# An Improved POD Model for Fast Semi-Quantitative Analysis of Carbendazim in Fruit by Surface Enhanced Raman Spectroscopy

^{1}

^{2}

^{3}

^{4}

^{5}

^{6}

^{*}

## Abstract

**:**

## 1. Introduction

## 2. Results

#### 2.1. Establishment of a Rapid Detection Method for CBZ in Apple

#### 2.1.1. Selection of Raman Characteristic Peaks

^{−1}, 728 cm

^{−1}, 1000 cm

^{−1}, 1218 cm

^{−1}, 1260 cm

^{−1}and 1315 cm

^{−1}) were observed for CBZ standard and spiked sample compared to the blank sample. The six Raman shift peaks can be used as the qualitative characteristic peaks of CBZ in apple due to the clear peak shapes and distinct intensities. According to the SERS spectrum of CBZ and referring to the relevant literature [34,35,36]., the assignment of Raman shift and Vibrational Description were calculated and shown in Table 1.

#### 2.1.2. Establishment of POD Model

#### 2.1.3. Consistency Evaluation of Qualitative Methods Based on POD Model

- (1)
- Consistency evaluation between qualitative method and reference method

- (2)
- Consistency evaluation of qualitative method among different labs

#### 2.2. Establishment of a Semi-Quantitative Analysis Method for Carbendazim in Apple

#### 2.2.1. Establishment and Screening of Semi-Quantitative Models

^{−1}fit well with the theoretical Gaussian curve, and the fitting degree of the remaining five characteristic peaks was poor. When the CI was higher than 95%, the semi-quantitative model had a Raman intensity threshold of 1.4 × 10

^{4}at 630 cm

^{−1}, indicating that the concentration of CBZ in apple was no less than 5 mg/kg when the intensity of characteristic peak at 630 cm

^{−1}of CBZ in apple was greater than 1.4 × 10

^{4}.

^{−1}can well distinguish 0.5 mg/kg, 2.5 mg/kg and 5 mg/kg, and the concentration distribution results discriminated by the model are consistent with the reality. The test sets of the semi-quantitative models established at the remaining characteristic peaks have different degrees of overlap, and the results of model discrimination do not match the actual ones.

^{−1}and 1315 cm

^{−1}are 97 and 100, respectively. However, the semi-quantitative model at 1315 cm

^{−1}cannot distinguish the spiked samples at 0.5 mg/kg and 2.5 mg/kg (Figure 6). The semi-quantitative model at 630 cm

^{−1}was well fitted (Figure 5) and the concentration distribution is consistent with reality, so the semi-quantitative model at 630 cm

^{−1}is chosen as the optimal model for the semi-quantitative analysis of CBZ in apple. The model has a false positive rate of 0 at 0.5 mg/kg, a false positive rate of 5% at 2.5 mg/kg, and a POD of 100% at 5 mg/kg.

#### 2.2.2. Consistency Evaluation of Semi-Quantitative Methods among Different Labs

^{−1}, and the results are not affected by the environment.

## 3. Discussion

^{−1}, 728 cm

^{−1}, 1000 cm

^{−1}, 1218 cm

^{−1}, 1260 cm

^{−1}and 1315 cm

^{−1}in the CBZ standard and spiked samples were obvious. Therefore, these six Raman shifts were selected as the Raman characteristic peaks of CBZ in apple. The characteristic peaks of CBZ obtained here are the same as those of existing research [34]. For example, the peak at 630 cm

^{−1}is related to the C–C–C in-plane bending and the peak at 728 cm

^{−1}is attributed to the out-of-plane bending of the C-H bond in the benzene ring. The LOD of CBZ was 0.5 mg/kg, which was lower than the MRL of CBZ in apple (5 mg/kg). At the same time, it is four times lower than the LOD (2 mg/kg) of previous research methods [40].

