ISSN 10619348, Journal of Analytical Chemistry, 2011, Vol. 66, No. 13, pp. 1227–1242. © Pleiades Publishing, Ltd., 2011. Original Russian Text © P.A. Kuzema, 2010, published in Massspektrometria, No. 7(4), pp. 243–260.

REVIEWS

SmallMolecule Analysis by SurfaceAssisted Laser Desorption/Ionization Mass Spectrometry1 P. A. Kuzema Chuiko Institute of Surface Chemistry, National Academy of Sciences of Ukraine, ul. Generala Naumova 17, Kiev, 03164 Ukraine email: sci[email protected] Received August 24, 2010; in final form, October 13, 2010

Abstract—In order to meet the challenges facing modern chemistry, biology, and medicine, methods are required capable of performing rapid and reliable analysis of both individual compounds and complex mix tures at the molecular level. Matrixassisted laser desorption/ionization mass spectrometry meets these requirements; however, some limitations complicate its application for the analysis of small molecules. Recently, smallmolecule analysis has greatly progressed owing to development of surfaceassisted laser de sorption/ionization mass spectrometry involving approaches which combine the unique properties of nano structured surface chemistry and morphology. This review examines such approaches and their specific appli cation in smallmolecule mass analysis. Keywords: mass spectrometry, MALDI, SALDI, small molecules, nanostructures DOI: 10.1134/S1061934811130065 1

CONTENT 1. Introduction 2. Mass Spectrometry with Direct and Matrix Assisted Laser Desorption/Ionization of Small Mole cules 3. SurfaceEnhanced Laser Desorption/Ionization 4. The Concept of SurfaceAssisted Laser Desorp tion/Ionization Mass Spectrometry 5. Analysis of LowMolecular Compounds by Sur faceAssisted Laser Desorption/Ionization Mass Spectrometry 5.1. Desorption/Ionization on Macroscopic Sur faces 5.2. Desorption/Ionization on Macroscopic Nanostructured Coatings 5.3. Desorption/Ionization on the Surface of Micro and Nanoparticles 6. Mechanism of SurfaceAssisted Laser Desorp tion/Ionization and Factors Affecting Its Efficiency 7. Comparative Efficiency of Various Materials and Surfaces in SmallMolecule Analysis by Surface Assisted Laser Desorption/Ionization Mass Spec trometry 8. Conclusions Abbreviations: DIOS, desorption/ionization on porous silicon; GALDI, graphiteassisted laser de sorption/ionization; IMS, imaging mass spectrome try; LDI, laser desorption/ionization; MALDI, matrix assisted laser desorption/ionization; MESALDI, 1 The article was translated by the author.

matrixenhanced surfaceassisted laser desorp tion/ionization; NALDI, nanowireassisted laser des orption/ionization; NIMS, nanostructureinitiator mass spectrometry; SALDI, surfaceassisted laser des orption/ionization; SEAC, surfaceenhanced affinity capture; SELDI, surfaceenhanced laser desorp tion/ionization; SEND, surfaceenhanced neat des orption; SPALDI, silicon nanoparticleassisted laser desorption/ionization; CHCA, αcyano4hydroxy cinnamic acid; TLC, thinlayer liquid chromatogra 2

phy; MPC, metalphthalocyanine.

1. INTRODUCTION Laser ion sources have been applicable in mass spectrometry since the 1960s [2]—several years after the laser was invented. Later, the process and method of ionized atom and molecule formation in the gas phase by means of direct sample exposure to laser irra diation have become known as laser desorption/ion ization (LDI) [3, 4]. Initially, LDI mass spectrometry had limitations related to complications in controlling the thermal decomposition of the analyte, low sensi tivity, and small variety of compounds classes (mainly, lowmolecular organic substances) capable of effi ciently absorbing radiation in the wavelength range of IR (1.06 and 10.6 μm) or UV (266 nm) lasers [5]. As a result of further investigations of the influence of laser irradiation on desorption/ionization of organic sub 2 Accepted

Russian abbreviations and terms related to laser mass spectrometry can be found in [1].

1227

1228

KUZEMA

stances, in the mid1980s, a variant of mass spectrom etry with matrixassisted laser desorption/ionization (MALDI) was proposed [6]. The use of a matrix—a lowmolecular organic substance facilitating analyte desorption/ionization under laser irradiation—has made it possible to significantly broaden the range of analyzed molecules. The MALDI mass spectrometry became popular in the late1980s after a demonstra tion of its capabilities (nearly simultaneously by research groups from Germany (Karas and Hillen kamp [7]) and Japan (Tanaka et al. [8])) in analyzing proteins and other macromolecules. MALDI has become a powerful analytical tool in chemical and biological investigations owing to such advantages as high efficiency and sensitivity, soft and efficient ionization of nonvolatile and fragile com pounds, tolerance to impurities, rapid mass determi nation, and relative simplicity of the mass spectra obtained [9, 10]. However, the interference of the matrix and analyte signals in a low molecular mass range, sensitivity of the matrix choice, and nonuni form distribution of the analyte during its cocrystalli zation with the matrix (low reproducibility) limit the application of conventional MALDI for analysis of lowmolecular compounds (M < 1500 Da)—so called “small molecules” [10]. The increasing rigidity of analysis of small molecules, both individually and in a mixture, has been an impetus for researchers to improve the LDI methodology while simultaneously improving LDI mass spectrometers. Methodical advances were mainly directed at suppressing matrix signals or using special highmolecular matrices for smallmolecule analysis [5]. The use of various surfaces facilitating laser desorp tion/ionization of the analyte has become the main milestone in LDI mass spectrometry for smallmole cule analysis. Further, this approach will be designated here as surfaceassisted laser desorption/ionization (SALDI) mass spectrometry—a term introduced in scientific use by Sunner et al. [11] in 1995, who applied graphite particles suspended in a mixture with glycerol, sucrose, and methanol for peptide analysis. This approach was implemented for the first time in 1988 by Tanaka et al. [8], who studied the possibility of applying cobalt nanoparticles suspended in glycerol to analyze proteins and polymers. Recently, SALDI mass spectrometric smallmolecule analysis has greatly progressed owing to approaches which combine the unique properties of nanostructured surface chemistry and morphology. The use of special surfaces made it possible not only to obviate the matrix effect, but also to increase the sensitivity and selectivity of analysis of certain small molecules. Development of the technology for synthesis and modification of nanostructured surfaces has stimu lated the development of various approaches in SALDI mass spectrometry. In this review, uptodate achievements, tendencies, and outlooks in the use of various approaches and materials to study appropriate

classes of lowmolecular compounds using laser, in particular, SALDI mass spectrometry are considered. The material, as well as the majority of the works examined, is devoted mainly to studies in the field of mass spectrometry with UV laser desorption/ioniza tion. 2. MASS SPECTROMETRY WITH DIRECT AND MATRIXASSISTED LASER DESORPTION/IONIZATION OF SMALL MOLECULES A conventional MALDI experiment envisages the deposition of a mixture containing dissolved analyte and matrix onto the surface of a target (standardized stainless steel substrate). After evaporation of the sol vent, matrix–analyte crystals are exposed to IR (usu ally λ = 10.6, 3.28, or 2.94 μm) or UV (usually λ = 355, 337, 266, or 193 nm) laser irradiation with a flu ence of about 106 W/cm2 and a pulse period of a few nanoseconds, which results in simultaneous desorp tion of matrix and analyte molecules into the gas phase and their ionization [5]. As a result of laser ablation, a large quantity of neutral associates and mainly singly charged ions are formed in the gas phase [12]. Since laser radiation is pulsed, it is optimal to couple MALDI with a timeofflight mass analyzer; the suc cess in smallmolecule analysis is defined by the possi bility of intact ion detection and depends significantly on the choice of matrix. The role of the matrix is to absorb the energy of laser radiation photons and to transfer it into the energy of system excitation, to transfer protons during ionization, and to dissolve the analyte, minimizing the aggregation of its molecules [5, 10]. Generally, the matrix is usable for smallmol ecule analysis if it provides efficient ionization and minimal fragmentation and causes no signal interfer ence in the mass spectrum [13]. Conventional matrices for MALDI are crystalline organic substances intensively absorbing UV radia tion. Among the large variety of substances used as matrices (see, for instance, [5, 12]), cinnamic acid derivatives and aromatic carbonyls are frequently used for smallmolecule analysis [14]. In particular, αcyano4hydroxycinnamic acid (CHCA) and 2,5dihydroxybenzoic acid efficient in the analysis of small peptides, aminoacids, purine and pyrimidine bases, macrocyclic porphyrin metal complexes, and phthalocyanines. However, their efficiency depends significantly on the matrix : analyte molar ratio, which is optimal within the range (10–1–103) : 1 [10]. Among other matrices, 9aminoacridine for quantitation of bile acids in plasma [15], trans3indoleacrylic acid for identification of catechin oligomers in apples [16], and 2',4',6'trihydroxyacetophenone for identification of isoflavones in soya samples [17] were successfully used. The limitations of conventional MALDI mass spectrometry complicate the task of lowmolecular

