Chemometrics

Subscriber access provided by OCCIDENTAL COLL

Technical Note

“Lab-on-Click” detection of illicit drugs in oral fluids by MicroNIR/Chemometrics Roberta Risoluti, Giuseppina Gullifa, Alfredo Battistini, and Stefano Materazzi Anal. Chem., Just Accepted Manuscript • Publication Date (Web): 29 Apr 2019 Downloaded from http://pubs.acs.org on April 29, 2019

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 11 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

“Lab-on-Click” detection of illicit drugs in oral fluids by MicroNIR/Chemometrics Roberta Risoluti§*, Giuseppina Gullifa§, Alfredo Battistini† and Stefano Materazzi§

§Department

of Chemistry, “Sapienza” University of Rome, p.le A.Moro 5, 00185 Rome Italy per la ricerca in agricoltura e l’analisi dell’economia agraria – Centro di Politiche e Bioeconomia, via Pò 14, 00198 Rome Italy †Consiglio

KEYWORDS: Illicit drugs, MicroNIR, Chemometrics, Screening ABSTRACT: A novel entirely automated MicroNIR/Chemometric platform was developed for the “Lab-on-Click” detection of illicit drugs in non-pretreated oral fluids and a novel tool for the first level test is proposed. Calibration of the method was achieved by collecting oral fluids specimens from volunteers and chemometric analysis was considered for the development of models of prediction for cocaine, amphetamine and Δ9-Tetrahydrocannabinol. In addition, a comprehensive model was optimized for the simultaneously prediction of positive/negative samples and the specific illicit drug used by abusers in single “click”. The detection ability of the method was checked for true positive and false positive outcomes and results were validated by GC-MS reference official method. The MicroNIR/Chemometric platform provided the simultaneously prediction of the three most frequently addictive drugs with the sensitivity and accuracy of the confirmatory analyses, offering the advantages of rapidity and simplicity and demonstrating to be a promising tool supporting public health surveillance.

Up to date, testing for illicit drugs in oral fluid is a customary procedure for determining recent drug abuse as the impairment with similar determinations in urine is long time demonstrated 1, 2. As a consequence, the interest in the development of accurate, sensitive and easy to use screening tests for workplace surveillance 3 and roadside controls 4 , is increasing more and more. A number of roadside screening tests are usually used by Police enforcement to detect illicit drug traces in oral fluids 5-7 and innovative approaches are proposed for the first level test including Raman spectroscopy 8, the involvement of gold nanorods 9 and molecularly imprinted polymers 10. Neverthless, the confirmatory analysis usually requires liquid chromatographic techniques associated to mass spectrometer detectors 11-13. Despite all the advances in the field, there is still a lack of accurate method achieving a suitable sensitivity to avoid true positive and false negative response to be use for roadside controls in a completely automated device 14-16. Nowadays, the detection of psychoactive drugs by spectroscopic techniques was found to be very challenging as non-destructive methods providing the accurate analysis of complex matrices 17-22. In addition, portable and one-touch devices ensured the possibility of timely monitoring drivers and workers under the influence of drugs 23. In this work, a “Lab-on-Click” platform is designed to detect illicit drugs in non-pretreated oral fluids by MicroNIR/Chemometric approach 24-29. This novel automated and miniaturized device permits collection, processing and prediction in a single “click” with the accuracy of the official confirmatory analyses.

EXPERIMENTAL SECTION Materials and samples collection Illicit drugs reference standards of cocaine (COC), amphetamine (AMP) and Δ9Tetrahydrocannabinol (THC) were purchased from SigmaAldrich (St. Louis, Missouri United States) as methanolic solution at the concentration of 1 mg-ml. All the standards used were pure at 99%. Oral fluids were collected from 50 anonimous voluteers by using the pad from QuantisalTM collection kit (Immunalysis Corporation, Pomona, CA). Different habits of volunteers were considered and subjects were asked for eating or drinking as usual, taking into consideration potential interferences in the platform response (caffeine, sugars, chewing-gum, smoke, alcohol). During model calibration, all these situations were represented by a pool of oral fluids analyzed as such and spiked with increasing amount of psychoactive substances in the range 1-100 ng-ml. MicroNIR measuremets did not required any sample pretreatment and OF were directly processed by the platform after collection, while all the samples were extracted and derivatized prior to GC-MS for confirmatory analysis according to reference procedures 30. MicroNIR/Chemometric platform The MicroNIR is an ultra compact and portable device (45 mm in diameter and 42 mm in height), developed and distributed by Viavi Solutions (JDSU Corporation, Milpitas, CA) and operates in the spectral region of 900− 1700 nm. It weighs about 60 g and it is entirely powered (5 V) and controlled by a USB port of a portable computer. In this work, the miniaturized MicroNIR platform

