**2. Applications of MSFIA to environmental monitoring**

MSFIA systems have been successfully applied to the determination of more than 20 species in environmental samples as illustrated in Tables 1 and 2.

Several applications were targeted to plant macronutrients, such as potassium and phosphorus, and also to micronutrients, including boron, iron and selenium (Table 1). These species were quantified in different types of water (natural and seawater) and also in soil extracts or even soil slurries when applying flame emission spectrometry for determination of potassium (Almeida et al., 2008). The introduction of aqueous samples in flow systems is rather trivial, while manipulation of samples containing suspended solids is not common, requiring a special manifold design employing larger commutation valves and large bore tubing (Almeida et al., 2008).

In fact, solid environmental samples were successfully handled within MSFIA systems. Extraction of potassium contained in 1.8 g of soil was performed in-line, prior to potentiometric determination. The soil was placed in a container where 9 ml of Morgan extractant solution was delivered automatically by one of the syringes. After 6 min, in-line filtration took place, and a small portion of filtrate (100 µl) was sent to the potentiometric detector after in-line addition of an ionic strength adjusting buffer. Different soils were analyzed consecutively without carry-over effects and the filtration unit was reutilized up to 10 times (Almeida et al., 2006).

Besides the determination of total extractable content, MSFIA systems have been employed to dynamic fractionation testing schemes, profiting from its inherent capabilities of controllable flow programming. In fact, Miró and co-workers developed a multiple stirredflow chamber assembly, containing up to three parallel extractors, to perform sequential extraction of readily mobilizable fractions of trace elements (Cu, Cd, Ni, Pb, Zn) in fly ashes (Boonjob et al., 2008). Though the detection step was performed off-line (not automated) on each 10 ml fraction collected, the MSFIA system still provided information about overall extractable pools in less than 2 hours, a drastic reduction of time when compared to 18 to 24 hours required per fraction in equilibrium leaching tests. Moreover, the implementation of a sequential leaching scheme was easily accommodated in MSFIA, due to its inherent flow features and also by housing different extracting solutions (water, 0.11 M acetic acid, 0.11 M acetic acid/acetate buffer) simultaneously in each syringe of the multisyringe burette.

Multisyringe Flow Injection

immediately after extraction.

2005a, 2005b) as far as nutrient analysis is concerned.

in water and soil samples (Oliveira et al., 2009).

al., 2006b) and determination of sulphide (Ferrer et al., 2005b).

Analysis for Environmental Monitoring: Applications and Recent Trends 287

after derivatization using 1,5-diphenylcarbazide and in-line adjustment of Cr(VI) concentration prior to determination. This last feature was attained by incorporating a dilution chamber (for extracts from highly contaminated samples) and another flowthrough column filled with multi-walled carbon nanotubes for preconcentration of Cr-1,5 diphenylcarbazide (for extracts or fractions with low Cr(VI) concentration). The MSFIA system was successfully applied to SRM 2701 soil, providing the extraction kinetics for sequential extraction using water and an acid rain surrogate. The automatic integration of extraction and Cr(VI) determination also allowed the minimization of interconversion between Cr oxidation states often observed when determination is not carried out

In fact, several MSFIA systems were developed for monitoring of environmental pollutants as indicated in Table 2. Both organic and inorganic species were targeted, with a focus on water analysis. In this context, in-line sample treatment is undeniably a requisite when devising monitoring schemes with real environmental samples. There are two main reasons for this. First, analytes, particularly pollutants, are generally present at low concentrations (ppt or ppb range) in these samples, requiring a preconcentration step in order to meet the linear working range offered by the available detection systems. Secondly, the target analyte may be strongly bound or entrapped in the sample matrix or it can present different forms concerning its oxidation state. Hence, solid phase extraction has been frequently implemented in-line, aiming the enrichment and selective uptake of analytes. It has been applied for the determination of trace levels of phosphate (5 – 50 µg l-1 of P) in natural waters combined to chemiluminescence detection (Morais et al., 2004), for determination of selenium (5.7 - 1290 µg l-1) in natural and seawater (Serra et al., 2010) and for determination of trace iron (0.05 – 8 µg l-1; 0.2 – 42 µg l-1) in waters (Pascoa et al., 2009; Pons et al., 2004,

