SUPPLEMENT TO March 2016 - PharmTechfiles.pharmtech.com/alfresco_images/pharma/2019/02/... ·...

Petroleum Distillate Characterization Using GCxGC with High-Resolution TOF-MS

LC–MS-MS Determination of Mycotoxins in Food

Optimizing ESI for LC–MS-MS with a Design of Experiments Approach

LC–MS-MS Analysis of a Narcotic Antagonist and Its Metabolite

March 2016

S U P P L E M E N T T O

LCGC North America | LCGC Europe | Spectroscopy

Shimadzu’s MALDI systems include:

Order consumables and accessories on-line at http://store.shimadzu.comShimadzu Scientific Instruments Inc., 7102 Riverwood Dr., Columbia, MD 21046, USA

Offering the Utmost in MALDI-TOF MS Performance and Flexibility

From a linear-only system to a unique QIT-TOF

configuration to a high-resolution, high-energy TOF-TOF,

Shimadzu’s suite of MALDI mass spectrometers easily

meets the needs of researchers with varying applications

and budgets. Mine your data using our comprehensive

software solutions for:

Q Polymers and copolymers

Q Proteomics

Q Tissue imaging

Q LC-MALDI

Q Microorganism ID

Q Biomarker discovery

Q Ultrahigh mass applications

Q QA/QC

Q AXIMA Assurance linear system – ideal for QA/QC applications and simple yet accurate mass measurements

Q AXIMA Confidence linear/reflectron system – incorporating the patented curved field reflectron for flexibility and reliable, simple MS/MS

Q AXIMA Performance TOF/TOF system – high-energy CID and excellent sensitivity for complex biomolecule and polymer characterization

Q AXIMA Resonance QIT-TOF MS – uniquely designed for the structural characterization and sequencing of biomolecules, not just mass measurement

Q MALDI-7090 – providing unparalleled MS/MS resolution, ultrafast acquisition speeds, and excellent mass accuracy

Q MegaTOF – ideally suited for ultrahigh mass applications, allowing outstanding detection of macromolecules up to 1,500 kDa

Q iDplus – identify over 1300 pathogens in less than two minutes and easily register new microorganisms

Learn more about Shimadzu’s series of MALDI-TOF instruments, Call (800) 477-1227 or visit us online at

www.ssi.shimadzu.com/MALDI

Shimadzu’s Suite Meets the Requirements of Every Bioanalytical and Chemical Challenge

4 Current Trends In Mass Spectrometry March 2016 www.spec t roscopyonl ine .com

®

MANUSCRIPTS: To discuss possible article topics or obtain manuscript preparation

guidelines, contact the editorial director at: (732) 346-3020, e-mail: [email protected].

Publishers assu me no responsibility for safety of artwork, photographs, or manuscripts.

Every caution is taken to ensure accuracy, but publishers cannot accept responsibility for

the information supplied herein or for any opinion expressed.

SUBSCRIPTIONS: For subscription information: Spectroscopy, P.O. Box 6196, Duluth, MN

55806-6196; (888) 527-7008, 7:00 a.m. to 6:00 p.m. CST. Outside the U.S., +1-218-740-

6477. Delivery of Spectroscopy outside the U.S. is 3–14 days after printing. Single-copy price:

U.S., $10.00 + $7.00 postage and handling ($17.00 total); Canada and Mexico, $12.00 +

$7.00 postage and handling ($19.00 total); Other international, $15.00 + $7.00 postage and

handling ($22.00 total).

CHANGE OF ADDRESS: Send change of address to Spectroscopy, P.O. Box 6196, Duluth,

MN 55806-6196; provide old mailing label as well as new address; include ZIP or postal

code. Allow 4–6 weeks for change. Alternately, go to the following URL for address

changes or subscription renewal: https://advanstar.replycentral.com/?PID=581

RETURN ALL UNDELIVERABLE CANADIAN ADDRESSES TO: IMEX Global Solutions,

P.O. Box 25542, London, ON N6C 6B2, CANADA. PUBLICATIONS MAIL AGREEMENT

No.40612608.

REPRINTS: Reprints of all ar t icles in this issue and past issues are available

(500 minimum). Call 877-652-5295 ext. 121 or e-mail [email protected]. Outside US, UK,

direct dial: 281-419-5725. Ext. 121

C.A.S.T. DATA AND LIST INFORMATION: Contact Ronda Hughes, (218) 464-4430;

e-mail: [email protected]

INTERNATIONAL LICENSING: Maureen Cannon, (440) 891-2742,

fax: (440) 891-2650; e-mail: [email protected].

© 2016 UBM. All rights reserved. No part of this publication may be reproduced or transmit-ted in any form or by any means, electronic or mechanical including by photocopy, recording, or information storage and retrieval without permission in writing from the publisher. Authori-zation to photocopy items for internal/educational or personal use, or the internal/educational or personal use of specific clients is granted by UBM for libraries and other users registered with the Copyright Clearance Center, 222 Rosewood Dr. Danvers, MA 01923, 978-750-8400 fax 978-646-8700 or visit http://www.copyright.com online. For uses beyond those listed above, please direct your written request to Permission Dept. fax 440-756-5255 or email: [email protected].

UBM Life Sciences provides certain customer contact data (such as customers’ names, addresses, phone numbers, and e-mail addresses) to third parties who wish to promote relevant products, services, and other opportunities that may be of interest to you. If you do not want UBM Life Sciences to make your contact information available to third parties for marketing purposes, simply call toll-free 866-529-2922 between the hours of 7:30 a.m. and 5 p.m. CST and a customer service representative will assist you in removing your name from UBM Life Sciences’ lists. Outside the U.S., please phone 218-740-6477.

Spectroscopy/LCGC does not verify any claims or other information appearing in any of the advertisements contained in the publication, and cannot take responsibility for any losses or other damages incurred by readers in reliance of such content.

Spectroscopy/LCGC welcomes unsolicited articles, manuscripts, photographs, illustra-tions and other materials but cannot be held responsible for their safekeeping or return.

To subscribe, call toll-free 888-527-7008. Outside the U.S. call 218-740-6477.

Authorization to photocopy items for internal or personal use or for the internal or personal use of specific clients is granted by Advanstar Communications for libraries and other users registered with the Copyright Clearance Center, 222 Rosewood Drive, Danvers, MA 01923, tel. 978- 750-8400, fax 978-646-8700; call for copying beyond that permitted by Sections 107 and 108 of the U.S. Copyright law. LCGC/Spectroscopy and the logos appearing on the cover of this magazine are registered trademarks of Advanstar Communications, Inc.

UBM Life Sciences (www.ubmadvanstar.com) is a leading worldwide media company providing integrated marketing solutions for the Fashion, Life Sciences and Powersports industries. UBM Life Sciences serves business professionals and consumers in these industries with its portfolio of 91 events, 67 publications and directories, 150 electronic publications and Web sites, as well as educational and direct marketing products and services. Market leading brands and a commitment to delivering innovative, quality products and services enables UBM Life Sciences to “Connect Our Customers With Theirs.” UBM Life Sciences has approximately 1000 employees and currently operates from multiple offices in North America and Europe.

PUBLISHING & SALES

485F US Highway One South, Suite 210, Iselin, NJ 08830(732) 596-0276,

Fax: (732) 647-1235

Michael J. Tessalone Vice President/Group Publisher, Pharma/Science Group

[email protected]

Edward Fantuzzi Publisher,

[email protected]

Stephanie Shaffer Associate Publisher

[email protected] (774) 249-1890

Lizzy Thomas Account Executive,

[email protected] (574) 276-2941

EDITORIAL

Laura Bush Editorial Director,

[email protected]

Megan L’Heureux Managing Editor,

[email protected]

Stephen A. Brown Group Technical Editor, [email protected]

Cindy Delonas Associate Editor,

[email protected]

Dan Ward Art Director,

[email protected]

Anne Lavigne Marketing Manager,

[email protected]

Ronda Hughes C.A.S.T. Data and List Information,

[email protected]

Wright’s Media Reprints,

[email protected]

Maureen Cannon Permissions,

[email protected]

Jesse Singer Production Manager,

[email protected]

Wendy Bong Audience Development Manager,

[email protected]

Ross Burns Audience Development Assistant Manager,

[email protected]

SPP speed. USLC® resolution.

A new species of column.!��

!��

!��

!��

��#��

Selectivity Accelerated.

www.restek.com/raptor

www.restek.com/raptorPure Chromatography

6 Current Trends in Mass Spectrometry March 2016 www.spec t roscopyonl ine .com

ArticlesOptimization of the Electrospray Ionization Source with the Use of the Design of Experiments Approach for the LC–MS-MS Determination of Selected Metabolites in Human Urine 8Oliwia Szerkus, Arlette Yumba Mpanga, Michał J. Markuszewski, Roman Kaliszan, and Danuta Siluk

This article presents a step-by-step DoE optimization strategy for an electrospray ionization source with the aim to quantitatively analyze 18 compounds in a metabolomic study of human urine.

A Rapid, Selective, and Sensitive Method for the Analysis of Naltrexone and 6β-Naltrexol in Urine 17Oneka T. Cummings, Erin C. Strickland, and Gregory L. McIntire

An LC–MS-MS method is presented that can assist with monitoring substance abuse patients who are prescribed naltrexone, a narcotic antagonist.

Light and Medium Petroleum Distillate Characterization Using Two-Dimensional Gas Chromatography–High-Resolution Time-of-Flight Mass Spectrometry and Enhanced Software Processing 21David E. Alonso, Joe Binkley, and Clecio Klitzke

Enhanced chromatographic resolution (GC×GC), increased resolving power with high-resolution TOF-MS, and software designed to leverage these attributes are combined to provide data that are easy to process and interpret.

Multi-Toxin Determination in Food: The Power of “Dilute and Shoot” Approaches in LC–MS–MS 27Alexandra Malachová, Michael Sulyok, Eduardo Beltrán, Franz Berthiller, and Rudolf Krska

This article provides useful tips for smooth validation of multi-analyte LC–MS-MS methods and summarizes important validation outcomes for 295 analytes, including more than 200 mycotoxins.

DepartmentsProducts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34

Cover image courtesy of Andy Sotiriou/Getty Images.

March 2016

powered by SPEMSGCHPLC IR

CHROMacademy is the world’s largesteLearning website for analytical scientists

Packed with practical information that will help increase your knowledge.

Lite members have access to less than 5% of our content.

Premier members get so much more !

Put your questions to our panel of experts.

Complex scenarios welcome.

Ask the Expert.Get answers to your questions within 24 hrs

Tutor assisted certified coursesHPLC Operator | HPLC Method Developer

GC Operator | GC Method Developer

To find out more about CHROMacademy Premier membership contact:

Glen Murry on +1 732 - 346 - 3056 | e-mail: [email protected]

www.chromacademy.com

w w w.spec t roscopyonl ine .com8 Current Trends in Mass Spectrometry March 2016

Oliwia Szerkus, Arlette Yumba Mpanga, Michał J. Markuszewski, Roman Kaliszan, and Danuta Siluk

Simultaneous quantification of metabolites in biological fluids, where they are present at differ-ent concentration levels, is usually a challenging analytical task. One of the steps that should be undertaken to increase analytical method efficiency is optimization of the electrospray ionization (ESI) source, which can be especially helpful in increasing method sensitivity of agents with poor ionization characteristics. We present here a step-by-step ESI optimization strategy with the use of the design of experiments (DoE) approach. This multivariate statistical approach allows effective evaluation of the effects of multiple factors and interactions among factors on a given response in a minimum number of experimental runs.

Since the discovery of a soft ionization technique in the late 1980s, electrospray ionization (ESI), mass spectrometry (MS) started to play a more significant role in the analysis

of biological samples. Nowadays, ESI is the most commonly applied method for the ionization of molecules in liquid form and thus, is compatible with the common chromatographic separation techniques, for example, high performance liq-uid chromatography (HPLC) and ultrahigh-pressure liquid chromatography (UHPLC) (1). Furthermore, it is an ioniza-tion technique with considerably low chemical specificity. Ions that are generated in the process are very stable and are not prone to degradation, which is observed in another soft

ionization technique, namely matrix-assisted laser desorption ionization (MALDI). Moreover, because of the generation of multiply charged ions there is practically no limit in mass of large molecules in the ionization process. All of these features, together with high ionization efficiency, make ESI the most wide spread ion source in bioanalysis fields such as metabolo-mics, proteomics, and environmental monitoring (2–5).

When using liquid chromatography coupled with tandem mass spectrometry (LC–MS-MS), one is faced with the diffi-culty of finding the most optimal ionization parameters for the analysis of tested compounds. Generally, in most cases, basic optimization of MS parameters is conducted by an autotuning

Optimization of the Electrospray Ionization Source with the Use of the Design of Experiments Approach for the LC–MS-MS Determination of Selected Metabolites in Human Urine

www.spec t roscopyonl ine .com March 2016 Current Trends in Mass Spectrometry 9

procedure, with the use of optimization software supplied by the instrument’s manufacturer. However, when there is a need for sensitivity intensification, further and more detailed optimization of MS parameters might be considered toward an increase of the detector signal (6).

Optimization of experimental pa-rameters in such a complex system may be time consuming or unsuccessful if not designed with a proper plan. Most frequently, optimization of MS param-eters is conducted according to a one-variable-at-a-time (OVAT) strategy, in which the effect of tested parameters is assessed by changing the level of a single factor (parameter) while keeping the other factors at nominal levels. This univariate procedure, however, is not recommended as only a small part of the experimental domain is tested and no potential interactions between the fac-tors are taken into account. An OVAT strategy also requires a high amount of experimental runs to be performed (7).

Another optimization strategy uses

chemometrics-based, multivariate tech-nique like design of experiments (DoE). DoE enables the determination of sig-nificant experimental variables, building mathematical models for the responses, screening factors that have an influence on the response, and final optimiza-tion of selected factors. The factors are simultaneously varied among experi-mental runs according to a defined plan, with the goal to obtain maximum information from a minimum number of experiments. The relation between parameters and responses is described by mathematical modeling and can be well understood by graphical visualiza-tion. All interactions between factors are taken into consideration and measured factor effects represent the experimental design space more effectively compared to OVAT (8,9).

