Imaging of a Tribolayer Formed from Ionic Liquids by Laser Desorption

Article pubs.acs.org/ac

Imaging of a Tribolayer Formed from Ionic Liquids by Laser Desorption/Ionization-Reflectron Time-of-Flight Mass Spectrometry Christoph Gabler,*,†,‡ Ernst Pittenauer,† Nicole Dörr,‡ and Günter Allmaier*,† †

Institute of Chemical Technologies and Analytics, Vienna University of Technology, Vienna, Austria Austrian Centre of Competence for Tribology − AC2T research GmbH, Wiener Neustadt, Austria

‡

ABSTRACT: For the first time, imaging using laser desorption/ionization (LDI) reflectron time-of-flight (RTOF) mass spectrometry (MS) was demonstrated to be a powerful tool for an offline monitoring of tribometrical experiments directly from disc specimen applying selected ammonium-, phosphonium-, and sulfonium-based ionic liquids (IL) with bis(trifluoromethylsulfonyl)imide as counterion for lubrication. The direct measurement of IL tribolayers by LDIMS allowed the visualization of the lubricants in the form of the distribution of their intact cations and the anion in and outside the wear scar after the tribometrical experiment with a low degree of in-source generated fragmentation. Besides, also, an oxidation product formed during a tribometrical experiment was detected and located exclusively in the wear track. Comparative data of identical wear tracks were obtained by X-ray photoelectron spectroscopy (XPS) imaging not only enabling the determination of elemental distributions of the IL across the area imaged but also corroborating the mass spectrometry imaging (MSI) data, thus generating multimodal images. Merging data from MSI and XPS imaging exhibited that areas, where iron−fluorine bonds were detected in the wear track, are corresponding to data from LDI-MS imaging showing absence of IL cations and anions.

R

development of more performant lubricants optimized for their respective applications. A limiting factor of surface analysis with XPS and TOF-SIMS is that they provide usually no comprehensive structural information and no molecular weight data about the chemical compounds on the surface,14 which would be a highly valuable contribution especially for applications employing novel lubricants like IL with their multitude of different upcoming structures.2,14−17 MALDI-MS, which is extensively used in organic, medical, and biological research,18−20 has the potential to provide this missing information. This technique is capable of two-dimensional imaging of solid surfaces providing a distribution pattern of intact molecular ions, known as MS imaging (MSI).21−23 Furthermore, recent papers have reported about the characterization of neat and degraded IL by LDI-TOF-MS analysis too.24,25 The amenability of IL to direct desorption and ionization without the application of MALDI-MS matrix molecules, because they are used also as MALDI-MS matrixes themselves, opens new opportunities in MSI due to a straightforward sample preparation and an easier interpretation of the data due to the absence of interfering background ions deriving from the MALDI matrix. Since the aim of this work was to investigate the distribution of the IL anionic and cationic moiety and possible degradation products on a disc specimen by direct measurement after the

eciprocating tribometers are essential tools for evaluating lubricants of organic, inorganic, and biological origin in model experiments for their capability to perform under varying parameters such as temperature, applied forces, duration, or friction partners. Chemical reactions occurring during these sliding experiments between two friction partners in contact with the lubricant are designated as tribochemical reactions.1 The products of these reactions are often the main components of the layers formed in the tribologically stressed area, commonly called tribolayers or tribofilms. Ionic liquids (IL) are receiving more and more attention as novel lubricants due to their unique combination of properties relevant to tribology, e.g., low volatility and nonflammability together with good lubricating properties.2,3 Additionally, IL possess also an important role as UV matrix-assisted laser desorption ionization (MALDI)-mass spectrometry (MS) matrixes, e.g., for peptides, proteins, or synthetic polymers.4−7 This situation can be seen as quite favorable for the present study. The investigation and characterization of tribologically stressed surfaces with X-ray photoelectron spectroscopy (XPS), X-ray absorption near edge structure spectroscopy (XANES), or time-of-flight secondary ion mass spectrometry (TOF-SIMS) can be seen as state of the art in the field of tribochemistry.8−13 The knowledge gained using these powerful surface analytical techniques to elucidate tribochemical reaction mechanisms is nowadays indispensable in order to understand which conditions lubricants and/or additives have to face in the tribological contact, e.g., with metal, ceramic, or polymer surfaces. This kind of surface analytical research is part of the © 2012 American Chemical Society

