Synlett 2023; 34(05): 414-422
DOI: 10.1055/s-0042-1753200
account

Adventures in 1,1-Diphenylacetone Dianion Chemistry

Peter Langer
a   Institut für Chemie, Universität Rostock, Albert-Einstein-Str. 3a, 18059 Rostock, Germany
b   Leibniz-Institut für Katalyse an der Universität Rostock e.V., Albert-Einstein-Str. 29a, 18059 Rostock, Germany
› Author Affiliations
 


Dedicated to Professor Dietmar Seyferth, in memoriam.

Abstract

The present personalized account highlights reactions of the dianion of 1,1-diphenylacetone and its applications. From a study directed towards the regioselectivity of such reactions a new allene synthesis via allene dianions was found by serendipity. In addition, reactions of allene dianions with various electrophiles have been studied.

1 Introduction

2 Cyclizations of the Dianion of 1,1-Diphenylacetone

2.1 Reaction with Bis(cyclopentadienyl)zirconium Dichloride

2.1 Reactions with Dichlorosilanes

2.1 Reactions with Difluorosilanes

2.1 Reactions with Ketones and Dielectrophiles

3 Regioselectivity of Dianion Reactions

4 Allene Dianions from Silyl Enol Ethers

5 Scope of the Reactions of Allene Dianions

6 Reactions of Allene Dianions with Ketones

7 Reactions of Allene Dianions with Nitriles

8 Conclusions


#

Biographical Sketch

Zoom Image

Peter Langer was born in Hannover (Germany) in 1969. He studied chemistry at the University of Hannover and at the Massachusetts Institute of Technology (MIT), and received his diploma under the guidance of Prof. Dietmar Seyferth in March 1994. In February 1997, he obtained his Ph.D. (Dr. rer. nat.) under the supervision of Prof. H. Martin R. Hoffmann at the University of Hannover. After postdoctoral studies with Prof. Steven V. Ley, FRS (1997–1998, Cambridge, UK), Peter moved to the University of Göttingen where he started his independent research associated to Prof. Armin de Meijere. He completed his habilitation in July 2001 and became private docent. In April 2002, he took a permanent position as a full professor (C4) at the University of Greifswald. In December 2004, Peter moved to a new position as a full professor (C4) at the University of Rostock. Prof. Langer is the co-author of about 780 research papers and reviews, and is the recipient of the following awards and scholarships: Studienstiftung des deutschen Volkes, Promotionsstipendium des Fonds der Chemischen Industrie, Feodor-Lynen-Stipendium, Liebig-Stipendium, and Heisenberg-Stipendium. He has received honorary doctorates, honorary professorships and medals from various universities. In addition, he is an elected member of the Academy of Sciences of the Republic of Armenia and of the Academy of Sciences of Pakistan. In 2015, he received the civil award ‘Sitara-i-Quaid-i-Azam’ from the President of the Islamic Republic of Pakistan.

1

Introduction

At a very early stage of my career, as an undergraduate student at the University of Hannover (Germany), I decided to carry out my specialization and diploma thesis (nowadays equivalent to a master’s thesis) in the field of organic chemistry. My mentor, Prof. H. M. R. Hoffmann, recommended that I carry out a research stay abroad, ideally in the USA.[1] [2] He was one of the early protagonists of international student exchange and had already established exchange programs for students in Hannover in the 1980s and 1990s. I remember that he once said, “We should not be egoistic and keep our best students at the place. If we send our best students abroad, we also get excellent students back”. Such an international student exchange was, in the first half of the 1990s when I was studying chemistry, not as common as it is today and made a deep impression on me. Prof. Hoffmann himself had worked many years abroad, his wife was from the United Kingdom, and he had a network of colleagues and friends working at the best universities all over the world. A number of students of the ­Hoffmann group had carried out research stays of one semester at Stanford University. However, I planned to undertake my diploma thesis at the Massachusetts Institute of Technology (MIT, Cambridge, USA). I was lucky to be accepted as an unsalaried research associate by Prof. Dietmar ­Seyferth, who worked in the field of metalorganic chemistry. Prof. Hoffmann agreed to act as the German supervisor of my diploma thesis to be submitted at the University of Hannover. Therefore, I did not have to enroll as a student at MIT and did not have to pay tuition fees. As a fellow of the German Academic Scholarship Foundation (Studienstiftung des deutschen Volkes), I was very lucky to receive a scholarship for my stay at MIT.

