2.2. Effect of Processing Conditions
Preliminary UV-VIS measurements were run on the reagents, i.e., N
) and 1-bromobutane (1-Br-But), used for the preparation of the precursor to reveal the features due to impurities. As stated above, such impurities (contained within the chemicals) represent the major contaminants for the precursor. Both chemicals were previously dissolved in ethanol (at volume ratios equal to 3:1 and 10:1 for PYR1
and 1-Br-But, respectively) to attenuate the spectrophotometric signal. The absorbance vs. wavelength dependence, reported in Figure 1
, shows a main feature (much larger with respect to the other ones) below 250 nm for both chemicals, clearly due to PYR1
(blue trace) and 1-Br-But (red trace), respectively. Much less pronounced, but still well defined, features (better evidenced in the magnification reported in panel b) are displayed for both PYR1
(290 nm) and 1-Br-But (270, 278, and 290 nm). These peaks can be attributed to the impurities within the reagents. Therefore, the features falling in the wavelength range from 270 to 600 nm of the absorbance response were assigned to impurities. The ethanol solvent (dotted trace) is seen to not interfere in the UV-VIS measurements.
The effect of the processing conditions (namely, temperature and time), as well as the amount of the sorbents on the (PYR14Br) precursor purity level was investigated by UV-VIS spectrophotometry measurements. Briefly, aqueous precursor samples were subjected to the purification route described below and, then, vacuum dried (in a rotary evaporator) to remove the water. Successively, the solid PYR14Br was dissolved in ethanol to obtain solutions with analogous concentration (in PYR14Br) with respect to the (initial) aqueous samples.
The sorbent materials (carbon and alumina in a weight ratio equal to 2:3) [38
] were intimately mixed with the pristine (not purified) aqueous precursor. After vacuum filtration (to separately collect the purified PYR14
Br aqueous phase), the sorbents were rinsed with deionized water (sorbents: H2
O weight ratio fixed to 2:1) [38
] to recover the trapped precursor (through the sorbents). Different mixing temperatures and times were investigated whereas the (C + Al2
): precursor weight ratio ranged from 0.75:1 to 2.25:1. Solutions of the pristine precursor (in ethanol) were investigated for comparison purposes.
The effect of the temperature and time is reported in Figure 2
. It is worth noting the impressive difference with respect to the response given by the pristine precursor (grey trace), highlighting for high efficiency of the purification route. The absorbance vs. wavelength trace obtained for the PYR14
Br sample purified at 23 °C (panel a) shows modest, but significant, enhancement of the purity level with respect to that processed at 4 °C which evidenced a more pronounced feature around 280 nm. Conversely, no practical improvement of the precursor quality is observed if the mixing time is increased from 3 to 20 h (panel b). On the basis of these results, the purification routes were carried out above 20 °C, whereas the mixing time was limited to 3 h.
depicts the results of the UV-Vis measurements performed on aqueous PYR14
Br samples subjected to purification routes run at different (C + Al2
): precursor weight ratios. The pristine PYR14
Br is reported for comparison purposes. All absorbance vs. wavelength traces are characterized by a main feature around 220 nm (i.e., much more relevant with respect to the other ones), likely ascribable to the precursor, and much weaker profiles, due to impurities (i.e., in agreement with the results of Figure 1
), detected in the 250–600 nm range.
The effect of the sorbents (Figure 3
a) is highlighted, in the profile of the purified precursor samples, by the disappearance of the feature around 470 nm observed in pristine PYR14
Br and dramatic reduction (with respect to pristine PYR14
Br) of the peaks recorded from 250 to 400 nm. It should be noted that the pristine PYR14
Br solution was previously diluted (in water) in a volume ratio of 5:1 (before to be subjected to UV-VIS measurements) to reduce the absorbance signal. This clearly indicates a drastic decrease of the impurity content in the aqueous precursor. The 470 nm feature falls within the visible wavelength range, thus explaining the yellow coloration of the pristine PYR14
Br sample (i.e., the purified precursor solutions are uncolored). A still relevant reduction in impurity content is evidenced in passing from a (C + Al2
): precursor weight ratio of 0.75:1 to 1.5:1 whereas a moderate decrease is found for a further increase (up to a (C + Al2
): precursor ratio equal to 2.25:1) of the sorbent amount.
