Gas-and Water-Phase PAHs Emitted from a Single Hydrogen-Oxygen PEM Fuel Cell

This study focuses on the comparison between gasand water-phase polycyclic aromatic hydrocarbons (PAHs) emitted from a single hydrogen–oxygen proton exchange membrane (PEM) fuel cell (FC) at different flowrates and temperatures. The results show that among 21 PAHs, the most and least dominant species were Nap and BeP, respectively. At 65°C, the concentrations of individual gasand water-phase PAHs decreased with increasing flowrate, and the PAH concentrations were lower at the anode than those at the cathode. The concentrations of gas-phase Total-PAHs and Total-BaPeq were slightly lower at 65°C than those at 90°C, but an opposite trend was observed for water-phase ones. The temperature influenced water-phase PAH concentration profiles more than gas-phase ones, and the gasand water-phase PAHs had different concentration profiles. The performance of membrane-electrode assembly (MEA) decreased with increasing flowrate or temperature. The emission factor (EF) sum (anode + cathode) for gasor water-phase Total-PAHs increased with increasing flowrate. This tendency was also true for gas-phase Total-PAHs EFs but not for water-phase ones when raising the temperature from 65°C to 90°C. At 65°C and 52/35 sccm, the EF sums of water-phase Total-PAHs and TotalBaPeq were 2.18 ± 0.04 and 0.09 ± 0.00 μg g-MEA, respectively—smaller than those of gas-phase ones (3.02 ± 0.09 and 0.12 ± 0.00 μg g-MEA, respectively). More environmental concern should be directed at emitted gas-phase PAHs than at water-phase ones because the anode and cathode water effluents are usually recycled during PEMFC operations.


INTRODUCTION
Hydrogen-oxygen proton exchange membrane fuel cells (PEMFCs), regarded as a zero-emission green power generators, are promising alternative power sources for stationary and mobile applications (Wang et al., 2006).The key advantages of PEMFCs are high energy conversion efficiencies, quick start-up, and zero by-product emissions (Tenson and Baby, 2017).An H 2 -O 2 PEMFC has the main components of membrane-electrode assembly (MEA), gasdiffusion layer (GDL), and bipolar/end plate (Huang et al., 2006;Lai et al., 2012;Tenson and Baby, 2017).It is common that a PEM is symmetrically sandwiched by anode and cathode to prepare the MEA; moreover, the electrodes usually contain precious metal catalysts (e.g., Pt) (Huang et al., 2006;Lai et al., 2012), Pt-based alloys (Zhang et al., 2017), or non-platinum group metals (Reshetenko, et al., 2016) supported by carbon materials (e.g., carbon black, fiber, and nanotube) which are also used for the fabrication of bipolar/end plates and GDLs (e.g., carbon paper or cloth) (Tran et al., 2015;Zeis, 2015;Lee and Lee, 2016).
Extensive study has concentrated on accelerating the sluggish kinetics of cathode catalyst, minimizing overall Pt content, and increasing membrane proton conductivity for PEMFCs (Scofield et al., 2015;Tenson and Baby, 2017).Furthermore, hydrogen powered fuel cell electric vehicles are classed as ultra-low emission vehicles when supposing water vapor as the only emission substance in exhaust; therefore, the use of hydrogen powered fuel cell electric vehicles can reduce the emissions of pollutants (which degrade local air quality) and carbon dioxide.However, little attention has been paid to the emission of polycyclic aromatic hydrocarbons (PAHs) from PEMFCs (Huang et al., 2015(Huang et al., , 2016)).
PAHs have long been concerned for their adverse impacts on health and the environment (Tiwari et al., 2015;Kavouras et al., 2015;Pongpiachan, 2016;Li et al., 2017).PAHs can be generated from pyrogenic, petrogenic, biogenic, or diagenetic (indoor/outdoor) sources (Kavouras et al., 2015;Mwangi et al., 2015;Stogiannidis and Laane, 2015;Zhang et al., 2016;Fan et al., 2017;Lai et al., 2017;Li et al., 2017).In our previous works, we reported the emissions of 21 PAHs from a H 2 -O 2 PEMFC (installed by a lab-prepared MEA with the same Pt loading of 0.5 mg cm -1 on anode and cathode) (Huang et al., 2015) and effect of operating conditions on PAHs emission from the PEMFC (Huang et al., 2016).In this study, we explored the emissions, concentration profiles, and emission factors of gas-vs.water-phase PAHs from the PEMFC under different flowrates and temperatures.

