Rapid Formation of Nanoclusters for Detection of Drugs in Urine Using Surface-Enhanced Raman Spectroscopy

nanomaterials

Article Rapid Formation of Nanoclusters for Detection of Drugs in Urine Using Surface-Enhanced Raman Spectroscopy

Yun-Chu Chen 1, Shang-Wen Hong 1, Huang-Hesin Wu 1, Yuh-Lin Wang 2 and Yih-Fan Chen 1,*

1 Institute of Biophotonics, National Yang Ming Chiao Tung University, Taipei 112, Taiwan; [email protected] (Y.-C.C.); [email protected] (S.-W.H.); [email protected] (H.-H.W.) 2 Institute of Atomic and Molecular Sciences, Academia Sinica, Taipei 106, Taiwan; [email protected] * Correspondence: [email protected]

Abstract: We developed a method based on surface-enhanced Raman spectroscopy (SERS) and a sample pretreatment process for rapid, sensitive, reproducible, multiplexed, and low-cost detection of illegal drugs in urine. The abuse of new psychoactive substances (NPS) has become an increasingly serious problem in many countries. However, immunoassay-based screening kits for NPS are usually not available because of the lack of corresponding antibodies. SERS has a great potential for rapid detection of NPS because it can simultaneously detect multiple kinds of drugs without the use of antibodies. To achieve highly sensitive SERS detection of drugs, sodium bromide was ﬁrst employed to induce the rapid formation of Ag nanoclusters by aggregating silver nanoparticles (AgNPs) in the extracted sample solution. SERS measurements were performed immediately after the sample

pretreatment without incubation. The three-dimensional SERS hot spots were believed to form signiﬁcantly within the nanoclusters, providing strong SERS enhancement effects. The displacement

Citation: Chen, Y.-C.; Hong, S.-W.; of citrate molecules on the surfaces of the AgNPs by bromide ions helped increase the adsorption Wu, H.-H.; Wang, Y.-L.; Chen, Y.-F. of drug molecules, increasing their areal density. We demonstrated the simultaneous detection of Rapid Formation of Nanoclusters for two kinds of NPS, methcathinone and 4-methylmethcathinone, in urine at a concentration as low as Detection of Drugs in Urine Using 0.01 ppm. Surface-Enhanced Raman Spectroscopy. Nanomaterials 2021, 11, 1789. https:// Keywords: surface-enhanced Raman spectroscopy (SERS); drug; new psychoactive substance; doi.org/10.3390/nano11071789 urine; nanocluster

Academic Editors: Mohsen Rahmani, Cuifeng Ying and Lei Xu 1. Introduction Received: 9 June 2021 Drug abuse has become an increasingly serious problem in many countries. The use of Accepted: 7 July 2021 Published: 9 July 2021 either traditional illegal drugs such as cocaine and heroin or new psychoactive substances (NPS) such as methcathinone (MC) and 4-methylmethcathinone (4-MMC, mephedrone)

Publisher’s Note: MDPI stays neutral poses a serious threat to health and public security. To detect the use of illegal drugs during with regard to jurisdictional claims in the preceding 1–4 days, urine samples are usually collected and analyzed using GC-MS or published maps and institutional affil- LC-MS/MS [1–3]. These methods can provide sensitive and specific detection of drugs; iations. however, they need to be performed by well-trained personnel in a laboratory and require delicate sample pretreatment processes and expensive reagents. Therefore, for the rapid, on-site detection of illegal drugs, lateral flow immunoassays are usually used instead [4,5]. These assays are low-cost and easy-to-use, but they usually suffer from a higher false- positive rate [4]. Rapid screening of NPS is especially difficult because immunoassay test Copyright: © 2021 by the authors. Licensee MDPI, Basel, Switzerland. kits are usually not available due to the lack of corresponding antibodies and aptamers. This article is an open access article Therefore, to prevent the spread of NPS, the development of a rapid, simple, sensitive, and distributed under the terms and low-cost method that can detect NPS in urine without the need of antibodies or aptamers conditions of the Creative Commons is urgent. Attribution (CC BY) license (https:// Surface-enhanced Raman spectroscopy (SERS) has been used to detect a variety of creativecommons.org/licenses/by/ molecules [6–10] because SERS-based methods only need a small sample and take a few sec- 4.0/). onds to achieve the multiplexed detection of molecules with high sensitivity. Many studies

Nanomaterials 2021, 11, 1789. https://doi.org/10.3390/nano11071789 https://www.mdpi.com/journal/nanomaterials Nanomaterials 2021, 11, 1789 2 of 14

have reported the SERS detection of drugs, including cannabinoid [11], cocaine [12–17], flu- nitrazepam [18], alpha-methyltryptamine [19], heroin [17,20], ketamine [21], morphine [13,22], benzodiazepine [23], methamphetamine [13,24–30], amphetamine [31], opioid [32], 3,4- methylenedioxymethamphetamine (MDMA) [25,27,28,33], 4-MMC [34,35], MC [25], bar- biturate [36], and sulfa drugs [37]. While the plasmonic properties of nanostructured substrates or nanoparticles used for SERS measurements is critical for the sensitive detection of molecules, it has been demonstrated that sample pretreatment is also very important for the SERS detection of drugs in biological samples such as urine [14,25,27,29,30,36] and serum [30]. It should be noted that, although antibodies or aptamers are sometimes used in SERS-based detection methods [12], they are not required in all SERS detection methods, which makes it useful for detecting NPS. Three-dimensional (3D) SERS substrates have drawn much attention in recent years [38]. Given that the laser focal volume is a 3D space, SERS substrates with SERS hot spots in all spatial directions can utilize the laser focal volume better than a 2D SERS substrate. In addition, having SERS hot spots in the z-axis allows for a higher tolerance for the position of the laser focal plane. Recent studies showed that 3D SERS substrates could be fabricated using nano-fabrication processes [39] or formed by metal colloidal nanoparticles [19,27,40–43]. For example, several studies demonstrated that a 3D SERS hot spot matrix could be formed by evaporating a droplet of metal nanoparticles to provide highly sensitive detection of molecules [19,40–43]. As the interparticle distance, which was critical to the SERS effect, gradually decreased during the evaporation process in those studies, the interparticle distance could decrease to an optimal value at a certain time point and resulted in strong SERS effects. On the other hand, instead of using the evaporation process, inducing the aggregation of nanoparticles was another strategy to generate 3D SERS hot spots [32,44]. The sensitive and quantitative SERS detection of molecules was demonstrated by adding a suitable kind and concentration of aggregating agent, such as an alkali halide salt [32,45], to colloidal metal nanoparticles to form nanoclusters with many SERS-active nanogaps. In this study, we developed a sensitive and reproducible SERS method for the detection of drugs in urine. A simple sample pretreatment process was developed to reduce interferences from other components in urine on the detection results. An SERS measurement was performed immediately after sodium bromide (NaBr), a kind of alkali halide salt, was added to a mixture of the pretreated sample and a silver nanoparticle (AgNP) colloidal solution. By using an alkali halide salt as an aggregating agent to induce the rapid formation of Ag nanoclusters with SERS hot spots in three dimensions, the sensitive and reproducible detection of drugs was achieved without incubation time and a nanofabricated SERS substrate. We demonstrated individual and simultaneous detection of two kinds of NPS, MC and 4-MMC, in urine at a concentration as low as 0.01 ppm.

