Faster Synthesis of Beta-Diketonate Ternary Europium Complexes: Elapsed Times & Reaction Yields

β-diketonates are customary bidentate ligands in highly luminescent ternary europium complexes, such as Eu(β-diketonate)3(L)2, where L stands for a nonionic ligand. Usually, the syntheses of these complexes start by adding, to an europium salt such as EuCl3(H2O)6, three equivalents of β-diketonate ligands to form the complexes Eu(β-diketonate)3(H2O)2. The nonionic ligands are subsequently added to form the target complexes Eu(β-diketonate)3(L)2. However, the Eu(β-diketonate)3(H2O)2 intermediates are frequently both difficult and slow to purify by recrystallization, a step which usually takes a long time, varying from days to several weeks, depending on the chosen β-diketonate. In this article, we advance a novel synthetic technique which does not use Eu(β-diketonate)3(H2O)2 as an intermediate. Instead, we start by adding 4 equivalents of a monodentate nonionic ligand L straight to EuCl3(H2O)6 to form a new intermediate: EuCl3(L)4(H2O)n, with n being either 3 or 4. The advantage is that these intermediates can now be easily, quickly, and efficiently purified. The β-diketonates are then carefully added to this intermediate to form the target complexes Eu(β-diketonate)3(L)2. For the cases studied, the 20-day average elapsed time reduced to 10 days for the faster synthesis, together with an improvement in the overall yield from 42% to 69%.


Introduction
Luminescence of europium complexes has a wide range of applications, such as pH sensing based on a modified nanoscale MOF [1]; in the determination of enterobacter cloacae labeled DNA probe based on time-resolved fluorescence [2]; as redox-active magnetic resonance imaging (MRI) sensors [3]; in the determination of DNA using the gatifloxacin europium (III) complex [4]; as sensor to visualize latent fingerprints [5]; as emissive nano-confined system for highly-luminescent bioimaging in vivo [6], etc.
In particular, β-diketonate ligands have proved to be efficient antennae leading to highly efficient complexes, with relevant and unique applications such as in emitting layers to fabricate white organic light-emitting diodes (WOLED) [7]; as triple-layer organic light emitting diodes (OLED) in UV portable dosimeter [8]; as red phosphors applied in light-emitting diodes (LED) [9][10][11][12]; as turn-on sensors for detecting basic molecules hosted in nanozeolite [13], as surfactant functionalized in polymer for luminescence integration [14], as highly sensitive temperature sensors [15], as two-photon sensitized luminescent probes for applications in biological imaging [16], etc.
Beta-diketonate lanthanide complexes are known since the end of the 19 th century when they were first synthesized and reported by Urbain and Budischovsky in 1897 [17]. Rare-earth beta-diketonates have been extensively reviewed by Binnemans, both as complexes [18] and, more especially so, as lanthanide-based organic-inorganic hybrid materials [19,20].
There are three main types of lanthanide beta-diketonate complexes: tris complexes, Lewis base adducts of the tris complexes which are the ternary lanthanide beta-diketonates and tetrakis complexes.
Synthesis of ternary europium β-diketonate complexes are usually carried out by first reacting β-diketonates with either hydrated chloride or hydrated nitrate salts. Unlike chloride ions, which are a good leaving group, nitrate ions coordinate more strongly, in a bidentate manner, which often leads to final complexes containing a coordinating nitrate group [18]. Therefore, most of the europium β-diketonate complexes are prepared using EuCl 3 (H 2 O) 6 , to yield Eu(βdiketonate) 3 (H 2 O) 2 [18]. Of course, the presence of the coordinated water molecules tends to quench the luminescence due to non-radiative processes, as expected from the presence of near resonance oscillators. Replacing the water molecules with other non-ionic ligands tends to intensify the quantum yield, rendering the europium complexes highly luminescent. Accordingly, the water molecules are subsequently displaced by other non-ionic ligands, such as sulfoxides, phosphine oxides, nitrogen-containing polycyclic aromatic compounds, etc, to obtain the target complexes.