JOURNAL OF ANALYTICAL CHEMISTRY

Vol. 66

No. 13

2011

SMALLMOLECULE ANALYSIS

compound analysis and stimulate the search for novel efficient matrices. The interference of matrix and ana lyte signals could be avoided by using matrices with a higher molecular weight. For instance, some porphy rins can be used as such matrices [18, 19]. Traditional acid matrices in the case of organometallic and coor dination compound analysis can generate mixtures of positive ion radicals and protonated species, which complicates interpretation of the data. The electron carrying aprotic matrix 2[(2E)3(4tertbutylphe nyl)2methylprop2enylidene]malononitrile was found to be efficient in the analysis of such compounds as acetylacetonate complexes with transition metals [20] generating mainly positive ion radicals. A successful choice of matrix is the main, but not the only success criterion in smallmolecule analysis by MALDI. The sample preparation procedure (including its preliminary concentration), the method of its deposition onto the substrate, and the technique of matrix signal suppression envisaging the optimiza tion of matrix : analyte ratio and the use of various additives decreasing the matrix signals, are also impor tant; they are described in detail in [10, 13]. These reports also contain a review on qualitative and quan titative determination of small proteins, carbohy drates, and some other classes of lowmolecular com pounds by MALDI mass spectrometry. Zhang et al. [21] proposed a novel strategy for smallmolecule analysis by MALDI mass spectrome try in the high mass range using metalphthalocya nines (MPCs) capable not only of absorbing laser energy, but also of forming matrix–analyte additive compounds (adducts). Compounds with a mass M = M(MPC) + M(analyte) and M = M(MPC) + M(ana lyte) – 1 are detected in the positive and in the negative ion mode, respectively. MPCs themselves are also detectable and can serve as an internal standard. Alu minum, gallium, and indium phthalocyanines with alkyl and phenyl groups (Fig. 1) have proven to be effi cient matrices forming adducts with various small molecules (aminoacids, peptides, fatty acids, etc.), both in the positive and negative ion detection mode. The detection limit is 17–75 fmol depending on the analyte—the higher the acidity, the lower is the detec tion limit, which indicates the important role of elec trostatic interactions in the formation of matrix–ana lyte adducts. The intensity of the analytical peak (I) in a mass spectrum significantly depends on the matrix : analyte molar ratio (the optimal value is from 1 : 10 to 10 : 1), metal (I(Al) Ⰷ I(Ga) > I(In)), and analyte pKa value (the lower pKa, the higher I). Mass spectrometry with direct (matrixless) laser desorption/ionization of the analyte (LDI method) remains traditionally one of the means of smallmole cule analysis. In the modern implementation of this method, conventional MALDI targets can be used; however, in contrast to MALDI, the sample prepara tion procedure excludes matrix addition and, corre spondingly, the LDI mass spectrum contains only sig JOURNAL OF ANALYTICAL CHEMISTRY

Vol. 66

1229

R N N R

N

N M N

N

R

N N R

M = Al3+, Ga3+, In3+

R = Alk, Ph

Fig. 1. Structural formula of phenyl and alkylsubstituted metalphthalocyanines [21].

nals related to ionized analyte molecules. LDI, how ever, has limitations related to the appropriate requirements on the analyte (in particular, high proton donating activity and absorptivity in the range of the laser radiation wavelength), increased probability of aggregation as well as fragmentation because of the need to apply a higher laser power to achieve desorp tion/ionization [22]. Despite this, the LDI method has found application for analysis of many classes of lowmolecular compounds. Thus, for instance, vari ous triarylboranes [23], oil porphyrins [24], substi tuted polycyclic carboxylic acids [25], fractions of pyrolytic lignines [26], polycyclic aromatic hydrocar bons [27], fullerenes [28], synthetic derivatives of fla vonol and their complexes with zinc and iron [29], zir conium azaborolinyl complexes [30], and copper phthalocyanine complex [31] have been successfully characterized by this method. By means of LDI mass spectrometry, it was possible to identify a number of polyphenolic compounds in vine samples [32], to reveal the correlations related to decoloration of “aging” polymers and the presence of coloring pig ments in their composition [33], to study the dissocia tion and photochemical behavior of various organic semiconducting materials (some triarylamines, substi tuted phenantrolines and aluminum quinolinates) [34], to perform the qualitative monitoring of organic and inorganic, heterogeneous, and multiphase reac tions in a system simulating reactions involving parti cles which exist in the troposphere [35]. The approach combining thinlayer chromatogra phy (TLC) with LDI mass spectrometry was shown in [36] with the example of an analysis of small alkaloid molecules berberine and palmatine. Using standard silica gel TLC plates, the authors first extracted and separated the mentioned alkaloids and then visualized them using a UV lamp (366 nm), cut the identified parts, and attached them, using adhesive film, to the modified MALDI target. The LDI mass spectra obtained were “clean” and contained mainly molecu lar cations of the analyte, the efficient desorption/ion No. 13

2011

1230

KUZEMA

CH3 OH

O

CH3

O O

O

CN

CN

CN

COOH

COOH

COOH

O

O

+

O

Cl

CN Methacryloyl chloride COOH

αCyano4hydroxy cinnamic acid

αCyano4methacryloyl Polyαcyano4methacryl oyloxycinnamic acid oxycinnamic acid

Fig. 2. Scheme of cinnamic acid functionalization and synthesis of matrix polymeric film.

ization being achieved at a relatively low power of the instrumental laser—comparable to that commonly used in conventional MALDI. For those lowmolecular compounds not detected directly by LDI method, the socalled “analyte derivatization” approach (chemical interaction involving the functional groups of the analyte and matrix molecules with subsequent detection of the analyte–matrix reaction products) has been proposed. For instance, small carbonyl compounds, derivatized via the condensation reaction involving the carbonyl group of the analyte and the aminogroup of the matrix 4dimethylamino6(4methoxy1naph thyl)1,3,5triazine2hydrazine, were studied using this approach [37]. 3. SURFACEENHANCED LASER DESORPTION/IONIZATION In 1993 Hutchens and Yip proposed an analytical strategy called “surfaceenhanced laser desorp tion/ionization” (SELDI) [38], the essence of which lies in use of targets with surfaces modified in a special way to increase the efficiency of the analyte laser des orption/ionization. One of the SELDI modifications, defined as surface enhanced neat desorption (SEND), envisages the use of targets with matrix molecules grafted to the surface—in this case the matrix signals in mass spectra could be avoided while maintaining a high efficiency of analyte desorption/ionization. The synthesis of a polyαcyano4methacryloyloxycin namic acid film on the substrate surface via the Schot ten–Baumann reaction [39] (Fig. 2) is a typical exam ple of such an approach. The second SELDI modification—surface enhanced affinity capture (SEAC)—includes meth ods to form surfaces for extraction and retention of appropriate analyte classes due to hydrophobic, elec trostatic, Lewis acid/base interactions, as well as coor dination binding [39]. Such molecules could be subse quently analyzed by both LDI and MALDI. This

approach is a peculiar sample preparation strategy combining the principles of MALDI mass spectrome try and liquid chromatography. Modified surfaces can contain hydrophilic, hydrophobic, cation and anion exchange groups, as well as metal ions [5]. The main field of SELDI application is proteom ics. In 1997 the SELDI technology was commercial ized by Ciphergen Biosystems as the ProteinChip sys tem [40]. A detailed review on achievements and spe cific applications of this technology in proteomics is given in [41]. The promise of SELDI strategy for smallmolecule analysis lies in combining the SEND and SEAC principles. Such a combination makes it possible not only to simplify sample preparation and avoid matrix signals, but also to increase the analysis sensitivity. The authors of [42] have suggested a method to synthesize hybrid polymeric films contain ing functional fragments responsible for facilitation of ionization (based on matrices of cinnamic and dihy drobenzoic acids, 2',6',6'trihydroxyacetophenone), signal enhancement (acrylic acid), specific interaction with analyte (stearyl methacrylate), stabilization of films and their cohesive and adhesive properties (tri methoxysilyl methacrylate). 4. THE CONCEPT OF SURFACEASSISTED LASER DESORPTION/IONIZATION MASS SPECTROMETRY The term SALDI was originally proposed by Sun ner et al. [11] in 1995, who used graphite microparti cles as ion emitters. Initially, it was implied that SALDI activity is mainly related to the physical prop erties of the materials and surfaces used, in particular, to electroconductivity and efficient absorptivity of radiation in the range of the laser wavelength. Later, it was shown (see below) that directed chemical modifi cation of the surface increases not only SALDI activity (thereby providing all the advantages of the SELDI technology), but also the storage and operation life of the materials used.

JOURNAL OF ANALYTICAL CHEMISTRY

Vol. 66

No. 13

2011

SMALLMOLECULE ANALYSIS

From the moment that SALDI was first mentioned in the scientific literature it was published, many arti cles describing studies on the influence of various sur faces on the LDI efficiency of many classes of com pounds. As a result of intensive study of LDI processes on porous surfaces, nanoparticles, and nanostructured coatings by laser mass spectrometry, many desorp tion/ionization methods have been suggested, such as DIOS (desorption/ionization on porous silicon) [43], SPALDI (silicon nanoparticleassisted laser desorp tion/ionization) [44], GALDI (graphiteassisted laser desorption/ionization) [45], and NALDI (nanowire assisted laser desorption/ionization) [46]. The large variety of materials and sample preparation tech niques complicates the classification of laser mass spectrometry methods. Some researchers consider SALDI as a matrixfree method, while others classify it as MALDI using an inorganic matrix. There with DIOS and NALDI, for instance, are considered as separate methods. The last version of IUPAC classifi cation [47] in relation to laser mass spectrometry des ignates only the terms LDI, MALDI, SELDI and SALDI; the latter is defined as MALDI using a partic ulate matrix (with reference to the work where the term SALDI was originally proposed). Thus, the ques tion on SALDI nomenclature remains open. In this review, the SALDI methodology is consid ered as a set of approaches including various technol ogies for materials synthesis, methods of forming coatings, methods of surface modification, and sam ple preparation techniques, the use of which promotes the activation of desorption/ionization of substances analyzed by laser mass spectrometry. Law and Larkin [48] proposed the following criteria by which a partic ular method can be related to SALDI: (1) the LDI efficiency should be much higher than in the case when sanded metal or silicon surface is used; (2) the laser fluence necessary to achieve LDI should be no higher than in the case of MALDI using conventional organic matrices; (3) it should be a soft ionization technique: molecular or quasimolecular ions of the analyte should prevail in the mass spectrum; (4) in the case of fragmentation, its pattern should be orderly and predictable; (5) the method should make it possi ble to analyze many classes of compounds. 5. ANALYSIS OF LOWMOLECULAR COMPOUNDS BY SURFACEASSISTED LASER DESORPTION/IONIZATION MASS SPECTROMETRY In the past decade, many articles have been pub lished that report on the activity of various surfaces in LDI processes of many classes of small molecules. Their diversity does not allow for unambiguous classi fication of the approaches neither by sample prepara tion techniques nor by the types of the materials used. In [48] it is suggested to classify SALDI approaches by elemental composition of the materials used (carbon JOURNAL OF ANALYTICAL CHEMISTRY

