JEANETTE ADAMS gets excited talking about all the things she can do with mass spectrometry at the Library of Congress. A scientist in its Preservation Research & Testing Division, Adams uses mass spectrometry to look for early signs of degradation in documents, photographs, and microfilm from the library's collections. In most cases she does so without even taking a sample, thanks to an ionization method called DART (direct analysis in real time).
DART is just one of a new generation of ambient ionization methods. Others include DESI (desorption electrospray ionization) and ASAP (atmospheric solids analysis probe). These new methods are expanding the already vast array of mass spectrometry applications.
"We have to think more broadly because now we're able to analyze the real world," says Gary J. Van Berkel, mass spectrometry group leader at Oak Ridge National Laboratory. "What are we going to look at?"
Mass spectrometrists are taking that question seriously and using these and related ionization techniques for a variety of applications, ranging from pharmaceutical quality assessment to forensic analysis to cultural heritage conservation.
Ionization is a necessary first step for any mass spectrometric analysis because the mass analyzer separates charged particles on the basis of their mass-to-charge ratio. Ionization has traditionally been done within the spectrometer's vacuum system. This newest crop of methods moves ionization into the open.
In DESI (invented by R. Graham Cooks and coworkers at Purdue University), a stream of charged solvent droplets bombards the sample surface, desorbing molecules into a stream of small droplets that goes into the mass spectrometer. (Science 2004, 306, 471)
DART and related ionization techniques are relatives of atmospheric pressure techniques such as atmospheric pressure chemical ionization (APCI) and atmospheric pressure photoionization (APPI).
In DART (invented by Robert B. Cody at instrument maker JEOL and James A. Laram??e at government contractor EAI Corp.), a glow-discharge plasma excites an inert gas that subsequently ionizes sample that's remote from the plasma (Anal. Chem. 2005, 77, 2297). In ASAP (invented by Charles N. McEwen and coworkers at DuPont), the sample is inserted directly into an APCI source and exposed to a corona discharge carried out in heated nitrogen. PADI (plasma-assisted desorption ionization, invented by scientists at the University of Nottingham and Hiden Analytical in England) uses a nonthermal plasma generated by a radio frequency (Anal. Chem. 2007, 79, 6094).
In each of these methods a plasma ionizes sample molecules (or molecules in the atmosphere that then proceed to ionize the sample). The main differences are in how the plasma is generated and where the sample is introduced.
Small companies are already commercializing DART and DESI. IonSense, located in Saugus, Mass., was founded to develop DART. Indianapolis-based Prosolia is commercializing DESI. For both companies, the challenge has been engineering ionization sources to fit on mass spectrometers from different manufacturers.
These ionization methods—whether commercial or homemade—are finding plenty of applications.
For instance, ambient ionization methods are helping Facundo Fernandez identify counterfeit pharmaceuticals. Fernandez, an assistant professor of chemistry at Georgia Institute of Technology, collaborates with scientists at the Centers for Disease Control & Prevention and the Wellcome Trust to investigate antimalarial medications collected in Southeast Asia.
Fernandez had already been working on the problem of counterfeit drugs for about two years when he first became aware of DART and DESI. He recognized immediately the potential of these methods to speed up his analyses. Defending against counterfeits requires analyzing many more tablets than does conventional quality control, Fernandez says. "We find a huge variety of counterfeits, even from the same blister pack. You really need to test as many samples as you can," he says. The minimal sample preparation required for ambient ionization is key to shortening the analysis time.
Now, Fernandez routinely uses both DART and DESI for their different advantages. The DART spectra are usually simpler than the DESI spectra, he says. But with DESI, by adding reactants to the spray, one can do chemistry to enhance ionization, improve sensitivity, and prevent fragmentation of the antimalarials.
