Representative experimental procedures for the addition of amine nucleophiles to tricarbonyl(tropone)iron and subsequent demetallation of the resulting complexes are presented in detail.
aza-Michael adducts of tricarbonyl(tropone)iron are synthesized by two different methods. Primary aliphatic amines and cyclic secondary amines participate in a direct aza-Michael reaction with tricarbonyl(tropone)iron under solvent-free conditions. Less nucleophilic aniline derivatives and more hindered secondary amines add efficiently to the cationic tropone complex formed by protonation of tricarbonyl(tropone)iron. While the protocol utilizing the cationic complex is less efficient overall for accessing the aza-Michael adducts than the direct, solvent-free addition to the neutral complex, it allows the use of a broader range of amine nucleophiles. Following protection of the amine of the aza-Michael adduct as a tert-butyl carbamate, the diene is decomplexed from the iron tricarbonyl fragment upon treatment with cerium(IV) ammonium nitrate to provide derivatives of 6-aminocyclohepta-2,4-dien-1-one. These products can serve as precursors to diverse compounds containing a seven-membered carbocyclic ring. Because the demetallation requires protection of the amine as a carbamate, the aza-Michael adducts of secondary amines cannot be decomplexed using the protocol described here.
Structurally complex amines containing a seven-membered carbocyclic ring are common to a number of biologically active molecules. Notable examples include the tropane alkaloids1 and several members of the Lycopodium2, Daphniphyllum3, and monoterpenoid indole alkaloid4 families. However, such compounds are often more difficult to synthesize compared to compounds of similar complexity containing only five- or six-membered rings. Thus, we sought to develop a new avenue towards such compounds by attaching diverse amine nucleophiles to tropone5. The resulting adduct contains several functional handles for subsequent synthetic elaboration to diverse complex seven-membered ring-containing scaffolds that would be otherwise difficult to access.
While previous work with tropone6,7 suggests that it would not be suitable for such a transformation, the related organometallic complex tricarbonyl(tropone)iron8 (1, Figure 1) has proven to be a versatile synthetic building block that has been utilized in the synthesis of a number of natural products and complex molecules9,10,11,12,13. Furthermore, the uncomplexed double bond of tricarbonyl(tropone)iron has been shown to behave similar to an α,β-unsaturated ketone in reactions with, for example, dienes14,15, tetrazines16, nitrile oxides17, diazoalkanes8,10, and organocopper reagents11. Thus, we envisioned that an aza-Michael reaction of tricarbonyl(tropone)iron would provide an efficient entry to synthetically valuable aminated tropone derivatives.
Eisenstadt had previously reported that, following protonation of tricarbonyl(tropone)iron, the resulting cationic complex 2 (Figure 1) could undergo nucleophilic attack by aniline or tert-butylamine to produce aminated derivatives of the tropone iron complex.18 However, the synthetic potential of this method remains unrealized. Indeed, no additions of other amines had been reported, and the demetallation of those products was not explored in Eisenstadt’s report. We have adapted this protocol to demonstrate the addition of a wide variety of amine nucleophiles.
We also describe a method for direct aza-Michael additions to tricarbonyl(tropone)iron (Figure 2), which does not require synthesis of the cationic complex and generally proceeds in higher yields compared to the previously reported method. We also report herein a protocol for the demetallation of the resulting adducts. Overall, this protocol provides formal aza-Michael adducts of tropone in four steps from tropone (and three steps from the known iron complex).
1. Synthesis of tricarbonyl(tropone)iron (1)19
2. Synthesis of tricarbonyl(5-ketocycloheptadienyl)iron tetrafluoroborate (2)21
3. Synthesis of aza-Michael adduct 4: Tricarbonyl[(2-5-h)-6-((2-phenylethyl)amino)cyclohepta-2,4-dien-1-one]iron
4. Synthesis of tricarbonyl[(2-5-h)-6-(2-methylanilino)cyclohepta-2,4-dien-1-one]iron (3)
5. Protection of amine 4 as a tert-butyl carbamate
6. Synthesis of tert-butyl (6-oxocyclohepta-2,4-dien-1-yl)(2-phenylethyl) carbamate (6)
All novel compounds in this study were characterized by 1H and 13C NMR spectroscopy and high resolution mass spectrometry. Previously reported compounds were characterized by 1H NMR spectroscopy. NMR data for representative compounds are described in this section.
The 1H NMR spectrum of tricarbonyl(tropone)iron is shown in Figure 3. The protons of the η4-diene ligand give rise to the signals at 6.39 ppm (2 H), 3.19 ppm, and 2.75 ppm. The protons from the uncomplexed double bond appear at 6.58 and 5.05 ppm.
The progress of the aza-Michael addition is monitored via 1H NMR by observing the disappearance of the signals from the uncomplexed double bond and a characteristic change in the chemical shift of the two furthest downfield η4-diene protons from around 6.4 ppm to two well separated signals that typically appear between 5.3 and 6.0 ppm (see Figure 3 and Figure 4). Furthermore, the aza-Michael adduct features signals corresponding to the two diastereotopic methylene protons (adjacent to the ketone within the seven-membered ring), which typically appear between 1.5 and 2.5 ppm.