However, the step involved in the purification through recrystallization of the intermediate Eu(β-diketonate) 3 (H 2 O) 2 , which sometimes is obtained in the form of a viscous oil [18], is normally too time consuming. Indeed the recrystallization of only a few milligrams may take from days to several weeks depending on the β-diketonate involved, more especially so for the case of complexes of highly fluorinated ligands [18]. Recently, β-diketonate lanthanide complexes were prepared by mixing a lanthanide salt LnX 3 (H 2 O) 6 with non-ionic nitrogen coordinating ligands, L, in methanol, to first obtain an intermediate of the generic formula Ln (L) l X 3 (CH 3 OH) m , where Ln may be Eu 3+ , Sm 3+ or Gd 3+ ; l may be either 1 or 2; X may be Clor NO 3 -; and m may be either 0 or 2. [21]. Subsequently, a methanol solution was added containing 1,1,1-trifluoro-2,4-pentanedione, acacF 3 H, and sodium hydroxide to a methanol solution of the intermediate complex to obtain the final product Ln(L) l X n (CH 3 OH) m (acacF 3 ) 3-n where n may be 0 or 1 [21]. These intermediates and final products are insoluble in methanol, a property that facilitates their purification. Therefore, in both steps of the syntheses, the respective intermediates and final products precipitate, are filtered out and subsequently purified by recrystallization in 2-propanol. However, acacF 3 H complexes of europium tend to lead to very small quantum yields, ranging from 0.28% to 1.5% [21]. On the other hand, ligands that lead to highly luminescent complexes are usually soluble in methanol or ethanol, a property that makes their purification more difficult. Much more arduous and critically so, is the case of the intermediate Eu(β-diketonate) 3 (H 2 O) 2 , whose purification may take up months.
In this article, we advance a strategy to considerably shorten the time involved in the overall synthesis of known highly luminescent β-diketonate europium complexes, whose photophysical properties have already been reported [22][23][24][25][26], by first adding the non-ionic ligand to the original salt, EuCl 3 (H 2 O) 6 [27], to obtain the intermediates EuCl 3 (L) 4 (H 2 O) n , which are soluble in solvents such as methanol or ethanol. Subsequently, the β-diketonate ligands are carefully added to the intermediate to obtain the final product Eu(β-diketonate) 3 Hence, a process, which usually takes a few weeks, can now be carried out in days. Finally, when compared to the usual synthesis, we show that the overall synthetic yield of this faster synthesis of europium β-diketonate complexes is significantly larger.

Faster Synthesis
Intermediate complex EuCl 3 (L) 4 (H 2 O) n . A solution of EuCl 3 (H 2 O) 6 (0.37 g, 1 mmol) in 250 mL of pure ethanol was prepared under stirring conditions. To this solution, we slowly added the non-ionic ligand L, DBSO (0.92 g, 4 mmol), PTSO (0.92 g, 4 mmol) or TPPO (1.11 g, 4 mmol), which was previously dissolved in 50 mL of ethanol. Then the system was left overnight being stirred under reflux at 78°C. Subsequently, the solvent was rotaevaporated. A solid of EuCl 3 (L) 4 (H 2 O) n (white for L = TPPO and PTSO, and yellow for L = DBSO) was obtained. The complex was recrystallized with a water/EtOH solution (10:1), the supernatant was removed with a Pasteur pipette, and the resulting solid was dried under vacuum for 24h. Yields: 89% for EuCl 3 (TPPO) 4 4 (0.25g, 0.2 mmol), in 30 mL of ethanol was prepared and, at the same time, the β-diketonateK, TTAK (0.16 g, 0.6 mmol) or DBMK (0.16 g, 0.6 mmol) salt, was prepared by dissolving β-diketonateH, TTAH (0.13 g, 0.6 mmol) or DBMH (0.13 g, 0.6 mmol), in ethanol separately and by slowly adding potassium hydroxide (0.03 g, 0.6 mmol) in ethanol/water solution (10:1) under stirring conditions. Once formed, the β-diketonateK ethanolic solution was added, drop by drop, over the EuCl 3 (L) 4 (H 2 O) n (n = 3 for L = TPPO and n = 4 for L = PTSO or DBSO) solution. The pH was checked and, if it was still acidic, a few more drops of the ethanolic solution of KOH were added until the pH stabilized around 6.0~6.5. After at least 30 minutes of stirring, a white precipitate of potassium chloride salt was formed, which was subsequently removed by filtration. If small amounts of the yellow complex were present in the filtered salt, they were washed out with~15mL of pure ethanol. The filtrate is a yellow solution of the complex and reactants. Next, the filtrate was stirred under reflux at 78°C for 24 h to complete the reaction until a yellow solution of Eu(β-diketonate) 3 (L) 2 is obtained. Then, the solvent was allowed to slowly evaporate at room temperature. A yellow solid of Eu(β-diketonate) 3 (L) 2 was obtained, which could still contain some potassium chloride. That is why it is important to wash the solid with~30mL of water to remove the remaining salt. In sequence, it is necessary to use boiling hexane to remove the released L ligands. Finally, the complex was recrystallized with hexane/acetone solution (10:1). The supernatant was removed with a Pasteur pipette, and the solid was dried under vacuum for 24h. Yields: 88% for Eu(TTA) 3 (TPPO) 2 , 86% for Eu(DBM) 3 (TPPO) 2 , 70% for Eu(TTA) 3 (DBSO) 2 , 68% for Eu(DBM) 3 (DBSO) 2 , 75% for Eu(TTA) 3 (PTSO) 2 , and 80% for Eu(DBM) 3 (PTSO) 2 .

