Detection of Papaverine for the Possible Identification of Illicit Opium

Article pubs.acs.org/ac

Detection of Papaverine for the Possible Identification of Illicit Opium Cultivation Rustin Y. Mirsafavi,† Kristine Lai,‡ Neal D. Kline,∥ Augustus W. Fountain III,∥ Carl D. Meinhart,*,‡ and Martin Moskovits*,§ †

Department of Biomolecular Science and Engineering, University of California Santa Barbara, Santa Barbara, California 93106, United States ‡ Department of Mechanical Engineering, University of California Santa Barbara, Santa Barbara, California 93106, United States § Department of Chemistry and Biochemistry, University of California Santa Barbara, Santa Barbara, California 93106, United States ∥ Edgewood Chemical and Biological Center, Edgewood, Maryland 21010, United States S Supporting Information *

ABSTRACT: Papaverine is a non-narcotic alkaloid found endemically and uniquely in the latex of the opium poppy. It is normally refined out of the opioids that the latex is typically collected for, hence its presence in a sample is strong prima facie evidence that the carrier from whom the sample was collected is implicated in the mass cultivation of poppies or the collection and handling of their latex. We describe an analysis technique combining surface-enhanced Raman spectroscopy (SERS) with microfluidics for detecting papaverine at low concentrations and show that its SERS spectrum has unique spectroscopic features that allows its detection at low concentrations among typical opioids. The analysis requires approximately 2.5 min from sample loading to results, which is compatible with field use. The weak acid properties of papaverine hydrochloride were investigated, and Raman bands belonging to the protonated and unprotonated forms of the isoquinoline ring of papaverine were identified.

T

significance, such as methamphetamine in saliva and ampicillin in milk, at concentrations as low as 10 ppb.8,9 Those approaches, however, do not efficiently distinguish the illicit user or trafficker of opiates from individuals involved in the cultivation of opiate based narcotics or their refinement from opium poppies, i.e., the individuals closest to the source. The harvesting process includes extracting and handling latex from the ripening plant pods, a task carried out largely by hand.10 The manual harvesting in combination with the low standard of hygiene associated with such work leads to the accumulation of papaverine on harvesters. To identify harvesters of opium, a quick and reliable system must be developed. Raman and SERS spectra of papaverine have been previously reported; however, the aim of those studies was the understanding of the vibrational spectrum rather than the development of a fieldable assay at low concentrations.11 The most common method of papaverine detection involves liquid or gas phase chromatography usually coupled with mass spectrometry.12−16 Sägmüller et al. did use SERS to detect

he opium poppy (Papaver somniferum) has been cultivated for its analgesic and narcotic properties since prehistoric times.1 The sap, or latex, of the poppy plant when dried to yield opium, contains multiple alkaloids including morphine, thebaine, codeine, papaverine, and noscapine (see Figure 1). While opium is the source for the active ingredients of several pain-relieving pharmaceuticals such as morphine, codeine, and hydrocodone, it is also the source of heroin, a highly addictive, illicit substance. Opiate addiction is a serious medical and social problem, with significant harmful economic consequences;2 and the opium trade is a serious global criminal enterprise, some of whose profits also fund international terrorist activities.3 Moreover, despite massive national and international policing efforts, those who traffic in opiates routinely find new ways to move these substances across borders. Developing the analytical means for credibly identifying the people responsible for the early stages of this complex trade, those who grow, harvest, and obtain the latex from the poppy, could, therefore, be useful in reducing opiate trafficking. Surface-enhanced Raman spectroscopy (SERS) has been used for detecting opiate analytes of forensic importance for quite some time.4−7 Microfluidic/SERS assays have also been previously utilized to detect other analytes of forensic © XXXX American Chemical Society

Received: September 26, 2016 Accepted: January 3, 2017 Published: January 3, 2017 A

DOI: 10.1021/acs.analchem.6b03797 Anal. Chem. XXXX, XXX, XXX−XXX

Article

Analytical Chemistry

the individual from whom the sample was obtained in the largescale cultivation of poppies and/or in handling its latex.