Received: August 30, 2012 Accepted: November 28, 2012 Published: November 28, 2012 10708

dx.doi.org/10.1021/ac302503a | Anal. Chem. 2012, 84, 10708−10714

Analytical Chemistry

Article

used for subsequent MS analysis, and the fourth one was exclusively for XPS in order to avoid an effect of the LDI laser beam on the chemical species forming the tribolayer. Immediately after the tribo-experiment, the cleaning procedure started, which consisted of immersion in toluene (analytical grade) in an ultrasonic bath for 15 min at room temperature, followed by 2-propanol (LC-MS Chromasolv grade) and petroleum ether (analytical grade) for the same duration. All solvents were purchased from Sigma Aldrich (St. Louis, MO, USA). After cleaning, the discs were stored in a desiccator until the beginning of the analysis. The selected IL (IL 1 to IL 5) were purchased from IoLiTec (Ionic Liquid Technologies, Heilbronn, Germany). The chosen IL structures, summarized in Table 1, cover a number of structurally different cationic moieties which are representative for compounds currently investigated in tribology, with the well-known anionic moiety bis(trifluoromethylsulfonyl)imide as counterion.2,3 Mass Spectrometry. Since one of the aims of this work was to measure unreacted IL and tribochemical reaction products with LDI-RTOF-MS directly from the disc specimen after the ball-on-disc experiment, the crucial part of this work was the necessity of adapting a common microtiter plate (MTP) format MALDI stainless steel target (96 rings (3.4 mm I.D.), 10 mm thick plate, plate type number DE2112TA, Shimadzu Kratos Analytical, Manchester, UK) as a sample holder for the disc specimens (for two shapes of standardized tribological discs: 24 mm diameter with a 8 mm thickness and 11 mm diameter with a thickness of 3 mm). This way, the prepared disc specimens could be directly supplied to the LDIRTOF-MS without the need of further alignment in the MS. Figure 2 illustrates how the disc specimens were mounted on the modified MALDI-MS target with sufficient accuracy to ensure the planarity of target and disc surface within single-digit micrometer range as the disc is the focus plane of the laser. This guarantees that the laser beam is in focus on the selected area to scan across the disc throughout the whole MSI experiment. As small deviations from the standard height between ±10 μm of the MALDI-MS target were still possible, the instrument was calibrated directly from the surface of the disc specimen in order to avoid wrong m/z assignments by a flight-time shift compared to the rest of the target plate. In the positive-ion reflectron mode, a three point calibration with Na+, K+, and the known cationic moiety of the investigated IL was performed. In the negative-ion reflectron mode, calibration was achieved through a two point calibration with the anionic moiety of the IL and the cyanide anion, which is always present in negative ion LDI-mass spectra of organic compounds containing nitrogen. The MSI experiments were performed on a Shimadzu Axima TOF2 MALDI/LDI tandem time-of-flight (linear TOF/RTOF) instrument (Shimadzu Kratos Analytical, Manchester, UK) with a curved field reflectron as MS2. The instrument was equipped with a nitrogen laser having an operating wavelength of λ = 337 nm and a repetition rate of 20 Hz. Twenty kilovolts was selected as acceleration voltage. For all the MSI experiments, an area of 4 mm2 (2 × 2 mm2) was defined, which was scanned with a step size of 30 μm, resulting in 4489 measurement spots, i.e., mass spectra, at an ion source pressure of roughly 5 × 10−6 mbar. The approximate laser beam diameter of 60 μm and the raster step size of 30 μm resulted in overlapping measurement spots. The laser power for these measurements was in the region of 55 to 65 (0−180 relative units). In order to achieve sufficiently good ion statistics, each acquired profile represents