Like Prof. Hoffmann, Prof. Seyferth was not only an outstanding scientist, but also a generous man with an excellent personality. He was born in Chemnitz in Germany and left as a child in the 1930s to go to the USA, where his father found a job as a chemist. He studied at Harvard University, worked as a postdoc at the Technical University of Munich in the field of metalorganic chemistry (with Prof. E. O. Fischer) and later became a professor at MIT. Dietmar ­Seyferth visited Munich nearly every year and before the start of my stay at MIT, in the summer of 1993, he invited me to meet with him personally and to acquaint myself with Munich. I travelled by train from Hannover to Munich, and following my arrival, Prof. Seyferth took me to an excellent restaurant (Hofbräuhaus) and provided me with some details of my future project. This meeting, thirty years ago, was very important to me as it made me more relaxed regarding my stay at MIT, which was planned to start a few weeks later. By the way, at that time, email and the internet did not exist, and urgent information had to be exchanged by fax, phone or in person (wherever possible).

My project was in the field of organosilicon chemistry. A Ph.D. student in the group, Tao Wang, used dianions[1] of 1,1-diphenylacetone and 1,3-diphenylacetone to react with dihalides of various elements to prepare a range of cyclic compounds. When I joined the group of Dietmar (I was soon allowed to call him by his first name, despite the age difference of exactly 40 years) I was only 24 years old and did not have much experience, neither in the lab, nor in the English language. Although the beginning was not easy, my stay was very fruitful and had a great impact on my future career. Based on serendipity, I was able to develop an unprecedented transformation of silyl enol ethers into allenes via allenyl dianions.

After completion of my diploma thesis, I returned to Hannover and undertook my Ph.D. (1994–1997) in the group of Prof. Hoffmann in a completely different area, namely, in the field of Cinchona alkaloids. I did my postdoc (1997–1998) with Steven V. Ley (Cambridge, UK), again in a different field: carbohydrate chemistry. Later, during my habilitation at the University of Göttingen (1998–2001), I developed several new applications of dianions and silyl enol ethers in organic chemistry, and also continued to study the allene synthesis already developed during my diploma thesis. Dietmar Seyferth and I were in contact until he died at the age of 90 in 2020. This present account aims to highlight my first experiences in the field of chemistry and what developed from these later.


# 2

Cyclizations of the Dianion of 1,1-Diphenylacetone

2.1

Reaction with Bis(cyclopentadienyl)zirconium Dichloride

My colleague Tao Wang studied cyclizations of the dianion of 1,1-diphenylacetone (1) with various dihalides. Treatment of 1 with potassium hydride and subsequent addition of n-butyllithium gave rise to the formation of ambident dianion A (Scheme [1]).[2] Dianion A is formed completely. Therefore, no reaction with a second molecule of 1 is possible. The reaction of dianion A with bis(cyclopentadienyl)zirconium dichloride afforded product 2, which was presumably formed by initial formation of zirconaoxetane B, a 1:1 cyclization product, and subsequent dimerization to give the isolated 2:2 product. In fact, in the solid state, this compound exists in the form of dimer 2, according to MS and X-ray structure analysis. In contrast, in solution, the compound resides, according to VPO (vapor pressure osmometry), in the form of monomer B.

Zoom Image
Scheme 1 Synthesis of 2. Conditions: (i) (1) KH, THF, 20 °C, 20 min; (2) nBuLi, 0 °C, 5 min; (3) Cp2ZrCl2, 0 to 20 °C, 5 h.

# 2.2

Reactions with Dichlorosilanes

The cyclization of the dianion of 1 with dichlorosilanes afforded eight-membered cyclic products 3ac in moderate to good yields (Scheme [2]).[3] The regioselectivity was different as compared to the formation of zirconaoxetane 2. While in the case of 2 both Zr atoms are bound to a carbon and an oxygen atom, the two silicon atoms of 3 occupy different chemical environments. One silicon is neighbored by two carbon atoms, while the other silicon is neighbored by two oxygen atoms. The formation of products 3ac can be explained by the SN2 reaction of dianion A with the dichlorosilane to give intermediate B. The latter undergoes cyclization via the oxygen atoms with a second molecule of the dichlorosilane.

