Parallel segmented outlet flow high performance liquid chromatography with multiplexed detection

Analytica Chimica Acta 803 (2013) 154–159

Contents lists available at ScienceDirect

Analytica Chimica Acta journal homepage: www.elsevier.com/locate/aca

Parallel segmented outlet flow high performance liquid chromatography with multiplexed detection Michelle Camenzuli a , Jessica M. Terry b , R. Andrew Shalliker a,∗∗ , Xavier A. Conlan b , Neil W. Barnett b , Paul S. Francis b,∗ a Australian Centre for Research on Separation Science (ACROSS), School of Science and Health, University of Western Sydney (Parramatta), Sydney, NSW, Australia b Centre for Chemistry and Biotechnology, School of Life and Environmental Sciences, Deakin University, Geelong, Victoria 3216, Australia

h i g h l i g h t s

g r a p h i c a l

a b s t r a c t

• Multiplexed detection for liquid chromatography.

• ‘Parallel segmented outlet flow’ distributes inner and outer portions of the analyte zone. • Three detectors were used simultaneously for the determination of opiate alkaloids.

Parallel segmented outlet flow

HPLC pump

Detectors UV

Column CL1

Injector

CL2 Pump R1

a r t i c l e

i n f o

Article history: Received 10 February 2013 Received in revised form 28 March 2013 Accepted 2 April 2013 Available online 16 April 2013 Keywords: High performance liquid chromatography Parallel segmented outlet flow Chemiluminescence detection Opiate alkaloids

R2

a b s t r a c t We describe a new approach to multiplex detection for HPLC, exploiting parallel segmented outlet flow – a new column technology that provides pressure-regulated control of eluate flow through multiple outlet channels, which minimises the additional dead volume associated with conventional post-column flow splitting. Using three detectors: one UV-absorbance and two chemiluminescence systems (tris(2,2 bipyridine)ruthenium(III) and permanganate), we examine the relative responses for six opium poppy (Papaver somniferum) alkaloids under conventional and multiplexed conditions, where approximately 30% of the eluate was distributed to each detector and the remaining solution directed to a collection vessel. The parallel segmented outlet flow mode of operation offers advantages in terms of solvent consumption, waste generation, total analysis time and solute band volume when applying multiple detectors to HPLC, but the manner in which each detection system is influenced by changes in solute concentration and solution flow rates must be carefully considered. © 2013 Elsevier B.V. All rights reserved.

1. Introduction The scope of liquid chromatography as an analytical tool depends on the capabilities of both separation and detection [1]. Innovation in stationary phase technologies [2–5] and the application of multidimensional separations [6–9] continue to expand

∗ Corresponding author. Tel.: +61 3 5227 1294; fax: +61 3 5227 1040. ∗∗ Corresponding author. Tel.: +61 2 9685 9951; fax: +61 2 9685 9915. E-mail addresses: [email protected] (R.A. Shalliker), [email protected], [email protected] (P.S. Francis). 0003-2670/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.aca.2013.04.014

the limits of chromatographic selectivity and separation power, but further advances in the speed and comprehensiveness of analysis can be achieved by combining multiple modes of detection with distinct selectivities. Moreover, this approach enables quantitation to be combined with assessments of the molecular structure [10,11] or (bio)activity [12,13] of individual sample components. Non-destructive detectors (e.g. absorbance or fluorescence) can be connected in series with a single destructive mode of detection (such as mass spectrometry [14] or chemiluminescence [15]). However, each successive detector suffers from the cumulative band broadening imparted by all previous components and the extra column connections must be considered in order to correctly align

