Real Time Monitoring of Containerless Microreactions in Acoustically

Aug 9, 2016 - (32) The microdroplet fusion method attributes the observed reaction acceleration mostly to droplet evaporation, resulting in extreme pH...
1 downloads 3 Views 3MB Size
Technical Note pubs.acs.org/ac

Real Time Monitoring of Containerless Microreactions in Acoustically Levitated Droplets via Ambient Ionization Mass Spectrometry Elizabeth A. Crawford,† Cemal Esen,‡ and Dietrich A. Volmer*,† †

Institute of Bioanalytical Chemistry, Saarland University, 66123 Saarbrücken, Germany Department of Mechanical Engineering, Ruhr-University Bochum, 44801 Bochum, Germany



S Supporting Information *

ABSTRACT: Direct in-droplet (in stillo) microreaction monitoring using acoustically levitated micro droplets has been achieved by combining acoustic (ultrasonic) levitation for the first time with real time ambient tandem mass spectrometry (MS/MS). The acoustic levitation and inherent mixing of microliter volumes of reactants (3 μL droplets), yielding total reaction volumes of 6 μL, supported monitoring the acidcatalyzed degradation reaction of erythromycin A. This reaction was chosen to demonstrate the proof-of-principle of directly monitoring in stillo microreactions via hyphenated acoustic levitation and ambient ionization mass spectrometry. The microreactions took place completely in stillo over 30, 60, and 120 s within the containerless stable central pressure node of an acoustic levitator, thus readily promoting reaction miniaturization. For the evaluation of the miniaturized in stillo reactions, the degradation reactions were also carried out in vials (in vitro) with a total reaction volume of 400 μL. The reacted in vitro mixtures (6 μL total) were similarly introduced into the acoustic levitator prior to ambient ionization MS/MS analysis. The in stillo miniaturized reactions provided immediate real-time snap-shots of the degradation process for more accurate reaction monitoring and used a fraction of the reactants, while the larger scale in vitro reactions only yielded general reaction information. coustic (ultrasonic) levitation as first described in 1933 by Bücks and Müller1 for levitation of small samples results from the stable pressure nodes generated by acoustic standing waves. The technique has emerged as the most practical method of object levitation among other levitation techniques such as magnetic, optical, electrostatic, and aerodynamic levitation.2,3 Ultrasound waves are capable of levitating even heavy tungsten balls and provide a “contact-free method” of holding samples in midair.4 Until now, ultrasound in analytical chemistry has been primarily associated with sample preparation steps, such as digestions, homogenization, dissolution, degassing of solutions, or general sonication based cleaning processes. Lesser known analytical applications of ultrasound include the coupling of liquid samples with analytical detection systems as reviewed by Priego-Capote and Luque de Castro.5 The authors categorized, ultrasound-assisted analytical couplings primarily into three main subcategories: ultrasonic nebulization, sample levitation, and assistance with electroanalytical techniques, with acoustic levitation presented as the most practical means of sample levitation. Liquid and solid samples can be acoustically levitated and this particular method of levitation can be applied to a wide range of sample cross sections (20 μm−2.5 mm) and easily achieved regardless of specific analyte properties, such as conductivity or magnetic traits. Most liquid samples are levitated via acoustic levitation if they exhibit sufficient specific density (ρs) and surface tension (σs).6

A

© XXXX American Chemical Society

The coupling of acoustic levitation with various analytical detection systems (termed “airborne” analytical systems) enabled the practical and unconstrained use of samples in the micro and nanoliter range for direct analytical measurements.7 The natural containerless environment created by the acoustic field provides an ideal setting for miniaturization of reactions and workflows, because major problems of poor surface-tovolume ratios and potential adsorption of analytes to container walls that are encountered in most miniaturized experiments are clearly avoided. Typically, nanoliter and microliter liquid volumes can be levitated in acoustic systems operating at the standard 58 kHz frequency. Advances in dispensing systems have enabled acoustic,8 piezoelectric,9 and electrokinetic10,11 injection of low nanoliter to picoliter volumes. Remote detection systems directly coupled with acoustically levitated samples include fluorescence imaging, right angle light scattering, Raman spectroscopy, surface-enhanced Raman scattering (SERS) spectroscopy, and X-ray diffraction.7 The main goal in all of these approaches was the efficient minimization of the required volume of reactants and analytes as well as solvent usage, while maintaining or improving the analytical performance of the detection method. Only recently has acoustic levitation been combined with mass spectrometry. Westphall et al.12 were the first group to Received: April 18, 2016 Accepted: August 9, 2016

