Organic nitrogen in the atmosphere —Where does it come from?A review of sources and methods
J.N.Cape a ,⁎,S.E.Cornell b ,T.D.Jickells c ,E.Nemitz a
a Centre for Ecology &Hydrology,Bush Estate,Penicuik,EH260QB UK
b Department of Earth Sciences,University of Bristol,Wills Memorial Building,Queens Road,Bristol,BS81RJ UK c
School of Environmental Sciences,University of East Anglia,Norwich,NR47TJ UK
a r t i c l e i n f o a
b s t r a
c t
Article history:
Received 7December 2010
Received in revised form 14July 2011Accepted 20July 2011This review considers the ways in which atmospheric organic nitrogen has been measured and linked to potential sources.Organic N exists in gas,particle and dissolved phases and represents a large (ca.30%)fraction of total airborne nitrogen,but with large variability in time and space.Although some components (e.g.amines)have been the subject of several studies,little information is available for the many other components of organic N that have been identified in individual measurements.Measurements of organic N in precipitation have been made for many decades,but both sampling and chemical analytical methods have changed,resulting in data that are not directly comparable.Nevertheless,it is clear that organic N is ubiquitous and chemically complex.We discuss some of the issues which have inhibited the widespread adoption of organic N as a routine analyte in atmospheric sampling,and identify current best practice.Correlation analysis is the most widely used technique for attributing likely sources,examining the co-variation in time and/or space of organic N with other components of precipitation or particulate matter,yet the shortcomings of such simple approaches are rarely recognised.Novel measurement techniques which can identify,if not yet quantify,many of the components of particulate or dissolved organic N greatly enhance the data richness,thereby permitting powerful statistical analyses of co-variation such as factor analysis,to be employed.However,these techniques also have their limitations,and whilst specific questions about th
e origin and fate of particular components of atmospheric organic N may now be addressed,attempts to quantify and attribute the whole suite of materials that comprise atmospheric organic N to their sources is still a distant goal.Recommendations are made as to the steps that need to be taken if a consistent and systematic approach in identifying and quantifying atmospheric organic N is to progress.Only once sources have been recognised can any necessary control measures to mitigate adverse effects of atmospheric organic N on human health or ecosystem function be determined.
©2011Elsevier B.V.All rights reserved.
Keywords:
Particulate matter Precipitation
Source attribution Nitrogen deposition
Measurement techniques
1.Introduction
The atmospheric deposition of nitrogen is a focus of both scienti fic and policy concern,because of th
e importance of nitrogen (N)in acidi fication of soils and aquatic systems,nutrient enrichment,tropospheric ozone processes,and parti-culate matter.The role of inorganic N (nitrate and ammonium)
is well understood in these contexts,and is probably dominant in the first two of these.Although organic N is known to be important,the details of the role of organic N are less well understood,especially in the latter two contexts.
Eriksson (1952)provided an early but comprehensive assessment of the scope of research into atmospheric deposi-tion,and traced the history of nitrogen deposition studies.The
Atmospheric Research 102(2011)30–48
⁎Corresponding author.Tel.:+441314458533;fax:+441314453943.E-mail address:jnc@ceh.ac.uk (J.N.
Cape).0169-8095/$–see front matter ©2011Elsevier B.V.All rights reserved.doi:
10.1016/j.atmosres.2011.07.009
Contents lists available at ScienceDirect
Atmospheric Research
j o u r n a l h om e p a g e :w w w.e l sev i e r.c o m /l o c a t e /a t m o s
more recent history of organic N measurement,focused on studies of rainwater,has been outlined by Cornell(in press). Early on,the deposition of organic N compounds was recognised as quantitatively significant(Eriksson reports studies from the United Kingdom,North America,India and New Zealand),
but assumptions about its source–locally recycled dust–and a focus on agricultural productivity meant that its role in the‘nitrogen economy of soil’was regarded as negligible at that time.