■

EXPERIMENTAL SECTION The flow-focusing microfluidic device was designed with three channels that merge at a junction at which analyte (middle channel), aqueous AgNPs, and LiCl solution (side channels) join, thereafter creating interfaces across which molecular species can diffuse (as shown in Figure 2). The fluid velocity magnitude is shown in Figure 2, which was estimated from numerical simulation of the Navier−Stokes equations using COMSOL Multiphysics V5.2 (Stockholm, SE). The red color represents a maximum velocity magnitude of approximately 1 mm/s. The solid black lines indicate the streamlines emanating from the analyte channel. At the channel cross-section, the black streamlines become focused toward the center of the channel. This design and its operation were previously described.9 Briefly, the AgNPs diffuse little on account of their size and mass, while the analyte in the middle channel diffuse into the silver colloid stream adsorbing on the nanoparticles. The LiCl, that acts as an AgNP aggregating agent, must diffuse across the middle channel before encountering the (now analyte-covered) nanoparticles, causing them to form aggregates that produce intense SERS signals. LiCl has previously been identified as yielding an order of magnitude increase in enhancement over NaCl.18 This scheme leads to highly controllable signal generation, governed solely by the relative rates of diffusion of the analyte molecules, the silver nanoparticles, and the ions derived from the aggregating salt, resulting in a gradually increasing SERS signal along the flow-wise direction of the channel. By properly designing the three channel microfluidic, ample time is provided for the analyte to diffuse into the nanoparticle channel and adsorb on the nanoparticles, while downstream the LiCl salt,19 which must cross two interfaces, leads to the nanoparticle aggregation that produces the strong SERS signals. The optimal location along the channel at which the SERS intensity is maximum, once found, remains constant provided the same flow rates and channel configurations are used. Turbulent mixing, which is often used to prepare samples for SERS, does not result in such controllable and reproducible assay conditions, responsible, in part, for the less than ideal reproducibility often encountered in SERS. The microfluidic flow is driven by vacuum (∼700 Torr) applied to the outlet. The relative rates of flow in the respective channels are managed by serpentine channels. The microfluidic device was fabricated using previously described micro-

Figure 1. Molecular structures of four of the alkaloids commonly found in the latex of the poppy plant. A major difference distinguishing papaverine from the opioids is the presence of an isoquinoline ring (highlighted) which produces a characteristic Raman signature. The nitrogen on the isoquinoline is the locus of papaverine’s weak acidic properties.

papaverine but not as an assay, rather, used SERS in place of mass spectrometry for species identification following highperformance liquid chromatography (HPLC).15 Chromatography and mass spectrometry are expensive and not easily field deployable. What we report is a SERS-based assay coupled with microfluidics that is able to detect papaverine at low concentration in a fieldable device. The microfluidic detection platform presented here is ideal for field-use as it is small, inexpensive, and yields highly reproducible results. The reagents used are readily available, are relatively inexpensive, and do not require sensitive storage and handling conditions. Miniaturized Raman systems have already been developed by several companies, e.g., Snowy Range Instruments which produce hand-held and small benchtop instruments compatible with microfluidic devices. Here we propose the molecule papaverine as an analyte that occurs exclusively in the opium poppy. Most of the papaverine in opium latex does not survive typical opium processing and occurs at negligible concentrations in the finished opiate.17 Although papaverine itself is not a narcotic, its presence on skin or clothes would be prima facie evidence that the individual so identified had handled opium poppies or their latex. Below we describe a sensing technique based on combining surface enhance Raman spectroscopy with microfluidics as a rapid and discriminating sensor for papaverine implicating, prima facie,

Figure 2. Simulation of the hydrodynamics in the flow-merging device showing velocity lines of fluid flow. Laminar flow is achieved resulting in diffusion-dominated processes that cause the analyte and the lithium chloride to cross sequentially into the stream containing the AgNPs, the former leading to analyte adsorption and the latter causing AgNP aggregation. B