ACS Paragon Plus Environment


Confirmatory GC-MS A Perkin Elmer (Waltham, MA) GC system interfaced with a mass selective detector was used for the analyses. Separation was accomplished by a HP-5MS (30 m x 0.25 mm x 0.25 mm) using helium as carrier gas at 1 mL/min. Injections were carried out in split mode (split ratio of 1:50) and an injector temperature of 250 °C. The oven temperature program was set as follows: 120°C for 1 min, ramped to 240°C at 30°C/min and then increased to 290°C at 10°C/min for 10 min. A post run was made at 300 °C for 5 min. The mass spectrometer parameters were as follows: transfer line at 250 °C, ion source at 230 °C and quadrupole at 150 °C. Mass spectral data was collected in the scan mode from m/z 44 to 450.

This particular assembling strategy was found to be very promising because it demonstrated to provide the optimal focal point between the radiation source and the collecting swab, resulting in a significant improvement in the spectral signal as all the surface of the swab was processed in a single measurement. Overlapped spectra of oral fluids and spiked oral fluids with COC, AMP and THC collected onto the pad are reported in Figure 1. 1.4

35

3.0 1

2.5 0.4

PC 2 (20 %)

was designed to easy collect spectra from swab. To this aim, a particular MicroNIR windowed collar accessory was used to get the optimal focal point of the radiation from the spectrometer, featuring an integrated anti-reflective coated Sapphire window. In addition, a linear variable filter (LVF), the dispersing element, was directly connected to a 128 pixel linear InGaAs array detector and two tungsten light bulbs were used as the radiation source. All the collected spectra were recorded in the reflectance mode and 6.25 nm was select as the most performing nominal spectral resolution. Spectralon was used as the NIR-reflectance standard (the blank) with a 99% diffuse reflectance, while the dark reference was obtained from a fixed place in the room. The acquisitions were performed with an integration time of 10 ms, resulting in a total measurement time of 2.5 s per sample. The MicroNIR Pro software (JDSU Corporation, Milpitas, CA) was used for the automated instrument control and all of the chemometric analysis were performed by VJDSU Unscrambler Lite (Camo software AS, Oslo, Norway). Four different models of prediction for illicit drugs abuse were provided in this study and a number of chemometric spectral pretreatments were investigated in order to provide the best separation among samples. To this aim, standard-normalvariate (SNV) transform, multiplicative scatter correction (MSC) as normalization were evaluated 31-33, whereas the Savitzky-Golay (SG) polynomial-derivative filter 34 was considered as spectral-derivation technique.

% Reflectance

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 11

2.0 0

1.5

-0.4

-1

-1.5

1.0

Blank OF Blank OF + COC OF + AMP Blank + COC OF + THC Blank + AMP

5.0 0 900

1000

1100

1200

1300

1400

1500

Blank + THC

1600

1700

Wavelenght (nm)

Figure 1. Overlapped spectra of oral fluids (OF, blue) and spiked oral fluids with COC (red), AMP (green) and THC (black) collected onto the pad.

For each dataset of measurements, Principal Component Analysis (PCA) was used as exploratory tool to investigate correlations among samples, while Partial Least Square regression algorithms were involved to develop three models of prediction for COC, AMP and THC in oral fluids. Results of the scores plots for each illicit drug are reported in Figure 2, where colors are used to identify blank and spiked OF with different amount of illicit drugs.

