After aerosol deposition from the atmosphere, black carbon (BC) takes part
in the snow albedo feedback contributing to the modification of the Arctic
radiative budget. With the initial goal of quantifying the concentration of
BC in the Arctic snow and subsequent climatic impacts, snow samples were
collected during the research vessel (R/V)
Black carbon (BC) aerosol, produced by incomplete combustion of biomass and fossil fuels, is transported from extensive mid-latitude source regions to the Arctic atmosphere (Schacht et al., 2019), where it influences the regional climate (Quinn et al., 2015). Once removed from the atmosphere, BC particles continue to affect the Arctic radiative budget by directly decreasing the snow albedo (Dou and Xiao, 2016) and promoting snow grain growth (Skiles and Painter, 2017). In turn, the acceleration of the melting rate leads to earlier exposure to the underlying surface. The overall process is usually called snow albedo feedback and might be considered among the strongest forcing mechanisms in the Arctic region (Hansen and Nazarenko, 2004; Flanner et al., 2007; Skiles et al., 2018).
Considering the climatic repercussions caused by BC in snow, the scientific community has been measuring the content of BC in snow across the Arctic for almost 4 decades (Clarke and Noone, 1985; Doherty et al., 2010; Dou and Xiao, 2016; Tørseth et al., 2019). Unfortunately, a standardized and universally accepted analytical technique for the measurement does not yet exist. Generally, the wide variety of analytical approaches to measure BC in snow can be divided into offline and online methods. Considering the offline approach, BC mass can be measured after the melting and filtration of the snow sample via thermal optical analysis (Hagler et al., 2007) or transmittance spectroscopy (Doherty et al., 2010). Alternatively, BC mass can be quantified, after the nebulization of the melted snow samples, with online techniques such as the photoacoustic technique (Schnaiter et al., 2019) or the laser-induced incandescence technique (Schwarz et al., 2012).
In recent years, the laser-induced incandescence technique, more specifically with the single particle soot photometer (SP2; Droplet Measurement Technologies, Longmont, CO, USA), was often used to quantify refractory BC (rBC) particle (Petzold et al., 2013) concentrations in the snow and ice in various regions of the Arctic (Khan et al., 2017; Macdonald et al., 2017; Jacobi et al., 2019; Mori et al., 2019; Zhang et al., 2020). The rBC analytical procedure now generally includes the following three steps: (1) melting of the snow and/or ice sample, (2) nebulization with the pneumatic concentric nebulizer equipped with a warming–cooling desolvating system (i.e., Marin-5 – Teledyne Technologies, Omaha, NE, USA; Apex Q – Elemental Scientific Inc., Omaha, NE, USA) and (3) sampling the resulting aerosol with the SP2. During nebulization, the melted sample is usually transported to the nebulizer at a constant flow rate via a peristaltic pump. The liquid is then broken into small droplets and suspended in a nebulization chamber by means of a pneumatic concentric nebulizer. Once suspended, the solvent in the droplets is evaporated and removed with a warming cooling cycle. Several studies addressed the issue of reducing the losses of rBC during the nebulization phases by controlling the liquid flow rate, gas flow rate and pressure and temperature cycle (Lim et al., 2014; Wendl et al., 2014; Mori et al., 2016; Katich et al., 2017). Overall, up to 75 % of rBC mass is suspended from the sample, transported through the nebulizer and, finally, detected by the SP2 without the addition of surfactants (Lim et al., 2014; Mori et al., 2016). Due to the reduction in water density and viscosity, the doping with isopropyl alcohol increases the rBC mass nebulization efficiency to values close to unity (Katich et al., 2017). It must be considered that, despite corrections for the nebulization efficiency, the degree of comparability with more traditional techniques such as the thermal optical method and the integrating sphere/integrating sandwich spectrophotometer is still variable (Schwarz et al., 2012; Lim et al., 2014; Mori et al., 2019).
In the Arctic region, many snow samples were collected in coastal regions and over sea ice (Tørseth et al., 2019) where the sea salt components often dominate the snow chemical composition, especially in summer in presence of open water (Krnavek et al., 2012; Jacobi et al., 2019). This is particularly relevant over sea ice, where sea salt aerosol, suspended as sea spray, can be deposited at the snow surface, and the capillary upward migration of sea salt from the sea ice can lead to a high salt concentration at the bottom of the snowpack (Domine et al., 2004). Our surface snow samples, collected over the sea-ice-covered Fram Strait in summer 2017 within the PASCAL (Physical Feedbacks of Arctic Boundary Layer, Sea Ice, Cloud and Aerosol) experiment, were highly affected by salt deposition and showed a wide range of salinity. The presence of salt might have broad impacts on snow analysis, including, via the influence of the nebulization of the sample, the analyte and solvent transport and even the analytical signal of certain analytical techniques, such as inductively coupled plasma atomic emission spectroscopy (e.g., Sharp, 1988; Todolí et al., 2002; Burgener and Makonnen, 2020). Such effects are commonly called matrix effects. Until now, the potential interference of sea salt during the analysis of rBC particles with the SP2 has not been assessed.