Membrane-Electrode Assembly (MEA) and Fuel Cell Emission Test
The membrane-electrode assemblies (MEAs) tested in this study were all E-TEK commercial MEAs purchased from Micro Power Fuel Cell Co., Ltd.These MEAs had the same catalyst loading in both anode and cathode (0.4 mg-Pt cm -2 ) which were separated with a proton exchange membrane .An MEA commonly consists of a proton exchange membrane (solid electrolyte), two dispersed catalyst layers, and two gas diffusion layers (GDL).The structure of MEA was addressed in our earlier papers (Huang et al., 2006;Lai et al., 2012;Huang et al., 2015Huang et al., , 2016)).
The single H 2 -O 2 PEMFC used in PAH emission tests had an MEA symmetrically sandwiched by Telfon gaskets, carbon (graphite) blocks (with gas flow channel), and copper current collectors.The single cell was installed on a standard fuel cell test station that worked at 65°C or 90°C (80°C and 70°C for the anode and cathode humidifiers, respectively) and 0.6 V for 48 h for PAH collection under different anode hydrogen and cathode oxygen flow rates which were regulated by two mass flow controllers.The details of sampling system connected to the single cell and PAH collection were provided in our previous works (Huang et al., 2015(Huang et al., , 2016)).For comparison, the PAH emissions from the single PEMFC were also tested at various anode/cathode flowrates of 30/18, 100/100, and 200/200 sccm.It is almost impossible to form PAHs under the operating conditions of this study because it is an electrochemical process and the only "fuel" was hydrogen, so the PAHs collected from the anode/cathode emission gases and effluents were mainly from the desorption of PAHs which originally absorbed on the carbon materials of MEA in the fuel cell during operation (Huang et al., 2015(Huang et al., , 2016)).

PAH Sampling, Analysis, and Quality Control
In this study, the PAH sampling system, connected to the PEMFC, simultaneously collected PAHs from the anode and cathode emission ports.For each sampling, a glass water collector was used to collect the cell effluent containing water-phase PAHs; the water collector was serially connected with a glass cartridge (packed with XAD-16 resin granules and supported by a polyurethane foam (PUF) plug) for the collection of gas-phase PAHs (Fig. 1).Triplicate PAH samplings (n = 3) were conducted in each experiment.After PAH collection, cell polarization tests were performed at different anode/cathode flowrates or cell temperatures.
The details of PAHs extraction by Soxhlet extraction method, the identification/quantification of PAHs (with blank correction) by gas chromatograph/mass spectrometry (GC/MS), and the quality assurance (QA)/quality control (QC) procedures for PAHs GC/MS analysis were stated elsewhere (Huang et al., 2015).The correlation coefficients of the calibration curves were 0.995-0.999.Ten consecutive injections of a PAH 610-M standard yielded an average relative standard deviation (RSD) of the GC/MS integration area of 3.0% with a range of 0.8-5.1%.The total recovery efficiencies of PAHs from seven consecutive injections ranged from 74 to 110%, while the average recoveries of the five internal standards were 85-93% across seven consecutive injections.
The coefficients of variation for repeat injections of the standard solution (containing PAH Mixture-610 M (16 PAHs) and five Merck PAH standards) were all less than 5% for all of the analyzed PAHs, whereas those obtained by replicate analysis were 2-10%.Analyses of serial dilutions of PAH standards revealed the limits of detection (LODs) of GC/MS to be between 0.010 and 0.474 ng for the 21 PAH compounds.The limit of quantification (LOQ) was defined as the LOD divided by the sampling volume.The LOQ values of the 21 PAH compounds for cell effluent and emission gas were 0.102-2.470ng L -1 and 0.076-0.175ng Nm -3 , respectively.