Vol. 66

1231

based, semiconductorbased, and metallicbased); however, the authors themselves admit that this classi fication is tentative since, e.g., carbonbased materi als, depending on molecular structure or atomic arrangement, can be dielectrics, semiconductors, or conductors. In addition, this classification is insuffi cient because it does not include other materials (e.g., oxides) which also possess SALDI activity. In the author’s opinion, it is more reasonable to classify the approaches by the types of the surfaces used—macro scopic surfaces (wafers with porous and nanostruc tured surfaces), macroscopic nanostructured coatings (wafers with grown nanostructures, with nanostruc tured films and other coatings), surfaces of micro and nanoparticles (here, the difference is rather in sample preparation envisaging the presence of the suspension of SALDI active particles in a mixture with the analyte on the surface of a SALDI inactive support), and within each type, by elemental composition (carbon and silicon materials, oxides, metals, and other mate rials). Of course, this classification is also tentative. 5.1. Desorption/ionization on Macroscopic Surfaces Porous silicon. In 1999 Siuzdak et al. [43] proposed a technique of desorption/ionization on porous sili con (DIOS). Owing to the unique combination of sur face morphology, providing the retention of the sol vent and analyte molecules, with high UV absorbtivity of porous silicon, realizing the mechanism of energy transfer to the analyte, the DIOS method was found to be convenient and useful for the analysis of a large variety of small molecules, in particular, peptides and carbohydrates [10], some organic acids and their salts [49], organometallic compounds [50]; the fragmenta tion of the analyte being controlled during desorp tion/ionization. DIOS performance depends essentially on the structure of silicon surface including pore diameter, depth, and form. In [51] the authors studied the influ ence of the conditions of silicon wafers electrochemi cal etching in ethanol solution of HF, as well as some other factors on the LDI efficiency of small peptides. The best results are related to porous silicon obtained from the wafers of ntype silicon (specific resistance 0.008–0.05 Ω/cm) at low current densities (4 mA/cm2) applied over a period of 1–2 min at medium intensity of white light illumination. It has been revealed that the crystal orientation does not play an important role in DIOS performance, as well as a change in the HF concentration in the etching solu tion from 15 to 35% at a standart of 25%. The pore diameter for such silicon supports was ~70–120 nm; the pore depth, up to 200 nm; the porosity, ~30–40% [52]. Optimized and commercially available DIOS supports for smallmolecule analysis (DIOS™, Waters corp.) contain pores with a ~50–100 nm diameter and 400–700 nm depth [53]. No. 13

2011

1232

KUZEMA

“Fresh” porous silicon wafers obtained via electro chemical etching in an alcoholic HF solution contain a considerable amount of surface hydride groups. During longterm storage in air, these groups gradually oxidize, which in general may lead to degradation of the DIOS characteristics [54]. Chemical modification of a silicon surface not only increases the lifetime of wafers, preventing their uncontrollable oxidation, but ensures a surface with desirable properties, thereby increasing the sensitivity and selectivity of the appro priate compounds analysis. Thus, it has been found [52] that covalent attachment of alkenes and alkynes via their hydrosilylation with freshly made porous sili con surface stabilizes the materials and, in general, does not substantially change its efficiency in DIOS processes. It was also shown that silylation of oxidized porous silicon surface using various functional silanes leads to a considerable increase in its DIOS activity. In particular, high efficiency of the selective adsorption approach in combination with DIOS mass spectrom etry is confirmed by successful analysis of the mixtures of protein metabolites, carbohydrates, and certain small nonpolar molecules using, respectively, trime thyl, 3aminopropyldimethyl, and 3perfluoropro pyldimethylsilylated silicon [55], as well as by the analysis of ionic dyes—methylene blue (containing anionexchange groups) and methyl orange (contain ing cationexchange groups)—using silicon with grafted cationexchange (alkylsulfonic acid) and anionexchange (propyloctadecyldimethylammo nium chloride) groups, respectively [56]. Further development of the DIOS approach for analyzing small molecules poorly detected by MALDI has led to the appearance of nanostructureinitiator mass spectrometry (NIMS) [57]—a method in which desorption/ionization of the adsorbed analyte occurs from the surface of materials/initiators captured by the nanostructured silicon surface. This method makes it possible to analyze a wide spectrum of biological mol ecules, and it is efficient in the imaging mass spec trometry (IMS) of biologically important metabolites in tissues. One of the variants—cationenhanced NIMS—has demonstrated the possibility of analyzing carbohydrates and steroids (small biomolecules having a low ionization efficiency) in the form of cation adducts with sodium and silver, respectively, with a sensitivity at a level of 800 and 100 fmol [58]. Ami noorganosilane and fluorinecontaining organodisi loxane were used as materialsinitiators. The high accuracy of this method was shown by an example of glucose and cholesterol quantification (within a con centration interval of 1–200 μM) in human blood serum. In order to improve the DIOS performance in IMS of various metabolites contained in tissues, the technique of matrixenhanced surfaceassisted laser desorption/ionization (MESALDI) [59, 60] has been proposed. Matrix sublimation onto the tissue used in MESALDI makes it possible to overcome the limita

tions related to criticality of tissue density and thick ness for efficient absorption of laser radiation by sili con surface. As a result, the sensitivity of metabolites determination grows substantially which was shown by an example of the analysis of mouse heart and brain tissues. Li and coauthors [61] have suggested obtaining of nanostructured silicon by means of laserassisted elec trochemical etching. In contrast to simple electro chemical etching, a nanostructured surface of micrometer depth is obtained by using this method and a smallmolecule desorption/ionization effi ciency on such supports increases owing to the enlargement of the specific surface area. As well, the nanostructured layer demonstrates a higher stability against laser irradiation exposure. The detection limit of the test compound—hexapeptide dalargin—was 0.32 ± 0.04 pmol. Alimpiev et al. elaborated a method for obtaining nanostructured silicon SALDI substrates by electrochemical etching in an oxidizing iodine containing electrolyte [62]. It has been shown that, owing to the ultrathin oxide layer formed on the sur face, such targets are efficient in SALDI mass spec trometry coupled with gas chromatography (e.g., for analyzing amphetaminelike compounds [63]) and their storage stability is higher in comparison to that for supports obtained by conventional electrochemical etching [64]. Carbon supports. After the demonstration of high SALDI efficiency of graphite particles by Sunner et al. [11], the approach was later developed to use the sur face of mesoporous activated carbon wafers to analyze such small molecules as glucose, caffeine, lysine, and bradykinin, whose detection limit level was no higher than 100 fmol [65]. Studies using graphite supports for analyzing certain polymeric materials in the mass range of 100–1000 Da are given in [66]. The authors of [67] used graphite supports for analyzing certain small macromolecules by means of SALDI mass spec trometry applying a visible wavelength (532 nm) laser. The ionization efficiency of small peptide bradykinin was shown to increase substantially with the addition of glycerol. The authors also noted that visible SALDI is a softer ionization technique as compared to ultravi olet MALDI. 5.2. Desorption/Ionization on Macroscopic Nanostructured Coatings Semiconductor films and nanostructures. The application of semiconductor thin films is one of the most efficient approaches to form surfaces active in LDI processes. Chemical vapor deposition makes it possible to apply nanostructured silicon coatings on various surfaces; here, the film thickness should be within 50–100 nm in order to achieve optimal desorp tion/ionization [68]. Amorphous silicon films with a thickness of no more than 100 nm applied on a micrometersized hybrid organic–inorganic silica

JOURNAL OF ANALYTICAL CHEMISTRY

Vol. 66

No. 13

2011

SMALLMOLECULE ANALYSIS

based material coating by this method were efficient in analyzing small peptides and drug molecules [69]. The femtomole detection limit level was achieved for bradykinin. It has been shown that, unlike films applied to a planar hybride material coating, the films on its micro and nanostructures generate a weaker and stronger signal, respectively. Amorphous silicon films obtained by radiofrequency sputtering were also studied as ion emitters for SALDI mass spectrometry [70]. Mass spectrometry coupled with gas chromatog raphy has shown the high efficiency of some phenyla lkylamines analysis, detection limit was 5– 150 pg/mL. The same coating application method was used to obtain ZnO films on silicon and glass sub strates [71]. The SALDI mass spectra of some drugs (atenolol, reserpine, and gramicidin) detected using these films contained mainly ions of protonated mol ecules with low background noise; the films them selves have demonstrated high stability and good reproducibility with a detection limit of indicated sub stances at a level of 10 fmol. Using the physical vapor deposition technology, supports with nanostructured germanium films were obtained, commercially avail able under the trademark QuickMass™ (NanoHori zons Inc. and Shimadzu corp.). As developers indi cate, these substrates are efficient in analyzing small molecules (in particular, drugs) and can be used and stored in air for a long time without essential loss of SALDI activity [72]. Promising results were obtained using the desorp tion/ionization on silicon nanowires [73]. Silicon nanowires were grown on silicon wafers from gas phase via silane decomposition at 480°C in the presence of Au nanoclusters (10–40 nm). Surface oxidation and fluorosilylation were performed aiming to achieve the desirable polarity and stability. Limit of bradykinin detection was 500 fmol which is six orders lower than in the case of silylated porous silicon used and studied in [55]. Owing to the large specific surface area, it is possible to use silicon nanowires for thinlayer liquid separation and then to perform analysis in different places of the wafer to identify various compounds. An optimal SALDI target contains silicon nanowires with diameters of 10–40 nm, lengths of several microns, and densities of 10–50 wires/μm2; it demonstrates far smaller laser energy thresholds necessary for ioniza tion than the DIOS surface and MALDI matrices [74]. Later, the supports with silicon nanowires were optimized (diameter 20 nm, length 100–500 nm and density > 100 wires/μm2) and commercialized under the trade name NALDI™ (Nanosys Inc. and Bruker Daltonics Inc.), and the researchers from the manu facturers have shown that the sensitivity of mass spec trometric analysis of small, relatively polar molecules using NALDI™ targets is two to ten times higher than that for the conventional MALDI method using CHCA; the possibility of analyzing complex biological mixtures was also demonstrated [75]. The efficiency of JOURNAL OF ANALYTICAL CHEMISTRY

