Continuous colorimetric determination of trace

Marine Chemistry 96 (2005) 73 – 85 www.elsevier.com/locate/marchem

Continuous colorimetric determination of trace ammonium in seawater with a long-path liquid waveguide capillary cell Qian Perry Lia,b,*, Jia-Zhong Zhanga, Frank J. Millerob, Dennis A. Hansellb a

Ocean Chemistry Division, Atlantic Oceanographic and Meteorological Laboratory, National Oceanic and Atmospheric Administration, 4301 Rickenbacker Causeway, Miami, FL 33149, USA b Division of Marine and Atmospheric Chemistry, Rosenstiel School of Marine and Atmospheric Science, University of Miami, 4600 Rickenbacker Causeway, Miami, FL 33149, USA Available online 16 March 2005

Abstract An automated method for routine determination of nanomolar ammonium in seawater has been developed using segmented flow analysis coupled with a 2-m-long liquid waveguide capillary cell. Conventional photometric detector and autosampler were modified for this method. The optimal concentrations of the reagents and parameters for the development of indophenol blue are discussed. The method has low detection limit (5 nM), high precision (5% at 10–100 nM) and the advantage of rapid analysis of a large number of samples. The method has been used to examine the distribution of ammonium in Florida Bay and Biscayne Bay. D 2004 Elsevier B.V. All rights reserved. Keywords: Ammonium; Seawater; Automated analysis; Liquid waveguide

1. Introduction Ammonia (NH3) is an important nitrogen species in the natural environment. As a dominant gaseous base in the air, it plays a very important role on the acid–base chemistry of the atmosphere and greatly influences the atmospheric sulfur cycle in the remote

* Corresponding author. MAC/RSMAS/University of Miami, 4600 Rickenbacker Causeway, Miami, FL 33149, USA. Tel.: +1 305 4214019; fax: +1 305 4214689. E-mail address: [email protected] (Q.P. Li). 0304-4203/$ - see front matter D 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.marchem.2004.12.001

marine boundary (Galloway, 1995; Quinn et al., 1996). Being a gaseous compound, ammonia exchanges at the air–sea interface although its flux is not well quantified in a variety of environments (Bouwman et al., 1997). Ammonia can easily dissolve in water and become ammonium ion (NH4+). In the ocean, ammonium is the dominant form, with ammonia as a minor component. Ammonium is also one of the most commonly used nutrients by marine phytoplankton. Compared to nitrate, phytoplankton generally prefer ammonium because additional energy is required for them to reduce nitrate to ammonium (D’Elia and DeBoer, 1978; Wheeler and Kokkinakis,

74

Q.P. Li et al. / Marine Chemistry 96 (2005) 73–85

1990; Harrison et al., 1996). Because ammonium is consumed by phytoplankton in the surface waters of the ocean, it is often well below micromolar concentrations and difficult to accurately quantify by conventional analytical techniques. To study the nutrient cycle in the oligotrophic ocean, nanomolar-level nutrient analytical methods are needed. Methods for nitrate and nitrite (Yao et al., 1998; Zhang, 2000; Masserini and Fanning, 2001), phosphate (Karl and Tien, 1992; Zhang and Chi, 2002), iron (Vink et al., 2000; Zhang et al., 2001a,b) and ammonium (Brzezinski, 1987; Jones, 1991; Kerouel and Aminot, 1997) at nanomolar concentrations have been developed. The application of these methods to field studies (Law et al., 2001; Woodward and Rees, 2001; Zhang et al., 2001a,b) has greatly improved our understanding of nutrient dynamics in the surface of the oligotrophic ocean. However, low-level ammonium determination still suffers from low sensitivity and high contamination (Aminot et al., 1997), particularly on shipboard measurements (Harrison et al., 1996). Therefore, it is desirable to develop a highly sensitive method for shipboard automated measurements of ammonium. The most popular technique for the determination of ammonium in aqueous samples is the colorimetric method based on the formation of indophenol blue (Solorzano, 1969; Hansen and Koroleff, 1999). Although this method is simple, economical and easy for automation, it is not sensitive enough for the determination of submicromolar concentrations of ammonium (Aminot et al., 1997). A selective electrode method was found easy to operate (Garside et al., 1978), but requires long equilibration times. Moreover, its detection limit of 0.2 AM is not sufficient for routine work in oligotrophic waters. To increase the sensitivity, a solvent extraction method (Brzezinski, 1987) was developed, but the procedure is time consuming and labor intensive and thus impossible for shipboard automated measurements. Although the fluorometric method (Jones, 1991; Kerouel and Aminot, 1997) has a detection limit of nanomolar concentrations for ammonium, the method often suffers from high background fluorescence and interference by methylamines. An ion chromatography method coupled with a flow injection gas diffusion technique has a reported detection

