Online Sol-gel Capillary Microextraction-Mass Spectrometry (CME-MS) Analysis of Illicit Drugs


Providing rapid and sensitive sample cleanup, sol-gel capillary microextraction (CME) is a form of solid phase microextraction (SPME). The capillary format of CME couples easily with mass spectrometry (MS) by employing sol-gel sorbent coatings in inexpensive fused silica capillaries. By leveraging the syringe pump and six-port valve readily available on the commercial MS, we can obviate the need for chromatography for samples as complex as urine in quantitative assays. Two different sol-gel materials were studied as microextraction sorbents: one with a single ligand of octadecyl (C18) and the other with a dual-ligand combination of C18 and phenyl (Phe) groups. The CME-MS method was optimized for flow rate and solvent desorption and studied for overall microextraction performance between the two sorbents studied. We extract illicit drugs including cocaine, heroin, amphetamine, methamphetamine, 3,4-methylenedioxymethamphetamine, and oxycodone, proving good run-to-run reproducibility (RSD% < 10%) and low detection limits (< 10 ng mL−1). The dual-ligand sorbent demonstrated superior performance due to typical hydrophobic properties of C18 as well as potential π-π interactions of the Phe functionality and the aromatic moiety common to many drugs. This study demonstrates the advantage of fine-tuning sol-gel sorbents for application-specific CME-MS. We apply our method to the analysis of various drugs in synthetic and human urine samples and show low carryover effect (~ 5%) and low matrix effect in the presence of the urine matrix. Thus, the sol-gel CME-MS technique described herein stands to be an attractive alternative to other SPME-MS techniques.


Since its inception in 1989 by Pawliszyn et al. [1], solid phase microextraction (SPME) has proven to be a highly effective sample preparation method. SPME has been coupled to various analytical techniques such as GC [2, 3], HPLC [4,5,6], capillary electrophoresis (CE) [7, 8], supercritical fluid chromatography (SFC) [9], and MS [9,10,11,12,13,14,15,16,17,18]. In particular, SPME-MS technologies [19] are attractive because they combine the rapid, inexpensive, selective sample preparation of SPME with the exquisite sensitivity and specificity of MS. If targeted development is pursued, resource-intensive chromatographic systems may be entirely avoided for some applications. However, unlike LC-MS, SPME-MS formats are not yet widely adopted for quantitative assays, because they are infrequently validated for reproducibility or repeatability, especially for implementations meant for single-use. High-throughput commercial LC-MS sequences routinely shunt early unretained constituents (e.g., salts) to waste using the native fluidic valve of the MS. Meanwhile, without such a valving mechanism, SPME-MS could be prone to MS contamination and, thus, instrument downtime. Finally, SPME is more vulnerable to matrix effects of co-desorbed species, especially with electrospray ionization (ESI). To alleviate some matrix effects, SPME has been coupled to many other ambient ionization techniques such as desorption electrospray ionization [12, 19,20,21,22,23], the commercial direct analysis in real time (DART) [24,25,26], desorption corona beam ionization [27], and dielectric barrier discharge ionization [28]. Notably, incorporating stable isotope-labeled internal standards can largely mitigate matrix effects in targeted quantitative applications.

Research has progressed beyond commercially available SPME fibers and toward new sorbents [29] and new geometries [19] which can be leveraged for targeted analyses. In particular, sol-gel capillary microextraction (CME), invented by Malik et al. [30], combines a sol-gel sorbent with an in-tube microextraction geometry. At the heart of CME platforms, fused silica capillaries provide a simple, inexpensive means to coupling with various analytical methods. Meanwhile, sol-gel fabrication imparts robust features of enhanced solvent stability through intrinsic chemical bonding between the fused silica and the sorbent. CME uniquely provides facile physical tunability of the sorbent morphology, including pore size and coating thickness. Furthermore, CME allows chemical tunability with vast numbers of novel sorbent types to provide enhanced selectivity, sensitivity, and microextraction performance. Currently, silica [31,32,33,34] and non-silica-based [35,36,37,38,39,40] materials are used with combinations of sol-gel active ligands [29, 41] yielding a vast array of hybrid organic–inorganic microextraction sorbents. Sol-gel-based CME has been applied to various target analytes in urine [36, 42], human serum [43], human plasma [44], saline [35], and water [45] samples, demonstrating a robust format against many matrices, even biological in nature. Remarkably, sol-gel CME has traditionally been coupled only to chromatographic systems.

In this work, without chromatography, we demonstrate coupling sol-gel-coated capillaries to an atmospheric pressure inlet MS (CME-MS) for simple quantitative assays. Sample is introduced by the MS’s native syringe pump, and the standard transfer line simply replaced with the CME, making deliberate use of the MS’s native six-port valve. We design and test two sol-gel coatings for the CME: (1) a single octadecyl ligand and (2) a dual-ligand including octadecyl and phenyl ligands. We also deliberately employ deuterium-isotope-labeled internal standards, as is common industry practice for LC-MS quantitation. Moreover, we employ only a single stage of mass spectrometry on a simple commercial but relatively inexpensive ion trap MS. Development of our CME-MS method is validated for online microextraction, desorption, and analysis of illicit drugs specifically from water and urine matrices. For the numerous runs performed, minimal matrix effects and carryover effects are observed. To our knowledge, this work is the first example of sol-gel CME-MS applications.



A Corning LSE vortex mixer (Corning Corporation, Corning, NY, USA) was used to properly mix the sol solution ingredients. For the centrifugation of the sol solutions, a Thermo IEC model Micromax microcentrifuge (Needham Heights, MA, USA) was utilized. The coated capillaries were conditioned in a Shimadzu (Tokyo, Japan) model GC-17A gas chromatograph. Infrared spectra were obtained from a Nicolet IR spectrometer (Thermo Fisher Scientific, USA), and a Hitachi-S800 (Hitachi, Tokyo, Japan) electron microscope was used for the scanning electron microscopy (SEM) investigations. Sol-gel CME-MS experiments were performed using an LTQ XL (Waltham, MA, USA) MS system. Xcalibur software (Thermo Scientific, San Jose, CA) was employed for the data acquisition, and OriginPro 2015 software (Originlab Corp., Northampton, MA) was used for plotting the graphs. Chemical structures were drawn in Chemdraw software (Perkin Elmer, USA).

