Membranes, Viruses, and Bacteria

TERS and SERS [101, 102] do not need labels to chemically characterize membranes as shown in a study with an artificial biotin-modified lipid bilayer and biotin-PS [103]. Using the very specific affinity of the covalent bond between streptavidin and biotin, it was possible to attach streptavidin molecules to the membrane. The aim of the study was the differentiation of lipid and protein containing areas. The distinction of the components was accomplished by analyzing specific marker bands in the TERS spectra. Marker bands for lipids are vibrational modes of the phosphate group (830 and 1080 cm ) and of the polar head group. 

Probing bacterial surfaces, it is important to have a high accuracy to differentiate pathogenic species from nonpathogenic species [92]. Accuracies of up to 97% are realistic as demonstrated by an SERS study with 14 different but closely related bacteria strands [109]. 

The first examined bacterium with TERS was bacterium Staphylococcus epidermidis [110—112]. More statistical measurements on S. epidermidis were performed by collecting 15 spectra at one point per cell and five cells in total [112]. From these spectra, a single average spectrum for each cell was generated and characteristic bands for teichoic acid (a (l-3)-linked poly(glycerol 

If the data set for TERS measurements becomes large ( 100 spectra), such an automated reduction of the data is most useful. However, due to the already discussed intrinsic spectral shifts and intensity variations in TERS (and also SERS), the position of an individual vibrational mode might be shifted depending on its orientation and environment. Hence, it is challengingto build a robust and reliable 

A differentiation is much more challenging if virions with similar size and shape are examined by TERS. Here, one can take advantage of the high specificity of Raman spectroscopy. A successful discrimination of single viruses is only feasible 

---