The source apportionment of aerosol iron (Fe), including natural and combustion Fe, is an important issue because aerosol Fe can enhance oceanic primary production in the surface ocean. Based on our previous finding that combustion Fe emitted by evaporation processes has Fe isotope ratios (δ⁵⁶Fe) that are approximately 4 ‰ lower than those of natural Fe, this study aimed to distinguish aerosol Fe sources over the northwestern Pacific using two size-fractionated marine aerosols. The δ⁵⁶Fe values of fine and coarse particles from the eastern or northern Pacific were found to be similar to each other, ranging from 0.0 ‰ to 0.4 ‰. Most of them were close to the crustal average, suggesting the dominance of natural Fe. On the other hand, particles from the direction of East Asia demonstrated lower δ⁵⁶Fe values in fine particles (−0.5 ‰ to −2.2 ‰) than in coarse particles (on average −0.02 ± 0.12 ‰). The correlations between the δ⁵⁶Fe values and the enrichment factors of lead and vanadium suggested that the low δ⁵⁶Fe values obtained were due to the presence of combustion Fe. The δ⁵⁶Fe values of the soluble component of fine particles in this region were lower than the total, indicating the preferential dissolution of combustion Fe. In addition, we found a negative correlation between the δ⁵⁶Fe value and the fractional Fe solubility in air masses from the direction of East Asia. These results suggest that the presence of combustion Fe is an important factor in controlling the fractional Fe solubility in air masses from the direction of East Asia, whereas other factors are more important in the other areas. By assuming typical δ⁵⁶Fe values for combustion and natural Fe, the contribution of combustion Fe to the total (acid-digested) Fe in aerosols was estimated to reach up to 50 % of fine and 21 % of bulk (coarse + fine) particles in air masses from the direction of East Asia, whereas its contribution was small in the other areas. The contribution of combustion Fe to the soluble Fe component estimated for one sample was approximately twice as large as the total, indicating the importance of combustion Fe as a soluble Fe source despite lower emissions than the natural. These isotope-based estimates were compared with those estimated using an atmospheric chemical transport model (IMPACT), in which the fractions of combustion Fe in fine particles, especially in air masses from the direction of East Asia, were consistent with each other. In contrast, the model estimated a relatively large contribution from combustion Fe in coarse particles, probably because of the different characteristics of combustion Fe that are included in the model calculation and the isotope-based estimation. This highlights the importance of observational data on δ⁵⁶Fe for size-fractionated aerosols to scale the combustion Fe emission by the model. The average deposition fluxes of soluble Fe to the surface ocean were 1.4 and 2.9 nmol m⁻² d⁻¹ from combustion and natural aerosols, respectively, in air masses from the direction of East Asia, which suggests that combustion Fe could be an important Fe source to the surface seawater among other Fe sources. Distinguishing Fe sources using the δ⁵⁶Fe values of marine aerosols and seawater is anticipated to lead to a more quantitative understanding of the Fe cycle in the atmosphere and surface ocean.

Atmospheric deposition is one of the main sources of dissolved iron (Fe) in the ocean surfaces. Atmospheric processes are recognized as controlling fractional Fe solubility (Fe_sol%) in marine aerosol particles. However, the impact of these processes on Fe_sol% remains unclear. One of the reasons for this is the lack of field observations focusing on the relationship between Fe_sol% and Fe species in marine aerosol particles. In particular, the effects of organic ligands on Fe_sol% have not been thoroughly investigated in observational studies. In this study, Fe species in size-fractionated aerosol particles in the Pacific Ocean were determined using X-ray absorption fine structure (XAFS) spectroscopy. The internal mixing states of Fe and organic carbon were investigated using scanning transmission X-ray microscopy (STXM). The effects of atmospheric processes on Fe_sol% in marine aerosol particles were investigated based on the speciation results. Iron in size-fractionated aerosol particles was mainly derived from mineral dust, regardless of aerosol diameter, because the enrichment factor of Fe was almost 1 in both coarse (PM_>1.3) and fine aerosol particles (PM_1.3). Approximately 80 % of the total Fe (insoluble + labile Fe) was present in PM_>1.3, whereas labile Fe was mainly present in PM_1.3. The Fe_sol% in PM_>1.3 was not significantly increased (2.56±2.53 %, 0.00 %–8.50 %, n=20) by the atmospheric processes because mineral dust was not acidified beyond the buffer capacity of calcite. In contrast, mineral dust in PM_1.3 was acidified beyond the buffer capacity of calcite. As a result, Fe_sol% in PM_1.3 (0.202 %–64.7 %, n=10) was an order of magnitude higher than that in PM_>1.3. The PM_1.3 contained ferric organic complexes with humic-like substances (Fe(III)-HULIS, but not Fe-oxalate complexes), and the abundance correlated with Fe_sol%. Iron(III)-HULIS was formed during transport in the Pacific Ocean because Fe(III)-HULIS was not found in aerosol particles in Beijing and Japan. The pH estimations of mineral dust in PM_1.3 established that Fe was solubilized by proton-promoted dissolution under highly acidic conditions (pH < 3.0), whereas Fe(III)-HULIS was stabilized under moderately acidic conditions (pH 3.0–6.0). Since the observed labile Fe concentration could not be reproduced by proton-promoted dissolution under moderately acidic conditions, the pH of mineral dust increased after proton-promoted dissolution. The cloud process in the marine atmosphere increases the mineral dust pH because the dust particles are covered with organic carbon and Na. The precipitation of ferrihydrite was suppressed by Fe(III)-HULIS owing to its high water solubility. Thus, the organic complexation of Fe with HULIS plays a significant role in the stabilization of Fe that was initially solubilized by proton-promoted dissolution.

Light‐absorbing iron oxide (FeO_x) aerosols such as magnetite contribute to shortwave atmospheric heating and possibly affect the biogeochemical cycle. However, their atmospheric abundance and emission sources are poorly understood. In this study, we quantified the abundance and mixing states of FeO_x at two urban sites in Tokyo and Chiba, Japan, using a modified single‐particle soot photometer and filter‐based instruments. At both sites, the majority of the FeO_x were of anthropogenic origin in the form of aggregated magnetite nanoparticles, and their concentrations generally correlated with those of black carbon (BC) and carbon monoxide. In Chiba, where the observatory was located near an integrated steel plant, we observed distinctly high FeO_x concentrations and high FeO_x/BC concentration ratios when the air mass passed through the plant. From the observed FeO_x plumes with the mass equivalent diameter range of 170–2,100 nm, we made an estimate of their emission flux to be approximately 0.012% of the crude steel production, assuming that a Gaussian plume approximation was applicable to our data analysis. Meanwhile, in Tokyo, where the observatory was 20–40 km northwest of steel plants, the FeO_x concentrations and FeO_x/BC ratios showed clear diurnal variations and depended little on wind direction. This indicates that other human activities also locally produce FeO_x aerosols in Tokyo. Our data imply that although steel plant activities emit a large amount of FeO_x, emissions from other anthropogenic sources, for example, motor vehicles, have a major contribution to the abundance of FeO_x aerosols at the regional and global scales.