Chemicals and Materials

Fused silica capillary (0.25 mm i.d.) was purchased from Polymicro Technologies (Phoenix, AZ, USA). Phenyltrimethoxysilane, octadecyltrimethoxysilane, and tetramethoxysilane (TMOS) (Table S1) were purchased from Gelest (Morrisville, PA, USA). Trifluoroacetic acid (TFA) was purchased from Acros (Morris Planes, NJ, USA). Drug standard solutions (Table S2) in 1000 mg L−1 and deuterated internal standards such as cocaine-d3, oxycodone-d3, (±)-3,4-methylenedioxymethamphetamine (MDMA)-d5, amphetamine-d3 were purchased from Cerilliant (Round Rock, TX, USA). LC-MS grade methanol, acetonitrile, isopropanol, ethanol, and water were purchased from Fisher Scientific (Fisher Scientific, Pittsburgh, PA, USA).

Preparation of Sol-gel-Coated Capillaries

Preparation of the sol-gel-coated capillaries is described in detail in the supplementary material, and a general representation of the process is provided in Figure 1. Briefly, the process consists of three main steps: (1) designing the sol-gel material, (2) coating and conditioning the coated capillaries, and (3) characterization. In this study, two types of silica-based sol-gel materials were designed: one with octadecyl ligands, the other with octadecyl and phenyl ligands using appropriate amounts (Table S1) of sol-gel precursors and co-precursor tetramethoxysilane (TMOS) with trifluoroacetic acid (TFA). The preparation of the dual-ligand sol-gel coating uses equimolar sol-gel co-precursors. Sol-gel material is prepared in microcentrifuge vials by vortexing sol-gel active components. While the mixture is still a sol, the vial is placed into a homemade filling system [46]. The fused silica capillary installed in the filling system is loaded with the sol by nitrogen pressure. After 30 min, the sol is removed and the capillary then purged with nitrogen. Subsequently, the coated capillary is conditioned in a GC oven in the presence of nitrogen and slow temperature programming.

Figure 1

General concept of CME-MS. Capillaries are prepared in (1)–(3) and ready for multiple rounds of microextraction and detection (4)

Experimental Setup and the Microextraction-Desorption-Detection Procedure

The CME experiments were carried out on a Thermo LTQ XL linear ion-trap system outfitted with the conventional electrospray ionization (ESI) source. The standard PEEK tubing between the ionization source and the syringe pump was replaced by the prepared 75-cm-long sol-gel-coated fused silica capillary and placed between the readily available Rheodyne valve of the MS system and the syringe. To accommodate the distinct extraction and desorption/detection steps, the load/inject function of the valve was employed to help automate the CME process because desorption could easily be initiated by switching the valve from the load to the inject position.

As seen in step 4 (CME-MS) of Figure 1, a syringe (250 μL) containing sample is connected to the coated fused silica capillary. The microextraction adsorption takes place by flowing the entire 250 μL sample through the capillary at a rate of 20 μL min−1 to a sample waste container after passing through the valve. In the desorption step, the sample syringe is simply replaced by a new syringe (syringe volume = 250 μL) filled with desorption solvent. To induce desorption, the valve is switched from “load” to “inject” (Figure 2), and the syringe pump begins to pump the desorption solvent at a rate of 100 μL min−1. The desorbed analytes are ionized through the ESI probe and then detected by the MS system, commonly eluting in under 3 min. After extraction of the urine samples, an extra cleaning step was added to the procedure to eliminate possible complications of the urine matrix. In this step, with the valve remaining in the load position, optima grade water flowed at a rate of 50 μL min−1 through the coated capillary for a total volume of 250 μL.

Figure 2

Usage of the six-port valve in load and inject positions for the CME-MS method

For each drug analysis, positive ion electrospray ionization with single ion monitoring (SIM) and full scan modes were utilized. The spray voltage was set to 5.0 kV and the capillary temperature was maintained at 275 °C. Nitrogen was utilized as a sheath gas set to 8 arbitrary units (AU). For full scan experiments, the mass range was kept at 150–400 m/z (except for amphetamine, which was 120–300 m/z). The number of microscans and the maximum inject time on the LTQ were set to 3 and 200 ms, respectively.

CME-MS of Illicit Drugs in Urine Media

The abovementioned synthetic urine samples were prepared based on Haddad’s patent [47] where 2 g L−1 potassium chloride, 2 g L−1 sodium sulfate, 0.85 g L−1 ammonium phosphate, 0.85 g L−1 ammonium diphosphate, 0.25 g L−1 calcium chloride and 0.5 g L−1 magnesium chloride, 75 mg dl−1 urea, and 5 mg dl−1 of creatinine used. Samples were prepared fresh before the CME-MS experiments. Post-mortem human urine samples were provided by the Hillsborough Medical Examiner’s Office (Tampa, FL, USA). Toxicology testing had previously verified the presence of illicit substances in these case subjects.

Synthetic urine samples were spiked with the target analytes (illicit drugs), and their deuterated isotopes. Calibration curves were built based on the ratio of the analyte/d-analyte signal and the concentration of the analyte. Urine samples were syringe-filtered and identical microextraction-desorption and detection procedures were followed for both synthetic urine and human urine samples. Human urine samples were spiked with the same concentration of deuterated analytes (10 ng mL−1) and, thus, quantified using the previously built calibration curves.

Carryover Studies in CME-MS Applications

A schematic representation of the carryover effect studies can be seen in Figure 3. To observe the possible carryover effects between CME-MS experiments, C18-Phe and C18 sol-gel-coated capillaries were tested for amphetamine, methamphetamine, cocaine, MDMA, heroin, and oxycodone (500 ng mL−1) in synthetic urine. Each analyte was run under identical conditions in triplicate as described in the experimental section. The signal from the microextraction/desorption was recorded after each run and later coated capillaries were washed with 50/50 methanol:water solution (250 μL at a rate of 50 μL min−1). Following that, only methanol was passed through the capillary in the inject position of the six-port valve and another spectrum was collected from the previously cleaned capillary. Analyte signals obtained after microextraction and signals obtained after a wash step (where no microextraction was performed) were used to calculate any potential carryover effect.

