SIMS2015 Session FN-TuP: Fundamentals Poster Session

Since time-of-flight secondary ion mass spectrometry (TOF-SIMS) is one of the most powerful surface analysis techniques and has been widely used in various fields, the TOF-SIMS working group (WG) was established in Surface Analysis Society of Japan (SASJ), which has contributed in standardization in Japan, on June 2007. One of the most important topics for the SASJ TOF-SIMS WG is identification of surface chemical species, and therefore accurate mass calibration is required. The SASJ TOF-SIMS WG has been working on the investigation of mass calibration and identification and organized TOF-SIMS inter-laboratory studies regarding the mass scale calibration of TOF-SIMS. On the other hand, some pioneering works on the TOF-SIMS inter-laboratory studies such as intensity repeatability, mass scale accuracy and relative quantification have been reported by the National Physical Laboratory (NPL) [1-5].

The SASJ TOF-SIMS WG considered those reports and ISO 13084 and also carried out inter-laboratory tests to find out a degree of variation of the mass accuracy, and worked a research of the variation due to conditions for mass calibration based on the results of the measurement. As a result, variation of relative mass accuracy was reduced from about 200 ppm to about 50 ppm [6] by sharing the knowledge obtained from five tests among the SASJ members. Especially, in the fourth inter-laboratory test, the variation of relative mass accuracy was studied using peak positions at a relatively high-mass region, which were not used in the previous tests and which are also described in ISO s tandard [7]. Moreover, a practical method for improving mass accuracy was also investigated. Then, we are referring to the idea of the ISO standard, and tried a new method of using the molecular ion peak for the mass calibration by adding a quaternary ammonium salt to the sample,. In this paper we report on the course of these studies that have been done by SASJ TOF-SIMS WG [8,9].

Molecular dynamics computer simulations are used to perform first atomistic simulations of cluster bombardment of a polymer material composed from C, H and O atoms. The ReaxFF potential for C, H and O has been implemented in a parallel MD code [1, 2]. Material ejection, energy transfer pathways and chemical reactions stimulated by keV C₆₀ and Ar₈₇₂ projectiles bombarding polylactic acid (PLA) at various kinetic energies and impact angles have been investigated. The details of the computational model and a summary of results will be discussed.

A wide angular spread of sputtered atoms may have a negative impact on the mass resolution in Secondary Ion Mass Spectrometry. It has been observed that impacts of large Ar cluster projectiles on flat metal surfaces along the surface normal lead to a strong off-normal material ejection [1]. This effect is known as lateral sputtering. It is attributed predominantly to blocking of substrate particles ejection along the surface normal by a cloud of Ar atoms hovering for a prolonged time above the point of projectile impact. Chernysh et al. have recently studied angular distributions of atoms ejected by 10 keV Ar₁₀₀₀ at normal incidence from surfaces of several metal polycrystals [2]. For Mo surface they observed that most atoms are emitted in directions close to the surface normal. This unusual behavior was hypothesized to be due to a spring like ejection caused by the large elastic constant of molybdenum.

Molecular dynamics computer simulations are employed to test this proposition and to investigate the effect of various projectile/sample properties on the angular emission. Argon projectiles with a kinetic energy of 10 keV and 20 keV at normal incidence are used to sputter surfaces of Mo and Ag samples, which have high and low elastic constants respectively. Argon clusters of various sizes (n=60, 250,500, 1000) are used to probe the influence of the projectile size. Finally, both flat and corrugated samples are bombarded to investigate the effect of the sample topography.

Our studies could not confirm the proposition of Chernysh et al. Even for 1000 impacts of 20 keV Ar₁₀₀₀ projectiles no ejection from flat Mo surface was recorded. This observation is attributed to a very large binding energy of Mo. Ejection is observed from Ag bombarded by Ar₁₀₀₀ but it is off-normal. Molybdenum begins to sputter when smaller 20 keV Ar_n projectiles are used. The angular distributions are found to be sensitive to the projectile size and the surface roughness. The shape of the angular spectra shifts towards the surface normal with a decrease of the projectile size. A similar trend is observed with increasing surface roughness. Reduction of the hovering cloud density of projectile atoms and material ejection taking place from randomly inclined surface areas are found to be responsible for such behavior. These effects occur, however, even on the Ag surface which has a low elastic constant.

ToF-SIMS is a powerful technique for imaging molecules with sub-µm lateral resolution. However, detection efficiency and sensitivity often are not high enough to obtain high-resolution images from large molecules. This disadvantage originates mainly from high fragmentation rates caused by high-energy primary ion bombardment as well as generally low ionization probabilities. Detection efficiency for larger molecules can be increased and fragmentation rates reduced by using Au_n⁺, Bi_n⁺, C₆₀⁺, or even large Ar_n⁺ clusters as primary cluster ions for increasing near-surface energy deposition. Another promising way would be to modify the surface with organic matrices using matrix-enhanced SIMS (ME-SIMS). In this approach, analyte molecules are embedded in low-weight organic matrices such as common MALDI matrices.