Article


Figure 3. SERS spectra obtained of aqueous papaverine HCl solutions of the indicated concentrations. Other than the reference spectrum (Ref.) the SERS spectra presented were collected from the microfluidic flow-merging device using identical interrogation and collection schemes. The red dashed lines indicate the Raman shifts that reoccur reproducibly in all of the spectra and are assigned to papaverine. The spectra shown are averaged and each normalized to the intensity of the most intense band in the spectrum.

fingers. The lower detection limit was experimentally found to be 6 μM. Because the solution was produced from the HCl salt of papaverine, the concentration of free base papaverine in solution is approximately 2% lower than the nominal values. Also the HCl salt releases H+ ions when dissolved producing solution of varying pH as a function of the papaverine HCl concentration such that solutions with concentrations of 3000, 300, 30, 15, and 6 μM had measured pH values of 4.23, 4.95, 5.44, 5.6, and 5.66, respectively. (We show below that these values suggest a pKa value of 6.0 for papaverine HCl, which agrees with an estimated pKa value of 6.03 (ChemAxon) for papaverine HCl but disagrees with other reported values.23 An ionic solution of 0.2 M LiCl was used as aggregating agent for all solutions except for those with concentrations of 3000 and 300 μM, which were found to produce a great deal of AgNP aggregation on their own, presumably on account of the “high” concentration of Cl− ions originating from the papaverine salt. Upon successful loading of the device with all reagents (20 μL each), the microfluidic device is mounted to the spectrometer stage and vacuum is applied to the outlet to induce flow. Spectra were collected using a confocal microRaman system (LabRam Aramis spectrometer (Horiba, Kyoto, JP)), with 0.84 mW of 633 nm laser, a 50× objective lens (Olympus MPLN50×), and 1 s acquisition time per point while

electromechanical systems (MEMS)-based lithographic processes.20−22 Briefly, a mold was created on a 4 in. silicon wafer using SU-8 photoresist. PDMS (Sylgard 184, Dow-Corning) was cast upon the mold and peeled off to yield microchannels. The cross-sectional area of the channels was 20 μm deep × 50 μm wide. The flow velocity was found to be approximately 1 mm/s, yielding a total volumetric flow rate of 1 μL/s. The PDMS pieces containing the microchannels were sealed with 200 μm thick Pyrex microscope coverslips (Fisher Scientific). Both PDMS and glass were ozonated (UVO-Cleaner 42, Jelight Company Inc.) prior to bonding to create a robust and permanent seal. Approximately 20 nm diameter citrate capped AgNPs (nanoComposix, product AGCB20-1M) in colloidal suspension were diluted 1:100 from 1 mg/mL stock solution. A concentrated papaverine HCl (MP Biomedicals, SKU 02190261) solution was prepared by dissolving as much papaverine as possible in room temperature DI water, which produces a solution of approximately 40 mM. From this stock solution, papaverine HCl solutions of 1000, 100, 10, 5, and 2 ppm, corresponding to 3000, 300, 30, 15, and 6 μM, respectively, were prepared. We envision various methods for sample acquisition including the use of a solvent-moistened sterile cotton swab to remove residues from the hand and C


Article

Analytical Chemistry the device was under flow. An optimal interrogation area of 60 μm × 250 μm located 100 μm downstream from the junction at a depth 10 μm was found for all concentrations of analyte. This interrogation scheme proved to be an optimal region for SERS signal and also allowed for quick generation of data (within 2.5 min) by rasterization of the region in the microchannel. The rasterized area yielded 150 points per acquisition. Each concentration of analyte was run in triplicate, interrogating the device four times in 2.5 min intervals. The cumulative data points for each concentration at 5 min of flow were averaged to yield the spectra in Figure 3.