^{−12}M and the Raman enhancement factor was as high as 10

^{8}. Wu et al. developed a simple and effective SERS tape based on biconical gold nanoparticles (BP-AuNPs) for monitoring methyl parathion residues on the surfaces of vegetables and fruits [21]. In real world applications, the screening of finite contaminants in food mainly depends on the MRL. If the added concentration of the sample is higher than the MRL, it is judged as a non-conforming product, otherwise it is judged as a qualified product. In order to fulfill the detection requirements of finite contaminants, we developed a threshold-based semi-quantitative analysis method for finite contaminants, which reduces the difficulty and cost of developing new materials and improves the detection efficiency. The model established at 630 cm

^{−1}was selected as the basis for semi-quantitative analysis of CBZ in apple after screening and verification. When the Raman intensity at 630 cm

^{−1}was greater than 1.4 × 10

^{4}, the concentration of CBZ in the sample was higher than the MRL (5 mg/kg). When the additive concentration was 5 mg/kg, the POD of this semi-quantitative method was 100%. The semi-quantitative method developed in this study only requires modelling based on a large number of samples from MRL, and the semi-quantitative results are determined by the Raman characteristic peak intensity. Compared to existing CBZ semi-quantitative analysis methods of CBZ, such as PLS-DA [42], this semi-quantitative does not require complex classification models and classification parameters such as the variable importance of variables in projection fraction, so it is simpler and more tractable. In order to ensure the SERS rapid detection method and semi-quantitative analysis method satisfy the evaluation standards of rapid detection methods (released by the State Food and Drug Administration in 2017). POD curve and dPOD curve were constructed for different methods or different labs within a certain concentration range, and the consistency of the methods was determined based on whether the POD is the same. The results showed that at the MRL level of CBZ (5 mg/kg), the SERS qualitative detection method was consistent among different labs, and the results were the same as the reference method. Compared with ‘Technology specification for the evaluation of food rapid detection products’ [37] , this evaluation method can show the change of sensitivity with concentration, and the LOD is well defined. It can compare the consistency of each concentration interval within the detection concentration range and can be applied to the consistency analysis between methods, environments, and instruments. The obtained POD curve can display the results visually, which is more statistically significant.

^{−1}in samples, which can improve the detection efficiency, and can be extended to other finite contaminants such as melamine in liquid milk and other pesticide residues in food. However, the training set of the semi-quantitative model requires a large amount of sample size (the number of samples >50), which leads to consume most of the time and energy before modeling. How to use less time to obtain the more sample information and reduce the preparation time is the bottleneck and future development direction of this research. Raman hyperspectral imaging technology is an advanced non-destructive testing technology that combines conventional imaging and spectroscopy to collect Raman spectral information of each pixel in space, so as to conduct qualitative, quantitative and localized analysis of samples [43]. Compared with SERS, the advantage of Raman hyperspectral imaging technology is that it can continuously collect a large number of spectral information through an automated sample platform, so as to obtain more sample information in less time. For example, Yang et al. applied Raman hyperspectral imaging technology to continuously collect spectral information of 100 pixels in 10 min [44]. However, it would have taken at least 50 min to collect the spectral information with SERS. Therefore, the semi-quantitative method developed in this study can be combined with the Raman hyperspectral imaging to shorten the sample information acquisition time and further improve the efficiency of sample screening.

## 4. Materials and Methods

#### 4.1. Samples, Reagents and Instruments

#### 4.2. Methods

#### 4.2.1. Sample Preparation

#### 4.2.2. SERS Detection

^{−1}, the integration time was 1 s, and the laser power was 200 mw. Spectral data were collected by Uspecral-PRO software. (Shanghai Oceanhood opto-electronics tech Co., Ltd., Shanghai, China).

#### 4.2.3. Data Processing

- (1)
- Screening of Raman characteristic peaks of CBZ in apple

^{−1}can be considered as the same characteristic peak.

