Quantitative extraction of chromium VI and III from tanned leather: a comparative study of pretreatment methods

In this study, seven pretreatment methods for chromium speciation in tanned leather were evaluated: acidic mineralization, ethylenediaminetetraacetic acid (EDTA) extraction, diethylenetriaminepentaacetic acid (DTPA) extraction, alkaline extraction (NH4OH), ammonium nitrate extraction (NH4NO3), water extraction, and phosphate buffer extraction. Acidic mineralization permitted the decomposition of the organic matter and ensured the complete digestion of leathers, giving access to the total content of chromium in each sample using inductively coupled plasma-atomic emission spectrometry (ICP-AES). From all the extractant media tested, EDTA proved to be the most efficient, allowing the extraction of Cr(VI) and Cr(III) as a Cr(III)-EDTA complex, quantitatively. Method validation is presented for EDTA extraction and direct mineralization. For the EDTA extraction, method detection limit (MDL) and method quantification limit (MQL) for total Cr in leather were 3.4 ppb and 11.2 ppb (µg of total Cr per L of extraction solution), respectively. Due to the lack of leather certified reference materials (CRMs) for Cr(VI), accuracy was evaluated by spiking leather samples with a Cr(VI) solution. The spike recovery of EDTA microwave assisted extraction ranged from 91.0 to 108.6%. Interday precision was also evaluated and all variation coefficients were below 5%, for both mineralization and EDTA extraction. This article provides an efficient procedure to extract quantitatively chromium from leather, while maintaining the speciation, which can be further followed by ion chromatography-inductively coupled plasma-mass spectrometry (IC-ICP-MS).