Figure 3

Carryover experiments were shown in six steps: 1. Extraction of drugs from urine medium; 2. Removal of urine salt by water wash; 3. Desorption of the drugs off the sol-gel coating; 4. Detection in MS; 5. Washing analyte residues, regenerating the coating; and 6. Testing the coating to observe any analyte remaining using methanol

Matrix Effect Experiments

Different matrices and their effects on the CME-MS performance were compared. In this study, methamphetamine, amphetamine, heroin, cocaine, oxycodone, and MDMA were used as test analytes and two different calibration curves were built. The first curve encompassed solutions of the illicit drugs, and their deuterated standards spiked in synthetic urine. The second curve encompassed standards spiking the same concentration of analyte and its internal standard but in LC-MS grade water. Equations of these calibration curves were used to calculate the matrix effect.

Results and Discussion

Creation and Characterization of Sol-gel Sorbents with Single and Dual-Ligands

Hydrolytic sol-gel reactions were performed in the presence of TFA serving as a catalyst. In the octadecyl-containing sorbent, C18- precursor underwent hydrolysis with co-precursor TMOS, eventually creating a 3D silica-based network. In the case of the dual-ligand-containing sorbent, equimolar amounts of phenyltrimethoxysilane and n-octadecyltrimethoxysilane precursors were hydrolyzed with TMOS. After preparation of the sols, the sorbents were created inside a fused silica capillary in one step in which the sol solutions transformed into gels upon chemical bonding between the silanol groups in the fused silica capillaries (Figure S1). Here, silanol groups were critically enriched prior to the coating following the pretreatment process [35], thus anchoring a uniform 3D sol-gel network to the fused silica capillaries. Polycondensation of the sol solution with the silanol groups evolves into the covalently bonded sol-gel coating. Consequently, octadecyl (C18-) or octadecyl-phenyl (C18-Phe) functionalized, surface bonded sol-gel coatings in fused silica capillaries were fabricated.

We validated incorporation of both the single (C18) and dual-ligand (C18-Phe)-containing sol-gel sorbents with FTIR spectra in Figure S2. Characteristic CH2 stretches (2921 cm−1 and 2852 cm−1) were found to be in agreement with the literature data [48] and exhibited higher intensity in the lone octadecyl ligand-containing sol-gel sorbent and lower intensity in the dual-ligand-containing sorbent. The peak at ~ 698 cm−1 in the C18-Phe sorbent spectrum, was attributed to the presence of a phenyl group and is absent in the spectrum of the C18- sol-gel material. Notably, the peak indicating Si–O–Si stretching, visible at 1044 cm−1, was observed in both spectra. The obtained infrared spectra thus indicate the successful creation of silica-based single and dual-ligand sol-gel materials with C18- and Phe- groups. SEM images were also collected from the single and dual-ligand sol-gel sorbents in capillaries and are depicted in Figure S3. It was confirmed that the sol-gel coating is present on the inner surface of the fused silica capillaries. Thickness measurements found that the dual-ligand and single ligand sorbent had thicknesses of 1.78 μm and 1.67 μm, respectively.

Optimization and Characterization of CME-MS

Microextraction Flow Rate Evaluation

CME is a dynamic extraction process where a sample solution containing target analytes flows through the fused silica capillary, during which analytes interact with the sorbent. Here, the flow was provided by the syringe pump of the LTQ XL system and we sought to determine an optimum flow rate for the microextraction. To achieve that, 100 ng mL−1 of scopolamine was used to test extraction using various flow rates in triplicate and 20 μL min−1 was chosen as the optimum flow rate. (Detailed information are available in supplemental material and Figure S4.)

Desorption of Analytes From Capillary After Microextraction

Desorption in CME applications refers to the use of organic solvents to further extract or elute the embedded analytes from the immobilized sorbent within the capillary. Four different solvents: methanol, ethanol, acetonitrile, and isopropanol were utilized under the same conditions to desorb the illicit drug targets. Initially, the most hydrophobic analyte among the used illicit drugs, cocaine, was tested with hydrophobic C18 coating (Figure S6). Considering C18-Phe is also hydrophobic in nature, we expected similar desorption performance from the solvents used.

As seen in Figure 4, for scopolamine, MDMA, heroin, and cocaine, methanol provided the highest signal intensity and isopropanol provided the lowest signal intensity. Notably, for the above-mentioned analytes, we observed a relationship between octanol-water coefficients (log P) of the alcohols and the signal intensities obtained after the extractions. When considering ethanol, methanol, and isopropanol, the highest log P is attributed to isopropanol (0.05) and the signals obtained after desorption by using this solvent were the lowest. On the other hand, methanol (log P = − 0.77) provided the highest signal intensities for such analytes. In other words, the lower hydrophobicity of the alcohol resulted in higher microextraction performance. Similarly, Hildebrand solubility parameters of the methanol (δ = 14.5) and isopropanol (δ = 11.5) also correlated with the observation that the difference in water solubility of these two solvents translated to microextraction performance. Interestingly, desorption with acetonitrile provided highly distorted peak shapes with higher baseline noise compared to alcohols.

Figure 4

CME-MS of illicit drugs (100 ng mL−1) in their aqueous solutions (n = 3), under identical conditions, by using various desorption solvents (a) ethanol; (b) isopropanol; (c) acetonitrile; and (d) methanol. The sol-gel coating used was C18

Reproducibility of the Sol-gel CME-MS in the Microextraction of Illicit Drugs

To characterize run-to-run reproducibility for the target analytes, each individual illicit drug in synthetic urine was prepared at a concentration of ~ 100 ng mL−1. The samples were then microextracted in triplicate; system conditions were kept identical in each run. From the obtained signal intensities (Figure S5), the RSD % values for the C18- sol-gel coating was 3.0 to 7.9%, and, for the dual-ligand coating, 3.7 to 9.2% (Table 1). Such low RSD % values demonstrate the excellent reproducibility of the CME-MS method and, also, the similarity in performance of SPME-MS and CME-HPLC or CME-GC applications [29]. Therefore, for routine analyses with MS, these sol-gel-coated capillaries appeared promising for on-line sample preparation prior to mass spectrometric detection.