Effect of Flowrate on Gas-phase PAHs Emission
For the tests of PEMFC PAH emissions at different flowrates, the anode/cathode flowrates of 52/35 sccm represents 1.5 and 2 times of the stoichiometric requirements for the anode and cathode feeding gases, respectively (Huang et al., 2006;Lai et al., 2012;Huang et al., 2015Huang et al., , 2016)).Fig. 2 shows the anode and cathode concentration profiles of gas-phase 21 PAHs collected from the exhaust of fuel cell operated at 65°C and different flowrates.At anode/cathode flowrates of 52/35 sccm, the concentrations of PAHs in anode and cathode emission gas ranged from not detectable (ND) (BeP) to 2.961 ± 0.005 (Nap) and ND (BeP) to 4.115 ± 0.005 (Nap) µg Nm -3 , respectively (Fig. 2(a)), while most of the other species had concentrations of 0.200-0.800µg Nm -3 .The same PAH species showed a higher concentration in the cathode emission gas than in the anode one.A similar trend was also found at the lower anode/cathode flowrates of 30/18 sccm (Fig. 2(b)).This phenomenon is mainly attributed to the fact that the sampling volume was greater at anode than at cathode but both electrode had similar or slightly different PAH mass emission, regardless of the difference in operation between anode and cathode.
The operating potential of anode was < 1.23 V (theoretical) vs. standard hydrogen electrode (SHE), while that of cathode was ~0 V vs. SHE.Moreover, the main electrochemical reaction at anode is hydrogen oxidation which can be expressed as the following reaction: On the other hand, the principle electrochemical reaction at cathode is oxygen reduction, described by reaction 2: Nevertheless, the emission concentration of the same PAH species was lower at 30/18 sccm than at 52/35 sccm, because the former emitted much less PAH mass than the latter, although the sampling volume was lower at 30/18 sccm than at 52/35 sccm (see more discussion in the last section).
Dissimilarly, the concentration of individual PAH species decreased, when the gas flowrate increased from 52/35 to 100/100 sccm, except NaP (Fig. 2(c)); hence, the emission of Nap was more influenced by flowrate than those of the other PAH species.This result is majorly attributable to the higher sampling volume at the greater flowrate, although more PAH mass was present in the anode and cathode emission gases at the greater flowrate (see more discussion in the last section).In our earlier study, we also observed this phenomenon when using the same sampling volume at different flowrates (52/35 and 100/100 sccm) or different sampling volumes (or time periods: 12, 24, and 36 hr) at the same flowrate (52/35 sccm) (Huang et al., 2016).When the gas flowrate further increased from 100/100 to 200/200 sccm, more PAHs desorption occurred and thus was collected in the anode or cathode emission gas (see more discussion in the last section) (Fig. 2(d)).Therefore, flowrate significantly influenced the desorption of PAHs from carbon black support during operation, despite the possible strong π-π interaction between PAHs and the graphitic structure of carbon materials (Kah et al., 2011) and the fact that PAHs are relatively easier to adsorb on than to desorb from the surface of carbon blacks (Moninot, 2010).
Of all the 21 PAHs, NaP was the most predominant species at different flowrates while BeP was only detected at 200/200 sccm.Some PAH species such as Flu, CHR, BkF, BbC, and COR were comparably dominant as PA at 30/18, 52/35 and 100/100 sccm, but these compounds were significantly less dominant than PA at 200/200 sccm.Consequently, flowrate also affected the desorption of individual PAH species and thus its dominance in the anode and cathode emission gases.This result is also probably associated with the electronegative functional groups of compounds, because aromatic compounds having strong electronegative functional groups generally form stronger bonds with the graphene surface of carbon materials (Ma and Uddin, 2013).

