摘要:

J. Environ. Monit. 1999 1 83N Focus Public environmental information Access or excess? Amongst the mass of academic papers and policy reports it is easy to forget that scientists and regulators are not the only ones concerned about the environment. The public has an interest too. Aided by modern technology and increasing legal rights the man in the street has access to environmental information as never before. In this article we review recent developments in public assess and what these mean for �the experts�. When UK pressure group Friends of the Earth (FOE) launched a website cataloguing industrial pollution data earlier this year it caught the attention of business and the public alike. The site called FactoryWatch allows people to examine pollution emissions within their area based on oYcial emissions data and to compile local maps and league tables (see Box 1).Of course much of this information has been in the public domain for a long time. But who except the ardent campaigner would go to the trouble of requesting data from a regulator? Putting the information on-line is a diVerent matter with modern technology a company�s environmental record can be put on show for all to see�as those highlighted by FOE as �filthy factories� soon found out. Companies suddenly began to realise how easy it can be for people to get environmental data in the age of the Net. Factory Watch is just one example of an increasing array of initiatives that provide individuals and communities with access to environmental data. Against a backdrop of an evolving legal framework and rapid developments in technology citizens have opportunities to access environmental information with an ease and speed never seen before.But what exactly do we mean by �access� and why is it important? Does the public really need all this data? A right to know Over recent years definitions of �environmental information� have broadened as the environmental debate itself has grown.1 At the most basic level the term refers to information arising from environmental legislation and other policy measures and which is held by public authorities. However much important environmental data is also held by private companies. With the growth of ideas of sustainability environmental information has come to mean all information relevant to decision-making on the environment embracing financial social and economic data as well as environmental data.So what is this information exactly? While some topics and types of data are requested more frequently diversity is a key characteristic of the public�s information needs. Surveys show people�s interests in accessing data include:2 $ Links between environmental conditions and human health especially within local communities; $ The condition location and trends of natural systems such as lakes rivers animal and plant populations; $ EVects of human activities on the environment such as industry agriculture transport and economic development; $ Costs and results of public initiatives to protect or remediate the environment; $ Learning and education on environmental issues and policies; $ Opportunities and conditions for recreations such as hunting fishing camping etc.A distinction can be made between rights to �active� information which public authorities or private companies are legally obliged to provide and �passive� information which the public has a right to obtain on request.1 Much of the public access debate focuses on making environmental information more transparent so that citizens have at least a passive right to all relevant data. Access is not to be confused with the issue of public participation in environmental decision-making although clearly the two are closely linked. Specifically there is assumed to be a causal link between access to information stakeholder awareness and improved environmental conditions citizens must be informed about policymaking so as to make reasoned decisions about it. Easy access to environmental information it is argued Box 1 Factory Watch Factory Watch (www.foe.co.uk/factorywatch) was launched by Friends of the Earth in February this year.Using oYcial data on pollution emissions collected by the Environment Agency the site provides online access to information on industrial chemicals releases for the first time. Users can compile league tables of polluting factories and make local maps showing the location of for example releases of cancer-causing chemicals. Most information relates to 1996 the most recent year for which full emission inventories are available but with some data for 1998. While the site marks a great step forward it also highlights major shortcomings in access to environmental information in the UK. For example $ The data only cover England andWales equivalent information for Scotland and Northern Ireland is not even collected; $ Data on smaller factories or other pollution sources such as transport is not readily available; $ Basic toxicity data needed to assess health impacts is lacking; $ Omissions in the reporting requirements and inconsistencies in the categories used; $ Inadequate checks on data quality and reliability.84N J. Environ. Monit. 1999 1 Focus should enable citizens to get more involved and give individuals and communities better tools for local action. Issues relating to public access are generally distinct from those of access by professionals or experts such as scientists government agencies and businesses. Key issues for this market include factors such as technical standards which are not discussed here.However it is important to note that professional private and public data users are a critical link in meeting the public�s information needs. Many environmental questions are too complex for most citizens to understand on their own and they look to professional users for help in assessment and interpretation. The international framework Many countries have enacted laws guaranteeing the public right of access to information on environmental issues. One of the first international commitments in this area was the European Union Directive 90/313/EEC on Freedom of Access to Environmental Information passed in 1990 which established for the first time a set of minimum standards within the EU.1 The Directive contains a general rule of access for any person without the need to show an interest in the information requested but also allows for a number of exemptions and sets limits on response times appeals and costs.All EU member states have implemented the Directive within national law although substantial variations still exist in rights and procedures between countries.3 Revisions to the Directive are likely following an international agreement signed by 35 European governments in June 1998.3,4 Negotiated under the UN Economic Commission for Europe (UNECE) the so-called Aarhus Convention commits the parties to legally binding rights for citizens in access to information and involvement in environmental decision-making. The convention defines minimum rights of access for citizens and NGOs enabling them to apply for information and take part in local or national decision-making processes.While welcoming the convention NGOs say that the mechanisms for enabling access to private sector information and for noncompliance are still too weak.5 Agenda 21 contains many references to information and public participation and Principle 10 of the Rio Declaration on Environment and Development stressed that �environmental issues are best handled with the participation of all concerned citizens at the relevant level�. One outcome of Agenda 21 has been the strengthening of INFOTERRA a network of information centres operated by the UN Environment Programme.6 Now spanning 176 countries INFOTERRA is one of the main actors in improving environmental information capacity in developing countries. Another important UN mechanism is GRID a network of 14 data centres providing geo-referenced data and information to support assessments reporting and policy formulation.This too is being strengthened as a result of the Agenda 21 process. Two cultures The United States wasnational information laws. The main legislation currently comprises the Emergency Planning and Community Right-to-Know Act (EPCRA) 1986 the Superfund legislation on contaminated sites and most recently the Chemical Right-to- Know Initiative 1996. EPCRA was introduced in response to the Bhopal disaster in India in 1984 and a similar release at a sister plant in West Virginia. Its main purpose is to inform communities and citizens of chemical hazards in their areas. Sections 311 and 312 of EPCRA require businesses to report the locations and quantities of chemicals stored on-site to state and local governments and section 313 requires manufacturers to report environmental releases of more than 600 designated toxic chemicals.EPA compiles this information into a publicly accessible on-line database the Toxics Release Inventory (TRI).7 All information is published and is freely available on the Internet�the dataset for the whole US accounts for around 200 MB of computer disk space. This vast new source of data is proving a powerful force for environmental improvement. By encouraging dialogue between individuals and local companies TRI has resulted in changes in operating Box 2 Making an EMPACT EMPACT�Environmental Monitoring for Public Access and Community Tracking�is an EPA programme to provide people with up-to-date environmental information they can understand and use every day.The Agency claims EMPACT represents a new approach to the delivery of accurate timely and useful environmental and public health information directly to communities and individuals. The initiative aims to apply the latest measurement information management and communications technologies in tracking environmental conditions. Information should not only be accurate but useful and should be presented in plain language directly to the communities concerned. All EMPACT projects are required to maintain a website and to adopt best practice in communicating their results to the public.11 Communication channels used include Internet CDs television radio newspapers fliers billboards local meetings community organisations and �environmental teller machines� (similar to those used in banks).Focusing on the largest metropolitan areas EPA has initiated a series of pilot projects that include $ Tracking of toxic air pollutants in the San Francisco and Cleveland areas; $ Real-time information on water quality at Los Angeles beaches; $ Monitoring of children�s lead exposure in the Boston area; $ Tracking of water quality in Long Island Sound and selected urban areas. Grants are available for similar projects that show innovative and eVective ways to track and deliver environmental information with an emphasis on community partnerships. Further information at www.epa.gov/empact J. Environ. Monit. 1999 1 85N Focus practices and improvements in environmental standards. Public interest groups use TRI to bring public pressure to bear on facilities and public oYcials.For trade unions it provides important information on workplace hazards. And there are commercial benefits too. Manufacturers have used TRI to cut costs and improve operations through reduced use of chemicals and to identify new business opportunities through cleaner safer or more cost-eVective alternatives. Law firms real estate companies and banks use TRI to identify liability issues. Most significantly the publicity arising from TRI data has caused many companies to volunteer pledges to reduce toxic chemical releases. In Europe the picture is more fragmented. Despite the 1990 directive significant variations in requirements and procedures still exist between countries and in general citizens have fewer rights of access than in the US.In the UK for example none of the pollution registers are available on-line and much of the emission data is protected by commercial confidentiality. The Environment Agency only set up on-line access to an emissions database earlier this year based on the same data that FoE uses in Factory Watch.8 Daily air quality reports are available online but only at a highly aggregated (regional ) level.9 As well as diVerences in legal rights the situation in Europe has been complicated by variations in definitions and collection methodologies between countries. EIONET�the European Environment Information and Observation Network�is a European Environment Agency initiative to standardise data collection and reporting procedures across Europe. EIONET links more than two hundred organisations in EEA member countries and Central and Eastern Europe many of which are already linked by computer networks.From this autumn the public will have access to this substantial information base through a single webbased gateway called E2RC� the European Environmental Reference Centre.10 The key diVerence between the US and Europe is one of culture. In the US there is a culture of disclosure whereas in Europe there is a culture of nondisclosure no matter what the spirit of the legislation. This climate of secrecy is deeply engrained in European governments institutions and companies and will take a long time to overturn. Such attitudes are not confined to environmental information but it is often the environmental field where they are in sharpest relief.With these cultural legal and methodological diVerences it is likely to be a long time before Europeans enjoy the same rights to environmental information as US citizens have already. Acting locally Even in the US reliable information on local environmental quality and associated risks has been lacking. Two recent programmes aim to help fill these gaps. EMPACT focuses on real-time information that communities and individuals can use on a day-to-day basis (see Box 2).11 Projects employ the latest in measurement technology and disseminate the data through channels such as the Internet newspapers broadcasting and community meetings. Chemical Right-to-Know (ChemRTK) is addressing the need for publicly available toxicity data for high production volume (HPV) commercial chemicals.12 EPA studies have shown even for relatively common chemicals a lack of basic hazard information needed for risk assessment.The ChemRTK Initiative is an ambitious eVort to tackle the problem by rapidly testing chemicals and making this important data available to scientists policy-makers industry and the public. Earlier this year chemical manufacturers and importers were invited to participate in a voluntary challenge programme to provide basic toxicity data on the HPV chemicals they produce (see JEM 1999 1(3) 42N). Chemicals not adopted will be tested separately by EPA and those of particular concern to children�s health are being singled out for more detailed and extensive testing. The most recent measure takes the whole approach to public communication a stage further�into the proactive delivery of information rather than waiting for active or passive enquiry.Under the measure EPA is mandating all municipal water authorities in the US to mail out �consumer confidence reports� to every home and business connected to public water supplies. The reports to be distributed by October this year list the concentrations of every substance present in the water supply and in some cases include additional information such as number of miles of water main number and location of hydrants and local contact numbers. Utility oYcials are having to engage in an elaborate communication exercise to reassure consumers they need not be alarmed by what to many may seem an alarmingly long list of contaminants. Trends and best practice What conclusions can we draw about the public�s information requirements and how as scientists and regulators can we aim to address these? Studies in Europe and the US show a strong correlation between pollution emissions poverty and human health.In its analysis of UK emission data for 1998 Friends of the Earth found 75% of the top 25 factories were located in areas of below average household income.13 In democratic societies public agencies have a responsibility to provide people with information to enable them to make informed decisions on their health and welnformation programmes that speak to the person in the street addressing the questions they ask and answering them in a way they understand. This has implications for both the design and implementation of public access policies. In terms of policy design key factors to be considered include:2 $ People want information not data.Raw data is of little value on its own. It must be presented in a way that enables users to understand the significance of data values i.e. with suitable context and qualification. Hence public access becomes an exercise in risk communication. $ People are often looking to answer specific questions such as �Is my drinking water safe?� �What is the air quality in this neighbourhood today?� or �How much has the government spent on local clean-up?�. Agencies need to understand these concerns and address them directly. $ Information requests are often personal and local in nature. For example people want to know about water quality or transport emission in their area or opportunities for personal recreation such as fishing.Data collection programmes need to be suYciently diverse and disaggregated � with obvious cost implications. $ Maps are a favoured format in many information requests emphasising 86N J. Environ. Monit. 1999 1 that people find spatial representations a convenient way to summarise information on the environment and natural resources. $ Real-time data presents quality control problems. Raw data that is collected and updated in real-time or near real-time can be prone to errors. Yet introduction of thorough screening and quality control can delay the release of data and so undermine its value. Management procedures need to provide appropriate safeguards and users should be warned of the potential for anomalies and errors. $ Information requirements cut across agencies and programmes.The public has no respect for administrative boundaries. To be successful therefore information programmes require a high degree of inter-agency co-operation and data-sharing between agencies and across diVerent levels of government. Turning to the medium rather than the message there are three main trends (1) The Internet is a key enabling technology In this sphere as elsewhere the Internet is bringing fundamental change. With a modest PC modem and browser software any user can now access environmental information in a way that was unimaginable even five years ago. These datasets include not only historical information on local emissions such as those accessible through Factory Watch and TRI but increasingly real-time data on local environmental conditions such as those developed under EMPACT.Highly localised online datasets will become an increasingly important information resource. (2) Mass media as a complement to online channels Despite its in the environmental field include electronic displays and information signs alerting drivers to changing traYc conditions. Notes 1 Public Access to Environmental Information Ralph E. Hallo European Environment Agency Experts Corner Number 1 1997. Available at www.eea.eu.int 2 Information Needs Assessment Summary of Survey Results Foundation Project Minnesota Planning St Paul Minnesota April 1998. 3 For a review of national policies and trends in the EU see Doors to Democracy Current T rends and Practices in Public Participation in Environmental Decisionmaking in Western Europe Regional Environmental Centre Budapest June 1998.Available at www.rec.org 4 T owards an ECE Convention on Access to Environmental Information and Public Participation in Environmental Decisionmaking Jeremy Wates European Environmental Bureau Brussels. 5 ENDS Daily 25th June 1998. See www.ends.co.uk 6 For information on INFOTERRA Agenda 21 and UNEP programmes see http://unephq.unep.org 7 TRI data is available online at www.epa.gov/opptintr/tri 8 For Environment Agency database see www.environment-agency.gov.uk and FactoryWatch see www.foe.co.uk/factorywatch 9 The UK National Air Quality Information Archive www.aeat.co.uk/netcen/airqual/ 10 No date available for launch of E2RC at time of going to press. For latest news see EEA website at www.eea.eu.int 11 EMPACT Information Management Handbook Environmental Protection Agency 1998.Available at www.epa.gov/empact/ 12 For details of ChemRTK see www.epa.gov/oppintr/chemrtk 13 Friends of the Earth Press Release 30th May 1999. Available at www.foe.co.uk Mike Sharpe phenomenal growth many people do not yet have Net access especially within the socially disadvantaged groups that may benefit most from local information. Hence it is important that other tools besides theWorld Wide Web are used to convey environmental information to the public. Mass media such as television radio and newspapers are still able to reach much larger audiences than the Web but have their own limitations. Television radio and to a lesser extent newspapers are not well suited to delivering complex information whereas web pages allow a more detailed presentation.Secondly mass media are not yet interactive and so do not allow users to tailor information to their own preferences or provide feedback. Thirdly mass media are not able to provide real-time data as readily as a website. In the longer term the Net itself will be accessible through these mass media through developments such as interactive TV and third generation mobile phones. (3) Innovative uses of existing technologies Existing technologies also oVer opportunities to access environmental information in new ways.10 In Sacramento for example residents are alerted about hazardous ozone concentrations by e-mail and telephone. A speech recognition system developed by MIT allows users to call in for weather information from over 500 locations. Under an EMPACT project called NEWNET �environmental teller machines� similar to bank teller machines use interactive computer screens to display information on radiation exposure. Other mature technologies that are b

ISSN:0960-7919    

DOI:10.1039/a907152k

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

J. Environ. Monit. 1999 1 87N News Legislation Finns unveil EU priorities Four new draft directives on air quality are amongst Finland�s top priorities for its presidency of the EU. The proposed measures include an air quality directive to limit benzene and carbon monoxide concentrations in ambient air and a proposal to set targets for ambient levels of ground level ozone. A national �emissions ceiling� directive will limit emissions of sulfur dioxide nitrogen dioxides volatile organic compounds and ammonia. The large combustion plants directive aims towards further reductions in emissions from power generation and other large facilities. (See JEM 1999 1(4) 59N). The air quality directive is likely to receive the quickest passage although member states remain split over limit values for benzene.Southern countries are likely to push for a limit higher than the 5 mg m-3 proposed by the Commission while Nordic countries are in favour of stricter limits. The emissions ceiling and LCP directives will be progressed in parallel followed by the ozone directive if time allows. Other major negotiations falling under the Finns responsibility include a proposal for mandatory strategic environmental assessment of infrastructure plans and programmes and a proposal to set minimum standards for environmental audits. Finland assumed the presidency which rotates between EU member states every six months from the beginning of July and holds it until the end of the year. Finnish Presidency http://presidency.finland.fi EPA cleans up waterways Revisions to the Clean Water Act regulations aimed at improving the quality of 20 000 waterways were announced by President Clinton in August.Under the EPA proposal the agency will work with state and local governments to develop more comprehensive assessments of water bodies so as to identify those at risk from pollution and/or failing water quality standards. Clean-up plans and timetables will be developed (within 5 years for �high priority� waters) where necessary with stricter discharge limits for both point and non-point sources. Implementation of the proposals is likely to involve changes to EPA�s Total Maximum Daily Load Water Quality Standards and National Pollutant Discharge Elimination System programs. Details at EPA OYce of Water www.epa.gov/owow/ tmdl/proprule.html Progress on Lake Superior plan Agencies in the US and Canada have published detailed targets for reducing pollution discharges into Lake Superior.The move is an important element of the Lake Superior Lakewide Management Plan (LaMP) which is part of the bilateral US-Canadian program to restore and protect the Lake Superior basin. LaMP Stage 2 targets pollutants which impair or are likely to impair human health and wildlife and beneficial uses of the lake. Critical pollutants are mercury PCBs dieldrin/ aldrin chlordane DDT and metabolites toxaphene dioxins hexachlorobenzene (HCB) and octachlorostyrene (OCS). These are all persistent in the environment and bio-accumulate in animals and humans causing cancers neurological diseases and chronic health problems. The report sets schedules for reducing these critical pollutants across the Lake Superior basin which will be updated and reported on in future stages.Environment Canada www.ec.gc.ca EU enforces water laws The European Commission has announced a new round of legal actions against EU member states for noncompliance with European laws on water quality. Firstly a series of infringement actions are being taken out for failing to implement the 1991 nitrates directive. Both Germany and Luxembourg are being referred to the European Court of Justice while the Netherlands France and Belgium have been sent �reasoned opinions� (final warnings). The directive which aims to reduce nitrate pollution from agriculture is one of the most poorly implemented of all EU environmental laws and infringement actions are already underway against 12 of the 15 member states. Another key problem area is the 1991 urban wastewater directive.Greece has been referred to the European Court for inadequate compliance with the directive. Along with Germany Greece is also to be sent a reasoned opinion on a 1998 amendment to the directive clarifying requirements for urban wastewater discharges in areas sensitive to eutrophication. Neither country has implemented the legislation within national law. Ireland Finland Portugal and the UK are also to be ticked oV for minor infringements in particular areas of water quality legislation. OP pesticide ban EPA has announced cancellation agreements and risk reduction strategies for two organophosphates which are amongst the oldest and most widely used pesticides. Specific uses of the pesticide methyl parathion are to be eliminated while allowable residues for azinphos methyl are to be lowered for a wide variety of fruit and vegetable products.The Agency says the move is aimed particularly at reducing risks to children�s health from pesticide exposure. EPA has set a rigorous 18-month schedule for completing its review of all 39 organophosphates. Several other older widely used pesticides are also being targeted for priority review within the next year and a half including atrazine aldicarb and carbofuran. The review is part of a move to more protective health-based standards as required under the 1996 Food Quality Protection Act (FQPA). EPA OYce of Pesticide Programs www.epa.gov/pesticides Cadmium tops EU agenda A draft directive banning cadmium in batteries was amongst the first 88N J. Environ.Monit. 1999 1 News environmental measures to be considered by the new EU commissioners when they started work in September. Under an extension to the 1991 directive on battery waste the Commission is seeking to introduce a ban on all batteries containing cadmium by 2008.Member states would also be required to ensure that 75%of consumer and 95%of industrial batteries are recycled. The main target is rechargeable batteries made of nickel cadmium. research shows nickel hydride batteries present an economically feasible substitute. Although currently 20�100% more expensive it says prices of nickel hydride technology should drop dramatically over the next five years. Other parts of the Commission are believed to be more sympathetic to the industry�s view however. International Cadmium Association tel.+32 2 7770560 Industry is resisting the move claiming that viable alternatives to rechargeable Ni�Cd batteries have yet to be developed�especially for high capacity applications such as power tools.It is arguing instead for the opportunity to set up an EU-wide battery collection scheme which would ensure the recycling of 75% of Ni�Cd batteries by 2004. The Commission�s Environment Directorate says the recycling target is unattainable whereas Environmental quality Europe�s �weather forecast� The European environment remains �under serious pressure� according to a major survey of environmental quality and trends published by the European Environment Agency (EEA). In what it calls a �weather forecast� through to 2010 the Agency says the unsustainable development of sectors such as transport energy agriculture and tourism represents a major barrier to environmental improvement.In transport and agriculture in particular the Agency finds �no indication of significant ecoe Yciency gains ...up to 2010�. Amongst the positive developments noted are the significant cuts in ozonedepleting substances and reductions in acidification and phosphorous discharges to rivers. Overall however the agency says there has been insuYcient decoupling of pollution emissions from GDP. The outlook foresees these trends continuing through to 2010 with particular problem areas in greenhouse gas emissions chemicals and waste. Lack of data continues to hinder projections in some areas such as soil quality biodiversity and pesticides in groundwater. Other areas remain uncertain due to socioeconomic developments.In chemicals the report notes that �total chemicals production is on a rising trend while minimal risk assessment analysis is not carried out for 75% of the large volume chemicals on the als like cadmium and copper and of hazardous chemicals from industry transport and agriculture such as pesticides are expected to rise. The study the fourth of the EEA�s �state-of-the-environment� reports describes primarily the 15 EU member states but also covers the EFTA countries and the 11 accession countries. The report diagnoses and measures the situation for most recent years available and also assesses future trends. The baseline �business-as-usual� scenario assumes a 45% increase in economic growth over the period 1990�2010 and a 50% increase in final consumption between 1995�2010.Environment in the European Union at the T urn of the Century European Environment Agency Copenhagen 1999. Available at www.eea.eu.int The State of the Environment Reporting Information System (SERIS) is a gateway to information about the state of the environment in European countries and regions www.eea.eu.int/ frdb.htm. Experts count cost of Balkans conflict Teams of UN experts visited Yugoslavia over the summer to assess the environmental impact of the Kosovo crisis. In July a team of 12 scientists visited industrial sites across the country taking soil and groundwater samples. The experts also collected data on earlier (pre-conflict) pollution incidents and gathered data such as maps and records of air pollution and hazardous wastes.Speaking to reporters at the end of the visit Pekka Haavisto chairman of the Balkans Task Force said it was too early to reach any firm conclusions on the overall impact of the conflict. Further missions in August looked at the impacts on the Danube the region�s biodiversity and human health. The BTF was due to report to the UN Secretary-General in September. Balkans Task Force www.grid.unep.ch/btf Equal right-to-know Data on toxic chemical releases across North America are available in a new report from the North American Commission for Environmental Cooperation (CEC). The report entitled Taking Stock presents an international picture of the releases and transfers of toxic chemicals by industrial facilities. This third annual report covers the situation in 1996 drawing on data from the US Toxics Release Inventory (TRI) and the Canadian National Pollutant Release Inventory (NPRI).Comparable data from Mexico�s environmental agencies will be included in future editions. At a recent CEC Council Meeting in BanV the three countries agreed to step up eVorts to improve the comparability of data between their right-to-know programs. North American Council for Environmental Cooperation www.cec.org Ireland feels pressure Environmental indicators issued by the Irish government highlight the increasing pressure on the country�s environment from the high rate of economic growth. In the first quantitative study of the country�s environmental performance the Irish EPA presents data for fifty indicators over a ten-year period. A series of twelve headline indicators is identified for which regular monitoring will be undertaken to provide a snapshot view of progress over time.These focus on key issues for national policy such as phosphorus levels in soil. In line with other recent studies (see JEM 1999 1(3) 61N) the results show particular pressures in controlling river pollution where eutrophication from J. Environ. Monit. 1999 1 89N News agriculture is a major problem. Waste management is also a significant problem with arisings of household and commercial wastes increasing by 62% since 1995. Irish EPA www.epa.ie German water management �unsustainable� Water management policy in Germany is becoming increasingly unsustainable according to a highly critical report from the country�s environment agency. In a study of prospects for water management through to 2020 Umweltbundesamt says the overall water pollution situation has not improved significantly over the last 20 years.The report goes on to point the finger firmly at non-industrial sources such as households transport and agriculture. Of the 820 000 tonnes of nitrogen entering watercourses in Germany each year 60% is from agricultural sources. This needs to be halved over the next 20 years as part of a sustainable water management regime. Other practices singled out in the report as �unsustainable� include the escalating compensation payments to farmers the increased use of pesticides and the further development of lignite mining. Although some have argued that privatisation would be the best way forward the agency sees public ownership as being more conducive for the necessary policy changes.Umweltbundesamt www.umweltbundesamt.de California plans groundwater inventory The State of California has launched an initiative to improve the quality of information on groundwater resources. Under the 3-year project the Department ofWater Resources (DWR) will prepare a statewide inventory of groundwater data develop a model groundwater regulation for local governments and create guidelines for evaluating local groundwater management plans. Data collected from local water agencies will be combined with well-drillers� reports and DWR information on the state�s 500 groundwater basins. DWR is organising statewide consultation from end of August with a draft report due by 2001 and a final report in 2002. California Department of Water Resources Debra Carlisle Project Manager carlisle@water.ca.gov Chemical hazards Momentum grows for phthalate ban Pressures for an EU-wide ban on phthalate plasticisers grew over the summer as two of the EU�s largest members Germany and France moved towards restrictions.France introduced a one-year ban from beginning of July on childcare items made from soft PVC containing phthalate softeners. The decree halts the sale import export and manufacture of items such as teethers and dummies and requires the removal of all articles currently in circulation. Germany is due to finalise a similar law by the autumn following a critical report from the country�s environment agency. In its study published in June Umweltbundesamt said the use of flexible PVC should be phased out completely because of risks from plasticising agents and that cadmium and lead should be phased out as stabilisers in rigid PVC.Representatives of Germany�s PVC industry said the report used old data and had not been subject to peer review. To date eight European countries have introduced or are in the process of introducing bans on these chemicals following increasing concerns over the migration of phthalates from PVC items into saliva (see JEM 1999 1(3�4)). These latest moves are significant in that Germany and France are the largest countries to adopt a ban. Britain and the Netherlands have opted instead for migration tests and the other largest members Spain and Italy are still undecided. The European Commission is due to pronounce a new policy covering all uses of PVC by the end of the year.Meanwhile the EU�s use of the precautionary principle in this and other areas of environmental policy has been criticised by a leading academic. In a pamphlet published by the European Science and Environment Forum (ESEF) an environmental science association Bill Durodie� of the London School of Economics argues that the EU�s approach to environmental policies is leading to a �paralysing sensitivity to risk�. Citing the phthalates case as an example Durodie� says there is �not a single shred of evidence that they have ever harmed a human being� and that recent moves towards legal restrictions are the result of �unsubstantiated scientific papers�. His wider argument is that the precautionary approach is not �a zero cost-option either financially or socially� and that an urgent and wide-ranging critique is needed of the usage costs and limitations of this approach.We will be looking at the phthalates debate in a future issue of JEM. Meanwhile why not tell us what you think about the scientific evidence and how it is being used and interpreted? �Fields of Action and Criteria for a Precautionary and Sustainable Substance Control Policy illustrated by the example of PVC� from www.umweltbundesamt.de; �Poisonous Dummies European Risk Regulation afries of the �Commonwealth of Independent States� (CIS) are having to face up to the legacy of inadequate management of toxic substances. In two separate meetings over the summer experts from these and other countries meet with international organisations to focus on solutions for public health and the environment.The first meeting in Moscow addressed technologies for treatment and destruction of polychlorinated biphenyls and obsolete pesticides. A subsequent meeting held in St Petersburg addressed data collection needs based on the Pollutant Release and Transfer Register (PRTR) model. Pilot PRTRs are already being set up in Kazakhstan the Russian Federation Ukraine and Uzbekistan. Both meetings were facilitated by UNEP as part of a major programme to 90N J. Environ. Monit. 1999 1 News trade in dangerous chemicals has been signed by more than 60 countries since its agreement last year. Under the PIC regime exporters of chemicals must obtain consent of the receiving country before trading in a listed substance.UNEP PIC website www.pic.org Further progress on TBT and asbestos Further progress has been made towards legislation reported in previous issues of JEM banning two classes of carcinogenic chemicals. At its meeting in London in July the International Maritime Organisation (IMO) called for a conference within two years to finalise a treaty banning tributyltin (TBT). At the meeting major shipping countries supported the adoption of an international treaty by 2001. Use of anti-fouling paints containing poisonous organotin compounds would be prohibited from 2003 with an outright ban from 2008. (see JEM 1999 1(1) 11N). In a separate move the European Commission has formalised its longproposed ban on white asbestos (see JEM 1999 1(4) 62N). The regulation in the form of an amendment to the dangerous substances directive (76/769) requires all EU states to phase out white asbestos also known as chrysotile by 2005.Nine EU countries already have restrictions and the main eVect is to harmonise regulations across the EU. However the move has been heavily criticised by Canada the main exporter of chrysotile as infringing WTO rules. IMO www.imo.org; European Commission http://europa.eu.int/comm strengthen national chemicals management in CIS countries. UNEP www.chem.unep.ch/prtr Extension to PIC list Two toxic agricultural chemicals have been added to the list subject to strict global trade controls under the UN�s prior informed consent (PIC) procedure. At a recent meeting signatories to the Rotterdam Convention voted to add binapacryl a fungicide and miticide and toxaphene an insecticide to a list of five industrial chemicals and 22 pesticides already on the list of PIC-controlled substances.Four other substances are also being considered for inclusion on the list�bromacil ethylene dichloride ethylene oxide and maleic hydrazide. The Convention which mainly aims to protect developing countries from Public and occupational health Diesel studies need more data On-going eVorts to estimate the risk of lung cancer from exposure to diesel engine emissions would benefit from improved exposure measurements and further analysis of existing worker studies. These are the main findings of a major study on diesel epidemiology issued by the US�s Health EVects Institute (HEI). For the study HEI�s Diesel Epidemiology Expert Panel reviewed recent literature on risk of lung cancer mortality associated with diesel exhaust exposure.The Panel noted that the data sets were complex and that assumptions made by some of the investigators should be re-examined. The Panel�s 10 recommendations to improve risk assessment in the future include improved models for historical reconstruction of exposures; development of an improved marker specific to diesel particulate matter; and reliable estimates of historical exposures. Non-cancer health outcomes and biomarker development should also be considered. Diesel Emissions and L ung Cancer Epidemiology and Quantitative Risk Assessment Health EVects Institute June 1999. Available at www.healtheVects.org PDBE study sparks exposure fear A study of brominated flame retardants has shown new evidence of occupational exposure for workers in the electronics equipment recycling industry.The study by Swedish scientists found levels of PBDE (polybrominated diphenyl ethers) in blood up to 70 times higher than normal for staV at an electronics dismantling plant. The study also found evidence for the bioavailability of BDE-209 a high molecular mass compound not previously thought to bioaccumulate. DecaBDE is a high volume production chemical produced in quantities of around 30 000 tonnes per year. The research team said the evidence for the bioavailability of BDE-209 should lead to greater eVorts to reduce environmental concentrations. The report will add to pressures in Sweden for a ban on PBDE and related compounds following a critical report from the National Chemical Inspectorate earlier this year (see JEM 1999 1(3) 40N).In the light of the findings the Bromine Science and Environment Forum representing the world bromine industry has launched its own investigation into how to minimise workers� exposure. The new study involves testing workers from recycling facilities across Europe and will be completed by the end of the year. �Flame retardant exposure polybrominated diphenyl ethers in blood from Swedish workers� Environmental Health Perspectives vol.107 no.8 August 1999. http://ehis.niehs.nih.gov Weak evidence for EMF cancers Risk of cancer and other diseases from electric and magnetic fields (EMFs) around power lines is weak according to a recent scientific review. The report by the US National Institute of Environmental Health Sciences (NIEHS) follows six years of intensive research.NIEHS looked at the extremely low frequency EMFs surrounding both power distribution lines and low voltage electric distribution systems in the home. While the report finds some evidence that EMF exposure �cannot be recognized as entirely safe� it concludes �The NIEHS believes that the probability that EMF exposure is truly a health hazard is currently small. The weak epidemiological associations and lack of any laboratory support for these associations provide only marginal scientific support that exposure to this agent is causing any degree of harm.� The strongest evidence for health eVects comes from statistical associations observed with childhood leukemia and chronic lymphocytic leukemia in occupationally exposed adults such as electric utility workers machinists and welders.NIEHS recommends that research on these �lingering concerns� should continue together with eVorts to reduce occupational exposures. T he NIEHS report is available online at www.niehs.nih.gov/emfrapid/ J. Environ. Monit. 1999 1 91N News Research activity Air quality innovation A new database using highly innovative analytical techniques will help measure airborne pollutants from manufacturing plants and other sources with greater accuracy. Developed by the US�s National Institute of Standards Technology (NIST) the Quantitative Infrared Database has been designed to calibrate and verify measurements made with infrared-based analytical instruments in field monitoring of hazardous air pollutants. According to NIST the database is the first to assess infrared spectra for priority volatile organic compounds (VOCs) using new Fourier transform infrared (FTIR) techniques approved by EPA.Measurements were made on primary gas standards prepared and verified at NIST. From an initial 21 VOCs researchers plan to expand the database to around 100 of the 189 compounds regulated under the Clean Air Act. The database is supplied on CD-ROM with updates available over the Internet. In addition to emissions monitoring NIST says the technique has potential applications in monitoring CFC gases in semiconductor manufacturing measuring the components of natural gas and detecting gaseous chemical weapons. NIST Standard Reference Data Program www.gases.nist.gov Shared water prize The 1999 StockholmWater Prize has been awarded jointly to scientists from the US and Switzerland.Jaalifornia Institute of Technology Pasadena and Werner Stumm emeritus professor at the Federal Institute of Technology Zu� rich share the prize for their outstanding contributions to aquatic chemistry. Their citation said that over the past two decades the two scholars have contributed significantly to understanding how phosphorus iron manganese and other substances react in water leading to many practical applications in water and wastewater filtration. They were also credited with fundamental discoveries that helped validate the concept of acid rain and understanding of how metals and other substances are transported in the environment. The prize honours outstanding research action or education in the field of water and is administered by the Stockholm InternationalWater Institute on behalf of the Stockholm Water Foundation.SIWI www.siwi.org New pesticides group for EPA EPA�s OYce of Pesticide Programs is to set up an expert group to advise on communication on pesticides. The Inert Disclosure StakeholderWorkgroup will advise the Pesticide Program Dialogue Committee (PPDC) on ways to make information on pesticide inert ingredients more widely available. An inert ingredient is any ingredient that is not �pesticidally active� as opposed to active ingredients which prevent destroy or repel pests. EPA has charged the Group to review current OPP policy provide a forum for open discussions on disclosure principles and examine options for communicating with the public. Appointees will be announced shortly drawn from a variety of public and business organisations.In a separate measure EPA is also consulting on various science policy issues relating to implementation of the Food Quality Protection Act (FQPA) and tolerance assessment. Current consultation covers application of the �FQPA 10-Fold Safety Factor�. Related policy issues and an expected timetable are listed at the website of the Tolerance Reassessment Advisory Committee (TRAC). EPA OYce of Pesticide Programs www.epa.gov/opp; TRAC www.epa.gov/ pesticides/trac/science Sampling methodology reduces clean-up costs A new methodology for identifying sampling locations for contaminated groundwater has been developed by researchers at the University of Illinois. The researchers claim the technique can reduce the costs of long-term sampling and monitoring of contaminated sites.To ensure compliance with legislation contaminated sites may require many hundreds of wells and determining the most cost-eVective number and placement can be problematic. The Illinois methodology consists of a groundwater fate-and-transport simulation and several plumeinterpolation modules. It also includes a monitoring plan selection process based on what the research team call a �genetic algorithm�. This identifies optimal subsets of monitoring wells through a process of �natural selection�. �Results have shown that our methodology is very eVective at both reducing monitoring costs and accurately quantifying the mass of contaminant in the plume� said team leader Barbara Minsker. Expert groups for Institute The UK�s Institute for Occupational Safety & Health (IOSH) has set up two new specialist groups in response to members� requirements.The Healthcare Specialist Group supports safety and risk practitioners working in the National Health Service or private healthcare sector. The Environmental Specialist Group is the first topic-based group within IOSH. The Institute says it is recognition that boundaries between OSH and environmental management are becoming increasingly blurred. The Institute has also issued a policy paper on education calling for occupational health and safety issues to be integrated into the school curriculum. IOSH www.iosh.co.uk HSE explains research The UK Health and Safety Executive has published guidelines clarifying the business context of its research activities. In the guidelines HSE describes the priority health and safety issues requiring future research.It also outlines some of the key influences on HSE�s business needs and on its research programmes over the next few years. As such the guidelines provide a link between annual statements on strategic planning and research strategy. Guidelines for HSE�s Research Programmes. Available at www.open.gov.uk/hse/hsehome.htm 92N J. Environ. Monit. 1999 1 News Publications Environmental expenditures A report by Eurostat on provides for the first time harmonised statistics on environmental protection expenditures (EPE) by government industry and economic-oriented response indicators. The report shows that total EPE expenditure in the EU accounted for 1% of GDP or 186 euro per capita in 1995. From 1992 to 1995 EU total expenditure as a percentage of GDP increased by 13% while EU expenditure per capita increased by 36%.Total EPE grew from 53.2 billions euro in 1992 to 65.5 billions euros in 1995�an increase of 23%. The report shows that Germany�s share was the highest with 27.5% in 1992 and 33.3% in 1995 followed by France with 14.3% in 1992 and 14.1% in 1995 and The Netherlands with a share falling from 7.8% in 1993 to 6.5% in 1995. The UK had the largest share of total EPE by industry with 21.3% in 1994 followed by Germany and The Netherlands Environment Protection Expenditure (EPE) in Member States 1988�96 Eurostat cat CA-16�98�732-EN-C 144 pp. e32 Consultation on occupational exposure limits The UK Health and Safety Commission has published a consultation document on proposed changes to the approved list of occupational exposure limits (OELs).The document lists 18 changes to Maximum Exposure Limits (MELs) and Occupational Exposure Standards (OESs). If approved these changes will come into force early in 2000. HSC has also published the 1999 supplement to EH64 its reference document on occupational exposure limits containing new summaries for 8 substances. Proposals for maxiumum exposure limits occupational exposure standards and biological monitoring guidance values CD150/99. See www.open.gov.uk/hse/ condocs/index.htm. EH64 Summary criteria for occupational exposure limits ISBN 0 7176 24651 X �10. Both available through HSE Books (tel +44 (0)1787 881165) or the Stationery OYce. Pesticides Trust Environmental pressure group the Pesticides Trust has published the latest issue of its newsletter (No.44) with articles on GM crops pesticide policy in the EU reproductive eVects of pesticides in women and a report on endosulfan in Australian beef. Full contents list and subscription details available at www.gn.apc.org/ pesticidestrust/ Events Environment Japan �99 T he 16th JETRO Import Fair. 20�23 October 1999 Osaka Japan. An international trade fair to be held alongside the New Earth �99 Conference. Details from www.environmentjapan.org T he 1999 Right-to-Know Conference and Exhibition. 25�28 October 1999 Denver CO. Organised by the National Environmental Health Association. Occupational Exposure Databases and T heir Application for the Next Millennium. 1�3 November 1999 London UK. Organised by the American Conference of Governmental Industrial Hygienists. Details at www.achig.org/events/oedb99.htm Measurement for success UK National Measurement Conference �99.2�4 November 1999 Brighton UK. Incorporates BEMC �99�The 9th International Conference on Electromagnetic Measurement. Details from NMP Conference Secretariat National Physical Laboratory. e-mail nmp_sec@npl.co.uk 1999 Eastern Analytical Symposium. 14�19 November 1999 Somerset NJ United States. Details from EAS Program Committee e-mail easinfo@aol.com Fourth Workshop on Biosensors and Biological T echniques in Environmental Analysis. 1�3 December 1999 Menorca Spain. Organised by the International Association of Environmental Analytical Chemistry and the Institut Menorqu� d�Estudis. Details from dbcqam@cid.csic.es Conservation 2000. 15�17 December 1999 New Orleans LO. Conference sponsored by US EPA and US Department of Agriculture presenting accomplishments and lessons learned from local state and federal programmes for soil and water quality.T ools for Urban Water Resource Management and Protection A National Conference. 7