This article presents a step-by-step DoE optimization strategy for an elec-trospray ionization source with the aim to quantitatively analyze 18 compounds in a metabolomic study of human urine. The chosen goal was to improve

responses of 7-methylguanine and gluc-uronic acid, which demonstrated in the developed chromatographic conditions the poorest ionization characteristics in positive and negative ionization modes, respectively. For this purpose, detailed research of factors with potential influ-ence on the response was performed by grouping them in sequential optimiza-tion procedures, starting from screening and finishing with the actual optimiza-tion phase. Ion source parameters that had significant effect on MS response were selected through a screening pro-cedure, after application of fractional factorial design (FFD). Consecutive optimization of those parameters was performed by applying face-centered central composite design (CCD) for pos-itive ionization mode and Box–Behnken design (BBD) for negative ionization mode. Response surface methodology (RSM) and response surface graphs were used for visual impact assessment of the tested parameters and for defining the most optimal MS settings. As a result of this complex optimization procedure,

www.spec t roscopyonl ine .com10 Current Trends in Mass Spectrometry March 2016

the significant increase in MS signal was obtained that allowed for the develop-ment of a sensitive bioanalytical method for the determination of 18 metabolites in human urine samples in a metabolo-mics study.

ExperimentalMaterial and Reagents

The compounds 7-methylguanine, 1,7-dimethylxanthine, xanthine, 1-me-

tyluric acid, 3,7-dimethyluric acid, tryptophan, taurine, hypoxanthine, glucuronic acid, gluconic acid, aconitic acid, 2-furoylglycine, N-acetylneur-aminic acid, citrulline, hippuric acid, trimethyllysine, suberic acid, acetylly-sine, adrenaline, acetic acid (≥ 99,7%), and 0.1 N sodium hydroxide (NaOH) were obtained from Sigma-Aldrich. NaOH micropills were obtained from POCH SA. Pseudouridine and uridine

were purchased from Jena Bioscience. LC–MS-grade acetonitrile and HPLC-grade ethanol were purchased from J.T. Baker. Deionized water was acquired from a Milli-QPLUS system (Millipore). ESI-L Low Concentration Tuning Mix was purchased from Agilent Technolo-gies. Finally, 0.2-μm nylon membrane filters were obtained from Agilent Technologies.

Standard Solutions

Standard stock solutions of 5 mg/mL were obtained by dissolving each com-pound separately in the appropriate sol-vent (water, ethanol, or a solution of 1 M NaOH ). Stock solutions of 1 mg/mL were prepared for 1,7-dimethylxanthine and 1-methyluric acid in a solution of 0.1 N and 1 M NaOH, respectively. After that, stock solutions were filtered with 0.2-μm nylon membrane filters (Agilent Technologies). Working solu-tions of 0.05 μg/mL and 1 μg/mL were prepared in water for 7-methylguanine and glucuronic acid, respectively. A mixture containing 0.1 μg/mL of each compound was also prepared by mix-ing stock solutions of all compounds and diluting with water to obtain the desired concentrations.

Instrumentation

The LC–MS-MS experiments were per-formed with Agilent 1260 series LC sys-tem (Agilent Technologies) coupled with a 6430 Series Triple Quadrupole (QqQ) mass spectrometer (Agilent Technolo-gies) equipped with an ESI source in positive or negative mode. The software used for system control and data pro-cessing was MassHunter Workstation (Agilent Technologies). Nitrogen was used as the collision gas. Mobile-phase A was 0.06% acetic acid in water, and mobile-phase B was 0.06% acetic acid in acetonitrile. To optimize the ion source for MS analyses, DoE methods were car-ried out with a 1-min sample run com-posed of 1% B in isocratic elution. The capillary voltage, nebulizer pressure, gas flow rate, and gas temperature were optimized in the range of 2000–4000 V, 10–50 psi, 4–12 L/min, and 200–340 °C, respectively. To optimize those param-eters for all compounds, the compounds that ionized the least in positive ioniza-

Figure 1: A half-normal quantile plot that describes the factors and interactions which were assessed as significant (marked as red circle). (a) Positive ionization, (b) negative ionization.

400

300

200

100Ab

sorb

an

ce c

on

trast

Ab

sorb

an

ce c

on

trast

0

0.0

7

6

5

4

3

2

1

0

-1

0.5 1.0

Half normal quantile

Half normal quantile

1.5 2.0 2.5

0.0 0.5 1.0 1.5 2.0 2.5

Gas temp

Gas flow

Gas flow

Gas temp

Gas temp*Nebulizer pressureNebulizer pressure

Gas flow*Gas temp

(a)

(b)


tion mode (7-methylguanine) and nega-tive ionization mode (glucuronic acid) were chosen. The analyses were carried out in a multiple reaction monitoring (MRM) mode. For the optimization step, MRM transitions monitored for 7-methylguanine and glucuronic acid, were m/z 166 A 149 and m/z 181 A 113, respectively. The precursor ion, product ion, fragmentor, and collision energy for all 18 compounds are shown in Table I. Those transitions were obtained by the use of an Optimizer for Agilent Mass-Hunter Software (Agilent Technologies), and those that could not be found were optimized manually.

After the optimization step, a mixture of all compounds was analyzed using obtained spectrometric parameters with an optimized chromatographic method. The chromatographic analyses were performed with a gradient elution of mobile-phases A and B on a 150 mm × 4.6 mm, 2.7-μm dp Ascentis Express C18 column with a 2-μm precolumn fil-ter (Supelco Analytical) at 35 °C, with a flow rate of 0.3 mL/min and an injection volume of 2 μL. Next, 20-min and 10-min sample runs were provided for the analyses in positive and negative mode, respectively, under gradient elution, as shown in Table II. The equilibration lasted 12 min and 7 min for the analyses in positive and negative mode, respec-tively. A urine sample was also analyzed to test the applicability of the method to a biological sample.

Experimental Data AnalysisDuring the screening step, two-level FFD was applied for the simultaneous investigation of various factors and in-teractions, to assess their level of signifi-cance. In this type of screening design, only a small number of experimental runs is required in contrast to the full factorial design, in which the total amount of experiments is represented by Lf, where L denotes the number of factor levels and f is the number of fac-tors. Such an amount of experiments is assumed as undesirable because of economic limitations and project time-frames. Two-level FFD 2f-v contains only a fraction of the full factorial design and allows users to examine f factors at two levels in N = 2f-v experiments, with 1/2v,

that denotes the fraction of the full fac-torial (where v = 1,2,3, . . .). Thus, despite the fact that the number of experiments is decreased, FFD will generally lead to the same conclusions, however, not all

main and interaction effects can be as-sessed separately (4,5).

A four-factor, two-level screening FFD was generated using JMP 8.0 sta-tistical software (SAS Institute, Inc.).

Figure 3: Screening design–prediction profiler that models the system and predicts the optimal combination of settings (shown in red above each factor). (a) Positive ionization, (b) negative ionization.

2000

1500

1000

500

0

35

25

15

5

-5

4 6 8

10

12

200

250

300

2000

2500

3000

3500

4000

10

20

30

40

50

4 6 8

10

12

200

250

300

2000

2500

3000

3500

4000

10

20

30

40

50

Gas flow

Peak

po

siti

ve

Peak

neg

ati

ve

Gas tempCapillaryvoltage

Nebulizerpressure

8

Gas flow

8,119

Gas temp

270,7

564,

2727

±147,2

624

10,1

5725

±1,3

321

2703000 30

Capillaryvoltage

Nebulizerpressure

3000 30

(b)

(a)

Figure 2: List of factors and interactions that were examined to provide contrasts. Those highlighted in black are significant within 90% confidence level and those additionally marked with an asterisk within 95% confidence. (a) Positive ionization, (b) negative ionization.

Term Contrast Lenth t-Ratio p-Value

Contrast Lenth t-Ratio p-Value

Capillary voltageGas flow*Gas flowGas flow*Gas temp

Gas flow*Nebulizer pressure

Gas temp*Nebulizer pressureNull 10

Null 11

Nebulizer pressure

Gas temp

Gas flow

Term

Capillary voltageGas flow*Gas flowGas flow*Gas temp

Gas flow*Nebulizer pressureGas temp*Nebulizer pressure

Null 10

Null 11

Nebulizer pressure

Gas temp

Gas flow

386.000

254.668

119.92557.244

-74.951198.596

63.854

113.31628.310

-2.151

6.11673

3.84188

2.87181

1.10438

-0.57266

1.81647

0.84641

3.12979

-0.34667

-0.27711

3.69

2.32

1.73

0.67

-0.35

1.10

0.51

1.89

-0.21

-0.17

0.0151*

0.0402*

0.0925

0.5159

0.7603

0.2437

0.6503

0.0733

0.8550

0.8839

3.43

2.27

1.070.51

-0.671.77

0.57

1.010.25

-0.02

0.0183*

0.0436*

0.25890.65270.51590.0888

0.6116

0.28530.8238

0.9873

(a)

(b)


The design table, including the studied parameters and their ranges, can be found in Table III. In the performed experiments, a 2-μL solution of 7-meth-ylguanine (0.05 μg/mL) and glucuronic acid (1 μg/mL) were injected onto the instrument. An addition of acetic acid to the mobile phase in a concentration of 0.06% provided an acceptable ioniza-tion response for all of the tested com-pounds in both ionization modes. Peak areas of the selected metabolites were chosen as the experimental response with the aim of maximization. Each two-factor interaction was taken into account in the study. The experiments were performed in randomized order to minimize the influence of experimental and systematic bias. After completion of the assigned experiments, peak areas for both metabolites were extracted from a single mass spectrum and entered into the statistical software program to de-termine the effect of each tested param-eters. The factors that were estimated as significant in screening design were afterward used as the main factors in a response surface design.

In this type of design, the most im-

portant factors found from screening are tested in more detail, at higher resolution. Those significant variables were tested using CCD (for positive ion-ization) and BBD (for negative ioniza-tion) design with narrower ranges for each parameter. The choice of the two different types of design were based on the number of significant factors, with the aim of the minimization of amount of experimental runs. An experimental matrix for both designs with narrower limits for significant parameters is pre-sented in Table IV.

Results and DiscussionScreening of Significant Variables

Using Fractional Factorial Design

In our study, two levels were chosen for each of the tested parameters, maximum (+1) and minimum (-1), which covered the operational and practical range. The center points at level 0 were also added to assess the linearity of the responses within the examined range and to evalu-ate system error. After incorporation of all information (factors, levels, and re-sponses) into the statistical program, the design resolution was chosen: resolution

IV. In practice, resolution denotes the degree of alias, which would be allowed in the study. Resolution IV means that main effects are not confounded with other main effects or two-factor interac-tions. However, two-factor interactions are confounded with other two-factor interactions. Afterwards, the DoE plan was generated with a list of experiments to be performed. Based on these experi-ments, each parameter’s effect was cal-culated and extrapolated to the entire response system. In our study, the gener-ated design covered only 11 experimen-tal runs with inclusion of three center points. Performance order was random-ized. The drying gas flow, temperature, capillary voltage, and nebulizer pressure parameters were examined in ranges, which reflected the operational settings used in our laboratory. The change of the response, peak areas of both ana-lytes, reflects the difference in the elec-trospray ionization, which results from various ion source factor settings.

The coefficients of the response function, their statistical significance and analysis of variance (ANOVA) for responses were evaluated at 95% con-fidence level by the method of least squares. Regression models (for both analytes) were found statistically signifi-cant at p < 0.05. The quality of the fit of the polynomial model equation was represented by the value of R2. The ob-tained values of R2 were 0.978 (positive ionization) and 0.993 (negative ioniza-tion). They indicated the reliability of the equation revealing that 97.8% and 99.3% of the variation could be ascribed to the independent variable. In addition, the lack-of-fit statistics for each model was determined. The lack-of-fit test shows the difference between the total error and “pure” error sum of squares. If the lack-of-fit is large, the model might not be appropriate for the data. The F-ratio determines whether the variation due to lack-of-fit is small enough to be accepted as a negligible portion of the pure error. The lack-of-fit analyses for our regression models did not reveal statistical significance, indicating that the models were appropriate in all cases.

The brief summary of screening sta-tistics is presented in Figures 1–4. The half normal probability plot (Figure 1)

Figure 4: Response surface design–prediction profiler that models the system and predicts the optimal combination of settings (shown in red above each factor). (a) Positive ionization, (b) negative ionization.

320

310

300

290

320Gas temp

9.95Gas flow

270Gas temp

34.869572Nebulizer

10Gas flow

1134.5

16

±27.9

7034

Peak a

rea

po

siti

ve

280

270

10

320

30

35

40

45

50

310

300

290

280

270

10

9.59

8.58

9.59

8.58

900

950

1000

1050

1100

1150

280.1

918

±9.8

97008

Peak a

rea

neg

ati

ve

190

210

230

250

270

290

(b)

(a)


shows the absolute value of the contrast (estimated factor ef-fect) against the normal quantiles for the absolute value of normal distribution. Significant effects, which contribute the most of the variance to the response, are separated from the line toward the upper right and are marked in red. Figure 2 presents the list of main factors and interactions with their corresponding contrasts, t-ratios and p-values (significant fac-tors are highlighted in black). A factor at p < 0.1 (90% confi-dence level) was considered statistically significant. Factors at p < 0.05 (95% confidence level) were marked with an asterisk. The contrasts present the degree of the impact onto response. Those that are large in magnitude value induced the greatest effect on the response. The sign for contrast (+ or -) indicates whether the settings of parameters should be maximized (+1) or minimized (-1) to achieve the best response. For the posi-tive ionization mode, two of the four main parameters were considered important, namely temperature and flow rate of the drying gas. For the ionization in the negative mode, the most critical variables were gas temperature, gas flow rate, and nebulizer pressure.

The identification of the most critical main factors for the ESI positive and negative ionization modes seem to be pre-dictable. The ions are formed in the solution phase and after-wards are changed into gas phase before entering into mass spectrometer. Because of that, the parameters such as drying gas temperature and drying gas flow rate are crucial for the desolvation process of the electrosprayed droplets, the forma-tion, and control of the size of these droplets. The nebulizer pressure determines nebulization process efficiency, which al-together with drying gas temperature and flow rate, play the leading role in the quality and quantity of the ions produced in the source. For the other variable, capillary voltage, which was applied to the heated capillary, only the partial effect was observed and the factor was not considered as significant for both ionization modes.

The gas flow rate was found to be the most significant main factor in terms of induced change to both studied responses. In addition, a few two-factor interactions were found significant, which would not be investigated in the OVAT procedure. For positive ionization, interaction between drying gas flow rate and its temperature was found important. For negative mode, two-factor interaction between gas temperature and nebulizer pressure was statistically significant.

Figure 3 presents a prediction profiler plot that visualizes how the levels of each factor affects the other, using standard least-squares analysis. Application of this plot allows the global response modeling, maximization and calculation of each fac-tor coefficients that cause rise to the maximum response. The values marked in red, above the factor names, characterize the globally predicted optimal settings.

Optimization of Significant Variables Using Response Surface MethodologyThe parameters, which were assessed as insignificant within the tested ranges, were kept constant at settings that were predicted in FFD to achieve the most optimal response (as seen in Figure 4). Thus, nebulizer pressure was set at 35 psi in the positive ioniza-

Figure 5: Response surface plots in function of main significant variable, drying gas flow rate. (a) Positive ionization, (b) negative ionization—in function with drying gas temperature, and (c) negative ionization—in function with nebulizer pressure.

