Marine Chemistry 166 (2014) 114–121

Contents lists available at ScienceDirect

Marine Chemistry

j ourna l homepage: www.e lsev ie r .com/ locate /marchem

Precise determination of dissolved platinum in seawater of the Japan Sea,Sea of Okhotsk and western North Pacific Ocean

Asami Suzuki a,⁎, Hajime Obata a, Ayako Okubo b,1, Toshitaka Gamo a

a Atmosphere and Ocean Research Institute, The University of Tokyo, Japanb Japan Atomic Energy Agency, Japan

⁎ Corresponding author.1 Present affiliation.

http://dx.doi.org/10.1016/j.marchem.2014.10.0030304-4203/© 2014 Elsevier B.V. All rights reserved.

a b s t r a c t

a r t i c l e i n f o

Article history:Received 30 December 2013Received in revised form 8 October 2014Accepted 9 October 2014Available online 17 October 2014

Keywords:PtAnthropogenicWestern North Pacific Ocean

Platinum is among the least abundant elements in the earth's crust and is now widely used in various productssuch as catalytic converters in automobiles and anticancer drugs. Consequently, the concentration of Pt in urbanaquatic environments is increasing. However, little is known about the distributions and geochemical cycles ofPt in the ocean owing to its low concentrations in seawater (b0.2 pmol/L). In this study, we report an improvedanalyticalmethod for determining sub-picomolar levels of Pt in seawater, and reveal the distributions of Pt in theJapan Sea, Sea of Okhotsk, and western North Pacific Ocean.For determining sub-picomolar levels of Pt in seawater, we used isotope-dilution ICP-MS (Inductively CoupledPlasmaMass Spectrometry) after columnpreconcentrationwith an anion exchange resin. Thismethod facilitatedhighly accurate analysis of Pt in seawater using small sample volumes (~1 L). The detection limit and proceduralblank value for this method were 0.015 and approximately 0.01 pmol/L, respectively. We obtained conservativevertical distributions, with nearly constant Pt concentrations between 0.19 and 0.25 pmol/L in the Japan Sea,Sea of Okhotsk, and western North Pacific Ocean. Judging from the constant dissolved Pt profiles, little anthropo-genic influence of Pt is apparent in the open oceans at the present time.

© 2014 Elsevier B.V. All rights reserved.

1. Introduction

Platinum (Pt) is a transitionmetal and because of its low reactivity isalso considered a noble metal. It is present in very low concentration inthe earth's crust (0.01 mg/kg; Wedepohl, 1995). Platinum has a highmelting point (1769 °C) and boiling point (4170 °C). Industrial demandfor Pt is increasing because of its use as an automobile catalyst. Catalyticconverters are installed in the exhaust system of cars in order to decom-pose nitrogen oxides and carbonmonoxide released from the engine aswell as hydrocarbons released by incomplete combustion of gasoline.Installation of catalytic converters in automobiles began in Japan in1973, in the USA in 1975, and in Europe in 1986. A reported increasein the accumulation of Pt in ice cores implies that anthropogenicallyreleased Pt has spread over the entire northern hemisphere (Barbanteet al., 2001). Moreover, Pt is also used in anticancer drugs, whichcould be related to the elevated Pt concentrations detected in aquaticenvironments (Kümmerer andHelmers, 1997). Consequently, the effectof anthropogenic Pt in marine environments should be investigatedfrom the start of the Pt contamination process.

The distribution and geochemical cycle of Pt in ocean water havebeen studied for a long time; however little is known because precise

determination of extremely low concentrations of Pt in seawater(10−15–10−12 mol/kg) is difficult. In addition, studies on Pt in seawaterhave shown that different oceanic regions show different types ofvertical profiles. For example, published work has demonstratedthe recycled type profile in the eastern North Pacific Ocean(0.46–1.17 pmol/L; Hodge et al., 1986); the scavenged type in theIndian Ocean (0.17–1.6 pmol/L; Van den Berg and Jacinto, 1988); andthe conservative type in the Atlantic and the western North PacificOceans (0.11–0.28 pmol/L; Colodner et al., 1993). These studiesemployed different analytical methods; therefore it is uncertain whichdata is the most reliable (Ravizza, 2001). Analytical artifacts shouldalso be considered. Specifically, Pt spikesmight not be fully equilibratedwith naturally occurring Pt in the sample utilized for isotope dilutionanalyses (Colodner et al., 1993), or the UV irradiation time for thesamples might affect the detected Pt concentrations for voltammetricanalyses (Obata et al., 2006). Moreover, the presence of an intensivematrix background or reagent-derived peaks might also interfere withvoltammetric analyses (Cobelo-Garcia et al., 2014).

In this study, we improved the analytical method for detectionof sub-picomolar levels of Pt in seawater using a column pre-concentration method with an anion exchange resin and isotope-dilution inductively coupled plasma mass spectrometry (ID-ICP-MS).Herein, we report the distributions of Pt in the western North PacificOcean and its marginal seas.

115A. Suzuki et al. / Marine Chemistry 166 (2014) 114–121

2. Methods

2.1. Samples

The sampling locations for this study are shown in Fig. 1. Thesampling stations covered areas of open ocean and marginal basins,CR-19 (North Pacific; 40°N, 144°E), CR-27 (western North Pacific;46°N, 159°E), CR-30 (Sea of Okhotsk; 45°N, 145°E), and CR-47 (JapanSea; 42°N, 138°E). The Japan Sea is a semi-closed marginal sea and ismore isolated than the Sea of Okhotsk. The water in the Japan Sea cancommunicate with the neighboring open ocean through four narrowstraits (Tsushima, Tsugaru, Soya, and Tatar) with sill depths of lessthan 130 m.

