Problems Associated with the EPA/ASTM Approved Method for Determination of Total Cyanide

By E.B. Milosavljevic and L. Solujic

Introduction

The EPA/ASTM approved method for determination of total cyanide is based on liberating HCN from a sample acidified with the sulfuric acid using an hour long reflux distillation procedure. The HCN gas formed is trapped by passing it through the high pH alkaline scrubbing solution. Cyanide concentration in this solution is then quantified by the spectrophotometric or potentiometric procedure.

Most of the problems associated with the total cyanide determination stem from the fact that the method utilizes extremely harsh analytical conditions (t>100 ° C and pH<0). In addition, the actual "color developing reaction" in the spectrophotometric quantitation procedure involves the addition of the very strong oxidant (chloramine-T). These conditions along with the complexity of the sample matrices, often produce complicated reaction pathways that can result in the production of cyanide (real or as an artifact) during distillation and associated colorimetric quantitation procedures.

Discussion

Overall Problems

The overall problems inherent in the standard total cyanide determination method may be illustrated by the remarkably low reproducibility and repeatability of the technique obtained when the "real world" samples are analyzed in the round-robin format. The results obtained by four commercial laboratories analyzing two samples are illustrated in Table I.1

Table I. An inter-laboratory comparison of cyanide values obtained with the EPA/ASTM approved total cyanide determination method.

Total CN- found (ppm)

Laboratory

Sample 1

Sample 2

A

0.120

8.80

B

0.091

7.18

C

0.009

15.8

D

<0.002

11.5

As may be seen from this table, over an order of magnitude difference in the cyanide levels were found for a given sample.

Similar observations were recently reported by Dr. Matz2 of Quantum Chemical Company. The company has problems in determination of cyanide in their syngas plant: "To determine if there was a problem with the method we sent our samples over a three month period to two outside labs that conduct cyanide analysis using the EPA approved protocol, and we analyzed them using the colorimetric method. During the three-month period, over 50 percent of our samples disagreed by more than a factor of 10. So, we determined that there was definitely a problem with the method. You can see from the data that the interlab and intralab consistency is pretty bad. When your standard deviation is bigger than your mean, you know that you are in trouble".

In addition, if one accepts the definition that the TOTAL cyanide method measures all of the available CN- in a given sample, it is obvious form the data summarized in Table II that the designation "TOTAL" is a misnomer. In essence, the TOTAL method, besides the "WAD cyanides", achieves complete cyanide recovery only from Fe-cyano species.

Table II. Species dependent cyanide recoveries (%) obtained with the EPA/ASTM approved TOTAL cyanide determination method.

Species

0.200 ppm

CN-

2.0 ppm

CN-

[Zn(CN)4]2-

99.5

104.4

[Cd(CN)4]2-

103.8

102.9

[Cu(CN)4]3-

97.7

98.0

[Ag(CN)]-

97.8

100.2

[Ni(CN)4]2-

104.2

97.1

[Hg(CN)4]2-

95.8

97.6

Hg(CN)2

98.0

97.3

[Fe(CN)6]4-

100.8

99.6

[Fe(CN)6]3-

104.0

99.5

[Pd(CN)4]2-

69.1

66.5

[Pt(CN)4]2-

0.0

0.0

[Pt(CN)6]2-

0.0

0.0

[Ru(CN)6]4-

50.1

51.0

[Au(CN)2]-

56.6

58.4

[Co(CN)6]3-

0.0

2.5

Interference Problems

Causes of positive bias. The analyses for total cyanide are subject to various types of interferences which produce artificially high cyanide values. These interferences may be caused by the so called cyanogenic compounds. Theses are the compounds that do not contain free or coordinated CN- ion, but that can produce cyanide under the harsh condition that exist in the distillation flask during a protracted heating procedure required for the total cyanide determination. For example, NO2- can react with a variety of organic compounds to produce HCN during the distillation procedure3. Under certain conditions nitrate ion can be reduced to nitrite which can then react with the organic carbon sources to produce hydrogen cyanide.

Another source of positive bias in the total cyanide determination method is the SCN- ion which is often present in the cyanide containing samples. The thiocyanate is much less toxic than cyanide and is not regulated by the EPA, hence it should not be determined by the regulatory methods designed to measure cyanide. However, under certain conditions, at high acidities that exist in the distillation flask thiocyanate does react with nitrate to produce free cyanide4.

When oxidizing agents are present in the cyanide containing sample, standard methods require the addition of the ascorbic acid. Unfortunately, the investigators at Santa Fe Springs, CA have recently established that if the given sample also contains ammonia and/or brine, colored complexes are developed during the colorimetric cyanide finish resulting in the variable false cyanide measurements5.

Another cause of positive bias in the total cyanide determination method is the so called spectrophotometric turbidity after distillation. This particular positive interference is probably caused by the fact that certain products of the oxidation of organic compounds formed during distillation are being purged into the absorbing solution which undergoes spectrophotometric finish. These products then react with chloramine-T producing colored species which interfere in the cyanide quantitation.