Table 1 RSD% (Run-to-Run), Obtained Signal Intensity at 100 ng mL−1 Concentration of the Analytes in Synthetic Urine Were Shown (n = 3). The LODs Were Calculated Based on the Calibration Curves Built Using Spiked Synthetic Urine Solutions of the Analytes

The detection limits of the sol-gel CME-MS technique for the illicit drugs (Table 1) were less than 10 ng mL−1. Notably, the Substance Abuse and Mental Health Services Administration (SAMHSA) proposed cutoff values of 150 ng mL−1 for cocaine, 2000 ng mL−1 for opiates, 250 ng mL−1 for amphetamines, and 250 ng mL−1 for MDMA [49]. In addition, cutoff values provided by Certified Laboratory References (CLR) were also met in the current study [50]. Our detection limits were remarkably lower than the values required from SAMSHA and CLR, which indicates that our novel sol-gel CME-MS approach is a promising technique for rapid detection of illicit drugs. The obtained LOD values were also comparable with recent SPME-MS applications where less than 1 ng mL−1 of LOD were reported for cocaine, heroin, methamphetamine, and oxycodone [26, 51, 52]. Additionally, in a recent application of SPME-Transmission Mode (SPME-TM) system coupled with a portable mass spectrometer, for oxycodone, cocaine, and heroin, detection limits less than 25 ng mL−1 were reported [53]. This could indicate that the CME-MS technique in this study could potentially be integrated with portable MS systems in the field for forensic applications.

Sorbent Performance of the Sol-gel-Coated Capillaries

In the current study, to observe the microextraction differences of the C18- and C18-Phe sol-gel coatings, triplicate extractions of the illicit drugs were performed. As shown in Table 1, an average improvement of ~ 56% in the signal intensity from C18- single ligand to C18-Phe dual-ligand was observed. Here, simply replacing half of the C18- amount in the sol solution with phenyl group resulted in a more effective sol-gel coating with dual-ligands. Incorporation of phenyl groups in the newly created sol-gel sorbent enhanced microextraction performance on illicit drugs due to the addition of the π-π interactions between the sorbent and the analytes (Figure 5).

Figure 5

CME-MS of amphetamine with (a) dual-ligand sol-gel sorbent and (b) C18 ligand. One hundred nanograms per milliliter of amphetamine aqueous solution with 250 μL total volume was used. During microextraction, 20 μL min −1 of sample flow was utilized. Desorption with methanol was employed

Caffeine extraction was ~ 4 times greater for the dual-ligand sorbent compared to the single ligand and represented the highest improvement in the microextraction among all the analytes studied (see Table S2), presumably due to caffeine low hydrophobicity. In other words, decreasing the hydrophobic octadecyl ligand amount in the sol-gel coating resulted in higher extraction performance for caffeine. At this point, it is worth noting that CME-MS is an ideal platform to test ligand effects on microextraction performance because the resulting direct and rapid data collection is faster than traditional CME coupled with chromatographs.

Microextraction of Illicit Drugs From Urine Samples

For real-life applications of the techniques mentioned above using sol-gel CME, chromatographic methods also serve as cleanup systems before final detection by a diode array detector (LC) or a flame ionization detector (GC) or MS. In our application, we study human urine or synthetic urine, containing various salts among other components. Without proper sample cleanup, the presence of salts in a urine matrix generally lead to ion suppression in MS. Therefore, in the current study, detection quality, preconcentration-microextraction, and analysis heavily depend on the type of the sol-gel coating, microextraction process, and the system parameters.

Investigation of Carryover Effect in CME-MS

Considering that the carryover effect from previous CME-MS experiments could provide false positive results, we investigated the potential carryover effect. Briefly, we obtained the signal of the target analyte after a CME-MS run and subsequent wash step, by simply running neat methanol (Figure 3). The carryover effect, calculated according to Eq. (1), was obtained from the signal ratios for four illicit drugs (MDMA, oxycodone, cocaine, and heroin) with 500 ng mL−1 solutions in synthetic urine matrix. Both the dual-ligand sol-gel coating and the C18 coating were tested for carryover effects:

$$ \mathrm{Carryover}\ \mathrm{effect}\%=\frac{\mathrm{Signal}\ \mathrm{after}\kern0.5em \mathrm{washstep}}{\mathrm{Signal}\ \mathrm{after}\ \mathrm{CME}-\mathrm{MS}}\times 100 $$

Experimental results showed that for the abovementioned drug analytes, dual-ligand sol-gel coating showed carryover effects ranging from 1.8 to 5.6% while the C18 coating ranged from 2.5 to 6.8% (Figure S1). We noticed that only MDMA was over 5%, the highest carryover effect among all the test analytes (Figure 6). In summary, carryover effects were generally lower than 5% indicating reasonable resistance to carryover for the CME-MS technique. Based on these results, we can infer that the potential to obtain false positive results is low in CME-MS. More detailed carryover results for both coatings can be seen in in Table S3.

