Quantitative and qualitative sensing techniques for biogenic volatile organic compounds and their oxidation products

The physiological production mechanisms of some of the organics in plants, commonly known as biogenic volatile organic compounds (BVOCs), have been known for more than a century. Some BVOCs are emitted to the atmosphere and play a signi ﬁ cant role in tropospheric photochemistry especially in ozone and secondary organic aerosol (SOA) productions as a result of interplays between BVOCs and atmospheric radicals such as hydroxyl radical (OH), ozone (O 3 ) and NO X (NO + NO 2 ). These ﬁ ndings have been drawn from comprehensive analysis of numerous ﬁ eld and laboratory studies that have characterized the ambient distribution of BVOCs and their oxidation products, and reaction kinetics between BVOCs and atmospheric oxidants. These investigations are limited by the capacity for identifying and quantifying these compounds. This review highlights the major analytical techniques that have been used to observe BVOCs and their oxidation products such as gas chromatography, mass spectrometry with hard and soft ionization methods, and optical techniques from laser induced ﬂ uorescence (LIF) to remote sensing. In addition, we discuss how new analytical techniques can advance our understanding of BVOC photochemical processes. The principles, advantages, and drawbacks of the analytical techniques are discussed along with speci ﬁ c examples of how the techniques were applied in ﬁ eld and laboratory measurements. Since a number of thorough review papers for each speci ﬁ c analytical technique are available, readers are referred to these publications rather than providing thorough descriptions of each technique. Therefore, the aim of this review is for readers to grasp the advantages and disadvantages of various sensing techniques for BVOCs and their oxidation products and to provide guidance for choosing the optimal technique for a speci ﬁ c research task. have served as a main analytical tool for BVOC quanti  cation since the early stage of BVOC quanti  cation research. In addition, many di ﬀ erent techniques such as mass spectrometry and spectroscopic methods have emerged. New insights from the emerging techniques have revealed new perspectives on BVOCs photo- chemistry to regional and global air quality. This critical review concisely summarizes the current understanding of BVOC emissions and their atmospheric photooxidation processes along with discussion about their impacts on air quality and climate. The discussion shows how new emerging techniques provide important cluesin understanding BVOCphotochemistryinthe troposphere. Inaddition,principles,advantagesand disadvantagesofa wide varietyofanalytical techniquesarediscussedinthemanuscript.Therefore, webelievethatthiscriticalreviewwillserveasimportantguidanceespecially foremerging scientistswho are just starting their career in biosphere – atmosphere interaction research and a useful guide for established scientists who would like to review the overall development of analytical techniques

Quantitative and qualitative sensing techniques for biogenic volatile organic compounds and their oxidation products Saewung Kim, * a Alex Guenther b and Eric Apel b The physiological production mechanisms of some of the organics in plants, commonly known as biogenic volatile organic compounds (BVOCs), have been known for more than a century. Some BVOCs are emitted to the atmosphere and play a significant role in tropospheric photochemistry especially in ozone and secondary organic aerosol (SOA) productions as a result of interplays between BVOCs and atmospheric radicals such as hydroxyl radical (OH), ozone (O 3 ) and NO X (NO + NO 2 ). These findings have been drawn from comprehensive analysis of numerous field and laboratory studies that have characterized the ambient distribution of BVOCs and their oxidation products, and reaction kinetics between BVOCs and atmospheric oxidants. These investigations are limited by the capacity for identifying and quantifying these compounds. This review highlights the major analytical techniques that have been used to observe BVOCs and their oxidation products such as gas chromatography, mass spectrometry with hard and soft ionization methods, and optical techniques from laser induced fluorescence (LIF) to remote sensing. In addition, we discuss how new analytical techniques can advance our understanding of BVOC photochemical processes. The principles, advantages, and drawbacks of the analytical techniques are discussed along with specific examples of how the techniques were applied in field and laboratory measurements. Since a number of thorough review papers for each specific analytical technique are available, readers are referred to these publications rather than providing thorough descriptions of each technique. Therefore, the aim of this review is for readers to grasp the advantages and disadvantages of various sensing techniques for BVOCs and their oxidation products and to provide guidance for choosing the optimal technique for a specific research task.

Environmental impact
The role of biogenic volatile organic compounds (BVOCs) in controlling oxidation capacity and aerosol formation has been highlighted over the past ve decades. Many breakthroughs on quantitative and qualitative analytical techniques have provided important clues to understand how the minor chemical constituents in the atmosphere drive atmospheric photochemistry governing ozone and secondary aerosol formation. Gas chromatography techniques have served as a main analytical tool for BVOC quantication since the early stage of BVOC quantication research. In addition, many different techniques such as mass spectrometry and spectroscopic methods have emerged. New insights from the emerging techniques have revealed new perspectives on BVOCs photochemistry to regional and global air quality. This critical review concisely summarizes the current understanding of BVOC emissions and their atmospheric photooxidation processes along with discussion about their impacts on air quality and climate. The discussion shows how new emerging techniques provide important clues in understanding BVOC photochemistry in the troposphere. In addition, principles, advantages and disadvantages of a wide variety of analytical techniques are discussed in the manuscript. Therefore, we believe that this critical review will serve as important guidance especially for emerging scientists who are just starting their career in biosphere-atmosphere interaction research and a useful guide for established scientists who would like to review the overall development of analytical techniques for BVOC observations.

Introduction and scope of this review
Atmospheric chemistry research in the past several decades reveals that (1) in the global perspective, the emissions of biogenic volatile organic compounds (BVOCs) are around an order of magnitude higher than the VOC emissions from anthropogenic activities, 1,2 (2) photochemical oxidation processes of BVOCs directly affect ozone and aerosol productions and thus they have signicant implications for regional air quality and global climate, 3,4 (3) BVOCs play a signicant role in air quality even in some urban areas with high anthropogenic VOC and NO X (NO + NO 2 ) emissions, 5 and (4) in very remote areas with little inuence of anthropogenic pollution, BVOC photochemistry amplies concentrations of the hydroxyl radical (OH), 6,7 which is a universal tropospheric oxidant and determines the lifetime of most trace gases including methane an important greenhouse gas. These expansions of our understanding have been achieved through the development and application of analytical techniques for BVOCs and their oxidation products in eld and laboratory studies.
