collec tetl by filtration a i d recrj stalli~edfrom ethanol t(J yieli 10.1 g (897,) of white crystals, mp 250-251". Ainal. Calcd for Cj0H24S2.2HCl: C, 49.0: 13, 10.7. S, 11.1. F o r u l ~ i : 49.0; Fr, 10.3; s,11 -2. (3,
Biological Section Methods and Results.-All of the alkylating agelib listed in Table I and I1 and caornpounds 22-24 werc cvaluat ed a$ inhibitors of reproduction in our oolony of houseflies (Jlusca doinesticn I,.). The method used was previously reported. .It 1 wt yo csoncwitratioii i i i the feed, only those compounds li.ted in Table 111 were active. All compounds werc iiiixed dry in lhc, fccd. Discussion.---The lack of effert oii reproduction of 1 lie h o u d l y of all of t h r hroiiioacetyl derivative,. iy11t hesized agree.. with the previously reported r e d t h ' arid indicates tlie rieceshity of a more actiw alkylating function before interference with reproduction 1achieved. With the aziridinyl derivat ivw, the ~iecesiity of having an S H function on the amide i:, emphaGzed by the lack of activity of 19-21. The neressity of having a bisaziridinyl functiori i q borne out by the 1:tc~kof activity of 24. Distanre between the aziridine groups is extremely important with regard to activity
Acknowledgment.-This
\\
ork was sulq)orted by
t-. S . Public Health Servicc Grant GM-11491. IVc n-ish to thank I*.Tovar for a4biatirc with the hiologi(d .tudic>.
Insect Chemosterilants. 111. 1-Aziridinylphosphine Oxides'
Entornology Reseurch Division, I'.S . Departvierit vj' ~ l q r i c i i l i ~ i r cUcltscillc, , .11 arylantl
Insert sterilizing activity of 40 varioue1)- substitriteti tri-] tli-, a n t i m o w ( l-aziriditi)-l)pho.phiiie oxides \\-:I> correlated with the degree of substitution on the nziridiiiyl ring atid with tjhe type of iioiiaziridinyl moieties att,ached to the phosphinylidyne group. All substitutions oii the aziridinyl carbon decreased the sterilizing activity of the parent compound. Among the di- and moiioaziridinyl c m q " i i d s , the most effective nonaziridinyl substituents were alkylamino groups, less effective were alkoxy groups, and least effective were aryl groiips. 111preparing strained aziridinylphosphine oxides, 3-om-6-azahit
Tris(l-:~ziridiiiyl)phospliine oxide (tepa) was one of the first chemicals found to effectively sterilize the males of many different insect species.2 During the years immediately following the discovery, variously substituted aziridinylphosphine oxides were synthesized or obtained from outside sources arid then screened for their insect sterilizing activity. I n attempts to correlabe the structure of substituted phosphine oxides with their chemosterilant activity, two approaches were investigated. One was the introduction of substituents onto one or both carbon atonis of the aziridinyl ring, and the other was the replacement of all aziridinyl moieties with various other groups. All substituted phosphine oxides described in this paper contained a t least one aziridinyl group linked to the phosphinylidyne group through its ring nitrogen. Triaziridinylphosphine Oxides.-The effects of progressive substitution of the hydrogens in tepa by methyl (1) Presented before the Symposium on Chemosterilants of the Division of Agricultural and Food Chemistry, 150th National Meeting of tlie American Chemical Society, Atlantic City, N. J., Sept 12-17, 1965. Previous p a p e r : 1.B. Boikovec and C . IT. tVoods, J . M e d . Chem.. 8, 645 (1H63). (2) A. €3. Roiko!-ec, Residue H e r . , 6, 87 (1964).
