Jump to main content.


Trace Organic Analysis

Regulatory Mass Spectrometry

William C. Brumley and James A. Sphon

Division of Chemistry and Physics, Food and Drug Administration, Washington, DC 20204, USA

Original citation:
W. C. Brumley and J. A. Sphon, "Regulatory Mass Spectrometry," Biomed. Mass Spectrom., 8, 390-396 (1981)

The history of regulatory mass spectrometry at the Food and Drug Administration began In the early 1960s and was initiated by J. N. Damico using a time-of-flight instrument with limited mass range, resolution and sensitivity. Early work involved confirmation of identity of compounds using direct probe introduction and lull mass scans. From 1964 to 1969 an important application of mass spectrometry was the analysis of pesticides and elucidation of their fragmentation pathways. Regulatory mass spectrometry was used to confirm manually trapped gas chromatographic peaks and it solved the problem of false positive identifications based solely on gas chromatographic retention time. In time the mass spectrometer became viewed as the ultimate gas chromatographic detector. Another important early regulatory case involved the question of Krebiozen as a cancer cure. Krebiozen was analyzed by mass spectrometry and found to consist of creatinine, which was known to have no activity as an anti-cancer drug. Such basic information as the identity of a substance demonstrated the indispensable regulatory use of mass spectrometry. The basic problem of identification has broadened in scope to include multicomponent analyses, trace level detection, quantitation, and newer ionization techniques. Two examples illustrate the continuing development of regulatory mass spectrometry. Negative ion chemical ionization mass spectrometry in the analysis of aflatoxin BI involves use of a newer ionization technique. The requirements of trace level detection and specificity are further stretched in the gas chromatographic mass spectrometric selected ion monitoring detection, confirmation and quantitation of tetrachlorodibenzo-p-dioxins.

INTRODUCTION

We would like to trace the historical development and applications of mass spectrometry at the Food and Drug Administration (FDA). Along this historic path, a central analytical problem concerns the identification of substances. But identification is a very broad subject and therefore often challenges our ingenuity and skill. Today's analyst may be faced with mixtures and matrices of enormous complexity, analytes at extremely low levels, and problems of liability and nonvolatility. Adding complexity to organic analysis is the sheer number of compounds, which now exceed several million according to the Chemical Abstracts Service registry. The regulatory environment places special emphasis on the necessity for unequivocal identification. In addition, several other factors must be considered, including practicality and reliability. Regulatory mass spectrometry at FDA attempts to integrate a rapidly developing science and technology into this environment while maintaining the primary mission of ensuring protection of the public health. Regulatory mass spectrometry is defined as the use of mass spectrometry to obtain an identification of a substance with sufficient specificity to support a charge of its presence in a court of law. Historically, certain guidelines have developed to make it possible to judge when an analytical technique has reached the level of confidence required for a regulatory assay. Thus, the theme of identification serves as the foundation of regulatory mass spectrometry and provides a basis from which to view its historical development. At one time, full scan electron impact mass spectra were widely and almost exclusively used as the benchmark for specific identification. Today, the mass spectrometrist has a myriad of techniques old and new to use in solving the identification problem. We will endeavor to point out some of our concerns and reservations on adopting new techniques and our difficulties in maintaining the requirements of regulatory mass spectrometry while involved in 'the search for zero.'

DEVELOPMENT OF REGULATORY MASS SPECTROMETRY

First instrument

Mass spectrometry at FDA began in 1960 with the purchase of a Bendix Model 14 time-of-flight mass spectrometer. This first mass spectrometer had a limited mass range of about 300 daltons and could achieve a resolution of about 200. Adequate full mass spectra could be obtained with milligram amounts of sample. Information from the mass spectrometer, consisting of mass spectra, was manipulated by hand after recording the data on photographic paper.

In 1961, Joseph N. Damico began his investigation into the applications of mass spectrometry. Mr. Damico's early laboratory notebooks contain the seeds of most of the problems our new laboratory is focused on today. Compounds he encountered included pesticides such as methyl parathion, drugs and biological substances such as creatinine, mycotoxins such as aflatoxins, volatile compounds from fish such as methylethyl ketone, polyaromatic hydrocarbons such as coronene, chlorinated compounds such as aldrin and dioxins, antioxidants such as butylated hydroxyanisole, nitrosamines such as dimethyl nitrosamine, and sterols and other essential oils and fats. He must be credited with the vision of the expanding role of mass spectrometry in so many areas of FDA activity.