Solid phase extraction has also been employed in more than half of the applications focusing on pollutants monitoring and it was implemented in several ways. In flow injection systems, sorbents are generally packed in flow-through columns, which are sequentially percolated by conditioning solution, sample, washing solution and eluent, fostering selective retention of target analyte(s), followed by its/their elution after sample matrix removal. This strategy has also been implemented in MSFIA for determination of total phenolics in waters, using Amberlite XAD-4 as sorbent and in-line derivatization with 4-aminoantipyrine (Oliveira et al., 2005). This MSFIA system was further improved and coupled to liquid chromatography, allowing on-line preconcentration and determination of eleven priority phenolic pollutants

Extraction membranes, containing different functional groups, have also been employed in MSFIA systems as they are an advantageous alternative to particulate sorbents because they allow higher flow rates (providing high determination throughputs) and low backpressure, avoiding leakages and clogging. Several applications have been reported, namely for preconcentration of nitrophenols and their determination after elution (Manera et al., 2007a) or using optosensing (Manera et al., 2007b) by probing the extraction membrane with a bifurcated optical fiber connected to a CCD spectrometer. In fact, the utilization of optosensors in MSFIA systems is simplified because all solutions required (sample, conditioning and regenerating solutions) can be automatically delivered by the multisyringe burette in a precise and timely way. Hence, optosensing has also been applied in MSFIA systems for trace level determination of 1-naphthylamine in water samples (Guzmán-Mar et


Table 1. MSFIA methods for nutrient assessment and monitoring

The same group proposed a fully automated strategy for fractionation of orthophosphate in soil samples and in-line spectrophotometric determination resorting to molybdenum blue reaction (Buanuam et al., 2007). The system integrated dynamic sequential extraction using 1 M ammonium chloride solution, 0.1 M sodium hydroxide solution and 0.5 M hydrochloric acid solution as extractants according to the Hieltjes–Lijklema scheme. Solid soil samples were placed in a flow-through customized dual conical chamber (Chomchoei et al., 2004), which could be filled with up to 300 mg of soil. Compared to conventional batch equilibrium procedures, the automatic dynamic fractionation system offered further knowledge on: (i) the extraction kinetics, (ii) the content of phosphorus in available pools, (iii) the efficiency of the leachants and (iv) the actual extractant volume required for quantitative release of orthophosphate. The automatic MSFIA extraction scheme was also validated through application to certified reference material SRM 2704 river sediment and SRM 2711 Montana soil.

Recently, a similar strategy was applied for automated dynamic extraction and determination of readily bioaccessible chromium(VI) in soils. Besides the extraction capabilities, the automatic MSFIA system also fostered in-line quantification of Cr(VI)

(available) Soil extract --- 15 (Gomes et al., 2005) Iron (available) Soil extract --- 34 (Gomes et al., 2005)

Fe(III) Water --- 60 (Pons et al., 2004)

(available) Soil extract --- 15 (Almeida et al., 2005)

Potassium Soil slurries --- 28 (Almeida et al., 2008) Selenium Sea lettuce --- 84 (Semenova et al.,

The same group proposed a fully automated strategy for fractionation of orthophosphate in soil samples and in-line spectrophotometric determination resorting to molybdenum blue reaction (Buanuam et al., 2007). The system integrated dynamic sequential extraction using 1 M ammonium chloride solution, 0.1 M sodium hydroxide solution and 0.5 M hydrochloric acid solution as extractants according to the Hieltjes–Lijklema scheme. Solid soil samples were placed in a flow-through customized dual conical chamber (Chomchoei et al., 2004), which could be filled with up to 300 mg of soil. Compared to conventional batch equilibrium procedures, the automatic dynamic fractionation system offered further knowledge on: (i) the extraction kinetics, (ii) the content of phosphorus in available pools, (iii) the efficiency of the leachants and (iv) the actual extractant volume required for quantitative release of orthophosphate. The automatic MSFIA extraction scheme was also validated through application to certified reference material SRM 2704 river sediment and

