Introduction

Functionalized nanoparticles (NPs) have many potential applications in analytical chemistry based on their unique magnetic, optical and electrochemical properties [1]. The functionalized superparamagnetic particles offer many advantages for the sample preparation and analyte preconcentration such as higher surface area, easy dispersion in a larger volume sample, and retrieval from sample using an external magnetic field without retaining residual magnetization after withdrawal of magnetic field [2, 3]. It is important that the magnetic particles should be easy to prepare, cheaper, nontoxic, stable under ambient conditions, should be superparamagnetic with higher magnetic saturation value, and easy to functionalize for the analyte specific applications. The magnetite (Fe3O4) NPs and its products of oxidation maghemite (γ-Fe2O3) have such properties and, therefore, used extensively not only in analytical chemistry but also for hosting the NPs for many other applications [4–7]. In general, the Fe3O4 NPs can be used after coating/anchoring desired functionality directly on their surfaces [8–11], immobilization in polymer matrix [12–16], or formation of ferrofluid [17, 18]. However, the most popular method appears to be the direct functionalization of the Fe3O4 NPs by physical coating or anchoring of functional groups by covalent linking [2, 3, 19–21].

Argonne National Laboratory has developed a magnetically assisted chemical separation (MACS) process for removal of transuranic elements for the waste management objectives [22]. Thereafter, several research papers have been published for the removal of actinides using different design and strategy of MACS [23–30]. However, the quantification of actinide requires the reproducible high extraction efficiency, selectivity towards the target analyte or a group of analytes, and reusability in a prevailing chemical environment. The quantification of the analyte preconcentrated on the Fe3O4 NPs is done either subjecting directly to tailored instrumental methods [31, 32], or eluting analyte in an appropriate solution which is subjected to conventional instrumental methods such as alpha spectrometry, ICP-AES etc. [33–35]. The radiation measurement based analytical methods are highly sensitive to radionuclides but lacks chemical selectivity, require constant radiation courting geometry, and often subjected to the sample preparation steps to avoid interferences and lowering detection limit by preconcentration [36, 37].

In the present work, the N,N,N′,N′-tetraoctyl diglycolamide (TODGA) and bis(2-ethylhexy)phosphoric acid (HDEHP) coated Fe3O4 NPs have been developed for the preconcentration of actinides from large volume samples followed by magnetic separation. These extractants are well known for their promising separation of lanthanides and actinides [38–40]. The coated particles are characterized by thermogravimetry, vibrating sample magnetometry, and their actinide sorption efficiencies. The amounts of actinides preconcentrated have been quantified either by decoating of the extracting phase on Fe3O4 NPs in scintillation cocktail which is subjected to alpha liquid scintillation counting, or by direct γ-spectrometry. The outline of the present work is shown in Scheme 1.

Scheme 1
scheme 1

The Outline of different steps involved in quantifications of actinides using extractants coated Fe3O4 NPs

Full size image

Experimental

Materials and methods

TODGA and Fe3O4 NPs (20–30 nm) were procured from Thermax, Pune, India and J. K. Impex, Mumbai, India, respectively, dodecane and tetra ethoxy silane (TEOS) from Sigma-Aldrich, methanol from S. D. Fine Chemicals, India, nitric acid (69–70%) received from Beaker chemicals Pvt. Ltd, Mumbai, India and liquid scintillation cocktail-O from Sisco Research Laboratory, India. JASCO 420 spectrometer was used for recording FTIR, and STARe system METLER TOLEDO instrument was used for thermo gravimetric analysis (TGA).

Coating of extractants on Fe3O4 NPs

The coating of TODGA on Fe3O4 (Fe3O4@TODGA) was carried out with and without pre-HNO3 treatment. For acid treatment, TODGA in dodecane (0.1 mol L−1) was equilibrated with equal volume of HNO3 (0.1–3 mol L−1) for 30 min prior to coating. It was then centrifuged and aqueous layer was discarded. Thereafter, Fe3O4 NPs (0.1 g) were dispersed in 10 mL of acid treated as well as in without acid treated solutions containing 0.1 mol L−1 TODGA in dodecane. These solutions containing Fe3O4 particles were subjected to constant agitation for 24 h at room temperature and 40 °C using a constant temperature shaking bath. Subsequently, the coated Fe3O4 particles were separated from solutions using an external magnet. Finally, the coated Fe3O4 particles (Fe3O4@TODGA) were washed 4–5 times with 5 mL of methanol, and dried under vacuum at room temp for overnight.

