Seagoing method for the determination of chromium(III) and total chromium in sea water by electron-capture detection gas chromatography

1

Analyhca Chunrca Acta, 2710993) l-9 Elsevler Science Pubhsbers B V , Amsterdam

Seagoing method for the determination of chromium(II1) and total chromium in sea water by electron-capture detection gas chromatography Robert K Mugo and Kristin J Orlans Departmentsof Chemistry and Oceanography, Unrversttyof BntBh Columbus, Vancouver, Bntuh Columbra MT II’6 (Canada)

(Received 7th May 1992)

Abstract A seagomg method for the determmatlon of Cr(III) and total chrommm m sea water IS presented The method employs electron-capture detection of the volatile tnfluoroacetylacetone denvatlve of 0011) formed via solvent extra&on unth toluene, total chrommm IS determmed as Cr(III) after reduction DetectIon hmlts at sea are 0 062 and 0 255 nM for C&II) and total chrommm, respectively Accuracy for total chrommm was verdied by the analysis of standard reference materials from the National Research Councd of Canada The procedure has a preclslon of 1 3% at 4 7 nM total chrommm, and has been apphed to stored samples m the laboratory m addltlon to Its use at sea Keywords Gas chromatography,

Chrommm,

Electron-capture

A variety of factors controls the drstrlbutlon of trace metals m sea water including the oxldatron state of the metal, which can mfluence mput and removal processes m the ocean. Chrommm exists m sea water m two different oxdatlon states, Cr(II1) and CrWI) The marme geochenustry of chrommm IS not very well understood, partly because the two chrommm oxldatlon states are characterlzed by dtierent chermcal behavlour and are difficult to analyze accurately wlthout mterconverslon durmg sample processmg Thermodynamic calculations predict that m oxygenated natural waters chrommm should crust almost exclusively as CrWI) with the predlcted species being chromate, CrOi-, and Cr(II1) exlstmg as the aquahydroxy species, Cr(OH),+(H,O), [ 11 However, the ratlo of Cr(II1) to CrWI) m natural waters has been found to vary from 0 02 to 0 99 [2] It was suggested that this variation and conCorrespondence to R IL Mugo, Departments of ChemlstIy and Oceanography, Umversity of Bntlsh Columbia, 2036 Mam Mall, Vancouver, Bntlsh Columbia V6T lY6 (Canada)

detectlon,

Sea watet, Solvent extractlon, Waters

tradlctlon with theory might be due to the m situ preclpltatlon of chromate only, with strontium or barmm sulphate [3] Cranston and Murray [4], however, have pomted out that speclatlon changes during sample handling, which vary depending on the technique used to detennme the chrommm species, might be responsible for these dlscrepanties The oxldatlon of Cr(II1) to CrWI) m natural waters IS known to be slow, wrth a reported half-hfe of several weeks [5] Early and Cannon [6] have attnbuted this to the kmetlc Inertness of aquated Cr(II1) species Various workers (cf Johnson and Xyla [7]) have Investigated the role of various manganese oxides as possible oxidants and have found that these phases are faster 0x1dants for Cr(II1) than 1s O2 Oxygen alone leads to oxldatlon half-lives of almost two years The increased rate observed m natural waters, therefore, suggests that manganese oxides, or possibly other mmeral oxides, are hkely to play an lmportant role m the environmental oxldatlon of G-011) Other factors which may influence the rate of