In a recwit quantitative study,4 tepa TWS found to be 13 times as active i n sterilizing niale houre flies (JIusca dowesticn I,.) :iq its V-nwthyl homolog, metepa. Table I sunini:~rizesthe sterilizirig properties of other te1t:i an:ilogs used to treat the house fly. Tepa and nietep:t arc iiicluded for comparison. Ilerivation of the valiic3 used 111 grading the sterilizing activity of the conipounds tested is discussed in the Experimental Sect ion. Conipouncls designated by literature references only \\-ere synthesized in our laboratory according to [nil)lirhed procedures. Introducing one higher alkyl groul) or plieriyl groul) into each aziridine ring drastically reduced the stcrilizing activity (Tablc I). In this respect, the effects were comparable to those obtained by introducing two or more methyl groups into the ring8 The reduced 1)liysiologicnl activity of the substituted compounds :ilq )cared to he related to their decreased susceptibility to nucleopliilic reagents. Simple measurements of g r o u p have heen 11rcviou4yrel)orted.
~ , 3 ) .I. B. Buikovec and C'. \ \ , \\'oods, .\dvances in Chemistry Series, KO. 4 1 , .\mrrican Cliemiral Society, \Vashington, I). C., 1963, 11 47. l i iC. Ciiang and \. T3. Roikovec, J . Eco7r. Entomol., 67, 488 (111641.
IKSECT CHEJIOSTERILAXTS. I11
July 1966
TABLEI HOUSEFLYSTERILIZING ACTIVITYOF TRIAZIRIDINYLPHOSPHINE OXIDES Compd
2
Structure
Graded activitya
T+
523
I). Unfortunately, none of the compounds demon-
Source
strated significant sterilizing activity. Introducing an additional heteroatom (oxygen) into the cycloalkenimines mas also tried. The base, 3oxa-6-azabicyclo [Xl.O]hexane, which represents a new heterocyclic system, was prepared according to a reaction sequence applied previously to the synthesis of cycloalkeniniines.*
b
C
4
d
The structure of the product was confirmed by infrared and nmr spectra (Figure 1). Treatment of the base with phosphorus oxychloride gave tris(3-oxa-6e 6 azabicyclo [3.1.O]hexan-6-yl)phosphine oxide, but its sterilizing activity was again negligible. It appears then that any substitution on the aziridinyl e 7 rings of tepa gives rise to compounds that are less effective in sterilizing insects than the parent compound. e 8 Diaziridinylphosphine Oxides.-Replacement of the entire aziridinyl moiety by another group resulted in e 9 compounds of surprisingly high sterilizing activity in spite of the total number of aziridinyl groups per molea +++, high activity (comparable to or higher than that of cule being reduced. Effects of the replacement of one tepa) ; + +, moderate activity (comparable to metepa); aziridinyl ring with aryl, alkoxy, and substitutedlow but significant activity; and -, insignificant or not detectamino groups are indicated in Table 11. Several of able activity a t any concentration tested. Commercial sources. InterA. T. Bottini, University of California, Davis, Calif. the listed Compounds also bear substituents on the chemical Corp. e See Experimental Section. two aziridinyl rings. Choice of compounds was governed by their availability rather than by intenthe stability of tepa and metepa in aqueous solutions5 tion. Most of the compounds were obtained from outsupported this proposition. I n order to gain further side sources or prepared according to published proevidence, tepa analogs, prepared so that the aziridine cedures. Fortunately, the sterilizing activities of the rings contained either electron-withdrawing substitparent triaziridinyl compounds mere known and thus uents or cyclic groups, were tested. (The cyclic the relative activities of the derivatives are of signifigroups could be expected to impart additional strain cance. The most striking feature of the diaziridines to the ring.) For those analogs with electron-withis the high activity of the amino-substituted comdrawing constituents, preparation of the intermediate pounds. Quantitative evaluationgof compounds 14-18 basic aziridine proved to be difficult. Only esters of revealed that their activity in sterilizing male house 2-aziridinecarboxylic acid6 were suitable for synthesis flies was higher than that of the parent compound tepa of the tepa analog. Ethyl 2-aziridinecarboxylate was and that 14, a methylaniino derivative, is one