insensitive to decay or decomposition, and gives greater stability and resistance to mechanical action and heat.
Even if hides are not directly treated with hexavalent chromium, which is a known carcinogen and mutagen [4], it has been proved that Cr(VI) can be released from chrome-tanned leather. Indeed, different environmental conditions (i.e. pH, light, temperature …) can induce trivalent chromium's oxidation into Cr(VI), as suggested by a recent publication from Fontaine et al. [5]. They studied the storage conditions that could influence the formation of Cr(VI) in four bovine leather samples, and found that a high temperature, a dry atmosphere, and presence of UV light had a great impact on Cr(VI) formation, with a predominant role of UV light. Consequently, these conditions should be considered when trying to predict the evolution of chromium in leather goods.
Furthermore, in direct contact with the skin, chrometanned leather products can cause chromium allergy and dermatitis [6,7]. For most leathers, Cr(III) is the dominant form of chromium that is released [8][9][10]. In a recent study, 10 Cr-allergic subjects and 22 controls were patch tested with Cr(III), Cr(VI), and leather samples. The results suggested that repeated exposure to Cr-tanned leather with predominantly Cr(III) release is sufficient to elicit allergic contact dermatitis in Crallergic individuals. Even if the majority of Cr-allergic subjects reacted to Cr(III) and Cr(VI), a lower dose of Cr(VI) was required for a stronger reaction [11]. Taking into account that throughout the life-cycle of leather chromium speciation has a dynamic behavior influenced by a large number of parameters, the one-time testing of Cr(VI) may be considered irrelevant if trivalent chromium is not tested as well. Moreover, no relationship has been found between released Cr(III) and Cr(VI) [10]. Besides, an article by Feng et al. [12], reported that both trivalent and hexavalent chromium induce genetic mutations and DNA damage in yeast cells, the effect of Cr(III) being greater than that of Cr(VI). Consequently, it is important to quantitatively determine both chromium species separately.
Regarding hexavalent chromium in leather, since 2015 the European Union limits its content to 3 mg/kg (of the total dry weight) in leather goods coming into direct contact with the skin [13]. However, the European Chemicals Agency has recently adopted an opinion to decrease the restriction limit of hexavalent chromium from 3 to 1 mg/ kg in leather, furs and hides. This shall come into effect 5 years after entry into force of the restriction [14]. Currently, the International Organization for Standardization (ISO) 17075 standard specifies a procedure for the selective extraction of hexavalent chromium in leather using K 2 HPO 4 •3H 2 O at pH 8.0. Further analysis of the extract is performed via colorimetric [15] or chromatographic analysis [16]. The colorimetric analysis consists of a reaction between Cr(VI) and 1,5-diphenylcarbazide (DPC), the resulting complex is detected by UV Visible (UV-Vis) spectroscopy at 540 nm. The chromatographic analysis can be performed by direct detection of the chromate peak at 372 nm, or through post-column reaction with DPC by measuring the absorption peak at 540 nm. The limit of quantification (LOQ) of this standard procedure is 3 mg/kg for both techniques. Table 1 shows a summary of the pretreatment and analytical methods that have been used for the determination of hexavalent chromium in leather samples. All of these techniques rely on a Cr(VI) extraction with phosphate buffer or NaOH, and then the complexation of hexavalent chromium with DPC for UV-Vis detection. Nevertheless, when working with leather matrices, the selectivity of the UV detection might be compromised as other compounds could absorb at the same wavelength as the target complex; it is important to remark that leather is frequently treated with a variety of chemicals (azo dyes, pigments, fungicides, bactericides, binders, cationic/anionic surfactants…).
However, it can be noticed that there is no current standard describing a method to extract and quantify Cr(III) and Cr(VI) simultaneously. A lot of studies have been conducted regarding chromium speciation in other matrices, for example foodstuffs, cosmetic products, environmental, and pharmaceutical samples [20], but as far as we know, the simultaneous extraction of both Cr species, while maintaining the speciation, has not been studied for leather samples. Fabregat-Cabello et al. [21] reported a focused microwave assisted extraction of Cr(VI) in several CRMs (soil containing chromite ore processing residue, welding dust loaded on a filter, and sandy clay soil) using EDTA 50 mmol/L at pH 10. According to their results, they were able to extract and measure Cr(VI) and Cr(III)-EDTA complex, simultaneously. The separation and analysis of the species was performed by IC-ICP-MS. DTPA has also been tested as chelating agent for trivalent chromium but it has shown lower kinetical stability than EDTA [22]. To extract Cr(VI) from solid matrices, alkaline extraction with NH 4 OH has been employed for dairy products, flour, chocolate, vegetables, fruits, meat, fish, eggs, and beverages. High-performance liquid chromatography-inductively coupled plasma-mass spectrometry (HPLC-ICP-MS) was used to quantify Cr(VI) after ultrafiltration of the extract [23]. Gürleyük and Wallschläger [24] reported the successful determination of Cr(III) and Cr(VI) in water by IC-ICP-MS using EDTA to derivatize Cr(III). However, as far as we know, the potential of EDTA to extract Cr species in leather samples has not been tested. For waters [24,25], bottom sediments [25], and traditional Chinese medicines [26], chromium species separation has been performed by chromatographic methods using ammonium nitrate as mobile phase after EDTA derivatization or phosphate buffer extraction. Therefore, there is a possibility that if the mobile phase serves to elute both chromium species, it can serve as extraction media as well.
Based on these studies, the objective of this work was to evaluate seven pretreatment methods in tanned leather: acidic mineralization (which permits the decomposition of the organic matter allowing the total Cr quantification by ICP-AES), EDTA extraction, DTPA extraction, NH 4 OH extraction, ammonium nitrate extraction (NH 4 NO 3 ), water extraction (for water-soluble chromium compounds), and phosphate buffer extraction (ISO 17075). The efforts in leather research are limited and, to the best of our knowledge, none of these extraction systems have been tried on leather samples. Method validation is presented for EDTA extraction and direct mineralization, and encompassed linearity, accuracy, interday precision, and method detection and quantification limits. IC-ICP-MS using Cr isotopic standards was performed in order to study the possible Cr interconversions resulting from the EDTA extraction conditions.

