Solar eclipse demonstrating the importance of photochemistry in new particle formation

Solar eclipses provide unique possibilities to investigate atmospheric processes, such as new particle formation (NPF), important to the global aerosol load and radiative balance. The temporary absence of solar radiation gives particular insight into different oxidation and clustering processes leading to NPF. This is crucial because our mechanistic understanding on how NPF is related to photochemistry is still rather limited. During a partial solar eclipse over Finland in 2015, we found that this phenomenon had prominent effects on atmospheric on-going NPF. During the eclipse, the sources of aerosol precursor gases, such as sulphuric acid and nitrogen- containing highly oxidised organic compounds, decreased considerably, which was followed by a reduced formation of small clusters and nanoparticles and thus termination of NPF. After the eclipse, aerosol precursor molecule concentrations recovered and re-initiated NPF. Our results provide direct evidence on the key role of the photochemical production of sulphuric acid and highly oxidized organic compounds in maintaining atmospheric NPF. Our results also explain the rare occurrence of this phenomenon under dark conditions, as well as its seemingly weak connection with atmospheric ions.

The Chemical Ionization Atmospheric Pressure interface Time Of Flight (CI-APi-TOF) mass spectrometer was measuring above canopy in the 35 m walk up tower. The used mass spectrometer was originally developed for precise sulphuric acid measurements (5) but nowadays the development of the technique utilizes high resolution time-of-flight mass spectrometers and is able to detect molecular sulphuric acid (6), sulphuric acid containing clusters (7) and highly oxidized organic molecules (8,9). HOMs measured with the CI-APi-TOF are estimated (calculated) to have low vapour pressures and this work reports exactly the same compound groups detected in previous laboratory experiments and publications, where compounds from C10H16O7 and higher oxygen numbers are classified as extremely low-volatile organic compounds (ELVOC) in 273 K temperatures (10,11). During the eclipse day 20 Mar 2016, temperature was around 273±2 K. Measurements in Fig.3 were measured during a spring campaign between 14 Mar and 11 Apr 2011 from close to ground level.
In previous studies (e.g. (11) all highly oxidized molecules measured by the CI-APi-TOF were somewhat incorrectly termed as ELVOCs. More recent studies, including both computational (12) and in laboratory (e.g. (13) suggest that many of the HOMs found by Ehn et al. are LVOCs rather than ELVOCs.
The instrument was run in the negative ion mode and with an active chemical ionization method using charged HNO3 ( NO3-) as a reagent ion. The sample is introduced via ~60 cm ¾" tube with a flow rate of 10 slpm (standard L/minute). The sample is surrounded with a sheath flow containing the reagent ions which are then guided to meet the sample using an electric field. Sample is charged either with proton transfer from the acidic sample molecule (AH) to the reagent ions (B) via AH + B - A -+ BH or via clustering AH + B - AHB -Using chemical ionization is a highly selective method for detection of low-volatility aerosol precursors and molecules from ambient air. Detailed information of the CI-APi-TOF can be obtained from Jokinen et al. 2012 (6).
The concentration of sample molecules is calculated as following: ,where [sample]neutral is the calculated concentration (in molecules/cm 3 ) of the neutral measured compound e.g. sulphuric acid, sampleis the measured signal of sample chemically charged sample charged via proton transfer, sample . NO3is the measured charged sample signal formed via clustering and C is a calibration factor. Reagent ions in the equation are the sum of signals from NO3 -, HNO3NO3and (HNO3)2NO3 -.
Calibration is done using a sulphuric acid generator described by Kürten et al., (2012) (14). Calibration factor determined for these measurements using the sulphuric acid generator was 2.2e10 molecules/cm 3 .

Particle Size Magnifier (PSM)
The Airmodus nano Condensation Nuclei Counter (nCNC)(15) is a mixing-type condensation particle counter (CPC) consisting of a Particle Size Magnifier (PSM) and a conventional CPC. Inside the PSM, diethylene glycol is used to activate and grow particles to 90 nm. After that, particles grow to detectable sizes by condensation of butanol inside the CPC. In this study, the cut-off size of the PSM was varied between approximately 1 and 3 nm (electrical mobility equivalent diameter) by altering the mixing ratio of the sample and saturator flow rates. Changing the supersaturation from the smallest to the highest value takes two minutes. The conversion from the PSM saturator flow rate to the cut-off size contains uncertainties related to the particle composition (16). Mixture compounds of sulfuric acid and nitrogen containing HOMs are observed to nucleate in Hyytiälä (17), and the range of supersaturations inside the PSM are adjusted accordingly. The data inversion was conducted by assuming Gaussian-shaped kernel functions and the concentration of clusters in three sub-3 nm size bin was obtained (18). Uncertainties in the inverted size bins reflect the small variations in the PSM cut-off sizes due to ambient water vapor concentration variations, and uncertainties in the particle activation due to unknown exact particle composition and resulting cut-off sizes. The data was inverted to size bins of < 1.5 nm, 1.5 -2 nm and 2 -3 nm +-0.4 nm with time resolution of 10 min. Note that in this article, the whole nCNC-system is referred to as the PSM, as is commonly done in the literature.

Neutral cluster and Air Ion Spectrometer (NAIS)
Neutral cluster and Air Ion Spectrometer (NAIS) (19,20) is an ion mobility spectrometer measuring the size distributions of ions between 0.8 nm and 42 nm (mobility equivalent diameter) and total particles, i.e. neutral and charged, between ~2 nm and 42 nm. When total particles are measured, the sample aerosol is charged with a corona charger. Two cylindrical differential mobility analysers classify positive and negative ions based on their electrical mobility and 21 electrometers on the outer cylinder detect the ions. High sample flow rate (54 l/min) is used to minimize the diffusion losses. Time resolution of the NAIS data is 3 minutes.

