6420
H. K. HALL,JR., l f . I(.BRANDT.\ND R.11.LfASON
IroI.
so
were heated at 10O-2OOo for 24 hours with litharge, potaschluroform :md trichloroethaiie. It \vas itisolublc iii 7)~sium carbonate, sodium hydride, tetraisopropyl titanate cresol, trifluoroacetic acid and 90% formic acid. Both polymers depolymerized t o the corresponding lacnnd 2,5-dichlorobenzenesulfonicacid. Lactones and cyclic ureas were heated a t 150-260" with water and sodium hy- tams when heated with a flame, the 1,3-isniner at lnwcr temdride for a similar period. peratures. Lactatn 1-1polymerized so quickly when heated to 200' Because of the possibility of isomerizatioii tluriiig polythat melting could not be completed before the polymer merization, these polymers may be mistures o f cis and trans congcalcd. .ittempts to moderate the reaction with milder forms. catalysts (see below) or b y conducting it in solution in ethyl Chelates. ~- h relationship between ease of forination of benzoate, y-butyrolxtone, pyrrolidone or 5.5-dimethylpyrchelate compounds and of bicyclic organic cnnipounds has rolitloiie,' gave no improvement in yield or molecular weight. been noted and discussed .61--63 Iii agreement with this conLactarn V polymerized at a much slower rate and addition cept, tlie feasibilit>- of forming cyclic :md bicyclic ureas of S-ncct!-lcaprolactam w:ts beneficial. Lactoiie IY pol>-- froin dianiiues could be assessed by seeing ivlicther or not iiierized a t about tlic same rate, confirming earlier ~ o r k . ~ they , ~ formed chelates with tnctal ions. Lactone S I pol\-mcrized slowly to give a low molecular The diamine ivns :tddctl t o an aqueous ~ i i l u ~ i ~o fi icupric i \\-eight, solvent-sensitive polymer. The polyuretlian ob- acetate. The appearance of an iiitetise violet color \vas tained from XI1 W:LS also sensitive to solvent and of low taken as :tii indication of chelate formation. The following ~nolecularweiglit. gave posirire tests: ethylenediatnirlr, t r i t n e ~ l l y l e t i e d i 3 ~ i i l i c , B~C~L tlie I Sreaction ~ of VI with sodium hydride procis-l,3-diamiiiocyclolies3iie. cis- and traIzs-l,-l-tliutni~ioci.et1i.d too rapidly for coiirenient control, a search was cyclohexane anti 1,8-diamino-p-menth:tlle gax-e negative innde for other catalysts. The following were ineffective tests. This order is in good accord with the observed tend(at 200" for 24 huurs): water, e-aminocaproic acid, sodium encies of these diamines t o form cyclic u r w s . This test should also apply t o a r n i n o c y c l o l i e ~ a ~ i ~ ~ l ~ . or potassium c:trbon:tte, sodium acetate, boric acid, sodium phosphite, litharge, antimony trioxide, tetraisopropyl tiFor the best results a rnetal atom should 1i:ive tllc s t m e ' ,ite, y-butyrolactone, toluenesulfonic acid and sodium stereiicheinistry as the c:irboti atoiii to viliicli it will correnide. .I11 salts were tested with and mithout 3. trace of spond . However, vie ha\-? usetl copper (sqmre pl:tn:ir'~ t o cr. Pliosphoric acid gavc a very small yield of polymer. correspond