- (2)
- Establishment and screening of semi-quantitative models

- The establishment of semi-quantitative models. Raman spectral data of spiked samples (number of samples > 50) at specific concentrations were collected as a training set for a semi-quantitative model. A histogram of the Raman intensity of each Raman characteristic peak at a specific additive concentration was obtained and the distribution of the intensity was viewed. If the intensity of the characteristic peak does not obey the Gaussian distribution, Raman intensity of the characteristic peak is not only affected by the random error of detection, so it is not suitable for the semi-quantitative model and should be eliminated. For the characteristic peaks whose Raman intensity follows a Gaussian distribution, calculate the intensity mean and standard deviation, determine the confidence level, and draw the confidence interval (CI). The Raman intensity corresponding to the lower limit of the confidence interval (CI > 95%) is used as the semi-quantitative threshold.
- Screening of semi-quantitative models. Raman data of spiked samples (number of samples > 20) at low concentration, half of specific concentration and specific concentration were collected as the test set for the semi-quantitative models of different characteristic peaks. The threshold value of the semi-quantitative model was used to determine whether the concentration in the sample exceeds a specific concentration, the POD was calculated, and the semi-quantitative model score under different characteristic peaks was computed. The higher the score of the model, the more accurate the semi-quantitative model will be. The model with the highest score and the Raman intensity conforming to the Gaussian distribution was selected as the optimal semi-quantitative model.

- 3.
- Result determination of semi-quantitative models. When the qualitative determination result is that the sample contains target substance, if the characteristic peak intensity of the target substance exceeds the semi-quantitative threshold, it is determined that the concentration of the target substance is not lower than a specific concentration. On the contrary, the concentration is lower than a specific concentration.
- (3)
- Evaluation of method based on POD model

