The Synthesis of Hypodiphosphoric Acid and Derivatives with P-P Bond, including Esters and Diphosphine Dioxides: A Review

The synthesis of hypodiphosphoric acid and its related compounds began in 1877, but no summary of the synthetic efforts has been reported. This review includes published papers related to the molecules containing the >P(=O)-P(=O)< fragment, which notably resembles the structure of the >P(=O)-O-P(=O)< moiety, the essential building block of many important molecules found in nature and in the field of medicinal chemistry. This review covers the strategies related to the synthesis of hypodiphosphoric acid (former name hypophosphoric acid), its ester form, and diphosphine dioxides. Finally, some properties and applications of these structures studied during this period are presented.


Introduction
Many organophosphorus molecules are biologically active and have been widely used as pesticides in agriculture, anticancer drugs and antiviral agents in medicines, and anthelmintics in the veterinary field. Understanding the mechanism of their reaction or inhibition process is critical to shedding light on the design of more active and efficient compounds in terms of medicinal treatment and agricultural production. All the mentioned molecules were insufficiently characterized in terms of spectroscopic characteristics and the procedure of synthesis, probably due to missing general synthetic methodologies, accompanying mechanisms, and/or some limitations in spectroscopic characterization. Other difficulties could be lacking proper structure elucidation for symmetric >P(O)- (O)P< type molecules, e.g., the ambiguity on the molecular formula of hypodiphosphoric acid (HO) 2 P(O)-(O)P(OH) 2 . In contrast, the unsymmetrical >P A (O)-(O)P B < (P A = P B ) type compounds are much easier to distinguish due to two inequivalent phosphorus atoms, thanks
Molecules 2021, 26,7286 2 of 30 unsymmetrical >PA(O)- (O)PB< (PA ≠ PB) type compounds are much easier to distinguish due to two inequivalent phosphorus atoms, thanks to the characteristic large 1 JP,P coupling visible in 31 P{ 1 H} NMR spectra and n JP,C couplings from both phosphorus atoms visible in 13 C{ 1 H} NMR spectra. Very interesting n JP,H couplings from both phosphorus atoms cannot be omitted. The NMR spectroscopy can also be used to differentiate diastereoisomers from each other if they are present. The present review is intended to comprehensively summarize recent advances in the process of P-P bond formation in the >P(O)- (O)P< type compounds (Figure 1), highlighting plausible mechanistic considerations, defining the scope and limitations, and raising interest in these exciting molecules.