tribometrical experiment and removal of surplus IL, without further sample preparation or manipulation, it was just necessary to transfer the discs directly into the MALDIreflectron time-of-flight (RTOF)-MS instrument. This was achieved by adapting a commercial MALDI-MS stainless steel target as a holder for disc specimens. A study26 presented recently indicated that direct LDI-TOF-MS measurements of a lubricant layer, which consisted of perfluoro-polyether (PFPE), applied on a magnetic recording disc is feasible. A significant difference between the present study and the above-mentioned one is that the PFPE layer could only be reasonably analyzed after coating the surface with a salt solution supporting the ion formation whereas here plain LDI-MS imaging has been applied. In order to supplement the data obtained from LDIRTOF-MS, XPS imaging data were provided from the disc specimens, too. To the best of our knowledge, this is the first time that LDI-RTOF-MS imaging data from tribolayers in general and IL tribolayers in particular were directly acquired from disc specimens and correlated with light microscopic images of the same disc.

■

MATERIALS AND METHODS Samples for Tribometrical Experiments. The samples for the imaging surface analytical experiments with LDI-RTOFMS and XPS were obtained through reciprocating ball-on-disc tribo-experiments, performed on a Schwing-Reib-Verschleiss (SRV) tribometer (Optimol Instruments Prüftechnik, Munich, Germany). Both friction partners (ball: Ø 10 mm, roughness Ra 0.01 μm, hardness 60−66 mm HRC; disc: Ø 24 mm, 7.9 mm thick, roughness 0.035 μ < Ra 0.05 μm, hardness 62 mm ±1 mm HRC) were 100Cr6 steel (purchased from Optimol Instruments Prüftechnik, Munich, Germany) and prepared according to the standard DIN 51834-2 providing a tribometrical method for the testing of lubricants.27 The assembly of this experiment is shown in Figure 1. All experiments were

Figure 1. (A) The assembly of the reciprocating tribometrical experiment. The upper specimen (ball) is oscillating on the lower specimen (disc) with a defined force (e.g., 100 N) applied. The IL (white area) serves as lubricant between these two specimens (amount approximately 20 μL). The original assembly is shown in (B).

performed at room temperature under ambient laboratory conditions with a load of 100 N for 30 min. The stroke (maximum deflection of the ball) was 1 mm, and the oscillating frequency was defined to be 50 Hz, resulting in an initial wear scar length of 1 mm. The average width of the wear scars was 400 μm. The neat IL lubricant was applied as droplet (approximately 20 μL) on the steel disc, to be spread by the movement of the oscillating ball. On each disc, four sliding experiments with one IL were performed. Three of those were 10709

dx.doi.org/10.1021/ac302503a | Anal. Chem. 2012, 84, 10708−10714

Analytical Chemistry

Article

Table 1. Overview of the Selected ILs and the Corresponding Monoisotopic Masses of the Ionic Moieties

the sum of five consecutive laser pulses on one spot, resulting in an average total experiment time of 30 min. Reference mass spectra of each selected IL were recorded prior to MSI in positive and negative-ion reflectron mode in order to investigate potential in-source fragmentation by the LDI process. For this purpose, 1 μL of a methanolic stock solution of each IL with a concentration range from 0.3 to 0.5% (v/v) was applied onto a MALDI stainless steel MTP target. The collected imaging mass spectra as well as reference mass spectra were processed with the Shimadzu MALDI-MS software v 2.8.5 (Shimadzu Kratos Analytical, Manchester, UK) package first (unsmoothed, no background subtraction). The MSI data were then exported to BioMap 3X 3.8.0.3 (Novartis Institute for BioMedical Research, Basel, Switzerland) for further processing. To enhance the contrast of the acquired images and to superimpose light microscopic pictures, taken with an Olympus SZX 16 (Olympus, Tokyo, Japan), of the disc surface with the MS-based images, the free GNU image manipulation program GIMP 2.6.12 was used. X-ray Photoelectron Spectroscopy (XPS). XPS analysis was performed with a Thermo Fisher Scientific Theta Probe