Direct aza-Michael additions to tricarbonyl(tropone)iron generally proceeded in 60-95% yield, depending on the amine substrate (see Discussion). Secondary cyclic amines tend to give somewhat higher yields than primary aliphatic amines, possibly due to a greater resistance to decomposition during purification.
1H NMR data for the cationic complex (in CD3CN) is shown in Figure 5 and features seven distinct multiplets. It should be noted that the complex decomposes over time in CD3CN. However, the dried solid tetrafluoroborate complex can be stored indefinitely under ambient conditions. Figure 6 shows 1H and 13C NMR data for the o-toluidine adduct 3, prepared via the cationic complex 2 (Figure 1), which contains the same features described above for the phenethylamine adduct 4.
Figure 7 shows 1H and 13C NMR spectra of tert-butyl carbamate 5. The 1H NMR spectrum is characterized by its broad peaks, caused by slow rotation of the carbamate C-N bond relative to the NMR time scale. In addition, the presence of the tert-butyl carbamate is evident from the large singlet at 1.5 ppm from the tert-butyl protons, as well as the signal at 154.3 ppm in the 13C NMR spectrum corresponding to the carbonyl carbon of the carbamate group.
Upon decomplexation of the diene from the iron, the most notable aspect of the 1H NMR spectrum (Figure 8) is the presence of four signals between 5.75 and 6.75 ppm, corresponding to the protons from the uncomplexed diene.
Figure 1. Synthesis of 3 from tricarbonyl(tropone)iron via cationic complex 2. Tricarbonyl(tropone)iron is converted to cationic complex 2 in two steps, which was followed by nucleophilic addition of ortho-toluidine to the complex. Please click here to view a larger version of this figure.
Figure 2. Synthesis of formal tropone aza-Michael adduct 6. Direct aza-Michael reaction of tricarbonyl(tropone)iron and phenethylamine was followed by amine protection and oxidative demetallation. Please click here to view a larger version of this figure.
Figure 3. 1H NMR spectrum (solvent: CDCl3) of tricarbonyl(tropone)iron 1. The peaks at 6.59 ppm and 5.05 ppm correspond to the uncomplexed alkene hydrogens while those 6.39 ppm (2H), 3.19 ppm, and 2.75 ppm arise from the iron-complexed diene. Please click here to view a larger version of this figure.
Figure 4. Spectral data for iron complex 4. (a) 1H NMR spectrum; (b) 13C NMR spectrum (solvent: CDCl3). Notable peaks in the 1H NMR spectrum include those from the iron-complexed diene (5.75, 5.48, 3.30, and 3.20 ppm) and the diastereotopic α-methylene protons (2.30 and 1.70 ppm). Please click here to view a larger version of this figure.
Figure 5. 1H NMR spectrum (solvent: CD3CN) of cationic iron complex 2. The most notable difference from the 1H NMR spectrum of 1 (the precursor to 2) is the signals arising from the diastereotopic α-methylene protons (2.85 and 2.23 ppm). Please click here to view a larger version of this figure.
Figure 6. Spectral data for iron complex 3. (a) 1H NMR spectrum; (b) 13C NMR spectrum (solvent: CDCl3). Similar to the 1H NMR spectrum of 4, the 1H NMR spectrum of 3 is characterized by signals arising from the iron-complexed diene (5.89, 5.51, 3.53, and 3.30 ppm) and the diastereotopic α-methylene protons (2.50 and 2.02 ppm). Please click here to view a larger version of this figure.
Figure 7. Spectral data for tert-butyl carbamate 5. (a) 1H NMR spectrum; (b) 13C NMR spectrum (solvent: CDCl3). The signal corresponding to the protons of the tert-butyl group of the carbamate appear at 1.52 ppm. Many signals also show characteristic broadening. Please click here to view a larger version of this figure.
Figure 8. Spectral data for demetallated diene 6. (a) 1H NMR spectrum; (b) 13C NMR spectrum (solvent: CDCl3). The most notable aspect of the 1H NMR spectrum compared to those of the iron complexes in Figure 4a, Figure 6a, and Figure 7a is that all of the signals corresponding to the diene protons now appear above 5.75 ppm (6.57, 6.34, 6.10, and 5.99 ppm). Please click here to view a larger version of this figure.
Whether the solvent-free protocol involving direct addition to tricarbonyl(tropone)iron (Figure 2) or the indirect method utilizing the corresponding cationic complex as the electrophile (Figure 1) is to be employed depends on the amine substrate used. In general, the direct addition method is preferable since it requires fewer steps to generate the aza-Michael adducts from tropone and the overall yields are generally higher. However, this more direct method is generally limited to reasonably unhindered primary aliphatic amines and cyclic secondary amines (e.g., piperidine). Less nucleophilic substrates such as arylamines or more sterically hindered amines such as acyclic secondary amines or tert-butylamine do not directly add to tricarbonyl(tropone)iron. On the other hand, these substrates efficiently add to the corresponding cationic complex (2, Figure 1). Thus, the two protocols complement one another in that the direct addition reaction is generally more efficient and higher yielding, while the addition to the cationic complex enjoys a broader substrate scope.
