Determination by gas chromatography GC

Once extracted and purified from matrix interferences, the chromatographic separation of the target compounds has to take place before analysis. Due to the semi-volatility of dioxins and related compounds, GC is used to separate the different congeners and to allow non-ambiguous identifications.

4.3.1 Injection

This paragraph summarizes different injection techniques that have been reported for POPs analysis. Splitless injection is the most frequently used technique for detecting ultra-traces of dioxins, PCBs and OCPs [163]. It is a robust and reliable injection technique. Generally, a volume of maximum 2 p.L is injected into a hot injector left at a temperature between 250°C and 300°C. On-column injection may be used as an alternative for OCPs [164] in particular for DDT that is easily degraded in hot injector above 150°C. Large-volume injectors (LVI) are becoming more popular taken into account that the low concentrations of dioxins and furans occur in biological materials required a high sensitive technique. There are three ways of injecting large volume onto the GC column: on-column injection, loop-type injection and Programmable Temperature Vaporizer (PTV) injection. On-column LVI is a less successful way of injection for dioxins compared to PTV injector [165-167], particularly because of the strong influence of solvent-impurities in the chromatogram and contamination of the column and detector. These techniques enable injection of volume as large as 100-200 p.L, but most dioxin applications by PTV injection use 5-30 pL. The main drawback of PTV is the number of parameters to optimize compared to simple split/splitless technique.

4.3.2 GC Separation techniques

4.3.2.1 Classical gas chromatography (GC). Capillary high-resolution GC (HRGC) columns ensure the required selectivity, especially for congeners of the same chlorination level. The separation characteristics and elution profiles on GC columns have been studied. Ryan et al. [168] published a comprehensive study on separation characteristics and elution profiles of different GC columns. An appropriate combination of column length, internal diameter and stationary phase polarity is needed. For the analysis of biological samples, non-polar GC columns are usually used. They allow separation between homologue groups and can separate 2,3,7,8-substituted congeners from each other [169], showing the efficiency of a non-polar (methyl polysiloxane with 5% phenyl) column to separate the 2,3,7,8 toxic congeners from the rest, especially for 2,3,7, 8-tetrachlorodibenzofuran (TCDF). The apolar GC column is the most widely used column for 2,3,7,8-PCDD/Fs and non-ortho PCBs analysis in foodstuffs, feedingstuffs and human samples. Usually 40-60 m columns with 0.18-0.25 mm internal diameter and 0.15-0.25 mm film thickness are selected.

With polar phase (cyanopropyl stationary phases), the resolving power specially improves for the separation of the 22 TCDD congeners, but separation of 2,3,7,8-TCDF and 1,2,3,7,8-PeCDF still remains incomplete [170]. This column is used for environmental matrices (e.g., fly ash, air, soils, sediments and biota) characterized by the presence of many PCDD/Fs congeners. The major drawback of the polar column is the non-linked phase and its low stability at elevated temperature (withstand temperature up to 275°C), which tends to produce significant bleed. For PCB fractions, the 8% phenyl polycarborane-siloxane is often used. The carborane group has a high affinity for PCBs with a low degree of ortho-substitution. Although this phase does not allow the separation of all the 209 PCB congeners, it separates some critical pairs of co-elutions present with other phases [171]. For example, indicator trichlorinated PCB-28 and PCB-31 (not followed), pentachlorinated mono-ortho PCB-123 and PCB-118, as well as hexachlorinated PCB-163 (not followed) and the indicator hexachlorinated PCB-138 are separated. A 25 m x 0.25 mm x 0.2 mm HT-8 seems to be a good compromise between the required resolving power and the GC run time of roughly 30min. It should be mentioned that till date, none of the existing stationary phases is capable of the separation of all the PCB congeners. Even emerging hyphenated methods such as comprehensive two-dimensional GC (GC x GC) coupled to time-of-flight mass spectrometry (TOF-MS) can at the most separate 192 congeners [172].

