High precision Pb isotope ratio analysis of wet depositions with low Pb concentration using multi-collector type inductively coupled plasma mass spectrometry and solid phase extraction

Abstract

Lead in wet deposition samples collected from remote mountainous areas in Japan, which exhibited low Pb concentrations (0.0020–1.94 µg L⁻¹) comparable to those found in Antarctic snow, was preconcentrated and separated from interfering components using a solid-phase extraction (SPE) procedure with a chelating resin under non-clean-room conditions. Subsequently, Pb isotope ratios were measured by MC-ICP-MS equipped with a desolvating nebulizer, applying mass discrimination correction based on Tl isotope ratio as an external standard. For snow samples, the relative standard deviation (RSD) of ²⁰⁸Pb/²⁰⁶Pb improved slightly from 0.0037% without the SPE to 0.0028% with the SPE. In contrast, the RSD of ²⁰⁷Pb/²⁰⁶Pb showed an improvement from 0.014% to 0.0017% with the SPE. Notably, the RSD of ²⁰⁴Pb/²⁰⁶Pb improved significantly from 0.19% without the SPE to 0.032% with the SPE. For rain samples, the RSDs were 0.0042% for ²⁰⁸Pb/²⁰⁶Pb, 0.0019% for ²⁰⁷Pb/²⁰⁶Pb, and 0.024% for ²⁰⁶Pb/²⁰⁴Pb. Without the SPE, the ²⁰⁶Pb/²⁰⁴Pb ratio exhibited a large error, making it difficult to distinguish between potential sources. However, with the SPE, the ²⁰⁶Pb/²⁰⁴Pb ratio was measured with sufficient precision to enable source discrimination.

---

1 Introduction

Lead is primarily released into the environment through human activities.¹ Due to its toxicity to humans and harmful effects on the environment, the behavior of Pb in the environment should be carefully monitored.² Atmospheric deposition is a major pathway for the transfer of substances from the atmosphere to terrestrial and marine surfaces,³ occurring via two primary mechanisms: wet deposition and dry deposition. For Pb, the wet deposition flux is greater than the dry deposition flux.^4,5 Furthermore, while dry deposition requires electricity for collection, wet deposition—such as rain and snow—can be collected passively, making it suitable for sampling in remote areas.⁶ Pb isotopic ratios in wet deposition from locations without anthropogenic activity provide valuable insights into the sources, long-range transport pathways, and relative contributions of airborne Pb.^7–12 However, precise measurement of Pb isotope ratios in such samples is generally difficult due to their low Pb concentrations.¹¹ Therefore requiring ultraclean sampling procedures, sample storage, pretreatment, and highly sensitive instrumentation.

For Antarctic snow samples ([Pb] 0.4–27.0 ng kg⁻¹), ²⁰⁶Pb/²⁰⁷Pb and ²⁰⁸Pb/²⁰⁷Pb were reported with relative standard deviations (RSDs) of 0.09–1.6% and 0.08–0.94% by MC-ICP-MS.¹² For another set of Antarctic snow samples ([Pb] 5 to 707 ng kg⁻¹), ²⁰⁷Pb/²⁰⁶Pb and ²⁰⁸Pb/²⁰⁶Pb were reported with RSDs of 0.023–0.23% and 0.014–0.12% by MC-ICP-MS combined with a torch integrated sample introduction system and a calibration method for mass discrimination correction based on Tl isotope standard (NIST SRM 997) doped into each sample.¹¹ In these studies, Pb was not chemically separated from coexisting materials, such as Na⁺, due to their low abundance in the snow samples.¹¹ Due to their lower Pb concentrations compared to other environmental samples, wet deposition samples from remote areas rarely yield reported cases of ²⁰⁴Pb-based isotope ratio measurements, even when analyzed using highly sensitive and precise instruments such as MC-ICP-MS or TIMS. It has been pointed out, however, that diagrams using ²⁰⁸Pb/²⁰⁶Pb versus²⁰⁷Pb/²⁰⁶Pb are unable to distinguish more than two sources of environmental Pb.¹³ This limitation arises because the linear trends in such plots are an inevitable consequence of the co-linearity of terrestrial Pb isotopes, and should not necessarily be interpreted as evidence of simple binary mixing of sources.^13,14 A more effective discrimination of lead isotope data in cases of multiple source mixing can be achieved using plots that incorporate ²⁰⁴Pb-based isotope ratios, which require measurement of the less abundant (1.4%) and non-radiogenic ²⁰⁴Pb isotope.^13–15 For the measurement of ²⁰⁴Pb in wet deposition samples from remote locations, pretreatment involving solid phase extraction (SPE)—such as preconcentration and removal of interfering components—is considered a viable solution. Nagaishi et al. reported that, for rock standard samples pretreated by SPE, MC-ICP-MS measurements using a desolvating nebulizer combined with a real-time online thallium addition system as an external standard enabled high-precision determination of ²⁰⁴Pb/²⁰⁶Pb at a Pb concentration of 1 µg L⁻¹, with a RSD of 0.008%.¹⁶ However, the effectiveness of SPE has not yet been validated for wet deposition samples with larger volumes and lower initial lead concentrations.