Technical breakthroughs in air sampling and quantication techniques indeed have expanded the horizon of our understanding of the atmospheric chemistry of volatile organic compounds. For example, cartridge-sampling techniques using solid sorbents (i.e. Tenax TA) allowed us to monitor VOC distributions in a wide spatial range so that BVOC measurements could be conducted in widespread geographical environments. The extensive observational results have served as an important basis in constructing BVOC emission inventories and emission models. 1 In addition, the development of a Proton Transfer Reaction-Mass Spectrometry (PTR-MS 8 ) technique, which can monitor a wide range of volatile organic compounds with high sensitivity in real time at a high sampling frequency, has enabled quantication of very reactive and low volatility compounds in the atmosphere, which have signicant implications for understanding tropospheric ozone and secondary aerosol formation processes.
In this review, we rst provide an overview of BVOCs and their photochemistry in the troposphere and discuss their implications in regional air quality and global climate. In the following sections, we review measurement techniques used for quantifying BVOCs and their oxidation products and discuss how the new observational dataset could expand our horizon in tropospheric BVOC photochemistry. We review most of the measurement techniques that have been used to quantify emission rates and ambient distributions of BVOCs and their oxidation products since 1960 when Went 9 rst suggested the importance of BVOCs in aerosol formation in remote forest regions. Since then, there have been a number of reviews on sensing techniques such as gas chromatography, chemical ionization mass spectrometry, and optical spectroscopy. Therefore the aim of this review is to succinctly introduce the currently available sensing techniques for BVOCs and their oxidation products, so that readers are fully informed about the advantages and disadvantages of each analytical technique and can choose the appropriate sensing techniques for their own research.
2 BVOCs in the regional air quality and the global climate system Haagen-Smit 10 rst presented a tropospheric photochemical reaction mechanism in an attempt to explain secondary photochemical product formations (ozone and aerosols) from anthropogenic precursors such as NO X (NO + NO 2 ), VOCs, and SO 2 . Due to a lack of understanding of the central role of the hydroxyl radical (OH) in tropospheric photochemistry, the proposed photochemical mechanisms speculated that molecular oxygen was the main oxidant in the troposphere although the mechanisms correctly contained the precursors (NO X and VOCs) and the products (ozone and organic aerosols) of photochemical smog. In 1960, Went speculated that BVOCs are a main source for the "blue haze" that had been observed since the sixteenth century (e.g. by Leonardo da Vinci) near forest areas especially during the summer season. This speculation was also supported by John Tyndall's (1820-1893) empirical nding that aerosol formation inside of a glass jar lled with organic vapor (alkene compounds) could be triggered by light and ozone. Went 9 argued that BVOCs, especially monoterpenes (C 10 H 16 ; MTs) with their multiple carbon-carbon double bonds in their chemical structure, are a major source of aerosols. MTs were widely known as important plant constituents long before 1960. 11 However, their existence in the atmosphere as an important gas phase component was controversial. Went 9 argued that the similarities between the ambient air smell near forest areas and smell from crushed pine needles were proof of MT's presence in the ambient air.
Quantication of BVOCs in various forest environments was rst reported by Rasmussen and Went. 12 They deployed a GC system coupled with a packed column (Diatoport S 60-80 mesh) and a hydrogen ame ionization detector in a mobile trailer. They drove the trailer around rural U.S. regions to quantify terpenoid species in the ambient air. The results showed that VOCs are present at up to a few ppbv in the summer season. In the following studies, Went et al. 13 demonstrated that BVOCs, especially terpenes, can be a signicant source for atmospheric particulate matter (Aitken condensation nuclei). The emissions of terpenoids from plants were conrmed by Rasmussen 14 from static leaf chamber measurements of ve different tree species (oak, cottonwood, sweetgum, eucalyptus and white spruce). In the study, the author used three different analytical methods (GC-FID, infrared spectrometry, and mass spectrometry) to conrm the presence of isoprene in complex organic vapor mixtures in the static chamber air.
In 1971, Levy 15 postulated that tropospheric OH present at very low levels (usually less than 10 7 molecules per cm 3 out of 2.5 Â 10 19 molecules per cm 3 at standard temperature and pressure) governs tropospheric photochemistry. This nding triggered a series of laboratory studies to characterize the reaction kinetics of atmospheric oxidants (e.g. OH and ozone) and BVOCs (e.g. Westberg and Rasmussen 16 ). In addition, comprehensive research was conducted in the 1970s to characterize ambient distributions of natural hydrocarbons, especially terpenoid species (isoprene, hemiterpenoid and monoterpene, monoterpenoids) to assess their fate in the atmosphere and impacts on regional and global air quality. Graedel 11 reviewed the understanding of BVOCs and their photochemistry in the atmosphere that had been accumulated by the late 1970s. At that time, the consensus of the scientic community was that isoprene and a few monoterpenes are the most dominant BVOCs in the atmosphere among the 20 organic compounds that had been conrmed as vegetation emissions. The unit of mg g À1 h À1 was typically used to express emission rates of those compounds at that time. 17,18 The laboratory characterization of oxidation reaction rates of selected BVOCs with atmospheric oxidants was rst conducted by Winer et al. 19 In addition, several laboratory smog chamber and theoretical studies elucidated reaction mechanisms, products and aerosol yields from BVOC oxidation. 16,[20][21][22][23][24] Those results showed that as oxidation proceeds, the vapor pressure of reaction products tends to be signicantly decreased. Those less volatile oxidation products, therefore, are more partitioned into the particle phase. Estimation by Duce 25 based on those laboratory observations indicated that global secondary organic aerosol mass burden (60-140 MT OC per year) is much higher than the estimated burden from the anthropogenic source ($20 MT OC per year). However, Weiss et al. 26 reported contradictory aerosol composition observations that indicated sulfate compositions, mostly from anthropogenic activities, dominated in several rural locations in the U.S. where one would expect high BVOC emissions to trigger organic aerosol formation. These contradicting results urged more comprehensive research including inter-phase partitioning characterizations. Another controversy, caused by insufficient understanding of gas phase oxidation mechanisms of BVOCs in the late 1970s, was the contribution of BVOCs to regional and global photochemistry, specically ozone formation. For example, Gay et al. 27 and Lonneman et al. 28 concluded that BVOCs are not important for local and regional photochemistry even in forest areas. In contrast, Zimmerman et al. 29 speculated that BVOCs were a signicant source of global CO and H 2 , by-products of BVOC oxidation processes, and thus important for atmospheric photochemistry.
