N: Dissolve 50 g NaOH in distilled
water and dilute to 1 liter with distilled water.
7.2 Cadmium rsrbonate: Powdered.
7.3 Ascorbic acid: Crystals.
7.4 Dilute sodium hydroxide solution, 0.25 N: Dilute 200 mL of sodium
hydroxide solution (7.1) to 1 liter with distilled water.
7.5 Zinc and lead acetate solution: Dissolve 50 g zinc acetate
Zn(C2H302)'2H20 and 86.4 g lead acetate Pb(C2H302)2'3H20 in 500 mL" ~
water and dilute to 1 liter.
7.6 Acetate buffer solution: Dissolve^ 410 g of sodium acetate tri-
hydrate (NaC2H302- 3H20) in 50Q._mL of water. Add glacial acetic
acid to pH 4.5, approximately 500mL .
7.7 Stock cyanide solution, 1 mg/nL : Dissolve 2.51 g KCN and 2 g KOH
in 1 liter of distilled waterV" Standardize with 0.0192 N AgN03.
Dilute to appropriate concentration so that imL = 1 mg CN~.
7.8 Standard cyanide solution, intermediate 50 mg/L: Dilute 50 mL of
stock solution (7.7) to 1 liter with distilled water.
7.9 Standard_ cyanide solution. 5 mg/L: Prepare fresh daily by diluting
100.0 mL of intermediate solution (7.8) to 1 liter with distilled
water and store in a glass-stoppered bottle.
7.10 Standard silver nitrate solution: Prepare by crushing approximately
5 g of AgNOs crystals and drying to constant weight at 40°C. Weigh
out 3.2647 g dried AgNOS, dissolve in distilled water, and dilute to
1 liter (l~mL = 1 mg CN~).
7.11 Rhodanine indicator: Dissolve 20 mg p—dimethylaminobenzalrhodanine
in lOOmL of acetone.
7.12 Chloramine-T solution: Dissolve 1 g of white, water-soluble chloramine-
T in 100 mL of distilled water and refrigerate until ready to use. Pre-
pare fresh weekly.
7.13 Color reagent - One of the following may be used:
7.13.1 Pyridine-barbituric acid reagent: Place 15 g of barbituric
acid in a 250-mL volumetric flask and add just enough distilled
water to wash the sides of the flask and wet the barbituric acid_.
B-5
-------
Add 75_mL_of pyridine and mix. Add 15 mL of HC1 (sp gr 1.19);
mix and cool to room temperature. Dilute to 250 ml with
distilled water and mix. „ This reagent is stable for approx-
imately six norths if stored in a cool, dark place.
7.13.2 Pyridine-pyrazolone solution:
7.13.2.1 3-methyl-l-phenyl-2i~T)yrazolin-5-one, reagent
saturated solution: Adi 0.25 g of this com-
pound to 50mL ; distilled watqr, head to 60°
C with stirring. Cool to room temperature.
7.13.2.2 3,3'-dimethyl-1,1'-diphenyl-(4,4'-bi-2-py-
razoline)-5,5'-dione (bispyrazqlone): Dissolve
0.01 g of bispyrazolone in 10 mL pyridine.
7.13.2.3 Pour solution (7.13.2.1) through non-acid-washed
filter paper and collect the filtrate. Through
the same filter pour solution (7.13.2-. 2) col-
lecting the filtrate in the same container as
the filtrate from (7.13.2.1). Mix until the
filtrates are homogeneous. The mixed reagent
reacts with cyanide to give a pink color.'
7.14 Methyl red indicator: Prepare 0.02 g dissolved in 60 mL water
and 40 mL_ acetic acid.
7.15 Boiling chips.
8. Procedure
8.1 Place 500 mL Of sampiej Or an aliquot diluted to 500 ir.L , in the •
1-liter boiling flask. Add 50 mL of sodium hydroxide (7.1) to the
absorbing tube and dilute if necessary with distilled water to
obtain an adequate depth of liquid in the absorber. Connect the
boiling flask, condenser, absorber, and trap in the vacuum train.
8.2 Adjust the vacuum so that approximately one bubble of air per
second enters the boiling flask through the air inlet tube.
CAUTION: The bubble rate will not remain constant after the
reagents have been added and while heat is being applied to the
flask. It will be necessary to readjust the air rate occasionally
to prevent the solution in the boiling flask from backing up into
the air inlet tube.
8.3 Add 2-3 drops of the methyl red indicator to the reaction flask.
8.4 Add 10 mL each of the acetate buffer (.7.6) and zinc-lead acetate
solutions (7.5) through the air inlet tube.
B-6
-------
8.5 Rinse the air inlet tube with a few ml of water and allow the air
flow to mix the contents of the flask. (If the solution is not pink,
add acetic acid dropwise through the air inlet tube until there is
a permanent color change).
8.6 After adding boiling chips to the flask and with the condenser
cooling water on, heat the solution to boiling, taking care to
prevent the solution from backing into the air inlet tube. (En-
sure that the solution remains pink during the distillation. The
dropwise addition of more acid or indicator may be necessary).
8.7 Reflux for 1 hour
8.8 Turn off the heat, but maintain the air flow for at least an
additional 15 minutes.
8.9 Quantitatively transfer the solution from the absorber into a
100-mL volumetric flask.
8.10 Analyze this solution, or an aliquot of this solution, for
cyanide using the titrimetric, colorimetric, or potentiometric
methods of analysis.
8.11 Titrimetric method of analysis.
8.11.1 If the sample contains more than 1 mg of CN~, transfer the
distillate, or a suitable aliquot diluted to 100 mL to a
500-mL Erlenmeyer flask. Add 10-12 drops of the rhodanine
indicator (7.11).
8.11.2 Titrate with standard silver nitrate (7.10) to the first
change in color from yellow to brownish-pink. Titrate a
distilled water blank using the same amount of sodium
hydroxide and indicator as in the sample.
8.11.3 The analyst should familiarize himself (herself) with the
end point of the titration and the amount of indicator
to be used before actually titrating the samples. A 5- or
10-mL microburet may be conveniently used to obtain a more
precise titration.
8.12 Colorimetric method of analysis
8.12.1 Withdraw 20 mL or less of the solution from the flask and
transfer to a 100- mL voJLumetric flask. If less than 20 mL
is taken, dilute to~~20 mL_with 0.25 N sodium hydroxide
solution (7.4). Add 6 mL.of sodium phosphate solution
(7.6) and mix.
8.12.1.1 Pyridine-barbituric acid method: Add 2 mL of
chloramine-T (7.12) and mix. After 1 to 2
minutes, add 5 mL of pyridine-barbituric acid
B-7
-------
solution (7.13.1) and mix. Dilute to mark with
distilled water and mix again. Allow 8 minutes
for color development, then read absorbance at
578 nm in a 1-cm cell within 15 minutes.
8.12.1.2 Pyridine-pyrazolone method: Add 0.5mL
chloramine-T (7_. 12) and mix. After 1 to 2
minutes, add 5 mL of pyridine-pyrazolone so-
lution (7.13.2) and mix. Dilute to mark with
distilled water and mix again. After 40 min-
utes read the absorbance at 620 nm in a 1-cm
cell. NOTE: More than 0.5 ml chloramine-T will
prevent the color from developing with pyri-
dine-pyrazolone .
8.12.2 Prepare a series of standards by pipeting suitable volumes
of standard solution into 100-niL _volumetric flasks. To each
standard add 50 roL of 1.25 N sodium hydroxide and dilute
to 100 mL with distilled water. Prepare as follows:
_mL of Standard Solution Concentration .CN~
rs'pg/mL" CN~1 (mg CN"/100mL )
1 0.005
2.0 0.010
5.0 0.025
10.0 0.050
15.0 0.075
20.0 0.100
8.12.2.1 Prepare a standard curve by plotting absorbance
of standard vs. cyanide concentration.
8.12.2.2 It is not necessary that all standards be dis-
tilled in the same manner as the samples. It
is recommended that at least two standards (a
high and low) be distilled and compared with
similar values on the curve to ensure that the
distillation technique is reliable. If dis-
tilled standards do not agree within + 10% of
the undistilled standards, the operator should
find the cause of the apparent error before
proceeding.
8.12.3 To check the efficiency of the sample distillation, add
an increment of cyanide from either the intermediate
standard (7.8) or the working standard (7.9) to ensure a
level of 20 yg/L or a significant increase in absorbance
value. Proceed with the analysis as in Section 8.1, using
the same flask and system from which the previous sample
was just distilled.
B-8
-------
.13 Potentiometric method of analysis.
