Issues in Explosives Residue Analysis A Primer for the Bar Frederic Whitehurst, Ph.D.
[Editor’s Note: This is a multi-part series deigned to educate the defense bar on important issues concerning explosive and explosive residue investigations]
Part 2: Back to the Basics: Was it the result of an explosive device in the first place? How do we know that?
Part 3: Daubert provides guidance and a means to expose limitations and evaluate explosive investigations, methods, and interpretation
Part 4: The Explosion Crime Scene: Sampling and Homogeneity Issues
Part 5: Disposition Homogeneity in explosive scene investigation
Part 6: Contamination and Cross Contamination in explosive scene investigation
Part 7: Contamination by “Render-Safe” acts of explosives
Part 8: Transportation and storage of evidence in explosive scene investigation
Part 9: Chemical analysis in explosive scene investigation
Part 10: Identifying Techniques in explosive scene investigation
Part 11: Interpretation of data in explosive scene investigation
Part 12: Experience: What makes for a proper expert in explosive scene investigation?
Part 13: Conclusion
At present there are a number of types of instruments that might be described as “identifying” techniques. The description is misleading and should be addressed especially in the area of explosives analysis. The “identifying” techniques include x-ray powder diffractometry, mass spectrometry, nuclear magnetic resonance and Fourier transform infrared spectrometry. Each of these techniques has very powerful analytical capabilities. However the court should not accept the techniques simply because of their complicated sounding names, their high cost or the apparent complexity of the data issuing from printers and plotters attached to the instruments.
At first overwhelming, the data is very simply explained by the expert who has a working knowledge of what the instruments do with the type of materials analyzed. Each of these instruments interacts with explosives in different ways. Following Imwinkelried  we must ask: 1. Are there any imprecisions in the formula the instrument uses to compute the degree of fit of the “identity” with the true analyte? 2. Is there an adequate comparative database to properly evaluate the statistical significance of the data? We can also add a third question: 3. Does the analytical technique itself change the analyte during the analysis?
We can quickly answer the second question. There are no databases from which to draw inferences concerning the identification of materials found in explosives debris which might or might not mimic explosives in various analyses. Each analysis must stand on its own. For instance the crime scene resulting from a bombing in a parking garage under a large skyscraper would most likely be like no other bombing crime scene ever experienced by a trace explosives residue analyst. The explosives ordnance disposal personnel on the scene might see a large hole in a building such as they had seen in the past. However the mixture of materials at the blast scene from a chemist’s point of view would most probably have not been seen before. There would be car exhaust on the walls of the parking garage. There would be car parts, oil, transmission fluid, burning tires, burning electrical components and plastics, broken sewage mains, fire extinguisher material, broken concrete and building materials, dust and dirt of unknown composition, and thousands of other unknown types of materials at that crime scene. And materials would be present that were brought in from outside of the building on car tires. For instance, in some cities, urea salt and calcium chloride are used to melt snow and ice on the roads in the wintertime. Those materials are certainly carried into the parking garages in those cities during the winter months. The finding of calcium ions could indicate the use of calcium nitrate based explosives in a terrorist bomb used in a garage. The detection of urea could indicate the presence of urea nitrate used in a terrorist bomb. However those materials could have other origins as we see. The bombing of an airliner would not necessarily result in the same types of materials being found after the bombing of a skyscraper. Pipe bombs which appear in mail packages give an even different signature.
There is no true database to compare with many of these crime scenes. In Porter  the court stated that “If experts cannot tie their assessment of data to known scientific conclusions, based on research or studies, then there is no comparison for the jury to evaluate and the experts’ testimony is not helpful to the jury.” The reader should note that if a small pipe bomb explodes in a mailbox, that crime scene can be relatively easily duplicated. The explosives residue expert can reconstruct the matrix in which the bomb was placed, fill various pipes with hypothesized energetic mixtures, initiate those bombs, collect the residue, and conduct analyses of the residues with various instruments. In that way she can validate her hypothesis concerning the type of explosive used. 
When a structure such as the World Trade Center is bombed, validation of the hypothesized explosive charge is virtually impossible. Reconstruction of the crime scene and reinitiation of explosives in that building under similar if not exactly the same types of circumstances would be cost prohibitive.