The rst comprehensive discussion about the impacts of BVOCs on regional air quality was presented by Altshuller. 30 Previously, Arnts and Meeks 31 concluded that BVOCs were a minor contribution to total VOCs even in rural places (e.g. Rio Blanco County Colorado and the Smoky Mountain Region in the U.S.) based on a relatively small measurement dataset. Lonneman et al. 28 also claimed that BVOCs did not contribute to ozone formation in Florida based on VOC measurements in multiple locations (e.g. St. Petersburg/Tampa and Miami, FL). The analysis by Altshuller 30 was conducted using BVOC and AVOC emission inventories in the United States. In addition, Altshuller 30 comprehensively reviewed a number of eld VOC measurement and laboratory (smog chamber) experiment results to assess the ozone and aerosol formation potential from the oxidation of major AVOCs and BVOCs. The emission inventories introduced in the early 1980s were mostly constructed using datasets from branch enclosure measurements. 17,18 Based on the small number of experimental points, emission rates from regional scales were estimated. Those bottom-up emission inventories were tested several times by top-down estimations from eld observations of chemical and micrometeorological parameters. The results indicated some general agreement but in some cases there was a major discrepancy between top-down and bottom-up BVOC emission estimations (À19% to 400% 32,33 ). Ambient measurements of both biogenic and anthropogenic VOCs in rural areas were also used to evaluate emission inventories of BVOCs and AVOCs. The analysis showed a higher degree of underestimated reactive AVOC (e.g. propene) emissions up to 100 times 28,31,34-36 compared to underestimated reactive BVOC emissions such as isoprene and a-pinene. Altshuller, 30 therefore, suggested the two possibilitieseither underestimated AVOC or overestimated BVOC emissions could cause these discrepancies although anecdotal evidence for unmeasured BVOCs especially oxygenated compounds was reported. 17,18,37 From the assessments of emissions and oxidation processes using the best available knowledge at that time, Altshuller 30 concluded that BVOC photochemistry contribution to ozone and aerosol levels in the U.S. was insignicant compared to contributions from AVOC photochemistry.
These conclusions, however, were not conrmed by follow up mechanistic photochemical modelling studies constrained by eld observations. Trainer et al. 38 interpreted a eld measurement data suite of NO X , VOCs both anthropogenic and biogenic, and ozone observations from a rural research site in summer (Scotia, Pennsylvania) using a one-dimensional photochemical model. The study presented three different model runs using the following ozone precursors: (1) NO X only, (2) NO X and BVOCs and (3) NO X , BVOCs and AVOCs and compared the model results to the observed ozone levels. The results indicated that most of the photochemical ozone formation was from BVOC photochemistry, which enhances HO 2 levels and other organic peroxy radicals (RO 2 ). This conclusion is quite contradictory to the previous several studies and the authors explained that non-linearity of photochemistry which has not been considered in the previous studies led to this new conclusion. In the following year, Chameides et al. 5 demonstrated that ozone formation in the urban area of Atlanta, Georgia U.S.A. was controlled by BVOC photochemistry. With detailed photochemical model calculations, the authors demonstrated that policy implementation for photochemical ozone pollution control should consider BVOC photochemistry. For example, a photochemical model scenario without BVOCs indicated that 30% reduction of AVOCs in the Atlanta area could reduce ozone to the EPA compliance level, 120 ppbv in one hour average but once BVOC photochemistry was taken into consideration then a 70% reduction was required to achieve the same goal. In addition, Chameides et al. 39 demonstrated that the concentration-based evaluations of impacts on tropospheric photochemistry from VOCs in the previous studies could be misleading because of a wide range of photochemical VOC reactivities. Chameides et al., 39 therefore, developed the concept of "propylene-equivalent concentration", which is a normalized concentration scale with respect to the propylene reactivity towards OH. The analysis of measurement datasets in various urban and remote locations using the propylene-equivalent concentration scale indicates that BVOC contributions, especially isoprene and MT, to photochemistry had been signicantly underestimated not only in forested rural areas but also in urban/suburban areas, where the importance of BVOCs in photochemistry had been ignored. In addition, Chameides et al. 39 also speculated about the possibility that there were signicant unmeasured VOCs, especially oxygenated compounds, that may play a signicant role in photochemistry.
As Went rst hypothesized, important roles of MT in aerosol formation have been conrmed by laboratory and eld experiments since 70s. However, the isoprene contributions to global aerosol formation had been ignored because most of the rst generation isoprene oxidation products (methyl vinyl ketone, methacrolein, and formaldehyde) are highly volatile and early aerosol chamber experiments observed no secondary aerosol formation. This assumption, however, has been overturned by recent eld 44-46 studies and laboratory [40][41][42][43] experiments with more realistic chamber experimental conditions. Considering the dominance of isoprene in BVOC emissions (about half of the global total 1 ), 2 a few percent of aerosol yields from isoprene is still signicant in the global scale. Based on the latest research results, Hallquist et al. 4 estimated the amount of secondary organic aerosol loading from BVOCs including isoprene and MT to be 88 TgC per year, which is signicantly higher than that from AVOCs (10 TgC per year).
A comprehensive isoprene photooxidation process was presented by Paulson and Seinfeld. 47 Atkinson and Arey 48 reviewed oxidation mechanisms and reaction rate constants of other BVOCs including monoterpenes for atmospheric oxidants such as OH, ozone and NO 3 . In the past few years, new ndings on the isoprene photo-oxidation mechanisms revealed isoprene oxidation products that previously have not been known or observed. For example, Paulot et al. 42,49 detected a number of new rst generation oxidation products of isoprene oxidation by OH such as C5-and C4-hydroxycarbonyl compounds, hydroxy acetone, glyoxal and methylglyoxal. Wolfe et al. 50 and Taraborrelli et al. 51 presented critical reviews on recent breakthroughs on isoprene oxidation mechanisms and products. These new ndings on isoprene oxidation chemistry suggest that further research is necessary for other relatively less studied BVOCs such as monoterpenes and sesquiterpenes.
Early research (1970s) probing BVOC photochemistry was mostly dependent on traditional GC techniques. As new techniques became available, including advanced GC techniques, chemical ionization mass spectrometry (CIMS), and optical techniques, new experimental ndings expanded our understanding of BVOC tropospheric photochemistry. In addition, calibration methods have been advanced to quantify semivolatile and very reactive compounds that have signicant implications for photochemistry and aerosol formations. In the following sections, we will review technical advances that have taken place during the past several decades for quantifying BVOCs and their oxidation products and how those advances have expanded our understanding of BVOC photochemistry in regional and global air quality and climate.