Seawater samples were collected in 12 L acid-precleaned X-Niskinbottles mounted on a SeaBird CTD-36 carousel array hung by atitanium-armored cable during a KH-10-2 R.V. Hakuho-maru cruise(June 21–July 14, 2010). The distance between the array and the sea-floor was monitored using a pinger (model 2216, Benthos). Immediate-ly after the recovery of theNiskin-X bottles, water sampleswere filteredthrough a capsulefilter (0.2 μmpore-size; Acropak; PALL Co.) in a closedspacefilledwith clean-air,whichwas passed through aHEPAfilter. Plat-inumcontamination from the capsulefilterwas evaluated bypassing 1 Lof Milli-Q water (MQW) through the filter, and the amount of Pt in theMQWwas determined. The concentration of Pt in the MQWwas belowthe detection limit (0.015 pmol/L). The filtered seawater samples wereacidified to 0.024 M with HCl, and stored until completion of the cruisefor analysis at the Atmosphere Ocean Research Institute, University ofTokyo. Auxiliary data (temperature and salinity) were obtained usingCTD sensors (Seabird, Model SBE-9-plus). In this study, we report dataof practical salinity, not absolute salinity. Dissolved oxygen data wereobtained with a DO sensor (SBE-43, Seabird).

2.2. Analytical methods

The Pt concentration in seawater was determined by the isotopedilution method according to previous studies (Colodner et al., 1993;Obata et al., 2006). The rationale for this method is based on Eq. (1) inwhich two isotopes (A and B) are present and the spike solution isenriched in isotope A. X is the number of moles of the sample, andSp is the number of moles of the added spike. Asp and An are the atomic

Fig. 1. Sampling locationsduringR.V.Hakuho-maruKH-10-2 cruise. StationsCR-19 and CR-27 arin the Japan Sea.

abundance (%) of isotope A in the spike and natural sample, respective-ly, and Bsp and Bn are the atomic abundance (%) of isotope B in the spikeand natural sample, respectively. R is the ratio of A/Bmeasured via ICP-MS. Assuming that Pt in the added spike solution and in natural samplesis completely equilibrated, R is obtained by using the following equa-tion:

R ¼ AB¼ AnX þ AspSp

BnX þ BspSpð1Þ

In this study, hydrochloric acid was added to seawater samples to be0.5 M HCl, and then the enriched spike solution was added to eachseawater sample. In seawater, spiked Pt is believed to form PtCl42−

or PtCl62− and is equilibrated with the Pt originally present in theseawater.

The number of moles of sample present in a solution spiked with anisotope-enriched tracer (X) in Eq. (1) is calculated using the followingequation:

X ¼ AspSp−RBspSpRBn−An

ð2Þ

We evaluated the suitability of the amount of spike added to theseawater sample by introducing an error multiplication factor (F), asdone in previous studies (Heumann, 1988; Ohata et al., 1998). Theprecision of the determinations ( dX

X

�� ��) is never better than that of theisotope ratio ( dR

R

�� ��) as shown in Eq. (3).

dXX

�������� ¼ F

dRR

��������: ð3Þ

The errormultiplication factor (F) can be calculated using Eq. (4) andEq. (5).

dXdR

¼ AnBsp−AspBn

RBn−Anð Þ2 � Sp ¼ X � AnBsp−AspBn

RBn−Anð Þ Asp−RBsp

� � ð4Þ

e located in thewesternNorth Pacific, CR-30 is in the southern Sea of Okhotsk, andCR-47 is

Fig. 2. Structure of the preconcentration column showing anion exchange resin packed ina Teflon tube with Teflon stoppers and quartz wool.

116 A. Suzuki et al. / Marine Chemistry 166 (2014) 114–121

F ¼dXXdRR

¼

An

Bn−

Asp

Bsp

!R

R−An

Bn

� �Asp

Bsp−R

! ð5Þ

Here, we differentiate F by the following equation;

dFdR

¼

An

Bn−

Asp

Bsp

!

R−An

Bn

� �Asp

Bsp−R

! 1− R

R−An

Bn

þ RAsp

Bsp−R

8>>><>>>:

9>>>=>>>;

ð6Þ

R is the value where dF/dR is zero, and the Pt concentration error isminimal.

The concentration of Pt in the oceanwater samples was expected tofall within the range 0.11–0.28 pmol/L as reported in previous studies.Therefore, appropriate amounts of the spike solution were added tothe seawater samples to give an F value of approximately 1. After themeasurements, we calculated the F value for each sample andconfirmed that the F value ranged between 1.172 and 1.212.

2.2.1. ReagentsAll experiments were carried out using ultrapure grade acids: ultra-

pure hydrochloric acid (HCl) (Tamapure-AA-100®, Tama Chemicals Co.Ltd.), ultrapure perchloric acid (HClO4) (Tamapure-AA-100®, TamaChemicals Co. Ltd.) and ultrapure nitric acid (HNO3) (Tamapure-AA-100®, Tama Chemicals Co. Ltd.). For the pre-concentration of Pt, weused the anion exchange resin, AG1-X8 (100-200 mesh, Cl-form;Bio-Rad Laboratories Inc.).

