This work has been digitalized and published in 2013 by Verlag Zeitschrift für Naturforschung in cooperation with the Max Planck Society for the Advancement of Science under a Creative Commons Attribution4.0 International License.

Dieses Werk wurde im Jahr 2013 vom Verlag Zeitschrift für Naturforschungin Zusammenarbeit mit der Max-Planck-Gesellschaft zur Förderung derWissenschaften e.V. digitalisiert und unter folgender Lizenz veröffentlicht:Creative Commons Namensnennung 4.0 Lizenz.

620 J . MIKOLA J CZ Y K E T A L . • CYCLIC ORGANOPHOSPHORUS COMPOUNDS

Cyclic Organophosphorus Compounds, V*

Cyclic Esters of Phosphorous — Phosphoric Anhydride as a Tool for Studying the “Abnormal”

Biphilic Reactivity of PH,-P V Anhydride System

J. M i k o l a j c z y k , J. M i c h a l s k i , and A. Z w i e r z a k

In stitu te of Organic Chemistry, Technical U niversity Lodz, Poland

(Z. Naturforsch. 28 b, 620-624 [1973]; received November 16, 1972/March 20, 1973)

Phosphorus, H udson’s concept, cyclic esters, anhydrides

A series of esters of phosphorous-phosphoric anhydride containing cyclic ester groups a t one or both phosphorus atoms were prepared. The typical biphilic character of these compounds has been found. Unexpectedly observed lack of reactiv ity of anhydrides con­taining cyclic ester groups a t trivalent phosphorus atom w ith respect to diethylphosphoric and acetic acids is the starting point for discussion on the mechanism of acidolysis. U nu­sual inertness of p m cyclic esters of phosphorous-phosphoric anhydride towards protic acids is discussed in term s of H u d s o n ’s concept of steric reta rda tion on quaternisation.

Although tetraalkyl esters of phosphorous- phosphoric anhydride (1 ) have been recognized as fairly good phosphorylating agents for acids1,2, the mechanism of acidolysis of an anhydride bond is so far contradictory and controversial2'5.

RO OR’\ /

P -O -P R = R ’; R * R ’./» \

RO 0 OR’1

The pronounced biphilic reactivity of acyclic Pm - P ' anhydride systems renders any detailed studies on the mechanism of phosphorylation practically impossible.

The present work is an attempt to use less reactive cyclic esters of phosphorous-phosphoric

Requests for reprin ts should be sent to Prof. Dr. A n d r z e j Z w i e r z a k , In stitu te of Organic Chemistry, Technical U niversity (Politechnika), Lodz, zwirki 36, Poland.

* P a rt I V : A. Z w i e r z a k , Tetrahedron [London] 2 5 ,

5177 [1969]. Prelim inary com munication: J . M i k o ­

l a j c z y k , J . M i c h a l s k i , and A. Z w i e r z a k , Chem. Comraun. 1 9 7 1 , 1257. Paper CLX X V II on Organo­phosphorus Compounds.

anhydride as more convenient models for studying the biphilic reactivity of P m -P Y anhydride systems.

ResultsCyclic esters of phosphorous-phosphoric an­

hydride ( la -e ) containing both tri- and tetracova- lent phosphorus atoms incorporated in five- or six- membered rings were obtained by the action of suitable chlorophosphites on pyridinium salts of diethyl hydrogen phosphate (compounds: la -c ) or 2-hydroxy-2-oxo-5,5-dimethyl-l,3,2-dioxaphospho- rinan (compounds: Id and le ).

0---1 xo —>. /0 —s.<Et0)2l?-0-p( (Et0)2P-0-Px > (Et0)2P-0-Px X

0 0—1 0 0—' 0 o—/

1a 1b 1c

>Os-°-p(°E'” >CrO1d le

According to expectations the mixed anhydrides ( 1 a-c) were found to be active electrophilic reagents. They reacted with typical nucleophiles, like water, alcohols and amines, readily and in one way only, the attack of the nucleophilic reagent being directed exclusively to the trivalent phosphorus atom.

J . M IKOLAJCZYK E T A L . CYCLIC ORGANOPHOSPHORUS COMPOUNDS 621

(Et0)2 FJ-0-Pv

0 0-(CH2)n ♦ N! — ► (Et0)2P-0“+ N- P (CH2)n

0 ° — '(«Ta-c

N = H 2 0 , R-O H , R -N H 2 2,3.

