Applied Surface Science 255 (2009) 6297–6302

UV laser irradiation of IR laser generated particles ablated from nitrobenzylalcohol

Xing Fan, Kermit K. Murray *

Department of Chemistry, Louisiana State University, Baton Rouge, LA 70803, United States

A R T I C L E I N F O

Article history:

Received 25 August 2008

Received in revised form 26 January 2009

Accepted 3 February 2009

Available online 12 February 2009

Keywords:

Laser

Desorption

Infrared

Ablation

Aerosol

Particle

A B S T R A C T

Particles generated by 2.94 mm pulsed IR laser ablation of liquid 3-nitrobenzyl alcohol were irradiated

with a 351 nm UV laser 3.5 mm above and parallel to the sample target. The size and concentration of the

ablated particles were measured with a light scattering particle sizer. The application of the UV laser

resulted in a reduction in the average particle size by one-half and an increase in the total particle

concentration by a factor of nine. The optimum delay between the IR and UV lasers was between 16 and

26 ms and was dependent on the fluence of the IR laser: higher fluence led to a more rapid appearance of

particulate. The ejection velocity of the particle plume, as determined by the delay time corresponding to

the maximum two-laser particle concentration signal, was 130 m/s at 1600 J/m2 IR laser fluence and

increased to 220 m/s at 2700 J/m2. The emission of particles extended for several ms. The observations

are consistent with a rapid phase change and emission of particulate, followed by an extended emission

of particles ablated from the target surface.

� 2009 Elsevier B.V. All rights reserved.

Contents lists available at ScienceDirect

Applied Surface Science

journa l homepage: www.e lsev ier .com/ locate /apsusc

1. Introduction

It has recently been shown that large quantities of material areablated from solid and liquid samples under the conditions usedfor matrix-assisted laser desorption ionization (MALDI). Thetypical wavelength used for MALDI is 337 nm, obtained from anitrogen laser. With this ultraviolet wavelength, particles rangingfrom nanometer [1,2] to micrometer [3] in size are produced fromcrystalline MALDI matrix materials at laser fluences on the order ofseveral hundred J/m2, a value similar to that required for ionformation [4]. The quantity of material removed as particulateunder these conditions is comparable to the total quantity ofmaterial depleted from the sample. Infrared lasers are well-knownfor removing large quantities of material when used for MALDIanalysis [5,6]. Due to the lower absorbance of the IR laser radiation,the fluence required for IR laser material removal and ionformation is approximately ten times higher than the UV.Furthermore, the quantity of material removed as particulate isstrongly dependent on the IR wavelength used [7–9].

The creation of particulate by laser ablation is generally notuseful for ionization because the particles carry away material that

* Corresponding author at: Department of Chemistry, Louisiana State University,

331 Choppin Hall, Baton Rouge, LA 70803, United States. Tel.: +1 225 578 3417; fax:

+1 225 578 3458.

E-mail address: kkmurray@lsu.edu (K.K. Murray).

0169-4332/$ – see front matter � 2009 Elsevier B.V. All rights reserved.

doi:10.1016/j.apsusc.2009.02.005

might otherwise be converted to ions. However, there are severalapproaches that can be used to generate ions from the ejectedparticulate. Ablated particles can be directed into a plasma, but thisleads to the production of atomic ions only [10,11]. Particles andassociated material directed into an electrospray form ions throughtheir interaction with the charged solvent droplets. Such particlescan be injected directly into the spray [12] or produced by theablation of material from a solid sample using a UV [13–15] or an IRlaser [16–18]. Ions can also be formed by irradiating the ablatedparticles with a second laser; particles that contain both matrix andanalyte can form ions by an aerosol MALDI mechanism [19]. Ions canbe formed from particles laser ablated from a surface using an IR [20]or UV laser [21] directed perpendicular to the surface.