The problems that potential interferences can cause during the total cyanide determination have been the focus of extensive investigative efforts of the Research Triangle Institute, Research Triangle Park, NC 27709 and U.S. Environmental Protection Agency, EMSL, Cincinnati, OH. In the first study6 by M.G. Goldberg, et al., it was stated that: "In brief, the problems reported included low recovery of cyanide spikes, suspected false positive results, and poor precision of replicate analysis." In the Q&A session after Ms. Goldberg’s presentation, one of the authors of this paper, Mr. Potter of the U.S. EPA, stated that their study has shown that the method (total cyanide method) is "disfunctionate" as written. In the same vein, Ms. Goldberg in answering one of the questions said: "It makes it very difficult to regulate on a method where you can get between zero and 1000 percent recovery." In responding to another question from the audience Ms. Goldberg said: "Well, I think, as Bill has said, the fact that we had 1000 percent recovery really is just showing that the method is disfunctional."

In the second study7 in the series, the authors have established that the interference effects are cyanide species dependent (the species investigated were KCN, K[Au(CN)2] and K4[Fe(CN)6]). They have developed a complicated predictive model which could potentially be applied to the analyses of the gold mining wastes. Unfortunately, for this model to work, the analyst should know, at least semiquantitatively, cyanide speciation and interferences present.

Conclusion

From the aforementioned it is obvious that total cyanide determination method often produces unreliable data. The method has low reproducibility and repeatability and serious interferences even from many ubiquitous species.

The problems associated with the approved methods for cyanide determination has prompted the EPA to look for solutions provided by different analytical methods. Thus, last year, the agency initiated the validation study of the EPA Method 1677 (Weak Acid Dissociable Cyanide by Ligand Exchange/Flow Injection/Amperometric Technique). The method tested is based on a recent study8 and has significant advantages over the EPA/ASTM approved procedures.

References

1 For more details please see: Gallagher, N.P., CYANIDE ANALYSES AT THE GOLDEN SUNLIGHT MINE, Proceedings of the Ninth Annual Conference of the Society of Mineral Analysts, Salt lake City, UT, April 1995, pp. 33-58; Milosavljevic, E.B.; Solujic, Lj.; Hendrix, J.L. PRESENT AND FUTURE OF "FREE CYANIDE" DETERMINATIONS, Proceedings of the 18th Annual EPA Conference on Analysis of Pollutants in the Environment, Norfolk, VA, May, 1995, p. 74; Solujic, Lj.; Milosavljevic, E.B.; Hendrix, J.L.; Straka, M.R.; Gallagher, N.P.; CYANIDE DETERMINATION METHODS: DISTILLATION vs. FLOW INJECTION ANALYSIS, Randol Gold Forum '96, Squaw Creek, CA, April, 1996, 167.

2 Matz, S.G.; THE DETERMINATION OF CYANIDE IN A CHEMICAL PLANT'S WASTE WATER USING ION CHROMATOGRAPHY, Proceedings of the 18th Annual EPA Conference on Analysis of Pollutants in the Environment, Norfolk, VA, May, 1995, p. 107.

3 STANDARD METHODS FOR THE EXAMINATION OF WATER AND WASTEWATER, APHA, AWWA AND WEF, Washington, D.C. (19th Edition, 1995) p.4-22.

4 Csikai, N.J.; Barnard, Jr., A.J. DETERMINATION OF TOTAL CYANIDE IN THIOCYANATE-CONTAINING WASTES, Anal. Chem. 1983, 55, 1677.

5 ASCORBIC ACID PRESERVATION MAY LEAD TO ERRONEOUS CYANIDE RESULTS, Water Environment Solutions, Sample issue, 1994.

6 Goldberg, M.M.; Clayton, A.A.; Potter, B.B. THE EFFECT OF MULTIPLE INTERFERNCES ON THE DETERMINATION OF TOTAL CYANIDE IN SIMULATED ELECTROPLATING WASTE BY EPA METHOD 335.4, Proceedings of the 17th Annual EPA Conference on Analysis of Pollutants in the Environment, Norfolk, VA, May, 1994, p. 395.

7 Goldberg, M.M.; Clayton, A.A. EFFECTS OF METALS, LIGANDS, AND OXIDANTS ON CYANIDE ANALYSIS: GOLD MINING WASTE CASE STUDY, Proceedings of the 18th Annual EPA Conference on Analysis of Pollutants in the Environment, Norfolk, VA, May, 199, p. 87.<

8 Milosavljevic, E.B.; Solujic, Lj.; Hendrix, J.L. RAPID DISTILLATIONLESS "FREE CYANIDE" DETERMINATION BY A FLOW INJECTION LIGAND EXCHANGE (FI/LE) METHOD, Environ. Sci. Technol. 1995, 29, 426.