Hypodiphosphoric Acid
Hypodiphosphoric acid (1) has been known about since the end of the 19th century. In 1877, Salzer published a synthetic procedure for the production of hypodiphosphoric acid salts (1′) (Scheme 1) [3][4][5][6][7]. Until now, no mechanism has been established for the formation of hypodiphosphoric acid (1). Acid 1 is a white crystalline molecule, stable as a hydrated form at room temperature. In aqueous solutions molecule 1 behaves as a weak tetrabasic acid at 25 °C (pKa1 = 2.2, pKa2 = 2.8, pKa3 = 7.3, pKa4 = 10.0) [8,9]. The hypodiphosphates (1′) undergo Until now, no mechanism has been established for the formation of hypodiphosphoric acid (1). Acid 1 is a white crystalline molecule, stable as a hydrated form at room temperature. In aqueous solutions molecule 1 behaves as a weak tetrabasic acid at 25 • C (pKa 1 = 2.2, pKa 2 = 2.8, pKa 3 = 7.3, pKa 4 = 10.0) [8,9]. The hypodiphosphates (1 ) undergo protonation to form compound 1 in strongly acidic aqueous solutions. However, acid 1 is unstable under acidic conditions and disproportionated to H 3 PO 4 and H 3 PO 3 [8,10]. One explanation of the instability of molecule 1 is due to a slow rearrangement to unsymmetrical The acid 2 is not stable and disproportionates to H 4 P 2 O 7 and H 4 P 2 O 5 . An alternative explanation of its instability could be the tautomerism of isohypodiphosphoric (2) with mixed phosphorus anhydride tautomer (2 ) (Scheme 2). The tautomeric equilibrium between the pentavalent central phosphorus atom in >P=(O)-H type molecules and the trivalent phosphorus in phosphinous acids R 2 P-OH is almost completely shifted to the side of the formation of a strong P=O bond (Scheme 2). The only known example of a thermally stable phosphinous acid is bis(trifluoromethyl)phosphinous acid, (CF 3 ) 2 P-OH [11], which was synthesized in 1960 by Burg and Griffiths [12]. Recently, Hoge et al. reported the solvent dependent tautomeric equilibrium between bis(pentafluorophenyl)phosphane oxide and the bis(pentafluorophenyl)phosphinous acid. The phosphinous acid is the dominated form in polar aprotic solvents, such as DMSO or DMF, but bis(pentafluorophenyl)phosphane oxide in toluene or CHCl 3 [13].
Gjikaj and co-workers synthesized the rubidium and cesium salts of 1 , the properties of which have been characterized by a combination of several techniques: FT-Raman and 31 P and 1 H MAS NMR [18]. They obtained the crystals of isotypic in the triclinic space group P-1 with one formula unit in the unit cell. The structures were built up by discrete (H 2 P 2 O 6 ) 2and (H 4 P 2 O 6 ) units in a staggered conformation for the P 2 O 6 skeleton and the corresponding alkali-metal cations. X-ray structural analysis confirmed the reported structures, which showed O•••HO-P hydrogen bonds between the (H 2 P 2 O 6 )] molecule. The P-P distance is 2.175 Å for Rb and 2.182 Å for Cs salts, the P-O bond lengths range from 1.501 to 1.569 Å, and both distances are consistent with the results of other works [13,16,24,[30][31][32]. Emami et al. synthesized organic-inorganic salts of hypodiphosphoric acid tetraalkylammonium cations, which have been characterized by X-ray crystallography, IR spectroscopy, and NMR measurements in the solid state [30]. The 31 P{ 1 H} CPMAS NMR measurements showed that the Na 4 P 2 O 6 •10H 2 O exhibits only one single resonance in 14.9 ppm, and for Rb salt at 12.5 and 13.5 ppm, and for Cs salt at 11.7 and 11.4 ppm. The 1 H MAS NMR spectra of Rb salt showed three well-resolved resonances signals at 16.1, 13.9, and 12.2 ppm, and for Cs salts at 16.1, 14.3, and 12.6 ppm. Based on the 31 P{ 1 H} CP-HETCOR spectrum of Rb salt, authors assigned the 31 P resonance at 13.5 ppm to the dihydrogenhypodiphosphate anion and the resonance at 12.5 ppm to the acid 1 moiety.
The acid 1 and its salts have attracted attention due to the discovery of the ferroelectricity properties of diammonium hypodiphosphate by Szklarz and co-workers [33,34]. Szklarz et al. reported the influence of the hydrogen bond for the ferroelectric phase transition shifts by about 20 K towards higher temperatures. They substituted hydrogen with deuterium in the molecule (NH 4 ) 2 H 2 P 2 O 6 , leading to compound (ND 4 ) 2 D 2 P 2 O 6 [35].
Generally, there are four routes for the syntheses of compounds 3 that contain the >P(O)-(O)P< moiety (Scheme 3). One route relies on converting dialkyl phosphorohalidate (5) (mainly chlorides) by alkali metals in heterogeneous (alkali metals) or homogeneous conditions (e.g., potassium naphthalenide, or alkali metals in liquid ammonia) (method A). The reaction of the electrophiles 5 with alkali metals is a complex transformation, and the data presented in the literature are often incompatible. Baudler [37,38] reported that the treatment of dialkyl phosphorochloridate 5 (formerly dialkyl chlorophosphate) with sodium in an inert solvent, such as xylene, toluene, or diethyl ether, at room temperature followed by gentle warming, resulted in a mixture of products consisting, in the main, of hypodiphosphoric acid esters (3) together with considerable amounts of anhydrides 10 (formerly pyrophosphoric acid esters). Similarly, the conversion of electron acceptors 5 by alkali metals in heterogeneous conditions often led to a complex reaction mixture [8]. This is not surprising because alkali metals can react with both starting material 5 and products (e.g., molecules 3 and 10), which are all electron acceptors [6].
Interestingly, method B presented in Scheme 3 relies on the conversion of dialkyl phosphonate 7 (formerly dialkylphosphite) with molecule 5, which produces tetraalkyl hypodiphosphates 3, tetraalkyl diphosphates (10) (formerly tetraalkylpyrophosphates), and mixed anhydrides 12 [39,40]. Comparing the reaction mixtures from both methods A and B, many similarities could be found, especially between the relative rate of product Generally, there are four routes for the syntheses of compounds 3 that contain the >P(O)-(O)P< moiety (Scheme 3). One route relies on converting dialkyl phosphorohalidate (5) (mainly chlorides) by alkali metals in heterogeneous (alkali metals) or homogeneous conditions (e.g., potassium naphthalenide, or alkali metals in liquid ammonia) (method A). The reaction of the electrophiles 5 with alkali metals is a complex transformation, and the data presented in the literature are often incompatible. Baudler [37,38] reported that the treatment of dialkyl phosphorochloridate 5 (formerly dialkyl chlorophosphate) with sodium in an inert solvent, such as xylene, toluene, or diethyl ether, at room temperature followed by gentle warming, resulted in a mixture of products consisting, in the main, of hypodiphosphoric acid esters (3) together with considerable amounts of anhydrides 10 (formerly pyrophosphoric acid esters). Similarly, the conversion of electron acceptors 5 by alkali metals in heterogeneous conditions often led to a complex reaction mixture [8]. This is not surprising because alkali metals can react with both starting material 5 and products (e.g., molecules 3 and 10), which are all electron acceptors [6].
Interestingly, method B presented in Scheme 3 relies on the conversion of dialkyl phosphonate 7 (formerly dialkylphosphite) with molecule 5, which produces tetraalkyl hypodiphosphates 3, tetraalkyl diphosphates (10) (formerly tetraalkylpyrophosphates), and mixed anhydrides 12 [39,40]. Comparing the reaction mixtures from both methods A and B, many similarities could be found, especially between the relative rate of product formation and their distribution profile [10]. For both routes, the mixtures formed are relatively complex.
In another paper, the same group of authors conducted a reaction between electrophile 5b and nucleophile 7a in benzene solution in the presence of triethylamine (Scheme 5). They isolated as the main constituent molecule 10b, contaminated with some amounts of anhydrides of phosphorus acids at different oxidation levels 12b and 14a. The authors claimed that compound 3a was not detected. In another paper, the same group of authors conducted a reaction between electrophile 5b and nucleophile 7a in benzene solution in the presence of triethylamine (Scheme 5). They isolated as the main constituent molecule 10b, contaminated with some amounts of anhydrides of phosphorus acids at different oxidation levels 12b and 14a. The authors claimed that compound 3a was not detected.
The authors in both experiments (Schemes 4 and 5) excluded the mixed anhydrides 12a or 12b as a key intermediate that could explain the presence of anhydrides 10a or 10b "oxidized" and 14a "reduced" with the absence of molecule 3a. It was found that under the conditions of the described reaction, the mixed anhydride 12b did not react with either nucleophile 7a in the presence of triethylamine at room temperature or electrophile 5a, which excluded its contribution to the reaction as an intermediate product, a precursor to the observed symmetric anhydrides 10 and 14. The authors claimed that in Scheme 5 the reaction went through the deoxygenation of diethyl phosphorochloridate (5b) by diethyl phosphonate (7a), which became diethyl hydrogen phosphate (16b). The acid 16b could react with the diethyl phosphorochloridate (5b), giving the main product tetraethyl diphosphate (10b). On the other hand, the diethyl phosphonate (7a) could react with deoxygenated diethyl chlorophosphite (20a), giving symmetric three coordinated anhydride tetraethyl diphosphite (14a). The acid 16b could react with electrophile 20a, giving mixed anhydride 12b (Scheme 6). These researchers demonstrated that anhydrides 10b and 14a were not the products of rearrangements of the originally arising anhydride 12b (Scheme 6). The authors in both experiments (Schemes 4 and 5) excluded the mixed anhydrides 12a or 12b as a key intermediate that could explain the presence of anhydrides 10a or 10b "oxidized" and 14a "reduced" with the absence of molecule 3a. It was found that under the conditions of the described reaction, the mixed anhydride 12b did not react with either nucleophile 7a in the presence of triethylamine at room temperature or electrophile 5a, which excluded its contribution to the reaction as an intermediate product, a precursor to the observed symmetric anhydrides 10 and 14. The authors claimed that in Scheme 5 the reaction went through the deoxygenation of diethyl phosphorochloridate (5b) by diethyl phosphonate (7a), which became diethyl hydrogen phosphate (16b). The acid 16b could react with the diethyl phosphorochloridate (5b), giving the main product tetraethyl diphosphate (10b). On the other hand, the diethyl phosphonate (7a) could react with deoxygenated diethyl chlorophosphite (20a), giving symmetric three coordinated anhydride tetraethyl diphosphite (14a). The acid 16b could react with electrophile 20a, giving mixed anhydride 12b (Scheme 6). These researchers demonstrated that anhydrides 10b and 14a were not the products of rearrangements of the originally arising anhydride 12b (Scheme 6). The >P-O − anion (the salts of molecules 7 and 8) is an ambident nucleophile having a hard centre on the oxygen and a soft one on the phosphorus atom. Accordingly, many examples can be indicated where such an anion reacted selectively, either via the phosphorus [42] or the oxygen [43], respectively (Scheme 7). Scheme 6. The possible reaction routes of the reaction between the electrophile 5b and nucleophile 7a.
The >P-O − anion (the salts of molecules 7 and 8) is an ambident nucleophile having a hard centre on the oxygen and a soft one on the phosphorus atom. Accordingly, many examples can be indicated where such an anion reacted selectively, either via the phosphorus [42] or the oxygen [43], respectively (Scheme 7). Scheme 6. The possible reaction routes of the reaction between the electrophile 5b and nucleophile 7a.
The >P-O − anion (the salts of molecules 7 and 8) is an ambident nucleophile having a hard centre on the oxygen and a soft one on the phosphorus atom. Accordingly, many examples can be indicated where such an anion reacted selectively, either via the phosphorus [42] or the oxygen [43], respectively (Scheme 7). Scheme 7. The dual reactivity of nucleophiles 7 and 8 (R 1 = R or RO).
Similar to the aforementioned reactivity, the ambident phosphorus nucleophiles 7 can react with phosphorus electrophiles 5 to produce both P-phosphorylated products, i.e., the tetraalkyl hypodiphosphoric acid esters 3, and O-phosphorylated products i.e., mixed anhydrides 12 (Scheme 8). The mixed anhydride 12 is a critical intermediate, explaining the formation of anhydrides 10 (Schemes 4 and 8). The pattern of the yield distribution depends on the substituent located on both phosphorus atoms (nucleophile and electrophile). The architecture of products 3 and 12, with the presence of single or double bonds between oxygen and phosphorus atoms, is controlled by steric and/or electronic factors of substituents R (Scheme 8). The reaction mixture analysis indicated that target Scheme 7. The dual reactivity of nucleophiles 7 and 8 (R 1 = R or RO).
Similar to the aforementioned reactivity, the ambident phosphorus nucleophiles 7 can react with phosphorus electrophiles 5 to produce both P-phosphorylated products, i.e., the tetraalkyl hypodiphosphoric acid esters 3, and O-phosphorylated products i.e., mixed anhydrides 12 (Scheme 8). The mixed anhydride 12 is a critical intermediate, explaining the formation of anhydrides 10 (Schemes 4 and 8). The pattern of the yield distribution depends on the substituent located on both phosphorus atoms (nucleophile and electrophile). The architecture of products 3 and 12, with the presence of single or double bonds between oxygen and phosphorus atoms, is controlled by steric and/or electronic factors of substituents R (Scheme 8). The reaction mixture analysis indicated that target products 3 were always accompanied by the "reduced" molecules 17 and/or 14 and "oxidized" compounds, resulting in anhydrides 10 (Scheme 8).
Steinberg conducted a chlorination reaction of diethyl phosphonate (7a) with CCl 4 in the presence of triethylamine (Scheme 10) [45]. He found the formation of high-boiling by-products, which, based on the refractive index values, were assigned a composition of 83% tetraethyl diphosphate (10b) and 17% tetraethyl hypodiphosphate (3a). The product composition was similar to earlier work undertaken by Nylen [46]. In 1932, Arbuzov et al. treated phosphate anion 7a′ with chlorine gas. In the first step of the reaction, a complex mixture of products was formed, which, after treatment with an excess of chlorine, turned into diethyl chlorophosphite (20a) (Scheme 11) [47]. Scheme 11. The reaction between nucleophile 7a′ with chlorine gas (no yield was given).
Zwierzak et al. carried out a reaction of sodium salt of diethyl phosphonate (7a′) and diethyl phosphorochloridate (5b) [39]. The authors obtained nucleophile 7a′ through treating diethyl phosphonate (7a) by metallic sodium. They noted that the use of 25% excess of nucleophilic reagent 7a′ avoided the formation of anhydride 10b. After the distillation of the reaction mixture, they obtained a forehead which, from a vigorous reaction with water, attributed to the mixed anhydride 12b, and the main fraction, based on the refractive coefficient considered to be a mixture of molecule 3a and anhydride 12b in a ratio of approximately 5:1 (Scheme 12). Zwierzak's team reported the efficient synthesis of mixed anhydrides 12 based on the reaction of dialkyl phosphonate 7 with dialkyl phosphorochloridate 5 in benzene, in the presence of pyridine (reaction A; Scheme 13) [48]. The products were purified by vacuum distillation. The authors mentioned that they were unable to purify mixed anhydrides 12g. From the present perspective, the likely explanation of this difficulty is the existence of mixed anhydrides 12c, 12d, and 12g as diastereoisomeric mixtures (meso and rac) among the products (reaction A; Scheme 13). In 1932, Arbuzov et al. treated phosphate anion 7a with chlorine gas. In the first step of the reaction, a complex mixture of products was formed, which, after treatment with an excess of chlorine, turned into diethyl chlorophosphite (20a) (Scheme 11) [47]. In 1932, Arbuzov et al. treated phosphate anion 7a′ with chlorine gas. In the first step of the reaction, a complex mixture of products was formed, which, after treatment with an excess of chlorine, turned into diethyl chlorophosphite (20a) (Scheme 11) [47]. Scheme 11. The reaction between nucleophile 7a′ with chlorine gas (no yield was given).
Zwierzak et al. carried out a reaction of sodium salt of diethyl phosphonate (7a′) and diethyl phosphorochloridate (5b) [39]. The authors obtained nucleophile 7a′ through treating diethyl phosphonate (7a) by metallic sodium. They noted that the use of 25% excess of nucleophilic reagent 7a′ avoided the formation of anhydride 10b. After the distillation of the reaction mixture, they obtained a forehead which, from a vigorous reaction with water, attributed to the mixed anhydride 12b, and the main fraction, based on the refractive coefficient considered to be a mixture of molecule 3a and anhydride 12b in a ratio of approximately 5:1 (Scheme 12). Zwierzak's team reported the efficient synthesis of mixed anhydrides 12 based on the reaction of dialkyl phosphonate 7 with dialkyl phosphorochloridate 5 in benzene, in the presence of pyridine (reaction A; Scheme 13) [48]. The products were purified by vacuum distillation. The authors mentioned that they were unable to purify mixed anhydrides 12g. From the present perspective, the likely explanation of this difficulty is the existence of mixed anhydrides 12c, 12d, and 12g as diastereoisomeric mixtures (meso and rac) among the products (reaction A; Scheme 13). Scheme 11. The reaction between nucleophile 7a with chlorine gas (no yield was given).
Zwierzak et al. carried out a reaction of sodium salt of diethyl phosphonate (7a ) and diethyl phosphorochloridate (5b) [39]. The authors obtained nucleophile 7a through treating diethyl phosphonate (7a) by metallic sodium. They noted that the use of 25% excess of nucleophilic reagent 7a avoided the formation of anhydride 10b. After the distillation of the reaction mixture, they obtained a forehead which, from a vigorous reaction with water, attributed to the mixed anhydride 12b, and the main fraction, based on the refractive coefficient considered to be a mixture of molecule 3a and anhydride 12b in a ratio of approximately 5:1 (Scheme 12). In 1932, Arbuzov et al. treated phosphate anion 7a′ with chlorine gas. In the first step of the reaction, a complex mixture of products was formed, which, after treatment with an excess of chlorine, turned into diethyl chlorophosphite (20a) (Scheme 11) [47]. Scheme 11. The reaction between nucleophile 7a′ with chlorine gas (no yield was given).
Zwierzak et al. carried out a reaction of sodium salt of diethyl phosphonate (7a′) and diethyl phosphorochloridate (5b) [39]. The authors obtained nucleophile 7a′ through treating diethyl phosphonate (7a) by metallic sodium. They noted that the use of 25% excess of nucleophilic reagent 7a′ avoided the formation of anhydride 10b. After the distillation of the reaction mixture, they obtained a forehead which, from a vigorous reaction with water, attributed to the mixed anhydride 12b, and the main fraction, based on the refractive coefficient considered to be a mixture of molecule 3a and anhydride 12b in a ratio of approximately 5:1 (Scheme 12). Zwierzak's team reported the efficient synthesis of mixed anhydrides 12 based on the reaction of dialkyl phosphonate 7 with dialkyl phosphorochloridate 5 in benzene, in the presence of pyridine (reaction A; Scheme 13) [48]. The products were purified by vacuum distillation. The authors mentioned that they were unable to purify mixed anhydrides 12g. From the present perspective, the likely explanation of this difficulty is the existence of mixed anhydrides 12c, 12d, and 12g as diastereoisomeric mixtures (meso and rac) among the products (reaction A; Scheme 13). Zwierzak's team reported the efficient synthesis of mixed anhydrides 12 based on the reaction of dialkyl phosphonate 7 with dialkyl phosphorochloridate 5 in benzene, in the presence of pyridine (reaction A; Scheme 13) [48]. The products were purified by vacuum distillation. The authors mentioned that they were unable to purify mixed anhydrides 12g.
From the present perspective, the likely explanation of this difficulty is the existence of mixed anhydrides 12c, 12d, and 12g as diastereoisomeric mixtures (meso and rac) among the products (reaction A; Scheme 13).
Three years later, in a brief communication, the same group of authors reported the synthesis of tetraethyl hypodiphosphate (3a) with a yield of 53% [49]. They reacted sodium salts of diethyl phosphonate (7a′) (50% excess) with diethyl phosphorochloridate (5a) in benzene solution at 0-5 °C (reaction B; Scheme 13). The product was isolated by distillation in vacuo and finally characterized by Raman spectroscopy. The characteristic frequency of the P-P bond was observed at 257 cm −l . Additionally, the authors realized that the esterification of anhydrous hypodiphosphoric acid (1) by diazoalkanes as an alternative procedure was unsuitable because of the poor accessibility of anhydrous acid 1.
All yields were referred to products isolated by crystallization and identified by chemical properties, IR spectra, and elemental analysis. The authors compared the stability of acyclic tetralkyl hypodiphosphoric acid esters 3 with obtained molecule 3b, and found that the acyclic molecules were relatively easily hydrolyzed compared to cyclic 3b, which remained almost intact when refluxed with water in tetrahydrofuran solution for 20 h.
Three years later, in a brief communication, the same group of authors reported the synthesis of tetraethyl hypodiphosphate (3a) with a yield of 53% [49]. They reacted sodium salts of diethyl phosphonate (7a ) (50% excess) with diethyl phosphorochloridate (5a) in benzene solution at 0-5 • C (reaction B; Scheme 13). The product was isolated by distillation in vacuo and finally characterized by Raman spectroscopy. The characteristic frequency of the P-P bond was observed at 257 cm −l . Additionally, the authors realized that the esterification of anhydrous hypodiphosphoric acid (1) by diazoalkanes as an alternative procedure was unsuitable because of the poor accessibility of anhydrous acid 1.
Three years later, in a brief communication, the same group of authors reported the synthesis of tetraethyl hypodiphosphate (3a) with a yield of 53% [49]. They reacted sodium salts of diethyl phosphonate (7a′) (50% excess) with diethyl phosphorochloridate (5a) in benzene solution at 0-5 °C (reaction B; Scheme 13). The product was isolated by distillation in vacuo and finally characterized by Raman spectroscopy. The characteristic frequency of the P-P bond was observed at 257 cm −l . Additionally, the authors realized that the esterification of anhydrous hypodiphosphoric acid (1) by diazoalkanes as an alternative procedure was unsuitable because of the poor accessibility of anhydrous acid 1.
All yields were referred to products isolated by crystallization and identified by chemical properties, IR spectra, and elemental analysis. The authors compared the stability of acyclic tetralkyl hypodiphosphoric acid esters 3 with obtained molecule 3b, and found that the acyclic molecules were relatively easily hydrolyzed compared to cyclic 3b, which remained almost intact when refluxed with water in tetrahydrofuran solution for 20 h.
Mehrotra et al. reported the syntheses of unsymmetric hypodiphosphoric acid esters i.e., diisopropyl dipropyl hypodiphosphate (3c), which was obtained in the reaction of Scheme 14. The reaction between nucleophile 7a and electrophile 5a.
All yields were referred to products isolated by crystallization and identified by chemical properties, IR spectra, and elemental analysis. The authors compared the stability of acyclic tetralkyl hypodiphosphoric acid esters 3 with obtained molecule 3b, and found that the acyclic molecules were relatively easily hydrolyzed compared to cyclic 3b, which remained almost intact when refluxed with water in tetrahydrofuran solution for 20 h.
Zhao's team used dialkyl phosphonates 5 (Alkyl = Me, Et, Pr i ) in the presence of CCl 4 and triethylamine to synthesize N-phosphorylated amino acids and dipeptides [51]. In examining this reaction, they carried out a blank test, i.e., the chlorination of phosphorus nucleophiles 7 in the absence of N-nucleophile. However, in the reaction mixture, in addition to the expected dialkyl phosphorochloridates 5, they observed tetraalkyl diphosphate 10 anhydrides. The reaction presented here is similar to the presented above in Scheme 10 by Steinberg [45]. The yields of products 5 and 10 depended on substitutes located on phosphorus atoms.
Rachoń et al. demonstrated the reduction of mixed anhydride 1,3-di-tert-butyl-1,3diphenyldiphosphoxane 1-oxide (mixture of meso and rac) 13a,b to acid anion 16c and tertbutyl(phenyl)phosphido anion 19a [43]. The obtained intermediates were subsequently treated by sulfur and methyl iodide. The final products tert-butyl(phenyl)phosphinic acid (16c) and methyl tert-butyl(phenyl)phosphinodithioate were isolated with good yields (Scheme 15). The presented cleavage is unequivocal and shows that the acid anion 16c is a good leaving group. diisopropyl phosphorochloridate (5c) with the diesters of phosphonic acids activated by the 1,3,2-dioxarsolan moiety ((RO)2P(O)-As(OCH2)2) [50]. Unfortunately, the authors did not present either the synthesis procedure or the characteristics of the resulting products. Zhao's team used dialkyl phosphonates 5 (Alkyl = Me, Et, Pr i ) in the presence of CCl4 and triethylamine to synthesize N-phosphorylated amino acids and dipeptides [51]. In examining this reaction, they carried out a blank test, i.e., the chlorination of phosphorus nucleophiles 7 in the absence of N-nucleophile. However, in the reaction mixture, in addition to the expected dialkyl phosphorochloridates 5, they observed tetraalkyl diphosphate 10 anhydrides. The reaction presented here is similar to the presented above in Scheme 10 by Steinberg [45]. The yields of products 5 and 10 depended on substitutes located on phosphorus atoms.
Rachoń et al. demonstrated the reduction of mixed anhydride 1,3-di-tert-butyl-1,3diphenyldiphosphoxane 1-oxide (mixture of meso and rac) 13a,b to acid anion 16c′ and tert-butyl(phenyl)phosphido anion 19a′ [43]. The obtained intermediates were subsequently treated by sulfur and methyl iodide. The final products tert-butyl(phenyl)phosphinic acid (16c) and methyl tert-butyl(phenyl)phosphinodithioate were isolated with good yields (Scheme 15). The presented cleavage is unequivocal and shows that the acid anion 16c′ is a good leaving group. The beginning of this section results from a reduction of dialkyl phosphorohalidate (5) by alkali metals in heterogeneous (alkali metals) or homogeneous conditions (e.g., potassium naphthalenide or alkali metals in liquid ammonia) were presented (method A; Scheme 3). The above results showed that mixed anhydrides 12 can be successfully reduced, and the reduction products were acid anion 16c′ and tert-butyl(phenyl)phosphido anion 19a′ (Scheme 15). Their presence can explain alternatively the mysterious origin of "oxidized" and "reduced" anhydrides 10 and 14, respectively, during the synthesis of molecules 3, which will be further explained.
The reactivity of ambident nucleophiles 7 is much less predictable than nucleophiles 19, having only one center on the phosphorus atom. The >P-O − 7′ nucleophilicity depends on the type of substituents located on the phosphorus atom. It is assumed that alkoxy, aryloxy substituents directly connected to phosphorus atoms, should increase their nucleophilicity because they are α-nucleophiles. The alpha effect is the interaction of the two lone pairs on adjacent oxygen atoms raises the HOMO of the anions and makes them better and softer nucleophiles.

