Reduced Sampling Size with Nanopipette for Tapping-Mode Scanning Probe Electrospray Ionization Mass Spectrometry Imaging

Abstract

Mass spectrometry imaging (MSI) with ambient sampling and ionization can rapidly and easily capture the distribution of chemical components in a solid sample. Because the spatial resolution of MSI is limited by the size of the sampling area, reducing sampling size is an important goal for high resolution MSI. Here, we report the first use of a nanopipette for sampling and ionization by tapping-mode scanning probe electrospray ionization (t-SPESI). The spot size of the sampling area of a dye molecular film on a glass substrate was decreased to 6 μm on average by using a nanopipette. On the other hand, ionization efficiency increased with decreasing solvent flow rate. Our results indicate the compatibility between a reduced sampling area and the ionization efficiency using a nanopipette. MSI of micropatterns of ink on a glass and a polymer substrate were also demonstrated.

INTRODUCTION

Mass spectrometry imaging (MSI) has attracted attention for analyzing the distribution of multiple chemical species in materials in a wide range of fields, including biomedical science, food chemistry, and materials chemistry. Since the early 2000s, an increasing number of studies of ambient sampling/ionization mass spectrometry have been reported.¹⁾ This is because using an ambient ionization technique with MSI can be performed quickly and easily under atmospheric conditions without special sample preparation. Desorption electrospray ionization (DESI), one of the most well-known techniques, was first reported in 2004.²⁾ Charged solvent droplets are generated by electrospray ionization (ESI),³⁾ and are then sprayed onto the sample surface with a flow of high-pressure nebulizing gas to extract and ionize the compounds in the sample. Because ESI is a soft ionization method, ambient ionization techniques that use ESI can be applied to high polarity, sparingly volatile, and thermally labile compounds. Other methods such as nanospray desorption electrospray ionization (nano-DESI),⁴⁾ liquid micro junction-surface sampling probe (LMJ/SSP),⁵⁾ liquid extract surface analysis (LESA)⁶⁾ and probe electrospray ionization (PESI)^7,8) have also been developed and applied to MSI.⁹⁾

In 2012, tapping-mode scanning probe electrospray ionization (t-SPESI), which is an ambient sampling/ionization technique that uses a single capillary probe, was reported.¹⁰⁾ In t-SPESI, sampling/ionization is performed by oscillating a capillary probe in which a charged solvent is flowing. Sampling occurs in a liquid bridge formed between the probe apex and the sample surface, and the solution, including the chemical components that are extracted from the sample, undergoes ESI on the probe as the probe approaches the inlet of the mass spectrometer (MS inlet). The timing of the liquid bridge formation and the ESI are separated by the probe oscillation. Tapping-mode scanning probe electrospray ionization mass spectrometry imaging (t-SPESI-MSI) of tissue sections of mouse brain,¹¹⁾ mouse pancreatic cancer tissue,¹²⁾ and a polymer film¹³⁾ has been reported.

The spatial resolution of t-SPESI-MSI depends on the size of the liquid bridge. Previously, t-SPESI-MSI of mouse brain tissue sections achieved a spatial resolution of 35 μm by optimizing the experimental parameters and using a capillary probe (commercial silica tip emitter) with a 20 μm inner diameter (i.d.).¹⁴⁾ On the other hand, MSI of a tissue sample by nano-DESI, that also utilized a liquid bridge for sampling, indicated a sampling width of less than 8 μm and a spatial resolution of 12 μm using two aligned capillary probes with a 23 μm i.d.¹⁵⁾ This is the best sampling size utilizing solvent flow that has been achieved using an ambient sampling/ionization technique.¹⁾

In this report, we show the size decrease of the sampling spot for t-SPESI MSI with a nanopipette. Previously, nanopipettes were used to reduce the size of the liquid bridge for the deposition of gold particles on a solid substrate.¹⁶⁾ In order to clarify the size of the sampling spot, a dye molecular film on a glass substrate was used. The average diameter of the sampling spot was approximately 6 μm when a single oscillating nanopipette was used. Furthermore, the ion intensity was increased by reducing the flow rate of the solvent. MSI of an ink micropattern with a nanopipette is also described.