0003-2670/93/$06 00 0 1993 - Elsewer Science Pubhshers B V All rights reserved

2

oxldatlon of 0011) to Cr(V1) were investigated by Pettme et al [8], using H,O, m NaCl and the major sea salts at pH 8 and 25°C They also looked at the effect of aging the Cr(II1) solutions before oxldatlon An increase m the borate concentration was found to increase the oxldatlon rate while agmg decreased the rate A method that allows the determmatlon of both chrommm species as quickly as possible after sample collection, has great potential m allowing the factors controllmg the marme geochemistry of this element to be studied Currently most methods for the determmatlon of trace metals in sea water involve an initial sample collection step, followed by preservation or at times by some preliminary sample preparation step (e g preconcentratlon) on board ship The bulk of the sample-handling procedure plus the analysis, however, are still carried out m a shore-based laboratory sometimes long after the nntlal sample collection step As pointed out by Measures and Edmond [9], the ability to do shlpboard determlnations also offers the important advantage of contammatlon control which 1s cntlcal for accurate trace metal determmatlons In cases where the element of interest can exist m more than one omdatlon state, as does chrommm, and where oxldatlon state mterconverslon can occur relatively easily during handling or storage, the lmportance of doing determmatlons at sea cannot be overemphasized Electron-capture detection gas chromatography (GC-ECD) of metal chelates has been applied to the determination of metals m various matrices and has shown advantages over other methods m terms of sensltlvlty [lo] and a reduction m the sample preparation procedure [9] The compact size, general msensltlvlty to motion and vlbratlons on board ships, and the relatively low cost of the mstrument make it ideal for use at sea The electron-capture detector 1s highly sensetlve to fluorinated metal chelates provided they are volatile enough to be chromatographed The mltlal problems associated with the gas chromatography of metal chelates had mainly to do wth finding suitable hgands [ll] The hgand used m this study, l,l,l-tnfluoroacetylacetone (HTFA), has been employed m studies dealing with the

RK Mugo andK.J Onans/Anal

Chm Acta 271 (1993) 1-9

determination of chrommm m various matrices Chrommm was measured m blologlcal samples [lo,121 and m lunar samples [13] Its apphcatlon to natural waters include the determmatlon m sea water of beryllmm [9] and alummmm [14] Its use for the determmatlon of chrommm m natural waters has been confined to total chrommm m stored samples often with long reaction times and considerably higher sample volume requirements [15,16], none of the methods has been adapted for use at sea This study was therefore aimed at the development of an accurate and rapid GC shipboard technique for the determmatlon of Cr(II1) and total chrommm m sea water, which would allow the analysis of these two chrommm species durmg oceanographic cruises Such a techmque allows collection of data on the dlstrlbutlon of chrommm redox species m various oceanographic environments and thus allows for the elucldatlon of the factors responsible for the marme geochemistry of this element

EXPERIMENTAL Materials and reagents Chelatrng agent l,l,l-Trlfluoroacetylacetone

(HTFA) (Aldrich, Milwaukee, WI) was purified by dlstlllatlon at atmospheric pressure m a Perfluoroalkoxy (PFA) still as described by Measures and Edmond [9] Internal standard 2,6-Dlchloroblphenyl (Chemical Service, Westchester, PA) was used as obtained without further purlflcatlon Solvent Toluene (BDH, ACS or Ommsolve grade) was purified by dlstlllatlon 2-3 times through a 4 ft X 15 m glass still The distillate was monitored for extraneous peaks by mJectlon mto the gas chromatograph The redistilled toluene was spiked with the internal standard at a concentration of approximately 100 ng ml-’ Acetone (BDH, ACS grade) was used untreated for rinsing the PFA reaction bottles and the separatory funnel during the extractlon procedure Buffer Doubly quartz-distilled acetic acid (Seastar Chemicals, Sidney, BC) diluted with