The Isomerism between Molecules Possess >P(=O)-P< and >P-O-P< Fragments
To understand the possibilities and predictability of the syntheses of hypodiphosphoric acid 1, their esters 3, diphosphine dioxides 4, and related compounds 4 with P-P bond, it is necessary to know which substituents stabilize the structure of >P The beginning of this section results from a reduction of dialkyl phosphorohalidate (5) by alkali metals in heterogeneous (alkali metals) or homogeneous conditions (e.g., potassium naphthalenide or alkali metals in liquid ammonia) were presented (method A; Scheme 3). The above results showed that mixed anhydrides 12 can be successfully reduced, and the reduction products were acid anion 16c and tert-butyl(phenyl)phosphido anion 19a (Scheme 15). Their presence can explain alternatively the mysterious origin of "oxidized" and "reduced" anhydrides 10 and 14, respectively, during the synthesis of molecules 3, which will be further explained.
The reactivity of ambident nucleophiles 7 is much less predictable than nucleophiles 19, having only one center on the phosphorus atom. The >P-O − 7 nucleophilicity depends on the type of substituents located on the phosphorus atom. It is assumed that alkoxy, aryloxy substituents directly connected to phosphorus atoms, should increase their nucleophilicity because they are α-nucleophiles. The alpha effect is the interaction of the two lone pairs on adjacent oxygen atoms raises the HOMO of the anions and makes them better and softer nucleophiles.