The coating of HDEHP was carried out in a similar fashion except that acid treatment was not given. Typically, 0.1 g of Fe3O4 NPs were dispersed in 10 mL of 0.1 mol L−1 HDEHP in ethanol and equilibrated under shaking condition for 24 h. Then the coated Fe3O4 was removed and washed 3 times with 5 mL methanol. It was then dried overnight at room temp.

Extraction experiments

The sorption/desorption of actinides in Fe3O4@TODGA was studied using 241Am and mixPu as representatives of actinides. Weighed amounts of Fe3O4@TODGA coated under different chemical conditions were equilibrated with 5 mL solution containing known activity of Am and Pu in 3 mol L−1 HNO3 for 2 h with a constant shaking at room temp. After equilibration, the aqueous phase and Am/Pu loaded Fe3O4@TODGA were separated by applying external magnet, and counted by γ-spectrometer consisting of NaI(Tl) detector coupled to a multichannel analyzer for 241Am, and a home built liquid scintillation counter was used for mixPu. The radioactivity of the aqueous phase before and after equilibration was used to determine the D value using following equation:

$$D = \frac{{(A_{i} - A_{f} )}}{W} \times \frac{V}{{A_{f} }}$$
(1)

where A i and A f represent γ/α -activity of radionuclide in aqueous phase initial and after equilibration, W and V are weight of the functionalized Fe3O4 and volume of equilibrating aqueous phase (5–10 mL), respectively. The uptake efficiency was calculated using following relation:

$${\text{Uptake}}\,{\text{efficiency}} = \frac{{[A_{i} - A_{f} ]}}{{[A_{i} ]}} \times 100$$
(2)

where Ai and Af are same as given in Eq 1. To study the reproducibility of coating procedure, four batches of Fe3O4@TODGA were prepared by the same procedure and equilibrated with 241Am as described above. Standard deviation was calculated on percentage extraction from different batches. Similar extraction experiments were also carried out with HDEHP coated Fe3O4 for Plutonium extraction at 3 mol L−1 HNO3.

Analytical applications

To explore analytical applications of TODGA@Fe3O4, the known amounts of 241Am activity (5.7–57 µCi) spiked in the 3 mol L−1 HNO3 and loaded in fixed weights of TODGA@Fe3O4 (100 mg) as described above for the sorption experiments. The 241Am activity loaded in Fe3O4 was counted by γ-spectrometry. For alpha liquid scintillation counting, the mixPu loaded TODGA@Fe3O4 particles were dispersed in 5 mL of different solvents like toluene, toluene based scintillator, isopropyl alcohol (IPA) and dioxane based scintillator for 24 h to remove TODGA coating in a scintillating glass vials. Toluene based scintillator contained 5% HDEHP in cocktail-O; whereas dioxane based scintillator contained 1% TOPO in cocktail-W.

Results and discussion

Characterizations of extractant coated Fe3O4 NPs

The morphology of the Fe3O4 NPs did not change after coating the extractant (TODGA/HDEHP) as shown in the representative FE-SEM image in Fig. 1. The size of particles were 20 ± 5 nm. In order to stabilise the magnetite nanoparticles in high nitric acid medium, the magnetite nanoparticles were coated with silica using tetra ethoxy silane (TEOS). The presence of extractant on Fe3O4 NPs was confirmed by FTIR spectroscopy. The absorption peaks at around 638 and 580 cm−1 observed in FTIR spectrum of the pristine Fe3O4 could be assigned to stretching mode of Fe–O. The absorption peaks at 1095 and 1210 cm−1 corresponds to Si–O bond, which shows that silica layer was successfully coated on Fe3O4 see Fig. 2. The additional absorption bands appeared at 2920 and 1375 cm−1 in Fe3O4@TODGA were due to C-H stretching and bending modes of the alkyl group, respectively, 1640 cm−1 was assigned to carbonyl group, and 1105 cm−1 assigned to stretching vibration of C–O–C groups present in TODGA see Fig. 2b. The characteristic phosphate bands were observed at 1370 and 960 cm−1 along with C–H stretching vibrations at 2926 and 1060 cm−1 in FTIR spectrum shown in Fig. 2c of the Fe3O4@HDEHP particles.

