NOTES
3424
bath. During the addition, the temperature of the acid solution was not allowed to rise above -2". The frothy precipitate that rose to the surface was collected by filtration and sucked dry; it was then slurried with 100 ml. of ice-water, filtered, and washed with three 100-ml. portions of ice-cold water. The product was dried in air to constant weight; the yield of offwhite, fluffy powder was 59.4g. (30%), m.p. 75-77'. There was no change in the melting point on storage a t 0" for 3 years. For analysis, a sample was crystallized twice from aqueous methanol, m.p. 76.Ck76.5". The infrared spectrum of this material (1% solution in carbon tetrachloride) showed absorptions a t 5.74 ( C = O ) and 6.64 p (N-N=O). The spectrum of N-nitroso-N-methylurea ( 1% solution in carbon tetrachloride) showed corresponding absorptions a t 5.70 and 6.68p . Anal. Calcd. for C4HgN302: C, 36.63; H, 6.92; N, 32.05. Found: C, 36.65; H, 6.93; K , 31.81. B . From N-n-Propyl-N '-n-butyr y1urea.-N-n-Propyl-N '-n-butyrylurea was prepared in 61% yield from butyramide and 0.5 equiv. of bromine, m.p. 102.5103.5" (lit,.g m.p. 99-102'). A mixture of 72.O g. (0.42mole) of N-n-propyl-N'-butyrylurea and 265 ml. of concentrated hydrochloric acid was heated on a steam bath for 10 min. The solution was then cooled to 5", diluted with 265 ml. of water, and a solution of 36.9 g. (0.53mole) of sodium nitrite in 250 ml. of water was added during 15 min. The product was collected by filtration, washed with ice-cold water, and air-dried. The yield was 25.0g. (46y0). Recrystallization from aqueous methanol gave a sample that had m.p.
76.C-76.5'. 1-Diazopropane.-N-Xitroso-K-n-propylurea ( 1.0114 g.) was added during 2 min. to a mixture of 10 ml. of ether and 3 ml. of 40% potassium hydroxide a t 0". The mixture was allowed to stand a t 0" for 30 min. and the ether layer was decanted onto potassium hydroxide pellets. After drying 2 hr. a t O", the ether solution was filtered into an ethereal benzoic acid solution. Titration6 indicated the presence of 0.28 g. 152%) of l-diazopropane. The titration solution was diluted with water and extracted with ether. The ether solution was dried and evaporated. The n.m.r. spectrumlo of the residue in carbon tetrachloride solution was superimposable on that of n-propyl benzoate. The yield of 1-diazopropane was slightly less (43-47%) if the drying period or the time of addition was extended.
VOL. 29 TABLE I EFFECT OF REACTION CONDITIONS O N THE PREPARATION
-
OF @,p,p',Pf-TETRAMETHYL-2,4-8,10-TETRAOXASPIRO [ 5,5]
VNDECANE-3,g-DIETHANOL( h ) Concn. of aldehyde, Ma
Time, hr.
Temp., "C.
4.5 4.5 4.5 4.5 4.5 4.5 2.0 6.0
73 f 3 73 f 3 73 f 3 75 f 3 53 *3 90 73 f 3 25
0.7 1.3 2.7 3.6 2.7 2.7 2.7 1.3
% yield Of
M.p.,b
IS
"C.
19 203-205 199-202 34 70 201-202c 75 201-202 40 194-195 43 188-193 23 199-202 7 193-194 a Pentaerythritol was added in stoichiometric amounts. * Determined on a Fisher-Johns melting point apparatus; J. R. Caldwell, R. Gilkey, and B. S. Meeks, Jr., U. S. Patent 2,945,008 (1960),give m.p. 197". All melting points are uncorrected. anal. Calcd. for CloHzsOs: C, 59.29; H, 9.27. Found: C, 59.47; H, 9.25.
