Chirogenesis and Pfeiffer Effect in Optically Inactive EuIII and TbIII Tris(β-diketonate) Upon Intermolecular Chirality Transfer From Poly- and Monosaccharide Alkyl Esters and α-Pinene: Emerging Circularly Polarized Luminescence (CPL) and Circular Dichroism (CD)

We report emerging circularly polarized luminescence (CPL) at 4f-4f transitions when lanthanide (EuIII and TbIII) tris(β-diketonate) embedded to cellulose triacetate (CTA), cellulose acetate butyrate (CABu), D-/L-glucose pentamethyl esters (D-/L-Glu), and D-/L-arabinose tetramethyl esters (D-/L-Ara) are in film states. Herein, 6,6,7,7,8,8,8-heptafluoro-2,2-dimethyl-3,5-octanedionate (fod) and 2,2,6,6-tetramethyl-3,5-heptanedione (dpm) were chosen as the β-diketonates. The glum value of Eu(fod)3 in CABu are +0.0671 at 593 nm (5D0→7F1) and −0.0059 at 613 nm (5D0→7F2), respectively, while those in CTA are +0.0463 and −0.0040 at these transitions, respectively. The glum value of Tb(fod)3 in CABu are −0.0029 at 490 nm (5D4→7F6), +0.0078 at 540 nm (5D4→7F5), and −0.0018 at 552 nm (5D4→7F5), respectively, while those in CTA are −0.0053, +0.0037, and −0.0059 at these transitions, respectively. D-/L-Glu and D-/L-Ara induced weaker glum values at 4f-4f transitions of Eu(fod)3, Tb(fod)3, and Tb(dpm)3. For comparison, Tb(dpm)3 in α-pinene showed clear CPL characteristics, though Eu(dpm)3 did not. A surplus charge neutralization hypothesis was applied to the origin of attractive intermolecular interactions between the ligands and saccharides. This idea was supported from the concomitant opposite tendency in upfield 19F-NMR and downfield 1H-NMR chemical shifts of Eu(fod)3 and the opposite Mulliken charges between F-C bonds (fod) and H-C bonds (CTA and D-/L-Glu). An analysis of CPL excitation (CPLE) and CPL spectra suggests that (+)- and (–)-sign CPL signals of EuIII and TbIII at different 4f-4f transitions in the visible region are the same with the (+)-and (–)-sign exhibited by CPLE bands at high energy levels of EuIII and TbIII in the near-UV region.

Herein, we showcase that Eu(fod) 3, Tb(fod) 3 , and Tb(dpm) 3 , except Eu(dpm) 3 commonly exhibited CPL signals at 4f-4f transitions and CD bands due to n-π * /π-π * transitions of the ligands when CABu, CTA, D-Glu/L-Glu, and D-Ara/L-Ara were employed as embedding films. In α-pinene, although Tb(dpm) 3 showed clear CPL characteristics, Eu(dpm) 3 did not. Unresolved inherent nature between Eu III and Tb III causes considerable differences of CPL characteristics as well as emission wavelengths at 4f-4f transitions. Although the current g lum values of our CPL-active lanthanide complex in the films are not very outstanding, the hypothesis of these non-covalent weak chiral intermolecular interactions in the ES and GS led us to freely design solution processible CPL-and/or CD-functioned composite films made of several optically inactive lanthanide complexes coordinated with achiral organic ligands upon chirality/helicity transfer of inexpensive soluble polysaccharide, oligosaccharide, and monosaccharide derivatives as the solidified chiral bioresources, in addition to a conventional multi-step synthesis of optically active lanthanide complexes coordinated with chiral ligands designed strategically.

Monosaccharide Permethylesters
L-(-)-Glu. To a mixture of L-(-)-glucose (0.5 g) and stoichiometric acetic anhydride (4 mL), freshly dried Cu(OTf) 2 (0.03 mol % of L-(-)-glucose) at 0 • C was added under nitrogen [Scheme S1 in Supplementary Materials (SM)]. The mixture was stirred for 1 h in an ice bath and then stirred at room temperature for 12 h. Methanol (5 mL) was slowly added to quench the acylation reaction, and the mixture was stirred for another 0.5 h, followed by evaporation under reduced pressure. Chloroform (25 mL) was added to dissolve the residue, and the mixture was consecutively washed twice with saturated NaHCO 3 brine and water, respectively. The organic layer was dried over anhydrous Na 2 SO 4 , and the mixture was filtered. The filtrate was concentrated in a vacuum and purified by silica gel column chromatography, and eluted with petroleum ether/ethyl acetate (v/v, 50/1) to yield a white powder (yield: 0.42 g, 70%; Analysis: calculated (%) for C 16  D-Ara and L-Ara. D-(-)-Arabinose tetramethyl ester and L-(+)-arabinose tetramethyl ester were synthesized utilizing the same procedure as described for the synthesis of L-Glu. A viscous liquid was obtained as D-Ara. Yield: 2.4 g, 32%. Analysis for D-Ara: Calcd (%) for C 13 H 18 O 9 : C, 49.06; H, 5.70, Found (%): C, 48.97; H, 5.56. L-Ara. Yield (2.4 g, 32%). Analysis: Calcd (%) for C 13 H 18 O 9 : C, 49.06; H, 5.70, Found (%): C, 49.01; H, 5.60. 1 H-NMR, FT-IR, and ESI-MS (positive mode) spectra of D-Ara and L-Ara are shown in Figures S6, S7, SM, respectively. The results of the elemental analysis of L-Ara and D-Ara are shown in Figure S8, SM.