The original Pt standard solution was purchased from SPEX Indus-tries, Inc. and the stable isotope-enriched spike (Table 1) was obtainedfrom Oak Ridge National Laboratories (USA). The stock solution ofenriched isotope 192Ptwas prepared frompuremetal powders by disso-lution in aqua regia. Nitrous oxide was expelled by repeated evapora-tion with concentrated HCl (Yi and Masuda, 1996). Finally the salt wasdissolved in 1 M HCl, and the solution was diluted to a suitable volumewith the same acid. The Pt isotope concentration in the spike solutionwas calibrated by adding the certified Pt standard solution to the spikesolution and determining the isotope ratios.

2.2.2. ApparatusPolyethylene labware devices were washed with 5% Extran MA01

(Merck and Co. Inc.), 3 M HCl, and MQW. Teflon funnels, tubes, andvessels were used in the preconcentration system to prevent contami-nation. Teflon vessels were cleaned by heating in aqua regia (a 3:1mixture of 12 M HCl and 16 M HNO3) and in MQW. The Teflon columnand funnel were washed with 3 M HCl and cleaned by heating in amixture of HNO3, H2SO4, and HClO4 acids (1:1:1, v/v/v), 6 M HCl, andMQW.

The column extraction method was applied for the removal of seasalts and pre-concentration of the Pt contained in the ocean water sam-ples. The anion exchange resin was packed in a Teflon column withquartz wool on the surface of the resin within the column (Fig. 2). The

Table 1Platinum isotopic composition in enriched spike compared with natural abundance.

Platinum isotope Natural abundance (%) Spike abundance (%)

190 0.014192 0.782 56.97194 32.97 26.16195 33.83 11.23196 25.24 4.74198 7.163

resin was cleaned with 0.1 M HCl, 0.05 M NaOH, and MQW prior topacking.

2.2.3. ProceduresSeawater samples were acidified with 0.5 M HCl, and the 192Pt

enriched spike solution was added to each seawater sample. Themixture was left to stand for one day to reach isotopic equilibrium.The solutionwas transferred to a 6 cm-long Teflon column (inner diam-eter 8 mm) filled with anion exchange resin at a flow rate of 2 mL/minto concentrate Pt in the form of PtCl42− or PtCl62− (Fig. 3). Before intro-duction of the sample, the column was cleaned using 30 mL of amixed eluent (5 M HNO3 and 5 M HClO4), and conditioned with20 mL of 0.5 M HCl. Seawater samples (1 L) were passed through thecolumn for the Pt pre-concentration. After the samples were passedthrough the columns, they were rinsed with 6 mL of 0.05 M HCl and6 mL of MQW to remove the salts and loosely bound metal ions.The Pt adsorbed on the anion exchange resin was eluted with 25 mLof 5 M HNO3 and 5 M HClO4 in the direction opposite to that of thesample introduction. Eluents were collected in Teflon beakers andevaporated until their volumes were reduced to less than 0.1 mL. Theconcentrated samples were diluted with 1.5 mL of 5% HCl. The Ptconcentrations were determined by analyzing 192Pt and 195Pt with aQuadrupole Inductively Coupled Plasma Mass Spectrometer (ICP-MS;Agilent 7700) (Table 2). The preconcentrated Pt solution from onealiquot of each sample was analyzed 10 times by ICP-QMS. The errorswere calculated from the relative standard deviation obtained by thereplicated measurements and the error multiplication factor (F).

Fig. 3. Schematic drawing of the preconcentration system. A Teflon column (Fig. 1)connected to a Teflon funnel. Seawater samples were passed through the Teflon columnwith a peristaltic pump.

Table 2ICP-MS operating conditions.

Plasma conditionsICP RF power (W) 1500Cooling chamber temperature (°C) 2Argon gas flow rate

Plasma (L/min) 0.6Carrier (L/min) 0.7Auxiliary (L/min) 0

Sampling conditionSampling cone diameter (mm) 1.0Skimmer cone diameter (mm) 0.4Torch vertical position (mm) −0.5Nebulizer flow rate (μL/min) 100

Data acquisitionDwell time per point (ms) 100Data acquisition time (s) 25.2

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0 20 40 60 80 100 120 140

Isot

ope

ratio

Time(h)

R=192/195

Fig. 4. Variation of Pt isotope ratio (192Pt/195Pt) in the eluent after different equilibrationtimes. After adding the spike solution, the seawater samples were allowed to stand for1 h, 12 h, 24 h, and 133 h. The isotope ratios of the preconcentrated Pt solutions weremeasured as described in the “Methods” section.

117A. Suzuki et al. / Marine Chemistry 166 (2014) 114–121

Analysis of four samples was performed in triplicate to confirmthe precision of the analysis. The precisions for Pt concentration inthe CR-19 samples were 0.23 ± 0.026 pmol/L at 26 m, 0.21 ±0.029 pmol/L at 745 m, 0.22 ± 0.021 pmol/L at 5880 m and 0.23 ±0.021 pmol/L at 7507 m.

-5

0

5

10

15

20

25

Pot.

Tem

p.(

C) Sea of Okhotsk

Dicothermal layer

Japan Sea

NPDW

JSPW

3. Results

3.1. Optimal conditions

The analytical method using anion exchange and ID-ICP-MS wasoriginally developed by Colodner et al. (1993) and was subsequentlymodified by Obata et al. (2006). In this study, we reduced the columnvolume to minimize contamination from the anion exchange resin.The original method (Colodner et al., 1993) used a column volume of12 mL, however for this study the volume was reduced to 3 mL. Therecovery of Pt from seawater using a smaller column was examinedby changing various parameters such as the concentration of acid inthe sample, column length, flow rate, and elution temperature andvolume. The optimal conditions (Table 3) were identified when therecovery of Pt was the highest and nearly constant. Under these condi-tions, 75–90% of Pt was recovered from the samples.