For the direct addition to tricarbonyl(tropone)iron, reaction times tend to be substrate-dependent. Some additions are complete within minutes as judged by 1H NMR analysis (e.g., unhindered primary amines) while some must be left overnight (e.g., morpholine). Upon completion, excess amine is removed via chromatography over Activity II/III alumina. However, for sufficiently volatile amine substrates, the excess amine may be removed via rotary evaporation and the crude material can then be subjected to protection as the corresponding carbamate (if applicable).
Adducts of primary aliphatic amines should be purified without delay and should be protected as carbamates as soon as is practicable, as we have generally experienced that such adducts will degrade over time. The degradation is generally accompanied by a color change from bright yellow to orange-brown. NMR analysis of such partially degraded samples showed the presence of tricarbonyl(tropone)iron, indicating that elimination of the amine had occurred.
We screened a variety of known protocols for removing the iron tricarbonyl group from the diene of the aza-Michael adducts22,23,24,25,26,27. The only successful protocol in our hands involved oxidative demetallation via treatment of the carbamate-protected adducts with cerium(IV) ammonium nitrate28. A representative result is described for demetallation of a tert-butyl carbamate-protected adduct. However, benzyl carbamates can also be demetallated using this protocol (no other carbamates were examined). Since tertiary amines cannot be protected as carbamates, we have thus far been unable to successfully demetallate those substrates despite extensive experimentation, including attempts to temporarily protect the nitrogen from oxidation by quantitatively protonating it with trifluoroacetic acid.
This protocol represents an extension of a method reported by Eisenstadt18 for addition of amines to cationic complex 2. However, addition of only two amines to the complex was reported, and demetallation of the complex was not described. The work described herein more fully explores the scope of addition to the cationic complex. Furthermore, the protocol for the direct addition of certain amines to tricarbonyl(tropone)iron constitutes a more efficient method for synthesizing such amine adducts. In addition, successful demetallation of the complexes opens the way for diverse subsequent reactions to access more complex molecular architectures containing a seven-membered carbocyclic ring. Notably, the addition of diverse amine nucleophiles with different functionalized side chains can potentially enable an even more diverse set of downstream reactions. Exploration of such newly-opened synthetic routes to complex alkaloid-like architectures is currently under investigation in our laboratory.
The authors have nothing to disclose.
Acknowledgement is made to the Donors of the American Chemical Society Petroleum Research Fund for support of this research. We acknowledge the Lafayette College Chemistry Department and the Lafayette College EXCEL Scholars program for financial support.
10 g SNAP Ultra silica gel columns | Biotage | for automated column chromatography | |
Acetic anhydride | Fisher Scientific | A10-500 | |
Acetone | Fisher Scientific | A-16S-20 | for cooling baths |
Acetonitrile-D3 | Sigma Aldrich | 366544 | |
Benzene, anhydrous, 99.8% | Sigma Aldrich | 401765 | |
Biotage Isolera Prime | Biotage | ISO-PSF | for automated chromatography |
Celite; 545 Filter Aid | Fisher Scientific | C212-500 | diatomaceous earth |
Cerium(IV) ammonium nitrate, ACS, 99+% | Alfa Aesar | 33254 | |
Chloroform-D | Acros | 209561000 | |
Di-tert-butyl dicarbonate, 99% | Acros | 194670250 | |
Ethyl acetate | Fisher Scientific | E145-4 | |
Ethyl alcohol, absolute – 200 proof | Greenfield Global | 111000200PL05 | |
Ethyl ether anhydrous | Fisher Scientific | E138-1 | |
Hexanes | Fisher Scientific | H302-4 | |
iron nonacarbonyl 99% | Strem | 26-2640 | air sensitive, synonymous with diiron nonacarbonyl |
Magnesium sulfate | Fisher Scientific | M65-500 | |
Methanol | EMD Millipore | MX0475-1 | |
Methylene chloride | Fisher Scientific | D37-4 | |
MP alumina, Act. II-III acc. To Brockmann | MP Biomedicals | 4691 | for column chromatography |
o-toluidine 98% | Sigma Aldrich | 466190 | |
Phenethylamine 99% | Sigma Aldrich | 128945 | distill prior to use if not colorless |
Sodium bicarbonate | Fisher Scientific | S233-500 | |
Sodium carbonate anhydrous | Fisher Scientific | S263-500 | |
Sodium chloride | Fisher Scientific | S271-500 | dissolved in deionized water to perpare a saturated aqueous solution |
Sodium sulfate anhydrous | Fisher Scientific | S415-500 | |
Sonicator | Branson | model 2510 | |
Sulfuric acid | Fisher Scientific | A300C-212 | |
Tetrafluoroboric acid solution, 48 wt.% | Sigma Aldrich | 207934 | aqueous solution |
TLC Aluminium oxide 60 F254, neutral | EMD Millipore | 1.05581.0001 | for thin layer chromatography |
Tropone 97% | Alfa Aesar | L004730-06 | Light sensitive |