M. Camenzuli et al. / Analytica Chimica Acta 803 (2013) 154–159

the chromatographic response. Alternatively, by splitting the postcolumn stream using a T-piece [13], several modes of detection can be performed in parallel at the expense of greater sample dilution and extra-column band broadening. We have recently described a new column technology that separates the mobile phase eluting from the radial centre of the chromatographic bed from that eluting near the column wall [16,17]. The new design comprises an annular frit located at the column outlet inside a multi-port outlet end fitting. The frit has a solid PEEK ring that divides two permeable frit zones (central and peripheral) which direct the flow to either a single central outlet, or three peripheral exit ports depending on the region traversed by the mobile phase. The ratio of flow (central/peripheral) can be controlled by the relative pressure at each port. This ‘parallel segmented outlet flow’ has several major benefits: (1) the portion of each solute that has migrated through the more efficient central region of the column is separated from the trailing portion in the peripheral zone that is influenced by the wall effect. As a result, solutes that exit the central port elute with a higher number of theoretical plates compared to the band as a whole, producing narrower bands [17,18]; (2) without dilution from the peripheral zone, the leading portion of solute from the central region is more concentrated, which can improve detection sensitivity [17,18]; and (3) flow-limited detectors such as mass spectrometers can be employed in conjunction with high separation flow rates without the losses in sensitivity and efficiency associated with traditional post-column stream splitting [18]. To date, parallel segmented outlet flow columns have been used with a UV-absorbance or mass spectrometry detector connected through the central outlet port, with the solution from the peripheral region either directed to a collection vessel or monitored with a second UV-absorbance detector. However, the ability to control the solution flow through four separate outlets, without additional extra-column dead volume or post-column flow splitters, offers great potential for the incorporation of multiple detection systems, particularly considering that the portion of solute eluting from the peripheral region exhibits less on-column band broadening than the solute in its entirety [19]. In this paper, we explore the use of parallel segmented outlet flow for multiplexed detection, using the determination of six opium poppy (Papaver somniferum) alkaloids as a model system that exploits the complementary selectivity of two widely used chemiluminescence reagents [20–22]. Coupling these reagent systems is of interest in applications such as monitoring the extraction of opiate alkaloids from poppy straw [23,24], and rapid screening for heroin in suspected drug samples [25–27]. 2. Materials and methods 2.1. Chemicals and reagents Deionised water and analytical grade reagents were used unless otherwise stated. Chemicals were obtained from the following sources: sodium polyphosphate (+80 mesh) and trifluoroacetic acid from Sigma–Aldrich (NSW, Australia); sodium thiosulfate from Fluka (NSW, Australia); potassium permanganate from ChemSupply (SA, Australia); lead dioxide and sodium perchlorate from Ajax Finechem (NSW, Australia); methanol and sulfuric acid from Merck (Vic., Australia); glacial acetic acid and perchloric acid (70%, w/v) from Univar (NSW, Australia); acetonitrile from Burdick & Jackson (MI, USA) and tris(2,2 -bipyridine)ruthenium(II) dichloride hexahydrate from Strem Chemicals (MN, USA). Heroin (3,6-diacetylmorphine), codeine, morphine, oripavine, papaverine and thebaine were provided by GlaxoSmithKline (Vic., Australia). Stock solutions of the opiate alkaloids (1 mM) were prepared in acidified deionised water. Heroin (1 mM) was prepared in 0.1% (v/v) acetic acid and diluted in 0.05% (v/v) acetic acid.