Instrumentation
Extraction and mineralization of samples were performed with a Milestone ETHOS 1 microwave digestion system equipped with a temperature sensor. TFM reactors were used with a maximum pressure of 100 bar. ICP-AES measurements were performed on an iCAP 6300 series instrument (Thermo Fisher Scientific) using the axial view. The quantification of chromium was performed at 283.5 nm. Chromatographic separation was performed using a Dionex ICS-6000 Capillary HPIC System (Thermo Fisher Scientific). A Dionex ADRS 600 Anion Dynamically Regenerated Suppressor (2 mm) (Thermo Fisher Scientific) was set before the entrance into ICP-MS to avoid salt deposition and was operated in the external regeneration mode. A constant flow of 0.6 mL/min of ultrapure water was supplied by means of a Dionex AXP auxiliary pump (Thermo Fisher Scientific). Two guard columns were placed before the analytical column: a Dionex IonPac NG1 (2 × 50 mm, particle size: 10 µm) and a Dionex IonPac AG7 (2 × 50 mm, particle size: 10 µm). Ions were separated with a Dionex IonPac AS7 column (2 × 250 mm, particle size: 10 µm). The analytical column was coupled to the nebulizer of the ICP-MS using a PEEK tubing (0.25 mm i.d., 150 cm long). Detection of chromium species was performed on an iCAP RQ ICP-MS instrument (Thermo Fisher Scientific) fitted with a collision cell. Two different measurement modes were used: standard mode (STD) and kinetic energy discrimination mode (KED). In the KED mode, He gas was injected in the collision cell and an energy discrimination barrier was applied in order to remove unwanted polyatomic interferences. Data processing was performed using ThermoFisher Scientific Qtegra software. As example, optimized conditions for the plasma and nebulizer for ICP-AES and ICP-MS are given in Table 2.
The pH of the solutions was measured using a PHM240 pH/Ion Meter (MeterLab, Radiometer Analytical, Villeurbanne, France). Centrifugation of the extracts was performed with an Eppendorf Centrifuge 5702. Alkaline and ammonium nitrate extractions were performed using an Elmasonic S ultrasonic bath (Elma, Singen, Germany).

Total chromium determination
Leather was either directly mineralized or extracted prior to mineralization of the supernatant to further quantify by ICP-AES the amount of chromium extracted by the different media (EDTA, DTPA, NH 4 OH, NH 4 NO 3 , water, phosphate buffer). After the mineralization process, samples were diluted and total chromium was determined by ICP-AES. The calibration standards were freshly prepared with potassium dichromate and they were acidified as the same level as the samples. A scheme of the methodology is shown in Fig. 1a.

Sample preparation
Dry leather samples were cut into small pieces (3 to 5 mm per side) in accordance with ISO 4044:2017, 6.3 [27]. They were cut in the laboratory and immediately after this, they were stored in resealable plastic bags. This was done for each sample in the same way during the study. For each pretreatment method, three samples were treated in three different days.

Mineralization (or acidic digestion) 2.5.1 Direct mineralization of leather
An amount of 200 mg of leather was digested with 2 mL of nitric acid at 150 °C in a microwave system: 10 min from room temperature, 21 °C, to 150 °C, 30 min at 150 °C. After the mineralization, the digest was diluted up to 50 mL with water. For total Cr determination by ICP-AES, 1.25 mL of the last solution were diluted up to 25 mL with water. Calibration standards were acidified at the same level as the samples.

Mineralization of the extracts
Two milliliters of the extract (EDTA, DTPA, NH 4 OH, NH 4 NO 3 , water, or phosphate buffer extraction) were digested with 2 mL of nitric acid at 150 °C in a microwave system: 10 min from room temperature, 21 °C, to 150 °C, 30 min at 150 °C. After the mineralization, the digest was diluted up to 50 mL with water. For total Cr determination by ICP-AES, 1.25 mL of the last solution were diluted up to 25 mL with water. Calibration standards were acidified at the same level as the samples.

EDTA extraction
An amount of 200 mg of leather was extracted with 4 mL of EDTA 50 mmol/L (adjusted to pH 10 with NH 4 OH) at 100 °C in a microwave system: 10 min from room temperature, 21 °C, to 100 °C, 30 min at 100 °C. The extract was afterwards centrifuged for 15 min at 4000 rpm, the supernatant was collected and mineralized for further total Cr quantification.

DTPA extraction
Similarly, for the DTPA extraction 200 mg of leather were extracted with 4 mL of DTPA 50 mmol/L (adjusted to pH 10 with NH 4 OH) at 100 °C in a microwave system: 10 min from room temperature, 21 °C, to 100 °C, 30 min at 100 °C. The extract was afterwards centrifuged for 15 min at 4000 rpm, the supernatant was collected and mineralized for further total Cr quantification.

