HOMOLYTIC AXD OTHERC O N ~ T A N IKT1~,:~-BEXZODIOXOLES
September 1 W S
92 1
Y
Y KO.
1
Y
H
11
I1 II Ir
H H H
Znn
Xu' b
ZapC
2a1 d
xu
d
liso
Obsdf
Cald'
log SR5
log YR5
00 00 00 00 1.40 1 386 00 00 -0.17 00 1.9; 52 0.09 -0.03 -0.31 2 029 0.03 0 42 0.23 00 1.88 0.11 2 023 70 0.11 0.43 0.23 00 2.19 02 2 2-52 0 . 13 0.47 00 2.48 0.78 0.68 I I1 13 24 2 480 0.79 0.40 €1 -0.27 00 2.44 0.28 2 141 11 -0 04 -0.78 6 0.84 2 258 0.06 0.46 00 2.28 Ir 0.22 7 c1 1 40 0.22 0.90 2 489 0.46 8 Br 11 2 04 00 2.44 0.30 -0.04 00 2.62 0.43 0.70 2 390 H -0.67 !I 0 66 c1 0.73 -0.04 0.51 -0.63 I1 00 2.66 2 824 IO 0 98 Br -0.23 2.73 0.96 0.51 0.87 0.01 11 so, 1% 2 770 0 20 1.10 1.01 -1.06 2.11 0.50 0.90 12 KO2 I3 1 870 0 94 1.1s 1.01 I1 0.5s -1.24 1.92 0.94 12 1 910 1 26 NO1 1.36 I3 0.94 1.56 1 .TO9 -1.99 1.38 1.58 14 XO, 0 48 1.68 0.12 0.92 00 2.34 0.44 2 80 1.i c1 c1 2 121 1.80 0.44 Br 0.92 00 1.59 0.60 Br 1.677 16 4 08 a K values ale from the phenoxyacetic acid bystem.'," * See text for definition of u'. c From ref 13. d From C. 1).Ititchie and W. F. Sager, Prog. Phrls. Org. Chem., 2, 334 (1964). e For only compounds 11-14 did it seem appropriate to use this constant. See discussion under method. f From ref 4. 8115 represents data from experiments where the ratio of synergist to insecticide r a s 5:l. Calculated using eq 7 1 2 3 4
0 0 1 0
beiizcnc ring gave deriw tivcs of increased activity. This iir: a well-linown characteristic of substituent effects on homolytic reactions with aroniatic conipounds. l 4 The two iiiost activating substituents in Table I are the nitro and iiiethoxy functions. It is well known that in nucleophilic or electrophilic substitutions these functions have opposite effects. However, in certain homolytic substitutions, nitro and niethoxy are among the strongest proiiioters of reaction.'C Relatively littlc work has been done using the Haniniett equation for hoiiiolytic reaction. Leffler arid Grunwald'6 point out that ionictimes radical reactions can be correlated uhing u. Pryor17 has pointed out that in a number of instances substituent effect< for electrophilic radicals are correlated by u + . The fact that we did not obtain a very high correlation using u + appeared t o indicate that an electrophilic-type radical might not be involved in our system and that it might be profitable to forniulate a Haminett-like constant from the relative rates of phenylation obtained by Hey and \ T i l l i a n i ~ ~for ~ ~ conipounds ~* of the type C6H,X. For the substituent X, u' is defined as log g$HK where K is the relative phenylation ratel4 for all positions on PhX. Since in our example of the 1,3-benzodioxoles each substituent is situated para to one and meta to the other. iiiethylenedioxy function, it seems inore appropriate to use the total rate constant K rather than a partial rate constant for a. ??zetaor para position. Equation 2 illustrates how the methylenedioxy function is admirably constructed to delocalize an odd electron re(14) G . H. Williams. "Homolytic Aromatic Substitution," Pergamon Press, New York, N. Y., 1960. (15) Reference 14, p 5 i . (16) J. E . Leffler a n d R . Grunwald, "Rates and Equilibria of Organic Reactions." J o h n Wiley a n d Sons, Inc., New York, N. Y., 1Y63. p 171. ( 1 7 ) \V. .I. Pryor, "Free Radicals," McGraa-Hill Book Co., Inc., K e w Y o r k , X . I.., 1Dti6. Chapter 12. (18) G. 11. iVilliains, Chem. Irrrl. (London), 128ti (1!161).
sulting from the homolytic reinoval of it niethylene hydrogen arid how a niethoxy group could participate in this process. The valuet5of u' for OCH3represents an
CH3
CH3
FH3
uncorrected value in which some attack of the type depicted in eq 3 mould make p$:(&H8K a little higher than that corrected for attack on the ring only. For our purposes this is very likely an advantage since the methoxyl group on the benzodioxole ring could react in a parallel manner in the biochemical system
FH3 A rather successful approach for correlating substituent effects in radical reactions has been recently developed by Yaniamoto and Otsu.'g These workers have shown that the relationship, log ( k / k o ) = pu ~ E can R be used to correlate a variety of radical
+
(lY) T. l n m a m u t u and T.