Vol. 66

1233

NALDI™ was evaluated by the authors of [76] for 40 different small molecules—organic acids, amines, aromatic and heteroaromatic compounds, esters, pep tides, aminoacids, carbohydrates, lipids, and nucleo side. Ninetyfive percent of these compounds were successfully analyzed in positive (90%) and negative ions modes with detection limits of 14–200 fmol. The authors did not succeed in analyzing certain test small molecules because of the interference of the desired signal and characteristic background signals of the tar get (m/z 235, 243, etc., corresponding to Aucontain ing ions) in the mass spectra [76]. Nonpolar, organo metallic, and ionic compounds were also studied as analytes for a NALDI™ target [77]. The authors have shown that, in comparison to MALDI, sensitivity using NALDI™ with the example of two porphyrins is three orders higher (the detection limit is in the 10 fmol range) and concluded that NALDI is a “harder” ionization technique as compared to MALDI. Metal coatings and nanostructures. Silicon sub strates coated with Pt nanoparticles were studied as SALDI targets in the analysis of peptides [78]. A sili con support with a platinum coating applied galvani cally from HF solutions was the most efficient. Rely ing on a model of ion formation under SALDI pro posed in [64], the authors of [78] assume that UV radiationstimulated surfacelocalized positive charges that form owing to charge separation (electron and hole) in platinum particles on silicon surfaces play an important role in the transfer of protons to peptides bound with hydroxyl groups of silicon surface via hydrogen bond, and in formation of protonated pep tide molecules: ≡Si–OH⋅⋅⋅NHnR + hν ≡Si+–OH⋅⋅⋅NHnR + e–, ≡Si+–OH⋅⋅⋅NHnR ≡Si–O + H–NHnR. Silicon substrates with a coating of Ag nanoparti cles obtained by reduction of Ag ions from solutions have shown higher stability, sensitivity, and reproduc ibility, in comparison to uncoated porous silicon wafers, in the analysis of such small molecules as tet rapyridineporphyrin, polyethylene glycol oligomers, and peptide oxytocin. The corresponding limits of detection were at the femto, pico and subpicomole levels [79]. Silicon substrates with a twotiered nanostructure comprised of a polyelectrolyte/SiO2 porous film coated with Au nanoparticles were successfully used for the characterization of penta(ethylene gly col)undecane selfassembled monolayers by LDI mass spectrometry [80]. The mass spectrometric analysis of peptides and aminoacids performed by the authors of [81] using glass supports with gold nanowires applied by lithographic electrodeposition was also successful. KCl or NaCl electrolytes were used as the cation source. The dominant peaks in the mass spectra of aminoacids were related to protonated molecules, whereas in the case of peptides, to [M–H+2Na]+ ions. No. 13

2011

1234

KUZEMA

Copper micro and nanostructures of cylindrical (diameter 0.1–2.0 μm, height up to 10 μm) and coni cal (base diameter 0.5–1.5 μm, height up to 5–10 μm) that form with a density of 108–109 cm–2 obtained as replicas of polymeric track membranes were studied as SALDI targets in analyzing the peptide gramicidin [82]. It has been shown that the analysis sensitivity is comparable to that for MALDI, and the ion current intensity of the analyte, as well as the probability of oli gomer ions formation increases with a decrease in diameter of surface structure and an increase in their density. The efficiency of the substrates with conical structures was five to ten times higher as compared to the substrates containing cylindrical surface struc tures. The authors explain this by the role of the effect of enhanced electrical field in the local area at the top of the conical structure whose gradient induces the dipole movement into this area and increases the effi ciency of desorption/ionization. Oxide coatings, films, and nanostructures. Oxi dized silicon wafers containing a micrometer oxide layer were obtained by means of electrochemical etch ing with subsequent hightemperature “wet” oxida tion and studied as SALDI targets for analyzing small biomolecules [83]. The high ionization efficiency and low background noise were demonstrated in analyzing small peptides and catecholamines with femto and picomole detection limit levels, respectively. Supports with a nanostructured porous silicon dioxide coating on a silicon substrate obtained by the template solgel method were also studied [84]. Operational suitability of these SALDI targets was demonstrated in analyzing certain aminoacids, peptides, and siderophores. Oxide films did not lose SALDI activity during storage in laboratory conditions. The template solgel method was also used to obtain stable porous SiO2, SiO2/TiO2, and TiO2 thin films on glass substrates, which were successfully used as SALDI targets for analyzing methylene blue dye and investigating its photodegra dation processes under UV irradiation of various dura tions [85]. Earlier publications on studies of small molecule desorption/ionization from the surface of various oxide films obtained via the solgel method were analyzed in [22]. Titanium supports with vertical TiO2 nanotubes at the surface layer (internal diameter ~70 nm, length ~250 nm) obtained via electrochemi cal etching have demonstrated high SALDI efficiency in analyzing small organic molecules and peptides with a detection limit in the low femtomole range [86]. The supports can be stored in air for a long time and do not require further modification after the etching pro cedure. Shin et al. [87] studied how the length of ZnO nanowires (diameter 50 nm) grown on Au/Ti/Si from gas phase affects the SALDI performance of certain small drug molecules. It has been shown that such tar gets demonstrate SALDI activity only in a certain interval of nanowire lengths with maximum efficiency in the 250 nm range. The authors explain the existence of this interval by the existence of limitations in the

efficient heat energy transfer to the analyte and its thermal desorption, when the nanowire length is lower than a certain value; when the nanowire length is higher than other certain value, a physical barrier is formed that obstructs efficient irradiation of the sur face and the release of desorbed molecules (ions). Nanostructured carbon coatings. It has been shown [88] that a coating containing carbon nanotubes is promising in SALDI mass spectrometry—a detection limit level of 10 fmol was achieved in analyzing small peptides. The steel supports, used in this study, with an aligned vertical nanotube layer (diameter 100 nm, length 4 μm) grown by gasphase chemical plasma deposition did not lose SALDI activity during longterm storage in air. Nanostructured diamond like 300nmthick carbon coatings applied to the sur face of a DVD by pulsed laser deposition have shown high SALDI efficiency and versatility in the analysis of various aminoacids, carbohydrates, lipids, peptides, and other small molecules [89]. The detection limit of small peptides and carbohydrates was at 10 and 1 fmol level, respectively, and the coatings were characterized by high stability and tolerance to salts. The authors refer the SALDI activity of such coatings to the pres ence of a large amount of vacancies, defects, and dan gling bonds. Various approaches to deriving nanocrys talline diamond coatings for selective extraction and analysis of appropriate molecules are described in detail in [90]. 5.3. Desorption/Ionization on the Surface of Micro and Nanoparticles In SALDI mass spectrometry, the role of nanopar ticles lies not only (and in most cases not so much) in the absorption of laser energy and its efficient transfer to the analyte. Their use ensures a low level of back ground noise in a low mass range, simplicity of sample preparation, flexibility in the sample application tech nique, and the possibility of achieving high selectivity, sensitivity, and reproducibility of the analysis due to the appropriate surface chemistry and morphology. A detailed review of the works up to 2007 inclusive devoted to studies of various nanoparticles as affinity probes (analyte sorbents) in LDI mass spectrometry is given in [91], which has analyzed the possibilities of applying carbon nanotubes and fullerenes, nanoparti cles of diamond, gold, silicon, silicon and titanium dioxide, and magnetic nanoparticles (unmodified, functionalized, and covered with various coatings) to analyze various biological molecules. In this section, we will consider subsequent works devoted mainly to the application of various nanoparticles in smallmol ecule analysis by SALDI mass spectrometry. Metal nanoparticles. Among metal nanoparticles, gold nanoparticles are the most extensively used in SALDI mass spectrometry. Their advantage lies in simplicity of synthesis and chemical modification, sta bility, and high absorption coefficients [91]. Au nano

JOURNAL OF ANALYTICAL CHEMISTRY

Vol. 66

No. 13

2011

SMALLMOLECULE ANALYSIS

1235

Table 1. Sensitivity of analysis of some small molecules by SALDI mass spectrometry using Au, Ag, and FePtCu nanopar ticles Particles/size, nm Au/14 Au/14 Au/10–14 Au/12–14

Citrate + Nile red Citrate Uncoated Citrate

Au/21–22 Au/12–15

CHCA 33Meric DNA oligonucle otides 4Mercaptobenzoic acid Ag + citrate

Au/14 Au⎯Ag/34–44

Analyte

Detection limit, fmol

Refer ence

Glutathione, cysteine, homocysteine Glutathione, cysteine, homocysteine Ribose, glucose, cellobiose, maltose Ribose, glucose, maltose, testoster one, progesterone Angiotensin Adenosine triphosphate

25, 54, 34 100, 2000, 500 82, 41, 144, 151 173, 203, 365, 188, 399

92 93 94 95

<20000 2100 (480 – with Au matrix) 1000 3, 30, 25, 15

96 98 99 100

2230, 230, 2110 50000 5

101 102 104

Coating

Captopril Paramomycin, neomycin, kanamycin A, gentamycin Ag/31–37 Citrate Estrone, estradiol, estriol Ag/3–4 Alkylcarboxylate, alkylamine Palmitic acid FePtCu/20–100 Uncoated Angiotensin I

particles are obtained by reduction of chloroaurates – (Au Cl 4 ) with trisodium citrate, and their size can be easily controlled by the amount of the citrate added [92]. Functionalized Au nanoparticles of various size are efficient in the qualitative and quantitative analysis of some aminothiols [92, 93], small carbohydrates [94, 95], steroids [95], peptides [96, 97], adenosine triphosphate [98], and captopril [99]. Quantitative analysis of aminoglycosides in human blood plasma was performed with silvercoated gold nanoparticles [100]. Silver nanoparticles are efficient in analyzing estrogens [101], fatty acids [102], and olefines [97, 103]. It was reported [104] that FePtCu nanoparticles obtained via reduction of metal salts in an aqueous solution using hydrazine are highly efficient ion emit ters for analyzing small peptides. It was shown that surface modification of such particles with sulfonate groups makes it possible to use them as affinity traps for detecting certain oligopeptides in mixtures, as well as lysozyme in human blood serum. The data on SALDI mass spectrometric analysis of some small molecules using Au, Ag, and FePtCu nanoparticles are given in Table 1. Oxide particles. Certain metal oxides can be used as ion emitters for smallmolecule analysis. ZnO nanoparticles of various forms and sizes (20–200 nm) have shown high SALDI activity in the analysis of veparamil and testosterone, as well as some oligosac charides, lipids, and small synthetic polymers with a sensitivity comparable to the detection limits when the conventional MALDI method is used [105]. Addition of citrate salts or immobilization of polyacrylic acid on the surface of iron oxide particles substantially increases their SALDI activity, which has been shown for the analysis of bradykinin, mellittin, and insulin JOURNAL OF ANALYTICAL CHEMISTRY

