Article pubs.acs.org/Langmuir
Mechanistic Study of Laser Desorption/Ionization of Small Molecules on Graphene Oxide Multilayer Films Young-Kwan Kim†,‡ and Dal-Hee Min*,†,‡ †
Center for RNA Research, Institute for Basic Science, and ‡Department of Chemistry, Seoul National University, Seoul 151-747, Republic of Korea S Supporting Information *
ABSTRACT: Graphene and graphene oxide (GO) films have been explored to develop an efficient laser desorption/ ionization mass spectrometry (LDI-MS) platform for the analysis of chemically and biologically important small molecules. The GO films were prepared by layer-by-layer (LBL) assembly cycles (one to ten layers) with precisely controlled thickness and surface roughness which are important structural factors for laser energy absorption capacity and laser energy transfer for efficient LDI-MS analysis. Amino acids, saccharides, and pyrenylated molecules were analyzed by LDI-MS on the LBL assembled GO films to reveal their structural influence on LDI-MS analysis of small molecules. Then, the structural influence of LBL assembled GO films on synergistic effect was investigated to develop an efficient and widely applicable LDI-MS analysis platform with an additional multiwalled carbon nanotube (MWCNT) layer. We found that the optimum number of GO film layers for LDI-MS analysis was dependent on the chemical structures of small molecules, and the laser energy threshold needed for LDI of small molecules on GO/MWCNT films could be lowered as the number of LBL assembled GO films increased underneath the MWCNT layer.
■
mechanical durability.10 Recently, our group reported that LDIMS efficiency of GO/MWCNT hybrid films could be further enhanced by layer-by-layer (LBL) assembly of GO and MWCNT which lead to increase of thickness, surface roughness, and laser energy absorption capacity of GO/ MWCNT hybrid films.11 The previous studies showed that the structural factors of GO/MWCNT hybrid films have great influence on LDI-MS efficiency and thus need to be systematically investigated to clearly understand the LDI process on their surfaces for the development of an efficient LDI-MS platform. Although GO films can also clearly provide basic information about LDI process based on the fine-tuning laser energy absorption capacity and flat morphology compared to GO/MWCNT hybrid films, their detailed structural influence has been not studied yet because of their lower LDI efficiency than GO/MWCNT hybrid films. The precise control of laser energy absorption capacity without much change of surface morphology could be a very attractive strategy to reveal the detailed LDI process of various small molecules depending on their chemical structures. Here, we prepared GO films by using the LBL assembly technique to precisely control their structural factors such as thickness, surface roughness, and laser energy absorption capacity12−14 and investigated the structural influence of GO films on LDI efficiency of small molecules. To the best of our knowledge, the present study is the first example of structural
INTRODUCTION Matrix-assisted laser desorption/ionization mass spectrometry (MALDI-MS) has been considered one of the most important analytical tools for oncology,1 chemical biology,2 and proteomics3 because it provides intact molecular weight of analytes with high sensitivity and resolution. The conventional organic matrix molecules induce soft ionization of large biomolecules (1−100 kDa) without much fragmentation, but they give significant signal interference in the low mass region derived from matrix ions.4 Therefore, MALDI-MS has been not actively applied to analysis of chemically and biologically important small molecules. To solve this problem, many efforts have been devoted to develop an efficient laser desorption/ ionization mass spectrometry (LDI-MS) platform by utilizing various nanostructures as a matrix instead of conventional organic matrices.5 Recently, carbon nanomaterials such as fullerene,6 carbon nanotube (CNT),7 graphene,8 and graphene oxide (GO)9 have been extensively investigated because they have many advantages as a matrix for laser desorption/ ionization mass spectrometry (LDI-MS) such as negligible interference in low mass region, high extinction in UV light, efficient energy transfer to analyte, and affinity to various small molecules. Our group previously demonstrated that the LDI-MS efficiency of small molecules was greatly enhanced on GO films by incorporation of multiwalled CNT (MWCNT).10 The GO/MWCNT hybrid films showed excellent performance as LDI-MS platforms for analysis of enzyme activity, small molecules, and mouse brain tissue mass imaging because of negligible interference, no sweet-spot, high salt tolerance, and © XXXX American Chemical Society
Received: July 17, 2014 Revised: September 6, 2014
A
dx.doi.org/10.1021/la5027653 | Langmuir XXXX, XXX, XXX−XXX
Langmuir
Article
Scheme 1. Fabrication of GO(1−10) and GO(1−10)/MWCNT-NH2 Films by LBL Assembly Process and Their Applications to LDIMS Analysis of Small Molecules for Investigation of Structural Influence of GO Films on the LDI-MS Process of Small Molecules
influence of GO films on LDI-MS behavior of small molecules by utilizing the LBL assembly process to control structures of GO films with single layer precision. To precisely control laser energy absorption capacity without much alteration in surface morphology, poly(allylamine hydrochloride) (PAAH) was employed instead of MWCNT-NH2 as a positively charged species for electrostatic LBL assembly of negatively charged GO sheets. The prepared GO(1−10) films, where the subscript number indicates the number of LBL assembly cycles, were applied to LDI-MS analysis of various small molecules having different chemical structures such as amino acids, saccharides, and pyrenylated molecules to reveal their structural influence on LDI-MS efficiency. Then, the GO(1−10) films were respectively harnessed as supports for incorporation of aminated MWCNT (MWCNT-NH2) to investigate the structural influence of GO films on synergistic effect of MWCNT-NH2 for LDI-MS analysis (Scheme 1). By applying GO(1−10) and GO(1−10)/MWCNT-NH films to LDI-MS analysis of various small molecules, the structural influence of GO films on LDI efficiency with synergistic effect of MWCNTNH2 incorporation was systematically investigated and demonstrated on the basis of thorough structural characterization of carbon nanomaterial films and chemical structures of small molecules.