EXPERIMENTAL

SPESI instrumentation

A schematic illustration of the t-SPESI instrument is shown in Fig. 1. A commercial silica tip emitter (10 μm i.d., 20 μm outer diameter (o.d.), New Objective, USA) or a homemade nanopipette was used as a capillary probe. The capillary probe was connected to a syringe through a PEEK union, a metal line filter (6010–52600, GL Science) and a silica capillary (50 μm i.d., 150 μm o.d., GL Science). The flow rate of the solvent was controlled by means of a syringe pump (YSP-201, YMC Co., Ltd., Japan). The MS inlet (SUS tube, 10 cm length, 0.5 mm i.d. 1/16 o.d.) was connected to a time-of-flight mass spectrometer (JMS-T100LP AccuTOF, JEOL, Japan) with a laboratory-built part. The capillary probe was observed by a digital camera (GR-i68c, Shodensha, Japan) for position alignment. The oscillation of the capillary probe was controlled with an external piezo actuator (PMF3030, Nihon Ceratec Co., Ltd., Japan) that was attached to the probe. A function generator (WF 1946, NF Corp., Japan) and a voltage amplifier (BA4825, NF Corp.) were used to drive the piezo actuator. The bias voltage and frequency of the sine waveform for the piezo actuator were adjusted to the resonance point of the probe oscillation. Mass spectra were analyzed by Mass Center (JEOL). The data were also converted to netCDF (network Common Data Form) for MSI and to analyze the signal intensity. Original software for MSI was developed and used (LabVIEW 13.0, National Instruments Corporation, Japan). A sample was mounted on the auto-stage (XA-E35L-50, SUS Corp., Japan) and moved in the X- and Y-directions at pitches of 200, 100, or 50 μm every 2 s using a custom program written by the authors. All measurements were taken in the positive ion mode. The bias voltage of the solvent was 4.5 kV. The temperature of the MS inlet was 300°C.

Fig. 1. Schematic illustration of the t-SPESI instrumentation used in this study. The capillary probe is oscillated by a piezo actuator at their oscillation frequency. The timing of liquid bridge formation and the ESI are separated by the dynamic oscillation of the capillary probe.

Reagents

Methanol, acetonitrile (ACN), tetrahydrofuran (THF), formic acid, sodium iodide and Rhodamine B were purchased from Wako Pure Chemical Industries, Ltd., and used as received. Pure water was obtained from an automatic water purifier (RFD250NB, ADVANTEC, Japan).

Dye molecular film

An ethanol solution of Rhodamine B (10 mM, 150 μL) was cast on a glass slide (25×25 mm), and a dye molecular film was prepared by spin coating (4000 rpm, 5 s). A sodium iodide solution (100 μM) in ACN/THF (1/1, v/v) was used as the measurement solvent. The flow rate was controlled from 5 to 80 nL/min when nanopipettes were being used. The flow rate was controlled from 30 to 200 nL/min when silica tip emitters were being used.

Ink patterns

Two types of ink patterns were used for MSI. One was prepared by drawing a pattern on a glass slide (1.0±0.05 mm thick, 76×26 mm, AS ONE Corporation, Japan) with colored pens (red and blue Permanent Marker Sharpie, Sanford Manufacturing Company, USA). The other was prepared by printing a grid pattern on an overhead projector (OHP) polyethyleneterephthalate (PET) film (VF-1101N, KOKUYO Co., Ltd., Japan) with an ink jet printer. A 1% solution of formic acid in methanol was used for the solvent. The flow rates for the ink pattern on the glass slide and on the OHP film were 5 and 50 nL/min, respectively.

Nanopipettes

The nanopipettes were fabricated by extending a heated glass tube (GD-1, 0.6 mm i.d., 1 mm o.d., Narishige Group, Japan) with a puller (PE-22 Puller, Narishige, Japan). The nanopipettes were used for t-SPESI without any treatment. The nanopipettes had a minimum diameter of approximately 150 nm and the tip apex was observed by scanning electron microscope (SEM) (JSM-7600F, JEOL) after platinum coating.

