Two-step one-pot speciation of chromium as Cr-APDC-and-NaEDTA complexes followed by ETA-AAS analysis

A two-step preconcentration method involving solid-and-liquid-phase extraction procedures has been proposed for the speciation of chromium in a one-pot system. The method used sodium ethylenediaminetetraacetic acid (NaEDTA) and ammonium pyrro-lidinedithiocarbamate (APDC). The Cr(VI)-APDC was extracted with ethyl acetate and digested with nitric acid. while the Cr(III)-EDTA complex ion was adsorbed onto Al 2 O 3 (neutral grade) with the aid of a tetrabutylammonium bromide (TBAB) pairing agent and desorbed with a hydrochloric acid solution. Sample splitting was not required. The ZEEnit 650P, electrothermal atomization atomic absorption spectrometer, (ETA-AAS) was employed for the analysis. The concentration of the chromium (III) in the samples ranged from 1.83 ± 0.00 (cid:181) g L − 1 to 106.28 ± 0.21 (cid:181) g L − 1 , while < LOD to 19.85 ± 0.12 (cid:181) g L − 1 was the range of Cr(VI) concentration in the samples. Recovery from spiked samples was between 83% and 117% Cr(III) and 79% and 99% Cr(VI). Precision (n = 6), was 1.87% for Cr(III) and 7.67% for Cr(VI). The limits of detection (LODs), calculated by the Aspect LS software (3 σ , n = 11) were 0.42 (cid:181) g L − 1 and 0.62 (cid:181) g L − 1 for Cr(III) and Cr(VI) respectively. The mean total chromium (n = 3), from the certified reference material (CRM), LGC6019 was 0.74 ± 0.61 (cid:181) g L − 1 , (certified value = 0.78 ± 0.20 (cid:181) g L − 1 ). The preconcentration factor: ratio of the Cr(III) content of CRM before and after preconcentration was 3.57. Parameters such as pH, time, temperature, and the amount of adsorbent were optimized by the isolation method. The method was validated and applied for the speciation analysis of chromium in water samples.


INTRODUCTION
The emphasis on chemical speciation analysis of metals rather than their total concentration in samples emerged from the knowledge that biological systems respond differently to chem-ical forms of metal [1][2][3].Thus, the physiological response of biological organisms to tri-and-hexavalent chromium represents the beneficial and the harmful roles associated with the element respectively [4,5].
The trivalent Cr being essential to humans plays a vital role in the metabolism of glucose and lipids, [2,[5][6][7], and the formation of deoxyribonucleic acid (DNA) in mammals [1,5].Con-trariwise, hexavalent chromium is a class 1 human carcinogen [8][9][10].Instances of contact with Cr(VI) through inhaling the dust, dermal contact, and ingestion of Cr(VI) contaminated substances have been linked to nasal septum, asthma, and inflammation of the larynx and liver.Humans and experimental animals have developed dermatitis, and skin ulceration, and shown mutagenic and genotoxic effects on exposure to hexavalent chromium [8,9,[11][12][13].
Due to its numerous industrial applications in wood preservation, leather tanning, electroplating, and steel alloy production, exposure to hexavalent Cr is unavoidable [14][15][16][17].When discharged into the environment, the atmosphere and aquatic systems serve as means of long-range chromium transport and distribution [9,11].
From the literature, ion pairing agents such as tetraethylammonium (TEA) and tetrabutylammonium, (TBA) bromides, chlorides and hydroxides have been employed in conjunction with EDTA derivatization for the speciation analysis of chromium by chromatography [27,28].Other authors have used the cloud point extraction (CPE), based on isopropyl 2-[(No Suggestions Available)disulfonyl] ethane-aided speciation of Cr in human blood [29].Shirkhanloo et al. [4] used Triton-X45 and graphene in a CPE for the speciation of Cr in water samples by the ETA-AAS technique and the concentration of each Cr specie was based on difference wit the total concentration.Chromium in inhaled human breath condensate was analysed by micro liquid chromatography coupled to inductively coupled plasma mass spectroscopy (µLC-ICP-MS) hyphenated system [2].
The literature indicated that using a combination of liquid and solid phase preconcentration techniques in one-pot speciation analysis of the element is rare.But, Honma [19] described a onepot two-step procedure which used APDC and diisobutyl ketone (DIBK).It relied on the rate of reaction of the Cr species with the ligand, (which is very slow with Cr(III)) at room temperature.The Cr(VI)-PDC was taken into the DIBK within a few minutes of the reaction, and the Cr(III)-PDC was extracted with a fresh portion of DIBK after a long time.
From other reports, EDTA has been employed in conjunction with APDC [30,31] and diethyldithiocarbamate (DDTC), for the speciation analysis of chromium where the EDTA functioned as a masking agent only.The concentration of Cr(III) was based on the difference with the total concentration and a redox step and sample splitting were required [32,33].Fasihi et al. [32] used 1-(2-pyridylazo)-2-naphthol as a complexing agent, while 4-hydroxy-2-[(E)-(4-sulfonato-1-naphthyl) diazenyl] naphthalene-1-sulfonate (azorubine) was employed by Tuzen et al. [34] for the analysis of Cr in water and beverages samples.In a related research, Khoshmaram and Mohammadi [35] employed 1,5 diphenyl carbazide for the speciation of Cr in surface and groundwater samples.
In this research, a proposed method that utilized NaEDTA and APDC for the speciation analysis of Cr has been described.The NaEDTA served as both a precursor to the preconcentration of Cr(III) and as a masking agent.The Cr(III) was separated as the Cr(III)-EDTA by SPE via the ion pairing principle and the Cr(VI) was preconcentrated as the Cr(VI)-APDC complex by LPE.Sample splitting and a redox step were not required.Hence, the proposed method may find application in microanalysis where the sample is in small quantity.Parameters such as the effect of temperature, reaction time, pH, effects of metal ions and the amounts of adsorbent, TBA, and ligands, were studied.The method was validated and applied to water samples.