Vol. 66

[106]. Using the example of polyethylene glycol degra dation, it was shown [107] that direct LDI mass spec trometric analysis of some additives or contaminants can be performed owing to their selective photocata lytic degradation using TiO2 nanoparticles modified with tetraethoxysilane. Porous oxide particles obtained via template syn thesis are also found to be efficient in smallmolecule SALDI processes. Mesoporous WO3–TiO2 (pore diameter ~10 nm, specific pore volume ~0.3 cm3/g, specific surface area 150 m2/g) was successfully used in the analysis of gramicidin [108], and mesoporous sili cate (SiO2 : Al2O3 = 25 : 1, specific pore volume ~1 cm3/g, specific surface area 818 m2/g) has demon strated high SALDI efficiency in the analysis of ami noacids (the detection limit of arginine is at the 10 pmol level, which is comparable to the MALDI sensitivity of 500 fmol) [109]. Silica with grafted fullerenes (pore size 30 nm) was efficient in the analy sis of some small peptides, carbohydrates, bile acids, steroids, phospholipids and aminoacids with a detec tion limit in the low picomole range [110]. The possi bility of using the porous silica gel and aluminum oxide particles with pore size 60 and 90 Å, respectively, in SALDI mass spectrometric analysis of peptides was studied in [111]. It has been shown that successful analysis is defined not only by pore size but also by the presence of residual solvent providing “wet” environ ment for the trapped analyte molecules. Carbon particles and nanotubes. High efficiency of the use of liquid substances (liquid matrices) in com bination with graphite for LDI mass spectrometric analysis of peptides and oligosaccharides has been shown in 1990s [11, 112]. Later, this approach was realized on planar silica surfaces using ultrafine carbon No. 13

2011

1236

KUZEMA

powder for the analysis of such small molecules as methylenephedrin and citosine [113]. More recently, colloid graphite with submicron particles size was suc cessfully used in SALDI mass spectrometric analysis of various plant metabolites (phospholipids, cerebro sides, oligosaccharides and flavonoids) [114], as well as powdered carbon aerogels (particle size 10– 30 nm)—for quantitative determination of free fatty acids in seed extracts (detection limit—~3 ng) [115]. Graphitized carbon soot particles were used for solid phase extraction of some organophosphate esters, sul fanilamides, and drugs with a subsequent successful their screening by SALDI mass spectrometry [116], whereas oxidized carbon soot nanoparticles, owing to presence of greater amount of carboxyl groups at their surface, are found to be more efficient ion emitters in the analysis of drug propanolol as compared to unoxi dized soot [117]. The authors of [118] have performed successful mass spectrometric analysis of 10 endoge nous steroids (detection limit 1.57–18.1 ng) using activated carbon particles (40–150 μm). Nanopowder containing carbon soot particles with hydrophobic sil ica coverage was used as an efficient chalking agent for fingerprints revelation and further enhancing agent in SALDI mass spectrometric analysis of their compo nents—some drugs and their metabolites [119], as well as nicotine and cotinine [120]. Carbon nanotubes are widely used in smallmole cule analysis by SALDI mass spectrometry. Studies of some small pharmaceutical molecules have shown that, in comparison to activated carbon, carbon nano tubes are more efficient in SALDI processes since they are highly hydrophobic, highly UV absorptive, and can be used as both adsorbents in solidphase extrac tion and efficient “transmitters” of absorbed energy to the analyte [121]. The advantages of using oxidized nanotubes (simplicity in sample preparation, forma tion of a more uniform and compact layer, better LDI efficiency, and reproducibility) over unoxidized nano tubes and conventional MALDI matrices are shown in [122] using quantitative analysis of some aminoacids in mixtures. Oxidized nanotubes were also used for quantitative monitoring of enzymatic activity of ace tylcholinesterase and for control of lowmolecular enzyme inhibitors using SALDI mass spectrometry [123]. Silicon nanoparticles. Silicon nanoparticles are efficient ion emitters in SALDI mass spectrometry of various small molecules. The laser energy necessary for desorption/ionization was found to be much lower than in MALDI and DIOS and close to that for silicon nanowires; this was shown by the analysis of certain drugs, peptides, pesticides, and acids using silicon nanoparticles (30 nm), which provided a sensitivity of propaphenone and veparamil determination at the low femtomole level [44]. It was also shown that in the presence of silicon nanoparticles (30 nm), the SALDI efficiency of nonvolatile organophosphates immobi lized on the surface of hydrophobic silica, diatomite,

alumina, polyethylene, talc, and corn starch particles increases by several orders (detection limit ~20 ng) [124]. Nanoparticles of other materials. Application of SiN nanoparticles for efficient desorption/ionization of some drug molecules was shown in [125]. Mesopo rous WTiO particles with aligned pore distribution (pore size 6 or 10 nm, specific surface area ~150 m2/g) [126], as well as ZnS nanoparticles (6–12 nm) modi fied with cysteine and doped with Mn2+ ions [127], were successfully used for the analysis of small pep tides (limit of gramicidin detection ~980 fmol). The authors of [128] studied the efficiency of ZnS nano particles with various coatings in SALDI mass spec trometric analysis of α, β, and γcyclodextrins. The detection limits at a level of 20–28, 26–34, 31–41, 37–49, and 40–55 fmol were determined respectively for 3mercaptopropane, citrate, cysteamine, 2mer captoethanesulfonate coverage, and unmodified ZnS particles. 6. MECHANISM OF SURFACEASSISTED LASER DESORPTION/IONIZATION AND FACTORS AFFECTING ITS EFFICIENCY At present, the most studied material in SALDI mass spectrometry is silicon; therefore, here, data on the mechanism of SALDI performance are addressed mainly to this material. The SALDI mechanism on nanostructured semiconductors is considered in detail in [48]. It may be tentatively divided into adsorption and retention of analyte by a surface, absorption of radiation by a surface and energy transfer to analyte, desorption of an analyte, and ionization reactions. Alimpiev et al. [64], based on studies of peptide ion formation on silicon supports obtained by various etching methods, as well as on amorphous films, sug gested a model of SALDI ion formation according to which (1) adsorption of neutral analyte molecules occurs via the formation of hydrogen bonds with sur face silanol groups; (2) electron excitation of support takes place under laser irradiation, as a result of which free electronhole pairs are formed (their relaxation leads to enrichment of the nearsurface layer in posi tive charges at sites with increased electron density (dangling bonds) causing an increase of silanol group acidity and proton transfer to analyte molecules); (3) the subsequent process is characterized by thermally activated dissociation of analyte ions from the surface, and activation energy of this process is ~70 kJ/mol. The authors of [64] also point out that both porous and nonporous silicon materials may possess SALDI activity; the internal excitation energy and the extent of fragmentation is higher in the case of IR in compar ison with UV laser application, the threshold of laser fluence necessary to achieve LDI being decreased in the sequence IR > visible range > UV; the presence of a highly disordered structure with high concentration

JOURNAL OF ANALYTICAL CHEMISTRY

Vol. 66

No. 13

2011

SMALLMOLECULE ANALYSIS

of dangling bonds at the nearsurface layer is the deter mining factor in SALDI performance. The factors affecting the DIOS efficiency were comprehensively studied in [129]. Investigations of tetra, hexa, and octadecylamines have shown that the DIOS process depends essentially on laser irradia tion energy and correlates with the formation of sur + face ions ( Si n and OSiH+); laser fluence threshold for molecular ion formation is 2.5 × 106 W/cm2 indepen dently of analyte volatility, which indicates that the key role in DIOS is played not only by the thermal process, but namely by surface restructuring; surface wettabil ity is not absolutely essential for producing the signal in DIOS; analyte desorption is mainly ensured by the high specific surface area, and the porous silicon sur face containing hydride groups is itself a source of pro tons for analyte ionization. The silicon surface during laser desorption/ioniza tion may be a donor of not only protons but also elec trons. This conclusion was drawn by the authors of [130] observed the reduction of copper(II) chloride and riboflavin under the conditions of DIOS experi ment. Besides, in this work it was shown that solvent (water) may be a proton source in DIOS process. Later it has been confirmed with the results obtained by the authors of [131]. By an example of triethylamine, they have shown that the main proton source in the DIOS process is local chemical environment of the analyte including the residual solvent and surface functional groups. These results indicate that the DIOS perfor mance could be improved by directed modification of substrate surface with functional groups having low pKa values and by choosing the analyte solvents with low pKa and vapor pressure values. The authors of [73], who observed a sevenfold decrease in laser fluence threshold necessary for de sorption/ionization in the case when silicon nanow ires are used, in comparison to DIOS and MALDI, explain it by the existence of the nearfield optical effect. Under laser irradiation, silicon nanowires act as tiny antennas generating significant field enhance ment in the vicinity of the nanowire spike. Thus, the laser energy is effectively focused on small area (com mensurable with the cross section of silicon nano wires). As a result, the focusing effect may cause the effect of field desorption of deposited molecules into the gas phase. The field enhancement phenomenon is similar to the light focusing effect observed in aper tureless nearfield scanning microscopy. The analysis of SALDI efficiency of unmodified and alkylfluorosilylated silicon nanoparticles in com parison with MALDI (using CHCA) was performed by the authors of [132], who calculated the internal energy deposition of molecular ions of benzylpyridin ium salts forming under LDI and studied the depen dence of the output of these ions on the laser energy. It has been shown that modified silicon nanoparticles are more efficient in SALDI processes because al JOURNAL OF ANALYTICAL CHEMISTRY