Sensing techniques before 1990
The most commonly used analytical technique for BVOC measurements has been GC. Two early qualication and quanti-cation studies of isoprene and monoterpenes 12,14 used a GC system with a packed column. This was coupled with a sampling loop for sample injections to the GC system. In these early studies, the big challenge in applying a GC method to BVOC sensing was identifying BVOC peaks from chromatograms. The challenge was addressed by applying analytical tools such as mass spectrometry and FT-IR spectroscopy. As shown in Fig. 1, the major peaks from branch enclosure samples were identied by the multiple qualitative analytical tools. Interestingly, the olfactory sense was useful for identifying BVOCs in these early studies. Rasmussen and Went 12 noted that "Some terpenes, commonly formed in plants, were tested to determine their retention times on the chromatographic columns used, making it possible to identify the odors found in free air, at least tentatively." Similarly, Rasmussen 14 wrote "Some of observations from smelling these accumulated foliage volatiles at concentrations of a few ppm are of note.
------The description given most oen was "smells like an oak forest." The response for smelling nonisoprene releasing foliages like maples, ash, birch, and alders was "smells like fresh leaves."" These studies used the Flame Ionization Detector (FID), which is still commonly used for VOC quantications. The sensitivity of the FID (in a standard conguration) to molecules of the light alkanes (C 1 to C 8 ) shows a nearly proportional response to the number of C atoms per mole. [52][53][54][55] The response for alkenes may be slightly altered but proportional to the carbon number within 10%. The response for alkynes may be dramatically different and dependent on the ame conditions. 54,56 Also there is a reduced response to oxidized groups such as carbonyl or carboxyl groups, 57 and the response to carbons attached to hydroxyl groups or halogens is reduced. Many investigators have taken advantage of the proportional response and simply report concentrations of a myriad of hydrocarbons referenced to a particular standard hydrocarbon (e.g., propane from the National Institute of Standards and Technology (NIST)). Other laboratories do not rely on this proportional response of the FID. These laboratories generate a standard for each compound that they wish to analyze, and calibrate their instruments accordingly.
Whitby and Coffey 58 reported ambient concentrations of speciated monoterpenoids (a-pinene, b-pinene, and p-cymene) in ambient air of the Adirondack Mountain region, New York USA. The GC instrumentation for this study was equipped with two nearly identical columns for back-ushing and a liquid nitrogen (LN 2 ) cryogenic concentrator. The sample concentration technique could provide more sensitivity so that relatively small amounts of MTs in the range of $64 pptv in the morning and $1.7 ppbv during the aernoon were quantied.
However, single-stage cryogenic concentrating methods have drawbacks of water condensation and aerosol formation. 59 The concentrated water in the trap can signicantly interfere with the FID signals. Therefore, multi-stage preconcentrating methods or solid sorbent concentration techniques that do not have the water interference issue have gained a wide interest in the VOC measurement community. An early study by Pellizzari et al. 59 tested several candidate sorbent materials (polymer beads, activated carbons, and liquid phases) and reported that Tenax GC (2,6-dimethyl-p-phenylene oxide) and Chromosorb 101 (styrene-divinyl benzene) showed good adsorption efficiencies for a wide range of compounds (NMHCs and oxygenated compounds) and ow rates (up to 9 slm). The study concluded that Tenax GC and Chromosorb are adequate substances for cartridge sampling. It was noted that Tenax GC has an advantage over other materials because of its stability at high temperature and relative insensitivity to water vapor. Brown and Purnell 60 investigated necessary sampling volumes as a function of sampling ow-rate, sample concentration, temperature, and relative humidity for a wide range of analytes (hydrocarbons, chlorinated hydrocarbons, other halogenated hydrocarbons, esters, aldehydes and ketones, alcohols, acids and anhydrides, and amines). The results indicated that optimal sampling volumes are a strong function of ow-rates, concentrations of VOCs, and sample temperatures. Helmig et al. 61 explored additional materials and conrmed again that Tenax shows the most ideal performance in terms of the desorption efficiency, the adsorption capacity, and minimized blanks. Since then there have been major advances in sorbent sampling technology. Sorbents having a wide range of characteristics including temperature stability, hydrophobicity/ hydrophilicity, and retention strength are now available. These may be used singly but are oen used in combinations in modern applications and can be tailored for specic measurement applications, e.g., OVOC, low, midrange, and semivolatile NMHC measurements, etc. For a comprehensive review of the subject and additional details see Woolfenden 62,63 and references therein.
Another important aspect to be considered for accurate VOC measurements using cartridge sampling is chemical transformation of the trapped and sorbent molecules while sampling is conducted. The reason is that the molecules adsorbed to the sorbent bed are exposed to the ambient airstream for a signicant time period so oxidants in ambient air can transform reactive VOCs to their oxidation products. Pellizzari and Krost 64 and Pellizzari et al. 65 reported laboratory experiment results of those surface reactions between oxidant and adsorbed molecules. The results indicated that a measurable amount of benzaldehyde, acetophenone, and phenol were detected from sorbent material oxidation by ozone and adsorbed analytes can be signicantly transformed to oxidized compounds by ozone. The degree of transformation was highly dependent on the reactivity of the analytes (e.g. alkenes showed more loss compared with alkanes). To inhibit those unintended reactions, Pellizzari et al. 65 tested several lter options and suggested that lters for oxidants should be carefully selected because the lter material may interfere with atmospheric VOCs to be sampled.
Canister samplers (whole air samplers) have also been used for BVOC sampling. 66 The canister sampling method has several advantages over the sorbent sampling method such as (1) sampling without breakthrough, and (2) no interactions with the sorbent material. 67 However, more care is required for canister preparations (e.g. samplings and canister clean up) than that for the sorbent sampling method. In addition, some compounds are not stable in some sampling canisters and should be carefully assessed by laboratory experiments. A major disadvantage is the much larger volume required to transport samples. In general, the canister sampling method has been used for compounds below C 11 , which include hemiterpenoid (isoprene and 2-methyl-3-buten-2-ol) and monoterpenoid BVOCs. 68 Although some studies extend the compound list up to C 13 , 69-71 Zielinska et al. 68 presented intercomparison results between canister and sorbent samplings that indicated signicant sampling losses in the compound ranges C 11 and higher when canister sampling was applied. Apel et al. 72 used electropolished SS canisters as the basis for their Non-Methane Hydrocarbon Intercomparison Experiment (NOMHICE). Individual canisters containing an ambient air sample were prepared and analyzed and then sent to participants for analysis and then reanalyzed upon return to the reference laboratory at the National Center for Atmospheric Research (NCAR) in Boulder, Colorado U.S.A. Compounds were found to be stable through C8 over a period of months but >C8 species tended to show some losses. Since that time, there have been advances in electropolishing technology (Don Blake, personal communication) and in deactivating the surface of the canisters with proprietary ceramic coatings to reduce the potential for chemical adsorption. Bags (mostly Tedlar or Teon) have also been used for whole air sampling but Wang et al. 73 reported the possibility of signicant sampling losses within 24 hours which must be considered when using this technique.