1150

8

8.5

9

9.5

320310

300

290

280

270

1150

1100

1050

1000

950

900

1100

1050

1000

Peak a

rea p

osi

tive

Peak a

rea p

osi

tive

Gas flow (8.10)

8.5

9

9.5

Gas flow (8

.10)

Gas temp (270.320)

310

300

290

280Gas

tem

p (270

.320

)

950

900

(a)

(c)

205

210

215

220

225

230

235

240

245

250

250

245

240

235

230

225

220

215

210

205

45

40

35

45

40

35

30

50

8

8.5

9

9.5Nebulizer (30.50)

Gas flow (8.10)

8.5

9

9.5

Gas flow (8

.10)

Neb

uliz

er (3

0.50

)

Peak a

rea n

eg

ati

ve

Peak a

rea n

eg

ati

ve

(b)

280

290

300

310

320

Gas

tem

p (270

.320

)

280

270

290

300

310320 Gas temp (270.320)

290

280

270

260

250

240

230

220

210

200

Peak a

rea n

eg

ati

ve

290

280

270

260

250

240

230

220

210

200

190

Peak a

rea n

eg

ati

ve

8.5

9

9.5

8

Gas flow (8.10)

8.5

9

9.5

Gas flow (8

.10)


tion mode while in both ionization modes the capillary voltage was set at 3000 V. Next, narrower ranges for statistically significant factors were set by limiting the values that were within 95.0% confi-dence interval—where the red horizontal

dashed line is crossed by blue dashed line (see Figure 4). The typed parameters with shorter ranges and design matrix were al-ready presented in Table IV. All experi-ments were carried out in randomized order with a standard solution of 0.05 μg/

mL and 1 μg/mL for 7-methylguanine and glucuronic acid, respectively.

Summarizing the statistics, an actual by predicted plot was generated to visu-alize the fitting of empirically obtained data to the data predicted by regression analysis. For both models the data are fit-ted very well, with R2 of 0.988 for CCD and 0.984 for BBD. In results of ANOVA, both models were considered significant (Prob > F 0.0006 for both) and the lack-of-fit test revealed that the obtained data were appropriate to the generated mod-els (Prob > F 0.5789 for BBD and Prob > F 0.8309 for CCD). ANOVA results for CCD showed that of the two examined factors, drying gas flow rate and tempera-ture, both appeared to be significant and the drying gas flow rate had the highest influence on the response. Moreover, the

Table I: MRM transitions, collision energy, fragmentor voltage, and polarity ion mode

Compound Precursor Ion Product Ion CE (V) Fragmentor (V) Polarity

7-Methylguanine 166149

124

20

20165 Positive

Pseudouridine 245125

191

15

795 Positive

Uridine 245113

70

8

2580 Positive

Xanthine 153110

136

15

1575 Positive

2-Furoylglycine 17095

124

20

790 Positive

1,7-Dimethylxanthine 18199

55

8

3670 Positive

Taurine 126108

85

10

585 Positive

N-Acetylneuraminic acid 310274

292

3

380 Positive

Citrulline 17670

159

24

875 Positive

Hippuric acid 180105

77

7

3570 Positive

Tryptophan 205188

146

8

1685 Positive

Trimethyllysine 189130

84

15

20110 Positive

Acetyllysine 18984

56

24

5295 Positive

Glucuronic acid 193 113 3 85 Negative

1-Methyluric acid 181138

83

10

20135 Negative

Gluconic acid 195129

75

9

15110 Negative

3,7-Dimethyluric acid 195180

124

15

18150 Negative

Aconitic acid 17385

129

7

355 Negative

Table II: Gradient elution program for LC–MS analysis in positive and negative ionization mode

Positive Ionization Mode Negative Ionization Mode

Time

(min)

Mobile-

Phase A (%)

Mobile-

Phase B (%)Time (min)

Mobile-Phase

A (%)

Mobile-

Phase B (%)

0 99 1 0 98 2

5 99 1 4.5 98 2

9 65 35 5 60 40

9.1 2 98 5.1 5 95

20 2 98 6 1 99

— — — 9 1 99


second order interaction of drying gas flow was also statistically important. For BBD, three of the tested factors (drying gas flow rate, gas temperature, and nebu-lizer pressure) were found significant as well as two-factor interaction between gas flow and gas temperature, which had the highest influence on the response, and quadratic interaction of nebulizer pres-sure and gas temperature. The prediction profiler was applied to achieve the com-bination of parameter settings to gain the best response (Figure 4). The most opti-mal combination is shown in red above each parameter.

Additionally, to visualize the level of each variable for maximum response, three-dimensional response surface plots were generated by plotting the response on the Z-axis against two parameters. The response surface plots for both re-sponses with the most influential main variable, gas flow rate, are presented in Figure 5. For CCD, it could be seen that peak area of 7-methylguanine increases together with an increase of gas flow and simultaneous increase of gas tempera-ture, which results from ESI operational principles. For BBD, it was found that the response obtained for glucuronic acid (peak area in negative ionization) increases with an increase of gas flow rate and decrease of gas temperature. In the function of gas flow rate and nebulizer pressure, the increase of response is ob-served when the gas flow rate is increased and the nebulizer pressure is decreased. The observations given above together with the most significant factors (gas flow and gas temperature) found to influence glucuronic acid peak area, are probably due to the fragility of the compound in the tested conditions. It should be noted that the optimal conditions predicted by the model in the screening design, match quite well with the results gained in re-sponse surface designs.

Final optimal conditions for the maxi-mum response of 7-methylguanine were as follows: drying gas temperature, 320 °C; drying gas flow rate, 10 L/min; capillary voltage, 3000 V; and nebulizer pressure, 30 psi. Final optimal conditions for gluc-uronic acid were as follows: drying gas temperature, 270 °C; drying gas flow rate, 10 L/min; capillary voltage, 3000 V; and nebulizer pressure, 35 psi. Those settings

Figure 6: Exemplary MRM chromatograms for (a) standards solution and (b) urine sample obtained in positive and negative (main and inset figures, respectively) mode. Peaks: 1 = 1,7-dimethylxanthine, 2 = trimethyllysine, 3 = acetyllysine, 4 = citrulline, 5 = taurine, 6 = pseudouridine, 7 = N-acetylneuraminic acid, 8 = 7-methylguanine, 9 = xanthine, 10 = uridine, 11 = 2-furoylglycine, 12 = tryptophan, 13 = hippuric acid, 14 = gluconic acid, 15 = glucuronic acid, 16 = aconitic acid, 17 = 1-methyluric acid, 18 = 3,7 dimethyluric acid.

3

1

0

1

2

3

4

04 5 6 7 8 9 10 11 12 13 14 15 16

4 5 6 7 8 9 10 11 12 13 14 15 16

1

2

3

4

5

1

2 34

5

6 7 89

10

14

15

16

17

18

12

11

13

2

4

Co

un

ts (

% X

10

1)

Co

un

ts (

% X

10

1)

Acquisition time (min)

5

67

910

1415

16

17

18

8

11

13

12

7-methylguanine

H2N

N N

N

H

7-methylguanine

HH

N

O

NCH

3

N

H2N

O

O

O OH

HO

HO

HO O

O OH

OHOH

Glucuronic acid

Glucuronic acid

HO

OHOH

CH3

(b)

(a)

Table III: Experimental plan for screening fractional factorial design and obtained responses

Run

NumberPattern

Factors

Drying Gas

Flow Rate

(L/min)

Drying Gas

Temperature

(ºC)

Capillary

Voltage (V)

Nebulizer

Pressure (psi)

1 ++ - - 12 340 2000 10

2 - - ++ 4 200 4000 50

3 ++++ 12 340 4000 50

4 0000 8 270 3000 30

5 0000 8 270 3000 30

6 - ++ - 4 340 4000 10

7 0000 8 270 3000 30

8 - - - - 4 200 2000 10

9 - + - + 4 340 2000 50

10 +- - -+ 12 200 2000 50

11 + - + - 12 200 4000 10


were applied in further metabolomics studies on selected metabolites in human urine samples. Exemplary MRM chro-matograms for 18 standards of metabo-lites and a urine sample, obtained under the final defined conditions are presented in Figure 6.

ConclusionsThe DoE approach is favored for global op-timization of complex systems character-

ized by multiple variables and settings. Op-timization of such complex experimental domains using the OVAT approach might be ineffective and economically unjusti-fied due to neglecting interactions between studied factors and the large number of ex-periments to be conducted.

A step-by-step optimization procedure utilized in the present article describes the systematic approach in method develop-ment with the aim to increase the poten-

tial of the ESI source. Comprehensive and planned experimental assessment allowed us to select the most significant variables and establish source settings to obtain maximized response, peak areas of two metabolites (7-methylguanine and gluc-uronic acid), and allowed further metabo-lomic studies of human urine. Afterward, the most optimal ESI source conditions were applied in a metabolomics study to quantify concentrations of 18 metabolites in human urine samples.

References:(1) M. Wilm, Mol. Cell Proteomics 10(7), 1–8

(2011).(2) R. Aebersold and M. Mann, Nature

422(6928), 198–207 (2003).(3) W. Struck, D. Siluk, A. Yumba-Mpanga,

M. Markuszewski, R. Kaliszan, and M.J. Markuszewski, J. Chromatogr. A 1283, 122–131 (2013).

(4) K. Macur, C. Temporini, G. Massolini, J. Grzenkowicz-Wydra, M. Obuchowski, and T. Baczek, Proteome Sci. 8(60), 1–8 (2010).

(5) R. Bujak, R. Gadzała-Kopciuch, A. Nowaczyk, J. Raczak-Gutknecht, M. Kordalewska, W. Struck-Lewicka, M.J. Markuszewski, and B. Buszewski, Talanta 146, 401–409 (2016).

(6) N. Kostic, Y. Dotsikas, A. Malenovic, B. J. Stojanovic, T. Rakic, D. Ivanovic, and M. Medenica, J. Mass Spectrom. 48(7), 875–884 (2013).

(7) D.L. Massart, B.G.M. Vandeginste, L.M.C. Buydens, S. de Jong, P.J. Lewi, and J. Smeyers-Verbeke, Handbook of Chemo-metrics and Qualimetrics Part A (Elsevier Science, Amsterdam, 1997), Chapter 21.

(8) B. Dejaegher and Y. Vander Heyden, J. Pharm. Biomed. Anal. 56(2), 141–158 (2011).

(9) O. Szerkus, J. Jacyna, A. Gibas, M. Siec-zkowski, D. Siluk, M. Matuszewski, R. Kaliszan, and M.J. Markuszewski, Anal. Chim. Acta, submitted (2016).

Oliwia Szerkus, Arlette Yumba Mpanga, Michał J. Markuszewski, Roman Kaliszan, and Danuta Siluk are with the Department of Biopharmaceutics and Pharmacodynamics, at the Medical University of Gdansk, in Gdansk, Poland. Direct correspondence to: [email protected] ◾

Table IV: Response surface experimental design table and obtained responses, central composite design

Run

NumberPattern

Factors Response

Drying Gas

Flow Rate

(L/min)

Drying Gas

Temperature

(ºC)

Peak Area (Positive Mode)

7-Methylguanine

1 ++ 10 320 1136

2 a0 8 295 969

3 - - 8 270 917

4 0A 9 320 1107

5 0a 9 270 1066

6 - + 8 320 978

7 00 9 295 1077

8 A0 10 295 1118

9 + - 10 270 1077

10 00 9 295 1101

Table V: Response surface experimental design table and obtained responses, Box-Behnken design

Run

NumberPattern

Factors Response

Drying Gas

Flow Rate

(L/min)

Drying Gas

Temperature

(ºC)

Nebulizer

Pressure

(psi)

Peak Area

(Negative

Mode)

Glucuronic Acid

1 + 0 - 10 295 30 242.8

2 - + 0 8 320 40 267.3

3 + - 0 10 270 40 280.9

4 000 9 295 40 237.0

5 0 - + 9 270 50 219.2

6 0 + - 9 320 30 240.5

7 000 9 295 40 228.6

8 - 0 + 8 295 50 214.3

9 - - 0 8 270 40 200.0

10 000 9 295 40 231.1

11 0 - - 9 270 30 235.3

12 ++0 10 320 40 227.1

13 - 0 - 8 295 30 234.6

14 0++ 9 320 50 230.7

15 +0+ 10 295 50 229.2


Oneka T. Cummings, Erin C. Strickland, and Gregory L. McIntire

Naltrexone (Depade, Re Via, and Trexan) is a potent narcotic antagonist structurally similar to oxymor-phone and naloxone. It blocks the subjective effects of heroin and other opiates and is primarily used in the management of opioid dependence and alcohol dependence. While conjugated 6β-naltrexol is the major urinary metabolite in man, conjugated naltrexone and free 6β-naltrexol are also major urinary species. To assist with monitoring substance abuse patients who are prescribed naltrexone, a rapid, selective, and sensitive liquid chromatography–tandem mass spectrometry (LC–MS-MS) method was developed to analyze for both naltrexone and 6β-naltrexol post-enzymatic hydrolysis. The validation of this method and some representative patient data are discussed in this article.

A Rapid, Selective, and Sensitive Method for the Analysis of Naltrexone and 6β-Naltrexol in Urine

N altrexone is a pure narcotic antagonist structurally similar to oxymorphone and naloxone. It blocks the subjective ef-fects of heroin and other opiates and is primarily used in

the management of opioid and alcohol dependence (1). The drug is supplied as 25-mg or 50-mg tablets (under the brands Depade, ReVia, and Trexan) (2) and as an extended-release injectable for-mulation (Vivitrol) (3), which is administered monthly as an intra-muscular injection (1). Naltrexone undergoes extensive first-pass metabolism, and up to 70% of a dose can be recovered in urine within 24 h (1,4). Conjugated 6β-naltrexol (naltrexol) is the major urinary metabolite in humans, and conjugated naltrexone and free naltrexol are other known major urinary species (5). Naltrexol has significant pharmacological activity with a half-life that is greater than three times its parent, naltrexone (6).

Many of the published analytical methods used to detect na-ltrexone and its metabolites were reported in plasma, which is a more complex matrix and usually requires more sample pro-cessing and cleanup when compared to urine (5,7–9). In a high-throughput laboratory environment, the emphasis is on develop-ing methods that are not only rapid but also sensitive and selective. In view of that, this work demonstrates a method that requires dilution with internal standard, hydrolysis, and centrifugation. The subsequent analysis of naltrexone demonstrates toxicologi-cally acceptable limits of quantitation (LOQs) and limits of detec-tion (LODs) and robust instrument performance that minimizes instrument utilization time when analyzing naltrexone in urine.