Usual Synthesis
Intermediate complex Eu(β-diketonate) 3 6 (0.37 g, 1 mmol) in 250 mL of pure ethanol was prepared under stirring conditions. At the same time, the β-diketonateK, TTAK (0.78 g, 3mmol) or DBMK (0.79 g, 3mmol) salt, were prepared by dissolving 3 mmol β-diketonateH, TTAH (0.67 g, 3mmol) or DBMH (0.67 g, 3mmol), in ethanol and by slowly adding of potassium hydroxide (0.17 g, 3mmol) in ethanol/water solution (10:1). Once formed, the DBMK ethanolic solution was added, drop by drop, to the EuCl 3 (H 2 O) 6 solution. The pH was checked and, if it was still acidic, a few more drops of the ethanolic solution of KOH were added until the pH stabilized around 6.0~6.5. After at least 30 minutes of stirring, a white precipitate of potassium chloride salt was formed, which was subsequently removed by filtration. If small amounts of the yellow complex were present in the filtered salt, they were washed out with~15mL of pure ethanol. The filtrate is a yellow solution of the complex and reactants. Then, the filtrate stirred under reflux at 78°C for 24 h to complete the reaction to obtain a solution of Eu(β-diketonate) 3 (H 2 O) 2 . Subsequently, the solvent was allowed to slowly evaporate at room temperature. A yellow solid of Eu(β-diketonate) 3 (H 2 O) 2 was then obtained, which can still contain some potassium chloride. That is why washing the solid with~30mL of water, was necessary. Finally, the complex was recrystallized with a hexane/acetone solution (10:1), the supernatant was removed with a Pasteur pipette, and the solid was dried under vacuum for 24h. Yields: 73% for Eu(TTA) 3 (H 2 O) 2 , and 60% for Eu (DBM) 3

Characterizations
All synthesized complexes, both intermediate and target compounds, have been characterized by elemental analysis, (Perkin-Elmer CHN 2400), infrared spectroscopy (samples of the complexes were prepared as KBr disks and the spectra were measured in a Bruker model IFS 66 spectrophotometer, 4000 cm -1 -400 cm -1 ), and by 1 H NMR, 19 F NMR, and 31 P NMR (all NMR spectra of all complexes were obtained in CDCl 3 solutions via a Varian Unity Plus 400 MHz, tuning the probe to each of the nuclei studied). The novel intermediate complexes were further characterized by MALDI-TOF mass spectroscopy taken on an Autoflex 3 Smart Beam Vertical spectrometer by Brucker Daltonics, with α-cyano-4-hydroxycinnamic acid as the matrix.
The Supporting Information presents all details of the experimental characterizations, data, spectra and all relevant spectral attributions for all complexes synthesized in this article.
Our faster synthesis involves the following sequence of reactions: The first step of the faster synthesis, reaction Eq (1) involves a rotary evaporation of the final ethanolic solution, is very straightforward, and can be carried out in one day. The resulting solution is subsequently dried in a vacuum desiccator for another day. This procedure has been carried out in all three cases studied. It is noteworthy that water is the only product of reaction Eq (1), and therefore its separation with the help of a vacuum desiccator is forthright. Purification of these intermediates is very simple, taking different times depending on the nonionic ligand, from two days for TPPO to less than one week for DBSO and PTSO, as described in the Experimental section.