■
Preparation of GO Films by LBL Assembly Process on a Substrate. APTES-treated glass substrates were immersed in aqueous GO suspension (1 mg/mL) for 1 h, washed with water and ethanol, and dried under a stream of nitrogen. This process resulted in formation of GO films with high surface coverage. The GO film coated glass substrates were then immersed in aqueous PAAH solution (1 mg/mL, pH 7.5) for 30 min, washed with water and ethanol, and dried under a stream of nitrogen. The immersing processes in GO suspension and PAAH solution were repeated up to 10 cycles to prepare GO(1−10) films with precisely controlled thickness and surface roughness; the subscript numbers indicate the applied LBL assembly cycles. Incorporation of MWCNT on LBL Assembled GO Films. The prepared GO(1−10) films were respectively immersed in aqueous MWCNT-NH2 suspension (120 μg/mL) for 1 h, washed with water and ethanol, dried under a stream of nitrogen, and baked at 150 °C for 15 min under a continuous stream of nitrogen. Synthesis of Benzylpyridinium Salt (BP). 12 mL of pyridine was reacted with benzyl chloride at a molar ratio 20:1 (pyridine/benzyl chloride) at 60 °C for 6 h with refluxing. The BP was collected by removal of excess pyridine with rotary vacuum evaporation. LDI-MS Analysis of Small Molecules. Small molecules such as cellobiose, Leu-enkephalin, glucose, lysine, leucine, and phenylalanine were dissolved in water at 1 nmol/μL by using a vortex. 1 μL of the prepared small molecule solutions was deposited on GO(1−10) and GO(1−10)/MWCNT-NH2 films, dried under ambient conditions, and subjected to LDI-MS analysis with 76.8 μJ laser power using a manual mode (this laser power was constantly applied to LDI-MS analysis unless otherwise notified). This sample preparation process was applied to monitoring LDI characteristics of BP as a thermometer molecule. For analysis of pyrene derivatives, GO(1−10) films were immersed in 1 mM ethanolic solution of pyrene derivatives for 12 h, washed with water and ethanol, dried under a stream of nitrogen, and subjected to LDI-MS analysis. Characterization. The atomic force microscopy (AFM) images, line profiles, and center-line average surface roughness of LBL assembled GO films with and without MWCNT-NH2 on silicon substrates were obtained with an XE-100 (Park System, Korea) with a backside gold-coated silicon SPM probe (M to N, Korea). All LDI-MS analyses on GO1-10 films with and without MWCNT-NH2 layer coated glass substrates were carried out using a Bruker Autoflex III (Bruker Daltonics, Germany) equipped with a Smartbeam laser (Nd:YAG, 355 nm, 120 μJ, 100 Hz, 50 μm of spot diameter at target plate) in a positive reflection mode. The accelerating voltage was 19 kV, and all spectra were obtained by averaging 500 laser shots with 76.8 μJ laser power unless otherwise indicated. The UV−vis spectra of GO1-10 films on quartz substrates were recorded with a UV-2550 (Shimadzu, Japan). Ellipsometric analysis was carried out with a L116S (Gaertner Scientific Corp., USA). Raman characterization was carried out by LabRAM HR UV/vis/NIR (Horiba Jobin Yvon, France) using an Ar ion CW laser (514.5 nm) as an excitation source focused through a BXFM confocal microscope equipped with an objective (50×, numerical aperture = 0.50). FT-IR spectra measurements of graphite oxide were performed with an EQUINOX55 (Bruker, Germany) using the KBr pellet method.