Zoom Image
Scheme 2 Synthesis of 3ac. Conditions: (i) (1) KH, THF, 20 °C, 20 min; (2) nBuLi, 0 °C, 5 min; (3) R2SiCl2, 0 to 20 °C, 12 h.

# 2.3

Reactions with Difluorosilanes

The reactions of 1 with diethyldifluorosilane and diphenyldifluorosilane afforded eight-membered rings 4a and 4b, respectively (Scheme [3]).[3] In these products both silicon atoms are neighbored by a carbon and an oxygen atom, and thus have a similar structure as present in 2, but a different structure as compared to 3ac. Thus, the cyclization of the dianion of 1 with dichloro and difluorosilanes proceeds with different regioselectivities. The formation of 4a,b might be explained by attack of the dianion on the difluorosilane to give pentacovalent intermediate A, cyclization to give silaoxetane B and subsequent dimerization. This mechanism is related to that suggested above for the formation of zirconaoxetane 2.

Zoom Image
Scheme 3 Synthesis of 4a,b. Conditions: (i) (1) KH, THF, 20 °C, 20 min; (2) nBuLi, 0 °C, 5 min; (3) R2SiF2, 0 to 20 °C, 12 h.

In the case of the reaction of the dianion of 1 with diphenylgermanium dichloride and diphenylgermanium difluoride the same regioselectivity was observed and the product possesses a structure related to that of silane 4b. This result might be explained by the fact that germanium, due to its size, more readily than silicon undergoes formation of a hypervalent species. Therefore, for both the difluoride and the dichloride, the product is formed via the mechanism depicted in Scheme [3] for the difluorosilanes.


# 2.4

Reactions with Ketones and Dielectrophiles

One-pot reactions of the dianion of 1 with ketones and subsequent cyclizations with various dielectrophiles have also been studied. The reaction of the dianion of 1 with benzophenone and subsequent addition of diphenyldichlorosilane afforded the six-membered ring product 5 in good yield (Scheme [4]).[4] The reaction proceeds by regioselective attack of the carbon atom of dianion A on the carbonyl group of the ketone to give dianionic intermediate B, and subsequent cyclization via the two oxygen atoms results in formation of the product.

Zoom Image
Scheme 4 Synthesis of 5. Conditions: (i) (1) KH, THF, 20 °C, 20 min; (2) nBuLi, 0 °C, 5 min; (3) Ph2CO, 2 h, 0 to 20 °C; (4) Ph2SiCl2, 0 to 20 °C, 12 h.

The reaction of 1 with benzophenone and bis(4-tolyl)ketone and subsequent cyclization with diethylmalonyl dichloride afforded the eight-membered rings 6a,b in moderate yields (Scheme [5]).[4] This result shows that the formation of medium-sized rings from 1 is not restricted to zirconium- and silicon-containing ring systems.

Zoom Image
Scheme 5 Synthesis of 6a,b. Conditions: (i) (1) KH, THF, 20 °C, 20 min; (2) nBuLi, 0 °C, 5 min; (3) R2CO, 2 h, 0 to 20 °C; (4) Et2C(COCl)2, 0 to 20 °C, 12 h.

The reaction of 1 with benzophenone and subsequent cyclization with phthaloyl dichloride afforded the spirocyclic compound 7 (89%) rather than the expected nine-membered ring product (Scheme [6]).[4] The result might be explained by the transformation of phthaloyl dichloride A into isophthaloyl dichloride B, which then undergoes the cyclization.

Zoom Image
Scheme 6 Synthesis of 7. Conditions: (i) (1) KH, THF, 20 °C, 20 min; (2) nBuLi, 0 °C, 5 min; (3) Ph2CO, 2 h, 0 to 20 °C; (4) phthaloyl dichloride, 0 to 20 °C, 12 h.