The resulting method reported herein has a 1.7-min run time and a 2.7-min cycle time using a liquid chromatography–tandem mass spectrometry (LC–MS-MS) system. Limits of quantitation

for both naltrexone and 6β-naltrexol were determined to be 5 ng/mL. The development and validation of this method are discussed in further detail in this article.

Methods and MaterialsChemicals

Reference standards for naltrexone, naltrexone-D3, 6β-naltrexol, oxymorphone glucuronide, and oxymorphone were purchased from Cerilliant. All solvents, including metha-nol (Optima grade), acetonitrile (Optima grade), isopropanol (Optima grade), 0.2 M phosphate buffer (pH 7.5), and formic acid (88%) were purchased from VWR. Drug free urine was obtained from Utak Laboratories. IMCSzyme, a recombinant β-glucuronidase, was purchased from IMCS.

Instrumentation

The analytical method was developed using an Agilent 6460 LC–MS-MS system. After testing several ultrahigh-pressure liquid chromatography (UHPLC) columns, a 50 mm × 4.6 mm, 100-Å, 2.6-μm dp Phenomenex Phenyl Hexyl column coupled with a Phenomenex guard cartridge (Phenyl) was used to validate this method. The column temperature was maintained at 50 °C during analysis.

Patient Sample Preparation

Patient urine specimens (50 μL) were diluted 10× with a mas-termix consisting of 1.6 μg/mL of naltrexone-D3 and 2000 units of a recombinant β-glucuronidase from IMCS (IMCSzyme) in 0.02 M phosphate buffer, pH 7.5. Samples were incubated in an


oven for 60 min at 60 °C for hydrolysis. Earlier studies using oxymorphone glu-curonide (naltrexone glucuronide is not commercially available) demonstrated the time course of hydrolysis for these analytes required >60 min for complete hydrolysis. It was determined that a hydrolysis time of 60 min provided stable and consistent results with a control. A correction fac-tor of 1.14—adjusting for approximately 88% hydrolysis efficiency—was applied to results to mimic complete hydrolysis. Samples were removed from the oven and centrifuged before analysis by LC–MS-MS.

Standard Preparation

Reference standards were diluted to appro-priate calibrator level concentrations (5, 10, 50, 250, 1000, and 5000 ng/mL) in normal human urine and were then diluted an ad-ditional 10× as detailed above. The stand-ards were meant to mimic the already 10×

diluted patient samples. An oxymorphone glucuronide hydrolysis control was used to ensure the enzyme–mastermix was accept-able for every batch in addition to a positive and negative control.

Method ValidationThis method was validated for naltrexone and naltrexol as per an internal standard operating procedure (SOP) based upon College of American Pathologist require-ments (10) and applied to authentic positive urine samples to evaluate efficacy for ad-herence drug testing. For a standard at any concentration level to be considered pass-ing, during the validation of this method, quantitation must be within ±25% of the expected value and ion ratios and chro-matographic requirements must pass for all analytes assessed. Quantitation values are based on a passing calibration curve (no points excluded and R2 value greater than

or equal to 0.99), which must be freshly prepared and analyzed on the same day that the particular validation study was conducted. Furthermore, the coefficient of variation (CV) between sets of replicates must be ≤15%. Experiments performed during the method validation process are explained in greater detail below.

Limits or Sensitivity

The lower and upper concentration lim-its at which the method could accurately identify and quantify analytes were evaluated. To establish reliable LODs and LOQs, four out of five replicates must pass for the necessary criteria when compared to a passing calibration curve (detailed above). The LOD and LOQ both require the analyte signal to demonstrate repro-ducible ion ratio stability, while the LOQ requires the analyte signal to demonstrate reproducible quantitation.

Linearity

Linearity is the reproducible regression or fit of the calibration curve compared to expected concentration values. Four of the five replicates of each linearity point—usually comprising the calibrators and a minimum of five additional points within the analytical range—must pass all valida-tion criteria.

Carryover

Carryover is the highest quantitative level of analytes present that do not produce a concentration level above the lower limit of quantitation in a subsequent blank injec-tion. Carryover is investigated to ensure false positives are alleviated in a production environment. This experiment requires four of five blank injections following each of the five replicates at the potential carryo-ver limit concentration to quantitate lower than the established LOQ.

Precision and Accuracy

The precision and accuracy test measures the capability of the method to yield re-producibly accurate results over a period of multiple days at concentrations span-ning the concentration range of interest. A total of 10 replicates of three concentra-tions (omitting calibration points) within the analytical range of interest are chosen and run on three separate days. Out of these 10 injections, nine must pass within

Table I: The mass spectrometric ion fragmentation transitions for naltrexone analysis method

Analyte Transition FragmentorCollision Energy

(CE, V)

Naltrexone342.2 A 270.0* 112 28

324.2 A 212.0† 112 52

Naltrexol344.2 A 55.0* 107 52

344.2 A 308.1† 107 28

Oxymorphone302.1 A 227.0* 102 52

302.1 A 198.0† 102 28

Internal Standard Transition FragmentorCollision Energy

(CE, V)

Naltrexone-D3345.2 A 58.0* 102 44

345.2 A 270† 102 28

Cell accelerator voltage = 6 for all analytes and the internal standard

*Quantifier ion; †qualifier ion

Table II: Parameters used for the liquid chromatography gradient

Step Time (min) Flow Rate (mL/min) %A* %B†

1 0.00

0.800

98 2

2 0.25 98 2

3 1.00 77 23

4 1.10 2 98

5 1.20 98 2

6 1.21 98 2

7 1.41 98 2

Stop time at 1.45 min

Post time = 0.25 min

*Solvent A = 5 mM ammonium formate in 10% methanol†Solvent B = 100% methanol


25% of the target concentration.

Interference–Selectivity

The ability of the method to be unaffected by the presence of other medications or compounds is the interference–selectivity. Interference would be considered con-firmed if a false signal met or exceeded LOQ requirements (that is, was higher than the LOQ and had a passing ion ratio) in greater than one of the five replicates that must be run for each individual test.

Matrix Effects

Matrix effects are the suppression or en-hancement of analyte signals of interest because of the presence of matrix—in this case urine. The effect of matrix on the method was evaluated by running the quality control (QC) concentration of 750 ng/mL both in matrix and in the LC start-ing conditions (98% 5 mM ammonium formate in 10% methanol: 2% methanol).

Patient Evaluation and Comparison

This test is used to confirm whether a new method accurately quantitates actual pa-tient samples when compared to the cur-rently in-use and accepted method for these analytes (10). The previously vali-dated method used a Waters Acquity TQD system and had been in regular daily use for more than a year. To make this com-parison, all recent patient positive urine samples for the analytes in this method were prepared using the new procedure and compared to the previous method.

Results and DiscussionDaily calibration curves (5, 10, 50, 250, 1000, and 5000 ng/mL) routinely demon-strated 100 ± 25% accuracy for each point and excellent goodness of fit (R2 > 0.99). Calibration curves for oxymorphone only included 250, 1000, and 5000 ng/mL as the hydrolysis control was set to liberate 1577.8 ng/mL. Consistent peak shapes and ion ra-tios were achieved for naltrexone, naltrexol, and oxymorphone. A typical chromato-gram is shown in Figure 1 for the 100 ng/mL calibrator demonstrating the separa-

tion achieved for this 1.7-min method; note that naltrexol, the latest eluted, is retained well before 1 min of the gradient. The longer run time (1.7 min) and cycle time (2.7 min) are necessary for column equi-libration and needle washing–carryover reduction, respectively. Details about the LC gradient and MS ion fragmentation transitions used are described in Tables I and II, respectively. Compiled results for all conducted method validation experiments are shown in Table III.

The LOD and LOQ for naltrexone and

Table III: Summary of validation data for the naltrexone method. The linearity results are compiled for five replicates of 5, 25, 50, 100, 250, 500, 1000, 2500, and 5000 ng/mL. Carryover was tested by running a matrix blank immediately following the 5000 ng/mL point. Precision and accuracy statistics were arrived at by compiling data from 10 replicates of three separate concentration stan-dards (75, 750, and 2500 ng/mL) prepared and analyzed on three separate days. Matrix data was calculated by comparing standards prepared in negative human urine versus a neat preparation in chromatographic starting conditions (98% A: 2% B). An extensive list of drug compound pain management medications and drugs structurally similar to naltrexone were assessed for potential interfer-ing signals, which may contribute to false positives.

Analyte

Linearity Carryover* Precision and Accuracy Matrix Interference

LOQ/

LOD

ng/mL

ULOL

ng/mLR2

Average

Concentration

ng/mL (n = 5)

Average % Target (n = 30) Average % CV (n = 30)%

Matrix

Effect

Interfering

compounds25 ng/

mL

750

ng/mL

2500

ng/mL

25 ng/

mL

750

ng/mL

2500

ng/mL

Naltrex-one

5 5000 0.996 0.00 97.1 102 99.7 6.6 1.9 2.0 -68.4 None

Naltrexol 5 5000 1.000 1.94 94.8 94.3 89.3 3.7 6.7 2.4 -54.9 None

Oxymor-phone

100 5000 1.000 4.58 N/A 97.9 102.5 13.2 4.8 4.3 -88.2 None

*At ULOLN/A = not applicable

Figure 1: Example chromatogram of a 100-ng/mL sample injection (internal standard excluded).

1.1

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

0

3.5

2.5

1.5

3

2

1

0

0.5

2.75

2.5

2.25

1.75

1.5

1.25

0.75

0.5

0.25

0.25 0.35 0.45 0.55 0.65 0.75 0.8 0.9 0.950.850.70.6Acquisition time (min)

Co

un

ts (

X10

3)

Co

un

ts (

X10

3)

Co

un

ts (

X10

4)

0.3 0.4 0.5

2

1

0

1

0.322

0.704

Oxymorphone

Naltrexone

Naltrexol

0.781


naltrexol were determined to be 5 ng/mL and 100 ng/mL for oxymorphone. No ap-preciable carryover was observed follow-ing the highest curve point of 5000 ng/mL (see Table III). The precision and accuracy was evaluated by assessing 10 replicates of standards at 75, 750, and 2500 ng/mL on three separate days—usually there is at least one skipped day between precision and accuracy experiments. The precision and accuracy results shown in Table III demonstrate the robustness of this method for all three analytes at the three concen-tration levels tested.

All analytes experienced suppressed signal response when dissolved in normal human urine compared to a neat prepara-tion in mobile phase initial conditions (98% 5 mM ammonium formate in 10% metha-nol: 2% methanol). Different methods can be used to calculate the effect of the ma-trix on the signal response. To calculate the percentages presented in Table III the rela-tionship between the absolute responses is used since it is believed to present a more thorough picture than comparison of rela-tive responses or calculated concentrations.

The interference study was con-ducted and there were no identified interferences for any of the analytes in this method. Interference analytes in-cluded structurally similar compounds such as naloxone, 6α-naloxol, codeine, oxycodone, noroxycodone, hydromor-phone, hydrocodone, and norhydroco-done as well as panels of common drugs

of abuse such as opiates, opioids, and benzodiazepines.

Analysis of Patient Samples

In addition to the validation, the patient samples tested showed 100% qualitative correlation between the original method and the newly validated method. Results from patient samples submitted who tested positive using this method and had known prescriptions for naltrexone (tablets or in-jectables) are listed in Table IV. Of the 7158 patient samples analyzed over a one-month period in a production setting using this method, 125 had known prescriptions and 90 tested positive while also passing all the specimen validity criteria (11,12). The sum-mary of this sample set (positive patients with a known prescription) is presented in Table IV.

Analysis of naltrexone dosages show a range of 15 ng/mL to 42,017 ng/mL with a mean of 4506 ng/mL for daily tablets (50 mg taken between once and thrice daily) and a narrower range of 18 ng/mL to 18,389 ng/mL for the injectable Vivit-rol formulation with a mean of 1381 ng/mL. It is difficult to assess a mean for these data because of the large discrep-ancies between measured concentration values for patients with both dosage forms. The median is a more meaning-ful measure for these data and is higher for patients taking the injectable form of naltrexone when compared to those tak-ing the daily tablets.

ConclusionThis work demonstrates that the analysis of naltrexone and its major metabolite, 6β-naltrexol, can be adequately assessed in a rapid LC–MS-MS quantitative method following hydrolysis. Key features of the method include hydrolysis using a recom-binant enzyme and a cycle time of 2.7 min while achieving the high level of selectivity and specificity required for naltrexone and naltrexol. The results from this method compare well with the same samples run with an LC–MS-MS method on a differ-ent system. This newly developed method uses different transitions and a comparable column and reduces the cycle time by ap-proximately 1 min.

References (1) R.C. Baselt, Disposition of Toxic Drugs

and Chemicals in Man, 10th Edition (Biomedical Publications, Seal Beach, California, 2014), pp. 1419–1420.

(2) Depade (package insert), Hazelwood, Missouri, Mallinckrodt Pharmaceuticals Inc. (2014).

(3) Vivitrol (package insert), Waltham, Mas-sachusetts, Alkermes Inc. (2010).

(4) K. Verebay, M.J. Kogan, A. de Pace, and S.J. Mule, J. Chrom. 118, 331–335 (1976).

(5) W. Huang, D.E. Moody, R.L. Foltz, and S.L. Walsh, J. Anal. Tox. 21, 252–257 (1997).

(6) E.J. Cone, C.W. Gorodetzky, and S.Y. Yeh, Drug Metab. Disp. 2, 506–512 (1974).

(7) M.H. Slawson et al., J. Anal. Tox. 31, 453–461 (2007).

(8) C. Clavijo et al., J. Chrom. B. 874, 33–41 (2008).

(9) A.F. Davidson, T.A. Emm, and H.J. Pien-iaszek, J. Pharmaceut. Biomed. Anal. 14, 1717–1725 (1996).

(10) J.R. Enders and G.L.McIntire, J. Anal. Toxi-col. 39(8), 662–667 (2015).

(11) D.M. Bush, Forensic Sci. Int. 174(2–3), 111–119 (2008).

(12) Substance Abuse and Mental Health Services Administration Department of Health and Human Services, “Mandatory Guidelines for Federal Workplace Drug Testing Programs,” Federal Register 69, 19643–19673 (2004).

Oneka T. Cummings, Erin C. Strickland, and Gregory L. McIntire are with Ameritox, Ltd., in Greensboro, North Carolina. Direct correspondence to: [email protected] ◾

Table IV: Data for patients who tested positive and had a known prescription for naltrexone tablets at 50 mg (dosed between one and three times daily) and Vivitrol at 380 mg each (dosed once monthly)

AnalyteStatistical

Measure

Tablets (50 mg,

QD, BID, TID)

Injectable (380 mg,

QMO)

Naltrexone

Sample size (n) 23 67

Min. concentration 15 18

Max concentration 42,017 18,389

Mean 4506 1380

Median 379 675

Naltrexol

Sample size (n) 23 67

Min. concentration 35 105

Max concentration 84,083 143,917

Mean 13,610 6349

Median 302 2257

QD = once daily, BID = twice daily, TID = thrice daily, QMO = once monthlyNote: Samples above the ULOQ are typically diluted further and reanalyzed to ex-trapolate more accurate quantitation.