Vol. 66

1237

kylfluorosilylation of silicon nanoparticle surfaces decreases the value of internal energy deposition (the longer the hydrocarbon chain, the lower this value: at C10 it already becomes lower than in the case of MALDI) and decreases the laser energy threshold when molecular ion detection becomes possible. Grafted surface groups with longer hydrocarbon chains having more degrees of freedom are better energy distributors providing “softer” desorption/ion ization. This makes it possible to study the kinetics of smallmolecule fast dissociation in carrying out the experiments on surfaceinduced dissociation in com bination with SALDI and peak form analysis [133]. 7. COMPARATIVE EFFICIENCY OF VARIOUS MATERIALS AND SURFACES IN SMALL MOLECULE ANALYSIS BY SURFACE ASSISTED LASER DESORPTION/IONIZATION MASS SPECTROMETRY Comparative studies on the efficiency of particles of various materials and size in the analysis of tetracy clineseries antibiotics by SALDI mass spectrometry were performed by the authors of [134]. Suspensions of Co (20 nm), TiN (36 nm), TiO2 (1 μm), Si (45 μm), and graphite (1–2 μm) particles in an alcohol solution of ethylene glycol were deposited onto a silica gel TLC strip attached to a modified MALDI target. The lowest background noise and the highest signal intensity in the mass spectrum were obtained with graphite parti cles. The performance of various semiconductor nanowires in SALDI mass spectrometric analysis of some small molecules was studied in [46]. Monocrys talline ZnO, SnO2, GaN, and SiC nanowires grown on a silicon substrate by chemical vapor deposition had a diameter of 50–100 nm and a length of tens of microns. Two approaches for sample preparation were used: (1) the sample was deposited onto the substrate with grown nanowires; (2) synthesized nanowires were removed from the substrate and added to the suspen sion with analyte before deposition onto the conven tional MALDI target. The diameter of sample “spots” in both cases was less than 200 μm. Mass spectra of the test compound (leucineenkephaline, M = 555 Da) for all the nanowire suspensions studied were “pure” and contained mainly MNa+ and MK+ ions. The laser fluence threshold necessary for the desorption/ioniza tion was 20, 24, 30, and 40% of 100 kW/pulse for ZnO, SiC, SnO2, and GaN, respectively. ZnO and SiC nanowires have also shown better uniformity in des orption/ionization characteristics and a lower response time. In quantitative measurements of the test compound using ZnO nanowires, it was shown that the method of sample deposition onto the sub strate with grown nanowires gives better reproducibil ity and linearity in comparison with “suspension” approach for sample preparation. No. 13

2011

1238

KUZEMA

Table 2. Efficiency of various nanoparticles used in analysis of peptides by SALDI mass spectrometry Nanomaterial/diameter, nm Au nanoparticles/12–16 TiO2 nanoparticles/4–6 Se nanoparticles/90–110 CdTe quantum dots/~3 Fe3O4 nanoparticles/10–16 Pt nanoporous particles/31–43

Analyte

Laser power, kW Detection limit, fmol Optimal mass range, Da

Glutathione Angiotensin Glutathione Angiotensin Glutathione Angiotensin Glutathione Angiotensin Glutathione Angiotensin Glutathione Angiotensin

Studies on the efficiency of Ag (10–30 nm), Au (~20 nm), Cu (10–20 nm), and Pt (~2 nm) nanopar ticles obtained by laser ablation as ion emitters in the analysis of peptides were performed in [135] using angiotensin (M = 1296.5 Da) as an example. Ioniza tion with Ag and Cu nanoparticles was not achieved even at the maximal power of the instrumental laser. Pt nanoparticles demonstrated the highest performance, which, in authors’ opinion, is related to their lower heat conductivity and the highest melting tempera ture. The authors of [136] performed comparative analysis of the efficiency of various nanoparticles in the analysis of glutathione (M = 307.3 Da), angio tensin, and other peptides. The results (for the mass range up to 1500 Da) are given in Table 2. The influence of physicochemical characteristics of various silicon substrates obtained by etching in the iodinecontaining solvent [62] on their efficiency as SALDI targets was studied by Law [137] by analysis of some small organic molecules in mixtures. It was shown that not so much the porosity, but the density of surface nanostructures and the depth of the rough layer are important for efficient desorption/ionization on silicon; the dependence of the intensity of the molecular ion current on the laser irradiation energy has at least two maxima (the smaller at 15, and the larger at about 40% of the instrumental laser power), attributed by the author to several energy transfer mechanisms in the SALDI process—electron and thermal, respectively, the latter being accompanied by surface restructuring. In general, these substrates, ini tially optimized for analysis with the gasphase sample deposition in an airfree atmosphere, were found to be not ideal for analysis of mixtures with “wet” sample deposition, mainly because the surface obtained after the etching is contaminated with atmospheric air impurities, which adversely affect the activity in SALDI processes. However, purposeful chemical

14.6 14.6 18.3 18.3 14.6 15.8 14.6 15.8 15.8 15.8 14.6 14.6

140 810 2200 83 33000 160 190 11 8300 98 23000 43

300–1200 1200–1500 1200–1500 300–1500 1200–1500 1200–1500

modification of a clean (purified) surface of silicon substrates, obtained by electrochemical etching in iodinecontaining electrolyte and having a relatively deep and dense nanostructured layer, makes it possible to significantly increase their performance and to make them suitable for analysis of complex mixtures of biological molecules in solution. This was confirmed by data [137] on analyzing a bacterial extract by SALDI mass spectrometry. Commercial substrates DIOS™ and QuickMass™ were also studied by Law [138]. Silicon substrate DIOS™ was characterized by a deep porous 20–200nmthick layer and a high con centration of fluorocarbon and silicon oxides, whereas QuickMass™ had a rough nonporous germanium ~100nmthick covering on stainless steel. Both tar gets have shown good quality of the mass spectra of a large variety of pharmaceuticals without background peaks and surface impurities. In contrast to DIOS™, which, owing to the extended fluorosilylated surface, ensured high ionization efficiency with a detection limit in the femtomole range, QuickMass™ demon strated much lower sensitivity and versatility, as well as low efficiency in the analysis of peptides and complex mixtures. The main advantage of the QuickMass™ is high stability to surface contamination. Using as an example the analysis of human blood plasma, human urine, and the extract of rat liver tissue, it was shown that DIOS™ target can successfully be used to analyze complex biological mixtures by laser mass spectrome try. Based on the results of further investigations of the DIOS mechanism, Law concludes that, in contrast to silicon SALDI substrates described in [137], in DIOS™, the combined influence of electron and ther mal processes is observed. Owing to this, as well as to improvements due to the change in the chemical and morphological properties of the surface, DIOS™ sub strates have shown the best LDI performance among all matrixfree substrates studied by Law.

JOURNAL OF ANALYTICAL CHEMISTRY

Vol. 66

No. 13

2011

SMALLMOLECULE ANALYSIS

From an analysis of certain synthetic model pep tides containing different aminoacid sequences and having different lengths (mass range 500–3000 Da), comparative studies were performed on the SALDI efficiency of four surfaces representing different mate rials and morphologies: commercial DIOS™ and NALDI™ targets, carbon powder (from pencil lead), and porous silica gel for chromatography (particle size 63–200 μm, pore size 6 nm) [139]. Carbon was depos ited by scratching a pencil on a conventional MALDI steel plate, and silica gel first mixed in water and then in suspension form was also deposited on a MALDI target. As compared to carbon and silica gel, which demonstrated roughly the same efficiency both in the quantity of peptides determined (95%) and in the sen sitivity (detection limit 10 pmol), the DIOS™ sub strate is less efficient (63% of determined peptides with a detection limit of 1 nmol), and NALDI™, more efficient (100% of determined peptides with a detec tion limit of 100 fmol). The SALDI activity of silica gel was higher than that of carbon because, with the former, in the mass spectra of a larger number of ana lyzed peptides, aside from MK+ and MNa+ ions, MH+ ions were observed. 8. CONCLUSIONS The surface chemistry and morphology, as well as the sample preparation conditions, are vital for suc cessful smallmolecule analysis by SALDI mass spec trometry. The role of surface chemical modification in enhancing its activity in SALDI processes is not only the formation of active sites for selective entrapment and concentration of analyte molecules or ensuring protons (cations) for more efficient analyte ioniza tion, but also ensuring the optimal conditions for absorbed energy transfer to an analyte and its efficient desorption. At present, substrates containing silicon nanowires at the surface layer are the most efficient in respect to sensitivity and versatility for smallmolecule analysis. Nanostructured diamondlike carbon coat ings, microporous silica gel, and silicon nanoparticles are also attractive ion emitters. Quantum dots (CdTe) and nanoparticles containing several different metals (FePtCu) are promising as well. In general, the choice of ion emitter to use in analysis by SALDI mass spec trometry is defined by numerous factors and depends on the analyte and research objectives. REFERENCES 1. List of Accepted and Suggested Abbreviations, Rus sian and English, Related to Mass Spectrometry, Massspectrometria, 2009, vol. 6, no. 4, p. 315. 2. Honig, R.E. and Woolston, J.R., Appl. Phys. Lett., 1963, vol. 2, no. 7, p. 138. 3. Zakett, D., Schoen, A.E., Cooks, R.G., and Hem berger, P.H., J. Am. Chem. Soc., 1981, vol. 103, no. 5, p. 1295. JOURNAL OF ANALYTICAL CHEMISTRY

