Graphene for Preconcentration of Trace Amounts of Ni in Water and Paraffin-embedded Tissues from Liver Loggerhead Turtles Specimens Prior to Flame Atomic Absorption Spectrometry

A new sensitive and simple method was developed for the preconcentration of trace amounts of Ni using 1-(2-pyridylazo)-2-naphthol (PAN) as chelating reagent prior to its determination by ûame atomic absorption spectrometry. The proposed method is based on the utilization of a column packed with graphene as sorbent. Several effective parameters on the extraction and complex formation were selected and optimized. Under optimum conditions, the calibration graph was linear in the concentration range of 5.0–240.0 μg L-1 with a detection limit of 0.36 μg L-1. The relative standard deviation for ten replicate measurements of 20.0 and 100.0 μg L-1 of Ni were 3.45 and 3.18%, respectively. Comparative studies showed that graphene is superior to other adsorbents including C18 silica, graphitic carbon, and singleand multi-walled carbon nanotubes for the extraction of Ni. In the present study, we report the application of preconcentration techniques still continues increasingly for trace metal determinations by flame atomic absorption spectrometry (FAAS) for quantification of Ni in Formalin-fixed paraffin-embedded (FFPE) tissues from Liver loggerhead turtles. The proposed method was successfully applied in the analysis of four real environmental water samples. Good spiked recoveries over the range of 95.8–102.6% were obtained. Key word: Graphene; Solid-phase extraction; Preconcentration; Flame atomic absorption spectrometry; Formalin-fixed paraffin-embedded (FFPE) ; tissues from Liver loggerhead turtles;Ni al,1990).This element is needed by plants at only very low levels and is toxic at higher levels. At these levels, Ni can bind to the cell membrane and hinder the transport process through the cell wall. Ni at nearly 40ng mL-1 is required for normal metabolism of many living organisms (Gomes-Gomes 1995; Unger et al, 1979). On the other hand, Ni is an important 310 RASHID & MOGHIMI, Orient. J. Chem., Vol. 30(1), 309-317 (2014) element in many industries. Thus, the development of new methods for selective separation, concentration and determination of it in sub-micro levels in different industrial, medicinal and environmental samples is of continuing interest. The determination of Ni is usually carried out by flame and graphite furnace atomic absorption spectrometry (AAS) . (Boudreau et al, 1989 ) as well as spectrometric methods (Bruening et al,1991; Mahmoud and Soliman,1997 a) . Solid phase extraction (SPE) methods are the best alternatives for traditional classic methods due to selective removal of trace amounts of metal ions from their matrices. Solid phase extraction determinations can be carried out on different efficient ways. One of the most appropriative performation features of SPE is achieved by using octadecyl silica membrane disks. SPE reduce the use of toxic solvent, disposal costs, and extraction time(Mahmoud et al, 1997 b,45; Mahmoud et al, 1997b).The octadecyl silica membrane disks involves shorter sample processing time and decreased plugging due to the large cross-sectional area of the disk and small pressure drop which allows higher flow-rates; reduced channeling resulting from the use of sorbent with smaller particle size and a greater mechanical stability of the sorbent bed(Tong et al,1990). In our previous attempts, we modified SPE membrane disks with suitable compounds for selective determination of chromium(Dadler et al,1987; Moghimi 2007) and lead(Mahmoud et al,1990). Meanwhile, other investigators have successfully utilized these sorbents for quantitative extraction and monitoring trace amounts of lead(Leyden et al,1976; Moghimi et al,2009; Liu et al,1992), copper (Liu et al,1996; Mishenina et al,1996; Wang et al,1999), silver( Wang et al,1997; Zhang et al,1982), mercury (Zhou et al,1983; Zargaran et al, 2008) , cadmium (Tabarzadi et al, 2010) , palladium (Shin et al, 2004), Ni (Moghimi et al, 2012) and UO2 (Mahmoud et al, 1998; Moghimi et al, 2006, Pramod et al and Rafeeq et al.,). To ease the retrieval procedure, the SPE using graphene as the absorbent in a column combined with ûame atomic absorption spectrometry (FAAS) has been demonstrated by our research group (Wang et al., 2012). We extend its application to other inorganic analyses. 1-(2-Pyridylazo)-2naphthol (PAN), a chelating agent which forms stable complexes with a number of metals and has found numerous applications in trace element separation and pre-concentration methods (Narin and Soylak., 2003; Shokouû et al., 2007), was used to extract Co (structure of the Ni-PAN complex is shown in Fig. 1). What is more, it possesses a benzene ring structure. Based on this, the Ni-PAN is considered to have formed a strong p-stacking interaction with graphene when the sample solution passes through the column during which the Nichelate is retained. The factors inûuencing the efûciency of SPE and FAAS determination were systematically studied. The proposed method has been applied for the determination the National Institute of Standards (Beijing, China). Working standard solutions were prepared daily through serial dilutions of the stock solution with deionized water prior to analysis. The chelating agent, 2.0 g L-1 PAN solution, was prepared by dissolving the appropriate amount of PAN (Shanghai Chemistry Reagent Company, Shanghai, China) in absolute ethanol. Stock solution of diverse elements was prepared from high purity compounds. Single-walled CNTs (SWCNTs, carFigure 1 of trace amounts of Ni in water samples with satisfactory results. It reveals great potential of graphene as an excellent sorbent material in analytical processes for metal ions once again.