R K Mugo and RJ Onans/AnaI

Chtm. Acta 271 (1993) 1-9

delomzed water to give approxunately 10% CH,COOH and 1 M NaAc-HAc were used for pH adjustments A solution of 1 M NaAc-HAc was prepared from analytlcal-reagent grade sodnun acetate (BDH), which had been recrystalhzed once to remove trace amounts of chrommm and other metal contammants present An alternate cleanmg procedure for the sodmm acetate involved treating the impure sodium acetate buffer solution m the same way as the sea-water samples and usmg the redrstdled HTFA lrgand to scavenge any chrommrn present Both procedures were equally effective m producmg NaAc which was clean enough for use m this study The recrystalhzatlon procedure was chosen for all subsequent work as it was less tune consummg Reducmg agent Sodmrn sulphlte (BDH, ACS grade) solution (1 MI, used for the reduction of (XVI) to (XIII), was cleaned via extraction with the redlstdled HTFA Thrs was done by adJustmg the pH of the solution to approxunately 6 with doubly d&filed acetic acid followed by solvent extractron vvlth toluene m a similar manner to the samples (described below) The solution was rmsed several times vvlth the redlstdled toluene to ensure that all the HTFA had been removed Wafer Delomzed water was obtamed from a Bamstead Nanopure Series 630 delomzatlon system Chromzum stczndurds Certified atormc absorptlon standards for Cr(III) and Cr(VI) (Aldrich, Milwaukee, WI> were used to prepare appropnate extractlon standards Solutions of 1000 pg ml-’ Cr(III) m 1 wt % HCI and 1000 pg ml-’ Cr(VI), as an ammomum dlchromate solution m water, were diluted with delomzed water to approprrate concentration ranges for use m this study The optumzatlon procedures for both the gas chromatographlc and the solvent extractlon steps were monitored on the basis of chrommm recoveries obtained by the analysis of chrommm trlfluoroacetylacetonate [Cr(TFA),] synthesized as described by Fay and Prper [17] Apparatus A Hewlett-Packard 5890 Series II gas chromatograph equipped with a 555 0 MBq 63N1electron-capture detector was used m this study The

3

GC-ECD apparatus was run m the spht mode vvlth a split ratio of approxunately 10 1 A J&W Sclentlfic DB 210 15 m X 0 25 mm o d capdlary cohunn Hnth a 05-pm film thickness was used The carrier gas, ultra high purity (UHP)_grade nitrogen, was purified further by passmg It through a molecular-sieve trap and a hydrocarbon trap The detector makeup gas, UHP-grade nitrogen, was purified via a heated tamer gas trap and an mdlcatmg oxygen trap Data handlmg from the GC runs was performed on an on-hne Hewlett-Packard personal computer equrpped with HP-Chemstatlon 3365 software pH measurements were performed usmg an Orion SA 520 pH meter equipped wrth a 91-02 general-purpose combination electrode Shakmg was accomphshed with a Burrell Model 75 wnst action shaker A Samsung MW257OUC home mlcrowave oven was employed to speed up the chromrum extractions A make-shift PFA separatory funnel for the extractlons was constructed as described by Measures and Edmond for alummium [ 141 Sea-water samples Chrommm determmatlons m the laboratory were conducted on two kmds of stored sea-water samples (1) samples that had been acldlfied after collection and b) sea-water samples that had been frozen unmedlately after collection wthout acldlflcatlon Sea-water samples from the central North Atlantic, near Bermuda, were collected usmg 5-l N&m bottles (General Oceamcs) mounted on a stamless steel hydrowu-e, by E A Boyle and colleagues from the Massachusetts Instltute of Technology The samples were filtered then acidtied to pH 2 f 0 1 and stored m aad-leached polyethylene bottles Under these condltrons all the chrommm 1s reduced to Cr(III) Hrlthm 24 h [18], and only total chrommm can be determmed after adjusting the sample pH to the extractlon pH (6 0 f 0 2) with the NaAc-HAc buffer Sea-water samples that had been frozen unmedlately after collectlon were collected m the Northeast Paclfx Ocean (approxtmately 30 km off Nootka Sound, 49”N, 127”W) usmg 30-I Go-F10 bottles (General Oceamcs) mounted on a Kevlar @