Laser ablation of particles in the plume can also be used tocreate particles more amenable to ion formation for inductivelycoupled plasma mass spectrometry. In a recent study, a 266 nmdual-pulse Nd: YAG laser was used to ablate material from asample and, after an adjustable delay, the second pulse wasdirected along the same path as the first beam to irradiate theplume of expanding material [22]. Shadowgraph images show thatthe double-pulse plume irradiation reduces the size of the ablatedparticulate. Smaller particles led to an improvement in intensity aswell as the precision in the detection of copper and zinc from abrass sample.

In this work, we have investigated the effect of a UV post-ablation laser on the size and concentration of IR laser generatedparticles. A 3 mm mid-IR laser system was used to ablate particles

Fig. 1. Particle size distribution for IR laser ablation and UV post-irradiation of 3-

nitrobenzyl alcohol at UV laser delay times of (a) 0 ms, (b) 18 ms, (c) 20 ms, (d)

28 ms, (e) 100 ms, (f) 1000 ms and (g) 15 ms. The IR and UV fluences were 1900 and

2400 J/m2, respectively.

X. Fan, K.K. Murray / Applied Surface Science 255 (2009) 6297–63026298

from a thin film of the liquid MALDI matrix 3-nitrobenzyl alcohol(NBA) [23] at atmospheric pressure. A 351 nm excimer laserdirected parallel to and a few millimeters above the target surfacewas used to irradiate the plume of desorbed material. The particleconcentration, mean particle size, and particle size distributionswere recorded as a function of the time delay between the IR andUV laser pulses. The mean initial velocities of the ejected particlesfrom sample target were obtained from the time delay plots.

2. Experimental

The particle ablation instrument has been described previouslyand consists of an atmospheric pressure stainless steel ablationchamber coupled to an aerodynamic particle sizer (Model 3321,TSI, Shoreview, MN) [3,7]. The chamber is a six-way cross with3.5 cm inner diameter and a volume of 240 cm [3]; the particlesizer inlet is 1.5 cm inner diameter. Under the conditions used inthis work, the inertial and sedimentation effects are negligible forparticles smaller than 10 mm in diameter and will not lead to asize-dependent sampling efficiency [24]. A 1 mm thick deposit of3-nitrobenzyl alcohol was prepared by dropping a 3.0 mL neat NBA(98%, Aldrich, Milwaukee, WI) onto a stainless steel sample target,which was placed at the center of the ablation chamber. Theresulting liquid sample spot on the target was approximately6 mm in diameter and was sufficiently viscous so as to stay in placethroughout the experiment. Particles were formed by laserablation of NBA from the target into the chamber that wassuffused with filtered compressed air at a flow rate of 5 L/min.Ejected particulate was pulled into the particle sizer, which wasmounted directly under the chamber.

The two laser systems used in this experiment were awavelength-tunable IR optical parametric oscillator (OPO) system(Mirage 3000B, Continuum, Santa Clara, CA) and a UV excimer laser(Optex, Lambda Physik AG, Gottingen, Germany). The OPO wasused as the primary particle formation laser and the excimer wasused as the particle irradiation laser. The OPO is tunable from 1.4 to4.0 mm; however, it was fixed at 2.94 mm for the work described inthis article. The excimer was operated using a xenon fluoride gasmix for operation at the wavelength of 351 nm. Both lasers wereattenuated by inserting optical flats into the beams. The temporalpulse width for OPO and excimer are 5 and 8 ns, respectively.

The IR beam was directed through a 254 mm focal length CaF2

lens that produced a 250 mm � 300 mm spot on the target asdetermined by laser burn paper. The sample target was irradiatedby the IR light at normal incidence, and a sapphire window wasused on the chamber. The UV laser beam was oriented parallel tothe plane of the sample target. The UV laser was focused to a spotsize of 1.4 mm � 0.6 mm directly above the target using a 254 mmfocal length cylindrical lens. The distance between the center of UVbeam and the target was 3.5 mm and was adjusted using a singleaxis stage (LMT-151, MDC, Hayward, CA).