Yamamoto et al.¹⁷ reported Pb concentrations of 3.1 ± 1.3 and 23.0 ± 13.7 ng kg⁻¹ in rain and fresh snow collected in a remote mountainous area at 1400 m elevation in Shikoku, Japan, which are comparable to those found Antarctic and Arctic snow. Pretreatment for isotope ratio measurements of such low Pb concentration samples is typically conducted in a clean room to avoid contamination; however, such facilities are costly and not widely accessible. The aim of this study is to establish a method for the pretreatment of natural samples with low Pb concentration by using solid phase extraction (SPE) in a non-clean-room environment to concentrate Pb and separate it from coexisting elements, particularly Tl.^16,18 To this end, we assessed the precision of Pb isotope ratios in wet deposition samples obtained from mountainous areas in western Japan, including Shikoku, through SPE pretreatment.

2 Experimental

2.1 Instrumentation

A double-focusing MC-ICP-MS (NEPTUNE, Thermo Fisher Scientific, MA, USA) with a desolvating nebulizer (Aridus IITM, Teledyne CETAC Technologies, NE, USA) was used for Pb isotope measurement. An auto sampler (SC-Micro, Elemental Scientific Inc., Germany) and a peristaltic pump (MP2, Elemental Scientific Inc., Germany) were used for sample introduction into the MC-ICP-MS. A quadrupole-type inductively coupled plasma mass spectrometer (ICP-QMS, X series II, Thermofisher Scientific, MA, USA) was used for the determination of Pb concentration in filtered rain and snow samples. An electrothermal atomic absorption spectrometer (ETAAS, Z-2710, Hitachi High-Tech, Japan) with a W-treated graphite furnace and Pd modifier, and a 100 µL sample injection volume (Imai et al., 1998) was used for the determination of Pb. The limit of detection (LOD) values of ICP-QMS and ETAAS were 0.5 ng L⁻¹ and 0.02 µg L⁻¹, respectively. A commercial syringe-type column of NOBIAS Chelate PA-1 resin (Hitachi High-Tech, Japan), containing iminodiacetic acid and ethylenediaminetriacetic acid groups on hydrophilic methacrylate substrate was used for solid phase extraction.¹⁹ The column was placed at the column holder in a solid phase extractor (SPE-100, HIRANUMA, Japan) equipped with a dual-plunger pump with PTFE plunger and glass cylinder. The column was repeatedly used for 10 rain samples and 7 snow samples. All preparations were carried out in a non-clean room laboratory with a clean air system using a HEPA filter, and an anteroom separated from inside public areas, where no outdoor shoes were allowed. Isotope measurements were carried out in a class-1000 clean room at the Kochi Core Center. Ultrapure water purified by Milli-Q Academic (Merck Millipore, MA, USA) and pure water produced by Elix (Essential, Merck Millipore, MA, USA) were used for all preparation. A polysulfone filter holder (KP-47S, ADVANTEC, Japan), and a mixed cellulose ester membrane filter (pore size 0.45 µm, diameter 47 mm, ADVANTEC, Japan) were used for filtration of rain and snow samples.

2.2 Solid phase extraction (SPE)

After adjusting the sample pH to 5.5 using 0.1 mol L⁻¹ ammonium acetate, SPE was performed according to Table 1. The sample and reagent solutions were placed in bottles sealed with screw caps containing a small hole at the top (diameter: 3 mm). The bottle materials are listed in Table 1. A PFA tube (diameter 2 mm, length 1 m) was inserted in each bottle through the hole, and the opposite side was connected to the solid phase extractor. The small gap between the PFA tube and the screw cap was sealed using parafilm®. Air was filtered through a 0.45 µm membrane filter prior to introduction into the solid phase extractor. The eluate was collected in a PFA vessel.

Table 1 Procedure of solid phase extraction

Stage	No.	Solution	Vial	Volume/mL	Flow rate/mL min^−1
Cleaning	1	3 mol L^−1 HNO3	PP	20	10
	2	Air		3	10
Conditioning	3	Acetone	Glass	5	10
	4	Ultrapure water	PP	10	10
	5	3 mol L^−1 HNO3	PP	10	10
	6	Ultrapure water	PP	10	10
	7	0.1 mol L^−1 CH3COONH4	PP	10	10
Sample loading	8	Sample	PE	100–2000	5
	9	Air		3	10
Column wash	10	Ultrapure water	PP	10	10
	11	Air		3	10
Elution	12	3 mol L^−1 HNO3	PP	6	1

---

The eluate was completely evaporated on an aluminum hot plate (EHP-170N, AS ONE, Japan). The hot plate was placed in a PTFE chamber (A-2, SAN-AI KAGAKU, Japan) lined with a PFA film. An air pump (Clean air Z, SAN-AI KAGAKU, Japan), equipped with a polypropylene filter (pore size 1.0 µm) and a silicon tube, was used to deliver filtered air into the chamber. After evaporation, the residue was redissolved in 2% nitric acid in a 7 mL PFA vessel (Savillex, MN, USA). The concentration of Pb in the nitric acid was adjusted to exceed 1 µg L⁻¹ for isotopic measurements and verified using ICP-QMS and ETAAS. The recovery was calculated based on the Pb concentrations before and after SPE treatment. The average recovery was 100 ± 10% (n = 17).

2.3 Isotope measurement

Lead isotopes were measured using the MC-ICP-MS equipped with a desolvating nebulizer. Instrumental parameters are listed in Table 2. Fig. 1 shows a schematic illustration of the Pb isotope measurement setup. Isobaric interference from ²⁰⁴Hg to ²⁰⁴Pb was corrected using the ²⁰⁴Hg/²⁰¹Hg ratio reported in the IUPAC Technical Report (0.521244).²⁰ Since ²⁰¹Hg intensity was less than 0.01% of the ²⁰⁴Pb signal for all samples, the interference from ²⁰⁴Hg to ²⁰⁴Pb was considered negligible. The Pb isotope standard solution was measured after every three to five samples. The concentration of Tl isotope standard solution after mixing with the sample solution was adjusted to 5 µg L⁻¹.