David E. Alonso, Joe Binkley, and Clecio Klitzke

The enhanced resolution of comprehensive two-dimensional gas chromatography (GC×GC) was com-bined with the increased resolving power, speed, and mass accuracy of high-resolution time-of-flight mass spectrometry to confidently characterize molecules in light cycle oil (LCO) and vacuum gas oil (VGO). Optimized chromatographic and mass spectrometry parameters were implemented to improve data acquisition, processing, and heteroatomic speciation of these light to midlevel petroleum fractions. Advanced software tools were used to process the data to facilitate robust compound identifications and resulted in comprehensive characterization of molecules in LCO and VGO samples. Compound classes consisted of, but were not limited to, alkanes, cycloalkanes, aromatics, benzothiophenes, and carbazoles. Selective detection and quantification of alkylbenzothiophenes and dibenzothiophenes were conducted by retrospectively processing data using rapid two-dimensional, accurate mass target analyte finding. Results were compared to those provided by more-traditional petroleum analysis methods.

Light and Medium Petroleum Distillate Characterization Using Two-Dimensional Gas Chromatography–High-Resolution Time-of-Flight Mass Spectrometry and Enhanced Software Processing

The world-wide demand for “clean” fuels and lubricants in combination with changing feedstock supplies has in-creased the need for novel processing technologies and

analysis protocols for heavier and more-sour crude oils (1). Cur-rent environmental concerns and updated global regulations have mandated ultralow sulfur fuels that are compliant with new emission control technology (2–4). The results of process-ing methods and thus the quality of fuels can be best assessed through determination of the molecular composition of these complex materials. Indeed, the hydrocarbon content and degree of saturation is of great importance to the overall value of fuel; however, just as important is the minimization of less desirable compounds containing sulfur, nitrogen, or oxygen because of their lower energy output upon combustion and negative impact on catalytic refinement processes (5).

Gas chromatography (GC) is a widely used technique for

the analysis of low- to mid-level petroleum fractions because of its resolution, speed, robustness, and reproducibility. It can be coupled to various detectors including thermal conductiv-ity, flame ionization, atomic emission, mass spectrometry (MS), high-resolution MS, sulfur chemiluminescence, vacuum ultra-violet (VUV), and others for qualitative and quantitative analyses of petroleum fractions. For example, Garcia and coworkers (6) investigated the utility of GC with sulfur chemiluminescence detection and GC–high-resolution MS for aromatic sulfur detec-tion in gas oils and concluded that they were excellent technolo-gies for profiling and characterizing oils.

Comprehensive two-dimensional GC coupled with time-of-flight mass spectrometry (GC×GC–TOF-MS) has been used for the analysis of light- to mid-level petroleum fractions since it results in structured chromatograms and better separation through the use of orthogonally related, high efficiency columns.


The columns provide larger peak capacities in two-dimensional (2D) planes, and increased sensitivity as a result of cryogenic focusing of effluents in the modulator connecting the first and second dimension columns (7). GC×GC–TOF-MS has also been used to analyze higher molecular weight hydrocarbons (up to C60) in heavier vacuum gas oil (VGO) samples (8).

In this study, GC×GC was combined with high-resolution TOF-MS to obtain comprehensive structural information for

the molecular constituents of samples for light cycle oil (LCO), as shown in Figure 1. GC×GC–high-resolution TOF-MS data resulted in unequivocal compound identifications through spectral similarity comparisons to large, well-established data-bases and formula determinations for high-resolution accurate mass fragment and molecular ions. The distinct advantage of GC×GC-high-resolution TOF-MS for complex material analysis is a direct result of its ability to reduce coelutions through supe-rior chromatographic separation and minimize mass spectral interferences with higher mass analyzer resolving power.

ExperimentalLCO and VGO samples were diluted in dichloromethane (100:1 and 10:1) and 0.5 mL aliquots were placed in 2-mL vials. Data acquisition was carried out using an Agilent 7890 gas chromato-graph with a split–splitless injector, a Gerstel MPS2 autosampler, and a LECO Pegasus GC-HRT 4D MS system. The GC system was equipped with a secondary oven, a liquid nitrogen quad-jet, a dual-stage thermal modulator (LECO Corporation) and a dual capillary column set (Restek Corporation): column 1: a nonpolar 30 m × 0.25 mm, 0.25-μm df Rxi-5ms; column 2: a mid-polar 0.6 m × 0.25 mm, 0.25-μm df Rxi-17Sil MS.

LCO (0.5 μL) was injected into the split–splitless injector using a split ratio of 200:1 and an inlet temperature of 280 °C. The sepa-ration was performed using a 1-mL/min constant flow of helium and a temperature program of 40 °C (2-min hold) to 330 °C at 10 °C/min (3-min hold). VGO sample introduction consisted of a 1-μL injection with a 20:1 split and an inlet temperature of 350 °C. Helium at a constant flow of 1 mL/min and an oven temperature program of 40 °C (2-min hold) to 350 °C at 3 °C/min (25-min hold) was used for VGO component separation. The secondary oven and modulator were maintained at +5 °C and +15 °C relative to the corresponding primary oven temperature programs used in LCO and VGO analysis. Modulation times of 2 s and 6 s were used for LCO and VGO, respectively. The Pega-sus HRT 4D mass spectrometer was run with an electron energy of 70 eV, an emission current of 1 mA, and a source temperature of 250 °C. Data were acquired at 200 spectra/s, and the mass range was 35–510 m/z. LECO ChromaTOF-HRT software was used to acquire, process, and display data.

Results and DiscussionA significant improvement in comprehensive LCO and VGO sample characterization was achieved through coupling 2D GC with high-resolution TOF-MS. Comprehensive processing of the LCO data resulted in confident identification of alkanes, cyclo-alkanes, aromatics (mono-, tri-, tetra-), and heteroatomic com-pounds through a combination of spectral similarity searches and robust formula determinations using high-resolution ac-curate mass (HRAM) ions. For example, Tables I and II list re-tention times, formulas, expected and observed molecular ion m/z values, mass accuracy values, and spectral similarity search results for representative sets of aromatic and polyaromatic hy-drocarbons (PAHs). The average spectral similarity value for the aromatic hydrocarbon set was 903/1000 and 904/1000 for PAHs. Average absolute mass accuracy values for these sets of compounds were 0.43 and 0.36 ppm, respectively.

Figure 1: Contour plot illustrating different classes of compounds in LCO.

Figure 2: (a) Contour plot (TIC) and (b) expanded surface plot (XIC) displaying oxygen and nitrogen compounds in LCO. Peak true mass spectra for (c) dibenzofuran and (d) carbazole.

Mid

-po

lar

colu

mn

Nonpolar column

Toluene

4-Methylindan

Tridecane

Naphthalene

Acenaphthene

Anthracene

Chrysene

Pyrene Pyrene, 4-methyl-

Naphthacene

Triaromatic

Tetraromatic

Diaromatic/naphthenic

Monoaromatic/naphthenic

Monoaromatic

Diaromatic

1.4

2

287 787

Linear Branched

Saturated alkanesand cycloalkanes

1287 1787

1.9

2

(a) (b)

Dibenzofuran

897

285 785 1285 1785

1097

1297

14971.59

1.79

2/0

0.19

Mid-polar column

Mid

-po

lar

colu

mn

1.535

2/0

0.035

Nonpolar column

Nonpolar column

13

9.0

54

16

168.056941000

900

800

700

600

500

400

300

200

100

0

960/1000

1.13 ppm

m/z

40 60 80 100 120 140 160 180 200

C12H8O C12H9N

C13H11N

C14H13N

C15H15N

C13H10O

C12H8O

13

9.0

54

26

167.072911000

900

800

700

600

500

400

300

200

100

0

955/1000

NH

O

-0.27 ppm

m/z

40 60 80 100 120 140 160 180 200

C12H9N

(c) (d)

Figure 3: ChromaTOF-HRT (a) target analyte finding list and (b) HRAM ion list.

(a) (b)


A major bottleneck for petroleum anal-ysis is data processing and interpretation. GC×GC–high-resolution TOF-MS con-tour plots provided quick views or “fin-gerprints” that displayed organized clus-ters of compounds in complex petroleum fractions. HRAM ions were used to easily extract information from complex chro-matograms as shown for oxygenates and nitrogen containing compounds in LCO (Figure 2). Peak true (deconvoluted) mass spectral data are displayed for dibenzofu-ran and carbazole. Spectral similarity and mass accuracy values for these heterocyclic aromatic compounds were 960/1000, 1.13 ppm and 955/1000, -0.27 ppm.

The software was used to quickly in-terrogate rich data sets retrospectively in a targeted manner after compounds of interest were identified using its com-prehensive Peak Find processing feature. GC×GC–high-resolution TOF-MS data from these complex petroleum fractions were processed by leveraging target com-pound HRAM ions and XY coordinates or regions (retention times) of chromato-graphic maps (contour plots). Figure 3

illustrates how dibenzothiophene and alkylated dibenzothiophenes (C1–C4) in VGO were targeted using a new soft-ware processing feature, target analyte finding (TAF). TAF methods allow for minimum peak height and area specifi-

cations, smoothing, and tabular input of compounds to locate in high-resolution TOF-MS comprehensive data files. Mo-lecular, fragment, or adduct search criteria can be specified in the “Expected Adduct” list of TAF processing methods. For VGO,

Figure 4: VGO GC×GC–high-resolution TOF-MS surface plot showing DBT and alkylated DBTs (C1–C4).

Mid-polar column

Nonpolar column

DBT

2601

3101

C1-C4DBT

3601

0.8

2.8

4.8

Find out more about CHROMacademy Premier membership contact:

Glen Murry on +1 732 - 346 - 3056 | e-mail: [email protected]

www.chromacademy.com/gc_troubleshooting.html

Interactive GC Troubleshooter

Overcome your instrument, separation and quantitation issues.

1. Select your chromatographic symptoms

2. Select your instrument symptoms

3. The troubleshooter returns a list of possible causes

and recommended solutions

Get answers fast, reduce downtime and improve efficiency

powered by

SPEMSGCHPLC IR BLS BIO


individual analyte names, target regions (first and second dimension times), and molecular formulas were entered into the “Analytes to Find” table (Figure 3a). The software automatically calculated the four most abundant ions in ion clusters. Alternatively, m/z values or additional ions can be added to the target list. In ad-dition, ion tolerance values or units (Da, mDa, or ppm) can be changed as well as the number of ions required for a positive

hit and the production of a peak marker in GC×GC–high-resolution TOF-MS contour plots. For example, a positive hit for a C3-dibenzothiophene (DBT) isomer (C15H14S) required both M+ (226.08162 ± 0.005) and [M+2]+ (228.07761 ± 0.005) ions for a peak in the specified region of the plot (Figure 3b).

The results of TAF for dibenzothio-phene (DBT) and alkylated dibenzo-thiophene (C1–C4 DBT) in VGO are

displayed on the surface plot in Figure 4. The peak numbers, names, retention times, areas, and formulas for diben-zothiophenes meeting the TAF criteria are listed in Table III.

The TAF processing feature was used to quickly target higher mo-lecular weight, sulfur-containing compounds (C17H12S to C20H14S) in a dif ferent region of the VGO contour plot as displayed in Fig-

Table I: Representative aromatic compounds in LCO

Name R.T. (s) Formula Expected m/z Observed m/z ppm Similarity

Ethylbenzene 389, 1.46 C8H10 106.07770 106.07776 0.57 934

Benzene, 1,3-dimethyl- 397, 1.45 C8H10 106.07770 106.07769 -0.13 969

Mesitylene 505, 1.44 C9H12 120.09335 120.09338 0.28 955

Benzene, 1-ethyl-2-methyl-

517, 1.46 C9H12 120.09335 120.09339 0.30 928

Benzene, 1,2,3-trime-thyl-

563, 1.49 C9H12 120.09335 120.09325 -0.85 957

Benzene, 1-ethynyl-4-methyl-

587, 1.63 C9H8 116.06205 116.06205 -0.03 896

Benzene, 1-methyl-3-propyl-

589, 1.43 C10H14 134.10900 134.10905 0.36 892

Benzene, 2-ethyl-1,4-dimethyl-

597, 1.43 C10H14 134.10900 134.10891 -0.67 939

Benzene, 1,4-diethyl- 601, 1.44 C10H14 134.10900 134.10906 0.41 847

Benzene, 1-methyl-2-propyl-

607, 1.43 C10H14 134.10900 134.10913 0.92 941


617, 1.45 C10H14 134.10900 134.10896 -0.30 960


623, 1.46 C10H14 134.10900 134.10893 -0.56 899

Benzene, 1,2,3,4-te-tramethyl-

645, 1.48 C10H14 134.10900 134.10900 -0.03 860

Benzene, 1,2,4,5-te-tramethyl-

655, 1.47 C10H14 134.10900 134.10903 0.20 956

Benzene, 1-methyl-4-butyl

687, 1.43 C11H16 148.12465 148.12475 0.64 881


689, 1.52 C10H14 134.10900 134.10904 0.26 847

Benzene, (1-methyl-2-cyclopropen-1-yl)-

693, 1.61 C10H10 130.07770 130.07775 0.39 963

Benzene, 1-methyl-4-(1-methylpropyl)-

699, 1.45 C11H16 148.12465 148.12475 0.69 905

Benzene, 1-methyl-4-(1-methylpropyl)-

709, 1.45 C11H16 148.12465 148.12465 -0.02 884

Benzene, pentamethyl- 729, 1.47 C11H16 148.12465 148.12457 -0.57 858

Benzene, (1,3-dimeth-ylbutyl)-

739, 1.41 C12H18 162.14030 162.14032 0.12 861

Benzene, 2,4-dimethyl-1-(1-methylpropyl)-

753, 1.41 C12H18 162.14030 162.14019 -0.69 855

Benzene, 1,4-dimethyl-2-(2-methylpropyl)-

769, 1.415 C12H18 162.14030 162.14047 1.03 833

Benzene, 1,3,5-trime-thyl-2-propyl-

779, 1.435 C12H18 162.14030 162.14036 0.38 855


ure 5. The mass accuracy values for three naphthenic benzothiophenes, (9,10-dihydro-6,7-dimethylbenzo[b]n a p h t h a [ 2 , 3 - d ] t h i o p h e n e , dinaphtho(1,2-b:2’,1’d)thiophene and 5,6-dihyrodinaphtho(1,2-b:1’,2’d)thio-phene), ranged from -0.54 to 1.13 ppm.

