Surface Desorption Dielectric-Barrier Discharge Ionization Mass

Jun 21, 2017 - A variant of dielectric-barrier discharge named surface desorption dielectric-barrier discharge ionization (SDDBDI) mass spectrometry w...
0 downloads 0 Views 888KB Size
Subscriber access provided by Binghamton University | Libraries

Article

Surface desorption dielectric-barrier discharge ionization mass spectrometry Hong Zhang, Jie Jiang, Na Li, Ming Li, Yingying Wang, Jing He, and Hong You Anal. Chem., Just Accepted Manuscript • Publication Date (Web): 21 Jun 2017 Downloaded from http://pubs.acs.org on June 21, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Analytical Chemistry is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 8

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Surface desorption dielectric-barrier discharge ionization mass spectrometry Hong Zhangb,c, Jie Jianga*, Na Lia, Ming Lid*, Yingying Wanga, Jing Hea, Hong Youa,b,c a

School of Marine Science and Technology, Harbin Institute of Technology at Weihai, Weihai, Shandong 264209, P. R. China b State Key Laboratory of Urban Water Resource and Environment, Harbin Institute of Technology, Harbin, Heilongjiang 150090, P. R. China c School of Municipal and Environmental Engineering, Harbin Institute of Technology, Harbin, Heilongjiang 150090, P. R. China d Division of Chemical Metrology and Analytical Science, National Institute of Metrology, Beijing 100029, P. R. China ABSTRACT: A variant of dielectric-barrier discharge named surface desorption dielectric-barrier discharge ionization (SDDBDI) mass spectrometry was developed for high-efficiency ion transmission and high spatial resolution imaging. In SDDBDI, a tungsten nanotip and the inlet of the mass spectrometer are used as electrodes, and a piece of cover slip is used as a sample plate as well as an insulating dielectric barrier, which simplifies the configuration of instrument and thus the operation. Different from volume dielectric-barrier discharge (VDBD), the microdischarges are generated on the surface at SDDBDI, and therefore the plasma density is extremely high. Analyte ions are guided directly into the MS inlet without any deflection. This configuration significantly improves the ion transmission efficiency and thus the sensitivity. The dependence of sensitivity and spatial resolution of the SDDBDI on the operation parameters were systematically investigated. The application of SDDBDI was successfully demonstrated by analysis of multiple species including amino acids, pharmaceuticals, putative cancer biomarkers, and mixtures of both fatty acids and hormones. Limits of detection (S/N = 3) were determined to be 0.84 and 0.18 pmol, respectively, for the analysis of L-alanine, and metronidazole. A spatial resolution of 22 µm was obtained for the analysis of an imprinted cyclophosphamide pattern, and imaging of a “T” character was successfully demonstrated under ambient conditions. These results indicate that SDDBDI has highefficiency ion transmission, high sensitivity, and high spatial resolution, which render it a potential tool for mass spectrometry imaging.

Mass spectrometry imaging (MSI) is well-suited for mapping the spatial distributions of a wide range of chemical compounds in their native environment1,2. Initially, spatiallylocalized sampling in mass spectrometry imaging is achieved by secondary ion mass spectrometry (SIMS)3, and matrixassisted laser desorption ionization (MALDI)4. Later, MSI is dominated with ambient ionization techniques including desorption electrospray ionization (DESI)5, laser ablate electrospray ionization (LAESI)6, low-temperature plasma probe (LTP)7, and nanotip ambient ionization (NAIMS)8. The application of ambient ionization techniques in MSI has been extensively investigated and well documented2,9. DESI has emerged as a popular platform in imaging applications10. A typical spatial resolution of DESI is 200 µm10, and approximate 40 µm is possible with careful optimization of experimental parameters11. Though DESI’s spatial resolution is lower than those of SIMS and MALDI (10 µm or better), it is sufficient for identification and localization of compounds on various samples. DESI is a label-free technique and the potential interference from matrix as for MALDI is avoided, especially in detection of small molecules12,13. Another approach is direct liquid extraction techniques2, whose spatial resolution approximately ranges from 10 to 200 µm.