The Isomerism between Molecules Possess >P(=O)-P< and >P-O-P< Fragments
To understand the possibilities and predictability of the syntheses of hypodiphosphoric acid 1, their esters 3, diphosphine dioxides 4, and related compounds 4 with P-P bond, it is necessary to know which substituents stabilize the structure of >P(O)-(O)P< type molecules. The best models for such a study are anhydrides 14, 15, and their isomers 17 and 18, respectively.
Anderson's team provided evidence that the diethyl phosphonate (7a) gave only O-phosphorylated products [52]. They developed the synthesis of tetraethyl diphosphite (14a) in a clean and fast (15 min) procedure. The obtained anhydride 14a with trivalent phosphorus atoms is a stable compound that was not transformed into isomeric diethyl (diethoxyphosphaneyl)phosphonate (17a) (Scheme 16).  [52]. They developed the synthesis of tetraethyl diphosphite (14a) in a clean and fast (15 min) procedure. The obtained anhydride 14a with trivalent phosphorus atoms is a stable compound that was not transformed into isomeric diethyl (diethoxyphosphaneyl)phosphonate (17a) (Scheme 16). Understanding the equilibrium between anhydride 14a and molecule 17a (Scheme 16) and their origin is the key to predict the efficiency of the synthesis of hypodiphosphoric acid esters 3, diphosphine dioxides 4, and related compounds 4 with P-P bond.
Anderson et al. obtained a stable unsymmetrical anhydride with trivalent phosphorus atom, namely 1,3,2-dioxaphospholan-2-yl diethyl phosphite (14b) [53]. The authors did not observe any isomerization processes which could produce molecule 17b or symmetrization transformations leading to symmetrical anhydrides 14a or 14c with identical substituents on both phosphorus atoms (Scheme 17).  [52]. They developed the synthesis of tetraethyl diphosphite (14a) in a clean and fast (15 min) procedure. The obtained anhydride 14a with trivalent phosphorus atoms is a stable compound that was not transformed into isomeric diethyl (diethoxyphosphaneyl)phosphonate (17a) (Scheme 16). Understanding the equilibrium between anhydride 14a and molecule 17a (Scheme 16) and their origin is the key to predict the efficiency of the synthesis of hypodiphosphoric acid esters 3, diphosphine dioxides 4, and related compounds 4 with P-P bond.
Anderson et al. obtained a stable unsymmetrical anhydride with trivalent phosphorus atom, namely 1,3,2-dioxaphospholan-2-yl diethyl phosphite (14b) [53]. The authors did not observe any isomerization processes which could produce molecule 17b or symmetrization transformations leading to symmetrical anhydrides 14a or 14c with identical substituents on both phosphorus atoms (Scheme 17). Crofts' team synthesized a stable anhydride, 2,2′-oxybis(2,3-dihydro-1H-isophosphindole) (15a) (Scheme 18) [54]. Crofts' team synthesized a stable anhydride, 2,2 -oxybis(2,3-dihydro-1H-isophosphindole) (15a) (Scheme 18) [54].    [56]. For the first time, they used reagents with dialkylamino substituents on phosphorus atoms and identified molecules 18 as the final products of this reaction (Method A, Scheme 20). However, when a nucleophile was used, which was generated in the reaction between molecules 8 and triethylamine, not only were the compounds 18 isolated, but anhydrides 15 were also obtained (Method B, Scheme 20). These researchers concluded that molecules 18 were the primary products, which were subsequently isomerized into the anhydrides 15 and back. It was found that the rise of temperature increased the concentration of anhydrides 15. These molecules are easy to be distinguished from each other. The molecules 18 possess characteristic large 1 J PP couplings in 125-290 Hz due to inequivalent phosphorus atoms. The symmetric anhydrides 15 (both phosphorus atoms are the same) were identified based on the singlets. In another work, the same team studied reactions in which one substrate had alkoxy residues, and the other had dialkylamino (Scheme 21) [57]. Nucleophilic reagents were generated in the reaction with triethylamine. The authors observed a complicated reaction mixture composed of monoxides in which both substituents on each phosphorus atom are different (e.g., 18a) and the same (e.g., 18b) molecules 18 and anhydrides 15. They explained the symmetrization process by the direct reaction of anhydrides 15 with electrophile 20d. These researchers concluded that molecules 18 were the primary products, which were subsequently isomerized into the anhydrides 15 and back. It was found that the rise of temperature increased the concentration of anhydrides 15. These molecules are easy to be distinguished from each other. The molecules 18 possess characteristic large 1 JPP couplings in 125-290 Hz due to inequivalent phosphorus atoms. The symmetric anhydrides 15 (both phosphorus atoms are the same) were identified based on the singlets. In another work, the same team studied reactions in which one substrate had alkoxy residues, and the other had dialkylamino (Scheme 21) [57]. Nucleophilic reagents were generated in the reaction with triethylamine. The authors observed a complicated reaction mixture composed of monoxides in which both substituents on each phosphorus atom are different (e.g., 18a) and the same (e.g., 18b) molecules 18 and anhydrides 15. They explained the symmetrization process by the direct reaction of anhydrides 15 with electrophile 20d. In 1984, the same team discussed the possibility of equilibrium between molecules 18 and 15 [58]. At the temperature of −78 °C, compounds 18 were formed, while anhydrides 15 were formed when a reaction was conducted under thermodynamic control. In this case, the nucleophile was generated by reacting molecule 8 with triethylamine at room temperature. Similar to previous research, the authors observed isomerization processes giving a complicated reaction mixture composed of symmetric and unsymmetric molecules 18 and anhydrides 15 [57]. This was confirmed by the fact that molecules 18 spontaneously isomerized into anhydrides 15, which were more thermodynamically stable [59]. In 1983, the same team in another paper described two reactions of a new type of the trivalent phosphorus electrophile, i.e., mixed anhydride 21a, with nucleophiles 7d and 7e, respectively [60]. The researchers received stable symmetric 14e and unsymmetric 14f These researchers concluded that molecules 18 were the primary products, which were subsequently isomerized into the anhydrides 15 and back. It was found that the rise of temperature increased the concentration of anhydrides 15. These molecules are easy to be distinguished from each other. The molecules 18 possess characteristic large 1 JPP couplings in 125-290 Hz due to inequivalent phosphorus atoms. The symmetric anhydrides 15 (both phosphorus atoms are the same) were identified based on the singlets. In another work, the same team studied reactions in which one substrate had alkoxy residues, and the other had dialkylamino (Scheme 21) [57]. Nucleophilic reagents were generated in the reaction with triethylamine. The authors observed a complicated reaction mixture composed of monoxides in which both substituents on each phosphorus atom are different (e.g., 18a) and the same (e.g., 18b) molecules 18 and anhydrides 15. They explained the symmetrization process by the direct reaction of anhydrides 15 with electrophile 20d. In 1984, the same team discussed the possibility of equilibrium between molecules 18 and 15 [58]. At the temperature of −78 °C, compounds 18 were formed, while anhydrides 15 were formed when a reaction was conducted under thermodynamic control. In this case, the nucleophile was generated by reacting molecule 8 with triethylamine at room temperature. Similar to previous research, the authors observed isomerization processes giving a complicated reaction mixture composed of symmetric and unsymmetric molecules 18 and anhydrides 15 [57]. This was confirmed by the fact that molecules 18 spontaneously isomerized into anhydrides 15, which were more thermodynamically stable [59]. In 1983, the same team in another paper described two reactions of a new type of the trivalent phosphorus electrophile, i.e., mixed anhydride 21a, with nucleophiles 7d and 7e, respectively [60]. The researchers received stable symmetric 14e and unsymmetric 14f In 1984, the same team discussed the possibility of equilibrium between molecules 18 and 15 [58]. At the temperature of −78 • C, compounds 18 were formed, while anhydrides 15 were formed when a reaction was conducted under thermodynamic control. In this case, the nucleophile was generated by reacting molecule 8 with triethylamine at room temperature. Similar to previous research, the authors observed isomerization processes giving a complicated reaction mixture composed of symmetric and unsymmetric molecules 18 and anhydrides 15 [57]. This was confirmed by the fact that molecules 18 spontaneously isomerized into anhydrides 15, which were more thermodynamically stable [59]. In 1983, the same team in another paper described two reactions of a new type of the trivalent phosphorus electrophile, i.e., mixed anhydride 21a, with nucleophiles 7d and 7e, respectively [60]. The researchers received stable symmetric 14e and unsymmetric 14f anhydrides, which did not isomerize into appropriate molecules 17 (Scheme 23). It is important to mention that 2-((4-methyl-1,3,2-dioxaphospholan-2-yl)oxy)benzo[d] [1][2][3]dioxaphosphole (14f) did not symmetrize.  The presence of halogen in the phosphorus electrophile constitution could influence side reactions, such as X-philic substitution [63], disproportionation, or symmetrization processes. Due to these drawbacks, another type of reaction of trivalent phosphorus electrophile, with 1H-imidazole as a leaving group, was studied, in which the 1H-imidazole cyclic phosphate's derivative 1-(5,5-dimethyl-1,3,2-dioxaphosphinan-2-yl)-1H-imidazole (21b) was selected (Scheme 26). The only identified product was a stable anhydride, 2,2′oxybis(5,5-dimethyl-1,3,2-dioxaphosphinane) (14h) [64].   The presence of halogen in the phosphorus electrophile constitution could influence side reactions, such as X-philic substitution [63], disproportionation, or symmetrization processes. Due to these drawbacks, another type of reaction of trivalent phosphorus electrophile, with 1H-imidazole as a leaving group, was studied, in which the 1H-imidazole cyclic phosphate's derivative 1-(5,5-dimethyl-1,3,2-dioxaphosphinan-2-yl)-1H-imidazole (21b) was selected (Scheme 26). The only identified product was a stable anhydride, 2,2′oxybis(5,5-dimethyl-1,3,2-dioxaphosphinane) (14h) [64].  The presence of halogen in the phosphorus electrophile constitution could influence side reactions, such as X-philic substitution [63], disproportionation, or symmetrization processes. Due to these drawbacks, another type of reaction of trivalent phosphorus electrophile, with 1H-imidazole as a leaving group, was studied, in which the 1H-imidazole cyclic phosphate's derivative 1-(5,5-dimethyl-1,3,2-dioxaphosphinan-2-yl)-1H-imidazole (21b) was selected (Scheme 26). The only identified product was a stable anhydride, 2,2′oxybis(5,5-dimethyl-1,3,2-dioxaphosphinane) (14h) [64]. The presence of halogen in the phosphorus electrophile constitution could influence side reactions, such as X-philic substitution [63], disproportionation, or symmetrization processes. Due to these drawbacks, another type of reaction of trivalent phosphorus electrophile, with 1H-imidazole as a leaving group, was studied, in which the 1H-imidazole cyclic phosphate's derivative 1-(5,5-dimethyl-1,3,2-dioxaphosphinan-2-yl)-1H-imidazole (21b) was selected (Scheme 26). The only identified product was a stable anhydride, 2,2 -oxybis(5,5-dimethyl-1,3,2-dioxaphosphinane) (14h) [64].