The early mass spectrometry work required extensive clean-up of samples, sometimes accompanied by manual trapping of peaks from gas chromatographs. Samples were usually run by solid probe introduction, ionized by electron impact and recorded using full mass scans.

We might better appreciate the early applications of mass spectrometry at FDA by noting certain historical conditions in 1961 as shown in Figure 1.

Chart of historical development - for further information contact brumley.william@epa.gov

Figure 1. Historical setting and development of mass spectrometry at FDA.

There was one mass spectrometry laboratory for all of FDA. There was no Environmental Protection Agency (EPA), although several existing government laboratories were equipped with mass spectrometers. There was no commercial instrument with direct coupling via a separator of the mass spectrometer and the gas chromatograph. Early data systems were extremely limited, and automatic searching of large libraries of mass spectra was unknown. Various newer ionization techniques were to come into use later in the decade or in the 1970s.

One of the earliest regulatory applications of mass spectrometry at FDA illustrates our subject.

Krebiozen

In 1963, a regulatory case concerned the proposed anti-cancer drug Krebiozen. Controversy abounded over what Krebiozen actually was and over its effectiveness in treating cancer [1]. The mass spectrum of authentic Krebiozen was obtained at FDA and other laboratories along with other spectroscopic data and compared with that of a well known compound, creatinine. The mass spectrum of Krebiozen clearly supported the contention that Krebiozen consisted of creatinine, which was known to possess no anti-cancer activity.

Creatinine

Image  of molecular line structure - contact brumley.william@epa.gov for further information

The spectrum was obtained as a full mass scan at low resolution and contained ions at m/z 113, 84, 69, 43 and 42, for example. The occurrence of ions of specific the primary information of a mass spectrum. The absence of response at various ions thus represents information as well. Later, we will return to the interrelation of ionization, fragmentation and the scanned mass range in determining specificity.

At that time mass spectral identifications were part of research procedures rather than regulatory methods. This meant that in a legal context the identification had to be defended on its own merits. As developed later in this paper, regulatory methods are validated, in part, through collaborative efforts.

Pesticides

Efforts in the drug analysis area continued but were perhaps overshadowed by application of mass spectrometry to the analysis of pesticides and the correlation of their fragmentation to structure during the period from 1964 to 1969. In the early part of this period, the major emphasis was on obtaining the basic data for a series of compounds and the relationship of structures to spectra. Later in this period, as instrumentation improved, and basic information became correlated, the emphasis began to shift to applications and how the data could be used to solve analytical problems such as identification of metabolites and unknown compounds.

Examples of pesticides investigated include chlorinated compounds such as pp'-DDT, carbamates such as carbaryl, and organophosphorus compounds such as methyl parathion. Usually, the sample was manually trapped from the gas chromatograph or otherwise purified and run as a solid probe sample with spectra recorded as full mass scans.

Methyl parathion
Carbaryl
Image  of molecular line structure - contact brumley.william@epa.gov for further information
Image  of molecular line structure - contact brumley.william@epa.gov for further information

Much of the work on pesticides through 1969 was summarized in the chapter on pesticides written by Mr. Damico in the book Biochemical Applications of Mass Spectrometry which was published in 1972 [2]. High resolution and some metastable studies reported in this work were performed by collaboration with other laboratories using their instrumentation. The spectrum of methyl parathion illustrated the utility of mass spectrometry in characterizing compound classes. Besides the molecular ion at m/z 263, an ion occurred at m/z 125. In this spectrum, m/z 125 was assigned a structure, [(CH3O)2P=S]+, which was supported by metastable data and high resolution determination of elemental composition. This ion is characteristic of organophosphorothioates.