to carbonyl carbini (trigrinal pl:in:tr) hcc:tusc of .It tlie boiling point of t h e lactaiii, sodium phenoxide arid easily observcd color changes 1111 chelatiiiil . __ c:irbonate were still ineffective. Poly-3-cycloliexanecarboxaniide was soluble in ?iz-cresol, ric acid, trifluoroacetic acid, 90 and 997@formic acids, trichloroethaiic~-~07@ formic acid and 60% chloroform-formic acid. The 4-isomer dissolved in sulfuric acid, ~VILMISGTOS98, DEL. 997$ formic acid, : t i i d in the mixtures of the latter with
1 C U S ~ I ' K L I J ~ I ' I c l S1'11 T i I E P l l J S E E R I S ( ; 1 Propiolactone polymerizes readily and y-butyrolactone does not, pet the two lactones hydrolyze a t cotiiparablc The suggestion of Carothers is therefore not generally valid. lic. ~wiiip(iu~id is a n i d Polyinerizability of a cation of strain in the rit and sirice the hydrolysis rate is not determined by strain in the ring, i t follows that the ring is riot broken in the rate-determining step; the latter must consist of addition of OH- to the ring, iii agreement with Bender's results for open-chain coiiipounds.iC Reactivity of Cyclic Compounds Relative to Acyc1ics.-Table 111 makes clear that only lactones (and possibly cyclic carbonates)8 exhibit a large ( 3 ) €1. K. Hall, J r . , TIXIS J O I I R N A I . , 81, 6412 (1959). ( t i ) P. A . Long and hI. 1-urchase. i b i d , 72, 3267 (1950). ( 7 : (a) (>. E. Vv'heland, "Ad\-dnced Organic Chem a n d Suns, Inc., K e w X'Lxk, S . Y . , 1949, p. 3 0 6 ; (b) I;. J . [\.in, Ou,ni.l. Rers., 12, 82 ( 1 S 3 S ) ; (c) 31. I.. B s A r , , 73, 102b (1951). and later i x ~ i x $ S > SI1 %ire1 and . I l. ' c ~ ~ h < , r >i l ~ s ,
TABLE I RATEDATA Acyl derivative
Xcetylcaprolactam
Concn. X 10s
ki X 108
kz
Xlethod
1.49 1.70 2.99 4.04 5.42 7.51 10.7 15.6 20.3
2.86 2.99 6.11 7.08 9.80 13.3 17.3 23.2 39.4
1.9 1.8 2.1 1.8 1.8 1.8 1.6 1.5 1.9
Cap. Cond. Cap. Cond. Pot. Cond. Pot. Pot. Pot.
10 17 26 33
13 21 34 54
3 6 1 8
3 1 7 8
1.3 1.2 1.3 1.6
Pot. Pot. Pot. Pot.
1 57 3 15 4 07 5 85 9 12 9 84 19 2 28 0 36 3
3.04 6.31 7.13 13.1 16.3 15.6 33.1 41.7 50.6
1.9 2.0 1.8 1.6 1.8 1.6 1.7 1.5 1.4
4 6 8 9 11 24 46
2 4 2 2
Cap. Cap. Cond. Pot. Cond. Pot. Pot. Pot. Pot.
9.63 8.77 9.28 21.3 24.4 40.4
2.3 1.4 1.3
1.77 4.86 6.22 7.50 16.3 39 9
3.90 8.43 10.0 10.8 20.7 46.8
2.2 1.7 1.6 1.4 1.3 1.2
N-Methylglutarimide
1.il 6.03 7.23 9.40 10.35 19.75 28.3
73 82 48 96 87 7 3
22 2i
63
i1 8 1 6
2 2 2 2 2
5 5 4 3 2
5 15 8.25 9.87 12.7 28 1
3.6 3.0 2.5 2.5 1.9
a S - B e i i z o ~ l c a p r o l a c t ~ ~ t 2i ~. 54 4.55
2 . 76 4.18
Cap. Cap. Cap. Cap. Pot.
1.76 3.52 6.81 9 23 10.7 17.0 17.6
5.13 11.0 18 3 24.7 33.0 52.5 45.7
2.12 4.16
3,72 7 75
3 3 2 2 3 3 2
9
1 7 i 1 1 6
COlld. Cond.
cap Cap. Cond. Pot. Cond. Cond. Pot.
2 . 8 9 xv. N-Methyldiacetaniide
1.8 1.9
0.14 .17 .14 .13 .15
Cap. Cap. Cap. Cap. Pot. Pot. Pot
1.81 Av. 9.71 12.7 15.4 20.3 25.2
N- Acetylethyleneurea
23.5 45.0 56.5 60.5 68.8 137.3
2.61 6.21 4.57 4.86 4.64 8.88
N-Acetyl-2-azabicyclo [2 :2 : 21 octan-3-one-(IV)
Pot. Cap. Cap. Cap. Pot.