8.13.1 Prepare a series of standards by jDipeting suitable volumes
of a standard solution into 100-JnL_volumetric flasks. To
.each standard add 50 mL "of 1.25 N NaOH and dilute to 100
_mL_ with distilled water. Prepare as follows:
Concentration CN
_mL of Standard Solution (mg CN~/10Q ml )
(0.5 Mg/_mL CN~)
2.0 0.0010
5.0 0.0025
10.0 0.005
(5
2.0 0.010
5.0 0.025
10.0 0.050
(50 ug/mL" CN-)
2.0 0.100
5.0 0.250
10.0 0.50
8.13.1.1 Transfer the standard solutions into 150-mL beakers
prerinsed with a small portion of the standard
being tested. Immerse the cyanide and double junc-
tion reference electrodes in the solution and mix
on a magentic stirrer. Maintain as closely as pos-
sible the same stirring rate and temperature for
all solutions.
8.13.1.2 After equilibrium is reached (at least 5 minutes
and not more than 10 minutes) record the millivolt
reading and plot the CN~ concentrations versus
millivolt reading on similogarithmic graph paper.
A straight line with a slope of 59 mV indicates
that the instrument is operating properly.
8.13.1.3 It is not imperative that all standards be distil-
led in the same manner as the samples. It is recom-
mended that at least two standards (a high and low)
be distilled and compared with similar values on the
curve to ensure that the distillation technique is
reliable. If the distilled standards do not agree
within rlO% of the undistilled standards, the
operator should find the cause of the apparent error
before proceeding.
B-9
-------
8.13.2 Place the absorption liquid into a 150-m\ beaker and proceed
with the analysis as in Section 8.11.1. Determine the CN~
concentration by observing the millivolt reading and refer-
ring to the calibration curve established in Section 8.11.1.
The method of known addition can be used for measuring
occasional samples, since the preparation of a calibration
curve is not required.
8.13.3 To check the efficiency of the sample distillation, add an
increment of cyanide from either the intermediate standard
(7.8) or the working standard (7.9) to ensure a level of
20 ug/L or a significant increase in values. Proceed with
the analysis as in Section 8.1, using the same flask and
system from which the previous sample was just distilled.
9. Calculation
9.1 Using the titrimetric procedure, calculate the concentration of
CN as follows:
(A-B) (1000) (100 mL )
CN (mg/L) = (mL of original sample) (mL of aliquot titrated)
where:
A = Volume of AgNO,, used for titration of sample
A = Volume of AgNO., used for titration of blank.
9.2 Using the colorimetric procedure, calculate the concentration of
CN~ as follows :
- f 0/i^ - (A) (1000) (100)
N (yg/L) - - - -
where:
A = pg CN~ read from standard curve
B = niL^ of original sample taken for distillation
C = mL of scrubber solution taken for colorimetric analysis.
9.3 Using the potentiometric procedure, calculate the concentration
of CN~ as follows:
or
where:
A = yg CN read from standard curve
B = mL of original sample taken for distillation
B-1Q
-------
COOLING WATER
SCREW CLAMP
INLET TUBE
TO LOW
VACUUM
SOURCE
DISTILLING FLASK
HEATING
MANTLE
o
SA-7B54-23R
FIGURE 1 CYANIDE DISTILLATION APPARATUS
3-11
-------
-------
Appendix C
EPA PROCEDURE FOR ANALYSIS OF CYANIDES AMENABLE TO CHLORINATION
C-l
-------
Scope and Application
1.1 This method is applicable to the determination of cyanides in
drinking, surface, and saline waters, and domestic and industrial
wastes.
1.2 The titration procedure using silver nitrate with p-dimethylamino-
benzal-rhodanine indicator is used for measuring concentrations of
cyanide exceeding 1 mg/L (0.1 mg/100 mL "of absorbing liquid).
1.3 The colorimetric procedure is used for concentrations below 1 mg/
mL and is sensitive to about 0.02 mg/L.
1.4 The potentiometric procedure is used for concentrations between
26 and 0.02 mg/L. The lower limit can be extended through the
use of a calibration curve to 0.002 mg/L. Higher concentrations
can be measured, but since these increase erosion of the membrane,
measurements above 26 mg/L cyanide should be done only occasionally.
Summary of method
2.1 A portion of the sample is chlorinated to decompose the cyanide
amenable to chlorination. Subsequently, total cyanide is determined
in both the chlorinated and original sample. The difference between
the total cyanide concentration found in the two parts is expressed
as "cyanides amenable to chlorination."
2.1.1 Cyanide as hydrocyanic acid (HCN) .is released from cyan-
ide compounds by means of a reflux distillation and ab-
sorbed in a sodium hydroxide solution. The cyanide in
the absorbing solution is determined volumetrically,
colorimetrically, or electrometrically.
2.2 The volumetric measurement uses a standard solution of silver
nitrate to titrate cyanide in the presence of a silver-sensitive
indicator.
2.3 In the colorimetric measurement the cyanide is converted to
cyanogen chloride (CNC1) by reaction with chloramine-T at a pH
less than 8. After the reaction is complete, color is formed
on the addition of a pyridine-pyrazalone or a pyridine-bar-
bituric acid reagent. The absorbance of the resulting colored
solution is read at 620 nm when using pyridine-pyrazalone and
578 nm when using the pyridine-barbituric acid reagent. It is
essential that both the sample and the standards have compara-
ble ionic strengths.
2.4 The potentiometric measurement uses a cyanide ion selective elec-
trode and double junction reference electrode to quantitate the
cyanide ion. It is essential that both the sample and the stan-
dard have comparable ionic strengths.
C-2
-------
3. Definitions
3.1 Cyanide is defined as cyanide ion and complex cyanide converted
to hydrocyanic acid by reaction in a reflux system of a mineral
acid in the presence of magnesium ion.
4. Sample handling and preservation
4.1 The sample should be collected in plastic or glass bottles of 1
liter or larger size. All bottles must be thoroughly cleansed
and rinsed to remove soluble material.
4.2 Oxidizing agents such as chlorine decompose most to the cyanides.
Test a drop of the sample with potassium iodide-starch test paper
(Kl-starch paper); a blue color indicates the need for treatment.
Add ascorbic acid, a few crystals at a time, until a drop of
sample produces no color on the indicator paper. Add an addition-
al 0.6 g ascorbic acid for each liter of sample volume. (Large
excesses of ascorbic acid may adversely affect the analysis by
production of a yellow color in the scrubber solution.)
4.3 Sulfides slowly convert the cyanide in the sample to thiocyanate.
The reaction rate is greatly increased at high pH. Sulfide there-
fore interferes and should be removed as soon as the sample is
collected and before adjustment of the pH. When sulfides are
present in the sample, it may be assumed that oxidizing agents
are absent. Test for the presence of sulfide by placing a drop
of the sample on a strip of lead acetate test paper that has
been previously moistened with the acetic acid solution. Dark-
ening of the test paper indicates the presence of sulfide.
4.3.1 Sulfide is removed by treating the sample with small in-
crements of powdered lead carbonate (PbCOs), cadmium car-
bonate (CdC03), or with the dropwise addition of lead
nitrate [Pb (1*103)2] solution. (When significant quantities
of sulfide must be removed, the addition of PbCOs, or
CdCOs is preferred. Pb(N03)2 may unduly depress the pH
and with Pb(OAc)2 additions, the acetic acid—which will
distill over—may neutralize too much NaOH in the absorber)
Black PbS precipitates in samples containing sulfide.
Repeat the operation until no more lead sulfide forms, as
indicated by testing the supernatant liquid with Pb(OAc)2
test paper. Immediately filter through dry paper into a
dry beaker and stabilize the sample by adjusting the pH.
(.Insoluble cyanides may be removed from the sample by this
procedure.)
4.4 Samples must be preserved with 2mL of ION sodium hydroxide per
liter of sample (pH 12) at the time of sample collection. (The
addition of more base may be required to ensure that the pH of
the sample is .>1>2.)
4.5 Samples should be analyzed as soon as possible. If storage is
required, the samples should be stored in a refrigerator or in
C-3
-------
an ice chest filled with water and ice to maintain the temperature
at 4°C.
4.6 Minimize the exposure of the samples to ultraviolet radiation.
Photodecomposition of the iron cyanides may significantly in-
crease the cyanide content of the sample. (Remove interferences
in the hood under incandescent light conditions, etc.)
5. Interferences
5.1 Interferences are eliminated or reduced by using the distillation
procedure described in Sections 8.1 through 8.5.
5.2 Fatty, acids, which distill and form soaps in the alkaline scrubber
solution, make quantitation by titrimetric or colorimetric means
difficult. (This is not a problem if the ion-selective electrode
is used) The fatty acids should be removed by extraction before
distillation. (Caution: This operation should be performed in
a fume hood and the sample left there until it can be made basic
again after extraction.) Acidify the sample with acetic acid
(1 + 9) to pH 6-7. Extract with iso-octane, hexane, or chloroform
(preference in order named) with a solvent volume equal to 20% of
the sample volume. One extraction usually is adequate to elimin-
ate the interference. Avoid multiple extractions or a long con-
tact time at low pH to minimize the loss of HCN. When the ex-
traction is completed, immediately raise the pH to ^12 with NaOH
solution.