limit of 20 nM (Gibb et al., 1995), but it requires expensive chromatographic equipment and a long diffusion time. A liquid waveguide capillary cell made out of AF-2400 Teflon has been applied to enhance the sensitivity of spectrophotometric analysis of trace concentrations of ferrous, chromate, nitrate and phosphate ions in aqueous samples (Waterbury et al., 1997; Yao et al., 1998; Zhang, 2000; Zhang and Chi, 2002). This newly developed liquid waveguide capillary cell (World Precision Instrument, Sarasota, FL, USA) has the advantage of low light attenuation, is easy to clean, and is, therefore, very suitable for low-level photometric measurements. Here, we incorporate a 2-m-long liquid waveguide capillary cell to a modified gas-segmented continuous flow auto-analyzer, thereby enhancing the sensitivity and the precision of ammonium determination in seawater by the indophenol blue method.

2. Experiment 2.1. Liquid waveguide capillary cell and spectra system UV–Vis detector The liquid waveguide capillary cell (LWCC) is an optical sample cell that uses the World Precision Instruments’ patented Aqueous Waveguide Technology (Liu, 1996). It offers an increased optical path length compared to a standard cuvette and a small sample volume for spectroscopy application. In this study, a 2-m-long LWCC made of quartz capillary tubing (550 Am ID) was used. A Spectra System UV–Vis detector (UV1000) was modified to adapt the LWCC to an auto-analyzer for continuous analysis. The conventional flow cell assembly (0.55 cm path length) was removed and replaced with two custom-made fiber optic connectors. The LWCC was connected to the detector by two fiber optical cables that transmit the source light from the lamp of the detector through the LWCC and to the photodiode detector. A detailed description of the coupling of an LWCC with a detector is given by Zhang (2000). In this study, before and after each run, the LWCC was cleaned with 10% HCl, 1 M NaOH and deionized water, respectively. To get a good signal, each step should last at least 10 min.


2.2. Automated analytical system 2.2.1. Autosampler Contamination has been reported to be a major problem that affects the precision and accuracy of ammonium determination in submicromolar concentrations. Aminot et al. (1997) have pointed out that a sample volume less than 50 ml is not suitable for routine measurements of trace ammonium in natural waters. However, most of the traditional high-speed autosamplers for gas-segmented flow systems are designed for small cups. Autosamplers that can accommodate large volume samples are mostly designed for FIA systems with relatively low sampling speeds, which can introduce large intersample bubbles. In FIA systems, intersample bubbles can be avoided by the automated control of an injection valve. In gas-segmented flow systems, bubbles are usually removed to a waste line by a debubbler before the stream flows into a detector. Extra large sizes of intersample bubbles can escape from the debubblers and get into the detector causing interferences. The other reason to remove the intersample bubbles is to reduce the air contamination. This will be further discussed in the Section 3.1. In this study, a traditional high-speed autosampler (WESTCO Scientific Instrument) was modified for large cups (50 ml) and a debubbler successfully removed the intersample bubbles generated by this sampler.

75

2.2.2. Manifold configuration A gas-segmented continuous flow colorimetric method was used for the analysis of ammonium in seawater. The flow diagram is shown in Fig. 1. The analytical method is based on the conventional indophenol blue method (Solorzano, 1969) as modified to a gas-segmented continuous flow system (Zhang et al., 1997). The chemical reaction takes place in two steps. Firstly, the addition of hypochlorite to the ammonium samples results in the formation of mono-chloramines. Secondly, phenol reacts with the mono-chloramines to produce an indophenol blue dye. The maximum absorbance of the indophenol blue is measured at 640 nm. To increase the speed of this reaction, a catalyst (sodium nitroferricyanide) and a heater were used in this study. 2.2.3. Data acquisition system Concentrations of ammonium in the samples are calculated from the linear regression, obtained from the standard curve in which the concentrations of the calibration standards are entered as the independent variable, and their corresponding peak heights are the dependent variable. The operation of the autosampler and the collection and analysis of data are simultaneously controlled by a computer based data acquisition system (SOFTPAC, Measurement Microsystems A-Z Inc.). Autosampler