155

The permanganate reagent was prepared by dissolving potassium permanganate (1.9 mM) in 1% (m/v) sodium polyphosphate, adjusting to pH 2.5 with sulfuric acid, and then adding sodium thiosulfate (0.6 mM), using a small volume of a 0.1 M solution [28]. The tris(2,2 -bipyridine)ruthenium(III) reagent was prepared by treating [Ru(bipy)3 ]Cl2 with sodium perchlorate in aqueous solution to yield a bright orange [Ru(bipy)3 ](ClO4 )2 precipitate, which was collected by vacuum filtration, washed twice with ice water, and dried over phosphorus pentoxide for 24 h [29]. The [Ru(bipy)3 ](ClO4 )2 crystals (1 mM) were then oxidised with lead dioxide (0.2 g/100 mL) in acetonitrile containing 0.05 M perchloric acid, which was observed as a change in the colour of the solution from orange to blue-green. The excess solid oxidant left in the reagent was prevented from entering the chemiluminescence detector by a filter (consisting of a small Pasteur pipette packed tightly with glass wool) fitted to the end of the tubing in the reagent reservoir. 2.2. High performance liquid chromatography (HPLC) Analyses were carried out on an Agilent Technologies 1200 series liquid chromatography system, equipped with a quaternary pump, solvent degasser system and autosampler, using a reversed phase Hypersil GOLD chromatography column (100 mm × 4.6 mm i.d., 5 ␮m, ThermoFisher Scientific, Cheshire, UK), with an injection volume of 20 ␮L, flow rate of 2.5 mL min−1 , and gradient elution using deionised water adjusted to pH 2.5 with trifluoroacetic acid (solvent A) and methanol (solvent B) as follows: 0–1 min: 5–10% B, 1–2 min: 10–25% B, 2–6 min: 25–35% B, 6–6.5 min: 35% B, 6.5–8 min: 35–5% B, 8–10 min: 5% B. Solvents and sample solutions were filtered through a 0.45 ␮m nylon membrane. 2.3. Conventional detection Separations were conducted with the column connected to each detection system individually. This included UV-absorbance at 280 nm (G1314A variable wavelength detector with standard flow-cell; 10 mm path length, 14 ␮L volume; Agilent Technologies), and two chemiluminescence detectors with permanganate and [Ru(bipy)3 ]3+ reagents, in which the column eluate (2.5 mL min−1 ) and reagent (1.0 mL min−1 ) merged immediately prior to or within the detection zone of the respective flow-cell (Fig. 1a). For the experiments with the permanganate reagent, we used a GloCel detector (Global FIA, WA, USA) with dual-inlet serpentine-channel flow-cell (fabricated from Teflon impregnated with glass microspheres) [30,31]. For experiments with the [Ru(bipy)3 ]3+ reagent, we used an in-house fabricated detector comprising a coil of transparent PTFE-PFA tubing (0.8 mm i.d.) that was glued into a spiral channel in an aluminium plate [31]. The flow-cell was mounted against an extended range photomultiplier module (Electron Tubes model P30A-05; ETP, NSW, Australia) within a light-tight housing. All tubing entering and exiting the detector was black PTFE (0.76 mm i.d., Global FIA). 2.4. Multiplexed detection Using the parallel segmented outlet flow column end-fitting (ThermoFisher Scientific), the column eluate (2.5 mL min−1 ) was divided in the following manner: 30% (0.74 mL min−1 ) was directed to the first chemiluminescence detector (permanganate reagent) via peripheral port 1; 30% (0.76 mL min−1 ) to the second chemiluminescence detector ([Ru(bipy)3 ]3+ reagent) via peripheral port 2; 27% (0.68 mL min−1 ) to the UV-absorbance detector via peripheral port 3; and the remaining 13% (0.32 mL min−1 ) to a collection vessel via the central port (Fig. 1b). The chemiluminescence

156

M. Camenzuli et al. / Analytica Chimica Acta 803 (2013) 154–159

(a)

(a) 10

O

Column

Detector CL1

Injector

Signal (arbitrary units)

8

HPLC pump

T

H P

6 C

M

Convenonal

4 O T

H

2

Pump

P C

M

Mulplexed

R1 0 0

(b)

1

2

3

(b) 15

4 Time (min)

5

6

7

8

O

Detectors

HPLC pump

CL1

Injector

CL2

Signal (arbitrary units)

UV Column

10

M

6-MAM Convenonal

5

Pump R1

O

M

6-MAM

R2

Mulplexed 0

Fig. 1. HPLC coupled with (a) a single chemiluminescence detector and (b) three detectors (UV-absorbance and two chemiluminescence systems) utilising parallel segmented outlet flow.