1.36 3.60 6.00 6.21 10.0 12.9 19.8 41.6
2.36 3.40 5.25 5.31 8.75 11.6 18.5 38.4
Cap. CaD.
0.13 Cap. .14 Cap. ,081 Cap. ,080 Pot. .068 Cap. ,065 Pot. __ ,094 Av. 1.7 0.94 .88 .86 .88 .90 .94 .92
Cap. Cap. Cap. Pot. Cap. Cap. Pot. Pot
0 , 9 0 -4v. N-Acetyl-6-azabicyclo[3:2: 11 octan-7-one-(111)
1 0.lv
S-.\cet~losazolidoiie
2.3 1.8 2.0 1.9 1.6 1.6 1.5
Cond. Pot.
0 .lv.
1 1 0 $10
3.90 10.7 14.1 17.5 16.2 31.3 42.7
70.7 73.7 111 154 174
2 . 7 AV. 3 2 2.8
Cap. Cap. Cap. Cap. Pot. Pot.
0. 146 Av. (0.753)'
Cap Cond. Cap. Cond. Pot. Cond. Pot.
2 . 3 6 Av. 1 45 2.77 3.97 5.09 14 9
1.2
N-Meth ylsuccinimide
1 70 A\,. S-.~cet~.l-2-p~-rrolidolle1 2 3 3 4 10 21
1.1 1.3
Cap Pot. Pot. Pot. Pot. Pot.
1.57 Xv.
1 . 3 5 Av. N-.~cet?.l-2-piperidone
4.25 6.29 7,14 19.2 19.3 32.6
1 . 5 4 Av. Xi-Xcetyltrimethylene urethan
1.80 Xv. Acetyl-5-methylcaprolactam
K-Isobutyrylpyrrolidone
642 1
HYDROLYSIS RATESAND XECHANISMS OF CYCLIC MONOMERS
Dec. 5 , 1958
1.08 1.98 2.85 2.86 4.41 5.41 7.47 13.2
5.40 6.71 8.53 9.47 14.3 21.8 24.2 42.4
5 ,0 3.4 3.0 3 .3 3.2 4.0 3.2 3.2
Cap. Cap. Pot. Call. Cap. Cap. Put.
Pot.
3.3 x v y-Butyrolactone
5.32 10.1 16.1 31.4 34.7
5.41 7.91 13.2 23.2 24.0
1.02 0.78 .82 .74 .69"
Put. Pot. Pot. Pot. Cond.
0.81 Av.il.ll"'i
Yd.so
H. K. HALL,JR., 11.K. URAXDT . ~ N DR. 11.1 2 1 . ~ ~ ~ TAuLs I Acyl derivative
(Conti,tucd)
Concn x 103
k, X lo3
TABLE 11 kz
Method
i -Ynlerol:ictonc 8-\'alerolactoiic
2 83 2 39 4 78
8 9 0 0 31(0.467d)C(md 32 7 13 7 Pot Pot. GG 3 13 9
t-Caprolactoiic
x
6 G1 16 5 29.1
13.8 .lv. 31 21 9 38 7
0 80
Pot.
.76 Pot. . r 3 Pot.
--
0.77 h v . .36 Pot. . 3 5 Pot. __ 0.35 -1v.(0 . 45-0.50") N-Etlios>c.irbon) 1p yrrolidonc
13 5 31 7
8.63 11.3
0.27 .33
20 2 38 5
9 0; lti 7
0 16
Pot. Pot.
0 30 .\v 43
Pot Pv:
0 44 .Iv (i-Osabicyclo[3: 2 : 11Octall-7-onc ( I )
9 14 2; 5 62 6
2 98 10 4 22 9
0 33 38
3;
Pot. Pot Put.