Hf oxides such as 179Hf16O might cause potential interference withthe Pt isotopes. To evaluate the potential interference of Hf oxides, weadded an appropriate amount of standard Hf solution (so that theconcentration was 56 pmol/L) to natural and blank seawater. Notably,this concentration was much higher than in open oceans (0.06–1 pmol/kg; Bruland and Lohan, 2006). The aforementioned procedurewas also applied with the Hf-enriched samples. We did not detect anyinterference by Hf in the determination of sub-picomolar Pt concentra-tions in natural or blank seawater, in accord with previous reports(Colodner et al., 1993).

The isotopic equilibrium experiment was performed to allow suffi-cient time for isotopic equilibrium after addition of the spike solution.We found that one day was required to reach isotopic equilibrium(Fig. 4). To confirm isotopic equilibrium, we performed an additionalexperiment. After addition of the spike, three aliquots of the seawatersample were heated to 80 °C for 24 h and the three other aliquotsweremaintained at room temperature for 24 h. No differences between

Table 3Optimal conditions for preconcentration of Pt in seawater with anion exchange resincolumn.

Acidification of sample 1000 mL of seawater containing 0.5 M HClColumn length 6 cmSample flow rate 2 mL/minElution 25 mL of 5 M HClO4 + 5 M HNO3 (room temperature)

the amounts of Pt determined for the samples stored at 80 °C and roomtemperature were observed.

3.2. Hydrography

The potential temperature–salinity diagram at CR-47, located inthe eastern Japan Basin, is characterized by a nearly invariant potentialtemperature (~0 °C) and salinity (~34) below a depth of 500 m(Fig. 5). The dissolved oxygen concentrations are high (~4.7 mL/L)compared to other stations in the same depth range (Fig. 6). Thedeep water in the Japan Sea has unique properties and is called“the Japan Sea Proper Water (JSPW).” At station CR-30 in the Sea ofOkhotsk, the water temperatures are lower than those in the westernNorth Pacific Ocean (CR-27) throughout the water column, withthe exception of the surface water. The water temperature between40 m and 150 m, also known as the “dicothermal layer” (Kitani,1973), was below 0 °C. The dicothermal layer is characteristic of theSea of Okhotsk and the North Pacific Subarctic Gyre. In the Sea ofOkhotsk, this cold water forms off Sakhalin Island via winter mixing(Miura et al., 2002).

32 33 34 35Salinity

CR-19 CR-27 CR-30 CR-47

Fig. 5. Potential temperature versus salinity diagram for water samples from the fourstations; CR-19, CR-27, CR-30, and CR-47. The major water masses North Pacific DeepWater (NPDW) and Japan Sea Proper Water (JSPW), are also indicated.

0

2000

4000

6000

8000

-2 4 10

Dep

th(m

)

P. temp(C))

CR-19

32 34 36Salinity

0 5 10DO(ml/l)

0.0 0.1 0.2 0.3Pt(pmol/l)

0

2000

4000

6000

-2 4 10

Dep

th(m

)

CR-27

32 34 36 0 5 10 0.0 0.1 0.2 0.3

0

1000

2000

3000

-2 4 10

Dep

th(m

)

CR-30

32 34 36 0 5 10 0.0 0.1 0.2 0.3

0

1000

2000

3000

4000

-2 4 10

Dep

th(m

)

CR-47

32 34 36 0 5 10 0.0 0.1 0.2 0.3

Fig. 6.Vertical profiles of potential temperature, salinity, dissolved oxygen, and dissolved Pt at stations CR-19 andCR-27 inwesternNorth Pacific, CR-30 in the Sea of Okhotsk, and CR-47 inthe Japan Sea.

118 A. Suzuki et al. / Marine Chemistry 166 (2014) 114–121

3.3. Dissolved Pt concentrations

Our improved method for the analysis of Pt concentration wasapplied to seawater samples collected in 2010 at four stations (CR-19,CR-27, CR-30, and CR-47). We discovered Pt concentrations rangingfrom 0.19 to 0.24 pmol/L (CR-19), 0.19 to 0.25 pmol/L (CR-27), 0.20 to0.24 pmol/L (CR-30), and0.20 to 0.25 pmol/L (CR-47). The vertical distri-butions were of the conservative type (Fig. 6). There was no significantdifference in the Pt concentrations determined by Friedman test (theP valuewas 0.134 for a critical rate of 5%) between the stations examined.

Water collected at a depth of ≥300 m at CR-47 had the same char-acteristics as those of the JSPW and almost constant Pt concentrations

in the range of 0.20–0.25 pmol/L. The Pt concentrations below a depthof 1000 m at CR-19 and CR-27 in the northwest Pacific Ocean werealso nearly constant at 0.19–0.24 pmol/L. At CR-30 in the Sea ofOkhotsk, the Pt concentration ranged from 0.20 to 0.24 pmol/Lthroughout the water column. In the Sea of Okhotsk, to which theAmur River supplies terrestrial materials (Nakatsuka et al., 2004), noclear difference in the Pt concentration was observed in low-salinitywaters shallower than 1000 m. In the dicothermal layer, we couldnot find any distinguishable feature in the Pt concentrations. Prior toour study, the maximum ocean depth at which Pt values were investi-gated was 5888 m (Colodner, 1991). In this study, the maximum depthwas extended to 7507 m at station CR-19 in the Japan Trench, and no

119A. Suzuki et al. / Marine Chemistry 166 (2014) 114–121

significant difference was observed in the Pt concentrations at depthsgreater than ~6000 m.

