Field of the Invention

The present invention relates to sample slides for laser

desorption ionisation (LDI) mass spectrometry and optical microscopy, and in particular to nanostructured indium-tin oxide (ITO) slides suitable for use in matrix-free LDI MS and imaging LDI MS,

Background of the Invention

Mass Spectrometry Imaging is a mass spectrometry based histology technique for the label-free analysis of proteins, metabolites, lipids or drugs on tissue samples . ^ll,2] A limitation of classic histology techniques is the limited the number and type of molecules that can be selectively visualized. On the contrary, mass spectrometry imaging has the advantage that it can detect simultaneously hundreds of analytes up to a mass limit of 30-40 kDa in a non-targeted fashion. ¹¹¹

While MALDI imaging has provided mass weighted images of

proteins, peptides and lipids, imaging of metabolites and small molecule drugs below 600 Da has been achieved by secondary ion mass spectrometry (SIMS) , ambient ionization methods like desorption electrospray ionization (DESI-MS) and laser ablation electrospray ionization (LAESI ) ^!3! or laser desorption ionization mass spectrometry (LDI-MS ) on nanostructured surfaces. In addition, the maximal image resolution obtained by matrix assisted tissue imaging mass spectrometry is determined by the size matrix crystals and can be further compromised by lateral analyte diffusion during matrix application . ^m Matrix-free LDI- MS based tissue imaging has the potential to produce tissue images with higher resolution than MALDI-Tof MS , higher

reproducibility and higher spatial fidelity as lateral metabolite diffusion during matrix application is avoided. The label-free imaging of the spatial distribution of metabolites and drugs on tissue sections by mass spectrometry has important potential applications in drug discovery, clinical histology, pharmacokinetics and basic molecular biology.

Small molecule tissue imaging by LDI-MS is compatible with instrumentation employed for MALDI-Tof MS and therefore highly complementary to standard MALDI-MS tissue imaging. From the large number of different LDI-active substrates only very few have been evaluated for imaging mass spectrometr . ^{[ ~6i} With one recent exception of a semi-transparent NI S-surface ^m , the reported materials cannot be used in standard histology protocols employing optical microscopy. An alternative are nanoparticle based approaches which substitute the organic matrix for a solution of suspended gold ^[8 > silver ^ί9ί or titanium oxide ^[10! nanoparticles or graphite . Although in these cases transparent ( ITO) sample slides can be used, sample preparation is tedious and can lead to the lateral analyte diffusion and heterogenous nanoparticle distribution notorious for matrix based methods. Indium tin oxide coated glass slides have been the material of choice for imaging MALDI-Tof MS application due to their

excellent conductivity and transparency required to analyse the same sample both by mass spectrometry and optical microscopy. Recently, Goodwin et al . showed that employing a highly sensitive 12T Fourier Transfer Ion Cyclotron Resonance (FTICR) mass spectrometer phospholipids could be detected from a tissue sample on commercial indium tin oxide slides without the application of matrix ^[12! , the sensitivity of conventional MALDI-Tof MS

instruments however is not sufficient for this application.

We have recently reported on the use of indium tin oxide coated slides for the preparation of glycan and protein microarrays for their analysis by MALDI-Tof mass spectrometry. ¹¹³¹ Summary of he Invention

Broadly, the present invention is based on the inventors' work to develop a cost-effective, transparent and conductive material for the matrix-free imaging mass spectrometry that is capable of use on a standard MALDI-Tof mass spectrometer. The present invention is based on investigations into the preparation of indium tin oxide coated glass slides by radio frequency (r . f , ) magnetron sputtering and their application as Li3I-active surfaces for small molecule analysis and tissue imaging.

Conventional (MA) LDI and FTICR MS are conceptually so different that it is safe to talk about two different techniques. FTICR is the most sensitive and precise mass spectrometry method currently available, we have shown that LDI-MS on a MALDI-Tof spectrometer equipped with a photomultiplier does not produce signals on commercial ITO slides. In this context the acquisition parameters are also important. We have included them in the supporting information. At higher fluence signals can be obtained from nearly any surface material, the clue is to have good signal under comparably soft ionization conditions which maintain the integrity of the analytes and do not evoke strong background signals coming from the surface of the material. Accordingly, in a first aspect, the present invention provides a sample slide for laser desorption ionisation (LDI) mass

spectrometry and optical microscopy, wherein the slide is formed from a glass substrate coated with a nanostructured indium tin oxide (ITO) surface layer, wherein the slide is capable of analysing a sample by matrix free laser desorption ionization

(LDI) mass spectroscopy and of visualising the sample by optical microscopy, wherein the indium tin oxide surface layer is between 10 nm and 150 nm in thickness, has a surface roughness (Rq) of at least 0.800 nm as determined by atomic force microscopy and has a resistivity between 1.0 and 20.0 microOhm/cm ² .

Additional and preferred properties and characteristics of the sample slides of the present invention are described in more detail below.

In a further aspect, the present invention provides a method of making a sample slide capable of use in laser desorption

ionization (LDI) mass spectrometry and optical microscopy, the method comprising sputtering a glass substrate with an indium tin oxide surface layer under conditions to produce a sample slide having an indium tin oxide surface layer between 10 nm and 150 nm in thickness, a surface roughness (Rq) of at least 0.800 nm as determined by atomic force microscopy and a resistivity between 1.0 and 20.0 microOhm/cm ² . In a further aspect, the present invention provides a sample slide obtainable by the methods described herein.

In a further aspect, the present invention provides the use of a sample slide of the present invention in the detection of an analyte using matrix-free laser desorption ionization mass spectrometry (LDI-MS ) and imaging laser desorption ionization mass spectrometry (LDI-MS ) .

In a further aspect, the present invention provides a method of detecting an analyte using matrix-free laser desorption

ionization mass spectrometry (LDI-MS) and imaging laser

desorption ionization mass spectrometry (LDI-MS) , the method comprising:

depositing the analyte onto the sample surface of a sample slide of the present invention;

irradiating the sample surface with a laser to desorb and ionize the analyte;

detecting the ionized analyte. Embodiments of the present invention will now be described by way of example and not limitation with reference to the accompanying figures. However various further aspects and embodiments of the present invention will be apparent to those skilled in the art in view of the present disclosure.

"and/or" where used herein is to be taken as specific disclosure of each of the two specified features or components with or without the other. For example "A and/or B" is to be taken as specific disclosure of each of (i) A, (ii) B and (iii) A and B, just as if each is set out individually herein. Unless context dictates otherwise, the descriptions and

definitions of the features set out above are not limited to any particular aspect or embodiment of the invention and apply equally to all aspects and embodiments which are described. Brief Description of the Figures

Figure 1A/B. LDI-MS spectra of l-aminohexaethylenglycol analyte (MW 281.33) in all assayed set of substrates, including the commercial ones, using same acquisition parameters. Figure 2. Selection of lower weight molecules analysed by LDI-MS on nanostructured ITO sample slides of the present invention.