of the converted with phosphorus oxychloride to the cormost potent male house fly sterilants known. Beresponding tris(2-carbethoxy-1-aziridiny1)phosphineoxcause the aziridinyl moiety is undoubtedly the carrier ide, but the sterilizing activity of the latter was very of sterilizing activity in this type of alkylating agents, low. one could expect that a reduction in the number of Although the tris(2,3-dimethyl-l-aziridinyl)phos- such groups in a molecule of a chemosterilant would phine oxide3 was a much weaker chemosterilant than always lead to a decrease in sterilizing activity. Almetepa, it was conceivable that if the two alkyl subthough the physical properties of the lower alkylamino stituents were linked together, thus exerting additional derivatives (14-17) do not appear to be different from strain on the already strained aziridine ring, the sterithose of tepa (l), the findings reported herein aclizing activity of the bicyclic derivative might be incentuate the importance of the carrier portion of the creased. Cycloalkenimines, previously synthesized,' molecule to its physiological activity. I n general, appeared suitable for this purpose. the sterilizing activity decreases with the increasing The treatment of 6-azabicyclo [3.1.0]hexanej 7 chain length of the alkyl group, the octyl compound azabicyclo [4.1.O]heptane, and 8-azabicyclo [5.1.O]oc(20) being only slightly active. Dialkyl substitution tane, respectively, with phosphorus oxychloride, yielded on the amino nitrogen (28-30) is also effective, particuthe expected trisubstituted phosphine oxides (Table larly when methyl groups are the substituents (28). e
+,
(5) M. Beroza and A. B. Boikovec, J . N e d . Chem., I , 44 (1964). (6) Hoffmann-LaRoche & Co., British Patent 847,205 (Sept 7, 1960). (7) (a) 0. E. Paris a n d P. E. F a n t a , J . A m . Chem. Soc., '74, 3007 (1952); (b) P. E. Fanta, J . Chem. Soc., 1441 (1957); (c) P. B. Talukdar a n d P. E. Fanta, J . Org. Chem., 24, 555 (1959).
(8) (a) A. Hassner and C Heathcock, Tetrahedron. 20, 1037 (1964). (b) of 3-oxa-6-azahiryclo[3.1.0]hexane by the \\enker method u a s announced: cf. P E F a n t a and E. N. 1% alsh, J . Org. Chem., 31, 59 (1966). (9) S. C. Cliang and A. B. Boikovec, J . Econ. Entomol., in press.
M liile this paper mas in preparation, the syntheuis
;24
B (2.0H) A (4.0H)
I
C
~
Figure 1.-Nmr
A
~
spectrum of 3-oxa-6-azabiryclo[3.l.O]hexanein CCld ( 13c; )
More precise deter in in at ion^,^ however, indicated a decrease in activity of the dimethyl compound 28 compared with that of the monomethyl conipound 14. Alkyls were not the only substituents to maintain a high activity (22, 24. 26, and 27), but they were the only ones giving rise to activities greater than the activity of tepa. Alkoxy (11) and aryloxy (12 arid 13) diaziridines were also active, but substantially less so than the parent triaziridinyl compounds. The aryl group (10) proved to be even less active as ,z substituent than the aryloxy group. Monoaziridinylphosphne Oxides.--The effects of substituents in the morioaxiridinyl series (Table 111) paralleled those in the diaziridiriyl series. Most of the compounds, with the exreption of 38, had either very low (33 and 39) or insignificant activity. Xttenipt> 1o prepare alkylamino derivatives analogous to 14, 16, and 17 were unsuccessful. In spite of several niodifications in the experinie~italprocedures only illdefined polymeric materials were obtained. Sevcral phosphine sulfides, arialogous to the phos~ihineoxides presented here. were tested."J Results substantiated the findings evident in Tables 1-111. Unfortunately, the sulfur coinkiounds often exhibited :i high toxicity to test insectb; their usefulness in structure-activity studies was thus severely limited. Comparison of the activities of 22 and 23, 33 and 34, us well as 38-40, further substantiates the effects of alkyl substituents on the aziridinyl ring previously observed3f4i n tepa (1) and nietepa (2). Results of these tests point to the following generalizations: (a) any substitution on the aziridinyl ring of a chemosterilant reduces thc sterilizing activity of the derivative compared with that of the parent compound, and (b) replacing an aziridinyl substituent in a poly(101 Vnpublished results.