EXPERIMENTAL SECTION
Materials. Graphite (FP, 99.95% pure) was purchased from Graphit Kropfmühl AG (Hauzenberg, Germany). MWCNT (diameter in 15 nm and length 20 μm) was purchased from Nanolab (USA). Sodium nitrate and hydrogen peroxide (30% in water) were purchased from Junsei (Japan). Cellobiose, Leu-enkephalin, glucose, lysine, leucine, phenylalanine, potassium permanganate, 3-aminopropyltriethoxysilane (APTES), ethylenediamine, anhydrous dimethylformamide (DMF), and PAAH (MW = 15 000) were purchased from Sigma-Aldrich (St. Louis, MO). Nitric acid, sulfuric acid, and thionyl chloride were purchased from Samchun (Seoul, Korea). Ethanol was purchased from Merck (Darmstadt, Germany). The #2 glass coverslip (∼0.22 mm in thickness), P++ Si substrates (500 μm in thickness), and 4 in. quartz wafers (500 μm in thickness) were purchased from Warner (Hamden, USA), STC (Japan), and i-Nexus (Stamford, USA), respectively. All chemicals were used without further purification. Preparation of APTES Functionalized Substrates. The #2 glass coverslips were immersed in piranha solution (sulfuric acid:hydrogen peroxide (30%) = 3:1) (warning: piranha solution is highly explosive and corrosive) for 10 min at 125 °C, washed with water and ethanol, and dried under a stream of nitrogen. The substrates were immersed in a 10 mM toluene solution of APTES for 30 min, sonicated in toluene for 2 min, rinsed with ethanol and water, and dried under a stream of nitrogen. The substrates were then baked at 150 °C for 15 min under a continuous stream of nitrogen. This process is equally applied to APTES functionalization of the other substrates such as Si and quartz substrates. B
dx.doi.org/10.1021/la5027653 | Langmuir XXXX, XXX, XXX−XXX
Langmuir
Article
Figure 1. AFM images of (a) GO(1−10) and (b) GO(1−10)/MWCNT-NH2 films fabricated by the LBL assembly process and subsequent MWCNTNH2 incorporation.
■
immersing those films in aqueous MWCNT-NH2 suspension, and this process resulted in adsorption of MWCNT-NH2 (Figure 1b).10,11 The thickness of GO films almost linearly increased from 1.9 to 27.0 nm through 1−10 LBL assembly cycles and almost equally increased by 7.8 ± 1.1 nm after incorporation of MWCNT-NH2, respectively (Figure 2a). This result implied that MWCNT-NH2 was incorporated onto the GO (1−10) films with similar surface coverage. Therefore, we can describe that GO films showed little increase in CLA roughness from 1.34 to 2.88 nm by 1−10 LBL assembly cycles with similar surface morphology. The CLA roughness values of GO(1−10) films also almost equally increased by 4.9 ± 1.1 nm after incorporation of MWCNT-NH2, respectively (Figure 2b). These results indicated that thickness of GO films could be precisely controlled without much change of surface morphology by using LBL assembly process, and thus, the fabricated GO(1−10) films could be successfully harnessed as supports for subsequent incorporation of MWCNT-NH2 with similar surface coverage. The correlation between structural factors of GO(1−10) films and laser energy absorption capacity was investigated by measuring their optical absorbance. The UV−vis spectra of GO(1−10) and GO(1−10)/MWCNT-NH2 films showed typical absorption peak at 230 nm from the π−π* transition of the
RESULTS AND DISCUSSION GO was synthesized by the modified Hummers’ method15 and dispersed in water at 1 mg/mL by sonication for 1 h (Figure S1 in the Supporting Information). GO dispersed in water presented a negative zeta-potential (−42.1 mV) because of the presence of oxygen-containing functional groups on its basal plane and edge. The synthesized GO was immobilized on APTES-treated glass substrates by electrostatic assembly, and then, the GO(1) films were used to fabricate GO(1−10) films by the LBL assembly process.10,11 The GO(1) film was composed of 1−3 layered GO sheets with high surface coverage (1.9 nm in ellipsometric thickness).10,11 The structural changes of GO(1) films such as thickness and surface roughness with LBL assembly process were monitored by using ellipsometer and AFM. The thickness and surface roughness of carbon nanomaterial films are important structural factors for LDIMS analysis because they are closely related to laser energy absorption capacity and interfacing area with analytes for efficient laser energy transfer.10,11 AFM images of GO(1−10) films showed high surface coverage with few folded and rippled structures which gradually increased with 1−10 LBL assembly cycles (Figure 1a). The prepared GO(1−10) films were harnessed as supports for MWCNT-NH2 incorporation by simply C
dx.doi.org/10.1021/la5027653 | Langmuir XXXX, XXX, XXX−XXX
Langmuir
Article
Figure 2. Changes of (a) ellipsometric thickness and (b) CLA surface roughness of GO(1−10) and GO(1−10)/MWCNT-NH2 films. The UV−vis spectra of (c) GO(1−10) and (d) GO(1−10)/MWCNT-NH2 films and their absorbance at (e) 230 and (f) 355 nm.