Figure 3. Selection of higher weight molecules analysed by LDI-MS on nanostructured ITO sample slides of the present invention.

Figure 4. LDI spectra of sodium and potassium adducts for 1000 pmol, 60 pmol, 25 pmol and 0.5 pmol of on surface spotted creatinine analyte. Figure 5. Performance parameters for lactose quantification in milk by isotopic dilution a) dynamic range b) limit of detection and limit of quantification

Figure 6. LDI-MS spectra of lactose, labeled ¹³ C-lactose, milk sample, milk sample with the addition of lactose standard and, the lower one, the lactose free milk with the standard.

Figure 7. MS spectra in positive and neqative mode of brain tissue sample on ITO_4 and hydrophobically modified ITO_4.

Figure 8. Spatial distribution of selected metabolite ions in mouse brain tissue sample at 50 μιη resolution with no washing procedure . Figure 5. A) Spatial distribution and tentative assignment of detected peaks by MSI in negative mode after washing procedure (right: scanned area) . B) Spatial distribution image of lipids distributed within the corpus callosum (right: scanned area) .

Figure 10. MS/MS spectra and proposed structure for metabolite ion at 888 Da.

Detailed Description

Nanostructured ITO Slides

The nanostructured ITO slides of the present invention and five commercial indium tin oxide slides from different sources

(Bruker, Hudson Technologies and VisionTek®) were tested as potential substrates for matrix free LDT-MS by measuring a range of different properties of the slides including elemental composition (by X-ray photoelectron spectroscopy, XPS) , topology (atomic force microscopy, AFM, and scanning electron microscopy, SEM) , optical and electrical properties (transmission/reflectance spectroscopy and ellipsometry) and hydrophobicity (contact anglemeter) . Furthermore, all were evaluated as mass

spectrometry sample slide for both MALDI-Tof and LDI-MS of small molecules .

A particularly preferred aim of these experiments was to find sample slides that are capable of use in both laser desorption ionisation (LDI) mass spectrometry, and optional also optical microscopy. The sample slides of the present invention have the advantage that there are capable of being used for the matrix free laser desorption ionization (LDI) mass spectroscopy

techniques, such as matrix-free laser desorption ionization mass spectrometry (LDI-MS) or imaging laser desorption ionization mass spectrometry (LDI-MS) , providing good signals under comparably soft ionization conditions which maintain the integrity of the analytes and do not evoke strong background signals coming from the surface material. This distinguishes the performance and use of the slides of the present invention from techniques, such as

Fourier Transform Ion Cyclotron Resonance (FTICR) mass spectroscopy employed by Goodwin et al (supra) . FTICR-MS differs significantly from other mass spectrometry techniques in that the ions are not detected by hitting a detector such as an electron multiplier but only by passing near detection plates.

Additionally, the masses are not resolved in space or time as with other techniques but only by the cyclotron (rotational) frequency that each ion produces as it rotates in a magnetic field. Thus, the different ions are not detected in different places as with sector instruments or at different times as with time-of-flight instruments but all ions are detected

simultaneously during the detection interval. This provides an increase in the observed signal to noise ratio owing to the principles of Fellgett ' s advantage. In FTICR-MS, resolution can be improved either by increasing the strength of the magnet (in teslas) or by increasing the detection duration.

From the studies described herein, none of the studied commercial slides showed any LDI performance at all even at high fluence and high analyte concentrations. Without wishing to be bound by any particular theory, the present inventors believe that the beneficial properties observed for the slides of the present invention, but not in any of the commercially slides is not the result of a single feature of the slides which explains this behaviour in sputtered substrates, but rather the effect is result of a beneficial combination of surface parameters.

As explained further below, the samples slides of the present invention are preferably formed from a glass substrate coated with a nanostructured indium tin oxide (ITO) surface layer.

While indium tin oxide (ITO) thin films can be obtained either by evaporation (thermal or e-beam) or sputtering (d.c. or r . f . ) , the films prepared by sputtering show the lowest resistivity and highest transmission. Consequently, and as explained below, preferably sputtering is the method of choice for making the ITO surfaces of the slides of the present invention. Preferably, in one set of surface characteristics of the samples slides of the present invention, the indium tin oxide surface layer is between 10 nm and 150 nm in thickness, the slide has a surface roughness (Rq) of at least 0.800 nm as determined by atomic force microscopy and has a resistivity between 1.0 and 20.0 microOhm/cm ² .

The thickness and/or resistivity of the surface layer of a sample slide can be determined using techniques well known to the skilled person, such as ellipsometry . The sample slide may have an indium tin oxide surface layer between 10 nm and 75 nm in thickness, more preferably has an indium tin oxide surface layer between 15 nm and 60 nm in thickness, and more preferably has an indium tin oxide surface layer between 20 nm and 50 nm in thickness. The sample slide may have a resistivity between 1.2 and 10.0 microOhm/cm ² and more preferably has a resistivity between 1.4 and 2.0 microOhm/cm ² . The resistivity of the samples slides is a property that reflects the need for the sample slides to have good electrical conductivity to enable their use in mass spectrometry. Sample slides having a low resistivity are achieved by increasing the thickness of the ITO layer to help to improve their properties when used in mass spectrometry.

However, optical microscopy requires good transparency and this property is improved with a thin ITO layer. Consequently, the thickness of the ITO layer and/or its resistivity are a balance between these competing needs.

The sample slide may have an average grain size between 40 nm and 100 nm, more preferably has an average grain size between 50 nm and 100 nm. The adjustment of the grain size can improve absorption at a desired wavelength.

A further preferred characteristic of the sample slides of the present invention is the extent to which surface layer of ITO is nanostructured, i.e. that it has a degree of surface roughness that is believed to assist in the matrix free ionisation of samples placed on the surface and ionised in a mass spectrometry experiment. This has the advantage of assisting in the energy transfer from the excitation source used to ionise the sample, and preferably means that this can be achieved under

comparatively soft ionisation conditions. The surface roughness (Rq) of a sample slide may be determined by atomic force

microscopy. Preferably, the sample slides have a surface roughness of at least 0.900 nm, more preferably a surface roughness (Rq) of at least 1.000 nm as determined by atomic force microscopy, more preferably a surface roughness (Rq) of at least 1.100 nm, and more preferably a surface roughness (Rq) of at least 1.200 nm. By way of example, surface topology of the sample slides of the present invention, including surface roughness may be characterised using an atomic force microscope such as a JPK NanoWizard® AFM in intermittent contact tapping mode using a TESP probe with a 0.01 - 0.025 Q/cm Antimony (n) doped Si, rectangular, 4 μιτι thick cantilever with a nominal resonant frequency, spring constant, length and width of 320 kHz, 42 N/m, 125 μπι and 40 μπι respectively. As mentioned above, preferably the sample slides of the present invention have a degree of optical transparency that enables them to be employed in optical microscopy. This enables the sample slides to be used for histological protocols as well as mass spectrometry. Optical transparency may be measured by

determining the optical transmission at a wavelength of visible light and determining the percentage of light transmitted through the slide, for example using transmission/reflectance

spectroscopy as demonstrated in the examples. Preferably, the sample slides of the present invention have an optical

transmission at 355 nm of at least 50%, more preferably an optical transmission at 355 nm of at least 55%.

