Portable cyanide detection systems are designed to be used in the field to evaluate for potential cyanide contamination of a water asset. These detection systems use one of two distinct analytical methods - either a colorimetric method or an ion selective method - to provide a quick, accurate cyanide measurement that does not require laboratory evaluation. This document discusses several portable cyanide detection systems that are currently on the market.

Aqueous cyanide chemistry can be complex. Various factors, including the water asset's pH and redox potential, can affect the toxicity of cyanide in that asset. While personnel using these cyanide detection devices do not need to have advanced knowledge of cyanide chemistry to successfully screen a water asset for cyanide, understanding aqueous cyanide chemistry can help users to interpret whether the asset's cyanide concentrations represent a potential threat. Therefore, a short summary of aqueous cyanide chemistry, including a discussion of cyanide toxicity, is provided below. For more information, the reader is referred to Standard Methods, 20th Edition, Method 4500-CN-.

Cyanide (CN-) is a toxic carbon-nitrogen organic compound that is the functional portion of the lethal gas hydrogen cyanide (HCN). The toxicity of aqueous cyanide varies depending on its form. At near-neutral pH, "free cyanide" (which is commonly designated as "CN-", although it is actually defined as the total of HCN and CN-) is the predominant cyanide form in water. Free cyanide is potentially toxic in its aqueous form (see discussion under Cyanide Toxicity below), although the primary concern regarding aqueous cyanide is that it could volatilize. Free cyanide is not highly volatile (it is less volatile than most VOCs, but its volatility increases as the pH decreases below 8). However, when free cyanide does volatilize, it volatilizes in its highly toxic gaseous form (gaseous HCN). As a general rule, metal-cyanide complexes are much less toxic than free cyanide because they do not volatilize unless the pH is low (more).

Analyses for cyanide in public water systems are often conducted in certified laboratories using various EPA-approved methods, such as the preliminary distillation procedure with subsequent analysis by a colorimetric, ion selective electrode, or flow injection methods. Lab analyses using these methods require careful sample preservation and pretreatment procedures and are generally expensive and time consuming. Using these methods, several cyanide fractions are typically defined:

As discussed above, these different methods yield various different cyanide measurements which may or may not give a complete picture of that sample's potential toxicity. For example, the "Total Cyanide" method includes cyanide complexed with metals, some of which will not contribute to cyanide toxicity unless the pH is out of the normal range. In contrast, the "WAD Cyanide" measurement includes metal-complexed cyanide that could become free cyanide at low pH, and "Amenable Cyanide" measurements include metal-complexed cyanide that could become free cyanide at high pH. Personnel using these kits should therefore be aware of the potential differences in actual cyanide toxicity versus the cyanide measured in the sample under different environmental conditions.

Ingestion of aqueous cyanide can result in numerous adverse heath effects and may be lethal. EPA's Maximum Contamination Level (MCL) (an enforceable drinking water standard maintained by EPA) for cyanide in drinking water is 0.2 mg/L (0.2 parts per million, or ppm). This MCL is based on free cyanide analysis per the "Amenable Cyanide" method described above (EPA has recognized that very stable metal-cyanide complexes such as iron-cyanide complex are non-toxic [unless exposed to significant UV irradiation], and these fractions are therefore not considered when defining cyanide toxicity). Ingestion of free cyanide at concentrations in excess of this MCL causes both acute effects (e.g. rapid breathing, tremors, and neurological symptoms) and chronic effects (e.g. weight loss, thyroid effects, and nerve damage). Under the current primary drinking water standards, public water systems are required to monitor their systems to minimize public exposure to cyanide levels in excess of the MCL.

HCN gas is also toxic, and the Office of Safety and Health Administration (OSHA) has set a permissible exposure limit (PEL) of 10 ppmv for HCN inhalation. HCN also has a strong, bitter, almond-like smell and an odor threshold of approximately 1 ppmv. Considering the fact that HCN is relatively non-volatile (see discussion of cyanide volatility under the Cyanide Chemistry section above), a slight cyanide odor emanating from a water sample suggests very high aqueous cyanide concentrations (greater than 10 to 50 mg/L), which is in the range of a lethal or near lethal dose with the ingestion of one pint of water.

The portable cyanide analysis kits currently on the market use either the ion selective electrode method or the colorimetric method to analyze for cyanide. The portable cyanide analyzers based on the ion selective electrode method are designed to measure only free cyanide (CN- + HCN) and do not measure any significant levels of weak or strong cyanide complexes or insoluble cyanide species in a water sample. In contrast, the colorimetric methods measure both free cyanide and some fraction of the weakly-complexed cyanide (more). These two analytical methods are discussed in more detail below:

The colorimetric method for measuring cyanide measures the intensity of color produced by the reaction of the reagent solutions and the sample. Color intensity is quantified using a portable photometer. The first step in using this method is to add reagent solutions to the test water sample (more). The photometer then measures the intensity of the color, which is proportional to the concentration of cyanide. The equivalent cyanide concentrations are estimated and displayed on the digital display of the photometer.

Ion selective electrode sensors utilize a solid-state combination reference and ion selective electrode containing a mixture of inorganic silver compounds bonded into the tip of an electrode body. The system is first calibrated using calibration solutions, the sample pH is then adjusted, and then the electrode is placed in the sample to be tested. When the electrode is in contact with cyanide-containing solution, silver ions dissolve from the membrane surface. Ions within the sensing element then move to the surface of the element, resulting in a concentration-dependent electrical response in the sensing element. The electrode sensor measures this potential difference and converts it to a cyanide concentration, which it can display in mg/L or PPM.