5.3 Organic thiocyanates will distill over and form turbid solutions
that adversely affect quantitation by titrimetric or colorimetric
means. These compounds also adversely affect quantitation by
electrochemical means. These compounds may be minimized by the
extraction procedure outlined for removal of fatty acids.
5.4 Aldehydes react with cyanide to form nitriles, which are further
hydrolyzed to their corresponding acids and ammonia under the
distillation conditions. Some of the aldehydes may be removed
by the extraction procedure outlined in Section 5.2.
5.5 Thiocyanates are decomposed during distillation to sulfide, which
interferes with quantitation of cyanide. The procedure outlined
in Section 4.3 can be used to remove sulfide from the scrubber
solution.
6. Apparatus
j*
6.1 The reflux distillation apparatus is shown in Figure 1. The
boiling flask should be of 1-liter size with an inlet tube and
provision for a condenser.
*
Figure 1 is given at the end of the Appendix.
C-4
-------
6.2 Microburet, 5.0m_L.(for titration) .
6.3 Spectrophotometer suitable for measurement at 578 nm or 620 nm
with a 1.0 cm cell or larger.
6.4 A cyanide ion selective electrode, a double junction reference
electrode, an expanded scale mV meter or specific ion meter and
a magnetic stirrer with TFE fluorocarbon-coated stirring bar.
Reagents
7.1 Sodium hydroxide solution, 1.25 N: Dissolve 50 g NaOH in dis-
tilled water and dilute to 1 liter with distilled water.
7.2 Cadmium carbonate: Powdered.
7.3 Ascorbic acid: Crystals.
7.4 Dilute sodium hydroxide solution, 0.25 N: Dilute 200 mL~~of sodium
hydroxide solution (7.1) to 1 liter with distilled water".
7.5 Sulfuric acid: Concentrated.
7.6 Sodium dihydrogenphosphate, 1M: Dissolve 138 g NaH2POi+.H20 in 1
liter of distilled water. Refrigerate this solution.
7.7 Stock cyanide solution, 1 mg/ml: Dissolve 2.51 g KCN and 2 g KOH
in 1 liter distilled water. Standardize with 0.0192N AgNOs-
Dilute to appropriate concentration so that 1 m_L = 1 mg CN .
7.8 Standard cyanide solution, intermediate: 50 mg/L: Dilute 50.0
mL of stock solution (7.7) to 1 liter with distilled water.
7.9 Standard cyanide solution, 5 mg/L: Prepare fresh daily by di-
luting lOO.OrnL of intermediate solution (7.8) to 1 liter with
distilled water and store in a glass-stoppered bottle.
7.10 Standard silver nitrate solution, 0.0192 N: Prepare by crushing
approximately 5 g AgNC>3 crystals and drying to constant weight
at 40°C. Weigh out 3.2647 g dried AgN03, dissolve in distilled
water, and dilute to 1 liter (l3L_ =1 mg CN~) .
7.11 Rhodanine indicator: Dissolve 20 mg p-dimethylaminobenzalrho-
danine in 100 mL of acetone.
7.12 Chloramine-T solution: Dissolve 1 g of white, water-soluble chlo-
ramine-T in 100 mL of distilled water and refrigerator until ready
to use. Prepare fresh weekly.
C-5
-------
7.13 Color reagent - One of the following may be used:
7.13.1 Pyridine-barbituric acid reagent: Place 15 g of barbituric
acid in a 250-mL volumetric flask and add just enough dis-
tilled water to wash the sides of the flask and wet the
the barbituric acid. Add 75 ml of pyridine and mix. Add
15_mL of HC1 (sp gr 1.19); mix and cool to room temperature.
Dilute to 250 mL with distilled water and mix. This rea-
gent is stable for approximately six months if stored in
a cool dark place.
7.13.2 Pyridine-pyrazolone solution:
7.13.2.1 3-methyl-l-phenyl-2-pyrazolin-5-in-one, reagent
saturated solution: Add 0.25 g of this compound
to 50 mL_ distilled water, heat to 60°C with
stirring. Cool to room temperature.
7.13.2.2 3,3'-dimethyl-L,1'-diphenyl-(4,4'-bi-2-py-
razoline)-5,5'-dione (bispyrazolone): Dissolve
0.01 g of bispyrazolone in 10 mL pyridine.
7.13.2.3 Pour solution (7.13.2.1) through non-acid-
washed filter paper. Collect the filtrate.
Pour solution through the same filter (7.13.2.2),
collecting the filtrate in the same container
as the filtrate from (7.13.2.1). Mix until
the filtrates are homogeneous. The mixed re-
agent develops a pink color, but this does not
affect the color production with cyanide if
used within 24 hours of preparation.
7.14 Magnesium chloride solution: Weigh 510 g of MgCl2'6H20 into a
1-liter flask; dissolve and dilute to 1 liter with distilled
water.
7.15 Calcium hypochlorite solution: Dissolve 5 g calcium hypochlorite
[Ca(OCl)2] in 100 _m_L'distilled water.
7.16 Boiling chips.
8. Procedure
8.1 Two sample aliquots are required to determine cyanides amenable to_
chlorination. To one 500-mL aliquot or a volume diluted to 500 mL 5
add calcium hypochlorite (7.15) dropwise while agitating and main-
taining the pH between 11 and 12 with sodium hydroxide (7.1).
8.2 Test for residual chlorine with Kl-starch test paper and maintain
this excess for 1 hour, continuing agitation. A distinct blue
color on the test paper indicates a sufficient chlorine level.
If necessary, add additional hypochlorite solution.
0-6
-------
8.3 After 1 hour, add 0.5 g ascorbic acid (7.3) until Kl-starch test
paper shows no residual chlorine. Add an additional 0.5 g of asc-
orbic acid to ensure the presence of excess reducing agent, (large
excesses of ascorbic acid may adversely affect analysis)
8.4 Test for total cyanide in both the chlorinated and unchlorinated
aliquots by the procedure outlined in the following sections.
8.5 Place 500 mL of sample, or an aliquot diluted to 500 mL , in the
1-liter boiling flask. Add 50 mL of sodium hydroxide (7.1) to the
absorbing tube and dilute if necessary with distilled water to
obtain an adequate depth of liquid in the absorber. Connect the
boiling flask, condenser, absorber, and trap in the vacuum train.
8.6 Adjust the vacuum so that approximately one bubble of air per second
enters the boiling flask through the air inlet tube.
CAUTION: The bubble rate will not remain constant after the re-
agents have been added and while heat is being applied to the
flask. It will be necessary to readjust the flow rate occasionally
to prevent the solution in the boiling flask from backing up into
the air inlet tube.
8.7 Slowly add 25 ml concentrated sulfuric acid (7.5) through the air
inlet tube. Rinse the tube with distilled water and allow the
airflow to mix the flask contents for 3 minutes. Pour 20 mL of
magnesium chloride solution (7.14) into the air inlet tube and
wash down with a stream of water. Add few boiling chips through
the air inlet tube and wash down with distilled water.
8.8 Heat the solution to boiling, taking care to prevent the solution
from backing up into and overflowing from the air inlet tube. Re-
flux for one hour. Turn off the heat and continue the airflow for
at least 15 minutes. After cooling the boiling flask, disconnect
the absorber and close off the vacuum source.
8.9 Quantitatively transfer the solution from the absorber into a 100-
mL volumetric flask.
8.10 Analyze this solution, or an aliquot of this solution, for cyanide
using the titrimetric, colorimetric, or potentiometric method of
analysis.
8.11 Titrimetric method of analysis
8.11.1 If the sample contains more than 1 mg of CN~, transfer the
distillate, or a suitable aliquot diluted to 100 mL to a
500-mL Erlenmeyer flask. Add 10-12 drops of the rhodanine
indicator (7.11).
8.11.2 Titrate with standard silver nitrate (7.10) to the first
change in color from yellow to brownish-pink. Titrate a
C-7
-------
distilled water blank using the same amount of sodium hy-
droxide and indicator as in the sample.
8.11.3 The analyst should familiarize himself (herself) with the
end point of the titration and the amount of indicator to
be used before actually titrating the samples, A 5- or
10-mL microburet may be conveniently used to obtain a more
precise titration.
8.12 Colorimetric method of analysis
8.12.1 Withdraw 20 mL or less of the solution from the flask and
transfer to a 100-mL_ volumetric flask. If less than 20iriL
is taken, dilute to ^20 mL with 0.25 N sodium hydroxide
solution (7.4). Add 6 mL^ of sodium phosphate solution (7.6)
and mix.