Alkaline extraction
For the alkaline extraction, 500 mg of leather were extracted with 10 mL of NH 4 OH at ultrasonic bath during 1 h, as previously described by Vacchina et de la Calle [23] for dry food samples. The extract was centrifuged for 15 min at 4000 rpm and filtered using a nylon 0.45 µm filter. Further digestion of the extract was performed before total Cr quantification.

Ammonium nitrate extraction
An amount of 500 mg of leather was extracted with 10 mL of NH 4 NO 3 0.1 M at ultrasonic bath for 1 h. After the ultrasonic assisted extraction, the solution was centrifuged for 15 min at 4000 rpm and filtered using a nylon 0.45 µm filter. The digestion of the extract was performed as formerly described.

Water extraction
An amount of 200 mg of leather were extracted with 4 mL of ultrapure water at 100 °C in a microwave system: 10 min from room temperature, 21 °C, to 100 °C, 30 min at 100 °C. The extract was centrifuged for 15 min at 4000 rpm and the supernatant was collected. Then, the supernatant was mineralized for total Cr quantification.

Phosphate buffer extraction
In order to evaluate the ISO 17075 standardized method, 500 mg of leather were extracted with 25 mL of K 2 HPO 4 •3H 2 O at pH 8.0 (adjusted with phosphoric acid) for three hours under magnetic stirring. The extracts were subsequently filtrated through nylon 0.45 µm filter and mineralized for total Cr determination.

Sample spiking 2.12.1 Sample spiking of mineralization
An amount of 200 mg of leather was spiked with 1, 3, and 5 mg of Cr(VI) coming from a concentrated solution of potassium dichromate freshly prepared. Subsequently, the samples were mineralized, diluted, and analyzed by ICP-AES for total Cr quantification.

Sample spiking of EDTA extraction
An amount of 200 mg of leather was spiked with 2, 6, and 10 mg of Cr(VI) with a concentrated solution of potassium dichromate freshly prepared. The samples were extracted with an EDTA solution (as specified in Sect. 2.6) and the supernatant was mineralized as previously specified. After proper dilution, the samples were analyzed by ICP-AES for total Cr quantification. This methodology is described in Fig. 1b.

Complexation
In order to follow possible Cr interconversions during the separation and determination of species through IC-ICP-MS, 400 µL of a Cr isotopic standard ( 53 Cr(III) or 50 Cr(VI)) was complexed with 4 mL of EDTA 0.5 mmol/L (diluted from the EDTA extraction solution, specified at 2.6) at 100 °C in a microwave system: 10 min from room temperature, 21 °C, to 100 °C, 30 min at 100 °C. Solutions were subsequently diluted and analyzed by IC-ICP-MS.

Analysis of the samples
Separation was achieved using NH 4 NO 3 75 mmol/L (adjusted to pH 3 with HNO 3 ) as mobile phase. Analysis were conducted on isocratic mode at a mobile phase flow