INTRODUCTION
Ni at trace concentrations acts as both a micronutrient and a toxicant in marine and fresh water systems (Izatt et al,1991;Izatt et al,1985;Izatt et al,1995;Blake et al,1996;Arca et al,2001;Ghoulipour et al,2002;Hashemi et al,2001; Shcherbinina et element in many industries.Thus, the development of new methods for selective separation, concentration and determination of it in sub-micro levels in different industrial, medicinal and environmental samples is of continuing interest.The determination of Ni is usually carried out by flame and graphite furnace atomic absorption spectrometry (AAS) .(Boudreau et al, 1989 ) as well as spectrometric methods (Bruening et al,1991; Mahmoud and Soliman,1997 a) .
Solid phase extraction (SPE) methods are the best alternatives for traditional classic methods due to selective removal of trace amounts of metal ions from their matrices.Solid phase extraction determinations can be carried out on different efficient ways.One of the most appropriative performation features of SPE is achieved by using octadecyl silica membrane disks.SPE reduce the use of toxic solvent, disposal costs, and extraction time (Mahmoud et al, 1997 b,45;Mahmoud et al, 1997b).The octadecyl silica membrane disks involves shorter sample processing time and decreased plugging due to the large cross-sectional area of the disk and small pressure drop which allows higher flow-rates; reduced channeling resulting from the use of sorbent with smaller particle size and a greater mechanical stability of the sorbent bed (Tong et al,1990).
In our previous attempts, we modified SPE membrane disks with suitable compounds for selective determination of chromium (Dadler et al,1987;Moghimi 2007) and lead (Mahmoud et al,1990) To ease the retrieval procedure, the SPE using graphene as the absorbent in a column combined with ûame atomic absorption spectrometry (FAAS) has been demonstrated by our research group (Wang et al., 2012).We extend its applica-tion to other inorganic analyses.1-(2-Pyridylazo)-2naphthol (PAN), a chelating agent which forms stable complexes with a number of metals and has found numerous applications in trace element separation and pre-concentration methods (Narin and Soylak., 2003;Shokouû et al., 2007), was used to extract Co (structure of the Ni-PAN complex is shown in Fig. 1).What is more, it possesses a benzene ring structure.Based on this, the Ni-PAN is considered to have formed a strong p-stacking interaction with graphene when the sample solution passes through the column during which the Ni-chelate is retained.The factors inûuencing the efûciency of SPE and FAAS determination were systematically studied.The proposed method has been applied for the determination the National Institute of Standards (Beijing, China).Working standard solutions were prepared daily through serial dilutions of the stock solution with deionized water prior to analysis.
The chelating agent, 2.0 g L-1 PAN solution, was prepared by dissolving the appropriate amount of PAN (Shanghai Chemistry Reagent Company, Shanghai, China) in absolute ethanol.Stock solution of diverse elements was prepared from high purity compounds.Single-walled CNTs (SWCNTs, car-Figure 1 of trace amounts of Ni in water samples with satisfactory re-sults.It reveals great potential of graphene as an excellent sor-bent material in analytical processes for metal ions once again.