When one asks the third question as to whether the instrument itself might possibly change the analytes found in explosives debris, one must consider the very nature of energetic materials. Any layperson can see that explosives are reactive materials and do not simply wait for detonation to react. The “thermal lability” or sensitivity of some explosives to heat has been a topic of discussion in the scientific community for years. ,  The reader should note that the studies involving decomposition have involved mostly relatively pure samples of explosives. (Again we are reminded of Berger’s  statements concerning the degradation or contamination of DNA samples.) When explosives residues are collected from debris, the residues are on surfaces that have many other types of materials present on them. The residues are collected by mechanical extraction (e.g., scraping or brushing of debris surface) or solvent wash and are therefore mixed with other materials which are on those surfaces. A question must be asked as to whether during analysis the explosives themselves react with other materials present in the matrix. Without knowing what those other materials might be, the answer is open to conjecture.
Some explosives, such as firecracker powders and black powder, are composed of separate oxidizers and fuels and some explosives, such as TNT and nitroglycerine, combine oxidizer and fuel in one material, in one molecule. Again, when analyzed, these materials may very well react with materials around them. For instance, if one injects residues containing explosives into a very hot chamber such as the injection port of a gas chromatograph, they may very well react with accompanying matrix material before they are eluted from the gas chromatographic column into the waiting mass spectrometer.
The gas chromatographic analysis of a multicomponent explosive mixture may result in the destruction of some of the components of the material and leave others resulting in data that leads the analyst to a wrong conclusion concerning the original explosive’s composition.
Before analyzing materials with the x-ray powder diffractometer, the materials are normally ground into a fine powder. If analytes are ground with too much force before being placed on the x-ray powder diffractometer, one may find that crystalline analytes metamorphose through solid phase metathetical reactions, trading ions between species. The analyst will find herself detecting the presence of materials she did not have before she started her analysis scheme.
How does one know if such reactions mentioned above have taken place leading to improper interpretations of data? Daubert spells out the answer in two different areas. “Scientific methodology today is based on generating hypotheses and testing them to see if they can be falsified.”  The initial determination of the presence of materials presents a hypothesis which is followed in most analytical schemes by an “orthogonal” technique, a very different method of analysis. The hypothesis is then tested by this orthogonal technique to see if the hypothesis can be falsified. The modern scientific method requires that we first make observations by collecting data. Residue evidence which is not visible to the naked eye requires that those observations be conducted with sensitive equipment that is capable of giving us our first pieces of data. However, the hypothesis from that data must be validated, tested. Without those tests we are simply presenting the trier of fact a hypothesis, not a tested and validated conclusion/opinion. For instance, if one utilizes the mass spectrometer to determine the presence of invisible residues of explosive “X” then a confirmation is necessary through the use of another technique.
During an International Symposium and Workshop on Explosives Residue Analysis held by the FBI at Quantico, Virginia in 1993, participants agreed that at least two different analytical techniques were necessary before conclusions with significant weight could be drawn from the data. This agreement came from a recognition that explosives are extremely reactive materials which present significant analytical problems with reactivity during analysis. Participants also recognized that confirmation is the normal course of business in valid scientific analyses.
The first question posed by Imwinkelried asks about the formula used by the instrument to arrive at its conclusion. Analytical chemists who have utilized instrumentation combined with modern computer analysis schemes recognize that these tools are just aids and do not give definitive answers. When one depends too heavily on these types of tools the results can be disquieting. The mass spectrometer’s spectral library computer database comparison algorithm may “identify” an analyte differently from analysis to analysis. There can be any number of reasons for this lack of specificity. The reader should remember that the explosives residue analyte does not exist alone but is extracted normally from the surface of debris with a lot of other materials.
Preparation of samples for the Fourier transform infrared spectrometer may result in changes in the spectrum which the computer data base comparison interprets to be different materials on two different analyses.
The material analyzed may also not be in the databases of these instruments and the computer is then forced to choose the “best” choice among those available.