~~~
B
aziridinyl chemosterilant does riot necessarily reduce the sterilizing activity of the derivative as compared to the parent conipound. Experimental Section" Biological Testing.-All compounds listed in Tables 1-111 were tested a t various concentrations as additives to the diet of house flies.I2 Gradation of activities was expressed by a series of plus sigiis. All feeding tests a i t h house flies were performed by the entomologists of the Entomology Research Division, Agricultural Research Service, U. 8. Department of Agriculture, Gainesville, Fla. Ethyl 2-Aziridinecarboxylate.---A procedure reported i i i a patent6 was followed, but the rate of mixing the reagents proved to be critical. To 500 ml of liquid ammonia was added 31 g of ethyl a-bromoacrylate during a period of 1 hr. If the additioii of the acrylate is too rapid it reacts with the intermediate 8amino-a-bromopropionate rather than with NH3 and the yields of the aziridine are drastically reduced. The present procedure yielded, upon distillation, 11 g ( 5 5 7 , ) of ethyl P-aziridinecarboxylate, bp 55-57' (12 mm) [lit.@ bp 53-54' (11mm)]. 3-Iodotetrahydrofuran-4-ethylcarbamate.-To a iolutiori of iodine isocyanate, prepared by the reactiou of 76 g (0.60 mole) of iodine with 50 g (0.33 mole) of freshly prepared silver cyanate in 500 ml of dry ether, was added dropwise 21 g (0.30 mole) of 2,5-dihydrofurari. The temperature was kept a t -70" during both reactions and was allowed to rise to room temperature after the addition of dihydrofuran had been completed. Precipitated silver iodide u as collected by filtration and the solution was concentrated by distillation under reduced pressure. The resulting isocyanate was not isolated but was converted to the carbamate by treatment with refluxing ethanol for 3 hr. This solution was evaporated t o 100 ml, diluted with 600 ml of water, and decolorized with a small quantitl of sodium bisulfite. The carbamate mas extracted into ether and the ether solution, after drying (MgS04), wa? further decolorized by passage through R ( 1 1 ) A11 melting and h i l i n g points are corrected. (12) (a) For detailed experimental procedures, see 1%.
li. Gouck, >I. 11. Crystal, A. R. Boikovec, and D. W. Meifert, J . Econ. Entomol., 66, 506 (1963). (b) For a discussion of concentration effects in oral application of rhemosterilants, see C . IG. Smith, G. C . LaBrecque, and A. B. Boikovec. A n n . Rev. Entomol., 9, 269 (1964).
INSECT CHEMOSTERILASTS. I11
July 1966
TABLEI1 HOUSEFLYSTERILIZING ACTIVITYOF DIAZIRIDINYLPHOSPHINE OXIDES Graded Compd
Structure
10
( C H 1 3 N ) m c,,H,
11
(bN):POOC2Hj
activitya
+ ++
Source
b
C
12
+
d
13
+
d
21
+
h
i
23
+
25
++ ++
26 27
29 30
( CN)POWH
j
j
+ +
h
++
9
k
N--N
31
a See footnote a, Table I. * Interchemical Corp. c Continental Oil Co. d Minnesota Mining and llanufacturing Co. e Prepared according to T. Nizuma, Y. Minaki, and S. Toyoshima, Yakugaku Zasshi, 81, 48 (1961). f Edwin Kuh, Rutgers University. See Experimental Section. American Cyanamid Co. Armour Pharmaceutical Co. 1 The Squibb Institute. National Institutes of Health. Q
column of alumina. The crystalline product obtained by evaporation of the ether solution and two recrystallizations from cyclohexane weighed 30 g (357,)and melted a t 81-84'. Anal. Calcd for C7HdNOa: C, 29.41; H, 4.51; I, 44.3; N, 4.90. Found: C, 29.70; H, 4.27;I, 44.75; N,5.01.