aromatic C−C bond (Figure 2c,d),10,11 and the absorbance of GO(1−10) films at 230 nm gradually increased by 1−10 LBL assembly cycles and subsequent incorporation of MWCNTNH2 on the GO(1−10) films (Figure 2e,f). This result was well matched with the ellipsometric and AFM analysis data. Along with increase of absorbance at 230 nm, the absorbance at 355 nm corresponding to laser wavelength equipped in a mass spectrometer also gradually increased with incremental LBL assembly cycles (Figure 2e,f). This result indicated that the increased thickness of GO (1−10) films and subsequent incorporation of MWCNT-NH2 were well correlated with the increase of laser energy absorption capacity, and thus, the LDIMS efficiency of GO films could be improved with the LBL assembly of GO and subsequent MWCNT-NH2 incorporation processes. To investigate the structural influence of GO(1−10) films on LDI-MS efficiency of small molecules, 1 nmol of cellobiose, Leu-enkephaline, phenylalanine, and glucose was deposited on GO(1−10) films and subjected to LDI-MS analysis. There was no mass signal on GO(1) films probably due to low-energy
absorption capacity. This result well concurred with our previous reports (Figure 3a).10,11 Although the thickness, surface roughness, and laser energy absorption capacity increased with LBL assembly cycles, the mass signal of cellobiose, phenylalanine, and glucose was not detected on GO(1−10) films whereas the background interference significantly increased with LBL assembly cycles (Figure 3a and see Figure S2 for all mass spectra obtained on GO(1−10) films). By contrast, Leu-enkephaline started to be detected on GO(8) films as sodium and potassium adducts with low background interference (Leu-enkephalin: m/z 576 [M + Na]+ and m/z 592 [M + K]+), and the mass peak intensities augmented on GO(9−10) films as LBL assembly cycles increased (Figure 3a and Figure S2). The cation adducts might be originated from the residual salts from synthesis of GO.10,11 This result suggested that compounds having relatively high molecular weight required higher laser energy to be detected and generated less background interference in low mass region than compounds having smaller molecular weight. The low background interference of high molecular weight compounds D
dx.doi.org/10.1021/la5027653 | Langmuir XXXX, XXX, XXX−XXX
Langmuir
Article
Figure 3. (a) Mass spectra of cellobiose, Leu-enkephalin, glucose, and phenylalanine obtained on GO(1,5,10) films (see Figure S2 for all spectra of small molecules obtained on GO(1−10) films). (b) Mass spectra of HER, PKA, and PPKA peptides on GO(1,5,10) films.