Multiple techniques such as electrospray-assisted laser desorption/ionization (ELDI)14, and laser ablation electrospray ionization (LAESI)6 offer advantages on spatial resolution both in depth and width compared to droplet-based methods5,15. With these techniques, the analyte is directly desorbed via laser beam and is subsequently extracted and ionized by an electrospray plume. MSI applications with these techniques range from small molecules such as metabolites to larger biomolecules such as proteins6, and from understanding the microorganisms communication16 to in situ cell-by-cell imaging17. These techniques also allow three-dimensional (3D) imaging of tissues under ambient conditions, and the reported depth resolution of LAESI is approximate 30 µm18. Plasma-based methods with polarity independence has been also widely used for direct surface analysis of various samples with different sizes and shapes7,8,19. LTP, for example, was used for the analysis of works of art without damage or contamination20, and also was used to identify the spatial arrangement of biomolecules such as capsaicin as well21. The lateral spatial resolution with the LTP probe is approximate 250 µm, which is comparable to that of DESI. NAIMS using a tungsten nanotip to generate a plasma also successfully located sample molecules under ambient conditions8. Different variants of plasma-based methods have been demonstrated by

ACS Paragon Plus Environment

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

modifying and adjusting the electrode configuration22. However, they haven’t been used for MS imaging so far. It is notable that a drastic decrease in MS sensitivity is often observed when spatial resolution improved during imaging of biological tissues23. However, effect of ion transmission efficiency from the ionization region to the mass spectrometer on the spatial resolution has been rarely reported. Actually, only a small portion of the ions are introduced into the mass spectrometer because ions plume generated during desorption/ionization process using a solvent spray, laser beam or plasma probe5-8, exceeds the tolerance of a capillary orifice. Herein, a surface desorption dielectric-barrier discharge ionization (SDDBDI) with advantages of high-efficiency ion transmission and high spatial resolution was developed. Different from volume dielectric barrier discharge ionization (VDBDI)19, ionized analytes on the surface are directly guided to the MS inlet without any deflection, which extremely improves the ion transmission efficiency and thus the sensitivity. Only a nanotip is used as a discharge electrode, and the MS inlet, instead of a metal plate, is used as another electrode. Therefore, the configuration simplifies the instrument and thus the operation. In this work, parameters for improving sensitivity and spatial resolution were systematically optimized. The characteristics and pilot applications of SDDBDI are presented. EXPERIMENTAL SECTION Chemicals and Reagents Methanol at MS-grade was purchased from Sigma-Aldrich (Darmstadt, Germany). Helium and argon (purity ≥ 99.999%) were purchased from Jinghua Industry Co. (Hangzhou, China). The tungsten nanotips with size of 0.5 and 10 µm were purchased from GGB Industries, Inc. (Naples, FL). The cover slip was purchased from Fisher Scientific. Amino acids were purchased from Sigma-Aldrich (St. Louis, MO, USA). Cyclophosphamide, 8-hydroxyquinoline, rhodamine B, adenosine, fatty acids, hormones, lipids, and cholesteryl sulfate were purchased from Sigma-Aldrich (Steinheim, Germany). Ultrapure water was obtained from a Milli-Q water purification system (Milford, MA). All the individual solutions were freshly prepared before use. Surface Desorption Dielectric-barrier Discharge Ionization (SDDBDI) As shown in Figure 1, SDDBDI consists of a tungsten nanotip electrode, and a piece of cover slip used as dielectric barrier as well as sample plate. The mass spectrometer cone was used as ground electrode. The nanotip electrode was positioned align with the MS inlet. The cover slip was vertically abutted to the nanotip. A custom-designed PEEK nozzle (ID 2 × 1 mm, OD 3 × 2 mm) positioned vertically to the axis of the two electrodes was used to introduce the discharge gases (He, Ar, or air). A sinusoidal signal with a frequency of 63.5 kHz generated from a waveform generator (Tektronix, Shanghai, China) was amplified by a power amplifier (Electronics and Innovation, Model 1020L, Rochester, USA). Subsequently, the sinusoidal voltage was transferred to a home-built coil to generate an alternating high voltage. Finally, an alternating high voltage of 4 kV was applied between the nanotip and the MS inlet to generate a plasma. A DC offset (Lion, HV-502P2, Tianjin, China) was applied to the nanotip to enhance ion transmission efficiency. Unless otherwise specified, the sample was prepared by pipetting 2 µL of methanol/water (v/v: 1/1) solution onto the glass surface and allowing it dry at room