2 m LWCC

Waste

0.41 Heater

Debubbler

0.41

150 cm 0.10

25 cm

Waste

Sample/ wash = 80s: 60s Nitroferricyanide

25 cm 0.10 25 cm 0.10

NaDTT Phenol

25 cm 1.01

Photodiode Lamp 640 nm Data Acquisition System

0.25

Nitrogen

0.32

Complexing reagent

UV/Vis Detector Manifold

Pump (ml/min)

Fig. 1. Schematic flow diagram of manifold configuration for the ammonium analysis with LWCC. The inner diameter of coils used here is 1.0 mm.

76


2.3. Low ammonium seawater and standards Low ammonium seawater was prepared by the removal of background ammonium in low-nutrient seawater (LNSW) collected from the surface of the Gulf Stream. Several drops of 1 M NaOH were added to the LNSW until a small amount of precipitation was observed. After that, it was swirled and heated to 60 8C. This solution was then sealed and naturally cooled to room temperature and finally filtered through a 0.45-Am filter. Ammonium stock

standard solutions (10 mM) were prepared from analytical reagent-grade pre-dried (105 8C for 2 h) ammonium sulfate ((NH4)2SO4) and stored at 48C in a refrigerator. Working standards were prepared from serial dilutions of stock solutions with the low ammonium seawater. Glass cups were found to be subject to ambient ammonium contamination, which might be caused by the adsorption of ammonium on the glass walls. Therefore, plastic cups made of polypropylene were used for both the samples and standards.

0.60

0.50

Absorbance

0.40

0.30

0.20

0.10

(a) 0.00 0.0

5.0

10.0

15.0

20.0

25.0

30.0

Phenol (mM)

0.60

0.50

Absorbance

0.40

0.30

0.20

0.10

(b)

0.00 0.0

5.0

10.0

15.0 Phenol (mM)

20.0

25.0

30.0

Fig. 2. (a) Relationship between phenol and blank. (b) Relationship between phenol and the net signal of a 600 nM ammonium standard solution. For (a) and (b), other reagents and temperature were fixed (NaDTT: 1.6 mM, FSCN: 1.4 mM, T=80 8C).


2.4. Reagents All the chemicals used in this study were of analytical reagent-grade. Deionized water (DIW) used for preparing reagents was purified by a distilling unit followed by a Millipore Super-Q Plus Water System that produces water with 18 Mg resistance. To avoid contamination in the analysis, the deionized water used was purified daily. All the samples and reagents were stored in high-density polypropylene bottles that were ultrasonicated in 1 M NaOH for 6 h at room temper-

77

ature and rinsed several times by deionized water prior to their use. Concentrations of phenol, sodium dichloroisocyanuric acid and sodium nitroferricyanide were varied over a wide range to define the optimal reaction concentration for each reagent. After optimization, the following recipe was found suitable for routine analyses of trace ammonium in seawater by LWCC. Complexing reagent: Dissolve 80 g of sodium citrate (Na 3 C 6 H 5 O 7 d 2H 2 O), 4.0 g of sodium hydroxide, and 10.0 g of EDTA in 1000 ml distilled water. The daily working solution is prepared by

0.60

0.50

Absorbance

0.40

0.30

0.20

0.10

(a) 0.00 0.00

0.50

1.00

1.50

2.00 2.50 NaDTT (mM)

3.00

3.50

4.00

4.50

0.60

0.50

Absorbance

0.40

0.30

0.20

0.10

0.00 0.00

(b) 0.50

1.00

1.50

2.00 2.50 NaDTT (mM)

3.00

3.50

4.00

4.50

Fig. 3. (a) Relationship between NaDTT and blank. (b) Relationship between NaDTT and the net signal of a 500 nM ammonium standard solution. For (a) and (b) other reagents and temperature were fixed (Phenol: 10.0 mM, FSCN: 1.8 mM, T=80 8C).