4.3.2.2 Comprehensive Two-dimensional Gas Chromatography (GC x GC). GC x GC

has been developed to meet an increasing need for complex sample analysis and to address limitations, such as peak capacity, dynamic range and restricted specificity of mono-dimensional (classical) GC systems (i.e., to improve the global efficiency of the separation). It is a chromatographic technique where the entirety of a sample is subjected to two orthogonal separation processes. This is the case when two independent separation mechanisms are used (orthogonality rule), and when the separation obtained in the first dimension is maintained in the second dimension (conservation rule) [173]. This technique is based on the fast sampling and transfer of the sample by a cryogenic modulator located between the first dimension (1D) column and the second dimension (2D) column connected in series.

The entire 1D chromatogram is continuously sampled following a modulation period (PM) of a few seconds and sent into 2D for a fast GC-type separation (Figure 2A). Because modulation occurs during the :D separation, the total GC run time of a GC x GC separation is the same as in classical GC. As in classical GC, a trace is monitored continuously at the detector located at the end of 2D. In reality, series of high-speed secondary chromatograms of a length equal to PM (3-10 s) are recorded one after another (Figure 2B). They can be computerized by specific software and combined to describe the multi-dimensional elution pattern by means of contour plots in the chromatographic separation space (Figure 2C).

In GC x GC the peak capacity (the maximum number of peaks that can be placed next to each other in the available chromatographic separation space at a given resolution) is equal to the product of the peak capacities of the two coupled systems making GC x GC well suited to accommodate complex mixtures of compounds. Compared to classical GC, the analytical speed (number of compounds separated per unit of time) is greatly enhanced. GC x GC also improves specificity (2 retention times (tR)), selectivity (phase combination) and sensitivity (peak compression resulting in signal enhancement because of mass conservation) [175,176].

4.3.3 Gas chromatography and mass spectrometry (GC-MS)

It was clear over 30 years ago that GC-MS was the instrumental method of choice for POP determinations and specially for PCDD-Fs. MS provides not only a very specific quantification but also ensures the unambiguous identification of target compounds. However, not all GC/MS can measure dioxins. As already mentioned, the complexity to measure these compounds is also related to the low levels at which they occur in environmental matrices but particularly in food and feed samples (parts-per-quadrillion: 10_15g 2,3,7,8-TCDD/g of sample to parts-per-billion levels: 10_9g 2,3,7,8-TCDD/g of sample). The required sensitivity is achieved by a combination of high-resolution and high-mass accuracy using double focusing magnetic sector instruments called high resolution mass spectrometer (HRMS). In the past few years, other MS-based detection techniques have been investigated as alternative to GC-HRMS for dioxins and related compounds, such as GC-ion trap in MS/MS mode, GC x GC-TOF-MS and GC x GC-HRMS.

4.3.3.1 The reference GC-HRMS method. The GC-HRMS method has been recently reviewed by Eppe et al. [177] and Reiner et al. [178].

The extreme sensitivity of HRMS instrument is gained by the ability to monitor specific characteristic ions in the mass spectrum of the target compound. It is called the selected ion monitoring (SIM) mode. By this scan mode, a few femtograms of 2,3,7,8-TCDD injected in GC-HRMS can be detected. These performances are unmatched by any other techniques and it makes GC-HMRS the reference or the ''gold standard'' method for dioxins. Even after an extensive clean up and a high resolution chromatographic separation, the risk of interferences is still high. For that reason, the resolution R (R = M/LM) of the mass spectrometer should be set at least at 10,000 (10% valley definition). This allows mass discrimination at the 0.03-0.05 mass unit (Da) level in the tetra- to octa-substituted congeners mass range. The number of ions, which can be measured at any one time, is generally limited because at least ten sampling points for each GC peak are needed in order to get a Gaussian peak for accurate integration and quantification. Selected ions are therefore grouped in various segments. In addition, to control the mass accuracy, a lock mass is measured in each cycle and a lock mass check is often included (e.g., perfluorokerosene, PFK). The lock mass is ideally located within the measured mass range.

As mentioned, within each segment, the number of compounds (i.e., the number of specific ions) is limited. A quantification mass (the most abundant peak from the ions cluster) and a confirmatory mass (based on relative isotopic ratio of naturally occurring chlorinated isotopes) are required for native as well as for internal standard (Table 4). Thus, four masses are needed for one target congener. A maximum of three to four compounds can therefore be measured. In order to overcome this limitation, the chromatogram is sliced in time windows by grouping target compounds based on their retention time. The chromatographic challenge is then to bring the compound by groups (chromatographic windows) with no overlap, which would result in loss of congener measurement. Hence, one of the major drawbacks of SIM mode is the necessity to redefine windows any time the chromatographic parameters are modified (i.e., cutting or changing the column).

The presence of congeners affected by different TEFs leads to severe requirements for isomeric separation, which is relied on HRGC. The retention time of native and labeled standard peaks must be within a range of 2 s. To control the chlorination level and therefore the identity and the absence of interfering compounds, the measurement of the isotopic composition of the two most intense ions of both native and 13C-labeled ion clusters must be +15% <-

Figure 2 Scheme of the column coupling in the GC x GC setup and how data are handled (not to scale). (A) The modulator allows rapid sampling of the analytes eluting out of 1D and reinjection in 2D. The modulation process is illustrated for two overlapping compounds (X and Y) coming out of 1D at a defined first-dimension retention time (1tR). As the modulation process occurs during a defined modulation period (PM), narrow bands of sampled analytes are entering D and appear to have different second-dimension retention times (2tRX and 2tRY). (B) Raw data signal as recorded by the detector through the entire separation process. (C) Construction of the two-dimensional contour plot from the collected high-speed secondary chromatograms of (B), in which similar signal intensities are connected by contour lines. Reproduced from Ref. [174].

ic

pi

o

t

o

is

r

for

s io

s?

ti

m

a

1

r

s

io

l al

ti

a

ic

r

ti

ic

e r

pi

o

o

e

t

h

o

T

is

e

im

ti

n

a

c

s

er

s)

t In

(m

e

im

ti

el

w

d

n Io

(ms)

8

8

8

8

8

8

5

5

l

l

l

l

4

.8

.8

.8

.8

.8

.8

.7

.7

.7

.7

.7

.7

.9

i

i

i

i

i

i

i

i

i

i

i

i

i

5

5

5

5

5

5

6

6

3

3

3

3

9

.6

.6

.6

.6

.6

.6

.5

.5

.5

.5

.5

.5

.6

0.

0.

0.

0.

0.

0.

0.

0.

0.

0.

0.

0.

0.

7

7

7

7

7

7

4

4

l

l

l

l

l

.7

.7

.7

.7

.7

.7

.6

.6

.6

.6

.6

.6

.8

0.

0.

0.

0.

0.

0.

0.

0.

0.

0.

0.

0.

0.

55 l8

000 5025550 l5lllll

4I

4I

4I

+

+

+

2

2

£7

2

2

2

2

2

2

2

S S

I

I

I

I

I

I

I

I

I

I

I

I

I

I

I

4

6

4

6

9

5

8

5

7

2

7

9

6

8

5

2

2

2

l

l

6

6

7

7

9

2

2

7

7

8

2

6

8

0

4

9

3

7

l

7

6

0

5

9

3

.9

.9

.9

.9

.9

.8

.9

.8

.9

.9

.8

.9

.8

.8

.8

9.

l.

6.

3.

5.

9.

l.

7.

9.

0.

7.

9.

3.

5.

l.

8

0

l

0

l

l

3

2

3

3

3

4

5

6

6

2

3

3

3

3

3

3

3

3

3

3

3

3

3

3

2I

2I

2I

2I

2I

2I

6I

2I

2I

2I

2I

2I

2I

2I

i—i

S

S

S

S

S

S

i-i

S

S

S s

I

I

I

I

4

7

4

7

9

6

9

8

4

7

2

7

0

6

9

5

9

9

2

8

8

3

3

7

0

0

9

9

0

4

4

l

l

5

8

9

3

9

3

0

8

2

7

5

0

5

9

4

.9

.9

.9

.8

.9

.8

.9

.9

.8

.9

.9

.8

.9

.8

.8

.8

l.

3.

6.

5.

7.

l.

3.

l.

5.

7.

0.

9.

l.

5.

7.

9.