0

1

2

(c) 20

3. Results and discussion 3.1. Separation conditions We initially examined a synthetic mixture of heroin and five naturally occurring opium poppy alkaloids (morphine, codeine, oripavine, thebaine, and papaverine) using a reversed-phase column and UV-absorbance detection. Using gradient elution with a total mobile phase flow rate of 2.5 mL min−1 , the six analytes were separated within 6.5 min (Fig. 2a) whilst maintaining relatively low column pressure (<200 bar). The separation was repeated using post-column chemiluminescence detection, first with the permanganate reagent (Fig. 2b) and then with [Ru(bipy)3 ]3+ (Fig. 2c). A reagent flow rate of 1.0 mL min−1 afforded the greatest chemiluminescence responses. In agreement with previous investigations [20–22,32], the two reagents exhibited markedly different selectivities: the nonphenolic morphinan alkaloids codeine, thebaine and heroin produced peaks with [Ru(bipy)3 ]3+ , whilst only their phenolic counterparts (morphine and oripavine) elicited a response with permanganate. The small peak at 3.0 min in Fig. 2b was identified as 6-monoacetylmorphine (by comparing elution time with that of a separate standard solution), which was present in the opiate alkaloid mixture due to degradation of heroin. Calibration curves were prepared using the separation conditions described above with the three individual modes of detection (Table 1). Limits of detection (S N−1 = 3) using UV-absorbance were between 5 × 10−8 M and 5 × 10−7 M. The chemiluminescence systems were far more

4 Time (min)

5

6

7

8

C

16

Signal (arbitrary units)

reagents were pumped to their respective detectors at 1 mL min−1 . For these experiments the analyses were completed simultaneously (i.e. a single injection enabled the collection of data from the three modes of detection).

3

12

T H

Convenonal

8 C 4

T H

Mulplexed

0 0

1

2

3

4 Time (min)

5

6

7

8

Fig. 2. HPLC separation of a mixture of opium poppy alkaloids (2.5 × 10−6 M) and heroin (1.25 × 10−5 M) with (a) UV-absorbance at 280 nm, (b) permanganate chemiluminescence, and (c) tris(2,2 -bipyridine)ruthenium(III) chemiluminescence detection systems, performed separately (conventional) or simultaneously, using parallel segmented outlet flow (multiplexed). Peaks: morphine (M), codeine (C), oripavine (O), thebaine (T), heroin (H) and papaverine (P). The small peak at 3.0 min in (b) was identified as 6-monoacetylmorphine (6-MAM), present due to the degradation of heroin.

sensitive towards certain analytes, with limits of detection between 1 × 10−9 M and 5 × 10−9 M for non-phenolic morphinan alkaloids with [Ru(bipy)3 ]3+ and between 7.5 × 10−10 M and 1 × 10−9 M for the phenolic morphinan alkaloids using permanganate. One consequence of this greater sensitivity in the chemiluminescence modes was superior precision (as represented by lower relative standard deviations) for peak areas at 1 × 10−6 M. The calibration functions obtained with all modes of detection approximated linearity.

M. Camenzuli et al. / Analytica Chimica Acta 803 (2013) 154–159

157

Table 1 Analytical figures of merit for the determination of heroin and opium poppy alkaloids using HPLC with three individual modes of detection. Analyte

Calibration functiona

UV-abs. (280 nm)

Morphine Codeine Oripavine Thebaine Heroin Papaverine

y = 6.13 × 10 x + 0.028 y = 6.59 × 105 x − 0.089 y = 3.27 × 106 x − 0.347 y = 3.10 × 106 x + 0.002 y = 7.16 × 105 x − 0.039 y = 2.37 × 106 x − 0.092

0.9988 0.9996 0.9998 0.9985 0.9965 0.9999

CL (KMnO4 )

Morphine Oripavine

y = 9.04 × 108 x − 1.319 y = 1.13 × 109 x + 0.178

CL ([Ru(bipy)3 ]3+ )