Typical reactions of anhydrides (la , b and d) with nucleophiles were discussed in our earlier publication6.

Contrary to the expectations it was found that anhydrides 1 a-c do not react with diethyl hydrogen phosphate as well as acetic-acid even when mixtures of the substrates are heated for several hours at 1 0 0 °C. Unchanged starting materials could be re­covered almost quantitatively from such mixtures. The observed resistance of anhydrides (la-c) con­taining trivalent phosphorus atom in a cyclic dioxaphospholan or diaxaphosphorinan system to the action of acids indicates that their nucleo- philicity is lower than that of analogous acyclic compounds 1 .

The behaviour of anhydrides 1 d and 1 e (contain­ing tetracovalent phosphorus or both phosphorus atoms in the cyclic dioxaphosphorinan system) in the reactions with diethyl hydrogen phosphate and acetic acid was different. In the case of diethyl hydrogen phosphate the reaction led to the equili­brium and 2-hydroxy-2-oxo-5,5-dimethyl-1,3,2-di- oxaphosphorinan (2 ) and the corresponding new mixed anhydride 3 could be isolated from the reac­tion mixtures in addition to the unchanged sub­strates.

o benzeneA + (Et0)2 P-0H ...........^

Ö " « room temp. O r OH + (EtO)2 P-OA

Ö (2)

A = P(0Et) 2 ( Id) ; y ie ld -65%

0 —VA = p( y (1e) ; yield - 28 %

0—

2-Acetoxy-2-oxo-5,5-dimethyl-1,3,2-dioxaphos- phorinan (4) and diethyl phosphite (5), both isolated in 70% yield, were formed in the reaction between the anhydride Id and acetic acid.

x T V o -' \ _rs II— 0 oP(0Et) 2 ♦ CH3-C-OH — )p-0-C-CH3+ (EtO)2P-H

Ö 0 0 0 0Id

Only 2-hydroxy-2-oxo-5,5-dimethyl-1,3,2-dioxa- phosphorinan (2 ) could be, however, isolated in 27%

yield from the mixture of products formed in the reaction of anhydride 1 e with acetic acid.

All the above described reactions of anhydrides Id and l e were taking place very readily at room temperature, which was not previously observed in the case of the acyclic analogues 1 .

The products were unambiguously characterised by comparison with authentic samples of standard preparations. The structure of 2-acetoxy-2-oxo-5,5- dimethyl-l,3,2-dioxaphosphorinan (4) inferred from the spectral data was confirmed by aminolysis with aniline, affording anilinium salt of 2 -hydroxy-2 - oxo-5,5-dimethyl-1,3,2-dioxaphosphorinan (6 ) and acetanilide (7).

X 7 -0 -c -iCH3 + PhNH2 x>°-N— 0 0

PhNH3 +CH3-C-NHPh 3 3 11

° (4)

Discussion

The observation7 that five-membered cyclic phosphite esters are more electrophilic than their acyclic analogues was recently supplemented by G r e e n h a l g h and H u d s o n 8 who found that they are also considerably less prone to quaternization and offered a plausible explanation of this pheno­menon.

The application of the “steric retardation” hypothesis to the reactions of cyclic phosphites with electrophilic reagents leads to definite mechanistic implications regarding the proved lack of reactivity of anhydrides (la-c) in acidolysis reactions. Since three different nucleophilic centres are present in the molecules of cyclic anhydrides la -c at least four different acidolysis paths (A-D) could be considered:

A. Protonation of tervalent phosphorus atom and subsequent attack of the anion A 0 on the tetra­covalent phosphorus atom :

XX/

(3)

(Et0)2 P-0-PT (CH2)n + H - A ^ AÜ

5 “ x o -----------J

1a-c

------1 '{Et0)2 P-0-P.r (C H2)n

0 H 0---(6 )

--- ► (Et0)2P-A + 0=P^ (CH2)n

Ö 0---- 1n = 2,3; A = (EtO)2P (O )O 0; CH3 C(O)O0.