#
# 3

Regioselectivity of Dianion Reactions

To study the regioselectivity of the reaction of the dianion of 1 with dichlorosilanes, reactions with two equivalents of monochlorosilanes were studied. The reaction of 1 with three different chlorosilanes afforded products 8ac in good yields (Scheme [7]).[3] The products contain a silyl enol ether and an allyl silane moiety.

Zoom Image
Scheme 7 Synthesis of 8ac. Conditions: (i) (1) KH, THF, 20 °C, 20 min; (2) nBuLi, 0 °C, 5 min; (3) chlorosilane (2 equiv.), 0 to 20 °C, 12 h.
Zoom Image
Scheme 8 Synthesis of 9ae. Conditions: (i) (1) KH, THF, 20 °C, 20 min; (2) nBuLi, 0 °C, 5 min; (3) 1st chlorosilane (1.0 equiv.), 0 to 20 °C, 30 min; (4) 2nd chlorosilane (1.0 equiv.), 0 to 20 °C, 12 h.

The reaction of the dianion of 1 with one equivalent of a monochlorosilane and subsequent addition of another one equivalent afforded products 9ae (Scheme [8]). The yields were good, except for product 9e. The results can be explained by initial attack of the carbon atom of dianion A on the chlorosilane to give intermediate B. The latter then reacts with the second chlorosilane to give the product. In dianion A, the carbon atom, which was deprotonated last, is more reactive than the oxygen atom. This is common for the chemistry of dianions.[1] The low yield of 9e can be explained by the fact that tert-butyldimethylchlorosilane, which was added in the first step, represents a sterically hindered reagent.

As mentioned above, dianions usually react with electrophiles at the center that was deprotonated last.[1] However, we considered that the situation might be different in the case of silylation reactions due to the oxophilicity of the chlorosilane and its hard character according to the HSAB principle. Although the formation of products 9 suggested that the dianion was first silylated at the carbon atom, attack at the oxygen and subsequent O → C rearrangement of the silyl group could not be excluded. In fact, such a rearrangement had been reported by Corey and Rücker who obtained α-triisopropylsilyl ketones from carbanions of triisopropylsilyl enol ethers of 4-tert-butylcyclohexanone derivatives.[5a] The driving force of the silyl migration was suggested to be the formation of a stable enolate from a less stable allylic anion. This transformation takes the opposite direction to that reported for electroneutral allylsilanes, which form the thermodynamically more stable silyl enol ethers (Brooke rearrangement).[5b] [c] This reaction usually takes place upon heating of the allylsilane or by the action of Lewis acids.


# 4

Allene Dianions from Silyl Enol Ethers

To study the possibility of a silyl migration from the oxygen to the carbon during the formation of products 9, silyl enol ether 10a was prepared by silylation of the monoanion of 1 (Scheme [9]). Deprotonation of the latter and subsequent addition of dimethylchlorosilane (Me2HSiCl) afforded a 1:1 mixture of 9a and 10b. The formation of 9a can be explained by deprotonation of 10a to give intermediate B, O → C rearrangement of the silyl group to give intermediate C and subsequent silylation of the oxygen atom upon addition of Me2HSiCl. Indeed, this result suggested that during the formation of 9a a rearrangement was involved, as earlier observed by Corey and Rücker. The formation of significant amounts of 10b was assumed to be a result of nucleophilic attack of LDA at the silicon atom of 10a, cleavage of the O–Si bond and extrusion of Me3SiNiPr2 to give enolate A, which was then silylated at the oxygen atom upon addition of Me2HSiCl. In fact, it is known that silyl enol ethers react with strong nucleophiles, such as methyllithium, to give lithium enolates.[6] In the present case, the action of LDA was surprising as it represents a sterically hindered reagent of rather low nucleophilicity.

Zoom Image
Scheme 9 Synthesis and deprotonation–silylation of 10a. Conditions: (i) (1) KH, THF, 20 °C, 20 min; (2) Me2HSiCl (1.1 equiv.), 0 to 20 °C, 12 h; (ii) (1) LDA (1.1 equiv.), 4 h, 0 °C; (2) Me2HSiCl (1.1 equiv.), 0 to 20 °C, 12 h.
Zoom Image
Scheme 10 Synthesis and deprotonation–silylation of 10c. Conditions: (i) (1) KH, THF, 20 °C, 20 min; (2) tBuMe2SiCl (1.1 equiv.), 0 to 20 °C, 12 h; (ii) (1) LDA (1.1 equiv.), 6 h, 0 °C; (2) Me3SiCl (1.1 equiv.), 0 to 20 °C, 12 h; (iii) (1) LDA (3.3 equiv.), 6 h, 0 °C; (2) Me3SiCl (3.5 equiv.), 0 to 20 °C, 12 h.