We have combined these two approaches to investigate the molecular yield of two peptides (bradykinin and melittin) as a function of primary ion species and cluster sizes as well as of preparation methods. Several different bismuth primary ions and large argon clusters in the mass range between 1000 and 2500 atoms per cluster were used to determine molecular yields and fragment rates. For the experiments, pure peptide solutions were spin-coated, and matrix-peptide mixtures were spray-coated onto silicon wafers. ME-SIMS samples were prepared using two classical MALDI matrices, 2,5-dihydroxybenzoic acid (DHB) and α-cyano-4-hydroxycinnamic acid (HCCA). With the data obtained, we were able to describe the molecular yield and the fragmentation rate of the analyzed peptides for various primary ion parameters. For bismuth primary cluster ions, the molecular yield was nearly constant in the ME-SIMS samples, while for the neat sample, an increase of the molecular yield with PI cluster size was observed. In contrast, with PI argon clusters the molecular yield decreases with the cluster size for neat samples, while it increases for ME-SIMS samples.

In terms of surface contamination analysis, an accurate mass scale for time-of-flight secondary ion mass spectrometry (TOF-SIMS) is needed to identify unknown high-mass peaks properly, although TOF-SIMS is powerful [1] for chemical analysis and usually has high mass resolution [2, 3]. Regarding mass scale calibration procedures, ISO 13084 formulated in 2011 is helpful. Typical TOF-SIMS spectra, however, do not contain secondary ions that meet ISO recommendations. For example, in order to identify mass m of an unknown high-mass peak, one of secondary ions for mass scale calibration should have mass m_A ≥ 0.55m. In this study, we developed a novel method for mass scale calibration in TOF-SIMS spectra using molecular ions of internal additives [4, 5].

Four kinds of target samples, solutions of didecyldimethylammonium bromide (Dc10dma), dioctadecyldimethylammonium chloride (Dc18dma), tetradecyltrimethylammonium chloride (C14TMA) and Tinuvin 770, individually, that easily generate high-mass molecular ion peaks were prepared. Then, solutions of octadecyltrimethylammonium chloride (C18TMA) and sodium hexadecyl sulfate (C16OSO3), individually, were prepared as the internal additives for positive or negative TOF-SIMS spectra. A target sample and an internal additive were selected one by one and were sequentially put on Si wafer by a dropper. TOF-SIMS measurements were carried out on TOF-SIMS 5 (ION-TOF, GmbH) with Bi₃⁺⁺ primary ion.

For instance, when C14TMA was employed as the target sample, it was added onto a Si wafer first and then C18TMA used as the internal additive was added to the sample. In this case, the molecular ion peak derived from C14TMA was not detected after the addition of C18TMA though it was highly detected in the positive TOF-SIMS spectrum before the addition. Since water solubility of C14TMA is higher than that of C18TMA, it is assumed that C14TMA is removed from the substrate when C18TMA was added. It should be confirmed that the target peak does not disappear from TOF-SIMS spectrum after addition of the internal additives. Finally, it is indicated that the internal additives are helpful for meeting ISO 13084 when an appropriate combination of internal additives for the target molecule is chosen.

[1] A. Hattori, J. Non-Cryst. Solids, (218), 1997, 196.

[2] I. S. Gilmore, F. M. Green and M. P. Seah, Surf. Interface Anal., (39), 2007, 817.

[3] Y. Abe, H. Itoh, S. Otomo and TOF-SIMS Working Group, J. Surf. Anal., (17), 2010, 186.

[4] D. Kobayashi, S. Otomo and H. Ito, J. Surf. Anal., (20), 2014, 187.

[5] D. Kobayashi, S. Otomo, S. Aoyagi and H. Ito, Surf. Interface Anal., (46), 2014, 229.

We have designed and implemented a variable temperature stage for a Cameca 7F-GEO SIMS with temperatures ranging from near liquid nitrogen temperatures to 300C with the primary use being to study the temperature dependent relocation of the Cs⁺ primary ion beam during SIMS analysis[1]. The variable temperature stage extends the usefulness of SIMS by operating across a wide temperature range. This capability makes it possible to analyze samples and perform studies that are not possible at room temperature, such as hydrated biological samples and fluid melt inclusions at cryogenic temperatures, high temperature phases of a material, and diffusion profiles requiring a wide range of temperatures. As a practical example, we will demonstrate the use of sample heating to quickly achieve low background levels of atmospherics (H, C, N, and O).

Even though this is not the first variable temperature stage on a Cameca[2–4], the complexity of the Cameca 7F-GEO main chamber and motorized stage provided new constraints for the stage design. Our variable temperature stage design is innovative and preserves the integrity of the original Cameca design, requiring minimal instrument modification and costing less than $5000. This work focuses on the design and implementation challenges for the variable temperature stage and a newly designed sample holder with high thermal conductivity for rapid thermal stabilization and excellent spot to spot repeatability.

[1] A. Giordani, J. Tuggle, C. Winkler, and J. Hunter, Surf. Interface Anal. 31 (2014).

[2] M. T. Bernius, S. Chandra, and G. H. Morrison, Rev. Sci. Instrum. 56, 1347 (1985).

[3] S. M. Hues, J. Vac. Sci. Technol. A Vacuum, Surfaces, Film. 4, 1942 (1986).

[4] M. Wiedenbeck, D. Rhede, R. Lieckefett, and H. Witzki, Appl. Surf. Sci. 231-232, 888 (2004).