Infrared MALDI

From Murray Mass Spectrometry Group

Jump to: navigation, search

Creating new tools for chemical analysis requires an intimate understanding of the processes involved. Laser ionization mass spectrometry of large biomolecules encompasses solid state energy transfer, supersonic expansion cooling, gas phase ion chemistry, and unimolecular dissociation reactions. A significant aspect of our research is the elucidation of the ionization process and exploitation of new physical and chemical discoveries for improving existing and creating new techniques in analytical and biological mass spectrometry.

Contents

Research

Infrared MALDESI

We have recently devised a new ambient ionization technique that combines infrared laser desorption with electrospray ionization.<ref>Dong, J.; Rezenom, Y. H.; Murray, K. K., Aerosol Desorption Electrospray Ionization. Proc. ASMS Conf. Mass Spectrom. 2007, 55, MP06.</ref><ref>Y. H. Rezenom, J. Dong, K. K. Murray. "Infrared Laser-Assisted Desorption Electrospray Ionization Mass Spectrometry" Analyst 133, 226-232 (2008).</ref>

We call this technique IR LDESI; the matrix-assisted version is called IR MALDESI. This ionization method can be used for direct mass spectrometry analysis of water-containing samples under ambient conditions. A schematic of the instrument setup is shown in in the diagram below. Infrared laser light from a pulsed mid-infrared is directed at the sample target in front of the sampling cone of an ion trap mass spectrometer. The material desorbed and ablated by the IR laser is ionized through interaction with a continuous stream of charged droplets from an electrosprayed solvent. The ions are mass separated in the ion trap instrument.

IR laser used for MALDESI/LDESI with an ion trap mass spectromeer.


A direct analysis of whole blood by IR LDESI is shown below. The blood was obtained from a volunteer using a lancet and the droplet was directly transferred to a stainless steel target for analysis. Two sets of peaks corresponding to distributions of multiply protonated ions are observed with charge distributions ranging from [M+11H]+11 to [M+21]+21. These distributions correspond to the hemoglobin α and β chains.

IR MALDI and LDI Wavelength Dependence

MALDI mass spectrometry is usually performed using a pulsed ultraviolet laser, typically a small nitrogen laser at 337 nm. The strong electronic absorption of small aromatic organic acids is used for many UV matrix materials. Infrared lasers can also be used for MALDI with the vibrational absorption of the matrix facilitating the process. The advantages of IR MALDI are 1) softer (less fragmentation) ionization that can be used for ionization of easily fragmented molecules, 2) greater material removal that can be useful for direct gel or cell and tissue ionization, 3) less low mass interference for small molecule mass spectrometry, 4) compatibility with solvent matrix or matrix-free laser desorption mass spectrometry.

Wavelength dependence of ionization threshold for 4-nitroaniline matrix.


Perhaps the most striking aspect of MALDI in the IR wavelength region is the strong wavelength dependence of the ion signal, mass resolution, and other ionization parameters. This is in stark contrast to UV MALDI where performance is comparable over a large wavelength range. With a tunable IR laser and with proper control over experimental conditions, MALDI wavelength response spectroscopy can be performed. An example of this is shown in the figure above (Sheffer & Murray, 1998). Here, the inverse of the threshold fluence for MALDI of the protein insulin is plotted as a function of IR wavelength. Plotting the threshold fluence in this way facilitates comparison of MALDI performance to IR absorption: higher absorption is expected to lead to lower fluence. The sticks in the figure are a crude representation of the peaks in a published FTIR absorption spectrum of crystalline 4-nitroaniline. The peaks centered on 3.0 µm correspond to amide NH stretching vibrations and the peaks centered on 3.45 µm correspond to CH stretching vibrations. The IR and MALDI response traces are comparable, although there is possibly an extra contribution to MALDI absorption from residual water near 2.9 µm, which may lower the threshold fluence at this wavelength.

Wavelength dependence of the threshold fluence for soft infrared laser desorption ionization of insulin. The solid trace is the IR absorption spectrum of an insulin thin film.


We have extended our wavelength studies to matrix-free soft infrared laser desorption ionization. Over the past several years, we have been developing methods for soft IR LDI; peptides and proteins up to 15 kDa in mass can be ionized without fragmentation and without a matrix (Bhattacharya, et al., 2002; Rousell, et al., 2004). This technique has been demonstrated with direct from gel IR-LDI (Xu, et al., 2004) and has great potential for the analysis of tissue sections without the spatial resolution issues associated with matrix addition (see Section II.C below). The nature of laser energy absorption in soft IR-LDI is an important issue since it has long been assumed that absorption of laser energy directly by the analyte will lead to fragmentation. To better understand this phenomenon, we have measured the threshold laser fluence as a function of wavelength for the soft IR-LDI of insulin. In the figure above, the filled circles are the inverse of the threshold fluence and the solid line is the IR absorption spectrum of a thin film of insulin. The ionization efficiency, as indicated by the inverse threshold fluence, tracks the IR absorption of the insulin film remarkably well. Maxima are at 3.0 µm (OH and NH stretch) and 3.4 µm (CH stretch). Our current hypothesis is that a fraction of the insulin molecules in the film are acting as a “sacrificial matrix�? that is aiding in the ionization of the remaining desorbed molecules.

Two laser IR/UV MALDI

MALDI using two UV laser pulses has been performed in the UV using a single laser and optical delay line or two lasers of the same wavelength (Moskovets & Vertes, 2002; and references therein). The goal of these experiments has been to study the mechanism of UV MALDI, particularly as it relates to excited electronic state processes. The idea is to take the total number of photons required for a single MALDI laser shot and split them into two packets. The first packet is delivered to the target followed in a few ps to ns by the second packet. The ion signal (or potentially some other relevant parameter) is then measured as a function of the time delay between the two packets.