Codeine Thebaine Heroin

y = 1.36 × 109 x + 0.863 y = 7.04 × 108 x + 5.190 y = 5.84 × 107 x + 0.678

Mode of detection

R2

% R.S.D. (area)b

% R.S.D. (r.t.)b

2.5 × 10 5.0 × 10−7 1.0 × 10−7 5.0 × 10−8 2.5 × 10−7 1.0 × 10−7

3.0 3.6 3.4 4.4 3.1 3.2

0.10 0.05 0.05 0.10 0.13 0.15

0.9998 0.9999

7.5 × 10−10 1.0 × 10−9

0.67 0.37

0.29 0.16

0.9993 0.9997 0.9994

1.0 × 10−9 2.5 × 10−9 5.0 × 10−9

0.57 0.89 0.74

0.09 0.19 0.06

LOD (M) −7

UV-abs.: 5 × 10−7 M to 7.5 × 10−6 M; CL: 5 × 10−8 M to 1 × 10−6 M. n = 5 (at 1 × 10−6 M).

The same separation was then performed utilising parallel segmented outlet flow, with approximately 30% of the eluate flow directed to each detector via the three peripheral outlets. The remaining solution was directed to a collection vessel via the central port. Splitting the flow at the end-column fitting to multiple detectors has two major effects: (1) the number of solute molecules entering each detector is reduced and (2) the flow rate through each detector is lower (in this case 0.7 mL min−1 compared to 2.5 mL min−1 ). Considering the UV-absorbance peak heights (Fig. 2a), the concentration of solute was also reduced under these flow splitting ratios. The lower post-column flow rate resulted in a slightly longer time for the solute band to reach the detector, as well as a longer residence time within the detector (see Fig. 3a). The increased residence time that results from the slower flow rate (0.7 mL min−1 ) results in an artificial broadening of the peak when compared to the separation in the conventional mode (at the higher flow rate of 2.5 mL min−1 ). A more complete visual representation can be seen by peak profiles illustrated in Fig. 3b, which represents the profiles as a function of peak volume, normalised to zero at the location of the peak maximum (aligned for visual clarity). When viewed in this context there is a substantial decrease in the volumetric peak width when a multiplex detection system is employed, collecting solute from the peripheral zone. The same tubing was used to couple these detectors to the traditional (single outlet) and parallel segmented outlet flow column-end-fittings, and therefore unlike conventional post-column flow-splitting approaches, additional contributions to post-column band-broadening were avoided [17]. It should be noted that even though post-column flow splitting is not required when using parallel segmented flow columns, care should be taken to minimise post-column dead volume since the peak width at elution following flow segmentation is greatly reduced. In the current set-up, only 27% of the flow entered the UV detector. This represents a substantial reduction in the peak volume since the flow to the detector was reduced by 73%. The origin of the tailing in the UV chromatographic profiles is likely caused by the large dead volume of the detector flow cell – 11 ␮L, since in essence, the parallel segmented flow column operated with a 27% outlet segmentation ratio operates in a manner akin to a 2.1 mm i.d. conventional column [33]. Studies recently conducted on a LC system with a reduced post column dead volume have shown almost no difference in peak symmetry between any of the outlet ports including the central port (data not yet published). The other two detection systems used in this study are based on fast chemiluminescence reactions [20,21], where the peak width may be less dependent on total residence time because the lightproducing reaction may be complete before the solution exits the detector [34,35], particularly under low flow rate conditions.

Therefore, although the lower post-column flow rate increased the time required for the solute to reach the detector in the parallel segmented flow mode, the peak widths were similar to those obtained with conventional detection (Fig. 4). The differences in peak heights and areas between the single and multiplexed (parallel segmented outlet flow) approaches arise from several factors, which in some cases are unique to this mode of detection: (i) The lower number of solute molecules entering the detector at a lower flow rate reduces the total number of emitting species and the frequency at which they are generated. (ii) Changes in flow rate affect the time required for the reacting mixture to travel from the confluence point to the detection zone and

(a) 0.9 0.8 0.7

Signal (arbitrary units)

3.2. Multiplexed detection

0.6 0.5 0.4

0.3 0.2 0.1 0 1.6

(b)

1.7

1.8

1.9 Time (min)

2

2.2

2.1

1 0.9 0.8 0.7

Signal (normalised)

a b

5

0.6 0.5 0.4 0.3 0.2 0.1 0 -0.25

-0.2

-0.15

-0.1

-0.05

0

0.05

0.1

0.15

0.2

0.25

0.3

Peak volume (mL)

Fig. 3. Peak shape according to (a) time and (b) volume, for morphine obtained using HPLC with UV-absorbance detection in single (solid line) or multiplexed (dotted line) systems.