To address the problem of the formation of 10b as a side product, we prepared tert-butyldimethylsilyl enol ether 10c, as silyl ethers containing bulky silyl groups, such as the tert-butyldimethylsilyl (TBDS), are known to be more stable against nucleophiles and electrophiles.[7] We assumed that this would be the case also for silyl enol ethers, that 10c would be more stable than 10b and that migration of the silyl group might not occur. Silylation of 1 with tert-butyldimethylsilyl chloride afforded silyl enol ether 10c (Scheme [10]). Treatment of 10c with LDA, stirring for 6 hours, and addition of Me3SiCl afforded a 1:2 mixture of allene 11a and substrate 10c. The surprising formation of an allene along with the recovery of a considerable amount of starting material suggested increasing the amount of LDA and Me3SiCl. Indeed, treatment of 10c with 3.3 equivalents of LDA, stirring for 6 hours and subsequent addition of 3.5 equivalents of Me3SiCl resulted in formation of allene 11a in an excellent 84% yield.[8]

The formation of allene 11a can be explained by two pathways. Treatment of 10c with LDA results in formation of intermediate A (Scheme [11]). Subsequent O → C rearrangement of the silyl group afforded intermediate B which then undergoes a Peterson elimination to give allene C (path A). The latter could have also been formed directly from A (path B). Given the stability of sterically hindered silyl enol ethers towards nucleophiles, we assume that the direct elimination (path B) takes place. It is well-known that allenes can be more readily deprotonated than alkenes. Thus, twofold deprotonation of C with LDA results in formation of dilithiated allene D, which is subsequently trapped by addition of Me3SiCl. Instead of LDA, nBuLi could also be used, but the yield of 11a dropped slightly to 70%. In fact, a small amount of starting material 10c was present in the reaction mixture. This can be explained by decomposition of nBuLi by reaction with the solvent THF. Despite the nucleophilicity of nBuLi, it obviously does not attack the silicon atom of 10c, because this would have resulted in the formation of the enolate of 1 and then 10a upon addition of Me3SiCl. However, 10a could not be detected in the reaction mixture.[9] In conclusion, we were convinced that the formation of allene C is a process that requires some time and is not a very rapid process. 1,1-Dilithiated allene D was previously obtained by double lithiation of 1,1-diphenylcyclopropene.[10]

Zoom Image
Scheme 11 A possible mechanism for the formation of 11a

The formation of allene 11a from silyl enol ether 10c constitutes an unprecedented allene synthesis and a new reaction mode of silyl enol ethers. In typical reactions of silyl enol ethers the oxygen–silicon bond is cleaved. In contrast, in the present case, the carbon–oxygen bond is cleaved. This reaction is related to the known base-mediated transformation of enol phosphates into alkynes[11] and allenes,[12] and to the conversion of alkenyl halides and triflates into alkynes.[13]

As mentioned above, the formation of allene 11a was observed only when the sterically hindered tert-butyldimethylsilyl group was employed. To further study this aspect, we also studied the sterically hindered triisopropylsilyl group. The reaction of the monoanion of 1 with triisopropylsilyl chloride afforded silyl enol ether 10d (Scheme [12]). Treatment of the latter with an excess of LDA and subsequent addition of Me3SiCl gave allene 11a in 82% yield. This result shows that employment of sterically hindered silyl enol ethers is indeed important to induce an elimination reaction.

Zoom Image
Scheme 12 Synthesis of 11a via 10d. Conditions: (i) (1) KH, THF, 20 °C, 20 min; (2) (iPr)3SiCl (1.1 equiv.), 0 to 20 °C, 12 h; (ii) (1) LDA (3.3 equiv.), 6 h, 0 °C; (2) Me3SiCl (3.5 equiv.), 0 to 20 °C, 12 h.