Evaluation of different pretreatment methods
Leather samples coming from the luxury industry were analyzed after going through seven different pretreatment procedures to evaluate their extraction rates. Samples used for these experiments had gone through different manufacturing processes, and were available in limited quantity, hence the content of chromium may vary from one sample to another (i.e. samples used for EDTA extraction were not the same as the samples used for NH 4 OH extraction). Additionally, considering that leather is not manufactured, it is less homogeneous than other materials and chromium content may vary from one area to another. Consequently, for each sample (sample 1, sample 2, sample 3), a specific leather piece was selected, cut into small pieces, and used for the full procedure (day 1, day 2, day 3). For all the extractions done hereafter (EDTA, DTPA, NH 4 OH, NH 4 NO 3 , water, phosphate buffer extraction), a direct and complete mineralization of the same sample was done to provide each time the reference amount of total chromium, and the extraction rate was calculated with Eq. (1).
The results are shown in Table 3 and are only valid for the conditions described on the experimental section. The highest extraction rate was achieved with the EDTA extraction, which ranged from 92.5 to 108.3% of total Cr (Table 3a). The advantage of the EDTA extraction is that it complexes trivalent chromium decreasing the probability of its conversion to Cr(VI), and allows the dissolution of insoluble forms of Cr(VI) [28]. Therefore, it permits the extraction of both Cr species. A slightly lower rate, from 78.7 to 96.1% of total Cr, as shown in Table 3b, was obtained with DTPA as extractant, as a consequence of a lower stability of the [Cr(DTPA)] 2− complex [22].
It is well-known that the behavior of chromium species in aqueous media is principally affected by pH and redox potential. The extraction of hexavalent chromium in environmental samples is generally performed in alkaline media. Under these conditions, redox potential decreases, which promotes the stabilization of Cr(VI) [29]. For leather samples, NH 4 OH reached an extraction rate from 50.3 to 55.5% (Table 3c), half of the total chromium content, which means that a relevant part of Cr(III) was also extracted. At pH above 6, Cr(III) precipitates as Cr(OH) 3 (s). At higher pH, above 9, this precipitate (1) Extraction rate% = mg of extracted chromium(extraction) mg of total chromium(direct mineralization) × 100 solubilizes as a Cr(OH) 4 − complex. Hence the possibility of extracting Cr(III) using NH 4 OH as extractant.
Regarding the ISO 17075 standard, the phosphate buffer extraction exhibited an extraction rate from 5.1 to 6.6% of total Cr. As can be seen in Table 3f, total extracted chromium ranges from 700 to 800 mg/kg which is much higher than the restriction limit for hexavalent chromium. These values reveal that even if the protocol permits to extract hexavalent chromium, it also allows the extraction of a portion of trivalent chromium, present at levels much more important than the restriction limit of Cr(VI). The selectivity of this method comes from the selective reaction of DPC with Cr(VI), and the further detection of the complex.
The results for water extraction are shown in Table 3e. An extraction rate of 4.2 to 8.8% of total Cr (up to 1130 mg/kg of total Cr) was found when using only water, which implies that by submerging the sample in water and heating it to 100 °C, it is possible to extract some chromium species from leather. Considering that total extracted Cr was far above the restriction limit for Cr(VI), it can be affirmed that Cr(III) has been extracted although not in its entirety. Moreover, it is well-known that while Cr(III) compounds are barely soluble in water, Cr(VI) compounds are readily soluble in water [30].
Therefore, there is a possibility that Cr(VI) has been extracted as well.
According to the results presented in Table 3d, the lowest extraction rate was obtained when using ammonium nitrate as extractant (1.4 to 3.3% of total Cr). Hence, even if ammonium nitrate can be used for the successful elution of chromium species in some chromatographic methods [24][25][26], it is not an adequate extractant for chromium species in leather.
As mentioned before, a mineralization step succeeded each extraction. Mineralization permits the decomposition of the organic matter in the extract and makes possible the analysis and quantification of total Cr by ICP-AES. In Fig. 2 we can observe the summary of the results from Table 3. Error bars represent the standard deviation of the three results (three different days) obtained for each sample.

Validation of the method
Due to the high extraction rate obtained with EDTA, and the possibility to preserve speciation from the extraction step until the chromatographic analysis by IC-ICP-MS for example, this method was validated together with  the mineralization method. Linearity, Method Detection Limit (MDL), Method Quantification Limit (MQL), interday precision, and recovery are presented in the following subsections.

Linearity
Linearity was studied through the analysis of calibration solutions of Cr(VI) from 0 to 10 mg/L (n = 6). Each time, the standards were prepared with the same amount  of nitric acid as the real samples and total Cr was determined in each solution. Calibration curve was done each day a sample was analyzed, and linearity was found to be 0.9998 for all ICP-AES analyses performed throughout this study.

Method detection limit and method quantification limit
MDL and MQL were estimated by the analysis of three blanks and three replicates per blank (9 measurements in total). Blanks consisted of the extraction solutions (nitric acid for the mineralization and EDTA 50 mmol/L pH 10 for the EDTA extraction) and had gone through all the steps of pretreatment and analysis, in the same way as leather samples. For the 9 measurements, a standard deviation was obtained. With this value, MDL was calculated as 3 times the standard deviation (3 σ ) and MQL was calculated as 10 times the standard deviation (10 σ ). For the mineralization method, MDL and MQL for total Cr in leather were 3.9 ppb and 13.1 ppb (µg of total Cr per L of mineralization solution), respectively. MDL and MQL for the EDTA extraction were 3.4 ppb and 11.2 ppb of total Cr (µg of total Cr per L of extraction solution), respectively.