158

M. Camenzuli et al. / Analytica Chimica Acta 803 (2013) 154–159

Table 2 Analytical figures of merit for the determination of heroin and opium poppy alkaloids using HPLC with parallel segmented outlet flow and multiplexed detection. Analyte

Calibration functiona

(UV-abs. (280 nm)

Morphine Codeine Oripavine Thebaine Heroin Papaverine

y = 6.83 × 10 x − 0.139 y = 6.57 × 105 x − 0.006 y = 3.26 × 106 x − 1.034 y = 2.89 × 106 x + 0.202 y = 5.94 × 105 x + 0.045 y = 2.45 × 106 x − 0.268

CL (KMnO4 )

Morphine Oripavine

CL ([Ru(bipy)3 ]3+ )

Codeine Thebaine Heroin

Mode of detection

a b

% R.S.D. (area)b

% R.S.D. (r.t.)b

0.9969 0.9986 0.9984 0.9961 0.9947 0.9985

−7

7.5 × 10 7.5 × 10−7 5.0 × 10−7 2.5 × 10−7 5.0 × 10−7 2.5 × 10−7

3.0 3.3 3.9 2.8 3.0 2.7

0.13 0.13 0.07 0.07 0.06 0.05

y = 1.41 × 108 x + 1.201 y = 1.38 × 108 x + 0.891

0.9999 0.9998

1.0 × 10−9 1.0 × 10−9

1.6 1.6

0.21 0.30

y = 5.05 × 108 x − 124.3 y = 5.29 × 108 x − 38.09 y = 2.85 × 107 x − 4.155

0.9999 0.9998 0.9956

1.0 × 10−8 1.0 × 10−8 5.0 × 10−8

1.3 1.1 1.2

0.09 0.06 0.07

5

LOD (M)

UV-abs. and CL ([Ru(bipy)3 ]3+ ): 5 × 10−7 M to 7.5 × 10−6 M; CL (KMnO4 ): 5 × 10−8 M to 1 × 10−6 M. n = 5 (at 1 × 10−6 M).

the efficiency of eluate/reagent mixing. Together with the kinetics of the reaction between the reagent and each analyte and the dimensions of the flow-cell, these factors determine the portion of the transient emission that is captured by the photodetector. In the GloCel detector used for permanganate chemiluminescence, the column eluate and reagent were merged within the detection zone, but this was not the case within the other detector. (iii) The relative flow rates of the column eluate and chemiluminescence reagent determine their volumetric ratio (and therefore influence their concentrations). Overall, these effects resulted in lower peak heights and peak areas for both chemiluminescence detection systems, by a factor of 1.4–3 for the [Ru(bipy)3 ]3+ reagent and 6–8 for the permanganate reagent (at 2.5 × 10−6 M; Fig. 2b and c), which translated to poorer limits of detection (Table 2). The discrepancy between the peak height/area reduction factors for the two systems can be ascribed to differences in the configurations of their detectors and the rates of their light-producing reactions. On the other hand, the simultaneous use of three modes of detection lowered sample/solvent consumption and waste generation, and reduced the total analysis time to a third of that required to run each mode of detection individually. The ability to maintain a consistent proportion of flow to the four end-column fitting outlets over an extended period of time is supported by the retention time precision (Tables 1 and 2), and an almost equivalent measure in precision of the peak areas at 1 × 10−6 M between conventional and parallel segmented modes of operation, where the small

1 0.9 0.8

Signal (arbitrary units)