RESULTS AND DISCUSSION MicroNIR analysis Providing analytical tools for fast and accurate prediction of illicit drugs abuse represents a challenging issue for the scientific research applied to forensic field and may contribute in addressing the problem of the frequency of accidents occurring after driving under the effects of psychoactive substances. With the aim of developing a tool to be used on-site, oral fluids from volunteers were collected onto swab and spectra in the NIR region were recorded by the mean of the a special collar directly connected with the MicroNIR. In order to provide a platform able to correctly identify illicit drugs in oral fluids, three models of prediction were developed separately according to the following procedure: spectra of oral fluids were recorded as such and spiked with increasing amount of cocaine (COC), amphetamine (AMP) and Δ9Tetrahydrocannabinol (THC) in the range 1-100 ng-ml. In a second stage, the same set of samples were deposited onto the pad by dipping 1 ml of each specimen and spectra were recorded by the MicroNIR provided by the special collar.


Page 3 of 11

Analytical Chemistry 3

COC

2

a)

a)

PC 2 (18 %)

1 0 -1

Blank Blank + 1 ng Blank + 10 ng Blank + 100 ng

-2 -3 -4

-4

4

-3

-2

AMP

-1 PC 1 (28 %)

0

1

2

COC

a)

AMP

b)

3

b)

b)

3

PC 2 (18 %)

2 1 0 -1 -2 -3

-4

0.03

-3

-2

-1

PC 1 (30 %)

0

1

2

3

THC

c)

c)

0.02

0.01

PC 2 (17 %)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

0

-0.01 -0.02 -0.03 -0.05

-0.04

-0.03

-0.02

-0.01 0 PC 1 (30 %)

0.01

0.02

0.03

0.04

0.05

Figure 2. Resulting scores plot of Blank OF (black) and spiked OF with 1 ng (blue), 10 ng (red) and 100 ng (green) of COC (a), AMP (b) and THC (c).

Figure 2 provides preliminary important information about samples correlation. In particular, all the samples belonging to the same class were found to be well grouped according to PC1 for all the three dataset of spiked samples, as the overall explained variance (expressed as the sum of PC1 and PC2) for COC, AMP and THC was about 46%, 43% and 56% respectively. In addition the MicroNIR/Chemometric approach simultaneously distinguishes samples according to the presence of the molecule and to the quantity of the illicit drug. This behaviour represents an important issue in forensic field and suggests to further investigate the possibility to develop models of prediction, able to accurately detect and quantify illicit drugs abuse at the first level test.

THC c) Blank Blank + 1 ng Blank + 10 ng Blank + 100 ng Figure 3. 3-D plots of the PLS regression models for COC (a), AMP (b) and THC (c).

To this aim, PLS regression was performed for the three data set and chemometric techniques were also used to identify the variables affecting most the separation of the samples, thus improving the accuracy of the prediction. Stepwise decorrelation of variables was applied to the spectral datasets in order to identify variables with the largest Fisher weight 35. In particular, spectra from oral fluids spiked with COC (Figure 3a) and AMP (Figure 3b) were baseline corrected and pretreated by first derivative transform followed by Standard Normal Variate (SNV) while spectra from oral fluids spiked with THC (Figure 3c) were baseline corrected and pretreated by first derivative transform. In addition, as not all the wavelengths contribute equally to samples separation, the range 908-1040 nm was selected in order to minimize the contribution of the matrix in the case of COC, while for AMP and THC dataset, the ranges 980-1030 nm and 914-1100 nm were considered, respectively.


Analytical Chemistry The 3-D plots of the models are reported in Figure 3, while the prediction ability expressed as the Root Mean Squared Error of Calibration (RMSEC), Root Mean Squared Error of CrossValidation (RMSECV) and Root Mean Squared Error of Prediction (RMSEP) 36, 37, the correlation (R2) and Minimum Detection Concentration (MDC) are summarized in Table 1.

the investigated molecule in single “click”. A suitable separation among samples belonging to different classes was achieved, as shown by the 3-D plot in Figure 5 and the resulting figures of merits were calculated for all the investigated classes. Data are reported in Table 2.