622 J . MIKOLA J CZ YK E T A L . ■ CYCLIC O RGAXOPHOSPHORUS COMPOUNDS

B. Protonation of the anhydride-bridge oxygen atom followed by the attack of anion A 0 on the tetracovalent phosphorus atom :

(Et0)2P-0-P. ICH2)n + H-A=f^=A;

o ” 0----)la-c

O r!.

+ H -H+(EtO)2P-A + HO-P. (C H 2 )n ± = ^H 0- P ' (CH2)n

0 0---- > +X0---->

Y/0-----10=P (CH2)n

0---- 1

C. Protonation of the anhydride-bridge oxygen atom followed by the attack of anion A 0 on the tervalent phosphorus atom:

(Et0)2P-0-R^ (CH2)n + H - A ^ A I

o xx 0---- >la-c

f---- ° \=s=±: <CH2>n /P - A +HO-P(OEt)2

‘ (-----° \ V ' +(CH2)n / P1 0-P(0Et)2

- 1---- 0 x< o(8 )

is due to steric factors. Paths A and B of acidolysis of cyclic anhydrides (la -c) are unfavourable from the thermodynamical point of view, and consequently these compounds do not react with diethyl hydro­gen phosphate and acetic acid.

The sensitivity of anhydrides 1 d and 1 e to acids becomes readily understandable when it is assumed that acidolysis can take place according to paths C or D. In all cases the position of equilibrium in these reactions is favourable as a result of separation of 2-hydroxy-2-oxo-5,5-dimethyl-1,3,2-dioxaphospho- rinan (2 ) from the reaction mixture in the form of a crystalline, practically insoluble precipitate. The formation of new mixed anhydrides 3 in the reac­tions of 1 d and 1 e with diethyl hydrogen phosphate additionally confirmes the possibility of a direct attack of acid anion A 0 on tervalent phosphorus atom.

The formation of acetylphosphate 4 and diethyl phosphite 5 in the reaction of cyclic anhydride Id with acetic acid is a result of the consecutive reaction between the primarily formed (according to Scheme C or D) diethyl acetylphosphite 9 and 2-hydroxy-2-oxo-5,5-dimethyl-1,3,2-dioxaphospho- rinan 2 .

D. Protonation of the phosphoryl oxygen atom followed by the attack of anion A © on the tervalent phosphorus atom:

(Et0)2P-0-P (CH2)n + h-a ^ a ;

o ’" x 0---- 1

' I—<C.H2>n / LPX1 P ( O E V 2

OH

1a-cr------o .

(9)

(CH2)n /P-A + 0=P(0Et)2__r\'OH

__ ;p -o -p (o E t)2 + ch3c-o h =

1d

{Cr

Paths C and D of acidolysis of cyclic anhydrides la -c seem to be improbable and ineffective, since in both cases the position of the equilibrium is un­favourable. This is undoubtedly due to the fact that diethyl hydrogen phosphate is a poor leaving group. Moreover, if acidolysis could take place according to these paths we should not observe any significant differences between the reactivities of cyclic an­hydrides 1 a-C and their acyclic analogues 1 since in this case none of the reaction stages involves the quaternisation of the ring phosphorus atom.

Since it has been found that cyclic anhydrides1 a-c are not sensitive to acids, it appears that this

■>cX} CH3-C-^LP(0Et)2

0 H

0H+ CH3-C-0-P(OEt)2 J II i

0 (10

V-0-C-CH3 + 0=P(0Et)2\ __n II II I

u 0 0 H

4 5

This course of the reaction is also indicated by the fact that a crystalline precipitate of 2 separates from the solution in the initial stage of the reaction.

Conclusions

The results of the present investigation are in­sufficient for deciding which one of the proposed paths of acidolysis of phosphorous-phosphoric anhydride esters 1 is the most effective.

However, the pronounced steric retardation effect observed in the case of cyclic anhydrides 1 a-c and considered in the light of the H u d so n ’s hypo­thesis makes it possible to give preference to paths A and B limiting the remaining possibilities to

J . M IKOLAJCZYK E T A L . CYCLIC ORGANOPHOSPHORUS COMPOUNDS 623

special cases in which the evolved phosphoric acid is a particularly good leaving group.

ExperimentalSolvents and reagents were purified by conven­

tional methods. All solutions were evaporated under reduced pressure. Boiling points and melting points (taken in capillaries) are uncorrected. IR spectra were recorded for liquid films or N u j o 1 mulls using an U R -10 spectrophotometer (C. Zeiss, Jena) or Perkin-Elmer Infracord 137 spectrophotometer.