Eriksson wrote at a time of growing awareness of N pollution in Europe,which led to the inception of atmospheric measurement and monitoring programmes,notably the Euro-pean Air Chemistry Network for air and precipitation chemistry established in1955.A rapid increase in studies of atmospheric nitrogen deposition,including studies of organic N,was a response to the requirements of the1979UNECE Convention on Long Range Transboundary Air Pollution(www. /env/lrtap/).The US National Atmospheric Deposition Program(NADP,nadp.sws.uiuc.edu/)was established in 1977,with a parallel organisation in Europe,the European Monitoring and Evaluation Programme(p. int),both with a remit of monitoring air quality for the management of acidification and eutrophication,but in both cases the monitoring programmes addressed only inorganic N. Since the1980s,studies have extended beyond a focus on local pollution and agriculture,and paid more attention to N deposition in the marine environment,and its role in natural biogeochemical cycles.A‘third wave’of organic N deposition studies has recently become apparent in the last1–2decades,as Asia and Latin America become more concerned about anthropogenic pollution.
This paperfills a gap—other studies have confirmed the quantitative importance,ubiquity,and global distribution of organic N,but have largely left open the question of sources, except for rather speculative attention.
2.What is‘organic nitrogen’?
Atmospheric organic N exists in gaseous,particulate and aqueous phases,and has become of increasing scientific in-terest as its quantitative importance to total airborne nitro-gen has been appreciated.In part,the lag in addressing the organic N,compared with the inorganic species,has been because of a lack of consistency in sampling and measure-ment methods and because of the large number of different compounds that contribute to the total organic N.This has led to an understandable caution in attributing measured organic N in precipitation to material actually deposited from the atmosphere,rather than formed in rain collectors by bio-logical transformations of inorganic N,or other artefacts of sampling.However,careful attention to sampling protocols, including storage conditions and chemical analytical meth-ods,has demonstrated that water-soluble organic N(WSON) measured in precipitation can be demonstrated to constitute a significant proportion(typically around30%)of total water-soluble N(Cape et al.,2001;Cornell et al.,2003;Keene et al., 2002;Neff et al.,2002;Scudlark et al.,1998;Zhang et al., 2008).Measurements from a review of r
ecent data(Cornell, in press)suggest that the absolute concentrations of organic N in precipitation may be increasing(Fig.1),but this apparent trend may simply represent the number of recent measure-ments in areas of the world(such as China)where concen-trations of all forms of atmospheric N are large,typically over 10times greater than those observed in Europe or North America(Zhang et al.,2008).
A major difficulty with addressing the biogeochemical and ecological role of atmospheric organic N is that there has been little practical consensus to date on how to define and measure it.Two strategies have been applied to its analysis,largely in parallel—bulk analysis,and the analysis of specific compounds in precipitation or aerosol,similar to the issues faced when measuring carbonaceous compounds,cf.Fig.4in Hallquist et al. (2009).Molecular or functional group analysis allows for particular components to be investigated,but there remains a concern that this approach is not a representative sampling of the organic matter present in the bulk samples,and given the diversity of possible compounds to analyse in this approach, there are few comparable studies of any given compound group.In general,analysis of specific compounds in rain or aerosol has only been able to account for a small proportion of the total organic N(Shi et al.,2010;Zhang et al.,2002a).Bulk analysis gives the total concentration of the rain or aerosol organic N,so it can only give insights into the‘colle
ctive behaviour’of the organic matter.It has proved important in assessing global budgets,but the value of these budget assessments is now constrained by the paucity of knowledge about what the organic matter is,where it comes from,and what it does in the environment.The categories within bulk analysis are operationally defined.For precipitation organic N,‘dissolved’material is what passes through afilter.For aerosols,‘particulate’material is the organic matter that is retained on a filter through which air is sampled.Different studies have used different kinds offilter,and different extraction and dissolution treatments,with or without pre-denuder sampling to minimise gas-phase artefacts,raising issues even about the representa-tiveness and comparability of bulk analysis.Aerosol samples are usually extracted into water to obtain water-soluble organic N (WSON);the interest in the water-soluble component arises from the assumption that solubility and bioavailability are linked,although some studies have noted that there may be a significant component of organic N that is collected by a particulatefilter which is not readily water Miyazaki et al.,2010;Russell et al.,2003).Water-soluble organic N is regarded as bio-available(Lipson and Näsholm, 2001;Paerl,1997;Peierls and Paerl,1997;Seitzinger and Sanders,1999;Timperley et al.,1985;Wedyan et al.,2007). Very little is known about the biological effects of the insoluble component.
Where deposition of inorganic N to terrestrial ecosystems exceeds the‘Critical Load’(Achermann and Bobbink,2003) the additional input of organic N,which is not included in the risk evaluation,may provide even greater pressures towards eutrophication or species composition change than predicted, and may pose a threat to systems where the Critical Load does not appear to be exceeded.
2.1.Chemical analytical methods for organic N
Despite much recent measurement activity,the chemical composition of WSON is still very poorly understood;‘organic N’in this context is usually defined as the difference between some measure of the total dissolved N in a sample(or the
31
J.N.Cape et al./Atmospheric Research102(2011)30–48
total water-soluble component of sampled particulates),TN,and the concentrations of measured inorganic N(ammonium,nitrate,nitrite =IN).Organic N =TN –IN
Organic N,including WSON,forms part of atmospheric organic carbon,as particulate matter and disso
lved in precipitation and cloud water.The complexity and range of different sources of this organic carbon are similar to those for organic N,and are the subject of much current research (e.g.Chow et al.,2010;Collett et al.,2008;Hallquist et al.,2009;Heald et al.,2008;Miller et al.,2009;Muller et al.,2008;Salma et al.,2007;Simoneit and Mazurek,2007).The N/C ratio in airborne organic matter may be a useful diagnostic for sources,as described below (Section 3.5.3).
Historically,concentrations of dissolved inorganic N (IN)have been made colorimetrically,which introduces additional uncertainties because of the response of such methods to some amino compounds (Cape et al.,2001).More recently ion chromatography (IC),which has less interference from organic material,has been used for measuring IN in precip-itation and aerosol samples.
Several techniques are used to determine total water-soluble N,including wet chemical oxidation (WCO)using strong chemical oxidants such as persulfate,peroxide or chromic acid,UV oxidation (Mace and Duce,2002),and high temperature oxidation with or without a catalyst (Bronk et al.,2000;Sharp et al.,2002).Adding oxidants always introduces the risk of adding contaminants,so WCO methods often have high blanks associated with them.There is also always the risk of not oxidising all the organic matter present,leading to an under-estimation of the organic N,and often resulting in high variability and low precision.Different oxidants appear to have different oxidation ef fi
ciencies for the various compo-nents of the organic N present in the samples,so not only would the N recoveries be partial with WCO methods,they may also be different and not easily comparable.Develop-ments in high-temperature oxidation (HTO)and catalytic oxidation (HTCO)methods in the 1980s (Sharp et al.,2002;Sugimura and Suzuki,1988;Suzuki et al.,1985)resulted in the direct determination of total N and C.In these methods,organic N is generally oxidised to NO,which is measured by chemiluminescence.In the early years of the HTCO technique,problems with instrument blanks were serious,affecting the detection limits for organic N.Although technical develop-ments have now resolved these problems,the high blanks were particularly problematic for dilute solutions like rain-water,and samples where the completeness of dissolution could not be assured,as for the aqueous extracts of aerosol.This,combined with the high cost of the HTCO instruments and analysis,meant that WCO methods have remained in use in atmospheric studies despite their known limitations.These techniques were investigated very thoroughly for marine science applications through the 1990s,where the new methods opened up new debates about organic matter in seawater.The techniques underpinned research into the role of organic matter in the marine ‘microbial loop ’,in trace nutrient (metal)availability,and so on.This field of research transformed understanding of marine organic matter,from the perception that it was dilute and refractory,to the rec-ognition that organic matter is a dynamic assemblage of mo-lecules intimate
ly involved in biogeochemical processes operating at multiple space and time scales (Hedges,2002;Sharp et al.,2002).
