*3.2. Anthocyanins*

Anthocyanins are important natural products leading to the various colors in different parts of plants, including fruits, flowers, and grains [101,102]. Anthocyanins are derivatives of the salts of 2-phenylbenzopyrylium, naturally present as glycosylated molecules [103,104]. Subtle changes of the molecule structure can be revealed in the SERS spectra [105], which demonstrated the advantages of SERS used to identificate anthocyanidins. Benzopyrylium is the common moiety of the molecule structure of anthocyanidins with different phenyl rings, and different kinds of anthocyanidins exist in aqueous extracting solutions with different pH. Compared to resonance Raman, which could not distinguish the similar species, SERS can be used to further identify anthocyanidins. Zaffino et al. [106] identified some kinds of anthocyanidins using SERS and studied the influence of the pH on the SERS spectrum of six main anthocyanidins. They further provided SERS procedures of identificating anthocyanin in plant extracts. Optimized procedures of sample extraction and preparation were selected for different plant species and then detected by SERS measurement [107]. Luca et al. [108] reported the SERS spectra of analytes extracted from mulberry, gromwell, rhubarb and so on. It was proposed that the SERS spectrum was correlated with the molecule structure.

### *3.3. Chinese Herbal Medicinal Ingredients*

SERS is also used to characterize herb extracts. Gu et al. [109] presented an analytical method for the rapid detection and identification of bioactive substances from Chinese herbs by combining thin layer chromatography (TLC) and SERS. The limits of detection were 0.05–0.10 μM, which were far more sensitive than the UV lamp based method. The established method further enabled predicting and uncovering of unknown substances from Chinese herbs. The TLC-SERS strategy for the sample preparation procedure of *n*-butanol was illustrated in Figure 4.

**Figure 4.** Schematic of the proposed TLC-SERS method for n-butanol extract detection [109]. Reproduced with permission from [109]. Copyright Elsevier, 2018.

Zhang et al. [110] employed SERS and fluorescence spectroscopy to test the interaction of herb molecules with human serum albumin (HSA). The SERS methods were applied to predicting the molecule conformation on colloidal AuNPs. Similar transformations were found for four ginsenosides when combined with HSA, while the glucose and aglycone were exposed to fit suitable sites. Zheng et al. [111] developed a novel SERS method to monitor 5-demethylnobiletin produced in citrus and the SERS methods based on substrate or solution were all well correlated with high performance liquid chromatography (HPLC). The solution-based SERS method separated nobiletin by applying a procedure like "affinity chromatography". The substrate-based SERS could simply and quickly collect the "fingerprint" spectra. SERS had more advantages than HPLC methods in convenient, rapid characterizition and quantification of the production of 5-demethylnobiletin.

### *3.4. Molecular Fingerprint Identification of Plant*

The fingerprinting method is widely applied to the determination of molecules or the bond behavior in samples. A new SERS method was proposed to ge<sup>t</sup> the spectrum of tea species to detect the sample purity involved in different types of planting and processing. The fingerprint SERS spectrum of seven kinds of tea samples was obtained. Data processing method of Principal Component Analysis (PCA) was used to separate tea species and several models for different tea samples. The combined method of fingerprinting-PCA was accurate and rapid for the evaluation of different tea species [112,113].

Pollen extracts from various plant species have different beneficial biological effects [114]. Seifert et al. [115,116] analyzed aqueous extracts from different kinds of pollen using SERS with AuNPs substrate. The SERS spectra of targets were specific, and the different species could be distinguished to classify the genus. The accurate identification of pollen was achieved by analyzing the intrinsic information in SERS data. Results showed that SERS had good potential to characterize and identify pollen species, and could improve the study of pollen physiology. Routine investigations of food composition and vitamin/nutrient contents are challenged by food matrix complexity and low analyte content in samples. Radu et al. [117] simultaneously detected two B-vitamins using the SERS fingerprint method, which could sensitively identify analyte molecules at a low cost.

### *3.5. Other Plant Components*

SERS spectra of DNA can offer an assessment of the genetic identity of different kinds of plants [118,119]. Knowledge of genetic resources with high diversity is a precondition for developing novel species. Muntean et al. measured the half bandwidths of SERS of genomic DNAs in vitro-grown tissues of apple leaf. Results showed that the SERS method could be applied to studying the dynamics of DNA approaching the surface of the metallic substrate, with good perspectives of analyzing interactions of DNA-ligand or changes of DNA structure under environmental stress conditions [120,121]. They extended the application of this SERS method to other plant leaves, such as chrysanthemum, common sundew, edelweiss and so on [122]. The SERS spectra of genomic DNAs from tomato plants was also collected and the structural changes of DNAs undergoing cryopreservation were discussed [123]. Based on these works, they put forward that interactions of plant DNA and ligand or the precision DNA structure when approaching metallic surface could be further studied using SERS.

SERS can be used to investigate samples with weak Raman signals, for example juices and pulp. Camerlingo et al. [124] studied apple juices and pulp to verify the existence of fructose and pectin, which were related to the quality evaluation of these products. A home-made substrate fabricated with a glass slide decorated with AuNPs was designed and applied to the SERS detection. The obtained SERS spectra with legible Raman features provided useful information for the characterization of products in food processing.

Peanuts are a main life-threatening food allergen [125,126]. Gezer et al. [127] applied a biodegradable SERS technique to detecting Ara h1, the main kind of allergen protein, with LOD of 0.14 mg/mL. By using anti-Ara h1 monoclonal antibodies to functionalize the sensor surface, high specificity was achieved.

Palanco et al. [128] used plasmonic structures of silver nanoaggregates or films to enhance the detection of the chemical components of an onion layer. Results showed a competitive adsorption of molecules of onion and reporter. Different spectra from different parts of the layer indicated the complicated molecule structure of the plant. Shen et al. [129] presented nondestructive imaging of the living leaf using micro-Raman spectroscopy by delivering the carbon-encapsulated SERS tags into the living leaf. In vivo SERS spectra were used to investigate the distribution of tags, which could avoid interfering from autofluorescence. The novel modality of imaging provided SERS attractive potential for noninvasive biochemical imaging of living plants.

Cepeda-Perez et al. [130] reported the distribution and interaction of quantum dots (QDs) in the microalgae extracellular matrix. Changes in the Raman spectra of *Haematococcus pluvialis* microalgae caused by the adsorption of QDs were found by applying nano-sensors with bare anisotropic gold structures for SERS effect. This research demonstrated early QDs accumulation in plant cells which would benefit understanding of the environmental influence [131].

### **4. Application on the Detection of Microorganism Original Biomolecules**

### *4.1. Bacteria Original Biomolecules*

Sensitive and accurate pathogen detection is a key measure for ensuring public health due to the rapidly increasing infectious disease rate globally. The correct identification of pathogens in clinical or food samples assures the proper selection of clinical treatments or food safety procedures [132]. SERS is well-suited for bacteria detection from molecular to cellular level due to its sensitivity, selectivity, and compatiblity with other techniques. A representative work was presented by Meng et al. [133] in which a new type of SERS chip, consisting a sophisticated sandwich graphene (G)-AgNP-silicon (Si) nanohybrids, has been developed. The chip system could achieve both molecular detection and cellular analysis in samples. A schematic illustration of the developed platform was showed in Figure 5. The chip could realize sensitive and accurate quantification of adenosine triphosphate (ATP), with LOD of about 1 pM, and can also simultaneously capture, discriminate, and inactivate the bacteria. The efficiency of bacteria capture was 54% at the bacteria concentration of 10<sup>8</sup> CFU/mL, and 93% antibacterial rate could be reached 24 h after treating.

**Figure 5.** Schematic of the chip system for the analysis of ATP and bacteria [133]. Reproduced with permission from [133]. American Chemical Society, 2018.

For the application of SERS in the bacteria sensing, most of the studies focus on the cell detection, which is through the recognition of biomolecule on the cell membrane or inside the cell using antibodies, molecularly imprinted polymers or aptamers. The bacteria SERS signal can be significantly enhanced by these specifically recognized molecules, and the bacteria could be identified directly and visually from the SERS spectra [134]. There have been several reviews about the bacteria cell detection using SERS. Efrima et al. [135] reviewed studies using SERS for bacteria identification based on analyzing the spectra according to the nature of active centers and their distribution in the bacterium. Chauvet et al. [136] reviewed and proposed main strategies, such as methods to prepare the sample, from the bacterial culture conditions to the analysis of the spectra over the last 20 years. These reviews introduced the most recent reports about SERS detection of bacteria cells, which were comprehensive and detailed. So here the application of SERS on the detection of bacteria original biomolecules, rather than cells, was reviewed.

### 4.1.1. Bacteria DNAs

Enteritidis caused by *Salmonella enterica* is a common foodborne disease increasingly rising globally [137]. Draz et al. [138] established an integrated assay for DNA detection of *Salmonella*. Using Au-nanoprobes, the LOD of the developed method (66 CFU/mL) was about 100-fold lower than the conventional PCR method.

Gracie et al. reported a novel detection method of three meningitis pathogens using lambda-Exonuclease digestion of double-stranded DNA and SERS detection. Two complementary DNA probes were simultaneously hybridized to a target sequence. The obtained double stranded DNA was digested using lambda-exonuclease and then detected by SERS. Meningitis pathogens were detected with LOD in the range of pico-molar [139].

Another simple and low-cost platform was designed to sensitively detect bacterial DNA by SERS using AuNPs modified as reporter probes, by using the separation and enrichment function of magnetic beads. A good linear relationship was gained of the DNA concentration range of 5 pM to 5 nM. The LOD for the detection of bacterial DNA was about 5 pM [140].