Figure 6

Carryover effect in the CME-MS applications. A sample of 500 ng mL−1 MDMA in urine was used for the tests. Dots represent the spectrum after a wash step and lines represent the spectrum after CME-MS procedure. This difference between chronograms shows that the wash step is effective in preventing false positive results

Evaluation of Matrix Effects in CME-MS

Detection and quantification of illicit drugs with LC-MS could be a challenging task when the sample matrix is urine [54]. Considering the lack of chromatography, evaluating the possibility of matrix effects (ME) in sol-gel CME-MS applications is crucial, because high ion suppression could affect the reliability and reproducibility of the results. To achieve that, ME were evaluated by using synthetic urine and water matrices. Calibration curves were plotted by using the solutions prepared by synthetic urine and water which were spiked by illicit drugs and their deuterated isotope standards. Observing the signal ratio changes between a more complex matrix and comparing it to a simpler matrix such as water can provide insight into the ME. Therefore, the slope of the best fit lines was obtained from the calibration curves (from both urine and water matrices) and input into Eq. (2) based on the methodology of Matuszewski and coworkers’ [55,56,57]:

$$ \mathrm{Matrix}\ \mathrm{effect}\%\frac{\mathrm{Slope}\ \mathrm{from}\ \mathrm{spiked}\kern0.5em \mathrm{urine}\ \mathrm{matrix}}{\mathrm{Slope}\ \mathrm{from}\ \mathrm{spiked}\ \mathrm{water}\ \mathrm{matrix}}\times 100 $$

Experimental data suggest that switching the matrix from water to urine caused a change in ratios of signalanalyte/signalstandard (Figure S8), and these changes were − 19.2% for methamphetamine, − 17.6% for heroin, + 16.8% for cocaine, − 11.9% for amphetamine, − 2.8% for oxycodone, and − 5.7% for MDMA, and in summary, all the variations were in ± 20% range. It should be noted that in SPME-LC-MS applications, typically 80–120% matrix effect is considered acceptable [54, 58]; therefore, the developed sol-gel CME-MS application provided promising results even with complex matrices without use of chromatography. The reproducibility of the experiments was also not affected with the urine medium, for which run-to-run (n = 3) RSD% values of less than 10% were recorded.

Quantitative Analysis of Illicit Drugs by CME-MS in Human Case Studies

For quantitative analysis of these illicit drugs in synthetic urine samples, calibration curves were built using signal ratios obtained for various concentrations of the target analytes and their deuterated isotopes. Addition of the deuterated isotopes of the drugs minimized the matrix effect in the quantification. Excellent R2 values (0.9909 to 0.9996) indicate the linearity of the sol-gel CME-MS in the microextraction of illicit drugs (Figure S9).

Tissue samples from ten deceased individuals had been previously screened by the Medical Examiner for various drugs. Five drugs in particular (methamphetamine, amphetamine, cocaine, oxycodone, and MDMA) had been detected in blood samples across these ten cases. Urine samples from these individuals had also been gathered and frozen but notably had not been tested for the presence of these substances. The urine samples were thawed and provided to our laboratory for testing by CME-MS. The case samples were individually spiked with the deuterated version of any analytes of interest, i.e., the drugs for which there was previously confirmed presence in the individual’s blood. Quantitative results of the urine by CME-MS are provided in Table 2 for four of the five analytes included in this study. While methamphetamine had been confirmed in the blood of cases 1, 2, and 3, it was not detected in the urine by CME-MS. Otherwise, the urine analyses were in agreement with the detected presence in blood. Because full scan MS data were collected for each case, we performed a cursory search for all the other drugs in our panel, but found no evidence. While numerous CME-MS runs from real human urine samples, we note that no blockage in the ESI was observed. Therefore, we may surmise that such blockages are sufficiently mitigated in the sample cleanup procedure employed herein for MS detection. Overall results obtained from the human urine samples show the potential of the sol-gel CME-MS technique in real-life applications.

Table 2 Illicit Drug Concentrations in Human Urine Samples and CME-MS Information. Samples Were Run in Triplicate and RSD% Values Were Calculated. Case Numbers Refer to Different People. Ratio Average Represents the Analyte Signal/d-Internal Standard Signal Obtained From Samples. Methamphetamine Was Not Detected in the Urine of Cases 1, 2, and 3 Despite Confirmed Presence in Blood


Direct coupling of sol-gel capillary microextraction with mass spectrometry (CME-MS) was successfully utilized to rapidly microextract illicit drugs. A simple sol-gel-coated-fused silica capillary was employed between a syringe pump and the six-port valve of an ion trap MS system. Two different sol-gel sorbents were studied: one with only octadecyl functionality, and the other with octadecyl-phenyl functionality. With the incorporation of aromatic groups in the structure of the sol-gel material, the microextraction performance visibly improved signal intensity obtained from the illicit drugs in a head-to-head comparison. The illicit drugs were additionally microextracted with good run-to-run and batch-to-batch reproducibility (< 10% RSD). Low nanograms per milliliter levels of detection limits were obtained and comparable to the reported literature values. Thus, the CME-MS technique was quite effective in quantifying the illicit drugs such as cocaine, oxycodone, amphetamine, and MDMA from human urine samples. Overall, we have demonstrated that the CME-MS technique, by leveraging the fine tuning properties of sol-gel materials, provides a highly effective yet inexpensive means to quantitative analysis where high-end LC-MS systems are unavailable. CME-MS conceivably lends itself to various analytical applications including environmental pollutants; other illicit drug classes; cancer biomarkers; and, additionally, with various sample media such as human blood plasma, saliva, or stomach fluids (Table 3).

Table 3 RSD% (Run-to-Run), Detection Limit Data and Obtained Signal Intensity at 100 ng mL−1 Concentration of the Analytes in Standard Solutions Were Shown. All Data Were Collected in Triplicate (n = 3) Under the Same Conditions


  1. 1.

    Belardi, R.P., Pawliszyn, J.B.: The application of chemically modified fused silica fibers in the extraction of organics from water matrix samples and their rapid transfer to capillary columns. Water Qual. Res. J. Can. 24, 179–191 (1989)

    Article  CAS  Google Scholar 

  2. 2.

    Zhang, Z., Pawliszyn, J.: Headspace solid-phase microextraction. Anal. Chem. 65, 1843–1852 (1993)

    Article  CAS  Google Scholar 

  3. 3.

    Rudnicka, J., Kowalkowski, T., Ligor, T., Buszewski, B.: Determination of volatile organic compounds as biomarkers of lung cancer by SPME–GC–TOF/MS and chemometrics. J. Chromatogr. B. 879, 3360–3366 (2011)

    Article  CAS  Google Scholar 

  4. 4.

    Chen, J., Pawliszyn, J.B.: Solid phase microextraction coupled to high-performance liquid chromatography. Anal. Chem. 67, 2530–2533 (1995)

    Article  CAS  Google Scholar 

  5. 5.