(East Grinstead, UK) with a monochromatic Al Kα X-ray source (hυ = 1486.6 eV). The base pressure during the measurements was consistently at 3 × 10−9 mbar. The reference binding energies of the IL were measured by depositing the sample fluid as thin film onto a polished steel surface, by distributing approximately 1 μL of the fluid with a pipet tip. Analysis spot size and pass energy for the reference detail spectra were 400 μm and 50 eV, respectively. The wear tracks were scanned with a step size and spot size of 100 μm. All acquired spectra were referenced to the C1s peak at 284.6 eV. The resulting analysis data were processed with the Avantage Data System software v 4.75 (Thermo Fisher Scientific, East Grinstead, UK) using Gaussian/Lorentzian peak fitting.

■

RESULTS AND DISCUSSION

LDI-RTOF-MS Analysis of Neat IL as References. Acquiring reference mass spectra of the five selected IL was necessary to elucidate their in-source fragmentation behavior during the LDI process and to obtain the optimal laser power 10710

dx.doi.org/10.1021/ac302503a | Anal. Chem. 2012, 84, 10708−10714

Analytical Chemistry

Article

Figure 2. Adapting the 10 mm thick MALDI-MS stainless steel target as sample holder for two types of disc specimens. (A) Target before modification from the top; (B) target after modification from the top, with 4 wide holes for large-sized disc specimens (fixation of discs by headless screws from the side, white arrows) and six small holes for small-sized discs (these discs are just dropped in and are without any fixation); (C) counter plate which served as gauge before mounting the discs; (D) MALDI-target lying with the front side on the counter plate, with one already mounted large-sized disc (right upper corner).

per area for desorption/ionization of the IL. This knowledge was fundamental for the further investigation of IL tribolayers. The anionic structure for all five investigated IL was bis(trifluoromethylsulfonyl)imide [(CF3SO2)2N]−. The obtained negative-ion mass spectrum (Figure 3) shows an intense Figure 4. Positive-ion LDI reflectron mass spectra of the reference IL (IL 1 to IL 5) acquired from a stainless steel target surface with a relative laser power of 71 (0−180 relative units): (A) 1-hexyl-3methyl-imidazolium cation (IL 1), (B) triethyl-sulfonium cation (IL 2), (C) 1-butyl-1-methyl-pyrrolidinium cation (IL 3), (D) butyltrimethyl-ammonium cation (IL 4), and (E) tributyl-methylphosphonium (IL 5).

m/z 58 could be observed. The cleavage of the alkyl moieties is base for this fragmentation path;24 hence, m/z 60 could be related to the loss of (C4H8). During formation of the m/z 58 fragment ion, the butyl side chain is leaving the IL cation. This causes the formation of a double bond between the nitrogen atom and one of the remaining methyl groups. Another alkyl chain loss was observed in the case of the 1-butyl-1-methylpyrrolidinium (IL 3) cation, which exhibits a minor abundant fragment ion at m/z 86. This mass corresponds to the loss of the butyl side chain. Another alkyl side chain loss was observed for the triethyl-sulfonium (IL 2) cation. This results in the formation of a fragment ion at m/z 91 which could be related to the loss of one ethyl side chain. No significant fragment ion caused by an alkyl side chain elimination was observable for the tributyl-methyl-phosphonium (IL 5) cation, even though this molecule exhibited three butyl side chains and with 71 units (0−180 relative units) the relative laser power was at the same level for all reference measurements. MSI of the Discs after Tribometrical Experiments. Through the modification of the thick MALDI-MS stainless steel target, the measurements of the disc surfaces were straightforward, particularly due to the positional fixation in the holder. A scanned area of 4 mm2 was defined around the worn area (see Figure 5 center of light microscopic image) to obtain