Table 1 Prediction ability of the PLS regression models for COC, AMP and THC. COC

AMP

THC

2.9 3.1 3.5

3.0 4.1 5.0

4.4 4.5 5.2

3 0.97 1

4 0.96 1

Figures of merit RMSEC RMSECV

RMSEP 3 LV* 0.98 R2 Validation 1 MDC** (ng) *Latent variables **Minimum detection concentration

Lab-on-click model On the basis of the previously attractive results and with the aim of providing an effective tool supporting forensic controls, a completely automated “Lab-onClick” platform was developed and validated for the on-site multiparametric detection of illicit drugs in oral fluids (COC, AMP and THC). Unsupervised Principal Component Analysis was used to simultaneously correlate all the collected samples. Baseline correction and first derivative transform followed by Standard Normal Variate (SNV) were applied as chemometric pretreatments to the entire dataset in the interval 900-1250 nm. As reported in Figure 4, the location of the samples in the plot of the scores shows a good accordance among spectra corresponding to the same class. In addition, the plot provides for two type of outcomes: first, moving along PC 2 (20 % of explained variance) the blank oral fluids were found to be well grouped in the upper side of the plot, while moving along PC 1 (50 % of explained variance) samples resulted separated according to molecule demonstrating to be a promising tool to simultaneously predict positive samples and the involved molecule. 1.4 1

0.4

PC 2 (20 %)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 11

0

Blank Blank + COC Blank + AMP Blank + THC Figure 5 3-D plots of the PLS regression models for COC (blue), AMP (red) and THC (green).

Table 2. Figures of merits of the PLS-DA model calculated for each class

Blank COC AMP THC

NER (%) 100 98.1 98.1 100

Calibration Sp RMSE (%) (%) 100 0.1 98.1 0.2 95.3 0.2 100 0.1

NER (%) 100 96.9 91.0 100

Validation Sp (%) 100 96.2 95.9 100

RMSE 0.1 0.2 0.2 0.1

NER (%) 100 100 100 100

Prediction Sp RMSE (%) (%) 100 0.1 100 0.2 100 0.3 100 0.1

The prediction ability of the model was expressed as the Non Error Rate (NER %), the specificity (Sp %) and the Root Mean Square Error (RMSE) in calibration, validation and prediction. Results reported in Table 2, indicates that the platform provides satisfactory accuracy in calibration and validation for all the processed classes (NER % not less than 90 %) resulting in a 100 % of correct classified samples when predicting the illicit drugs abuse. In addition, the specificity of the platform ranges from 95 % (for AMP) to 100 % (THC and blank) and the RMSE is not higher than 0.1 % in prediction.

-0.4

Blank Blank + COC Blank + AMP Blank + THC

-1

-1.5

-3

-2

-1

0

1

2

3

PC 1 (50 %)

Figure 4. Resulting scores plot of Blank OF (black) and spiked OF with COC (blue), AMP (red) and THC ng (green)

Therefore, calibration and validation of a comprehensive model of prediction for COC, AMP and THC was developed based on Partial Least Square - Discriminant Analysis (PLSDA) and optimized for the fast and accurate detection of all

Real samples prediction A number of 11 real samples were provided by anonimous volunteers and processed by the validated MicroNIR/Chemometric platform. Among the investigated samples, two subjects were found to be positive to cocaine, one subject was found to be positive to AMP, while the remaining 7 were found to be predicted as positive to THC (3 subjects) and negative (4 subjects). In order to check the false positive and false negative response of the platform, extraction and analysis by GC-MS were performed according to the reference procedures 30. A good correlation among the data from the two methods was observed since all the samples were correctly predicted by the model with no


Page 5 of 11 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


miscalssification results. In addition, in such cases, metabolites as Benzoylecgonine (BEG) and 11-hydroxytetrahydrocannabinol (OH-THC) were recovered in the unknown oral fluids by GC-MS analysis, leading to confirm the cocaine and THC abuse. Interestingly, even in these cases, samples were correctly predicted by the MicroNIR platform as positive to cocaine and THC. This is not surprising because the NIR spectroscopy provides different signal as a function of the different structure of the molecule and the three investigated illicit drugs belong to three different classes, as well as their metabolites. Therefore, the MicroNIR/Chemometric platform would represent a promising novel tool as screening method that provides the accuracy of the reference procedure in one-touch, nondestructive and solvent-less analysis. CONCLUSIONS MicroNIR/Chemometric approach was investigated for the first time to develop an analytical platform able to detect illicit drugs abuse in non pre-treated oral fluids. Three model of prediction for the detection of cocaine (COC), amphetamine (AMP) and thetraidrocannabinol (THC) were optimized on collected oral fluids and a comprehensive one-touch platform was proposed for the simultaneously detection of illicit drugs abuse. This novel platform permits collection, processing and prediction of illicit drugs in a single “click” with the accuracy of the official confirmatory analyses and offering the advantages of rapidity and simplicity and representing a real innovative tool to be included into conventional procedures.