The fundamental knowledge of fragmentation had an important application in the structural identification of metabolites of pesticides and was illustrated in the case of a metabolite of carbaryl. Carbaryl (mol. wt 201) is one of the N-methylcarbamate pesticides. Their spectra are characterized by the loss of the methyl isocyanate moiety, CH3-NCO, resulting in the appearance of the [M - 57]+. ion. A characteristic loss of CO from the [M- 57]+. ion is also exhibited. A metabolite of carbaryl isolated from cow urine and milk gave a mass spectrum which paralleled that of carbaryl with some additional features. Losses of water occurred from the molecular ion, the [M - 57]+. rearrangement ion, and the [M - 57-CO]+. ion. The metabolite was identified by mass spectrometry as 5,6-dihydro-5,6-dihydroxycarbaryl (mol. wt= 235), and the isomer assignment was supported by infrared and nuclear magnetic resonance data as well. The assignment of structure to an adulterant of food is important before toxicological and other studies may begin.

These applications indicated the potential of full scan mass spectrometry for specificity in identifications. In addition, applications of metastable ion studies to fragmentation and isomer problems were initiated. These early studies were concerned with establishing that fragments arose by successive decompositions from the original molecular ion. The idea of ion decomposition observed by metastable techniques was developed later and applied by others to problems of differentiating isomers and identifying ion structures.

This early work in the mass spectrometry of pesticides brought to the attention of analysts the potential problem of false positives (incorrect identification) in pesticide analysis. At that time gas chromatography with flame ionization detection was often used to confirm the identity of a pesticide in a sample primarily on the basis of its gas chromatographic retention time. However, it was suspected that some of these identifications could be false positives due to heretofore unknown interferences.

Later, a highly sensitive and specific detector for gas chromatography was introduced and became known as the electron capture detector. Nevertheless, potential interferences and unidentified peaks remained a real problem even with selective detection. Mass spectrometry identified some of these previously unknown substances as polychlorinated biphenyls.

IMPROVED TECHNIQUES

Mass spectrometry, with its inherent high specificity, rapidly became recognized for its power, due in part to new or improved techniques such as the combined gas chromatograph mass spectrometry instrument. At that time the unpublished or unformalized requirements of regulatory mass spectrometry were being satisfied by improvements in the reliability and practicality of Instrumentation. The inherent specificity of mass spectra was more fully appreciated and utilized as the underlying processes producing spectra became better understood.

Newer ionization techniques such as field ionization and later field desorption offered the possibility of simplifying the interpretation of data by reducing fragmentation and giving clear molecular weight information. However, the issues of specificity and practicality became real concerns, and spectra were not always simple to interpret.

Regulatory guidelines-veterinary drugs

Around 1975 some of these concerns about the acceptability of regulatory methods began to crystallize, and a need for some type of guidelines was recognized. In terms of instrumentation at that time, gas chromatographic mass spectrometric technology was firmly in place. Data systems were becoming commonplace, and were recognized as indispensable in the processing of data generated by gas chromatographic mass spectrometric analyses. Data systems had further significance in the control they provided over instrumentation, in the computer searching of mass spectral libraries, and in the automatic storage of data.

Newer ionization techniques became available to expand capabilities and supplement the historical use of electron impact ionization. These techniques included positive ion and negative ion chemical ionization mass spectrometry; field ionization and field desorption mass spectrometry; 252 desorption; and atmospheric pressure ionization. Another area developing rapidly in the application of mass spectrometry to analysis involved metastable ion techniques such as collision-induced decomposition of ions and mass spectrometry mass spectrometry.

Some of these mass spectrometric developments emphasized the need to question objectively the appropriateness of any method of identification. Whereas the early mass spectrometry work employed electron impact ionization and full mass scans, contemporary work may involve a newer ionization technique with few fragment ions or a limited selection of masses over which to scan (often called multiple ion detection).

The introduction of quadrupole mass spectrometers and chemical ionization mass spectrometry complicated the evaluation of mass spectral data. In the past when all data were obtained on magnetic sector instruments tinder electron impact ionization, the results were largely transferable from laboratory to laboratory. The different performance characteristics of the quadrupole mass spectrometer introduced a new variable in the evaluation of data and in computer searching of mass spectral libraries. The use of chemical ionization mass spectrometry, with its dependence on source pressure and temperature and on reagent gas, also introduced new variables and complexity to the overall considerations of mass spectral data.