Colorimetric photometers and ion selective electrode sensors have different performance attributes that may make one of these measurement devices more useful than the other for certain applications. In general, colorimetric photometers have shorter test times and higher accuracies than do ion selective electrode sensors, and they are generally less expensive than cyanide selective electrode sensors. However, high or low temperatures in the samples can affect cyanide chemical reaction rates, decreasing the accuracy of the calibration and thus decreasing the accuracy of the measurement. In contrast, the calibration standards for ion selective electrode sensors can be maintained at temperatures similar to the sample temperature, which eliminates this potential problem. In addition, ion selective electrode sensors tend to be more precise than colorimetric photometer methods.

The EPA ETV program has evaluated six portable analyzers for detecting cyanide in water samples, including four kits that use colorimetric methods, and two electrode sensors. Each product was evaluated for accuracy, precision, linearity, method detection limit (MDL), inter-unit reproducibility, lethal/near-lethal dose response, and other factors. Specific interferences for each method were not determined in the ETV program, but the data show that matrix interferences occurred when comparing the kit measurements to the laboratory reference method (more).

EPA evaluated solutions ranging from 0.03 to 0.8 and 0.03 to 25 mg/L cyanide using the colorimetric photometer and cyanide selective electrode sensor technologies, respectively. This represents a wide range of detection capability, ranging from well below the MCL to well above the MCL. Following is a summary of the ETV program verification results for the six products analyzed. Detailed discussion of these test results is located at http://www.epa.gov/etv/verifications/vcenter1-23.html.

Accuracy is a measure of how close a measurement is to the true value. The measured difference between sample reading and the true value is referred to as "bias." The ETV documentation reported bias relative to a laboratory reference method (LRM) measuring "Amenable Cyanide" by the shortcut direct colorimetric method (more). ETV data reported all bias results as positive values (absolute values), and the raw data must be inspected to determine whether the measured bias was positive or negative (i.e., whether the kit recovered more cyanide than the LRM or the kit did not recover as much cyanide as the LRM). A bias of exactly 100 percent was reported for samples where the method measurement was non-detect and the LRM returned an "Amenable Cyanide" concentration above the MDL of the kit.

In general, EPA found that the bias for the colorimetric methods ranged from 2 to 100 percent (more). All of the colorimetric photometers and cyanide selective electrode sensors evaluated were able to effectively measure cyanide at lethal or near-lethal cyanide levels. However, the cyanide selective electrode sensors were biased high for many of the samples tested, and the overall biases for the cyanide selective electrode sensor methods were slightly higher than those observed in the colorimetric photometer methods, with a maximum bias of 128 percent.

Precision measures the repeatability of a measurement (e.g., the bias in measuring the same sample a number of times). EPA's ETV program expresses precision as the Relative Standard Deviation (RSD) of replicate analyses. The RSD for the colorimetric photometers generally ranged from zero to 100 percent. The RSD for the cyanide selective electrode sensors were significantly lower than the colorimetric photometers, ranging from zero to 23 percent.

The colorimetric photometer method MDLs verified by the ETV program were significantly lower than the cyanide MCL and ranged from 0.004 to 0.034 mg/L. However, of the two cyanide selective electrode sensors evaluated, only the Thermo Orion Model 9606 Cyanide Electrode with Model 290 A+Ion Selective Electrode Meter is capable of detecting cyanide concentrations near the MCL. WTW Measurement Systems' Cyanide Electrode CN 501 has an MDL of 0.271 mg/L, and thus this device can only detect cyanide concentrations at levels that are potentially toxic (50 to 250 mg/L).

Although specific interferences were not evaluated during ETV program testing, it is clear that matrix effects for many of the water samples caused bias when comparing the colorimetric photometer method analyses to a LRM that analyzed for "Amenable Cyanide." Potential interferences that could bias the colorimetric photometer method low include sulfides (which are often present in wastewater samples), organic aldehydes (e.g. formaldehyde), and heavy metals. The colorimetric method can be biased high by the presence of thiocyanate or cyanogen chloride (CNCl). However, thiocyanate would not be expected in drinking water samples unless contamination from an unusual industrial source occurs. In contrast, CNCl can be formed in drinking water as a result of the oxidation of free cyanide by chlorine (active disinfectant in drinking water), and thus the presence of CNCl would be a good indicator of cyanide contamination.

For the cyanide selective electrode sensor methods, potential interferences that would cause low bias include heavy metals, sulfide, and iodide. Unless there is an unusual application of iodine (I2) as a drinking water disinfectant, none of these interferences would be expected to be problematic for drinking water samples. However, as described above for the colorimetric method, sulfide is often present in wastewater samples and could result in significant low bias of any cyanide measurements.

A summary of the detection parameters achieved by the kits and electrodes tested in the ETV program is provided below.

Costs for portable cyanide analyzers vary depending on the technology. In general, costs for colorimetric photometer kits range between $400 and $1,000. These costs include all equipment necessary to conduct the analysis, and generally include a photometer and the necessary reagents to analyze between 30 and 50 samples.

Cyanide selective electrode sensors cost approximately $1,600 to $1,800. These costs generally include the ion meter, an electrode, the reagents necessary for pH adjustments, the electrode fill solution, and a carrying case or stand.

Individual costs for the products evaluated by the EPA ETV program are summarized in Table 2 below:

The EPA ETV program has verified six portable cyanide analyzers. Following is the list of vendors for these six analyzers. Additional information on these technologies can be found at http://www.epa.gov/ETV/verifications/vcenter1-23.html.