4

RK Mugo and RJ O~ns/Anal

lme The samples were filtered then stored m acid-cleaned polyethylene bottles and frozen lmmedtately on board ship Analyses of C1-011) and total chromium were subsequently performed on these samples m the laboratory Chrommm determmatlons at sea durmg an oceanographic cruise were performed on seawater samples collected using either lo- or 30-l Go-F10 bottles mounted on a Kevlar@ line Processtng

Sample processmg m the laboratory took place m a filtered-air environment wthm a lammar flow bench m an effort to reduce possible contammatlon from the surroundmgs The extractions were carried out m PFA PFA bottles Before their first use, these 60-ml bottles and the PFA separatory funnel were leached m 4 M hydrochlorlc acid at 60°C for three days followed by a dilute doubly dlstllled HNO, (approximately 1%) leach for about a week During usual laboratory runs the PFA reaction bottles and separatory funnel were cleaned between extractions by rmsmg three tunes with approxunately 5-10 ml acetone A portable high-efficiency particle an- (HEPA) filter cabmet was used at sea to reduce possible contammatlon from the ship environment General procedure Chromzum(ZZZ)Sea-water samples (15 ml) for Cr(II1) determmatlon are measured accurately into the reaction bottles using an adjustable lo-ml Eppendorf plpet The pH is adjusted from the natural sea water pH of 7 5-8 3 to the extraction pH of 6 0 f 0 2 with 20 ~1 10% quartz-d&filed CH,COOH, 100 ~1 of the purified hgand are then added, followed by the addition of 1 ml toluene The bottles are shaken manually for 5 s to ensure rmxlng of the various reagents, then placed m the nucrowave and heated four at a time for 3 mm To reduce pressure build-up mslde the bottles durmg microwave heatmg, bottles are pamally deflated by squeezing the walls prior to capping, allowmg room for expansion The bottles are removed from the nucrowave and shaken for 5 s and then returned for another 3

Chun Acta 271(1993) l-9

mm at the same power level At the end of this period the sample temperatures are 65-70°C During coohng, samples are shaken for 10 mm on the mechanical wst action shaker, fully cooled to room temperature, then carefully transferred to the PFA make-shift separatory funnel where the aqueous layer 1s separated and discarded The organic layer 1s shaken for 10 s with 1 ml of delomzed water to help preventing the formation of emulsions from calcium and magnesium hydroxldes which can occur when NaOH 1s added m the next step The layers are allowed to separate and the organic layer 1s then shaken for 20 s with 1 ml of 1 M NaOH This washmg step wrth base 1s critical as it destroys the excess hgand which would oversaturate the detector if not removed After separation the organic layer 1s rinsed two tunes wrth a total of 2 ml of delomzed water to remove traces of NaOH The extract 1s then transferred to a clean glass vial with a PFA-lined cap (once chelate formation 1s complete and the excess hgand has been removed, the organic extracts are virtually immune from any chromium contammatlon, contact with glassware 1s therefore not a problem at this stage) The extracted sample 1s ready for mjectlon mto the gas chromatograph at this stage or it can be stored at room condltlons for l-2 days For long-term storage, the extracts are stored m the freezer at - 15°C and are stable for several weeks Cr(II1) standards are treated m the same manner Total chromzum The reaction between the hgand, 1,&l-tnfluoroacetylacetone, and chrommm 1s specific for Cr(III), for total chrommm determination it 1s necessary to convert CrWI) to the reduced state before chelation can occur For samples stored frozen and unacldlfled, 20 ~1 of

TABLE I Gas chromatographlc condltlons for the analysis of chrommm as Cr(TFA), InJectIon port temperature Oven temperature Detector temperature Column head pressure Hydrogen carrier gas flow-rate Nitrogen makeup gas flow-rate

200°C 130°C 350°C 15 PSI 26mlmm-’ 49 ml mm-’

RR Mugo and RI Onans/AnaI

5

Uum Acta 271 (1993) I-9

the 10% CH,COOH and 200 ~1 of the 1 M sodmm sulphlte reducmg agent are added to 15 ml of thawed sample In samples which have been stored acldlfied at pH 2, CrWI) 1s reduced within 24 h and thus no reducing agent 1s added These samples are brought to the extraction pH by the addition of 2 ml (for each 15 ml of sample) of the 1 M NaAc-HAc buffer In both cases the samples are then treated m a snmlar manner as described for the Cr(II1) samples above 100

l.

.

l

’ b)

RESULTS AND DISCUSSION

Optunuatwn Gas chromatography The GC condltlons were selected to ensure the highest possible sensltlvlty of the electron-capture detector for the Cr(TFA), chelate, as well as good resolution and reasonable retention times for the peaks of interest Table 1 shows the optlmlzed parameters used for all quantitative work A typical chromatogram 1s shown m Fig 1 Chrommm elutes as two peaks for the tram and CIS Isomers of the WTFA), chelate Quantltatlve determmatlons were per-

2-I 0

I

1

3

Time (mln)

Fig 1 Typical chromatogram of a chronuum determmatlon m sea water Cr(TFA)3 elutes as two well resolved peaks wrth retentron times of 2 681 and 3 351 mm The peak wth the shorter retention time IS asslgned to the trans Isomer (less polar and therefore less mteractlon with the polar stationary phase) The mtemal standard, 2,6-dlchloroblphenyl, has a retention time of 2042 mm, the peak wtth retention time 1572 mm corresponds to the Al(TFA), chelate formed from the reactlon of the element m sea water wrath the hgand (Identified by comparison with the results of analyst of the pure chelate), other peaks are not yet ldentlfied W condotlons as In text

Fig 2 Total chromium recovery as a function of (a) pH and (b) hgand volume Samples were 15 ml of sea water, spiked wth chrommm and allowed to eqmhbrate for several days Each pomt IS the mean of two rephcate analyses All other condltlons are as m the general procedure

formed by summmg up the areas of the two peaks Extractron condztwns The chromnnn complexatlon and solvent extraction procedure was optlmlzed with respect to the pH, the hgand concentratlon, and the temperature/ reaction tnne The highest recovery of chromium was achieved at pH range 5 6-6 5 (Fig 2a) In general use the samples were adjusted to pH 6 0 f 0 2 The extraction of chromnnn 1s dependent on the amount of hgand added At constant pH, reaction tnne and temperature, the extraction efficiency increased Hrlth hgand volume m the range 20-80 ~1 but was independent of hgand volume at higher levels (Fig 2b) Hamm et al [19] observed that m solutions contammg the hexaaqua Cr(II1) ion and organic acid anions, the reaction rate and the pH effect were mdependent of the nature of the amon present or its concentration so long as an excess was mamtamed, due to the slow chrommm complexatlon kmetlcs A volume of 100 ~1 of the pure hgand was used for all extractrons m this study, and this was suffl-

6

R K Mugo and ILl Orms /Anal Chun Acta 271(1993) I-9 100

.

l

(4

80

.

P p 8 60

3

’

.

.

; 20..

.

07 0

loo-

ShaLng time Zroom te%lp (hrm)’ l

.

l

l

.

.

b) 30.-

E

Eso-0 z

action mrxture A microwave oven was chosen for this purpose and resulted m excellent recoveries of chrommm being obtained (100 f 4%) m a short time Moreover, the reaction tnne and the power level could be controlled precisely thus allowmg for good reproduclblhty As can be seen m Fig 3b, chrommm recovery was quantltatrve after the microwave procedure followed by 10 mm or less of shakmg the sample on the wrist action shaker