INSTRUMENTATION
The analysis of chromium was achieved with the ZEEnit 650P GF-AAS (Analytik Jena, Germany), equipped with an MPE 60 auto-sampler, a transversely heated graphite tube atomizer, and a Zeeman-2-way background correction system.The radiation source was a 4.0 mA current from a Cr hollow cathode lamp (Analytik Jena AG, Germany).The analysis was performed on the liquid platform set up, 357.90 nm analytical line, slit width 0.8 nm, and PMT 300 v.The drying temperature and time were 80 • C -110 • C, and 50 s respectively, Pyrolysis temperature and time were respectively 350 • C -1300 • C, and 30 s, atomization temperature and time were 2300 • C, and 5s correspondingly, clean up temperature and time were 2450 • C, and 4 s, in that order and spectral measurements were performed within 5 s of atomization.The Fisherbrand FB70155 (Fisher Scientific, UK) pump and SPE set were used for the solid phase preconcentration process.The pH meter was the Metler-Toledo pH meter, (FiveEasy F20, Switzerland), and the shaking-water bath was the Julabo SW22 (Julabo GmbH, Germany) while centrifugation was achieved by the KUBOKU 4200 centrifuge, (Kuboku, Japan).

SAMPLES
Samples of wastewater were obtained from the wastewater treatment plant of the Sungai Buloh Hospital, Selangor, Malaysia.Tap and drinking water samples were from Bangunan Makmal Kimia of the University Malaya, Kuala Lumpur.Samples were collected in 1 L high-density polyethene (HDPE) containers and stored as recommended by the USEPA [9,13].

Preliminary study
The preliminary study of the SPE and LPE steps was made by the UV-Vis (Shimadzu UV-1800, Japan) and aimed to ascertain the specificity of the NaEDTA and APDC to Cr(III) and Cr(VI) respectively in the mixture.The premier investigation involved three test tubes.In test tubes, (A) and (B) were 2 mL each of   The purple-coloured Cr(III)-EDTA complex was eluted with 5% (v/v) HCl into a test tube and the UV-Vis spectrum was obtained.The UV-Vis spectra of Cr(III)-EDTA, from test tubes (A) and (C) were compared.Similarly, the spectra of the ethyl acetate extracts containing Cr(VI)-PDC, from test tubes (B) and (C) were compared (Figure 2).
Parameters were optimized by the isolation method (one parameter varied at a time), and determinations were made by the ETA-AAS.An optimum condition became a constant parameter while optimizing the next parameter.Variable amounts of Cr(III) and Cr(VI) species standards were used for the optimization [36,38].Since the SPE tank and vacuum pump were not equipped with the facility to control the flow rate of samples, the SPE system was manually calibrated and the flow rate was fixed at 1.0±0.2mL min −1 and used through the solid phase preconcentration process.