Tb(fod) 3
The starting material, 0.19 g (0.74 mmol) of TbCl 3 (99% purity, Sigma-Aldrich, now Merck) was dissolved in the minimum amount of methanol (3.0 mL), and 1,1,1,2,2,3,3-heptafluoro-7,7dimethyl-4,6-octanedione (Hfod, 0.636 g (2.15 mmol, 0.5 mL), TCI) was adjusted to pH 5-6 by adding the required amount of aqueous NaOH solution. The above two solutions were mixed by vigorous stirring with a magnetic stir bar for 10 min, followed by addition of 200 mL distilled water dropwise into the solution. A pale-yellow Tb(fod) 3 was precipitated under vigorous stirring with a magnetic bar for 12 h. The crude product adhered to the bottom of the reaction vessel was purified by a short-column silica gel chromatography (Wakogel C-200, Chart S1 in SI) as shown below, with chloroform as an eluent to yield a pale yellow oil, followed by drying in a vacuum oven 90 • C to obtain a white solid. Yield, 200 mg (40%). The purification of Tb(fod) 3 using a 5 mL pipette tip made of polypropylene, as a shortcolumn apparatus, is illustrated below. Analysis: Calcd (%) for C 30 H 30 F 21 O 6 Tb: C, 34.50; H, 2.90. Found (%): C, 34.58; H, 2.80. The elemental analysis suggested that Tb(fod) 3 has no water adducts.
Before we utilized the purification method (see Chart S1, SM) for Tb(fod) 3 , we were aware of some impurity peaks approximately 450 nm in the PL spectrum of Tb(fod) 3 . After the purification by silica gel column chromatography and elution with chloroform, we obtained a pure Tb(fod) 3 (based on the expected fluorescent emission spectrum). 1 H-NMR (CDCl 3 ), 19 F-NMR (CDCl 3 ), FT-IR (CaF 2 ), and ESI-MS (positive mode) spectra and elemental analysis of Tb(fod) 3 are shown in Figure S9, SM.

Instrumentation
The UV-vis and CD spectra of the solutions were measured with a JASCO J-820 spectropolarimeter (Hachioji-Tokyo, Japan) equipped with Peltier-controlled housing units. Synthetic quartz (SQ) cuvette with a 10-mm path length (scanning rate: 100 nm min −1 ; bandwidth: 1.0 nm; response time: 1.0 s; 0.5-nm interval sampling; single accumulation) at 25 • C were used. To avoid second-and third-order stray light due to diffraction grating, CPL and PL spectra were recorded on a JASCO CPL-200, that was designed as a prism-based spectrofluoropolarimeter with a forward scattering of 0 • angle equipped with focusing and collecting lenses, and a manually movable film holder onto an optical rail enables to adjust the best focal point to maximize CPL/PL signal amplitudes. Measurement conditions were bandwidths of 10 nm for excitation and emission, a scanning rate of 100 nm min −1 , and a data sampling of 0.5 nm interval. IR spectra were measured on CaF 2 plate using a Horiba FT-730 Fourier-transform (FT) infrared (IR) spectrometer (Horiba, Kyoto, Japan) over a wavenumber range between 800 and 4,000 cm −1 with a resolution of 2 cm −1 and a scanning speed of 5 mm s −1 for 128 scans and a Perkin-Elmer Spectrum One/100 FT-IR spectrometer (Winter Street Waltham, MA 02451, USA) over a wavenumber range between 900 and 4,000 cm −1 with a resolution of 4 cm −1 for 64 scans. Electrospray ionization mass spectrometry (ESI-MS) was conducted with a JEOL (Akishima, Tokyo, Japan) AccuTOF JMS-T100 LC mass spectrometer (accelerating voltage, 10 kV). Electron ionization mass spectrometry with highresolution (HR-EI-MS) mode was recorded with a JEOL JMS-700 double-focusing mass spectrometer (accelerating voltage, 10 kV). The ionic species were often attached with Na + ion. The hybridized polymers were characterized by a JEOL JNM-ECX400 cross-polarization (CP) magic-angle-spinning (MAS) solid-state (ss)-13 C{H}-FT-NMR spectrometer (resonance frequency 100.5 MHz, contact time 2.0 ms, 550 scans, relaxation delay 5.0 s, spinning 8.0 kHz, repetition time 5.05 s). Elemental analysis was performed on a Perkin-Elmer 2400II CHNS/O. The solution 1 H-and 19 F-FT-NMR spectra were recorded on the JEOL ECP-400 spectrometer. The resonance frequencies of 19 F-and 1 H-NMR are 376 MHz and 400 MHz, respectively. Representative measurement conditions for 19 F-NMR spectra had an acquisition time of 0.432 sec, 64 acquisitions, a relaxation delay of 4.0 sec, at a temperature of ∼20 • C, a pulse angle of 45 • and a pulse width of 7.0 sec were used. Raw NMR data were processed and analyzed by JEOL Delta (Ver. 5) software. Hexafluorobenzene (HFB, −163.0 ppm) and tetramethylsilane (Me 4 Si, 0.0 ppm) were used as internal standards for the 19 Fand 1 H-NMR measurements, respectively. Photodynamic decay of six solid films (Eu(fod) 3 with CTA and CABu (detected at 610-620 nm), Eu(dpm) 3 with CTA and CABu (detected at 610-620 nm), and Tb(dpm) 3 with CTA and CABu (detected at 542-551 nm) excited by an N 2 laser (Usho KEC-160; wavelength 337.1 nm; pulse width 600 ps; 10 Hz) were measured with the help of streak camera (Hamamatsu, picosecond fluorescence measurement system C4780 with Grating 150 lines per mm and slit width 100 µm). The 337.1 nm of N 2 laser source was used to excite shoulder UV/CD signals of the lanthanide complexes. Photodynamic measurements of other Eu III and Tb III complexes in the D-/L-Glu and D-/L-Ara films excited at 337.1 nm were not successful. For simplicity, the emission lifetime was evaluated by single exponential decay analysis. Quantum yields of the Eu III and Tb III complexes in the solid films were not obtained due to lack of an integrating sphere. The all processed data saved as raw text data were re-organized by KaleidaGraph ver. 4.53 (Synergy software, Reading, PA 19606, USA).