Page 2 of 8

temperature. All the mass spectra were recorded using Xcalibur® software supplied with the mass spectrometer. Mass Spectrometry Imaging A 2-D motorized stage (Chuo Seiki, Kōchi, Japan) was used to move the holder of the dielectric barrier. The position of the dielectric barrier was controlled using a custom-designed LabVIEW software (version 11.0, National Instruments, Austin, Texas, USA). The nanotip, 2-D stage, and the holder were assembled together with an extruded aluminum base. The experiments for optimization of spatial resolution were performed by measuring the desorption width of Rhodamine B deposited on a cover slip under an optical microscope (Phenix, Jiangxi, China). Average datum of triple measurements was used as a final result. The Rhodamine B grid containing cyclophosphamide was prepared as described elsewhere24. The use of Rhodamine B was for a visual observation of scanning area. The grid was imprinted on a cover slip surface by gently pressing the cover slip against an electroformed mesh (300 lpi, Precision Eforming LLC., Cortland, USA) placed on a glass slide surface. The width of the grid lines was 25 µm, and the spacing between them was 60 µm. For the grid imaging, step sizes in the x- and y-dimension were set at 5 and 60 µm, respectively. For imaging a “T” character, a mixture solution of cyclophosphamide and L-alanine in methanol/water (v/v: 1/1) was drawn on a glass surface, generating a “T” character. The “T” character was scanned with a rate of 100 µm/s. To simplify data analysis and obtain satisfied imaging quality, step sizes in both x- and y-dimension were set at 100 µm. An area of approximate 6 × 8 mm was scanned. Data were manually exported from the Xcalibur® software and plotted in a false-color scale using Origin® 9.0 software (OriginLab Co., Northampton, MA) to determine the 2D distribution of analytes.

Figure 1. Schematic of SDDBDI. Mass Spectrometry Experiments All the experiments were implemented on a LTQ/Orbitrap mass spectrometer (Thermo Fisher Scientific, Waltham, MA). The ion maximum injection time of the Orbitrap was set at 500 ms. The capillary temperature was maintained at 275°C. The tube lens voltages in positive and negative mode were set at 65 and -241 V, respectively. The capillary voltage was blocked with a rocker switch to prevent damage of the electronic device from discharge. RESULTS AND DISCUSSION Procedures of SDDBDI-MS In SDDBDI, the solution samples are directly dried on a glass slide. Similar to DBD19, insulated glass was used as dielectric barrier in SDDBDI. When a conductive glass used, a

ACS Paragon Plus Environment

Page 3 of 8

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

filiform plasma was generated close to the glass surface, resulting in an unstable plasma and a low signal intensity (Figure S1). Other insulated dielectrics such as ceramic and polytetrafluoroethylene (PTFE) were also investigated (Figure S2). Compared to insulated glass, similar mass spectra and sensitivity were obtained. For convenience consideration, a glass slide was used as dielectric barrier as well as sample plate in this work. To avoid the damage of the nanotip caused by scratching the glass slide, the distance between the nanotip and the glass slide was set at 0.2 ~ 0.4 mm. Within this distance, the plasma was stable and similar signal intensity was obtained (Figure S3). However, when the distance was above 0.4 mm, an unstable plasma was generated between the glass slide and the MS inlet, resulting in a low signal intensity, as shown in Figure S3. Effect of DC voltage Besides the mass spectrometer vacuum, the potential difference between the nanotip and the capillary orifice pushes the analyte ions from atmospheric pressure into the MS. Ion trajectory simulations were performed using SIMION® software and the result is shown in Figure S4. The ions are focused into a narrow beam when DC voltage increased. Dependence of MS signal on DC voltages was investigated by monitoring the protonated L-alanine (0.18 µg on glass surface). Not surprisingly, MS signal increased about five folds when DC voltage increased (Figure 2). However, the signal intensity did not continuously increase when DC voltage is above 200 V. A possible reason for this phenomenon is that most of the target ions expanded at atmosphere pressure were pulled into the MS inlet with 200 V DC applied. Other amino acids were also tested with or without DC voltage applied. MS signals were improved 2 ~ 6 folds with 200 V DC voltage applied (Table S1). In addition, the abundances of fragments for all amino acids tested with or without 200 V DC voltage applied were listed in Table S1, and less than 5 % variation on the abundances of the fragments was observed; while they were reported to be different in conventional DBDI19. These results prove that the DC voltage only improves ion transmission efficiency and does not cause fragmentation.

