A preliminary methods comparison for measurement of dissolved organic nitrogen in seawater
Introduction
Although attempts to measure dissolved organic nitrogen (DON) in seawater extend over a century (at least as far back as Putter, 1909), it could probably be argued that modern methods began with publication of a method based on ultraviolet (UV) light oxidation (Armstrong et al., 1966). Because of apparent problems with incomplete oxidation of recalcitrant organic compounds by UV, use of the strong chemical oxidant, potassium persulfate, was introduced in the 1970s (see Koroleff, 1976) as an alternate method. There was indication that the persulfate oxidation (PO) method gave better recovery of standard chemicals and slightly higher values for DON in natural marine samples than the UV method D'Elia et al., 1977, Solorzano and Sharp, 1980. Similar higher values by PO than UV have been demonstrated for rainwater Cornell et al., 1995, Scudlark et al., 1998. Nonetheless, many laboratories have used some variant of either UV or PO methods for the past 20–30 years. Sometimes, UV methods use higher energy UV sources and different geometry than the original Armstrong et al.'s instrument; however, some current routine methods use essentially the same system as the original.
A high temperature combustion (HTC) method was described in the 1980s that appeared to measure considerably higher concentrations of DON in seawater (Suzuki et al., 1985) than either wet chemical method. While this method created considerable interest, other analysts were not able to reproduce the high DON values using HTC methods (e.g., Koike and Tupas, 1993, Hansell, 1993, Lopez-Vernoni and Cifuentes, 1995) and the method as originally published was eventually withdrawn (Suzuki, 1993). Essentially, all other HTC methods used since the early 1990s report DON values within the same range as the wet chemical methods. However, most comparisons have been indirect.
A direct comparison performed several years ago in one laboratory showed comparable results with both the UV and HTC methods and apparently comparable results with PO for oceanic samples (Walsh, 1989). Also, good comparability between HTC and PO methods was demonstrated for river and domestic wastewater samples (Ammann et al., 2000).
The DON subgroup report from the Seattle DOC/DON workshop evaluated published reports of various DON methods and unpublished results from the participants; they concluded that there was no strong indication of wet chemical methods giving lower yields than HTC methods (Hopkinson et al., 1993). A direct comparison was also made at that workshop by 13 analysts using a variety of HTC instruments and two wet chemical methods (Hedges et al., 1993). While agreement among the analyses was not great, there was no indication of wet chemical methods measuring less DON than HTC (Hedges et al., 1993). With data from the 13 analyses, the reproducibilities of analyses gave coefficients of variation values (see Section 2.2 below) of 9–49% for the four samples.
A direct comparison between a commercial HTC analyzer and PO and UV methods was done in Bronk's laboratory recently (Bronk et al., 2000). In that comparison, differences between methods were found for standard chemicals and for natural samples with differing distinctions. For example, the PO method gave best recovery of most standard chemicals, but not all, and the HTC method gave higher values for oceanic samples than wet chemical methods but did not for nearshore samples. The UV method, in its standard format, appeared not to give complete recovery for standard compounds and gave lower values for natural samples. When modified using persulfate as an added oxygen source instead of peroxide, the UV method gave better results. Most of the natural samples in this comparison had relatively high DON content. One oceanic sample appeared to give comparable results by HTC and PO but lower values by UV even with the modified chemistry. It was concluded that different methods have advantages for different types of samples (Bronk et al., 2000).
At this time, large differences in DON concentrations are not expected by different methods, and there is an inherent assumption that all methods are acceptable. However, recent publications give indirect comparison of large scatter for subsurface oceanic DON concentrations. Examining some of the recent literature, DON concentrations for oceanic samples below 200 m ranged from less than 2 to over 8 μM N Hansell and Waterhouse, 1997, Kahler et al., 1997, Wheeler et al., 1997, Ogawa et al., 1999. These reports represent use of UV, PO, and two hybrid HTC methods. Thus, it appears that a broad community evaluation of DON methodology is timely.
Several of us involved in this paper (Sharp, Hansell, Burdige, Cauwet, and Ogawa) started a small-scale methods comparison recently; results of which have been reported in national meetings. Samples from an oceanic depth profile and from an estuarine gradient were analyzed by UV, PO, and HTC methods. For the depth profile, there was relatively good agreement for total dissolved nitrogen (TDN) analysis but after subtracting the large dissolved inorganic nitrogen (DIN) concentration, the variability for the small remaining DON pool was large in relative terms. In the estuarine gradient with TDN concentrations in the range of 10–50 and DON in the range of 7–9, good agreement was found (unpublished data). A larger community effort was thought to be the best next step.
This study involves a comparison of a small number of samples by a large number of analysts, i.e., a broad community comparison with replicate samples sent to all interested participants. The comparison represented here is different from recent DOC comparisons (Sharp et al., 2002) in that it included only five samples, all seawater, and no blank or pure water references were included. Also, the majority of analysts participating in the DOC comparison used HTC methods; the majority for this DON comparison used wet chemical methods.