Configuration for L2-MALDI: The IR laser preps the surface for enhanced UV MALDI after an adjustable delay


We have extended the two-laser approach to two wavelengths using infrared and ultraviolet matrix-assisted laser desorption ionization. The hypothesis is that an IR pulse followed by a UV laser pulse will enhance ionization efficiency for the large quantity of material ablated. It has been shown that at least 100 times more material is removed by IR MALDI compared to UV MALDI for crystalline matrices, yet the ion signal obtained is nearly identical (Kampmeier, et al., 1997). Our initial study using 10.6 µm and 337 nm lasers was published in the Journal of Mass Spectrometry (Little, et al., 2003). The results are consistent simple thermal model with the IR laser heating the sample on a microsecond time scale, which heats the matrix and may even cause matrix melting. The heated sample improves the efficiency of the subsequent UV MALDI. The use of a laser pre-pulse to modify the MALDI sample opens new possibilities for MALDI analysis.

Two-laser IR/UV MALDI using a 2.94 µm OPO followed by a 337 nm nitrogen laser after a selectable delay.


Additional experiments have been performed at different wavelength combinations. The figure above shows a result that was recently obtained using a tunable 3 µm mid-IR laser in conjunction with a 337 nm UV laser. Both lasers are adjusted such that their energies are sufficiently low that no ions are produced by either laser alone. The IR laser is fired first and the UV laser second, after an adjustable delay. The insulin signal rises immediately at t=0 (the time jitter is approximately 50 ns) and approaches a maximum near 200 ns. The signal drops again to zero after 500 ns. We interpret these results as transient heating, as with the 10.6 µm results. We are currently expanding these results to different matrix and analyte combinations.

References

<references />

Rousell, D. J.; Dutta, S. M.; Little, M. W.; Murray, K. K., Matrix-free infrared soft laser desorption/ionization Mass Spectrom 2004, 39, 1182.
Laboy, J. L.; Murray, K. K., Characterization of infrared matrix-assisted laser desorption ionization samples by Fourier transform infrared attenuated total reflection spectroscopy. Appl Spectrosc 2004, 58, 451.
Little, M. W.; Kim, J. K.; Murray, K. K., Two-laser infrared and ultraviolet matrix-assisted laser desorption/ionization. J Mass Spectrom 2003, 38, 772.
Rousell, D. J.; Dutta, S. M.; Little, M. W.; Murray, K. K., [http://dx.doi.orS. H. Bhattacharya, T. J. Raiford, and K. K. Murray, “Infrared Laser Desorption/Ionization on Silicon�? Analytical Chemistry 74, 2228 (2002).
J. Kampmeier, K. Dreisewerd, M. Schurenberg, and K. Strupat, “Investigations of 2,5-DHB and succinic acid as matrixes for IR and UV MALDI. Part I: UV and IR laser ablation in the MALDI process�? Int. J. Mass Spectrom. Ion Processes 169/170, 31 (1997).
M. W. Little, J. K. Kim, and K. K. Murray, “Two-laser infrared and ultraviolet matrix-assisted laser desorption/ionization�? J Mass Spectrom 38, 772 (2003).
E. Moskovets and A. Vertes, “Fast dynamics of ionization in ultraviolet matrix-assisted laser desorption ionization of biomolecules�? J. Phys. Chem. B 106, 3301 (2002).
D. J. Rousell, S. M. Dutta, M. W. Little, and K. K. Murray, “Matrix-free Infrared Soft Laser Desorption Ionization,�? J. Mass Spectrom. 39, 1182 (2004).

Presentations

ASMS 2005: The Role of Coarse Particles in Desorption Ionization
ASMS 2005: Wavelength Dependence of Infrared Soft Laser Desorption and Ionization
ASMS 2003: Laser Ablation Particle Ejection from MALDI Matrices
ABRF 2003: Modification of a Commercial Mass Spectrometer for Employment of Infrared Desorption/Ionisation on Silicon

Publications

Rousell, D. J.; Dutta, S. M.; Little, M. W.; Murray, K. K., Matrix-free infrared soft laser desorption/ionization Mass Spectrom 2004, 39, 1182.
Laboy, J. L.; Murray, K. K., Characterization of infrared matrix-assisted laser desorption ionization samples by Fourier transform infrared attenuated total reflection spectroscopy. Appl Spectrosc '2004, 58, 451.
Little, M. W.; Kim, J. K.; Murray, K. K., Two-laser infrared and ultraviolet matrix-assisted laser desorption/ionization. J Mass Spectrom 2003, 38, 772.
Jackson, S. N.; Murray, K. K., Infrared matrix-assisted laser desorption/ionization of polycyclic aromatic hydrocarbons with a sulfolane matrix. Rapid Commun. Mass Spectrom. 2001, 15, 1448.
Sheffer, J. D.; Murray, K. K., Infrared matrix-assisted laser desorption/ionization using a frozen alcohol matrix. J. Mass Spectrom. 2000, 35, (1), 95-97.
Sheffer, J. D.; Murray, K. K., Infrared matrix-assisted laser desorption ionization using OH, NH and CH vibrational absorption. Rapid Commun. Mass Spectrom. 1998, 12, (22), 1685-1690.
Caldwell, K. L.; Murray, K. K., Mid-infrared matrix assisted laser desorption ionization with a water/glycerol matrix. Appl. Surf. Sci. 1998, 127-129, (1-4), 242-247.
Caldwell, K. L.; McGarity, D. R.; Murray, K. K., Matrix-assisted Laser Desorption/Ionization with a Tunable Mid-infrared Optical Parametric Oscillator. Journal of Mass Spectrometry 1997, 32, 1374-1377.

Bibliography

Connotea: MALDI
Connotea: Infrared MALDI
Personal tools