# 5

Scope of the Reactions of Allene Dianions

To study the scope of the allene synthesis the electrophiles were varied.[14] The reaction of 10c with dimethyl- and methylphenyl-chlorosilane afforded silylated allenes 11b and 11c (Figure [1]). The yield of 11c was only moderate because of the steric hindrance of the silane. The reaction with trimethyltin chloride gave 11d. The yield was moderate because of competing reduction. Trapping of the dianion with ethanol gave allene 11e, albeit in low yield, while addition of dimethyl sulfate gave alkyne 11f. The regioselectivity can be explained based on the HSAB principle.[15]

Zoom Image
Figure 1 Synthesis of compounds 11bf

The reaction of 1,3-diphenylacetone (12) with tert-butyldimethylsilyl chloride afforded silyl enol ether 13 in 82% yield (Scheme [13]).[14] Treatment of the latter with 3.3 equivalents of LDA and subsequent addition of 3.5 equivalents of Me3SiCl afforded bis-silylated allene 14a in 82% yield. The formation of the product can be explained by a mechanism related to that discussed above for silyl enol ether 10c and involves the formation of dianion A. The reaction of the corresponding silyl enol ether, carrying a triisopropylsilyl instead of a tert-butyldimethylsilyl group, proceeded equally well. The reaction of 13 with trimethyltin chloride gave allene 14b, while addition of ethanol gave cyclobutane 14c, which was formed by dimerization of the corresponding allene. The reaction with dimethyl sulfate gave a 1:2 mixture of allene 14d and alkyne 14e.

Zoom Image
Scheme 13 Synthesis of compounds 14ae

The reaction of 2-quinolylacetone (15) with tert-butyldimethylsilyl chloride afforded silyl enol ether 16 in 84% yield (Figure [2]).[14] Treatment of the latter with 4.4 equivalents of LDA and subsequent addition of 4.5 equivalents of Me3SiCl afforded tris-silylated allene 17 in 85% yield. The formation of the product can be explained by the generation of an intermediate allenyl trianion. The reactions of substrates containing a 2-pyridyl or a phenyl instead of the 2-quinolyl group proceeded sluggishly. The corresponding allenes were formed, but could not be isolated. The corresponding reactions starting with silyl enol ethers derived from acetone or pentan-3-one failed and did not result in formation of the desired allenes. These results show that the presence of an aryl group, preferably two, is important for formation of the allene.

Zoom Image
Figure 2 Synthesis of compound 17

# 6

Reactions of Allene Dianions with Ketones

The reaction of 10c with various benzophenones 18ae afforded sterically encumbered bis(hydroxymethyl)allenes 19ae in 62–80% yields (Scheme [14]).[14] [16] Heating of a toluene solution of the latter in the presence of p-toluenesulfonic acid (PTSA) gave rise to the formation of arenes 20ae. The formation of these products can be explained by extrusion of water to give cation A, cyclization via the central carbon atom of the allene, extrusion of a second molecule of water to give cation B and cyclization via the phenyl group derived from the silyl enol ether. In most cases, symmetrical benzophenones were employed. However, in one example, an unsymmetrical benzophenone was used. The first cyclization proceeded with good regioselectivity via the more electron rich p-methoxyphenyl group to give product 20b.

Zoom Image
Scheme 14 Synthesis of 19ae and 20ae. Conditions: (i) (1) LDA (3.3 equiv.), 6 h, 0 °C; (2) 18ae (2.5 equiv.), –78 to 20 °C, 12 h; (3) water; (ii) PTSA, toluene, 80 °C, 2 h.

The reaction of 10c with fluorenone and xanthone afforded bis(hydroxymethyl)allenes 21 and 22 in good yields, respectively (Figure [3]).[14] [16] Heating of a toluene solution of 21 and 22 in the presence of p-toluenesulfonic acid (PTSA) afforded cumulenes 23 and 24. The formation of the products can be explained by acid-mediated extrusion of water and of one molecule of the ketone. The change of the mode of reaction in comparison to the corresponding reactions of bis(hydroxymethyl)allenes 19 is presumably a result of the rigid character of the ketones and the anti-aromatic character of the 9-fluorenyl cation.