Table 2 Instrumental parameters for MC-ICP-MS

ICP ion source
RF frequencies	27.12 MHz
RF power	1200 W
Interface pump	OnToolBooster 150
Sampling cone	Jet cone (Ni)
Skimmer cone	X-skimmer cone (Ni)
Ar gas flow rate
Cooling gas	15 L min^−1
Auxiliary gas	0.70 L min^−1
Sample gas	0.90 L min^−1
Data acquisition parameters
Integration time	8 s
Number of cycles	30 cycles
Cup configuration	L4: ^201Hg; L3: ^203Tl; L2: ^204Pb; L1: ^205Tl;
	C: ^206Pb; H1: ^207Pb; H2: ^208Pb
Resistivity of the pre-amplifier	L3, L1, C, H1: 10^11Ω; H1, H2: 10^12Ω
Desolvating nebulizer
Nebulizer	C-flow PFA
Spray chamber	PFA
Sample uptake gas flow	80 µL min^−1 (40 µL min^−1 both for sample and Tl)
Ar sweep gas flow rate	6–7 L min^−1
N2 add gas flow rate	1–3 L min^−1
Spray chamber temperature	110 °C
Membrane desolvator temperature	160 to 100 °C

---

Fig. 1 Schematic illustration of the Pb isotope measurement set up.

2.4 Materials

Standard materials of Pb (NIST SRM 981, NIST, MD, USA) and Tl (NIST SRM 997) were used for calibration of mass discrimination during the isotope measurements. The certified values for the calibration were 16.937 ± 0.0011 for ²⁰⁶Pb/²⁰⁴Pb, 0.91464 ± 0.00033 for ²⁰⁷Pb/²⁰⁶Pb, 2.1681 ± 0.0008 for ²⁰⁸Pb/²⁰⁶Pb,²¹ and 2.38714 ± 0.00101 for ²⁰⁵Tl/²⁰³Tl.²² A commercially available Pb standard solution (1000 mg L⁻¹, Cica-reagent, KANTO CHEMICAL, Japan) was used to prepare an in-house Pb isotope standard solution (20 µg L⁻¹). Ultra-pure nitric acid (69%, Ultrapur-100, KANTO CHEMICAL, Japan), acetone (99.8%, environmental analysis grade, FUJIFILM Wako Pure Chemical, Japan), and ammonium acetate (Cica-Reagent, KANTO CHEMICAL, Japan) were used. Lead standard solution (1000 mg L⁻¹, for atomic absorption spectrometry, NAKARAI TESQUE, Japan) was used for ICP-QMS and ETAAS analyses. A palladium matrix modifier (10 000 mg L⁻¹, Cica-Reagent for atomic absorption spectrometry, KANTO CHEMICAL, Japan) and sodium tungstate(VI) dihydrate (Cica-Reagent, KANTO CHEMICAL, Japan) were used for ETAAS analysis. Reagent-grade nitric acid (60%, Extra pure reagent, NAKARAI TESQUE, Japan) and poly(oxyethylene)alkylether (Contaminon® L, Wako Pure Chemicals, Japan) were used to clean instruments and containers prior to use.

2.5 Rain and snow samples

Sampling site locations were listed in Table 3. Rain was collected at sampling sites 1, 2, and 3, and fresh snow was collected at all locations except site 3. Rain was collected in a low-density polyethylene (LDPE) bottle (Nalgene, Thermo Fisher Sci., MA, USA) using a polyethylene (PE) funnel (diameter: 300 mm) and a PFA tube (Takano et al., 2021). A PE mesh was placed over the top of the PFA tube to exclude larger particles. The funnel was positioned 1.5 m above the ground. These components were installed at each sampling site one day prior to the rainfall event and retrieved within 24 hours after the event. Rain samples were transported to the laboratory using a heat-insulated container with a refrigerant. The samples were prepared by vacuum filtration using a lidded filter holder.

Table 3 Locations of sampling sites

	Location	Longitude		Latitude		Altitude/m
St. 1	Mt. Kajigamori, Kochi Pref.	133.75°	E	33.76°	N	1400
St. 2	Mt. Suihamine, Ehime Pref.	133.54°	E	33.94°	N	870
St. 3	Okawa, Kochi Pref.	133.47°	E	33.83°	N	830
St. 4	Besshiyama, Ehime Pref.	133.43°	E	33.86°	N	810
St. 5	Mt. Osorakan, Hiroshima Pref.	132.13°	E	34.60°	N	1260
St. 6	Shobara, Nanatsuka, Hiroshima Pref.	132.98°	E	34.82°	N	350
St. 7	Mt. Shimokura, Iwate Pref.	140.93°	E	39.89°	N	1200
St. 8	Mt. Nokaike, Tokushima Pref.	133.71°	E	33.85°	N	900

---

Fresh snow was collected as soon as possible (within 24 h) after each snow fall event, avoiding melting and compaction. At sites 1, 2, 3, and 8, fresh snow was collected from a PE container (length 500 mm, width 350 mm, height 150 mm) positioned 1 m above the ground. At other sites, fresh snow was collected within 10 cm of the top of the accumulated snow, after removing the uppermost 1 cm layer. Each sample was collected in a Ziploc® freezer bag (Asahi Kasei, Japan) using a polycarbonate shovel (ASONE, Japan). The samples were transported to the laboratory using a heat-insulated container with a refrigerant. Each sample was melted at room temperature in a lidded filter holder, and subsequently prepared by vacuum filtration. Blanks for sampling containers were determined using 500 mL of ultrapure water adjusted to pH 4 with nitric acid, corresponding to the pH of rain and melted snow. All container blanks of Pb were below the limit of detection (LOD) of ICP-QMS (0.5 ng L⁻¹; equivalent to 0.25 ng per sample).