Adsorbents efficiency
Four adsorbents were compared to study the efficiency of the solid phase preconcentration of the Cr-EDTA complex.The result of the comparison presented in Figure 3 indicated that the Al 2 O 3 was a better adsorbent of the Cr-EDTA complex.

SPECIATION ANALYSIS OF CHROMIUM PROCEDURE
Chromium species were determined by the ETA-AAS using the setup outlined under Instrumentation.The external calibration method was used for the instrument calibration.A 30 mL portion of the dilute sample, 40 mL CRM or 20 mL of suitable Cr standard, was placed in a 50 mL HDPE centrifuge tube. 2 mL of 0.3% (v/w) Na 2 EDTA solution was added and the pH was adjusted with NaOH and HNO 3 to between pH 5.0 and pH 6.0 using the Metler-Toledo pH meter.The tube was capped loosely and placed in the Julabo SW22 shaking water bath at 80 • C for 10 minutes.The mixture was cooled to room temperature in the open air, and then 2.5 mL of 0.5% (w/v) APDC solution was added and allowed to stand for 3 min.A 3 mL portion of ethyl acetate was added and covered securely before shaking the mixture vigorously and centrifuged for 3 min at 4500 rpm with the KUBOKU centrifuge.The organic phase was pipetted into a 25 mL conical flask containing 10 mL, 5% (v/v) HNO 3 covered with a watch glass and then digested on a hot plate at ≤ 90 • C. The digest was transferred into a 15 mL HDPE test tube and made up to 10 mL with deionized water before analysis for Cr(VI) concentration.
The aqueous phase was treated with 3.5 mL 0.3% (w/v) TBAB solution and diluted to 40 mL with deionized water (except the CRM), before adjusting the pH (pH 5 to pH 6), with HNO 3 and NaOH solutions.The mixture was then loaded onto a 3 mL SPE cartridge holding 500 mg Al 2 O 3 at about 1.0 ± 0.2 mL min −1 .The Cr-EDTA complex was eluted with 6 mL 5% (v/v) HCl solution into a 15 mL centrifuge tube and made up to 10 mL with deionized water before ETA-AAS analysis.

UV-VIS ANALYSIS OF THE COMPLEXES
The spectral analysis of extracts, eluates, and contents from test tubes (A), (B) and (C) clarified the feasibility of the one-pot speciation analysis of Cr(III) and Cr(VI) by EDTA and APDC chelating agents due to the similarity of the spectra, (Figure 2).The proposed mechanism of the chromium -EDTA reaction has been reported by Cerar [36] and a study on the general ligand-metal behaviour during complex formation was reported by Eigen and Wilkins [41].

THE EFFECT OF PH ON THE CHROMIUM-EDTA-AND-APDC COMPLEXES FORMATION
Even though the chromium-EDTA-and-APDC complexation is well known, studying the effect of pH on the formation was important because a combination of complexation agents was involved in this study.It was studied using standards of the species fortified in DI water and treated at various pH before preconcentration at pH 5. Figure 4 depicts the effect of pH on the formation of the Cr(III)-EDTA and Cr(VI)-APDC complexes.The Cr(III)-EDTA complexation was affected at higher pH values probably due to the continuous replacement of the water molecules in the [Cr(H 2 O) 6 ] 3+ by OH − and the subsequent formation of solid [Cr(OH) 3 (H 2 O) 3 ] complex that makes the OH − replacement difficult as well as deplete the concentration of reactive Cr(III) species in the solution, thus accounting for the low preconcentration of the analyte for complexation.