more preferably an optical transmission at 355 nm of at least 60%, and most preferably an optical transmission at 355 nm of at least 65%. By way of illustration, optical transmission may be determined from the transmission spectra of the TTO-coated glass slides using a UV Beckman Coulter DU 800 UV/VIS/NIR

spectrophotometer equipped with an adapter for the sample slide in transmission mode from 200 up to 800 nm. The wavelength interval as well as the scan speed may be set at 0.2 nm and 240 nm/min, respectively. In one preferred embodiment, the present invention provides a sample slide with an advantageous combination of characteristics reported in the examples below as ITO_4. This sample slide has an indium tin oxide surface layer between about 25 nm and about 35 nm in thickness, a surface roughness of about 1.2 nm, a resistivity of between 1.5-1.9 microOhm/cm ² and an optical transmission at 355nm of greater than 60%.

For some applications it may be useful for the surface of the sample slide to comprise free hydroxyl groups capable of surface functionalization by covalent reaction with one or more

components present in a sample. The presence of free hydroxyl groups may be determined by measuring the degree of

hydrophobicity of the surface of the sample slide, using

apparatus such as a contact anglemeter and expressed in terms of the contact angle. A smaller contact angle and higher

hydrophilicity observed is consistent with a higher hydroxyl group content of a surface. Preferably, the sample slides of the present invention have a surface hydrophobicity contact angle (Θ) of 25 degrees or less, and more preferably have a surface hydrophobicity contact angle (Θ) of 20 degrees or less.

Preparation of Nanostructured ITO Slides

The samples slides of the present invention are formed from a glass substrate coated with a nanostructured indium tin oxide (ITO) surface layer to provide the surface with the

characteristics that mean it is capable of being used for laser desorption ionisation (LDI) mass spectrometry, and optionally also optical microscopy. While indium tin oxide (ITO) thin films can be obtained either by evaporation (thermal or e-beam) or sputtering (d.c. or r . f . ) , the present invention provides a method of producing sample slides having ITO surface layers by sputtering. Preferably, this results in sample slides which combine a relatively low resistivity and relatively high optical transmission. The present inventors also found that ITO sample slides produced by r.f. sputtering show high homogeneity and reproducibility, important parameters for surface-based mass spectrometry applications.

Sputtering is a technique for depositing thin films on

substrates, see htt : //en . wikipedia . org/wiki/Sputter deposition. In a preferred embodiment, the ITO films were deposited on glass substrates by a r.f. magnetron sputtering system (AJA ATC 1800

UHV) on glass substrates, such as Corning 2948. A two inch, 90% Ιη2θ3 and 10% Sn02 in weight (99.99% purity) target was used as sputtering source . The sputtering chamber was evacuated pr i or to sputtering to a base pressure of around 10 ^~8 mTorr , and then backfilled with Argon gas. The rotation speed (80 rpm) and temperature (room temperature) were kept constant for all the set of experiments. The skilled person can select sputtering parameters suitable for producing the sample slides based on the experiments described herein. Preferably, the sputtering employs a radiofrequency (R.F.) power (W) of 100W or less, more

preferably a radiofrequency (R.F.) power (W) of 60W or less and most preferably a radiofrequency (R.F.) power (W) of SOW or less. Generally, the sputtering employs an Argon pressure of 6 mTorr or less, or an Argon pressure of 3 mTorr or less. The sputtering flux may be a flux of 50 seem or less, and more preferably a flux of 25 seem or less. Deposition t i mes are generally in the range of 30 minutes to 300 minutes , and may be 30 minutes to 180 minutes. Preferably, the sputtering is carried out in the absence of oxygen.

Perivatisation to produce a microarray

Microarrays are now a commonly used format for most high- throughput screening applications in genomics, proteomics and glycomics and the ability to analyse them by more than a single readout method is desirable to broaden their applications. US 61/777202, and US 2014/0274771 Al (US 14/203,611) filed on 11 March 2014 that claims the priority of US 61/777202, which are herein incorporated by reference in their entirety, describe novel microarrays and methods of making these novel microarrays which may be assembled on a sample slide according to the present invention to provide a microarray suitable for use in SALDI-MS methods as described herein.

Accordingly, in a further aspect, the present invention provides a method of making a microarray on a surface of a sample slide according to the present invention, the method comprising:

(a) providing sample according to the present invention;

(b) forming a support layer of hydrophobic molecules attached to the surface of the sample slide;

(c) forming a layer of linker molecules on the surface of the sample slide, wherein the linker molecules comprise a hydrophobic group capable of non-covalently binding to the support layer and a reactive functional group; and

(d) printing a plurality of binding agents at a plurality of locations on the sample slide, wherein the binding agents comprise a functional group capable of reacting in situ on the microarray with the reactive functional group of the linker molecules to covalently link the binding agents to the linker molecules immobilized on the sample slide, thereby forming the microarray.

The present invention further provides a microarray supported on a sample slide according to the present invention as obtainable by this method or any other method described in US2014/0274771 Al (US 14/203,611) filed on 11 March 2014 that claims the priority of US 61/777202 filed on 12 March 2013, the contents of which are expressly incorporated by reference.

The method may further comprise derivatising the surface of the sample slide prior to adsorbing a sample. Use of the sample slides in kits with isotopically tagged glycan standards

There exists an unmet need for improved methods for rapidly and easily analysing the content of released glycan mixtures and solutions. The identification and quantification of glycans and particular glycan signatures may be useful for the diagnosis of numerous diseases and disorders.

WO2014/161960, which claims priority from GB1305986.0, which is herein incorporated by reference in its entirety, describes methods and materials for the identification and quantification of the glycan content of sample using isotopically-labelled glycan standards (tagged standards), and provides methods for the preparation of suitable isotopologues for use in these methods. In particular, WO2014/161960 provides a kit for identifying a glycan in a sample. These standard, or corresponding ones for other analytes such as clinical metabolites, drugs, small molecules or biopharmaceuticals , may be provided in a kit comprising a sample slide of the present invention and one or more further reagents for carrying out laser desorption

ionization (LDI) mass spectrometry, these reagents optionally including :

(a) one or more isotopically tagged standards of analytes suspected of being present in a sample analysed using the sample slide; and/or

(b) instructions for doping a sample suspected of

containing one or more of the analytes with the tagged

standard (s) to obtain a doped sample and analysing the doped sample using mass spectrometry.

