Deconvolution of Soft Ionization Mass Spectra of Chlorinated

Juliane GlügeLena SchinkelKonrad HungerbühlerRonan CariouChristian Bogdal. Environmental ... their use for quantification. Jannik Sprengel , Walter ...
0 downloads 0 Views 609KB Size
Subscriber access provided by United Arab Emirates University | Libraries Deanship

Article

Deconvolution of Soft Ionization Mass Spectra of Chlorinated Paraffins to Resolve Congener Groups Bo Yuan, Tomas Alsberg, Christian Bogdal, Matthew MacLeod, Urs Berger, Wei Gao, Yawei Wang, and Cynthia A. de Wit Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.6b01172 • Publication Date (Web): 17 Aug 2016 Downloaded from http://pubs.acs.org on August 27, 2016

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

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

Page 1 of 22

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

Analytical Chemistry

1

Deconvolution of Soft Ionization Mass Spectra of Chlorinated Paraffins to

2

Resolve Congener Groups

3 4 5

Bo Yuan1*, Tomas Alsberg1, Christian Bogdal2,3, Matthew MacLeod1, Urs Berger4, Wei Gao5, Yawei Wang5,

6

Cynthia A. de Wit1.

7 8 9

1

Department of Environmental Science and Analytical Chemistry, Stockholm University, Svante Arrhenius väg

10

8, SE-10691 Stockholm, Sweden

11

2

12

Prelog-Weg 1, CH-8093 Zürich, Switzerland

13

3

Institute for Sustainability Sciences, Agroscope, Reckenholzstrasse 191, CH-8046 Zürich, Switzerland

14

4

Department Analytical Chemistry, Helmholtz Centre for Environmental Research - UFZ, Permoserstraße 15,

15

DE-04318, Leipzig, Germany

16

5

17

Sciences, Chinese Academy of Sciences, Shuangqing Road 18, CN-100085 Beijing, China

Institute for Chemical and Bioengineering, Swiss Federal Institute of Technology, ETH Zürich, Vladimir-

State Key Laboratory of Environmental Chemistry and Ecotoxicology, Research Center for Eco-Environmental

1 ACS Paragon Plus Environment

Analytical Chemistry

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

18



19

We describe and illustrate a three-step data processing approach that enables individual congener groups of

20

chlorinated paraffins (CPs) to be resolved in mass spectra obtained from either of two soft ionization methods,

21

electron capture negative ionization mass spectrometry (ECNI-MS) or atmospheric pressure chemical ionization

22

mass spectrometry (APCI-MS). In the first step, general fragmentation pathways of CPs are deduced from

23

analysis of mass spectra of individual CP congeners. In the second step, all possible fragment ions in the general

24

fragmentation pathways of CPs with 10 to 20 carbon atoms are enumerated and compared to mass spectra of CP

25

mixture standards and a deconvolution algorithm is applied to identify fragment ions that are actually observed.

26

In the third step, isotope permutations of the observed fragment ions are calculated and used to identify isobaric

27

overlaps, so that mass intensities of individual CP congener groups can be deconvolved from the unresolved

28

isobaric ion signal intensities in mass spectra. For a specific instrument, the three steps only need to be done

29

once to enable deconvolution of CPs in unknown samples. This approach enables congener group-level

30

resolution of CP mixtures in environmental samples, and opens up the possibility for quantification of congener

31

groups.

ABSTRACT

32 33



34

Chlorinated paraffins (CPs) are mixtures of chlorinated n-alkanes. They are widely used product additives, with

35

numerous applications that include use as flame retardants and plasticizers in plastics, paints, sealants, and

36

adhesives, as coolants and lubricants in metal-cutting, and as fatliquors in leather processing.1 Global production

37

of CPs has increased to more than 1 million t/a in 2009.2 CP mixtures can be divided into short-chain (SCCPs,

38

C10 – C13), medium-chain (MCCPs, C14 – C17), and long-chain CPs (LCCPs, C≥18). The Persistent Organic

39

Pollutants (POPs) Review Committee of the UN Stockholm Convention has listed SCCPs as a POP-candidate,

40

citing evidence that SCCPs occur in remote areas such as the Arctic, are toxic to aquatic organisms, and are

41

potentially carcinogenic.3 However, scientific studies on the occurrence and behavior of CPs in humans and the

42

environment are still rare. The major reason for this paucity of data is the challenge posed by CP analysis, even

43

using modern separation and detection techniques.