Preparation of the Hybridized Films
In fabricating the hybridized film, 10 mg of lanthanide complexes and 20 mg of saccharide derivatives (Glu and Ara) or cellulose derivatives (CABu and CTA) were completely dissolved in 1.0 mL of the desired solvent (chloroform or tetrahydrofuran (THF)) at ambient temperature. The hybridized film was deposited onto a polished circular quartz plate or borosilicate glass (Tempax Float R , Schott AG. Germany) (25 mm in diameter and 1 mm in thickness) by spin coating using a spin coater (MIKASA, MS-B100, Tokyo, Japan), then, 800 µL of the solution was placed onto the center of the plate and spun at 1,500 rpm for 60 s. The films on the glass were attached on both sides (front and back surfaces) to ensure an optically symmetrical geometry with air-sample-(quartz or borosilicate substrate)-sample-air contact by spin coating chloroform or THF solutions that consist of saccharides (chiral host) and lanthanide complex (achiral guest) (Guo et al., 2017(Guo et al., , 2018Yamada et al., 2018). Although the film thicknesses of both sides were not determined, we assumed to be on the order of several µm for each. The hybridized double-side coating films were scattering-free and transparent by the naked eye. To measure CPL/PL/CPLE/PLE/CD/UVvisible spectra, the optical density of the double-side coating specimen was controlled to 0.3-1.0 in the range of 280 and 330 nm. CD, CPL, and CPLE spectra of the hybridized films were measured at ambient temperature (24-26 • C). The double-sided coating in the symmetrical optical geometry avoids chiroptical inversion artifacts that could be originated from linear dichroism induced by mechanical stress on anisotropic specimens due to spin coating. Based on our experience, single-side coating in the dissymmetrical optical geometry can often cause artifact inversion in signs of CPL and CD signals. In the case of singleside coating, the probability of the chiroptical sign inversion was approximately 2-3 out of 10, while double-side coating prevented the artifact origin sign inversion.

RESULTS AND DISCUSSION
The chirogenesis characteristics of oligo-/polyfluorenes originate from rotatable C-C bonds between fluorene rings and from C-O/C-C bonds of the cellulose derivatives (Figure 1) (Guo et al., 2017(Guo et al., , 2018Yamada et al., 2018) in the GS and ES because rotational barrier heights of the single bonds are as small as 1.5-2.5 kcal mol −1 . Eu III and Tb III complexes with three fod ligands should coexist as racemic mixtures of D-/L-species of C 3 -symmetrical facial (fac)-and C 1 -meridional (mer) motifs (Brittain and Richardson, 1977a;Jalilah et al., 2018), while even Eu III and Tb III with three dpm should coexist as a racemic mixture of D-/L-species of D 3 -geometry. Although barrier heights of D-L stereomutation and/or fac-mer isomerisms are considerably high on the order of 10-20 kcal mol −1 (Glover-Fischer et al., 1998;Carr et al., 2012;Miyake, 2014), multiple intermolecular C-H/O-C, C-H/π, and C-H/F-C interactions (Murray-Rust et al., 1983;Nishio et al., 1998;Desiraju and Steiner, 1999;Tsuzuki et al., 2003;Yuasa et al., 2011;Koiso et al., 2017;Jalilah et al., 2018) should overcome the barriers when solidified matrices are employed. Note that solidified matrices are regarded as solid-like solvents with a very high viscosity. In this work, we applied a double-side, spin-coating technique (Guo et al., 2017(Guo et al., , 2018 to fabricate CPL-/CD-functioned films deposited onto fused quartz and/or borosilicate glass to obtain artifact-free CPL/photoluminescence (PL), CPLE/PL excitation (PLE), and CD/UV-visible spectra. CPL and CPL spectral characterizations of Eu III and Tb III at 4f-4f transitions were assigned based on the literature (Fulgêncio et al., 2012;Tanner, 2013;Binnemans, 2015;de Queiroz et al., 2015;Xue et al., 2015;Yang et al., 2017). Dimensionless Kuhn's anisotropic ratios in the ES and GS, being popularly known as g lum and g abs , were manually evaluated at a specific extreme wavelength (λ ext ) of the corresponding CPL and CD spectral profiles in line with the literature (Eliel and Wilen, 1994). All CPL characteristics (g lum value at λ ext ) of Eu III and Tb III complexes are summarized in Table 1. 1 | CPL characteristics (dissymmetry ratio, g lum in 10 −2 at specific wavelength) of Eu III and Tb III coordinated with three β-diketonates as achiral ligands embedded in two polysaccharide alkyl esters (CABu and CTA), D-/L-glucose pentamethyl esters (D-/L-Glu), and D-/L-Arabinose tetramethyl esters (D-/L-Ara).