78


adding 1 ml Brij-35 (ICI Americas) to 200 ml of this complexing reagent. Phenol reagent: Dissolve 1.0 g of solid phenol (C6H5OH) in 1000 ml distilled water. Hypochlorite reagent (NaDTT): Dissolve 0.35 g of dichloroisocyanuric acid sodium salt (NaC3Cl2N3O3) in 1000 ml distilled water. Catalyst: Dissolve 0.55 g of sodium nitroferricyanide (Na2Fe(CN)5NOd 2H2O) in 1000 ml distilled water.

All the reagents were prepared fresh daily except the complexing reagent that was prepared each week.

3. Results and discussion 3.1. The optimization of the flow configuration For the gas segmented continuous flow analysis, the injection of bubbles is critical to the final output

0.60

0.50

Absorbance

0.40

0.30

0.20

0.10

(a) 0.00 0.00

0.50

1.00

1.50 FSCN (mM)

2.00

2.50

3.00

0.60

0.50

Absorbance

0.40

0.30

0.20

0.10

(b) 0.00 0.00

0.50

1.00

1.50

2.00

2.50

3.00

FSCN (mM)

Fig. 4. (a) Relationship between nitroferricyanide (FSCN) and blank. (b) Relationship between nitroferricyanide (FSCN) and the net signal of a 500 nM ammonium standard solution. For (a) and (b), other reagents and temperature were fixed (Phenol: 10.0 mM, NaDTT: 1.6 mM, T=80 8C).


79

3.2. The influence of reagent concentrations

of the signal. The mixing of reagents and the sample is achieved through the performance of the bubble system, which contributes to the high precision of gas-segmented flow analysis. Segmented bubbles are usually injected from a pump tube by pumping air to the flow stream. High erratic peaks observed in trace measurements are possibly due to ammonia contamination in the ambient air. To minimize ammonia contamination, pure nitrogen was used as the segmentation gas. However, intersample bubbles will inevitably bring ambient air into the flow system. Therefore, a debubbler is introduced after the autosampler to remove the intersample bubbles and to avoid the potential contamination. Although the surfactant is not involved in the chemical reaction, it does influence the baseline signal. Therefore, the amount of surfactant should be minimal. In this study, 1 ml of Brij-35 in 200 ml of working complexing reagent was found to be enough to keep a regular pattern for the bubble stream. To get a smooth baseline, 1 M NaOH solution followed by deionized water was pumped through the system before the experiment to clean the trace amounts of ammonium left in the system.

According to the Lambert–Beer law, the absorbance is proportional to the concentration of analyte and the path length of the light in the sample solution. An increase of the path length of the cell will directly enhance the sensitivity of spectrophotometry, which has been applied to improve the determination of various ions in natural waters (Waterbury et al., 1997; Yao et al., 1998; Zhang et al., 2001a,b; Zhang and Chi, 2002). However, the increase of light path proportionally enlarges both the reagent blank and sample signals. For measurements whose reagent blanks are very low (e.g., nitrite and iron), the final blanks after enhanced by the long flow cell still allow sufficient source light to reach the detector. However, the reagent blank for ammonium analysis is high even in the conventional short-cell colorimetric measurement (Aminot et al., 1997). In this case, the reagent blank in the long flow cell absorbs so much source light that there is not enough light to reach the detector. To effectively reduce the reagent blanks in trace ammonium analysis, a series of experiments were designed to investigate the optimal concentrations for reagents with minimal reagent blanks and maximal signals.

0.60

0.50

Absorbance

0.40

0.30

0.20

0.10

0.00 25

30

35

40

45

50

55

60

65

70

75

80

85

90

T(°C)

Fig. 5. Relationship of temperature and the net signal of 600 nM ammonium standard solution (Phenol: 10.0 mM, NaDTT: 1.6 mM, FSCN: 1.4 mM).