Counsel should look very closely at this data and ask specifically why the expert arrived at a particular conclusion. Counsel may find that the “expert” has no scientific basis for opinions and has just accepted the results of the computer analysis blindly. Or when offered a set of possible choices by the algorithm, the analyst may choose that answer which best “fits the scenario.” If counsel lets that reasoning go unquestioned, his client has hired the wrong advocate. The computer analysis algorithms in analytical tools such as mass spectrometry, x-ray powder diffractometry and Fourier transform infrared spectrometry are extremely complex and require substantial amounts of time to understand.
Counsel may again find that “Plaintiff argues that the net effect of this scrutiny of his expert evidence is to put the claim beyond his financial ability to pursue.”  However Rule 702 and Daubert require these expenditures if they are needed to prove the reliability of the scientific techniques used. The second area where Daubert gives direction is in the quote “Ordinarily, a key question to be answered in determining whether a theory or technique is scientific knowledge that will assist the trier of fact will be whether it can be (and has been) tested.”  Here Daubert teaches us that we must test our theories/hypotheses in order for them to be scientific knowledge that will assist the trier of fact.
The “orthogonal” method test then is appropriately required by the court and should be looked for when analytical data and interpretation of that data is presented. As examples of orthogonal testing of hypotheses one may follow a mass spectrometric analysis with a Fourier transform infrared analysis. The mass spectrometric analysis subjects the analytes to the severe environments of high heat and extremely low pressures in which they do not at times survive intact and may react with other analytes present in the sample. The analyst may innocently inject material “A” into a mass spectrometer, and before it reaches the mass spectrometric analysis unit, it has become “B” by decomposition or chemical reaction with components of the matrix that was injected with it. The analyst then interprets the data to mean that “B” is what he injected. By following this analysis with the infrared analysis, the severe analytical conditions are avoided, a very different kind of “interrogation” of the specimen is conducted, and the required need to test the hypothesis generated by data from the mass spectrometer is satisfied.
An explosive which exemplifies this situation is ammonium picrate. This explosive, when subjected to the mass spectrometer’s conditions, results in data consistent with the presence of picric acid.  And yet when the hypothesis of picric acid is presented to the Fourier transform infrared spectrometer, there is disagreement. The Fourier transform infrared spectrometer gives results consistent with the presence of ammonium picrate. However with the two instruments seemingly contradicting each other, further testing must be conducted. If that testing is conducted with either x-ray powder diffractometry or nuclear magnetic resonance spectrometry, one begins to realize that the ammonium picrate “fell apart” in the severe conditions of the mass spectrometer resulting in the production of picric acid and ammonium. The significance of this is great when one realizes that picric acid is a relatively sensitive explosive and ammonium picrate is extremely insensitive. Were one trying to find a source of picric acid as opposed to ammonium picrate, the search would lead to very different conclusions for the investigator and therefore the trier of fact. One might suggest that the Fourier transform infrared spectrometer should be utilized by itself if it is superior to the mass spectrometer in its analytical conditions. The infrared spectrometer however can lead the analyst to believe that material is composed of only those components that the spectrometer can detect. This technique suffers also from the failing that components present in large amounts can give signals which overwhelm minor components. If one is asking whether two materials differ, significant differences might be found in minor components which the infrared technique can miss. We are again led to the conclusion that following the modern scientific method is most appropriate and necessary in order that the trier of fact not be misled. The modern scientific method requires validation of hypotheses.