Otsu, i b i d . , i 8 i (1YBi)
+ 1 !)ti3 log SI35 = 0 070r + 0 g l i a + 2 Ot50
log SKT,
log SI15
=
=
0 910u
-0
ll.ix?
+2
I4(i
log SI = - 0
I!).-)+
0 :j4xn 0 (ii0.lr
+I
/I
,
4
1:2
0 li03 0 ;iX4
(1)
1X
O(i3h
OBKh
(-5)
1:2
o.io0
0:M) ((i)
1;2
0 H29 0 171
+ t
; < l ( < u -t
1 til2
(7)
serticiclc \va- ,-).I, I I atatids ior the nuiiiher ot data point s uqed in the r e g r e G m , :iiitl s represents the stari( h i d cleviat ion froni regrehiion. Coiiipounds 12-14 v ere riot iricluded for reasons cliscussetl under t h c .\lethod Sectioii. u is L: rdic:il coiiitaiit dehiiecl 111 the *ect i o i i o i i iiiethod Equatioiis 1-6give poor corrclatioii-. .ldditig :iii elcctroriic tcriii to eq B j.icltlh ccl i I\ hicli gives :I vvrj good coirelat ion, uccouritirig for ,\(io; ot t l i r variaiicc in t he data. Taking the partial (It.i-ivative,b log SIG ' b a , and setting thiq equal t o zero, gives ti value of 1 . i 2 (1 41-1.96) for no. This reprew i t s the ideal lipophilic character for the b u i ~of the subst i t i ~ e r i t s . ~The figures in parentheses define the !IT,C, wiifidence interval 011 this constant. The posiI ivci coefficient with u indicates that the electronic aspect of the substittierit effect parallels that of the honio1) tic pheiiylatioii of simple l~e~izeiie derivative*. ILOI
I{rirrriir~Ih
11
218
September 1968
HOMOLYTIC AND OTHER
To explore the possibility that higher order equations might give better correlations, we added a term in T U to eq 12. A significant reduction in variance was not obtained. The same is true for the addition of a term in ~ 2 . However, adding two terms does result in a quite significant reduction in variance (eq 15). How
+
r
rL
S
log SRS = -0 1 2 3 ~ ~ 0.633~ - l.S23u"+ 3 . 1 6 2 ~' 0 . 7 9 6 ( ~ ~4-' ) 0.639Es 1.450 16 0.991 0.074 (15)
+
much validity can be attached to eq 15 remains to be seen. One would certainly like a good many more data points before niaking any strong statements about the meaning of the additional terms, especially in view of the fact that neither terni alone caused a significant iniprovenient . From a smaller group of compounds (Table 11) for which the synergistic ratio was defined in a slightly different fashion, we obtain eq 16. From eq 16 me find
+
r
n
+ +
log S125 = - 0 . 5 7 6 ~ ~ 1.90SA 3.50lu. 0.233
6
S
0.934 0.404 (16)
A" = 1.66, in good agreement with the other work. However, the confidence interval on this figure cannot be defined.21 While the coefficient with u ' is much higher for this different type of fly, we cannot say that the difference in u' is due to the different fly because of the uncertainty in the value of the coefficient with u'. The 95% confidence interval is wide (h4.29). This is partly because of the few points and partly because of more scatter in the data.
COXSTANTS IX
C.\LCULATED SYXERGISTIC
ACTIVITY FOR
~IOKOSUBSTITUTED1,3-BENZODIOXOLES WTH CARBARYL I K FLIES Substituenl
0 .
7r
CH30
NOn &Butyl CH3 c1 H From ref 8.
0
0.40 -0.04 0.24 0.47 1.68 -0.06 0.09 0.82 0.70 0.03 0 0 Calculated using eq 17.