solutions of PTG, PDPG, and PPKA for 12 h to fully induce absorption of pyrenylated analytes by π−π interaction and subjected to LDI-MS analysis. In contrast to cellobiose, Leuenkephaline, phenylalanine, and glucose, PTG, PDPG, and PPKA peptides were successfully detected on GO(1) films as protonated ions with low background interference (PTG: m/z 421 [M + H]+; PDPG: m/z 507 [M + H]+; and PPKA: m/z 1041 [M + H]) (Figure S3). This result confirmed that the pyrene pendent groups successfully served as an anchor and subsequently LDI efficiency enhancer. Interestingly, the peak intensities corresponding to PTG, PDPG, and PPKA peptide respectively reached to maximum value on GO(4), GO(5), and GO(8) films and diminished without significant increase of background interference signal with further LBL assembly cycles (Figure 4b). This tendency is particularly interesting because our previous reports showed only increased or saturated mass peak intensities of small molecules with increasing background interference as LBL assembly cycles of GO and MWCNT increased. In the present study, the LDI-MS results of pyrenylated compounds on GO(1−10) films clearly showed that the optimum thickness of GO films existed depending on the chemical structure of analytes for successful mass spectrometric analysis. This discordance might be originated from the different laser energy absorption capacity of MWCNT and PAAH which were respectively used as a positively charged component to fabricate GO LBL assembled films. When MWCNT was used to fabricate LBL assembled GO films, the laser energy absorption capacity was rapidly enhanced by each LBL assembly cycle, which make it difficult to find the optimum number of GO LBL assembly cycles because of fragmentation-induced background interference. Compared to MWCNT, PAAH has negligible laser energy absorption capacity, and thus, LBL assembled GO and PAAH could precisely control the laser absorption capacity of the
might be attributed to their high threshold laser energy compared to low molecular weight compounds because the high molecular weight compounds could sufficiently harness the increasing laser energy absorbed by LBL assembled GO(1−10) films for their LDI process. By contrast, the low molecular weight compounds could not fully harness the increasing laser energy absorbed by LBL assembled GO(1−10) films, and thus, the excess laser energy could induce background interference in the low mass region by fragmentation. To further prove this assumption, small peptides such as HER peptide (AHHAHHAAD), protein kinase A (PKA), substrate peptide (LRRASLG), and pyrenylated PKA (PPKA) peptide were chosen as analytes because they have larger molecular weights than above tested small molecules and different amino acid sequence than each other. 1 nmol of HER, PKA, and PPKA peptides was deposited on GO(1,5,10) films and subjected to LDI-MS analysis. The HRE, PKA, and PPKA peptides were only detected on GO(10) films as protonated ions and sodium and/or potassium adducts (HRE peptide: m/z 1003 [M + Na]+ and m/z 1019 [M + K]+; PKA peptide: m/z 836 [M + Na]+ and m/z 852 [M + K]+; and PPKA peptide: m/ z 1041 [M + H]+, m/z 1063 [M + Na]+, and m/z 1079 [M + K]+). This result also suggested that the increased thickness resulted in enhancement of LDI-MS efficiency of analytes with molecular weights over 300 Da (Figure 3b). To precisely investigate the structural influence of GO films on LDI-MS analysis, the pyrenylated triethylene glycol (PTG), pyrenylated diaminopentaethylene glycol (PDPG), and pyrenylated PKA (PPKA) peptide were synthesized and used as analytes (Figure 4a). The pyrene pendent groups of the compounds provide high LDI efficiency16 and strong affinity to GO films for efficient energy transfer as well as homogeneous absorption of analytes on GO(1−10) films.17 The prepared GO(1−10) films were respectively immersed in 1 mM ethanolic E
dx.doi.org/10.1021/la5027653 | Langmuir XXXX, XXX, XXX−XXX
Langmuir
Article
Figure 4. (a) Chemical structures of pyrenylated compounds and (b) the normalized mass signal intensities of pyrenylated compounds obtained on GO(1−10) films fabricated by the LBL assembly process. The desorption efficiency and surface yield of BP obtained on (c) GO(1−10) and (d) GO(1−10)/MWCNT-NH2 films fabricated by the LBL assembly process. The error bars indicated standard deviation calculated from at least three measurements at different positions of each substrate.
resulting GO films. Therefore, the proper laser energy absorption capacity of GO films could be found, and this result suggests that the structural factors of GO films should be considered on the basis of the molecular weights of analytes for efficient LDI-MS analysis. To explore the detailed LDI-MS process of small molecules on GO(1−10) films, benzylpyridinium salt (BP) was harnessed as a model compound to obtain important two parameters for investigation of LDI-MS process−desorption efficiency and survival yield which are respectively estimated from summing absolute peak intensity of parent BP ion and fragmented BP ion and from dividing parent BP ion intensity by total peak intensity of parent BP ion and fragmented BP ion.18 The desorption efficiency of BP on GO(1) films was 80 ± 52, and this value increased to 335 ± 129 by one LBL assembly cycle. Despite the increasing LBL assembly cycles from 2 to 8, the desorption efficiency of GO(2−8) films was almost maintained around 748 ± 684, but further LBL assembly cycles resulted in rapid increase of desorption efficiency to 10 598 ± 1372 and 21 150 ± 8056 on GO(9) and GO(10) films, respectively (Figure 4c and Figure S4). The nonlinear increase of desorption efficiency was interesting considering the linear increase of thickness and laser energy absorption capacity of GO films with 1−10 LBL assembly cycles. The results indicated that the efficient LDI of BP on GO films required threshold laser energy