INTRODUCTION

2 ACS Paragon Plus Environment

Page 2 of 22

Page 3 of 22

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

Analytical Chemistry

44

Technical CP mixtures contain thousands of congeners with the general elemental composition CnH2n+2-mClm

45

(hereafter abbreviated as CnClm). Complete separation of individual CPs has so far not been achieved,4 although

46

two dimensional gas-chromatography (GC×GC) techniques show considerable promise.5 The identification of

47

CP congener groups (i.e., CPs with a given number of carbon and chlorine atoms) relies on detection of pseudo-

48

molecular ions by soft ionization mass spectrometry.6 The large number of isomers, chlorine isotope patterns and

49

multiple fragmentations of different CP congeners all contribute to an extremely complex mass spectrum and

50

result in mass interference for pseudo-molecular as well as possible confirmation ions, both within and between

51

congener groups. High instrument resolving power is required to separate interfering ions of the same nominal

52

mass(i.e. isobaric mass).7 For instance, to separate [C10Cl8 – Cl] – (12C101H1435Cl637Cl.-, m/z 380.8886 amu) from

53

[C10Cl6 + Cl] – (12C101H1635Cl7.-, m/z 380.9072 amu), a resolving power of at least 20 500 (10% valley definition)

54

is required. The required resolving powers are significantly higher than those of commonly used magnetic sector

55

or time-of-flight (TOF) high-resolution instruments in routine resolution mode (typically 10 000). In addition,

56

certain fragment ion species, such as the three species [C10Cl8 – Cl]

57

possess identical chemical sum formulas (in this case, C10H14Cl7), and therefore cannot be resolved by mass

58

spectrometry, even with infinite resolving power.

59

Mathematical deconvolution techniques have been applied to resolve signals of high-mass molecules in the field

60

of protein mass spectrometry and to overcome the limitations of instruments with poor resolution.8-12

61

Deconvolution can allow unique identification of individual ion signals within a cluster of overlapping

62

masses.13,14 Some deconvolution methods determine the relative amounts of an individual ion by reconstructing

63

the unresolved mass spectra from the isolated isotopic patterns of the ion candidates.15 However, these

64

applications depend on knowledge of possible ion candidates, and complete knowledge of overlapping ions of

65

CPs is not available due to their compositional complexity and the lack of information about the content of

66

commercial CP products.

67

Recently, Bogdal and coworkers introduced a method for analysis of CP mixtures in environmental samples

68

based on the identification of chloride adducts by deconvolution of chlorine-enhanced atmospheric pressure

69

chemical ionization (APCI) mass spectra.7 However, possible mass interferences due to overlapping fragment

70

ions were not discussed. The present study is a logical extension of the method presented by Bogdal et al. Here,

– ,

[C10Cl7 – H] – and [C10Cl9 – 2Cl + H] –,

3 ACS Paragon Plus Environment

Analytical Chemistry

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

71

we describe a data processing approach to identify individual CnClm isobaric mass signals in two commonly

72

applied soft ionization mass spectrometry methods, electron capture negative ionization mass spectrometry

73

(ECNI-MS) and APCI-MS. For either ionization method, all possible fragment ions of C10 – C20 CPs can be

74

deduced progressively by mass spectral analysis of CP chemical standards. Isotopic distributions of all the

75

possible ions are enumerated to identify isobaric overlaps. Thereafter, CnClm (i.e., specific congener groups) can

76

be identified from the isobaric ion signals unresolved by the instrument. Typical applications are described and

77

requirements to extend the method to quantify concentrations of congener groups are identified.

78



79

CP Standards. We purchased eight individual CP congeners with defined molecular structure, including carbon

80

chain lengths 10, 11, 12 and 14, with 4 to 8 chlorine atoms. We also purchased three CP chain length standards

81

with a defined carbon chain length (C12) but varying Cl substitution. Finally, seven CP mixture standards with

82

variable chain length and chlorination degree between 42% and 63% were selected for this study. For a detailed

83

list of standards see the supporting information.

84

Instruments. Three different GC/ECNI-MS and one LC/APCI-MS were used in this study: (1) GC/ECNI-

85

quadrupole time-of-flight (QTOF) HRMS (Agilent 7200, Santa Clara, USA); (2) GC/ECNI-magnetic sector

86

HRMS (MS: Thermo Finnigan MAT95, Bremen, Germany); (3) GC/ECNI-quadrupole LRMS (Thermo DSQ II,

87

Bremen, Germany); (4) APCI-QTOF (Waters QTOF Premier, Manchester, UK). For detailed instrumental

88

methods see the supporting information. The GC was only used to separate solvents from the CP analytes, before

89

introducing the CPs into the ECNI-ion source. For the APCI system, the autoinjector of the LC system was used

90

for direct injection into the ion source without chromatography.