Sensing techniques aer 1990
As important roles of BVOCs in regional photochemical pollution were elucidated in the late 1980s, several intensive measurement campaigns to understand the roles of BVOCs in regional air-quality were conducted in the 1990s (e.g. Southern Oxidants Study (SOS)). Although GC-FID and GC-FID-MS techniques were used as the main tools to identify and quantify BVOCs in the atmosphere, growing interest in quantifying oxygenated compounds required new analytical techniques. GC-based techniques for oxygenated compounds are challenging due to their low volatility and surface reactivity. In addition, importance of constraining semi-volatile BVOCs such as sesquiterpenoids (C 15 H 24 ) grew as they have signicant implications for secondary organic aerosol formation. 74 To quantify those compounds, sensing techniques should meet several conditions such as (1) minimal contact of samples with the sampling and analysis interface to minimize wall loss, (2) real time measurement to avoid any chemical transformation, and (3) fast response for applications such as eddy-covariance ux observations and airborne observations.
3.2.1 Advanced GC techniques. Quantication of oxygenated compounds using a GC technique was well described by Goldan et al. 75 The main technical details that they adapted were: A Teon sampling inlet to prevent the loss of hydrophilic and reactive compounds.
In-line removal of ozone to prevent sample loss during the cryo-trap process, especially, alkene compounds.
An in-line water trap to prevent the loss of water soluble compounds on the column (polar compounds especially oxygenated compounds).
A water trap to remove potential interferences was installed as 1/4 00 o.d. Teon tubing (50 cm), and maintained at À50 C with dry ice as shown in Fig. 2. Goldan et al. 75 reported the possible sample loss for more polar compounds such as light alcohols could be up to 15% due to the water trap. Other water removal techniques such as using a Naon dryer have also been used but those techniques potentially cause the signicant chemical structure rearrangement of analytes. 76 Goldan et al. 75 also investigated the possible sample loss for BVOCs by ozonolysis reactions. The test results indicated that reactive BVOCs such as isoprene ($500 pptv) and alpha-pinene ($500 pptv) can be completely scrubbed away by ozonolysis with one liter of ambient air that contains 100 ppbv of ozone. To prevent sample loss from ozone, Goldan et al. 75 used a Na 2 SO 3 lter to remove ozone. Detailed description of various alternative ozone scrubbing methods can also be found in Helmig. 77 Singh et al. 78,79 developed an aircra-based gas chromatographic technique to measure acetone, ethanol, methanol, methyl ethyl ketone, acetonitrile, hydrogen cyanide, acetaldehyde and propanal. These species were measured on a column coupled with a photoionization detector (PID) and a reduction gas detector (RGD) placed in series. Typically, a 200 mL aliquot of air was cryogenically trapped at À140 C prior to analysis. Moisture was greatly reduced by passing air through a water trap held at À40 C during sampling and À50 C between samples. Laboratory tests were performed to ensure the integrity of oxygenates during this drying process. The calibration standards were added to the ambient air stream in the main manifold and were analyzed in a manner that was identical to normal ambient sampling. This procedure was designed to compensate for any line losses. An on-board dilution system allowed varied concentrations to be prepared. The limit of detection of reactive nitrogen species was estimated to be 1 ppt, while that of other oxygenates was 5-20 ppt. There was an indication of artifact OVOC formation under high ozone concentrations in the stratosphere.
Apel et al. 72 developed a fast GC-MS technique to measure oxygenated organic compounds for deployments on aircra platforms. The system used a cryogenic water trap, an enrichment trap and a cryofocusing trap to effectively concentrate the analytes. The key for fast analysis is to synchronize the sample preparation time (cryotrapping, water removal, cryofocusing) with the analysis time, which includes GC run-time and cool down time. A custom-built cryogenic system and low thermal mass GC were developed to accomplish the instrument objectives. For the GC cycle, the authors adopted a temperature program starting at 30 C for 10 seconds and ramping up to 120 C at the rate of 100 C min À1 followed by a fast cool down procedure. Possible interferences for light oxygenated hydrocarbon compounds were thoroughly investigated. The investigators set up a glass manifold system and generated a number of light oxygenated compounds using a dynamic dilution system between a compressed standard mixture and zero air. In the length of $2 m manifolds, Apel et al. (2003) added a number of possible chemical interferences such as ozone, NO X , SO 2 and water vapor. The series of tests indicated that the Teon diaphragm in some pumps can be a source for ethanol from degassing. This problem was easily addressed by replacing the pump with a metal bellows pump. In addition, this study reported signicant inferences from ozone for acetaldehyde, propanal, acetone and butanal quantications. On the other hand, alcohols and methyl ethyl ketone were not interfered by ozone. The source of the inferences appeared to be the rotor material in the VICI valves and they found that the Teon (PTFE) composite material was most inert. The degree of the interferences was a strong function of ozone concentrations so concurrent ozone measurement data were used to correct the ambient OVOC observational dataset. Second and third generation instruments were subsequently developed by Apel et al. 80 resulting in a technique that is able to provide full GC-MS analyses of over 45 trace species every 2 minutes at near or sub-pptv levels. Measurements include C 2 to C 9 carbonyl compounds, alcohols, and additional 25 or more compounds such as NMHCs, nitriles and halogenated compounds. 80 The system has been deployed in a series of aircra eld campaigns to characterize distributions of VOC species in the troposphere and chemical transformation of VOCs on regional and global scales. 72,80 Recently, the system has been deployed at a groundbased site measuring primarily biogenic species including terpenoids that can vary considerably over short time periods. 