Section snippets
Samples, sample preparation, and analysts
Five samples were used, from estuarine, coastal, and open ocean surface and deep environments. Most samples were aged for several months to allow oxidation of more reduced inorganic ions (NH4+ and NO2−) to the highest oxidation state (NO3−) and decomposition of labile dissolved organic matter. The purpose of the aging, in 5-gal glass carboys closed with silicone rubber stoppers, was to render the samples stable for the period of time necessary to include a large number of analysts (about 6
Results
A number of laboratories also performed analysis of N&N prior to treatment for the TDN analysis and reported those data along with the TDN data. Some of them also analyzed for NO2− and NH4+. Results of all of these DIN data are listed in Table 4. For the 12 N&N analyses, the agreement is not very good (%cv values ranging from 13% to 73%). In addition, some samples by a few analysts had higher initial N&N than the final TDN for that same sample, potentially indicating error in initial DIN
Discussion
The three families of methods (HTC, PO, and UV) differ in chemistry of conversion of TDN and in complexity of analysis. The wet chemical methods entail oxidation of all nitrogen to the highest oxidation state in aqueous medium while the HTC methods use pyrolysis to yield an intermediate gaseous form that is not in the highest oxidation state. The HTC methods are more automatic with less handling while the other two methods involve more manual manipulation and, hence, have more potential for
Acknowledgements
This research was partially supported by NSF grants OCE 96-33296 and OCE 00-82238 to JHS, who also thanks Douglas Miller for assistance with statistical analysis of the data and John Hedges for data from the 1991 Seattle workshop. The results reported here were contributed by many individuals; in addition to the authors, all others listed in Table 2 made valuable contributions to the success of the research.
Associate editor: Dr. Edward T. Peltzer.
References (30)
- et al.
Simultaneous determination of dissolved organic carbon and total dissolved nitrogen in seawater by high temperature catalytic oxidation: conditions for precise shipboard measurements
Mar. Chem.
(1998) - et al.
Simultaneous determination of TOC and TN in surface and wastewater by optimised high temperature catalytic combustion
Water Res.
(2000) - et al.
Total dissolved nitrogen analysis: comparisons between persulfate, UV, and high temperature oxidation methods
Mar. Chem.
(2000) A chemiluminescent technique for the determination of nanomolar concentrations of nitrate and nitrite in seawater
Mar. Chem.
(1982)Results and observations from the measurement of DOC and DON in seawater using a high-temperature catalytic oxidation technique
Mar. Chem.
(1993)- et al.
Controls on the distribution of organic carbon and nitrogen in the eastern Pacific Ocean
Deep-Sea Res., Part I
(1997) - et al.
Comparative analyses of DOC and DON in natural waters
Mar. Chem.
(1993) - et al.
DON subgroup report
Mar. Chem.
(1993) - et al.
Dissolved organic matter and its utilization by bacteria during spring in the Southern Ocean
Deep-Sea Res., Part II
(1997) - et al.
Total dissolved nitrogen in the Northern North Pacific assessed by a high-temperature combustion method
Mar. Chem.
(1993)
Vertical distributions of dissolved organic carbon and nitrogen in the Southern Ocean
Deep-Sea Res., Part I
Wet-oxidation and automated colorimetry for simultaneous determination of organic carbon, nitrogen, and phosphorus dissolved in seawater
Mar. Chem.
Organic nitrogen in precipitation at the mid-Atlantic U.S. coast—methods evaluation and preliminary measurements
Atmos. Environ.
Marine dissolved organic carbon: are the older values correct?
Mar. Chem.
Final dissolved organic carbon broad community intercalibration and preliminary use of DOC reference materials
Mar. Chem.
Cited by (84)
Detecting trace dissolved organic nitrogen (DON) in freshwater by converting DON rapidly into nitrate with VUV/H<inf>2</inf>O<inf>2</inf> pretreatment
2022, ChemosphereCitation Excerpt :As for high temperature combustion (HTC) method, although it does not add agents, its transformation product (nitrogen oxide) exists in gaseous form and its MDL is high (49‒100 μg-N/L) too (Chen et al., 2019; Saunders et al., 2017). So, accurate and convenient measurement of sub μg-N/L level of DON is desired but impossible currently (Johnes and Heathwaite, 1992; Sharp et al., 2002; Lee and Westerhoff, 2005; Li et al., 2018). In an earlier review, it was reported that approximately 30% of surface water, 76% of shallow groundwater, and 81% of deep groundwater samples do not have DON data because they contain DON below the MDLs of existing TDN analytical methods (Westerhoff and Mash, 2002).
Submarine groundwater-derived inorganic and organic nutrients vs. mariculture discharge and river contributions in a typical mariculture bay
2022, Journal of HydrologyCitation Excerpt :The DIN concentration was expressed as the sum of the NO2−, NO3−, and NH4+ concentrations, while NOx is expressed as the sum of the NO3− and NO2−. The TDN and TDP were measured using high-temperature combustion methods (Sharp et al., 2002; Lee and Westerhoff, 2005). The concentrations of DON and dissolved organic phosphorus (DOP) are obtained by subtracting the dissolved inorganic nutrients from the total dissolved nutrients.
Identifying locations of sewage pollution within a Hawaiian watershed for coastal water quality management actions
2021, Journal of Hydrology: Regional StudiesA green and robust method to measure nanomolar dissolved organic nitrogen (DON) by vacuum ultraviolet
2019, Chemical Engineering Journal