spiro-l,&dioxane. Recently, Cohen and Lavin4 described the condensation of water-soluble dialdehydes with pentaerythritol to polymeric 1,&dioxanes. Our work, which is summarized in eq. 1-3, features the use of water alone as the medium in which aliphatic aldehydes and l73-dio1s are converted to 1,3dioxanes. n-Butyraldehyde and 2,2-dimethyl-3-hyZRCHO
HOCH2,
+
,CHzOH
--+
C
(1)
HOCH~' 'CH~OH R-CH
/O--CHz \
\C/CHz-
O-CHz'
O\ /CH-R
'CH2-0
Ia, R=HOCHzC(CH3)2 b, R = CH3CHzCH2
(9) M . Montagne, Bull. soc. chim. France, 125 (1947). (10) A Varian A 4 0 instrument was used: tetramethylailane served BB internal reference.
RICH0
+
HO-CHa,
HO-CH/
Formation of 1,3-Dioxanes in Water
,,Rz
C
/O-CHz\ +
/Rz
RiCH
'R$
(2) 'O-CHz
/c\R3
I1 F. R. GALIANO, D. RANKIN, A N D G. J. MANTELL Spencer Research Center, Spencer Chemical Company, Merriam, Kansas
IIa, R1
= HOCHzC(CHa)z; Rz = HOCHz; Ra = CHI b, Ri = HOCHzC(CH3)z; Rz R3 CHa CHa C, R1 = H ; Rz = HOCH,; Ra d, Ri = HOCHzC(CH8)z; Rz = HOCHz; R3 = CHoCHz
Received September SO, 1963
Although Skrabal' showed by kinetic studies that 1,3dioxanes are highly resistant to acid hydrolysis, few workers have investigated their preparation in water, a practical, but unappreciated solvent for the synthesis of derivatives of 1,3-dioxane. Conrad and co-workers2 prepared 1,&dioxanes by condensing water-soluble aldehydes with 2,2-dimethylpropanediol-l,3 and 2-hydroxymethylpropanediol-l,3; water-insoluble aldehydes underwent the same reaction in 1:1 dioxanewater. Read3 formed spiro-l,3-dioxanes by condensing pentaerythritol in sulfuric acid (30-5070) at room temperature with aliphatic and aromatic, unhydroxylated aldehydes. Read also used crotonaldehyde to give a (1) A . Skrabal and M. Zlateaa, 2. P h y s i k . Chem. (Leipzig). 119, 305 (1926). (2) W . E. Conrad, B. D. Gesner, L. .4. Levasseur, R . F. Murphy, and H. M . Conrad. J . O r g . C h e m . , 16, 3571 (1961). (3) J. Read, J . C h e m . S o c . , 101, 2090 (1912).
CH3'
'CHzOH
HOCH2\ ,CHz.-O
C
CHI'
'cH--(cH~),-cH /
0 -CHz, / 'O-CHz'
,CH2OH
C'CHS
\CHa-O IIIa, n = 2
b, n = 3
droxypropionaldehyde, both only sparingly soluble in water, give high yields of products; dioxane is not required. Furthermore, the water-soluble products IIIa and I I I b are obtained in more than 90% yield; therefore, product insolubility is not a driving force for the reaction. The drastic conditions employed by Read are not necessary for the preparation of acetals from aliphatic saturated aldehydes but may be needed for reaction of a,&unsaturated aldehydes. Crotonaldehyde would not condense with pentaerythritol under (1) R. M. Cohen and E. Lavin, J . A p p l . Polymer S c i . . 6 , 503 (1962).
KOTES
NOVEMBER, 1964 CONDITIONS FOR
THE
TABLE I1 PREPARATION OF VARIOCS CYCLICACETALS I N WATER Concn. of alcohol,
Alcohol
Aldehyde
2,2-Dimethyl-3-hydroxypropionaldehyde n-But yraldehyde Crotonaldehyde 2,2-Dimethyl-3-hydroxypropionaldehyde Formaldehydeb
3425
2-Hydroxymethyl-2-methylpropanediol-1,3 Pentaerythritol Pentaerythritol 2,2-Dimethylpropanediol-1,3
Time, hr.
Temp.,
"C.
Product
2.1
4.5
75
IIa
50
1.4 1.4 5 7
4.0 2.5 16.0
60 45-65 75
Ib
48 0 88
Ma
70
...