A

DOI: 10.1021/acs.analchem.6b01519 Anal. Chem. XXXX, XXX, XXX−XXX

Technical Note

Analytical Chemistry

provides a more in-depth reaction monitoring timeline, yielding information on a seconds time scale, in comparison to the minutes time scale of the same reaction performed in vitro using a conventional reaction vessel (400 μL total reaction volume). Acid-catalyzed degradation of EA frequently occurs as a result of improper storage or errors in processing during manufacturing and readily degrades even under mildly acidic conditions, yielding the pharmacologically inactive main decomposition product anhydroerythromycin A (AEA).38 An acid stable internal standard, namely, clarithromycin (CL), was employed as it provided very stable performance even at up to 5% acid addition. The major resulting degradant of EA, AEA, is no longer active against bacterial infections and therefore renders the pharmaceutical product inactive.39 Actively monitoring for such changes in a pharmaceutical end product using a miniaturized microscale analytical setup is attractive, as miniaturized assays require less material, solvent, reagents, and analytical standards. Miniaturization also permits the pharmaceutical product to be sampled more frequently, because of the reduced amount of sample required and the major decrease in the analytical measurement times, which was within 30 s per measurement in our study. The real time acid-catalyzed degradation of EA was directly initiated within single liquid droplets (6 μL total droplet volume) in an acoustic field for in stillo microscale degradation reactions, while a scaled-up in vitro (400 μL reaction volume, 6 μL sampled) reaction was carried out for method comparison and control.

employ acoustic levitation as a containerless sample holder integrated into their charge and matrix-assisted laser desorption/ionization (CALDI)-time-of-flight (TOF) instrument. The levitated liquid sample (5 μL) contained a mixture of analyte and ionic matrix and direct desorption was achieved via laser ablation at 337 nm. Subsequently, Leiterer et al.13 reported on reaction monitoring for the first time, studying a polymerization reaction directly in acoustically levitated drops. Analyte characterization was carried out offline via matrixassisted laser desorption ionization (MALDI)-TOF and gel permeation chromatography (GPC). In addition, Stindt et al.14 and Warschat et al.15 combined laser ablation (using CO2 and IR lasers) with TOF mass spectrometry to directly analyze liquid samples in a containerless environment. Ambient ionization, usually applied as a means of direct sample ionization for MS analyses, has developed extensively in recent years, because it is able to interrogate samples directly from their native physical state. Numerous ambient techniques have evolved, mostly derived from the two well-established techniques of desorption electrospray ionization (DESI)16 and direct analysis in real time (DART),17 with additional methods developed independently or as combinations of multiple techniques for enhanced analyte coverage.18 Early development of soft ionization methods included fieldinduced droplet ionization34 first established by Grimm and Beauchamp in 2003 and in later publications35−37 was explored for single droplet mass spectrometry and in particular looking at gas−liquid interfacial reactions in microliter-ranged droplets. This method first employed a vibrating orifice aerosol generator, and in later work a stainless steel capillary introduced the sample as a hanging droplet to the high-voltage pulse after exposure to gas for seconds to minute reaction time scale. Recent efforts in the scope of online reaction monitoring and reaction miniaturization via ambient ionization,19 and more specifically via spray-based ionization methods, have explored DESI,20−22 paper spray (PS),23,24 PS coupled with a gravitydriven microfluidic chip,25 ultrasonic-assisted spray ionization (UASI),26,27 gravitational sampling electrospray ionization (GSESI),28 extractive electrospray (EESI)29 and, most recently, droplet spray ionization from a glass microscope slide30 and microdroplet fusion mass spectrometry.31−33 All of these methods involve spraying microdroplets in various ways toward the inlet of the mass spectrometer, for example, by bouncing the spray off of solid surfaces or by combining two spray plumes in midair, generating typical droplets with volumes on the order of 0.5 pL−2 nL.32 The microdroplet fusion method attributes the observed reaction acceleration mostly to droplet evaporation, resulting in extreme pH changes in the droplets and increased confinement of the reagents within the shrinking droplets.32 In these types of experiments, the total reacted volumes ranged from 20 to 30 μL of solvent, plus 10 μL total of reagents for paper spray-based experiments. Typical experimental flow rates of 15−30 μL/min were employed for microdroplet fusion MS. In this work, in stillo (in droplet) microreaction progress was directly monitored by ambient ionization mass spectrometry for the first time via direct analysis in real time tandem mass spectrometry (DART-MS/MS). The acid-catalyzed degradation of the antibiotic erythromycin A (EA) was used as model reaction with a total reaction volume of 6 μL, to demonstrate the proof-of-principle of the new analytical technique. Under the in stillo reaction conditions, the microreaction occurs completely within the acoustic field of an acoustic levitator. It