ConclusionThe combination of enhanced chromato-graphic resolution (GC×GC), increased resolving power of high-resolution TOF-MS, and software designed to le-verage these attributes resulted in data that were easy to process and interpret. Superlative structural characterization of low to mid-level petroleum fractions

was facilitated by the production of high quality spectral data obtained with un-matched spectral acquisition rates for proper chromatographic peak definition. Confident molecular identifications were made via spectral similarity and formula searches of high-resolution accurate mass data. In addition, rapid speciation of various compound classes was accom-plished through retrospective analysis of comprehensive GC×GC–high-resolution TOF-MS data using TAF.

References (1) H. Muller, F.M. Adam, S.K. Panda, H.H.

Al-Jawad, and A.A. Al-Hajji, J. Am. Soc. Mass Spectrom. 23, 806–815 (2012).

(2) A.T. Nawaf, S.A. Gheni, A.T. Jarullah, and I.M. Mujtaba, Fuel Process. Tech-nol. 138, 337–343 (2015).

(3) U. Arellano, J.M. Shen, J.A. Wang, M.T. Timko, L.F. Chen, J.T. Vazquez Rodriguez, M. Asomoza, A. Estrella, O.A. Gonzalez Vargas Gonzales, and M.E. Llanos, Fuel (2014). Avail-able at: http://dx.doi.org/10.106/j.fuel.2014.11.001.

(4) B. Jiang, H. Yang, L. Zhan, R. Zhang, Y. Sun, and Y. Huang, Chem. Eng. J. 283, 89–96 (2016).

(5) F. Adam, F. Bertoncini, D. Thiebaut, S. Esnault, D. Espinat, and M.C. Hen-nion, J. of Chrom. Science 5, 643–649 (2007).

Table II: Representative PAHs in LCO

Name R.T. (s) Formula Expected m/z Observed m/z ppm Similarity

Naphthalene 723, 1.72 C10H8 128.06205 128.06212 0.57 957

Naphthalene, 1-ethyl- 901, 1.65 C12H12 156.09335 156.09335 0.01 960

Naphthalene, 1,2-dime-thyl-

949, 1.72 C12H12 156.09335 156.09337 0.10 956

Naphthalene, 1-propyl- 973, 1.62 C13H14 170.10900 170.10890 -0.58 923

Acenaphthene 973, 1.80 C12H10 154.07770 154.07760 -0.64 933

Naphthalene, 2-(1-methylethyl)-

979, 1.63 C13H14 170.10900 170.10900 -0.03 905

1-Isopropenylnaphtha-lene

993, 1.72 C13H12 168.09335 168.09339 0.26 816

Naphthalene, 1,6,7-tri-methyl-

1029, 1.67 C13H14 170.10900 170.10906 0.32 958

Fluorene 1045, 1.80 C13H10 166.07770 166.07754 -0.96 933

Naphthalene, 2-methyl-1-propyl-

1053, 1.59 C14H16 184.12465 184.12464 -0.05 903

Chamazulene 1065, 1.60 C14H16 184.12465 184.12470 0.29 856

1,4,5,8-Tetramethyl-naphthalene

1069, 1.61 C14H16 184.12465 184.12454 -0.59 865

Naphthalene, 1-methyl-7-(1-methylethyl)-

1097, 1.67 C14H16 184.12465 184.12461 -0.23 846

Naphthalene, 1,6-dime-thyl-4-(1-methylethyl)-

1109, 1.58 C15H18 198.14030 198.14027 -0.18 842

9-Ethylfluorene 1111, 1.75 C15H14 194.10900 194.10896 -0.20 887

Anthracene 1183, 1.92 C14H10 178.07770 178.07767 -0.18 952

Phenanthrene, 1-me-thyl-

1257, 1.88 C15H12 192.09335 192.09335 -0.01 939

Anthracene, 9-methyl- 1293, 1.95 C15H12 192.09335 192.09331 -0.22 852

Phenanthrene, 3,6-dimethyl-

1317, 1.83 C16H14 206.10900 206.10902 0.11 925

Pyrene 1387, 3.07 C16H10 202.07770 202.07761 -0.47 883

Phenanthrene, 2,4,5,7-tetramethyl-

1447, 1.77 C18H18 234.14030 234.14031 0.04 844

Pyrene, 1-methyl- 1451, 3.03 C17H12 216.09335 216.09310 -1.15 950

Naphthacene 1559, 3.11 C18H12 228.09335 228.09343 0.34 890

Chrysene 1565, 3.14 C18H12 228.09335 228.09310 -1.12 920


(6) C.L. Garcia, M. Becchi, M.F. Grenier-Loustalot, O. Paisse, and R. Szyman-ski, Anal. Chem. 74, 3849–3857 (2002).

(7) H. Zheng, F. Zou, E. Lehne, J.Y. Zuo, and D. Zhang, in Advanced Gas Chro-matography–Progress in Agricultural, Biomedicine and Industrial Applica-tions, M.A. Mohd, Ed. (InTech, Rijeka, Croatia, 2012), pp. 363–388.

(8) T. Dutriez, M. Courtiade, D. Thiebaut, H. Dulot, F. Bertoncini, J. Vial, and M.-C. Hennion, J. Chromatogr. A 1216, 2905–2913 (2009).

David E. Alonso, Clecio Klitzke, and Joe Binkley are with LECO Corporation in St. Joseph, Michigan. Direct correspondence to: [email protected] ◾

Table III: GCxGC TAF results: DBT and C1–C4 DBTs with their corresponding total peak areas

Peak Name R.T. (s) Area Formula Peak Name R.T. (s) Area Formula

1 DBT 2817, 4.52 2,629,779 C12H8S 6 C2DBT 3219, 4.19 4,437,186 C14H12S

Total 2,629,779 7 C2DBT 3249, 4.10 1,676,725 C14H12S

8 C2DBT 3255, 4.14 2,119,964 C14H12S

Peak Name R.T. (s) Area Formula 9 C2DBT 3261, 4.19 5,531,624 C14H12S

2 C1DBT 3021, 4.33 4,295,769 C13H10S 10 C2DBT 3291, 4.13 572,273 C14H12S

3 C1DBT 3057, 4.32 470,704 C13H10S 11 C2DBT 3309, 4.31 5,135,844 C14H12S

4 C1DBT 3063, 4.36 3,788,441 C13H10S 12 C2DBT 3333, 4.36 4,784,801 C14H12S

5 C1DBT 3111, 4.51 2,561,581 C13H10S 13 C2DBT 3363, 4.32 694,955 C14H12S

Total 11,116,495 17 C2DBT 3417, 4.48 153,497 C14H12S

Total 25,106,869

Peak Name R.T. (s) Area Formula

14 C3DBT 3369, 4.06 69,087 C15H14S Peak Name R.T. (s) Area Formula

15 C3DBT 3387, 4.05 1,519,258 C15H14S 23 C4DBT 3543, 3.93 688,290 C16H16S

16 C3DBT 3405, 3.98 293,316 C15H14S 25 C4DBT 3579, 3.92 231,063 C16H16S

18 C3DBT 3435, 4.01 3,317,105 C15H14S 28 C4DBT 3597, 3.87 1,322,518 C16H16S

19 C3DBT 3477, 4.02 4,586,246 C15H14S 30 C4DBT 3615, 4.02 232,753 C16H16S

20 C3DBT 3489, 4.09 726,991 C15H14S 32 C4DBT 3627, 3.86 341,210 C16H16S

21 C3DBT 3495, 4.16 3,394,187 C15H14S 34 C4DBT 3639, 3.85 994,656 C16H16S

22 C3DBT 3525, 4.19 6,619,829 C15H14S 35 C4DBT 3651, 4.06 1,590,787 C16H16S

24 C3DBT 3555, 4.20 4,463,060 C15H14S 36 C4DBT 3669, 3.95 896,010 C16H16S

26 C3DBT 3585, 4.19 923,265 C15H14S 37 C4DBT 3699, 3.99 2,588,975 C16H16S

27 C3DBT 3591, 4.31 570,484 C15H14S 38 C4DBT 3777, 4.10 1,076,716 C16H16S

29 C3DBT 3603, 4.3 842,817 C15H14S 39 C4DBT 3789, 4.08 123,869 C16H16S

31 C3DBT3621, 4.295

141,328 C15H14S 40 C4DBT 3801, 4.17 865,625 C16H16S

33 C3DBT 3633, 4.35 234,361 C15H14S 41 C4DBT 3831, 4.17 211,106 C16H16S

Total 27,701,332 Total 11163577

Figure 5: VGO (a) contour plot, (b) contour plot expansion showing higher molecular weight sulfur-containing compounds (C17H12S to C20H14S) and (c–d) peak true mass spectral data with molecular ion mass accuracies for naphthenic benzothiophenes.

Mid

-po

lar

colu

mn

3.24

285 2285 4285 6285

5.24

Mid

-po

lar

colu

mn

4.275

5.275

6/0

Nonpolar column

C17H12S

C20H12S

C20H14SC18H14S

C18H16S

C17H16SC17H14S

11

7.0

24

60

235.0

5739

249.0

7316

264.096601000

900

800

700

600

500

400

300

200

100

0

50 100 150 200 250 300

S -0.48 ppm

m/z

C18H16S

14

2.0

32

69

284.065741000

900

800

700

600

500

400

300

200

100

0

50 100 150 200 250 300

S

1.13 ppm

m/z

C20H12S

14

2.0

32

68

26

9.0

42

06

286.080921000

900

800

700

600

500

400

300

200

100

0

50 100 150 200 250 300

S-0.54 ppm

m/z

C20H14S

(a)

(b)(c)

(d) (e)


Alexandra Malachová, Michael Sulyok, Eduardo Beltrán, Franz Berthiller, and Rudolf Krska

A liquid chromatography–tandem mass spectrometry (LC–MS-MS) “dilute and shoot” method for the determination of 331 (toxic) secondary metabolites of fungi and bacteria has recently been optimized and validated for different food matrices (1) according to the guidelines established in the Directorate General for Health and Consumer Affairs of the European Commission (SANCO) document No. 12495/2011. This article provides useful tips for smooth validation of multi-analyte LC–MS-MS methods and summarizes important validation outcomes for 295 analytes, including over 200 mycotoxins. The second part of this article focuses on the performance of the method in proficiency testing with an emphasis on difficult matrices. The z-scores obtained with this method were between -2 and 2 in 368 out of 408 cases for maize, wheat, triticale, bran, nuts, baby food, rai-sins, figs, coffee, liquorice, and hot pepper. From these results it can be concluded that quantitative determination of mycotoxins by LC–MS-MS based on a “dilute and shoot” approach is also feasible in the case of complex matrices such as such as pepper, coffee, or liquorice.

Multi-Toxin Determination in Food—The Power of “Dilute and Shoot” Approaches in LC–MS-MS

Spoilage of food and feed is still a global problem, leading to enormous annual losses of several hundred million tonnes. In relation to that, microscopic filamentous fungi

are considered as one of the most feared agents because of my-cotoxin production (2). Mycotoxins are low-molecular-weight, secondary metabolites of mold, and are toxic to animals and humans even in low concentrations (3). The European Union (and many other countries) has therefore established maxi-mum levels for the most important mycotoxins in food (4). Such regulations can be laid down or amended only based on long-term occurrence monitoring and availability of toxico-logical data followed by exposure risk assessment. The Euro-pean Food Safety Authority (EFSA) organizes a continuous call for data collection about the occurrence of already regu-lated (aflatoxins, ochratoxin A, deoxynivalenol, HT-2 toxin, and T-2 toxin) and so-far unregulated mycotoxins (nivalenol and ergot alkaloids) (5). Further collection of accurate data

is also needed for enniatins, beauvericin, sterigmatocystin, Alternaria toxins, citrinin, and phomopsins for human and animal exposure assessment (6).

The increasing requirements for accurate and reliable oc-currence data for more than one toxin has led to great progress in the development of highly selective, sensitive, and accurate liquid chromatography–tandem mass spectrometry (LC–MS-MS) methods for multi-mycotoxin determination in food and feed (7). A range of methods have been published on the iden-tification and accurate quantification of single or chemically related mycotoxins in several matrices (8). However, the num-ber of validated multi-analyte methods covering mycotoxins of different classes is limited for several reasons, beginning with sample preparation. The extraction solvent used has to solve a broad range of chemically unrelated toxins, and sophis-ticated cleanups have to be omitted to avoid discrimination of some compounds. Unspecific extraction and the reduction of


sample cleanup to a minimum can re-sult in suppression or enhancement of analyte responses during the ionization process. So-called matrix effects are a major challenge in successful develop-ment of reliable, quantitative multiana-lyte methods. Therefore, considerable efforts to control matrix effects should be performed to obtain accurate results. However, common approaches for counteracting matrix effects (unspecific QuEChERS cleanup, standard addition, matrix-matched standard addition, sta-ble isotope dilution assay [SIDA]) are not feasible for LC–MS-MS methods cover-ing more than 200 mycotoxins. In this case, injection of diluted extracts on a highly sensitive LC–MS-MS instrument has been shown to be the most effective approach (1).

Performing in-house validation is necessary for reliable quantification at a high level of trueness, preferably accord-ing to international guidelines. However,

validation of each matrix that should be analyzed for around 300 substances is costly and time-consuming. The Direc-torate General for Health and Consumer Affairs of the European Commission (SANCO) document for development of multi-analyte methods in pesticides residue analysis recommends that at least one representative commodity from each commodity group should be validated and evidence for fitness of purpose should be provided (9). This ap-proach, which was successfully applied in pesticide residues analysis, has also been followed for the validation of the developed multi-mycotoxin LC–MS-MS method (1).

Another difficulty in quality assurance of multi-mycotoxin LC–MS-MS data is the limited or even non-availability of certified reference materials containing multiple structurally unrelated mycotox-ins. Therefore, participation in proficiency testing is the only option to prove the

long-term performance of the method.The objective of this article is to pro-

vide useful tips for smooth validation of multi-analyte LC–MS-MS methods, summarize the important validation outcomes, and present the long-term performance of this “dilute and shoot” method in the proficiency testing with a special focus on “difficult” matrices.

ExperimentalChemicals and Consumables

Methanol, acetonitrile, ammonium acetate (all LC–MS-grade quality), and glacial acetic acid (p.a.) were purchased from Sigma Aldrich. A Purelab Ultra system (ELGA LabWater) was used for further purification of reverse osmosis water.

Standards of fungal and bacterial metabolites were obtained either as gifts from various research groups or from the following commercial sources: Romer Labs Inc., Sigma-Aldrich, Iris

Figure 1: Distribution of apparent recoveries of 295 analytes in different matrices (a) at the lowest spiking level and (b) at the highest spiking level, based on (1).

Figure 2: Distribution of matrix effects of 295 analytes in different matrices (a) at the lowest spiking level and (b) at the highest spiking level, based on (1).