2. Materials and Methods 2.1. Materials S(-)-methcathinone hydrochloride solution, 4-methylmethcathinone (mephedrone) hydrochloride solution, n-hexane, sodium chloride (NaCl), potassium bromide (KBr), and sodium borohydride (NaBH4) were purchased from Sigma-Aldrich. Sodium bromide (NaBr) and silver nitrate (AgNO3) were purchased from J.T. Baker. Trisodium citrate dihydrate and sodium hydroxide (NaOH) were purchased from Showa Chemical Industry. Ultrapure water was produced by a Millipore water system.

2.2. Synthesis of AgNPs The glassware used for the synthesis of AgNPs was cleaned with aqua regia and was then rinsed with ultrapure water before use. To synthesize 4-nanometer AgNP seeds, 47.5 mL of 0.2% trisodium citrate, 0.85 mL of 60 mM AgNO3, and 1 mL of 0.1% NaBH4 were mixed at 70 ◦C in a reﬂux setup for 60 min. Next, to increase the size of the AgNPs, 5 mL of the synthesized 4-nanometer AgNP seeds, 1 mL of 1% trisodium citrate, 37.5 mL of ultrapure water, and 0.85 mL of 60 mM AgNO3 were mixed and boiled in a reﬂux Nanomaterials 2021, 11, x FOR PEER REVIEW 3 of 15

Nanomaterials 2021, 11, 1789 mixed at 70 °C in a reflux setup for 60 min. Next, to increase the size of the AgNPs,3 5 of mL 14 of the synthesized 4-nanometer AgNP seeds, 1 mL of 1% trisodium citrate, 37.5 mL of ultrapure water, and 0.85 mL of 60 mM AgNO3 were mixed and boiled in a reflux setup for 60 min. Then, 1 mL of 1% trisodium citrate and 0.85 mL of 60 mM AgNO3 were added setupto the forboiled 60 min. mixture, Then, and 1 mLthe ofmixture 1% trisodium was boiled citrate for andanother 0.85 60 mL min. of 60The mM addition AgNO of3 were added to the boiled mixture, and the mixture was boiled for another 60 min. The trisodium citrate and AgNO3 and the boiling process were repeated two more times. Fi- additionnally, the of mixture trisodium was citrate cooled and to AgNO room 3tempand theerature, boiling and process the AgNPs were repeatedwere ready two to more use. times.Transmission Finally, electron the mixture microscopy was cooled (TEM) to roomimages temperature, showed that and the thediameter AgNPs of were the AgNPs ready towas use. ~28 Transmission nm. (Supplementary electron microscopy material Figure (TEM) S1a). images The showedlocalized that surface the diameter plasmon of reso- the AgNPs was ~28 nm. (Supplementary material Figure S1a). The localized surface plasmon nance (LSPR) wavelengths of the 4-nanometer AgNP seeds and the 28-nanometer AgNPs resonance (LSPR) wavelengths of the 4-nanometer AgNP seeds and the 28-nanometer were 391.9 nm and 395.1 nm, respectively (Supplementary material Figure S1b). AgNPs were 391.9 nm and 395.1 nm, respectively (Supplementary material Figure S1b).

2.3.2.3. UrineUrine SampleSample PretreatmentPretreatment TheThe samplesample pretreatmentpretreatment process,process, whichwhich waswas modiﬁedmodified fromfrom aa processprocess reportedreported inin thethe literatureliterature [[25],25], isis illustratedillustrated inin SchemeScheme1 1.. Brieﬂy, Briefly, human human urine urine (1 (1 mL) mL) was was mixed mixed with with 0.680.68 gg ofof NaClNaCl andand 200200 µμLL ofof 5%5% NaOHNaOH forfor 3030 s.s. Then,Then, n-hexanen-hexane (200(200 µμL)L) waswas addedadded toto thethe mixture,mixture, andand thethe mixturemixture waswas vigorouslyvigorously mixedmixed byby aa vortexvortex mixermixer forfor 3030 s.s. AfterAfter thethe mixing,mixing, thethe mixturemixture waswas centrifugedcentrifuged atat 95009500 gg forfor 22 min,min, andand thenthen thethe upperupper organicorganic phasephase waswas transferredtransferred toto anotheranother centrifugecentrifuge tube.tube. Finally,Finally, thethe collectedcollected n-hexanen-hexane solutionsolution waswas drieddried atat 7070 ◦°C,C, andand ultrapureultrapure waterwater (10(10 µμL)L) waswas addedadded toto thethe centrifugecentrifuge tubetube toto resuspendresuspend thethe extractedextracted drugsdrugs forfor SERSSERS measurements.measurements.

SchemeScheme 1.1.Illustration Illustration ofof thethe urineurine samplesample pretreatmentpretreatment and and thethe formationformation ofof AgAg nanoclustersnanoclusters forfor SERSSERS detectiondetection ofof drugsdrugs inin urine.urine. AA humanhuman urineurine samplesample isis mixedmixed withwith NaCl,NaCl, NaOH,NaOH, andand n-hexanen-hexane forfor drug drug extraction. extraction. NaBrNaBr waswas addedadded toto thethe extractedextracted solution solution to to induce induce rapid rapid formation formation of of Ag Ag nanoclustersnanoclusters and and to toincrease increase the the adsorption adsorption of of molecules molecules on on the the AgNPsAgNPs forfor SERS SERS detection detection of of drugs. drugs. A A droplet droplet of of the the mixture mixture of AgNPs,of AgNPs, NaBr, NaBr, and and the extractedthe extracted sample sample solution solution was droppedwas dropped onto aonto silicon a silicon wafer wafer for SERS for measurements.SERS measurements.

2.4. SERS Measurement 2.4. SERS Measurement To induce the rapid formation of Ag nanoclusters for SERS measurements, 0.5 μL of 250 mMTo induce NaBr thewas rapid mixed formation with 5 μ ofL Agof the nanoclusters synthesized for 28 SERS nm measurements,AgNPs and 5 μ 0.5L µofL the of 250sample mM solution NaBr was (standard mixed with solution 5 µL of or the extrac synthesizedted urine 28 sample). nm AgNPs The and final 5 µ concentrationL of the sample of solutionNaBr in (standardthe mixture solution was ~12 or mM. extracted After urinethe mixing, sample). 5 μL The of the ﬁnal mixture concentration was dropped of NaBr onto in thea silicon mixture wafer was for ~12 measurements. mM. After the A mixing, 532-na 5nometer-RamanµL of the mixture system was dropped (B&W TEK, onto i-Raman- a silicon wafer532S) forwas measurements. used to measure A the 532-nanometer-Raman SERS spectra. The laser system from (B&W the Raman TEK, i-Raman-532S)probe was directed was usedinto a to side measure port of the an SERS upright spectra. microscope The laser (Olympus, from the RamanBX51) and probe was was focused directed by intoa 20X a sideobjective port oflens an (N.A. upright 0.7) microscope on a droplet (Olympus, of the mixture BX51) during and was SERS focused measurements. by a 20X objective lens (N.A. 0.7) on a droplet of the mixture during SERS measurements.