The second step of the faster synthesis is also very straightforward. A side solution of β-diketone, stoichiometrically deprotonated by adding potassium hydroxide, is prepared, and then added to a solution of the intermediate complex EuCl 3 (L) 4 (H 2 O) n as in reaction Eq (2). Finally, the pH of the reaction medium was adjusted to 6.5 by carefully adding a warm diluted potassium hydroxide ethanolic solution. Despite the fact that this step leads to three other by products (KCl, L, and H 2 O), these can be easily removed. Indeed, when the reaction completes, the  solution is transferred to a beaker, and the solvent (ethanol) is allowed to completely evaporate. Water is then added, and the supernatant with the dissolved KCl, is subsequently removed by means of a Pasteur pipette. The remaining solid precipitate, in the beaker, is then washed thrice with boiling hexane, to dissolve the unbound non-ionic ligands L. The supernatant is once again carefully removed by means of a Pasteur pipette. The purified wet solid in the beaker is finally dried out in a vacuum desiccator to remove the final traces of hexane and water that may have been left out.
In order to prove that our faster synthesis is indeed superior to the usual one, exactly the same target complexes Eu(β-diketonate) 3 (L) 2 were also synthesized by the usual route, according to Fig 4. As such, we carried out the usual synthesis of complexes of the type Eu(β-diketonate) 3 (L) 2 according to the following sequence of reactions: The intermediates in the usual synthesis, Eu(β-diketonate) 3 (H 2 O) 2 (step 3), are soluble in ethanol and insoluble in water. On the other hand, the formed salt (in our case KCl) is insoluble in ethanol and soluble in water. Since ethanol is the solvent used in the first step of the usual synthesis, it is important to keep filtrating the formed salt as many times as possible during the reaction. At this point, the precipitated salt is contaminated with the formed complex, and thus it luminesces. To decontaminate the salt and recover the remnants of the formed complex, the salt is washed three times with ethanol. The ethanolic filtrate is then poured into the reaction flask, which is under reflux. After 24h the reaction is completed and the solvent (ethanol) is then allowed to evaporate. When the β-diketonate DBM is used, sometimes an even more meaningful amount of the intermediate product Eu(DBM) 3 (H 2 O) 2 precipitates with the salt while the remainder stays in solution. Then, this solid must be washed with water to remove the remaining salt, keeping only the intermediate product. What we found out is that if too much salt is left out mixed with the intermediate, when water is added, a white colloidal suspension tends to form, hindering the separation, sometimes to the point of making it impossible to be accomplished.
The second step of the usual synthesis, reaction 4, is carried out as usual, with the two coordinating water molecules being replaced by two non-ionic ligands L. The solvent is allowed to evaporate, and the remaining solid is recrystallized in a mixture of hexane 10:1 acetone. The resulting purified solid is dried out in vacuum.
Thus, having performed the synthesis of all these 6 complexes, via both the faster synthesis and via the usual route, we are now in a position to compare them, both in terms of elapsed times involved as well as in terms of overall yields. Table 1 shows the elapsed days involved in both the usual and faster syntheses. The second step in both usual and faster syntheses lead, of course, to the same final product with the same degree of difficulty in purifying. The reason the faster synthesis is much rapid is due to its first step, which is the displacement reaction of water by the non-ionic ligand L, leading to the easy to purify intermediate EuCl 3 (L) 4 (H 2 O) n , a procedure which takes, on average, 3 days. Indeed, 3 days is an interval more than 4 times shorter than the one required by the first step of the usual synthesis. And the first step of the usual synthesis, which is the displacement of the chloride ions by the β-diketonate ions, leads to difficult purifications of the intermediates Eu(β-diketonate) 3 (H 2 O) 2 , a procedure which takes an average of 14 days. Therefore, in this step, the faster synthesis represents a reduction of about 80% of the elapsed time needed to obtain the corresponding pure intermediate.
Being fast is more powerful when combined with greater efficacy. Thus, we also examined the reaction yields of both steps of the usual and faster syntheses. Results are shown in Table 2 where, clearly, the faster synthesis displays higher yields when compared to the usual one. Indeed, in average, the usual synthesis leads to an overall yield of 42%, whereas the faster one leads to 69%.  Besides, with the sole exception of step 2 of the synthesis of complex Eu(TTA) 3 (DBSO) 2 , each one of the steps of the faster syntheses lead to higher yields when compared to any of the steps of the usual synthesis.