With the flow rate of discharge gas of 0.6 L/min, the signal intensity is the maximum as shown in Figure 3a. When the flow rate was set at 0.2, 0.4, 0.8, and 1.0 L/min, the plasma was not very stable and lower signal intensity was obtained. As shown in Figure 3b, the optimal tip-to-cone distance was found to be 3 mm. When the tip-to-cone distance was set at 2 mm, the homogeneous discharge transformed to filamentarymode discharge which was characterized with low ionization efficiency25. When the tip-to-cone distance was 4, 5 or 6 mm, the plasma was visually unstable and even disappeared. The position of gas nozzle was crucial for this source. The signal intensity was largest when the nozzle-to-glass distance was set at 0.5 mm, as shown in Figure 3c. The plasma was unstable at 0 mm distance. The signal intensity was lower because the microdischarges on the cover slip became weakly ionized plasma channels and could not develop into a surface discharge26. When the nozzle-to-glass distance was above 0.5 mm, the plasma on the sample surface was small, resulting in poor sample desorption efficiency and having low signal intensity. The optimal distance of nozzle-to-axis was found to be 1.5 mm, as shown in Figure 3d. A dramatic increase of signal intensity from 1.15 to 1.5 mm, and decrease of signal intensity from 1.5 to 3.0 mm, was observed. The gas composition varies along direction of nozzle-to-axis due to radial diffusion of the discharge gas. When the distance of nozzle-to-axis is larger than 1.5 mm, the gas mixture contains too much nitrogen, and the density of helium metastables (Hem) is too low to form enough reagent ions for target sample ionized.

Figure 3. Optimization of operating parameters: (a) flow rate of discharge gas, (b) tip-to-cone distance, (c) nozzle-to-glass distance, and (d) nozzle-to-axis distance. The signal intensity was obtained by monitoring protonated L-alanine (m/z 90) (0.18 µg on glass) with helium as discharge gas. Figure 2. Dependence of signal intensity on DC voltage applied to the nanotip. Error bars represent standard deviation from three measurements.

Parameters optimization The flow rate of the discharge gas, tip-to-cone distance, nozzle-to-glass distance, and nozzle-to-axis distance were carefully optimized to achieve better sensitivity. A diagram (Figure S5) was used for visualization of these terminologies and the results are shown in Figure 3.

Theoretical model A theoretical model was proposed to explain the behavior that the signal intensity increased approximate 6 times with both 200 V DC voltage and discharge gas applied (Figure S6 and Figure 2). Without DC voltage applied, the target ions (red balls) are radially dispersed away from the tip-to-cone axis (Figure S6a and S6b). With DC voltage applied, the target ions are radially constrained along the tip-to-cone axis (Figure S6c). This was confirmed by analysis of amine acids with and without DC voltage applied (Figure 2 and Table S1). Without discharge gas added, the neutral molecules (e.g. O2, N2, H2O)

ACS Paragon Plus Environment

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

exist in the tip-to-cone axis (Figure S6a), while these neutrals are pushed out of the tip-to-cone axis with discharge gas flow (Figure S6b and S6c). This avoids some collisions and reactions between these neutrals and energy species. This was confirmed by the experiments with air as discharge gas. With a flow rate of air increased, the mass spectrum of L-serine (0.5 µg on glass surface) is getting clearer, as shown in Figure S7. Performance of SDDBDI i) Limit of detection The limit of detection (LOD) of SDDBDI-MS was investigated by analysis of L-alanine and metronidazole. The ions of m/z 90 and 172 were used for estimating the calibration curves of L-alanine and metronidazole, respectively. As shown in Figure 4, a good correlation coefficient was obtained. The LOD (S/N = 3) for analysis of L-alanine was calculated to be 0.84 pmol, which is lower than the LOD obtained from volume DBDI (3.5 pmol)19. Also, the LOD for metronidazole was determined to be 0.18 pmol (also expressed as 0.09 µmol/L), which is lower than the LOD (0.15 mmol/L) for DBDI27. These results demonstrate that SDDBDI-MS with advantage of high-efficiency ion transmission is a very sensitive method.