80


3.2.1. Complexing reagent Citrate is the most common complexing reagent used for the ammonium analysis in seawater to prevent the precipitation of metal hydroxides such as Mg(OH)2 and Ca(OH)2. Some authors use EDTA as the complexing reagent (Gibb et al., 1995), while others use EDTA and citrate together (Aminot et al., 1997; Zhang et al., 1997). It has been argued that EDTA should be excluded from reagents because it may reduce available chlorine (Kempers and Kok, 1989). Laboratory studies showed that the effective pH at which citrate and EDTA work best is quite different (Gibb et al., 1995). Citrate is effective only in pHb11

and EDTA works at pHN12. For ammonium analysis by colorimetric methods, a pH range of 10.5–11.5 was reported to give satisfactory results for the development of indophenol blue (Patton and Crouch, 1977; Hansen and Koroleff, 1999; Aminot et al., 1997). A higher pH (N12) was used to increase the reaction rate, especially in automated systems (Aminot et al., 1997; Zhang et al., 1997). Therefore, for a pH change between 11.0 and 12 or above, both citrate and EDTA should be used. In this study, final concentrations of 54 mM of citrate (after dilution by reagents and sample), together with 5.4 mM of EDTA, was enough to prevent the precipitation of divalent metal ions in

2.00

1.60

y = 0.0008x + 0.0033

Absorbance

2 R = 0.9997

1.20

0.80

0.40

(a) 0.00 0

500

1000

1500

2000

2500

Ammonia (nM) 0.100

0.080

Absorbance

y = 0.0008x - 0.0004 R2 = 0.9979

0.060

0.040

0.020

(b) 0.000 0

10

20

30

40 50 60 Ammonia (nM)

70

80

90

100

Fig. 6. (a) The linear dynamic range of ammonium determination by gas-segmented continuous-flow analysis with the LWCC. The open circles are data beyond the linearity. (b) The calibration curve for ammonia (10–100 nM).


seawater in a pH range of 11–12. Large precipitates were only observed when the citrate level was below 10 mM. The hydrolysis of the magnesium-citrate complex was found to change the final pH of seawater and interfere with the color formation (Pai et al., 2001), and thus, an increased amount of NaOH in this study can overcome the buffer capacity of this complex. 3.2.2. Phenol Phenol, 2-methylphenol and 2-chlorophenol are currently the most satisfactory reagents for the Berthelot reaction (Patton and Crouch, 1977), for their high sensitivity involved in the development of indophenol blue. Due to the toxicity of phenol, some people used salicylate as a substitute to measure ammonium in seawater, but the sensitivity is significantly decreased (Bower and Holm-Hansen, 1980; Kempers and Kok, 1989). The relationships of phenol concentrations with blanks in low ammonia seawater and the net absorbance of a 600 nM ammonium sample are shown in Fig. 2. The blank is very sensitive to the amount of phenol used at low concentrations and reaches a stable value at phenol concentrations greater than 10.0 mM (Fig. 2a). The net signal of sample is calculated by the difference between the absorbance of samples and the value of the blank. The optimal concentration of phenol is 8.0–10.0 mM, at which the blank is low and the net signal of sample is high (Fig. 2a and b).

Hansen, 1980), a more stable chlorine donor, sodium dichloroisocyanurate (NaDTT), was used. However, higher temperatures are required in order to liberate its chlorine (Kempers and Kok, 1989). The concentration of NaDTT was optimized by examining signals of 500 nM ammonium samples after varying the concentration levels of NaDTT, while keeping the concentrations of all other reagents constant. A plot of absorbance vs. concentration of NaDTT shows the optimum of NaDTT to be between 1.0 mM and 2.0 mM (Fig. 3). The slight increase of the sample signal with increasing chlorine concentration is in agreement with the work of Kempers and Kok (1989). Due to the reaction with seawater constituents, the available chlorine is lower than the chlorine originally added to the system. In order to achieve the same sensitivity, the chlorine concentration required for the reaction in seawater is about four times higher than that in pure water. 3.2.4. Catalyst Without an appropriate catalyst, the reaction rate for the formation of indophenol blue is very slow. Generally, nitroprusside (NP) is used as a catalyst in the IPB method, but in basic medium it becomes nitroferricyanide (NF) and produces aquopentacyanoferrate (AqF). AqF was indeed the actual catalyst (Patton and Crouch, 1977), but it usually needs ultraviolet radiation to activate and is sensitive to the change of pH. Therefore, sodium nitroferricyanide is used as a catalyst in this study without radiation. To study the effect of nitroferricyanide on the chemical

0.00

Absorbance

0.05

3.2.3. Hypochlorite Because sodium hypochlorite is sensitive to the changes of light and temperature (Bower and Holm-

81

0

10

20

Time (minute) Fig. 7. The typical output of signals of 10–50 nM of ammonium analysis by LWCC, Peaks are two replicates samples with ammonium concentrations of 0, 10, 20, 30, 40 and 50 nM, respectively.