papaverine.25 An even lower amount of sample may be necessary as the papaverine content of opium can be as great as 9% w/w.17 Other studies have reported detection limits as low as 50 ng utilizing a complex reverse phase liquid chromatography system consisting of multiple pumps, gradient makers, and UV−vis detectors.14 Peaks near 1614 and 1401 cm−1 also occur in the spectra of opiates and have been assigned to various phenyl ring stretching modes, while the peaks 369 cm−1, 730 cm−1, and 769 cm−1 do not occur in opiates. The single peak at 369 cm−1 and the doublet peak at 730 and 769 cm−1 have both been previously assigned to modes associated with the isoquinoline ring in papaverine.11 The relative intensities of some SERS papaverine bands associated with its isoquinoline ring appear to be dependent on the analyte’s concentration. We ascribe this to the varying ratio of the protonated and deprotonated forms of papaverine due to the change in solution pH with varying analyte concentration (the higher the concentration, the lower the pH). Specifically, the intensity ratio of the peaks at 730 and 769 cm−1 vary systematically as the papaverine concentration is varied. The area of each of the two overlapping SERS bands was determined as a function of pH by fitting a Lorentzian to each of the overlapping SERS features (ipf.m, O’Haver, University of Maryland). We find (Figure 3) that the ratio of the 730 cm−1 to 769 cm−1 band intensities increases as the concentration of papaverine decreases, implying that the 730 cm−1 band belongs to the deprotonated form of the papaverine and implies that papaverine behaves as a weak acid. A straightforward weak acid analysis was carried out (reported in detail in the Supporting Information) which yielded a calculated pH value at each value of concentration assuming various parametric values of pKa. A selection of the results is shown in Figure 4, where we show that the measured pH values of papaverine HCl dissolved in DI water agrees well with what we calculate assuming pKa = 6. This accords with one

■

RESULTS AND DISCUSSION Various reference and control spectra were measured. Flowing DI H2O through the center (“analyte”) channel produced spectra that contained only PDMS Raman bands. Raman spectra of solid papaverine HCl powder and concentrated papaverine HCl dissolved in a AgNP solution agreed well with each other as well as with previously reported Raman spectra.11,24 For example, Figure 3 shows the Raman spectrum of 40 mM papaverine HCl dissolved in a 1:1 ratio with 0.1 mg/ mL AgNP solution (denoted as “Ref.”). A droplet of the solution was placed on a glass slide and was interrogated at a single point with a 633 nm laser at 3.8 mW for 1 s. While the SERS peaks are consistent with those measured for the various concentrations of papaverine HCl, this reference sample produced a spectrum approximately 20 times more intense than what we measured for the 3000 μM solution flowed in the device. Such a large difference in intensity was due the reference solution being much more concentrated as well as the necessary attenuation of the laser power in the device, as papaverine decomposes under high laser powers. Analysis of the raw spectra shown in Figure 3 involved the averaging of counts from different device runs at the same concentration of analyte as well as the normalization of each spectra to the intensity of the most intense band in the range of 300 cm−1 to 1700 cm−1. At all concentrations of papaverine, the most intense band occurred at 730 cm−1. Although the SERS experiments were carried out in a PDMS device, the Raman peaks of PDMS were normally much weaker than the intensities of the most intense papaverine SERS bands of 369, 730, 769, 1401, and 1614 cm−1. These correspond well with previously reported and assigned papaverine bands.6,11 Many of these are also easily distinguished from known features in the spectra of opiates such as morphine, codeine, and hydrocodone4 (see Figure 1) allowing one to detect papaverine in the presence of other components of the latex which would also likely be present in samples collected from a suspected poppy or latex handler. We stress that opiates such as hydrocodone, codeine, and morphine are structurally different from papaverine; hence their Raman spectra can be easily distinguished from that of papaverine even in a mixture. For example, papaverine has well resolved features at 369 cm−1, 730 cm−1, and 769 cm−1, which are not shared with any of the opiates. The opiates all have rather similar Raman spectra with bands at or near 445 cm−1, 630 cm−1, and 1450 cm−1, which papaverine does not share.4 What is special about papaverine is the fact that, although it is not a banned substance, its detected presence on a person strongly indicates contact with Papaver somniferum. The detection limit achieved is compatible with field deployment. Only 3 μg of latex sample from a suspect would suffice to create 20 μL of a 6 μM papaverine solution, assuming the typical opium sample contains 1.5% w/w