RESULTS AND DISCUSSION

Shape and oscillation of nanopipette

Figure 2A shows a digital camera image of the nanopipette. The shank and shoulder were formed by pulling a heated glass tube. The shank length of the nanopipette was approximately 7 mm. SEM images of a nanopipette are shown in Figs. 2B and C. The average i.d. of freshly prepared nanopipettes used in this study was approximately 150 nm. Figure 2D shows the oscillation of a nanopipette observed by the digital camera. The nanopipette was oscillated at a frequency of 2104 Hz, with the node in the middle of shank. Length of the nanopipette from the node to the antinode was about 5 mm. Figure 2E shows the relationship between the amplitude of oscillation and the drive frequency of the piezo actuator. A different nanopipette was used in this measurement. The oscillation amplitude was measured from the peak to peak value from the camera image. The maximum amplitude was measured at an input signal frequency of 4093 Hz. The length of the nanopipette from the node to the antinode was about 3 mm. This frequency corresponds to the resonance frequency of the nanopipette, and it can be varied depending on the size of the shank. The oscillation frequency was adjusted for each nanopipette and the oscillation amplitude was adjusted to about 1 mm in this study. The oscillation of the probe needs to be at its resonant frequency in order to obtain the stable large amplitude. The large amplitude is an important factor in guiding ions from the electrospray into the MS inlet in this study.

Fig. 2. (A) Nanopipette observed by a digital microscope. (1) Heated MS inlet. (2) Shank and (3) shoulder of the nanopipette. (B), (C) SEM image of the nanopipette. (D) Oscillation of the nanopipette observed by digital microscope. The arrows show the oscillation of the apex of the nanopipette. The oscillation frequency was 2104 Hz. (E) The oscillation amplitude of the nanopipette vs. drive frequency of the nanopipette.

Sampling and ionization of the Rhodamine B film

The size of the sampling spot and ionization efficiency were examined by using a Rhodamine B film. The Rhodamine B film was sampled and ionized at a pitch of 200 or 100 μm in a two-dimensional area. Figure 3A shows a schematic illustration of the measurements. There were 15 sampling positions in the X-direction and 4 in the Y-direction. The accumulated mass spectrum of all the measurement points is shown in Fig. 3B. The peak at m/z 443 was assigned to the molecular ion [M−Cl]⁺. The peaks at m/z 399 and 371 were tentatively assigned to the fragment ions [M−Cl−COOH+H]⁺ and [M−Cl−CH₃−CH₂+H]⁺, respectively. The fragment ions were generated by the degradation of the Rhodamine B film during its storage in air.

Fig. 3. (A) Schematic illustration of the measurement of a Rhodamine B film by t-SPESI using the nanopipette. The sampling pitch was 200 μm for the silica tip emitter or 100 μm for the nanopipette. (B) Mass spectrum of Rhodamine B. (C), (D) Relationship between the flow rate and the spot diameter with the silica tip emitter and the nanopipette, respectively. The optical image of typical sampling spot is shown inside. The representative sampling spot is indicated by the arrow. Scale bar: 100 μm. (E) Correlation between the flow rate and the signal ion intensity for t-SPESI with a nanopipette.