- (1)
- when x = 0,

- (2)
- When $\mathrm{x}=\mathrm{N}$,

- (3)
- When $0\mathrm{x}\mathrm{N}$,

## 5. Conclusions

## 6. Patents

## Author Contributions

## Funding

## Institutional Review Board Statement

## Informed Consent Statement

## Data Availability Statement

## Acknowledgments

## Conflicts of Interest

## Sample Availability

## References

- Liu, Z.Y.; Chen, Y.; Han, J.H.; Chen, D.; Yang, G.Q.; Lan, T.T.; Li, J.M.; Zhang, K.K. Determination, dissipation dynamics, terminal residues and dietary risk assessment of thiophanate-methyl and its metabolite carbendazim in cowpeas collected from different locations in China under field conditions. J. Sci. Food Agric.
**2021**, 101, 5498–5507. [Google Scholar] [CrossRef] [PubMed] - Singh, S.; Singh, N.; Kumar, V.; Datta, S.; Wani, A.B.; Singh, D.; Singh, K.; Singh, J. Toxicity, monitoring and biodegradation of the fungicide carbendazim. Environ. Chem. Lett.
**2016**, 14, 317–329. [Google Scholar] [CrossRef] - Prashantkumar, W.; Sethi, R.S.; Pathak, D.; Rampal, S.; Saini, S.P.S. Testicular damage after chronic exposure to carbendazim in male goats. Toxicol. Environ. Chem.
**2012**, 94, 1433–1442. [Google Scholar] [CrossRef] - Daundkar, P.S.; Rampal, S. Evaluation of ameliorative potential of selenium on carbendazim induced oxidative stress in male goats. Environ. Toxicol. Pharmacol.
**2014**, 38, 711–719. [Google Scholar] [CrossRef] [PubMed] - GB 2763-2021; National Food Safety Standard—Maximum Residue Limits for Pesticides in Food. National Health Commission: Beijing, China, 2021.
- EU. On Maximum Residue Levels of Pesticides in or on Food and Feed of Plant and Animal Origin and Amending Council Directive 91/414/EEC; (EC) NO 396/2005; EU: Brussels, Belgium, 2005. [Google Scholar]
- Zhao, K.; Che, J.; Huang, A.; Shi, F. Determination of Residual Carbendazim in the Sugar Orange by HPLC. Hubei Agric. Sci.
**2010**, 49, 1193–1195. [Google Scholar] - Economou, A.; Botitsi, H.; Antoniou, S.; Tsipi, D. Determination of multi-class pesticides in wines by solid-phase extraction and liquid chromatography-tandem mass spectrometry. J. Chromatogr. A
**2009**, 1216, 5856–5867. [Google Scholar] [CrossRef] - de Macedo, J.F.; Alves, A.A.C.; Sant’Anna, M.V.S.; Cunha, F.G.C.; Oliveira, G.d.A.R.; Liao, L.M.; Sussuchi, E.M. Electrochemical determination of carbendazim in grapes and their derivatives by an ionic liquid-modified carbon paste electrode. J. Appl. Electrochem.
**2022**, 52, 729–742. [Google Scholar] [CrossRef] - Jiang, X.; Li, D.; Xu, X.; Ying, Y.; Li, Y.; Ye, Z.; Wang, J. Immunosensors for detection of pesticide residues. Biosens. Bioelectron.
**2008**, 23, 1577–1587. [Google Scholar] [CrossRef] - Lee, H.S.; Rahman, M.M.; Chung, H.S.; Kabir, H.; Yoon, K.S.; Cho, S.K.; Abd El-Aty, A.M.; Shim, J.H. An effective methodology for simultaneous quantification of thiophanate-methyl, and its metabolite carbendazim in pear, using LC-MS/MS. J. Chromatogr. b-Anal. Technol. Biomed. Life Sci.
**2018**, 1095, 1–7. [Google Scholar] [CrossRef] - Liu, Z.; Liu, W.; Wu, Q.; Zang, X.; Zhou, X.; Zeng, X.; Wang, Z. Determination of carbendazim and thiabendazole in apple juice by hollow fibre-based liquid phase microextraction-high performance liquid chromatography with fluorescence detection. Int. J. Environ. Anal. Chem.
