Evaluation of Time-of-Flight Secondary Ion Mass Spectrometry Spectra of Peptides by Random Forest with Amino Acid Labels: Results from a Versailles Project on Advanced Materials and Standards Interlaboratory Study.

We report the results of a VAMAS (Versailles Project on Advanced Materials and Standards) interlaboratory study on the identification of peptide sample TOF-SIMS spectra by machine learning. More than 1000 time-of-flight secondary ion mass spectrometry (TOF-SIMS) spectra of six peptide model samples (one of them was a test sample) were collected using 27 TOF-SIMS instruments from 25 institutes of six countries, the U. S., the U. K., Germany, China, South Korea, and Japan. Because peptides have systematic and simple chemical structures, they were selected as model samples. The intensity of peaks in every TOF-SIMS spectrum was extracted using the same peak list and normalized to the total ion count. The spectra of the test peptide sample were predicted by Random Forest with 20 amino acid labels. The accuracy of the prediction for the test spectra was 0.88. Although the prediction of an unknown peptide was not perfect, it was shown that all of the amino acids in an unknown peptide can be determined by Random Forest prediction and the TOF-SIMS spectra. Moreover, the prediction of peptides, which are included in the training spectra, was almost perfect. Random Forest also suggests specific fragment ions from an amino acid residue Q, whose fragment ions detected by TOF-SIMS have not been reported, in the important features. This study indicated that the analysis using Random Forest, which enables translation of the mathematical relationships to chemical relationships, and the multi labels representing monomer chemical structures, is useful to predict the TOF-SIMS spectra of an unknown peptide.

Chemical Imaging of Buried Interfaces in Organic-Inorganic Devices Using Focused Ion Beam-Time-of-Flight-Secondary-Ion Mass Spectrometry.

Organic-inorganic hybrid materials enable the design and fabrication of new materials with enhanced properties. The interface between the organic and inorganic materials is often critical to the device's performance; therefore, chemical characterization is of significant interest. Because the interfaces are often buried, milling by focused ion beams (FIBs) to expose the interface is becoming increasingly popular. Chemical imaging can subsequently be obtained using secondary-ion mass spectrometry (SIMS). However, the FIB milling process damages the organic material. In this study, we make an organic-inorganic test structure to develop a detailed understanding of the processes involved in FIB milling and SIMS imaging. We provide an analysis methodology that involves a "clean-up" process using sputtering with an argon gas cluster ion source to remove the FIB-induced damage. The methodology is evaluated for two additive manufactured devices, an encapsulated strain sensor containing silver tracks embedded in a polymeric material and a copper track on a flexible polymeric substrate created using a novel nanoparticle sintering technique.

Embedding-Free Method for Preparation of Cross-Sections of Organic Materials for Micro Chemical Analysis Using Gas Cluster Ion Beam Sputtering.

We present a novel in situ mask method for the preparation of cross-sections of organic materials such as polymer multilayer films suitable for chemical imaging of buried interfaces. We demonstrate this method on a model buried interface system consisting of a piece of Scotch tape adhered to a PET substrate and a protective film used in consumer packaging. A high dose of gallium from a focused ion beam (FIB) was used to produce a damaged overlayer on the surface of the organic sample. The damaged layer has a significantly slower sputter rate compared to the native undamaged organic material. Therefore, during gas cluster ion beam (GCIB) depth profiling experiments the damaged layer functions as a mask, protecting the sample beneath and producing a cross-section at the edge of the mask. The FIB itself cannot be used directly to prepare the cross-section since the organic materials are easily damaged. A four step workflow is described including a final cleaning procedure to remove redeposited material from the cross-section. The workflow is completed in a few hours for samples up to 100 µm thickness. The method does not require sample embedding and is suited to automated analysis, which can be important benefits for industrial analysis where a variety of samples are analyzed routinely.

Identification and separation of protein, contaminant and substrate peaks using gentle-secondary ion mass spectrometry and the g-ogram.

RATIONALE: Secondary ion mass spectrometry (SIMS) is an important technique for the characterization of proteins at surfaces. However, interpretation of the mass spectra is complicated owing to confusion with peaks from contaminants and the substrate which is further compounded by complex fragmentation mechanisms. We test a new development of the G-SIMS method called the g-ogram to separate out spectral components without a priori information about which peaks to include in the analysis and which peaks relate to each component. METHODS: The effectiveness of the g-ogram method is investigated using a model system of lysozyme adsorbed onto a silicon wafer and indium tin oxide substrates. In the method, two SIMS spectra are acquired using Bi(+) and Mn(+) primary ions which create lower and higher fragmentation in the spectra, respectively. The g-ogram separates out components using a separation parameter that is related to the fragmentation energy. RESULTS: The g-ogram separates the spectrum of lysozyme adsorbed onto a silicon wafer into three components: (i) the substrate and PDMS contamination; (ii) a second, but unexpected, contaminant; and (iii) peaks from the protein amino acids. Similar results are achieved for the indium tin oxide substrate. In addition, evidence of fragments from plural amino acids with two candidate peaks at 140.12 Da and 185.08 Da is observed. CONCLUSIONS: The g-ogram method effectively separates out mass peaks relating to the substrate, contamination and protein without any a priori information or subjective decisions about which peaks to include in the analysis (so called 'peak picking'). This is a great help to analysts. We find two possible peaks from plural amino acids but no evidence of pluralities is found for peaks above 240 Da that are generated from when using Bi or Mn primary ions.

Sebaceoma and related neoplasms with sebaceous differentiation: a clinicopathologic study of 30 cases.

The classification of benign sebaceous neoplasms has been challenged both by the assertion that sebaceous adenomas are really carcinomas and by difficulties in drawing the boundaries between sebaceomas and other lesions. We performed a clinicopathologic study of 30 cases of basaloid neoplasms with sebaceous differentiation, excluding cases of definite sebaceous carcinoma with severe nuclear atypia invading deep within the subcutaneous tissue and those of ocular sebaceous carcinoma. We tried to classify sebaceous neoplasms in six categories with defined histopathologic criteria. All the neoplasms were characterized by aggregations of basaloid cells admixed with sebocytes and sebaceous duct-like structures located in the dermis with or without connection to the epidermis. The categories were 1) sebaceoma (14 cases); 2) trichoblastoma with sebaceous differentiation (3 cases); 3) apocrine poroma with sebaceous differentiation (2 cases); 4) low-grade sebaceous carcinoma (6 cases); 5) sebaceous carcinoma (4 cases); and 6) basal cell carcinoma with sebaceous differentiation (1 case). The sebaceoma was further subclassified as classic type (12 cases) or verruca/seborrheic keratosis type (2 cases). Although most sebaceomas can be distinguished from other lesions, there are problematic cases. We discuss the histopathologic diagnostic problems associated with sebaceoma and also argue in favor of the concept of sebaceous adenoma.