The correlation between the type of probe and the sampling area was also examined. The size of the sampling spot formed by the silica tip emitter or the nanopipette was compared to elucidate the effect of the diameter of the opening of the capillary probe on the sampling spot size. In this study, we used a commercial silica tip emitter with a 10 μm inner diameter. The silica tip emitter is commercially available, and stable ionization is reported.¹⁴⁾ Sampling and ionization were performed at different solvent flow rates and the trace of the sampling area was measured with an optical microscope. The relationship between flow rate and spot diameter of the sampling area for the silica tip emitter is shown in Fig. 3C. The optical image of the sampling traces on the Rhodamine B film at a flow rate of 30 nL/min with the silica tip emitter is shown inside. The spot diameter was increased with increasing flow rate in the range 30–200 nL/min. Stable sampling was difficult at a flow rate 10 nL/min. Figure 3D shows that, for the nanopipette, the spot diameter of the sampling area was nearly constant as the flow rate increased in the range of 5–80 nL/min. An optical image of the sampling traces on the Rhodamine B film at a flow rate of 5 nL/min with the nanopipette is shown inside. This result shows that the spot diameter was decreased down by the nanopipette to 5–6 μm. This value is the smallest spot size ever reported for a single capillary probe as far as we know. The 5–6 μm spot size by using a nanopipette was confirmed, but this volume is much larger than the initial aperture of the nanopipettes. This can be attributed to the tip of the nanopipette being cracked. After the experiment, the apex of the nanopipette was observed by SEM. The aperture diameter of the nanopipette was approximately 2 μm, indicating that the nanopipette had been mechanically broken during the experiment. Because the fine apex of a nanopipette is fragile, the precise control of the loading force of the nanopipette to the sample surface is clearly needed. A feedback control system designed to maintain the oscillation amplitude to prevent stress to the tip of nanopipette will be reported elsewhere. The constant spot diameter obtained with the nanopipette, irrespective of solvent flow rate, can be attributed to the suppression of the spread of the solvent on the sample surface by the oscillation of the probe. The interval between sampling and ionization can be estimated from the oscillation frequency.¹⁰⁾ In this study, it was assumed to be approximately 244 μs, based on the oscillation frequency of the nanopipette (4093 Hz). This interval is much shorter than that of the silica tip emitter. The interval for the silica tip emitter (360 Hz) was estimated to be 2.78 ms. The increased number of ionizations per unit time owing to the high speed oscillation of the nanopipette would increase the consumption of the liquid solvent to the gas phase. As a result, the volume of the solvent in the liquid bridge was reduced. Moreover, because of the increased number of the liquid bridge formation, the solvent between the sample surface and the nanopipette per unit time would prevent the solvent from spreading on the surface. The results with the nanopipette were different from those with a silica tip emitter. The longer sampling/ionization interval and the increased outflow rate of the solvent due to the larger opening of the silica tip emitter would allow the diameter of the sampling area to increase with an increase in solvent flow rate.

The correlation between flow rate and ion intensity for t-SPESI with a nanopipette is shown in Fig. 3E. Ion intensity was increased by reducing the solvent flow rate in the range of 5–100 nL/min. The increase in ionization efficiency for the nanopipette with decreasing solvent flow rate can be attributed to the size of the charged droplets. The generation of smaller charged droplets at the beginning of the ESI process is essential for enhancing ionization efficiency as nano-ESI with lower flow rates.¹⁷⁾ In this study, the initial size of the charged droplets from a nanopipette would also be decreased with decreasing solvent flow rate. Moreover, the solution concentration of extracted chemicals at the distal end of the nanopipette would increase with decreasing solvent volume. This would also have an effect on the reduction of charged droplets.¹⁸⁾ The large error in the ion intensity at lower solvent flow rates would be due to fluctuations in solvent flow by the syringe pump and/or fluctuations in probe oscillation. The XY stage showed a slight fluctuation in the Z-direction, during the X–Y scanning. Therefore, we assume that the amplitude of the probe oscillation changes slightly during the measurement. At that time, the position of ESI was varied and hence, the amount of ions introduced to the MS inlet would be expected to fluctuate at each measurement position.

MSI of ink patterns

MSI of two ink patterns on a glass substrate and the OHP film were carried out in order to visualize multiple components contained in the mixed material. Figure 4A shows an optical image of the ink pattern after t-SPESI-MSI. The ink pattern was sampled and ionized at a 100 μm pitch in a two-dimensional area (X-direction 40 pixel, Y-direction 30 pixel). The spot diameter of the sampling area was 6.5–7 μm in this experiment. The accumulated mass spectrum of all measurement points is shown in Fig. 4B. There were three peaks in the mass spectrum. The ion peaks at m/z 478 and 434 were assigned as [M−Cl]⁺ and [M−Cl−NHCH₂CH₃]⁺, corresponding to basic blue 7 (blue Sharpie marker).¹⁹⁾ The ion peak at m/z 399 was assigned as [M−Cl−COOH+H]⁺, corresponding to Rhodamine B (red Sharpie marker).¹⁹⁾ The MS image of m/z 478 and 399 and their overlaid image are shown in Figs. 4C–E, respectively. The overlaid image corresponds to the optical image.