AUTHOR INFORMATION Corresponding Author * Roberta Risoluti, Department of Chemistry, Sapienza University of Rome, Piazzale Aldo Moro 5, 00185 Rome, Italy; Tel +390649913616 fax: +390649387137 e-mail address: [email protected]

Author Contributions RR and SM conceived the study and wrote the manuscript. Data were obtained through contributions of all authors. All authors have given approval to the final version of the manuscript.

Notes The authors declare no competing financial interest.

REFERENCES (1) Toennes, S.W.; Kauert, G.F.; Steinmeyer, S.; Moeller, M.R. Driving under the influence of drugs evaluation of analytical data of drugs in oral fluid, serum and urine, and correlation with impairment symptoms. Forensic Science International 2005, 152, 149–155. (2) Fierro, I.; González-Luque, J.C.; Alvarez, F.J. The relationship between observed signs of impairment and THC concentration in oral fluid. Drug and Alcohol Dependence 2014, 144, 231–238 (3) Substance Abuse and Mental Health Administration. Revised mandatory guidelines by secretary of health and human services (SAMHSA, HHS) 2004, 69, 19673–19732.

(4) Verstraete, A.G. Proceedings of the 17th International Conference on Alcohol, Drugs Traffic Safety, Glasgow, Scotland, 2004. (5) Möller, M.R.; Steinmeyer, S.; Aberl, F. Deliverable operational, user and legal requirements across EU member states for roadside drug testing equipment: Rosita Deliverables. Rosita Consortium, Gent, 1999, 1–59. (6) Cooper, G.; Wilson, L.; Reid, C.; Main, L.; Hand, C. Evaluation of the Cozart1 RapiScan drug test system for opiates and cocaine in oral fluid. Forensic Science International 2005, 150, 239–243. (7) Wille, S.M.R.; Samyn, N.; del Mar Ramirez-Fernandez, M.; De Boeck, G. Evaluation of on-site oral fluid screening using Drugwipe-5(+), RapidSTAT and Drug Test 5000 for the detection of drugs of abuse in drivers. Forensic Science International 2010, 198, 2–6. (8) D'Elia, V.; Montalvo, G.; Garcia Ruiz, C.; Ermolenkov, V.V.; Ahmed, Y.; Lednev, I.K. Ultraviolet resonance Raman spectroscopy for the detection of cocaine in oral fluid. Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 2018, 188 , 338–340. (9) D'Elia, V.; Rubio-Retama, J.;. Ortega-Ojeda, F.E.; GarciaRuiz, C.; Montalvo, G. Gold nanorods as SERS substrate for the ultratrace detection of cocaine in non-pretreated oral fluid samples. Colloids and Surfaces A 2018, 557, 43–50. (10) Chantada–Vázquez, M.P.; de–Becerra–Sanchez, C.; Fernández–del–Río, A.; Sánchez–González, J.; Bermejo, A.M.; Bermejo–Barrera,P.; Moreda–Pineiro, A. Development and application of molecularly imprinted polymer - Mn-doped ZnS quantum dot fluorescent optosensing for cocaine screening in oral fluid and serum. Talanta 2018, 181, 232–238. (11) Pujadas, M.; Pichini, S.; Civit, E.; Santamarina, E.; Perez, K.; de la Torre, R. A simple and reliable procedure for the determination of psychoactive drugs in oral fluid by gas chromatography-mass spectrometry. Journal of Pharmaceutical and Biomedical Analysis 2007, 44, 594–601. (12) Wylie, F.M.; Torrance, H.; Anderson, R.A.; Oliver, J.S. Drugs in Oral Fluid Part I.Validation of an analytical procedure for illicit and illicit drugs in oral fluid. Forensic Science International 2005, 150, 191–198. (13) Milman, G.; Barnes, A.J.; Lowe1, R.H.; HuestiS, M.A. Simultaneous quantification of cannabinoids and metabolites in oral fluid by two-dimensional gas chromatography mass spectrometry. Journal of Chromatography A 2010, 1217, 1513– 1521. (14) Musshoff, F.; GroBe Hokamp, E.; Bott, U.; Madea, B. Performance evaluation of on-site oral fluid drug screening devices in normal police procedure in Germany. Forensic Science International 2014, 238, 120–124. (15) Van der Linden, T.; Wille, S.M.R; Ramirez-Fernandez, M.; Verstraete, A.G.; Samyn, N. Roadside drug testing: comparison of two legal approaches in Belgium. Forensic Science International 2015, 249, 148–155. (16) Houwing, S.; Hagenzieker, M.; Mathijssen, R.P.M.; Legrand, S.A.; Verstraete, A.G.; Hels, T.; Bernhoft, I.M.; Simonsen, K.W.; Lillsunde, P.; Favretto, D.; Ferrara, S.D.; Caplinskiene, M.; Movig, K.L.L.; Brookhuis, K,A. Random and systematic errors in case-control studies calculating the injury risk of driving under the influence of psychoactive substances. Accident Analysis and Prevention 2013, 52, 144–153. (17) Pereira, L.S.A.; Lisboa, F.L.C.; Neto, J.C.; Valladão, F.N.; Sena, M.M. Direct classification of new psychoactive substances in seized blotter papers by ATR-FTIR and multivariate discriminant analysis. Microchemical Journal 2017, 133, 96–103. (18) Pereira, L.S.A.; Lisboa, F.L.C.; Neto, J.C.; Valladão, F.N.; Sena, M.M. Screening method for rapid classification of psychoactive substances in illicit tablets using mid infrared


Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

spectroscopy and PLS-DA. Forensic Science International 2018, 288, 227–235. (19) Risoluti, R.; Materazzi, S.; Gregori, A.; Ripani, L. Early detection of emerging street drugs by near infrared spectroscopy and chemometrics. Talanta 2016, 153, 407–413. (20) Materazzi, S.; Peluso, G.; Ripani, L.; Risoluti, R. Highthroughput prediction of AKB48 in emerging illicit products by NIR spectroscopy and Chemometrics. Microchemical Journal 2017, 134, 277–283. (21) Materazzi, S.; Gregori, A.; Ripani, L.; Apriceno, A.; Risoluti, R. Cocaine profiling: implementation of a predictive model by ATR-FTIR coupled with chemometrics in forensic chemistry. Talanta 2017, 166, 328-335. (22) de la Asuncion-Nadal, V.; Armenta, S.; Garrigues, S.; de la Guardia, M. Talanta 2017, 167, 344–351. (23) Schulze, H.S.M.; Urmeew, R.; Auerbach, K.; Alvarez, J.; Bernhoft, I.M.; de Gier, H.; Hagenzieker, M.; Houwing, S.; Knoche, A.; Pilgerstorfer, M.; Zlender, B. Finding From the DRUID Project, EMCDDA, Lisbon, 2012. (24) Baca-Bocanegra, B.; Hernández-Hierro, J.M.; NogalesBueno, J.; Heredia, F.J. Feasibility study on the use of a portable micro near infrared spectroscopy device for the "in vineyard" screening of extractable polyphenols in red grape skins. Talanta 2019, 192, 353–359. (25) da Silva Oliveira, V.; Honorato, R.S.; Honorato, F.A.; Pereira, C.F. Authenticity assessment of banknotes using portable near infrared spectrometer and chemometrics. Forensic Science International 2018, 286, 121–127. (26) Risoluti, R.; Materazzi, S.; Tau, F.; Russo, A.; Romolo, F.S. Towards innovation in paper dating: a MicroNIR analytical platform and chemometrics. Analyst 2018, 143, 4394-4399. (27) Risoluti, R.; Materazzi, S. MicroNIR/Chemometrics assessement of occupational exposure to hydroxyurea. Frontiers in Chemistry 2018, 6, 228. (28) Risoluti, R.; Gregori, A.; Schiavone, S.; Materazzi, S. “Click and Screen" technology for explosives detection on human hands by portable MicroNIR/Chemometrics platform. Analytical Chemistry 2018, 90, 4288-4292. (29) Pereira, J.F.Q.; Silva, C.S.; Vieira, M.J.L.; Pimentel, M.F.; Braz, A.; Honorato, R.S. Evaluation and identification of blood stains with handheld NIR spectrometer. Microchemical Journal 2017, 133, 561–566. (30) Pichini, S.; Pacifici, R. Linee guida per la determinazione di sostanza d’abuso nella saliva. Istituto Superiore di Sanità 2013. (31) Barnes, R.J., Dhanoa, M.S., Lister, S.J. Standard Normal Variate Transformation and De-Trending of Near-Infrared Diffuse Reflectance Spectra. Spectra. Appl. Spectros. 1989, 43, 772-777. (32) Geladi, P.; MacDougall, D.; Martens, H. Linearization and Scatter-Correction for Near-infrared reflectance spectra of meat. Appl. Spectrosc. 1985, 39, 491-500. (33) Wold, S.; Sjöström, M. American Chemical Society Symposium Series 1977, 52, 243-282A. (34) Savitzky, M.J.E. Golay. Smoothing and differentiation of data by simplified least squares procedures. Anal.Chem. 1964, 36, 1627-1639. (35) Wold, S.; Sjöström, M. SIMCA: A Method for Analyzing Chemical Data in Terms of Similarity and Analogy. Chemom.: Theory Appl. 1977, 12, 243–282. (36) Mark, H.; Workman, J. Chemometrics in Spectroscopy, Elsevier/Academic Press, Amsterdam, 2007. (37) Miller, J.N.; Miller, J.C.; Statistics and Chemometrics for Analytical Chemistry, Prentice Hall, Harlow, England, 2000.