EXPERIMENTAL SECTION Chemicals. Erythromycin A (EA) (98%) and clarithromycin (CL) (pharmaceutical secondary standard) were purchased from Sigma-Aldrich (Steinheim, Germany). Anhydroerythromycin A (AEA) was prepared by incubating a solution of EA (10 μM) with 1% formic acid (FA) at room temperature (22 °C) overnight. Formic acid (98−100%) was from AppliChem (Darmstadt, Germany), methanol (Chromasolv HPLC grade >99.9%) from Sigma-Aldrich. Purified water was generated via a Millipore (Bedford MA) filtration system. Acid-Catalyzed Degradation Reactions. For in vitro reactions, each acidified methanol solution (200 μL; prepared by spiking with formic acid) was mixed 1:1 with 200 μL of a 20 μM EA-CL working solution (1:1 molar ratio of EA and CL in purified water using stock solutions, EA 500 μM in 100% H2O; CL 200 μM in 1:1 acetone/H2O v/v), to give reaction mixtures at concentrations of 10 μM. The final acid concentrations of the solutions were 0, 0.01, 0.05, 0.1, 0.5, 1, and 5%; all in vitro reaction solutions were stored in brown plastic 1.5 mL microtubes for protection from light. Each solution was prepared immediately before MS analysis and solution pH immediately read after vortex mixing using a micro pH probe (SevenCompact pH meter, Mettler Toledo, Giessen, Germany). The determined apparent m,wpH values for methanol/ water final reaction solutions were in the range of 2.4−3.8; these values were averages of three measurements. In total, the degradation reaction time for in vitro samples was approximately 20 min from reaction-start to MS measurement. A total volume of 6 μL of reacted mixture was directly analyzed as a hanging drop via DART-MS/MS in triplicate for each acid concentration. After 48 h of storage at 6 °C, the same in vitro samples were reanalyzed under identical conditions. The in stillo single droplet reactions were carried out in the acoustic levitator after adding 3 μL of acidified methanol directly to 3 μL of EA-CL working solution. The final acid B

DOI: 10.1021/acs.analchem.6b01519 Anal. Chem. XXXX, XXX, XXX−XXX

Technical Note

Analytical Chemistry

Figure 1. Experimental setup of coupling acoustically levitated samples with real time ambient mass spectrometry. (a) The sample droplet (3 or 6 μL) is injected into the acoustic field; (b) detailed positioning of the microprobe for sample desorption/ionization. (c) The reacted micro droplet is directly positioned into the ionization region of the DART source for MS/MS analysis.

(HV) electrode, −3000 V; discharge electrode, +350 V; grid voltage, +350 V; and helium flow, 0.55 L/min (minimum setting). The gas heater was optimized at 375 °C, which yielded the optimum intensities for both analytes and internal standard. The source was coupled to a Bruker (Bremen, Germany) Esquire HCT ion trap mass spectrometer via a Vapur interface (IonSense) held at 430 mbar. Figure 1 shows the experimental setup, where the exit of the DART source was axially positioned 8 mm away from the entrance of the ceramic transfer tube in the Vapur interface. The metal sample probe with hanging droplet sample was perpendicularly positioned ≈2 mm in front of the DART source exit to the center and approximately 6 mm from the ceramic tube. The sample droplets were desorbed within 30 s of being placed in the desorption region. Data dependent tandem MS was performed for the EA, AEA, and internal standard CL. For identification, an optimized averaged fragmentation amplitude of 0.18 V was used for collision induced dissociation (CID). The m/z range set for all experiments was m/z 450−775 with ion charge control (ICC) turned on, smart target value of 30 000, maximum accumulation time of 100 ms, and averaging of 5 mass spectra. The peak area under the first 10 data points of the collected DART signal was manually integrated for each analyte and internal standard.

concentrations of the mixtures were the same as for the in vitro reactions; however, the reaction volume was only 6 μL. The microreaction mixture remained in the levitator for 30, 60, and 120 s reaction times, and then the droplets were immediately introduced to the DART ionization source as a hanging droplet. Three replicates were recorded per time point and acid concentration. Acoustic Levitator and Sample Loading. A commercial acoustic levitator (Dantec/Invent Measurements Technology, Erlangen, Germany) with flat transducer and concave reflector operating at standard frequency of 58 kHz was used for all experiments. Droplets were injected into the acoustic levitator using a 1000 μL gastight syringe (Hamilton Bonaduz, Bonaduz, Switzerland) coupled with a 0.5 mm i.d. × 24 mm long beveled stainless steel needle via a NanoJet syringe pump (Chemyx, Stafford, TX) at 1300 μL/min for a fixed volume of either 3 μL (in stillo experiments) or 6 μL (in vitro experiments). The dispensing syringe pump employed was limited to a lower volume of 1 μL based on the combination of syringe diameter and syringe tip that was used. The droplets remained levitated in the acoustic pressure node for 30, 60, or 120 s (in stillo microreactions) or ≈10 s (in vitro reactions) and were then transferred from the acoustic field using a microprobe (0.25 mm diameter Hamilton stainless steel wire probe). During injection of the liquid droplets into the acoustic field, the power of the ultrasonic generator was set to 4.0−4.4 W. Once the droplets were suspended in the acoustic field, the power was reduced to 2.5−2.8 W, yielding a stable acoustic environment for 3−6 μL droplets. Droplet Evaporation Measurement Experiments. Levitated droplets (6 μL) were collected onto the tray of a Sartorius Pro 11 microbalance (Sartorius AG, Göttingen, Germany) after levitating for 0, 30, 60, and 120 s in stillo reaction times. The measured droplet weights were recorded and converted to volumes based on a methanol weight density of 40%40 in the 1:1 methanol/water v/v levitated solution. DART Ion Source and Mass Spectrometry. The direct analysis in real time (DART) ionization source used was an IonSense (Saugus, MA) DART-100 source, which was operated using the following conditions: positive ion mode; high voltage