Nanomaterials 2021, 11, 1789 4 of 14

3. Results and Discussion 3.1. Formation of SERS-Active Nanoclusters The sample pretreatment and measurement processes for the SERS detection of drugs in urine are shown in Scheme1. To extract drugs from urine, hexane and a large amount of NaCl were added to the samples first. NaCl was used to reduce the solubility of the drugs, and hexane was used to extract the drugs from the urine. After shaking the mixture of urine and hexane vigorously for 30 s, the upper phase containing hexane and the extracted drugs was transferred to another microcentrifuge tube and was then dried by heating. Finally, a small amount of water was added to the microcentrifuge tube to obtain a concentrated, resuspend drug solution for measurement. The whole sample pretreatment process could be finished within 20 min. To detect drugs using SERS, an alkali halide salt, which functioned as an aggregating agent, and AgNPs were added to a sample solution before the SERS measurements. It has been known that alkali halide ions can replace citrate molecules on the surfaces of AgNPs and reduce the stability of colloidal nanoparticles [32]. Depending on the kind and the concentration of the alkali halide salt mixed with the AgNPs, the AgNPs may form many small nanoclusters or large aggregates. However, for sensitive SERS detection of molecules, small nanoclusters are generally preferred over large aggregates because the molecules of interest can access the nanogaps among nanoparticles more easily. To find a suitable kind and concentration of alkali halide salt to induce the formation of nanoclusters for SERS detection, we measured 1 ppm MC using three kinds of alkali halide salts, NaCl, NaBr, and KBr, with different concentrations, as shown in Figure1a,c,e. The range of the concentrations tested was different for the three kinds of alkali halide salts because they had different effects on the stability of the AgNPs. For each SERS measurement, we quickly mixed the drug solution, the 28-nanometer AgNPs, and one of the kinds of alkali halide salt, and then put a droplet of the mixture on a silicon wafer. The SERS spectra were measured immediately after the mixing without incubation. The droplet was not covered by a coverslip, and the 532-nanometer laser of the Raman system was focused on the droplet for measuring the SERS spectra. Figure1b,d,f show that the intensity of the characteristic Raman peak of MC at 1000 cm−1 varied with the kind and the concentration of the alkali halide salt. Among the nine experimental conditions tested, we found that the intensity of the 1000 cm−1 peak was highest when 12 mM NaBr was mixed with the AgNPs and the sample solution. Therefore, we used this amount of NaBr to induce the formation of Ag nanoclusters for the other experiments of this study. The rapid formation of nanoclusters in a sample solution allowed us to detect molecules using SERS without incubation time. According to the extinction spectra of the AgNPs, the LSPR peak of the AgNP solution increased from 395.1 to 399.7 nm right after the addition of NaBr but then remained almost unchanged afterwards (Supplementary material Figure S1c,d). The dynamic light scattering (DLS) measurements also showed that the addition of NaBr caused an increase in the hydrodynamic diameter of the AgNPs (Supplementary material Figure S2a,b). Both the red shift in the LSPR peak and the increase in the hydrodynamic diameter indicated the formation of nanoclusters. A time-lapse observation of the SERS spectra of 1 ppm MC after the addition of 12 mM NaBr showed a slight increase in the Raman signal over time (Supplementary material Figure S3), which was likely caused by the evaporation-induced concentration of molecules. As the rate of evaporation depends on several factors, such as the humidity and temperature of the surrounding environment, to avoid the effect of evaporation on the detection, we performed the SERS measurements immediately after we put a droplet of the sample on a silicon wafer. It should be noted that mixing the AgNPs with NaBr also helped to increase the adsorption of the drug molecules to the AgNPs because the citrate molecules capped on the surfaces of the AgNPs were displaced with bromide ions [32]. The Raman signals of the drug molecules adsorbed on the AgNPs could be significantly enhanced because of the SERS effects of the nanoclusters. On the other hand, the intensity of the characteristic Raman peak of citrate molecules at ~1400 cm−1 was not very strong in the measured Nanomaterials 2021, 11, 1789 5 of 14

Nanomaterials 2021, 11, x FOR PEER spectra.REVIEW It was most likely because the majority of the citrate molecules on the5 surfaces of 15 of the AgNPs were displaced by bromide ions during the displacement and aggregation processes that led to the formation of SERS hot spots.

Figure 1. Effects of the kind and the concentration of alkali halide salt on SERS detection of drugs. (a,c,e) SERS spectra Figure 1. Effects of the kind and the concentration of alkali halide salt on SERS detection of drugs. (a,c,e) SERS spectra of of1 1 ppm ppm MC MC standard standard solution solution measured measured using using the the nanoclusters nanoclusters induced induced by by(a) ( aNaBr,) NaBr, (b) ( bKBr,) KBr, and and (c) ( cNaCl) NaCl of ofdifferent different −1 concentrations.concentrations. TheThe characteristiccharacteristic Raman Raman peak peak of of MC MC at at1000 1000 cm cm−1 is markedis marked with with“*”. (b “*”.,d,f) (Theb,d ,fintensities) The intensities of the char- of the characteristicacteristic Raman Raman peak peak of MC of MCat 1000 at 1000 cm−1 cm measured−1 measured using usingthe nanostructures the nanostructures induced induced by (b) NaBr, by (b ()d NaBr,) KBr, (andd) KBr, (f) NaCl and (f) NaClof different of different concentrations. concentrations. Error Error bars show bars showstandard standard deviations deviations with n with = 30. n = 30.

Nanomaterials 2021, 11, 1789 6 of 14

3.2. Detection of MC and 4-MMC in Water After ﬁnding the suitable kind and concentration of the aggregating agent, we used the optimal experimental condition to measure MC and 4-MMC standard solutions of different concentrations. Figure2a shows the SERS spectra of MC with a concentration between 0.01 and 1 ppm, and Figure2b shows the relationship between the concentration of MC and the intensity of the characteristic Raman peak at 1000 cm−1. Similarly, Figure2c shows the SERS spectra of 4-MMC with a concentration between 0.01 and 1 ppm, and Figure2d shows the relationships between the concentration of 4-MMC and the intensities Nanomaterials 2021, 11, x FOR PEER REVIEW − 7 of 15 of the three characteristic Raman peaks at 799, 970, 1212 cm 1, respectively. Our results showed that MC and 4-MMC could be detected within 1 min with high sensitivity using the SERS method developed in this study.