THE EFFECT OF PH ON THE CHROMIUM-EDTA-AND-APDC COMPLEXES PRECONCENTRATION
The pH is significant in the preconcentration of Cr(III)-EDTA as it plays a major role in determining the magnitude and stability of the charge of the Cr-EDTA complex ion [36].The role of pH in the preconcentration techniques was studied by treating mixed analytes in DI water at the optimum pH of formation and other optimized conditions, followed by preconcentration at different pH values.The effect of pH on preconcentration techniques as presented in Figure 5 indicated that solid-phase preconcentration of the Cr-EDTA seems unaffected.At pH 5 to pH 7, similar behaviour was observed in previous research [36,40].Earlier studies revealed that under slightly acidic conditions, the monovalent [Cr(EDTA)] − ion predominates [36].The data also indicate that the univalent [Cr(EDTA)] − ion is favourably paired with the equally univalent (R 4 N) + ion of the TBAB.Nevertheless, the preconcentration of the Cr-APDC complex was relatively stable between pH 2 to pH 6. Hence pH 5 to pH 6 was chosen for both the liquid and solid phase preconcentration of the Cr complexes.

THE EFFECT OF TEMPERATURE ON THE CHROMIUM-EDTA-AND-APDC COMPLEXATION
From the preliminary studies and the literature, the formation of the Cr(III)-EDTA complex is slow at room temperature while Cr(VI)-APDC forms instantaneously under the same condition.This observation made it necessary to study the effect of temperature on the complexation process.The standards of Cr species fortified in DI water and kept at pH 5 to pH 6 were treated in a water bathe at temperatures ranging from 40 • C to 100 • C and the liquid-and-solid phase preconcentration processes were carried out at the same pH range.The concentration-temperature curve (Figure 6), showed that 80 • C is the optimum temperature for Cr-EDTA complexation.On the other hand, the Cr(VI)-APDC complexation seems unaffected by temperature in this study.

THE EFFECT OF TIME ON THE CR(III)-EDTA AND CR(VI) -APDC COMPLEXATION
The standards of chromium species were mixed, and treated at optimum temperature (80 • C), and pH, (pH 5 to pH 6), at various times, and then preconcentrated with the LPE-and-SPE procedures before analysis with the ETA-AAS.The formation of Cr-PDC complexes was fast while the Cr-EDTA formation took 10 min to complete under the conditions as shown in Figure 7.

EFFECT OF THE AMOUNT OF SOLVENTS, LIGANDS, AND PAIRING AGENTS ON THE RECOVERY OF CR(III) AND CR(VI)
From the preliminary study, the NaEDTA and APDC are specific with respect to Cr(III) and Cr(VI) species respectively.However, a study has shown that they are complex with other metals as well [36].Besides, real sample compositions are heterogeneous, particularly in terms of metal content.For this reason, the optimization of the amounts of solvent, ligands, and pairing agent was particularly important to ensure significant analytes recovery in real samples despite competition.The optimization process was achieved by varying the volume of these reagents one at a time and computing the recoveries.In Figure S1 (Supplementary), the recovery of Cr(VI) was stable from 1.5 mL.ethyl acetate.Therefore, 3 mL to 4 mL ethyl acetate was used in the subsequent preconcentration of Cr(VI)-PDC.On the other hand, the recovery of eluted Cr-EDTA was relatively stable from 6 mL HCl as shown in Figure S3 (Supplementary).For subsequent elution, 6 mL to 10 mL (5% v/v) HCl was chosen for the elution of the Cr(III)-EDTA complex.While 2 mL (0.5% w/v) EDTA and 2.5 mL (0.5% w/v), APDC were sufficient for the complexation of 100 µg L −1 each of Cr(III) and Cr(VI) respectively as shown in Figures S2 and S4 (supplementary) and were used as optimum amounts of ligands in further sample preparation.Nevertheless, 4 mL of TBAB (0.3% w/v), was the optimal volume for pairing 100 µg L −1 Cr(III)-EDTA for the SPE preconcentration, Figure S5 (Supplementary).