3 Results and discussion

3.1 Blanks

The blank of ultrapure water, reagents used in this study were below the LOD of ICP-QMS (0.5 ng L⁻¹). The blanks of the sample containers, evaluated by adding ultrapure water to each container and storing for 24 hours, were also below the LOD of ICP-QMS. The blank for the SPE procedure was assessed using 250 mL of ultrapure water adjusted to pH 5.5. The Pb blank in the eluate after the SPE procedure was below the LOD of ETAAS (0.02 µg L⁻¹; equivalent to 0.12 ng of Pb per 250 mL).

3.2 Conditions for the SPE procedure

The conditions for pH, sample volume, and sample flow rate in the SPE procedure were systematically optimized. The effectiveness of SPE treatment in removing Tl was confirmed using snow samples collected in 2017. Details are provided in the SI (Table S1 and Fig. S1–S3). As a result, the optimal conditions were determined to be pH 5.5, a sample volume of 2000 mL or less, and a sample flow rate of 5 mL min⁻¹. On average, 96.3 ± 1.5% of Tl was removed from the snow samples following SPE treatment. The residual Tl in the samples after SPE treatment was less than 0.018% of the Tl added as an isotope standard.

3.3 Concentrations and isotope ratios

Lead concentrations in rain were from 0.0020 to 0.408 µg L⁻¹, whereas those in snow were from 0.159 to 1.94 µg L⁻¹ (Table 4). Concentration factors for rain by SPE ranged from 10-fold to 100-fold, whereas those for snow ranged from 2-fold to 20-fold. Only R5 did not reach 1 µg L⁻¹ by SPE due to its small sample volume. Since SPE was not conducted from S8 to S17, Pb isotopes were measured at concentrations below 1 µg L⁻¹, except for S10 and S11.

Table 4 Initial concentration and isotope ratios of Pb in rain and snow samples

No.	Date	Location	Pb/µg L^−1	^206Pb/^204Pb				^207Pb/^206Pb				^208Pb/^206Pb
							RSD/%				RSD/%				RSD/%
Rain samples with SPE
R1	2018, Dec. 19	St. 1	0.0403	18.0812	±	0.0022	0.0123	0.862517	±	0.000014	0.0017	2.10947	±	0.00010	0.0047
R2	2019, Jan. 31	St. 1	0.408	18.1607	±	0.0028	0.0155	0.859571	±	0.000011	0.0013	2.10551	±	0.00006	0.0031
R3	2019, Mar. 4	St. 1	0.0207	17.9766	±	0.0018	0.0100	0.867593	±	0.000018	0.0021	2.12250	±	0.00009	0.0041
R4	2019, Mar. 7	St. 1	0.0520	18.0203	±	0.0023	0.0126	0.865254	±	0.000014	0.0016	2.10570	±	0.00028	0.0132
R5	2019, Jun. 8	St. 2	0.0335	18.0243	±	0.0197	0.1093	0.864657	±	0.000027	0.0032	2.06412	±	0.00007	0.0032
R6	2019, Jun. 8	St. 3	0.0223	18.1584	±	0.0020	0.0108	0.860061	±	0.000016	0.0019	2.10668	±	0.00006	0.0028
R7	2019, Jun. 8	St. 1	0.0079	18.0726	±	0.0023	0.0126	0.863083	±	0.000018	0.0021	2.11203	±	0.00007	0.0034
R8	2019, Jul. 1	St. 2	0.190	18.1852	±	0.0062	0.0338	0.858531	±	0.000014	0.0017	2.10101	±	0.00006	0.0028
R9	2019, Jul. 1	St. 3	0.0020	18.0611	±	0.0016	0.0091	0.864155	±	0.000017	0.0020	2.10151	±	0.00005	0.0022
R10	2019, Jul. 1	St. 1	0.239	18.2028	±	0.0032	0.0173	0.858011	±	0.000016	0.0018	2.11618	±	0.00006	0.0029
Snow samples with SPE
S1	2018, Dec. 15	St. 1	1.76	17.8834	±	0.0057	0.0320	0.871013	±	0.000015	0.0018	2.11006	±	0.00005	0.0024
S2	2018, Dec. 29	St. 4	1.33	17.5897	±	0.0051	0.0291	0.883994	±	0.000015	0.0017	2.10899	±	0.00006	0.0030
S3	2018, Dec. 30	St. 5	0.952	17.9936	±	0.0049	0.0273	0.867233	±	0.000013	0.0015	2.09967	±	0.00006	0.0029
S4	2019, Jan. 31	St. 1	0.159	17.9944	±	0.0054	0.0302	0.866815	±	0.000015	0.0018	2.10044	±	0.00006	0.0030
S5	2019, Feb. 10	St. 6	0.425	17.9798	±	0.0055	0.0307	0.867315	±	0.000014	0.0016	2.10899	±	0.00006	0.0030
S6	2019, Mar. 8	St. 7	0.207	17.9817	±	0.0068	0.0380	0.866305	±	0.000018	0.0021	2.11610	±	0.00006	0.0030
S7	2019, Mar. 8	St. 1	1.24	18.0669	±	0.0063	0.0346	0.862929	±	0.000011	0.0012	2.09905	±	0.00005	0.0022
Snow samples without SPE
S8	2018, Jan.24	St. 1	0.937	18.0908	±	0.0798	0.4412	0.862662	±	0.000407	0.0472	2.11006	±	0.00005	0.0024
S9	2018, Jan.24	St. 2	0.252	18.0018	±	0.1253	0.6961	0.865764	±	0.000157	0.0181	2.10570	±	0.00028	0.0132
S10	2018, Jan.24	St. 8	1.94	17.7293	±	0.0206	0.1163	0.878112	±	0.000025	0.0029	2.10917	±	0.00008	0.0038
S11	2018, Jan.24	St. 6	1.41	17.9514	±	0.0020	0.0114	0.867809	±	0.000023	0.0027	2.13116	±	0.00004	0.0020
S12	2018, Jan.25	St. 5	0.583	18.0522	±	0.0262	0.1452	0.864448	±	0.000037	0.0043	2.11490	±	0.00006	0.0030
S13	2018, Feb. 1	St. 5	0.391	17.9320	±	0.0109	0.0607	0.867890	±	0.000088	0.0101	2.11005	±	0.00006	0.0029
S14	2018, Feb. 2	St. 2	0.233	17.9851	±	0.0502	0.2794	0.859369	±	0.000256	0.0298	2.11572	±	0.00005	0.0024
S15	2018, Feb. 2	St. 8	0.199	18.1572	±	0.0229	0.1260	0.858290	±	0.000195	0.0227	2.11610	±	0.00006	0.0030
S16	2018, Dec. 29	St. 1	0.955	17.7776	±	0.0019	0.0104	0.876082	±	0.000019	0.0022	2.13902	±	0.00003	0.0016
S17	2019. Feb. 16	St. 1	0.227	18.1170	±	0.0076	0.0419	0.861053	±	0.000036	0.0042	2.11757	±	0.00007	0.0031