Figure 2. SERS detection of MC and 4-MMC in water. (a,c) SERS spectra of (a) MC and (c) 4-MMC in water with Figure 2. SERS detection of MC and 4-MMC in water. (a,c) SERS spectra of (a) MC and (c) 4-MMC−1 in water with concen- concentrationstrations ranging ranging from from 0.01 0.01to 1 ppm. to 1 ppm. The ch Thearacteristic characteristic Raman Raman peak of peak MC at of 1000 MC atcm 1000−1 and cm the characteristicand the characteristic Raman peak Raman −1 −1 peakof of 4-MMC 4-MMC at at1212 1212 cm cm−1 are markedare marked with with“*” and “*” “+”, and respectively. “+”, respectively. (b) A plot (b )of A the plot intensity of the intensityof the 1000 of cm the−1 peak 1000 versus cm peak versusMC MC concentration. concentration. Error Error bars barsshow show standa standardrd deviations deviations with n with = 30.n The = 30. straight The straight line is the line linear is the regression linear regression line, and line,the and σ σ −1 the dotteddotted line indicates mean mean plus plus 3 3standard standard deviations deviations (mean (mean + 3 +, 3 σis, theσ is standard the standard deviation) deviation) of the 1000 of the cm 1000 peak cm of−1 thepeak of blank sample. (d) A plot of the intensities of the 796, the 970, and the 1212 cm−1 peaks versus 4-MMC concentration. Error the blank sample. (d) A plot of the intensities of the 796, the 970, and the 1212 cm−1 peaks versus 4-MMC concentration. bars show standard deviations with n = 30. The straight lines are the linear regression lines, and the blue, gray, and red Errordotted bars showlines indicate standard mean deviations plus 3σ of with the 796, n = 30.the The970, and straight the 1212 lines cm are−1 peaks, the linear respectively, regression of the lines, blank and sample. the blue, gray, and red dotted lines indicate mean plus 3σ of the 796, the 970, and the 1212 cm−1 peaks, respectively, of the blank sample. 3.3. Simultaneous Detection of MC and 4-MMC in Water 3.3. SimultaneousIn addition to Detection demonstrating of MCand the 4-MMCSERS detection in Water of each of the two kinds of drugs, we Inalso addition tested the to simultaneous demonstrating detection the SERS of both detection MC and of 4-MMC. each of Compared the two kinds with ofim- drugs, wemunoassay also tested methods, the simultaneous simultaneous detection detection of of bothmultiple MC kinds and 4-MMC.of drugs Comparedis an advantage with im- munoassayfor SERS-based methods, method simultaneous because it does detection not require of multiple multiple kinds kinds of of drugs antibodies is an for advantage detec- for tion. Multiple drugs can be detected simultaneously using SERS as long as their characteristic Raman peaks do not fully overlap with each other. To test the simultaneous detection of MC and 4-MMC, we measured the SERS spectra of three kinds of drug mixtures, including 5 ppm 4-MMC plus 1 ppm MC, 1 ppm 4-MMC plus 5 ppm MC, and 1 ppm 4- MMC plus 1 ppm MC. We also measured 1 ppm MC, 5 ppm MC, 1 ppm 4-MMC, and 5 ppm 4-MMC for comparison. The results, as shown in Figure 3, revealed that the SERS spectra of the mixtures had the characteristic Raman peaks of both drugs. It should be noted that, because the adsorption of one kind of molecule to the AgNPs could interfere

Nanomaterials 2021, 11, 1789 7 of 14

SERS-based method because it does not require multiple kinds of antibodies for detection. Multiple drugs can be detected simultaneously using SERS as long as their characteristic Raman peaks do not fully overlap with each other. To test the simultaneous detection of MC and 4-MMC, we measured the SERS spectra of three kinds of drug mixtures, including 5 ppm 4-MMC plus 1 ppm MC, 1 ppm 4-MMC plus 5 ppm MC, and 1 ppm 4-MMC plus 1 Nanomaterials 2021, 11, x FOR PEER REVIEW 8 of 15 ppm MC. We also measured 1 ppm MC, 5 ppm MC, 1 ppm 4-MMC, and 5 ppm 4-MMC for comparison. The results, as shown in Figure3, revealed that the SERS spectra of the mixtures had the characteristic Raman peaks of both drugs. It should be noted that, because with the adsorption of of other one kindmolecules, of molecule the SERS to thespectra AgNPs measured could interferefor the mixtures with the of adsorption of MC andother 4-MMC molecules, showed the weaker SERS characteristic spectra measured Raman forpeaks the than mixtures the SERS of spectra MC and of 4-MMCin- showed dividualweaker drugs characteristic of the same concentration. Raman peaks In addition, than the our SERS results spectra showed of individual that the 4-MMC drugs of the same moleculeconcentration. appeared to have In addition, stronger adsorpti our resultson to showed the AgNPs that than the the 4-MMC MC molecule molecule be- appeared to cause the intensities of its characteristic Raman peaks decreased less during the simulta- have stronger adsorption to the AgNPs than the MC molecule because the intensities of its neous detection of 4-MMC and MC. To determine the limits of detection during simulta- neouscharacteristic detection, we Ramanperformed peaks SERS decreased measurements less duringof 4-MMC the and simultaneous MC of various detection con- of 4-MMC centrations.and MC. Figure To determine 4a shows the SERS limits spec of detectiontra of the duringmixtures simultaneous of MC and 4-MMC, detection, and we performed FigureSERS 4b shows measurements the relationships of 4-MMC between and the MC concentration of various of concentrations. each drug in the mixture Figure4 a shows the and theSERS intensities spectra of of the the characteristic mixtures of Raman MC and peaks. 4-MMC, The results and Figure showed4b showsthat, although the relationships be- the intensitiestween the of concentration the characteristic of eachRaman drug peaks in theof MC mixture and 4-MMC and the were intensities weaker ofwhen the characteristic the twoRaman drugs peaks. were detected The results simultaneously, showed that, we althoughcould still thedetect intensities both drugs of theat a characteristicconcen- Raman trationpeaks as low of as MC 0.01 and ppm 4-MMC in the mixture. were weaker when the two drugs were detected simultaneously, we could still detect both drugs at a concentration as low as 0.01 ppm in the mixture.

FigureFigure 3. SERS 3. detectionSERS detection of MC, 4-MMC, of MC, 4-MMC,and mixtures and of mixtures MC and of 4-MMC MC and in water. 4-MMC SERS in water.spectra SERS spectra of 1 of 1 ppmppm MC, MC, 5 ppm 5 ppm MC, MC, 1 ppm 1 ppm 4-MMC, 4-MMC, 5 ppm 5 ppm4-MMC, 4-MMC, 5 ppm 54-MMC ppm 4-MMC plus 1 ppm plus MC, 1 ppm 1 ppm MC, 1 ppm 4-MMC 4-MMCplus plus 5 ppm5 ppm MC, MC, and and 1 1 ppm ppm 4-MMC 4-MMC plusplus 11 ppmppm MC. The characteristic characteristic Raman Raman peak peak of of MC at 1000 cm−1 −1 −1 MC at 1000 cm and the characteristic Raman peak of 4-MMC at 1212−1 cm are marked with “*” and “+”,and respectively. the characteristic Raman peak of 4-MMC at 1212 cm are marked with “*” and “+”, respectively.

NanomaterialsNanomaterials 20212021, 11, 11, x, FOR 1789 PEER REVIEW 8 of9 14of 15

FigureFigure 4. 4. SimultaneousSimultaneous SERS SERS detection detection of of MC and 4-MMC inin water.water. ((aa)) SERSSERS spectra spectra of of mixtures mixtures ofof MC MC and and 4-MMC. 4-MMC. The The concentration concentration of of both both MC MC and and 4-MMC 4-MMC was was 5, 5, 0.5, 0.5, 0.05, 0.05, 0.01, 0.01, and 0 ppm for thefor SERS the SERS spectrum spectrum (1), (2), (1), (3), (2), (4), (3), and (4), and(5), (5),respectively. respectively. The The characterist characteristicic Raman Raman peak peak of ofMC MC at − − 1000at 1000 cm− cm1 and1 andthe characteristic the characteristic Raman Raman peak peak of 4-MMC of 4-MMC at 1212 at 1212 cm cm−1 are1 aremarked marked with with “*” “*” and and “+”, respectively.“+”, respectively. (b) A ( bplot) A plotof the of intensity the intensity of the of the1000 1000 cm cm−1 peak−1 peak versus versus MC MC concentration concentration and and the the intensityintensity of of the the 1212 1212 cm cm−1 1peakpeak versus versus 4-MMC 4-MMC concentration. ErrorError bars bars show show standard standard deviations deviations withwith n n= =30. 30. The The straight straight lines lines are are the the linear linear re regressiongression lines, lines, and and the the black black and and red red dotted dotted lines lines indicateindicate mean mean plus plus 3 3σσ ofof the 1000 andand thethe1212 1212 cm cm−−11 peak, respectively,respectively, of of the the blank blank sample. sample.