Considering the high salinity of the snow samples collected in the Fram
Strait, a series of laboratory experiments were conducted to quantify the
impact of sea salt on nebulization and rBC detection with the SP2
instrument. This work aims to identify the importance of the salt matrix
effect, especially in the context of the MOSAiC (Multidisciplinary
drifting Observatory for the Study of Arctic Climate;
The PASCAL expedition (Flores and Macke, 2018), organized within
the framework of the (AC)
The experimental setup used for the snow sample analysis is schematized in
Fig. 1. First, the snow samples were melted in a thermostatic bath at 25
Schematic of the instrumental setup deployed to analyze the PASCAL snow samples and to perform the laboratory test experiments. ALABAMA (aircraft-based laser ablation aerosol mass spectrometer) was not available for laboratory test experiments.
The sample was then fed via a peristaltic pump and through a liquid flow
meter (SLI liquid flow meter; Sensirion AG, Staefa, Switzerland) to a
Marin-5 nebulizer (Teledyne Technologies, Omaha, NE, USA). In the Marin-5, the liquid sample was aerosolized into an airflow by a concentric pneumatic
nebulizer, and the resulting droplets were dehydrated by a heating–cooling
cycle (110–5
Transport losses of aerosol particles were calculated using the Particle
Loss Calculator software, which treats aerosol diffusion and sedimentation as well as turbulent inertial deposition and inertial deposition in the bends and contractions of tubing (von der Weiden et al., 2009). The software was developed at the Max Planck Institute for Chemistry in Mainz (Germany) and is
available at
Warm conditions were encountered during the sampling campaign, with the air
temperature increasing from approximately
Properties of snow samples grouped in conductivity classes.
Conductivity of melted samples measured between 20 and 25
Here we investigate the potential relationship between electrical conductivity, which is used here as a proxy for salinity, and the aerosolization of particles. Note that
Under fixed nebulization conditions (constant liquid sample flow and gas
flow), a large number of particles (droplet residues) were suspended by the
nebulization process.
The ALABAMA measurements confirmed the predominant presence of seawater
components, such as sodium chloride (NaCl) and magnesium (Mg), over other
chemical species (shown in Fig. 3). The particle fraction (PF) of NaCl-containing particles increased from roughly 30 % for the lowest salinity class to 60 %–80 % of all analyzed particles for S2, S3, S4 and S5. A similar increase was observed for Mg-containing particles. Other particle species, e.g., non-sea-salt (nss) nitrate and sulfate, were only relatively abundant in samples with low conductivity. For salinity class S4, the fraction of NaCl- and Mg-containing particles is significantly lower compared to other classes with
The number fraction of analyzed particles (PF) containing the given chemical species measured by ALABAMA in the PASCAL snow samples as a function of the salinity classes (Sn). The selected species are sodium chloride (NaCl), non-sea-salt (nss) nitrate, nss sulfate, magnesium (Mg), levoglucosan and dicarboxylic acids, organic carbon (OC) and elemental carbon (EC). The chemical composition is measured for particles in the 110–5000 nm diameter range.
SP2 measurements indicated a monotonic decrease in
These results make it clear that rBC-containing particles represent the
small minority of the aerosol population nebulized from the snow samples. In
fact, the number fraction of detected rBC particles (
The laboratory experiments aimed to reproduce BC snow concentrations representative of generic Arctic conditions and the specific salinity conditions representative of PASCAL snow samples. It is important to note that the salt concentrations explored in the present work do not represent realistic conditions encountered in continental or mountain regions where sea salt aerosol deposition is not dominant. The ability to reproduce such conditions and practical limitations is presented in the following.
Fullerene soot (FS; Alfa Aesar; lot no. W08A039) was used as a proxy for
ambient black carbon. FS is a well-characterized standard for SP2
calibration (Gysel et al., 2011; Laborde et al., 2012a) and is accepted as the reference standard for ambient black carbon (Moteki and Kondo, 2010; Baumgardner et al., 2012). A total of three different inorganic salts were chosen to replicate the conductivity array of the snow samples, i.e., sodium
chloride (NaCl; Honeywell International Inc), potassium chloride (KCl; Honeywell International Inc) and ammonium
sulfate ((NH
The proportionality between electrical conductivity and the mass
concentration of the salts was assessed (Fig. S2). All the saline
solutions showed a linear relationship between conductivity and mass
concentration, with a high correlation coefficient (
Comparison of the number concentration of aerosolized particles produced from inorganic salt solutions (sodium chloride, potassium chloride and ammonium sulfate) and snow samples as a function of salinity.
Comparison of the diameter of aerosolized particles produced from inorganic salt solutions (sodium chloride, potassium chloride and ammonium sulfate) and snow samples as a function of salinity. The diameter is expressed as the geometric mean diameter calculated from the number size distribution of aerosolized particles.