Figure 4. Measured (large red point) and calculated (lines) pH values of aqueous solution of papaverine hydrochloride as a function of concentration. Calculated values for various assumed values of pKa are shown. The measured pH values agree well with pKa = 6. Also shown are the log10 ratios of the intensities of the SERS bands at 730 cm−1 to 769 cm−1 (large blue point) as a function of concentration. The ratio decreases with decreasing solution pH indicating that the 730 cm−1 band belong to the protonated form of papaverine in solution, while the 769 cm−1 band belongs to the deprotonated form. On the basis of this analysis it is clear that details of the SERS spectrum of papaverine that is observed would vary systematically according to the substances that a suspect might have handled prior or subsequent to his/her coming into contact with the papaverine. D



■

ACKNOWLEDGMENTS We would like to thank Jason Guicheteau, Ralph Moon, and Steve Christensen for their many helpful discussions. This research was supported by the Institute for Collaborative Biotechnologies through Grant W911NF-09-0001 from the U.S. Army Research Office. The content of the information does not necessarily reflect the position or the policy of the Government, and no official endorsement should be inferred.

of the previously estimated pKa values for papaverine HCl (ChemAxon) but contrasts with other reports.23 The aforementioned 730 cm−1 to 769 cm−1 band intensity ratio is shown in the Supporting Information section to depend on pH as follows: log10

SPH + [σP] = log10 − log10 K a + pH SP [σPH+]

(S7)

■

,in which SPH+, SP, σPH+, and σP, respectively, are the SERS intensities of the protonated and deprotonated forms of papaverine and the cross sections of the SERS bands of the protonated and deprotonated papaverine. Equation S7 predicts S + that log10 PH should track the pH with a slope of one if one S [σP] [σ PH+]

− log10 K a as an adjustable additive number,

whose value was found to be 4.9 (Figure 4) suggesting that log10

[σP] [σ PH+]

= −1.1. Figure 4 shows that the ratio does indeed

decrease with decreasing pH (and hence with increasing papaverine concentration) but less rapidly than predicted by the analysis, perhaps signaling the fact that more than two bands contribute to the SERS feature observed in the range 700−780 cm−1. Nevertheless, the self-consistency of these results strongly supports the overall picture we present for the dynamics and equilibria of papaverine hydrochloride in aqueous solution.

■

CONCLUSION Papaverine is proposed as a discriminating indicator of possible drug trafficking and production. SERS, in conjunction with a microfluidic platform that allows the detection of low concentration of papaverine, was successfully able to distinguish papaverine spectroscopically. The intensity ratio of the SERS bands at 730 and 769 cm−1 were shown to originate from the isoquinoline ring in papaverine. The pH dependence of that ratio was used to show that the former band belonged to the protonated form of the isoquinoline ring, while the latter indicated the unprotonated form.

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.analchem.6b03797.