A digital delay generator (Model DG535, Stanford ResearchSystem, Sunnyvale, CA) was triggered by the Q-switch synchro-nization output of the OPO pump laser at a repetition rate of 2 Hz.After a selected delay, the pulser triggered the excimer laser at thesame repletion rate. The time delay between the lasers is indicatedbelow by Dt, where Dt = tUV � tIR. The pulse energies of the IR andUV lasers were measured using a pyroelectric detector (ED-104AX,Gentec, Palo Alto, CA).

The concentration and size distribution of ablated particulateswere measured using the light scattering particle sizer. Particleswith aerodynamic diameters between 500 nm and 20 mm weremeasured to a precision of 30 nm. Measurements were initiated10 s after target irradiation was initiated to allow the system toreach a steady state of particle production. The measurementsignals were acquired for 10 s (20 laser shots).

3. Results

The size distributions of ablated particles for different delaytimes are show in Fig. 1. The IR and UV fluences were 1900 and2400 J/m2, respectively. The height of each vertical bar indicatesthe concentration of ejected particles measured in the indicatedaerodynamic diameter range and the y-axis in each plot extendsfrom zero concentration to 30 cm�3. The total particle concentra-tion, obtained by summing the counts in each of the size ranges, isgiven in Table 1. This table includes both the particle concentrationin terms of particles per unit volume as well as the mass-weightedconcentration in terms of total mass of particulate per unit volume.The maximum total particle concentration is 400 cm�3, which isobserved at 20 ms delay time (Fig. 1c). This value is more than 5times the baseline concentration with the IR laser alone (Fig. 1a).

A plot of the UV fluence dependence of the particle concentra-tion is shown in Fig. 2. The IR laser fluence was 1900 J/m2 and thedelay time was 20 ms. The threshold for two-laser particleformation is near 250 J/m2. Between this fluence and 900 J/m2,the particle concentration increases linearly with a slope of 25particles/cm3 per 1000 J/m2. Between 900 and 1600 J/m2, the rise islinear, but the slope increases by a factor of 25. Above 1600 J/m2,the slope decreases to near the initial value.

The particle concentration and the size distribution of theparticles change with the delay time. The peak particle sizes atvarious delay times are listed in Table 1. The peak particle size wasobtained by determining the centroid of the distribution for all pointsgreater than 60% of the maximum. At zero delay, the peak particle

Table 1Concentration and average size of ablated particles at different delay times at IR and UV fluences of 1900 and 2400 J/m2, respectively.

Delay time (ms) Particle

concentration (cm�3)

Mean particle

size (mm)

Peak particle

size (mm)

Mass-weighted

concentration (mg/m3)

Mean mass-weighted

particle size (mm)

0 70 2.12 1.17 2.05 9.88

15 100 1.88 0.85 2.03 9.67

18 280 1.27 0.70 2.06 9.20

20 400 1.20 0.70 2.09 9.02

24 330 1.24 0.70 2.08 9.24

37 230 1.36 0.70 2.07 9.51

100 160 1.54 0.70 2.01 9.63

300 100 1.83 0.75 2.09 9.73

1,000 90 1.90 0.78 2.02 9.81

7,000 80 2.01 0.88 2.08 9.79

15,000 70 2.09 1.09 2.04 9.97

Fig. 2. UV fluence dependence of the ablated particle concentration at an IR fluence

of 1900 J/m2 and 20 ms delay.

Fig. 3. Delay time dependence of particle concentration at different IR fluences ((&)

2700 J/m2, (&) 2100 J/m2, (&) 1900 J/m2, (*) 1600 J/m2) at delay times from zero

to (a) 1000 ms and (b) 100 ms. The UV laser fluence was 2400 J/m2.