ISSN:0960-7919    

DOI:10.1039/a907153i

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

94N J. Environ. Monit. 1999 1 US Focus Contaminated groundwater likely to sink clean air additive The US Environmental Protection Agency plans to urge Congress to revoke laws that require a widely used controversial gasoline additive which is meant to reduce air pollution because it pollutes water. The agency is taking the action because an EPA-appointed expert panel has concluded that use of the controversial ingredient,MTBE (methyl tert-butyl ether) should be immediately �reduced substantially�.1 Several panel members recommended thatMTBE a possible human carcinogen should be phased out completely. The panel recommended the policy change because thanks to its high solubility in water MTBE is seeping into drinking water supplies across the country. When the report was released on July 27 EPA administrator Carol Browner said she would urge Congress to change rules passed in 1990 which require oil companies to add an oxygenate to gasoline.2 Oil companies can choose which oxygenate to use and most choose MTBE.Ethanol comes a distant secondplace. Other oxygenates in limited commercial use include methanol ethyl tert-butyl ether (ETBE) tert-amyl methyl ether (TAME) and diisopropyl ether (DIPE).MTBE producers are already fighting the panel�s recommendations. MTBEwas introduced in theUS in 1979 as an octane booster in unleaded gasoline. But under the Clean Air Amendments of 1990 use of the additive skyrocketed from 9.7 billion litres in 1980 to about 350 billion litres in 1997.3 This is because the act required use of oxygenates such as MTBE in areas with poor air quality.By 1995,MTBE was added to gasoline in 17 states to help meet air-quality goals.MTBE is now the largest volume end-use chemical produced in the US.4 Following the widespread introduction of the additive to gasoline many areas of the nation saw an increase in reports of symptoms such as headaches skin and eye irritation and dizziness. But a 1996 government report while not discounting these claims concluded that such problems were unlikely from exposure to common levels ofMTBE.5 But water contamination initially picked up by routine groundwater monitoring is the unexpected consequence that appears to spell the end forMTBE. The panel report estimates that in places where oxygenates are used 5 to 10% of drinking-water supplies show detectable amounts ofMTBE. The additive�s presence in drinking water sources is a problem because it makes water smell bad and it may have adverse health eVects.Even at low concentrations MTBE�s odor is so strong that most people refuse to drink it. EPA has not set a health-based standard but the agency issued a drinking-water advisory limit of 20�40 mg L-1 based on taste and odor. Based on its own taste and odor study California set an aesthetic standard of 5 mg L-1 and a 14 mg L-1 public health goal for drinking water. EPA considersMTBE to be a possible human carcinogen because it has been shown to cause cancer in animals and it is included in an agency list of contaminants that may require regulations in the future. Monitoring by the United States Geological Survey (USGS) since 1993 has built up a picture of increasing groundwater contamination.The additive has been detected in 21%of 480 wells located in areas that use MTBE to abate air pollution at a reporting level of 0.2 mg L-1. In the rest of the country MTBE detection was 2%. Gasoline aromatic hydrocarbons (benzene toluene ethylbenzene and xylenes collectively known as BTEX) were detected much less frequently 4% than MTBE in areas where the additive is used.Most detections in groundwater 97% have been less than the EPA advisory level.2 USGS uses purge and trap gas chromatography�mass spectrometry to identify the fuel oxygenates.6,7 Although USGS�sMTBE monitoring and research is still in a formative state researchers have identified several possible entryways forMTBE into the hydrologic cycle.MTBE-oxygenated gasoline contains between 10 and 15% MTBE by volume.Leaking storage tanks or pipelines can lead to very high levels of MTBE in water (at 25 �C thewater solubility of MBTE is about 5000 mg L-1 compared to the total hydrocarbon solubility of nonoxygenated gas which is about 120 mg L-1).8 These major releases have caused the loss of some community water supplies.MTBE may also aVect public water supplies by its introduction into reservoirs and lakes from the exhaust and noncombusted gasoline losses of recreational watercraft such as jet skis. Emissions and evaporative losses from automobiles also releaseMTBE to the atmosphere and land surfaces where it can move with water in the hydrologic cycle. OnceMTBE is in the environment several characteristics make it an environmental problem.MTBE sorbs only weakly to soil and aquifer materials so that the additive moves along with the groundwater.Once in groundwater MTBE resists biological degradation. It also moves quickly out of the atmosphere and into the water cycle. In the atmosphere the half-life ofMTBE can be as short as three days in a regional airshed. The panel�s recommendation puts EPA in synch with eVorts by many states. Last October Maine received EPA approval to stop the mandatory sale of reformulated gasoline in exchange developing an alternative plan for meeting air quality standards. InMarch California ordered oil companies to phase outMTBE by 2002. New Jersey recently won permission from EPA to stop using extra MTBE in gasoline in winter because it has solved its winter carbon monoxide problem and New Hampshire in July asked to be exempted from the program.At this point it remains unclear how air quality goals will be accomplished without MTBE. Other ethers (ETBE TAME and DIPE) have similar chemical properties and are thought to have similar drawbacks although the panel recommends further research. Ethanol which comes from corn biodegrades rapidly but production is currently far too small to replaceMTBE. Cleanerburning low-sulfur fuels may also take on more prominence. J. Environ. Monit. 1999 1 95N US Focus Notes 1 United States Environmental Protection Agency 1999. Blue Ribbon Panel on Oxygenates in Gasoline Executive Summary and Recommendations. 27 July 1999. http://www.epa.gov/oms/consumer/fuels/ oxypanel/rec721.pdf 2 Carol Browner press release. http://www.epa.gov/oms/consumer/fuels/ oxypanel/browner.pdf 3 Zogorski J.S. et al. 1998. MTBE Summary of Findings and Research by the U.S. Geological Survey Proceedings of the 1998 Annual Conference of the American Water Works Association June 21�25 1998 Dallas Texas. 287�309 AmericanWater Works Association Denver CO. 4 MTBE in the Nation�s GroundWater NationalWater Quality Assessment (NAWQA) Program Results Squillace P.J. United States Geological Survey April 29 1999. http://wwwsd.cr.usgs.gov/nawqa/vocns/ brp-pjs-handout.html 5 Interagency Assessment of Oxygenated Fuels�Final Report OYce of Science and Technology Policy (OSTP) Executive OYce of the President. 6 US Geological Survey LaboratoryMethod forMethyl tert-Butyl Ether and Other Fuel Oxygenates October 1995. US Geological Survey Fact Sheet FS-219�95. 7 LaboratoryMethod for Analysis of Small Concentrations of Methyl tert-Butyl Ether and Other Ether Gasoline Oxygenates in Water. September 1998. USGS Fact Sheet 086�98. 8 Squillance P. J. et al. 1998. Environmental Behavior and Fate ofMethyl tert-Butyl Ether (MTBE) United States Gological Survey Fact Sheet FS-203�96. Rebecca Renner Science writer and editor based in the US Tel+1 570 321 8640; Fax+1 570 321