Another example of the need for confirmation of results before conclusions can be drawn is found with x-ray powder diffractometry. This technique distinguishes materials based upon the orientation of their respective atoms in space. By bombarding solid crystalline materials with x-rays and detecting how those x-rays are reflected from the materials, comparisons can be made between this data and libraries of knowns. The x-ray diffractometer theoretically, in an ideal situation, can absolutely identify materials. However, as in any other instrumental technique, factors ranging from sample preparation to the complexity of multicomponent samples lead the x-ray diffractometer to venture its best “guess” concerning the identity of the materials. Dr. Ron Jenkins, of the International Center for Diffraction Data, advises that he would not make a call with x-ray powder diffractometry unless that call was confirmed with an elemental analysis using energy dispersive x-ray analysis or x-ray fluorescence spectrometry . Jenkins is simply doing his science well, recognizing that “Scientific methodology today is based on generating hypotheses and testing them to see if they can be falsified.”  The x-ray powder diffractometer can be as fallible as the mass spectrometer. Quite often the forensic explosives analyst is tasked with determining the components of a highly complex sample. In order to separate the components of the sample, solvent extractions can be used. The analyst might place portions of the sample in water to remove all the water soluble materials. This technique is followed by the FBI and others in explosives residue analysis protocols. Following aqueous extraction, the water is evaporated off in order that crystalline material may be submitted to the x-ray powder diffractometer. This extraction technique immediately leads to indecision in the interpretation of the data. Materials which are removed from evidence in this matter may be involved in metathetical reactions. These reactions are best exemplified by imagining residue containing potassium chloride salt, which is found prevalently in nature, and ammonium nitrate, a component of explosives and fertilizers. When placed in water, the potassium, chloride, ammonium and nitrate ions separate, pulled apart by the action of the water about them. When the water evaporates, those separate components can recombine as they originally existed as well as “change partners” and become potassium nitrate and ammonium chloride. Without understanding this, the examiner could very well interpret the data as meaning that the original explosive was a potassium nitrate-based explosive as opposed to an ammonium nitrate-based explosive. 
As we have seen above, the method of handling evidence during analysis, the “protocol/technique”, is extremely important in understanding the data that is generated. “Rule 702’s ‘helpfulness’ standard requires a valid scientific connection to the pertinent inquiry as a precondition to admissibility.”  A mass spectrometer or Fourier transform infrared spectrometer may have been proven again and again to be the appropriate tool to use in a particular situation. However that same tool may not be appropriate for the samples considered in the residue analysis. Stout  represents just such a situation. Though the ruling in the case depends on Frye  which has been superseded by Daubert the case is still on point. In Stout an expert presented evidence based upon neutron activation analysis of blood. Neutron activation analysis had been proven to be a useful tool in the analysis of other types of forensic evidence. However the expert was the only individual who had any experience with the method’s use on blood, and he was not prepared to show its proven usefulness in the analysis of blood. The court refused to accept it. Daubert states that “scientific validity for one purpose is not necessarily scientific validity for other, unrelated purposes.”  Determining that “appropriateness” and therefore the “valid scientific connection to the pertinent inquiry” requires confirmation and validation of protocols. With an understanding that one needs validated protocols, an understanding that no one analytical technique “identifies” materials, and the “scientific knowledge” standard of evidentiary reliability, the court has the tools to properly review the scientific data and inferences from the data presented by the forensic explosives residue expert. The attorney reviewing analytical data should then ask if all analytical conditions are described, if results from one analytical technique have been validated by at least another technique and should review carefully the protocol used by the examiner to acquire the data presented.
 Executive Director, Forensic Justice Project, Washington, D.C., B.S. Chemistry, 1974, East Carolina University, Ph.D. in Chemistry, 1980, Duke University, J.D., 1996, Georgetown University School of Law. (202)342-6980.
 Porter v. Whitehall Laboratories, Inc., 9 F.2d 607 614(7th Cir. 1993).
 Jehuda Yinon, Federal Bureau Of Investigation, U.S. Department Of Justice, Analysis of Explosives By LC/MS, Proceedings of The International Symposium on The Analysis And Detection Of Explosives 227 (1983).
 Yinon and Zitrin, supra note 11, at 134, note that the explosive tetryl decomposes during its gas chromatographic analysis and gives mass spectral data corresponding to that of N-methylpicramide. They hypothesize that this is the result of hydrolysis of tetryl, possibly at the injector of the gas chromatograph.
 Berger, supra note 5.
 Stanczyk v. Black & Decker, 836 F.Supp 565 568 (1993).
 Daubert, supra note 17 at 2796.
 This data is from the author’s own experience with this explosive material
 Ron Jenkins taught a class in x-ray powder diffractometry at the FBI Laboratory in 1989 in which he made this statement to the class.
 Daubert, supra note 17 at 2796.
 The author has demonstrated this phenomenon, which has been known for many years by scientists in the field, by experiment in the FBI Laboratory.
 Daubert, at 2796.
 State v. Stout, 478 S.W.2d 368 (1972).
 Frye v. United States, 293 F. 1013 (1923).
 Daubert, supra note 17 at 2796.