Figure 3. Negative-ion LDI reflectron mass spectrum of IL 4 showing [(CF3SO2)2N]− as base peak. This mass spectrum is representative for the negative-ion mode for all five selected IL.

molecular ion peak at m/z 279.9 (calculated m/z 279.9) as base peak accompanied by a low abundant in-source generated fragment ion at m/z 146.9 (calculated m/z 147.0). The latter ion represents the fragment ion formed through the neutral loss of one (CF3SO2)-side chain, which was already reported previously.25 Figure 4 shows the positive-ion mass spectra of the IL 1 to IL 5 (for calculated m/z values, see Table 1) corresponding to their investigated cationic moieties, all of which exhibit abundant molecular ion peaks as base peaks. In the case of the 1-hexyl-3-methyl-imidazolium (IL 1) cation, an intense fragment ion is present at m/z 83. This fragment ion could be related to the loss of the hexyl side chain. A previous report about LDI-TOF-MS of IL also describes the formation of the m/z 83 fragment ion for 1-butyl-methyl-imidazolium bis(trifluoromethylsulfonyl)imide, which suggests that imidazolium cations with longer side chains fragment under LDI by a cleavage between the imidazolium core and the aliphatic side chain.25 Fragmentation also occurred in the case of the butyltrimethyl-ammonium (IL 4) cation where ions at m/z 60 and 10711

dx.doi.org/10.1021/ac302503a | Anal. Chem. 2012, 84, 10708−10714

Analytical Chemistry

Article

Figure 6. Positive-ion LDI reflectron mass spectra from butyltrimethyl-ammonium (IL 4) where (A) is showing the reference mass spectrum (obtained at elevated laser power per area values compared to (B)) with the characteristic fragment ions at m/z 58.1 and m/z 60.1 and (B) represents the mass spectrum from the disc surface obtained after tribometrical experiment without these fragment ions.

detectable for the desorption/ionization from the disc surface. This decreased fragmentation could be found for all investigated IL under the same conditions. One of the major aims of this study was to visualize the distribution of the IL moieties in and around the worn area after a tribometrical experiment. The first results are presented in Figure 5, where the distribution of the cationic and anionic moieties of IL 1 is shown. For a better understanding of the IL localization with respect to the worn area, a light microscopic picture of this area is shown in Figure 5 A and the same image is superimposed by a mass spectrometric image from this region (Figure 5B,C). The distribution of the intensity maxima (red color) of the anionic and cationic moieties of IL 1 implies that the majority of the IL is bound to the surface of the disc outside the worn area. Furthermore, there is a boundary line along the long edge of the wear scar where no IL was detected, whereas at the reversal points of the steel ball, the deposition of the IL occurred directly next to the wear scar. The same was observed for the intensity distributions of IL 2 and IL 3. The areas, especially in the wear scar, with no mass spectrometric signals related to the IL ions, can be corroborated by previous studies of IL tribolayers with XPS reporting the formation of iron fluoride in the tribologically stressed area under lubrication with fluorine-containing IL.28−30 A specific model for the behavior of IL based on the bis(trifluoromethylsulfonyl)imide anion was proposed by Minami et al.,28 where they describe the formation of a boundary film composed of metal fluoride under similar tribological conditions as selected for the present study. Further, they suggested that there is no cationic effect to be expected for the formation of the boundary film which means that it is dominated by the anionic species for the selected parameters. Figure 7 illustrates the distribution of the inorganic fluoride in the case of IL 1. It clearly shows that the fluorides are predominantly located in the wear track, especially at the reversal points of the ball, but also outside, in the region of the boundary line, they were detected to a certain extent. This implies that in areas where such inorganic compounds were located, the desorption/ionization of the IL moieties was not possible anymore due to their degradation. Another reason could be that the tribochemical reaction products hinder or weaken the physisorption/chemisorption of the neat IL due to a stronger binding to the metal surface and the IL could be washed away during the cleaning procedure.