Apparatus
A Shimadzu (Kyoto, Japan) Model AA-6300C atomic absorp-tion spectrometer equipped with deuterium background correction and a Ni hollow-cathode lamp as the radiation source were used for absorbance measurements at a wavelength of 240.7 nm.All measurements were carried out in an air/acetylene ûame.The instrumental parameters were adjusted according to the manufacturer's recommendations.A pH3-3C digital pH meter equipped with a combined glass-calomel electrode (Hangzhou Dongxing Instrument Factory, Hangzhou, China) was used for pH adjustment.The SPE experiments were performed on an Agilent vacuum manifold processing station with a Gast vacuum pump (Tegent Technology Ltd. Shanghai, China).The empty SPE columns (3.0 mL) and SPE frits were purchased from Agilent.
Nitric acid (0.1 mol L -1 ) was used to adjust the pH in the 2-3 range, and ammonium acetate buffers (0.2 mol L -1 ) were prepared by adding an appropriate amount of acetic acid to ammonium acetate solutions resulting in solutions with a pH range of 4.0-5.0.For a pH range of 6.0-8.0, a phosphate buf-fer solution (0.2 mol L -1 ) was prepared by adding an appro-priate amount of disodium hydrogen phosphate to sodium dihydrogen phosphate.Ammonium chloride buffer solutions (0.2 mol L -1 ) were prepared by adding an appropriate amount of ammonia to ammonium chloride solutions, resulting in solutions with a pH range of 9.0-10.0.
All reagents used were of analytical reagent grade.Deion-ized water was used in the preparation of all solutions.All glassware and columns were kept in 10% nitric acid for at least 24 h and subsequently washed four times with deionized water before application.

Synthesis and characterization of graphene
Graphene nanoparticles were synthesized according to our previously reported study (Wang et al., 2012).The size and morphology of G was observed by scanning electron micros-copy (SEM) using an S-3000N microscope and X-ray diffrac-tion (XRD) measurement was carried out using a Rigaku D/ max-rB diffractometer with Cu Ka radiation.In Fig. 2a, the SEM image shows the graphene agglomerate, consisting of al-most transparent carbon nanosheets with thin wrinkled and silk-like structures.XRD patterns in Fig. 2b reveal that the graphene nanosheets' peak at 2h = 26.2°,which is the charac-teristic peak of graphene (Rao et al., 2009).

Column preparation
Graphene (30.0 mg) was placed in a 3.0 mL SPE column using an upper frit and a lower frit to avoid adsorbent loss.Prior to extraction, the column was preconditioned with 10.0 mL meth-anol and 10.0 mL deionized water, respectively.The column was then conditioned to the desired pH with 5.0 mL of 0.2 mol L-1 acetate buffer solution.

Recommended procedure
100.0 mL of the sample solution containing 100.0 µg L -1 of Ni was prepared.2.5 mL of PAN (2.0 g L -1 ) solution was added and the pH value was adjusted to 5.0 with acetate buffer solu-tion.The resulting sample solution was passed through the col-umn at a flow rate of 2.0 mL min -1 .After the solution passed through it completely, the column was rinsed with 5.0 mL of deionised water, and the analytes retained on the column were

Sampling
Tap, sea, and river water samples used for the development of the method were collected in polytetrafluoroethylene (PTFE) containers from the Hebei Province.Before the anal-ysis, the organic content of the water samples was oxidized in the presence of 1% H 2 O 2 and then concentrated nitric acid was added.These water samples were then ûltered using a 0.45 lm pore size membrane ûlter to remove suspended par-ticulate matter and stored in a refrigerator in the dark before analysis.

Analysis of sample paraffin-embedded tissues from liver loggerhead turtles Specimens
Selected areas from fresh frozen tissues from liver loggerhead turtles specimens were sliced in three pieces (numbered as 1, 2 and 3) of approximately 10 mm × 5 mm × 2 mm each.Sets of pieces of set 1 (controls), were placed into a vacuum chamber at 50 ºC overnight to dry (until a constant weight was obtained) and the sets 2 and 3 were subjected to the standard 10 % buffered formalin fixation and paraffin embedding31 histological process using a tissue processor (Tissue-Tek VIP, Sakura Finetek USA Inc., Torrance, CA).After the paraffin embedding process, tissues were subsequently excised from the blocks with a titanium knife and deparaffinized in xylene at 55 º C for 1 h in the tissue processor (the set 2), or with hexane at 20 ºC for 1 week with frequent changes of the solvent in handling-based procedure (the set 3). Xylene was of a grade routinely used for the FFPE process and hexane was of ''Optima'' grade (Fisher Scientific).Upon deparaffinization, the tissue samples were dried in a vacuum chamber until constant weight was obtained.Each dried sample (of the sets 1-3) was divided into three portions (5-10 mg each) to be further analyzed as triplicates.

Effect of pH
Sample pH had a critical effect on the adsorption of target compounds by affecting the existing form of target com-pounds, the charge species and density on the sorbents surface (Jiao et al., 2012).A series of experiments was performed by adjusting the pH from 2.0 to 10.0 with nitric acid, ammonium acetate, phosphate and ammonium chloride.The results illus-trated in Fig. 3 reveals that the absorbance is nearly constant in the pH range of 5.0-7.0.The progressive decrease in the extraction of Ni at low pH is due to the competition of the hydrogen ion with the analyte for the reaction with PAN.Accordingly, pH 5.0 was selected for subsequent work and real sample analysis.

Influence of the amount of PAN
The effect of the amount of PAN on the absorption was stud-ied using various volumes of the reagent ranging from 0.5 mL to 3.5 mL.The signal of Ni was increased with the increase of PAN volume up to 2.0 mL, and then kept constant.Hence, 2.5 mL of 2.0 g L-1 PAN solution was chosen to account for other extractable species that might potentially interfere with the assaying of Ni.Sample and eluent solution, 2.0 mL min -1

Effect of flow rates of sample and eluent solution
The efficiency of metal preconcentration essentially depends on the flow rate of the sample solution to pass through the micro-column, whereas the ûow of eluent solutions affects the recov-eries.The time taken is also a considerable factor.Therefore, the effect of the ûow rate of the sample and eluent solutions on the recoveries of Ni on graphene was examined in the range of 0.5-5.0mL min -1 .The flow rate of the sample and eluent solutions had no obvious influence on the quantitative recover-ies of analytes at the range of 0.5-4.0 and 0.5-2.0mL min-1, respectively.A 2.0 mL min -1 flow rate of the sample and eluent solutions was chosen in subsequent experiments.

breakthrough volume
The measurement of breakthrough volume is important in so-lid phase extraction because breakthrough volume represents the sample volume that can be preconcentrated without the loss of  In the present study the possible preconcentra-tion factor was 200.

Adsorption capacity
In order to evaluate the adsorptive capacity of graphene, a batch method was used.100.0 mL of solution containing 1.0 mg of metal ion at pH 5.0 was added to 30.0 mg sorbent.The mixture was ûltered, after shaking for 10 min.10.0 mL of the supernatant solution was determined by FAAS.The capacity of the sorbent for Ni was found to be 20.6 mg g -1 .

Eluent type and its volume
In order to choose the best solvent for desorption of the ad-sorbed analytes on graphene, many reagent solutions were investigated.The results were given in Table 1.As can be seen, the recoveries of Ni were not so satisfactory when HCl, HNO 3 , H 2 SO 4 , CH 3 COOH and NaOH were used as eluents solely.Good quantitative recovery for analyte ions could be obtained with 2.0 mol L -1 HNO 3 in methanol.The effect of eluent volume on the recovery of Ni was also studied by using 2.0 mol L-1 HNO 3 in methanol; it was found that quantitative recoveries could be obtained with 2.0-4.0 mL of 2.0 mol L -1 HNO 3 in methanol.Therefore, the volume of 2.0 mL of 2.0 mol L-1 HNO 3 in methanol was used in the following experiments.

Column reuse
To investigate the stability and the potential regeneration of the sorbent, several extraction and elution operation cycles were carried out following the column procedure.It was found that the column can be reused after being regenerated with 10.0 mL methanol and 10.0 mL double distilled water, and it is stable up to at least 50 adsorption-elution cycles without a signiûcant decrease in the recovery of Ni.This is of great importance especially when the sorbent does the job with sat-isfactory analytical performance after using several circles in solid-phase extraction.

Effects of coexisting ions
The effects of common coexisting ions on the adsorption of Ni on graphene were investigated.In these experiments, 100.0 mL solutions containing 100.0 µg L-1 of Ni and various amounts of interfering ions were treated according to the recommended procedure.The tolerance limit is deûned as the ion concentra-tion causing a relative error smaller than ±5% related to the preconcentration and determination of analytes.The results, summarized in Table 2, show that the presence of major cations and anions in natural water has no signiûcant inûuence on the adsorption of Ni ion under the selected conditions.

Analytical performance
Under the optimized conditions, the calibration curve was obtained by preconcentrating a series of the solutions accord-ing to procedure under experiment.The curve was linear from 5.0 to 240.0 µg L-1 for Ni.The calibration equation is A = 3.16 • 10-3C + 0.0042 with a correlation coefûcient of 0.9992, where A is the atomic absorbance of Ni, obtained by peak height, in the eluent at 240.5 nm and C is its concentra-tion in the sample solution (µg L-1).The limits of detection and quantiûcation deûned as 3SB/m and 10SB/m (where SB is standard deviation of the blank and m is the slope of the cal-ibration graph) were 0.36 and 1.20 µg L-1, respectively.The relative standard deviation (RSD) for ten replicate measure-ments of 20.0 and 100.0 µg L-1 of Ni were 3.45% and 3.18%, respectively.