82


reaction, two experiments were carried out in which deionized water was used as a wash. In one study, the change of absorbance of reagent blank was examined with a series of different concentrations of catalyst. The other study examined the response of a 500 nM ammonium sample to the same series of catalysts. The net signal of the samples was calculated by subtracting the blank from the value of sample. The best

concentration of nitroferricyanide is 1.6–2.0 mM, as shown in Fig. 4. The best temperature for the formation of indophenol blue for this system is 808C (Fig. 5). It should be noted that different systems might have a different optimal temperature, because heat transfer efficiencies may vary. These variations are related to the length of the heating coil and the performance of the heater. We

25.8 B1 B2 B3

25.6

B4

Florida Bay and Biscayne Bay Survey (9/15~16/2004)

B16

B15

B5 B6 B14

B7 B8

25.4 B9 B10 B11 B12

10 25.2

1

2 13 12

8

9

10

7

11

6 33

14

32

25.0

34

31 35

29 28 27 30

4

5

26

37

6

38

4

20

23 39

36

15

24

8

21

22

25

B13

3

40

2

19

1

18

0.8

17

0.6

16

Ammonium (uM) 24.8

0.4 0.2

-81.0

-80.8

-80.6

-80.4

-80.2

Fig. 8. Spatial distribution of ammonium in Florida Bay and Biscayne Bay during September 2004 survey. Numbers 1–40 and B1–B16 are the stations in Florida Bay and Biscayne Bay, respectively.


3.3. Linear dynamic range and detection limit Using the manifold configuration shown in Fig. 1 and the above optimal recipe for reagents, the upper limits of the linear dynamic range of ammonium analysis is 1000 nM (Fig. 6). The calibration curve was calculated from the average of six independent runs with the standard deviation less than 5%. A typical output signal of automated trace ammonium analysis is shown in Fig. 7. A linear absorbance response to ammonium concentrations below 1000 nM can be obtained as Absorbance ¼ ð0:0033F0:0013Þ þ ð0:000 8005F0:000 0065Þ NHþ 4 ðnMÞ with r 2=0.9997 (n=20). Above this concentration range, the measured absorbance is lower than that predicted from the linear relationship as shown in open circles in Fig. 6. The linear dynamic range of the ammonium analysis can be extended by either using a shorter LWCC or by diluting the sample with low ammonium seawater by adding a dilution line to the sample flow. The detection limit of this method is 5 nM, which is estimated as three times the standard deviation of measurement blanks. For ammonium analysis in seawater, the correction of refractive index interference is important (Aminot et al., 1997). Because the refractive index signal is much smaller than the analytical signal, it is usually qualified by measuring the absorbance of samples with different salinities relative to deionized water in the absence of color formation. For ammonium analysis, this is achieved by using a series of water samples with different salinities as ammonium samples and deionized water as the wash solution, with the exception of the sodium nitroferricyanide being replaced by deionized water. The resultant absorbance was converted to ammonium concentrations. The relationship of measured refractive index with different salinities is: þ NH4 ri ¼ 0:233 þ 0:376S; r2 ¼ 0:998; n ¼ 8 Where [NH4+]ri is a correction for refractive index for ammonium sample in nanomolar and S is the salinity

of sample. To avoid the significant refractive interference for low-level ammonium samples, it is necessary to match the salinity of the wash solution with that of the sample. Low ammonium seawater is recommended.