IIb
yield
IIC" 2-Hy droxy methy l-2-methyl19.0 18.0 85 55 propanediol-1,3 IId 81 2-Hy droxymethy l-2-ethyl1.5 4.5 75 2,2-Dimethyl-3-hydroxypropanediol-1,3 propionaldehyde IIIad 94 2-H ydroxymethyl-2-methyl5.0 16.0 75 Diethoxytetrahydrofuran propanediol-1,3 IIIbd 95 2-H y droxymethyl-2-methyl6.7 16.0 75 Glutaraldehyde propanediol-l,3 . . Water-soluble liquid, isolated by fractional dis1, As a 377, aqueous solution. a Aldehyde was added in stoichiometric amounts. Water-solublesolid, extracted tillation after neutralization; L. Coes, Jr., U.S.Patent 2,526,601 (1950), gives b.p. 72-84' (0.8 mm.). from the reaction mixture with chloroform.
TABLE I11 PROPERTIES OF CYCLIC ACETALB Product
Formula
M.p., oC.a
-----Calcd.----
C
n
-Found--
C
H
IIa CioHzoO4 128-130b 58.82 9.83 58.67 10.11 Ib C13H2404 58-59" ... ... ... ... IIb C~oHz003 63-6jd ... ... ... ... IIC C6HlZO3 ... 54.52 9.15 53.98 8.66 IId CIIHZZO~ 120-121e 60.55 10.09 60.45 10.27 IIIa CuH2806 94.5-96' 57.93 8.97 58.03 8.62 IIIb Cl6H2806 109-110' 59.21 9.21 59.50 8.95 Recrystallized from 1 : 1 isopropyl alcoholTaken on a Fisher-Johns melting point apparatus. * Recrystallized from benzene. E. Spath and I. von Szilbgyi, Ber., 76B, 949 water; V. G. Mkhitaryan, J. Gen. Chem. USSR, 9, 1923 (1939), gives m.p. 59-60". (1943), gives m.p. 65-66". e Caldwell, et al., Table I, footnote b, gives m.p. 123'. f Recrystallized from heptane.
conditions wherein n-butyraldehyde readily condensed. This observation is in accord with data obtained by Kreevoy and Taft5 who found that acetals of a,Punsaturated aldehydes are much less resistant to acid hydrolysis than those of the corresponding saturated aldehydes. The data in Table I show the effect of time, temperature, and concentration on the preparation of acetal Ia. At constant time and concentration, the yield at 73" (70%) was higher than at either 53 (40%) or 90" (43oj,), and it dropped to 23% when the tinie was reduced from 4.5 to 2.0 hr. Although the yield rose rapidly with an increase in concentration, an unstirrable slurry formed a t a concentration of 3.6 moles/l. of aldehyde. Obviously, the formation of a slurry limits the concentration that can be used practically. Conducting the condensation in water simplifies the preparation of Ia, because the starting aldehyde and alcohol, which are prepared in water from acetaldehyde, isobutyraldehyde, and formaldehyde, need not be isolated. For example, an aldol condensation of formaldehyde and isobutyraldehyde gives 2,2-dimethyl-3hydroxypropionaldehyde. Another aldol condensation of formaldehyde and acetaldehyde, followed by a Cannizzaro reaction that uses another mole of formaldehyde, forms pentaerythritol. After removal of some of the water, a mixture of the two solutions containing 1.7 moles/l. of aldehyde and 0.85 moles/l. of alcohol is
heated at 70-75" for 4.5 hr. The yield of I a is 387, which, according to Table I, is close to that expected from a condensation started with these concentrations of pure alcohol and aldehyde. Table 11, which summarizes our work with six aldehydes and four alcohols, shows the generality of the condensation of saturated, aliphatic mono- and dialdehydes with 1,3-diols. The melting points and chemical analyses of the 1,3-dioxanes prepared in this work appear in Table 111.
(5) M. M. Kreevoy and R . W . Taft, Jr.. J . A m . Chem. Soc.. 11, 5590 (1955).