4. Discussion

4.1. Analytical figures of merit

Themethod developed in this study allowed a facile andhighly accu-rate analysis of Pt concentrations in seawater using smaller samplevolumes (~1 L) and eluents (25 mL) than those used in previous proto-cols (~2 L and 30mL, respectively; Colodner et al., 1993). The procedur-al blank value was calculated by determining the Pt concentration inPt-free seawater (1 L). Three aliquots fromone oceanic seawater samplewere prepared; the firstwas passed through the acid-conditioned anionexchange resin columnonce, the second aliquotwas passed through thecolumn twice, and the third was passed through the column threetimes. The same procedure was applied for the Pt-removed seawaterthat was previously used for the standard seawater samples. Using theseawater that had been passed through the column two or threetimes, we obtained procedure blank values as low as 0.01 pmol/L(Table 4). The Pt content of the Pt-free seawater was comparable tothat of the eluent alone, which indicates that the Pt contaminationfrom an anion exchange resin column was successfully reduced. Basedon repeated analyses of Pt concentrations, the detection limit was calcu-lated to be 0.015 pmol/L. The detection limit and the blank value obtain-ed in this study are lower than those reported previously (Colodneret al., 1993), namely 0.07 and 0.05 pmol/kg, respectively, for ~2 L ofseawater. The low blank value obtained in this study was attributed tothe reduction in size of the anion exchange resin column.

4.2. Comparison of Pt with other trace metals in the western North Pacificand marginal seas

Somemetals such as Pt, Ag, and Cd are known to form chloride com-plexes in aquatic environments. In seawater, Ag exists as AgCl0, AgCl2−,and AgCl32−, Cd exists as CdCl20 (Turner et al., 1981), and Pt (Pt2+ orPt4+) forms stable chloride complexes (Reith et al., 2014). At a typicalseawater pH (~8.2), the dominant inorganic forms of Pt2+ and Pt4+

are PtCl42− and PtCl5OH2−, respectively (Cosden and Byrne, 2003).Recently, Cobelo-García et al. (2013) calculated Pt speciation in naturalwaters using thermodynamic data, and concluded that Pt4+

(PtCl5OH2−) predominates over Pt2+ (PtCl42−) in typical seawater(7.5–8.4). Therefore, it is important to compare the distributions of Pt,Ag, and Cd in seawater in different oceanic basins to understand thegeochemical cycle of these elements in aquatic environments.

Vertical profiles of Ag and Cd in the Japan Sea and the Sea of Okhotskwere previously reported (Zhang et al., 2001; Abe, 2002, 2005). The ver-tical profiles of Ag and Cdwere of the nutrient-type in the Japan Sea, Seaof Okhotsk, and western North Pacific Ocean, and were very similar tothose of Si and P (Zhang et al., 2001; Abe, 2002, 2005). JSPW has a rela-tively young water residence time (c.a. 300 years: Gamo and Horibe,1983). Ag and Cd concentrations in JSPW were lower than those inthe western North Pacific Ocean (Zhang et al., 2001; Abe, 2005),whose 14C age was estimated as ~1000 years (Matsumoto, 2007). Thedifference in Ag and Cd concentrations between the Japan Sea and thewestern North Pacific Ocean reflects the age of the water masses,

Table 4Analytical blank values during the Pt analyses. Open ocean seawater samples are passed throu

Sample

SeawaterSeawater passed through the anion exchange resin column onceSeawater passed through the anion exchange resin column twiceSeawater passed through the anion exchange resin column 3 timeseluent

similarly to the relative concentrations of Si and P in these waters(Zhang et al., 2001; Abe, 2005). Conversely, the Pt concentrationswere nearly constant in these two water masses.

Terrestrial substances are abundantly supplied into marginal seasvia rivers, the atmosphere, and continental shelf sediments; thesesupplies can elevate lithogenic trace metals in seawater. For example,the Japan Sea and Sea of Okhotsk have a higher concentration ofdissolved aluminum and indium in seawater than the open oceanbecause of the high supply of lithogenic substances (Obata et al.,2007). However, Pt concentrations were constant for these differentbasins, indicating that the supply of Pt from their associated terrestrialareas is smaller than the relative Pt content in seawater.

In semi-closed basins such as the Japan Sea, the activity of thorium-230, a decay product of uranium-234, is lower than that in the openocean because of enhanced bottom scavenging (Nozaki and Yamada,1987; Okubo et al., 2007). In contrast, the Pt concentrations wererelatively consistent throughout the different basins, implying that theparticle reactivity of Pt is too weak for its concentration to be alteredin the water column.