Preparation of the anhydrides 1 a, 1 b and 1 c6 as well as the compound 1 e9 wTas described previously.

P-Diethoxyphosphoryl-P’-(5,5-dimethyl-l,2,3- dioxaphosphorinanyl)oxide (lc).

To the solution of diethyl hydrogen phosphate (30.8 g, 0.2 mole) and pyridine (15.8 g, 0.2 mole) in benzene (100 ml) 2-chloro-5,5-dimethyl-1,2,3-dioxa- phosphorinan (33.7 g, 0.2 mole) was added dropwise with stirring and efficient cooling. The resulting mixture was stirred for 1 hour at room temperature. Pyridine hydrochloride was then filtered off and washed with benzene ( 1 0 0 ml).

After evaporation of solvent the residue was distilled in vacuo to give pure cyclic ester lc , b.p. 90 °C/0.01 mm Hg, nf^ - 1.4456; yield: - 45.2 g (79%).

C9 H 2 0 O6 P 2

Calcd.: C 37.8 H 7.0,Found: C 37.8 H 7.0.

IR spectrum: (Film) P = O 1270 cm-1.

Reaction of cyclic esters of phosphorous-phos­phoric anhydride 1 a-e with diethyl hydrogen phosphate. General procedure.

The mixture of the corresponding cyclic ester la - e (0 . 0 2 mole) and diethyl hydrogen phosphate (3.08 g, 0.02 mole) was protected against moisture and heated at 50 °C for 2 hours. Liquid products were separated and purified by distillation in vacuo. Solid compounds were isolated by filtration and purified by crystallization.

Cyclic esters (la -c ) did not react with diethyl hydrogen phosphate even when heated at 1 0 0 °C for several hours. Distillation in vacuo afforded quantitative amounts of recovered starting ma­terials.

Cyclic ester ld (5.72 g) reacted with diethyl hydrogen phosphate (3.1 g) in benzene (20 ml) at room temperature. After 24 hours the crystalline precipitate was filtered off and identified as 2 - hydroxy-2-oxo-5,5-dimethyl-1,3,2-dioxaphospho- rinan (2) (IR, m.p. - 172-174 °C, undepressed on admixture of an authentic specimen). Yield: - 2 . 1 6 g ( 6 6 % ) .

Distillation of the filtrate afforded tetraethyl ester of phosphorous-phosphoric anhydride (1, R = Et). Yield - 3.57 g (65%). B.p. 74 °C/0.01 mm Hg,

ri$ - 1.4245. (Lit.10: b.p. 74 °C/0.01 mm Hg, - 1.4254). With 2 m equivalents of aniline in ether (1, R = Et) gave anilinium salt of diethyl hydrogen phosphate in 69% yield. M.p. (from ether) and m. m.p. - 69-70 °C (Lit.11: m.p. 69-70 °C). The residue left in the distillation flask was found to contain the unreacted cyclic ester ld , which afforded the acid 1 2 on hydrolysis with water.

Cyclic ester l e (5.96 g) reacted with diethyl hydrogen phosphate (3.1 g) in benzene (40 ml) at 40 °C for 6 hours. 2-Hydroxy-2-oxo-5,5-dimethyl- 1,3,2-dioxaphosphorinan (2 ), which crystallized on cooling, was filtered off and washed with benzene ( 2 0 ml).

Yield - 0.93 g (28%), m.p. and m .m .p. - 170- 172 °C (Lit.12: m.p. 174-176 °C). After removal of solvent the residue was distilled in vacuo to give the fraction boiled at 92 °C/0.01 mm Hg, n $ - 1.4450, yield - 1.6 g (28%), identified (IR) as cyclic ester lc .

The residue left in the distillation flask contained the unreacted ester le . It afforded quantitatively the acid 2 when hydrolyzed with water.

Reactions of cyclic esters of phosphorous-phos­phoric anhydride 1 a-e with acetic acid. General procedure.

The mixture of the corresponding cyclic ester la - e (0 . 0 2 mole) and acetic acid (1 . 2 g, 0 . 0 2 mole) was protected against moisture and heated for2 hours at 50 °C. Liquid products were separated and purified by distillation in vacuo. Solid com­pounds were isolated by filtration and purified by crystallization.