3.4. Urine Sample Pretreatment While we already demonstrated the sensitive detection of MC and 4-MMC standard solutions, to detect drugs in human urine samples, we had to use a suitable sample

Nanomaterials 2021, 11, 1789 9 of 14

Nanomaterials 2021, 11, x FOR PEER REVIEW 10 of 15

3.4. Urine Sample Pretreatment

pretreatmentWhile we method already demonstratedbefore the SERS the measur sensitiveements detection to reduce of MC the and interferences 4-MMC standard from solutions,other components to detect in drugsurine. inAs humanshown in urine Figure samples, 5a, when we using had our to use SERS a suitablemethod to sample meas- pretreatment method before the SERS measurements to reduce the interferences from other ure urine samples containing 0.1 ppm MC, which was already 10 times higher than the components in urine. As shown in Figure5a, when using our SERS method to measure lowest detected concentration shown in Figure 2a, we could not detect the 1000 cm−1 Ra- urine samples containing 0.1 ppm MC, which was already 10 times higher than the lowest man peak of MC unless the urine samples were processed using the sample pretreatment detected concentration shown in Figure2a, we could not detect the 1000 cm −1 Raman peak method illustrated in Scheme 1. The results confirmed that using the sample pretreatment of MC unless the urine samples were processed using the sample pretreatment method process to extract and concentrate drug molecules was essential for achieving sensitive illustrated in Scheme1. The results conﬁrmed that using the sample pretreatment process detection of drugs in urine samples. Figure 5b shows that, when the time duration for to extract and concentrate drug molecules was essential for achieving sensitive detection of mixing hexane and the urine samples containing 0.1 ppm MC increased from 30 s to 3 min drugs in urine samples. Figure5b shows that, when the time duration for mixing hexane and 5 min, the intensity of the characteristic Raman peak at 1000 cm−1 did not significantly and the urine samples containing 0.1 ppm MC increased from 30 s to 3 min and 5 min, increase with the mixing time. Therefore, to reduce the time of the sample pretreatment the intensity of the characteristic Raman peak at 1000 cm−1 did not signiﬁcantly increase as much as possible, we only mixed hexane with urine samples for 30 s for latter experi- with the mixing time. Therefore, to reduce the time of the sample pretreatment as much as possible,ments. we only mixed hexane with urine samples for 30 s for latter experiments.

Figure 5. Effects of the sample pretreatment process on detection of drugs in urine. (a) SERS spectra of 0.1 ppm MC in urine measuredFigure 5. (1)Effects after of and the (2) sample before pretreatment sample pretreatment process andon detection SERS spectra of drugs of blank in urine. urine ( samplesa) SERS measuredspectra of (3)0.1 afterppm and MC (4) in urine measured (1) after and (2) before sample pretreatment and SERS spectra of blank urine samples measured (3) after before sample pretreatment. (b) SERS detection of 0.1 ppm MC in urine with different extraction time. The characteristic and (4) before sample pretreatment. (b) SERS detection of 0.1 ppm MC in urine with different extraction time. The char- Raman peak of MC at 1000 cm−1 is marked with “*”. acteristic Raman peak of MC at 1000 cm−1 is marked with “*”.

3.5.3.5. SimultaneousSimultaneous DetectionDetection ofof MCMC andand 4-MMC4-MMC inin UrineUrine WeWe furtherfurther testedtested thethe samplesample pretreatmentpretreatment methodmethod withwith standardstandard solutionssolutions andand urineurine samplessamples containingcontaining drugsdrugs ofof differentdifferent concentrations.concentrations. We ﬁrstfirst preparedprepared 1,1, 0.1,0.1, andand 0.010.01 ppmppm MCMC standardstandard solutionssolutions andand processedprocessed themthem usingusing thethe samplesample pretreatment method.method. AlthoughAlthough thesethese samplessamples werewere notnot urineurine samples,samples, wewe pretreatedpretreated themthem beforebefore measuringmeasuring theirtheir SERS spectra spectra to to understand understand how how the the dete detectionction results results would would be affected be affected by the by sam- the sampleple pretreatment pretreatment process. process. As Asshown shown in inFigure Figure 6a,6a, the the results results of the SERSSERS measurementsmeasurements revealedrevealed thatthat 0.01 0.01 ppm ppm MC MC could could still still be be detected detected after after the the sample sample pretreatment. pretreatment.Figure Figure6b shows6b shows that that the the intensity intensity of theof the 1000 1000 cm cm−1−1peak peak increased increased with withthe the concentrationconcentration ofof MC.MC. Then,Then, toto testtest the the detection detection of of drugs drugs in in urine urine samples, samples, we we prepared prepared urine urine samples samples containing contain- 1,ing 0.1, 1, and0.1, and 0.01 0.01 ppm ppm MC MC and and then then pretreated pretreated the samplesthe samples using using the samethe same method method before be- thefore SERS the SERS measurements. measurements. Figure Figure7a,b show 7a,b thatshow we that could we successfullycould successfully detect MCdetect in urineMC in samplesurine samples containing containing 0.01 ppm0.01 ppm MC orMC more. or more. The The relation relation between between the the intensity intensity of of thethe 10001000 cmcm−11 peakpeak and and MC MC concentration concentration shown shown in inFigu Figurere 7b7 wasb was slightly slightly different different from from that thatshown shown in Figure in Figure 6b because6b because they they were were measured measured with with urine urine samples samples and andwith with water water sam- samples,ples, respectively. respectively. To understa To understandnd the thereproducibility reproducibility of the of SERS the SERS measurements, measurements, we pre- we preparedpared five ﬁve urine urine samples samples containing containing 0.01 0.01ppm ppm MC and MC then and performed then performed 10 repeated 10 repeated measurements for each sample. Figure 7c shows the intensities of the 1000 cm−1 peak in the SERS spectra (Supplementary material Figure S4) of the five urine samples and the relative

Nanomaterials 2021, 11, 1789 10 of 14

Nanomaterials 2021, 11, x FOR PEER REVIEW 11 of 15

measurements for each sample. Figure7c shows the intensities of the 1000 cm −1 peak in thestandard SERS spectra deviation (Supplementary (RSD) of the results, material whic Figureh was S4)13.3%. of the Finally, ﬁve urinewe prepared samples urine and sam- the relativeples containing standard both deviation MC and (RSD) 4-MMC of the to results,test the whichsimultaneous was 13.3%. extraction Finally, and we detection prepared of urineboth samplesdrugs in containingurine. Figure both 7d MCshows and that 4-MMC the SERS to test spectra the simultaneousof the extracted extraction solutions and pre- detectionsented the of characteristic both drugs in Raman urine. Figurepeaks of7d both shows drugs, that including the SERS spectrathe 1000 of cm the−1 peak extracted of MC solutions presented the characteristic Raman peaks of both drugs, including the 1000 cm−1 and the 1212 cm−1 peak of 4-MMC. The concentrations of MC and 4-MMC in the tested peak of MC and the 1212 cm−1 peak of 4-MMC. The concentrations of MC and 4-MMC in urine samples were in the range of 0.01–5 ppm. In contrast, the SERS spectra of the blank the tested urine samples were in the range of 0.01–5 ppm. In contrast, the SERS spectra of sample, which was the urine sample without drug additives, did not present the charac- the blank sample, which was the urine sample without drug additives, did not present the teristic Raman peaks of the drugs. Figure 7e shows that the intensity of the 1000 and the characteristic Raman peaks of the drugs. Figure7e shows that the intensity of the 1000 and 1212 cm−1 peak increased with the concentration of MC and 4-MMC. the 1212 cm−1 peak increased with the concentration of MC and 4-MMC.

FigureFigure 6. 6.SERS SERS detection detection of of MC MC in in water water after after sample sample pretreatment. pretreatment. (a ()a SERS) SERS spectra spectra of of MC MC in in water water measured measured after after sample sample pretreatment.pretreatment. The The characteristic characteristic RamanRaman peakpeak ofof MC at 1000 cm −−1 1isis marked marked with with “*”. “*”. (b) ( bA) plot A plot of the of the intensity intensity of the of the1000 1000cm− cm1 peak−1 peak versus versus MC concentration. MC concentration. Error Error bars show bars show standard standard deviations deviations with withn = 30.n The= 30. straight The straight line isline the islinear the linearregres- σ −1 regressionsion line, line,and the and dotted the dotted line indicates line indicates mean mean plus plus3 of 3 σtheof 1000 the 1000 cm cm peak−1 peakof the of blank the blank sample. sample.

Our results show that we could simultaneously detect MC and 4-MMC in urine at a concentration as low as 0.01 ppm. In comparison, Muhamadali et al. demonstrated the detection of 4-MMC in water and in urine with limits of detection of ~800 ppm by mixing AgNPs with sample solutions [35]. No sample pretreatment method was used in their study. Han et al. reported SERS detection of 10 ppm MC and two other kinds of drugs in urine using self-assembled gold nanorod arrays, but they did not determine the limit of detection for MC [25]. Lee et al. showed that their polymer-stabilized Ag nanoparticle ﬁlms could be used as SERS substrates to detect 5.7 µg of MC using a swabbing method [34]. In addition to the SERS-based methods, an electrochemical method was used to detect 4-MMC with a limit of detection of 11.8 ppm [46]. In comparison with these results, the SERS method developed in this study was one of the most sensitive rapid screening methods reported in the literature for the detection of MC and 4-MMC in urine.

NanomaterialsNanomaterials2021 2021, 11, ,11 1789, x FOR PEER REVIEW 12 of11 15 of 14

Figure 7. SERS detection of MC and 4-MMC in urine. (a) SERS spectra of MC in urine measured after sample pretreatment. TheFigure characteristic 7. SERS detection Raman peakof MC of and MC 4-MMC at 1000 in cm urine.−1 is (a marked) SERS spectra with “*”. of MC (b) in A urine plot ofmeasured the intensity after sample of the pretreatment. 1000 cm−1 peak The characteristic Raman peak of MC at 1000 cm−1 is marked with “*”. (b) A plot of the intensity of the 1000 cm−1 peak versus MC concentration. Error bars show standard deviations with n = 30. The straight line is the linear regression line, versus MC concentration. Error bars show standard deviations with n = 30. The straight line is the linear regression line, −1 andand the the dotted dotted line line indicates indicates meanmean plus 3 3σσ ofof the the 1000 1000 cm cm−1 peakpeak of the of blank the blank sample. sample. (c) The (c intensities) The intensities of the 1000 of the cm 1000−1 −1 cmpeakpeak measured measured from from 5 urine 5 urine samples samples containing containing 0.01 ppm 0.01 MC ppm. Error MC. bars Error show bars standard show standard deviations deviations with n = with10. (d n) SERS = 10. (d) SERSspectra spectra of urine of urine samples samples containing containing both MC both and MC 4-MMC and 4-MMC measured measured after sample after pretreatment. sample pretreatment. The concentration The concentration of both of bothMC and MC 4-MMC and 4-MMC was 5, was 0.5, 5,0.05, 0.5, 0.01, 0.05, and 0.01, 0 ppm and for 0 ppm the SERS for the spec SERStrum spectrum (1), (2), (3), (1), (4), (2), and (3), (5), (4), respectively. and (5), respectively. The charac- The −1 −1 characteristicteristic Raman Raman peak peak of MC of MCat 1000 at 1000 cm cm and−1 theand characteristic the characteristic Raman Raman peak peakof 4-MMC of 4-MMC at 1212 at cm 1212 are cm −marked1 are marked with “*” with and “+”, respectively. (e) A plot of the intensity of the t 1000 cm−1 peak versus MC concentration and the intensity of the “*” and “+”, respectively. (e) A plot of the intensity of the t 1000 cm−1 peak versus MC concentration and the intensity of 1212 cm−1 peak versus 4-MMC concentration. Error bars show standard deviations with n = 10. The straight lines are the the 1212 cm−1 peak versus 4-MMC concentration. Error bars show standard deviations with n = 10. The straight lines are the linear regression lines, and the black and red dotted lines indicate mean plus 3σ of the 1000 and the 1212 cm−1 peak, respectively, of the blank sample.

Nanomaterials 2021, 11, 1789 12 of 14

4. Conclusions In conclusion, we have successfully developed a rapid, low-cost, and sensitive method for the detection of MC and 4-MMC in urine. To reduce the interferences from other components in urine on the detection results, we used a simple sample pretreatment process that could extract and concentrate drugs from urine within 20 min before SERS measurements. To achieve the sensitive detection of drugs, we optimized the kind and concentration of the alkali halide salt used to quickly form Ag nanoclusters with strong SERS effects. Our results showed that, by using NaBr as the aggregating agents for AgNPs, we could simultaneously detect both MC and 4-MMC in urine at a concentration as low as 0.01 ppm after sample pretreatment. The whole measurement process could be ﬁnished within 1 min because we measured the SERS spectra immediately after we added the AgNPs and NaBr to the extracted sample solution without incubation. We believe that the SERS-based detection method and the sample pretreatment process developed in this study have a great potential to be used for rapid screening of illegal drugs in urine.

Supplementary Materials: The following are available online at https://www.mdpi.com/article/ 10.3390/nano11071789/s1, Figure S1: Extinction spectra and TEM image of the AgNPs, Figure S2: Hydrodynamic diameter of the AgNPs determined using dynamic light scattering (DLS), Figure S3: Time-lapse observation of SERS spectra after the addition of NaBr. Figure S4: SERS spectra of 5 urine samples containing 0.01 ppm MC. Author Contributions: Conceptualization, Y.-C.C., and Y.-F.C.; methodology, Y.-F.C.; investigation and data curation, Y.-C.C., S.-W.H., and H.-H.W.; writing—original draft, Y.-C.C.; writing—review and editing, Y.-L.W., and Y.-F.C.; supervision, Y.-L.W., and Y.-F.C.; funding acquisition, Y.-L.W., and Y.-F.C. All authors have read and agreed to the published version of the manuscript. Funding: This work was supported by the Ministry of Science and Technology, Taiwan (MOST 107-2622-E-010-002-CC2, MOST 108-2628-E-010-001-MY3) and Taiwan Advance Bio-Pharmaceutical Inc. Data Availability Statement: The data presented in this study are available on request from the corresponding author. Conﬂicts of Interest: This work was ﬁnancially supported in part by Taiwan Advance Bio-Pharmaceutical Inc.