40.20.07 0

Fig 3 Total chronuum recovery as a fun&on of (a) shakmg tune at room temperature and (b) shakmg time after rmcrowave Irradiation Samples were 15 ml of sea water, spiked with chromium and allowed to eqmhbrate for several days Each point 1s the mean of two replrcate analyses All other condltlons are as m the text

clent for recovery of all the chrommm m a U-ml sample plus any added spikes during the optlmlzatlon procedure The complexatlon kmetlcs of (XII) with HTFA are very slow at room conditions Initial attempts to obtam quantltatlve extractron of chrommm by reactlon with the hgand at room condltlons showed that at least 3-4 h shakmg time were required Figure 3a shows the percentage extraction efficiency of chromnun as a function of shakmg tnne at room temperature Lovett and Lee [15] have reported a room temperature shakmg time of at least 2 h for chrommm concentrations below 10 pg ml-’ with the hgand m great excess (0 164 M m benzene) m order to obtain maximum chrommm extraction Measures [20] estimates 87% chrommm recovery using 15 ml sea-water samples after shakmg at room temperature for 2 h using 100 ,ul of HTFA In order to reduce the reaction time needed for quantitative extraction of chrommm, It was necessary to increase the temperature of the re-

Analytical figures of ment The preclslon of the technique

was evaluated by performing replicate total chrommm analyses on a stored acidified sea-water sample Six rephcate analyses gave a relative standard devlatlon of 13%at47nM Absolute recovery studies of both CrGII) and Cr(V1) spikes added to sea water were performed by splkmg known amounts of the two chrommm species into sea-water samples and then subjectmg the samples to the full extraction procedure after allowing eqmhbratlon for 2-3 days Two replicates for each sample were analyzed and as shown m Table 2, excellent recoveries of both species were achieved The excellent recovery of both chrommm species shows that reduction of Cr(VI) to Cr(II1) and the extractlon procedure are both quantltatlve WIthout reduction, however, Cr(VI)-spiked sea-water samples (stored frozen) show no detectable increase m slgnal This verifies that

TABLE 2 Recovery of chrommm spikes from sea-water samples a Imtial total chromium (nM)

Sp&e chrommm added (nM)

Chrommm species added

Total chrommm recovered b (nM)

Recwery (%I

470*008 470*008 470*008 470*008

215 431 385 769

Cr(III) Cr(III)

662*019 899*004 862&021 123*003

97 100 101 99

Cr(VI) CrWI)

’ Samples were U-ml ahquots of sea water collected at 49”N 127”W m the North PacAc Ocean (see text for detads) b Mean of two replicates

R L Mugo and LJ Omm/Anal

Chun Acta 271 (1993) I-9

chrommm reduction durmg sample handling 1s insIgnificant The accuracy of the techmque was assessed by the analysis of chrommm m trace metal sea-water reference standards from the National Research Council of Canada Two reference sea-water samples, the Nearshore Seawater Reference Material (CASS-2) and the Open Ocean Seawater Reference Material (NASS3) were analyzed for total dissolved chronuum usmg the technique developed m this study The results are shown m Table 3 The good agreement between the certified values for the reference materials and the values obtamed by this technique to the samples provide proof for the accuracy of the method The hmrt of detection was estunated by the replicate analysis at sea for Cr(II1) and total dissolved chrommrn m delonlzed water (n = 6 m each case) using the procedure developed here DetectIon limits (3s) for Cr(II1) and total chrommm of 0 062 and 0 255 nM, respectively, were obtamed Determmatlons at sea Shipboard determmatlon of Cr(II1) and total chrommm m the North Pacific were performed on board the C S S “Endeavour” at two stations, P20 (Station “Papa”, 50”OO’N, 145”OO’W) and P26 (49”34’N, 138”4O’W) m October 1991 Cr(II1) and total chrommm analysis was completed for each statlon within 8 h after collection Figure 4a and b shows the chrommm concentration versus depth profiles at the two stations Analysis of filtered and unfiltered sea-water samples for chrommm was carried out Samples were filtered using 0 4-pm Nucleopore polycarbonate membrane filters, m an acid-cleaned fil-