P-Centered Radicals Dimerization
The study of reactions involving organophosphorus radicals has a long history. However, the chemistry of the organophosphorus radical dimerization in terms of mechanism, variation of substitution groups, and reaction conditions is still not sufficiently illustrated [65,66]. The phosphinoyl and phosphonyl radicals are non-planar and, as a result, had a variable degree of s-character [66]. From a theoretical point of view, the synthesis of molecules 3 could be realized by the radical dimerization (Scheme 3; Method C). Romakhin et al. investigated the reactivity of phosphorus radicals generated electrochemically in acetonitrile solution (ACN) [67] and found that the dimerization process could produce two different types of compounds 3 and 12 (Scheme 27). The metal (sodium and lithium) salts of dialkyl phosphonates 7′ could be electrochemically oxidized, in which the one-electron oxidation of metal salts 7′ was allowed to generate phosphorus radicals >P(O) • (Scheme 27). The authors used acetonitrile as a solvent for the electrochemical oxidation processes. It was noticed that the free radicals were generated in the presence of nucleophiles 7′. A complex product mixture was identified, including anhydrides 10, 12, and 14, and molecules 3 in the minority, which have been confirmed by 31 P { 1 H} NMR. The origin of anhydrides 14 was the result of the reaction between nucleophiles 7′ and anhydrides 12 (Scheme 28), which has already been proved by Nycz et al. [68]. The chemical shifts of molecules 3 are given: tetraethyl hypodiphosphate (3a) δ = 8 ppm; tetrapropyl hypodiphosphate (3c) δ= 10 ppm, tetraisopropyl hypodiphosphate (3d) δ = 6 ppm and tetraisobutyl hypodiphosphate (3e) δ = 11 ppm. The authors found that lithium cation could stimulate the formation of P-P bonds, resulting in higher yields of compound 3. Scheme 28. The reaction between nucleophiles 7′ and anhydrides 12 (M = Na, Li).
Zhou and co-workers showed the first copper-catalyzed P-P bond-forming reactions [69]. They found that the copper-catalyzed P-P bond coupling reaction was highly Scheme 26. The synthesis of molecule 14h.

P-Centered Radicals Dimerization
The study of reactions involving organophosphorus radicals has a long history. However, the chemistry of the organophosphorus radical dimerization in terms of mechanism, variation of substitution groups, and reaction conditions is still not sufficiently illustrated [65,66]. The phosphinoyl and phosphonyl radicals are non-planar and, as a result, had a variable degree of s-character [66]. From a theoretical point of view, the synthesis of molecules 3 could be realized by the radical dimerization (Scheme 3; Method C). Romakhin et al. investigated the reactivity of phosphorus radicals generated electrochemically in acetonitrile solution (ACN) [67] and found that the dimerization process could produce two different types of compounds 3 and 12 (Scheme 27).

P-Centered Radicals Dimerization
The study of reactions involving organophosphorus radicals has a long history. However, the chemistry of the organophosphorus radical dimerization in terms of mechanism, variation of substitution groups, and reaction conditions is still not sufficiently illustrated [65,66]. The phosphinoyl and phosphonyl radicals are non-planar and, as a result, had a variable degree of s-character [66]. From a theoretical point of view, the synthesis of molecules 3 could be realized by the radical dimerization (Scheme 3; Method C). Romakhin et al. investigated the reactivity of phosphorus radicals generated electrochemically in acetonitrile solution (ACN) [67] and found that the dimerization process could produce two different types of compounds 3 and 12 (Scheme 27). The metal (sodium and lithium) salts of dialkyl phosphonates 7′ could be electrochemically oxidized, in which the one-electron oxidation of metal salts 7′ was allowed to generate phosphorus radicals >P(O) • (Scheme 27). The authors used acetonitrile as a solvent for the electrochemical oxidation processes. It was noticed that the free radicals were generated in the presence of nucleophiles 7′. A complex product mixture was identified, including anhydrides 10, 12, and 14, and molecules 3 in the minority, which have been confirmed by 31 P { 1 H} NMR. The origin of anhydrides 14 was the result of the reaction between nucleophiles 7′ and anhydrides 12 (Scheme 28), which has already been proved by Nycz et al. [68]. The chemical shifts of molecules 3 are given: tetraethyl hypodiphosphate (3a) δ = 8 ppm; tetrapropyl hypodiphosphate (3c) δ= 10 ppm, tetraisopropyl hypodiphosphate (3d) δ = 6 ppm and tetraisobutyl hypodiphosphate (3e) δ = 11 ppm. The authors found that lithium cation could stimulate the formation of P-P bonds, resulting in higher yields of compound 3. Scheme 28. The reaction between nucleophiles 7′ and anhydrides 12 (M = Na, Li).
Zhou and co-workers showed the first copper-catalyzed P-P bond-forming reactions [69]. They found that the copper-catalyzed P-P bond coupling reaction was highly The metal (sodium and lithium) salts of dialkyl phosphonates 7 could be electrochemically oxidized, in which the one-electron oxidation of metal salts 7 was allowed to generate phosphorus radicals >P(O) • (Scheme 27). The authors used acetonitrile as a solvent for the electrochemical oxidation processes. It was noticed that the free radicals were generated in the presence of nucleophiles 7 . A complex product mixture was identified, including anhydrides 10, 12, and 14, and molecules 3 in the minority, which have been confirmed by 31 P { 1 H} NMR. The origin of anhydrides 14 was the result of the reaction between nucleophiles 7 and anhydrides 12 (Scheme 28), which has already been proved by Nycz et al. [68]. The chemical shifts of molecules 3 are given: tetraethyl hypodiphosphate (3a) δ = 8 ppm; tetrapropyl hypodiphosphate (3c) δ= 10 ppm, tetraisopropyl hypodiphosphate (3d) δ = 6 ppm and tetraisobutyl hypodiphosphate (3e) δ = 11 ppm. The authors found that lithium cation could stimulate the formation of P-P bonds, resulting in higher yields of compound 3.

P-Centered Radicals Dimerization
The study of reactions involving organophosphorus radicals has a long history. However, the chemistry of the organophosphorus radical dimerization in terms of mechanism, variation of substitution groups, and reaction conditions is still not sufficiently illustrated [65,66]. The phosphinoyl and phosphonyl radicals are non-planar and, as a result, had a variable degree of s-character [66]. From a theoretical point of view, the synthesis of molecules 3 could be realized by the radical dimerization (Scheme 3; Method C). Romakhin et al. investigated the reactivity of phosphorus radicals generated electrochemically in acetonitrile solution (ACN) [67] and found that the dimerization process could produce two different types of compounds 3 and 12 (Scheme 27). The metal (sodium and lithium) salts of dialkyl phosphonates 7′ could be electrochemically oxidized, in which the one-electron oxidation of metal salts 7′ was allowed to generate phosphorus radicals >P(O) • (Scheme 27). The authors used acetonitrile as a solvent for the electrochemical oxidation processes. It was noticed that the free radicals were generated in the presence of nucleophiles 7′. A complex product mixture was identified, including anhydrides 10, 12, and 14, and molecules 3 in the minority, which have been confirmed by 31 P { 1 H} NMR. The origin of anhydrides 14 was the result of the reaction between nucleophiles 7′ and anhydrides 12 (Scheme 28), which has already been proved by Nycz et al. [68]. The chemical shifts of molecules 3 are given: tetraethyl hypodiphosphate (3a) δ = 8 ppm; tetrapropyl hypodiphosphate (3c) δ= 10 ppm, tetraisopropyl hypodiphosphate (3d) δ = 6 ppm and tetraisobutyl hypodiphosphate (3e) δ = 11 ppm. The authors found that lithium cation could stimulate the formation of P-P bonds, resulting in higher yields of compound 3. Scheme 28. The reaction between nucleophiles 7′ and anhydrides 12 (M = Na, Li).

The Diphosphine Dioxides
The hypodiphosphoric acid esters 3 and the diphosphine dioxides 4 are related molecules, and they are differentiated by the types of substituents located on phosphorus atoms. The hypodiphosphoric acid esters 3 possess alkoxy and/or aryloxy substituents, and diphosphine dioxides 4 have alkyl and/or aryl (Figure 1). Other molecules like hypodiphosphonates 4 have mixed substituents, i.e., alkyl or aryl and alkoxy or aryloxy or others. The substitution groups disposed of direct impact on the feasibility and efficiency of the reactions. Even nowadays, we lack efficient methods for the synthesis of these Scheme 30. Synthesis of molecules 3 and 4 via P-centered radicals dimerization (M = Na, Li).

The Diphosphine Dioxides
The hypodiphosphoric acid esters 3 and the diphosphine dioxides 4 are related molecules, and they are differentiated by the types of substituents located on phosphorus atoms. The hypodiphosphoric acid esters 3 possess alkoxy and/or aryloxy substituents, and diphosphine dioxides 4 have alkyl and/or aryl (Figure 1). Other molecules like hypodiphosphonates 4 have mixed substituents, i.e., alkyl or aryl and alkoxy or aryloxy or others. The substitution groups disposed of direct impact on the feasibility and efficiency of the reactions. Even nowadays, we lack efficient methods for the synthesis of these compounds.
The presented data are often incompatible and sometimes contain notable mistakes, e.g., 31 P NMR chemical shifts of both meso-and rac-di-1,2-tert-butyl-1,2-diphenyldiphosphine 1,2-dioxides (4b,c) as the same value, namely 48.9 ppm [70,73]. However, their chemical shifts are very far from each other (δ rac. = 50.7 and δ meso = 39.8 ppm) [68]. It should be noted here that the syntheses procedures of related diphosphines 23 of the R 2 P-PR 2 type (R = Alkyl, Aryl) have been described much better, and the results reported in the literature outlined the application, scope, and limitations of the existing protocols [74][75][76].
The difficulty and even the efficiency of the synthesis were attributed to the equilibrium between molecules 15 and 18, where trivalent central phosphorus atom underwent pentavalent. However, the diphosphoxane (CF 3 ) 2 P-O-P(CF 3 ) 2 was only one exception, which was almost completely shifted to the side of the phosphinous acids, with both trivalent phosphorus atoms. The diphosphoxane represented a unique compound showing no rearrangement to a phosphane oxide [13]. The diphosphoxane could be obtained in the reaction between (CF 3 ) 2 PI with silver carbonate giving the final product an 80% yield [13].
The difficulty and even the efficiency of the synthesis were attributed to the equilibrium between molecules 15 and 18, where trivalent central phosphorus atom underwent pentavalent. However, the diphosphoxane (CF3)2P-O-P(CF3)2 was only one exception, which was almost completely shifted to the side of the phosphinous acids, with both trivalent phosphorus atoms. The diphosphoxane represented a unique compound showing no rearrangement to a phosphane oxide [13]. The diphosphoxane could be obtained in the reaction between (CF3)2PI with silver carbonate giving the final product an 80% yield [13].

Some Synthetic Approaches of Diphosphine Dioxides
The diphosphine dioxides 4 can be obtained through the same synthetic routes as previously described in hypodiphosphoric acid esters 3 (Scheme 3). The first method is a reaction between electrophiles 6, with electron donors such as the alkali metals in heterogeneous (Li, Na, K) or homogenous conditions (potassium naphthalenide, Li, Na, K in NH3liq.). The first application of the Wurtz-Fittig type synthesis for the P-P bond formation was the coupling of PI3 with mercury to diphosphorus tetraiodide, reported by Besson at the end of the XIX century [80]. Other metals, such as lithium [81], sodium [82][83][84][85][86], potassium [85,86], and magnesium [87,88], in organic solvents have been successfully used for the coupling of the less reactive but more accessible phosphinous halides.
The reaction of compounds 6 with alkali metals is complicated, and the data presented in the literature are often inconsistent with each other. Horner et al. isolated, in a poor yield, 1,1,2,2-tetraphenyldiphosphane 1,2-dioxide (4a) from the reaction mixture of diphenylphosphinic chloride (6a) and Li/Hg [89]. On the other hand, in the reaction of compound 6a with sodium in toluene, the formation of sodium diphenylphosphinite (8b') or diphenylphosphido anion (19b') was observed, depending on the amount of sodium used in this reaction [90]. Inamoto and co-workers conducted several experiments where electrophile 6a was treated by alkali metals and metal salts in THF [91]. Sasaki et al. described the reduction of electrophile 6a with SmI2, at room temperature in THF. They smoothly converted molecule 6a into the corresponding compound 8b. As a by-product, molecule 4a was isolated with 7% yield (Scheme 33) [92]. The presented evidence indicates that the cleavage of P-Cl bonds of molecules 6 with electron donors may occur either by one-electron or two-electron pathways. The reductive cleavage of P-C, as well as P-O bonds, may occur. Rachoń's team observed the treatment of compounds 6 in THF with one equivalent of potassium metal (with or without the catalysts: naphthalene or 4,4′-ditertbutylbiphenyl) as well as potassium naphthalenide resulted in a complex mixture of products [42,93]. From the theoretical point of view, the electrophile 5 and 6 can accept an electron to form an anion radical, which should collapse into a phosphonyl radical and a chlorine anion. The phosphonyl radical, similar to the phosphorus radicals presented in Scheme 27, as a very reactive species, may dimerize to form P-P or P-O-P bond and subsequently accept an electron to form a >P-Oanion followed by further fragmentation, according to the substituents on the phosphorus and the reaction conditions. At this point, it is worth adding that the reagents of the electrophilic, nucleophilic, and radical characters can exist in such a reaction mixture. Tordo's team also demonstrated that diarylphosphonyl radical can be reduced into diarylphosphinite anion by the electron-rich olefin [65].
The second method of synthesizing compounds 4 is the reaction of electrophiles 6 with appropriate nucleophiles 8, i.e., diarylphosphine oxides or dialkylphosphine oxides. Scheme 32. Synthesis of molecules 4b,c via metal complexes 9c,d decomposition.

Some Synthetic Approaches of Diphosphine Dioxides
The diphosphine dioxides 4 can be obtained through the same synthetic routes as previously described in hypodiphosphoric acid esters 3 (Scheme 3). The first method is a reaction between electrophiles 6, with electron donors such as the alkali metals in heterogeneous (Li, Na, K) or homogenous conditions (potassium naphthalenide, Li, Na, K in NH 3liq .). The first application of the Wurtz-Fittig type synthesis for the P-P bond formation was the coupling of PI 3 with mercury to diphosphorus tetraiodide, reported by Besson at the end of the XIX century [80]. Other metals, such as lithium [81], sodium [82][83][84][85][86], potassium [85,86], and magnesium [87,88], in organic solvents have been successfully used for the coupling of the less reactive but more accessible phosphinous halides.
The reaction of compounds 6 with alkali metals is complicated, and the data presented in the literature are often inconsistent with each other. Horner et al. isolated, in a poor yield, 1,1,2,2-tetraphenyldiphosphane 1,2-dioxide (4a) from the reaction mixture of diphenylphosphinic chloride (6a) and Li/Hg [89]. On the other hand, in the reaction of compound 6a with sodium in toluene, the formation of sodium diphenylphosphinite (8b') or diphenylphosphido anion (19b') was observed, depending on the amount of sodium used in this reaction [90]. Inamoto and co-workers conducted several experiments where electrophile 6a was treated by alkali metals and metal salts in THF [91]. Sasaki et al. described the reduction of electrophile 6a with SmI 2 , at room temperature in THF. They smoothly converted molecule 6a into the corresponding compound 8b. As a by-product, molecule 4a was isolated with 7% yield (Scheme 33) [92].