WITH THE AMOUNT of counterfeit drugs in the market, Fernandez would like to see ambient ionization methods move into the field, where they could be used by the people on the ground to monitor the integrity of their drug supplies. Last year, Fernandez analyzed the medications used to treat a patient who died of malaria. The entire supply of antimalarials at the field hospital turned out to be fake.
In addition to quality analysis, ambient ionization methods are also finding use in drug discovery applications. Christopher Petucci, a mass spectrometrist at Wyeth Research in Collegeville, Pa., has used both DART and ASAP. He points out that DART's special ion source increases its cost relative to ASAP, which requires only minor modifications to an existing APCI source.
However, DART and ASAP are "similar in their ability to directly ionize a wide range of small molecules in our lab, from nonpolar compounds like naphthalenes to highly polar compounds such as warfarin," he says. He has used both techniques to monitor simple organic reactions without chromatographic separation (Anal. Chem. 2007, 79, 5064). He currently uses ASAP to quickly ionize nonpolar, high-melting-point compounds that do not ionize by electrospray or APCI.
Besides pharmaceutical applications, direct ionization methods can also be used to analyze solid biological samples. Renato Zenobi, a chemistry professor at the Swiss Federal Institute of Technology, Zurich, and coworkers use uncharged streams of nitrogen gas to sample biological surfaces for analysis by a method called extractive electrospray, which is turning out to be closely related to DESI. Zenobi believes, however, that the neutral nitrogen stream is much gentler than the high-voltage electrospray stream used in DESI.
Using neutral desorption, Zenobi and coworkers have analyzed a variety of biological samples (Angew. Chem. Int. Ed., DOI: 10.1002/anie.200702200). For example, they obtained metabolic fingerprints that included nicotine and caffeine from the skin of a male smoker before and after coffee consumption.
FOOD SAFETY ANALYSIS will be a key application, Zenobi believes. Using neutral-desorption sampling, he and his coworkers detected biogenic amines typically associated with meat spoilage, including histamine, putrescine, and cadaverine, in fish exposed to room temperature for one to two days. They also detected Escherichia coli bacterial contamination in iceberg lettuce.
DESI is also being developed as a method for mass spectrometric imaging, the use of mass spectral data to visualize a sample. Such imaging is usually done with matrix-assisted laser desorption/ionization (MALDI) or secondary ion mass spectrometry. Cooks's group, in collaboration with Richard M. Caprioli at Vanderbilt University, has demonstrated DESI images of phospholipid profiles in a variety of tissues (Angew. Chem. Int. Ed. 2005, 44, 7094). Prosolia is developing a commercial prototype for DESI imaging.
Van Berkel's group likewise is developing DESI for mass spectrometric imaging, particularly as a replacement for whole-body autoradiography. His team is using DESI to visualize drug distribution in mice that have been dosed with a drug. DESI gives information that is molecularly specific and spatially resolved about both the parent drugs and their metabolites without the need for radioactive tags.
But the sensitivity isn't good enough yet for general tissue-imaging applications, Van Berkel says. "It may be easy to find the drug when administered at a physiological dose, but the metabolites are at lower levels. We find it much more difficult to get good pictures of the metabolites," he says. "Detection levels have to improve by at least a couple of orders of magnitude."
The operating room is where Zoltán Takáts would like to see DESI. Takáts, a mass spectrometrist at Semmelweis University in Budapest, Hungary, is one of the coinventors of DESI. He aims to bring DESI and a related technique called JeDI (jet desorption ionization) out of the lab and into clinical settings. DESI can be used to analyze tissue surfaces, either on the skin or inside an incision. JeDI, a destructive technique in which a liquid jet erodes the surface, can be used simultaneously to cut tissue and ionize biomolecules.
Both methods can be used to perform real-time histological identification of tissues. "The most important application is the in situ identification of cancerous tissue, which will help surgeons to localize tumors, increase tumor-removal efficiency, and minimize the amount of healthy tissue removed," Takáts says. He and his colleagues are currently developing a device for electrosurgery with DESI and JeDI, which has already been tested in mouse and rat models. "It's capable of identifying even traces of infiltrating tumors in healthy tissue," he says.