Chirality Transfer Capability From Cellulose Alkyl Esters to Eu(fod) 3
The normalized CD and UV-visible spectra of Eu(fod) 3 in CABu and CTA films are shown in Figures 2A,B, respectively. For comparison, the original raw CD and UV-visible spectra of the Eu(fod) 3 -hybridized films were given in Figure S13A, SM. Bisignate profile at Cotton CD bands at 290 and 310 nm between CABu and CTA films are obviously opposite. These Cotton CD bands at 290 and 310 nm, however, do not originate from CABu and CTA. Broad monosignate CD bands due to n-π * transition from alkyl esters of CABu and CTA thin films appeared at ∼215 nm with (+)-sign and ∼205 nm with (-)-sign, respectively (Guo et al., 2018). These (+)and (-)-sign CD bands at 205 and 215 nm in the solid film reflect from left-handed helicity of CABu and right-handed helicity of CTA in solutions, respectively (Dubois et al., 1998;Onofrei et al., 2015), though CABu and CTA are β-(1→ 4) linked polymers made of D-glucose framework as a common repeating unit. Although the alkyl ester itself does not have a stereogenic center, the ester can adopt particular chiral conformational geometry by the direct connection of the D-glucose ring. Two lone pairs at ethereal "-O-" and two C-H groups at "-CH 2 -" are no longer to be equal because of C-O-C single bonds in R-C(=O)-O-CH 2 -side group act as pseudochiral stereogenic bonds, similar to gauche n-butane. The unequal lone pairs at ethereal oxygen and unequal CH 2 groups may be responsible for the induction of chiral intermolecular C-O/H-C and C-H/F-C interactions between the alkyl ester moieties and lanthanide ligands. However, any CD/CPL signals of Eu(fod) 3 in the presence of CABu and CTA in dilute chloroform solutions (∼10 −3 M) were not able to detect because the postulated chiral intermolecular C-O/H-C and C-H/F-C interactions are inherently weak in the fluidic solution. The postulated chiral alkyl esters of CABu and CTA can thus act efficiently and differently in the solidified films only as external chirality inducible scaffoldings toward optically inactive Eu(fod) 3 and several lanthanide complexes, as discussed in later sections.
The g abs values at λ ext of Eu(fod) 3 at 290 and 316 nm in CABu film are +3.5 × 10 −4 at 280 nm and −3.5 × 10 −4 at 310 nm, respectively, while those in CTA film are −1.6 × 10 −4 at 290 nm and +1.1 × 10 −4 at 300 nm, respectively. Note that the λ max values at non-polarized UV-visible spectra of Eu(fod) 3 in CABu and CTA films are commonly ∼291 nm. Although these CD bands at ∼290 nm and ∼310 nm are ascribed to n-π * / 1 π-3 π * bands of the three fod ligands, their signs appear to be determined solely by preferential helix sense and/or local chirality of multiple alkyl esters of CABu and CTA.
To confirm the apparent inconsistency between the retention in CPL bands at 4f-4f transitions in CABu and CTA films and between the inversion in CD bands at n-π/ 1 π-3 π * transitions in CABu and CTA films, we applied CPLE and PLE spectroscopy (Duong and Fujiki, 2017) by monitoring at 5 D 0 → 7 F 1 and 5 D 0 → 7 F 2 transitions in CABu and CTA films, as shown in Figures 2E,F, respectively. Disregard of CABu and CTA, it is obvious that the CPLE band at 310 nm is commonly (+)sign monitored at 5 D 0 → 7 F 1 transition and that the CPLE band at 310 nm is commonly (-)-sign monitored at 5 D 0 → 7 F 2 transition. The magnitudes of the CPLE bands in CABu film are +4.86 × 10 −2 monitored at 589 nm and −0.30 × 10 −2 monitored at 615 nm, respectively. Similarly, the magnitude of the CPLE bands in CTA film somewhat weaken, and +2.96 × 10 −2 monitored at 590 nm and −0.33 x 10 −2 monitored at 615 nm, respectively.
The origin of the inconsistency between the sign at the first Cotton CD band (310 nm) and the opposite CPLE sign at this wavelength that depends on the wavelengths monitored at 5 D 0 → 7 F 1 and 5 D 0 → 7 F 2 transitions is an unresolved question and obscure. However, the n-π/ 1 π-3 π * bands at ∼310 nm with the opposite sign originates from the three fod ligands, is obviously degenerative, and is responsible for LMCT (from the ligands to high energy levels of Eu III , for example, 5 D 2 state), leading to 5 D 0 → 7 F 1 and 5 D 0 → 7 F 2 transitions with the opposite CPL sign. The broader ∼310 nm transition is likely to be a convolution of two nearly degenerate transitions with an opposite chirality; one (+)-sign band at ∼310 nm is responsible for 5 D 0 → 7 F 1 and another (-)-sign band at ∼310 nm for 5 D 0 → 7 F 2 bands. The 5 D 2 state of Eu III is close to the lowest photoexcited T 1 states of the ligands. When one excite simultaneously at couplet-like 1 π-3 π * transitions (∼310 nm) of CD-active Eu(fod) 3 using monochromated non-polarized light, the photoexcited Eu(fod) 3 decays into the 7 F 1 and 7 F 2 states with two different pathways because 5 D 0 → 7 F 1 and 5 D 0 → 7 F 2 states are magnetic dipole (MD) allowed transition with an electric dipole (ED) forbidden transition and MD forbidden transition with a forced induced ED transition, so-called, hypersensitive transition, respectively (Tanner, 2013;Binnemans, 2015).

Chirality Transfer Capability From Cellulose Alkyl Esters to Tb(fod) 3
The normalized CD and UV-visible spectra of Tb(fod) 3 in CTA and CABu films are compared in Figure 3A. For comparison, the original CD and UV-visible spectra of the films were given in Figure S13B, SM. Unlikely to the case of Eu(fod) 3 , trisignate profile at CD bands appeared at (-)-sign at 315 nm, (+)-sign at 292 nm and (-)-sign at 273 nm in CABu while (+)-sign at 322 nm, (+)-sign at 300 nm and (-)-sign at 275 nm in CTA. The apparent g abs values at the first Cotton band of Tb(fod) 3 in CABu and CTA films are −3.5 × 10 −4 at 315 nm and +2.0 × 10 −4 at 324 nm, respectively. CABu efficiently induced the Cotton CD band of Tb(fod) 3 rather than CTA, similar to the case of Eu(fod) 3 .
Chirality Transfer Capability From Cellulose Alkyl Esters to Eu(dpm) 3 and Tb(dpm) 3 The dpm ligand is a symmetrical β-diketonate, in which two methyl groups of acetylacetonate are replaced by two electrondonating (ED) tert-butyl groups. Therefore, Tb(dpm) 3 and Eu(dpm) 3 can adopt a single D 3 -symmetrical configuration (Brittain and Richardson, 1977a;Brittain, 1980). In a recent paper (Jalilah et al., 2018) and the present work, we confirmed that Eu(fod) 3 in α-pinene has shown clear CPL signals due to the presence of electron-withdrawing (EW) fluoroalkyl groups. Eu(dpm) 3 in α-pinene shows no detectable CPL signals due to the lack of EW fluoroalkyl groups (Figure S14B, SM).
Tb(dpm) 3 in CABu and CTA films shows clear CD and UVvisible spectra in the range of 200 nm and 340 nm, as shown in Figure 4A. For comparison, the original CD and UV-visible spectra of the films are given in Figure S13C, SM. Tb(dpm) 3 in CABu and CTA films, however, showed detectable but weak CPL spectra at 4f-4f transitions, as shown in Figure 4B. The g lum values in CABu film are −0.53 × 10 −3 at 5 D 4 → 7 F 6 (491 nm), +0.37 × 10 −3 at 5 D 4 → 7 F 5 (537 nm), −0.59 × 10 −3 at 5 D 4 → 7 F 5 (547 nm), while the g lum values in CTA are −0.44 × 10 −3 at 5 D 4 → 7 F 6 (489 nm) and −0.80 × 10 −3 at 5 D 4 → 7 F 5 (547 nm) ( Table 1). The absolute magnitudes of g lum values with dpm ligands greatly diminished compared to Tb III complexes with fod ligands. Tb(dpm) 3 and Eu(dpm) 3 showed different behaviors toward external chiral chemical perturbations regardless of the same fod and dpm as the ligands. Eu(dpm) 3 in CABu and CTA did not demonstrate obvious CPL spectra although the corresponding PL spectra are evident (Figures S14A,B, SM).
We ascertained many times that there were no detectable CPL signals of Eu(dpm) 3 in CABu films. This could be because Eu(dpm) 3 is lack of EW-fluoroalkyl groups that can cause efficient chiral C-F/H-C interactions. Chiral C-O/H-C interactions seem not efficient to induce the chiral perturbation.

Chirality Transfer Capability From Monosaccharide Permethyl Esters to Eu(fod) 3 and Tb(fod) 3
Kipping and Pope found that preferential crystallization of L-NaClO 3 in the presence of naturally occurring D-glucose and D-mannitol (Kipping and Pope, 1898). Currently, non-naturally occurring L-glucose is available commercially, although it is costly. L-(+)-and D-(-)-arabinose are also available, but the L-form is more abundant in nature than the D-form due to unknown reasons. To verify whether chirogenesis of Eu(fod) 3 is solely determined by point chirality of monosaccharides, we measured CD and CPL spectra of Eu(fod) 3 embedded in D-/L-Glu and D-/L-Ara films, as displayed in Figures 5A-D. Firstly, CD and UV-visible spectra between Eu(fod) 3 in Dand L-Glu films are compared in Figure 5A. The original CD and UV-visible spectra of the films are given in Figure S13D, SM. Eu(fod) 3 showed nearly mirror-image bisignate CD bands, though the value of λ ext at the first and second Cotton bands are considerably different from each other. The g abs values of Eu(fod) 3 in D-Glu film are +0.89 × 10 −4 at 321 nm and −0.42 × 10 −4 at 284 nm, while in L-Glu film, these values are −0.80 × 10 −4 at 315 nm and +0.44 × 10 −4 at 282 nm ( Table 1).
CD and UV-visible spectra between Eu(fod) 3 in D-and L-Ara films are compared in Figure 5B. The original CD and UV-visible spectra of the films are given in Figure S13E, SM. Similarly, Eu(fod) 3 shows nearly mirror-image bisignate CD bands, though the value of λ ext at the first and second Cotton bands are subtly different. The g abs values of Eu(fod) 3 in D-Ara are −0.96 × 10 −4 at 321 nm and +0.55 × 10 −4 at 281 nm, respectively, while in L-Glu film, these values are +1.08 × 10 −4 at 316 nm and −0.39 × 10 −4 at 275 nm, respectively (Table 1).
CPL and PL spectra of Tb(fod) 3 excited at 315 nm in D-/L-Glu and D-/L-Ara films are given in Figures 5E,F, respectively. Regardless of Glu and Ara, although Tb(fod) 3 shows very weak CPL bands at 4f-4f transitions, CPL signs are likely to depend on the chirality of Glu and Ara. Because the absolute g lum values at these 4f-4f transitions considerably weaken, the g lum values cannot be precisely evaluated.
Although L-cellulose is not available on earth, we can conclude that D-chirality of CTA, CABu, Glu, and Ara determines the (+)-and (-)-sign CPL characteristics of Eu(fod) 3 at 5 D 0 → 7 F 1 and 5 D 0 → 7 F 2 transitions. Conversely, (-)-and (+)-signs at these transitions are from the L-chirality of Glu and Ara, although the inversion in the bisignate CD bands of Eu(fod) 3 at 280-290 nm and 300-310 nm are considerably dependent on the nature of the alkyl groups of CTA and CABu. Similarly, D-chirality of CTA, CABu, Glu, and Ara determines the (+)and (-)-sign CPL characteristics of Tb(fod) 3 at two 5 D 4 → 7 F 5 transitions. Conversely, L-chirality of Glu and Ara determines the (-)-and (+)-signs at the transitions.
We further verified these CPL signals of Tb(dpm) 3 by the broad CPLE spectra centered at ∼300 nm associated with the corresponding PLE spectra with several shoulders, as marked by blue, pink, and green bars in Figures 4D,E. The signs of the CPLE spectra depend on the signs at monitor wavelengths (535 and 548 nm) and α-pinene chirality; (+)-sign in CPLE spectrum is identical to (+)-sign CPL signal at 546 nm in (R)-pinene, while (+)-sign in CPLE spectrum is the same of (+)-sign CPL signal at 535 nm in (S)-pinene. The broad CPLE/PLE spectra may arise from at least three different origins of n-π * / 1 π-3 π * transitions of three dpm ligands associated with high energy levels (e.g., 5 D 1 and 5 D 2 ) of Tb III (Fulgêncio et al., 2012;de Queiroz et al., 2015;Xue et al., 2015). A similar tendency can be seen in the broad CPLE/PLE spectra of Tb(fod) 3 in CABu, as marked in blue, pink, and green bars in Figures 3C,D.
Unresolved factors between Tb III and Eu III associated with ligands (fod and dpm) and chiral matrices (α-pinene and monosaccharide alkyl esters) are other critical parameters to generate CPL signals and boost g lum characteristics. Although the inherent nature of the differences between Tb III and Eu III is unresolved, we assume that the interactions of C-H/π between C(δ-)-H(δ+) bonds at dpm of Tb(dpm) 3 and C(δ+)-C(δ-) double bond of α-pinene are more crucial while the C-F/H-C interactions between C(δ+)-F(δ-) bonds of fod ligands and H(δ+)-C(δ-) bonds of α-pinene are crucial (Jalilah et al., 2018).