Major problems began in the regulatory evaluation of methods using mass spectrometry with the promotion of the very powerful and sensitive techniques called multiple ion detection or multiple ion monitoring and its corollary, single ion monitoring. The application of single ion monitoring again raised the possibility of false identification. For example, Table 1 [3] lists compounds which have similar retention times on OV-101 according to the Pesticide Analytical Manual [4]. The polychlorinated biphenyl compounds occur in commercial mixtures known as Aroclors 1242, 1248 and 1262. False identifications are marked by X's if only the ion current corresponding to the most abundant ion in the molecular weight region of the compound to be detected was monitored. Thus, for example, the occurrence of octachlor epoxide in an environmental sample would give a false positive result for sulphenone determined by monitoring m/z 252. In fact, octachlor epoxide would interfere with the single ion monitoring of the other six compounds at the appropriate mass listed along the top of the table. Furthermore, monitoring several ions in the [M]+. cluster of a chlorinated compound does not guarantee that the desired specificity has been obtained. The 13C isotope peaks for the [M- Cl.]+ cluster of octachlor epoxide (m/z 386, 388, 390, etc.) give an almost identical response to the [M]+. cluster of heptachlor epoxide if only the major ions in the heptachlor epoxide cluster are monitored as illustrated in Figure 2.

Table 1. Common peaks observed in the mass spectra of coeluting compounds under Pesticide Analytical Manual [4] conditions[a]

Compound (mol. Wt)
m/z 422
m/z 388
m/z 358
m/z 326
m/z 320
m/z 292
m/z 252
octachlor epoxide (420)
M+2
X
X
X
X
X
X
heptachlor epoxide (386)
-
M+2
-
-
X
X
X
chlorfen-vinphos (358)
-
-
M
-
-
X
X
pentachlor-obiphenyls (324)
-
-
-
M+2
-
X
-
phenoate (320)
-
-
-
-
M
-
-
tetrachloro-biphenyls (290)
-
-
-
-
-
M+2
-
sulphenone (252)
-
-
-
-
-
-
M
a The X entries represent ions from the compound to the left that could produce false identifications for other compounds based in the single ion monitoring. The M or M+2 entries designate the ion of the greatest relative abundance in the Molecular ion cluster of the compound to the left.

Image of chromatograph - contact brumley.william@epa.gov for further information

Image of chromatograph - contact brumley.william@epa.gov for further information

Figure 2. Electron impact mass spectra of (a) heptachlor epoxide (mol. wt = 386) and (b) octachlor epoxide (mol. wt = 420); CH5-DF; 70 eV; ion source 200 C; direct introduction.

In contrast, full or limited mass scan data clearly differentiate the two compounds. For these reasons the interrelation of sample origin and clean-up to the method of analysis must be considered carefully. The absence of response for certain ions may be as important to know as the presence of responses for certain ions.

Recourse to high resolution single ion monitoring does not necessarily provide adequate specificity but is dependent on the context of analysis and the particular experimental manner in which the high resolution experiment is carried out. Lack of control over environmental and food samples restricts possible assumptions about potential interferences. On the other hand, assumptions may be practical in biological assays such as feeding studies carried out under carefully controlled conditions. There are several operational approaches to carrying out the high resolution experiment. Among these is the experimental technique of peak profiling, which reveals peak centroids over a limited mass range at high resolution. Such a technique is advantageous in that it directly relates instrument resolution to the observed sample responses.

The analytical clean-up itself presents the analyst with a limited set of compounds from the sample. The specificity of this clean-up and the contaminant level are intimately related to the overall selection of method for regulatory confirmation. The term regulatory confirmation refers to a method or procedure such as mass spectrometry for the specific identification of a compound or substance.

In practice, regulatory confirmation bases its identification on some chemical or physical index or set of indices which serves to distinguish one compound from any other. An analogy has been drawn between chemical identification by mass spectrometry, for example, and personal identification by means of fingerprints. To state that the mass spectrum of a compound is a fingerprint to its identity is clearly an oversimplification, but to continue the analogy for a moment, when the mass spectrum is limited to a few specific ions, then we have available only a partial fingerprint.

A set of guidelines has appeared in the literature which tries to come to grips with the idea of how much information is required for an identification [5]. Three choices were discussed in animal drug analysis and are listed in order of preference: full scan data, limited scan data, or multiple ion detection down to a limit of three ions. As an example, consider the mass spectrum of sulfamethazine obtained using electron impact ionization. The spectrum is dominated by the [M - 64]+. and [M-65]+. ions at m/z 214 and 213, and a peak corresponding to the molecular ion is absent. An ion at m/z 123 likely arises from the pyrimidyl portion of the molecule, while ions at m/z 156, 140, 108, 92, 80 and 65 presumably derive from the sulfanil portion of the molecule. Even without the molecular ion, the full spectrum is highly specific.