The results reported in Figure 3
b show a beneficial, even if moderate, effect due to Al2
as evidenced from the comparison (blue vs. red trace) with the precursor purified without alumina. This is in agreement with results reported in literature [34
], which show a synergic effect of Al2
with activated carbon. Once more, the solvent gives no interference.
2.3. Absorbance Dependence on the Impurity Content
The demand to obtain precursors with high purity level has pushed towards suitable analytical methodologies able to reveal even low impurity contents. The UV-VIS spectrophotometry was selected for its reliability, easy management (also in preparing samples) and low cost. The purpose of this investigation is to find a correlation between the absorbance of the precursor samples and their impurity content.
The aqueous phase as obtained from the synthetic route (containing the pristine precursor and called hereafter “mother solution
”) shows concentration values, referred to PYR14
Br and overall impurities (coming from the PYR1
and 1-Br-But reagents) equal to 780 ± 4 and 12.490 ± 0.070 g L−1
(corresponding to an impurity content, referred to pure PYR14
Br, equal to 16,020 ± 80 ppm), respectively. These values were calculated taking into account: (i) the stoichiometry (Scheme 2
) and the yield (100%) [38
] of the reaction for obtaining the PYR14
Br precursor; (ii) the proper mass (included a PYR1
excess equal to 0.5% in weight with respect to the stoichiometric amount) of the reagents and solvent (deionized water) used in the synthesis route; (iii) the impurity content (i.e., 2 wt% and 1 wt% for PYR1
and 1-Br-But, respectively) of the reagents (used as received); (iv) the density (1.175 ± 0.001 g mL−1
) of the mother solution
. Therefore, it was possible to calculate the mass of the PYR14
Br precursor, obtained from the synthesis process (Scheme 2
), and the overall impurity amount (i.e., allowing to determine the overall impurity content in the pristine precursor before to be purified). The precursor and overall impurity concentrations in the mother solution
were determined through its overall mass (reagents + water) and density values.
Various fractions of the mother solution
were diluted (in deionized water) according to volume ratios from 5:1 to 5000:1 in order to prepare aqueous PYR14
Br samples at different impurity concentrations (mg L−1
, i.e., calculated considering the impurity concentration of the mother solution
and dilution volume ratios) as reported in Table 1
. The overall impurity content (i.e., calculated by the dilution volume ratios) with respect to the pure precursor mass is also reported.
UV-VIS measurements were performed (at room temperature) on the precursor solutions listed in Table 1
. The results (reported in Figure 4
) reveal, apart the features around 200 nm due to PYR14
Br and the PYR1
excess, significant profiles within the 260–450 nm wavelength range, which became progressively weaker with the increase of the dilution ratio. These features, as observed in Figure 1
for the UV-VIS measurements run on reagents, are to be assigned to impurities. In addition, as better evidenced in the magnification of the 250–400 nm range depicted in panel (b), evident spectrophotometric signals are observed even at high dilution ratios (i.e., very low content of impurities). Therefore, the absorbance due to the impurities in the precursor aqueous phase is seen (as expected) to depend on their concentration and, consequently, on the precursor purity level.
2.4. Determination of Precursor Content in Aqueous Phase
The “aqueous” synthesis process allows of obtaining a precursor dissolved in water phase. In order to determinate both the yield of the process (after preparation and purification) and the purity level of the (precursor) sample, the concentration (and the overall amount) of the precursor in the aqueous phase has to be determined.