9

0

l

0

l

2

3

3

2

3

3

3

5

5

6

5

2

3

3

3

3

3

3

3

3

3

3

3

3

3

3

CC TT

DDD DDD CCCCC TTTTT

FF DF DF

CO CO v

01 01 o

Ph PH J

DDDD CCCC

01 m PH

FF DD

01 01 P

o

o

o

o

1-H

1-H

ON

ON

ON

ON

ON

<N

<N

<N

<N

O

p

p

p

1

1

1

1

1

1

1

1

1

1

1

1

1

ON

ON

ON

ON

ON

oo

oo

oo

oo

LC

LC

LC

LC

oo

oo

oo

co

ci

ci

ci

ci

ci

ci

ci

ci

ci

ci

ö

ö

ö

1-H

1-H

1-H

ON

ON

ON

ON

co

co

co

co

co

p

p

p

p

co

co

co

co

ö

ci

ö

Ö

ö

1-H

1-H

ci

Ö

Ö

'dh

'dh

'dh

'dh

_1_

_1_

^dh

_J_

£7

s s

£7

1-1

1-1

1-1

1-1

1-1

1-1

1-1

1-1

1-1

1-1

1-1

1-1

1-1

1-1

1-1

1-1

1-1

oo

o

oo

o

o

o

oo

o

o

oo

o

oo

o

oo

oo

oo

(N

CO

oo

ON

CO

(N

oo

(N

CO

(N

tv

LC

1-H

LC

tv

1-H

1-H

tv

CO

tv

^

oo

tv

0(3

ON

co

co

0(3

co

ON

0(3

oo

ON

IN.

tv

ON

ci

Ö

LC

1-H

ci

Ö

ON

1-H

LC

Ö

ON

ci

oo

oo

ON

o

oo

o

(N

(N

CO

CO

LC

■"t

LC

CO

CO

CO

CO

CO

CO

_(_

_J_

<N

<N

_1_

<N

£7

^7

s s

^7

^7

1-1

1-1

'-1

'-1

1-1

'-1

'-1

'-1

1—1

'—1

1—1

'—1

1—1

'—1

1—1

'—1

1—1

o

o

ON

o

oo

o

ON

oo

oo

O

oo

oo

1-H

o

LC

LC

1-H

(N

(N

LC

ON

O

(N

oo

(N

LC

oo

(N

CO

CO

oo

tv

oo

ON

co

oo

oo

oo

ON

oo

oo

ON

tv

tv

ON

1-H

c^

ci

Lfi

ON

Ö

ON

ci

Lfi

C^

ON

ci

Lfi

oo

oo

oo

o

oo

o

(N

CO

CO

LC

LC

CO

CO

CO

CO

CO

CO

U tn to

u^uuuu^yyyy^

x gx x x x gCPCPCPCPg

around the theoretical ion abundance ratio (Table 4). Any deviation out of this range will cause rejection of the congener's result.

The power of MS in quantitative analysis is, in fact, further enhanced by the isotopic dilution technique. This technique consists of spiking samples with an ideal internal standard, which is the isotopically labeled standard (e.g., 13C12 2,3,7,8-TCDD), showing almost identical characteristics to the compound of interest (e.g., 12C12 2,3,7,8-TCDD). The small mass difference (e.g., m/z 12) enables the discrimination between the compound of interest and its internal standard (Table 4). A calibration performed for all the PCDD/Fs and DL-PCBs with known amounts of native and internal standard congeners allows calculating the Relative Response Factor (RRF). The RRF takes into account the discrepancy that can be observed during MS ionization between natives and internal labeled standards. Thus, the RRF value directly affects the congener quantification as indicated in the following equation:

(Alabeled,i + Alabeled,,)x RRFi x m where [congener],- is the concentration of the congener i (e.g., ng/kg), areas Anative i and A2native , are the areas of the quantitation and confirmation ions for the native congener i, A1abeled , and A2abeled , the areas of the quantitation and confirmation ions for its corresponding labelled compound i, Q, is the amount of the corresponding recovery standard i spiked (e.g., ng) in the sample, RRF,- the relative response factor of the congener i and m the weight of the sample (e.g., kg).