X. Fan, K.K. Murray / Applied Surface Science 255 (2009) 6297–6302 6299

size is 1.17 mm. The peak particle size decreases to 0.70 mm after20 ms and stays constant from 18 to 100 ms before it begins toincrease. After 15 ms, the peak particle size has increased to 1.09 mm.

Shown in Fig. 3 is the delay time dependence of the total particleconcentration at different IR fluences of 1600, 1900, 2100, and2700 J/m2 with the UV laser fluence held constant at 2400 J/m2.Fig. 3a shows the time behavior between 0 and 1 ms and Fig. 3b isan expansion of the time range between 0 and 100 ms. At zerodelay there is no UV laser interaction with the ablated material andthe concentration of particles (baseline) increases with IR fluence.At delay times between 10 and 30 ms, the plume of ablatedmaterial has moved into the path of the UV laser beam andirradiation with the UV laser leads to particle break-up and agreater particle count. The concentration increase that results fromthe UV laser continues beyond 200 ms, suggesting that particulateis emitted from the surface long after the IR laser has irradiated it.

Fig. 4 shows the mass-weighted concentration of ablatedparticles at different IR laser fluences with the IR laser alone andwith the UV laser at a fluence of 2400 J/m2 and at the delay timecorresponding to maximum particle count at each fluence.Although the UV laser has a large effect on the total particleconcentration (see Fig. 3), the total mass of particulate is notsignificantly affected by the second laser.

Plots of the mean particle size at different delay times areshown in Fig. 5. The UV fluence was 2400 J/m2 and the IR fluencewas 600, 1900, 2100, and 2700 J/m2. In all cases, there is a rapiddecrease in particle size as the plume passes the UV laser beam atdelays near 20 ms. There is a minimum for particle size near 26, 20,18, and 16 ms when the IR fluence is 1600, 1900, 2100, and 2700 J/m2, respectively. These time delays correspond to the positions ofthe maxima of particle concentration (see Fig. 3). As with theparticle concentration results, the decrease in particle size isobserved for delays of several milliseconds between the IR laserand UV laser firing, suggesting delayed ejection of material from

Fig. 4. Mass-weighted concentration with IR laser (*) and IR plus UV laser (&) at

the peak particle concentration. The UV laser fluence was 2400 J/m2.

Fig. 5. Particle diameter as a function of delay time at IR ablation fluences of (*)

1600 J/m2, (&) 1900 J/m2, (&) 2100 J/m2, and (&) 2700 J/m2 at delay times from

zero to (a) 5000 ms and (b) 100 ms. The UV fluence was 2400 J/m2.

Fig. 6. The ejected particle velocity as a function of IR fluence dependence.

X. Fan, K.K. Murray / Applied Surface Science 255 (2009) 6297–63026300

the surface. The mean particle sizes at different delay times for anIR laser fluence of 1900 J/m2 are indicated in Table 1.

Since the distance between the target surface and the center ofthe UV laser beam was fixed at 3.5 mm, the average velocities ofplume front can be obtained from the delay time corresponding tothe maximum of the particle concentration plot of Fig. 3. Fig. 6shows a plot of the particle plume velocity obtained from themaximum particle concentration. The plume front velocityincreases linearly from 130 m/s at 1600 J/m2 to 220 m/s at2700 J/m2.

4. Discussion

The above results demonstrate that it is possible to affect asignificant break-up of the micrometer-sized particulate ablatedfrom a sample using a UV laser. The particle concentration can beincreased by nearly an order of magnitude, but the total mass ofejected material is not greatly affected, suggesting that the UV laseris breaking the large particles into smaller ones rather than causingcomplete vaporization.