absorption capacity provided by more than eight LBL assembly cycles. This change of desorption efficiency of BP was well matched with the LDI-MS analysis results of Leu-enkephalin which started to be detected on GO(8) films. On the other hand, the survival yield of BP on GO(1) films significantly decreased from 69 ± 4% on GO(1) to 42.5 ± 1% on GO(2) film, and this value showed no significant change on GO(3−10) films with further LBL assembly cycles (the survival yield of BP on GO(10) was 38 ± 9%) (Figure 4c and Figure S4). The decreased survival yield on GO films with increasing LBL assembly cycles might be attributed to very high, local heat generation induced by increased laser energy absorption capacity which resulted in significant background interference in low mass region of small molecule mass spectra.11 Next, the survival yield and desorption efficiency of BP on GO(1−10)/MWCNT-NH2 films were obtained to reveal the effect of MWCNT-NH2 incorporation. The desorption efficiency of BP on GO(1) film increased from 80 ± 52 to 40 950 ± 16125 on GO(1)/MWCNT-NH2 films, and this value gradually increased to 51 506 ± 8222 with 1−3 LBL assembly cycles. The desorption efficiency rapidly increased to 141 994 ± 10 980 by four LBL assembly cycles and plateaued with further LBL assembly cycles (the desorption efficiency of BP on GO(10)/MWCNT-NH2 films was 13 6297 ± 15 414) (Figure 4d and Figure S5). This result also indicated that the present GO(1−10)/MWCNT-NH2 films provide F
dx.doi.org/10.1021/la5027653 | Langmuir XXXX, XXX, XXX−XXX
Langmuir
Article
Figure 5. Mass spectra of cellobiose, Leu-enkephalin, glutamine, glucose, leucine, lysine, mannitol, and phenylalanine obtained on (a) GO(1)/ MWCNT-NH2 and (b) GO(10)/MWCNT-NH2 films.
Figure 6. Mass spectra of cellobiose, Leu-enkephalin, glutamine, glucose, leucine, lysine, mannitol, and phenylalanine obtained on (a) GO(1)/ MWCNT-NH2 and (b) GO(10)/MWCNT-NH2 films with 69.6 μJ laser power.
desorption efficiency, the survival yield of BP on GO(1) film (68 ± 4%) slightly decreased to 60 ± 2% on GO(1)/MWCNT-
reproducible mass signal intensity from their overall surfaces without sweet spots. Despite the significantly increased G
dx.doi.org/10.1021/la5027653 | Langmuir XXXX, XXX, XXX−XXX
Langmuir
Article
NH2 film, and this value did not notably change with 1−10 LBL assembly cycles although accompanying slight fluctuation (the survival yield of BP on GO(10)/MWCNT-NH2 was 63 ± 1%) (Figure 4d and Figure S5). This result showed the importance of MWCNT-NH2 incorporated on GO(1−10) films as a thermally conductive bridge and laser energy absorber to improve desorption efficiency and survival yield for efficient and reproducible LDI-MS analysis.10,11 Finally, the structural influence of GO films on the synergistic effect with MWCNT-NH2 for LDI-MS analysis of small molecules was investigated by analyzing cellobiose, Leuenkephaline, glutamine, glucose, leucine, lysine, D-mannitol, and phenylalanine on GO(1−10)/MWCNT-NH2 films. All of small molecules were clearly detected on GO(1)/MWCNTNH2 films as protonated ions and sodium and/or potassium adducts with negligible background interference (cellobiose: m/ z 365 [M + Na]+; Leu-enkephalin: m/z 577 [M + Na]+ and m/ z 593 [M + K]+; glutamine: m/z 148 [M + H]+ and m/z 170 [M + Na]+; glucose: m/z 203 [M + Na]+ and m/z 219 [M + K]+; leucine: m/z 154 [M + Na]+; lysine: m/z 169 [M + Na]+; + + D-mannitol: m/z 205 [M + Na] and m/z 221 [M + K] ; and + phenylalanine: m/z 166 [M + H] , m/z 188 [M + Na]+, and m/z 204 [M + K]+) (Figure 5a). Along with the LBL assembly cycles of GO(2−10) films underneath MWCNT-NH2, all of mass peak intensities corresponding to each analyte were considerably enhanced with significant background interference (Figure 5b and see Figure S6 for all spectra obtained on GO(1−10)/MWCNT-NH2 films). Interestingly, the degree of background interference in mass spectra of glutamine, glucose, leucine, lysine, D-mannitol, and phenylalanine was higher than that in mass spectra of cellobiose and Leu-enkaphaline (Figure 5 and Figure S6). This difference in the degree of background interference was well matched with the LDI-MS result of small molecules on GO(1−10) films which showed that the high molecular weight compounds required high laser energy to be detected and generated less background interference in the low mass region than low molecular weight compounds. This result clearly indicated that the structural factors of GO films influenced on not only LDI-MS efficiency of small molecules but also synergistic effect with MWCNT-NH2 incorporation. To further investigate the structural influence of GO films on synergistic effect with MWCNT-NH2 for LDI-MS analysis, LDI-MS analysis of small molecules was carried out on GO(1−10)/MWCNT-NH2 films under lower laser power (69.6 μJ) than conventionally used laser power throughout this study (76.8 μJ). While no mass signal of small molecules was detected on GO(1)/MWCNT-NH2 films (Figure 6a), small molecules started to be detected on GO(4)/MWCNT-NH2 films (see Figure S7 for all spectra obtained on GO(1−10)/MWCNT-NH2 films), and the mass peaks of all small molecules were enhanced on GO(10)/MWCNT-NH2 films with LBL assembly cycles without much background interference (Figure 6b and Figure S7). This result indicated that the background interference resulted from the thermal fragmentation of carbon nanomaterials and small molecules, and threshold laser power for LDIMS analysis of small molecules could be lowered by controlling structural factors of GO films underneath of MWCNT-NH2.