81 A critical aspect for quantifying reactive and semi-volatile organic compounds is the preparation of accurate reference calibration gases. In general, for oxygenated compounds, compressed gas mixtures have been used for calibration standards. Goldan et al. 75 tested 30 VOCs of different classes (alkanes, oxygenated hydrocarbons, aromatics, alkenes, monoterpenes, and cyclic hydrocarbons), which were gravimetrically prepared in an AculifeÔ treated aluminum cylinder with a batch of 4-6 compounds. They were prepared at 10 ppmv levels and a dynamic dilution system allowed multi-point calibration in the range of 0.2 to 2 ppbv. Apel et al. 82 described the preparation of OVOC standards in high-pressure cylinders in the ppmv range and their subsequent validation. Another popular way to prepare standards is using diffusion sources. To generate commercially unavailable compounds using pure liquids, Helmig et al. 83 developed a capillary diffusion calibration system. They built a constant temperature housing for multi-channel diffusion sources for a number of sesquiterpene and oxygenated-sesquiterpene compounds. Those compounds were kept in glass vials connected to glass capillaries. Since diffusion rates from the vials is a function of pressure and temperature, one can expect constant diffusion rates under the constant temperature and pressure environment. A nitrogen dilution ow swept the diffused ow to the outlet, branched to an online GC system monitoring the concentrations of the generated standard samples. The calibration system was used to evaluate analytical characteristics and potential interferences by ozone in sesquiterpene sampling of sorbent cartridges. 84 Results from the study clearly indicated that with proper ozone removal, sesquiterpenoid compounds can be quantied by conventional sorbent cartridge sampling techniques and a number of studies have clearly demonstrated sesquiterpenoid quantications with GC techniques especially for branch enclosure samples. 85 Finally, Bouvier-Brown et al. 86 presented a eld deployable GC system for ambient sesquiterpenoid measurement. The whole analytical system is kept in a temperature controlled container ($50 C) to prevent the wall loss of semi-volatile compounds. The system was also equipped with an ozone removal lter and an on-line ozone monitor to prevent chemical loss. Bouvier-Brown et al. 86 used a micro-syringe injection method for the on-site sensitivity calibration of sesquiterpenoid compounds. A small amount of liquid (0.25-1 microliter) was injected upstream of the zero air to produce calibration standard samples in mixing ratios of 2-100 pptv.
The potential of GC Â GC technology for the measurement of VOCs in ambient air was rst demonstrated by Lewis et al. 87 showing that a large number of atmospheric species go undetected by conventional chromatographic methods; more than 500 individual components were separated in an urban air sample. GC Â GC instrumentation consists of an injector, primary column, modulator, secondary column, detector and data processor. In essence the GC Â GC separation is a normal GC separation (primary) followed by a steady repetition of secondary GC separations. The selectivity of the primary and secondary stationary phases is chosen to be different and the timescale of the secondary separations (the modulation period) is small enough to only minimally diminish the primary separation resolution. Thus, a consistent portion of each peak emerging from the primary stationary phase is transferred to the secondary phase (column) so that the total area of the secondary phase is representative of the component concentration. The resulting 2-D chromatogram has peaks scattered about a plane rather than along a line. Good modulation, which is critical to the performance of GC Â GC systems, is achieved through the precise control of ow. A great deal of effort has been devoted to developing precise, reliable and robust modulators which have come to fruition and are key components in recently introduced GC Â GC products from a variety of vendors. Xu et al. 88 applied the technique to the measurement of C7-C14 organic compounds including monoterpenes. To aid in peak identication, the GC Â GC system was coupled with a time-of-ight (ToF) mass spectrometer: approximately 650 peaks were identied. Recently, the technique has also been used to analyze the organic composition of aerosols. 89,90 Bartenbach et al. 91 recently deployed a GC Â GC-FID technique in a rural agricultural and forested area in southern Germany and found good quantitative agreement with a conventional GC-MS technique in the measurement of both anthropogenic and biogenic VOCs. In addition, for light oxygenated-VOC (OVOC) quantications in the ambient air, a high-performance liquid chromatography (HPLC) technique has also been utilized. 92 This application requires stripping light OVOCs into a solvent stream that is analyzed by HPLC for its ion contents. 93 3.2.2 Chemical ionization mass spectrometry (CIMS). CIMS techniques have been popular for a wide range of pure and applied chemical research such as ion chemistry, biochemistry, medical and environmental applications. Chemical ionization is a representative low-energy ionization technique (so ionization) with no or mild fragmentation. This analytical characteristic can provide clearer molecular information and precise quantication. 94 Although the technique was introduced in 1966, 95 the application in atmospheric chemistry started in the late 1970s. 96,97 Since then, a number of ion chemistry and ion-neutral reaction chamber congurations have been developed to measure inorganic (such as nitric acid, OH/HO 2 , and H 2 SO 4 ) and organic (such as alkenes, aromatic compounds, peroxy radicals, acetonitrile, and more) compounds in the atmosphere with very low detection limits (tens of pptv or lower) and a very fast time response ($1 s or less). 98 Protons Transfer Reaction-Mass Spectrometry (PTR-MS) is one of the most widely used CIMS techniques for VOC quantications. This technique utilizes proton transfer reaction by applying an excess amount of protonated water ions (reagent ion; H 3 O + ). 99 A very comprehensive review of the PTR-MS technique including principles of the technique and various applications can be found in Blake et al. 100 In addition, a critical review on atmospheric chemistry applications is given by de Gouw and Warneke. 8 From an ion thermochemistry perspective, the PTR method is applicable for any gas molecules with a higher proton affinity than water (691 kJ mol À1 ). Most VOCs except alkanes have a higher proton affinity than water. This universal detecting ability gives PTR technology a great advantage over other selective CIMS techniques. From a reaction kinetics perspective, the sensitivity of VOCs depends on proton transfer reaction rates between volatile organic compounds and hydronium ions. The reaction rate constants of some of the stable compounds can be assessed by empirical methods using standard samples such as standard gas mixtures or diffusion sources. However, the growing interest in the quantication of very reactive or semi-volatile compounds, which are difficult to prepare as conventional standard samples, has resulted in the use of an alternative theoretical method to calculate reaction rate constants. The average dipole moment (ADO) theory is the most widely used method. This method estimates reaction rates by considering ion induced dipole interactions with both polar and non-polar molecules. 101 Several studies have assessed the sensitivity of molecules towards PTR-MS and other CIMS techniques using the ADO theory. [102][103][104] Notably, Zhao and Zhang 105 and Paulot et al. 42 presented comprehensive intercomparison results between theoretical and empirical values of polarizability (a) and permanent dipole moment (m D ) for the molecules, particularly relevant with tropospheric photochemistry.
Quadrupole-Secondary Electron Multiplier (SEM) has been used as a detection method for the protonated ions (PTR-QMS; Fig. 3a). The ion-throughputs of the detection unit decrease signicantly towards higher mass ions. The nature and the mathematical description of the mass discrimination were well established in the 1970s 106,107 and several studies have characterized the mass discrimination behavior of PTR-MS using empirical methods. Early publications 108,109 monitored the ratios between H 3 O + and VOC$H + signals by standard sample additions to characterize the mass discrimination behavior. However, several recent studies 103,110 demonstrated a method obtaining a transmission curve for the mass range (m/z) 70 + to 205 + using aromatic compound standard gas mixtures. Those compounds show no pronounced fragmentation and have uniform proton transfer reaction rate constants. Therefore, normalized signal distributions over the mass range can be directly interpreted as sensitivity differences from mass discrimination.
In addition, since mass-spectrometry techniques using a quadrupole mass lter detect compounds only as a function of molecular mass, signicant overlaps may occur if there are multiple compounds corresponding to one given nominal mass. In addition, potential fragmentation of high mass compounds (m/z À100 and higher) could hinder the identications of molecules. For this reason, a number of studies have explored the possible interference from isobaric compounds or fragments for the conventionally monitored masses. One way to address this issue is by interfacing a GC upstream of the PTR-QMS system. [111][112][113][114] These GC-PTR-MS studies explored common masses that were routinely monitored in urban and rural areas for anthropogenic and biogenic VOCs and their oxidation products. The results indicate that users should take extra care in interpretation of the datasets due to interferences from isobaric compounds. Comprehensive information about the potential interferences is provided by de Gouw and Warneke. 8 Also, several recent studies have reported fragmentation of high mass compounds and their possible interferences on other masses. For example, Lee et al. 115 presented aerosol chamber observation results of monoterpenoid and sesquiterpenoid compounds by PTR-QMS. Fragmentation patterns of the terpenoid compounds and their oxidation products are described and they discussed possible implications for ambient measurements to identify unknown peaks in ambient air. Kim et al. 116 presented laboratory experiment results about fragmentation patterns of seven different sesquiterpene species and their possible interferences on other masses as a function of analytical parameters such as relative humidity and energy level in the ion reactor (dri tube).
Recently, a new PTR application was introduced using a Time-of-Flight (ToF) detector. 117,118 This adaptation could overcome most of the drawbacks of the quadrupole-SEM detector unit by taking advantage of greatly enhanced mass resolution (m/Dm ¼ 6000) and no mass discrimination towards higher mass ions of the ToF detector (Fig. 3b). These signicant advantages were demonstrated in eddy covariance (EC) ux measurements for BVOCs and their oxidation products (17 compounds) in a mass range of 33 to 205 (m/z) over a grass land. 112,119 This emerging analytical technique, however, requires signicant data mining efforts to deduce ambient concentrations from raw data. Cappellin et al. 120,121 described physical and mathematical descriptions of each step for data mining from raw datasets.
Other ionization methods with different reagent ions have also been used for sensing BVOCs in the atmosphere. Slusher et al. 122 demonstrated that the I À ion is a suitable reagent ion to detect peroxyacyl nitrates (PANs) such as peroxyacetyl nitrate (PAN), peroxypropionyl nitrate (PPN) and peroxymethacryloyl nitrate (MPAN). Recently Veres et al. 123 developed negative-ion proton-transfer chemical ionization mass spectrometry (NI-PT-CIMS) to analyze organic acids in the atmosphere. The ion chemistry uses an acetate ion to take hydrogen away from organic acids as follows Acetate ion + H-R / acetic acid + R À Thermochemically, any acids with a lower acidity than acetic acid can be quantied by the technique. Considering the high acidity of acetic acid, the ion chemistry can be applicable to many different organic acids and an application for inorganic acids was recently presented. 124 Paulot et al. 42 presented a CIMS technique using CF 3 O À . CF 3 O À is a relatively non-selective reagent ion, previously used for peroxide analysis. 125,126 The authors characterized two different ion-neural reaction pathways for isoprene oxidation product detections: Paulot et al. 42 identied 38 oxidation products of isoprene that have not been routinely quantied by other techniques from laboratory isoprene oxidation experiments, which have signicant implications to SOA formation in the background atmosphere 49 and oxidation capacity.

Optical techniques.
Optical techniques such as chemiluminescence, fourier transform infrared spectroscopy (FT-IR), cavity ring-down laser absorption spectroscopy (CRLAS), tunable diode laser spectroscopy (TDLS), differential optical absorption spectrometry (DOAS) and laser induced orescence (LIF) have been used mostly to quantify relatively small oxygenated VOCs that are not easily detected by GC or mass spectrometry techniques.