Zoom Image
Figure 3 Synthesis of compounds 2124

# 7

Reactions of Allene Dianions with Nitriles

The reaction of 10c with 4.5 equivalents of benzonitrile afforded the unexpected product 25 in 51% yield (Scheme [15]).[17] [18] In the course of this reaction, four molecules of benzonitrile were consumed and incorporated into the final product. Therefore, the reaction represents a domino process and it is termed a ‘multiple anion capture reaction’. The formation of the product can be explained by attack of allenyl dianion A on two molecules of the nitrile to give intermediate B, addition of a third molecule of the nitrile to give C, cyclization to give intermediate D, 1,2-rearrangement of a benzonitrile moiety to give E, addition of a fourth molecule of the nitrile to give intermediate F and a second cyclization to give intermediate G. The product 25 is formed upon addition of water during the aqueous work-up. An imidazole was obtained as a side product in 12% yield, which was formed by hydrolysis of intermediate E. In the case of the employment of pivalonitrile, a product derived from E was exclusively formed.

Zoom Image
Scheme 15 Synthesis of compound 25. Conditions: (i) (1) LDA (3.3 equiv.), 6 h, 0 °C; (2) benzonitrile (4.5 equiv.), –78 to 20 °C, 12 h; (3) water.

# 8

Conclusions

The present personalized account highlights reactions of the dianion of 1,1-diphenylacetone and its applications. From a study directed towards the regioselectivity of such reactions a new allene synthesis was found by serendipity. Applications of this reaction were studied. The chemistry highlighted herein, carried out during the early stage of my career, was the starting point for the development of a great variety of cyclization reactions of free and masked dianions with electrophiles in the following years. This chemistry has been reviewed and includes cyclization reactions of dianions in general,[19] cyclizations of dianions and dinucleophiles with oxalic acid–bis(imidoyl)dichlorides,[20] cyclizations of oxime and hydrazine dianions[21] and the synthesis of 2-(alkylidene)tetrahydrofurans by cyclizations of dianions.[22] In addition, cyclizations of (2,4-dioxobutylidene)phosphoranes, as equivalents of 1,3-dicarbonyl dianions, have been reviewed.[23] 1,3-Bis(silyloxy)-1,3-butadienes and 1,3-bis(silyl enol ethers) also represent electroneutral equivalents of 1,3-dicarbonyl dianions (masked dianions). They can be reacted with electrophiles in the presence or absence of Lewis acids.[24] Applications of 1,3-bis(trimethylsilyloxy)-1,3-butadienes in cyclizations with oxalyl chloride,[25] 3-silyloxy- and 3-alkoxy-2-en-1-ones,[26] iminium salts,[27] benzopyrylium triflates,[28] and acid chlorides[29] have been reviewed. Reactions of sulfur-,[30] fluorine-[31] and chlorine-containing[32] (free and masked) dianions and 1,1-bis(trimethylsilyloxy)ketene acetals[33] have also been studied. All these reactions provide a convenient access to hetero- and carbocyclic compounds, which are of relevance in medicine and materials science.


#
#

Conflict of Interest

The author declares no conflict of interest.

Acknowledgment

P.L. is grateful to the late Professor Dietmar Seyferth for his generous support.


Corresponding Author

Peter Langer
Institut für Chemie, Universität Rostock
Albert-Einstein-Str. 3a, 18059 Rostock
Germany   