The mechanism of IR laser material ejection can be understoodin terms of the thermomechanical response of the nitrobenzylalcohol matrix material to the pulsed IR laser [25]. The energy perunit volume deposited in the nitrobenzyl alcohol by the IR OPO isgiven by

E

V¼ 2:3 ð1� RÞencnH0 (1)

where R is the surface reflectance, en is the molar decadicabsorption coefficient, cn is the molar concentration, and H0 is the

laser fluence [26]. The molar decadic absorption coefficient of NBAis not known, but measurements made for glycerol and otheralcohols suggests that it is in the range of 100 L/mol cm [27]. The IRlaser fluence was in the range of 2000 J/m2 for these studies and thereflectance is estimated to be 0.1. The molar concentration of theNBA liquid is 8500 mol/m3. These values lead to an energy densityof 4 � 108 J/m3 or 40 kJ/mol. The heat of vaporization of NBA is notavailable; however, this value can be compared to the heat ofvaporization for benzyl alcohol of 60 kJ/mol [28]. The values are ofthe same magnitude, suggesting that the volumetric energydensity of the NBA following IR laser irradiation is close to thatrequired for a phase transition.

The temperature rise in the absence of phase transition can alsobe calculated based on the energy absorbed and assuming thatthere is no phase change, the increase in temperature is given by

DT ¼ E

V

1

cvr(2)

where cv is the heat capacity of NBA and r is the density. Taking theheat capacity to be 200 J/mol K, the predicted temperature rise is210 8C to a final temperature of 235 8C. This is well above theboiling point between 175 and 180 8C and approximately 75% ofthe estimated critical temperature [29]. If the sample is heatedrapidly to approximately 90% of the critical temperature, thesuperheated material can undergo an explosive phase transitionthat results in volume ejection of the upper layers of the sample[25,30–32]. It appears likely that these conditions may be in effect,especially for higher laser energy levels.

The phase change can occur in either the stress or thermalconfinement regimes depending on the rate of energy relaxation.The time for thermal energy dissipation in a liquid MALDI matrix ison the order of microseconds [33], thus the system is most likelywithin the thermal confinement regime under all of the experi-mental conditions described above. The characteristic time foracoustic energy dissipation is given by

tac ¼1

ac(3)

where a is the IR absorption (�200/mm) and c is the speed ofsound in NBA (�1000 m/s). When energy is added to the system ona time scale less than tac, the system is in the stress confinementregime. Using the above values, the acoustic relaxation time is�5 ns and nearly identical to the laser pulse width. This puts thesystem at the stress confinement limit.

Time-resolved imaging has been used to study the ejection ofmaterial from liquid samples under vacuum using mid-IR lasers.Leisner et al. used dark field images to measure plume density andright angle scattering to image ejected particulate from a glycerolsample [34]. They found different plume dynamics for lasersoperating in the thermal and stress confinement regimes. A 5 nsOPO, similar to the one used in this work, was found to generate adenser plume with more particulate compared to a 100 ns Er: YAGlaser. Pressure confinement was not achieved with the longer pulsewidth Er: YAG at 5200 J/m2 fluence, but it was at or near thatregime with the OPO at 2800 J/m2 fluence. The OPO ejectedmaterial plume showed the lift-off of the entire glycerol layer thatsubsequently disintegrated into particulate. In contrast, theEr:YAG-induced plume resulted in smaller particles that coalescedto form larger particles as the plume expanded. Rohlfing et al.performed a photoacoustic analysis of glycerol IR ablation undersimilar conditions and found a distinct contrast in the pressurepulse signals of the OPO compared to the Er:YAG [33]. The OPO wasfound to generate more than an order of magnitude greaterthermoelastic stress and this led to a rapid phase explosion and lift-off of the entire irradiated volume.