Interday precision
Interday precision was determined by analyzing three different leather samples in three different days. It can be noticed in Table 4 that for each sample the interday precision provided a relative standard deviation (RSD) below 5%, for EDTA extraction and mineralization. Low RSD values show that both methods exhibit excellent interday precision.

Spike recovery
Due to the lack of leather certified reference materials for Cr(VI), accuracy was evaluated by spiking leather samples with a concentrated Cr(VI) solution prepared with potassium dichromate. Sample spiking was performed at three different levels of spike (low, medium, high), three different samples per level, the same day. For each level, total Cr was determined by ICP-AES after proper extraction, mineralization, and dilution. This allowed to measure the matrix effects. As observed in Table 5, the spike recovery for the mineralization ranged from 80.1 to 95.2%, so the recovery was good at all spike levels. Same tendency was found for EDTA extraction, as recovery ranged from 91.0 to 108.6% for all spike levels. Both techniques provided acceptable results, which proved that EDTA extraction was efficient to quantitatively extract the hexavalent chromium that was used to spike the samples, as well as the total Cr content from leather.

Speciation by IC-ICP-MS
Regarding chromium speciation, IC-ICP-MS analyses were performed in order to verify if the EDTA extraction conditions permit to preserve chromium speciation. This was demonstrated by analyzing Cr isotopic standards. The analyses were carried out in the standard mode. KED mode was also performed but not differences were perceived. Figure 3a shows the chromatogram obtained from a 50 Cr(VI) solution. As can be seen, Cr(VI) elutes at 920 s and there is no other peak in the chromatogram. Figure 3b corresponds to a Chromium(III) isotopic reference standard enriched in 53 Cr, which was complexed with EDTA at the same conditions as the EDTA extraction used for leather. In this case, it is possible to observe that a 53 Cr(III)-EDTA complex was formed and it eluted at 230 s. An unidentified small peak with a retention time of 150 s was also noticed. Nevertheless, no peak was observed around 920 s. Thus, these conditions did not induce a conversion from Cr(III) to Cr(VI). Figure 3c corresponds to a Chromium(VI) isotopic reference standard enriched in 50 Cr. This solution was submitted to the same conditions as the EDTA extraction used for leather. As it was expected, no complex was formed as EDTA complexes only with Cr(III). Anyhow, it is possible to see, when zooming in, two small peaks with a retention time of 150 and 230 s. Each of these peaks represent 1.5% of the principal peak. The peak with a retention time of 150 s has not been identified. On the other hand, the retention time of the second peak corresponds to that of the Cr(III)-EDTA complex. Consequently, in the reverse direction, there is a conversion from Cr(VI) to Cr(III) under these extraction conditions, but this conversion is minimal. From this study, we can conclude that the EDTA extraction conditions allowed quantitative chromium extraction while limiting chromium interconversions. An ongoing study focuses on different leathers samples (chrome-tanned, colored, satin, or glossy leather), extracted with EDTA as presented in this study, and analyzed by IC-ICP-MS. Different analysis conditions are tested to give access to the amount of both species (Cr(III) and Cr(VI)), with the same protocol (extraction and chromatographic method).

Conclusions
This work focused on the extraction of chromium from leather samples. On the one hand, the acidic digestion allowed the complete mineralization and gave access to the total chromium content in each sample, but such conditions do not preserve chromium speciation as Cr(VI) is reduced to Cr(III). Among the extraction conditions tested (i.e. EDTA, DTPA, NH 4 OH, NH 4 NO 3 , water, phosphate buffer), EDTA proved to be the most efficient and ensured quantitative extraction of Cr(VI) and Cr(III) as a Cr(III)-EDTA complex. The content of total chromium obtained by the EDTA extraction was similar to the content obtained by direct mineralization. The method was validated in terms of linearity, recovery, interday precision, method detection limit, and method quantification limit. The separation of the extracted species was studied through IC-ICP-MS using Cr isotopic standards, which proved that the EDTA extraction conditions permit a quantitative extraction while limiting the interconversions. No interconversion from Cr(III) to Cr(VI) was noticed, and a limited portion (1.5%) of Cr(VI) was transformed into Cr(III) under the extraction conditions used.