ISSN:0960-7919    

DOI:10.1039/a907154g

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

Personal ammonia sensor for industrial environments C. Malins,*a A. Doyle,b B. D. MacCraith,b F. Kvasnik,a M. Landl,c P. S¢§ imon,c L. Kalvoda,d R. Lukas¢§,d K. Puflere and I. Babus©¥¢¥kf aDepartment of Instrumentation and Analytical Science, UMIST, Manchester, UK M60 1QD. E-mail: chris.malins@umist.ac.uk; Fax:+44.161.2004911; Tel:+44.161.2008939 bOptical Sensors Group, Dublin City University, Glasnevin, Dublin 9, Ireland cDepartment of Physical Chemistry, Slovak Technical University, Radlinske¢¥ho 9, 812 37 Bratislava, Slovak Republic dLUKAS¢§ Industrial Research, Doubrav©¥¢¥nova 1/219, Prague, 163 00, Czech Republic eCKD Technical Laboratories, Na Harfe 7, Prague, 190 02, Czech Republic fSlovak Hydrometerological Institute, Jeseniova 17, Bratislava, 833 15, Slovak Republic Received 17th June 1999, Accepted 29th July 1999 The realisation of an opto-chemical ammonia sensor suitable for personal monitoring tasks is described, comprising a cyanine dye immobilised in a microporous glass thin film.The fabrication of sensor platforms incorporating embossed grating couplers provides a compact optical design with eVective waveguiding characteristics, resulting in reversible ammonia sensitivity in the 5.100 ppm range in under 2 min.Cross-sensitivity of sensor response with water and other potential interferents is considered. compact ammonia sensor device with low power demand that 1. Introduction could be used for in situ real-time analysis. Ammonia is widely used in the production of explosives, A variety of optrode (optical electrode) devices have been fertilisers and as an industrial coolant.The toxic qualities of described in the literature that utilise the reaction of dissolved this gas are well documented, and acute poisoning can result ammonia vapour with a pH-dependent dye material, which from inhalation of only small doses of ammonia vapour. undergoes a suitable colour change.7.19 In general, these are Exposure limits of 25 ppm over an 8 h period and 35 ppm based on monitoring the absorption or fluorescence characterover a 10 min period have been recommended and have istics of indicator dyes entrapped within a membrane deposited recently been legislated for.1 Locations employing industrially onto a waveguiding substrate. Ammonia interacts with the sized cooling systems, such as food production and storage immobilised indicator, resulting in changes in absorbance or plants, report a number of exposure incidents every year.As emission spectra, which are monitored using a suitable detector industry is becoming increasingly safety conscious, eYcient module via an optical fibre or planar waveguide. Evanescent sensor devices to monitor personal exposure of workers under- wave absorption is an eVective technique for performing such taking any risk of contact with ammonia are desirable, as analysis.20 Light energy associated with the guided mode rapid evacuation of personnel from contaminated areas may penetrates into a coating of lower refractive index than the be all that is required to prevent serious illness.substrate.This evanescent field is able to interact with dye In the field of air quality monitoring, a wide variety of materials contained within the coating, and has been discussed instrumentation can be employed for ammonia analysis. previously in terms of application to absorption based detec- Devices employing spectroscopic2 or electrochemical3,4 tion of ammonia vapour.21,22 methods are usually very accurate, sensitive and selective. The sol.gel process is an ambient temperature method that However, they are also expensive, static instruments, requiring may be used to fabricate thin glass films that have well-defined the presence of an experienced operator.Simpler detector optical and physical characteristics, such as refractive index. systems based on the semiconductivity of SnO2 thin films5 do Details of the influence of varying process parameters on the exist for monitoring ammonia leakage in the working environ- properties of the resultant silica films are described elsement, and have been commercialised as personal monitoring where,23,24 but, in general, factors such as the pH and ratio systems.However, these devices come at a much higher cost, of water to silicone alkoxide precursor can be used to define are not selective towards ammonia and display a restricted the nature of the glass.Analyte-sensitive reagents may be active sensor lifetime. added at the precursor stage, and will subsequently become Only recently has a miniaturised electrochemical sensor entrapped in the porous coating, to yield dye-doped sensor device become available for ammonia detection.6 This device films.uses interchangeable sensor heads biased towards the detection An array of cyanine, triphenylmethane and phthalate dyeof a specific gas, with the added requirement of peripheral stuVs were prepared during the course of this research, and filters to achieve eVective selectivity towards ammonia. A the properties of some of these materials have been described diVerent approach can be applied when considering a dedicated elsewhere.25,26 The aim of the synthesis was to produce an ammonia sensor, which similarly takes advantage of miniat- ammonia-sensitive dye material that was able to withstand the urised, low power components, with the potential for rapid, low pH environment of the sol.gel route for microporous selective response to ammonia in the required range.Opto- glass production. This robust and chemically stable support material can be used to fabricate dye-doped thin films of well- chemical transduction is a way to develop a non-consumptive, J. Environ. Monit., 1999, 1, 417.422 417defined optical characteristics, which can then be examined which separated on cooling, was repeatedly recrystallised (acetone, acetic acid) to give green crystals.using the evanescent absorption interrogation technique. In designing a personal monitoring device, emphasis has to 2.2. Optical fibre preparation be placed on producing a compact configuration with a simple and eYcient opto-electronic design.We have recently described A plastic-clad silica multi-mode optical fibre of core diameter a planar waveguide approach to evanescent absorbance moni- 600 mm (Ceramoptec, Bonn, Germany) was cut into 12 cm toring.27,28 This technique involves a surface relief grating lengths and both ends were polished (PM2A polishing rig, being imprinted upon the surface of a planar waveguide. 9–0.1 mm, Al2O3 suspensions, Logitech, Aberdeen, UK).A EYcient in- and out-coupling of interrogating light is achieved 5 cm portion of the outer coating and cladding was removed using these embossed gratings, and as the ammonia sensitive by mechanical and chemical means, respectively, and the fibres dyestuV is also immobilised on the surface of the waveguide, were cleaned (acetone, methanol and water) before dye-doped the detection of ammonia is facilitated by monitoring the out- sol–gel glass coatings were fabricated along the length of the coupled light.Along the length of the waveguide the evanescent declad fibre. A small amount of epoxy resin was placed on fields of the guided modes interrogate the absorption spectrum the end of the fibre to prevent coverage during the dip coating of the ammonia-sensitive reagent.Any variation in the evanprocedure. escent absorption of the sensing layer in the bandwidth covered by the light source results in a change in the intensity of out- 2.3. Grating coupler fabrication coupled light. After upwards of 15 years of research and development, Titanium tetrabutoxide (2.0 g) was added dropwise to a stirred solution of methyltriethoxysilane (3.6 g) in ethanol (6.6 g).optical chemical sensors are approaching maturity. Commercial devices are currently available for tasks such as Silicon tetrachloride (0.6 g) was then added and the resultant solution was stirred for 2 h before being left overnight prior dissolved oxygen and carbon dioxide detection, as well as determination of pH, for water quality and pollution monitor- to use.This high refractive index sol30 was then dip coated onto the ends of a glass slide. Immediately following depos- ing. However, no opto-chemical ammonia sensing device is currently available that could be described as a personal ition, an aluminium-coated surface relief master grating (1200 lines mm-1, Ealing Electro-Optics, Watford, UK) was pressed exposure monitor for use in the workplace.Here we outline a prototype ammonia monitoring device for this task that against the sol–gel layer for 1 min, using a pressure embossing system. The resultant embossed grating couplers31–35 were then employs all solid-state components, is highly compact and has a low power demand. left to dry at room temperature for 3 d before dye-doped ammonia sensitive films were coated on the opposing face of the waveguide. 2. Experimental 2.4. Sol–gel film fabrication All chemicals used during dye synthesis, glass preparation and waveguide fabrication were obtained from Sigma-Aldrich To a stirred solution of the dye (10 mg) and aqueous HCl (Gillingham, UK) unless otherwise stated. (pH 1, 1.6 g) in ethanol (3.0 g), tetraethoxysilane (4.6 g) was added.Stirring was continued for 1 h, and the sol was left at 2.1. Synthesis of 1,3,3-trimethyl-2-methyleneroom temperature overnight, after which films of an approxi- [4-{1,3,3-trimethyl-2(1H,3H)indolylidene-1-propenyl}-5-oxo- mate thickness of 400 nm were prepared either on optical 2-furfurylidene]-3(H)indolinium perchlorate25 (Fig. 1) fibres or on the reverse side of grating-embossed waveguides, using a dip coating procedure.36–39 Coated samples were dried The Vilsmeier-Haack reagent was prepared by the dropwise overnight (70 °C) and stored in ambient conditions away from addition of POCl3 (40.12 g, 0.26 mol) to a stirred, cooled direct sunlight for 1 month prior to examination.This storage solution (0–5 °C) of dimethylformamide (19.13 g, 0.26 mol), period ensures that post-fabrication structural evolution of the after which the mixture was left to stand at room temperature silica matrix is complete.24 for a further 20 min. 2(5H)-Furanone (10 g, 0.12 mol) was then added with further cooling (0–5 °C), over a period 2.5. Optical fibre characterisation system26 of 30 min, after which the resultant mixture was heated (60–70 °C) for 3 h. The cooled residue was then dissolved in The sensing characteristics of the coated fibres were evaluated ethanol (50 ml ) and treated with perchloric acid (11.95 g, in a gas flow cell of volume 10 ml mounted on a positioning 0.12 mol).The precipitate was filtered oV and recrystallised stage. Light from a 50W tungsten–halogen source was (glacial acetic acid) to yield yellow needles of 4-(dimethyl- launched via a microscope objective (numerical aperture, 0.4) aminomethylene)-5-oxo-2-furfurylidenedimethyliminium fitted with an annular beam mask20 to block lower order light perchlorate.29 modes, thereby improving the evanescent wave absorption of A mixture of 1,2,3,3-tetramethyl(3H)indolium iodide the dye-coated fibre.40 The light transmitted by the sensing (2.05 g, 6.8 mmol), 4-(dimethylaminomethylene)-5-oxo-2-fur- fibre was collected by a 400 mm silica fibre and launched into furylidenedimethyliminium perchlorate (1.0 g, 3.4 mmol) and a miniature CCD array spectrometer (Ocean Optics S2000, anhydrous potassium acetate (0.66 g, 6.8 mol ) in acetic anhy- Dunedin, FL, USA).dride (20 ml ) was heated under reflux for 1 h. The product, Gas flow was regulated by the use of dedicated mass flow controllers (Unit Instruments, Dublin, Ireland).Mixtures of nitrogen and 1% ammonia in nitrogen (BOC, Guildford, Surrey, UK) were employed, such that the total flow rate of the gas mixture was maintained at 500 ml min-1 at all times. To achieve changes in relative humidity, the carrier gas was split into two streams, one of which was bubbled through water.Relative humidity was measured (TH210 Relative Humidity Meter, Eirelec, Dundalk, Ireland) after the two components were recombined and mixed with 1% ammonia Fig. 1 Structure of 1,3,3-trimethyl-2-methylene-[4-{1,3,3-trimethylprior to delivery to the flow cell. All gas mixtures used were 2(1H,3H)indolylidene-1-propenyl}-5-oxo-2-furfurylidene]-3(H)indolinium perchlorate.of 0% relative humidity unless otherwise stated. 418 J. Environ. Monit., 1999, 1, 417–4222.6. Grating embossed planar waveguide characterisation The apparatus used to characterise the ammonia sensitivity of grating-embossed planar waveguides is shown in Fig. 2. A photodiode with dedicated lock-in circuitry41 was used to measure the intensity of light from a green LED (Toshiba TLPGA159P, lmax=562 nm, Radionics, Dublin, Ireland) coupled into and out of the waveguide.A small flow cell was attached to the waveguide such that only the dye-doped layer was exposed to the analyte gas. Gas flow was regulated in the manner described in Section 2.5. 2.7. Cross-sensitivity studies Reflectance spectra of glass slides dip coated with dye-doped glass thin films were measured using a microspectrometer Fig. 3 Variation of absorption with increasing pH for ethanolic dye solution. (Ocean Optics S1000, Dunedin, FL, USA) during exposure to 100 ppb ozone in air, produced by an arc generator, at a carrier gas flow rate of 1 l min-1. Carbon dioxide and nitrogen glass thin films, where acidic preparative conditions are dioxide exposure studies were undertaken using a similar required.Furthermore, diVerential scanning calorimetry system. NO2 headspace was created by reacting nitric acid measurements indicate that the dye is stable in air up to with copper wire. Cross-sensitivity to other possible inter- temperatures of 150 °C, and therefore would not be adversely ferents (butane, hexane, methane, acetylene, ethene, diethyl aVected by the drying stages involved in glass film production. ether, hydrogen, toluene and nitrogen monoxide) was exam- The mechanism of the interaction of this dye with ammonia ined by monitoring the evanescent wave absorption spectra of in solution involves the lone electron pair of ammonia attackdye- doped glass-coated fibres using the system detailed in ing the positive charge delocalised at the p-electron system of Section 2.5.the dye. Formation of a bond between ammonia and the carbon skeleton breaks this conjugation, and is manifest by changes in the absorption spectrum. This also means that 3. Results and discussion other electron donors, such as the hydroxyl ion or amines, 3.1. Dye behaviour in solution can also interact with the dye in the same manner. 1,3,3-Trimethyl-2-methylene-[4-{1,3,3-trimethyl-2(1H,3H)ind- 3.2. Sol–gel immobilisation of the dye reagent olylidene-1-propenyl}-5-oxo-2-furfurylidene]-3(H)indolinium perchlorate displays absorption bands in ethanolic solution at Initial characterisation of the ammonia-sensitive dye was wavelengths of 570 and 730 nm. Reversible diminution of undertaken by immobilisation onto optical fibre substrates.these peaks, alongside the appearance of an absorption band Spectral variations across the whole visible region can then at 500 nm, occurs on exposure to ammonia. The original be assessed which provide important information, dictating absorption spectrum of the dye can be recovered by adding factors such as light source selection for the planar waveguide hydrochloric acid.This behaviour demonstrates that the dye assembly. can be used as a pH indicator in the basic region, as large Due to the polarity of the dye, a hydrophilic support matrix spectral variations are observed between pH 8 and pH 12, but was required to achieve eVective dye encapsulation. Sol–gel little change is noted below pH 7. The pH dependence of the derived glass films were fabricated using tetraethoxysilane dye spectrum in water–ethanol solution (151 v/v) is shown in (TEOS) as the glass precursor.The pore surface of this type Fig. 3. The stability of the dye in an environment of pH 2 is of membrane is populated by silanol (Si-OH) groups, which also illustrated here. Such robust behaviour was rarely encoun- are able to form hydrogen bonds and electrostatic interactions tered with the range of dyes prepared during the course of with polar species.However, the conditions employed to this work and rapid degradation of other analogous dyestuVs perform acid hydrolysis of the glass precursor required to in aqueous solutions of below pH 4 was regularly encountered. form the gel are often suYcient to adversely aVect the molecu- The stability of this material to low pH environments makes lar stability of the dye to be immobilised within the final glass.it an ideal candidate for incorporation into sol–gel derived In this case, we found that the dye was not only stable in the low pH environment required to fabricate the thin glass films, but was also readily compatible with this type of support matrix.In Fig. 4 we can see the evanescent absorption spectrum of a dye-doped TEOS-coated optical fibre. There are two large absorbance bands centred at 535 and 700 nm, and the increase of these peaks on exposure to ammonia vapour is highlighted. In comparison with the spectra obtained for the dye in solution, we can see that the absorption maxima are blue shifted by 30–40 nm, possibly as a result of dye–silanol interactions.It is also evident that the behaviour of the immobilised dye is remarkably diVerent to that shown when in solution. Exposure to ammonia now results in an increase in absorption, with no evidence of the evolution of the peak associated with the unconjugated dye, found at 500 nm in solution. This suggests that the glass surface also interacts strongly with the p-system in the dye, in a manner that dramatically aVects the mechanism of the dye reaction with ammonia, such that it is Fig. 2 Grating-embossed planar waveguide sensor device. no longer analogous with that occurring in solution. J. Environ. Monit., 1999, 1, 417–422 419displayed using fibre substrates, the average values of T90 being 60 s and 200 s, respectively.The normalised sensor response to varying concentrations of ammonia is shown in Fig. 6, from which a limit of detection of 5 ppm was calculated in the manner described previously. The line of best fit is a first-order function, with a correlation coeYcient of 0.988. This behaviour is also consistent with the notion of complex dye–glass interactions existing within the sensor membrane, as the classical theory for the type of indicator response is no longer adhered to.14 Although the signal variations found on exposure to ammonia using this method are smaller than with the optical fibre configuration due to the reduced path length, we do, however, see good reproducibility and signal stability.Further Fig. 4 Evanescent absorption spectrum of dye-doped TEOS coated advantages over the fibre system are evident when considering fibre on exposure to 100 ppm ammonia in nitrogen.the size of the optical components. The laboratory-scale prototype instrument is all solid state, and can be battery powered. For the immobilised dye, the most significant spectral region This configuration would be ideally suited to on-site monitor- existed at 535 nm, and the absorbance at this wavelength was ing tasks.However, other factors that may influence the in sequentially monitored during exposure to ammonia vapour situ performance of this sensor have to be examined. under a dry gas atmosphere, as shown in Fig. 5. We can see that there is significant sensitivity in the desired range 3.4. Evaluation of cross-sensitivity of ammonia concentration, and repeated exposure to this ammonia–nitrogen cycle allows the construction of a cali- Moisture sensitivity is a ubiquitous concern when evaluating bration plot of absorbance against ammonia concentration for optical chemical sensors.Variations of ambient or membrane a single coated fibre. A limit of detection42 of 3 ppm ammonia humidity will always result in spectral changes of varying can be calculated (that ammonia concentration equivalent to magnitude.Indeed, dye-doped thin films have been reported the absorption under nitrogen minus 3s of the data set). The as sensors devoted solely to humidity detection,43 and we times taken to achieve 90% of the average total signal change would expect to see some moisture reliance with our dye– (T90) were found to be 70 and 140 s for response and recovery, matrix combination. Therefore, the eVect of the relative respectively.humidity of the gas mixture on the sensitivity of the dye immobilised in TEOS-derived glass was investigated. The 3.3. Grating-embossed planar waveguide sensor nitrogen carrier gas was humidified by bubbling through water before delivery to the flow cell, and the amount of moisture Interrogation of the evanescent absorption spectrum of the in the final gas mixture was regulated by mixing humid and dye-doped film on optical fibres requires bulky optical compodry nitrogen gas streams. nents, such as micropositioning stages, to ensure eYcient in- The evanescent absorption of the immobilised dye increases and out-coupling of light to the optical fibre substrate.systematically with the relative humidity of the carrier gas as Consequently, whilst an ideal laboratory-based characterisexpected, and this also has a pronounced eVect on the ammonia ation system, this arrangement is unsuitable for a miniature sensitivity of the thin film. The absorption variation at 535 nm personal monitoring device.Integration of eYcient lighton exposure to 50 ppm ammonia was examined on a coated coupling elements onto the sensor substrate itself oVers an optical fibre, for a range of carrier gas humidity, and is alternative to such bulky optical components. displayed in Fig. 7. Here we see that the sensitivity towards The grating-coupled configuration detailed in Section 2.6 ammonia is gradually reduced on humidification of the carrier and Fig. 2 was employed using a dye-doped TEOS layer for gas, being retarded by up to 50% for a carrier gas of 100% ammonia sensitivity. The immobilised dye absorption band relative humidity. However, it was found that the response centred at 535 nm, displayed in Fig. 4, overlaps with the times remained unchanged during humidity variation.emission from a green LED, and this illumination can be In terms of equating this humidity sensitivity to that found launched into the waveguide. The transmission signal of light towards ammonia, on going from 0 to 100% humidity there out-coupled from the waveguide was then monitored on is an increase of over 0.06 absorption units. Compared to the repeated exposure to ammonia in nitrogen.An increase in dye change in evanescent absorption on exposure to 100 ppm absorption in the bandwidth of the LED on exposure to ammonia (~0.02), this does indeed appear to be significant. ammonia results in a decrease of out-coupled light from the waveguide. Response and recovery times are similar to those Fig. 6 Calibration plot for the ammonia response of the dye-doped Fig. 5 Variation in evanescent absorption at 535 nm at various ammonia concentrations of dye-doped TEOS-coated fibre. TEOS-coated planar waveguide sensor. 420 J. Environ. Monit., 1999, 1, 417–422If we consider the target application for this personal monitoring device (within refrigeration plants), the temperature profile would be relatively unchanged at close to, or below, freezing point.Under such conditions, the vapour pressure of any volatile amines resultant from the deterioration of foodstuVs would be negligible, as would the concentrations of the reactive gases mentioned above. Taking into account all of these factors, we can look forward to the prospect of this personal monitoring device for ammonia exposure being optimised further with a view to undergoing field trials. 4. Conclusions We have described the characterisation of a prototype device Fig. 7 Variation of absorption change of dye-doped TEOS-coated optical fibre encountered upon exposure to 50 ppm ammonia for to be used for the personalised monitoring of ammonia vapour. carrier gas of increasing relative humidity. An opto-chemical sensor approach employs integrated optics and compact illumination and detection components that can be driven by a battery source.The dye, when immobilised in a sol–gel glass thin film, shows sensitivity in the required However, normal ambient conditions usually range between range, with a limit of detection of 5 ppm. Response is in terms 50 and 80% humidity, and a variation of this extent would of seconds, and is fully reversible.Cross-sensitivity to a range equate to the signal change encountered on exposure to of other pollutants is minimal, as is dye photo-bleaching. We 100 ppm ammonia. It is therefore vitally important to consider have detailed the realisation of a prototype personalised moni- the conditions of humidity when interpreting the data supplied toring device for ammonia sensing employing an eVective by the instrument.Adequate provision has to be made in combination of substrate design and dye–matrix selection. terms of calibrating for atmospheric humidity to give meaningful values of ammonia concentration. Fortunately, relative Acknowledgements humidity measuring devices are both compact and inexpensive, and information from an integrated humidity sensor could be We would like to thank the European Commission for funding readily incorporated to provide an adequate calibration this research (contract number ERBIC15CT960819).system. The prospect of realising a personal monitoring device References brings with it the necessity to consider a wider range of contaminants that might influence the behaviour of the dye- 1 Health and Safety Executive, Guidance Note EH40/93 doped films; therefore, the sensitivity of the immobilised dye Occupational Exposure Limits 1993, HMSO, London, 1993. 2 A. Verdin, Gas Analysis Instrumentation, MacMillan, London, towards some common atmospheric pollutants was examined 1973. in two diVerent ways. 3 F. Bailescu and V. V. Cosofret, Applications of Ion-Selective Firstly, the reflectance spectra of dye-doped sol–gel thin Membrane Electrodes in Organic Analysis, Ellis Horwood, films on planar substrates were examined on exposure to Chichester, 1978. ozone, Cl2, NO2 and carbon dioxide in air, in the absence of 4 W.E. Morf, in The Principles of Ion-Selective Electrodes and daylight. It was found that, while the immobilised dye was Membrane Transport, ed. E. Pungor and S.W. Incsedy, Elsevier, Amsterdam, 1981, pp. 402–406. insensitive to 100% CO2, it was irreversibly aVected by high 5 T. Seiyama, A. Kato, K. Fujiishi and M. Nagatani, Anal. Chem., concentrations of Cl2 and NO2 and by 100 ppb ozone. It is 1962, 34, 1502. known that these reactive species have great aYnity for 6 http://www.draeger.com conjugated p-electron systems,44 and here this is manifest in 7 J.F.Giuliani, H.Wohltjen andN. L. Jarvis, Opt. Lett., 1983, 8, 54. photo-bleaching of the dye. In the field of optical chemical 8 M. A. Arnold and T. J. Ostler, Anal. Chem., 1986, 58, 1137. 9 P.C� aglar and R. Narayanaswamy, Analyst, 1987, 112, 1285. sensors, this is a common problem which ultimately limits the 10 M. R. Shahriara, Q. Zhou and G. H. Sigel, Opt.Lett., 1988, lifetime of the sensor substrate. However, we found that dye- 13, 407. doped TEOS films were essentially photo-stable under normal 11 Q. Zhou, D. Kritz, L. Bonnell and G. H. Sigel, Appl. Opt., 1989, laboratory conditions, even in direct sunlight. Indeed, as the 28, 2022. photo-stability of the dye in solution was found to be poor, 12 W. Sellien, R. Czolk, J. Reichert and H.J. Ache, Anal. Chim. we must again speculate that the dye–glass interactions have Acta, 1992, 269, 83. 13 R. Klein and E. Voges, Sens. Actuators B, 1993, 11, 221. a beneficial eVect on the overall sensor performance, whilst at 14 A. Mills, L. Wild and Q. Chang, Mikrochim. Acta, 1995, 121, 225. the same time making mechanistic interpretation very complex. 15 T. Werner, I. Klimant and O.S. Wolfbeis, Analyst, 1995, 120, Further tests were also performed on dye-doped TEOS- 1627. coated optical fibre substrates with other species often found 16 A. A. Vaughan, M. G. Baron and R. Narayanaswamy, Anal. in industrial environments. Negligible changes in evanescent Commun., 1996, 33, 393. absorption spectra were found on exposure to large concen- 17 R. A. Potyrailo, L.A. Mikheenko, P. S. Borsuk, S. P. Golubkov and P. M. Talanchuk, Sens. Actuators B, 1994, 21, 65. trations (thousands of ppm) of methane, hexane, diethyl ether, 18 Y. Sadaoka, Y. Sakai and M. Yamada, J. Mater. Chem., 1993, hydrogen, acetylene, toluene, ethene, butane and nitrogen 3, 877. monoxide vapours. Long-term stability tests have also been 19 A. Brandenburg, R. Edelhauser, T.Weiner, H. He and O. S. undertaken which found that a decrease in sensitivity of only Wolfbeis, Mikrochim. Acta, 1995, 121, 95. 25% of the dye-doped TEOS-coated fibres was encountered 20 O. S. Wolfbeis, Fibre Optical Chemical Sensors and Biosensors after 10 weeks of exposure to ambient laboratory conditions. Vol. I & II, CRC Press, Boca Raton, FL, 1991, p. 43. 21 T.M. Butler, B. D. MacCraith and C. M. McDonagh, Proc. In combination, these studies suggest that the ammonia- SPIE, 1995, 2508, 168. sensitive films described here are best employed in environ- 22 B. D. MacCraith, Sens. Actuators B, 1993, 11, 29. ments free of basic species, such as amines, or highly reactive 23 C. J. Brinker and G. W. Scherer, Sol–Gel Science, Academic Press, gases, such as Cl2, NO2 or ozone, but can operate successfully San Diego, 1990.for a number of weeks with appropriate account taken of 24 C. McDonagh, F. Sheridan, T. Butler and B. D. MacCraith, J. Non-Cryst. Sol., 1996, 194, 72. moisture variations in the ambient environment. J. Environ. Monit., 1999, 1, 417–422 42125 M. Landl, P. S¡ imon and F. Kvasnik, Sens. Actuators B, 1998, 37 L. E. Scriven, Better Ceramics Through Chemistry III, Materials 51, 114. Research Society, Pittsburgh, 1988, pp. 717–729. 26 C. Malins, M. Landl, P. S¡ imon and B. D. MacCraith, Sens. 38 I. Strawbridge and P. F. James, J. Non-Cryst. Sol., 1986, 86, 381. Actuators B, 1998, 51, 359. 39 C. J. Brinker, A. J. Hurd, P. R. Schunk, G. C. Frye and C. S. 27 A. Doyle and B. D. MacCraith, Proc. SPIE, 1997, 3105, 61. Ashley, J. Non-Cryst. Sol., 1992, 147–148, 424. 28 A. Doyle and B. D. MacCraith, Proc. SPIE, 1998, 3540, 136. 40 V. Ruddy, B. D. MacCraith and J. A. Murphy, J. Appl. Phys., 29 T. Gracza, M. Landl and D. Ve� gh, Chem. Papers, 1994, 48, 410. 1990, 67, 6070. 30 I. Yang, S. Saavedra, N. R. Armstrong and J. Hayes, Anal. Chem., 41 P. C. Hauser and S. S. Tan, Analyst, 1993, 118, 991. 1994, 66, 1254. 42 J. N. Miller and J. C. Miller, Statistics for Analytical Chemistry, 31 W. Lukosz and K. Tiefenthaler, Opt. Lett., 1983, 8, 537. Ellis Horwood, Chichester, 1988, p. 115. 32 K. Heuberger and W. Lukosz, Appl. Opt., 1986, 25, 1499. 43 P. J. Skrdla, S. S. Saavedra, N. R. Armstrong, S. B. Mendes and 33 R. L. Roncone, L. A. Weller-Brophy, L. Weisenbach and B. N. Peyghambarian, Anal. Chem., 1999, 71, 1332. J. J. Zelinski, J. Non-Cryst. Sol., 1991, 128, 111. 44 Q. Zhou and R. D. Gould, Thin Solid Films, 1998, 317, 436. 34 B. L. Ramos, N. F. Fell and S. J. Chouquette, Anal. Chem., 1996, 68, 1245. 35 L. Li and M. C. Gupta, Appl. Opt., 1990, 29, 5320. 36 I. M. Thomas, in Sol–Gel Optics: Processing and Applications, ed. Paper 9/04846D L. C. Klein, Kluwer, London, 1994, pp. 141–158. 422 J. Environ. Monit., 1999, 1, 417&ndash

ISSN:0960-7919    

DOI:10.1039/a904846d

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

Development of a personal monitoring method for nitrogen dioxide and sulfur dioxide with Sep–Pak C18 cartridge sampling and ion chromatographic determination YunWang, Andrew Allen, David Mark and Roy M. Harrison* Institute of Public and Environmental Health, University of Birmingham, Birmingham, UK B15 2TT Received 1st July 1999, Accepted 18th August 1999 Personal monitoring methods for the determination of hourly integrated concentrations of NO2 and SO2 in ambient air have been developed.Triethanolamine (TEA)-impregnated C18 Sep-Pak cartridges were used to collect NO2 and SO2 simultaneously. After sampling, NO2 and SO2 as their nitrite, nitrate, sulfite and sulfate analogues were stripped from the cartridges with a solution of 5% methanol in distilled, deionized water (DDW) and then determined by ion chromatography.Laboratory tests were conducted to evaluate the sampling rate, collection and recovery eYciencies, breakthrough volumes, absorption capacity, interference and sample stability on the cartridge during storage. NO2 and SO2 detection limits of 0.3 and 0.4 ppb respectively for 1 h samples were obtained. Recoveries for both NO2 and SO2 exceeded 85%.commercially prepared Sep-Pak C18 and silica cartridges were 1. Introduction obtained from Waters Ltd (Watford, UK) with 1.0 g weight, Nitrogen dioxide and sulfur dioxide are major air pollutants 0.35 g net weight, 0.9 cm net length and 1.0 cm diameter. arising mainly from combustion processes, such as the combustion of fossil fuels, oil and coal, which in the UK are used 2.2.Apparatus primarily by motor transport and non-nuclear power stations. Personal exposure monitoring is used to obtain an accurate A Dionex DX 500 chromatography system (Dionex, Sunnyvale, CA, USA), consisting of a CD20 conductivity estimate of the subject’s true exposure to these pollutants.1 For this purpose and to assist in modelling studies of pollutant detector, LC20 chromatography enclosure, UI20 universal interface, AS40 automated sampler and GP40 gradient pump, exposure, it is necessary to obtain at least hourly resolved measurements of personal exposure to NO2 and SO2 through equipped with an AS4A-SC analytical column, was employed.A Scintrex Model LMA-3 NO2 analyser (Scintrex Ltd., the day. Reported methods for personal monitoring of NO2 and SO2 involve the use of a triethanolamine (TEA)-impreg- Ontario, Canada), an API Model 100 fluorescence SO2 analyser (Advanced Pollution Instrumentation Ltd., CA, USA) nated molecular sieve as a solid sorbent,2,3 TEA-impregnated glass fibre filters4 and a TEA-impregnated passive diVusive and an Environnement (France) chemiluminescent NO–NO2–NOx analyser were used to monitor the concen- sampler.5–7 TEA is a compound widely used for trapping of NO2 and SO2 due to its high collection eYciency.5 Ion trations of NO2, NO and SO2 in the test gases.A Model 1923 gas blender (Analysis Automation Ltd., Oxford, UK) was chromatography (IC) is capable of the highly sensitive determination of nitrite, nitrate, sulfite and sulfate,2,5 derived from used for the preparation of diVerent known levels of the test gases.A Vortex Hi Flow personal air-sampling pump (Casella collected NO2 and SO2. The disadvantage of most published methods is that they can only be used at low flow rates, hence Ltd, Bedford, UK) was used for field sampling. requiring long sampling periods. We report a procedure for the simultaneous sampling of NO2 and SO2 at a high flow 2.3.Preparation of the sampling cartridges and filters rate over short sampling intervals to obtain hourly resolved Laboratory self-prepared cartridges were filled with glass beads measurements of low atmospheric concentrations of these or charcoal using diVerent lengths and diameters of PTFE gases. The methodology described is simple, reliable and light tubes with glass wool inserted at both ends to secure the in weight for convenience of use.sorbent. Self-prepared and Sep-Pak cartridges were loaded with 2 ml of TEA solution using a multiple port vacuum 2. Experimental section device (Supelco Ltd, Poole, UK), and excess TEA solution was drained out. After loading, clean dry air (compressed air 2.1. Reagents and materials passed through a charcoal and molecular sieve scrubber) was flushed through the cartridge to further dry the sorbent at The chemical reagents were of analytical or special grade from Aldrich (Gillingham, UK) and Merck (Poole, UK).The TEA 4 l min-1 for 2 min, and then the TEA impregnated cartridges were stored in a 500 ml glass bottle with a tight screw top at solution used in the experiments performed in this study was prepared by dissolving TEA (25 g) and ethylene glycol (4 g) 4 °C before use.The Sep-Pak cartridge can be reused. The used Sep-Pak in acetone (25 ml ) and diluting to 100 ml with distilled deionized water (DDW). Standard gases, nitrogen dioxide, nitrogen cartridge was cleaned by flushing through 15 ml of DDW and 2 ml of acetone, and then passing through dry clean air at a oxide and sulfur dioxide, 45, 50 and 100 ppm respectively in nitrogen, were obtained from BOC (Guildford, UK).Both flow rate of 4 l min-1 for drying for 2 min. The cleaned Sep- J. Environ. Monit., 1999, 1, 423–426 423Pak cartridge was loaded with 2 ml of TEA solution under was calculated from the total amounts of NO2- and NO3- ions, and that of SO2 from the total amounts of SO32- and vacuum, and excess TEA solution was drained out as above.A comparison of new and reused Sep-Pak C18 cartridges was SO42- ions detected. performed. The results showed that there was no significant diVerence between the new TEA impregnated Sep-Pak C18 3. Results and discussion cartridge and the reused TEA impregnated Sep-Pak C18 cartridge for sampling NO2 and SO2. 3.1. Preliminary experiments Filters were prepared in 60 ml polypropylene wide-neck Preliminary experiments were carried out to select an eYcient plastic bottles by rinsing four times with 40 ml of DDW. The sampling and analytical procedure for the determination of pre-washed filters were covered with 40 ml of TEA solution. NO2 and SO2 in air. Four diVerent techniques for collecting After soaking for 5 min the TEA solution was poured away NO2, and two for collecting SO2 were tested: (i) TEA-coated and the bottles were shaken to remove as much liquid as Whatman 41 paper filters; (ii) TEA-coated glass beads in possible.Bottles (minus lids)+filter (wet) were dried over PTFE tubes of diVerent lengths and diameters; (iii) TEA- silica gel in a vacuum desiccator over a 48 h period.On coated charcoal (20–40 mesh) in PTFE tubes of diVerent completion of the 48 h drying time, the lids were placed tightly lengths and diameters; (iv) TEA-coated Sep-Pak silica and C18 on the bottles. The filters were stored in this way until needed. solid sorbent cartridges. The procedure used to test the collection eYciencies (CE) 2.4. Preparation of standard gases and test atmospheres (NO2, of NO2 and SO2 was as follows.Each sampler was tested by NO and SO2) using known concentrations of NO2 and SO2 to evaluate CE Known concentration test atmospheres of NO2, NO and SO2 at a range of flow rates. The CE values were then determined (50–200 ppb) were generated at 0.2–4 l min-1 by using certi- by comparing the concentrations of the test gas, measured fied cylinders of compressed NO2, NO and SO2 gases with upstream and downstream of the collection device using dilution with compressed air in a two-channel gas calibration continuous instruments (Scintrex NO2 and API SO2 analysers).system equipped with twin mass flow controllers. Stable gas Results are presented in Table 1. Recovery eYciencies (RE) concentrations of NO2 and SO2 were obtained from the above were determined by comparing the total amounts of NO2-, standard gas preparation system after 11 h.Up to 25% of NO3-, SO3 2- and SO42- measured by ion chromatographic NO2 or SO2 could be lost to internal surfaces of the system analysis of the extracted samples with the amounts expected during the first 6 h. The system was therefore allowed to based on the known levels of NO2 and SO2 input. Results stabilize overnight to ensure accurate concentrations of stan- from NO2 and SO2 recovery experiments are also listed dard gases.in Table 1. As seen in Table 1, both TEA-impregnated filters and glass 2.5. Sampling beads for collecting NO2 and SO2 at flow rates over 1 l min-1 were unsuitable in view of their low collection eYciencies.Air samples of 12–240 l were sampled at 0.2–4 l min-1 through Although the charcoal sorbent impregnated with TEA was TEA-coated cartridges or filters after filtration through a used successfully to sample for NO2 and SO2, it was found to 37 mm PTFE filter mounted behind a PM10 inlet to collect contain high blank levels of Cl- and NO3-. Results showed PM10 particles.The exposed cartridge was manually changed that the use of the impregnated Sep-Pak cartridges for simul- at the end of every hour. taneously sampling NO2 and SO2 provided a high collection eYciency (100%), sampling at a high flow rate, compact size 2.6. Analytical procedure and low blanks. The TEA-impregnated Sep-Pak C18 cartridge After sampling, a solution of 5% methanol in DDW (5 ml ) was therefore selected as the sampling method in this study.was flushed through the cartridge using a 5 ml PTFE-lined glass syringe. The eluate of 4.5 ml was collected and analysed 3.2. Laboratory tests to refine the method for NO2-, NO3 -, SO32- and SO42- using ion chromatography. The anionic substances were identified by their retention Breakthrough. Experiments to determine the breakthrough volumes were performed.Known concentrations of NO2 and times and quantified by peak area. The concentration of NO2 Table 1 Results of collection and recovery eYciencies for the diVerent samplers CE (%) RE (%) Flow rate/ Duration/ NO2/ SO2/ Sampler type l min-1 h ppb ppb NO2 SO2 NO2 SO2 37 mm Whatman 41 filter coated with 25% TEA 4.0 1 50 80 <10 <15 NA NA 2.0 1 50 80 <10 <30 NA NA 1.0 1 50 80 >25 >40 NA NA 0.2 1 50 80 >70 >75 NA NA PTFE tubes packed with glass beads coated with 25% TEA 18 cm×1/4 id 1.0 1 50 80 >30 NA NA NA 9 cm ×1/4 id 1.0 1 50 80 >20 NA NA NA 18 cm×1/2 id 0.2 1 50 80 >85 NA NA NA 9 cm ×1/2 id 0.2 1 50 80 >75 NA NA NA PTFE tubes packed with charcoal coated with 25% TEA 9 cm ×1/4 id 2.5 1 50 80 100 100 NA NA 9 cm ×1/2 id 2.5 1 50 80 100 100 NA NA Sep-Pak C18 solid sorbent cartridge coated with 25% TEA 3.0 1 50 80 100 100 93–102 90–101 (n=3) (n=3) Sep-Pak silica solid sorbent cartridge coated with 25% TEA 3.0 1 50 80 100 100 86–93 87–98 (n=3) (n=3) NA, not available; n=3, three replicates. 424 J. Environ. Monit., 1999, 1, 423–426Table 2 Recoveries of NO2 and SO2 Recovery (%) Relative standard deviation (%) Amount of sample collected NO2 SO2 NO2 SO2 20.28 mg of NO2+31.33 mg of SO2 82.0, 80.5, 93.5, 86.7 98.1, 107.6, 86.4, 97.3 ±5.1 ±7.5 40.56 mg of NO2+62.66 mg of SO2 102.5, 94.0, 104.5, 93.2 101.1, 87.6, 92.8, 98.1 ±5.0 ±5.1 SO2 were drawn through a Sep-Pak C18 cartridge impregnated months, if stored in a closed container at 4 °C.To evaluate the stability of collected NO2 and SO2, five parallel samples with TEA.The cartridges contained 0.35 g of C18 solid sorbent, and retained approximately 0.2 g of coating materials. were collected. One sample was analysed immediately; the others were capped and stored in a closed container at 4 °C. Breakthrough was considered to have occurred when NO2 or SO2 could be detected by the Scintrex NO2 analyser or API The results are listed in Table 3.The results in Table 3 indicate that less than 10% of NO2 and SO2 was lost from the sample SO2 analyser at the outlet of the cartridge. It was found that the breakthrough volumes at 20 °C for sampling NO2 and SO2 after 14 days. at an inlet concentration of 100 ppb on the TEA-impregnated C18 Sep-Pak cartridge were 3750 l for NO2 and 1830 l for SO2.Interference. Interference tests were performed using The absorption capacity of the TEA-impregnated Sep-Pak C18 prepared mixtures containing NO, NO2 and SO2 in air, in the cartridge for sampling NO2 and SO2 was also determined. The range of 50–150 ppb. NO, NO2 and SO2 concentrations were absorption capacities for NO2 and SO2 were 745 and 481 mg monitored continuously at both the inlet and outlet of the respectively, calculated as the recovered masses of the anacartridge.There was almost no NO absorption, indicated by lytes expressed as the loading of NO2 or SO2 on the caridentical NO concentrations measured at the inlet and outlet tridge at the time of breakthrough. It has been previously of the cartridge. The results indicated that the TEA-impreg- reported that 800 mg of NO2 can be collected with 770 mg of nated Sep-Pak C18 cartridge was very eVective for collecting TEA-impregnated solid sorbent.2 NO2 and SO2, but would not collect NO.An earlier study has also shown that changes of NO concentration were negligible Detection limit and analyte recovery. The blank test was when NO was passed through a TEA-coated annular denuder.8 performed similarly with the impregnated Sep-Pak C18 car- Although both peroxyacetylnitrate (PAN) and HNO2 may tridges exposed to zero air.The mean (± standard deviation; interfere with the determination of NO2,4,9 these interferences n=6) blank levels for NO2-, NO3-, SO32- and SO42- were are considered to be negligible in the current application due 0.10±0.04, 0.11±0.04, 0.14±0.054 and 0.12±0.033 mg to the generally low concentrations of ambient PAN and respectively.On the basis of the precision of the blanks, NO2 HNO2 in the UK. Interferences from particulate nitrate and and SO2 detection limits of 0.30 and 0.36 ppb under the sulfate in air were prevented by using a PTFE pre-filter. The sampling conditions of 3 l min-1 flow rate and 1 h duration particle collection eYciency of the PTFE pre-filter was 99.3%, were calculated.NO2 and SO2 recoveries for the method were measured using a Condensation Particle Counter (TSI 3022A). studied at diVerent concentration levels. Both NO2 and SO2 Interference from gaseous sulfur trioxide is not expected, as were collected at 3 l min-1 for 1 h on one sampling tube and this gas is not stable in ambient air.analysed. The results are shown in Table 2. Each value in Table 2 represents an average of four measurements. The recoveries of NO2 and SO2 were calculated by measuring the 3.3. Field tests total of NO2-+NO3-, and SO32-+SO42-. Recoveries for both NO2 and SO2 exceeded 85%. Several field samples were obtained in Birmingham to test this method, and the results are presented in Tables 4 and 5.As a Stability during storage. The unexposed TEA-impregnated means of calibrating the system, a direct reading Scintrex Sep-Pak C18 cartridge itself remains usable for at least 2 luminol NO2 analyser operated from a battery was fixed on a military radio frame which can be worn by volunteers. This Table 3 Stability of NO2 and SO2 collected on the TEA-impregnated provides a method of continuous monitoring of the personal Sep-Pak C18 cartridge exposure to NO2.A comparison of the Sep-Pak C18 cartridge sampling method for the determination of NO2 (method A) Storage time/days NO2/mg SO2 /mg in indoor air with the direct chemiluminescence NO2 reading 0 6.51 5.45 (method B) was conducted. It can be seen in Fig. 1 that the 4 6.68 5.11 agreement between method A and method B is fairly good 7 6.24 5.30 (r2=0.92). The cartridge method has now been used success- 14 5.92 5.00 fully for several months in a study of personal exposure to 30 5.75 4.48 nitrogen dioxide. Table 4 Personal exposure monitoring of NO2 and SO2 indoors in Birmingham [NO2] [SO2 ] Sampling date and period Sampling site Flow rate/l min-1 Duration/min ppb mg m-3 ppb mg m-3 24 November 1998 14555–15555 In oYce and laboratory 1.5 60 36.1 70.1 3.1 8.4 24 November 1998 16505–17505 In oYce and laboratory 1.5 90 40.9 79.5 4.5 12.2 25 November 1998 20520–21540 In home (gas cooker) 1.5 100 35.6 69.2 3.2 8.7 25 November 1998 21545–22535 In home (gas cooker) 1.5 50 45.2 87.8 4.5 12.2 26 November 1998 10510–11510 In oYce and laboratory 1.5 60 31.4 61.0 1.1 3.0 26 November 1998 10510–11510 In oYce and laboratory 1.5 20 25.3 49.2 2.1 5.7 27 November 1998 10500–11532 In oYce and laboratory 1.5 92 37.3 72.5 4.0 10.8 J.Environ. Monit., 1999, 1, 423–426 425Table 5 Personal exposure monitoring of NO2 and SO2 outdoors in Birmingham [NO2] [SO2 ] Flow rate/ Duration/ Sampling date and period Sampling site l min-1 min ppb mg m-3 ppb mg m-3 26 November 1998 8550–10500 Walk along street (Quinton and Harborne) 1.5 70 48.0 93.3 4.4 11.9 27 November 1998 8550–9555 Walk along street (Quinton and Harborne) 1.5 65 72.3 140.5 5.6 15.1 27 November 1998 18520–19525 Walk along street (Quinton and Harborne) 1.5 65 61.9 120.3 8.1 21.9 29 November 1998 11530–12555 In front garden of house (Wolverhampton Rd) 1.5 85 57.5 111.7 3.9 10.5 29 November 1998 13515–14500 In front garden of house (Wolverhampton Rd) 1.5 45 60.9 118.3 3.8 10.3 30 November 1998 9515–11520 Walk along street (Harborne high street) 1.5 125 80.6 156.6 14.1 38.1 Acknowledgements This research was supported by the UK Department of Environment, Transport and the Regions through contract No.EPG 1/3/111.We thank Dr. Jiping Shi at the Institute of Public and Environmental Health at the University of Birmingham for measuring the particle collection eYciency in this study. References 1 N. A. H. Janssen, G. Hoek and H. Harssema, J. Exposure Anal. Environ. Epid., 1998, 8, 31. Fig. 1 Comparison of Sep-Pak cartridge and chemiluminescence for 2 D. V. Vinjamoori and C. S. Ling, Anal. Chem., 1981, 53, 1689. determination of NO2 in indoor air. 3 Y. Nishikawa and K. Taguchi, J. Chromatogr., 1987, 396, 251. 4 J. E. Sickles, P. M. Grohse, L. L. Hodson, C. A. Salmons, K. W. Cox and A. R. Turner, Anal. Chem., 1990, 62, 338. 4. Conclusion 5 D. Krochmal and A. Kalina, Atmos. Environ., 1997, 31, 3473. Personal monitoring methods for measuring NO2 and SO2 6 M. R. Heal and J. N. Cape, Atmos. Environ., 1997, 31, 1911. 7 S. O. Baek, Y. S. Kim and R. Perry, Atmos. Environ., 1997, 31, 529. have been established and validated. The methodology 8 J. P. Shi, The chemistry of nitrogen oxides in urban air, PhD Thesis, described is simple, reliable and light in weight for convenience University of Birmingham, 1996. of use. Its good precision and sensitivity make the method 9 M. W. M. Hisham and D. Grosjean, Atmos. Environ., 1990, 24A, suitable for sampling NO2 and SO2 over short periods (around 2523. 1 h). It is therefore ideal for high time resolution personal sampling, and oVers an advance over earlier methods which Paper 9/05305K were capable of only low flow rates. 426 J. Environ. Monit., 1999, 1, 423–426