Br and Br−
have identical molarity, the precursor concentration was determined by checking the bromide content in the aqueous phase. Water solutions of highly pure KBr (Carlo Erba, ≥99.4 wt%) at different molarities (KBr and Br−
have identical molar concentrations) were prepared and subjected to X-ray fluorescence analysis (performed at room temperature). The Br−
concentration (mmol L−1
) of the KBr solution samples was plotted as a function of the count number recorded during the measurements to obtain a calibration curve as reported in Figure 5
, which shows a linear trend. The results are well represented by Equation (1) (regression coefficient = 0.999) where y
(error bar equal to 0.005) represent the Br−
concentration and the count number, respectively:
Therefore, the molar concentration of PYR14Br can be determined, through the Equation (1), by X-ray fluorescence measurements carried out on the precursor aqueous phases (previously diluted in water). The overall amount of PYR14Br can be easily calculated by the volume of the aqueous phase (e.g., obtained by mass and density measurements run on the aqueous phase).
The UV-VIS measurements allowed to reveal the absorbance vs. wavelength range due to impurities (Figure 1
) and to define the processing conditions (Figure 2
) and (C + Al2
): precursor weight ratio (Figure 3
) for the purification route. In addition, a linear relationship between the precursor concentration and count number, recorded by X-ray fluorescence analysis, was found (Figure 4
). Therefore, it was possible to determine the efficiency of the precursor recovery process, carried out by rinsing of the sorbent materials after purification route, and the impurity content in the purified precursor.
2.5. Efficiency of the Precursor Recover Process
The pristine precursor, obtained from the aqueous synthesis process, has to be addressed to the purification route carried out through sorbent materials (i.e., successively, the purified precursor is addressed to the anion exchange process to obtain the ionic liquid). Despite vacuum filtration, a precursor fraction is trapped through sorbents (activated charcoal and alumina), which (upon separation from the aqueous precursor phase) has to be rinsed with deionized water (in weight ratio equal to 2:1 with respect to the sorbent mass) to recover the retained PYR14
Br. However, release of impurities (previously retained by sorbents) should be avoided and/or minimized. This issue was investigated by UV-VIS measurements run on the purified aqueous PYR14
Br both directly obtained through vacuum filtration and collected after rinsing of sorbents. The (C + Al2
Br and C: Al2
weight ratios were fixed equal to 1.5:1 and 2:3, respectively. The results, reported in Figure 6
as absorbance vs. wavelength traces, evidence a much lower impurity content in the aqueous phase obtained by sorbent rinsing (blue curve) with respect to that directly obtained from vacuum filtration (red curve), e.g., revealed by the much smaller features in the 250–400 nm range (conversely, the precursor is revealed by the main feature around 220 nm). Therefore, no relevant impurity release (from sorbent materials) occurs during the rinsing step for a water: (C + Al2
) weight ratio equal to 2:1.
For determining the efficiency of the recovery process of precursor (during the purification route), the aqueous pristine PYR14
Br samples were purified through different (C + Al2
Br weight ratios, i.e., from 0.75:1 to 3:1, whereas the C: Al2
weight ratio was fixed equal to 2:3. The water: (C + Al2
) weight ratio was kept to 2:1 during the rinsing steps. X-ray fluorescence measurements were performed on the purified precursor samples; the aqueous phases collected after both (direct) vacuum filtration (of the slurry containing aqueous PYR14
Br and sorbents) and the rinsing step (of sorbents) were analyzed. The pristine PYR14
Br samples were tested for comparison purpose. All samples were diluted with deionized water according to a volume ratio from 20:1 to 100:1, i.e., depending on the precursor concentration, in order to avoid that signal intensity falls out of the instrumental full scale. The PYR14
Br concentration of each sample was determined, on the basis of the count number recorded during the X-ray fluorescence measurements, using the Equation (1). Finally, the PYR14
Br amount (for each sample) was calculated by the dilution ratio and by determining the volume of the sample. The results, summarized in Table 2
, show a decrease of the PYR14
Br fraction obtained by direct filtration from 91 to 69 wt% in passing from a PYR14
weight ratio from 1:0.30:0.45 to 1:1.20:1.80, e.g., indicating how the fraction of retained precursor raises with increasing the sorbent amount (due to the increase of the overall volume and surface area of sorbents with respect to the precursor amount). However, the rinsing step can recover almost the whole precursor fraction retained by sorbents, allowing to minimize the loss of PYR14
Br (during the purification route) and, therefore, to increase the yield of the purification route up to 100%.