R2

0.7 0.6 0.5 0.4 0.3

differences could be ascribed to the greater sensitivity obtained in the conventional mode of operation. 4. Conclusions Parallel segmented outlet flow clearly offers advantages in terms of solvent consumption, waste generation, total analysis time and the reduction in band volume when applying multiple detection systems to HPLC, but the manner that each detection system is influenced by changes in solute concentration and solution flow rates needs to be considered. Although advantageous for flow-limited detectors such as mass spectrometers, the reduced post-column flow rates imparted losses in sensitivity for detection systems based on chemiluminescence reactions, for which rapid and efficient mixing of relatively large volumes of reactants generate the greatest signal intensities. Here we chose two relatively fast chemiluminescence systems, for which high eluate flow rates are known to be advantageous [24], but irrespectively, impressive gains in analysis through-put were obtained with sufficient sensitivity. We anticipate that other commonly used chemiluminescence systems (such as the longer-lived luminol and peroxyoxalate reactions [15,36]) will be more influenced by total flow-cell volume and residence time, resulting in fewer losses or even improvements in sensitivity using the parallel segmented outlet flow approach. Nevertheless, even in the above example, the losses in emission intensity and limits of detection were less than an order of magnitude, and far exceeded by the greater inherent sensitivity of the chemiluminescence systems compared to UV-absorbance detection. This approach allows convenient manipulation of the ratio of solution distributed from the column, which we set at ∼30% (0.7 mL min−1 ) to each detector. The remaining solution (0.3 mL min−1 ) was sent directly to a collection vessel through the central port, but this also provides the opportunity to couple an additional detector such as a mass spectrometer to further enhance the effective peak capacity and total information derived from each chromatographic separation. Acknowledgements

0.2

0.1 0 1.5

1.6

1.7

1.8 1.9 Time (min)

2

2.1

2.2

Fig. 4. Peak profiles of morphine obtained using the permanganate chemiluminescence detection using an individual detector (solid line) or as part of multiplexed detection (dotted line). The multiplexed detection response was multiplied by a factor of 6 for comparison purposes. The fluctuations in the peak shape arise from the pulsing flow of the peristaltic pump that was used to propel the chemiluminescence reagent to the detector.

JMT acknowledges the support of an Australian Postgraduate Award. MC acknowledges the support of a UWS ACROSS Postgraduate Research Award. This research was supported by the Australian Research Council (DP0987318 and FT100100646). References [1] M. Swartz, J. Liq. Chromatogr. Relat. Technol. 33 (2010) 1130. [2] F. Svec, J. Chromatogr. A 1228 (2012) 250. [3] O. Nunez, H. Gallart-Ayala, C.P.B. Martins, P. Lucci, J. Chromatogr. A 1228 (2012) 298.

M. Camenzuli et al. / Analytica Chimica Acta 803 (2013) 154–159 [4] Y. Zhang, H. Luo, P.W. Carr, J. Chromatogr. A 1228 (2012) 110. [5] D. Guillarme, J. Ruta, S. Rudaz, J.-L. Veuthey, Anal. Bioanal. Chem. 397 (2010) 1069. [6] D.R. Stoll, Anal. Bioanal. Chem. 397 (2010) 979. [7] H. Malerod, E. Lundanes, T. Greibrokk, Anal. Methods 2 (2010) 110. [8] I. Francois, K. Sandra, P. Sandra, Anal. Chim. Acta 641 (2009) 14. [9] D.R. Stoll, X. Li, X. Wang, P.W. Carr, S.E.G. Porter, S.C. Rutan, J. Chromatogr. A 1168 (2007) 3. [10] H.H. Maurer, Anal. Bioanal. Chem. 388 (2007) 1315. [11] R. Brkljaca, S. Urban, J. Liq. Chromatogr. Relat. Technol. 34 (2011) 1063. [12] H.A.G. Niederländer, T.A. van Beek, A. Bartasiute, I.I. Koleva, J. Chromatogr. A 1210 (2008) 121. [13] M. Mnatsakanyan, T.A. Goodie, X.A. Conlan, P.S. Francis, G.P. McDermott, N.W. Barnett, D. Shock, F. Gritti, G. Guiochon, R.A. Shalliker, Talanta 81 (2010) 837. [14] X.A. Conlan, in: R.A. Shalliker (Ed.), Hyphenated and Alternative Methods of Detection in Chromatography, CRC Press, Boca Raton, FL, 2011, p. 1. [15] P.S. Francis, J.L. Adcock, in: R.A. Shalliker (Ed.), Hyphenated and Alternative Methods of Detection in Chromatography, CRC Press, Boca Raton, FL, 2012, p. 221. [16] M. Camenzuli, H.J. Ritchie, J.R. Ladine, R.A. Shalliker, Analyst 136 (2011) 5127. [17] M. Camenzuli, H.J. Ritchie, J.R. Ladine, R.A. Shalliker, J. Chromatogr. A 1232 (2012) 47. [18] M. Camenzuli, T.A. Goodie, D.N. Bassanese, P.S. Francis, N.W. Barnett, H. Ritchie, J. LaDine, R.A. Shalliker, X.A. Conlan, Rapid Commun. Mass Spectrom. 26 (2012) 943. [19] M. Camenzuli, H.J. Ritchie, J.R. Ladine, R.A. Shalliker, J. Liq. Chromatogr. Relat. Technol. 36 (2013) 1379. [20] B.A. Gorman, P.S. Francis, N.W. Barnett, Analyst 131 (2006) 616.