91

Data Acquisition. Mass spectra of the CPs using ECNI-QTOF MS and APCI-QTOF MS (full scan acquisition,

92

centroid mode, background subtracted) were obtained by the MassHunter B.06.00 and the MassLynx V4.1 data

93

system, respectively. The magnetic sector HRMS was operated in selected ion monitoring (SIM) mode using the

94

Xcalibur V2.2 data system. The flow injection peak of an individual m/z in APCI-MS was integrated by an

95

automatic algorithm to avoid inconsistencies of manual integration. For integration of “humps” resulting from

96

the gas chromatographic introduction see the supporting information including Figure S-1. In QTOF MS data

CHEMICALS AND INSTRUMENTS

4 ACS Paragon Plus Environment

Page 4 of 22

Page 5 of 22

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

Analytical Chemistry

97

acquisition, signals were recorded as centroid m/z including all ions that were unresolved within a given mass

98

range. The mass range can be calculated from the resolving power; for example, an observed centroid m/z

99

444.8706 amu using a QTOF instrument with a resolving power of 10 000 represents the ions within m/z

100

{444.8706 – 444.8706/10 000, 444.8706 + 444.8706/10 000} = {444.8261, 444.9151} amu.

101



102

We processed the data acquired for the three types of CP standards (individual CP congeners, CP chain length

103

standards, CP mixture standards) in three sequential steps with repeated application of deconvolution to find

104

isobaric overlaps that occur at each centroid m/z. The approach is summarized in the flowchart in Figure 1. For a

105

specific instrument, these three steps only need to be done once, then the output of step 3 can be applied for

106

resolving CnClm in unknown samples. The principles of the deconvolution and the data processing steps are

107

described below.

DATA PROCESSING APPROACH

5 ACS Paragon Plus Environment

Analytical Chemistry

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

STEP 1 INPUT Mass spectra of individual CP congeners

STEP 2 PROCESS INPUT Deconvolution (Equation 1) Identified fragment ions Fragment ion types

Mass spectra of reference standards: • CP chain length standards • SCCP mixture standards • MCCP mixture standards • LCCP mixture standards

PROCESS OUTPUT General fragmentation pathways

General fragmentation pathways Possible fragment ions of all CnClm Deconvolution (Equation 1)

STEP 3 INPUT

OUTPUT

Identified fragment ions

Identified fragment ions of all CnClm

PROCESS Isotope permutation Isotopic distributions of all ions

INPUT Resolving power of instrument

Isobaric overlaps checking

PROCESS

OUTPUT CnClm that have isobaric overlaps at each resolvable centroid mass

Isobaric overlaps to the confirmation ions of each CnClm

OUTPUT

108 109

Instrument-specific deconvolution matrix of each CnClm (Equation 1)

Figure 1. Flowchart summary of the data processing steps.

6 ACS Paragon Plus Environment

Page 6 of 22

Page 7 of 22

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

Analytical Chemistry

110

Principle of Deconvolution

111

We apply mathematical deconvolution to resolve individual abundances of overlapping ions. Its principle has

112

been described elsewhere,15,16 and is briefly presented here. The intensity of the m/z observed in mass spectra is

113

equal to the sum of the intensities of the individual overlapping ions at the same centroid m/z, which in matrix

114

notation is

115 116

 

  …  = … 





 …



…   …  ∙   →  =   ∙  … … …  … 

(1)

where {I} is a vector containing the intensities of measured m/z in the unresolved spectra (input), i is the number

117

of measured m/z, and j is the number of possible ions (i ≥ j). {A} is an i × j matrix containing the isotope pattern

118

abundances of all ions (calculated). Solution by least square optimization generates the relative amounts of

119

individual ions in overlapping mass spectra in the vector {X} (output, unknown of the equation system). For the

120

isotopic calculation (matrix A) and the least-square algorithm, see the supporting information.

121

Quality control of the deconvolution consists of mass spectrum reconstruction and centroid mass shift analysis.

122

The former calculates the percent difference value (DV) between the reconstructed and the measured mass

123

spectra in order to evaluate the performance of deconvolution. The equation used to determine DV is16

124

 = 100 ∙ 

125

Centroid mass shift analysis compares the observed centroid mass and the calculated one (based on the

126

knowledge of instrumental mass resolution) to verify the deconvolution result {X}. The calculated centroid mass

127

is approximated from the weighted average of the overlapping m/z.15

128

G/I =

129

where m/zj is the exact mass-to-charge ratio of the fragment ion j.

130

Data Processing Steps

131

Step 1: Deduction of General Fragmentation Pathways. The general fragmentation pathway for a molecule

132

specifies the fragment ions that could be observed using mass spectrometry. The first step in our approach is to

 ! "#$%&'($) *+,-./012,0+3 456789:6;