ISSN:0960-7919    

DOI:10.1039/a905305k

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

Measuring strategies and monitoring of the indoor environment† Thomas Schneider National Institute of Occupational Health, Lersø Parkalle 105, DK-2100 Copenhagen, Denmark. E-mail: ts@ami.dk This paper presents key elements on which to base the design of a strategy for a particular indoor environment measurement task. Starting from a general strategy for responding to occupant complaints and conducting epidemiological and intervention studies the paper concentrates on selected key elements, with focus on air contaminants.Strategies for grouping of occupants with the purpose of exposure assessment should aim at optimising exposure contrast between groups, and grouping according to buildings may not be the proper choice. Occupant exposure can be measured by personal sampling or constructed from measured concentrations in microenvironments and activity patterns.The cause of temporal and spatial variability is discussed and guidance is given on the number of samples needed to detect a given change in true concentration level. The sick building syndrome (SBS) is defined as those situations where a high proportion of building occupants are complaining about mucous membrane or skin irritation, general symptoms.When investigating SBS, questionnaires should be used to collect structured information from occupants on perception of environmental conditions, psychosocial factors and symptoms. There is as yet no consensus on choice of length of reference period for reporting symptoms and for estimating exposure. Finally, methods for identifying and quantification of sources of air contaminants are discussed, including lab-scale or full-scale simulation, on-site measurement of source emission, air monitoring, surface sampling, and modelling.to modify the indoor environment). Sensory irritation in eyes, Introduction nose and throat must be dominating while systemic symptoms Complaints about the quality of non-industrial, oYce like (e.g.from stomach) must be infrequent. It is important to environments are common and air contaminants are usually recognise that since this definition of SBS is based on prevaconsidered as one of the prime causes for complaints. lence the unit of observation is a group of occupants. There Contaminants may be present, which do not lead to complaints is no general consensus on how to describe SBS at the but that have potential to cause chronic health eVects, such individual level, i.e.how to identify a case.4 as radon and legionella. Mølhave1 summarises the eVects It was already suggested byWHO2 that psychological factors of indoor air contaminants on human health as: adverse/ could be involved in SBS. Psychological and social factors irreversible or chronic eVects, including cancer, reproductive have since been shown to be important determinants of SBS eVects, and allergic sensitisation; hyperreactivity and non- and may be dominating if exposure to physical factors and air specific hypersensitivity reactions; non-adverse/reversible contaminants is low.5 Recent research suggests that physicoe Vects including general symptoms such as headache; inflam- chemical and psychosocial eVect mechanisms are mutually matory irritation, rashes etc., and perceptions including interacting.6 Thus, design of strategies for measuring the odours, sensory irritation and feeling of dryness of skin, eye indoor environment quality is a great challenge.or nose. A measurement strategy must always be designed for the The term sick building syndrome (SBS) was originally particular task and prevailing conditions.General recommencoined by WHO.2 It is a characteristic of SBS that no obvious dations for strategies thus can only be given in rather broad cause can be found even after extensive investigation of air terms. This paper briefly presents a general strategy and then contaminants, ventilation system performance and the building concentrates on selected key elements, on which to base the structure itself.A broad spectrum of symptoms is reported in design of a particular strategy, with focus on air contaminants. sick buildings, but the symptoms have many common features. Berglund et al.3 distinguish between building related illness (BRI) which has a known etiology, definite symptoms and General strategies can be diagnosed, e.g.by the presence of specific antibodies Indoor environment investigations may be undertaken for a to causative agents in the serum of exposed occupants, and range of reasons. 1, As a response to complaints by room SBS. In SBS situations aVected occupants report non-specific occupants or suspicion that exposure factors that give rise to symptoms occurring only during the time when they are at health concern are present.Buildings representing such prob- work, and where a high proportion of building occupants are lems are frequently termed ‘problem buildings’. 2, As part of aVected. The reactions are: acute physiological or sensory an epidemiological study. 3, Assessing eVectiveness of inter- reactions (sensory irritation of mucous membranes or skin; ventions. 4, Need to determine whether specified limits or general malaise, headache and reduced performance; unspecific guideline values are being maintained.hypersensitivity reactions, dryness of skin; odour or taste From a measurement strategy, point 4 is the least complex complaints) and psychosocial reactions (decreased productivity as the measurand is known and strategies for measuring or absenteeism; in contact with health care system; initiatives specific factors are often specified.An example is mineral fibres VDI.7 The eVectiveness of interventions (point 3) can be docu- †Presented at AIRMON ’99, Geilo, Norway, February 10–14, 1999. J. Environ. Monit., 1999, 1, 427–434 427mented using two diVerent types of end-points: etiologic where example assumed to be one month. During this month, person number i in this group has received a true integrated exposure the endpoint is prevalence of manifestations and prevention eVectiveness where the endpoint is exposure.The design during a day, Eij, where j=1, 2, ..., m, and m is the number of days per month.The day integrated exposure averaged over requirements for each type of intervention types are diVerent. Etiologic intervention studies usually need large study popu- the month considered is for person number i, see e.g.:11 lations and large exposure contrasts. Exposure and prevalence of manifestations need to be measured. Ideally, randomisation mi= 1 m . m j=1 Eij and blinding should be applied.EVectiveness of prevention may on the other hand be documented in small study popu- The group mean day integrated exposure is given by: lations with exposure as the only outcome variable and no need for randomisation and blinding.8 m= 1 N . N i=1 mi= 1 Nm . N i=1 . m j=1 Eij When investigating problem buildings (point 1) it is prudent to think in terms of multi-factorial causes.However, a stepwise The within-person (w) number i variance (day-to-day variance approach will be most cost-eVective, since nevertheless, simple for person number i ) is: explanations and solutions to problems can often be found. Several stepwise approaches have been published among them swi2= 1 m-1 . m j=1 (Eij-mi)2 the following by Kukkonen et al.:9 1, Check whether operational conditions of e.g.the HVAC system are normal. Instruct those The average within-person variance is: who complain about the possibilities for individual control. 2, Determine type and extent of problem by using a standardised sw*2= 1 N . N i=1 swi2 ‘sick building’ questionnaire or if there are few employees using structured interviews. 3, Make a technical survey for assessing The between-person (B) variance (within group variance of risk factors inherent in the building or its use and operation. 4, one month average day integrated exposure) is: Assess building and its fittings and furniture, quality of cleaning, moisture damage and mold growth, temperature, air movements sB2= 1 N-1 .N i=1 (mi-m)2 using smoke tubes, and carbon dioxide concentration.Estimate degree of recirculation of air and possible contamination of Cost and practical diYculties and inconvenience limit the total intake air. 5, Measure ventilation eYciencies. 6, Make detailed number of persons and samples and thus direct measurement assessment of contaminant sources and concentration. 7, Make of every Eij for each individual is rarely possible. Therefore, m clinical examination of aZicted persons and additional occuhas to be estimated from measurements, which do not cover pational hygiene investigations.all person and all days. As a consequence, only a sub-sample Kukkonen et al.9 discuss for all common types of physical, of all Eij for a given month is available. We will consider the chemical and microbiological exposures reasons for measurecase where n out of N persons have been sampled, and each ment, and describe how to measure these factors.They also sampled r out of m days, and that sampling is random. The propose which levels should be considered as low, medium or Eij for the given person and days are assumed to be known high. For volatile organic compounds Mølhave1 has proposed without error. The group mean m is then estimated as: four ranges: comfort range (no irritation or discomfort); range of multi-factorial exposures (irritation and discomfort possible m�= 1 nr .n i=1 . r j=1 Eij if other exposures interact); range of discomfort; and range of toxic eVects. The variance of this estimator is: Exposure measurement Var(m� )= 1 n A1 r - 1 mBsw*2+A1 n - 1 NBsB2 Defining exposure The ultimate definition of exposure is given by Ott10 as # 1 nr sw*2+ 1 n sB2 (for mIr and NIn) ‘Exposure is the joint occurrence of event A (person is present at location x, y at time t) and event B (concentration C(t) is In this approach exposure has been estimated for each person present at location x, y at time t)’.It is rarely possible to using a day as the basic unit.For epidemiological studies of assess exposure with this ultimate resolution. Necessary shortubiquitous exposures, days would include all days. In studies cuts blur the picture, such as averaging both event A and of SBS only work-days would be included. Dust mite, cat and event B over a room, averaging time over a day, representing other allergens and mould in the homes of residences could individual exposure by a group mean, etc.be confounders and should be quantified. Another approach is to construct cumulative exposure math- Estimation of group exposure ematically by combining information on the time persons Integrated, or cumulative exposure, Ei, for occupant number spend in various locations with the concentrations expected to i, for a time period [tia,tib] of duration relevant to the eVect is occur there.12 As an example, let a building be divided into defined as: small geographical entities or microenvironments within which concentration is uniform (according to a specified degree of Ei=Ptib tia Ci(t)dt, uniformity).Suppose person number i spends the time interval [ta,tb] in e diVerent microenvironments. The e-dimensional where Ci is the concentration in the breathing zone of occupant vector Di=(Di1, ..., Dis, ..., Die), where Dis is the total time number i.A basic task in exposure assessment for epidemiology person number i spends in microenvironment s during the or intervention studies is to assign Ei for a given time period, time period [tia,tib] is referred to as the time budget. Then to a priori defined groups of individuals. Suppose exposure of a defined group of N persons has to be characterised.A Ei= . e j=1 CsDis temporal resolution of one day is chosen in this example, but could be one hour or any other suitable period. The exposure where Cs is the concentration in room s. When studying SBS, Ei can be subdivided into cumulative exposure originating of the group is followed for a relevant time period in this 428 J.Environ. Monit., 1999, 1, 427–434from those micro-environments where symptom prevalence is during the daytime in homes was more than 50% above concurrent concentration measurements by stationary indoor high and low, respectively, as this can give clues for likely causes of the symptoms. In occupational hygiene the term task samplers.This increase in particle concentration was dominated by coarse particles (above 2.5 mm), and it was suggested time weights is used.13 Much of the within person variance of integrated exposure is caused by variations in Di.14 Di can be that in part they were particles resuspended from carpets and furniture by motion. determined by questionnaire or time activity diaries.Electronic microenvironment sensors could be designed for measuring The mixing caused by turbulent diVusion in the turbulent core of the room is not instantaneous. This causes short-term temperature, light, noise, etc., and be trained to recognise typical micro-environments.15 variations in concentration while the turbulent mixing process of the pollutant plume with the surrounding air is proceeding.Grouping This is seen as ‘noise’ on the output of a fast response direct reading instrument. Suppose an exposure-response study has to be designed for The time it takes until mixing is complete has been investithe relation between SBS symptom prevalence and concen- gated by Baughman et al.23 They defined the characteristic tration of airborne micro-organisms, treating other exposures mixing time as the time period required for an instantaneous as confounders or co-variates.Prevalence of symptoms can be release from a point source to become uniform to within a determined by distributing questionnaires to all occupants. relative standard deviation of <10%. For stirring induced by Exposure to micro-organisms can be determined by personal active heating sources and solar load they found characteristic samplers or by stationary samplers in defined microenviron- mixing times of a tracer gas of less than 15 min.For a typical ments. Due to cost limitations, exposure usually has to be mechanical ventilation scenario resulting in Nair=1.25 h-1, a estimated from a sub-sample of occupants or microenviron- characteristic mixing time of 7.4 min has been found.24 ments.Therefore occupants or microenvironments have to be Temporal variations on a larger time-scale are caused by divided into well-defined groups from which a random sub- variations in source position, output and removal rates. The sample can be drawn. If groups are formed in a way that response of a room to a change in source output rate is deterwithin group variance is large compared with between group mined by the total removal rate Ntot=Nair+Ne, where Nair is air variance, a possible dose-response for micro-organisms will be exchange rate, and Ne is removal rate to all other sinks, such as attenuated or go undetected.16 This would be the case if e.g.surface deposition. Ne is termed the equivalent air exchange rate.persons were grouped according to buildings, and the between For a given and constant Ntot the concentration (C) after inactiroom variance was large compared with the between building vation of all sources at time t=0 will decay as: variance. A pilot study may have to be performed to estimate the relative magnitude of the within and between building C(t)=C(0)exp(-Ntott) variance in order to design an eVective grouping and sampling Then strategy.Obviously, if all Ei have been determined grouping can be made a posteriori. Considerable knowledge on the role Pt=2 t=0 C(t)dt= 1 Ntot C(0) of uncertainty in exposure estimation and attenuation has accumulated in industrial epidemiology.17 For the indoor 1/Ntot is the average residence time in the air of a molecule or environment criteria for eVective grouping should be estabparticle entering the room air at t=0.Large particles may lished, in particular, the merits of using building characteristics. have a residence time in the air shorter than the characteristic mixing time. Particles less than 5 mm, have residence times Measurement of concentrations 1/Ne>45 min (neglecting Nair), thus there will be suYcient In industrial hygiene it is well documented that airborne time for mixing.For particles larger than 10 mm 1/Ne<10 min concentrations may exhibit large spatial and temporal variad there would not be suYcient time for mixing before the bility. Data on spatial and temporal variability in the indoor particles deposit on the floor.This is a simple model and does environment are scarcer. This may be because design of not include eVects of the buoyant thermal plume generated by measurement strategies was influenced by environmental pol- persons and that the generation of large airborne particles will lution thinking, as might be suggested by the following often coincide with high activity and thus with increased quotation from ISO.18 ‘The indoor environment is rarely static stirring.Field data are needed on the mixing of nonsince the substance concentration may be constantly altered respirable particles. by the strength of the source, human activity, ventilation rate, Roach25 (and several others) have studied concentration external or internal climatic conditions, chemical reactions and variability using a time series approach.He assumed to have possible sinks... A special situation often arises in the indoor a source injecting contaminants (varying as white noise around environment because of the short distance between source and a constant mean) into a room and being removed at a rate receptor. Also the composition of indoor air may vary within Ntot. Then for a sample of duration T the measured concenand between rooms and be less homogeneous than the outdoor tration Cav would vary with a variance sCav 2 given as: air surrounding the building.’ sCav 2= K V 2Ntot f (Ntot, T ) Factors causing temporal and within room variability.While a point source is active, a concentration gradient is maintained where in the air due to the turbulent diVusion.For intermittent sources, the average concentration gradient is reduced as the f (NtotT )= NtotT+exp(-NtotT )-1 Ntot2T 2 ratio of on to oV periods decreases.19 The convective plume around the body carries dust (skin scales, textile fibres, bacteria, etc.) originating from the body K is a constant and V is the room volume. It is seen that the variance is inversely proportional to the room volume squared as well as dust resuspended by the persons activity or contaminants (respectively clean air) from a layer close to the floor to and depends on total removal rate and sampling time as shown in Fig. 1. Notice that complete and immediate mixing the breathing zone.20,21 Thus a significant dust source may ‘follow’ the person, introducing an ever-present concentration in the entire room is assumed.Rappaport and Spear26 interpreted Ntot as a local dilution rate in e.g. the breathing zone. gradient. This phenomenon has been termed ‘personal cloud’22 and it was found that exposure measured by personal samplers The above characterised the short and medium scale spatial J. Environ. Monit., 1999, 1, 427–434 429Table 2 Number of work-days ND to sample during one year (250 d) to catch at least one event with confidence p.E is number of eventdays during a year E 6 12 26 52 p=95% 97 54 26 13 p=90% 79 43 20 10 p=80% 58 31 15 7 pling during simulation of events, may considerably reduce sampling eVorts. Recommendations by international bodies/groups. An ISO working group intends to provide guidance on sources and sampling strategies, and some preliminarymaterial is available.18 Sampling times up to one hour are referred to as short-term Fig. 1 Plot of the function f (Ntot, T ) for total removal rate Ntot= sampling, and sampling times ranging from several hours to 0.01, 1, and 10 h-1. several days as long-term sampling. If using short-term sampling for obtaining representative results it is recommended that sevand temporal within-room variability.Data on the relative eral hours should be allowed after a step change, such as opening magnitude of day-to-day, between room and between building a window or starting the ventilation system, for establishment variance is scarce, even though it is important. Such data are of equilibrium (i.e. steady state). If heating, ventilation and airneeded for choice of measurement strategies, for the statistical conditioning (HVAC) is used the operating parameters and state power of an intervention study, and for grouping strategies. of maintenance must be included in the report.If HVAC oper- Kildesø et al.27 measured total and respirable airborne dust ation is intermittent or restricted at least three hours should every Thursday for 26 weeks in an oYce building, using elapse with the HVAC performing normally.stationary samplers as part of an intervention study. Kildesø The concept of representativity is useful, but can also be et al.28 used the results to estimate the population mean and misleading. If concentration is variable, then deviations from standard deviation of airborne and surface dust concentration.the average are normal and should be quantified. If representa- Given this standard deviation the necessary sample size, N, tive means ‘close to the long-term average’ then it is inherent was then calculated for measurements before and after an in the definition that it is possible by a priori criteria to decide intervention to detect a change in true mean of 10%, 25%, if the day selected for measurement will give exposures close and 50%, setting level of significance, a=0.05 and power, to the long-term average.Experience from industry indicates 1-b=0.8. The results are given in Table 1. that it can be diYcult for workers to correctly judge if a given day is representative regarding exposure.30 Thus several repeat Probability of detecting single events.Contaminants may be measurements may be needed to correctly assess if a day was released in an unpredictable and episodic way. How certain representative. This is very often impractical or outside the can the investigator be that an event is included during an air resources of the study and as an alternative as much inforsampling program? Consider a total number of days, D, E of mation as possible about the exposure modifiers is collected which include an episode.Assume that ND days are measured, with the assumption that this information can be used to and that events are independent of measurement days. Then determine the ‘degree of normality’. Several check-lists for the probability p of having included at least one event is:29 information to be recorded during indoor air measurement have been published.9,18 However, if it turns out that the p=1- (D-E)! (D-E-ND)! (D-ND)! D! measurement day was not normal no guidance is given on how to use this information to estimate the degree of deviation from long-term average exposure.If the value of exposure Table 2 shows ND calculated for p=0.8, 0.9, and 0.95. It is seen that cost of air sampling for detecting random events modifiers during the period of sampling has been quantified, and there are means of modelling their eVect on exposure, the with a high degree of certainty can become prohibitive. Targeted sampling of identifiable events or alternatively, sam- exposure prediction interval for the long-term average expo- Table 1 Necessary number of single measurements before and after intervention to detect change in average dust levels d.Level of significance, a=0.05, power, 1-b=0.8. From Kildesø et al.28 Number of individual samples for d= Population standard Type of measurement Mean value deviation 10% 25% 50% Airborne— Total dusta 57 21 145 23 6 Respirable dusta 46 18 173 28 7 Carpets— Walking zoneb 0.88 1.1 1651 264 66 Non-walking zoneb 0.54 0.40 590 94 24 Furniture, hard surfaces— Close to personc 2.5 2.1 807 129 32 Easily accessiblec 2.2 1.7 673 108 27 Othersb 4.4 4.9 1370 219 55 amg m-3.bDust index as measured with Carpettester. cArea of surface covered with dust, in percent. 430 J. Environ. Monit., 1999, 1, 427–434sure can be tightened. In large-scale surveys statistical modelling can be used to assess the eVect of the modifiers.To have credibility the resulting statistical model parameter values should not be at odds with physics.31 ISO18 discuss suitable sampling locations recommending the centre of the room and 1–1.5 m above the floor (breathing zone). However, the inhomogeneity caused by sources, forced thermal movement and prevailing air movements are stressed.It is also recommended to take outdoor samples in the vicinity of the building, but not closer than 1 m. If the building is equipped with an HVAC system, the outdoor air sample should be taken close to the intake. It should be noticed that large concentration gradients can be present both vertically and horizontally in the street canyons. Perceived environmental conditions, symptoms and Fig. 2 95% upper confidence limit for number of persons reporting signs symptoms that are due to background only, given total number of persons in a group. Two levels of background: 20% and 30% are shown. Questionnaires have been used successfully to collect structured information from occupants regarding perception of environmental conditions (temperature, noise, draught, dry air, odour, used once a week, a four-week duration of an intervention etc.) and symptoms (fatigue, burning eyes, sore throat, etc.) and the corresponding control period is optimal. Both shorter psychosocial factors (such as job satisfaction and job stress) and longer periods increase the variance of the individuals and personal factors (such as gender, age, smoking status and mean marking position on the scale. Marking positions for a allergy).5,32 Jaakkola4 has pointed out that the term symptom group of persons deviate strongly from a normal distribution.is used inconsistently and stressed the diVerence between However, the diVerence for each person between average (subjective) symptoms and objective symptoms (signs). Also marking positions during intervention and during control were the way of collecting information, i.e.using questionnaires normally distributed, opening up the use of parametric statistias opposed to observation or measurement has sometimes cal tests. The authors have calculated the number of completed determined if the outcome has been denoted symptom. VAS questionnaires needed before and after an intervention Questionnaires may readily become too extensive if all aspects to detect a true diVerence d=0.05, 0.1, and 0.2, based on the in a multidisciplinary approach had to be included.Current estimated values of the within-person variance, setting level of research tries to identify which subsets of questions contain significance, a=0.05 and power, 1-b=0.8. Table 3 shows the most information relevant for identifying causes of SBS.It results for the question on perceived air quality. should be noticed that to a psychologist or sociologist, use of Clinical measurements of e.g. tear film stability are often questionnaires is also a measurement. performed in a room diVerent from the occupant’s usual room, Questionnaires have their inherent error sources and like thereby placing the occupant in an artificial condition.Wyon35 any other measurement method a questionnaire must be used simple methods for measuring objective symptoms that validated. As an example, prevalence will depend on how the could be applied at the work site with no interference from question about symptom frequency is asked and what reference measurement instruments. While less precise these methods period is used.Whether a person will answer that at least one eliminate several sources of potential bias if acute eVects are symptom has been experienced within a reference period will present that could be caused or modified by the artificial depend on the persons recollection of past symptoms and on environment in the test room and the presence of the measurethe probability that an exposure triggering the symptom ment instrument.occurred during the reference period. There is no consensus on length of reference period and periods ranging from three Temporal aspects months,33 over two months,9 one month5 to 20 min34 have been used. Design of questionnaires is a job for specialists and The outcomes range from chronic health eVects with latency times of several decades, over allergic sensitisation and ‘good SBS questionnaire design practice’ should be followed.32 For a questionnaire to be useful a well-documented baseline inflammatory irritation to sensory perception.In the case of odour perception the reaction is immediate, but also very obtained with that questionnaire must be available to be used as a reference as established by e.g.Andersson.33 much influenced by olfactory fatigue on prolonged exposures. Interaction and adaptation processes are characteristic of the Prevalence of symptoms is a stochastic variable. Suppose the background prevalence is 20%. In a group of, say, 35 sensory systems involved in perception of odour and mucosal irritation. Therefore, the duration of exposure influences the persons more than 0.2×35=7 persons could have symptoms, and still would not exceed the background.The upper 95% perception.3 Mølhave1 proposes that in order to assess health eVects from multi-factorial air contaminant exposures esti- confidence level has been calculated using the binomial distribution and for groups up to 100 and for two levels of mates of long-term average as well as 24 h average, peak exposure (10–30 s average) and peak frequency would ideally background levels 20% and 30%, respectively (Fig. 2). Visual Analog Scale (VAS) is a special type of questionnaire, be needed. On sampling duration, ISO18 state that for substances having acute eVects short-term sampling should be which oVers the possibility of obtaining an indication of symptom intensity. On a linear scale, each person indicates, used, whereas long-term sampling should be used for substances having a chronic eVect.For a comprehensive assess- by placing a mark between two statements concerning a specific condition, how he or she is feeling at the moment. ment, both a short-term and a long-term sample should be collected. Sometimes the sampling duration is determined by These questionnaires are easily administered, quick to complete, and the result does not depend on the person’s recollec- the limit of quantification of the available analytical method.Hodgson et al.34 used questionnaires including questions tion of past symptoms. Wyon35 used the VAS approach to document the eVect of interventions in a hospital.asking about how persons were feeling right now and measured air contaminants with direct reading instruments during the Kildesø et al.36 have shown that if a VAS questionnaire is J. Environ. Monit., 1999, 1, 427–434 431Table 3 VAS question: good air quality, bad air quality. Necessary number of persons to detect average change in scale recordings d. Level of significance, a=0.05, power, 1-b=0.8.From Kildesø et al.27 Number of persons for d= Duration Number of Number of Population period/weeks persons questionnaires received variance (s2) 0.05 0.1 0.2 14 70 1220 0.124 614 153 38 8 67 675 0.120 594 148 37 6 64 513 0.142 703 176 44 4 51 312 0.183 906 226 57 20 min it took to fill in the questionnaires. Fortmann et al.37 fluorescein tagged dust in a hospital room and found a decay made measurements during one week and administered the rate in the order 5–7 d in floor dust swept daily.This illustrates questionnaire at the end of the measurement week. Mølhave1 the total ‘memory’ of a room. Allott et al.46 found residence recommends that recording of symptoms should be made times of 29 d of dust in mats close to the entrance in a private within the time period covered by the sampling of air contami- home being the combined result of daily vacuuming, track-in nants but that a certain range outside this period is acceptable and redistribution by footwear, deposition from the air and for practical reasons for eVects that are not acute.resuspension. In the related area of biological monitoring a basic question The ratio between airborne (mg m-3) and surface (mg m-2) is when to take the biological sample in relation to a time concentration (resuspension factor in units of m-1) ranges varying exposure and how much ‘memory’ of past exposure orders of magnitude depending on the scenario and is not incidents there is in a biological sample.Physiologically based predictable. For an oYce, being an example of a specific pharmacokinetic modelling can be used for determining proper scenario, Kildesø et al.47 found a correlation (R2=0.46, time constants.38 The sampling task can be formulated in an p=0.05) between dust accumulation rate (measured in % alternative way: air sampling should reflect the variability in surface area covered by dust) and coarse airborne aerosol the target organ.The basic concept is physiological damping (PM10–PM2.5; PM=particulate matter). expressed e.g. as the ratio R:39 Surface dust levels show large day-to-day variability. In an intervention study on cleaning Kildesø et al.27 measured surface dust every Thursday before cleaning for 26 weeks in R= Asx mxBNAsC mCB an oYce building. Dust from hard surfaces was sampled by gelatine foils and the percentage area covered by dust particles where s/m is the relative standard deviation for target organ measured by light scattering from a circular area of diameter burden, x, and for air concentration, C, respectively. The 15 mm.Carpets were evaluated by a method based on sampling requirement for air sampling averaging time can then be stated of dust by a standardised vacuum mouthpiece, collection on as: choose averaging time so that R=1. Roach25 showed that gelatine foil by impaction and measurement by light scattering. R is close to one if the sampling averaging time T is 2 to 3 The result is given as a dust index ranging from 0–100%.times the time constant (b) in the target organ. Thus selection Kildesø et al.28 used the results to calculate the necessary of proper averaging times for measuring air contaminants can number of individual measurements to detect a change in true be formulated in terms of relevant time constants for indoor level of surface dust of 10%, 25%, and 50%, setting level of environment eVects.significance, a=0.05 and power, 1-b=0.8. The results are given in Table 1. Characterisation of contaminant sources Some surface sampling methods intend to sample the contaminant with highest possible recovery.Thus these methods Indoor sources of airborne contaminants are plenty and quantify the total mass in the surface contamination layer. include consumer and commercial products, building sources, Another purpose is to assess the transport, such as transfer of personal sources, and outdoor sources.40 Sources should pesticides to skin or resupension.The transport will depend always be identified and if reasonably practicable be reduced on the external driving force and thus one approach is to use or eliminated. a method defined driving force. As an example Kildesø et al.48 Room and furniture surfaces constitute important sources have developed the STEPP tester, which by dropping a hemi- of contaminants.Emission of gases and vapours can be sphere from a well defined height on the surface and collecting measured by an emission cell, termed FLEC.41 The cell is all resuspended dust, simulates resuspension from a carpet placed tightly on the surface and a well controlled flow of gas caused by a walking person.Resuspension caused by air is directed over the surface and sampled. The cell will become movements is simulated in the SMAIR tester.49 In this instru- European standard.42 By this method the emission rate from ment a well-defined air jet is directed at a surface contaminant room surfaces, furniture, and carpets, can be quantified in situ. layer and the resupended particles collected on a filter.Surfaces also act as sinks for vapours, which can be released Vacuuming techniques would simulate resuspension caused by if external conditions change. strong air velocities provided the nozzle does not touch Surfaces accumulate dust, which represents a secondary the surface. A fixed geometry must be maintained during source of airborne particles, if activated by mechanical impact vacuuming, such as obtained in the sampler developed by (disturbance by room occupants) and to a lesser degree by air Roberts et al.50 movements.Surface dust also acts as an adsorbent of volatile Transfer of surface contamination such as pesticide residues and semi-volatile organic vapours (VOC, SVOC) and as a to the fingers or other parts of the skin is an important carrier of e.g.detergents,43 pesticides and microbial toxins. As exposure source particularly for toddlers in homes where dust accumulates on surfaces the surface dust stores inforpesticides are used. The exposure risk can be estimated by mation of episodic releases of specific contaminants. It has using a fixed pressure transferable residue sampler, such as been shown that surface sampling could demonstrate the the polyurethane foam (PUF) roller which intends to simulate presence of synthetic vitreous fibres in rooms where air sampling was below the detection limit.44 Lowbury45 dispersed the contact between skin and surface.51 432 J.Environ. Monit., 1999, 1, 427–434Schneider et al.52 provided the statistical framework for outdoor, decay rate due to surface deposition, and source output rate from cooking, smoking, and ‘other sources’. designing a sampling strategy for surface dust concentrations.First the micro-scale correlation structure of surface dust Deterministic models exploit the physics and chemistry of air contaminant generation to predict airborne concentrations. concentration measured as percentage of surface covered by particles was determined.The results were then used to select It is beyond the scope of the present paper to summarise published models but it is stressed that only models, which the surface sampling spot size for dust that was large enough to provide suYcient sample and small enough to reflect micro- have been validated against measurement should be used. Used with prudence, deterministic models for air concentration scale variability.Given this spot size, large surface areas were mapped to determine the macro-scale correlation structure. can: Facilitate design of cost-eVective exposure measurement strategies including estimation of order of magnitude of time One parameter is the range of influence (the distance between two measuring spot positions beyond which there is no corre- constants.Assist interpretation of recorded external factors in relation to concentration measurements. Be used for data lation). This range was then used to determine the proper distance between sampling grid points. Individual surface dust interpolation, interpretation and reduction. Support design and implementation of eVective control strategies based on concentrations were found to be log-normally distributed and this has implications for estimating the average (or total ) source reduction, ventilation and cleaning.Support extrapolation of laboratory scale test results to full scale. surface concentration. For sample sizes n<20, the arithmetic mean of the n results is superior to the maximum likelihood estimated mean, but inferior if n>50.53 Conclusions When designing sampling strategies, selecting sampling methods and interpreting the role of surface dust in relation Key elements of methodologies for quantifying low-level, multi-factorial exposures and for characterising eVects of such to indoor air quality it is important to make the following distinctions: Surface dust accumulated only via deposition exposures on building occupants have been presented.The definition of observational units or groups of both from the airborne state and on surfaces that are not disturbed reflects what has been in the air during the entire period since occupants and microenvironments needs careful consideration. The task is to obtain small within group and large between the surface was last cleaned.Surface dust (tracked in or deposited from the air) on surfaces that become agitated group variance. EVective grouping criteria need to be developed. Development of easy to use and aVordable personal represent a secondary source of airborne dust but not what has been in the air. The amount of surface dust indicates the monitors that could be worn by a large proportion of persons in a study would reduce this need, however.eYciency and time since last cleaning. During exposure measurements use of the rooms should be The physics of contaminant generation and removal can be used to select proper sampling positions and timing. The normal. It only makes sense to make dust measurements in unoccupied rooms if one wants to determine the contribution necessary number of samples to be taken to detect episodes may be large.The variability in air and surface dust concen- by outdoor/ventilation air or some active dust sources, such as printers/copiers. Alternatively, conditions of usage, which tration is considerable and for detecting a change in surface dust level of, say, 25%, more than 200 samples in total would induce resuspension should be simulated.The VDI 3492 guideline7 describes a strategy, which is not intended to are needed. For investigation of SBS, questionnaires should be used that predict long-term changes in fibre concentrations but to bring about and determine short-term maximum fibre concen- collect data on perceived environmental conditions, psychosocial factors, and personal factors.Unfortunately question- trations. Rules are given for selecting number and position of samples. A key part of the strategy is the ‘simulation of the naires are not standardised, making comparison between studies diYcult. conditions of use’ with the intention to activate secondary sources. Activation is obtained by: Air movement using a Criteria for selecting reference periods for symptom reporting and exposure measurement are not based on a solid blower generating a given velocity at the surface.Producing vibrations by bouncing a ball or slamming a door in a defined scientific background. A criterion is proposed which relates the averaging (or sampling) time for air contaminants to way. If there are carpeted floors by dropping an object typical for that room from the height of 1 m.relevant time constants for indoor environment eVects. Source characterisation, such as quantification of emission of vapours from building materials and resuspension of dust Modelling from surfaces, is a useful tool for assessing exposure risk. Large-scale probability based studies of personal exposures Since external forces are needed to resuspend surface dust, have been conducted, such as the PTEAM study.22,54 In this resuspension has to be assessed using method defined driving study, 178 persons were selected using three-stage probability forces.sampling, representing 139 000 non-smoking residents in a Statistical modelling can be used to extract source specific geographical area. Personal dust samples PM10 and PM2.5 data from large-scale studies of personal exposures.Deterwere used to collect two consecutive 12 h personal samples ministic modelling is useful for designing measurement stra- (day/night). Stationary samples, indoors and outdoors, were tegies, interpreting results and supporting design of control also taken. Air exchange rates were measured during both strategies. 12 h periods.O� zkaynak et al.22 used statistical modelling to Several gaps remain in the methodologies for characterising quantify the relative contribution of various sources to the low-level, multi-factorial exposures and their eVects. personal dust concentration. The model was PM10=k0+k1Coutdoor+k2smoking+k3cooking References 1 L. Mølhave, Indoor Air, 1998, Suppl. 4, 17. +k4N+k5volume+k6dirt level+error 2 WHO, Indoor air pollutants: exposure and health eVects, Copenhagen, World Health Organisation Regional OYce for where Coutdoor is outdoor concentration, smoking and cooking Europe, 1983.are binary variables, volume is room volume, dirt level is a 3 B. Berglund, B. Brunekreef, H. Kno� ppel, T. Lindvall, M. Maroni visual assessment on a 7-point scale and k0 ,…, k6 are and L.Mølhave, Indoor Air, 1992, 2, 2. regression coeYcients. A mass balance approach was used to 4 J. J. K. Jaakkola, Indoor Air, 1998, Suppl. 4, 7. 5 A. Hedge, W. A. Erickson and G. Rubin, Environ. Int., 1996, 22, 3. estimate for each substance (or element) penetration from J. Environ. Monit., 1999, 1, 427–434 4336 M. Lahtinen, P. Huuhtanen and K. Reijula, Indoor Air, 1998, 30 E.Olsen and B. Jensen, Appl. Occup. Environ. Hyg., 1994, 9, 245. 31 R. W. Hornung, A. L. Grelfe, L. T. Stayner, N. K. Steenland, Suppl. 4, 71. 7 VDI, Indoor Air Pollution Measurement. Measurement of inorganic R. F. Herrick, L. J. Elliott, V. L. Ringenburg and J. Morawetz, Am. J. Ind. Med., 1994, 25, 825. fibrous particles. Measurement planning and procedure. Scanning electron microscopy method, VDI 3492, part 2, Beuth Verlag 32 G.Raw and M. S. Roys, Environ. Int., 1999, 22, 61. 33 K. Andersson, Indoor Air, 1998, Suppl. 4, 32. GmbH, Berlin, 1994. 8 T. Skov and T. S. Kristensen, Am. J. Ind. Med., 1996, 29, 378. 34 M. J. Hodgson, J. Frohliger, E. Permar, C. Tidwell, N. D. Travem, S. A. Ohlenchock and M. Karpf, J. Occup. Med., 1991, 33, 527. 9 E. Kukkonen, E. Ska°ret, J. Sundell and O. Valbjørn, Indoor climate problems. Investigation and remedial measures, NT TECHN 35 D. Wyon, Environ. Tech., 1992, 13, 313. 36 J. Kildesø, D. Wyon, T. Skov and T. Schneider, Scand. J. Work REPORT 204, Nordtest report, Nordtest, Espoo, 1993. 10 W. R. Ott, J. Exp. Anal. Environ. Epidemiol., 1995, 5, 449. Environ. Health, 1999, in the press. 37 R.Fortmann, R. Clayton, V. R. Highsmith and C. J. Nelson, 11 T. Schneider and E. Holst, Occup. Hyg., 1996, 3, 59. 12 W. Ott, L. Wallace, D. Mage, G. Akland, R. Lewis, H. Sauls, Results of the initital survey of randomly selected GSA buildings, Report EPA/600/A-94/157, US Department of Commerce, NTIS, C. Rodes, D. KleVman, D. Kuroda and K. Morehouse, Environ. Int., 1986, 12, 475.Springfield, Virginia, 1994. 38 P. O. Droz, M. Berode and M. M. Wu, Appl. Occup. Environ. 13 M. Nicas and R. C. Spear, Am. Ind. Hyg. Assoc. J., 1993, 54(5), 211. Hyg., 1991, 6, 465. 39 S. A. Roach, Am. Ind. Hyg. Assoc. J., 1966, 27, 1. 14 R. A. Stone, G. Marsh, A. Youk, T. J. Smith and M. M. Quinn, Occup. Hyg., 1996, 3, 91. 40 M. Maroni, B. Seifert and T. Lindvall, Indoor air quality, Elsevier Science B.V., 1995. 15 M. Jantunen, Time activity patterns in exposure assessment, Report number 6, Report EUR 15892 EN, ed. U. Ackerman-Liebrich, G. 41 P.WolkoV, GefahrstoVe-Reinhaltung der Luft, 1996, 56, 151. 42 CEN, Building products-Determination of the emission of volatile Viegi and C. Nolan, European Commission, Brussels, 1995, p. 63. 16 M. Gardiner, Occup. Environ.Med., 1995, 52, 705. organic compounds-Part 2: Emission test cell method, prENV 13419–2, European Prestandard, European Committee for 17 D. Loomis, A. Salvan, H. Kromhout and D. Kriebel, Occup. Hyg., 1999, 5, 73. Standardization, Brussels, 1999. 43 P. A. Clausen, K. Wilkins and P. WolkoV, J. Chromatogr., A, 18 ISO, Indoor air. General aspects of sampling strategy, ISO/TC 1998, 814, 161. 146/SC 6 N 100, 1998, unpublished. 44 T. Schneider, O. Nielsen, P. BredsdorV and P. Linde, Scand. 19 D. T. Mage and W. R. Ott, Models for Predicting Source and Sink J. Work Environ. Health, 1990, 16, 434. Behavior, ed. A. B. Tichenor, ASTM STP 1287, 1996, p. 263. 45 E. J. Lowbury, J. Hyg. Camb., 1950, 48, 1. 20 R. P. Clark and R. N. Cox, in Airborne transmission and airborne 46 R. W. Allott, M. Kelly and C. N. Hewitt, Atmos. Environ., 1994, infection, ed. J. F. Hers and K. C. Winkler, Osthoek Publishing 28, 679. Co, Utrecht, 1973, p. 413. 47 J. Kildesø, J. Vallarino, J. D. Spengler, H. S. Brightman and T. 21 H. Brohus, Personal Exposure to Contaminant Sources in Schneider, Atmos. Environ., 1999, 33, 699. Ventilated Rooms, PhD Thesis, Aalborg University, Aalborg, 48 J. Kildesø, P. Vinzents, N. P. Kloch and T. Schneider, Textile Res. Denmark, 1997. J., 1999, 69, 169. 22 H. O� zkaynak, J. Xue, J. Spengler, L. Wallace, E. Pellizzari and 49 G. W. Royster and B. R. Fish, Surf. Contam., Proc. Symp., ed. P. Jenkins, J. Exp. Anal. Environ. Epidemiol., 1996, 6, 57. B. R. Fish, Pergamon Press, 1967, p. 201. 23 A. V. Baughman, A. J. Gadgil and W. W. NazaroV, Indoor Air, 50 J. W. Roberts, W. T. Budd, D. E. Camann, R. C. Fortmann, 1994, 4, 114. R. G. Lewis, M. G. R. R. Stamper and L. S. Sheldon, Proc. 24 A. C. Drescher, C. Lobascio, A. J. Gadgil and W. W. NazaroV, Ann. Meeting Air Waste Manage. Assoc., Vancouver, June 1991, Indoor Air, 1995, 5, 204. Air and Waste Management Association, Pittsburgh, PA, 1991, 25 S. A. Roach, Ann. Occup. Hyg., 1977, 20, 65. vol. 15, p. 150.2. 26 S. M. Rappaport and R. C. Spear, Ann. Occup. Hyg., 1987, 32, 21. 51 S. A. Ness, Surface and Dermal Monitoring for Toxic Exposures, 27 J. Kildesø, L. Tornvig, P. Skov and T. Schneider, Indoor Air, 1998, Van Nostrand Reinhold, New York, 1994. 8, 12. 52 T. Schneider, O. H. Petersen, A. Aasbjerg Nielsen and K. 28 J. Kildesø, T. Schneider, D. Wyon, P. Skov, P. Pinholt, Windfeld, J. Aerosol Sci., 1990, 21, 555. L. Tornvig, A.-L. Schjønning and H. Skjerbæk, in Design of an 53 S. Selvin and S. Rappaport, Am. Ind. Hyg. Assoc. J., 1989, 50, 627. intervention study to evaluate the eVect of improved cleaning 54 E. D. Pellizzari, K. W. Thomas, C. A. Clayton, R. W. Whitmore, methods, ed. T. Kujala, Finnish Association of Cleaning R. C. Shores, H. S. Zelon and R. L. Peritt, EPA Project Summary. Technology, Helsinki, 1995, pp. 138–142. Particle Total Exposure Assessment Methodology (PTEAM): 29 N. A. Leidel, K. A. Busch and J. R. Lynch, Occupational Exposure Riverside, California Pilot Study, vol. 1, US Environmental Sampling Strategy Manual, DHEW (NIOSH) Publication Protection Agency, Cincinnati, OH, 1993. No. 77–173, US Department of Health, Education, and Welfare, National Institute of Occupational Safety and Health, Cincinnati, OH, 1977. Paper 9/02884F 434 J. Environ. Monit., 1999, 1, 427–434