**2012**, 92, 582–591. [Google Scholar] [CrossRef] - Langer, J.; Jimenez de Aberasturi, D.; Aizpurua, J.; Alvarez-Puebla, R.A.; Auguié, B.; Baumberg, J.J.; Bazan, G.C.; Bell, S.E.J.; Boisen, A.; Brolo, A.G.; et al. Present and Future of Surface-Enhanced Raman Scattering. ACS Nano
**2020**, 14, 28–117. [Google Scholar] [CrossRef] [PubMed] [Green Version] - Zhu, J.; Agyekum, A.A.; Kutsanedzie, F.Y.H.; Li, H.; Chen, Q.; Ouyang, Q.; Jiang, H. Qualitative and quantitative analysis of chlorpyrifos residues in tea by surface-enhanced Raman spectroscopy (SERS) combined with chemometric models. LWT
**2018**, 97, 760–769. [Google Scholar] [CrossRef] - Craig, A.P.; Franca, A.S.; Irudayaraj, J. Surface-enhanced Raman spectroscopy applied to food safety. Annu. Rev. Food Sci. Technol.
**2013**, 4, 369–380. [Google Scholar] [CrossRef] [PubMed] - Jiang, L.; Hassan, M.M.; Ali, S.; Li, H.; Sheng, R.; Chen, Q. Evolving trends in SERS-based techniques for food quality and safety: A review. Trends Food Sci. Technol.
**2021**, 112, 225–240. [Google Scholar] [CrossRef] - Pilot, R. SERS detection of food contaminants by means of portable Raman instruments. J. Raman Spectrosc.
**2018**, 49, 954–981. [Google Scholar] [CrossRef] - Liu, B.; Zheng, S.; Li, H.; Xu, J.; Tang, H.; Wang, Y.; Wang, Y.; Sun, F.; Zhao, X. Ultrasensitive and facile detection of multiple trace antibiotics with magnetic nanoparticles and core-shell nanostar SERS nanotags. Talanta
**2022**, 237, 122955. [Google Scholar] [CrossRef] - Dugandžić, V.; Kupfer, S.; Jahn, M.; Henkel, T.; Weber, K.; Cialla-May, D.; Popp, J. A SERS-based molecular sensor for selective detection and quantification of copper(II) ions. Sens. Actuators B Chem.
**2019**, 279, 230–237. [Google Scholar] [CrossRef] - Wei, C.; Li, M.; Zhao, X. Surface-Enhanced Raman Scattering (SERS) With Silver Nano Substrates Synthesized by Microwave for Rapid Detection of Foodborne Pathogens. Front. Microbiol.
**2018**, 9, 2857. [Google Scholar] [CrossRef] - Wu, H.; Luo, Y.; Hou, C.; Huo, D.; Zhou, Y.; Zou, S.; Zhao, J.; Lei, Y. Flexible bipyramid-AuNPs based SERS tape sensing strategy for detecting methyl parathion on vegetable and fruit surface. Sens. Actuators B Chem.
**2019**, 285, 123–128. [Google Scholar] [CrossRef] - Chen, X.; Lin, M.; Sun, L.; Xu, T.; Lai, K.; Huang, M.; Lin, H. Detection and quantification of carbendazim in Oolong tea by surface-enhanced Raman spectroscopy and gold nanoparticle substrates. Food Chem.
**2019**, 293, 271–277. [Google Scholar] [CrossRef] - Zhou, H.L.; Li, X.D.; Wang, L.H.; Liang, Y.F.; Jialading, A.; Wang, Z.S.; Zhang, J.G. Application of SERS quantitative analysis method in food safety detection. Rev. Anal. Chem.
**2021**, 40, 173–186. [Google Scholar] [CrossRef] - Li, D.; Yao, D.; Li, C.; Luo, Y.; Liang, A.; Wen, G.; Jiang, Z. Nanosol SERS quantitative analytical method: A review. TrAC Trends Anal. Chem.
**2020**, 127, 115885. [Google Scholar] [CrossRef] - Long, Y.; Li, H.; Du, Z.; Geng, M.; Liu, Z. Confined Gaussian-distributed electromagnetic field of tin(II) chloride-sensitized surface-enhanced Raman scattering (SERS) optical fiber probe: From localized surface plasmon resonance (LSPR) to waveguide propagation. J. Colloid Interface Sci.
**2021**, 581, 698–708. [Google Scholar] [CrossRef] [PubMed] - Wilrich, P.-T. The determination of precision of qualitative measurement methods by interlaboratory experiments. Accredit. Qual. Assur.
**2010**, 15, 439–444. [Google Scholar] [CrossRef] - Fleiss, J.L.; Levin, B.; Paik, M.C. Statistical Methods for Rates and Proportions, 2nd ed.; John Wiley & Sons: Hoboken, NJ, USA, 2013. [Google Scholar]
- Wehling, P.; LaBudde, R.A.; Brunelle, S.L.; Nelson, M.T. Probability of detection (POD) as a statistical model for the validation of qualitative methods. J. AOAC Int.
