Synthesis, Characterization, and Application of Europium(III) Complexes as Luminescent Markers of Banknotes

In this work, three complexes were synthesized from the trivalent europium ion (Eu), using the picrate anion (pic), and delta-valerolactam (DVL), epsilon-caprolactam (EPK), and oenantholactam (OEN). The synthesized complexes [Eu(pic)3∙(DVL)3], [Eu(pic)3∙(EPK)3], and [Eu(pic)3∙(OEN)3] were studied as luminescent markers for application as security elements in Brazilian banknotes. All complexes showed red color emission with absorption at 397 nm and emission at 614 nm. Qualitative luminescence tests were performed on R$10, R$20, R$50, and R$100 Brazilian banknotes. The complexes were applied on the surface of the banknotes and were exposed to different wavelengths of 254, 312, 365, and 320-400 nm. The chemical profiles of the complexes were identified on the banknotes employing the laser desorption ionization mass spectrometry (LDI (±) MS) technique. Generally, tests were promising, and can thus provide a simple, fast, and easy method to identify the authenticity of questioned documents, with an average cost of R$0.65 per mg.


Introduction
Documentoscopy or detection of forgery is the branch of forensic chemistry dealing with the study of questioned documents to verify their authenticity. It has a role in criminalistics, not only to determine the veracity of documents but also to discover the authorship of falsifications and the means that were employed. [1][2][3] The falsification of documents is characterized as a crime of intelligence, and within the field of documentoscopy, banknotes, stamps, national driving licenses, identity cards, vehicle registration certificates, vehicle licensing certificates, credit cards, and checks stand out as primary targets of frauds. 4 Within this context, the use of different analytical techniques is reported in the literature to determinate authenticity, such as gas chromatography coupled to mass spectrometry (GC-MS) and Fourier transform infrared (FTIR) spectroscopy to determine the ages of documents, 5,6 video spectral comparator (VSC) to establish trace release order, 7 and ultraviolet-visible (UV-Vis) spectroscopy for the differentiation of italic letters. 8 The advancement of digital technologies in copying equipment, printing, and image processing has facilitated the falsification of documents, and the quality of falsified documents can be practically indistinguishable from the original document. 9 Therefore, the development of efficient and rapid analytical techniques to determine authenticity is necessary. Typically, the techniques used include atomic force microscopy, 10 FTIR spectroscopy, 11 Raman spectroscopy, [12][13][14] GC-MS, 15 ESI MS (electrospray ionization mass spectrometry), 16 LDI MS (laser desorption ionization mass spectrometry), [17][18][19] and ambient ionization mass spectrometry techniques EASI MS (easy ambient sonic-spray ionization mass spectrometry), PSI MS (paper spray ionization mass spectrometry), DESI MS (desorption electrospray ionization mass spectrometry), and DART MS (direct analysis in real time mass spectrometry). [20][21][22][23][24] Among the types of forgery employed, the counterfeiting of banknotes deserves particular attention, since this is an increasingly common financial crime, both in the sense of increasing the number of falsified records and in the diversity of the falsification methods used by the falsifiers. 25,26 In addition, this type of crime can affect the economy of several countries. Therefore, the use of efficient security items such as security papers, latent images, watermarks, magnetic strips, special printing techniques, holograms, and areas with infrared (IR) or ultraviolet (UV) light responses has been adopted as a form of fingerprinting for authentic banknotes. 27 It is of great importance to use security items that are easily and quickly identified by the population to prevent counterfeiting of banknotes. For this purpose, the use of new photoluminescent materials that emit light in the presence of electromagnetic radiation in the IR or UV region gains prominence because these materials meet the requirement of simple and rapid identification. New materials incorporating lanthanides 28-37 may be potential photoluminescent markers for the authenticity of banknotes.