4. Field application This method has been applied to water samples from a Florida Bay and Biscayne Bay survey conducted in September 2004. Samples were collected from 40 stations in Florida Bay and 16 stations in Biscayne Bay, as is shown in Fig. 8. The ammonium samples were preserved by adding several drops of chloroform and sent back to laboratory for analyses at days end. These samples were first measured by the conventional auto-analyzer with a 0.55-cm cell. Samples with ammonium concentration around or below 1.0 AM were redetermined by the above LWCC method. A comparison of these two methods is shown in Fig. 9. The agreement is quite good. Fig. 8 shows the spatial distribution of ammonium in Florida Bay and Biscayne Bay. The ammonium concentrations in these two bays vary widely from above 10 AM to several hundred nanomolar, which are agreeable with the long-term observation of ammonium in the same area (Boyer et al., 1999). We also found that there are 1.40 1.20 Liquid waveguide method

did observe a decrease in the optimal temperature when the length of the heating coil was increased.

83

y =1.0299x 2 R = 0.9816

1.00 0.80 0.60 0.40 0.20 0.00 0.00

0.20

0.40 0.60 0.80 1.00 1.20 Colorimetric method (0.55cm)

1.40

Fig. 9. Comparison of liquid waveguide method with the conventional colorimetric method. The unit of ammonium concentration used here is micromolar.

84


decreasing concentrations of ammonium from the coast to open ocean both in Florida Bay and Biscayne Bay.

Acknowledgements We thank Christ Kelble for the collection of samples and two anonymous reviewers for helpful comments on the first draft of this paper. NOAA’s South Florida Ecosystem Restoration Prediction and Modeling Program under the Coastal Ocean Program is acknowledged. Additional support comes from the US National Science Foundation (OCE0241340) to DAH. This research was carried out under the auspices of the Cooperative Institute of Marine and Atmospheric Studies, a joint institute of the University of Miami and the National Oceanic and Atmospheric Administration (Contract no. NA67RJ0149). Frank J. Millero also acknowledges the support of the Oceanographic Section of the National Science Foundation.

References Aminot, A., Kirkwood, D.S., Kerouel, R., 1997. Determination of ammonia in seawater by the indophenol-blue method: evaluation of the ICES NUTS1/C5 questionnaire. Mar. Chem. 56, 59 – 75. Bouwman, A.F., Lee, D.S., Asman, W.A.H., Dentener, F.J., 1997. A global high-resolution emission inventories for ammonia. Global Biogeochem. Cycles 11, 561 – 587. Bower, C.E., Holm-Hansen, T., 1980. A salicylate-hypochlorite method for determining ammonia in seawater. Can. J. Fish. Aquat. Sci. 37, 794 – 798. Boyer, J.N., Fourqurean, J.W., Jones, R.D., 1999. Temporal trends in water chemistry of Florida Bay (1989–1997). Estuaries 22, 417 – 430. Brzezinski, M.A., 1987. Colorimetric determination of nanomolar concentration of ammonium in seawater using solvent extraction. Mar. Chem. 20, 277 – 288. D’Elia, C.F., DeBoer, J.A., 1978. Nutritional studies of two red algae: kinetics of ammonium and nitrate uptake. J. Phycol. 14, 266 – 272. Galloway, J.W., 1995. Acid deposition: perspectives in the time and space. Water Air Soil Pollut. 85, 15 – 24. Garside, C., Hull, G., Murray, S., 1978. Determination of submicromolar concentration of ammonia in natural waters by a standard addition method using a gas-sensing electrode. Limnol. Oceanogr. 23, 1073 – 1076. Gibb, S.W., Mantoura, R.F., Liss, P.S., 1995. Analysis of ammonia and methylamines in natural waters by flow injection gas