Vol. 66

1239

4. Cotter, R.J., Anal. Chem., 1984, vol. 56, no. 3, p. 485A. 5. Dass, C., Fundamentals of Contemporary Mass Spec trometry, New Jersey: Wiley, 2007. 6. Karas, M., Bachmann, D., and Hillenkamp, F., Anal. Chem., 1985, vol. 57, no. 14, p. 2935. 7. Karas, M. and Hillenkamp, F., Anal. Chem., 1988, vol. 60, no. 20, p. 2299. 8. Tanaka, K., Waki, H., Ido, Y., et al., Rapid Commun. Mass Spectrom., 1988, vol. 2, no. 8, p. 151. 9. Henderson, W. and McIndoe, J.S., Mass Spectrometry of Inorganic, Coordination and Organometallic Com pounds, Chichester: Wiley, 2005. 10. Cohen, L., Go, E.P., and Siuzdak, G., in MALDI MS: A Practical Guide to Instrumentation, Methods and Applications, Hillenkamp, F. and PeterKatalinic, J., Eds., Weinheim: Wiley, 2007, p. 299. 11. Sunner, J., Dratz, E., and Chen, Y.C., Anal. Chem., 1995, vol. 67, no. 23, p. 4335. 12. Hillenkamp, F. and Karas, M. in MALDI MS: A Prac tical Guide to Instrumentation, Methods and Applica tions, Hillenkamp, F. and PeterKatalinic, J., Eds., Weinheim: Wiley, 2007, p. 1. 13. Cohen, L.H. and Gusev, A.I., Anal. Bioanal. Chem., 2002, vol. 373, no. 7, p. 571. 14. Krause, J., Stoeckli, M., and Schlunegger, U.P., Rapid Commun. Mass Spectrom., 1996, vol. 10, no. 15, p. 1927. 15. Mims, D. and Hercules, D., Anal. Bioanal. Chem., 2004, vol. 378, no. 5, p. 1322. 16. OhnishiKameyama, M., Yanagida, A., Kanda, T., and Nagata, T., Rapid Commun. Mass Spectrom., 1997, vol. 11, no. 1, p. 31. 17. Wang, J. and Sporns, P., J. Agric. Food Chem., 2000, vol. 48, no. 12, p. 5887. 18. Jones, R.M., Lamb, J.H., and Lim, C.K., Rapid Com mun. Mass Spectrom., 1995, vol. 9, no. 10, p. 968. 19. Ayorinde, F.O., Hambright, P., Porter, T.N., and Keith, Q.L., Rapid Commun. Mass Spectrom., 1999, vol. 13, no. 24, p. 2474. 20. Wyatt, M.F., Havard, S., Stein, B.K., and Brenton, G., Rapid Commun. Mass Spectrom., 2008, vol. 22, no. 1, p. 11. 21. Zhang, S., Liu, J., Chen, Y., et al., J. Am. Soc. Mass Spectrom., 2010, vol. 21, no. 1, p. 154. 22. Peterson, D.S., Mass Spectrom. Rev., 2007, vol. 26, no. 1, p. 19. 23. Ramsey, B.G. and Bier, M.E., J. Organomet. Chem., 2005, vol. 690, no. 4, p. 962. 24. Xu, H., Yu, D., and Que, G., Fuel, 2005, vol. 84, no. 6, p. 647. 25. Chevrier, M.R., Ryan, A.E., Lee, D. Y.W., et al., Clin. Diagn. Lab. Immunol., 2005, vol. 12, no. 5, p. 575. 26. Bayerbach, R., Nguyen, V.D., Schurr, U., and Meier, D., J. Anal. Appl. Pyrolysis, 2006, vol. 77, no. 2, p. 95. 27. Apicella, B., Carpentieri, A., Alfe, M., et al., Proc. Combust. Inst., 2007, vol. 31, no. 1, p. 547. 28. Apicella, B., Alfè, M., Amoresano, A., Galano, E., Barbella, R., and Ciajolo, A., in Proc. 32nd Combustion Meeting (Naples, 2009), Ragucci, R., Ed., Naples: Naples Univ, 2009, p. I111, I117. No. 13

2011

1240

KUZEMA >

29. Fesenko, T., Laguta, I., Kuzema, P., and Stavinskaya, O., Materials Sci. (Medz iagotyra), 2010, vol. 16, no. 3, p. 272. 30. Duderstadt, R.E., Tsuie, B.M., Macha, S.F., and Lim bach, P.A., Anal. Chim. Acta, 2007, vol. 596, no. 1, p. 124. 31. Xia, D., Yu, S., Shen, R., et al., Dyes Pigm., 2008, vol. 78, no. 1, p. 84. 32. Spacil, Z., Shariatgorji, M., Amini, N., et al., Rapid Commun. Mass Spectrom., 2009, vol. 23, no. 12, p. 1834. 33. Exposito, J., Becker, C., Ruch, D., and Aubriet, F., Res. Lett. Phys. Chem., 2007, vol. 2007, p. 1. 34. Scholz, S., Corten, C., Walzer, K., et al., Org. Elec tron., 2007, vol. 8, no. 6, p. 709. 35. Haddrell, A.E., Feng, X., Nassar, R., et al., Aerosol Sci., 2005, vol. 36, no. 4, p. 521. 36. Shariatgorji, M., Spacil, Z., Maddalo, G., et al., Rapid Commun. Mass Spectrom., 2009, vol. 23, no. 23, p. 3655. 37. Mugo, S.M. and Bottaro, C.S., J. Mass Spectrom., 2007, vol. 42, no. 2, p. 206. 38. Hutchens, T.W. and Yip, T.T., Rapid Commun. Mass Spectrom., 1993, vol. no. 7, p. 576. 39. Weinberger, S.R., Lomas, L., Fung, E., and Ender wick, C., in Spectral Techniques in Proteomics, Sem, D.S., Ed., Boca Raton, FL: Taylor & Francis, 2007, p. 101. 40. SurfaceEnhanced Laser Desorption/Ionization. http://en.wikipedia.org/wiki/Surfaceenhanced_laser_ desorption/ionization. Accessed October 9, 2010. 41. Tang, N., Tornatore, P., and Weinberger, S.R., Mass Spectrom. Rev., 2004, vol. 23, no. 1, p. 34. 42. Lin, R., Viner, R., Kitagava, N., Chang, D., Tang, N., and Weinberger, S., in Proc. 51st ASMS Conference on Mass Spectrometry and Allied Topics (Montreal, 2003). 43. Wei, J., Buriak, J.M., and Siuzdak, G., Nature, 1999, vol. 399, no. 6733, p. 243. 44. Wen, X., Dagan, S., and Wysocki, V.H., Anal. Chem., 2007, vol. 79, no. 2, p. 434. 45. Cha, S.W. and Yeung, E.S., Anal. Chem., 2007, vol. 79, no. 6, p. 2373. 46. Kang, M.J., Pyun, J.C., Lee, J.C., et al., Rapid Commun. Mass Spectrom., 2005, vol. 19, no. 21, p. 3166. 47. IUPAC (2004) Project: Standard Definitions of Terms Relating to Mass Spectrometry. http://www.iupac.org/ web/ins/20030562500. Accessed October 10, 2010. 48. Law, K.P. and Larkin, J.R., Anal. Bioanal. Chem. (in press). doi: 10.1007/s0021601040633. 49. Kawasaki, H., Shimomae, Y., Watanabe, T., and Arakawa, R., Colloids Surf., A, 2009, vol. 347, nos. 1– 3, p. 220. 50. Okuno, S., Oka, K., and Arakawa, R., Anal. Sci., 2005, vol. 21, no. 12, p. 1449. 51. Lewis, W.G., Shen, Z., Finn, M.G., and Siuzdak, G., Int. J. Mass Spectrom., 2003, vol. 226, no. 1, p. 107. 52. Shen, Z., Thomas, J.J., Averbuj, C., et al., Anal. Chem., 2001, vol. 73, no. 3, p. 612.

53. Credo, G.M., Hewitson, H.B., Fountain, K.J., Gilar, M., Finch, J.W., Stumpf, C.L., Benevides, C.C., Bouvier, E.S.P., Compton, B.J., Shen, Z., and Siuzdak, G., in Proc. 51st ASMS Conference on Mass Spectrometry and Allied Topics (Montreal, 2003). 54. Go, E.P., Apon, J.V., Uritboonthai, W., Compton, B.J., Bouvier, E.S.P., Finn, M.G., Shen, Z., and Siuzdak, G., in Proc. 52nd ASMS Conference on Mass Spectrometry and Allied Topics (Nashville, 2004). 55. Trauger, S.A., Go, E.P., Shen, Z., et al., Anal. Chem., 2004, vol. 76, no. 15, p. 4484. 56. Shmigol, I.V., Alekseev, S.A., Lavrynenko, O.Yu., et al., J. Mass. Spectrom., 2009, vol. 44, no. 8, p. 1234. 57. Northen, T.R., Yanes, O., Northen M.T., et al., Nature, 2007, vol. 449, no. 7165, p. 1033. 58. Patti, G.J., Woo, H.K., Yanes, O., et al., Anal. Chem., 2010, vol. 82, no. 1, p. 121. 59. Liu, Q., Xiao, Y., PaganMiranda, C., et al., J. Am. Soc. Mass Spectrom., 2009, vol. 20, no. 1, p. 80. 60. Liu, Q. and He, L., J. Am. Soc. Mass Spectrom., 2009, vol. 20, no. 12, p. 2229. 61. Li, J., Lu, C., Hu, X.K., et al., Int. J. Mass Spectrom., 2009, vol. 285, no. 3, p. 137. 62. Alimpiev, S.S., Nikiforov, S.M., Grechnikov, A.A., Karavanskii, V.A., and Sunner, J.A., RF Patent 2217840, 2003. 63. Alimpiev, S., Grechnikov, A., Sunner, J., et al., Anal. Chem., 2009, vol. 81, no. 3, p. 1255. 64. Alimpiev, S., Grechnikov, A., Sunner, J., et al., J. Chem. Phys. 2008, vol. 128, no. 1, p. 014711. 65. Han, M. and Sunner, J., J. Am. Soc. Mass. Spectrom., 2000, vol. 11, no. 7, p. 644. 66. Kim, H.J., Lee, J.K., Park, S.J., et al., Anal. Chem., 2000, vol. 72, no. 22, p. 5673. 67. Kim, J., Paek, K., and Kang, W., Bull. Korean Chem. Soc., 2002, vol. 23, no. 2, p. 315. 68. Cuiffi, J.D., Hayes, D.J., Fonash, S.J., et al., Anal. Chem., 2001, vol. 73, no. 6, p. 1292. 69. Jokinen, V., Aura, S., Luosujärvi, L., et al., J. Am. Soc. Mass Spectrom., 2009, vol. 20, no. 9, p. 1723. 70. Grechnikov, A.A., Borodkov, A.S., Alimpiev, S.S., et al., Massspectrometria, 2010, vol. 7, no. 1, p. 53. 71. Grechnikov, A.A., Georgieva, V.V., Alimpiev, S.S., et al., J. Phys. Conf. Ser., 2010, vol. 223, no. 1, p. 012038. 72. MatrixLess MALDI Now Available for Fast, Cost Effective Analysis of Small Molecules; Shimadzu Bio thech. http://www.shimadzubiotech.net/pages/news/ 1/press_releases/2004_01_21matrixless.php. Accessed August 3, 2010. 73. Go, E.P., Apon, J.V., Luo, G., et al., Anal. Chem., 2005, vol. 77, no. 6, p. 1641. 74. Luo, G., Chen, Y., Daniels, H., et al., J. Phys. Chem. B, 2006, vol. 110, no. 27, p. 13381. 75. Daniels, R.H., Dikler, S., Li, E., and Stacey, C., J. Assoc. Lab. Automation, 2008, vol. 13, no. 6, p. 314. 76. Guenin, E., Lecouvey, M., and Hardouin, J., Rapid Commun. Mass Spectrom., 2009, vol. 23, no. 9, p. 1395. 77. Wyatt, M.F., Ding, S., Stein, B.K., et al., J. Am. Soc. Mass Spectrom., 2010, vol. 21, no. 7, p. 1256.