Effect of Flowrate on Water-Phase PAHs Emission
For comparison, the water-phase PAHs emitted from anode and cathode gases were also collected under the same .Again, among the 21 PAHs, NaP had the highest concentration, followed by Flu, COR, PA, BbC, CHR, and FL, while BeP was not detected under all tested flowrates.However, the concentration difference between the highest (NaP) and the second highest was less for water-phase PAHs than for gas-phase PAHs.This result should be associated with the fact that the 21 PAHs have low water solubility.Roughly, the water solubility of PAHs decreases as the aromatic ring number increases; for example, the solubility data of Nap, Pyr, and BaP at 25°C are 31.0,0.132, and 0.0038 mg L -1 , respectively (ER Wiki, 2017).The NaP concentrations in anode effluents at 30/18, 52/35, 100/100, and 200/200 sccm were 1.759 ± 0.057, 0.766 ± 0.122, 0.419 ± 0.053, and 0.220 ± 0.028 µg L -1 , respectively, while the corresponding data in cathode effluents were 8.530 ± 0.178, 3.871 ± 0.253, 2.325 ± 0.061, and 0.954 ± 0.060 µg L -1 , respectively.BeP was not detectable (ND) in the anode and cathode effluents at all the tested flowrates.It was also observed that the concentrations of individual water-phase PAHs were smaller at anode/cathode flowrate = 52/35 sccm than at anode/cathode flowrate = 30/18 sccm, mainly because the collected water volume was greater at 52/35 sccm (Fig. 3(a)) than at 30/18 sccm (Fig. 3(b)), but their emitted PAH masses were comparable (see more discussion in the last section).For the same reason, the concentrations of individual PAHs were greater at 52/35 sccm than at 100/100 sccm (Fig. 3(c)).This tendency is also true when increasing the flowrate to 200/200 sccm (Fig. 3(d)).

Effect of Temperature on Gas-and Water-Phase PAHs
Fig. 4(a) shows the concentration profiles of 21 individual PAHs in anode and cathode emission gases at 90°C and 52/35 sccm.Again, the concentration of BeP was ND in anode and cathode emission gases.This result is also true in the anode and cathode effluents (Fig. 4(b)).In anode emission gas, the concentrations of individual PAHs ranged from 0.102 ± 0.111 (DBA) to 3.055 ± 0.358 (NaP) µg Nm -3 ,  excluding BeP; the corresponding values were from 0.152 ± 0.001 (DBA) to 4.585 ± 0.718 (NaP) µg Nm -3 in cathode emission gas.However, they ranged from 0.168 ± 0.003 (DBA) to 0.557 ± 0.106 (NaP) µg L -1 in anode effluent and from 0.192 ± 0.001 (BghiP) to 2.856 ± 0.205 (NaP) µg L -1 in cathode effluent.The concentrations of individual PAHs were also higher in the anode side than in the cathode side, except that of water-phase BkF at 90°C.At 52/35 sccm, the concentrations of gas-phase PAHs were similar at both 65°C (Fig. 2(a)) and 90°C (Fig. 4(a)), while the water-phase PAH concentrations were greater at 65°C (Fig. 3(a)) than at 90°C (Fig. 4(b)), consistent with the fact that the water solubility of a PAH species was lower when the temperature is higher.