---

Isotope ratios (²⁰⁶Pb/²⁰⁴Pb, ²⁰⁷Pb/²⁰⁶Pb, and ²⁰⁸Pb/²⁰⁶Pb) were listed in Table 4. The ²⁰⁶Pb/²⁰⁴Pb, ²⁰⁷Pb/²⁰⁶Pb and ²⁰⁸Pb/²⁰⁶Pb values of rain measured by MC-ICP-MS range from 17.9766 to 18.2028, 0.85801 to 0.86759 and 2.06412 to 2.12250, respectively. Those values of snow measured by MC-ICP-MS ranged from 17.5897 to 18.1572, 0.85829 to 0.88399, and 2.09905 to 2.13902. Measured values of NIST SRM981 by MC-ICP-MS, used for the calibration of mass discrimination, were 16.9326, 0.914483 and 2.16614. The average ²⁰⁵Tl/²⁰³Tl ratio was 2.41050 ± 0.00075 and remained nearly constant throughout the measurements (Fig. S4). No significant differences were observed in the ²⁰⁵Tl/²⁰³Tl ratio between samples with and without SPE treatment, implying that serious interference from initial Tl in snow samples was negligible even without SPE.

Fig. 2a shows the relative standard deviation (RSD) of ²⁰⁶Pb/²⁰⁴Pb for rain, snow, and NIST SRM 981. For rain samples, R5 (0.109%) and R8 (0.0338%) exhibited relatively high RSDs. In the former case, the post-SPE concentration was below 1 µg L⁻¹, and both ²⁰⁴Pb and ²⁰⁶Pb signal intensities were lower than those of the other samples, which is considered to be the reason for the high RSD. In the latter case, the sample was the first measured among all samples and exhibited a slightly higher ²⁰⁵Tl/²⁰³Tl ratio (Fig. S4). It was likely measured prior to full stabilization of the MC-ICP-MS, which may have contributed to the high RSD. The RSDs of other rain samples ranged from 0.0091 to 0.0173%. For snow samples processed with the SPE, the RSDs ranged from 0.0273 to 0.0380%. In contrast, snow samples without the SPE exhibited RSDs ranging from 0.0104 to 0.125%, with greater variability than that observed in SPE-treated samples. The RSDs for both rain and snow were greater than that of the standard reference material (0.0047%). Fig. 2b shows the RSDs of ²⁰⁷Pb/²⁰⁶Pb. For rain samples, the RSDs ranged from 0.0013 to 0.0032%, which were lower than those observed for ²⁰⁶Pb/²⁰⁴Pb. For snow samples processed with the SPE, the RSDs ranged from 0.012 to 0.0018%, with less variation than that without the SPE. In contrast, snow samples without the SPE exhibited RSDs ranging from 0.0022 to 0.0472%. Although the RSDs were lower than those observed for ²⁰⁶Pb/²⁰⁴Pb, some samples still showed relatively high RSD values. The RSDs of rain and snow samples processed with the SPE were as low as that of the NIST SRM (0.0016%). Fig. 2c shows the RSDs of ²⁰⁸Pb/²⁰⁶Pb. For rain samples, the RSDs ranged from 0.0022 to 0.0047%, except for R4, which exhibited a higher RSD of 0.0132%. The RSDs of ²⁰⁸Pb/²⁰⁶Pb were lower than those of ²⁰⁶Pb/²⁰⁴Pb and comparable to those of ²⁰⁷Pb/²⁰⁶Pb. For snow samples, the RSDs ranged from 0.0022 to 0.0030% with the SPE, and from 0.0016 to 0.0038% without the SPE, except for S9, which exhibited a higher RSD of 0.0132%. The RSDs for ²⁰⁸Pb/²⁰⁶Pb were comparable to that of the NIST SRM (0.0024%) regardless of the SPE treatment.