Figure 5. Light microscopic picture (A) (5× magnification) of the IL 1 wear scar superimposed by the corresponding mass spectrometric images from the cationic moiety with m/z 167.2 (positive ion mode) (B) and the anionic moiety with m/z 279.9 (negative ion mode) (C). The length of the wear scar is 1 mm, and the width is 0.4 mm. Blue = low intensity; red = high intensity; transparent/black = no ions detected.

representative information also outside the wear scar, which was essential in order to comprehend the distribution of the IL moieties after the tribometrical experiment. The first measurements from the disc surface revealed that acquiring high quality mass spectra in positive and negative-ion reflectron mode with high resolution (R = 2600−3500 at fwhm) required about 10 to 20% less laser power per area (55− 65 arbitrary unit between 0 and 180) than was needed for the reference LDI-MS measurements. This could be explained by the formation of a bulky layer (multilayer) of reference IL due to the application of a relatively highly concentrated methanolic stock solution on the unmodified MALDI-MS target. Due to this thick layer, more laser power per area was required for the desorption/ionization process than in the case of the disc specimens where comparatively thin IL layers were present. The decrease in laser power resulted further in a reduced insource fragmentation, as shown for the butyl-trimethylammonium cation (IL 4) in Figure 6. The comparison of the data acquired from the reference IL and the disc specimen obtained after the tribometrical experiment clearly shows that the characteristic fragment ions at m/z 58 and m/z 60 were not 10712

dx.doi.org/10.1021/ac302503a | Anal. Chem. 2012, 84, 10708−10714

Analytical Chemistry

Article

Figure 7. XPS image superimposed on light microscopic picture of the wear scar (5× magnification) showing the intensity distribution of F 1s (684.2 eV) after the tribometrical experiment with IL 1. The unit for the color scale at the top of the image is atomic percent (at %, proportional to detected elements on the surface). The wear track is marked by the black dashed line.

Figure 8. Comparison of the distribution of m/z 217 corresponding to IL 5 cation (A) with the distribution of m/z 231 (B), a tribochemical reaction product. (C) is showing the corresponding mass spectrum obtained from all acquired mass spectra across the whole disc surface. The light microscopic picture was taken at a 5× magnification and superimposed by the mass spectrometric images. The length of the wear scar is 1 mm, and the width is 0.4 mm. Blue = low intensity; red = high intensity; transparent/black = no ions detected.

The intensity distribution of the IL 4 ions (not shown) differs from the above-mentioned as high intensities are observed also in the worn area. A possible explanation for this behavior could be the low content of inorganic fluorine of about 10%, compared to 40% of the total fluorine yield, in the case of the other IL, as the XPS analysis of the wear scars revealed. This would mean that less degradation occurred in the tribological contact area for IL 4. Because only the cationic moieties were changed between the experiments, we propose that this behavior can be related to a cationic effect, but for a closer investigation of the mechanism behind this effect, further tribometrical experiments with varying parameters such as load and temperature are required. In the case of the tributyl-methyl-phosphonium cation of IL 5, the boundary line was not observed. Small areas of higher intensities for the cation were found inside the worn area, as shown in Figure 8. Utmost interesting was that especially in these regions a mass of m/z 231, which was not observable in the reference spectrum, was also located. Figure 8 shows the intensity distribution of m/z 231 and the mass spectra of IL 5 from the disc specimen. The mass difference between the cationic mass m/z 217 and m/z 231 is 14 Da which most likely correspond to a single oxidation of an alkyl side chain through formation of a carbonyl group. This finding proves that it is also possible to detect and visualize tribochemical reaction products by LDI-MS imaging, which could be developed as an important tool in the field of tribochemistry. XPS of the Discs after the Tribometrical Experiments. Comparative data from the cleaned disc surface were acquired by XPS measurements to support the hypothesis that, in the disc regions where no IL-related ions were detected on the surface by LDI-RTOF-MS, the IL were degraded by simultaneous formation of inorganic fluorine compounds. Therefore, reference binding energies of all selected IL were acquired to relate the binding energies detected on the disc surface to the respective IL. The reference binding energy measured for the F 1s photoelectron peak of the anionic trifluoromethyl group was 688.3 (±0.2) eV. After the tribometrical experiments, a peak at this reference binding energy was found on the disc specimens inside the wear scar as well as outside the worn area for all investigated IL. In addition to the binding energy of the trifluoromethyl group, a binding energy of 684.2 (±0.2) eV was