    Wang, C., Zhou, W., Liao, X., Zhang, W., Chen, Z.: An etched polyether ether ketone tube covered with immobilized graphene oxide for online solid phase microextraction of quaternary alkaloids prior to their quantitation by HPLC-MS/MS. Microchim. Acta. 184, 2715–2721 (2017)

    Article  CAS  Google Scholar 

  6. 6.

    Piri-Moghadam, H., Lendor, S., Pawliszyn, J.: Development of a biocompatible in-tube solid-phase microextraction device: a sensitive approach for direct analysis of single drops of complex matrixes. Anal. Chem. 88, 12188–12195 (2016)

    Article  CAS  PubMed  Google Scholar 

  7. 7.

    Whang, C.-W., Pawliszyn, J.: Solid phase microextraction coupled to capillary electrophoresis. Anal. Commun. 35, 353–356 (1998)

    Article  CAS  Google Scholar 

  8. 8.

    Zhang, W., Li, Y., Chen, Z.: Selective and sensitive determination of protoberberines by capillary electrophoresis coupled with molecularly imprinted microextraction. J. Sep. Sci. 38, 3969–3975 (2015)

    Article  CAS  PubMed  Google Scholar 

  9. 9.

    Lesellier, E.: Extraction and analysis of polycyclic aromatic hydrocarbons (PAHs) by solid phase micro-extraction/supercritical fluid chromatography (SPME/SFC). Analusis. 27, 363–368 (1999)

    Article  CAS  Google Scholar 

  10. 10.

    Fang, L., Deng, J., Yang, Y., Wang, X., Chen, B., Liu, H., Zhou, H., Ouyang, G., Luan, T.: Coupling solid-phase microextraction with ambient mass spectrometry: strategies and applications. TrAC Trends Anal. Chem. 85, 61–72 (2016)

    Article  CAS  Google Scholar 

  11. 11.

    Gómez-Ríos, G.A., Pawliszyn, J.: Solid phase microextraction (SPME)-transmission mode (TM) pushes down detection limits in direct analysis in real time (DART). Chem. Commun. 50, 12937–12940 (2014)

    Article  CAS  Google Scholar 

  12. 12.

    Kennedy, J.H., Aurand, C., Shirey, R., Laughlin, B.C., Wiseman, J.M.: Coupling desorption electrospray ionization with solid-phase microextraction for screening and quantitative analysis of drugs in urine. Anal. Chem. 82, 7502–7508 (2010)

    Article  CAS  PubMed  Google Scholar 

  13. 13.

    Deng, J., Yu, T., Yao, Y., Peng, Q., Luo, L., Chen, B., Wang, X., Yang, Y., Luan, T.: Surface-coated wooden-tip electrospray ionization mass spectrometry for determination of trace fluoroquinolone and macrolide antibiotics in water. Anal. Chim. Acta. 954, 52–59 (2017)

    Article  CAS  PubMed  Google Scholar 

  14. 14.

    Mirabelli, M.F., Wolf, J.-C., Zenobi, R.: Direct coupling of solid-phase microextraction with mass spectrometry: sub-pg/g sensitivity achieved using a dielectric barrier discharge ionization source. Anal. Chem. 88, 7252–7258 (2016)

    Article  CAS  PubMed  Google Scholar 

  15. 15.

    Gómez-Ríos, G.A., Tascon, M., Reyes-Garcés, N., Boyacı, E., Poole, J.J., Pawliszyn, J.: Rapid determination of immunosuppressive drug concentrations in whole blood by coated blade spray-tandem mass spectrometry (CBS-MS/MS). Anal. Chim. Acta. 999, 69–75 (2018)

    Article  CAS  PubMed  Google Scholar 

  16. 16.

    Tascon, M., Gómez-Ríos, G.A., Reyes-Garcés, N., Poole, J., Boyacı, E., Pawliszyn, J.: Ultra-fast quantitation of voriconazole in human plasma by coated blade spray mass spectrometry. J. Pharm. Biomed. Anal.

  17. 17.

    Yang, B.-C., Fang, S.-F., Wan, X.-J., Luo, Y., Zhou, J.-Y., Li, Y., Li, Y.-J., Wang, F., Huang, O.-P.: Quantification of monohydroxylated polycyclic aromatic hydrocarbons in human urine samples using solid-phase microextraction coupled with glass-capillary nanoelectrospray ionization mass spectrometry. Anal. Chim. Acta.

  18. 18.

    Deng, J., Yang, Y., Wang, X., Luan, T.: Strategies for coupling solid-phase microextraction with mass spectrometry. TrAC Trends Anal. Chem. 55, 55–67 (2014)

    Article  CAS  Google Scholar 

  19. 19.

    Reyes-Garcés, N., Gionfriddo, E., Gómez-Ríos, G.A., Alam, M.N., Boyacı, E., Bojko, B., Singh, V., Grandy, J., Pawliszyn, J.: Advances in solid phase microextraction and perspective on future directions. Anal. Chem. 90, 302–360 (2018)

    Article  CAS  PubMed  Google Scholar 

  20. 20.

    D’Agostino, P.A., Chenier, C.L., Hancock, J.R., Lepage, C.R.: Jackson: desorption electrospray ionisation mass spectrometric analysis of chemical warfare agents from solid-phase microextraction fibers. Rapid Commun. Mass Spectrom. 21, 543–549 (2007)

    Article  CAS  Google Scholar 

  21. 21.

    D’Agostino, P.A., Chenier, C.L.: Desorption electrospray ionization mass spectrometric analysis of organophosphorus chemical warfare agents using ion mobility and tandem mass spectrometry. Rapid Commun. Mass Spectrom. 24, 1617–1624 (2010)

    Article  CAS  PubMed  Google Scholar 

  22. 22.

    Huang, G., Chen, H., Zhang, X., Cooks, R.G., Ouyang, Z.: Rapid screening of anabolic steroids in urine by reactive desorption electrospray ionization. Anal. Chem. 79, 8327–8332 (2007)

    Article  CAS  PubMed  Google Scholar 

  23. 23.