ISSN:0960-7919    

DOI:10.1039/a902884f

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

Measurements of fractionated gaseous mercury concentrations over northwestern and central Europe, 1995–99† Jonas Sommar,* Xinbin Feng, Katarina Ga°rdfeldt and Oliver Lindqvist Inorganic Chemistry, Department of Chemistry, Go�teborg University, 412 96 Go�teborg, Sweden. E-mail: sommar@inoc.chalmers.se Received 1st April 1999, Accepted 12th July 1999 Although it makes up only a few per cent. of total gaseous mercury (TGM) in the atmosphere, the fraction of oxidised (divalent) mercury plays a major role in the biogeochemical cycle of mercury due to its high aYnity for water and surfaces.Quantitative knowledge of this fraction present in mixing ratios in the parts-per-1015 (ppq) range is currently very scarce. This work is based on #220 data for divalent gaseous mercury (DGM) collected during 1995–99 in ambient air.Over the course of the measurements, the sampling and analytical methods were modified and improved. This is described here in detail and includes transition from wet leaching and reduction procedures to thermo-reductive desorption, the use of annular as well as tubular denuders and adoption of an automated sampling system. The concentration of DGM exhibited a strong seasonal behaviour in contrast to atomic gaseous mercury, with low values in winter and maximum values in summer.The DGM/TGM ratios were frequently found to be below the detection limit (1%) and in the range 1–5%. A trend of diurnal DGM patterns was observed and implies photolytically induced sources. Scavenging of DGM during rain events was also noticed.Dry adsorption methods to trap oxidised mercury in flue Aim of investigation gases rely on the same principle. Oxidised mercury, ‘HgCl2’, Mercury is the only air pollutant predominantly present in is complexed to [HgCl3]- or [HgCl4]2- and retained as the atomic form (Hg0). Owing to high detection sensitivity, it is complex anion in a KCl matrix. The diVusion denuder possible to monitor roughly background sub-ppt concen- method10 separates gases from particles based on the fact that trations in the atmosphere with direct absorption methods.1 they diVuse much faster than particles.For sampling purposes, For oxidised mercury forms present in the parts-per-1015 (ppq) the gas has to be sucked along a surface that acts as a sink range in ambient air, accumulative sampling is required. for the specific gas of interest.Numerous coatings can act as Besides CH3HgX, (CH3)2Hg and Hg0, speciation of gaseous a sink for gaseous divalent mercury compounds but not for mercury is not obtained, only fractionation. Recently, eVorts elemental mercury, such as alkali metal hydroxides, halides have been made to determine the fraction of oxidised (divalent) and chromates.11 The use of KCl is mainly due to its high gaseous mercury in the atmosphere.2–5 Normally, this is rep- deliquescence point.The compound traps DGM compounds resented by reactive gaseous mercury (RGM). The term is, such as HgCl2 and CH3HgCl.2 however, media-dependent and accordingly depends on HgIIa We have previously published a paper on the applicability in the aqueous phase, defined by Brosset.6 In the following, of KCl denuders to sample and determine RGM in ambient divalent gaseous mercury (DGM) will be used instead and air.2 In the present paper, a data set of ambient air RGM, only when it is appropriate will the term RGM be given.Even DGM and TGM samples from 1995 to date is presented. Over though DGM makes up only a small portion of the total the course of the measurements, the method was modified and gaseous mercury (TGM) in the atmosphere, it plays an improved in order to make the sampling less strenuous and important role in the biogeochemical cycle of mercury in the time-consuming and achieve higher time resolution.This environment due to its high solubility in water and easy includes transition from wet leaching and reduction procedures deposition back to terrestrial ecosystems by both wet and dry to thermo-reductive desorption,12 the use of annular as well processes.DGM exhibits dry deposition velocities similar to as tubular denuders and modifications of the technique aimed those of HNO34 while the corresponding figure for Hg0 is at integration and automation.The diVerent techniques are several orders of magnitude lower. Modelling studies over briefly described and commented upon, being used in overlap various spatial scales have shown that even trace amounts of during transition periods. DGM species may control the overall deposition of mercury.7,8 However, further advancement of mercury modelling is limited by the lack of data on the mercury composition in the Description of experimental procedures atmosphere. KCl coated denuders The earliest evidence for the presence of ppq concentrations of RGM in ambient air was presented by Brosset.9 Using a The denuders were made from quartz or borosilicate glass high flow refluxing mist chamber, Stratton and Lindberg3,4 tubes.In addition to tubular denuders (Ø#6 mm), annular were able to obtain diurnal-resolved RGM in ambient air.denuders consisting of two coaxial quartz tubes with #1 mm Representative samples collected corresponded to a detection spacing and with sandblasted annulus walls were also limit of 1 ppq RGM in ambient air. employed. The cleaning of the tubes and application of a methanolic KCl solution to the walls were performed as described previously.2 Both ends were cleaned with doubly †Presented at AIRMON ’99, Geilo, Norway, February 10–14, 1999.J. Environ. Monit., 1999, 1, 435–439 435de-ionised water (Millipore, Milli-Q). The eVective KCl coated always present. The generator eZuent was then diluted to a suitable concentration by adding carrier gas. length was 50 and 20 cm for tubular and annular denuders, respectively. The denuders were repeatedly ‘blanked’ by heating in mercury-free inert gas until low long-term blanks were Sampling set-up obtained.Denuders that were to be exposed were generally The collection eYciency of the divalent species has been shown ‘blanked’ shortly before collection started. to be near-quantitative.2 Therefore, most of the DGM and RGM measurements were performed with single collectors Determination of mercury only.Usually, about 1–1.5 m3 samples were collected in the Divalent gaseous mercury. Mercury trapped in the denuders KCl denuder. The tubular and annular denuders were mainwas determined either by leaching or thermal desorption. tained at a flow of 0.7–1 and 6–9 l min-1, respectively. The Borosilicate denuders were only analysed by the former method denuders were mounted vertically with a downturned funnel whereas denuders made from quartz were subject to testing at the inlet to protect them from precipitation.In about 60% with both methods. The extraction solution, suprapure grade of the sampling events TGM was co-sampled with DGM. HCl (Merck), was diluted to 1.0 M with Milli-Q water and further purified following the procedure previously described2 Manual sampling.The assembly for manual sampling of before being used. Divalent mercury molecules were extracted divalent mercury is given in Fig. 1. Denuders mounted vertias their chloro complexes. The leachates were analysed for cally at least 1 m above ground were generally heated at both RGM and DGM. DGM was obtained after oxidising 40–50 °C to avoid KCl deliquescencing during sampling.By the leachate with BrCl followed by reduction with having a negligible pressure drop, the volume exposed to NH2OH·HCl and eventually with SnCl2 and determined by the denuder was obtained by simply using a gas meter cold vapour atomic fluorescence spectrometry (CVAFS) (Schlumberger) connected downstream.(Brooks Rand CVAFS-2 Mercury Analyzer or PSA Millennium Merlin Mercury Analyser). In the former case, a Automated sampling. An automated sampling system for dual amalgamation step was employed. DGM based on the thermo-reduction desorption principle was The set-up used to determine mercury by thermal desorption also used. The configuration shown in Fig. 2 is generally is described below.During analyor ‘blanking’ of the KCl adopted from the Tekran Model 1130P system (Tekran, denuder, it was heated externally from a NiCr resistance Toronto, Canada). A heated (50 °C) annular denuder with the ribbon. The temperature (<700 °C) and time were set to the eVective KCl coated length enclosed in an oven equipped with grade of contamination of the denuder.By desorbing at 450 °C, a cooling fan was used to collect DGM. The denuder was generally >90% of mercury trapped was released in a few connected with short, cleaned Teflon transfer lines to a pump minutes.12 Samples were generally treated at 450 °C for 10 min. and CVAFS (Tekran 2537A Mercury Vapour Analyzer) unit. Collectors made of borosilicate were also heated to a few The working principle of the CVAFS system has been described hundred degrees to attain low blanks.During sample analysis, by Schro�der et al.14 The surface temperature of the denuder mercury species desorbed were pyrolysed before preconcen- was controlled with a set-point controller (Eurotherm 2416) tration on an analytical column with gold-coated quartz beads including time-regulated heating and cooling plug-ins.The secured with quartz wool plugs. The pyrolyser consisted of flows through the denuder were synchronised with the temperaquartz beads in a quartz column maintained at 900 °C. The ture as well as with the CVAFS unit’s sampling cycles using gold trap was heated in a short pulse to 500 °C and atomic Tekran accessories (1110 Synchronized two port sampling mercury released was detected by CVAFS (Brooks Rand unit, 1120 Standard Addition Controller).This unit was CVAFS-2 Mercury Analyzer or Tekran 2537A Mercury continuously sampling ambient air at 1.5 l min-1. During Vapour Analyzer). The output signal from the detector of the preconcentration of DGM, when an additional gas flow of PSA or Tekran instrument was acquired on a portable #4.5 l min-1 passes through the denuder, the CVAFS unit computer.simply detects Hg0. A few cycles before desorption, the inlet of the denuder was blocked with mercury-scrubbed ambient Total gaseous mercury. TGM was collected on gold filled air. Hence, the detector only experienced the influence of quartz tubes (gold traps) either manually or automatically by a field-portable instrument (Ekoservis Gardis-1A or Tekran 2537A Mercury Vapour Analyzer).The gold traps were analysed as described above. Calibration of the instruments was achieved by injecting certain amounts of elemental mercury into the analytical system. Mercury was delivered as acidic aliquots into the PSA instrument and as nitrogen gas saturated with mercury into the other instruments. The Tekran instrument also exhibits an internal calibration system with a permeation source.Laboratory testing of KCl denuders A detailed description of the HgCl2 generation system has been given previously.2 The generation rate is highly dependent on the temperature and flow of the carrier gas. The parts of the system upstream of the delivery point have to be completely heated to a temperature exceeding that of the bath to avoid condensation.The composition and stability of the source were tested by connection to an on-line fractionation system normally used for flue gases.13 The gas stream was pyrolysed at 600 °C in a quartz cell and detected by Zeeman-eVect electrothermal AAS (Semtech 2000 Mercury Analyzer) at the Fig. 1 Schematic diagram of set-up for manual sampling with denuders.same temperature. A few per cent. of elemental mercury was 436 J. Environ. Monit., 1999, 1, 435–439Fig. 2 Schematic diagram of set-up for automated fractionation of gaseous mercury. The flow indicated by the dark-coloured lines to the left of the denuder is active during preconcentration while the flow indicated by the light-coloured lines acts to block the denuder from ambient air during thermal desorption.mercury zero air. During desorption, the temperature of the denuder was ramped to 450 °C and, consequently, DGM trapped was released, pyrolysed and detected. After the denuder had been allowed to cool down, preconcentration was resumed. The internal blank of the denuder was intermittently checked by passage of mercury zero air during the whole Fig. 3 Locations of European sampling sites. sampling period followed by thermal desorption. In the most recent measurements, deposition of coarse particles on the denuder was prevented by passing ambient air through a Analytical performance of the methods cyclone (2.5 mm aerodynamic diameter cut-oV ). During 1996, two parallel sampling lines were run with dual leaching analysis and the variations were in the range up to Sampling sites 30%.The DGM method blank is generally 15–35 pg and as Air samples were taken manually from 1995 and automatically low as 10 pg can be obtained for RGM if care is taken to since 1998. Sampling was performed during two international avoid every possible contamination. Leaching of denuders intercomparison exercises at Mace Head, County Galway, treated with or without repeated heating after a new coating Ireland, and at Sasseta, Tuscany, Italy, during September 1995 had been applied showed that heating was necessary to eliminand June-July 1998, respectively, involving groups from North ate initial contamination.The use of borosilicate tubes, how- America and Europe. ever, prevented high temperatures from being used.As it was The sampling sites in Sweden were located around the city a component of all the solutions used, distillation of the of Go�teborg (58°N 12° E) situated on the west coast. About suprapure grade HCl from mercury induced by SnCl2 addition 65 samples were exposed at the Brottka�rr site 15 km south- was also found to be important.2 Contribution from reagents west during 1995–97.About 25 samples were collected outside gives the DGM method a 5–20 pg higher blank compared the building of the Department of Chemistry in the urban with that of RGM. The detection limit, based on 3s of the area of Chalmers. The S:t Jo�rgen site is located 15 km north, method blanks, varies between 5 and 15 pg, corresponding to where a combustion simulator is operated with mercury injec- air concentrations of <2 ppq.tion for research purposes. Samples were taken outside the The analytical precision of field samples by the thermal simulator building. Measurements of DGM were also per- method was similar to that described above (5–40% variability, formed within the Mercury Over Europe/Mediterranean triplicate samples, n=7).The regenerative use of thermal Atmospheric Mercury Cycling (MOE/MAMCS) Project. denuders makes the blank smaller and less variable without MOE/MAMCS is a multi-year project set-up in 1998, which influencing the collection eYciency. Typical values of the blank includes air sampling at ten sites in Western and Southern and detection limit were a few picograms for each based on Europe.Some results are reported from the Ro�rvik site located field blanks. The internal blank of the annular denuder 35 km south-west of Go�teborg. was higher, but could be suppressed to 5–10 pg during auto- The locations of the European sampling sites are shown mated sampling with repeated heating cycles. The detection in Fig. 3. limit of the thermal methods corresponds typically to DGM concentrations of <1 ppq.The sampling data were screened from outliers. Those Results and discussion generally encountered resulted from passivation of gold traps during field measurements and mainly influenced TGM (also In the following, the concentrations given were calculated based on the molar weight of atomic mercury; hence, the DGM during automated sampling), giving reduced concentrations of <100 ppq of the former.About 10% of the mixing ratios are slightly overestimated for RGM and DGM. TGM data presented here were averaged to cover RGM/DGM automated and <2% of the manually collected RGM/DGM samples were rejected. sampling periods. J. Environ. Monit., 1999, 1, 435–439 437Table 1 Summary of mixing ratios of RGM, DGM and TGM obtained in ambient air at European sites, 1995–1999.The values in italicte that most of the single samples grouped together were below the detection limit given in Fig. 4 RGM(ppq) DGM(ppq) DGMa/TGM (%) Site Season Method x: ±s n x:±s n x:±s n Brottka�rr All Manual leaching 3.4±2.9 45 4.2±4.0 19 1.7±1.6 32 Summer 4.3±3.1 30 5.0±4.2 15 2.2±1.7 19 Winter 1.5±1.0 15 1.3±0.6 4 0.8±0.7 13 Ro� rvik Winter Manual heating 2.6±2.4 30 1.1±0.9 20 Winter Automated heating 1.0±1.0 34 0.5±0.7 34 Summer Manual heating 3.0±3.1 14 1.7±1.8 13 Summer Automated heating 3.9±2.1 21 2.5±1.3 21 Chalmers Summer Manual leaching 16.0±2.1 3 6.9±2.1 3 Summer Manual heating 5.3±2.5 10 Winter Automated heating 1.0±0.4 13 0.4±0.2 13 S:t Jo�rgen Summer Manual leaching 8.7±5.4b 20b Mace Head Summer Manual leaching 9.1±2.9 4 3.0±0.9 4 Sasseta Summer Manual heating 2.1±0.9b 14b aIn Brottka�rr data, RGM/TGM is displayed.bIncludes duplicate or triplicate exposed samples. Intercomparison with other methods to determine oxidised the samples were analysed for both RGM and DGM. The latter was found to be higher using a one-way analysis of gaseous mercury variance test (F=0.46, P<0.06).RGM and DGM were well The methods currently available to fractionate/speciate correlated with each other (r=0.92, P<0.0001) but not with mercury in ambient air—denuder-based techniques and mist TGM (P=0.21). As shown in Table 1, the diVerence was more chambers—were intercompared during the start-up phase of pronounced during the summer period.The RGM fraction is the MOE/MAMCS project in Sasseta. Owing to the scarce considered to represent inorganic compounds, since even the data set, it was diYcult to rank the diVerent methods. However, most stable inorganic mercury compounds in aqueous solution it can be concluded that they generally produce relatively have been found to be easily reduced by SnCl2. On the other similar results with variability up to 30–40%.The complete hand, CH3HgCl(aq) is not reduced by SnCl2 and does not data set including TGM and particulate Hg data will be belong to the RGM fraction.15 This implies the presence of published elsewhere.15 The results obtained by manual tubular gaseous monomethylated mercury species in ambient air. and automated annular denuders were intercompared during However, other methods which can exclusively identify such one campaign at Ro� rvik.The data were significantly correlated compounds in air (mist chambers15 and graphitised carbon (P<0.10, n=9) but diVered occasionally in magnitude (within traps16,17) have to be used during such speciation. a factor of 2). The results obtained with annular denuders The concentration of oxidised gaseous mercury exhibited a were consistently lower.strong seasonal behaviour (P<0.01) with low values in winter and maximum values in summer. The concentration during Concentrations of RGM and DGM in air the winter period often dropped below the detection limit (1–2 ppq). This trend has also been observed by Lindberg In Table 1, the complete set of over 200 data is categorised and Stratton4 and was attributed to seasonal diVerences in into location, season and analytical method.The most extensair stagnation and chemical kinetics. Extensive continental ive data set is from Brottka�rr and covers three years, and is European TGM measurements imply that Hg0 is predomi- displayed in Fig. 4. Exposed denuders were analysed after nantly sink- rather than source-modulated,6,9,18–20 which is manual leaching.RGM was determined throughout the whole characteristic of atmospheric trace gases removed by oxidation campaign while DGM was determined occasionally. Samples processes. Higher oxidant concentrations during summer lead were usually exposed on a diurnal basis with a flow rate of to faster oxidation and to a summer minimum of TGM.The about 700 ml min-1. The average level of DGM measured intensified oxidation process contributes to increased concen- was 4.2 ppq compared with 3.4 ppq for RGM. Nineteen of trations of DGM and subsequently enhanced deposition21,22 of the highly surface-reactive oxidised mercury forms. However, the role of diVerent photolytically induced gas-phase oxidants has not been shown in any detail.23 In its extreme manifestation, when high concentrations of oxidants build up during polar sunrise in the Arctic, fast depletion of elemental gaseous mercury occurs.24 During winter periods, the frequency of precipitation increases from 20 to 40% and leads to enhanced scavenging of oxidised mercury.RGM/DGM was measured during three rainfall events during summer 1996 and indeed oxidised gaseous mercury was found to decrease on average by #60% for both fractions.TGM showed no consistent variation during the same period of time. The limited amount of oxidised mercury collected on the tubular denuders did not allow any diurnal resolution of the measurements. The introduction of high-flow annular denuders enabled the same amount to be collected during a few hours.Fig. 5 shows automated 6 h measurements of DGM and Hg0 at the Chalmers site during Fig. 4 Time–concentration behaviour of RGM/DGM at Brottka�rr the shift between January and February 1999 (a data series during 1995–97. The bars given indicate detection limit for each fraction. without any significant passivation of gold traps). 438 J. Environ. Monit., 1999, 1, 435–439definitely not the case for RGM/DGM. As it is being measured extensively within the MOE/MAMCS campaign and elsewhere, quantitative knowledge of RGM/DGM is, however, likely to be extended in the near future.Acknowledgements This study was initiated by Zifan Xiao, whom we greatly acknowledge together with Shiqiang Wei. This research was a contribution to the ‘Mercury species over Europe’ project being carried out in the specific RTD program sponsored by the European Community under contract number ENV4-ET97–0595.References Fig. 5 Fractionation of mercury in air samples at Chalmers during 1 H. Edner and S. Svanberg,Water, Air Soil Pollut., 1991, 56, 131. Jan.–Feb. 1999. 2 Z. Xiao, J. Sommar, S. Wei and O. Lindqvist, Fresenius’ J.Anal. Chem., 1997, 358, 386. 3 W. J. Stratton and S. E. Lindberg, Water, Air Soil Pollut., 1995, 80, 1269. 4 S. E. Lindberg and W. J. Stratton, Environ. Sci. Technol., 1998, 32, 49. 5 R. K. Stevens, F. A. Scha�dlich, D. R. Schneeberger, E. M. Prestbo, S. E. Lindberg and G. Keeler, Book of Abstracts, 5th International Conference on Mercury as a Global Pollutant, Rio de Janeiro, May 23–28, 1999, ed.J. P. Barbosa and R. Melamed, CETEM—Center for Mineral Technology, Rio de Janeiro, 1999, p. 7. 6 C. Brosset, Water, Air Soil Pollut., 1987, 34, 145. 7 G. Petersen, J. Munthe and R. Bloxam, in Global and Regional Mercury Cycles: Sources, Fluxes and Mass Balances, ed. W. Baeyens, R. Ebinghaus and O. Vasiliev, Kluwer, Dordrecht, 1996, pp. 191–217. 8 G. Petersen, Belastung von Nord- und Ostee durch O� kologisch gefa�hrliche StaVe am Beispiel atmospha�rischer Quecksilberverbindungen, GKSS Forschunzentrum Geesthacht, 1992. 9 C. Brosset, Water, Air Soil Pollut., 1982, 17, 37. 10 K. Larjava, J. Reith and D. Klockow, Int. J. Environ. Anal. Chem., 1990, 38, 31. 11 V. Siemens, personal communication. 12 X. Feng, J. Sommar, K. Ga°rdfeldt and O. Lindqvist, Fresenius’ Fig. 6 Diurnal variation in automated DGM samples at Ro� rvik J. Anal. Chem., submitted. during 03/05/99–09/05/99. 13 X. Feng, M. Abul-Milh, J. Sommar, D. Stro�mberg and O. Lindqvist, unpublished work. 14 W. H. Schro� der, G. Keeler, H. Kock, P. Roussel, D. Schneeberger Lindberg and Stratton4 reported higher TGM figures as and F. Scha�dlich, Water, Air Soil Pollut., 1995, 80, 611.well as RGM/TGM ratios at temperate latitudes influenced 15 J. Munthe, personal communication. by point sources. High RGM/TGM means similar to those 16 N. S. Bloom and W. F. Fitzgerald, Anal. Chim. Acta, 1988, 208, reported in the literature4 were obtained in our observations 151. aand S:t Jo� rgen. The main local 17 J. Sommar, X. Feng and O. Lindqvist, Appl. Organomet.Chem., 1999, 13, 441. source of DGM in Go�teborg is a municipal waste incinerator. 18 A° . Iverfeldt, J. Munthe, C. Brosset and J. Pacyna, Water, Air Soil Diurnal cycles with statistically high significance, a factor of Pollut., 1995, 80, 227. >3 between extreme figures, were reported in the US data.4 19 F. Slemr, in Global and Regional Mercury Cycles: Sources, Fluxes A similar pattern, shown in Fig. 6, was observed at Ro� rvik and Mass Balances, ed. W. Baeyens, R. Ebinghaus and O. Vasiliev, during one week in the early summer with prevailing clear sky Kluwer, Dordrecht, 1996, pp. 33–84. conditions. Our limited winter data under conditions without 20 F. Slemr and H. E. Scheel, Atmos. Environ., 1998, 32, 845. 21 A. Jensen and A° . Iverfeldt, in Mercury Pollution: Integration and precipitation (>75% of a period) did not reveal any highly Synthesis, ed. C. J. Watras and J. W. Huckabee, Lewis Publishers, significant diurnal variation. Boca Raton, FL, 1994, pp. 221–229. The impact of DGM on the deposition of oxidised mercury 22 M. Høyer, J. Burke and G. Keeler, Water, Air Soil Pollut., 1996, based on precipitation scavenging has been assessed.7 Direct 80, 199. scavenging of a mixing ratio of 5 ppq DGM gives wet 23 C.-J. Lin and S. O. Pehkonen, Atmos. Environ., 1999, 33, 2067. deposition concentrations typical for south-west Sweden. As 24 W. H. Schro� der, K. G. Anlauf, L. A. Barrie, J. Y. Lu, A. SteVen, D. R. Schneeberger and T. Berg, Nature (London), 1998, 394, 331. it is highly surface-active, DGM is also likely to influence areas of vegetation significantly by dry deposition. The current database of TGM is now fairly well established, which is Paper 9/02729G J. Environ. Monit., 1999, 1, 435&ndash

ISSN:0960-7919    

DOI:10.1039/a902729g

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

Determination of acrylic acid in air by using diVusion denuder tubes combined with HPLC technique R. Zanella,† M. Schilling and D. Klockow* ISAS—Institut fu�r Spektrochemie und Angewandte Spektroskopie, Bunsen-KirchhoV-Str. 11, D-44139 Dortmund, Germany Received 8th June 1999, Accepted 26th August 1999 A procedure for the determination of atmospheric acrylic acid in air at the mg m-3 to mg m-3 level is described.DiVusion-based sampling, designed to discriminate gaseous analytes from their particulate counterparts, has been used. Acrylic acid is collected with an eYciency of >98% in tubular denuders coated with sodium hydroxide–barium hydroxide–hydroquinone monomethyl ether, and analyzed by high performance liquid chromatography with UV absorbance detection. The detection limit is 2.9 mg m-3 at a flow rate of 0.5 L min-1 and 30 min sampling time. A precision ( p=0.95, n=10) of 7.5% of the overall procedure was achieved at the 100 mg m-3 level.Results of laboratory studies concerning the eVect of the coating reagent and the relative humidity on the sampling eYciency as well as possible interferences, in particular by ozone, and the elimination of these interferences are discussed.This method was developed to monitor workplace atmospheres as well as ambient air in industrial environments. while particles pass the denuder without essential losses Introduction because of their dramatically lower diVusion coeYcients. Acrylic acid (2-propenoic acid) and its alkyl esters are of great In this paper the use of tubular denuders coated with an industrial importance for the production of a wide range of appropriate alkaline reagent and a polymerization inhibitor9 polymers, lacquers, coatings, and resins.1–3 The acid is known is described as a collection method for acrylic acid from as an acute local irritant, and exposure to its vapors can cause ambient air.eVects to respiratory tract, skin and eyes.4,5 The American Conference of Governmental Industrial Hygienists (ACGIH) has adopted a threshold limit value for an 8 h time-weighted Methods and materials average of 6 mg m-3 (2 ppmv) for acrylic acid.6 Therefore, HPLC apparatus permanent control of work place air quality in relevant industries is essential due to health reasons. The high-performance liquid chromatograph (Model FR-30, One important aspect in air measurements is to obtain Knauer, Berlin, Germany) was equipped with an injector reliable information about the gas-particle distribution of (Model 7125, Rheodyne, Redwood, CA, USA), a variable organic air pollutants because of the diVerent depositional wavelength UV-VIS detector (Model PU 4020, Philips, behaviour of gaseous and particulate compounds.Because of Eindhoven, The Netherlands) and a recorder (BBC Model Se its relatively high vapor pressure at ambient temperatures, 120, supplied by Riedl, Traunreut, Germany).An analytical acrylic acid can be expected to occur mainly in the gas phase, column (Nucleosil 100 C18 5 mm [250+40]×4 mm id, but the absorption of the compound on other atmospheric Machery-Nagel, Du� ren, Germany) was used.An aqueous particulate material seems to be possible too. solution of phosphoric acid (1.5 mmol L-1) was used as Up to now, airborne acrylic acid and acrylates were mobile phase at a flow rate of 0.8 mL min-1. determined by preconcentration on activated charcoal,7 polymer material,8 or coated silica gel,9 followed by extraction and Preparation of denuder chromatographic analysis.These preconcentration methods, however, are not able to distinguish between gaseous and Denuder tubes were made from brown colored borosilicate particulate compounds. Therefore, a denuder sampling system glass, in order to protect the sampled acrylic acid against was applied. polymerization by sunlight. Prior to coating, the tubes of Because of their simplicity, tubular denuders have been used 600 mm length and 6 mm id were pretreated with acetone, for many years in a variety of applications related to sampling rinsed with doubly distilled water, then treated with 40% HF, and analysis of ambient air pollutants.10–15 Most commonly rinsed again with doubly distilled water and finally dried at they are applied to remove selected vapor phase components 80 °C for 1 h.An alkaline coating was applied by introducing from an air stream while particles and other vapor components 1 mL of the coating reagent solution (40.0 mg sodium are transmitted to subsequent stages of the sampling device. hydroxide–315.5 mg barium hydroxide octahydrate–2.0 mg Generally, the denuder technique is based on the diVusion of hydroquinone monomethyl ether (HQMM) in 100 mL doubly gaseous species to the inner wall of the denuder tube, when distilled water) into each tube, which was slowly rotated and air is passed through the tube under laminar flow conditions, inclined to ensure complete wetting of the walls.The tubes were then rotated continuously in a horizontal position at 80 °C and dried with a stream of clean nitrogen in a specially †Present address: Departamento de Quý�mica da UFSM, 97119-900 Santa Maria-RS, Brazil.designated apparatus, to obtain a smooth coating. J. Environ. Monit., 1999, 1, 441–443 441Preparation of ozone scrubber in front of the sampling device. During sampling, the denuder arrangement was placed in vertical position to avoid gravi- AMnO2 coating was deposited on the walls of tubes of 60 mm tational losses of particles in the tubes.The ozone scrubber length and 6 mm id by injection of 1 mL potassium permanga- and denuder can be used only once. nate solution 0.1 mol L-1 and of 1 mL manganese sulfate solution 0.15 mol L-1 into each tube, which was manually Sample extraction and analysis rotated to obtain a film of manganese dioxide.The solution After air sampling, the denuder and the scrubber tube were was discarded and the procedure repeated. The tube was extracted together with 1.5 mL of 0.01 mol L-1 KH2PO4 solu- finally rinsed with doubly distilled water and dried at 80 °C. tion, adjusted to pH 2.5 with phosphoric acid. The particle Denuders and scrubbers thus prepared were sealed at the filter was extracted with 1 mL of the same phosphate buVer ends with Eppendorf tubes and stored until they were used. in an Eppendorf tube.All extracts were centrifuged for 10 min and finally analyzed by RP HPLC with UV detection at Generation of acrylic acid test gas 195 nm under the following conditions: column, Nucleosil Test gas mixtures with known concentrations of acrylic acid 100–5 (250×4 mm); eluent, H3PO4 (1.5 mmol L-1); flow rate, in nitrogen as a carrier gas were generated by employing a 0.8 mL min-1.permeation technique.16,17 A glass tube filled with pure acrylic acid was capped with a porous PTFE membrane (active Testing of ozone interferences surface 95 mm2, thickness 2.0 mm, porosity 1 mm). This tube Ozone was generated by a simple device, in which synthetic was stored in a double-walled vessel thermostated at 21 °C air (flow rate 0.2–2 L min-1) was irradiated with UV light and flushed permanently with pure nitrogen (10–100 L h-1), (Hg lamp).The concentration of ozone in the test gas was which picked up the the vapour of acrylic acid permeating adjusted by the air flow rate and the intensity of the UV light, through the PTFE membrane.The permeation rate under the and was measured by an ozone monitor (DASIBI, 1008 ozone given experimental conditions was measured using an impinger analyzer, Environment Corp., Glendale, CA). filled with aqueous 0.01 mol L-1 NaOH and was found to be 174±5 ngmin-1. The gas phase concentration could be adjusted via the flow rate of the carrier gas. Results and discussion Sampling eYciency Air sampling The sampling eYciency was calculated as: Air samples were collected with a sampling train consisting of an ozone scrubber, a denuder with alkaline coating and a E (%)=(nd/nTG)×100 PTFE filter of 37 mm diameter inserted into a PTFE filter where nd=the amount of acrylic acid, determined in the holder at the end, as shown in Fig. 1. Connections were made denuder tube and nTG=the amount ylic acid, released with PTFE couplers.Air was sucked through the arrangement by the test gas source during the collection period. Because of by a membrane pump. The flow rate was set to 0.5 L min-1 the acidic character of the analyte, an alkaline coating was by a critical orifice and controlled by a gas meter. To ensure necessary to achieve a quantitative and irreversible absorption a laminar flow at the entrance of the first denuder tube, a of acrylic acid in the denuder tube.Investigations were made conical PTFE inlet (id, 6 mm; length, 50 mm) was mounted with two alkaline substances (NaOH; Ba(OH)2) as coating material. (1) Sodium hydroxide: a tube coated with pure NaOH showed a practically quantitative absorption eYciency (>98%).Because of the hygroscopic character of NaOH, however, the mechanical stability of the coating film got worse with increasing relative humidites. (2) Barium hydroxide: this coating showed a good stability even at high relative humidities and long sampling times, but the sampling eYciency was lower (~90%) compared to a NaOH denuder. The best compromise was a mixture of NaOH and Ba(OH)2.An amount of 10 mmol of each base per denuder was suYcient for a highly eYcient absorption (>98%). Such a denuder showed no changes of the coating during a sampling period of 4 h at a relative humidity of >95%. Its capacity allows the preconcentration of 1.5 mmol (108 mg) of acrylic acid with a high eYciency at a flow rate of 0.5 L min-1. To avoid polymerization of the collected acrylic acid through the influence of light or radicals, hydroquinone monomethyl ether (HQMM) was added to the coating (20 mg per denuder) as an inhibitor.This reagent had no influence, either on the preconcentration or on the chromatographic analysis. Interferences Possible interferences of the determination of acrylic acid in air can be as follows. (1) Reaction of collected acrylic acid with atmospheric ozone, which can be found in ambient air under certain pollution and solar radiation conditions in concentrations higher than 100 ppbv.Fig. 1 Scheme of the sampling arrangement. a, Denuder inlet; b, The occurrence of acrylic acid oxidation due to ozone during ozone scrubber; c, connection; d, coated denuder; e, filter holder; f, critical orifice; g, membrane pump; and h, gas meter.sampling was investigated by passing first acrylic acid test gas 442 J. Environ. Monit., 1999, 1, 441–44315 to 30 min. After sampling the denuders were closed with Eppendorf tubes and analysed in the laboratory afterwards. In all cases, the concentrations of acrylic acid found were far below the threshold limit adopted by the ACGIH.Back up filters were stored in Eppendorf tubes until extraction and analysis; in all filter samples acrylic acid could not be detected. Conclusions This paper describes a method for the determination of airborne acrylic acid at the mg m-3 to mg m-3 level. A tubular denuder tube with alkaline coating is used to preconcentrate the gaseous acid, while possible particulate counterparts are collected on a back-up filter.After extraction of the tubes (ozone scrubber and denuder) and the back-up filter, the acid Fig. 2 Acrylic acid fraction in the denuder remaining unchanged after is determined by HPLC with UV detection. exposure to diVerent ozone concentrations. Because of the simple construction and preparation of the denuder tube assembly the method appears suitable for monitoring workplace atmospheres as well as ambient air.It can (347 mg m-3) and then ozone air mixtures (20–200 mg m-3 of easily be used for control of the threshold limit set by O3) through a denuder, each flow for 15 min at 0.5 L min-1. the ACGIH. The results of these experiments are shown in Fig. 2. As it could be expected, the losses of acrylic acid increased with increasing ozone concentration.Acknowledgement To avoid this interference, a MnO2 coated glass tube (id 6 mm; length, 60 mm) was placed in front of the denuder. This R. Zanella thanks the DAAD, Germany, and CNPq, Brazil, scrubber destroys the ozone very eVectively. However, a small for providing a scholarship. Financial support by Hu� ls AG, amount of acrylic acid (~5%) is retained in this scrubber; Marl, Germany, is also gratefully acknowledged.therefore, it is necessary after each sampling to extract the scrubber together with the denuder tube for final analysis. References (2) Formation of acrylic acid in the denuder because of alkaline hydrolysis of absorbed acrylic acid esters. 1 S. S. Cutie, G. J. Kallos and P. B. Smith, J. Chromatogr., 1987, The hydrolysis of acrylic acid esters at the alkaline coating 408, 349. 2 A. B. Kinney and A. B. Scranton, Polym. Mater. Sci. Eng., 1993, of the denuder was investigated by passing an ester test gas 69, 487. mixture corresponding to absolute amounts of 54.6 nmol 3 T. P. Davis, Plast. Eng. (N. Y.), 1997, 41, 203. methyl acrylate, 50.7 nmol ethyl acrylate, 45.3 nmol isobutyl 4 J. W. Nemec and W.Bauer, in Kirk-Othmer Encyclopedia of acrylate, 49.2 nmol butyl acrylate, and 44.7 nmol 2-ethylhexyl Chemical Technology, Interscience, New York, 3rd edn., 1978, acrylate for 10 min through the denuder arrangement, vol. 1, pp. 330–354. extracting the denuder, and analysing the extract. 5 H. Abe, T. Yoshimura, S. Kanaya, Y. Takahashi, Y. Miyashita and S. I. Sasaki, Anal.Chim. Acta, 1987, 194, 1. It was found, that 1–2% of the total amount of esters 6 ACGIH Threshold Limit Values and Biological Exposure Indices applied could be determined as acrylic acid. This interference for 1991–1992, ACGIH, Cincinnati, OH, 1991. can be tolerated as long as the concentrations of acrylic acid 7 S. Czerczak, Chem. Anal. (Warsaw), 1983, 28, 35. and of the acrylic acid esters appear in the same order of 8 K.Ventura, P. Prihoda and J. Churacek, J. Chromatogr., 1995, magnitude. 710, 167. 9 W. J. Vincent and V. Guient, Jr., Am. Ind. Hyg. Assoc. J., 1982, 43, 499. Statistical characterization of the described method 10 R. Niessner and D. Klockow, Int. J. Environ. Anal. Chem., 1980, The detection limit achieved at 30 min sampling time and 8, 163. 11 E. E. Lewin and K. A. Hansen, Anal. Chem., 1984, 56, 842. under the experimental conditions described above was 12 D. J. Eathough, V. F. White, L. D. Hansen, N. L. Eathough and 40.3 nmol m-3 (2.9 mg m-3) of acrylic acid. The precision E. C. Ellis, Anal. Chem., 1985, 57, 743. ( p=0.95, n=10) of the overall procedure (sampling and 13 R. S. Braman, M. A. de la Cantera and Q. X. Han, Anal. Chem., analysis) was 7.5% at 114 mg m-3 acrylic acid. The calibration 1986, 58, 1537. curve is linear up to 120 nmol per denuder (576 mg m-3 at 14 D. M. Murphy and D. W. Fahey, Anal. Chem., 1987, 59, 2753. 30 min sampling time). 15 S. Sioutas and P. Koutrakis, in The Handbook of Environmental Chemistry, Vol. 4D: Airborne Particulate Matter, ed. O. Hutzinger; Springer, Berlin-Heidelberg-New York, 1995, pp. 201–231. Application 16 J. Namiesnik, J. Chromatogr., 1984, 300, 79. 17 M. Schilling, PhD Thesis, University of Dortmund, 1988. Field investigations with the method described were carried out at diVerent locations in the vicinity of an industrial plant for the production of acrylates. Sampling times varied from Paper 9/04577E J. Environ. Monit., 1999, 1, 441–443 4