4.3. Conservative-type trace metals in the ocean

The vertical profiles of Pt in the Japan Sea, the Sea of Okhotsk, andthe western North Pacific Ocean were determined in this study to bethe conservative type. Vertical distributions of trace metals such asmolybdenum, tungsten, rhenium, cesium, and rubidium are known tobe constant throughout the water column. Cesium and rubidium existas relatively unreactive monovalent cations in seawater. Molybdenum,tungsten, and rhenium exist in seawater as oxyanions such as molyb-date (MoO2

2−, Sohrin et al., 1987), tungstate (WO22−, Sohrin et al.,

1987) and perrhenate (ReO4−, Anbar et al., 1992). Negatively charged

anions such as MoO22−, WO2

2−, and ReO4− have a relatively low particle

affinity at the slightly basic pH of seawater. These trace metals withconservative-type vertical profiles might be involved in biogeochemicalprocesses, such as adsorption onto biogenic particles and regenerationand desorption from biogenic particles. However, these processes arenegligible because of the low concentrations of these trace metals(Bruland and Lohan, 2006).

Among the trace metals capable of forming chloride complexes, Cdand Ag show nutrient-type vertical distributions in the ocean (Brulandand Lohan, 2006). Cd and Ag strongly interact with marine biota(Martin and Knauer, 1973; Fisher and Wente, 1993; Ratte, 1998; Laneand Morel, 2000; Ho et al., 2003), which induces low concentrationsof both elements in surface waters. Conversely, Pt exists as PtCl42− andPtCl5OH2− in seawater (Cosden and Byrne, 2003; Cobelo-García et al.,2013) as previously described, but there is little research concerningthe relationship between Pt and organisms (Ravindra et al., 2004;Mulholland and Turner, 2011; Shams et al., 2014). The conservative-type vertical profile implies that there are relatively few interactionsbetween Pt and biota. The negatively charged anions, PtCl42− andPtCl5OH2− (Cosden and Byrne, 2003; Cobelo-García et al., 2013),might also have a relatively low particle affinity in seawater. However,the kinetics of the interaction between each chemical species of Pt andparticles in seawater are not well known. Therefore, additional studiesare required to understand the biogeochemical cycling and speciationof Pt in the oceans.

gh the anion exchange resin column as described in the “Methods” section.

Content (pg) Concentration (pmol/L)

53 ± 5.0 0.270 ± 0.0266.3 ± 1.8 0.033 ± 0.0091.8 ± 0.9 0.009 ± 0.0041.9 ± 0.8 0.010 ± 0.0042.3 ± 0.5

120 A. Suzuki et al. / Marine Chemistry 166 (2014) 114–121

Judging from the conservative vertical profiles in this study and thework of Colodner et al. (1993), oceanic residence times of Pt might begreater than the ocean mixing time. However, different vertical profilesof Pt have been reported in different oceanic basins, namely therecycled-type in the eastern North Pacific Ocean (Hodge et al., 1986)and the scavenged-type in the Indian Ocean (Van den Berg andJacinto, 1988). In order to thoroughly understand the global cycles ofPt in the ocean, Pt distributions in various oceanic regions should beinvestigated.

4.4. Anthropogenic influence on Pt in the western North Pacific Ocean

Fig. 7 shows the vertical profiles of Pt at CR-19, CR-27 and thosereported previously (Colodner, 1991) in the Pacific Ocean (24°N,164°E). Similar vertical profiles were observed at all of these stations.A Friedman test was performed on the data for water samples obtainedat the three stations, and a P value of 0.368 was obtained when thecritical rate was 5%, indicating that there was no significant differencebetween the data. The obtained Pt concentrations determined in thisstudy were nearly identical to those reported more than 20 years ago(Colodner, 1991). From reports on recent increases in the accumulationof Pt in ice cores, it is evident that anthropogenically released Pt hasspread over the entire northern hemisphere (e.g., Barbante et al.,2001; Rauch et al., 2005; Moldovan et al., 2007; Soyol-Erden et al.,2011). The Pt released could also be deposited on the ocean surface,although the amount of deposited Pt that can be dissolved in seawaterremains to be elucidated.

Among the Pt group elements (PGE), anthropogenic emission of Osis known to cause global-scale atmospheric contamination (Chenet al., 2009). From the isotope ratio (187Os/188Os) and concentrationsof Os in precipitation samples from North America, Europe, Asia, andAntarctica, it was estimated that anthropogenic Os with a lower187Os/188Os ratio was released from the current production of Pt fromPGE ores (Chen et al., 2009). The 187Os/188Os ratios in some surface

0

1000

2000

3000

4000

5000

6000

7000

80000.0 0.2 0.4

Dep

th(m

)

Pt(pmol/L)

CR-19

CR-27

Colodner(1991)

Fig. 7. Comparison of vertical Pt profiles in the North Pacific. The profiles at CR-19 andCR-27 were obtained in this study. Another profile of Pt in the North Pacific was reportedby Colodner (1991).

seawaters were found to be lower than those in deep waters (Chenet al., 2009; Levasseur et al., 1999; Martin et al., 2001), which mightbe caused by the precipitation of anthropogenic Os. However, Os iseasily volatilized as OsO4 (boiling point, 135 °C) during PGE oreprocessing whereas Pt is not. The difference in the surface seawaterresponse between Os and Pt suggests that atmospherically-derivedanthropogenic Pt has little effect on dissolved Pt at the open oceansurface and/or is not dissolved via precipitation to the extent of Os.

Conversely, Pt was shown to have a remarkable anthropogenicimpact on land (Ravindra et al., 2004). Water from the Tama River, anurban river in the Tokyo district flowing into Tokyo Bay, was shown tocontain higher levels of Pt (1.3–4.7 pmol/L) than water from TokyoBay (0.3–1.5 pmol/L; Obata et al., 2006). The Pt concentration in theTama River was higher than that recently reported for the Lérez Riverin Spain (0.04–0.62 pmol/L; Cobelo-García et al., 2013), which is notlocated near any urban areas. This shows that anthropogenic Pt maybe emitted in highly populated urban areas. Platinum transported viaurban rivers will be supplied to coastal waters and ultimately to theopen ocean, which may take a long time.