EFFECT OF METAL IONS ON THE RECOVERY OF CR(III) AND CR(VI)
Metal ions of Cu, Fe, Co, and Ni among others, chelate EDTA and APDC [36,40], which makes it necessary to study the effect of other metals that may compete for the ligands.A multielement standard XVI (MS XVI) (Merck, Darmstadt, Germany), made of 21 elements from different parts of the periodic table at equal concentration was used.Though it is not expected that the metal ions be present in equal amounts naturally, the composition of MS XVI placed it close to the real environment.To accomplish studying the effects of metals, portions of DI water containing 50 µg L −1 each of Cr(III) and Cr(VI), were spiked with 2.5 µg L −1 , 7.5 µg L −1 and 15.0 µg L −1 of MS XVI and treated as in the procedure.The recovery studies (Table 1), were between 102% and 114% Cr(III) and 80% to 99% Cr(VI).The recoveries were within the acceptable range (80% to 120%) despite the presumed competition for the ligands [9].Nevertheless, the observed increase in recovery of Cr(III) with increasing concentration of MS XVI is presumably a response to the presence of Cr(III) in the MS XVI standard.In a real application, the wastewater samples were diluted before sample preparation and the results were calculated appropriately.

EFFECT OF ADSORBENT, AL 2 O 3 DOSAGE ON THE RECOVERY OF CR(III) AND CR(VI)
The effect of the adsorbent dosage was studied and the result is presented in Figure 8 where it can be seen that 400 mg adsorbent was preferred for the adsorption of Cr species.The accuracy of the method was studied by monitoring percentage recovery.Four concentration levels, (10 µg L −1 , 20 µg L −1 , 50 µg L −1 , and 100 µg L −1 ), of the analytes in DI water, were subjected to treatment and analysis as described in the procedure and the accuracy was calculated for each level using Equation (1).The result in Table 2 showed that recoveries from the analysis fell within the acceptable range of 80% to 120% [9].The mass balance compares the sum of chromium recovered from speciation analysis with the total determined from a fortified sample containing an equal concentration of the analyte mixture as that used for the speciation analysis.To determine the total concentration, the APDC complex of the mixture was preconcentrated in ethyl acetate, digested with HNO 3 and diluted to a volume equal to that used for speciation analysis before the ETA-AAS analysis.The results in Table 2, showed agreement between the sum from the speciation analysis and the total Cr determined.

METHOD VALIDATION
The method precision was studied following the method described by Narola et al. [42].Six portions of each of the DI water samples fortified with 100 µg L −1 Cr(III) and 50 µg L −1 Cr(VI), were subjected to sample treatment and analysis as in the procedure to obtain six replicate determinations.The RSD% was calculated using Equation (2).The result in Table S1 (Supplementary) recorded 1.87% and 7.67% precision for Cr(III) and Cr(VI) respectively and agree with the literature of ≤10.00% relative standard deviation [20].

LOD AND LOQ
The limits of detection and quantitation represent the concentration of analytes detectable without and with accuracy respectively [7,20].The LOD and LOQ were determined following the fortified blank method [43][44][45].Deionized water was spiked with a very low concentration of analytes and passed through sample preparation.The ZEEnit 650P GF-AAS was programmed to analysed and calculate the limits.Eleven replicates were made from where the standard deviation (σ), and the limits were calculated by the Aspect LS software.By this software, the LODs were 3σ (0.42 µg L −1 and 0.62 µg L −1 ) Cr(III) Cr(VI) and the LOQs were 9σ (1.27 µg L −1 and 1.87 µg L −1 ) for Cr(III) and Cr(VI) respectively, (Table 3).

MATRIX EFFECT
Analytes signals are often enhanced or suppressed by the presence of in the matrix (endogenous substances), substances introduced into the matrix (exogenous substances) or substances within the analysis environment called the extraneous substances.The matrix effect affects the quality of analytical results.The extent of suppression or enhancement of the signal was studied by the calibration slope method as described by Lehotay et al. [46] and Leito [47] with some modifications and reported as a percentage relative matrix effect RME(%).The RME% compared the slopes of the matrix-matched samples or the reagents (used for digestion or elution), to that of the fortified DI water which is often used for dilution purposes (Equation (3)).This method for studying the ME was suitable due to the non-availability of the commercial Cr-EDTA and Cr-APDC complexes.The Aspect LS software was programmed to calibrate by dilution and 0 µg L −1 , 10 µg L −1 , 20 µg L −1 , and 30 µg L −1 were prepared by programmed dilution from 40 µg L −1 matrix-matched stock solutions and fortified DI water.The solutions were prepared in DI water, HCl (5%), HNO 3 (10%) and filtered samples diluted five times.The RME(%) above 100% indicated signal enhancement and below 100% showed signal suppression.
The RME% as presented in Table 4 was between 96% and 105% representing signal enhancement and suppression between -5% and +5% respectively, (<10% in both cases).This may be indicative of the non-detrimental effect of the RME on the overall analytical result.The sensitivity of the instrument as measured by the characteristic concentration was between 0.2 µg L −1 (1%A) −1 and 0.5 µg L −1 (1%A) −1 .The calibration curves of the fortified matrix-matched samples, digestion and elution reagents as well as fortified DI water are shown in Figure S6 (Supplementary).