References 1. Feng, J.; Wang, L.; Dai, I.; Harmon, T.; Bernert, J.T. Simultaneous determination of multiple drugs of abuse and relevant metabolites in urine by LC-MS-MS. J. Anal. Toxicol. 2007, 31, 359–368. [CrossRef] 2. Wang, J.M.; Yang, Z.; Lechago, J. Rapid and simultaneous determination of multiple classes of abused drugs and metabolites in human urine by a robust LC-MS/MS method—Application to urine drug testing in pain clinics. Biomed. Chromatogr. 2013, 27, 1463–1480. [CrossRef] 3. Fiorentin, T.R.; Fogarty, M.; Limberger, R.P.; Logan, B.K. Determination of cutting agents in seized cocaine samples using GC-MS, GC-TMS and LC-MS/MS. Forensic Sci. Int. 2019, 295, 199–206. [CrossRef] 4. Saitman, A.; Park, H.D.; Fitzgerald, R.L. False-positive interferences of common urine drug screen immunoassays: A review. J. Anal. Toxicol. 2014, 38, 387–396. [CrossRef] 5. Angelini, D.J.; Biggs, T.D.; Prugh, A.M.; Smith, J.A.; Hanburger, J.A.; Llano, B.; Avelar, R.; Ellis, A.; Lusk, B.; Naanaa, A.; et al. Detection of fentanyl and derivatives using a lateral flow immunoassay. Forensic Chem. 2021, 23, 100309. [CrossRef] 6. Ruperez, A.; Montes, R.; Laserna, J.J. Identification of stimulant-drugs by surface-enhanced Raman-spectrometry on colloidal silver. Vib. Spectrosc 1991, 2, 145–154. [CrossRef] 7. Guerrini, L.; Garcia-Ramos, J.V.; Domingo, C.; Sanchez-Cortes, S. Sensing polycyclic aromatic hydrocarbons with dithiocarbamate- functionalized Ag nanoparticles by surface-enhanced Raman scattering. Anal. Chem. 2009, 81, 953–960. [CrossRef][PubMed] 8. Wu, L.-A.; Li, W.-E.; Lin, D.-Z.; Chen, Y.-F. Three-dimensional SERS substrates formed with plasmonic core-satellite nanostructures. Sci. Rep. 2017, 7, 13066. [CrossRef][PubMed] 9. Dugandzic, 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. [CrossRef] 10. D’Agostino, A.; Giovannozzi, A.M.; Mandrile, L.; Sacco, A.; Rossi, A.M.; Taglietti, A. In situ seed-growth synthesis of silver nanoplates on glass for the detection of food contaminants by surface enhanced Raman scattering. Talanta 2020, 216, 120936. [CrossRef] Nanomaterials 2021, 11, 1789 13 of 14

11. Deriu, C.; Conticello, I.; Mebel, A.M.; McCord, B. Micro solid phase extraction surface-enhanced Raman spectroscopy (µ- SPE/SERS) screening test for the detection of the synthetic cannabinoid JWH-018 in oral fluid. Anal. Chem. 2019, 91, 4780–4789. [CrossRef] 12. Li, Y.; Zeng, Y.; Mao, Y.; Lei, C.; Zhang, S. Proximity-dependent isothermal cycle amplification for small-molecule detection based on surface enhanced Raman scattering. Biosens. Bioelectron. 2014, 51, 304–309. [CrossRef] 13. Kline, N.D.; Tripathi, A.; Mirsafavi, R.; Pardoe, I.; Moskovits, M.; Meinhart, C.; Guicheteau, J.A.; Christesen, S.D.; Fountain, A.W. Optimization of surface-enhanced Raman spectroscopy conditions for implementation into a microfluidic device for drug detection. Anal. Chem. 2016, 88, 10513–10522. [CrossRef][PubMed] 14. Meng, J.; Tang, X.H.; Zhou, B.B.; Xie, Q.W.; Yang, L.B. Designing of ordered two-dimensional gold nanoparticles film for cocaine detection in human urine using surface-enhanced Raman spectroscopy. Talanta 2017, 164, 693–699. [CrossRef] 15. Kong, X.; Chong, X.; Squire, K.; Wang, A.X. Microfluidic diatomite analytical devices for illicit drug sensing with ppb-Level sensitivity. Sens. Actuators B-Chem. 2018, 259, 587–595. [CrossRef] 16. Ameku, W.A.; Gonçalves, J.M.; Ataide, V.N.; Ferreira Santos, M.S.; Gutz, I.G.R.; Araki, K.; Paixão, T.R.L.C. Combined colorimetric and electrochemical measurement paper-based device for chemometric proof-of-concept analysis of cocaine samples. ACS Omega 2021, 6, 594–605. [CrossRef][PubMed] 17. Masterson, A.N.; Hati, S.; Ren, G.; Liyanage, T.; Manicke, N.E.; Goodpaster, J.V.; Sardar, R. Enhancing nonfouling and sensitivity of surface-enhanced Raman scattering substrates for potent plasma via fabrication of a flexible plasmonic patch. Anal. Chem. 2021, 93, 2578–2588. [CrossRef][PubMed] 18. Segawa, H.; Fukuoka, T.; Itoh, T.; Imai, Y.; Iwata, Y.T.; Yamamuro, T.; Kuwayama, K.; Tsujikawa, K.; Kanamori, T.; Inoue, H. Rapid detection of hypnotics using surface-enhanced Raman scattering based on gold nanoparticle co-aggregation in a wet system. Analyst 2019, 144, 2158–2165. [CrossRef][PubMed] 19. Yan, X.; Li, P.; Zhou, B.; Tang, X.; Li, X.; Weng, S.; Yang, L.; Liu, J. Optimal hotspots of dynamic surfaced-enhanced Raman spectroscopy for drugs quantitative detection. Anal. Chem. 2017, 89, 4875–4881. [CrossRef][PubMed] 20. Dies, H.; Raveendran, J.; Escobedo, C.; Docoslis, A. Rapid identification and quantification of illicit drugs on nanodendritic surface-enhanced Raman scattering substrates. Sens. Actuators B-Chem. 2018, 257, 382–388. [CrossRef] 21. Yang, Y.; Li, Z.-Y.; Yamaguchi, K.; Tanemura, M.; Huang, Z.; Jiang, D.; Chen, Y.; Zhou, F.; Nogami, M. Controlled fabrication of silver nanoneedles array for SERS and their application in rapid detection of narcotics. Nanoscale 2012, 4, 2663–2669. [CrossRef] 22. Yu, B.; Cao, C.; Li, P.; Mao, M.; Xie, Q.; Yang, L. Sensitive and simple determination of zwitterionic morphine in human urine based on liquid-liquid micro-extraction coupled with surface-enhanced Raman spectroscopy. Talanta 2018, 186, 427–432. [CrossRef] 23. Wang, C.; Xu, X.; Qiu, G.; Ye, W.; Li, Y.; Harris, R.A.; Jiang, C. Group-targeting SERS screening of total benzodiazepines based on large-size (111) faceted silver nanosheets decorated with zinc oxide nanoparticles. Anal. Chem. 2021, 93, 3403–3410. [CrossRef] [PubMed] 24. Andreou, C.; Hoonejani, M.R.; Barmi, M.R.; Moskovits, M.; Meinhart, C.D. Rapid detection of drugs of abuse in saliva using surface enhanced Raman spectroscopy and microfluidics. ACS Nano 2013, 7, 7157–7164. [CrossRef] 25. Han, Z.Z.; Liu, H.L.; Meng, J.; Yang, L.B.; Liu, J.; Liu, J.H. Portable kit for identification and detection of drugs in human urine using surface-enhanced Raman spectroscopy. Anal. Chem. 2015, 87, 9500–9506. [CrossRef][PubMed] 26. Dong, R.; Weng, S.; Yang, L.; Liu, J. Detection and direct readout of drugs in human urine using dynamic surface-enhanced Raman spectroscopy and support vector machines. Anal. Chem. 2015, 87, 2937–2944. [CrossRef] 27. Han, Z.; Liu, H.; Wang, B.; Weng, S.; Yang, L.; Liu, J. Three-dimensional surface-enhanced Raman scattering hotspots in spherical colloidal superstructure for identification and detection of drugs in human urine. Anal. Chem. 2015, 87, 4821–4828. [CrossRef] 28. Ma, Y.; Liu, H.; Mao, M.; Meng, J.; Yang, L.; Liu, J. Surface-enhanced Raman spectroscopy on liquid interfacial nanoparticle arrays for multiplex detecting drugs in urine. Anal. Chem. 2016, 88, 8145–8151. [CrossRef] 29. Nuntawong, N.; Eiamchai, P.; Somrang, W.; Denchitcharoen, S.; Limwichean, S.; Horprathum, M.; Patthanasettakul, V.; Chaiya, S.; Leelapojanaporn, A.; Saiseng, S. Detection of methamphetamine/amphetamine in human urine based on surface-enhanced Raman spectroscopy and acidulation treatments. Sens. Actuators B-Chem. 2017, 239, 139–146. [CrossRef] 30. Fang, W.; Zhang, B.; Han, F.Y.; Qin, Z.N.; Feng, Y.Q.; Hu, J.M.; Shen, A.G. On-site and quantitative detection of trace methamphetamine in urine/serum samples with a surface-enhanced Raman scattering-active microcavity and rapid pretreatment device. Anal. Chem. 2020, 92, 13539–13549. [CrossRef][PubMed] 31. Faulds, K.; Smith, W.E.; Graham, D.; Lacey, R.J. Assessment of silver and gold substrates for the detection of amphetamine sulfate by surface enhanced Raman scattering (SERS). Analyst 2002, 127, 282–286. [CrossRef] 32. Turzhitsky, V.; Zhang, L.; Horowitz, G.L.; Vitkin, E.; Khan, U.; Zakharov, Y.; Qiu, L.; Itzkan, I.; Perelman, L.T. Picoanalysis of drugs in biofluids with quantitative label-free surface-enhanced Raman spectroscopy. Small 2018, 14, 1802392. [CrossRef][PubMed] 33. Stewart, A.; Bell, S.E.J. Modification of Ag nanoparticles with mixed thiols for improved SERS detection of poorly adsorbing target molecules: Detection of MDMA. Chem. Commun. 2011, 47, 4523–4525. [CrossRef][PubMed] 34. Lee, W.W.Y.; Silverson, V.A.D.; Jones, L.E.; Ho, Y.C.; Fletcher, N.C.; McNaul, M.; Peters, K.L.; Speers, S.J.; Bell, S.E.J. Surface- enhanced Raman spectroscopy of novel psychoactive substances using polymer-stabilized Ag nanoparticle aggregates. Chem. Commun. 2016, 52, 493–496. [CrossRef][PubMed] Nanomaterials 2021, 11, 1789 14 of 14