Page 4 of 8

at m/z 146, 128, and 162 corresponded to the protonated 8hydroxyquinoline, [M-H2O+H]+ and [M+O+H]+, respectively. The formation of [M+O+H]+ was due to the oxidation of 8hydroxyquinoline during ionization. Such oxidation reaction was commonly observed for compounds containing the aromatic group in plasma-based ionization process such as LTP28 and flowing atmospheric pressure afterglow (FAPA)29. No reduction species formed by dihydrogenation was detected. This compared well with previous studies that reduction by dihydrogenation was observed for arenes using LTP, but not detected for aromatic heterocycle compounds28.

Figure 5. Analysis of pharmaceuticals on glass surface using SDDBDI. Mass spectra of (a) cyclophosphamide (0.14 µg) and (b) 8-hydroxyquinoline (0.15 µg).

Figure 4. Calibration curves of (a) L-alanine, and (b) metronidazole. The LODs for L-alanine and metronidazole were calculated to be 0.84 and 0.18 pmol, respectively. Error bars represent standard deviation from three measurements.

ii) Pharmaceutical analysis As demonstration for SDDBDI application, pharmaceuticals including cyclophosphamide and 8-hydroxyquinoline were analyzed when He was used as discharge gas, and the results are shown in Figure 5. The protonated cyclophosphamide at m/z 261 and a major fragment at m/z 142 were observed, as shown in Figure 5a. The fragment is assigned as [C4H9Cl2N+H]+, formed via the cleavage of N-P bond, which is confirmed by the characteristic chlorine isotopic envelope (inset of Figure 5a).The mass spectrum of 8-hydroxyquinoline is shown in Figure 5b. The signals

iii) Biological molecules analysis SDDBDI can also be greatly beneficial for applications in analysis of biological molecules in both positive and negative ion modes. To investigate the applicability of SDDBDI, positive and negative ion modes were performed for the analysis of adenosine and cholesteryl sulfate, respectively, reported as a putative tumor30 and prostate cancer biomarker31. In the mass spectrum of adenosine (Figure 6a), besides the signal of protonated adenosine at m/z 268, a base peak signal of m/z 136 was observed. This is similar with the spectrum using LTP32, while contrasted with the spectrum using ESI where signal at m/z 268 is the base peak32. The characteristic peak of m/z 136 was tentatively assigned as the fragment by neutral loss of C5O4H8 from protonated adenosine. Besides operated in positive ion mode, SDDBDI is also suitable for operation in negative ion mode. This was demonstrated by analysis of cholesteryl sulfate. The deprotonated cholesteryl sulfate and a characteristic fragment gave signals of m/z 465 and 97, respectively (Figure 6b). The characteristic fragment at m/z 97 was assigned as HSO4-31. Low intensities obtained for analysis of these compounds may be related to their low vapor pressure, resulting in poor sample desorption. This was also observed in DBDI, and

ACS Paragon Plus Environment

Page 5 of 8

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Figure 6. Mass spectra of biological molecules. (a) adenosine (9 ng); (b) cholesteryl sulfate (12 ng) (inset MS2 data); (c) mixture of palmitic acid, linoleic acid and oleic acid with each concentration of 0.4 µg; (d) mixture of juvenile hornone III, testosterone and progesterone with each concentration of 0.2 µg.