Recently, a similar strategy was applied for automated dynamic extraction and determination of readily bioaccessible chromium(VI) in soils. Besides the extraction capabilities, the automatic MSFIA system also fostered in-line quantification of Cr(VI)

Solid phase

Solid phase

Determination

extraction 12 (Pascoa et al., 2009)

extraction 5 - 10 (Pons et al., 2004,

extraction 11 (Morais et al., 2004)

extraction Not given (Buanuam et al., 2007)

digestion 12 (Almeida et al., 2004)

in-line filtration 13 (Almeida et al., 2006)

extraction 8 (Serra et al., 2010)

2003)

throughput (h-1) Reference

2005a, 2005b)

treatment

Analyzed

Iron (total) and

Iron (total) and

Phosphorus

Fe(III)

Boron

species Sample type In-line sample

Iron (total) Water Solid phase

Water (natural and seawater)

Phosphate Soil extract In-line sequential

Phosphate Water Solid phase

Phosphorus Water Microwave

Selenium Water (natural

SRM 2711 Montana soil.

Potassium Soil extracts Extraction and

and seawater)

Table 1. MSFIA methods for nutrient assessment and monitoring

after derivatization using 1,5-diphenylcarbazide and in-line adjustment of Cr(VI) concentration prior to determination. This last feature was attained by incorporating a dilution chamber (for extracts from highly contaminated samples) and another flowthrough column filled with multi-walled carbon nanotubes for preconcentration of Cr-1,5 diphenylcarbazide (for extracts or fractions with low Cr(VI) concentration). The MSFIA system was successfully applied to SRM 2701 soil, providing the extraction kinetics for sequential extraction using water and an acid rain surrogate. The automatic integration of extraction and Cr(VI) determination also allowed the minimization of interconversion between Cr oxidation states often observed when determination is not carried out immediately after extraction.

In fact, several MSFIA systems were developed for monitoring of environmental pollutants as indicated in Table 2. Both organic and inorganic species were targeted, with a focus on water analysis. In this context, in-line sample treatment is undeniably a requisite when devising monitoring schemes with real environmental samples. There are two main reasons for this. First, analytes, particularly pollutants, are generally present at low concentrations (ppt or ppb range) in these samples, requiring a preconcentration step in order to meet the linear working range offered by the available detection systems. Secondly, the target analyte may be strongly bound or entrapped in the sample matrix or it can present different forms concerning its oxidation state. Hence, solid phase extraction has been frequently implemented in-line, aiming the enrichment and selective uptake of analytes. It has been applied for the determination of trace levels of phosphate (5 – 50 µg l-1 of P) in natural waters combined to chemiluminescence detection (Morais et al., 2004), for determination of selenium (5.7 - 1290 µg l-1) in natural and seawater (Serra et al., 2010) and for determination of trace iron (0.05 – 8 µg l-1; 0.2 – 42 µg l-1) in waters (Pascoa et al., 2009; Pons et al., 2004, 2005a, 2005b) as far as nutrient analysis is concerned.

Solid phase extraction has also been employed in more than half of the applications focusing on pollutants monitoring and it was implemented in several ways. In flow injection systems, sorbents are generally packed in flow-through columns, which are sequentially percolated by conditioning solution, sample, washing solution and eluent, fostering selective retention of target analyte(s), followed by its/their elution after sample matrix removal. This strategy has also been implemented in MSFIA for determination of total phenolics in waters, using Amberlite XAD-4 as sorbent and in-line derivatization with 4-aminoantipyrine (Oliveira et al., 2005). This MSFIA system was further improved and coupled to liquid chromatography, allowing on-line preconcentration and determination of eleven priority phenolic pollutants in water and soil samples (Oliveira et al., 2009).