8.12.1.1 Pyridine-barbituric acid method: Add 2 mL of
chloramine-T (7.12) and mix. After 1 to 2 minutes,
add 4 ml of pyridine-barbituric acid solution
(7.13.1) and mix. Dilute to mark with distilled
water and mix again. Allow 8 minutes for color
development, then read absorbance at 578 nm in a
1-cm cell within 15 minutes.
8.12.2
8.12.1.2 Pyridine-pyrazolone method: Add 0 .5 mL chloramine
T (7.12) and mix. After 1 to 2 minutes, add 5 mL
of pyridine-pyrazolone solution (7.13.2) and mix.
Dilute to mark with distilled water and mix again.
After 40 minutes read the absorbance at 620 nm
in a 1-cm cell. NOTE: More than 0.5_mL chlor-
amine-T will prevent the color from developing
with pyridine-pyrazolone.
Prepare a series of standards by pipeting suitable volumes
of standard solution into 100-mL volumetric flasks. To
each standard add 50 mL of 1.25 N sodium hydroxide and
dilute to lOOmL with distilled water. Prepare as follows:
mL of Standard Solution
(5 '
1
2.0
5.0
10.0
15.0
20.0
Concentration CN
(mg CN~/100 mL )
0.005
0.010
0.025
0.050
0.075
0.100
8.12.2.1 Prepare a standard curve by plotting absorbance.
of standard versus cyanide concentration.
C-8
-------
8.12.2.2 It is not necessary that all standards be distil-
led in the same manner as the samples. It is
recommended that at least two standards (a high
and low) be distilled and compared to similar
values on the curve to ensure that the distilla-
tion technique is reliable. If distilled stan-
dards do not agree within ±10% of the undistilled
standards, the operator should find the cause of
the apparent error before proceeding.
8.12.3 To check the efficiency of the sample distillation, add an
increment of cyanide from either the intermediate standard
(7.8) or the working standard (7.9) to ensure a level of
20 yg/L or a significant increase in absorbance value.
Proceed with the analysis as in Section 8.1, using the same
flask and system from which the previous sample was just
distilled.
8.13 Potentiometric method of analysis.
8.13.1 Prepare a series of standards by pipeting suitable volumes
'of a standard solution into 100-mL volumetric flasks. To
each standard add 50 "ml" of 1.25 N NaOH and dilute to 100 mL"
with distilled water. Prepare as follows:
Concentration CN
ml of Standard Solution (mg CN-/100 ml )
(0.5 ;;g/mL CN~)
2.0 • 0.0010
5.0 0.0025
10.0 0.005
(5 ug/mL CN~)
2.0 0.010
5.0 0.025
10.0 0.050
(50
2.0 0.100
5.0 0.250
10.0 0.50
8.13.1.1 Transfer the standard solutions into 150-rnL-
beakers prerinsed with a small portion of the
standard being tested. Immerse the cyanide and
double junction reference electrodes in the solu-
tion and mix well on a magnetic stirrer. Main-
tain as closely as possible the same stirring
rate and temperature for all solutions.
C-9
-------
8.13.1.2 After equilbrium is reached (at least 5 minutes
and not less than 10 minutes), record the milli-
volt reading and plot the CN~ concentrations
versus millivolt reading on semilogarithmic graph
paper. A straight line with a slope of 59 mV
indicates that the instrument is operating properly.
8.13.1.3 It is not imperative that all standards be distil-
led in the same manner as the samples. It is recom-
mended that at least two standards (a high and
low) be distilled and compared to similar values
on the curve to ensure that the distillation tech-
nique is reliable. If the distilled standards
do not agree within ±10% of the undistilled stan-
dards, the operator should find the cause of the
apparent error before proceeding.
8.13.2 Place the absorption liquid into a 150-mL^ beaker and pro-
ceed with the analysis as in Section 8.^.3.1_. Determine the
CN~~ concentration by observing the millivolt reading and
referring to the calibration curve established in Section
8.13.1. The method of known addition can be used for measur-
ing occasional samples since the preparation of a calibra-
tion curve is not required.
8.13.3 To check the efficiency of the sample distillation, add an
increment of cyanide from either the intermediate standard
(7.8) or the working standard (7.9) to ensure a level of
20 yg/L or a significant increase in values. Proceed with
the analysis as in Section 8.1 using the same flask and
system from which the previous sample was just distilled.
9. Calculations
9.1 Using the titrimetric procedure, calculate the concentration of CN~
as follows:
GIT (mg/L) = (A-B)(1000)(100mL.)
(ml of original sample)(raL of aliquot titrated)
where:
A = Volume of AgNC>3 used for titration of sample
B = Volume of AgNC>3 used for titration of blank.
9.2 Using the colorimetric procedure, calculate the concentration of
CN~ as follows:
™- t Om (A) (100) (100)
CN (Mg/L) = (B)(C)
where:
C-10
-------
A = yg_ CN read from standard curve
B = mL of original sample taken for distillation
C = mL of scrubber solution taken for colorimetric analysis.
9.3 Using the potentiometric procedure, calculate the concentration of
CN~ as follows :
ra- (ug/L) = (
where :
A = ug CN read from standard curve
B = mL of original sample taken for distillation
C-ll
-------
COOLING WATER
INLET TUBE
SCREW CLAMP
DISTILLING FLASK
HEATING
MANTLE
o
TO LOW
VACUUM
SOURCE
SA-7854-23R
FIGURE 1 CYANIDE DISTILLATION APPARATUS
C-12
-------
Appendix D
LIGAND-EXCHANGE METHOD OF ANALYSIS FOR TOTAL CYANIDE
D-l
-------
1. Scope and application
1.1 This method is applicable to the determination of cyanides in drink-
ing, surface, and saline waters, and domestic and industrial wastes.
1.2 The titration procedure using silver nitrate with p-dimethylamino-
benzal-rhodaine indicator is used for measuring concentrations of
cyanide exceeding 1 mg/L (0.1 mg/100 mL of absorbing liquid).
1.3 The colorimetric procedure is used for concentrations below 1 ing/mL
and is sensitive to about 0.02 mg/L.
1.4 The potentiometric procedure is used for concentrations between 26
and 0.02 mg/L. The lower limit can be extended through the use of
a calibration curve to 0.002 mg/L. Higher concentrations can be
measured but since these increase erosion of the membrane, measure-
ments above 26 mg/L cyanide should be made only occasionally.
2. Summary of method
2.1 The cyanide is released from cyanide complexes by means of a ligand
exchange reaction during a reflux distillation. The cyanide, as
hydrocyanic acid (HCN), is aerated from the sample and absorbed in
a sodium hydroxide scrubber solution. The cyanide in the absorbing
solution is then quantitated by volumetric titration, colorimetry,
or potentiometry.
2.2 The titrimetric measurement uses a standard solution of silver
nitrate to titrate cyanide in the presence of a silver-sensitive
indicator.
2.3 In the colorimetric procedure the cyanide is converted to cyanogen
chloride (CNC1) by reaction with chloramine-T at a pH less than 8.
After the reaction is complete, color is formed on the addition of
a pryidine-pyrazalone or a pyridine-barbituric acid reagent. The
absorbance of the resulting colored solution is read at 620 nm when
using pyridine-pyrazalone and 578 nm when using the pyridine-
barbituric acid reagent. It is essential that both the sample and
the standards be of comparable ionic strength.
2.4 The potentiometric measurement uses a cyanide ion selective elec-
trode and double junction reference electrode to quantitate the
cyanide ion. It is essential that both the sample and the standard
have comparable ionic strengths.
3. Definitions
3.1 Cyanide is defined as free cyanide ion and cyanide from metal
complexes that is converted to hydrocyanic acid by reaction in this
system.
D-2
-------
Sampling handling and preservation
4.1 The sample should be collected in plastic or glass bottles of 1
liter or larger size. All bottles must be thoroughly cleansed and
rinsed to remove soluble material.
4.2 Oxidizing agents such as chlorine decompose most to the cyanides.
Test a drop of the sample with potassium iodide-starch test paper
(Kl-starch paper); a blue color indicates the need for treatment.
Add ascorbic acid, a few crystals at a time, until a drop of sam-
ple produces no color on the indicator paper. Add an additional
0.6 g ascorbic acid for each liter of sample volume.
4.3 Sulfides slowly convert the cyanide in the sample to thiocyanate.
The reaction rate is greatly increased at high pH. Sulfide there-
fore interferes and should be removed as soon as the sample is col-
lected and before adjustment of the pH. When sulfides are present
in the sample, it may be assumed that oxidizing agents are absent.
Test for the presence of sulfide by placing a drop of the sample
on a strip of lead acetate test paper that has been previously
moistened with the acetic acid solution. Darkening of the test
paper indicates the presence of sulfide.