the sensitivity could be improved by heating the substrate33. The characteristic fragments obtained for these species can be used to directly determine the structure without additional MS/MS experiments. For example, full mass spectrum (Figure 6b) of cholesteryl sulfate shows the same characteristic peak of m/z 97 as MS2 spectrum (inset of Figure 6b). Extensive fragmentations of analytes were also observed using other plasma methods such as LTP, DBDI, and PADI7,19,22. In addition, mixtures of biologically-relevant compounds such as fatty acids (negative ion mode) and hormones (positive ion mode) were also investigated. In Figure 6c, the mass spectrum was dominated by the deprotonated palmitic acid (m/z 255), linoleic acid (m/z 279), and oleic acid (m/z 281). In Figure 6d, the peaks of protonated juvenile hornone III (m/z 267), testosterone (m/z 289), and progesterone (m/z 315) were clearly observed. These results indicate that SDDBDI is suitable for analysis of complex mixtures. Detection of some lipids (listed in Table S2) in both positive and negative ion modes were performed, but no parent ion peaks were observed (data not shown). To our best knowledge, little data about lipids were available using plasma-based methods (DBDI, LTP, and PADI and FAPA)34-36. SDDBDI is more suitable for analysis of small molecules. iv) Spatial resolution Carryover effect was investigated, because it influences both spatial resolution and imaging quality. 1-mm ink bars with 1-mm spacings (Figure S8a) prepared by an ink-jet printer was scanned by SDDBDI and the extracted ion chromatogram is shown in Figure S8b. The extracted ion current profile of protonated cyclophosphamide (m/z 261) was nearly a rectangle and no trailing was observed. The results show that no carryover effect exists in SDDBDI. To optimize spatial resolution, the parameters including AC voltage, the size of nanotip, and the thickness of dielectric

barrier were carefully investigated. As shown in Figure 7a, the spatial resolution is improved with AC voltage decreased, because lower voltage generates a smaller-sized plasma and thus a smaller desorption size. The spatial resolution is improved with nanotip size decreased, which is similar to the NAIMS8. The spatial resolution is improved with the thickness of dielectric barrier decreased. It was failed to ignite a plasma when the thickness of dielectric barrier was increased to 0.6 mm with discharge voltages of 3.6 and 3.8 kV, because an ignition threshold exists with a thicker dielectric barrier used. Thus, an AC voltage of 3.6 kV, a nanotip size of 0.5 µm, and a dielectric barrier width of 0.3 mm were used to achieve the optimal spatial resolution. Finally, a desorption width of approximate 22 µm after the plasma crossing the Rhodamine B surface was obtained at optimization conditions (Figure 7b). In addition, a Rhodamine B grid containing cyclophosphamide imprinted onto a cover slip surface was used to evaluate the spatial resolution of SDDBDI. The optical image of the grid is shown in Figure 7c. The corresponding ion image obtained for the protonated cyclophosphamide (m/z 261) in the white frame in Figure 7c is shown in Figure 7d. For convenience consideration, the scanning was began at the middle of one spacing between the lines. The step size in y-dimension was set at 60 µm which is same as the spacing between the lines. Therefore, no horizontal line was observed in the ion image. By using the distance required for a signal peak (m/z 261) rising from 20 to 80% as a measure of the spatial resolution8,24, the spatial resolution of SDDBDI was calculated to be approximate 22 µm (Figure 7e). This is better than the spatial resolution of LTP (250 µm)20. Mass spectrometry imaging As a proof-of-principle experiment, SDDBDI imaging of a “T” character was carried out. The character “T” was visually observed after drying at room temperature (Figure 8b). When

ACS Paragon Plus Environment

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 8

Figure 7. Spatial resolution of SDDBDI. (a) Optimization of spatial resolution; (b) desorption width on Rhodamine B surface; (c) optical image of grid on a cover slip surface; (d) SDDBDI image of cyclophosphamide at m/z 261 of the white frame in c, step size in the x- and ydimension set at 5 and 60 µm, respectively; (d) profile of ion intensity along the white-dotted line in d.

Figure 8. SDDBDI imaging of a “T” character. (a) mass spectrum of an individual pixel; (b) optical image of the character “T”; SDDBDI images of (c) m/z 261, and (d) m/z 90.

the microplasma crossed an individual pixel, the mass spectrum was dominated by the signal of protonated cyclophosphmides (m/z 261) and protonated L-alanines (m/z 90) (Figure 8a). With a pixel size of 100 × 100 µm, total acquisition time was 81 min for SDDBDI imaging. The corresponding ion images obtained for m/z 261 and 90 are shown in Figure 8c and 8d, respectively. These SDDBDI images match well with the optical image (Figure 8b). The distribution of these two chemicals is almost the same. These results demonstrate that SDDBDI is capable of detecting and locating the molecules under ambient conditions with no or little pretreatment. CONCLUSIONS In conclusion, a variant of dielectric-barrier discharge named SDDBDI was developed for direct surface analysis with high-efficiency ion transmission and high spatial resolution. A nanotip and the MS inlet, applied with an alternating