ISSN:0960-7919    

DOI:10.1039/a904577e

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

Determination of selected microbial volatile organic compounds by diVusive sampling and dual-column capillary GC-FID—a new feasible approach for the detection of an exposure to indoor mould fungi? Katja Elke,a Jutta Begerow,*a Hanna Oppermann,b Ursula Kra�mer,a Erich Jermanna and Lothar Dunemanna aMedizinisches Institut fu�r Umwelthygiene, Auf ’m Hennekamp 50, D-40225 Du�sseldorf, Germany bHygieneinstitut Sachsen-Anhalt, Wallonerberg 2/3, D-39104 Magdeburg, Germany Received 16th April 1999, Accepted 5th July 1999 A new, analytically valid procedure is described to assess the exposure of human beings to the so-called microbial volatile organic compounds (MVOCs) in air.The method can be used routinely for large sample numbers and is especially valuable as a basis for further research on the correlation between single MVOCs and indoor mould growth. The procedure is based on the fact that fungi produce a variety of volatile organic compounds, such as 3-methylbutan-1-ol, 3-methylbutan-2-ol, fenchone, heptan-2-one, hexan-2-one, octan-3-one, octan-3-ol, pentan- 2-ol, a-terpineol, and thujopsene, which they emit into the indoor environment.Using diVusive samplers, these MVOCs are adsorbed onto charcoal during a sampling interval of four weeks.The described method is thus superior to existing methods which use short-term active sampling. After desorption with carbon disulfide, the MVOCs were determined by dual-column gas chromatography with flame ionization detection using the largevolume injection technique for sample introduction.The detection limits ranged between 0.15 and 0.53 mg m-3, within-series precision was found to range between 6.5 and 19.0%, and recovery was between 77 and 118%. The procedure has been sucessfully applied in the context of a large field study to measure the indoor MVOC exposure in children’s rooms of 132 dwellings. The objective of the study was to examine the relation between indoor mould growth, the indoor MVOC exposure and the prevalence of adverse health eVects.Information about mould formation has been obtained by a questionnaire and by the determination of colony forming units of mould fungi in mattress dust. With the exception of 3-methylbutan-2-ol, fenchone, nonan-2-one, octan-2-one, and thujopsene, indoor air concentrations of all MVOCs under investigation were significantly higher inside damp and mouldy dwellings.From the primary MVOCs under investigation, 3-methylbutan-1-ol, hexan-2-one, heptan-2-one, and octan-3-ol were found to be most reliable indicators for mould formation. A correlation was also found between selected MVOCs and the occurance of mould species in mattress dust. Aspergillus sp. correlated with heptan-2-one, hexan-2-one, octan-3-ol, octan-3-one, and a-terpineol, while the occurrance of Eurotium sp.was correlated with higher indoor air concentrations of 3-methylbutan-1-ol, 3-methylbutan-2-ol, heptan-2-one, hexan-2-one, octan-3-ol, and thujopsene. Children living in dwellings with elevated MVOC levels had a higher prevalence of asthma, hay fever, wheezing, and irritations of the eyes.These positive associations persisted after controlling for confounding factors such as age, sex, body-mass index, number of siblings, social status, passive smoking, type of heating, and ventilation habits. However, they were not statistically significant. This lack of significance may be a result of the small number of investigated samples. Swab samples can be taken, if mould growth is visible on Introduction building materials or furniture.Dust precipitate samples can Mould fungi and dampness in houses are a common problem. be obtained with a vacuum cleaner and filters,8 which can be Constructional faults and low ventilation rates lead to high collected from the surface of furniture, from the floor, from humidity favouring mould growth inside such buildings.carpets, or mattresses. The isolated quantities of mould fungi Numerous investigations indicate that mould fungi may play largely depend on the power of the vacuum cleaner, the an important role as indoor allergens, and that they may be characteristics of the used filter type, from the presence or responsible for the ‘sick building syndrome’ (SBS). Possible absence of humidity, and on the temperature inside the room.health implications caused by fungi are respiratory health The material of which the mattress is made, the age of the problems and irritations of skin and eyes.1–6 mattress, the culture conditions and the kind of the culture The exposure to mould fungi in indoor environments is medium have also an eVect on the species composition.The traditionally assessed microbiologically by air, dust precipitate mould concentration in dust samples is less aVected by fluctu- or swab sampling with subsequent incubation of the exposed ations in the course of the day and by cleaning and other Petri dishes followed by quantification and identification of physical activities inside the dwelling than the concentration the colony-forming units (CFU).The CFU is defined as the of aerial spores.8 In living rooms equipped with carpets Hunter number of colonies formed under defined conditions on or in et al.,9 for example, found an increase in the aerial spore a defined culture medium. burden up to a factor of 17 following cleaning activities. Dust For air sampling diVerent samplers are available, most often Andersen samplers or Reuter centrifugal samplers are used.7 precipitate samples are thus more suitable to assess the chronic J.Environ. Monit., 1999, 1, 445–452 445fungal burden, because they are more independent from the used synthetic media, which are not comparable with real-life conditions typically found inside buildings. Pasanen et al.10,25 time of sampling.8 According to Beguin8 some fungal species cannot be detected in mattress dust, which is attributed to the and Sunesson et al.16 determined MVOCs of fungal species which were cultivated on diVerent building materials as sub- adverse conditions in the matrix.On the other hand, some species can be more frequently detected in mattresses than in strates. They found diVerences in the relative proportions of volatile metabolites cultivated on various media, indicating air.8 Since everyone is exposed to mattress dust for several hours per day, it is a suitable matrix to estimate the extent of the diYculties existing for identification of mould fungi growing on a natural substrate with an undefined composition. fungal exposure.The already mentioned microbiological methods work quite In this paper a new, simple and analytically reliable procedure is described for the determination of MVOCs using well, if the spore source is in direct contact with the indoor air, but fail, if the microorganisms grow behind wallpapers long-term passive sampling, solvent extraction, and dualcolumn capillary GC with flame ionization detector (FID).and covers, because the spores cannot penetrate through such materials.Additionally, mould can only be detected during The described procedure was applied to the determination of MVOCs in dwellings of persons who reported mould growth the sporulation period. The microbiological techniques used to quantify fungi in indoor environments are thus diYcult to and/or dampness and in control dwellings.The aim of this study was to select characteristic MVOCs which are a reliable standardize.1 A relatively new approach is the determination of indicator of indoor mould growth. Mould formation has been recorded by a standardized interview and by the determination microbially produced volatile organic compounds as an indicator for mould growth inside buildings.10–16 Fungi produce of CFU of mould fungi in mattress dust.a wide range of volatile products, including alcohols, aldehydes, ketones, hydrocarbons and esters. Some of them are specific for microorganisms, but some can also be emitted Experimental from other sources such as building materials, furniture, paints, Passive sampling adhesives, essential oils, and other household articles. Thus, extensive and systematic studies are needed on the selection Indoor air sampling was carried out with diVusiveers of characteristic indicator MVOCs.First proposals were made (3500 OVM, 3M, Neuss, Germany) according to a procedure by Dewey et al.,12 Sagunski13 and Keller et al.17 In indoor air, which has been already described in former papers.26,27 The oct-1-ene-3-ol, hexan-2-one and heptan-2-one were found to samplers were placed inside the children’s bedroom at 1.5 m be the most reliable markers.12,13 These authors12,13,17 used height and at a distance of at least 1 m from the walls to analytical procedures for the determination of MVOCs in guarantee a suYcient air circulation (see Field studies).After indoor air which were similar to those for the determination an exposure interval of 4 weeks, the monitors were closed with of environmentally relevant anthropogenic volatile organic an impermeable cap.They were shipped to the laboratory compounds (VOCs) such as benzene, toluene, and o-, m-, p- protected with pads of activated charcoal to prevent them xylene. The MVOCs are adsorbed onto adsorbents such as from contamination and were then stored at +4 °C until Tenax or activated charcoal by active short-term sampling analysis.techniques.11,12,16,17 After solvent extraction or thermal desorption the MVOCs were determined by gas chromatogra- Active sampling phy (GC) with mass spectrometric detection (MSD).11,12,17 According to Keller et al.17 thermodesorption was found to The results obtained with the passive sampling method were be 100 times more sensitive than solvent extraction followed validated by comparison with active sampling.According to by liquid injection of an aliquot of the extract, but has the NIOSH28 charcoal tubes (GK 26–16, SKC, Eighty Four, PA, disadvantage that each sample can only be analyzed once. USA) containing 800 mg of charcoal in the sample section When applying short-term sampling for the determination of and 200 mg in the back-up section were used in combination MVOCs, it is also possible that the extent of mould formation with a sampling pump model P-4000 (Dupont, Wilmington, is severely over- or underestimated.18 This problem could be DE, USA) which was operated at a constant flow of only overcome by extending the sampling period, which is 25 ml min-1.diYcult to achieve with active sampling techniques. Long-term sampling is a domain of passive sampling techniques. Sample preparation In contrast to the determination of the colony-forming units, MVOCs which the fungi produce during their metabolism, Before use, all materials and chemicals coming into contact with the samples or standards were continuously checked for can be detected even if no visible mould occurs.The volatile metabolites are able to diVuse through the wallpaper and contamination. Glass vials and septa were heated for 24 h at 150 °C and 80 °C, respectively, and stored in a clean bench other materials which cannot be penetrated by fungal spores. Thus, both methods complement each other. Additionally, equipped with activated charcoal filters (Bleymehl, Ju� lich, Germany) before use.MVOC production is restricted to the growth period of the fungi whereas the number of CFU can be determined only After exposure, the MVOCs were desorbed with 1.5 ml carbon disulfide (CS2, for the analysis of aromatic hydro- during sporulation. Due to the dependency of spore release on temperature and humidity,19,20 the number of CFU thus carbons, Promochem, Wesel, Germany) containing 1% methanol (p.a., Riedel-de Hae�n, Seelze, Germany).After addition depends on the climatic conditions inside the dwellings.9 Hunter et al.9 found that domestic activities in a room have of the solvent, the monitors were closed and mechanically shaken for 30 min. The extract was then decanted into GC also a great influence on spore burden in indoor air.Classification of fungi species can be performed by autosampler glass vials and analysed.26,27 The sample and back-up sections of the charcoal tubes microscopic examination. In contrast to this, the identification of single fungal species by the determination of the MVOC which were used for active sampling were separately removed from the tubes and extracted with 3 ml and 1.5 ml CS2 pattern is not possible, because the produced MVOC pattern is not only dependent on the fungal species but also on the (containing 1% of methanol ), respectively, by mechanical agitation for 60 min.The extract was centrifuged (5 min, substrate (chemical composition, pH, water content, temperature). 16,19,20 Several authors identified diVerences in the metab- 4000 rpm), then decanted into glass vials and analysed.A breakthrough was not observed in any case.28 olite pattern for diVerent types of mould fungi,21–24 but they 446 J. Environ. Monit., 1999, 1, 445–452Table 1 Calculation constants A for the determination of selected Capillary gas chromatography MVOCs by passive sampling The gas chromatographic separation of the MVOCs and the Calculation constant detection were carried out with a dual-column HRGC 5300 MVOC A/1000 min ml-1 (Carlo Erba, Hofheim/Taunus, Germany) gas chromatograph equipped with two flame ionization detectors (FIDs).The 3-Methylbutan-1-ol 32.29 dual-column system which was used for this application has 3-Methylbutan-2-ol 31.97 been already described in detail in a former paper.27 In short: Fenchone 44.57 the carrier gas flow (helium, purity 5.6, flow: 2 ml min-1) was Heptan-2-one 36.01 Hexan-2-one 33.45 led through a precolumn (2.5 m×0.32 mm id methyl silicone Nonan-2-one 40.79 coated capillary column, Chrompack) and split via a YOctan- 3-ol 37.86 connector (SGE, Weiterstadt, Germany) at a ratio of 151 onto Octan-3-one 38.52 two columns of diVerent polarity, a 60 m×0.32 mm id Pentan-2-ol 32.13 DB-1701 (14% cyanopropyl phenyl silicone, 1 mm film a-Terpineol 48.71 thickness, J&W Scientific, Cologne, Germany) and a Thujopsene 58.43 60 m×0.32 mm id DB-5 (5% phenyl, 95% methylsilicone, 1 mm film thickness, J&W Scientific).Each column was connected to an FID which was held at 330 °C. The following temperature Detection limits and reliability criteria program was used for both columns: 35 °C for 5 min, ramp at 4 °Cmin-1 to 200 °C, which was held for 30 min.It is usual practice that detection limits are calculated as the Sample injection was controlled by a temperature three-fold standard deviation of replicate measurements of the programmable multi-injector MFA 515 (Carlo Erba).The analytical blanks. In this case, this procedure was not possible, sample was injected by the splitless large-volume technique because the MVOC blank values were too low to be registered. (sample size: 10 ml ) into an insert filled with SE30 (poly(dime- Thus, the three-fold standard deviation of replicate measurethylpolysiloxane) gum, S+H Analytik, Mo�nchengladbach, ments (n=8) of a low-level standard solution with a concen- Germany) onto which the analytes were adsorbed. After tration level around 0.5 mg m-3 for each MVOC was used for injection, the temperature of the multi-injector was kept at calculation.The detection limits are expressed in terms 50 °C, while the solvent was removed via the split vent at a of mg m-3 related to an exposure interval of four weeks.flow rate of 15 ml min-1 for 10 s. Then the split was closed Desorption eYciencies were determined by spiking the for 60 s and the multi-injector was ballistically heated up to charcoal pads of unexposed samplers with known concen- 280 °C to thermally desorb the analytes and transfer them at trations of the analytes,31 waiting for 24 h and extracting them a ratio of 151 to the two capillary columns. with CS2 (containing 1% methanol ) as described above.Precision was tested by replicate (n=10) analysis of a real sample within-series and on diVerent days, respectively. Calibration and calculation The accuracy of the passive sampling procedure was checked by comparison with active sampling. For this, four active The MVOCs 3-methylbutan-2-ol, pentan-2-ol, 3-methylbutansamplers and four passive samplers were exposed simul- 1-ol, hexan-2-one, heptan-2-one, octan-3-one, octan-3-ol, taneously for two weeks in a room with visible mould nonan, fenchone and a-terpineol were purchased from formation.Aldrich (Steinheim, Germany), thujopsene was obtained from Roth (Karlsruhe, Germany). Storage stability Calibration was performed with a set of three standard solutions containing 0.4 to 1.6 mg ml-1 of each MVOC.The Storage stability was tested by exposing six diVusive samplers standard solutions and a blank were analysed before and after simultaneously in the same room close (approximately 10 cm) each analytical section which consisted of 20 samples. The to each other. After exposure two monitors were analysed means of the two runs of the standard solutions were used for immediately and two monitors were analysed after storage calibration.Standards were made from high-purity reagents periods of 1 month and 3 months at +4 °C, respectively. by diluting them first with methanol and then with ‘lowbenzene’ CS2 resulting in a solution finally containing 1% of Microbiological determination of mould fungi in mattress dust methanol.The stock solution in methanol can be used for at Mattress dust samples were taken by trained persons using a least 6 months, the standards in CS2 were prepared freshly vacuum cleaner (Weltstar, Mini-Turbo 1410 I Elektronik, just before use. 1400 W) equipped with a filter box and an intake manifold The results of the gas chromatographic determination, which (Schleicher & Schuell, Dassel, Germany). After clothing was are given in mg ml-1, were corrected for the blank value, removed, the whole surface of the child’s mattress was vacuum- converted to mg absolute and then to mg m-3 according to the cleaned for 5 min in longitudinal and transverse directions.following equation:29 Subsequently, the dust samples were weighed.The dust samples were immediately sent to the laboratory. Aliquots (100 to c (mg m-3)=m·A/r·t·1000 500 mg) of the dust samples were diluted tenfold with 0.9% (physiological ) sodium chloride solution and agitated on a with m=mass adsorbed onto the charcoal pad of the passive sampler given in mg; A=calculation constant=1000/sampling Vortex mixer type G 560 E (Scientific Industries Inc., USA) for 5 min.The suspensions were again diluted 1+9 and 1+99. rate given in 1000 min ml-1; r=recovery coeYcient; t=sampling interval given in min. The starting suspension and the 1+9 and 1+99 diluted suspensions were pipetted onto Petri dishes with Applying this equation it is necessary to know the sampling rate A, which diVers from analyte to analyte.These constants dichlorane–glycerine(DG-18)-agar as culture medium and incubated for up to 7 d at 25 °C. CFU were counted out after can be either determined experimentally29 or can be calculated mathematically according to Hirschfelder et al.30 In this work 2, 3, 4, 5, and 7 d. Identification of the mould fungi was performed microscopically according to their morphological the Hirschfelder equation was applied.The resulting calculation constants which were used for the single MVOCs are characteristics. The slide preparations were dyed with Lactophenol-Blue (Merck, Darmstadt, Germany). listed in Table 1. J. Environ. Monit., 1999, 1, 445–452 447Field studies Results and discussion Indoor MVOC concentrations were determined in the context The described analytical procedure based on diVusive sampling of a large field study conducted between February and May and solvent extraction in combination with dual-column capil- 1997, which has been described in detail in a former paper.32 lary GC-FID permits the separation and determination of 11 The study was carried out in Halle/Saale and Essen/Ruhr, two selected MVOCs of microbial origin in indoor air samples at industrialized cities in Eastern and Western Germany, and in normal and elevated concentrations.The 11 MVOCs under Osterburg, a small East-German town located in a rural area. investigation were detected in all dwellings, in those with The children (5- to 7-years old) were examined by a physician mould and in those without mould formation. All MVOCs on the occasion of their school entrance examination.showed a tendency to higher values in dwellings with mould Participation at the study was on a voluntary basis. The formation, but, however, not always at a statistically significant parents were also asked to answer a questionnaire. This was level. Elevated MVOC levels can thus serve as a first hint for sent along with the invitation and controlled by a physician the presence of (non-visible) mould fungi.The MVOC backat the day of investigation. It asked for such information as ground levels in dwellings without known mould formation complaints and symptoms of the children (bronchitis, asthma, may be a result of non-visible indoor mould growth, of allergies, frequency of colds, irritation of eyes, skin and nose, penetration of MVOCs via outdoor air originating from etc.) and confounding factors such as age, sex, body-mass outdoor mould or by unspecific emission from other sources index, educational level of the parents, and number of siblings, besides mould such as essential oils, building materials, paints, type of heating, and presence of pets.A German translation adhesives, and other household articles.of ISAAC questions33 was used for assessing symptoms of The dual-column configuration was used for validation allergies. purposes. Nine of the MVOCs listed in Table 1 can be deter- Air samples were taken in a total of 132 dwellings of those mined without interferences with both types of columns. Due families who had accepted to participate. The monitors were to overlap with oct-1-ene-3-ol, the determination of octanplaced in the children’s room (as described above).Dust 3-one is only possible with the non-polar DB-5 column. The samples of the child’s mattress were taken by trained persons same is true for heptan-2-one due to peak overlap with an visiting the children’s home who brought with them a vacuum unknown compound. The determination of oct-1-ene-3-ol, for cleaner (see Microbiological determination of mould fungi in which other authors10–12 were able to show an association mattress dust).The measurement of relative humidity was not with the extent of indoor mould growth, is neither possible possible, because it is not feasible to monitor the humidity with the DB-5 nor with the DB-1701 column due to peak simultaneously in 132 dwellings over the sampling period of 4 overlap with octan-3-one (DB-1701) and 1,3,5- weeks.Short-term measurements have a relatively high degree trimethylbenzene (DB-5). This MVOC was thus excluded from of uncertainty due to the variation in the course of the day. further investigations which were continued for the remaining Thus, information about dampness and mould formation was 11 MVOCs.assessed by the standardized interview done by trained persons. The figures of merit of the described analytical procedure Additional information about the building/dwelling (venti- are also summarized in Table 2. Related to a 4 week sampling lation, renovation, smoking of the parents, age of the house, period, the detection limits applying FID detection ranged etc.) were gained in this interview, too.In 15 of the 132 between 0.15 and 0.53 mg m-3 for the individual MVOCs. dwellings (11.4%) a visible mould formation was reported for They can be additionally improved by using an MSD instead the children’s room. In 19 cases (14.4%) the children’s rooms of the FID detector. Our detection limits for MVOCs are thus were indicated as damp. comparable with those which were reported for ‘normal’ VOCs, such as benzene, toluene and the xylenes.Porstmann Statistical analysis et al.34 reported detection limits between 0.6 and 0.8 mg m-3 for benzene and toluene using passive sampling and determi- Linear regression was used to test the influence of dampness, nation by GC-FID. Begerow et al.27 found detection limits mould formation, and occurrence of mould species in mattress ranging between 0.1 and 0.8 mg m-3 using a similar procedure.dust on the MVOC concentrations in indoor air. Logistic Our detection limits were slightly lower than those reported regression was used to investigate the influence of MVOCs on by Porstmann et al.34 and Begerow et al.27 because we used health endpoints after adjusting for confounders. The confounthe splitless large-volume injection technique (sample size: ders were selected according to the experience of the large field study.32 10 ml ) instead of the conventional split injection technique of Table 2 Precisions, recoveries and detection limits of the described passive sampling procedure Precision Concentration in real-life Within-series Day-to-day Recoverya Detection limita standard MVOC sample/mg m-3 (%) n=10 (%) n=10 standard 2 (%) n=4 1/mg m-3 (n=8) 3-Methylbutan-1-ol 1.42 11.8 11.9 55 0.53 3-Methylbutan-2-ol 0.82 6.5 16.7 102 0.17 Fenchone 1.27 10.1 21.1 111 0.24 Heptan-2-one 0.23 16.3 20.2 110 0.15 Hexan-2-one 0.26 16.5 25.9 98 0.16 Nonan-2-one 0.60 11.5 30.4 118 0.15 Octan-3-ol 2.43 13.9 17.5 84 0.23 Octan-3-one 1.86 19.0 15.7 104 0.15 Pentan-2-ol 0.54 11.4 15.2 77 0.23 a-Terpineol <0.23 — — 83 0.23 Thujopsene <0.23 — — 99 0.23 aConcentration level of standard 1: around 0.5 mg m-3 for each MVOC (range: 0.4–0.9 mg m-3); concentration level of standard 2: around 1 mg m-3 for each MVOC (range: 0.8–1.8 mg m-3). 448 J. Environ. Monit., 1999, 1, 445–452of 3-methylbutan-1-ol, the mean recovery (n=4) of the MVOCs was at least 77% (Table 2).The relatively low recovery rate of 55% for 3-methylbutan-1-ol may be explained by the high polarity of this MVOC and its incomplete desorption by CS2. Sunesson et al.36 who tested diVerent adsorbents (Tenax, Chromosorb, Carbotrap, Carbopack, Anasorb, Porasil C) for the sampling of MVOCs followed by thermal desorption found Tenax to give the best recoveries.For a concentration of 1 mg m-3 of each MVOC the recovery was between 81 and 107%. Only for propan-2-ol was a high breakthrough observed. Wesse�n and Schoeps11 reported MVOC recoveries in the range of 95–177% using Anasorb. Using passive sampling devices with activated charcoal, Begerow et al.27 found recoveries between 93 and 109% for 26 VOCs of environmental relevance.Only the recoveries for the less volatile naphthalene and the polar pyridine were distinctly lower and were about 50%.27 For validation purposes, samples were taken simultaneously by the passive and active sampling technique in a room with visible mould growth. The corresponding results are summarized in Table 3. With the exception of fenchone and nonan- 2-one, the relative percentage diVerences for passive sampling related to active sampling were found to range between -30% and +30%. These deviations are slightly higher than for other VOCs,26,37,38 which may be explained by the fact that the MVOC concentrations were distinctly lower than the VOC concentrations.Regarding the MVOCs, losses or contaminations were neither observed during storage of exposed samplers for one month at+4 °C nor during storage at+4 °C for three months.The results are illustrated in Fig. 2. The relative diVerences were comparable with those for the day-to-day precision. Field studies Our investigations have shown that indoor 3-methylbutan-1-ol, heptan-2-one, hexan-2-one, octan-3-ol, pentan-2-ol, and a-terpineol concentrations were significantly higher in those children’s rooms in which visible mould growth occurred according to the interview data (see Fig. 3). The geometric means, geometric standard deviations, ranges, and 95th percentiles of the indoor MVOC concentrations in dwellings with and without a known mould growth are listed in Table 4. Significantly higher octan-3-ol and 3-methylbutan-1-ol were also found in those children’s rooms which were regarded as Fig. 1 Comparison of chromatograms (column: DB-5) of a real-life damp (n=19) without a visible mould formation. This associ- sample obtained with (a) the splitless large-volume (sample loading: ation may be explained by the fact that dampness favours the 10 ml ) and (b) the split injection (sample loading: 2 ml; split: growth of microorganisms.Thus, in some of the damp dwell- 10 ml min-1) techniques; (c) shows the chromatogram of a low level standard solution containing between 0.4–0.9 mg m-3 of each analyte ings non-visible mould formation may have occurred resulting using splitless large-volume injection (attenuation in all three chroma- in elevated concentrations of octan-3-ol and 3-methylbutantograms is identical ).MVOC concentrations of the real-life sample: 1-ol in indoor air. These two MVOCs seem to be the most 1: 3-methylbutan-2-ol (2.06 mg m-3); 2: pentan-2-ol (0.47 mg m-3); 3: sensitive and reliable indicators for indoor mould growth. 3-methylbutan-1-ol (1.16 mg m-3); 4: hexan-2-one (0.32 mg m-3); 5: Our results are in accordance with those of Dewey et al.12 heptan-2-one (0.74 mg m-3); 6: octan-3-one (5.87 mg m-3); 7: who found a correlation beween hexan-2-one, heptan-2-one, oct-1-ene-3-ol (not determinable due to peak overlap with 1,3,5-trimethylbenzene); 8: octan-3-ol (6.12 mg m-3); 9: nonan-2-one and oct-1-ene-3-ol with the extent of indoor mould formation.(2.10 mg m-3); 10: fenchone (0.63 mg m-3); 11: a-terpineol Sagunski13 also reported an association with the extent of (2.06 mg m-3); 12: thujopsene (<0.23 mg m-3).mould growth for oct-1-ene-3-ol, hexan-2-one, heptan-2-one, and to a lesser degree to 2-methylpropan-1-ol, butan-1-ol, oct-2-ene-1-ol and 3-methylfuran. As already explained, the a 2 ml aliquot for sample introduction. A comparison of the chromatograms obtained with the large-volume and the split determination of oct-1-ene-3-ol was not possible under the given analytical conditions of the study presented here due to technique is given in Fig. 1 indicating that the large-volume injection technique leads to a distinct increase in sensitivity. peak overlap on both types of columns. A statistically significant correlation was also found between Within-series precision was between 6.5% and 19.0%, and day-to-day precision was in the range of 11.9–30.4% (Table 2).selected MVOCs and the occurrance of mould species in mattress dust. Aspergillus sp. correlated significantly on the These data were obtained with a real sample (concentrations are given in Table 2), which was analysed tenfold, and are in P<0.1 level with heptan-2-one, hexan-2-one, octan-3-ol, octan-3-one, and a-terpineol, while the occurrance of Eurotium good agreement with those which were found for other VOCs, such as BTX, for example.26,27,35 Results for a-terpineol and sp.was statistically significant (P<0.1) correlated with higher indoor air concentrations of 3-methylbutan-1-ol, thujopsene are not available, because their concentrations were below the detection limit in this sample.With the exception 3-methylbutan-2-ol, heptan-2-one, hexan-2-one, octan-3-ol, J. Environ. Monit., 1999, 1, 445–452 449Table 3 Method validation by comparison between passive and active sampling (AM=arithmetic mean) Passive sampling AM Active sampling AM DiVerence MVOC (n=4)/mg m-3 (n=4)/mg m-3 (%) 3-Methylbutan-1-ol 2.0 1.8 +11.1 3-Methylbutan-2-ol 2.6 1.9 +36.8 Fenchone 0.35 0.93 -62.3 Heptan-2-one 1.3 1.4 -7.1 Hexan-2-one 0.48 0.39 +23.1 Nonan-2-one 1.3 2.2 -40.9 Octan-3-ol 15.0 11.9 +26.1 Octan-3-one 5.7 5.9 -3.4 Pentan-2-ol 0.27 0.40 -32.5 a-Terpineol 0.90 1.3 -30.8 Thujopsene 0.65 n.d.a — aNot determined.Table 4 Indoor MVOC concentration in children’s rooms with (n= 15) and without (n=117) mould formation (GM=geometric mean; GSD=geometric standard deviation) Mould GM/ Range/ P95 a / MVOC formation mg m-3 GSD mg m-3 mg m-3 3-Methylbutan-1-ol No 0.7 1.9 0.4–5.3 2.4 Yes 1.3 1.9 0.4–3.1 3.1 3-Methylbutan-2-ol No 0.5 2.1 0.1–4.8 1.9 Yes 0.7 2.5 0.1–2.5 2.5 Fenchone No 0.3 2.2 0.2–3.5 1.5 Yes 0.4 3.4 0.2–6.5 6.5 Heptan-2-one No 0.2 1.8 0.2–0.9 0.8 Yes 0.3 2.3 0.3–0.8 0.8 Hexan-2-one No 0.1 1.5 0.1–0.9 0.4 Yes 0.2 1.6 0.1–0.4 0.4 Nonan-2-one No 0.4 2.4 0.1–3.8 1.7 Yes 0.8 3.9 0.1–33.5 33.5 Octan-3-ol No 1.8 2.4 0.2–15.7 7.4 Yes 4.6 2.4 1.2–22.9 22.9 Octan-3-one No 1.4 2.3 0.1–7.5 5.2 Fig. 2 Storage stability of exposed ssive samplers (arithmetric mean, Yes 2.0 2.8 0.2–13.8 13.8 n=2). Pentan-2-ol No 0.2 1.7 0.2–1.1 1.1 Yes 0.3 1.8 0.2–0.8 0.8 a-Terpineol No 0.5 2.3 0.2–3.8 1.7 and thujopsene. The correlation coeYcients r and the signifi- Yes 0.9 2.5 0.2–4.6 4.6 cance levels P for each MVOC under investigation are Thujopsene No 0.2 1.5 0.2–1.6 0.4 Yes 0.2 1.7 0.2–0.5 0.5 summarized in Table 5.The duration of daily ventilation was found to be negatively a95th percentile. correlated with the MVOC concentration in indoor air (r= -0.313). This finding can be easily explained by the fact that Table 5 Association between the MVOC concentration in the air of ventilation leads to a dilution of the MVOC loaded air by the children’s rooms and in mattress dust; correlation coeYcients r outdoor air.and significance levels P Aspergillus sp. Eurotium sp. MVOC r P r P 3-Methylbutan-1-ol 0.123 0.166 0.186 0.036a 3-Methylbutan-2-ol 0.010 0.263 0.154 0.084a Fenchone 0.014 0.88 0.049 0.582 Heptan-2-one 0.185 0.036a 0.257 0.003a Hexan-2-one 0.262 0.003a 0.237 0.007a Nonan-2-one -0.002 0.978 0.145 0.102 Octan-3-ol 0.156 0.079a 0.251 0.004a Octan-3-one 0.182 0.040a 0.144 0.105 Pentan-2-ol 0.131 0.139 0.029 0.744 a-Terpineol 0.158 0.076a 0.134 0.132 Thujopsene 0.109 0.221 0.193 0.029a aSignificant at the P<0.1 level.Our results indicate that there is an association between the prevalence of symptoms in the children and the indoor MVOC Fig. 3 MVOC concentrations in children’s rooms with and without concentrations. This is illustrated in Fig. 4(a) and 4(b) on the mould formation (geometric means and 95% confidence intervals, example of 3-methylbutan-1-ol and octan-3-ol. Children living significant diVerences of the t-test (P<0.05) are indicated by an asterisk.in dwellings with higher MVOC levels had a higher prevalence 450 J. Environ. Monit., 1999, 1, 445–452determination of MVOCs in indoor air, should be taken. Due to fluctuations in the indoor MVOC concentrations in the course of the day and due to the fact that MVOCs are only produced during the growth phase, long-term sampling is recommended. The detection limits can be further improved by using a more sensitive detector, such as an MSD instead of an FID.Nevertheless, further basic studies are needed on the selection of suitable indicator MVOCs and on their dependence on indoor mould growth and on possible interferences and confounders. More intensive research is needed about the dependence of the MVOC concentrations on various parameters such the fungal species, the fungal growth period, and the substrate.These data are an essential prerequisite for a future implementation of reference and threshold values. Acknowledgements We thank Ms. Rossi for data-management and help with the statistical analysis and all participating families for their collaboration. The field work of the study was partly financed by the Ministry of Social AVairs Saxony-Anhalt.References 1 B. Flannigan, E. M. McCabe and F. McGarry, J. Appl. Bact. Symp., 1990, (Suppl.) 70, 61S. 2 S. Gravesen, Allergy, 1979, 34, 135. 3 R. S. Bernstein, W. G. Sorensen, D. Garabrant, C. Reaux and R. D. Treitman, Am. Ind. Hyg. Assoc. J., 1983, 44, 161. 4 H. Repp and D. Mu�ller-Wening, Allergy, 1989, 12, 54. Fig. 4 Correlation of the indoor 3-methylbutan-1-ol (a) and octan-3-ol 5 J. Lacey and J. Dutkiewicz, J. Aerosol Sci., 1994, 24, 1371. (b) concentration with adverse health eVects. 6 S. D. Platt, C. J. Martin, S. M. Hunt and C. W. Lewis, Br. Med. J., 1989, 298, 1673. 7 J. D. Miller, Atmos. Environ., 1992, 26A, 2163. of asthma, hay fever, wheezing, and irritations of the eyes. 8 H. Beguin, Aerobiologia, 1995, 11, 3.These correlations remained positive after controlling for 9 C. A. Hunter, C. Grant, B. Flannigan and A. F. Bravery, Int. confounding factors such as age, sex, body-mass index, number Biodeterior., 1988, 24, 81. of siblings, social status, passive smoking, type of heating, and 10 A.-L. Pasanen, S. Lappalainen and P. Pasanen, Analyst, 1996, ventilation habits.However, these associations were not stat- 121, 1949. 11 B. Wesse�n and K.-O. Schoeps, Analyst, 1996, 121, 1203. istically significant. This lack of significance may be a result 12 S. Dewey, H. Sagunski. U. Palmgren and B. Wildeboer, Zbl. Hyg., of the low sample number or may indicate that there is no 1995, 197, 504. true association. Therefore the investigations will be replicated 13 H.Sagunski, Umweltmed. Forsch. Prax., 1997, 2, 95. on a broader scale in an ongoing study. 14 D. Norba�ck, I. Michel and J. Wildstro�m, Scand. J. Work Environ. Our results are in accordance with the findings of Platt Health, 1990, 16, 121. et al.6 that children living in damp and mouldy houses had a 15 J. D. Miller, A. M. Laflamme, Y. Sobol, P. Lafontaine and R. Greenhalgh, Int.Biodeterior., 1988, 24, 103. higher prevalence of respiratory symptoms, such as wheeze, 16 A.-L. Sunesson, C.-A. Nilsson, B. Anderson and G. Blomquist, sore throat, runny nose, and headache and fever, compared Ann. Occup. Hyg., 1996, 40, 397. with those living in dry dwellings. In the study of Platt et al.6 17 R. Keller, R. So�nnichsen and H. Ohgke, Umweltmed. Forsch. assessment of dampness and type and location of mould Prax., 1997, 2, 265.formation was carried out by surveyers who visited the dwell- 18 J. Bjurman, E. Nordstrand and J. Kristensson, Indoor Air, 1997, ings. They took air samples from rooms, which were handed 7, 2. 19 A.-L. Pasanen, P. Kalliokoski, P. Pasanen, M. S. Jantunen and over to a microbiologist who estimated the spore counts and A.Nevelainen, Environ. Int., 1991, 17, 225. identified the fungi. A determination of the indoor MVOC 20 A.-L. Pasanen, P. Pasanen, M. J. Jantunen and P. Kalliokoski, level was not carried out by Platt et al.6 Atmos. Environ., 1991, 25A, 459. 21 T. O. Larsen and J. C. Frisvad, Mycol. Res., 1995, 99, 1153. 22 A.-L. Sunesson, W. H. J. Vaes, C.-A. Nilsson, G. Blomquist, Conclusions B.Andersson and R. Carlson, Appl. Environ. Microbiol., 1995, 61, 2911. Our results indicate that MVOC determination in indoor air 23 K. Wilkins and K. Larsen, Chemosphere, 1995, 31, 3225. is an encouraging approach to identify mould formation in 24 J. Bjurman and J. Kristensson, Mycopathologia, 1992, 118, 173. dwellings. A significant association with the extent of mould 25 P. Pasanen, A. Korpi, P. Kalliokoski and A.-L. Pasanen, Environ. formation has been found for selected MVOCs such as Int., 1997, 23, 425. 3-methylbutan-1-ol and octan-3-ol. From the primary MVOCs 26 J. Begerow, E. Jermann, T. Keles, U. Ranft and L. Dunemann, under investigation, 3-methylbutan-1-ol, hexan-2-one, heptan- Fresenius’ J. Anal. Chem., 1995, 351, 549. 27 J. Begerow, E. Jermann, T. Keles, T. Koch and L. Dunemann, 2-one, and octan-3-ol were found to be most reliable indicators J. Chromatogr., A, 1996, 749, 181. for mould formation. The MVOC determination is especially 28 NIOSH manual of analytical methods, ed. P. M. Eller, DHHS valuable for screening purposes, because MVOCs and other (NIOSH) Publication, Cincinnati, OH, 3rd edn., 1984, vol. 1, environmentally relevant anthropogenic VOCs, such as ben- pp. 84–100. zene, toluene, ethylbenzene, and the xylenes can be determined 29 K.-H. Pannwitz, Dra� gerheft, 1983, 325, 1. using the same sample. To obtain reliable results, samples 30 J. O. Hirschfelder, R. B. Bird and E. L. Spotz, Trans. Am. Soc. Mech. Eng., 1949, 71, 921. both for the determination of the number of CFU and for the J. Environ. Monit., 1999, 1, 445–452 45131 S. T. Rodriguez, D. W. Gosselink and H. E. Mullins, Am. Ind. 35 H. C. Shields and C. J. Weschler, JAPCA, 1987, 37, 1039. 36 A.-L. Sunesson, C.-A. Nilsson and B. Andersson, J. Chromatogr., Hyg. Assoc. J., 1982, 43, 569. A, 1995, 699, 203. 32 U. Kra�mer, H. Behrendt, U. Ranft, J. Ring and 37 W. J. Lautenberger, E. V. Kring and J. A. Morello, Am. Ind. Hyg. H. W. Schlipko� ter, Int. J. Epidemiol., in the press. Assoc. J., 1989, 4, 737. 33 M. I. Asher, U. Keil and H. R. Anderson, Eur. Resp. J., 1995, 38 M. A. Cohen, B. Ryan, Y. Yanagisawa and S. K. Hammond, 8, 483. J. AirWaste Manage., 1990, 40stmann, J. Bo� ke, S. Hartwig, R. Kaaden, R. Rosenlehner, A. Schupp, T. Stiller and H. E. Wichmann, Staub—Reinhalt. Luft, 1994, 54, 147. Paper 9/03034D 452 J. Environ. Monit., 1999, 1, 445–4

