Orientation and Training
Food and Drug Administration
I. Simplify the problem if possible
If the preliminary information indicates that a certain contaminant (or type of contaminant) is present, then analytical efforts should target that material. Initial observations and odors can provide important clues. For example, the odor of the sample may suggest the presence of bleach, ammonia, amines or a petroleum product such as gasoline, thinner or a vehicle for an emulsifiable concentrate of a pesticide. Analytical methodology can be obtained from compendia (9, 10, 11) or the general chemical literature (12, 21) but often these procedures are modified to accommodate differences in the sample matrix. The analyst may need guidance from an experienced analyst for method modifications and validations. Method validation studies will demonstrate the method is usable to support the conclusions of the analysis for the sample and the situation to which it was applied.
II. When there isn't much to go on
The remainder of this section addresses situations where there is little information to provide focus for the analyst.
If moderate shaking of the sample (see Section 14.6 before shaking) produces a persistent foam, then the presence of a surfactant which might be associated with a cleaning product is indicated. If the foam remains when a small portion of the sample is acidified, then the surfactant is probably a detergent. This form of tampering is relatively common and there are a variety of tests to determine which class of surfactant is present (13, 14). A soap is indicated when the foam produced by moderate shaking does not remain after acidification.
Exposing the sample to ultraviolet light may indicate that a component of the contaminant is a fluorescent material like many of the dyes which are associated with antifreeze. This would suggest that the sample be further examined for the presence of ethylene glycol.
A good early chemical test is an estimate of pH with test paper (e.g. pHydrion paper (1 to 12), MicroEssential Laboratory or colorpHast pH 0 –14 strips). A shift in pH relative to a control sample indicates a change in the sample which should be explored. If the pH is less than 2 or greater than 10, then the presence of a corrosive acid or base is likely. Further investigation might include examination by ion chromatography (IC) for anions, such as chloride or nitrate, or inductively-coupled plasma atomic emission spectroscopy (ICP-AES) or mass spectroscopy (ICP-MS) to indicate the presence of a metalloid and/or complex anion (such as PO42-). IC can also be used to detect some organic acid; gas chromatography – mass spectrometry (GC-MS) can be used to detect organic acids or bases.
Other simple chemical tests might include cyanide screen (Cyantesmo test paper) and sulfide screen with lead acetate paper.
Spot tests for the presence of oxidizing agents using diphenylamine in sulfuric acid (15) or Starch/Iodide paper can provide useful information.
Experience with actual tampering cases has shown that the contaminant is typically present in relatively large amounts. However, when the examination reveals no remarkable differences, it is important to have a sense of the adequacy of the screening procedure. The toxicity of a wide variety of substances has been described based upon ranges for the probable oral lethal dose (4). If we allow for a margin of safety of 3% of the probable oral lethal dose, then a warning concentration for a contaminant in a product can be defined by the following equation:
Warning = 0.03 X Probable Oral X Body Weight X 1
Concentration Lethal Dose Portion
(mg/ml)(mg/kg body wt.) (kg) (ml) -1
Using a body weight of 70 kg (average adult) with a portion size of 355 mL (e.g. a 12 oz. beverage), the table below can serve as a guide to the needed sensitivity of the analytical methods. Bear in mind that significant clinical illness can be expected at doses on the order of 10% of the probable lethal dose. This equation can be adjusted in keeping with additional information. For example, there may be a good estimate of the amount of product ingested by a child of known body weight and an identified poison with a known lethal dose may be suspected.
|Description||Toxicity Class||Probable Oral
(mg / kg body wt.)
|Slightly Toxic||2||5000 to 15000||30000|
|Moderately Toxic||3||500 to 5000||3000|
|Very Toxic||4||50 to 500||300|
|Extremely Toxic||5||5 to 50||30|
|Super Toxic||6||< 5||1|
NOTE: The minimum lethal dose for some super toxic materials can be as low as 0.1 (mg/kg body weight) and this is used to calculate this number.
III. Application of instrumental techniques
The following is a brief overview of the use of instrumental methods to compare the suspect sample and the control sample. It is recognized that not all of the instrumentation that will be discussed is found in every laboratory. A basic understanding of each technique on the part of the analyst is assumed.
Sample preparation should be minimized in order to speed the progress of analysis and to retain information. Whenever possible, direct analysis is preferred to extraction/clean-up. Good judgment on the part of the analyst(s) should be exercised when deciding which methodologies are to be use and the order in which selected methodologies should be executed.
A. Static Headspace Sampling combined with Capillary Gas Chromatography - Mass Spectrometry (GC-MS) for Volatile Materials
The technique (16) is directed at the detection and characterization of volatile materials (boiling points approximately below 200°C). It is very useful for detecting solvents (e.g. alcohols, chlorinated hydrocarbons), fragrances associated with cleaning products which frequently appear as contaminants, and petroleum products. Petroleum products may occur as contaminants in their own right or may be associated with pesticides in emulsifiable concentrates.
A small portion (10 to 500 mg) of the sample is placed in a vial which is sealed with a teflon-lined septum-cap. The vial is incubated at an elevated temperature for about 10 min and a portion of the vapor in the vial (the headspace) is withdrawn through the septum and injected into the GC-MS for analysis. Limit of Detections (LODs) for typical analytes extend to the low mg/kg range.
Mass spectra which are associated with observed differences between the suspect sample and control sample are compared to reference spectra (17) to obtain tentative identification. This may be subsequently confirmed and quantified through the analysis of standards and demonstrated recovery of a standard which has been spiked into the control sample.