4.3.1 Sulfide is removed by treating the sample with small incre-
ments of powdered lead carbonate (PbC03), cadmium carbonate
(CdC03), or with the dropwise addition of lead nitrate
[Pb(N03)2] solution. (When significant quantities of sulfide
must be removed, the addition of PbCOs , or CdCOs is preferred.
Pb(N03)2 may unduly depress the pH and with Pb(OAc)2 addi-
tions, the acetic acid that will distill over may neutra-
lize too much NaOH in the absorber.) Black PbS precipitates
in samples containing sulfide. Repeat the operation until
no more lead sulfide forms, as indicated by testing the
supernatant liquid with Pb(OAc)2 test paper. It is not
necessary to filter this liquid since S=, as PbS, will not
interfere.
4.4 Samples must be preserved with 2 ml of ION sodium hydroxide per
liter of sample (pH 12) at the time of sample collection. (It may
be necessary to add additional base to ensure that the pH of the
sample is ^12 .)
4.5 Samples should be analyzed as soon as possible. If storage is
required, the samples should be stored in a refrigerator or in an
ice chest filled with water and ice to maintain the temperature at
4°C.
4.6 Minimize exposure of the samples to ultraviolet radiation. Photo-
decomposition of the iron cyanides may significantly increase the
cyanide content of the sample. (Remove interferences in the hood
under incandescent light, etc.)
D-3
-------
5. Interferences
5.1 Interferences are eliminated or reduced by using the distillation
procedure described in Sections 8.1 through 8.5.
5.2 Fatty acids, which distill and form soaps in the alkaline scrubber
solution, make quantitation by titrimetric or colorimetric means
difficult. (This problem is not encountered if the ion selective
electrode is used.) The fatty acids should be removed by extraction
before distillation. (Caution: This operation should be performed
in a fume hood and the sample left there until it can be made basic
again after extraction.) Acidify the sample with acetic acid
(1 + 9) to pH 6-7. Extract with iso-octane, hexane, or chloro-
form (preference in order named) with a solvent volume equal to 20%
of the sample volume. One extraction usually is adequate to reduce
the fatty acid concentration below the interference. Avoid multiple
extractions or a long contact time at low pH to minimize the loss
of HCN. When the extraction is completed, immediately raise the
pH to ^12 with NaOH solution.
5.3 Organic thiocyanates will distill over and form turbid solutions
that adversely affect quantitation by titrimetric or colorimetric
finishes. These compounds also adversely affect quantitation by
electrochemical means. These compounds may be minimized by the
extraction procedure outlined for removal of fatty acids.
5.4 Aldehydes may react with the cyanide to produce nitriles, which
are further hydrolyzed to the corresponding acids and ammonia. Some
of the aldehydes may be removed by the extraction procedure out-
lined in Section 5.2.
5.5 Other possible interferences include substances that might contri-
bute color or turbidity. In most cases, the distillation procedure
will remove these.
6. Apparatus
*
6.1 The reflux distillation apparatus is shown in Figure 1. The boil-
ing flask should be of 1-liter size with an inlet tube and provision
for a condenser.
6.2 Microburet, 5.0mL (for titration) .
6.3 Spectrophotometer suitable for measurement at 578 nm or 620 nm with
a 1.0 cm cell or larger.
6.4 A cyanide ion selective electrode, a double junction reference elec-
trode, an expanded scale mV meter or specific ion meter, and a mag-
netic stirrer with TFE fluorocarbon-coated stirring bar.
Figure 1 is given at the end of the Appendix.
D-4
-------
7. Reagents
7.1 Sodium hydroxide solution, 1.25 N: Dissolve 50 g NaOH is distilled
water and dilute to 1 liter with distilled water.
7.2 Cadmium carbonate: Powdered.
7.3 Ascorbic acid: Crystals.
7.4 Dilute sodium hydroxide solution, 0.25 N: Dilute 200 mL of sodium
hydroxide solution (7.1) to 1 liter with distilled water.
7.5 Methyl red: Dissolve 0.2 g in 60 mL water and 50 mL glacial acetic
acid.
7.6 Sodium dihydrogenphosphate, 1 M: Dissolve 138 g Na^PO^-f^O in
1 liter of distilled water. Refrigerate this solution.
7.7 Stock cyanide solution, 1 mg/mL: Dissolve 2.51 g KCN and 2 g KOH
in 1 liter distilled water. Standardize with 0.0192N AgN03. Dilute
to appropriate concentration so that 1 mL = 1 mg CN~.
7.8 Standard cyanide solution, intermediate, 50 mg/L: Dilute 50.0 mL '•
of stock solution (7.7) to 1 liter with distilled water.
7.9 Standard cyanide solution, 5 mg/L; Prepare fresh daily by diluting
100. OmL of intermediate solution (7.8) to 1 liter with distilled
. water and store in a glass-stoppered bottle.
7.10 Standard silver nitrate solution, 0.0192 N: Prepared by crushing
approximately 5 g AgNOg crystals and drying to constant weight at
50°C. Weigh out 3.2647 g_ dried AgN03, dissolve in distilled water,
and dilute to 1 liter (1 mL =1 mg CN~) .
7.11 Rhodanine indicator: Dissolve 20 mg p-dimethylaminobenzalrho-
danine in 100 mL of acetone.
7.12 Chloramine-T_ solution: Dissolve 1 g of white, water-soluble chlora-
mine in 100 mL of distilled water and refrigerate until ready to
use. Prepare fresh weekly.
7.13 Color reagent - One of the following may be used:
7.13.1 Pyridine-barbituric acid reagent: Place 15 g of barbituric
acid in a 250-niL volumetric flask and add just enough dis-
tilled water to wash the sides of the flask and wet the
barbituric acid. Add 75 mL of pyridine and mix. Add 15 mL
of HC1 (sp gr 1^19); mix and cool to room temperature.
Dilute to 250 mL_with distilled water and mix. This reagent
is stable for approximately six months if stored in a cool
dark place.
D-5
-------
7.13.2 Pyridine-pyrazolone solution:
7.13.2.1 3-methyl-l-phenyl-2-pyrazolin-5-one reagent satu-
rated solution: Add 0.25 g of this compound to
50 mL_ distilled water, heat to 60°C with stirring.
Cool to room temperature.
7.13.2.2 3,3'-dimethyl-1,1'-diphenyl-(4,4'-bi-2-pyrazoline)-
5,5'-dione (bispyrazolone): Dissolve 0.01 g of
bispyrazolone in 10 mL pyridine.
7.13.2.3 Pour solution (7.13.2.1) through non-acid-washed
filter paper. Collect the filtrate. Through
the same filter pour solution (7.13.2.2), collec-
ting the filtrate in the same container as the
filtrate from (7.13.2.1). Mix until the filtrates
are homogeneous. The mixed reagent develops a
pink color but this does not affect the color
production with cyanide if used within 24 hours
of preparation.
7.14 Acetate buffer solution: Dissolve 410 g of sodium acetate trihy-
drate (NaC2H302'3H20) in 500_mL of water. Add glacial acetic acid
to pH 4.5 (approximately 500mL ).
7.15 TIRON solution: Dissolve 200 g of l,2-dihydroxy-3,5-benzenedi-
sulfonic acid, disodium salt monohydrate in 1 liter of water.
7.16 TEP solution: Adjust the pH of 250 g of tetraethylenepentamine to
5 with a HC1 solution (3+1). Dilute to 1 liter with water.
7.17 Lead acetate solution: Dissolve 90 g lead acetate trihydrate
(Pb[C2H302]2'3H20) in 1 liter of water.
7.18 Boiling chips.
8. Procedure
8.1 Place 500 ml of sample, or an aliquot diluted to 500_mLin the
1-liter boiling flask. Add 50 mL of sodium hydroxide (7.1) to the
absorbing tube and dilute if necessary with distilled water to ob-
tain an adequate depth of liquid in the absorber. Connect the
boiling flask, condenser, absorber, and trap in the vacuum train.
8.2 Adjust the vacuum so that approximately one bubble of air per second
enters the boiling flask through the air inlet tube. CAUTION: The
bubbler rate will not remain constant after the reagents have been
added and while heat is being applied to the flask. It will be
necessary to readjust the air rate occasionally to prevent the
solution in the boiling flask from backing up into the air inlet
tube.
D-6
-------
8.3 Add through the_air inlet tube the following reagents in jthe order
presented: 5 _mL of the lead acetate solution (7.17); 10 mL of _the
TIRON solution (7.15); 5 mL of the TEP solution (7.16), and 10 ~mL
of the acetate buffer solution (7.14). After each addition, rinse
the tube with water and allow the air flow to mix the flask con-
tents. Add a few boiling chips through the air inlet tube and wash
down with water.
8.4 Heat the solution to boiling, taking care to prevent the solution
from backing up into and overflowing from the air inlet tube. Re-
flux for one-half hour. Turn off the heat and continue the air
flow for at least 15 minutes. After cooling the boiling flask, dis-
connect the absorber and close off the vacuum source.