■

REFERENCES

(1) Merlin, M. D. Econ. Bot. 2003, 57, 295−323. (2) Degenhardt, L.; Charlson, F.; Mathers, B.; Hall, W. D.; Flaxman, A. D.; Johns, N.; Vos, T. Addiction 2014, 109, 1320−1333. (3) Steinitz, M. S. Washington Quarterly 1985, 8, 141−153. (4) Rana, V.; Cañamares, M. V.; Kubic, T.; Leona, M.; Lombardi, J. R. J. Forensic Sci. 2011, 56, 200−207. (5) Yang, Y.; Li, Z.-Y.; Yamaguchi, K.; Tanemura, M.; Huang, Z.; Jiang, D.; Chen, Y.; Zhou, F.; Nogami, M. Nanoscale 2012, 4, 2663− 2669. (6) Ryder, A. G. Curr. Opin. Chem. Biol. 2005, 9, 489−493. (7) Farquharson, S.; Shende, C.; Sengupta, A.; Huang, H.; Inscore, F. Pharmaceutics 2011, 3, 425−439. (8) Andreou, C.; Mirsafavi, R.; Moskovits, M.; Meinhart, C. D. Analyst 2015, 140, 5003−5005. (9) Andreou, C.; Hoonejani, M. R.; Barmi, M. R.; Moskovits, M.; Meinhart, C. D. ACS Nano 2013, 7, 7157−7164. (10) Zerell, U.; Ahrens, B.; Gerz, P. Bulletin on Narcotics: Science in Drug Control: The Role of Laboratory and Scientific Expertise, Vol. LVII; United Nations Office on Drugs and Crime, 2005 . (11) Cinta Pinzaru, S.; Leopold, N.; Pavel, I.; Kiefer, W. Spectrochim. Acta, Part A 2004, 60, 2021−2028. (12) Paul, B. D.; Dreka, C.; Knight, E. S.; Smith, M. L. Planta Med. 1996, 62, 544−547. (13) Cordero, R.; Paterson, S. J. Chromatogr. B: Anal. Technol. Biomed. Life Sci. 2007, 850, 423−431. (14) Boberić-Borojević, D.; Radulović, D.; Ivanović, D.; Ristić, P. J. Pharm. Biomed. Anal. 1999, 21, 15−22. (15) Sägmüller, B.; Schwarze, B.; Brehm, G.; Trachta, G.; Schneider, S. J. Mol. Struct. 2003, 661−662, 279−290. (16) Bogusz, M. J.; Maier, R. D.; Erkens, M.; Kohls, U. J. Anal. Toxicol. 2001, 25, 431−438. (17) Zerell, U.; Ahrens, B.; Gerz, P. Bull. Narc. 2005, 57, 11−31. (18) Koo, T. W.; Chan, S.; Sun, L.; Su, X.; Zhang, J.; Berlin, A. A. Appl. Spectrosc. 2004, 58, 1401−1407. (19) Tang, L. Cem. Concr. Res. 1999, 29, 1469−1474. (20) Qin, D.; Xia, Y.; Whitesides, G. M. Nat. Protoc. 2010, 5, 491− 502. (21) Flores, A.; Wang, M. R. Soft Lithographic Fabrication of Micro Optic and Guided Wave Devices. In Lithography; InTech: Rijeka, Croatia, 2010; Chapter 19, DOI: 10.5772/8185 (22) Whitesides, G. M.; Ostuni, E.; Takayama, S.; Jiang, X.; Ingber, D. E. Annu. Rev. Biomed. Eng. 2001, 3, 335−373. (23) O’Neil, M. J., Ed. Merck Index: An Encyclopedia of Chemicals, Drugs, and Biologicals, 14th ed.; Merck: Whitehouse Station, NJ, 2006; p 1210 (24) Leopold, N.; Baena, J. R.; Bolboacǎ, M.; Cozar, O.; Kiefer, W.; Lendl, B. Vib. Spectrosc. 2004, 36, 47−55. (25) Kalant, H. Addiction 1997, 92, 267−277.

P

treats log10

Article

Derivation of weak acid analysis equations used for determination of results in Figure 4 (PDF)

AUTHOR INFORMATION

Corresponding Authors

*E-mail: [email protected]. *E-mail: [email protected]. ORCID

Augustus W. Fountain III: 0000-0003-2271-8999 Carl D. Meinhart: 0000-0003-0701-2728 Martin Moskovits: 0000-0002-0212-108X Notes

The authors declare no competing financial interest. E


Detection of Papaverine for the Possible Identification of Illicit Opium

Recommend Documents