determined for the F 1s photoelectron line inside the wear scars. This binding energy is attributed to the formation of inorganic fluorine compounds, e.g., iron fluoride, corroborating the results from previous investigations about fluorinecontaining IL tribolayers.28−30 The average concentration of inorganic fluorine, in contrast to the total amount [defined as 100%] of fluorine within the wear scar for IL 1, IL 2, IL 3, and IL 5, was approximately 40%. The wear scar of IL 4 contained only 10% inorganic fluorine compounds, compared to 90% of organic fluorine compounds.

■

CONCLUSIONS The present study demonstrates the feasibility of acquiring mass spectrometry imaging data from disc specimens as used in tribometrical experiments by direct LDI-RTOF-MS measurement for the first time. The investigated IL tribolayers provided the opportunity to perform just LDI-MS, i.e., no sample preparation by deposition of MALDI-MS matrix. It was possible to show that this technique has the capability to visualize the distribution of IL anionic and cationic moieties after the tribometrical experiment directly from the disc and therefore provides information about the molecular mass and to a limited extent structural information based on in-source formed fragment ions of organic compounds on the disc surface. Tandem MS/MS experiments for detailed structure elucidation can be now considered to be feasible. The XPS data of the worn area provided complementary information about the inorganic compounds (both elements and binding energies) on the surface which supported the proposed hypothesis for the distributions of the investigated IL. A further feature of this method was the possibility to elucidate the localization of a tribochemical reaction product for IL 5, which is a remarkable new contribution to the field of tribochemistry, since the generally used methods such as XPS and TOF-SIMS provide no comprehensive information about molecular mass and structure of organic compounds generated 10713

dx.doi.org/10.1021/ac302503a | Anal. Chem. 2012, 84, 10708−10714

Analytical Chemistry

Article

(24) Pisarova, L.; Gabler, C.; Dörr, N.; Pittenauer, E.; Allmaier, G. Tribol. Int. 2012, 46, 73−83. (25) Zabet-Moghaddam, M.; Krüger, R.; Heinzle, E.; Tholey, A. J. Mass Spectrom. 2004, 39, 1494−1505. (26) Kudo, T.; Macht, M.; Kuroda, M. Anal. Chem. 2011, 83, 5563− 5569. (27) DIN-51834-2; Testing of lubricants − Tribological test in the translatory oscillation apparatus − Part 2: Determination of friction and wear data for lubricating oils; DIN Deutsches Institut für Normung e.V.: Berlin, 2004. (28) Minami, I.; Inada, T.; Sasaki, R.; Nanao, H. Tribol. Lett. 2010, 40, 225−235. (29) Kamimura, H.; Kubo, T.; Minami, I.; Mori, S. Tribol. Lett. 2007, 40, 620−625. (30) Lu, R.; Mori, S.; Kabayashi, K.; Nanao, H. Appl. Surf. Sci. 2009, 255, 8965−8971.

on the investigated surface obtained from the tribometrical experiment.