Intermolecular Interactions Between Eu(fod) 3 and CTA/Glu by Solid-State 13 C{ 1 H}-NMR and Solution 1 H/ 19 F-NMR Spectra
We do not yet know what kinds of intermolecular noncovalent interactions exist between the lanthanide tris(β-diketonate) and the poly-/monosaccharide alkyl esters. Among lanthanide tris(βdiketonate)s, we chose Eu(fod) 3 for simplicity and excellent solubility in CDCl 3 . The three fod ligands have 1 H, 19 F, and 13 C-NMR active elements to discuss the possible interactions. Additionally, we chose CTA and D-/L-Glu for simplicity in these solid-state (ss)-13 C-NMR and solution 1 H-/ 19 F-NMR spectra. According to the hard-soft-acid-base theory proposed by Pearson (Pearson, 1963), the lone pairs of the hard base O atom(s) of ester groups can coordinate with the hard acid Eu III of Eu(fod) 3 . A marked chemical shift in 13 C-NMR spectra of ester (C=O and O) and ethereal C-O-C atoms was expected.
The ss-13 C{ 1 H}-NMR spectra of Eu(fod) 3 , CTA, and a mixture of Eu(fod) 3 and CTA in 1/1 (w/w) are compared in Figures 6A,B. The ss-13 C-NMR spectra of the Eu(fod) 3 -CTA mixture are merely a convolution of those of CTA and Eu(fod) 3 . Any noticeable chemical shift in the 13 C-NMR spectra was not seen. 13 C-NMR chemical shift at ∼170 ppm due to O=C-O of CTA was unchanged after mixing with Eu(fod) 3 .
The ss-13 C{ 1 H}-NMR spectra of Eu(fod) 3 , D-Glu, and a mixture of Eu(fod) 3 and D-Glu in 1/1 (w/w) are compared in Figures 6C-G. Similarly, the ss-13 C-NMR spectra of the Eu(fod) 3 -D-Glu mixture are merely a convolution of those of D-Glu and Eu(fod) 3 . Any remarkable chemical shift in the 13 C-NMR spectra was not seen. Even five well-resolved O=C-O peaks of D-Glu pentamethyl ester ranging of ∼169 ppm and ∼173 ppm showed no detectable chemical shifts (Figure 6D). Similarly, no noticeable chemical shifts of three well-resolved methyl groups ranging from ∼19 ppm to ∼22 ppm were observed (Figure 6E). Among five well-resolved 13 C peaks assignable to O = C-O of five esters, C-O-C pyranose ring ranging from ∼64 to ∼72 ppm and at ∼83 ppm does not show remarkable chemical shifts of D-glucose ring. These ss-13 C{ 1 H}-NMR data led us to the conclusion that any O atom(s) of glucose ring and ester moieties do not coordinate directly to Eu III ions.
On the other hand, alterations in the chemical shifts of the solution 19 F-NMR spectra of Eu(fod) 3 in the absence and presence of CTA, D-Glu, and L-Glu in CDCl 3 are apparent, as shown in Figures 7A-C. The outer CF 3 signal of Eu(fod) 3 in CDCl 3 resonates at −81.73 ppm as a single peak, indicating C 3symmetrical geometry ( Figure 7B). The single peak resonates at −82.21 ppm and −82.26 ppm in the absence and presence of L-Glu and D-Glu, respectively, and shifts upfield by 0.48 ppm and 0.53 ppm, respectively ( Figure 7B). In the presence of CTA, the CF 3 signal resonates at −82.31 ppm and shifts upfield by 0.58 ppm (Figure 7B). Although the middle CF 2 signal of fod ligand resonates broadly at −126 ppm, in the presence of L-Glu and D-Glu, this signal resonates at −127.4 ppm and −127.5 ppm, that is shifted upfield by 1.4 ppm and 1.5 ppm, respectively ( Figure 7C). Similarly, the middle CF 2 peak at −127.5 ppm shifts upfield by 1.5 ppm in the presence of CTA (Figure 7C). The inner CF 2 of fod shows a broad resonance at −129.2 ppm, but this signal appears at −130.1 ppm in the presence of L-and D-Glu and shifts upfield by 0.9 ppm (Figure 7C). Similarly, the signal appears at −130.3 ppm in the presence of CTA and shifts upfield by 1.1 ppm. These upfield shifts in the 19 F-NMR spectra indicate intermolecular interactions between the F atoms and CTA, L-Glu, and D-Glu, possibly, C-F(δ-) (of the three fod ligands)/H(δ+)-C (of CTA and Glu) interactions.
It is thus evident that there were no marked alterations in ss-13 C-NMR spectra between CPL-inactive and CPL-active Eu(fod) 3 in CTA, L-Glu, and D-Glu, while the remarkable upfield chemical shifts in 19 F-NMR of fod ligands in dilute CDCl 3 solution was distinct. These characteristics should arise from outer or second-sphere effects perturbed by chiral chemicals that are non-coordinating to Eu III , the so-called 'Pfeiffer effect'. The multiple H(δ+)-C(δ-) bonds of chiral chemical species (CTA, D-/L-Glu, possibly, D-/L-Ara and CABu) interact with multiple F(δ-)-C(δ+) bonds of the three fod ligands but do not directly coordinate with Eu III .
For comparison, FT-IR spectra between Eu(fod) 3 , D-Glu, and Eu(fod) 3 mixed with D-Glu in the ranges of 2,700 and 3,700 cm −1 , 2,800 and 3,100 cm −1 , and 1,000 and 2,000 cm −1 are shown in Figures S15A-C, SM. We observed no noticeable frequency shifts in ν(C-H) at 2,850-3,000 cm −1 and ν(C-F) at 1,250-1,050 cm −1 . Since the postulated C-F/H-C and C-H/O-C interactions are very weak, the resulting frequency shifts might be minimal, possibly, within 10 cm −1 . One ν as (C-H) at 2973 cm −1 and ν s (C-H) at 2873 cm −1 due to methyl groups of fod in the absence of D-Glu shift to lower frequencies at 2967 cm −1 by 7 cm −1 and 2871 by 2 cm −1 , respectively (Figure S15B, SM). These small shifts may be the consequence of the C-H/O-C interactions. On the other hand, ν(C=O) at ∼1,750 cm −1 characteristic of five ester group of D-Glu does not coordinate to Eu III directly because there are no noticeable frequency shifts (Figure S15C, FIGURE 8 | Comparison of 1 H-NMR spectra in CDCl 3 at room temperature between (A) D-Glu (blue) and Eu(fod) 3 with D-Glu (red) and between (B) CTA (blue) and Eu(fod) 3 with CTA (red). Sharper and broader 1 H-NMR peaks at ∼1.5 ppm are assumed to be free water in CDCl 3 and bounded waters at Eu(fod) 3 , respectively. Although 20 mg of D-Glu and 10 mg of Eu(fod) 3 were able to co-dissolve in 0.6 mL of CDCl 3 , 10 mg of CTA and 10 mg of Eu(fod) 3 were co-dissolved in 0.6 mL of CDCl 3 because of the limited solubility of 20 mg CTA in 0.6 ml of CDCl 3 . 2 | Comparison of chemical shifts in the 1 H-NMR spectra in CDCl 3 at room temperature between (a) D-Glu without and with Eu(fod) 3 and between (b) CTA without and with Eu(fod) 3 , whereas (+)-sign stands for the downfield shift. SM). Although, in the absence of D-Glu, Eu(fod) 3 has one broad and one shoulder ν(C=O) band at 1621 cm −1 and 1594 cm −1 due to the β-diketonate, in the presence of D-Glu, the shoulder ν(C=O) may disappear and merge to 1621 cm −1 or shift to 1642 cm −1 due to specific alteration of β-diketonates. Other frequency shifts such as ν(C-O-C) at ∼1200 cm −1 and 1150 cm −1 of the ester groups and glucose rings of D-Glu are not apparent because of significant overlapping with other intense ν(C-F) bands. The C-F/H-C interactions are not obvious due to the significant overlapping. Figures S16A-C, SM compare the FT-IR spectra between Eu(fod) 3 , CABu, and Eu(fod) 3 with CABu in the ranges of 2,700 and 3,700 cm −1 , 2,800 and 3,100 cm −1 , and 1,000 and 2,000 cm −1 . Similarly, minimal frequency shifts in ν(C-H) at 2,850-3,000 cm −1 can be seen due to the postulated C-F/H-C and C-H/O-C interactions (Figure S16B, SM). The ν as (C-H) band at 2,973 cm −1 and ν s (C-H) at 2,873 cm −1 of fod methyl groups in the absence of CABu shift to lower frequencies at 2,967 cm −1 by 7 cm −1 and conversely higher frequency of 2,878 cm −1 by 5 cm −1 , respectively (Figure S16B, SM). These small shifts may arise from the C-H/O-C interactions. On the other hand, a broad ν(C=O) band at ∼1,630 cm −1 characteristic of the β-diketonate split into two ν(C=O) bands at 1,643 cm −1 and 1,623 cm −1 , suggesting specific structural alterations of the β-diketonates by the ester groups and/or ethers of CABu. However, no noticeable frequency shifts of ν(C-O-C) at ∼1,200 cm −1 and 1,150 cm −1 of the ester groups of CABu are not seen because of the spectral overlapping with the intense ν(C-F) bands (Figure S16C, SM).

