Synthesis of a Naphthalocyanine-Like Dye: The First Report on Zn(II)-1,6-methano[10]annulenecyanine

The synthesis of the new dye 1,6-methano[10]annulenecyanine is described. For this purpose, the 3,4-dicyano-1,6-methano[10]annulene and 3,4-carboxyimide-1,6-methano[10]annulene buildings blocks were synthesized in six to eight steps. In both cases, these building blocks were then cyclotetramerized to furnish a new Zn(II)-1,6-methano[10]annulenecyanine which presents a strong red-shifted absorption band at 800 nm and high solubility in common organic solvents.

These classes of compounds are not naturally occurring, unlike their analogues porphyrins, chlorins and bacteriochlorins [24,25]. Phthalocyanines were accidentally discovered at the beginning of the last century and only systematically studied in the 1930s by Linstead and Robertson, who finally established their structure and a general synthetic methodology [26][27][28][29][30][31]. Since then their thermal stability, low solubility and intense color took up attention. Over the past few decades many improved synthetic approaches for phthalocyanine and also naphthalocyanine derivatives have been established, with a relevant molecular diversity which allows the modulation of both chemical and photochemical properties of these dyes [26]. In general, the classic synthetic approach to obtain phthalocyanines and naphthalocyanines is the cyclotetramerization of their building blocks phthalonitriles and naphthalonitriles, respectively [26][27][28][29][30][31]. Depending on the building block substitution pattern, it is possible to modulate the photophysical properties of the phthalocyanine/naphthalocyanine of interest, fine-tuning the dye for the desired application. The literature furnishes a vast number of interesting and creative molecules using this approach [26][27][28][29][30][31].
Seeking for new NIR building blocks our group decided to explore a class of molecules called 1,x-methano [10]annulenes, especially the 1,6-methano [10]annulene ( Figure 1) [32]. These compounds have fascinated a number of scientists since Emanuel Vogel's discovery in 1964 [33]. Vogel and coworkers achieved the required planarity of fully conjugated 10-membered carbon rings by inserting a carbon bridge between C-1 and C-6 ( Figure 1), thus furnishing 1,6-methano [10]annulene as an example of an aromatic cyclodecapentaene ring. Up to now, this class of compounds was extensively studied in synthetic, theoretical and biological areas [34]. However, to our knowledge, suitable derivatives of suitable derivatives of these compounds have not been described in the literature as building blocks for phthalocyanine and naphthalocyanine-like compounds. Herein, we describe two approaches for the synthesis of a new naphthalocyanine-like compound which we name as Zn(II)-1,6-methano [10]annulenecyanine (10), using annulenenitrile 9 and annulenemaleimide 13 as building blocks (Schemes 1-5).

Synthesis
The synthesis of the building block 3,4-dicyano-1,6-methano [10]annulene (9) was carried out as shown in Scheme 1 [35]. First, cycloheptatriene (1) was mono-acetylated at low temperature by a Friedel-Crafts reaction furnishing compound 2 in 60% yield. Then, the resulting ketone 2 was acetylated again, yielding the 1,3-diketone 3 in 55% yield. The diacid 4 was obtained in 62% yield by submitting compound 3 to the haloform reaction. Subsequently, the bis-Weinreb amide 5 was prepared using N-O-dimethylhydroxylamine hydrochloride (72% yield), and the dialdehyde 6 obtained after reduction with LiAlH4 at -78 °C (69% yield). The next steps required selective olefinations of dialdehyde 6 with two different olefination agents (Scheme 1). First, compound 7 was obtained in 57% yield by a chemoselective Horner-Wadsworth-Emmons olefination of dialdehyde 6. The literature describes an olefination under phase-transfer catalysis conditions that could give compound 8, but, in our hands only HBr elimination was observed [36]. Therefore, a modified procedure to reach 8 was tested and furnished the desired compound in 30% yield. Attempts to optimize this yield were tested, but unsuccessfully. The annulenenitrile 9 was obtained by a one-pot electrocyclization followed by HBr elimination/aromatization (52% yield) (Scheme 1). The overall yield for the eight steps is 0.90%. Scheme 1. Synthetic approach to the annulenenitrile 9 building block. Herein, we describe two approaches for the synthesis of a new naphthalocyanine-like compound which we name as Zn(II)-1,6-methano [10]annulenecyanine (10), using annulenenitrile 9 and annulenemaleimide 13 as building blocks (Schemes 1-5).