ART CONSERVATION is another area where DART is making a difference. The Library of Congress' Adams is exploring the use of DART to diagnose degradation of materials in its collection.
Many materials could benefit from nondestructive, sensitive detection of marker compounds, such as the acetic acid formed when cellulose acetate films degrade. Called "vinegar syndrome," the reaction is recognizable by the telltale odor. By the time the acetic acid concentration is high enough to smell, further deterioration will occur rapidly. "You want to be able to detect it at lower levels than your nose can," Adams says.
Adams is also using DART to analyze paper. DART is sensitive enough that even when samples must be removed, they can be tiny. "I can tweeze a sample of paper that's the size of the dot on an i in a 12-point Times New Roman font," she says. "I can get chemical fingerprints from something that small."
One of the things Adams is doing is revisiting the specifications for materials used to store the Library's collection. "I'm reanalyzing the materials by DART to ensure that we're seeing everything we can see." For example, she analyzed protective sleeves made of Mylar polyester—which usually meet the library's specifications for storing collection items—that were made in DuPont's factory in China. She detected erucamide, a long-chain fatty acid amide used as a slip agent to keep plastic sheets from sticking together. "That's stuff you don't want to have sitting on your collection item," Adams says. "DART picked it up. The same samples looked great" by Fourier transform infrared spectroscopy. It turned out that the erucamide came from materials the Mylar sleeves were shipped in rather than the Mylar itself.
Adams collaborates with conservators such as Adrienne Lundgren, who works with photographs in the library's collection, to help determine how objects should be conserved. They would like to know what materials were used in creating particular photographs, both to understand the artists' methods and to make informed treatment decisions.
For example, platinum prints were coated with a variety of organic substances, Lundgren says. The type of coating dictates the proper treatment. "You don't want to introduce any sort of treatment that will in some way change the artistic integrity of the item," she says. "DART will be helpful because we could determine the coating on each of these prints."
Using DART, they also hope to learn more about the characteristics of photographic papers and dyes from different periods. Such information could help with dating photographs. For example, the use of optical brightening agents is a hallmark of photographic paper made after the 1950s, Lundgren says. Other manufacturing changes have not been well-documented, and much of that proprietary information is being lost as companies go out of business, Lundgren says. "It's important for us as connoisseurs of photography and as conservators who treat photography to understand those changes and when they occurred."
DART can even help conservators preserve digital prints. The paper for digital prints contains a "receiving layer" that helps the inks adhere. "If you use a paper from one company and an ink from another company, the print may be more likely to fade," Lundgren says. DART can help them characterize these proprietary materials.
In addition to their use as stand-alone techniques, these ionization methods could make it easier to interface mass spectrometry with planar separation techniques such as thin-layer chromatography. TLC separations can be monitored by simply placing the plate in front of the source. "You have an analyte ready to be analyzed on a surface," Van Berkel says.
Although these ionization methods are already being used for a variety of applications, scientists still want to understand the mechanisms better.
"You always want to know the basics: for controlling it better, for optimizing it, for pushing up the ion yield, for understanding which compounds get ionized and which are discriminated against," Zenobi says. The fundamentals could have practical implications. "It could be that there's a combination of experimental conditions that is a total sweet spot and no one has found it yet for some reason."
"It's very easy to rush and just work on the applications," Fernandez says. "We shouldn't lose sight of the fundamentals." For example, he says, knowledge of the energy transfer during ionization could lead to modifications in the sources that can ionize labile compounds without fragmentation or to improvements in ionization efficiency that enable analysis of low-concentration metabolites.
Information about the fundamentals is guiding technology development of DESI sources, according to Justin Wiseman, a scientist at Prosolia and one of the inventors of DESI. "We see a large need to understand better the chemistry that can happen at the surface to introduce new applications" of DESI, he says.