SPECIATION ANALYSIS OF CHROMIUM
The result of the speciation analysis of chromium by the method is present in Table 5 from where the concentration of Cr(III) as Cr(EDTA) ranges from 1.83 µg L −1 to 106 µg L −1 , while that of Cr(VI) as Cr(PDC) 3 and Cr(PDC) 3 (OPDC) was between <LOD (0.62 µg L −1 ) to 19.85 µg L −1 .Studies have shown that redox reaction may be triggered in the presence of organic matter and some metals especially Mn and Fe resulting in the inter-conversion of Cr(VI) and Cr(III) which more often than not favour the Cr(VI) to Cr(III) conversion [22].This could be responsible for the general higher amounts of Cr(III) compared to Cr(VI) species in the wastewater samples.
From the literature, Table 3, the LOD of Cr(III) ranged from 0.0041 µg L −1 to 7.7 µg L −1 , while that of Cr(VI) is between 0.002 µg L −1 and 3.5 µg L −1 .Although analysts employed different techniques for the determination of Cr species, the LODs (0.423 µg L −1 Cr(III) and 0.624 µg L −1 Cr(VI)), from this work indicates that the sensitivity of the method and instrument compared favourably with the literature.
The certified value of Cr in the CRM is 0.78 µg L −1 Cr(III), the amount of the Cr(III) found after sample treatment was higher.
That could be attributed to the preconcentration process as the species was harnessed into 10 mL from a 40 mL sample before analysis.Thus, the preconcentration factor (PF), calculated as the ratio of Cr(III) content of CRM after and before preconcentration is 3.57, (Equation ( 4)).Due to the non-availability of the CRM of the species, the PF determination could not be applied to Cr(VI).

CONCLUSION
The selectivity of EDTA for Cr(III) and the solubility of Cr(VI)-APDC complex were exploited for the speciation of Cr(III) and Cr(VI) in this proposed method.The ETA-AAS analysis revealed that the LOD of was Cr(III) 0.423 µg L −1 and that of Cr(VI) was 0.624 µg L −1 .The recovery of Cr(III) was between 83% and 117% and that of Cr(VI) was 79 and 99%, indicating that the accuracy of the method is within the analytically acceptable limit.The relative matrix effect was ±5% and the R 2 from the study of the matrix effect ranged from 0.99298 to 0.99997, further confirming the reliability of detection.The total chromium (0.74 µg L −1 ), determined from the raw CRM (LGC6019), agreed with the certified value of 0.78±0.20 µg L −1 of the CRM.Due to the non-availability of the Cr(VI) reference material, the preconcentration factor, ratio of the Cr(III) content of CRM before and after preconcentration was 3.57.The proposed method requires no redox reaction (to convert one species Cr to the other), this saves both reagents and the cost of analysis as well as possible matrix contamination.More so, this proposed method needs no sample splitting and, therefore, may be applied at the micro-scale level with different detection techniques similar to the ETA-AAS.Hence, the proposed method is comparable to earlier methods of chromium speciation involving a two-step procedure.

Figure 2 .
Figure 2. The UV-Vis spectra (a) extracts of Cr(III)-EDTA from test tubes (A) and (C), and (b) extracts of Cr(VI)-PDC from test tubes (B) and (C).

Figure 3 .
Figure 3.Comparison of adsorbents for the adsorption of 100 µg L −1 chromium at optimum conditions.