Comparison of Gas-and Water-phase PAH Emission Profiles
In this study, the concentrations of gas-and water-phase individual PAHs in anode and cathode emission gases were normalized to those of Total-PAHs for the comparison of their emission profiles (fingerprints) which denoted the fraction of a PAH species (PAH i , i = 1-21) using the following calculation: Fraction (%) of PAH i = 100 × (concentration of PAH i / concentration of Total-PAHs) (3).
For the anode emission under various flowrates or temperatures, the concentration profiles of gas-phase individual PAHs were similar (Fig. 5(a)).A similar trend was also observed for the gas-phase individual PAHs from cathode emission (Fig. 5(b)).Moreover, the concentration profiles of individual PAHs from anode and cathode emission gases were similar.Therefore, the flowrate or temperature in the tested ranges did not significantly influence the concentration profile of gas-phase individual PAHs.On the other hand, the concentration profiles of water-phase individual PAHs from anode effluent looked various in the sector from BkF to DBA at the lowest and highest feeding gas flowrates (30 and 200 sccm, respectively), although the profiles in the other sector were similar (Fig. 5(c)).This tendency was also observed for the PAH concentration profiles of cathode effluents in the sector from BaA to DBA; in addition to feeding gas flowrate, temperature (90°C) also affected the pattern of profile (Fig. 5(d)).However, the PAH concentration profiles of cathode effluents were not similar to those of anode effluents in the sector from BaA to DBA.These findings might be related to the effect of feeding gas flowrate or temperature on the PAH water solubility, particularly for the PAHs with relatively low water solubility (BaA, CHR, BbF, BkF, PER, IND, and DBA) (ER Wiki, 2017).Moreover, the concentration profiles of water-phase PAHs were largely dissimilar to those of gas-phase PAHs.The partitioning of individual PAH species between gas and water phases should be responsible for this difference (Lee et al., 2004).

Polarization Curves and Emission Factors
According to the polarization (current-potential) curves shown in Fig. 8, the performance of MEA installed in the PEMFC decreased when flowrate increased at 65°C.Furthermore, the MEA performance was even worse at 90°C, which had quick potential drop at current density smaller than 100 mA cm -2 , probably because the Nafion-117 membrane was drier to have lower conductivity at the higher temperature.However, the MEA performance was better at 30/18 sccm than at 52/35 sccm, although the former had lower stoichiometric requirements of feeding gases than the latter (Huang et al., 2015(Huang et al., , 2016)).
Table 1 summarizes the emission factors (EFs) of gasphase and water-phase Total-PAHs and Total-BaP eq for the anode and cathode and their sums under different operating flowrates or temperatures.The EF of anode gas-phase Total-PAHs increased with increasing flowrate (0.72 ± 0.05, 1.51 ± 0.01, 2.73 ± 0.37, and 5.11 ± 0.18 µg g-MEA -1 at 30/18, 52/35, 100/100, and 200/200 sccm, respectively, at 65°C) or temperature (1.51 ± 0.01 and 1.59 ± 0.07 µg g-MEA -1 at 65°C and 90°C respectively, at 52/35 sccm).A similar phenomenon was also observed for the EFs of cathode gasphase Total-PAHs ((0.66 ± 0.03)-(4.72 ± 0.30) µg g-MEA -1 ) and thus the sums of anode and cathode gas-phase Total-PAHs EFs (1.73 ± 0.05, 3.02 ± 0.09, 5.01 ± 0.47, and 9.83 ± 0.35 µg g-MEA -1 at 30/18, 52/35, 100/100, and 200/200 sccm, respectively, at 65°C and 3.19 ± 0.13 µg g-MEA -1 at 90°C and 52/35 sccm).This result is attributed to the more mass of gas-phase PAHs desorbed from carbon material support at higher flowrate or temperature, and this PAH desorption behavior was more enhanced by increasing flowrate than by increasing temperature.Nevertheless, the EFs of anode and cathode gas-phase Total-PAHs were similar at 52/35 sccm regardless of difference in operating temperature, whereas the EFs of anode gas-phase Total-PAHs were higher than those of cathode ones.This result is possibly associated with the presence of the product (water) generated from the reaction 2 at cathode).The sums of anode and cathode gas-phase Total-PAHs EFs for a commercial E-TEK MEA in this study (3.02 ± 0.09 µg g-MEA -1 ) was lower than that of our previous study using a lab-prepared MEA (21.4 ± 1.28 µg g-MEA -1 ) (Huang et al., 2016), primarily due to the differences in MEA weight and structure.However, it is not appropriate to directly compare the PAH emission factors obtained in this study with those of fuel consumption or driving distance (mileage) based data reported in literature.
It is known that at equilibrium, the air-water partition coefficient (K AW ) of a PAH species is equal to H/RT, where H, R, and T are Henry's constant, gas constant, and absolute temperature, respectively (Shiu and Mackay, 1997;Bamford et al., 1999).PAHs usually show a trend of diminishing H or K AW with increasing molecular mass at constant temperature (Shiu and Mackay, 1997); however, the K AW increases when temperature rises (Bamford et al., 1999).In this study, the fuel cell system might possibly reach steady-state in the operation at fixed flowrate and temperature, but it was not the case of equilibrium.Therefore, it is not appropriate to use our gas-and water-phase PAH data for K AW calculation.