Fig. 1
figure 1

FE-SEM image of extractant (TODGA) coated Fe3O4 particles

Full size image
Fig. 2
figure 2

FTIR spectra of pristine Fe3O4, Fe3O4@SiO2, Fe3O4@HDEHP and Fe3O4@TODGA nanoparticles respectively

Full size image

The effect of extractants coatings on the superparamagnetic properties of Fe3O4 was studied by using vibrating sample magnetometer (VSM). It is seen from the magnetization curves given in Fig. 3 that the TODGA and HDEHP coatings did not reduce the saturation magnetization to a significant extent with respect to that of the pristine Fe3O4 particles. This seems to suggest that the extractants coating on Fe3O4 was thin as expected from the FE-SEM images.

Fig. 3
figure 3

VSM magnetization curves of pristine Fe3O4, and TODGA/HDEHP coated Fe3O4 particles

Full size image

The amounts of extractants coated on the Fe3O4 particles were determined by thermo gravimetric analysis (TGA) performed under nitrogen atmosphere with the heating rate of 15 °C min−1, and thus obtained thermograms are given in Fig. 4. As can be seen from Fig. 4, 3% weight loss was observed because of adsorbed water in the pristine Fe3O4. For HDEHP and TODGA coated Fe3O4 particles, the initial weight loss about 3 wt% was observed below 200 °C attributed to the loss of water molecules and additional weight losses of 6 wt% from 200 to 600 °C for TODGA coated Fe3O4 and 8% weight loss from 200 to 700 °C for HDEHP coated Fe3O4 were observed. The organic extractants were burned to gases during analysis. Thus, 6 and 8 wt% losses could be correlated to the amounts of extractants coated on the Fe3O4 particles. The higher coating of HDEHP on the Fe3O4 particles could be due to mono acidic phosphate group in HDHEP which has higher affinity towards Fe3+/Fe2+ sites of Fe3O4.

Fig. 4
figure 4

TGA curves of pure Fe3O4 (a), Fe3O4@TODGA (b) and Fe3O4@HDEHP (c) nanoparticles

Full size image

Actinide sorptions and desorptions

The actinide extraction properties of diglycolamides are attributed to their aggregations induced by HNO3 [41–45]. These class of extractants take up tri and tetravalent actinides such as Am(III) and Pu(III,IV) from 3 to 4 mol L−1 HNO3 solutions [46, 47]. To understand effect of HNO3 induced aggregation of TODGA, 0.1 mol L−1 TODGA in dodecane was pre-equilibrated with HNO3 and then coated on the Fe3O4 particles at different temp. As can be seen from distribution coefficients values (D) given in Table 1, the pre-acid treated Fe3O4@TODGA extracted Am3+ quantitatively irrespective of coating temp but no extraction was observed for non-acid treated TODGA coated particles. It was also evident from data given in Table 2 that the Fe3O4@TODGA formed by the coatings of lower concentration of HNO3 acid pre-equilibrated TODGA did not have significant extraction efficiency toward Am3+ ions. This is in accordance with the literature which reported the critical concentration of acid required for 0.1 mol L−1 TODGA in n-alkane to form TODGA reverse micelles was 0.7 M mol L−1 HNO3 [48]. HDEHP exists normally in a dimer form without any pre-treatment [49, 50], therefore was coated as such using conditions described in experimental section. The depicted chemical structure and schematic representation of HDEHP/TODGA coatings is shown in Scheme 2. But in aqueous solution the hydrocarbon chains collapses and coils around magnetite nanoparticles.