Fig. 4. (A) Optical image of the ink pattern on the glass substrate. Dotted line shows the area of MSI by t-SPESI. Scale bar: 1 mm. (B) Mass spectrum of the ink pattern. (C)–(E) MS image of m/z 478 and m/z 399 and their overlaid image, respectively.

The MSI of the ink pattern on the OHP film was also carried out. Figure 5A shows the optical image of the ink pattern. The line width and their spacing were 213 and 370 μm, respectively. The ink pattern on the OHP film was sampled and ionized at a 50 μm pitch in a two-dimensional area (X-direction 40 pixel, Y-direction 30 pixel). The accumulated mass spectrum of all measurement points is shown in Fig. 5B. Four different chemical groups with peaks separated by 44 Da were detected in the range of m/z 500–1000. The 44 Da intervals of the ion peak correspond to the molecular weight of the repeating unit (CH₂–CH₂O) of polyethylene glycol, which is used as an anti-drying agent for ink preparations. MS images for each of the chemical groups are shown in Figs. 5C–F. Ion intensities of m/z 683, 793, 825, and 897 were used for MSI because these four peaks show the highest ion intensities for each chemical group. The four MSI patterns are similar to the optical images. This result demonstrates the visualization of a polymer pattern by t-SPESI with a sampling spot diameter of 6–15 μm on the polymer film.

Fig. 5. (A) Optical image of the ink pattern on the OHP film. (B) Mass spectrum of the ink pattern. (C)–(F) MS images of m/z 683, m/z 793, m/z 825, and m/z 897, respectively.

CONCLUSION

The size of the sampling spot in t-SPESI could successfully be reduced by using a nanopipette. The diameter of the sampling spots on a dye molecular film was approximately 5–6 μm, which is smaller than the diameter achieved with a commercial glass capillary. The little dependence of spot size on the flow rate and an increase in ion intensity with decreasing the flow rate were confirmed. The reason for the constant spot diameter can be attributed to the increased consumption of the liquid solvent to the gas phase, thus reducing the volume of the liquid bridge and the increased capturing of the solvent between the sample surface and the apex of the nanopipette per unit time with the high speed oscillation of a nanopipette. The increase in ionization efficiency with the decrease in the solvent flow rate can be explained as the size reduction of charged droplets like nano-ESI. The t-SPESI-MSI of ink patterns on glass and polymer substrates was also demonstrated. Our results would indicate the potential of this technique for advancing the spatial resolution of ambient sampling/ionization MSI by t-SPESI via the use of a nanopipette.

Acknowledgments

We wish to thank Mr. T. Shirasawa of Shizuoka University for fabricating the nanopipettes. This study was partly supported by Grants-in-Aid for Scientific Research (No. 26505011) from the Ministry of Education, Culture, Sports, Science and Technology (MEXT) of Japan and by the Center for the Global Study of Culture Heritage and Culture, Kansai University. This work was partially supported by JSPS KAKENHI Grant Number JP16K21143.