250

n.d.

< 50

50-70

70-120

> 120

n.d.

< 50

50-70

70-120

> 120

Apple Hazelnut Maize GreenPeper


(b)(a)

Nu

mb

er

of

An

aly

tes

200

150

100

50

0

250

Nu

mb

er

of

An

aly

tes

200

150

100

50

0

n.d.

< 50

50-90

90-110

110-120

> 120

n.d.

< 50

50-90

90-110

110-120

> 120



(b)(a)

Nu

mb

er

of

An

aly

tes

200

150

100

50

0

200

150

100

50

0

Nu

mb

er

of

An

aly

tes


Table I: Overview of some proficiency testing (Bipea) results on “difficult” matrices

Matrix/PT ID AnalyteReported Result

(μg/kg)

Assigned Value

(μg/kg)

Standard Devia-

tion (μg/kg)z-Score

Peanutcake05-0231

Aflatoxin B1 495.4 467.3 98.7 0.28

Aflatoxin B2 96.8 55.5 14.1 2.93

Aflatoxin G1 127.5 117.6 28.8 0.34

Aflatoxin G2 21.3 11.3 3.1 3.23

Sum of aflatoxins 741.0 657.9 136.9 0.61

Ochratoxin A 5.1 7.2 2.1 -1.00

Triticale02-1931

Ochratoxin A 1.0 1.8 0.6 -1.55

Deoxynivalenol 593.0 578.0 142.0 0.11

Nivalenol 161.0 139.0 36.0 0.61

T-2 toxin 55.6 55.0 15.0 0.04

HT-2 toxin 64.5 64.0 18.0 0.03

Sum of T-2 and HT-2 toxins 120.1 118.0 31.0 0.07

Zearalenone 127.0 125.0 33.0 0.06

Marzipan01-3831

Aflatoxin B1 20.6 19.0 5.4 0.29

Aflatoxin B2 38.2 32.6 8.5 0.66

Aflatoxin G1 5.6 7.1 2.2 -0.70

Aflatoxin G2 4.2 2.3 0.7 2.67


Ochratoxin A 2.0 2.2 0.6 -0.30

Bran01-3631

Ochratoxin A 1.4 4.4 1.3 -2.28

Deoxynivalenol 660.0 620.0 151.0 0.26

Nivalenol 93.0 111.0 29.0 -0.62

T-2 toxin 27.9 59.0 16.0 -1.94

HT-2 toxin 43.2 79.0 21.0 -1.70

Sum of T-2 and HT-2 toxins 71.1 129.0 34.0 -1.70

Zearalenone 105.0 99.0 26.0 0.23

Hot pepper02-1031

Aflatoxin B1 14.0 7.9 2.2 2.77

Aflatoxin B2 3.7 2.9 0.9 0.89

Aflatoxin G1 60.2 32.7 8.4 3.27

Aflatoxin G2 22.6 11.0 3.0 3.87


Ochratoxin A 25.7 19.6 5.2 1.17

Raisins03-3131

Aflatoxin B1 4.8 2.4 0.7 3.43

Aflatoxin B2 35.1 23.5 6.1 1.90

Aflatoxin G1 4.4 1.2 0.4 8.00

Aflatoxin G2 92.9 30.0 7.8 8.06


Ochratoxin A 5.2 5.7 1.7 -0.29


Biotech GmbH, Axxora Europe, and LGC Promochem GmbH.

Method Validation According

to the SANCO Document

Detailed information about method

validation is provided in reference 1. Briefly, four different model matrices were chosen for spiking experiments—apple puree (high water-containing matrix), hazelnuts (high fat content), maize (high starch content), and green

pepper (complex or “difficult” matrix) (9). Blank ground samples of 0.50 g were spiked with solutions containing all analytes and left overnight in darkness. Samples were then extracted with 2 mL of 79:20:1 (v/v/v) acetonitrile–water–ace-

Table I: Overview of some.... (continued)

Matrix/PT ID AnalyteReported Result

(μg/kg)

Assigned Value

(μg/kg)

Standard

Deviation (μg/kg)z-Score

Coffee03-1731

Aflatoxin B1 12.1 18.5 4.9 -1.31

Aflatoxin B2 5.3 11.4 3.1 -1.95

Aflatoxin G1 4.1 7.6 2.2 -1.60

Aflatoxin G2 6.7 6.6 1.9 0.07

Sum of aflatoxins 28.2 44.1 11.3 -1.41

Ochratoxin A 2.1 2.2 1.3 -0.06

Liquorice01-4131

Aflatoxin B1 6.8 5.8 1.7 0.61

Aflatoxin B2 3.2 5.1 1.5 -1.27

Aflatoxin G1 6.3 6.9 2.0 -0.30

Aflatoxin G2 5.9 3.6 1.1 2.09


Ochratoxin A 6.4 5.0 1.5 0.93

Pistachiopaste04-1431

Aflatoxin B1 17.0 20.8 5.5 -0.69

Aflatoxin B2 17.3 17.6 4.7 -0.06

Aflatoxin G1 4.3 4.5 1.4 -0.14

Aflatoxin G2 8.1 6.9 2.0 0.60


Ochratoxin A 3.9 4.2 1.3 -0.23

Raisins04-3131

Aflatoxin B1 22.6 15.3 4.1 1.78

Aflatoxin B2 10.1 9.2 2.6 0.35

Aflatoxin G1 7.2 8.6 2.4 -0.58

Aflatoxin G2 4.6 4.6 1.4 0.00


Ochratoxin A 5.0 5.9 1.7 -0.53

Hot pepper03-1031

Aflatoxin B1 18.0 13.3 3.6 1.31

Aflatoxin B2 15.4 11.0 3.0 1.47

Aflatoxin G1 11.1 9.6 2.7 0.54

Aflatoxin G2 19.8 9.1 2.6 4.13


Ochratoxin A 29.8 20.3 5.4 1.76

Figs02-2631

Aflatoxin B1 13.6 17.7 4.7 -0.88

Aflatoxin B2 22.9 26.7 7.0 -0.55

Aflatoxin G1 15.6 17.0 4.5 -0.30

Aflatoxin G2 8.9 9.1 2.6 -0.08


Ochratoxin A 0.7 2.1 0.7 -1.96


tic acid for 90 min on a rotary shaker (GFL) and subsequently centrifuged for 2 min at 3000 rpm. The extracts were transferred into glass vials using Pas-teur pipettes, and 350-μL aliquots were diluted with the same volume of 20:79:1 (v/v/v) acetonitrile–water–acetic acid. Finally, 5 μL of the diluted extract was injected into a LC–MS-MS system with-out further pretreatment. Spiking ex-periments were performed at four levels in five repetitions of each, resulting in relative concentrations of 1:3:10:30 of the final diluted extracts. External and matrix-matched calibrations were pre-pared by dilution of the final working solution with 49.5:49.5:1 (v/v/v) aceto-nitrile–water–acetic acid and diluted blank extracts of each model matrices, respectively, at the levels correspond-ing to spiked samples. In addition, one calibration point below and one above the spiking concentration range were prepared in anticipation of the expected matrix induced signal suppression and enhancement of target analytes.

The peaks were integrated and linear, 1/x weighted, calibration curves were constructed from the data obtained

from the analysis of each sample type (spiked sample, neat solvent standard, spiked extract) using MultiQuant 2.0.2 software (Sciex) to evaluate the linearity of the method. Further data evaluation was performed in Microsoft Excel 2007. All the other performance character-istics of the method (recovery, appar-ent recovery, repeatability, and matrix effects) were evaluated at each spiking level for all model matrices (1).

Instrumental Parameters

Detection and quantification was per-formed with a QTrap 5500 MS-MS system (Sciex) equipped with a TurboV electrospray ionization (ESI) source and a 1290 series ultrahigh-pressure liquid chromatography (UHPLC) system (Ag-ilent Technologies). Chromatographic separation was performed at 25 °C on a 150 mm × 4.6 mm, 5-μm Gemini C18 column equipped with a 4 mm × 3 mm C18 security guard cartridge (Phe-nomenex). Elution was carried out in bi-nary gradient mode. Both mobile phases contained 5 mM ammonium acetate and were composed of 10:89:1 (v/v/v) methanol–water–acetic acid (eluent A)

and 97:2:1 (v/v/v) methanol–water–ace-tic acid (eluent B), respectively. After an initial time of 2 min at 100% A, the proportion of B was increased linearly to 50% within 3 min. Further linear in-crease of B to 100% within 9 min was followed by a hold-time of 4 min at 100% B and 2.5-min column re-equilibration at 100% A. The flow rate was 1000 μL/min. ESI–MS-MS was performed in the scheduled multiple reaction monitor-ing (sMRM) mode both in positive and negative polarities in two separate chro-matographic runs. The sMRM detection window of each analyte was set to the respective retention time ±27 s and ±42 s in positive and in negative mode, respec-tively. The target scan time was set to 1 s. The settings of the ESI source were as follows: source temperature, 550 °C; cur-tain gas, 30 psi; ion source gas 1 (sheath gas), 80 psi; ion source gas 2 (drying gas), 80 psi; ion-spray voltage, −4500 V and +5500 V, respectively; and collision gas, (nitrogen) medium. Confirmatory identification was obtained through the acquisition of two sMRMs per analyte (with the exception of moniliformin and 3-nitropropionic acid that each exhibit

Table II: Performance of “dilute and shoot” LC–MS-MS method in the proficiency testing of multianalyte methods organized by ISPA-CNR

Matrix AnalyteReported Result

(μg/kg)

Assigned Value

(μg/kg)

Standard Deviation

(μg/kg)z-Score

Maize

Deoxynivalenol 1419.6 1264.0 195.0 0.8

Fumonisin B1 1062.1 1305.0 201.0 -1.2

Fumonisin B2 338.9 350.0 65.6 -0.2

Zearalenone 18.7 21.7 0.6 -0.6

HT-2 toxin 24.7 30.7 12.0 -0.9

T-2 toxin 52 54.4 6.8 -0.2

Ochratoxin A 2.7 2.73 4.8 -0.1

Aflatoxin B1 1.6 1.35 0.3 0.8

Aflatoxin G1 0.65 0.63 0.14 0.1

Aflatoxin B2 LOQ n.e n.e. n.e.

Aflatoxin G2 LOQ n.e n.e. n.e.

Wheat

Deoxynivalenol 1332.2 1298.0 200.0 0.2

Zearalenone 170 148 31.5 0.7

HT-2 toxin 51.5 58.8 12.9 -0.6

T-2 toxin 6.2 8.3 1.8 -1.1

Ochratoxin A 9.0 7.2 1.59 1.1

n.e. = not evaluated


only one fragment ion) (1), which yields 4.0 identification points according to commission decision 2002/657/EC (10).

Method Trueness

The trueness of the method is shown by participation in proficiency tests organized by Bipea (Gennevilliers, France) and by the Institute of Science of Food Production of the National Research Council of Italy (ISPA-CNR, Bari, Italy) (15).

Results and DiscussionUseful Tips for the Experimental

Set-up of the Validation Process

and Validation Data Evaluation

As no guidelines and recommendations suitable for validation of multi-myco-toxin methods (or multi-analyte meth-ods in general) have been laid down so far, the recommendations provided by SANCO for pesticides have been fol-lowed (9). However, slight amendments were necessary to improve validation outcomes, minimize working time for sample preparation, and lower valida-tion costs.

Spiking Concentrations

SANCO suggests validation at two con-centration levels using five repetitions, one level close to the limit of quantifica-tion (LOQ) and one higher. As the final working mixture of analytical standards contained 331 analytes at various con-

centrations, we decided to validate four levels, each at five repetitions. Such ex-perimental setup amendment prevented a potential nonlinearity of the method for some analytes and discrimination of analytes with high LOQs.

Miniaturization

The whole sample preparation proce-dure can be miniaturized for valida-tion purposes to decrease the amount of standards needed for spiking. For example, we used 0.5 g of blank sample material plus 2 mL of extraction solvent for validation, but in routine analysis and proficiency testing, 5 g of sample was extracted with 20 mL of extraction solvent instead. Be careful: High sugar content matrices (for example, raisins) cannot be miniaturized to 0.5 g because of inhomogeneity (1).

Method Accuracy

This was evaluated based on apparent recoveries instead of extraction recov-ery because we use neat-solvent calibra-tion and spiked samples in daily routine analysis rather than matrix-matched standards. The reason for this is that it is very difficult to find a blank matrix for all analytes involved in the method. Apparent recovery expresses both ex-traction efficiency and matrix effects, and therefore it contains more relevant and real-life information (1). In routine

analysis, data are also corrected for ap-parent recovery.

Matrix Effects

As SANCO provides no recommenda-tions about the acceptable range for ma-trix effects, the range of 90–110% was considered to have no effect. Matrix effect evaluation should be performed using two approaches: “traditional slope” and “one calibration point.” Traditional slope evalu-ation is commonly used during validation. The slope obtained from matrix-matched standards is compared to the one obtained from neat solvent standards. However, evaluation of matrix effects by compari-son of each calibration point instead of the slope better reflects the real extent of ma-trix effect across the whole concentration range. Therefore, we recommend using both approaches during validation (1).

Limit of quantification

Evaluation of the limit of quantification (LOQ) according to SANCO is highly recommended for multi-analyte meth-ods. LOQ is estimated as the lowest spiking level (LL) that allows reliable de-tection of all five replicates meeting the performance criteria of relative standard deviation (RSD) < 20% and mean recov-ery of 70–120% at both selected-reaction monitoring (SRM) transitions. LOQs ob-tained by the LL approach did not differ significantly from LOQs calculated from signal-to-noise in our experience (1).

Validation Summary

Following the guidelines laid down by SANCO for pesticide analysis, conve-nient validation data of our method were obtained for 295 out of 331 analytes. The remainder of the compounds could not be reliably validated for several reasons: non-availability of analytical standards; instability of an analyte in the final work-ing solution; and low concentration of an-alytical standard for spiking. Validation for some analytes was done only semi-quantitatively; a fungal extract was used for spiking instead of analytical standards that were not commercially available (1).

The distribution of apparent recover-ies and matrix effects throughout the set of 295 analytes in apple puree, hazelnuts, maize, and green pepper at the lowest and the highest spiking levels is displayed in

Figure 3: Compilation of the z-scores obtained by the multi-mycotoxin LC–MS-MS method in routine proficiency testing organized by Bipea (green lines: borders of acceptable range of z-scores; red lines: borders of questionable range of z-scores, area out of red lines: unacceptable values) (9).

4

3

2

1

0

-1

-2

-3

-4Maize Nuts Baby

foodRaisins, pepper,coffee, milk...