Fig. 2 Box plot for the relative standard deviation of ²⁰⁶Pb/²⁰⁴Pb, ²⁰⁷Pb/²⁰⁶Pb, and ²⁰⁸Pb/²⁰⁶Pb in rain and snow samples. (a) ²⁰⁶Pb/²⁰⁴Pb, (b) ²⁰⁷Pb/²⁰⁶Pb, and (c) ²⁰⁸Pb/²⁰⁶Pb. Open circles indicate outliers, which were identified based on the first and third quartiles (Q1 and Q3) and the interquartile (IQR). Closed circles indicate the NIST SRM 981.

3.4 Comparison of isotopic precision with previous studies

The analytical precision of ²⁰⁶Pb/²⁰⁴Pb of rain samples with the SPE (RSD: 0.009 to 0.11%) was lower than that for rain collected in the Tokyo urban area measured by MC-ICP-MS (RSD: 0.1 to 0.3%).²³ The RSD values of rain samples were comparable to those of JA-1 (0.008% at 1 µg L⁻¹ Pb) and JMn-1 (0.009% at 1 µg L⁻¹ Pb),¹⁶ which are reference materials with high Pb concentration obtained from the Geological Survey of Japan (GSJ) and prepared using SPE treatment, except for R5 and R8. These results confirm that the preconcentration of Pb and removal of interfering elements in rain samples using the SPE treatment employed in this study are sufficient for precise ²⁰⁴Pb measurement. The RSD values of ²⁰⁶Pb/²⁰⁴Pb in snow samples with the SPE were higher than those in rain samples, except for R5 and R8. The Pb concentration in snow samples after the SPE is nearly identical to those in rain samples, suggesting that the difference in RSDs is attributable to coexisting components present at relatively higher levels in snow samples, even after the SPE treatment. Interference from these components is most pronounced in snow samples without the SPE, resulting in relatively high RSDs.

The RSD of ²⁰⁷Pb/²⁰⁶Pb of rain were lower than those reported for rain from Taiwan measured by sector-field (SF) ICP-MS (RSD 0.1%),²⁴ rain from Tokyo measured by MC-ICP-MS (RSD 0.05–0.15%),²³ and rain from São Paulo measured by TIMS (RSD 0.1%).²⁵ Similarly, the RSD of snow samples with the SPE were lower than those reported for snow from a remote mountain in the Czech Republic measured by SF-ICP-MS (RSD 0.4%)¹⁰ and snow from the French Alps measured by TIMS (RSD 0.05–0.19%).⁸ The RSD of ²⁰⁷Pb/²⁰⁶Pb in rain and snow samples with the SPE were nearly identical to that of NIST SRM, which was measured without any interference, indicating that SPE effectively removes coexisting components that interfere with ²⁰⁷Pb/²⁰⁶Pb measurements.

The RSD of ²⁰⁸Pb/²⁰⁶Pb in rain and snow samples, both with and without the SPE, were lower than those for Antarctic snow measured by MC-ICP-MS (RSD 0.04%).¹¹ Since the RSDs in snow samples without the SPE, except for S9, were comparable to that of the NIST SRM, suggesting that interference from coexisting components did not significantly affect the ²⁰⁸Pb/²⁰⁶Pb measurement.

3.5 Evaluation of isotopic precision for sources discrimination

Reported ²⁰⁸Pb/²⁰⁶Pb ratios of Chinese loess²⁶ and Chinese coal,²⁷ considered potential sources of lead in the rain and snow samples collected in this study, are 2.1340 and 2.0720, respectively, yielding a difference of 0.0620. The mean error (2SD) of ²⁰⁸Pb/²⁰⁶Pb obtained in this study was 0.00018 for rain with the SPE, 0.00012 for snow with the SPE, and 0.00016 for snow without the SPE, demonstrating sufficient analytical precision to distinguish between natural and anthropogenic sources. Comparable analytical precision has also been reported in previous studies for Antarctic snow measured by MC-ICP-MS without SPE.¹¹ Conversely, the analysis of Greenland snow conducted with ICP-QMS exhibited relatively large 2SD values,²⁸ which were insufficient to distinguish between natural and anthropogenic sources. Thus, the measurements of ²⁰⁸Pb/²⁰⁶Pb demonstrate that, for precise determination of this isotope ratio in natural water samples with low concentrations of Pb and interfering components, the use of an analytical instrument with high mass resolution is more critical than pretreatment procedures.