voltage, are used as electrodes. This configuration simplifies the instrument and thus the operation. The ions generated on the sample surface are guided directly into the MS inlet without any deflection resulting in high-efficiency ion transmission. The parameters for improving MS sensitivity and spatial resolution of the SDDBDI were systematically studied, providing a guidance for better instrument design. SDDBDI was applied for the analysis of amino acids, pharmaceuticals, putative cancer biomarkers, and mixtures such as fatty acids and hormones. SDDBDI using plasma to desorb and ionize analytes avoids the interferences from solvent. The LODs were determined to be 0.84 pmol for the analysis of L-alanine and 0.18 pmol for metronidazole. A spatial resolution of approximate 22 µm was obtained, and directly imaging a “T” character was successfully demonstrated. For high-resolution MS imaging using other ionization techniques, the sensitivity may not always be sufficient because the desorption size is small. While SDDBDI can be applied because it has the advantages of high spatial resolution without significantly sacrificing the MS sensitivity. These results indicate that SDDBDI has high sensitivity, and high spatial resolution, and therefore, it is a potential tool for mass spectrometry imaging of small molecules.

ASSOCIATED CONTENT Supporting Information Mass spectra with different dielectrics, gas flows, DC voltages; Effect of nanotip-to-dielectric distance; Simulated ion trajectory; Distance diagram; Theoretical model; Carryover effect; Lipids analyzed; AUTHOR INFORMATION

Corresponding Author *E-mail: [email protected]. Fax: + (86)-631-5685-359; [email protected]. Fax: +(86)-10-64526792;

Notes The authors declare no competing financial interest.

ACS Paragon Plus Environment

Page 7 of 8

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

ACKNOWLEDGMENTS This work was supported by the 13th Five Year Key Research & Development Program of China (No. 2016YFF0100302), Natural Science Foundation of Shandong (ZR2016BM11, ZR2016BP01) and Key Research & Development Program of HITWH (No. 2013DXGJ01, HIT.NSRIF.201710).

REFERENCES (1) Eberlin, L. S.; Margulis, K.; Planell-Mendez, I.; Zare, R. N.; Tibshirani, R.; Longacre, T. A.; Jalali, M.; Norton, J. A.; Poultsides, G. A. PLos Med. 2016, 13, 1-21. (2) Laskin, J.; Lanekoff, I. Anal. Chem. 2016, 88, 52-73. (3) Belu, A. M.; Graham, D. J.; Castner, D. G. Biomaterials 2003, 24, 3635-3653. (4) Caprioli, R. M.; Farmer, T. B.; Gile, J. Anal. Chem. 1997, 69, 4751-4760. (5) Takats, Z.; Wiseman, J. M.; Gologan, B.; Cooks, R. G. Science 2004, 306, 471-473. (6) Nemes, P.; Vertes, A. Anal. Chem. 2007, 79, 8098-8106. (7) Harper, J. D.; Charipar, N. A.; Mulligan, C. C.; Zhang, X. R.; Cooks, R. G.; Ouyang, Z. Anal. Chem. 2008, 80, 9097-9104. (8) Zhou, Z.; Lee, J. K.; Kim, S. C.; Zare, R. N. Anal. Chem. 2016, 88, 5542-5548. (9) Wu, C. P.; Dill, A. L.; Eberlin, L. S.; Cooks, R. G.; Ifa, D. R. Mass Spectrom. Rev. 2013, 32, 218-243. (10) Ifa, D. R.; Wiseman, J. M.; Song, Q. Y.; Cooks, R. G. Int. J. Mass Spectrom. 2007, 259, 8-15. (11) Kertesz, V.; Van Berkel, G. J. Rapid Commun. Mass Spectrom. 2008, 22, 2639-2644. (12) Guo, Z.; Zhang, Q. C.; Zou, H. F.; Guo, B. C.; Ni, J. Y. Anal. Chem. 2002, 74, 1637-1641. (13) Eberlin, L. S.; Ferreira, C. R.; Dill, A. L.; Ifa, D. R.; Cooks, R. G. Biochim. Biophys. Acta Mol. Cell Biol. Lipids 2011, 1811, 946-960. (14) Shiea, J.; Huang, M. Z.; Hsu, H. J.; Lee, C. Y.; Yuan, C. H.; Beech, I.; Sunner, J. Rapid Commun. Mass Spectrom. 2005, 19, 37013704. (15) Roach, P. J.; Laskin, J.; Laskin, A. Analyst 2010, 135, 22332236. (16) Stopka, S. A.; Mansour, T. R.; Shrestha, B.; Marechal, E.; Falconet, D.; Vertes, A. Anal. Chim. Acta 2016, 902, 1-7.