The luminescent property of lanthanide ions mainly originates from transitions involving partially occupied 4f orbitals. The luminescence phenomenon using direct excitation of the lanthanide ion is inefficient because it does not have high molar absorptivity. The light is absorbed by an organic molecule (ligand), which transfers energy to the lanthanide ion, which then emits luminescence resulting from intra-configuration transitions of orbital 4f and usually observed in the visible region. Hence, there is an intra-molecular energy transfer from the ligand to the metal ion known as the "antenna effect". 38,39 This luminescent ability of the lanthanide complexes, to absorb and emit radiation with characteristic wavelengths, defines these materials as light conversion molecular devices (DMCLs). The efficiency of the transfer of energy from the binder to the lanthanide ion depends on the chemical nature of the coordinated ligand. 40 Among the lanthanide(III) ions, the elements europium, terbium, and thulium emit in the regions of visible red, green, and blue light, respectively. 41,42 The luminescent properties of materials derived from these metals have broad applicability in several areas, including their use as bio-detectors, 43,44 films, 45 solar cells, 46 and organic light-emitting diode (OLED), 47 and, in the forensic field, in the detection of explosives, 28 fingerprints, 48,49 and gunshot residues. 50,51 A series of lanthanide(III) picrate (pic) complexes with different organic molecules as coligands are reported in the literature. Silva et al. 52  This work reports the synthesis, characterization, and application of three new europium(III) picrate complexes, using three different lactams (delta-valerolactam, epsiloncaprolactam, and oenantholactam) as coligands, for luminescent markers of Brazilian R$10, R$20, R$50, and R$100 banknotes.
The chemical stoichiometry of the complexes was determined through complexometric titration with 0.01 mol L -1 EDTA standard solution 58 and elemental analysis using a Thermo Fisher Scientific Flash 1112-CHNS-O (Waltham, Massachusetts, USA). The infrared spectra were obtained in transmittance mode with an attenuated total reflectance accessory (ATR) in the region of 4000-650 cm -1 , using a PerkinElmer FTIR Spectrum 400 MID/NIR spectrometer (Waltham, Massachusetts, USA) at room temperature. UV-Vis region spectroscopy analyses were performed for the solid-state complexes using a PerkinElmer spectrometer (Waltham, Massachusetts, USA), in the range of 220-800 nm. Excitation spectra were obtained at room temperature in the range of 250-550 nm, with a slot opening of 0.75, monitoring the intensity of the 5 D 0 → 7 F 2 transition at 616 nm. Emission spectra were obtained in the range of 550-750 nm at room temperature, with a slot opening of 0.75 and excitation at 397 nm, using a Quanta Master 40 spectrofluorometer (Edison, New Jersey, USA) with a 75 W xenon lamp. Determinations of the exact masses of the complexes were made by laser desorption ionization mass spectrometry in both ionization modes, LDI (±) MS, using an FT-ICR model 9.4 T Solarix mass spectrometer, Bruker Daltonics (Bremen, Germany), equipped with a Smartbeam-II TM (355 nm) laser. LDI (±) MS data were acquired with 16 scans with a frequency of 200 Hz in the range of m/z 200-1200, using 100 laser shots per pixel, a small (ca. 30 µm) laser focus setting, and laser power ranging from 13 to 15%.

Synthesis of hydrated europium(III) picrate [Eu(pic) 3 •(H 2 O) 11 ]
Initially, hydrated basic carbonate of europium(III), EuCO 3 (OH)•xH 2 O, was obtained from Eu 2 O 3 (5.0 g) and concentrated hydrochloric acid was added to the oxide suspension in water (800 mL). The solution was heated (85-90 °C) and urea was added until the solution reached a pH of approximately 7. Basic carbonate hydrate of europium(III) with a yield of 91% was obtained as a product, and 2.0 g of this compound was then suspended in an aqueous medium (100 mL). This solution was heated (85-90 °C) and picric acid was added until all basic carbonate was consumed. The resulting solution was filtered and allowed to stand at room temperature for crystallization. The obtained crystals had yellow coloration and were dried at room temperature and stored in an amber bottle. 59  ν ass NO 2 , 1335 (s) ν s NO 2 , 797 (m) γCH, for free picric acid ( Figure 1a) 1526 (s) ν ass NO 2 , 1539 (s) ν s NO 2 , 782 (m) γCH.