Fig. 3 presents the ²⁰⁷Pb/²⁰⁶Pb and ²⁰⁶Pb/²⁰⁴Pb isotope ratio plots for snow samples with and without the SPE, as well as those for potential source materials.^29–34 Snow samples with and without the SPE were both plotted in proximity to urban atmospheric sources from Shanghai and Xiamen, incineration fly ash, and unleaded gasoline. With the SPE, the average errors for ²⁰⁷Pb/²⁰⁶Pb and ²⁰⁶Pb/²⁰⁴Pb were 0.00003 and 0.0113, respectively, enabling significant differentiation between the isotope ratios of individual samples and those of potential sources. In contrast, without the SPE, the average errors were 0.00025 and 0.0695, and the ²⁰⁶Pb/²⁰⁴Pb ratios could not be used to distinguish the candidate sources. The analysis of ²⁰⁴Pb is technically challenging due to its low natural abundance, which has limited the practical application of ²⁰⁴Pb-based isotope ratios despite their considerable analytical significance. The results of this study demonstrate that the SPE enables high-precision measurement of the ²⁰⁶Pb/²⁰⁴Pb ratio in natural water samples with low lead concentrations (as low as 0.002 µg L⁻¹). Furthermore, because the SPE can be conducted in non-cleanroom environments, it offers high throughput for sample pretreatment. This methodological advancement is expected to facilitate further research on identifying lead sources in remote regions.

Fig. 3 Comparison of Pb isotopic composition for rain and snow samples with and without the SPE. (a) Red diamonds indicate rain samples with the SPE, (b) blue diamonds indicate snow samples with the SPE, and (c) green diamonds indicate snow samples without the SPE. Gray circles represent fly ash from an Australian coal power plant,²⁹ and light blue circles represent Japanese waste incineration fly ash.³⁰ The pink area indicates unleaded gasoline;³¹ the purple, orange, and green areas indicate urban aerosols from Shanghai,³² Xiamen,³³ and Nanjing,³⁴ respectively.

4 Conclusions

Pb isotope ratios (²⁰⁶Pb/²⁰⁴Pb, ²⁰⁷Pb/²⁰⁶Pb, and ²⁰⁸Pb/²⁰⁶Pb) of rain and snow at concentrations below 0.1 µg L⁻¹, collected from remote areas in Japan, were precisely determined by MC-ICP-MS with a desolvating nebulizer, and mass discrimination correction based on Tl isotopes. Sample pretreatment processed using SPE with a chelating resin was conducted in a non-cleanroom laboratory. The Pb blanks from containers and pretreatment were below the LOD of the measuring instrument, indicating that contamination was sufficiently minimized even when pretreatment was conducted in a non-cleanroom environment. The RSD value of ²⁰⁶Pb/²⁰⁴Pb of rain at a concentration of 0.0020 µg L⁻¹ was 0.0091%, whereas that of snow at a concentration of 0.159 µg L⁻¹ was 0.0302%. Analytical method of this study was able to measure Pb isotope ratios in wet depositions with equal or better precision than previous reports for rain and snow.

Although rain and snow collected from remote areas have fewer interfering coexisting compositions than other natural water samples, when using a desolvationg nebulizer, the precision of their Pb isotope ratios without the SPE would be degraded by a small amount of coexisting compositions, especially for ²⁰⁴Pb. While the SPE pre-treatment resulted in no observable change in the precision of the ²⁰⁸Pb/²⁰⁶Pb ratio and a slight improvement in that of the ²⁰⁷Pb/²⁰⁶Pb ratio, these variations were not substantial enough to affect the discrimination of candidate source isotope ratios. In contrast, the precision of the ²⁰⁶Pb/²⁰⁴Pb ratio without the SPE was inadequate for discrimination of candidate sources, whereas SPE treatment enhanced the analytical precision sufficiently to allow for reliable source attribution. The results of this study indicated that separation by SPE is effective for high-precision isotope measurements, even in natural water samples presumed to contain minimal interfering constituents. The precise determination of ²⁰⁴Pb isotope ratios facilitates detailed source attribution of low-level Pb in remote regions, offering valuable insights into the environmental behavior of Pb.

Author contributions

Y. Yamamoto wrote the initial manuscript, complied, analyzed and visualized the data. R. Nakada and K. Nagaishi performed Pb isotopic analyses. S. Tokoro performed rain and snow sampling, SPE treatment, and Pb concentration analyses. Y. Kikuchi, and J. Nishimoto performed snow sampling. S. Imai ensured funding, supervised the project and reviewed the manuscript. All authors contributed to the discussion and interpretation of the results.

Conflicts of interest

The authors declare no conflict of interest.

Data availability

Fundamental data supporting the findings of this study, including Pb concentrations and isotope ratios obtained by MC-ICP-MS, ICP-QMS, and ETAAS, are available in the main text and supplementary information (SI). Supplementary information: Pb recovery rates and Tl removal efficiencies during SPE treatment, as well as measurement data for the Tl isotope ratio. See DOI: https://doi.org/10.1039/d5ja00408j.

Additional data may be made available from the corresponding author upon reasonable request.

Acknowledgements

This study was supported by JSPS KAKENHI Grant Number JP 26340084 and 20K05585.