(17) Shrestha, B.; Patt, J. M.; Vertes, A. Anal. Chem. 2011, 83, 2947-2955. (18) Nemes, P.; Barton, A. A.; Vertes, A. Anal. Chem. 2009, 81, 6668-6675. (19) Na, N.; Zhao, M. X.; Zhang, S. C.; Yang, C. D.; Zhang, X. R. J. Am. Soc. Mass Spectrom. 2007, 18, 1859-1862. (20) Liu, Y.; Ma, X.; Lin, Z.; He, M.; Han, G.; Yang, C.; Xing, Z.; Zhang, S.; Zhang, X. Angew. Chem.-Int. Edit. 2010, 49, 4435-4437. (21) Maldonado-Torres, M.; Lopez-Hernandez, J. F.; JimenezSandoval, P.; Winkler, R. J. Proteomics 2014, 102, 60-65. (22) Ratcliffe, L. V.; Rutten, F. J. M.; Barrett, D. A.; Whitmore, T.; Seymour, D.; Greenwood, C.; Aranda-Gonzalvo, Y.; Robinson, S.; McCoustra, M. Anal. Chem. 2007, 79, 6094-6101. (23) Bartels, B.; Svatos, A. Front. Plant Sci. 2015, 6, 1-7. (24) Laskin, J.; Heath, B. S.; Roach, P. J.; Cazares, L.; Semmes, O. J. Anal. Chem. 2012, 84, 141-148. (25) Meyer, C.; Muller, S.; Gilbert-Lopez, B.; Franzke, J. Anal. Bioanal. Chem. 2013, 405, 4729-4735. (26) Kogelschatz, U.; Eliasson, B.; Egli, W. J. Phys. IV 1997, 7, 47-66. (27) Zhang, Y.; Ai, W.; Bai, Y.; Zhou, Y.; Wen, L.; Zhang, X.; Liu, H. Anal Bioanal Chem 2016, 408, 8655-8661. (28) Na, N.; Xia, Y.; Zhu, Z. L.; Zhang, X. R.; Cooks, R. G. Angew. Chem.-Int. Edit. 2009, 48, 2017-2019. (29) Zeiri, O. M.; Storey, A. P.; Ray, S. J.; Hieftje, G. M. Anal. Chim. Acta 2017, 952, 1-8. (30) Clayton, A.; Al-Taei, S.; Webber, J.; Mason, M. D.; Tabi, Z. J. Immunol. 2011, 187, 676-683. (31) Eberlin, L. S.; Dill, A. L.; Costa, A. B.; Ifa, D. R.; Cheng, L.; Masterson, T.; Koch, M.; Ratliff, T. L.; Cooks, R. G. Anal. Chem. 2010, 82, 3430-3434. (32) Albert, A.; Engelhard, C. Anal. Chem. 2012, 84, 10657-10664. (33) Gilbert-Lopez, B.; Schilling, M.; Ahlmann, N.; Michels, A.; Hayen, H.; Molina-Diaz, A.; Garcia-Reyes, J. F.; Franzke, J. Anal. Chem. 2013, 85, 3174-3182. (34) Brandt, S.; Klute, F. D.; Schuetz, A.; Franzke, J. Anal. Chim. Acta 2017, 951, 16-31. (35) Guo, C. A.; Tang, F.; Chen, J.; Wang, X. H.; Zhang, S. C.; Zhang, X. R. Anal. Bioanal. Chem. 2015, 407, 2345-2364. (36) Ding, X. L.; Duan, Y. X. Mass Spectrom. Rev. 2015, 34, 449473.

ACS Paragon Plus Environment

Analytical Chemistry

Page 8 of 8

TOC Graphic

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

8

ACS Paragon Plus Environment