Synthesis
The synthesis of the building block 3,4-dicyano-1,6-methano [10]annulene (9) was carried out as shown in Scheme 1 [35]. First, cycloheptatriene (1) was mono-acetylated at low temperature by a Friedel-Crafts reaction furnishing compound 2 in 60% yield. Then, the resulting ketone 2 was acetylated again, yielding the 1,3-diketone 3 in 55% yield. The diacid 4 was obtained in 62% yield by submitting compound 3 to the haloform reaction. Subsequently, the bis-Weinreb amide 5 was prepared using N-O-dimethylhydroxylamine hydrochloride (72% yield), and the dialdehyde 6 obtained after reduction with LiAlH 4 at −78 • C (69% yield). The next steps required selective olefinations of dialdehyde 6 with two different olefination agents (Scheme 1). First, compound 7 was obtained in 57% yield by a chemoselective Horner-Wadsworth-Emmons olefination of dialdehyde 6. The literature describes an olefination under phase-transfer catalysis conditions that could give compound 8, but, in our hands only HBr elimination was observed [36]. Therefore, a modified procedure to reach 8 was tested and furnished the desired compound in 30% yield. Attempts to optimize this yield were tested, but unsuccessfully. The annulenenitrile 9 was obtained by a one-pot electrocyclization followed by HBr elimination/aromatization (52% yield) (Scheme 1). The overall yield for the eight steps is 0.90%.
Molecules 2020, 25, 2164 2 of 13 suitable derivatives of these compounds have not been described in the literature as building blocks for phthalocyanine and naphthalocyanine-like compounds. Herein, we describe two approaches for the synthesis of a new naphthalocyanine-like compound which we name as Zn(II)-1,6-methano [10]annulenecyanine (10), using annulenenitrile 9 and annulenemaleimide 13 as building blocks (Schemes 1-5).

Synthesis
The synthesis of the building block 3,4-dicyano-1,6-methano [10]annulene (9) was carried out as shown in Scheme 1 [35]. First, cycloheptatriene (1) was mono-acetylated at low temperature by a Friedel-Crafts reaction furnishing compound 2 in 60% yield. Then, the resulting ketone 2 was acetylated again, yielding the 1,3-diketone 3 in 55% yield. The diacid 4 was obtained in 62% yield by submitting compound 3 to the haloform reaction. Subsequently, the bis-Weinreb amide 5 was prepared using N-O-dimethylhydroxylamine hydrochloride (72% yield), and the dialdehyde 6 obtained after reduction with LiAlH4 at -78 °C (69% yield). The next steps required selective olefinations of dialdehyde 6 with two different olefination agents (Scheme 1). First, compound 7 was obtained in 57% yield by a chemoselective Horner-Wadsworth-Emmons olefination of dialdehyde 6. The literature describes an olefination under phase-transfer catalysis conditions that could give compound 8, but, in our hands only HBr elimination was observed [36]. Therefore, a modified procedure to reach 8 was tested and furnished the desired compound in 30% yield. Attempts to optimize this yield were tested, but unsuccessfully. The annulenenitrile 9 was obtained by a one-pot electrocyclization followed by HBr elimination/aromatization (52% yield) (Scheme 1). The overall yield for the eight steps is 0.90%.