Photodynamics of Eu(fod) 3 , Eu(dpm) 3 , and Tb(dpm) 3 in CTA and CABu Films
Lifetimes of Eu(fod) 3 , Eu(dpm) 3 , and Tb(dpm) 3 species embedded to CTA and CABu films excited at an N 2 pulsed laser 337.1 nm are summarized in Table 3 based on decay curves (semilog and linear plots) of these emitters (Figures S17A-K, SM). The decay times (τ ) of these emitters in CABu are somewhat long by the magnitude of 15-56 % compared to those in CTA. Possibly, these emitters have differently interacted with CTA and CABu, that depends on the nature of alkyl esters. Alternatively, regardless of CTA and CABu, the values of τ belong to in the order of Eu(fod) 3 , Eu(dpm) 3 , and Tb(dpm) 3 , depending on the nature of lanthanides and ligands.
Mulliken Charges of Sc III Tris(β-diketonate) as Models of Eu III /Tb III Tris(β-diketonate), D-Glu and D-Glu Dimer as a Model of CTA Obtained With MP2 (6-311G) Calculation To theoretically discuss possible intermolecular interactions, we calculated Mulliken charges (Mulliken, 1955) by the Møller-Plesset second-order perturbation theory (MP2) (Møller and Plesset, 1934;Head-Gordon et al., 1988) (6-311G basis set) method of the model compounds optimized by MM (UFF force filed), followed by DFT [6-31G(d)] methods (Frisch et al., 2013). Time-consuming MP2 calculation allows for reliable Mulliken charges compared to DFT calculation. Figure 9 and Figure  In the previous paper (Jalilah et al., 2018), we proposed that the surplus charge neutralization obtained with Mulliken charges is a driving force of attractive forces between several ligands in the lanthanide complexes and CPL-inducible chiral substances. The fluorine atoms of Sc(fod) 3 have negative Mulliken charges ranging from −0.341 to −0.373; conversely, the hydrogen atoms of Sc(fod) 3 and Sc(dpm) 3 have positive Mulliken charges ranging from +0.153 to +0.193.

The Degree of Chirogenesis and the Pfeiffer Effects
Since the serendipitous finding by an anomaly in an optical rotation of chiral substances in the presence of optically inactive labile metal ions in aqueous solutions Quehl, 1931, 1932), the chirogenesis in the GS and ES from optically inactive labile metal complexes induced by chiral additives has been often appeared in the titles of several papers in the past and currently: e.g., Pfeiffer effect (Kirschner and Ahmad, 1968;Mayer and Brasted, 1973;Schipper, 1978;Brittain, 1982Brittain, , 1984Kirschner and Bakkar, 1982;Lunkley et al., 2018); outer-sphere coordination and complexation (Mason and Norman, 1965;Madaras and Brittain, 1980;Kirschner and Bakkar, 1982); second-sphere coordination (Colquhoun et al., 1986). Pfeiffer effect and/or outer-sphere/second-sphere coordination are mainly investigated in their solution states of the metal complexes.
An equilibrium shift from dynamic racemic mixtures ( :Λ = 50/50) of labile metal complexes is responsible for the Pfeiffer effect and chirogenesis by outer-sphere/secondsphere coordination. A barrier height of racemization should be rather small to permit dynamic racemization at ambient temperatures. In 1975, Schipper theoretically discussed chemical discrimination between racemic substances (A' and A") and chiral substance B surrounded by achiral solvent (Schipper, 1975) as a model of the Pfeiffer effect. In the hypothetical system, an exothermic enthalpic gain h /T is acquired to compensate an entropic loss s . The long-range interactions in dynamically dissociate system is needed to overcome thermal fluctuation k B T.
When a similar analysis was applied to Eu(fod) 3 and nonrigid chiral CTA, we obtained K b = 0.09 M −1 (Figures S21A,B), SM). This K b value is more significant than that of α-pinene by two orders of magnitude and it is reasonable because a chiral repeating unit in CTA (though non-rigid and floppy) contains five oxygen atoms with the substantial (-)-Mulliken charges responsible for multiple pseudo chiral O/H-C interactions with achiral ligands (Figure 9 and Figure S18, SM). When solidified film and high concentrations of CTA were employed as chirality inducible scaffolds and platforms, the degree of chirogenesis was characterizable as g lum values from CPL spectral characteristics. This idea can be extended to CABu and other four monosaccharide alkyl esters ( Table 1).