**2011**, 94, 335–347. [Google Scholar] [CrossRef] - van Wieringen, W.N.; de Mast, J. Measurement System Analysis for Binary Data. Technometrics
**2008**, 50, 468–478. [Google Scholar] [CrossRef] - Gondim, C.d.S.; Junqueira, R.G.; de Souza, S.V.C.; Callao, M.P.; Ruisánchez, I. Determining performance parameters in qualitative multivariate methods using probability of detection (POD) curves. Case Study Two Common Milk Adulterants. Talanta
**2017**, 168, 23–30. [Google Scholar] [CrossRef] - Jarvis, B.; Wilrich, C.; Wilrich, P.T. Estimation of the POD Function and the LOD of a Binary Microbiological Measurement Method from an Interlaboratory Experiment. J. AOAC Int.
**2019**, 102, 1617–1623. [Google Scholar] [CrossRef] - Uhlig, S.; Frost, K.; Colson, B.; Simon, K.; Mäde, D.; Reiting, R.; Gowik, P.; Grohmann, L. Validation of qualitative PCR methods on the basis of mathematical–statistical modelling of the probability of detection. Accredit. Qual. Assur.
**2015**, 20, 75–83. [Google Scholar] [CrossRef] - Syed Akbar Ali, M.S.; Kumar, A.; Rao, P.B.; Tammana, J.; Balasubramaniam, K.; Rajagopal, P. Bayesian synthesis for simulation-based generation of probability of detection (PoD) curves. Ultrasonics
**2018**, 84, 210–222. [Google Scholar] [CrossRef] - Furini, L.N.; Sanchez-Cortes, S.; López-Tocón, I.; Otero, J.C.; Aroca, R.F.; Constantino, C.J.L. Detection and quantitative analysis of carbendazim herbicide on Ag nanoparticles via surface-enhanced Raman scattering. J. Raman Spectrosc.
**2015**, 46, 1095–1101. [Google Scholar] [CrossRef] - Ren, X.; Feng, X.; Li, X.; Li, X. Preparation of silver with an ultrathin molecular imprinted layer for detection of carbendazim by SERS. Chem. Pap.
**2021**, 75, 6477–6485. [Google Scholar] [CrossRef] - Sharma, V.; Krishnan, V. Fabrication of highly sensitive biomimetic SERS substrates for detection of herbicides in trace concentration. Sens. Actuators B Chem.
**2018**, 262, 710–719. [Google Scholar] [CrossRef] - DB36/T 1334-2020; Technology Specification for the Evaluation of Food Rapid Detection Products. Jiangxi Provincial Administration for Market Regulation: Nanchang, China, 2020.
- GB/T 23380-2009; Determination of Carbendazim Residues in Fruits and Vegetables—HPLC Method. Standardization Administration of China: Beijing, China, 2009.
- Neng, J.; Zhang, Q.; Sun, P.L. Application of surface-enhanced Raman spectroscopy in fast detection of toxic and harmful substances in food. Biosens. Bioelectron.
**2020**, 167, 112480. [Google Scholar] [CrossRef] - Shen, Z.D.; Fan, Q.Z.; Yu, Q.; Wang, R.; Wang, H.; Kong, X.M. Facile detection of carbendazim in food using TLC-SERS on diatomite thin layer chromatography. Spectrochim. Acta Part A-Mol. Biomol. Spectrosc.
**2021**, 247, 119037. [Google Scholar] [CrossRef] - Sivashanmugan, K.; Nguyen, V.-H.; Nguyen, B.-S. Tailoring a novel Au nanodot arrays on graphene substrate for a highly active Surface-Enhanced Raman Scattering (SERS). Mater. Lett.
**2020**, 271, 127807. [Google Scholar] [CrossRef] - Valderrama, L.; Valderrama, P.; Carasek, E. A semi-quantitative model through PLS-DA in the evaluation of carbendazim in grape juices. Food Chem.
**2022**, 368, 130742. [Google Scholar] [CrossRef] - Lohumi, S.; Lee, H.; Kim, M.S.; Qin, J.; Cho, B.-K. Raman hyperspectral imaging and spectral similarity analysis for quantitative detection of multiple adulterants in wheat flour. Biosyst. Eng.
**2019**, 181, 103–113. [Google Scholar] [CrossRef] - Yang, Q.; Niu, B.; Gu, S.; Ma, J.; Zhao, C.; Chen, Q.; Guo, D.; Deng, X.; Yu, Y.; Zhang, F. Rapid Detection of Nonprotein Nitrogen Adulterants in Milk Powder Using Point-Scan Raman Hyperspectral Imaging Technology. ACS Omega
**2022**, 7, 2064–2073. [Google Scholar] [CrossRef] - Brain, C.W.; Mi, J. On some properties of the quantiles of the chi-square distribution and their applications to interval estimation. Commun. Stat. Theory Methods
**2001**, 30, 1851–1867. [Google Scholar] [CrossRef]