TABLE 3 Accuracy of chrommm determmatlon

on sea-water samples

Sea-water sample

Chrommm (nM) Found a

Certified

CASS-2 NASS3

2331kOO68 3333*0011

2327i.0308 3365+0192

a Mean of 7 rephcates

Chromium

(nM)

Silicate

(pM)

Chromium

(nM)

Chromwm

(nM)

:

. . ”

Fig 4 Concentratrons versus depth profiles (m) Cr(III), (0) Cr(VI) and (A) total chrommm concentrations at (a) Station P26 (“PAPA”, 50”00’N, 145VO’WJ and (b) Station P2O (49”34’N, 138YO’W) Determmatlons were performed on board the C S S “Endeavour” (c) (v) Dissolved skate, given for comparmon, at station P26 (5O”OO’N, 145VO’W, same as “PAPA” [21]) and ( v) m the central north Atlantic Ocean (26”20’N, 33”4O’W) (d) (A) Total chrommm 111the central north Atlantic Ocean (26“20’N, 33”4O’W), from stored acldlfied samples

tratlon unit hooked up to a vacuum pump Fdtratlon had no effect on the chrommm content of the samples mdlcatmg that either particulate chrommm concentration m the samples was negligible, undetectable by this technique, or that the particles were so small that they passed through the filters Chromium depth dutnbutwn Shipboard determmatlon of chrommm at the two stations revealed that the Cr(II1) concentration at both statlons was very low and was generally at or below the detection lnmt of 0 06 nM Total chrommm, on the other hand, was conslderably higher at 2 3-4 3 nM These results agree with the predlctlon from theoretical calculations,

8

Cr(V1) was the dominant chrommm species at these statlons (2 95% of total chromtum) The profiles show a slight surface depletion with a modest increase at depth, suggesting that chrommm may be a nutrient--type (blo-mtermediate) element It has been suggested [5,18] that the depth dlstrlbutlon of total chromrum m both the Atlantic and Pacific Ocean 1s similar to that of s&ate m these ocean basms, and that chromnun, hke sdlcate, might be mvolved m a deep regeneration cycle Figure 4c shows the s&ate dlstributlon at Station “Papa” [21] obtained m 1987 Whereas slhcate 1s nearly depleted m the upper waters, total chrommm still has a significant surface concentration Moreover, the shght increase m the concentration of chrommm with depth does not m any way match the increase m silicate concentrations with depth Analysis of stored sea-water samples from the central North Atlantic for total chrommm (Fig 4d) has not produced profiles similar to those of silica (Frg 4c) either The claim that chrommm has a sea-water dlstrlbutlon slmllar to that of silica remains largely unsubstantiated The explanation for the increase m total chrommm with depth 1s still unclear, as the mechanism of uptake of chromate (CrOi-> by opal, CaCO, or orgamc matter has not been well established Mayer and Schick [22] m then study of the removal of Cr(VI) from estuarme waters by model and natural substrates reported a removal dependence on sediment concentrations and sahnlty They suggested a possible reductive adsorption mechamsm Naturally occurrmg levels of phosphate and slhcate showed negligible effects on chromate removal The solublhty control of Cr(II1) m natural waters 1s not very well understood Three candidates for this role have been suggested, and include Cr(OH), (s), chromate (FeCr,O,) and the mixed hydroxide (Cr,Fe,_,) (OH), (s) [5,8] In addltlon Cr(II1) 1s reported to bmd strongly to particles and organic material 123,241, which may be responsible for the low dissolved concentrations observed The role of orgamc hgands m Cr(II1) speclatlon and the forrnatlon of polynuclear species are not well understood