    Strittmatter, N., Düring, R.-A., Takáts, Z.: Analysis of wastewater samples by direct combination of thin-film microextraction and desorption electrospray ionization mass spectrometry. Analyst. 137, 4037–4044 (2012)

    Article  CAS  PubMed  Google Scholar 

  24. 24.

    LaPointe, J., Musselman, B., O’Neill, T., Shepard, J.R.E.: Detection of “bath salt” synthetic cathinones and metabolites in urine via DART-MS and solid phase microextraction. J. Am. Soc. Mass Spectrom. 26, 159–165 (2015)

    Article  CAS  PubMed  Google Scholar 

  25. 25.

    Wang, X., Li, X., Li, Z., Zhang, Y., Bai, Y., Liu, H.: Online coupling of in-tube solid-phase microextraction with direct analysis in real time mass spectrometry for rapid determination of triazine herbicides in water using carbon-nanotubes-incorporated polymer monolith. Anal. Chem. 86, 4739–4747 (2014)

    Article  CAS  PubMed  Google Scholar 

  26. 26.

    Rodriguez-Lafuente, A., Mirnaghi, F.S., Pawliszyn, J.: Determination of cocaine and methadone in urine samples by thin-film solid-phase microextraction and direct analysis in real time (DART) coupled with tandem mass spectrometry. Anal. Bioanal. Chem. 405, 9723–9727 (2013)

    Article  CAS  PubMed  Google Scholar 

  27. 27.

    Li, X., Wang, H., Sun, W., Ding, L.: Desorption corona beam ionization coupled with a poly (dimethylsiloxane) substrate: broadening the application of ambient ionization for water samples. Anal. Chem. 82, 9188–9193 (2010)

    Article  CAS  PubMed  Google Scholar 

  28. 28.

    Almasian, M.R., Yang, C., Xing, Z., Zhang, S., Zhang, X.: Development of a graphite low-temperature plasma source with dual-mode in-source fragmentation for ambient mass spectrometry. Rapid Commun. Mass Spectrom. 24, 742–748 (2010)

    Article  CAS  PubMed  Google Scholar 

  29. 29.

    Amiri, A.: Solid-phase microextraction-based sol–gel technique. TrAC Trends Anal. Chem. 75, 57–74 (2016)

    Article  CAS  Google Scholar 

  30. 30.

    Bigham, S., Medlar, J., Kabir, A., Shende, C., Alli, A., Malik, A.: Sol−gel capillary microextraction. Anal. Chem. 74, 752–761 (2002)

    Article  CAS  PubMed  Google Scholar 

  31. 31.

    Kulkarni, S., Fang, L., Alhooshani, K., Malik, A.: Sol–gel immobilized cyano-polydimethylsiloxane coating for capillary microextraction of aqueous trace analytes ranging from polycyclic aromatic hydrocarbons to free fatty acids. J. Chromatogr. A. 1124, 205–216 (2006)

    Article  CAS  PubMed  Google Scholar 

  32. 32.

    Zeng, J., Chen, J., Lin, Z., Chen, W., Chen, X., Wang, X.: Development of polymethylphenylsiloxane-coated fiber for solid-phase microextraction and its analytical application of qualitative and semi-quantitative of organochlorine and pyrethroid pesticides in vegetables. Anal. Chim. Acta. 619, 59–66 (2008)

    Article  CAS  PubMed  Google Scholar 

  33. 33.

    Lopes, A.L., Augusto, F.: Preparation and characterization of polydimethylsiloxane/poly (vinylalcohol) coated solid phase microextraction fibers using sol–gel technology. J. Chromatogr. A. 1056, 13–19 (2004)

    Article  CAS  PubMed  Google Scholar 

  34. 34.

    da Costa Silva, R.G., Augusto, F.: Sol–gel molecular imprinted ormosil for solid-phase extraction of methylxanthines. J. Chromatogr. A. 1114, 216–223 (2006)

    Article  CAS  PubMed  Google Scholar 

  35. 35.

    Seyyal, E., Malik, A.: Silica- and germania-based dual-ligand sol-gel organic-inorganic hybrid sorbents combining superhydrophobicity and π-π interaction. The role of inorganic substrate in sol-gel capillary microextraction. Anal. Chim. Acta. 964, 96–111 (2017)

    Article  CAS  PubMed  Google Scholar 

  36. 36.

    Alhendal, A., Mengis, S., Matthews, J., Malik, A.: Nonhydrolytic sol-gel approach to facile creation of surface-bonded zirconia organic-inorganic hybrid coatings for sample preparation. Ι. Capillary microextraction of catecholamine neurotransmitters. J. Chromatogr. A. 1468, 23–32 (2016)

    Article  CAS  PubMed  Google Scholar 

  37. 37.

    Tran, M., Turner, E.B., Segro, S.S., Fang, L., Seyyal, E., Malik, A.: Tantala-based sol-gel coating for capillary microextraction on-line coupled to high-performance liquid chromatography. J. Chromatogr. A. 1522, 38–47 (2017)

    Article  CAS  PubMed  Google Scholar 

  38. 38.

    Sheshanka, K., Abdul, M.: Sol–gel niobia sorbent with a positively charged octadecyl ligand providing enhanced enrichment of nucleotides and organophosphorus pesticides in capillary microextraction for online HPLC analysis. J. Sep. Sci. 0, (2017).

  39. 39.

    Kim, T.-Y., Alhooshani, K., Kabir, A., Fries, D.P., Malik, A.: High pH-resistant, surface-bonded sol–gel titania hybrid organic–inorganic coating for effective on-line hyphenation of capillary microextraction (in-tube solid-phase microextraction) with high-performance liquid chromatography. J. Chromatogr. A. 1047, 165–174 (2004)

    Article  CAS  PubMed  Google Scholar 

  40. 40.

    Fang, L., Kulkarni, S., Alhooshani, K., Malik, A.: Germania-based, sol−gel hybrid organic−inorganic coatings for capillary microextraction and gas chromatography. Anal. Chem. 79, 9441–9451 (2007)

    Article  CAS  PubMed  Google Scholar 

  41. 41.