**Figure 3.**The detection situation of CBZ in apple by Raman qualitative method and reference method. Note: (

**A**,

**B**) are the POD curve and the dPOD curve of the Raman qualitative detection method and the reference method, respectively.

**Figure 4.**The detection situation of CBZ in apple by Raman qualitative method in different labs. Note: (

**A**,

**B**) are the POD curve and the dPOD curve of the Raman qualitative detection method between different labs, respectively.

**Figure 5.**Semi-quantitative model of CBZ in apple at the MRL (5 mg/kg). Note: (

**A**–

**F**) represent the semi-quantitative models established by different Raman shifts, respectively. (

**A**) 630 cm

^{−1}; (

**B**) 728 cm

^{−1}; (

**C**) 1000 cm

^{−1}; (

**D**) 1218 cm

^{−1}; (

**E**) 1260 cm

^{−1}; (

**F**) 1315 cm

^{−1}. SD: Standard deviation.

**Figure 6.**Validation of the semi-quantitative model on the MRL (5 mg/kg) of CBZ in apple. (

**A**–

**F**) represent the validation of semi-quantitative models established by different Raman shifts, respectively. (

**A**) 630 cm

^{−1}; (

**B**) 728 cm

^{−1}; (

**C**) 1000 cm

^{−1}; (

**D**) 1218 cm

^{−1}; (

**E**) 1260 cm

^{−1}; (

**F**) 1315 cm

^{−1}. SD: Standard deviation.

**Figure 7.**The detection situation of CBZ in apple by semi-quantitative method in different labs. Note: (

**A**,

**B**) are the POD curve and dPOD curve of semi-quantitative detection method between different labs, respectively.

Wavenumber (cm^{−1}) | Vibrational Description | Wavenumber (cm^{−1}) | Vibrational Description |
---|---|---|---|

630 | ring stretching and C-C bending | 1315 | ring stretching |

728 | C-C bending and C-O-CH_{3} bending | 1370 | C—N stretch |

1000 | C-N bending and C-C bending and C-O-CH3 stretching | 1460 | N-H bending and C-H bending |

1218 | C-C stretch, C-C bending and N-H bending | 1523 | N-H bending and C-N stretch |

1260 | C-H bending and N-H bending |

Concentration (mg/kg) | x | N | POD | LCL | UCL |
---|---|---|---|---|---|

0.000 | 0 | 10 | 0.000 | 0.000 | 0.280 |

0.050 | 0 | 50 | 0.000 | 0.000 | 0.070 |

0.100 | 40 | 50 | 0.800 | 0.670 | 0.890 |

0.500 | 39 | 40 | 0.975 | 0.871 | 1.000 |

1.000 | 40 | 40 | 1.000 | 0.910 | 1.000 |

2.500 | 40 | 40 | 1.000 | 0.910 | 1.000 |

5.000 | 40 | 40 | 1.000 | 0.910 | 1.000 |

**Table 3.**Evaluation of the consistency between the Raman qualitative method and the reference method for CBZ in apple.

Method | SERS | HPLC | Difference in POD (dPOD) | ||||||||
---|---|---|---|---|---|---|---|---|---|---|---|

Concentration (mg/kg) | x | N | POD | LCL | UCL | x | N | POD | LCL | UCL | |

0.00 | 0 | 10 | 0.000 | 0.000 | 0.280 | 0 | 10 | 0 | 0.000 | 0.280 | 0.000 |

0.01 | 0 | 50 | 0.000 | 0.000 | 0.070 | 0 | 50 | 0 | 0.000 | 0.070 | 0.000 |

0.10 | 40 | 50 | 0.800 | 0.670 | 0.890 | 50 | 50 | 1 | 0.930 | 1.000 | 0.200 |

0.50 | 39 | 40 | 0.975 | 0.870 | 1.000 | 40 | 40 | 1 | 0.910 | 1.000 | 0.025 |

2.50 | 40 | 40 | 1.000 | 0.910 | 1.000 | 40 | 40 | 1 | 0.910 | 1.000 | 0.000 |

5.00 | 40 | 40 | 1.000 | 0.910 | 1.00 | 40 | 40 | 1 | 0.910 | 1.000 | 0.000 |

Lab | Ⅰ | Ⅱ | Difference in POD (dPOD) | ||||||||
---|---|---|---|---|---|---|---|---|---|---|---|