Extraction membranes, containing different functional groups, have also been employed in MSFIA systems as they are an advantageous alternative to particulate sorbents because they allow higher flow rates (providing high determination throughputs) and low backpressure, avoiding leakages and clogging. Several applications have been reported, namely for preconcentration of nitrophenols and their determination after elution (Manera et al., 2007a) or using optosensing (Manera et al., 2007b) by probing the extraction membrane with a bifurcated optical fiber connected to a CCD spectrometer. In fact, the utilization of optosensors in MSFIA systems is simplified because all solutions required (sample, conditioning and regenerating solutions) can be automatically delivered by the multisyringe burette in a precise and timely way. Hence, optosensing has also been applied in MSFIA systems for trace level determination of 1-naphthylamine in water samples (Guzmán-Mar et al., 2006b) and determination of sulphide (Ferrer et al., 2005b).

Multisyringe Flow Injection

16 ng l-1.

Analysis for Environmental Monitoring: Applications and Recent Trends 289

MSFIA systems were also devised for speciation of arsenic (Leal et al., 2006b) and mercury (Serra et al., 2009) in environmental samples. For arsenic, As(III) was quantified by atomic fluorescence spectrometry while As(V) was assessed by difference from total As content, determined for the same sample after in-line reduction of As(V) to As(III) by automatic addition of potassium iodide and ascorbic acid. For mercury, atomic fluorescence spectrometry was also employed and speciation between inorganic and organic (methylmercury) forms was performed by selectively retaining mercury tetrachloro complex in an anion exchange membrane, while organic mercury was directed towards a flow-through UV digestor before detection. Inorganic mercury was later eluted from the membrane by in-line reduction with tin chloride, allowing a limit of detection of

Besides the reduction of intervention from laboratory technicians in the analytical operations, automation of environmental assays by MSFIA provided also an acceptable determination throughput, ranging from 7.5 to 108 determinations per hour in automatic systems where no sample treatment was required or where it was performed off-line. For MSFIA systems comprising in-line sample treatment, determination throughputs ranged from 3 to 30 determinations per hour, which are excellent figures. For instance, the determination of total phosphorus in waste water samples was carried out with a determination throughput of 12 determinations per hour by implementing in-line microwave digestion of samples (Almeida et al., 2004). This is a significant reduction of the assay time when compared to the conventional

As mentioned before, environmental samples comprise complex matrices where target analytes are not generally amenable to direct determination by instrumental analysis, requiring sample treatments. Regarding this aspect, solid phase extraction is undoubtedly the most common treatment applied in MSFIA systems as shown in Tables 1 and 2. Besides the examples presented before in the text, MSFIA capabilities have been recently exploited to perform solid phase extraction using bead injection (BI) prior to chromatographic analysis. The bead injection concept consists of handling solid suspensions in a fully automatic fashion, where the solid-phase sorbent, presented as micrometric beads, is renewed in each individual analytical cycle, rendering a fresh portion of sorbent for each analysis. Moreover, bead injection allows the simultaneous monitoring of both effluent and solid phase itself (optosensing) in real time, which brings complementary and enhanced insight into the solid

The bead injection concept is often associated to the lab-on-valve (LOV) platform. The LOV module comprises a monolithic structure with microconduits machined in a polymethylmethacrylate or polyetherimide unit, which is mounted atop a multiposition valve (Fig. 2), representing a step forward towards automation and miniaturization of flow injection systems. The LOV-BI approach offers two main advantages, not matched by any other automatic, flow-based solid phase extraction scheme: (i) the automatic renewal of sorbent, without any intervention of operator or replacement of devices or physical parts of the system, so as to circumvent the progressive deactivation and tighter packing of permanent in-line solid phase extraction cartridges; and (ii) the accurate metering of sorbent and eluate quantities by resorting to bi-directional programmable flow, as precisely

batch digestion that took about 2 hours for quantitative measurements.

phase extraction procedure in a single assay (Gutzman et al., 2006).

controlled by the multisyringe burette (Miró et al., 2011).