(6) E. Stiller, S. Harris, J. Finkelstein, J. Keresatesy, and K. Folkers, i b i d . , 69, 1785 (1940).
Experimental The general method.for the preparation of cyclic acetals in water is illustrated by the following examples. p,p,5,5-Tetramethyl-2-m-dioxaneethanol (IIb).-2,2-Dimethylpropanediol-1,3 (104 g., 1.O mole) and 2,2-dimethyl-3-hydroxypropionaldehyde (102 g., 1.0 mole) were added to 175 ml. of water in a round-bottom flask equipped with stirrer, reflux condenser, and thermometer. The pH of the solution was adjusted to 3 with p-toluenesulfonic acid hydrate and heated for 16 hr. at 75". The oil layer that formed during this time solidified on cooling to room temperature. The solid was removed by filtration, washed with 100 ml. of cool water, and dried. The product (165.5 g., 88% yield) melted a t 63-65'. Preparation of p,p,p',~'-Tetramethy1-2,4-8,lO-tetraoxaspiro[SSjundecane-3,9-diethanol (Ia) from Acetaldehyde, Isobutyraldehyde, and Formaldehyde.-2,2-Dimethyl-3-hydroxypropionaldehyde (0.15 mole) was prepared as described by Stiller, et al.6 Potassium carbonate (8.28 g., 0.06 mole) was added to a mixture of formalin (12.5 g. of 3774, 0.15 mole) and isobutyraldehyde (10.2 g., 0.15 mole) a t such a rate as to maintain a temperature of
NOTES
3426
VOL. 29
30 =t3". The reaction mixture was stirred for 2 hr. and neutralized with a solution of 5.5 g. of sulfuric acid in 40 ml. of water. In another flask, pentaerythritol(O.075 mole) was prepared by adding a mixture of acetaldehyde (3.30 g., 0.075 mole) and formalin (30.3 g. of 37%, 1% methanol, 0.375 mole) to a slurry of calcium hydroxide (3.90 g., 0.052 mole) in 45 ml. of water. The mixture was heated a t 50" for 1 hr. and allowed to cool slowly to room temperature. Additional calcium hydroxide (4.40 g., 0.06 mole) and 4.2 ml. of 307, hydrogen peroxide were added, and the mixture was stirred for 1 hr. a t room temperature to remove unreacted formaldehyde. After neutralization with solid oxalic acid, the mixture was filtered and evaporated to about 40 ml. and added to the neutralized aldehyde solution. After the pH was adjusted to 3 with p-toluenesulfonic acid, the mixture was heated a t 7C-75" for 4.5 hr. The solid material was removed by filtration, washed, and dried. The yield of product was 8.72 g. (37.8%), m.p. 185-194'.
residual bonded hydroxyl absorption was observed3 at 0.023 M in carbon disulfide solution prompted us to re-examine this system a t greater dilution. Infrared spectra of tropine and pseudotropine were recorded in carbon disulfide, carbon tetrachloride, and tetrachloroethylene solutions at 2 X M , using a high-resolution grating spectrophotometer. Under these conditions, it was observed that (1) the bonded 0-H stretching absorption is completely eliminated in both alcohols, and (2) the free 0-H stretching band of tropine occurs a t a slightly higher frequency (5 e m - ' in carbon tetrachloride) than that of pseudotropine (see Table I ) . Furthermore, when examined on an expanded
A Re-examination of the Conformational Analysis
FREEOH STRETCHING MAXIMA, pK,, AND G.L.c. DATAFOR TROPINE AND PSEUDOTROPINE
TABLE I
of the Tropine-Pseudotropine System' HERBERT S.AARONA N D CHARLES P. RADER
Compd.
Received June 1, 1964
Although the configurations of pseudotropine (I) and tropine (11) have been unequivocally proven, the conformational equilibrium for the ground state of the piperidinol ring in these two compounds has not been rigorously established. The first interpretations according to the then emerging principles of conformational analysis led to divided opinions as to whether the chair (11) or the boat (111) form predominates in this ring system.2
Me N . MeN . -
I\
HI
H
I
cm. 1CClr
CnCl,
pKa"
Tropine 3613 3626 3627 10.44 Pseudotropine 3609 3621 3623 9.98 a Ionic strength 0.005 a t 30". * Column (10 ft. X of Carbowax 20 M (15%) on Gas-Chrom P (60-80) and 120 ml./min. (He). The value for 3-tropinone Chemical Co.) was 4.9 min. under these conditions.