Even with recent analytical improvements,however,this de finition of organic N ‘by difference ’leads to potentially large relative uncertainties in organic N concentrations associated with the aggregation of errors.The precision problem is most acute for samples with low concentrations of organic N and high concentrations of inorganic N,where the coef ficient of variation of the organic N analysis can easily be the same magnitude as the concentration (Cornell et al.,2003;Mace and Duce,2002).The final ‘systemic ’challenge,which
applies
Fig.1.Organic N concentration in wet deposition from 90sites,53publications,drawn from a recent compilation of the literature (Cornell,in press ).
deposition32J.N.Cape et al./Atmospheric Research 102(2011)30–48
to all methods,is that without knowing what organic N is,it is difficult to select standard compounds to test the oxidation efficiency of methods adequately.Typically a suite of organic compounds is urea,amino acids,amines,peptides), selected because they occur in nature,have particular chem-ical or photochemical properties of interest,or have previ-ously been used as standards(Mace and Duce,2002).There have been some method comparisons for rainwater and aerosol organic N(Cornell and Jickells,1999;Scudlark et al., 1998)for UV and persulfate oxidation and for UV,persulfate oxidation and HTO(Cape et al.,2001).Despite all the caveats, the WCO methods are largely effective,with recoveries often 90%or higher for known standards.Differences in technique efficiency appear to be method-and instrument-specific, rather than reflecting robust differences between the oxida-tion approaches,presenting further challenges for the aggregation and cross-comparison of published organic N data.Nevertheless,the organic N analyses of the20th century are
still an important data resource,because they relate to a period of very substantial changes in N deposition.
Onefinal challenge is that the WCO methods are prone to producing negative values for organic N.This may be partly because of the cumulative analytical uncertainty in calculat-ing organic N as the difference between a measured‘total’N and the inorganic N,but the difference(TN-IN)is often much larger than can be explained by analytical‘noise’,indicating that there are significant losses during oxidation(to a gaseous species or other unmeasured product),or perhaps problems with specificity of methods for IN analysis.Problems with interference in IN analysis are less likely as IC methods became the norm,because(in principle)amines and amino acids are separated from ammonium during analysis(but see Husted et al.,2000).The newer high-temperature(HTCO and HTO)methods appear to have fewer problems in this respect, but may still not capture the full range of organic N in a sample.For example,heterocyclic molecules or those with a cyanide(CN triple bond)produce molecular nitrogen(N2) rather than NO and so are not detected by chemilumines-cence(Yan et al.,2007).
2.2.Analysis of organic N components
The definition of WSON as the difference between total and inorganic N may be described as a‘top-down’approach.The alternative‘bottom-up’approach to the problem is to identify the individual components,and measure them separately. For WSON in the aqueous phase(rain or cloud),such analyses have been limited to amines,amino compounds(such as amino acids)and urea(Gorzelska et al.,1992;Kieber et al.,2005;Mace et al.,2003b,2003c;McGregor and Anastasio,2001;Zhang and Anastasio,2003b),and oxidised N compounds such as nitrophenols(Asman et al.,2005;Harrison et al.,2005;Lüttke et al.,1997).Recent developments in mass spectrometry have allowed the detection of a wide range of organic N compounds in precipitation(Altieri et al.,2009a,2009b).Intermediate in complexity are studies of organic functional groups,by infra-red and nuclear magnetic resonance spectrometry(Herckes et al.,2007),but these provide qualitative,rather than quantita-tive,information.None of these individually,however,ac-counts for most of the WSON,a large fraction of which appears to be associated with high molecular weight oligomers or polymers(Chen et al.,2010),described as‘humic like sub-stances’or HULIS(Kieber et al.,2005).
In the gas phase,this‘bottom up’approach is the usual course,with few studies on total gas-phase WSON(González Benítez et al.,2010;Zhang et al.,2002b).Gaseous reduced nitrogen compounds(Gronberg et al.,1992;Huang et al., 2009),nitrophenols(Delhomme et al.,2010;Römpp et
al., 2000;Schussler and Nitschke,2001),organic nitrates(both alkyl and acyl)(Buhr et al.,1990;Grosjean,2003;Jacobi et al., 1999;McFadyen and Cape,2005;Pippin et al.,2001;Zhang et al.,2009)have been measured.However,the largest suite of organic N compounds has been identified in the particulate phase,through the use of mass spectrometric techniques (Aiken et al.,2009;Ozel et al.,2010;Pratt et al.,2009;Ulbrich et al.,2009);some of the water-soluble components have been quantified either by direct sampling of aerosol into liquid(Lin et al.,2010;Sorooshian et al.,2008),or by disso-lution of particulate material trapped onfilters(see Sec-tion3.5.1).No single technique provides a full analysis of organic N in particulate matter,and the range of techniques applied to a‘bottom-up’approach is wide,varying from colorimetric analysis of one class of amines) to near-real-time analysis of total N using aerosol mass spectrometry.Some idea of the scope of current approaches is shown in Fig.2,which is based on a description of methods used for analysing particulate organic carbon(Hallquist et al., 2009).