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected] (C.G.); [email protected] (G.A.). Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was funded from the “Austrian COMET-Program” in the frame of K2 XTribology and has been carried out within the “Austrian Excellence Centre of Tribology” − AC2T research GmbH. The surface analytical work with XPS was supported with EFRE funding and with support of the country of Lower Austria within the project “Onlab”. All LDI-MS imaging experiments were performed on an instrument made available by the Vienna University of Technology (to G.A.). The authors are grateful to H. Störi (Institute of Applied Physics, Vienna University of Technology, Vienna, Austria) and his co-workers in the machine shop for fruitful discussions and adaptation of the MALDI-MS target.

■

REFERENCES

(1) Pawlak, Z. Tribochemistry of Lubricating Oils; Elisevier B.V: Amsterdam, NL, 2003. (2) Minami, I. Molecules 2009, 14, 2286−2305. (3) Palacio, M.; Bhushan, B. Tribol. Lett. 2010, 40, 247−268. (4) Armstrong, D. W.; Zhang, L. K.; He, L.; Gross, M. L. Anal. Chem. 2001, 73, 3679−3686. (5) Crank, J. A.; Armstrong, D. W. J. Am. Soc. Mass Spectrom. 2009, 20, 1970. (6) Li, Y. L.; Gross, M. L. J. Am. Soc. Mass Spectrom. 2004, 15, 1833. (7) Li, Y. L.; Gross, M. L. J. Am. Soc. Mass Spectrom. 2005, 16, 679. (8) Kubo, T.; Nanao, H.; Mori, S.; Enomoto, Y.; Nie, H.; Nomura, H. Wear 2010, 268, 1225−1229. (9) Reichelt, M.; Gunst, U.; Wolf, T.; Mayer, J.; Arlinghaus, H. F.; Gold, P. W. Wear 2010, 268, 1205−1213. (10) Mistry, K. K.; Morina, A.; Neville, A. Wear 2011, 271, 1739− 1744. (11) Philippon, D.; Barros-Bouchet, M. I.; Le Mogne, T.; Lerasle, O.; Bouffet, A.; Martin, J. M. Tribol. Int. 2011, 44, 684−691. (12) Kim, B.; Mourhatch, R.; Aswath, P. B. Wear 2010, 268, 579− 591. (13) Mourhatch, R.; Aswath, P. B. Tribol. Int. 2011, 44, 187−200. (14) Friedbacher, G.; Bubert, H. Surface and Thin Film Analysis; Wiley-VCH: Weinheim, Germany, 2011. (15) Zhou, F.; Liang, Y.; Liu, W. Chem. Soc. Rev. 2009, 38, 2590− 2599. (16) Bermûdez, M. D.; Jimênez, A. E.; Sanes, J.; Carriôn, F. J. Molecules 2009, 14, 2888−2908. (17) Weingärtner, H. Angew. Chem., Int. Ed. 2008, 47, 654−670. (18) Hillenkamp, F.; Karas, M.; Beavis, R. C.; Chait, B. T. Anal. Chem. 1991, 63, 1193A−1203A. (19) Macht, M. Bioanalysis 2009, 1, 1131−1148. (20) Resemann, A.; Wunderlich, D.; Rothbauer, U.; Warscheid, B.; Leonhardt, H.; Fuchser, J.; Kuhlmann, K.; Suckau, D. Anal. Chem. 2010, 82, 3283−3292. (21) Chaurand, P.; Schwartz, S. A.; Reyzer, M. L.; Caprioli, R. M. Toxicol. Pathol. 2005, 33, 92−101. (22) Rubakhin, S.; Sweedler, V. Mass Spectrometry Imaging: Principles and Protocols; Springer: London, UK, 2010. (23) Setou, M. Imaging Mass Spectrometry: Protocols for Mass Microscopy; Springer: Tokyo, Japan, 2010. 10714

dx.doi.org/10.1021/ac302503a | Anal. Chem. 2012, 84, 10708−10714