8.5 Quantitatively transfer the solution from the absorber into a 100-mL
volumetric flask.
8.6 This solution, or an aliquot of this solution, is then analyzed for
cyanide using the titrimetric, colorimetric, or potentiometric
methods of analysis.
8.7 Titrimetric method of analysis
8.7.1 If the sample contains more than 1 mg of CN~, transfer the
distillate, or a suitable aliquot diluted to lOOmL to a
500-mL E;rlenmeyer flask. Add 10-12 drops of. the rhodanine
indicator (7.11).
8.7.2 Titrate with standard silver nitrate (7.10) to the first
change in color from yellow to brownish-pink. Titrate a
distilled water blank using the same amount of sodium
hydroxide and indicator as in the sample.
8.7.3 The analyst should familiarize himself (herself) with the
end point of the titration and the amount of indicator to
be used before actually titrating the samples. A 5- or
10-mL microburet may be conveniently used to obtain a more
precise titration.
8.8 Colorimetric method of analysis
8.8.1 Withdraw 20 ml or less of the solution from the flask and
transfer to a 100-mL_volumetric flask. If less than 20 mL
is taken, dilute to 20 mL with 0.25 N sodium hydroxide solu-
tion (7.4). Add 6 mL of sodium phosphate solution (7.6)
and mix.
8.8.1.1 Pyridine-barbituric acid method: Add 2 mL of
chloramine-T (7.12) and mix. After 1 to 2 minutes,
add 5 mL of pyridine-barbituric acid solution
(7.13.1) and mix. Dilute to mark with distilled
water and mix again. Allow 8 minutes for color
D-7
-------
development, and read absorbance at 578 nm in a
1-cm cell within 15 minutes.
8.8.1.2 Pyridine-pyrazolone method: Add 0.5..mL chlora-
mine-T (7.12) and mix. After 1 to 2 minutes, add
5mL_of pyridine-pyrazolone solution (7.13.2) and
mix. Dilute to mark with distilled water and
mix again. After 40 minutes, read the absorbance
at 620 nm in a 1-cm cell. NOTE: More than 0.5 ml
chloramine-T will prevent the color from develop-
ing with pyridine-pyrazolone.
8.8.2 Prepare a series of standards by pipeting suitable volumes
of standard solution into 100-mL volumetric flasks. To
each standard add 50 mL _of 1.25 N sodium hydroxide and di-
lute to lOOJiL with distilled water. Prepare as follows:
mL of Standard Solution Concentration CN~
(5 pg/mL CN~) (mg CN'/lOO ml.)
1 0.005
2.0 0.010
5.0 0.025
10.0 0.050
15.0 0.075
20.0 0.100
8.8.2.1 Prepare a standard curve by plotting absorbance
of standard versus cyanide concentration.
8.8.2.2 It is not necessary that all standards be distil-
led in the same manner as the samples. It is
recommended that at least two standards (a high
and low) be distilled and compared to similar
values on the curve to ensure that the distilla-
tion technique is reliable. If distilled stan-
dards do not agree within ±10% of the undistilled
standards, the operator should find the cause of
the apparent error before proceeding.
8.8.3 To check the efficiency of the sample distillation, add an
increment of cyanide from either the intermediate standard
(7.8) or the working standard (7.9) to ensure a level of
20 ug/L or a significant increase in absorbance value. Pro-
ceed with the analysis as in Section 8.1, using the same
flask and system from which the previous sample was just
distilled.
8.9 Potentiometric method of analysis
8.9.1 Prepare a series of standards by_pipeting suitable volumes
of a standard solution into 100- ml volumetric flasks. To
D-8
-------
each standard add 50 ml of 1.25 N NaOH and dilute to 100 mL
with distilled water. Prepare as follows:
Concentration CN
..mL of Standard Solution (mg CN~/100 mL)
(0.5 yg/'mL CN~)
2.0 0.0010
5.0 0.0025
10.0 0.005
(5 yg/ml CN")
2.0 0.010
5.0 0.025
10.0 0.050
(50 yg/mL~CN~)
2.0 0.100
5.0 0.250
10.0 0.50
8.9.1.1 Transfer the standard solutions into 150- mL
beakers prerinsed with a small portion of the
standard being tested. Immerse the cyanide and
double junction reference electrodes in the solu-
tion and mix well on a magnetic stirrer. Main-
tain as closely as possible the same stirring
rate and temperature for all solutions.
8.9.1.2 After equilibrium is reached (at least 5 minutes
and not more than 10 minutes), record the milli-
volt reading and plot the CN~ concentrations
versus millivolt reading on semilogarithmic graph
paper. A straight line with a slope of 59 mV
indicates that the instrument is operating properly.
8.9.1.3 It is not necessary that all standards be distil-
led in the same manner as the samples. It is
recommended that at least two standards (a high
and low) be distilled and compared with similar
values on the curve to ensure that the distillation
technique is reliable. If the distilled standards
do not agree within ±10% of the undistilled stan-
dards, the operator should find the cause of the
apparent error before proceeding.
8.9.2 Place the absorption liquid into a 150-mL beaker and proceed
with the analysis as in Section 8.9.1. Determine the CN~
concentration by observing the millivolt reading and refer-
ring to the calibration curve established in Section 8.9.1.
D-9
-------
The method of known addition can be used for measuring
occasional samples since the preparation of a calibration
curve is not required.
8.9.3 To check the efficiency of the sample distillation, add an
increment of cyanide from either the intermediate standard
(7.8) or the working standard (7.9) to ensure a level of
20 yg/L or a significant increase in values. Proceed with
the analysis as in Section 8.1, using the same flask and
system from which the previous sample was just distilled.
9. Calculations
9.1 Using the titrimetric procedure, calculate the concentration of CN"
as follows :
CN- (mg/L) = _ (A-B) (1000) (100) _
(mL of original sample) (mL . of aliquot titrated)
where:
A = Volume of AgN03 for titration of sample
B = Volume of AgN03 for titration of blank.
9.2 Using the colorimetric procedure, calculate the concentration of
CN~ as follows:
CN- (yg/L) - (
where:
A = ug CN~ read from standard curve
B = mL of original sample taken for distillation
C = mL of scrubber solution taken for colorimetric analysis.
9.3 Using the potentiometric procedure, calculate the concentration of
CN~ as follows :
(yg/L)
where :
A = yg CN~ read from standard curve
B = mL of original sample taken for distillation
D-10
-------
COOLING WATER
SCREW CLAMP
INLET TUBE
TO LOW
VACUUM
SOURCE
DISTILLING FLASK
HEATING
MANTLE
FIGURE 1 CYANIDE DISTILLATION APPARATUS
SA-7854-23R
D-ll
-------
-------
Appendix E
EPA PROCEDURE FOR ANALYSIS OF TOTAL CYANIDE
E-l
-------
1. Scope and application
1.1 This method is applicable to the determination of cyanides in drink-
ing, surface, and saline waters, and domestic and industrial wastes.
1.2 The titration procedure using silver nitrate with p-dimethylamino-
benzal-rhodanine indicator is used for measuring concentrations of
cyanide exceeding 1 mg/L (0.1 mg/100_mL:of absorbing liquid).
1.3 The colorimetric procedure is used for concentrations below 1 mg/flL
and is sensitive to about 0.02 mg/L.
1.4 The potentiometric procedure is used for concentrations between 26
and 0.02 mg/L. The lower limit can be extended through the use of
a calibration curve to 0.002 mg/L. Higher concentrations can be
measured but since these increase erosion of the membrane, measure-
ments above 26 mg/L cyanide should be done only occasionally.
2. Summary of method
2.1 The cyanide as hydrocyanic acid (HCN) is released from cyanide com-
plexes by means of a reflux distillation operation and absorbed in
a scrubber containing sodium hydroxide solution. The cyanide ion
in the absorbing solution is then quantitated by volumetric titra-
tion, colorimetry, or potentiometry.
2.2 The titrimetric measurement uses a standard solution of silver ni-
trate to titrate cyanide in the presence of a silver-sensitive
indicator.
2.3 In the colorimetric measurement the cyanide is converted to cyanogen
chloride (CNC1) by reaction with chloramine-T at a pH less than 8.
After the reaction is complete, color is formed on the addition of
a pyridine-pyrazalone or a pyridine-barbituric acid reagent. The
absorbance of the resulting colored solution is read at 620 nm when
using pyridine-pyrazalone and 578 nm when using the pyridine-barbi-
turic acid reagent. It is essential that both the sample and the
standards be of comparable ionic strength.
2.4 The potentiometric measurement uses a cyanide ion selective electrode
and double junction reference electrode to quantitate the cyanide
ion. It is essential that both the sample and the standard have
comparable ionic strengths.