Ohsda log SR5
log SR5
Calcdh
1.91 1.95 1.57 1.50 1.48 0
2.293 1.592 1.374 1.381 -0.223
1.546
It is of interest to coiiipare the results obtained with the single constants u D ,u+, and u' with the more complex constants (u E R ) of Yaiiianioto and Otsu (see the llethod section). Equations 17 and 18 can be compared with eq 7. I n eq 17 we have siniply added the
+
+ +
log SR5 = -0. 160n2 0.386~ 1.883E~ 1.777 log SR5 = - 0 . 1 8 0 ~ ~ 0 . 5 6 6 ~ 3 . 1 7 6 E ~0 847u, 1 58.5
+
+ +
+
n
r
S
13 0.843 0.249 (17)
13 0 944 0.162 (18)
923
resonance interaction term (ER)to eq 6. While eq 17 does result in considerable reduction in variance, when compared to eq 6 it is not nearly as good as eq 7. Equation 18, containing an additional variable and an additional disposable constant, gives a very slight but not significant improvement over eq 7. While the u ER constants have been shown to correlate a variety of radical reactions, we feel that a t least in the present instance the greater simplicity of u ' is a distinct advantage. Log Po for eq 18 is 1.57 (1.10-1.91), in good agreement with the value obtained from eq 7. The fact that two different sets of radical constants give comparable results for the benzodioxole synergists strongly supports the idea that these molecules are involved in radical reactions. In the study by Wilkinson, Metcalf, and Fukuto from which the data in Table I1 are taken, activities for a large number of derivatives were reported. Unfortunately, substituent constants are not available to analyze their relative activities. From an inspection of these data one is inclined to suspect that flatness of the synergist may be important for maximum activity. The question arises, what does the fact that electronic effects of substituents in the synergists parallel those in honiolytic phenylation mean? As mentioned in the introduction, the hydrogens on the CH2 group of the methylenedioxy group are strongly implicated in the synergistic action and synergists seen1 to operate by inhibiting microsomal attack on the insecticides. Thus, it seems that abstraction of a hydrogen by a microsomal enzyme from the synergist according to eq 2 could generate a relatively stable free radical which could act as an inhibitor of a free-radical generating enzyme responsible for oxidation of C-H bonds in lipophilic drugs. ,4 relatively stable radical, if sufficiently lipophilic, could remain bound to an enzyme, thus inhibiting its further action. I n the typical enzymatic reaction sequence shown in eq 19 we have foundz2that k, can be
+
TABLE I1 OBSERVED B S D
1,3-BEh.ZODIOXOLES
E
+S
ki
k3
ES +E i"1
+ product
(19)
greatly reduced by lipophilic binding between product and E. The great dependence of activity of the synergists on u' and the rather poor correlations obtained using Q + and UI would indicate that H removal by homolytic cleavage is a better postulate on which to base discussion than hydride removal suggested by Hennessy.'O Turning to eq 12-14, we find r0to be 1.71, 1.79, and 1.75, respectively. Taking advantage of the additive character of A and log P we can calculate log Po, the ideal partition coefficient for a synergist. Log P for 3,4-niethylenedioxybenzy1 alcohol is 1.05; subtracting log P for benzyl alcohol (1.10) from this yields i~ (-0.05) for the -OCH,O- moiety. Addition of this to the log P of benzene (2.13) gives 2.08 for 1 in Table I. Taking TO to be 1.75, we find log Po of 3.S for the ideal lipophilic character of a synergist for carbaryl in flies. Keeping in mind that the molecules of highest molecular weight should be rated more active on a niolar basis (see the Method section), we would expect log Po for the set to be close to 4. I t is of interest to note that this is (21) C. Hensch, .I R . Stenard, S RI Anderson. and D. L BentleJ, J . M e d . Ckem., 11, 1 (1968). (22) C Hansch. E. \V. Deutsch. and R N Smith, J . A m C h e m Soc 87, 2738 (1965)
Es tra-l,3,5(10), 15-tetraenes.
1.
Birch Keduc tioii
The :ttlditioii ( i f org;iiiolit hiiiin rcicgeiilh t,ii :bmeilici~yes~ i.tt-l,:j,:i( Io), I ~ ~ - t ~ ~ ~ i ~ ~ i 1 , t3I~> ii t~s l -i l l( d~ :I - i*orit,< ~i~t 1i-suhstit~tt,edAis deiivat,ives (6). Birch reduction of 6a and 6b at -7s" 1t:d t o ruliiciioii of iiilg .4 \viihout, r!hduction of the A15 double bond. Oppenaiier oaidatiori at, room lenipertttiirc of Ihc iiitermedittt e :I-rnet,hoxyesf ra2,5(10),15-trien-l7~-ol (10) afforded the ketone 11 which was coiiverted to a scricn of wtive progt\tirrs. 'rhc hypoeholesteremic and estrogenic activities of the intermediate aromatic steroid. arc i q o r t e d . .A himpic prcic~rt11ir.c for ethynylation of base-sensitive ketoiies is described.
As part of a synthetic program leading t o modified steroidal estrogens arid their derivatives, some re:ictions of estra-1,3,5(10),15-tet~raenes,in part'icular 3methoxycstra-l,3,3( lo),15-tetraen-17-one (3), were es:tmincd.' The A15-17-one 3 had previously been prepared i i i five steps from estroiie methyl et'her (1) b!. Johnson :uid Johns. We used essentially the same procedui,e, but reduced the number of st'eps to four by direct bromination of estrone methyl ether with CuBrZ3 (Scheme I). 111 :in effort to reduce the number of steps even further, t'he direct, dehydrobromination of bromo Icetolie 2 w:ts reesamined. I n a related series, Pappo, et ( I ) E. \ \ , Cantrail. R. Littell, and W. Bernatein, .I. Org. Ciiem.. 29, 6 1 l!t6.i]. liar e iiwd 3 to prepare a series of 16-s11I~tituted derivative,, ( 2 ) \\., 5 . .lolinron a n d \\-, F. Joli V I . C ' / ~ w i .Sor.. 79, 20005 ( 1 9 > i t . ( ; $ I L. it. (ila,,ier, .I. O r g . Ciiem.. , I ) I ? . I'ai,pu. H 11. 1ilo01~1a n d \\. P. ~ l o l ~ n s o r.I. i , . l n i . C'//ctti. li31J 1 I!J561. !
~