Sulfamethazine

Image  of molecular line structure - contact brumley.william@epa.gov for further information

The full scan spectrum under chemical ionization conditions using methane as reagent gas is rather devoid of relatively abundant ions except for three: two positive and one negative ion. The positive ion at m/z 124 arises from the pyrimidyl portion of the molecule. The [M+ H]+ ion at m/z 279 establishes the molecular weight of the molecule. The negative ion at m/z 155 characterizes the sulfanil portion of the molecule. Finally, the multiple ion detection gas chromatographic mass spectrometric confirmation of sulfamethazine using these same three ions produced by positive and negative ion chemical ionization mass spectrometry may be an acceptable confirmation of the compound in the context of the analysis at hand, but note that an enormous amount of information is lost by excluding all other masses from the scan. Certainly the full scan negative ion spectrum by itself would not provide sufficient specificity for the unambiguous identification of sulfamethazine. In addition, the issue of isomer specificity remains unaddressed.

Three ions were proposed as a bare minimum for multiple ion detection or for full scan data when there are a limited number of fragments as occurs with newer ionization techniques such as chemical ionization or field desorption mass spectrometry. Three ions and their relative abundances may by appropriate for a specific identification, but the analyst's experience in the context of the problem should guide him to what is adequate. For example, in the analysis of sulfamethazine the choice of the [M+H]+ ion and adducts at [M+29]+ and [M+41]+ appears inadequate to identify sulfamethazine. In general, the greater the number of ions monitored, the greater the confidence in the identification.

Veterinary drugs and the determination of their residues in food by mass spectrometry is a relatively new endeavor in the FDA. During the past several years mass spectrometrists have been requested to review veterinary drug petitions and comment on the adequacy of proposed confirmatory methods which at this stage are considered research procedures. If deemed adequate, the confirmatory procedures, along with the proposed determinative procedure, are subjected to method trials which are collaborative efforts, and are evaluated by five criteria developed for acceptance: dependability; practicability; accuracy; a reliable, sufficient limit of detection known as the lowest limit of reliable measurement and specificity [5]. As mentioned earlier, specificity must satisfy a level of uniqueness such that the mass spectrum must without question be due to the compound being measured. After method trial, the research procedure may be adopted as a regulatory method.

Although it became necessary to develop regulatory guidelines for the evaluation of these proposed methods, no rigid set of rules has been set down. We believe it is preferable to leave experimental details to the judgment of the scientist who is developing the method. In this way, a certain fairness is maintained by allowing a variety of choices of techniques, instrumentation, and economics in the analysis. It must be realistically added here that regulatory mass spectrometry lags behind mass spectrometric research frontiers by at least five years. The reasons for this are practical ones. Instrumentation must be commercially available to satisfy practicability requirements. The lag period is not necessarily bad. It provides time for a body of experience to build up using the new technique. After the lag period, it is easier to make judgments as to whether or not the technique satisfies the requirements of regulatory mass spectrometry.

EXAMPLES OF CURRENT DEVELOPMENTS

We would now like to turn to two recent examples of continuing development of regulatory mass spectrometry. The two examples, aflatoxin B1 and tetrachlorodibenzo-p-dioxin (TCDD), involve analytical and instrumental developments which cause us to confront the dilemma created by the search for zero on one hand and the requirements of the regulatory method on the other. Currently, we are using research procedures to analyze for these compounds and are investigating the use of newer techniques such as collision induced decomposition of ions.

The occurrence in foods of aflatoxin B1, a potent carcinogen which arises as a metabolite of certain molds is a growing concern. The presence of this compound in peanut butter, for example, may represent a significant health risk to the young due to their high consumption of this foodstuff. Traditionally, the confirmation of aflatoxin B1 has been based on the chick embryo assay, which requires microgram amounts of material isolated by thin-layer chromatography at some considerable effort, time and expense [7].