Synthesis of complexes
The compounds were prepared by dissolving the hydrated europium(III) picrate in ethanol with an ethanolic solution of the lactam (molar ratio 1:3, lactam = DVL, EPK, and OEN). Triethyl orthoformate, which is used as a dehydrating agent, was then added, contributing to the removal of hydration waters from the europium(III) picrate. The system was stirred until a yellow solid appeared. The solid obtained was washed with ethyl ether, dried at room temperature, and stored in an amber bottle. 60 The results of experimental and calculated elemental analysis and FTIR analysis of the complexes were as follows: Yield 63%; anal. calcd. (%) for C 33 (4), ν ass NO 2 (7), ν as NO 2  The syntheses of the three complexes are represented schematically in Figure 2.

Application of complexes in Brazilian banknotes
A total of six Brazilian banknotes were used: one each in the values of R$10, R$20, and R$100, and three R$50 notes. The notes were purchased from a Brazilian bank. The R$10 and R$20 banknotes were chosen because they are more commonly used; the R$50 and R$100 banknotes were chosen because they have greater value and are major targets of counterfeiting. For the R$50 banknotes, approximately 1 mg of each complex was applied using a swab and deposited on the surface of the banknotes at the top of the region that contained the value of each banknote. Later, the complex that showed the best photoluminescence (in this case the [Eu(pic) 3

Synthesis
Hydrated basic carbonates of europium(III) were prepared by precipitation from homogeneous solutions via the hydrolysis of urea without the addition of an auxiliary anion, 61 allowing the formation of hydrated europium(III) picrate by the direct reaction of hydrated basic carbonate of europium(III) with picric acid. This process obtained a high yield and a non-hygroscopic compound.
The stoichiometry of the complexes was obtained through elemental analysis (CHN), which confirmed the 1:3 molar ratio ([Eu(pic) 3 •(H 2 O) 11 ]:lactam). FTIR spectra (Figure 1) identified the presence of asymmetric stretching frequencies (ν ass NO 2 (region 7)) and symmetrical frequencies (ν s NO 2 (region 8)) for the picrate ion, demonstrating the coordination of this ion to the metallic center of the europium(III). The split and shifted ν s NO 2 vibration suggested that the picrate ions are coordinated to the metallic center in a bidentate form, through the phenolic oxygen atom and the o-nitro group oxygen atom. The disappearance of the out-of-plane vibration of the OH (region 10) group at 1151 cm -1 , corresponding to free picric acid, indicates that hydrogen has been replaced by the Eu III ion and that phenolic oxygen coordination to Eu III has occurred. The νC-O (region 9) vibration shift from 1260 to 1274 cm -1 suggests that the substitution of phenolic hydrogen (OH) by Eu III increases the π character of the C-O bond. [62][63][64] In addition, the νC=O (regions 2, 4, and 6) vibration shift for lower energy regions and decreased intensity of νNH (regions 1, 3, and 5) stretching of lactams, concerning the free ligand, suggests the coordination of carbonyl to the Eu III ion. 65