REFERENCES

• 1) J. Laskin, I. Lanekoff. Ambient mass spectrometry imaging using direct liquid extraction techniques. Anal. Chem. 88: 52–73, 2016.
• 2) Z. Takáts, J. M. Wiseman, B. Gologan, R. G. Cooks. Mass spectrometry sampling under ambient conditions with desorption electrospray ionization. Science 306: 471–473, 2004.
• 3) M. Yamashita, J. B. Fenn. Electrospray ion source: Another variation on the free-jet theme. J. Phys. Chem. 88: 4451–4459, 1984.
• 4) P. J. Roach, J. Laskin, A. Laskin. Nanospray desorption electrospray ionization: An ambient method for liquid-extraction surface sampling in mass spectrometry. Analyst (Lond.) 135: 2233–2236, 2010.
• 5) G. J. Van Berkel, A. D. Sanchez, J. M. E. Quirke. Thin-layer chromatography and electrospray mass spectrometry coupled using a surface sampling probe. Anal. Chem. 74: 6216–6223, 2002.
• 6) T. Wachs, J. Henion. Electrospray device for coupling microscale separations and other miniaturized devices with electrospray mass spectrometry. Anal. Chem. 73: 632–638, 2001.
• 7) K. Hiraoka, K. Nishidate, K. Mori, D. Asakawa, S. Suzuki. Development of probe electrospray using a solid needle. Rapid Commun. Mass Spectrom. 21: 3139–3144, 2007.
• 8) L. C. Chen, K. Nishidate, Y. Saito, K. Mori, D. Asakawa, S. Takeda, T. Kubota, H. Hori, K. Hiraoka. Characteristics of probe electrospray generated from a solid needle. J. Phys. Chem. B 112: 11164–11170, 2008.
• 9) C. Wu, A. L. Dill, L. S. Eberlin, R. G. Cooks, D. R. Ifa. Mass spectrometry imaging under ambient conditions. Mass Spectrom. (Tokyo) 32: 218–243, 2013.
• 10) Y. Otsuka, S. Shide, J. Naito, M. Kyogaku, H. Hashimoto, R. Arakawa. Scanning probe electrospray ionization for ambient mass spectrometry. Rapid Commun. Mass Spectrom. 26: 2725–2732, 2012.
• 11) Y. Otsuka, J. Naito, S. Satoh, M. Kyogaku, H. Hashimoto, R. Arakawa. Imaging mass spectrometry of a mouse brain by tapping-mode scanning probe electrospray ionization. Analyst (Lond.) 139: 2336–2341, 2014.
• 12) Y. Otsuka, S. Satoh, J. Naito, M. Kyogaku, H. Hashimoto. Visualization of cancer-related chemical components in mouse pancreas tissue by tapping-mode scanning probe electrospray ionization mass spectrometry. J. Mass Spectrom. 50: 1157–1162, 2015.
• 13) R. Shimazu, Y. Yamoto, T. Kosaka, H. Kawasaki, R. Arakawa. Application of tapping-mode scanning probe electrospray ionization to mass spectrometry imaging of additives in polymer films. Mass Spectrom. (Tokyo) 3: S0050, 2014.
• 14) M. Lorenz, O. S. Ovchinnikova, V. Kertesz, G. J. Van Berkel. Controlled-resonant surface tapping-mode scanning probe electrospray ionization mass spectrometry imaging. Anal. Chem. 86: 3146–3152, 2014.
• 15) J. Laskin, B. S. Heath, P. J. Roach, L. Cazares, O. J. Semmes. Tissue imaging using nanospray desorption electrospray ionization mass spectrometry. Anal. Chem. 84: 141–148, 2012.
• 16) F. Iwata, S. Nagami, Y. Sumiya, A. Sasaki. Nanometre-scale deposition of colloidal Au particles using electrophoresis in a nanopipette probe. Nanotechnology 18: 105301, 2007.
• 17) A. Schmidt, M. Karas, T. Dülcks. Effect of different solution flow rates on analyte ion signals in nano-ESI MS, or: When does ESI turn into nano-ESI? J. Am. Soc. Mass Spectrom. 14: 492–500, 2003.
• 18) J. Fernández De La Mora, I. G. Loscertales. The current emitted by highly conducting Taylor cones. J. Fluid Mech. 260: 155–184, 1994.
• 19) J. F. Cahill, V. Kertesz, G. J. Van Berkel. Characterization and application of a hybrid optical microscopy/laser ablation liquid vortex capture/electrospray ionization system for mass spectrometry imaging with sub-micrometer spatial resolution. Anal. Chem. 87: 11113–11121, 2015.