JOURNAL OF ANALYTICAL CHEMISTRY

Vol. 66

No. 13

2011

SMALLMOLECULE ANALYSIS 78. Yao, T., Kawasaki, H., Watanabe, T., and Arakawa, R., Int. J. Mass Spectrom., 2010, vol. 291, no. 3, p. 145. 79. Yan, H., Xu, N., Huang, W.Y., et al., Int. J. Mass Spectrom., 2009, vol. 281, no. 1, p. 1. 80. Hong, M., Qiu, F., Huang, L.L., and Zhu, J., J. Phys. Chem., C, 2008, vol. 112, no. 30, p. 11078. 81. Colaianni, L., Kung, S.C., Taggart, D.K., et al., Sens. Lett., 2010, vol. 8, no. 3, p. 539. 82. Oleinikov, V.A., Zagorski, D.L., Bedin, S.A., et al., Radiat. Meas., 2008, vol. 43, no. 1, p. S635. 83. GóreckaDrzazga, A., Bargiel, S., Walczak, R., et al., Sens. Actuators, B, 2004, vol. 103, nos. 1–2, p. 206. 84. Dattelbaum, A.M., Hicks, R.K., Shelley, J., et al., Microporous Mesoporous Mater., 2008, vol. 114, nos. 1–3, p. 193. 85. Fesenko, T.V., Kosevich, M.V., Surovtseva, N.I., et al., MassSpectrometria, 2007, vol. 4, no. 4, p. 289. 86. Lo, C.Y.,Lin, J.Y., Chen, W.Y., et al., J. Am. Soc. Massspectrometria, 2008, vol. 19, no. 7, no. 1014. 87. Shin, W.J., Shin, J.H., Song, J.Y., and Han, S.Y., J. Am. Soc. Mass Spectrom., 2010, vol. 21, no. 6, p. 989. 88. Shin, S.J., Choi, D.W., Kwak, H.S., et al., Bull. Korean Chem. Soc., 2006, vol. 27, no. 4, p. 581. 89. NajamulHaq, M., Rainer, M., Huck, C.W., et al., Anal. Chem., 2008, vol. 80, no. 19, p. 7467. 90. NajamulHaq, M., Rainer, M., Szabó, Z., et al., J. Biochem. Biophys. Methods, 2007, vol. 70, no. 2, p. 319. 91. Chiu, T.C., Huang, L.S., Lin, P.C., et al., Recent Pat. Nanotechnol., 2007, vol. 1, no. 2, p. 99. 92. Huang, Y.F. and Chang, H.T., Anal. Chem., 2006, vol. 78, no. 5, p. 1485. 93. Chiang, C.K., Lin, Y.W., Chen, W.T., and Chang, H.T., Nanomed. Nanotechnol. Biol. Med., 2010, vol. 6, no. 4, p. 530. 94. Su, C.L. and Tseng, W.L., Anal. Chem., 2007, vol. 79, no. 4, p. 1626. 95. Wu, H.P., Yu, C.J., Lin, C.Y., et al., J. Am. Soc. Mass Spectrom., 2009, vol. 20, no. 5, p. 875. 96. Duan, J., Linman, M.J., Chen, C.Y., and Cheng, Q.J., J. Am. Soc. Mass Spectrom., 2009, vol. 20, no. 8, p. 1530. 97. Castellana, E.T., Sherrod, S.D., and Russell, D.H., J. Assoc. Lab. Automation, 2008, vol. 13, no. 6, p. 330. 98. Huang, Y.F. and Chang, H.T., Anal. Chem., 2007, vol. 79, no. 13, p. 4852. 99. Chen, W.T., Chiang, C.K., Lin, Y.W., and Chang, H.T., J. Am. Soc. Mass Spectrom., 2010, vol. 21, no. 5, p. 864. 100. Wang, M.T., Liu, M.H., Wang, C.R.C., and Chang, S.Y., J. Am. Soc. Mass Spectrom., 2009, vol. 20, no. 10, p. 1925. 101. Chiu, T.C., Chang, L.C., Chiang, C.K., and Chang, H.T., J. Am. Soc. Mass Spectrom., 2008, vol. 19, no. 9, p. 1343. JOURNAL OF ANALYTICAL CHEMISTRY

Vol. 66

1241

102. Hayasaka, T., GotoInoue, N., Zaima, N., et al., J. Am. Soc. Mass Spectrom., 2010, vol. 21, no. 8, p. 1446. 103. Sherrod, S.D., Diaz, A.J., Russell, W.K., et al., Anal. Chem., 2008, vol. 80, no. 17, p. 6796. 104. Kawasaki, H., Akira, T., Watanabe, T., et al., Anal. Bio anal. Chem., 2009, vol. 395, no. 5, p. 1423. 105. Watanabe, T., Kawasaki, H., Yonezawa, T., and Arakawa, R., J. Mass Spectrom., 2008, vol. 43, no. 8, p. 1063. 106. Chiu, Y.C. and Chen, Y.C., Anal. Lett., 2008, vol. 41, no. 2, p. 260. 107. Watanabe, T., Okumura, K., Nozaki, K., et al., Rapid Commun. Mass Spectrom., 2009, vol. 23, no. 23, p. 3886. 108. Yuan, M., Shan, Z., Tian, B., et al., Microporous Mesoporous Mater., 2005, vol. 78, no. 1, p. 37. 109. Lee, C.S., Kang, K.K., Kim, J.H., et al., Microporous Mesoporous Mater., 2007, vol. 98, nos. 1– 3, p. 200. 110. Szabo, Z., Vallant, R.M., Takatsy, A., et al., J. Mass. Spectrom., 2010, vol. 45, no. 5, p. 545. 111. Shenar, N., Martinez, J., and Enjalbal, C., J. Am. Soc. Mass Spectrom., 2008, vol. 19, no. 5, p. 632. 112. Dale, M.J., Knochenmuss, R., and Zenobi, R., Anal. Chem., 1996, vol. 68, no. 19, p. 3321. 113. Chen, Y.C. and Wu, J.Y., Rapid Commun. Mass Spectrom., 2001, vol. 15, no. 20, p. 1899. 114. Cha, S., Zhang, H., Ilarslan, H.I., et al., Plant J., 2008, vol. 55, no. 2, p. 348. 115. Borissova, M., Palk, K., and Vaher, M., Procedia Chemistry, 2010, vol. 2, no. 1, p. 174. 116. Shariatgorji, M., Amini, N., Thorsén, G., et al., Anal. Chem., 2008, vol. 80, no. 14, p. 5515. 117. Amini, N., Shariatgorji, M., and Thorsén, G., J. Am. Soc. Mass Spectrom., 2009, vol. 20, no. 6, p. 1207. 118. Guild, G.E., Lenehan, C.E., and Walker, G.S., Int. J. Mass Spectrom., 2010, vol. 294, no. 1, p. 16. 119. Rowell, F., Hudson, K., and Seviour, J., Analyst, 2009, vol. 134, no. 4, p. 701. 120. Benton, M., Chua, M.J., Gu, F., et al., Forensic Sci. Int., 2010, vol. 200, nos. 1–3, p. 28. 121. Pan, C., Xu, S., Zou, H., et al., J. Am. Soc. Mass Spec trom., 2005, vol. 16, no. 2, p. 263. 122. Pan, C., Xu, S., Hu, L., et al., J. Am. Soc. Mass Spec trom., 2005, vol. 16, no. 6, p. 883. 123. Hu, L. and Jiang, G., J. Am. Soc. Mass Spectrom., 2006, vol. 17, no. 11, p. 1616. 124. Hagan, N.A., Cornish, T.J., Pilato, R.S., et al., Int. J. Mass Spectrom., 2008, vol. 278, nos. 2–3, p. 158. 125. Shariatgorji, M., Amini, N., and Ilag, L.L., J. Nano part. Res., 2009, vol. 11, no. 6, p. 1509. 126. Shan, Z., Han, L., Yuan, M., et al., Anal. Chim. Acta, 2007, vol. 593, no. 1, p. 13. 127. Kailasa, S.K. and Wu, H.F., Analyst, 2010, vol. 135, no. 5, p. 1115. No. 13

2011

1242

KUZEMA

128. Kailasa, S.K., Kiran, K., and Wu, H.F., Anal. Chem., 2008, vol. 80, no. 24, p. 9681. 129. Northen, T.R., Woo, H.K., Northen, M.T., et al., J. Am. Soc. Mass Spectrom., 2007, vol. 18, no. 11, p. 1945. 130. Okuno, S., Arakawa, R., and Wada, Y., J. Mass Spec trom. Soc. Jpn., 2004, vol. 52, no. 1, p. 13. 131. Liu, Q. and He, L., J. Am. Soc. Mass Spectrom., 2008, vol. 19, no. 1, p. 8. 132. Dagan, S., Hua, Y., Boday, D.J., et al., Int. J. Mass Spectrom., 2009, vol. 283, nos. 1–3, p. 200. 133. Yoon, S.H., Gamage, C.M., Gillig, K.J., and Wysocki, V.H., J. Am. Soc. Mass Spectrom., 2009, vol. 20, no. 6, p. 957.

134. Crecelius, A., Clench, M.R., Richards, D.S., and Parr, V., J. Chromatogr. A, 2002, vol. 958, nos. 1–2, p. 249. 135. Yonezawa, T., Kawasaki, H., Akira, T., et al., Anal. Sci., 2009, vol. 25, no. 3, p. 339. 136. Chiang, C.K., Chiang, N.C., Lin, Z.H., et al., J. Am. Soc. Mass Spectrom., 2010, vol. 21, no. 7, p. 1204 137. Law, K.P., Int. J. Mass Spectrom., 2010, vol. 290, no. 1, p. 47. 138. Law, K.P., Int. J. Mass Spectrom., 2010, vol. 290, no. 1, p. 72. 139. Shenar, N., Cantel, S., Martinez, J., and Enjalbal, C., Rapid Commun. Mass Spectrom., 2009, vol. 23, no. 15, p. 2371.

JOURNAL OF ANALYTICAL CHEMISTRY

Vol. 66

No. 13

2011