Emission and excitation analysis
The UV-Vis absorption spectra of [Eu(pic) 3 •(DVL) 3 ] in Figure S1a (Supplementary Information (SI) section) and [Eu(pic) 3 •(EPK) 3 ] complexes in Figure S1c (SI section) show bands with maximum absorption at 323 and 324 nm, respectively, which can be attributed to intra-ligand π→π*type electronic transitions, and a band with maximum intensity at 391 nm, which was observed for both, attributed to n→π*-type transitions. For the [Eu(pic) 3 •(OEN) 3 ] complex in Figure S1b (SI section), a single broad band with a maximum intensity at 361 nm was observed and can be attributed to the overlapping π→π* and n→π* transitions. In addition, a low-intensity band around 740 nm was observed for all complexes and is assigned to the f-f transitions of the 7 F 5 → 5 D 0 type for the Eu III ion. 66 The excitation spectra in Figure S2 (SI section) presented a broad band between 250-500 nm, which can be attributed to the central ligand (maximum at 397 nm) and characterized as the transition band of the ligands (S 0 → S 1 ). The excitation at 397 nm is very close to the 5 L 6 level of the Eu III (394 nm). The excitation spectra exhibit a broad band between 250-500 nm, overlapping the 5 L 6 level. Therefore, the 5 L 6 level can also be an efficient channel for the photoluminescence of the compounds, although lanthanide ions have low absorption coefficients.
The emission spectra in Figure 3 of the compounds at room temperature showed the regions of the f-f transitions from the excited level 5 D 0 → 7 F J (J = 0, 1, 2, 3, and 4): 5 D 0 → 7 F 0 (576 nm), 5 D 0 → 7 F 1 (590 nm), 5 D 0 → 7 F 2 (614 nm), 5 D 0 → 7 F 3 (650 nm), and 5 D 0 → 7 F 4 (696 nm), with maximum emission at 614 nm. The band corresponding to the 5 D 0 → 7 F 0 transition was observed to have weak intensity. The presence of this band suggests that the Eu 3+ ion is involved in a low-symmetry chemical environment, which may be of C n , C s , or C nv type. 67 The 5 D 0 → 7 F 2 transition is a hypersensitive transition with a predominantly electric dipole character, and when its intensity is higher than that of the 5 D 0 → 7 F 1 transition, this indicates that the compounds have no inversion center. The low intensity of the 5 D 0 → 7 F 0 transition associated with the intensity presented by the 5 D 0 → 7 F 4 transition can be interpreted in terms of the symmetry of the Eu III ion chemical environment, suggesting that it is in a highly polarizable chemical environment with distorted symmetry. 68 The experimental intensity parameters Ω λ (λ = 2 and 4) in Table 1 were determined from the emission spectra for the compounds based on the 5 D 0 → 7 F 2 and 5 D 0 → 7 F 4 transitions, with the 5 D 0 → 7 F 1 magnetic-dipoleallowed transition as the reference because this transition is practically insensitive to changes in the chemical environment. They are estimated according to equation 1: 69 (1) where e is the electronic charge, ω is the angular frequency of the transition, ћ is Planck's constant divided by 2π, c is the velocity of light, and χ = n(n 2 + 2) 2 /9 is the Lorentz local field correction term with a refraction index n of 1.5. The squared reduced matrix elements are 〈 5 D 0 ||U (2) || 7 F 2 〉 2 = 0.0032 and 〈 5 D 0 ||U (4) || 7 F 4 〉 2 = 0.0023. 70,71 The Ω 6 parameter was not determined because the 5 D 0 → 7 F 6 transition could not be experimentally detected. It occurs in the near IR (ca. 840 nm) and thus beyond the detection range of the experimental setup. The Ω 2 and Ω 4 intensity parameters for the complexes are presented in Table 1. In this case, the Einstein coefficient values of spontaneous emission between 5 D 0 → 7 F J (A 0λ ) are obtained using equation 2: where S 0λ is the area under the emission curve related to the 5 D 0 → 7 F J transitions obtained from the spectral data, and ν 0λ is the energy barycenter of the transitions from the 5 D 0 excited state to the 7 F 1 , 7 F 2 , and 7 F 4 ground states (in cm -1 ). The 5 D 0 → 7 F 1 magnetic-dipole-allowed transition is almost independent of the crystal field environment around the Eu III ion (the A 01 value is estimated to be approximately 50 s -1 ). 72 The experimental intensity parameter Ω 2 shows different values for the compounds in Table 1, indicating that the Eu III ions are in different chemical environments and that a highly polarizable chemical environment exists around the Eu III . According to the literature, 73 the Ω 2 value is most influenced by small angular changes in the local geometry. This effect, together with changes in the polarizability of the ligating atom (α), has been used to explain the hypersensitivity of certain 4f-4f transitions to changes in the chemical environment. Borges et al. 74 reported a new anionic complex containing 1-ethyl-3-methylimidazolium (EMIm) with the composition (EMIm) 2 [Eu(Pic) 4 •(H 2 O) 2 ]Pic and found values for Ω 2 (16.7 × 10 -20 cm 2 ) and Ω 4 (7.7 × 10 -20 cm 2 ). This is consistent with the values found for Ω 2 and Ω 4 in the present study for the compounds. The red color of the synthesized compounds was verified using the chromaticity diagram in Figure 3 Figure 3h. The values found for the chromaticity coordinates are in agreement with the standard values for red luminophores (x = 0.64; y = 0.33). 75,76 Photoluminescence analyses were qualitatively performed on the banknotes to visually assess the light emission of each complex present on the R$50 banknotes. Figure 4 shows the light emission of each complex under excitation at different wavelengths.
The emission of red light, corresponding to each complex present in the banknotes, under excitation at different wavelengths (254, 312, 365, and 320-400 nm (Lumatec Spritelite 400)) was easily visualized. These excitation wavelengths were visually evaluated by observation of the emitted light from the complexes. At 365 and 320-400 nm, which are wavelengths close to that obtained in the excitation spectra (397 nm), the observed red light emission was more intense than that at the 254 and 312 nm wavelengths. In addition to the luminescence from the complexes added to the banknote surfaces, Figures 4i-4k also show a region of emitted green light, which is attributable to one of the security elements already present in the banknote itself.
While all the complexes produced light emissions that allowed their visual identification on banknotes, the [Eu(pic) 3 •(EPK) 3 ] complex was selected and applied on the surface of the R$10, R$20, and R$100 banknotes. Figure 5 shows the red light emission of the [Eu(pic) 3 •(EPK) 3 ] complex under excitation at different wavelengths on the R$10, R$20, and R$100 banknotes.
Based on the results in Figure 5, it was observed that the luminescence of the [Eu(pic) 3   noticeable at different wavelengths. The 365 nm wavelength emits more intense luminescence because it is closer to the maximum excitation wavelength (397 nm) of the [Eu(pic) 3 •(EPK) 3 ] complex.