Grains


Figures 1 and 2, respectively. The high-est validation level corresponds to a 1:10 dilution of the final working solution of the analytical standards. As the lowest validated level, the lowest spiking level re-liably detectable at all five repetitions (RSD < 20) on both SRM transitions was con-sidered. As expected, green pepper was the most difficult matrix. Only 20% of ana-lytes met acceptable apparent recoveries of 70–120%, as recommended by SANCO (9). Moreover, 25% of the compounds were not detected in green pepper at any level lower than the highest validated level. With re-gards to the other matrices, more than 50% of compounds were in the acceptable range of 70–120%. For the analytes out of this range, either high matrix suppression or enhancement or low extraction efficiency was observed. More detailed information is provided in reference 1.

Any influence of matrix effects was observed for all matrices, albeit to a vary-ing extent, which was strongly dependent on the analyte–matrix combination (see Figure 2). Both matrix suppression and enhancement were observed. In general, the lowest matrix effects were obtained in apple puree and the highest in green pep-per. Only 10% of analytes were not affected by matrix effects in green pepper.

Method precision was proven by repeat-ability of five repetitions at the highest and lowest level. Less than 5% of compounds were above the RSD of 20% in apple puree, maize, and hazelnuts. In green pepper, the RSD was not below 20% for 11% of all vali-dated analytes.

Method Trueness—Proficiency Testing

To check for the trueness of the results obtained by our “dilute and shoot” LC–MS-MS approach, we have been participat-ing in a number of laboratory proficiency testing (PT) schemes organized by vari-ous institutions (Bipea, CODA-CERVA, FAPAS) since 2010. Regular PT partici-pation enables method drawbacks to be highlighted, helps to detect if something is out of control (for example, spoilage of a standard), and assists in the improve-ment of the overall method performance. Furthermore, analysis of uncommon and difficult matrices such as raisins or spices adds information about method accuracy in complex or even nonvalidated matri-ces. Generally, the spectrum of analytical

methods used by participating laboratories varies from enzyme-linked immunosor-bent assay (ELISA) through conventional high performance liquid chromatography (HPLC) coupled to fluorescence detec-tion (FLD) to advanced LC–MS-MS or LC–high-resolution mass spectrometry (HRMS)-based methods. Therefore, each successful PT participation has proven that the developed multi-mycotoxin method can reliably compete with other methods in terms of accuracy. A few examples of PTs are discussed below.

A total of 408 z-scores have been cal-culated out of the reported data within our continuous PT participation, and 368 values were in the acceptable range of -2 < z < 2. Only 23 z-scores were considered as questionable (between -3 and 3) and 17 were unacceptable (z < -3, z > 3). Figure 3 summarizes the z-scores of our multi-mycotoxin LC–MS-MS method in Bipea PT (11), where multiple mycotoxins were determined in maize, grains, nuts, baby food, and “difficult” matrices. The distri-bution shows that the majority of outliers (z-scores) were calculated for baby food, which exhibited concentrations very near the LOQ, and difficult matrices.

The performance of the method in some “difficult” matrices is presented in Table I. Although, matrices such as rai-sins, liquorice, or figs have never been fully validated, most of the obtained results were within the acceptable z-score range (-2 < z < 2). However, the z-scores for aflatoxins in hot pepper (02-1031) were out of the accept-able range. The reported results were evalu-ated on the external neat-solvent calibra-tion curve and corrected on the apparent recovery obtained from spiking of the vali-dated results of green pepper (12). Another very complex matrix was raisins. The sam-ple of raisins with identification number of 03-3131 was not successfully analyzed for aflatoxins. In this case, the results reported by laboratories were dispersed widely. Dur-ing the homogeneity testing and checks be-tween bags, some differences for aflatoxins G1 were found (13). Obviously, most of the participants did not homogenize the whole obtained batch. The assigned value was not calculated from the data of participat-ing laboratories, but the reference labora-tory measurements instead. It should be noted that reported results for aflatoxins in raisins (04-3131) analyzed later were in

agreement with other laboratories, yielding acceptable z-scores (14).

In 2014, ISPA-CNR organized a profi-ciency test on mycotoxins (15). It was aimed at laboratories willing to test the trueness of their multi-mycotoxin methods. Therefore, mainly LC–MS-MS methods were used for the quantitative determination of deoxyni-valenol, zearalenone, HT-2 and T-2 toxins, fumonisins, aflatoxins and ochratoxin A in maize and deoxynivalenol, zearalenone, HT-2 and T-2 toxins, and ochratoxin A in wheat within this proficiency testing scheme. The performance of our method is summarized in Table II. The developed LC–MS-MS multi-mycotoxin method was successful, yielding z-scores within the range of -2 < z < 2 (15). Concerning af-latoxin B2 and aflatoxin G2 in maize, no statistical evaluation was reported because of a lack of sufficient data.

ConclusionsThis article presents information on the background and the personal experiences that have been gained during the validation of a “dilute and shoot” multianalyte LC–MS-MS method for the determination of 331 (toxic) secondary metabolites of fungi and bacteria in food. In addition, guidance on proper method validation of a multi-analyte LC–MS-MS method is provided.

The performance of our “dilute and shoot” method in proficiency tests has been satisfactory. The routine participation in proficiency testing has proven that the de-veloped and validated LC–MS-MS method is able to provide accurate results for regu-lated mycotoxins in the case of “difficult” matrices such as raisins, figs, or liquorice.

References(1) A. Malachová, M. Sulyok, E. Beltrán, F.

Berthiller, and R. Krska, J. Chromatogr. A 1362, 145–156 (2014).

(2) http://www.mycotoxins.org/node/56 (accessed on 15 July 2015).

(3) J.W. Bennett and M. Klich, Mycotoxins Clin. Microbiol. Rev. 16, 497–516 (2003).

(4) Commission regulation (EC) No. 1881/2006 http://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=OJ:L:2006:364:0005:0024:EN:PDF (accessed on 15 July 2015).

(5) EFSA. http://www.efsa.europa.eu/en/search.htm?search=call&text=call (ac-cessed on 15 July 2015).


(6) CEN mandates, Call for tender for project leaders for the development of standardized methods for the analysis of mycotoxins in food (available online since 10 May 2013) file:///C:/Users/ataml/Downloads/Call_520_CENweb-site_2013_05_10.pdf

(7) Web of Science, Thomson Reuters.(8) P. Zöllner and B. Mayer-Helm, J. Chro-

matogr. A 1136, 123–169 (2006).(9) Document SANCO/12495/2011. http://

ec.europa.eu/food/plant/plant protec-tion products/guidance documents/docs/qualcontrol en.pdf (accessed on 10 December 2013).

(10) 2002/657/EC. http://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=OJ:L:2002:221:0008:0036:EN:PDF (accessed on 15 July 2015).

(11) Bipea, Provider of Proficiency Testing Programs. http://www.bipea.org/con-tent/pt-programs (accessed on 15 July 2015)

(12) Bipea, Interlaboratory comparisons report on hot pepper (02-1031), 31a-133-

Mycotoxins-Aflatoxins and Ochratoxins, 12 April 2013.

(13) Bipea, Interlaboratory comparisons report on raisins (03-3131), 31a-134-Mycotoxins-Aflatoxins and Ochratoxins, 24 July 2013.

(14) Bipea, Interlaboratory comparisons report on raisins (04-3131), 31a-140-My-cotoxins-Dried Fruits, Spices and Other Products, 16 October 2014.

(15) A. De Girolamo, B. Ciasca, J. Stroka, S. Bratinova, A. Visconti, and V.M.T. Lattan-zio, Report of the 2014 Proficiency Test for LC-MS(MS) multi-mycotoxin meth-ods, June 2015. http://www.ispacnr.it/?p=479.

Alexandra Malachová is a postdoc in the Christian Doppler Laboratory for Mycotoxin Metabolism and the Center for Analytical Chemistry at the Department of Agrobiotechnology (IFA-Tulln) of the University of Natural Resources and Life Sciences (BOKU) in Vienna, Austria. Michael Sulyok is head of the

Laboratory for Multi-Toxin Screening at the Department of Agrobiotechnology (IFA-Tulln) at BOKU. Eduardo Beltrán obtained an International PhD degree at University Jaume I in Castellon, Spain, in April 2014. He has performed pre-doctoral and post-doctoral research at the Center for Analytical Chemistry-IFA in Tulln, Austria. His current research is mainly focused on the application of LC–MS for the development of analytical methods for quantification and confirmation of different compounds (toxins, hormones, pesticides) in different matrices at very low levels. Franz Berthiller is an associate professor at the University of Natural Resources and Life Sciences, Vienna (BOKU) and Head of the Christian Doppler Laboratory for Mycotoxin Metabolism. Rudolf Krska is full professor for (Bio-)Analytics and Organic Trace Analysis and is head of the IFA-Tulln with more than 180 co-workers at the University of Natural Resources and Life Sciences, Vienna (BOKU). ◾

PRODUCTS & RESOURCESSample preparation systemA filtration option for Gerstel’s MultiPurpose Sampler sam-ple preparation system is designed to enable efficient automated cleanup of up to 98 samples or extracts com-bined with other sample preparation steps and sample introduction to an LC–MS-MS or GC–MS system. According to the company, liquid transfer is performed with exact con-trol of flow and volume.Gerstel GmbH & Co., KG,

Linthicum, MD; www.gerstel.com

LC columnsRestek’s FluoroPhenyl Phase Raptor columns are designed to run in reversed-phase or hydrophilic interaction mode for analyzing a variety of compounds. According to the company, because of their efficiency with acidic mobile phases, the columns also are suitable for LC–MS. Restek Corporation,

Bellefonte, PA; www.restek.com

Monolithic reflectron lensesMonolithic reflectron lenses from Photonis are designed to replace tra-ditional stacked ring reflectron lenses currently used in mass spectrom-eters. According to the company, the technology enables instrument designers to produce nonlinear and dynamic fields within the lens. Photonis, Inc.,

Sturbridge, MA; www.photonis.com

Mass spectrometerThe LCMS-8060 triple-quadru-pole mass spectrometer from Shimadzu is designed to pro-vide a scan speed of 30,000 u/s while maintaining mass accuracy and multiple reaction monitoring speeds of 555 ch/s. According to the com-pany, the instrument has a polarity switching speed of 5 ms. Shimadzu Scientific Instruments,

Columbia, MD; www.shimadzu.com

ADVERTISEMENT Mass Spectrometry 35

The sensitive, selective, and real-time analysis

characteristic of the SIFT-MS technique provides

simple, robust, and continuous analysis of extremely

diverse odor compounds at trace levels in air. This

application note illustrates instantaneous, broad-

spectrum odor analysis with monitoring data from a

chicken meat production facility.

Odor compounds are chemically very diverse, making

comprehensive analysis very challenging using conventional

analytical technologies. Furthermore, odors tend to be dynamic,

requiring fast response, whereas odor panels or lab-based analyses

are expensive and based on time-averaged samples.

Alternatively, selected ion flow tube mass spectrometry (SIFT-

MS) simultaneously detects and quantifies both organic and

inorganic odorous compounds in real-time in air to pptv levels with

no sample preparation (1,2). SIFT-MS is a direct mass spectrometry

technique that utilizes multiple, rapidly switchable chemical

ionization agents to deliver unparalleled real-time selectivity,

absolute quantitation, and wide linear and dynamic ranges.

This application note illustrates how readily a SIFT-MS

instrument can be applied on-site for comprehensive real-time

odor analysis.

Experimental Conditions

A van-mounted Syft Technologies’ Voice200 SIFT-MS instrument

was deployed at several chicken production facilities in South East

Queensland, Australia, during September and October 2013.

Air from the poultry shed was sampled continuously by the

SIFT-MS instrument using a flow-past configuration. A sampling

pump drew air at a flow rate of a few liters per min through Teflon

tubing and the SIFT-MS subsampled it at a flow rate of 25 sccm.

A diverse range of odorous compounds was targeted based on a

study using GC–MS (3), augmented with other odorous compounds

that are difficult to detect using traditional chromatographic methods

(for example, ammonia and hydrogen sulfide).

Results

Figure 1 shows a 13-h snapshot of real-time data obtained while

chickens were harvested and the shed cleared of used litter. A

diverse range of odor compounds were detected and quantified,

including ammonia, amines, hydrogen sulfide, organosulfur

compounds, ketones, aldehydes, and organic acids. The odor

profile changes dramatically over the analysis period, illustrating

the value of continuous, broad-spectrum monitoring using SIFT-

MS compared to time-averaged methods.

Conclusions

The real-time, broad-spectrum analysis provided by SIFT-MS

makes it ideally suited to continuous monitoring of odorous

organic and inorganic compounds even at the very low

concentrations required for investigation of odor complaints. The

Syft Voice200ultra SIFT-MS solution provides a robust, easily

deployed and operated package for on-site monitoring, but is

equally at home in contract and research laboratories.

References

(1) B.J. Prince, D.B. Milligan, and M.J. McEwan, Rapid Commun. Mass

Spectrom. 24, 1763–1769 (2010).

(2) V.S. Langford, I. Graves, and M.J. McEwan, Rapid Commun. Mass Spec-

trom. 28, 10–18 (2014).

(3) K.R. Murphy, G. Parcsi, and R.M. Stuetz, Chemosphere 95, 423–432

(2014).

Real-Time, Broad-Spectrum Odor Analysis Using SIFT-MSVaughan Langford, PhD, Syft Technologies Ltd

Syft Technologies

3 Craft Place, Christchurch 8024, New Zealand

tel. +64-3-338 6701, fax +64-3-338 6704

Website: www.syft.com

Figure 1: Comprehensive, continuous odor monitoring using SIFT-MS during harvesting of meat chickens and the subsequent clean out of the production shed.

www.gerstel.com

Heading home on time?

Or always working late?Don’t work day and night. Rely on the GERSTEL

MultiPurpose Sampler (MPS) to do that for you.

To automate your analysis, talk to us. We listen

and adapt our modular solutions to your needs.

Our sample preparation solutions for GC/MS and

LC/MS offer PrepAhead productivity for non-stop

throughput, perfect timing and a job well done.

Just set things in motion and be home on time.

What can we do for you?

Derivatization, Addition of Standards, Dilution, Calibration

Extraction, Filtration, SPE, Evaporation (mVAP)

SPME, SBSE, Thermal Desorption

Headspace, Dynamic Headspace (DHS) and PYRO

Automated Liner Exchange (ALEX)

Example: MPS Dual Head WorkStation with Centrifuge, Filtration, and Vortex (mVORX) options

Example: ALEX-GC-MS/MS-Systemfor QuEChERS, Metabolomics ….

SUPPLEMENT TO March 2016 - PharmTechfiles.pharmtech.com/alfresco_images/pharma/2019/02/... ·...

Documents

Transcript of SUPPLEMENT TO March 2016 - PharmTechfiles.pharmtech.com/alfresco_images/pharma/2019/02/... ·...