Concentration (mg/kg) | x | N | POD | LCL | UCL | x | N | POD | LCL | UCL | |

0.00 | 0 | 10 | 0.000 | 0.000 | 0.280 | 0 | 10 | 0.000 | 0.000 | 0.280 | 0.000 |

0.05 | 0 | 50 | 0.000 | 0.000 | 0.070 | 0 | 50 | 0.000 | 0.000 | 0.070 | 0.000 |

0.10 | 40 | 50 | 0.800 | 0.670 | 0.890 | 48 | 50 | 0.960 | 0.865 | 0.989 | 0.160 |

0.50 | 39 | 40 | 0.975 | 0.870 | 1.000 | 40 | 40 | 1.000 | 0.910 | 1.000 | 0.025 |

2.50 | 40 | 40 | 1.000 | 0.910 | 1.000 | 40 | 40 | 1.000 | 0.910 | 1.000 | 0.000 |

5.00 | 40 | 40 | 1.000 | 0.910 | 1.000 | 40 | 40 | 1.000 | 0.910 | 1.000 | 0.000 |

Concentration (mg/kg) | 0.5 | 2.5 | 5 | Score (S) | |||
---|---|---|---|---|---|---|---|

Peaks (cm^{−1}) | x | POD1 | x | POD2 | x | POD3 | |

630 | 0 | 0 | 2 | 0.050 | 40 | 1 | 97 |

728 | 0 | 0 | 11 | 0.275 | 40 | 1 | 83 |

1000 | 8 | 0.200 | 39 | 0.975 | 40 | 1 | 44 |

1218 | 0 | 0 | 29 | 0.725 | 40 | 1 | 60 |

1260 | 9 | 0.225 | 40 | 1 | 40 | 1 | 43 |

1315 | 0 | 0 | 0 | 0 | 40 | 1 | 100 |

Lab | Ⅰ | Ⅱ | Difference in POD (dPOD) | ||||||||
---|---|---|---|---|---|---|---|---|---|---|---|

Concentration (mg/kg) | x | N | POD | LCL | UCL | x | N | POD | LCL | UCL | |

0.50 | 0 | 40 | 0.000 | 0 | 0.088 | 7 | 40 | 0.175 | 0.087 | 0.320 | 0.175 |

2.50 | 25 | 40 | 0.625 | 0.470 | 0.758 | 34 | 40 | 0.850 | 0.710 | 0.930 | 0.225 |

5.00 | 40 | 40 | 1.000 | 0.912 | 1.000 | 40 | 40 | 1.000 | 0.912 | 1.000 | 0.000 |

Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations. |

© 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).

## Share and Cite

**MDPI and ACS Style**

Yang, Q.; Lin, H.; Ma, J.; Chen, N.; Zhao, C.; Guo, D.; Niu, B.; Zhao, Z.; Deng, X.; Chen, Q.
An Improved POD Model for Fast Semi-Quantitative Analysis of Carbendazim in Fruit by Surface Enhanced Raman Spectroscopy. *Molecules* **2022**, *27*, 4230.
https://doi.org/10.3390/molecules27134230

**AMA Style**

Yang Q, Lin H, Ma J, Chen N, Zhao C, Guo D, Niu B, Zhao Z, Deng X, Chen Q.
An Improved POD Model for Fast Semi-Quantitative Analysis of Carbendazim in Fruit by Surface Enhanced Raman Spectroscopy. *Molecules*. 2022; 27(13):4230.
https://doi.org/10.3390/molecules27134230

**Chicago/Turabian Style**

Yang, Qiaoling, Hong Lin, Jinge Ma, Niannian Chen, Chaomin Zhao, Dehua Guo, Bing Niu, Zhihui Zhao, Xiaojun Deng, and Qin Chen.
2022. "An Improved POD Model for Fast Semi-Quantitative Analysis of Carbendazim in Fruit by Surface Enhanced Raman Spectroscopy" *Molecules* 27, no. 13: 4230.
https://doi.org/10.3390/molecules27134230