Publication History

Received: 24 November 2022

Accepted after revision: 04 December 2022

Article published online:
05 January 2023

© 2022. Thieme. All rights reserved

Georg Thieme Verlag KG
Rüdigerstraße 14, 70469 Stuttgart, Germany


Zoom Image
Zoom Image
Scheme 1 Synthesis of 2. Conditions: (i) (1) KH, THF, 20 °C, 20 min; (2) nBuLi, 0 °C, 5 min; (3) Cp2ZrCl2, 0 to 20 °C, 5 h.
Zoom Image
Scheme 2 Synthesis of 3ac. Conditions: (i) (1) KH, THF, 20 °C, 20 min; (2) nBuLi, 0 °C, 5 min; (3) R2SiCl2, 0 to 20 °C, 12 h.
Zoom Image
Scheme 3 Synthesis of 4a,b. Conditions: (i) (1) KH, THF, 20 °C, 20 min; (2) nBuLi, 0 °C, 5 min; (3) R2SiF2, 0 to 20 °C, 12 h.
Zoom Image
Scheme 4 Synthesis of 5. Conditions: (i) (1) KH, THF, 20 °C, 20 min; (2) nBuLi, 0 °C, 5 min; (3) Ph2CO, 2 h, 0 to 20 °C; (4) Ph2SiCl2, 0 to 20 °C, 12 h.
Zoom Image
Scheme 5 Synthesis of 6a,b. Conditions: (i) (1) KH, THF, 20 °C, 20 min; (2) nBuLi, 0 °C, 5 min; (3) R2CO, 2 h, 0 to 20 °C; (4) Et2C(COCl)2, 0 to 20 °C, 12 h.
Zoom Image
Scheme 6 Synthesis of 7. Conditions: (i) (1) KH, THF, 20 °C, 20 min; (2) nBuLi, 0 °C, 5 min; (3) Ph2CO, 2 h, 0 to 20 °C; (4) phthaloyl dichloride, 0 to 20 °C, 12 h.
Zoom Image
Scheme 7 Synthesis of 8ac. Conditions: (i) (1) KH, THF, 20 °C, 20 min; (2) nBuLi, 0 °C, 5 min; (3) chlorosilane (2 equiv.), 0 to 20 °C, 12 h.
Zoom Image
Scheme 8 Synthesis of 9ae. Conditions: (i) (1) KH, THF, 20 °C, 20 min; (2) nBuLi, 0 °C, 5 min; (3) 1st chlorosilane (1.0 equiv.), 0 to 20 °C, 30 min; (4) 2nd chlorosilane (1.0 equiv.), 0 to 20 °C, 12 h.
Zoom Image
Scheme 9 Synthesis and deprotonation–silylation of 10a. Conditions: (i) (1) KH, THF, 20 °C, 20 min; (2) Me2HSiCl (1.1 equiv.), 0 to 20 °C, 12 h; (ii) (1) LDA (1.1 equiv.), 4 h, 0 °C; (2) Me2HSiCl (1.1 equiv.), 0 to 20 °C, 12 h.
Zoom Image
Scheme 10 Synthesis and deprotonation–silylation of 10c. Conditions: (i) (1) KH, THF, 20 °C, 20 min; (2) tBuMe2SiCl (1.1 equiv.), 0 to 20 °C, 12 h; (ii) (1) LDA (1.1 equiv.), 6 h, 0 °C; (2) Me3SiCl (1.1 equiv.), 0 to 20 °C, 12 h; (iii) (1) LDA (3.3 equiv.), 6 h, 0 °C; (2) Me3SiCl (3.5 equiv.), 0 to 20 °C, 12 h.
Zoom Image
Scheme 11 A possible mechanism for the formation of 11a
Zoom Image
Scheme 12 Synthesis of 11a via 10d. Conditions: (i) (1) KH, THF, 20 °C, 20 min; (2) (iPr)3SiCl (1.1 equiv.), 0 to 20 °C, 12 h; (ii) (1) LDA (3.3 equiv.), 6 h, 0 °C; (2) Me3SiCl (3.5 equiv.), 0 to 20 °C, 12 h.
Zoom Image
Figure 1 Synthesis of compounds 11bf
Zoom Image
Scheme 13 Synthesis of compounds 14ae
Zoom Image
Figure 2 Synthesis of compound 17
Zoom Image
Scheme 14 Synthesis of 19ae and 20ae. Conditions: (i) (1) LDA (3.3 equiv.), 6 h, 0 °C; (2) 18ae (2.5 equiv.), –78 to 20 °C, 12 h; (3) water; (ii) PTSA, toluene, 80 °C, 2 h.
Zoom Image
Figure 3 Synthesis of compounds 2124
Zoom Image
Scheme 15 Synthesis of compound 25. Conditions: (i) (1) LDA (3.3 equiv.), 6 h, 0 °C; (2) benzonitrile (4.5 equiv.), –78 to 20 °C, 12 h; (3) water.