We now turn to quantify the statistical significance of this finding by verifying whether the mean yield of the faster synthesis, 69%, is truly larger than the mean yield of the usual synthesis, 42%. The t-statistic for both sets of data is 4.65, which, for 5 degrees of freedom, does indicate that the mean for the faster synthesis is indeed larger than the mean for the usual synthesis within a 99.7% confidence level. This result strengthens the fact that the faster synthesis is not only speedier but that it is also overall much more effective.

Characterization
Elemental Analysis and MALDI-TOF/MS. Tables 3 and 4 show the results of the elemental analysis carried out for all synthesized compounds, both intermediate and target complexes. By examining Table 4, the mean absolute error for %C of the 6 complexes synthesized by the usual route is 0.14 amu whereas the equivalent value for the same complexes synthesized by the fast route is 0.09 amu. The equivalent numbers for %H for both sets of complexes are 0.08 amu and 0.11 amu, indicating they are comparable numbers, within the usual accepted range of purity, implying they refer to the same compounds. Table 5 shows MALDI-TOF data, together with elemental analysis results for the novel intermediates obtained by the faster synthesis of europium β-diketonate complexes advanced in this article.    4 , and of the target complex Eu(DBM) 3 (DBSO) 2 , both prepared by the faster synthesis, the corresponding chemical shifts of the CH 2 group are: δ 3.90 ppm and δ 4.72 ppm, respectively. Table 6 shows the 1 H chemical shifts, δ, of the methine group of the β-diketonate ligands coordinated to the europium ion for all synthesized complexes. Clearly, the average chemical shifts of the methine hydrogen of the DBM ligands are unshielded to an average of δ 16.85 ppm. However, the average chemical shifts of the methine hydrogen for TTA is δ 10.24 ppm. Apparently, this is due to the electron withdrawing effect of the single CF 3 group of TTA. We propose that this CF 3 group, being electronegative, pulls the electrons from the Faster Synthesis of Beta-Diketonate Ternary Europium Complexes carbonyl oxygens-more distinctly so from its closest carbonyl oxygen-weakening the coordination of the TTA ligand to the europium ion, leading to a shielding on the methine hydrogen. On the other hand, DBM, which contains two identical aromatic phenyl groups, is more capable of balancing the coordinating bond strengths of its two carbonyl oxygens to the europium ion in a more efficient manner, leading to a more pronounced unshielding of the methine hydrogen. 19 F NMR Spectroscopy. The presence of the TTA ligand in the structure of the complexes can be confirmed by 19 F NMR spectroscopy. For example, while the δ 19 F in the TTA free ligand is -75.80 ppm, for the three TTA containing complexes synthesized in this article, the observed signals are: δ -79.97 ppm for Eu(TTA) 3 (TPPO) 2 , δ -80.38 ppm for Eu(TTA) 3 (DBSO) 2 , and δ -80.51 ppm for Eu(TTA) 3 (PTSO) 2 . Since the compounds are pure, as indicated by elemental analysis, the very few other low intensity and broadened signals in the 19 3 , the phosphorus signal appears at δ 30 ppm, albeit broadened by the paramagnetism of the europium ion, indicating complexation. Similarly, for complex Eu (DBM) 3 (TPPO) 2 , the phosphorus chemical shift also appears in the same region at δ 25 ppm.

Conclusions
The difficulty involved in the purification of the synthetic intermediates Eu(β-diketonate) 3 (H 2 O) 2 , which may take up to a few weeks, prompted us to devise a new strategy for the syntheses of highly luminescent europium complexes of the general formula Eu(β-diketonate) 3 (L) 2 , where L stands for a nonionic ligand. Experimental proof illustrating the presumed all-encompassing efficacy of the new strategy was provided by a synthetic combinatorial approach. All Eu(β-diketonate) 3 (L) 2 complexes made up by any of the β-diketonate ligands, DBM or TTA, and by any of the non-ionic ligands DBSO, PTSO, or TPPO, were synthesized both by the usual and by the faster synthesis advanced in this article.
Indeed, by simply switching the order in which ionic and non-ionic ligands are usually added to the coordination polyhedron of the europium ion, that is, by first coordinating the non-ionic ligands, and by only subsequently coordinating the ionic ligands, we were able to halve the average elapsed time for the preparation of these highly luminescent complexes from 20 days to 10 days, while boosting the average overall yield of the syntheses from 42% to 69%.