Zimmerman and Hills 127 demonstrated a fast isoprene sensing technique using chemiluminescence reactions: An instrument with a sampling frequency of $2 Hz was developed to quantify the photons from relaxation of excited formaldehyde (CH 2 O*) and glyoxal (HCOCHO*). 127 This fast response system was utilized for canopy scale ux measurements (EC). 128 FT-IR is a multiplex instrument that is a single channel instrument with an ability to simultaneously scan over a wide wavelength range. This technique has advantages over IR spectroscopy such as higher throughput and wavelength accuracy (Skoog and Leary, 1992 (ref. 129)). The applications of the FT-IR technique in atmospheric chemistry are thoroughly reviewed by Tuazon et al., 130,131 Marshall et al., 132 [136][137][138] The references contain specic information about chamber preparations and quanti-cations of product yields and reaction rates. The compilations of the laboratory experiment results on BVOC oxidation reaction rates and products can be found in Atkinson et al. 48 TDLS uses a laser light source that can be tunable over a small wavenumber range with a very narrow line width. One of the representative TDLS applications for BVOC related research is the quantication of formaldehyde (CH 2 O). Fried et al. 139 presented TDLS instrumentation for CH 2 O measurement utilizing two wavenumber absorption in 2912.0918 cm À1 and 2914.4598 cm À1 . The path length was achieved at $100 m and interferences from atmospheric water vapors were carefully characterized. The method has been successfully deployed in both ground and aircra measurements. [140][141][142] Recently, CH 2 O quantication by LIF was applied for the canopy scale ux observations by the EC technique. 143 The technique demonstrated an excellent lower detection limit ($300 pptv in 1 second average). A laser induced phosphorescence (LIP) instrument was developed by Huisman et al. 144 for atmospheric glyoxal observations. They used the laser induced phosphorescence (LIP) instrument to achieve a limit of detection of below 10 pptv (1 min average, S/N ¼ 3). The method has been deployed in chamber and eld studies to characterize aerosol formation 145 from BVOCs and explore photochemistry inside of forest canopies. 146 The DOAS technique adopts broadband light sources such as a high-pressure Xe, incandescent quartz-iodine lamp, a broadband laser, the sun, and the moon. Spectral radiation through the path length is monitored to calculate mixing ratios of atmospheric constituents using the Beer-Lambert law. The instrumentation and its application for atmospheric chemistry are thoroughly reviewed by Platt 147 and Plane and Smith. 148 From the 1990s, the technique has been used to report ambient concentrations of CH 2 O. Recently, Volkamer et al. 149 developed a method to quantify ambient glyoxal (CHOCHO) using long path DOAS (LP-DOAS). Glyoxal is a primary oxidation product of many different VOCs and the direct measurement of glyoxal can constrain photochemical models to assess regional and global photochemical O 3 and SOA formations from VOCs. Indeed, ambient glyoxal observational results in Mexico City clearly indicated that our current understanding of SOA formation in the urban area signicantly underestimated the observed aerosol loading. 150 A global modeling study 151 demonstrated that almost half of the glyoxal is coming from biogenic VOC oxidation processes on the global scale and urged more observations in various environments. CRLAS quanties spectral absorptions from specic analytes. It improves the sensitivity of a conventional infrared spectrometer by trapping a laser pulse in a highly reective detection cavity resulting in an effective path length of many kilometers. CRLAS techniques 152,153 have been used to quantify small BVOC molecules such as ethene, formaldehyde, and glyoxal at the levels typically observed in BVOC emission enclosure systems and at the high end of the range of ambient levels. This can be accomplished using CRLAS without any preconcentration. However, further improvements in sensitivity are needed to make this technique more widely applicable for BVOC applications. In addition, LIF and CRDS NO 2 (ref. 154 and 155) quantication techniques have been utilized to quantify peroxy radicals 156 and alkyl nitrates 157 in the ambient air by applying chemical conversions up-stream of the sample air.
Optical sensors deployed in space-based platforms are providing an important constraint for global distributions of some important BVOCs. Satellite derived formaldehyde column data have been used in a number of studies 158 to infer emissions of isoprene, the globally dominant BVOC. Formaldehyde is a high yield product of isoprene and variations in formaldehyde in many regions are dominated by isoprene oxidation. Satellite observations of glyoxal, an oxidation product of many VOCs, also have a considerable potential for constraining BVOC emissions and distributions. 149 Global methanol distributions have recently been derived from the Tropospheric Emission Spectrometer on the EOS Aura satellite. 159 These satellite based spectrometers are greatly improving the information available for constraining emissions and concentrations of certain BVOCs although it should be noted that considerable uncertainties are associated with these estimates and more in situ observations are needed to assess these data.
3.2.4 Undetected compounds. Finally, the possibility that conventional analytical techniques may not detect all nonmethane organic compounds has long been speculated. 39,160 Indeed, several experiments 87,161,162 to quantify total nonmethane hydrocarbons in the atmosphere have shown that as air masses are more photochemically oxidized, the ratio of total NMHCs and sum of speciated NMHCs could reach up to a factor of 2-3. The methodology used for those studies was a conversion of total NMHCs to CO 2 (ref. 161 and 162) or a 2D GC technique. 87 Total OH reactivity, the reciprocal of OH chemical lifetime (s À1 ), provides quantitative information about the total amount of atmospheric constituents in the reactivity scale. Therefore, the comparison between measured OH reactivity and calculated OH reactivity from an observational dataset such as CO, CH 4 , NO 2 , and VOCs can provide a unique opportunity to assess our ability to measure all reactive atmospheric constituents. Several ambient total OH reactivity measurement results have consistently indicated signicant "missing OH reactivity", which cannot be explained by observational datasets of known trace gases. The degree of discrepancies tends to be higher in BVOC dominant environments. 163 The cause of missing OH reactivity has been speculated: unknown/unmeasured BVOC emissions 164 or sources of unknown/unmeasured oxidation products of known BVOCs. 165 A similar approach can be used to quantify the total VOC reactivity with ozone and other oxidants and measurement systems are being developed to accomplish this.

Summary and outlook
We extensively reviewed analytical techniques that have been widely used to characterize emissions and ambient distributions of BVOCs and their oxidation products. For the past ve decades, GC techniques have served as the main analytical methods for most BVOCs and their oxidation products. These applications have been expanded to oxygenated and semivolatile organic compounds, which were not routinely measured by GC techniques but have tremendous implications for SOA formation and tropospheric photochemistry. In addition, some cryofocusing techniques have become available to expedite GC analysis times on the order of a few minutes. This technology makes GC techniques suitable for fast time response applications such as airborne measurements. CIMS techniques have provided very selective and fast response measurements for BVOCs and their oxygenated products. Since mass spectrometry with a so ionization interface can provide molecular mass information of analytes, it has been effectively used for identication of unknown compounds. Also, optical techniques have been and are being developed for light oxygenated compounds, such as glyoxal and methyl glyoxal, that cannot be measured by conventional techniques.
Although the development of new sensing techniques provided new insights on BVOC emission and oxidation processes in the atmosphere, relatively few intercomparison efforts have been reported among different analytical techniques and for identical techniques with different congurations. Apel et al. 166 reported extensive instrument intercomparison results among 15 groups with some overlaps of analytical techniques (e.g. GC, HPLC, DOAS and PTR-MS). The campaign was mostly focused on oxygenated VOCs. The results indicate that the reported concentrations from each group of the tested compounds, especially low volatility compounds, vary signicantly and the authors concluded reliable measurements require "a high-quality instrument, and experience with the instrument including strict attention to analytical procedures such as zeroing the instrument and calibration". This statement should be noted by all investigators. As Kim et al. 116 discussed, signicant uncertainty associated with the measurement of known compounds such as isoprene can cause issues in constraining total reactive BVOCs in the atmosphere because a few dominant BVOC species contribute most of the reactivity towards oxidants in many instances. Therefore, we as a community should concentrate on both new technique development and promote assessment, intercomparison, and characterization of the existing sensing techniques (e.g. Kaser et al. 81 ). As the past decades of history indicates, new perspectives from well characterized and reliable measurements can expand our understanding of the importance of BVOCs and their oxidation processes in regional and global air quality and climate.