RESULTS AND DISCUSSION Acoustic Levitation-Microreactor-DART-MS Setup. The newly hyphenated acoustic levitation-mass spectrometry technique allowed direct interrogation of the chemical content of dissolved molecules in droplets by means of ambient ionization. Unfortunately, the sonic field surrounding the levitated droplets prevented helium from directly entering the acoustic trap, necessitating an indirect means of exposing a hanging droplet to the beam of metastable helium for DARTMS. This was achieved by suspending the droplet from a stainless steel needle and immediately exposing the hanging droplet to metastable atom bombardment (Figure 1). This setup allowed samples in the range of 1−8 μL to be directly levitated and analyzed in real time via DART-MS. The DART ionization source thermally desorbs samples (from liquid or C

DOI: 10.1021/acs.analchem.6b01519 Anal. Chem. XXXX, XXX, XXX−XXX

Technical Note

Analytical Chemistry

Figure 2. Results of the reproducibility study across varying sample droplet volumes for erythromycin A and its major degradant species anhydroerythromycin A. (Average peak areas were normalized to the internal standard.)

Figure 3. Full scan mass spectra for in stillo acid-catalyzed reactions showing the shift in the relative protonated analyte signal intensity as a function of increasing formic acid concentration (0.01−5%).

permits microamounts of sample to be analyzed with minimal or no sample preparation. As a result of the contactless microenvironment within the acoustic levitator, no crosscontamination during the reaction occurs. The specificity and sensitivity of MS/MS permits direct monitoring of the microreaction and its progress in real time.

solid states) and ionizes molecules under ambient conditions, via Penning ionization using metastable helium.17 Analytes are usually protonated in positive ion mode by means of primary reagent ions such as ionized water clusters. The process is very fast and also permits direct ionization from solid samples. The combination of fast DART analysis and acoustic levitation D

DOI: 10.1021/acs.analchem.6b01519 Anal. Chem. XXXX, XXX, XXX−XXX

Technical Note

Analytical Chemistry Optimization of Droplet Volume. In stillo acid-catalyzed degradation of EA was carried out completely within the acoustic field of the acoustic levitator and directly monitored in real time via DART-MS/MS. For the miniaturization of the reaction, low microliter volumes of reactants were suitable for significant miniaturization of the reaction as compared to hundreds of microliters or milliliters of reactants required for reactions at in vitro levels. The volume of levitated drops was optimized by repeatedly injecting various droplet volumes (2, 4, 6, and 8 μL) directly into the acoustic levitator, followed by direct DART-MS. Figure 2 displays the results for EA and its major degradant AEA produced by in vitro reaction and 0.01% formic acid addition after 20 min. Overall, the 6 μL levitated droplet volume yielded the best repeatability based on the lowest achieved standard deviation as compared with the other experimental droplet volumes. The normalized peak areas for EA and AEA were consistent across the four droplet volumes and did not distinctly influence the optimal droplet volume selection. Figure S-1 (Supporting Information) provides details of the performed repeatability measurements across 10 suspended sample replicates, where RSD was within 10−19% for the extracted ion chromatograms for 6 μL volume levitated samples. In terms of the physical droplet stability in the acoustic field, the 8 μL droplets were often ejected from the pressure node due to unstable perturbations in the heavier droplets. On the basis of these optimization results, the 6 μL droplet was selected as the optimal experimental droplet volume and employed for all further in vitro acoustic sampling and in stillo miniaturized experiments. It is important to emphasize that even though the in vitro bulk reactions were carried out with a total reaction volume of 400 μL within microtubes, the amount sampled by DART-MS for direct comparison with the in stillo reactions was only 6 μL. The bulk reacted mixture was briefly (