5. Conclusions

We have developed an improved analytical method to determinesub-picomolar levels of Pt in seawater by modifying a pre-concentration method using anion exchange resin and ICP-MS mea-surements. This method allowed a facile and highly accurate analysisof Pt concentrations in seawater using smaller sample volumes (~1 L)and eluents than previous methods. The detection limit was0.015 pmol/L and the blank value was ~0.01 pmol/L.

In 2010, the concentrations of dissolved Pt in the northwesternPacific Ocean, Sea of Okhotsk, and Japan Sea were between 0.19 and0.25 pmol/L. These Pt levels are nearly identical to those reported20 years ago in the western North Pacific Ocean, suggesting littleanthropogenic influence on dissolved Pt in the open oceans.

Acknowledgments

Wewould like to thankCaptainU. Fujita, and the officers and crewofthe R.V. Hakuho-Maru for their kind help in sampling. We are alsograteful to Prof. J. Zhang, the chief scientist. This study was supportedin part by Grants-in-aid for Scientific Research (A) (Nos. 19253006and 23253001) from Monkasho (the Ministry of Education, Culture,Sports, Science and Technology: MEXT).

References

Abe, K., 2002. Preformed Cd and PO4 and the relationship between the two elements inthe northwestern Pacific and the Okhotsk Sea. Mar. Chem. 79, 27–36.

Abe, K., 2005. Concentration and regeneration of Cd in the Japan Sea proper water. J.Oceanogr. 61, 455–463.

Anbar, A.D., Creaser, R.A., Papanastassiou, D.A., Wasserburg, G.J., 1992. Rhenium isseawater-confirmation of generally conservative behavior. Geochim. Cosmochim.Acta 56, 4099–4103.

Barbante, C., Veysseyre, A., Ferrari, C., Van de Velde, K., Morel, C., Capodaglio, G., Cescon, P.,Scarponi, G., Boutron, C., 2001. Greenland snow evidence of large scale atmosphericcontamination for platinum, palladium, and rhodium. Environ. Sci. Technol. 35,835–839.

Bruland, K.W., Lohan, M.C., 2006. Controls of trace metals in seawater. Oceans Mar.Geochem. 23–47.

Chen, C., Sedwick, P.N., Sharma, M., 2009. Anthropogenic osmium in rain and snowreveals global-scale atmospheric contamination. Proc. Natl. Acad. Sci. U.S.A. 106,7724–7728.

Cobelo-García, A., Lopez-Sanchez, D.E., Almecija, C., Santos-Echeandia, J., 2013. Behaviorof platinum during estuarine mixing (Pontevedra Ria, NW Iberian Peninsula. Mar.Chem. 150, 11–18.

Cobelo-Garcia, A., Santos-Echeandia, J., Lopez-Sanchez, D.E., Almecija, C., Omanovic, D.,2014. Improving the voltammetric quantification of ill-defined peaks using secondderivative signal transformation: example of the determination of platinum inwater and sediments. Anal. Chem. 86, 2308–2313.

Colodner, D.C., 1991. The marine geochemistry of rhenium, iridium and platinum (PhDThesis) Massachusetts Institute of Technology.

121A. Suzuki et al. / Marine Chemistry 166 (2014) 114–121

Colodner, D.C., Boyle, E.A., Edmond, J.E., 1993. Determination of rhenium and plati-num in natural waters and sediments, and iridium in sediments by flow injectionisotope dilution inductively coupled plasma mass spectrometry. Anal. Chem. 65,1419–1425.

Cosden, J.M., Byrne, R.H., 2003. Comparative geochemistries of PdII and PtII: formation ofmixed hydroxychloro and chlorocarbonato-complexes in seawater. Geochim.Cosmochim. Acta 67, 1331–1338.

Fisher, N.S.,Wente, M., 1993. The release of trace elements by dyingmarine phytoplankton.Deep-Sea Res. 40, 671–694.

Gamo, T., Horibe, Y., 1983. Abyssal circulation in the Japan Sea. J. Oceanogr. Soc. Jpn 39,220–230.

Heumann, K.G., 1988. Isotope dilution mass spectrometry. In: Adams, F., Gijbels, R.,Van Grieken, R. (Eds.), Inorganic Mass Spectrometry. Wiley, New York, pp. 301–376.

Ho, T.Y., Ouigg, A., Finkel, Z.V., Milligan, A.J., Wyman, K., Falkowski, P.G., Morel, F.M.M.,2003. The elemental composition of some marine phytoplankton. J. Phycol. 39,1145–1159.

Hodge, V., Stallard, M., Koide, M., Goldberg, E.D., 1986. Determination of platinum andiridium in marine waters, sediments, and organisms. Anal. Chem. 58, 616–620.

Kitani, K., 1973. An oceanographic study of the Okhotsk Sea—particularly in regard to coldwaters. Bull. Far Seas Fish. Res. Lab. 9, 45–76.

Kümmerer, K., Helmers, E., 1997. Hospital effluents as a source for platinum in theenvironment. Sci. Total Environ. 193, 179–184.

Lane, T.W., Morel, F.M.M., 2000. A biological function for cadmium in marine diatoms.Proc. Natl. Acad. Sci. U. S. A. 97, 4627–4631.