4.3.3.2 GC-QIST/MS. Tandem MS as ion trap (Quadrupole ionstorage/MS, QIST/MS) MS/MS have been developed and used to analyse dioxins and related compounds. Dioxin analysis in MS/MS mode by ion trap MS is a wonderful application and illustration of ion trap theory [179-183]. Basically, the lack of selectivity in full scan mode with these benchtop instruments (i.e., mass unit resolution) is compensated by operating the instrument in MS/MS mode. The dioxin or furan precursor ion loses a fragment of COCl* which is characteristic of these molecules. No other similar halogenated organic compound fragments in this way, improving considerably the selectivity of the method. A global overview of MS/MS scan functions occurring in-time is shown in Figure 3 for the TCDD example [181]. The abscissa axis represents the time in milliseconds and the ordinate represents the amplitude of the voltages. RF corresponds to the potential applied to the ring electrode and the supplementary alternating voltages applied to the end-cap electrodes in dipolar fashion, which are referred to as waveforms. The first step consists of isolating the two most intense ions in the precursor ion cluster (e.g., m/z 320 and 322), that is the predominant transition [M]+* and [M+2]+*. Ions with m/z <320 are ejected using the mass selective axial instability mode by ramping the rf amplitude voltage. The ions' ejection is facilitated by the concurrent application of axial modulation with amplitude of 3 V. When ions of m/ z 320 arrive close to the instability region (qz = 0.908), the rf amplitude is modulated moderately in order to avoid

Cl 3

ipoa

^Cl 7Cl

Analytical RF ramp

Cl 3

^Cl 7Cl

Analytical RF ramp s

m/z 165

ionization!

m/z 165

m/z 280

Waveform

Broadband Waveform

CID Waveform -

Axial Modulation

Mass Analysis

Figure 3 Scan function for MS/MS of dioxin (TCDD). Reproduced from Ref. [181].

the ejection of the selected ions. Then, when ions of m/z < 320 are ejected, the rf is a little bit decreased and the ejection of ions with m/z >322 can start. It is achieved by applying a broadband waveform. Ions are ejected by matching the frequency (500 Hz steps) with the secular frequency of ions of a higher m/z ratio. Once isolation of the selected ions is completed, the rf voltage is dropped to obtain a qz value of 0.4. Ions of m/z 322 migrate on the left side of the axial qz axis to a more stable region of the stability diagram. In MS/MS, collision induced dissociation (CID) process can be affected in four modes: (1) single frequency irradiation (SFI), (2) multi-frequency irradiation (MFI), (3) secular-frequency modulation and (4) non-resonant excitation. The first three resonant modes were investigated and compared by Plombey and March [180] for dioxin application. They concluded that the tuning requirements of MFI and the duration of irradiation were compatible with the gas chromatographic time scale.

During CID process, PCDDs fragmentation is characterized by losses of Cl*, COCr, 2COCl* and PCDFs fragmentation by losses of Cl*, COCl*, COCl2 and COCl*. The main fragment used for quantification by isotopic dilution technique is the loss of COCl* for both PCDDs and PCDFs, whereas the loss of Cl2 characterizes the main fragment used for PCBs quantification.

s ionization!

4.3.3.3 Isotopic dilution technique for quantification by MS/MS technique. The principle is exactly the same as already reported for HRMS. The main difference is characterized here by the fact that instead of following the two most abundant molecular ions at 10,000 resolution in SIM mode, the two product ions [M+2—CO35Cl*] and [M+2—CO37Cl'] are monitored for both native and labelled molecular ions. It is called a selected reaction monitoring (SRM). Within a time window, the instrument alternatively scans the native and the label congener. Only 2,3,7,8-chloro-substituted congeners are followed. As for HRMS, the chromatogram is sliced into seven time windows from tetra-through-octa chlorinated dioxins and furans on a Rtx-5MS as shown in Figure 4 and Table 5. Compounds are numbered from 1 to 17 from TCDF to OCDF (see Table 5) for numbering correspondences. As native congeners co-elute with its corresponding labelled 13C12 isomer, the native peak maxima should fall within 3 s of their corresponding 13C labelled analogues for identification.

In windows 1, 2 and 4, the native compound and its corresponding labelled internal standard are monitored in MRM mode. For windows 3, 5, 6 and 7, MRM is performed by monitoring alternatively four molecular ions. After isolation, the molecular ion is replaced in the stability diagram at a value of qz — 0.45. The optimum voltage applied to the end-cap electrodes during CID varies between 5.5 and 6 V, whereas CID time has been optimized to 30 ms.