diffusion coupled to ion chromatography. Anal. Chim. Acta 316, 291 – 304. Hansen, H.P., Koroleff, F., 1999. Determination of nutrients. In: Grasshoff, K., Kremling, K., Ehrhardt, M. (Eds.), Methods of Seawater Analysis, 3rd ed. Wiley-VCH, Weinheim, ISBN: 3-527-29589-5pp. 159 – 228. Harrison, W.G., Harris, L.R., Irwin, B.D., 1996. The kinetics of nitrogen utilization in the oceanic mixed layer: nitrate and ammonium interactions at nanomolar concentrations. Limnol. Oceanogr. 41, 16 – 32. Jones, R.D., 1991. An improved fluorescence method for the determination of nanomolar concentration of ammonium in natural waters. Limnol. Oceanogr. 36, 814 – 819. Karl, D., Tien, G., 1992. MAGIC: a sensitive and precise method for measuring dissolved phosphorus in aquatic environments. Limnol. Oceanogr. 37, 105 – 116. Kempers, A.J., Kok, C.J., 1989. Re-examination of the determination of ammonium as the indophenol blue complex using salicylate. Anal. Chim. Acta 221, 147 – 155. Kerouel, R., Aminot, A., 1997. Fluorometric determination of ammonia in sea and estuarine waters by direct segmented flow analysis. Mar. Chem. 57, 265 – 275. Law, C.S., et al., 2001. A lagrangian SF6 tracer study of an anticyclonic eddy in the North Atlantic patch evolution, vertical mixing and nutrient supply to the mixed layer. Deep-Sea Res. II 48, 705 – 724. Liu, S.Y., 1996. Improved aqueous fluid core waveguide. US Patent 5,570,447. Masserini, R.T., Fanning, K.A., 2001. A sensor package for the simultaneous determination of nanomolar concentrations of nitrite, nitrate and ammonia in seawater by fluorescence detection. Mar. Chem. 68, 323 – 333. Pai, S.C., Tsau, Y.J., Yang, T.I., 2001. pH and buffering capacity problems involved in the determination of ammonia in saline water using the indophenol blue spectrophotometric method. Anal. Chim. Acta 434, 209 – 216. Patton, C.J., Crouch, S.R., 1977. Spectrophotometric and kinetic investigation of the Berthelot reaction for the determination of ammonia. Anal. Chem. 49, 464 – 469. Quinn, P.K., et al., 1996. Estimation of the air/sea exchange of ammonia for the North Atlantic basin. Biogeochemistry 35, 275 – 304. Solorzano, L., 1969. Determination of ammonia in natural waters by phenol hypochlorite method. Limnol. Oceanogr. 14, 799 – 801. Vink, S., et al., 2000. Automated high resolution determination of the trace elements iron and aluminum in the surface ocean using a towed Fish coupled to flow injection analysis. Deep Sea Res. II 47, 1141 – 1156. Waterbury, R.D., Yao, W., Byrne, R.H., 1997. Long pathlength absorbance spectroscopy: trace analysis of Fe(II) using a 4.5 m liquid core waveguide. Anal. Chem. Acta 357, 99 – 102. Wheeler, P.A., Kokkinakis, A., 1990. Ammonium recycling limits nitrate use in the oceanic sub-Arctic Pacific. Limnol. Oceanogr. 35, 1267 – 1278. Woodward, E.M.S., Rees, A.P., 2001. Nutrient distributions in an anticyclonic eddy in the northeast Atlantic Ocean, with

Q.P. Li et al. / Marine Chemistry 96 (2005) 73–85 reference to nanomolar ammonium concentrations. Deep Sea Res. II 48, 775 – 793. Yao, W., Byrne, R.H., Waterbury, R.D., 1998. Determination of nanomolar concentrations of nitrate, nitrite in natural waters using long path length absorbance spectroscopy. Environ. Sci. Technol. 32, 2646 – 2649. Zhang, J.Z., 2000. Shipboard automated determination of trace concentrations of nitrite and nitrate in oligotrophic water by gas segmented continuous flow analysis with a liquid waveguide capillary flow cell. Deep Sea Res. I 47, 1157 – 1171. Zhang, J.Z., Chi, J., 2002. Automated analysis of nanomolar concentrations of phosphate in natural waters with liquid waveguide. Environ. Sci. Technol. 36, 1048 – 1053.

85

Zhang, J.Z., et al., 1997. Determination of ammonia in estuarine and coastal waters by gas segmented continuous flow colorimetric analysis. Methods for determination of chemical substances in marine and estuarine environmental matrices, 2nd ed. EPA/7664-41-7. Zhang, J.Z., Kelble, C., Millero, F.J., 2001a. Gas segmented continuous flow analysis of iron in water with a long liquid waveguide capillary flow cell. Anal. Chim. Acta 438, 49 – 57. Zhang, J.Z., Wanninkhof, R., Lee, K., 2001b. Enhanced new production observed from the diurnal cycle of nitrate in an oligotrophic anticyclonic eddy. Geophys. Res. Lett. 28, 1579 – 1582.