Chemical Research Division, Chemical Research and Development Lab oratories, Edgewood Arsenal, Maryland
c--
-Cmmax CS?
I1
Evidence for the boat conformation rests mainly on infrared spectral data which have been interpreteda (and continue to be cited4)to indicate that intramolecular hydrogen bonding exists between the nitrogen and the hydroxyl group of the pseudotropine system. The data, however, do not meet established criteria5 for intramolecular hydrogen bonding, because the intensity of the bonded hydroxyl stretching band is concentration dependent, a characteristic typical of intermolecularly hydrogen-bonded systems. However, the fact that (1) Presented in part at the 145th National Meeting of the American Chemical Society, New York. 9 . Y . , Sept.. 1963, Abstracts of Papers, p. 42Q. (2) G. Fodor, "The Alkaloids," Vol. V I , R. H. F. Manske, Ed., Academic Press. New York, N. Y . , 1960, Chapter 5. (3) B. L. Zenitz, C. M. Martini, M. Prianar, and F. C. Nachod, J . Am. C h e m . Soc., 74,5564 (1952). (4) G. L. Closs, ihid.. 81, 5456 (1959): G. Hite, E. E. Smissman, and R. West, ibid., 82, 1207 (1960); W, .4. M. Davies, J. B. Jones, and A. R. Pinder, J . C h e m . S O C . ,3504 (1960); & Balasubramanian 'I. and N. Padma. Tetrahedron, 19,2135 (1963). ( 5 ) A. R. H. Cole, "Technique of Organic Chemistry," Vol. XI (part l ) , A. Weissberger, Ed., Interscience Publishers, Inc., New York, N. Y . , 1963, Chapter 3.
G.1.c. retention time, min.*
6.5 8.0 0.25 in.) a t 205' (Aldrich
abscissa (5 cm.-'/cm.), the tropine absorption appears as a single, highly symmetrical band, while that of pseudotropine appears as an unsymmetrical band, apparently an unresolved doublet. Comparable results were obtained for the first overtone of the hydroxyl stretching absorptions. The absence of any detectable intramolecular hydrogen bonding6 indicates that the piperidinol ring of the pseudotropine system must exist in a chair conformation, and the per cent of molecules in a boat conformation must be smaller (undoubtedly less than 2%; see Experimental) than can be detected by infrared methods. Tropine, of course, cannot bond intramolecularly in either a chair or a boat conformation. Here, however, the position and shape of the free 0-H stretching band may be used as criteria for conformational assignment. Thus, it has been shown that in dilute carbon tetrachloride solution, the axial hydroxyl group will have a symmetrical stretching band' which occurs~~* about 5-10 cm.-l higher then that of its equatorial epimer, which has an unsymmetrical band. These distinctions are apparently due to the relative populations of isomers which correspond to rotational conformations of the hydroxyl group about the C-0 bond. In this respect, the band of tropine in a boat conformation (111) should resemble that of an equatorial alcohol. The lack of any significant population of tropine molecules in a boat conformation, therefore, is suggested by the highly symmetrical shape of its hydroxyl stretching band. (6) (a) The erroneous report3 of intramolecular hydrogen bonding in ~ a paper which appeared pseudotropine was corrected by House, et G Z . , ~ in simultaneously with our earlier presentation1 of these results. (b) H. 0. House, H. C. Muller, C. G. Pitt, and P. P. Wickham, J . Org. Chem., 28, 2407 (1963). (7) H. S. Aaron and C. P. Rader, J . Am. Chem. Soc., 86, 3046 (1963). (8) I. L. Allsop. A. R . H. Cole, D. E. White, and R. L. S. Willix, J . C h e m . SOC.,4868 (1956).