35. Muhamadali, H.; Watt, A.; Xu, Y.; Chisanga, M.; Subaihi, A.; Jones, C.; Ellis, D.I.; Sutcliffe, O.B.; Goodacre, R. Rapid detection and quantification of novel psychoactive substances (NPS) using Raman spectroscopy and surface-enhanced Raman scattering. Front. Chem. 2019, 7, 412. [CrossRef][PubMed] 36. Farquharson, S.; Lee, Y.H. Trace drug analysis by surface-enhanced Raman spectroscopy. Proc. SPIE 2000, 4200, 89–95. 37. Sutherland, W.S.; Laserna, J.J.; Angebranndt, M.J.; Winefordner, J.D. Surface-enhanced Raman analysis of sulfa drugs on colloidal silver dispersion. Anal. Chem. 1990, 62, 689–693. [CrossRef][PubMed] 38. Liu, H.L.; Yang, L.B.; Liu, J.H. Three-dimensional SERS hot spots for chemical sensing: Towards developing a practical analyzer. TrAC-Trends Anal. Chem. 2016, 80, 364–372. [CrossRef] 39. Zhang, Q.; Lee, Y.H.; Phang, I.Y.; Lee, C.K.; Ling, X.Y. Hierarchical 3D SERS substrates fabricated by integrating photolithographic microstructures and self-assembly of silver nanoparticles. Small 2014, 10, 2703–2711. [CrossRef] 40. Liu, H.L.; Yang, Z.L.; Meng, L.Y.; Sun, Y.D.; Wang, J.; Yang, L.B.; Liu, J.H.; Tian, Z.Q. Three-dimensional and time-ordered surface-enhanced Raman scattering hotspot matrix. J. Am. Chem. Soc. 2014, 136, 5332–5341. [CrossRef] 41. Wang, H.Y.; Fang, J.M.; Xu, J.F.; Wang, F.; Sun, B.; He, S.N.; Sun, G.P.; Liu, H.L. A hanging plasmonic droplet: Three-dimensional SERS hotspots for a highly sensitive multiplex detection of amino acids. Analyst 2015, 140, 2973–2978. [CrossRef][PubMed] 42. Sun, Y.D.; Han, Z.Z.; Liu, H.L.; He, S.N.; Yang, L.B.; Liu, J.H. Three-dimensional hotspots in evaporating nanoparticle sols for ultrahigh Raman scattering: Solid-liquid interface effects. Nanoscale 2015, 7, 6619–6626. [CrossRef][PubMed] 43. Wang, Y.K.; Wei, Z.Y.; Zhang, Y.; Chen, Y.F. Glycerol-assisted construction of long-life three-dimensional surface-enhanced Raman scattering hot spot matrix. Langmuir 2019, 35, 15795–15804. [CrossRef][PubMed] 44. Panigrahi, S.; Praharaj, S.; Basu, S.; Ghosh, S.K.; Jana, S.; Pande, S.; Vo-Dinh, T.; Jiang, H.; Pal, T. Self-assembly of silver nanoparticles: Synthesis, stabilization, optical properties, and application in surface-enhanced Raman scattering. J. Phys. Chem. B 2006, 110, 13436–13444. [CrossRef][PubMed] 45. Koo, T.W.; Chan, S.; Sun, L.; Su, X.; Zhang, J.W.; Berlin, A.A. Specific chemical effects on surface-enhanced Raman spectroscopy for ultra-sensitive detection of biological molecules. Appl. Spectrosc. 2004, 58, 1401–1407. [CrossRef] 46. Smith, J.P.; Metters, J.P.; Khreit, O.I.G.; Sutcliffe, O.B.; Banks, C.E. Forensic electrochemistry applied to the sensing of new psychoactive substances: Electroanalytical sensing of synthetic cathinones and analytical validation in the quantification of seized street samples. Anal. Chem. 2014, 86, 9985–9992. [CrossRef]