Some Synthetic Approaches of Diphosphine Dioxides
The diphosphine dioxides 4 can be obtained through the same synthetic routes as previously described in hypodiphosphoric acid esters 3 (Scheme 3). The first method is a reaction between electrophiles 6, with electron donors such as the alkali metals in heterogeneous (Li, Na, K) or homogenous conditions (potassium naphthalenide, Li, Na, K in NH3liq.). The first application of the Wurtz-Fittig type synthesis for the P-P bond formation was the coupling of PI3 with mercury to diphosphorus tetraiodide, reported by Besson at the end of the XIX century [80]. Other metals, such as lithium [81], sodium [82][83][84][85][86], potassium [85,86], and magnesium [87,88], in organic solvents have been successfully used for the coupling of the less reactive but more accessible phosphinous halides.
The reaction of compounds 6 with alkali metals is complicated, and the data presented in the literature are often inconsistent with each other. Horner et al. isolated, in a poor yield, 1,1,2,2-tetraphenyldiphosphane 1,2-dioxide (4a) from the reaction mixture of diphenylphosphinic chloride (6a) and Li/Hg [89]. On the other hand, in the reaction of compound 6a with sodium in toluene, the formation of sodium diphenylphosphinite (8b') or diphenylphosphido anion (19b') was observed, depending on the amount of sodium used in this reaction [90]. Inamoto and co-workers conducted several experiments where electrophile 6a was treated by alkali metals and metal salts in THF [91]. Sasaki et al. described the reduction of electrophile 6a with SmI2, at room temperature in THF. They smoothly converted molecule 6a into the corresponding compound 8b. As a by-product, molecule 4a was isolated with 7% yield (Scheme 33) [92]. The presented evidence indicates that the cleavage of P-Cl bonds of molecules 6 with electron donors may occur either by one-electron or two-electron pathways. The reductive cleavage of P-C, as well as P-O bonds, may occur. Rachoń's team observed the treatment of compounds 6 in THF with one equivalent of potassium metal (with or without the catalysts: naphthalene or 4,4′-ditertbutylbiphenyl) as well as potassium naphthalenide resulted in a complex mixture of products [42,93]. From the theoretical point of view, the electrophile 5 and 6 can accept an electron to form an anion radical, which should collapse into a phosphonyl radical and a chlorine anion. The phosphonyl radical, similar to the phosphorus radicals presented in Scheme 27, as a very reactive species, may dimerize to form P-P or P-O-P bond and subsequently accept an electron to form a >P-Oanion followed by further fragmentation, according to the substituents on the phosphorus and the reaction conditions. At this point, it is worth adding that the reagents of the electrophilic, nucleophilic, and radical characters can exist in such a reaction mixture. Tordo's team also demonstrated that diarylphosphonyl radical can be reduced into diarylphosphinite anion by the electron-rich olefin [65].
The second method of synthesizing compounds 4 is the reaction of electrophiles 6 with appropriate nucleophiles 8, i.e., diarylphosphine oxides or dialkylphosphine oxides. The presented evidence indicates that the cleavage of P-Cl bonds of molecules 6 with electron donors may occur either by one-electron or two-electron pathways. The reductive cleavage of P-C, as well as P-O bonds, may occur. Rachoń's team observed the treatment of compounds 6 in THF with one equivalent of potassium metal (with or without the catalysts: naphthalene or 4,4 -ditertbutylbiphenyl) as well as potassium naphthalenide resulted in a complex mixture of products [42,93]. From the theoretical point of view, the electrophile 5 and 6 can accept an electron to form an anion radical, which should collapse into a phosphonyl radical and a chlorine anion. The phosphonyl radical, similar to the phosphorus radicals presented in Scheme 27, as a very reactive species, may dimerize to form P-P or P-O-P bond and subsequently accept an electron to form a >P-Oanion followed by further fragmentation, according to the substituents on the phosphorus and the reaction conditions. At this point, it is worth adding that the reagents of the electrophilic, nucleophilic, and radical characters can exist in such a reaction mixture. Tordo's team also demonstrated that diarylphosphonyl radical can be reduced into diarylphosphinite anion by the electron-rich olefin [65].
The second method of synthesizing compounds 4 is the reaction of electrophiles 6 with appropriate nucleophiles 8, i.e., diarylphosphine oxides or dialkylphosphine oxides. This chemical transformation is similar to the reaction described above for hypodiphosphoric acid esters 3, but with a stronger base, e.g., NaH, alkyllithium, aryllithium, etc. (Scheme 3; method B).
Hunter et al. described the reaction between chlorodiphenylphosphane (20g) and the nucleophile 8b in the presence and absence of base [94]. The electrophile 20g reacted with nucleophile 8b in the presence of base, giving 1,1,2,2-tetraphenyldiphosphane 1-oxide (18d). Based on the 31 P NMR spectroscopy, the authors proved that these reagents reacted to each other in the absence of base, leading to the initial formation of the molecule 18d, which had a characteristic double-doublet with chemical coupling 1 J PP = 224 Hz. The compound 18d underwent further transformations with electrophile 20g to form tetraphenyldiphosphine (23a) and molecule 6a. The electrophile 6a could be further deoxygenated by compound 8b to electrophile 20g, and the molecule 8b was oxygenated to diphenylphosphinic acid (16d). The origin of compounds 16d and 23a was explained to be the side reaction products between 18d and 20g (Scheme 34). These results are consistent with the earlier work of Stec et al. This chemical transformation is similar to the reaction described above for hypodiphosphoric acid esters 3, but with a stronger base, e.g., NaH, alkyllithium, aryllithium, etc. (Scheme 3; method B). Hunter et al. described the reaction between chlorodiphenylphosphane (20g) and the nucleophile 8b in the presence and absence of base [94]. The electrophile 20g reacted with nucleophile 8b in the presence of base, giving 1,1,2,2-tetraphenyldiphosphane 1-oxide (18d). Based on the 31 P NMR spectroscopy, the authors proved that these reagents reacted to each other in the absence of base, leading to the initial formation of the molecule 18d, which had a characteristic double-doublet with chemical coupling 1 JPP = 224 Hz. The compound 18d underwent further transformations with electrophile 20g to form tetraphenyldiphosphine (23a) and molecule 6a. The electrophile 6a could be further deoxygenated by compound 8b to electrophile 20g, and the molecule 8b was oxygenated to diphenylphosphinic acid (16d). The origin of compounds 16d and 23a was explained to be the side reaction products between 18d and 20g (Scheme 34). These results are consistent with the earlier work of Stec et al. Majewski et al. reported similar observations that diethylphosphine oxide (8c) reacted with in situ produced diethylphosphinic chloride (6b) in the reaction of molecule 8c with CCl4 [95]. Among final products, diethylphosphinic anhydride (11a) was observed, but 1,1,2,2-tetraethyldiphosphane 1,2-dioxide (4f) was not identified. These results are consistent with the earlier work of Steinberg (Scheme 10) [45].
Quin et al. presented the efficient synthesis of symmetrical tetraaryldiphosphine dioxides 4 [96]. The appropriate chlorodiarylphosphanes 20 in the presence of tertiary amines and exposed to oxygen and water produced the target compounds 4 (Scheme 35). However, when the air was excluded, no compounds 4 were observed. This suggests that the direct reaction product was diphosphine monoxides 18, which gave target compounds Scheme 35. Synthesis of molecules 4. The possible reaction routes of the reaction between the nucleophile 8b and electrophile 20g (no yield of 13c was given).
Quin et al. were the first to recognize the duality in secondary phosphine oxides 8′s chemical character because only the existence of O-phosphorylation could explain the obtained anhydrides 11b and 13c [95]. Ambident phosphorus nucleophiles 8′ of the type >P-Ocould react through either the P or O nucleophilic centre with phosphorus electrophiles 6 or 20 to produce both P-phosphorylated products 4 or O-phosphorylated compounds 11 mixed with anhydrides 13. The pattern of the yield distribution depends on the substituent located on the phosphorus atom, which was controlled by steric and/or electronic factors.
Many results reported above showed mysterious "oxidized" and "reduced" anhydrides 11 and 13, respectively, as the products of the syntheses of molecules 4. To explain this puzzle, Rachoń's team performed the reaction of mixed anhydrides 13 with the metal salts of nucleophile 8′, having the phosphorus atom surrounded by identical substituents. They performed reactions between the mixed anhydrides 13a,b, and 5,5-dimethyl-1,3,2dioxaphosphinan-2-yl tert-butyl(phenyl)phosphinate (13d) and the potassium salts of nucleophiles 7b′ and 8b′, which possessed the identical substituents located on the trivalent phosphorus atom in the constitution of mixed anhydrides 13 (Scheme 36) [68]. It is important to mention that Stec et al. found that the mixed anhydride 12b did not react with either nucleophile 7a in the presence of triethylamine at room temperature or electrophile 5a. The metal salts of nucleophiles 7 and 8 are stronger reagents than those generated in the reaction between molecules 7 and 8 and triethylamine. Quin et al. were the first to recognize the duality in secondary phosphine oxides 8 s chemical character because only the existence of O-phosphorylation could explain the obtained anhydrides 11b and 13c [95]. Ambident phosphorus nucleophiles 8 of the type >P-Ocould react through either the P or O nucleophilic centre with phosphorus electrophiles 6 or 20 to produce both P-phosphorylated products 4 or O-phosphorylated compounds 11 mixed with anhydrides 13. The pattern of the yield distribution depends on the substituent located on the phosphorus atom, which was controlled by steric and/or electronic factors.
Many results reported above showed mysterious "oxidized" and "reduced" anhydrides 11 and 13, respectively, as the products of the syntheses of molecules 4. To explain this puzzle, Rachoń's team performed the reaction of mixed anhydrides 13 with the metal salts of nucleophile 8 , having the phosphorus atom surrounded by identical substituents. They performed reactions between the mixed anhydrides 13a,b, and 5,5-dimethyl-1,3,2dioxaphosphinan-2-yl tert-butyl(phenyl)phosphinate (13d) and the potassium salts of nucleophiles 7b and 8b , which possessed the identical substituents located on the trivalent phosphorus atom in the constitution of mixed anhydrides 13 (Scheme 36) [68]. It is important to mention that Stec et al. found that the mixed anhydride 12b did not react with either nucleophile 7a in the presence of triethylamine at room temperature or electrophile 5a. The metal salts of nucleophiles 7 and 8 are stronger reagents than those generated in the reaction between molecules 7 and 8 and triethylamine.
The reactions between the metal salts of nucleophiles 7 and 8 with mixed anhydrides 13 were very selective. The trivalent phosphorus atoms were always attacked by nucleophiles 7b or 8c , giving Por O-phosphorylated products 18e,f, and 14h, respectively, and an acid anion 16c' as a leaving group (Scheme 36). Depending on the nature of substituents, the diphosphine monoxides 18e,f, or the anhydride 14h were obtained. Hydrolysis, oxygenation, or sulfuration gave the final products in similar molar yields in both cases. The trivalent phosphorus atom of diphosphine monoxides 18e,f were oxidized or sulfurated to produced molecules 4b,c or 4k,l with good yields. In the next step, acid anions 16c can be reacted easily with phosphorus electrophiles 6 to produce phosphorus acid anhydrides 11 as "oxidized" products. These results are compatible and consistent with the earlier work of Zwierzak et al. (Schemes 4-6) [40]. The reactions between the metal salts of nucleophiles 7 and 8 with mixed anhydrides 13 were very selective. The trivalent phosphorus atoms were always attacked by nucleophiles 7b′ or 8c′, giving P-or O-phosphorylated products 18e,f, and 14h, respectively, and an acid anion 16c' as a leaving group (Scheme 36). Depending on the nature of substituents, the diphosphine monoxides 18e,f, or the anhydride 14h were obtained. Hydrolysis, oxygenation, or sulfuration gave the final products in similar molar yields in both cases. The trivalent phosphorus atom of diphosphine monoxides 18e,f were oxidized or sulfurated to produced molecules 4b,c or 4k,l with good yields. In the next step, acid anions 16c′ can be reacted easily with phosphorus electrophiles 6 to produce phosphorus acid anhydrides 11 as "oxidized" products. These results are compatible and consistent with the earlier work of Zwierzak et al. (Schemes 4-6) [40].
In the aforementioned reaction between nucleophile 8′ and electrophiles 6, two nucleophiles coexisted in the reaction mixture: 8′ and 16′. The >P-O − anion, nucleophile 8′ as an α-nucleophile should react faster with electrophiles 6 than the acid anion 16′. Thus, the increase of nucleophile 6′ concentration should increase the amount of molecule 4. These results are in line with the earlier work of Michalski et al. (25% excess of nucleophilic reagent 7a′) [39] and (50% excess sodium salts of diethyl phosphonate 7a′) [49]. According to Pearson's hard and soft acid and base principle, leaving the group could affect the ratio of P-vs. O-phosphorylated product [97]. Thus, nucleophilic substitution proceeds faster when the nucleophile and the leaving group are either hard or soft. Consequently, obtaining higher yields of compounds 4 requires modifying the softness of the phosphorus atom substituents. This can be achieved by changing the halogen atom bound to the phosphorus in electrophiles 6, from Cl through Br to I. A reaction of nucleophiles 7 or 8 with iodine seems to be an exciting idea, which guarantees a temporary excess of the nucleophile. Moreover, iodine, as opposite to bromine, did not quantitatively react with compounds 3.
Quite surprisingly, the reaction of nucleophiles 7′ or 8′ and iodine in THT or THF/NH3liq. yielded compounds 3 and 4 efficiently, which obtained diphosphine dioxides 4 in 98% and hypophosphoric acid esters 3 in 48% yields in a facile and convenient way. The reaction of iodine and nucleophiles 7′ or 8′ anion generated in situ proceeded efficiently in both THF and liquid ammonia. Moreover, it appeared that the dissolution rate of solid iodine limited the reactivity of iodine in liquid ammonia at −78 °C. In the reaction mixture tert-butyl(phenyl)phosphinic iodide 6b was isolated with 1% yield. The results, especially  [39] and (50% excess sodium salts of diethyl phosphonate 7a ) [49]. According to Pearson's hard and soft acid and base principle, leaving the group could affect the ratio of Pvs. O-phosphorylated product [97]. Thus, nucleophilic substitution proceeds faster when the nucleophile and the leaving group are either hard or soft. Consequently, obtaining higher yields of compounds 4 requires modifying the softness of the phosphorus atom substituents. This can be achieved by changing the halogen atom bound to the phosphorus in electrophiles 6, from Cl through Br to I. A reaction of nucleophiles 7 or 8 with iodine seems to be an exciting idea, which guarantees a temporary excess of the nucleophile. Moreover, iodine, as opposite to bromine, did not quantitatively react with compounds 3. Quite surprisingly, the reaction of nucleophiles 7 or 8 and iodine in THT or THF/NH 3liq . yielded compounds 3 and 4 efficiently, which obtained diphosphine dioxides 4 in 98% and hypophosphoric acid esters 3 in 48% yields in a facile and convenient way. The reaction of iodine and nucleophiles 7 or 8 anion generated in situ proceeded efficiently in both THF and liquid ammonia. Moreover, it appeared that the dissolution rate of solid iodine limited the reactivity of iodine in liquid ammonia at −78 • C. In the reaction mixture tertbutyl(phenyl)phosphinic iodide 6b was isolated with 1% yield. The results, especially from liquid ammonia, suggest that the reaction proceeded via radical mechanism rather than pure nucleophilic substitution. The alkoxy group as a phosphorus substituent reduced the efficiency of compounds 3 because the iodide anion could react with alkoxy substituents (see Arbuzov's reaction).
The synthesis of unsymmetrical compounds 3 or 4, in which one fragment would come from nucleophilic molecule 7 or 8 and the second from electrophilic 5 or 6, was much more complicated. Both reagents can undergo X-philic substitution [63], which leads to the exchange of characters of reagents, and in effect, symmetrical and unsymmetrical compounds 4 (Scheme 37). duced the efficiency of compounds 3 because the iodide anion could react with alkoxy substituents (see Arbuzov's reaction).
The synthesis of unsymmetrical compounds 3 or 4, in which one fragment would come from nucleophilic molecule 7 or 8 and the second from electrophilic 5 or 6, was much more complicated. Both reagents can undergo X-philic substitution [63], which leads to the exchange of characters of reagents, and in effect, symmetrical and unsymmetrical compounds 4 (Scheme 37). Scheme 37. The synthesis of unsymmetrical compounds 4. The X-philic substitution reaction between the nucleophile 8 and electrophile 6.
In 1983, Foss et al. carried out a thermal inducted rearrangement of unsymmetrical molecule 18g [98] and determined the equilibrium between compounds 18 and 15. Some interesting rearrangement and symmetrization similar to the above-described reactivity were observed (Scheme 38). It was suggested that monoxides 18 preferred the alkyl substituents, but anhydrides 15 alkoxyl. Scheme 38. The thermal inducted rearrangement of diphosphine monoxide 18g.