References

1. M. Komárek, V. Ettler, V. Chrastný and M. Mihailjevič, Environ. Internat., 2008, 34, 562.
2. A. L. Wani, A. Ara and J. A. Usmani, Interdiscip. Toxicol., 2015, 8, 55.
3. J. M. Prospero, R. J. Charlson, V. Mohnen, R. Jaenicke, A. C. Delany, J. Moyers, W. Zoller and K. Rahn, Rev. Geophys., 1983, 21, 1607.
4. R. A. Duce, P. S. Liss, J. T. Merrill, E. L. Atlas, P. Buat-Menard, B. B. Hicks, J. M. Millertl, J. M. Prospero, R. Arimoto, T. M. Church, W. Ellis, J. N. Galloway, L. Hansen, T. D. Jickells, A. H. Knap, K. H. Reinhardt, B. Schneider, A. Soudine, J. J. Tokos, S. Tsunogai, R. Wollast and M. Zhou, Glob. Biogeochem. Cycles, 1991, 5, 193.
5. R. Chance, T. D. Jickells and A. R. Baker, Mar. Chem., 2015, 177, 45.
6. Y. Yamamoto, K. Oka, T. Shunichi, N. Nishii, Y. Kikuchi, J. Nishimoto and S. Imai, Anal. Sci., 2023, 39, 679.
7. J. G. Farmer, L. J. Eades, M. C. Graham and J. R. Bacon, J. Environ. Monit., 2000, 2, 49.
8. A. M. Veysseyre, A. F. Bollhöfer, K. J. R. Rosman, C. P. Ferrari and C. F. Boutron, Environ. Sci. Technol., 2001, 35, 4463.
9. J. M. Luck and D. B. Othman, Chem. Geol., 2002, 182, 443.
10. N. Cimova, M. Novak, V. Chrastny, J. Curik, F. Veselovsky, V. Blaha, E. Prechova, J. Pasava, M. Houskova, L. Bohdalkova, M. Stepanova, J. Mikova, M. Krachler and A. Komarek, Atmos. Environ., 2016, 143, 51.
11. A. Bazzano, K. Latruwe, M. Grotti and F. Vanhaecke, J. Anal. At. Spectrom., 2015, 30, 1322.
12. S. Bertinetti, F. Ardini, M. A. Vecchio, L. Caiazzo and M. Grotti, Chemosphere, 2020, 255, 126858.
13. R. M. Ellam, Sci. Tot. Environ., 2020, 408, 3490.
14. B. Gulson, G. D. Kamenov, W. Manton and M. Rabinowitz, Int. J. Environ. Res. Public Health, 2018, 15, 723.
15. M. Grotti, M. A. Vecchio, D. Gobbato, M. Mataloni and F. Ardini, J. Anal. At. Spectrom., 2023, 38, 1057.
16. K. Nagaishi, R. Nakada and T. Ishikawa, Geochem. J., 2021, 55, 1.
17. Y. Yamamoto, Y. Sanagawa, Y. Kurumi, A. Saito, J. Nishimoto, Y. Kikuchi and S. Imai, Bunseki Kagaku, 2019, 68, 51.
18. M. Tanimizu and T. Ishikawa, Geochem. J., 2006, 40, 121.
19. Y. Sohrin, S. Urushihara, S. Nakatsuka, T. Kono, E. Higo, T. Minami, K. Norisuye and S. Umetani, Anal. Chem., 2008, 80, 6267.
20. J. R. de Laeter, J. K. Böhlke, P. de Bièvre, H. Hidaka, H. S. Peiser, K. J. R. Rosman and P. D. P. Taylor, Pure Appl. Chem., 2003, 75, 683.
21. E. J. Catanzaro, T. J. Murphy, W. R. Shields and E. L. Garner, J. Res. Natl. Bur. Stand. A Phys. Chem., 1968, 72A, 261.
22. L. P. Dunstan, J. W. Gramlich, I. L. Barnes and W. C. Purdy, J. Res. Natl. Bur. Stand., 1980, 85, 1.
23. T. Shimamura, S. Iijima, M. Iwashita, M. Hattori and Y. Takaku, Atmos. Environ., 2007, 41, 3797.
24. M. C. Cheng, C. F. You, F. J. Lin, K. F. Huang and C. H. Chung, Atmos. Environ., 2011, 45, 1919.
25. M. Babinski, C. Aily, I. R. Ruiz and K. Sato, J. Phys. IV France, 2003, 107, 87.
26. F. Wu, S. S. H. Ho, Q. Sun and S. H. S. Ip, Terr. Atmos. Ocean. Sci., 2011, 22, 305.
27. X. Y. Bi, Z. G. Li, S. X. Wang, L. Zhang, R. Xu, J. L. Liu, H. M. Yang and M. Z. Guo, Environ. Sci. Technol., 2017, 51, 13502.
28. B. Astray, A. Sípkova, D. Baragano, J. Pechar, R. Krejci, M. Komarek and V. Chrastný, Environ. Pollut., 2024, 345, 123457.
29. M. Chiaradia, B. E. Chenhall, A. M. Depersb, B. L. Gulson and B. G. Jones, Sci. Tot. Environ., 1997, 205, 107.
30. M. Sakata, M. Kurata and N. Tanaka, Geochem. J., 2000, 34, 23.
31. P. H. Yao, G. S. Shyu, Y. F. Chang, Y. C. Chou, C. C. Shen, C. S. Chou and T. K. Chang, Int. J. Environ. Res. Public Health, 2015, 12, 4602.
32. J. Zheng, M. Tan, Y. Shibata, A. Tanaka, Y. Li, G. Zhang, Y. Zhang and Z. Shan, Atmos. Environ., 2004, 38, 1191.
33. L. Zhu, J. Tang, B. Lee, Y. Zhang and F. Zhang, Mar. Pollut. Bull., 2010, 60, 1946.
34. L. Hua, H. Ma and J. Ji, Environ. Monit. Assess, 2011, 173, 361.

---