LDI (±) MS
To identify the chemical profile of the complexes deposited on the banknote surfaces while preserving their integrity, the LDI (±) MS technique was used, in which an analysis is made from the incidence of a laser beam (which may be pulsed ultraviolet or IR) focusing on a surface, allowing analyte desorption and ionization without the necessity of previous sample preparation. 77 Figure 6a shows the chemical profile of the banknote with no complexes applied. It was observed that an ion of m/z 575.08132 is present. This ion is found abundantly in different regions of the R$50 banknotes before application of the complexes and thus can be characterized as a natural chemical marker to recognize the authenticity of banknotes. Eberlin et al. 22 reported an analysis of Brazilian banknotes using DESI MS and EASI MS, identifying ions of m/z 391, 413, 429, 803 and 819 as natural markers that characterize the chemical profile of the authentic banknotes. Schmidt et al. 81 also reported the study of a second family of real banknotes using EASI (+) MS, where the ion of m/z 443 was used as a fingerprint of authenticity.
LDI (+) MS spectra in Figures 6a-6d were obtained with mass resolution m/Δm 50% ca. 381783, where Δm 50% is the full peak width at half-maximum peak height and m/z ca. 400, 82 and showed mass errors ranging from 0.76 to 4.64 ppm as shown in Table 2.   83 For the LDI (±) mass spectra of the banknotes marked with complexes, the signal of the ion of m/z 575.08132 (unmarked banknote) has been suppressed by signals coming from the complexes.

Conclusions
In this study, the [Eu(pic) 3 3 ] complexes were shown as new potential luminescent security materials for application in questioned documents. The identification of the complexes on the banknotes can be easily performed because wavelengths in the UV region can be used to produce light emission by the complexes. It was also possible to identify the complexes present on the surface of the banknotes using the non-destructive LDI (±) MS technique. Finally, the average cost for the synthesis of the complexes was R$0.65 per mg.