3. Definitions
3.1 Cyanide is defined as cyanide ion and complex cyanide converted to
hydrocyanic acid by reaction -in a reflux system of a mineral acid
in the presence of magnesium ion.
E-2
-------
4. Sample handling and preservation
4.1 The sample should be collected in plastic or glass bottles of 1 liter
or larger size. All bottles must be thoroughly cleansed and rinsed
to remove soluble material.
4.2 Oxidizing agents such as chlorine decompose most to the cyanides.
Test a drop of the sample with potassium iodide-starch test paper
(Kl-starch paper)' a blue color indicates the need for treatment.
Add ascorbic acid, a few crystals at a time, until a drop of sample
produces no color on the indicator paper. Add an additional 0.6 g
ascorbic acid for each liter of sample volume. (Large excesses of
ascorbic acid may produce a yellow-colored scrubber solution that
interferes with the colorimetric finish.)
4.3 Sulfides slowly convert the cyanide in the sample to thiocyanate.
The reaction rate is greatly increased at high pH. Sulfide there-
fore interferes and should be removed as soon as the sample is col-
lected and before adjustment of pH. When sulfides are present in
the sample, it may be assumed that oxidizing agents are absent.
Test for the presence of sulfide by placing a drop of the sample on
a strip of lead acetate test paper that has been previously moistened
with the acetic acid solution. Darkening of the test paper indicates
the presence of sulfide.
4.3.1 Sulfide is removed by treating the sample with small incre-
ments of powdered lead carbonate (PbCOs) , cadmium carbonate
(CdC03), or with the dropwise addition of lead nitrate
[Pb(N03)2l solution. (When significant quantities of sulfide
must be removed, the addition of PbCOs, or CdCO 3 is prefer-
red. Pb(NOs)2 may unduly depress the pH and with Pb(OAc)2
additions, the acetic acid that will distill over may neu-
tralize too much NaOH in the absorber.) Black PbS precipi-
tates in samples containing sulfide. Repeat the operation
until no more lead sulfide forms, as indicated by testing
the supernatant liquid with Pb(OAc)2 test paper. Immediately
filter through dry paper into a dry beaker and stabilize the
sample by adjusting the pH. (This may have the adverse
effect of removing insoluble cyanides, thereby resulting
in abnormally low results.)
4.4 Samples must be preserved with 2 mL of ION sodium hydroxide per
liter of sample (pH 12) at the time of sample collection.
(Additional base may be needed to ensure that the pH of the sample
is
4.5 Samples should be analyzed as soon as possible. If storage is re-
quired, the samples should be stored in a refrigerator or in an ice
chest filled with water and ice to maintain the temperature at 4°C.
4.6 Minimize exposure of the samples to ultraviolet radiation. Photo-
decomposition of the iron cyanides may significantly increase the
E-3
-------
cyanide content of the sample. (Remove interferences in the hood
under incandescent light conditions, etc.)
5. Interferences
5.1 Interferences are eliminated or reduced by using the distillation
procedure described in Sections 8..1 through 8.5.
5.2 Fatty acids that distill and form soaps in the alkaline scrubber
solution make quantitation by titrimetric or colorimetric means
difficult. (This is not a problem if the ion-selective electrode
is used.) The fatty acids should be removed by extraction before
distillation. (Caution: This operation should be performed in a
fume hood and the sample left there until it can be made basic
again after extraction.) Acidify the sample with acetic acid
(1+9) to pH 6-7. Extract with iso-octane, hexane, or chloroform
(preference in order named) with a solvent volume equal to 20% of
the sample volume. One extraction usually is adequate to reduce
the fatty acid concentration below the interference. Avoid multiple
extractions or a longer contact time at low pH to minimize the loss
of HCN. When the extraction is completed, immediately raise the
pH to 2:12 with NaOH solution.
5.3 Organic thiocyanates will distill over and form turbid solutions
that adversely affect quantitation by titrimetric or colorimetric
finishes. These compounds also adversely affect quantitation by
electrochemical means. These compounds may be minimized by the
extraction procedure outlined for removal of fatty acids.
5.4 Aldehydes react with cyanide to produce nitriles, which are further
hydrolyzed to their corresponding acids and ammonia. Some of the
aldehydes may be removed by the extraction procedure outlined in
Section 5.2.
5.5 Thiocyanates are decomposed during the distillation to sulfide,
which interferes with quantitation. The procedure outlined in
Section 4.3 can be used to remove sulfide from the scrubber solution.
5.6 Other possible interferences include substances that might contri-
bute color or turbidity. In most cases, the distillation procedure
will remove these.
6. Apparatus
*
6.1 The reflux distillation apparatus is shown in Figure 1. The boil-
ing flask should be of 1 liter size with an inlet tube and provision
for a condenser.
6.2 Microburet, 5.0 ml (for titration).
*
Figure 1 is given at the end of the Appendix.
E-4
-------
6.3 Spectrophotometer suitable for measurement at 578 nm or 620 nm with
a 1.0 cm cell or larger.
6.4 A cyanide ion selective electrode, a double junction reference elec-
trode, an expanded scale mV meter or specific ion meter, and a mag-
netic stirrer with TFE fluorocarbon-coated stirring bar.
Reagents
7.1 Sodium hydroxide solution, 1.25 N: Dissolve 50 g NaOH in distilled
water and dilute to 1 liter with distilled water.
7.2 Cadmium carbonate: Powdered.
7.3 Ascorbic acid: Crystals.
7.4 Dilute sodium hydroxide solution, 0.25 N: Dilute 200_mL_of sodium
hydroxide solution (7.1) to 1 liter with distilled water.
7.5 " Sulfuric acid: Concentrated.
7.6 Sodium dihydrogenphosphate, 1 M: Dissolve 138 g NaF^PO^-t^O in
1 liter of distilled water. Refrigerate this solution.
7.7 Stock cyanide solution, 1 mg/mL Dissolve 2.51 g KCN and 2 g KOH
in 1 liter distilled water. Standardize with 0.0192 N AgN03. Dilute
to appropriate concentration so that lm_L = 1 mg CN~.
7.8 Standard cyanide solution, intermediate, 50 mg/L: Dilute SO.OmL
of stock solution (7.7) to 1 liter with distilled water.
7.9 Standard cyanide solution, 5 mg/L: Prepare fresh daily by diluting
lOO.O.mL'of intermediate solution (7.8) to 1 liter with distilled
water in a glass-stoppered bottle.
7.10 Standard silver nitrate solution, 0.0192 N: Prepared by crushing
approximately 5 g AgNOs crystals and drying to constant weight at
50°C. Weigh out 3.2647 g dried AgN03, dissolve in distilled water,
and dilute to 1 liter (lmL_ = 1 mg CN~) .
7.11 Pxhodanine indicator: Dissolve 20 mg p-dimethylaminobenzalrhod-
anine in 100 mL of acetone.
7.12 Chloramine-T solution: Dissolve 1 g of white, water-soluble chlor-
amine-T in 100 mL of distilled water and refrigerate until ready
to use. Prepare fresh daily.
7.13 Color reagent - One of the following may be used:
7.13.1 Pyridine-barbituric acid reagent: Place 15 g of barbituric
acid in a 250-mL volumetric flask and add just enough dis-
tilled water to wash the sides of the flask and wet the
barbituric acid. Add 75 mL_of pyridine and mix. Add 15 mL
E-5
-------
of HC1 (sp gr 1.19), mix and cool to room temperature. Di-
lute to 250 mL with distilled water and mix. This reagent
is stable for approximately six months if stored in a cool
dark place.
7.13.2 Pyridine-pyrazolone solution
7.13.2.1 3-methyl-l-phenyl-2-pyrazolin-5-one reagent satu-
rated solution: Add 0.25 g of this compound to
50mL_ distilled water, heat to 60°C with stirring.
Cool to room temperature.
7.13.2.2 3,3'-dimethyl-l,ll-diphenyl-(4,4'-bi-2-pyrazoline)-
5,5'-dione (bispyrazplone): Dissolve 0.01 g of
bispyrazolone in 10 mL pyridine.
7.13.2.3 Pour solution (7.13.2.1) through non-acid-washed
filter paper. Collect the filtrate. Through the
same filter pour solution (7.13.2.2), collecting
the filtrate in the same container as the filtrate
from (7.13.2.1). Mix until the filtrates are homo-
geneous . The mixed reagent develops a pink color
but does not affect the color production with
cyanide if used within 25 hours of preparation.
7.14 Magnesium chloride solution: Weigh 510 g of MgCl2'6H20 into a
1-liter flask, dissolve, and dilute to 1 liter with distilled water.
