Forensics and LIBS

We recently talked to Jose Almirall, PhD, professor in the Department of Chemistry and Biochemistry and the director of the International Forensic Research Institute at Florida International University (Miami, Florida) about the application of laser-induced breakdown spectroscopy (LIBS) to forensic applications.

What makes LIBS well suited for forensics?

JA: Laser-induced breakdown spectroscopy (LIBS) does not require much (if any) sample preparation, can analyze solids directly, is fast, provides simultaneous multielemental information, is less expensive to acquire and to maintain than the competing techniques, and can be miniaturized or packaged to the point where it can be used in the field, so that measurements can be taken close to the sampling site. There are now several commercial offerings in LIBS instrumentation for the laboratory and even for the field.

What types of elements or materials can LIBS detect?

JA: LIBS is particularly well suited for the analysis of glass, which is often encountered in hit-and-run accidents and can aid in associating a vehicle to an accident event and can even place a driver inside the car after the accident, even if the driver has fled the scene of the accident. Our research group has developed methods for the analysis of other forensic materials such as paint, soil, cotton, paper, and ink on paper. We can detect many elements of interest at concentrations of 10–50 μg/g in the solid, and we have used a variety of spot sizes, in the order of 50–150 μm.

What is the current state of the application of LIBS to forensics?

JA: LIBS is moving into a position to compete with other, widely used elemental analysis methods in forensic analysis. Of the approximately 380 forensic laboratories in the US, ~100 maintain trace-evidence sections that analyze materials such as glass, fibers, paint, and soil. At the moment, the most widely used technique for elemental analysis is micro-X-ray fluorescence (μXRF), followed by laser ablation–inductively coupled plasma–mass spectrometry (LA-ICP-MS). LIBS offers advantages over these two methods and produces similar probative information. There are only three forensic labs with LIBS at the moment, but the number of labs adopting LIBS is growing. Adoption of new technology is slow in forensic labs even when there are advantages. Several research groups around the world have made important contributions to the analytical chemistry literature that are making a convincing case for LIBS as a viable analytical method for forensic application.

How does LIBS's performance in forensic applications compare to other methods, such as LA-ICP-MS or μXRF?

JA: It depends on the application. For discriminating among different potential sources of glass, for example, the performance is very similar, so we can use any of the three methods equally well. If we use LIBS to analyze glass from a crime scene, and compare it to glass from a known source, we can determine with a fair amount of confidence that the glass samples share a common manufacturing origin — that they were made in the same plant within some time of manufacture.

We've also used LIBS for the analysis of soil, and it works very well for that application also. It provides very good performance for comparing soil samples and associating soil samples of similar origin. Its performance here is very similar to that of LA-ICP-MS.

We are also using LIBS for determining the sources of paper, and of ink that has been deposited on paper. Because LIBS is a laser microanalysis tool, we can analyze the composition of the ink right off the paper. We can do that with LA-ICP-MS, too, of course. We see a little bit better discrimination with laser ablation, but the performance is still very good with LIBS.

What about limits of detection?

JA: LA-ICP-MS can detect some substances, such as metals, at lower concentrations than LIBS can. But for many applications, you don't need the sensitivity that LA-ICP-MS can afford. For example, LIBS performs very similarly to LA-ICP-MS for discriminating between two different pieces of glass or different sources of soil. LIBS is sufficiently sensitive to do that kind of test.

How do the costs compare?

JA: It would take about $500,000 to equip a laboratory to do LA-ICP-MS, because you need not only the instrumentation, but also a special room and very experienced users. Very few forensic labs can afford that. Commercial LIBS systems costing around $90,000–100,000 are currently available, making LIBS much more affordable for the typical forensic laboratory.

Can you tell us about your study comparing the analysis of automobile glass at various laboratories using LIBS, LA-ICP-MS, and μXRF?

JA: The National Institute of Justice has funded a working group to investigate how operational laboratories perform in the analysis of glass. A lot of the work that has been done to date on the analysis of glass with LIBS has been done by academic laboratories. So with this study, we are examining two questions: First, can LIBS and other elemental analysis methods for glass analysis be conducted by forensic scientists in a typical crime lab environment that conducts casework, rather than by academic researchers? Second, can we conduct these analyses using commercial equipment rather than research-grade laboratory-built instruments?

We are using glass standards that we buy from NIST, of known concentration and from known manufacturing facilities. We send them out as blind samples and ask the laboratories to identify them using commercial instruments. Then we compare the level of agreement between the laboratories in terms of the performance. We are doing round-robin studies, using different kinds of glass and different experimental designs to test different things about the methods.

We have already conducted five round robin trials, and it's going extremely well. We have very good agreement between the labs in terms of the results. We expect to complete the study and publish the results by the end of the year.

This project is also giving us a chance to compare the performance of the different methods, and to begin standardizing the methods to do these analyses. Until now, there were no standard methods for this application for LA-ICP-MS, LIBS, or μXRF. So we are also working on establishing standards. For example, if we consider match criteria, when do we say we have a match versus not? Likewise, we need to standardize operating criteria.

Recently, you used LIBS to analyze glass from an attempted bank robbery. Can you tell us about that?

JA: The case was an attempted robbery of a bank in Maryland and the suspect had fired a gun through the glass door of the bank, causing glass to spray onto the suspect. When two suspects were captured by police, one of them had glass on his clothing, suggesting he was the one that fired the gun. The glass was analyzed and compared to the glass from the bank door and was found to be indistinguishable; therefore we associated the person with the glass on his clothing with the broken glass on the door.

What is the process for the courts to accept evidence produced by a new analytical method like LIBS?

JA: This process takes time. There have not been any court cases yet where evidence produced by LIBS has been presented through the entire process. The bank robbery case did not go to court, because the defendant pled guilty.

Before evidence using any new analytical technique can be accepted by the courts, the method has to go through the normal process of scrutiny by the scientific community. That starts when researchers write scientific publications showing the validity of the method. Then there is further scrutiny, and you have to show that you can duplicate the methods and standardize the analysis — that you will get similar results no matter what laboratory does the analysis. It's the same for any method. The same process was followed for accepting DNA analysis in the courts. It was necessary to lay the scientific underpinnings showing that DNA analysis works, then show that people could apply it correctly.

Three forensic labs have just acquired LIBS systems. Very soon they will be validating those techniques and using them for analysis. So we may start to see LIBS analyses presented in the courts in a couple of years.