X. Fan, K.K. Murray / Applied Surface Science 255 (2009) 6297–6302 6301

Recent high-speed photographic studies of 2.94 mm Er:YAGlaser ablation of water reveal a two-step process involving therapid boiling and ejection of water molecules at 1000 m/s that isfollowed by the lift-off of particles at velocities about an order ofmagnitude slower [32]. The slower particle ejection process iscaused by the recoil of the shock wave induced by the initial laser-induced phase transition. Although the particulate jet extendsseveral millimeters after a few tens of microseconds, theparticulate continues to be expelled for as long as 500 ms afterthe IR laser fires. The onset of shock wave induced material ejectionoccurred at 5400 J/m2, which corresponds to an irradiance of1011 W/m2, given the 70 ns pulse width of the laser. This can becompared to the results presented above that were obtained atfluences of 1000–2000 J/m2, which correspond to irradiances of(2–4) � 1011 W/m2. The higher irradiance of the OPO leads tovolume ejection of material [32–34].

The UV laser used in our experiments appears to be irradiation aplume of particulate similar to that ablated from water [32] andglycerol [34]. The maximum concentration (Fig. 3) and minimumparticle diameter (Fig. 5) are consistent with the irradiation ofparticulate material that has been ejected in the initial formation ofdroplets after explosive vaporization, followed by irradiation ofparticulate material at longer time delays that results from therecoil-induced material expulsion process. Although extendedmaterial emission to more than 3 ms has been observed with the IROPO laser system irradiating a glycerol sample at 3000–5000 J/m2,emission beyond several tens of microseconds was not observed[34]. Under a vacuum, the plume expands freely without collision;the lack of a confined and slowly expanding plume could explainthe lack of extended material emission.

The velocity of ions emitted from MALDI matrix material undervacuum has been reported at 5000–1000 m/s [35], consistent withthe high velocity of primary material ejection. By contrast, ionsformed by irradiation of ablated particles with a second laser areformed for tens of microseconds after the IR irradiation of thetarget [20,21]. This is consistent with an initial phase transitionthat ejects molecules and clusters that is followed by a shock wave-induced volume ablation of material that is observed in fastphotography studies. The rapid rise in the particle concentrationseen in Fig. 3 and drop in particle diameter between 10 and 20 msafter the IR laser pulse indicates the arrival of the front of theparticle component of the ejected plume. The slow decay of thesignal indicates the delayed emission of material from the sample.The rise in concentration drops below half its maximum within50 ms and a return to the baseline is observed after more than300 ms delay. The late time signal may be due to recoil-inducedexpulsion as well as coagulation of ejected particles above thetarget. High-speed photography of IR laser induced water ablationshows the formation of a column of material above the surface formore than 300 ms [27]. The column of ablated material is rapidlyexpelled and after 50 ms collapses into a cone and disintegrates.This is consistent with the strong effect observed in Figs. 3 and 5between 10 and 50 ms that probes the column of ablated material,followed by the weak effect after 50 ms that probes thedisintegrating column.

Molecular modeling has been used to simulate the break-up ofparticles irradiated with a pulsed laser [36,37]. The simulationsindicate that a weakly absorbing particle that is uniformly heatedwill be completely evaporated. In the case of a strongly absorbingparticle, the energy is absorbed on the side facing the laser. Theirradiated material evaporates while the material that has not beenheated breaks up into smaller particles. The UV absorptioncoefficient of NBA has not been reported, but it is known that atypical MALDI matrix has an absorption of 6 � 104 cm�1 [38]. Thisindicates a penetration depth of about 160 nm, which isqualitatively consistent with the observed particle size reduction.

The sigmoidal shape of the concentration plot of Fig. 2 can beinterpreted in light of this proposed mechanism. According to thismodel, the UV fluence range up to 1000 J/m2 represents removal ofmaterial from the surface of the particle by evaporation andlimited break-up of the material that has not been irradiated. Usingthe estimated absorption from above, the energy density in theNBA particle will exceed 60 kJ/mol above 1000 J/m2 UV fluence.We postulate that the concentration increases rapidly between1000 and 1800 J/m2 because there is sufficient volumetric energyto achieve a rapid phase transition throughout the irradiatedelement. This results in a more efficient break-up of the portion ofthe particle that has not been irradiated and therefore a highermeasured particle concentration. Above 1800 J/m2, the phaseexplosion effect is saturated and additional fluence does not lead toa large increase in particle concentration.