Table 1 Extraction behaviour of Fe3O4@TODGA NPs toward 241Am(III) from 3 mol L−1 HNO3 under different coating conditions of 0.1 mol L−1 TODGA in n-dodecane
Full size table
Table 2 Extraction behavior of Fe3O4@TODGA formed by coating TODGA in dodecane pre-equilibrated with different acidity towards 241Am(III) from 3 mol L−1 HNO3
Full size table
Scheme 2
scheme 2

Surface modifications of Fe3O4 with TODGA and HDEHP

Full size image

To study the reproducibility of TODGA coating, four batches of Fe3O4@TODGA were prepared by the same procedure. Fe3O4@TODGA (0.1 g) was equilibrated with 25 mL solution containing 241Am in 3 mol L−1 HNO3 for 3 h under constant shaking at 25 °C. The extraction efficiency was found to quite reproducible as 87 ± 3%. It was also observed that Pu(III,IV) sorption in Fe3O4@TODGA was similar as expected from literature [41–47].

Unlike Fe3O4@TODGA particles, the Pu(IV) ions were selectively adsorbed in the Fe3O4@HDEHP particles at higher HNO3 concentration as shown in Table 3. Thus, it is possible to quantify Pu(IV) selectively in the presence of Am(III) using the Fe3O4@HDEHP particles. The reusability of Fe3O4@TODGA particles was tested for Am(III) ions. First the Fe3O4@TODGA particles were loaded with Am(III) ions from 25 mL of 3 mol L−1 HNO3 for 3 h; and subsequently de-loaded quantitatively (>99%) with 3 mL of 0.02 mol L−1 disodium salt of EDTA. The cycle of extraction and stripping was repeated for three times. It was observed that the percentage of extraction remains same even after three cycles. The same is also true for Fe3O4@HDEHP particles toward Pu(IV) ions. Where Pu was deloaded with 0.02 mol L−1 disodium salt of EDTA and again equilibrated with same amount of plutonium. The re-usability cycles given in Fig. 5. Which indicates the extractant coatings on Fe3O4 particles did not deteriorate after multiple uses.

Table 3 Extraction behaviour of Fe3O4@HDEHP particles toward Pu(IV) and Am(III) ions as a function of HNO3 conc
Full size table
Fig. 5
figure 5

The reusability cycles of Fe3O4@HDEHP particles towards Plutonium

Full size image

Analytical applications

The quantification of 241Am, which emits 59.54 keV γ-rays with 35.9% abundance, was carried out by direct γ-spectrometry of 241Am(III) loaded Fe3O4@TODGA particles. For this, the Fe3O4@TODGA particles (0.01 g) were equilibrated with different amount of 241Am activity (5.7–57 µCi) spiked in the 3 mol L−1 HNO3 and kept under shaking condition for 2 h at room temp. As can be seen from Fig. 6, the gamma activity in Fe3O4@TODGA particles varied linearly as a function of amount of activity spiked. Thus, the direct γ-spectrometry of 241Am(III) loaded Fe3O4@TODGA particles could be used for Am(III) quantification using calibration plot or by the standard comparison method.

Fig. 6
figure 6

Variation in γ-activity of 241Am loaded on Fe3O4@TODGA nanoparticles as a function of Am(III) amount spiked in the solution. Fe3O4@TODGA nanoparticles were equilibrated with different amount of 241Am activity (5.7–57 µCi) spiked in the 3 mol L−1 HNO3 for 2 h under shaking conditions at room temp

Full size image

The isotopes of Pu are weak γ-emitters and, therefore, their quantification with γ-spectrometry at ultra trace conc is not feasible. However, the α-counting using alpha spectrometry, proportional counters and scintillation counters etc. could be used for the quantification of Pu after appropriate sample manipulation [36, 37, 51–53]. In the present work, the Pu(IV)-loaded Fe3O4@TODGA particles were dispersed in a liquid scintillation cocktail and subjected to scintillation counting. It was observed that Fe3O4@TODGA particles did not remain homogeneously distributed and settled at bottom of the vial. This made alpha scintillation counting non-reproducible and lower than that obtained by spiking the same quantity of Pu in the liquid scintillation cocktail. Therefore, the removal of coating in the liquid scintillation cocktail having different solvents was studied. For this, 241Am was used as it is good γ and α emitter which make it possible to corroborate liquid scintillation counting results with γ-spectrometry.