ISSN:0960-7919    

DOI:10.1039/a903034d

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

A catalogue of urban hydrocarbons for the city of Leeds: atmospheric monitoring of volatile organic compounds by thermal desorption-gas chromatography Suzanne Hassoun,*† Michael J. Pilling and Keith D. Bartle School of Chemistry, University of Leeds, Leeds, UK LS2 9JT Received 18th June 1999, Accepted 10th August 1999 A method has been developed for the speciation and quantitative determination of hydrocarbons in urban air in the city of Leeds.Hydrocarbons were pre-concentrated by adsorbent tube air sampling and analysed using thermal desorption and gas chromatography with flame ionization detection and structural confirmation by mass spectrometric detection. While automated volatile organic compound (VOC) analysers produced data for a maximum of about 30 compounds simultaneously, with the method described here, a total of 68 C6–C12 hydrocarbons were measured simultaneously in one analysis at parts per billion (ppb) levels. Several monitoring surveys were performed, one during the winter of 1993 and the other in the summer of 1994, at a number of sites to investigate the levels of VOCs identified in the urban air of Leeds.pounds.11 Wider surveys, on a less frequent basis, are needed Introduction to assess the full range of VOCs, as illustrated by those The atmospheric oxidation of volatile organic compounds conducted in Rome12 which indicate a much larger number of (VOCs), in the presence of nitrogen oxides and sunlight, leads VOC species present in urban air.to the formation of ozone and other photochemical oxi- The main objective of this study was therefore to develop a dants.1–4 In addition, some VOCs are carcinogenic.5 Various sampling and analysis method, using thermal desorption-gas approaches, such as environmental chamber experiments6 and chromatography, to enable compilation of a comprehensive the use of photochemical ozone creation potentials in model- inventory of the occurrence and concentrations of hydroling studies,7 have demonstrated that VOC structure is a major carbons in urban air.With the method described here, a total factor in determining their rates and mechanisms of reaction of 68 hydrocarbons at ppb levels were monitored simulin the atmosphere and hence their diVering abilities to produce taneously. In addition to the larger number of hydrocarbons photochemical oxidants. The development and application of monitored per analysis run, this method was also able to techniques to monitor speciated VOCs in the atmosphere is, provide, when necessary, spatial hydrocarbon data, completherefore, of importance. menting the single site measurements from the DETR HCN.Emissions inventories for the United Kingdom (UK) indicate that mobile sources and solvent usage are the two largest source categories, accounting for about 70% of total Experimental UK emissions of VOCs.8 Current tropospheric photochemical Instrumentation oxidant models rely on VOC emission levels taken from emissions inventories that have not been tested in detail against The equipment used for VOC adsorbent tube analysis consisted measurements in urban environments owing to the lack of of a thermal desorption unit autosampler (TDAS 5000, Carlo VOC monitoring performed.Up until the 1990s, very few Erba Instruments, Fisons, Milan, Italy) coupled to a gas measurements were made in urban areas in the UK to monitor chromatograph (GC, Carlo Erba Mega series 5600) via an air quality. Before VOC control strategies can be developed intermediate fused silica capillary interface and cold trap and implemented, an accurate understanding of the species, (Carlo Erba Multi Function Actuator MFA 515).levels and sources of VOCs present in urban air is required. Since the publication of the UK’s Environment White Adsorbent tubes and air sampling Paper,9 substantial progress has been made in urban air Stainless steel tubes, packed with 0.200 g of Tenax TA 60/80 pollution monitoring.The routine measurement of UK urban mesh adsorbent (Chrompack, London, UK), were used for concentrations of VOCs is now performed through the air sampling of VOCs. Before use, tubes were cleaned in the Department of the Environment, Transport and the Regions thermal desorption unit (TDAS). To prevent contamination (DETR) Hydrocarbon Network (HCN).This method is based during storage, tubes were sealed using Swagelok end caps on adsorbent tube air sampling followed by thermal desorp- fitted with Teflon ferrules and were also stored in a vacuum tion-gas chromatographic analysis.10 These systems provide desiccator. Sampled tubes were stored in a refrigerator at 5 °C continuous VOC data, but the number of VOC species moniand were analysed as soon as possible, always within 24 h of tored is necessarily limited to only approximately 30 comsample collection.The sample volume was 3 l collected at a typical flow rate of 150 ml min-1 using a portable constant †Present address: Environmental Health, Institute of Public and flow sampling pump ( Universal Sample Pump-224-PCXR7, Environmental Health, The University of Birmingham, Edgbaston, SKC, Blandford Forum, Dorset, UK).The sampling height Birmingham, UK B15 2TT. E-mail: S.HASSOUN@bham.ac.uk; Fax: +44 (0)121 4143709. was 1.5 m above ground level. J. Environ. Monit., 1999, 1, 453–458 453VOC sample analysis The sampled tube was purged with helium in the TDAS at ambient temperature to remove any air and water vapour remaining after sampling and was then pre-heated to attain the desorption temperature of 250 °C.The VOCs were then desorbed from the tube in a stream of helium at 5 ml min-1 for 10 min. The TDAS switching valve, transfer line (fused silica capillary BP1, 0.32 mm id, 5 mm film thickness, SGE, Milton Keynes, Buckinghamshire, UK) and GC injector were held at 250 °C. During desorption, sample components were focused by cooling the end section of the capillary transfer line to -150 °C.At the end of desorption, the cold trap was rapidly heated to 250 °C to inject the VOCs onto the analytical column which was a non-polar methylpolysiloxane column, 007–1, 50 m, 0.25 mm id, 1 mm film thickness (Quadrex, Fig. 1 Retention volume (VR) plot for hexane on Tenax. Weybridge, Surrey, UK).The helium flow rate was 1.5 ml min-1. VOCs were separated using a temperature protube at diVerent temperatures. A graph of log(VR) against 1/T gramme -50 °C for 2 min, 49.9 °Cmin-1 to 35 °C for 10 min, (Fig. 1) for hexane, the most volatile VOC studied, was linear, 3.0 °Cmin-1 to 200 °C for 2 min and 40.0 °C to 240 °C for which gives log(VR/ml )=3622K/T-8.3, from which the VR 30 min.Compounds were identified using a combination of for hexane at 20 °C may be calculated. The BTV or safe information: according to their retention times by direct sampling volume has been defined as 50% of the retention injection of standard VOC mixtures prepared in the laboratory volume.14 A safe sampling volume (1/2VR) for hexane was (Aldrich Chemicals, Gillingham, Dorset, UK) with detection 5.4 l (per 0.200 g of Tenax at 20 °C), which exceeds the using a flame ionization detector (FID-40, Carlo-Erba) at experimental sampling volume of 3 l so that no losses of VOCs 290 °C, with a hydrogen flow of 30 ml min-1 and an air flow were expected from the adsorbent tube during sampling.All of 300 ml min-1; a mass spectrometer (VG Trio-1000, VG samples were collected at air temperatures of 20 °C or less.To Instruments, East Grinstead, West Sussex, UK) was employed confirm the maximum sampling volume during air sampling, for positive peak identification in the EI+ mode (70 eV), two Tenax adsorbent tubes were connected in series and 3 l scanning 5–200 amu with a scan time of 1 s and a scan interval air samples were collected as above at 150 ml min-1 of air.of 0.15 s; the ionization temperature was 200 °C and the No VOCs were observed on the second back-up tube. detector multiplier voltage was 500 V. Hydrocarbon breakthrough from the intermediate cold trap The external standard method of calibration was used for during desorption was also investigated at several desorption the quantification of hydrocarbons.A standard was prepared volumes for hexane using the minimum cold trapping tempera- containing two hydrocarbons, heptane and toluene, in pentane. ture of -150 °C. Desorption eYciencies of hydrocarbons from Serial dilution of this stock solution produced a series of the Tenax tubes versus desorption temperature and time were standards in the nanogram range.Tenax tubes were repeatedly investigated for several compounds with a range of volatilities, loaded with each concentration of standard via injection of i.e. octane, o-xylene and 2-methylnonane. Fig. 2(a) confirms the standard into a stream of helium flowing through a heated that, for the desorption volume of 50 ml helium used during packed column gas chromatograph injection port onto the analysis, there were no losses from the cold trap during Tenax tube which was then analysed as above.Most of the desorption of hexane, the lightest VOC sample component. pentane solvent was lost from the tube after injection owing Fig. 2(b) and 2(c) indicate incomplete desorption of VOCs to breakthrough. Calibration graphs of signal versus concenfrom the sample tube at values below the analysis desorption tration were plotted; the regression equations obtained were temperature (250 °C) and desorption time (10 min).used to determine the unknown hydrocarbon concentrations in urban air samples. For heptane, y=161152x+6515 (R2= 0.999) and, for toluene, y=144015x-80736 (R2=0.999). Results and discussion Since the responses of most hydrocarbons in the flame Identification of hydrocarbons in Leeds air ionization detector (FID) are similar, i.e.they are approximately proportional to the total mass of carbon and hydrogen Fig. 3 shows a chromatogram of a typical 3 l urban air VOC in the analyte, it was therefore possible to use the response sample. The sample was collected at a busy roadside location factor of heptane to quantify aliphatic compounds and the in the city centre of Leeds during the summer of 1994.A total toluene response factor to quantify the aromatics, thus simpli- of 68 C6–C12 hydrocarbons were identified. Peak identifications fying the calibration process. Relative responses (RR) of a are given in Table 1. range of hydrocarbons in the FID have been quoted by Before any VOC air samples were analysed, an estimation Smith13 as 97–102 for aliphatics with an RR of 100 for heptane of the compounds expected to be found in urban air was and 103–112 for aromatics with an RR of 107 for toluene.carried out by studying UK VOC source emissions profiles From the above RR values, using heptane to quantify ali- for the major VOC sources expected in urban air, i.e.petrol phatics produced a relative standard deviation (RSD) in exhaust, diesel exhaust.8 Standards of pure hydrocarbon comhydrocarbon concentrations of 3% and, using toluene to pounds were analysed individually or as group mixtures, e.g. quantify aromatics, the RSD was 4%. n-alkanes, to obtain a GC retention time for each compound. The hydrocarbon peaks in unknown air samples were identified Method development in the usual way by comparing their retention times with those of known hydrocarbon peaks in the standard mixes.The maximum sampling volume was established to prevent losses of the most volatile hydrocarbon from the Tenax tube For approximately half of the compounds listed in Table 1, pure hydrocarbon standards were not available in the labora- during sampling.The tube was treated as a short chromatographic column in the GC oven. Safe sampling volumes or tory. These compounds were identified by a combination of mass spectral information, referral to literature chromato- breakthrough volumes (BTV) were investigated by measuring the retention volumes (VR) of hexane (C6) on the adsorbent grams of urban air samples and petroleum standard mix- 454 J.Environ. Monit., 1999, 1, 453–458Fig. 3 Chromatogram of C6–C12 hydrocarbons in the urban air of Leeds (for peak identifications, see Table 1). of VOCs from the GC column was also determined when necessary by subtraction of the two mass spectra obtained for a given peak. The majority of hydrocarbons observed were alkanes (straight chain, branched chain, cyclic) and aromatic compounds, including benzene.Reference to the UK VOC source emissions profiles8 confirms that the major sources of the above compounds are petrol and diesel engine emissions and the measured hydrocarbon species profile closely resembles that of petroleum hydrocarbon analysis. There is some evidence for solvent emissions through the detection of methylcyclohexane and styrene.Leeds has comparatively little industry, so that a solvent contribution well below the national Fig. 2 (a) Trapping eYciency of hexane in the coated capillary cold trap. (b) Desorption eYciencies of hydrocarbons from Tenax versus temperature. (c) Desorption eYciencies of hydrocarbons from Tenax versus time. tures12,15 and comparison of calculated linear temperature programmed retention indices (RI ) of sample hydrocarbons with RI values in the literature12,15–17 as shown in Table 1.The mass spectrum fragmentation pattern of the unknown compound showed the types of groups present in the hydrocarbon and in some cases the molecular weight could be obtained as shown in Fig. 4(a) and 4(b) for methylcyclohexane and styrene respectively. Indications as to the identities of these unknown compounds were then obtained by referring to the elution order and relative concentrations of hydrocarbons in chromatograms taken from the literature of urban air samples and petroleum standard mixtures.Identifications were confirmed by calculation of a retention index for each sample hydrocarbon. A plot of n-alkane retention index (100×carbon number) versus adjusted retention time produced a linear temperature programmed retention index equation18 of RI= 376.88tR¾+13.81 (R2=0.994). Calculated RIs were in agree- Fig. 4 Selected mass spectra for hydrocarbons in Leeds urban air: (top) methylcyclohexane; (bottom) styrene. ment with literature RIs with an error of 1–3%. Any co-elution J. Environ. Monit., 1999, 1, 453–458 455Table 1 Peak identifications (see chromatogram, Fig. 1) for hydrocarbons in Leeds urban air No. Hydrocarbon Formula Method of identification Retention index 1 Hexane C6H14 Standard mix retention time 581.51 2 Hexene C6H12 MS-mass spec molecular weight&fragmentation ions 584.00 3 2-Methyl-2-pentene C6H12 Standard mix retention time 587.18 4 3-Methyl-c-2-pentene C6H12 Standard mix retention time 589.52 5 Hexene C6H12 MS 596.84 6 3-Methyl-t-2-pentene C6H12 Standard mix retention time 603.75 7 Methylcyclopentane C6H12 MS 610.93 8 2,4-Dimethylpentane C7H16 MS 614.79 9 3-Methylcyclopentene C6H10 MS 637.44 10 Benzene C6H6 Standard mix retention time 640.20 11 Cyclohexane C6H12 MS 648.90 12 2-Methylhexane C7H16 Standard mix retention time 658.98 13 2,3-Dimethylpentane C7H16 Standard mix retention time 661.05 14 3-Methylhexane C7H16 Standard mix retention time 669.20 15 c-1,3-Dimethylcyclopentane C7H14 MS 677.48 16 t-1,3-Dimethylcyclopentane C7H14 MS 680.66 17 2,2,4-Trimethylpentane C8H18 Standard mix retention time 684.80 18 Heptane C7H16 Standard mix retention time 698.19 19 Methylcyclohexane C7H14 MS 723.05 20 2,5-Dimethylhexane C8H18 MS+Elution order from literature chromatogram 736.30 21 2,4-Dimethylhexane C8H18 MS+Elution order from literature chromatogram 738.65 22 2,3,4-Trimethylpentane C8H18 Standard mix retention time 757.70 23 2,3,3,-Trimethylpentane C8H18 MS+Elution order from literature chromatogram 763.09 24 Toluene C7H8 Standard mix retention time 765.30 25 2,3-Dimethylhexane C8H18 MS+Elution order from literature chromatogram 769.99 26 2-Methylheptane C8H18 Standard mix retention time 776.21 27 4-Methylheptane C8H18 MS 777.73 28 3-Methylheptane C8H18 Standard mix retention time 784.63 29 c-1,3-Dimethylcyclohexane C8H16 MS 785.32 30 1,1-Dimethylcyclohexane C8H16 MS 791.12 31 2,2,5-Trimethylhexane C9H20 MS 797.06 32 1,4-Dimethylcyclohexane C8H16 MS 811.42 33 Octane C8H18 Standard mix retention time 814.18 34 Dimethylcyclohexane C8H16 MS 819.01 35 Dimethylheptane C9H20 MS 846.77 36 Dimethylheptane C9H20 MS 854.22 37 Ethylbenzene C8H10 Standard mix retention time 870.79 38 m&p-Xylenes C8H10 Standard mix retention time 879.49 39 2&4-Methyloctane C9H20 MS 884.05 40 3-Methyloctane C9H20 Standard mix retention time 891.50 41 Styrene C8H8 MS 900.02 42 o-Xylene C8H10 Standard mix retention time 903.66 43 Nonane C9H20 Standard mix retention time 919.12 44 i-Propylbenzene C9H12 MS 936.10 45 Propylcyclohexane C9H18 MS 950.33 46 Dimethyloctanes C10H22 MS 953.09 47 a-Pinene C10H16 MS 955.02 48 Propylbenzene C9H12 MS+Elution order from literature chromatogram 964.41 49 m-Ethyltoluene C9H12 Standard mix retention time 971.04 50 p-Ethyltoluene C9H12 Standard mix retention time 972.97 51 135-Trimethylbenzene C9H12 Standard mix retention time 978.22 52 2-Methylnonane C10H22 Standard mix retention time 980.70 53 o-Ethyltoluene C9H12 Standard mix retention time 988.16 54 1,2,4-Trimethylbenzene C9H12 Standard mix retention time 1001.97 55 Decane C10H22 Standard mix retention time 1012.05 56 1,2,3-Trimethylbenzene C9H12 Standard mix retention time 1027.65 57 Indane C9H10 Standard mix retention time 1039.25 58 1,3-Diethylbenzene C10H14 MS+Elution order from literature chromatogram 1049.74 59 1-Methyl-3-n-propylbenzene C10H14 MS 1057.61 60 n-Butylbenzene C10H14 Standard mix retention time 1066.87 61 1,4-Dimethyl-2-ethylbenzene C10H14 MS+Elution order from literature chromatogram 1074.60 62 1,3-Dimethyl-4-ethylbenzene C10H14 MS+Elution order from literature chromatogram 1076.67 63 1,2-Dimethyl-4-ethylbenzene C10H14 MS+Elution order from literature chromatogram 1091.78 64 Undecane C11H24 Standard mix retention time 1097.66 65 1,2,4,5-Tetramethylbenzene C10H14 Standard mix retention time 1110.08 66 1,2,3,5-Tetramethylbenzene C10H14 MS+Elution order from literature chromatogram 1112.85 67 Naphthalene C10H8 Standard mix retention time 1165.73 68 Dodecane C12H26 Standard mix retention time 1177.33 emissions average is to be expected.Finally, the detection of During the sampling period, summer 1994, the DETR HCN station in Leeds was monitoring hourly concentrations of a-pinene, a biogenic hydrocarbon, was not expected in urban air. The concentrations were, however, very small compared approximately 30 hydrocarbons in urban air ranging from C2–C9.19 The additional hydrocarbons identified in this work to those of alkanes and aromatics and may derive from localized parkland.enhance the hydrocarbon database available for Leeds and 456 J. Environ. Monit., 1999, 1, 453–458Table 2 Concentrations/ppb (n=21) of hydrocarbons in Leeds urban air Mean Concentration concentration range in No.Hydrocarbon in Leeds air/ppb Leeds air/ppb 1 Hexane 0.26 0.05–0.50 2 Hexene 0.03 0.01–0.07 3 2-Methyl-2-pentene 0.05 0.04–0.12 4 3-Methyl-c-2-pentene 0.06 0.01–0.12 5 Hexene 0.05 0.02–0.10 6 3-Methyl-t-2-pentene 0.05 0.01–0.11 7 Methylcyclopentane 0.13 0.03–0.22 8 2,4-Dimethylpentane 0.05 0.01–0.11 9 3-Methylcyclopentene 0.05 0.01–0.09 10 Benzene 0.97 0.22–1.94 11 Cyclohexane 0.03 0.01–0.05 12 2-Methylhexane 0.16 0.04–0.33 13 2,3-Dimethylpentane 0.09 0.02–0.19 Fig. 5 Concentrations of hydrocarbons identified in Leeds urban air. 14 3-Methylhexane 0.17 0.04–0.34 15 c-1,3-Dimethylcyclopentane 0.03 0.01–0.06 16 t-1,3-Dimethylcyclopentane 0.04 0.01–0.06 17 2,2,4-Trimethylpentane 0.37 0.09–0.80 18 Heptane 0.11 0.03–0.23 include, for example, methylhexanes, heptanes, octanes and 19 Methylcyclohexane 0.07 0.01–0.14 nonanes, n-alkanes C8–C12 and ethyltoluenes. Many of these 20 2,5-Dimethylhexane 0.06 0.02–0.13 compounds are significant in the generation of photochemical 21 2,4-Dimethylhexane 0.08 0.02–0.17 oxidants in the presence of nitric oxide.Aromatics and alkanes 22 2,3,4-Trimethylpentane 0.19 0.05–0.40 have been shown to have the highest photochemical ozone 23 2,3,3,-Trimethylpentane 0.08 0.02–0.19 24 Toluene 1.59 0.39–3.26 creation potentials (POCP) and, in terms of POCP weighted 25 2,3-Dimethylhexane 0.11 0.02–0.24 emissions, petrol exhaust is the largest source category for 26 2-Methylheptane 0.07 0.01–0.16 photochemical ozone generation.20 27 4-Methylheptane 0.03 0.01–0.07 28 3-Methylheptane 0.07 0.01–0.14 Quantification of hydrocarbons in Leeds air 29 c-1,3-Dimethylcyclohexane 0.02 0.01–0.05 30 1,1-Dimethylcyclohexane 0.01 0.01–0.02 Atmospheric concentrations of compounds were in the low 31 2,2,5-Trimethylhexane 0.02 0.01–0.05 ppb range as shown in Table 2 (number of samples, n=21). 32 1,4-Dimethylcyclohexane 0.02 0.01–0.04 The concentrations of non-aromatic hydrocarbons, such as 33 Octane 0.05 0.02–0.10 34 Dimethylcyclohexane 0.02 0.01–0.03 hexane and methylhexanes, were generally found to be lower 35 Dimethylheptane 0.01 0.01–0.02 (0.16–0.26 ppb) than those of the aromatics such as benzene, 36 Dimethylheptane 0.01 0.01–0.02 toluene and xylenes (0.97–1.59 ppb).Limits of detection of 37 Ethylbenzene 0.30 0.06–0.62 the analytical method were 0.1–0.2 ppb (3s) with an 38 m&p-Xylenes 1.07 0.22–2.21 on-column detection limit of 0.01 ppb with a signal to noise 39 2&4-Methyloctane 0.07 0.01–0.15 ratio of 351.DiVerences in sample concentrations between 40 3-Methyloctane 0.03 0.01–0.06 41 Styrene 0.04 0.01–0.11 compounds were thought to be caused by diVerences in source 42 o-Xylene 0.48 0.11–0.95 emissions and atmospheric behaviour for each compound. 43 Nonane 0.04 0.01–0.08 However, the pattern of variation in VOC concentrations with 44 i-Propylbenzene 0.02 0.01–0.04 time for all compounds is very similar, as shown in Fig. 5, 45 Propylcyclohexane 0.01 0.01–0.03 suggesting common factors (mobile sources and meteorology) 46 Dimethyloctanes 0.01 0.01–0.03 aVecting VOC behaviour equally. 47 a-Pinene 0.01 0.01–0.02 48 Propylbenzene 0.07 0.02–0.16 49 m-Ethyltoluene 0.25 0.05–0.50 Acknowledgement 50 p-Ethyltoluene 0.15 0.04–0.29 51 135-Trimethylbenzene 0.19 0.05–0.36 We would like to thank the UK Department of the 52 2-Methylnonane 0.05 0.02–0.10 Environment for financial support of this work (PECD 53 o-Ethyltoluene 0.13 0.02–0.25 7/12/111). 54 1,2,4-Trimethylbenzene 0.53 0.11–1.05 55 Decane 0.10 0.05–0.16 56 1,2,3-Trimethylbenzene 0.11 0.02–0.22 References 57 Indane 0.05 0.01–0.13 58 1,3-Diethylbenzene 0.02 0.01–0.05 1 A.J. Haagen-Smit, Ind. Eng. Chem., 1952, 44, 1342. 59 1-Methyl-3-n-propylbenzene 0.04 0.01–0.11 2 M. Lippmann, Environ. Sci. Technol., 1991, 25, 1954. 60 n-Butylbenzene 0.07 0.01–0.19 3 D. I. Tingey and G.E. Taylor, Jr, EVects of Gaseous Air Pollution 61 1,4-Dimethyl-2-ethylbenzene 0.05 0.01–0.11 in Agriculture and Horticulture, Butterworths, London, 1982, 62 1,3-Dimethyl-4-ethylbenzene 0.04 0.01–0.09 pp. 27–42. 63 1,2-Dimethyl-4-ethylbenzene 0.04 0.01–0.13 4 Department of the Environment, Transport and the Regions, The 64 Undecane 0.05 0.01–0.08 Fourth Report of the Photochemical Oxidants Review Group, 65 1,2,4,5-Tetramethylbenzene 0.03 0.01–0.08 London, 1997. 66 1,2,3,5-Tetramethylbenzene 0.05 0.01–0.14 5 Expert Panel on Air Quality Standards (EPAQS), Benzene, 67 Naphthalene 0.05 0.01–0.14 Department of the Environment, HMSO, London, 1994. 68 Dodecane 0.03 0.01–0.08 6 W. P. L. Carter, J. A. Pierce, L. Dougmin and I. L. Malkina, Atmos. Environ., 1995, 29, 2499. 7 R.G. Derwent, Environ. Pollution, 1990, 63, 299. 8 First Report of the Quality of Urban Air Review Group (QUARG), Urban Air Quality in the United Kingdom, Department of the Environment, London, 1993. J. Environ. Monit., 1999, 1, 453–458 4579 Her Majesty’s Government, This Common Inheritance: Britain’s 15 C. M. White, J. Hackett, R. R. Anderson, S. Kail and P. S. Spock, J. High Res. Chromatog., 1992, 15, 105. Environmental Strategy, HMSO, London, 1990. 16 D. Estel, M. Mohnke, F. Biermans and H. Rotzsche, J. High Res. 10 J. Derwent, P. Dumitrean, J. Chandier, T. J. Davies, Chromatog., 1995, 18, 403. R. G. Derwent, G. J. Dollard, M. Delaney, B. M. R. Jones and 17 E. Kivi-Etelatalo, O. Kostiainenand M. Kokko, J. Chromatogr. P. D. Nason, Report No. AEA/CS/1 8358030/005, AEA A, 1997, 787, 205. Technology, Culham, Oxon., UK, 1994. 18 L. S. Ettre, Anal. Chem., 1964, 36, 31A. 11 J. Murlis, Proc. Int. Conf. on VOCs in the Environment, 19 R. H. Partridge, Proc. Int. Conf. on VOCs in the Environment, 27–28 October, London, 1993, pp. 85–98. 27–28 October, London, 1993, pp. 361–373. 12 P. Ciccioli, A. Cecinato, E. Brancaleoni, M. Frattoni and 20 R. G. Derwent, M. E. Jenkin and S. M. Saunders, Atmos. A. Liberti, J. High Res. Chromatog., 1992, 15, 75. Environ., 1996, 30, 181. 13 R. M. Smith, Gas and Liquid Chromatography in Analytical Chemistry, Wiley, Chichester, UK, 1988. 14 R. H. Brown and C. J. Purnell, J. Chromatog., 1979, 178, 79. Paper 9/04879K 458 J. Environ. Monit., 1999, 1, 453–458

ISSN:0960-7919    

DOI:10.1039/a904879k

出版商:RSC    

年代:1999

数据来源: RSC