Cellulose mini-membranes modified with TiO2 for separation, determination, and speciation of arsenates and selenites

Sorptive and selective mini-membranes based on TiO2 directly synthesized onto cellulose filters (TiO2@cellulose) have been developed. The in situ synthesis of TiO2@cellulose applied is simple and economically advantageous. The obtained membranes can be useful for (1) separating arsenic(V) and selenium(IV) from other ions and organic matter, (2) speciation of arsenic and selenium, and (3) determining ulratraces of these ions in water samples. The membranes exhibit good stability and high maximum adsorption capacities for Se(IV) (71 mg g−1) and As(V) (41 mg g−1). A monolayer chemical adsorption of analytes on the membranes was confirmed. The structure of membranes was examined with scanning electron microscopy, x-ray diffractometry, and micro energy-dispersive x-ray fluorescence spectrometry (μ-EDXRF). The membranes were characterized by homogenous distribution of TiO2 onto cellulose. The TiO2@cellulose was used as a new sorbent in micro-solid phase extraction for determination of Se(IV) and As(V) by EDXRF. Using direct analysis of mini-membranes after sorption of analytes avoids the elution step. Thus, the proposed procedure is an attractive and solvent-free option for quantitative monitoring of Se(IV) and As(V) in different materials. Both analytes were quantitatively and simultaneously separated/determined from samples at pH 2 with very good recovery (close to 100%), precision (4.5%), and detection limits (0.4 ng mL−1 Se and 0.25 ng mL−1 As). TiO2@cellulose membranes were applied to water analysis. Graphical-abstract Effective method for determination of ultra trace arsenates and selenites using cellulose-based sorbent Electronic supplementary material The online version of this article (10.1007/s00604-020-04387-4) contains supplementary material, which is available to authorized users.

The regulations for water as well as toxicity of arsenic and selenium are given by United States Environmental Protection Agency (USEPA) [3,4], World Health Organization (WHO) [5], European Union (EU) [6]. Toxicity of arsenic and selenium are described inter alia by [7,8]. Maximum contaminant levels (MCL) of arsenic was established by USEPA at 10 ng mL -1 . USEPA established also the maximum acceptable concentration level of selenium in surface waters at 5.0 μg/L. As reported by WHO, EU, and USEPA selenium limit in drinking water was set at 40 μg L -1 , 10 μg L -1 and 50 μg L -1 , respectively.

Characterization of TiO 2 @cellulose
Figure S1. Mapping of TiO 2 @cellulose (collimator -0.3 millimeter diameter). Figure S2. The dependence of Se and As species on pH and the effect of pH on the recovery of Se(IV), Se(VI) and As(III), As(V) on TiO 2 @cellulose.

Adsorption isotherms
The adsorption of Se(IV) and As(V) ions on TiO 2 @cellulose at pH 2 was simulated using Langmuir [9,10] and Freundlich [11] isotherm models: where q max is the maximum amount of ions adsorbed per unit weight of TiO 2 at the high equilibrium ion concentration (mg g -1 ), K L is the constant related to the free energy of adsorption (L mg -1 ), C e is the equilibrium concentration, which is achieved during stabilization of the system (mg·L -1 ), K F (mg 1-n L n g -1 ) and n are Freundlich constants related to the adsorption capacity and the adsorption intensity, respectively. Isotherm parameters are included in Table S1. The better fitting the adsorption equilibrium data to the Langmuir model than the Freundlich model suggests a chemical adsorption process.

The influence of contact time and the sample volume on the adsorption of Se(IV) and As(V).
Figure S3. Effect of sample volume and contact time for TiO 2 @cellulose and Se(IV) (a) as well as As(V) (b) (x and z axis are sample volume and contact time, respectively).

Interferent study
The concentrations of potential interferents were chosen according to a literature study [3]: 20-200 μg mL -1 (Na + , K + , Mg 2+ , Ca 2+ ), 0.5-10.0 μg mL -1 (Al 3+ , Fe 3+ ), 0.01-250 μg mL -1 of SO 4 2-, 0.01-1.0 μg mL -1 of PO 4 3-, 100-500 µg mL-1 of Cl -, 100-800 µg mL -1 of NO 3 and 0.5 to 5.0 µg mL -1 of HA. The obtained results (see Table S2 and Figure 4S) prove that the metal cations, such as: Na + , K + , Mg 2+ , Ca 2+ , Al 3+ as well as anions such as: Cl -, NO 3 -, SO 4 2and HA in the whole studied concentration range does not influence the analytes sorption, which are adsorbed with recovery values in the range of 92-108 %. Only, the impact of Al 3+ and SO 4 2at high concentration levels (10 µg mL -1 Al 3+ and 250 µg mL -1 SO 4 2-) on arsenic sorption can be observed (Recovery of 85 %, and 62 %, respectively). Reducing the Al 3+ and SO 4 2concentration to 5 µg mL -1 and 50 µg mL -1 , respectively, allows obtaining 99-95% recovery of As(V). However, the US EPA recommends that aluminum levels for secondary drinking water do not exceed of 200 µg L -1 so, it is not a critical issue for As(V) determination in this type of water. Sulfate is not classified under the primary standards for drinking water. The secondary maximum contaminant level (SMCL) for sulfate is 250 µg mL -1 . The influence of 2.5 µg mL -1 Fe 3+ the recovery of both analytes is not observed. SMCL for iron is 0.3 µg mL -1 . The concentration of phosphates in the water sample should not exceed the concentration of 0.25 µg mL -1 (secondary standards) and 0.1 µg mL -1 (primary standards), so that the impact on recovery of As and Se is not observed. However, the maximum acceptable level of phosphorous in water to avoid accelerated eutrophication is 0.1 µg mL -1 . Sum up, the acidic solution (pH 2) enables the quantitative adsorption of Se (IV) and As (V), while preventing the precipitation of some elements found in real waters. Table S2. Effect of potential interferences on recovery of Se(IV) and As(V) on TiO 2 @cellulose.  Figure S4. Effect of foreign ions on recovery of Se(IV) and As(V) on TiO 2 @cellulose. *Se(VI) = Se(total)-Se(IV); **As(III) = As(total)-As(V)