Levasseur, S., Birck, J.L., Allegre, C.J., 1999. Osmium isotopic composition of Corals:evidences for multiple sources. Geochim. Cosmochim. Acta 63, 1335–1343.

Martin, J.H., Knauer, G.A., 1973. The elemental composition of plankton. Geochim.Cosmochim. Acta 37, 1639–1653.

Martin, C.E., Peucker-Ehrenbrink, B., Brunskill, G., Szymczak, R., 2001. Osmium isotopegeochemistry of a tropical estuary. Geochim. Cosmochim. Acta 65, 3193–3200.

Matsumoto, K., 2007. Radiocarbon-based circulation age of the world oceans. J. Geophys.Res. 112, C09004.

Miura, T., Suga, T., Hanawa, K., 2002. Winter mixed layer and formation of dichothermalwater in the Bering Sea. J. Oceanogr. 58, 815–823.

Moldovan, M., Veschambre, S., Amouroux, D., Benech, B., Donard, O.F.X., 2007. Platinum,palladium, and rhodium in fresh snow from the Aspe Valley (Pyrenees Mountains,France). Environ. Sci. Technol. 41, 66–73.

Mulholland, R., Turner, A., 2011. Accumulation of platinum group elements by themarinegastropod Littorina littorea. Environ. Pollut. 159, 977–982.

Nakatsuka, T., Toda, M., Kawamura, K., Wakatsuchi, M., 2004. Dissolved and particulateorganic carbon in the Sea of Okhotsk: transport from continental shelf to oceaninterior. J. Geophys. Res. 109, C09S14.

Nozaki, Y., Yamada, M., 1987. Thorium and protactinium isotope distributions inwaters ofthe Japan Sea. Deep-Sea Res. 34, 1417–1430.

Obata, H., Yoshida, T., Ogawa, H., 2006. Determination of picomolar levels of platinum inestuarine waters: a comparison of cathodic stripping voltammetry and isotope

dilution-inductively coupled plasma mass spectrometry. Anal. Chim. Acta. 580,32–38.

Obata, H., Alibo, D.S., Nozaki, Y., 2007. Dissolved aluminum, indium, and cerium in the Seaof Japan and the Sea of Okhotsk: comparison to the marginal seas of the westernNorth Pacific. J. Geophys. Res. Oceans 112, C12003.

Ohata, M., Ichinose, T., Furuta, N., Shinohara, A., Chiba, M., 1998. Isotope dilution analysisof Se in human blood serum by using high power nitrogen microwave-inducedplasma mass spectrometry coupled with a hydride generation technique. Anal.Chem. 70, 2726–2730.

Okubo, A., Obata, H., Gamo, T., Minami, H., Yamada, M., 2007. Scavenging of 230Th in theSulu Sea. Deep-Sea Res. II 54, 50–59.

Ratte, H.T., 1998. Bioaccumulation and toxicity of silver compounds: a review. Environ.Toxicol. Chem. 18, 89–108.

Rauch, S., Hemond, H.F., Barbante, C., Owari, M., Morrison, G.M., Peucker-Ehrenbrink, B.,Wass, U., 2005. Importance of automobile exhaust catalyst emissions for the deposi-tion of platinum, palladium, and rhodium in the Northern hemisphere. Environ. Sci.Technol. 39, 8156–8162.

Ravindra, K., Bencs, L., Van Grieken, R., 2004. Platinum group elements in the environ-ment and their health risk. Sci. Total Environ. 318, 1–43.

Ravizza, G.E., 2001. Platinum group elements and their isotopes in the ocean. Encycl.Ocean Sci. 2208–2217.

Reith, F., Campbell, S.G., Ball, A.S., Pring, A., Southam, G., 2014. Platinum in Earth surfaceenvironments. Earth-Sci. Rev. 131, 1–21.

Shams, L., Turner, A., Millward, G.E., Brown, M.T., 2014. Extra- and intra-cellular accumu-lation of platinum group elements by the marine microalga, Chlorella stigmatophora.Water Res. 50, 431–440.

Sohrin, Y., Issiki, K., Kuwamoto, T., Nakayama, E., 1987. Tungsten in North Pacific waters.Mar. Chem. 22, 95–103.

Soyol-Erden, T.O., Huh, Y., Hong, S., Hur, S.D., 2011. A 50-year record of platinum, iridium,and rhodium in Antarctic snow: volcanic and anthropogenic sources. Environ. Sci.Technol. 45, 5929–5935.

Turner, D.R., Whitfield, M., Dickson, A.G., 1981. The equilibrium speciation of dissolvedcomponents in freshwater and seawater at 25 °C and 1 atm pressure. Geochim.Cosmochim. Acta 45, 855–881.

Van den Berg, C.M.G., Jacinto, G.S., 1988. Different behavior of platinum in the Indian andPacific Oceans. Anal. Chim. Acta. 211, 129–139.

Wedepohl, K.H., 1995. The composition of continental crust. Geochim. Cosmochim. Acta59 (7), 1217–1232.

Yi, Y.V., Masuda, A., 1996. Simultaneous determination of ruthenium, palladium, iridium,and platinum at ultratrace levels by isotope dilution inductively coupled plasmamassspectrometry in geological samples. Anal. Chem. 68, 1444–1450.

Zhang, Y., Amakawa, H., Nozaki, Y., 2001. Oceanic profiles of dissolved silver: precisemeasurements in the basins of western North Pacific, Sea of Okhotsk, and the JapanSea. Mar. Chem. 75, 151–163.