RL Mugo and RJ Onans /Anal

Chun Acta 2710993) 1-9

The frozen samples collected off Nootka Sound yielded values for Cr(II1) which were higher (0 71-117 nM) than those found m open-ocean samples analyzed at sea, but still m agreement wrth other reported values [4,5] This may be due to artifacts resultmg from freezmg and storage or to a coastal versus open-ocean effect Concluswn The method described m this paper not only allows for the rapld and accurate determmatlon of Cr(II1) and total chrommm m sea water m the laboratory but, more Importantly, allows thex determmatlon at sea Prehmmary results obtamed at sea mdlcate Cr(III)/Cr(VII) ratios between 0 02 and 0 05, which are at the low end of the range reported m the literature (0 02-O 99 [2]) Further mvestlgatlons are clearly needed to provide a complete understandmg of the geochemical cycle of this element In partxular, the use of natural laboratones such as anoxlc and seasonally anoxlc basins to monitor Cr(II1) * Cr(V1) mterconverslon as a result of changes m oxygen concentrations, coupled with kmetlc and speclatlon studies m the laboratory will be Important Additional work on the concentration and dlstrlbutlon of the two chrommm species m hydrothermal solutions and buoyant plumes is also planned It 1s hoped that these studies wrll help us elucrdate the factors responsible for the marme geochemistry of chrommm The authors are grateful to EA Boyle and colleagues at Massachusetts Institute of Technology (MIT) for the central North Atlantic samples, and to R D Bellegay and the Instttute of Ocean Sciences (IOS), Sidney, BC, for provldmg shlptime on the C S S “Endeavour” We also thank C I Measures (Umveraty of Hawau) for helpful dlscusslons and suggestions throughout the course of this project

REFERENCES 1 H Elderfield, Earth Planet SCI Lett , 9 (1970) 10 2 TM Florence and GE Batley, CRC Crlt Rev Anal Chem , 9 (1980) 219

RK Mugo and RJ Onans/Anal

Chun Acta 271 (1993) l-9

3 G Arrhemus and E Bonattl, Prog Oceanogr, 3 (1965) 7 4 R E Cranston and J W Murray, Anal Chum Acta, 99 (1978) 275 5 J W Murray, B Spell and B Paul, m C S Wong, E A Boyle, KW Bruland, JD Burton and ED Goldberg (Eds), Trace Metals m Seawater, Plenum Press, New York, 1983, pp 643-669 6 J E Earley and R D Cannon, Trans Metal Chem , 1 (1965) 34 7 CA Johnson and A G Xyla, Geochlm Cosmochlm Acta, 55 (1991) 2861 8 M Pettme, F J Mdlero and T La Note, Mar Chem , 34 (1991) 29 9 C I Measures and J M Edmond, Anal Chem , 58 (1986) 2065 10 J Savory, P Mushak, F W Sundermann, EM Esters and N 0 Rozel, Anal Chem , 42 (1970) 294 11 R W Moshler and R E Bevers, Gas Chromatography of Metal Chelates, Pergamon, Oxford, 1965 12 G H Booth, Jr and W J Darby, Anal Chem , 43 (1971) 831 13 T L Isenhour, NM Frew and J J Leary, Anal Chem ,44 (1972) 665

9 14 C I Measures and J M Edmond, Anal Chem, 61 (1989) 544 15 R J Lovett and G F Lee, Environ Sa Technol , 10 (1976) 69 16 K.W M Sm, ME Bednas and S S Berman, Anal Chem, 55 (1983) 473 17 R C Fay and T S Piper, J Am Chem Sot ,85 (1963) 500 18 J A Campbell and P A Yeats, Earth Planet Scl L&t, 53 (1981) 427 19 R E Hamm, R L Johnson, R H Perkms and R E Davis, J Am Chem See , 80 (1959) 4469 20 C I Measures, personal commumcatlons, 1991 21 JH Martin, RM Gordon, S Fltzwater and W W Broenkow, Deep-Sea Research, 36 (1989) 649 22 L M Mayer and L L Schmck, Environ Scl Technol , 15 (1981) 1482 23 S Osalu, T OS&, M Setoyama and Y Takashlma, J Chromatogr ,257 (1983) 180 24 G S Douglas, G L Mills and J G Quinn, Mar Chem , 19 (1986) 161