films were precisely controlled by LBL assembly of GO and PAAH while maintaining their surface morphology. The optical and physical properties were thoroughly characterized to clearly reveal the correlation of those structural factors with both LDIMS efficiency of small molecules with and without MWCNTNH2 incorporation. On the basis of the results, we found several important factors in LDI-MS analysis of small molecules. First, we found that the threshold energy for LDIMS analysis of small molecules was dependent on their chemical structures. Second, the optimal number of GO(1−10) films is also dependent on the chemical structures of analytes. Third, the high molecular weight compounds are less fragmented than low molecular weight compounds under the same LDI-MS analysis conditions. Fourth, the threshold laser energy for LDI-MS analysis on GO(1−10)/MWCNT-NH2 films could be lowered by increasing number of GO films underneath incorporated MWCNT-NH2. Taken together, the GO films underneath of MWCNT top layer are proper components to optimize the structural design of GO and MWCNT hybrid films by the LBL assembly process for LDI-MS analysis of small molecules, and thus, it is suggested that when designing carbon nanomaterial-based LDI-MS platform, the chemical structures of analytes such as molecular weight, functional groups, and thermal stability should be considered to optimize LDI-MS analysis condition for each analyte without fragmentationinduced background interference in the low mass region. We believe that the present study could provide a meaningful and significant knowledge for the development of efficient LDI-MS platform by using carbon nanomaterials for important small molecules having liable structures to be fragmented during LDI-MS analysis based on the correlation between structural factors of LDI substrates and chemical structures and molecular weights of analytes. We are currently investigating carbon nanomaterial-based LDI-MS platforms for specific samples, for example, oligosaccharides, hydrophobic molecules, and peptides, for various bioanalytical applications.
CONCLUSION In conclusion, we investigated the structural influence of GO films on LDI-MS efficiency of small molecules and synergistic effect of additional MWCNT-NH2 incorporation. For this study, the thickness and laser energy absorption capacity of GO
(1) Gemoll, T.; Roblick, U. J.; Habermann, J. K. MALDI Mass Spectrometry Imaging in Oncology. Mol. Med. Rep. 2011, 4, 1045− 1051. (2) Jungbauer, L. M.; Cavagnero, S. Characterization of Protein Expression and Folding in Cell-Free Systems by MALDI-TOF Mass Spectrometry. Anal. Chem. 2006, 78, 2841−2852.
■
ASSOCIATED CONTENT
* Supporting Information S
Experimental details and Figures S1−S7. This material is available free of charge via the Internet at http://pubs.acs.org.
■
AUTHOR INFORMATION
Corresponding Author
*Fax +82-2-875-6636; Tel +82-2-880-4338; e-mail dalheemin@ snu.ac.kr (D.-H.M.). Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS This work was supported by the Basic Science Research Program (2011-0017356) through the National Research Foundation of Korea (NRF) and by the Research Center Program (IBS-R008-D1) funded by the Korean government.