We have investigated the analysis of aflatoxin B1 by negative ion chemical ionization mass spectrometry using resonance electron capture under chemical ionization conditions with methane as the reagent gas. In this instance we are using mass spectrometry to support monitoring activities for a naturally occurring food contaminant. A standard of aflatoxin B1 provided a spectrum relatively devoid of ions except at m/z 312 [M]-., 311 [M-H.]- and 297 [M-CH3.]- in the full scan mode. A naturally contaminated sample of peanut butter made from the 1980 peanut crop was subjected to extraction clean-up, in part, using two-dimensional thin-layer chromatography. The sample extract afforded the spectrum shown in Figure 3. The spectrum of aflatoxin B1 from the sample (10.5 ppb level) closely resembles that of a standard. The absence of large relative response on all ions scanned except the three mentioned is pertinent. The specificity of the confirmation is limited in that only three ions are observed, which is the minimum number recommended in the guidelines. In addition, there is not yet a large number of reference chemical ionization spectra with which to assess uniqueness. The specificity inherent to resonance electron capture ionization is itself an important factor, but reliability and ruggedness of the technique have not yet been demonstrated by collaborative studies, for example. We note that B1 decomposes in the usual gas chromatographic techniques and is run by solids probe introduction. Thus, the additional specificity of a gas chromatographic retention time is lost.

Aflatoxin B1

Image  of molecular line structure - contact brumley.william@epa.gov for further information

Image of chromatograph - contact brumley.william@epa.gov for further information

Figure 3. Negative ion chemical ionization mass spectrum of aflatoxin B1, in an extract of peanut butter (about 10.5 ppb); Finnigan 3300 F; CH4 1 Torr; ion source 150 C; direct introduction.

As is generally known, the response from an actual sample may not be exactly like that of a standard, and may exhibit additional ion responses not found in the spectrum of the standard. However, it is left to the discretion of the experienced scientist analyzing a sample whether a number of additional ion responses, for example, would invalidate the confirmation. This situation illustrates the limitations of multiple ion detection where responses at unmonitored ions are not detected and therefore not even considered. It also emphasizes the need to leave some latitude of decision-making by the scientist involved. This latitude for decision-making should be built into the proposed method, for example, in reasonable error limits for observed relative abundances of ions. In this way, the proposed method is not so restrictive as to be impractical.

Another example of the extension of limits of detection and specificity concerns the multiple ion detection gas chromatographic mass spectrometric confirmation and quantitation of TCDD in fish. In this determination, a quadrupole mass spectrometer is confined to scan only twelve ions (m/z 257, 259, 261, 305, 307, 320, 321, 322, 324, 326, 332 and 334). The determination currently makes use of a high resolution fused silica capillary column about 25 m in length. Figure 4 shows the response from an extract of a spiked environmental fish (sucker) sample which was subjected to an extensive extraction-clean-up procedure. This sample was particularly difficult to analyze due to the presence of many environmental contaminants including polychlorinated biphenyls and plasticizers. The sample was quantitated at 40 ppt based on 9.5 g of fish.

Tetrachlorodibenzo-p-dioxin

Image  of molecular line structure - contact brumley.william@epa.gov for further information

Image of chromatograph - contact brumley.william@epa.gov for further information

Figure 4. Electron impact multiple ion detection mass spectrum of TCDD in a fish (sucker) extract (about 40 ppt); Finnigan 3300F; 70ev; unheated source: gas chromatographic introduction.

The presentation of this particular procedure is not an attempt to be dogmatic concerning confirmation of TCDD. There are other gas chromatographic mass spectrometric and extraction-clean-up procedures in use which are undoubtedly acceptable in the hands of the experienced analyst. Some of these procedures use high resolution mass spectrometry. Most recently, we compared the use of low resolution limited mass scanning and collision-induced decomposition of ions on an instrument of high sensitivity and low limit of detection.

The limited mass scan technique (m/z 250-350) was used to confirm TCDD in the gas chromatographic mass spectrometric analysis of a fish (sucker) sample (about 35 ppt), similar to that discussed earlier using a 25 m, fused silica capillary column coated with methyl silicone. As an alternative technique, collision induced decomposition of TCDD was examined. TCDD produces the main beam representing the molecular ion at m/z 320 and major decompositions corresponding to [M - CL]+, [M-COCl.]+, and [M- 2COCl.]+ occurring at m/z 285, 257 and 194. If high sensitivity is required, only these three most intense decompositions are monitored in analogy to multiple ion detection. The appearance of appropriate responses for the decompositions will provide specificity for confirming TCDD in the sample.