The wide range of composition of atmospheric organic N means that measurements approach the analysis and interpre-tation of data from very different viewpoints(human health, eutrophication,nutrient supply,aerosol formation and visibil-ity),leading to great complexity in the literature,and only partial information from any one study.The complexity can be illustrated by contras
ting studies of proteins in particulate matter(Menetrez et al.,2007;Zhang and Anastasio,2003b),in which the focus is on the contribution of proteins measured in particulate matter(PM)to human health issues,and studies of the volatility of primary amines(Pratt et al.,2009),where the focus is on the chemical and physical behaviour of organic N in the atmosphere,and its role in particle formation.However, aspects from each study may prove useful in understanding the major sources and other factors that determine the distribution (in time and space)that are applicable more generally.
2.3.Data quality
2.3.1.Precipitation
Despite clear critiques of the problems of sampling,storing and analysing the organic N content of precipitation(Cape et al., 2001;Cornell et al.,2003;González Benítez et al.,2009; Gorzelska et al.,1992;Keene et al.,2002;Kieber et al.,2005; Markaki et al.,2010;Rolff et al.,2008;Scudlark et al.,1998; Violaki et al.,2010),many studies are compromised by the possibility of contamination during sampling,changes in com-position during sampling and storage,and differences in analytical methods used for estimating WSON.Precipitation may be sampled in bulk,or in wet-only collectors w
hich reduce the dry deposition of atmospheric gases and particles(includ-ing WSON)on the collector surfaces during dry periods.Where direct comparison of wet-only and bulk collectors has been made(González Benítez et al.,2009;Ham and Tamiya,2006;
33
J.N.Cape et al./Atmospheric Research102(2011)30–48
Ham et al.,2007;Jassby et al.,1994;Kram,2008;Markaki et al.,2010;Violaki et al.,2010;Zhang et al.,2008),there is usually a signi ficant contribution of ‘dry ’WSON to the total measured in the bulk collector,sometimes small (Kram,2008),and sometimes up to half the total (González Benítez et al.,2009).Although this leads to uncertainty in the true WSON content of precipitation,it indicates that there are major potential dry deposition sources of organic N,whether as gases or particles,that are even less well quanti fied than organic N in precipita-tion.Perhaps surprisingly,very few authors document their methods for eliminating obviously contaminated samples from their results.Precipitation collectors are favourite perches for birds,with the inevitable consequence of samples contam-inated by faeces,which are rich in ammonium,organic N,phosphate and potassium.Given the sensitivity of precipitation samples in a study of WSON to contamination from th
is source,clear criteria for eliminating contaminated samples should be developed and applied before further statistical analysis of the data.The presence of phosphate at concentrations above 0.1mg P litre −1is a useful diagnostic of contamination by bird droppings (UKEAP,2010),but other criteria are used
(Dämmgen et al.,2005;González Benítez et al.,2009).With the likelihood of contamination,and its importance to studies of organic N,it is surprising how few studies use replicate samplers at a single site,given the known variability in pre-cipitation sampling (Cape et al.,2001;Erisman et al.,2003;González Benítez et al.,2009).Even for identical bulk samplers sited within a few metres of each other,between-sampler variability can be surprisingly large.In one study of 38week-ly bulk samplers with 3replicates the standard deviation represented 8%of the mean for total N,and 25%for WSON (González Benítez et al.,2009).