CONCLUSIONS
In this study, at anode/cathode flowrates of 52/35 sccm, the concentrations of PAHs in anode and cathode emission gases ranged from not detectable (ND) (BeP) to 2.961 ± 0.005 (Nap) and ND (BeP) to 4.115 ± 0.005 (Nap) µg Nm -3 .Roughly, the concentrations of individual gas-and waterphase PAHs decreased when feeding gas flowrate increased, and the anode PAH concentrations were lower than the cathode ones at 65°C.At 52/35 sccm, the concentrations of gas-phase PAHs were similar at both 65°C and 90°C, while the water-phase PAH concentrations were greater at 65°C than those at 90°C.Similar concentration profiles were observed for anode and cathode gas-phase PAHs at various flowrates or temperatures; however, those for water-phase PAHs were partially different, and the gas-and waterphase PAHs had different concentration profiles.
The concentrations of PAHs classified by molecular weight (MW) in anode and cathode emission gases varied at different flow rates (except 18 sccm) and temperatures according to the order LMM-> HMM-> MMM-PAHs, and the gas-phase Total-PAHs and Total-BaP eq concentrations decreased with an increase of flowrate (except at 30/18 sccm).At 52/35 sccm, the concentrations of gas-phase LMW-, MMW-, and HMW-PAHs at 65°C were similar to those at 90°C; therefore, the gas-phase Total-PAHs or Total-BaP eq concentrations were also similar at both temperatures (Total-PAHs: 8.892 ± 0.079 and 13.18 ± 0.792 µg Nm -3 for the anode and the cathode, respectively; Total-BaP eq : 0.365 ± 0.001 and 0.546 ± 0.004 µg Nm -3 for the anode and the cathode, respectively, at 65°C).The concentrations of water-phase Total-PAHs were higher at 65°C (7.940 ± 0.149 µg L -1 and 32.09 ± 0.185 µg L -1 for anode and cathode effluents, respectively) than at 90°C.The Total-BaP eq concentrations in anode and cathode effluents were 0.440 ± 0.003 and 0.893 ± 0.005 µg L -1 , respectively, at 65°C.
The performance of MEA decreased with increasing flowrate or temperature.The EF sum (anode + cathode) for gas-or water-phase Total-PAHs increased with increasing flowrate.This tendency is also true for gas-phase Total-PAHs EFs but not for water-phase ones when raising the temperature from 65°C to 90°C.At 65°C and 52/35 sccm, the EF sums of water-phase Total-PAHs and Total-BaP eq were 2.18 ± 0.04 and 0.09 ± 0.00 µg g-MEA -1 , respectively, smaller than those of gas-phase ones (3.02 ± 0.09 and 0.12 ± 0.00 µg g-MEA -1 , respectively).
Fig. 1.The scheme of single PEMFC for PAH emission tests (W: water-phase PAHs collector, X: gas-phase PAHs collector (cartridge), Y: flow meter, and Z: pump).

Table 1 .
Emission factors (EFs) of gas-phase and water-phase Total-PAHs and Total-BaP eq for the anodes (A), cathodes (C), and sums (A + C) under different operating flowrates or temperatures.