0 . 3 6 -Iv,
L'-Oxal,icyclii [ 2 : 2 : 21 octatl-3-oIle (11)
88 8 85 6
1 09 2 15
0 028
025
Acyliactatri
k,,,,,
N-Acetyl-2-pyrrolidone 0.64 0.77 ?j-AAcetyl-2-piperidoi~e 1.30 h~-.~cetylcaprolactam ~ - ' - A c e t y 1 - 5 - m e t l i ~ l e ~ ~ ~ r ~ ~ l ~ ~ c1~.~0 i9i n N-Isobutyrylpyrrolidone 0.86 N-Benzoylpyrrolidoiir 7 0.39 X-Benzoylcaprolactani 0.004> S-Acetylimidazolidone 1.25 A~-.4cetylosazolidoii~ S-.lcetyltetrahydroos3zirlolle 0.33 .. S-hlethpldiacetarnide 0.15 N-Meth ylsucciniinide 1.81 S-Methylglutariinide S-.~ce:yl-O-azabicyclo[3:2 : ljoctan-73.33 one (I) ~-..rcetyl-2-azabic).clo[2 : 2 . 2 )octan-30.31 cine (11) G-Osabicyclo[3:2:l]octan-7-0ne (111, . 36 .027 2-Oxabicycloj2: 2 :2]octaii-3-one (IT) .44 S-~Zetlianesulfonylp>-rrolidoiic S-Ethuxycarbonylpprroli~one .30 I .
I., ,
\
Li2 0.9s .31 33 1 83
0.87 .81 ,0114 1 ci4
1.24 1.54 ..
.. O,l5> 05!l . .
..
..
pair atom dipole of oxygen.10--12 I n open-chain tyans esters C this interaction cannot be avoided, so esters react abnormally slowly. Anhydrides react equally quickly in the cyclic (D) or acyclic (E)
Pot. Pot.
0.027 Av. D. S. Mcgan and J. H. Wolfenden, J . Chem. SOC.,508 (1939), give an extrapolated value of 1.11 1. mole-' set.-'; M. Gordon, Thesis, Manchester, 1950, gives a lower value for butyrolactone at 0" (0.124 against 0.196). * Ethylene bromohydrin: C. L. McCabe and J. C. &'arner, THIS J O U R V A L , 70, 4031 (1948), give a n extrapolated value of 0.4&0.50 1. mole-' see.-'. c N-Methylsuccinimide: Miol.iti and Longo, ref. g, Table 111. These authors report thcir rate constants as Ak,, where A is the initial concentration of both imide and base. For A--methj lsuccinimide we find k2 = 0.753 1. mole-' set.-' by dividing by 0.005 X 60. The data of other investigators (S. S G. Sircar, J . Chem. SOC., 602, 1252 (1927); i V . Huckel and H . Muller, B e y . , 64B, 1981 (1931)) are t o be multiplied by 10 and divided by the 'ibove factor. See also Sircar, J . Chem. SOC.,898 (1928). 7-Valerolactone: Hegan and IVolfenden, ref. a , give 0.467 1. mole-' set.-'. Again, Gordon's value (ref. a ) is slightly lower at O " , 0.0719 zw. 0.0824. a
compounds, for the lone pair electrons cannot shicltl both carbonyl groups in the latter siniultaneously. The same factor may well account for the depressed reactivity of diiiiethyl ph3sphate anion relative to ethplenephosphate anion. The above postulate of rearward attack on ester carbonyl groups by hydroxide ion is supported by the results for 6-oxabicyclo[X ' 2 l]octan-'7-one (I) and for 2-oxabicyclo[2 : 2 : 2]octan-3-one (11). The conformational forniulas show that rcarward attack as described above is possible for I over the 3-
0
rate enhancement over their acyclic analogs (~f. Table I11 of ref. 16). Other cyclic compounds react at comparable rates to the open chain forms. The reason appears to be the following. Ballard and Bamfordg suggest that N-carboanI I1 hydrides are attacked by hydroxide ion via the transition state shown in A. inembered ring, and the rate constant approximates those for other 5-membered lactones. RearH R ward attack is impossible for 11, however, and the rate constant is drastically lowered to a value near 0 that of an aliphatic ester. I n the transition statc for hydrolysis of this hindered lactone, the hydroxyl ion attacks the p-orbital of the carbonyl group, with A Applying this concept to lactones B, we see that hydroxyl ion, when attacking the &-lactone, avoids unfavorable electrostatic repulsion from the lone ($1) U.