Immediate freezing of the sample,or use of an effective biocide (e.g.thymol,rather than one that evaporates,such as chloroform),appear to be most effective for sample storage prior to analysis,but freezing may not always be effective (Gorzelska et al.,1992;Kieber et al.,2005).Prior to analysis,some researchers filter samples,others do not.Organic N prob-ably exists in a continuum from true solution through colloidal material to suspended particles,so the content of a solution after
filtration will be de fined by the filter used.However,the interpretation of the resultant ‘WSON ’may re flect greater or lesser degrees of solubility —there is no ‘correct ’method,but any filtration process should be fully documented.
2.3.2.Fog and cloud
Few studies record problems with sampling or storing fog and cloud water,although a range of different sampling tech-niques,both active and passive,have been used.Apart from the possibility of dry deposition of material on sampler surfaces during periods without fog or cloud,similar to issues of bulk precipitation sampling,there is evidence of problems with degradation of WSON during storage of frozen fog samples (Zhang and Anastasio,2001).
2.3.3.Particulates
Filter-based techniques are subject to the risk that material can react on the filter or impactor surfaces with gases such as amines,ammonia,nitric acid,peroxyacetyl nitrate or ozone,thereby changing their composition in situ,but in general any attempts to avoid subsequent reaction are not reported;the problem is probably less severe than for changes in precipita-tion samples,but no data are available.A denuder may be used to remove gaseous components upstream of the filter or other
analyzer (Lin et al.,2010),which would prevent adsorption of organic N on filters and previously collected hods used to sample organic carbon (Chow et al.,2010;Cornell,in press and references therein).Particulate organic N may be partly volatile (Pratt et al.,2009),and could be lost during sampling and/or storage.Organic N is also susceptible to oxidation (Angelino et al.,2001;Malloy et al.,2009;Murphy et al.,2007;Zhang and Anastasio,2003a ),which could occur during or after sampling,yet few precautions are recorded in the literature.
2.3.4.Gases
Gas-phase measurements of organic N are subject to the same problems as sampling any gas-phase component of the atmosphere.Care must be taken in ensuring exclusion of particulate material,with the possibility that volatilisation of particles may occur from pre-filters,as noted above.If samples are trapped prior to of fline a reactive
100
s i n g l e c l a s f e w c l a s s e m a n y c l a s s e m a s s f r a g m e n t c l a s s e s
f u l l i d e n t i f i c a t i o c o m p l e t e n e s s (% o r
g a n i c N a n a l y z e d )
Functional groups of ON
Fig.2.Representation of the range of analytical methods employed for the characterization of organic N.Those in blue are applied to discrete filter or solution samples;those in orange are applied in near-real-time.‘Class ’refers to molecular functional amines or nitrophenols.Completeness (vertical axis)is expressed as a%of the total organic N in the samples (precipitation,filter,filter extract,or direct sampling)where that is known.Based on Hallquist et al.(2009).Abbreviations:FT-ICR-MS:Fourier Transform Ion Cyclotron Resonance Mass Spectrometry;HR-ToF-AMS:High Resolution Time-of-Flight Aerosol Mass Spectrometry;GC –MS:Gas chromatography –mass spectrometry;GC –GC-ToFMS:2-dimensional Gas Chromatography Time-of-Flight Mass Spectrometry;LC-GC-MS:Liquid column Chromatography-Gas chromatography –mass spectrometry;IR:Infra-Red spectro
metry;NMR:Nuclear Magnetic Resonance spectrometry;ESI-MS-MS:ElectroSpray Injection-2dimensional Mass Spectrometry;IC:Ion Chromatography;HPLC:High Performance Liquid Chromatography;ATR-FTIR:Attenuated Re flectance –Fourier Transform Infra Red spectrome-try;Py-GC-MS:Pyrolysis Gas Chromatography Mass Spectrometry;Fluor:Fluorimetry;Color:Colorimetry;GC-ECD:Gas Chromatography-Electron Capture Detection.
34J.N.Cape et al./Atmospheric Research 102(2011)30–48
版权声明：本站内容均来自互联网，仅供演示用，请勿用于商业和其他非法用途。如果侵犯了您的权益请与我们联系QQ:729038198，我们将在24小时内删除。
发表评论