Cyclic esters (la -c) did not react when heated with acetic acid at 1 0 0 °C for several hours. In all cases only unchanged starting materials could be recovered quantitatively from the reaction mixture by distillation in vacuo.

Cyclic ester (ld) (5.72 g) reacted with acetic acid (1.2 g) at room temperature. When equimolar amounts of starting materials were mixed together in benzene (15 ml) a white, crystalline precipitate was formed during the initial stage of the reaction. A sample of this precipitate was taken off from the reaction mixture and identified (IR, m.p. and m. m.p. - 171-173 °C) as 2-hydroxy-2-oxo-5,5-dime- thyl-l,3,2-dioxaphosphorinan (2). The reaction mixture was maintained at room temperature for 48 hours Crystalline precipitate filtered off and washed with benzene ( 2 0 ml) was identified as 2-acetoxy-2-oxo-5,5-dimethyl-l,3,2-dioxaphospho- rinan (if, m.p. 116-118 °C, yield - 2.91 g (70%).

C7 H 1 3 0 5PCalcd.: C 40.4 H 6.3 P 14.9,Found: C 39.9 H 6.2 P 15.5.

IR spectrum: (N ujol) C = 0 1760 cm-1, P = 0 1305 cm-1. The anhydride (1) reacted with 2 M equivalents of aniline in benzene (50 ml) at room temperature to give acetanilide (7) in 70% yield;

624 J . M IKOLAJCZYK E T A L . • CYCLIC ORGANOPHOSPHORUS COMPOUNDS

m.p. and m .m .p. - 113-114 °C, and anilinium salt of 2-hydroxy-2-oxo-5,5-dimethyl-1,3,2-dioxaphos- phorinan (6) in 98% yield; m.p. and m .m.p. - 184- 186 °C. The filtrate obtained after removal of the anhydride 4 was distilled in vacuo to give the frac­tion boiled at 48-50 °C/3 mm Hg, nff - 1.4058, yield - 1.93 g (70%), identified (IR, TLC) as diethyl phosphite. The residue, which was found to contain the unreacted cyclic ester Id, afforded the acid 2

1 J . M i c h a l s k i and T. M o d r o , Bull. Acad. Polon. Sei., Ser. sei. chim. 10, 327 [1962].

2 J . M i c h a l s k i and T. M o d r o , Chem. B o r . 95, 1629[1962].

3 L. J . M i z r a k h , U. P. E v d a k o v , and L. Yu. S a n d a - l o v a , Zh. Obshch. Khim. 35, 1871 [1965].

4 K. A. P e t r o v , E. Y e . N i f a n t ’e v , and J . J . S o p i - k o v a , Dokl. Akad. Nauk. SSSR 151, 859 [1968].

5 J . F. B r a z i e r and R. B u r g a d a , B u l l . Soc. Chim. France 1966, 2109.

6 J . M i c h a l s k i and J . M i k o l a j c z y k , Bull. Acad. Polon. Sei., Ser. sei. chim. 14, 829 [1966].

when treated with water. The reaction of cyclic ester l e (5.96 g) with acetic acid (1.2 g) was carried out in benzene (40 ml) at 40 °C for 6 hours. Crystal­line precipitate which separated on cooling was filtered off and washed with benzene (20 ml). It was identified as the acid 2, yield - 0.9 g (27%), m.p. and m .m .p. - 170-172 °C (Lit12. : m.p. 174-176 °C). No pure identifiable compounds could be, however, isolated by high-vacuum distillation of the filtrate.

7 G . A k s n e s and R. E r i k s e n , Acta chem. scand. 20, 2463 [1966].

8 R. G r e e n h a l g h and R. F . H u d s o n , Chem. Com­mun. 1968, 1300.

9 W. S t e c and A. Z w i e r z a k , Canad. J . Chem. 45, 2513 [1967].

10 J . M i c h a l s k i , T. M o d r o , and A. Z w i e k z a k , J . chem. Soc. [London] 1961, 4904.

11 A. Z w i e r z a k , Roczniki Chem. 39, 1411 [1965].12 R . L. M c C o n n e l l and J . H . W. C o o v e r , J . org.

Chemistry 24, 630 [1959].