2.7. Investigation on Sorbents:Precursor Ratio
On the basis of the results reported in Figure 5
and Figure 7
, we have tried to optimize the amount of sorbent materials to be used in the purification route. Different PYR14
Br batches were prepared and purified, according to the route described in Materials and Methods, through different (C + Al2
): precursor weight ratios as reported in Table 3
. Upon vacuum filtering of the aqueous PYR14
Br phase (and sorbent separation), the sorbents were rinsed with a deionized water amount twice with respect to the overall amount of C and Al2
(e.g., in weight ratio equal to 2:3). The processing temperature and time were fixed to 23 °C and 3 h, respectively.
The purified (aqueous) PYR14Br samples, obtained by vacuum filtration and sorbent rinsing steps, were investigated by UV-VIS spectrophotometry. Additionally, pristine samples were examined for comparison purpose. It should be noted that the precursor samples obtained from the purification route are more diluted (in PYR14Br) than the pristine ones (i.e., water is properly added either to pristine aqueous PYR14Br before to be treated with sorbents or to the sorbents for recovering the trapped precursor). In addition, the dilution increases from batch I trough IV because progressively higher amounts of sorbents and, consequently, of water were added. Therefore, the pristine precursor solutions were previously diluted in water (before to be subjected to UV-VIS measurements) in order to match the molar concentration of the purified PYR14Br samples. The dilution volume ratio ranged from 1.57:1 (batch I) to 3.29:1 (batch IV).
illustrates the absorbance vs. wavelength dependence of the pristine and purified PYR14
Br solution samples of batches I (panel a) and IV (panel b). The absorbance traces were normalized with respect to the same dilution volume ratio used for the pristine samples. The feasibility of the purification route is clearly observed at lower (C + Al2
): precursor ratios (batch I) as highlighted by the comparison among the pristine (black trace) and purified (red and blue traces) precursor profiles reported in panel a. In good agreement with the results shown in Figure 6
, lower impurity concentration is detected in the samples collected after sorbent rinsing step (blue trace) with respect to those obtained by “direct” vacuum filtration (red trace), once more indicating the feasibility to fully recover the precursor fraction trapped through sorbents without further contaminating the PYR14
Br samples. The comparison among the results shown in panels (a) and (b) indicate higher purity levels in the samples coming from batch IV with respect to batch I due to the larger sorbent amount used for precursor mass unit, e.g., in agreement with the data reported in Figure 3
. Therefore, the value of area, delimited by the absorbance profiles (falling in the 260–600 nm range) of the (pristine and purified) samples of the batches I to IV towards the X axis, was calculated. Conversely, the area value lying below the absorbance curves of the purified samples coming from sorbent rinsing step (blues trace) is too small to be estimated, indicating that the residual impurities (upon purification route) are practically located only in the (aqueous) precursor fraction obtained by “direct” vacuum filtration. The impurity content in the pristine and purified precursor samples of the batches I to IV (differing for the (C + Al2
): precursor weight ratio) was determined, on the basis of the area values, through Equation (2). It should be noted that all pristine samples exhibit the same impurity content, this allowing the evaluation of the purity level of the PYR14
Br precursor after treatment with different amounts of sorbents.
plots the residual fraction of the impurities, normalized with respect to the initial content, as a function of the (C + Al2
Br weight ratio. A relevant and progressive reduction of the impurity level with the increase of the sorbent amount is seen, i.e., from about 13 (above 2000 ppm) to less than 5% (below 800 ppm) with respect to the initial concentration of impurities is detected moving from a (C + Al2
Br weight ratio of 0.75:1 to 1.5:1. A further increase of the sorbent amount does not lead to any practical improvement in terms of the purity level of the precursor. We can likely hypothesize that, when the impurity content is very low, the precursor is preferentially adsorbed onto the sorbent surface (with respect to impurity), this hindering the adsorption of further impurities even in the presence of larger sorbent amounts. It should be noted that the possibility to use lower amounts of sorbents leads to reduction of the time/cost of the purification route and softer rinsing/recycling steps.