5. Conclusions

Particles ablated from a sample of 3-nitrobenzyl alcohol can bebroken up by irradiation with a UV laser directed at the plume ofexpanding material. The particle plume expands at a velocity ofseveral hundred m/s and is sufficiently concentrated that the UVlaser is able to increase the particle concentration three to ninetimes while decreasing the particle size by one-third to one-half.The plume velocity is dependent on the IR laser fluence and rangedfrom 130 m/s at 1600 J/m2 to 220 m/s at 2700 J/m2. Ejection ofmaterial from the sample occurred over a prolonged time period;the effect of the UV laser on the particle plume were observed forseveral ms after the IR laser was fired at the sample. The observedparticle emission dynamics is consistent with a rapid phasetransition and plume expansion, followed by a prolonged shockwave material removal process.

The UV laser break-up of IR ablated particles is important forimproving analytical applications and in understanding themechanisms of pulsed laser material removal. The performanceof inductively coupled plasma mass spectrometry or atomicemission spectroscopy is strongly dependent on the size andcomposition of the ablated material [10,11]. The addition of a post-ablation laser opens up the possibility of additional processing theablated material and potentially improving analytical perfor-mance. The technique of infrared laser matrix-assisted laserdesorption electrospray ionization (IR-MALDESI) may also benefitfrom the use of a post-ablation processing laser [16–18]. From amechanistic standpoint, the ability to probe the plume of ablatedmaterial in size and composition is an important tool forunderstanding laser material removal. Although this process canbe observed using high-speed photography techniques [34], theseare not always sensitive to small particles at a small numberdensity.

Acknowledgments

This work was supported by National Science Foundation grantCHE-0415360 and Louisiana State University.

References

[1] M. Handschuh, S. Nettesheim, R. Zenobi, Appl. Surf. Sci. 137 (1–4) (1999) 125–135.[2] S. Alves, M. Kalberer, R. Zenobi, Rapid Commun. Mass Spectrom. 17 (18) (2003)

2034–2038.[3] S.N. Jackson, S. Mishra, K.K. Murray, J. Phys. Chem. B 107 (47) (2003) 13106–

13110.[4] K. Dreisewerd, Chem. Rev. 103 (2) (2003) 395–425.[5] K. Dreisewerd, S. Berkenkamp, A. Leisner, A. Rohlfing, C. Menzel, Int. J. Mass.

Spectrom. 226 (1) (2003) 189–209.[6] K.K. Murray, Infrared MALDI, in: R.M. Caprioli, M.L. Gross (Eds.), Encyclopedia of

Mass Spectrometry, vol. 6, Elsevier, 2006.[7] S.N. Jackson, J.-K. Kim, J.L. Laboy, K.K. Murray, Rapid Commun. Mass Spectrom. 20

(8) (2006) 1299–1304.

X. Fan, K.K. Murray / Applied Surface Science 255 (2009) 6297–63026302

[8] M.W. Little, J. Laboy, K.K. Murray, J. Phys. Chem. C 111 (3) (2007) 1412–1416.[9] X. Fan, M.W. Little, K.K. Murray, Appl. Surf. Sci. 255 (2008) 1699–1704.

[10] D. Guenther, B. Hattendorf, TrAC, Trends Anal. Chem. 24 (3) (2005) 255–265.[11] R. Hergenroeder, Spectrochim. Acta Part B: At. Spectrosc. 61B (3) (2006) 284–300.[12] J. Dong, Y.H. Rezenom, K.K. Murray, Rapid Commun. Mass Spectrom. 21 (24)

(2007) 3995–4000.[13] J. Shiea, M.Z. Huang, H.J. Hsu, C.Y. Lee, C.H. Yuan, I. Beech, J. Sunner, Rapid