Unfortunately, the specificity is not without limitations. A tetrachloromethoxybiphenyl compound also exhibits two of these same decomposition ions and has a similar gas chromatographic retention time. Thus, monitoring the additional [M-CH3.]+ decomposition from the biphenyl compound would be required to indicate the possible interferences. This is analogous to monitoring the m/z 305 ion in the multiple ion detection procedure discussed earlier. Note that the limited scan technique automatically reveals the [M-CH3.]+ response of the potential interferences.

The approach to the analysis of TCDD is characterized by a great diversity. Possible approaches include limited mass scanning, low resolution multiple ion detection, high resolution multiple ion detection, and metastable techniques. Diversity is an essential feature of scientific investigations and stimulates a great deal of development work. Flexibility in guidelines for mass spectrometry protects this diversity of approaches.

Improvements in instrumentation and analytical methodology have pushed detection limits for TCDD to part per trillion and sub part per trillion levels. In some instances, a lower limit of detection has been achieved at the expense of specificity. The full mass scan spectrum of TCDD is a highly specific confirmation. Obviously, full scan spectra are not used at part per trillion levels, but the alternative approaches in use attempt to maintain high specificity. From a regulatory point of view, it is more important to confirm conclusively the presence of a contaminant with slightly higher detection limits than to achieve lower detection limits with inadequate specificity.

Thus, we confront our dilemma. The 'search for zero' requires a push to ultimate detection limits. The requirements of regulatory mass spectrometry demand that this be accomplished while providing an unequivocal identification.

In conclusion, the history of regulatory mass spectrometry at FDA began with instrumentation of limited capabilities. Yet, early in the application of mass spectrometry, such as in the Krebiozen case, the indispensable power of the technique to solve the problem of identification became apparent. The technique has continued with rapid development of its sensitivity, versatility, and applicability. It remains unrivaled by any other single analytical technique in many areas of trace organic analysis identification. As was the ease in the past scientists concerned with regulatory mass spectrometry have felt compelled to be involved in the cutting edge of the technique, helping to advance its capabilities and applications. This research effort is an integral part of their mission to help safeguard the public health by maintaining their integrity as skilled and experienced scientists. Regulatory mass spectrometry is intended to serve the public. It is not instituted to condemn various efforts and viewpoints but to heighten awareness and foster understanding. It is both to such past efforts in protecting the public health and to those of the future that the Joseph N. Damico laboratory is dedicated.

REFERENCES

  1. J. F. Holland, J. Am. Med. Assoc. 200:213 (1967).
  2. J. N. Damico, in Biochemical Applications of Mass Spectrornetry, edited by G. R. Waller, Chapt. 23, pp. 623-653. Wiley, New York (1972).
  3. J. A. Sphon and W. C. Brumley, in Biochemical Applications of Mass Spectrornetry, First Supplementary Volume, ed. by G. R. Waller and O. C. Dermer, Chap. 23, pp. 713-749. Wiley- Interscience, New York (1980).
  4. U.S. Food and Drug Administration Pesticide Analytical Manual, Vol. I., Appendix, Transmittal no, 78-1, 63/08/78.
  5. J. A. Sphon, J. Assoc. Off. Anal. Chem. 61:1247 (1978).
  6. J. A. G. Roach, J. A. Sphon, D. F. Hunt and F. W. Crow, J. Assoc. Off. Anal. Chem. 63:452 (1980).
  7. Official Methods of Analysis of the Association of Official Analytical Chemists, edited by W. Horwitz, 13th Edn, pp. 426-427. Association of Official Analytical Chemists, Washington, DC (1980).
  8. J. Sphon, D. Andrzejewski, W. Brumley, P. Dreifuss and J. Roach, Adv. Mass Spectrom. 8B:1490 (1980).

Trace Organic Analysis Home Page
Analytical Environmental Chemistry
Environmental Sciences | Office of Research Development
National Exposure Research Laboratory
Author: William C. Brumley
email: William C. Brumley


Local Navigation


Jump to main content.