7.15 Boiling chips.
Procedure
8.1 Place 500 mL_ of sample, or an aliquot diluted to 500 mL in the
1-liter boiling flask. Add SOmL, °f sodium hydroxide (7.1) to the
absorbing tube and dilute if necessary with distilled water to ob-
tain an adequate depth of liquid in the absorber. Connect the boil-
ing flask, condenser, absorber, and trap in the vacuum train.
8.2 Adjust the vacuum so that approximately one bubble of air per second
enters the boiling flask through the air inlet tube. CAUTION: The
bubble rate will not remain constant after the reagents have been
added and while heat is being applied to the flask. It will be
necessary to readjust the air rate occasionally to prevent the solu-
tion in the boiling flask from backing up into the air inlet tube.
8.3 Slowly add 25 mL concentrated sulfuric acid (7.5) through the air
inlet tube. Rinse the tube with distilled water and allow the air
flow to mix the flask' contents for 3 minutes. Pour 20 roL_ of mag-
nesium chloride solution (7.14) into the air inlet tube and wash
down with a stream of water. Add a few boiling chips through the
air inlet tube and wash down with distilled water.
E-6
-------
8.4 Heat the solution to boiling, taking care to prevent the solution
from backing up into and overflowing from the air inlet tube. Re-
flux for one-half hour. Turn off the heat and continue the air
flow for at least 15 minutes. After cooling the boiling flask,
disconnect the absorber and close off the vacuum source.
8.5 Quantitatively transfer the solution from the absorber into a
100-mL volumetric flask.
8.6 This solution, or an aliquot of this solution, is then analyzed for
cyanide using the titrimetric, colorimetric, or potentiometric
methods of analysis.
8.7 Titrimetric method of analysis
8.7.1 If the sample contains more than 1 mg of CX , transfer the
distillate, or a suitable aliquot diluted to lOOmL to a
500-mL" Erlenmeyer flask. Add 10-12 drops of the rhodanine
indicator (7.11).
8.7.2 Titrate with standard silver nitrate (7.10) to the first
change in color from yellow to brownish-pink. Titrate a
distilled water blank using the same amount of sodium hy-
droxide and indicator as in the sample.
8.7.3 The analyst should familiarize himself (herself) with the
end point of the titration and the amount of indicator to
be used before actually titrating the samples. A 5- or 10-mL
microburet may be conveniently used to obtain a more precise
titration.
8.8 Colorimetric method of analysis
8.8.1 Withdraw 20 ml or less of the solution from the flask and
transfer to a 100-mL volumetric flask. If less than 20mL
is taken, dilute to 20 mL with 0.25 N sodium hydroxide
solution (7.4). Add 6 mL of sodium phsophate solution (7.6)
and mix.
8.8.1.1 Pyridine-barbituric acid method: Add 2mL of
chloramine-T (7.12) and mix. After 1 to 2 minutes,
add 5 mL of pyridine-barbituric acid solution
(7.13.1) and mix. Dilute to mark with distilled
water and mix again. Allow 8 minutes for color
development, then read absorbance at 578 nm in a
1-cm cell within 15 minutes.
8.8.1.2 Pyridine-pyrazolone method: Add 0.5 mL chloramine-
T (7.12) and mix. After 1 to 2 minutes, add 5 ml
of pyridine-pyrazolone solution (7.13.2) and mix.
Dilute to mark with distilled water and mix again.
After 40 minutes read the absorbance at 620 nm
E-7
-------
in a 1-cm cell. NOTE: More than 0.5mL chlor-
amine-T will prevent the color from developing
with pyridine-pyrazalone.
8.8.2 Prepare a series of standards by pipeting suitable volumes
of standard solution into 100-mL volumetric flasks. To
each standard add 50 mL of 1.25 N sodium hydroxide and
dilute to 100 ml with distilled water. Prepare as follows;
mL_ of Standard Solution Concentration CN~
(5 yg/mL CN~) (mg CN-/100 mL.)
1 0.005
2.0 0.010
5.0 0.025
10.0 0.050
15.0 0.075
20.0 • 0.100
8.8.2.1 Prepare a standard curve by plotting absorbance
of standard versus cyanide concentration.
8.8.2.2 It is not imperative that all standards be dis-
tilled in the same manner as the samples. It is
recommended that at least two standards (a high
and low) be distilled and compared to similar
values on the curve to ensure that the distilla-
tion technique is reliable. If distilled stan-
dards do not agree within ±10% of the undistilled
standards, the operator should find the cause of
the apparent error before proceeding.
8.8.3 To check the efficiency of the sample distillation, add an
increment of cyanide from either the intermediate standard
(7.8) or the working standard (7.9) to ensure a level of
20 yg/L or a significant increase in absorbance value. Pro-
ceed with the analysis as in Section 8.1, using the same
flask and system from which the previous sample was just
distilled.
8.9 Potentiometric method of analysis
8.9.1 Prepare a series of standards by pipeting suitable volumes
of a standard solution into 100-mL volumetric flasks. To
each standard add 50mL of 1.25 N~NaOH and dilute to 100 mL
with distilled water. Prepare as follows:
E-8
-------
Concentration CN"
mL of Standard Solution (mg CN~/10QmL )
(0.5 ug/faL "CN-)
2.0 0.001
5.0 0.0025
10.0 0.005
(5 pg/lnL 'CN-)
2.0 0.010
5.0 0.025
10.0 0.075
(50 pg/mL'-CN")
2.0 0.100
5.0 0.25
10.0 0.50
8.9.1.1 Transfer the standard solutions into 150-inL
beakers prerinsed with a small portion of the
standard being tested. Immerse the cyanide and
double junction reference electrodes in the solu-
tion and mix well on a magnetic stirrer. Main-
tain as closely as possible the same stirring
rate and temperature for all solutions.
8.9.1.2 After equilibrium is reached (at least 5 minutes
and not more than 10 minutes), record the milli-
volt reading and plot the GN~ concentrations ver-
sus millivolt reading on semilogarithmic paper.
A straight line with a slope of 59 mV indicates
that the instrument is operating properly.
8.9.1.3 It is not necessary that all standards be distilled
in the same manner as the samples. It is recom-
mended that at least two standards (a high and
low) be distilled and compared to similar values
on the curve to ensure that the distillation
technique is reliable. If the distilled standards
do not agree within ±10% of the undistilled stan-
dards , the operator should find the cause of the
apparent error before proceeding.
8.9.2 Place the absorption liquid into a 150-mL beaker and pro-
ceed with the analysis as in Section 8.9.1. Determine the
CN concentration'by observing the millivolt reading and
referring to the calibration curve established in Section
8.9.1. The method of known addition can be used for measur-
ing occasional samples as the preparation of a caliabration
curve is not required.
E-9
-------
COOLING WATER
In
Out
-Its
INLET TUBE
SCREW CLAMP
TO LOW
VACUUM
SOURCE
DISTILLING FLASK
HEATING
MANTLE
o
SA-7854-23B
FIGURE 1 CYANIDE DISTILLATION APPARATUS
E-ll
-------
Appendix F
STATISTICAL PROTOCOL
F-l
-------
1. Introduction
1.1 This appendix outlines the steps of the statistical protocol used
in this study. The purpose of the statistical analysis is to
estimate the total precision error of an analytical method.
2. Definitions and Symbols
2.1 Mean—arithmetic average (X), defined as the sum of the observations
divided by the number of observations (n).
n
' X.
X =
2.2 Variance—a2, defined as the sum; of the squares of the deviations
from their mean (X) divided by one less than the number of obser-
vations (n - 1).
o
li-x
2 - i~l
'n - 1
2.3 Standard deviation—a, defined as the positive square root of the
variance (a2).
a = /a2
2.4 Relative standard deviation (or coefficient of variation)—defined
as the standard deviation divided by the mean.
Relative standard deviation = -^-
X
2.5 Grubb's test for rejection of an observation is applied to determine
if one of the observations should be rejected as being an outlier.
The following equation was used for the test:
x-x
where:
X = observation being tested (most distant from mean)
X = mean of n observations
a = standard deviation based on n - 1 degrees of freedom.
F-2
-------
For any six observations, a value can be rejected if Bj' >_ 1.944.
The B]/ limit is based on a 1% significance level (i.e., a BI" calcu-
lated from the data can be expected to exceed 1.944 only 1% of the
time if the observation is a legitimate one conforming to the under-
lying theory).
2.6 Lower limit of detection—defined as that level of cyanide in a sam-
ple, as KCN, where the percent relative standard deviation in a
series of replicate analyses is found to be above 10% but below 50%.
0.1 .<_ relative standard deviation >_ 0 . 5
In practice, the detection limit is governed not by the digestion/
distillation step but by the finish used (i.e., ion-selective elec-
trode, colorimetry or titrimetry). As such, the lower limit of
detection reported here reflects the limitations of finishes used
and their detection limits. It should be possible to extend the
useful range of these procedures by suitable modification of the
finish used here or by using a more sensitive finish.
F-3
------- |