Page 6 of 11

Page 7 of 11 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


TOC of the manuscript

MicroNIR

Sampling

GC-MS

Results


Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Page 8 of 11

Page 9 of 11

3.50E+01% 35

1.4

Blank% Blank%+%TNT% Blank%+%RDX%% Blank%+%PETN%

3.00E+01%

3.0 1 2.5 2.50E+01%

0.4

% Reflectance

PC 2 (20 %) %%Reﬂectance%

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41


2.00E+01% 2.0

1.5 1.50E+01% 0

-0.4

1.00E+01% 1.0

Blank OF Blank OF + COC OF + AMP Blank + COC OF + THC Blank + AMP

-1 5.00E+00% 5.0 0.00E+00% Blank + THC 0 900% 1000% 1100% 1200% 1300% 1400% 1500% 1600% 1700% 900 1000 1100 1200 1300 1400 1500 1600 1700 -1.5 Wavelength%(nm)% Wavelenght (nm) ACS Paragon Plus Environment

3

2

COC


Page 10 of 11

a)

PC 2 (18 %)

1 2 1 3 4 5 0 6 7 8 -1 9 10 11 -2 12 13 14 -3 15 16 17 -4 -4 18 19 4 20 21 3 22 23 24 2 25 26 27 1 28 29 0 30 31 -1 32 33 34 -2 35 36 37 -3 -4 38 39 40 0.03 41 42 43 0.02 44 45 46 47 0.01 48 49 0 50 51 52 53 -0.01 54 55 56 57-0.02 58 59 60-0.03

Blank Blank + 1 ng Blank + 10 ng Blank + 100 ng

-3

-2

-1 0 PC 1 (28 %)

AMP

1

2

3

PC 2 (18 %)

b)

-3

-2

-1

0 PC 1 (30 %)

THC

1

2

3

PC 2 (17 %)

c)

-0.05


-0.04 -0.03 -0.02 -0.01 0 0.01 0.02 0.03 0.04 0.05 PC 1 (30 %)

Page 11 of 11

PC 2 (20 %)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41

1.4


1

0.4

0 -0.4

Blank Blank + COC Blank + AMP Blank + THC

-1 -1.5 -3

-2

-1

0

1

PC 1 (50 %) ACS Paragon Plus Environment

2

3

Chemometrics

Recommend Documents