Analytical application
The proposed method was used for the determination of Ni in several water samples.The results, along with the recovery for the spiked samples, are given in Table 3.The recoveries for the addition of different concentrations of Ni to water samples were in the range of 95.8-97.6%.
To verify the accuracy of the proposed procedure, the method was used to determine of the content of Ni in the National Standard Reference Mate-rial for Environment Water (GSBZ 50030-94) after the appro-priate dilution and development of a methodology for the determination of Ni 2+ in FFPE tissue.The results for this test are presented in Table 3.A good agreement between the determined values and the certiûed values was obtained.

Comparison with other sorbent materials
In this work, we report a comparison between graphene with several commonly used reserved-phase sorbent materials including C18 silica, graphitic carbon, and CNTs.For this purpose, the same amount (30.0 mg) of different adsorbents was packed in 3.0 mL SPE columns.The columns were loaded with 100.0 mL of sample solutions containing 100.0 µg L-1 of Ni.All the work is done under the optimized conditions of graphene selected above.The C18 silica was evacuated from a Supelclean LC-18 SPE tube (Shanghai Chuding Instrument Company, Shanghai, China).The Ni in the flow-through, washing solution, and eluate were all determined.
As shown in Table 4, the graphene-packed column yields the highest recoveries (96.2%) among the studied adsorbents.This result definitely justiûes the worth of graphene as an SPE adsorbent.Ni could be detected in the flow-through and washing solution after loading on a C18 column, indicat-ing that 30.0 mg C18 silica is insufûcient for the retention of chelates.To obtain acceptable results with C18, more adsor-bent should be packed in the column to enhance the adsorp-tion capacity.For instance, with a C18 column packed with 300.0 mg C18 silica, the recoveries of Ni can reach 93.5%.However, increasing the adsorbent amount will add to the cost of analysis and is unfavorable for instrument miniaturization.Graphitic carbon performed even more poorly than C18.It was proposed that graphitic carbon did not give the expected extraction efûciency because of its large size and blank volume and less active sites for adsorption (Zhou et al., 2006).So it is noted that adsorption capacity of the adsorbents water sample volume, 100.0 mL; PAN was generally in the following order: Graphene >C18 silica >Graphitic car-bon.For MWCNTs, the recovery was approximately 78.3%, which is evidently inferior to that of graphene.Recoveries for columns using SWCNTs were higher than MWCNTs, but still inferior to graphene.No Ni was present in the ûow-throughs and washing solutions for MWCNTs and SWCNTs, indicating that CNTs also have good sorption capacities for Ni.Thus, the lower recoveries on CNTs should be ascribed to extremely stable adsorption and incomplete elution.Increasing the volume of the eluent solvent can improve the recovery, e.g., with 5.0 mL HNO 3 as eluent solvent, the recov-ery of Ni on SWCNT column can reach 92.3%.Nevertheless, increasing the volume of the eluent solvent will reduce the pre-concentration factor.
The advantage of graphene over C18 and graphitic carbon mainly lies in its higher sorption capacity.In addition, com-pared with CNTs, achieving complete elution with graphene is more facile.The above experimental results indicated that graphene is a very promising adsorbent material.

CONCLUSIONS
In conclusion, the proposed method reveals the great potential of graphene as an advantageous sorbent material in SPE.Using Ni as model analyte, the graphene-packed SPE columns showed reliable and attractive analytical performance in the analysis of environmental water samples.Higher recoveries were achieved with graphene than with other adsorbents including C18 silica, graphitic carbon, and CNTs, owing to the large surface area and unique chemical structure of graph-ene.Some other advantages of graphene as an SPE adsorbent have also been demonstrated, such as high sorption capacity, good reusability, and ûne reproducibility.Although the ob-tained results of this research were related to the Ni determi-nation, the system could be a considerable potential guide for the preconcentration and determination of other metals.

Table 1 : Effect of type and concentration of eluting agent on recovery of Ni
a Average of five determinations ± standard deviation

Table 3 : Analytical results of Ni determination (dissolved fraction) in certified reference materials and spiked natural water samples with the SPE-FAAS method (n = 3) Sample Ni 2+ added (µg) Ni 2+ determined(ng.mL -1 ) ICP-AES
Values in parentheses are %RSDs based on five individual replicate analysis b Not detected. a