Methods of synthesis and preparation of suitable isotopically- labelled glycan standards are provided more fully in GB1305986.0.

(a) one or more isotopically tagged standards of analytes suspected of being present in a sample analysed using the sample slide; and/or

(b) instructions for doping a sample suspected of

containing one or more of the analytes with the tagged standard ( s ) to obtain a doped sample and analysing the doped sample using mass spectrometry.

In a further aspect, the present invention provides a kit as described in GB1305986.0 further comprising a sample slide formed as described herein according to any described embodiment of the present invention. In some preferred embodiments, the sample slide comprises a hydrophobic layer, for example, an organic silane, e.g. an alkyl silane such as octadecylsilane (OTS) , to facilitate analysis of the glycan mixtures. It is envisaged that such a kit, comprising one or more tagged standards and a sample slide formed as described herein may have utility as a convenient and disposable kit for the rapid analysis of samples suspected of containing glycans .

Applications for the nanostructured ITO sample slides

The nanostructured indium-tin oxide (ITO) sample slides of the present invention are well adapted for a range of applications, in particular those in which laser desorption ionisation (LDI) mass spectrometry and optical microscopy are used in combination. The slides are suitable for use in matrix-free LDI MS and Imaging LDI MS. The samples slides of the present invention may be used for the matrix-free imaging mass spectrometry of small molecules, conventional matrix-assisted imaging of larger lipids and proteins and traditional histology by optical microscopy. Other applications include the general matrix-free analysis of small molecules, metabolites and drugs on standard MA.LDI-Tof mass spectrometers, at a fraction of the time and price of ESI-tof spectrometers working in solution phase. Further

functionalization of the surface with capture ligands or enzyme substrates permits the development of array based high-throughput assays for applications in diagnostics and drug discovery.

Other applications include:

Chemoselective targeted metabolomics , glycan, lipid, peptide or other small molecule arrays for enzyme inhibitor screening. Screening for new enzyme functions for applications in

biotechnology cellulose, starch and other polysaccharide

processing enzymes.

Analysis of samples for applications in the field of forensics.

Detection and identification of metabolites in biofluids. Lactose quantification for applications in the dairy industry.

Applications in the field food composition analysis, for example wines, olives etc. Glycan analysis in biopharmaceutical development of recombinant proteins, such as polypeptides, antibodies.

Specific uses in combination with optical properties: Rapid screening of potential binders by fluorescence and follow- up on hits by on-chip digestion (in the case of proteins) , and mass spectrometric identification of diagnostic peptides.

Imaging of the same tissue sample with 1. fluorescent probes or by optical microscopy and 2. by mass spectrometry.

Combination of mass spectrometry and confocal optical microscopy and mass spectrometry and Raman imaging as hybrid techniques on the same sample plate, in particular for applications in life and material science.

Examples

General methods

The commercial ITO slides (75 mm x 25 mm) were purchased from Hudson Surface Technology, Inc. (NJ, US), Bruker Daltonics

(Bremen, Germany) and VisionTek Systems Ltd (Cheshire, UK) and used for comparison with the sample slides of the present invention .

Deposition of ITO films

ITO films were deposited by a r.f. magnetron sputtering system (AJA ATC 1800 UHV) on glass substrates {Corning 2948) . A two inch, 90% ln ₂ 0 ₃ and 10% Sn0 ₂ in weight (99.99% purity), target was used as sputtering source. The sputtering chamber was evacuated prior to sputtering to a base pressure of around 10 ^~8 mTorr, and then backfilled with argon gas. The rotation speed (80 rpm) and temperature (room temperature) were kept constant for all the set of experiments.

Structure and morphology characterization

The surface topology was characterized with an atomic force microscope (JPK NanoWizard® AFM) in intermittent contact tapping mode using a TESP probe with a 0.01 - 0.025 Ω/cm Antimony (n) doped Si, rectangular, 4 μιη thick cantilever with a nominal resonant frequency, spring constant, length and width of 320 kHz, 42 N/m, 125 xm and 40 um respectively. Images were obtained with a line rate of 0.6 Hz, initial setpoint of 500 mV, integral gain of 40.0 Hz and proportional gain of 0.002 with a pixel resolution of 512 x 512. An area of 8 x 8 μπι was scanned.

Surface images were edited with JPK Data Processing software and the roughness measured as Rq (RMS) , Ra (average) and peak-to- valley value in the 8 8 μπι scanned images.

In order to compare the surface roughness, the cross-sections obtained from AFM images were traced in a vertical section from the upper left to the bottom right edge of the 8 x 8 μπι image.

Scanning electron microscopy ( JEOL 6490LV) was also used to characterize the surface topography of the samples. Images were taken at 25 kV acceleration voltage, working distance of 10 mm and magnification of x25000. Grain Size Calculation

The average grain diameter was calculated using the WSxM 4.0 software package, anotec Electronica S.L., Spain. To obtain area values, the Flooding option was used, which determines the number of hits (grains) in an image and the total area they represent at a certain height. By dividing the total detected area by the number of hits, the average grain diameter was obtained.

Elemental composition by X-ray photoelectron spectroscopy

XPS experiments were performed in a SPECS Sage HR 100

spectrometer with a non monochromatic X-ray source (Magnesium K« line of 1253.6 eV energy and 250 W) and calibrated using the 3ds/2 line of Ag with a full width at half maximum (FWHM) of 1.1 eV. The selected resolution for the spectra was 30 eV of Pass Energy and 0.5 eV/step for the general survey spectra and 15 eV of Pass Energy and 0.15 eV/step for the detailed spectra of the different elements. All measurements were made in an ultra high vacuum

(UHV) chamber at a pressure below 5 ·10 ^~θ mbar. In the fittings Gaussian Lorentzian functions were used (after a Shirley

background correction) where the FWHM of all the peaks were constrained while the peak positions and areas were set free.

Ellipsometry and Hydrophobici ty

The thickness of the ITO films deposited on silicon wafer was measured with a rotating spectroscopy ellipsometer (W2000V J. A

Woollam Co., Inc.) at several angles of incidence between 50 and 70 degrees. Analysis of the ellipsometry measurements was carried out using an ITO parameterized function with a model dielectric function having a combination of Lorentz oscillators and Drude term. The contact angle of each ITO deposited film was measured per triplicate using a Kriiss DSA 100 instrument. The dosing volume was 5 μΐ .