The Analogs of Biologically Active Compounds That Possess a P-P bond
The synthesis of molecules 3 depended on the type of substituents on the phosphorus atom, as is shown above. Elaborate procedures helped to obtain these interesting molecules. However, the syntheses of analogs of biologically active compounds that possess a P-P bond are still challenging.
than pure nucleophilic substitution. The alkoxy group as a phosphorus substituen duced the efficiency of compounds 3 because the iodide anion could react with alk substituents (see Arbuzov's reaction).
The synthesis of unsymmetrical compounds 3 or 4, in which one fragment w come from nucleophilic molecule 7 or 8 and the second from electrophilic 5 or 6, was m more complicated. Both reagents can undergo X-philic substitution [63], which lead the exchange of characters of reagents, and in effect, symmetrical and unsymmetrical c pounds 4 (Scheme 37). In 1983, Foss et al. carried out a thermal inducted rearrangement of unsymmet molecule 18g [98] and determined the equilibrium between compounds 18 and 15. S interesting rearrangement and symmetrization similar to the above-described react were observed (Scheme 38). It was suggested that monoxides 18 preferred the alkyl stituents, but anhydrides 15 alkoxyl.

The Analogs of Biologically Active Compounds That Possess a P-P bond
The synthesis of molecules 3 depended on the type of substituents on the phosph atom, as is shown above. Elaborate procedures helped to obtain these interesting m cules. However, the syntheses of analogs of biologically active compounds that posse P-P bond are still challenging.

The Analogs of Biologically Active Compounds That Possess a P-P bond
The synthesis of molecules 3 depended on the type of substituents on the phosphorus atom, as is shown above. Elaborate procedures helped to obtain these interesting molecules. However, the syntheses of analogs of biologically active compounds that possess a P-P bond are still challenging.
Dirheimer succeeded to obtained hypodiphosphorylated derivatives of nucleotides The antitumor activity of hypodiphosphoric acid 3 is potentially crucial in cancer treatment because the synthesis of deoxythymidine monophosphate (dTMP) is critical during rapid cell proliferation [99]. In general, most normal mammalian cells grow slowly and require less dTMP, hence the interruption of its synthesis can selectively kill cancer cells.
Dirheimer succeeded to obtained hypodiphosphorylated derivatives of nucleotides from hypodiphosphoric acid salts (1). The limited characterization data about the structure of hypodiphosphoric acid (1) hindered the clarity of the results, mainly because of the unavailability of 31 P NMR data [100,101].
Recently, Sepulveda-Boza published a synthetic procedure about the modified nucleotide of cytidine containing three directly connected phosphorus atoms, i.e., the PA-PB-PC system [8]. The absence of coupling constants between phosphorus atoms in the presented data left room for further clarification about these results. Coupling constants should be distinguished between each connected phosphorus atom.
The antitumor activity of hypodiphosphoric acid 3 is potentially crucial in cancer treatment because the synthesis of deoxythymidine monophosphate (dTMP) is critical during rapid cell proliferation [99]. In general, most normal mammalian cells grow slowly and require less dTMP, hence the interruption of its synthesis can selectively kill cancer cells.
Dirheimer succeeded to obtained hypodiphosphorylated derivatives of nucleotides from hypodiphosphoric acid salts (1). The limited characterization data about the structure of hypodiphosphoric acid (1) hindered the clarity of the results, mainly because of the unavailability of 31 P NMR data [100,101].
Recently, Sepulveda-Boza published a synthetic procedure about the modified nucleotide of cytidine containing three directly connected phosphorus atoms, i.e., the PA-PB-PC system [8]. The absence of coupling constants between phosphorus atoms in the presented data left room for further clarification about these results. Coupling constants should be distinguished between each connected phosphorus atom.
The antitumor activity of hypodiphosphoric acid 3 is potentially crucial in cancer treatment because the synthesis of deoxythymidine monophosphate (dTMP) is critical during rapid cell proliferation [99]. In general, most normal mammalian cells grow slowly and require less dTMP, hence the interruption of its synthesis can selectively kill cancer cells.
Dirheimer succeeded to obtained hypodiphosphorylated derivatives of nucleotides from hypodiphosphoric acid salts (1). The limited characterization data about the structure of hypodiphosphoric acid (1) hindered the clarity of the results, mainly because of the unavailability of 31 P NMR data [100,101].
Recently, Sepulveda-Boza published a synthetic procedure about the modified nucleotide of cytidine containing three directly connected phosphorus atoms, i.e., the PA-PB-PC system [8]. The absence of coupling constants between phosphorus atoms in the presented data left room for further clarification about these results. Coupling constants should be distinguished between each connected phosphorus atom.

Conclusions
In this review, the syntheses of the molecules 3 and 4 containing >P(=O)-P(=O)< fragment are described. Despite the lack of a well-developed mechanism, the well-known Salzer's synthetic procedure of hypodiphosphoric acid (1) is constantly used. The acid 1 is unstable under acidic conditions and disproportionate to H 3 PO 4 and H 3 PO 3 . The P-P distance is close to 2.182 Å, and the P-O bond lengths range from 1.501 to 1.569 Å. The 31 P{ 1 H} CPMAS NMR measurements showed that the hypodiphosphates 1 exhibit a single resonance in the range from 14.9 ppm to 11.4 ppm. The acid 1 and its salts show ferroelectricity properties.
The structure of related molecules, the hypodiphosphoric acid esters 3, and the diphosphine dioxides 4 are stabilized by alkyl, aryl, and cyclic substituents which are directly connected to the phosphorus atom. The use of an excess of nucleophiles during the synthesis of molecules 3 and 4 is also beneficial. The lithium cation could better stimulate the formation of P-P bonds than other alkali metals, resulting in higher yields of compounds 3 and 4, which requires modifying the softness of the phosphorus atom substituents. This can be achieved by changing the substituents, including the halogen atom bound to the phosphorus in the constitution of electrophiles 5 and 6, from Cl through Br to I.
The oxathiaphospholane method of P-P bond formation during the synthesis of biologically active analogs is an interesting idea that allows the syntheses of further compounds with potential biological activities. However, some authors have concluded that natural proteins did not recognize compounds with P-P bonds because of their non-existing living systems.
The P-centered phosphorus radicals could dimerize to produce compounds 3 and 4, which are considered precursors of >P(=O) • type radicals. The studies of reactions involving organophosphorus radicals have great potential in organophosphorus and medicinal chemistry.
Funding: This research received no external funding.

Conflicts of Interest:
The author declare no conflict of interest.