The specificity of MS/MS is achieved by monitoring two product ions and by checking their isotopic ratios. If the most abundant ion from the isotopic cluster is selected for MS/MS (i.e., [M+2] for TCDD), then the isotopic ratio for product ions does not follow the natural abundance of N-1 Cl (i.e., 3 Cl). Indeed, there is

Figure 4 Retention time of the seven toxic PCDDs and ten PCDFs on a Rtx5-MS column.
Table 5 Main parameters optimized for MS/MS analysis of dioxins and furans. The congener's classification corresponds to the elution order on Rtx5-MS 40 m column

Peak

Compounds

Window (min)

Molecular ions

CID (V)

Collision time (ms)

q value

Product ions

Isotopic ratios

1

2,3,7,8-TCDF

20-21.4

306

[M+2]

5.5

30

0.45

241/243

0.33

2,3,7,8-TCDF 13C12

318

[M+2]

5.5

30

0.45

252/254

0.33

2

2,3,7,8-TCDD

21.4—21.95

322

[M+2]

5

30

0.45

257/259

0.33

2,3,7,8-TCDD 13C12

334

[M+2]

5

30

0.45

268/270

0.33

3

1,2,3,7,8-

PeCDF

21.95-25.7

340

[M+2]

6

30

0.45

275/277

0.25

1,2,3,7,8-

PeCDF 13Ci2

352

[M+2]

6

30

0.45

286/288

0.25

4

2,3,4,7,8-

PeCDF

340

[M+2]

6

30

0.45

275/277

0.25

2,3,4,7,8-

PeCDF 13C12

352

[M+2]

6

30

0.45

286/288

0.25

5

1,2,3,7,8-

PeCDD

25.7-29

356

[M+2]

6

30

0.45

291/293

0.25

1,2,3,7,8-

PeCDD 13Ci2

368

[M+2]

6

30

0.45

302/304

0.25

6

1,2,3,4,7,

3-HxCDF

29-33.5

374

[M+2]

6

30

0.45

309/311

0.20

1,2,3,4,7,

3-HxCDF 13C12

386

[M+2]

6

30

0.45

320/322

0.20

7

1,2,3,6,7,

3-HxCDF

374

[M+2]

6

30

0.45

309/311

0.20

1,2,3,6,7,

3-HxCDF 13C12

386

[M+2]

6

30

0.45

320/322

0.20

8

2,3,4,6,7,

3-HxCDF

374

[M+2]

6

30

0.45

309/311

0.20

2,3,4,6,7,

3-HxCDF 13Ci2

386

[M+2]

6

30

0.45

320/322

0.20

9

1,2,3,4,7,

3-HxCDD

390

[M+2]

6

30

0.45

325/327

0.20

1,2,3,4,7,

3-HxCDD 13C12

402

[M+2]

6

30

0.45

336/338

0.20

10

1,2,3,6,7,

3-HxCDD

390

[M+2]

6

30

0.45

325/327

0.20

1,2,3,6,7,

3-HxCDD 13C12

402

[M+2]

6

30

0.45

336/338

0.20

11

1,2,3,7,8,

9-HxCDD

390

[M+2]

6

30

0.45

325/327

0.20

1,2,3,7,8,

9-HxCDD 113Ci2

402

[M+2]

6

30

0.45

336/338

0.20

12

1,2,3,7,8,

9-HxCDF

374

[M+2]

6

30

0.45

309/311

0.20

1,2,3,7,8,

9-HxCDF 13C12

386

[M+2]

6

30

0.45

320/322

TJ c

TJ c

O CL

o

o

o

o

o

o

CO

CO

CO

CO

^

CO

Ci

Ci

Ci

ci

ci

ci

ci

ci

ci

ci

ci

ci

ci

oo

CO

^

oo

oo

1-H

(N

LO

lO

LO

ON

o

oo

ON

CO

CO

CO

CO

CO

CO

CO

^

CO

CO

LC

lO

(N

LC

lO

LO

lO

ON

o

LO

lO

LO

ON

o

ON

CO

CO

CO

CO

CO

CO

CO

^

CO

CO

LC

LO

LO

LO

LO

LO

LO

LC

LC

LC

ci

ci

ci

ci

ci

ci

ci

ci

ci

ci

o

o

o

o

o

o

o

o

o

o

CO

CO

CO

CO

CO

CO

CO

CO

CO

LO co

s

s

o

(N O

(N

^ lO

(N lO

^ LC

0 0

Post a comment