    Kabir, A., Furton, K.G., Malik, A.: Innovations in sol-gel microextraction phases for solvent-free sample preparation in analytical chemistry. TrAC Trends Anal. Chem. 45, 197–218 (2013)

    Article  CAS  Google Scholar 

  42. 42.

    Bagheri, H., Piri-Moghadam, H., Bayat, P., Balalaie, S.: Application of sol–gel based molecularly imprinted xerogel for on-line capillary microextraction of fentanyl from urine and plasma samples. Anal. Methods. 5, 7096–7101 (2013)

    Article  CAS  Google Scholar 

  43. 43.

    Chen, B., Peng, H., Zheng, F., Hu, B., He, M., Zhao, W., Pang, D.: Immunoaffinity monolithic capillary microextraction coupled with ICP-MS for immunoassay with quantum dot labels. J. Anal. At. Spectrom. 25, 1674–1681 (2010)

    Article  CAS  Google Scholar 

  44. 44.

    Mu, L., Hu, X., Wen, J., Zhou, Q.: Robust aptamer sol–gel solid phase microextraction of very polar adenosine from human plasma. J. Chromatogr. A. 1279, 7–12 (2013)

    Article  CAS  PubMed  Google Scholar 

  45. 45.

    Sarafraz-Yazdi, A., Vatani, H.: A solid phase microextraction coating based on ionic liquid sol–gel technique for determination of benzene, toluene, ethylbenzene and o-xylene in water samples using gas chromatography flame ionization detector. J. Chromatogr. A. 1300, 104–111 (2013)

    Article  CAS  PubMed  Google Scholar 

  46. 46.

    Hayes, J.D., Malik, A.: Sol–gel chemistry-based Ucon-coated columns for capillary electrophoresis. J. Chromatogr. B Biomed. Sci. Appl. 695, 3–13 (1997)

    Article  CAS  PubMed  Google Scholar 

  47. 47.

    Haddad, L.: Synthetic urine and method of making same,, (2006). Accessed 02 April 2018

  48. 48.

    Ou, D.L., Seddon, A.B.: Near- and mid-infrared spectroscopy of sol–gel derived ormosils: vinyl and phenyl silicates. J. Non-Cryst. Solids. 210, 187–203 (1997)

    Article  CAS  Google Scholar 

  49. 49.

    Analytes and their cutoffs, Accessed 19 July 2018

  50. 50.

    Clinical Laboratory Reference. Cutoff and toxicity levels for drugs-of-abuse testing, Accessed 19 July 2018

  51. 51.

    Piri-Moghadam, H., Ahmadi, F., Gómez-Ríos, G.A., Boyacı, E., Reyes-Garcés, N., Aghakhani, A., Bojko, B., Pawliszyn, J.: Fast quantitation of target analytes in small volumes of complex samples by matrix-compatible solid-phase microextraction devices. Angew. Chem. Int. Ed. 55, 7510–7514.

  52. 52.

    Tascon, M., Gómez-Ríos, G.A., Reyes-Garcés, N., Poole, J., Boyacı, E., Pawliszyn, J.: High-throughput screening and quantitation of target compounds in biofluids by coated blade spray-mass spectrometry. Anal. Chem. 89, 8421–8428 (2017)

    Article  CAS  PubMed  Google Scholar 

  53. 53.

    Augusto Gómez-Ríos, G., Vasiljevic, T., Gionfriddo, E., Yu, M., Pawliszyn, J.: Towards on-site analysis of complex matrices by solid-phase microextraction-transmission mode coupled to a portable mass spectrometer via direct analysis in real time. Analyst. 142, 2928–2935 (2017)

    Article  Google Scholar 

  54. 54.

    Boyacı, E., Gorynski, K., Rodriguez-Lafuente, A., Bojko, B., Pawliszyn, J.: Introduction of solid-phase microextraction as a high-throughput sample preparation tool in laboratory analysis of prohibited substances. Anal. Chim. Acta. 809, 69–81 (2014)

    Article  CAS  PubMed  Google Scholar 

  55. 55.

    Matuszewski, B.K.: Standard line slopes as a measure of a relative matrix effect in quantitative HPLC–MS bioanalysis. J. Chromatogr. B. 830, 293–300 (2006)

    Article  CAS  Google Scholar 

  56. 56.

    Matuszewski, B.K., Constanzer, M.L., Chavez-Eng, C.M.: Strategies for the assessment of matrix effect in quantitative bioanalytical methods based on HPLC−MS/MS. Anal. Chem. 75, 3019–3030 (2003)

    Article  CAS  PubMed  Google Scholar 

  57. 57.

    Isaguirre, A.C., Olsina, R.A., Martinez, L.D., Lapierre, A.V., Cerutti, S.: Development of solid phase extraction strategies to minimize the effect of human urine matrix effect on the response of carnitine by UPLC–MS/MS. Microchem. J. 129, 362–367 (2016)

    Article  CAS  Google Scholar 

  58. 58.

    Mirnaghi, F.S., Pawliszyn, J.: Development of coatings for automated 96-blade solid phase microextraction-liquid chromatography–tandem mass spectrometry system, capable of extracting a wide polarity range of analytes from biological fluids. J. Chromatogr. A. 1261, 91–98 (2012)

    Article  CAS  PubMed  Google Scholar 

Download references


We are grateful to Ms. Dina Swanson of the Hillsborough County Medical Examiner’s Office for provision of the case samples. Furthermore, E.S. thanks Timothy Vazquez, Nathan Grimes, Benjimen Batazhan, and Merve Kacar for valuable help in the preparation of the manuscript.

Author information



Corresponding author

Correspondence to Theresa Evans-Nguyen.

Electronic supplementary material


(DOCX 2438 kb)

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Seyyal, E., Evans-Nguyen, T. Online Sol-gel Capillary Microextraction-Mass Spectrometry (CME-MS) Analysis of Illicit Drugs. J. Am. Soc. Mass Spectrom. 30, 595–604 (2019).

Download citation


  • Solid phase microextraction
  • Sample preparation
  • Urine analysis
  • Forensics
  • Drugs
  • Sol-gel