B. Capillary Gas Chromatography with Mass Spectrometry (GC-MS) for Volatile and Semi-volatile Materials
Direct injection of the sample or an extract/solution of the sample in methanol is the first choice.
If the nature of the sample precludes this or if some pre-concentration is needed, then the sample may be extracted with acidified aqueous acetonitrile (pH = 3) and the acetonitrile subsequently isolated by “salting out” for analysis by GC-Flame Ionization Detector (FID) or GC-MS. Isolation of a basic extract is obtained using the same procedure with basic aqueous acetonitrile (pH = 10). Additional sensitivity can be obtained by evaporative concentration of the acetonitrile extracts. This protocol is modeled upon a multi-residue pesticide method (18) in conjunction with an extraction procedure from a guide to forensic analysis of pharmaceuticals (15).
The range of applicability can be extended to functionally non-volatile materials by silylating the extracts after solvent exchange into pyridine by the addition of N,O-bis-(trimethylsilyl)-trifluoroacetamide which contains 1% trimethylchlorosilane with incubation at 60 degrees C for at least 15 min and repeating the analysis.
The suspect sample and the control sample are compared to expose differences. The mass spectra associated with these differences are compared to reference spectra to provide tentative identification. Occasionally, the tentative identification proceeds from first principles (19).
C. Ion Chromatography (IC)
The role of this technique is to detect anions such as those in the table, below. There is some overlap with GC-MS and Inductively-Coupled Plasma (ICP) methodology but ICP occupies an important niche. Unless circumstances dictate otherwise, analysis of a sample is performed on a ten-fold dilution of the sample in water using a conventional anion column (eg. Dionex AS9-HC, or equivalent) in suppressed mode with a 9 mM sodium carbonate buffer. Additional sample preparation such as further dilution or the use of sample preparation cartridges such as C-18, Ba (for sulfate removal), or H+(neutralization) may be needed for some samples. More sophisticated chromatography such as the use of hydroxide gradients may be needed with the most complicated samples. It is worth noting that a reactive contaminant, such as bleach, may produce additional peaks in a suspect sample that can only be properly characterized through additional experiments which monitor the impact of the addition of the nominal contaminant to a control sample through time. The probable presence of bleach in the suspect sample at some point in time is implied by elevations in the chloride and chlorate levels along with high pH and the presence of excess sodium.
Some Potentially Toxic Anions and the Associated Warning Level (WL)
|Toxicity Class 6||Toxicity Class 5||Toxicity Class 4|
|WL = 1 mg/l||WL = 30 mg/l||WL = 300 mg/l|
D. High Performance Liquid Chromatography - Mass spectrometry (HPLC-MS)
A wealth of methodology using HPLC and Thin Layer Chromatography (TLC) for analyzing certain compounds or classes of compounds is found in compendia (10, 11) and selected references (15). However, because of reduced information content of UV-Vis spectra with respect to MS data, HPLC techniques are somewhat inefficient for general screening. Advances in the application of mass spectrometry to liquid chromatography have made HPLC-MS the method of choice for detecting non-volatile or thermally unstable compounds (e.g. cardiac, glycosides, alkaloids, proteins) which are not amenable to determination by GC-MS. HPLC-MS is especially useful as a screening tool if a comparison sample is available. There are some restrictions on mobile phase composition which must be considered, and interpretation of results requires a degree of skill and experience due to lack of spectral libraries for HPLC-MS.
E. Inductively-Coupled Plasma Atomic Emission Spectroscopy (ICP-AES) for the Detection of Metals, Metalloids and Some Complex Anions
Many liquids and some water soluble solids can be analyzed directly, or with a single dilution, by ICP-AES and ICP-MS. However, in general, samples are solubilized or digested to remove organic components with concentrated acid. Modifications may be needed to digest the sample properly in order to detect certain elements (e.g. Hg). (20)
Selected Metals with Toxicity Class and Warning Level (WL)
|Toxicity Class 6||Toxicity Class 5||Toxicity Class 4|
|WL = 1 mg/l||WL = 30 mg/l||WL = 300 mg/l|
The form of the element defines its toxicity. This can create difficulties in the interpretation of comparative results and limit the usefulness of elemental analysis in some cases. For example, Phosphorous, as Yellow Phosphorous (Toxicity Class 6) is a significant hazard but Phosphorous, as Phosphate (Toxicity Class 3) is commonly encountered in foods. Inorganic arsenic is highly toxic (toxicity class 6), but arsenic as arsenobetaine, which is found in seafood, is considered non-toxic.
F. Fourier Transform Infrared Spectrophotometry(FT-IR)
Lack of sensitivity limits the usefulness of this procedure for the screening of solutions except in instances of gross contamination. However, FT-IR may be utilized in conjunction with a microscope for the examination of individual particles or for characterization of substances isolated by physical means. FT-IR can provide identification through matching spectral features with compilations of the IR spectra of reference compounds or it can provide indications of the presence of functional groups to supplement other analytical information.
G. UV-Visible Spectrophotometry (UV-VIS)
With some product types, compare the UV-Visible spectrum of the suspect sample dissolved in a solvent system of ethanol: 0.05M aqueous hydrochloric acid (1:1 v/v) with that of the control sample. A similar method is also applied in which the solvent system is ethanol: 0.05M aqueous sodium hydroxide. In ideal circumstances, these comparisons are capable of revealing the presence of a variety of drugs and other bio-active materials at the mg/l level (15).
However, this technique is most often encountered as an adjunct to modern HPLC instrumentation with photodiode array detectors that can spectrally characterize components during the course of a separation.