◆ 1- Microbial sampling: As many microorganisms can grow repeatedly on agar plates, the first step is to find the correct culture conditions. The simple and common technique of culturing cells on agar plates enables convenient sample preparation. For example, bacteria can be cultured in quadrants of different cell densities on agar plates for 16-24 hours to obtain very reliable results. Collect colonies only from the third quadrant. The study found that the key to the success of this method lies in the standardization of collection conditions and the consistency of bacterial growth on agar plates. Improvements can also be made, such as switching to liquid culture medium or changing the cultivation time/temperature. But it is necessary to ensure that all measurements are conducted under the same conditions to obtain reliable results.

◆ 2-Bacterial measurement technology: The technology used for microbial identification needs to be simple and highly reproducible. When microbial solutions slowly dry, microbial cells produce solid biofilms. By optimizing this process, an easy to operate sample preparation procedure was obtained that meets the repeatability requirements for infrared measurement. Infrared light penetrates biological membranes and is absorbed by different molecular components present in cells, producing unique chemical fingerprints for each tested strain.

The most concerning issue for FT-IR microbial identification is the accuracy of the results, which depends crucially on the quality of the reference spectral library. Microbial spectral libraries are different from general spectral libraries, and in order to ensure the reliability of results, a systematic spectral library needs to be established under standardized conditions.

Firstly, it is necessary to ensure the reproducibility of microbial cultivation. The culture medium, culture conditions (such as time, temperature), and sample pretreatment all have an impact on infrared spectroscopy, so standardized operating procedures need to be developed for these conditions.

Secondly, non homogeneity and variability during the sample pretreatment process (such as sample size, dryness of bacterial film, etc.) as well as measurement factors (such as light scattering caused by the particle shape of the sample) can affect the infrared spectrum. Therefore, multiple measurements of the same sample are required to ensure the reproducibility of the infrared spectrum.

Thirdly, the reliability of unknown samples for infrared identification is based on the reliability of the spectral library of the reference spectrum. Therefore, reliable methods must be used to identify reference strains. Otherwise, incorrect information in the spectral library can lead to deviations in the final identification results. Combining multiple significant spectral regions can construct a database of known bacterial strains. In order to classify and identify unknown bacteria, the spectral information is compared with a reference database of known bacteria, displaying microorganisms whose spectral patterns match the unknown strains. The similarity of spectra of different microorganisms can also be graphically described in the form of a tree diagram.