To summarize, a proper purification procedure (carried out at room temperature and for a processing time of 3 h), which suitably combines activated carbon and alumina as the sorbent materials and uses water as the only processing solvent, can remove more than 95% of the overall impurity content in pristine precursors as obtained from synthesis routes. The residual impurity content (below 5%) is expected to be removed through cleaning steps performed at the ionic liquid level. Work is in progress in our laboratory aiming to investigate this issue.
The effect of the treatment through sorbents on the purity level of ionic liquids is shown in Figure S1a (Supplementary Materials)
, which compares the results obtained from UV-VIS measurements carried out at room temperature on pristine (not purified) PYR14
TFSI (black trace) and on the same RTIL batch (red trace) after purification route performed at the precursor level according to the procedure followed for batch I. The samples were previously diluted in ethanol before running the UV-VIS measurements. The much weaker features, observed in the absorbance vs. wavelength trace of purified PYR14
TFSI with respect to those of the pristine RTIL, clearly indicate a remarkable reduction of the impurity content during the purification route through the sorbents. This is also confirmed by sharp coloration turning from the pristine (yellowish, A) to purified PYR14
TFSI (uncolored, B), as reported in the picture of Figure S1b
Finally, the effect of the precursor impurity level on the electrochemical performance, e.g., in terms of ionic conductivity and cathodic stability, of the PYR14
TFSI ionic liquid was investigated. The tested RTIL samples were synthesized after having subjected the PYR14
Br precursor to different purification routes, e.g., according to the procedure described for batches I (e.g., the precursor was purified using a (C + Al2
Br weight ratio equal to 0.75:1) and IV (PYR14
Br:sorbents weight ratio equal to 3:1), respectively, as reported in Table 3
. Therefore, the PYR14
TFSI ionic liquid samples were prepared from PYR14
Br precursors at different impurity contents (Figure 9
). The results are depicted in Figure S4 as Supplementary Materials
. The PYR14
TFSI sample (Figure S4a
) obtained from unpurified PYR14
Br (open black squares), i.e., exhibiting an impurity content > 16,000 ppm, shows slightly different conductivity values with respect to that coming from precursor (i.e., purified through the procedure followed for the batch IV) containing less than 800 ppm (red full squares). No variation of the melting point, evidenced by a conduction jump of more than four orders of magnitude around −7 °C [39
], is observed. Therefore, the impurity content does not relevantly affect the ion transport properties of the ionic liquids both in the solid and molten state. Figure S4b
compares the cathodic voltammetry trace of two PYR14
TFSI samples obtained from PYR14
Br precursor having an impurity level above 2000 ppm (black dotted trace) and below 800 ppm (red solid trace). The less pure RTIL shows well-pronounced, large features, likely ascribable to the reduction of impurities, around −1.5 V, −2.5 V, and −3.3 V vs. Ag°/Ag+
, respectively, prior to the massive reduction of PYR14
] which takes place around 3.8 V. Additionally, impurity traces may catalyze the RTIL reduction, which is detrimental in electrochemical systems as it reduces the operating voltage range and, consequently, the performance of the device. Conversely, a practically flat voltammetry curve is seen for the purer PYR14
TFSI, which also exhibits superior cathodic break-down potential. Therefore, the purity level of the precursor plays a key role in the electrochemical stability of RTILs, especially in view of application in practical devices.