The known quantity 241Am(III)-loaded Fe3O4@TODGA NPs were dispersed in different organic solvents such as toluene, dioxane and isopropyl alcohol. As can be seen from Table 4, there was no decoating of the TODGA organic phase from 241Am(III)-loaded Fe3O4@TODGA after equilibration for 24 h in toluene. However, the quantitative decoating of TODGA bearing organic phase in dioxane as well as isopropyl alcohol was observed. The dioxane decoated solvent was transparent, but white coagulation was observed in isopropyl alcohol. Therefore, dioxane based liquid scintillation cocktail was tested for decoating of organic phase of 241Am(III)-loaded Fe3O4@TODGA and subsequent alpha scintillation counting. It was observed that scintillation counts rate thus obtained was within ±1% of that obtained by direct spiking of the same quantity of 241Am in the same dioxane based liquid scintillation cocktail. There was no residual γ-activity in the decoated Fe3O4, and did not sorb the Am(III) on reuse.

Table 4 Decoating of TODGA bearing organic phase from the 241Am(III) loaded Fe3O4@TODGA in different organic solvents. Am(III) was loaded from 3 mol L−1 HNO3
Full size table

Finally, the fixed amount (0.1 g) of Fe3O4@TODGA and Fe3O4@HDEHP particles were equilibrated with solution containing varying amount of Pu(IV) ions and fixed amount of 241Am at 3 mol L−1 HNO3. The scintillation counts rates obtained after decoating in the dioxane based liquid cocktail varied linearly with the Pu(IV) amount spiked in the equilibrating solutions. The scintillation counts rate obtained by using the Fe3O4@TODGA particles were systematically higher than that obtained using the Fe3O4@HDEHP particles corresponding to fixed amount of 241Am, see Fig. 7. This was attributed to a fact that the Fe3O4@HDEHP particles sorbed Pu(IV) ions only, while Fe3O4@TODGA takes of both Am(III) and Pu(IV) ions.

Fig. 7
figure 7

a Variation in α-scintillation count rate of mixPu loaded on Fe3O4@TODGA nanoparticles as a function of Pu(IV) amount spiked in the solution. Fe3O4@TODGA nanoparticles were equilibrated with different amount of 239Pu activity (20–114 nCi) spiked in the 3 mol L−1 HNO3 for 2 h under shaking conditions at room temp. Pu loaded Fe3O4@TODGA nanoparticles were then dispersed in 5 mL dioxane based scintillation cocktail and then counted. b Variation in α-scintillation count rate obtained by equilibrating Fe3O4@TODGA and Fe3O4@HDEHP particles in solution having varying amount of Pu(IV) and fixed amount of 241Am (scintillation count rate 35,000 counts min−1) under similar conditions described for Fig. 6a

Full size image

Conclusions

TODGA and HDEHP extractants were successfully coated on superparamagnetic Fe3O4 nanoparticles using the optimized chemical conditions. The TODGA coated Fe3O4 nanoparticles showed quantitative uptake of Am3+ and Pu4+ ions. However, HDEHP coated Fe3O4 nanoparticles were found to take up Pu4+ ions only from the solution having 3 mol L−1 HNO3, which are normally encountered in the nuclear reprocessing plants. For liquid scintillation counting, the removal of coating in dioxane based scintillator was found to be suitable for the quantification of preconcentrated actinides on the extractant coated Fe3O4 nanoparticles. This provides a potential application of this method for monitoring the ultra-trace concentration of radioactivity in large volume of aqueous sample discharges from nuclear facilities using magnetically assisted separation followed by quantification by alpha scintillation counting.