Commun. Mass Spectrom. 19 (24) (2005) 3701–3704.[14] M.Z. Huang, H.J. Hsu, J.Y. Lee, J. Jeng, J. Shiea, J. Proteome Res. 5 (5) (2006) 1107–1116.[15] J.S. Sampson, A.M. Hawkridge, D.C. Muddiman, J. Am. Soc. Mass Spectrom. 17 (12)

(2006) 1712–1716.[16] J. Dong, Y.H. Rezenom, K.K. Murray, Proc. ASMS Conf. Mass Spectrom. 55 (2007)

MP06.[17] P. Nemes, A. Vertes, Anal. Chem. 79 (21) (2007) 8098–8106.[18] Y.H. Rezenom, J. Dong, K.K. Murray, Analyst 133 (2008) 226–232.[19] K.K. Murray, D.H. Russell, J. Am. Soc. Mass Spectrom. 5 (1994) 1–9.[20] A. Leisner, A. Rohlfing, S. Berkenkamp, F. Hillenkamp, K. Dreisewerd, J. Am. Soc.

Mass Spectrom. 15 (6) (2004) 934–941.[21] F. Huang, X. Fan, K.K. Murray, Int. J. Mass Spectrom. 274 (1–3) (2008) 21–24.[22] J. Gonzalez, C. Liu, J. Yoo, X. Mao, R.E. Russo, Spectrochim. Acta Part B 60B (1)

(2005) 27–31.[23] T.-W.D. Chan, A.W. Colburn, P.J. Derrick, Org. Mass Spectrom. 26 (1991) 342–344.[24] C. Davies, Br. J. Appl. Phys. 1 (1968) 921.

[25] A. Vogel, V. Venugopalan, Chem. Rev. 103 (2) (2003) 577–644.[26] D. Feldhaus, C. Menzel, S. Berkenkamp, F. Hillenkamp, K. Dreisewerd, J. Mass

Spectrom. 35 (11) (2000) 1320–1328.[27] C. Menzel, K. Dreisewerd, S. Berkenkamp, F. Hillenkamp, Int. J. Mass Spectrom.

207 (1/2) (2001) 73–96.[28] T.S. Papina, S.M. Pimenova, V.A. Luk’yanova, V.P. Kolesov, Russ. J. Phys. Chem.

(Engl. Transl.) 69 (1995) 1951–1953.[29] K.M. Klincewicz, R.C. Reid, AIChE J. 30 (1) (1984) 137–142.[30] A. Miotello, R. Kelly, Appl. Phys. A 69 (0) (1999) S67–S73.[31] L.V. Zhigilei, B.J. Garrison, J. Appl. Phys. 88 (3) (2000) 1281–1298.[32] I. Apitz, A. Vogel, Appl. Phys. A 81 (2) (2005) 329–338.[33] A. Rohlfing, C. Menzel, L.M. Kukreja, F. Hillenkamp, K. Dreisewerd, J. Phys. Chem. B

107 (44) (2003) 12275–12286.[34] A. Leisner, A. Rohlfing, U. Rohling, K. Dreisewerd, F. Hillenkamp, J. Phys. Chem. B

109 (23) (2005) 11661–11666.[35] S. Berkenkamp, C. Menzel, F. Hillenkamp, K. Dreisewerd, J. Am. Soc. Mass

Spectrom. 13 (3) (2002) 209–220.[36] T.A. Schoolcraft, G.S. Constable, L.V. Zhigilei, B.J. Garrison, Anal. Chem. 72 (21)

(2000) 5143–5150.[37] L.V. Zhigilei, Y.G. Yingling, T.E. Itina, T.A. Schoolcraft, B.J. Garrison, Int. J. Mass.

Spectrom. 226 (1) (2003) 85–106.[38] X. Chen, J.A. Carroll, R.C. Beavis, J. Am. Soc. Mass Spectrom. 9 (9) (1998) 885–891.