As the synthesis of annulenenitrile 9 involves eight steps we decided to scale-up some reactions in order to obtain gram-scale amounts of intermediates like 4 (Scheme 2) and 6 (Scheme 3). First, we considered the linear synthesis of 4 (Scheme 2) starting from 3.00 g of 1 and isolating each intermediate by column chromatography, thus obtaining 4 in a 0.95 g-scale (20.5% overall yield after three steps). Attempts to synthesize 4 from 1 with no intermediate purification by chromatography were performed starting from 50.0 g of cycloheptatriene (1) and, to our delight, the diacid 4 was obtained in 16.4% overall yield in a very short time. Additionally, only a simple final crystallization yielded 16.0 g of 4.
Molecules 2020, 25, 2164 3 of 13 As the synthesis of annulenenitrile 9 involves eight steps we decided to scale-up some reactions in order to obtain gram-scale amounts of intermediates like 4 (Scheme 2) and 6 (Scheme 3). First, we considered the linear synthesis of 4 (Scheme 2) starting from 3.00 g of 1 and isolating each intermediate by column chromatography, thus obtaining 4 in a 0.95 g-scale (20.5% overall yield after three steps). Attempts to synthesize 4 from 1 with no intermediate purification by chromatography were performed starting from 50.0 g of cycloheptatriene (1) and, to our delight, the diacid 4 was obtained in 16.4% overall yield in a very short time. Additionally, only a simple final crystallization yielded 16.0 g of 4. Next, we tested two additional optimizations. First, the synthesis of the bis-Weinreb amide 5 was carried out starting from 12.0 g of 4, thus obtaining 5 in 76% (13.4 g -Scheme 3). Furthermore, the scaled-up reduction of 3.0 g of 5 was successfully achieved, obtaining the di-aldehyde 6 in 1.16 g (69% yield). It is important to mention that the reduction of Weinreb amides requires a finely temperature-controlled reaction, with difficult scalabilities in batch conditions. After achieving improved conditions for obtaining annulenonitrile 9 in practical amounts, we tested the first cyclotetramerization condition for obtaining the desired Zn(II)-1,6methano [10]annulenecyanine (10) (Scheme 4) using N,N-dimethylethanolamine (DMAE) and zinc acetate dihydrate under thermal conditions. However, the new dye 10 was obtained in only trace amounts as determined by UV-Vis spectroscopy [5,12]. Fortunately, using zinc triflate in Next, we tested two additional optimizations. First, the synthesis of the bis-Weinreb amide 5 was carried out starting from 12.0 g of 4, thus obtaining 5 in 76% (13.4 g -Scheme 3). Furthermore, the scaled-up reduction of 3.0 g of 5 was successfully achieved, obtaining the di-aldehyde 6 in 1.16 g (69% yield). It is important to mention that the reduction of Weinreb amides requires a finely temperature-controlled reaction, with difficult scalabilities in batch conditions. Molecules 2020, 25, 2164 3 of 13 As the synthesis of annulenenitrile 9 involves eight steps we decided to scale-up some reactions in order to obtain gram-scale amounts of intermediates like 4 (Scheme 2) and 6 (Scheme 3). First, we considered the linear synthesis of 4 (Scheme 2) starting from 3.00 g of 1 and isolating each intermediate by column chromatography, thus obtaining 4 in a 0.95 g-scale (20.5% overall yield after three steps). Attempts to synthesize 4 from 1 with no intermediate purification by chromatography were performed starting from 50.0 g of cycloheptatriene (1) and, to our delight, the diacid 4 was obtained in 16.4% overall yield in a very short time. Additionally, only a simple final crystallization yielded 16.0 g of 4.
Scheme 2. Scale-up of the intermediate 4.
