Novel Fluorinated Indanone, Tetralone and Naphthone Derivatives: Synthesis and Unique Structural Features

Several fluorinated and trifluoromethylated indanone, tetralone and naphthone derivatives have been prepared via Claisen condensations and selective fluorinations in yields ranging from 22–60%. In addition, we report the synthesis of new, selectively fluorinated bindones in yields ranging from 72–92%. Of particular interest is the fluorination and trifluoroacetylation regiochemistry observed in these fluorinated products. We also note unusual transformations including a novel one pot, dual trifluoroacetylation, trifluoroacetylnaphthone synthesis via a deacetylation as well as an acetyl-trifluoroacetyl group exchange. Solid-state structural features exhibited by these compounds were investigated using crystallographic methods. Crystallographic results, supported by spectroscopic data, show that trifluoroacetylated ketones prefer a chelated cis-enol form whereas fluorinated bindone products exist primarily as the cross-conjugated triketo form.


Introduction
Molecules which have medicinal, industrial and herbicidal properties are of continued interest to the pharmaceutical, chemical and agrochemical communities.For example, indanone derivatives have anticoagulant properties and are used in elaborating latent fingerprints, bindone variants comprise components of near infrared dyes while certain tetralones and naphthones, ketones similar in structure to those shown in Figure 1, have demonstrated bioactive properties [1][2][3][4][5][6].Since bioactivity is known to be enhanced in many classes of fluorinated molecules [3,7], it is desirous to prepare fluorine-containing molecules with similar architecture and gain a better understanding of their structure-property relationships.Previously, we reported the preparation and structure-property relationships of acyclic fluorinated and trifluoromethylated β-diketones, precursors to a variety of heterocyclic molecules [8][9][10].While the syntheses and properties of these molecules have been investigated thoroughly, the preparation and study of selectively fluorinated, cyclic ketones containing the structural features of the molecules depicted in Figure 1 remains relatively limited [11].

Chemicals
All chemicals were obtained from the Aldrich Chemical Company, Eastman Kodak, or Fisher Chemical Company.All solvents (spectrophotometric grade) and starting materials were checked for purity by mass spectrometry prior to use.

Instrumentation
Melting points were obtained on a Mel-Temp melting point apparatus and are uncorrected.NMR data were collected using a Varian VXR-200 spectrometer with a broad band probe operating at 200.0 MHz for 1 H, 188.2 MHz for 19 F and 50.3 MHz for 13 C, and/or a Brüker Avance 300 spectrometer operating at 300.0 MHz for 1 H, 282.0 MHz for 19 F and 75.4 MHz for 13 C. Unless otherwise noted, CDCl 3 was used as the solvent and internal standard for 1 H and 13 C NMR experiments while CFCl 3 served as the internal standard for 19 F NMR experiments.All X-ray measurements were made on a Bruker-Nonius X8 Apex2 diffractometer.See the appendix for complete crystallographic experimental details.+ 5a (26%)

General Procedure for the Preparation of Trifluoromethyl-β-Diketones and Triketones [12]
A 100 mL round bottom flask equipped with a magnetic stirrer is charged with 50 mL diethyl ether and 60 mmol of sodium methoxide is added slowly.Then, 1eq (60 mmol) of trifluoromethyl ethyl acetate is added dropwise slowly while stirring.After 5 minutes, 1eq (60 mmol) of the ketone is added dropwise and stirred overnight at room temperature under a calcium chloride drying tube.The resulting solution is evaporated to dryness under reduced pressure and the solid residue dissolved in 30 mL 3M sulfuric acid.This solution is extracted with ether, and the organic layer dried over Na 2 SO 4 .The solvent is evaporated under reduced pressure and the crude diketone purified by radial chromatography.[10,13] A 100 mL round bottom flask equipped with a magnetic stirrer is charged with 40 mL CH 3 CN and the ketone (1 eq: 1-10 mmol).Then, Selectfluor ® (1-3 eq (3-30 mmol)) dissolved in 30 mL CH 3 CN is added slowly while stirring.The solution is either allowed to stir at room temperature or refluxed as required.Times range from 10-30 h.The resulting solution is evaporated to dryness under reduced pressure and the solid residue taken up in distilled water.This solution is extracted with CH 2 Cl 2 , and the organic layer dried over Na 2 SO 4 .The solvent is evaporated under reduced pressure and the crude fluorinated ketone purified by radial chromatography.

Synthesis
Compounds 1a-5a are commercially available and were used without further purification.Compounds 1b and 1c have been previously described, but were prepared according to a different method [14,15].Compound 1d is known and was synthesized via a previously described method [12,16].The remaining compounds were prepared using a modified Claisen condensation or direct fluorination with Selectfluor® [10,12,13,18,19].See Scheme 1.
Recent work by our group showed that regioselective monofluorination and geminal difluorination of acyclic β-diketones could be effected with Selectfluor ® under mild conditions without the necessity of specialized glassware or safety precautions [10,14,15].The current synthetic investigation sought to take advantage of this earlier work by probing Selectfluor ® 's efficiency and effectiveness in the monoand difluorination of 1,3-indanedione and bindone.Our efforts revealed some unexpected findings.The monofluorination of 1a proceeded with little difficulty to give 2-fluoro-1,3-indanedione (1b) in the diketonic form (as evidenced by a doublet signal (J F-H = 51.1 Hz) in the 19 F NMR at −207.3 ppm), albeit in slightly lower yield compared to fluorination achieved with 5% F 2 in N 2 [10].Diketone 1b was also successfully fluorinated (as evidenced by a singlet signal in the 19 F NMR at −125.9 ppm), delivering the geminally difluorinated product 1c in good overall yield.
We then examined whether bindone, the aldol self-condensation product of 1,3-indanedione, would react similarly to treatment with Selectfluor ® .As expected, monofluorination was achieved in high yield to give 1e as an enantiomeric triketone pair ( 19 F NMR: −182.4 ppm, J F-H = 46.0Hz), but the site of fluorination was the α-carbon adjacent to the isolated ketone rather than fluorination between the 1,3-diketone residue.Subsequent fluorination of 1e likewise yielded interesting results.Particularly noteworthy were the fluorination regioselectivity and alkene rearrangement observed during the formation of triketone 1f.We expected an outcome similar to the fluorination of 1b, but the occurrence of two distinct signals in the 19 F NMR at −137.3 ppm and −176.7 ppm ruled out geminal difluorination.Evidently, the alkene in 1e retains sufficient nucleophilic nature to permit electrophilic fluorination between the β-dicarbonyl residue.This addition, coupled with a concomitant E1-like elimination leads to 1f, rather than formation of [Δ1,2'-Biindan]-2,2-difluoro-1',3,3'-trione, shown in Scheme 1.
While preparing 2b and 2c, the previously undescribed one-pot, twin trifluoroacetylation of 2-indanone gave the dual exocyclic enol 2c in moderate yield (confirmed by the presence of a single 19 F NMR resonance at −68.5 ppm) and no 2c', Figure 2. In this case, the ethoxide base present following the condensation apparently deprotonates the unsubstituted benzylic α-hydrogen (H 3 ) rather than the more acidic α-hydrogen H 1 .
There are several possible explanations for the formation of 2c and the failure to obtain 2c'.The most plausible scenario involves initial formation of 2c'.Given the basic reaction conditions, however, we surmise that upon attachment of the second COCF 3 group, 2c' may undergo nucleophilic acyl substitution by ethoxide, reverting 2c' back to enolate A. A second possibility for the failure to obtain 2c' may be larger steric demands in the transition state leading to enolate A formation relative to that leading to enolate B. Finally, a base-promoted tautomerization from enolate A to enolate B could occur before formation of 2c', ultimately leading to 2c.We then attempted a similar strategy with β-tetralone (3a) to ascertain whether this ditrifluoroacetylation methodology could be generalized to other ketones with two acidic α-hydrogen sets.Sequential treatment of 3a with two equivalents of ethyl trifluoroacetate followed by neutralization at room temperature led to a mixture of the 3-and 1-trifluoroacetyl-2-tetralone endocyclic enols 3b and 3c, respectively; the formation of 1,3-ditrifluoroacetyl-2-tetralone was not observed.Assignment of the endocyclic enolic structures was based on the observation of a single 19 F NMR resonance at −70.6 ppm for 3b and −67.7 ppm for 3c.When the reaction workup conditions were modified by subjecting the enols 3b and 3c to an additional equivalent of base and ethyl trifluoroacetate followed by in vacuo removal of solvent at elevated temperature, we were surprised to find that aromatization occurred to give the trifluoroacetylated naphthol 3d in moderate overall yield. Figure 3 depicts a plausible route to naphthol 3d.We surmise that deprotonation of the less sterically hindered α-hydrogen enroute to 3b occurs rather than abstraction of the more acidic, benzylic α-hydrogen.Detrifluoroacetylation of triketone I followed by tautomerization of diketone II under acidic workup provides naphthol 3d.Application of Light and Hauser's method to 4a and 5a produced the cross-conjugated, dienolic 1,3,5-triketones 4b and 5b in modest yields [17].Assignment of the enolic structures was based on a combination of resonances found in their NMR spectra-(4b) 1 H: an alkene proton signal @ 6.76 ppm (1H), a broad, unresolvable singlet corresponding to the enol protons @ 15.68 ppm (2H) and 19 F: a singlet @ −72.0 ppm (3F); for (5b) 1 H: an alkene proton signal @ 6.90 ppm (1H), a singlet corresponding to the phenolic enol proton @ 14.44 (1H), a broader singlet @ 15.68 ppm (1H) corresponding to the enol adjacent to the CF 3 group and 19 F: a singlet @ −71.6 ppm (3F).Addition of D 2 O to the NMR samples of 4b and 5b resulted in rapid diminuation of the exchangeable enolic protons in the 1 H NMR. Increasing the molar ratio of ethyl trifluoroacetate:diketone to >2:1, although necessary for triketone product formation, also led to O-trifluoroacetylated by-products.Fortunately, these were easily separated by chromatography from the desired 1,3,5-triketones.Additionally, we found that when 4a was subjected to standard Claisen reaction conditions, an unintended acetyl-trifluoroacetyl group exchange occurred to give 2-trifluoroacetyl-1-tetralone (4c) in good yield along with, to our surprise, naphthol 5a.A process similar to the detrifluoroacetylation depicted in Figure 2 may be operating in these cases as well.
Treatment of 5a with Selectfluor ® demonstrated the fluorination preference of activated aromatic substrates over acetyl groups [19][20][21].The fluorinated naphthols 5c and 5d were achieved in good overall yield and a 5:1 ratio of the para:meta isomers, respectively.Ring fluorination was confirmed by the observation of resonances in the 19 F NMR spectra as singlets: −134.0 ppm (1F) and −134.2 ppm (1F) for 5c and 5d, respectively.Preferential para fluorination is in accord with the o-p directing ability of the hydroxyl group.Use of up to 5 equivalents of Selectfluor ® to effect fluorination at the acetyl carbon provided only the monofluorinated naphthols 5c and 5d.

Solid State Structural Features: X-ray Crystallography
Several of the target molecules (1d, 2c, 3c and 4c) were examined by x-ray crystallography.Crystal data and structure refinement information for 1d and 2c are recorded in Figure 4 and Table 1 while Figure 5 and Table 2 contain data for 3d and 4c.Critical bond information is listed in Table 3.The crystallographic information files for these molecules have been uploaded to the Cambridge Crystallographic Data Center and have the following control numbers: 1d: 854704, 2c: 854697, 3d: 854705 and 4c: 854706.In the case of compound 1d, the small O 3 -C 10 -C 18 -C 3 dihedral angle of −2.4° shows bindone to be nearly planar across the ring bridge in the solid state.The C-O and C 3 -C 10 bond lengths are consistent with those of typical carbonyls and alkenes, respectively and identify 1d as a cross-conjugated triketone in the solid state.For 2c, x-ray crystallography confirms the preference of a previously unreported structure in the solid state: a planar, exocyclic dienol shown in Figure 4.The weak intramolecular H-bonding normally observed in cyclic triketones is clearly supported for 2c by the interatomic O …. H-O distances of 1.75Å and 1.81Å and very short O-H bond lengths of 0.88Å and 0.90Å [11,22].The small O 1 -C 1 -C 2 -C 10 dihedral angle of −1.91° attests to the planar nature of the cyclopentanone residue.
Spectral data provided in the experimental section supports the solid-state structural data presented herein [8][9][10][23][24][25][26][27][28].A detailed examination of the absorption, vibrational and magnetic resonance spectroscopy of these molecules is underway to discern the keto-enol and enol-enol behavior of these di-and triketones in various solvent systems and where applicable in the solid-state and/or neat liquid.Those results, along with a comparative ab initio component, will be presented in a future communication.Data Collection and Processing.The sample 1d was submitted by Joseph Sloop of the Sloop research group at Georgia Gwinnett College.The sample was mounted on a nylon loop with a small amount of NVH immersion oil.All X-ray measurements were made on a Bruker-Nonius X8 Apex2 diffractometer at a temperature of 173 K.The unit cell dimensions were determined from a symmetry constrained fit of 9975 reflections with 5.0° < 2θ < 56.84°.The data collection strategy was a number of ω and ϕ scans which collected data up to 58.24° (2θ).The frame integration was performed using SAINT [29].The resulting raw data was scaled and absorption corrected using a multi-scan averaging of symmetry equivalent data using SADABS [30].
Structure Solution and Refinement.The structure was solved by direct methods using the SIR92 program [31].All non-hydrogen atoms were obtained from the initial E-map.The hydrogen atoms were introduced at idealized positions and were allowed to refine isotropically.The structural model was fit to the data using full matrix least-squares based on F 2 .The calculated structure factors included corrections for anomalous dispersion from the usual tabulation.The structure was refined using the XL program from SHELXTL [32], graphic plots were produced using the NRCVAX crystallographic program suite.Additional information and other relevant literature references can be found in the reference section of the Facility's Web page (http://www.xray.ncsu.edu).
SAINT+ [29].The resulting raw data was scaled and absorption corrected using a multi-scan averaging of symmetry equivalent data using SADABS [30].
Structure Solution and Refinement.The structure was solved by direct methods using the SIR92 program [31].All non-hydrogen atoms were obtained from the initial E-map.The hydrogen atoms were introduced at idealized positions and were allowed to refine isotropically.The structural model was fit to the data using full matrix least-squares based on F. The calculated structure factors included corrections for anomalous dispersion from the usual tabulation.The structure was refined using the LSTSQ program from NRCVAX [33], graphic plots were produced using the NRCVAX crystallographic program suite.Additional information and other relevant literature references can be found in the reference section of the Facility's Web page (http://www.xray.ncsu.edu).
2 ) / (No. of reflns.− No. of params.) ] ½       amount of NVH immersion oil.All X-ray measurements were made on a Bruker-Nonius X8 Apex2 diffractometer at a temperature of 110 K.The unit cell dimensions were determined from a symmetry constrained fit of 5859 reflections with 5.44° < 2θ < 52.66°.The data collection strategy was a number of ω and ϕ scans which collected data up to 62.92° (2θ).The frame integration was performed using SAINT [29].The resulting raw data was scaled and absorption corrected using a multi-scan averaging of symmetry equivalent data using SADABS [30].
Structure Solution and Refinement.The structure was solved by direct methods using the SIR92 program [31].All non-hydrogen atoms were obtained from the initial solution.The carbon bound hydrogen atoms were introduced at idealized positions while the hydroxy hydrogen atom position was obtained from a diffeence Fourier map.All hydrogen atoms were allowed to refine isotropically.The structural model was fit to the data using full matrix least-squares based on F 2 .The calculated structure factors included corrections for anomalous dispersion from the usual tabulation.The space group is achiral, therefore the structure has an absolute sense to it.However, the anomalous scattering signal is weak due to the absence of any atoms heavier than F, and the absolute structure could not be definitively determined.The structure was refined using the XL program from SHELXTL [32], graphic plots were produced using the NRCVAX crystallographic program suite.Additional information and other relevant literature references can be found in the reference section of the Facility's Web page (http://www.xray.ncsu.edu).Data Collection and Processing.The sample 4c was submitted by Joseph Sloop of the Sloop research group at Georgia Gwinnett College.The sample was mounted on a nylon loop with a small amount of Paratone N oil.All X-ray measurements were made on a Bruker-Nonius Kappa Axis X8 Apex2 diffractometer at a temperature of 110 K.The unit cell dimensions were determined from a symmetry constrained fit of 6416 reflections with 5.78° < 2θ < 71.38°.The data collection strategy was a number of ω and ϕ scans which collected data up to 71.58° (2θ).The frame integration was performed using SAINT [29].The resulting raw data was scaled and absorption corrected using a multi-scan averaging of symmetry equivalent data using SADABS [30].
Structure Solution and Refinement.The structure was solved by direct methods using the XS program [31].All non-hydrogen atoms were obtained from the initial solution.The hydrogen atoms were located from a difference Fourier map and were allowed to refine isotropically.The structural model was fit to the data using full matrix least-squares based on F 2 .The calculated structure factors included corrections for anomalous dispersion from the usual tabulation.The structure was refined using the XL program from SHELXTL [32], graphic plots were produced using the NRCVAX crystallographic program suite.Additional information and other relevant literature references can be found in the reference section of the Facility's Web page (http://www.xray.ncsu.edu).

Figure 4 .
Figure 4. ORTEP drawings of 1d and 2c.Ellipsoids are at the 50% probability level and hydrogen atoms were drawn with arbitrary radii for clarity.

Figure 5 .
Figure 5. ORTEP drawings of 3d and 4c.Ellipsoids are at the 50% probability level and hydrogen atoms were drawn with arbitrary radii for clarity.

Figure 6 .
Figure 6.ORTEP drawing of 1d molecule A showing naming and numbering scheme.Ellipsoids are at the 50% probability level and hydrogen atoms were drawn with arbitrary radii for clarity.

Figure 7 .
Figure 7. ORTEP drawing of 1d molecule A. Ellipsoids are at the 50% probability level and hydrogen atoms were drawn with arbitrary radii for clarity.

Figure 8 .
Figure 8. Stereoscopic ORTEP drawing of 1d molecule A. Ellipsoids are at the 50% probability level and hydrogen atoms were drawn with arbitrary radii for clarity.

Figure 9 .
Figure 9. ORTEP drawing of 1d molecule B showing naming and numbering scheme.Ellipsoids are at the 50% probability level and hydrogen atoms were drawn with arbitrary radii for clarity.

Figure 10 .
Figure 10.ORTEP drawing of 1d molecule B. Ellipsoids are at the 50% probability level and hydrogen atoms were drawn with arbitrary radii for clarity.

Figure 11 .
Figure 11.Stereoscopic ORTEP drawing of 1d molecule B. Ellipsoids are at the 50% probability level and hydrogen atoms were drawn with arbitrary radii for clarity.

Figure 12 .
Figure 12.ORTEP drawing of 2c showing naming and numbering scheme.Ellipsoids are at the 50%.

Figure 13 .
Figure 13.ORTEP drawing of 2c.Ellipsoids are at the 50% probability level and hydrogen atoms were drawn with arbitrary radii for clarity.

Figure 14 .
Figure 14.Stereoscopic ORTEP drawing of 2c.Ellipsoids are at the 50% probability level and hydrogen atoms were drawn with arbitrary radii for clarity.

Figure 15 .
Figure 15.ORTEP drawing of 3d showing naming and numbering scheme.Ellipsoids are at the 50% probability level and hydrogen atoms were drawn with arbitrary radii for clarity.

Figure 16 .
Figure 16.ORTEP drawing of 3d.Ellipsoids are at the 50% probability level and hydrogen atoms were drawn with arbitrary radii for clarity.

Figure 17 .
Figure 17.Stereoscopic ORTEP drawing of 3d.Ellipsoids are at the 50% probability level and hydrogen atoms were drawn with arbitrary radii for clarity.

Figure 18 .
Figure 18.ORTEP drawing of 4c showing naming and numbering scheme.Ellipsoids are at the 50% probability level and hydrogen atoms were drawn with arbitrary radii for clarity.

Figure 19 .
Figure 19.ORTEP drawing of 4c.Ellipsoids are at the 50% probability level and hydrogen atoms were drawn with arbitrary radii for clarity.

Figure 20 .
Figure 20.Stereoscopic ORTEP drawing of 4c.Ellipsoids are at the 50% probability level and hydrogen atoms were drawn with arbitrary radii for clarity.

Table 1 .
Crystal data and structure refinement for 1d and 2c.

Table 2 .
Crystal data and structure refinement for 3d and 4c.

Table 4 .
Summary of Crystal Data for 1d.

Table 10 .
Summary of Crystal Data for 2c.

Table 16 .
Summary of Crystal Data for 3d.

Table 20 .
Bond Angles for

Table 22 .
Summary of Crystal Data for 4c.