Mass spectrometry: instrumentation and conditions

Mass spectra were recorded on an Ultraflextreme III time-of- flight mass spectrometer equipped with a pulsed Nd : YAG laser (λ =

355 nm) and controlled by FlexControl 3.3 software (Bruker Daltonics, Bremen , Germany) . The acquisitions (total of 3000) were carried out in positive or negative reflector ion mode with pulse duration of 50 ns and frequency of 1000 Hz, laser fluence of 40-60 % and the following laser focus settings: Offset = 0 %, Range = 100 % and Value = 12 % . The m/z range was chosen according to the mass of the sample. The accumulated spectra were then treated with the FlexAnalysis v3.3 software.

For lactose analysis, the measurements were performed always in the same way: total of 3000 random walk shots per replica, ion source 1 = 25.18 kV, ion source 2 = 22.33 kV, lens voltage = 7.53 kV, reflector voltage = 26.37 kV, reflector voltage 2 = 13.50 kV, and positive reflector mode in a mass range of 100-600 Da. The laser energy has been optimized to ~ 55 pJ/pulse.

LDT-MS images were constructed using Flexlmaging 2.1 software with a spatial resolution of 20 - 70 μπι, in both positive and negative reflector ion mode, 500 shots per spectrum, a laser fluence of 55% and the following laser focus settings: Offset = 0 %, Range = 100 % and Value = 85 %.

Lactose quantification

The validation of the analytical method using ¹³ C fully labeled lactose as internal standard was determined with the linearity range, the reproducibility, the limit of detection (LOD) and the limit of quantification (LOQ) . For the linearity range, LOD and LOQ, 6 different concentrations of labeled lactose were prepared (336, 168, 129, 60, 17, 8.5 mg/1) in a milk working solution. The working solution was prepared by 1 to 10 dilution of the sample (100 μΐ to 1 ml) with water followed by a second 1 to 10 dilution with 70% of acetonitrile (1:100 final dilution) . The final dilution was centrifuged {13000 x g for 2 min) and the supernatant recovered. These solutions were directly spotted (0.5 ul) by triplicates over the ITO coated glass slides. Once dried, they were submitted to LDI analysis. A 1 πιΜ stock of "G ^a lactose standard (Omicron Biochemicals Inc.) was prepared in 70% acetonitrile . To 10 μΐ of lactose-free milk, 25 μΐ of 1 mM of standard and 65 μΐ of water were added. 10 μΐ of this solution were first diluted to 200 μΐ with 70% of

acetonitrile, making a 1:200 milk dilution with a ¹ C-lactose standard concentration of 12.5 μΜ. Six sample preparations were measured in triplicate to examine the reproducibility of the method. The lactose quantification was performed considering the area under the peaks of the sodium and potassium adducts of the analyte and the internal standard.

Imaging, surface modification

LDI-MS imaging tissue preparation. A Leica CM-3050S cryostat was used to slice mouse brain tissue. Tissues were sliced to 4 and 10 μιη at -20°C and deposited on the surfaces. Optimal tissue thickness was found to be 4 μιχι.

ITO coated slides were hydrophobically modified as described in Beloqui et a 1. , 2013. After the deposition of the tissue slices, they were stored overnight into a vacuum chamber. These samples were directly used for LDI-MS experiments. Hydrophobically coated slides were subjected to aqueous washing three steps of 15 sec of sonication, dried and measured. Metabolite Identification

Metabolites (adenosine diphosphate (ADP) , stearic acid, glycerol monophosphate, palmitic acid) were tentatively assigned on the basis of exact mass (m/z) queries using the databases Human

Metabolome Project

(http://redpoll.pharmacy.ualberta.ca/hmdb/HMDB/) and ETLIN

(http://metlin.scripps.edu/). The MS/MS spectra were obtained in negative ionization mode from accumulation of at least 2000 laser shots and using a pulsed ion extraction of 70 ns . The parent mass was selected manually at m/z 888 (monoisotopic peak) , and spectra were acquired in the range of m/z 50-900. The following voltages were used for the analysis: acceleration voltages of 7.5 kV for IS1 and 6.85 kV for IS2, reflector voltages of 29.5 kV for Rl and 13.95 for R2, and LIFT voltages of 19 kV and 3.3 kV for LI and L2, respectively. Results

Sample slide production

Indium tin oxide thin films can be obtained either by evaporation (thermal or e-beam) or sputtering (d.c. or r.f.) but films prepared by sputtering show the lowest resistivity and highest transmission ¹¹⁴ ' . Sputtering parameters were optimised to obtain thin films with improved transmission and desired photoelectric properties. The control of the sputtering process to produce nanostructured ITO surfaces that enable matrix-free LDI -MS has not been previously described.

Parameters for magnetron sputtering depositions on glass

substrate (Corning® 2948-75x25) were firstly set up for eight different tin oxide thin films, designated ITO 1 to ITO 8, with a targeted layer thickness of around 20-50 nm. The deposition time was adjusted to the chosen values for r.f. power and Ar gas or

Rr/O-2 mixture pressure. A second batch of slides with a thicker layer of indium tin oxide, by increasing the sputtering time to 300 minutes, was performed and specified as ITO 10-13. Table 1 summarizes the detailed sputtering parameters employed for the preparation of samples I 0_1 to I O__13. Changes in these parameters had a strong effect on the surface hydrophobicity, optical transmission, roughness, resistivity and performance in MALDI-Tof and LDI -MS .

Sputtering parameters

Sample R.F. Working Deposition Deposition

P Ar 02

Power distance rate time

(mTorr) (seem) (seem)

(W) (mm) (nm/H) (min)

ITO_l 30 30 3 24 48 50

ITO_2 60 30 3 24 114 21

ITO_3 60 30 3 10 20 28 40

ITO_4 20 30 3 24 15-21 100

ITO_5 10 30 3 24 10 180

ITO_6 10 10 3 24 10.3 180

ITO__7 10 30 3 24 1 4.0 180

ITO_8 30 30 6 24 — 40 60

ITO_10 10 30 3 24 10.2 300

ITO_ll 20 30 3 24 23.2 300

ITO_12 30 30 3 24 43.2 300

ITO_13 60 30 3 24 97.0 180

Table 1 . , Sputtering parameters for the preparation of surfaces ITOl-13

Trying to understand this effect , and being aware that there are not other examples in the literature showing LDI. -MS activity in

ITO surfaces, five commercial indium tin oxide slides from

different sources (Bruker, Hudson Technologies and VisionTek®) that covered a range of measurable properties (thickness,

roughness, resistivity, UV absorption) were also tested as

potential substrates for LDI-MS . For that, all surfaces,

including the commercial ITO-slides, were characterized in terms of elemental composition (by X-ray photoelectron spectroscopy,

XPS) , topology (atomic force microscopy, AFM, and scanning

electron microscopy, SEM) , optical and electrical properties

(transmission/reflectance spectroscopy and ellipsometry) and

hydrophobicity (contact anglemeter) . Furthermore, all were

evaluated as mass spectrometry sample slides for both ALDI-Tof and LDI-MS of small molecules (results in Table 2 and Figure 1) . LDI Average

Contact

MM Ellipsometr Spectroscopy Perform Grain

Angle

ance Size

Sample

Resistivity Hydropho

Rq Thickness Transmission

(rim) ( 10 ^"3 bicity d (am)

(nm) at 355 ran, %

ohm ·cm ² ) (Θ)

Hudson 3.343 130 0.1 66.1 19.9 198

Bruker 0.564 25 l.i 78.5 n.d.

Visiontek

9.079

4Q ~ 420 0.06 62.9 n.d. 95

Visiontek

1.865 130 0.6 64.8 n.d. 328 15Q

Visiontek

0.585 20 1.1 75.2. n.d.

100Q

Corning

1.085 n.d. n.d. 89.9 n.d.

Glass

ITO 1 0.871 40 1.4-2 81.0 17.5 42

ITO 2 0.865 40 1.6-1. 84.5 17.9 46

ITO 4 1.200 25-35 1.5-1.9 66.8 17.7 60

ITO 5 0.987 30 9-10 86.4 23.8 52

ITO 6 1.003 31 13-14 85.3 22.5 44

ITO 8 1.407 40 10-15 79.7 7.3 44

ITO 10 3.021 51 54.1 n.d. 45

ITO 11 2.182 116 1.1-1.4 48.1 n.d. 62

ITO 12 1.990 216 0.3-0.5 17.7 n.d. 74

ITO 13 4.083 291 0.0 n.d. 65

Table 2. Properties and characteristics of sample slides

according to the present invention as compared to commercially available slides.

In general, a higher surface roughness measured by atomic force microscopy (AFM) and visualized by scanning electron microscopy (SEM) was obtained for lower r.f. power (lower indium tin oxide concentration) and lower Argon pressure (lower deposition rate) both favoring crystal growth to larger grain sizes. Addition of oxygen as a reactive gas during sputtering (ITO_3 and I O_7 ) led to surfaces that were either nonconductive (ITO_3) or showed no LDI activity at all (ITO_3 and ITO_7) . Elemental surface composition analysis by XPS and subsequently fitting of the 0 Is peak showed a higher percentage of free hydroxyl functions for slides ITO 4, 5 and 8, an important factor for later covalent surface functionalization . The smaller contact angle and higher hydrophilicity observed was in agreement with the higher hydroxyl group content of theses surfaces.

C- Ό,

Substrate Sn In Ototal In-O-H, oo c

Si -0

ITOl 4.0 37 , .5 47. ,5 4. 7 14.0 11, .0

IT02 4.6 38 , , 8 46. , 0 4. 7 8.1 10. , 6

IT03 3.5 37, .4 46, , 9 3. 8 9.3 12. .2

IT04 4.0 37, .8 46. , 6 5. 5 15.0 11, .7

IT05 3.6 36. ,5 46. .2 4. 3 15.4 13, .7

IT06 3.5 35. .9 46. ,2 4. 2 13.2 14 , .3

IT07 — — —

IT08 3,3 46. ,8 46. ,8 5. 1 16.9 16. ,0

ITO10 2.6 32. , 1 50, , 1 4. 69 18.9 15. ,2

ITOll 2.9 33. .9 50. .6 4. 43 17.0 12. , 6

IT012 2.7 32. ,2 49. .9 5. 27 18.2 15. ,2

IT013 2.5 30, .8 48 , , 2 4. 1 18.1 18. , 5

Hudson 3.1 37. .6 46. , 4 3. 3 19.9 12. , 9

Bruker 3.6 35, , 3 47 , .5 3. 5 13.0 13. .5

VisionTek

3.3 30 , , 5 43. .9 4. 0 13.5 22, ,2

4Ω

VisionTek

4.1 33. , 2 46. .5 3. 7 13.4 16. , 2

15Ω

VisionTek

3.5 35 , , 2 47. ,5 3. 4 12.8 13. ,8

100Ω Table 3. Elemental compositions measured by XPS (in at . % ) From these studies, none of the studied commercial slides showed any LDI performance at all even at high fluence and high analyte concentrations. Nevertheless, while not wishing to be bound by any particular theory, the present inventors did not find that a single feature explains this behaviour in sputtered substrates, but rather the effect is rather due to a beneficial combination of surface parameters, e.g. found in the slide IT04. No other commercial or custom-made slide showed the same combination of surface roughness of 1.2 nm, 25-35 nm thickness of the coating, a low resistivity of 1.5-1.9 microOhm/cm ² , an average grain size of 60 nm and medium transmission of 66% at the laser wavelength, while still maintaining a superb transparency for optical microscopy applications. Both thin (30 nm) and thicker layers (up to 291 nm) , in a range of Rq values for roughness (0.6 to 3.2 nm) and resistivity (from 0.2 to 10 "10 ³ ohm-cm ² ) showed LDI activity in sputtered

substrates whereas the commercial slides with similar values (thickness of 25-400 nm, roughness of 0.4 to 6.7 nm and

resistivity of 0.1-1.8 -10 ^"3 ohm-cm ² ) did not. When comparing thin surfaces (ITO_l to ITO_8) , sample slide ITO showed an increase of 23 % in absorbance at 355 nm (MALDI-Tof used laser wavelength of 355 nm) . Furthermore, ITO_ll to ITO_13 slides showed a strongly decreased transmission at 355 nm related to the indium tin oxide thickness, between 116 and 291 nm, as determined by ellipsometry . Moreover, an important increase in LDI activity, as demonstrated by a higher S/N ratio for the analysis of our set of standard anaiytes, was found in particular for the sample ITO_13. Performance of slides ITO_l-13 in LDI-analysis is shown in Figure 1 and Table 2.

LDI performance of all substrates

The whole set of substrates was initially tested using 1- aminohexaethylenglycol analyte (0.1 mM, m/z = 282.19 for [M+H] ⁺ , 304.17 for [M+Na] ⁺ and 320.21 for [M+K] ⁺ ) . Obtained spectra are shown in the next Figure 1. With different S/N values, the peaks corresponding to the analyte in all the sputtered substrates were observed (m/z = 282.19 for [M+H] , 304.17 for [M+Na] ⁺ and 320.21 for [M+K] ⁺ ) without the addition of matrix to the sample. In contrast, using the same conditions for the sample preparation and measurement, the commercial substrates did not work as LDI surfaces.

The performance of the sample slides of the present invention was then compared with the commercially available ITO slides. The results of experiments to detect lactose, leukine, enkephaline, hexaethyleneglycol (HEG) and 1-amino-HEG are shown in Table 4.

S/N values for [M+Na] ⁺ species

Table 4. Comparison of S/N values obtained by LDI-MS of peaks

corresponding to 0.04, 0.4 and 4 mg/ml of lactose and 1 mg/ml

leukine-enkephaline . With a S/N ratio of 537 for the Leukine-Enkephaline (standard analyte for LDI evaluation) sodium adduct ITO_ 13 showed by far the highest LDI-MS performance for this analyte at 1 mg/ml. With an S/N of 100 to 300, surfaces ITO_2, ITO_4 and thicker

substrates ITO_ll and IT0_12 showed similar the highest

performance for the same analyte, but accompanied by loss of transparency for I O_12 and ITO_13.

Based on these preliminary findings, one of the preferred sample slides of the present invention was tested in a general screening experiment using surface ITC_4 and a library of small molecules. For that, and to assess the general scope of the method, a selection of picomole quantities of low mass analytes covering metabolites, carbohydrates, lipids and peptides were spotted onto the sample slide, dried and analysed by LDI-MS.

Figures 2 and 3 show representative spectra for the compounds creatinine (m/z= 136.08 [M+Na] ⁺ , 152.06 [M+K] ⁺ ) , taurine (m/z = 148.01 [M+Na] ⁺ , lami udine (m/z= 252.08, [M+Na] ⁺ , 268.06 [M+K] ⁺ ) , laminarin hexaose (m/z= 1013.55 [M+Na] ⁺ ) , di hexadecyl -sn-giycero- phosphoethanolamine (m/z= 662.85 [M-H] ^~ ) , n-octyl-p-thioglucoside (m/z = 331.26 [M+Na] ⁺ , 347.25 [M+K] ⁺ ) , hexaethylenglycol (305.30 [M+Na] ⁺ , 321.28 [M+K] ) and the Giy-Cys dipeptide (m/z = 201.06 for [M+Na] ⁺ ) . Other substrates analysed on this surface included the lipid fraction of human milk (palmitic acid m/z = 281.5 [M-H] ^" , oleic acid m/z = 281.5 [M-H] ^~ , among other lipids), a synthetic C5-amino modified N-glycan ^[15! (m/z= 1059.99 [M+Na] ⁺ , 1076.04

[M+K] ⁺ ) and immunoglobulin N-glycans enzymatically cleaved by PNGase treatment, (fucosyiated biantenary N-glycan, m/z = 1486.0; fucosyiated monogalactosylated biantenary N-glycan, m/z 1648.2 and fucosyiated bisgaiactosy.iat.ed biantenary N-glycan, m/z = 1810.4) .

Most analytes ionized as sodium and potassium adducts and clean spectra with little or no interference from background ions were usually obtained. For compounds which were less effectively ionized (e.g. taurine, laminarin hexasaccharide , N-glycans) indium tin oxide related and sodium and potassium background ions could be observed (enter background ions) . Peaks with m/z of 115.05 and 360.70, which are most likely corresponding to indium (In) and indium oxide (!¾()) species, respectively, are the highest intensity background peaks.

Background Putative

peak (m/z) specie

115.05 In

229.81 ΙΠ2

246.61 InOSn

360.70 ln ₃ 0

376.70 ln ₃ 0 ₂

381.70 In ₂ 0 ₂ Sn

491.61 ln ₄ 0 ₂

492.60 In ₃ Sn0 ₂

Table 5. Assignment of detected background peaks Sensitivity

A dilution series was performed (from 1000 to 0.5 pmol) for creatinine (m/z= 136.04 [M+Na] ⁺ ) to measure the sensitivity of the LDI-detection method. The method produced good quality spectra down to 0.5 pmol of deposited analyte with a signal to noise ratio of 13 as shown in Figure 4, comparable to the sensitivity of reported nanostructured gold ⁵¹⁶¹ and metal oxide LDI methods ^il7] .

Lactose quantification

Lactose quantification in so called lactose-poor and lactose-free milk products is an important analytical application in the dairy industry and inexpensive and high-throughput methods that substitute the time-consuming chromatographic methods and enzymatic cascade assays are desirable . ^il8"22! The utility of the matrix-free LDI-sample slides of the present invention was performed for the quantification of lactose in milk samples by isotopic dilution employing ¹³ C labelled lactose as an internal isotopic standard. Five milk samples with varying fat and lactose content were diluted up to 100-fold with water and acetonitrile and their lactose content quantified by relating the integrated peak area for lactose to the internal standard of known concentration. The dynamic range of the methods, limit of detection and limit of quantification of the methods were determined by adding known amounts of isotopically labelled standard to the milk matrix and plotting the peak intensity against the analyte concentration, (Figure 5 and 6) .

Lactose in diluted milk samples can be quantified down 10 mg/L, sufficient for the analysis of lactose-free labelled products (lOOmg/L) with unprecedented sample speed and no other sample preparation than dilution with water and acetonitrile. The sensitivity of our method is comparable with the reported limit of quantifications of 6 mg/L established LC-methods employing UV detection but avoiding the required extensive sample preparation and sampling times of around 25 minutes . ^!19!

MS Imaging in sputtered ITO slides

As a final application we employed the nanostructured ITO slides for imaging mass spectrometry of mouse brain tissue slices. For this application, the indium tin oxide surface was

hydrophobically functionalized sequentially with

aminopropyltriethoxysilane (APTES) and stearylsuccinimide ester ^f23 to improve the extraction of major often lipophilic components from the tissue sample. Although the surface modification with d i - and triethoxyalkylsilanes has been reported to potentially suppress LDI ^i24] , for our application in general, an increase in metabolite coverage and sensitivity was observed both in positive and negative mode after a hydrophobic coating of the ITO slides, (see Figure 7) .

Frozen mouse brain tissue slices of 4 and 10 μπι thickness (in frozen state) were thaw-mounted onto sample slides, dried and stored over night under vacuum and analysed either directly by

LDI-MS or after removal of the tissue by combined washing and sonication (stamping method) . ^t5] While direct LDI-MS of tissue slices produced appreciable analyte signal strength only for the 4μπι slices, the analysis of the hydrophobic surface impregnated with metabolites after tissue removal resulted in spectra with much improved signal intensity.

Analyte ionization during the analysis of tissue samples by LDI- MS occurs either after tissue ablation by repeated laser shots ⁱ²⁵¹ down to the underlying nanostructure or possibly through the presence of light absorbing metabolites that can act as an internal matrix. ^[26]

Repeated washing and sonication efficiently removed the tissue from the slide without touching metabolites extracted from the biological matrix by the hydrophobic surface.

Figure 8 and 9 show the spatial distribution of a selection of metabolites at 75 μπι resolution and negative mode on an axial section of a mouse brain. Ions were tentatively assigned based on f agmentation patterns by MS/MS, database queries and the literature. Even though the axial brain section employed in this proof of concept experiment showed little structural

heterogeneity by visual inspection under the microscope, LDI-MS imaging was able to trace a spatial differentiated distribution of certain metabolites to the rhinal fissure (m/z=420 , 834, 888, 890, 891, 131), the external capsule (m/z=834, 888, 890, 891, 131) or the temporal cortex (e.g. m/z= 387 (cholesterol), 554, 281 (oleic acid) . The metabolite ion 888 Da was assigned to a sulphated

galactocerebroside on the basis of LIFT-based fragmentation.

Figure 10 shows the MS/MS spectra and proposed metabolite structure for this ion. Metabolite distribution in mouse brain by mass spectrometry has been reported by Setou et al . employing T1O2 nanoparticles as a matrix with good metabolite coverage ¹¹⁰¹ , but this procedure adds a sample preparation step and requires specialized instrumentation for spray-coating the tissue sample with nanoparticles .

Conclusions

The experiments described herein demonstrate that a precise control of sputtering parameters can produce nanostructured indium tin oxide thin films with improved surface roughness and absorption profile for applications in small molecule LDI-MS. In comparison no LDT-activity was found for commercially available ITO slides on standard MALDI-Tof spectrometers. Samples obtained by the r.f, sputtering procedure of the present invention are conductive, nanostructured and transparent and show high

potential for the matrix-free imaging mass spectrometry of small molecules, conventional matrix-assisted imaging of larger lipids and proteins and traditional histology by optical microscopy. Nanostructuring of the indium tin oxide surface therefore broadens the applications of ITO slides in imaging mass

spectrometry to include matrix-free small molecule imaging, without compromising transparency or performance in matrix- assisted imaging.

Other applications include the general matrix-free analysis of small molecules, metabolites and drugs on standard MALDI-Tof mass spectrometers, at a fraction of the time and price of ESI-Tof spectrometers working in solution phase. ITO slides produced by r.f. sputtering show high homogeneity and reproducibility, important parameters for surface-based mass spectrometry

applications. Further functionalization of the surface with capture ligands or enzyme substrates will permit the development of array based high-throughput assays for applications in diagnostics and drug discovery. References :

All publications, patent and patent applications cited herein or filed with this application, including references filed as part of an Information Disclosure Statement are incorporated by reference in their entirety.

[I] A. Rompp, S. Guenther, Y. Schober, 0. Schulz, Z. Takats, .

Kummer, B. Spengler, Angew. Chem. Int. Ed. Engl. 2010, 49, 3834-8.

[2] K. Chughtai, R. M. A. Heeren, Chem. Rev. 2010, 110, 3237-77.

[3] P. Nernes, A. Vertes, Methods Mol. Biol. 2010, 656, 159-71.

[4] M. P. Gre ing, G. J. Patti, G. Siuzdak, Anal. Chem. 2011, 83, 2-7.

[5] V. Vi ova, P. Novak, M. Strohalm, J. Pol, V. Havlicek, M.

Volny, Anal. Chem. 2010, 82, 4994-7.

[6] Q. Liu, Z. Guo, L. He, Anal. Chem. 2007, 79, 3535-41.

[7] J. G. Forsythe, J. A. Broussard, J. L. Lawrie, M. Kliman, Y.

Jiao, S. M. Weiss, D. J. Webb, J. A. McLean, Anal. Chem.

2012, 84, 10665-70.

[8] N. Goto-Inoue, T. Hayasaka, N. Zaima, Y. Kashiwagi, M.

Yamamoto, M. Nakamoto, M. Setou, J. Am. Soc. Mass Spectrom.

2010, 21, 1940-3.

[9] T. Hayasaka, N. Goto-Inoue, N . Zaima, K. Shrivas , Y.

Kashiwagi, M. Yamamoto, M. Nakamoto, M. Setou, J. Am. Soc. Mass Spectrom. 2010, 21, 1446-54.

[10] K. Shrivas, T. Hayasaka, Y. Sugiura, M. Setou, Analytical Chemistry 2011.

[II] S. Cha, E. S. Yeung, Anal. Chem. 2007, 79, 2373-85.

[12] R. J. A. Goodwin, A. R. Pitt, D. Harrison, S. K. Weidt, P.

R. R. Langr i dge-Smith , M. P. Barrett, C. Logan Mackay, Rapid Commun . Mass Spectrom. 2011, 25, 969-72.

[13] A. Beloqui, J. Calvo, S. Serna, S. Yan, I. B. H. Wilson, M.

Martin-Lomas , N. C. Reichardt, Angew. Chem. Int. Ed. Engl.

2013, 52, 7477-81.

[14] R. Joshi, V. Singh, J. McClure, Thin Solid Films 1995, 257, 32-35.

[15] S. Serna, J. Etxebarria, N. Ruiz, M. Martin-Lomas, N.-C.

Reichardt, Chemistry 2010, 16, 13163-75. [16] R. Nayak, D. R. Knapp, Anal. Chem. 2010, 82, 7772-8.

[17] C. R. McAlpin, K. J. Voor ees, A. R. Corpuz, R. M. Richards,

Anal. Chem. 2012, 84, 7677-83.

[18] D. Biggs, L. Szijarto, Journal of Dairy Science 1963, 46,

1196-1200.

[19] S. Erich, T. Anzmann, L. Fischer, Food Chem 2012, 135, 2393- 6.

[20] C. Brons, C. Olieman, J. Chromatogr . 1983, 259, 79-86.

[21] S. Fornera, K. Yazawa, P. alde, Analytical and

bioanalytical chemistry 2011, 401, 2307-2310.

[22] T. R. Cataldi, M. Angelot i, G. Bianco, Analytica chimica acta 2003, 485, 43-49.

[23] A. Sanchez-Ruiz, S. Serna, N. Ruiz, M. Martin-Lomas , N . C.

Rei chardt , Angewandte Chenie International Edition 2011, 50,

1801-1804.

[24] S. A. Trauger, E. P. Go, Z. Shen, J. V. Apon, B. J. Compton, E. S. P. Bouvier, M. G. Finn, G. Siuzdak, Anal. Chem. 2004, 76, 4484-9.

[25] 0. Yanes, H.-K. Woo, T. R. Northen, S. R. Oppenheimer, L.

Shriver , J. Apon, M. N . Estrada, M. J. Potchoiba, R.

Steenwyk, M. Manchester, et al . , Anal. Chem. 2009, 81, 2969- 75.

[26] D. HoIscher, R. Shroff, K. Knop, M. Gottschaldt, A.

Crecelius, B. Schneider, D. G. Heckel, U. S. Schubert, A.

SvatoXvs, The Plant Journal 2009, 60, 907-918.