■
■
H
REFERENCES
dx.doi.org/10.1021/la5027653 | Langmuir XXXX, XXX, XXX−XXX
Langmuir
Article
(3) Aebersold, R.; Goodlett, D. R. Mass Spectrometry in Proteomics. Chem. Rev. 2001, 101, 269−296. (4) Guo, Z.; Zhang, Q.; Zou, H.; Guo, B.; Ni, J. A Method for the Analysis of Low-Mass Molecules by MALDI-TOF Mass Spectrometry. Anal. Chem. 2002, 74, 1637−1641. (5) Chiang, C. K.; Chen, W. T.; Chang, H. T. Nanoparticle-Based Mass Spectrometry for the Analysis of Biomolecules. Chem. Soc. Rev. 2011, 40, 1269−1281. (6) Shiea, J. T.; Huang, J. P.; Teng, C. F.; Jeng, J. Y.; Wang, L. Y.; Chiang, L. Y. Use of a Water-Soluble Fullerene Derivative as Precipitating Reagent and Matrix-Assisted Laser Desorption/Ionization Matrix to Selectively Detect Charged Species in Aqueous Solutions. Anal. Chem. 2003, 75, 3587−3595. (7) Xu, S.; Li, Y.; Zou, H.; Qiu, J.; Guo, Z.; Guo, B. Carbon Nanotubes as Assisted Matrix for Laser Desorption/Ionization Timeof-Flight Mass Spectrometry. Anal. Chem. 2003, 75, 6191−6195. (8) Dong, X.; Cheng, J.; Li, J.; Wang, Y. Graphene as a Novel Matrix for the Analysis of Small Molecules by MALDI-TOF MS. Anal. Chem. 2010, 82, 6208−6214. (9) Gulbakan, B.; Yasun, E.; Shukoor, M. I.; Zhu, Z.; You, M.; Tan, X.; Sanchez, H.; Powell, D. H.; Dai, H.; Tan, W. A Dual Platform for Selective Analyte Enrichment and Ionization in Mass Spectrometry using Aptamer-Conjugated Graphene Oxide. J. Am. Chem. Soc. 2010, 132, 17408−17410. (10) Kim, Y.-K.; Na, H.-K.; Kwack, S.-J.; Ryoo, S.-R.; Lee, Y.; Hong, S.; Hong, S.; Jeong, Y.; Min, D.-H. Synergistic Effect of Graphene Oxide/MWCNT Films in Laser Desorption/Ionization Mass Spectrometry of Small Molecules and Tissue Imaging. ACS Nano 2011, 5, 4550−4561. (11) Kim, Y.-K.; Min, D.-H. Fabrication of Alternating Multilayer Films of Graphene Oxide and Carbon Nanotube and Its Application in Mechanistic Study of Laser Desorption/Ionization of Small Molecules. ACS Appl. Mater. Interfaces 2012, 4, 2088−2095. (12) Hong, J.; Han, J. Y.; Yoon, H.; Joo, P.; Lee, T.; Seo, E.; Char, K.; Kim, B. S. Carbon-Based Layer-by-Layer Nanostructures: from Films to Hollow Capsules. Nanoscale 2011, 3, 4515−4531. (13) Srivastava, S.; Kotov, N. A. Composite Layer-by-Layer (LBL) Assembly with Inorganic Nanoparticles and Nanowires. Acc. Chem. Res. 2008, 41, 1831−1841. (14) Lee, S. W.; Kim, B. S.; Chen, S.; S. H, Y.; Hammond, P. T. Layer-by-Layer Assembly of All Carbon Nanotube Ultrathin Films for Electrochemical Applications. J. Am. Chem. Soc. 2009, 131, 671−679. (15) Hummers, W. S.; Offeman, R. E. Preparation of Graphitic Oxide. J. Am. Chem. Soc. 1958, 80, 1339. (16) Cabrera-Pardo, J. R.; Chai, D. I.; Liu, S.; Mrksich, M.; Kozmin, S. A. Label-Assisted Mass Spectrometry for the Acceleration of Reaction Discovery and Optimization. Nat. Chem. 2013, 5, 423−427. (17) Xu, Y.; Bai, H.; Lu, G.; Li, C.; Shi, G. Flexible graphene films via the filtration of water-soluble noncovalent functionalized graphene sheets. J. Am. Chem. Soc. 2008, 130, 5856−5857. (18) Tang, H. W.; Ng, K. M.; Lu, W.; Che, C. M. Ion Desorption Efficiency and Internal Energy Transfer in Carbon-Based SurfaceAssisted Laser Desorption/Ionization Mass Spectrometry: Desorption Mechanism(s) and the Design of SALDI Substrates. Anal. Chem. 2009, 81, 4720−4729.
I
dx.doi.org/10.1021/la5027653 | Langmuir XXXX, XXX, XXX−XXX