3-Oxa-6-azabicyclo[3.l.O]hexane.-A solution of 23.7 g (0.083mole) of 3-iodotetrahydrofuran-4-ethylcarbamate and 14.1 g (0.25mole) of KOH in 400 ml of water was heated a t reflux temperature for 2.5 hr. The solution was saturated with NaCl and continuously extracted with CH& overnight. Concen-
52ti
1
L
A
r
i
t
See f o ~ ~ t i i o CL, t e Table I. Iiiterc*hemicalCo. See ExperiP. E. Sonriet and ii. R. Ri~fkovec~, iinpi~hlished Sectioti.
meiit:il 1131:1.
tratioii of the extract followed by distillation at reduced pressure gave 3.8 g ( 5 4 5 ) of the aziridiiie, bp 48-51' ( 2 2 mm), n Z Z1.4694. ~ .Inn[. Cdcd for C4HiNO: C, 56.45; €1, S.29; Tu', 16.46. Fiii~iid:C, 56.26; 11, 8.45; Y , 18.20.
Tris(6-azabicyclo[3.1.0]hexan-6-y1)phosphine Oxide (6).To 1.4 g (0.017 mole) of 6-azabicyrlo[3.1.0]hexaiieii> and 1.7 g (0.017 mole) of triethylamine iii 25 nil of benzene was added 0.76 g (U.005 niole) of POC13 iii 14 nil of benzene. The reaction iiiix1ure was cooled i n ail ice bath duririg the addition arid then kept at 25' for 3 hr. The precipitated triethylarriitie hydrochloride waa collected by filtration, arid the filtrate was evaporated t o :til oily, c~ryatallitiemass. Two recrystallizations from pen1:riie gave 0.9 g ( 6 0 5 ) of idorless crystals, inp 105-107". A n d . Calcd for C I J & ~ N ~ O PC, : 61.42; If, 8.23; S , 14.32. Foiiiid: C, 61.17; II,S.34; N, 14.52. Tris(7-azabicyclo[4.1.0]heptan-7-y1)phosphine oxide ( 7 ) as prepared from 7-azabicyclo[4.1.0]heptariei~in the same iiiatiiier as the honiolog 6. The product wa$ obtained iri a d after recrystallization from pentane and melted a t Chlcd for C1,&S30P: C, 64.45; 11, 9.01; N, 12.53. C, 64.20; H, 8.82; N, 12.67. Tris(8-azabicyclo[5.1.0]octan-8-yl)phosphineoxide (8) was prepared from 8-azabicyclo [5.1.0]octane7~in the same niariiier :is 6. The product was obtained iri a 60y0 yield after recrystallization from pentane arid melted a t 89-90'. 1: nu/. Crtlcd for C I ~ I ~ ~ ~ XC,~ 66.81 O P : ; 11, 0.61 : 3,11.1:3. Foriiid; C, 66.67; H, 9.75; N, 11.25. Tris(3-oxa-6-azabicyclo[3.1.0] hexan-6-y1)phosphineoxide ( 9 ) was prepared from 3-oxa-6-azabicyclo [3.1.O]hexane in the same iii:iiiiier :is 6. The product was obtained iri 757;, yield and riielted at 14(3-1X0 after recrystallization from ethyl acetate. A d , Calrd for CI2HlsN3O4P:C, 4s.16; 11, 6.06; S , 14.04. Foiititl: C, 4'7.86; [T, 6.07; N, 14.12.
.Id.
Foiititl:
