*3.2. Other Effects of MOFs for Improving Selectivity and Sensitivity*

Owing to the large specific surface area [20], uniform porosity [54], structural adaptability and flexibility [73,114], and ease of functionalization [63], MOFs as shells can enhance the selectivity and specificity of SERS substrates through physical adsorption and chemical recognition [72,110]. The introduction of additional functional layers, such as aptamers and antibodies, can ensure selective adsorption through multiple molecular interactions [23]. In addition, other factors, such as the thickness of MOFs, the nature of the metal species, and the functional groups afforded by the organic linkers [23,126], can also affect the selectivity.

The MOF shell over the plasmonic SERS substrate can serve as a sieve to allow only the target of interest to diffuse to hotspots, as well as to facilitate an efficient reaction with the Raman label by prolonging the contact time. VOCs are important biomarkers for early diagnoses of diseases [127], but the low concentration and high mobility of gaseous molecules result in insufficient collisions between gas molecules and SERS substrates, significantly compromising the detection sensitivity [128]. Wang and colleagues [64] used a GSPs@ZIF-8 SERS substrate to selectively detect aldehydes, a lung cancer biomarker in patients' exhalation. The detection of gaseous aldehydes is currently limited by its small Raman cross section and poor adsorptivity on SERS substrates. The ZIF-8 shell allows the small vapor-phase aromatic compounds such as benzaldehyde, glutaricdiadehyde, and 4-ethylbenzaldehyde, but not 2-naphthaldehyde, to diffuse into the channel due to the sieving effect (Figure 7c–e). The prolonged contact time between gaseous molecules and the GSPs hotspots can facilitate the reaction of the analyte with the Raman-active probe molecule p-aminothiophenol (4-ATP) through the Schiff base reaction to generate a distinguished Raman signal. Apart from the physical adsorption, the MOFs can also

selectively adsorb analytes through chemical recognition. Yusuke and colleagues [72] prepared hierarchical mesoporous Au films coated with homochiral MOF, which were able to realize ultrasensitive and enantioselective sensing of pseudoephedrine (PE) in complex bio-samples. The chiral ligands (a kind of alanine derivative) were used to endow the MOF with the homochiral property to distinguish (+)-PE. For analysis of PE in blood serum, the matrix effect was reduced by taking the advantage of the pore size limit to prevent large molecules from entering the MOF shell. This smart SERS substrate enabled an extremely low detection limit (10−<sup>12</sup> M) in the complex biomatrix without preliminary separation.

In reality, the size of most of the biomarkers does not match that of the MOF channel. Efforts have been made to detect the metabolites as an indicator of the amount of biomarkers. By using an in situ reduction strategy, Zhou and colleagues [69] constructed a AuNPs@MIL-101@GOx (or AuNPs@MIL-101@LOx) nanoplatform for the detection of glucose/lactate, which are important neurochemicals associated with many physiological and pathological brain functions, such as ischemia, learning, and memory. The Ramaninactive reporter leucomalachite green (LMG) was oxidized into the active malachite green (MG) through a cascade of catalytic processes, and the signal intensity was used to indicate the amount of glucose/lactate. This nanoplatform has also been used to evaluate the therapeutic effects of astaxanthin for the purpose of alleviating cerebral ischemic injuries.

The metal-MOF-based SERS substrates have also been used in drug delivery and bioimaging due to the high surface area and porosity, excellent biocompatibility, and stability of MOFs [5]. The exposed active sites on the surface of the MOFs can be used for functionalization with recognition units, and the large specific surface area provides sufficient accessible binding sites for target molecules [80]. For example, the carboxyl group on the surface of the Au@Cu3(BTC)2 nanoparticles was used for functionalization with the aptamers [5]. In another example, the ZIF-8 shell on the Au@Ag surface was used to conjugate with IgG antibodies and recombinant nanobodies [70], based on the high affinity of polyhistidine toward transition metal ions [129]. The Au@Ag@ZIF-8 substrate was used to detect CD44 and EGFR biomarkers in mixed cell cultures, indicating the potential of the nanoprobes for SERS imaging and multiplexed bio-detection.

Farha and colleagues [113] reported the controlled encapsulation of gold nanorods (AuNRs) with a scu-topology Zr-MOF (NU-901) via the room-temperature assembly of MOF on AuNRs seeds. After incubating AuNR@NU-901 with a mixture of thiolated polystyrene (PST-SH; Mw = 5000 g/mol) and 4 -mercaptobiphenylcarbonitrile (BPTCN) molecules (roughly 15 × 7 Å along the thiol−CN axis and phenyl axis), the resulting spectrum closely matched with that of BPTCN alone. This result demonstrated that the prepared AuNR@MOFs were able to take-up molecules with suitable sizes and block large molecules from the pores, thus facilitating highly selective SERS detection at the AuNR ends.

#### *3.3. Enhancement of the Stability, Homogeneity, and Reproducibility of SERS Substrates*

The instability and reproducibility of the plasmonic nanoparticles under harsh environment represent inherent challenges in SERS detection [4,63,65,72]; therefore, a protective layer is necessary. Due to their excellent chemical and thermal stability [130], as well as their mechanical robustness [131], MOFs are an ideal candidate to act as a stabilizing layer. For example, Liang and colleagues [65] prepared a dense MIL-101(Cr) film on the rough titanium oxide foil via a secondary growth method, and then the Ag<sup>+</sup> was reduced to Ag on the surface of the film to form the Ag@MIL-101(Cr) film SERS substrate. In such a way, the excellent SERS effect and the high reproducibility of the SERS substrate were achieved to realize the detection of nitrofurantoin (down to 10−<sup>7</sup> M) without any complex subsequent procedures. Li and colleagues reported a highly sensitive and continuously stable 3D substrate (Cu2O@SiO2@ZIF-8@Ag) for SERS detection of phenol red with a LOD of 5.76 × <sup>10</sup>−<sup>12</sup> mol·L−<sup>1</sup> and LOQ of 1.92 × <sup>10</sup>−<sup>12</sup> mol·L<sup>−</sup>1, and a high enhancement factor of 1.7 × 107 was achieved even after 35 days [73].

The sensitivity and quantification performance of the SERS technique often contradict one another due to the modest reproducibility of the SERS substrate. Yang and colleagues constructed an integrated SERS platform with analyte enrichment and analyte filtration functions (referred to as AEF-SERS) to simultaneously achieve a good quantification performance and ultra-high sensitivity (Figure 9) [75]. In their work, single Au NRs were separated from each other through the coating of a thick ZIF-8 shell to form a AuNR@ZIF-8 submicroscale truncated rhombic dodecahedron (TRD); thus, a homogeneous SERS substrate was produced to improve the reproducibility of the detection. The separation of Au NRs may reduce the number of hotspots, thus compromising the sensitivity. However, the authors were still able to successfully realize a highly sensitive detection by constructing a polydimethylsiloxane (PDMS) brush surface that was capable of shrinking the analyte dispersion area by a million-fold in order to enrich the analyte.

**Figure 9.** (**a**) Schematic illustration of the working principle of the AEF-SERS platform, showing application of the test solution (I), the enrichment and sieving effect (II), and the formation of analyte-SERS substrate aggregates (III) for selective and sensitive quantification of the analyte. (**b**) SERS spectra of 4-NBT at different concentrations using AuNR@ZIF-8 TRDs on the PDMS brush surface. (**c**) The calibration curve of 4-NBT at 1329 cm−1. (**d**) SERS mapping of 10 nM 4-NBT at 1329 cm−<sup>1</sup> using AuNR@ZIF-8 TRDs. (**e**) Intensity variations at 1329 cm−<sup>1</sup> from randomly chosen 40 SERS spectra using AuNR@ZIF-8 TRDs [75]. Copyright 2020, American Chemical Society.

The integration of noble metals and MOFs can speed up the development of the SERS technique. The high sensitivity can be partially explained by pre-concentration of the analyte through physical adsorption and chemical recognition. Aptamers, antibodies, and other recognition units can be easily used to modify MOFs in order to increase the molecular recognition specificity. In addition, when used in complex environment such as the biomatrix, the MOF shells provide a physical defense to improve the stability and reproducibility of the substrates, as well as to reduce the nonspecific adsorption and, thus, improving the detection sensitivity.

#### **4. Conclusions and Future Perspectives**

Applications of MOFs in the analytical and bioanalytical fields have experienced rapid growth due to their unique structural features. This mini-review summarized the advances of MOF-based optical detection methods, including luminescence and SERS, from the following aspects: the development of MOF-based luminophores, including the single luminophore signal, ratiometric signal and multi-modality signals; and the SERS effect of including MOFs as enhancement substrates and as auxiliary moieties for target molecule concentration, selective separation, and SERS substrate homogeneity for the purpose of improving the method's robustness.

Compared to detection based on single parameter, the multiplexed detection with which multiple target or parameter detection is achieved in one sample volume can be more informational and can help to draw solid conclusions in the analysis of biological samples. Optical-based analytical methods have been used in the field of multiplexed detection due to their non-invasiveness, excellent spatiotemporal resolution, and, most importantly, their multiple coding elements, including intensity, wavelength, lifetime, location, and combinations of the above. Luminescent MOFs should be developed as an excellent type of multiplexing probe, because the broad choice of guest molecules and the structural diversity of MOFs provide diversified coding elements. However, the multiplexing capability of photoluminescent MOFs has been less frequently studied, if at all. Therefore, more efforts should be contributed to designing luminescent MOFs with multiple signal sources to facilitate the necessary analytical and bioanalytical applications.

Lanthanide ion-based MOFs, including mixed Ln ions, exhibit tunable luminescence peaks and lifetimes, making them suitable for ratiometric, multiplexed, and multi-modal measurements. However, as luminescent nanoprobes, the modest luminescence quantum yield in aqueous media impedes the application of these MOF-based luminescence sensors in aqueous solution as well as in biological samples. Future efforts should also include the engineering of the building elements of MOF structures to create more MOF-based optical nanosensors with improved performance in terms of factors such as physical and chemical stability, photostability, easiness in functionalization, quantum yield, red/near infrared emission wavelength, and tuned luminescence lifetime.

For the applications of MOFs in SERS measurements, the SERS mechanism of MOFs needs to be explored in more depth and breadth to rationally achieve the maximum SERS sensitivity. Further novel and facile approaches are expected to produce distinctive Raman signals via chemistry of the MOF with the analyte of interest. Efforts towards reproducible MOF substrates with high enhancement factors will be crucial to applications of MOF-based SERS in practical samples.

In both the photoluminescence and SERS fields, the breadth of applications need to be further explored in order to most effectively utilize the excellent physical and chemical properties of MOFs. The aforementioned needs represent challenges, but they also represent opportunities for MOF-based optical nanosensors to play a more significant role in bioanalytical applications.

**Author Contributions:** Conceptualization, writing and editing, N.L.; discussion, writing and editing, C.W.; discussion and editing, R.L.; discussion and editing, X.C. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by National Natural Science Foundation of China grant number [21974006] and National Natural Science Foundation of China grant number [22134005].

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** Not applicable.

**Conflicts of Interest:** The authors declare no conflict of interest.