159

[21] J.L. Adcock, P.S. Francis, N.W. Barnett, Anal. Chim. Acta 601 (2007) 36. [22] P.S. Francis, J.L. Adcock, J.W. Costin, S.D. Purcell, F.M. Pfeffer, N.W. Barnett, J. Pharm. Biomed. Anal. 48 (2008) 508. [23] Y. Gao, Q. Xiang, Y. Xu, Y. Tian, E. Wang, Electrophoresis 27 (2006) 4842. [24] J.W. Costin, S.W. Lewis, S.D. Purcell, L.R. Waddell, P.S. Francis, N.W. Barnett, Anal. Chim. Acta 597 (2007) 19. [25] K.M. Agg, A.F. Craddock, R. Bos, P.S. Francis, S.W. Lewis, N.W. Barnett, J. Forensic Sci. 51 (2006) 1080. [26] L.A. Hill, C.E. Lenehan, P.S. Francis, J.L. Adcock, M.E. Gange, F.M. Pfeffer, N.W. Barnett, Talanta 76 (2008) 674. [27] Y. Zhuang, D. Zhang, H. Ju, Analyst 130 (2005) 534. [28] P.S. Francis, C.M. Hindson, J.M. Terry, Z.M. Smith, T. Slezak, J.L. Adcock, B.L. Fox, N.W. Barnett, Analyst 136 (2011) 64. [29] G.P. McDermott, P. Jones, N.W. Barnett, D.N. Donaldson, P.S. Francis, Anal. Chem. 83 (2011) 5453. [30] J.M. Terry, J.L. Adcock, D.C. Olson, D.K. Wolcott, C. Schwanger, L.A. Hill, N.W. Barnett, P.S. Francis, Anal. Chem. 80 (2008) 9817. [31] J.M. Terry, S. Mohr, P.R. Fielden, N.J. Goddard, N.W. Barnett, D.C. Olson, D.K. Wolcott, P.S. Francis, Anal. Bioanal. Chem. 403 (2012) 2353. [32] G.M. Greenway, A.W. Knight, P.J. Knight, Analyst 120 (1995) 2549. [33] R.A. Shalliker, M. Camenzuli, L. Pereira, H.J. Ritchie, J. Chromatogr. A 1262 (2012) 64. ˜ W.R.G. Baeyens (Eds.), [34] D. Pérez-Bendito, M. Silva, in: A.M. García-Campana, Chemiluminescence in Analytical Chemistry, Marcel Dekker, New York, 2001, p. 175. [35] J.M. Terry, E.M. Zammit, T. Slezak, N.W. Barnett, D.C. Olson, D.K. Wolcott, D.L. Edwards, P.S. Francis, Analyst 136 (2011) 913. [36] A.G. Hadd, A. Seeber, J.W. Birks, J. Org. Chem. 65 (2000) 2675.