**3. Recent trends for sample treatment** 


Table 2. MSFIA methods for assessment and monitoring of pollutants

**Determination** 

extraction 9 (Long et al., 2006)

extraction Not given (Boonjob et al., 2011;

extraction Not given (Rosende et al., 2011)

extraction 9 (Maya et al., 2008)

extraction 30 (Serra et al., 2008)

extraction 14 (Serra et al., 2009)

extraction <sup>14</sup>(Guzmán-Mar et al.,

extraction 11 (Miró et al., 2001)

extraction 12 (Maya et al., 2010)

extraction 4 – 16 (Oliveira et al., 2005)

extraction 4 – 10 (Oliveira et al., 2009)

extraction Not given (Quintana et al., 2009)

extraction 5 - 8 (Ferrer et al., 2005a,

extraction 12 (de Armas et al., 2002)

extraction 7 (Oliveira et al., 2010)

extraction Not given (Quintana et al., 2006)

and liver --- 108 (Semenova et al., 2002)

**throughput (h-1) Reference** 

2009)

2006a)

2006b)

13 - 20 (de Armas et al., 2004; Maya et al., 2007)

2005b; Ferrer et al., 2006)

(Horstkotte et al., 2008; Manera et al., 2007a; Manera et al., 2007b)

Boonjob et al., 2010)

**treatment**

inorganic) and As(III) Water, sea lettuce --- 10 (Leal et al., 2006b) Azinphos methyl Water Hydrolysis 7 (Ornelas-Soto et al.,

Solid phase

Solid phase

Mercury Water, fish muscle --- 44 (Leal et al., 2006a)

Solid phase

1-Naphthylamine Water --- 90 (Guzmán-Mar et al.,

Solid phase

Solid phase

Solid phase

Sulphide Water --- 45 (Ferrer et al., 2004)

Solid phase

Sulphonated azo dyes Water --- 7.5 (Fernandez et al., 2010)

Solid phase

Analyte separation by gas diffusion

extraction 3 - 11

**Analyzed species Sample type In-line sample** 

Water and soil extracts

Water and leachates

leachates

and organic) Water, fish muscle Solid phase

1-Naphthylamine Water Solid phase

Nitrophenols Water Liquid-liquid

residues (NSAIDs) Water Solid phase

(total) Water Solid phase

Phenolic compounds Water and soil Solid phase

Solid waste leachates

Waters (fresh, seawater and wastewater)

Water (seawater and swimming

Table 2. MSFIA methods for assessment and monitoring of pollutants

Warfarin Water Solid phase

pool)

Water, seawater and waste leachates

Water and solid waste leachates

Arsenic Water Solid phase

Chromium(VI) Soil leachates In-line

Arsenic Water, fish muscle

Arsenic (total

Chlorotriazine herbicides

carbons

Halogenated organic

Mercury (inorganic

Nitrophenols

Pharmaceutical

Pharmaceutical residues (thiazide diuretics)

Polychlorinated biphenyls

Sulphide

UV filters

Phenolic compounds

Sulphide Wastewater

Mercury Water and

MSFIA systems were also devised for speciation of arsenic (Leal et al., 2006b) and mercury (Serra et al., 2009) in environmental samples. For arsenic, As(III) was quantified by atomic fluorescence spectrometry while As(V) was assessed by difference from total As content, determined for the same sample after in-line reduction of As(V) to As(III) by automatic addition of potassium iodide and ascorbic acid. For mercury, atomic fluorescence spectrometry was also employed and speciation between inorganic and organic (methylmercury) forms was performed by selectively retaining mercury tetrachloro complex in an anion exchange membrane, while organic mercury was directed towards a flow-through UV digestor before detection. Inorganic mercury was later eluted from the membrane by in-line reduction with tin chloride, allowing a limit of detection of 16 ng l-1.

Besides the reduction of intervention from laboratory technicians in the analytical operations, automation of environmental assays by MSFIA provided also an acceptable determination throughput, ranging from 7.5 to 108 determinations per hour in automatic systems where no sample treatment was required or where it was performed off-line. For MSFIA systems comprising in-line sample treatment, determination throughputs ranged from 3 to 30 determinations per hour, which are excellent figures. For instance, the determination of total phosphorus in waste water samples was carried out with a determination throughput of 12 determinations per hour by implementing in-line microwave digestion of samples (Almeida et al., 2004). This is a significant reduction of the assay time when compared to the conventional batch digestion that took about 2 hours for quantitative measurements.