"Ultimately, our industrial customers want a tool to solve a particular problem and that has been our focus at Prosolia," says Kevin Boscacci, president of Prosolia. "However, understanding the fundamental aspects of the technique allows us to deliver the best product to meet this goal."
THE UNDERSTANDING of the DESI mechanism has improved greatly since the first report three years ago. At that time, Cooks and his coworkers proposed that the creation of analyte-containing charged droplets was a coulombic process. As they continue to study the DESI mechanism, they are finding instead that the process is almost purely physical.
Simulations show that desorption from the surface is actually based on momentum transfer, Cooks says. "The small solvent droplet hits a thin film on the sample surface, splashes against it, and creates a whole crew of progeny droplets," he says. At some point during this process, the charge is transferred from the droplet to the analyte.
"The DESI process is really the transfer of the compound on the surface into the liquid film and then its transport" in the tiny progeny droplets, Cooks adds. The film can even be a layer of the solvent stream itself. DESI, therefore, is turning out to be a micro- or even nanoextraction experiment, Cooks says.
The selectivity of DESI for particular sample components can be adjusted by using a different solvent or by adding new reagents to the original one, which is typically a mixture of methanol and water. For example, adding hydroxylamine to the solvent spray improved the signal-to-noise ratio of the mass spectra of anabolic steroids tenfold, Cooks says.
The ionization mechanisms for DART and ASAP are less well-understood, but McEwen suspects they use basically the same gas-phase ion-molecule chemistry common to APCI without the ion suppression caused by the use of solvents in the latter technique. McEwen has been able to ionize the same kinds of samples with ASAP as have been reported for DART and has produced nearly identical mass spectra. "You get the same results, so it has to be pretty close to the same analyte ionization mechanism," he says.
Cody uses cholesterol to demonstrate how ionization conditions affect the ions that are formed. "If you ionize cholesterol by DART with the conditions we've been using most of the time, you get M+H minus water, so it looks like cholestadiene," he says. In contrast, using fluorobenzene as a dopant in the plasma yields an M+• ion. "Now we have a way to form odd-electron ions," Cody says. Odd-electron ions result in spectra that can be compared directly with those in mass spectral libraries, which are based on electron ionization, another method that forms odd-electron ions. In addition, odd-electron and even-electron ions fragment differently, thereby providing different structural information about the analyte.
Nevertheless, the methods certainly aren't perfect. Van Berkel believes a more complete understanding of interdependent factors like ionization selectivity, surface chemistry, and matrix effects will be required before these ionization methods are accepted for sensitive applications such as homeland security and food safety. An insufficient understanding of these factors can lead to two types of incorrect answers: false positives, in which something is identified that is not really present, and false negatives. Although false positives can be inconvenient and can slow down screening, the selectivity of mass spectrometry or the use of tandem MS can eliminate them as a concern.
"The real problem is with false negatives. You let a bomb through. The food is poisoned, and you don't know it," Van Berkel says. "More fundamental understanding and experience will be required to judge these techniques in terms of false-negative rate for these critical one-off, quick-look-in-the-field analyses," he says.
The next step for an open-air ionization method is taking it into the field. "Putting anything in the field is a big challenge, but I don't think we're very far actually," Fernandez says. He and Zenobi both want to see these ionization sources on portable mass spectrometers. Fernandez suggests that they could be used with ion-mobility spectrometers, as well, eliminating the need for a vacuum system.
Developing a DESI miniature mass spectrometer is a priority for his lab for the next two years, Cooks says. His group recently published its first paper in the area, describing how they coupled DESI to a portable mass spectrometer (Anal. Chem., DOI: 10.1021/ac071114x).
Who knows what applications lie in the future for these ionization methods? The world is wide open. "We may have completely different uses for these techniques than the traditional analytical analyses we're doing now," Van Berkel says. "We're all doing the easy things. There will be whole new applications that we haven't even thought of."