Although ambient salinity conditions were successfully reproduced, the SP2 was not able to operate stably under the most concentrated salt conditions. For the ambient samples, the SP2 was only exposed to the highest concentrations intermittently and for short times. However, for more focused laboratory conditions here, the range of BC and salt concentration that the SP2 was exposed to was limited, compared with ambient conditions, to ensure stable SP2 operation.
During the snow sample analysis, the large light extinction of the dense
aerosol produced from one extremely saline sample (
The prolonged sampling of the rich saline solution affected the flow system of the SP2. The sampling time needed to acquire a minimum of 30 000 recorded
particles substantially increased with decreasing FS concentration (Fig. S3). The prolonged nebulization (longer than 30 min) of 1 and 5
In this section, we investigate the possible impact of salt on the SP2 rBC
quantification, using the laboratory-generated snow sample proxies. When
nebulizing liquid samples, the overall mass quantification efficiency of rBC
(
Inorganic salt can alter water's physical properties, such as viscosity and
surface tension, modifying the size distribution of droplets produced in the
spray chamber and affecting the nebulization efficiency of various analytes (Todolí et al., 2002). High surface tension and viscosity cause an increase in the mean diameter of the liquid droplets suspended by pneumatic nebulizers, decreasing their transport efficiency (Sharp, 1988). NaCl concentrations above 1 g L
The transport losses were estimated for the SP2 sampling line, which was 30 cm long (distance from the Marin-5 exhaust) and composed of two different
sections. The first section was 24 cm long, with an internal diameter of 4.82 mm and an airflow rate of 1 L min
In this section, we will investigate the consequences of high number
particle density transiting the SP2 laser beam in its data acquisition
system at different acquisition settings. Additionally, the potential
quenching of incandescence caused by the presence of thick salt coatings on
rBC cores will be addressed.
A size-dependent
The frequency of recorded scattering events increased up to 17 500 counts per second for saline solutions with 250
First, the result obtained by triggering on the scattering detector (the
typical setup when operating the SP2 for atmospheric observations) will be
discussed.
Mass detection efficiency of the SP2 for different SP2 settings and
increasing electrical conductivity. Data are acquired from the analysis of a
fullerene soot suspension at 10
Triggering only off the incandescence signal reported 20 %–50 % more rBC
mass compared to triggering only off the scattering signal at higher
salinity (
rBC mass detection efficiency as a function of
Very thick coatings encapsulating rBC cores might not permit the SP2 laser to penetrate the coating, warm the rBC core, evaporate the coating and, finally, allow correct detection by bringing the rBC to vaporization temperature. This phenomenon will be called incandescence quenching. During the nebulization of saline samples, the salt contained in each droplet will remain on the rBC component after the water is evaporated and create thick coatings encapsulating the rBC.
As a rough estimation, we calculated that the theoretical coating thickness and coating–rBC mass ratio of spherical rBC cores have a diameter of 100,
200, 300, 400 and 500 nm as a function of salinity. We assumed the presence of a single rBC core per droplet, a concentric core shell geometry and NaCl density of 2170 kg m
To further investigate the potential quenching effect of salt, FS
suspensions (FS concentration of 10
Laboratory experiments were conducted to assess the interference caused by
inorganic salt on SP2 quantification of rBC mass and size distribution in
saline snow samples nebulized with a Marin-5. These experiments were
designed to reproduce the salinity conditions of snow samples collected over
the sea-ice-covered Fram Strait in summer 2017 during the PASCAL drift
shipborne campaign. Such salt concentrations might be exclusively
encountered in snow collected in coastal areas or over sea ice in the
vicinity of open water. The total mass quantification efficiency
(
The SP2 data were analyzed with PSI Toolkit single particle soot photometer (SP2), version 4.110. Contact Droplet Measurement Technologies to download the software.
Data are available upon request.
The supplement related to this article is available online at:
MZ collected the snow samples and performed the SP2, SMPS and physical measurements. OE performed the ALABAMA measurements. MZ drafted the paper, with contributions from all co-authors.
The authors declare that they have no conflict of interest.
This article is part of the special issue “Arctic mixed-phase clouds as studied during the ACLOUD/PASCAL campaigns in the framework of (AC)
We gratefully acknowledge the funding by the Deutsche Forschungsgemeinschaft (DFG – German Research Foundation; project ID 268020496 – TRR 172) within the Transregional Collaborative Research Center project of “ArctiC Amplification: Climate Relevant Atmospheric and SurfaCe Processes, and Feedback Mechanisms (AC)
This research has been supported by the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) (grant no. Projektnummer 268020496 – TRR 172).The article processing charges for this open-access publication were covered by the Alfred Wegener Institute, Helmholtz Centre for Polar and Marine Research (AWI).
This paper was edited by Jost Heintzenberg and reviewed by two anonymous referees.