Aerosol Chemical Speciation Monitor (ACSM)
The ACSM (21) measures the non refractory aerosol chemical composition, e.g. the mass concentration of organics, sulfates, nitrates, ammonium and chlorides within the vacuum aerodynamic diameter range of 75 -650 nanometers (22). The sample air is focused with aerodynamic lenses, flash vaporized in 600 ˚C and electron impact ionized (EI) with 70 electronvolts. The mass analyzer is a residual gas analyzer (RGA) type quadrupole.

Growth rates
The growth rate of sub-3nm clusters was determined using the NAIS ion data by fitting a Gaussian distribution to the concentration time series at a certain size to determine the moment of maximum concentration. The growth rate was obtained by a linear fit to the moments of maximum concentrations and the corresponding geometric mean diameters of the particles. For the comparison of this method to other growth rate methods, see Yli-Juuti et al. (2011) (23).

Formation rates
The cluster formation rates at 1.5 nm ( . ), and at 2 nm ( ) were calculated from PSM data according to Kulmala et al. (2012) (19): Here is the cluster concentration, refers to the coagulation sink, and is the cluster growth rate. Thus, Eq. (2) takes into account the losses of clusters due to coagulation and growth out of the studied size range.
The formation rates of positive and negative ions at 2 nm ( ± ) were calculated from NAIS ion mode data (19): In addition to the losses due to coagulation and growth, the loss of ions due to ion-ion recombination and the source of ions due to charging of neutral clusters were taken into account. The ion-ion recombination coefficient, , and ion-neutral attachment coefficient, , were assumed to equal 1.6•10 −6 cm 3 s −1 and 0.01•10 −6 cm 3 s −1 , respectively.

Proxy for OH
The proxy for the concentration of hydroxyl radical (OH) was calculated by scaling the measured UV-B radiation with empirically derived factors according to Petäjä et al. (2009) (24): Simple model calculations Simple model calculations were conducted by assuming that the time evolution of the concentration of sulfuric acid and N-HOMs during the eclipse can be expressed as: Here is measured UV-B radiation and CS is the condensation sink.
The value for the coefficient was obtained by assuming that in the beginning of the eclipse the vapor production rate equals the loss rate of vapor by condensation: For sulfuric acid we also tested another approach, where the dependence of vapor production rate on SO2 concentation was taken into account. In this case, the time evolution of sulfuric acid concentration is obtained from: Correspondingly, the value for the coefficient B was calculated from: Furthermore, to see the effect of CS on the modelled concentration, we performed additional simulations where CS was set to zero after the minimum of the measured sulfuric acid concentration.
Supporting results from the simple model The sulphuric acid and N-HOM concentrations as a function of UV-B intensity show that the values measured during the decreasing phase of UV-radiation were somewhat higher than those during the increasing phase of UV-radiation as seen in Figure S5 and S6. This feature is consistent with their assumed photochemical production pathways and their removal processes. The decrease in the concentrations of low-volatile vapours during the eclipse can be explained by their condensation onto existing aerosol particles. Throughout the eclipse, the condensation sink (CS) was around 0.0004 s -1 and stayed almost constant ( Figure S9). However, when modelling the production of condensing vapours during the recovery from the solar eclipse, we see that the pre-existing aerosols do not accommodate sulphuric acid as effectively as possible ( Figure S11), especially when compared with a day without a solar eclipse (see Figure S11). The recovery of the main N-HOM concentration from the solar eclipse is simulated surprisingly well with the simple model, indicating its irreversible condensation onto the measured pre-existing particles.

Shoot-level gas exchange and electron transport rate measurements
Gas-exchange (CO2, H2O, VOCs) of Scots pine shoots were measured using four dynamic branch enclosures with an automated sampling and gas-analyser setup. The enclosures were installed to unshaded top-canopy conditions and there were 48-72 closures per day for each enclosure. The approach is described in detail by Hari et al. 1999 (25) and Aalto et al. 2015 (26).
Photosynthetic energy conversion and electron transport rate were measured using a PAM chlorophjyll fluorescence monitoring system (Monitoring PAM, Walz GmbH, Germany). The system included five fluorometers distributed across three Scots pine trees (3 in top canopy and 2 in low canopy needles). Measurements were carried out at a frequency of 30 minutes. At each measuring point PAR, air temperature, needle-level instantaneous fluorescence yield (F') and maximal fluorescence yield (F'm) are retrieved. The rate of electron transport in photosystem II (PSII) is estimated as: Where A is leaf absorptance (equal to 0.82 for our pine needles), aII is the relative absorption cross section of PSII (assumed to be 0.5) (27). Further details on these measurements can be found in Porcar-Castell et al. 2008 (28) and Porcar-Castell 2011 (29).  Figure S3: Dynamics of electron transport rate (ETR) of PSII in needles at five different locations within the Scots pine canopy (A). CO2 exchange rate and transpiration of three to four top-canopy Scots pine branches (B and C). ETR is a measure of photosynthetic rate at the level of light reactions of photosynthesis. In contrast to gas exchange, ETR is not affected by leaf respiration or photorespiration, thus providing a complementary view of gross photosynthesis. The decrease in radiation during the eclipse exerted an instantaneous effect on the electron transport rate of photosystem II at all five measuring locations. Decreased ETR resulted in subsequent reduction in CO2 assimilation which was also accompanied with a decrease in transpiration. The data reflects the strong and rapid control of photosynthesis by PAR at this temporal scale, as during the solar eclipse. Red line: maximum phase, dashed black lines: beginning and end time of the eclipse.