G. €1. Ballrrd .ind C. €I. B."ord,
J . C h e w h c . , 365 (Jr32SJ
(10) C . A. Coulson, "Valence," Clarendon Press, London, 1952, pp. 207-211. (11) T.Piash,J . A p b l . C h f , n . , 6 , 300 (lY5fl). (12) W. J. Orville-Thomas, C h e m . Re;'s.,S T , 1193 (lS27). (131 J. Kumamoto, J. R . C o x , Jr., and F. H. Westheimer, ' l ' n i b J < , L . R N A I , , 78,
1 S . X (1!)56).
HYDROLYSIS RATESAND MECHANISMS OF CYCLIC MONOMERS
Dec. 5, 1958
G423
TABLE I11 HYDROLYSIS RATESOF CYCLICCOMPOUNDS Rate factora
Reaction
+ +
Temp.,
OC.
Acyclic rate constant
5-Ring rate constant
&Ring rate constant
7-King rate constant
25.5 0.04 15 550 Lactone OH-* 10-2 0 0.84 8.7 1.92 1.8 Lactam OH-' 10-4 75 N-Methyllactam OH-' 10-4 75 1.92 0.700 4.78 ... 1.39 1.54 1.02 0 . 77 1 25 N-Acetyllactam OH-d Imides OH-"' 1 25 0.92 3.16 0.63 ... S-Methylimides OH-d,p 1 25 1.54 0.15 1.81 ... Anhydrides HzOh 1073 20 1.90 1.83 2.00 ... Anhydrides OH" 103 0 14.3 5.54 9.10 ... N-Acetylurethanes OH-d 1 25 ... 1.24 0.314 ... Multiply the rate constant given in table by Chis factor to get actual value in 1. mole-' set.-'. * R. Huisgen, et al., .4ngezel. Chem., 69,345 (1957); C. W. Matuszak and 1% Schechter, Abstracts of 132nd A.C.S. Meeting, Sew York, S.Y., Present work. e Corrected for ionization of the imide. September 1957, p. 12P. 0 M. Gordon, Thesis, Manchester, 1950. A Miolati, 7. T . Edward and K. A. Terry, J . Chem. Soc., 3527 (1957); J. T. Edward and S.Nielsen, ibid., 5080 (1957) Atli della reale Accad nazionale Lincei, [5] 3, 515 (1894); A Miolati and E. Longo, ibid., [5] 3, 597 (1896); J. Koskikdlio, Ann. Aced. Sei. Fennicae, Series A , KO.57 (1954).
+
+ +
+ + +
+
a consequent electrostatic repulsion from the atomic dipole of the 0 atom. I n the hydrolysis of these lactones by water,14 where electrostatic considerations are minimized because the attacking reagent is now uncharged, the lactones react a t comparable rates. However this reaction may have a quite different mechanism, and more work on this problem would be helpful. The N-acetyl lactams present a different picture. N-Acetyl-6-azabicyclo[3 :2 : 11-octan-7-one (111), undergoes ring cleavage ten times as rapidly as Nacetyl-2-azabicyclo[2 : 2: 21-octan-3-one (IV). COCH,
&' \
I
1
IV
Hccause the lone pair electrons on the N atom strongly interact with the carbonyl group, the necessity for rearward attack by hydroxyl is apparently not as stringent as for the lactones. Thus IV undergoes ring cleavage a t a rate comparable to other imides. It will be recalled that both of these lactanis undergo sodium-catalyzed polymerization with facility, wherein the key step is the attack of a lactam anion on the N-acyl lactam. Anionic reagents, therefore, can attack both these lactams and their N-acyl derivatives a t appreciable rates. The facile ring cleavage of 111 probably is caused by ininteraction by the bridged hibition of the N-C=O structure. I t is of interest that the bicyclic lactone V hydrolyzes quickly in water, whereas the N-acyl lactone V I is much more stable.15 A direct field effect of the positive pole on the carbonyl group may be responsible. O\,
C-
0
9\C-
0
LE?
RC-N
0
V
VI
( 1 4 ) R. R. G r e w , C. Heinke and C. Sommer, Ckem. Ber., 89, 1978 (1956). (15) B. Witkop trnd R. Poltz, THISJOuRN.41., 79, 1'34 (1927).
c.
Effect of Hetero Atoms in the Ring.-Replacement of methylene by oxygen adjacent to the carbonyl group has negligible effect on the rates. There is little interaction of the type shown in F.