Next, we tested two additional optimizations. First, the synthesis of the bis-Weinreb amide 5 was carried out starting from 12.0 g of 4, thus obtaining 5 in 76% (13.4 g -Scheme 3). Furthermore, the scaled-up reduction of 3.0 g of 5 was successfully achieved, obtaining the di-aldehyde 6 in 1.16 g (69% yield). It is important to mention that the reduction of Weinreb amides requires a finely temperature-controlled reaction, with difficult scalabilities in batch conditions. After achieving improved conditions for obtaining annulenonitrile 9 in practical amounts, we tested the first cyclotetramerization condition for obtaining the desired Zn(II)-1,6methano [10]annulenecyanine (10) (Scheme 4) using N,N-dimethylethanolamine (DMAE) and zinc acetate dihydrate under thermal conditions. However, the new dye 10 was obtained in only trace amounts as determined by UV-Vis spectroscopy [5,12]. Fortunately, using zinc triflate in After achieving improved conditions for obtaining annulenonitrile 9 in practical amounts, we tested the first cyclotetramerization condition for obtaining the desired Zn(II)-1,6-methano [10]annulenecyanine (10) (Scheme 4) using N,N-dimethylethanolamine (DMAE) and zinc acetate dihydrate under thermal conditions. However, the new dye 10 was obtained in only trace amounts as determined by UV-Vis spectroscopy [5,12]. Fortunately, using zinc triflate in hexamethyldisilazane (HMDS) and N,N-dimethylformamide (DMF) at 100 • C for 24 h yielded 10 in 63% yield after purification by chromatography [37]. hexamethyldisilazane (HMDS) and N,N-dimethylformamide (DMF) at 100 °C for 24h yielded 10 in 63% yield after purification by chromatography [37].     Attempts to perform the separation of the diastereoisomers of 10 were tested by HPLC, but unsuccessfully, making the NMR characterizations even more difficult. We tested different deuterated solvents (THF-d8, DMF-d7, acetone-d6 and CDCl3) but in all cases aggregation and complex mixtures of signals were observed making it difficult to complete assignments. Despite this, it was possible to identify in the 1 H-NMR signals at the aromatic region 7.50-8.30 ppm corresponding to the hydrogens of the macrocycle periphery, and multiplets at 0.82-0.92 ppm which are consistent with the expected for CH2-bridged [10]annulenes (signals at upfield region-see examples of the precursors 9 and 13 in the supporting information). It is important to comment that difficulties for characterizations by NMR of naphthalocyanine and phthalocyanine derivatives are well-known in the literature, and the evidence presented here is consistent with the expected for the new dye 10.
We then decided to find improvements for our original linear synthetic approach (Scheme 1 + 4, 0.57% overall yield, nine steps).
Our alternative approach for the synthesis of 10 uses the precursor annulenemaleimide 13 (Scheme 5). For the synthesis of 13 we decided to use the same di-aldehyde intermediate 6, which in a one-pot Witting olefination with the ylide 12 and a 10π electrocyclization and dehydration yielded 13 in 56% yield [38,39]. Subsequently, compound 13 was submitted to the same cyclotetramerization conditions and furnished 10 in 16% yield. Attempts to optimize this last step were carried out, but no better results were achieved. Overall, the diastereomeric mixture of dye 10 is now obtained in seven steps and 0.91% overall yield.
The fluorescence quantum yield (ΦF) of 10 was determined using as standard the corresponding zinc naphthalocyanine by exiting both at 350 nm ( Figure 4). We found the ΦF for 10 as being 0.01 with an emission band at 820 nm, a common value for organic compounds which present aggregation and dissipate energy in solution by non-radiant processes [40,41]. Therefore, the fluorescence technique Attempts to perform the separation of the diastereoisomers of 10 were tested by HPLC, but unsuccessfully, making the NMR characterizations even more difficult. We tested different deuterated solvents (THF-d 8 , DMF-d 7 , acetone-d 6 and CDCl 3 ) but in all cases aggregation and complex mixtures of signals were observed making it difficult to complete assignments. Despite this, it was possible to identify in the 1 H-NMR signals at the aromatic region 7.50-8.30 ppm corresponding to the hydrogens of the macrocycle periphery, and multiplets at 0.82-0.92 ppm which are consistent with the expected for CH 2 -bridged [10]annulenes (signals at upfield region-see examples of the precursors 9 and 13 in the supporting information). It is important to comment that difficulties for characterizations by NMR of naphthalocyanine and phthalocyanine derivatives are well-known in the literature, and the evidence presented here is consistent with the expected for the new dye 10.
We then decided to find improvements for our original linear synthetic approach (Schemes 1-4, 0.57% overall yield, nine steps).
Our alternative approach for the synthesis of 10 uses the precursor annulenemaleimide 13 (Scheme 5). For the synthesis of 13 we decided to use the same di-aldehyde intermediate 6, which in a one-pot Witting olefination with the ylide 12 and a 10π electrocyclization and dehydration yielded 13 in 56% yield [38,39]. Subsequently, compound 13 was submitted to the same cyclotetramerization conditions and furnished 10 in 16% yield. Attempts to optimize this last step were carried out, but no better results were achieved. Overall, the diastereomeric mixture of dye 10 is now obtained in seven steps and 0.91% overall yield.
The fluorescence quantum yield (Φ F ) of 10 was determined using as standard the corresponding zinc naphthalocyanine by exiting both at 350 nm ( Figure 4). We found the Φ F for 10 as being 0.01 with an emission band at 820 nm, a common value for organic compounds which present aggregation and dissipate energy in solution by non-radiant processes [40,41]. Therefore, the fluorescence technique reinforces that comprehensive studies on aggregation should be conducted in the future, in order to understand better the self-association properties of 10 (being a mixture of diastereoisomers) in solutions. These additional studies will be essential for using this new dye as a photosensitizer and for NIR applications. reinforces that comprehensive studies on aggregation should be conducted in the future, in order to understand better the self-association properties of 10 (being a mixture of diastereoisomers) in solutions. These additional studies will be essential for using this new dye as a photosensitizer and for NIR applications. Overall, this first report on the dye 10 opens up many possibilities for the synthesis of hybrid systems with modulated chemical, photochemical and photophysical properties.  Overall, this first report on the dye 10 opens up many possibilities for the synthesis of hybrid systems with modulated chemical, photochemical and photophysical properties. reinforces that comprehensive studies on aggregation should be conducted in the future, in order to understand better the self-association properties of 10 (being a mixture of diastereoisomers) in solutions. These additional studies will be essential for using this new dye as a photosensitizer and for NIR applications. Overall, this first report on the dye 10 opens up many possibilities for the synthesis of hybrid systems with modulated chemical, photochemical and photophysical properties.

Materials and Methods
All reagents and starting materials and solvents were purchased from commercial sources and used as received or purified when necessary. Some reactions were carried out under an argon atmosphere as specified in the experimental procedures (see the supporting information). For NMR spectra (performed in CDCl 3 or in DMSO-d 6 solutions) tetramethylsilane was used as internal reference for 1 H (0 ppm), and C-D coupling signal as internal reference for 13 C (CDCl 3 -77.0 ppm and DMSO-39.5 ppm).
Flash chromatography was carried out using silica gel (230-400 mesh). Infrared spectra were registered using KBr cells for liquid (films) and KBr pellets for solids. Fluorescence emission spectra were recorded using 1 cm optical length cuvettes at 25 • C and N,N-dimethylformamide as solvent.
Analytical TLC was carried out on precoated aluminum sheets with silica gel (0.2 mm thick). UV-Vis analyses were performed using a double beam spectrometer with 0.1 nm of resolution. High resolution mass spectrometry was carried out on a MALDI-TOF for compound 10, and ESI-TOF for compounds 2, 3, 4, 5, 6, 7, 8, 9 and 13.
Experimental details, spectroscopic and spectrometric data of all key compounds are available online in the supporting information. 1,1'-(cyclohepta-3,5,7-triene-1,3-diyl)diethanone (3): In a suspension containing 1.49 g(11.2 mmol) of aluminium chloride and 6.00 mL of dry dichloromethane at 0 • C, 0.79 mL (870 mg; 11.2 mmol) of acetyl chloride was added under an argon atmosphere. After 5 min, 500 mg (3.73 mmol) of 2 was added. Then, the cooling bath was removed and the reaction mixture heated to 55 • C for 3 h. After this period, the reaction mixture was cooled to 0 • C and 5 mL of water at 5 • C was added. The reaction mixture was neutralized with sodium bicarbonate solution and filtered in a sintered funnel. The filtrate was washed with ethyl acetate (5 × 50 mL). The solvent was removed under vacuum and the organic residue (a brown oil) was purified by column chromatography on silica gel using toluene: ethyl acetate (95:5) as eluent, furnishing a yellow oil as product. Yield: 50% (333 mg; 1.89 mmol). Cyclohepta-3,5,7-triene-1,3-dicarbaldehyde (6): To a suspension containing 2.73 g (71.9 mmol) of lithium aluminum hydride and 180 mL of anhydrous tetrahydrofuran at −78 • C under an argon atmosphere, amide 5 (3.00 g; 11.3 mmol) previously dissolved in 80 mL of anhydrous tetrahydrofuran was added dropwise (ca 80 min). The reaction was stirred for 20 min, 100 mL of 0.50 mM aqueous solution of potassium bisulphate was added and the reaction allowed to reach 0 • C. Then 200 mL of 5% aqueous solution (wt/v) of citric acid was added and the reaction mixture was extracted with dichloromethane (3 × 100 mL). The organic layer was dried over anhydrous sodium sulphate, filtered and the solvent removed under vacuum (thermal bath at 20 • C). The aldehyde 6 was crystalized in an ethyl acetate/pentane mixture, furnishing a yellow-pale solid (999 mg; 6.73 mmol). The mother liquor was purified by chromatography on silica gel using CH 2 Cl 2 → CH 2 Cl 2 :AcOEt (   Cl and the reaction mixture was allowed to reach room temperature. The reaction mixture was extracted with dichloromethane (3 × 100 mL) and the organic layer washed with 50 mL of water. The organic layer was dried over anhydrous Na 2 SO 4 , filtered and the solvent removed under vacuum. The organic residue was purified by column chromatography on silica gel using the gradient hexane to hexane: AcOEt (9.5: 0.5) as eluent, furnishing a yellow solid as product. Zn(II)-1,6-methano [10]annulenecyanine (10) via annulenonitrile (9): To a high pressure glass tube under an argon atmosphere were added 50.0 mg (260 µmol) annulenonitrile 9, 24.2 mg (60.0 µmol) of zinc (II) triflate -Zn(OTf) 2 , 114 µL (87.8 mg; 540 µmol) of hexamethyldisilazane (HMDS) and 266 µL (253 mg; 3.46 mmol) of N,N-dimethylformamide (DMF). The reaction mixture was stirred at 120 • C for 24 h. After this period, the solvent was removed and the organic residue was purified by chromatography on silica gel utilizing CH 2 Cl 2 : MeOH (9.5:0.5) as eluent furnishing a green solid as product. For additional purification, it was necessary to utilize preparative TLC utilizing CH 2 Cl 2 :MeOH (9:1) as eluent. Yield: 63% (34.

Conclusions
We have developed two different approaches for the first total synthesis of Zn(II)-1,6-methano [10]annulenecyanine (10). Multistep synthetic approaches for naphthalocyanine and phthalocyanine derivatives always presents low overall yields but are very necessary for dye-discovery with improved photochemical properties, particularly compounds with NIR absorption bands. This synthesis is the first part of the dye-discovery process and many additional photochemical and photophysical studies are necessary to achieve all the potential of 10 and related compounds. It is important to highlight that only preliminary studies on aggregation are reported in this communication. However, as mentioned before, this phenomenon should be evaluated in different solvents, with different techniques and, if possible, using the separated diastereoisomers, before presenting conclusions on self-association properties of 10 in solutions.
Herein, we have also demonstrated for the first time the potential of bridged-annulene derivatives as precursors for phthalocyanine and naphthalocyanine-like dyes, thus opening up many possibilities for the synthesis of hybrid structures with common precursors like phthalimides and phthalonitriles.
Supplementary Materials: The following are available online. It contains details and characterization data along with copies of the 1 H-NMR and 13 C-NMR spectra and high-resolution mass spectra for compounds.