CONCLUSION
Two polysaccharide alkyl esters (CTA and CABu) as the films, two enantiopairs of monosaccharide permethyl esters (D-/L-Glu and D-/L-Ara) as the films, and (1S)-/(1R)-α-pinene in solution were capable of transferring their chirality to several optically inactive Eu III and Tb III tris(β-diketonate) (= fod and dpm), which impart the shining CPL characteristics at 4f-4f transitions. The greatest g lum values at 5 D 0 → 7 F 1 transitions (λ ex = 315 nm) of Eu(fod) 3 in CABu and CTA films are +0.067 and +0.046, respectively. Tb(fod) 3 in CABu and CTA exhibited moderately large g lum values of +0.008 and +0.004 at 5 D 4 → 7 F 5 transitions (λ ex = 315 nm), respectively. Meanwhile, D-/L-Glu and D-/L-Ara films induced weaker g lum values for Eu(fod) 3 , Tb(fod) 3 , and Tb(dpm) 3 . CTA and CABu induced CPL signals more efficiently for Eu(fod) 3 than D-/L-Glu and D-/L-Ara. Noticeably, the chirality of α-pinene enabled Tb(dpm) 3 to shine similar CPL characteristics of Tb(fod) 3 in CABu. However, Eu(dpm) 3 in CABu films and α-pinene did not reveal CPL. From the analyses of solution 1 H-/ 19 F-NMR, solid-state 13 C-NMR with the help of MP2 (6-311G basis set) calculation, we propose that the surplus charge neutralization evaluated by the opposite Mulliken charges between H(δ+)-C(δ-) bonds of the poly-and monosaccharides and F(δ-)-C(δ+) bonds of the fluorinated ligands are the attractive driving forces to induce the CPL characteristics of Tb(fod) 3 and Eu(fod) 3 . The present knowledge should enable the fabrication of films, sheets, fibers, and nanocomposites that emit Eu III -origin red-color and Tb III -origin green-color CPL spectra with narrow spectral bandwidths. As demonstrated, Eu III (fod) 3 and Tb III (dpm) 3 containing transparent CTA films deposited on the Tempax substrate displayed clear Eu III -origin red-color and Tb III -origin green-color emissions upon 365-nm excitation (see, photographs in Figure S22, SM). These materials were obtainable by a chiral ligand-free process at room temperature by co-mixing soluble polysaccharide derivatives (and bacterial cellulose) and several optically inactive Eu III /Tb III complexes. The challenging issue remains to boost the rather small g lum values [Eu(fod) 3 : 6 × 10 −2 at 593 nm and Tb(fod) 3 : 0.8 × 10 −2 at 540 nm] toward an ultimate g lum = ±2.0, i.e., obtaining purely left-or right-CPL forms (Eliel and Wilen, 1994). Symmetry-oriented designing of emitters should be considered by precisely controlling topological shape associated with an efficient lens and an optofluidic effect (Wang et al., 2007;Di Pietro and Di Bari, 2012;Kruk et al., 2014;Khorasaninejad et al., 2016;Yeung et al., 2017;Tanaka et al., 2018;Zhou et al., 2019). A deeper understanding of the Pfeiffer effect in the GS and ES, magnetic dipole transitions of Eu III and Tb III complexes, colloidal aggregations, hybridization by other chromophores/luminophores, chain-like polymers, supramolecular motifs and polymers, and nature of oligo-and polysaccharides with conformational freedom are the next challenges (Wormald et al., 2002;Zou et al., 2019) in addition to several approaches to elaborate CPL and CD functions as polymeric colloids, revealing moderately high |g lum | and |g abs | values (>10 −2 -10 −1 ) in the range of 300 and 800 nm (Nakano and Fujiki, 2011;Duong and Fujiki, 2017;Fujiki and Yoshimoto, 2017;Wang et al., 2017).
However, our approaches of chirogenesis from optically inactive labile Ln III tris(β-diketonate) (Ln: lanthanide) induced by soluble chiral biomaterials is very limited to common organic solvents of Ln III complexes and biomaterials. If watersoluble optically inactive labile Ln III complexes are designed in the future, our approaches are applicable as a thin film state to sense and detect various water-soluble chiral substances including biomaterials, drug, medicine, pesticide, and virus consisting of illness-causing single-strand (ss)/double-strand (ds) RNA.

DATA AVAILABILITY STATEMENT
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation. Requests for the original CPL/CPLE/CD/UV-visible/NMR/IR spectral and photodynamic data sets, followed by the processed data (#.qpc with #.qda and #.txt) using KaleidaGraph (mac, ver 4.53), and the calculation results (#.com, #.log, and #.chk up to 20 GB) of Gaussian09 (mac) to support the conclusion of this article should be sent to MF (fujikim@ms.naist.jp).

AUTHOR CONTRIBUTIONS
All the authors co-designed this work. MF, LW, and AJ co-wrote the paper. LW, NO, AJ, and MF co-measured and co-analyzed the CPL, CPLE, CD, UV-visible, PL, and PLE spectra of Tb(fod) 3 , Eu(fod) 3 , Tb(dpm) 3 , and Eu(dpm) 3 and other several lanthanide complexes in the presence of chiral additives and chiral solvents. LW and FA co-acquired and co-analyzed ss-13 C{ 1 H}-FT-NMR and solution 1 H-NMR/ 19 F-NMR spectra. FA conducted the elemental analysis of the products. MF performed MP2 and DFT calculations. LW, NO, AJ, AO, SO, HK, and MF contributed to a joint project of emerging CPL spectra from achiral organic, polymeric, and lanthanide luminophores endowed with chiral polymers and chiral solvents. All authors discussed the data and commented on the manuscript. All authors have given approval to the final version of the manuscript. These authors contributed equally. The manuscript was written through contributions of all authors.

ACKNOWLEDGMENTS
MF and AJ owe a debt of gratitude to Prof. Victor Borovcov (South-Central University for Nationalities, Wuhan, China) for giving us the opportunity to contribute to this special issue. LW and MF acknowledge Prof. Wei Zhang (Soochow University, China) for providing D-Glu. We thank Dr. Sibo Guo for assisting the chiroptical measurements. We thank Yoshiko Nishikawa for measuring and analyzing mass spectral data sets of D-/L-Glu, D-/L-Ara, Tb(fod) 3 , Eu(fod) 3 , Tb(dpm) 3 , and Eu(dpm) 3 , Yasuo Okajima for measuring photodynamics of Eu(fod) 3 , Eu(dpm) 3 , and Tb(dpm) 3 embedded to CTA and CABu specimens, and Prof. Tsuyoshi Ando for the permission to use Perkin Elmer FT-IR spectrometer.